1 files changed, 1 insertions, 0 deletions
diff --git a/drivers/video/fbdev/sstfb.c b/drivers/video/fbdev/sstfb.c
index f0cb279ef333..4e22ae383c87 100644
--- a/drivers/video/fbdev/sstfb.c
+++ b/drivers/video/fbdev/sstfb.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0-only
 /*
  * linux/drivers/video/sstfb.c -- voodoo graphics frame buffer
  *
diff --git a/drivers/Kconfig b/drivers/Kconfig
index 622fa266b29e..c70d6e45dc10 100644
--- a/drivers/Kconfig
+++ b/drivers/Kconfig
@@ -1,5 +1,7 @@
 menu "Device Drivers"
 
+source "drivers/amba/Kconfig"
+
 source "drivers/base/Kconfig"
 
 source "drivers/bus/Kconfig"
@@ -132,8 +134,6 @@ source "drivers/staging/Kconfig"
 
 source "drivers/platform/Kconfig"
 
-source "drivers/soc/Kconfig"
-
 source "drivers/clk/Kconfig"
 
 source "drivers/hwspinlock/Kconfig"
@@ -148,6 +148,8 @@ source "drivers/remoteproc/Kconfig"
 
 source "drivers/rpmsg/Kconfig"
 
+source "drivers/soc/Kconfig"
+
 source "drivers/devfreq/Kconfig"
 
 source "drivers/extcon/Kconfig"
@@ -180,4 +182,6 @@ source "drivers/ras/Kconfig"
 
 source "drivers/thunderbolt/Kconfig"
 
+source "drivers/android/Kconfig"
+
 endmenu
diff --git a/drivers/Makefile b/drivers/Makefile
index ebee55537a05..527a6da8d539 100644
--- a/drivers/Makefile
+++ b/drivers/Makefile
@@ -50,7 +50,10 @@ obj-$(CONFIG_RESET_CONTROLLER)	+= reset/
 obj-y				+= tty/
 obj-y				+= char/
 
-# gpu/ comes after char for AGP vs DRM startup
+# iommu/ comes before gpu as gpu are using iommu controllers
+obj-$(CONFIG_IOMMU_SUPPORT)	+= iommu/
+
+# gpu/ comes after char for AGP vs DRM startup and after iommu
 obj-y				+= gpu/
 
 obj-$(CONFIG_CONNECTOR)		+= connector/
@@ -141,7 +144,6 @@ obj-y				+= clk/
 
 obj-$(CONFIG_MAILBOX)		+= mailbox/
 obj-$(CONFIG_HWSPINLOCK)	+= hwspinlock/
-obj-$(CONFIG_IOMMU_SUPPORT)	+= iommu/
 obj-$(CONFIG_REMOTEPROC)	+= remoteproc/
 obj-$(CONFIG_RPMSG)		+= rpmsg/
 
@@ -161,3 +163,5 @@ obj-$(CONFIG_POWERCAP)		+= powercap/
 obj-$(CONFIG_MCB)		+= mcb/
 obj-$(CONFIG_RAS)		+= ras/
 obj-$(CONFIG_THUNDERBOLT)	+= thunderbolt/
+obj-$(CONFIG_CORESIGHT)		+= coresight/
+obj-$(CONFIG_ANDROID)		+= android/
diff --git a/drivers/acpi/Kconfig b/drivers/acpi/Kconfig
index d0f3265fb85d..8951cefb0a96 100644
--- a/drivers/acpi/Kconfig
+++ b/drivers/acpi/Kconfig
@@ -144,7 +144,7 @@ config ACPI_VIDEO
 
 config ACPI_FAN
 	tristate "Fan"
-	select THERMAL
+	depends on THERMAL
 	default y
 	help
 	  This driver supports ACPI fan devices, allowing user-mode
@@ -360,15 +360,14 @@ config ACPI_BGRT
 config ACPI_REDUCED_HARDWARE_ONLY
 	bool "Hardware-reduced ACPI support only" if EXPERT
 	def_bool n
-	depends on ACPI
 	help
-	This config item changes the way the ACPI code is built.  When this
-	option is selected, the kernel will use a specialized version of
-	ACPICA that ONLY supports the ACPI "reduced hardware" mode.  The
-	resulting kernel will be smaller but it will also be restricted to
-	running in ACPI reduced hardware mode ONLY.
+	  This config item changes the way the ACPI code is built.  When this
+	  option is selected, the kernel will use a specialized version of
+	  ACPICA that ONLY supports the ACPI "reduced hardware" mode.  The
+	  resulting kernel will be smaller but it will also be restricted to
+	  running in ACPI reduced hardware mode ONLY.
 
-	If you are unsure what to do, do not enable this option.
+	  If you are unsure what to do, do not enable this option.
 
 source "drivers/acpi/apei/Kconfig"
 
@@ -394,4 +393,27 @@ config ACPI_EXTLOG
 	  driver adds support for that functionality with corresponding
 	  tracepoint which carries that information to userspace.
 
+menuconfig PMIC_OPREGION
+	bool "PMIC (Power Management Integrated Circuit) operation region support"
+	help
+	  Select this option to enable support for ACPI operation
+	  region of the PMIC chip. The operation region can be used
+	  to control power rails and sensor reading/writing on the
+	  PMIC chip.
+
+if PMIC_OPREGION
+config CRC_PMIC_OPREGION
+	bool "ACPI operation region support for CrystalCove PMIC"
+	depends on INTEL_SOC_PMIC
+	help
+	  This config adds ACPI operation region support for CrystalCove PMIC.
+
+config XPOWER_PMIC_OPREGION
+	bool "ACPI operation region support for XPower AXP288 PMIC"
+	depends on AXP288_ADC = y
+	help
+	  This config adds ACPI operation region support for XPower AXP288 PMIC.
+
+endif
+
 endif	# ACPI
diff --git a/drivers/acpi/Makefile b/drivers/acpi/Makefile
index 505d4d79fe3e..f74317cc1ca9 100644
--- a/drivers/acpi/Makefile
+++ b/drivers/acpi/Makefile
@@ -43,9 +43,11 @@ acpi-y				+= pci_root.o pci_link.o pci_irq.o
 acpi-y				+= acpi_lpss.o
 acpi-y				+= acpi_platform.o
 acpi-y				+= acpi_pnp.o
+acpi-y				+= int340x_thermal.o
 acpi-y				+= power.o
 acpi-y				+= event.o
 acpi-y				+= sysfs.o
+acpi-y				+= property.o
 acpi-$(CONFIG_X86)		+= acpi_cmos_rtc.o
 acpi-$(CONFIG_DEBUG_FS)		+= debugfs.o
 acpi-$(CONFIG_ACPI_NUMA)	+= numa.o
@@ -86,3 +88,7 @@ obj-$(CONFIG_ACPI_PROCESSOR_AGGREGATOR) += acpi_pad.o
 obj-$(CONFIG_ACPI_APEI)		+= apei/
 
 obj-$(CONFIG_ACPI_EXTLOG)	+= acpi_extlog.o
+
+obj-$(CONFIG_PMIC_OPREGION)	+= pmic/intel_pmic.o
+obj-$(CONFIG_CRC_PMIC_OPREGION) += pmic/intel_pmic_crc.o
+obj-$(CONFIG_XPOWER_PMIC_OPREGION) += pmic/intel_pmic_xpower.o
diff --git a/drivers/acpi/acpi_lpss.c b/drivers/acpi/acpi_lpss.c
index b0ea767c8696..4f3febf8a589 100644
--- a/drivers/acpi/acpi_lpss.c
+++ b/drivers/acpi/acpi_lpss.c
@@ -1,7 +1,7 @@
 /*
  * ACPI support for Intel Lynxpoint LPSS.
  *
- * Copyright (C) 2013, Intel Corporation
+ * Copyright (C) 2013, 2014, Intel Corporation
  * Authors: Mika Westerberg <mika.westerberg@linux.intel.com>
  *          Rafael J. Wysocki <rafael.j.wysocki@intel.com>
  *
@@ -54,55 +54,62 @@ ACPI_MODULE_NAME("acpi_lpss");
 
 #define LPSS_PRV_REG_COUNT		9
 
-struct lpss_shared_clock {
-	const char *name;
-	unsigned long rate;
-	struct clk *clk;
-};
+/* LPSS Flags */
+#define LPSS_CLK			BIT(0)
+#define LPSS_CLK_GATE			BIT(1)
+#define LPSS_CLK_DIVIDER		BIT(2)
+#define LPSS_LTR			BIT(3)
+#define LPSS_SAVE_CTX			BIT(4)
+#define LPSS_DEV_PROXY			BIT(5)
+#define LPSS_PROXY_REQ			BIT(6)
 
 struct lpss_private_data;
 
 struct lpss_device_desc {
-	bool clk_required;
-	const char *clkdev_name;
-	bool ltr_required;
+	unsigned int flags;
 	unsigned int prv_offset;
 	size_t prv_size_override;
-	bool clk_divider;
-	bool clk_gate;
-	bool save_ctx;
-	struct lpss_shared_clock *shared_clock;
 	void (*setup)(struct lpss_private_data *pdata);
 };
 
+static struct device *proxy_device;
+
 static struct lpss_device_desc lpss_dma_desc = {
-	.clk_required = true,
-	.clkdev_name = "hclk",
+	.flags = LPSS_CLK | LPSS_PROXY_REQ,
 };
 
 struct lpss_private_data {
 	void __iomem *mmio_base;
 	resource_size_t mmio_size;
+	unsigned int fixed_clk_rate;
 	struct clk *clk;
 	const struct lpss_device_desc *dev_desc;
 	u32 prv_reg_ctx[LPSS_PRV_REG_COUNT];
 };
 
+/* UART Component Parameter Register */
+#define LPSS_UART_CPR			0xF4
+#define LPSS_UART_CPR_AFCE		BIT(4)
+
 static void lpss_uart_setup(struct lpss_private_data *pdata)
 {
 	unsigned int offset;
-	u32 reg;
+	u32 val;
 
 	offset = pdata->dev_desc->prv_offset + LPSS_TX_INT;
-	reg = readl(pdata->mmio_base + offset);
-	writel(reg | LPSS_TX_INT_MASK, pdata->mmio_base + offset);
-
-	offset = pdata->dev_desc->prv_offset + LPSS_GENERAL;
-	reg = readl(pdata->mmio_base + offset);
-	writel(reg | LPSS_GENERAL_UART_RTS_OVRD, pdata->mmio_base + offset);
+	val = readl(pdata->mmio_base + offset);
+	writel(val | LPSS_TX_INT_MASK, pdata->mmio_base + offset);
+
+	val = readl(pdata->mmio_base + LPSS_UART_CPR);
+	if (!(val & LPSS_UART_CPR_AFCE)) {
+		offset = pdata->dev_desc->prv_offset + LPSS_GENERAL;
+		val = readl(pdata->mmio_base + offset);
+		val |= LPSS_GENERAL_UART_RTS_OVRD;
+		writel(val, pdata->mmio_base + offset);
+	}
 }
 
-static void lpss_i2c_setup(struct lpss_private_data *pdata)
+static void byt_i2c_setup(struct lpss_private_data *pdata)
 {
 	unsigned int offset;
 	u32 val;
@@ -111,100 +118,58 @@ static void lpss_i2c_setup(struct lpss_private_data *pdata)
 	val = readl(pdata->mmio_base + offset);
 	val |= LPSS_RESETS_RESET_APB | LPSS_RESETS_RESET_FUNC;
 	writel(val, pdata->mmio_base + offset);
-}
 
-static struct lpss_device_desc wpt_dev_desc = {
-	.clk_required = true,
-	.prv_offset = 0x800,
-	.ltr_required = true,
-	.clk_divider = true,
-	.clk_gate = true,
-};
+	if (readl(pdata->mmio_base + pdata->dev_desc->prv_offset))
+		pdata->fixed_clk_rate = 133000000;
+}
 
 static struct lpss_device_desc lpt_dev_desc = {
-	.clk_required = true,
+	.flags = LPSS_CLK | LPSS_CLK_GATE | LPSS_CLK_DIVIDER | LPSS_LTR,
 	.prv_offset = 0x800,
-	.ltr_required = true,
-	.clk_divider = true,
-	.clk_gate = true,
 };
 
 static struct lpss_device_desc lpt_i2c_dev_desc = {
-	.clk_required = true,
+	.flags = LPSS_CLK | LPSS_CLK_GATE | LPSS_LTR,
 	.prv_offset = 0x800,
-	.ltr_required = true,
-	.clk_gate = true,
 };
 
 static struct lpss_device_desc lpt_uart_dev_desc = {
-	.clk_required = true,
+	.flags = LPSS_CLK | LPSS_CLK_GATE | LPSS_CLK_DIVIDER | LPSS_LTR,
 	.prv_offset = 0x800,
-	.ltr_required = true,
-	.clk_divider = true,
-	.clk_gate = true,
 	.setup = lpss_uart_setup,
 };
 
 static struct lpss_device_desc lpt_sdio_dev_desc = {
+	.flags = LPSS_LTR,
 	.prv_offset = 0x1000,
 	.prv_size_override = 0x1018,
-	.ltr_required = true,
-};
-
-static struct lpss_shared_clock pwm_clock = {
-	.name = "pwm_clk",
-	.rate = 25000000,
 };
 
 static struct lpss_device_desc byt_pwm_dev_desc = {
-	.clk_required = true,
-	.save_ctx = true,
-	.shared_clock = &pwm_clock,
+	.flags = LPSS_SAVE_CTX,
 };
 
 static struct lpss_device_desc byt_uart_dev_desc = {
-	.clk_required = true,
+	.flags = LPSS_CLK | LPSS_CLK_GATE | LPSS_CLK_DIVIDER | LPSS_SAVE_CTX |
+		 LPSS_DEV_PROXY,
 	.prv_offset = 0x800,
-	.clk_divider = true,
-	.clk_gate = true,
-	.save_ctx = true,
 	.setup = lpss_uart_setup,
 };
 
 static struct lpss_device_desc byt_spi_dev_desc = {
-	.clk_required = true,
+	.flags = LPSS_CLK | LPSS_CLK_GATE | LPSS_CLK_DIVIDER | LPSS_SAVE_CTX |
+		 LPSS_DEV_PROXY,
 	.prv_offset = 0x400,
-	.clk_divider = true,
-	.clk_gate = true,
-	.save_ctx = true,
 };
 
 static struct lpss_device_desc byt_sdio_dev_desc = {
-	.clk_required = true,
-};
-
-static struct lpss_shared_clock i2c_clock = {
-	.name = "i2c_clk",
-	.rate = 100000000,
+	.flags = LPSS_CLK | LPSS_DEV_PROXY,
 };
 
 static struct lpss_device_desc byt_i2c_dev_desc = {
-	.clk_required = true,
+	.flags = LPSS_CLK | LPSS_SAVE_CTX | LPSS_DEV_PROXY,
 	.prv_offset = 0x800,
-	.save_ctx = true,
-	.shared_clock = &i2c_clock,
-	.setup = lpss_i2c_setup,
-};
-
-static struct lpss_shared_clock bsw_pwm_clock = {
-	.name = "pwm_clk",
-	.rate = 19200000,
-};
-
-static struct lpss_device_desc bsw_pwm_dev_desc = {
-	.clk_required = true,
-	.save_ctx = true,
-	.shared_clock = &bsw_pwm_clock,
+	.setup = byt_i2c_setup,
 };
 
 #else
@@ -237,7 +202,7 @@ static const struct acpi_device_id acpi_lpss_device_ids[] = {
 	{ "INT33FC", },
 
 	/* Braswell LPSS devices */
-	{ "80862288", LPSS_ADDR(bsw_pwm_dev_desc) },
+	{ "80862288", LPSS_ADDR(byt_pwm_dev_desc) },
 	{ "8086228A", LPSS_ADDR(byt_uart_dev_desc) },
 	{ "8086228E", LPSS_ADDR(byt_spi_dev_desc) },
 	{ "808622C1", LPSS_ADDR(byt_i2c_dev_desc) },
@@ -251,7 +216,8 @@ static const struct acpi_device_id acpi_lpss_device_ids[] = {
 	{ "INT3436", LPSS_ADDR(lpt_sdio_dev_desc) },
 	{ "INT3437", },
 
-	{ "INT3438", LPSS_ADDR(wpt_dev_desc) },
+	/* Wildcat Point LPSS devices */
+	{ "INT3438", LPSS_ADDR(lpt_dev_desc) },
 
 	{ }
 };
@@ -276,7 +242,6 @@ static int register_device_clock(struct acpi_device *adev,
 				 struct lpss_private_data *pdata)
 {
 	const struct lpss_device_desc *dev_desc = pdata->dev_desc;
-	struct lpss_shared_clock *shared_clock = dev_desc->shared_clock;
 	const char *devname = dev_name(&adev->dev);
 	struct clk *clk = ERR_PTR(-ENODEV);
 	struct lpss_clk_data *clk_data;
@@ -289,12 +254,7 @@ static int register_device_clock(struct acpi_device *adev,
 	clk_data = platform_get_drvdata(lpss_clk_dev);
 	if (!clk_data)
 		return -ENODEV;
-
-	if (dev_desc->clkdev_name) {
-		clk_register_clkdev(clk_data->clk, dev_desc->clkdev_name,
-				    devname);
-		return 0;
-	}
+	clk = clk_data->clk;
 
 	if (!pdata->mmio_base
 	    || pdata->mmio_size < dev_desc->prv_offset + LPSS_CLK_SIZE)
@@ -303,24 +263,19 @@ static int register_device_clock(struct acpi_device *adev,
 	parent = clk_data->name;
 	prv_base = pdata->mmio_base + dev_desc->prv_offset;
 
-	if (shared_clock) {
-		clk = shared_clock->clk;
-		if (!clk) {
-			clk = clk_register_fixed_rate(NULL, shared_clock->name,
-						      "lpss_clk", 0,
-						      shared_clock->rate);
-			shared_clock->clk = clk;
-		}
-		parent = shared_clock->name;
+	if (pdata->fixed_clk_rate) {
+		clk = clk_register_fixed_rate(NULL, devname, parent, 0,
+					      pdata->fixed_clk_rate);
+		goto out;
 	}
 
-	if (dev_desc->clk_gate) {
+	if (dev_desc->flags & LPSS_CLK_GATE) {
 		clk = clk_register_gate(NULL, devname, parent, 0,
 					prv_base, 0, 0, NULL);
 		parent = devname;
 	}
 
-	if (dev_desc->clk_divider) {
+	if (dev_desc->flags & LPSS_CLK_DIVIDER) {
 		/* Prevent division by zero */
 		if (!readl(prv_base))
 			writel(LPSS_CLK_DIVIDER_DEF_MASK, prv_base);
@@ -344,7 +299,7 @@ static int register_device_clock(struct acpi_device *adev,
 		kfree(parent);
 		kfree(clk_name);
 	}
-
+out:
 	if (IS_ERR(clk))
 		return PTR_ERR(clk);
 
@@ -392,7 +347,10 @@ static int acpi_lpss_create_device(struct acpi_device *adev,
 
 	pdata->dev_desc = dev_desc;
 
-	if (dev_desc->clk_required) {
+	if (dev_desc->setup)
+		dev_desc->setup(pdata);
+
+	if (dev_desc->flags & LPSS_CLK) {
 		ret = register_device_clock(adev, pdata);
 		if (ret) {
 			/* Skip the device, but continue the namespace scan. */
@@ -413,12 +371,11 @@ static int acpi_lpss_create_device(struct acpi_device *adev,
 		goto err_out;
 	}
 
-	if (dev_desc->setup)
-		dev_desc->setup(pdata);
-
 	adev->driver_data = pdata;
 	pdev = acpi_create_platform_device(adev);
 	if (!IS_ERR_OR_NULL(pdev)) {
+		if (!proxy_device && dev_desc->flags & LPSS_DEV_PROXY)
+			proxy_device = &pdev->dev;
 		return 1;
 	}
 
@@ -550,14 +507,15 @@ static void acpi_lpss_set_ltr(struct device *dev, s32 val)
 /**
  * acpi_lpss_save_ctx() - Save the private registers of LPSS device
  * @dev: LPSS device
+ * @pdata: pointer to the private data of the LPSS device
  *
  * Most LPSS devices have private registers which may loose their context when
  * the device is powered down. acpi_lpss_save_ctx() saves those registers into
  * prv_reg_ctx array.
  */
-static void acpi_lpss_save_ctx(struct device *dev)
+static void acpi_lpss_save_ctx(struct device *dev,
+			       struct lpss_private_data *pdata)
 {
-	struct lpss_private_data *pdata = acpi_driver_data(ACPI_COMPANION(dev));
 	unsigned int i;
 
 	for (i = 0; i < LPSS_PRV_REG_COUNT; i++) {
@@ -572,12 +530,13 @@ static void acpi_lpss_save_ctx(struct device *dev)
 /**
  * acpi_lpss_restore_ctx() - Restore the private registers of LPSS device
  * @dev: LPSS device
+ * @pdata: pointer to the private data of the LPSS device
  *
  * Restores the registers that were previously stored with acpi_lpss_save_ctx().
  */
-static void acpi_lpss_restore_ctx(struct device *dev)
+static void acpi_lpss_restore_ctx(struct device *dev,
+				  struct lpss_private_data *pdata)
 {
-	struct lpss_private_data *pdata = acpi_driver_data(ACPI_COMPANION(dev));
 	unsigned int i;
 
 	/*
@@ -600,54 +559,82 @@ static void acpi_lpss_restore_ctx(struct device *dev)
 #ifdef CONFIG_PM_SLEEP
 static int acpi_lpss_suspend_late(struct device *dev)
 {
-	int ret = pm_generic_suspend_late(dev);
+	struct lpss_private_data *pdata = acpi_driver_data(ACPI_COMPANION(dev));
+	int ret;
 
+	ret = pm_generic_suspend_late(dev);
 	if (ret)
 		return ret;
 
-	acpi_lpss_save_ctx(dev);
+	if (pdata->dev_desc->flags & LPSS_SAVE_CTX)
+		acpi_lpss_save_ctx(dev, pdata);
+
 	return acpi_dev_suspend_late(dev);
 }
 
 static int acpi_lpss_resume_early(struct device *dev)
 {
-	int ret = acpi_dev_resume_early(dev);
+	struct lpss_private_data *pdata = acpi_driver_data(ACPI_COMPANION(dev));
+	int ret;
 
+	ret = acpi_dev_resume_early(dev);
 	if (ret)
 		return ret;
 
-	acpi_lpss_restore_ctx(dev);
+	if (pdata->dev_desc->flags & LPSS_SAVE_CTX)
+		acpi_lpss_restore_ctx(dev, pdata);
+
 	return pm_generic_resume_early(dev);
 }
 #endif /* CONFIG_PM_SLEEP */
 
-#ifdef CONFIG_PM_RUNTIME
 static int acpi_lpss_runtime_suspend(struct device *dev)
 {
-	int ret = pm_generic_runtime_suspend(dev);
+	struct lpss_private_data *pdata = acpi_driver_data(ACPI_COMPANION(dev));
+	int ret;
 
+	ret = pm_generic_runtime_suspend(dev);
 	if (ret)
 		return ret;
 
-	acpi_lpss_save_ctx(dev);
-	return acpi_dev_runtime_suspend(dev);
+	if (pdata->dev_desc->flags & LPSS_SAVE_CTX)
+		acpi_lpss_save_ctx(dev, pdata);
+
+	ret = acpi_dev_runtime_suspend(dev);
+	if (ret)
+		return ret;
+
+	if (pdata->dev_desc->flags & LPSS_PROXY_REQ && proxy_device)
+		return pm_runtime_put_sync_suspend(proxy_device);
+
+	return 0;
 }
 
 static int acpi_lpss_runtime_resume(struct device *dev)
 {
-	int ret = acpi_dev_runtime_resume(dev);
+	struct lpss_private_data *pdata = acpi_driver_data(ACPI_COMPANION(dev));
+	int ret;
+
+	if (pdata->dev_desc->flags & LPSS_PROXY_REQ && proxy_device) {
+		ret = pm_runtime_get_sync(proxy_device);
+		if (ret)
+			return ret;
+	}
 
+	ret = acpi_dev_runtime_resume(dev);
 	if (ret)
 		return ret;
 
-	acpi_lpss_restore_ctx(dev);
+	if (pdata->dev_desc->flags & LPSS_SAVE_CTX)
+		acpi_lpss_restore_ctx(dev, pdata);
+
 	return pm_generic_runtime_resume(dev);
 }
-#endif /* CONFIG_PM_RUNTIME */
 #endif /* CONFIG_PM */
 
 static struct dev_pm_domain acpi_lpss_pm_domain = {
 	.ops = {
+#ifdef CONFIG_PM
 #ifdef CONFIG_PM_SLEEP
 		.prepare = acpi_subsys_prepare,
 		.complete = acpi_subsys_complete,
@@ -659,7 +646,6 @@ static struct dev_pm_domain acpi_lpss_pm_domain = {
 		.poweroff_late = acpi_lpss_suspend_late,
 		.restore_early = acpi_lpss_resume_early,
 #endif
-#ifdef CONFIG_PM_RUNTIME
 		.runtime_suspend = acpi_lpss_runtime_suspend,
 		.runtime_resume = acpi_lpss_runtime_resume,
 #endif
@@ -682,30 +668,27 @@ static int acpi_lpss_platform_notify(struct notifier_block *nb,
 		return 0;
 
 	pdata = acpi_driver_data(adev);
-	if (!pdata || !pdata->mmio_base)
+	if (!pdata)
 		return 0;
 
-	if (pdata->mmio_size < pdata->dev_desc->prv_offset + LPSS_LTR_SIZE) {
+	if (pdata->mmio_base &&
+	    pdata->mmio_size < pdata->dev_desc->prv_offset + LPSS_LTR_SIZE) {
 		dev_err(&pdev->dev, "MMIO size insufficient to access LTR\n");
 		return 0;
 	}
 
 	switch (action) {
-	case BUS_NOTIFY_BOUND_DRIVER:
-		if (pdata->dev_desc->save_ctx)
-			pdev->dev.pm_domain = &acpi_lpss_pm_domain;
-		break;
-	case BUS_NOTIFY_UNBOUND_DRIVER:
-		if (pdata->dev_desc->save_ctx)
-			pdev->dev.pm_domain = NULL;
-		break;
 	case BUS_NOTIFY_ADD_DEVICE:
-		if (pdata->dev_desc->ltr_required)
+		pdev->dev.pm_domain = &acpi_lpss_pm_domain;
+		if (pdata->dev_desc->flags & LPSS_LTR)
 			return sysfs_create_group(&pdev->dev.kobj,
 						  &lpss_attr_group);
+		break;
 	case BUS_NOTIFY_DEL_DEVICE:
-		if (pdata->dev_desc->ltr_required)
+		if (pdata->dev_desc->flags & LPSS_LTR)
 			sysfs_remove_group(&pdev->dev.kobj, &lpss_attr_group);
+		pdev->dev.pm_domain = NULL;
+		break;
 	default:
 		break;
 	}
@@ -721,7 +704,7 @@ static void acpi_lpss_bind(struct device *dev)
 {
 	struct lpss_private_data *pdata = acpi_driver_data(ACPI_COMPANION(dev));
 
-	if (!pdata || !pdata->mmio_base || !pdata->dev_desc->ltr_required)
+	if (!pdata || !pdata->mmio_base || !(pdata->dev_desc->flags & LPSS_LTR))
 		return;
 
 	if (pdata->mmio_size >= pdata->dev_desc->prv_offset + LPSS_LTR_SIZE)
diff --git a/drivers/acpi/acpi_pad.c b/drivers/acpi/acpi_pad.c
index f148a0580e04..c7b105c0e1d3 100644
--- a/drivers/acpi/acpi_pad.c
+++ b/drivers/acpi/acpi_pad.c
@@ -350,12 +350,10 @@ static ssize_t acpi_pad_idlecpus_store(struct device *dev,
 static ssize_t acpi_pad_idlecpus_show(struct device *dev,
 	struct device_attribute *attr, char *buf)
 {
-	int n = 0;
-	n = cpumask_scnprintf(buf, PAGE_SIZE-2, to_cpumask(pad_busy_cpus_bits));
-	buf[n++] = '\n';
-	buf[n] = '\0';
-	return n;
+	return cpumap_print_to_pagebuf(false, buf,
+				       to_cpumask(pad_busy_cpus_bits));
 }
+
 static DEVICE_ATTR(idlecpus, S_IRUGO|S_IWUSR,
 	acpi_pad_idlecpus_show,
 	acpi_pad_idlecpus_store);
diff --git a/drivers/acpi/acpi_platform.c b/drivers/acpi/acpi_platform.c
index 2bf9082f7523..6ba8beb6b9d2 100644
--- a/drivers/acpi/acpi_platform.c
+++ b/drivers/acpi/acpi_platform.c
@@ -16,6 +16,7 @@
 #include <linux/err.h>
 #include <linux/kernel.h>
 #include <linux/module.h>
+#include <linux/dma-mapping.h>
 #include <linux/platform_device.h>
 
 #include "internal.h"
@@ -102,6 +103,7 @@ struct platform_device *acpi_create_platform_device(struct acpi_device *adev)
 	pdevinfo.res = resources;
 	pdevinfo.num_res = count;
 	pdevinfo.acpi_node.companion = adev;
+	pdevinfo.dma_mask = DMA_BIT_MASK(32);
 	pdev = platform_device_register_full(&pdevinfo);
 	if (IS_ERR(pdev))
 		dev_err(&adev->dev, "platform device creation failed: %ld\n",
@@ -113,3 +115,4 @@ struct platform_device *acpi_create_platform_device(struct acpi_device *adev)
 	kfree(resources);
 	return pdev;
 }
+EXPORT_SYMBOL_GPL(acpi_create_platform_device);
diff --git a/drivers/acpi/acpi_pnp.c b/drivers/acpi/acpi_pnp.c
index 1f8b20496f32..b193f8425999 100644
--- a/drivers/acpi/acpi_pnp.c
+++ b/drivers/acpi/acpi_pnp.c
@@ -130,10 +130,6 @@ static const struct acpi_device_id acpi_pnp_device_ids[] = {
 	{"PNP0401"},		/* ECP Printer Port */
 	/* apple-gmux */
 	{"APP000B"},
-	/* fujitsu-laptop.c */
-	{"FUJ02bf"},
-	{"FUJ02B1"},
-	{"FUJ02E3"},
 	/* system */
 	{"PNP0c02"},		/* General ID for reserving resources */
 	{"PNP0c01"},		/* memory controller */
diff --git a/drivers/acpi/acpi_processor.c b/drivers/acpi/acpi_processor.c
index 1fdf5e07a1c7..1020b1b53a17 100644
--- a/drivers/acpi/acpi_processor.c
+++ b/drivers/acpi/acpi_processor.c
@@ -170,7 +170,7 @@ static int acpi_processor_hotadd_init(struct acpi_processor *pr)
 	acpi_status status;
 	int ret;
 
-	if (pr->apic_id == -1)
+	if (pr->phys_id == -1)
 		return -ENODEV;
 
 	status = acpi_evaluate_integer(pr->handle, "_STA", NULL, &sta);
@@ -180,13 +180,13 @@ static int acpi_processor_hotadd_init(struct acpi_processor *pr)
 	cpu_maps_update_begin();
 	cpu_hotplug_begin();
 
-	ret = acpi_map_lsapic(pr->handle, pr->apic_id, &pr->id);
+	ret = acpi_map_cpu(pr->handle, pr->phys_id, &pr->id);
 	if (ret)
 		goto out;
 
 	ret = arch_register_cpu(pr->id);
 	if (ret) {
-		acpi_unmap_lsapic(pr->id);
+		acpi_unmap_cpu(pr->id);
 		goto out;
 	}
 
@@ -215,7 +215,7 @@ static int acpi_processor_get_info(struct acpi_device *device)
 	union acpi_object object = { 0 };
 	struct acpi_buffer buffer = { sizeof(union acpi_object), &object };
 	struct acpi_processor *pr = acpi_driver_data(device);
-	int apic_id, cpu_index, device_declaration = 0;
+	int phys_id, cpu_index, device_declaration = 0;
 	acpi_status status = AE_OK;
 	static int cpu0_initialized;
 	unsigned long long value;
@@ -262,15 +262,18 @@ static int acpi_processor_get_info(struct acpi_device *device)
 		pr->acpi_id = value;
 	}
 
-	apic_id = acpi_get_apicid(pr->handle, device_declaration, pr->acpi_id);
-	if (apic_id < 0)
-		acpi_handle_debug(pr->handle, "failed to get CPU APIC ID.\n");
-	pr->apic_id = apic_id;
+	phys_id = acpi_get_phys_id(pr->handle, device_declaration, pr->acpi_id);
+	if (phys_id < 0)
+		acpi_handle_debug(pr->handle, "failed to get CPU physical ID.\n");
+	pr->phys_id = phys_id;
 
-	cpu_index = acpi_map_cpuid(pr->apic_id, pr->acpi_id);
+	cpu_index = acpi_map_cpuid(pr->phys_id, pr->acpi_id);
 	if (!cpu0_initialized && !acpi_has_cpu_in_madt()) {
 		cpu0_initialized = 1;
-		/* Handle UP system running SMP kernel, with no LAPIC in MADT */
+		/*
+		 * Handle UP system running SMP kernel, with no CPU
+		 * entry in MADT
+		 */
 		if ((cpu_index == -1) && (num_online_cpus() == 1))
 			cpu_index = 0;
 	}
@@ -458,7 +461,7 @@ static void acpi_processor_remove(struct acpi_device *device)
 
 	/* Remove the CPU. */
 	arch_unregister_cpu(pr->id);
-	acpi_unmap_lsapic(pr->id);
+	acpi_unmap_cpu(pr->id);
 
 	cpu_hotplug_done();
 	cpu_maps_update_done();
diff --git a/drivers/acpi/acpica/acglobal.h b/drivers/acpi/acpica/acglobal.h
index ebf02cc10a43..7f60582d0c8c 100644
--- a/drivers/acpi/acpica/acglobal.h
+++ b/drivers/acpi/acpica/acglobal.h
@@ -305,6 +305,7 @@ ACPI_INIT_GLOBAL(u8, acpi_gbl_db_output_flags, ACPI_DB_CONSOLE_OUTPUT);
 
 ACPI_INIT_GLOBAL(u8, acpi_gbl_no_resource_disassembly, FALSE);
 ACPI_INIT_GLOBAL(u8, acpi_gbl_ignore_noop_operator, FALSE);
+ACPI_INIT_GLOBAL(u8, acpi_gbl_cstyle_disassembly, TRUE);
 
 ACPI_GLOBAL(u8, acpi_gbl_db_opt_disasm);
 ACPI_GLOBAL(u8, acpi_gbl_db_opt_verbose);
diff --git a/drivers/acpi/acpica/achware.h b/drivers/acpi/acpica/achware.h
index 2ad2351a9833..c318d3e27893 100644
--- a/drivers/acpi/acpica/achware.h
+++ b/drivers/acpi/acpica/achware.h
@@ -127,7 +127,7 @@ acpi_hw_clear_gpe_block(struct acpi_gpe_xrupt_info *gpe_xrupt_info,
 
 acpi_status
 acpi_hw_get_gpe_status(struct acpi_gpe_event_info *gpe_event_info,
-		       acpi_event_status * event_status);
+		       acpi_event_status *event_status);
 
 acpi_status acpi_hw_disable_all_gpes(void);
 
diff --git a/drivers/acpi/acpica/aclocal.h b/drivers/acpi/acpica/aclocal.h
index 2747279fbe3c..680d23bbae7c 100644
--- a/drivers/acpi/acpica/aclocal.h
+++ b/drivers/acpi/acpica/aclocal.h
@@ -413,8 +413,8 @@ struct acpi_gpe_handler_info {
 	acpi_gpe_handler address;	/* Address of handler, if any */
 	void *context;		/* Context to be passed to handler */
 	struct acpi_namespace_node *method_node;	/* Method node for this GPE level (saved) */
-	u8 original_flags;      /* Original (pre-handler) GPE info */
-	u8 originally_enabled;  /* True if GPE was originally enabled */
+	u8 original_flags;	/* Original (pre-handler) GPE info */
+	u8 originally_enabled;	/* True if GPE was originally enabled */
 };
 
 /* Notify info for implicit notify, multiple device objects */
@@ -454,6 +454,7 @@ struct acpi_gpe_register_info {
 	u16 base_gpe_number;	/* Base GPE number for this register */
 	u8 enable_for_wake;	/* GPEs to keep enabled when sleeping */
 	u8 enable_for_run;	/* GPEs to keep enabled when running */
+	u8 enable_mask;		/* Current mask of enabled GPEs */
 };
 
 /*
@@ -722,6 +723,7 @@ union acpi_parse_value {
 	ACPI_DISASM_ONLY_MEMBERS (\
 	u8                              disasm_flags;   /* Used during AML disassembly */\
 	u8                              disasm_opcode;  /* Subtype used for disassembly */\
+	char                            *operator_symbol;/* Used for C-style operator name strings */\
 	char                            aml_op_name[16])	/* Op name (debug only) */
 
 /* Flags for disasm_flags field above */
@@ -827,6 +829,8 @@ struct acpi_parse_state {
 #define ACPI_PARSEOP_EMPTY_TERMLIST     0x04
 #define ACPI_PARSEOP_PREDEF_CHECKED     0x08
 #define ACPI_PARSEOP_SPECIAL            0x10
+#define ACPI_PARSEOP_COMPOUND           0x20
+#define ACPI_PARSEOP_ASSIGNMENT         0x40
 
 /*****************************************************************************
  *
diff --git a/drivers/acpi/acpica/actables.h b/drivers/acpi/acpica/actables.h
index f14882788eee..1afe46e44dac 100644
--- a/drivers/acpi/acpica/actables.h
+++ b/drivers/acpi/acpica/actables.h
@@ -49,6 +49,8 @@ acpi_status acpi_allocate_root_table(u32 initial_table_count);
 /*
  * tbxfroot - Root pointer utilities
  */
+u32 acpi_tb_get_rsdp_length(struct acpi_table_rsdp *rsdp);
+
 acpi_status acpi_tb_validate_rsdp(struct acpi_table_rsdp *rsdp);
 
 u8 *acpi_tb_scan_memory_for_rsdp(u8 *start_address, u32 length);
diff --git a/drivers/acpi/acpica/amlresrc.h b/drivers/acpi/acpica/amlresrc.h
index f3f834408441..3a0beeb86ba5 100644
--- a/drivers/acpi/acpica/amlresrc.h
+++ b/drivers/acpi/acpica/amlresrc.h
@@ -117,6 +117,12 @@ struct asl_resource_node {
 	struct asl_resource_node *next;
 };
 
+struct asl_resource_info {
+	union acpi_parse_object *descriptor_type_op;	/* Resource descriptor parse node */
+	union acpi_parse_object *mapping_op;	/* Used for mapfile support */
+	u32 current_byte_offset;	/* Offset in resource template */
+};
+
 /* Macros used to generate AML resource length fields */
 
 #define ACPI_AML_SIZE_LARGE(r)      (sizeof (r) - sizeof (struct aml_resource_large_header))
@@ -449,4 +455,32 @@ union aml_resource {
 	u8 byte_item;
 };
 
+/* Interfaces used by both the disassembler and compiler */
+
+void
+mp_save_gpio_info(union acpi_parse_object *op,
+		  union aml_resource *resource,
+		  u32 pin_count, u16 *pin_list, char *device_name);
+
+void
+mp_save_serial_info(union acpi_parse_object *op,
+		    union aml_resource *resource, char *device_name);
+
+char *mp_get_hid_from_parse_tree(struct acpi_namespace_node *hid_node);
+
+char *mp_get_hid_via_namestring(char *device_name);
+
+char *mp_get_connection_info(union acpi_parse_object *op,
+			     u32 pin_index,
+			     struct acpi_namespace_node **target_node,
+			     char **target_name);
+
+char *mp_get_parent_device_hid(union acpi_parse_object *op,
+			       struct acpi_namespace_node **target_node,
+			       char **parent_device_name);
+
+char *mp_get_ddn_value(char *device_name);
+
+char *mp_get_hid_value(struct acpi_namespace_node *device_node);
+
 #endif
diff --git a/drivers/acpi/acpica/evgpe.c b/drivers/acpi/acpica/evgpe.c
index e4ba4dec86af..aa70154cf4fa 100644
--- a/drivers/acpi/acpica/evgpe.c
+++ b/drivers/acpi/acpica/evgpe.c
@@ -100,13 +100,14 @@ acpi_ev_update_gpe_enable_mask(struct acpi_gpe_event_info *gpe_event_info)
  *
  * FUNCTION:    acpi_ev_enable_gpe
  *
- * PARAMETERS:  gpe_event_info  - GPE to enable
+ * PARAMETERS:  gpe_event_info          - GPE to enable
  *
  * RETURN:      Status
  *
  * DESCRIPTION: Clear a GPE of stale events and enable it.
  *
  ******************************************************************************/
+
 acpi_status acpi_ev_enable_gpe(struct acpi_gpe_event_info *gpe_event_info)
 {
 	acpi_status status;
@@ -125,6 +126,7 @@ acpi_status acpi_ev_enable_gpe(struct acpi_gpe_event_info *gpe_event_info)
 	}
 
 	/* Clear the GPE (of stale events) */
+
 	status = acpi_hw_clear_gpe(gpe_event_info);
 	if (ACPI_FAILURE(status)) {
 		return_ACPI_STATUS(status);
@@ -132,11 +134,10 @@ acpi_status acpi_ev_enable_gpe(struct acpi_gpe_event_info *gpe_event_info)
 
 	/* Enable the requested GPE */
 
-	status = acpi_hw_low_set_gpe(gpe_event_info, ACPI_GPE_ENABLE);
+	status = acpi_hw_low_set_gpe(gpe_event_info, ACPI_GPE_ENABLE_SAVE);
 	return_ACPI_STATUS(status);
 }
 
-
 /*******************************************************************************
  *
  * FUNCTION:    acpi_ev_add_gpe_reference
@@ -212,7 +213,7 @@ acpi_ev_remove_gpe_reference(struct acpi_gpe_event_info *gpe_event_info)
 		if (ACPI_SUCCESS(status)) {
 			status =
 			    acpi_hw_low_set_gpe(gpe_event_info,
-						     ACPI_GPE_DISABLE);
+						ACPI_GPE_DISABLE_SAVE);
 		}
 
 		if (ACPI_FAILURE(status)) {
@@ -334,7 +335,7 @@ struct acpi_gpe_event_info *acpi_ev_get_gpe_event_info(acpi_handle gpe_device,
  *
  ******************************************************************************/
 
-u32 acpi_ev_gpe_detect(struct acpi_gpe_xrupt_info * gpe_xrupt_list)
+u32 acpi_ev_gpe_detect(struct acpi_gpe_xrupt_info *gpe_xrupt_list)
 {
 	acpi_status status;
 	struct acpi_gpe_block_info *gpe_block;
@@ -427,7 +428,7 @@ u32 acpi_ev_gpe_detect(struct acpi_gpe_xrupt_info * gpe_xrupt_list)
 
 			/* Check if there is anything active at all in this register */
 
-			enabled_status_byte = (u8) (status_reg & enable_reg);
+			enabled_status_byte = (u8)(status_reg & enable_reg);
 			if (!enabled_status_byte) {
 
 				/* No active GPEs in this register, move on */
@@ -450,7 +451,7 @@ u32 acpi_ev_gpe_detect(struct acpi_gpe_xrupt_info * gpe_xrupt_list)
 					    acpi_ev_gpe_dispatch(gpe_block->
 								 node,
 								 &gpe_block->
-						event_info[((acpi_size) i * ACPI_GPE_REGISTER_WIDTH) + j], j + gpe_register_info->base_gpe_number);
+								 event_info[((acpi_size) i * ACPI_GPE_REGISTER_WIDTH) + j], j + gpe_register_info->base_gpe_number);
 				}
 			}
 		}
@@ -615,8 +616,11 @@ static void ACPI_SYSTEM_XFACE acpi_ev_asynch_execute_gpe_method(void *context)
 static void ACPI_SYSTEM_XFACE acpi_ev_asynch_enable_gpe(void *context)
 {
 	struct acpi_gpe_event_info *gpe_event_info = context;
+	acpi_cpu_flags flags;
 
+	flags = acpi_os_acquire_lock(acpi_gbl_gpe_lock);
 	(void)acpi_ev_finish_gpe(gpe_event_info);
+	acpi_os_release_lock(acpi_gbl_gpe_lock, flags);
 
 	ACPI_FREE(gpe_event_info);
 	return;
@@ -636,7 +640,7 @@ static void ACPI_SYSTEM_XFACE acpi_ev_asynch_enable_gpe(void *context)
  *
  ******************************************************************************/
 
-acpi_status acpi_ev_finish_gpe(struct acpi_gpe_event_info *gpe_event_info)
+acpi_status acpi_ev_finish_gpe(struct acpi_gpe_event_info * gpe_event_info)
 {
 	acpi_status status;
 
@@ -654,7 +658,7 @@ acpi_status acpi_ev_finish_gpe(struct acpi_gpe_event_info *gpe_event_info)
 
 	/*
 	 * Enable this GPE, conditionally. This means that the GPE will
-	 * only be physically enabled if the enable_for_run bit is set
+	 * only be physically enabled if the enable_mask bit is set
 	 * in the event_info.
 	 */
 	(void)acpi_hw_low_set_gpe(gpe_event_info, ACPI_GPE_CONDITIONAL_ENABLE);
@@ -666,9 +670,9 @@ acpi_status acpi_ev_finish_gpe(struct acpi_gpe_event_info *gpe_event_info)
  *
  * FUNCTION:    acpi_ev_gpe_dispatch
  *
- * PARAMETERS:  gpe_device      - Device node. NULL for GPE0/GPE1
- *              gpe_event_info  - Info for this GPE
- *              gpe_number      - Number relative to the parent GPE block
+ * PARAMETERS:  gpe_device          - Device node. NULL for GPE0/GPE1
+ *              gpe_event_info      - Info for this GPE
+ *              gpe_number          - Number relative to the parent GPE block
  *
  * RETURN:      INTERRUPT_HANDLED or INTERRUPT_NOT_HANDLED
  *
@@ -681,7 +685,7 @@ acpi_status acpi_ev_finish_gpe(struct acpi_gpe_event_info *gpe_event_info)
 
 u32
 acpi_ev_gpe_dispatch(struct acpi_namespace_node *gpe_device,
-		    struct acpi_gpe_event_info *gpe_event_info, u32 gpe_number)
+		     struct acpi_gpe_event_info *gpe_event_info, u32 gpe_number)
 {
 	acpi_status status;
 	u32 return_value;
diff --git a/drivers/acpi/acpica/evgpeinit.c b/drivers/acpi/acpica/evgpeinit.c
index 49fc7effd961..7be928379879 100644
--- a/drivers/acpi/acpica/evgpeinit.c
+++ b/drivers/acpi/acpica/evgpeinit.c
@@ -424,6 +424,7 @@ acpi_ev_match_gpe_method(acpi_handle obj_handle,
 	}
 
 	/* Disable the GPE in case it's been enabled already. */
+
 	(void)acpi_hw_low_set_gpe(gpe_event_info, ACPI_GPE_DISABLE);
 
 	/*
diff --git a/drivers/acpi/acpica/evxface.c b/drivers/acpi/acpica/evxface.c
index 11e5803b8b41..55a58f3ec8df 100644
--- a/drivers/acpi/acpica/evxface.c
+++ b/drivers/acpi/acpica/evxface.c
@@ -786,18 +786,26 @@ acpi_install_gpe_handler(acpi_handle gpe_device,
 	handler->method_node = gpe_event_info->dispatch.method_node;
 	handler->original_flags = (u8)(gpe_event_info->flags &
 				       (ACPI_GPE_XRUPT_TYPE_MASK |
-				        ACPI_GPE_DISPATCH_MASK));
+					ACPI_GPE_DISPATCH_MASK));
 
 	/*
 	 * If the GPE is associated with a method, it may have been enabled
 	 * automatically during initialization, in which case it has to be
 	 * disabled now to avoid spurious execution of the handler.
 	 */
-
-	if ((handler->original_flags & ACPI_GPE_DISPATCH_METHOD)
-	    && gpe_event_info->runtime_count) {
-		handler->originally_enabled = 1;
+	if (((handler->original_flags & ACPI_GPE_DISPATCH_METHOD) ||
+	     (handler->original_flags & ACPI_GPE_DISPATCH_NOTIFY)) &&
+	    gpe_event_info->runtime_count) {
+		handler->originally_enabled = TRUE;
 		(void)acpi_ev_remove_gpe_reference(gpe_event_info);
+
+		/* Sanity check of original type against new type */
+
+		if (type !=
+		    (u32)(gpe_event_info->flags & ACPI_GPE_XRUPT_TYPE_MASK)) {
+			ACPI_WARNING((AE_INFO,
+				      "GPE type mismatch (level/edge)"));
+		}
 	}
 
 	/* Install the handler */
@@ -808,7 +816,7 @@ acpi_install_gpe_handler(acpi_handle gpe_device,
 
 	gpe_event_info->flags &=
 	    ~(ACPI_GPE_XRUPT_TYPE_MASK | ACPI_GPE_DISPATCH_MASK);
-	gpe_event_info->flags |= (u8) (type | ACPI_GPE_DISPATCH_HANDLER);
+	gpe_event_info->flags |= (u8)(type | ACPI_GPE_DISPATCH_HANDLER);
 
 	acpi_os_release_lock(acpi_gbl_gpe_lock, flags);
 
@@ -893,7 +901,7 @@ acpi_remove_gpe_handler(acpi_handle gpe_device,
 
 	gpe_event_info->dispatch.method_node = handler->method_node;
 	gpe_event_info->flags &=
-		~(ACPI_GPE_XRUPT_TYPE_MASK | ACPI_GPE_DISPATCH_MASK);
+	    ~(ACPI_GPE_XRUPT_TYPE_MASK | ACPI_GPE_DISPATCH_MASK);
 	gpe_event_info->flags |= handler->original_flags;
 
 	/*
@@ -901,7 +909,8 @@ acpi_remove_gpe_handler(acpi_handle gpe_device,
 	 * enabled, it should be enabled at this point to restore the
 	 * post-initialization configuration.
 	 */
-	if ((handler->original_flags & ACPI_GPE_DISPATCH_METHOD) &&
+	if (((handler->original_flags & ACPI_GPE_DISPATCH_METHOD) ||
+	     (handler->original_flags & ACPI_GPE_DISPATCH_NOTIFY)) &&
 	    handler->originally_enabled) {
 		(void)acpi_ev_add_gpe_reference(gpe_event_info);
 	}
@@ -946,7 +955,7 @@ ACPI_EXPORT_SYMBOL(acpi_remove_gpe_handler)
  * handle is returned.
  *
  ******************************************************************************/
-acpi_status acpi_acquire_global_lock(u16 timeout, u32 * handle)
+acpi_status acpi_acquire_global_lock(u16 timeout, u32 *handle)
 {
 	acpi_status status;
 
diff --git a/drivers/acpi/acpica/evxfevnt.c b/drivers/acpi/acpica/evxfevnt.c
index e286640ad4ff..bb8cbf5961bf 100644
--- a/drivers/acpi/acpica/evxfevnt.c
+++ b/drivers/acpi/acpica/evxfevnt.c
@@ -324,8 +324,9 @@ ACPI_EXPORT_SYMBOL(acpi_clear_event)
  ******************************************************************************/
 acpi_status acpi_get_event_status(u32 event, acpi_event_status * event_status)
 {
-	acpi_status status = AE_OK;
-	u32 value;
+	acpi_status status;
+	acpi_event_status local_event_status = 0;
+	u32 in_byte;
 
 	ACPI_FUNCTION_TRACE(acpi_get_event_status);
 
@@ -339,29 +340,40 @@ acpi_status acpi_get_event_status(u32 event, acpi_event_status * event_status)
 		return_ACPI_STATUS(AE_BAD_PARAMETER);
 	}
 
-	/* Get the status of the requested fixed event */
+	/* Fixed event currently can be dispatched? */
+
+	if (acpi_gbl_fixed_event_handlers[event].handler) {
+		local_event_status |= ACPI_EVENT_FLAG_HAS_HANDLER;
+	}
+
+	/* Fixed event currently enabled? */
 
 	status =
 	    acpi_read_bit_register(acpi_gbl_fixed_event_info[event].
-			      enable_register_id, &value);
-	if (ACPI_FAILURE(status))
+				   enable_register_id, &in_byte);
+	if (ACPI_FAILURE(status)) {
 		return_ACPI_STATUS(status);
+	}
 
-	*event_status = value;
+	if (in_byte) {
+		local_event_status |= ACPI_EVENT_FLAG_ENABLED;
+	}
+
+	/* Fixed event currently active? */
 
 	status =
 	    acpi_read_bit_register(acpi_gbl_fixed_event_info[event].
-			      status_register_id, &value);
-	if (ACPI_FAILURE(status))
+				   status_register_id, &in_byte);
+	if (ACPI_FAILURE(status)) {
 		return_ACPI_STATUS(status);
+	}
 
-	if (value)
-		*event_status |= ACPI_EVENT_FLAG_SET;
-
-	if (acpi_gbl_fixed_event_handlers[event].handler)
-		*event_status |= ACPI_EVENT_FLAG_HANDLE;
+	if (in_byte) {
+		local_event_status |= ACPI_EVENT_FLAG_SET;
+	}
 
-	return_ACPI_STATUS(status);
+	(*event_status) = local_event_status;
+	return_ACPI_STATUS(AE_OK);
 }
 
 ACPI_EXPORT_SYMBOL(acpi_get_event_status)
diff --git a/drivers/acpi/acpica/evxfgpe.c b/drivers/acpi/acpica/evxfgpe.c
index 0cf159cc6e6d..e889a5304abd 100644
--- a/drivers/acpi/acpica/evxfgpe.c
+++ b/drivers/acpi/acpica/evxfgpe.c
@@ -106,8 +106,8 @@ ACPI_EXPORT_SYMBOL(acpi_update_all_gpes)
  *
  * FUNCTION:    acpi_enable_gpe
  *
- * PARAMETERS:  gpe_device      - Parent GPE Device. NULL for GPE0/GPE1
- *              gpe_number      - GPE level within the GPE block
+ * PARAMETERS:  gpe_device          - Parent GPE Device. NULL for GPE0/GPE1
+ *              gpe_number          - GPE level within the GPE block
  *
  * RETURN:      Status
  *
@@ -115,7 +115,6 @@ ACPI_EXPORT_SYMBOL(acpi_update_all_gpes)
  *              hardware-enabled.
  *
  ******************************************************************************/
-
 acpi_status acpi_enable_gpe(acpi_handle gpe_device, u32 gpe_number)
 {
 	acpi_status status = AE_BAD_PARAMETER;
@@ -490,8 +489,8 @@ ACPI_EXPORT_SYMBOL(acpi_clear_gpe)
  *
  * FUNCTION:    acpi_get_gpe_status
  *
- * PARAMETERS:  gpe_device      - Parent GPE Device. NULL for GPE0/GPE1
- *              gpe_number      - GPE level within the GPE block
+ * PARAMETERS:  gpe_device          - Parent GPE Device. NULL for GPE0/GPE1
+ *              gpe_number          - GPE level within the GPE block
  *              event_status        - Where the current status of the event
  *                                    will be returned
  *
@@ -524,9 +523,6 @@ acpi_get_gpe_status(acpi_handle gpe_device,
 
 	status = acpi_hw_get_gpe_status(gpe_event_info, event_status);
 
-	if (gpe_event_info->flags & ACPI_GPE_DISPATCH_MASK)
-		*event_status |= ACPI_EVENT_FLAG_HANDLE;
-
 unlock_and_exit:
 	acpi_os_release_lock(acpi_gbl_gpe_lock, flags);
 	return_ACPI_STATUS(status);
@@ -596,6 +592,38 @@ acpi_status acpi_enable_all_runtime_gpes(void)
 
 ACPI_EXPORT_SYMBOL(acpi_enable_all_runtime_gpes)
 
+/******************************************************************************
+ *
+ * FUNCTION:    acpi_enable_all_wakeup_gpes
+ *
+ * PARAMETERS:  None
+ *
+ * RETURN:      Status
+ *
+ * DESCRIPTION: Enable all "wakeup" GPEs and disable all of the other GPEs, in
+ *              all GPE blocks.
+ *
+ ******************************************************************************/
+
+acpi_status acpi_enable_all_wakeup_gpes(void)
+{
+	acpi_status status;
+
+	ACPI_FUNCTION_TRACE(acpi_enable_all_wakeup_gpes);
+
+	status = acpi_ut_acquire_mutex(ACPI_MTX_EVENTS);
+	if (ACPI_FAILURE(status)) {
+		return_ACPI_STATUS(status);
+	}
+
+	status = acpi_hw_enable_all_wakeup_gpes();
+	(void)acpi_ut_release_mutex(ACPI_MTX_EVENTS);
+
+	return_ACPI_STATUS(status);
+}
+
+ACPI_EXPORT_SYMBOL(acpi_enable_all_wakeup_gpes)
+
 /*******************************************************************************
  *
  * FUNCTION:    acpi_install_gpe_block
diff --git a/drivers/acpi/acpica/hwgpe.c b/drivers/acpi/acpica/hwgpe.c
index 2e6caabba07a..494027f5c067 100644
--- a/drivers/acpi/acpica/hwgpe.c
+++ b/drivers/acpi/acpica/hwgpe.c
@@ -115,12 +115,12 @@ acpi_hw_low_set_gpe(struct acpi_gpe_event_info *gpe_event_info, u32 action)
 	/* Set or clear just the bit that corresponds to this GPE */
 
 	register_bit = acpi_hw_get_gpe_register_bit(gpe_event_info);
-	switch (action) {
+	switch (action & ~ACPI_GPE_SAVE_MASK) {
 	case ACPI_GPE_CONDITIONAL_ENABLE:
 
-		/* Only enable if the enable_for_run bit is set */
+		/* Only enable if the corresponding enable_mask bit is set */
 
-		if (!(register_bit & gpe_register_info->enable_for_run)) {
+		if (!(register_bit & gpe_register_info->enable_mask)) {
 			return (AE_BAD_PARAMETER);
 		}
 
@@ -145,6 +145,9 @@ acpi_hw_low_set_gpe(struct acpi_gpe_event_info *gpe_event_info, u32 action)
 	/* Write the updated enable mask */
 
 	status = acpi_hw_write(enable_mask, &gpe_register_info->enable_address);
+	if (ACPI_SUCCESS(status) && (action & ACPI_GPE_SAVE_MASK)) {
+		gpe_register_info->enable_mask = enable_mask;
+	}
 	return (status);
 }
 
@@ -202,7 +205,7 @@ acpi_status acpi_hw_clear_gpe(struct acpi_gpe_event_info * gpe_event_info)
 
 acpi_status
 acpi_hw_get_gpe_status(struct acpi_gpe_event_info * gpe_event_info,
-		       acpi_event_status * event_status)
+		       acpi_event_status *event_status)
 {
 	u32 in_byte;
 	u32 register_bit;
@@ -216,6 +219,13 @@ acpi_hw_get_gpe_status(struct acpi_gpe_event_info * gpe_event_info,
 		return (AE_BAD_PARAMETER);
 	}
 
+	/* GPE currently handled? */
+
+	if ((gpe_event_info->flags & ACPI_GPE_DISPATCH_MASK) !=
+	    ACPI_GPE_DISPATCH_NONE) {
+		local_event_status |= ACPI_EVENT_FLAG_HAS_HANDLER;
+	}
+
 	/* Get the info block for the entire GPE register */
 
 	gpe_register_info = gpe_event_info->register_info;
@@ -255,6 +265,32 @@ acpi_hw_get_gpe_status(struct acpi_gpe_event_info * gpe_event_info,
 
 /******************************************************************************
  *
+ * FUNCTION:    acpi_hw_gpe_enable_write
+ *
+ * PARAMETERS:  enable_mask         - Bit mask to write to the GPE register
+ *              gpe_register_info   - Gpe Register info
+ *
+ * RETURN:      Status
+ *
+ * DESCRIPTION: Write the enable mask byte to the given GPE register.
+ *
+ ******************************************************************************/
+
+static acpi_status
+acpi_hw_gpe_enable_write(u8 enable_mask,
+			 struct acpi_gpe_register_info *gpe_register_info)
+{
+	acpi_status status;
+
+	status = acpi_hw_write(enable_mask, &gpe_register_info->enable_address);
+	if (ACPI_SUCCESS(status)) {
+		gpe_register_info->enable_mask = enable_mask;
+	}
+	return (status);
+}
+
+/******************************************************************************
+ *
  * FUNCTION:    acpi_hw_disable_gpe_block
  *
  * PARAMETERS:  gpe_xrupt_info      - GPE Interrupt info
@@ -280,8 +316,8 @@ acpi_hw_disable_gpe_block(struct acpi_gpe_xrupt_info *gpe_xrupt_info,
 		/* Disable all GPEs in this register */
 
 		status =
-		    acpi_hw_write(0x00,
-				  &gpe_block->register_info[i].enable_address);
+		    acpi_hw_gpe_enable_write(0x00,
+					     &gpe_block->register_info[i]);
 		if (ACPI_FAILURE(status)) {
 			return (status);
 		}
@@ -348,21 +384,23 @@ acpi_hw_enable_runtime_gpe_block(struct acpi_gpe_xrupt_info *gpe_xrupt_info,
 {
 	u32 i;
 	acpi_status status;
+	struct acpi_gpe_register_info *gpe_register_info;
 
 	/* NOTE: assumes that all GPEs are currently disabled */
 
 	/* Examine each GPE Register within the block */
 
 	for (i = 0; i < gpe_block->register_count; i++) {
-		if (!gpe_block->register_info[i].enable_for_run) {
+		gpe_register_info = &gpe_block->register_info[i];
+		if (!gpe_register_info->enable_for_run) {
 			continue;
 		}
 
 		/* Enable all "runtime" GPEs in this register */
 
 		status =
-		    acpi_hw_write(gpe_block->register_info[i].enable_for_run,
-				  &gpe_block->register_info[i].enable_address);
+		    acpi_hw_gpe_enable_write(gpe_register_info->enable_for_run,
+					     gpe_register_info);
 		if (ACPI_FAILURE(status)) {
 			return (status);
 		}
@@ -392,19 +430,21 @@ acpi_hw_enable_wakeup_gpe_block(struct acpi_gpe_xrupt_info *gpe_xrupt_info,
 {
 	u32 i;
 	acpi_status status;
+	struct acpi_gpe_register_info *gpe_register_info;
 
 	/* Examine each GPE Register within the block */
 
 	for (i = 0; i < gpe_block->register_count; i++) {
-		if (!gpe_block->register_info[i].enable_for_wake) {
-			continue;
-		}
+		gpe_register_info = &gpe_block->register_info[i];
 
-		/* Enable all "wake" GPEs in this register */
+		/*
+		 * Enable all "wake" GPEs in this register and disable the
+		 * remaining ones.
+		 */
 
 		status =
-		    acpi_hw_write(gpe_block->register_info[i].enable_for_wake,
-				  &gpe_block->register_info[i].enable_address);
+		    acpi_hw_gpe_enable_write(gpe_register_info->enable_for_wake,
+					     gpe_register_info);
 		if (ACPI_FAILURE(status)) {
 			return (status);
 		}
diff --git a/drivers/acpi/acpica/tbxfroot.c b/drivers/acpi/acpica/tbxfroot.c
index 65ab8fed3d5e..43a54af2b548 100644
--- a/drivers/acpi/acpica/tbxfroot.c
+++ b/drivers/acpi/acpica/tbxfroot.c
@@ -50,6 +50,36 @@ ACPI_MODULE_NAME("tbxfroot")
 
 /*******************************************************************************
  *
+ * FUNCTION:    acpi_tb_get_rsdp_length
+ *
+ * PARAMETERS:  rsdp                - Pointer to RSDP
+ *
+ * RETURN:      Table length
+ *
+ * DESCRIPTION: Get the length of the RSDP
+ *
+ ******************************************************************************/
+u32 acpi_tb_get_rsdp_length(struct acpi_table_rsdp *rsdp)
+{
+
+	if (!ACPI_VALIDATE_RSDP_SIG(rsdp->signature)) {
+
+		/* BAD Signature */
+
+		return (0);
+	}
+
+	/* "Length" field is available if table version >= 2 */
+
+	if (rsdp->revision >= 2) {
+		return (rsdp->length);
+	} else {
+		return (ACPI_RSDP_CHECKSUM_LENGTH);
+	}
+}
+
+/*******************************************************************************
+ *
  * FUNCTION:    acpi_tb_validate_rsdp
  *
  * PARAMETERS:  rsdp                - Pointer to unvalidated RSDP
@@ -59,7 +89,8 @@ ACPI_MODULE_NAME("tbxfroot")
  * DESCRIPTION: Validate the RSDP (ptr)
  *
  ******************************************************************************/
-acpi_status acpi_tb_validate_rsdp(struct acpi_table_rsdp *rsdp)
+
+acpi_status acpi_tb_validate_rsdp(struct acpi_table_rsdp * rsdp)
 {
 
 	/*
diff --git a/drivers/acpi/acpica/utresrc.c b/drivers/acpi/acpica/utresrc.c
index 14cb6c0c8be2..bc1ff820c7dd 100644
--- a/drivers/acpi/acpica/utresrc.c
+++ b/drivers/acpi/acpica/utresrc.c
@@ -87,7 +87,9 @@ const char *acpi_gbl_io_decode[] = {
 
 const char *acpi_gbl_ll_decode[] = {
 	"ActiveHigh",
-	"ActiveLow"
+	"ActiveLow",
+	"ActiveBoth",
+	"Reserved"
 };
 
 const char *acpi_gbl_max_decode[] = {
@@ -261,7 +263,7 @@ const char *acpi_gbl_bpb_decode[] = {
 /* UART serial bus stop bits */
 
 const char *acpi_gbl_sb_decode[] = {
-	"StopBitsNone",
+	"StopBitsZero",
 	"StopBitsOne",
 	"StopBitsOnePlusHalf",
 	"StopBitsTwo"
diff --git a/drivers/acpi/acpica/utxface.c b/drivers/acpi/acpica/utxface.c
index 502a8492dc83..49c873c68756 100644
--- a/drivers/acpi/acpica/utxface.c
+++ b/drivers/acpi/acpica/utxface.c
@@ -531,7 +531,9 @@ acpi_decode_pld_buffer(u8 *in_buffer,
 	ACPI_MOVE_32_TO_32(&dword, &buffer[0]);
 	pld_info->revision = ACPI_PLD_GET_REVISION(&dword);
 	pld_info->ignore_color = ACPI_PLD_GET_IGNORE_COLOR(&dword);
-	pld_info->color = ACPI_PLD_GET_COLOR(&dword);
+	pld_info->red = ACPI_PLD_GET_RED(&dword);
+	pld_info->green = ACPI_PLD_GET_GREEN(&dword);
+	pld_info->blue = ACPI_PLD_GET_BLUE(&dword);
 
 	/* Second 32-bit DWord */
 
diff --git a/drivers/acpi/acpica/utxfinit.c b/drivers/acpi/acpica/utxfinit.c
index 13380d818462..b1fd6886e439 100644
--- a/drivers/acpi/acpica/utxfinit.c
+++ b/drivers/acpi/acpica/utxfinit.c
@@ -53,6 +53,9 @@
 #define _COMPONENT          ACPI_UTILITIES
 ACPI_MODULE_NAME("utxfinit")
 
+/* For acpi_exec only */
+void ae_do_object_overrides(void);
+
 /*******************************************************************************
  *
  * FUNCTION:    acpi_initialize_subsystem
@@ -65,6 +68,7 @@ ACPI_MODULE_NAME("utxfinit")
  *              called, so any early initialization belongs here.
  *
  ******************************************************************************/
+
 acpi_status __init acpi_initialize_subsystem(void)
 {
 	acpi_status status;
@@ -275,6 +279,13 @@ acpi_status __init acpi_initialize_objects(u32 flags)
 			return_ACPI_STATUS(status);
 		}
 	}
+#ifdef ACPI_EXEC_APP
+	/*
+	 * This call implements the "initialization file" option for acpi_exec.
+	 * This is the precise point that we want to perform the overrides.
+	 */
+	ae_do_object_overrides();
+#endif
 
 	/*
 	 * Execute any module-level code that was detected during the table load
diff --git a/drivers/acpi/apei/ghes.c b/drivers/acpi/apei/ghes.c
index fc5f780bb61d..e82d0976a5d0 100644
--- a/drivers/acpi/apei/ghes.c
+++ b/drivers/acpi/apei/ghes.c
@@ -128,7 +128,7 @@ static DEFINE_SPINLOCK(ghes_ioremap_lock_irq);
 static struct gen_pool *ghes_estatus_pool;
 static unsigned long ghes_estatus_pool_size_request;
 
-struct ghes_estatus_cache *ghes_estatus_caches[GHES_ESTATUS_CACHES_SIZE];
+static struct ghes_estatus_cache *ghes_estatus_caches[GHES_ESTATUS_CACHES_SIZE];
 static atomic_t ghes_estatus_cache_alloced;
 
 static int ghes_ioremap_init(void)
@@ -738,20 +738,6 @@ static LIST_HEAD(ghes_nmi);
 
 static int ghes_panic_timeout	__read_mostly = 30;
 
-static struct llist_node *llist_nodes_reverse(struct llist_node *llnode)
-{
-	struct llist_node *next, *tail = NULL;
-
-	while (llnode) {
-		next = llnode->next;
-		llnode->next = tail;
-		tail = llnode;
-		llnode = next;
-	}
-
-	return tail;
-}
-
 static void ghes_proc_in_irq(struct irq_work *irq_work)
 {
 	struct llist_node *llnode, *next;
@@ -765,7 +751,7 @@ static void ghes_proc_in_irq(struct irq_work *irq_work)
 	 * Because the time order of estatus in list is reversed,
 	 * revert it back to proper order.
 	 */
-	llnode = llist_nodes_reverse(llnode);
+	llnode = llist_reverse_order(llnode);
 	while (llnode) {
 		next = llnode->next;
 		estatus_node = llist_entry(llnode, struct ghes_estatus_node,
@@ -798,7 +784,7 @@ static void ghes_print_queued_estatus(void)
 	 * Because the time order of estatus in list is reversed,
 	 * revert it back to proper order.
 	 */
-	llnode = llist_nodes_reverse(llnode);
+	llnode = llist_reverse_order(llnode);
 	while (llnode) {
 		estatus_node = llist_entry(llnode, struct ghes_estatus_node,
 					   llnode);
@@ -1101,7 +1087,6 @@ static int ghes_remove(struct platform_device *ghes_dev)
 static struct platform_driver ghes_platform_driver = {
 	.driver		= {
 		.name	= "GHES",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= ghes_probe,
 	.remove		= ghes_remove,
diff --git a/drivers/acpi/battery.c b/drivers/acpi/battery.c
index 5fdfe65fe165..d98ba4355819 100644
--- a/drivers/acpi/battery.c
+++ b/drivers/acpi/battery.c
@@ -695,7 +695,7 @@ static void acpi_battery_quirks(struct acpi_battery *battery)
 	if (battery->power_unit && dmi_name_in_vendors("LENOVO")) {
 		const char *s;
 		s = dmi_get_system_info(DMI_PRODUCT_VERSION);
-		if (s && !strnicmp(s, "ThinkPad", 8)) {
+		if (s && !strncasecmp(s, "ThinkPad", 8)) {
 			dmi_walk(find_battery, battery);
 			if (test_bit(ACPI_BATTERY_QUIRK_THINKPAD_MAH,
 				     &battery->flags) &&
@@ -1180,6 +1180,10 @@ static int acpi_battery_add(struct acpi_device *device)
 
 	if (!device)
 		return -EINVAL;
+
+	if (device->dep_unmet)
+		return -EPROBE_DEFER;
+
 	battery = kzalloc(sizeof(struct acpi_battery), GFP_KERNEL);
 	if (!battery)
 		return -ENOMEM;
diff --git a/drivers/acpi/blacklist.c b/drivers/acpi/blacklist.c
index 36eb42e3b0bb..9b693d54c743 100644
--- a/drivers/acpi/blacklist.c
+++ b/drivers/acpi/blacklist.c
@@ -247,8 +247,8 @@ static struct dmi_system_id acpi_osi_dmi_table[] __initdata = {
 	},
 
 	/*
-	 * These machines will power on immediately after shutdown when
-	 * reporting the Windows 2012 OSI.
+	 * The wireless hotkey does not work on those machines when
+	 * returning true for _OSI("Windows 2012")
 	 */
 	{
 	.callback = dmi_disable_osi_win8,
@@ -258,66 +258,52 @@ static struct dmi_system_id acpi_osi_dmi_table[] __initdata = {
 		    DMI_MATCH(DMI_PRODUCT_NAME, "Inspiron 7737"),
 		},
 	},
-
-	/*
-	 * BIOS invocation of _OSI(Linux) is almost always a BIOS bug.
-	 * Linux ignores it, except for the machines enumerated below.
-	 */
-
-	/*
-	 * Lenovo has a mix of systems OSI(Linux) situations
-	 * and thus we can not wildcard the vendor.
-	 *
-	 * _OSI(Linux) helps sound
-	 * DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad R61"),
-	 * DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad T61"),
-	 * T400, T500
-	 * _OSI(Linux) has Linux specific hooks
-	 * DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad X61"),
-	 * _OSI(Linux) is a NOP:
-	 * DMI_MATCH(DMI_PRODUCT_VERSION, "3000 N100"),
-	 * DMI_MATCH(DMI_PRODUCT_VERSION, "LENOVO3000 V100"),
-	 */
 	{
-	.callback = dmi_enable_osi_linux,
-	.ident = "Lenovo ThinkPad R61",
+	.callback = dmi_disable_osi_win8,
+	.ident = "Dell Inspiron 7537",
 	.matches = {
-		     DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-		     DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad R61"),
+		    DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+		    DMI_MATCH(DMI_PRODUCT_NAME, "Inspiron 7537"),
 		},
 	},
 	{
-	.callback = dmi_enable_osi_linux,
-	.ident = "Lenovo ThinkPad T61",
+	.callback = dmi_disable_osi_win8,
+	.ident = "Dell Inspiron 5437",
 	.matches = {
-		     DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-		     DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad T61"),
+		    DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+		    DMI_MATCH(DMI_PRODUCT_NAME, "Inspiron 5437"),
 		},
 	},
 	{
-	.callback = dmi_enable_osi_linux,
-	.ident = "Lenovo ThinkPad X61",
+	.callback = dmi_disable_osi_win8,
+	.ident = "Dell Inspiron 3437",
 	.matches = {
-		     DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-		     DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad X61"),
+		    DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+		    DMI_MATCH(DMI_PRODUCT_NAME, "Inspiron 3437"),
 		},
 	},
 	{
-	.callback = dmi_enable_osi_linux,
-	.ident = "Lenovo ThinkPad T400",
+	.callback = dmi_disable_osi_win8,
+	.ident = "Dell Vostro 3446",
 	.matches = {
-		     DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-		     DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad T400"),
+		    DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+		    DMI_MATCH(DMI_PRODUCT_NAME, "Vostro 3446"),
 		},
 	},
 	{
-	.callback = dmi_enable_osi_linux,
-	.ident = "Lenovo ThinkPad T500",
+	.callback = dmi_disable_osi_win8,
+	.ident = "Dell Vostro 3546",
 	.matches = {
-		     DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-		     DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad T500"),
+		    DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+		    DMI_MATCH(DMI_PRODUCT_NAME, "Vostro 3546"),
 		},
 	},
+
+	/*
+	 * BIOS invocation of _OSI(Linux) is almost always a BIOS bug.
+	 * Linux ignores it, except for the machines enumerated below.
+	 */
+
 	/*
 	 * Without this this EEEpc exports a non working WMI interface, with
 	 * this it exports a working "good old" eeepc_laptop interface, fixing
diff --git a/drivers/acpi/device_pm.c b/drivers/acpi/device_pm.c
index 67075f800e34..c0d44d394ca3 100644
--- a/drivers/acpi/device_pm.c
+++ b/drivers/acpi/device_pm.c
@@ -201,7 +201,7 @@ int acpi_device_set_power(struct acpi_device *device, int state)
 	 * Transition Power
 	 * ----------------
 	 * In accordance with the ACPI specification first apply power (via
-	 * power resources) and then evalute _PSx.
+	 * power resources) and then evaluate _PSx.
 	 */
 	if (device->power.flags.power_resources) {
 		result = acpi_power_transition(device, state);
@@ -257,7 +257,7 @@ int acpi_bus_init_power(struct acpi_device *device)
 
 	device->power.state = ACPI_STATE_UNKNOWN;
 	if (!acpi_device_is_present(device))
-		return 0;
+		return -ENXIO;
 
 	result = acpi_device_get_power(device, &state);
 	if (result)
@@ -343,6 +343,7 @@ int acpi_device_update_power(struct acpi_device *device, int *state_p)
 
 	return 0;
 }
+EXPORT_SYMBOL_GPL(acpi_device_update_power);
 
 int acpi_bus_update_power(acpi_handle handle, int *state_p)
 {
@@ -679,19 +680,26 @@ static int acpi_device_wakeup(struct acpi_device *adev, u32 target_state,
 		if (error)
 			return error;
 
+		if (adev->wakeup.flags.enabled)
+			return 0;
+
 		res = acpi_enable_gpe(wakeup->gpe_device, wakeup->gpe_number);
-		if (ACPI_FAILURE(res)) {
+		if (ACPI_SUCCESS(res)) {
+			adev->wakeup.flags.enabled = 1;
+		} else {
 			acpi_disable_wakeup_device_power(adev);
 			return -EIO;
 		}
 	} else {
-		acpi_disable_gpe(wakeup->gpe_device, wakeup->gpe_number);
+		if (adev->wakeup.flags.enabled) {
+			acpi_disable_gpe(wakeup->gpe_device, wakeup->gpe_number);
+			adev->wakeup.flags.enabled = 0;
+		}
 		acpi_disable_wakeup_device_power(adev);
 	}
 	return 0;
 }
 
-#ifdef CONFIG_PM_RUNTIME
 /**
  * acpi_pm_device_run_wake - Enable/disable remote wakeup for given device.
  * @dev: Device to enable/disable the platform to wake up.
@@ -710,10 +718,9 @@ int acpi_pm_device_run_wake(struct device *phys_dev, bool enable)
 		return -ENODEV;
 	}
 
-	return acpi_device_wakeup(adev, enable, ACPI_STATE_S0);
+	return acpi_device_wakeup(adev, ACPI_STATE_S0, enable);
 }
 EXPORT_SYMBOL(acpi_pm_device_run_wake);
-#endif /* CONFIG_PM_RUNTIME */
 
 #ifdef CONFIG_PM_SLEEP
 /**
@@ -772,7 +779,6 @@ static int acpi_dev_pm_full_power(struct acpi_device *adev)
 		acpi_device_set_power(adev, ACPI_STATE_D0) : 0;
 }
 
-#ifdef CONFIG_PM_RUNTIME
 /**
  * acpi_dev_runtime_suspend - Put device into a low-power state using ACPI.
  * @dev: Device to put into a low-power state.
@@ -854,7 +860,6 @@ int acpi_subsys_runtime_resume(struct device *dev)
 	return ret ? ret : pm_generic_runtime_resume(dev);
 }
 EXPORT_SYMBOL_GPL(acpi_subsys_runtime_resume);
-#endif /* CONFIG_PM_RUNTIME */
 
 #ifdef CONFIG_PM_SLEEP
 /**
@@ -877,7 +882,7 @@ int acpi_dev_suspend_late(struct device *dev)
 		return 0;
 
 	target_state = acpi_target_system_state();
-	wakeup = device_may_wakeup(dev);
+	wakeup = device_may_wakeup(dev) && acpi_device_can_wakeup(adev);
 	error = acpi_device_wakeup(adev, target_state, wakeup);
 	if (wakeup && error)
 		return error;
@@ -1022,10 +1027,9 @@ EXPORT_SYMBOL_GPL(acpi_subsys_freeze);
 
 static struct dev_pm_domain acpi_general_pm_domain = {
 	.ops = {
-#ifdef CONFIG_PM_RUNTIME
+#ifdef CONFIG_PM
 		.runtime_suspend = acpi_subsys_runtime_suspend,
 		.runtime_resume = acpi_subsys_runtime_resume,
-#endif
 #ifdef CONFIG_PM_SLEEP
 		.prepare = acpi_subsys_prepare,
 		.complete = acpi_subsys_complete,
@@ -1037,10 +1041,45 @@ static struct dev_pm_domain acpi_general_pm_domain = {
 		.poweroff_late = acpi_subsys_suspend_late,
 		.restore_early = acpi_subsys_resume_early,
 #endif
+#endif
 	},
 };
 
 /**
+ * acpi_dev_pm_detach - Remove ACPI power management from the device.
+ * @dev: Device to take care of.
+ * @power_off: Whether or not to try to remove power from the device.
+ *
+ * Remove the device from the general ACPI PM domain and remove its wakeup
+ * notifier.  If @power_off is set, additionally remove power from the device if
+ * possible.
+ *
+ * Callers must ensure proper synchronization of this function with power
+ * management callbacks.
+ */
+static void acpi_dev_pm_detach(struct device *dev, bool power_off)
+{
+	struct acpi_device *adev = ACPI_COMPANION(dev);
+
+	if (adev && dev->pm_domain == &acpi_general_pm_domain) {
+		dev->pm_domain = NULL;
+		acpi_remove_pm_notifier(adev);
+		if (power_off) {
+			/*
+			 * If the device's PM QoS resume latency limit or flags
+			 * have been exposed to user space, they have to be
+			 * hidden at this point, so that they don't affect the
+			 * choice of the low-power state to put the device into.
+			 */
+			dev_pm_qos_hide_latency_limit(dev);
+			dev_pm_qos_hide_flags(dev);
+			acpi_device_wakeup(adev, ACPI_STATE_S0, false);
+			acpi_dev_pm_low_power(dev, adev, ACPI_STATE_S0);
+		}
+	}
+}
+
+/**
  * acpi_dev_pm_attach - Prepare device for ACPI power management.
  * @dev: Device to prepare.
  * @power_on: Whether or not to power on the device.
@@ -1072,42 +1111,9 @@ int acpi_dev_pm_attach(struct device *dev, bool power_on)
 		acpi_dev_pm_full_power(adev);
 		acpi_device_wakeup(adev, ACPI_STATE_S0, false);
 	}
+
+	dev->pm_domain->detach = acpi_dev_pm_detach;
 	return 0;
 }
 EXPORT_SYMBOL_GPL(acpi_dev_pm_attach);
-
-/**
- * acpi_dev_pm_detach - Remove ACPI power management from the device.
- * @dev: Device to take care of.
- * @power_off: Whether or not to try to remove power from the device.
- *
- * Remove the device from the general ACPI PM domain and remove its wakeup
- * notifier.  If @power_off is set, additionally remove power from the device if
- * possible.
- *
- * Callers must ensure proper synchronization of this function with power
- * management callbacks.
- */
-void acpi_dev_pm_detach(struct device *dev, bool power_off)
-{
-	struct acpi_device *adev = ACPI_COMPANION(dev);
-
-	if (adev && dev->pm_domain == &acpi_general_pm_domain) {
-		dev->pm_domain = NULL;
-		acpi_remove_pm_notifier(adev);
-		if (power_off) {
-			/*
-			 * If the device's PM QoS resume latency limit or flags
-			 * have been exposed to user space, they have to be
-			 * hidden at this point, so that they don't affect the
-			 * choice of the low-power state to put the device into.
-			 */
-			dev_pm_qos_hide_latency_limit(dev);
-			dev_pm_qos_hide_flags(dev);
-			acpi_device_wakeup(adev, ACPI_STATE_S0, false);
-			acpi_dev_pm_low_power(dev, adev, ACPI_STATE_S0);
-		}
-	}
-}
-EXPORT_SYMBOL_GPL(acpi_dev_pm_detach);
 #endif /* CONFIG_PM */
diff --git a/drivers/acpi/ec.c b/drivers/acpi/ec.c
index cb6066c809ea..1b5853f384e2 100644
--- a/drivers/acpi/ec.c
+++ b/drivers/acpi/ec.c
@@ -126,14 +126,16 @@ static int EC_FLAGS_MSI; /* Out-of-spec MSI controller */
 static int EC_FLAGS_VALIDATE_ECDT; /* ASUStec ECDTs need to be validated */
 static int EC_FLAGS_SKIP_DSDT_SCAN; /* Not all BIOS survive early DSDT scan */
 static int EC_FLAGS_CLEAR_ON_RESUME; /* Needs acpi_ec_clear() on boot/resume */
+static int EC_FLAGS_QUERY_HANDSHAKE; /* Needs QR_EC issued when SCI_EVT set */
 
 /* --------------------------------------------------------------------------
-                             Transaction Management
-   -------------------------------------------------------------------------- */
+ *                           Transaction Management
+ * -------------------------------------------------------------------------- */
 
 static inline u8 acpi_ec_read_status(struct acpi_ec *ec)
 {
 	u8 x = inb(ec->command_addr);
+
 	pr_debug("EC_SC(R) = 0x%2.2x "
 		 "SCI_EVT=%d BURST=%d CMD=%d IBF=%d OBF=%d\n",
 		 x,
@@ -148,6 +150,7 @@ static inline u8 acpi_ec_read_status(struct acpi_ec *ec)
 static inline u8 acpi_ec_read_data(struct acpi_ec *ec)
 {
 	u8 x = inb(ec->data_addr);
+
 	pr_debug("EC_DATA(R) = 0x%2.2x\n", x);
 	return x;
 }
@@ -164,10 +167,32 @@ static inline void acpi_ec_write_data(struct acpi_ec *ec, u8 data)
 	outb(data, ec->data_addr);
 }
 
+#ifdef DEBUG
+static const char *acpi_ec_cmd_string(u8 cmd)
+{
+	switch (cmd) {
+	case 0x80:
+		return "RD_EC";
+	case 0x81:
+		return "WR_EC";
+	case 0x82:
+		return "BE_EC";
+	case 0x83:
+		return "BD_EC";
+	case 0x84:
+		return "QR_EC";
+	}
+	return "UNKNOWN";
+}
+#else
+#define acpi_ec_cmd_string(cmd)		"UNDEF"
+#endif
+
 static int ec_transaction_completed(struct acpi_ec *ec)
 {
 	unsigned long flags;
 	int ret = 0;
+
 	spin_lock_irqsave(&ec->lock, flags);
 	if (ec->curr && (ec->curr->flags & ACPI_EC_COMMAND_COMPLETE))
 		ret = 1;
@@ -181,7 +206,8 @@ static bool advance_transaction(struct acpi_ec *ec)
 	u8 status;
 	bool wakeup = false;
 
-	pr_debug("===== %s =====\n", in_interrupt() ? "IRQ" : "TASK");
+	pr_debug("===== %s (%d) =====\n",
+		 in_interrupt() ? "IRQ" : "TASK", smp_processor_id());
 	status = acpi_ec_read_status(ec);
 	t = ec->curr;
 	if (!t)
@@ -198,7 +224,8 @@ static bool advance_transaction(struct acpi_ec *ec)
 				if (t->rlen == t->ri) {
 					t->flags |= ACPI_EC_COMMAND_COMPLETE;
 					if (t->command == ACPI_EC_COMMAND_QUERY)
-						pr_debug("hardware QR_EC completion\n");
+						pr_debug("***** Command(%s) hardware completion *****\n",
+							 acpi_ec_cmd_string(t->command));
 					wakeup = true;
 				}
 			} else
@@ -210,18 +237,14 @@ static bool advance_transaction(struct acpi_ec *ec)
 		}
 		return wakeup;
 	} else {
-		/*
-		 * There is firmware refusing to respond QR_EC when SCI_EVT
-		 * is not set, for which case, we complete the QR_EC
-		 * without issuing it to the firmware.
-		 * https://bugzilla.kernel.org/show_bug.cgi?id=86211
-		 */
-		if (!(status & ACPI_EC_FLAG_SCI) &&
+		if (EC_FLAGS_QUERY_HANDSHAKE &&
+		    !(status & ACPI_EC_FLAG_SCI) &&
 		    (t->command == ACPI_EC_COMMAND_QUERY)) {
 			t->flags |= ACPI_EC_COMMAND_POLL;
 			t->rdata[t->ri++] = 0x00;
 			t->flags |= ACPI_EC_COMMAND_COMPLETE;
-			pr_debug("software QR_EC completion\n");
+			pr_debug("***** Command(%s) software completion *****\n",
+				 acpi_ec_cmd_string(t->command));
 			wakeup = true;
 		} else if ((status & ACPI_EC_FLAG_IBF) == 0) {
 			acpi_ec_write_cmd(ec, t->command);
@@ -264,6 +287,7 @@ static int ec_poll(struct acpi_ec *ec)
 {
 	unsigned long flags;
 	int repeat = 5; /* number of command restarts */
+
 	while (repeat--) {
 		unsigned long delay = jiffies +
 			msecs_to_jiffies(ec_delay);
@@ -296,18 +320,25 @@ static int acpi_ec_transaction_unlocked(struct acpi_ec *ec,
 {
 	unsigned long tmp;
 	int ret = 0;
+
 	if (EC_FLAGS_MSI)
 		udelay(ACPI_EC_MSI_UDELAY);
 	/* start transaction */
 	spin_lock_irqsave(&ec->lock, tmp);
 	/* following two actions should be kept atomic */
 	ec->curr = t;
+	pr_debug("***** Command(%s) started *****\n",
+		 acpi_ec_cmd_string(t->command));
 	start_transaction(ec);
+	if (ec->curr->command == ACPI_EC_COMMAND_QUERY) {
+		clear_bit(EC_FLAGS_QUERY_PENDING, &ec->flags);
+		pr_debug("***** Event stopped *****\n");
+	}
 	spin_unlock_irqrestore(&ec->lock, tmp);
 	ret = ec_poll(ec);
 	spin_lock_irqsave(&ec->lock, tmp);
-	if (ec->curr->command == ACPI_EC_COMMAND_QUERY)
-		clear_bit(EC_FLAGS_QUERY_PENDING, &ec->flags);
+	pr_debug("***** Command(%s) stopped *****\n",
+		 acpi_ec_cmd_string(t->command));
 	ec->curr = NULL;
 	spin_unlock_irqrestore(&ec->lock, tmp);
 	return ret;
@@ -317,6 +348,7 @@ static int acpi_ec_transaction(struct acpi_ec *ec, struct transaction *t)
 {
 	int status;
 	u32 glk;
+
 	if (!ec || (!t) || (t->wlen && !t->wdata) || (t->rlen && !t->rdata))
 		return -EINVAL;
 	if (t->rdata)
@@ -333,8 +365,6 @@ static int acpi_ec_transaction(struct acpi_ec *ec, struct transaction *t)
 			goto unlock;
 		}
 	}
-	pr_debug("transaction start (cmd=0x%02x, addr=0x%02x)\n",
-			t->command, t->wdata ? t->wdata[0] : 0);
 	/* disable GPE during transaction if storm is detected */
 	if (test_bit(EC_FLAGS_GPE_STORM, &ec->flags)) {
 		/* It has to be disabled, so that it doesn't trigger. */
@@ -355,7 +385,6 @@ static int acpi_ec_transaction(struct acpi_ec *ec, struct transaction *t)
 			t->irq_count);
 		set_bit(EC_FLAGS_GPE_STORM, &ec->flags);
 	}
-	pr_debug("transaction end\n");
 	if (ec->global_lock)
 		acpi_release_global_lock(glk);
 unlock:
@@ -383,7 +412,7 @@ static int acpi_ec_burst_disable(struct acpi_ec *ec)
 				acpi_ec_transaction(ec, &t) : 0;
 }
 
-static int acpi_ec_read(struct acpi_ec *ec, u8 address, u8 * data)
+static int acpi_ec_read(struct acpi_ec *ec, u8 address, u8 *data)
 {
 	int result;
 	u8 d;
@@ -419,10 +448,9 @@ int ec_read(u8 addr, u8 *val)
 	if (!err) {
 		*val = temp_data;
 		return 0;
-	} else
-		return err;
+	}
+	return err;
 }
-
 EXPORT_SYMBOL(ec_read);
 
 int ec_write(u8 addr, u8 val)
@@ -436,22 +464,21 @@ int ec_write(u8 addr, u8 val)
 
 	return err;
 }
-
 EXPORT_SYMBOL(ec_write);
 
 int ec_transaction(u8 command,
-		   const u8 * wdata, unsigned wdata_len,
-		   u8 * rdata, unsigned rdata_len)
+		   const u8 *wdata, unsigned wdata_len,
+		   u8 *rdata, unsigned rdata_len)
 {
 	struct transaction t = {.command = command,
 				.wdata = wdata, .rdata = rdata,
 				.wlen = wdata_len, .rlen = rdata_len};
+
 	if (!first_ec)
 		return -ENODEV;
 
 	return acpi_ec_transaction(first_ec, &t);
 }
-
 EXPORT_SYMBOL(ec_transaction);
 
 /* Get the handle to the EC device */
@@ -461,7 +488,6 @@ acpi_handle ec_get_handle(void)
 		return NULL;
 	return first_ec->handle;
 }
-
 EXPORT_SYMBOL(ec_get_handle);
 
 /*
@@ -525,13 +551,14 @@ void acpi_ec_unblock_transactions_early(void)
 		clear_bit(EC_FLAGS_BLOCKED, &first_ec->flags);
 }
 
-static int acpi_ec_query_unlocked(struct acpi_ec *ec, u8 * data)
+static int acpi_ec_query_unlocked(struct acpi_ec *ec, u8 *data)
 {
 	int result;
 	u8 d;
 	struct transaction t = {.command = ACPI_EC_COMMAND_QUERY,
 				.wdata = NULL, .rdata = &d,
 				.wlen = 0, .rlen = 1};
+
 	if (!ec || !data)
 		return -EINVAL;
 	/*
@@ -557,6 +584,7 @@ int acpi_ec_add_query_handler(struct acpi_ec *ec, u8 query_bit,
 {
 	struct acpi_ec_query_handler *handler =
 	    kzalloc(sizeof(struct acpi_ec_query_handler), GFP_KERNEL);
+
 	if (!handler)
 		return -ENOMEM;
 
@@ -569,12 +597,12 @@ int acpi_ec_add_query_handler(struct acpi_ec *ec, u8 query_bit,
 	mutex_unlock(&ec->mutex);
 	return 0;
 }
-
 EXPORT_SYMBOL_GPL(acpi_ec_add_query_handler);
 
 void acpi_ec_remove_query_handler(struct acpi_ec *ec, u8 query_bit)
 {
 	struct acpi_ec_query_handler *handler, *tmp;
+
 	mutex_lock(&ec->mutex);
 	list_for_each_entry_safe(handler, tmp, &ec->list, node) {
 		if (query_bit == handler->query_bit) {
@@ -584,20 +612,20 @@ void acpi_ec_remove_query_handler(struct acpi_ec *ec, u8 query_bit)
 	}
 	mutex_unlock(&ec->mutex);
 }
-
 EXPORT_SYMBOL_GPL(acpi_ec_remove_query_handler);
 
 static void acpi_ec_run(void *cxt)
 {
 	struct acpi_ec_query_handler *handler = cxt;
+
 	if (!handler)
 		return;
-	pr_debug("start query execution\n");
+	pr_debug("##### Query(0x%02x) started #####\n", handler->query_bit);
 	if (handler->func)
 		handler->func(handler->data);
 	else if (handler->handle)
 		acpi_evaluate_object(handler->handle, NULL, NULL, NULL);
-	pr_debug("stop query execution\n");
+	pr_debug("##### Query(0x%02x) stopped #####\n", handler->query_bit);
 	kfree(handler);
 }
 
@@ -620,8 +648,8 @@ static int acpi_ec_sync_query(struct acpi_ec *ec, u8 *data)
 			if (!copy)
 				return -ENOMEM;
 			memcpy(copy, handler, sizeof(*copy));
-			pr_debug("push query execution (0x%2x) on queue\n",
-				value);
+			pr_debug("##### Query(0x%02x) scheduled #####\n",
+				 handler->query_bit);
 			return acpi_os_execute((copy->func) ?
 				OSL_NOTIFY_HANDLER : OSL_GPE_HANDLER,
 				acpi_ec_run, copy);
@@ -633,6 +661,7 @@ static int acpi_ec_sync_query(struct acpi_ec *ec, u8 *data)
 static void acpi_ec_gpe_query(void *ec_cxt)
 {
 	struct acpi_ec *ec = ec_cxt;
+
 	if (!ec)
 		return;
 	mutex_lock(&ec->mutex);
@@ -644,7 +673,7 @@ static int ec_check_sci(struct acpi_ec *ec, u8 state)
 {
 	if (state & ACPI_EC_FLAG_SCI) {
 		if (!test_and_set_bit(EC_FLAGS_QUERY_PENDING, &ec->flags)) {
-			pr_debug("push gpe query to the queue\n");
+			pr_debug("***** Event started *****\n");
 			return acpi_os_execute(OSL_NOTIFY_HANDLER,
 				acpi_ec_gpe_query, ec);
 		}
@@ -667,8 +696,8 @@ static u32 acpi_ec_gpe_handler(acpi_handle gpe_device,
 }
 
 /* --------------------------------------------------------------------------
-                             Address Space Management
-   -------------------------------------------------------------------------- */
+ *                           Address Space Management
+ * -------------------------------------------------------------------------- */
 
 static acpi_status
 acpi_ec_space_handler(u32 function, acpi_physical_address address,
@@ -699,27 +728,26 @@ acpi_ec_space_handler(u32 function, acpi_physical_address address,
 	switch (result) {
 	case -EINVAL:
 		return AE_BAD_PARAMETER;
-		break;
 	case -ENODEV:
 		return AE_NOT_FOUND;
-		break;
 	case -ETIME:
 		return AE_TIME;
-		break;
 	default:
 		return AE_OK;
 	}
 }
 
 /* --------------------------------------------------------------------------
-                               Driver Interface
-   -------------------------------------------------------------------------- */
+ *                             Driver Interface
+ * -------------------------------------------------------------------------- */
+
 static acpi_status
 ec_parse_io_ports(struct acpi_resource *resource, void *context);
 
 static struct acpi_ec *make_acpi_ec(void)
 {
 	struct acpi_ec *ec = kzalloc(sizeof(struct acpi_ec), GFP_KERNEL);
+
 	if (!ec)
 		return NULL;
 	ec->flags = 1 << EC_FLAGS_QUERY_PENDING;
@@ -742,9 +770,8 @@ acpi_ec_register_query_methods(acpi_handle handle, u32 level,
 
 	status = acpi_get_name(handle, ACPI_SINGLE_NAME, &buffer);
 
-	if (ACPI_SUCCESS(status) && sscanf(node_name, "_Q%x", &value) == 1) {
+	if (ACPI_SUCCESS(status) && sscanf(node_name, "_Q%x", &value) == 1)
 		acpi_ec_add_query_handler(ec, value, handle, NULL, NULL);
-	}
 	return AE_OK;
 }
 
@@ -753,7 +780,6 @@ ec_parse_device(acpi_handle handle, u32 Level, void *context, void **retval)
 {
 	acpi_status status;
 	unsigned long long tmp = 0;
-
 	struct acpi_ec *ec = context;
 
 	/* clear addr values, ec_parse_io_ports depend on it */
@@ -781,6 +807,7 @@ ec_parse_device(acpi_handle handle, u32 Level, void *context, void **retval)
 static int ec_install_handlers(struct acpi_ec *ec)
 {
 	acpi_status status;
+
 	if (test_bit(EC_FLAGS_HANDLERS_INSTALLED, &ec->flags))
 		return 0;
 	status = acpi_install_gpe_handler(NULL, ec->gpe,
@@ -817,6 +844,8 @@ static int ec_install_handlers(struct acpi_ec *ec)
 
 static void ec_remove_handlers(struct acpi_ec *ec)
 {
+	if (!test_bit(EC_FLAGS_HANDLERS_INSTALLED, &ec->flags))
+		return;
 	acpi_disable_gpe(NULL, ec->gpe);
 	if (ACPI_FAILURE(acpi_remove_address_space_handler(ec->handle,
 				ACPI_ADR_SPACE_EC, &acpi_ec_space_handler)))
@@ -981,6 +1010,18 @@ static int ec_enlarge_storm_threshold(const struct dmi_system_id *id)
 }
 
 /*
+ * Acer EC firmware refuses to respond QR_EC when SCI_EVT is not set, for
+ * which case, we complete the QR_EC without issuing it to the firmware.
+ * https://bugzilla.kernel.org/show_bug.cgi?id=86211
+ */
+static int ec_flag_query_handshake(const struct dmi_system_id *id)
+{
+	pr_debug("Detected the EC firmware requiring QR_EC issued when SCI_EVT set\n");
+	EC_FLAGS_QUERY_HANDSHAKE = 1;
+	return 0;
+}
+
+/*
  * On some hardware it is necessary to clear events accumulated by the EC during
  * sleep. These ECs stop reporting GPEs until they are manually polled, if too
  * many events are accumulated. (e.g. Samsung Series 5/9 notebooks)
@@ -1054,6 +1095,9 @@ static struct dmi_system_id ec_dmi_table[] __initdata = {
 	{
 	ec_clear_on_resume, "Samsung hardware", {
 	DMI_MATCH(DMI_SYS_VENDOR, "SAMSUNG ELECTRONICS CO., LTD.")}, NULL},
+	{
+	ec_flag_query_handshake, "Acer hardware", {
+	DMI_MATCH(DMI_SYS_VENDOR, "Acer"), }, NULL},
 	{},
 };
 
@@ -1078,7 +1122,8 @@ int __init acpi_ec_ecdt_probe(void)
 		boot_ec->data_addr = ecdt_ptr->data.address;
 		boot_ec->gpe = ecdt_ptr->gpe;
 		boot_ec->handle = ACPI_ROOT_OBJECT;
-		acpi_get_handle(ACPI_ROOT_OBJECT, ecdt_ptr->id, &boot_ec->handle);
+		acpi_get_handle(ACPI_ROOT_OBJECT, ecdt_ptr->id,
+				&boot_ec->handle);
 		/* Don't trust ECDT, which comes from ASUSTek */
 		if (!EC_FLAGS_VALIDATE_ECDT)
 			goto install;
@@ -1162,6 +1207,5 @@ static void __exit acpi_ec_exit(void)
 {
 
 	acpi_bus_unregister_driver(&acpi_ec_driver);
-	return;
 }
 #endif	/* 0 */
diff --git a/drivers/acpi/fan.c b/drivers/acpi/fan.c
index 8acf53e62966..7a36f02598a6 100644
--- a/drivers/acpi/fan.c
+++ b/drivers/acpi/fan.c
@@ -27,27 +27,22 @@
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/types.h>
-#include <asm/uaccess.h>
+#include <linux/uaccess.h>
 #include <linux/thermal.h>
 #include <linux/acpi.h>
-
-#define PREFIX "ACPI: "
-
-#define ACPI_FAN_CLASS			"fan"
-#define ACPI_FAN_FILE_STATE		"state"
-
-#define _COMPONENT		ACPI_FAN_COMPONENT
-ACPI_MODULE_NAME("fan");
+#include <linux/platform_device.h>
+#include <linux/sort.h>
 
 MODULE_AUTHOR("Paul Diefenbaugh");
 MODULE_DESCRIPTION("ACPI Fan Driver");
 MODULE_LICENSE("GPL");
 
-static int acpi_fan_add(struct acpi_device *device);
-static int acpi_fan_remove(struct acpi_device *device);
+static int acpi_fan_probe(struct platform_device *pdev);
+static int acpi_fan_remove(struct platform_device *pdev);
 
 static const struct acpi_device_id fan_device_ids[] = {
 	{"PNP0C0B", 0},
+	{"INT3404", 0},
 	{"", 0},
 };
 MODULE_DEVICE_TABLE(acpi, fan_device_ids);
@@ -66,37 +61,100 @@ static struct dev_pm_ops acpi_fan_pm = {
 #define FAN_PM_OPS_PTR NULL
 #endif
 
-static struct acpi_driver acpi_fan_driver = {
-	.name = "fan",
-	.class = ACPI_FAN_CLASS,
-	.ids = fan_device_ids,
-	.ops = {
-		.add = acpi_fan_add,
-		.remove = acpi_fan_remove,
-		},
-	.drv.pm = FAN_PM_OPS_PTR,
+struct acpi_fan_fps {
+	u64 control;
+	u64 trip_point;
+	u64 speed;
+	u64 noise_level;
+	u64 power;
+};
+
+struct acpi_fan_fif {
+	u64 revision;
+	u64 fine_grain_ctrl;
+	u64 step_size;
+	u64 low_speed_notification;
+};
+
+struct acpi_fan {
+	bool acpi4;
+	struct acpi_fan_fif fif;
+	struct acpi_fan_fps *fps;
+	int fps_count;
+	struct thermal_cooling_device *cdev;
+};
+
+static struct platform_driver acpi_fan_driver = {
+	.probe = acpi_fan_probe,
+	.remove = acpi_fan_remove,
+	.driver = {
+		.name = "acpi-fan",
+		.acpi_match_table = fan_device_ids,
+		.pm = FAN_PM_OPS_PTR,
+	},
 };
 
 /* thermal cooling device callbacks */
 static int fan_get_max_state(struct thermal_cooling_device *cdev, unsigned long
 			     *state)
 {
-	/* ACPI fan device only support two states: ON/OFF */
-	*state = 1;
+	struct acpi_device *device = cdev->devdata;
+	struct acpi_fan *fan = acpi_driver_data(device);
+
+	if (fan->acpi4)
+		*state = fan->fps_count - 1;
+	else
+		*state = 1;
 	return 0;
 }
 
-static int fan_get_cur_state(struct thermal_cooling_device *cdev, unsigned long
-			     *state)
+static int fan_get_state_acpi4(struct acpi_device *device, unsigned long *state)
+{
+	struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL };
+	struct acpi_fan *fan = acpi_driver_data(device);
+	union acpi_object *obj;
+	acpi_status status;
+	int control, i;
+
+	status = acpi_evaluate_object(device->handle, "_FST", NULL, &buffer);
+	if (ACPI_FAILURE(status)) {
+		dev_err(&device->dev, "Get fan state failed\n");
+		return status;
+	}
+
+	obj = buffer.pointer;
+	if (!obj || obj->type != ACPI_TYPE_PACKAGE ||
+	    obj->package.count != 3 ||
+	    obj->package.elements[1].type != ACPI_TYPE_INTEGER) {
+		dev_err(&device->dev, "Invalid _FST data\n");
+		status = -EINVAL;
+		goto err;
+	}
+
+	control = obj->package.elements[1].integer.value;
+	for (i = 0; i < fan->fps_count; i++) {
+		if (control == fan->fps[i].control)
+			break;
+	}
+	if (i == fan->fps_count) {
+		dev_dbg(&device->dev, "Invalid control value returned\n");
+		status = -EINVAL;
+		goto err;
+	}
+
+	*state = i;
+
+err:
+	kfree(obj);
+	return status;
+}
+
+static int fan_get_state(struct acpi_device *device, unsigned long *state)
 {
-	struct acpi_device *device = cdev->devdata;
 	int result;
 	int acpi_state = ACPI_STATE_D0;
 
-	if (!device)
-		return -EINVAL;
-
-	result = acpi_bus_update_power(device->handle, &acpi_state);
+	result = acpi_device_update_power(device, &acpi_state);
 	if (result)
 		return result;
 
@@ -105,21 +163,57 @@ static int fan_get_cur_state(struct thermal_cooling_device *cdev, unsigned long
 	return 0;
 }
 
-static int
-fan_set_cur_state(struct thermal_cooling_device *cdev, unsigned long state)
+static int fan_get_cur_state(struct thermal_cooling_device *cdev, unsigned long
+			     *state)
 {
 	struct acpi_device *device = cdev->devdata;
-	int result;
+	struct acpi_fan *fan = acpi_driver_data(device);
 
-	if (!device || (state != 0 && state != 1))
+	if (fan->acpi4)
+		return fan_get_state_acpi4(device, state);
+	else
+		return fan_get_state(device, state);
+}
+
+static int fan_set_state(struct acpi_device *device, unsigned long state)
+{
+	if (state != 0 && state != 1)
 		return -EINVAL;
 
-	result = acpi_bus_set_power(device->handle,
-				state ? ACPI_STATE_D0 : ACPI_STATE_D3_COLD);
+	return acpi_device_set_power(device,
+				     state ? ACPI_STATE_D0 : ACPI_STATE_D3_COLD);
+}
 
-	return result;
+static int fan_set_state_acpi4(struct acpi_device *device, unsigned long state)
+{
+	struct acpi_fan *fan = acpi_driver_data(device);
+	acpi_status status;
+
+	if (state >= fan->fps_count)
+		return -EINVAL;
+
+	status = acpi_execute_simple_method(device->handle, "_FSL",
+					    fan->fps[state].control);
+	if (ACPI_FAILURE(status)) {
+		dev_dbg(&device->dev, "Failed to set state by _FSL\n");
+		return status;
+	}
+
+	return 0;
 }
 
+static int
+fan_set_cur_state(struct thermal_cooling_device *cdev, unsigned long state)
+{
+	struct acpi_device *device = cdev->devdata;
+	struct acpi_fan *fan = acpi_driver_data(device);
+
+	if (fan->acpi4)
+		return fan_set_state_acpi4(device, state);
+	else
+		return fan_set_state(device, state);
+ }
+
 static const struct thermal_cooling_device_ops fan_cooling_ops = {
 	.get_max_state = fan_get_max_state,
 	.get_cur_state = fan_get_cur_state,
@@ -127,72 +221,170 @@ static const struct thermal_cooling_device_ops fan_cooling_ops = {
 };
 
 /* --------------------------------------------------------------------------
-                                 Driver Interface
-   -------------------------------------------------------------------------- */
+ *                               Driver Interface
+ * --------------------------------------------------------------------------
+*/
 
-static int acpi_fan_add(struct acpi_device *device)
+static bool acpi_fan_is_acpi4(struct acpi_device *device)
 {
-	int result = 0;
-	struct thermal_cooling_device *cdev;
+	return acpi_has_method(device->handle, "_FIF") &&
+	       acpi_has_method(device->handle, "_FPS") &&
+	       acpi_has_method(device->handle, "_FSL") &&
+	       acpi_has_method(device->handle, "_FST");
+}
 
-	if (!device)
-		return -EINVAL;
+static int acpi_fan_get_fif(struct acpi_device *device)
+{
+	struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL };
+	struct acpi_fan *fan = acpi_driver_data(device);
+	struct acpi_buffer format = { sizeof("NNNN"), "NNNN" };
+	struct acpi_buffer fif = { sizeof(fan->fif), &fan->fif };
+	union acpi_object *obj;
+	acpi_status status;
+
+	status = acpi_evaluate_object(device->handle, "_FIF", NULL, &buffer);
+	if (ACPI_FAILURE(status))
+		return status;
+
+	obj = buffer.pointer;
+	if (!obj || obj->type != ACPI_TYPE_PACKAGE) {
+		dev_err(&device->dev, "Invalid _FIF data\n");
+		status = -EINVAL;
+		goto err;
+	}
 
-	strcpy(acpi_device_name(device), "Fan");
-	strcpy(acpi_device_class(device), ACPI_FAN_CLASS);
+	status = acpi_extract_package(obj, &format, &fif);
+	if (ACPI_FAILURE(status)) {
+		dev_err(&device->dev, "Invalid _FIF element\n");
+		status = -EINVAL;
+	}
 
-	result = acpi_bus_update_power(device->handle, NULL);
-	if (result) {
-		printk(KERN_ERR PREFIX "Setting initial power state\n");
-		goto end;
+err:
+	kfree(obj);
+	return status;
+}
+
+static int acpi_fan_speed_cmp(const void *a, const void *b)
+{
+	const struct acpi_fan_fps *fps1 = a;
+	const struct acpi_fan_fps *fps2 = b;
+	return fps1->speed - fps2->speed;
+}
+
+static int acpi_fan_get_fps(struct acpi_device *device)
+{
+	struct acpi_fan *fan = acpi_driver_data(device);
+	struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL };
+	union acpi_object *obj;
+	acpi_status status;
+	int i;
+
+	status = acpi_evaluate_object(device->handle, "_FPS", NULL, &buffer);
+	if (ACPI_FAILURE(status))
+		return status;
+
+	obj = buffer.pointer;
+	if (!obj || obj->type != ACPI_TYPE_PACKAGE || obj->package.count < 2) {
+		dev_err(&device->dev, "Invalid _FPS data\n");
+		status = -EINVAL;
+		goto err;
 	}
 
-	cdev = thermal_cooling_device_register("Fan", device,
+	fan->fps_count = obj->package.count - 1; /* minus revision field */
+	fan->fps = devm_kzalloc(&device->dev,
+				fan->fps_count * sizeof(struct acpi_fan_fps),
+				GFP_KERNEL);
+	if (!fan->fps) {
+		dev_err(&device->dev, "Not enough memory\n");
+		status = -ENOMEM;
+		goto err;
+	}
+	for (i = 0; i < fan->fps_count; i++) {
+		struct acpi_buffer format = { sizeof("NNNNN"), "NNNNN" };
+		struct acpi_buffer fps = { sizeof(fan->fps[i]), &fan->fps[i] };
+		status = acpi_extract_package(&obj->package.elements[i + 1],
+					      &format, &fps);
+		if (ACPI_FAILURE(status)) {
+			dev_err(&device->dev, "Invalid _FPS element\n");
+			break;
+		}
+	}
+
+	/* sort the state array according to fan speed in increase order */
+	sort(fan->fps, fan->fps_count, sizeof(*fan->fps),
+	     acpi_fan_speed_cmp, NULL);
+
+err:
+	kfree(obj);
+	return status;
+}
+
+static int acpi_fan_probe(struct platform_device *pdev)
+{
+	int result = 0;
+	struct thermal_cooling_device *cdev;
+	struct acpi_fan *fan;
+	struct acpi_device *device = ACPI_COMPANION(&pdev->dev);
+	char *name;
+
+	fan = devm_kzalloc(&pdev->dev, sizeof(*fan), GFP_KERNEL);
+	if (!fan) {
+		dev_err(&device->dev, "No memory for fan\n");
+		return -ENOMEM;
+	}
+	device->driver_data = fan;
+	platform_set_drvdata(pdev, fan);
+
+	if (acpi_fan_is_acpi4(device)) {
+		if (acpi_fan_get_fif(device) || acpi_fan_get_fps(device))
+			goto end;
+		fan->acpi4 = true;
+	} else {
+		result = acpi_device_update_power(device, NULL);
+		if (result) {
+			dev_err(&device->dev, "Setting initial power state\n");
+			goto end;
+		}
+	}
+
+	if (!strncmp(pdev->name, "PNP0C0B", strlen("PNP0C0B")))
+		name = "Fan";
+	else
+		name = acpi_device_bid(device);
+
+	cdev = thermal_cooling_device_register(name, device,
 						&fan_cooling_ops);
 	if (IS_ERR(cdev)) {
 		result = PTR_ERR(cdev);
 		goto end;
 	}
 
-	dev_dbg(&device->dev, "registered as cooling_device%d\n", cdev->id);
+	dev_dbg(&pdev->dev, "registered as cooling_device%d\n", cdev->id);
 
-	device->driver_data = cdev;
-	result = sysfs_create_link(&device->dev.kobj,
+	fan->cdev = cdev;
+	result = sysfs_create_link(&pdev->dev.kobj,
 				   &cdev->device.kobj,
 				   "thermal_cooling");
 	if (result)
-		dev_err(&device->dev, "Failed to create sysfs link "
-			"'thermal_cooling'\n");
+		dev_err(&pdev->dev, "Failed to create sysfs link 'thermal_cooling'\n");
 
 	result = sysfs_create_link(&cdev->device.kobj,
-				   &device->dev.kobj,
+				   &pdev->dev.kobj,
 				   "device");
 	if (result)
-		dev_err(&device->dev, "Failed to create sysfs link "
-			"'device'\n");
-
-	printk(KERN_INFO PREFIX "%s [%s] (%s)\n",
-	       acpi_device_name(device), acpi_device_bid(device),
-	       !device->power.state ? "on" : "off");
+		dev_err(&pdev->dev, "Failed to create sysfs link 'device'\n");
 
 end:
 	return result;
 }
 
-static int acpi_fan_remove(struct acpi_device *device)
+static int acpi_fan_remove(struct platform_device *pdev)
 {
-	struct thermal_cooling_device *cdev;
-
-	if (!device)
-		return -EINVAL;
-
-	cdev =  acpi_driver_data(device);
-	if (!cdev)
-		return -EINVAL;
+	struct acpi_fan *fan = platform_get_drvdata(pdev);
 
-	sysfs_remove_link(&device->dev.kobj, "thermal_cooling");
-	sysfs_remove_link(&cdev->device.kobj, "device");
-	thermal_cooling_device_unregister(cdev);
+	sysfs_remove_link(&pdev->dev.kobj, "thermal_cooling");
+	sysfs_remove_link(&fan->cdev->device.kobj, "device");
+	thermal_cooling_device_unregister(fan->cdev);
 
 	return 0;
 }
@@ -200,10 +392,11 @@ static int acpi_fan_remove(struct acpi_device *device)
 #ifdef CONFIG_PM_SLEEP
 static int acpi_fan_suspend(struct device *dev)
 {
-	if (!dev)
-		return -EINVAL;
+	struct acpi_fan *fan = dev_get_drvdata(dev);
+	if (fan->acpi4)
+		return 0;
 
-	acpi_bus_set_power(to_acpi_device(dev)->handle, ACPI_STATE_D0);
+	acpi_device_set_power(ACPI_COMPANION(dev), ACPI_STATE_D0);
 
 	return AE_OK;
 }
@@ -211,16 +404,17 @@ static int acpi_fan_suspend(struct device *dev)
 static int acpi_fan_resume(struct device *dev)
 {
 	int result;
+	struct acpi_fan *fan = dev_get_drvdata(dev);
 
-	if (!dev)
-		return -EINVAL;
+	if (fan->acpi4)
+		return 0;
 
-	result = acpi_bus_update_power(to_acpi_device(dev)->handle, NULL);
+	result = acpi_device_update_power(ACPI_COMPANION(dev), NULL);
 	if (result)
-		printk(KERN_ERR PREFIX "Error updating fan power state\n");
+		dev_err(dev, "Error updating fan power state\n");
 
 	return result;
 }
 #endif
 
-module_acpi_driver(acpi_fan_driver);
+module_platform_driver(acpi_fan_driver);
diff --git a/drivers/acpi/int340x_thermal.c b/drivers/acpi/int340x_thermal.c
new file mode 100644
index 000000000000..9dcf83682e36
--- /dev/null
+++ b/drivers/acpi/int340x_thermal.c
@@ -0,0 +1,54 @@
+/*
+ * ACPI support for int340x thermal drivers
+ *
+ * Copyright (C) 2014, Intel Corporation
+ * Authors: Zhang Rui <rui.zhang@intel.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/acpi.h>
+#include <linux/module.h>
+
+#include "internal.h"
+
+#define INT3401_DEVICE 0X01
+static const struct acpi_device_id int340x_thermal_device_ids[] = {
+	{"INT3400"},
+	{"INT3401", INT3401_DEVICE},
+	{"INT3402"},
+	{"INT3403"},
+	{"INT3404"},
+	{"INT3406"},
+	{"INT3407"},
+	{"INT3408"},
+	{"INT3409"},
+	{"INT340A"},
+	{"INT340B"},
+	{""},
+};
+
+static int int340x_thermal_handler_attach(struct acpi_device *adev,
+					const struct acpi_device_id *id)
+{
+#if defined(CONFIG_INT340X_THERMAL) || defined(CONFIG_INT340X_THERMAL_MODULE)
+	acpi_create_platform_device(adev);
+#elif defined(INTEL_SOC_DTS_THERMAL) || defined(INTEL_SOC_DTS_THERMAL_MODULE)
+	/* Intel SoC DTS thermal driver needs INT3401 to set IRQ descriptor */
+	if (id->driver_data == INT3401_DEVICE)
+		acpi_create_platform_device(adev);
+#endif
+	return 1;
+}
+
+static struct acpi_scan_handler int340x_thermal_handler = {
+	.ids = int340x_thermal_device_ids,
+	.attach = int340x_thermal_handler_attach,
+};
+
+void __init acpi_int340x_thermal_init(void)
+{
+	acpi_scan_add_handler(&int340x_thermal_handler);
+}
diff --git a/drivers/acpi/internal.h b/drivers/acpi/internal.h
index 4c5cf77e7576..163e82f536fa 100644
--- a/drivers/acpi/internal.h
+++ b/drivers/acpi/internal.h
@@ -31,6 +31,7 @@ void acpi_pci_link_init(void);
 void acpi_processor_init(void);
 void acpi_platform_init(void);
 void acpi_pnp_init(void);
+void acpi_int340x_thermal_init(void);
 int acpi_sysfs_init(void);
 void acpi_container_init(void);
 void acpi_memory_hotplug_init(void);
@@ -103,8 +104,6 @@ int acpi_power_get_inferred_state(struct acpi_device *device, int *state);
 int acpi_power_on_resources(struct acpi_device *device, int state);
 int acpi_power_transition(struct acpi_device *device, int state);
 
-int acpi_device_update_power(struct acpi_device *device, int *state_p);
-
 int acpi_wakeup_device_init(void);
 
 #ifdef CONFIG_ARCH_MIGHT_HAVE_ACPI_PDC
@@ -168,17 +167,16 @@ static inline void suspend_nvs_restore(void) {}
 #endif
 
 /*--------------------------------------------------------------------------
-				Platform bus support
-  -------------------------------------------------------------------------- */
-struct platform_device;
-
-struct platform_device *acpi_create_platform_device(struct acpi_device *adev);
-
-/*--------------------------------------------------------------------------
 					Video
   -------------------------------------------------------------------------- */
 #if defined(CONFIG_ACPI_VIDEO) || defined(CONFIG_ACPI_VIDEO_MODULE)
 bool acpi_osi_is_win8(void);
 #endif
 
+/*--------------------------------------------------------------------------
+				Device properties
+  -------------------------------------------------------------------------- */
+void acpi_init_properties(struct acpi_device *adev);
+void acpi_free_properties(struct acpi_device *adev);
+
 #endif /* _ACPI_INTERNAL_H_ */
diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c
index 3abe9b223ba7..f9eeae871593 100644
--- a/drivers/acpi/osl.c
+++ b/drivers/acpi/osl.c
@@ -152,6 +152,16 @@ static u32 acpi_osi_handler(acpi_string interface, u32 supported)
 			osi_linux.dmi ? " via DMI" : "");
 	}
 
+	if (!strcmp("Darwin", interface)) {
+		/*
+		 * Apple firmware will behave poorly if it receives positive
+		 * answers to "Darwin" and any other OS. Respond positively
+		 * to Darwin and then disable all other vendor strings.
+		 */
+		acpi_update_interfaces(ACPI_DISABLE_ALL_VENDOR_STRINGS);
+		supported = ACPI_UINT32_MAX;
+	}
+
 	return supported;
 }
 
@@ -426,7 +436,7 @@ static void acpi_os_drop_map_ref(struct acpi_ioremap *map)
 static void acpi_os_map_cleanup(struct acpi_ioremap *map)
 {
 	if (!map->refcount) {
-		synchronize_rcu();
+		synchronize_rcu_expedited();
 		acpi_unmap(map->phys, map->virt);
 		kfree(map);
 	}
@@ -825,7 +835,7 @@ acpi_os_install_interrupt_handler(u32 gsi, acpi_osd_handler handler,
 
 	acpi_irq_handler = handler;
 	acpi_irq_context = context;
-	if (request_irq(irq, acpi_irq, IRQF_SHARED | IRQF_NO_SUSPEND, "acpi", acpi_irq)) {
+	if (request_irq(irq, acpi_irq, IRQF_SHARED, "acpi", acpi_irq)) {
 		printk(KERN_ERR PREFIX "SCI (IRQ%d) allocation failed\n", irq);
 		acpi_irq_handler = NULL;
 		return AE_NOT_ACQUIRED;
@@ -1178,6 +1188,12 @@ EXPORT_SYMBOL(acpi_os_execute);
 
 void acpi_os_wait_events_complete(void)
 {
+	/*
+	 * Make sure the GPE handler or the fixed event handler is not used
+	 * on another CPU after removal.
+	 */
+	if (acpi_irq_handler)
+		synchronize_hardirq(acpi_gbl_FADT.sci_interrupt);
 	flush_workqueue(kacpid_wq);
 	flush_workqueue(kacpi_notify_wq);
 }
diff --git a/drivers/acpi/pci_irq.c b/drivers/acpi/pci_irq.c
index 6e6b80eb0bba..b1def411c0b8 100644
--- a/drivers/acpi/pci_irq.c
+++ b/drivers/acpi/pci_irq.c
@@ -413,6 +413,9 @@ int acpi_pci_irq_enable(struct pci_dev *dev)
 		return 0;
 	}
 
+	if (dev->irq_managed && dev->irq > 0)
+		return 0;
+
 	entry = acpi_pci_irq_lookup(dev, pin);
 	if (!entry) {
 		/*
@@ -456,6 +459,7 @@ int acpi_pci_irq_enable(struct pci_dev *dev)
 		return rc;
 	}
 	dev->irq = rc;
+	dev->irq_managed = 1;
 
 	if (link)
 		snprintf(link_desc, sizeof(link_desc), " -> Link[%s]", link);
@@ -478,13 +482,13 @@ void acpi_pci_irq_disable(struct pci_dev *dev)
 	u8 pin;
 
 	pin = dev->pin;
-	if (!pin)
+	if (!pin || !dev->irq_managed || dev->irq <= 0)
 		return;
 
 	/* Keep IOAPIC pin configuration when suspending */
 	if (dev->dev.power.is_prepared)
 		return;
-#ifdef	CONFIG_PM_RUNTIME
+#ifdef	CONFIG_PM
 	if (dev->dev.power.runtime_status == RPM_SUSPENDING)
 		return;
 #endif
@@ -506,6 +510,8 @@ void acpi_pci_irq_disable(struct pci_dev *dev)
 	 */
 
 	dev_dbg(&dev->dev, "PCI INT %c disabled\n", pin_name(pin));
-	if (gsi >= 0 && dev->irq > 0)
+	if (gsi >= 0) {
 		acpi_unregister_gsi(gsi);
+		dev->irq_managed = 0;
+	}
 }
diff --git a/drivers/acpi/pci_root.c b/drivers/acpi/pci_root.c
index e6ae603ed1a1..c6bcb8c719d8 100644
--- a/drivers/acpi/pci_root.c
+++ b/drivers/acpi/pci_root.c
@@ -33,8 +33,10 @@
 #include <linux/pci.h>
 #include <linux/pci-acpi.h>
 #include <linux/pci-aspm.h>
+#include <linux/dmar.h>
 #include <linux/acpi.h>
 #include <linux/slab.h>
+#include <linux/dmi.h>
 #include <acpi/apei.h>	/* for acpi_hest_init() */
 
 #include "internal.h"
@@ -430,6 +432,19 @@ static void negotiate_os_control(struct acpi_pci_root *root, int *no_aspm,
 	acpi_handle handle = device->handle;
 
 	/*
+	 * Apple always return failure on _OSC calls when _OSI("Darwin") has
+	 * been called successfully. We know the feature set supported by the
+	 * platform, so avoid calling _OSC at all
+	 */
+
+	if (dmi_match(DMI_SYS_VENDOR, "Apple Inc.")) {
+		root->osc_control_set = ~OSC_PCI_EXPRESS_PME_CONTROL;
+		decode_osc_control(root, "OS assumes control of",
+				   root->osc_control_set);
+		return;
+	}
+
+	/*
 	 * All supported architectures that use ACPI have support for
 	 * PCI domains, so we indicate this in _OSC support capabilities.
 	 */
@@ -511,6 +526,7 @@ static int acpi_pci_root_add(struct acpi_device *device,
 	struct acpi_pci_root *root;
 	acpi_handle handle = device->handle;
 	int no_aspm = 0, clear_aspm = 0;
+	bool hotadd = system_state != SYSTEM_BOOTING;
 
 	root = kzalloc(sizeof(struct acpi_pci_root), GFP_KERNEL);
 	if (!root)
@@ -557,6 +573,11 @@ static int acpi_pci_root_add(struct acpi_device *device,
 	strcpy(acpi_device_class(device), ACPI_PCI_ROOT_CLASS);
 	device->driver_data = root;
 
+	if (hotadd && dmar_device_add(handle)) {
+		result = -ENXIO;
+		goto end;
+	}
+
 	pr_info(PREFIX "%s [%s] (domain %04x %pR)\n",
 	       acpi_device_name(device), acpi_device_bid(device),
 	       root->segment, &root->secondary);
@@ -583,7 +604,7 @@ static int acpi_pci_root_add(struct acpi_device *device,
 			root->segment, (unsigned int)root->secondary.start);
 		device->driver_data = NULL;
 		result = -ENODEV;
-		goto end;
+		goto remove_dmar;
 	}
 
 	if (clear_aspm) {
@@ -597,7 +618,7 @@ static int acpi_pci_root_add(struct acpi_device *device,
 	if (device->wakeup.flags.run_wake)
 		device_set_run_wake(root->bus->bridge, true);
 
-	if (system_state != SYSTEM_BOOTING) {
+	if (hotadd) {
 		pcibios_resource_survey_bus(root->bus);
 		pci_assign_unassigned_root_bus_resources(root->bus);
 	}
@@ -607,6 +628,9 @@ static int acpi_pci_root_add(struct acpi_device *device,
 	pci_unlock_rescan_remove();
 	return 1;
 
+remove_dmar:
+	if (hotadd)
+		dmar_device_remove(handle);
 end:
 	kfree(root);
 	return result;
@@ -625,6 +649,8 @@ static void acpi_pci_root_remove(struct acpi_device *device)
 
 	pci_remove_root_bus(root->bus);
 
+	dmar_device_remove(device->handle);
+
 	pci_unlock_rescan_remove();
 
 	kfree(root);
diff --git a/drivers/acpi/pmic/intel_pmic.c b/drivers/acpi/pmic/intel_pmic.c
new file mode 100644
index 000000000000..a732e5d7e322
--- /dev/null
+++ b/drivers/acpi/pmic/intel_pmic.c
@@ -0,0 +1,354 @@
+/*
+ * intel_pmic.c - Intel PMIC operation region driver
+ *
+ * Copyright (C) 2014 Intel Corporation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version
+ * 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/module.h>
+#include <linux/acpi.h>
+#include <linux/regmap.h>
+#include "intel_pmic.h"
+
+#define PMIC_POWER_OPREGION_ID		0x8d
+#define PMIC_THERMAL_OPREGION_ID	0x8c
+
+struct acpi_lpat {
+	int temp;
+	int raw;
+};
+
+struct intel_pmic_opregion {
+	struct mutex lock;
+	struct acpi_lpat *lpat;
+	int lpat_count;
+	struct regmap *regmap;
+	struct intel_pmic_opregion_data *data;
+};
+
+static int pmic_get_reg_bit(int address, struct pmic_table *table,
+			    int count, int *reg, int *bit)
+{
+	int i;
+
+	for (i = 0; i < count; i++) {
+		if (table[i].address == address) {
+			*reg = table[i].reg;
+			if (bit)
+				*bit = table[i].bit;
+			return 0;
+		}
+	}
+	return -ENOENT;
+}
+
+/**
+ * raw_to_temp(): Return temperature from raw value through LPAT table
+ *
+ * @lpat: the temperature_raw mapping table
+ * @count: the count of the above mapping table
+ * @raw: the raw value, used as a key to get the temerature from the
+ *       above mapping table
+ *
+ * A positive value will be returned on success, a negative errno will
+ * be returned in error cases.
+ */
+static int raw_to_temp(struct acpi_lpat *lpat, int count, int raw)
+{
+	int i, delta_temp, delta_raw, temp;
+
+	for (i = 0; i < count - 1; i++) {
+		if ((raw >= lpat[i].raw && raw <= lpat[i+1].raw) ||
+		    (raw <= lpat[i].raw && raw >= lpat[i+1].raw))
+			break;
+	}
+
+	if (i == count - 1)
+		return -ENOENT;
+
+	delta_temp = lpat[i+1].temp - lpat[i].temp;
+	delta_raw = lpat[i+1].raw - lpat[i].raw;
+	temp = lpat[i].temp + (raw - lpat[i].raw) * delta_temp / delta_raw;
+
+	return temp;
+}
+
+/**
+ * temp_to_raw(): Return raw value from temperature through LPAT table
+ *
+ * @lpat: the temperature_raw mapping table
+ * @count: the count of the above mapping table
+ * @temp: the temperature, used as a key to get the raw value from the
+ *        above mapping table
+ *
+ * A positive value will be returned on success, a negative errno will
+ * be returned in error cases.
+ */
+static int temp_to_raw(struct acpi_lpat *lpat, int count, int temp)
+{
+	int i, delta_temp, delta_raw, raw;
+
+	for (i = 0; i < count - 1; i++) {
+		if (temp >= lpat[i].temp && temp <= lpat[i+1].temp)
+			break;
+	}
+
+	if (i == count - 1)
+		return -ENOENT;
+
+	delta_temp = lpat[i+1].temp - lpat[i].temp;
+	delta_raw = lpat[i+1].raw - lpat[i].raw;
+	raw = lpat[i].raw + (temp - lpat[i].temp) * delta_raw / delta_temp;
+
+	return raw;
+}
+
+static void pmic_thermal_lpat(struct intel_pmic_opregion *opregion,
+			      acpi_handle handle, struct device *dev)
+{
+	struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL };
+	union acpi_object *obj_p, *obj_e;
+	int *lpat, i;
+	acpi_status status;
+
+	status = acpi_evaluate_object(handle, "LPAT", NULL, &buffer);
+	if (ACPI_FAILURE(status))
+		return;
+
+	obj_p = (union acpi_object *)buffer.pointer;
+	if (!obj_p || (obj_p->type != ACPI_TYPE_PACKAGE) ||
+	    (obj_p->package.count % 2) || (obj_p->package.count < 4))
+		goto out;
+
+	lpat = devm_kmalloc(dev, sizeof(int) * obj_p->package.count,
+			    GFP_KERNEL);
+	if (!lpat)
+		goto out;
+
+	for (i = 0; i < obj_p->package.count; i++) {
+		obj_e = &obj_p->package.elements[i];
+		if (obj_e->type != ACPI_TYPE_INTEGER) {
+			devm_kfree(dev, lpat);
+			goto out;
+		}
+		lpat[i] = (s64)obj_e->integer.value;
+	}
+
+	opregion->lpat = (struct acpi_lpat *)lpat;
+	opregion->lpat_count = obj_p->package.count / 2;
+
+out:
+	kfree(buffer.pointer);
+}
+
+static acpi_status intel_pmic_power_handler(u32 function,
+		acpi_physical_address address, u32 bits, u64 *value64,
+		void *handler_context, void *region_context)
+{
+	struct intel_pmic_opregion *opregion = region_context;
+	struct regmap *regmap = opregion->regmap;
+	struct intel_pmic_opregion_data *d = opregion->data;
+	int reg, bit, result;
+
+	if (bits != 32 || !value64)
+		return AE_BAD_PARAMETER;
+
+	if (function == ACPI_WRITE && !(*value64 == 0 || *value64 == 1))
+		return AE_BAD_PARAMETER;
+
+	result = pmic_get_reg_bit(address, d->power_table,
+				  d->power_table_count, &reg, &bit);
+	if (result == -ENOENT)
+		return AE_BAD_PARAMETER;
+
+	mutex_lock(&opregion->lock);
+
+	result = function == ACPI_READ ?
+		d->get_power(regmap, reg, bit, value64) :
+		d->update_power(regmap, reg, bit, *value64 == 1);
+
+	mutex_unlock(&opregion->lock);
+
+	return result ? AE_ERROR : AE_OK;
+}
+
+static int pmic_read_temp(struct intel_pmic_opregion *opregion,
+			  int reg, u64 *value)
+{
+	int raw_temp, temp;
+
+	if (!opregion->data->get_raw_temp)
+		return -ENXIO;
+
+	raw_temp = opregion->data->get_raw_temp(opregion->regmap, reg);
+	if (raw_temp < 0)
+		return raw_temp;
+
+	if (!opregion->lpat) {
+		*value = raw_temp;
+		return 0;
+	}
+
+	temp = raw_to_temp(opregion->lpat, opregion->lpat_count, raw_temp);
+	if (temp < 0)
+		return temp;
+
+	*value = temp;
+	return 0;
+}
+
+static int pmic_thermal_temp(struct intel_pmic_opregion *opregion, int reg,
+			     u32 function, u64 *value)
+{
+	return function == ACPI_READ ?
+		pmic_read_temp(opregion, reg, value) : -EINVAL;
+}
+
+static int pmic_thermal_aux(struct intel_pmic_opregion *opregion, int reg,
+			    u32 function, u64 *value)
+{
+	int raw_temp;
+
+	if (function == ACPI_READ)
+		return pmic_read_temp(opregion, reg, value);
+
+	if (!opregion->data->update_aux)
+		return -ENXIO;
+
+	if (opregion->lpat) {
+		raw_temp = temp_to_raw(opregion->lpat, opregion->lpat_count,
+				       *value);
+		if (raw_temp < 0)
+			return raw_temp;
+	} else {
+		raw_temp = *value;
+	}
+
+	return opregion->data->update_aux(opregion->regmap, reg, raw_temp);
+}
+
+static int pmic_thermal_pen(struct intel_pmic_opregion *opregion, int reg,
+			    u32 function, u64 *value)
+{
+	struct intel_pmic_opregion_data *d = opregion->data;
+	struct regmap *regmap = opregion->regmap;
+
+	if (!d->get_policy || !d->update_policy)
+		return -ENXIO;
+
+	if (function == ACPI_READ)
+		return d->get_policy(regmap, reg, value);
+
+	if (*value != 0 && *value != 1)
+		return -EINVAL;
+
+	return d->update_policy(regmap, reg, *value);
+}
+
+static bool pmic_thermal_is_temp(int address)
+{
+	return (address <= 0x3c) && !(address % 12);
+}
+
+static bool pmic_thermal_is_aux(int address)
+{
+	return (address >= 4 && address <= 0x40 && !((address - 4) % 12)) ||
+	       (address >= 8 && address <= 0x44 && !((address - 8) % 12));
+}
+
+static bool pmic_thermal_is_pen(int address)
+{
+	return address >= 0x48 && address <= 0x5c;
+}
+
+static acpi_status intel_pmic_thermal_handler(u32 function,
+		acpi_physical_address address, u32 bits, u64 *value64,
+		void *handler_context, void *region_context)
+{
+	struct intel_pmic_opregion *opregion = region_context;
+	struct intel_pmic_opregion_data *d = opregion->data;
+	int reg, result;
+
+	if (bits != 32 || !value64)
+		return AE_BAD_PARAMETER;
+
+	result = pmic_get_reg_bit(address, d->thermal_table,
+				  d->thermal_table_count, &reg, NULL);
+	if (result == -ENOENT)
+		return AE_BAD_PARAMETER;
+
+	mutex_lock(&opregion->lock);
+
+	if (pmic_thermal_is_temp(address))
+		result = pmic_thermal_temp(opregion, reg, function, value64);
+	else if (pmic_thermal_is_aux(address))
+		result = pmic_thermal_aux(opregion, reg, function, value64);
+	else if (pmic_thermal_is_pen(address))
+		result = pmic_thermal_pen(opregion, reg, function, value64);
+	else
+		result = -EINVAL;
+
+	mutex_unlock(&opregion->lock);
+
+	if (result < 0) {
+		if (result == -EINVAL)
+			return AE_BAD_PARAMETER;
+		else
+			return AE_ERROR;
+	}
+
+	return AE_OK;
+}
+
+int intel_pmic_install_opregion_handler(struct device *dev, acpi_handle handle,
+					struct regmap *regmap,
+					struct intel_pmic_opregion_data *d)
+{
+	acpi_status status;
+	struct intel_pmic_opregion *opregion;
+
+	if (!dev || !regmap || !d)
+		return -EINVAL;
+
+	if (!handle)
+		return -ENODEV;
+
+	opregion = devm_kzalloc(dev, sizeof(*opregion), GFP_KERNEL);
+	if (!opregion)
+		return -ENOMEM;
+
+	mutex_init(&opregion->lock);
+	opregion->regmap = regmap;
+	pmic_thermal_lpat(opregion, handle, dev);
+
+	status = acpi_install_address_space_handler(handle,
+						    PMIC_POWER_OPREGION_ID,
+						    intel_pmic_power_handler,
+						    NULL, opregion);
+	if (ACPI_FAILURE(status))
+		return -ENODEV;
+
+	status = acpi_install_address_space_handler(handle,
+						    PMIC_THERMAL_OPREGION_ID,
+						    intel_pmic_thermal_handler,
+						    NULL, opregion);
+	if (ACPI_FAILURE(status)) {
+		acpi_remove_address_space_handler(handle, PMIC_POWER_OPREGION_ID,
+						  intel_pmic_power_handler);
+		return -ENODEV;
+	}
+
+	opregion->data = d;
+	return 0;
+}
+EXPORT_SYMBOL_GPL(intel_pmic_install_opregion_handler);
+
+MODULE_LICENSE("GPL");
diff --git a/drivers/acpi/pmic/intel_pmic.h b/drivers/acpi/pmic/intel_pmic.h
new file mode 100644
index 000000000000..d4e90af8f0dd
--- /dev/null
+++ b/drivers/acpi/pmic/intel_pmic.h
@@ -0,0 +1,25 @@
+#ifndef __INTEL_PMIC_H
+#define __INTEL_PMIC_H
+
+struct pmic_table {
+	int address;	/* operation region address */
+	int reg;	/* corresponding thermal register */
+	int bit;	/* control bit for power */
+};
+
+struct intel_pmic_opregion_data {
+	int (*get_power)(struct regmap *r, int reg, int bit, u64 *value);
+	int (*update_power)(struct regmap *r, int reg, int bit, bool on);
+	int (*get_raw_temp)(struct regmap *r, int reg);
+	int (*update_aux)(struct regmap *r, int reg, int raw_temp);
+	int (*get_policy)(struct regmap *r, int reg, u64 *value);
+	int (*update_policy)(struct regmap *r, int reg, int enable);
+	struct pmic_table *power_table;
+	int power_table_count;
+	struct pmic_table *thermal_table;
+	int thermal_table_count;
+};
+
+int intel_pmic_install_opregion_handler(struct device *dev, acpi_handle handle, struct regmap *regmap, struct intel_pmic_opregion_data *d);
+
+#endif
diff --git a/drivers/acpi/pmic/intel_pmic_crc.c b/drivers/acpi/pmic/intel_pmic_crc.c
new file mode 100644
index 000000000000..ef7d8ff95abe
--- /dev/null
+++ b/drivers/acpi/pmic/intel_pmic_crc.c
@@ -0,0 +1,211 @@
+/*
+ * intel_pmic_crc.c - Intel CrystalCove PMIC operation region driver
+ *
+ * Copyright (C) 2014 Intel Corporation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version
+ * 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/module.h>
+#include <linux/acpi.h>
+#include <linux/mfd/intel_soc_pmic.h>
+#include <linux/regmap.h>
+#include <linux/platform_device.h>
+#include "intel_pmic.h"
+
+#define PWR_SOURCE_SELECT	BIT(1)
+
+#define PMIC_A0LOCK_REG		0xc5
+
+static struct pmic_table power_table[] = {
+	{
+		.address = 0x24,
+		.reg = 0x66,
+		.bit = 0x00,
+	},
+	{
+		.address = 0x48,
+		.reg = 0x5d,
+		.bit = 0x00,
+	},
+};
+
+static struct pmic_table thermal_table[] = {
+	{
+		.address = 0x00,
+		.reg = 0x75
+	},
+	{
+		.address = 0x04,
+		.reg = 0x95
+	},
+	{
+		.address = 0x08,
+		.reg = 0x97
+	},
+	{
+		.address = 0x0c,
+		.reg = 0x77
+	},
+	{
+		.address = 0x10,
+		.reg = 0x9a
+	},
+	{
+		.address = 0x14,
+		.reg = 0x9c
+	},
+	{
+		.address = 0x18,
+		.reg = 0x79
+	},
+	{
+		.address = 0x1c,
+		.reg = 0x9f
+	},
+	{
+		.address = 0x20,
+		.reg = 0xa1
+	},
+	{
+		.address = 0x48,
+		.reg = 0x94
+	},
+	{
+		.address = 0x4c,
+		.reg = 0x99
+	},
+	{
+		.address = 0x50,
+		.reg = 0x9e
+	},
+};
+
+static int intel_crc_pmic_get_power(struct regmap *regmap, int reg,
+				    int bit, u64 *value)
+{
+	int data;
+
+	if (regmap_read(regmap, reg, &data))
+		return -EIO;
+
+	*value = (data & PWR_SOURCE_SELECT) && (data & BIT(bit)) ? 1 : 0;
+	return 0;
+}
+
+static int intel_crc_pmic_update_power(struct regmap *regmap, int reg,
+				       int bit, bool on)
+{
+	int data;
+
+	if (regmap_read(regmap, reg, &data))
+		return -EIO;
+
+	if (on) {
+		data |= PWR_SOURCE_SELECT | BIT(bit);
+	} else {
+		data &= ~BIT(bit);
+		data |= PWR_SOURCE_SELECT;
+	}
+
+	if (regmap_write(regmap, reg, data))
+		return -EIO;
+	return 0;
+}
+
+static int intel_crc_pmic_get_raw_temp(struct regmap *regmap, int reg)
+{
+	int temp_l, temp_h;
+
+	/*
+	 * Raw temperature value is 10bits: 8bits in reg
+	 * and 2bits in reg-1: bit0,1
+	 */
+	if (regmap_read(regmap, reg, &temp_l) ||
+	    regmap_read(regmap, reg - 1, &temp_h))
+		return -EIO;
+
+	return temp_l | (temp_h & 0x3) << 8;
+}
+
+static int intel_crc_pmic_update_aux(struct regmap *regmap, int reg, int raw)
+{
+	return regmap_write(regmap, reg, raw) ||
+		regmap_update_bits(regmap, reg - 1, 0x3, raw >> 8) ? -EIO : 0;
+}
+
+static int intel_crc_pmic_get_policy(struct regmap *regmap, int reg, u64 *value)
+{
+	int pen;
+
+	if (regmap_read(regmap, reg, &pen))
+		return -EIO;
+	*value = pen >> 7;
+	return 0;
+}
+
+static int intel_crc_pmic_update_policy(struct regmap *regmap,
+					int reg, int enable)
+{
+	int alert0;
+
+	/* Update to policy enable bit requires unlocking a0lock */
+	if (regmap_read(regmap, PMIC_A0LOCK_REG, &alert0))
+		return -EIO;
+
+	if (regmap_update_bits(regmap, PMIC_A0LOCK_REG, 0x01, 0))
+		return -EIO;
+
+	if (regmap_update_bits(regmap, reg, 0x80, enable << 7))
+		return -EIO;
+
+	/* restore alert0 */
+	if (regmap_write(regmap, PMIC_A0LOCK_REG, alert0))
+		return -EIO;
+
+	return 0;
+}
+
+static struct intel_pmic_opregion_data intel_crc_pmic_opregion_data = {
+	.get_power	= intel_crc_pmic_get_power,
+	.update_power	= intel_crc_pmic_update_power,
+	.get_raw_temp	= intel_crc_pmic_get_raw_temp,
+	.update_aux	= intel_crc_pmic_update_aux,
+	.get_policy	= intel_crc_pmic_get_policy,
+	.update_policy	= intel_crc_pmic_update_policy,
+	.power_table	= power_table,
+	.power_table_count= ARRAY_SIZE(power_table),
+	.thermal_table	= thermal_table,
+	.thermal_table_count = ARRAY_SIZE(thermal_table),
+};
+
+static int intel_crc_pmic_opregion_probe(struct platform_device *pdev)
+{
+	struct intel_soc_pmic *pmic = dev_get_drvdata(pdev->dev.parent);
+	return intel_pmic_install_opregion_handler(&pdev->dev,
+			ACPI_HANDLE(pdev->dev.parent), pmic->regmap,
+			&intel_crc_pmic_opregion_data);
+}
+
+static struct platform_driver intel_crc_pmic_opregion_driver = {
+	.probe = intel_crc_pmic_opregion_probe,
+	.driver = {
+		.name = "crystal_cove_pmic",
+	},
+};
+
+static int __init intel_crc_pmic_opregion_driver_init(void)
+{
+	return platform_driver_register(&intel_crc_pmic_opregion_driver);
+}
+module_init(intel_crc_pmic_opregion_driver_init);
+
+MODULE_DESCRIPTION("CrystalCove ACPI opration region driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/acpi/pmic/intel_pmic_xpower.c b/drivers/acpi/pmic/intel_pmic_xpower.c
new file mode 100644
index 000000000000..6a082d4de12c
--- /dev/null
+++ b/drivers/acpi/pmic/intel_pmic_xpower.c
@@ -0,0 +1,268 @@
+/*
+ * intel_pmic_xpower.c - XPower AXP288 PMIC operation region driver
+ *
+ * Copyright (C) 2014 Intel Corporation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version
+ * 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/module.h>
+#include <linux/acpi.h>
+#include <linux/mfd/axp20x.h>
+#include <linux/regmap.h>
+#include <linux/platform_device.h>
+#include <linux/iio/consumer.h>
+#include "intel_pmic.h"
+
+#define XPOWER_GPADC_LOW	0x5b
+
+static struct pmic_table power_table[] = {
+	{
+		.address = 0x00,
+		.reg = 0x13,
+		.bit = 0x05,
+	},
+	{
+		.address = 0x04,
+		.reg = 0x13,
+		.bit = 0x06,
+	},
+	{
+		.address = 0x08,
+		.reg = 0x13,
+		.bit = 0x07,
+	},
+	{
+		.address = 0x0c,
+		.reg = 0x12,
+		.bit = 0x03,
+	},
+	{
+		.address = 0x10,
+		.reg = 0x12,
+		.bit = 0x04,
+	},
+	{
+		.address = 0x14,
+		.reg = 0x12,
+		.bit = 0x05,
+	},
+	{
+		.address = 0x18,
+		.reg = 0x12,
+		.bit = 0x06,
+	},
+	{
+		.address = 0x1c,
+		.reg = 0x12,
+		.bit = 0x00,
+	},
+	{
+		.address = 0x20,
+		.reg = 0x12,
+		.bit = 0x01,
+	},
+	{
+		.address = 0x24,
+		.reg = 0x12,
+		.bit = 0x02,
+	},
+	{
+		.address = 0x28,
+		.reg = 0x13,
+		.bit = 0x02,
+	},
+	{
+		.address = 0x2c,
+		.reg = 0x13,
+		.bit = 0x03,
+	},
+	{
+		.address = 0x30,
+		.reg = 0x13,
+		.bit = 0x04,
+	},
+	{
+		.address = 0x38,
+		.reg = 0x10,
+		.bit = 0x03,
+	},
+	{
+		.address = 0x3c,
+		.reg = 0x10,
+		.bit = 0x06,
+	},
+	{
+		.address = 0x40,
+		.reg = 0x10,
+		.bit = 0x05,
+	},
+	{
+		.address = 0x44,
+		.reg = 0x10,
+		.bit = 0x04,
+	},
+	{
+		.address = 0x48,
+		.reg = 0x10,
+		.bit = 0x01,
+	},
+	{
+		.address = 0x4c,
+		.reg = 0x10,
+		.bit = 0x00
+	},
+};
+
+/* TMP0 - TMP5 are the same, all from GPADC */
+static struct pmic_table thermal_table[] = {
+	{
+		.address = 0x00,
+		.reg = XPOWER_GPADC_LOW
+	},
+	{
+		.address = 0x0c,
+		.reg = XPOWER_GPADC_LOW
+	},
+	{
+		.address = 0x18,
+		.reg = XPOWER_GPADC_LOW
+	},
+	{
+		.address = 0x24,
+		.reg = XPOWER_GPADC_LOW
+	},
+	{
+		.address = 0x30,
+		.reg = XPOWER_GPADC_LOW
+	},
+	{
+		.address = 0x3c,
+		.reg = XPOWER_GPADC_LOW
+	},
+};
+
+static int intel_xpower_pmic_get_power(struct regmap *regmap, int reg,
+				       int bit, u64 *value)
+{
+	int data;
+
+	if (regmap_read(regmap, reg, &data))
+		return -EIO;
+
+	*value = (data & BIT(bit)) ? 1 : 0;
+	return 0;
+}
+
+static int intel_xpower_pmic_update_power(struct regmap *regmap, int reg,
+					  int bit, bool on)
+{
+	int data;
+
+	if (regmap_read(regmap, reg, &data))
+		return -EIO;
+
+	if (on)
+		data |= BIT(bit);
+	else
+		data &= ~BIT(bit);
+
+	if (regmap_write(regmap, reg, data))
+		return -EIO;
+
+	return 0;
+}
+
+/**
+ * intel_xpower_pmic_get_raw_temp(): Get raw temperature reading from the PMIC
+ *
+ * @regmap: regmap of the PMIC device
+ * @reg: register to get the reading
+ *
+ * We could get the sensor value by manipulating the HW regs here, but since
+ * the axp288 IIO driver may also access the same regs at the same time, the
+ * APIs provided by IIO subsystem are used here instead to avoid problems. As
+ * a result, the two passed in params are of no actual use.
+ *
+ * Return a positive value on success, errno on failure.
+ */
+static int intel_xpower_pmic_get_raw_temp(struct regmap *regmap, int reg)
+{
+	struct iio_channel *gpadc_chan;
+	int ret, val;
+
+	gpadc_chan = iio_channel_get(NULL, "axp288-system-temp");
+	if (IS_ERR_OR_NULL(gpadc_chan))
+		return -EACCES;
+
+	ret = iio_read_channel_raw(gpadc_chan, &val);
+	if (ret < 0)
+		val = ret;
+
+	iio_channel_release(gpadc_chan);
+	return val;
+}
+
+static struct intel_pmic_opregion_data intel_xpower_pmic_opregion_data = {
+	.get_power = intel_xpower_pmic_get_power,
+	.update_power = intel_xpower_pmic_update_power,
+	.get_raw_temp = intel_xpower_pmic_get_raw_temp,
+	.power_table = power_table,
+	.power_table_count = ARRAY_SIZE(power_table),
+	.thermal_table = thermal_table,
+	.thermal_table_count = ARRAY_SIZE(thermal_table),
+};
+
+static acpi_status intel_xpower_pmic_gpio_handler(u32 function,
+		acpi_physical_address address, u32 bit_width, u64 *value,
+		void *handler_context, void *region_context)
+{
+	return AE_OK;
+}
+
+static int intel_xpower_pmic_opregion_probe(struct platform_device *pdev)
+{
+	struct device *parent = pdev->dev.parent;
+	struct axp20x_dev *axp20x = dev_get_drvdata(parent);
+	acpi_status status;
+	int result;
+
+	status = acpi_install_address_space_handler(ACPI_HANDLE(parent),
+			ACPI_ADR_SPACE_GPIO, intel_xpower_pmic_gpio_handler,
+			NULL, NULL);
+	if (ACPI_FAILURE(status))
+		return -ENODEV;
+
+	result = intel_pmic_install_opregion_handler(&pdev->dev,
+					ACPI_HANDLE(parent), axp20x->regmap,
+					&intel_xpower_pmic_opregion_data);
+	if (result)
+		acpi_remove_address_space_handler(ACPI_HANDLE(parent),
+						  ACPI_ADR_SPACE_GPIO,
+						  intel_xpower_pmic_gpio_handler);
+
+	return result;
+}
+
+static struct platform_driver intel_xpower_pmic_opregion_driver = {
+	.probe = intel_xpower_pmic_opregion_probe,
+	.driver = {
+		.name = "axp288_pmic_acpi",
+	},
+};
+
+static int __init intel_xpower_pmic_opregion_driver_init(void)
+{
+	return platform_driver_register(&intel_xpower_pmic_opregion_driver);
+}
+module_init(intel_xpower_pmic_opregion_driver_init);
+
+MODULE_DESCRIPTION("XPower AXP288 ACPI operation region driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/acpi/processor_core.c b/drivers/acpi/processor_core.c
index e32321ce9a5c..02e48394276c 100644
--- a/drivers/acpi/processor_core.c
+++ b/drivers/acpi/processor_core.c
@@ -16,7 +16,7 @@ static int map_lapic_id(struct acpi_subtable_header *entry,
 		 u32 acpi_id, int *apic_id)
 {
 	struct acpi_madt_local_apic *lapic =
-		(struct acpi_madt_local_apic *)entry;
+		container_of(entry, struct acpi_madt_local_apic, header);
 
 	if (!(lapic->lapic_flags & ACPI_MADT_ENABLED))
 		return -ENODEV;
@@ -32,7 +32,7 @@ static int map_x2apic_id(struct acpi_subtable_header *entry,
 			 int device_declaration, u32 acpi_id, int *apic_id)
 {
 	struct acpi_madt_local_x2apic *apic =
-		(struct acpi_madt_local_x2apic *)entry;
+		container_of(entry, struct acpi_madt_local_x2apic, header);
 
 	if (!(apic->lapic_flags & ACPI_MADT_ENABLED))
 		return -ENODEV;
@@ -49,7 +49,7 @@ static int map_lsapic_id(struct acpi_subtable_header *entry,
 		int device_declaration, u32 acpi_id, int *apic_id)
 {
 	struct acpi_madt_local_sapic *lsapic =
-		(struct acpi_madt_local_sapic *)entry;
+		container_of(entry, struct acpi_madt_local_sapic, header);
 
 	if (!(lsapic->lapic_flags & ACPI_MADT_ENABLED))
 		return -ENODEV;
@@ -69,7 +69,7 @@ static int map_madt_entry(int type, u32 acpi_id)
 	unsigned long madt_end, entry;
 	static struct acpi_table_madt *madt;
 	static int read_madt;
-	int apic_id = -1;
+	int phys_id = -1;	/* CPU hardware ID */
 
 	if (!read_madt) {
 		if (ACPI_FAILURE(acpi_get_table(ACPI_SIG_MADT, 0,
@@ -79,7 +79,7 @@ static int map_madt_entry(int type, u32 acpi_id)
 	}
 
 	if (!madt)
-		return apic_id;
+		return phys_id;
 
 	entry = (unsigned long)madt;
 	madt_end = entry + madt->header.length;
@@ -91,18 +91,18 @@ static int map_madt_entry(int type, u32 acpi_id)
 		struct acpi_subtable_header *header =
 			(struct acpi_subtable_header *)entry;
 		if (header->type == ACPI_MADT_TYPE_LOCAL_APIC) {
-			if (!map_lapic_id(header, acpi_id, &apic_id))
+			if (!map_lapic_id(header, acpi_id, &phys_id))
 				break;
 		} else if (header->type == ACPI_MADT_TYPE_LOCAL_X2APIC) {
-			if (!map_x2apic_id(header, type, acpi_id, &apic_id))
+			if (!map_x2apic_id(header, type, acpi_id, &phys_id))
 				break;
 		} else if (header->type == ACPI_MADT_TYPE_LOCAL_SAPIC) {
-			if (!map_lsapic_id(header, type, acpi_id, &apic_id))
+			if (!map_lsapic_id(header, type, acpi_id, &phys_id))
 				break;
 		}
 		entry += header->length;
 	}
-	return apic_id;
+	return phys_id;
 }
 
 static int map_mat_entry(acpi_handle handle, int type, u32 acpi_id)
@@ -110,7 +110,7 @@ static int map_mat_entry(acpi_handle handle, int type, u32 acpi_id)
 	struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL };
 	union acpi_object *obj;
 	struct acpi_subtable_header *header;
-	int apic_id = -1;
+	int phys_id = -1;
 
 	if (ACPI_FAILURE(acpi_evaluate_object(handle, "_MAT", NULL, &buffer)))
 		goto exit;
@@ -125,53 +125,52 @@ static int map_mat_entry(acpi_handle handle, int type, u32 acpi_id)
 	}
 
 	header = (struct acpi_subtable_header *)obj->buffer.pointer;
-	if (header->type == ACPI_MADT_TYPE_LOCAL_APIC) {
-		map_lapic_id(header, acpi_id, &apic_id);
-	} else if (header->type == ACPI_MADT_TYPE_LOCAL_SAPIC) {
-		map_lsapic_id(header, type, acpi_id, &apic_id);
-	} else if (header->type == ACPI_MADT_TYPE_LOCAL_X2APIC) {
-		map_x2apic_id(header, type, acpi_id, &apic_id);
-	}
+	if (header->type == ACPI_MADT_TYPE_LOCAL_APIC)
+		map_lapic_id(header, acpi_id, &phys_id);
+	else if (header->type == ACPI_MADT_TYPE_LOCAL_SAPIC)
+		map_lsapic_id(header, type, acpi_id, &phys_id);
+	else if (header->type == ACPI_MADT_TYPE_LOCAL_X2APIC)
+		map_x2apic_id(header, type, acpi_id, &phys_id);
 
 exit:
 	kfree(buffer.pointer);
-	return apic_id;
+	return phys_id;
 }
 
-int acpi_get_apicid(acpi_handle handle, int type, u32 acpi_id)
+int acpi_get_phys_id(acpi_handle handle, int type, u32 acpi_id)
 {
-	int apic_id;
+	int phys_id;
 
-	apic_id = map_mat_entry(handle, type, acpi_id);
-	if (apic_id == -1)
-		apic_id = map_madt_entry(type, acpi_id);
+	phys_id = map_mat_entry(handle, type, acpi_id);
+	if (phys_id == -1)
+		phys_id = map_madt_entry(type, acpi_id);
 
-	return apic_id;
+	return phys_id;
 }
 
-int acpi_map_cpuid(int apic_id, u32 acpi_id)
+int acpi_map_cpuid(int phys_id, u32 acpi_id)
 {
 #ifdef CONFIG_SMP
 	int i;
 #endif
 
-	if (apic_id == -1) {
+	if (phys_id == -1) {
 		/*
 		 * On UP processor, there is no _MAT or MADT table.
-		 * So above apic_id is always set to -1.
+		 * So above phys_id is always set to -1.
 		 *
 		 * BIOS may define multiple CPU handles even for UP processor.
 		 * For example,
 		 *
 		 * Scope (_PR)
-                 * {
+		 * {
 		 *     Processor (CPU0, 0x00, 0x00000410, 0x06) {}
 		 *     Processor (CPU1, 0x01, 0x00000410, 0x06) {}
 		 *     Processor (CPU2, 0x02, 0x00000410, 0x06) {}
 		 *     Processor (CPU3, 0x03, 0x00000410, 0x06) {}
 		 * }
 		 *
-		 * Ignores apic_id and always returns 0 for the processor
+		 * Ignores phys_id and always returns 0 for the processor
 		 * handle with acpi id 0 if nr_cpu_ids is 1.
 		 * This should be the case if SMP tables are not found.
 		 * Return -1 for other CPU's handle.
@@ -179,28 +178,28 @@ int acpi_map_cpuid(int apic_id, u32 acpi_id)
 		if (nr_cpu_ids <= 1 && acpi_id == 0)
 			return acpi_id;
 		else
-			return apic_id;
+			return phys_id;
 	}
 
 #ifdef CONFIG_SMP
 	for_each_possible_cpu(i) {
-		if (cpu_physical_id(i) == apic_id)
+		if (cpu_physical_id(i) == phys_id)
 			return i;
 	}
 #else
 	/* In UP kernel, only processor 0 is valid */
-	if (apic_id == 0)
-		return apic_id;
+	if (phys_id == 0)
+		return phys_id;
 #endif
 	return -1;
 }
 
 int acpi_get_cpuid(acpi_handle handle, int type, u32 acpi_id)
 {
-	int apic_id;
+	int phys_id;
 
-	apic_id = acpi_get_apicid(handle, type, acpi_id);
+	phys_id = acpi_get_phys_id(handle, type, acpi_id);
 
-	return acpi_map_cpuid(apic_id, acpi_id);
+	return acpi_map_cpuid(phys_id, acpi_id);
 }
 EXPORT_SYMBOL_GPL(acpi_get_cpuid);
diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c
index 17f9ec501972..87b704e41877 100644
--- a/drivers/acpi/processor_idle.c
+++ b/drivers/acpi/processor_idle.c
@@ -334,10 +334,10 @@ static int acpi_processor_get_power_info_default(struct acpi_processor *pr)
 
 static int acpi_processor_get_power_info_cst(struct acpi_processor *pr)
 {
-	acpi_status status = 0;
+	acpi_status status;
 	u64 count;
 	int current_count;
-	int i;
+	int i, ret = 0;
 	struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL };
 	union acpi_object *cst;
 
@@ -358,7 +358,7 @@ static int acpi_processor_get_power_info_cst(struct acpi_processor *pr)
 	/* There must be at least 2 elements */
 	if (!cst || (cst->type != ACPI_TYPE_PACKAGE) || cst->package.count < 2) {
 		printk(KERN_ERR PREFIX "not enough elements in _CST\n");
-		status = -EFAULT;
+		ret = -EFAULT;
 		goto end;
 	}
 
@@ -367,7 +367,7 @@ static int acpi_processor_get_power_info_cst(struct acpi_processor *pr)
 	/* Validate number of power states. */
 	if (count < 1 || count != cst->package.count - 1) {
 		printk(KERN_ERR PREFIX "count given by _CST is not valid\n");
-		status = -EFAULT;
+		ret = -EFAULT;
 		goto end;
 	}
 
@@ -489,12 +489,12 @@ static int acpi_processor_get_power_info_cst(struct acpi_processor *pr)
 
 	/* Validate number of power states discovered */
 	if (current_count < 2)
-		status = -EFAULT;
+		ret = -EFAULT;
 
       end:
 	kfree(buffer.pointer);
 
-	return status;
+	return ret;
 }
 
 static void acpi_processor_power_verify_c3(struct acpi_processor *pr,
@@ -985,8 +985,6 @@ static int acpi_processor_setup_cpuidle_states(struct acpi_processor *pr)
 		state->flags = 0;
 		switch (cx->type) {
 			case ACPI_STATE_C1:
-			if (cx->entry_method == ACPI_CSTATE_FFH)
-				state->flags |= CPUIDLE_FLAG_TIME_VALID;
 
 			state->enter = acpi_idle_enter_c1;
 			state->enter_dead = acpi_idle_play_dead;
@@ -994,14 +992,12 @@ static int acpi_processor_setup_cpuidle_states(struct acpi_processor *pr)
 			break;
 
 			case ACPI_STATE_C2:
-			state->flags |= CPUIDLE_FLAG_TIME_VALID;
 			state->enter = acpi_idle_enter_simple;
 			state->enter_dead = acpi_idle_play_dead;
 			drv->safe_state_index = count;
 			break;
 
 			case ACPI_STATE_C3:
-			state->flags |= CPUIDLE_FLAG_TIME_VALID;
 			state->enter = pr->flags.bm_check ?
 					acpi_idle_enter_bm :
 					acpi_idle_enter_simple;
@@ -1111,7 +1107,7 @@ static int acpi_processor_registered;
 
 int acpi_processor_power_init(struct acpi_processor *pr)
 {
-	acpi_status status = 0;
+	acpi_status status;
 	int retval;
 	struct cpuidle_device *dev;
 	static int first_run;
diff --git a/drivers/acpi/property.c b/drivers/acpi/property.c
new file mode 100644
index 000000000000..0d083736e25b
--- /dev/null
+++ b/drivers/acpi/property.c
@@ -0,0 +1,551 @@
+/*
+ * ACPI device specific properties support.
+ *
+ * Copyright (C) 2014, Intel Corporation
+ * All rights reserved.
+ *
+ * Authors: Mika Westerberg <mika.westerberg@linux.intel.com>
+ *          Darren Hart <dvhart@linux.intel.com>
+ *          Rafael J. Wysocki <rafael.j.wysocki@intel.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/acpi.h>
+#include <linux/device.h>
+#include <linux/export.h>
+
+#include "internal.h"
+
+/* ACPI _DSD device properties UUID: daffd814-6eba-4d8c-8a91-bc9bbf4aa301 */
+static const u8 prp_uuid[16] = {
+	0x14, 0xd8, 0xff, 0xda, 0xba, 0x6e, 0x8c, 0x4d,
+	0x8a, 0x91, 0xbc, 0x9b, 0xbf, 0x4a, 0xa3, 0x01
+};
+
+static bool acpi_property_value_ok(const union acpi_object *value)
+{
+	int j;
+
+	/*
+	 * The value must be an integer, a string, a reference, or a package
+	 * whose every element must be an integer, a string, or a reference.
+	 */
+	switch (value->type) {
+	case ACPI_TYPE_INTEGER:
+	case ACPI_TYPE_STRING:
+	case ACPI_TYPE_LOCAL_REFERENCE:
+		return true;
+
+	case ACPI_TYPE_PACKAGE:
+		for (j = 0; j < value->package.count; j++)
+			switch (value->package.elements[j].type) {
+			case ACPI_TYPE_INTEGER:
+			case ACPI_TYPE_STRING:
+			case ACPI_TYPE_LOCAL_REFERENCE:
+				continue;
+
+			default:
+				return false;
+			}
+
+		return true;
+	}
+	return false;
+}
+
+static bool acpi_properties_format_valid(const union acpi_object *properties)
+{
+	int i;
+
+	for (i = 0; i < properties->package.count; i++) {
+		const union acpi_object *property;
+
+		property = &properties->package.elements[i];
+		/*
+		 * Only two elements allowed, the first one must be a string and
+		 * the second one has to satisfy certain conditions.
+		 */
+		if (property->package.count != 2
+		    || property->package.elements[0].type != ACPI_TYPE_STRING
+		    || !acpi_property_value_ok(&property->package.elements[1]))
+			return false;
+	}
+	return true;
+}
+
+static void acpi_init_of_compatible(struct acpi_device *adev)
+{
+	const union acpi_object *of_compatible;
+	struct acpi_hardware_id *hwid;
+	bool acpi_of = false;
+	int ret;
+
+	/*
+	 * Check if the special PRP0001 ACPI ID is present and in that
+	 * case we fill in Device Tree compatible properties for this
+	 * device.
+	 */
+	list_for_each_entry(hwid, &adev->pnp.ids, list) {
+		if (!strcmp(hwid->id, "PRP0001")) {
+			acpi_of = true;
+			break;
+		}
+	}
+
+	if (!acpi_of)
+		return;
+
+	ret = acpi_dev_get_property_array(adev, "compatible", ACPI_TYPE_STRING,
+					  &of_compatible);
+	if (ret) {
+		ret = acpi_dev_get_property(adev, "compatible",
+					    ACPI_TYPE_STRING, &of_compatible);
+		if (ret) {
+			acpi_handle_warn(adev->handle,
+					 "PRP0001 requires compatible property\n");
+			return;
+		}
+	}
+	adev->data.of_compatible = of_compatible;
+}
+
+void acpi_init_properties(struct acpi_device *adev)
+{
+	struct acpi_buffer buf = { ACPI_ALLOCATE_BUFFER };
+	const union acpi_object *desc;
+	acpi_status status;
+	int i;
+
+	status = acpi_evaluate_object_typed(adev->handle, "_DSD", NULL, &buf,
+					    ACPI_TYPE_PACKAGE);
+	if (ACPI_FAILURE(status))
+		return;
+
+	desc = buf.pointer;
+	if (desc->package.count % 2)
+		goto fail;
+
+	/* Look for the device properties UUID. */
+	for (i = 0; i < desc->package.count; i += 2) {
+		const union acpi_object *uuid, *properties;
+
+		uuid = &desc->package.elements[i];
+		properties = &desc->package.elements[i + 1];
+
+		/*
+		 * The first element must be a UUID and the second one must be
+		 * a package.
+		 */
+		if (uuid->type != ACPI_TYPE_BUFFER || uuid->buffer.length != 16
+		    || properties->type != ACPI_TYPE_PACKAGE)
+			break;
+
+		if (memcmp(uuid->buffer.pointer, prp_uuid, sizeof(prp_uuid)))
+			continue;
+
+		/*
+		 * We found the matching UUID. Now validate the format of the
+		 * package immediately following it.
+		 */
+		if (!acpi_properties_format_valid(properties))
+			break;
+
+		adev->data.pointer = buf.pointer;
+		adev->data.properties = properties;
+
+		acpi_init_of_compatible(adev);
+		return;
+	}
+
+ fail:
+	dev_warn(&adev->dev, "Returned _DSD data is not valid, skipping\n");
+	ACPI_FREE(buf.pointer);
+}
+
+void acpi_free_properties(struct acpi_device *adev)
+{
+	ACPI_FREE((void *)adev->data.pointer);
+	adev->data.of_compatible = NULL;
+	adev->data.pointer = NULL;
+	adev->data.properties = NULL;
+}
+
+/**
+ * acpi_dev_get_property - return an ACPI property with given name
+ * @adev: ACPI device to get property
+ * @name: Name of the property
+ * @type: Expected property type
+ * @obj: Location to store the property value (if not %NULL)
+ *
+ * Look up a property with @name and store a pointer to the resulting ACPI
+ * object at the location pointed to by @obj if found.
+ *
+ * Callers must not attempt to free the returned objects.  These objects will be
+ * freed by the ACPI core automatically during the removal of @adev.
+ *
+ * Return: %0 if property with @name has been found (success),
+ *         %-EINVAL if the arguments are invalid,
+ *         %-ENODATA if the property doesn't exist,
+ *         %-EPROTO if the property value type doesn't match @type.
+ */
+int acpi_dev_get_property(struct acpi_device *adev, const char *name,
+			  acpi_object_type type, const union acpi_object **obj)
+{
+	const union acpi_object *properties;
+	int i;
+
+	if (!adev || !name)
+		return -EINVAL;
+
+	if (!adev->data.pointer || !adev->data.properties)
+		return -ENODATA;
+
+	properties = adev->data.properties;
+	for (i = 0; i < properties->package.count; i++) {
+		const union acpi_object *propname, *propvalue;
+		const union acpi_object *property;
+
+		property = &properties->package.elements[i];
+
+		propname = &property->package.elements[0];
+		propvalue = &property->package.elements[1];
+
+		if (!strcmp(name, propname->string.pointer)) {
+			if (type != ACPI_TYPE_ANY && propvalue->type != type)
+				return -EPROTO;
+			else if (obj)
+				*obj = propvalue;
+
+			return 0;
+		}
+	}
+	return -ENODATA;
+}
+EXPORT_SYMBOL_GPL(acpi_dev_get_property);
+
+/**
+ * acpi_dev_get_property_array - return an ACPI array property with given name
+ * @adev: ACPI device to get property
+ * @name: Name of the property
+ * @type: Expected type of array elements
+ * @obj: Location to store a pointer to the property value (if not NULL)
+ *
+ * Look up an array property with @name and store a pointer to the resulting
+ * ACPI object at the location pointed to by @obj if found.
+ *
+ * Callers must not attempt to free the returned objects.  Those objects will be
+ * freed by the ACPI core automatically during the removal of @adev.
+ *
+ * Return: %0 if array property (package) with @name has been found (success),
+ *         %-EINVAL if the arguments are invalid,
+ *         %-ENODATA if the property doesn't exist,
+ *         %-EPROTO if the property is not a package or the type of its elements
+ *           doesn't match @type.
+ */
+int acpi_dev_get_property_array(struct acpi_device *adev, const char *name,
+				acpi_object_type type,
+				const union acpi_object **obj)
+{
+	const union acpi_object *prop;
+	int ret, i;
+
+	ret = acpi_dev_get_property(adev, name, ACPI_TYPE_PACKAGE, &prop);
+	if (ret)
+		return ret;
+
+	if (type != ACPI_TYPE_ANY) {
+		/* Check that all elements are of correct type. */
+		for (i = 0; i < prop->package.count; i++)
+			if (prop->package.elements[i].type != type)
+				return -EPROTO;
+	}
+	if (obj)
+		*obj = prop;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(acpi_dev_get_property_array);
+
+/**
+ * acpi_dev_get_property_reference - returns handle to the referenced object
+ * @adev: ACPI device to get property
+ * @name: Name of the property
+ * @index: Index of the reference to return
+ * @args: Location to store the returned reference with optional arguments
+ *
+ * Find property with @name, verifify that it is a package containing at least
+ * one object reference and if so, store the ACPI device object pointer to the
+ * target object in @args->adev.  If the reference includes arguments, store
+ * them in the @args->args[] array.
+ *
+ * If there's more than one reference in the property value package, @index is
+ * used to select the one to return.
+ *
+ * Return: %0 on success, negative error code on failure.
+ */
+int acpi_dev_get_property_reference(struct acpi_device *adev,
+				    const char *name, size_t index,
+				    struct acpi_reference_args *args)
+{
+	const union acpi_object *element, *end;
+	const union acpi_object *obj;
+	struct acpi_device *device;
+	int ret, idx = 0;
+
+	ret = acpi_dev_get_property(adev, name, ACPI_TYPE_ANY, &obj);
+	if (ret)
+		return ret;
+
+	/*
+	 * The simplest case is when the value is a single reference.  Just
+	 * return that reference then.
+	 */
+	if (obj->type == ACPI_TYPE_LOCAL_REFERENCE) {
+		if (index)
+			return -EINVAL;
+
+		ret = acpi_bus_get_device(obj->reference.handle, &device);
+		if (ret)
+			return ret;
+
+		args->adev = device;
+		args->nargs = 0;
+		return 0;
+	}
+
+	/*
+	 * If it is not a single reference, then it is a package of
+	 * references followed by number of ints as follows:
+	 *
+	 *  Package () { REF, INT, REF, INT, INT }
+	 *
+	 * The index argument is then used to determine which reference
+	 * the caller wants (along with the arguments).
+	 */
+	if (obj->type != ACPI_TYPE_PACKAGE || index >= obj->package.count)
+		return -EPROTO;
+
+	element = obj->package.elements;
+	end = element + obj->package.count;
+
+	while (element < end) {
+		u32 nargs, i;
+
+		if (element->type != ACPI_TYPE_LOCAL_REFERENCE)
+			return -EPROTO;
+
+		ret = acpi_bus_get_device(element->reference.handle, &device);
+		if (ret)
+			return -ENODEV;
+
+		element++;
+		nargs = 0;
+
+		/* assume following integer elements are all args */
+		for (i = 0; element + i < end; i++) {
+			int type = element[i].type;
+
+			if (type == ACPI_TYPE_INTEGER)
+				nargs++;
+			else if (type == ACPI_TYPE_LOCAL_REFERENCE)
+				break;
+			else
+				return -EPROTO;
+		}
+
+		if (idx++ == index) {
+			args->adev = device;
+			args->nargs = nargs;
+			for (i = 0; i < nargs; i++)
+				args->args[i] = element[i].integer.value;
+
+			return 0;
+		}
+
+		element += nargs;
+	}
+
+	return -EPROTO;
+}
+EXPORT_SYMBOL_GPL(acpi_dev_get_property_reference);
+
+int acpi_dev_prop_get(struct acpi_device *adev, const char *propname,
+		      void **valptr)
+{
+	return acpi_dev_get_property(adev, propname, ACPI_TYPE_ANY,
+				     (const union acpi_object **)valptr);
+}
+
+int acpi_dev_prop_read_single(struct acpi_device *adev, const char *propname,
+			      enum dev_prop_type proptype, void *val)
+{
+	const union acpi_object *obj;
+	int ret;
+
+	if (!val)
+		return -EINVAL;
+
+	if (proptype >= DEV_PROP_U8 && proptype <= DEV_PROP_U64) {
+		ret = acpi_dev_get_property(adev, propname, ACPI_TYPE_INTEGER, &obj);
+		if (ret)
+			return ret;
+
+		switch (proptype) {
+		case DEV_PROP_U8:
+			if (obj->integer.value > U8_MAX)
+				return -EOVERFLOW;
+			*(u8 *)val = obj->integer.value;
+			break;
+		case DEV_PROP_U16:
+			if (obj->integer.value > U16_MAX)
+				return -EOVERFLOW;
+			*(u16 *)val = obj->integer.value;
+			break;
+		case DEV_PROP_U32:
+			if (obj->integer.value > U32_MAX)
+				return -EOVERFLOW;
+			*(u32 *)val = obj->integer.value;
+			break;
+		default:
+			*(u64 *)val = obj->integer.value;
+			break;
+		}
+	} else if (proptype == DEV_PROP_STRING) {
+		ret = acpi_dev_get_property(adev, propname, ACPI_TYPE_STRING, &obj);
+		if (ret)
+			return ret;
+
+		*(char **)val = obj->string.pointer;
+	} else {
+		ret = -EINVAL;
+	}
+	return ret;
+}
+
+static int acpi_copy_property_array_u8(const union acpi_object *items, u8 *val,
+				       size_t nval)
+{
+	int i;
+
+	for (i = 0; i < nval; i++) {
+		if (items[i].type != ACPI_TYPE_INTEGER)
+			return -EPROTO;
+		if (items[i].integer.value > U8_MAX)
+			return -EOVERFLOW;
+
+		val[i] = items[i].integer.value;
+	}
+	return 0;
+}
+
+static int acpi_copy_property_array_u16(const union acpi_object *items,
+					u16 *val, size_t nval)
+{
+	int i;
+
+	for (i = 0; i < nval; i++) {
+		if (items[i].type != ACPI_TYPE_INTEGER)
+			return -EPROTO;
+		if (items[i].integer.value > U16_MAX)
+			return -EOVERFLOW;
+
+		val[i] = items[i].integer.value;
+	}
+	return 0;
+}
+
+static int acpi_copy_property_array_u32(const union acpi_object *items,
+					u32 *val, size_t nval)
+{
+	int i;
+
+	for (i = 0; i < nval; i++) {
+		if (items[i].type != ACPI_TYPE_INTEGER)
+			return -EPROTO;
+		if (items[i].integer.value > U32_MAX)
+			return -EOVERFLOW;
+
+		val[i] = items[i].integer.value;
+	}
+	return 0;
+}
+
+static int acpi_copy_property_array_u64(const union acpi_object *items,
+					u64 *val, size_t nval)
+{
+	int i;
+
+	for (i = 0; i < nval; i++) {
+		if (items[i].type != ACPI_TYPE_INTEGER)
+			return -EPROTO;
+
+		val[i] = items[i].integer.value;
+	}
+	return 0;
+}
+
+static int acpi_copy_property_array_string(const union acpi_object *items,
+					   char **val, size_t nval)
+{
+	int i;
+
+	for (i = 0; i < nval; i++) {
+		if (items[i].type != ACPI_TYPE_STRING)
+			return -EPROTO;
+
+		val[i] = items[i].string.pointer;
+	}
+	return 0;
+}
+
+int acpi_dev_prop_read(struct acpi_device *adev, const char *propname,
+		       enum dev_prop_type proptype, void *val, size_t nval)
+{
+	const union acpi_object *obj;
+	const union acpi_object *items;
+	int ret;
+
+	if (val && nval == 1) {
+		ret = acpi_dev_prop_read_single(adev, propname, proptype, val);
+		if (!ret)
+			return ret;
+	}
+
+	ret = acpi_dev_get_property_array(adev, propname, ACPI_TYPE_ANY, &obj);
+	if (ret)
+		return ret;
+
+	if (!val)
+		return obj->package.count;
+	else if (nval <= 0)
+		return -EINVAL;
+
+	if (nval > obj->package.count)
+		return -EOVERFLOW;
+
+	items = obj->package.elements;
+	switch (proptype) {
+	case DEV_PROP_U8:
+		ret = acpi_copy_property_array_u8(items, (u8 *)val, nval);
+		break;
+	case DEV_PROP_U16:
+		ret = acpi_copy_property_array_u16(items, (u16 *)val, nval);
+		break;
+	case DEV_PROP_U32:
+		ret = acpi_copy_property_array_u32(items, (u32 *)val, nval);
+		break;
+	case DEV_PROP_U64:
+		ret = acpi_copy_property_array_u64(items, (u64 *)val, nval);
+		break;
+	case DEV_PROP_STRING:
+		ret = acpi_copy_property_array_string(items, (char **)val, nval);
+		break;
+	default:
+		ret = -EINVAL;
+		break;
+	}
+	return ret;
+}
diff --git a/drivers/acpi/resource.c b/drivers/acpi/resource.c
index 2ba8f02ced36..782a0d15c25f 100644
--- a/drivers/acpi/resource.c
+++ b/drivers/acpi/resource.c
@@ -200,7 +200,7 @@ bool acpi_dev_resource_address_space(struct acpi_resource *ares,
 
 	status = acpi_resource_to_address64(ares, &addr);
 	if (ACPI_FAILURE(status))
-		return true;
+		return false;
 
 	res->start = addr.minimum;
 	res->end = addr.maximum;
diff --git a/drivers/acpi/sbs.c b/drivers/acpi/sbs.c
index 366ca40a6f70..a7a3edd28beb 100644
--- a/drivers/acpi/sbs.c
+++ b/drivers/acpi/sbs.c
@@ -35,6 +35,7 @@
 #include <linux/jiffies.h>
 #include <linux/delay.h>
 #include <linux/power_supply.h>
+#include <linux/dmi.h>
 
 #include "sbshc.h"
 #include "battery.h"
@@ -61,6 +62,8 @@ static unsigned int cache_time = 1000;
 module_param(cache_time, uint, 0644);
 MODULE_PARM_DESC(cache_time, "cache time in milliseconds");
 
+static bool sbs_manager_broken;
+
 #define MAX_SBS_BAT			4
 #define ACPI_SBS_BLOCK_MAX		32
 
@@ -109,6 +112,7 @@ struct acpi_sbs {
 	u8 batteries_supported:4;
 	u8 manager_present:1;
 	u8 charger_present:1;
+	u8 charger_exists:1;
 };
 
 #define to_acpi_sbs(x) container_of(x, struct acpi_sbs, charger)
@@ -429,9 +433,19 @@ static int acpi_ac_get_present(struct acpi_sbs *sbs)
 
 	result = acpi_smbus_read(sbs->hc, SMBUS_READ_WORD, ACPI_SBS_CHARGER,
 				 0x13, (u8 *) & status);
-	if (!result)
-		sbs->charger_present = (status >> 15) & 0x1;
-	return result;
+
+	if (result)
+		return result;
+
+	/*
+	 * The spec requires that bit 4 always be 1. If it's not set, assume
+	 * that the implementation doesn't support an SBS charger
+	 */
+	if (!((status >> 4) & 0x1))
+		return -ENODEV;
+
+	sbs->charger_present = (status >> 15) & 0x1;
+	return 0;
 }
 
 static ssize_t acpi_battery_alarm_show(struct device *dev,
@@ -483,16 +497,21 @@ static int acpi_battery_read(struct acpi_battery *battery)
 				  ACPI_SBS_MANAGER, 0x01, (u8 *)&state, 2);
 	} else if (battery->id == 0)
 		battery->present = 1;
+
 	if (result || !battery->present)
 		return result;
 
 	if (saved_present != battery->present) {
 		battery->update_time = 0;
 		result = acpi_battery_get_info(battery);
-		if (result)
+		if (result) {
+			battery->present = 0;
 			return result;
+		}
 	}
 	result = acpi_battery_get_state(battery);
+	if (result)
+		battery->present = 0;
 	return result;
 }
 
@@ -524,6 +543,7 @@ static int acpi_battery_add(struct acpi_sbs *sbs, int id)
 	result = power_supply_register(&sbs->device->dev, &battery->bat);
 	if (result)
 		goto end;
+
 	result = device_create_file(battery->bat.dev, &alarm_attr);
 	if (result)
 		goto end;
@@ -554,6 +574,7 @@ static int acpi_charger_add(struct acpi_sbs *sbs)
 	if (result)
 		goto end;
 
+	sbs->charger_exists = 1;
 	sbs->charger.name = "sbs-charger";
 	sbs->charger.type = POWER_SUPPLY_TYPE_MAINS;
 	sbs->charger.properties = sbs_ac_props;
@@ -580,9 +601,12 @@ static void acpi_sbs_callback(void *context)
 	struct acpi_battery *bat;
 	u8 saved_charger_state = sbs->charger_present;
 	u8 saved_battery_state;
-	acpi_ac_get_present(sbs);
-	if (sbs->charger_present != saved_charger_state)
-		kobject_uevent(&sbs->charger.dev->kobj, KOBJ_CHANGE);
+
+	if (sbs->charger_exists) {
+		acpi_ac_get_present(sbs);
+		if (sbs->charger_present != saved_charger_state)
+			kobject_uevent(&sbs->charger.dev->kobj, KOBJ_CHANGE);
+	}
 
 	if (sbs->manager_present) {
 		for (id = 0; id < MAX_SBS_BAT; ++id) {
@@ -598,12 +622,31 @@ static void acpi_sbs_callback(void *context)
 	}
 }
 
+static int disable_sbs_manager(const struct dmi_system_id *d)
+{
+	sbs_manager_broken = true;
+	return 0;
+}
+
+static struct dmi_system_id acpi_sbs_dmi_table[] = {
+	{
+		.callback = disable_sbs_manager,
+		.ident = "Apple",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc.")
+		},
+	},
+	{ },
+};
+
 static int acpi_sbs_add(struct acpi_device *device)
 {
 	struct acpi_sbs *sbs;
 	int result = 0;
 	int id;
 
+	dmi_check_system(acpi_sbs_dmi_table);
+
 	sbs = kzalloc(sizeof(struct acpi_sbs), GFP_KERNEL);
 	if (!sbs) {
 		result = -ENOMEM;
@@ -619,17 +662,24 @@ static int acpi_sbs_add(struct acpi_device *device)
 	device->driver_data = sbs;
 
 	result = acpi_charger_add(sbs);
-	if (result)
+	if (result && result != -ENODEV)
 		goto end;
 
-	result = acpi_manager_get_info(sbs);
-	if (!result) {
-		sbs->manager_present = 1;
-		for (id = 0; id < MAX_SBS_BAT; ++id)
-			if ((sbs->batteries_supported & (1 << id)))
-				acpi_battery_add(sbs, id);
-	} else
+	result = 0;
+
+	if (!sbs_manager_broken) {
+		result = acpi_manager_get_info(sbs);
+		if (!result) {
+			sbs->manager_present = 0;
+			for (id = 0; id < MAX_SBS_BAT; ++id)
+				if ((sbs->batteries_supported & (1 << id)))
+					acpi_battery_add(sbs, id);
+		}
+	}
+
+	if (!sbs->manager_present)
 		acpi_battery_add(sbs, 0);
+
 	acpi_smbus_register_callback(sbs->hc, acpi_sbs_callback, sbs);
       end:
 	if (result)
diff --git a/drivers/acpi/scan.c b/drivers/acpi/scan.c
index ae44d8654c82..dc4d8960684a 100644
--- a/drivers/acpi/scan.c
+++ b/drivers/acpi/scan.c
@@ -36,6 +36,8 @@ bool acpi_force_hot_remove;
 
 static const char *dummy_hid = "device";
 
+static LIST_HEAD(acpi_dep_list);
+static DEFINE_MUTEX(acpi_dep_list_lock);
 static LIST_HEAD(acpi_bus_id_list);
 static DEFINE_MUTEX(acpi_scan_lock);
 static LIST_HEAD(acpi_scan_handlers_list);
@@ -43,6 +45,12 @@ DEFINE_MUTEX(acpi_device_lock);
 LIST_HEAD(acpi_wakeup_device_list);
 static DEFINE_MUTEX(acpi_hp_context_lock);
 
+struct acpi_dep_data {
+	struct list_head node;
+	acpi_handle master;
+	acpi_handle slave;
+};
+
 struct acpi_device_bus_id{
 	char bus_id[15];
 	unsigned int instance_no;
@@ -124,17 +132,56 @@ static int create_modalias(struct acpi_device *acpi_dev, char *modalias,
 	if (list_empty(&acpi_dev->pnp.ids))
 		return 0;
 
-	len = snprintf(modalias, size, "acpi:");
-	size -= len;
-
-	list_for_each_entry(id, &acpi_dev->pnp.ids, list) {
-		count = snprintf(&modalias[len], size, "%s:", id->id);
-		if (count < 0)
-			return -EINVAL;
-		if (count >= size)
-			return -ENOMEM;
-		len += count;
-		size -= count;
+	/*
+	 * If the device has PRP0001 we expose DT compatible modalias
+	 * instead in form of of:NnameTCcompatible.
+	 */
+	if (acpi_dev->data.of_compatible) {
+		struct acpi_buffer buf = { ACPI_ALLOCATE_BUFFER };
+		const union acpi_object *of_compatible, *obj;
+		int i, nval;
+		char *c;
+
+		acpi_get_name(acpi_dev->handle, ACPI_SINGLE_NAME, &buf);
+		/* DT strings are all in lower case */
+		for (c = buf.pointer; *c != '\0'; c++)
+			*c = tolower(*c);
+
+		len = snprintf(modalias, size, "of:N%sT", (char *)buf.pointer);
+		ACPI_FREE(buf.pointer);
+
+		of_compatible = acpi_dev->data.of_compatible;
+		if (of_compatible->type == ACPI_TYPE_PACKAGE) {
+			nval = of_compatible->package.count;
+			obj = of_compatible->package.elements;
+		} else { /* Must be ACPI_TYPE_STRING. */
+			nval = 1;
+			obj = of_compatible;
+		}
+		for (i = 0; i < nval; i++, obj++) {
+			count = snprintf(&modalias[len], size, "C%s",
+					 obj->string.pointer);
+			if (count < 0)
+				return -EINVAL;
+			if (count >= size)
+				return -ENOMEM;
+
+			len += count;
+			size -= count;
+		}
+	} else {
+		len = snprintf(modalias, size, "acpi:");
+		size -= len;
+
+		list_for_each_entry(id, &acpi_dev->pnp.ids, list) {
+			count = snprintf(&modalias[len], size, "%s:", id->id);
+			if (count < 0)
+				return -EINVAL;
+			if (count >= size)
+				return -ENOMEM;
+			len += count;
+			size -= count;
+		}
 	}
 
 	modalias[len] = '\0';
@@ -142,6 +189,53 @@ static int create_modalias(struct acpi_device *acpi_dev, char *modalias,
 }
 
 /*
+ * acpi_companion_match() - Can we match via ACPI companion device
+ * @dev: Device in question
+ *
+ * Check if the given device has an ACPI companion and if that companion has
+ * a valid list of PNP IDs, and if the device is the first (primary) physical
+ * device associated with it.
+ *
+ * If multiple physical devices are attached to a single ACPI companion, we need
+ * to be careful.  The usage scenario for this kind of relationship is that all
+ * of the physical devices in question use resources provided by the ACPI
+ * companion.  A typical case is an MFD device where all the sub-devices share
+ * the parent's ACPI companion.  In such cases we can only allow the primary
+ * (first) physical device to be matched with the help of the companion's PNP
+ * IDs.
+ *
+ * Additional physical devices sharing the ACPI companion can still use
+ * resources available from it but they will be matched normally using functions
+ * provided by their bus types (and analogously for their modalias).
+ */
+static bool acpi_companion_match(const struct device *dev)
+{
+	struct acpi_device *adev;
+	bool ret;
+
+	adev = ACPI_COMPANION(dev);
+	if (!adev)
+		return false;
+
+	if (list_empty(&adev->pnp.ids))
+		return false;
+
+	mutex_lock(&adev->physical_node_lock);
+	if (list_empty(&adev->physical_node_list)) {
+		ret = false;
+	} else {
+		const struct acpi_device_physical_node *node;
+
+		node = list_first_entry(&adev->physical_node_list,
+					struct acpi_device_physical_node, node);
+		ret = node->dev == dev;
+	}
+	mutex_unlock(&adev->physical_node_lock);
+
+	return ret;
+}
+
+/*
  * Creates uevent modalias field for ACPI enumerated devices.
  * Because the other buses does not support ACPI HIDs & CIDs.
  * e.g. for a device with hid:IBM0001 and cid:ACPI0001 you get:
@@ -149,20 +243,14 @@ static int create_modalias(struct acpi_device *acpi_dev, char *modalias,
  */
 int acpi_device_uevent_modalias(struct device *dev, struct kobj_uevent_env *env)
 {
-	struct acpi_device *acpi_dev;
 	int len;
 
-	acpi_dev = ACPI_COMPANION(dev);
-	if (!acpi_dev)
-		return -ENODEV;
-
-	/* Fall back to bus specific way of modalias exporting */
-	if (list_empty(&acpi_dev->pnp.ids))
+	if (!acpi_companion_match(dev))
 		return -ENODEV;
 
 	if (add_uevent_var(env, "MODALIAS="))
 		return -ENOMEM;
-	len = create_modalias(acpi_dev, &env->buf[env->buflen - 1],
+	len = create_modalias(ACPI_COMPANION(dev), &env->buf[env->buflen - 1],
 				sizeof(env->buf) - env->buflen);
 	if (len <= 0)
 		return len;
@@ -179,18 +267,12 @@ EXPORT_SYMBOL_GPL(acpi_device_uevent_modalias);
  */
 int acpi_device_modalias(struct device *dev, char *buf, int size)
 {
-	struct acpi_device *acpi_dev;
 	int len;
 
-	acpi_dev = ACPI_COMPANION(dev);
-	if (!acpi_dev)
-		return -ENODEV;
-
-	/* Fall back to bus specific way of modalias exporting */
-	if (list_empty(&acpi_dev->pnp.ids))
+	if (!acpi_companion_match(dev))
 		return -ENODEV;
 
-	len = create_modalias(acpi_dev, buf, size -1);
+	len = create_modalias(ACPI_COMPANION(dev), buf, size -1);
 	if (len <= 0)
 		return len;
 	buf[len++] = '\n';
@@ -853,6 +935,9 @@ const struct acpi_device_id *acpi_match_device(const struct acpi_device_id *ids,
 	if (!ids || !handle || acpi_bus_get_device(handle, &adev))
 		return NULL;
 
+	if (!acpi_companion_match(dev))
+		return NULL;
+
 	return __acpi_match_device(adev, ids);
 }
 EXPORT_SYMBOL_GPL(acpi_match_device);
@@ -864,6 +949,51 @@ int acpi_match_device_ids(struct acpi_device *device,
 }
 EXPORT_SYMBOL(acpi_match_device_ids);
 
+/* Performs match against special "PRP0001" shoehorn ACPI ID */
+static bool acpi_of_driver_match_device(struct device *dev,
+					const struct device_driver *drv)
+{
+	const union acpi_object *of_compatible, *obj;
+	struct acpi_device *adev;
+	int i, nval;
+
+	adev = ACPI_COMPANION(dev);
+	if (!adev)
+		return false;
+
+	of_compatible = adev->data.of_compatible;
+	if (!drv->of_match_table || !of_compatible)
+		return false;
+
+	if (of_compatible->type == ACPI_TYPE_PACKAGE) {
+		nval = of_compatible->package.count;
+		obj = of_compatible->package.elements;
+	} else { /* Must be ACPI_TYPE_STRING. */
+		nval = 1;
+		obj = of_compatible;
+	}
+	/* Now we can look for the driver DT compatible strings */
+	for (i = 0; i < nval; i++, obj++) {
+		const struct of_device_id *id;
+
+		for (id = drv->of_match_table; id->compatible[0]; id++)
+			if (!strcasecmp(obj->string.pointer, id->compatible))
+				return true;
+	}
+
+	return false;
+}
+
+bool acpi_driver_match_device(struct device *dev,
+			      const struct device_driver *drv)
+{
+	if (!drv->acpi_match_table)
+		return acpi_of_driver_match_device(dev, drv);
+
+	return !!acpi_match_device(drv->acpi_match_table, dev);
+}
+EXPORT_SYMBOL_GPL(acpi_driver_match_device);
+
 static void acpi_free_power_resources_lists(struct acpi_device *device)
 {
 	int i;
@@ -871,7 +1001,7 @@ static void acpi_free_power_resources_lists(struct acpi_device *device)
 	if (device->wakeup.flags.valid)
 		acpi_power_resources_list_free(&device->wakeup.resources);
 
-	if (!device->flags.power_manageable)
+	if (!device->power.flags.power_resources)
 		return;
 
 	for (i = ACPI_STATE_D0; i <= ACPI_STATE_D3_HOT; i++) {
@@ -884,6 +1014,7 @@ static void acpi_device_release(struct device *dev)
 {
 	struct acpi_device *acpi_dev = to_acpi_device(dev);
 
+	acpi_free_properties(acpi_dev);
 	acpi_free_pnp_ids(&acpi_dev->pnp);
 	acpi_free_power_resources_lists(acpi_dev);
 	kfree(acpi_dev);
@@ -1266,6 +1397,26 @@ int acpi_device_add(struct acpi_device *device,
 	return result;
 }
 
+struct acpi_device *acpi_get_next_child(struct device *dev,
+					struct acpi_device *child)
+{
+	struct acpi_device *adev = ACPI_COMPANION(dev);
+	struct list_head *head, *next;
+
+	if (!adev)
+		return NULL;
+
+	head = &adev->children;
+	if (list_empty(head))
+		return NULL;
+
+	if (!child)
+		return list_first_entry(head, struct acpi_device, node);
+
+	next = child->node.next;
+	return next == head ? NULL : list_entry(next, struct acpi_device, node);
+}
+
 /* --------------------------------------------------------------------------
                                  Driver Management
    -------------------------------------------------------------------------- */
@@ -1470,7 +1621,7 @@ static void acpi_wakeup_gpe_init(struct acpi_device *device)
 	if (ACPI_FAILURE(status))
 		return;
 
-	wakeup->flags.run_wake = !!(event_status & ACPI_EVENT_FLAG_HANDLE);
+	wakeup->flags.run_wake = !!(event_status & ACPI_EVENT_FLAG_HAS_HANDLER);
 }
 
 static void acpi_bus_get_wakeup_device_flags(struct acpi_device *device)
@@ -1593,10 +1744,8 @@ static void acpi_bus_get_power_flags(struct acpi_device *device)
 			device->power.flags.power_resources)
 		device->power.states[ACPI_STATE_D3_COLD].flags.os_accessible = 1;
 
-	if (acpi_bus_init_power(device)) {
-		acpi_free_power_resources_lists(device);
+	if (acpi_bus_init_power(device))
 		device->flags.power_manageable = 0;
-	}
 }
 
 static void acpi_bus_get_flags(struct acpi_device *device)
@@ -1885,9 +2034,11 @@ void acpi_init_device_object(struct acpi_device *device, acpi_handle handle,
 	device->device_type = type;
 	device->handle = handle;
 	device->parent = acpi_bus_get_parent(handle);
+	device->fwnode.type = FWNODE_ACPI;
 	acpi_set_device_status(device, sta);
 	acpi_device_get_busid(device);
 	acpi_set_pnp_ids(handle, &device->pnp, type);
+	acpi_init_properties(device);
 	acpi_bus_get_flags(device);
 	device->flags.match_driver = false;
 	device->flags.initialized = true;
@@ -2048,6 +2199,59 @@ static void acpi_scan_init_hotplug(struct acpi_device *adev)
 	}
 }
 
+static void acpi_device_dep_initialize(struct acpi_device *adev)
+{
+	struct acpi_dep_data *dep;
+	struct acpi_handle_list dep_devices;
+	acpi_status status;
+	int i;
+
+	if (!acpi_has_method(adev->handle, "_DEP"))
+		return;
+
+	status = acpi_evaluate_reference(adev->handle, "_DEP", NULL,
+					&dep_devices);
+	if (ACPI_FAILURE(status)) {
+		dev_dbg(&adev->dev, "Failed to evaluate _DEP.\n");
+		return;
+	}
+
+	for (i = 0; i < dep_devices.count; i++) {
+		struct acpi_device_info *info;
+		int skip;
+
+		status = acpi_get_object_info(dep_devices.handles[i], &info);
+		if (ACPI_FAILURE(status)) {
+			dev_dbg(&adev->dev, "Error reading _DEP device info\n");
+			continue;
+		}
+
+		/*
+		 * Skip the dependency of Windows System Power
+		 * Management Controller
+		 */
+		skip = info->valid & ACPI_VALID_HID &&
+			!strcmp(info->hardware_id.string, "INT3396");
+
+		kfree(info);
+
+		if (skip)
+			continue;
+
+		dep = kzalloc(sizeof(struct acpi_dep_data), GFP_KERNEL);
+		if (!dep)
+			return;
+
+		dep->master = dep_devices.handles[i];
+		dep->slave  = adev->handle;
+		adev->dep_unmet++;
+
+		mutex_lock(&acpi_dep_list_lock);
+		list_add_tail(&dep->node , &acpi_dep_list);
+		mutex_unlock(&acpi_dep_list_lock);
+	}
+}
+
 static acpi_status acpi_bus_check_add(acpi_handle handle, u32 lvl_not_used,
 				      void *not_used, void **return_value)
 {
@@ -2074,6 +2278,7 @@ static acpi_status acpi_bus_check_add(acpi_handle handle, u32 lvl_not_used,
 		return AE_CTRL_DEPTH;
 
 	acpi_scan_init_hotplug(device);
+	acpi_device_dep_initialize(device);
 
  out:
 	if (!*return_value)
@@ -2164,13 +2369,18 @@ static void acpi_bus_attach(struct acpi_device *device)
 	/* Skip devices that are not present. */
 	if (!acpi_device_is_present(device)) {
 		device->flags.visited = false;
+		device->flags.power_manageable = 0;
 		return;
 	}
 	if (device->handler)
 		goto ok;
 
 	if (!device->flags.initialized) {
-		acpi_bus_update_power(device, NULL);
+		device->flags.power_manageable =
+			device->power.states[ACPI_STATE_D0].flags.valid;
+		if (acpi_bus_init_power(device))
+			device->flags.power_manageable = 0;
+
 		device->flags.initialized = true;
 	}
 	device->flags.visited = false;
@@ -2194,6 +2404,29 @@ static void acpi_bus_attach(struct acpi_device *device)
 		device->handler->hotplug.notify_online(device);
 }
 
+void acpi_walk_dep_device_list(acpi_handle handle)
+{
+	struct acpi_dep_data *dep, *tmp;
+	struct acpi_device *adev;
+
+	mutex_lock(&acpi_dep_list_lock);
+	list_for_each_entry_safe(dep, tmp, &acpi_dep_list, node) {
+		if (dep->master == handle) {
+			acpi_bus_get_device(dep->slave, &adev);
+			if (!adev)
+				continue;
+
+			adev->dep_unmet--;
+			if (!adev->dep_unmet)
+				acpi_bus_attach(adev);
+			list_del(&dep->node);
+			kfree(dep);
+		}
+	}
+	mutex_unlock(&acpi_dep_list_lock);
+}
+EXPORT_SYMBOL_GPL(acpi_walk_dep_device_list);
+
 /**
  * acpi_bus_scan - Add ACPI device node objects in a given namespace scope.
  * @handle: Root of the namespace scope to scan.
@@ -2315,6 +2548,7 @@ int __init acpi_scan_init(void)
 	acpi_container_init();
 	acpi_memory_hotplug_init();
 	acpi_pnp_init();
+	acpi_int340x_thermal_init();
 
 	mutex_lock(&acpi_scan_lock);
 	/*
diff --git a/drivers/acpi/sleep.c b/drivers/acpi/sleep.c
index 54da4a3fe65e..8aa9254a387f 100644
--- a/drivers/acpi/sleep.c
+++ b/drivers/acpi/sleep.c
@@ -14,6 +14,7 @@
 #include <linux/irq.h>
 #include <linux/dmi.h>
 #include <linux/device.h>
+#include <linux/interrupt.h>
 #include <linux/suspend.h>
 #include <linux/reboot.h>
 #include <linux/acpi.h>
@@ -626,6 +627,20 @@ static int acpi_freeze_begin(void)
 	return 0;
 }
 
+static int acpi_freeze_prepare(void)
+{
+	acpi_enable_all_wakeup_gpes();
+	acpi_os_wait_events_complete();
+	enable_irq_wake(acpi_gbl_FADT.sci_interrupt);
+	return 0;
+}
+
+static void acpi_freeze_restore(void)
+{
+	disable_irq_wake(acpi_gbl_FADT.sci_interrupt);
+	acpi_enable_all_runtime_gpes();
+}
+
 static void acpi_freeze_end(void)
 {
 	acpi_scan_lock_release();
@@ -633,6 +648,8 @@ static void acpi_freeze_end(void)
 
 static const struct platform_freeze_ops acpi_freeze_ops = {
 	.begin = acpi_freeze_begin,
+	.prepare = acpi_freeze_prepare,
+	.restore = acpi_freeze_restore,
 	.end = acpi_freeze_end,
 };
 
@@ -809,6 +826,7 @@ static void acpi_power_off_prepare(void)
 	/* Prepare to power off the system */
 	acpi_sleep_prepare(ACPI_STATE_S5);
 	acpi_disable_all_gpes();
+	acpi_os_wait_events_complete();
 }
 
 static void acpi_power_off(void)
diff --git a/drivers/acpi/sysfs.c b/drivers/acpi/sysfs.c
index 38cb9782d4b8..13e577c80201 100644
--- a/drivers/acpi/sysfs.c
+++ b/drivers/acpi/sysfs.c
@@ -537,7 +537,7 @@ static ssize_t counter_show(struct kobject *kobj,
 	if (result)
 		goto end;
 
-	if (!(status & ACPI_EVENT_FLAG_HANDLE))
+	if (!(status & ACPI_EVENT_FLAG_HAS_HANDLER))
 		size += sprintf(buf + size, "   invalid");
 	else if (status & ACPI_EVENT_FLAG_ENABLED)
 		size += sprintf(buf + size, "   enabled");
@@ -581,7 +581,7 @@ static ssize_t counter_set(struct kobject *kobj,
 	if (result)
 		goto end;
 
-	if (!(status & ACPI_EVENT_FLAG_HANDLE)) {
+	if (!(status & ACPI_EVENT_FLAG_HAS_HANDLER)) {
 		printk(KERN_WARNING PREFIX
 		       "Can not change Invalid GPE/Fixed Event status\n");
 		return -EINVAL;
diff --git a/drivers/acpi/tables.c b/drivers/acpi/tables.c
index 6d5a6cda0734..93b81523a2fe 100644
--- a/drivers/acpi/tables.c
+++ b/drivers/acpi/tables.c
@@ -190,30 +190,24 @@ void acpi_table_print_madt_entry(struct acpi_subtable_header *header)
 	}
 }
 
-
 int __init
-acpi_table_parse_entries(char *id,
-			     unsigned long table_size,
-			     int entry_id,
-			     acpi_tbl_entry_handler handler,
-			     unsigned int max_entries)
+acpi_parse_entries(char *id, unsigned long table_size,
+		acpi_tbl_entry_handler handler,
+		struct acpi_table_header *table_header,
+		int entry_id, unsigned int max_entries)
 {
-	struct acpi_table_header *table_header = NULL;
 	struct acpi_subtable_header *entry;
-	unsigned int count = 0;
+	int count = 0;
 	unsigned long table_end;
-	acpi_size tbl_size;
 
 	if (acpi_disabled)
 		return -ENODEV;
 
-	if (!handler)
+	if (!id || !handler)
 		return -EINVAL;
 
-	if (strncmp(id, ACPI_SIG_MADT, 4) == 0)
-		acpi_get_table_with_size(id, acpi_apic_instance, &table_header, &tbl_size);
-	else
-		acpi_get_table_with_size(id, 0, &table_header, &tbl_size);
+	if (!table_size)
+		return -EINVAL;
 
 	if (!table_header) {
 		pr_warn("%4.4s not present\n", id);
@@ -230,9 +224,12 @@ acpi_table_parse_entries(char *id,
 	while (((unsigned long)entry) + sizeof(struct acpi_subtable_header) <
 	       table_end) {
 		if (entry->type == entry_id
-		    && (!max_entries || count++ < max_entries))
+		    && (!max_entries || count < max_entries)) {
 			if (handler(entry, table_end))
-				goto err;
+				return -EINVAL;
+
+			count++;
+		}
 
 		/*
 		 * If entry->length is 0, break from this loop to avoid
@@ -240,22 +237,53 @@ acpi_table_parse_entries(char *id,
 		 */
 		if (entry->length == 0) {
 			pr_err("[%4.4s:0x%02x] Invalid zero length\n", id, entry_id);
-			goto err;
+			return -EINVAL;
 		}
 
 		entry = (struct acpi_subtable_header *)
 		    ((unsigned long)entry + entry->length);
 	}
+
 	if (max_entries && count > max_entries) {
 		pr_warn("[%4.4s:0x%02x] ignored %i entries of %i found\n",
 			id, entry_id, count - max_entries, count);
 	}
 
-	early_acpi_os_unmap_memory((char *)table_header, tbl_size);
 	return count;
-err:
+}
+
+int __init
+acpi_table_parse_entries(char *id,
+			 unsigned long table_size,
+			 int entry_id,
+			 acpi_tbl_entry_handler handler,
+			 unsigned int max_entries)
+{
+	struct acpi_table_header *table_header = NULL;
+	acpi_size tbl_size;
+	int count;
+	u32 instance = 0;
+
+	if (acpi_disabled)
+		return -ENODEV;
+
+	if (!id || !handler)
+		return -EINVAL;
+
+	if (!strncmp(id, ACPI_SIG_MADT, 4))
+		instance = acpi_apic_instance;
+
+	acpi_get_table_with_size(id, instance, &table_header, &tbl_size);
+	if (!table_header) {
+		pr_warn("%4.4s not present\n", id);
+		return -ENODEV;
+	}
+
+	count = acpi_parse_entries(id, table_size, handler, table_header,
+			entry_id, max_entries);
+
 	early_acpi_os_unmap_memory((char *)table_header, tbl_size);
-	return -EINVAL;
+	return count;
 }
 
 int __init
diff --git a/drivers/acpi/thermal.c b/drivers/acpi/thermal.c
index 112817e963e0..d24fa1964eb8 100644
--- a/drivers/acpi/thermal.c
+++ b/drivers/acpi/thermal.c
@@ -528,7 +528,6 @@ static void acpi_thermal_check(void *data)
 }
 
 /* sys I/F for generic thermal sysfs support */
-#define KELVIN_TO_MILLICELSIUS(t, off) (((t) - (off)) * 100)
 
 static int thermal_get_temp(struct thermal_zone_device *thermal,
 			    unsigned long *temp)
@@ -543,7 +542,8 @@ static int thermal_get_temp(struct thermal_zone_device *thermal,
 	if (result)
 		return result;
 
-	*temp = KELVIN_TO_MILLICELSIUS(tz->temperature, tz->kelvin_offset);
+	*temp = DECI_KELVIN_TO_MILLICELSIUS_WITH_OFFSET(tz->temperature,
+							tz->kelvin_offset);
 	return 0;
 }
 
@@ -647,7 +647,7 @@ static int thermal_get_trip_temp(struct thermal_zone_device *thermal,
 
 	if (tz->trips.critical.flags.valid) {
 		if (!trip) {
-			*temp = KELVIN_TO_MILLICELSIUS(
+			*temp = DECI_KELVIN_TO_MILLICELSIUS_WITH_OFFSET(
 				tz->trips.critical.temperature,
 				tz->kelvin_offset);
 			return 0;
@@ -657,7 +657,7 @@ static int thermal_get_trip_temp(struct thermal_zone_device *thermal,
 
 	if (tz->trips.hot.flags.valid) {
 		if (!trip) {
-			*temp = KELVIN_TO_MILLICELSIUS(
+			*temp = DECI_KELVIN_TO_MILLICELSIUS_WITH_OFFSET(
 				tz->trips.hot.temperature,
 				tz->kelvin_offset);
 			return 0;
@@ -667,7 +667,7 @@ static int thermal_get_trip_temp(struct thermal_zone_device *thermal,
 
 	if (tz->trips.passive.flags.valid) {
 		if (!trip) {
-			*temp = KELVIN_TO_MILLICELSIUS(
+			*temp = DECI_KELVIN_TO_MILLICELSIUS_WITH_OFFSET(
 				tz->trips.passive.temperature,
 				tz->kelvin_offset);
 			return 0;
@@ -678,7 +678,7 @@ static int thermal_get_trip_temp(struct thermal_zone_device *thermal,
 	for (i = 0; i < ACPI_THERMAL_MAX_ACTIVE &&
 		tz->trips.active[i].flags.valid; i++) {
 		if (!trip) {
-			*temp = KELVIN_TO_MILLICELSIUS(
+			*temp = DECI_KELVIN_TO_MILLICELSIUS_WITH_OFFSET(
 				tz->trips.active[i].temperature,
 				tz->kelvin_offset);
 			return 0;
@@ -694,7 +694,7 @@ static int thermal_get_crit_temp(struct thermal_zone_device *thermal,
 	struct acpi_thermal *tz = thermal->devdata;
 
 	if (tz->trips.critical.flags.valid) {
-		*temperature = KELVIN_TO_MILLICELSIUS(
+		*temperature = DECI_KELVIN_TO_MILLICELSIUS_WITH_OFFSET(
 				tz->trips.critical.temperature,
 				tz->kelvin_offset);
 		return 0;
@@ -714,8 +714,8 @@ static int thermal_get_trend(struct thermal_zone_device *thermal,
 
 	if (type == THERMAL_TRIP_ACTIVE) {
 		unsigned long trip_temp;
-		unsigned long temp = KELVIN_TO_MILLICELSIUS(tz->temperature,
-							tz->kelvin_offset);
+		unsigned long temp = DECI_KELVIN_TO_MILLICELSIUS_WITH_OFFSET(
+					tz->temperature, tz->kelvin_offset);
 		if (thermal_get_trip_temp(thermal, trip, &trip_temp))
 			return -EINVAL;
 
diff --git a/drivers/acpi/utils.c b/drivers/acpi/utils.c
index 07c8c5a5ee95..cd49a3982b6a 100644
--- a/drivers/acpi/utils.c
+++ b/drivers/acpi/utils.c
@@ -136,8 +136,7 @@ acpi_extract_package(union acpi_object *package,
 				break;
 			case 'B':
 				size_required +=
-				    sizeof(u8 *) +
-				    (element->buffer.length * sizeof(u8));
+				    sizeof(u8 *) + element->buffer.length;
 				tail_offset += sizeof(u8 *);
 				break;
 			default:
@@ -149,6 +148,21 @@ acpi_extract_package(union acpi_object *package,
 				break;
 			}
 			break;
+		case ACPI_TYPE_LOCAL_REFERENCE:
+			switch (format_string[i]) {
+			case 'R':
+				size_required += sizeof(void *);
+				tail_offset += sizeof(void *);
+				break;
+			default:
+				printk(KERN_WARNING PREFIX "Invalid package element"
+					      " [%d] got reference,"
+					      " expecting [%c]\n",
+					      i, format_string[i]);
+				return AE_BAD_DATA;
+				break;
+			}
+			break;
 
 		case ACPI_TYPE_PACKAGE:
 		default:
@@ -240,14 +254,25 @@ acpi_extract_package(union acpi_object *package,
 				memcpy(tail, element->buffer.pointer,
 				       element->buffer.length);
 				head += sizeof(u8 *);
-				tail += element->buffer.length * sizeof(u8);
+				tail += element->buffer.length;
+				break;
+			default:
+				/* Should never get here */
+				break;
+			}
+			break;
+		case ACPI_TYPE_LOCAL_REFERENCE:
+			switch (format_string[i]) {
+			case 'R':
+				*(void **)head =
+				    (void *)element->reference.handle;
+				head += sizeof(void *);
 				break;
 			default:
 				/* Should never get here */
 				break;
 			}
 			break;
-
 		case ACPI_TYPE_PACKAGE:
 			/* TBD: handle nested packages... */
 		default:
@@ -321,22 +346,16 @@ acpi_evaluate_reference(acpi_handle handle,
 	package = buffer.pointer;
 
 	if ((buffer.length == 0) || !package) {
-		printk(KERN_ERR PREFIX "No return object (len %X ptr %p)\n",
-			    (unsigned)buffer.length, package);
 		status = AE_BAD_DATA;
 		acpi_util_eval_error(handle, pathname, status);
 		goto end;
 	}
 	if (package->type != ACPI_TYPE_PACKAGE) {
-		printk(KERN_ERR PREFIX "Expecting a [Package], found type %X\n",
-			    package->type);
 		status = AE_BAD_DATA;
 		acpi_util_eval_error(handle, pathname, status);
 		goto end;
 	}
 	if (!package->package.count) {
-		printk(KERN_ERR PREFIX "[Package] has zero elements (%p)\n",
-			    package);
 		status = AE_BAD_DATA;
 		acpi_util_eval_error(handle, pathname, status);
 		goto end;
@@ -355,17 +374,13 @@ acpi_evaluate_reference(acpi_handle handle,
 
 		if (element->type != ACPI_TYPE_LOCAL_REFERENCE) {
 			status = AE_BAD_DATA;
-			printk(KERN_ERR PREFIX
-				    "Expecting a [Reference] package element, found type %X\n",
-				    element->type);
 			acpi_util_eval_error(handle, pathname, status);
 			break;
 		}
 
 		if (!element->reference.handle) {
-			printk(KERN_WARNING PREFIX "Invalid reference in"
-			       " package %s\n", pathname);
 			status = AE_NULL_ENTRY;
+			acpi_util_eval_error(handle, pathname, status);
 			break;
 		}
 		/* Get the  acpi_handle. */
@@ -661,7 +676,6 @@ EXPORT_SYMBOL(acpi_evaluate_dsm);
  * @uuid: UUID of requested functions, should be 16 bytes at least
  * @rev: revision number of requested functions
  * @funcs: bitmap of requested functions
- * @exclude: excluding special value, used to support i915 and nouveau
  *
  * Evaluate device's _DSM method to check whether it supports requested
  * functions. Currently only support 64 functions at maximum, should be
diff --git a/drivers/acpi/video.c b/drivers/acpi/video.c
index 8e7e18567ae6..032db459370f 100644
--- a/drivers/acpi/video.c
+++ b/drivers/acpi/video.c
@@ -155,6 +155,7 @@ struct acpi_video_bus {
 	u8 dos_setting;
 	struct acpi_video_enumerated_device *attached_array;
 	u8 attached_count;
+	u8 child_count;
 	struct acpi_video_bus_cap cap;
 	struct acpi_video_bus_flags flags;
 	struct list_head video_device_list;
@@ -411,12 +412,6 @@ static int __init video_set_bqc_offset(const struct dmi_system_id *d)
 	return 0;
 }
 
-static int __init video_set_use_native_backlight(const struct dmi_system_id *d)
-{
-	use_native_backlight_dmi = true;
-	return 0;
-}
-
 static int __init video_disable_native_backlight(const struct dmi_system_id *d)
 {
 	use_native_backlight_dmi = false;
@@ -467,265 +462,6 @@ static struct dmi_system_id video_dmi_table[] __initdata = {
 		DMI_MATCH(DMI_PRODUCT_NAME, "Aspire 7720"),
 		},
 	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "ThinkPad X230",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-		DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad X230"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "ThinkPad T430 and T430s",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-		DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad T430"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "ThinkPad T430",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-		DMI_MATCH(DMI_PRODUCT_VERSION, "2349D15"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "ThinkPad T431s",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-		DMI_MATCH(DMI_PRODUCT_VERSION, "20AACTO1WW"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "ThinkPad Edge E530",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-		DMI_MATCH(DMI_PRODUCT_VERSION, "3259A2G"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "ThinkPad Edge E530",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-		DMI_MATCH(DMI_PRODUCT_VERSION, "3259CTO"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "ThinkPad Edge E530",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-		DMI_MATCH(DMI_PRODUCT_VERSION, "3259HJG"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "ThinkPad W530",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-		DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad W530"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	.ident = "ThinkPad X1 Carbon",
-	.matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-		DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad X1 Carbon"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "Lenovo Yoga 13",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-		DMI_MATCH(DMI_PRODUCT_VERSION, "Lenovo IdeaPad Yoga 13"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "Lenovo Yoga 2 11",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-		DMI_MATCH(DMI_PRODUCT_VERSION, "Lenovo Yoga 2 11"),
-		},
-	},
-	{
-	.callback = video_set_use_native_backlight,
-	.ident = "Thinkpad Helix",
-	.matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-		DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad Helix"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "Dell Inspiron 7520",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
-		DMI_MATCH(DMI_PRODUCT_NAME, "Inspiron 7520"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "Acer Aspire 5733Z",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Acer"),
-		DMI_MATCH(DMI_PRODUCT_NAME, "Aspire 5733Z"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "Acer Aspire 5742G",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Acer"),
-		DMI_MATCH(DMI_PRODUCT_NAME, "Aspire 5742G"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "Acer Aspire V5-171",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Acer"),
-		DMI_MATCH(DMI_PRODUCT_NAME, "V5-171"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "Acer Aspire V5-431",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Acer"),
-		DMI_MATCH(DMI_PRODUCT_NAME, "Aspire V5-431"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "Acer Aspire V5-471G",
-	 .matches = {
-		DMI_MATCH(DMI_BOARD_VENDOR, "Acer"),
-		DMI_MATCH(DMI_PRODUCT_NAME, "Aspire V5-471G"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "Acer TravelMate B113",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Acer"),
-		DMI_MATCH(DMI_PRODUCT_NAME, "TravelMate B113"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "Acer Aspire V5-572G",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Acer Aspire"),
-		DMI_MATCH(DMI_PRODUCT_VERSION, "V5-572G/Dazzle_CX"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "Acer Aspire V5-573G",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Acer Aspire"),
-		DMI_MATCH(DMI_PRODUCT_VERSION, "V5-573G/Dazzle_HW"),
-		},
-	},
-	{
-	 .callback = video_set_use_native_backlight,
-	 .ident = "ASUS Zenbook Prime UX31A",
-	 .matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
-		DMI_MATCH(DMI_PRODUCT_NAME, "UX31A"),
-		},
-	},
-	{
-	.callback = video_set_use_native_backlight,
-	.ident = "HP ProBook 4340s",
-	.matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"),
-		DMI_MATCH(DMI_PRODUCT_VERSION, "HP ProBook 4340s"),
-		},
-	},
-	{
-	.callback = video_set_use_native_backlight,
-	.ident = "HP ProBook 4540s",
-	.matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"),
-		DMI_MATCH(DMI_PRODUCT_VERSION, "HP ProBook 4540s"),
-		},
-	},
-	{
-	.callback = video_set_use_native_backlight,
-	.ident = "HP ProBook 2013 models",
-	.matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"),
-		DMI_MATCH(DMI_PRODUCT_NAME, "HP ProBook "),
-		DMI_MATCH(DMI_PRODUCT_NAME, " G1"),
-		},
-	},
-	{
-	.callback = video_set_use_native_backlight,
-	.ident = "HP EliteBook 2013 models",
-	.matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"),
-		DMI_MATCH(DMI_PRODUCT_NAME, "HP EliteBook "),
-		DMI_MATCH(DMI_PRODUCT_NAME, " G1"),
-		},
-	},
-	{
-	.callback = video_set_use_native_backlight,
-	.ident = "HP EliteBook 2014 models",
-	.matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"),
-		DMI_MATCH(DMI_PRODUCT_NAME, "HP EliteBook "),
-		DMI_MATCH(DMI_PRODUCT_NAME, " G2"),
-		},
-	},
-	{
-	.callback = video_set_use_native_backlight,
-	.ident = "HP ZBook 14",
-	.matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"),
-		DMI_MATCH(DMI_PRODUCT_NAME, "HP ZBook 14"),
-		},
-	},
-	{
-	.callback = video_set_use_native_backlight,
-	.ident = "HP ZBook 15",
-	.matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"),
-		DMI_MATCH(DMI_PRODUCT_NAME, "HP ZBook 15"),
-		},
-	},
-	{
-	.callback = video_set_use_native_backlight,
-	.ident = "HP ZBook 17",
-	.matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"),
-		DMI_MATCH(DMI_PRODUCT_NAME, "HP ZBook 17"),
-		},
-	},
-	{
-	.callback = video_set_use_native_backlight,
-	.ident = "HP EliteBook 8470p",
-	.matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"),
-		DMI_MATCH(DMI_PRODUCT_NAME, "HP EliteBook 8470p"),
-		},
-	},
-	{
-	.callback = video_set_use_native_backlight,
-	.ident = "HP EliteBook 8780w",
-	.matches = {
-		DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"),
-		DMI_MATCH(DMI_PRODUCT_NAME, "HP EliteBook 8780w"),
-		},
-	},
 
 	/*
 	 * These models have a working acpi_video backlight control, and using
@@ -769,6 +505,33 @@ static struct dmi_system_id video_dmi_table[] __initdata = {
 		DMI_MATCH(DMI_PRODUCT_NAME, "HP ENVY 15 Notebook PC"),
 		},
 	},
+
+	{
+	 .callback = video_disable_native_backlight,
+	 .ident = "SAMSUNG 870Z5E/880Z5E/680Z5E",
+	 .matches = {
+		DMI_MATCH(DMI_SYS_VENDOR, "SAMSUNG ELECTRONICS CO., LTD."),
+		DMI_MATCH(DMI_PRODUCT_NAME, "870Z5E/880Z5E/680Z5E"),
+		},
+	},
+	{
+	 .callback = video_disable_native_backlight,
+	 .ident = "SAMSUNG 370R4E/370R4V/370R5E/3570RE/370R5V",
+	 .matches = {
+		DMI_MATCH(DMI_SYS_VENDOR, "SAMSUNG ELECTRONICS CO., LTD."),
+		DMI_MATCH(DMI_PRODUCT_NAME, "370R4E/370R4V/370R5E/3570RE/370R5V"),
+		},
+	},
+
+	{
+	 /* https://bugzilla.redhat.com/show_bug.cgi?id=1163574 */
+	 .callback = video_disable_native_backlight,
+	 .ident = "Dell XPS15 L521X",
+	 .matches = {
+		DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+		DMI_MATCH(DMI_PRODUCT_NAME, "XPS L521X"),
+		},
+	},
 	{}
 };
 
@@ -1419,6 +1182,28 @@ acpi_video_device_bind(struct acpi_video_bus *video,
 	}
 }
 
+static bool acpi_video_device_in_dod(struct acpi_video_device *device)
+{
+	struct acpi_video_bus *video = device->video;
+	int i;
+
+	/*
+	 * If we have a broken _DOD or we have more than 8 output devices
+	 * under the graphics controller node that we can't proper deal with
+	 * in the operation region code currently, no need to test.
+	 */
+	if (!video->attached_count || video->child_count > 8)
+		return true;
+
+	for (i = 0; i < video->attached_count; i++) {
+		if ((video->attached_array[i].value.int_val & 0xfff) ==
+		    (device->device_id & 0xfff))
+			return true;
+	}
+
+	return false;
+}
+
 /*
  *  Arg:
  *	video	: video bus device
@@ -1660,6 +1445,7 @@ acpi_video_bus_get_devices(struct acpi_video_bus *video,
 			dev_err(&dev->dev, "Can't attach device\n");
 			break;
 		}
+		video->child_count++;
 	}
 	return status;
 }
@@ -1858,6 +1644,15 @@ static void acpi_video_dev_register_backlight(struct acpi_video_device *device)
 	static int count;
 	char *name;
 
+	/*
+	 * Do not create backlight device for video output
+	 * device that is not in the enumerated list.
+	 */
+	if (!acpi_video_device_in_dod(device)) {
+		dev_dbg(&device->dev->dev, "not in _DOD list, ignore\n");
+		return;
+	}
+
 	result = acpi_video_init_brightness(device);
 	if (result)
 		return;
@@ -1919,6 +1714,19 @@ static void acpi_video_dev_register_backlight(struct acpi_video_device *device)
 		printk(KERN_ERR PREFIX "Create sysfs link\n");
 }
 
+static void acpi_video_run_bcl_for_osi(struct acpi_video_bus *video)
+{
+	struct acpi_video_device *dev;
+	union acpi_object *levels;
+
+	mutex_lock(&video->device_list_lock);
+	list_for_each_entry(dev, &video->video_device_list, entry) {
+		if (!acpi_video_device_lcd_query_levels(dev, &levels))
+			kfree(levels);
+	}
+	mutex_unlock(&video->device_list_lock);
+}
+
 static int acpi_video_bus_register_backlight(struct acpi_video_bus *video)
 {
 	struct acpi_video_device *dev;
@@ -1926,6 +1734,8 @@ static int acpi_video_bus_register_backlight(struct acpi_video_bus *video)
 	if (video->backlight_registered)
 		return 0;
 
+	acpi_video_run_bcl_for_osi(video);
+
 	if (!acpi_video_verify_backlight_support())
 		return 0;
 
diff --git a/drivers/acpi/video_detect.c b/drivers/acpi/video_detect.c
index c42feb2bacd0..27c43499977a 100644
--- a/drivers/acpi/video_detect.c
+++ b/drivers/acpi/video_detect.c
@@ -174,6 +174,14 @@ static struct dmi_system_id video_detect_dmi_table[] = {
 		DMI_MATCH(DMI_PRODUCT_NAME, "Inspiron 5737"),
 		},
 	},
+	{
+	.callback = video_detect_force_vendor,
+	.ident = "Lenovo IdeaPad Z570",
+	.matches = {
+		DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
+		DMI_MATCH(DMI_PRODUCT_VERSION, "Ideapad Z570"),
+		},
+	},
 	{ },
 };
 
diff --git a/drivers/amba/Kconfig b/drivers/amba/Kconfig
new file mode 100644
index 000000000000..4a5c9d279059
--- /dev/null
+++ b/drivers/amba/Kconfig
@@ -0,0 +1,14 @@
+config ARM_AMBA
+	bool
+
+if ARM_AMBA
+
+config TEGRA_AHB
+	bool "Enable AHB driver for NVIDIA Tegra SoCs"
+	default y if ARCH_TEGRA
+	help
+	  Adds AHB configuration functionality for NVIDIA Tegra SoCs,
+	  which controls AHB bus master arbitration and some performance
+	  parameters (priority, prefetch size).
+
+endif
diff --git a/drivers/amba/bus.c b/drivers/amba/bus.c
index 3cf61a127ee5..52ddd9fbb55e 100644
--- a/drivers/amba/bus.c
+++ b/drivers/amba/bus.c
@@ -15,6 +15,7 @@
 #include <linux/io.h>
 #include <linux/pm.h>
 #include <linux/pm_runtime.h>
+#include <linux/pm_domain.h>
 #include <linux/amba/bus.h>
 #include <linux/sizes.h>
 
@@ -94,8 +95,12 @@ static int amba_pm_runtime_suspend(struct device *dev)
 	struct amba_device *pcdev = to_amba_device(dev);
 	int ret = pm_generic_runtime_suspend(dev);
 
-	if (ret == 0 && dev->driver)
-		clk_disable_unprepare(pcdev->pclk);
+	if (ret == 0 && dev->driver) {
+		if (pm_runtime_is_irq_safe(dev))
+			clk_disable(pcdev->pclk);
+		else
+			clk_disable_unprepare(pcdev->pclk);
+	}
 
 	return ret;
 }
@@ -106,7 +111,10 @@ static int amba_pm_runtime_resume(struct device *dev)
 	int ret;
 
 	if (dev->driver) {
-		ret = clk_prepare_enable(pcdev->pclk);
+		if (pm_runtime_is_irq_safe(dev))
+			ret = clk_enable(pcdev->pclk);
+		else
+			ret = clk_prepare_enable(pcdev->pclk);
 		/* Failure is probably fatal to the system, but... */
 		if (ret)
 			return ret;
@@ -114,7 +122,7 @@ static int amba_pm_runtime_resume(struct device *dev)
 
 	return pm_generic_runtime_resume(dev);
 }
-#endif
+#endif /* CONFIG_PM */
 
 static const struct dev_pm_ops amba_pm = {
 	.suspend	= pm_generic_suspend,
@@ -123,7 +131,7 @@ static const struct dev_pm_ops amba_pm = {
 	.thaw		= pm_generic_thaw,
 	.poweroff	= pm_generic_poweroff,
 	.restore	= pm_generic_restore,
-	SET_PM_RUNTIME_PM_OPS(
+	SET_RUNTIME_PM_OPS(
 		amba_pm_runtime_suspend,
 		amba_pm_runtime_resume,
 		NULL
@@ -182,9 +190,15 @@ static int amba_probe(struct device *dev)
 	int ret;
 
 	do {
+		ret = dev_pm_domain_attach(dev, true);
+		if (ret == -EPROBE_DEFER)
+			break;
+
 		ret = amba_get_enable_pclk(pcdev);
-		if (ret)
+		if (ret) {
+			dev_pm_domain_detach(dev, true);
 			break;
+		}
 
 		pm_runtime_get_noresume(dev);
 		pm_runtime_set_active(dev);
@@ -199,6 +213,7 @@ static int amba_probe(struct device *dev)
 		pm_runtime_put_noidle(dev);
 
 		amba_put_disable_pclk(pcdev);
+		dev_pm_domain_detach(dev, true);
 	} while (0);
 
 	return ret;
@@ -220,6 +235,7 @@ static int amba_remove(struct device *dev)
 	pm_runtime_put_noidle(dev);
 
 	amba_put_disable_pclk(pcdev);
+	dev_pm_domain_detach(dev, true);
 
 	return ret;
 }
@@ -327,7 +343,7 @@ int amba_device_add(struct amba_device *dev, struct resource *parent)
 
 		amba_put_disable_pclk(dev);
 
-		if (cid == AMBA_CID)
+		if (cid == AMBA_CID || cid == CORESIGHT_CID)
 			dev->periphid = pid;
 
 		if (!dev->periphid)
diff --git a/drivers/amba/tegra-ahb.c b/drivers/amba/tegra-ahb.c
index d8961ef4d2e7..c6dc3548e5d1 100644
--- a/drivers/amba/tegra-ahb.c
+++ b/drivers/amba/tegra-ahb.c
@@ -277,7 +277,6 @@ static struct platform_driver tegra_ahb_driver = {
 	.probe = tegra_ahb_probe,
 	.driver = {
 		.name = DRV_NAME,
-		.owner = THIS_MODULE,
 		.of_match_table = tegra_ahb_of_match,
 		.pm = &tegra_ahb_pm,
 	},
diff --git a/drivers/android/Kconfig b/drivers/android/Kconfig
new file mode 100644
index 000000000000..bdfc6c6f4f5a
--- /dev/null
+++ b/drivers/android/Kconfig
@@ -0,0 +1,37 @@
+menu "Android"
+
+config ANDROID
+	bool "Android Drivers"
+	---help---
+	  Enable support for various drivers needed on the Android platform
+
+if ANDROID
+
+config ANDROID_BINDER_IPC
+	bool "Android Binder IPC Driver"
+	depends on MMU
+	default n
+	---help---
+	  Binder is used in Android for both communication between processes,
+	  and remote method invocation.
+
+	  This means one Android process can call a method/routine in another
+	  Android process, using Binder to identify, invoke and pass arguments
+	  between said processes.
+
+config ANDROID_BINDER_IPC_32BIT
+	bool
+	depends on !64BIT && ANDROID_BINDER_IPC
+	default y
+	---help---
+	  The Binder API has been changed to support both 32 and 64bit
+	  applications in a mixed environment.
+
+	  Enable this to support an old 32-bit Android user-space (v4.4 and
+	  earlier).
+
+	  Note that enabling this will break newer Android user-space.
+
+endif # if ANDROID
+
+endmenu
diff --git a/drivers/android/Makefile b/drivers/android/Makefile
new file mode 100644
index 000000000000..3b7e4b072c58
--- /dev/null
+++ b/drivers/android/Makefile
@@ -0,0 +1,3 @@
+ccflags-y += -I$(src)			# needed for trace events
+
+obj-$(CONFIG_ANDROID_BINDER_IPC)	+= binder.o
diff --git a/drivers/staging/android/binder.c b/drivers/android/binder.c
index 4f34dc0095b5..8c43521d3f11 100644
--- a/drivers/staging/android/binder.c
+++ b/drivers/android/binder.c
@@ -38,7 +38,11 @@
 #include <linux/slab.h>
 #include <linux/pid_namespace.h>
 
-#include "binder.h"
+#ifdef CONFIG_ANDROID_BINDER_IPC_32BIT
+#define BINDER_IPC_32BIT 1
+#endif
+
+#include <uapi/linux/android/binder.h>
 #include "binder_trace.h"
 
 static DEFINE_MUTEX(binder_main_lock);
@@ -1198,7 +1202,8 @@ static void binder_send_failed_reply(struct binder_transaction *t,
 			if (target_thread->return_error == BR_OK) {
 				binder_debug(BINDER_DEBUG_FAILED_TRANSACTION,
 					     "send failed reply for transaction %d to %d:%d\n",
-					      t->debug_id, target_thread->proc->pid,
+					      t->debug_id,
+					      target_thread->proc->pid,
 					      target_thread->pid);
 
 				binder_pop_transaction(target_thread, t);
@@ -2198,12 +2203,16 @@ retry:
 		struct binder_work *w;
 		struct binder_transaction *t = NULL;
 
-		if (!list_empty(&thread->todo))
-			w = list_first_entry(&thread->todo, struct binder_work, entry);
-		else if (!list_empty(&proc->todo) && wait_for_proc_work)
-			w = list_first_entry(&proc->todo, struct binder_work, entry);
-		else {
-			if (ptr - buffer == 4 && !(thread->looper & BINDER_LOOPER_STATE_NEED_RETURN)) /* no data added */
+		if (!list_empty(&thread->todo)) {
+			w = list_first_entry(&thread->todo, struct binder_work,
+					     entry);
+		} else if (!list_empty(&proc->todo) && wait_for_proc_work) {
+			w = list_first_entry(&proc->todo, struct binder_work,
+					     entry);
+		} else {
+			/* no data added */
+			if (ptr - buffer == 4 &&
+			    !(thread->looper & BINDER_LOOPER_STATE_NEED_RETURN))
 				goto retry;
 			break;
 		}
diff --git a/drivers/staging/android/binder_trace.h b/drivers/android/binder_trace.h
index 7f20f3dc8369..7f20f3dc8369 100644
--- a/drivers/staging/android/binder_trace.h
+++ b/drivers/android/binder_trace.h
diff --git a/drivers/ata/Kconfig b/drivers/ata/Kconfig
index e1b92788c225..5f601553b9b0 100644
--- a/drivers/ata/Kconfig
+++ b/drivers/ata/Kconfig
@@ -61,7 +61,7 @@ config ATA_ACPI
 
 config SATA_ZPODD
 	bool "SATA Zero Power Optical Disc Drive (ZPODD) support"
-	depends on ATA_ACPI && PM_RUNTIME
+	depends on ATA_ACPI && PM
 	default n
 	help
 	  This option adds support for SATA Zero Power Optical Disc
@@ -299,7 +299,7 @@ config SATA_HIGHBANK
 
 config SATA_MV
 	tristate "Marvell SATA support"
-	depends on PCI || ARCH_DOVE || ARCH_KIRKWOOD || ARCH_MV78XX0 || \
+	depends on PCI || ARCH_DOVE || ARCH_MV78XX0 || \
 		   ARCH_MVEBU || ARCH_ORION5X || COMPILE_TEST
 	select GENERIC_PHY
 	help
@@ -835,6 +835,7 @@ config PATA_AT32
 config PATA_AT91
 	tristate "PATA support for AT91SAM9260"
 	depends on ARM && SOC_AT91SAM9
+	depends on !ARCH_MULTIPLATFORM
 	help
 	  This option enables support for IDE devices on the Atmel AT91SAM9260 SoC.
 
diff --git a/drivers/ata/acard-ahci.c b/drivers/ata/acard-ahci.c
index 25d0ac32e721..c962886d7e71 100644
--- a/drivers/ata/acard-ahci.c
+++ b/drivers/ata/acard-ahci.c
@@ -498,8 +498,7 @@ static int acard_ahci_init_one(struct pci_dev *pdev, const struct pci_device_id
 	acard_ahci_pci_print_info(host);
 
 	pci_set_master(pdev);
-	return ata_host_activate(host, pdev->irq, ahci_interrupt, IRQF_SHARED,
-				 &acard_ahci_sht);
+	return ahci_host_activate(host, pdev->irq, &acard_ahci_sht);
 }
 
 module_pci_driver(acard_ahci_pci_driver);
diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index a0cc0edafc78..33bb06e006c9 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -60,6 +60,7 @@ enum board_ids {
 	/* board IDs by feature in alphabetical order */
 	board_ahci,
 	board_ahci_ign_iferr,
+	board_ahci_nomsi,
 	board_ahci_noncq,
 	board_ahci_nosntf,
 	board_ahci_yes_fbs,
@@ -121,6 +122,13 @@ static const struct ata_port_info ahci_port_info[] = {
 		.udma_mask	= ATA_UDMA6,
 		.port_ops	= &ahci_ops,
 	},
+	[board_ahci_nomsi] = {
+		AHCI_HFLAGS	(AHCI_HFLAG_NO_MSI),
+		.flags		= AHCI_FLAG_COMMON,
+		.pio_mask	= ATA_PIO4,
+		.udma_mask	= ATA_UDMA6,
+		.port_ops	= &ahci_ops,
+	},
 	[board_ahci_noncq] = {
 		AHCI_HFLAGS	(AHCI_HFLAG_NO_NCQ),
 		.flags		= AHCI_FLAG_COMMON,
@@ -313,6 +321,13 @@ static const struct pci_device_id ahci_pci_tbl[] = {
 	{ PCI_VDEVICE(INTEL, 0x8c87), board_ahci }, /* 9 Series RAID */
 	{ PCI_VDEVICE(INTEL, 0x8c8e), board_ahci }, /* 9 Series RAID */
 	{ PCI_VDEVICE(INTEL, 0x8c8f), board_ahci }, /* 9 Series RAID */
+	{ PCI_VDEVICE(INTEL, 0x9d03), board_ahci }, /* Sunrise Point-LP AHCI */
+	{ PCI_VDEVICE(INTEL, 0x9d05), board_ahci }, /* Sunrise Point-LP RAID */
+	{ PCI_VDEVICE(INTEL, 0x9d07), board_ahci }, /* Sunrise Point-LP RAID */
+	{ PCI_VDEVICE(INTEL, 0xa103), board_ahci }, /* Sunrise Point-H AHCI */
+	{ PCI_VDEVICE(INTEL, 0xa105), board_ahci }, /* Sunrise Point-H RAID */
+	{ PCI_VDEVICE(INTEL, 0xa107), board_ahci }, /* Sunrise Point-H RAID */
+	{ PCI_VDEVICE(INTEL, 0xa10f), board_ahci }, /* Sunrise Point-H RAID */
 
 	/* JMicron 360/1/3/5/6, match class to avoid IDE function */
 	{ PCI_VENDOR_ID_JMICRON, PCI_ANY_ID, PCI_ANY_ID, PCI_ANY_ID,
@@ -475,10 +490,11 @@ static const struct pci_device_id ahci_pci_tbl[] = {
 	{ PCI_VDEVICE(ASMEDIA, 0x0612), board_ahci },	/* ASM1062 */
 
 	/*
-	 * Samsung SSDs found on some macbooks.  NCQ times out.
-	 * https://bugzilla.kernel.org/show_bug.cgi?id=60731
+	 * Samsung SSDs found on some macbooks.  NCQ times out if MSI is
+	 * enabled.  https://bugzilla.kernel.org/show_bug.cgi?id=60731
 	 */
-	{ PCI_VDEVICE(SAMSUNG, 0x1600), board_ahci_noncq },
+	{ PCI_VDEVICE(SAMSUNG, 0x1600), board_ahci_nomsi },
+	{ PCI_VDEVICE(SAMSUNG, 0xa800), board_ahci_nomsi },
 
 	/* Enmotus */
 	{ PCI_DEVICE(0x1c44, 0x8000), board_ahci },
@@ -514,12 +530,9 @@ MODULE_PARM_DESC(marvell_enable, "Marvell SATA via AHCI (1 = enabled)");
 static void ahci_pci_save_initial_config(struct pci_dev *pdev,
 					 struct ahci_host_priv *hpriv)
 {
-	unsigned int force_port_map = 0;
-	unsigned int mask_port_map = 0;
-
 	if (pdev->vendor == PCI_VENDOR_ID_JMICRON && pdev->device == 0x2361) {
 		dev_info(&pdev->dev, "JMB361 has only one port\n");
-		force_port_map = 1;
+		hpriv->force_port_map = 1;
 	}
 
 	/*
@@ -529,9 +542,9 @@ static void ahci_pci_save_initial_config(struct pci_dev *pdev,
 	 */
 	if (hpriv->flags & AHCI_HFLAG_MV_PATA) {
 		if (pdev->device == 0x6121)
-			mask_port_map = 0x3;
+			hpriv->mask_port_map = 0x3;
 		else
-			mask_port_map = 0xf;
+			hpriv->mask_port_map = 0xf;
 		dev_info(&pdev->dev,
 			  "Disabling your PATA port. Use the boot option 'ahci.marvell_enable=0' to avoid this.\n");
 	}
@@ -790,7 +803,7 @@ static void ahci_pci_print_info(struct ata_host *host)
  */
 static void ahci_p5wdh_workaround(struct ata_host *host)
 {
-	static struct dmi_system_id sysids[] = {
+	static const struct dmi_system_id sysids[] = {
 		{
 			.ident = "P5W DH Deluxe",
 			.matches = {
@@ -1221,6 +1234,9 @@ static int ahci_init_interrupts(struct pci_dev *pdev, unsigned int n_ports,
 		goto single_msi;
 	}
 
+	if (nvec > 1)
+		hpriv->flags |= AHCI_HFLAG_MULTI_MSI;
+
 	return nvec;
 
 single_msi:
@@ -1233,71 +1249,6 @@ intx:
 	return 0;
 }
 
-/**
- *	ahci_host_activate - start AHCI host, request IRQs and register it
- *	@host: target ATA host
- *	@irq: base IRQ number to request
- *	@n_msis: number of MSIs allocated for this host
- *	@irq_handler: irq_handler used when requesting IRQs
- *	@irq_flags: irq_flags used when requesting IRQs
- *
- *	Similar to ata_host_activate, but requests IRQs according to AHCI-1.1
- *	when multiple MSIs were allocated. That is one MSI per port, starting
- *	from @irq.
- *
- *	LOCKING:
- *	Inherited from calling layer (may sleep).
- *
- *	RETURNS:
- *	0 on success, -errno otherwise.
- */
-int ahci_host_activate(struct ata_host *host, int irq, unsigned int n_msis)
-{
-	int i, rc;
-
-	/* Sharing Last Message among several ports is not supported */
-	if (n_msis < host->n_ports)
-		return -EINVAL;
-
-	rc = ata_host_start(host);
-	if (rc)
-		return rc;
-
-	for (i = 0; i < host->n_ports; i++) {
-		struct ahci_port_priv *pp = host->ports[i]->private_data;
-
-		/* Do not receive interrupts sent by dummy ports */
-		if (!pp) {
-			disable_irq(irq + i);
-			continue;
-		}
-
-		rc = devm_request_threaded_irq(host->dev, irq + i,
-					       ahci_hw_interrupt,
-					       ahci_thread_fn, IRQF_SHARED,
-					       pp->irq_desc, host->ports[i]);
-		if (rc)
-			goto out_free_irqs;
-	}
-
-	for (i = 0; i < host->n_ports; i++)
-		ata_port_desc(host->ports[i], "irq %d", irq + i);
-
-	rc = ata_host_register(host, &ahci_sht);
-	if (rc)
-		goto out_free_all_irqs;
-
-	return 0;
-
-out_free_all_irqs:
-	i = host->n_ports;
-out_free_irqs:
-	for (i--; i >= 0; i--)
-		devm_free_irq(host->dev, irq + i, host->ports[i]);
-
-	return rc;
-}
-
 static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 {
 	unsigned int board_id = ent->driver_data;
@@ -1306,7 +1257,7 @@ static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	struct device *dev = &pdev->dev;
 	struct ahci_host_priv *hpriv;
 	struct ata_host *host;
-	int n_ports, n_msis, i, rc;
+	int n_ports, i, rc;
 	int ahci_pci_bar = AHCI_PCI_BAR_STANDARD;
 
 	VPRINTK("ENTER\n");
@@ -1459,9 +1410,7 @@ static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	 */
 	n_ports = max(ahci_nr_ports(hpriv->cap), fls(hpriv->port_map));
 
-	n_msis = ahci_init_interrupts(pdev, n_ports, hpriv);
-	if (n_msis > 1)
-		hpriv->flags |= AHCI_HFLAG_MULTI_MSI;
+	ahci_init_interrupts(pdev, n_ports, hpriv);
 
 	host = ata_host_alloc_pinfo(&pdev->dev, ppi, n_ports);
 	if (!host)
@@ -1513,11 +1462,7 @@ static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 
 	pci_set_master(pdev);
 
-	if (hpriv->flags & AHCI_HFLAG_MULTI_MSI)
-		return ahci_host_activate(host, pdev->irq, n_msis);
-
-	return ata_host_activate(host, pdev->irq, ahci_interrupt, IRQF_SHARED,
-				 &ahci_sht);
+	return ahci_host_activate(host, pdev->irq, &ahci_sht);
 }
 
 module_pci_driver(ahci_pci_driver);
diff --git a/drivers/ata/ahci.h b/drivers/ata/ahci.h
index 59ae0ee00149..40f0e34f17af 100644
--- a/drivers/ata/ahci.h
+++ b/drivers/ata/ahci.h
@@ -53,7 +53,7 @@
 
 enum {
 	AHCI_MAX_PORTS		= 32,
-	AHCI_MAX_CLKS		= 4,
+	AHCI_MAX_CLKS		= 5,
 	AHCI_MAX_SG		= 168, /* hardware max is 64K */
 	AHCI_DMA_BOUNDARY	= 0xffffffff,
 	AHCI_MAX_CMDS		= 32,
@@ -304,7 +304,7 @@ struct ahci_port_priv {
 	unsigned int		ncq_saw_d2h:1;
 	unsigned int		ncq_saw_dmas:1;
 	unsigned int		ncq_saw_sdb:1;
-	u32			intr_status;	/* interrupts to handle */
+	atomic_t		intr_status;	/* interrupts to handle */
 	spinlock_t		lock;		/* protects parent ata_port */
 	u32 			intr_mask;	/* interrupts to enable */
 	bool			fbs_supported;	/* set iff FBS is supported */
@@ -388,11 +388,9 @@ int ahci_port_resume(struct ata_port *ap);
 void ahci_set_em_messages(struct ahci_host_priv *hpriv,
 			  struct ata_port_info *pi);
 int ahci_reset_em(struct ata_host *host);
-irqreturn_t ahci_interrupt(int irq, void *dev_instance);
-irqreturn_t ahci_hw_interrupt(int irq, void *dev_instance);
-irqreturn_t ahci_thread_fn(int irq, void *dev_instance);
 void ahci_print_info(struct ata_host *host, const char *scc_s);
-int ahci_host_activate(struct ata_host *host, int irq, unsigned int n_msis);
+int ahci_host_activate(struct ata_host *host, int irq,
+		       struct scsi_host_template *sht);
 void ahci_error_handler(struct ata_port *ap);
 
 static inline void __iomem *__ahci_port_base(struct ata_host *host,
diff --git a/drivers/ata/ahci_da850.c b/drivers/ata/ahci_da850.c
index ad1e71ec10cf..ce8a7a6d6c7f 100644
--- a/drivers/ata/ahci_da850.c
+++ b/drivers/ata/ahci_da850.c
@@ -103,7 +103,6 @@ static struct platform_driver ahci_da850_driver = {
 	.remove = ata_platform_remove_one,
 	.driver = {
 		.name = "ahci_da850",
-		.owner = THIS_MODULE,
 		.pm = &ahci_da850_pm_ops,
 	},
 };
diff --git a/drivers/ata/ahci_imx.c b/drivers/ata/ahci_imx.c
index f3970b4ed889..35d51c59a370 100644
--- a/drivers/ata/ahci_imx.c
+++ b/drivers/ata/ahci_imx.c
@@ -679,7 +679,6 @@ static struct platform_driver imx_ahci_driver = {
 	.remove = ata_platform_remove_one,
 	.driver = {
 		.name = "ahci-imx",
-		.owner = THIS_MODULE,
 		.of_match_table = imx_ahci_of_match,
 		.pm = &ahci_imx_pm_ops,
 	},
diff --git a/drivers/ata/ahci_mvebu.c b/drivers/ata/ahci_mvebu.c
index 68672d2692ee..64bb08432b69 100644
--- a/drivers/ata/ahci_mvebu.c
+++ b/drivers/ata/ahci_mvebu.c
@@ -115,7 +115,6 @@ static struct platform_driver ahci_mvebu_driver = {
 	.remove = ata_platform_remove_one,
 	.driver = {
 		.name = "ahci-mvebu",
-		.owner = THIS_MODULE,
 		.of_match_table = ahci_mvebu_of_match,
 	},
 };
diff --git a/drivers/ata/ahci_platform.c b/drivers/ata/ahci_platform.c
index f61ddb9146d6..18d539837045 100644
--- a/drivers/ata/ahci_platform.c
+++ b/drivers/ata/ahci_platform.c
@@ -32,7 +32,6 @@ static const struct ata_port_info ahci_port_info = {
 static int ahci_probe(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
-	struct ahci_platform_data *pdata = dev_get_platdata(dev);
 	struct ahci_host_priv *hpriv;
 	int rc;
 
@@ -44,29 +43,14 @@ static int ahci_probe(struct platform_device *pdev)
 	if (rc)
 		return rc;
 
-	/*
-	 * Some platforms might need to prepare for mmio region access,
-	 * which could be done in the following init call. So, the mmio
-	 * region shouldn't be accessed before init (if provided) has
-	 * returned successfully.
-	 */
-	if (pdata && pdata->init) {
-		rc = pdata->init(dev, hpriv->mmio);
-		if (rc)
-			goto disable_resources;
-	}
-
 	if (of_device_is_compatible(dev->of_node, "hisilicon,hisi-ahci"))
 		hpriv->flags |= AHCI_HFLAG_NO_FBS | AHCI_HFLAG_NO_NCQ;
 
 	rc = ahci_platform_init_host(pdev, hpriv, &ahci_port_info);
 	if (rc)
-		goto pdata_exit;
+		goto disable_resources;
 
 	return 0;
-pdata_exit:
-	if (pdata && pdata->exit)
-		pdata->exit(dev);
 disable_resources:
 	ahci_platform_disable_resources(hpriv);
 	return rc;
@@ -92,7 +76,6 @@ static struct platform_driver ahci_driver = {
 	.remove = ata_platform_remove_one,
 	.driver = {
 		.name = "ahci",
-		.owner = THIS_MODULE,
 		.of_match_table = ahci_of_match,
 		.pm = &ahci_pm_ops,
 	},
diff --git a/drivers/ata/ahci_st.c b/drivers/ata/ahci_st.c
index 835d6eea84fd..2f9e8317cc16 100644
--- a/drivers/ata/ahci_st.c
+++ b/drivers/ata/ahci_st.c
@@ -230,7 +230,6 @@ MODULE_DEVICE_TABLE(of, st_ahci_match);
 static struct platform_driver st_ahci_driver = {
 	.driver = {
 		.name = "st_ahci",
-		.owner = THIS_MODULE,
 		.pm = &st_ahci_pm_ops,
 		.of_match_table = of_match_ptr(st_ahci_match),
 	},
diff --git a/drivers/ata/ahci_sunxi.c b/drivers/ata/ahci_sunxi.c
index e44d675a30ec..e2e0da539a2f 100644
--- a/drivers/ata/ahci_sunxi.c
+++ b/drivers/ata/ahci_sunxi.c
@@ -27,6 +27,12 @@
 #include <linux/regulator/consumer.h>
 #include "ahci.h"
 
+/* Insmod parameters */
+static bool enable_pmp;
+module_param(enable_pmp, bool, 0);
+MODULE_PARM_DESC(enable_pmp,
+	"Enable support for sata port multipliers, only use if you use a pmp!");
+
 #define AHCI_BISTAFR	0x00a0
 #define AHCI_BISTCR	0x00a4
 #define AHCI_BISTFCTR	0x00a8
@@ -184,7 +190,15 @@ static int ahci_sunxi_probe(struct platform_device *pdev)
 		goto disable_resources;
 
 	hpriv->flags = AHCI_HFLAG_32BIT_ONLY | AHCI_HFLAG_NO_MSI |
-		       AHCI_HFLAG_NO_PMP | AHCI_HFLAG_YES_NCQ;
+		       AHCI_HFLAG_YES_NCQ;
+
+	/*
+	 * The sunxi sata controller seems to be unable to successfully do a
+	 * soft reset if no pmp is attached, so disable pmp use unless
+	 * requested, otherwise directly attached disks do not work.
+	 */
+	if (!enable_pmp)
+		hpriv->flags |= AHCI_HFLAG_NO_PMP;
 
 	rc = ahci_platform_init_host(pdev, hpriv, &ahci_sunxi_port_info);
 	if (rc)
@@ -238,7 +252,6 @@ static struct platform_driver ahci_sunxi_driver = {
 	.remove = ata_platform_remove_one,
 	.driver = {
 		.name = "ahci-sunxi",
-		.owner = THIS_MODULE,
 		.of_match_table = ahci_sunxi_of_match,
 		.pm = &ahci_sunxi_pm_ops,
 	},
diff --git a/drivers/ata/ahci_xgene.c b/drivers/ata/ahci_xgene.c
index f03aab187f4d..cbcd20810355 100644
--- a/drivers/ata/ahci_xgene.c
+++ b/drivers/ata/ahci_xgene.c
@@ -125,10 +125,11 @@ static int xgene_ahci_restart_engine(struct ata_port *ap)
  * xgene_ahci_qc_issue - Issue commands to the device
  * @qc: Command to issue
  *
- * Due to Hardware errata for IDENTIFY DEVICE command, the controller cannot
- * clear the BSY bit after receiving the PIO setup FIS. This results in the dma
- * state machine goes into the CMFatalErrorUpdate state and locks up. By
- * restarting the dma engine, it removes the controller out of lock up state.
+ * Due to Hardware errata for IDENTIFY DEVICE command and PACKET
+ * command of ATAPI protocol set, the controller cannot clear the BSY bit
+ * after receiving the PIO setup FIS. This results in the DMA state machine
+ * going into the CMFatalErrorUpdate state and locks up. By restarting the
+ * DMA engine, it removes the controller out of lock up state.
  */
 static unsigned int xgene_ahci_qc_issue(struct ata_queued_cmd *qc)
 {
@@ -137,7 +138,8 @@ static unsigned int xgene_ahci_qc_issue(struct ata_queued_cmd *qc)
 	struct xgene_ahci_context *ctx = hpriv->plat_data;
 	int rc = 0;
 
-	if (unlikely(ctx->last_cmd[ap->port_no] == ATA_CMD_ID_ATA))
+	if (unlikely((ctx->last_cmd[ap->port_no] == ATA_CMD_ID_ATA) ||
+	    (ctx->last_cmd[ap->port_no] == ATA_CMD_PACKET)))
 		xgene_ahci_restart_engine(ap);
 
 	rc = ahci_qc_issue(qc);
@@ -188,7 +190,7 @@ static unsigned int xgene_ahci_read_id(struct ata_device *dev,
 	 *
 	 * Clear reserved bit 8 (DEVSLP bit) as we don't support DEVSLP
 	 */
-	id[ATA_ID_FEATURE_SUPP] &= ~(1 << 8);
+	id[ATA_ID_FEATURE_SUPP] &= cpu_to_le16(~(1 << 8));
 
 	return 0;
 }
@@ -434,7 +436,7 @@ static int xgene_ahci_mux_select(struct xgene_ahci_context *ctx)
 	u32 val;
 
 	/* Check for optional MUX resource */
-	if (IS_ERR(ctx->csr_mux))
+	if (!ctx->csr_mux)
 		return 0;
 
 	val = readl(ctx->csr_mux + SATA_ENET_CONFIG_REG);
@@ -484,7 +486,13 @@ static int xgene_ahci_probe(struct platform_device *pdev)
 
 	/* Retrieve the optional IP mux resource */
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 4);
-	ctx->csr_mux = devm_ioremap_resource(dev, res);
+	if (res) {
+		void __iomem *csr = devm_ioremap_resource(dev, res);
+		if (IS_ERR(csr))
+			return PTR_ERR(csr);
+
+		ctx->csr_mux = csr;
+	}
 
 	dev_dbg(dev, "VAddr 0x%p Mmio VAddr 0x%p\n", ctx->csr_core,
 		hpriv->mmio);
@@ -538,7 +546,6 @@ static struct platform_driver xgene_ahci_driver = {
 	.remove = ata_platform_remove_one,
 	.driver = {
 		.name = "xgene-ahci",
-		.owner = THIS_MODULE,
 		.of_match_table = xgene_ahci_of_match,
 	},
 };
diff --git a/drivers/ata/libahci.c b/drivers/ata/libahci.c
index b784e9de426a..61a9c07e0dff 100644
--- a/drivers/ata/libahci.c
+++ b/drivers/ata/libahci.c
@@ -1789,19 +1789,16 @@ static void ahci_port_intr(struct ata_port *ap)
 	ahci_handle_port_interrupt(ap, port_mmio, status);
 }
 
-irqreturn_t ahci_thread_fn(int irq, void *dev_instance)
+static irqreturn_t ahci_port_thread_fn(int irq, void *dev_instance)
 {
 	struct ata_port *ap = dev_instance;
 	struct ahci_port_priv *pp = ap->private_data;
 	void __iomem *port_mmio = ahci_port_base(ap);
-	unsigned long flags;
 	u32 status;
 
-	spin_lock_irqsave(&ap->host->lock, flags);
-	status = pp->intr_status;
-	if (status)
-		pp->intr_status = 0;
-	spin_unlock_irqrestore(&ap->host->lock, flags);
+	status = atomic_xchg(&pp->intr_status, 0);
+	if (!status)
+		return IRQ_NONE;
 
 	spin_lock_bh(ap->lock);
 	ahci_handle_port_interrupt(ap, port_mmio, status);
@@ -1809,77 +1806,27 @@ irqreturn_t ahci_thread_fn(int irq, void *dev_instance)
 
 	return IRQ_HANDLED;
 }
-EXPORT_SYMBOL_GPL(ahci_thread_fn);
 
-static void ahci_hw_port_interrupt(struct ata_port *ap)
+static irqreturn_t ahci_multi_irqs_intr(int irq, void *dev_instance)
 {
+	struct ata_port *ap = dev_instance;
 	void __iomem *port_mmio = ahci_port_base(ap);
 	struct ahci_port_priv *pp = ap->private_data;
 	u32 status;
 
-	status = readl(port_mmio + PORT_IRQ_STAT);
-	writel(status, port_mmio + PORT_IRQ_STAT);
-
-	pp->intr_status |= status;
-}
-
-irqreturn_t ahci_hw_interrupt(int irq, void *dev_instance)
-{
-	struct ata_port *ap_this = dev_instance;
-	struct ahci_port_priv *pp = ap_this->private_data;
-	struct ata_host *host = ap_this->host;
-	struct ahci_host_priv *hpriv = host->private_data;
-	void __iomem *mmio = hpriv->mmio;
-	unsigned int i;
-	u32 irq_stat, irq_masked;
-
 	VPRINTK("ENTER\n");
 
-	spin_lock(&host->lock);
-
-	irq_stat = readl(mmio + HOST_IRQ_STAT);
-
-	if (!irq_stat) {
-		u32 status = pp->intr_status;
-
-		spin_unlock(&host->lock);
-
-		VPRINTK("EXIT\n");
-
-		return status ? IRQ_WAKE_THREAD : IRQ_NONE;
-	}
-
-	irq_masked = irq_stat & hpriv->port_map;
-
-	for (i = 0; i < host->n_ports; i++) {
-		struct ata_port *ap;
-
-		if (!(irq_masked & (1 << i)))
-			continue;
-
-		ap = host->ports[i];
-		if (ap) {
-			ahci_hw_port_interrupt(ap);
-			VPRINTK("port %u\n", i);
-		} else {
-			VPRINTK("port %u (no irq)\n", i);
-			if (ata_ratelimit())
-				dev_warn(host->dev,
-					 "interrupt on disabled port %u\n", i);
-		}
-	}
-
-	writel(irq_stat, mmio + HOST_IRQ_STAT);
+	status = readl(port_mmio + PORT_IRQ_STAT);
+	writel(status, port_mmio + PORT_IRQ_STAT);
 
-	spin_unlock(&host->lock);
+	atomic_or(status, &pp->intr_status);
 
 	VPRINTK("EXIT\n");
 
 	return IRQ_WAKE_THREAD;
 }
-EXPORT_SYMBOL_GPL(ahci_hw_interrupt);
 
-irqreturn_t ahci_interrupt(int irq, void *dev_instance)
+static irqreturn_t ahci_single_irq_intr(int irq, void *dev_instance)
 {
 	struct ata_host *host = dev_instance;
 	struct ahci_host_priv *hpriv;
@@ -1938,7 +1885,6 @@ irqreturn_t ahci_interrupt(int irq, void *dev_instance)
 
 	return IRQ_RETVAL(handled);
 }
-EXPORT_SYMBOL_GPL(ahci_interrupt);
 
 unsigned int ahci_qc_issue(struct ata_queued_cmd *qc)
 {
@@ -2057,7 +2003,7 @@ static void ahci_set_aggressive_devslp(struct ata_port *ap, bool sleep)
 
 	devslp = readl(port_mmio + PORT_DEVSLP);
 	if (!(devslp & PORT_DEVSLP_DSP)) {
-		dev_err(ap->host->dev, "port does not support device sleep\n");
+		dev_info(ap->host->dev, "port does not support device sleep\n");
 		return;
 	}
 
@@ -2472,6 +2418,81 @@ void ahci_set_em_messages(struct ahci_host_priv *hpriv,
 }
 EXPORT_SYMBOL_GPL(ahci_set_em_messages);
 
+static int ahci_host_activate_multi_irqs(struct ata_host *host, int irq,
+					 struct scsi_host_template *sht)
+{
+	int i, rc;
+
+	rc = ata_host_start(host);
+	if (rc)
+		return rc;
+
+	for (i = 0; i < host->n_ports; i++) {
+		struct ahci_port_priv *pp = host->ports[i]->private_data;
+
+		/* Do not receive interrupts sent by dummy ports */
+		if (!pp) {
+			disable_irq(irq + i);
+			continue;
+		}
+
+		rc = devm_request_threaded_irq(host->dev, irq + i,
+					       ahci_multi_irqs_intr,
+					       ahci_port_thread_fn, IRQF_SHARED,
+					       pp->irq_desc, host->ports[i]);
+		if (rc)
+			goto out_free_irqs;
+	}
+
+	for (i = 0; i < host->n_ports; i++)
+		ata_port_desc(host->ports[i], "irq %d", irq + i);
+
+	rc = ata_host_register(host, sht);
+	if (rc)
+		goto out_free_all_irqs;
+
+	return 0;
+
+out_free_all_irqs:
+	i = host->n_ports;
+out_free_irqs:
+	for (i--; i >= 0; i--)
+		devm_free_irq(host->dev, irq + i, host->ports[i]);
+
+	return rc;
+}
+
+/**
+ *	ahci_host_activate - start AHCI host, request IRQs and register it
+ *	@host: target ATA host
+ *	@irq: base IRQ number to request
+ *	@sht: scsi_host_template to use when registering the host
+ *
+ *	Similar to ata_host_activate, but requests IRQs according to AHCI-1.1
+ *	when multiple MSIs were allocated. That is one MSI per port, starting
+ *	from @irq.
+ *
+ *	LOCKING:
+ *	Inherited from calling layer (may sleep).
+ *
+ *	RETURNS:
+ *	0 on success, -errno otherwise.
+ */
+int ahci_host_activate(struct ata_host *host, int irq,
+		       struct scsi_host_template *sht)
+{
+	struct ahci_host_priv *hpriv = host->private_data;
+	int rc;
+
+	if (hpriv->flags & AHCI_HFLAG_MULTI_MSI)
+		rc = ahci_host_activate_multi_irqs(host, irq, sht);
+	else
+		rc = ata_host_activate(host, irq, ahci_single_irq_intr,
+				       IRQF_SHARED, sht);
+	return rc;
+}
+EXPORT_SYMBOL_GPL(ahci_host_activate);
+
 MODULE_AUTHOR("Jeff Garzik");
 MODULE_DESCRIPTION("Common AHCI SATA low-level routines");
 MODULE_LICENSE("GPL");
diff --git a/drivers/ata/libahci_platform.c b/drivers/ata/libahci_platform.c
index 5b92c290e6c6..0b03f9056692 100644
--- a/drivers/ata/libahci_platform.c
+++ b/drivers/ata/libahci_platform.c
@@ -49,7 +49,7 @@ static struct scsi_host_template ahci_platform_sht = {
  * RETURNS:
  * 0 on success otherwise a negative error code
  */
-int ahci_platform_enable_phys(struct ahci_host_priv *hpriv)
+static int ahci_platform_enable_phys(struct ahci_host_priv *hpriv)
 {
 	int rc, i;
 
@@ -77,7 +77,6 @@ disable_phys:
 	}
 	return rc;
 }
-EXPORT_SYMBOL_GPL(ahci_platform_enable_phys);
 
 /**
  * ahci_platform_disable_phys - Disable PHYs
@@ -85,7 +84,7 @@ EXPORT_SYMBOL_GPL(ahci_platform_enable_phys);
  *
  * This function disables all PHYs found in hpriv->phys.
  */
-void ahci_platform_disable_phys(struct ahci_host_priv *hpriv)
+static void ahci_platform_disable_phys(struct ahci_host_priv *hpriv)
 {
 	int i;
 
@@ -97,7 +96,6 @@ void ahci_platform_disable_phys(struct ahci_host_priv *hpriv)
 		phy_exit(hpriv->phys[i]);
 	}
 }
-EXPORT_SYMBOL_GPL(ahci_platform_disable_phys);
 
 /**
  * ahci_platform_enable_clks - Enable platform clocks
@@ -495,20 +493,14 @@ int ahci_platform_init_host(struct platform_device *pdev,
 	ahci_init_controller(host);
 	ahci_print_info(host, "platform");
 
-	return ata_host_activate(host, irq, ahci_interrupt, IRQF_SHARED,
-				 &ahci_platform_sht);
+	return ahci_host_activate(host, irq, &ahci_platform_sht);
 }
 EXPORT_SYMBOL_GPL(ahci_platform_init_host);
 
 static void ahci_host_stop(struct ata_host *host)
 {
-	struct device *dev = host->dev;
-	struct ahci_platform_data *pdata = dev_get_platdata(dev);
 	struct ahci_host_priv *hpriv = host->private_data;
 
-	if (pdata && pdata->exit)
-		pdata->exit(dev);
-
 	ahci_platform_disable_resources(hpriv);
 }
 
@@ -592,7 +584,6 @@ EXPORT_SYMBOL_GPL(ahci_platform_resume_host);
  */
 int ahci_platform_suspend(struct device *dev)
 {
-	struct ahci_platform_data *pdata = dev_get_platdata(dev);
 	struct ata_host *host = dev_get_drvdata(dev);
 	struct ahci_host_priv *hpriv = host->private_data;
 	int rc;
@@ -601,19 +592,9 @@ int ahci_platform_suspend(struct device *dev)
 	if (rc)
 		return rc;
 
-	if (pdata && pdata->suspend) {
-		rc = pdata->suspend(dev);
-		if (rc)
-			goto resume_host;
-	}
-
 	ahci_platform_disable_resources(hpriv);
 
 	return 0;
-
-resume_host:
-	ahci_platform_resume_host(dev);
-	return rc;
 }
 EXPORT_SYMBOL_GPL(ahci_platform_suspend);
 
@@ -629,7 +610,6 @@ EXPORT_SYMBOL_GPL(ahci_platform_suspend);
  */
 int ahci_platform_resume(struct device *dev)
 {
-	struct ahci_platform_data *pdata = dev_get_platdata(dev);
 	struct ata_host *host = dev_get_drvdata(dev);
 	struct ahci_host_priv *hpriv = host->private_data;
 	int rc;
@@ -638,12 +618,6 @@ int ahci_platform_resume(struct device *dev)
 	if (rc)
 		return rc;
 
-	if (pdata && pdata->resume) {
-		rc = pdata->resume(dev);
-		if (rc)
-			goto disable_resources;
-	}
-
 	rc = ahci_platform_resume_host(dev);
 	if (rc)
 		goto disable_resources;
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index f3e7b9f894cd..d1a05f9bb91f 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -1043,8 +1043,8 @@ const char *sata_spd_string(unsigned int spd)
  *	None.
  *
  *	RETURNS:
- *	Device type, %ATA_DEV_ATA, %ATA_DEV_ATAPI, %ATA_DEV_PMP or
- *	%ATA_DEV_UNKNOWN the event of failure.
+ *	Device type, %ATA_DEV_ATA, %ATA_DEV_ATAPI, %ATA_DEV_PMP,
+ *	%ATA_DEV_ZAC, or %ATA_DEV_UNKNOWN the event of failure.
  */
 unsigned int ata_dev_classify(const struct ata_taskfile *tf)
 {
@@ -1089,6 +1089,11 @@ unsigned int ata_dev_classify(const struct ata_taskfile *tf)
 		return ATA_DEV_SEMB;
 	}
 
+	if ((tf->lbam == 0xcd) && (tf->lbah == 0xab)) {
+		DPRINTK("found ZAC device by sig\n");
+		return ATA_DEV_ZAC;
+	}
+
 	DPRINTK("unknown device\n");
 	return ATA_DEV_UNKNOWN;
 }
@@ -1329,7 +1334,7 @@ static int ata_hpa_resize(struct ata_device *dev)
 	int rc;
 
 	/* do we need to do it? */
-	if (dev->class != ATA_DEV_ATA ||
+	if ((dev->class != ATA_DEV_ATA && dev->class != ATA_DEV_ZAC) ||
 	    !ata_id_has_lba(dev->id) || !ata_id_hpa_enabled(dev->id) ||
 	    (dev->horkage & ATA_HORKAGE_BROKEN_HPA))
 		return 0;
@@ -1889,6 +1894,7 @@ retry:
 	case ATA_DEV_SEMB:
 		class = ATA_DEV_ATA;	/* some hard drives report SEMB sig */
 	case ATA_DEV_ATA:
+	case ATA_DEV_ZAC:
 		tf.command = ATA_CMD_ID_ATA;
 		break;
 	case ATA_DEV_ATAPI:
@@ -1980,7 +1986,7 @@ retry:
 	rc = -EINVAL;
 	reason = "device reports invalid type";
 
-	if (class == ATA_DEV_ATA) {
+	if (class == ATA_DEV_ATA || class == ATA_DEV_ZAC) {
 		if (!ata_id_is_ata(id) && !ata_id_is_cfa(id))
 			goto err_out;
 		if (ap->host->flags & ATA_HOST_IGNORE_ATA &&
@@ -2015,7 +2021,8 @@ retry:
 			goto retry;
 	}
 
-	if ((flags & ATA_READID_POSTRESET) && class == ATA_DEV_ATA) {
+	if ((flags & ATA_READID_POSTRESET) &&
+	    (class == ATA_DEV_ATA || class == ATA_DEV_ZAC)) {
 		/*
 		 * The exact sequence expected by certain pre-ATA4 drives is:
 		 * SRST RESET
@@ -2280,7 +2287,7 @@ int ata_dev_configure(struct ata_device *dev)
 			sizeof(modelbuf));
 
 	/* ATA-specific feature tests */
-	if (dev->class == ATA_DEV_ATA) {
+	if (dev->class == ATA_DEV_ATA || dev->class == ATA_DEV_ZAC) {
 		if (ata_id_is_cfa(id)) {
 			/* CPRM may make this media unusable */
 			if (id[ATA_ID_CFA_KEY_MGMT] & 1)
@@ -4033,6 +4040,7 @@ int ata_dev_revalidate(struct ata_device *dev, unsigned int new_class,
 	if (ata_class_enabled(new_class) &&
 	    new_class != ATA_DEV_ATA &&
 	    new_class != ATA_DEV_ATAPI &&
+	    new_class != ATA_DEV_ZAC &&
 	    new_class != ATA_DEV_SEMB) {
 		ata_dev_info(dev, "class mismatch %u != %u\n",
 			     dev->class, new_class);
@@ -4225,10 +4233,33 @@ static const struct ata_blacklist_entry ata_device_blacklist [] = {
 	{ "PIONEER DVD-RW  DVR-216D",	NULL,	ATA_HORKAGE_NOSETXFER },
 
 	/* devices that don't properly handle queued TRIM commands */
-	{ "Micron_M500*",		NULL,	ATA_HORKAGE_NO_NCQ_TRIM, },
-	{ "Crucial_CT???M500SSD*",	NULL,	ATA_HORKAGE_NO_NCQ_TRIM, },
-	{ "Micron_M550*",		NULL,	ATA_HORKAGE_NO_NCQ_TRIM, },
-	{ "Crucial_CT*M550SSD*",	NULL,	ATA_HORKAGE_NO_NCQ_TRIM, },
+	{ "Micron_M[56]*",		NULL,	ATA_HORKAGE_NO_NCQ_TRIM |
+						ATA_HORKAGE_ZERO_AFTER_TRIM, },
+	{ "Crucial_CT*SSD*",		NULL,	ATA_HORKAGE_NO_NCQ_TRIM, },
+
+	/*
+	 * As defined, the DRAT (Deterministic Read After Trim) and RZAT
+	 * (Return Zero After Trim) flags in the ATA Command Set are
+	 * unreliable in the sense that they only define what happens if
+	 * the device successfully executed the DSM TRIM command. TRIM
+	 * is only advisory, however, and the device is free to silently
+	 * ignore all or parts of the request.
+	 *
+	 * Whitelist drives that are known to reliably return zeroes
+	 * after TRIM.
+	 */
+
+	/*
+	 * The intel 510 drive has buggy DRAT/RZAT. Explicitly exclude
+	 * that model before whitelisting all other intel SSDs.
+	 */
+	{ "INTEL*SSDSC2MH*",		NULL,	0, },
+
+	{ "INTEL*SSD*", 		NULL,	ATA_HORKAGE_ZERO_AFTER_TRIM, },
+	{ "SSD*INTEL*",			NULL,	ATA_HORKAGE_ZERO_AFTER_TRIM, },
+	{ "Samsung*SSD*",		NULL,	ATA_HORKAGE_ZERO_AFTER_TRIM, },
+	{ "SAMSUNG*SSD*",		NULL,	ATA_HORKAGE_ZERO_AFTER_TRIM, },
+	{ "ST[1248][0248]0[FH]*",	NULL,	ATA_HORKAGE_ZERO_AFTER_TRIM, },
 
 	/*
 	 * Some WD SATA-I drives spin up and down erratically when the link
@@ -4261,10 +4292,10 @@ static unsigned long ata_dev_blacklisted(const struct ata_device *dev)
 	ata_id_c_string(dev->id, model_rev, ATA_ID_FW_REV, sizeof(model_rev));
 
 	while (ad->model_num) {
-		if (glob_match(model_num, ad->model_num)) {
+		if (glob_match(ad->model_num, model_num)) {
 			if (ad->model_rev == NULL)
 				return ad->horkage;
-			if (glob_match(model_rev, ad->model_rev))
+			if (glob_match(ad->model_rev, model_rev))
 				return ad->horkage;
 		}
 		ad++;
@@ -4740,7 +4771,10 @@ static struct ata_queued_cmd *ata_qc_new(struct ata_port *ap)
 		return NULL;
 
 	for (i = 0, tag = ap->last_tag + 1; i < max_queue; i++, tag++) {
-		tag = tag < max_queue ? tag : 0;
+		if (ap->flags & ATA_FLAG_LOWTAG)
+			tag = i;
+		else
+			tag = tag < max_queue ? tag : 0;
 
 		/* the last tag is reserved for internal command. */
 		if (tag == ATA_TAG_INTERNAL)
@@ -6227,7 +6261,7 @@ int ata_host_activate(struct ata_host *host, int irq,
 	}
 
 	rc = devm_request_irq(host->dev, irq, irq_handler, irq_flags,
-			      dev_driver_string(host->dev), host);
+			      dev_name(host->dev), host);
 	if (rc)
 		return rc;
 
@@ -6772,32 +6806,28 @@ const struct ata_port_info ata_dummy_port_info = {
 /*
  * Utility print functions
  */
-int ata_port_printk(const struct ata_port *ap, const char *level,
-		    const char *fmt, ...)
+void ata_port_printk(const struct ata_port *ap, const char *level,
+		     const char *fmt, ...)
 {
 	struct va_format vaf;
 	va_list args;
-	int r;
 
 	va_start(args, fmt);
 
 	vaf.fmt = fmt;
 	vaf.va = &args;
 
-	r = printk("%sata%u: %pV", level, ap->print_id, &vaf);
+	printk("%sata%u: %pV", level, ap->print_id, &vaf);
 
 	va_end(args);
-
-	return r;
 }
 EXPORT_SYMBOL(ata_port_printk);
 
-int ata_link_printk(const struct ata_link *link, const char *level,
-		    const char *fmt, ...)
+void ata_link_printk(const struct ata_link *link, const char *level,
+		     const char *fmt, ...)
 {
 	struct va_format vaf;
 	va_list args;
-	int r;
 
 	va_start(args, fmt);
 
@@ -6805,37 +6835,32 @@ int ata_link_printk(const struct ata_link *link, const char *level,
 	vaf.va = &args;
 
 	if (sata_pmp_attached(link->ap) || link->ap->slave_link)
-		r = printk("%sata%u.%02u: %pV",
-			   level, link->ap->print_id, link->pmp, &vaf);
+		printk("%sata%u.%02u: %pV",
+		       level, link->ap->print_id, link->pmp, &vaf);
 	else
-		r = printk("%sata%u: %pV",
-			   level, link->ap->print_id, &vaf);
+		printk("%sata%u: %pV",
+		       level, link->ap->print_id, &vaf);
 
 	va_end(args);
-
-	return r;
 }
 EXPORT_SYMBOL(ata_link_printk);
 
-int ata_dev_printk(const struct ata_device *dev, const char *level,
+void ata_dev_printk(const struct ata_device *dev, const char *level,
 		    const char *fmt, ...)
 {
 	struct va_format vaf;
 	va_list args;
-	int r;
 
 	va_start(args, fmt);
 
 	vaf.fmt = fmt;
 	vaf.va = &args;
 
-	r = printk("%sata%u.%02u: %pV",
-		   level, dev->link->ap->print_id, dev->link->pmp + dev->devno,
-		   &vaf);
+	printk("%sata%u.%02u: %pV",
+	       level, dev->link->ap->print_id, dev->link->pmp + dev->devno,
+	       &vaf);
 
 	va_end(args);
-
-	return r;
 }
 EXPORT_SYMBOL(ata_dev_printk);
 
diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c
index dad83df555c4..8d00c2638bed 100644
--- a/drivers/ata/libata-eh.c
+++ b/drivers/ata/libata-eh.c
@@ -1809,6 +1809,7 @@ static unsigned int ata_eh_analyze_tf(struct ata_queued_cmd *qc,
 
 	switch (qc->dev->class) {
 	case ATA_DEV_ATA:
+	case ATA_DEV_ZAC:
 		if (err & ATA_ICRC)
 			qc->err_mask |= AC_ERR_ATA_BUS;
 		if (err & (ATA_UNC | ATA_AMNF))
@@ -2388,6 +2389,7 @@ const char *ata_get_cmd_descript(u8 command)
 
 	return NULL;
 }
+EXPORT_SYMBOL_GPL(ata_get_cmd_descript);
 
 /**
  *	ata_eh_link_report - report error handling to user
@@ -3792,7 +3794,8 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset,
 				struct ata_eh_context *ehc = &link->eh_context;
 				unsigned long tmp;
 
-				if (dev->class != ATA_DEV_ATA)
+				if (dev->class != ATA_DEV_ATA &&
+				    dev->class != ATA_DEV_ZAC)
 					continue;
 				if (!(ehc->i.dev_action[dev->devno] &
 				      ATA_EH_PARK))
@@ -3873,7 +3876,8 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset,
 
 		/* retry flush if necessary */
 		ata_for_each_dev(dev, link, ALL) {
-			if (dev->class != ATA_DEV_ATA)
+			if (dev->class != ATA_DEV_ATA &&
+			    dev->class != ATA_DEV_ZAC)
 				continue;
 			rc = ata_eh_maybe_retry_flush(dev);
 			if (rc)
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index 0586f66d70fa..6abd17a85b13 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -235,7 +235,8 @@ static ssize_t ata_scsi_park_store(struct device *device,
 		rc = -ENODEV;
 		goto unlock;
 	}
-	if (dev->class != ATA_DEV_ATA) {
+	if (dev->class != ATA_DEV_ATA &&
+	    dev->class != ATA_DEV_ZAC) {
 		rc = -EOPNOTSUPP;
 		goto unlock;
 	}
@@ -1164,7 +1165,7 @@ static int ata_scsi_dev_config(struct scsi_device *sdev,
 
 		depth = min(sdev->host->can_queue, ata_id_queue_depth(dev->id));
 		depth = min(ATA_MAX_QUEUE - 1, depth);
-		scsi_adjust_queue_depth(sdev, MSG_SIMPLE_TAG, depth);
+		scsi_change_queue_depth(sdev, depth);
 	}
 
 	blk_queue_flush_queueable(q, false);
@@ -1243,21 +1244,17 @@ void ata_scsi_slave_destroy(struct scsi_device *sdev)
  *	@ap: ATA port to which the device change the queue depth
  *	@sdev: SCSI device to configure queue depth for
  *	@queue_depth: new queue depth
- *	@reason: calling context
  *
  *	libsas and libata have different approaches for associating a sdev to
  *	its ata_port.
  *
  */
 int __ata_change_queue_depth(struct ata_port *ap, struct scsi_device *sdev,
-			     int queue_depth, int reason)
+			     int queue_depth)
 {
 	struct ata_device *dev;
 	unsigned long flags;
 
-	if (reason != SCSI_QDEPTH_DEFAULT)
-		return -EOPNOTSUPP;
-
 	if (queue_depth < 1 || queue_depth == sdev->queue_depth)
 		return sdev->queue_depth;
 
@@ -1282,15 +1279,13 @@ int __ata_change_queue_depth(struct ata_port *ap, struct scsi_device *sdev,
 	if (sdev->queue_depth == queue_depth)
 		return -EINVAL;
 
-	scsi_adjust_queue_depth(sdev, MSG_SIMPLE_TAG, queue_depth);
-	return queue_depth;
+	return scsi_change_queue_depth(sdev, queue_depth);
 }
 
 /**
  *	ata_scsi_change_queue_depth - SCSI callback for queue depth config
  *	@sdev: SCSI device to configure queue depth for
  *	@queue_depth: new queue depth
- *	@reason: calling context
  *
  *	This is libata standard hostt->change_queue_depth callback.
  *	SCSI will call into this callback when user tries to set queue
@@ -1302,12 +1297,11 @@ int __ata_change_queue_depth(struct ata_port *ap, struct scsi_device *sdev,
  *	RETURNS:
  *	Newly configured queue depth.
  */
-int ata_scsi_change_queue_depth(struct scsi_device *sdev, int queue_depth,
-				int reason)
+int ata_scsi_change_queue_depth(struct scsi_device *sdev, int queue_depth)
 {
 	struct ata_port *ap = ata_shost_to_port(sdev->host);
 
-	return __ata_change_queue_depth(ap, sdev, queue_depth, reason);
+	return __ata_change_queue_depth(ap, sdev, queue_depth);
 }
 
 /**
@@ -1968,6 +1962,7 @@ static void ata_scsi_rbuf_fill(struct ata_scsi_args *args,
 static unsigned int ata_scsiop_inq_std(struct ata_scsi_args *args, u8 *rbuf)
 {
 	const u8 versions[] = {
+		0x00,
 		0x60,	/* SAM-3 (no version claimed) */
 
 		0x03,
@@ -1976,6 +1971,20 @@ static unsigned int ata_scsiop_inq_std(struct ata_scsi_args *args, u8 *rbuf)
 		0x02,
 		0x60	/* SPC-3 (no version claimed) */
 	};
+	const u8 versions_zbc[] = {
+		0x00,
+		0xA0,	/* SAM-5 (no version claimed) */
+
+		0x04,
+		0xC0,	/* SBC-3 (no version claimed) */
+
+		0x04,
+		0x60,	/* SPC-4 (no version claimed) */
+
+		0x60,
+		0x20,   /* ZBC (no version claimed) */
+	};
+
 	u8 hdr[] = {
 		TYPE_DISK,
 		0,
@@ -1990,6 +1999,11 @@ static unsigned int ata_scsiop_inq_std(struct ata_scsi_args *args, u8 *rbuf)
 	if (ata_id_removeable(args->id))
 		hdr[1] |= (1 << 7);
 
+	if (args->dev->class == ATA_DEV_ZAC) {
+		hdr[0] = TYPE_ZBC;
+		hdr[2] = 0x6; /* ZBC is defined in SPC-4 */
+	}
+
 	memcpy(rbuf, hdr, sizeof(hdr));
 	memcpy(&rbuf[8], "ATA     ", 8);
 	ata_id_string(args->id, &rbuf[16], ATA_ID_PROD, 16);
@@ -2002,7 +2016,10 @@ static unsigned int ata_scsiop_inq_std(struct ata_scsi_args *args, u8 *rbuf)
 	if (rbuf[32] == 0 || rbuf[32] == ' ')
 		memcpy(&rbuf[32], "n/a ", 4);
 
-	memcpy(rbuf + 59, versions, sizeof(versions));
+	if (args->dev->class == ATA_DEV_ZAC)
+		memcpy(rbuf + 58, versions_zbc, sizeof(versions_zbc));
+	else
+		memcpy(rbuf + 58, versions, sizeof(versions));
 
 	return 0;
 }
@@ -2515,13 +2532,15 @@ static unsigned int ata_scsiop_read_cap(struct ata_scsi_args *args, u8 *rbuf)
 		rbuf[15] = lowest_aligned;
 
 		if (ata_id_has_trim(args->id)) {
-			rbuf[14] |= 0x80; /* TPE */
+			rbuf[14] |= 0x80; /* LBPME */
 
-			if (ata_id_has_zero_after_trim(args->id))
-				rbuf[14] |= 0x40; /* TPRZ */
+			if (ata_id_has_zero_after_trim(args->id) &&
+			    dev->horkage & ATA_HORKAGE_ZERO_AFTER_TRIM) {
+				ata_dev_info(dev, "Enabling discard_zeroes_data\n");
+				rbuf[14] |= 0x40; /* LBPRZ */
+			}
 		}
 	}
-
 	return 0;
 }
 
@@ -2571,7 +2590,6 @@ static void atapi_request_sense(struct ata_queued_cmd *qc)
 
 	DPRINTK("ATAPI request sense\n");
 
-	/* FIXME: is this needed? */
 	memset(cmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE);
 
 #ifdef CONFIG_ATA_SFF
@@ -3412,7 +3430,7 @@ static inline int __ata_scsi_queuecmd(struct scsi_cmnd *scmd,
 	ata_xlat_func_t xlat_func;
 	int rc = 0;
 
-	if (dev->class == ATA_DEV_ATA) {
+	if (dev->class == ATA_DEV_ATA || dev->class == ATA_DEV_ZAC) {
 		if (unlikely(!scmd->cmd_len || scmd->cmd_len > dev->cdb_len))
 			goto bad_cdb_len;
 
@@ -3570,7 +3588,7 @@ void ata_scsi_simulate(struct ata_device *dev, struct scsi_cmnd *cmd)
 		ata_scsi_rbuf_fill(&args, ata_scsiop_read_cap);
 		break;
 
-	case SERVICE_ACTION_IN:
+	case SERVICE_ACTION_IN_16:
 		if ((scsicmd[1] & 0x1f) == SAI_READ_CAPACITY_16)
 			ata_scsi_rbuf_fill(&args, ata_scsiop_read_cap);
 		else
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index 1121153f1ecd..2e86e3b85266 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -1333,7 +1333,19 @@ void ata_sff_flush_pio_task(struct ata_port *ap)
 	DPRINTK("ENTER\n");
 
 	cancel_delayed_work_sync(&ap->sff_pio_task);
+
+	/*
+	 * We wanna reset the HSM state to IDLE.  If we do so without
+	 * grabbing the port lock, critical sections protected by it which
+	 * expect the HSM state to stay stable may get surprised.  For
+	 * example, we may set IDLE in between the time
+	 * __ata_sff_port_intr() checks for HSM_ST_IDLE and before it calls
+	 * ata_sff_hsm_move() causing ata_sff_hsm_move() to BUG().
+	 */
+	spin_lock_irq(ap->lock);
 	ap->hsm_task_state = HSM_ST_IDLE;
+	spin_unlock_irq(ap->lock);
+
 	ap->sff_pio_task_link = NULL;
 
 	if (ata_msg_ctl(ap))
@@ -2008,13 +2020,15 @@ static int ata_bus_softreset(struct ata_port *ap, unsigned int devmask,
 
 	DPRINTK("ata%u: bus reset via SRST\n", ap->print_id);
 
-	/* software reset.  causes dev0 to be selected */
-	iowrite8(ap->ctl, ioaddr->ctl_addr);
-	udelay(20);	/* FIXME: flush */
-	iowrite8(ap->ctl | ATA_SRST, ioaddr->ctl_addr);
-	udelay(20);	/* FIXME: flush */
-	iowrite8(ap->ctl, ioaddr->ctl_addr);
-	ap->last_ctl = ap->ctl;
+	if (ap->ioaddr.ctl_addr) {
+		/* software reset.  causes dev0 to be selected */
+		iowrite8(ap->ctl, ioaddr->ctl_addr);
+		udelay(20);	/* FIXME: flush */
+		iowrite8(ap->ctl | ATA_SRST, ioaddr->ctl_addr);
+		udelay(20);	/* FIXME: flush */
+		iowrite8(ap->ctl, ioaddr->ctl_addr);
+		ap->last_ctl = ap->ctl;
+	}
 
 	/* wait the port to become ready */
 	return ata_sff_wait_after_reset(&ap->link, devmask, deadline);
@@ -2215,10 +2229,6 @@ void ata_sff_error_handler(struct ata_port *ap)
 
 	spin_unlock_irqrestore(ap->lock, flags);
 
-	/* ignore ata_sff_softreset if ctl isn't accessible */
-	if (softreset == ata_sff_softreset && !ap->ioaddr.ctl_addr)
-		softreset = NULL;
-
 	/* ignore built-in hardresets if SCR access is not available */
 	if ((hardreset == sata_std_hardreset ||
 	     hardreset == sata_sff_hardreset) && !sata_scr_valid(&ap->link))
diff --git a/drivers/ata/libata-transport.c b/drivers/ata/libata-transport.c
index e37413228228..3227b7c8a05f 100644
--- a/drivers/ata/libata-transport.c
+++ b/drivers/ata/libata-transport.c
@@ -143,6 +143,7 @@ static struct {
 	{ ATA_DEV_PMP_UNSUP,		"pmp" },
 	{ ATA_DEV_SEMB,			"semb" },
 	{ ATA_DEV_SEMB_UNSUP,		"semb" },
+	{ ATA_DEV_ZAC,			"zac" },
 	{ ATA_DEV_NONE,			"none" }
 };
 ata_bitfield_name_search(class, ata_class_names)
diff --git a/drivers/ata/pata_arasan_cf.c b/drivers/ata/pata_arasan_cf.c
index 4edb1a81f63f..a9b0c820f2eb 100644
--- a/drivers/ata/pata_arasan_cf.c
+++ b/drivers/ata/pata_arasan_cf.c
@@ -420,7 +420,7 @@ dma_xfer(struct arasan_cf_dev *acdev, dma_addr_t src, dma_addr_t dest, u32 len)
 
 	/* Wait for DMA to complete */
 	if (!wait_for_completion_timeout(&acdev->dma_completion, TIMEOUT)) {
-		chan->device->device_control(chan, DMA_TERMINATE_ALL, 0);
+		dmaengine_terminate_all(chan);
 		dev_err(acdev->host->dev, "wait_for_completion_timeout\n");
 		return -ETIMEDOUT;
 	}
@@ -928,8 +928,7 @@ static int arasan_cf_suspend(struct device *dev)
 	struct arasan_cf_dev *acdev = host->ports[0]->private_data;
 
 	if (acdev->dma_chan)
-		acdev->dma_chan->device->device_control(acdev->dma_chan,
-				DMA_TERMINATE_ALL, 0);
+		dmaengine_terminate_all(acdev->dma_chan);
 
 	cf_exit(acdev);
 	return ata_host_suspend(host, PMSG_SUSPEND);
@@ -962,7 +961,6 @@ static struct platform_driver arasan_cf_driver = {
 	.remove		= arasan_cf_remove,
 	.driver		= {
 		.name	= DRIVER_NAME,
-		.owner	= THIS_MODULE,
 		.pm	= &arasan_cf_pm_ops,
 		.of_match_table = of_match_ptr(arasan_cf_id_table),
 	},
diff --git a/drivers/ata/pata_at32.c b/drivers/ata/pata_at32.c
index d59d5239405f..9aeb7a6dd4d4 100644
--- a/drivers/ata/pata_at32.c
+++ b/drivers/ata/pata_at32.c
@@ -389,7 +389,6 @@ static struct platform_driver pata_at32_driver = {
 	.remove	       = __exit_p(pata_at32_remove),
 	.driver	       = {
 		.name  = "at32_ide",
-		.owner = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/ata/pata_at91.c b/drivers/ata/pata_at91.c
index 8a66f23af4c4..9e85937d36a9 100644
--- a/drivers/ata/pata_at91.c
+++ b/drivers/ata/pata_at91.c
@@ -444,7 +444,6 @@ static struct platform_driver pata_at91_driver = {
 	.remove		= pata_at91_remove,
 	.driver		= {
 		.name		= DRV_NAME,
-		.owner		= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/ata/pata_bf54x.c b/drivers/ata/pata_bf54x.c
index 03f2f2bc83bd..dd7410019d15 100644
--- a/drivers/ata/pata_bf54x.c
+++ b/drivers/ata/pata_bf54x.c
@@ -1657,7 +1657,6 @@ static struct platform_driver bfin_atapi_driver = {
 	.resume			= bfin_atapi_resume,
 	.driver = {
 		.name		= DRV_NAME,
-		.owner		= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/ata/pata_ep93xx.c b/drivers/ata/pata_ep93xx.c
index 4d37c5415fc7..bd6b089c67a3 100644
--- a/drivers/ata/pata_ep93xx.c
+++ b/drivers/ata/pata_ep93xx.c
@@ -1021,7 +1021,6 @@ static int ep93xx_pata_remove(struct platform_device *pdev)
 static struct platform_driver ep93xx_pata_platform_driver = {
 	.driver = {
 		.name = DRV_NAME,
-		.owner = THIS_MODULE,
 	},
 	.probe = ep93xx_pata_probe,
 	.remove = ep93xx_pata_remove,
diff --git a/drivers/ata/pata_imx.c b/drivers/ata/pata_imx.c
index af424573c2ff..139d20778b29 100644
--- a/drivers/ata/pata_imx.c
+++ b/drivers/ata/pata_imx.c
@@ -221,13 +221,10 @@ static int pata_imx_resume(struct device *dev)
 
 	return 0;
 }
-
-static const struct dev_pm_ops pata_imx_pm_ops = {
-	.suspend	= pata_imx_suspend,
-	.resume		= pata_imx_resume,
-};
 #endif
 
+static SIMPLE_DEV_PM_OPS(pata_imx_pm_ops, pata_imx_suspend, pata_imx_resume);
+
 static const struct of_device_id imx_pata_dt_ids[] = {
 	{
 		.compatible = "fsl,imx27-pata",
@@ -243,10 +240,7 @@ static struct platform_driver pata_imx_driver = {
 	.driver = {
 		.name		= DRV_NAME,
 		.of_match_table	= imx_pata_dt_ids,
-		.owner		= THIS_MODULE,
-#ifdef CONFIG_PM_SLEEP
 		.pm		= &pata_imx_pm_ops,
-#endif
 	},
 };
 
diff --git a/drivers/ata/pata_ixp4xx_cf.c b/drivers/ata/pata_ixp4xx_cf.c
index ddf470c2341d..abda44183512 100644
--- a/drivers/ata/pata_ixp4xx_cf.c
+++ b/drivers/ata/pata_ixp4xx_cf.c
@@ -193,7 +193,6 @@ static int ixp4xx_pata_probe(struct platform_device *pdev)
 static struct platform_driver ixp4xx_pata_platform_driver = {
 	.driver	 = {
 		.name   = DRV_NAME,
-		.owner  = THIS_MODULE,
 	},
 	.probe		= ixp4xx_pata_probe,
 	.remove		= ata_platform_remove_one,
diff --git a/drivers/ata/pata_mpc52xx.c b/drivers/ata/pata_mpc52xx.c
index ccd1c83a05cc..252ba27fa63b 100644
--- a/drivers/ata/pata_mpc52xx.c
+++ b/drivers/ata/pata_mpc52xx.c
@@ -863,7 +863,6 @@ static struct platform_driver mpc52xx_ata_of_platform_driver = {
 #endif
 	.driver		= {
 		.name	= DRV_NAME,
-		.owner	= THIS_MODULE,
 		.of_match_table = mpc52xx_ata_of_match,
 	},
 };
diff --git a/drivers/ata/pata_octeon_cf.c b/drivers/ata/pata_octeon_cf.c
index 2a97d3a531ec..80a80548ad0a 100644
--- a/drivers/ata/pata_octeon_cf.c
+++ b/drivers/ata/pata_octeon_cf.c
@@ -1059,7 +1059,6 @@ static struct platform_driver octeon_cf_driver = {
 	.probe		= octeon_cf_probe,
 	.driver		= {
 		.name	= DRV_NAME,
-		.owner	= THIS_MODULE,
 		.of_match_table = octeon_cf_match,
 		.shutdown = octeon_cf_shutdown
 	},
diff --git a/drivers/ata/pata_of_platform.c b/drivers/ata/pata_of_platform.c
index a7e95a54c782..dcc408abe171 100644
--- a/drivers/ata/pata_of_platform.c
+++ b/drivers/ata/pata_of_platform.c
@@ -35,25 +35,14 @@ static int pata_of_platform_probe(struct platform_device *ofdev)
 		return -EINVAL;
 	}
 
-	if (of_device_is_compatible(dn, "electra-ide")) {
-		/* Altstatus is really at offset 0x3f6 from the primary window
-		 * on electra-ide. Adjust ctl_res and io_res accordingly.
-		 */
-		ctl_res = io_res;
-		ctl_res.start = ctl_res.start+0x3f6;
-		io_res.end = ctl_res.start-1;
-	} else {
-		ret = of_address_to_resource(dn, 1, &ctl_res);
-		if (ret) {
-			dev_err(&ofdev->dev, "can't get CTL address from "
-				"device tree\n");
-			return -EINVAL;
-		}
+	ret = of_address_to_resource(dn, 1, &ctl_res);
+	if (ret) {
+		dev_err(&ofdev->dev, "can't get CTL address from "
+			"device tree\n");
+		return -EINVAL;
 	}
 
 	irq_res = platform_get_resource(ofdev, IORESOURCE_IRQ, 0);
-	if (irq_res)
-		irq_res->flags = 0;
 
 	prop = of_get_property(dn, "reg-shift", NULL);
 	if (prop)
@@ -79,15 +68,13 @@ static int pata_of_platform_probe(struct platform_device *ofdev)
 
 static struct of_device_id pata_of_platform_match[] = {
 	{ .compatible = "ata-generic", },
-	{ .compatible = "electra-ide", },
-	{},
+	{ },
 };
 MODULE_DEVICE_TABLE(of, pata_of_platform_match);
 
 static struct platform_driver pata_of_platform_driver = {
 	.driver = {
 		.name = "pata_of_platform",
-		.owner = THIS_MODULE,
 		.of_match_table = pata_of_platform_match,
 	},
 	.probe		= pata_of_platform_probe,
diff --git a/drivers/ata/pata_palmld.c b/drivers/ata/pata_palmld.c
index df2bb7504fc8..8c0d7d736b7a 100644
--- a/drivers/ata/pata_palmld.c
+++ b/drivers/ata/pata_palmld.c
@@ -124,7 +124,6 @@ static int palmld_pata_remove(struct platform_device *dev)
 static struct platform_driver palmld_pata_platform_driver = {
 	.driver	 = {
 		.name   = DRV_NAME,
-		.owner  = THIS_MODULE,
 	},
 	.probe		= palmld_pata_probe,
 	.remove		= palmld_pata_remove,
diff --git a/drivers/ata/pata_platform.c b/drivers/ata/pata_platform.c
index a5579b55e332..1eedfe46d7c8 100644
--- a/drivers/ata/pata_platform.c
+++ b/drivers/ata/pata_platform.c
@@ -118,7 +118,7 @@ int __pata_platform_probe(struct device *dev, struct resource *io_res,
 	 */
 	if (irq_res && irq_res->start > 0) {
 		irq = irq_res->start;
-		irq_flags = irq_res->flags;
+		irq_flags = irq_res->flags & IRQF_TRIGGER_MASK;
 	}
 
 	/*
@@ -213,8 +213,6 @@ static int pata_platform_probe(struct platform_device *pdev)
 	 * And the IRQ
 	 */
 	irq_res = platform_get_resource(pdev, IORESOURCE_IRQ, 0);
-	if (irq_res)
-		irq_res->flags = pp_info ? pp_info->irq_flags : 0;
 
 	return __pata_platform_probe(&pdev->dev, io_res, ctl_res, irq_res,
 				     pp_info ? pp_info->ioport_shift : 0,
@@ -226,7 +224,6 @@ static struct platform_driver pata_platform_driver = {
 	.remove		= ata_platform_remove_one,
 	.driver = {
 		.name		= DRV_NAME,
-		.owner		= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/ata/pata_pxa.c b/drivers/ata/pata_pxa.c
index 73259bfda1e3..c36b3e6531d8 100644
--- a/drivers/ata/pata_pxa.c
+++ b/drivers/ata/pata_pxa.c
@@ -385,7 +385,6 @@ static struct platform_driver pxa_ata_driver = {
 	.remove		= pxa_ata_remove,
 	.driver		= {
 		.name		= DRV_NAME,
-		.owner		= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/ata/pata_rb532_cf.c b/drivers/ata/pata_rb532_cf.c
index 3c5eb8fa6bd1..6d08446b877c 100644
--- a/drivers/ata/pata_rb532_cf.c
+++ b/drivers/ata/pata_rb532_cf.c
@@ -193,7 +193,6 @@ static struct platform_driver rb532_pata_platform_driver = {
 	.remove		= rb532_pata_driver_remove,
 	.driver	 = {
 		.name   = DRV_NAME,
-		.owner  = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/ata/pata_samsung_cf.c b/drivers/ata/pata_samsung_cf.c
index 1a24a5dc3940..fa44eb2872db 100644
--- a/drivers/ata/pata_samsung_cf.c
+++ b/drivers/ata/pata_samsung_cf.c
@@ -656,7 +656,6 @@ static struct platform_driver pata_s3c_driver = {
 	.id_table	= pata_s3c_driver_ids,
 	.driver		= {
 		.name	= DRV_NAME,
-		.owner	= THIS_MODULE,
 #ifdef CONFIG_PM_SLEEP
 		.pm	= &pata_s3c_pm_ops,
 #endif
diff --git a/drivers/ata/pata_serverworks.c b/drivers/ata/pata_serverworks.c
index fc5f31d4828e..57de02123c4c 100644
--- a/drivers/ata/pata_serverworks.c
+++ b/drivers/ata/pata_serverworks.c
@@ -251,12 +251,18 @@ static void serverworks_set_dmamode(struct ata_port *ap, struct ata_device *adev
 	pci_write_config_byte(pdev, 0x54, ultra_cfg);
 }
 
-static struct scsi_host_template serverworks_sht = {
+static struct scsi_host_template serverworks_osb4_sht = {
+	ATA_BMDMA_SHT(DRV_NAME),
+	.sg_tablesize	= LIBATA_DUMB_MAX_PRD,
+};
+
+static struct scsi_host_template serverworks_csb_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations serverworks_osb4_port_ops = {
 	.inherits	= &ata_bmdma_port_ops,
+	.qc_prep	= ata_bmdma_dumb_qc_prep,
 	.cable_detect	= serverworks_cable_detect,
 	.mode_filter	= serverworks_osb4_filter,
 	.set_piomode	= serverworks_set_piomode,
@@ -265,6 +271,7 @@ static struct ata_port_operations serverworks_osb4_port_ops = {
 
 static struct ata_port_operations serverworks_csb_port_ops = {
 	.inherits	= &serverworks_osb4_port_ops,
+	.qc_prep	= ata_bmdma_qc_prep,
 	.mode_filter	= serverworks_csb_filter,
 };
 
@@ -404,6 +411,7 @@ static int serverworks_init_one(struct pci_dev *pdev, const struct pci_device_id
 		}
 	};
 	const struct ata_port_info *ppi[] = { &info[id->driver_data], NULL };
+	struct scsi_host_template *sht = &serverworks_csb_sht;
 	int rc;
 
 	rc = pcim_enable_device(pdev);
@@ -417,6 +425,7 @@ static int serverworks_init_one(struct pci_dev *pdev, const struct pci_device_id
 		/* Select non UDMA capable OSB4 if we can't do fixups */
 		if (rc < 0)
 			ppi[0] = &info[1];
+		sht = &serverworks_osb4_sht;
 	}
 	/* setup CSB5/CSB6 : South Bridge and IDE option RAID */
 	else if ((pdev->device == PCI_DEVICE_ID_SERVERWORKS_CSB5IDE) ||
@@ -433,7 +442,7 @@ static int serverworks_init_one(struct pci_dev *pdev, const struct pci_device_id
 			ppi[1] = &ata_dummy_port_info;
 	}
 
-	return ata_pci_bmdma_init_one(pdev, ppi, &serverworks_sht, NULL, 0);
+	return ata_pci_bmdma_init_one(pdev, ppi, sht, NULL, 0);
 }
 
 #ifdef CONFIG_PM_SLEEP
diff --git a/drivers/ata/sata_dwc_460ex.c b/drivers/ata/sata_dwc_460ex.c
index 0bb2cabd2197..8e8248179d20 100644
--- a/drivers/ata/sata_dwc_460ex.c
+++ b/drivers/ata/sata_dwc_460ex.c
@@ -797,7 +797,7 @@ static int dma_dwc_init(struct sata_dwc_device *hsdev, int irq)
 	if (err) {
 		dev_err(host_pvt.dwc_dev, "%s: dma_request_interrupts returns"
 			" %d\n", __func__, err);
-		goto error_out;
+		return err;
 	}
 
 	/* Enabe DMA */
@@ -808,11 +808,6 @@ static int dma_dwc_init(struct sata_dwc_device *hsdev, int irq)
 		sata_dma_regs);
 
 	return 0;
-
-error_out:
-	dma_dwc_exit(hsdev);
-
-	return err;
 }
 
 static int sata_dwc_scr_read(struct ata_link *link, unsigned int scr, u32 *val)
@@ -1662,7 +1657,7 @@ static int sata_dwc_probe(struct platform_device *ofdev)
 	char *ver = (char *)&versionr;
 	u8 *base = NULL;
 	int err = 0;
-	int irq, rc;
+	int irq;
 	struct ata_host *host;
 	struct ata_port_info pi = sata_dwc_port_info[0];
 	const struct ata_port_info *ppi[] = { &pi, NULL };
@@ -1725,7 +1720,7 @@ static int sata_dwc_probe(struct platform_device *ofdev)
 	if (irq == NO_IRQ) {
 		dev_err(&ofdev->dev, "no SATA DMA irq\n");
 		err = -ENODEV;
-		goto error_out;
+		goto error_iomap;
 	}
 
 	/* Get physical SATA DMA register base address */
@@ -1734,14 +1729,16 @@ static int sata_dwc_probe(struct platform_device *ofdev)
 		dev_err(&ofdev->dev, "ioremap failed for AHBDMA register"
 			" address\n");
 		err = -ENODEV;
-		goto error_out;
+		goto error_iomap;
 	}
 
 	/* Save dev for later use in dev_xxx() routines */
 	host_pvt.dwc_dev = &ofdev->dev;
 
 	/* Initialize AHB DMAC */
-	dma_dwc_init(hsdev, irq);
+	err = dma_dwc_init(hsdev, irq);
+	if (err)
+		goto error_dma_iomap;
 
 	/* Enable SATA Interrupts */
 	sata_dwc_enable_interrupts(hsdev);
@@ -1759,9 +1756,8 @@ static int sata_dwc_probe(struct platform_device *ofdev)
 	 * device discovery process, invoking our port_start() handler &
 	 * error_handler() to execute a dummy Softreset EH session
 	 */
-	rc = ata_host_activate(host, irq, sata_dwc_isr, 0, &sata_dwc_sht);
-
-	if (rc != 0)
+	err = ata_host_activate(host, irq, sata_dwc_isr, 0, &sata_dwc_sht);
+	if (err)
 		dev_err(&ofdev->dev, "failed to activate host");
 
 	dev_set_drvdata(&ofdev->dev, host);
@@ -1770,7 +1766,8 @@ static int sata_dwc_probe(struct platform_device *ofdev)
 error_out:
 	/* Free SATA DMA resources */
 	dma_dwc_exit(hsdev);
-
+error_dma_iomap:
+	iounmap((void __iomem *)host_pvt.sata_dma_regs);
 error_iomap:
 	iounmap(base);
 error_kmalloc:
@@ -1791,6 +1788,7 @@ static int sata_dwc_remove(struct platform_device *ofdev)
 	/* Free SATA DMA resources */
 	dma_dwc_exit(hsdev);
 
+	iounmap((void __iomem *)host_pvt.sata_dma_regs);
 	iounmap(hsdev->reg_base);
 	kfree(hsdev);
 	kfree(host);
@@ -1807,7 +1805,6 @@ MODULE_DEVICE_TABLE(of, sata_dwc_match);
 static struct platform_driver sata_dwc_driver = {
 	.driver = {
 		.name = DRV_NAME,
-		.owner = THIS_MODULE,
 		.of_match_table = sata_dwc_match,
 	},
 	.probe = sata_dwc_probe,
diff --git a/drivers/ata/sata_fsl.c b/drivers/ata/sata_fsl.c
index 07bc7e4dbd04..f9054cd36a72 100644
--- a/drivers/ata/sata_fsl.c
+++ b/drivers/ata/sata_fsl.c
@@ -1488,7 +1488,7 @@ static int sata_fsl_probe(struct platform_device *ofdev)
 	host_priv->csr_base = csr_base;
 
 	irq = irq_of_parse_and_map(ofdev->dev.of_node, 0);
-	if (irq < 0) {
+	if (!irq) {
 		dev_err(&ofdev->dev, "invalid irq from platform\n");
 		goto error_exit_with_cleanup;
 	}
@@ -1624,7 +1624,6 @@ MODULE_DEVICE_TABLE(of, fsl_sata_match);
 static struct platform_driver fsl_sata_driver = {
 	.driver = {
 		.name = "fsl-sata",
-		.owner = THIS_MODULE,
 		.of_match_table = fsl_sata_match,
 	},
 	.probe		= sata_fsl_probe,
diff --git a/drivers/ata/sata_highbank.c b/drivers/ata/sata_highbank.c
index da3bc2709c63..24e311fe2c1c 100644
--- a/drivers/ata/sata_highbank.c
+++ b/drivers/ata/sata_highbank.c
@@ -568,8 +568,7 @@ static int ahci_highbank_probe(struct platform_device *pdev)
 	ahci_init_controller(host);
 	ahci_print_info(host, "platform");
 
-	rc = ata_host_activate(host, irq, ahci_interrupt, 0,
-					&ahci_highbank_platform_sht);
+	rc = ahci_host_activate(host, irq, &ahci_highbank_platform_sht);
 	if (rc)
 		goto err0;
 
@@ -635,7 +634,6 @@ static struct platform_driver ahci_highbank_driver = {
 	.remove = ata_platform_remove_one,
         .driver = {
                 .name = "highbank-ahci",
-                .owner = THIS_MODULE,
                 .of_match_table = ahci_of_match,
                 .pm = &ahci_highbank_pm_ops,
         },
diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c
index 391cfda1e83f..f9a0e34eb111 100644
--- a/drivers/ata/sata_mv.c
+++ b/drivers/ata/sata_mv.c
@@ -4280,7 +4280,6 @@ static struct platform_driver mv_platform_driver = {
 	.resume		= mv_platform_resume,
 	.driver		= {
 		.name = DRV_NAME,
-		.owner = THIS_MODULE,
 		.of_match_table = of_match_ptr(mv_sata_dt_ids),
 	},
 };
diff --git a/drivers/ata/sata_nv.c b/drivers/ata/sata_nv.c
index cdf99fac139a..1db6f5ce5e89 100644
--- a/drivers/ata/sata_nv.c
+++ b/drivers/ata/sata_nv.c
@@ -1951,7 +1951,7 @@ static int nv_swncq_slave_config(struct scsi_device *sdev)
 	ata_id_c_string(dev->id, model_num, ATA_ID_PROD, sizeof(model_num));
 
 	if (strncmp(model_num, "Maxtor", 6) == 0) {
-		ata_scsi_change_queue_depth(sdev, 1, SCSI_QDEPTH_DEFAULT);
+		ata_scsi_change_queue_depth(sdev, 1);
 		ata_dev_notice(dev, "Disabling SWNCQ mode (depth %x)\n",
 			       sdev->queue_depth);
 	}
diff --git a/drivers/ata/sata_rcar.c b/drivers/ata/sata_rcar.c
index 61eb6d77dac7..cb0d2e644af5 100644
--- a/drivers/ata/sata_rcar.c
+++ b/drivers/ata/sata_rcar.c
@@ -146,6 +146,7 @@
 enum sata_rcar_type {
 	RCAR_GEN1_SATA,
 	RCAR_GEN2_SATA,
+	RCAR_R8A7790_ES1_SATA,
 };
 
 struct sata_rcar_priv {
@@ -763,6 +764,9 @@ static void sata_rcar_setup_port(struct ata_host *host)
 	ap->udma_mask	= ATA_UDMA6;
 	ap->flags	|= ATA_FLAG_SATA;
 
+	if (priv->type == RCAR_R8A7790_ES1_SATA)
+		ap->flags	|= ATA_FLAG_NO_DIPM;
+
 	ioaddr->cmd_addr = base + SDATA_REG;
 	ioaddr->ctl_addr = base + SSDEVCON_REG;
 	ioaddr->scr_addr = base + SCRSSTS_REG;
@@ -792,6 +796,7 @@ static void sata_rcar_init_controller(struct ata_host *host)
 		sata_rcar_gen1_phy_init(priv);
 		break;
 	case RCAR_GEN2_SATA:
+	case RCAR_R8A7790_ES1_SATA:
 		sata_rcar_gen2_phy_init(priv);
 		break;
 	default:
@@ -838,9 +843,17 @@ static struct of_device_id sata_rcar_match[] = {
 		.data = (void *)RCAR_GEN2_SATA
 	},
 	{
+		.compatible = "renesas,sata-r8a7790-es1",
+		.data = (void *)RCAR_R8A7790_ES1_SATA
+	},
+	{
 		.compatible = "renesas,sata-r8a7791",
 		.data = (void *)RCAR_GEN2_SATA
 	},
+	{
+		.compatible = "renesas,sata-r8a7793",
+		.data = (void *)RCAR_GEN2_SATA
+	},
 	{ },
 };
 MODULE_DEVICE_TABLE(of, sata_rcar_match);
@@ -849,7 +862,9 @@ static const struct platform_device_id sata_rcar_id_table[] = {
 	{ "sata_rcar", RCAR_GEN1_SATA }, /* Deprecated by "sata-r8a7779" */
 	{ "sata-r8a7779", RCAR_GEN1_SATA },
 	{ "sata-r8a7790", RCAR_GEN2_SATA },
+	{ "sata-r8a7790-es1", RCAR_R8A7790_ES1_SATA },
 	{ "sata-r8a7791", RCAR_GEN2_SATA },
+	{ "sata-r8a7793", RCAR_GEN2_SATA },
 	{ },
 };
 MODULE_DEVICE_TABLE(platform, sata_rcar_id_table);
@@ -989,7 +1004,6 @@ static struct platform_driver sata_rcar_driver = {
 	.id_table	= sata_rcar_id_table,
 	.driver = {
 		.name		= DRV_NAME,
-		.owner		= THIS_MODULE,
 		.of_match_table	= sata_rcar_match,
 #ifdef CONFIG_PM_SLEEP
 		.pm		= &sata_rcar_pm_ops,
diff --git a/drivers/ata/sata_sil24.c b/drivers/ata/sata_sil24.c
index d81b20ddb527..ea655949023f 100644
--- a/drivers/ata/sata_sil24.c
+++ b/drivers/ata/sata_sil24.c
@@ -246,7 +246,7 @@ enum {
 	/* host flags */
 	SIL24_COMMON_FLAGS	= ATA_FLAG_SATA | ATA_FLAG_PIO_DMA |
 				  ATA_FLAG_NCQ | ATA_FLAG_ACPI_SATA |
-				  ATA_FLAG_AN | ATA_FLAG_PMP,
+				  ATA_FLAG_AN | ATA_FLAG_PMP | ATA_FLAG_LOWTAG,
 	SIL24_FLAG_PCIX_IRQ_WOC	= (1 << 24), /* IRQ loss errata on PCI-X */
 
 	IRQ_STAT_4PORTS		= 0xf,
diff --git a/drivers/atm/eni.c b/drivers/atm/eni.c
index d65975aba4ec..c7fab3ee14ee 100644
--- a/drivers/atm/eni.c
+++ b/drivers/atm/eni.c
@@ -356,6 +356,8 @@ static int do_rx_dma(struct atm_vcc *vcc,struct sk_buff *skb,
 	if (skb) {
 		paddr = pci_map_single(eni_dev->pci_dev,skb->data,skb->len,
 		    PCI_DMA_FROMDEVICE);
+		if (pci_dma_mapping_error(eni_dev->pci_dev, paddr))
+			goto dma_map_error;
 		ENI_PRV_PADDR(skb) = paddr;
 		if (paddr & 3)
 			printk(KERN_CRIT DEV_LABEL "(itf %d): VCI %d has "
@@ -481,6 +483,7 @@ trouble:
 	if (paddr)
 		pci_unmap_single(eni_dev->pci_dev,paddr,skb->len,
 		    PCI_DMA_FROMDEVICE);
+dma_map_error:
 	if (skb) dev_kfree_skb_irq(skb);
 	return -1;
 }
diff --git a/drivers/atm/fore200e.c b/drivers/atm/fore200e.c
index d4725fc0395d..d5d9eafbbfcf 100644
--- a/drivers/atm/fore200e.c
+++ b/drivers/atm/fore200e.c
@@ -2687,7 +2687,6 @@ MODULE_DEVICE_TABLE(of, fore200e_sba_match);
 static struct platform_driver fore200e_sba_driver = {
 	.driver = {
 		.name = "fore_200e",
-		.owner = THIS_MODULE,
 		.of_match_table = fore200e_sba_match,
 	},
 	.probe		= fore200e_sba_probe,
diff --git a/drivers/atm/lanai.c b/drivers/atm/lanai.c
index fa7d701933ba..93eaf8d94492 100644
--- a/drivers/atm/lanai.c
+++ b/drivers/atm/lanai.c
@@ -2614,27 +2614,7 @@ static struct pci_driver lanai_driver = {
 	.probe    = lanai_init_one,
 };
 
-static int __init lanai_module_init(void)
-{
-	int x;
-
-	x = pci_register_driver(&lanai_driver);
-	if (x != 0)
-		printk(KERN_ERR DEV_LABEL ": no adapter found\n");
-	return x;
-}
-
-static void __exit lanai_module_exit(void)
-{
-	/* We'll only get called when all the interfaces are already
-	 * gone, so there isn't much to do
-	 */
-	DPRINTK("cleanup_module()\n");
-	pci_unregister_driver(&lanai_driver);
-}
-
-module_init(lanai_module_init);
-module_exit(lanai_module_exit);
+module_pci_driver(lanai_driver);
 
 MODULE_AUTHOR("Mitchell Blank Jr <mitch@sfgoth.com>");
 MODULE_DESCRIPTION("Efficient Networks Speedstream 3010 driver");
diff --git a/drivers/atm/solos-pci.c b/drivers/atm/solos-pci.c
index 7652e8dc188f..21b0bc6a9c96 100644
--- a/drivers/atm/solos-pci.c
+++ b/drivers/atm/solos-pci.c
@@ -1225,11 +1225,13 @@ static int fpga_probe(struct pci_dev *dev, const struct pci_device_id *id)
 	card->config_regs = pci_iomap(dev, 0, CONFIG_RAM_SIZE);
 	if (!card->config_regs) {
 		dev_warn(&dev->dev, "Failed to ioremap config registers\n");
+		err = -ENOMEM;
 		goto out_release_regions;
 	}
 	card->buffers = pci_iomap(dev, 1, DATA_RAM_SIZE);
 	if (!card->buffers) {
 		dev_warn(&dev->dev, "Failed to ioremap data buffers\n");
+		err = -ENOMEM;
 		goto out_unmap_config;
 	}
 
diff --git a/drivers/base/Kconfig b/drivers/base/Kconfig
index 4e7f0ff83ae7..98504ec99c7d 100644
--- a/drivers/base/Kconfig
+++ b/drivers/base/Kconfig
@@ -165,6 +165,30 @@ config FW_LOADER_USER_HELPER_FALLBACK
 
 	  If you are unsure about this, say N here.
 
+config WANT_DEV_COREDUMP
+	bool
+	help
+	  Drivers should "select" this option if they desire to use the
+	  device coredump mechanism.
+
+config ALLOW_DEV_COREDUMP
+	bool "Allow device coredump" if EXPERT
+	default y
+	help
+	  This option controls if the device coredump mechanism is available or
+	  not; if disabled, the mechanism will be omitted even if drivers that
+	  can use it are enabled.
+	  Say 'N' for more sensitive systems or systems that don't want
+	  to ever access the information to not have the code, nor keep any
+	  data.
+
+	  If unsure, say Y.
+
+config DEV_COREDUMP
+	bool
+	default y if WANT_DEV_COREDUMP
+	depends on ALLOW_DEV_COREDUMP
+
 config DEBUG_DRIVER
 	bool "Driver Core verbose debug messages"
 	depends on DEBUG_KERNEL
@@ -231,6 +255,9 @@ config DMA_CMA
 	  to allocate big physically-contiguous blocks of memory for use with
 	  hardware components that do not support I/O map nor scatter-gather.
 
+	  You can disable CMA by specifying "cma=0" on the kernel's command
+	  line.
+
 	  For more information see <include/linux/dma-contiguous.h>.
 	  If unsure, say "n".
 
@@ -240,18 +267,24 @@ comment "Default contiguous memory area size:"
 config CMA_SIZE_MBYTES
 	int "Size in Mega Bytes"
 	depends on !CMA_SIZE_SEL_PERCENTAGE
+	default 0 if X86
 	default 16
 	help
 	  Defines the size (in MiB) of the default memory area for Contiguous
-	  Memory Allocator.
+	  Memory Allocator.  If the size of 0 is selected, CMA is disabled by
+	  default, but it can be enabled by passing cma=size[MG] to the kernel.
+
 
 config CMA_SIZE_PERCENTAGE
 	int "Percentage of total memory"
 	depends on !CMA_SIZE_SEL_MBYTES
+	default 0 if X86
 	default 10
 	help
 	  Defines the size of the default memory area for Contiguous Memory
 	  Allocator as a percentage of the total memory in the system.
+	  If 0 percent is selected, CMA is disabled by default, but it can be
+	  enabled by passing cma=size[MG] to the kernel.
 
 choice
 	prompt "Selected region size"
diff --git a/drivers/base/Makefile b/drivers/base/Makefile
index 4aab26ec0292..527d291706e8 100644
--- a/drivers/base/Makefile
+++ b/drivers/base/Makefile
@@ -4,7 +4,7 @@ obj-y			:= component.o core.o bus.o dd.o syscore.o \
 			   driver.o class.o platform.o \
 			   cpu.o firmware.o init.o map.o devres.o \
 			   attribute_container.o transport_class.o \
-			   topology.o container.o
+			   topology.o container.o property.o cacheinfo.o
 obj-$(CONFIG_DEVTMPFS)	+= devtmpfs.o
 obj-$(CONFIG_DMA_CMA) += dma-contiguous.o
 obj-y			+= power/
@@ -21,6 +21,7 @@ obj-$(CONFIG_SYS_HYPERVISOR) += hypervisor.o
 obj-$(CONFIG_REGMAP)	+= regmap/
 obj-$(CONFIG_SOC_BUS) += soc.o
 obj-$(CONFIG_PINCTRL) += pinctrl.o
+obj-$(CONFIG_DEV_COREDUMP) += devcoredump.o
 
 ccflags-$(CONFIG_DEBUG_DRIVER) := -DDEBUG
 
diff --git a/drivers/base/attribute_container.c b/drivers/base/attribute_container.c
index b84ca8f13f9e..3ead3af4be61 100644
--- a/drivers/base/attribute_container.c
+++ b/drivers/base/attribute_container.c
@@ -74,9 +74,9 @@ int
 attribute_container_register(struct attribute_container *cont)
 {
 	INIT_LIST_HEAD(&cont->node);
-	klist_init(&cont->containers,internal_container_klist_get,
+	klist_init(&cont->containers, internal_container_klist_get,
 		   internal_container_klist_put);
-		
+
 	mutex_lock(&attribute_container_mutex);
 	list_add_tail(&cont->node, &attribute_container_list);
 	mutex_unlock(&attribute_container_mutex);
@@ -104,14 +104,14 @@ attribute_container_unregister(struct attribute_container *cont)
 	spin_unlock(&cont->containers.k_lock);
 	mutex_unlock(&attribute_container_mutex);
 	return retval;
-		
+
 }
 EXPORT_SYMBOL_GPL(attribute_container_unregister);
 
 /* private function used as class release */
 static void attribute_container_release(struct device *classdev)
 {
-	struct internal_container *ic 
+	struct internal_container *ic
 		= container_of(classdev, struct internal_container, classdev);
 	struct device *dev = classdev->parent;
 
@@ -184,8 +184,8 @@ attribute_container_add_device(struct device *dev,
 		struct klist_node *n = klist_next(iter); \
 		n ? container_of(n, typeof(*pos), member) : \
 			({ klist_iter_exit(iter) ; NULL; }); \
-	}) ) != NULL; )
-			
+	})) != NULL;)
+
 
 /**
  * attribute_container_remove_device - make device eligible for removal.
@@ -247,7 +247,7 @@ attribute_container_remove_device(struct device *dev,
  * container, then use attribute_container_trigger() instead.
  */
 void
-attribute_container_device_trigger(struct device *dev, 
+attribute_container_device_trigger(struct device *dev,
 				   int (*fn)(struct attribute_container *,
 					     struct device *,
 					     struct device *))
diff --git a/drivers/base/bus.c b/drivers/base/bus.c
index 83e910a57563..876bae5ade33 100644
--- a/drivers/base/bus.c
+++ b/drivers/base/bus.c
@@ -254,13 +254,15 @@ static ssize_t store_drivers_probe(struct bus_type *bus,
 				   const char *buf, size_t count)
 {
 	struct device *dev;
+	int err = -EINVAL;
 
 	dev = bus_find_device_by_name(bus, NULL, buf);
 	if (!dev)
 		return -ENODEV;
-	if (bus_rescan_devices_helper(dev, NULL) != 0)
-		return -EINVAL;
-	return count;
+	if (bus_rescan_devices_helper(dev, NULL) == 0)
+		err = count;
+	put_device(dev);
+	return err;
 }
 
 static struct device *next_device(struct klist_iter *i)
diff --git a/drivers/base/cacheinfo.c b/drivers/base/cacheinfo.c
new file mode 100644
index 000000000000..6e64563361f0
--- /dev/null
+++ b/drivers/base/cacheinfo.c
@@ -0,0 +1,539 @@
+/*
+ * cacheinfo support - processor cache information via sysfs
+ *
+ * Based on arch/x86/kernel/cpu/intel_cacheinfo.c
+ * Author: Sudeep Holla <sudeep.holla@arm.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed "as is" WITHOUT ANY WARRANTY of any
+ * kind, whether express or implied; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+#include <linux/bitops.h>
+#include <linux/cacheinfo.h>
+#include <linux/compiler.h>
+#include <linux/cpu.h>
+#include <linux/device.h>
+#include <linux/init.h>
+#include <linux/of.h>
+#include <linux/sched.h>
+#include <linux/slab.h>
+#include <linux/smp.h>
+#include <linux/sysfs.h>
+
+/* pointer to per cpu cacheinfo */
+static DEFINE_PER_CPU(struct cpu_cacheinfo, ci_cpu_cacheinfo);
+#define ci_cacheinfo(cpu)	(&per_cpu(ci_cpu_cacheinfo, cpu))
+#define cache_leaves(cpu)	(ci_cacheinfo(cpu)->num_leaves)
+#define per_cpu_cacheinfo(cpu)	(ci_cacheinfo(cpu)->info_list)
+
+struct cpu_cacheinfo *get_cpu_cacheinfo(unsigned int cpu)
+{
+	return ci_cacheinfo(cpu);
+}
+
+#ifdef CONFIG_OF
+static int cache_setup_of_node(unsigned int cpu)
+{
+	struct device_node *np;
+	struct cacheinfo *this_leaf;
+	struct device *cpu_dev = get_cpu_device(cpu);
+	struct cpu_cacheinfo *this_cpu_ci = get_cpu_cacheinfo(cpu);
+	unsigned int index = 0;
+
+	/* skip if of_node is already populated */
+	if (this_cpu_ci->info_list->of_node)
+		return 0;
+
+	if (!cpu_dev) {
+		pr_err("No cpu device for CPU %d\n", cpu);
+		return -ENODEV;
+	}
+	np = cpu_dev->of_node;
+	if (!np) {
+		pr_err("Failed to find cpu%d device node\n", cpu);
+		return -ENOENT;
+	}
+
+	while (np && index < cache_leaves(cpu)) {
+		this_leaf = this_cpu_ci->info_list + index;
+		if (this_leaf->level != 1)
+			np = of_find_next_cache_node(np);
+		else
+			np = of_node_get(np);/* cpu node itself */
+		this_leaf->of_node = np;
+		index++;
+	}
+	return 0;
+}
+
+static inline bool cache_leaves_are_shared(struct cacheinfo *this_leaf,
+					   struct cacheinfo *sib_leaf)
+{
+	return sib_leaf->of_node == this_leaf->of_node;
+}
+#else
+static inline int cache_setup_of_node(unsigned int cpu) { return 0; }
+static inline bool cache_leaves_are_shared(struct cacheinfo *this_leaf,
+					   struct cacheinfo *sib_leaf)
+{
+	/*
+	 * For non-DT systems, assume unique level 1 cache, system-wide
+	 * shared caches for all other levels. This will be used only if
+	 * arch specific code has not populated shared_cpu_map
+	 */
+	return !(this_leaf->level == 1);
+}
+#endif
+
+static int cache_shared_cpu_map_setup(unsigned int cpu)
+{
+	struct cpu_cacheinfo *this_cpu_ci = get_cpu_cacheinfo(cpu);
+	struct cacheinfo *this_leaf, *sib_leaf;
+	unsigned int index;
+	int ret;
+
+	ret = cache_setup_of_node(cpu);
+	if (ret)
+		return ret;
+
+	for (index = 0; index < cache_leaves(cpu); index++) {
+		unsigned int i;
+
+		this_leaf = this_cpu_ci->info_list + index;
+		/* skip if shared_cpu_map is already populated */
+		if (!cpumask_empty(&this_leaf->shared_cpu_map))
+			continue;
+
+		cpumask_set_cpu(cpu, &this_leaf->shared_cpu_map);
+		for_each_online_cpu(i) {
+			struct cpu_cacheinfo *sib_cpu_ci = get_cpu_cacheinfo(i);
+
+			if (i == cpu || !sib_cpu_ci->info_list)
+				continue;/* skip if itself or no cacheinfo */
+			sib_leaf = sib_cpu_ci->info_list + index;
+			if (cache_leaves_are_shared(this_leaf, sib_leaf)) {
+				cpumask_set_cpu(cpu, &sib_leaf->shared_cpu_map);
+				cpumask_set_cpu(i, &this_leaf->shared_cpu_map);
+			}
+		}
+	}
+
+	return 0;
+}
+
+static void cache_shared_cpu_map_remove(unsigned int cpu)
+{
+	struct cpu_cacheinfo *this_cpu_ci = get_cpu_cacheinfo(cpu);
+	struct cacheinfo *this_leaf, *sib_leaf;
+	unsigned int sibling, index;
+
+	for (index = 0; index < cache_leaves(cpu); index++) {
+		this_leaf = this_cpu_ci->info_list + index;
+		for_each_cpu(sibling, &this_leaf->shared_cpu_map) {
+			struct cpu_cacheinfo *sib_cpu_ci;
+
+			if (sibling == cpu) /* skip itself */
+				continue;
+			sib_cpu_ci = get_cpu_cacheinfo(sibling);
+			sib_leaf = sib_cpu_ci->info_list + index;
+			cpumask_clear_cpu(cpu, &sib_leaf->shared_cpu_map);
+			cpumask_clear_cpu(sibling, &this_leaf->shared_cpu_map);
+		}
+		of_node_put(this_leaf->of_node);
+	}
+}
+
+static void free_cache_attributes(unsigned int cpu)
+{
+	cache_shared_cpu_map_remove(cpu);
+
+	kfree(per_cpu_cacheinfo(cpu));
+	per_cpu_cacheinfo(cpu) = NULL;
+}
+
+int __weak init_cache_level(unsigned int cpu)
+{
+	return -ENOENT;
+}
+
+int __weak populate_cache_leaves(unsigned int cpu)
+{
+	return -ENOENT;
+}
+
+static int detect_cache_attributes(unsigned int cpu)
+{
+	int ret;
+
+	if (init_cache_level(cpu))
+		return -ENOENT;
+
+	per_cpu_cacheinfo(cpu) = kcalloc(cache_leaves(cpu),
+					 sizeof(struct cacheinfo), GFP_KERNEL);
+	if (per_cpu_cacheinfo(cpu) == NULL)
+		return -ENOMEM;
+
+	ret = populate_cache_leaves(cpu);
+	if (ret)
+		goto free_ci;
+	/*
+	 * For systems using DT for cache hierarcy, of_node and shared_cpu_map
+	 * will be set up here only if they are not populated already
+	 */
+	ret = cache_shared_cpu_map_setup(cpu);
+	if (ret)
+		goto free_ci;
+	return 0;
+
+free_ci:
+	free_cache_attributes(cpu);
+	return ret;
+}
+
+/* pointer to cpuX/cache device */
+static DEFINE_PER_CPU(struct device *, ci_cache_dev);
+#define per_cpu_cache_dev(cpu)	(per_cpu(ci_cache_dev, cpu))
+
+static cpumask_t cache_dev_map;
+
+/* pointer to array of devices for cpuX/cache/indexY */
+static DEFINE_PER_CPU(struct device **, ci_index_dev);
+#define per_cpu_index_dev(cpu)	(per_cpu(ci_index_dev, cpu))
+#define per_cache_index_dev(cpu, idx)	((per_cpu_index_dev(cpu))[idx])
+
+#define show_one(file_name, object)				\
+static ssize_t file_name##_show(struct device *dev,		\
+		struct device_attribute *attr, char *buf)	\
+{								\
+	struct cacheinfo *this_leaf = dev_get_drvdata(dev);	\
+	return sprintf(buf, "%u\n", this_leaf->object);		\
+}
+
+show_one(level, level);
+show_one(coherency_line_size, coherency_line_size);
+show_one(number_of_sets, number_of_sets);
+show_one(physical_line_partition, physical_line_partition);
+show_one(ways_of_associativity, ways_of_associativity);
+
+static ssize_t size_show(struct device *dev,
+			 struct device_attribute *attr, char *buf)
+{
+	struct cacheinfo *this_leaf = dev_get_drvdata(dev);
+
+	return sprintf(buf, "%uK\n", this_leaf->size >> 10);
+}
+
+static ssize_t shared_cpumap_show_func(struct device *dev, bool list, char *buf)
+{
+	struct cacheinfo *this_leaf = dev_get_drvdata(dev);
+	const struct cpumask *mask = &this_leaf->shared_cpu_map;
+
+	return cpumap_print_to_pagebuf(list, buf, mask);
+}
+
+static ssize_t shared_cpu_map_show(struct device *dev,
+				   struct device_attribute *attr, char *buf)
+{
+	return shared_cpumap_show_func(dev, false, buf);
+}
+
+static ssize_t shared_cpu_list_show(struct device *dev,
+				    struct device_attribute *attr, char *buf)
+{
+	return shared_cpumap_show_func(dev, true, buf);
+}
+
+static ssize_t type_show(struct device *dev,
+			 struct device_attribute *attr, char *buf)
+{
+	struct cacheinfo *this_leaf = dev_get_drvdata(dev);
+
+	switch (this_leaf->type) {
+	case CACHE_TYPE_DATA:
+		return sprintf(buf, "Data\n");
+	case CACHE_TYPE_INST:
+		return sprintf(buf, "Instruction\n");
+	case CACHE_TYPE_UNIFIED:
+		return sprintf(buf, "Unified\n");
+	default:
+		return -EINVAL;
+	}
+}
+
+static ssize_t allocation_policy_show(struct device *dev,
+				      struct device_attribute *attr, char *buf)
+{
+	struct cacheinfo *this_leaf = dev_get_drvdata(dev);
+	unsigned int ci_attr = this_leaf->attributes;
+	int n = 0;
+
+	if ((ci_attr & CACHE_READ_ALLOCATE) && (ci_attr & CACHE_WRITE_ALLOCATE))
+		n = sprintf(buf, "ReadWriteAllocate\n");
+	else if (ci_attr & CACHE_READ_ALLOCATE)
+		n = sprintf(buf, "ReadAllocate\n");
+	else if (ci_attr & CACHE_WRITE_ALLOCATE)
+		n = sprintf(buf, "WriteAllocate\n");
+	return n;
+}
+
+static ssize_t write_policy_show(struct device *dev,
+				 struct device_attribute *attr, char *buf)
+{
+	struct cacheinfo *this_leaf = dev_get_drvdata(dev);
+	unsigned int ci_attr = this_leaf->attributes;
+	int n = 0;
+
+	if (ci_attr & CACHE_WRITE_THROUGH)
+		n = sprintf(buf, "WriteThrough\n");
+	else if (ci_attr & CACHE_WRITE_BACK)
+		n = sprintf(buf, "WriteBack\n");
+	return n;
+}
+
+static DEVICE_ATTR_RO(level);
+static DEVICE_ATTR_RO(type);
+static DEVICE_ATTR_RO(coherency_line_size);
+static DEVICE_ATTR_RO(ways_of_associativity);
+static DEVICE_ATTR_RO(number_of_sets);
+static DEVICE_ATTR_RO(size);
+static DEVICE_ATTR_RO(allocation_policy);
+static DEVICE_ATTR_RO(write_policy);
+static DEVICE_ATTR_RO(shared_cpu_map);
+static DEVICE_ATTR_RO(shared_cpu_list);
+static DEVICE_ATTR_RO(physical_line_partition);
+
+static struct attribute *cache_default_attrs[] = {
+	&dev_attr_type.attr,
+	&dev_attr_level.attr,
+	&dev_attr_shared_cpu_map.attr,
+	&dev_attr_shared_cpu_list.attr,
+	&dev_attr_coherency_line_size.attr,
+	&dev_attr_ways_of_associativity.attr,
+	&dev_attr_number_of_sets.attr,
+	&dev_attr_size.attr,
+	&dev_attr_allocation_policy.attr,
+	&dev_attr_write_policy.attr,
+	&dev_attr_physical_line_partition.attr,
+	NULL
+};
+
+static umode_t
+cache_default_attrs_is_visible(struct kobject *kobj,
+			       struct attribute *attr, int unused)
+{
+	struct device *dev = kobj_to_dev(kobj);
+	struct cacheinfo *this_leaf = dev_get_drvdata(dev);
+	const struct cpumask *mask = &this_leaf->shared_cpu_map;
+	umode_t mode = attr->mode;
+
+	if ((attr == &dev_attr_type.attr) && this_leaf->type)
+		return mode;
+	if ((attr == &dev_attr_level.attr) && this_leaf->level)
+		return mode;
+	if ((attr == &dev_attr_shared_cpu_map.attr) && !cpumask_empty(mask))
+		return mode;
+	if ((attr == &dev_attr_shared_cpu_list.attr) && !cpumask_empty(mask))
+		return mode;
+	if ((attr == &dev_attr_coherency_line_size.attr) &&
+	    this_leaf->coherency_line_size)
+		return mode;
+	if ((attr == &dev_attr_ways_of_associativity.attr) &&
+	    this_leaf->size) /* allow 0 = full associativity */
+		return mode;
+	if ((attr == &dev_attr_number_of_sets.attr) &&
+	    this_leaf->number_of_sets)
+		return mode;
+	if ((attr == &dev_attr_size.attr) && this_leaf->size)
+		return mode;
+	if ((attr == &dev_attr_write_policy.attr) &&
+	    (this_leaf->attributes & CACHE_WRITE_POLICY_MASK))
+		return mode;
+	if ((attr == &dev_attr_allocation_policy.attr) &&
+	    (this_leaf->attributes & CACHE_ALLOCATE_POLICY_MASK))
+		return mode;
+	if ((attr == &dev_attr_physical_line_partition.attr) &&
+	    this_leaf->physical_line_partition)
+		return mode;
+
+	return 0;
+}
+
+static const struct attribute_group cache_default_group = {
+	.attrs = cache_default_attrs,
+	.is_visible = cache_default_attrs_is_visible,
+};
+
+static const struct attribute_group *cache_default_groups[] = {
+	&cache_default_group,
+	NULL,
+};
+
+static const struct attribute_group *cache_private_groups[] = {
+	&cache_default_group,
+	NULL, /* Place holder for private group */
+	NULL,
+};
+
+const struct attribute_group *
+__weak cache_get_priv_group(struct cacheinfo *this_leaf)
+{
+	return NULL;
+}
+
+static const struct attribute_group **
+cache_get_attribute_groups(struct cacheinfo *this_leaf)
+{
+	const struct attribute_group *priv_group =
+			cache_get_priv_group(this_leaf);
+
+	if (!priv_group)
+		return cache_default_groups;
+
+	if (!cache_private_groups[1])
+		cache_private_groups[1] = priv_group;
+
+	return cache_private_groups;
+}
+
+/* Add/Remove cache interface for CPU device */
+static void cpu_cache_sysfs_exit(unsigned int cpu)
+{
+	int i;
+	struct device *ci_dev;
+
+	if (per_cpu_index_dev(cpu)) {
+		for (i = 0; i < cache_leaves(cpu); i++) {
+			ci_dev = per_cache_index_dev(cpu, i);
+			if (!ci_dev)
+				continue;
+			device_unregister(ci_dev);
+		}
+		kfree(per_cpu_index_dev(cpu));
+		per_cpu_index_dev(cpu) = NULL;
+	}
+	device_unregister(per_cpu_cache_dev(cpu));
+	per_cpu_cache_dev(cpu) = NULL;
+}
+
+static int cpu_cache_sysfs_init(unsigned int cpu)
+{
+	struct device *dev = get_cpu_device(cpu);
+
+	if (per_cpu_cacheinfo(cpu) == NULL)
+		return -ENOENT;
+
+	per_cpu_cache_dev(cpu) = cpu_device_create(dev, NULL, NULL, "cache");
+	if (IS_ERR(per_cpu_cache_dev(cpu)))
+		return PTR_ERR(per_cpu_cache_dev(cpu));
+
+	/* Allocate all required memory */
+	per_cpu_index_dev(cpu) = kcalloc(cache_leaves(cpu),
+					 sizeof(struct device *), GFP_KERNEL);
+	if (unlikely(per_cpu_index_dev(cpu) == NULL))
+		goto err_out;
+
+	return 0;
+
+err_out:
+	cpu_cache_sysfs_exit(cpu);
+	return -ENOMEM;
+}
+
+static int cache_add_dev(unsigned int cpu)
+{
+	unsigned int i;
+	int rc;
+	struct device *ci_dev, *parent;
+	struct cacheinfo *this_leaf;
+	struct cpu_cacheinfo *this_cpu_ci = get_cpu_cacheinfo(cpu);
+	const struct attribute_group **cache_groups;
+
+	rc = cpu_cache_sysfs_init(cpu);
+	if (unlikely(rc < 0))
+		return rc;
+
+	parent = per_cpu_cache_dev(cpu);
+	for (i = 0; i < cache_leaves(cpu); i++) {
+		this_leaf = this_cpu_ci->info_list + i;
+		if (this_leaf->disable_sysfs)
+			continue;
+		cache_groups = cache_get_attribute_groups(this_leaf);
+		ci_dev = cpu_device_create(parent, this_leaf, cache_groups,
+					   "index%1u", i);
+		if (IS_ERR(ci_dev)) {
+			rc = PTR_ERR(ci_dev);
+			goto err;
+		}
+		per_cache_index_dev(cpu, i) = ci_dev;
+	}
+	cpumask_set_cpu(cpu, &cache_dev_map);
+
+	return 0;
+err:
+	cpu_cache_sysfs_exit(cpu);
+	return rc;
+}
+
+static void cache_remove_dev(unsigned int cpu)
+{
+	if (!cpumask_test_cpu(cpu, &cache_dev_map))
+		return;
+	cpumask_clear_cpu(cpu, &cache_dev_map);
+
+	cpu_cache_sysfs_exit(cpu);
+}
+
+static int cacheinfo_cpu_callback(struct notifier_block *nfb,
+				  unsigned long action, void *hcpu)
+{
+	unsigned int cpu = (unsigned long)hcpu;
+	int rc = 0;
+
+	switch (action & ~CPU_TASKS_FROZEN) {
+	case CPU_ONLINE:
+		rc = detect_cache_attributes(cpu);
+		if (!rc)
+			rc = cache_add_dev(cpu);
+		break;
+	case CPU_DEAD:
+		cache_remove_dev(cpu);
+		if (per_cpu_cacheinfo(cpu))
+			free_cache_attributes(cpu);
+		break;
+	}
+	return notifier_from_errno(rc);
+}
+
+static int __init cacheinfo_sysfs_init(void)
+{
+	int cpu, rc = 0;
+
+	cpu_notifier_register_begin();
+
+	for_each_online_cpu(cpu) {
+		rc = detect_cache_attributes(cpu);
+		if (rc)
+			goto out;
+		rc = cache_add_dev(cpu);
+		if (rc) {
+			free_cache_attributes(cpu);
+			pr_err("error populating cacheinfo..cpu%d\n", cpu);
+			goto out;
+		}
+	}
+	__hotcpu_notifier(cacheinfo_cpu_callback, 0);
+
+out:
+	cpu_notifier_register_done();
+	return rc;
+}
+
+device_initcall(cacheinfo_sysfs_init);
diff --git a/drivers/base/core.c b/drivers/base/core.c
index 20da3ad1696b..97e2baf6e5d8 100644
--- a/drivers/base/core.c
+++ b/drivers/base/core.c
@@ -724,12 +724,12 @@ class_dir_create_and_add(struct class *class, struct kobject *parent_kobj)
 	return &dir->kobj;
 }
 
+static DEFINE_MUTEX(gdp_mutex);
 
 static struct kobject *get_device_parent(struct device *dev,
 					 struct device *parent)
 {
 	if (dev->class) {
-		static DEFINE_MUTEX(gdp_mutex);
 		struct kobject *kobj = NULL;
 		struct kobject *parent_kobj;
 		struct kobject *k;
@@ -793,7 +793,9 @@ static void cleanup_glue_dir(struct device *dev, struct kobject *glue_dir)
 	    glue_dir->kset != &dev->class->p->glue_dirs)
 		return;
 
+	mutex_lock(&gdp_mutex);
 	kobject_put(glue_dir);
+	mutex_unlock(&gdp_mutex);
 }
 
 static void cleanup_device_parent(struct device *dev)
@@ -1019,18 +1021,6 @@ int device_add(struct device *dev)
 	if (error)
 		goto attrError;
 
-	if (MAJOR(dev->devt)) {
-		error = device_create_file(dev, &dev_attr_dev);
-		if (error)
-			goto ueventattrError;
-
-		error = device_create_sys_dev_entry(dev);
-		if (error)
-			goto devtattrError;
-
-		devtmpfs_create_node(dev);
-	}
-
 	error = device_add_class_symlinks(dev);
 	if (error)
 		goto SymlinkError;
@@ -1045,6 +1035,18 @@ int device_add(struct device *dev)
 		goto DPMError;
 	device_pm_add(dev);
 
+	if (MAJOR(dev->devt)) {
+		error = device_create_file(dev, &dev_attr_dev);
+		if (error)
+			goto DevAttrError;
+
+		error = device_create_sys_dev_entry(dev);
+		if (error)
+			goto SysEntryError;
+
+		devtmpfs_create_node(dev);
+	}
+
 	/* Notify clients of device addition.  This call must come
 	 * after dpm_sysfs_add() and before kobject_uevent().
 	 */
@@ -1074,6 +1076,12 @@ int device_add(struct device *dev)
 done:
 	put_device(dev);
 	return error;
+ SysEntryError:
+	if (MAJOR(dev->devt))
+		device_remove_file(dev, &dev_attr_dev);
+ DevAttrError:
+	device_pm_remove(dev);
+	dpm_sysfs_remove(dev);
  DPMError:
 	bus_remove_device(dev);
  BusError:
@@ -1081,14 +1089,6 @@ done:
  AttrsError:
 	device_remove_class_symlinks(dev);
  SymlinkError:
-	if (MAJOR(dev->devt))
-		devtmpfs_delete_node(dev);
-	if (MAJOR(dev->devt))
-		device_remove_sys_dev_entry(dev);
- devtattrError:
-	if (MAJOR(dev->devt))
-		device_remove_file(dev, &dev_attr_dev);
- ueventattrError:
 	device_remove_file(dev, &dev_attr_uevent);
  attrError:
 	kobject_uevent(&dev->kobj, KOBJ_REMOVE);
@@ -1211,6 +1211,9 @@ void device_del(struct device *dev)
 	 */
 	if (platform_notify_remove)
 		platform_notify_remove(dev);
+	if (dev->bus)
+		blocking_notifier_call_chain(&dev->bus->p->bus_notifier,
+					     BUS_NOTIFY_REMOVED_DEVICE, dev);
 	kobject_uevent(&dev->kobj, KOBJ_REMOVE);
 	cleanup_device_parent(dev);
 	kobject_del(&dev->kobj);
@@ -2007,6 +2010,8 @@ create_syslog_header(const struct device *dev, char *hdr, size_t hdrlen)
 		return 0;
 
 	pos += snprintf(hdr + pos, hdrlen - pos, "SUBSYSTEM=%s", subsys);
+	if (pos >= hdrlen)
+		goto overflow;
 
 	/*
 	 * Add device identifier DEVICE=:
@@ -2038,7 +2043,14 @@ create_syslog_header(const struct device *dev, char *hdr, size_t hdrlen)
 				"DEVICE=+%s:%s", subsys, dev_name(dev));
 	}
 
+	if (pos >= hdrlen)
+		goto overflow;
+
 	return pos;
+
+overflow:
+	dev_WARN(dev, "device/subsystem name too long");
+	return 0;
 }
 
 int dev_vprintk_emit(int level, const struct device *dev,
diff --git a/drivers/base/cpu.c b/drivers/base/cpu.c
index 006b1bc5297d..f829a4c71749 100644
--- a/drivers/base/cpu.c
+++ b/drivers/base/cpu.c
@@ -207,11 +207,8 @@ static ssize_t show_cpus_attr(struct device *dev,
 			      char *buf)
 {
 	struct cpu_attr *ca = container_of(attr, struct cpu_attr, attr);
-	int n = cpulist_scnprintf(buf, PAGE_SIZE-2, *(ca->map));
 
-	buf[n++] = '\n';
-	buf[n] = '\0';
-	return n;
+	return cpumap_print_to_pagebuf(true, buf, *ca->map);
 }
 
 #define _CPU_ATTR(name, map) \
@@ -366,6 +363,60 @@ struct device *get_cpu_device(unsigned cpu)
 }
 EXPORT_SYMBOL_GPL(get_cpu_device);
 
+static void device_create_release(struct device *dev)
+{
+	kfree(dev);
+}
+
+static struct device *
+__cpu_device_create(struct device *parent, void *drvdata,
+		    const struct attribute_group **groups,
+		    const char *fmt, va_list args)
+{
+	struct device *dev = NULL;
+	int retval = -ENODEV;
+
+	dev = kzalloc(sizeof(*dev), GFP_KERNEL);
+	if (!dev) {
+		retval = -ENOMEM;
+		goto error;
+	}
+
+	device_initialize(dev);
+	dev->parent = parent;
+	dev->groups = groups;
+	dev->release = device_create_release;
+	dev_set_drvdata(dev, drvdata);
+
+	retval = kobject_set_name_vargs(&dev->kobj, fmt, args);
+	if (retval)
+		goto error;
+
+	retval = device_add(dev);
+	if (retval)
+		goto error;
+
+	return dev;
+
+error:
+	put_device(dev);
+	return ERR_PTR(retval);
+}
+
+struct device *cpu_device_create(struct device *parent, void *drvdata,
+				 const struct attribute_group **groups,
+				 const char *fmt, ...)
+{
+	va_list vargs;
+	struct device *dev;
+
+	va_start(vargs, fmt);
+	dev = __cpu_device_create(parent, drvdata, groups, fmt, vargs);
+	va_end(vargs);
+	return dev;
+}
+EXPORT_SYMBOL_GPL(cpu_device_create);
+
 #ifdef CONFIG_GENERIC_CPU_AUTOPROBE
 static DEVICE_ATTR(modalias, 0444, print_cpu_modalias, NULL);
 #endif
diff --git a/drivers/base/dd.c b/drivers/base/dd.c
index e4ffbcf2f519..cdc779cf79a3 100644
--- a/drivers/base/dd.c
+++ b/drivers/base/dd.c
@@ -54,7 +54,7 @@ static LIST_HEAD(deferred_probe_active_list);
 static struct workqueue_struct *deferred_wq;
 static atomic_t deferred_trigger_count = ATOMIC_INIT(0);
 
-/**
+/*
  * deferred_probe_work_func() - Retry probing devices in the active list.
  */
 static void deferred_probe_work_func(struct work_struct *work)
diff --git a/drivers/base/devcoredump.c b/drivers/base/devcoredump.c
new file mode 100644
index 000000000000..1bd120a0b084
--- /dev/null
+++ b/drivers/base/devcoredump.c
@@ -0,0 +1,305 @@
+/*
+ * This file is provided under the GPLv2 license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ * Copyright(c) 2014 Intel Mobile Communications GmbH
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * The full GNU General Public License is included in this distribution
+ * in the file called COPYING.
+ *
+ * Contact Information:
+ *  Intel Linux Wireless <ilw@linux.intel.com>
+ * Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
+ *
+ * Author: Johannes Berg <johannes@sipsolutions.net>
+ */
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/devcoredump.h>
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <linux/fs.h>
+#include <linux/workqueue.h>
+
+static struct class devcd_class;
+
+/* global disable flag, for security purposes */
+static bool devcd_disabled;
+
+/* if data isn't read by userspace after 5 minutes then delete it */
+#define DEVCD_TIMEOUT	(HZ * 60 * 5)
+
+struct devcd_entry {
+	struct device devcd_dev;
+	const void *data;
+	size_t datalen;
+	struct module *owner;
+	ssize_t (*read)(char *buffer, loff_t offset, size_t count,
+			const void *data, size_t datalen);
+	void (*free)(const void *data);
+	struct delayed_work del_wk;
+	struct device *failing_dev;
+};
+
+static struct devcd_entry *dev_to_devcd(struct device *dev)
+{
+	return container_of(dev, struct devcd_entry, devcd_dev);
+}
+
+static void devcd_dev_release(struct device *dev)
+{
+	struct devcd_entry *devcd = dev_to_devcd(dev);
+
+	devcd->free(devcd->data);
+	module_put(devcd->owner);
+
+	/*
+	 * this seems racy, but I don't see a notifier or such on
+	 * a struct device to know when it goes away?
+	 */
+	if (devcd->failing_dev->kobj.sd)
+		sysfs_delete_link(&devcd->failing_dev->kobj, &dev->kobj,
+				  "devcoredump");
+
+	put_device(devcd->failing_dev);
+	kfree(devcd);
+}
+
+static void devcd_del(struct work_struct *wk)
+{
+	struct devcd_entry *devcd;
+
+	devcd = container_of(wk, struct devcd_entry, del_wk.work);
+
+	device_del(&devcd->devcd_dev);
+	put_device(&devcd->devcd_dev);
+}
+
+static ssize_t devcd_data_read(struct file *filp, struct kobject *kobj,
+			       struct bin_attribute *bin_attr,
+			       char *buffer, loff_t offset, size_t count)
+{
+	struct device *dev = kobj_to_dev(kobj);
+	struct devcd_entry *devcd = dev_to_devcd(dev);
+
+	return devcd->read(buffer, offset, count, devcd->data, devcd->datalen);
+}
+
+static ssize_t devcd_data_write(struct file *filp, struct kobject *kobj,
+				struct bin_attribute *bin_attr,
+				char *buffer, loff_t offset, size_t count)
+{
+	struct device *dev = kobj_to_dev(kobj);
+	struct devcd_entry *devcd = dev_to_devcd(dev);
+
+	mod_delayed_work(system_wq, &devcd->del_wk, 0);
+
+	return count;
+}
+
+static struct bin_attribute devcd_attr_data = {
+	.attr = { .name = "data", .mode = S_IRUSR | S_IWUSR, },
+	.size = 0,
+	.read = devcd_data_read,
+	.write = devcd_data_write,
+};
+
+static struct bin_attribute *devcd_dev_bin_attrs[] = {
+	&devcd_attr_data, NULL,
+};
+
+static const struct attribute_group devcd_dev_group = {
+	.bin_attrs = devcd_dev_bin_attrs,
+};
+
+static const struct attribute_group *devcd_dev_groups[] = {
+	&devcd_dev_group, NULL,
+};
+
+static int devcd_free(struct device *dev, void *data)
+{
+	struct devcd_entry *devcd = dev_to_devcd(dev);
+
+	flush_delayed_work(&devcd->del_wk);
+	return 0;
+}
+
+static ssize_t disabled_show(struct class *class, struct class_attribute *attr,
+			     char *buf)
+{
+	return sprintf(buf, "%d\n", devcd_disabled);
+}
+
+static ssize_t disabled_store(struct class *class, struct class_attribute *attr,
+			      const char *buf, size_t count)
+{
+	long tmp = simple_strtol(buf, NULL, 10);
+
+	/*
+	 * This essentially makes the attribute write-once, since you can't
+	 * go back to not having it disabled. This is intentional, it serves
+	 * as a system lockdown feature.
+	 */
+	if (tmp != 1)
+		return -EINVAL;
+
+	devcd_disabled = true;
+
+	class_for_each_device(&devcd_class, NULL, NULL, devcd_free);
+
+	return count;
+}
+
+static struct class_attribute devcd_class_attrs[] = {
+	__ATTR_RW(disabled),
+	__ATTR_NULL
+};
+
+static struct class devcd_class = {
+	.name		= "devcoredump",
+	.owner		= THIS_MODULE,
+	.dev_release	= devcd_dev_release,
+	.dev_groups	= devcd_dev_groups,
+	.class_attrs	= devcd_class_attrs,
+};
+
+static ssize_t devcd_readv(char *buffer, loff_t offset, size_t count,
+			   const void *data, size_t datalen)
+{
+	if (offset > datalen)
+		return -EINVAL;
+
+	if (offset + count > datalen)
+		count = datalen - offset;
+
+	if (count)
+		memcpy(buffer, ((u8 *)data) + offset, count);
+
+	return count;
+}
+
+/**
+ * dev_coredumpv - create device coredump with vmalloc data
+ * @dev: the struct device for the crashed device
+ * @data: vmalloc data containing the device coredump
+ * @datalen: length of the data
+ * @gfp: allocation flags
+ *
+ * This function takes ownership of the vmalloc'ed data and will free
+ * it when it is no longer used. See dev_coredumpm() for more information.
+ */
+void dev_coredumpv(struct device *dev, const void *data, size_t datalen,
+		   gfp_t gfp)
+{
+	dev_coredumpm(dev, NULL, data, datalen, gfp, devcd_readv, vfree);
+}
+EXPORT_SYMBOL_GPL(dev_coredumpv);
+
+static int devcd_match_failing(struct device *dev, const void *failing)
+{
+	struct devcd_entry *devcd = dev_to_devcd(dev);
+
+	return devcd->failing_dev == failing;
+}
+
+/**
+ * dev_coredumpm - create device coredump with read/free methods
+ * @dev: the struct device for the crashed device
+ * @owner: the module that contains the read/free functions, use %THIS_MODULE
+ * @data: data cookie for the @read/@free functions
+ * @datalen: length of the data
+ * @gfp: allocation flags
+ * @read: function to read from the given buffer
+ * @free: function to free the given buffer
+ *
+ * Creates a new device coredump for the given device. If a previous one hasn't
+ * been read yet, the new coredump is discarded. The data lifetime is determined
+ * by the device coredump framework and when it is no longer needed the @free
+ * function will be called to free the data.
+ */
+void dev_coredumpm(struct device *dev, struct module *owner,
+		   const void *data, size_t datalen, gfp_t gfp,
+		   ssize_t (*read)(char *buffer, loff_t offset, size_t count,
+				   const void *data, size_t datalen),
+		   void (*free)(const void *data))
+{
+	static atomic_t devcd_count = ATOMIC_INIT(0);
+	struct devcd_entry *devcd;
+	struct device *existing;
+
+	if (devcd_disabled)
+		goto free;
+
+	existing = class_find_device(&devcd_class, NULL, dev,
+				     devcd_match_failing);
+	if (existing) {
+		put_device(existing);
+		goto free;
+	}
+
+	if (!try_module_get(owner))
+		goto free;
+
+	devcd = kzalloc(sizeof(*devcd), gfp);
+	if (!devcd)
+		goto put_module;
+
+	devcd->owner = owner;
+	devcd->data = data;
+	devcd->datalen = datalen;
+	devcd->read = read;
+	devcd->free = free;
+	devcd->failing_dev = get_device(dev);
+
+	device_initialize(&devcd->devcd_dev);
+
+	dev_set_name(&devcd->devcd_dev, "devcd%d",
+		     atomic_inc_return(&devcd_count));
+	devcd->devcd_dev.class = &devcd_class;
+
+	if (device_add(&devcd->devcd_dev))
+		goto put_device;
+
+	if (sysfs_create_link(&devcd->devcd_dev.kobj, &dev->kobj,
+			      "failing_device"))
+		/* nothing - symlink will be missing */;
+
+	if (sysfs_create_link(&dev->kobj, &devcd->devcd_dev.kobj,
+			      "devcoredump"))
+		/* nothing - symlink will be missing */;
+
+	INIT_DELAYED_WORK(&devcd->del_wk, devcd_del);
+	schedule_delayed_work(&devcd->del_wk, DEVCD_TIMEOUT);
+
+	return;
+ put_device:
+	put_device(&devcd->devcd_dev);
+ put_module:
+	module_put(owner);
+ free:
+	free(data);
+}
+EXPORT_SYMBOL_GPL(dev_coredumpm);
+
+static int __init devcoredump_init(void)
+{
+	return class_register(&devcd_class);
+}
+__initcall(devcoredump_init);
+
+static void __exit devcoredump_exit(void)
+{
+	class_for_each_device(&devcd_class, NULL, NULL, devcd_free);
+	class_unregister(&devcd_class);
+}
+__exitcall(devcoredump_exit);
diff --git a/drivers/base/devres.c b/drivers/base/devres.c
index 69d9b0c89a01..c8a53d1e019f 100644
--- a/drivers/base/devres.c
+++ b/drivers/base/devres.c
@@ -817,13 +817,13 @@ char *devm_kstrdup(struct device *dev, const char *s, gfp_t gfp)
 EXPORT_SYMBOL_GPL(devm_kstrdup);
 
 /**
- * devm_kvasprintf - Allocate resource managed space
- *			for the formatted string.
+ * devm_kvasprintf - Allocate resource managed space and format a string
+ *		     into that.
  * @dev: Device to allocate memory for
  * @gfp: the GFP mask used in the devm_kmalloc() call when
  *       allocating memory
- * @fmt: the formatted string to duplicate
- * @ap: the list of tokens to be placed in the formatted string
+ * @fmt: The printf()-style format string
+ * @ap: Arguments for the format string
  * RETURNS:
  * Pointer to allocated string on success, NULL on failure.
  */
@@ -849,12 +849,13 @@ char *devm_kvasprintf(struct device *dev, gfp_t gfp, const char *fmt,
 EXPORT_SYMBOL(devm_kvasprintf);
 
 /**
- * devm_kasprintf - Allocate resource managed space
- *		and copy an existing formatted string into that
+ * devm_kasprintf - Allocate resource managed space and format a string
+ *		    into that.
  * @dev: Device to allocate memory for
  * @gfp: the GFP mask used in the devm_kmalloc() call when
  *       allocating memory
- * @fmt: the string to duplicate
+ * @fmt: The printf()-style format string
+ * @...: Arguments for the format string
  * RETURNS:
  * Pointer to allocated string on success, NULL on failure.
  */
diff --git a/drivers/base/dma-coherent.c b/drivers/base/dma-coherent.c
index 7d6e84a51424..55b83983a9c0 100644
--- a/drivers/base/dma-coherent.c
+++ b/drivers/base/dma-coherent.c
@@ -14,11 +14,14 @@ struct dma_coherent_mem {
 	int		size;
 	int		flags;
 	unsigned long	*bitmap;
+	spinlock_t	spinlock;
 };
 
-int dma_declare_coherent_memory(struct device *dev, phys_addr_t phys_addr,
-				dma_addr_t device_addr, size_t size, int flags)
+static int dma_init_coherent_memory(phys_addr_t phys_addr, dma_addr_t device_addr,
+			     size_t size, int flags,
+			     struct dma_coherent_mem **mem)
 {
+	struct dma_coherent_mem *dma_mem = NULL;
 	void __iomem *mem_base = NULL;
 	int pages = size >> PAGE_SHIFT;
 	int bitmap_size = BITS_TO_LONGS(pages) * sizeof(long);
@@ -27,40 +30,77 @@ int dma_declare_coherent_memory(struct device *dev, phys_addr_t phys_addr,
 		goto out;
 	if (!size)
 		goto out;
-	if (dev->dma_mem)
-		goto out;
-
-	/* FIXME: this routine just ignores DMA_MEMORY_INCLUDES_CHILDREN */
 
 	mem_base = ioremap(phys_addr, size);
 	if (!mem_base)
 		goto out;
 
-	dev->dma_mem = kzalloc(sizeof(struct dma_coherent_mem), GFP_KERNEL);
-	if (!dev->dma_mem)
+	dma_mem = kzalloc(sizeof(struct dma_coherent_mem), GFP_KERNEL);
+	if (!dma_mem)
 		goto out;
-	dev->dma_mem->bitmap = kzalloc(bitmap_size, GFP_KERNEL);
-	if (!dev->dma_mem->bitmap)
-		goto free1_out;
+	dma_mem->bitmap = kzalloc(bitmap_size, GFP_KERNEL);
+	if (!dma_mem->bitmap)
+		goto out;
+
+	dma_mem->virt_base = mem_base;
+	dma_mem->device_base = device_addr;
+	dma_mem->pfn_base = PFN_DOWN(phys_addr);
+	dma_mem->size = pages;
+	dma_mem->flags = flags;
+	spin_lock_init(&dma_mem->spinlock);
 
-	dev->dma_mem->virt_base = mem_base;
-	dev->dma_mem->device_base = device_addr;
-	dev->dma_mem->pfn_base = PFN_DOWN(phys_addr);
-	dev->dma_mem->size = pages;
-	dev->dma_mem->flags = flags;
+	*mem = dma_mem;
 
 	if (flags & DMA_MEMORY_MAP)
 		return DMA_MEMORY_MAP;
 
 	return DMA_MEMORY_IO;
 
- free1_out:
-	kfree(dev->dma_mem);
- out:
+out:
+	kfree(dma_mem);
 	if (mem_base)
 		iounmap(mem_base);
 	return 0;
 }
+
+static void dma_release_coherent_memory(struct dma_coherent_mem *mem)
+{
+	if (!mem)
+		return;
+	iounmap(mem->virt_base);
+	kfree(mem->bitmap);
+	kfree(mem);
+}
+
+static int dma_assign_coherent_memory(struct device *dev,
+				      struct dma_coherent_mem *mem)
+{
+	if (dev->dma_mem)
+		return -EBUSY;
+
+	dev->dma_mem = mem;
+	/* FIXME: this routine just ignores DMA_MEMORY_INCLUDES_CHILDREN */
+
+	return 0;
+}
+
+int dma_declare_coherent_memory(struct device *dev, phys_addr_t phys_addr,
+				dma_addr_t device_addr, size_t size, int flags)
+{
+	struct dma_coherent_mem *mem;
+	int ret;
+
+	ret = dma_init_coherent_memory(phys_addr, device_addr, size, flags,
+				       &mem);
+	if (ret == 0)
+		return 0;
+
+	if (dma_assign_coherent_memory(dev, mem) == 0)
+		return ret;
+
+	dma_release_coherent_memory(mem);
+	return 0;
+}
 EXPORT_SYMBOL(dma_declare_coherent_memory);
 
 void dma_release_declared_memory(struct device *dev)
@@ -69,10 +109,8 @@ void dma_release_declared_memory(struct device *dev)
 
 	if (!mem)
 		return;
+	dma_release_coherent_memory(mem);
 	dev->dma_mem = NULL;
-	iounmap(mem->virt_base);
-	kfree(mem->bitmap);
-	kfree(mem);
 }
 EXPORT_SYMBOL(dma_release_declared_memory);
 
@@ -80,6 +118,7 @@ void *dma_mark_declared_memory_occupied(struct device *dev,
 					dma_addr_t device_addr, size_t size)
 {
 	struct dma_coherent_mem *mem = dev->dma_mem;
+	unsigned long flags;
 	int pos, err;
 
 	size += device_addr & ~PAGE_MASK;
@@ -87,8 +126,11 @@ void *dma_mark_declared_memory_occupied(struct device *dev,
 	if (!mem)
 		return ERR_PTR(-EINVAL);
 
+	spin_lock_irqsave(&mem->spinlock, flags);
 	pos = (device_addr - mem->device_base) >> PAGE_SHIFT;
 	err = bitmap_allocate_region(mem->bitmap, pos, get_order(size));
+	spin_unlock_irqrestore(&mem->spinlock, flags);
+
 	if (err != 0)
 		return ERR_PTR(err);
 	return mem->virt_base + (pos << PAGE_SHIFT);
@@ -115,6 +157,7 @@ int dma_alloc_from_coherent(struct device *dev, ssize_t size,
 {
 	struct dma_coherent_mem *mem;
 	int order = get_order(size);
+	unsigned long flags;
 	int pageno;
 
 	if (!dev)
@@ -124,6 +167,7 @@ int dma_alloc_from_coherent(struct device *dev, ssize_t size,
 		return 0;
 
 	*ret = NULL;
+	spin_lock_irqsave(&mem->spinlock, flags);
 
 	if (unlikely(size > (mem->size << PAGE_SHIFT)))
 		goto err;
@@ -138,10 +182,12 @@ int dma_alloc_from_coherent(struct device *dev, ssize_t size,
 	*dma_handle = mem->device_base + (pageno << PAGE_SHIFT);
 	*ret = mem->virt_base + (pageno << PAGE_SHIFT);
 	memset(*ret, 0, size);
+	spin_unlock_irqrestore(&mem->spinlock, flags);
 
 	return 1;
 
 err:
+	spin_unlock_irqrestore(&mem->spinlock, flags);
 	/*
 	 * In the case where the allocation can not be satisfied from the
 	 * per-device area, try to fall back to generic memory if the
@@ -171,8 +217,11 @@ int dma_release_from_coherent(struct device *dev, int order, void *vaddr)
 	if (mem && vaddr >= mem->virt_base && vaddr <
 		   (mem->virt_base + (mem->size << PAGE_SHIFT))) {
 		int page = (vaddr - mem->virt_base) >> PAGE_SHIFT;
+		unsigned long flags;
 
+		spin_lock_irqsave(&mem->spinlock, flags);
 		bitmap_release_region(mem->bitmap, page, order);
+		spin_unlock_irqrestore(&mem->spinlock, flags);
 		return 1;
 	}
 	return 0;
@@ -218,3 +267,61 @@ int dma_mmap_from_coherent(struct device *dev, struct vm_area_struct *vma,
 	return 0;
 }
 EXPORT_SYMBOL(dma_mmap_from_coherent);
+
+/*
+ * Support for reserved memory regions defined in device tree
+ */
+#ifdef CONFIG_OF_RESERVED_MEM
+#include <linux/of.h>
+#include <linux/of_fdt.h>
+#include <linux/of_reserved_mem.h>
+
+static int rmem_dma_device_init(struct reserved_mem *rmem, struct device *dev)
+{
+	struct dma_coherent_mem *mem = rmem->priv;
+
+	if (!mem &&
+	    dma_init_coherent_memory(rmem->base, rmem->base, rmem->size,
+				     DMA_MEMORY_MAP | DMA_MEMORY_EXCLUSIVE,
+				     &mem) != DMA_MEMORY_MAP) {
+		pr_err("Reserved memory: failed to init DMA memory pool at %pa, size %ld MiB\n",
+			&rmem->base, (unsigned long)rmem->size / SZ_1M);
+		return -ENODEV;
+	}
+	rmem->priv = mem;
+	dma_assign_coherent_memory(dev, mem);
+	return 0;
+}
+
+static void rmem_dma_device_release(struct reserved_mem *rmem,
+				    struct device *dev)
+{
+	dev->dma_mem = NULL;
+}
+
+static const struct reserved_mem_ops rmem_dma_ops = {
+	.device_init	= rmem_dma_device_init,
+	.device_release	= rmem_dma_device_release,
+};
+
+static int __init rmem_dma_setup(struct reserved_mem *rmem)
+{
+	unsigned long node = rmem->fdt_node;
+
+	if (of_get_flat_dt_prop(node, "reusable", NULL))
+		return -EINVAL;
+
+#ifdef CONFIG_ARM
+	if (!of_get_flat_dt_prop(node, "no-map", NULL)) {
+		pr_err("Reserved memory: regions without no-map are not yet supported\n");
+		return -EINVAL;
+	}
+#endif
+
+	rmem->ops = &rmem_dma_ops;
+	pr_info("Reserved memory: created DMA memory pool at %pa, size %ld MiB\n",
+		&rmem->base, (unsigned long)rmem->size / SZ_1M);
+	return 0;
+}
+RESERVEDMEM_OF_DECLARE(dma, "shared-dma-pool", rmem_dma_setup);
+#endif
diff --git a/drivers/base/dma-contiguous.c b/drivers/base/dma-contiguous.c
index 6606abdf880c..950fff9ce453 100644
--- a/drivers/base/dma-contiguous.c
+++ b/drivers/base/dma-contiguous.c
@@ -211,3 +211,70 @@ bool dma_release_from_contiguous(struct device *dev, struct page *pages,
 {
 	return cma_release(dev_get_cma_area(dev), pages, count);
 }
+
+/*
+ * Support for reserved memory regions defined in device tree
+ */
+#ifdef CONFIG_OF_RESERVED_MEM
+#include <linux/of.h>
+#include <linux/of_fdt.h>
+#include <linux/of_reserved_mem.h>
+
+#undef pr_fmt
+#define pr_fmt(fmt) fmt
+
+static int rmem_cma_device_init(struct reserved_mem *rmem, struct device *dev)
+{
+	dev_set_cma_area(dev, rmem->priv);
+	return 0;
+}
+
+static void rmem_cma_device_release(struct reserved_mem *rmem,
+				    struct device *dev)
+{
+	dev_set_cma_area(dev, NULL);
+}
+
+static const struct reserved_mem_ops rmem_cma_ops = {
+	.device_init	= rmem_cma_device_init,
+	.device_release = rmem_cma_device_release,
+};
+
+static int __init rmem_cma_setup(struct reserved_mem *rmem)
+{
+	phys_addr_t align = PAGE_SIZE << max(MAX_ORDER - 1, pageblock_order);
+	phys_addr_t mask = align - 1;
+	unsigned long node = rmem->fdt_node;
+	struct cma *cma;
+	int err;
+
+	if (!of_get_flat_dt_prop(node, "reusable", NULL) ||
+	    of_get_flat_dt_prop(node, "no-map", NULL))
+		return -EINVAL;
+
+	if ((rmem->base & mask) || (rmem->size & mask)) {
+		pr_err("Reserved memory: incorrect alignment of CMA region\n");
+		return -EINVAL;
+	}
+
+	err = cma_init_reserved_mem(rmem->base, rmem->size, 0, &cma);
+	if (err) {
+		pr_err("Reserved memory: unable to setup CMA region\n");
+		return err;
+	}
+	/* Architecture specific contiguous memory fixup. */
+	dma_contiguous_early_fixup(rmem->base, rmem->size);
+
+	if (of_get_flat_dt_prop(node, "linux,cma-default", NULL))
+		dma_contiguous_set_default(cma);
+
+	rmem->ops = &rmem_cma_ops;
+	rmem->priv = cma;
+
+	pr_info("Reserved memory: created CMA memory pool at %pa, size %ld MiB\n",
+		&rmem->base, (unsigned long)rmem->size / SZ_1M);
+
+	return 0;
+}
+RESERVEDMEM_OF_DECLARE(cma, "shared-dma-pool", rmem_cma_setup);
+#endif
diff --git a/drivers/base/dma-mapping.c b/drivers/base/dma-mapping.c
index 6cd08e145bfa..9e8bbdd470ca 100644
--- a/drivers/base/dma-mapping.c
+++ b/drivers/base/dma-mapping.c
@@ -10,6 +10,8 @@
 #include <linux/dma-mapping.h>
 #include <linux/export.h>
 #include <linux/gfp.h>
+#include <linux/slab.h>
+#include <linux/vmalloc.h>
 #include <asm-generic/dma-coherent.h>
 
 /*
@@ -267,3 +269,73 @@ int dma_common_mmap(struct device *dev, struct vm_area_struct *vma,
 	return ret;
 }
 EXPORT_SYMBOL(dma_common_mmap);
+
+#ifdef CONFIG_MMU
+/*
+ * remaps an array of PAGE_SIZE pages into another vm_area
+ * Cannot be used in non-sleeping contexts
+ */
+void *dma_common_pages_remap(struct page **pages, size_t size,
+			unsigned long vm_flags, pgprot_t prot,
+			const void *caller)
+{
+	struct vm_struct *area;
+
+	area = get_vm_area_caller(size, vm_flags, caller);
+	if (!area)
+		return NULL;
+
+	area->pages = pages;
+
+	if (map_vm_area(area, prot, pages)) {
+		vunmap(area->addr);
+		return NULL;
+	}
+
+	return area->addr;
+}
+
+/*
+ * remaps an allocated contiguous region into another vm_area.
+ * Cannot be used in non-sleeping contexts
+ */
+
+void *dma_common_contiguous_remap(struct page *page, size_t size,
+			unsigned long vm_flags,
+			pgprot_t prot, const void *caller)
+{
+	int i;
+	struct page **pages;
+	void *ptr;
+	unsigned long pfn;
+
+	pages = kmalloc(sizeof(struct page *) << get_order(size), GFP_KERNEL);
+	if (!pages)
+		return NULL;
+
+	for (i = 0, pfn = page_to_pfn(page); i < (size >> PAGE_SHIFT); i++)
+		pages[i] = pfn_to_page(pfn + i);
+
+	ptr = dma_common_pages_remap(pages, size, vm_flags, prot, caller);
+
+	kfree(pages);
+
+	return ptr;
+}
+
+/*
+ * unmaps a range previously mapped by dma_common_*_remap
+ */
+void dma_common_free_remap(void *cpu_addr, size_t size, unsigned long vm_flags)
+{
+	struct vm_struct *area = find_vm_area(cpu_addr);
+
+	if (!area || (area->flags & vm_flags) != vm_flags) {
+		WARN(1, "trying to free invalid coherent area: %p\n", cpu_addr);
+		return;
+	}
+
+	unmap_kernel_range((unsigned long)cpu_addr, size);
+	vunmap(cpu_addr);
+}
+#endif
diff --git a/drivers/base/firmware_class.c b/drivers/base/firmware_class.c
index bf424305f3dc..58470c395301 100644
--- a/drivers/base/firmware_class.c
+++ b/drivers/base/firmware_class.c
@@ -591,8 +591,7 @@ static int fw_map_pages_buf(struct firmware_buf *buf)
 	if (!buf->is_paged_buf)
 		return 0;
 
-	if (buf->data)
-		vunmap(buf->data);
+	vunmap(buf->data);
 	buf->data = vmap(buf->pages, buf->nr_pages, 0, PAGE_KERNEL_RO);
 	if (!buf->data)
 		return -ENOMEM;
@@ -925,7 +924,7 @@ static int _request_firmware_load(struct firmware_priv *fw_priv,
 		kobject_uevent(&fw_priv->dev.kobj, KOBJ_ADD);
 	}
 
-	wait_for_completion(&buf->completion);
+	retval = wait_for_completion_interruptible(&buf->completion);
 
 	cancel_delayed_work_sync(&fw_priv->timeout_work);
 	if (is_fw_load_aborted(buf))
@@ -1004,7 +1003,7 @@ static int sync_cached_firmware_buf(struct firmware_buf *buf)
 			break;
 		}
 		mutex_unlock(&fw_lock);
-		wait_for_completion(&buf->completion);
+		ret = wait_for_completion_interruptible(&buf->completion);
 		mutex_lock(&fw_lock);
 	}
 	mutex_unlock(&fw_lock);
@@ -1105,6 +1104,9 @@ _request_firmware(const struct firmware **firmware_p, const char *name,
 	if (!firmware_p)
 		return -EINVAL;
 
+	if (!name || name[0] == '\0')
+		return -EINVAL;
+
 	ret = _request_firmware_prepare(&fw, name, device);
 	if (ret <= 0) /* error or already assigned */
 		goto out;
diff --git a/drivers/base/memory.c b/drivers/base/memory.c
index a2e13e250bba..85be040a21c8 100644
--- a/drivers/base/memory.c
+++ b/drivers/base/memory.c
@@ -228,8 +228,8 @@ memory_block_action(unsigned long phys_index, unsigned long action, int online_t
 	struct page *first_page;
 	int ret;
 
-	first_page = pfn_to_page(phys_index << PFN_SECTION_SHIFT);
-	start_pfn = page_to_pfn(first_page);
+	start_pfn = phys_index << PFN_SECTION_SHIFT;
+	first_page = pfn_to_page(start_pfn);
 
 	switch (action) {
 		case MEM_ONLINE:
@@ -373,6 +373,45 @@ static ssize_t show_phys_device(struct device *dev,
 	return sprintf(buf, "%d\n", mem->phys_device);
 }
 
+#ifdef CONFIG_MEMORY_HOTREMOVE
+static ssize_t show_valid_zones(struct device *dev,
+				struct device_attribute *attr, char *buf)
+{
+	struct memory_block *mem = to_memory_block(dev);
+	unsigned long start_pfn, end_pfn;
+	unsigned long nr_pages = PAGES_PER_SECTION * sections_per_block;
+	struct page *first_page;
+	struct zone *zone;
+
+	start_pfn = section_nr_to_pfn(mem->start_section_nr);
+	end_pfn = start_pfn + nr_pages;
+	first_page = pfn_to_page(start_pfn);
+
+	/* The block contains more than one zone can not be offlined. */
+	if (!test_pages_in_a_zone(start_pfn, end_pfn))
+		return sprintf(buf, "none\n");
+
+	zone = page_zone(first_page);
+
+	if (zone_idx(zone) == ZONE_MOVABLE - 1) {
+		/*The mem block is the last memoryblock of this zone.*/
+		if (end_pfn == zone_end_pfn(zone))
+			return sprintf(buf, "%s %s\n",
+					zone->name, (zone + 1)->name);
+	}
+
+	if (zone_idx(zone) == ZONE_MOVABLE) {
+		/*The mem block is the first memoryblock of ZONE_MOVABLE.*/
+		if (start_pfn == zone->zone_start_pfn)
+			return sprintf(buf, "%s %s\n",
+					zone->name, (zone - 1)->name);
+	}
+
+	return sprintf(buf, "%s\n", zone->name);
+}
+static DEVICE_ATTR(valid_zones, 0444, show_valid_zones, NULL);
+#endif
+
 static DEVICE_ATTR(phys_index, 0444, show_mem_start_phys_index, NULL);
 static DEVICE_ATTR(state, 0644, show_mem_state, store_mem_state);
 static DEVICE_ATTR(phys_device, 0444, show_phys_device, NULL);
@@ -523,6 +562,9 @@ static struct attribute *memory_memblk_attrs[] = {
 	&dev_attr_state.attr,
 	&dev_attr_phys_device.attr,
 	&dev_attr_removable.attr,
+#ifdef CONFIG_MEMORY_HOTREMOVE
+	&dev_attr_valid_zones.attr,
+#endif
 	NULL
 };
 
diff --git a/drivers/base/node.c b/drivers/base/node.c
index c6d3ae05f1ca..a3b82e9c7f20 100644
--- a/drivers/base/node.c
+++ b/drivers/base/node.c
@@ -25,32 +25,26 @@ static struct bus_type node_subsys = {
 };
 
 
-static ssize_t node_read_cpumap(struct device *dev, int type, char *buf)
+static ssize_t node_read_cpumap(struct device *dev, bool list, char *buf)
 {
 	struct node *node_dev = to_node(dev);
 	const struct cpumask *mask = cpumask_of_node(node_dev->dev.id);
-	int len;
 
 	/* 2008/04/07: buf currently PAGE_SIZE, need 9 chars per 32 bits. */
 	BUILD_BUG_ON((NR_CPUS/32 * 9) > (PAGE_SIZE-1));
 
-	len = type?
-		cpulist_scnprintf(buf, PAGE_SIZE-2, mask) :
-		cpumask_scnprintf(buf, PAGE_SIZE-2, mask);
- 	buf[len++] = '\n';
- 	buf[len] = '\0';
-	return len;
+	return cpumap_print_to_pagebuf(list, buf, mask);
 }
 
 static inline ssize_t node_read_cpumask(struct device *dev,
 				struct device_attribute *attr, char *buf)
 {
-	return node_read_cpumap(dev, 0, buf);
+	return node_read_cpumap(dev, false, buf);
 }
 static inline ssize_t node_read_cpulist(struct device *dev,
 				struct device_attribute *attr, char *buf)
 {
-	return node_read_cpumap(dev, 1, buf);
+	return node_read_cpumap(dev, true, buf);
 }
 
 static DEVICE_ATTR(cpumap,  S_IRUGO, node_read_cpumask, NULL);
@@ -289,8 +283,6 @@ static int register_node(struct node *node, int num, struct node *parent)
 		device_create_file(&node->dev, &dev_attr_distance);
 		device_create_file(&node->dev, &dev_attr_vmstat);
 
-		scan_unevictable_register_node(node);
-
 		hugetlb_register_node(node);
 
 		compaction_register_node(node);
@@ -314,7 +306,6 @@ void unregister_node(struct node *node)
 	device_remove_file(&node->dev, &dev_attr_distance);
 	device_remove_file(&node->dev, &dev_attr_vmstat);
 
-	scan_unevictable_unregister_node(node);
 	hugetlb_unregister_node(node);		/* no-op, if memoryless node */
 
 	device_unregister(&node->dev);
@@ -603,7 +594,6 @@ void unregister_one_node(int nid)
 		return;
 
 	unregister_node(node_devices[nid]);
-	kfree(node_devices[nid]);
 	node_devices[nid] = NULL;
 }
 
diff --git a/drivers/base/platform.c b/drivers/base/platform.c
index ab4f4ce02722..9421fed40905 100644
--- a/drivers/base/platform.c
+++ b/drivers/base/platform.c
@@ -21,6 +21,7 @@
 #include <linux/err.h>
 #include <linux/slab.h>
 #include <linux/pm_runtime.h>
+#include <linux/pm_domain.h>
 #include <linux/idr.h>
 #include <linux/acpi.h>
 #include <linux/clk/clk-conf.h>
@@ -506,11 +507,12 @@ static int platform_drv_probe(struct device *_dev)
 	if (ret < 0)
 		return ret;
 
-	acpi_dev_pm_attach(_dev, true);
-
-	ret = drv->probe(dev);
-	if (ret)
-		acpi_dev_pm_detach(_dev, true);
+	ret = dev_pm_domain_attach(_dev, true);
+	if (ret != -EPROBE_DEFER) {
+		ret = drv->probe(dev);
+		if (ret)
+			dev_pm_domain_detach(_dev, true);
+	}
 
 	if (drv->prevent_deferred_probe && ret == -EPROBE_DEFER) {
 		dev_warn(_dev, "probe deferral not supported\n");
@@ -532,7 +534,7 @@ static int platform_drv_remove(struct device *_dev)
 	int ret;
 
 	ret = drv->remove(dev);
-	acpi_dev_pm_detach(_dev, true);
+	dev_pm_domain_detach(_dev, true);
 
 	return ret;
 }
@@ -543,7 +545,7 @@ static void platform_drv_shutdown(struct device *_dev)
 	struct platform_device *dev = to_platform_device(_dev);
 
 	drv->shutdown(dev);
-	acpi_dev_pm_detach(_dev, true);
+	dev_pm_domain_detach(_dev, true);
 }
 
 /**
@@ -578,9 +580,10 @@ void platform_driver_unregister(struct platform_driver *drv)
 EXPORT_SYMBOL_GPL(platform_driver_unregister);
 
 /**
- * platform_driver_probe - register driver for non-hotpluggable device
+ * __platform_driver_probe - register driver for non-hotpluggable device
  * @drv: platform driver structure
  * @probe: the driver probe routine, probably from an __init section
+ * @module: module which will be the owner of the driver
  *
  * Use this instead of platform_driver_register() when you know the device
  * is not hotpluggable and has already been registered, and you want to
@@ -596,8 +599,8 @@ EXPORT_SYMBOL_GPL(platform_driver_unregister);
  * Returns zero if the driver registered and bound to a device, else returns
  * a negative error code and with the driver not registered.
  */
-int __init_or_module platform_driver_probe(struct platform_driver *drv,
-		int (*probe)(struct platform_device *))
+int __init_or_module __platform_driver_probe(struct platform_driver *drv,
+		int (*probe)(struct platform_device *), struct module *module)
 {
 	int retval, code;
 
@@ -612,7 +615,7 @@ int __init_or_module platform_driver_probe(struct platform_driver *drv,
 
 	/* temporary section violation during probe() */
 	drv->probe = probe;
-	retval = code = platform_driver_register(drv);
+	retval = code = __platform_driver_register(drv, module);
 
 	/*
 	 * Fixup that section violation, being paranoid about code scanning
@@ -631,27 +634,28 @@ int __init_or_module platform_driver_probe(struct platform_driver *drv,
 		platform_driver_unregister(drv);
 	return retval;
 }
-EXPORT_SYMBOL_GPL(platform_driver_probe);
+EXPORT_SYMBOL_GPL(__platform_driver_probe);
 
 /**
- * platform_create_bundle - register driver and create corresponding device
+ * __platform_create_bundle - register driver and create corresponding device
  * @driver: platform driver structure
  * @probe: the driver probe routine, probably from an __init section
  * @res: set of resources that needs to be allocated for the device
  * @n_res: number of resources
  * @data: platform specific data for this platform device
  * @size: size of platform specific data
+ * @module: module which will be the owner of the driver
  *
  * Use this in legacy-style modules that probe hardware directly and
  * register a single platform device and corresponding platform driver.
  *
  * Returns &struct platform_device pointer on success, or ERR_PTR() on error.
  */
-struct platform_device * __init_or_module platform_create_bundle(
+struct platform_device * __init_or_module __platform_create_bundle(
 			struct platform_driver *driver,
 			int (*probe)(struct platform_device *),
 			struct resource *res, unsigned int n_res,
-			const void *data, size_t size)
+			const void *data, size_t size, struct module *module)
 {
 	struct platform_device *pdev;
 	int error;
@@ -674,7 +678,7 @@ struct platform_device * __init_or_module platform_create_bundle(
 	if (error)
 		goto err_pdev_put;
 
-	error = platform_driver_probe(driver, probe);
+	error = __platform_driver_probe(driver, probe, module);
 	if (error)
 		goto err_pdev_del;
 
@@ -687,7 +691,7 @@ err_pdev_put:
 err_out:
 	return ERR_PTR(error);
 }
-EXPORT_SYMBOL_GPL(platform_create_bundle);
+EXPORT_SYMBOL_GPL(__platform_create_bundle);
 
 /* modalias support enables more hands-off userspace setup:
  * (a) environment variable lets new-style hotplug events work once system is
@@ -1004,6 +1008,7 @@ int __init platform_bus_init(void)
 	error =  bus_register(&platform_bus_type);
 	if (error)
 		device_unregister(&platform_bus);
+	of_platform_register_reconfig_notifier();
 	return error;
 }
 
diff --git a/drivers/base/power/clock_ops.c b/drivers/base/power/clock_ops.c
index b99e6c06ee67..d626576a4f75 100644
--- a/drivers/base/power/clock_ops.c
+++ b/drivers/base/power/clock_ops.c
@@ -12,6 +12,7 @@
 #include <linux/pm.h>
 #include <linux/pm_clock.h>
 #include <linux/clk.h>
+#include <linux/clkdev.h>
 #include <linux/slab.h>
 #include <linux/err.h>
 
@@ -34,14 +35,20 @@ struct pm_clock_entry {
 /**
  * pm_clk_enable - Enable a clock, reporting any errors
  * @dev: The device for the given clock
- * @clk: The clock being enabled.
+ * @ce: PM clock entry corresponding to the clock.
  */
-static inline int __pm_clk_enable(struct device *dev, struct clk *clk)
+static inline int __pm_clk_enable(struct device *dev, struct pm_clock_entry *ce)
 {
-	int ret = clk_enable(clk);
-	if (ret)
-		dev_err(dev, "%s: failed to enable clk %p, error %d\n",
-			__func__, clk, ret);
+	int ret;
+
+	if (ce->status < PCE_STATUS_ERROR) {
+		ret = clk_enable(ce->clk);
+		if (!ret)
+			ce->status = PCE_STATUS_ENABLED;
+		else
+			dev_err(dev, "%s: failed to enable clk %p, error %d\n",
+				__func__, ce->clk, ret);
+	}
 
 	return ret;
 }
@@ -53,7 +60,8 @@ static inline int __pm_clk_enable(struct device *dev, struct clk *clk)
  */
 static void pm_clk_acquire(struct device *dev, struct pm_clock_entry *ce)
 {
-	ce->clk = clk_get(dev, ce->con_id);
+	if (!ce->clk)
+		ce->clk = clk_get(dev, ce->con_id);
 	if (IS_ERR(ce->clk)) {
 		ce->status = PCE_STATUS_ERROR;
 	} else {
@@ -63,15 +71,8 @@ static void pm_clk_acquire(struct device *dev, struct pm_clock_entry *ce)
 	}
 }
 
-/**
- * pm_clk_add - Start using a device clock for power management.
- * @dev: Device whose clock is going to be used for power management.
- * @con_id: Connection ID of the clock.
- *
- * Add the clock represented by @con_id to the list of clocks used for
- * the power management of @dev.
- */
-int pm_clk_add(struct device *dev, const char *con_id)
+static int __pm_clk_add(struct device *dev, const char *con_id,
+			struct clk *clk)
 {
 	struct pm_subsys_data *psd = dev_to_psd(dev);
 	struct pm_clock_entry *ce;
@@ -93,6 +94,12 @@ int pm_clk_add(struct device *dev, const char *con_id)
 			kfree(ce);
 			return -ENOMEM;
 		}
+	} else {
+		if (IS_ERR(ce->clk) || !__clk_get(clk)) {
+			kfree(ce);
+			return -ENOENT;
+		}
+		ce->clk = clk;
 	}
 
 	pm_clk_acquire(dev, ce);
@@ -104,6 +111,32 @@ int pm_clk_add(struct device *dev, const char *con_id)
 }
 
 /**
+ * pm_clk_add - Start using a device clock for power management.
+ * @dev: Device whose clock is going to be used for power management.
+ * @con_id: Connection ID of the clock.
+ *
+ * Add the clock represented by @con_id to the list of clocks used for
+ * the power management of @dev.
+ */
+int pm_clk_add(struct device *dev, const char *con_id)
+{
+	return __pm_clk_add(dev, con_id, NULL);
+}
+
+/**
+ * pm_clk_add_clk - Start using a device clock for power management.
+ * @dev: Device whose clock is going to be used for power management.
+ * @clk: Clock pointer
+ *
+ * Add the clock to the list of clocks used for the power management of @dev.
+ * It will increment refcount on clock pointer, use clk_put() on it when done.
+ */
+int pm_clk_add_clk(struct device *dev, struct clk *clk)
+{
+	return __pm_clk_add(dev, NULL, clk);
+}
+
+/**
  * __pm_clk_remove - Destroy PM clock entry.
  * @ce: PM clock entry to destroy.
  */
@@ -223,10 +256,6 @@ void pm_clk_destroy(struct device *dev)
 	}
 }
 
-#endif /* CONFIG_PM */
-
-#ifdef CONFIG_PM_RUNTIME
-
 /**
  * pm_clk_suspend - Disable clocks in a device's PM clock list.
  * @dev: Device to disable the clocks for.
@@ -266,7 +295,6 @@ int pm_clk_resume(struct device *dev)
 	struct pm_subsys_data *psd = dev_to_psd(dev);
 	struct pm_clock_entry *ce;
 	unsigned long flags;
-	int ret;
 
 	dev_dbg(dev, "%s()\n", __func__);
 
@@ -275,13 +303,8 @@ int pm_clk_resume(struct device *dev)
 
 	spin_lock_irqsave(&psd->lock, flags);
 
-	list_for_each_entry(ce, &psd->clock_list, node) {
-		if (ce->status < PCE_STATUS_ERROR) {
-			ret = __pm_clk_enable(dev, ce->clk);
-			if (!ret)
-				ce->status = PCE_STATUS_ENABLED;
-		}
-	}
+	list_for_each_entry(ce, &psd->clock_list, node)
+		__pm_clk_enable(dev, ce);
 
 	spin_unlock_irqrestore(&psd->lock, flags);
 
@@ -346,63 +369,7 @@ static int pm_clk_notify(struct notifier_block *nb,
 	return 0;
 }
 
-#else /* !CONFIG_PM_RUNTIME */
-
-#ifdef CONFIG_PM
-
-/**
- * pm_clk_suspend - Disable clocks in a device's PM clock list.
- * @dev: Device to disable the clocks for.
- */
-int pm_clk_suspend(struct device *dev)
-{
-	struct pm_subsys_data *psd = dev_to_psd(dev);
-	struct pm_clock_entry *ce;
-	unsigned long flags;
-
-	dev_dbg(dev, "%s()\n", __func__);
-
-	/* If there is no driver, the clocks are already disabled. */
-	if (!psd || !dev->driver)
-		return 0;
-
-	spin_lock_irqsave(&psd->lock, flags);
-
-	list_for_each_entry_reverse(ce, &psd->clock_list, node)
-		clk_disable(ce->clk);
-
-	spin_unlock_irqrestore(&psd->lock, flags);
-
-	return 0;
-}
-
-/**
- * pm_clk_resume - Enable clocks in a device's PM clock list.
- * @dev: Device to enable the clocks for.
- */
-int pm_clk_resume(struct device *dev)
-{
-	struct pm_subsys_data *psd = dev_to_psd(dev);
-	struct pm_clock_entry *ce;
-	unsigned long flags;
-
-	dev_dbg(dev, "%s()\n", __func__);
-
-	/* If there is no driver, the clocks should remain disabled. */
-	if (!psd || !dev->driver)
-		return 0;
-
-	spin_lock_irqsave(&psd->lock, flags);
-
-	list_for_each_entry(ce, &psd->clock_list, node)
-		__pm_clk_enable(dev, ce->clk);
-
-	spin_unlock_irqrestore(&psd->lock, flags);
-
-	return 0;
-}
-
-#endif /* CONFIG_PM */
+#else /* !CONFIG_PM */
 
 /**
  * enable_clock - Enable a device clock.
@@ -482,7 +449,7 @@ static int pm_clk_notify(struct notifier_block *nb,
 	return 0;
 }
 
-#endif /* !CONFIG_PM_RUNTIME */
+#endif /* !CONFIG_PM */
 
 /**
  * pm_clk_add_notifier - Add bus type notifier for power management clocks.
diff --git a/drivers/base/power/common.c b/drivers/base/power/common.c
index df2e5eeaeb05..b0f138806bbc 100644
--- a/drivers/base/power/common.c
+++ b/drivers/base/power/common.c
@@ -11,6 +11,8 @@
 #include <linux/export.h>
 #include <linux/slab.h>
 #include <linux/pm_clock.h>
+#include <linux/acpi.h>
+#include <linux/pm_domain.h>
 
 /**
  * dev_pm_get_subsys_data - Create or refcount power.subsys_data for device.
@@ -82,3 +84,53 @@ int dev_pm_put_subsys_data(struct device *dev)
 	return ret;
 }
 EXPORT_SYMBOL_GPL(dev_pm_put_subsys_data);
+
+/**
+ * dev_pm_domain_attach - Attach a device to its PM domain.
+ * @dev: Device to attach.
+ * @power_on: Used to indicate whether we should power on the device.
+ *
+ * The @dev may only be attached to a single PM domain. By iterating through
+ * the available alternatives we try to find a valid PM domain for the device.
+ * As attachment succeeds, the ->detach() callback in the struct dev_pm_domain
+ * should be assigned by the corresponding attach function.
+ *
+ * This function should typically be invoked from subsystem level code during
+ * the probe phase. Especially for those that holds devices which requires
+ * power management through PM domains.
+ *
+ * Callers must ensure proper synchronization of this function with power
+ * management callbacks.
+ *
+ * Returns 0 on successfully attached PM domain or negative error code.
+ */
+int dev_pm_domain_attach(struct device *dev, bool power_on)
+{
+	int ret;
+
+	ret = acpi_dev_pm_attach(dev, power_on);
+	if (ret)
+		ret = genpd_dev_pm_attach(dev);
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(dev_pm_domain_attach);
+
+/**
+ * dev_pm_domain_detach - Detach a device from its PM domain.
+ * @dev: Device to attach.
+ * @power_off: Used to indicate whether we should power off the device.
+ *
+ * This functions will reverse the actions from dev_pm_domain_attach() and thus
+ * try to detach the @dev from its PM domain. Typically it should be invoked
+ * from subsystem level code during the remove phase.
+ *
+ * Callers must ensure proper synchronization of this function with power
+ * management callbacks.
+ */
+void dev_pm_domain_detach(struct device *dev, bool power_off)
+{
+	if (dev->pm_domain && dev->pm_domain->detach)
+		dev->pm_domain->detach(dev, power_off);
+}
+EXPORT_SYMBOL_GPL(dev_pm_domain_detach);
diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c
index eee55c1e5fde..0d8780c04a5e 100644
--- a/drivers/base/power/domain.c
+++ b/drivers/base/power/domain.c
@@ -8,9 +8,11 @@
 
 #include <linux/kernel.h>
 #include <linux/io.h>
+#include <linux/platform_device.h>
 #include <linux/pm_runtime.h>
 #include <linux/pm_domain.h>
 #include <linux/pm_qos.h>
+#include <linux/pm_clock.h>
 #include <linux/slab.h>
 #include <linux/err.h>
 #include <linux/sched.h>
@@ -25,10 +27,6 @@
 	__routine = genpd->dev_ops.callback; 			\
 	if (__routine) {					\
 		__ret = __routine(dev); 			\
-	} else {						\
-		__routine = dev_gpd_data(dev)->ops.callback;	\
-		if (__routine) 					\
-			__ret = __routine(dev);			\
 	}							\
 	__ret;							\
 })
@@ -70,8 +68,6 @@ static struct generic_pm_domain *pm_genpd_lookup_name(const char *domain_name)
 	return genpd;
 }
 
-#ifdef CONFIG_PM
-
 struct generic_pm_domain *dev_to_genpd(struct device *dev)
 {
 	if (IS_ERR_OR_NULL(dev->pm_domain))
@@ -147,13 +143,66 @@ static void genpd_recalc_cpu_exit_latency(struct generic_pm_domain *genpd)
 {
 	s64 usecs64;
 
-	if (!genpd->cpu_data)
+	if (!genpd->cpuidle_data)
 		return;
 
 	usecs64 = genpd->power_on_latency_ns;
 	do_div(usecs64, NSEC_PER_USEC);
-	usecs64 += genpd->cpu_data->saved_exit_latency;
-	genpd->cpu_data->idle_state->exit_latency = usecs64;
+	usecs64 += genpd->cpuidle_data->saved_exit_latency;
+	genpd->cpuidle_data->idle_state->exit_latency = usecs64;
+}
+
+static int genpd_power_on(struct generic_pm_domain *genpd)
+{
+	ktime_t time_start;
+	s64 elapsed_ns;
+	int ret;
+
+	if (!genpd->power_on)
+		return 0;
+
+	time_start = ktime_get();
+	ret = genpd->power_on(genpd);
+	if (ret)
+		return ret;
+
+	elapsed_ns = ktime_to_ns(ktime_sub(ktime_get(), time_start));
+	if (elapsed_ns <= genpd->power_on_latency_ns)
+		return ret;
+
+	genpd->power_on_latency_ns = elapsed_ns;
+	genpd->max_off_time_changed = true;
+	genpd_recalc_cpu_exit_latency(genpd);
+	pr_warn("%s: Power-%s latency exceeded, new value %lld ns\n",
+		genpd->name, "on", elapsed_ns);
+
+	return ret;
+}
+
+static int genpd_power_off(struct generic_pm_domain *genpd)
+{
+	ktime_t time_start;
+	s64 elapsed_ns;
+	int ret;
+
+	if (!genpd->power_off)
+		return 0;
+
+	time_start = ktime_get();
+	ret = genpd->power_off(genpd);
+	if (ret == -EBUSY)
+		return ret;
+
+	elapsed_ns = ktime_to_ns(ktime_sub(ktime_get(), time_start));
+	if (elapsed_ns <= genpd->power_off_latency_ns)
+		return ret;
+
+	genpd->power_off_latency_ns = elapsed_ns;
+	genpd->max_off_time_changed = true;
+	pr_warn("%s: Power-%s latency exceeded, new value %lld ns\n",
+		genpd->name, "off", elapsed_ns);
+
+	return ret;
 }
 
 /**
@@ -193,9 +242,9 @@ static int __pm_genpd_poweron(struct generic_pm_domain *genpd)
 		return 0;
 	}
 
-	if (genpd->cpu_data) {
+	if (genpd->cpuidle_data) {
 		cpuidle_pause_and_lock();
-		genpd->cpu_data->idle_state->disabled = true;
+		genpd->cpuidle_data->idle_state->disabled = true;
 		cpuidle_resume_and_unlock();
 		goto out;
 	}
@@ -227,25 +276,9 @@ static int __pm_genpd_poweron(struct generic_pm_domain *genpd)
 		}
 	}
 
-	if (genpd->power_on) {
-		ktime_t time_start = ktime_get();
-		s64 elapsed_ns;
-
-		ret = genpd->power_on(genpd);
-		if (ret)
-			goto err;
-
-		elapsed_ns = ktime_to_ns(ktime_sub(ktime_get(), time_start));
-		if (elapsed_ns > genpd->power_on_latency_ns) {
-			genpd->power_on_latency_ns = elapsed_ns;
-			genpd->max_off_time_changed = true;
-			genpd_recalc_cpu_exit_latency(genpd);
-			if (genpd->name)
-				pr_warning("%s: Power-on latency exceeded, "
-					"new value %lld ns\n", genpd->name,
-					elapsed_ns);
-		}
-	}
+	ret = genpd_power_on(genpd);
+	if (ret)
+		goto err;
 
  out:
 	genpd_set_active(genpd);
@@ -285,10 +318,6 @@ int pm_genpd_name_poweron(const char *domain_name)
 	return genpd ? pm_genpd_poweron(genpd) : -EINVAL;
 }
 
-#endif /* CONFIG_PM */
-
-#ifdef CONFIG_PM_RUNTIME
-
 static int genpd_start_dev_no_timing(struct generic_pm_domain *genpd,
 				     struct device *dev)
 {
@@ -368,9 +397,19 @@ static int __pm_genpd_save_device(struct pm_domain_data *pdd,
 	struct device *dev = pdd->dev;
 	int ret = 0;
 
-	if (gpd_data->need_restore)
+	if (gpd_data->need_restore > 0)
 		return 0;
 
+	/*
+	 * If the value of the need_restore flag is still unknown at this point,
+	 * we trust that pm_genpd_poweroff() has verified that the device is
+	 * already runtime PM suspended.
+	 */
+	if (gpd_data->need_restore < 0) {
+		gpd_data->need_restore = 1;
+		return 0;
+	}
+
 	mutex_unlock(&genpd->lock);
 
 	genpd_start_dev(genpd, dev);
@@ -380,7 +419,7 @@ static int __pm_genpd_save_device(struct pm_domain_data *pdd,
 	mutex_lock(&genpd->lock);
 
 	if (!ret)
-		gpd_data->need_restore = true;
+		gpd_data->need_restore = 1;
 
 	return ret;
 }
@@ -396,12 +435,17 @@ static void __pm_genpd_restore_device(struct pm_domain_data *pdd,
 {
 	struct generic_pm_domain_data *gpd_data = to_gpd_data(pdd);
 	struct device *dev = pdd->dev;
-	bool need_restore = gpd_data->need_restore;
+	int need_restore = gpd_data->need_restore;
 
-	gpd_data->need_restore = false;
+	gpd_data->need_restore = 0;
 	mutex_unlock(&genpd->lock);
 
 	genpd_start_dev(genpd, dev);
+
+	/*
+	 * Call genpd_restore_dev() for recently added devices too (need_restore
+	 * is negative then).
+	 */
 	if (need_restore)
 		genpd_restore_dev(genpd, dev);
 
@@ -430,7 +474,7 @@ static bool genpd_abort_poweroff(struct generic_pm_domain *genpd)
  * Queue up the execution of pm_genpd_poweroff() unless it's already been done
  * before.
  */
-void genpd_queue_power_off_work(struct generic_pm_domain *genpd)
+static void genpd_queue_power_off_work(struct generic_pm_domain *genpd)
 {
 	queue_work(pm_wq, &genpd->power_off_work);
 }
@@ -520,32 +564,27 @@ static int pm_genpd_poweroff(struct generic_pm_domain *genpd)
 		}
 	}
 
-	if (genpd->cpu_data) {
+	if (genpd->cpuidle_data) {
 		/*
-		 * If cpu_data is set, cpuidle should turn the domain off when
-		 * the CPU in it is idle.  In that case we don't decrement the
-		 * subdomain counts of the master domains, so that power is not
-		 * removed from the current domain prematurely as a result of
-		 * cutting off the masters' power.
+		 * If cpuidle_data is set, cpuidle should turn the domain off
+		 * when the CPU in it is idle.  In that case we don't decrement
+		 * the subdomain counts of the master domains, so that power is
+		 * not removed from the current domain prematurely as a result
+		 * of cutting off the masters' power.
 		 */
 		genpd->status = GPD_STATE_POWER_OFF;
 		cpuidle_pause_and_lock();
-		genpd->cpu_data->idle_state->disabled = false;
+		genpd->cpuidle_data->idle_state->disabled = false;
 		cpuidle_resume_and_unlock();
 		goto out;
 	}
 
 	if (genpd->power_off) {
-		ktime_t time_start;
-		s64 elapsed_ns;
-
 		if (atomic_read(&genpd->sd_count) > 0) {
 			ret = -EBUSY;
 			goto out;
 		}
 
-		time_start = ktime_get();
-
 		/*
 		 * If sd_count > 0 at this point, one of the subdomains hasn't
 		 * managed to call pm_genpd_poweron() for the master yet after
@@ -554,21 +593,11 @@ static int pm_genpd_poweroff(struct generic_pm_domain *genpd)
 		 * the pm_genpd_poweron() restore power for us (this shouldn't
 		 * happen very often).
 		 */
-		ret = genpd->power_off(genpd);
+		ret = genpd_power_off(genpd);
 		if (ret == -EBUSY) {
 			genpd_set_active(genpd);
 			goto out;
 		}
-
-		elapsed_ns = ktime_to_ns(ktime_sub(ktime_get(), time_start));
-		if (elapsed_ns > genpd->power_off_latency_ns) {
-			genpd->power_off_latency_ns = elapsed_ns;
-			genpd->max_off_time_changed = true;
-			if (genpd->name)
-				pr_warning("%s: Power-off latency exceeded, "
-					"new value %lld ns\n", genpd->name,
-					elapsed_ns);
-		}
 	}
 
 	genpd->status = GPD_STATE_POWER_OFF;
@@ -610,6 +639,7 @@ static void genpd_power_off_work_fn(struct work_struct *work)
 static int pm_genpd_runtime_suspend(struct device *dev)
 {
 	struct generic_pm_domain *genpd;
+	struct generic_pm_domain_data *gpd_data;
 	bool (*stop_ok)(struct device *__dev);
 	int ret;
 
@@ -619,8 +649,6 @@ static int pm_genpd_runtime_suspend(struct device *dev)
 	if (IS_ERR(genpd))
 		return -EINVAL;
 
-	might_sleep_if(!genpd->dev_irq_safe);
-
 	stop_ok = genpd->gov ? genpd->gov->stop_ok : NULL;
 	if (stop_ok && !stop_ok(dev))
 		return -EBUSY;
@@ -637,6 +665,16 @@ static int pm_genpd_runtime_suspend(struct device *dev)
 		return 0;
 
 	mutex_lock(&genpd->lock);
+
+	/*
+	 * If we have an unknown state of the need_restore flag, it means none
+	 * of the runtime PM callbacks has been invoked yet. Let's update the
+	 * flag to reflect that the current state is active.
+	 */
+	gpd_data = to_gpd_data(dev->power.subsys_data->domain_data);
+	if (gpd_data->need_restore < 0)
+		gpd_data->need_restore = 0;
+
 	genpd->in_progress++;
 	pm_genpd_poweroff(genpd);
 	genpd->in_progress--;
@@ -665,8 +703,6 @@ static int pm_genpd_runtime_resume(struct device *dev)
 	if (IS_ERR(genpd))
 		return -EINVAL;
 
-	might_sleep_if(!genpd->dev_irq_safe);
-
 	/* If power.irq_safe, the PM domain is never powered off. */
 	if (dev->power.irq_safe)
 		return genpd_start_dev_no_timing(genpd, dev);
@@ -733,20 +769,12 @@ void pm_genpd_poweroff_unused(void)
 	mutex_unlock(&gpd_list_lock);
 }
 
-#else
-
-static inline int genpd_dev_pm_qos_notifier(struct notifier_block *nb,
-					    unsigned long val, void *ptr)
+static int __init genpd_poweroff_unused(void)
 {
-	return NOTIFY_DONE;
+	pm_genpd_poweroff_unused();
+	return 0;
 }
-
-static inline void genpd_power_off_work_fn(struct work_struct *work) {}
-
-#define pm_genpd_runtime_suspend	NULL
-#define pm_genpd_runtime_resume		NULL
-
-#endif /* CONFIG_PM_RUNTIME */
+late_initcall(genpd_poweroff_unused);
 
 #ifdef CONFIG_PM_SLEEP
 
@@ -754,9 +782,9 @@ static inline void genpd_power_off_work_fn(struct work_struct *work) {}
  * pm_genpd_present - Check if the given PM domain has been initialized.
  * @genpd: PM domain to check.
  */
-static bool pm_genpd_present(struct generic_pm_domain *genpd)
+static bool pm_genpd_present(const struct generic_pm_domain *genpd)
 {
-	struct generic_pm_domain *gpd;
+	const struct generic_pm_domain *gpd;
 
 	if (IS_ERR_OR_NULL(genpd))
 		return false;
@@ -774,46 +802,6 @@ static bool genpd_dev_active_wakeup(struct generic_pm_domain *genpd,
 	return GENPD_DEV_CALLBACK(genpd, bool, active_wakeup, dev);
 }
 
-static int genpd_suspend_dev(struct generic_pm_domain *genpd, struct device *dev)
-{
-	return GENPD_DEV_CALLBACK(genpd, int, suspend, dev);
-}
-
-static int genpd_suspend_late(struct generic_pm_domain *genpd, struct device *dev)
-{
-	return GENPD_DEV_CALLBACK(genpd, int, suspend_late, dev);
-}
-
-static int genpd_resume_early(struct generic_pm_domain *genpd, struct device *dev)
-{
-	return GENPD_DEV_CALLBACK(genpd, int, resume_early, dev);
-}
-
-static int genpd_resume_dev(struct generic_pm_domain *genpd, struct device *dev)
-{
-	return GENPD_DEV_CALLBACK(genpd, int, resume, dev);
-}
-
-static int genpd_freeze_dev(struct generic_pm_domain *genpd, struct device *dev)
-{
-	return GENPD_DEV_CALLBACK(genpd, int, freeze, dev);
-}
-
-static int genpd_freeze_late(struct generic_pm_domain *genpd, struct device *dev)
-{
-	return GENPD_DEV_CALLBACK(genpd, int, freeze_late, dev);
-}
-
-static int genpd_thaw_early(struct generic_pm_domain *genpd, struct device *dev)
-{
-	return GENPD_DEV_CALLBACK(genpd, int, thaw_early, dev);
-}
-
-static int genpd_thaw_dev(struct generic_pm_domain *genpd, struct device *dev)
-{
-	return GENPD_DEV_CALLBACK(genpd, int, thaw, dev);
-}
-
 /**
  * pm_genpd_sync_poweroff - Synchronously power off a PM domain and its masters.
  * @genpd: PM domain to power off, if possible.
@@ -837,8 +825,7 @@ static void pm_genpd_sync_poweroff(struct generic_pm_domain *genpd)
 	    || atomic_read(&genpd->sd_count) > 0)
 		return;
 
-	if (genpd->power_off)
-		genpd->power_off(genpd);
+	genpd_power_off(genpd);
 
 	genpd->status = GPD_STATE_POWER_OFF;
 
@@ -869,8 +856,7 @@ static void pm_genpd_sync_poweron(struct generic_pm_domain *genpd)
 		genpd_sd_counter_inc(link->master);
 	}
 
-	if (genpd->power_on)
-		genpd->power_on(genpd);
+	genpd_power_on(genpd);
 
 	genpd->status = GPD_STATE_ACTIVE;
 }
@@ -995,7 +981,7 @@ static int pm_genpd_suspend(struct device *dev)
 	if (IS_ERR(genpd))
 		return -EINVAL;
 
-	return genpd->suspend_power_off ? 0 : genpd_suspend_dev(genpd, dev);
+	return genpd->suspend_power_off ? 0 : pm_generic_suspend(dev);
 }
 
 /**
@@ -1016,7 +1002,7 @@ static int pm_genpd_suspend_late(struct device *dev)
 	if (IS_ERR(genpd))
 		return -EINVAL;
 
-	return genpd->suspend_power_off ? 0 : genpd_suspend_late(genpd, dev);
+	return genpd->suspend_power_off ? 0 : pm_generic_suspend_late(dev);
 }
 
 /**
@@ -1103,7 +1089,7 @@ static int pm_genpd_resume_early(struct device *dev)
 	if (IS_ERR(genpd))
 		return -EINVAL;
 
-	return genpd->suspend_power_off ? 0 : genpd_resume_early(genpd, dev);
+	return genpd->suspend_power_off ? 0 : pm_generic_resume_early(dev);
 }
 
 /**
@@ -1124,7 +1110,7 @@ static int pm_genpd_resume(struct device *dev)
 	if (IS_ERR(genpd))
 		return -EINVAL;
 
-	return genpd->suspend_power_off ? 0 : genpd_resume_dev(genpd, dev);
+	return genpd->suspend_power_off ? 0 : pm_generic_resume(dev);
 }
 
 /**
@@ -1145,7 +1131,7 @@ static int pm_genpd_freeze(struct device *dev)
 	if (IS_ERR(genpd))
 		return -EINVAL;
 
-	return genpd->suspend_power_off ? 0 : genpd_freeze_dev(genpd, dev);
+	return genpd->suspend_power_off ? 0 : pm_generic_freeze(dev);
 }
 
 /**
@@ -1167,7 +1153,7 @@ static int pm_genpd_freeze_late(struct device *dev)
 	if (IS_ERR(genpd))
 		return -EINVAL;
 
-	return genpd->suspend_power_off ? 0 : genpd_freeze_late(genpd, dev);
+	return genpd->suspend_power_off ? 0 : pm_generic_freeze_late(dev);
 }
 
 /**
@@ -1231,7 +1217,7 @@ static int pm_genpd_thaw_early(struct device *dev)
 	if (IS_ERR(genpd))
 		return -EINVAL;
 
-	return genpd->suspend_power_off ? 0 : genpd_thaw_early(genpd, dev);
+	return genpd->suspend_power_off ? 0 : pm_generic_thaw_early(dev);
 }
 
 /**
@@ -1252,7 +1238,7 @@ static int pm_genpd_thaw(struct device *dev)
 	if (IS_ERR(genpd))
 		return -EINVAL;
 
-	return genpd->suspend_power_off ? 0 : genpd_thaw_dev(genpd, dev);
+	return genpd->suspend_power_off ? 0 : pm_generic_thaw(dev);
 }
 
 /**
@@ -1292,8 +1278,7 @@ static int pm_genpd_restore_noirq(struct device *dev)
 			 * If the domain was off before the hibernation, make
 			 * sure it will be off going forward.
 			 */
-			if (genpd->power_off)
-				genpd->power_off(genpd);
+			genpd_power_off(genpd);
 
 			return 0;
 		}
@@ -1344,13 +1329,13 @@ static void pm_genpd_complete(struct device *dev)
 }
 
 /**
- * pm_genpd_syscore_switch - Switch power during system core suspend or resume.
+ * genpd_syscore_switch - Switch power during system core suspend or resume.
  * @dev: Device that normally is marked as "always on" to switch power for.
  *
  * This routine may only be called during the system core (syscore) suspend or
  * resume phase for devices whose "always on" flags are set.
  */
-void pm_genpd_syscore_switch(struct device *dev, bool suspend)
+static void genpd_syscore_switch(struct device *dev, bool suspend)
 {
 	struct generic_pm_domain *genpd;
 
@@ -1366,9 +1351,20 @@ void pm_genpd_syscore_switch(struct device *dev, bool suspend)
 		genpd->suspended_count--;
 	}
 }
-EXPORT_SYMBOL_GPL(pm_genpd_syscore_switch);
 
-#else
+void pm_genpd_syscore_poweroff(struct device *dev)
+{
+	genpd_syscore_switch(dev, true);
+}
+EXPORT_SYMBOL_GPL(pm_genpd_syscore_poweroff);
+
+void pm_genpd_syscore_poweron(struct device *dev)
+{
+	genpd_syscore_switch(dev, false);
+}
+EXPORT_SYMBOL_GPL(pm_genpd_syscore_poweron);
+
+#else /* !CONFIG_PM_SLEEP */
 
 #define pm_genpd_prepare		NULL
 #define pm_genpd_suspend		NULL
@@ -1466,10 +1462,13 @@ int __pm_genpd_add_device(struct generic_pm_domain *genpd, struct device *dev,
 
 	spin_unlock_irq(&dev->power.lock);
 
+	if (genpd->attach_dev)
+		genpd->attach_dev(genpd, dev);
+
 	mutex_lock(&gpd_data->lock);
 	gpd_data->base.dev = dev;
 	list_add_tail(&gpd_data->base.list_node, &genpd->dev_list);
-	gpd_data->need_restore = genpd->status == GPD_STATE_POWER_OFF;
+	gpd_data->need_restore = -1;
 	gpd_data->td.constraint_changed = true;
 	gpd_data->td.effective_constraint_ns = -1;
 	mutex_unlock(&gpd_data->lock);
@@ -1484,39 +1483,6 @@ int __pm_genpd_add_device(struct generic_pm_domain *genpd, struct device *dev,
 }
 
 /**
- * __pm_genpd_of_add_device - Add a device to an I/O PM domain.
- * @genpd_node: Device tree node pointer representing a PM domain to which the
- *   the device is added to.
- * @dev: Device to be added.
- * @td: Set of PM QoS timing parameters to attach to the device.
- */
-int __pm_genpd_of_add_device(struct device_node *genpd_node, struct device *dev,
-			     struct gpd_timing_data *td)
-{
-	struct generic_pm_domain *genpd = NULL, *gpd;
-
-	dev_dbg(dev, "%s()\n", __func__);
-
-	if (IS_ERR_OR_NULL(genpd_node) || IS_ERR_OR_NULL(dev))
-		return -EINVAL;
-
-	mutex_lock(&gpd_list_lock);
-	list_for_each_entry(gpd, &gpd_list, gpd_list_node) {
-		if (gpd->of_node == genpd_node) {
-			genpd = gpd;
-			break;
-		}
-	}
-	mutex_unlock(&gpd_list_lock);
-
-	if (!genpd)
-		return -EINVAL;
-
-	return __pm_genpd_add_device(genpd, dev, td);
-}
-
-
-/**
  * __pm_genpd_name_add_device - Find I/O PM domain and add a device to it.
  * @domain_name: Name of the PM domain to add the device to.
  * @dev: Device to be added.
@@ -1558,6 +1524,9 @@ int pm_genpd_remove_device(struct generic_pm_domain *genpd,
 	genpd->device_count--;
 	genpd->max_off_time_changed = true;
 
+	if (genpd->detach_dev)
+		genpd->detach_dev(genpd, dev);
+
 	spin_lock_irq(&dev->power.lock);
 
 	dev->pm_domain = NULL;
@@ -1603,7 +1572,7 @@ void pm_genpd_dev_need_restore(struct device *dev, bool val)
 
 	psd = dev_to_psd(dev);
 	if (psd && psd->domain_data)
-		to_gpd_data(psd->domain_data)->need_restore = val;
+		to_gpd_data(psd->domain_data)->need_restore = val ? 1 : 0;
 
 	spin_unlock_irqrestore(&dev->power.lock, flags);
 }
@@ -1744,112 +1713,6 @@ int pm_genpd_remove_subdomain(struct generic_pm_domain *genpd,
 }
 
 /**
- * pm_genpd_add_callbacks - Add PM domain callbacks to a given device.
- * @dev: Device to add the callbacks to.
- * @ops: Set of callbacks to add.
- * @td: Timing data to add to the device along with the callbacks (optional).
- *
- * Every call to this routine should be balanced with a call to
- * __pm_genpd_remove_callbacks() and they must not be nested.
- */
-int pm_genpd_add_callbacks(struct device *dev, struct gpd_dev_ops *ops,
-			   struct gpd_timing_data *td)
-{
-	struct generic_pm_domain_data *gpd_data_new, *gpd_data = NULL;
-	int ret = 0;
-
-	if (!(dev && ops))
-		return -EINVAL;
-
-	gpd_data_new = __pm_genpd_alloc_dev_data(dev);
-	if (!gpd_data_new)
-		return -ENOMEM;
-
-	pm_runtime_disable(dev);
-	device_pm_lock();
-
-	ret = dev_pm_get_subsys_data(dev);
-	if (ret)
-		goto out;
-
-	spin_lock_irq(&dev->power.lock);
-
-	if (dev->power.subsys_data->domain_data) {
-		gpd_data = to_gpd_data(dev->power.subsys_data->domain_data);
-	} else {
-		gpd_data = gpd_data_new;
-		dev->power.subsys_data->domain_data = &gpd_data->base;
-	}
-	gpd_data->refcount++;
-	gpd_data->ops = *ops;
-	if (td)
-		gpd_data->td = *td;
-
-	spin_unlock_irq(&dev->power.lock);
-
- out:
-	device_pm_unlock();
-	pm_runtime_enable(dev);
-
-	if (gpd_data != gpd_data_new)
-		__pm_genpd_free_dev_data(dev, gpd_data_new);
-
-	return ret;
-}
-EXPORT_SYMBOL_GPL(pm_genpd_add_callbacks);
-
-/**
- * __pm_genpd_remove_callbacks - Remove PM domain callbacks from a given device.
- * @dev: Device to remove the callbacks from.
- * @clear_td: If set, clear the device's timing data too.
- *
- * This routine can only be called after pm_genpd_add_callbacks().
- */
-int __pm_genpd_remove_callbacks(struct device *dev, bool clear_td)
-{
-	struct generic_pm_domain_data *gpd_data = NULL;
-	bool remove = false;
-	int ret = 0;
-
-	if (!(dev && dev->power.subsys_data))
-		return -EINVAL;
-
-	pm_runtime_disable(dev);
-	device_pm_lock();
-
-	spin_lock_irq(&dev->power.lock);
-
-	if (dev->power.subsys_data->domain_data) {
-		gpd_data = to_gpd_data(dev->power.subsys_data->domain_data);
-		gpd_data->ops = (struct gpd_dev_ops){ NULL };
-		if (clear_td)
-			gpd_data->td = (struct gpd_timing_data){ 0 };
-
-		if (--gpd_data->refcount == 0) {
-			dev->power.subsys_data->domain_data = NULL;
-			remove = true;
-		}
-	} else {
-		ret = -EINVAL;
-	}
-
-	spin_unlock_irq(&dev->power.lock);
-
-	device_pm_unlock();
-	pm_runtime_enable(dev);
-
-	if (ret)
-		return ret;
-
-	dev_pm_put_subsys_data(dev);
-	if (remove)
-		__pm_genpd_free_dev_data(dev, gpd_data);
-
-	return 0;
-}
-EXPORT_SYMBOL_GPL(__pm_genpd_remove_callbacks);
-
-/**
  * pm_genpd_attach_cpuidle - Connect the given PM domain with cpuidle.
  * @genpd: PM domain to be connected with cpuidle.
  * @state: cpuidle state this domain can disable/enable.
@@ -1861,7 +1724,7 @@ EXPORT_SYMBOL_GPL(__pm_genpd_remove_callbacks);
 int pm_genpd_attach_cpuidle(struct generic_pm_domain *genpd, int state)
 {
 	struct cpuidle_driver *cpuidle_drv;
-	struct gpd_cpu_data *cpu_data;
+	struct gpd_cpuidle_data *cpuidle_data;
 	struct cpuidle_state *idle_state;
 	int ret = 0;
 
@@ -1870,12 +1733,12 @@ int pm_genpd_attach_cpuidle(struct generic_pm_domain *genpd, int state)
 
 	genpd_acquire_lock(genpd);
 
-	if (genpd->cpu_data) {
+	if (genpd->cpuidle_data) {
 		ret = -EEXIST;
 		goto out;
 	}
-	cpu_data = kzalloc(sizeof(*cpu_data), GFP_KERNEL);
-	if (!cpu_data) {
+	cpuidle_data = kzalloc(sizeof(*cpuidle_data), GFP_KERNEL);
+	if (!cpuidle_data) {
 		ret = -ENOMEM;
 		goto out;
 	}
@@ -1893,9 +1756,9 @@ int pm_genpd_attach_cpuidle(struct generic_pm_domain *genpd, int state)
 		ret = -EAGAIN;
 		goto err;
 	}
-	cpu_data->idle_state = idle_state;
-	cpu_data->saved_exit_latency = idle_state->exit_latency;
-	genpd->cpu_data = cpu_data;
+	cpuidle_data->idle_state = idle_state;
+	cpuidle_data->saved_exit_latency = idle_state->exit_latency;
+	genpd->cpuidle_data = cpuidle_data;
 	genpd_recalc_cpu_exit_latency(genpd);
 
  out:
@@ -1906,7 +1769,7 @@ int pm_genpd_attach_cpuidle(struct generic_pm_domain *genpd, int state)
 	cpuidle_driver_unref();
 
  err_drv:
-	kfree(cpu_data);
+	kfree(cpuidle_data);
 	goto out;
 }
 
@@ -1929,7 +1792,7 @@ int pm_genpd_name_attach_cpuidle(const char *name, int state)
  */
 int pm_genpd_detach_cpuidle(struct generic_pm_domain *genpd)
 {
-	struct gpd_cpu_data *cpu_data;
+	struct gpd_cpuidle_data *cpuidle_data;
 	struct cpuidle_state *idle_state;
 	int ret = 0;
 
@@ -1938,20 +1801,20 @@ int pm_genpd_detach_cpuidle(struct generic_pm_domain *genpd)
 
 	genpd_acquire_lock(genpd);
 
-	cpu_data = genpd->cpu_data;
-	if (!cpu_data) {
+	cpuidle_data = genpd->cpuidle_data;
+	if (!cpuidle_data) {
 		ret = -ENODEV;
 		goto out;
 	}
-	idle_state = cpu_data->idle_state;
+	idle_state = cpuidle_data->idle_state;
 	if (!idle_state->disabled) {
 		ret = -EAGAIN;
 		goto out;
 	}
-	idle_state->exit_latency = cpu_data->saved_exit_latency;
+	idle_state->exit_latency = cpuidle_data->saved_exit_latency;
 	cpuidle_driver_unref();
-	genpd->cpu_data = NULL;
-	kfree(cpu_data);
+	genpd->cpuidle_data = NULL;
+	kfree(cpuidle_data);
 
  out:
 	genpd_release_lock(genpd);
@@ -1970,17 +1833,13 @@ int pm_genpd_name_detach_cpuidle(const char *name)
 /* Default device callbacks for generic PM domains. */
 
 /**
- * pm_genpd_default_save_state - Default "save device state" for PM domians.
+ * pm_genpd_default_save_state - Default "save device state" for PM domains.
  * @dev: Device to handle.
  */
 static int pm_genpd_default_save_state(struct device *dev)
 {
 	int (*cb)(struct device *__dev);
 
-	cb = dev_gpd_data(dev)->ops.save_state;
-	if (cb)
-		return cb(dev);
-
 	if (dev->type && dev->type->pm)
 		cb = dev->type->pm->runtime_suspend;
 	else if (dev->class && dev->class->pm)
@@ -1997,17 +1856,13 @@ static int pm_genpd_default_save_state(struct device *dev)
 }
 
 /**
- * pm_genpd_default_restore_state - Default PM domians "restore device state".
+ * pm_genpd_default_restore_state - Default PM domains "restore device state".
  * @dev: Device to handle.
  */
 static int pm_genpd_default_restore_state(struct device *dev)
 {
 	int (*cb)(struct device *__dev);
 
-	cb = dev_gpd_data(dev)->ops.restore_state;
-	if (cb)
-		return cb(dev);
-
 	if (dev->type && dev->type->pm)
 		cb = dev->type->pm->runtime_resume;
 	else if (dev->class && dev->class->pm)
@@ -2023,109 +1878,6 @@ static int pm_genpd_default_restore_state(struct device *dev)
 	return cb ? cb(dev) : 0;
 }
 
-#ifdef CONFIG_PM_SLEEP
-
-/**
- * pm_genpd_default_suspend - Default "device suspend" for PM domians.
- * @dev: Device to handle.
- */
-static int pm_genpd_default_suspend(struct device *dev)
-{
-	int (*cb)(struct device *__dev) = dev_gpd_data(dev)->ops.suspend;
-
-	return cb ? cb(dev) : pm_generic_suspend(dev);
-}
-
-/**
- * pm_genpd_default_suspend_late - Default "late device suspend" for PM domians.
- * @dev: Device to handle.
- */
-static int pm_genpd_default_suspend_late(struct device *dev)
-{
-	int (*cb)(struct device *__dev) = dev_gpd_data(dev)->ops.suspend_late;
-
-	return cb ? cb(dev) : pm_generic_suspend_late(dev);
-}
-
-/**
- * pm_genpd_default_resume_early - Default "early device resume" for PM domians.
- * @dev: Device to handle.
- */
-static int pm_genpd_default_resume_early(struct device *dev)
-{
-	int (*cb)(struct device *__dev) = dev_gpd_data(dev)->ops.resume_early;
-
-	return cb ? cb(dev) : pm_generic_resume_early(dev);
-}
-
-/**
- * pm_genpd_default_resume - Default "device resume" for PM domians.
- * @dev: Device to handle.
- */
-static int pm_genpd_default_resume(struct device *dev)
-{
-	int (*cb)(struct device *__dev) = dev_gpd_data(dev)->ops.resume;
-
-	return cb ? cb(dev) : pm_generic_resume(dev);
-}
-
-/**
- * pm_genpd_default_freeze - Default "device freeze" for PM domians.
- * @dev: Device to handle.
- */
-static int pm_genpd_default_freeze(struct device *dev)
-{
-	int (*cb)(struct device *__dev) = dev_gpd_data(dev)->ops.freeze;
-
-	return cb ? cb(dev) : pm_generic_freeze(dev);
-}
-
-/**
- * pm_genpd_default_freeze_late - Default "late device freeze" for PM domians.
- * @dev: Device to handle.
- */
-static int pm_genpd_default_freeze_late(struct device *dev)
-{
-	int (*cb)(struct device *__dev) = dev_gpd_data(dev)->ops.freeze_late;
-
-	return cb ? cb(dev) : pm_generic_freeze_late(dev);
-}
-
-/**
- * pm_genpd_default_thaw_early - Default "early device thaw" for PM domians.
- * @dev: Device to handle.
- */
-static int pm_genpd_default_thaw_early(struct device *dev)
-{
-	int (*cb)(struct device *__dev) = dev_gpd_data(dev)->ops.thaw_early;
-
-	return cb ? cb(dev) : pm_generic_thaw_early(dev);
-}
-
-/**
- * pm_genpd_default_thaw - Default "device thaw" for PM domians.
- * @dev: Device to handle.
- */
-static int pm_genpd_default_thaw(struct device *dev)
-{
-	int (*cb)(struct device *__dev) = dev_gpd_data(dev)->ops.thaw;
-
-	return cb ? cb(dev) : pm_generic_thaw(dev);
-}
-
-#else /* !CONFIG_PM_SLEEP */
-
-#define pm_genpd_default_suspend	NULL
-#define pm_genpd_default_suspend_late	NULL
-#define pm_genpd_default_resume_early	NULL
-#define pm_genpd_default_resume		NULL
-#define pm_genpd_default_freeze		NULL
-#define pm_genpd_default_freeze_late	NULL
-#define pm_genpd_default_thaw_early	NULL
-#define pm_genpd_default_thaw		NULL
-
-#endif /* !CONFIG_PM_SLEEP */
-
 /**
  * pm_genpd_init - Initialize a generic I/O PM domain object.
  * @genpd: PM domain object to initialize.
@@ -2177,15 +1929,452 @@ void pm_genpd_init(struct generic_pm_domain *genpd,
 	genpd->domain.ops.complete = pm_genpd_complete;
 	genpd->dev_ops.save_state = pm_genpd_default_save_state;
 	genpd->dev_ops.restore_state = pm_genpd_default_restore_state;
-	genpd->dev_ops.suspend = pm_genpd_default_suspend;
-	genpd->dev_ops.suspend_late = pm_genpd_default_suspend_late;
-	genpd->dev_ops.resume_early = pm_genpd_default_resume_early;
-	genpd->dev_ops.resume = pm_genpd_default_resume;
-	genpd->dev_ops.freeze = pm_genpd_default_freeze;
-	genpd->dev_ops.freeze_late = pm_genpd_default_freeze_late;
-	genpd->dev_ops.thaw_early = pm_genpd_default_thaw_early;
-	genpd->dev_ops.thaw = pm_genpd_default_thaw;
+
+	if (genpd->flags & GENPD_FLAG_PM_CLK) {
+		genpd->dev_ops.stop = pm_clk_suspend;
+		genpd->dev_ops.start = pm_clk_resume;
+	}
+
 	mutex_lock(&gpd_list_lock);
 	list_add(&genpd->gpd_list_node, &gpd_list);
 	mutex_unlock(&gpd_list_lock);
 }
+
+#ifdef CONFIG_PM_GENERIC_DOMAINS_OF
+/*
+ * Device Tree based PM domain providers.
+ *
+ * The code below implements generic device tree based PM domain providers that
+ * bind device tree nodes with generic PM domains registered in the system.
+ *
+ * Any driver that registers generic PM domains and needs to support binding of
+ * devices to these domains is supposed to register a PM domain provider, which
+ * maps a PM domain specifier retrieved from the device tree to a PM domain.
+ *
+ * Two simple mapping functions have been provided for convenience:
+ *  - __of_genpd_xlate_simple() for 1:1 device tree node to PM domain mapping.
+ *  - __of_genpd_xlate_onecell() for mapping of multiple PM domains per node by
+ *    index.
+ */
+
+/**
+ * struct of_genpd_provider - PM domain provider registration structure
+ * @link: Entry in global list of PM domain providers
+ * @node: Pointer to device tree node of PM domain provider
+ * @xlate: Provider-specific xlate callback mapping a set of specifier cells
+ *         into a PM domain.
+ * @data: context pointer to be passed into @xlate callback
+ */
+struct of_genpd_provider {
+	struct list_head link;
+	struct device_node *node;
+	genpd_xlate_t xlate;
+	void *data;
+};
+
+/* List of registered PM domain providers. */
+static LIST_HEAD(of_genpd_providers);
+/* Mutex to protect the list above. */
+static DEFINE_MUTEX(of_genpd_mutex);
+
+/**
+ * __of_genpd_xlate_simple() - Xlate function for direct node-domain mapping
+ * @genpdspec: OF phandle args to map into a PM domain
+ * @data: xlate function private data - pointer to struct generic_pm_domain
+ *
+ * This is a generic xlate function that can be used to model PM domains that
+ * have their own device tree nodes. The private data of xlate function needs
+ * to be a valid pointer to struct generic_pm_domain.
+ */
+struct generic_pm_domain *__of_genpd_xlate_simple(
+					struct of_phandle_args *genpdspec,
+					void *data)
+{
+	if (genpdspec->args_count != 0)
+		return ERR_PTR(-EINVAL);
+	return data;
+}
+EXPORT_SYMBOL_GPL(__of_genpd_xlate_simple);
+
+/**
+ * __of_genpd_xlate_onecell() - Xlate function using a single index.
+ * @genpdspec: OF phandle args to map into a PM domain
+ * @data: xlate function private data - pointer to struct genpd_onecell_data
+ *
+ * This is a generic xlate function that can be used to model simple PM domain
+ * controllers that have one device tree node and provide multiple PM domains.
+ * A single cell is used as an index into an array of PM domains specified in
+ * the genpd_onecell_data struct when registering the provider.
+ */
+struct generic_pm_domain *__of_genpd_xlate_onecell(
+					struct of_phandle_args *genpdspec,
+					void *data)
+{
+	struct genpd_onecell_data *genpd_data = data;
+	unsigned int idx = genpdspec->args[0];
+
+	if (genpdspec->args_count != 1)
+		return ERR_PTR(-EINVAL);
+
+	if (idx >= genpd_data->num_domains) {
+		pr_err("%s: invalid domain index %u\n", __func__, idx);
+		return ERR_PTR(-EINVAL);
+	}
+
+	if (!genpd_data->domains[idx])
+		return ERR_PTR(-ENOENT);
+
+	return genpd_data->domains[idx];
+}
+EXPORT_SYMBOL_GPL(__of_genpd_xlate_onecell);
+
+/**
+ * __of_genpd_add_provider() - Register a PM domain provider for a node
+ * @np: Device node pointer associated with the PM domain provider.
+ * @xlate: Callback for decoding PM domain from phandle arguments.
+ * @data: Context pointer for @xlate callback.
+ */
+int __of_genpd_add_provider(struct device_node *np, genpd_xlate_t xlate,
+			void *data)
+{
+	struct of_genpd_provider *cp;
+
+	cp = kzalloc(sizeof(*cp), GFP_KERNEL);
+	if (!cp)
+		return -ENOMEM;
+
+	cp->node = of_node_get(np);
+	cp->data = data;
+	cp->xlate = xlate;
+
+	mutex_lock(&of_genpd_mutex);
+	list_add(&cp->link, &of_genpd_providers);
+	mutex_unlock(&of_genpd_mutex);
+	pr_debug("Added domain provider from %s\n", np->full_name);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(__of_genpd_add_provider);
+
+/**
+ * of_genpd_del_provider() - Remove a previously registered PM domain provider
+ * @np: Device node pointer associated with the PM domain provider
+ */
+void of_genpd_del_provider(struct device_node *np)
+{
+	struct of_genpd_provider *cp;
+
+	mutex_lock(&of_genpd_mutex);
+	list_for_each_entry(cp, &of_genpd_providers, link) {
+		if (cp->node == np) {
+			list_del(&cp->link);
+			of_node_put(cp->node);
+			kfree(cp);
+			break;
+		}
+	}
+	mutex_unlock(&of_genpd_mutex);
+}
+EXPORT_SYMBOL_GPL(of_genpd_del_provider);
+
+/**
+ * of_genpd_get_from_provider() - Look-up PM domain
+ * @genpdspec: OF phandle args to use for look-up
+ *
+ * Looks for a PM domain provider under the node specified by @genpdspec and if
+ * found, uses xlate function of the provider to map phandle args to a PM
+ * domain.
+ *
+ * Returns a valid pointer to struct generic_pm_domain on success or ERR_PTR()
+ * on failure.
+ */
+struct generic_pm_domain *of_genpd_get_from_provider(
+					struct of_phandle_args *genpdspec)
+{
+	struct generic_pm_domain *genpd = ERR_PTR(-ENOENT);
+	struct of_genpd_provider *provider;
+
+	mutex_lock(&of_genpd_mutex);
+
+	/* Check if we have such a provider in our array */
+	list_for_each_entry(provider, &of_genpd_providers, link) {
+		if (provider->node == genpdspec->np)
+			genpd = provider->xlate(genpdspec, provider->data);
+		if (!IS_ERR(genpd))
+			break;
+	}
+
+	mutex_unlock(&of_genpd_mutex);
+
+	return genpd;
+}
+EXPORT_SYMBOL_GPL(of_genpd_get_from_provider);
+
+/**
+ * genpd_dev_pm_detach - Detach a device from its PM domain.
+ * @dev: Device to attach.
+ * @power_off: Currently not used
+ *
+ * Try to locate a corresponding generic PM domain, which the device was
+ * attached to previously. If such is found, the device is detached from it.
+ */
+static void genpd_dev_pm_detach(struct device *dev, bool power_off)
+{
+	struct generic_pm_domain *pd = NULL, *gpd;
+	int ret = 0;
+
+	if (!dev->pm_domain)
+		return;
+
+	mutex_lock(&gpd_list_lock);
+	list_for_each_entry(gpd, &gpd_list, gpd_list_node) {
+		if (&gpd->domain == dev->pm_domain) {
+			pd = gpd;
+			break;
+		}
+	}
+	mutex_unlock(&gpd_list_lock);
+
+	if (!pd)
+		return;
+
+	dev_dbg(dev, "removing from PM domain %s\n", pd->name);
+
+	while (1) {
+		ret = pm_genpd_remove_device(pd, dev);
+		if (ret != -EAGAIN)
+			break;
+		cond_resched();
+	}
+
+	if (ret < 0) {
+		dev_err(dev, "failed to remove from PM domain %s: %d",
+			pd->name, ret);
+		return;
+	}
+
+	/* Check if PM domain can be powered off after removing this device. */
+	genpd_queue_power_off_work(pd);
+}
+
+/**
+ * genpd_dev_pm_attach - Attach a device to its PM domain using DT.
+ * @dev: Device to attach.
+ *
+ * Parse device's OF node to find a PM domain specifier. If such is found,
+ * attaches the device to retrieved pm_domain ops.
+ *
+ * Both generic and legacy Samsung-specific DT bindings are supported to keep
+ * backwards compatibility with existing DTBs.
+ *
+ * Returns 0 on successfully attached PM domain or negative error code.
+ */
+int genpd_dev_pm_attach(struct device *dev)
+{
+	struct of_phandle_args pd_args;
+	struct generic_pm_domain *pd;
+	int ret;
+
+	if (!dev->of_node)
+		return -ENODEV;
+
+	if (dev->pm_domain)
+		return -EEXIST;
+
+	ret = of_parse_phandle_with_args(dev->of_node, "power-domains",
+					"#power-domain-cells", 0, &pd_args);
+	if (ret < 0) {
+		if (ret != -ENOENT)
+			return ret;
+
+		/*
+		 * Try legacy Samsung-specific bindings
+		 * (for backwards compatibility of DT ABI)
+		 */
+		pd_args.args_count = 0;
+		pd_args.np = of_parse_phandle(dev->of_node,
+						"samsung,power-domain", 0);
+		if (!pd_args.np)
+			return -ENOENT;
+	}
+
+	pd = of_genpd_get_from_provider(&pd_args);
+	if (IS_ERR(pd)) {
+		dev_dbg(dev, "%s() failed to find PM domain: %ld\n",
+			__func__, PTR_ERR(pd));
+		of_node_put(dev->of_node);
+		return PTR_ERR(pd);
+	}
+
+	dev_dbg(dev, "adding to PM domain %s\n", pd->name);
+
+	while (1) {
+		ret = pm_genpd_add_device(pd, dev);
+		if (ret != -EAGAIN)
+			break;
+		cond_resched();
+	}
+
+	if (ret < 0) {
+		dev_err(dev, "failed to add to PM domain %s: %d",
+			pd->name, ret);
+		of_node_put(dev->of_node);
+		return ret;
+	}
+
+	dev->pm_domain->detach = genpd_dev_pm_detach;
+	pm_genpd_poweron(pd);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(genpd_dev_pm_attach);
+#endif /* CONFIG_PM_GENERIC_DOMAINS_OF */
+
+
+/***        debugfs support        ***/
+
+#ifdef CONFIG_PM_ADVANCED_DEBUG
+#include <linux/pm.h>
+#include <linux/device.h>
+#include <linux/debugfs.h>
+#include <linux/seq_file.h>
+#include <linux/init.h>
+#include <linux/kobject.h>
+static struct dentry *pm_genpd_debugfs_dir;
+
+/*
+ * TODO: This function is a slightly modified version of rtpm_status_show
+ * from sysfs.c, so generalize it.
+ */
+static void rtpm_status_str(struct seq_file *s, struct device *dev)
+{
+	static const char * const status_lookup[] = {
+		[RPM_ACTIVE] = "active",
+		[RPM_RESUMING] = "resuming",
+		[RPM_SUSPENDED] = "suspended",
+		[RPM_SUSPENDING] = "suspending"
+	};
+	const char *p = "";
+
+	if (dev->power.runtime_error)
+		p = "error";
+	else if (dev->power.disable_depth)
+		p = "unsupported";
+	else if (dev->power.runtime_status < ARRAY_SIZE(status_lookup))
+		p = status_lookup[dev->power.runtime_status];
+	else
+		WARN_ON(1);
+
+	seq_puts(s, p);
+}
+
+static int pm_genpd_summary_one(struct seq_file *s,
+		struct generic_pm_domain *gpd)
+{
+	static const char * const status_lookup[] = {
+		[GPD_STATE_ACTIVE] = "on",
+		[GPD_STATE_WAIT_MASTER] = "wait-master",
+		[GPD_STATE_BUSY] = "busy",
+		[GPD_STATE_REPEAT] = "off-in-progress",
+		[GPD_STATE_POWER_OFF] = "off"
+	};
+	struct pm_domain_data *pm_data;
+	const char *kobj_path;
+	struct gpd_link *link;
+	int ret;
+
+	ret = mutex_lock_interruptible(&gpd->lock);
+	if (ret)
+		return -ERESTARTSYS;
+
+	if (WARN_ON(gpd->status >= ARRAY_SIZE(status_lookup)))
+		goto exit;
+	seq_printf(s, "%-30s  %-15s  ", gpd->name, status_lookup[gpd->status]);
+
+	/*
+	 * Modifications on the list require holding locks on both
+	 * master and slave, so we are safe.
+	 * Also gpd->name is immutable.
+	 */
+	list_for_each_entry(link, &gpd->master_links, master_node) {
+		seq_printf(s, "%s", link->slave->name);
+		if (!list_is_last(&link->master_node, &gpd->master_links))
+			seq_puts(s, ", ");
+	}
+
+	list_for_each_entry(pm_data, &gpd->dev_list, list_node) {
+		kobj_path = kobject_get_path(&pm_data->dev->kobj, GFP_KERNEL);
+		if (kobj_path == NULL)
+			continue;
+
+		seq_printf(s, "\n    %-50s  ", kobj_path);
+		rtpm_status_str(s, pm_data->dev);
+		kfree(kobj_path);
+	}
+
+	seq_puts(s, "\n");
+exit:
+	mutex_unlock(&gpd->lock);
+
+	return 0;
+}
+
+static int pm_genpd_summary_show(struct seq_file *s, void *data)
+{
+	struct generic_pm_domain *gpd;
+	int ret = 0;
+
+	seq_puts(s, "    domain                      status         slaves\n");
+	seq_puts(s, "           /device                                      runtime status\n");
+	seq_puts(s, "----------------------------------------------------------------------\n");
+
+	ret = mutex_lock_interruptible(&gpd_list_lock);
+	if (ret)
+		return -ERESTARTSYS;
+
+	list_for_each_entry(gpd, &gpd_list, gpd_list_node) {
+		ret = pm_genpd_summary_one(s, gpd);
+		if (ret)
+			break;
+	}
+	mutex_unlock(&gpd_list_lock);
+
+	return ret;
+}
+
+static int pm_genpd_summary_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, pm_genpd_summary_show, NULL);
+}
+
+static const struct file_operations pm_genpd_summary_fops = {
+	.open = pm_genpd_summary_open,
+	.read = seq_read,
+	.llseek = seq_lseek,
+	.release = single_release,
+};
+
+static int __init pm_genpd_debug_init(void)
+{
+	struct dentry *d;
+
+	pm_genpd_debugfs_dir = debugfs_create_dir("pm_genpd", NULL);
+
+	if (!pm_genpd_debugfs_dir)
+		return -ENOMEM;
+
+	d = debugfs_create_file("pm_genpd_summary", S_IRUGO,
+			pm_genpd_debugfs_dir, NULL, &pm_genpd_summary_fops);
+	if (!d)
+		return -ENOMEM;
+
+	return 0;
+}
+late_initcall(pm_genpd_debug_init);
+
+static void __exit pm_genpd_debug_exit(void)
+{
+	debugfs_remove_recursive(pm_genpd_debugfs_dir);
+}
+__exitcall(pm_genpd_debug_exit);
+#endif /* CONFIG_PM_ADVANCED_DEBUG */
diff --git a/drivers/base/power/domain_governor.c b/drivers/base/power/domain_governor.c
index a089e3bcdfbc..2a4154a09e4d 100644
--- a/drivers/base/power/domain_governor.c
+++ b/drivers/base/power/domain_governor.c
@@ -11,8 +11,6 @@
 #include <linux/pm_qos.h>
 #include <linux/hrtimer.h>
 
-#ifdef CONFIG_PM_RUNTIME
-
 static int dev_update_qos_constraint(struct device *dev, void *data)
 {
 	s64 *constraint_ns_p = data;
@@ -42,7 +40,7 @@ static int dev_update_qos_constraint(struct device *dev, void *data)
  * default_stop_ok - Default PM domain governor routine for stopping devices.
  * @dev: Device to check.
  */
-bool default_stop_ok(struct device *dev)
+static bool default_stop_ok(struct device *dev)
 {
 	struct gpd_timing_data *td = &dev_gpd_data(dev)->td;
 	unsigned long flags;
@@ -227,18 +225,6 @@ static bool always_on_power_down_ok(struct dev_pm_domain *domain)
 	return false;
 }
 
-#else /* !CONFIG_PM_RUNTIME */
-
-bool default_stop_ok(struct device *dev)
-{
-	return false;
-}
-
-#define default_power_down_ok	NULL
-#define always_on_power_down_ok	NULL
-
-#endif /* !CONFIG_PM_RUNTIME */
-
 struct dev_power_governor simple_qos_governor = {
 	.stop_ok = default_stop_ok,
 	.power_down_ok = default_power_down_ok,
diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c
index b67d9aef9fe4..9717d5f20139 100644
--- a/drivers/base/power/main.c
+++ b/drivers/base/power/main.c
@@ -540,7 +540,7 @@ static void async_resume_noirq(void *data, async_cookie_t cookie)
  * Call the "noirq" resume handlers for all devices in dpm_noirq_list and
  * enable device drivers to receive interrupts.
  */
-static void dpm_resume_noirq(pm_message_t state)
+void dpm_resume_noirq(pm_message_t state)
 {
 	struct device *dev;
 	ktime_t starttime = ktime_get();
@@ -662,7 +662,7 @@ static void async_resume_early(void *data, async_cookie_t cookie)
  * dpm_resume_early - Execute "early resume" callbacks for all devices.
  * @state: PM transition of the system being carried out.
  */
-static void dpm_resume_early(pm_message_t state)
+void dpm_resume_early(pm_message_t state)
 {
 	struct device *dev;
 	ktime_t starttime = ktime_get();
@@ -1093,7 +1093,7 @@ static int device_suspend_noirq(struct device *dev)
  * Prevent device drivers from receiving interrupts and call the "noirq" suspend
  * handlers for all non-sysdev devices.
  */
-static int dpm_suspend_noirq(pm_message_t state)
+int dpm_suspend_noirq(pm_message_t state)
 {
 	ktime_t starttime = ktime_get();
 	int error = 0;
@@ -1232,7 +1232,7 @@ static int device_suspend_late(struct device *dev)
  * dpm_suspend_late - Execute "late suspend" callbacks for all devices.
  * @state: PM transition of the system being carried out.
  */
-static int dpm_suspend_late(pm_message_t state)
+int dpm_suspend_late(pm_message_t state)
 {
 	ktime_t starttime = ktime_get();
 	int error = 0;
@@ -1266,6 +1266,8 @@ static int dpm_suspend_late(pm_message_t state)
 	}
 	mutex_unlock(&dpm_list_mtx);
 	async_synchronize_full();
+	if (!error)
+		error = async_error;
 	if (error) {
 		suspend_stats.failed_suspend_late++;
 		dpm_save_failed_step(SUSPEND_SUSPEND_LATE);
diff --git a/drivers/base/power/opp.c b/drivers/base/power/opp.c
index 89ced955fafa..106c69359306 100644
--- a/drivers/base/power/opp.c
+++ b/drivers/base/power/opp.c
@@ -49,11 +49,12 @@
  *		are protected by the dev_opp_list_lock for integrity.
  *		IMPORTANT: the opp nodes should be maintained in increasing
  *		order.
+ * @dynamic:	not-created from static DT entries.
  * @available:	true/false - marks if this OPP as available or not
  * @rate:	Frequency in hertz
  * @u_volt:	Nominal voltage in microvolts corresponding to this OPP
  * @dev_opp:	points back to the device_opp struct this opp belongs to
- * @head:	RCU callback head used for deferred freeing
+ * @rcu_head:	RCU callback head used for deferred freeing
  *
  * This structure stores the OPP information for a given device.
  */
@@ -61,11 +62,12 @@ struct dev_pm_opp {
 	struct list_head node;
 
 	bool available;
+	bool dynamic;
 	unsigned long rate;
 	unsigned long u_volt;
 
 	struct device_opp *dev_opp;
-	struct rcu_head head;
+	struct rcu_head rcu_head;
 };
 
 /**
@@ -76,18 +78,24 @@ struct dev_pm_opp {
  *		RCU usage: nodes are not modified in the list of device_opp,
  *		however addition is possible and is secured by dev_opp_list_lock
  * @dev:	device pointer
- * @head:	notifier head to notify the OPP availability changes.
+ * @srcu_head:	notifier head to notify the OPP availability changes.
+ * @rcu_head:	RCU callback head used for deferred freeing
  * @opp_list:	list of opps
  *
  * This is an internal data structure maintaining the link to opps attached to
  * a device. This structure is not meant to be shared to users as it is
- * meant for book keeping and private to OPP library
+ * meant for book keeping and private to OPP library.
+ *
+ * Because the opp structures can be used from both rcu and srcu readers, we
+ * need to wait for the grace period of both of them before freeing any
+ * resources. And so we have used kfree_rcu() from within call_srcu() handlers.
  */
 struct device_opp {
 	struct list_head node;
 
 	struct device *dev;
-	struct srcu_notifier_head head;
+	struct srcu_notifier_head srcu_head;
+	struct rcu_head rcu_head;
 	struct list_head opp_list;
 };
 
@@ -100,6 +108,14 @@ static LIST_HEAD(dev_opp_list);
 /* Lock to allow exclusive modification to the device and opp lists */
 static DEFINE_MUTEX(dev_opp_list_lock);
 
+#define opp_rcu_lockdep_assert()					\
+do {									\
+	rcu_lockdep_assert(rcu_read_lock_held() ||			\
+				lockdep_is_held(&dev_opp_list_lock),	\
+			   "Missing rcu_read_lock() or "		\
+			   "dev_opp_list_lock protection");		\
+} while (0)
+
 /**
  * find_device_opp() - find device_opp struct using device pointer
  * @dev:	device pointer used to lookup device OPPs
@@ -200,9 +216,7 @@ EXPORT_SYMBOL_GPL(dev_pm_opp_get_freq);
  * This function returns the number of available opps if there are any,
  * else returns 0 if none or the corresponding error value.
  *
- * Locking: This function must be called under rcu_read_lock(). This function
- * internally references two RCU protected structures: device_opp and opp which
- * are safe as long as we are under a common RCU locked section.
+ * Locking: This function takes rcu_read_lock().
  */
 int dev_pm_opp_get_opp_count(struct device *dev)
 {
@@ -210,11 +224,14 @@ int dev_pm_opp_get_opp_count(struct device *dev)
 	struct dev_pm_opp *temp_opp;
 	int count = 0;
 
+	rcu_read_lock();
+
 	dev_opp = find_device_opp(dev);
 	if (IS_ERR(dev_opp)) {
-		int r = PTR_ERR(dev_opp);
-		dev_err(dev, "%s: device OPP not found (%d)\n", __func__, r);
-		return r;
+		count = PTR_ERR(dev_opp);
+		dev_err(dev, "%s: device OPP not found (%d)\n",
+			__func__, count);
+		goto out_unlock;
 	}
 
 	list_for_each_entry_rcu(temp_opp, &dev_opp->opp_list, node) {
@@ -222,6 +239,8 @@ int dev_pm_opp_get_opp_count(struct device *dev)
 			count++;
 	}
 
+out_unlock:
+	rcu_read_unlock();
 	return count;
 }
 EXPORT_SYMBOL_GPL(dev_pm_opp_get_opp_count);
@@ -259,6 +278,8 @@ struct dev_pm_opp *dev_pm_opp_find_freq_exact(struct device *dev,
 	struct device_opp *dev_opp;
 	struct dev_pm_opp *temp_opp, *opp = ERR_PTR(-ERANGE);
 
+	opp_rcu_lockdep_assert();
+
 	dev_opp = find_device_opp(dev);
 	if (IS_ERR(dev_opp)) {
 		int r = PTR_ERR(dev_opp);
@@ -305,6 +326,8 @@ struct dev_pm_opp *dev_pm_opp_find_freq_ceil(struct device *dev,
 	struct device_opp *dev_opp;
 	struct dev_pm_opp *temp_opp, *opp = ERR_PTR(-ERANGE);
 
+	opp_rcu_lockdep_assert();
+
 	if (!dev || !freq) {
 		dev_err(dev, "%s: Invalid argument freq=%p\n", __func__, freq);
 		return ERR_PTR(-EINVAL);
@@ -353,6 +376,8 @@ struct dev_pm_opp *dev_pm_opp_find_freq_floor(struct device *dev,
 	struct device_opp *dev_opp;
 	struct dev_pm_opp *temp_opp, *opp = ERR_PTR(-ERANGE);
 
+	opp_rcu_lockdep_assert();
+
 	if (!dev || !freq) {
 		dev_err(dev, "%s: Invalid argument freq=%p\n", __func__, freq);
 		return ERR_PTR(-EINVAL);
@@ -378,34 +403,34 @@ struct dev_pm_opp *dev_pm_opp_find_freq_floor(struct device *dev,
 }
 EXPORT_SYMBOL_GPL(dev_pm_opp_find_freq_floor);
 
-/**
- * dev_pm_opp_add()  - Add an OPP table from a table definitions
- * @dev:	device for which we do this operation
- * @freq:	Frequency in Hz for this OPP
- * @u_volt:	Voltage in uVolts for this OPP
- *
- * This function adds an opp definition to the opp list and returns status.
- * The opp is made available by default and it can be controlled using
- * dev_pm_opp_enable/disable functions.
- *
- * Locking: The internal device_opp and opp structures are RCU protected.
- * Hence this function internally uses RCU updater strategy with mutex locks
- * to keep the integrity of the internal data structures. Callers should ensure
- * that this function is *NOT* called under RCU protection or in contexts where
- * mutex cannot be locked.
- *
- * Return:
- * 0:		On success OR
- *		Duplicate OPPs (both freq and volt are same) and opp->available
- * -EEXIST:	Freq are same and volt are different OR
- *		Duplicate OPPs (both freq and volt are same) and !opp->available
- * -ENOMEM:	Memory allocation failure
- */
-int dev_pm_opp_add(struct device *dev, unsigned long freq, unsigned long u_volt)
+static struct device_opp *add_device_opp(struct device *dev)
+{
+	struct device_opp *dev_opp;
+
+	/*
+	 * Allocate a new device OPP table. In the infrequent case where a new
+	 * device is needed to be added, we pay this penalty.
+	 */
+	dev_opp = kzalloc(sizeof(*dev_opp), GFP_KERNEL);
+	if (!dev_opp)
+		return NULL;
+
+	dev_opp->dev = dev;
+	srcu_init_notifier_head(&dev_opp->srcu_head);
+	INIT_LIST_HEAD(&dev_opp->opp_list);
+
+	/* Secure the device list modification */
+	list_add_rcu(&dev_opp->node, &dev_opp_list);
+	return dev_opp;
+}
+
+static int dev_pm_opp_add_dynamic(struct device *dev, unsigned long freq,
+				  unsigned long u_volt, bool dynamic)
 {
 	struct device_opp *dev_opp = NULL;
 	struct dev_pm_opp *opp, *new_opp;
 	struct list_head *head;
+	int ret;
 
 	/* allocate new OPP node */
 	new_opp = kzalloc(sizeof(*new_opp), GFP_KERNEL);
@@ -417,38 +442,25 @@ int dev_pm_opp_add(struct device *dev, unsigned long freq, unsigned long u_volt)
 	/* Hold our list modification lock here */
 	mutex_lock(&dev_opp_list_lock);
 
+	/* populate the opp table */
+	new_opp->rate = freq;
+	new_opp->u_volt = u_volt;
+	new_opp->available = true;
+	new_opp->dynamic = dynamic;
+
 	/* Check for existing list for 'dev' */
 	dev_opp = find_device_opp(dev);
 	if (IS_ERR(dev_opp)) {
-		/*
-		 * Allocate a new device OPP table. In the infrequent case
-		 * where a new device is needed to be added, we pay this
-		 * penalty.
-		 */
-		dev_opp = kzalloc(sizeof(struct device_opp), GFP_KERNEL);
+		dev_opp = add_device_opp(dev);
 		if (!dev_opp) {
-			mutex_unlock(&dev_opp_list_lock);
-			kfree(new_opp);
-			dev_warn(dev,
-				"%s: Unable to create device OPP structure\n",
-				__func__);
-			return -ENOMEM;
+			ret = -ENOMEM;
+			goto free_opp;
 		}
 
-		dev_opp->dev = dev;
-		srcu_init_notifier_head(&dev_opp->head);
-		INIT_LIST_HEAD(&dev_opp->opp_list);
-
-		/* Secure the device list modification */
-		list_add_rcu(&dev_opp->node, &dev_opp_list);
+		head = &dev_opp->opp_list;
+		goto list_add;
 	}
 
-	/* populate the opp table */
-	new_opp->dev_opp = dev_opp;
-	new_opp->rate = freq;
-	new_opp->u_volt = u_volt;
-	new_opp->available = true;
-
 	/*
 	 * Insert new OPP in order of increasing frequency
 	 * and discard if already present
@@ -463,17 +475,17 @@ int dev_pm_opp_add(struct device *dev, unsigned long freq, unsigned long u_volt)
 
 	/* Duplicate OPPs ? */
 	if (new_opp->rate == opp->rate) {
-		int ret = opp->available && new_opp->u_volt == opp->u_volt ?
+		ret = opp->available && new_opp->u_volt == opp->u_volt ?
 			0 : -EEXIST;
 
 		dev_warn(dev, "%s: duplicate OPPs detected. Existing: freq: %lu, volt: %lu, enabled: %d. New: freq: %lu, volt: %lu, enabled: %d\n",
 			 __func__, opp->rate, opp->u_volt, opp->available,
 			 new_opp->rate, new_opp->u_volt, new_opp->available);
-		mutex_unlock(&dev_opp_list_lock);
-		kfree(new_opp);
-		return ret;
+		goto free_opp;
 	}
 
+list_add:
+	new_opp->dev_opp = dev_opp;
 	list_add_rcu(&new_opp->node, head);
 	mutex_unlock(&dev_opp_list_lock);
 
@@ -481,11 +493,115 @@ int dev_pm_opp_add(struct device *dev, unsigned long freq, unsigned long u_volt)
 	 * Notify the changes in the availability of the operable
 	 * frequency/voltage list.
 	 */
-	srcu_notifier_call_chain(&dev_opp->head, OPP_EVENT_ADD, new_opp);
+	srcu_notifier_call_chain(&dev_opp->srcu_head, OPP_EVENT_ADD, new_opp);
 	return 0;
+
+free_opp:
+	mutex_unlock(&dev_opp_list_lock);
+	kfree(new_opp);
+	return ret;
+}
+
+/**
+ * dev_pm_opp_add()  - Add an OPP table from a table definitions
+ * @dev:	device for which we do this operation
+ * @freq:	Frequency in Hz for this OPP
+ * @u_volt:	Voltage in uVolts for this OPP
+ *
+ * This function adds an opp definition to the opp list and returns status.
+ * The opp is made available by default and it can be controlled using
+ * dev_pm_opp_enable/disable functions.
+ *
+ * Locking: The internal device_opp and opp structures are RCU protected.
+ * Hence this function internally uses RCU updater strategy with mutex locks
+ * to keep the integrity of the internal data structures. Callers should ensure
+ * that this function is *NOT* called under RCU protection or in contexts where
+ * mutex cannot be locked.
+ *
+ * Return:
+ * 0:		On success OR
+ *		Duplicate OPPs (both freq and volt are same) and opp->available
+ * -EEXIST:	Freq are same and volt are different OR
+ *		Duplicate OPPs (both freq and volt are same) and !opp->available
+ * -ENOMEM:	Memory allocation failure
+ */
+int dev_pm_opp_add(struct device *dev, unsigned long freq, unsigned long u_volt)
+{
+	return dev_pm_opp_add_dynamic(dev, freq, u_volt, true);
 }
 EXPORT_SYMBOL_GPL(dev_pm_opp_add);
 
+static void kfree_opp_rcu(struct rcu_head *head)
+{
+	struct dev_pm_opp *opp = container_of(head, struct dev_pm_opp, rcu_head);
+
+	kfree_rcu(opp, rcu_head);
+}
+
+static void kfree_device_rcu(struct rcu_head *head)
+{
+	struct device_opp *device_opp = container_of(head, struct device_opp, rcu_head);
+
+	kfree_rcu(device_opp, rcu_head);
+}
+
+static void __dev_pm_opp_remove(struct device_opp *dev_opp,
+				struct dev_pm_opp *opp)
+{
+	/*
+	 * Notify the changes in the availability of the operable
+	 * frequency/voltage list.
+	 */
+	srcu_notifier_call_chain(&dev_opp->srcu_head, OPP_EVENT_REMOVE, opp);
+	list_del_rcu(&opp->node);
+	call_srcu(&dev_opp->srcu_head.srcu, &opp->rcu_head, kfree_opp_rcu);
+
+	if (list_empty(&dev_opp->opp_list)) {
+		list_del_rcu(&dev_opp->node);
+		call_srcu(&dev_opp->srcu_head.srcu, &dev_opp->rcu_head,
+			  kfree_device_rcu);
+	}
+}
+
+/**
+ * dev_pm_opp_remove()  - Remove an OPP from OPP list
+ * @dev:	device for which we do this operation
+ * @freq:	OPP to remove with matching 'freq'
+ *
+ * This function removes an opp from the opp list.
+ */
+void dev_pm_opp_remove(struct device *dev, unsigned long freq)
+{
+	struct dev_pm_opp *opp;
+	struct device_opp *dev_opp;
+	bool found = false;
+
+	/* Hold our list modification lock here */
+	mutex_lock(&dev_opp_list_lock);
+
+	dev_opp = find_device_opp(dev);
+	if (IS_ERR(dev_opp))
+		goto unlock;
+
+	list_for_each_entry(opp, &dev_opp->opp_list, node) {
+		if (opp->rate == freq) {
+			found = true;
+			break;
+		}
+	}
+
+	if (!found) {
+		dev_warn(dev, "%s: Couldn't find OPP with freq: %lu\n",
+			 __func__, freq);
+		goto unlock;
+	}
+
+	__dev_pm_opp_remove(dev_opp, opp);
+unlock:
+	mutex_unlock(&dev_opp_list_lock);
+}
+EXPORT_SYMBOL_GPL(dev_pm_opp_remove);
+
 /**
  * opp_set_availability() - helper to set the availability of an opp
  * @dev:		device for which we do this operation
@@ -508,7 +624,7 @@ EXPORT_SYMBOL_GPL(dev_pm_opp_add);
 static int opp_set_availability(struct device *dev, unsigned long freq,
 		bool availability_req)
 {
-	struct device_opp *tmp_dev_opp, *dev_opp = ERR_PTR(-ENODEV);
+	struct device_opp *dev_opp;
 	struct dev_pm_opp *new_opp, *tmp_opp, *opp = ERR_PTR(-ENODEV);
 	int r = 0;
 
@@ -522,12 +638,7 @@ static int opp_set_availability(struct device *dev, unsigned long freq,
 	mutex_lock(&dev_opp_list_lock);
 
 	/* Find the device_opp */
-	list_for_each_entry(tmp_dev_opp, &dev_opp_list, node) {
-		if (dev == tmp_dev_opp->dev) {
-			dev_opp = tmp_dev_opp;
-			break;
-		}
-	}
+	dev_opp = find_device_opp(dev);
 	if (IS_ERR(dev_opp)) {
 		r = PTR_ERR(dev_opp);
 		dev_warn(dev, "%s: Device OPP not found (%d)\n", __func__, r);
@@ -557,14 +668,14 @@ static int opp_set_availability(struct device *dev, unsigned long freq,
 
 	list_replace_rcu(&opp->node, &new_opp->node);
 	mutex_unlock(&dev_opp_list_lock);
-	kfree_rcu(opp, head);
+	call_srcu(&dev_opp->srcu_head.srcu, &opp->rcu_head, kfree_opp_rcu);
 
 	/* Notify the change of the OPP availability */
 	if (availability_req)
-		srcu_notifier_call_chain(&dev_opp->head, OPP_EVENT_ENABLE,
+		srcu_notifier_call_chain(&dev_opp->srcu_head, OPP_EVENT_ENABLE,
 					 new_opp);
 	else
-		srcu_notifier_call_chain(&dev_opp->head, OPP_EVENT_DISABLE,
+		srcu_notifier_call_chain(&dev_opp->srcu_head, OPP_EVENT_DISABLE,
 					 new_opp);
 
 	return 0;
@@ -629,7 +740,7 @@ struct srcu_notifier_head *dev_pm_opp_get_notifier(struct device *dev)
 	if (IS_ERR(dev_opp))
 		return ERR_CAST(dev_opp); /* matching type */
 
-	return &dev_opp->head;
+	return &dev_opp->srcu_head;
 }
 
 #ifdef CONFIG_OF
@@ -666,7 +777,7 @@ int of_init_opp_table(struct device *dev)
 		unsigned long freq = be32_to_cpup(val++) * 1000;
 		unsigned long volt = be32_to_cpup(val++);
 
-		if (dev_pm_opp_add(dev, freq, volt))
+		if (dev_pm_opp_add_dynamic(dev, freq, volt, false))
 			dev_warn(dev, "%s: Failed to add OPP %ld\n",
 				 __func__, freq);
 		nr -= 2;
@@ -675,4 +786,40 @@ int of_init_opp_table(struct device *dev)
 	return 0;
 }
 EXPORT_SYMBOL_GPL(of_init_opp_table);
+
+/**
+ * of_free_opp_table() - Free OPP table entries created from static DT entries
+ * @dev:	device pointer used to lookup device OPPs.
+ *
+ * Free OPPs created using static entries present in DT.
+ */
+void of_free_opp_table(struct device *dev)
+{
+	struct device_opp *dev_opp;
+	struct dev_pm_opp *opp, *tmp;
+
+	/* Check for existing list for 'dev' */
+	dev_opp = find_device_opp(dev);
+	if (IS_ERR(dev_opp)) {
+		int error = PTR_ERR(dev_opp);
+		if (error != -ENODEV)
+			WARN(1, "%s: dev_opp: %d\n",
+			     IS_ERR_OR_NULL(dev) ?
+					"Invalid device" : dev_name(dev),
+			     error);
+		return;
+	}
+
+	/* Hold our list modification lock here */
+	mutex_lock(&dev_opp_list_lock);
+
+	/* Free static OPPs */
+	list_for_each_entry_safe(opp, tmp, &dev_opp->opp_list, node) {
+		if (!opp->dynamic)
+			__dev_pm_opp_remove(dev_opp, opp);
+	}
+
+	mutex_unlock(&dev_opp_list_lock);
+}
+EXPORT_SYMBOL_GPL(of_free_opp_table);
 #endif
diff --git a/drivers/base/power/power.h b/drivers/base/power/power.h
index a21223d95926..b6b8a273c5da 100644
--- a/drivers/base/power/power.h
+++ b/drivers/base/power/power.h
@@ -9,7 +9,7 @@ static inline void device_pm_init_common(struct device *dev)
 	}
 }
 
-#ifdef CONFIG_PM_RUNTIME
+#ifdef CONFIG_PM
 
 static inline void pm_runtime_early_init(struct device *dev)
 {
@@ -20,7 +20,21 @@ static inline void pm_runtime_early_init(struct device *dev)
 extern void pm_runtime_init(struct device *dev);
 extern void pm_runtime_remove(struct device *dev);
 
-#else /* !CONFIG_PM_RUNTIME */
+/*
+ * sysfs.c
+ */
+
+extern int dpm_sysfs_add(struct device *dev);
+extern void dpm_sysfs_remove(struct device *dev);
+extern void rpm_sysfs_remove(struct device *dev);
+extern int wakeup_sysfs_add(struct device *dev);
+extern void wakeup_sysfs_remove(struct device *dev);
+extern int pm_qos_sysfs_add_resume_latency(struct device *dev);
+extern void pm_qos_sysfs_remove_resume_latency(struct device *dev);
+extern int pm_qos_sysfs_add_flags(struct device *dev);
+extern void pm_qos_sysfs_remove_flags(struct device *dev);
+
+#else /* CONFIG_PM */
 
 static inline void pm_runtime_early_init(struct device *dev)
 {
@@ -30,7 +44,15 @@ static inline void pm_runtime_early_init(struct device *dev)
 static inline void pm_runtime_init(struct device *dev) {}
 static inline void pm_runtime_remove(struct device *dev) {}
 
-#endif /* !CONFIG_PM_RUNTIME */
+static inline int dpm_sysfs_add(struct device *dev) { return 0; }
+static inline void dpm_sysfs_remove(struct device *dev) {}
+static inline void rpm_sysfs_remove(struct device *dev) {}
+static inline int wakeup_sysfs_add(struct device *dev) { return 0; }
+static inline void wakeup_sysfs_remove(struct device *dev) {}
+static inline int pm_qos_sysfs_add(struct device *dev) { return 0; }
+static inline void pm_qos_sysfs_remove(struct device *dev) {}
+
+#endif
 
 #ifdef CONFIG_PM_SLEEP
 
@@ -77,31 +99,3 @@ static inline void device_pm_init(struct device *dev)
 	device_pm_sleep_init(dev);
 	pm_runtime_init(dev);
 }
-
-#ifdef CONFIG_PM
-
-/*
- * sysfs.c
- */
-
-extern int dpm_sysfs_add(struct device *dev);
-extern void dpm_sysfs_remove(struct device *dev);
-extern void rpm_sysfs_remove(struct device *dev);
-extern int wakeup_sysfs_add(struct device *dev);
-extern void wakeup_sysfs_remove(struct device *dev);
-extern int pm_qos_sysfs_add_resume_latency(struct device *dev);
-extern void pm_qos_sysfs_remove_resume_latency(struct device *dev);
-extern int pm_qos_sysfs_add_flags(struct device *dev);
-extern void pm_qos_sysfs_remove_flags(struct device *dev);
-
-#else /* CONFIG_PM */
-
-static inline int dpm_sysfs_add(struct device *dev) { return 0; }
-static inline void dpm_sysfs_remove(struct device *dev) {}
-static inline void rpm_sysfs_remove(struct device *dev) {}
-static inline int wakeup_sysfs_add(struct device *dev) { return 0; }
-static inline void wakeup_sysfs_remove(struct device *dev) {}
-static inline int pm_qos_sysfs_add(struct device *dev) { return 0; }
-static inline void pm_qos_sysfs_remove(struct device *dev) {}
-
-#endif
diff --git a/drivers/base/power/qos.c b/drivers/base/power/qos.c
index 36b9eb4862cb..a8fe4c1a8d07 100644
--- a/drivers/base/power/qos.c
+++ b/drivers/base/power/qos.c
@@ -599,7 +599,6 @@ int dev_pm_qos_add_ancestor_request(struct device *dev,
 }
 EXPORT_SYMBOL_GPL(dev_pm_qos_add_ancestor_request);
 
-#ifdef CONFIG_PM_RUNTIME
 static void __dev_pm_qos_drop_user_request(struct device *dev,
 					   enum dev_pm_qos_req_type type)
 {
@@ -880,7 +879,3 @@ int dev_pm_qos_update_user_latency_tolerance(struct device *dev, s32 val)
 	mutex_unlock(&dev_pm_qos_mtx);
 	return ret;
 }
-#else /* !CONFIG_PM_RUNTIME */
-static void __dev_pm_qos_hide_latency_limit(struct device *dev) {}
-static void __dev_pm_qos_hide_flags(struct device *dev) {}
-#endif /* CONFIG_PM_RUNTIME */
diff --git a/drivers/base/power/runtime.c b/drivers/base/power/runtime.c
index 67c7938e430b..5070c4fe8542 100644
--- a/drivers/base/power/runtime.c
+++ b/drivers/base/power/runtime.c
@@ -13,43 +13,38 @@
 #include <trace/events/rpm.h>
 #include "power.h"
 
-#define RPM_GET_CALLBACK(dev, cb)				\
-({								\
-	int (*__rpm_cb)(struct device *__d);			\
-								\
-	if (dev->pm_domain)					\
-		__rpm_cb = dev->pm_domain->ops.cb;		\
-	else if (dev->type && dev->type->pm)			\
-		__rpm_cb = dev->type->pm->cb;			\
-	else if (dev->class && dev->class->pm)			\
-		__rpm_cb = dev->class->pm->cb;			\
-	else if (dev->bus && dev->bus->pm)			\
-		__rpm_cb = dev->bus->pm->cb;			\
-	else							\
-		__rpm_cb = NULL;				\
-								\
-	if (!__rpm_cb && dev->driver && dev->driver->pm)	\
-		__rpm_cb = dev->driver->pm->cb;			\
-								\
-	__rpm_cb;						\
-})
-
-static int (*rpm_get_suspend_cb(struct device *dev))(struct device *)
-{
-	return RPM_GET_CALLBACK(dev, runtime_suspend);
-}
+typedef int (*pm_callback_t)(struct device *);
 
-static int (*rpm_get_resume_cb(struct device *dev))(struct device *)
+static pm_callback_t __rpm_get_callback(struct device *dev, size_t cb_offset)
 {
-	return RPM_GET_CALLBACK(dev, runtime_resume);
-}
+	pm_callback_t cb;
+	const struct dev_pm_ops *ops;
+
+	if (dev->pm_domain)
+		ops = &dev->pm_domain->ops;
+	else if (dev->type && dev->type->pm)
+		ops = dev->type->pm;
+	else if (dev->class && dev->class->pm)
+		ops = dev->class->pm;
+	else if (dev->bus && dev->bus->pm)
+		ops = dev->bus->pm;
+	else
+		ops = NULL;
 
-#ifdef CONFIG_PM_RUNTIME
-static int (*rpm_get_idle_cb(struct device *dev))(struct device *)
-{
-	return RPM_GET_CALLBACK(dev, runtime_idle);
+	if (ops)
+		cb = *(pm_callback_t *)((void *)ops + cb_offset);
+	else
+		cb = NULL;
+
+	if (!cb && dev->driver && dev->driver->pm)
+		cb = *(pm_callback_t *)((void *)dev->driver->pm + cb_offset);
+
+	return cb;
 }
 
+#define RPM_GET_CALLBACK(dev, callback) \
+		__rpm_get_callback(dev, offsetof(struct dev_pm_ops, callback))
+
 static int rpm_resume(struct device *dev, int rpmflags);
 static int rpm_suspend(struct device *dev, int rpmflags);
 
@@ -347,7 +342,7 @@ static int rpm_idle(struct device *dev, int rpmflags)
 
 	dev->power.idle_notification = true;
 
-	callback = rpm_get_idle_cb(dev);
+	callback = RPM_GET_CALLBACK(dev, runtime_idle);
 
 	if (callback)
 		retval = __rpm_callback(callback, dev);
@@ -517,7 +512,7 @@ static int rpm_suspend(struct device *dev, int rpmflags)
 
 	__update_runtime_status(dev, RPM_SUSPENDING);
 
-	callback = rpm_get_suspend_cb(dev);
+	callback = RPM_GET_CALLBACK(dev, runtime_suspend);
 
 	retval = rpm_callback(callback, dev);
 	if (retval)
@@ -737,7 +732,7 @@ static int rpm_resume(struct device *dev, int rpmflags)
 
 	__update_runtime_status(dev, RPM_RESUMING);
 
-	callback = rpm_get_resume_cb(dev);
+	callback = RPM_GET_CALLBACK(dev, runtime_resume);
 
 	retval = rpm_callback(callback, dev);
 	if (retval) {
@@ -1402,7 +1397,6 @@ void pm_runtime_remove(struct device *dev)
 	if (dev->power.irq_safe && dev->parent)
 		pm_runtime_put(dev->parent);
 }
-#endif
 
 /**
  * pm_runtime_force_suspend - Force a device into suspend state if needed.
@@ -1422,16 +1416,10 @@ int pm_runtime_force_suspend(struct device *dev)
 	int ret = 0;
 
 	pm_runtime_disable(dev);
-
-	/*
-	 * Note that pm_runtime_status_suspended() returns false while
-	 * !CONFIG_PM_RUNTIME, which means the device will be put into low
-	 * power state.
-	 */
 	if (pm_runtime_status_suspended(dev))
 		return 0;
 
-	callback = rpm_get_suspend_cb(dev);
+	callback = RPM_GET_CALLBACK(dev, runtime_suspend);
 
 	if (!callback) {
 		ret = -ENOSYS;
@@ -1467,7 +1455,7 @@ int pm_runtime_force_resume(struct device *dev)
 	int (*callback)(struct device *);
 	int ret = 0;
 
-	callback = rpm_get_resume_cb(dev);
+	callback = RPM_GET_CALLBACK(dev, runtime_resume);
 
 	if (!callback) {
 		ret = -ENOSYS;
diff --git a/drivers/base/power/sysfs.c b/drivers/base/power/sysfs.c
index 95b181d1ca6d..d2be3f9c211c 100644
--- a/drivers/base/power/sysfs.c
+++ b/drivers/base/power/sysfs.c
@@ -92,13 +92,9 @@
  *	wakeup_count - Report the number of wakeup events related to the device
  */
 
-static const char enabled[] = "enabled";
-static const char disabled[] = "disabled";
-
 const char power_group_name[] = "power";
 EXPORT_SYMBOL_GPL(power_group_name);
 
-#ifdef CONFIG_PM_RUNTIME
 static const char ctrl_auto[] = "auto";
 static const char ctrl_on[] = "on";
 
@@ -333,14 +329,16 @@ static ssize_t pm_qos_remote_wakeup_store(struct device *dev,
 
 static DEVICE_ATTR(pm_qos_remote_wakeup, 0644,
 		   pm_qos_remote_wakeup_show, pm_qos_remote_wakeup_store);
-#endif /* CONFIG_PM_RUNTIME */
 
 #ifdef CONFIG_PM_SLEEP
+static const char _enabled[] = "enabled";
+static const char _disabled[] = "disabled";
+
 static ssize_t
 wake_show(struct device * dev, struct device_attribute *attr, char * buf)
 {
 	return sprintf(buf, "%s\n", device_can_wakeup(dev)
-		? (device_may_wakeup(dev) ? enabled : disabled)
+		? (device_may_wakeup(dev) ? _enabled : _disabled)
 		: "");
 }
 
@@ -357,11 +355,11 @@ wake_store(struct device * dev, struct device_attribute *attr,
 	cp = memchr(buf, '\n', n);
 	if (cp)
 		len = cp - buf;
-	if (len == sizeof enabled - 1
-			&& strncmp(buf, enabled, sizeof enabled - 1) == 0)
+	if (len == sizeof _enabled - 1
+			&& strncmp(buf, _enabled, sizeof _enabled - 1) == 0)
 		device_set_wakeup_enable(dev, 1);
-	else if (len == sizeof disabled - 1
-			&& strncmp(buf, disabled, sizeof disabled - 1) == 0)
+	else if (len == sizeof _disabled - 1
+			&& strncmp(buf, _disabled, sizeof _disabled - 1) == 0)
 		device_set_wakeup_enable(dev, 0);
 	else
 		return -EINVAL;
@@ -531,8 +529,6 @@ static DEVICE_ATTR(wakeup_prevent_sleep_time_ms, 0444,
 #endif /* CONFIG_PM_SLEEP */
 
 #ifdef CONFIG_PM_ADVANCED_DEBUG
-#ifdef CONFIG_PM_RUNTIME
-
 static ssize_t rtpm_usagecount_show(struct device *dev,
 				    struct device_attribute *attr, char *buf)
 {
@@ -562,15 +558,13 @@ static DEVICE_ATTR(runtime_usage, 0444, rtpm_usagecount_show, NULL);
 static DEVICE_ATTR(runtime_active_kids, 0444, rtpm_children_show, NULL);
 static DEVICE_ATTR(runtime_enabled, 0444, rtpm_enabled_show, NULL);
 
-#endif
-
 #ifdef CONFIG_PM_SLEEP
-
 static ssize_t async_show(struct device *dev, struct device_attribute *attr,
 			  char *buf)
 {
 	return sprintf(buf, "%s\n",
-			device_async_suspend_enabled(dev) ? enabled : disabled);
+			device_async_suspend_enabled(dev) ?
+				_enabled : _disabled);
 }
 
 static ssize_t async_store(struct device *dev, struct device_attribute *attr,
@@ -582,9 +576,10 @@ static ssize_t async_store(struct device *dev, struct device_attribute *attr,
 	cp = memchr(buf, '\n', n);
 	if (cp)
 		len = cp - buf;
-	if (len == sizeof enabled - 1 && strncmp(buf, enabled, len) == 0)
+	if (len == sizeof _enabled - 1 && strncmp(buf, _enabled, len) == 0)
 		device_enable_async_suspend(dev);
-	else if (len == sizeof disabled - 1 && strncmp(buf, disabled, len) == 0)
+	else if (len == sizeof _disabled - 1 &&
+		 strncmp(buf, _disabled, len) == 0)
 		device_disable_async_suspend(dev);
 	else
 		return -EINVAL;
@@ -593,7 +588,7 @@ static ssize_t async_store(struct device *dev, struct device_attribute *attr,
 
 static DEVICE_ATTR(async, 0644, async_show, async_store);
 
-#endif
+#endif /* CONFIG_PM_SLEEP */
 #endif /* CONFIG_PM_ADVANCED_DEBUG */
 
 static struct attribute *power_attrs[] = {
@@ -601,12 +596,10 @@ static struct attribute *power_attrs[] = {
 #ifdef CONFIG_PM_SLEEP
 	&dev_attr_async.attr,
 #endif
-#ifdef CONFIG_PM_RUNTIME
 	&dev_attr_runtime_status.attr,
 	&dev_attr_runtime_usage.attr,
 	&dev_attr_runtime_active_kids.attr,
 	&dev_attr_runtime_enabled.attr,
-#endif
 #endif /* CONFIG_PM_ADVANCED_DEBUG */
 	NULL,
 };
@@ -638,7 +631,6 @@ static struct attribute_group pm_wakeup_attr_group = {
 };
 
 static struct attribute *runtime_attrs[] = {
-#ifdef CONFIG_PM_RUNTIME
 #ifndef CONFIG_PM_ADVANCED_DEBUG
 	&dev_attr_runtime_status.attr,
 #endif
@@ -646,7 +638,6 @@ static struct attribute *runtime_attrs[] = {
 	&dev_attr_runtime_suspended_time.attr,
 	&dev_attr_runtime_active_time.attr,
 	&dev_attr_autosuspend_delay_ms.attr,
-#endif /* CONFIG_PM_RUNTIME */
 	NULL,
 };
 static struct attribute_group pm_runtime_attr_group = {
@@ -655,9 +646,7 @@ static struct attribute_group pm_runtime_attr_group = {
 };
 
 static struct attribute *pm_qos_resume_latency_attrs[] = {
-#ifdef CONFIG_PM_RUNTIME
 	&dev_attr_pm_qos_resume_latency_us.attr,
-#endif /* CONFIG_PM_RUNTIME */
 	NULL,
 };
 static struct attribute_group pm_qos_resume_latency_attr_group = {
@@ -666,9 +655,7 @@ static struct attribute_group pm_qos_resume_latency_attr_group = {
 };
 
 static struct attribute *pm_qos_latency_tolerance_attrs[] = {
-#ifdef CONFIG_PM_RUNTIME
 	&dev_attr_pm_qos_latency_tolerance_us.attr,
-#endif /* CONFIG_PM_RUNTIME */
 	NULL,
 };
 static struct attribute_group pm_qos_latency_tolerance_attr_group = {
@@ -677,10 +664,8 @@ static struct attribute_group pm_qos_latency_tolerance_attr_group = {
 };
 
 static struct attribute *pm_qos_flags_attrs[] = {
-#ifdef CONFIG_PM_RUNTIME
 	&dev_attr_pm_qos_no_power_off.attr,
 	&dev_attr_pm_qos_remote_wakeup.attr,
-#endif /* CONFIG_PM_RUNTIME */
 	NULL,
 };
 static struct attribute_group pm_qos_flags_attr_group = {
diff --git a/drivers/base/power/wakeup.c b/drivers/base/power/wakeup.c
index eb1bd2ecad8b..c2744b30d5d9 100644
--- a/drivers/base/power/wakeup.c
+++ b/drivers/base/power/wakeup.c
@@ -24,6 +24,9 @@
  */
 bool events_check_enabled __read_mostly;
 
+/* If set and the system is suspending, terminate the suspend. */
+static bool pm_abort_suspend __read_mostly;
+
 /*
  * Combined counters of registered wakeup events and wakeup events in progress.
  * They need to be modified together atomically, so it's better to use one
@@ -719,7 +722,18 @@ bool pm_wakeup_pending(void)
 		pm_print_active_wakeup_sources();
 	}
 
-	return ret;
+	return ret || pm_abort_suspend;
+}
+
+void pm_system_wakeup(void)
+{
+	pm_abort_suspend = true;
+	freeze_wake();
+}
+
+void pm_wakeup_clear(void)
+{
+	pm_abort_suspend = false;
 }
 
 /**
diff --git a/drivers/base/property.c b/drivers/base/property.c
new file mode 100644
index 000000000000..c45845874d4f
--- /dev/null
+++ b/drivers/base/property.c
@@ -0,0 +1,431 @@
+/*
+ * property.c - Unified device property interface.
+ *
+ * Copyright (C) 2014, Intel Corporation
+ * Authors: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
+ *          Mika Westerberg <mika.westerberg@linux.intel.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/property.h>
+#include <linux/export.h>
+#include <linux/acpi.h>
+#include <linux/of.h>
+
+/**
+ * device_property_present - check if a property of a device is present
+ * @dev: Device whose property is being checked
+ * @propname: Name of the property
+ *
+ * Check if property @propname is present in the device firmware description.
+ */
+bool device_property_present(struct device *dev, const char *propname)
+{
+	if (IS_ENABLED(CONFIG_OF) && dev->of_node)
+		return of_property_read_bool(dev->of_node, propname);
+
+	return !acpi_dev_prop_get(ACPI_COMPANION(dev), propname, NULL);
+}
+EXPORT_SYMBOL_GPL(device_property_present);
+
+/**
+ * fwnode_property_present - check if a property of a firmware node is present
+ * @fwnode: Firmware node whose property to check
+ * @propname: Name of the property
+ */
+bool fwnode_property_present(struct fwnode_handle *fwnode, const char *propname)
+{
+	if (is_of_node(fwnode))
+		return of_property_read_bool(of_node(fwnode), propname);
+	else if (is_acpi_node(fwnode))
+		return !acpi_dev_prop_get(acpi_node(fwnode), propname, NULL);
+
+	return false;
+}
+EXPORT_SYMBOL_GPL(fwnode_property_present);
+
+#define OF_DEV_PROP_READ_ARRAY(node, propname, type, val, nval) \
+	(val) ? of_property_read_##type##_array((node), (propname), (val), (nval)) \
+	      : of_property_count_elems_of_size((node), (propname), sizeof(type))
+
+#define DEV_PROP_READ_ARRAY(_dev_, _propname_, _type_, _proptype_, _val_, _nval_) \
+	IS_ENABLED(CONFIG_OF) && _dev_->of_node ? \
+		(OF_DEV_PROP_READ_ARRAY(_dev_->of_node, _propname_, _type_, \
+					_val_, _nval_)) : \
+		acpi_dev_prop_read(ACPI_COMPANION(_dev_), _propname_, \
+				   _proptype_, _val_, _nval_)
+
+/**
+ * device_property_read_u8_array - return a u8 array property of a device
+ * @dev: Device to get the property of
+ * @propname: Name of the property
+ * @val: The values are stored here
+ * @nval: Size of the @val array
+ *
+ * Function reads an array of u8 properties with @propname from the device
+ * firmware description and stores them to @val if found.
+ *
+ * Return: %0 if the property was found (success),
+ *	   %-EINVAL if given arguments are not valid,
+ *	   %-ENODATA if the property does not have a value,
+ *	   %-EPROTO if the property is not an array of numbers,
+ *	   %-EOVERFLOW if the size of the property is not as expected.
+ */
+int device_property_read_u8_array(struct device *dev, const char *propname,
+				  u8 *val, size_t nval)
+{
+	return DEV_PROP_READ_ARRAY(dev, propname, u8, DEV_PROP_U8, val, nval);
+}
+EXPORT_SYMBOL_GPL(device_property_read_u8_array);
+
+/**
+ * device_property_read_u16_array - return a u16 array property of a device
+ * @dev: Device to get the property of
+ * @propname: Name of the property
+ * @val: The values are stored here
+ * @nval: Size of the @val array
+ *
+ * Function reads an array of u16 properties with @propname from the device
+ * firmware description and stores them to @val if found.
+ *
+ * Return: %0 if the property was found (success),
+ *	   %-EINVAL if given arguments are not valid,
+ *	   %-ENODATA if the property does not have a value,
+ *	   %-EPROTO if the property is not an array of numbers,
+ *	   %-EOVERFLOW if the size of the property is not as expected.
+ */
+int device_property_read_u16_array(struct device *dev, const char *propname,
+				   u16 *val, size_t nval)
+{
+	return DEV_PROP_READ_ARRAY(dev, propname, u16, DEV_PROP_U16, val, nval);
+}
+EXPORT_SYMBOL_GPL(device_property_read_u16_array);
+
+/**
+ * device_property_read_u32_array - return a u32 array property of a device
+ * @dev: Device to get the property of
+ * @propname: Name of the property
+ * @val: The values are stored here
+ * @nval: Size of the @val array
+ *
+ * Function reads an array of u32 properties with @propname from the device
+ * firmware description and stores them to @val if found.
+ *
+ * Return: %0 if the property was found (success),
+ *	   %-EINVAL if given arguments are not valid,
+ *	   %-ENODATA if the property does not have a value,
+ *	   %-EPROTO if the property is not an array of numbers,
+ *	   %-EOVERFLOW if the size of the property is not as expected.
+ */
+int device_property_read_u32_array(struct device *dev, const char *propname,
+				   u32 *val, size_t nval)
+{
+	return DEV_PROP_READ_ARRAY(dev, propname, u32, DEV_PROP_U32, val, nval);
+}
+EXPORT_SYMBOL_GPL(device_property_read_u32_array);
+
+/**
+ * device_property_read_u64_array - return a u64 array property of a device
+ * @dev: Device to get the property of
+ * @propname: Name of the property
+ * @val: The values are stored here
+ * @nval: Size of the @val array
+ *
+ * Function reads an array of u64 properties with @propname from the device
+ * firmware description and stores them to @val if found.
+ *
+ * Return: %0 if the property was found (success),
+ *	   %-EINVAL if given arguments are not valid,
+ *	   %-ENODATA if the property does not have a value,
+ *	   %-EPROTO if the property is not an array of numbers,
+ *	   %-EOVERFLOW if the size of the property is not as expected.
+ */
+int device_property_read_u64_array(struct device *dev, const char *propname,
+				   u64 *val, size_t nval)
+{
+	return DEV_PROP_READ_ARRAY(dev, propname, u64, DEV_PROP_U64, val, nval);
+}
+EXPORT_SYMBOL_GPL(device_property_read_u64_array);
+
+/**
+ * device_property_read_string_array - return a string array property of device
+ * @dev: Device to get the property of
+ * @propname: Name of the property
+ * @val: The values are stored here
+ * @nval: Size of the @val array
+ *
+ * Function reads an array of string properties with @propname from the device
+ * firmware description and stores them to @val if found.
+ *
+ * Return: %0 if the property was found (success),
+ *	   %-EINVAL if given arguments are not valid,
+ *	   %-ENODATA if the property does not have a value,
+ *	   %-EPROTO or %-EILSEQ if the property is not an array of strings,
+ *	   %-EOVERFLOW if the size of the property is not as expected.
+ */
+int device_property_read_string_array(struct device *dev, const char *propname,
+				      const char **val, size_t nval)
+{
+	return IS_ENABLED(CONFIG_OF) && dev->of_node ?
+		of_property_read_string_array(dev->of_node, propname, val, nval) :
+		acpi_dev_prop_read(ACPI_COMPANION(dev), propname,
+				   DEV_PROP_STRING, val, nval);
+}
+EXPORT_SYMBOL_GPL(device_property_read_string_array);
+
+/**
+ * device_property_read_string - return a string property of a device
+ * @dev: Device to get the property of
+ * @propname: Name of the property
+ * @val: The value is stored here
+ *
+ * Function reads property @propname from the device firmware description and
+ * stores the value into @val if found. The value is checked to be a string.
+ *
+ * Return: %0 if the property was found (success),
+ *	   %-EINVAL if given arguments are not valid,
+ *	   %-ENODATA if the property does not have a value,
+ *	   %-EPROTO or %-EILSEQ if the property type is not a string.
+ */
+int device_property_read_string(struct device *dev, const char *propname,
+				const char **val)
+{
+	return IS_ENABLED(CONFIG_OF) && dev->of_node ?
+		of_property_read_string(dev->of_node, propname, val) :
+		acpi_dev_prop_read(ACPI_COMPANION(dev), propname,
+				   DEV_PROP_STRING, val, 1);
+}
+EXPORT_SYMBOL_GPL(device_property_read_string);
+
+#define FWNODE_PROP_READ_ARRAY(_fwnode_, _propname_, _type_, _proptype_, _val_, _nval_) \
+({ \
+	int _ret_; \
+	if (is_of_node(_fwnode_)) \
+		_ret_ = OF_DEV_PROP_READ_ARRAY(of_node(_fwnode_), _propname_, \
+					       _type_, _val_, _nval_); \
+	else if (is_acpi_node(_fwnode_)) \
+		_ret_ = acpi_dev_prop_read(acpi_node(_fwnode_), _propname_, \
+					   _proptype_, _val_, _nval_); \
+	else \
+		_ret_ = -ENXIO; \
+	_ret_; \
+})
+
+/**
+ * fwnode_property_read_u8_array - return a u8 array property of firmware node
+ * @fwnode: Firmware node to get the property of
+ * @propname: Name of the property
+ * @val: The values are stored here
+ * @nval: Size of the @val array
+ *
+ * Read an array of u8 properties with @propname from @fwnode and stores them to
+ * @val if found.
+ *
+ * Return: %0 if the property was found (success),
+ *	   %-EINVAL if given arguments are not valid,
+ *	   %-ENODATA if the property does not have a value,
+ *	   %-EPROTO if the property is not an array of numbers,
+ *	   %-EOVERFLOW if the size of the property is not as expected,
+ *	   %-ENXIO if no suitable firmware interface is present.
+ */
+int fwnode_property_read_u8_array(struct fwnode_handle *fwnode,
+				  const char *propname, u8 *val, size_t nval)
+{
+	return FWNODE_PROP_READ_ARRAY(fwnode, propname, u8, DEV_PROP_U8,
+				      val, nval);
+}
+EXPORT_SYMBOL_GPL(fwnode_property_read_u8_array);
+
+/**
+ * fwnode_property_read_u16_array - return a u16 array property of firmware node
+ * @fwnode: Firmware node to get the property of
+ * @propname: Name of the property
+ * @val: The values are stored here
+ * @nval: Size of the @val array
+ *
+ * Read an array of u16 properties with @propname from @fwnode and store them to
+ * @val if found.
+ *
+ * Return: %0 if the property was found (success),
+ *	   %-EINVAL if given arguments are not valid,
+ *	   %-ENODATA if the property does not have a value,
+ *	   %-EPROTO if the property is not an array of numbers,
+ *	   %-EOVERFLOW if the size of the property is not as expected,
+ *	   %-ENXIO if no suitable firmware interface is present.
+ */
+int fwnode_property_read_u16_array(struct fwnode_handle *fwnode,
+				   const char *propname, u16 *val, size_t nval)
+{
+	return FWNODE_PROP_READ_ARRAY(fwnode, propname, u16, DEV_PROP_U16,
+				      val, nval);
+}
+EXPORT_SYMBOL_GPL(fwnode_property_read_u16_array);
+
+/**
+ * fwnode_property_read_u32_array - return a u32 array property of firmware node
+ * @fwnode: Firmware node to get the property of
+ * @propname: Name of the property
+ * @val: The values are stored here
+ * @nval: Size of the @val array
+ *
+ * Read an array of u32 properties with @propname from @fwnode store them to
+ * @val if found.
+ *
+ * Return: %0 if the property was found (success),
+ *	   %-EINVAL if given arguments are not valid,
+ *	   %-ENODATA if the property does not have a value,
+ *	   %-EPROTO if the property is not an array of numbers,
+ *	   %-EOVERFLOW if the size of the property is not as expected,
+ *	   %-ENXIO if no suitable firmware interface is present.
+ */
+int fwnode_property_read_u32_array(struct fwnode_handle *fwnode,
+				   const char *propname, u32 *val, size_t nval)
+{
+	return FWNODE_PROP_READ_ARRAY(fwnode, propname, u32, DEV_PROP_U32,
+				      val, nval);
+}
+EXPORT_SYMBOL_GPL(fwnode_property_read_u32_array);
+
+/**
+ * fwnode_property_read_u64_array - return a u64 array property firmware node
+ * @fwnode: Firmware node to get the property of
+ * @propname: Name of the property
+ * @val: The values are stored here
+ * @nval: Size of the @val array
+ *
+ * Read an array of u64 properties with @propname from @fwnode and store them to
+ * @val if found.
+ *
+ * Return: %0 if the property was found (success),
+ *	   %-EINVAL if given arguments are not valid,
+ *	   %-ENODATA if the property does not have a value,
+ *	   %-EPROTO if the property is not an array of numbers,
+ *	   %-EOVERFLOW if the size of the property is not as expected,
+ *	   %-ENXIO if no suitable firmware interface is present.
+ */
+int fwnode_property_read_u64_array(struct fwnode_handle *fwnode,
+				   const char *propname, u64 *val, size_t nval)
+{
+	return FWNODE_PROP_READ_ARRAY(fwnode, propname, u64, DEV_PROP_U64,
+				      val, nval);
+}
+EXPORT_SYMBOL_GPL(fwnode_property_read_u64_array);
+
+/**
+ * fwnode_property_read_string_array - return string array property of a node
+ * @fwnode: Firmware node to get the property of
+ * @propname: Name of the property
+ * @val: The values are stored here
+ * @nval: Size of the @val array
+ *
+ * Read an string list property @propname from the given firmware node and store
+ * them to @val if found.
+ *
+ * Return: %0 if the property was found (success),
+ *	   %-EINVAL if given arguments are not valid,
+ *	   %-ENODATA if the property does not have a value,
+ *	   %-EPROTO if the property is not an array of strings,
+ *	   %-EOVERFLOW if the size of the property is not as expected,
+ *	   %-ENXIO if no suitable firmware interface is present.
+ */
+int fwnode_property_read_string_array(struct fwnode_handle *fwnode,
+				      const char *propname, const char **val,
+				      size_t nval)
+{
+	if (is_of_node(fwnode))
+		return of_property_read_string_array(of_node(fwnode), propname,
+						     val, nval);
+	else if (is_acpi_node(fwnode))
+		return acpi_dev_prop_read(acpi_node(fwnode), propname,
+					  DEV_PROP_STRING, val, nval);
+
+	return -ENXIO;
+}
+EXPORT_SYMBOL_GPL(fwnode_property_read_string_array);
+
+/**
+ * fwnode_property_read_string - return a string property of a firmware node
+ * @fwnode: Firmware node to get the property of
+ * @propname: Name of the property
+ * @val: The value is stored here
+ *
+ * Read property @propname from the given firmware node and store the value into
+ * @val if found.  The value is checked to be a string.
+ *
+ * Return: %0 if the property was found (success),
+ *	   %-EINVAL if given arguments are not valid,
+ *	   %-ENODATA if the property does not have a value,
+ *	   %-EPROTO or %-EILSEQ if the property is not a string,
+ *	   %-ENXIO if no suitable firmware interface is present.
+ */
+int fwnode_property_read_string(struct fwnode_handle *fwnode,
+				const char *propname, const char **val)
+{
+	if (is_of_node(fwnode))
+		return of_property_read_string(of_node(fwnode),propname, val);
+	else if (is_acpi_node(fwnode))
+		return acpi_dev_prop_read(acpi_node(fwnode), propname,
+					  DEV_PROP_STRING, val, 1);
+
+	return -ENXIO;
+}
+EXPORT_SYMBOL_GPL(fwnode_property_read_string);
+
+/**
+ * device_get_next_child_node - Return the next child node handle for a device
+ * @dev: Device to find the next child node for.
+ * @child: Handle to one of the device's child nodes or a null handle.
+ */
+struct fwnode_handle *device_get_next_child_node(struct device *dev,
+						 struct fwnode_handle *child)
+{
+	if (IS_ENABLED(CONFIG_OF) && dev->of_node) {
+		struct device_node *node;
+
+		node = of_get_next_available_child(dev->of_node, of_node(child));
+		if (node)
+			return &node->fwnode;
+	} else if (IS_ENABLED(CONFIG_ACPI)) {
+		struct acpi_device *node;
+
+		node = acpi_get_next_child(dev, acpi_node(child));
+		if (node)
+			return acpi_fwnode_handle(node);
+	}
+	return NULL;
+}
+EXPORT_SYMBOL_GPL(device_get_next_child_node);
+
+/**
+ * fwnode_handle_put - Drop reference to a device node
+ * @fwnode: Pointer to the device node to drop the reference to.
+ *
+ * This has to be used when terminating device_for_each_child_node() iteration
+ * with break or return to prevent stale device node references from being left
+ * behind.
+ */
+void fwnode_handle_put(struct fwnode_handle *fwnode)
+{
+	if (is_of_node(fwnode))
+		of_node_put(of_node(fwnode));
+}
+EXPORT_SYMBOL_GPL(fwnode_handle_put);
+
+/**
+ * device_get_child_node_count - return the number of child nodes for device
+ * @dev: Device to cound the child nodes for
+ */
+unsigned int device_get_child_node_count(struct device *dev)
+{
+	struct fwnode_handle *child;
+	unsigned int count = 0;
+
+	device_for_each_child_node(dev, child)
+		count++;
+
+	return count;
+}
+EXPORT_SYMBOL_GPL(device_get_child_node_count);
diff --git a/drivers/base/regmap/Kconfig b/drivers/base/regmap/Kconfig
index 4251570610c9..db9d00c36a3e 100644
--- a/drivers/base/regmap/Kconfig
+++ b/drivers/base/regmap/Kconfig
@@ -3,20 +3,26 @@
 # subsystems should select the appropriate symbols.
 
 config REGMAP
-	default y if (REGMAP_I2C || REGMAP_SPI || REGMAP_SPMI || REGMAP_MMIO || REGMAP_IRQ)
+	default y if (REGMAP_I2C || REGMAP_SPI || REGMAP_SPMI || REGMAP_AC97 || REGMAP_MMIO || REGMAP_IRQ)
 	select LZO_COMPRESS
 	select LZO_DECOMPRESS
 	select IRQ_DOMAIN if REGMAP_IRQ
 	bool
 
+config REGMAP_AC97
+	tristate
+
 config REGMAP_I2C
 	tristate
+	depends on I2C
 
 config REGMAP_SPI
 	tristate
+	depends on SPI
 
 config REGMAP_SPMI
 	tristate
+	depends on SPMI
 
 config REGMAP_MMIO
 	tristate
diff --git a/drivers/base/regmap/Makefile b/drivers/base/regmap/Makefile
index a7c670b4123a..0a533653ef3b 100644
--- a/drivers/base/regmap/Makefile
+++ b/drivers/base/regmap/Makefile
@@ -1,6 +1,7 @@
 obj-$(CONFIG_REGMAP) += regmap.o regcache.o
 obj-$(CONFIG_REGMAP) += regcache-rbtree.o regcache-lzo.o regcache-flat.o
 obj-$(CONFIG_DEBUG_FS) += regmap-debugfs.o
+obj-$(CONFIG_REGMAP_AC97) += regmap-ac97.o
 obj-$(CONFIG_REGMAP_I2C) += regmap-i2c.o
 obj-$(CONFIG_REGMAP_SPI) += regmap-spi.o
 obj-$(CONFIG_REGMAP_SPMI) += regmap-spmi.o
diff --git a/drivers/base/regmap/internal.h b/drivers/base/regmap/internal.h
index bfc90b8547f2..0da5865df5b1 100644
--- a/drivers/base/regmap/internal.h
+++ b/drivers/base/regmap/internal.h
@@ -49,8 +49,10 @@ struct regmap_async {
 };
 
 struct regmap {
-	struct mutex mutex;
-	spinlock_t spinlock;
+	union {
+		struct mutex mutex;
+		spinlock_t spinlock;
+	};
 	unsigned long spinlock_flags;
 	regmap_lock lock;
 	regmap_unlock unlock;
diff --git a/drivers/base/regmap/regcache-flat.c b/drivers/base/regmap/regcache-flat.c
index d9762e41959b..0246f44ded74 100644
--- a/drivers/base/regmap/regcache-flat.c
+++ b/drivers/base/regmap/regcache-flat.c
@@ -10,9 +10,9 @@
  * published by the Free Software Foundation.
  */
 
-#include <linux/slab.h>
 #include <linux/device.h>
 #include <linux/seq_file.h>
+#include <linux/slab.h>
 
 #include "internal.h"
 
diff --git a/drivers/base/regmap/regcache-lzo.c b/drivers/base/regmap/regcache-lzo.c
index e210a6d1406a..2d53f6f138e1 100644
--- a/drivers/base/regmap/regcache-lzo.c
+++ b/drivers/base/regmap/regcache-lzo.c
@@ -10,9 +10,9 @@
  * published by the Free Software Foundation.
  */
 
-#include <linux/slab.h>
 #include <linux/device.h>
 #include <linux/lzo.h>
+#include <linux/slab.h>
 
 #include "internal.h"
 
diff --git a/drivers/base/regmap/regcache-rbtree.c b/drivers/base/regmap/regcache-rbtree.c
index f3e8fe0cc650..d453a2c98ad0 100644
--- a/drivers/base/regmap/regcache-rbtree.c
+++ b/drivers/base/regmap/regcache-rbtree.c
@@ -10,11 +10,11 @@
  * published by the Free Software Foundation.
  */
 
-#include <linux/slab.h>
-#include <linux/device.h>
 #include <linux/debugfs.h>
+#include <linux/device.h>
 #include <linux/rbtree.h>
 #include <linux/seq_file.h>
+#include <linux/slab.h>
 
 #include "internal.h"
 
diff --git a/drivers/base/regmap/regcache.c b/drivers/base/regmap/regcache.c
index 5617da6dc898..f373c35f9e1d 100644
--- a/drivers/base/regmap/regcache.c
+++ b/drivers/base/regmap/regcache.c
@@ -10,12 +10,12 @@
  * published by the Free Software Foundation.
  */
 
-#include <linux/slab.h>
-#include <linux/export.h>
-#include <linux/device.h>
-#include <trace/events/regmap.h>
 #include <linux/bsearch.h>
+#include <linux/device.h>
+#include <linux/export.h>
+#include <linux/slab.h>
 #include <linux/sort.h>
+#include <trace/events/regmap.h>
 
 #include "internal.h"
 
@@ -36,6 +36,23 @@ static int regcache_hw_init(struct regmap *map)
 	if (!map->num_reg_defaults_raw)
 		return -EINVAL;
 
+	/* calculate the size of reg_defaults */
+	for (count = 0, i = 0; i < map->num_reg_defaults_raw; i++)
+		if (!regmap_volatile(map, i * map->reg_stride))
+			count++;
+
+	/* all registers are volatile, so just bypass */
+	if (!count) {
+		map->cache_bypass = true;
+		return 0;
+	}
+
+	map->num_reg_defaults = count;
+	map->reg_defaults = kmalloc_array(count, sizeof(struct reg_default),
+					  GFP_KERNEL);
+	if (!map->reg_defaults)
+		return -ENOMEM;
+
 	if (!map->reg_defaults_raw) {
 		u32 cache_bypass = map->cache_bypass;
 		dev_warn(map->dev, "No cache defaults, reading back from HW\n");
@@ -43,40 +60,25 @@ static int regcache_hw_init(struct regmap *map)
 		/* Bypass the cache access till data read from HW*/
 		map->cache_bypass = 1;
 		tmp_buf = kmalloc(map->cache_size_raw, GFP_KERNEL);
-		if (!tmp_buf)
-			return -EINVAL;
+		if (!tmp_buf) {
+			ret = -ENOMEM;
+			goto err_free;
+		}
 		ret = regmap_raw_read(map, 0, tmp_buf,
 				      map->num_reg_defaults_raw);
 		map->cache_bypass = cache_bypass;
-		if (ret < 0) {
-			kfree(tmp_buf);
-			return ret;
-		}
+		if (ret < 0)
+			goto err_cache_free;
+
 		map->reg_defaults_raw = tmp_buf;
 		map->cache_free = 1;
 	}
 
-	/* calculate the size of reg_defaults */
-	for (count = 0, i = 0; i < map->num_reg_defaults_raw; i++) {
-		val = regcache_get_val(map, map->reg_defaults_raw, i);
-		if (regmap_volatile(map, i * map->reg_stride))
-			continue;
-		count++;
-	}
-
-	map->reg_defaults = kmalloc(count * sizeof(struct reg_default),
-				      GFP_KERNEL);
-	if (!map->reg_defaults) {
-		ret = -ENOMEM;
-		goto err_free;
-	}
-
 	/* fill the reg_defaults */
-	map->num_reg_defaults = count;
 	for (i = 0, j = 0; i < map->num_reg_defaults_raw; i++) {
-		val = regcache_get_val(map, map->reg_defaults_raw, i);
 		if (regmap_volatile(map, i * map->reg_stride))
 			continue;
+		val = regcache_get_val(map, map->reg_defaults_raw, i);
 		map->reg_defaults[j].reg = i * map->reg_stride;
 		map->reg_defaults[j].def = val;
 		j++;
@@ -84,9 +86,10 @@ static int regcache_hw_init(struct regmap *map)
 
 	return 0;
 
+err_cache_free:
+	kfree(tmp_buf);
 err_free:
-	if (map->cache_free)
-		kfree(map->reg_defaults_raw);
+	kfree(map->reg_defaults);
 
 	return ret;
 }
@@ -150,6 +153,8 @@ int regcache_init(struct regmap *map, const struct regmap_config *config)
 		ret = regcache_hw_init(map);
 		if (ret < 0)
 			return ret;
+		if (map->cache_bypass)
+			return 0;
 	}
 
 	if (!map->max_register)
@@ -269,8 +274,11 @@ static int regcache_default_sync(struct regmap *map, unsigned int min,
 		map->cache_bypass = 1;
 		ret = _regmap_write(map, reg, val);
 		map->cache_bypass = 0;
-		if (ret)
+		if (ret) {
+			dev_err(map->dev, "Unable to sync register %#x. %d\n",
+				reg, ret);
 			return ret;
+		}
 		dev_dbg(map->dev, "Synced register %#x, value %#x\n", reg, val);
 	}
 
@@ -615,8 +623,11 @@ static int regcache_sync_block_single(struct regmap *map, void *block,
 		ret = _regmap_write(map, regtmp, val);
 
 		map->cache_bypass = 0;
-		if (ret != 0)
+		if (ret != 0) {
+			dev_err(map->dev, "Unable to sync register %#x. %d\n",
+				regtmp, ret);
 			return ret;
+		}
 		dev_dbg(map->dev, "Synced register %#x, value %#x\n",
 			regtmp, val);
 	}
@@ -641,6 +652,9 @@ static int regcache_sync_block_raw_flush(struct regmap *map, const void **data,
 	map->cache_bypass = 1;
 
 	ret = _regmap_raw_write(map, base, *data, count * val_bytes);
+	if (ret)
+		dev_err(map->dev, "Unable to sync registers %#x-%#x. %d\n",
+			base, cur - map->reg_stride, ret);
 
 	map->cache_bypass = 0;
 
diff --git a/drivers/base/regmap/regmap-ac97.c b/drivers/base/regmap/regmap-ac97.c
new file mode 100644
index 000000000000..e4c45d2299c1
--- /dev/null
+++ b/drivers/base/regmap/regmap-ac97.c
@@ -0,0 +1,114 @@
+/*
+ * Register map access API - AC'97 support
+ *
+ * Copyright 2013 Linaro Ltd.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/clk.h>
+#include <linux/err.h>
+#include <linux/init.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/regmap.h>
+#include <linux/slab.h>
+
+#include <sound/ac97_codec.h>
+
+bool regmap_ac97_default_volatile(struct device *dev, unsigned int reg)
+{
+	switch (reg) {
+	case AC97_RESET:
+	case AC97_POWERDOWN:
+	case AC97_INT_PAGING:
+	case AC97_EXTENDED_ID:
+	case AC97_EXTENDED_STATUS:
+	case AC97_EXTENDED_MID:
+	case AC97_EXTENDED_MSTATUS:
+	case AC97_GPIO_STATUS:
+	case AC97_MISC_AFE:
+	case AC97_VENDOR_ID1:
+	case AC97_VENDOR_ID2:
+	case AC97_CODEC_CLASS_REV:
+	case AC97_PCI_SVID:
+	case AC97_PCI_SID:
+	case AC97_FUNC_SELECT:
+	case AC97_FUNC_INFO:
+	case AC97_SENSE_INFO:
+		return true;
+	default:
+		return false;
+	}
+}
+EXPORT_SYMBOL_GPL(regmap_ac97_default_volatile);
+
+static int regmap_ac97_reg_read(void *context, unsigned int reg,
+	unsigned int *val)
+{
+	struct snd_ac97 *ac97 = context;
+
+	*val = ac97->bus->ops->read(ac97, reg);
+
+	return 0;
+}
+
+static int regmap_ac97_reg_write(void *context, unsigned int reg,
+	unsigned int val)
+{
+	struct snd_ac97 *ac97 = context;
+
+	ac97->bus->ops->write(ac97, reg, val);
+
+	return 0;
+}
+
+static const struct regmap_bus ac97_regmap_bus = {
+		.reg_write = regmap_ac97_reg_write,
+		.reg_read = regmap_ac97_reg_read,
+};
+
+/**
+ * regmap_init_ac97(): Initialise AC'97 register map
+ *
+ * @ac97: Device that will be interacted with
+ * @config: Configuration for register map
+ *
+ * The return value will be an ERR_PTR() on error or a valid pointer to
+ * a struct regmap.
+ */
+struct regmap *regmap_init_ac97(struct snd_ac97 *ac97,
+				const struct regmap_config *config)
+{
+	return regmap_init(&ac97->dev, &ac97_regmap_bus, ac97, config);
+}
+EXPORT_SYMBOL_GPL(regmap_init_ac97);
+
+/**
+ * devm_regmap_init_ac97(): Initialise AC'97 register map
+ *
+ * @ac97: Device that will be interacted with
+ * @config: Configuration for register map
+ *
+ * The return value will be an ERR_PTR() on error or a valid pointer
+ * to a struct regmap.  The regmap will be automatically freed by the
+ * device management code.
+ */
+struct regmap *devm_regmap_init_ac97(struct snd_ac97 *ac97,
+				     const struct regmap_config *config)
+{
+	return devm_regmap_init(&ac97->dev, &ac97_regmap_bus, ac97, config);
+}
+EXPORT_SYMBOL_GPL(devm_regmap_init_ac97);
+
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/base/regmap/regmap-debugfs.c b/drivers/base/regmap/regmap-debugfs.c
index 0c94b661c16f..5799a0b9e6cc 100644
--- a/drivers/base/regmap/regmap-debugfs.c
+++ b/drivers/base/regmap/regmap-debugfs.c
@@ -473,6 +473,7 @@ void regmap_debugfs_init(struct regmap *map, const char *name)
 {
 	struct rb_node *next;
 	struct regmap_range_node *range_node;
+	const char *devname = "dummy";
 
 	/* If we don't have the debugfs root yet, postpone init */
 	if (!regmap_debugfs_root) {
@@ -491,12 +492,15 @@ void regmap_debugfs_init(struct regmap *map, const char *name)
 	INIT_LIST_HEAD(&map->debugfs_off_cache);
 	mutex_init(&map->cache_lock);
 
+	if (map->dev)
+		devname = dev_name(map->dev);
+
 	if (name) {
 		map->debugfs_name = kasprintf(GFP_KERNEL, "%s-%s",
-					      dev_name(map->dev), name);
+					      devname, name);
 		name = map->debugfs_name;
 	} else {
-		name = dev_name(map->dev);
+		name = devname;
 	}
 
 	map->debugfs = debugfs_create_dir(name, regmap_debugfs_root);
diff --git a/drivers/base/regmap/regmap-i2c.c b/drivers/base/regmap/regmap-i2c.c
index ca193d1ef47c..053150a7f9f2 100644
--- a/drivers/base/regmap/regmap-i2c.c
+++ b/drivers/base/regmap/regmap-i2c.c
@@ -168,6 +168,8 @@ static struct regmap_bus regmap_i2c = {
 	.write = regmap_i2c_write,
 	.gather_write = regmap_i2c_gather_write,
 	.read = regmap_i2c_read,
+	.reg_format_endian_default = REGMAP_ENDIAN_BIG,
+	.val_format_endian_default = REGMAP_ENDIAN_BIG,
 };
 
 static const struct regmap_bus *regmap_get_i2c_bus(struct i2c_client *i2c,
diff --git a/drivers/base/regmap/regmap-spi.c b/drivers/base/regmap/regmap-spi.c
index 0eb3097c0d76..53d1148e80a0 100644
--- a/drivers/base/regmap/regmap-spi.c
+++ b/drivers/base/regmap/regmap-spi.c
@@ -109,6 +109,8 @@ static struct regmap_bus regmap_spi = {
 	.async_alloc = regmap_spi_async_alloc,
 	.read = regmap_spi_read,
 	.read_flag_mask = 0x80,
+	.reg_format_endian_default = REGMAP_ENDIAN_BIG,
+	.val_format_endian_default = REGMAP_ENDIAN_BIG,
 };
 
 /**
diff --git a/drivers/base/regmap/regmap.c b/drivers/base/regmap/regmap.c
index 1cf427bc0d4a..d2f8a818d200 100644
--- a/drivers/base/regmap/regmap.c
+++ b/drivers/base/regmap/regmap.c
@@ -15,6 +15,7 @@
 #include <linux/export.h>
 #include <linux/mutex.h>
 #include <linux/err.h>
+#include <linux/of.h>
 #include <linux/rbtree.h>
 #include <linux/sched.h>
 
@@ -448,6 +449,71 @@ int regmap_attach_dev(struct device *dev, struct regmap *map,
 }
 EXPORT_SYMBOL_GPL(regmap_attach_dev);
 
+static enum regmap_endian regmap_get_reg_endian(const struct regmap_bus *bus,
+					const struct regmap_config *config)
+{
+	enum regmap_endian endian;
+
+	/* Retrieve the endianness specification from the regmap config */
+	endian = config->reg_format_endian;
+
+	/* If the regmap config specified a non-default value, use that */
+	if (endian != REGMAP_ENDIAN_DEFAULT)
+		return endian;
+
+	/* Retrieve the endianness specification from the bus config */
+	if (bus && bus->reg_format_endian_default)
+		endian = bus->reg_format_endian_default;
+
+	/* If the bus specified a non-default value, use that */
+	if (endian != REGMAP_ENDIAN_DEFAULT)
+		return endian;
+
+	/* Use this if no other value was found */
+	return REGMAP_ENDIAN_BIG;
+}
+
+static enum regmap_endian regmap_get_val_endian(struct device *dev,
+					const struct regmap_bus *bus,
+					const struct regmap_config *config)
+{
+	struct device_node *np;
+	enum regmap_endian endian;
+
+	/* Retrieve the endianness specification from the regmap config */
+	endian = config->val_format_endian;
+
+	/* If the regmap config specified a non-default value, use that */
+	if (endian != REGMAP_ENDIAN_DEFAULT)
+		return endian;
+
+	/* If the dev and dev->of_node exist try to get endianness from DT */
+	if (dev && dev->of_node) {
+		np = dev->of_node;
+
+		/* Parse the device's DT node for an endianness specification */
+		if (of_property_read_bool(np, "big-endian"))
+			endian = REGMAP_ENDIAN_BIG;
+		else if (of_property_read_bool(np, "little-endian"))
+			endian = REGMAP_ENDIAN_LITTLE;
+
+		/* If the endianness was specified in DT, use that */
+		if (endian != REGMAP_ENDIAN_DEFAULT)
+			return endian;
+	}
+
+	/* Retrieve the endianness specification from the bus config */
+	if (bus && bus->val_format_endian_default)
+		endian = bus->val_format_endian_default;
+
+	/* If the bus specified a non-default value, use that */
+	if (endian != REGMAP_ENDIAN_DEFAULT)
+		return endian;
+
+	/* Use this if no other value was found */
+	return REGMAP_ENDIAN_BIG;
+}
+
 /**
  * regmap_init(): Initialise register map
  *
@@ -551,17 +617,8 @@ struct regmap *regmap_init(struct device *dev,
 		map->reg_read  = _regmap_bus_read;
 	}
 
-	reg_endian = config->reg_format_endian;
-	if (reg_endian == REGMAP_ENDIAN_DEFAULT)
-		reg_endian = bus->reg_format_endian_default;
-	if (reg_endian == REGMAP_ENDIAN_DEFAULT)
-		reg_endian = REGMAP_ENDIAN_BIG;
-
-	val_endian = config->val_format_endian;
-	if (val_endian == REGMAP_ENDIAN_DEFAULT)
-		val_endian = bus->val_format_endian_default;
-	if (val_endian == REGMAP_ENDIAN_DEFAULT)
-		val_endian = REGMAP_ENDIAN_BIG;
+	reg_endian = regmap_get_reg_endian(bus, config);
+	val_endian = regmap_get_val_endian(dev, bus, config);
 
 	switch (config->reg_bits + map->reg_shift) {
 	case 2:
@@ -1408,7 +1465,7 @@ int _regmap_write(struct regmap *map, unsigned int reg,
 	}
 
 #ifdef LOG_DEVICE
-	if (strcmp(dev_name(map->dev), LOG_DEVICE) == 0)
+	if (map->dev && strcmp(dev_name(map->dev), LOG_DEVICE) == 0)
 		dev_info(map->dev, "%x <= %x\n", reg, val);
 #endif
 
@@ -1659,6 +1716,9 @@ out:
 	} else {
 		void *wval;
 
+		if (!val_count)
+			return -EINVAL;
+
 		wval = kmemdup(val, val_count * val_bytes, GFP_KERNEL);
 		if (!wval) {
 			dev_err(map->dev, "Error in memory allocation\n");
@@ -2058,7 +2118,7 @@ static int _regmap_read(struct regmap *map, unsigned int reg,
 	ret = map->reg_read(context, reg, val);
 	if (ret == 0) {
 #ifdef LOG_DEVICE
-		if (strcmp(dev_name(map->dev), LOG_DEVICE) == 0)
+		if (map->dev && strcmp(dev_name(map->dev), LOG_DEVICE) == 0)
 			dev_info(map->dev, "%x => %x\n", reg, *val);
 #endif
 
diff --git a/drivers/base/syscore.c b/drivers/base/syscore.c
index dbb8350ea8dc..8d98a329f6ea 100644
--- a/drivers/base/syscore.c
+++ b/drivers/base/syscore.c
@@ -9,7 +9,7 @@
 #include <linux/syscore_ops.h>
 #include <linux/mutex.h>
 #include <linux/module.h>
-#include <linux/interrupt.h>
+#include <linux/suspend.h>
 #include <trace/events/power.h>
 
 static LIST_HEAD(syscore_ops_list);
@@ -54,9 +54,8 @@ int syscore_suspend(void)
 	pr_debug("Checking wakeup interrupts\n");
 
 	/* Return error code if there are any wakeup interrupts pending. */
-	ret = check_wakeup_irqs();
-	if (ret)
-		return ret;
+	if (pm_wakeup_pending())
+		return -EBUSY;
 
 	WARN_ONCE(!irqs_disabled(),
 		"Interrupts enabled before system core suspend.\n");
diff --git a/drivers/base/topology.c b/drivers/base/topology.c
index be7c1fb7c0c9..6491f45200a7 100644
--- a/drivers/base/topology.c
+++ b/drivers/base/topology.c
@@ -29,75 +29,52 @@
 #include <linux/hardirq.h>
 #include <linux/topology.h>
 
-#define define_one_ro_named(_name, _func)				\
-	static DEVICE_ATTR(_name, 0444, _func, NULL)
-
-#define define_one_ro(_name)				\
-	static DEVICE_ATTR(_name, 0444, show_##_name, NULL)
-
 #define define_id_show_func(name)				\
-static ssize_t show_##name(struct device *dev,			\
+static ssize_t name##_show(struct device *dev,			\
 		struct device_attribute *attr, char *buf)	\
 {								\
 	return sprintf(buf, "%d\n", topology_##name(dev->id));	\
 }
 
-#if defined(topology_thread_cpumask) || defined(topology_core_cpumask) || \
-    defined(topology_book_cpumask)
-static ssize_t show_cpumap(int type, const struct cpumask *mask, char *buf)
-{
-	ptrdiff_t len = PTR_ALIGN(buf + PAGE_SIZE - 1, PAGE_SIZE) - buf;
-	int n = 0;
-
-	if (len > 1) {
-		n = type?
-			cpulist_scnprintf(buf, len-2, mask) :
-			cpumask_scnprintf(buf, len-2, mask);
-		buf[n++] = '\n';
-		buf[n] = '\0';
-	}
-	return n;
-}
-#endif
-
-#define define_siblings_show_map(name)					\
-static ssize_t show_##name(struct device *dev,				\
+#define define_siblings_show_map(name, mask)				\
+static ssize_t name##_show(struct device *dev,				\
 			   struct device_attribute *attr, char *buf)	\
 {									\
-	return show_cpumap(0, topology_##name(dev->id), buf);		\
+	return cpumap_print_to_pagebuf(false, buf, topology_##mask(dev->id));\
 }
 
-#define define_siblings_show_list(name)					\
-static ssize_t show_##name##_list(struct device *dev,			\
-				  struct device_attribute *attr,	\
-				  char *buf)				\
+#define define_siblings_show_list(name, mask)				\
+static ssize_t name##_list_show(struct device *dev,			\
+				struct device_attribute *attr,		\
+				char *buf)				\
 {									\
-	return show_cpumap(1, topology_##name(dev->id), buf);		\
+	return cpumap_print_to_pagebuf(true, buf, topology_##mask(dev->id));\
 }
 
-#define define_siblings_show_func(name)		\
-	define_siblings_show_map(name); define_siblings_show_list(name)
+#define define_siblings_show_func(name, mask)	\
+	define_siblings_show_map(name, mask);	\
+	define_siblings_show_list(name, mask)
 
 define_id_show_func(physical_package_id);
-define_one_ro(physical_package_id);
+static DEVICE_ATTR_RO(physical_package_id);
 
 define_id_show_func(core_id);
-define_one_ro(core_id);
+static DEVICE_ATTR_RO(core_id);
 
-define_siblings_show_func(thread_cpumask);
-define_one_ro_named(thread_siblings, show_thread_cpumask);
-define_one_ro_named(thread_siblings_list, show_thread_cpumask_list);
+define_siblings_show_func(thread_siblings, thread_cpumask);
+static DEVICE_ATTR_RO(thread_siblings);
+static DEVICE_ATTR_RO(thread_siblings_list);
 
-define_siblings_show_func(core_cpumask);
-define_one_ro_named(core_siblings, show_core_cpumask);
-define_one_ro_named(core_siblings_list, show_core_cpumask_list);
+define_siblings_show_func(core_siblings, core_cpumask);
+static DEVICE_ATTR_RO(core_siblings);
+static DEVICE_ATTR_RO(core_siblings_list);
 
 #ifdef CONFIG_SCHED_BOOK
 define_id_show_func(book_id);
-define_one_ro(book_id);
-define_siblings_show_func(book_cpumask);
-define_one_ro_named(book_siblings, show_book_cpumask);
-define_one_ro_named(book_siblings_list, show_book_cpumask_list);
+static DEVICE_ATTR_RO(book_id);
+define_siblings_show_func(book_siblings, book_cpumask);
+static DEVICE_ATTR_RO(book_siblings);
+static DEVICE_ATTR_RO(book_siblings_list);
 #endif
 
 static struct attribute *default_attrs[] = {
diff --git a/drivers/bcma/Makefile b/drivers/bcma/Makefile
index 91290f7f61b8..838b4b9d352f 100644
--- a/drivers/bcma/Makefile
+++ b/drivers/bcma/Makefile
@@ -1,5 +1,6 @@
 bcma-y					+= main.o scan.o core.o sprom.o
 bcma-y					+= driver_chipcommon.o driver_chipcommon_pmu.o
+bcma-y					+= driver_chipcommon_b.o
 bcma-$(CONFIG_BCMA_SFLASH)		+= driver_chipcommon_sflash.o
 bcma-$(CONFIG_BCMA_NFLASH)		+= driver_chipcommon_nflash.o
 bcma-y					+= driver_pci.o
diff --git a/drivers/bcma/bcma_private.h b/drivers/bcma/bcma_private.h
index 09b632ad0fe2..314ae4032f3e 100644
--- a/drivers/bcma/bcma_private.h
+++ b/drivers/bcma/bcma_private.h
@@ -24,6 +24,7 @@ struct bcma_bus;
 /* main.c */
 bool bcma_wait_value(struct bcma_device *core, u16 reg, u32 mask, u32 value,
 		     int timeout);
+void bcma_prepare_core(struct bcma_bus *bus, struct bcma_device *core);
 int bcma_bus_register(struct bcma_bus *bus);
 void bcma_bus_unregister(struct bcma_bus *bus);
 int __init bcma_bus_early_register(struct bcma_bus *bus,
@@ -50,6 +51,10 @@ void bcma_chipco_serial_init(struct bcma_drv_cc *cc);
 extern struct platform_device bcma_pflash_dev;
 #endif /* CONFIG_BCMA_DRIVER_MIPS */
 
+/* driver_chipcommon_b.c */
+int bcma_core_chipcommon_b_init(struct bcma_drv_cc_b *ccb);
+void bcma_core_chipcommon_b_free(struct bcma_drv_cc_b *ccb);
+
 /* driver_chipcommon_pmu.c */
 u32 bcma_pmu_get_alp_clock(struct bcma_drv_cc *cc);
 u32 bcma_pmu_get_cpu_clock(struct bcma_drv_cc *cc);
@@ -84,6 +89,20 @@ extern int __init bcma_host_pci_init(void);
 extern void __exit bcma_host_pci_exit(void);
 #endif /* CONFIG_BCMA_HOST_PCI */
 
+/* host_soc.c */
+#if defined(CONFIG_BCMA_HOST_SOC) && defined(CONFIG_OF)
+extern int __init bcma_host_soc_register_driver(void);
+extern void __exit bcma_host_soc_unregister_driver(void);
+#else
+static inline int __init bcma_host_soc_register_driver(void)
+{
+	return 0;
+}
+static inline void __exit bcma_host_soc_unregister_driver(void)
+{
+}
+#endif /* CONFIG_BCMA_HOST_SOC && CONFIG_OF */
+
 /* driver_pci.c */
 u32 bcma_pcie_read(struct bcma_drv_pci *pc, u32 address);
 
diff --git a/drivers/bcma/driver_chipcommon.c b/drivers/bcma/driver_chipcommon.c
index b068f98920a8..19f679667ca4 100644
--- a/drivers/bcma/driver_chipcommon.c
+++ b/drivers/bcma/driver_chipcommon.c
@@ -339,7 +339,7 @@ void bcma_chipco_serial_init(struct bcma_drv_cc *cc)
 		return;
 	}
 
-	irq = bcma_core_irq(cc->core);
+	irq = bcma_core_irq(cc->core, 0);
 
 	/* Determine the registers of the UARTs */
 	cc->nr_serial_ports = (cc->capabilities & BCMA_CC_CAP_NRUART);
diff --git a/drivers/bcma/driver_chipcommon_b.c b/drivers/bcma/driver_chipcommon_b.c
new file mode 100644
index 000000000000..c20b5f4ff290
--- /dev/null
+++ b/drivers/bcma/driver_chipcommon_b.c
@@ -0,0 +1,61 @@
+/*
+ * Broadcom specific AMBA
+ * ChipCommon B Unit driver
+ *
+ * Copyright 2014, Hauke Mehrtens <hauke@hauke-m.de>
+ *
+ * Licensed under the GNU/GPL. See COPYING for details.
+ */
+
+#include "bcma_private.h"
+#include <linux/export.h>
+#include <linux/bcma/bcma.h>
+
+static bool bcma_wait_reg(struct bcma_bus *bus, void __iomem *addr, u32 mask,
+			  u32 value, int timeout)
+{
+	unsigned long deadline = jiffies + timeout;
+	u32 val;
+
+	do {
+		val = readl(addr);
+		if ((val & mask) == value)
+			return true;
+		cpu_relax();
+		udelay(10);
+	} while (!time_after_eq(jiffies, deadline));
+
+	bcma_err(bus, "Timeout waiting for register %p\n", addr);
+
+	return false;
+}
+
+void bcma_chipco_b_mii_write(struct bcma_drv_cc_b *ccb, u32 offset, u32 value)
+{
+	struct bcma_bus *bus = ccb->core->bus;
+
+	writel(offset, ccb->mii + 0x00);
+	bcma_wait_reg(bus, ccb->mii + 0x00, 0x0100, 0x0000, 100);
+	writel(value, ccb->mii + 0x04);
+	bcma_wait_reg(bus, ccb->mii + 0x00, 0x0100, 0x0000, 100);
+}
+EXPORT_SYMBOL_GPL(bcma_chipco_b_mii_write);
+
+int bcma_core_chipcommon_b_init(struct bcma_drv_cc_b *ccb)
+{
+	if (ccb->setup_done)
+		return 0;
+
+	ccb->setup_done = 1;
+	ccb->mii = ioremap_nocache(ccb->core->addr_s[1], BCMA_CORE_SIZE);
+	if (!ccb->mii)
+		return -ENOMEM;
+
+	return 0;
+}
+
+void bcma_core_chipcommon_b_free(struct bcma_drv_cc_b *ccb)
+{
+	if (ccb->mii)
+		iounmap(ccb->mii);
+}
diff --git a/drivers/bcma/driver_gpio.c b/drivers/bcma/driver_gpio.c
index aec9f850b4a8..598a6cd9028a 100644
--- a/drivers/bcma/driver_gpio.c
+++ b/drivers/bcma/driver_gpio.c
@@ -76,7 +76,7 @@ static void bcma_gpio_free(struct gpio_chip *chip, unsigned gpio)
 	bcma_chipco_gpio_pullup(cc, 1 << gpio, 0);
 }
 
-#if IS_BUILTIN(CONFIG_BCMA_HOST_SOC)
+#if IS_BUILTIN(CONFIG_BCM47XX)
 static int bcma_gpio_to_irq(struct gpio_chip *chip, unsigned gpio)
 {
 	struct bcma_drv_cc *cc = bcma_gpio_get_cc(chip);
@@ -152,7 +152,7 @@ static int bcma_gpio_irq_domain_init(struct bcma_drv_cc *cc)
 					 handle_simple_irq);
 	}
 
-	hwirq = bcma_core_irq(cc->core);
+	hwirq = bcma_core_irq(cc->core, 0);
 	err = request_irq(hwirq, bcma_gpio_irq_handler, IRQF_SHARED, "gpio",
 			  cc);
 	if (err)
@@ -183,7 +183,7 @@ static void bcma_gpio_irq_domain_exit(struct bcma_drv_cc *cc)
 		return;
 
 	bcma_cc_mask32(cc, BCMA_CC_IRQMASK, ~BCMA_CC_IRQ_GPIO);
-	free_irq(bcma_core_irq(cc->core), cc);
+	free_irq(bcma_core_irq(cc->core, 0), cc);
 	for (gpio = 0; gpio < chip->ngpio; gpio++) {
 		int irq = irq_find_mapping(cc->irq_domain, gpio);
 
@@ -215,9 +215,13 @@ int bcma_gpio_init(struct bcma_drv_cc *cc)
 	chip->set		= bcma_gpio_set_value;
 	chip->direction_input	= bcma_gpio_direction_input;
 	chip->direction_output	= bcma_gpio_direction_output;
-#if IS_BUILTIN(CONFIG_BCMA_HOST_SOC)
+#if IS_BUILTIN(CONFIG_BCM47XX)
 	chip->to_irq		= bcma_gpio_to_irq;
 #endif
+#if IS_BUILTIN(CONFIG_OF)
+	if (cc->core->bus->hosttype == BCMA_HOSTTYPE_SOC)
+		chip->of_node	= cc->core->dev.of_node;
+#endif
 	switch (cc->core->bus->chipinfo.id) {
 	case BCMA_CHIP_ID_BCM5357:
 	case BCMA_CHIP_ID_BCM53572:
@@ -251,5 +255,6 @@ int bcma_gpio_init(struct bcma_drv_cc *cc)
 int bcma_gpio_unregister(struct bcma_drv_cc *cc)
 {
 	bcma_gpio_irq_domain_exit(cc);
-	return gpiochip_remove(&cc->gpio);
+	gpiochip_remove(&cc->gpio);
+	return 0;
 }
diff --git a/drivers/bcma/driver_mips.c b/drivers/bcma/driver_mips.c
index 11115bbe115c..04faf6df959f 100644
--- a/drivers/bcma/driver_mips.c
+++ b/drivers/bcma/driver_mips.c
@@ -20,6 +20,17 @@
 #include <linux/serial_core.h>
 #include <linux/serial_reg.h>
 #include <linux/time.h>
+#ifdef CONFIG_BCM47XX
+#include <bcm47xx_nvram.h>
+#endif
+
+enum bcma_boot_dev {
+	BCMA_BOOT_DEV_UNK = 0,
+	BCMA_BOOT_DEV_ROM,
+	BCMA_BOOT_DEV_PARALLEL,
+	BCMA_BOOT_DEV_SERIAL,
+	BCMA_BOOT_DEV_NAND,
+};
 
 static const char * const part_probes[] = { "bcm47xxpart", NULL };
 
@@ -107,7 +118,7 @@ static u32 bcma_core_mips_irqflag(struct bcma_device *dev)
  * If disabled, 5 is returned.
  * If not supported, 6 is returned.
  */
-static unsigned int bcma_core_mips_irq(struct bcma_device *dev)
+unsigned int bcma_core_mips_irq(struct bcma_device *dev)
 {
 	struct bcma_device *mdev = dev->bus->drv_mips.core;
 	u32 irqflag;
@@ -125,13 +136,6 @@ static unsigned int bcma_core_mips_irq(struct bcma_device *dev)
 	return 5;
 }
 
-unsigned int bcma_core_irq(struct bcma_device *dev)
-{
-	unsigned int mips_irq = bcma_core_mips_irq(dev);
-	return mips_irq <= 4 ? mips_irq + 2 : 0;
-}
-EXPORT_SYMBOL(bcma_core_irq);
-
 static void bcma_core_mips_set_irq(struct bcma_device *dev, unsigned int irq)
 {
 	unsigned int oldirq = bcma_core_mips_irq(dev);
@@ -229,11 +233,51 @@ u32 bcma_cpu_clock(struct bcma_drv_mips *mcore)
 }
 EXPORT_SYMBOL(bcma_cpu_clock);
 
+static enum bcma_boot_dev bcma_boot_dev(struct bcma_bus *bus)
+{
+	struct bcma_drv_cc *cc = &bus->drv_cc;
+	u8 cc_rev = cc->core->id.rev;
+
+	if (cc_rev == 42) {
+		struct bcma_device *core;
+
+		core = bcma_find_core(bus, BCMA_CORE_NS_ROM);
+		if (core) {
+			switch (bcma_aread32(core, BCMA_IOST) &
+				BCMA_NS_ROM_IOST_BOOT_DEV_MASK) {
+			case BCMA_NS_ROM_IOST_BOOT_DEV_NOR:
+				return BCMA_BOOT_DEV_SERIAL;
+			case BCMA_NS_ROM_IOST_BOOT_DEV_NAND:
+				return BCMA_BOOT_DEV_NAND;
+			case BCMA_NS_ROM_IOST_BOOT_DEV_ROM:
+			default:
+				return BCMA_BOOT_DEV_ROM;
+			}
+		}
+	} else {
+		if (cc_rev == 38) {
+			if (cc->status & BCMA_CC_CHIPST_5357_NAND_BOOT)
+				return BCMA_BOOT_DEV_NAND;
+			else if (cc->status & BIT(5))
+				return BCMA_BOOT_DEV_ROM;
+		}
+
+		if ((cc->capabilities & BCMA_CC_CAP_FLASHT) ==
+		    BCMA_CC_FLASHT_PARA)
+			return BCMA_BOOT_DEV_PARALLEL;
+		else
+			return BCMA_BOOT_DEV_SERIAL;
+	}
+
+	return BCMA_BOOT_DEV_SERIAL;
+}
+
 static void bcma_core_mips_flash_detect(struct bcma_drv_mips *mcore)
 {
 	struct bcma_bus *bus = mcore->core->bus;
 	struct bcma_drv_cc *cc = &bus->drv_cc;
 	struct bcma_pflash *pflash = &cc->pflash;
+	enum bcma_boot_dev boot_dev;
 
 	switch (cc->capabilities & BCMA_CC_CAP_FLASHT) {
 	case BCMA_CC_FLASHT_STSER:
@@ -269,6 +313,26 @@ static void bcma_core_mips_flash_detect(struct bcma_drv_mips *mcore)
 			bcma_nflash_init(cc);
 		}
 	}
+
+	/* Determine flash type this SoC boots from */
+	boot_dev = bcma_boot_dev(bus);
+	switch (boot_dev) {
+	case BCMA_BOOT_DEV_PARALLEL:
+	case BCMA_BOOT_DEV_SERIAL:
+#ifdef CONFIG_BCM47XX
+		bcm47xx_nvram_init_from_mem(BCMA_SOC_FLASH2,
+					    BCMA_SOC_FLASH2_SZ);
+#endif
+		break;
+	case BCMA_BOOT_DEV_NAND:
+#ifdef CONFIG_BCM47XX
+		bcm47xx_nvram_init_from_mem(BCMA_SOC_FLASH1,
+					    BCMA_SOC_FLASH1_SZ);
+#endif
+		break;
+	default:
+		break;
+	}
 }
 
 void bcma_core_mips_early_init(struct bcma_drv_mips *mcore)
@@ -361,7 +425,7 @@ void bcma_core_mips_init(struct bcma_drv_mips *mcore)
 		break;
 	default:
 		list_for_each_entry(core, &bus->cores, list) {
-			core->irq = bcma_core_irq(core);
+			core->irq = bcma_core_irq(core, 0);
 		}
 		bcma_err(bus,
 			 "Unknown device (0x%x) found, can not configure IRQs\n",
diff --git a/drivers/bcma/driver_pci_host.c b/drivers/bcma/driver_pci_host.c
index c3d7b03c2fdc..c8a6b741967b 100644
--- a/drivers/bcma/driver_pci_host.c
+++ b/drivers/bcma/driver_pci_host.c
@@ -593,7 +593,7 @@ int bcma_core_pci_plat_dev_init(struct pci_dev *dev)
 	pr_info("PCI: Fixing up device %s\n", pci_name(dev));
 
 	/* Fix up interrupt lines */
-	dev->irq = bcma_core_irq(pc_host->pdev->core);
+	dev->irq = bcma_core_irq(pc_host->pdev->core, 0);
 	pci_write_config_byte(dev, PCI_INTERRUPT_LINE, dev->irq);
 
 	readrq = pcie_get_readrq(dev);
@@ -617,6 +617,6 @@ int bcma_core_pci_pcibios_map_irq(const struct pci_dev *dev)
 
 	pc_host = container_of(dev->bus->ops, struct bcma_drv_pci_host,
 			       pci_ops);
-	return bcma_core_irq(pc_host->pdev->core);
+	return bcma_core_irq(pc_host->pdev->core, 0);
 }
 EXPORT_SYMBOL(bcma_core_pci_pcibios_map_irq);
diff --git a/drivers/bcma/host_pci.c b/drivers/bcma/host_pci.c
index f032ed6dd459..cd9161a8b3a1 100644
--- a/drivers/bcma/host_pci.c
+++ b/drivers/bcma/host_pci.c
@@ -208,6 +208,9 @@ static int bcma_host_pci_probe(struct pci_dev *dev,
 	bus->boardinfo.vendor = bus->host_pci->subsystem_vendor;
 	bus->boardinfo.type = bus->host_pci->subsystem_device;
 
+	/* Initialize struct, detect chip */
+	bcma_init_bus(bus);
+
 	/* Register */
 	err = bcma_bus_register(bus);
 	if (err)
@@ -272,7 +275,7 @@ static SIMPLE_DEV_PM_OPS(bcma_pm_ops, bcma_host_pci_suspend,
 static const struct pci_device_id bcma_pci_bridge_tbl[] = {
 	{ PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, 0x0576) },
 	{ PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, 0x4313) },
-	{ PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, 43224) },
+	{ PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, 43224) },	/* 0xa8d8 */
 	{ PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, 0x4331) },
 	{ PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, 0x4353) },
 	{ PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, 0x4357) },
@@ -282,7 +285,8 @@ static const struct pci_device_id bcma_pci_bridge_tbl[] = {
 	{ PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, 0x43a9) },
 	{ PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, 0x43aa) },
 	{ PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, 0x4727) },
-	{ PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, 43227) },	/* 0xA8DB */
+	{ PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, 43227) },	/* 0xa8db, BCM43217 (sic!) */
+	{ PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, 43228) },	/* 0xa8dc */
 	{ 0, },
 };
 MODULE_DEVICE_TABLE(pci, bcma_pci_bridge_tbl);
diff --git a/drivers/bcma/host_soc.c b/drivers/bcma/host_soc.c
index 3475e600011a..335cbcfd945b 100644
--- a/drivers/bcma/host_soc.c
+++ b/drivers/bcma/host_soc.c
@@ -7,6 +7,9 @@
 
 #include "bcma_private.h"
 #include "scan.h"
+#include <linux/slab.h>
+#include <linux/module.h>
+#include <linux/of_address.h>
 #include <linux/bcma/bcma.h>
 #include <linux/bcma/bcma_soc.h>
 
@@ -134,12 +137,16 @@ static void bcma_host_soc_block_write(struct bcma_device *core,
 
 static u32 bcma_host_soc_aread32(struct bcma_device *core, u16 offset)
 {
+	if (WARN_ONCE(!core->io_wrap, "Accessed core has no wrapper/agent\n"))
+		return ~0;
 	return readl(core->io_wrap + offset);
 }
 
 static void bcma_host_soc_awrite32(struct bcma_device *core, u16 offset,
 				  u32 value)
 {
+	if (WARN_ONCE(!core->io_wrap, "Accessed core has no wrapper/agent\n"))
+		return;
 	writel(value, core->io_wrap + offset);
 }
 
@@ -161,7 +168,6 @@ static const struct bcma_host_ops bcma_host_soc_ops = {
 int __init bcma_host_soc_register(struct bcma_soc *soc)
 {
 	struct bcma_bus *bus = &soc->bus;
-	int err;
 
 	/* iomap only first core. We have to read some register on this core
 	 * to scan the bus.
@@ -173,11 +179,100 @@ int __init bcma_host_soc_register(struct bcma_soc *soc)
 	/* Host specific */
 	bus->hosttype = BCMA_HOSTTYPE_SOC;
 	bus->ops = &bcma_host_soc_ops;
+	bus->host_pdev = NULL;
 
-	/* Register */
+	/* Initialize struct, detect chip */
+	bcma_init_bus(bus);
+
+	return 0;
+}
+
+int __init bcma_host_soc_init(struct bcma_soc *soc)
+{
+	struct bcma_bus *bus = &soc->bus;
+	int err;
+
+	/* Scan bus and initialize it */
 	err = bcma_bus_early_register(bus, &soc->core_cc, &soc->core_mips);
 	if (err)
 		iounmap(bus->mmio);
 
 	return err;
 }
+
+#ifdef CONFIG_OF
+static int bcma_host_soc_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct device_node *np = dev->of_node;
+	struct bcma_bus *bus;
+	int err;
+
+	/* Alloc */
+	bus = devm_kzalloc(dev, sizeof(*bus), GFP_KERNEL);
+	if (!bus)
+		return -ENOMEM;
+
+	/* Map MMIO */
+	bus->mmio = of_iomap(np, 0);
+	if (!bus->mmio)
+		return -ENOMEM;
+
+	/* Host specific */
+	bus->hosttype = BCMA_HOSTTYPE_SOC;
+	bus->ops = &bcma_host_soc_ops;
+	bus->host_pdev = pdev;
+
+	/* Initialize struct, detect chip */
+	bcma_init_bus(bus);
+
+	/* Register */
+	err = bcma_bus_register(bus);
+	if (err)
+		goto err_unmap_mmio;
+
+	platform_set_drvdata(pdev, bus);
+
+	return err;
+
+err_unmap_mmio:
+	iounmap(bus->mmio);
+	return err;
+}
+
+static int bcma_host_soc_remove(struct platform_device *pdev)
+{
+	struct bcma_bus *bus = platform_get_drvdata(pdev);
+
+	bcma_bus_unregister(bus);
+	iounmap(bus->mmio);
+	platform_set_drvdata(pdev, NULL);
+
+	return 0;
+}
+
+static const struct of_device_id bcma_host_soc_of_match[] = {
+	{ .compatible = "brcm,bus-axi", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, bcma_host_soc_of_match);
+
+static struct platform_driver bcma_host_soc_driver = {
+	.driver = {
+		.name = "bcma-host-soc",
+		.of_match_table = bcma_host_soc_of_match,
+	},
+	.probe		= bcma_host_soc_probe,
+	.remove		= bcma_host_soc_remove,
+};
+
+int __init bcma_host_soc_register_driver(void)
+{
+	return platform_driver_register(&bcma_host_soc_driver);
+}
+
+void __exit bcma_host_soc_unregister_driver(void)
+{
+	platform_driver_unregister(&bcma_host_soc_driver);
+}
+#endif /* CONFIG_OF */
diff --git a/drivers/bcma/main.c b/drivers/bcma/main.c
index 0ff8d58831ef..534e1337766d 100644
--- a/drivers/bcma/main.c
+++ b/drivers/bcma/main.c
@@ -10,6 +10,8 @@
 #include <linux/platform_device.h>
 #include <linux/bcma/bcma.h>
 #include <linux/slab.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
 
 MODULE_DESCRIPTION("Broadcom's specific AMBA driver");
 MODULE_LICENSE("GPL");
@@ -120,16 +122,177 @@ static void bcma_release_core_dev(struct device *dev)
 	kfree(core);
 }
 
-static int bcma_register_cores(struct bcma_bus *bus)
+static bool bcma_is_core_needed_early(u16 core_id)
+{
+	switch (core_id) {
+	case BCMA_CORE_NS_NAND:
+	case BCMA_CORE_NS_QSPI:
+		return true;
+	}
+
+	return false;
+}
+
+#if defined(CONFIG_OF) && defined(CONFIG_OF_ADDRESS)
+static struct device_node *bcma_of_find_child_device(struct platform_device *parent,
+						     struct bcma_device *core)
+{
+	struct device_node *node;
+	u64 size;
+	const __be32 *reg;
+
+	if (!parent || !parent->dev.of_node)
+		return NULL;
+
+	for_each_child_of_node(parent->dev.of_node, node) {
+		reg = of_get_address(node, 0, &size, NULL);
+		if (!reg)
+			continue;
+		if (of_translate_address(node, reg) == core->addr)
+			return node;
+	}
+	return NULL;
+}
+
+static int bcma_of_irq_parse(struct platform_device *parent,
+			     struct bcma_device *core,
+			     struct of_phandle_args *out_irq, int num)
+{
+	__be32 laddr[1];
+	int rc;
+
+	if (core->dev.of_node) {
+		rc = of_irq_parse_one(core->dev.of_node, num, out_irq);
+		if (!rc)
+			return rc;
+	}
+
+	out_irq->np = parent->dev.of_node;
+	out_irq->args_count = 1;
+	out_irq->args[0] = num;
+
+	laddr[0] = cpu_to_be32(core->addr);
+	return of_irq_parse_raw(laddr, out_irq);
+}
+
+static unsigned int bcma_of_get_irq(struct platform_device *parent,
+				    struct bcma_device *core, int num)
+{
+	struct of_phandle_args out_irq;
+	int ret;
+
+	if (!parent || !parent->dev.of_node)
+		return 0;
+
+	ret = bcma_of_irq_parse(parent, core, &out_irq, num);
+	if (ret) {
+		bcma_debug(core->bus, "bcma_of_get_irq() failed with rc=%d\n",
+			   ret);
+		return 0;
+	}
+
+	return irq_create_of_mapping(&out_irq);
+}
+
+static void bcma_of_fill_device(struct platform_device *parent,
+				struct bcma_device *core)
+{
+	struct device_node *node;
+
+	node = bcma_of_find_child_device(parent, core);
+	if (node)
+		core->dev.of_node = node;
+
+	core->irq = bcma_of_get_irq(parent, core, 0);
+}
+#else
+static void bcma_of_fill_device(struct platform_device *parent,
+				struct bcma_device *core)
+{
+}
+static inline unsigned int bcma_of_get_irq(struct platform_device *parent,
+					   struct bcma_device *core, int num)
+{
+	return 0;
+}
+#endif /* CONFIG_OF */
+
+unsigned int bcma_core_irq(struct bcma_device *core, int num)
+{
+	struct bcma_bus *bus = core->bus;
+	unsigned int mips_irq;
+
+	switch (bus->hosttype) {
+	case BCMA_HOSTTYPE_PCI:
+		return bus->host_pci->irq;
+	case BCMA_HOSTTYPE_SOC:
+		if (bus->drv_mips.core && num == 0) {
+			mips_irq = bcma_core_mips_irq(core);
+			return mips_irq <= 4 ? mips_irq + 2 : 0;
+		}
+		if (bus->host_pdev)
+			return bcma_of_get_irq(bus->host_pdev, core, num);
+		return 0;
+	case BCMA_HOSTTYPE_SDIO:
+		return 0;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL(bcma_core_irq);
+
+void bcma_prepare_core(struct bcma_bus *bus, struct bcma_device *core)
+{
+	core->dev.release = bcma_release_core_dev;
+	core->dev.bus = &bcma_bus_type;
+	dev_set_name(&core->dev, "bcma%d:%d", bus->num, core->core_index);
+
+	switch (bus->hosttype) {
+	case BCMA_HOSTTYPE_PCI:
+		core->dev.parent = &bus->host_pci->dev;
+		core->dma_dev = &bus->host_pci->dev;
+		core->irq = bus->host_pci->irq;
+		break;
+	case BCMA_HOSTTYPE_SOC:
+		core->dev.dma_mask = &core->dev.coherent_dma_mask;
+		if (bus->host_pdev) {
+			core->dma_dev = &bus->host_pdev->dev;
+			core->dev.parent = &bus->host_pdev->dev;
+			bcma_of_fill_device(bus->host_pdev, core);
+		} else {
+			core->dma_dev = &core->dev;
+		}
+		break;
+	case BCMA_HOSTTYPE_SDIO:
+		break;
+	}
+}
+
+static void bcma_register_core(struct bcma_bus *bus, struct bcma_device *core)
+{
+	int err;
+
+	err = device_register(&core->dev);
+	if (err) {
+		bcma_err(bus, "Could not register dev for core 0x%03X\n",
+			 core->id.id);
+		put_device(&core->dev);
+		return;
+	}
+	core->dev_registered = true;
+}
+
+static int bcma_register_devices(struct bcma_bus *bus)
 {
 	struct bcma_device *core;
-	int err, dev_id = 0;
+	int err;
 
 	list_for_each_entry(core, &bus->cores, list) {
 		/* We support that cores ourself */
 		switch (core->id.id) {
 		case BCMA_CORE_4706_CHIPCOMMON:
 		case BCMA_CORE_CHIPCOMMON:
+		case BCMA_CORE_NS_CHIPCOMMON_B:
 		case BCMA_CORE_PCI:
 		case BCMA_CORE_PCIE:
 		case BCMA_CORE_PCIE2:
@@ -138,39 +301,16 @@ static int bcma_register_cores(struct bcma_bus *bus)
 			continue;
 		}
 
+		/* Early cores were already registered */
+		if (bcma_is_core_needed_early(core->id.id))
+			continue;
+
 		/* Only first GMAC core on BCM4706 is connected and working */
 		if (core->id.id == BCMA_CORE_4706_MAC_GBIT &&
 		    core->core_unit > 0)
 			continue;
 
-		core->dev.release = bcma_release_core_dev;
-		core->dev.bus = &bcma_bus_type;
-		dev_set_name(&core->dev, "bcma%d:%d", bus->num, dev_id);
-
-		switch (bus->hosttype) {
-		case BCMA_HOSTTYPE_PCI:
-			core->dev.parent = &bus->host_pci->dev;
-			core->dma_dev = &bus->host_pci->dev;
-			core->irq = bus->host_pci->irq;
-			break;
-		case BCMA_HOSTTYPE_SOC:
-			core->dev.dma_mask = &core->dev.coherent_dma_mask;
-			core->dma_dev = &core->dev;
-			break;
-		case BCMA_HOSTTYPE_SDIO:
-			break;
-		}
-
-		err = device_register(&core->dev);
-		if (err) {
-			bcma_err(bus,
-				 "Could not register dev for core 0x%03X\n",
-				 core->id.id);
-			put_device(&core->dev);
-			continue;
-		}
-		core->dev_registered = true;
-		dev_id++;
+		bcma_register_core(bus, core);
 	}
 
 #ifdef CONFIG_BCMA_DRIVER_MIPS
@@ -247,6 +387,12 @@ int bcma_bus_register(struct bcma_bus *bus)
 		bcma_core_chipcommon_early_init(&bus->drv_cc);
 	}
 
+	/* Cores providing flash access go before SPROM init */
+	list_for_each_entry(core, &bus->cores, list) {
+		if (bcma_is_core_needed_early(core->id.id))
+			bcma_register_core(bus, core);
+	}
+
 	/* Try to get SPROM */
 	err = bcma_sprom_get(bus);
 	if (err == -ENOENT) {
@@ -261,6 +407,13 @@ int bcma_bus_register(struct bcma_bus *bus)
 		bcma_core_chipcommon_init(&bus->drv_cc);
 	}
 
+	/* Init CC core */
+	core = bcma_find_core(bus, BCMA_CORE_NS_CHIPCOMMON_B);
+	if (core) {
+		bus->drv_cc_b.core = core;
+		bcma_core_chipcommon_b_init(&bus->drv_cc_b);
+	}
+
 	/* Init MIPS core */
 	core = bcma_find_core(bus, BCMA_CORE_MIPS_74K);
 	if (core) {
@@ -297,7 +450,7 @@ int bcma_bus_register(struct bcma_bus *bus)
 	}
 
 	/* Register found cores */
-	bcma_register_cores(bus);
+	bcma_register_devices(bus);
 
 	bcma_info(bus, "Bus registered\n");
 
@@ -315,6 +468,8 @@ void bcma_bus_unregister(struct bcma_bus *bus)
 	else if (err)
 		bcma_err(bus, "Can not unregister GPIO driver: %i\n", err);
 
+	bcma_core_chipcommon_b_free(&bus->drv_cc_b);
+
 	cores[0] = bcma_find_core(bus, BCMA_CORE_MIPS_74K);
 	cores[1] = bcma_find_core(bus, BCMA_CORE_PCIE);
 	cores[2] = bcma_find_core(bus, BCMA_CORE_4706_MAC_GBIT_COMMON);
@@ -334,8 +489,6 @@ int __init bcma_bus_early_register(struct bcma_bus *bus,
 	struct bcma_device *core;
 	struct bcma_device_id match;
 
-	bcma_init_bus(bus);
-
 	match.manuf = BCMA_MANUF_BCM;
 	match.id = bcma_cc_core_id(bus);
 	match.class = BCMA_CL_SIM;
@@ -494,6 +647,11 @@ static int __init bcma_modinit(void)
 	if (err)
 		return err;
 
+	err = bcma_host_soc_register_driver();
+	if (err) {
+		pr_err("SoC host initialization failed\n");
+		err = 0;
+	}
 #ifdef CONFIG_BCMA_HOST_PCI
 	err = bcma_host_pci_init();
 	if (err) {
@@ -511,6 +669,7 @@ static void __exit bcma_modexit(void)
 #ifdef CONFIG_BCMA_HOST_PCI
 	bcma_host_pci_exit();
 #endif
+	bcma_host_soc_unregister_driver();
 	bus_unregister(&bcma_bus_type);
 }
 module_exit(bcma_modexit)
diff --git a/drivers/bcma/scan.c b/drivers/bcma/scan.c
index b4764c6bcf17..917520776879 100644
--- a/drivers/bcma/scan.c
+++ b/drivers/bcma/scan.c
@@ -276,7 +276,7 @@ static int bcma_get_next_core(struct bcma_bus *bus, u32 __iomem **eromptr,
 			      struct bcma_device *core)
 {
 	u32 tmp;
-	u8 i, j;
+	u8 i, j, k;
 	s32 cia, cib;
 	u8 ports[2], wrappers[2];
 
@@ -314,6 +314,7 @@ static int bcma_get_next_core(struct bcma_bus *bus, u32 __iomem **eromptr,
 		/* Some specific cores don't need wrappers */
 		switch (core->id.id) {
 		case BCMA_CORE_4706_MAC_GBIT_COMMON:
+		case BCMA_CORE_NS_CHIPCOMMON_B:
 		/* Not used yet: case BCMA_CORE_OOB_ROUTER: */
 			break;
 		default:
@@ -367,6 +368,7 @@ static int bcma_get_next_core(struct bcma_bus *bus, u32 __iomem **eromptr,
 	core->addr = tmp;
 
 	/* get & parse slave ports */
+	k = 0;
 	for (i = 0; i < ports[1]; i++) {
 		for (j = 0; ; j++) {
 			tmp = bcma_erom_get_addr_desc(bus, eromptr,
@@ -376,9 +378,9 @@ static int bcma_get_next_core(struct bcma_bus *bus, u32 __iomem **eromptr,
 				/* pr_debug("erom: slave port %d "
 				 * "has %d descriptors\n", i, j); */
 				break;
-			} else {
-				if (i == 0 && j == 0)
-					core->addr1 = tmp;
+			} else if (k < ARRAY_SIZE(core->addr_s)) {
+				core->addr_s[k] = tmp;
+				k++;
 			}
 		}
 	}
@@ -421,10 +423,13 @@ static int bcma_get_next_core(struct bcma_bus *bus, u32 __iomem **eromptr,
 		core->io_addr = ioremap_nocache(core->addr, BCMA_CORE_SIZE);
 		if (!core->io_addr)
 			return -ENOMEM;
-		core->io_wrap = ioremap_nocache(core->wrap, BCMA_CORE_SIZE);
-		if (!core->io_wrap) {
-			iounmap(core->io_addr);
-			return -ENOMEM;
+		if (core->wrap) {
+			core->io_wrap = ioremap_nocache(core->wrap,
+							BCMA_CORE_SIZE);
+			if (!core->io_wrap) {
+				iounmap(core->io_addr);
+				return -ENOMEM;
+			}
 		}
 	}
 	return 0;
@@ -434,9 +439,7 @@ void bcma_init_bus(struct bcma_bus *bus)
 {
 	s32 tmp;
 	struct bcma_chipinfo *chipinfo = &(bus->chipinfo);
-
-	if (bus->init_done)
-		return;
+	char chip_id[8];
 
 	INIT_LIST_HEAD(&bus->cores);
 	bus->nr_cores = 0;
@@ -447,10 +450,11 @@ void bcma_init_bus(struct bcma_bus *bus)
 	chipinfo->id = (tmp & BCMA_CC_ID_ID) >> BCMA_CC_ID_ID_SHIFT;
 	chipinfo->rev = (tmp & BCMA_CC_ID_REV) >> BCMA_CC_ID_REV_SHIFT;
 	chipinfo->pkg = (tmp & BCMA_CC_ID_PKG) >> BCMA_CC_ID_PKG_SHIFT;
-	bcma_info(bus, "Found chip with id 0x%04X, rev 0x%02X and package 0x%02X\n",
-		  chipinfo->id, chipinfo->rev, chipinfo->pkg);
 
-	bus->init_done = true;
+	snprintf(chip_id, ARRAY_SIZE(chip_id),
+		 (chipinfo->id > 0x9999) ? "%d" : "0x%04X", chipinfo->id);
+	bcma_info(bus, "Found chip with id %s, rev 0x%02X and package 0x%02X\n",
+		  chip_id, chipinfo->rev, chipinfo->pkg);
 }
 
 int bcma_bus_scan(struct bcma_bus *bus)
@@ -460,8 +464,6 @@ int bcma_bus_scan(struct bcma_bus *bus)
 
 	int err, core_num = 0;
 
-	bcma_init_bus(bus);
-
 	erombase = bcma_scan_read32(bus, 0, BCMA_CC_EROM);
 	if (bus->hosttype == BCMA_HOSTTYPE_SOC) {
 		eromptr = ioremap_nocache(erombase, BCMA_CORE_SIZE);
@@ -503,6 +505,7 @@ int bcma_bus_scan(struct bcma_bus *bus)
 		bus->nr_cores++;
 		other_core = bcma_find_core_reverse(bus, core->id.id);
 		core->core_unit = (other_core == NULL) ? 0 : other_core->core_unit + 1;
+		bcma_prepare_core(bus, core);
 
 		bcma_info(bus, "Core %d found: %s (manuf 0x%03X, id 0x%03X, rev 0x%02X, class 0x%X)\n",
 			  core->core_index, bcma_device_name(&core->id),
diff --git a/drivers/block/amiflop.c b/drivers/block/amiflop.c
index 758da2287d9a..5fd50a284168 100644
--- a/drivers/block/amiflop.c
+++ b/drivers/block/amiflop.c
@@ -1864,7 +1864,6 @@ static int __exit amiga_floppy_remove(struct platform_device *pdev)
 static struct platform_driver amiga_floppy_driver = {
 	.driver   = {
 		.name	= "amiga-floppy",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/block/aoe/aoeblk.c b/drivers/block/aoe/aoeblk.c
index dd73e1ff1759..46c282fff104 100644
--- a/drivers/block/aoe/aoeblk.c
+++ b/drivers/block/aoe/aoeblk.c
@@ -395,7 +395,7 @@ aoeblk_gdalloc(void *vp)
 	WARN_ON(d->flags & DEVFL_TKILL);
 	WARN_ON(d->gd);
 	WARN_ON(d->flags & DEVFL_UP);
-	blk_queue_max_hw_sectors(q, BLK_DEF_MAX_SECTORS);
+	blk_queue_max_hw_sectors(q, 1024);
 	q->backing_dev_info.name = "aoe";
 	q->backing_dev_info.ra_pages = READ_AHEAD / PAGE_CACHE_SIZE;
 	d->bufpool = mp;
diff --git a/drivers/block/drbd/drbd_actlog.c b/drivers/block/drbd/drbd_actlog.c
index d26a3fa63688..1318e3217cb0 100644
--- a/drivers/block/drbd/drbd_actlog.c
+++ b/drivers/block/drbd/drbd_actlog.c
@@ -158,14 +158,14 @@ static int _drbd_md_sync_page_io(struct drbd_device *device,
 	if (bio_add_page(bio, device->md_io.page, size, 0) != size)
 		goto out;
 	bio->bi_private = device;
-	bio->bi_end_io = drbd_md_io_complete;
+	bio->bi_end_io = drbd_md_endio;
 	bio->bi_rw = rw;
 
 	if (!(rw & WRITE) && device->state.disk == D_DISKLESS && device->ldev == NULL)
 		/* special case, drbd_md_read() during drbd_adm_attach(): no get_ldev */
 		;
 	else if (!get_ldev_if_state(device, D_ATTACHING)) {
-		/* Corresponding put_ldev in drbd_md_io_complete() */
+		/* Corresponding put_ldev in drbd_md_endio() */
 		drbd_err(device, "ASSERT FAILED: get_ldev_if_state() == 1 in _drbd_md_sync_page_io()\n");
 		err = -ENODEV;
 		goto out;
@@ -827,8 +827,7 @@ static int update_sync_bits(struct drbd_device *device,
  *
  */
 int __drbd_change_sync(struct drbd_device *device, sector_t sector, int size,
-		enum update_sync_bits_mode mode,
-		const char *file, const unsigned int line)
+		enum update_sync_bits_mode mode)
 {
 	/* Is called from worker and receiver context _only_ */
 	unsigned long sbnr, ebnr, lbnr;
diff --git a/drivers/block/drbd/drbd_bitmap.c b/drivers/block/drbd/drbd_bitmap.c
index 426c97aef900..434c77dcc99e 100644
--- a/drivers/block/drbd/drbd_bitmap.c
+++ b/drivers/block/drbd/drbd_bitmap.c
@@ -941,7 +941,7 @@ static void drbd_bm_aio_ctx_destroy(struct kref *kref)
 }
 
 /* bv_page may be a copy, or may be the original */
-static void bm_async_io_complete(struct bio *bio, int error)
+static void drbd_bm_endio(struct bio *bio, int error)
 {
 	struct drbd_bm_aio_ctx *ctx = bio->bi_private;
 	struct drbd_device *device = ctx->device;
@@ -1027,7 +1027,7 @@ static void bm_page_io_async(struct drbd_bm_aio_ctx *ctx, int page_nr) __must_ho
 	 * according to api.  Do we want to assert that? */
 	bio_add_page(bio, page, len, 0);
 	bio->bi_private = ctx;
-	bio->bi_end_io = bm_async_io_complete;
+	bio->bi_end_io = drbd_bm_endio;
 
 	if (drbd_insert_fault(device, (rw & WRITE) ? DRBD_FAULT_MD_WR : DRBD_FAULT_MD_RD)) {
 		bio->bi_rw |= rw;
@@ -1125,7 +1125,7 @@ static int bm_rw(struct drbd_device *device, const unsigned int flags, unsigned
 	}
 
 	/*
-	 * We initialize ctx->in_flight to one to make sure bm_async_io_complete
+	 * We initialize ctx->in_flight to one to make sure drbd_bm_endio
 	 * will not set ctx->done early, and decrement / test it here.  If there
 	 * are still some bios in flight, we need to wait for them here.
 	 * If all IO is done already (or nothing had been submitted), there is
diff --git a/drivers/block/drbd/drbd_debugfs.c b/drivers/block/drbd/drbd_debugfs.c
index 5c20b18540b8..9a950022ff88 100644
--- a/drivers/block/drbd/drbd_debugfs.c
+++ b/drivers/block/drbd/drbd_debugfs.c
@@ -419,7 +419,7 @@ static int in_flight_summary_show(struct seq_file *m, void *pos)
 	return 0;
 }
 
-/* simple_positive(file->f_dentry) respectively debugfs_positive(),
+/* simple_positive(file->f_path.dentry) respectively debugfs_positive(),
  * but neither is "reachable" from here.
  * So we have our own inline version of it above.  :-( */
 static inline int debugfs_positive(struct dentry *dentry)
@@ -437,14 +437,14 @@ static int drbd_single_open(struct file *file, int (*show)(struct seq_file *, vo
 
 	/* Are we still linked,
 	 * or has debugfs_remove() already been called? */
-	parent = file->f_dentry->d_parent;
+	parent = file->f_path.dentry->d_parent;
 	/* not sure if this can happen: */
 	if (!parent || !parent->d_inode)
 		goto out;
 	/* serialize with d_delete() */
 	mutex_lock(&parent->d_inode->i_mutex);
 	/* Make sure the object is still alive */
-	if (debugfs_positive(file->f_dentry)
+	if (debugfs_positive(file->f_path.dentry)
 	&& kref_get_unless_zero(kref))
 		ret = 0;
 	mutex_unlock(&parent->d_inode->i_mutex);
@@ -695,7 +695,7 @@ static void resync_dump_detail(struct seq_file *m, struct lc_element *e)
 {
        struct bm_extent *bme = lc_entry(e, struct bm_extent, lce);
 
-       seq_printf(m, "%5d %s %s %s\n", bme->rs_left,
+       seq_printf(m, "%5d %s %s %s", bme->rs_left,
 		  test_bit(BME_NO_WRITES, &bme->flags) ? "NO_WRITES" : "---------",
 		  test_bit(BME_LOCKED, &bme->flags) ? "LOCKED" : "------",
 		  test_bit(BME_PRIORITY, &bme->flags) ? "PRIORITY" : "--------"
diff --git a/drivers/block/drbd/drbd_int.h b/drivers/block/drbd/drbd_int.h
index 1a000016ccdf..b905e9888b88 100644
--- a/drivers/block/drbd/drbd_int.h
+++ b/drivers/block/drbd/drbd_int.h
@@ -61,8 +61,6 @@
 # define __must_hold(x)
 #endif
 
-#define __no_warn(lock, stmt) do { __acquire(lock); stmt; __release(lock); } while (0)
-
 /* module parameter, defined in drbd_main.c */
 extern unsigned int minor_count;
 extern bool disable_sendpage;
@@ -1456,7 +1454,6 @@ extern int is_valid_ar_handle(struct drbd_request *, sector_t);
 
 
 /* drbd_nl.c */
-extern int drbd_msg_put_info(struct sk_buff *skb, const char *info);
 extern void drbd_suspend_io(struct drbd_device *device);
 extern void drbd_resume_io(struct drbd_device *device);
 extern char *ppsize(char *buf, unsigned long long size);
@@ -1483,7 +1480,7 @@ extern int drbd_khelper(struct drbd_device *device, char *cmd);
 
 /* drbd_worker.c */
 /* bi_end_io handlers */
-extern void drbd_md_io_complete(struct bio *bio, int error);
+extern void drbd_md_endio(struct bio *bio, int error);
 extern void drbd_peer_request_endio(struct bio *bio, int error);
 extern void drbd_request_endio(struct bio *bio, int error);
 extern int drbd_worker(struct drbd_thread *thi);
@@ -1560,52 +1557,31 @@ extern void drbd_set_recv_tcq(struct drbd_device *device, int tcq_enabled);
 extern void _drbd_clear_done_ee(struct drbd_device *device, struct list_head *to_be_freed);
 extern int drbd_connected(struct drbd_peer_device *);
 
-/* Yes, there is kernel_setsockopt, but only since 2.6.18.
- * So we have our own copy of it here. */
-static inline int drbd_setsockopt(struct socket *sock, int level, int optname,
-				  char *optval, int optlen)
-{
-	mm_segment_t oldfs = get_fs();
-	char __user *uoptval;
-	int err;
-
-	uoptval = (char __user __force *)optval;
-
-	set_fs(KERNEL_DS);
-	if (level == SOL_SOCKET)
-		err = sock_setsockopt(sock, level, optname, uoptval, optlen);
-	else
-		err = sock->ops->setsockopt(sock, level, optname, uoptval,
-					    optlen);
-	set_fs(oldfs);
-	return err;
-}
-
 static inline void drbd_tcp_cork(struct socket *sock)
 {
 	int val = 1;
-	(void) drbd_setsockopt(sock, SOL_TCP, TCP_CORK,
+	(void) kernel_setsockopt(sock, SOL_TCP, TCP_CORK,
 			(char*)&val, sizeof(val));
 }
 
 static inline void drbd_tcp_uncork(struct socket *sock)
 {
 	int val = 0;
-	(void) drbd_setsockopt(sock, SOL_TCP, TCP_CORK,
+	(void) kernel_setsockopt(sock, SOL_TCP, TCP_CORK,
 			(char*)&val, sizeof(val));
 }
 
 static inline void drbd_tcp_nodelay(struct socket *sock)
 {
 	int val = 1;
-	(void) drbd_setsockopt(sock, SOL_TCP, TCP_NODELAY,
+	(void) kernel_setsockopt(sock, SOL_TCP, TCP_NODELAY,
 			(char*)&val, sizeof(val));
 }
 
 static inline void drbd_tcp_quickack(struct socket *sock)
 {
 	int val = 2;
-	(void) drbd_setsockopt(sock, SOL_TCP, TCP_QUICKACK,
+	(void) kernel_setsockopt(sock, SOL_TCP, TCP_QUICKACK,
 			(char*)&val, sizeof(val));
 }
 
@@ -1664,14 +1640,13 @@ extern void drbd_advance_rs_marks(struct drbd_device *device, unsigned long stil
 
 enum update_sync_bits_mode { RECORD_RS_FAILED, SET_OUT_OF_SYNC, SET_IN_SYNC };
 extern int __drbd_change_sync(struct drbd_device *device, sector_t sector, int size,
-		enum update_sync_bits_mode mode,
-		const char *file, const unsigned int line);
+		enum update_sync_bits_mode mode);
 #define drbd_set_in_sync(device, sector, size) \
-	__drbd_change_sync(device, sector, size, SET_IN_SYNC, __FILE__, __LINE__)
+	__drbd_change_sync(device, sector, size, SET_IN_SYNC)
 #define drbd_set_out_of_sync(device, sector, size) \
-	__drbd_change_sync(device, sector, size, SET_OUT_OF_SYNC, __FILE__, __LINE__)
+	__drbd_change_sync(device, sector, size, SET_OUT_OF_SYNC)
 #define drbd_rs_failed_io(device, sector, size) \
-	__drbd_change_sync(device, sector, size, RECORD_RS_FAILED, __FILE__, __LINE__)
+	__drbd_change_sync(device, sector, size, RECORD_RS_FAILED)
 extern void drbd_al_shrink(struct drbd_device *device);
 extern int drbd_initialize_al(struct drbd_device *, void *);
 
@@ -2100,16 +2075,19 @@ static inline bool is_sync_state(enum drbd_conns connection_state)
 
 /**
  * get_ldev() - Increase the ref count on device->ldev. Returns 0 if there is no ldev
- * @M:		DRBD device.
+ * @_device:		DRBD device.
+ * @_min_state:		Minimum device state required for success.
  *
  * You have to call put_ldev() when finished working with device->ldev.
  */
-#define get_ldev(M) __cond_lock(local, _get_ldev_if_state(M,D_INCONSISTENT))
-#define get_ldev_if_state(M,MINS) __cond_lock(local, _get_ldev_if_state(M,MINS))
+#define get_ldev_if_state(_device, _min_state)				\
+	(_get_ldev_if_state((_device), (_min_state)) ?			\
+	 ({ __acquire(x); true; }) : false)
+#define get_ldev(_device) get_ldev_if_state(_device, D_INCONSISTENT)
 
 static inline void put_ldev(struct drbd_device *device)
 {
-	enum drbd_disk_state ds = device->state.disk;
+	enum drbd_disk_state disk_state = device->state.disk;
 	/* We must check the state *before* the atomic_dec becomes visible,
 	 * or we have a theoretical race where someone hitting zero,
 	 * while state still D_FAILED, will then see D_DISKLESS in the
@@ -2122,10 +2100,10 @@ static inline void put_ldev(struct drbd_device *device)
 	__release(local);
 	D_ASSERT(device, i >= 0);
 	if (i == 0) {
-		if (ds == D_DISKLESS)
+		if (disk_state == D_DISKLESS)
 			/* even internal references gone, safe to destroy */
 			drbd_device_post_work(device, DESTROY_DISK);
-		if (ds == D_FAILED)
+		if (disk_state == D_FAILED)
 			/* all application IO references gone. */
 			if (!test_and_set_bit(GOING_DISKLESS, &device->flags))
 				drbd_device_post_work(device, GO_DISKLESS);
diff --git a/drivers/block/drbd/drbd_interval.c b/drivers/block/drbd/drbd_interval.c
index 89c497c630b4..51b25ad85251 100644
--- a/drivers/block/drbd/drbd_interval.c
+++ b/drivers/block/drbd/drbd_interval.c
@@ -37,40 +37,8 @@ compute_subtree_last(struct drbd_interval *node)
 	return max;
 }
 
-static void augment_propagate(struct rb_node *rb, struct rb_node *stop)
-{
-	while (rb != stop) {
-		struct drbd_interval *node = rb_entry(rb, struct drbd_interval, rb);
-		sector_t subtree_last = compute_subtree_last(node);
-		if (node->end == subtree_last)
-			break;
-		node->end = subtree_last;
-		rb = rb_parent(&node->rb);
-	}
-}
-
-static void augment_copy(struct rb_node *rb_old, struct rb_node *rb_new)
-{
-	struct drbd_interval *old = rb_entry(rb_old, struct drbd_interval, rb);
-	struct drbd_interval *new = rb_entry(rb_new, struct drbd_interval, rb);
-
-	new->end = old->end;
-}
-
-static void augment_rotate(struct rb_node *rb_old, struct rb_node *rb_new)
-{
-	struct drbd_interval *old = rb_entry(rb_old, struct drbd_interval, rb);
-	struct drbd_interval *new = rb_entry(rb_new, struct drbd_interval, rb);
-
-	new->end = old->end;
-	old->end = compute_subtree_last(old);
-}
-
-static const struct rb_augment_callbacks augment_callbacks = {
-	augment_propagate,
-	augment_copy,
-	augment_rotate,
-};
+RB_DECLARE_CALLBACKS(static, augment_callbacks, struct drbd_interval, rb,
+		     sector_t, end, compute_subtree_last);
 
 /**
  * drbd_insert_interval  -  insert a new interval into a tree
@@ -79,6 +47,7 @@ bool
 drbd_insert_interval(struct rb_root *root, struct drbd_interval *this)
 {
 	struct rb_node **new = &root->rb_node, *parent = NULL;
+	sector_t this_end = this->sector + (this->size >> 9);
 
 	BUG_ON(!IS_ALIGNED(this->size, 512));
 
@@ -87,6 +56,8 @@ drbd_insert_interval(struct rb_root *root, struct drbd_interval *this)
 			rb_entry(*new, struct drbd_interval, rb);
 
 		parent = *new;
+		if (here->end < this_end)
+			here->end = this_end;
 		if (this->sector < here->sector)
 			new = &(*new)->rb_left;
 		else if (this->sector > here->sector)
@@ -99,6 +70,7 @@ drbd_insert_interval(struct rb_root *root, struct drbd_interval *this)
 			return false;
 	}
 
+	this->end = this_end;
 	rb_link_node(&this->rb, parent, new);
 	rb_insert_augmented(&this->rb, root, &augment_callbacks);
 	return true;
diff --git a/drivers/block/drbd/drbd_main.c b/drivers/block/drbd/drbd_main.c
index 9b465bb68487..1fc83427199c 100644
--- a/drivers/block/drbd/drbd_main.c
+++ b/drivers/block/drbd/drbd_main.c
@@ -1622,13 +1622,13 @@ int drbd_send_dblock(struct drbd_peer_device *peer_device, struct drbd_request *
 	struct drbd_socket *sock;
 	struct p_data *p;
 	unsigned int dp_flags = 0;
-	int dgs;
+	int digest_size;
 	int err;
 
 	sock = &peer_device->connection->data;
 	p = drbd_prepare_command(peer_device, sock);
-	dgs = peer_device->connection->integrity_tfm ?
-	      crypto_hash_digestsize(peer_device->connection->integrity_tfm) : 0;
+	digest_size = peer_device->connection->integrity_tfm ?
+		      crypto_hash_digestsize(peer_device->connection->integrity_tfm) : 0;
 
 	if (!p)
 		return -EIO;
@@ -1659,9 +1659,9 @@ int drbd_send_dblock(struct drbd_peer_device *peer_device, struct drbd_request *
 
 	/* our digest is still only over the payload.
 	 * TRIM does not carry any payload. */
-	if (dgs)
+	if (digest_size)
 		drbd_csum_bio(peer_device->connection->integrity_tfm, req->master_bio, p + 1);
-	err = __send_command(peer_device->connection, device->vnr, sock, P_DATA, sizeof(*p) + dgs, NULL, req->i.size);
+	err = __send_command(peer_device->connection, device->vnr, sock, P_DATA, sizeof(*p) + digest_size, NULL, req->i.size);
 	if (!err) {
 		/* For protocol A, we have to memcpy the payload into
 		 * socket buffers, as we may complete right away
@@ -1674,23 +1674,23 @@ int drbd_send_dblock(struct drbd_peer_device *peer_device, struct drbd_request *
 		 * out ok after sending on this side, but does not fit on the
 		 * receiving side, we sure have detected corruption elsewhere.
 		 */
-		if (!(req->rq_state & (RQ_EXP_RECEIVE_ACK | RQ_EXP_WRITE_ACK)) || dgs)
+		if (!(req->rq_state & (RQ_EXP_RECEIVE_ACK | RQ_EXP_WRITE_ACK)) || digest_size)
 			err = _drbd_send_bio(peer_device, req->master_bio);
 		else
 			err = _drbd_send_zc_bio(peer_device, req->master_bio);
 
 		/* double check digest, sometimes buffers have been modified in flight. */
-		if (dgs > 0 && dgs <= 64) {
+		if (digest_size > 0 && digest_size <= 64) {
 			/* 64 byte, 512 bit, is the largest digest size
 			 * currently supported in kernel crypto. */
 			unsigned char digest[64];
 			drbd_csum_bio(peer_device->connection->integrity_tfm, req->master_bio, digest);
-			if (memcmp(p + 1, digest, dgs)) {
+			if (memcmp(p + 1, digest, digest_size)) {
 				drbd_warn(device,
 					"Digest mismatch, buffer modified by upper layers during write: %llus +%u\n",
 					(unsigned long long)req->i.sector, req->i.size);
 			}
-		} /* else if (dgs > 64) {
+		} /* else if (digest_size > 64) {
 		     ... Be noisy about digest too large ...
 		} */
 	}
@@ -1711,13 +1711,13 @@ int drbd_send_block(struct drbd_peer_device *peer_device, enum drbd_packet cmd,
 	struct drbd_socket *sock;
 	struct p_data *p;
 	int err;
-	int dgs;
+	int digest_size;
 
 	sock = &peer_device->connection->data;
 	p = drbd_prepare_command(peer_device, sock);
 
-	dgs = peer_device->connection->integrity_tfm ?
-	      crypto_hash_digestsize(peer_device->connection->integrity_tfm) : 0;
+	digest_size = peer_device->connection->integrity_tfm ?
+		      crypto_hash_digestsize(peer_device->connection->integrity_tfm) : 0;
 
 	if (!p)
 		return -EIO;
@@ -1725,9 +1725,9 @@ int drbd_send_block(struct drbd_peer_device *peer_device, enum drbd_packet cmd,
 	p->block_id = peer_req->block_id;
 	p->seq_num = 0;  /* unused */
 	p->dp_flags = 0;
-	if (dgs)
+	if (digest_size)
 		drbd_csum_ee(peer_device->connection->integrity_tfm, peer_req, p + 1);
-	err = __send_command(peer_device->connection, device->vnr, sock, cmd, sizeof(*p) + dgs, NULL, peer_req->i.size);
+	err = __send_command(peer_device->connection, device->vnr, sock, cmd, sizeof(*p) + digest_size, NULL, peer_req->i.size);
 	if (!err)
 		err = _drbd_send_zc_ee(peer_device, peer_req);
 	mutex_unlock(&sock->mutex);  /* locked by drbd_prepare_command() */
@@ -2532,10 +2532,6 @@ int set_resource_options(struct drbd_resource *resource, struct res_opts *res_op
 
 	if (!zalloc_cpumask_var(&new_cpu_mask, GFP_KERNEL))
 		return -ENOMEM;
-		/*
-		retcode = ERR_NOMEM;
-		drbd_msg_put_info("unable to allocate cpumask");
-		*/
 
 	/* silently ignore cpu mask on UP kernel */
 	if (nr_cpu_ids > 1 && res_opts->cpu_mask[0] != 0) {
@@ -2731,7 +2727,7 @@ enum drbd_ret_code drbd_create_device(struct drbd_config_context *adm_ctx, unsig
 
 	device = minor_to_device(minor);
 	if (device)
-		return ERR_MINOR_EXISTS;
+		return ERR_MINOR_OR_VOLUME_EXISTS;
 
 	/* GFP_KERNEL, we are outside of all write-out paths */
 	device = kzalloc(sizeof(struct drbd_device), GFP_KERNEL);
@@ -2793,20 +2789,16 @@ enum drbd_ret_code drbd_create_device(struct drbd_config_context *adm_ctx, unsig
 
 	id = idr_alloc(&drbd_devices, device, minor, minor + 1, GFP_KERNEL);
 	if (id < 0) {
-		if (id == -ENOSPC) {
-			err = ERR_MINOR_EXISTS;
-			drbd_msg_put_info(adm_ctx->reply_skb, "requested minor exists already");
-		}
+		if (id == -ENOSPC)
+			err = ERR_MINOR_OR_VOLUME_EXISTS;
 		goto out_no_minor_idr;
 	}
 	kref_get(&device->kref);
 
 	id = idr_alloc(&resource->devices, device, vnr, vnr + 1, GFP_KERNEL);
 	if (id < 0) {
-		if (id == -ENOSPC) {
-			err = ERR_MINOR_EXISTS;
-			drbd_msg_put_info(adm_ctx->reply_skb, "requested minor exists already");
-		}
+		if (id == -ENOSPC)
+			err = ERR_MINOR_OR_VOLUME_EXISTS;
 		goto out_idr_remove_minor;
 	}
 	kref_get(&device->kref);
@@ -2825,10 +2817,8 @@ enum drbd_ret_code drbd_create_device(struct drbd_config_context *adm_ctx, unsig
 
 		id = idr_alloc(&connection->peer_devices, peer_device, vnr, vnr + 1, GFP_KERNEL);
 		if (id < 0) {
-			if (id == -ENOSPC) {
+			if (id == -ENOSPC)
 				err = ERR_INVALID_REQUEST;
-				drbd_msg_put_info(adm_ctx->reply_skb, "requested volume exists already");
-			}
 			goto out_idr_remove_from_resource;
 		}
 		kref_get(&connection->kref);
@@ -2836,7 +2826,6 @@ enum drbd_ret_code drbd_create_device(struct drbd_config_context *adm_ctx, unsig
 
 	if (init_submitter(device)) {
 		err = ERR_NOMEM;
-		drbd_msg_put_info(adm_ctx->reply_skb, "unable to create submit workqueue");
 		goto out_idr_remove_vol;
 	}
 
diff --git a/drivers/block/drbd/drbd_nl.c b/drivers/block/drbd/drbd_nl.c
index 1cd47df44bda..74df8cfad414 100644
--- a/drivers/block/drbd/drbd_nl.c
+++ b/drivers/block/drbd/drbd_nl.c
@@ -92,7 +92,7 @@ static void drbd_adm_send_reply(struct sk_buff *skb, struct genl_info *info)
 
 /* Used on a fresh "drbd_adm_prepare"d reply_skb, this cannot fail: The only
  * reason it could fail was no space in skb, and there are 4k available. */
-int drbd_msg_put_info(struct sk_buff *skb, const char *info)
+static int drbd_msg_put_info(struct sk_buff *skb, const char *info)
 {
 	struct nlattr *nla;
 	int err = -EMSGSIZE;
@@ -588,7 +588,7 @@ drbd_set_role(struct drbd_device *const device, enum drbd_role new_role, int for
 	val.i  = 0; val.role  = new_role;
 
 	while (try++ < max_tries) {
-		rv = _drbd_request_state(device, mask, val, CS_WAIT_COMPLETE);
+		rv = _drbd_request_state_holding_state_mutex(device, mask, val, CS_WAIT_COMPLETE);
 
 		/* in case we first succeeded to outdate,
 		 * but now suddenly could establish a connection */
@@ -2052,7 +2052,7 @@ check_net_options(struct drbd_connection *connection, struct net_conf *new_net_c
 	rv = _check_net_options(connection, rcu_dereference(connection->net_conf), new_net_conf);
 	rcu_read_unlock();
 
-	/* connection->volumes protected by genl_lock() here */
+	/* connection->peer_devices protected by genl_lock() here */
 	idr_for_each_entry(&connection->peer_devices, peer_device, i) {
 		struct drbd_device *device = peer_device->device;
 		if (!device->bitmap) {
@@ -3483,7 +3483,7 @@ int drbd_adm_new_minor(struct sk_buff *skb, struct genl_info *info)
 	 * that first_peer_device(device)->connection and device->vnr match the request. */
 	if (adm_ctx.device) {
 		if (info->nlhdr->nlmsg_flags & NLM_F_EXCL)
-			retcode = ERR_MINOR_EXISTS;
+			retcode = ERR_MINOR_OR_VOLUME_EXISTS;
 		/* else: still NO_ERROR */
 		goto out;
 	}
@@ -3530,6 +3530,27 @@ out:
 	return 0;
 }
 
+static int adm_del_resource(struct drbd_resource *resource)
+{
+	struct drbd_connection *connection;
+
+	for_each_connection(connection, resource) {
+		if (connection->cstate > C_STANDALONE)
+			return ERR_NET_CONFIGURED;
+	}
+	if (!idr_is_empty(&resource->devices))
+		return ERR_RES_IN_USE;
+
+	list_del_rcu(&resource->resources);
+	/* Make sure all threads have actually stopped: state handling only
+	 * does drbd_thread_stop_nowait(). */
+	list_for_each_entry(connection, &resource->connections, connections)
+		drbd_thread_stop(&connection->worker);
+	synchronize_rcu();
+	drbd_free_resource(resource);
+	return NO_ERROR;
+}
+
 int drbd_adm_down(struct sk_buff *skb, struct genl_info *info)
 {
 	struct drbd_config_context adm_ctx;
@@ -3575,14 +3596,6 @@ int drbd_adm_down(struct sk_buff *skb, struct genl_info *info)
 		}
 	}
 
-	/* If we reach this, all volumes (of this connection) are Secondary,
-	 * Disconnected, Diskless, aka Unconfigured. Make sure all threads have
-	 * actually stopped, state handling only does drbd_thread_stop_nowait(). */
-	for_each_connection(connection, resource)
-		drbd_thread_stop(&connection->worker);
-
-	/* Now, nothing can fail anymore */
-
 	/* delete volumes */
 	idr_for_each_entry(&resource->devices, device, i) {
 		retcode = adm_del_minor(device);
@@ -3593,10 +3606,7 @@ int drbd_adm_down(struct sk_buff *skb, struct genl_info *info)
 		}
 	}
 
-	list_del_rcu(&resource->resources);
-	synchronize_rcu();
-	drbd_free_resource(resource);
-	retcode = NO_ERROR;
+	retcode = adm_del_resource(resource);
 out:
 	mutex_unlock(&resource->adm_mutex);
 finish:
@@ -3608,7 +3618,6 @@ int drbd_adm_del_resource(struct sk_buff *skb, struct genl_info *info)
 {
 	struct drbd_config_context adm_ctx;
 	struct drbd_resource *resource;
-	struct drbd_connection *connection;
 	enum drbd_ret_code retcode;
 
 	retcode = drbd_adm_prepare(&adm_ctx, skb, info, DRBD_ADM_NEED_RESOURCE);
@@ -3616,27 +3625,10 @@ int drbd_adm_del_resource(struct sk_buff *skb, struct genl_info *info)
 		return retcode;
 	if (retcode != NO_ERROR)
 		goto finish;
-
 	resource = adm_ctx.resource;
-	mutex_lock(&resource->adm_mutex);
-	for_each_connection(connection, resource) {
-		if (connection->cstate > C_STANDALONE) {
-			retcode = ERR_NET_CONFIGURED;
-			goto out;
-		}
-	}
-	if (!idr_is_empty(&resource->devices)) {
-		retcode = ERR_RES_IN_USE;
-		goto out;
-	}
 
-	list_del_rcu(&resource->resources);
-	for_each_connection(connection, resource)
-		drbd_thread_stop(&connection->worker);
-	synchronize_rcu();
-	drbd_free_resource(resource);
-	retcode = NO_ERROR;
-out:
+	mutex_lock(&resource->adm_mutex);
+	retcode = adm_del_resource(resource);
 	mutex_unlock(&resource->adm_mutex);
 finish:
 	drbd_adm_finish(&adm_ctx, info, retcode);
diff --git a/drivers/block/drbd/drbd_proc.c b/drivers/block/drbd/drbd_proc.c
index 06e6147c7601..3b10fa6cb039 100644
--- a/drivers/block/drbd/drbd_proc.c
+++ b/drivers/block/drbd/drbd_proc.c
@@ -142,10 +142,12 @@ static void drbd_syncer_progress(struct drbd_device *device, struct seq_file *se
 			    (unsigned long) Bit2KB(rs_left >> 10),
 			    (unsigned long) Bit2KB(rs_total >> 10));
 	else
-		seq_printf(seq, "(%lu/%lu)K\n\t",
+		seq_printf(seq, "(%lu/%lu)K",
 			    (unsigned long) Bit2KB(rs_left),
 			    (unsigned long) Bit2KB(rs_total));
 
+	seq_printf(seq, "\n\t");
+
 	/* see drivers/md/md.c
 	 * We do not want to overflow, so the order of operands and
 	 * the * 100 / 100 trick are important. We do a +1 to be
diff --git a/drivers/block/drbd/drbd_receiver.c b/drivers/block/drbd/drbd_receiver.c
index 9342b8da73ab..d169b4a79267 100644
--- a/drivers/block/drbd/drbd_receiver.c
+++ b/drivers/block/drbd/drbd_receiver.c
@@ -1371,9 +1371,9 @@ int drbd_submit_peer_request(struct drbd_device *device,
 	struct bio *bio;
 	struct page *page = peer_req->pages;
 	sector_t sector = peer_req->i.sector;
-	unsigned ds = peer_req->i.size;
+	unsigned data_size = peer_req->i.size;
 	unsigned n_bios = 0;
-	unsigned nr_pages = (ds + PAGE_SIZE -1) >> PAGE_SHIFT;
+	unsigned nr_pages = (data_size + PAGE_SIZE -1) >> PAGE_SHIFT;
 	int err = -ENOMEM;
 
 	if (peer_req->flags & EE_IS_TRIM_USE_ZEROOUT) {
@@ -1388,7 +1388,7 @@ int drbd_submit_peer_request(struct drbd_device *device,
 		list_add_tail(&peer_req->w.list, &device->active_ee);
 		spin_unlock_irq(&device->resource->req_lock);
 		if (blkdev_issue_zeroout(device->ldev->backing_bdev,
-			sector, ds >> 9, GFP_NOIO))
+			sector, data_size >> 9, GFP_NOIO))
 			peer_req->flags |= EE_WAS_ERROR;
 		drbd_endio_write_sec_final(peer_req);
 		return 0;
@@ -1426,12 +1426,12 @@ next_bio:
 	++n_bios;
 
 	if (rw & REQ_DISCARD) {
-		bio->bi_iter.bi_size = ds;
+		bio->bi_iter.bi_size = data_size;
 		goto submit;
 	}
 
 	page_chain_for_each(page) {
-		unsigned len = min_t(unsigned, ds, PAGE_SIZE);
+		unsigned len = min_t(unsigned, data_size, PAGE_SIZE);
 		if (!bio_add_page(bio, page, len, 0)) {
 			/* A single page must always be possible!
 			 * But in case it fails anyways,
@@ -1446,11 +1446,11 @@ next_bio:
 			}
 			goto next_bio;
 		}
-		ds -= len;
+		data_size -= len;
 		sector += len >> 9;
 		--nr_pages;
 	}
-	D_ASSERT(device, ds == 0);
+	D_ASSERT(device, data_size == 0);
 submit:
 	D_ASSERT(device, page == NULL);
 
@@ -1591,24 +1591,24 @@ read_in_block(struct drbd_peer_device *peer_device, u64 id, sector_t sector,
 	const sector_t capacity = drbd_get_capacity(device->this_bdev);
 	struct drbd_peer_request *peer_req;
 	struct page *page;
-	int dgs, ds, err;
-	unsigned int data_size = pi->size;
+	int digest_size, err;
+	unsigned int data_size = pi->size, ds;
 	void *dig_in = peer_device->connection->int_dig_in;
 	void *dig_vv = peer_device->connection->int_dig_vv;
 	unsigned long *data;
 	struct p_trim *trim = (pi->cmd == P_TRIM) ? pi->data : NULL;
 
-	dgs = 0;
+	digest_size = 0;
 	if (!trim && peer_device->connection->peer_integrity_tfm) {
-		dgs = crypto_hash_digestsize(peer_device->connection->peer_integrity_tfm);
+		digest_size = crypto_hash_digestsize(peer_device->connection->peer_integrity_tfm);
 		/*
 		 * FIXME: Receive the incoming digest into the receive buffer
 		 *	  here, together with its struct p_data?
 		 */
-		err = drbd_recv_all_warn(peer_device->connection, dig_in, dgs);
+		err = drbd_recv_all_warn(peer_device->connection, dig_in, digest_size);
 		if (err)
 			return NULL;
-		data_size -= dgs;
+		data_size -= digest_size;
 	}
 
 	if (trim) {
@@ -1661,16 +1661,16 @@ read_in_block(struct drbd_peer_device *peer_device, u64 id, sector_t sector,
 		ds -= len;
 	}
 
-	if (dgs) {
+	if (digest_size) {
 		drbd_csum_ee(peer_device->connection->peer_integrity_tfm, peer_req, dig_vv);
-		if (memcmp(dig_in, dig_vv, dgs)) {
+		if (memcmp(dig_in, dig_vv, digest_size)) {
 			drbd_err(device, "Digest integrity check FAILED: %llus +%u\n",
 				(unsigned long long)sector, data_size);
 			drbd_free_peer_req(device, peer_req);
 			return NULL;
 		}
 	}
-	device->recv_cnt += data_size>>9;
+	device->recv_cnt += data_size >> 9;
 	return peer_req;
 }
 
@@ -1708,17 +1708,17 @@ static int recv_dless_read(struct drbd_peer_device *peer_device, struct drbd_req
 	struct bio_vec bvec;
 	struct bvec_iter iter;
 	struct bio *bio;
-	int dgs, err, expect;
+	int digest_size, err, expect;
 	void *dig_in = peer_device->connection->int_dig_in;
 	void *dig_vv = peer_device->connection->int_dig_vv;
 
-	dgs = 0;
+	digest_size = 0;
 	if (peer_device->connection->peer_integrity_tfm) {
-		dgs = crypto_hash_digestsize(peer_device->connection->peer_integrity_tfm);
-		err = drbd_recv_all_warn(peer_device->connection, dig_in, dgs);
+		digest_size = crypto_hash_digestsize(peer_device->connection->peer_integrity_tfm);
+		err = drbd_recv_all_warn(peer_device->connection, dig_in, digest_size);
 		if (err)
 			return err;
-		data_size -= dgs;
+		data_size -= digest_size;
 	}
 
 	/* optimistically update recv_cnt.  if receiving fails below,
@@ -1738,9 +1738,9 @@ static int recv_dless_read(struct drbd_peer_device *peer_device, struct drbd_req
 		data_size -= expect;
 	}
 
-	if (dgs) {
+	if (digest_size) {
 		drbd_csum_bio(peer_device->connection->peer_integrity_tfm, bio, dig_vv);
-		if (memcmp(dig_in, dig_vv, dgs)) {
+		if (memcmp(dig_in, dig_vv, digest_size)) {
 			drbd_err(peer_device, "Digest integrity check FAILED. Broken NICs?\n");
 			return -EINVAL;
 		}
@@ -2482,7 +2482,7 @@ bool drbd_rs_c_min_rate_throttle(struct drbd_device *device)
 			atomic_read(&device->rs_sect_ev);
 
 	if (atomic_read(&device->ap_actlog_cnt)
-	    || !device->rs_last_events || curr_events - device->rs_last_events > 64) {
+	    || curr_events - device->rs_last_events > 64) {
 		unsigned long rs_left;
 		int i;
 
@@ -5561,6 +5561,7 @@ int drbd_asender(struct drbd_thread *thi)
 		 * rv <  expected: "woken" by signal during receive
 		 * rv == 0	 : "connection shut down by peer"
 		 */
+received_more:
 		if (likely(rv > 0)) {
 			received += rv;
 			buf	 += rv;
@@ -5636,6 +5637,11 @@ int drbd_asender(struct drbd_thread *thi)
 			expect	 = header_size;
 			cmd	 = NULL;
 		}
+		if (test_bit(SEND_PING, &connection->flags))
+			continue;
+		rv = drbd_recv_short(connection->meta.socket, buf, expect-received, MSG_DONTWAIT);
+		if (rv > 0)
+			goto received_more;
 	}
 
 	if (0) {
diff --git a/drivers/block/drbd/drbd_req.c b/drivers/block/drbd/drbd_req.c
index c67717d572d1..34f2f0ba409b 100644
--- a/drivers/block/drbd/drbd_req.c
+++ b/drivers/block/drbd/drbd_req.c
@@ -36,29 +36,15 @@ static bool drbd_may_do_local_read(struct drbd_device *device, sector_t sector,
 /* Update disk stats at start of I/O request */
 static void _drbd_start_io_acct(struct drbd_device *device, struct drbd_request *req)
 {
-	const int rw = bio_data_dir(req->master_bio);
-	int cpu;
-	cpu = part_stat_lock();
-	part_round_stats(cpu, &device->vdisk->part0);
-	part_stat_inc(cpu, &device->vdisk->part0, ios[rw]);
-	part_stat_add(cpu, &device->vdisk->part0, sectors[rw], req->i.size >> 9);
-	(void) cpu; /* The macro invocations above want the cpu argument, I do not like
-		       the compiler warning about cpu only assigned but never used... */
-	part_inc_in_flight(&device->vdisk->part0, rw);
-	part_stat_unlock();
+	generic_start_io_acct(bio_data_dir(req->master_bio), req->i.size >> 9,
+			      &device->vdisk->part0);
 }
 
 /* Update disk stats when completing request upwards */
 static void _drbd_end_io_acct(struct drbd_device *device, struct drbd_request *req)
 {
-	int rw = bio_data_dir(req->master_bio);
-	unsigned long duration = jiffies - req->start_jif;
-	int cpu;
-	cpu = part_stat_lock();
-	part_stat_add(cpu, &device->vdisk->part0, ticks[rw], duration);
-	part_round_stats(cpu, &device->vdisk->part0);
-	part_dec_in_flight(&device->vdisk->part0, rw);
-	part_stat_unlock();
+	generic_end_io_acct(bio_data_dir(req->master_bio),
+			    &device->vdisk->part0, req->start_jif);
 }
 
 static struct drbd_request *drbd_req_new(struct drbd_device *device,
@@ -1545,6 +1531,7 @@ int drbd_merge_bvec(struct request_queue *q, struct bvec_merge_data *bvm, struct
 		struct request_queue * const b =
 			device->ldev->backing_bdev->bd_disk->queue;
 		if (b->merge_bvec_fn) {
+			bvm->bi_bdev = device->ldev->backing_bdev;
 			backing_limit = b->merge_bvec_fn(b, bvm, bvec);
 			limit = min(limit, backing_limit);
 		}
@@ -1628,7 +1615,7 @@ void request_timer_fn(unsigned long data)
 		 time_after(now, req_peer->pre_send_jif + ent) &&
 		!time_in_range(now, connection->last_reconnect_jif, connection->last_reconnect_jif + ent)) {
 		drbd_warn(device, "Remote failed to finish a request within ko-count * timeout\n");
-		_drbd_set_state(_NS(device, conn, C_TIMEOUT), CS_VERBOSE | CS_HARD, NULL);
+		_conn_request_state(connection, NS(conn, C_TIMEOUT), CS_VERBOSE | CS_HARD);
 	}
 	if (dt && oldest_submit_jif != now &&
 		 time_after(now, oldest_submit_jif + dt) &&
@@ -1645,6 +1632,6 @@ void request_timer_fn(unsigned long data)
 		? oldest_submit_jif + dt : now + et;
 	nt = time_before(ent, dt) ? ent : dt;
 out:
-	spin_unlock_irq(&connection->resource->req_lock);
+	spin_unlock_irq(&device->resource->req_lock);
 	mod_timer(&device->request_timer, nt);
 }
diff --git a/drivers/block/drbd/drbd_state.c b/drivers/block/drbd/drbd_state.c
index c35c0f001bb7..2d7dd269b6a8 100644
--- a/drivers/block/drbd/drbd_state.c
+++ b/drivers/block/drbd/drbd_state.c
@@ -136,50 +136,50 @@ enum drbd_role conn_highest_peer(struct drbd_connection *connection)
 
 enum drbd_disk_state conn_highest_disk(struct drbd_connection *connection)
 {
-	enum drbd_disk_state ds = D_DISKLESS;
+	enum drbd_disk_state disk_state = D_DISKLESS;
 	struct drbd_peer_device *peer_device;
 	int vnr;
 
 	rcu_read_lock();
 	idr_for_each_entry(&connection->peer_devices, peer_device, vnr) {
 		struct drbd_device *device = peer_device->device;
-		ds = max_t(enum drbd_disk_state, ds, device->state.disk);
+		disk_state = max_t(enum drbd_disk_state, disk_state, device->state.disk);
 	}
 	rcu_read_unlock();
 
-	return ds;
+	return disk_state;
 }
 
 enum drbd_disk_state conn_lowest_disk(struct drbd_connection *connection)
 {
-	enum drbd_disk_state ds = D_MASK;
+	enum drbd_disk_state disk_state = D_MASK;
 	struct drbd_peer_device *peer_device;
 	int vnr;
 
 	rcu_read_lock();
 	idr_for_each_entry(&connection->peer_devices, peer_device, vnr) {
 		struct drbd_device *device = peer_device->device;
-		ds = min_t(enum drbd_disk_state, ds, device->state.disk);
+		disk_state = min_t(enum drbd_disk_state, disk_state, device->state.disk);
 	}
 	rcu_read_unlock();
 
-	return ds;
+	return disk_state;
 }
 
 enum drbd_disk_state conn_highest_pdsk(struct drbd_connection *connection)
 {
-	enum drbd_disk_state ds = D_DISKLESS;
+	enum drbd_disk_state disk_state = D_DISKLESS;
 	struct drbd_peer_device *peer_device;
 	int vnr;
 
 	rcu_read_lock();
 	idr_for_each_entry(&connection->peer_devices, peer_device, vnr) {
 		struct drbd_device *device = peer_device->device;
-		ds = max_t(enum drbd_disk_state, ds, device->state.pdsk);
+		disk_state = max_t(enum drbd_disk_state, disk_state, device->state.pdsk);
 	}
 	rcu_read_unlock();
 
-	return ds;
+	return disk_state;
 }
 
 enum drbd_conns conn_lowest_conn(struct drbd_connection *connection)
@@ -215,6 +215,18 @@ static bool no_peer_wf_report_params(struct drbd_connection *connection)
 	return rv;
 }
 
+static void wake_up_all_devices(struct drbd_connection *connection)
+{
+	struct drbd_peer_device *peer_device;
+	int vnr;
+
+	rcu_read_lock();
+	idr_for_each_entry(&connection->peer_devices, peer_device, vnr)
+		wake_up(&peer_device->device->state_wait);
+	rcu_read_unlock();
+
+}
+
 
 /**
  * cl_wide_st_chg() - true if the state change is a cluster wide one
@@ -410,6 +422,22 @@ _drbd_request_state(struct drbd_device *device, union drbd_state mask,
 	return rv;
 }
 
+enum drbd_state_rv
+_drbd_request_state_holding_state_mutex(struct drbd_device *device, union drbd_state mask,
+		    union drbd_state val, enum chg_state_flags f)
+{
+	enum drbd_state_rv rv;
+
+	BUG_ON(f & CS_SERIALIZE);
+
+	wait_event_cmd(device->state_wait,
+		       (rv = drbd_req_state(device, mask, val, f)) != SS_IN_TRANSIENT_STATE,
+		       mutex_unlock(device->state_mutex),
+		       mutex_lock(device->state_mutex));
+
+	return rv;
+}
+
 static void print_st(struct drbd_device *device, const char *name, union drbd_state ns)
 {
 	drbd_err(device, " %s = { cs:%s ro:%s/%s ds:%s/%s %c%c%c%c%c%c }\n",
@@ -629,14 +657,11 @@ is_valid_soft_transition(union drbd_state os, union drbd_state ns, struct drbd_c
 	if (ns.conn == C_DISCONNECTING && os.conn == C_UNCONNECTED)
 		rv = SS_IN_TRANSIENT_STATE;
 
-	/* if (ns.conn == os.conn && ns.conn == C_WF_REPORT_PARAMS)
-	   rv = SS_IN_TRANSIENT_STATE; */
-
 	/* While establishing a connection only allow cstate to change.
-	   Delay/refuse role changes, detach attach etc... */
+	   Delay/refuse role changes, detach attach etc... (they do not touch cstate) */
 	if (test_bit(STATE_SENT, &connection->flags) &&
-	    !(os.conn == C_WF_REPORT_PARAMS ||
-	      (ns.conn == C_WF_REPORT_PARAMS && os.conn == C_WF_CONNECTION)))
+	    !((ns.conn == C_WF_REPORT_PARAMS && os.conn == C_WF_CONNECTION) ||
+	      (ns.conn >= C_CONNECTED && os.conn == C_WF_REPORT_PARAMS)))
 		rv = SS_IN_TRANSIENT_STATE;
 
 	if ((ns.conn == C_VERIFY_S || ns.conn == C_VERIFY_T) && os.conn < C_CONNECTED)
@@ -1032,8 +1057,10 @@ __drbd_set_state(struct drbd_device *device, union drbd_state ns,
 
 	/* Wake up role changes, that were delayed because of connection establishing */
 	if (os.conn == C_WF_REPORT_PARAMS && ns.conn != C_WF_REPORT_PARAMS &&
-	    no_peer_wf_report_params(connection))
+	    no_peer_wf_report_params(connection)) {
 		clear_bit(STATE_SENT, &connection->flags);
+		wake_up_all_devices(connection);
+	}
 
 	wake_up(&device->misc_wait);
 	wake_up(&device->state_wait);
@@ -1072,7 +1099,6 @@ __drbd_set_state(struct drbd_device *device, union drbd_state ns,
 
 		set_ov_position(device, ns.conn);
 		device->rs_start = now;
-		device->rs_last_events = 0;
 		device->rs_last_sect_ev = 0;
 		device->ov_last_oos_size = 0;
 		device->ov_last_oos_start = 0;
diff --git a/drivers/block/drbd/drbd_state.h b/drivers/block/drbd/drbd_state.h
index cc41605ba21c..7f53c40823cd 100644
--- a/drivers/block/drbd/drbd_state.h
+++ b/drivers/block/drbd/drbd_state.h
@@ -117,6 +117,11 @@ extern enum drbd_state_rv _drbd_request_state(struct drbd_device *,
 					      union drbd_state,
 					      union drbd_state,
 					      enum chg_state_flags);
+
+extern enum drbd_state_rv
+_drbd_request_state_holding_state_mutex(struct drbd_device *, union drbd_state,
+					union drbd_state, enum chg_state_flags);
+
 extern enum drbd_state_rv __drbd_set_state(struct drbd_device *, union drbd_state,
 					   enum chg_state_flags,
 					   struct completion *done);
diff --git a/drivers/block/drbd/drbd_worker.c b/drivers/block/drbd/drbd_worker.c
index 50776b362828..d0fae55d871d 100644
--- a/drivers/block/drbd/drbd_worker.c
+++ b/drivers/block/drbd/drbd_worker.c
@@ -43,10 +43,10 @@ static int make_ov_request(struct drbd_device *, int);
 static int make_resync_request(struct drbd_device *, int);
 
 /* endio handlers:
- *   drbd_md_io_complete (defined here)
+ *   drbd_md_endio (defined here)
  *   drbd_request_endio (defined here)
  *   drbd_peer_request_endio (defined here)
- *   bm_async_io_complete (defined in drbd_bitmap.c)
+ *   drbd_bm_endio (defined in drbd_bitmap.c)
  *
  * For all these callbacks, note the following:
  * The callbacks will be called in irq context by the IDE drivers,
@@ -65,7 +65,7 @@ rwlock_t global_state_lock;
 /* used for synchronous meta data and bitmap IO
  * submitted by drbd_md_sync_page_io()
  */
-void drbd_md_io_complete(struct bio *bio, int error)
+void drbd_md_endio(struct bio *bio, int error)
 {
 	struct drbd_device *device;
 
@@ -1592,11 +1592,15 @@ void drbd_resync_after_changed(struct drbd_device *device)
 
 void drbd_rs_controller_reset(struct drbd_device *device)
 {
+	struct gendisk *disk = device->ldev->backing_bdev->bd_contains->bd_disk;
 	struct fifo_buffer *plan;
 
 	atomic_set(&device->rs_sect_in, 0);
 	atomic_set(&device->rs_sect_ev, 0);
 	device->rs_in_flight = 0;
+	device->rs_last_events =
+		(int)part_stat_read(&disk->part0, sectors[0]) +
+		(int)part_stat_read(&disk->part0, sectors[1]);
 
 	/* Updating the RCU protected object in place is necessary since
 	   this function gets called from atomic context.
@@ -1743,7 +1747,6 @@ void drbd_start_resync(struct drbd_device *device, enum drbd_conns side)
 		device->rs_failed    = 0;
 		device->rs_paused    = 0;
 		device->rs_same_csum = 0;
-		device->rs_last_events = 0;
 		device->rs_last_sect_ev = 0;
 		device->rs_total     = tw;
 		device->rs_start     = now;
@@ -1853,9 +1856,12 @@ static void drbd_ldev_destroy(struct drbd_device *device)
 	device->resync = NULL;
 	lc_destroy(device->act_log);
 	device->act_log = NULL;
-	__no_warn(local,
-		drbd_free_ldev(device->ldev);
-		device->ldev = NULL;);
+
+	__acquire(local);
+	drbd_free_ldev(device->ldev);
+	device->ldev = NULL;
+	__release(local);
+
 	clear_bit(GOING_DISKLESS, &device->flags);
 	wake_up(&device->misc_wait);
 }
@@ -1928,19 +1934,18 @@ void __update_timing_details(
 	++(*cb_nr);
 }
 
-#define WORK_PENDING(work_bit, todo)	(todo & (1UL << work_bit))
 static void do_device_work(struct drbd_device *device, const unsigned long todo)
 {
-	if (WORK_PENDING(MD_SYNC, todo))
+	if (test_bit(MD_SYNC, &todo))
 		do_md_sync(device);
-	if (WORK_PENDING(RS_DONE, todo) ||
-	    WORK_PENDING(RS_PROGRESS, todo))
-		update_on_disk_bitmap(device, WORK_PENDING(RS_DONE, todo));
-	if (WORK_PENDING(GO_DISKLESS, todo))
+	if (test_bit(RS_DONE, &todo) ||
+	    test_bit(RS_PROGRESS, &todo))
+		update_on_disk_bitmap(device, test_bit(RS_DONE, &todo));
+	if (test_bit(GO_DISKLESS, &todo))
 		go_diskless(device);
-	if (WORK_PENDING(DESTROY_DISK, todo))
+	if (test_bit(DESTROY_DISK, &todo))
 		drbd_ldev_destroy(device);
-	if (WORK_PENDING(RS_START, todo))
+	if (test_bit(RS_START, &todo))
 		do_start_resync(device);
 }
 
@@ -1992,22 +1997,13 @@ static bool dequeue_work_batch(struct drbd_work_queue *queue, struct list_head *
 	return !list_empty(work_list);
 }
 
-static bool dequeue_work_item(struct drbd_work_queue *queue, struct list_head *work_list)
-{
-	spin_lock_irq(&queue->q_lock);
-	if (!list_empty(&queue->q))
-		list_move(queue->q.next, work_list);
-	spin_unlock_irq(&queue->q_lock);
-	return !list_empty(work_list);
-}
-
 static void wait_for_work(struct drbd_connection *connection, struct list_head *work_list)
 {
 	DEFINE_WAIT(wait);
 	struct net_conf *nc;
 	int uncork, cork;
 
-	dequeue_work_item(&connection->sender_work, work_list);
+	dequeue_work_batch(&connection->sender_work, work_list);
 	if (!list_empty(work_list))
 		return;
 
@@ -2033,8 +2029,6 @@ static void wait_for_work(struct drbd_connection *connection, struct list_head *
 		prepare_to_wait(&connection->sender_work.q_wait, &wait, TASK_INTERRUPTIBLE);
 		spin_lock_irq(&connection->resource->req_lock);
 		spin_lock(&connection->sender_work.q_lock);	/* FIXME get rid of this one? */
-		/* dequeue single item only,
-		 * we still use drbd_queue_work_front() in some places */
 		if (!list_empty(&connection->sender_work.q))
 			list_splice_tail_init(&connection->sender_work.q, work_list);
 		spin_unlock(&connection->sender_work.q_lock);	/* FIXME get rid of this one? */
@@ -2121,7 +2115,7 @@ int drbd_worker(struct drbd_thread *thi)
 		if (get_t_state(thi) != RUNNING)
 			break;
 
-		while (!list_empty(&work_list)) {
+		if (!list_empty(&work_list)) {
 			w = list_first_entry(&work_list, struct drbd_work, list);
 			list_del_init(&w->list);
 			update_worker_timing_details(connection, w->cb);
@@ -2137,13 +2131,13 @@ int drbd_worker(struct drbd_thread *thi)
 			update_worker_timing_details(connection, do_unqueued_work);
 			do_unqueued_work(connection);
 		}
-		while (!list_empty(&work_list)) {
+		if (!list_empty(&work_list)) {
 			w = list_first_entry(&work_list, struct drbd_work, list);
 			list_del_init(&w->list);
 			update_worker_timing_details(connection, w->cb);
 			w->cb(w, 1);
-		}
-		dequeue_work_batch(&connection->sender_work, &work_list);
+		} else
+			dequeue_work_batch(&connection->sender_work, &work_list);
 	} while (!list_empty(&work_list) || test_bit(DEVICE_WORK_PENDING, &connection->flags));
 
 	rcu_read_lock();
diff --git a/drivers/block/hd.c b/drivers/block/hd.c
index 8a290c08262f..3abb121825bc 100644
--- a/drivers/block/hd.c
+++ b/drivers/block/hd.c
@@ -694,16 +694,6 @@ static const struct block_device_operations hd_fops = {
 	.getgeo =	hd_getgeo,
 };
 
-/*
- * This is the hard disk IRQ description. The IRQF_DISABLED in sa_flags
- * means we run the IRQ-handler with interrupts disabled:  this is bad for
- * interrupt latency, but anything else has led to problems on some
- * machines.
- *
- * We enable interrupts in some of the routines after making sure it's
- * safe.
- */
-
 static int __init hd_init(void)
 {
 	int drive;
@@ -761,7 +751,7 @@ static int __init hd_init(void)
 			p->cyl, p->head, p->sect);
 	}
 
-	if (request_irq(HD_IRQ, hd_interrupt, IRQF_DISABLED, "hd", NULL)) {
+	if (request_irq(HD_IRQ, hd_interrupt, 0, "hd", NULL)) {
 		printk("hd: unable to get IRQ%d for the hard disk driver\n",
 			HD_IRQ);
 		goto out1;
diff --git a/drivers/block/mg_disk.c b/drivers/block/mg_disk.c
index e352cac707e8..145ce2aa2e78 100644
--- a/drivers/block/mg_disk.c
+++ b/drivers/block/mg_disk.c
@@ -1082,7 +1082,6 @@ static struct platform_driver mg_disk_driver = {
 	.remove = mg_remove,
 	.driver = {
 		.name = MG_DEV_NAME,
-		.owner = THIS_MODULE,
 		.pm = &mg_pm,
 	}
 };
diff --git a/drivers/block/mtip32xx/mtip32xx.c b/drivers/block/mtip32xx/mtip32xx.c
index 5c8e7fe07745..3bd7ca9853a8 100644
--- a/drivers/block/mtip32xx/mtip32xx.c
+++ b/drivers/block/mtip32xx/mtip32xx.c
@@ -247,7 +247,7 @@ static void mtip_async_complete(struct mtip_port *port,
 	if (unlikely(cmd->unaligned))
 		up(&port->cmd_slot_unal);
 
-	blk_mq_end_io(rq, status ? -EIO : 0);
+	blk_mq_end_request(rq, status ? -EIO : 0);
 }
 
 /*
@@ -3739,7 +3739,7 @@ static int mtip_submit_request(struct blk_mq_hw_ctx *hctx, struct request *rq)
 		int err;
 
 		err = mtip_send_trim(dd, blk_rq_pos(rq), blk_rq_sectors(rq));
-		blk_mq_end_io(rq, err);
+		blk_mq_end_request(rq, err);
 		return 0;
 	}
 
@@ -3775,13 +3775,17 @@ static bool mtip_check_unal_depth(struct blk_mq_hw_ctx *hctx,
 	return false;
 }
 
-static int mtip_queue_rq(struct blk_mq_hw_ctx *hctx, struct request *rq)
+static int mtip_queue_rq(struct blk_mq_hw_ctx *hctx,
+			 const struct blk_mq_queue_data *bd)
 {
+	struct request *rq = bd->rq;
 	int ret;
 
 	if (unlikely(mtip_check_unal_depth(hctx, rq)))
 		return BLK_MQ_RQ_QUEUE_BUSY;
 
+	blk_mq_start_request(rq);
+
 	ret = mtip_submit_request(hctx, rq);
 	if (likely(!ret))
 		return BLK_MQ_RQ_QUEUE_OK;
@@ -3951,6 +3955,7 @@ skip_create_disk:
 
 	/* Set device limits. */
 	set_bit(QUEUE_FLAG_NONROT, &dd->queue->queue_flags);
+	clear_bit(QUEUE_FLAG_ADD_RANDOM, &dd->queue->queue_flags);
 	blk_queue_max_segments(dd->queue, MTIP_MAX_SG);
 	blk_queue_physical_block_size(dd->queue, 4096);
 	blk_queue_max_hw_sectors(dd->queue, 0xffff);
diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
index fb31b8ee4372..4bc2a5cb9935 100644
--- a/drivers/block/nbd.c
+++ b/drivers/block/nbd.c
@@ -847,6 +847,7 @@ static int __init nbd_init(void)
 		 * Tell the block layer that we are not a rotational device
 		 */
 		queue_flag_set_unlocked(QUEUE_FLAG_NONROT, disk->queue);
+		queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, disk->queue);
 		disk->queue->limits.discard_granularity = 512;
 		disk->queue->limits.max_discard_sectors = UINT_MAX;
 		disk->queue->limits.discard_zeroes_data = 0;
diff --git a/drivers/block/null_blk.c b/drivers/block/null_blk.c
index 00d469c7f9f7..aa2224aa7caa 100644
--- a/drivers/block/null_blk.c
+++ b/drivers/block/null_blk.c
@@ -78,7 +78,33 @@ module_param(home_node, int, S_IRUGO);
 MODULE_PARM_DESC(home_node, "Home node for the device");
 
 static int queue_mode = NULL_Q_MQ;
-module_param(queue_mode, int, S_IRUGO);
+
+static int null_param_store_val(const char *str, int *val, int min, int max)
+{
+	int ret, new_val;
+
+	ret = kstrtoint(str, 10, &new_val);
+	if (ret)
+		return -EINVAL;
+
+	if (new_val < min || new_val > max)
+		return -EINVAL;
+
+	*val = new_val;
+	return 0;
+}
+
+static int null_set_queue_mode(const char *str, const struct kernel_param *kp)
+{
+	return null_param_store_val(str, &queue_mode, NULL_Q_BIO, NULL_Q_MQ);
+}
+
+static struct kernel_param_ops null_queue_mode_param_ops = {
+	.set	= null_set_queue_mode,
+	.get	= param_get_int,
+};
+
+device_param_cb(queue_mode, &null_queue_mode_param_ops, &queue_mode, S_IRUGO);
 MODULE_PARM_DESC(queue_mode, "Block interface to use (0=bio,1=rq,2=multiqueue)");
 
 static int gb = 250;
@@ -94,7 +120,19 @@ module_param(nr_devices, int, S_IRUGO);
 MODULE_PARM_DESC(nr_devices, "Number of devices to register");
 
 static int irqmode = NULL_IRQ_SOFTIRQ;
-module_param(irqmode, int, S_IRUGO);
+
+static int null_set_irqmode(const char *str, const struct kernel_param *kp)
+{
+	return null_param_store_val(str, &irqmode, NULL_IRQ_NONE,
+					NULL_IRQ_TIMER);
+}
+
+static struct kernel_param_ops null_irqmode_param_ops = {
+	.set	= null_set_irqmode,
+	.get	= param_get_int,
+};
+
+device_param_cb(irqmode, &null_irqmode_param_ops, &irqmode, S_IRUGO);
 MODULE_PARM_DESC(irqmode, "IRQ completion handler. 0-none, 1-softirq, 2-timer");
 
 static int completion_nsec = 10000;
@@ -177,7 +215,7 @@ static void end_cmd(struct nullb_cmd *cmd)
 {
 	switch (queue_mode)  {
 	case NULL_Q_MQ:
-		blk_mq_end_io(cmd->rq, 0);
+		blk_mq_end_request(cmd->rq, 0);
 		return;
 	case NULL_Q_RQ:
 		INIT_LIST_HEAD(&cmd->rq->queuelist);
@@ -313,13 +351,16 @@ static void null_request_fn(struct request_queue *q)
 	}
 }
 
-static int null_queue_rq(struct blk_mq_hw_ctx *hctx, struct request *rq)
+static int null_queue_rq(struct blk_mq_hw_ctx *hctx,
+			 const struct blk_mq_queue_data *bd)
 {
-	struct nullb_cmd *cmd = blk_mq_rq_to_pdu(rq);
+	struct nullb_cmd *cmd = blk_mq_rq_to_pdu(bd->rq);
 
-	cmd->rq = rq;
+	cmd->rq = bd->rq;
 	cmd->nq = hctx->driver_data;
 
+	blk_mq_start_request(bd->rq);
+
 	null_handle_cmd(cmd);
 	return BLK_MQ_RQ_QUEUE_OK;
 }
@@ -447,14 +488,10 @@ static int init_driver_queues(struct nullb *nullb)
 
 		ret = setup_commands(nq);
 		if (ret)
-			goto err_queue;
+			return ret;
 		nullb->nr_queues++;
 	}
-
 	return 0;
-err_queue:
-	cleanup_queues(nullb);
-	return ret;
 }
 
 static int null_add_dev(void)
@@ -493,7 +530,7 @@ static int null_add_dev(void)
 			goto out_cleanup_queues;
 
 		nullb->q = blk_mq_init_queue(&nullb->tag_set);
-		if (!nullb->q) {
+		if (IS_ERR(nullb->q)) {
 			rv = -ENOMEM;
 			goto out_cleanup_tags;
 		}
@@ -504,7 +541,9 @@ static int null_add_dev(void)
 			goto out_cleanup_queues;
 		}
 		blk_queue_make_request(nullb->q, null_queue_bio);
-		init_driver_queues(nullb);
+		rv = init_driver_queues(nullb);
+		if (rv)
+			goto out_cleanup_blk_queue;
 	} else {
 		nullb->q = blk_init_queue_node(null_request_fn, &nullb->lock, home_node);
 		if (!nullb->q) {
@@ -513,11 +552,14 @@ static int null_add_dev(void)
 		}
 		blk_queue_prep_rq(nullb->q, null_rq_prep_fn);
 		blk_queue_softirq_done(nullb->q, null_softirq_done_fn);
-		init_driver_queues(nullb);
+		rv = init_driver_queues(nullb);
+		if (rv)
+			goto out_cleanup_blk_queue;
 	}
 
 	nullb->q->queuedata = nullb;
 	queue_flag_set_unlocked(QUEUE_FLAG_NONROT, nullb->q);
+	queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, nullb->q);
 
 	disk = nullb->disk = alloc_disk_node(1, home_node);
 	if (!disk) {
diff --git a/drivers/block/nvme-core.c b/drivers/block/nvme-core.c
index 02351e217165..d826bf3e62c8 100644
--- a/drivers/block/nvme-core.c
+++ b/drivers/block/nvme-core.c
@@ -13,9 +13,9 @@
  */
 
 #include <linux/nvme.h>
-#include <linux/bio.h>
 #include <linux/bitops.h>
 #include <linux/blkdev.h>
+#include <linux/blk-mq.h>
 #include <linux/cpu.h>
 #include <linux/delay.h>
 #include <linux/errno.h>
@@ -33,7 +33,6 @@
 #include <linux/module.h>
 #include <linux/moduleparam.h>
 #include <linux/pci.h>
-#include <linux/percpu.h>
 #include <linux/poison.h>
 #include <linux/ptrace.h>
 #include <linux/sched.h>
@@ -42,12 +41,12 @@
 #include <scsi/sg.h>
 #include <asm-generic/io-64-nonatomic-lo-hi.h>
 
-#include <trace/events/block.h>
-
 #define NVME_Q_DEPTH		1024
+#define NVME_AQ_DEPTH		64
 #define SQ_SIZE(depth)		(depth * sizeof(struct nvme_command))
 #define CQ_SIZE(depth)		(depth * sizeof(struct nvme_completion))
 #define ADMIN_TIMEOUT		(admin_timeout * HZ)
+#define SHUTDOWN_TIMEOUT	(shutdown_timeout * HZ)
 #define IOD_TIMEOUT		(retry_time * HZ)
 
 static unsigned char admin_timeout = 60;
@@ -62,6 +61,10 @@ static unsigned char retry_time = 30;
 module_param(retry_time, byte, 0644);
 MODULE_PARM_DESC(retry_time, "time in seconds to retry failed I/O");
 
+static unsigned char shutdown_timeout = 5;
+module_param(shutdown_timeout, byte, 0644);
+MODULE_PARM_DESC(shutdown_timeout, "timeout in seconds for controller shutdown");
+
 static int nvme_major;
 module_param(nvme_major, int, 0);
 
@@ -76,10 +79,12 @@ static wait_queue_head_t nvme_kthread_wait;
 static struct notifier_block nvme_nb;
 
 static void nvme_reset_failed_dev(struct work_struct *ws);
+static int nvme_process_cq(struct nvme_queue *nvmeq);
 
 struct async_cmd_info {
 	struct kthread_work work;
 	struct kthread_worker *worker;
+	struct request *req;
 	u32 result;
 	int status;
 	void *ctx;
@@ -90,7 +95,7 @@ struct async_cmd_info {
  * commands and one for I/O commands).
  */
 struct nvme_queue {
-	struct rcu_head r_head;
+	struct llist_node node;
 	struct device *q_dmadev;
 	struct nvme_dev *dev;
 	char irqname[24];	/* nvme4294967295-65535\0 */
@@ -99,23 +104,17 @@ struct nvme_queue {
 	volatile struct nvme_completion *cqes;
 	dma_addr_t sq_dma_addr;
 	dma_addr_t cq_dma_addr;
-	wait_queue_head_t sq_full;
-	wait_queue_t sq_cong_wait;
-	struct bio_list sq_cong;
-	struct list_head iod_bio;
 	u32 __iomem *q_db;
 	u16 q_depth;
-	u16 cq_vector;
+	s16 cq_vector;
 	u16 sq_head;
 	u16 sq_tail;
 	u16 cq_head;
 	u16 qid;
 	u8 cq_phase;
 	u8 cqe_seen;
-	u8 q_suspended;
-	cpumask_var_t cpu_mask;
 	struct async_cmd_info cmdinfo;
-	unsigned long cmdid_data[];
+	struct blk_mq_hw_ctx *hctx;
 };
 
 /*
@@ -143,62 +142,80 @@ typedef void (*nvme_completion_fn)(struct nvme_queue *, void *,
 struct nvme_cmd_info {
 	nvme_completion_fn fn;
 	void *ctx;
-	unsigned long timeout;
 	int aborted;
+	struct nvme_queue *nvmeq;
 };
 
-static struct nvme_cmd_info *nvme_cmd_info(struct nvme_queue *nvmeq)
+static int nvme_admin_init_hctx(struct blk_mq_hw_ctx *hctx, void *data,
+				unsigned int hctx_idx)
 {
-	return (void *)&nvmeq->cmdid_data[BITS_TO_LONGS(nvmeq->q_depth)];
+	struct nvme_dev *dev = data;
+	struct nvme_queue *nvmeq = dev->queues[0];
+
+	WARN_ON(nvmeq->hctx);
+	nvmeq->hctx = hctx;
+	hctx->driver_data = nvmeq;
+	return 0;
 }
 
-static unsigned nvme_queue_extra(int depth)
+static int nvme_admin_init_request(void *data, struct request *req,
+				unsigned int hctx_idx, unsigned int rq_idx,
+				unsigned int numa_node)
 {
-	return DIV_ROUND_UP(depth, 8) + (depth * sizeof(struct nvme_cmd_info));
+	struct nvme_dev *dev = data;
+	struct nvme_cmd_info *cmd = blk_mq_rq_to_pdu(req);
+	struct nvme_queue *nvmeq = dev->queues[0];
+
+	BUG_ON(!nvmeq);
+	cmd->nvmeq = nvmeq;
+	return 0;
 }
 
-/**
- * alloc_cmdid() - Allocate a Command ID
- * @nvmeq: The queue that will be used for this command
- * @ctx: A pointer that will be passed to the handler
- * @handler: The function to call on completion
- *
- * Allocate a Command ID for a queue.  The data passed in will
- * be passed to the completion handler.  This is implemented by using
- * the bottom two bits of the ctx pointer to store the handler ID.
- * Passing in a pointer that's not 4-byte aligned will cause a BUG.
- * We can change this if it becomes a problem.
- *
- * May be called with local interrupts disabled and the q_lock held,
- * or with interrupts enabled and no locks held.
- */
-static int alloc_cmdid(struct nvme_queue *nvmeq, void *ctx,
-				nvme_completion_fn handler, unsigned timeout)
+static void nvme_exit_hctx(struct blk_mq_hw_ctx *hctx, unsigned int hctx_idx)
 {
-	int depth = nvmeq->q_depth - 1;
-	struct nvme_cmd_info *info = nvme_cmd_info(nvmeq);
-	int cmdid;
+	struct nvme_queue *nvmeq = hctx->driver_data;
 
-	do {
-		cmdid = find_first_zero_bit(nvmeq->cmdid_data, depth);
-		if (cmdid >= depth)
-			return -EBUSY;
-	} while (test_and_set_bit(cmdid, nvmeq->cmdid_data));
+	nvmeq->hctx = NULL;
+}
+
+static int nvme_init_hctx(struct blk_mq_hw_ctx *hctx, void *data,
+			  unsigned int hctx_idx)
+{
+	struct nvme_dev *dev = data;
+	struct nvme_queue *nvmeq = dev->queues[
+					(hctx_idx % dev->queue_count) + 1];
 
-	info[cmdid].fn = handler;
-	info[cmdid].ctx = ctx;
-	info[cmdid].timeout = jiffies + timeout;
-	info[cmdid].aborted = 0;
-	return cmdid;
+	if (!nvmeq->hctx)
+		nvmeq->hctx = hctx;
+
+	/* nvmeq queues are shared between namespaces. We assume here that
+	 * blk-mq map the tags so they match up with the nvme queue tags. */
+	WARN_ON(nvmeq->hctx->tags != hctx->tags);
+
+	hctx->driver_data = nvmeq;
+	return 0;
+}
+
+static int nvme_init_request(void *data, struct request *req,
+				unsigned int hctx_idx, unsigned int rq_idx,
+				unsigned int numa_node)
+{
+	struct nvme_dev *dev = data;
+	struct nvme_cmd_info *cmd = blk_mq_rq_to_pdu(req);
+	struct nvme_queue *nvmeq = dev->queues[hctx_idx + 1];
+
+	BUG_ON(!nvmeq);
+	cmd->nvmeq = nvmeq;
+	return 0;
 }
 
-static int alloc_cmdid_killable(struct nvme_queue *nvmeq, void *ctx,
-				nvme_completion_fn handler, unsigned timeout)
+static void nvme_set_info(struct nvme_cmd_info *cmd, void *ctx,
+				nvme_completion_fn handler)
 {
-	int cmdid;
-	wait_event_killable(nvmeq->sq_full,
-		(cmdid = alloc_cmdid(nvmeq, ctx, handler, timeout)) >= 0);
-	return (cmdid < 0) ? -EINTR : cmdid;
+	cmd->fn = handler;
+	cmd->ctx = ctx;
+	cmd->aborted = 0;
+	blk_mq_start_request(blk_mq_rq_from_pdu(cmd));
 }
 
 /* Special values must be less than 0x1000 */
@@ -206,17 +223,12 @@ static int alloc_cmdid_killable(struct nvme_queue *nvmeq, void *ctx,
 #define CMD_CTX_CANCELLED	(0x30C + CMD_CTX_BASE)
 #define CMD_CTX_COMPLETED	(0x310 + CMD_CTX_BASE)
 #define CMD_CTX_INVALID		(0x314 + CMD_CTX_BASE)
-#define CMD_CTX_ABORT		(0x318 + CMD_CTX_BASE)
 
 static void special_completion(struct nvme_queue *nvmeq, void *ctx,
 						struct nvme_completion *cqe)
 {
 	if (ctx == CMD_CTX_CANCELLED)
 		return;
-	if (ctx == CMD_CTX_ABORT) {
-		++nvmeq->dev->abort_limit;
-		return;
-	}
 	if (ctx == CMD_CTX_COMPLETED) {
 		dev_warn(nvmeq->q_dmadev,
 				"completed id %d twice on queue %d\n",
@@ -229,99 +241,89 @@ static void special_completion(struct nvme_queue *nvmeq, void *ctx,
 				cqe->command_id, le16_to_cpup(&cqe->sq_id));
 		return;
 	}
-
 	dev_warn(nvmeq->q_dmadev, "Unknown special completion %p\n", ctx);
 }
 
-static void async_completion(struct nvme_queue *nvmeq, void *ctx,
-						struct nvme_completion *cqe)
-{
-	struct async_cmd_info *cmdinfo = ctx;
-	cmdinfo->result = le32_to_cpup(&cqe->result);
-	cmdinfo->status = le16_to_cpup(&cqe->status) >> 1;
-	queue_kthread_work(cmdinfo->worker, &cmdinfo->work);
-}
-
-/*
- * Called with local interrupts disabled and the q_lock held.  May not sleep.
- */
-static void *free_cmdid(struct nvme_queue *nvmeq, int cmdid,
-						nvme_completion_fn *fn)
+static void *cancel_cmd_info(struct nvme_cmd_info *cmd, nvme_completion_fn *fn)
 {
 	void *ctx;
-	struct nvme_cmd_info *info = nvme_cmd_info(nvmeq);
 
-	if (cmdid >= nvmeq->q_depth || !info[cmdid].fn) {
-		if (fn)
-			*fn = special_completion;
-		return CMD_CTX_INVALID;
-	}
 	if (fn)
-		*fn = info[cmdid].fn;
-	ctx = info[cmdid].ctx;
-	info[cmdid].fn = special_completion;
-	info[cmdid].ctx = CMD_CTX_COMPLETED;
-	clear_bit(cmdid, nvmeq->cmdid_data);
-	wake_up(&nvmeq->sq_full);
+		*fn = cmd->fn;
+	ctx = cmd->ctx;
+	cmd->fn = special_completion;
+	cmd->ctx = CMD_CTX_CANCELLED;
 	return ctx;
 }
 
-static void *cancel_cmdid(struct nvme_queue *nvmeq, int cmdid,
-						nvme_completion_fn *fn)
+static void async_req_completion(struct nvme_queue *nvmeq, void *ctx,
+						struct nvme_completion *cqe)
 {
-	void *ctx;
-	struct nvme_cmd_info *info = nvme_cmd_info(nvmeq);
-	if (fn)
-		*fn = info[cmdid].fn;
-	ctx = info[cmdid].ctx;
-	info[cmdid].fn = special_completion;
-	info[cmdid].ctx = CMD_CTX_CANCELLED;
-	return ctx;
-}
+	struct request *req = ctx;
 
-static struct nvme_queue *raw_nvmeq(struct nvme_dev *dev, int qid)
-{
-	return rcu_dereference_raw(dev->queues[qid]);
+	u32 result = le32_to_cpup(&cqe->result);
+	u16 status = le16_to_cpup(&cqe->status) >> 1;
+
+	if (status == NVME_SC_SUCCESS || status == NVME_SC_ABORT_REQ)
+		++nvmeq->dev->event_limit;
+	if (status == NVME_SC_SUCCESS)
+		dev_warn(nvmeq->q_dmadev,
+			"async event result %08x\n", result);
+
+	blk_mq_free_hctx_request(nvmeq->hctx, req);
 }
 
-static struct nvme_queue *get_nvmeq(struct nvme_dev *dev) __acquires(RCU)
+static void abort_completion(struct nvme_queue *nvmeq, void *ctx,
+						struct nvme_completion *cqe)
 {
-	struct nvme_queue *nvmeq;
-	unsigned queue_id = get_cpu_var(*dev->io_queue);
+	struct request *req = ctx;
 
-	rcu_read_lock();
-	nvmeq = rcu_dereference(dev->queues[queue_id]);
-	if (nvmeq)
-		return nvmeq;
+	u16 status = le16_to_cpup(&cqe->status) >> 1;
+	u32 result = le32_to_cpup(&cqe->result);
 
-	rcu_read_unlock();
-	put_cpu_var(*dev->io_queue);
-	return NULL;
+	blk_mq_free_hctx_request(nvmeq->hctx, req);
+
+	dev_warn(nvmeq->q_dmadev, "Abort status:%x result:%x", status, result);
+	++nvmeq->dev->abort_limit;
 }
 
-static void put_nvmeq(struct nvme_queue *nvmeq) __releases(RCU)
+static void async_completion(struct nvme_queue *nvmeq, void *ctx,
+						struct nvme_completion *cqe)
 {
-	rcu_read_unlock();
-	put_cpu_var(nvmeq->dev->io_queue);
+	struct async_cmd_info *cmdinfo = ctx;
+	cmdinfo->result = le32_to_cpup(&cqe->result);
+	cmdinfo->status = le16_to_cpup(&cqe->status) >> 1;
+	queue_kthread_work(cmdinfo->worker, &cmdinfo->work);
+	blk_mq_free_hctx_request(nvmeq->hctx, cmdinfo->req);
 }
 
-static struct nvme_queue *lock_nvmeq(struct nvme_dev *dev, int q_idx)
-							__acquires(RCU)
+static inline struct nvme_cmd_info *get_cmd_from_tag(struct nvme_queue *nvmeq,
+				  unsigned int tag)
 {
-	struct nvme_queue *nvmeq;
+	struct blk_mq_hw_ctx *hctx = nvmeq->hctx;
+	struct request *req = blk_mq_tag_to_rq(hctx->tags, tag);
 
-	rcu_read_lock();
-	nvmeq = rcu_dereference(dev->queues[q_idx]);
-	if (nvmeq)
-		return nvmeq;
-
-	rcu_read_unlock();
-	return NULL;
+	return blk_mq_rq_to_pdu(req);
 }
 
-static void unlock_nvmeq(struct nvme_queue *nvmeq) __releases(RCU)
+/*
+ * Called with local interrupts disabled and the q_lock held.  May not sleep.
+ */
+static void *nvme_finish_cmd(struct nvme_queue *nvmeq, int tag,
+						nvme_completion_fn *fn)
 {
-	rcu_read_unlock();
+	struct nvme_cmd_info *cmd = get_cmd_from_tag(nvmeq, tag);
+	void *ctx;
+	if (tag >= nvmeq->q_depth) {
+		*fn = special_completion;
+		return CMD_CTX_INVALID;
+	}
+	if (fn)
+		*fn = cmd->fn;
+	ctx = cmd->ctx;
+	cmd->fn = special_completion;
+	cmd->ctx = CMD_CTX_COMPLETED;
+	return ctx;
 }
 
 /**
@@ -331,26 +333,29 @@ static void unlock_nvmeq(struct nvme_queue *nvmeq) __releases(RCU)
  *
  * Safe to use from interrupt context
  */
-static int nvme_submit_cmd(struct nvme_queue *nvmeq, struct nvme_command *cmd)
+static int __nvme_submit_cmd(struct nvme_queue *nvmeq, struct nvme_command *cmd)
 {
-	unsigned long flags;
-	u16 tail;
-	spin_lock_irqsave(&nvmeq->q_lock, flags);
-	if (nvmeq->q_suspended) {
-		spin_unlock_irqrestore(&nvmeq->q_lock, flags);
-		return -EBUSY;
-	}
-	tail = nvmeq->sq_tail;
+	u16 tail = nvmeq->sq_tail;
+
 	memcpy(&nvmeq->sq_cmds[tail], cmd, sizeof(*cmd));
 	if (++tail == nvmeq->q_depth)
 		tail = 0;
 	writel(tail, nvmeq->q_db);
 	nvmeq->sq_tail = tail;
-	spin_unlock_irqrestore(&nvmeq->q_lock, flags);
 
 	return 0;
 }
 
+static int nvme_submit_cmd(struct nvme_queue *nvmeq, struct nvme_command *cmd)
+{
+	unsigned long flags;
+	int ret;
+	spin_lock_irqsave(&nvmeq->q_lock, flags);
+	ret = __nvme_submit_cmd(nvmeq, cmd);
+	spin_unlock_irqrestore(&nvmeq->q_lock, flags);
+	return ret;
+}
+
 static __le64 **iod_list(struct nvme_iod *iod)
 {
 	return ((void *)iod) + iod->offset;
@@ -361,17 +366,17 @@ static __le64 **iod_list(struct nvme_iod *iod)
  * as it only leads to a small amount of wasted memory for the lifetime of
  * the I/O.
  */
-static int nvme_npages(unsigned size)
+static int nvme_npages(unsigned size, struct nvme_dev *dev)
 {
-	unsigned nprps = DIV_ROUND_UP(size + PAGE_SIZE, PAGE_SIZE);
-	return DIV_ROUND_UP(8 * nprps, PAGE_SIZE - 8);
+	unsigned nprps = DIV_ROUND_UP(size + dev->page_size, dev->page_size);
+	return DIV_ROUND_UP(8 * nprps, dev->page_size - 8);
 }
 
 static struct nvme_iod *
-nvme_alloc_iod(unsigned nseg, unsigned nbytes, gfp_t gfp)
+nvme_alloc_iod(unsigned nseg, unsigned nbytes, struct nvme_dev *dev, gfp_t gfp)
 {
 	struct nvme_iod *iod = kmalloc(sizeof(struct nvme_iod) +
-				sizeof(__le64 *) * nvme_npages(nbytes) +
+				sizeof(__le64 *) * nvme_npages(nbytes, dev) +
 				sizeof(struct scatterlist) * nseg, gfp);
 
 	if (iod) {
@@ -380,7 +385,6 @@ nvme_alloc_iod(unsigned nseg, unsigned nbytes, gfp_t gfp)
 		iod->length = nbytes;
 		iod->nents = 0;
 		iod->first_dma = 0ULL;
-		iod->start_time = jiffies;
 	}
 
 	return iod;
@@ -388,7 +392,7 @@ nvme_alloc_iod(unsigned nseg, unsigned nbytes, gfp_t gfp)
 
 void nvme_free_iod(struct nvme_dev *dev, struct nvme_iod *iod)
 {
-	const int last_prp = PAGE_SIZE / 8 - 1;
+	const int last_prp = dev->page_size / 8 - 1;
 	int i;
 	__le64 **list = iod_list(iod);
 	dma_addr_t prp_dma = iod->first_dma;
@@ -404,65 +408,54 @@ void nvme_free_iod(struct nvme_dev *dev, struct nvme_iod *iod)
 	kfree(iod);
 }
 
-static void nvme_start_io_acct(struct bio *bio)
-{
-	struct gendisk *disk = bio->bi_bdev->bd_disk;
-	if (blk_queue_io_stat(disk->queue)) {
-		const int rw = bio_data_dir(bio);
-		int cpu = part_stat_lock();
-		part_round_stats(cpu, &disk->part0);
-		part_stat_inc(cpu, &disk->part0, ios[rw]);
-		part_stat_add(cpu, &disk->part0, sectors[rw],
-							bio_sectors(bio));
-		part_inc_in_flight(&disk->part0, rw);
-		part_stat_unlock();
-	}
-}
-
-static void nvme_end_io_acct(struct bio *bio, unsigned long start_time)
+static int nvme_error_status(u16 status)
 {
-	struct gendisk *disk = bio->bi_bdev->bd_disk;
-	if (blk_queue_io_stat(disk->queue)) {
-		const int rw = bio_data_dir(bio);
-		unsigned long duration = jiffies - start_time;
-		int cpu = part_stat_lock();
-		part_stat_add(cpu, &disk->part0, ticks[rw], duration);
-		part_round_stats(cpu, &disk->part0);
-		part_dec_in_flight(&disk->part0, rw);
-		part_stat_unlock();
+	switch (status & 0x7ff) {
+	case NVME_SC_SUCCESS:
+		return 0;
+	case NVME_SC_CAP_EXCEEDED:
+		return -ENOSPC;
+	default:
+		return -EIO;
 	}
 }
 
-static void bio_completion(struct nvme_queue *nvmeq, void *ctx,
+static void req_completion(struct nvme_queue *nvmeq, void *ctx,
 						struct nvme_completion *cqe)
 {
 	struct nvme_iod *iod = ctx;
-	struct bio *bio = iod->private;
+	struct request *req = iod->private;
+	struct nvme_cmd_info *cmd_rq = blk_mq_rq_to_pdu(req);
+
 	u16 status = le16_to_cpup(&cqe->status) >> 1;
-	int error = 0;
 
 	if (unlikely(status)) {
-		if (!(status & NVME_SC_DNR ||
-				bio->bi_rw & REQ_FAILFAST_MASK) &&
-				(jiffies - iod->start_time) < IOD_TIMEOUT) {
-			if (!waitqueue_active(&nvmeq->sq_full))
-				add_wait_queue(&nvmeq->sq_full,
-							&nvmeq->sq_cong_wait);
-			list_add_tail(&iod->node, &nvmeq->iod_bio);
-			wake_up(&nvmeq->sq_full);
+		if (!(status & NVME_SC_DNR || blk_noretry_request(req))
+		    && (jiffies - req->start_time) < req->timeout) {
+			unsigned long flags;
+
+			blk_mq_requeue_request(req);
+			spin_lock_irqsave(req->q->queue_lock, flags);
+			if (!blk_queue_stopped(req->q))
+				blk_mq_kick_requeue_list(req->q);
+			spin_unlock_irqrestore(req->q->queue_lock, flags);
 			return;
 		}
-		error = -EIO;
-	}
-	if (iod->nents) {
-		dma_unmap_sg(nvmeq->q_dmadev, iod->sg, iod->nents,
-			bio_data_dir(bio) ? DMA_TO_DEVICE : DMA_FROM_DEVICE);
-		nvme_end_io_acct(bio, iod->start_time);
-	}
+		req->errors = nvme_error_status(status);
+	} else
+		req->errors = 0;
+
+	if (cmd_rq->aborted)
+		dev_warn(&nvmeq->dev->pci_dev->dev,
+			"completing aborted command with status:%04x\n",
+			status);
+
+	if (iod->nents)
+		dma_unmap_sg(&nvmeq->dev->pci_dev->dev, iod->sg, iod->nents,
+			rq_data_dir(req) ? DMA_TO_DEVICE : DMA_FROM_DEVICE);
 	nvme_free_iod(nvmeq->dev, iod);
 
-	trace_block_bio_complete(bdev_get_queue(bio->bi_bdev), bio, error);
-	bio_endio(bio, error);
+	blk_mq_complete_request(req);
 }
 
 /* length is in bytes.  gfp flags indicates whether we may sleep. */
@@ -479,26 +472,27 @@ int nvme_setup_prps(struct nvme_dev *dev, struct nvme_iod *iod, int total_len,
 	__le64 **list = iod_list(iod);
 	dma_addr_t prp_dma;
 	int nprps, i;
+	u32 page_size = dev->page_size;
 
-	length -= (PAGE_SIZE - offset);
+	length -= (page_size - offset);
 	if (length <= 0)
 		return total_len;
 
-	dma_len -= (PAGE_SIZE - offset);
+	dma_len -= (page_size - offset);
 	if (dma_len) {
-		dma_addr += (PAGE_SIZE - offset);
+		dma_addr += (page_size - offset);
 	} else {
 		sg = sg_next(sg);
 		dma_addr = sg_dma_address(sg);
 		dma_len = sg_dma_len(sg);
 	}
 
-	if (length <= PAGE_SIZE) {
+	if (length <= page_size) {
 		iod->first_dma = dma_addr;
 		return total_len;
 	}
 
-	nprps = DIV_ROUND_UP(length, PAGE_SIZE);
+	nprps = DIV_ROUND_UP(length, page_size);
 	if (nprps <= (256 / 8)) {
 		pool = dev->prp_small_pool;
 		iod->npages = 0;
@@ -511,13 +505,13 @@ int nvme_setup_prps(struct nvme_dev *dev, struct nvme_iod *iod, int total_len,
 	if (!prp_list) {
 		iod->first_dma = dma_addr;
 		iod->npages = -1;
-		return (total_len - length) + PAGE_SIZE;
+		return (total_len - length) + page_size;
 	}
 	list[0] = prp_list;
 	iod->first_dma = prp_dma;
 	i = 0;
 	for (;;) {
-		if (i == PAGE_SIZE / 8) {
+		if (i == page_size >> 3) {
 			__le64 *old_prp_list = prp_list;
 			prp_list = dma_pool_alloc(pool, gfp, &prp_dma);
 			if (!prp_list)
@@ -528,9 +522,9 @@ int nvme_setup_prps(struct nvme_dev *dev, struct nvme_iod *iod, int total_len,
 			i = 1;
 		}
 		prp_list[i++] = cpu_to_le64(dma_addr);
-		dma_len -= PAGE_SIZE;
-		dma_addr += PAGE_SIZE;
-		length -= PAGE_SIZE;
+		dma_len -= page_size;
+		dma_addr += page_size;
+		length -= page_size;
 		if (length <= 0)
 			break;
 		if (dma_len > 0)
@@ -544,88 +538,25 @@ int nvme_setup_prps(struct nvme_dev *dev, struct nvme_iod *iod, int total_len,
 	return total_len;
 }
 
-static int nvme_split_and_submit(struct bio *bio, struct nvme_queue *nvmeq,
-				 int len)
-{
-	struct bio *split = bio_split(bio, len >> 9, GFP_ATOMIC, NULL);
-	if (!split)
-		return -ENOMEM;
-
-	trace_block_split(bdev_get_queue(bio->bi_bdev), bio,
-					split->bi_iter.bi_sector);
-	bio_chain(split, bio);
-
-	if (!waitqueue_active(&nvmeq->sq_full))
-		add_wait_queue(&nvmeq->sq_full, &nvmeq->sq_cong_wait);
-	bio_list_add(&nvmeq->sq_cong, split);
-	bio_list_add(&nvmeq->sq_cong, bio);
-	wake_up(&nvmeq->sq_full);
-
-	return 0;
-}
-
-/* NVMe scatterlists require no holes in the virtual address */
-#define BIOVEC_NOT_VIRT_MERGEABLE(vec1, vec2)	((vec2)->bv_offset || \
-			(((vec1)->bv_offset + (vec1)->bv_len) % PAGE_SIZE))
-
-static int nvme_map_bio(struct nvme_queue *nvmeq, struct nvme_iod *iod,
-		struct bio *bio, enum dma_data_direction dma_dir, int psegs)
-{
-	struct bio_vec bvec, bvprv;
-	struct bvec_iter iter;
-	struct scatterlist *sg = NULL;
-	int length = 0, nsegs = 0, split_len = bio->bi_iter.bi_size;
-	int first = 1;
-
-	if (nvmeq->dev->stripe_size)
-		split_len = nvmeq->dev->stripe_size -
-			((bio->bi_iter.bi_sector << 9) &
-			 (nvmeq->dev->stripe_size - 1));
-
-	sg_init_table(iod->sg, psegs);
-	bio_for_each_segment(bvec, bio, iter) {
-		if (!first && BIOVEC_PHYS_MERGEABLE(&bvprv, &bvec)) {
-			sg->length += bvec.bv_len;
-		} else {
-			if (!first && BIOVEC_NOT_VIRT_MERGEABLE(&bvprv, &bvec))
-				return nvme_split_and_submit(bio, nvmeq,
-							     length);
-
-			sg = sg ? sg + 1 : iod->sg;
-			sg_set_page(sg, bvec.bv_page,
-				    bvec.bv_len, bvec.bv_offset);
-			nsegs++;
-		}
-
-		if (split_len - length < bvec.bv_len)
-			return nvme_split_and_submit(bio, nvmeq, split_len);
-		length += bvec.bv_len;
-		bvprv = bvec;
-		first = 0;
-	}
-	iod->nents = nsegs;
-	sg_mark_end(sg);
-	if (dma_map_sg(nvmeq->q_dmadev, iod->sg, iod->nents, dma_dir) == 0)
-		return -ENOMEM;
-
-	BUG_ON(length != bio->bi_iter.bi_size);
-	return length;
-}
-
-static int nvme_submit_discard(struct nvme_queue *nvmeq, struct nvme_ns *ns,
-		struct bio *bio, struct nvme_iod *iod, int cmdid)
+/*
+ * We reuse the small pool to allocate the 16-byte range here as it is not
+ * worth having a special pool for these or additional cases to handle freeing
+ * the iod.
+ */
+static void nvme_submit_discard(struct nvme_queue *nvmeq, struct nvme_ns *ns,
+		struct request *req, struct nvme_iod *iod)
 {
 	struct nvme_dsm_range *range =
 				(struct nvme_dsm_range *)iod_list(iod)[0];
 	struct nvme_command *cmnd = &nvmeq->sq_cmds[nvmeq->sq_tail];
 
 	range->cattr = cpu_to_le32(0);
-	range->nlb = cpu_to_le32(bio->bi_iter.bi_size >> ns->lba_shift);
-	range->slba = cpu_to_le64(nvme_block_nr(ns, bio->bi_iter.bi_sector));
+	range->nlb = cpu_to_le32(blk_rq_bytes(req) >> ns->lba_shift);
+	range->slba = cpu_to_le64(nvme_block_nr(ns, blk_rq_pos(req)));
 
 	memset(cmnd, 0, sizeof(*cmnd));
 	cmnd->dsm.opcode = nvme_cmd_dsm;
-	cmnd->dsm.command_id = cmdid;
+	cmnd->dsm.command_id = req->tag;
 	cmnd->dsm.nsid = cpu_to_le32(ns->ns_id);
 	cmnd->dsm.prp1 = cpu_to_le64(iod->first_dma);
 	cmnd->dsm.nr = 0;
@@ -634,11 +565,9 @@ static int nvme_submit_discard(struct nvme_queue *nvmeq, struct nvme_ns *ns,
 	if (++nvmeq->sq_tail == nvmeq->q_depth)
 		nvmeq->sq_tail = 0;
 	writel(nvmeq->sq_tail, nvmeq->q_db);
-
-	return 0;
 }
 
-static int nvme_submit_flush(struct nvme_queue *nvmeq, struct nvme_ns *ns,
+static void nvme_submit_flush(struct nvme_queue *nvmeq, struct nvme_ns *ns,
 								int cmdid)
 {
 	struct nvme_command *cmnd = &nvmeq->sq_cmds[nvmeq->sq_tail];
@@ -651,49 +580,34 @@ static int nvme_submit_flush(struct nvme_queue *nvmeq, struct nvme_ns *ns,
 	if (++nvmeq->sq_tail == nvmeq->q_depth)
 		nvmeq->sq_tail = 0;
 	writel(nvmeq->sq_tail, nvmeq->q_db);
-
-	return 0;
 }
 
-static int nvme_submit_iod(struct nvme_queue *nvmeq, struct nvme_iod *iod)
+static int nvme_submit_iod(struct nvme_queue *nvmeq, struct nvme_iod *iod,
+							struct nvme_ns *ns)
 {
-	struct bio *bio = iod->private;
-	struct nvme_ns *ns = bio->bi_bdev->bd_disk->private_data;
+	struct request *req = iod->private;
 	struct nvme_command *cmnd;
-	int cmdid;
-	u16 control;
-	u32 dsmgmt;
-
-	cmdid = alloc_cmdid(nvmeq, iod, bio_completion, NVME_IO_TIMEOUT);
-	if (unlikely(cmdid < 0))
-		return cmdid;
+	u16 control = 0;
+	u32 dsmgmt = 0;
 
-	if (bio->bi_rw & REQ_DISCARD)
-		return nvme_submit_discard(nvmeq, ns, bio, iod, cmdid);
-	if (bio->bi_rw & REQ_FLUSH)
-		return nvme_submit_flush(nvmeq, ns, cmdid);
-
-	control = 0;
-	if (bio->bi_rw & REQ_FUA)
+	if (req->cmd_flags & REQ_FUA)
 		control |= NVME_RW_FUA;
-	if (bio->bi_rw & (REQ_FAILFAST_DEV | REQ_RAHEAD))
+	if (req->cmd_flags & (REQ_FAILFAST_DEV | REQ_RAHEAD))
 		control |= NVME_RW_LR;
 
-	dsmgmt = 0;
-	if (bio->bi_rw & REQ_RAHEAD)
+	if (req->cmd_flags & REQ_RAHEAD)
 		dsmgmt |= NVME_RW_DSM_FREQ_PREFETCH;
 
 	cmnd = &nvmeq->sq_cmds[nvmeq->sq_tail];
 	memset(cmnd, 0, sizeof(*cmnd));
 
-	cmnd->rw.opcode = bio_data_dir(bio) ? nvme_cmd_write : nvme_cmd_read;
-	cmnd->rw.command_id = cmdid;
+	cmnd->rw.opcode = (rq_data_dir(req) ? nvme_cmd_write : nvme_cmd_read);
+	cmnd->rw.command_id = req->tag;
 	cmnd->rw.nsid = cpu_to_le32(ns->ns_id);
 	cmnd->rw.prp1 = cpu_to_le64(sg_dma_address(iod->sg));
 	cmnd->rw.prp2 = cpu_to_le64(iod->first_dma);
-	cmnd->rw.slba = cpu_to_le64(nvme_block_nr(ns, bio->bi_iter.bi_sector));
-	cmnd->rw.length =
-		cpu_to_le16((bio->bi_iter.bi_size >> ns->lba_shift) - 1);
+	cmnd->rw.slba = cpu_to_le64(nvme_block_nr(ns, blk_rq_pos(req)));
+	cmnd->rw.length = cpu_to_le16((blk_rq_bytes(req) >> ns->lba_shift) - 1);
 	cmnd->rw.control = cpu_to_le16(control);
 	cmnd->rw.dsmgmt = cpu_to_le32(dsmgmt);
 
@@ -704,45 +618,26 @@ static int nvme_submit_iod(struct nvme_queue *nvmeq, struct nvme_iod *iod)
 	return 0;
 }
 
-static int nvme_split_flush_data(struct nvme_queue *nvmeq, struct bio *bio)
-{
-	struct bio *split = bio_clone(bio, GFP_ATOMIC);
-	if (!split)
-		return -ENOMEM;
-
-	split->bi_iter.bi_size = 0;
-	split->bi_phys_segments = 0;
-	bio->bi_rw &= ~REQ_FLUSH;
-	bio_chain(split, bio);
-
-	if (!waitqueue_active(&nvmeq->sq_full))
-		add_wait_queue(&nvmeq->sq_full, &nvmeq->sq_cong_wait);
-	bio_list_add(&nvmeq->sq_cong, split);
-	bio_list_add(&nvmeq->sq_cong, bio);
-	wake_up_process(nvme_thread);
-
-	return 0;
-}
-
-/*
- * Called with local interrupts disabled and the q_lock held.  May not sleep.
- */
-static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns,
-								struct bio *bio)
+static int nvme_queue_rq(struct blk_mq_hw_ctx *hctx,
+			 const struct blk_mq_queue_data *bd)
 {
+	struct nvme_ns *ns = hctx->queue->queuedata;
+	struct nvme_queue *nvmeq = hctx->driver_data;
+	struct request *req = bd->rq;
+	struct nvme_cmd_info *cmd = blk_mq_rq_to_pdu(req);
 	struct nvme_iod *iod;
-	int psegs = bio_phys_segments(ns->queue, bio);
-	int result;
-
-	if ((bio->bi_rw & REQ_FLUSH) && psegs)
-		return nvme_split_flush_data(nvmeq, bio);
+	int psegs = req->nr_phys_segments;
+	enum dma_data_direction dma_dir;
+	unsigned size = !(req->cmd_flags & REQ_DISCARD) ? blk_rq_bytes(req) :
+						sizeof(struct nvme_dsm_range);
 
-	iod = nvme_alloc_iod(psegs, bio->bi_iter.bi_size, GFP_ATOMIC);
+	iod = nvme_alloc_iod(psegs, size, ns->dev, GFP_ATOMIC);
 	if (!iod)
-		return -ENOMEM;
+		return BLK_MQ_RQ_QUEUE_BUSY;
+
+	iod->private = req;
 
-	iod->private = bio;
-	if (bio->bi_rw & REQ_DISCARD) {
+	if (req->cmd_flags & REQ_DISCARD) {
 		void *range;
 		/*
 		 * We reuse the small pool to allocate the 16-byte range here
@@ -752,35 +647,48 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns,
 		range = dma_pool_alloc(nvmeq->dev->prp_small_pool,
 						GFP_ATOMIC,
 						&iod->first_dma);
-		if (!range) {
-			result = -ENOMEM;
-			goto free_iod;
-		}
+		if (!range)
+			goto retry_cmd;
 		iod_list(iod)[0] = (__le64 *)range;
 		iod->npages = 0;
 	} else if (psegs) {
-		result = nvme_map_bio(nvmeq, iod, bio,
-			bio_data_dir(bio) ? DMA_TO_DEVICE : DMA_FROM_DEVICE,
-			psegs);
-		if (result <= 0)
-			goto free_iod;
-		if (nvme_setup_prps(nvmeq->dev, iod, result, GFP_ATOMIC) !=
-								result) {
-			result = -ENOMEM;
-			goto free_iod;
+		dma_dir = rq_data_dir(req) ? DMA_TO_DEVICE : DMA_FROM_DEVICE;
+
+		sg_init_table(iod->sg, psegs);
+		iod->nents = blk_rq_map_sg(req->q, req, iod->sg);
+		if (!iod->nents)
+			goto error_cmd;
+
+		if (!dma_map_sg(nvmeq->q_dmadev, iod->sg, iod->nents, dma_dir))
+			goto retry_cmd;
+
+		if (blk_rq_bytes(req) !=
+                    nvme_setup_prps(nvmeq->dev, iod, blk_rq_bytes(req), GFP_ATOMIC)) {
+			dma_unmap_sg(&nvmeq->dev->pci_dev->dev, iod->sg,
+					iod->nents, dma_dir);
+			goto retry_cmd;
 		}
-		nvme_start_io_acct(bio);
 	}
-	if (unlikely(nvme_submit_iod(nvmeq, iod))) {
-		if (!waitqueue_active(&nvmeq->sq_full))
-			add_wait_queue(&nvmeq->sq_full, &nvmeq->sq_cong_wait);
-		list_add_tail(&iod->node, &nvmeq->iod_bio);
-	}
-	return 0;
 
- free_iod:
+	nvme_set_info(cmd, iod, req_completion);
+	spin_lock_irq(&nvmeq->q_lock);
+	if (req->cmd_flags & REQ_DISCARD)
+		nvme_submit_discard(nvmeq, ns, req, iod);
+	else if (req->cmd_flags & REQ_FLUSH)
+		nvme_submit_flush(nvmeq, ns, req->tag);
+	else
+		nvme_submit_iod(nvmeq, iod, ns);
+
+	nvme_process_cq(nvmeq);
+	spin_unlock_irq(&nvmeq->q_lock);
+	return BLK_MQ_RQ_QUEUE_OK;
+
+ error_cmd:
 	nvme_free_iod(nvmeq->dev, iod);
-	return result;
+	return BLK_MQ_RQ_QUEUE_ERROR;
+ retry_cmd:
+	nvme_free_iod(nvmeq->dev, iod);
+	return BLK_MQ_RQ_QUEUE_BUSY;
 }
 
 static int nvme_process_cq(struct nvme_queue *nvmeq)
@@ -801,8 +709,7 @@ static int nvme_process_cq(struct nvme_queue *nvmeq)
 			head = 0;
 			phase = !phase;
 		}
-
-		ctx = free_cmdid(nvmeq, cqe.command_id, &fn);
+		ctx = nvme_finish_cmd(nvmeq, cqe.command_id, &fn);
 		fn(nvmeq, ctx, &cqe);
 	}
 
@@ -823,29 +730,13 @@ static int nvme_process_cq(struct nvme_queue *nvmeq)
 	return 1;
 }
 
-static void nvme_make_request(struct request_queue *q, struct bio *bio)
+/* Admin queue isn't initialized as a request queue. If at some point this
+ * happens anyway, make sure to notify the user */
+static int nvme_admin_queue_rq(struct blk_mq_hw_ctx *hctx,
+			       const struct blk_mq_queue_data *bd)
 {
-	struct nvme_ns *ns = q->queuedata;
-	struct nvme_queue *nvmeq = get_nvmeq(ns->dev);
-	int result = -EBUSY;
-
-	if (!nvmeq) {
-		bio_endio(bio, -EIO);
-		return;
-	}
-
-	spin_lock_irq(&nvmeq->q_lock);
-	if (!nvmeq->q_suspended && bio_list_empty(&nvmeq->sq_cong))
-		result = nvme_submit_bio_queue(nvmeq, ns, bio);
-	if (unlikely(result)) {
-		if (!waitqueue_active(&nvmeq->sq_full))
-			add_wait_queue(&nvmeq->sq_full, &nvmeq->sq_cong_wait);
-		bio_list_add(&nvmeq->sq_cong, bio);
-	}
-
-	nvme_process_cq(nvmeq);
-	spin_unlock_irq(&nvmeq->q_lock);
-	put_nvmeq(nvmeq);
+	WARN_ON_ONCE(1);
+	return BLK_MQ_RQ_QUEUE_ERROR;
 }
 
 static irqreturn_t nvme_irq(int irq, void *data)
@@ -869,10 +760,11 @@ static irqreturn_t nvme_irq_check(int irq, void *data)
 	return IRQ_WAKE_THREAD;
 }
 
-static void nvme_abort_command(struct nvme_queue *nvmeq, int cmdid)
+static void nvme_abort_cmd_info(struct nvme_queue *nvmeq, struct nvme_cmd_info *
+								cmd_info)
 {
 	spin_lock_irq(&nvmeq->q_lock);
-	cancel_cmdid(nvmeq, cmdid, NULL);
+	cancel_cmd_info(cmd_info, NULL);
 	spin_unlock_irq(&nvmeq->q_lock);
 }
 
@@ -895,47 +787,40 @@ static void sync_completion(struct nvme_queue *nvmeq, void *ctx,
  * Returns 0 on success.  If the result is negative, it's a Linux error code;
  * if the result is positive, it's an NVM Express status code
  */
-static int nvme_submit_sync_cmd(struct nvme_dev *dev, int q_idx,
-						struct nvme_command *cmd,
+static int nvme_submit_sync_cmd(struct request *req, struct nvme_command *cmd,
 						u32 *result, unsigned timeout)
 {
-	int cmdid, ret;
+	int ret;
 	struct sync_cmd_info cmdinfo;
-	struct nvme_queue *nvmeq;
-
-	nvmeq = lock_nvmeq(dev, q_idx);
-	if (!nvmeq)
-		return -ENODEV;
+	struct nvme_cmd_info *cmd_rq = blk_mq_rq_to_pdu(req);
+	struct nvme_queue *nvmeq = cmd_rq->nvmeq;
 
 	cmdinfo.task = current;
 	cmdinfo.status = -EINTR;
 
-	cmdid = alloc_cmdid(nvmeq, &cmdinfo, sync_completion, timeout);
-	if (cmdid < 0) {
-		unlock_nvmeq(nvmeq);
-		return cmdid;
-	}
-	cmd->common.command_id = cmdid;
+	cmd->common.command_id = req->tag;
+
+	nvme_set_info(cmd_rq, &cmdinfo, sync_completion);
 
 	set_current_state(TASK_KILLABLE);
 	ret = nvme_submit_cmd(nvmeq, cmd);
 	if (ret) {
-		free_cmdid(nvmeq, cmdid, NULL);
-		unlock_nvmeq(nvmeq);
+		nvme_finish_cmd(nvmeq, req->tag, NULL);
 		set_current_state(TASK_RUNNING);
-		return ret;
 	}
-	unlock_nvmeq(nvmeq);
-	schedule_timeout(timeout);
+	ret = schedule_timeout(timeout);
 
-	if (cmdinfo.status == -EINTR) {
-		nvmeq = lock_nvmeq(dev, q_idx);
-		if (nvmeq) {
-			nvme_abort_command(nvmeq, cmdid);
-			unlock_nvmeq(nvmeq);
-		}
+	/*
+	 * Ensure that sync_completion has either run, or that it will
+	 * never run.
+	 */
+	nvme_abort_cmd_info(nvmeq, blk_mq_rq_to_pdu(req));
+
+	/*
+	 * We never got the completion
+	 */
+	if (cmdinfo.status == -EINTR)
 		return -EINTR;
-	}
 
 	if (result)
 		*result = cmdinfo.result;
@@ -943,59 +828,100 @@ static int nvme_submit_sync_cmd(struct nvme_dev *dev, int q_idx,
 	return cmdinfo.status;
 }
 
-static int nvme_submit_async_cmd(struct nvme_queue *nvmeq,
+static int nvme_submit_async_admin_req(struct nvme_dev *dev)
+{
+	struct nvme_queue *nvmeq = dev->queues[0];
+	struct nvme_command c;
+	struct nvme_cmd_info *cmd_info;
+	struct request *req;
+
+	req = blk_mq_alloc_request(dev->admin_q, WRITE, GFP_ATOMIC, false);
+	if (IS_ERR(req))
+		return PTR_ERR(req);
+
+	req->cmd_flags |= REQ_NO_TIMEOUT;
+	cmd_info = blk_mq_rq_to_pdu(req);
+	nvme_set_info(cmd_info, req, async_req_completion);
+
+	memset(&c, 0, sizeof(c));
+	c.common.opcode = nvme_admin_async_event;
+	c.common.command_id = req->tag;
+
+	return __nvme_submit_cmd(nvmeq, &c);
+}
+
+static int nvme_submit_admin_async_cmd(struct nvme_dev *dev,
 			struct nvme_command *cmd,
 			struct async_cmd_info *cmdinfo, unsigned timeout)
 {
-	int cmdid;
+	struct nvme_queue *nvmeq = dev->queues[0];
+	struct request *req;
+	struct nvme_cmd_info *cmd_rq;
+
+	req = blk_mq_alloc_request(dev->admin_q, WRITE, GFP_KERNEL, false);
+	if (IS_ERR(req))
+		return PTR_ERR(req);
 
-	cmdid = alloc_cmdid_killable(nvmeq, cmdinfo, async_completion, timeout);
-	if (cmdid < 0)
-		return cmdid;
+	req->timeout = timeout;
+	cmd_rq = blk_mq_rq_to_pdu(req);
+	cmdinfo->req = req;
+	nvme_set_info(cmd_rq, cmdinfo, async_completion);
 	cmdinfo->status = -EINTR;
-	cmd->common.command_id = cmdid;
+
+	cmd->common.command_id = req->tag;
+
 	return nvme_submit_cmd(nvmeq, cmd);
 }
 
-int nvme_submit_admin_cmd(struct nvme_dev *dev, struct nvme_command *cmd,
-								u32 *result)
+static int __nvme_submit_admin_cmd(struct nvme_dev *dev, struct nvme_command *cmd,
+						u32 *result, unsigned timeout)
 {
-	return nvme_submit_sync_cmd(dev, 0, cmd, result, ADMIN_TIMEOUT);
+	int res;
+	struct request *req;
+
+	req = blk_mq_alloc_request(dev->admin_q, WRITE, GFP_KERNEL, false);
+	if (IS_ERR(req))
+		return PTR_ERR(req);
+	res = nvme_submit_sync_cmd(req, cmd, result, timeout);
+	blk_mq_free_request(req);
+	return res;
 }
 
-int nvme_submit_io_cmd(struct nvme_dev *dev, struct nvme_command *cmd,
+int nvme_submit_admin_cmd(struct nvme_dev *dev, struct nvme_command *cmd,
 								u32 *result)
 {
-	return nvme_submit_sync_cmd(dev, smp_processor_id() + 1, cmd, result,
-							NVME_IO_TIMEOUT);
+	return __nvme_submit_admin_cmd(dev, cmd, result, ADMIN_TIMEOUT);
 }
 
-static int nvme_submit_admin_cmd_async(struct nvme_dev *dev,
-		struct nvme_command *cmd, struct async_cmd_info *cmdinfo)
+int nvme_submit_io_cmd(struct nvme_dev *dev, struct nvme_ns *ns,
+					struct nvme_command *cmd, u32 *result)
 {
-	return nvme_submit_async_cmd(raw_nvmeq(dev, 0), cmd, cmdinfo,
-								ADMIN_TIMEOUT);
+	int res;
+	struct request *req;
+
+	req = blk_mq_alloc_request(ns->queue, WRITE, (GFP_KERNEL|__GFP_WAIT),
+									false);
+	if (IS_ERR(req))
+		return PTR_ERR(req);
+	res = nvme_submit_sync_cmd(req, cmd, result, NVME_IO_TIMEOUT);
+	blk_mq_free_request(req);
+	return res;
 }
 
 static int adapter_delete_queue(struct nvme_dev *dev, u8 opcode, u16 id)
 {
-	int status;
 	struct nvme_command c;
 
 	memset(&c, 0, sizeof(c));
 	c.delete_queue.opcode = opcode;
 	c.delete_queue.qid = cpu_to_le16(id);
 
-	status = nvme_submit_admin_cmd(dev, &c, NULL);
-	if (status)
-		return -EIO;
-	return 0;
+	return nvme_submit_admin_cmd(dev, &c, NULL);
 }
 
 static int adapter_alloc_cq(struct nvme_dev *dev, u16 qid,
 						struct nvme_queue *nvmeq)
 {
-	int status;
 	struct nvme_command c;
 	int flags = NVME_QUEUE_PHYS_CONTIG | NVME_CQ_IRQ_ENABLED;
 
@@ -1007,16 +933,12 @@ static int adapter_alloc_cq(struct nvme_dev *dev, u16 qid,
 	c.create_cq.cq_flags = cpu_to_le16(flags);
 	c.create_cq.irq_vector = cpu_to_le16(nvmeq->cq_vector);
 
-	status = nvme_submit_admin_cmd(dev, &c, NULL);
-	if (status)
-		return -EIO;
-	return 0;
+	return nvme_submit_admin_cmd(dev, &c, NULL);
 }
 
 static int adapter_alloc_sq(struct nvme_dev *dev, u16 qid,
 						struct nvme_queue *nvmeq)
 {
-	int status;
 	struct nvme_command c;
 	int flags = NVME_QUEUE_PHYS_CONTIG | NVME_SQ_PRIO_MEDIUM;
 
@@ -1028,10 +950,7 @@ static int adapter_alloc_sq(struct nvme_dev *dev, u16 qid,
 	c.create_sq.sq_flags = cpu_to_le16(flags);
 	c.create_sq.cqid = cpu_to_le16(qid);
 
-	status = nvme_submit_admin_cmd(dev, &c, NULL);
-	if (status)
-		return -EIO;
-	return 0;
+	return nvme_submit_admin_cmd(dev, &c, NULL);
 }
 
 static int adapter_delete_cq(struct nvme_dev *dev, u16 cqid)
@@ -1087,151 +1006,170 @@ int nvme_set_features(struct nvme_dev *dev, unsigned fid, unsigned dword11,
 }
 
 /**
- * nvme_abort_cmd - Attempt aborting a command
- * @cmdid: Command id of a timed out IO
- * @queue: The queue with timed out IO
+ * nvme_abort_req - Attempt aborting a request
  *
  * Schedule controller reset if the command was already aborted once before and
  * still hasn't been returned to the driver, or if this is the admin queue.
  */
-static void nvme_abort_cmd(int cmdid, struct nvme_queue *nvmeq)
+static void nvme_abort_req(struct request *req)
 {
-	int a_cmdid;
-	struct nvme_command cmd;
+	struct nvme_cmd_info *cmd_rq = blk_mq_rq_to_pdu(req);
+	struct nvme_queue *nvmeq = cmd_rq->nvmeq;
 	struct nvme_dev *dev = nvmeq->dev;
-	struct nvme_cmd_info *info = nvme_cmd_info(nvmeq);
-	struct nvme_queue *adminq;
+	struct request *abort_req;
+	struct nvme_cmd_info *abort_cmd;
+	struct nvme_command cmd;
+
+	if (!nvmeq->qid || cmd_rq->aborted) {
+		unsigned long flags;
 
-	if (!nvmeq->qid || info[cmdid].aborted) {
+		spin_lock_irqsave(&dev_list_lock, flags);
 		if (work_busy(&dev->reset_work))
-			return;
+			goto out;
 		list_del_init(&dev->node);
 		dev_warn(&dev->pci_dev->dev,
-			"I/O %d QID %d timeout, reset controller\n", cmdid,
-								nvmeq->qid);
+			"I/O %d QID %d timeout, reset controller\n",
+							req->tag, nvmeq->qid);
 		dev->reset_workfn = nvme_reset_failed_dev;
 		queue_work(nvme_workq, &dev->reset_work);
+ out:
+		spin_unlock_irqrestore(&dev_list_lock, flags);
 		return;
 	}
 
 	if (!dev->abort_limit)
 		return;
 
-	adminq = rcu_dereference(dev->queues[0]);
-	a_cmdid = alloc_cmdid(adminq, CMD_CTX_ABORT, special_completion,
-								ADMIN_TIMEOUT);
-	if (a_cmdid < 0)
+	abort_req = blk_mq_alloc_request(dev->admin_q, WRITE, GFP_ATOMIC,
+									false);
+	if (IS_ERR(abort_req))
 		return;
 
+	abort_cmd = blk_mq_rq_to_pdu(abort_req);
+	nvme_set_info(abort_cmd, abort_req, abort_completion);
+
 	memset(&cmd, 0, sizeof(cmd));
 	cmd.abort.opcode = nvme_admin_abort_cmd;
-	cmd.abort.cid = cmdid;
+	cmd.abort.cid = req->tag;
 	cmd.abort.sqid = cpu_to_le16(nvmeq->qid);
-	cmd.abort.command_id = a_cmdid;
+	cmd.abort.command_id = abort_req->tag;
 
 	--dev->abort_limit;
-	info[cmdid].aborted = 1;
-	info[cmdid].timeout = jiffies + ADMIN_TIMEOUT;
+	cmd_rq->aborted = 1;
 
-	dev_warn(nvmeq->q_dmadev, "Aborting I/O %d QID %d\n", cmdid,
+	dev_warn(nvmeq->q_dmadev, "Aborting I/O %d QID %d\n", req->tag,
 							nvmeq->qid);
-	nvme_submit_cmd(adminq, &cmd);
+	if (nvme_submit_cmd(dev->queues[0], &cmd) < 0) {
+		dev_warn(nvmeq->q_dmadev,
+				"Could not abort I/O %d QID %d",
+				req->tag, nvmeq->qid);
+		blk_mq_free_request(abort_req);
+	}
 }
 
-/**
- * nvme_cancel_ios - Cancel outstanding I/Os
- * @queue: The queue to cancel I/Os on
- * @timeout: True to only cancel I/Os which have timed out
- */
-static void nvme_cancel_ios(struct nvme_queue *nvmeq, bool timeout)
+static void nvme_cancel_queue_ios(struct blk_mq_hw_ctx *hctx,
+				struct request *req, void *data, bool reserved)
 {
-	int depth = nvmeq->q_depth - 1;
-	struct nvme_cmd_info *info = nvme_cmd_info(nvmeq);
-	unsigned long now = jiffies;
-	int cmdid;
+	struct nvme_queue *nvmeq = data;
+	void *ctx;
+	nvme_completion_fn fn;
+	struct nvme_cmd_info *cmd;
+	struct nvme_completion cqe;
 
-	for_each_set_bit(cmdid, nvmeq->cmdid_data, depth) {
-		void *ctx;
-		nvme_completion_fn fn;
-		static struct nvme_completion cqe = {
-			.status = cpu_to_le16(NVME_SC_ABORT_REQ << 1),
-		};
+	if (!blk_mq_request_started(req))
+		return;
 
-		if (timeout && !time_after(now, info[cmdid].timeout))
-			continue;
-		if (info[cmdid].ctx == CMD_CTX_CANCELLED)
-			continue;
-		if (timeout && nvmeq->dev->initialized) {
-			nvme_abort_cmd(cmdid, nvmeq);
-			continue;
-		}
-		dev_warn(nvmeq->q_dmadev, "Cancelling I/O %d QID %d\n", cmdid,
-								nvmeq->qid);
-		ctx = cancel_cmdid(nvmeq, cmdid, &fn);
-		fn(nvmeq, ctx, &cqe);
-	}
+	cmd = blk_mq_rq_to_pdu(req);
+
+	if (cmd->ctx == CMD_CTX_CANCELLED)
+		return;
+
+	if (blk_queue_dying(req->q))
+		cqe.status = cpu_to_le16((NVME_SC_ABORT_REQ | NVME_SC_DNR) << 1);
+	else
+		cqe.status = cpu_to_le16(NVME_SC_ABORT_REQ << 1);
+
+
+	dev_warn(nvmeq->q_dmadev, "Cancelling I/O %d QID %d\n",
+						req->tag, nvmeq->qid);
+	ctx = cancel_cmd_info(cmd, &fn);
+	fn(nvmeq, ctx, &cqe);
 }
 
-static void nvme_free_queue(struct rcu_head *r)
+static enum blk_eh_timer_return nvme_timeout(struct request *req, bool reserved)
 {
-	struct nvme_queue *nvmeq = container_of(r, struct nvme_queue, r_head);
+	struct nvme_cmd_info *cmd = blk_mq_rq_to_pdu(req);
+	struct nvme_queue *nvmeq = cmd->nvmeq;
+
+	/*
+	 * The aborted req will be completed on receiving the abort req.
+	 * We enable the timer again. If hit twice, it'll cause a device reset,
+	 * as the device then is in a faulty state.
+	 */
+	int ret = BLK_EH_RESET_TIMER;
+
+	dev_warn(nvmeq->q_dmadev, "Timeout I/O %d QID %d\n", req->tag,
+							nvmeq->qid);
 
 	spin_lock_irq(&nvmeq->q_lock);
-	while (bio_list_peek(&nvmeq->sq_cong)) {
-		struct bio *bio = bio_list_pop(&nvmeq->sq_cong);
-		bio_endio(bio, -EIO);
-	}
-	while (!list_empty(&nvmeq->iod_bio)) {
-		static struct nvme_completion cqe = {
-			.status = cpu_to_le16(
-				(NVME_SC_ABORT_REQ | NVME_SC_DNR) << 1),
-		};
-		struct nvme_iod *iod = list_first_entry(&nvmeq->iod_bio,
-							struct nvme_iod,
-							node);
-		list_del(&iod->node);
-		bio_completion(nvmeq, iod, &cqe);
-	}
+	if (!nvmeq->dev->initialized) {
+		/*
+		 * Force cancelled command frees the request, which requires we
+		 * return BLK_EH_NOT_HANDLED.
+		 */
+		nvme_cancel_queue_ios(nvmeq->hctx, req, nvmeq, reserved);
+		ret = BLK_EH_NOT_HANDLED;
+	} else
+		nvme_abort_req(req);
 	spin_unlock_irq(&nvmeq->q_lock);
 
+	return ret;
+}
+
+static void nvme_free_queue(struct nvme_queue *nvmeq)
+{
 	dma_free_coherent(nvmeq->q_dmadev, CQ_SIZE(nvmeq->q_depth),
 				(void *)nvmeq->cqes, nvmeq->cq_dma_addr);
 	dma_free_coherent(nvmeq->q_dmadev, SQ_SIZE(nvmeq->q_depth),
 					nvmeq->sq_cmds, nvmeq->sq_dma_addr);
-	if (nvmeq->qid)
-		free_cpumask_var(nvmeq->cpu_mask);
 	kfree(nvmeq);
 }
 
 static void nvme_free_queues(struct nvme_dev *dev, int lowest)
 {
+	LLIST_HEAD(q_list);
+	struct nvme_queue *nvmeq, *next;
+	struct llist_node *entry;
 	int i;
 
 	for (i = dev->queue_count - 1; i >= lowest; i--) {
-		struct nvme_queue *nvmeq = raw_nvmeq(dev, i);
-		rcu_assign_pointer(dev->queues[i], NULL);
-		call_rcu(&nvmeq->r_head, nvme_free_queue);
+		struct nvme_queue *nvmeq = dev->queues[i];
+		llist_add(&nvmeq->node, &q_list);
 		dev->queue_count--;
+		dev->queues[i] = NULL;
 	}
+	synchronize_rcu();
+	entry = llist_del_all(&q_list);
+	llist_for_each_entry_safe(nvmeq, next, entry, node)
+		nvme_free_queue(nvmeq);
 }
 
 /**
  * nvme_suspend_queue - put queue into suspended state
  * @nvmeq - queue to suspend
- *
- * Returns 1 if already suspended, 0 otherwise.
  */
 static int nvme_suspend_queue(struct nvme_queue *nvmeq)
 {
-	int vector = nvmeq->dev->entry[nvmeq->cq_vector].vector;
+	int vector;
 
 	spin_lock_irq(&nvmeq->q_lock);
-	if (nvmeq->q_suspended) {
+	if (nvmeq->cq_vector == -1) {
 		spin_unlock_irq(&nvmeq->q_lock);
 		return 1;
 	}
-	nvmeq->q_suspended = 1;
+	vector = nvmeq->dev->entry[nvmeq->cq_vector].vector;
 	nvmeq->dev->online_queues--;
+	nvmeq->cq_vector = -1;
 	spin_unlock_irq(&nvmeq->q_lock);
 
 	irq_set_affinity_hint(vector, NULL);
@@ -1242,15 +1180,18 @@ static int nvme_suspend_queue(struct nvme_queue *nvmeq)
 
 static void nvme_clear_queue(struct nvme_queue *nvmeq)
 {
+	struct blk_mq_hw_ctx *hctx = nvmeq->hctx;
+
 	spin_lock_irq(&nvmeq->q_lock);
 	nvme_process_cq(nvmeq);
-	nvme_cancel_ios(nvmeq, false);
+	if (hctx && hctx->tags)
+		blk_mq_tag_busy_iter(hctx, nvme_cancel_queue_ios, nvmeq);
 	spin_unlock_irq(&nvmeq->q_lock);
 }
 
 static void nvme_disable_queue(struct nvme_dev *dev, int qid)
 {
-	struct nvme_queue *nvmeq = raw_nvmeq(dev, qid);
+	struct nvme_queue *nvmeq = dev->queues[qid];
 
 	if (!nvmeq)
 		return;
@@ -1263,32 +1204,29 @@ static void nvme_disable_queue(struct nvme_dev *dev, int qid)
 		adapter_delete_sq(dev, qid);
 		adapter_delete_cq(dev, qid);
 	}
+	if (!qid && dev->admin_q)
+		blk_mq_freeze_queue_start(dev->admin_q);
 	nvme_clear_queue(nvmeq);
 }
 
 static struct nvme_queue *nvme_alloc_queue(struct nvme_dev *dev, int qid,
-							int depth, int vector)
+							int depth)
 {
 	struct device *dmadev = &dev->pci_dev->dev;
-	unsigned extra = nvme_queue_extra(depth);
-	struct nvme_queue *nvmeq = kzalloc(sizeof(*nvmeq) + extra, GFP_KERNEL);
+	struct nvme_queue *nvmeq = kzalloc(sizeof(*nvmeq), GFP_KERNEL);
 	if (!nvmeq)
 		return NULL;
 
-	nvmeq->cqes = dma_alloc_coherent(dmadev, CQ_SIZE(depth),
-					&nvmeq->cq_dma_addr, GFP_KERNEL);
+	nvmeq->cqes = dma_zalloc_coherent(dmadev, CQ_SIZE(depth),
+					  &nvmeq->cq_dma_addr, GFP_KERNEL);
 	if (!nvmeq->cqes)
 		goto free_nvmeq;
-	memset((void *)nvmeq->cqes, 0, CQ_SIZE(depth));
 
 	nvmeq->sq_cmds = dma_alloc_coherent(dmadev, SQ_SIZE(depth),
 					&nvmeq->sq_dma_addr, GFP_KERNEL);
 	if (!nvmeq->sq_cmds)
 		goto free_cqdma;
 
-	if (qid && !zalloc_cpumask_var(&nvmeq->cpu_mask, GFP_KERNEL))
-		goto free_sqdma;
-
 	nvmeq->q_dmadev = dmadev;
 	nvmeq->dev = dev;
 	snprintf(nvmeq->irqname, sizeof(nvmeq->irqname), "nvme%dq%d",
@@ -1296,23 +1234,14 @@ static struct nvme_queue *nvme_alloc_queue(struct nvme_dev *dev, int qid,
 	spin_lock_init(&nvmeq->q_lock);
 	nvmeq->cq_head = 0;
 	nvmeq->cq_phase = 1;
-	init_waitqueue_head(&nvmeq->sq_full);
-	init_waitqueue_entry(&nvmeq->sq_cong_wait, nvme_thread);
-	bio_list_init(&nvmeq->sq_cong);
-	INIT_LIST_HEAD(&nvmeq->iod_bio);
 	nvmeq->q_db = &dev->dbs[qid * 2 * dev->db_stride];
 	nvmeq->q_depth = depth;
-	nvmeq->cq_vector = vector;
 	nvmeq->qid = qid;
-	nvmeq->q_suspended = 1;
 	dev->queue_count++;
-	rcu_assign_pointer(dev->queues[qid], nvmeq);
+	dev->queues[qid] = nvmeq;
 
 	return nvmeq;
 
- free_sqdma:
-	dma_free_coherent(dmadev, SQ_SIZE(depth), (void *)nvmeq->sq_cmds,
-							nvmeq->sq_dma_addr);
  free_cqdma:
 	dma_free_coherent(dmadev, CQ_SIZE(depth), (void *)nvmeq->cqes,
 							nvmeq->cq_dma_addr);
@@ -1335,17 +1264,15 @@ static int queue_request_irq(struct nvme_dev *dev, struct nvme_queue *nvmeq,
 static void nvme_init_queue(struct nvme_queue *nvmeq, u16 qid)
 {
 	struct nvme_dev *dev = nvmeq->dev;
-	unsigned extra = nvme_queue_extra(nvmeq->q_depth);
 
+	spin_lock_irq(&nvmeq->q_lock);
 	nvmeq->sq_tail = 0;
 	nvmeq->cq_head = 0;
 	nvmeq->cq_phase = 1;
 	nvmeq->q_db = &dev->dbs[qid * 2 * dev->db_stride];
-	memset(nvmeq->cmdid_data, 0, extra);
 	memset((void *)nvmeq->cqes, 0, CQ_SIZE(nvmeq->q_depth));
-	nvme_cancel_ios(nvmeq, false);
-	nvmeq->q_suspended = 0;
 	dev->online_queues++;
+	spin_unlock_irq(&nvmeq->q_lock);
 }
 
 static int nvme_create_queue(struct nvme_queue *nvmeq, int qid)
@@ -1353,6 +1280,7 @@ static int nvme_create_queue(struct nvme_queue *nvmeq, int qid)
 	struct nvme_dev *dev = nvmeq->dev;
 	int result;
 
+	nvmeq->cq_vector = qid - 1;
 	result = adapter_alloc_cq(dev, qid, nvmeq);
 	if (result < 0)
 		return result;
@@ -1365,10 +1293,7 @@ static int nvme_create_queue(struct nvme_queue *nvmeq, int qid)
 	if (result < 0)
 		goto release_sq;
 
-	spin_lock_irq(&nvmeq->q_lock);
 	nvme_init_queue(nvmeq, qid);
-	spin_unlock_irq(&nvmeq->q_lock);
-
 	return result;
 
  release_sq:
@@ -1408,27 +1333,32 @@ static int nvme_wait_ready(struct nvme_dev *dev, u64 cap, bool enabled)
  */
 static int nvme_disable_ctrl(struct nvme_dev *dev, u64 cap)
 {
-	u32 cc = readl(&dev->bar->cc);
+	dev->ctrl_config &= ~NVME_CC_SHN_MASK;
+	dev->ctrl_config &= ~NVME_CC_ENABLE;
+	writel(dev->ctrl_config, &dev->bar->cc);
 
-	if (cc & NVME_CC_ENABLE)
-		writel(cc & ~NVME_CC_ENABLE, &dev->bar->cc);
 	return nvme_wait_ready(dev, cap, false);
 }
 
 static int nvme_enable_ctrl(struct nvme_dev *dev, u64 cap)
 {
+	dev->ctrl_config &= ~NVME_CC_SHN_MASK;
+	dev->ctrl_config |= NVME_CC_ENABLE;
+	writel(dev->ctrl_config, &dev->bar->cc);
+
 	return nvme_wait_ready(dev, cap, true);
 }
 
 static int nvme_shutdown_ctrl(struct nvme_dev *dev)
 {
 	unsigned long timeout;
-	u32 cc;
 
-	cc = (readl(&dev->bar->cc) & ~NVME_CC_SHN_MASK) | NVME_CC_SHN_NORMAL;
-	writel(cc, &dev->bar->cc);
+	dev->ctrl_config &= ~NVME_CC_SHN_MASK;
+	dev->ctrl_config |= NVME_CC_SHN_NORMAL;
 
-	timeout = 2 * HZ + jiffies;
+	writel(dev->ctrl_config, &dev->bar->cc);
+
+	timeout = SHUTDOWN_TIMEOUT + jiffies;
 	while ((readl(&dev->bar->csts) & NVME_CSTS_SHST_MASK) !=
 							NVME_CSTS_SHST_CMPLT) {
 		msleep(100);
@@ -1444,20 +1374,93 @@ static int nvme_shutdown_ctrl(struct nvme_dev *dev)
 	return 0;
 }
 
+static struct blk_mq_ops nvme_mq_admin_ops = {
+	.queue_rq	= nvme_admin_queue_rq,
+	.map_queue	= blk_mq_map_queue,
+	.init_hctx	= nvme_admin_init_hctx,
+	.exit_hctx	= nvme_exit_hctx,
+	.init_request	= nvme_admin_init_request,
+	.timeout	= nvme_timeout,
+};
+
+static struct blk_mq_ops nvme_mq_ops = {
+	.queue_rq	= nvme_queue_rq,
+	.map_queue	= blk_mq_map_queue,
+	.init_hctx	= nvme_init_hctx,
+	.exit_hctx	= nvme_exit_hctx,
+	.init_request	= nvme_init_request,
+	.timeout	= nvme_timeout,
+};
+
+static void nvme_dev_remove_admin(struct nvme_dev *dev)
+{
+	if (dev->admin_q && !blk_queue_dying(dev->admin_q)) {
+		blk_cleanup_queue(dev->admin_q);
+		blk_mq_free_tag_set(&dev->admin_tagset);
+	}
+}
+
+static int nvme_alloc_admin_tags(struct nvme_dev *dev)
+{
+	if (!dev->admin_q) {
+		dev->admin_tagset.ops = &nvme_mq_admin_ops;
+		dev->admin_tagset.nr_hw_queues = 1;
+		dev->admin_tagset.queue_depth = NVME_AQ_DEPTH - 1;
+		dev->admin_tagset.timeout = ADMIN_TIMEOUT;
+		dev->admin_tagset.numa_node = dev_to_node(&dev->pci_dev->dev);
+		dev->admin_tagset.cmd_size = sizeof(struct nvme_cmd_info);
+		dev->admin_tagset.driver_data = dev;
+
+		if (blk_mq_alloc_tag_set(&dev->admin_tagset))
+			return -ENOMEM;
+
+		dev->admin_q = blk_mq_init_queue(&dev->admin_tagset);
+		if (IS_ERR(dev->admin_q)) {
+			blk_mq_free_tag_set(&dev->admin_tagset);
+			return -ENOMEM;
+		}
+		if (!blk_get_queue(dev->admin_q)) {
+			nvme_dev_remove_admin(dev);
+			return -ENODEV;
+		}
+	} else
+		blk_mq_unfreeze_queue(dev->admin_q);
+
+	return 0;
+}
+
 static int nvme_configure_admin_queue(struct nvme_dev *dev)
 {
 	int result;
 	u32 aqa;
 	u64 cap = readq(&dev->bar->cap);
 	struct nvme_queue *nvmeq;
+	unsigned page_shift = PAGE_SHIFT;
+	unsigned dev_page_min = NVME_CAP_MPSMIN(cap) + 12;
+	unsigned dev_page_max = NVME_CAP_MPSMAX(cap) + 12;
+
+	if (page_shift < dev_page_min) {
+		dev_err(&dev->pci_dev->dev,
+				"Minimum device page size (%u) too large for "
+				"host (%u)\n", 1 << dev_page_min,
+				1 << page_shift);
+		return -ENODEV;
+	}
+	if (page_shift > dev_page_max) {
+		dev_info(&dev->pci_dev->dev,
+				"Device maximum page size (%u) smaller than "
+				"host (%u); enabling work-around\n",
+				1 << dev_page_max, 1 << page_shift);
+		page_shift = dev_page_max;
+	}
 
 	result = nvme_disable_ctrl(dev, cap);
 	if (result < 0)
 		return result;
 
-	nvmeq = raw_nvmeq(dev, 0);
+	nvmeq = dev->queues[0];
 	if (!nvmeq) {
-		nvmeq = nvme_alloc_queue(dev, 0, 64, 0);
+		nvmeq = nvme_alloc_queue(dev, 0, NVME_AQ_DEPTH);
 		if (!nvmeq)
 			return -ENOMEM;
 	}
@@ -1465,27 +1468,30 @@ static int nvme_configure_admin_queue(struct nvme_dev *dev)
 	aqa = nvmeq->q_depth - 1;
 	aqa |= aqa << 16;
 
-	dev->ctrl_config = NVME_CC_ENABLE | NVME_CC_CSS_NVM;
-	dev->ctrl_config |= (PAGE_SHIFT - 12) << NVME_CC_MPS_SHIFT;
+	dev->page_size = 1 << page_shift;
+
+	dev->ctrl_config = NVME_CC_CSS_NVM;
+	dev->ctrl_config |= (page_shift - 12) << NVME_CC_MPS_SHIFT;
 	dev->ctrl_config |= NVME_CC_ARB_RR | NVME_CC_SHN_NONE;
 	dev->ctrl_config |= NVME_CC_IOSQES | NVME_CC_IOCQES;
 
 	writel(aqa, &dev->bar->aqa);
 	writeq(nvmeq->sq_dma_addr, &dev->bar->asq);
 	writeq(nvmeq->cq_dma_addr, &dev->bar->acq);
-	writel(dev->ctrl_config, &dev->bar->cc);
 
 	result = nvme_enable_ctrl(dev, cap);
 	if (result)
-		return result;
+		goto free_nvmeq;
 
+	nvmeq->cq_vector = 0;
 	result = queue_request_irq(dev, nvmeq, nvmeq->irqname);
 	if (result)
-		return result;
+		goto free_nvmeq;
 
-	spin_lock_irq(&nvmeq->q_lock);
-	nvme_init_queue(nvmeq, 0);
-	spin_unlock_irq(&nvmeq->q_lock);
+	return result;
+
+ free_nvmeq:
+	nvme_free_queues(dev, 0);
 	return result;
 }
 
@@ -1516,7 +1522,7 @@ struct nvme_iod *nvme_map_user_pages(struct nvme_dev *dev, int write,
 	}
 
 	err = -ENOMEM;
-	iod = nvme_alloc_iod(count, length, GFP_KERNEL);
+	iod = nvme_alloc_iod(count, length, dev, GFP_KERNEL);
 	if (!iod)
 		goto put_pages;
 
@@ -1644,7 +1650,7 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio)
 	if (length != (io.nblocks + 1) << ns->lba_shift)
 		status = -ENOMEM;
 	else
-		status = nvme_submit_io_cmd(dev, &c, NULL);
+		status = nvme_submit_io_cmd(dev, ns, &c, NULL);
 
 	if (meta_len) {
 		if (status == NVME_SC_SUCCESS && !(io.opcode & 1)) {
@@ -1676,10 +1682,10 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio)
 	return status;
 }
 
-static int nvme_user_admin_cmd(struct nvme_dev *dev,
-					struct nvme_admin_cmd __user *ucmd)
+static int nvme_user_cmd(struct nvme_dev *dev, struct nvme_ns *ns,
+			struct nvme_passthru_cmd __user *ucmd)
 {
-	struct nvme_admin_cmd cmd;
+	struct nvme_passthru_cmd cmd;
 	struct nvme_command c;
 	int status, length;
 	struct nvme_iod *uninitialized_var(iod);
@@ -1716,10 +1722,23 @@ static int nvme_user_admin_cmd(struct nvme_dev *dev,
 
 	timeout = cmd.timeout_ms ? msecs_to_jiffies(cmd.timeout_ms) :
 								ADMIN_TIMEOUT;
+
 	if (length != cmd.data_len)
 		status = -ENOMEM;
-	else
-		status = nvme_submit_sync_cmd(dev, 0, &c, &cmd.result, timeout);
+	else if (ns) {
+		struct request *req;
+
+		req = blk_mq_alloc_request(ns->queue, WRITE,
+						(GFP_KERNEL|__GFP_WAIT), false);
+		if (IS_ERR(req))
+			status = PTR_ERR(req);
+		else {
+			status = nvme_submit_sync_cmd(req, &c, &cmd.result,
+								timeout);
+			blk_mq_free_request(req);
+		}
+	} else
+		status = __nvme_submit_admin_cmd(dev, &c, &cmd.result, timeout);
 
 	if (cmd.data_len) {
 		nvme_unmap_user_pages(dev, cmd.opcode & 1, iod);
@@ -1743,7 +1762,9 @@ static int nvme_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd,
 		force_successful_syscall_return();
 		return ns->ns_id;
 	case NVME_IOCTL_ADMIN_CMD:
-		return nvme_user_admin_cmd(ns->dev, (void __user *)arg);
+		return nvme_user_cmd(ns->dev, NULL, (void __user *)arg);
+	case NVME_IOCTL_IO_CMD:
+		return nvme_user_cmd(ns->dev, ns, (void __user *)arg);
 	case NVME_IOCTL_SUBMIT_IO:
 		return nvme_submit_io(ns, (void __user *)arg);
 	case SG_GET_VERSION_NUM:
@@ -1759,11 +1780,9 @@ static int nvme_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd,
 static int nvme_compat_ioctl(struct block_device *bdev, fmode_t mode,
 					unsigned int cmd, unsigned long arg)
 {
-	struct nvme_ns *ns = bdev->bd_disk->private_data;
-
 	switch (cmd) {
 	case SG_IO:
-		return nvme_sg_io32(ns, arg);
+		return -ENOIOCTLCMD;
 	}
 	return nvme_ioctl(bdev, mode, cmd, arg);
 }
@@ -1773,11 +1792,18 @@ static int nvme_compat_ioctl(struct block_device *bdev, fmode_t mode,
 
 static int nvme_open(struct block_device *bdev, fmode_t mode)
 {
-	struct nvme_ns *ns = bdev->bd_disk->private_data;
-	struct nvme_dev *dev = ns->dev;
+	int ret = 0;
+	struct nvme_ns *ns;
 
-	kref_get(&dev->kref);
-	return 0;
+	spin_lock(&dev_list_lock);
+	ns = bdev->bd_disk->private_data;
+	if (!ns)
+		ret = -ENXIO;
+	else if (!kref_get_unless_zero(&ns->dev->kref))
+		ret = -ENXIO;
+	spin_unlock(&dev_list_lock);
+
+	return ret;
 }
 
 static void nvme_free_dev(struct kref *kref);
@@ -1799,6 +1825,35 @@ static int nvme_getgeo(struct block_device *bd, struct hd_geometry *geo)
 	return 0;
 }
 
+static int nvme_revalidate_disk(struct gendisk *disk)
+{
+	struct nvme_ns *ns = disk->private_data;
+	struct nvme_dev *dev = ns->dev;
+	struct nvme_id_ns *id;
+	dma_addr_t dma_addr;
+	int lbaf;
+
+	id = dma_alloc_coherent(&dev->pci_dev->dev, 4096, &dma_addr,
+								GFP_KERNEL);
+	if (!id) {
+		dev_warn(&dev->pci_dev->dev, "%s: Memory alocation failure\n",
+								__func__);
+		return 0;
+	}
+
+	if (nvme_identify(dev, ns->ns_id, 0, dma_addr))
+		goto free;
+
+	lbaf = id->flbas & 0xf;
+	ns->lba_shift = id->lbaf[lbaf].ds;
+
+	blk_queue_logical_block_size(ns->queue, 1 << ns->lba_shift);
+	set_capacity(disk, le64_to_cpup(&id->nsze) << (ns->lba_shift - 9));
+ free:
+	dma_free_coherent(&dev->pci_dev->dev, 4096, id, dma_addr);
+	return 0;
+}
+
 static const struct block_device_operations nvme_fops = {
 	.owner		= THIS_MODULE,
 	.ioctl		= nvme_ioctl,
@@ -1806,43 +1861,9 @@ static const struct block_device_operations nvme_fops = {
 	.open		= nvme_open,
 	.release	= nvme_release,
 	.getgeo		= nvme_getgeo,
+	.revalidate_disk= nvme_revalidate_disk,
 };
 
-static void nvme_resubmit_iods(struct nvme_queue *nvmeq)
-{
-	struct nvme_iod *iod, *next;
-
-	list_for_each_entry_safe(iod, next, &nvmeq->iod_bio, node) {
-		if (unlikely(nvme_submit_iod(nvmeq, iod)))
-			break;
-		list_del(&iod->node);
-		if (bio_list_empty(&nvmeq->sq_cong) &&
-						list_empty(&nvmeq->iod_bio))
-			remove_wait_queue(&nvmeq->sq_full,
-						&nvmeq->sq_cong_wait);
-	}
-}
-
-static void nvme_resubmit_bios(struct nvme_queue *nvmeq)
-{
-	while (bio_list_peek(&nvmeq->sq_cong)) {
-		struct bio *bio = bio_list_pop(&nvmeq->sq_cong);
-		struct nvme_ns *ns = bio->bi_bdev->bd_disk->private_data;
-
-		if (bio_list_empty(&nvmeq->sq_cong) &&
-						list_empty(&nvmeq->iod_bio))
-			remove_wait_queue(&nvmeq->sq_full,
-							&nvmeq->sq_cong_wait);
-		if (nvme_submit_bio_queue(nvmeq, ns, bio)) {
-			if (!waitqueue_active(&nvmeq->sq_full))
-				add_wait_queue(&nvmeq->sq_full,
-							&nvmeq->sq_cong_wait);
-			bio_list_add_head(&nvmeq->sq_cong, bio);
-			break;
-		}
-	}
-}
-
 static int nvme_kthread(void *data)
 {
 	struct nvme_dev *dev, *next;
@@ -1858,28 +1879,26 @@ static int nvme_kthread(void *data)
 					continue;
 				list_del_init(&dev->node);
 				dev_warn(&dev->pci_dev->dev,
-					"Failed status, reset controller\n");
+					"Failed status: %x, reset controller\n",
+					readl(&dev->bar->csts));
 				dev->reset_workfn = nvme_reset_failed_dev;
 				queue_work(nvme_workq, &dev->reset_work);
 				continue;
 			}
-			rcu_read_lock();
 			for (i = 0; i < dev->queue_count; i++) {
-				struct nvme_queue *nvmeq =
-						rcu_dereference(dev->queues[i]);
+				struct nvme_queue *nvmeq = dev->queues[i];
 				if (!nvmeq)
 					continue;
 				spin_lock_irq(&nvmeq->q_lock);
-				if (nvmeq->q_suspended)
-					goto unlock;
 				nvme_process_cq(nvmeq);
-				nvme_cancel_ios(nvmeq, true);
-				nvme_resubmit_bios(nvmeq);
-				nvme_resubmit_iods(nvmeq);
- unlock:
+
+				while ((i == 0) && (dev->event_limit > 0)) {
+					if (nvme_submit_async_admin_req(dev))
+						break;
+					dev->event_limit--;
+				}
 				spin_unlock_irq(&nvmeq->q_lock);
 			}
-			rcu_read_unlock();
 		}
 		spin_unlock(&dev_list_lock);
 		schedule_timeout(round_jiffies_relative(HZ));
@@ -1902,27 +1921,28 @@ static struct nvme_ns *nvme_alloc_ns(struct nvme_dev *dev, unsigned nsid,
 {
 	struct nvme_ns *ns;
 	struct gendisk *disk;
+	int node = dev_to_node(&dev->pci_dev->dev);
 	int lbaf;
 
 	if (rt->attributes & NVME_LBART_ATTRIB_HIDE)
 		return NULL;
 
-	ns = kzalloc(sizeof(*ns), GFP_KERNEL);
+	ns = kzalloc_node(sizeof(*ns), GFP_KERNEL, node);
 	if (!ns)
 		return NULL;
-	ns->queue = blk_alloc_queue(GFP_KERNEL);
-	if (!ns->queue)
+	ns->queue = blk_mq_init_queue(&dev->tagset);
+	if (IS_ERR(ns->queue))
 		goto out_free_ns;
-	ns->queue->queue_flags = QUEUE_FLAG_DEFAULT;
 	queue_flag_set_unlocked(QUEUE_FLAG_NOMERGES, ns->queue);
 	queue_flag_set_unlocked(QUEUE_FLAG_NONROT, ns->queue);
-	blk_queue_make_request(ns->queue, nvme_make_request);
+	queue_flag_set_unlocked(QUEUE_FLAG_SG_GAPS, ns->queue);
 	ns->dev = dev;
 	ns->queue->queuedata = ns;
 
-	disk = alloc_disk(0);
+	disk = alloc_disk_node(0, node);
 	if (!disk)
 		goto out_free_queue;
+
 	ns->ns_id = nsid;
 	ns->disk = disk;
 	lbaf = id->flbas & 0xf;
@@ -1931,6 +1951,8 @@ static struct nvme_ns *nvme_alloc_ns(struct nvme_dev *dev, unsigned nsid,
 	blk_queue_logical_block_size(ns->queue, 1 << ns->lba_shift);
 	if (dev->max_hw_sectors)
 		blk_queue_max_hw_sectors(ns->queue, dev->max_hw_sectors);
+	if (dev->stripe_size)
+		blk_queue_chunk_sectors(ns->queue, dev->stripe_size >> 9);
 	if (dev->vwc & NVME_CTRL_VWC_PRESENT)
 		blk_queue_flush(ns->queue, REQ_FLUSH | REQ_FUA);
 
@@ -1956,141 +1978,17 @@ static struct nvme_ns *nvme_alloc_ns(struct nvme_dev *dev, unsigned nsid,
 	return NULL;
 }
 
-static int nvme_find_closest_node(int node)
-{
-	int n, val, min_val = INT_MAX, best_node = node;
-
-	for_each_online_node(n) {
-		if (n == node)
-			continue;
-		val = node_distance(node, n);
-		if (val < min_val) {
-			min_val = val;
-			best_node = n;
-		}
-	}
-	return best_node;
-}
-
-static void nvme_set_queue_cpus(cpumask_t *qmask, struct nvme_queue *nvmeq,
-								int count)
-{
-	int cpu;
-	for_each_cpu(cpu, qmask) {
-		if (cpumask_weight(nvmeq->cpu_mask) >= count)
-			break;
-		if (!cpumask_test_and_set_cpu(cpu, nvmeq->cpu_mask))
-			*per_cpu_ptr(nvmeq->dev->io_queue, cpu) = nvmeq->qid;
-	}
-}
-
-static void nvme_add_cpus(cpumask_t *mask, const cpumask_t *unassigned_cpus,
-	const cpumask_t *new_mask, struct nvme_queue *nvmeq, int cpus_per_queue)
-{
-	int next_cpu;
-	for_each_cpu(next_cpu, new_mask) {
-		cpumask_or(mask, mask, get_cpu_mask(next_cpu));
-		cpumask_or(mask, mask, topology_thread_cpumask(next_cpu));
-		cpumask_and(mask, mask, unassigned_cpus);
-		nvme_set_queue_cpus(mask, nvmeq, cpus_per_queue);
-	}
-}
-
 static void nvme_create_io_queues(struct nvme_dev *dev)
 {
-	unsigned i, max;
-
-	max = min(dev->max_qid, num_online_cpus());
-	for (i = dev->queue_count; i <= max; i++)
-		if (!nvme_alloc_queue(dev, i, dev->q_depth, i - 1))
-			break;
+	unsigned i;
 
-	max = min(dev->queue_count - 1, num_online_cpus());
-	for (i = dev->online_queues; i <= max; i++)
-		if (nvme_create_queue(raw_nvmeq(dev, i), i))
+	for (i = dev->queue_count; i <= dev->max_qid; i++)
+		if (!nvme_alloc_queue(dev, i, dev->q_depth))
 			break;
-}
-
-/*
- * If there are fewer queues than online cpus, this will try to optimally
- * assign a queue to multiple cpus by grouping cpus that are "close" together:
- * thread siblings, core, socket, closest node, then whatever else is
- * available.
- */
-static void nvme_assign_io_queues(struct nvme_dev *dev)
-{
-	unsigned cpu, cpus_per_queue, queues, remainder, i;
-	cpumask_var_t unassigned_cpus;
-
-	nvme_create_io_queues(dev);
-
-	queues = min(dev->online_queues - 1, num_online_cpus());
-	if (!queues)
-		return;
 
-	cpus_per_queue = num_online_cpus() / queues;
-	remainder = queues - (num_online_cpus() - queues * cpus_per_queue);
-
-	if (!alloc_cpumask_var(&unassigned_cpus, GFP_KERNEL))
-		return;
-
-	cpumask_copy(unassigned_cpus, cpu_online_mask);
-	cpu = cpumask_first(unassigned_cpus);
-	for (i = 1; i <= queues; i++) {
-		struct nvme_queue *nvmeq = lock_nvmeq(dev, i);
-		cpumask_t mask;
-
-		cpumask_clear(nvmeq->cpu_mask);
-		if (!cpumask_weight(unassigned_cpus)) {
-			unlock_nvmeq(nvmeq);
+	for (i = dev->online_queues; i <= dev->queue_count - 1; i++)
+		if (nvme_create_queue(dev->queues[i], i))
 			break;
-		}
-
-		mask = *get_cpu_mask(cpu);
-		nvme_set_queue_cpus(&mask, nvmeq, cpus_per_queue);
-		if (cpus_weight(mask) < cpus_per_queue)
-			nvme_add_cpus(&mask, unassigned_cpus,
-				topology_thread_cpumask(cpu),
-				nvmeq, cpus_per_queue);
-		if (cpus_weight(mask) < cpus_per_queue)
-			nvme_add_cpus(&mask, unassigned_cpus,
-				topology_core_cpumask(cpu),
-				nvmeq, cpus_per_queue);
-		if (cpus_weight(mask) < cpus_per_queue)
-			nvme_add_cpus(&mask, unassigned_cpus,
-				cpumask_of_node(cpu_to_node(cpu)),
-				nvmeq, cpus_per_queue);
-		if (cpus_weight(mask) < cpus_per_queue)
-			nvme_add_cpus(&mask, unassigned_cpus,
-				cpumask_of_node(
-					nvme_find_closest_node(
-						cpu_to_node(cpu))),
-				nvmeq, cpus_per_queue);
-		if (cpus_weight(mask) < cpus_per_queue)
-			nvme_add_cpus(&mask, unassigned_cpus,
-				unassigned_cpus,
-				nvmeq, cpus_per_queue);
-
-		WARN(cpumask_weight(nvmeq->cpu_mask) != cpus_per_queue,
-			"nvme%d qid:%d mis-matched queue-to-cpu assignment\n",
-			dev->instance, i);
-
-		irq_set_affinity_hint(dev->entry[nvmeq->cq_vector].vector,
-							nvmeq->cpu_mask);
-		cpumask_andnot(unassigned_cpus, unassigned_cpus,
-						nvmeq->cpu_mask);
-		cpu = cpumask_next(cpu, unassigned_cpus);
-		if (remainder && !--remainder)
-			cpus_per_queue++;
-		unlock_nvmeq(nvmeq);
-	}
-	WARN(cpumask_weight(unassigned_cpus), "nvme%d unassigned online cpus\n",
-								dev->instance);
-	i = 0;
-	cpumask_andnot(unassigned_cpus, cpu_possible_mask, cpu_online_mask);
-	for_each_cpu(cpu, unassigned_cpus)
-		*per_cpu_ptr(dev->io_queue, cpu) = (i++ % queues) + 1;
-	free_cpumask_var(unassigned_cpus);
 }
 
 static int set_queue_count(struct nvme_dev *dev, int count)
@@ -2106,7 +2004,7 @@ static int set_queue_count(struct nvme_dev *dev, int count)
 	if (status > 0) {
 		dev_err(&dev->pci_dev->dev, "Could not set queue count (%d)\n",
 									status);
-		return -EBUSY;
+		return 0;
 	}
 	return min(result & 0xffff, result >> 16) + 1;
 }
@@ -2116,39 +2014,15 @@ static size_t db_bar_size(struct nvme_dev *dev, unsigned nr_io_queues)
 	return 4096 + ((nr_io_queues + 1) * 8 * dev->db_stride);
 }
 
-static void nvme_cpu_workfn(struct work_struct *work)
-{
-	struct nvme_dev *dev = container_of(work, struct nvme_dev, cpu_work);
-	if (dev->initialized)
-		nvme_assign_io_queues(dev);
-}
-
-static int nvme_cpu_notify(struct notifier_block *self,
-				unsigned long action, void *hcpu)
-{
-	struct nvme_dev *dev;
-
-	switch (action) {
-	case CPU_ONLINE:
-	case CPU_DEAD:
-		spin_lock(&dev_list_lock);
-		list_for_each_entry(dev, &dev_list, node)
-			schedule_work(&dev->cpu_work);
-		spin_unlock(&dev_list_lock);
-		break;
-	}
-	return NOTIFY_OK;
-}
-
 static int nvme_setup_io_queues(struct nvme_dev *dev)
 {
-	struct nvme_queue *adminq = raw_nvmeq(dev, 0);
+	struct nvme_queue *adminq = dev->queues[0];
 	struct pci_dev *pdev = dev->pci_dev;
 	int result, i, vecs, nr_io_queues, size;
 
 	nr_io_queues = num_possible_cpus();
 	result = set_queue_count(dev, nr_io_queues);
-	if (result < 0)
+	if (result <= 0)
 		return result;
 	if (result < nr_io_queues)
 		nr_io_queues = result;
@@ -2171,6 +2045,13 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 	/* Deregister the admin queue's interrupt */
 	free_irq(dev->entry[0].vector, adminq);
 
+	/*
+	 * If we enable msix early due to not intx, disable it again before
+	 * setting up the full range we need.
+	 */
+	if (!pdev->irq)
+		pci_disable_msix(pdev);
+
 	for (i = 0; i < nr_io_queues; i++)
 		dev->entry[i].entry = i;
 	vecs = pci_enable_msix_range(pdev, dev->entry, 1, nr_io_queues);
@@ -2194,14 +2075,12 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 	dev->max_qid = nr_io_queues;
 
 	result = queue_request_irq(dev, adminq, adminq->irqname);
-	if (result) {
-		adminq->q_suspended = 1;
+	if (result)
 		goto free_queues;
-	}
 
 	/* Free previously allocated queues that are no longer usable */
 	nvme_free_queues(dev, nr_io_queues + 1);
-	nvme_assign_io_queues(dev);
+	nvme_create_io_queues(dev);
 
 	return 0;
 
@@ -2244,14 +2123,37 @@ static int nvme_dev_add(struct nvme_dev *dev)
 	dev->oncs = le16_to_cpup(&ctrl->oncs);
 	dev->abort_limit = ctrl->acl + 1;
 	dev->vwc = ctrl->vwc;
+	dev->event_limit = min(ctrl->aerl + 1, 8);
 	memcpy(dev->serial, ctrl->sn, sizeof(ctrl->sn));
 	memcpy(dev->model, ctrl->mn, sizeof(ctrl->mn));
 	memcpy(dev->firmware_rev, ctrl->fr, sizeof(ctrl->fr));
 	if (ctrl->mdts)
 		dev->max_hw_sectors = 1 << (ctrl->mdts + shift - 9);
 	if ((pdev->vendor == PCI_VENDOR_ID_INTEL) &&
-			(pdev->device == 0x0953) && ctrl->vs[3])
+			(pdev->device == 0x0953) && ctrl->vs[3]) {
+		unsigned int max_hw_sectors;
+
 		dev->stripe_size = 1 << (ctrl->vs[3] + shift);
+		max_hw_sectors = dev->stripe_size >> (shift - 9);
+		if (dev->max_hw_sectors) {
+			dev->max_hw_sectors = min(max_hw_sectors,
+							dev->max_hw_sectors);
+		} else
+			dev->max_hw_sectors = max_hw_sectors;
+	}
+
+	dev->tagset.ops = &nvme_mq_ops;
+	dev->tagset.nr_hw_queues = dev->online_queues - 1;
+	dev->tagset.timeout = NVME_IO_TIMEOUT;
+	dev->tagset.numa_node = dev_to_node(&dev->pci_dev->dev);
+	dev->tagset.queue_depth =
+				min_t(int, dev->q_depth, BLK_MQ_MAX_DEPTH) - 1;
+	dev->tagset.cmd_size = sizeof(struct nvme_cmd_info);
+	dev->tagset.flags = BLK_MQ_F_SHOULD_MERGE;
+	dev->tagset.driver_data = dev;
+
+	if (blk_mq_alloc_tag_set(&dev->tagset))
+		goto out;
 
 	id_ns = mem;
 	for (i = 1; i <= nn; i++) {
@@ -2292,6 +2194,9 @@ static int nvme_dev_map(struct nvme_dev *dev)
 	dev->entry[0].vector = pdev->irq;
 	pci_set_master(pdev);
 	bars = pci_select_bars(pdev, IORESOURCE_MEM);
+	if (!bars)
+		goto disable_pci;
+
 	if (pci_request_selected_regions(pdev, bars, "nvme"))
 		goto disable_pci;
 
@@ -2302,10 +2207,22 @@ static int nvme_dev_map(struct nvme_dev *dev)
 	dev->bar = ioremap(pci_resource_start(pdev, 0), 8192);
 	if (!dev->bar)
 		goto disable;
+
 	if (readl(&dev->bar->csts) == -1) {
 		result = -ENODEV;
 		goto unmap;
 	}
+
+	/*
+	 * Some devices don't advertse INTx interrupts, pre-enable a single
+	 * MSIX vec for setup. We'll adjust this later.
+	 */
+	if (!pdev->irq) {
+		result = pci_enable_msix(pdev, dev->entry, 1);
+		if (result < 0)
+			goto unmap;
+	}
+
 	cap = readq(&dev->bar->cap);
 	dev->q_depth = min_t(int, NVME_CAP_MQES(cap) + 1, NVME_Q_DEPTH);
 	dev->db_stride = 1 << NVME_CAP_STRIDE(cap);
@@ -2357,13 +2274,18 @@ static void nvme_wait_dq(struct nvme_delq_ctx *dq, struct nvme_dev *dev)
 			break;
 		if (!schedule_timeout(ADMIN_TIMEOUT) ||
 					fatal_signal_pending(current)) {
+			/*
+			 * Disable the controller first since we can't trust it
+			 * at this point, but leave the admin queue enabled
+			 * until all queue deletion requests are flushed.
+			 * FIXME: This may take a while if there are more h/w
+			 * queues than admin tags.
+			 */
 			set_current_state(TASK_RUNNING);
-
 			nvme_disable_ctrl(dev, readq(&dev->bar->cap));
-			nvme_disable_queue(dev, 0);
-
-			send_sig(SIGKILL, dq->worker->task, 1);
+			nvme_clear_queue(dev->queues[0]);
 			flush_kthread_worker(dq->worker);
+			nvme_disable_queue(dev, 0);
 			return;
 		}
 	}
@@ -2401,7 +2323,8 @@ static int adapter_async_del_queue(struct nvme_queue *nvmeq, u8 opcode,
 	c.delete_queue.qid = cpu_to_le16(nvmeq->qid);
 
 	init_kthread_work(&nvmeq->cmdinfo.work, fn);
-	return nvme_submit_admin_cmd_async(nvmeq->dev, &c, &nvmeq->cmdinfo);
+	return nvme_submit_admin_async_cmd(nvmeq->dev, &c, &nvmeq->cmdinfo,
+								ADMIN_TIMEOUT);
 }
 
 static void nvme_del_cq_work_handler(struct kthread_work *work)
@@ -2439,7 +2362,6 @@ static void nvme_del_queue_start(struct kthread_work *work)
 {
 	struct nvme_queue *nvmeq = container_of(work, struct nvme_queue,
 							cmdinfo.work);
-	allow_signal(SIGKILL);
 	if (nvme_delete_sq(nvmeq))
 		nvme_del_queue_end(nvmeq);
 }
@@ -2464,7 +2386,7 @@ static void nvme_disable_io_queues(struct nvme_dev *dev)
 	atomic_set(&dq.refcount, 0);
 	dq.worker = &worker;
 	for (i = dev->queue_count - 1; i > 0; i--) {
-		struct nvme_queue *nvmeq = raw_nvmeq(dev, i);
+		struct nvme_queue *nvmeq = dev->queues[i];
 
 		if (nvme_suspend_queue(nvmeq))
 			continue;
@@ -2497,16 +2419,49 @@ static void nvme_dev_list_remove(struct nvme_dev *dev)
 		kthread_stop(tmp);
 }
 
+static void nvme_freeze_queues(struct nvme_dev *dev)
+{
+	struct nvme_ns *ns;
+
+	list_for_each_entry(ns, &dev->namespaces, list) {
+		blk_mq_freeze_queue_start(ns->queue);
+
+		spin_lock(ns->queue->queue_lock);
+		queue_flag_set(QUEUE_FLAG_STOPPED, ns->queue);
+		spin_unlock(ns->queue->queue_lock);
+
+		blk_mq_cancel_requeue_work(ns->queue);
+		blk_mq_stop_hw_queues(ns->queue);
+	}
+}
+
+static void nvme_unfreeze_queues(struct nvme_dev *dev)
+{
+	struct nvme_ns *ns;
+
+	list_for_each_entry(ns, &dev->namespaces, list) {
+		queue_flag_clear_unlocked(QUEUE_FLAG_STOPPED, ns->queue);
+		blk_mq_unfreeze_queue(ns->queue);
+		blk_mq_start_stopped_hw_queues(ns->queue, true);
+		blk_mq_kick_requeue_list(ns->queue);
+	}
+}
+
 static void nvme_dev_shutdown(struct nvme_dev *dev)
 {
 	int i;
+	u32 csts = -1;
 
 	dev->initialized = 0;
 	nvme_dev_list_remove(dev);
 
-	if (!dev->bar || (dev->bar && readl(&dev->bar->csts) == -1)) {
+	if (dev->bar) {
+		nvme_freeze_queues(dev);
+		csts = readl(&dev->bar->csts);
+	}
+	if (csts & NVME_CSTS_CFS || !(csts & NVME_CSTS_RDY)) {
 		for (i = dev->queue_count - 1; i >= 0; i--) {
-			struct nvme_queue *nvmeq = raw_nvmeq(dev, i);
+			struct nvme_queue *nvmeq = dev->queues[i];
 			nvme_suspend_queue(nvmeq);
 			nvme_clear_queue(nvmeq);
 		}
@@ -2525,8 +2480,10 @@ static void nvme_dev_remove(struct nvme_dev *dev)
 	list_for_each_entry(ns, &dev->namespaces, list) {
 		if (ns->disk->flags & GENHD_FL_UP)
 			del_gendisk(ns->disk);
-		if (!blk_queue_dying(ns->queue))
+		if (!blk_queue_dying(ns->queue)) {
+			blk_mq_abort_requeue_list(ns->queue);
 			blk_cleanup_queue(ns->queue);
+		}
 	}
 }
 
@@ -2589,6 +2546,11 @@ static void nvme_free_namespaces(struct nvme_dev *dev)
 
 	list_for_each_entry_safe(ns, next, &dev->namespaces, list) {
 		list_del(&ns->list);
+
+		spin_lock(&dev_list_lock);
+		ns->disk->private_data = NULL;
+		spin_unlock(&dev_list_lock);
+
 		put_disk(ns->disk);
 		kfree(ns);
 	}
@@ -2598,8 +2560,11 @@ static void nvme_free_dev(struct kref *kref)
 {
 	struct nvme_dev *dev = container_of(kref, struct nvme_dev, kref);
 
+	pci_dev_put(dev->pci_dev);
 	nvme_free_namespaces(dev);
-	free_percpu(dev->io_queue);
+	nvme_release_instance(dev);
+	blk_mq_free_tag_set(&dev->tagset);
+	blk_put_queue(dev->admin_q);
 	kfree(dev->queues);
 	kfree(dev->entry);
 	kfree(dev);
@@ -2624,9 +2589,16 @@ static int nvme_dev_release(struct inode *inode, struct file *f)
 static long nvme_dev_ioctl(struct file *f, unsigned int cmd, unsigned long arg)
 {
 	struct nvme_dev *dev = f->private_data;
+	struct nvme_ns *ns;
+
 	switch (cmd) {
 	case NVME_IOCTL_ADMIN_CMD:
-		return nvme_user_admin_cmd(dev, (void __user *)arg);
+		return nvme_user_cmd(dev, NULL, (void __user *)arg);
+	case NVME_IOCTL_IO_CMD:
+		if (list_empty(&dev->namespaces))
+			return -ENOTTY;
+		ns = list_first_entry(&dev->namespaces, struct nvme_ns, list);
+		return nvme_user_cmd(dev, ns, (void __user *)arg);
 	default:
 		return -ENOTTY;
 	}
@@ -2640,6 +2612,22 @@ static const struct file_operations nvme_dev_fops = {
 	.compat_ioctl	= nvme_dev_ioctl,
 };
 
+static void nvme_set_irq_hints(struct nvme_dev *dev)
+{
+	struct nvme_queue *nvmeq;
+	int i;
+
+	for (i = 0; i < dev->online_queues; i++) {
+		nvmeq = dev->queues[i];
+
+		if (!nvmeq->hctx)
+			continue;
+
+		irq_set_affinity_hint(dev->entry[nvmeq->cq_vector].vector,
+							nvmeq->hctx->cpumask);
+	}
+}
+
 static int nvme_dev_start(struct nvme_dev *dev)
 {
 	int result;
@@ -2663,7 +2651,7 @@ static int nvme_dev_start(struct nvme_dev *dev)
 
 	if (start_thread) {
 		nvme_thread = kthread_run(nvme_kthread, NULL, "nvme");
-		wake_up(&nvme_kthread_wait);
+		wake_up_all(&nvme_kthread_wait);
 	} else
 		wait_event_killable(nvme_kthread_wait, nvme_thread);
 
@@ -2672,12 +2660,21 @@ static int nvme_dev_start(struct nvme_dev *dev)
 		goto disable;
 	}
 
-	result = nvme_setup_io_queues(dev);
-	if (result && result != -EBUSY)
+	nvme_init_queue(dev->queues[0], 0);
+	result = nvme_alloc_admin_tags(dev);
+	if (result)
 		goto disable;
 
+	result = nvme_setup_io_queues(dev);
+	if (result)
+		goto free_tags;
+
+	nvme_set_irq_hints(dev);
+
 	return result;
 
+ free_tags:
+	nvme_dev_remove_admin(dev);
  disable:
 	nvme_disable_queue(dev, 0);
 	nvme_dev_list_remove(dev);
@@ -2692,7 +2689,7 @@ static int nvme_remove_dead_ctrl(void *arg)
 	struct pci_dev *pdev = dev->pci_dev;
 
 	if (pci_get_drvdata(pdev))
-		pci_stop_and_remove_bus_device(pdev);
+		pci_stop_and_remove_bus_device_locked(pdev);
 	kref_put(&dev->kref, nvme_free_dev);
 	return 0;
 }
@@ -2701,8 +2698,8 @@ static void nvme_remove_disks(struct work_struct *ws)
 {
 	struct nvme_dev *dev = container_of(ws, struct nvme_dev, reset_work);
 
-	nvme_dev_remove(dev);
 	nvme_free_queues(dev, 1);
+	nvme_dev_remove(dev);
 }
 
 static int nvme_dev_resume(struct nvme_dev *dev)
@@ -2710,13 +2707,16 @@ static int nvme_dev_resume(struct nvme_dev *dev)
 	int ret;
 
 	ret = nvme_dev_start(dev);
-	if (ret && ret != -EBUSY)
+	if (ret)
 		return ret;
-	if (ret == -EBUSY) {
+	if (dev->online_queues < 2) {
 		spin_lock(&dev_list_lock);
 		dev->reset_workfn = nvme_remove_disks;
 		queue_work(nvme_workq, &dev->reset_work);
 		spin_unlock(&dev_list_lock);
+	} else {
+		nvme_unfreeze_queues(dev);
+		nvme_set_irq_hints(dev);
 	}
 	dev->initialized = 1;
 	return 0;
@@ -2726,7 +2726,7 @@ static void nvme_dev_reset(struct nvme_dev *dev)
 {
 	nvme_dev_shutdown(dev);
 	if (nvme_dev_resume(dev)) {
-		dev_err(&dev->pci_dev->dev, "Device failed to resume\n");
+		dev_warn(&dev->pci_dev->dev, "Device failed to resume\n");
 		kref_get(&dev->kref);
 		if (IS_ERR(kthread_run(nvme_remove_dead_ctrl, dev, "nvme%d",
 							dev->instance))) {
@@ -2751,33 +2751,33 @@ static void nvme_reset_workfn(struct work_struct *work)
 
 static int nvme_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 {
-	int result = -ENOMEM;
+	int node, result = -ENOMEM;
 	struct nvme_dev *dev;
 
-	dev = kzalloc(sizeof(*dev), GFP_KERNEL);
+	node = dev_to_node(&pdev->dev);
+	if (node == NUMA_NO_NODE)
+		set_dev_node(&pdev->dev, 0);
+
+	dev = kzalloc_node(sizeof(*dev), GFP_KERNEL, node);
 	if (!dev)
 		return -ENOMEM;
-	dev->entry = kcalloc(num_possible_cpus(), sizeof(*dev->entry),
-								GFP_KERNEL);
+	dev->entry = kzalloc_node(num_possible_cpus() * sizeof(*dev->entry),
+							GFP_KERNEL, node);
 	if (!dev->entry)
 		goto free;
-	dev->queues = kcalloc(num_possible_cpus() + 1, sizeof(void *),
-								GFP_KERNEL);
+	dev->queues = kzalloc_node((num_possible_cpus() + 1) * sizeof(void *),
+							GFP_KERNEL, node);
 	if (!dev->queues)
 		goto free;
-	dev->io_queue = alloc_percpu(unsigned short);
-	if (!dev->io_queue)
-		goto free;
 
 	INIT_LIST_HEAD(&dev->namespaces);
 	dev->reset_workfn = nvme_reset_failed_dev;
 	INIT_WORK(&dev->reset_work, nvme_reset_workfn);
-	INIT_WORK(&dev->cpu_work, nvme_cpu_workfn);
-	dev->pci_dev = pdev;
+	dev->pci_dev = pci_dev_get(pdev);
 	pci_set_drvdata(pdev, dev);
 	result = nvme_set_instance(dev);
 	if (result)
-		goto free;
+		goto put_pci;
 
 	result = nvme_setup_prp_pools(dev);
 	if (result)
@@ -2785,17 +2785,14 @@ static int nvme_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 
 	kref_init(&dev->kref);
 	result = nvme_dev_start(dev);
-	if (result) {
-		if (result == -EBUSY)
-			goto create_cdev;
+	if (result)
 		goto release_pools;
-	}
 
-	result = nvme_dev_add(dev);
+	if (dev->online_queues > 1)
+		result = nvme_dev_add(dev);
 	if (result)
 		goto shutdown;
 
- create_cdev:
 	scnprintf(dev->name, sizeof(dev->name), "nvme%d", dev->instance);
 	dev->miscdev.minor = MISC_DYNAMIC_MINOR;
 	dev->miscdev.parent = &pdev->dev;
@@ -2805,11 +2802,14 @@ static int nvme_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	if (result)
 		goto remove;
 
+	nvme_set_irq_hints(dev);
+
 	dev->initialized = 1;
 	return 0;
 
  remove:
 	nvme_dev_remove(dev);
+	nvme_dev_remove_admin(dev);
 	nvme_free_namespaces(dev);
  shutdown:
 	nvme_dev_shutdown(dev);
@@ -2818,8 +2818,9 @@ static int nvme_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	nvme_release_prp_pools(dev);
  release:
 	nvme_release_instance(dev);
+ put_pci:
+	pci_dev_put(dev->pci_dev);
  free:
-	free_percpu(dev->io_queue);
 	kfree(dev->queues);
 	kfree(dev->entry);
 	kfree(dev);
@@ -2828,12 +2829,12 @@ static int nvme_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 
 static void nvme_reset_notify(struct pci_dev *pdev, bool prepare)
 {
-       struct nvme_dev *dev = pci_get_drvdata(pdev);
+	struct nvme_dev *dev = pci_get_drvdata(pdev);
 
-       if (prepare)
-               nvme_dev_shutdown(dev);
-       else
-               nvme_dev_resume(dev);
+	if (prepare)
+		nvme_dev_shutdown(dev);
+	else
+		nvme_dev_resume(dev);
 }
 
 static void nvme_shutdown(struct pci_dev *pdev)
@@ -2852,13 +2853,11 @@ static void nvme_remove(struct pci_dev *pdev)
 
 	pci_set_drvdata(pdev, NULL);
 	flush_work(&dev->reset_work);
-	flush_work(&dev->cpu_work);
 	misc_deregister(&dev->miscdev);
-	nvme_dev_remove(dev);
 	nvme_dev_shutdown(dev);
+	nvme_dev_remove(dev);
+	nvme_dev_remove_admin(dev);
 	nvme_free_queues(dev, 0);
-	rcu_barrier();
-	nvme_release_instance(dev);
 	nvme_release_prp_pools(dev);
 	kref_put(&dev->kref, nvme_free_dev);
 }
@@ -2941,18 +2940,11 @@ static int __init nvme_init(void)
 	else if (result > 0)
 		nvme_major = result;
 
-	nvme_nb.notifier_call = &nvme_cpu_notify;
-	result = register_hotcpu_notifier(&nvme_nb);
-	if (result)
-		goto unregister_blkdev;
-
 	result = pci_register_driver(&nvme_driver);
 	if (result)
-		goto unregister_hotcpu;
+		goto unregister_blkdev;
 	return 0;
 
- unregister_hotcpu:
-	unregister_hotcpu_notifier(&nvme_nb);
  unregister_blkdev:
 	unregister_blkdev(nvme_major, "nvme");
  kill_workq:
@@ -2972,6 +2964,6 @@ static void __exit nvme_exit(void)
 
 MODULE_AUTHOR("Matthew Wilcox <willy@linux.intel.com>");
 MODULE_LICENSE("GPL");
-MODULE_VERSION("0.9");
+MODULE_VERSION("1.0");
 module_init(nvme_init);
 module_exit(nvme_exit);
diff --git a/drivers/block/nvme-scsi.c b/drivers/block/nvme-scsi.c
index a4cd6d691c63..5e78568026c3 100644
--- a/drivers/block/nvme-scsi.c
+++ b/drivers/block/nvme-scsi.c
@@ -329,7 +329,7 @@ INQUIRY_EVPD_BIT_MASK) ? 1 : 0)
 (GET_U32_FROM_CDB(cdb, READ_CAP_16_CDB_ALLOC_LENGTH_OFFSET))
 
 #define IS_READ_CAP_16(cdb)					\
-((cdb[0] == SERVICE_ACTION_IN && cdb[1] == SAI_READ_CAPACITY_16) ? 1 : 0)
+((cdb[0] == SERVICE_ACTION_IN_16 && cdb[1] == SAI_READ_CAPACITY_16) ? 1 : 0)
 
 /* Request Sense Helper Macros */
 #define GET_REQUEST_SENSE_ALLOC_LENGTH(cdb)			\
@@ -2105,7 +2105,7 @@ static int nvme_trans_do_nvme_io(struct nvme_ns *ns, struct sg_io_hdr *hdr,
 
 		nvme_offset += unit_num_blocks;
 
-		nvme_sc = nvme_submit_io_cmd(dev, &c, NULL);
+		nvme_sc = nvme_submit_io_cmd(dev, ns, &c, NULL);
 		if (nvme_sc != NVME_SC_SUCCESS) {
 			nvme_unmap_user_pages(dev,
 				(is_write) ? DMA_TO_DEVICE : DMA_FROM_DEVICE,
@@ -2658,7 +2658,7 @@ static int nvme_trans_start_stop(struct nvme_ns *ns, struct sg_io_hdr *hdr,
 			c.common.opcode = nvme_cmd_flush;
 			c.common.nsid = cpu_to_le32(ns->ns_id);
 
-			nvme_sc = nvme_submit_io_cmd(ns->dev, &c, NULL);
+			nvme_sc = nvme_submit_io_cmd(ns->dev, ns, &c, NULL);
 			res = nvme_trans_status_code(hdr, nvme_sc);
 			if (res)
 				goto out;
@@ -2686,7 +2686,7 @@ static int nvme_trans_synchronize_cache(struct nvme_ns *ns,
 	c.common.opcode = nvme_cmd_flush;
 	c.common.nsid = cpu_to_le32(ns->ns_id);
 
-	nvme_sc = nvme_submit_io_cmd(ns->dev, &c, NULL);
+	nvme_sc = nvme_submit_io_cmd(ns->dev, ns, &c, NULL);
 
 	res = nvme_trans_status_code(hdr, nvme_sc);
 	if (res)
@@ -2894,7 +2894,7 @@ static int nvme_trans_unmap(struct nvme_ns *ns, struct sg_io_hdr *hdr,
 	c.dsm.nr = cpu_to_le32(ndesc - 1);
 	c.dsm.attributes = cpu_to_le32(NVME_DSMGMT_AD);
 
-	nvme_sc = nvme_submit_io_cmd(dev, &c, NULL);
+	nvme_sc = nvme_submit_io_cmd(dev, ns, &c, NULL);
 	res = nvme_trans_status_code(hdr, nvme_sc);
 
 	dma_free_coherent(&dev->pci_dev->dev, ndesc * sizeof(*range),
@@ -2915,6 +2915,14 @@ static int nvme_scsi_translate(struct nvme_ns *ns, struct sg_io_hdr *hdr)
 	if (copy_from_user(cmd, hdr->cmdp, hdr->cmd_len))
 		return -EFAULT;
 
+	/*
+	 * Prime the hdr with good status for scsi commands that don't require
+	 * an nvme command for translation.
+	 */
+	retcode = nvme_trans_status_code(hdr, NVME_SC_SUCCESS);
+	if (retcode)
+		return retcode;
+
 	opcode = cmd[0];
 
 	switch (opcode) {
@@ -2947,7 +2955,7 @@ static int nvme_scsi_translate(struct nvme_ns *ns, struct sg_io_hdr *hdr)
 	case READ_CAPACITY:
 		retcode = nvme_trans_read_capacity(ns, hdr, cmd);
 		break;
-	case SERVICE_ACTION_IN:
+	case SERVICE_ACTION_IN_16:
 		if (IS_READ_CAP_16(cmd))
 			retcode = nvme_trans_read_capacity(ns, hdr, cmd);
 		else
@@ -3016,152 +3024,6 @@ int nvme_sg_io(struct nvme_ns *ns, struct sg_io_hdr __user *u_hdr)
 	return retcode;
 }
 
-#ifdef CONFIG_COMPAT
-typedef struct sg_io_hdr32 {
-	compat_int_t interface_id;	/* [i] 'S' for SCSI generic (required) */
-	compat_int_t dxfer_direction;	/* [i] data transfer direction  */
-	unsigned char cmd_len;		/* [i] SCSI command length ( <= 16 bytes) */
-	unsigned char mx_sb_len;		/* [i] max length to write to sbp */
-	unsigned short iovec_count;	/* [i] 0 implies no scatter gather */
-	compat_uint_t dxfer_len;		/* [i] byte count of data transfer */
-	compat_uint_t dxferp;		/* [i], [*io] points to data transfer memory
-					      or scatter gather list */
-	compat_uptr_t cmdp;		/* [i], [*i] points to command to perform */
-	compat_uptr_t sbp;		/* [i], [*o] points to sense_buffer memory */
-	compat_uint_t timeout;		/* [i] MAX_UINT->no timeout (unit: millisec) */
-	compat_uint_t flags;		/* [i] 0 -> default, see SG_FLAG... */
-	compat_int_t pack_id;		/* [i->o] unused internally (normally) */
-	compat_uptr_t usr_ptr;		/* [i->o] unused internally */
-	unsigned char status;		/* [o] scsi status */
-	unsigned char masked_status;	/* [o] shifted, masked scsi status */
-	unsigned char msg_status;		/* [o] messaging level data (optional) */
-	unsigned char sb_len_wr;		/* [o] byte count actually written to sbp */
-	unsigned short host_status;	/* [o] errors from host adapter */
-	unsigned short driver_status;	/* [o] errors from software driver */
-	compat_int_t resid;		/* [o] dxfer_len - actual_transferred */
-	compat_uint_t duration;		/* [o] time taken by cmd (unit: millisec) */
-	compat_uint_t info;		/* [o] auxiliary information */
-} sg_io_hdr32_t;  /* 64 bytes long (on sparc32) */
-
-typedef struct sg_iovec32 {
-	compat_uint_t iov_base;
-	compat_uint_t iov_len;
-} sg_iovec32_t;
-
-static int sg_build_iovec(sg_io_hdr_t __user *sgio, void __user *dxferp, u16 iovec_count)
-{
-	sg_iovec_t __user *iov = (sg_iovec_t __user *) (sgio + 1);
-	sg_iovec32_t __user *iov32 = dxferp;
-	int i;
-
-	for (i = 0; i < iovec_count; i++) {
-		u32 base, len;
-
-		if (get_user(base, &iov32[i].iov_base) ||
-		    get_user(len, &iov32[i].iov_len) ||
-		    put_user(compat_ptr(base), &iov[i].iov_base) ||
-		    put_user(len, &iov[i].iov_len))
-			return -EFAULT;
-	}
-
-	if (put_user(iov, &sgio->dxferp))
-		return -EFAULT;
-	return 0;
-}
-
-int nvme_sg_io32(struct nvme_ns *ns, unsigned long arg)
-{
-	sg_io_hdr32_t __user *sgio32 = (sg_io_hdr32_t __user *)arg;
-	sg_io_hdr_t __user *sgio;
-	u16 iovec_count;
-	u32 data;
-	void __user *dxferp;
-	int err;
-	int interface_id;
-
-	if (get_user(interface_id, &sgio32->interface_id))
-		return -EFAULT;
-	if (interface_id != 'S')
-		return -EINVAL;
-
-	if (get_user(iovec_count, &sgio32->iovec_count))
-		return -EFAULT;
-
-	{
-		void __user *top = compat_alloc_user_space(0);
-		void __user *new = compat_alloc_user_space(sizeof(sg_io_hdr_t) +
-				       (iovec_count * sizeof(sg_iovec_t)));
-		if (new > top)
-			return -EINVAL;
-
-		sgio = new;
-	}
-
-	/* Ok, now construct.  */
-	if (copy_in_user(&sgio->interface_id, &sgio32->interface_id,
-			 (2 * sizeof(int)) +
-			 (2 * sizeof(unsigned char)) +
-			 (1 * sizeof(unsigned short)) +
-			 (1 * sizeof(unsigned int))))
-		return -EFAULT;
-
-	if (get_user(data, &sgio32->dxferp))
-		return -EFAULT;
-	dxferp = compat_ptr(data);
-	if (iovec_count) {
-		if (sg_build_iovec(sgio, dxferp, iovec_count))
-			return -EFAULT;
-	} else {
-		if (put_user(dxferp, &sgio->dxferp))
-			return -EFAULT;
-	}
-
-	{
-		unsigned char __user *cmdp;
-		unsigned char __user *sbp;
-
-		if (get_user(data, &sgio32->cmdp))
-			return -EFAULT;
-		cmdp = compat_ptr(data);
-
-		if (get_user(data, &sgio32->sbp))
-			return -EFAULT;
-		sbp = compat_ptr(data);
-
-		if (put_user(cmdp, &sgio->cmdp) ||
-		    put_user(sbp, &sgio->sbp))
-			return -EFAULT;
-	}
-
-	if (copy_in_user(&sgio->timeout, &sgio32->timeout,
-			 3 * sizeof(int)))
-		return -EFAULT;
-
-	if (get_user(data, &sgio32->usr_ptr))
-		return -EFAULT;
-	if (put_user(compat_ptr(data), &sgio->usr_ptr))
-		return -EFAULT;
-
-	err = nvme_sg_io(ns, sgio);
-	if (err >= 0) {
-		void __user *datap;
-
-		if (copy_in_user(&sgio32->pack_id, &sgio->pack_id,
-				 sizeof(int)) ||
-		    get_user(datap, &sgio->usr_ptr) ||
-		    put_user((u32)(unsigned long)datap,
-			     &sgio32->usr_ptr) ||
-		    copy_in_user(&sgio32->status, &sgio->status,
-				 (4 * sizeof(unsigned char)) +
-				 (2 * sizeof(unsigned short)) +
-				 (3 * sizeof(int))))
-			err = -EFAULT;
-	}
-
-	return err;
-}
-#endif
-
 int nvme_sg_get_version_num(int __user *ip)
 {
 	return put_user(sg_version_num, ip);
diff --git a/drivers/block/paride/pcd.c b/drivers/block/paride/pcd.c
index 719cb1bc1640..3b7c9f1be484 100644
--- a/drivers/block/paride/pcd.c
+++ b/drivers/block/paride/pcd.c
@@ -69,8 +69,8 @@
             nice        This parameter controls the driver's use of
                         idle CPU time, at the expense of some speed.
  
-	If this driver is built into the kernel, you can use kernel
-        the following command line parameters, with the same values
+	If this driver is built into the kernel, you can use the
+        following kernel command line parameters, with the same values
         as the corresponding module parameters listed above:
 
 	    pcd.drive0
diff --git a/drivers/block/paride/pd.c b/drivers/block/paride/pd.c
index fea7e76a00de..d48715b287e6 100644
--- a/drivers/block/paride/pd.c
+++ b/drivers/block/paride/pd.c
@@ -722,6 +722,8 @@ static int pd_special_command(struct pd_unit *disk,
 	int err = 0;
 
 	rq = blk_get_request(disk->gd->queue, READ, __GFP_WAIT);
+	if (IS_ERR(rq))
+		return PTR_ERR(rq);
 
 	rq->cmd_type = REQ_TYPE_SPECIAL;
 	rq->special = func;
diff --git a/drivers/block/pktcdvd.c b/drivers/block/pktcdvd.c
index 758ac442c5b5..09e628dafd9d 100644
--- a/drivers/block/pktcdvd.c
+++ b/drivers/block/pktcdvd.c
@@ -704,6 +704,8 @@ static int pkt_generic_packet(struct pktcdvd_device *pd, struct packet_command *
 
 	rq = blk_get_request(q, (cgc->data_direction == CGC_DATA_WRITE) ?
 			     WRITE : READ, __GFP_WAIT);
+	if (IS_ERR(rq))
+		return PTR_ERR(rq);
 	blk_rq_set_block_pc(rq);
 
 	if (cgc->buflen) {
diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c
index 4b97baf8afa3..8a86b62466f7 100644
--- a/drivers/block/rbd.c
+++ b/drivers/block/rbd.c
@@ -210,6 +210,12 @@ enum obj_request_type {
 	OBJ_REQUEST_NODATA, OBJ_REQUEST_BIO, OBJ_REQUEST_PAGES
 };
 
+enum obj_operation_type {
+	OBJ_OP_WRITE,
+	OBJ_OP_READ,
+	OBJ_OP_DISCARD,
+};
+
 enum obj_req_flags {
 	OBJ_REQ_DONE,		/* completion flag: not done = 0, done = 1 */
 	OBJ_REQ_IMG_DATA,	/* object usage: standalone = 0, image = 1 */
@@ -276,6 +282,7 @@ enum img_req_flags {
 	IMG_REQ_WRITE,		/* I/O direction: read = 0, write = 1 */
 	IMG_REQ_CHILD,		/* initiator: block = 0, child image = 1 */
 	IMG_REQ_LAYERED,	/* ENOENT handling: normal = 0, layered = 1 */
+	IMG_REQ_DISCARD,	/* discard: normal = 0, discard request = 1 */
 };
 
 struct rbd_img_request {
@@ -335,7 +342,6 @@ struct rbd_device {
 
 	struct list_head	rq_queue;	/* incoming rq queue */
 	spinlock_t		lock;		/* queue, flags, open_count */
-	struct workqueue_struct	*rq_wq;
 	struct work_struct	rq_work;
 
 	struct rbd_image_header	header;
@@ -395,6 +401,8 @@ static struct kmem_cache	*rbd_segment_name_cache;
 static int rbd_major;
 static DEFINE_IDA(rbd_dev_id_ida);
 
+static struct workqueue_struct *rbd_wq;
+
 /*
  * Default to false for now, as single-major requires >= 0.75 version of
  * userspace rbd utility.
@@ -785,6 +793,20 @@ static int parse_rbd_opts_token(char *c, void *private)
 	return 0;
 }
 
+static char* obj_op_name(enum obj_operation_type op_type)
+{
+	switch (op_type) {
+	case OBJ_OP_READ:
+		return "read";
+	case OBJ_OP_WRITE:
+		return "write";
+	case OBJ_OP_DISCARD:
+		return "discard";
+	default:
+		return "???";
+	}
+}
+
 /*
  * Get a ceph client with specific addr and configuration, if one does
  * not exist create it.  Either way, ceph_opts is consumed by this
@@ -1600,6 +1622,21 @@ static bool img_request_write_test(struct rbd_img_request *img_request)
 	return test_bit(IMG_REQ_WRITE, &img_request->flags) != 0;
 }
 
+/*
+ * Set the discard flag when the img_request is an discard request
+ */
+static void img_request_discard_set(struct rbd_img_request *img_request)
+{
+	set_bit(IMG_REQ_DISCARD, &img_request->flags);
+	smp_mb();
+}
+
+static bool img_request_discard_test(struct rbd_img_request *img_request)
+{
+	smp_mb();
+	return test_bit(IMG_REQ_DISCARD, &img_request->flags) != 0;
+}
+
 static void img_request_child_set(struct rbd_img_request *img_request)
 {
 	set_bit(IMG_REQ_CHILD, &img_request->flags);
@@ -1636,6 +1673,17 @@ static bool img_request_layered_test(struct rbd_img_request *img_request)
 	return test_bit(IMG_REQ_LAYERED, &img_request->flags) != 0;
 }
 
+static enum obj_operation_type
+rbd_img_request_op_type(struct rbd_img_request *img_request)
+{
+	if (img_request_write_test(img_request))
+		return OBJ_OP_WRITE;
+	else if (img_request_discard_test(img_request))
+		return OBJ_OP_DISCARD;
+	else
+		return OBJ_OP_READ;
+}
+
 static void
 rbd_img_obj_request_read_callback(struct rbd_obj_request *obj_request)
 {
@@ -1722,6 +1770,21 @@ static void rbd_osd_write_callback(struct rbd_obj_request *obj_request)
 	obj_request_done_set(obj_request);
 }
 
+static void rbd_osd_discard_callback(struct rbd_obj_request *obj_request)
+{
+	dout("%s: obj %p result %d %llu\n", __func__, obj_request,
+		obj_request->result, obj_request->length);
+	/*
+	 * There is no such thing as a successful short discard.  Set
+	 * it to our originally-requested length.
+	 */
+	obj_request->xferred = obj_request->length;
+	/* discarding a non-existent object is not a problem */
+	if (obj_request->result == -ENOENT)
+		obj_request->result = 0;
+	obj_request_done_set(obj_request);
+}
+
 /*
  * For a simple stat call there's nothing to do.  We'll do more if
  * this is part of a write sequence for a layered image.
@@ -1773,6 +1836,11 @@ static void rbd_osd_req_callback(struct ceph_osd_request *osd_req,
 	case CEPH_OSD_OP_STAT:
 		rbd_osd_stat_callback(obj_request);
 		break;
+	case CEPH_OSD_OP_DELETE:
+	case CEPH_OSD_OP_TRUNCATE:
+	case CEPH_OSD_OP_ZERO:
+		rbd_osd_discard_callback(obj_request);
+		break;
 	case CEPH_OSD_OP_CALL:
 	case CEPH_OSD_OP_NOTIFY_ACK:
 	case CEPH_OSD_OP_WATCH:
@@ -1823,7 +1891,7 @@ static void rbd_osd_req_format_write(struct rbd_obj_request *obj_request)
  */
 static struct ceph_osd_request *rbd_osd_req_create(
 					struct rbd_device *rbd_dev,
-					bool write_request,
+					enum obj_operation_type op_type,
 					unsigned int num_ops,
 					struct rbd_obj_request *obj_request)
 {
@@ -1831,16 +1899,18 @@ static struct ceph_osd_request *rbd_osd_req_create(
 	struct ceph_osd_client *osdc;
 	struct ceph_osd_request *osd_req;
 
-	if (obj_request_img_data_test(obj_request)) {
+	if (obj_request_img_data_test(obj_request) &&
+		(op_type == OBJ_OP_DISCARD || op_type == OBJ_OP_WRITE)) {
 		struct rbd_img_request *img_request = obj_request->img_request;
-
-		rbd_assert(write_request ==
-				img_request_write_test(img_request));
-		if (write_request)
-			snapc = img_request->snapc;
+		if (op_type == OBJ_OP_WRITE) {
+			rbd_assert(img_request_write_test(img_request));
+		} else {
+			rbd_assert(img_request_discard_test(img_request));
+		}
+		snapc = img_request->snapc;
 	}
 
-	rbd_assert(num_ops == 1 || (write_request && num_ops == 2));
+	rbd_assert(num_ops == 1 || ((op_type == OBJ_OP_WRITE) && num_ops == 2));
 
 	/* Allocate and initialize the request, for the num_ops ops */
 
@@ -1850,7 +1920,7 @@ static struct ceph_osd_request *rbd_osd_req_create(
 	if (!osd_req)
 		return NULL;	/* ENOMEM */
 
-	if (write_request)
+	if (op_type == OBJ_OP_WRITE || op_type == OBJ_OP_DISCARD)
 		osd_req->r_flags = CEPH_OSD_FLAG_WRITE | CEPH_OSD_FLAG_ONDISK;
 	else
 		osd_req->r_flags = CEPH_OSD_FLAG_READ;
@@ -1865,9 +1935,10 @@ static struct ceph_osd_request *rbd_osd_req_create(
 }
 
 /*
- * Create a copyup osd request based on the information in the
- * object request supplied.  A copyup request has three osd ops,
- * a copyup method call, a hint op, and a write op.
+ * Create a copyup osd request based on the information in the object
+ * request supplied.  A copyup request has two or three osd ops, a
+ * copyup method call, potentially a hint op, and a write or truncate
+ * or zero op.
  */
 static struct ceph_osd_request *
 rbd_osd_req_create_copyup(struct rbd_obj_request *obj_request)
@@ -1877,18 +1948,24 @@ rbd_osd_req_create_copyup(struct rbd_obj_request *obj_request)
 	struct rbd_device *rbd_dev;
 	struct ceph_osd_client *osdc;
 	struct ceph_osd_request *osd_req;
+	int num_osd_ops = 3;
 
 	rbd_assert(obj_request_img_data_test(obj_request));
 	img_request = obj_request->img_request;
 	rbd_assert(img_request);
-	rbd_assert(img_request_write_test(img_request));
+	rbd_assert(img_request_write_test(img_request) ||
+			img_request_discard_test(img_request));
 
-	/* Allocate and initialize the request, for the three ops */
+	if (img_request_discard_test(img_request))
+		num_osd_ops = 2;
+
+	/* Allocate and initialize the request, for all the ops */
 
 	snapc = img_request->snapc;
 	rbd_dev = img_request->rbd_dev;
 	osdc = &rbd_dev->rbd_client->client->osdc;
-	osd_req = ceph_osdc_alloc_request(osdc, snapc, 3, false, GFP_ATOMIC);
+	osd_req = ceph_osdc_alloc_request(osdc, snapc, num_osd_ops,
+						false, GFP_ATOMIC);
 	if (!osd_req)
 		return NULL;	/* ENOMEM */
 
@@ -2021,32 +2098,26 @@ static void rbd_dev_parent_put(struct rbd_device *rbd_dev)
  * If an image has a non-zero parent overlap, get a reference to its
  * parent.
  *
- * We must get the reference before checking for the overlap to
- * coordinate properly with zeroing the parent overlap in
- * rbd_dev_v2_parent_info() when an image gets flattened.  We
- * drop it again if there is no overlap.
- *
  * Returns true if the rbd device has a parent with a non-zero
  * overlap and a reference for it was successfully taken, or
  * false otherwise.
  */
 static bool rbd_dev_parent_get(struct rbd_device *rbd_dev)
 {
-	int counter;
+	int counter = 0;
 
 	if (!rbd_dev->parent_spec)
 		return false;
 
-	counter = atomic_inc_return_safe(&rbd_dev->parent_ref);
-	if (counter > 0 && rbd_dev->parent_overlap)
-		return true;
-
-	/* Image was flattened, but parent is not yet torn down */
+	down_read(&rbd_dev->header_rwsem);
+	if (rbd_dev->parent_overlap)
+		counter = atomic_inc_return_safe(&rbd_dev->parent_ref);
+	up_read(&rbd_dev->header_rwsem);
 
 	if (counter < 0)
 		rbd_warn(rbd_dev, "parent reference overflow");
 
-	return false;
+	return counter > 0;
 }
 
 /*
@@ -2057,7 +2128,8 @@ static bool rbd_dev_parent_get(struct rbd_device *rbd_dev)
 static struct rbd_img_request *rbd_img_request_create(
 					struct rbd_device *rbd_dev,
 					u64 offset, u64 length,
-					bool write_request)
+					enum obj_operation_type op_type,
+					struct ceph_snap_context *snapc)
 {
 	struct rbd_img_request *img_request;
 
@@ -2065,20 +2137,17 @@ static struct rbd_img_request *rbd_img_request_create(
 	if (!img_request)
 		return NULL;
 
-	if (write_request) {
-		down_read(&rbd_dev->header_rwsem);
-		ceph_get_snap_context(rbd_dev->header.snapc);
-		up_read(&rbd_dev->header_rwsem);
-	}
-
 	img_request->rq = NULL;
 	img_request->rbd_dev = rbd_dev;
 	img_request->offset = offset;
 	img_request->length = length;
 	img_request->flags = 0;
-	if (write_request) {
+	if (op_type == OBJ_OP_DISCARD) {
+		img_request_discard_set(img_request);
+		img_request->snapc = snapc;
+	} else if (op_type == OBJ_OP_WRITE) {
 		img_request_write_set(img_request);
-		img_request->snapc = rbd_dev->header.snapc;
+		img_request->snapc = snapc;
 	} else {
 		img_request->snap_id = rbd_dev->spec->snap_id;
 	}
@@ -2093,8 +2162,7 @@ static struct rbd_img_request *rbd_img_request_create(
 	kref_init(&img_request->kref);
 
 	dout("%s: rbd_dev %p %s %llu/%llu -> img %p\n", __func__, rbd_dev,
-		write_request ? "write" : "read", offset, length,
-		img_request);
+		obj_op_name(op_type), offset, length, img_request);
 
 	return img_request;
 }
@@ -2118,7 +2186,8 @@ static void rbd_img_request_destroy(struct kref *kref)
 		rbd_dev_parent_put(img_request->rbd_dev);
 	}
 
-	if (img_request_write_test(img_request))
+	if (img_request_write_test(img_request) ||
+		img_request_discard_test(img_request))
 		ceph_put_snap_context(img_request->snapc);
 
 	kmem_cache_free(rbd_img_request_cache, img_request);
@@ -2134,8 +2203,8 @@ static struct rbd_img_request *rbd_parent_request_create(
 	rbd_assert(obj_request->img_request);
 	rbd_dev = obj_request->img_request->rbd_dev;
 
-	parent_request = rbd_img_request_create(rbd_dev->parent,
-						img_offset, length, false);
+	parent_request = rbd_img_request_create(rbd_dev->parent, img_offset,
+						length, OBJ_OP_READ, NULL);
 	if (!parent_request)
 		return NULL;
 
@@ -2176,11 +2245,18 @@ static bool rbd_img_obj_end_request(struct rbd_obj_request *obj_request)
 	result = obj_request->result;
 	if (result) {
 		struct rbd_device *rbd_dev = img_request->rbd_dev;
+		enum obj_operation_type op_type;
+
+		if (img_request_discard_test(img_request))
+			op_type = OBJ_OP_DISCARD;
+		else if (img_request_write_test(img_request))
+			op_type = OBJ_OP_WRITE;
+		else
+			op_type = OBJ_OP_READ;
 
 		rbd_warn(rbd_dev, "%s %llx at %llx (%llx)",
-			img_request_write_test(img_request) ? "write" : "read",
-			obj_request->length, obj_request->img_offset,
-			obj_request->offset);
+			obj_op_name(op_type), obj_request->length,
+			obj_request->img_offset, obj_request->offset);
 		rbd_warn(rbd_dev, "  result %d xferred %x",
 			result, xferred);
 		if (!img_request->result)
@@ -2245,6 +2321,71 @@ out:
 }
 
 /*
+ * Add individual osd ops to the given ceph_osd_request and prepare
+ * them for submission. num_ops is the current number of
+ * osd operations already to the object request.
+ */
+static void rbd_img_obj_request_fill(struct rbd_obj_request *obj_request,
+				struct ceph_osd_request *osd_request,
+				enum obj_operation_type op_type,
+				unsigned int num_ops)
+{
+	struct rbd_img_request *img_request = obj_request->img_request;
+	struct rbd_device *rbd_dev = img_request->rbd_dev;
+	u64 object_size = rbd_obj_bytes(&rbd_dev->header);
+	u64 offset = obj_request->offset;
+	u64 length = obj_request->length;
+	u64 img_end;
+	u16 opcode;
+
+	if (op_type == OBJ_OP_DISCARD) {
+		if (!offset && length == object_size &&
+		    (!img_request_layered_test(img_request) ||
+		     !obj_request_overlaps_parent(obj_request))) {
+			opcode = CEPH_OSD_OP_DELETE;
+		} else if ((offset + length == object_size)) {
+			opcode = CEPH_OSD_OP_TRUNCATE;
+		} else {
+			down_read(&rbd_dev->header_rwsem);
+			img_end = rbd_dev->header.image_size;
+			up_read(&rbd_dev->header_rwsem);
+
+			if (obj_request->img_offset + length == img_end)
+				opcode = CEPH_OSD_OP_TRUNCATE;
+			else
+				opcode = CEPH_OSD_OP_ZERO;
+		}
+	} else if (op_type == OBJ_OP_WRITE) {
+		opcode = CEPH_OSD_OP_WRITE;
+		osd_req_op_alloc_hint_init(osd_request, num_ops,
+					object_size, object_size);
+		num_ops++;
+	} else {
+		opcode = CEPH_OSD_OP_READ;
+	}
+
+	if (opcode == CEPH_OSD_OP_DELETE)
+		osd_req_op_init(osd_request, num_ops, opcode);
+	else
+		osd_req_op_extent_init(osd_request, num_ops, opcode,
+				       offset, length, 0, 0);
+
+	if (obj_request->type == OBJ_REQUEST_BIO)
+		osd_req_op_extent_osd_data_bio(osd_request, num_ops,
+					obj_request->bio_list, length);
+	else if (obj_request->type == OBJ_REQUEST_PAGES)
+		osd_req_op_extent_osd_data_pages(osd_request, num_ops,
+					obj_request->pages, length,
+					offset & ~PAGE_MASK, false, false);
+
+	/* Discards are also writes */
+	if (op_type == OBJ_OP_WRITE || op_type == OBJ_OP_DISCARD)
+		rbd_osd_req_format_write(obj_request);
+	else
+		rbd_osd_req_format_read(obj_request);
+}
+
+/*
  * Split up an image request into one or more object requests, each
  * to a different object.  The "type" parameter indicates whether
  * "data_desc" is the pointer to the head of a list of bio
@@ -2259,28 +2400,26 @@ static int rbd_img_request_fill(struct rbd_img_request *img_request,
 	struct rbd_device *rbd_dev = img_request->rbd_dev;
 	struct rbd_obj_request *obj_request = NULL;
 	struct rbd_obj_request *next_obj_request;
-	bool write_request = img_request_write_test(img_request);
 	struct bio *bio_list = NULL;
 	unsigned int bio_offset = 0;
 	struct page **pages = NULL;
+	enum obj_operation_type op_type;
 	u64 img_offset;
 	u64 resid;
-	u16 opcode;
 
 	dout("%s: img %p type %d data_desc %p\n", __func__, img_request,
 		(int)type, data_desc);
 
-	opcode = write_request ? CEPH_OSD_OP_WRITE : CEPH_OSD_OP_READ;
 	img_offset = img_request->offset;
 	resid = img_request->length;
 	rbd_assert(resid > 0);
+	op_type = rbd_img_request_op_type(img_request);
 
 	if (type == OBJ_REQUEST_BIO) {
 		bio_list = data_desc;
 		rbd_assert(img_offset ==
 			   bio_list->bi_iter.bi_sector << SECTOR_SHIFT);
-	} else {
-		rbd_assert(type == OBJ_REQUEST_PAGES);
+	} else if (type == OBJ_REQUEST_PAGES) {
 		pages = data_desc;
 	}
 
@@ -2289,7 +2428,6 @@ static int rbd_img_request_fill(struct rbd_img_request *img_request,
 		const char *object_name;
 		u64 offset;
 		u64 length;
-		unsigned int which = 0;
 
 		object_name = rbd_segment_name(rbd_dev, img_offset);
 		if (!object_name)
@@ -2321,7 +2459,7 @@ static int rbd_img_request_fill(struct rbd_img_request *img_request,
 								GFP_ATOMIC);
 			if (!obj_request->bio_list)
 				goto out_unwind;
-		} else {
+		} else if (type == OBJ_REQUEST_PAGES) {
 			unsigned int page_count;
 
 			obj_request->pages = pages;
@@ -2332,38 +2470,19 @@ static int rbd_img_request_fill(struct rbd_img_request *img_request,
 			pages += page_count;
 		}
 
-		osd_req = rbd_osd_req_create(rbd_dev, write_request,
-					     (write_request ? 2 : 1),
-					     obj_request);
+		osd_req = rbd_osd_req_create(rbd_dev, op_type,
+					(op_type == OBJ_OP_WRITE) ? 2 : 1,
+					obj_request);
 		if (!osd_req)
 			goto out_unwind;
+
 		obj_request->osd_req = osd_req;
 		obj_request->callback = rbd_img_obj_callback;
-		rbd_img_request_get(img_request);
-
-		if (write_request) {
-			osd_req_op_alloc_hint_init(osd_req, which,
-					     rbd_obj_bytes(&rbd_dev->header),
-					     rbd_obj_bytes(&rbd_dev->header));
-			which++;
-		}
-
-		osd_req_op_extent_init(osd_req, which, opcode, offset, length,
-				       0, 0);
-		if (type == OBJ_REQUEST_BIO)
-			osd_req_op_extent_osd_data_bio(osd_req, which,
-					obj_request->bio_list, length);
-		else
-			osd_req_op_extent_osd_data_pages(osd_req, which,
-					obj_request->pages, length,
-					offset & ~PAGE_MASK, false, false);
+		obj_request->img_offset = img_offset;
 
-		if (write_request)
-			rbd_osd_req_format_write(obj_request);
-		else
-			rbd_osd_req_format_read(obj_request);
+		rbd_img_obj_request_fill(obj_request, osd_req, op_type, 0);
 
-		obj_request->img_offset = img_offset;
+		rbd_img_request_get(img_request);
 
 		img_offset += length;
 		resid -= length;
@@ -2386,7 +2505,8 @@ rbd_img_obj_copyup_callback(struct rbd_obj_request *obj_request)
 	struct page **pages;
 	u32 page_count;
 
-	rbd_assert(obj_request->type == OBJ_REQUEST_BIO);
+	rbd_assert(obj_request->type == OBJ_REQUEST_BIO ||
+		obj_request->type == OBJ_REQUEST_NODATA);
 	rbd_assert(obj_request_img_data_test(obj_request));
 	img_request = obj_request->img_request;
 	rbd_assert(img_request);
@@ -2424,11 +2544,10 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request)
 	struct ceph_osd_client *osdc;
 	struct rbd_device *rbd_dev;
 	struct page **pages;
+	enum obj_operation_type op_type;
 	u32 page_count;
 	int img_result;
 	u64 parent_length;
-	u64 offset;
-	u64 length;
 
 	rbd_assert(img_request_child_test(img_request));
 
@@ -2492,26 +2611,10 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request)
 	osd_req_op_cls_request_data_pages(osd_req, 0, pages, parent_length, 0,
 						false, false);
 
-	/* Then the hint op */
-
-	osd_req_op_alloc_hint_init(osd_req, 1, rbd_obj_bytes(&rbd_dev->header),
-				   rbd_obj_bytes(&rbd_dev->header));
-
-	/* And the original write request op */
-
-	offset = orig_request->offset;
-	length = orig_request->length;
-	osd_req_op_extent_init(osd_req, 2, CEPH_OSD_OP_WRITE,
-					offset, length, 0, 0);
-	if (orig_request->type == OBJ_REQUEST_BIO)
-		osd_req_op_extent_osd_data_bio(osd_req, 2,
-					orig_request->bio_list, length);
-	else
-		osd_req_op_extent_osd_data_pages(osd_req, 2,
-					orig_request->pages, length,
-					offset & ~PAGE_MASK, false, false);
+	/* Add the other op(s) */
 
-	rbd_osd_req_format_write(orig_request);
+	op_type = rbd_img_request_op_type(orig_request->img_request);
+	rbd_img_obj_request_fill(orig_request, osd_req, op_type, 1);
 
 	/* All set, send it off. */
 
@@ -2728,7 +2831,7 @@ static int rbd_img_obj_exists_submit(struct rbd_obj_request *obj_request)
 
 	rbd_assert(obj_request->img_request);
 	rbd_dev = obj_request->img_request->rbd_dev;
-	stat_request->osd_req = rbd_osd_req_create(rbd_dev, false, 1,
+	stat_request->osd_req = rbd_osd_req_create(rbd_dev, OBJ_OP_READ, 1,
 						   stat_request);
 	if (!stat_request->osd_req)
 		goto out;
@@ -2748,11 +2851,10 @@ out:
 	return ret;
 }
 
-static int rbd_img_obj_request_submit(struct rbd_obj_request *obj_request)
+static bool img_obj_request_simple(struct rbd_obj_request *obj_request)
 {
 	struct rbd_img_request *img_request;
 	struct rbd_device *rbd_dev;
-	bool known;
 
 	rbd_assert(obj_request_img_data_test(obj_request));
 
@@ -2760,22 +2862,44 @@ static int rbd_img_obj_request_submit(struct rbd_obj_request *obj_request)
 	rbd_assert(img_request);
 	rbd_dev = img_request->rbd_dev;
 
+	/* Reads */
+	if (!img_request_write_test(img_request) &&
+	    !img_request_discard_test(img_request))
+		return true;
+
+	/* Non-layered writes */
+	if (!img_request_layered_test(img_request))
+		return true;
+
+	/*
+	 * Layered writes outside of the parent overlap range don't
+	 * share any data with the parent.
+	 */
+	if (!obj_request_overlaps_parent(obj_request))
+		return true;
+
 	/*
-	 * Only writes to layered images need special handling.
-	 * Reads and non-layered writes are simple object requests.
-	 * Layered writes that start beyond the end of the overlap
-	 * with the parent have no parent data, so they too are
-	 * simple object requests.  Finally, if the target object is
-	 * known to already exist, its parent data has already been
-	 * copied, so a write to the object can also be handled as a
-	 * simple object request.
+	 * Entire-object layered writes - we will overwrite whatever
+	 * parent data there is anyway.
 	 */
-	if (!img_request_write_test(img_request) ||
-		!img_request_layered_test(img_request) ||
-		!obj_request_overlaps_parent(obj_request) ||
-		((known = obj_request_known_test(obj_request)) &&
-			obj_request_exists_test(obj_request))) {
+	if (!obj_request->offset &&
+	    obj_request->length == rbd_obj_bytes(&rbd_dev->header))
+		return true;
+
+	/*
+	 * If the object is known to already exist, its parent data has
+	 * already been copied.
+	 */
+	if (obj_request_known_test(obj_request) &&
+	    obj_request_exists_test(obj_request))
+		return true;
 
+	return false;
+}
+
+static int rbd_img_obj_request_submit(struct rbd_obj_request *obj_request)
+{
+	if (img_obj_request_simple(obj_request)) {
 		struct rbd_device *rbd_dev;
 		struct ceph_osd_client *osdc;
 
@@ -2791,7 +2915,7 @@ static int rbd_img_obj_request_submit(struct rbd_obj_request *obj_request)
 	 * start by reading the data for the full target object from
 	 * the parent so we can use it for a copyup to the target.
 	 */
-	if (known)
+	if (obj_request_known_test(obj_request))
 		return rbd_img_obj_parent_read_full(obj_request);
 
 	/* We don't know whether the target exists.  Go find out. */
@@ -2932,7 +3056,7 @@ static int rbd_obj_notify_ack_sync(struct rbd_device *rbd_dev, u64 notify_id)
 		return -ENOMEM;
 
 	ret = -ENOMEM;
-	obj_request->osd_req = rbd_osd_req_create(rbd_dev, false, 1,
+	obj_request->osd_req = rbd_osd_req_create(rbd_dev, OBJ_OP_READ, 1,
 						  obj_request);
 	if (!obj_request->osd_req)
 		goto out;
@@ -2995,7 +3119,7 @@ static struct rbd_obj_request *rbd_obj_watch_request_helper(
 	if (!obj_request)
 		return ERR_PTR(-ENOMEM);
 
-	obj_request->osd_req = rbd_osd_req_create(rbd_dev, true, 1,
+	obj_request->osd_req = rbd_osd_req_create(rbd_dev, OBJ_OP_WRITE, 1,
 						  obj_request);
 	if (!obj_request->osd_req) {
 		ret = -ENOMEM;
@@ -3133,7 +3257,7 @@ static int rbd_obj_method_sync(struct rbd_device *rbd_dev,
 	obj_request->pages = pages;
 	obj_request->page_count = page_count;
 
-	obj_request->osd_req = rbd_osd_req_create(rbd_dev, false, 1,
+	obj_request->osd_req = rbd_osd_req_create(rbd_dev, OBJ_OP_READ, 1,
 						  obj_request);
 	if (!obj_request->osd_req)
 		goto out;
@@ -3183,11 +3307,20 @@ out:
 static void rbd_handle_request(struct rbd_device *rbd_dev, struct request *rq)
 {
 	struct rbd_img_request *img_request;
+	struct ceph_snap_context *snapc = NULL;
 	u64 offset = (u64)blk_rq_pos(rq) << SECTOR_SHIFT;
 	u64 length = blk_rq_bytes(rq);
-	bool wr = rq_data_dir(rq) == WRITE;
+	enum obj_operation_type op_type;
+	u64 mapping_size;
 	int result;
 
+	if (rq->cmd_flags & REQ_DISCARD)
+		op_type = OBJ_OP_DISCARD;
+	else if (rq->cmd_flags & REQ_WRITE)
+		op_type = OBJ_OP_WRITE;
+	else
+		op_type = OBJ_OP_READ;
+
 	/* Ignore/skip any zero-length requests */
 
 	if (!length) {
@@ -3196,9 +3329,9 @@ static void rbd_handle_request(struct rbd_device *rbd_dev, struct request *rq)
 		goto err_rq;
 	}
 
-	/* Disallow writes to a read-only device */
+	/* Only reads are allowed to a read-only device */
 
-	if (wr) {
+	if (op_type != OBJ_OP_READ) {
 		if (rbd_dev->mapping.read_only) {
 			result = -EROFS;
 			goto err_rq;
@@ -3226,21 +3359,35 @@ static void rbd_handle_request(struct rbd_device *rbd_dev, struct request *rq)
 		goto err_rq;	/* Shouldn't happen */
 	}
 
-	if (offset + length > rbd_dev->mapping.size) {
+	down_read(&rbd_dev->header_rwsem);
+	mapping_size = rbd_dev->mapping.size;
+	if (op_type != OBJ_OP_READ) {
+		snapc = rbd_dev->header.snapc;
+		ceph_get_snap_context(snapc);
+	}
+	up_read(&rbd_dev->header_rwsem);
+
+	if (offset + length > mapping_size) {
 		rbd_warn(rbd_dev, "beyond EOD (%llu~%llu > %llu)", offset,
-			 length, rbd_dev->mapping.size);
+			 length, mapping_size);
 		result = -EIO;
 		goto err_rq;
 	}
 
-	img_request = rbd_img_request_create(rbd_dev, offset, length, wr);
+	img_request = rbd_img_request_create(rbd_dev, offset, length, op_type,
+					     snapc);
 	if (!img_request) {
 		result = -ENOMEM;
 		goto err_rq;
 	}
 	img_request->rq = rq;
 
-	result = rbd_img_request_fill(img_request, OBJ_REQUEST_BIO, rq->bio);
+	if (op_type == OBJ_OP_DISCARD)
+		result = rbd_img_request_fill(img_request, OBJ_REQUEST_NODATA,
+					      NULL);
+	else
+		result = rbd_img_request_fill(img_request, OBJ_REQUEST_BIO,
+					      rq->bio);
 	if (result)
 		goto err_img_request;
 
@@ -3255,7 +3402,8 @@ err_img_request:
 err_rq:
 	if (result)
 		rbd_warn(rbd_dev, "%s %llx at %llx result %d",
-			 wr ? "write" : "read", length, offset, result);
+			 obj_op_name(op_type), length, offset, result);
+	ceph_put_snap_context(snapc);
 	blk_end_request_all(rq, result);
 }
 
@@ -3302,7 +3450,7 @@ static void rbd_request_fn(struct request_queue *q)
 	}
 
 	if (queued)
-		queue_work(rbd_dev->rq_wq, &rbd_dev->rq_work);
+		queue_work(rbd_wq, &rbd_dev->rq_work);
 }
 
 /*
@@ -3382,7 +3530,7 @@ static int rbd_obj_read_sync(struct rbd_device *rbd_dev,
 	page_count = (u32) calc_pages_for(offset, length);
 	pages = ceph_alloc_page_vector(page_count, GFP_KERNEL);
 	if (IS_ERR(pages))
-		ret = PTR_ERR(pages);
+		return PTR_ERR(pages);
 
 	ret = -ENOMEM;
 	obj_request = rbd_obj_request_create(object_name, offset, length,
@@ -3393,7 +3541,7 @@ static int rbd_obj_read_sync(struct rbd_device *rbd_dev,
 	obj_request->pages = pages;
 	obj_request->page_count = page_count;
 
-	obj_request->osd_req = rbd_osd_req_create(rbd_dev, false, 1,
+	obj_request->osd_req = rbd_osd_req_create(rbd_dev, OBJ_OP_READ, 1,
 						  obj_request);
 	if (!obj_request->osd_req)
 		goto out;
@@ -3610,6 +3758,13 @@ static int rbd_init_disk(struct rbd_device *rbd_dev)
 	blk_queue_io_min(q, segment_size);
 	blk_queue_io_opt(q, segment_size);
 
+	/* enable the discard support */
+	queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, q);
+	q->limits.discard_granularity = segment_size;
+	q->limits.discard_alignment = segment_size;
+	q->limits.max_discard_sectors = segment_size / SECTOR_SIZE;
+	q->limits.discard_zeroes_data = 1;
+
 	blk_queue_merge_bvec(q, rbd_merge_bvec);
 	disk->queue = q;
 
@@ -4078,7 +4233,6 @@ static int rbd_dev_v2_parent_info(struct rbd_device *rbd_dev)
 		 */
 		if (rbd_dev->parent_overlap) {
 			rbd_dev->parent_overlap = 0;
-			smp_mb();
 			rbd_dev_parent_put(rbd_dev);
 			pr_info("%s: clone image has been flattened\n",
 				rbd_dev->disk->disk_name);
@@ -4124,7 +4278,6 @@ static int rbd_dev_v2_parent_info(struct rbd_device *rbd_dev)
 	 * treat it specially.
 	 */
 	rbd_dev->parent_overlap = overlap;
-	smp_mb();
 	if (!overlap) {
 
 		/* A null parent_spec indicates it's the initial probe */
@@ -4924,7 +5077,7 @@ static int rbd_dev_image_id(struct rbd_device *rbd_dev)
 		ret = image_id ? 0 : -ENOMEM;
 		if (!ret)
 			rbd_dev->image_format = 1;
-	} else if (ret > sizeof (__le32)) {
+	} else if (ret >= 0) {
 		void *p = response;
 
 		image_id = ceph_extract_encoded_string(&p, p + ret,
@@ -4932,8 +5085,6 @@ static int rbd_dev_image_id(struct rbd_device *rbd_dev)
 		ret = PTR_ERR_OR_ZERO(image_id);
 		if (!ret)
 			rbd_dev->image_format = 2;
-	} else {
-		ret = -EINVAL;
 	}
 
 	if (!ret) {
@@ -4955,10 +5106,7 @@ static void rbd_dev_unprobe(struct rbd_device *rbd_dev)
 {
 	struct rbd_image_header	*header;
 
-	/* Drop parent reference unless it's already been done (or none) */
-
-	if (rbd_dev->parent_overlap)
-		rbd_dev_parent_put(rbd_dev);
+	rbd_dev_parent_put(rbd_dev);
 
 	/* Free dynamic fields from the header, then zero it out */
 
@@ -5087,15 +5235,9 @@ static int rbd_dev_device_setup(struct rbd_device *rbd_dev)
 	set_capacity(rbd_dev->disk, rbd_dev->mapping.size / SECTOR_SIZE);
 	set_disk_ro(rbd_dev->disk, rbd_dev->mapping.read_only);
 
-	rbd_dev->rq_wq = alloc_workqueue("%s", 0, 0, rbd_dev->disk->disk_name);
-	if (!rbd_dev->rq_wq) {
-		ret = -ENOMEM;
-		goto err_out_mapping;
-	}
-
 	ret = rbd_bus_add_dev(rbd_dev);
 	if (ret)
-		goto err_out_workqueue;
+		goto err_out_mapping;
 
 	/* Everything's ready.  Announce the disk to the world. */
 
@@ -5107,9 +5249,6 @@ static int rbd_dev_device_setup(struct rbd_device *rbd_dev)
 
 	return ret;
 
-err_out_workqueue:
-	destroy_workqueue(rbd_dev->rq_wq);
-	rbd_dev->rq_wq = NULL;
 err_out_mapping:
 	rbd_dev_mapping_clear(rbd_dev);
 err_out_disk:
@@ -5356,7 +5495,6 @@ static void rbd_dev_device_release(struct device *dev)
 {
 	struct rbd_device *rbd_dev = dev_to_rbd_dev(dev);
 
-	destroy_workqueue(rbd_dev->rq_wq);
 	rbd_free_disk(rbd_dev);
 	clear_bit(RBD_DEV_FLAG_EXISTS, &rbd_dev->flags);
 	rbd_dev_mapping_clear(rbd_dev);
@@ -5560,11 +5698,21 @@ static int __init rbd_init(void)
 	if (rc)
 		return rc;
 
+	/*
+	 * The number of active work items is limited by the number of
+	 * rbd devices, so leave @max_active at default.
+	 */
+	rbd_wq = alloc_workqueue(RBD_DRV_NAME, WQ_MEM_RECLAIM, 0);
+	if (!rbd_wq) {
+		rc = -ENOMEM;
+		goto err_out_slab;
+	}
+
 	if (single_major) {
 		rbd_major = register_blkdev(0, RBD_DRV_NAME);
 		if (rbd_major < 0) {
 			rc = rbd_major;
-			goto err_out_slab;
+			goto err_out_wq;
 		}
 	}
 
@@ -5582,6 +5730,8 @@ static int __init rbd_init(void)
 err_out_blkdev:
 	if (single_major)
 		unregister_blkdev(rbd_major, RBD_DRV_NAME);
+err_out_wq:
+	destroy_workqueue(rbd_wq);
 err_out_slab:
 	rbd_slab_exit();
 	return rc;
@@ -5593,6 +5743,7 @@ static void __exit rbd_exit(void)
 	rbd_sysfs_cleanup();
 	if (single_major)
 		unregister_blkdev(rbd_major, RBD_DRV_NAME);
+	destroy_workqueue(rbd_wq);
 	rbd_slab_exit();
 }
 
diff --git a/drivers/block/rsxx/core.c b/drivers/block/rsxx/core.c
index 820b4009d5f7..d8b2488aaade 100644
--- a/drivers/block/rsxx/core.c
+++ b/drivers/block/rsxx/core.c
@@ -62,12 +62,6 @@ static DEFINE_SPINLOCK(rsxx_ida_lock);
 
 /* --------------------Debugfs Setup ------------------- */
 
-struct rsxx_cram {
-	u32 f_pos;
-	u32 offset;
-	void *i_private;
-};
-
 static int rsxx_attr_pci_regs_show(struct seq_file *m, void *p)
 {
 	struct rsxx_cardinfo *card = m->private;
@@ -184,93 +178,50 @@ static int rsxx_attr_pci_regs_open(struct inode *inode, struct file *file)
 static ssize_t rsxx_cram_read(struct file *fp, char __user *ubuf,
 			      size_t cnt, loff_t *ppos)
 {
-	struct rsxx_cram *info = fp->private_data;
-	struct rsxx_cardinfo *card = info->i_private;
+	struct rsxx_cardinfo *card = file_inode(fp)->i_private;
 	char *buf;
-	int st;
+	ssize_t st;
 
-	buf = kzalloc(sizeof(*buf) * cnt, GFP_KERNEL);
+	buf = kzalloc(cnt, GFP_KERNEL);
 	if (!buf)
 		return -ENOMEM;
 
-	info->f_pos = (u32)*ppos + info->offset;
-
-	st = rsxx_creg_read(card, CREG_ADD_CRAM + info->f_pos, cnt, buf, 1);
-	if (st)
-		return st;
-
-	st = copy_to_user(ubuf, buf, cnt);
+	st = rsxx_creg_read(card, CREG_ADD_CRAM + (u32)*ppos, cnt, buf, 1);
+	if (!st)
+		st = copy_to_user(ubuf, buf, cnt);
+	kfree(buf);
 	if (st)
 		return st;
-
-	info->offset += cnt;
-
-	kfree(buf);
-
+	*ppos += cnt;
 	return cnt;
 }
 
 static ssize_t rsxx_cram_write(struct file *fp, const char __user *ubuf,
 			       size_t cnt, loff_t *ppos)
 {
-	struct rsxx_cram *info = fp->private_data;
-	struct rsxx_cardinfo *card = info->i_private;
+	struct rsxx_cardinfo *card = file_inode(fp)->i_private;
 	char *buf;
-	int st;
+	ssize_t st;
 
-	buf = kzalloc(sizeof(*buf) * cnt, GFP_KERNEL);
+	buf = kzalloc(cnt, GFP_KERNEL);
 	if (!buf)
 		return -ENOMEM;
 
 	st = copy_from_user(buf, ubuf, cnt);
+	if (!st)
+		st = rsxx_creg_write(card, CREG_ADD_CRAM + (u32)*ppos, cnt,
+				     buf, 1);
+	kfree(buf);
 	if (st)
 		return st;
-
-	info->f_pos = (u32)*ppos + info->offset;
-
-	st = rsxx_creg_write(card, CREG_ADD_CRAM + info->f_pos, cnt, buf, 1);
-	if (st)
-		return st;
-
-	info->offset += cnt;
-
-	kfree(buf);
-
+	*ppos += cnt;
 	return cnt;
 }
 
-static int rsxx_cram_open(struct inode *inode, struct file *file)
-{
-	struct rsxx_cram *info = kzalloc(sizeof(*info), GFP_KERNEL);
-	if (!info)
-		return -ENOMEM;
-
-	info->i_private = inode->i_private;
-	info->f_pos = file->f_pos;
-	file->private_data = info;
-
-	return 0;
-}
-
-static int rsxx_cram_release(struct inode *inode, struct file *file)
-{
-	struct rsxx_cram *info = file->private_data;
-
-	if (!info)
-		return 0;
-
-	kfree(info);
-	file->private_data = NULL;
-
-	return 0;
-}
-
 static const struct file_operations debugfs_cram_fops = {
 	.owner		= THIS_MODULE,
-	.open		= rsxx_cram_open,
 	.read		= rsxx_cram_read,
 	.write		= rsxx_cram_write,
-	.release	= rsxx_cram_release,
 };
 
 static const struct file_operations debugfs_stats_fops = {
@@ -886,7 +837,7 @@ static int rsxx_pci_probe(struct pci_dev *dev,
 				"Failed to enable MSI\n");
 	}
 
-	st = request_irq(dev->irq, rsxx_isr, IRQF_DISABLED | IRQF_SHARED,
+	st = request_irq(dev->irq, rsxx_isr, IRQF_SHARED,
 			 DRIVER_NAME, card);
 	if (st) {
 		dev_err(CARD_TO_DEV(card),
diff --git a/drivers/block/rsxx/dev.c b/drivers/block/rsxx/dev.c
index 2839d37e5af7..ac8c62cb4875 100644
--- a/drivers/block/rsxx/dev.c
+++ b/drivers/block/rsxx/dev.c
@@ -112,37 +112,16 @@ static const struct block_device_operations rsxx_fops = {
 
 static void disk_stats_start(struct rsxx_cardinfo *card, struct bio *bio)
 {
-	struct hd_struct *part0 = &card->gendisk->part0;
-	int rw = bio_data_dir(bio);
-	int cpu;
-
-	cpu = part_stat_lock();
-
-	part_round_stats(cpu, part0);
-	part_inc_in_flight(part0, rw);
-
-	part_stat_unlock();
+	generic_start_io_acct(bio_data_dir(bio), bio_sectors(bio),
+			     &card->gendisk->part0);
 }
 
 static void disk_stats_complete(struct rsxx_cardinfo *card,
 				struct bio *bio,
 				unsigned long start_time)
 {
-	struct hd_struct *part0 = &card->gendisk->part0;
-	unsigned long duration = jiffies - start_time;
-	int rw = bio_data_dir(bio);
-	int cpu;
-
-	cpu = part_stat_lock();
-
-	part_stat_add(cpu, part0, sectors[rw], bio_sectors(bio));
-	part_stat_inc(cpu, part0, ios[rw]);
-	part_stat_add(cpu, part0, ticks[rw], duration);
-
-	part_round_stats(cpu, part0);
-	part_dec_in_flight(part0, rw);
-
-	part_stat_unlock();
+	generic_end_io_acct(bio_data_dir(bio), &card->gendisk->part0,
+			   start_time);
 }
 
 static void bio_dma_done_cb(struct rsxx_cardinfo *card,
@@ -307,6 +286,7 @@ int rsxx_setup_dev(struct rsxx_cardinfo *card)
 	blk_queue_physical_block_size(card->queue, RSXX_HW_BLK_SIZE);
 
 	queue_flag_set_unlocked(QUEUE_FLAG_NONROT, card->queue);
+	queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, card->queue);
 	if (rsxx_discard_supported(card)) {
 		queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, card->queue);
 		blk_queue_max_discard_sectors(card->queue,
diff --git a/drivers/block/skd_main.c b/drivers/block/skd_main.c
index 8fcdcfb4b472..1e46eb2305c0 100644
--- a/drivers/block/skd_main.c
+++ b/drivers/block/skd_main.c
@@ -4426,6 +4426,7 @@ static int skd_cons_disk(struct skd_device *skdev)
 	q->limits.discard_zeroes_data = 1;
 	queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, q);
 	queue_flag_set_unlocked(QUEUE_FLAG_NONROT, q);
+	queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, q);
 
 	spin_lock_irqsave(&skdev->lock, flags);
 	pr_debug("%s:%s:%d stopping %s queue\n",
diff --git a/drivers/block/sunvdc.c b/drivers/block/sunvdc.c
index 5814deb6963d..4b911ed96ea3 100644
--- a/drivers/block/sunvdc.c
+++ b/drivers/block/sunvdc.c
@@ -9,6 +9,7 @@
 #include <linux/blkdev.h>
 #include <linux/hdreg.h>
 #include <linux/genhd.h>
+#include <linux/cdrom.h>
 #include <linux/slab.h>
 #include <linux/spinlock.h>
 #include <linux/completion.h>
@@ -22,8 +23,8 @@
 
 #define DRV_MODULE_NAME		"sunvdc"
 #define PFX DRV_MODULE_NAME	": "
-#define DRV_MODULE_VERSION	"1.0"
-#define DRV_MODULE_RELDATE	"June 25, 2007"
+#define DRV_MODULE_VERSION	"1.2"
+#define DRV_MODULE_RELDATE	"November 24, 2014"
 
 static char version[] =
 	DRV_MODULE_NAME ".c:v" DRV_MODULE_VERSION " (" DRV_MODULE_RELDATE ")\n";
@@ -32,13 +33,15 @@ MODULE_DESCRIPTION("Sun LDOM virtual disk client driver");
 MODULE_LICENSE("GPL");
 MODULE_VERSION(DRV_MODULE_VERSION);
 
-#define VDC_TX_RING_SIZE	256
+#define VDC_TX_RING_SIZE	512
 
 #define WAITING_FOR_LINK_UP	0x01
 #define WAITING_FOR_TX_SPACE	0x02
 #define WAITING_FOR_GEN_CMD	0x04
 #define WAITING_FOR_ANY		-1
 
+static struct workqueue_struct *sunvdc_wq;
+
 struct vdc_req_entry {
 	struct request		*req;
 };
@@ -59,19 +62,25 @@ struct vdc_port {
 	u64			max_xfer_size;
 	u32			vdisk_block_size;
 
+	u64			ldc_timeout;
+	struct timer_list	ldc_reset_timer;
+	struct work_struct	ldc_reset_work;
+
 	/* The server fills these in for us in the disk attribute
 	 * ACK packet.
 	 */
 	u64			operations;
 	u32			vdisk_size;
 	u8			vdisk_type;
+	u8			vdisk_mtype;
 
 	char			disk_name[32];
-
-	struct vio_disk_geom	geom;
-	struct vio_disk_vtoc	label;
 };
 
+static void vdc_ldc_reset(struct vdc_port *port);
+static void vdc_ldc_reset_work(struct work_struct *work);
+static void vdc_ldc_reset_timer(unsigned long _arg);
+
 static inline struct vdc_port *to_vdc_port(struct vio_driver_state *vio)
 {
 	return container_of(vio, struct vdc_port, vio);
@@ -79,9 +88,16 @@ static inline struct vdc_port *to_vdc_port(struct vio_driver_state *vio)
 
 /* Ordered from largest major to lowest */
 static struct vio_version vdc_versions[] = {
+	{ .major = 1, .minor = 1 },
 	{ .major = 1, .minor = 0 },
 };
 
+static inline int vdc_version_supported(struct vdc_port *port,
+					u16 major, u16 minor)
+{
+	return port->vio.ver.major == major && port->vio.ver.minor >= minor;
+}
+
 #define VDCBLK_NAME	"vdisk"
 static int vdc_major;
 #define PARTITION_SHIFT	3
@@ -94,20 +110,71 @@ static inline u32 vdc_tx_dring_avail(struct vio_dring_state *dr)
 static int vdc_getgeo(struct block_device *bdev, struct hd_geometry *geo)
 {
 	struct gendisk *disk = bdev->bd_disk;
-	struct vdc_port *port = disk->private_data;
+	sector_t nsect = get_capacity(disk);
+	sector_t cylinders = nsect;
 
-	geo->heads = (u8) port->geom.num_hd;
-	geo->sectors = (u8) port->geom.num_sec;
-	geo->cylinders = port->geom.num_cyl;
+	geo->heads = 0xff;
+	geo->sectors = 0x3f;
+	sector_div(cylinders, geo->heads * geo->sectors);
+	geo->cylinders = cylinders;
+	if ((sector_t)(geo->cylinders + 1) * geo->heads * geo->sectors < nsect)
+		geo->cylinders = 0xffff;
 
 	return 0;
 }
 
+/* Add ioctl/CDROM_GET_CAPABILITY to support cdrom_id in udev
+ * when vdisk_mtype is VD_MEDIA_TYPE_CD or VD_MEDIA_TYPE_DVD.
+ * Needed to be able to install inside an ldom from an iso image.
+ */
+static int vdc_ioctl(struct block_device *bdev, fmode_t mode,
+		     unsigned command, unsigned long argument)
+{
+	int i;
+	struct gendisk *disk;
+
+	switch (command) {
+	case CDROMMULTISESSION:
+		pr_debug(PFX "Multisession CDs not supported\n");
+		for (i = 0; i < sizeof(struct cdrom_multisession); i++)
+			if (put_user(0, (char __user *)(argument + i)))
+				return -EFAULT;
+		return 0;
+
+	case CDROM_GET_CAPABILITY:
+		disk = bdev->bd_disk;
+
+		if (bdev->bd_disk && (disk->flags & GENHD_FL_CD))
+			return 0;
+		return -EINVAL;
+
+	default:
+		pr_debug(PFX "ioctl %08x not supported\n", command);
+		return -EINVAL;
+	}
+}
+
 static const struct block_device_operations vdc_fops = {
 	.owner		= THIS_MODULE,
 	.getgeo		= vdc_getgeo,
+	.ioctl		= vdc_ioctl,
 };
 
+static void vdc_blk_queue_start(struct vdc_port *port)
+{
+	struct vio_dring_state *dr = &port->vio.drings[VIO_DRIVER_TX_RING];
+
+	/* restart blk queue when ring is half emptied. also called after
+	 * handshake completes, so check for initial handshake before we've
+	 * allocated a disk.
+	 */
+	if (port->disk && blk_queue_stopped(port->disk->queue) &&
+	    vdc_tx_dring_avail(dr) * 100 / VDC_TX_RING_SIZE >= 50) {
+		blk_start_queue(port->disk->queue);
+	}
+
+}
+
 static void vdc_finish(struct vio_driver_state *vio, int err, int waiting_for)
 {
 	if (vio->cmp &&
@@ -121,7 +188,11 @@ static void vdc_finish(struct vio_driver_state *vio, int err, int waiting_for)
 
 static void vdc_handshake_complete(struct vio_driver_state *vio)
 {
+	struct vdc_port *port = to_vdc_port(vio);
+
+	del_timer(&port->ldc_reset_timer);
 	vdc_finish(vio, 0, WAITING_FOR_LINK_UP);
+	vdc_blk_queue_start(port);
 }
 
 static int vdc_handle_unknown(struct vdc_port *port, void *arg)
@@ -165,9 +236,9 @@ static int vdc_handle_attr(struct vio_driver_state *vio, void *arg)
 	struct vio_disk_attr_info *pkt = arg;
 
 	viodbg(HS, "GOT ATTR stype[0x%x] ops[%llx] disk_size[%llu] disk_type[%x] "
-	       "xfer_mode[0x%x] blksz[%u] max_xfer[%llu]\n",
+	       "mtype[0x%x] xfer_mode[0x%x] blksz[%u] max_xfer[%llu]\n",
 	       pkt->tag.stype, pkt->operations,
-	       pkt->vdisk_size, pkt->vdisk_type,
+	       pkt->vdisk_size, pkt->vdisk_type, pkt->vdisk_mtype,
 	       pkt->xfer_mode, pkt->vdisk_block_size,
 	       pkt->max_xfer_size);
 
@@ -192,8 +263,11 @@ static int vdc_handle_attr(struct vio_driver_state *vio, void *arg)
 		}
 
 		port->operations = pkt->operations;
-		port->vdisk_size = pkt->vdisk_size;
 		port->vdisk_type = pkt->vdisk_type;
+		if (vdc_version_supported(port, 1, 1)) {
+			port->vdisk_size = pkt->vdisk_size;
+			port->vdisk_mtype = pkt->vdisk_mtype;
+		}
 		if (pkt->max_xfer_size < port->max_xfer_size)
 			port->max_xfer_size = pkt->max_xfer_size;
 		port->vdisk_block_size = pkt->vdisk_block_size;
@@ -224,7 +298,7 @@ static void vdc_end_one(struct vdc_port *port, struct vio_dring_state *dr,
 
 	ldc_unmap(port->vio.lp, desc->cookies, desc->ncookies);
 	desc->hdr.state = VIO_DESC_FREE;
-	dr->cons = (index + 1) & (VDC_TX_RING_SIZE - 1);
+	dr->cons = vio_dring_next(dr, index);
 
 	req = rqe->req;
 	if (req == NULL) {
@@ -236,8 +310,7 @@ static void vdc_end_one(struct vdc_port *port, struct vio_dring_state *dr,
 
 	__blk_end_request(req, (desc->status ? -EIO : 0), desc->size);
 
-	if (blk_queue_stopped(port->disk->queue))
-		blk_start_queue(port->disk->queue);
+	vdc_blk_queue_start(port);
 }
 
 static int vdc_ack(struct vdc_port *port, void *msgbuf)
@@ -270,17 +343,20 @@ static void vdc_event(void *arg, int event)
 
 	spin_lock_irqsave(&vio->lock, flags);
 
-	if (unlikely(event == LDC_EVENT_RESET ||
-		     event == LDC_EVENT_UP)) {
+	if (unlikely(event == LDC_EVENT_RESET)) {
 		vio_link_state_change(vio, event);
-		spin_unlock_irqrestore(&vio->lock, flags);
-		return;
+		queue_work(sunvdc_wq, &port->ldc_reset_work);
+		goto out;
+	}
+
+	if (unlikely(event == LDC_EVENT_UP)) {
+		vio_link_state_change(vio, event);
+		goto out;
 	}
 
 	if (unlikely(event != LDC_EVENT_DATA_READY)) {
-		printk(KERN_WARNING PFX "Unexpected LDC event %d\n", event);
-		spin_unlock_irqrestore(&vio->lock, flags);
-		return;
+		pr_warn(PFX "Unexpected LDC event %d\n", event);
+		goto out;
 	}
 
 	err = 0;
@@ -324,6 +400,7 @@ static void vdc_event(void *arg, int event)
 	}
 	if (err < 0)
 		vdc_finish(&port->vio, err, WAITING_FOR_ANY);
+out:
 	spin_unlock_irqrestore(&vio->lock, flags);
 }
 
@@ -356,6 +433,8 @@ static int __vdc_tx_trigger(struct vdc_port *port)
 			delay = 128;
 	} while (err == -EAGAIN);
 
+	if (err == -ENOTCONN)
+		vdc_ldc_reset(port);
 	return err;
 }
 
@@ -388,12 +467,6 @@ static int __send_request(struct request *req)
 	for (i = 0; i < nsg; i++)
 		len += sg[i].length;
 
-	if (unlikely(vdc_tx_dring_avail(dr) < 1)) {
-		blk_stop_queue(port->disk->queue);
-		err = -ENOMEM;
-		goto out;
-	}
-
 	desc = vio_dring_cur(dr);
 
 	err = ldc_map_sg(port->vio.lp, sg, nsg,
@@ -431,23 +504,34 @@ static int __send_request(struct request *req)
 		printk(KERN_ERR PFX "vdc_tx_trigger() failure, err=%d\n", err);
 	} else {
 		port->req_id++;
-		dr->prod = (dr->prod + 1) & (VDC_TX_RING_SIZE - 1);
+		dr->prod = vio_dring_next(dr, dr->prod);
 	}
-out:
 
 	return err;
 }
 
-static void do_vdc_request(struct request_queue *q)
+static void do_vdc_request(struct request_queue *rq)
 {
-	while (1) {
-		struct request *req = blk_fetch_request(q);
+	struct request *req;
 
-		if (!req)
-			break;
+	while ((req = blk_peek_request(rq)) != NULL) {
+		struct vdc_port *port;
+		struct vio_dring_state *dr;
+
+		port = req->rq_disk->private_data;
+		dr = &port->vio.drings[VIO_DRIVER_TX_RING];
+		if (unlikely(vdc_tx_dring_avail(dr) < 1))
+			goto wait;
+
+		blk_start_request(req);
 
-		if (__send_request(req) < 0)
-			__blk_end_request_all(req, -EIO);
+		if (__send_request(req) < 0) {
+			blk_requeue_request(rq, req);
+wait:
+			/* Avoid pointless unplugs. */
+			blk_stop_queue(rq);
+			break;
+		}
 	}
 }
 
@@ -574,7 +658,7 @@ static int generic_request(struct vdc_port *port, u8 op, void *buf, int len)
 	err = __vdc_tx_trigger(port);
 	if (err >= 0) {
 		port->req_id++;
-		dr->prod = (dr->prod + 1) & (VDC_TX_RING_SIZE - 1);
+		dr->prod = vio_dring_next(dr, dr->prod);
 		spin_unlock_irqrestore(&port->vio.lock, flags);
 
 		wait_for_completion(&comp.com);
@@ -638,12 +722,9 @@ static void vdc_free_tx_ring(struct vdc_port *port)
 	}
 }
 
-static int probe_disk(struct vdc_port *port)
+static int vdc_port_up(struct vdc_port *port)
 {
 	struct vio_completion comp;
-	struct request_queue *q;
-	struct gendisk *g;
-	int err;
 
 	init_completion(&comp.com);
 	comp.err = 0;
@@ -651,29 +732,48 @@ static int probe_disk(struct vdc_port *port)
 	port->vio.cmp = &comp;
 
 	vio_port_up(&port->vio);
-
 	wait_for_completion(&comp.com);
-	if (comp.err)
-		return comp.err;
+	return comp.err;
+}
 
-	err = generic_request(port, VD_OP_GET_VTOC,
-			      &port->label, sizeof(port->label));
-	if (err < 0) {
-		printk(KERN_ERR PFX "VD_OP_GET_VTOC returns error %d\n", err);
-		return err;
-	}
+static void vdc_port_down(struct vdc_port *port)
+{
+	ldc_disconnect(port->vio.lp);
+	ldc_unbind(port->vio.lp);
+	vdc_free_tx_ring(port);
+	vio_ldc_free(&port->vio);
+}
 
-	err = generic_request(port, VD_OP_GET_DISKGEOM,
-			      &port->geom, sizeof(port->geom));
-	if (err < 0) {
-		printk(KERN_ERR PFX "VD_OP_GET_DISKGEOM returns "
-		       "error %d\n", err);
+static int probe_disk(struct vdc_port *port)
+{
+	struct request_queue *q;
+	struct gendisk *g;
+	int err;
+
+	err = vdc_port_up(port);
+	if (err)
 		return err;
-	}
 
-	port->vdisk_size = ((u64)port->geom.num_cyl *
-			    (u64)port->geom.num_hd *
-			    (u64)port->geom.num_sec);
+	if (vdc_version_supported(port, 1, 1)) {
+		/* vdisk_size should be set during the handshake, if it wasn't
+		 * then the underlying disk is reserved by another system
+		 */
+		if (port->vdisk_size == -1)
+			return -ENODEV;
+	} else {
+		struct vio_disk_geom geom;
+
+		err = generic_request(port, VD_OP_GET_DISKGEOM,
+				      &geom, sizeof(geom));
+		if (err < 0) {
+			printk(KERN_ERR PFX "VD_OP_GET_DISKGEOM returns "
+			       "error %d\n", err);
+			return err;
+		}
+		port->vdisk_size = ((u64)geom.num_cyl *
+				    (u64)geom.num_hd *
+				    (u64)geom.num_sec);
+	}
 
 	q = blk_init_queue(do_vdc_request, &port->vio.lock);
 	if (!q) {
@@ -691,6 +791,10 @@ static int probe_disk(struct vdc_port *port)
 
 	port->disk = g;
 
+	/* Each segment in a request is up to an aligned page in size. */
+	blk_queue_segment_boundary(q, PAGE_SIZE - 1);
+	blk_queue_max_segment_size(q, PAGE_SIZE);
+
 	blk_queue_max_segments(q, port->ring_cookies);
 	blk_queue_max_hw_sectors(q, port->max_xfer_size);
 	g->major = vdc_major;
@@ -704,9 +808,32 @@ static int probe_disk(struct vdc_port *port)
 
 	set_capacity(g, port->vdisk_size);
 
-	printk(KERN_INFO PFX "%s: %u sectors (%u MB)\n",
+	if (vdc_version_supported(port, 1, 1)) {
+		switch (port->vdisk_mtype) {
+		case VD_MEDIA_TYPE_CD:
+			pr_info(PFX "Virtual CDROM %s\n", port->disk_name);
+			g->flags |= GENHD_FL_CD;
+			g->flags |= GENHD_FL_REMOVABLE;
+			set_disk_ro(g, 1);
+			break;
+
+		case VD_MEDIA_TYPE_DVD:
+			pr_info(PFX "Virtual DVD %s\n", port->disk_name);
+			g->flags |= GENHD_FL_CD;
+			g->flags |= GENHD_FL_REMOVABLE;
+			set_disk_ro(g, 1);
+			break;
+
+		case VD_MEDIA_TYPE_FIXED:
+			pr_info(PFX "Virtual Hard disk %s\n", port->disk_name);
+			break;
+		}
+	}
+
+	pr_info(PFX "%s: %u sectors (%u MB) protocol %d.%d\n",
 	       g->disk_name,
-	       port->vdisk_size, (port->vdisk_size >> (20 - 9)));
+	       port->vdisk_size, (port->vdisk_size >> (20 - 9)),
+	       port->vio.ver.major, port->vio.ver.minor);
 
 	add_disk(g);
 
@@ -738,6 +865,7 @@ static int vdc_port_probe(struct vio_dev *vdev, const struct vio_device_id *id)
 	struct mdesc_handle *hp;
 	struct vdc_port *port;
 	int err;
+	const u64 *ldc_timeout;
 
 	print_version();
 
@@ -765,6 +893,17 @@ static int vdc_port_probe(struct vio_dev *vdev, const struct vio_device_id *id)
 	else
 		snprintf(port->disk_name, sizeof(port->disk_name),
 			 VDCBLK_NAME "%c", 'a' + ((int)vdev->dev_no % 26));
+	port->vdisk_size = -1;
+
+	/* Actual wall time may be double due to do_generic_file_read() doing
+	 * a readahead I/O first, and once that fails it will try to read a
+	 * single page.
+	 */
+	ldc_timeout = mdesc_get_property(hp, vdev->mp, "vdc-timeout", NULL);
+	port->ldc_timeout = ldc_timeout ? *ldc_timeout : 0;
+	setup_timer(&port->ldc_reset_timer, vdc_ldc_reset_timer,
+		    (unsigned long)port);
+	INIT_WORK(&port->ldc_reset_work, vdc_ldc_reset_work);
 
 	err = vio_driver_init(&port->vio, vdev, VDEV_DISK,
 			      vdc_versions, ARRAY_SIZE(vdc_versions),
@@ -814,8 +953,21 @@ static int vdc_port_remove(struct vio_dev *vdev)
 	struct vdc_port *port = dev_get_drvdata(&vdev->dev);
 
 	if (port) {
+		unsigned long flags;
+
+		spin_lock_irqsave(&port->vio.lock, flags);
+		blk_stop_queue(port->disk->queue);
+		spin_unlock_irqrestore(&port->vio.lock, flags);
+
+		flush_work(&port->ldc_reset_work);
+		del_timer_sync(&port->ldc_reset_timer);
 		del_timer_sync(&port->vio.timer);
 
+		del_gendisk(port->disk);
+		blk_cleanup_queue(port->disk->queue);
+		put_disk(port->disk);
+		port->disk = NULL;
+
 		vdc_free_tx_ring(port);
 		vio_ldc_free(&port->vio);
 
@@ -826,6 +978,102 @@ static int vdc_port_remove(struct vio_dev *vdev)
 	return 0;
 }
 
+static void vdc_requeue_inflight(struct vdc_port *port)
+{
+	struct vio_dring_state *dr = &port->vio.drings[VIO_DRIVER_TX_RING];
+	u32 idx;
+
+	for (idx = dr->cons; idx != dr->prod; idx = vio_dring_next(dr, idx)) {
+		struct vio_disk_desc *desc = vio_dring_entry(dr, idx);
+		struct vdc_req_entry *rqe = &port->rq_arr[idx];
+		struct request *req;
+
+		ldc_unmap(port->vio.lp, desc->cookies, desc->ncookies);
+		desc->hdr.state = VIO_DESC_FREE;
+		dr->cons = vio_dring_next(dr, idx);
+
+		req = rqe->req;
+		if (req == NULL) {
+			vdc_end_special(port, desc);
+			continue;
+		}
+
+		rqe->req = NULL;
+		blk_requeue_request(port->disk->queue, req);
+	}
+}
+
+static void vdc_queue_drain(struct vdc_port *port)
+{
+	struct request *req;
+
+	while ((req = blk_fetch_request(port->disk->queue)) != NULL)
+		__blk_end_request_all(req, -EIO);
+}
+
+static void vdc_ldc_reset_timer(unsigned long _arg)
+{
+	struct vdc_port *port = (struct vdc_port *) _arg;
+	struct vio_driver_state *vio = &port->vio;
+	unsigned long flags;
+
+	spin_lock_irqsave(&vio->lock, flags);
+	if (!(port->vio.hs_state & VIO_HS_COMPLETE)) {
+		pr_warn(PFX "%s ldc down %llu seconds, draining queue\n",
+			port->disk_name, port->ldc_timeout);
+		vdc_queue_drain(port);
+		vdc_blk_queue_start(port);
+	}
+	spin_unlock_irqrestore(&vio->lock, flags);
+}
+
+static void vdc_ldc_reset_work(struct work_struct *work)
+{
+	struct vdc_port *port;
+	struct vio_driver_state *vio;
+	unsigned long flags;
+
+	port = container_of(work, struct vdc_port, ldc_reset_work);
+	vio = &port->vio;
+
+	spin_lock_irqsave(&vio->lock, flags);
+	vdc_ldc_reset(port);
+	spin_unlock_irqrestore(&vio->lock, flags);
+}
+
+static void vdc_ldc_reset(struct vdc_port *port)
+{
+	int err;
+
+	assert_spin_locked(&port->vio.lock);
+
+	pr_warn(PFX "%s ldc link reset\n", port->disk_name);
+	blk_stop_queue(port->disk->queue);
+	vdc_requeue_inflight(port);
+	vdc_port_down(port);
+
+	err = vio_ldc_alloc(&port->vio, &vdc_ldc_cfg, port);
+	if (err) {
+		pr_err(PFX "%s vio_ldc_alloc:%d\n", port->disk_name, err);
+		return;
+	}
+
+	err = vdc_alloc_tx_ring(port);
+	if (err) {
+		pr_err(PFX "%s vio_alloc_tx_ring:%d\n", port->disk_name, err);
+		goto err_free_ldc;
+	}
+
+	if (port->ldc_timeout)
+		mod_timer(&port->ldc_reset_timer,
+			  round_jiffies(jiffies + HZ * port->ldc_timeout));
+	mod_timer(&port->vio.timer, round_jiffies(jiffies + HZ));
+	return;
+
+err_free_ldc:
+	vio_ldc_free(&port->vio);
+}
+
 static const struct vio_device_id vdc_port_match[] = {
 	{
 		.type = "vdc-port",
@@ -845,9 +1093,13 @@ static int __init vdc_init(void)
 {
 	int err;
 
+	sunvdc_wq = alloc_workqueue("sunvdc", 0, 0);
+	if (!sunvdc_wq)
+		return -ENOMEM;
+
 	err = register_blkdev(0, VDCBLK_NAME);
 	if (err < 0)
-		goto out_err;
+		goto out_free_wq;
 
 	vdc_major = err;
 
@@ -861,7 +1113,8 @@ out_unregister_blkdev:
 	unregister_blkdev(vdc_major, VDCBLK_NAME);
 	vdc_major = 0;
 
-out_err:
+out_free_wq:
+	destroy_workqueue(sunvdc_wq);
 	return err;
 }
 
@@ -869,6 +1122,7 @@ static void __exit vdc_exit(void)
 {
 	vio_unregister_driver(&vdc_port_driver);
 	unregister_blkdev(vdc_major, VDCBLK_NAME);
+	destroy_workqueue(sunvdc_wq);
 }
 
 module_init(vdc_init);
diff --git a/drivers/block/swim.c b/drivers/block/swim.c
index 6b44bbe528b7..b5afd495d482 100644
--- a/drivers/block/swim.c
+++ b/drivers/block/swim.c
@@ -971,7 +971,6 @@ static struct platform_driver swim_driver = {
 	.remove = swim_remove,
 	.driver   = {
 		.name	= CARDNAME,
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/block/sx8.c b/drivers/block/sx8.c
index d5e2d12b9d9e..5d552857de41 100644
--- a/drivers/block/sx8.c
+++ b/drivers/block/sx8.c
@@ -568,7 +568,7 @@ static struct carm_request *carm_get_special(struct carm_host *host)
 		return NULL;
 
 	rq = blk_get_request(host->oob_q, WRITE /* bogus */, GFP_KERNEL);
-	if (!rq) {
+	if (IS_ERR(rq)) {
 		spin_lock_irqsave(&host->lock, flags);
 		carm_put_request(host, crq);
 		spin_unlock_irqrestore(&host->lock, flags);
diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c
index 0a581400de0f..cdfbd21e3597 100644
--- a/drivers/block/virtio_blk.c
+++ b/drivers/block/virtio_blk.c
@@ -41,12 +41,6 @@ struct virtio_blk
 	/* Process context for config space updates */
 	struct work_struct config_work;
 
-	/* Lock for config space updates */
-	struct mutex config_lock;
-
-	/* enable config space updates */
-	bool config_enable;
-
 	/* What host tells us, plus 2 for header & tailer. */
 	unsigned int sg_elems;
 
@@ -86,7 +80,7 @@ static int __virtblk_add_req(struct virtqueue *vq,
 {
 	struct scatterlist hdr, status, cmd, sense, inhdr, *sgs[6];
 	unsigned int num_out = 0, num_in = 0;
-	int type = vbr->out_hdr.type & ~VIRTIO_BLK_T_OUT;
+	__virtio32 type = vbr->out_hdr.type & ~cpu_to_virtio32(vq->vdev, VIRTIO_BLK_T_OUT);
 
 	sg_init_one(&hdr, &vbr->out_hdr, sizeof(vbr->out_hdr));
 	sgs[num_out++] = &hdr;
@@ -97,19 +91,19 @@ static int __virtblk_add_req(struct virtqueue *vq,
 	 * block, and before the normal inhdr we put the sense data and the
 	 * inhdr with additional status information.
 	 */
-	if (type == VIRTIO_BLK_T_SCSI_CMD) {
+	if (type == cpu_to_virtio32(vq->vdev, VIRTIO_BLK_T_SCSI_CMD)) {
 		sg_init_one(&cmd, vbr->req->cmd, vbr->req->cmd_len);
 		sgs[num_out++] = &cmd;
 	}
 
 	if (have_data) {
-		if (vbr->out_hdr.type & VIRTIO_BLK_T_OUT)
+		if (vbr->out_hdr.type & cpu_to_virtio32(vq->vdev, VIRTIO_BLK_T_OUT))
 			sgs[num_out++] = data_sg;
 		else
 			sgs[num_out + num_in++] = data_sg;
 	}
 
-	if (type == VIRTIO_BLK_T_SCSI_CMD) {
+	if (type == cpu_to_virtio32(vq->vdev, VIRTIO_BLK_T_SCSI_CMD)) {
 		sg_init_one(&sense, vbr->req->sense, SCSI_SENSE_BUFFERSIZE);
 		sgs[num_out + num_in++] = &sense;
 		sg_init_one(&inhdr, &vbr->in_hdr, sizeof(vbr->in_hdr));
@@ -125,17 +119,18 @@ static int __virtblk_add_req(struct virtqueue *vq,
 static inline void virtblk_request_done(struct request *req)
 {
 	struct virtblk_req *vbr = blk_mq_rq_to_pdu(req);
+	struct virtio_blk *vblk = req->q->queuedata;
 	int error = virtblk_result(vbr);
 
 	if (req->cmd_type == REQ_TYPE_BLOCK_PC) {
-		req->resid_len = vbr->in_hdr.residual;
-		req->sense_len = vbr->in_hdr.sense_len;
-		req->errors = vbr->in_hdr.errors;
+		req->resid_len = virtio32_to_cpu(vblk->vdev, vbr->in_hdr.residual);
+		req->sense_len = virtio32_to_cpu(vblk->vdev, vbr->in_hdr.sense_len);
+		req->errors = virtio32_to_cpu(vblk->vdev, vbr->in_hdr.errors);
 	} else if (req->cmd_type == REQ_TYPE_SPECIAL) {
 		req->errors = (error != 0);
 	}
 
-	blk_mq_end_io(req, error);
+	blk_mq_end_request(req, error);
 }
 
 static void virtblk_done(struct virtqueue *vq)
@@ -164,14 +159,15 @@ static void virtblk_done(struct virtqueue *vq)
 	spin_unlock_irqrestore(&vblk->vqs[qid].lock, flags);
 }
 
-static int virtio_queue_rq(struct blk_mq_hw_ctx *hctx, struct request *req)
+static int virtio_queue_rq(struct blk_mq_hw_ctx *hctx,
+			   const struct blk_mq_queue_data *bd)
 {
 	struct virtio_blk *vblk = hctx->queue->queuedata;
+	struct request *req = bd->rq;
 	struct virtblk_req *vbr = blk_mq_rq_to_pdu(req);
 	unsigned long flags;
 	unsigned int num;
 	int qid = hctx->queue_num;
-	const bool last = (req->cmd_flags & REQ_END) != 0;
 	int err;
 	bool notify = false;
 
@@ -179,25 +175,25 @@ static int virtio_queue_rq(struct blk_mq_hw_ctx *hctx, struct request *req)
 
 	vbr->req = req;
 	if (req->cmd_flags & REQ_FLUSH) {
-		vbr->out_hdr.type = VIRTIO_BLK_T_FLUSH;
+		vbr->out_hdr.type = cpu_to_virtio32(vblk->vdev, VIRTIO_BLK_T_FLUSH);
 		vbr->out_hdr.sector = 0;
-		vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
+		vbr->out_hdr.ioprio = cpu_to_virtio32(vblk->vdev, req_get_ioprio(vbr->req));
 	} else {
 		switch (req->cmd_type) {
 		case REQ_TYPE_FS:
 			vbr->out_hdr.type = 0;
-			vbr->out_hdr.sector = blk_rq_pos(vbr->req);
-			vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
+			vbr->out_hdr.sector = cpu_to_virtio64(vblk->vdev, blk_rq_pos(vbr->req));
+			vbr->out_hdr.ioprio = cpu_to_virtio32(vblk->vdev, req_get_ioprio(vbr->req));
 			break;
 		case REQ_TYPE_BLOCK_PC:
-			vbr->out_hdr.type = VIRTIO_BLK_T_SCSI_CMD;
+			vbr->out_hdr.type = cpu_to_virtio32(vblk->vdev, VIRTIO_BLK_T_SCSI_CMD);
 			vbr->out_hdr.sector = 0;
-			vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
+			vbr->out_hdr.ioprio = cpu_to_virtio32(vblk->vdev, req_get_ioprio(vbr->req));
 			break;
 		case REQ_TYPE_SPECIAL:
-			vbr->out_hdr.type = VIRTIO_BLK_T_GET_ID;
+			vbr->out_hdr.type = cpu_to_virtio32(vblk->vdev, VIRTIO_BLK_T_GET_ID);
 			vbr->out_hdr.sector = 0;
-			vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
+			vbr->out_hdr.ioprio = cpu_to_virtio32(vblk->vdev, req_get_ioprio(vbr->req));
 			break;
 		default:
 			/* We don't put anything else in the queue. */
@@ -205,12 +201,14 @@ static int virtio_queue_rq(struct blk_mq_hw_ctx *hctx, struct request *req)
 		}
 	}
 
+	blk_mq_start_request(req);
+
 	num = blk_rq_map_sg(hctx->queue, vbr->req, vbr->sg);
 	if (num) {
 		if (rq_data_dir(vbr->req) == WRITE)
-			vbr->out_hdr.type |= VIRTIO_BLK_T_OUT;
+			vbr->out_hdr.type |= cpu_to_virtio32(vblk->vdev, VIRTIO_BLK_T_OUT);
 		else
-			vbr->out_hdr.type |= VIRTIO_BLK_T_IN;
+			vbr->out_hdr.type |= cpu_to_virtio32(vblk->vdev, VIRTIO_BLK_T_IN);
 	}
 
 	spin_lock_irqsave(&vblk->vqs[qid].lock, flags);
@@ -226,7 +224,7 @@ static int virtio_queue_rq(struct blk_mq_hw_ctx *hctx, struct request *req)
 		return BLK_MQ_RQ_QUEUE_ERROR;
 	}
 
-	if (last && virtqueue_kick_prepare(vblk->vqs[qid].vq))
+	if (bd->last && virtqueue_kick_prepare(vblk->vqs[qid].vq))
 		notify = true;
 	spin_unlock_irqrestore(&vblk->vqs[qid].lock, flags);
 
@@ -335,7 +333,8 @@ static ssize_t virtblk_serial_show(struct device *dev,
 
 	return err;
 }
-DEVICE_ATTR(serial, S_IRUGO, virtblk_serial_show, NULL);
+
+static DEVICE_ATTR(serial, S_IRUGO, virtblk_serial_show, NULL);
 
 static void virtblk_config_changed_work(struct work_struct *work)
 {
@@ -347,10 +346,6 @@ static void virtblk_config_changed_work(struct work_struct *work)
 	char *envp[] = { "RESIZE=1", NULL };
 	u64 capacity, size;
 
-	mutex_lock(&vblk->config_lock);
-	if (!vblk->config_enable)
-		goto done;
-
 	/* Host must always specify the capacity. */
 	virtio_cread(vdev, struct virtio_blk_config, capacity, &capacity);
 
@@ -374,8 +369,6 @@ static void virtblk_config_changed_work(struct work_struct *work)
 	set_capacity(vblk->disk, capacity);
 	revalidate_disk(vblk->disk);
 	kobject_uevent_env(&disk_to_dev(vblk->disk)->kobj, KOBJ_CHANGE, envp);
-done:
-	mutex_unlock(&vblk->config_lock);
 }
 
 static void virtblk_config_changed(struct virtio_device *vdev)
@@ -486,7 +479,8 @@ static int virtblk_get_cache_mode(struct virtio_device *vdev)
 				   struct virtio_blk_config, wce,
 				   &writeback);
 	if (err)
-		writeback = virtio_has_feature(vdev, VIRTIO_BLK_F_WCE);
+		writeback = virtio_has_feature(vdev, VIRTIO_BLK_F_WCE) ||
+		            virtio_has_feature(vdev, VIRTIO_F_VERSION_1);
 
 	return writeback;
 }
@@ -606,10 +600,8 @@ static int virtblk_probe(struct virtio_device *vdev)
 
 	vblk->vdev = vdev;
 	vblk->sg_elems = sg_elems;
-	mutex_init(&vblk->config_lock);
 
 	INIT_WORK(&vblk->config_work, virtblk_config_changed_work);
-	vblk->config_enable = true;
 
 	err = init_vq(vblk);
 	if (err)
@@ -646,7 +638,7 @@ static int virtblk_probe(struct virtio_device *vdev)
 		goto out_put_disk;
 
 	q = vblk->disk->queue = blk_mq_init_queue(&vblk->tag_set);
-	if (!q) {
+	if (IS_ERR(q)) {
 		err = -ENOMEM;
 		goto out_free_tags;
 	}
@@ -733,6 +725,8 @@ static int virtblk_probe(struct virtio_device *vdev)
 	if (!err && opt_io_size)
 		blk_queue_io_opt(q, blk_size * opt_io_size);
 
+	virtio_device_ready(vdev);
+
 	add_disk(vblk->disk);
 	err = device_create_file(disk_to_dev(vblk->disk), &dev_attr_serial);
 	if (err)
@@ -771,10 +765,8 @@ static void virtblk_remove(struct virtio_device *vdev)
 	int index = vblk->index;
 	int refc;
 
-	/* Prevent config work handler from accessing the device. */
-	mutex_lock(&vblk->config_lock);
-	vblk->config_enable = false;
-	mutex_unlock(&vblk->config_lock);
+	/* Make sure no work handler is accessing the device. */
+	flush_work(&vblk->config_work);
 
 	del_gendisk(vblk->disk);
 	blk_cleanup_queue(vblk->disk->queue);
@@ -784,8 +776,6 @@ static void virtblk_remove(struct virtio_device *vdev)
 	/* Stop all the virtqueues. */
 	vdev->config->reset(vdev);
 
-	flush_work(&vblk->config_work);
-
 	refc = atomic_read(&disk_to_dev(vblk->disk)->kobj.kref.refcount);
 	put_disk(vblk->disk);
 	vdev->config->del_vqs(vdev);
@@ -805,11 +795,7 @@ static int virtblk_freeze(struct virtio_device *vdev)
 	/* Ensure we don't receive any more interrupts */
 	vdev->config->reset(vdev);
 
-	/* Prevent config work handler from accessing the device. */
-	mutex_lock(&vblk->config_lock);
-	vblk->config_enable = false;
-	mutex_unlock(&vblk->config_lock);
-
+	/* Make sure no work handler is accessing the device. */
 	flush_work(&vblk->config_work);
 
 	blk_mq_stop_hw_queues(vblk->disk->queue);
@@ -823,12 +809,14 @@ static int virtblk_restore(struct virtio_device *vdev)
 	struct virtio_blk *vblk = vdev->priv;
 	int ret;
 
-	vblk->config_enable = true;
 	ret = init_vq(vdev->priv);
-	if (!ret)
-		blk_mq_start_stopped_hw_queues(vblk->disk->queue, true);
+	if (ret)
+		return ret;
+
+	virtio_device_ready(vdev);
 
-	return ret;
+	blk_mq_start_stopped_hw_queues(vblk->disk->queue, true);
+	return 0;
 }
 #endif
 
@@ -837,25 +825,34 @@ static const struct virtio_device_id id_table[] = {
 	{ 0 },
 };
 
-static unsigned int features[] = {
+static unsigned int features_legacy[] = {
 	VIRTIO_BLK_F_SEG_MAX, VIRTIO_BLK_F_SIZE_MAX, VIRTIO_BLK_F_GEOMETRY,
 	VIRTIO_BLK_F_RO, VIRTIO_BLK_F_BLK_SIZE, VIRTIO_BLK_F_SCSI,
 	VIRTIO_BLK_F_WCE, VIRTIO_BLK_F_TOPOLOGY, VIRTIO_BLK_F_CONFIG_WCE,
 	VIRTIO_BLK_F_MQ,
+}
+;
+static unsigned int features[] = {
+	VIRTIO_BLK_F_SEG_MAX, VIRTIO_BLK_F_SIZE_MAX, VIRTIO_BLK_F_GEOMETRY,
+	VIRTIO_BLK_F_RO, VIRTIO_BLK_F_BLK_SIZE,
+	VIRTIO_BLK_F_TOPOLOGY,
+	VIRTIO_BLK_F_MQ,
 };
 
 static struct virtio_driver virtio_blk = {
-	.feature_table		= features,
-	.feature_table_size	= ARRAY_SIZE(features),
-	.driver.name		= KBUILD_MODNAME,
-	.driver.owner		= THIS_MODULE,
-	.id_table		= id_table,
-	.probe			= virtblk_probe,
-	.remove			= virtblk_remove,
-	.config_changed		= virtblk_config_changed,
+	.feature_table			= features,
+	.feature_table_size		= ARRAY_SIZE(features),
+	.feature_table_legacy		= features_legacy,
+	.feature_table_size_legacy	= ARRAY_SIZE(features_legacy),
+	.driver.name			= KBUILD_MODNAME,
+	.driver.owner			= THIS_MODULE,
+	.id_table			= id_table,
+	.probe				= virtblk_probe,
+	.remove				= virtblk_remove,
+	.config_changed			= virtblk_config_changed,
 #ifdef CONFIG_PM_SLEEP
-	.freeze			= virtblk_freeze,
-	.restore		= virtblk_restore,
+	.freeze				= virtblk_freeze,
+	.restore			= virtblk_restore,
 #endif
 };
 
@@ -887,8 +884,8 @@ out_destroy_workqueue:
 
 static void __exit fini(void)
 {
-	unregister_blkdev(major, "virtblk");
 	unregister_virtio_driver(&virtio_blk);
+	unregister_blkdev(major, "virtblk");
 	destroy_workqueue(virtblk_wq);
 }
 module_init(init);
diff --git a/drivers/block/xen-blkback/blkback.c b/drivers/block/xen-blkback/blkback.c
index 64c60edcdfbc..63fc7f06a014 100644
--- a/drivers/block/xen-blkback/blkback.c
+++ b/drivers/block/xen-blkback/blkback.c
@@ -763,6 +763,7 @@ again:
 			BUG_ON(new_map_idx >= segs_to_map);
 			if (unlikely(map[new_map_idx].status != 0)) {
 				pr_debug(DRV_PFX "invalid buffer -- could not remap it\n");
+				put_free_pages(blkif, &pages[seg_idx]->page, 1);
 				pages[seg_idx]->handle = BLKBACK_INVALID_HANDLE;
 				ret |= 1;
 				goto next;
diff --git a/drivers/block/xen-blkback/xenbus.c b/drivers/block/xen-blkback/xenbus.c
index 3a8b810b4980..630a489e757d 100644
--- a/drivers/block/xen-blkback/xenbus.c
+++ b/drivers/block/xen-blkback/xenbus.c
@@ -270,6 +270,9 @@ static int xen_blkif_disconnect(struct xen_blkif *blkif)
 		blkif->blk_rings.common.sring = NULL;
 	}
 
+	/* Remove all persistent grants and the cache of ballooned pages. */
+	xen_blkbk_free_caches(blkif);
+
 	return 0;
 }
 
@@ -281,9 +284,6 @@ static void xen_blkif_free(struct xen_blkif *blkif)
 	xen_blkif_disconnect(blkif);
 	xen_vbd_free(&blkif->vbd);
 
-	/* Remove all persistent grants and the cache of ballooned pages. */
-	xen_blkbk_free_caches(blkif);
-
 	/* Make sure everything is drained before shutting down */
 	BUG_ON(blkif->persistent_gnt_c != 0);
 	BUG_ON(atomic_read(&blkif->persistent_gnt_in_use) != 0);
@@ -907,22 +907,17 @@ static int connect_ring(struct backend_info *be)
 	return 0;
 }
 
-
-/* ** Driver Registration ** */
-
-
 static const struct xenbus_device_id xen_blkbk_ids[] = {
 	{ "vbd" },
 	{ "" }
 };
 
-
-static DEFINE_XENBUS_DRIVER(xen_blkbk, ,
+static struct xenbus_driver xen_blkbk_driver = {
+	.ids  = xen_blkbk_ids,
 	.probe = xen_blkbk_probe,
 	.remove = xen_blkbk_remove,
 	.otherend_changed = frontend_changed
-);
-
+};
 
 int xen_blkif_xenbus_init(void)
 {
diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c
index 5deb235bd18f..2236c6f31608 100644
--- a/drivers/block/xen-blkfront.c
+++ b/drivers/block/xen-blkfront.c
@@ -126,7 +126,6 @@ struct blkfront_info
 	unsigned int persistent_gnts_c;
 	unsigned long shadow_free;
 	unsigned int feature_flush;
-	unsigned int flush_op;
 	unsigned int feature_discard:1;
 	unsigned int feature_secdiscard:1;
 	unsigned int discard_granularity;
@@ -479,7 +478,19 @@ static int blkif_queue_request(struct request *req)
 				 * way.  (It's also a FLUSH+FUA, since it is
 				 * guaranteed ordered WRT previous writes.)
 				 */
-				ring_req->operation = info->flush_op;
+				switch (info->feature_flush &
+					((REQ_FLUSH|REQ_FUA))) {
+				case REQ_FLUSH|REQ_FUA:
+					ring_req->operation =
+						BLKIF_OP_WRITE_BARRIER;
+					break;
+				case REQ_FLUSH:
+					ring_req->operation =
+						BLKIF_OP_FLUSH_DISKCACHE;
+					break;
+				default:
+					ring_req->operation = 0;
+				}
 			}
 			ring_req->u.rw.nr_segments = nseg;
 		}
@@ -582,6 +593,16 @@ static inline void flush_requests(struct blkfront_info *info)
 		notify_remote_via_irq(info->irq);
 }
 
+static inline bool blkif_request_flush_invalid(struct request *req,
+					       struct blkfront_info *info)
+{
+	return ((req->cmd_type != REQ_TYPE_FS) ||
+		((req->cmd_flags & REQ_FLUSH) &&
+		 !(info->feature_flush & REQ_FLUSH)) ||
+		((req->cmd_flags & REQ_FUA) &&
+		 !(info->feature_flush & REQ_FUA)));
+}
+
 /*
  * do_blkif_request
  *  read a block; request is in a request queue
@@ -604,10 +625,8 @@ static void do_blkif_request(struct request_queue *rq)
 
 		blk_start_request(req);
 
-		if ((req->cmd_type != REQ_TYPE_FS) ||
-		    ((req->cmd_flags & (REQ_FLUSH | REQ_FUA)) &&
-		    !info->flush_op)) {
-			__blk_end_request_all(req, -EIO);
+		if (blkif_request_flush_invalid(req, info)) {
+			__blk_end_request_all(req, -EOPNOTSUPP);
 			continue;
 		}
 
@@ -677,20 +696,26 @@ static int xlvbd_init_blk_queue(struct gendisk *gd, u16 sector_size,
 	return 0;
 }
 
+static const char *flush_info(unsigned int feature_flush)
+{
+	switch (feature_flush & ((REQ_FLUSH | REQ_FUA))) {
+	case REQ_FLUSH|REQ_FUA:
+		return "barrier: enabled;";
+	case REQ_FLUSH:
+		return "flush diskcache: enabled;";
+	default:
+		return "barrier or flush: disabled;";
+	}
+}
 
 static void xlvbd_flush(struct blkfront_info *info)
 {
 	blk_queue_flush(info->rq, info->feature_flush);
-	printk(KERN_INFO "blkfront: %s: %s: %s %s %s %s %s\n",
-	       info->gd->disk_name,
-	       info->flush_op == BLKIF_OP_WRITE_BARRIER ?
-		"barrier" : (info->flush_op == BLKIF_OP_FLUSH_DISKCACHE ?
-		"flush diskcache" : "barrier or flush"),
-	       info->feature_flush ? "enabled;" : "disabled;",
-	       "persistent grants:",
-	       info->feature_persistent ? "enabled;" : "disabled;",
-	       "indirect descriptors:",
-	       info->max_indirect_segments ? "enabled;" : "disabled;");
+	pr_info("blkfront: %s: %s %s %s %s %s\n",
+		info->gd->disk_name, flush_info(info->feature_flush),
+		"persistent grants:", info->feature_persistent ?
+		"enabled;" : "disabled;", "indirect descriptors:",
+		info->max_indirect_segments ? "enabled;" : "disabled;");
 }
 
 static int xen_translate_vdev(int vdevice, int *minor, unsigned int *offset)
@@ -1182,7 +1207,6 @@ static irqreturn_t blkif_interrupt(int irq, void *dev_id)
 				if (error == -EOPNOTSUPP)
 					error = 0;
 				info->feature_flush = 0;
-				info->flush_op = 0;
 				xlvbd_flush(info);
 			}
 			/* fall through */
@@ -1802,7 +1826,6 @@ static void blkfront_connect(struct blkfront_info *info)
 		physical_sector_size = sector_size;
 
 	info->feature_flush = 0;
-	info->flush_op = 0;
 
 	err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
 			    "feature-barrier", "%d", &barrier,
@@ -1815,10 +1838,8 @@ static void blkfront_connect(struct blkfront_info *info)
 	 *
 	 * If there are barriers, then we use flush.
 	 */
-	if (!err && barrier) {
+	if (!err && barrier)
 		info->feature_flush = REQ_FLUSH | REQ_FUA;
-		info->flush_op = BLKIF_OP_WRITE_BARRIER;
-	}
 	/*
 	 * And if there is "feature-flush-cache" use that above
 	 * barriers.
@@ -1827,10 +1848,8 @@ static void blkfront_connect(struct blkfront_info *info)
 			    "feature-flush-cache", "%d", &flush,
 			    NULL);
 
-	if (!err && flush) {
+	if (!err && flush)
 		info->feature_flush = REQ_FLUSH;
-		info->flush_op = BLKIF_OP_FLUSH_DISKCACHE;
-	}
 
 	err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
 			    "feature-discard", "%d", &discard,
@@ -2055,13 +2074,14 @@ static const struct xenbus_device_id blkfront_ids[] = {
 	{ "" }
 };
 
-static DEFINE_XENBUS_DRIVER(blkfront, ,
+static struct xenbus_driver blkfront_driver = {
+	.ids  = blkfront_ids,
 	.probe = blkfront_probe,
 	.remove = blkfront_remove,
 	.resume = blkfront_resume,
 	.otherend_changed = blkback_changed,
 	.is_ready = blkfront_is_ready,
-);
+};
 
 static int __init xlblk_init(void)
 {
diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c
index d00831c3d731..bd8bda386e02 100644
--- a/drivers/block/zram/zram_drv.c
+++ b/drivers/block/zram/zram_drv.c
@@ -44,15 +44,14 @@ static const char *default_compressor = "lzo";
 static unsigned int num_devices = 1;
 
 #define ZRAM_ATTR_RO(name)						\
-static ssize_t zram_attr_##name##_show(struct device *d,		\
+static ssize_t name##_show(struct device *d,		\
 				struct device_attribute *attr, char *b)	\
 {									\
 	struct zram *zram = dev_to_zram(d);				\
 	return scnprintf(b, PAGE_SIZE, "%llu\n",			\
 		(u64)atomic64_read(&zram->stats.name));			\
 }									\
-static struct device_attribute dev_attr_##name =			\
-	__ATTR(name, S_IRUGO, zram_attr_##name##_show, NULL);
+static DEVICE_ATTR_RO(name);
 
 static inline int init_done(struct zram *zram)
 {
@@ -99,14 +98,15 @@ static ssize_t mem_used_total_show(struct device *dev,
 {
 	u64 val = 0;
 	struct zram *zram = dev_to_zram(dev);
-	struct zram_meta *meta = zram->meta;
 
 	down_read(&zram->init_lock);
-	if (init_done(zram))
-		val = zs_get_total_size_bytes(meta->mem_pool);
+	if (init_done(zram)) {
+		struct zram_meta *meta = zram->meta;
+		val = zs_get_total_pages(meta->mem_pool);
+	}
 	up_read(&zram->init_lock);
 
-	return scnprintf(buf, PAGE_SIZE, "%llu\n", val);
+	return scnprintf(buf, PAGE_SIZE, "%llu\n", val << PAGE_SHIFT);
 }
 
 static ssize_t max_comp_streams_show(struct device *dev,
@@ -122,6 +122,73 @@ static ssize_t max_comp_streams_show(struct device *dev,
 	return scnprintf(buf, PAGE_SIZE, "%d\n", val);
 }
 
+static ssize_t mem_limit_show(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	u64 val;
+	struct zram *zram = dev_to_zram(dev);
+
+	down_read(&zram->init_lock);
+	val = zram->limit_pages;
+	up_read(&zram->init_lock);
+
+	return scnprintf(buf, PAGE_SIZE, "%llu\n", val << PAGE_SHIFT);
+}
+
+static ssize_t mem_limit_store(struct device *dev,
+		struct device_attribute *attr, const char *buf, size_t len)
+{
+	u64 limit;
+	char *tmp;
+	struct zram *zram = dev_to_zram(dev);
+
+	limit = memparse(buf, &tmp);
+	if (buf == tmp) /* no chars parsed, invalid input */
+		return -EINVAL;
+
+	down_write(&zram->init_lock);
+	zram->limit_pages = PAGE_ALIGN(limit) >> PAGE_SHIFT;
+	up_write(&zram->init_lock);
+
+	return len;
+}
+
+static ssize_t mem_used_max_show(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	u64 val = 0;
+	struct zram *zram = dev_to_zram(dev);
+
+	down_read(&zram->init_lock);
+	if (init_done(zram))
+		val = atomic_long_read(&zram->stats.max_used_pages);
+	up_read(&zram->init_lock);
+
+	return scnprintf(buf, PAGE_SIZE, "%llu\n", val << PAGE_SHIFT);
+}
+
+static ssize_t mem_used_max_store(struct device *dev,
+		struct device_attribute *attr, const char *buf, size_t len)
+{
+	int err;
+	unsigned long val;
+	struct zram *zram = dev_to_zram(dev);
+
+	err = kstrtoul(buf, 10, &val);
+	if (err || val != 0)
+		return -EINVAL;
+
+	down_read(&zram->init_lock);
+	if (init_done(zram)) {
+		struct zram_meta *meta = zram->meta;
+		atomic_long_set(&zram->stats.max_used_pages,
+				zs_get_total_pages(meta->mem_pool));
+	}
+	up_read(&zram->init_lock);
+
+	return len;
+}
+
 static ssize_t max_comp_streams_store(struct device *dev,
 		struct device_attribute *attr, const char *buf, size_t len)
 {
@@ -219,19 +286,18 @@ static inline int is_partial_io(struct bio_vec *bvec)
 /*
  * Check if request is within bounds and aligned on zram logical blocks.
  */
-static inline int valid_io_request(struct zram *zram, struct bio *bio)
+static inline int valid_io_request(struct zram *zram,
+		sector_t start, unsigned int size)
 {
-	u64 start, end, bound;
+	u64 end, bound;
 
 	/* unaligned request */
-	if (unlikely(bio->bi_iter.bi_sector &
-		     (ZRAM_SECTOR_PER_LOGICAL_BLOCK - 1)))
+	if (unlikely(start & (ZRAM_SECTOR_PER_LOGICAL_BLOCK - 1)))
 		return 0;
-	if (unlikely(bio->bi_iter.bi_size & (ZRAM_LOGICAL_BLOCK_SIZE - 1)))
+	if (unlikely(size & (ZRAM_LOGICAL_BLOCK_SIZE - 1)))
 		return 0;
 
-	start = bio->bi_iter.bi_sector;
-	end = start + (bio->bi_iter.bi_size >> SECTOR_SHIFT);
+	end = start + (size >> SECTOR_SHIFT);
 	bound = zram->disksize >> SECTOR_SHIFT;
 	/* out of range range */
 	if (unlikely(start >= bound || end > bound || start > end))
@@ -385,7 +451,7 @@ static int zram_decompress_page(struct zram *zram, char *mem, u32 index)
 }
 
 static int zram_bvec_read(struct zram *zram, struct bio_vec *bvec,
-			  u32 index, int offset, struct bio *bio)
+			  u32 index, int offset)
 {
 	int ret;
 	struct page *page;
@@ -434,6 +500,21 @@ out_cleanup:
 	return ret;
 }
 
+static inline void update_used_max(struct zram *zram,
+					const unsigned long pages)
+{
+	int old_max, cur_max;
+
+	old_max = atomic_long_read(&zram->stats.max_used_pages);
+
+	do {
+		cur_max = old_max;
+		if (pages > cur_max)
+			old_max = atomic_long_cmpxchg(
+				&zram->stats.max_used_pages, cur_max, pages);
+	} while (old_max != cur_max);
+}
+
 static int zram_bvec_write(struct zram *zram, struct bio_vec *bvec, u32 index,
 			   int offset)
 {
@@ -445,6 +526,7 @@ static int zram_bvec_write(struct zram *zram, struct bio_vec *bvec, u32 index,
 	struct zram_meta *meta = zram->meta;
 	struct zcomp_strm *zstrm;
 	bool locked = false;
+	unsigned long alloced_pages;
 
 	page = bvec->bv_page;
 	if (is_partial_io(bvec)) {
@@ -476,7 +558,8 @@ static int zram_bvec_write(struct zram *zram, struct bio_vec *bvec, u32 index,
 	}
 
 	if (page_zero_filled(uncmem)) {
-		kunmap_atomic(user_mem);
+		if (user_mem)
+			kunmap_atomic(user_mem);
 		/* Free memory associated with this sector now. */
 		bit_spin_lock(ZRAM_ACCESS, &meta->table[index].value);
 		zram_free_page(zram, index);
@@ -513,6 +596,16 @@ static int zram_bvec_write(struct zram *zram, struct bio_vec *bvec, u32 index,
 		ret = -ENOMEM;
 		goto out;
 	}
+
+	alloced_pages = zs_get_total_pages(meta->mem_pool);
+	if (zram->limit_pages && alloced_pages > zram->limit_pages) {
+		zs_free(meta->mem_pool, handle);
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	update_used_max(zram, alloced_pages);
+
 	cmem = zs_map_object(meta->mem_pool, handle, ZS_MM_WO);
 
 	if ((clen == PAGE_SIZE) && !is_partial_io(bvec)) {
@@ -550,14 +643,13 @@ out:
 }
 
 static int zram_bvec_rw(struct zram *zram, struct bio_vec *bvec, u32 index,
-			int offset, struct bio *bio)
+			int offset, int rw)
 {
 	int ret;
-	int rw = bio_data_dir(bio);
 
 	if (rw == READ) {
 		atomic64_inc(&zram->stats.num_reads);
-		ret = zram_bvec_read(zram, bvec, index, offset, bio);
+		ret = zram_bvec_read(zram, bvec, index, offset);
 	} else {
 		atomic64_inc(&zram->stats.num_writes);
 		ret = zram_bvec_write(zram, bvec, index, offset);
@@ -606,6 +698,7 @@ static void zram_bio_discard(struct zram *zram, u32 index,
 		bit_spin_lock(ZRAM_ACCESS, &meta->table[index].value);
 		zram_free_page(zram, index);
 		bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value);
+		atomic64_inc(&zram->stats.notify_free);
 		index++;
 		n -= PAGE_SIZE;
 	}
@@ -617,6 +710,9 @@ static void zram_reset_device(struct zram *zram, bool reset_capacity)
 	struct zram_meta *meta;
 
 	down_write(&zram->init_lock);
+
+	zram->limit_pages = 0;
+
 	if (!init_done(zram)) {
 		up_write(&zram->init_lock);
 		return;
@@ -754,7 +850,7 @@ out:
 
 static void __zram_make_request(struct zram *zram, struct bio *bio)
 {
-	int offset;
+	int offset, rw;
 	u32 index;
 	struct bio_vec bvec;
 	struct bvec_iter iter;
@@ -769,6 +865,7 @@ static void __zram_make_request(struct zram *zram, struct bio *bio)
 		return;
 	}
 
+	rw = bio_data_dir(bio);
 	bio_for_each_segment(bvec, bio, iter) {
 		int max_transfer_size = PAGE_SIZE - offset;
 
@@ -783,15 +880,15 @@ static void __zram_make_request(struct zram *zram, struct bio *bio)
 			bv.bv_len = max_transfer_size;
 			bv.bv_offset = bvec.bv_offset;
 
-			if (zram_bvec_rw(zram, &bv, index, offset, bio) < 0)
+			if (zram_bvec_rw(zram, &bv, index, offset, rw) < 0)
 				goto out;
 
 			bv.bv_len = bvec.bv_len - max_transfer_size;
 			bv.bv_offset += max_transfer_size;
-			if (zram_bvec_rw(zram, &bv, index + 1, 0, bio) < 0)
+			if (zram_bvec_rw(zram, &bv, index + 1, 0, rw) < 0)
 				goto out;
 		} else
-			if (zram_bvec_rw(zram, &bvec, index, offset, bio) < 0)
+			if (zram_bvec_rw(zram, &bvec, index, offset, rw) < 0)
 				goto out;
 
 		update_position(&index, &offset, &bvec);
@@ -816,7 +913,8 @@ static void zram_make_request(struct request_queue *queue, struct bio *bio)
 	if (unlikely(!init_done(zram)))
 		goto error;
 
-	if (!valid_io_request(zram, bio)) {
+	if (!valid_io_request(zram, bio->bi_iter.bi_sector,
+					bio->bi_iter.bi_size)) {
 		atomic64_inc(&zram->stats.invalid_io);
 		goto error;
 	}
@@ -846,21 +944,64 @@ static void zram_slot_free_notify(struct block_device *bdev,
 	atomic64_inc(&zram->stats.notify_free);
 }
 
+static int zram_rw_page(struct block_device *bdev, sector_t sector,
+		       struct page *page, int rw)
+{
+	int offset, err;
+	u32 index;
+	struct zram *zram;
+	struct bio_vec bv;
+
+	zram = bdev->bd_disk->private_data;
+	if (!valid_io_request(zram, sector, PAGE_SIZE)) {
+		atomic64_inc(&zram->stats.invalid_io);
+		return -EINVAL;
+	}
+
+	down_read(&zram->init_lock);
+	if (unlikely(!init_done(zram))) {
+		err = -EIO;
+		goto out_unlock;
+	}
+
+	index = sector >> SECTORS_PER_PAGE_SHIFT;
+	offset = sector & (SECTORS_PER_PAGE - 1) << SECTOR_SHIFT;
+
+	bv.bv_page = page;
+	bv.bv_len = PAGE_SIZE;
+	bv.bv_offset = 0;
+
+	err = zram_bvec_rw(zram, &bv, index, offset, rw);
+out_unlock:
+	up_read(&zram->init_lock);
+	/*
+	 * If I/O fails, just return error(ie, non-zero) without
+	 * calling page_endio.
+	 * It causes resubmit the I/O with bio request by upper functions
+	 * of rw_page(e.g., swap_readpage, __swap_writepage) and
+	 * bio->bi_end_io does things to handle the error
+	 * (e.g., SetPageError, set_page_dirty and extra works).
+	 */
+	if (err == 0)
+		page_endio(page, rw, 0);
+	return err;
+}
+
 static const struct block_device_operations zram_devops = {
 	.swap_slot_free_notify = zram_slot_free_notify,
+	.rw_page = zram_rw_page,
 	.owner = THIS_MODULE
 };
 
-static DEVICE_ATTR(disksize, S_IRUGO | S_IWUSR,
-		disksize_show, disksize_store);
-static DEVICE_ATTR(initstate, S_IRUGO, initstate_show, NULL);
-static DEVICE_ATTR(reset, S_IWUSR, NULL, reset_store);
-static DEVICE_ATTR(orig_data_size, S_IRUGO, orig_data_size_show, NULL);
-static DEVICE_ATTR(mem_used_total, S_IRUGO, mem_used_total_show, NULL);
-static DEVICE_ATTR(max_comp_streams, S_IRUGO | S_IWUSR,
-		max_comp_streams_show, max_comp_streams_store);
-static DEVICE_ATTR(comp_algorithm, S_IRUGO | S_IWUSR,
-		comp_algorithm_show, comp_algorithm_store);
+static DEVICE_ATTR_RW(disksize);
+static DEVICE_ATTR_RO(initstate);
+static DEVICE_ATTR_WO(reset);
+static DEVICE_ATTR_RO(orig_data_size);
+static DEVICE_ATTR_RO(mem_used_total);
+static DEVICE_ATTR_RW(mem_limit);
+static DEVICE_ATTR_RW(mem_used_max);
+static DEVICE_ATTR_RW(max_comp_streams);
+static DEVICE_ATTR_RW(comp_algorithm);
 
 ZRAM_ATTR_RO(num_reads);
 ZRAM_ATTR_RO(num_writes);
@@ -885,6 +1026,8 @@ static struct attribute *zram_disk_attrs[] = {
 	&dev_attr_orig_data_size.attr,
 	&dev_attr_compr_data_size.attr,
 	&dev_attr_mem_used_total.attr,
+	&dev_attr_mem_limit.attr,
+	&dev_attr_mem_used_max.attr,
 	&dev_attr_max_comp_streams.attr,
 	&dev_attr_comp_algorithm.attr,
 	NULL,
@@ -929,6 +1072,7 @@ static int create_device(struct zram *zram, int device_id)
 	set_capacity(zram->disk, 0);
 	/* zram devices sort of resembles non-rotational disks */
 	queue_flag_set_unlocked(QUEUE_FLAG_NONROT, zram->disk->queue);
+	queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, zram->disk->queue);
 	/*
 	 * To ensure that we always get PAGE_SIZE aligned
 	 * and n*PAGE_SIZED sized I/O requests.
diff --git a/drivers/block/zram/zram_drv.h b/drivers/block/zram/zram_drv.h
index e0f725c87cc6..b05a816b09ac 100644
--- a/drivers/block/zram/zram_drv.h
+++ b/drivers/block/zram/zram_drv.h
@@ -66,8 +66,8 @@ static const size_t max_zpage_size = PAGE_SIZE / 4 * 3;
 /* Flags for zram pages (table[page_no].value) */
 enum zram_pageflags {
 	/* Page consists entirely of zeros */
-	ZRAM_ZERO = ZRAM_FLAG_SHIFT + 1,
-	ZRAM_ACCESS,	/* page in now accessed */
+	ZRAM_ZERO = ZRAM_FLAG_SHIFT,
+	ZRAM_ACCESS,	/* page is now accessed */
 
 	__NR_ZRAM_PAGEFLAGS,
 };
@@ -90,6 +90,7 @@ struct zram_stats {
 	atomic64_t notify_free;	/* no. of swap slot free notifications */
 	atomic64_t zero_pages;		/* no. of zero filled pages */
 	atomic64_t pages_stored;	/* no. of pages currently stored */
+	atomic_long_t max_used_pages;	/* no. of maximum pages stored */
 };
 
 struct zram_meta {
@@ -112,6 +113,11 @@ struct zram {
 	u64 disksize;	/* bytes */
 	int max_comp_streams;
 	struct zram_stats stats;
+	/*
+	 * the number of pages zram can consume for storing compressed data
+	 */
+	unsigned long limit_pages;
+
 	char compressor[10];
 };
 #endif
diff --git a/drivers/bluetooth/Kconfig b/drivers/bluetooth/Kconfig
index fa7fd62ddffa..364f080768d0 100644
--- a/drivers/bluetooth/Kconfig
+++ b/drivers/bluetooth/Kconfig
@@ -201,7 +201,7 @@ config BT_MRVL
 	  The core driver to support Marvell Bluetooth devices.
 
 	  This driver is required if you want to support
-	  Marvell Bluetooth devices, such as 8688/8787/8797/8897.
+	  Marvell Bluetooth devices, such as 8688/8787/8797/8887/8897.
 
 	  Say Y here to compile Marvell Bluetooth driver
 	  into the kernel or say M to compile it as module.
@@ -210,11 +210,12 @@ config BT_MRVL_SDIO
 	tristate "Marvell BT-over-SDIO driver"
 	depends on BT_MRVL && MMC
 	select FW_LOADER
+	select WANT_DEV_COREDUMP
 	help
 	  The driver for Marvell Bluetooth chipsets with SDIO interface.
 
 	  This driver is required if you want to use Marvell Bluetooth
-	  devices with SDIO interface. Currently SD8688/SD8787/SD8797/SD8897
+	  devices with SDIO interface. Currently SD8688/SD8787/SD8797/SD8887/SD8897
 	  chipsets are supported.
 
 	  Say Y here to compile support for Marvell BT-over-SDIO driver
diff --git a/drivers/bluetooth/ath3k.c b/drivers/bluetooth/ath3k.c
index a0d7355ef127..1ee27ac18de0 100644
--- a/drivers/bluetooth/ath3k.c
+++ b/drivers/bluetooth/ath3k.c
@@ -79,6 +79,7 @@ static const struct usb_device_id ath3k_table[] = {
 	{ USB_DEVICE(0x0489, 0xe057) },
 	{ USB_DEVICE(0x0489, 0xe056) },
 	{ USB_DEVICE(0x0489, 0xe05f) },
+	{ USB_DEVICE(0x0489, 0xe078) },
 	{ USB_DEVICE(0x04c5, 0x1330) },
 	{ USB_DEVICE(0x04CA, 0x3004) },
 	{ USB_DEVICE(0x04CA, 0x3005) },
@@ -86,8 +87,10 @@ static const struct usb_device_id ath3k_table[] = {
 	{ USB_DEVICE(0x04CA, 0x3007) },
 	{ USB_DEVICE(0x04CA, 0x3008) },
 	{ USB_DEVICE(0x04CA, 0x300b) },
+	{ USB_DEVICE(0x04CA, 0x3010) },
 	{ USB_DEVICE(0x0930, 0x0219) },
 	{ USB_DEVICE(0x0930, 0x0220) },
+	{ USB_DEVICE(0x0930, 0x0227) },
 	{ USB_DEVICE(0x0b05, 0x17d0) },
 	{ USB_DEVICE(0x0CF3, 0x0036) },
 	{ USB_DEVICE(0x0CF3, 0x3004) },
@@ -104,6 +107,7 @@ static const struct usb_device_id ath3k_table[] = {
 	{ USB_DEVICE(0x13d3, 0x3375) },
 	{ USB_DEVICE(0x13d3, 0x3393) },
 	{ USB_DEVICE(0x13d3, 0x3402) },
+	{ USB_DEVICE(0x13d3, 0x3408) },
 	{ USB_DEVICE(0x13d3, 0x3432) },
 
 	/* Atheros AR5BBU12 with sflash firmware */
@@ -129,6 +133,7 @@ static const struct usb_device_id ath3k_blist_tbl[] = {
 	{ USB_DEVICE(0x0489, 0xe056), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0489, 0xe057), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0489, 0xe05f), .driver_info = BTUSB_ATH3012 },
+	{ USB_DEVICE(0x0489, 0xe078), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x04c5, 0x1330), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x04ca, 0x3004), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x04ca, 0x3005), .driver_info = BTUSB_ATH3012 },
@@ -136,8 +141,10 @@ static const struct usb_device_id ath3k_blist_tbl[] = {
 	{ USB_DEVICE(0x04ca, 0x3007), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x04ca, 0x3008), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x04ca, 0x300b), .driver_info = BTUSB_ATH3012 },
+	{ USB_DEVICE(0x04ca, 0x3010), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0930, 0x0219), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0930, 0x0220), .driver_info = BTUSB_ATH3012 },
+	{ USB_DEVICE(0x0930, 0x0227), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0b05, 0x17d0), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0CF3, 0x0036), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0cf3, 0x3004), .driver_info = BTUSB_ATH3012 },
@@ -154,6 +161,7 @@ static const struct usb_device_id ath3k_blist_tbl[] = {
 	{ USB_DEVICE(0x13d3, 0x3375), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x13d3, 0x3393), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x13d3, 0x3402), .driver_info = BTUSB_ATH3012 },
+	{ USB_DEVICE(0x13d3, 0x3408), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x13d3, 0x3432), .driver_info = BTUSB_ATH3012 },
 
 	/* Atheros AR5BBU22 with sflash firmware */
diff --git a/drivers/bluetooth/bluecard_cs.c b/drivers/bluetooth/bluecard_cs.c
index dfa5043e68ba..35e63aaa6f80 100644
--- a/drivers/bluetooth/bluecard_cs.c
+++ b/drivers/bluetooth/bluecard_cs.c
@@ -61,7 +61,7 @@ MODULE_LICENSE("GPL");
 /* ======================== Local structures ======================== */
 
 
-typedef struct bluecard_info_t {
+struct bluecard_info {
 	struct pcmcia_device *p_dev;
 
 	struct hci_dev *hdev;
@@ -78,7 +78,7 @@ typedef struct bluecard_info_t {
 
 	unsigned char ctrl_reg;
 	unsigned long hw_state;		/* Status of the hardware and LED control */
-} bluecard_info_t;
+};
 
 
 static int bluecard_config(struct pcmcia_device *link);
@@ -157,7 +157,7 @@ static void bluecard_detach(struct pcmcia_device *p_dev);
 
 static void bluecard_activity_led_timeout(u_long arg)
 {
-	bluecard_info_t *info = (bluecard_info_t *)arg;
+	struct bluecard_info *info = (struct bluecard_info *)arg;
 	unsigned int iobase = info->p_dev->resource[0]->start;
 
 	if (!test_bit(CARD_HAS_PCCARD_ID, &(info->hw_state)))
@@ -173,7 +173,7 @@ static void bluecard_activity_led_timeout(u_long arg)
 }
 
 
-static void bluecard_enable_activity_led(bluecard_info_t *info)
+static void bluecard_enable_activity_led(struct bluecard_info *info)
 {
 	unsigned int iobase = info->p_dev->resource[0]->start;
 
@@ -215,7 +215,7 @@ static int bluecard_write(unsigned int iobase, unsigned int offset, __u8 *buf, i
 }
 
 
-static void bluecard_write_wakeup(bluecard_info_t *info)
+static void bluecard_write_wakeup(struct bluecard_info *info)
 {
 	if (!info) {
 		BT_ERR("Unknown device");
@@ -368,7 +368,8 @@ static int bluecard_read(unsigned int iobase, unsigned int offset, __u8 *buf, in
 }
 
 
-static void bluecard_receive(bluecard_info_t *info, unsigned int offset)
+static void bluecard_receive(struct bluecard_info *info,
+			     unsigned int offset)
 {
 	unsigned int iobase;
 	unsigned char buf[31];
@@ -497,7 +498,7 @@ static void bluecard_receive(bluecard_info_t *info, unsigned int offset)
 
 static irqreturn_t bluecard_interrupt(int irq, void *dev_inst)
 {
-	bluecard_info_t *info = dev_inst;
+	struct bluecard_info *info = dev_inst;
 	unsigned int iobase;
 	unsigned char reg;
 
@@ -562,7 +563,7 @@ static irqreturn_t bluecard_interrupt(int irq, void *dev_inst)
 
 static int bluecard_hci_set_baud_rate(struct hci_dev *hdev, int baud)
 {
-	bluecard_info_t *info = hci_get_drvdata(hdev);
+	struct bluecard_info *info = hci_get_drvdata(hdev);
 	struct sk_buff *skb;
 
 	/* Ericsson baud rate command */
@@ -611,7 +612,7 @@ static int bluecard_hci_set_baud_rate(struct hci_dev *hdev, int baud)
 
 static int bluecard_hci_flush(struct hci_dev *hdev)
 {
-	bluecard_info_t *info = hci_get_drvdata(hdev);
+	struct bluecard_info *info = hci_get_drvdata(hdev);
 
 	/* Drop TX queue */
 	skb_queue_purge(&(info->txq));
@@ -622,7 +623,7 @@ static int bluecard_hci_flush(struct hci_dev *hdev)
 
 static int bluecard_hci_open(struct hci_dev *hdev)
 {
-	bluecard_info_t *info = hci_get_drvdata(hdev);
+	struct bluecard_info *info = hci_get_drvdata(hdev);
 
 	if (test_bit(CARD_HAS_PCCARD_ID, &(info->hw_state)))
 		bluecard_hci_set_baud_rate(hdev, DEFAULT_BAUD_RATE);
@@ -643,7 +644,7 @@ static int bluecard_hci_open(struct hci_dev *hdev)
 
 static int bluecard_hci_close(struct hci_dev *hdev)
 {
-	bluecard_info_t *info = hci_get_drvdata(hdev);
+	struct bluecard_info *info = hci_get_drvdata(hdev);
 
 	if (!test_and_clear_bit(HCI_RUNNING, &(hdev->flags)))
 		return 0;
@@ -663,7 +664,7 @@ static int bluecard_hci_close(struct hci_dev *hdev)
 
 static int bluecard_hci_send_frame(struct hci_dev *hdev, struct sk_buff *skb)
 {
-	bluecard_info_t *info = hci_get_drvdata(hdev);
+	struct bluecard_info *info = hci_get_drvdata(hdev);
 
 	switch (bt_cb(skb)->pkt_type) {
 	case HCI_COMMAND_PKT:
@@ -691,7 +692,7 @@ static int bluecard_hci_send_frame(struct hci_dev *hdev, struct sk_buff *skb)
 /* ======================== Card services HCI interaction ======================== */
 
 
-static int bluecard_open(bluecard_info_t *info)
+static int bluecard_open(struct bluecard_info *info)
 {
 	unsigned int iobase = info->p_dev->resource[0]->start;
 	struct hci_dev *hdev;
@@ -806,7 +807,7 @@ static int bluecard_open(bluecard_info_t *info)
 }
 
 
-static int bluecard_close(bluecard_info_t *info)
+static int bluecard_close(struct bluecard_info *info)
 {
 	unsigned int iobase = info->p_dev->resource[0]->start;
 	struct hci_dev *hdev = info->hdev;
@@ -833,7 +834,7 @@ static int bluecard_close(bluecard_info_t *info)
 
 static int bluecard_probe(struct pcmcia_device *link)
 {
-	bluecard_info_t *info;
+	struct bluecard_info *info;
 
 	/* Create new info device */
 	info = devm_kzalloc(&link->dev, sizeof(*info), GFP_KERNEL);
@@ -857,7 +858,7 @@ static void bluecard_detach(struct pcmcia_device *link)
 
 static int bluecard_config(struct pcmcia_device *link)
 {
-	bluecard_info_t *info = link->priv;
+	struct bluecard_info *info = link->priv;
 	int i, n;
 
 	link->config_index = 0x20;
@@ -897,7 +898,7 @@ failed:
 
 static void bluecard_release(struct pcmcia_device *link)
 {
-	bluecard_info_t *info = link->priv;
+	struct bluecard_info *info = link->priv;
 
 	bluecard_close(info);
 
diff --git a/drivers/bluetooth/bt3c_cs.c b/drivers/bluetooth/bt3c_cs.c
index 1d82721cf9c6..4f7e8d400bc0 100644
--- a/drivers/bluetooth/bt3c_cs.c
+++ b/drivers/bluetooth/bt3c_cs.c
@@ -67,7 +67,7 @@ MODULE_FIRMWARE("BT3CPCC.bin");
 /* ======================== Local structures ======================== */
 
 
-typedef struct bt3c_info_t {
+struct bt3c_info {
 	struct pcmcia_device *p_dev;
 
 	struct hci_dev *hdev;
@@ -80,7 +80,7 @@ typedef struct bt3c_info_t {
 	unsigned long rx_state;
 	unsigned long rx_count;
 	struct sk_buff *rx_skb;
-} bt3c_info_t;
+};
 
 
 static int bt3c_config(struct pcmcia_device *link);
@@ -175,7 +175,7 @@ static int bt3c_write(unsigned int iobase, int fifo_size, __u8 *buf, int len)
 }
 
 
-static void bt3c_write_wakeup(bt3c_info_t *info)
+static void bt3c_write_wakeup(struct bt3c_info *info)
 {
 	if (!info) {
 		BT_ERR("Unknown device");
@@ -214,7 +214,7 @@ static void bt3c_write_wakeup(bt3c_info_t *info)
 }
 
 
-static void bt3c_receive(bt3c_info_t *info)
+static void bt3c_receive(struct bt3c_info *info)
 {
 	unsigned int iobase;
 	int size = 0, avail;
@@ -336,7 +336,7 @@ static void bt3c_receive(bt3c_info_t *info)
 
 static irqreturn_t bt3c_interrupt(int irq, void *dev_inst)
 {
-	bt3c_info_t *info = dev_inst;
+	struct bt3c_info *info = dev_inst;
 	unsigned int iobase;
 	int iir;
 	irqreturn_t r = IRQ_NONE;
@@ -388,7 +388,7 @@ static irqreturn_t bt3c_interrupt(int irq, void *dev_inst)
 
 static int bt3c_hci_flush(struct hci_dev *hdev)
 {
-	bt3c_info_t *info = hci_get_drvdata(hdev);
+	struct bt3c_info *info = hci_get_drvdata(hdev);
 
 	/* Drop TX queue */
 	skb_queue_purge(&(info->txq));
@@ -418,7 +418,7 @@ static int bt3c_hci_close(struct hci_dev *hdev)
 
 static int bt3c_hci_send_frame(struct hci_dev *hdev, struct sk_buff *skb)
 {
-	bt3c_info_t *info = hci_get_drvdata(hdev);
+	struct bt3c_info *info = hci_get_drvdata(hdev);
 	unsigned long flags;
 
 	switch (bt_cb(skb)->pkt_type) {
@@ -451,7 +451,8 @@ static int bt3c_hci_send_frame(struct hci_dev *hdev, struct sk_buff *skb)
 /* ======================== Card services HCI interaction ======================== */
 
 
-static int bt3c_load_firmware(bt3c_info_t *info, const unsigned char *firmware,
+static int bt3c_load_firmware(struct bt3c_info *info,
+			      const unsigned char *firmware,
 			      int count)
 {
 	char *ptr = (char *) firmware;
@@ -536,7 +537,7 @@ error:
 }
 
 
-static int bt3c_open(bt3c_info_t *info)
+static int bt3c_open(struct bt3c_info *info)
 {
 	const struct firmware *firmware;
 	struct hci_dev *hdev;
@@ -603,7 +604,7 @@ error:
 }
 
 
-static int bt3c_close(bt3c_info_t *info)
+static int bt3c_close(struct bt3c_info *info)
 {
 	struct hci_dev *hdev = info->hdev;
 
@@ -620,7 +621,7 @@ static int bt3c_close(bt3c_info_t *info)
 
 static int bt3c_probe(struct pcmcia_device *link)
 {
-	bt3c_info_t *info;
+	struct bt3c_info *info;
 
 	/* Create new info device */
 	info = devm_kzalloc(&link->dev, sizeof(*info), GFP_KERNEL);
@@ -683,7 +684,7 @@ static int bt3c_check_config_notpicky(struct pcmcia_device *p_dev,
 
 static int bt3c_config(struct pcmcia_device *link)
 {
-	bt3c_info_t *info = link->priv;
+	struct bt3c_info *info = link->priv;
 	int i;
 	unsigned long try;
 
@@ -724,7 +725,7 @@ failed:
 
 static void bt3c_release(struct pcmcia_device *link)
 {
-	bt3c_info_t *info = link->priv;
+	struct bt3c_info *info = link->priv;
 
 	bt3c_close(info);
 
diff --git a/drivers/bluetooth/btmrvl_debugfs.c b/drivers/bluetooth/btmrvl_debugfs.c
index 023d35e3c7a7..1828ed8cae7a 100644
--- a/drivers/bluetooth/btmrvl_debugfs.c
+++ b/drivers/bluetooth/btmrvl_debugfs.c
@@ -167,6 +167,35 @@ static const struct file_operations btmrvl_hscmd_fops = {
 	.llseek = default_llseek,
 };
 
+static ssize_t btmrvl_fwdump_write(struct file *file, const char __user *ubuf,
+				   size_t count, loff_t *ppos)
+{
+	struct btmrvl_private *priv = file->private_data;
+	char buf[16];
+	bool result;
+
+	memset(buf, 0, sizeof(buf));
+
+	if (copy_from_user(&buf, ubuf, min_t(size_t, sizeof(buf) - 1, count)))
+		return -EFAULT;
+
+	if (strtobool(buf, &result))
+		return -EINVAL;
+
+	if (!result)
+		return -EINVAL;
+
+	btmrvl_firmware_dump(priv);
+
+	return count;
+}
+
+static const struct file_operations btmrvl_fwdump_fops = {
+	.write	= btmrvl_fwdump_write,
+	.open	= simple_open,
+	.llseek = default_llseek,
+};
+
 void btmrvl_debugfs_init(struct hci_dev *hdev)
 {
 	struct btmrvl_private *priv = hci_get_drvdata(hdev);
@@ -197,6 +226,8 @@ void btmrvl_debugfs_init(struct hci_dev *hdev)
 			    priv, &btmrvl_hscmd_fops);
 	debugfs_create_file("hscfgcmd", 0644, dbg->config_dir,
 			    priv, &btmrvl_hscfgcmd_fops);
+	debugfs_create_file("fw_dump", 0200, dbg->config_dir,
+			    priv, &btmrvl_fwdump_fops);
 
 	dbg->status_dir = debugfs_create_dir("status", hdev->debugfs);
 	debugfs_create_u8("curpsmode", 0444, dbg->status_dir,
diff --git a/drivers/bluetooth/btmrvl_drv.h b/drivers/bluetooth/btmrvl_drv.h
index 38ad66289ad6..330f8f84928d 100644
--- a/drivers/bluetooth/btmrvl_drv.h
+++ b/drivers/bluetooth/btmrvl_drv.h
@@ -32,6 +32,24 @@
 /* Time to wait for command response in millisecond */
 #define WAIT_UNTIL_CMD_RESP		5000
 
+enum rdwr_status {
+	RDWR_STATUS_SUCCESS = 0,
+	RDWR_STATUS_FAILURE = 1,
+	RDWR_STATUS_DONE = 2
+};
+
+#define FW_DUMP_MAX_NAME_LEN    8
+#define FW_DUMP_HOST_READY      0xEE
+#define FW_DUMP_DONE            0xFF
+#define FW_DUMP_READ_DONE       0xFE
+
+struct memory_type_mapping {
+	u8 mem_name[FW_DUMP_MAX_NAME_LEN];
+	u8 *mem_ptr;
+	u32 mem_size;
+	u8 done_flag;
+};
+
 struct btmrvl_thread {
 	struct task_struct *task;
 	wait_queue_head_t wait_q;
@@ -81,6 +99,7 @@ struct btmrvl_private {
 				u8 *payload, u16 nb);
 	int (*hw_wakeup_firmware) (struct btmrvl_private *priv);
 	int (*hw_process_int_status) (struct btmrvl_private *priv);
+	void (*firmware_dump)(struct btmrvl_private *priv);
 	spinlock_t driver_lock;		/* spinlock used by driver */
 #ifdef CONFIG_DEBUG_FS
 	void *debugfs_data;
@@ -151,6 +170,7 @@ int btmrvl_send_hscfg_cmd(struct btmrvl_private *priv);
 int btmrvl_enable_ps(struct btmrvl_private *priv);
 int btmrvl_prepare_command(struct btmrvl_private *priv);
 int btmrvl_enable_hs(struct btmrvl_private *priv);
+void btmrvl_firmware_dump(struct btmrvl_private *priv);
 
 #ifdef CONFIG_DEBUG_FS
 void btmrvl_debugfs_init(struct hci_dev *hdev);
diff --git a/drivers/bluetooth/btmrvl_main.c b/drivers/bluetooth/btmrvl_main.c
index 1d7db2064889..30939c993d94 100644
--- a/drivers/bluetooth/btmrvl_main.c
+++ b/drivers/bluetooth/btmrvl_main.c
@@ -22,6 +22,7 @@
 #include <linux/of.h>
 #include <net/bluetooth/bluetooth.h>
 #include <net/bluetooth/hci_core.h>
+#include <linux/mmc/sdio_func.h>
 
 #include "btmrvl_drv.h"
 #include "btmrvl_sdio.h"
@@ -41,6 +42,11 @@ void btmrvl_interrupt(struct btmrvl_private *priv)
 
 	priv->adapter->int_count++;
 
+	if (priv->adapter->hs_state == HS_ACTIVATED) {
+		BT_DBG("BT: HS DEACTIVATED in ISR!");
+		priv->adapter->hs_state = HS_DEACTIVATED;
+	}
+
 	wake_up_interruptible(&priv->main_thread.wait_q);
 }
 EXPORT_SYMBOL_GPL(btmrvl_interrupt);
@@ -209,7 +215,7 @@ int btmrvl_send_module_cfg_cmd(struct btmrvl_private *priv, u8 subcmd)
 
 	ret = btmrvl_send_sync_cmd(priv, BT_CMD_MODULE_CFG_REQ, &subcmd, 1);
 	if (ret)
-		BT_ERR("module_cfg_cmd(%x) failed\n", subcmd);
+		BT_ERR("module_cfg_cmd(%x) failed", subcmd);
 
 	return ret;
 }
@@ -245,7 +251,7 @@ int btmrvl_send_hscfg_cmd(struct btmrvl_private *priv)
 
 	ret = btmrvl_send_sync_cmd(priv, BT_CMD_HOST_SLEEP_CONFIG, param, 2);
 	if (ret)
-		BT_ERR("HSCFG command failed\n");
+		BT_ERR("HSCFG command failed");
 
 	return ret;
 }
@@ -263,7 +269,7 @@ int btmrvl_enable_ps(struct btmrvl_private *priv)
 
 	ret = btmrvl_send_sync_cmd(priv, BT_CMD_AUTO_SLEEP_MODE, &param, 1);
 	if (ret)
-		BT_ERR("PSMODE command failed\n");
+		BT_ERR("PSMODE command failed");
 
 	return 0;
 }
@@ -276,7 +282,7 @@ int btmrvl_enable_hs(struct btmrvl_private *priv)
 
 	ret = btmrvl_send_sync_cmd(priv, BT_CMD_HOST_SLEEP_ENABLE, NULL, 0);
 	if (ret) {
-		BT_ERR("Host sleep enable command failed\n");
+		BT_ERR("Host sleep enable command failed");
 		return ret;
 	}
 
@@ -323,12 +329,19 @@ int btmrvl_prepare_command(struct btmrvl_private *priv)
 		} else {
 			ret = priv->hw_wakeup_firmware(priv);
 			priv->adapter->hs_state = HS_DEACTIVATED;
+			BT_DBG("BT: HS DEACTIVATED due to host activity!");
 		}
 	}
 
 	return ret;
 }
 
+void btmrvl_firmware_dump(struct btmrvl_private *priv)
+{
+	if (priv->firmware_dump)
+		priv->firmware_dump(priv);
+}
+
 static int btmrvl_tx_pkt(struct btmrvl_private *priv, struct sk_buff *skb)
 {
 	int ret = 0;
@@ -487,34 +500,36 @@ static int btmrvl_download_cal_data(struct btmrvl_private *priv,
 	ret = btmrvl_send_sync_cmd(priv, BT_CMD_LOAD_CONFIG_DATA, data,
 				   BT_CAL_HDR_LEN + len);
 	if (ret)
-		BT_ERR("Failed to download caibration data\n");
+		BT_ERR("Failed to download caibration data");
 
 	return 0;
 }
 
-static int btmrvl_cal_data_dt(struct btmrvl_private *priv)
+static int btmrvl_check_device_tree(struct btmrvl_private *priv)
 {
 	struct device_node *dt_node;
 	u8 cal_data[BT_CAL_HDR_LEN + BT_CAL_DATA_SIZE];
-	const char name[] = "btmrvl_caldata";
-	const char property[] = "btmrvl,caldata";
 	int ret;
-
-	dt_node = of_find_node_by_name(NULL, name);
-	if (!dt_node)
-		return -ENODEV;
-
-	ret = of_property_read_u8_array(dt_node, property,
-					cal_data + BT_CAL_HDR_LEN,
-					BT_CAL_DATA_SIZE);
-	if (ret)
-		return ret;
-
-	BT_DBG("Use cal data from device tree");
-	ret = btmrvl_download_cal_data(priv, cal_data, BT_CAL_DATA_SIZE);
-	if (ret) {
-		BT_ERR("Fail to download calibrate data");
-		return ret;
+	u32 val;
+
+	for_each_compatible_node(dt_node, NULL, "btmrvl,cfgdata") {
+		ret = of_property_read_u32(dt_node, "btmrvl,gpio-gap", &val);
+		if (!ret)
+			priv->btmrvl_dev.gpio_gap = val;
+
+		ret = of_property_read_u8_array(dt_node, "btmrvl,cal-data",
+						cal_data + BT_CAL_HDR_LEN,
+						BT_CAL_DATA_SIZE);
+		if (ret)
+			return ret;
+
+		BT_DBG("Use cal data from device tree");
+		ret = btmrvl_download_cal_data(priv, cal_data,
+					       BT_CAL_DATA_SIZE);
+		if (ret) {
+			BT_ERR("Fail to download calibrate data");
+			return ret;
+		}
 	}
 
 	return 0;
@@ -526,14 +541,15 @@ static int btmrvl_setup(struct hci_dev *hdev)
 
 	btmrvl_send_module_cfg_cmd(priv, MODULE_BRINGUP_REQ);
 
-	btmrvl_cal_data_dt(priv);
+	priv->btmrvl_dev.gpio_gap = 0xffff;
+
+	btmrvl_check_device_tree(priv);
 
 	btmrvl_pscan_window_reporting(priv, 0x01);
 
 	priv->btmrvl_dev.psmode = 1;
 	btmrvl_enable_ps(priv);
 
-	priv->btmrvl_dev.gpio_gap = 0xffff;
 	btmrvl_send_hscfg_cmd(priv);
 
 	return 0;
diff --git a/drivers/bluetooth/btmrvl_sdio.c b/drivers/bluetooth/btmrvl_sdio.c
index 3e683b153259..0057c0b7a776 100644
--- a/drivers/bluetooth/btmrvl_sdio.c
+++ b/drivers/bluetooth/btmrvl_sdio.c
@@ -24,6 +24,7 @@
 #include <linux/mmc/sdio_ids.h>
 #include <linux/mmc/sdio_func.h>
 #include <linux/module.h>
+#include <linux/devcoredump.h>
 
 #include <net/bluetooth/bluetooth.h>
 #include <net/bluetooth/hci_core.h>
@@ -33,6 +34,24 @@
 
 #define VERSION "1.0"
 
+static struct memory_type_mapping mem_type_mapping_tbl[] = {
+	{"ITCM", NULL, 0, 0xF0},
+	{"DTCM", NULL, 0, 0xF1},
+	{"SQRAM", NULL, 0, 0xF2},
+	{"APU", NULL, 0, 0xF3},
+	{"CIU", NULL, 0, 0xF4},
+	{"ICU", NULL, 0, 0xF5},
+	{"MAC", NULL, 0, 0xF6},
+	{"EXT7", NULL, 0, 0xF7},
+	{"EXT8", NULL, 0, 0xF8},
+	{"EXT9", NULL, 0, 0xF9},
+	{"EXT10", NULL, 0, 0xFA},
+	{"EXT11", NULL, 0, 0xFB},
+	{"EXT12", NULL, 0, 0xFC},
+	{"EXT13", NULL, 0, 0xFD},
+	{"EXTLAST", NULL, 0, 0xFE},
+};
+
 /* The btmrvl_sdio_remove() callback function is called
  * when user removes this module from kernel space or ejects
  * the card from the slot. The driver handles these 2 cases
@@ -84,7 +103,27 @@ static const struct btmrvl_sdio_card_reg btmrvl_reg_87xx = {
 	.int_read_to_clear = false,
 };
 
-static const struct btmrvl_sdio_card_reg btmrvl_reg_88xx = {
+static const struct btmrvl_sdio_card_reg btmrvl_reg_8887 = {
+	.cfg = 0x00,
+	.host_int_mask = 0x08,
+	.host_intstatus = 0x0C,
+	.card_status = 0x5C,
+	.sq_read_base_addr_a0 = 0x6C,
+	.sq_read_base_addr_a1 = 0x6D,
+	.card_revision = 0xC8,
+	.card_fw_status0 = 0x88,
+	.card_fw_status1 = 0x89,
+	.card_rx_len = 0x8A,
+	.card_rx_unit = 0x8B,
+	.io_port_0 = 0xE4,
+	.io_port_1 = 0xE5,
+	.io_port_2 = 0xE6,
+	.int_read_to_clear = true,
+	.host_int_rsr = 0x04,
+	.card_misc_cfg = 0xD8,
+};
+
+static const struct btmrvl_sdio_card_reg btmrvl_reg_8897 = {
 	.cfg = 0x00,
 	.host_int_mask = 0x02,
 	.host_intstatus = 0x03,
@@ -102,6 +141,9 @@ static const struct btmrvl_sdio_card_reg btmrvl_reg_88xx = {
 	.int_read_to_clear = true,
 	.host_int_rsr = 0x01,
 	.card_misc_cfg = 0xcc,
+	.fw_dump_ctrl = 0xe2,
+	.fw_dump_start = 0xe3,
+	.fw_dump_end = 0xea,
 };
 
 static const struct btmrvl_sdio_device btmrvl_sdio_sd8688 = {
@@ -110,6 +152,7 @@ static const struct btmrvl_sdio_device btmrvl_sdio_sd8688 = {
 	.reg		= &btmrvl_reg_8688,
 	.support_pscan_win_report = false,
 	.sd_blksz_fw_dl	= 64,
+	.supports_fw_dump = false,
 };
 
 static const struct btmrvl_sdio_device btmrvl_sdio_sd8787 = {
@@ -118,6 +161,7 @@ static const struct btmrvl_sdio_device btmrvl_sdio_sd8787 = {
 	.reg		= &btmrvl_reg_87xx,
 	.support_pscan_win_report = false,
 	.sd_blksz_fw_dl	= 256,
+	.supports_fw_dump = false,
 };
 
 static const struct btmrvl_sdio_device btmrvl_sdio_sd8797 = {
@@ -126,14 +170,25 @@ static const struct btmrvl_sdio_device btmrvl_sdio_sd8797 = {
 	.reg		= &btmrvl_reg_87xx,
 	.support_pscan_win_report = false,
 	.sd_blksz_fw_dl	= 256,
+	.supports_fw_dump = false,
+};
+
+static const struct btmrvl_sdio_device btmrvl_sdio_sd8887 = {
+	.helper		= NULL,
+	.firmware	= "mrvl/sd8887_uapsta.bin",
+	.reg		= &btmrvl_reg_8887,
+	.support_pscan_win_report = true,
+	.sd_blksz_fw_dl	= 256,
+	.supports_fw_dump = false,
 };
 
 static const struct btmrvl_sdio_device btmrvl_sdio_sd8897 = {
 	.helper		= NULL,
 	.firmware	= "mrvl/sd8897_uapsta.bin",
-	.reg		= &btmrvl_reg_88xx,
+	.reg		= &btmrvl_reg_8897,
 	.support_pscan_win_report = true,
 	.sd_blksz_fw_dl	= 256,
+	.supports_fw_dump = true,
 };
 
 static const struct sdio_device_id btmrvl_sdio_ids[] = {
@@ -149,6 +204,9 @@ static const struct sdio_device_id btmrvl_sdio_ids[] = {
 	/* Marvell SD8797 Bluetooth device */
 	{ SDIO_DEVICE(SDIO_VENDOR_ID_MARVELL, 0x912A),
 			.driver_data = (unsigned long) &btmrvl_sdio_sd8797 },
+	/* Marvell SD8887 Bluetooth device */
+	{ SDIO_DEVICE(SDIO_VENDOR_ID_MARVELL, 0x9136),
+			.driver_data = (unsigned long)&btmrvl_sdio_sd8887 },
 	/* Marvell SD8897 Bluetooth device */
 	{ SDIO_DEVICE(SDIO_VENDOR_ID_MARVELL, 0x912E),
 			.driver_data = (unsigned long) &btmrvl_sdio_sd8897 },
@@ -733,8 +791,8 @@ static void btmrvl_sdio_interrupt(struct sdio_func *func)
 
 	card = sdio_get_drvdata(func);
 	if (!card || !card->priv) {
-		BT_ERR("sbi_interrupt(%p) card or priv is "
-				"NULL, card=%p\n", func, card);
+		BT_ERR("sbi_interrupt(%p) card or priv is NULL, card=%p",
+		       func, card);
 		return;
 	}
 
@@ -1049,6 +1107,277 @@ static int btmrvl_sdio_wakeup_fw(struct btmrvl_private *priv)
 	return ret;
 }
 
+static void btmrvl_sdio_dump_regs(struct btmrvl_private *priv)
+{
+	struct btmrvl_sdio_card *card = priv->btmrvl_dev.card;
+	int ret = 0;
+	unsigned int reg, reg_start, reg_end;
+	char buf[256], *ptr;
+	u8 loop, func, data;
+	int MAX_LOOP = 2;
+
+	btmrvl_sdio_wakeup_fw(priv);
+	sdio_claim_host(card->func);
+
+	for (loop = 0; loop < MAX_LOOP; loop++) {
+		memset(buf, 0, sizeof(buf));
+		ptr = buf;
+
+		if (loop == 0) {
+			/* Read the registers of SDIO function0 */
+			func = loop;
+			reg_start = 0;
+			reg_end = 9;
+		} else {
+			func = 2;
+			reg_start = 0;
+			reg_end = 0x09;
+		}
+
+		ptr += sprintf(ptr, "SDIO Func%d (%#x-%#x): ",
+			       func, reg_start, reg_end);
+		for (reg = reg_start; reg <= reg_end; reg++) {
+			if (func == 0)
+				data = sdio_f0_readb(card->func, reg, &ret);
+			else
+				data = sdio_readb(card->func, reg, &ret);
+
+			if (!ret) {
+				ptr += sprintf(ptr, "%02x ", data);
+			} else {
+				ptr += sprintf(ptr, "ERR");
+				break;
+			}
+		}
+
+		BT_INFO("%s", buf);
+	}
+
+	sdio_release_host(card->func);
+}
+
+/* This function read/write firmware */
+static enum
+rdwr_status btmrvl_sdio_rdwr_firmware(struct btmrvl_private *priv,
+				      u8 doneflag)
+{
+	struct btmrvl_sdio_card *card = priv->btmrvl_dev.card;
+	int ret, tries;
+	u8 ctrl_data = 0;
+
+	sdio_writeb(card->func, FW_DUMP_HOST_READY, card->reg->fw_dump_ctrl,
+		    &ret);
+
+	if (ret) {
+		BT_ERR("SDIO write err");
+		return RDWR_STATUS_FAILURE;
+	}
+
+	for (tries = 0; tries < MAX_POLL_TRIES; tries++) {
+		ctrl_data = sdio_readb(card->func, card->reg->fw_dump_ctrl,
+				       &ret);
+
+		if (ret) {
+			BT_ERR("SDIO read err");
+			return RDWR_STATUS_FAILURE;
+		}
+
+		if (ctrl_data == FW_DUMP_DONE)
+			break;
+		if (doneflag && ctrl_data == doneflag)
+			return RDWR_STATUS_DONE;
+		if (ctrl_data != FW_DUMP_HOST_READY) {
+			BT_INFO("The ctrl reg was changed, re-try again!");
+			sdio_writeb(card->func, FW_DUMP_HOST_READY,
+				    card->reg->fw_dump_ctrl, &ret);
+			if (ret) {
+				BT_ERR("SDIO write err");
+				return RDWR_STATUS_FAILURE;
+			}
+		}
+		usleep_range(100, 200);
+	}
+
+	if (ctrl_data == FW_DUMP_HOST_READY) {
+		BT_ERR("Fail to pull ctrl_data");
+		return RDWR_STATUS_FAILURE;
+	}
+
+	return RDWR_STATUS_SUCCESS;
+}
+
+/* This function dump sdio register and memory data */
+static void btmrvl_sdio_dump_firmware(struct btmrvl_private *priv)
+{
+	struct btmrvl_sdio_card *card = priv->btmrvl_dev.card;
+	int ret = 0;
+	unsigned int reg, reg_start, reg_end;
+	enum rdwr_status stat;
+	u8 *dbg_ptr, *end_ptr, *fw_dump_data, *fw_dump_ptr;
+	u8 dump_num, idx, i, read_reg, doneflag = 0;
+	u32 memory_size, fw_dump_len = 0;
+
+	/* dump sdio register first */
+	btmrvl_sdio_dump_regs(priv);
+
+	if (!card->supports_fw_dump) {
+		BT_ERR("Firmware dump not supported for this card!");
+		return;
+	}
+
+	for (idx = 0; idx < ARRAY_SIZE(mem_type_mapping_tbl); idx++) {
+		struct memory_type_mapping *entry = &mem_type_mapping_tbl[idx];
+
+		if (entry->mem_ptr) {
+			vfree(entry->mem_ptr);
+			entry->mem_ptr = NULL;
+		}
+		entry->mem_size = 0;
+	}
+
+	btmrvl_sdio_wakeup_fw(priv);
+	sdio_claim_host(card->func);
+
+	BT_INFO("== btmrvl firmware dump start ==");
+
+	stat = btmrvl_sdio_rdwr_firmware(priv, doneflag);
+	if (stat == RDWR_STATUS_FAILURE)
+		goto done;
+
+	reg = card->reg->fw_dump_start;
+	/* Read the number of the memories which will dump */
+	dump_num = sdio_readb(card->func, reg, &ret);
+
+	if (ret) {
+		BT_ERR("SDIO read memory length err");
+		goto done;
+	}
+
+	/* Read the length of every memory which will dump */
+	for (idx = 0; idx < dump_num; idx++) {
+		struct memory_type_mapping *entry = &mem_type_mapping_tbl[idx];
+
+		stat = btmrvl_sdio_rdwr_firmware(priv, doneflag);
+		if (stat == RDWR_STATUS_FAILURE)
+			goto done;
+
+		memory_size = 0;
+		reg = card->reg->fw_dump_start;
+		for (i = 0; i < 4; i++) {
+			read_reg = sdio_readb(card->func, reg, &ret);
+			if (ret) {
+				BT_ERR("SDIO read err");
+				goto done;
+			}
+			memory_size |= (read_reg << i*8);
+			reg++;
+		}
+
+		if (memory_size == 0) {
+			BT_INFO("Firmware dump finished!");
+			break;
+		}
+
+		BT_INFO("%s_SIZE=0x%x", entry->mem_name, memory_size);
+		entry->mem_ptr = vzalloc(memory_size + 1);
+		entry->mem_size = memory_size;
+		if (!entry->mem_ptr) {
+			BT_ERR("Vzalloc %s failed", entry->mem_name);
+			goto done;
+		}
+
+		fw_dump_len += (strlen("========Start dump ") +
+				strlen(entry->mem_name) +
+				strlen("========\n") +
+				(memory_size + 1) +
+				strlen("\n========End dump========\n"));
+
+		dbg_ptr = entry->mem_ptr;
+		end_ptr = dbg_ptr + memory_size;
+
+		doneflag = entry->done_flag;
+		BT_INFO("Start %s output, please wait...",
+			entry->mem_name);
+
+		do {
+			stat = btmrvl_sdio_rdwr_firmware(priv, doneflag);
+			if (stat == RDWR_STATUS_FAILURE)
+				goto done;
+
+			reg_start = card->reg->fw_dump_start;
+			reg_end = card->reg->fw_dump_end;
+			for (reg = reg_start; reg <= reg_end; reg++) {
+				*dbg_ptr = sdio_readb(card->func, reg, &ret);
+				if (ret) {
+					BT_ERR("SDIO read err");
+					goto done;
+				}
+				if (dbg_ptr < end_ptr)
+					dbg_ptr++;
+				else
+					BT_ERR("Allocated buffer not enough");
+			}
+
+			if (stat != RDWR_STATUS_DONE) {
+				continue;
+			} else {
+				BT_INFO("%s done: size=0x%tx",
+					entry->mem_name,
+					dbg_ptr - entry->mem_ptr);
+				break;
+			}
+		} while (1);
+	}
+
+	BT_INFO("== btmrvl firmware dump end ==");
+
+done:
+	sdio_release_host(card->func);
+
+	if (fw_dump_len == 0)
+		return;
+
+	fw_dump_data = vzalloc(fw_dump_len+1);
+	if (!fw_dump_data) {
+		BT_ERR("Vzalloc fw_dump_data fail!");
+		return;
+	}
+	fw_dump_ptr = fw_dump_data;
+
+	/* Dump all the memory data into single file, a userspace script will
+	   be used to split all the memory data to multiple files*/
+	BT_INFO("== btmrvl firmware dump to /sys/class/devcoredump start");
+	for (idx = 0; idx < dump_num; idx++) {
+		struct memory_type_mapping *entry = &mem_type_mapping_tbl[idx];
+
+		if (entry->mem_ptr) {
+			strcpy(fw_dump_ptr, "========Start dump ");
+			fw_dump_ptr += strlen("========Start dump ");
+
+			strcpy(fw_dump_ptr, entry->mem_name);
+			fw_dump_ptr += strlen(entry->mem_name);
+
+			strcpy(fw_dump_ptr, "========\n");
+			fw_dump_ptr += strlen("========\n");
+
+			memcpy(fw_dump_ptr, entry->mem_ptr, entry->mem_size);
+			fw_dump_ptr += entry->mem_size;
+
+			strcpy(fw_dump_ptr, "\n========End dump========\n");
+			fw_dump_ptr += strlen("\n========End dump========\n");
+
+			vfree(mem_type_mapping_tbl[idx].mem_ptr);
+			mem_type_mapping_tbl[idx].mem_ptr = NULL;
+		}
+	}
+
+	/* fw_dump_data will be free in device coredump release function
+	   after 5 min*/
+	dev_coredumpv(&priv->btmrvl_dev.hcidev->dev, fw_dump_data,
+		      fw_dump_len, GFP_KERNEL);
+	BT_INFO("== btmrvl firmware dump to /sys/class/devcoredump end");
+}
+
 static int btmrvl_sdio_probe(struct sdio_func *func,
 					const struct sdio_device_id *id)
 {
@@ -1072,6 +1401,7 @@ static int btmrvl_sdio_probe(struct sdio_func *func,
 		card->reg = data->reg;
 		card->sd_blksz_fw_dl = data->sd_blksz_fw_dl;
 		card->support_pscan_win_report = data->support_pscan_win_report;
+		card->supports_fw_dump = data->supports_fw_dump;
 	}
 
 	if (btmrvl_sdio_register_dev(card) < 0) {
@@ -1103,6 +1433,7 @@ static int btmrvl_sdio_probe(struct sdio_func *func,
 	priv->hw_host_to_card = btmrvl_sdio_host_to_card;
 	priv->hw_wakeup_firmware = btmrvl_sdio_wakeup_fw;
 	priv->hw_process_int_status = btmrvl_sdio_process_int_status;
+	priv->firmware_dump = btmrvl_sdio_dump_firmware;
 
 	if (btmrvl_register_hdev(priv)) {
 		BT_ERR("Register hdev failed!");
@@ -1280,4 +1611,5 @@ MODULE_FIRMWARE("mrvl/sd8688_helper.bin");
 MODULE_FIRMWARE("mrvl/sd8688.bin");
 MODULE_FIRMWARE("mrvl/sd8787_uapsta.bin");
 MODULE_FIRMWARE("mrvl/sd8797_uapsta.bin");
+MODULE_FIRMWARE("mrvl/sd8887_uapsta.bin");
 MODULE_FIRMWARE("mrvl/sd8897_uapsta.bin");
diff --git a/drivers/bluetooth/btmrvl_sdio.h b/drivers/bluetooth/btmrvl_sdio.h
index 453559f98a75..1a3bd064c442 100644
--- a/drivers/bluetooth/btmrvl_sdio.h
+++ b/drivers/bluetooth/btmrvl_sdio.h
@@ -81,6 +81,9 @@ struct btmrvl_sdio_card_reg {
 	bool int_read_to_clear;
 	u8 host_int_rsr;
 	u8 card_misc_cfg;
+	u8 fw_dump_ctrl;
+	u8 fw_dump_start;
+	u8 fw_dump_end;
 };
 
 struct btmrvl_sdio_card {
@@ -90,6 +93,7 @@ struct btmrvl_sdio_card {
 	const char *firmware;
 	const struct btmrvl_sdio_card_reg *reg;
 	bool support_pscan_win_report;
+	bool supports_fw_dump;
 	u16 sd_blksz_fw_dl;
 	u8 rx_unit;
 	struct btmrvl_private *priv;
@@ -101,6 +105,7 @@ struct btmrvl_sdio_device {
 	const struct btmrvl_sdio_card_reg *reg;
 	const bool support_pscan_win_report;
 	u16 sd_blksz_fw_dl;
+	bool supports_fw_dump;
 };
 
 
diff --git a/drivers/bluetooth/btuart_cs.c b/drivers/bluetooth/btuart_cs.c
index fb948f02eda5..abb4d2106db4 100644
--- a/drivers/bluetooth/btuart_cs.c
+++ b/drivers/bluetooth/btuart_cs.c
@@ -62,7 +62,7 @@ MODULE_LICENSE("GPL");
 /* ======================== Local structures ======================== */
 
 
-typedef struct btuart_info_t {
+struct btuart_info {
 	struct pcmcia_device *p_dev;
 
 	struct hci_dev *hdev;
@@ -75,7 +75,7 @@ typedef struct btuart_info_t {
 	unsigned long rx_state;
 	unsigned long rx_count;
 	struct sk_buff *rx_skb;
-} btuart_info_t;
+};
 
 
 static int btuart_config(struct pcmcia_device *link);
@@ -127,7 +127,7 @@ static int btuart_write(unsigned int iobase, int fifo_size, __u8 *buf, int len)
 }
 
 
-static void btuart_write_wakeup(btuart_info_t *info)
+static void btuart_write_wakeup(struct btuart_info *info)
 {
 	if (!info) {
 		BT_ERR("Unknown device");
@@ -172,7 +172,7 @@ static void btuart_write_wakeup(btuart_info_t *info)
 }
 
 
-static void btuart_receive(btuart_info_t *info)
+static void btuart_receive(struct btuart_info *info)
 {
 	unsigned int iobase;
 	int boguscount = 0;
@@ -286,7 +286,7 @@ static void btuart_receive(btuart_info_t *info)
 
 static irqreturn_t btuart_interrupt(int irq, void *dev_inst)
 {
-	btuart_info_t *info = dev_inst;
+	struct btuart_info *info = dev_inst;
 	unsigned int iobase;
 	int boguscount = 0;
 	int iir, lsr;
@@ -340,7 +340,8 @@ static irqreturn_t btuart_interrupt(int irq, void *dev_inst)
 }
 
 
-static void btuart_change_speed(btuart_info_t *info, unsigned int speed)
+static void btuart_change_speed(struct btuart_info *info,
+				unsigned int speed)
 {
 	unsigned long flags;
 	unsigned int iobase;
@@ -397,7 +398,7 @@ static void btuart_change_speed(btuart_info_t *info, unsigned int speed)
 
 static int btuart_hci_flush(struct hci_dev *hdev)
 {
-	btuart_info_t *info = hci_get_drvdata(hdev);
+	struct btuart_info *info = hci_get_drvdata(hdev);
 
 	/* Drop TX queue */
 	skb_queue_purge(&(info->txq));
@@ -427,7 +428,7 @@ static int btuart_hci_close(struct hci_dev *hdev)
 
 static int btuart_hci_send_frame(struct hci_dev *hdev, struct sk_buff *skb)
 {
-	btuart_info_t *info = hci_get_drvdata(hdev);
+	struct btuart_info *info = hci_get_drvdata(hdev);
 
 	switch (bt_cb(skb)->pkt_type) {
 	case HCI_COMMAND_PKT:
@@ -455,7 +456,7 @@ static int btuart_hci_send_frame(struct hci_dev *hdev, struct sk_buff *skb)
 /* ======================== Card services HCI interaction ======================== */
 
 
-static int btuart_open(btuart_info_t *info)
+static int btuart_open(struct btuart_info *info)
 {
 	unsigned long flags;
 	unsigned int iobase = info->p_dev->resource[0]->start;
@@ -521,7 +522,7 @@ static int btuart_open(btuart_info_t *info)
 }
 
 
-static int btuart_close(btuart_info_t *info)
+static int btuart_close(struct btuart_info *info)
 {
 	unsigned long flags;
 	unsigned int iobase = info->p_dev->resource[0]->start;
@@ -550,7 +551,7 @@ static int btuart_close(btuart_info_t *info)
 
 static int btuart_probe(struct pcmcia_device *link)
 {
-	btuart_info_t *info;
+	struct btuart_info *info;
 
 	/* Create new info device */
 	info = devm_kzalloc(&link->dev, sizeof(*info), GFP_KERNEL);
@@ -613,7 +614,7 @@ static int btuart_check_config_notpicky(struct pcmcia_device *p_dev,
 
 static int btuart_config(struct pcmcia_device *link)
 {
-	btuart_info_t *info = link->priv;
+	struct btuart_info *info = link->priv;
 	int i;
 	int try;
 
@@ -654,7 +655,7 @@ failed:
 
 static void btuart_release(struct pcmcia_device *link)
 {
-	btuart_info_t *info = link->priv;
+	struct btuart_info *info = link->priv;
 
 	btuart_close(info);
 
diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c
index 292c38e8aa17..19cf2cf22e87 100644
--- a/drivers/bluetooth/btusb.c
+++ b/drivers/bluetooth/btusb.c
@@ -106,9 +106,12 @@ static const struct usb_device_id btusb_table[] = {
 	{ USB_DEVICE(0x0b05, 0x17b5) },
 	{ USB_DEVICE(0x0b05, 0x17cb) },
 	{ USB_DEVICE(0x413c, 0x8197) },
+	{ USB_DEVICE(0x13d3, 0x3404),
+	  .driver_info = BTUSB_BCM_PATCHRAM },
 
 	/* Foxconn - Hon Hai */
-	{ USB_VENDOR_AND_INTERFACE_INFO(0x0489, 0xff, 0x01, 0x01) },
+	{ USB_VENDOR_AND_INTERFACE_INFO(0x0489, 0xff, 0x01, 0x01),
+	  .driver_info = BTUSB_BCM_PATCHRAM },
 
 	/* Broadcom devices with vendor specific id */
 	{ USB_VENDOR_AND_INTERFACE_INFO(0x0a5c, 0xff, 0x01, 0x01),
@@ -156,6 +159,7 @@ static const struct usb_device_id blacklist_table[] = {
 	{ USB_DEVICE(0x0489, 0xe056), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0489, 0xe057), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0489, 0xe05f), .driver_info = BTUSB_ATH3012 },
+	{ USB_DEVICE(0x0489, 0xe078), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x04c5, 0x1330), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x04ca, 0x3004), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x04ca, 0x3005), .driver_info = BTUSB_ATH3012 },
@@ -163,8 +167,10 @@ static const struct usb_device_id blacklist_table[] = {
 	{ USB_DEVICE(0x04ca, 0x3007), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x04ca, 0x3008), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x04ca, 0x300b), .driver_info = BTUSB_ATH3012 },
+	{ USB_DEVICE(0x04ca, 0x3010), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0930, 0x0219), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0930, 0x0220), .driver_info = BTUSB_ATH3012 },
+	{ USB_DEVICE(0x0930, 0x0227), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0b05, 0x17d0), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0cf3, 0x0036), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0cf3, 0x3004), .driver_info = BTUSB_ATH3012 },
@@ -181,6 +187,7 @@ static const struct usb_device_id blacklist_table[] = {
 	{ USB_DEVICE(0x13d3, 0x3375), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x13d3, 0x3393), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x13d3, 0x3402), .driver_info = BTUSB_ATH3012 },
+	{ USB_DEVICE(0x13d3, 0x3408), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x13d3, 0x3432), .driver_info = BTUSB_ATH3012 },
 
 	/* Atheros AR5BBU12 with sflash firmware */
@@ -267,20 +274,24 @@ struct btusb_data {
 	struct usb_interface *intf;
 	struct usb_interface *isoc;
 
-	spinlock_t lock;
-
 	unsigned long flags;
 
 	struct work_struct work;
 	struct work_struct waker;
 
+	struct usb_anchor deferred;
 	struct usb_anchor tx_anchor;
+	int tx_in_flight;
+	spinlock_t txlock;
+
 	struct usb_anchor intr_anchor;
 	struct usb_anchor bulk_anchor;
 	struct usb_anchor isoc_anchor;
-	struct usb_anchor deferred;
-	int tx_in_flight;
-	spinlock_t txlock;
+	spinlock_t rxlock;
+
+	struct sk_buff *evt_skb;
+	struct sk_buff *acl_skb;
+	struct sk_buff *sco_skb;
 
 	struct usb_endpoint_descriptor *intr_ep;
 	struct usb_endpoint_descriptor *bulk_tx_ep;
@@ -293,20 +304,193 @@ struct btusb_data {
 	unsigned int sco_num;
 	int isoc_altsetting;
 	int suspend_count;
+
+	int (*recv_bulk)(struct btusb_data *data, void *buffer, int count);
 };
 
-static int inc_tx(struct btusb_data *data)
+static inline void btusb_free_frags(struct btusb_data *data)
 {
 	unsigned long flags;
-	int rv;
 
-	spin_lock_irqsave(&data->txlock, flags);
-	rv = test_bit(BTUSB_SUSPENDING, &data->flags);
-	if (!rv)
-		data->tx_in_flight++;
-	spin_unlock_irqrestore(&data->txlock, flags);
+	spin_lock_irqsave(&data->rxlock, flags);
+
+	kfree_skb(data->evt_skb);
+	data->evt_skb = NULL;
 
-	return rv;
+	kfree_skb(data->acl_skb);
+	data->acl_skb = NULL;
+
+	kfree_skb(data->sco_skb);
+	data->sco_skb = NULL;
+
+	spin_unlock_irqrestore(&data->rxlock, flags);
+}
+
+static int btusb_recv_intr(struct btusb_data *data, void *buffer, int count)
+{
+	struct sk_buff *skb;
+	int err = 0;
+
+	spin_lock(&data->rxlock);
+	skb = data->evt_skb;
+
+	while (count) {
+		int len;
+
+		if (!skb) {
+			skb = bt_skb_alloc(HCI_MAX_EVENT_SIZE, GFP_ATOMIC);
+			if (!skb) {
+				err = -ENOMEM;
+				break;
+			}
+
+			bt_cb(skb)->pkt_type = HCI_EVENT_PKT;
+			bt_cb(skb)->expect = HCI_EVENT_HDR_SIZE;
+		}
+
+		len = min_t(uint, bt_cb(skb)->expect, count);
+		memcpy(skb_put(skb, len), buffer, len);
+
+		count -= len;
+		buffer += len;
+		bt_cb(skb)->expect -= len;
+
+		if (skb->len == HCI_EVENT_HDR_SIZE) {
+			/* Complete event header */
+			bt_cb(skb)->expect = hci_event_hdr(skb)->plen;
+
+			if (skb_tailroom(skb) < bt_cb(skb)->expect) {
+				kfree_skb(skb);
+				skb = NULL;
+
+				err = -EILSEQ;
+				break;
+			}
+		}
+
+		if (bt_cb(skb)->expect == 0) {
+			/* Complete frame */
+			hci_recv_frame(data->hdev, skb);
+			skb = NULL;
+		}
+	}
+
+	data->evt_skb = skb;
+	spin_unlock(&data->rxlock);
+
+	return err;
+}
+
+static int btusb_recv_bulk(struct btusb_data *data, void *buffer, int count)
+{
+	struct sk_buff *skb;
+	int err = 0;
+
+	spin_lock(&data->rxlock);
+	skb = data->acl_skb;
+
+	while (count) {
+		int len;
+
+		if (!skb) {
+			skb = bt_skb_alloc(HCI_MAX_FRAME_SIZE, GFP_ATOMIC);
+			if (!skb) {
+				err = -ENOMEM;
+				break;
+			}
+
+			bt_cb(skb)->pkt_type = HCI_ACLDATA_PKT;
+			bt_cb(skb)->expect = HCI_ACL_HDR_SIZE;
+		}
+
+		len = min_t(uint, bt_cb(skb)->expect, count);
+		memcpy(skb_put(skb, len), buffer, len);
+
+		count -= len;
+		buffer += len;
+		bt_cb(skb)->expect -= len;
+
+		if (skb->len == HCI_ACL_HDR_SIZE) {
+			__le16 dlen = hci_acl_hdr(skb)->dlen;
+
+			/* Complete ACL header */
+			bt_cb(skb)->expect = __le16_to_cpu(dlen);
+
+			if (skb_tailroom(skb) < bt_cb(skb)->expect) {
+				kfree_skb(skb);
+				skb = NULL;
+
+				err = -EILSEQ;
+				break;
+			}
+		}
+
+		if (bt_cb(skb)->expect == 0) {
+			/* Complete frame */
+			hci_recv_frame(data->hdev, skb);
+			skb = NULL;
+		}
+	}
+
+	data->acl_skb = skb;
+	spin_unlock(&data->rxlock);
+
+	return err;
+}
+
+static int btusb_recv_isoc(struct btusb_data *data, void *buffer, int count)
+{
+	struct sk_buff *skb;
+	int err = 0;
+
+	spin_lock(&data->rxlock);
+	skb = data->sco_skb;
+
+	while (count) {
+		int len;
+
+		if (!skb) {
+			skb = bt_skb_alloc(HCI_MAX_SCO_SIZE, GFP_ATOMIC);
+			if (!skb) {
+				err = -ENOMEM;
+				break;
+			}
+
+			bt_cb(skb)->pkt_type = HCI_SCODATA_PKT;
+			bt_cb(skb)->expect = HCI_SCO_HDR_SIZE;
+		}
+
+		len = min_t(uint, bt_cb(skb)->expect, count);
+		memcpy(skb_put(skb, len), buffer, len);
+
+		count -= len;
+		buffer += len;
+		bt_cb(skb)->expect -= len;
+
+		if (skb->len == HCI_SCO_HDR_SIZE) {
+			/* Complete SCO header */
+			bt_cb(skb)->expect = hci_sco_hdr(skb)->dlen;
+
+			if (skb_tailroom(skb) < bt_cb(skb)->expect) {
+				kfree_skb(skb);
+				skb = NULL;
+
+				err = -EILSEQ;
+				break;
+			}
+		}
+
+		if (bt_cb(skb)->expect == 0) {
+			/* Complete frame */
+			hci_recv_frame(data->hdev, skb);
+			skb = NULL;
+		}
+	}
+
+	data->sco_skb = skb;
+	spin_unlock(&data->rxlock);
+
+	return err;
 }
 
 static void btusb_intr_complete(struct urb *urb)
@@ -315,8 +499,8 @@ static void btusb_intr_complete(struct urb *urb)
 	struct btusb_data *data = hci_get_drvdata(hdev);
 	int err;
 
-	BT_DBG("%s urb %p status %d count %d", hdev->name,
-					urb, urb->status, urb->actual_length);
+	BT_DBG("%s urb %p status %d count %d", hdev->name, urb, urb->status,
+	       urb->actual_length);
 
 	if (!test_bit(HCI_RUNNING, &hdev->flags))
 		return;
@@ -324,12 +508,14 @@ static void btusb_intr_complete(struct urb *urb)
 	if (urb->status == 0) {
 		hdev->stat.byte_rx += urb->actual_length;
 
-		if (hci_recv_fragment(hdev, HCI_EVENT_PKT,
-						urb->transfer_buffer,
-						urb->actual_length) < 0) {
+		if (btusb_recv_intr(data, urb->transfer_buffer,
+				    urb->actual_length) < 0) {
 			BT_ERR("%s corrupted event packet", hdev->name);
 			hdev->stat.err_rx++;
 		}
+	} else if (urb->status == -ENOENT) {
+		/* Avoid suspend failed when usb_kill_urb */
+		return;
 	}
 
 	if (!test_bit(BTUSB_INTR_RUNNING, &data->flags))
@@ -344,7 +530,7 @@ static void btusb_intr_complete(struct urb *urb)
 		 * -ENODEV: device got disconnected */
 		if (err != -EPERM && err != -ENODEV)
 			BT_ERR("%s urb %p failed to resubmit (%d)",
-						hdev->name, urb, -err);
+			       hdev->name, urb, -err);
 		usb_unanchor_urb(urb);
 	}
 }
@@ -377,8 +563,7 @@ static int btusb_submit_intr_urb(struct hci_dev *hdev, gfp_t mem_flags)
 	pipe = usb_rcvintpipe(data->udev, data->intr_ep->bEndpointAddress);
 
 	usb_fill_int_urb(urb, data->udev, pipe, buf, size,
-						btusb_intr_complete, hdev,
-						data->intr_ep->bInterval);
+			 btusb_intr_complete, hdev, data->intr_ep->bInterval);
 
 	urb->transfer_flags |= URB_FREE_BUFFER;
 
@@ -388,7 +573,7 @@ static int btusb_submit_intr_urb(struct hci_dev *hdev, gfp_t mem_flags)
 	if (err < 0) {
 		if (err != -EPERM && err != -ENODEV)
 			BT_ERR("%s urb %p submission failed (%d)",
-						hdev->name, urb, -err);
+			       hdev->name, urb, -err);
 		usb_unanchor_urb(urb);
 	}
 
@@ -403,8 +588,8 @@ static void btusb_bulk_complete(struct urb *urb)
 	struct btusb_data *data = hci_get_drvdata(hdev);
 	int err;
 
-	BT_DBG("%s urb %p status %d count %d", hdev->name,
-					urb, urb->status, urb->actual_length);
+	BT_DBG("%s urb %p status %d count %d", hdev->name, urb, urb->status,
+	       urb->actual_length);
 
 	if (!test_bit(HCI_RUNNING, &hdev->flags))
 		return;
@@ -412,12 +597,14 @@ static void btusb_bulk_complete(struct urb *urb)
 	if (urb->status == 0) {
 		hdev->stat.byte_rx += urb->actual_length;
 
-		if (hci_recv_fragment(hdev, HCI_ACLDATA_PKT,
-						urb->transfer_buffer,
-						urb->actual_length) < 0) {
+		if (data->recv_bulk(data, urb->transfer_buffer,
+				    urb->actual_length) < 0) {
 			BT_ERR("%s corrupted ACL packet", hdev->name);
 			hdev->stat.err_rx++;
 		}
+	} else if (urb->status == -ENOENT) {
+		/* Avoid suspend failed when usb_kill_urb */
+		return;
 	}
 
 	if (!test_bit(BTUSB_BULK_RUNNING, &data->flags))
@@ -432,7 +619,7 @@ static void btusb_bulk_complete(struct urb *urb)
 		 * -ENODEV: device got disconnected */
 		if (err != -EPERM && err != -ENODEV)
 			BT_ERR("%s urb %p failed to resubmit (%d)",
-						hdev->name, urb, -err);
+			       hdev->name, urb, -err);
 		usb_unanchor_urb(urb);
 	}
 }
@@ -462,8 +649,8 @@ static int btusb_submit_bulk_urb(struct hci_dev *hdev, gfp_t mem_flags)
 
 	pipe = usb_rcvbulkpipe(data->udev, data->bulk_rx_ep->bEndpointAddress);
 
-	usb_fill_bulk_urb(urb, data->udev, pipe,
-					buf, size, btusb_bulk_complete, hdev);
+	usb_fill_bulk_urb(urb, data->udev, pipe, buf, size,
+			  btusb_bulk_complete, hdev);
 
 	urb->transfer_flags |= URB_FREE_BUFFER;
 
@@ -474,7 +661,7 @@ static int btusb_submit_bulk_urb(struct hci_dev *hdev, gfp_t mem_flags)
 	if (err < 0) {
 		if (err != -EPERM && err != -ENODEV)
 			BT_ERR("%s urb %p submission failed (%d)",
-						hdev->name, urb, -err);
+			       hdev->name, urb, -err);
 		usb_unanchor_urb(urb);
 	}
 
@@ -489,8 +676,8 @@ static void btusb_isoc_complete(struct urb *urb)
 	struct btusb_data *data = hci_get_drvdata(hdev);
 	int i, err;
 
-	BT_DBG("%s urb %p status %d count %d", hdev->name,
-					urb, urb->status, urb->actual_length);
+	BT_DBG("%s urb %p status %d count %d", hdev->name, urb, urb->status,
+	       urb->actual_length);
 
 	if (!test_bit(HCI_RUNNING, &hdev->flags))
 		return;
@@ -505,13 +692,15 @@ static void btusb_isoc_complete(struct urb *urb)
 
 			hdev->stat.byte_rx += length;
 
-			if (hci_recv_fragment(hdev, HCI_SCODATA_PKT,
-						urb->transfer_buffer + offset,
-								length) < 0) {
+			if (btusb_recv_isoc(data, urb->transfer_buffer + offset,
+					    length) < 0) {
 				BT_ERR("%s corrupted SCO packet", hdev->name);
 				hdev->stat.err_rx++;
 			}
 		}
+	} else if (urb->status == -ENOENT) {
+		/* Avoid suspend failed when usb_kill_urb */
+		return;
 	}
 
 	if (!test_bit(BTUSB_ISOC_RUNNING, &data->flags))
@@ -525,7 +714,7 @@ static void btusb_isoc_complete(struct urb *urb)
 		 * -ENODEV: device got disconnected */
 		if (err != -EPERM && err != -ENODEV)
 			BT_ERR("%s urb %p failed to resubmit (%d)",
-						hdev->name, urb, -err);
+			       hdev->name, urb, -err);
 		usb_unanchor_urb(urb);
 	}
 }
@@ -580,12 +769,12 @@ static int btusb_submit_isoc_urb(struct hci_dev *hdev, gfp_t mem_flags)
 	pipe = usb_rcvisocpipe(data->udev, data->isoc_rx_ep->bEndpointAddress);
 
 	usb_fill_int_urb(urb, data->udev, pipe, buf, size, btusb_isoc_complete,
-				hdev, data->isoc_rx_ep->bInterval);
+			 hdev, data->isoc_rx_ep->bInterval);
 
-	urb->transfer_flags  = URB_FREE_BUFFER | URB_ISO_ASAP;
+	urb->transfer_flags = URB_FREE_BUFFER | URB_ISO_ASAP;
 
 	__fill_isoc_descriptor(urb, size,
-			le16_to_cpu(data->isoc_rx_ep->wMaxPacketSize));
+			       le16_to_cpu(data->isoc_rx_ep->wMaxPacketSize));
 
 	usb_anchor_urb(urb, &data->isoc_anchor);
 
@@ -593,7 +782,7 @@ static int btusb_submit_isoc_urb(struct hci_dev *hdev, gfp_t mem_flags)
 	if (err < 0) {
 		if (err != -EPERM && err != -ENODEV)
 			BT_ERR("%s urb %p submission failed (%d)",
-						hdev->name, urb, -err);
+			       hdev->name, urb, -err);
 		usb_unanchor_urb(urb);
 	}
 
@@ -605,11 +794,11 @@ static int btusb_submit_isoc_urb(struct hci_dev *hdev, gfp_t mem_flags)
 static void btusb_tx_complete(struct urb *urb)
 {
 	struct sk_buff *skb = urb->context;
-	struct hci_dev *hdev = (struct hci_dev *) skb->dev;
+	struct hci_dev *hdev = (struct hci_dev *)skb->dev;
 	struct btusb_data *data = hci_get_drvdata(hdev);
 
-	BT_DBG("%s urb %p status %d count %d", hdev->name,
-					urb, urb->status, urb->actual_length);
+	BT_DBG("%s urb %p status %d count %d", hdev->name, urb, urb->status,
+	       urb->actual_length);
 
 	if (!test_bit(HCI_RUNNING, &hdev->flags))
 		goto done;
@@ -632,10 +821,10 @@ done:
 static void btusb_isoc_tx_complete(struct urb *urb)
 {
 	struct sk_buff *skb = urb->context;
-	struct hci_dev *hdev = (struct hci_dev *) skb->dev;
+	struct hci_dev *hdev = (struct hci_dev *)skb->dev;
 
-	BT_DBG("%s urb %p status %d count %d", hdev->name,
-					urb, urb->status, urb->actual_length);
+	BT_DBG("%s urb %p status %d count %d", hdev->name, urb, urb->status,
+	       urb->actual_length);
 
 	if (!test_bit(HCI_RUNNING, &hdev->flags))
 		goto done;
@@ -719,6 +908,8 @@ static int btusb_close(struct hci_dev *hdev)
 	clear_bit(BTUSB_INTR_RUNNING, &data->flags);
 
 	btusb_stop_traffic(data);
+	btusb_free_frags(data);
+
 	err = usb_autopm_get_interface(data->intf);
 	if (err < 0)
 		goto failed;
@@ -738,122 +929,181 @@ static int btusb_flush(struct hci_dev *hdev)
 	BT_DBG("%s", hdev->name);
 
 	usb_kill_anchored_urbs(&data->tx_anchor);
+	btusb_free_frags(data);
 
 	return 0;
 }
 
-static int btusb_send_frame(struct hci_dev *hdev, struct sk_buff *skb)
+static struct urb *alloc_ctrl_urb(struct hci_dev *hdev, struct sk_buff *skb)
 {
 	struct btusb_data *data = hci_get_drvdata(hdev);
 	struct usb_ctrlrequest *dr;
 	struct urb *urb;
 	unsigned int pipe;
-	int err;
 
-	BT_DBG("%s", hdev->name);
+	urb = usb_alloc_urb(0, GFP_KERNEL);
+	if (!urb)
+		return ERR_PTR(-ENOMEM);
 
-	if (!test_bit(HCI_RUNNING, &hdev->flags))
-		return -EBUSY;
+	dr = kmalloc(sizeof(*dr), GFP_KERNEL);
+	if (!dr) {
+		usb_free_urb(urb);
+		return ERR_PTR(-ENOMEM);
+	}
 
-	skb->dev = (void *) hdev;
+	dr->bRequestType = data->cmdreq_type;
+	dr->bRequest     = 0;
+	dr->wIndex       = 0;
+	dr->wValue       = 0;
+	dr->wLength      = __cpu_to_le16(skb->len);
 
-	switch (bt_cb(skb)->pkt_type) {
-	case HCI_COMMAND_PKT:
-		urb = usb_alloc_urb(0, GFP_ATOMIC);
-		if (!urb)
-			return -ENOMEM;
-
-		dr = kmalloc(sizeof(*dr), GFP_ATOMIC);
-		if (!dr) {
-			usb_free_urb(urb);
-			return -ENOMEM;
-		}
+	pipe = usb_sndctrlpipe(data->udev, 0x00);
 
-		dr->bRequestType = data->cmdreq_type;
-		dr->bRequest     = 0;
-		dr->wIndex       = 0;
-		dr->wValue       = 0;
-		dr->wLength      = __cpu_to_le16(skb->len);
+	usb_fill_control_urb(urb, data->udev, pipe, (void *)dr,
+			     skb->data, skb->len, btusb_tx_complete, skb);
 
-		pipe = usb_sndctrlpipe(data->udev, 0x00);
+	skb->dev = (void *)hdev;
 
-		usb_fill_control_urb(urb, data->udev, pipe, (void *) dr,
-				skb->data, skb->len, btusb_tx_complete, skb);
+	return urb;
+}
 
-		hdev->stat.cmd_tx++;
-		break;
+static struct urb *alloc_bulk_urb(struct hci_dev *hdev, struct sk_buff *skb)
+{
+	struct btusb_data *data = hci_get_drvdata(hdev);
+	struct urb *urb;
+	unsigned int pipe;
 
-	case HCI_ACLDATA_PKT:
-		if (!data->bulk_tx_ep)
-			return -ENODEV;
+	if (!data->bulk_tx_ep)
+		return ERR_PTR(-ENODEV);
 
-		urb = usb_alloc_urb(0, GFP_ATOMIC);
-		if (!urb)
-			return -ENOMEM;
+	urb = usb_alloc_urb(0, GFP_KERNEL);
+	if (!urb)
+		return ERR_PTR(-ENOMEM);
 
-		pipe = usb_sndbulkpipe(data->udev,
-					data->bulk_tx_ep->bEndpointAddress);
+	pipe = usb_sndbulkpipe(data->udev, data->bulk_tx_ep->bEndpointAddress);
 
-		usb_fill_bulk_urb(urb, data->udev, pipe,
-				skb->data, skb->len, btusb_tx_complete, skb);
+	usb_fill_bulk_urb(urb, data->udev, pipe,
+			  skb->data, skb->len, btusb_tx_complete, skb);
 
-		hdev->stat.acl_tx++;
-		break;
+	skb->dev = (void *)hdev;
 
-	case HCI_SCODATA_PKT:
-		if (!data->isoc_tx_ep || hci_conn_num(hdev, SCO_LINK) < 1)
-			return -ENODEV;
+	return urb;
+}
 
-		urb = usb_alloc_urb(BTUSB_MAX_ISOC_FRAMES, GFP_ATOMIC);
-		if (!urb)
-			return -ENOMEM;
+static struct urb *alloc_isoc_urb(struct hci_dev *hdev, struct sk_buff *skb)
+{
+	struct btusb_data *data = hci_get_drvdata(hdev);
+	struct urb *urb;
+	unsigned int pipe;
 
-		pipe = usb_sndisocpipe(data->udev,
-					data->isoc_tx_ep->bEndpointAddress);
+	if (!data->isoc_tx_ep)
+		return ERR_PTR(-ENODEV);
 
-		usb_fill_int_urb(urb, data->udev, pipe,
-				skb->data, skb->len, btusb_isoc_tx_complete,
-				skb, data->isoc_tx_ep->bInterval);
+	urb = usb_alloc_urb(BTUSB_MAX_ISOC_FRAMES, GFP_KERNEL);
+	if (!urb)
+		return ERR_PTR(-ENOMEM);
 
-		urb->transfer_flags  = URB_ISO_ASAP;
+	pipe = usb_sndisocpipe(data->udev, data->isoc_tx_ep->bEndpointAddress);
 
-		__fill_isoc_descriptor(urb, skb->len,
-				le16_to_cpu(data->isoc_tx_ep->wMaxPacketSize));
+	usb_fill_int_urb(urb, data->udev, pipe,
+			 skb->data, skb->len, btusb_isoc_tx_complete,
+			 skb, data->isoc_tx_ep->bInterval);
 
-		hdev->stat.sco_tx++;
-		goto skip_waking;
+	urb->transfer_flags  = URB_ISO_ASAP;
 
-	default:
-		return -EILSEQ;
-	}
+	__fill_isoc_descriptor(urb, skb->len,
+			       le16_to_cpu(data->isoc_tx_ep->wMaxPacketSize));
 
-	err = inc_tx(data);
-	if (err) {
-		usb_anchor_urb(urb, &data->deferred);
-		schedule_work(&data->waker);
-		err = 0;
-		goto done;
-	}
+	skb->dev = (void *)hdev;
+
+	return urb;
+}
+
+static int submit_tx_urb(struct hci_dev *hdev, struct urb *urb)
+{
+	struct btusb_data *data = hci_get_drvdata(hdev);
+	int err;
 
-skip_waking:
 	usb_anchor_urb(urb, &data->tx_anchor);
 
-	err = usb_submit_urb(urb, GFP_ATOMIC);
+	err = usb_submit_urb(urb, GFP_KERNEL);
 	if (err < 0) {
 		if (err != -EPERM && err != -ENODEV)
 			BT_ERR("%s urb %p submission failed (%d)",
-						hdev->name, urb, -err);
+			       hdev->name, urb, -err);
 		kfree(urb->setup_packet);
 		usb_unanchor_urb(urb);
 	} else {
 		usb_mark_last_busy(data->udev);
 	}
 
-done:
 	usb_free_urb(urb);
 	return err;
 }
 
+static int submit_or_queue_tx_urb(struct hci_dev *hdev, struct urb *urb)
+{
+	struct btusb_data *data = hci_get_drvdata(hdev);
+	unsigned long flags;
+	bool suspending;
+
+	spin_lock_irqsave(&data->txlock, flags);
+	suspending = test_bit(BTUSB_SUSPENDING, &data->flags);
+	if (!suspending)
+		data->tx_in_flight++;
+	spin_unlock_irqrestore(&data->txlock, flags);
+
+	if (!suspending)
+		return submit_tx_urb(hdev, urb);
+
+	usb_anchor_urb(urb, &data->deferred);
+	schedule_work(&data->waker);
+
+	usb_free_urb(urb);
+	return 0;
+}
+
+static int btusb_send_frame(struct hci_dev *hdev, struct sk_buff *skb)
+{
+	struct urb *urb;
+
+	BT_DBG("%s", hdev->name);
+
+	if (!test_bit(HCI_RUNNING, &hdev->flags))
+		return -EBUSY;
+
+	switch (bt_cb(skb)->pkt_type) {
+	case HCI_COMMAND_PKT:
+		urb = alloc_ctrl_urb(hdev, skb);
+		if (IS_ERR(urb))
+			return PTR_ERR(urb);
+
+		hdev->stat.cmd_tx++;
+		return submit_or_queue_tx_urb(hdev, urb);
+
+	case HCI_ACLDATA_PKT:
+		urb = alloc_bulk_urb(hdev, skb);
+		if (IS_ERR(urb))
+			return PTR_ERR(urb);
+
+		hdev->stat.acl_tx++;
+		return submit_or_queue_tx_urb(hdev, urb);
+
+	case HCI_SCODATA_PKT:
+		if (hci_conn_num(hdev, SCO_LINK) < 1)
+			return -ENODEV;
+
+		urb = alloc_isoc_urb(hdev, skb);
+		if (IS_ERR(urb))
+			return PTR_ERR(urb);
+
+		hdev->stat.sco_tx++;
+		return submit_tx_urb(hdev, urb);
+	}
+
+	return -EILSEQ;
+}
+
 static void btusb_notify(struct hci_dev *hdev, unsigned int evt)
 {
 	struct btusb_data *data = hci_get_drvdata(hdev);
@@ -930,6 +1180,7 @@ static void btusb_work(struct work_struct *work)
 
 		if (hdev->voice_setting & 0x0020) {
 			static const int alts[3] = { 2, 4, 5 };
+
 			new_alts = alts[data->sco_num - 1];
 		} else {
 			new_alts = data->sco_num;
@@ -1002,7 +1253,7 @@ static int btusb_setup_csr(struct hci_dev *hdev)
 		return -PTR_ERR(skb);
 	}
 
-	rp = (struct hci_rp_read_local_version *) skb->data;
+	rp = (struct hci_rp_read_local_version *)skb->data;
 
 	if (!rp->status) {
 		if (le16_to_cpu(rp->manufacturer) != 10) {
@@ -1040,7 +1291,7 @@ struct intel_version {
 } __packed;
 
 static const struct firmware *btusb_setup_intel_get_fw(struct hci_dev *hdev,
-						struct intel_version *ver)
+						       struct intel_version *ver)
 {
 	const struct firmware *fw;
 	char fwname[64];
@@ -1216,7 +1467,7 @@ static int btusb_check_bdaddr_intel(struct hci_dev *hdev)
 		return -EIO;
 	}
 
-	rp = (struct hci_rp_read_bd_addr *) skb->data;
+	rp = (struct hci_rp_read_bd_addr *)skb->data;
 	if (rp->status) {
 		BT_ERR("%s Intel device address result failed (%02x)",
 		       hdev->name, rp->status);
@@ -1346,6 +1597,7 @@ static int btusb_setup_intel(struct hci_dev *hdev)
 
 	if (skb->data[0]) {
 		u8 evt_status = skb->data[0];
+
 		BT_ERR("%s enable Intel manufacturer mode event failed (%02x)",
 		       hdev->name, evt_status);
 		kfree_skb(skb);
@@ -1455,7 +1707,7 @@ static int btusb_set_bdaddr_intel(struct hci_dev *hdev, const bdaddr_t *bdaddr)
 	if (IS_ERR(skb)) {
 		ret = PTR_ERR(skb);
 		BT_ERR("%s: changing Intel device address failed (%ld)",
-			hdev->name, ret);
+		       hdev->name, ret);
 		return ret;
 	}
 	kfree_skb(skb);
@@ -1530,19 +1782,19 @@ static int btusb_setup_bcm_patchram(struct hci_dev *hdev)
 	if (IS_ERR(skb)) {
 		ret = PTR_ERR(skb);
 		BT_ERR("%s: HCI_OP_READ_LOCAL_VERSION failed (%ld)",
-			hdev->name, ret);
+		       hdev->name, ret);
 		goto done;
 	}
 
 	if (skb->len != sizeof(*ver)) {
 		BT_ERR("%s: HCI_OP_READ_LOCAL_VERSION event length mismatch",
-			hdev->name);
+		       hdev->name);
 		kfree_skb(skb);
 		ret = -EIO;
 		goto done;
 	}
 
-	ver = (struct hci_rp_read_local_version *) skb->data;
+	ver = (struct hci_rp_read_local_version *)skb->data;
 	BT_INFO("%s: BCM: patching hci_ver=%02x hci_rev=%04x lmp_ver=%02x "
 		"lmp_subver=%04x", hdev->name, ver->hci_ver, ver->hci_rev,
 		ver->lmp_ver, ver->lmp_subver);
@@ -1553,7 +1805,7 @@ static int btusb_setup_bcm_patchram(struct hci_dev *hdev)
 	if (IS_ERR(skb)) {
 		ret = PTR_ERR(skb);
 		BT_ERR("%s: BCM: Download Minidrv command failed (%ld)",
-			hdev->name, ret);
+		       hdev->name, ret);
 		goto reset_fw;
 	}
 	kfree_skb(skb);
@@ -1565,13 +1817,13 @@ static int btusb_setup_bcm_patchram(struct hci_dev *hdev)
 	fw_size = fw->size;
 
 	while (fw_size >= sizeof(*cmd)) {
-		cmd = (struct hci_command_hdr *) fw_ptr;
+		cmd = (struct hci_command_hdr *)fw_ptr;
 		fw_ptr += sizeof(*cmd);
 		fw_size -= sizeof(*cmd);
 
 		if (fw_size < cmd->plen) {
 			BT_ERR("%s: BCM: patch %s is corrupted",
-				hdev->name, fw_name);
+			       hdev->name, fw_name);
 			ret = -EINVAL;
 			goto reset_fw;
 		}
@@ -1587,7 +1839,7 @@ static int btusb_setup_bcm_patchram(struct hci_dev *hdev)
 		if (IS_ERR(skb)) {
 			ret = PTR_ERR(skb);
 			BT_ERR("%s: BCM: patch command %04x failed (%ld)",
-				hdev->name, opcode, ret);
+			       hdev->name, opcode, ret);
 			goto reset_fw;
 		}
 		kfree_skb(skb);
@@ -1612,19 +1864,19 @@ reset_fw:
 	if (IS_ERR(skb)) {
 		ret = PTR_ERR(skb);
 		BT_ERR("%s: HCI_OP_READ_LOCAL_VERSION failed (%ld)",
-			hdev->name, ret);
+		       hdev->name, ret);
 		goto done;
 	}
 
 	if (skb->len != sizeof(*ver)) {
 		BT_ERR("%s: HCI_OP_READ_LOCAL_VERSION event length mismatch",
-			hdev->name);
+		       hdev->name);
 		kfree_skb(skb);
 		ret = -EIO;
 		goto done;
 	}
 
-	ver = (struct hci_rp_read_local_version *) skb->data;
+	ver = (struct hci_rp_read_local_version *)skb->data;
 	BT_INFO("%s: BCM: firmware hci_ver=%02x hci_rev=%04x lmp_ver=%02x "
 		"lmp_subver=%04x", hdev->name, ver->hci_ver, ver->hci_rev,
 		ver->lmp_ver, ver->lmp_subver);
@@ -1636,19 +1888,19 @@ reset_fw:
 	if (IS_ERR(skb)) {
 		ret = PTR_ERR(skb);
 		BT_ERR("%s: HCI_OP_READ_BD_ADDR failed (%ld)",
-			hdev->name, ret);
+		       hdev->name, ret);
 		goto done;
 	}
 
 	if (skb->len != sizeof(*bda)) {
 		BT_ERR("%s: HCI_OP_READ_BD_ADDR event length mismatch",
-			hdev->name);
+		       hdev->name);
 		kfree_skb(skb);
 		ret = -EIO;
 		goto done;
 	}
 
-	bda = (struct hci_rp_read_bd_addr *) skb->data;
+	bda = (struct hci_rp_read_bd_addr *)skb->data;
 	if (bda->status) {
 		BT_ERR("%s: HCI_OP_READ_BD_ADDR error status (%02x)",
 		       hdev->name, bda->status);
@@ -1683,7 +1935,7 @@ static int btusb_set_bdaddr_bcm(struct hci_dev *hdev, const bdaddr_t *bdaddr)
 	if (IS_ERR(skb)) {
 		ret = PTR_ERR(skb);
 		BT_ERR("%s: BCM: Change address command failed (%ld)",
-			hdev->name, ret);
+		       hdev->name, ret);
 		return ret;
 	}
 	kfree_skb(skb);
@@ -1692,7 +1944,7 @@ static int btusb_set_bdaddr_bcm(struct hci_dev *hdev, const bdaddr_t *bdaddr)
 }
 
 static int btusb_probe(struct usb_interface *intf,
-				const struct usb_device_id *id)
+		       const struct usb_device_id *id)
 {
 	struct usb_endpoint_descriptor *ep_desc;
 	struct btusb_data *data;
@@ -1707,6 +1959,7 @@ static int btusb_probe(struct usb_interface *intf,
 
 	if (!id->driver_info) {
 		const struct usb_device_id *match;
+
 		match = usb_match_id(intf, blacklist_table);
 		if (match)
 			id = match;
@@ -1755,17 +2008,18 @@ static int btusb_probe(struct usb_interface *intf,
 	data->udev = interface_to_usbdev(intf);
 	data->intf = intf;
 
-	spin_lock_init(&data->lock);
-
 	INIT_WORK(&data->work, btusb_work);
 	INIT_WORK(&data->waker, btusb_waker);
+	init_usb_anchor(&data->deferred);
+	init_usb_anchor(&data->tx_anchor);
 	spin_lock_init(&data->txlock);
 
-	init_usb_anchor(&data->tx_anchor);
 	init_usb_anchor(&data->intr_anchor);
 	init_usb_anchor(&data->bulk_anchor);
 	init_usb_anchor(&data->isoc_anchor);
-	init_usb_anchor(&data->deferred);
+	spin_lock_init(&data->rxlock);
+
+	data->recv_bulk = btusb_recv_bulk;
 
 	hdev = hci_alloc_dev();
 	if (!hdev)
@@ -1790,6 +2044,7 @@ static int btusb_probe(struct usb_interface *intf,
 	if (id->driver_info & BTUSB_BCM_PATCHRAM) {
 		hdev->setup = btusb_setup_bcm_patchram;
 		hdev->set_bdaddr = btusb_set_bdaddr_bcm;
+		set_bit(HCI_QUIRK_STRICT_DUPLICATE_FILTER, &hdev->quirks);
 	}
 
 	if (id->driver_info & BTUSB_INTEL) {
@@ -1857,7 +2112,7 @@ static int btusb_probe(struct usb_interface *intf,
 
 	if (data->isoc) {
 		err = usb_driver_claim_interface(&btusb_driver,
-							data->isoc, data);
+						 data->isoc, data);
 		if (err < 0) {
 			hci_free_dev(hdev);
 			return err;
@@ -1898,6 +2153,7 @@ static void btusb_disconnect(struct usb_interface *intf)
 	else if (data->isoc)
 		usb_driver_release_interface(&btusb_driver, data->isoc);
 
+	btusb_free_frags(data);
 	hci_free_dev(hdev);
 }
 
diff --git a/drivers/bluetooth/btwilink.c b/drivers/bluetooth/btwilink.c
index f038dba19e36..55c135b7757a 100644
--- a/drivers/bluetooth/btwilink.c
+++ b/drivers/bluetooth/btwilink.c
@@ -349,7 +349,6 @@ static struct platform_driver btwilink_driver = {
 	.remove = bt_ti_remove,
 	.driver = {
 		.name = "btwilink",
-		.owner = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/bluetooth/dtl1_cs.c b/drivers/bluetooth/dtl1_cs.c
index 2bd8fad17206..78e10f0c65b2 100644
--- a/drivers/bluetooth/dtl1_cs.c
+++ b/drivers/bluetooth/dtl1_cs.c
@@ -62,7 +62,7 @@ MODULE_LICENSE("GPL");
 /* ======================== Local structures ======================== */
 
 
-typedef struct dtl1_info_t {
+struct dtl1_info {
 	struct pcmcia_device *p_dev;
 
 	struct hci_dev *hdev;
@@ -78,7 +78,7 @@ typedef struct dtl1_info_t {
 	unsigned long rx_state;
 	unsigned long rx_count;
 	struct sk_buff *rx_skb;
-} dtl1_info_t;
+};
 
 
 static int dtl1_config(struct pcmcia_device *link);
@@ -94,11 +94,11 @@ static int dtl1_config(struct pcmcia_device *link);
 #define RECV_WAIT_DATA  1
 
 
-typedef struct {
+struct nsh {
 	u8 type;
 	u8 zero;
 	u16 len;
-} __packed nsh_t;	/* Nokia Specific Header */
+} __packed;	/* Nokia Specific Header */
 
 #define NSHL  4				/* Nokia Specific Header Length */
 
@@ -126,7 +126,7 @@ static int dtl1_write(unsigned int iobase, int fifo_size, __u8 *buf, int len)
 }
 
 
-static void dtl1_write_wakeup(dtl1_info_t *info)
+static void dtl1_write_wakeup(struct dtl1_info *info)
 {
 	if (!info) {
 		BT_ERR("Unknown device");
@@ -176,7 +176,7 @@ static void dtl1_write_wakeup(dtl1_info_t *info)
 }
 
 
-static void dtl1_control(dtl1_info_t *info, struct sk_buff *skb)
+static void dtl1_control(struct dtl1_info *info, struct sk_buff *skb)
 {
 	u8 flowmask = *(u8 *)skb->data;
 	int i;
@@ -199,10 +199,10 @@ static void dtl1_control(dtl1_info_t *info, struct sk_buff *skb)
 }
 
 
-static void dtl1_receive(dtl1_info_t *info)
+static void dtl1_receive(struct dtl1_info *info)
 {
 	unsigned int iobase;
-	nsh_t *nsh;
+	struct nsh *nsh;
 	int boguscount = 0;
 
 	if (!info) {
@@ -227,7 +227,7 @@ static void dtl1_receive(dtl1_info_t *info)
 		}
 
 		*skb_put(info->rx_skb, 1) = inb(iobase + UART_RX);
-		nsh = (nsh_t *)info->rx_skb->data;
+		nsh = (struct nsh *)info->rx_skb->data;
 
 		info->rx_count--;
 
@@ -287,7 +287,7 @@ static void dtl1_receive(dtl1_info_t *info)
 
 static irqreturn_t dtl1_interrupt(int irq, void *dev_inst)
 {
-	dtl1_info_t *info = dev_inst;
+	struct dtl1_info *info = dev_inst;
 	unsigned int iobase;
 	unsigned char msr;
 	int boguscount = 0;
@@ -365,7 +365,7 @@ static int dtl1_hci_open(struct hci_dev *hdev)
 
 static int dtl1_hci_flush(struct hci_dev *hdev)
 {
-	dtl1_info_t *info = hci_get_drvdata(hdev);
+	struct dtl1_info *info = hci_get_drvdata(hdev);
 
 	/* Drop TX queue */
 	skb_queue_purge(&(info->txq));
@@ -387,9 +387,9 @@ static int dtl1_hci_close(struct hci_dev *hdev)
 
 static int dtl1_hci_send_frame(struct hci_dev *hdev, struct sk_buff *skb)
 {
-	dtl1_info_t *info = hci_get_drvdata(hdev);
+	struct dtl1_info *info = hci_get_drvdata(hdev);
 	struct sk_buff *s;
-	nsh_t nsh;
+	struct nsh nsh;
 
 	switch (bt_cb(skb)->pkt_type) {
 	case HCI_COMMAND_PKT:
@@ -436,7 +436,7 @@ static int dtl1_hci_send_frame(struct hci_dev *hdev, struct sk_buff *skb)
 /* ======================== Card services HCI interaction ======================== */
 
 
-static int dtl1_open(dtl1_info_t *info)
+static int dtl1_open(struct dtl1_info *info)
 {
 	unsigned long flags;
 	unsigned int iobase = info->p_dev->resource[0]->start;
@@ -505,7 +505,7 @@ static int dtl1_open(dtl1_info_t *info)
 }
 
 
-static int dtl1_close(dtl1_info_t *info)
+static int dtl1_close(struct dtl1_info *info)
 {
 	unsigned long flags;
 	unsigned int iobase = info->p_dev->resource[0]->start;
@@ -534,7 +534,7 @@ static int dtl1_close(dtl1_info_t *info)
 
 static int dtl1_probe(struct pcmcia_device *link)
 {
-	dtl1_info_t *info;
+	struct dtl1_info *info;
 
 	/* Create new info device */
 	info = devm_kzalloc(&link->dev, sizeof(*info), GFP_KERNEL);
@@ -552,7 +552,7 @@ static int dtl1_probe(struct pcmcia_device *link)
 
 static void dtl1_detach(struct pcmcia_device *link)
 {
-	dtl1_info_t *info = link->priv;
+	struct dtl1_info *info = link->priv;
 
 	dtl1_close(info);
 	pcmcia_disable_device(link);
@@ -571,7 +571,7 @@ static int dtl1_confcheck(struct pcmcia_device *p_dev, void *priv_data)
 
 static int dtl1_config(struct pcmcia_device *link)
 {
-	dtl1_info_t *info = link->priv;
+	struct dtl1_info *info = link->priv;
 	int ret;
 
 	/* Look for a generic full-sized window */
diff --git a/drivers/bluetooth/hci_ath.c b/drivers/bluetooth/hci_ath.c
index 0bc8a6a6a148..2ff6dfd2d3f0 100644
--- a/drivers/bluetooth/hci_ath.c
+++ b/drivers/bluetooth/hci_ath.c
@@ -74,7 +74,7 @@ static int ath_wakeup_ar3k(struct tty_struct *tty)
 
 	status = tty->driver->ops->tiocmget(tty);
 
-	/* Disable Automatic RTSCTS */
+	/* Enable Automatic RTSCTS */
 	ktermios.c_cflag |= CRTSCTS;
 	status = tty_set_termios(tty, &ktermios);
 
diff --git a/drivers/bluetooth/hci_h5.c b/drivers/bluetooth/hci_h5.c
index caacb422995d..ec0fa7732c0d 100644
--- a/drivers/bluetooth/hci_h5.c
+++ b/drivers/bluetooth/hci_h5.c
@@ -168,6 +168,27 @@ wakeup:
 	hci_uart_tx_wakeup(hu);
 }
 
+static void h5_peer_reset(struct hci_uart *hu)
+{
+	struct h5 *h5 = hu->priv;
+
+	BT_ERR("Peer device has reset");
+
+	h5->state = H5_UNINITIALIZED;
+
+	del_timer(&h5->timer);
+
+	skb_queue_purge(&h5->rel);
+	skb_queue_purge(&h5->unrel);
+	skb_queue_purge(&h5->unack);
+
+	h5->tx_seq = 0;
+	h5->tx_ack = 0;
+
+	/* Send reset request to upper stack */
+	hci_reset_dev(hu->hdev);
+}
+
 static int h5_open(struct hci_uart *hu)
 {
 	struct h5 *h5;
@@ -237,7 +258,7 @@ static void h5_pkt_cull(struct h5 *h5)
 			break;
 
 		to_remove--;
-		seq = (seq - 1) % 8;
+		seq = (seq - 1) & 0x07;
 	}
 
 	if (seq != h5->rx_ack)
@@ -283,8 +304,12 @@ static void h5_handle_internal_rx(struct hci_uart *hu)
 	conf_req[2] = h5_cfg_field(h5);
 
 	if (memcmp(data, sync_req, 2) == 0) {
+		if (h5->state == H5_ACTIVE)
+			h5_peer_reset(hu);
 		h5_link_control(hu, sync_rsp, 2);
 	} else if (memcmp(data, sync_rsp, 2) == 0) {
+		if (h5->state == H5_ACTIVE)
+			h5_peer_reset(hu);
 		h5->state = H5_INITIALIZED;
 		h5_link_control(hu, conf_req, 3);
 	} else if (memcmp(data, conf_req, 2) == 0) {
diff --git a/drivers/bluetooth/hci_vhci.c b/drivers/bluetooth/hci_vhci.c
index 5bb5872ffee6..6653473f2757 100644
--- a/drivers/bluetooth/hci_vhci.c
+++ b/drivers/bluetooth/hci_vhci.c
@@ -160,13 +160,11 @@ static int vhci_create_device(struct vhci_data *data, __u8 opcode)
 }
 
 static inline ssize_t vhci_get_user(struct vhci_data *data,
-				    const struct iovec *iov,
-				    unsigned long count)
+				    struct iov_iter *from)
 {
-	size_t len = iov_length(iov, count);
+	size_t len = iov_iter_count(from);
 	struct sk_buff *skb;
 	__u8 pkt_type, opcode;
-	unsigned long i;
 	int ret;
 
 	if (len < 2 || len > HCI_MAX_FRAME_SIZE)
@@ -176,12 +174,9 @@ static inline ssize_t vhci_get_user(struct vhci_data *data,
 	if (!skb)
 		return -ENOMEM;
 
-	for (i = 0; i < count; i++) {
-		if (copy_from_user(skb_put(skb, iov[i].iov_len),
-				   iov[i].iov_base, iov[i].iov_len)) {
-			kfree_skb(skb);
-			return -EFAULT;
-		}
+	if (copy_from_iter(skb_put(skb, len), len, from) != len) {
+		kfree_skb(skb);
+		return -EFAULT;
 	}
 
 	pkt_type = *((__u8 *) skb->data);
@@ -294,13 +289,12 @@ static ssize_t vhci_read(struct file *file,
 	return ret;
 }
 
-static ssize_t vhci_write(struct kiocb *iocb, const struct iovec *iov,
-			  unsigned long count, loff_t pos)
+static ssize_t vhci_write(struct kiocb *iocb, struct iov_iter *from)
 {
 	struct file *file = iocb->ki_filp;
 	struct vhci_data *data = file->private_data;
 
-	return vhci_get_user(data, iov, count);
+	return vhci_get_user(data, from);
 }
 
 static unsigned int vhci_poll(struct file *file, poll_table *wait)
@@ -365,7 +359,7 @@ static int vhci_release(struct inode *inode, struct file *file)
 static const struct file_operations vhci_fops = {
 	.owner		= THIS_MODULE,
 	.read		= vhci_read,
-	.aio_write	= vhci_write,
+	.write_iter	= vhci_write,
 	.poll		= vhci_poll,
 	.open		= vhci_open,
 	.release	= vhci_release,
diff --git a/drivers/bus/Kconfig b/drivers/bus/Kconfig
index 603eb1be4f6a..b99729e36860 100644
--- a/drivers/bus/Kconfig
+++ b/drivers/bus/Kconfig
@@ -6,7 +6,7 @@ menu "Bus devices"
 
 config BRCMSTB_GISB_ARB
 	bool "Broadcom STB GISB bus arbiter"
-	depends on ARM
+	depends on ARM || MIPS
 	help
 	  Driver for the Broadcom Set Top Box System-on-a-chip internal bus
 	  arbiter. This driver provides timeout and target abort error handling
diff --git a/drivers/bus/arm-cci.c b/drivers/bus/arm-cci.c
index 7af78df241f2..0ce5e2d65a06 100644
--- a/drivers/bus/arm-cci.c
+++ b/drivers/bus/arm-cci.c
@@ -16,17 +16,17 @@
 
 #include <linux/arm-cci.h>
 #include <linux/io.h>
+#include <linux/interrupt.h>
 #include <linux/module.h>
 #include <linux/of_address.h>
 #include <linux/of_irq.h>
 #include <linux/of_platform.h>
+#include <linux/perf_event.h>
 #include <linux/platform_device.h>
 #include <linux/slab.h>
 #include <linux/spinlock.h>
 
 #include <asm/cacheflush.h>
-#include <asm/irq_regs.h>
-#include <asm/pmu.h>
 #include <asm/smp_plat.h>
 
 #define DRIVER_NAME		"CCI-400"
@@ -98,6 +98,8 @@ static unsigned long cci_ctrl_phys;
 
 #define CCI_PMU_CNTR_BASE(idx)	((idx) * SZ_4K)
 
+#define CCI_PMU_CNTR_MASK	((1ULL << 32) -1)
+
 /*
  * Instead of an event id to monitor CCI cycles, a dedicated counter is
  * provided. Use 0xff to represent CCI cycles and hope that no future revisions
@@ -170,18 +172,29 @@ static char *const pmu_names[] = {
 	[CCI_REV_R1] = "CCI_400_r1",
 };
 
-struct cci_pmu_drv_data {
+struct cci_pmu_hw_events {
+	struct perf_event *events[CCI_PMU_MAX_HW_EVENTS];
+	unsigned long used_mask[BITS_TO_LONGS(CCI_PMU_MAX_HW_EVENTS)];
+	raw_spinlock_t pmu_lock;
+};
+
+struct cci_pmu {
 	void __iomem *base;
-	struct arm_pmu *cci_pmu;
+	struct pmu pmu;
 	int nr_irqs;
 	int irqs[CCI_PMU_MAX_HW_EVENTS];
 	unsigned long active_irqs;
-	struct perf_event *events[CCI_PMU_MAX_HW_EVENTS];
-	unsigned long used_mask[BITS_TO_LONGS(CCI_PMU_MAX_HW_EVENTS)];
 	struct pmu_port_event_ranges *port_ranges;
-	struct pmu_hw_events hw_events;
+	struct cci_pmu_hw_events hw_events;
+	struct platform_device *plat_device;
+	int num_events;
+	atomic_t active_events;
+	struct mutex reserve_mutex;
+	cpumask_t cpus;
 };
-static struct cci_pmu_drv_data *pmu;
+static struct cci_pmu *pmu;
+
+#define to_cci_pmu(c)	(container_of(c, struct cci_pmu, pmu))
 
 static bool is_duplicate_irq(int irq, int *irqs, int nr_irqs)
 {
@@ -252,7 +265,7 @@ static int pmu_validate_hw_event(u8 hw_event)
 	return -ENOENT;
 }
 
-static int pmu_is_valid_counter(struct arm_pmu *cci_pmu, int idx)
+static int pmu_is_valid_counter(struct cci_pmu *cci_pmu, int idx)
 {
 	return CCI_PMU_CYCLE_CNTR_IDX <= idx &&
 		idx <= CCI_PMU_CNTR_LAST(cci_pmu);
@@ -293,14 +306,9 @@ static u32 pmu_get_max_counters(void)
 	return n_cnts + 1;
 }
 
-static struct pmu_hw_events *pmu_get_hw_events(void)
+static int pmu_get_event_idx(struct cci_pmu_hw_events *hw, struct perf_event *event)
 {
-	return &pmu->hw_events;
-}
-
-static int pmu_get_event_idx(struct pmu_hw_events *hw, struct perf_event *event)
-{
-	struct arm_pmu *cci_pmu = to_arm_pmu(event->pmu);
+	struct cci_pmu *cci_pmu = to_cci_pmu(event->pmu);
 	struct hw_perf_event *hw_event = &event->hw;
 	unsigned long cci_event = hw_event->config_base & CCI_PMU_EVENT_MASK;
 	int idx;
@@ -336,7 +344,7 @@ static int pmu_map_event(struct perf_event *event)
 	return mapping;
 }
 
-static int pmu_request_irq(struct arm_pmu *cci_pmu, irq_handler_t handler)
+static int pmu_request_irq(struct cci_pmu *cci_pmu, irq_handler_t handler)
 {
 	int i;
 	struct platform_device *pmu_device = cci_pmu->plat_device;
@@ -371,17 +379,91 @@ static int pmu_request_irq(struct arm_pmu *cci_pmu, irq_handler_t handler)
 	return 0;
 }
 
+static void pmu_free_irq(struct cci_pmu *cci_pmu)
+{
+	int i;
+
+	for (i = 0; i < pmu->nr_irqs; i++) {
+		if (!test_and_clear_bit(i, &pmu->active_irqs))
+			continue;
+
+		free_irq(pmu->irqs[i], cci_pmu);
+	}
+}
+
+static u32 pmu_read_counter(struct perf_event *event)
+{
+	struct cci_pmu *cci_pmu = to_cci_pmu(event->pmu);
+	struct hw_perf_event *hw_counter = &event->hw;
+	int idx = hw_counter->idx;
+	u32 value;
+
+	if (unlikely(!pmu_is_valid_counter(cci_pmu, idx))) {
+		dev_err(&cci_pmu->plat_device->dev, "Invalid CCI PMU counter %d\n", idx);
+		return 0;
+	}
+	value = pmu_read_register(idx, CCI_PMU_CNTR);
+
+	return value;
+}
+
+static void pmu_write_counter(struct perf_event *event, u32 value)
+{
+	struct cci_pmu *cci_pmu = to_cci_pmu(event->pmu);
+	struct hw_perf_event *hw_counter = &event->hw;
+	int idx = hw_counter->idx;
+
+	if (unlikely(!pmu_is_valid_counter(cci_pmu, idx)))
+		dev_err(&cci_pmu->plat_device->dev, "Invalid CCI PMU counter %d\n", idx);
+	else
+		pmu_write_register(value, idx, CCI_PMU_CNTR);
+}
+
+static u64 pmu_event_update(struct perf_event *event)
+{
+	struct hw_perf_event *hwc = &event->hw;
+	u64 delta, prev_raw_count, new_raw_count;
+
+	do {
+		prev_raw_count = local64_read(&hwc->prev_count);
+		new_raw_count = pmu_read_counter(event);
+	} while (local64_cmpxchg(&hwc->prev_count, prev_raw_count,
+		 new_raw_count) != prev_raw_count);
+
+	delta = (new_raw_count - prev_raw_count) & CCI_PMU_CNTR_MASK;
+
+	local64_add(delta, &event->count);
+
+	return new_raw_count;
+}
+
+static void pmu_read(struct perf_event *event)
+{
+	pmu_event_update(event);
+}
+
+void pmu_event_set_period(struct perf_event *event)
+{
+	struct hw_perf_event *hwc = &event->hw;
+	/*
+	 * The CCI PMU counters have a period of 2^32. To account for the
+	 * possiblity of extreme interrupt latency we program for a period of
+	 * half that. Hopefully we can handle the interrupt before another 2^31
+	 * events occur and the counter overtakes its previous value.
+	 */
+	u64 val = 1ULL << 31;
+	local64_set(&hwc->prev_count, val);
+	pmu_write_counter(event, val);
+}
+
 static irqreturn_t pmu_handle_irq(int irq_num, void *dev)
 {
 	unsigned long flags;
-	struct arm_pmu *cci_pmu = (struct arm_pmu *)dev;
-	struct pmu_hw_events *events = cci_pmu->get_hw_events();
-	struct perf_sample_data data;
-	struct pt_regs *regs;
+	struct cci_pmu *cci_pmu = dev;
+	struct cci_pmu_hw_events *events = &pmu->hw_events;
 	int idx, handled = IRQ_NONE;
 
 	raw_spin_lock_irqsave(&events->pmu_lock, flags);
-	regs = get_irq_regs();
 	/*
 	 * Iterate over counters and update the corresponding perf events.
 	 * This should work regardless of whether we have per-counter overflow
@@ -403,154 +485,407 @@ static irqreturn_t pmu_handle_irq(int irq_num, void *dev)
 
 		pmu_write_register(CCI_PMU_OVRFLW_FLAG, idx, CCI_PMU_OVRFLW);
 
+		pmu_event_update(event);
+		pmu_event_set_period(event);
 		handled = IRQ_HANDLED;
-
-		armpmu_event_update(event);
-		perf_sample_data_init(&data, 0, hw_counter->last_period);
-		if (!armpmu_event_set_period(event))
-			continue;
-
-		if (perf_event_overflow(event, &data, regs))
-			cci_pmu->disable(event);
 	}
 	raw_spin_unlock_irqrestore(&events->pmu_lock, flags);
 
 	return IRQ_RETVAL(handled);
 }
 
-static void pmu_free_irq(struct arm_pmu *cci_pmu)
+static int cci_pmu_get_hw(struct cci_pmu *cci_pmu)
 {
-	int i;
+	int ret = pmu_request_irq(cci_pmu, pmu_handle_irq);
+	if (ret) {
+		pmu_free_irq(cci_pmu);
+		return ret;
+	}
+	return 0;
+}
 
-	for (i = 0; i < pmu->nr_irqs; i++) {
-		if (!test_and_clear_bit(i, &pmu->active_irqs))
-			continue;
+static void cci_pmu_put_hw(struct cci_pmu *cci_pmu)
+{
+	pmu_free_irq(cci_pmu);
+}
 
-		free_irq(pmu->irqs[i], cci_pmu);
+static void hw_perf_event_destroy(struct perf_event *event)
+{
+	struct cci_pmu *cci_pmu = to_cci_pmu(event->pmu);
+	atomic_t *active_events = &cci_pmu->active_events;
+	struct mutex *reserve_mutex = &cci_pmu->reserve_mutex;
+
+	if (atomic_dec_and_mutex_lock(active_events, reserve_mutex)) {
+		cci_pmu_put_hw(cci_pmu);
+		mutex_unlock(reserve_mutex);
 	}
 }
 
-static void pmu_enable_event(struct perf_event *event)
+static void cci_pmu_enable(struct pmu *pmu)
 {
+	struct cci_pmu *cci_pmu = to_cci_pmu(pmu);
+	struct cci_pmu_hw_events *hw_events = &cci_pmu->hw_events;
+	int enabled = bitmap_weight(hw_events->used_mask, cci_pmu->num_events);
 	unsigned long flags;
-	struct arm_pmu *cci_pmu = to_arm_pmu(event->pmu);
-	struct pmu_hw_events *events = cci_pmu->get_hw_events();
-	struct hw_perf_event *hw_counter = &event->hw;
-	int idx = hw_counter->idx;
+	u32 val;
+
+	if (!enabled)
+		return;
+
+	raw_spin_lock_irqsave(&hw_events->pmu_lock, flags);
+
+	/* Enable all the PMU counters. */
+	val = readl_relaxed(cci_ctrl_base + CCI_PMCR) | CCI_PMCR_CEN;
+	writel(val, cci_ctrl_base + CCI_PMCR);
+	raw_spin_unlock_irqrestore(&hw_events->pmu_lock, flags);
+
+}
+
+static void cci_pmu_disable(struct pmu *pmu)
+{
+	struct cci_pmu *cci_pmu = to_cci_pmu(pmu);
+	struct cci_pmu_hw_events *hw_events = &cci_pmu->hw_events;
+	unsigned long flags;
+	u32 val;
+
+	raw_spin_lock_irqsave(&hw_events->pmu_lock, flags);
+
+	/* Disable all the PMU counters. */
+	val = readl_relaxed(cci_ctrl_base + CCI_PMCR) & ~CCI_PMCR_CEN;
+	writel(val, cci_ctrl_base + CCI_PMCR);
+	raw_spin_unlock_irqrestore(&hw_events->pmu_lock, flags);
+}
+
+static void cci_pmu_start(struct perf_event *event, int pmu_flags)
+{
+	struct cci_pmu *cci_pmu = to_cci_pmu(event->pmu);
+	struct cci_pmu_hw_events *hw_events = &cci_pmu->hw_events;
+	struct hw_perf_event *hwc = &event->hw;
+	int idx = hwc->idx;
+	unsigned long flags;
+
+	/*
+	 * To handle interrupt latency, we always reprogram the period
+	 * regardlesss of PERF_EF_RELOAD.
+	 */
+	if (pmu_flags & PERF_EF_RELOAD)
+		WARN_ON_ONCE(!(hwc->state & PERF_HES_UPTODATE));
+
+	hwc->state = 0;
 
 	if (unlikely(!pmu_is_valid_counter(cci_pmu, idx))) {
 		dev_err(&cci_pmu->plat_device->dev, "Invalid CCI PMU counter %d\n", idx);
 		return;
 	}
 
-	raw_spin_lock_irqsave(&events->pmu_lock, flags);
+	raw_spin_lock_irqsave(&hw_events->pmu_lock, flags);
 
 	/* Configure the event to count, unless you are counting cycles */
 	if (idx != CCI_PMU_CYCLE_CNTR_IDX)
-		pmu_set_event(idx, hw_counter->config_base);
+		pmu_set_event(idx, hwc->config_base);
 
+	pmu_event_set_period(event);
 	pmu_enable_counter(idx);
 
-	raw_spin_unlock_irqrestore(&events->pmu_lock, flags);
+	raw_spin_unlock_irqrestore(&hw_events->pmu_lock, flags);
 }
 
-static void pmu_disable_event(struct perf_event *event)
+static void cci_pmu_stop(struct perf_event *event, int pmu_flags)
 {
-	struct arm_pmu *cci_pmu = to_arm_pmu(event->pmu);
-	struct hw_perf_event *hw_counter = &event->hw;
-	int idx = hw_counter->idx;
+	struct cci_pmu *cci_pmu = to_cci_pmu(event->pmu);
+	struct hw_perf_event *hwc = &event->hw;
+	int idx = hwc->idx;
+
+	if (hwc->state & PERF_HES_STOPPED)
+		return;
 
 	if (unlikely(!pmu_is_valid_counter(cci_pmu, idx))) {
 		dev_err(&cci_pmu->plat_device->dev, "Invalid CCI PMU counter %d\n", idx);
 		return;
 	}
 
+	/*
+	 * We always reprogram the counter, so ignore PERF_EF_UPDATE. See
+	 * cci_pmu_start()
+	 */
 	pmu_disable_counter(idx);
+	pmu_event_update(event);
+	hwc->state |= PERF_HES_STOPPED | PERF_HES_UPTODATE;
 }
 
-static void pmu_start(struct arm_pmu *cci_pmu)
+static int cci_pmu_add(struct perf_event *event, int flags)
 {
-	u32 val;
-	unsigned long flags;
-	struct pmu_hw_events *events = cci_pmu->get_hw_events();
+	struct cci_pmu *cci_pmu = to_cci_pmu(event->pmu);
+	struct cci_pmu_hw_events *hw_events = &cci_pmu->hw_events;
+	struct hw_perf_event *hwc = &event->hw;
+	int idx;
+	int err = 0;
 
-	raw_spin_lock_irqsave(&events->pmu_lock, flags);
+	perf_pmu_disable(event->pmu);
 
-	/* Enable all the PMU counters. */
-	val = readl_relaxed(cci_ctrl_base + CCI_PMCR) | CCI_PMCR_CEN;
-	writel(val, cci_ctrl_base + CCI_PMCR);
+	/* If we don't have a space for the counter then finish early. */
+	idx = pmu_get_event_idx(hw_events, event);
+	if (idx < 0) {
+		err = idx;
+		goto out;
+	}
 
-	raw_spin_unlock_irqrestore(&events->pmu_lock, flags);
+	event->hw.idx = idx;
+	hw_events->events[idx] = event;
+
+	hwc->state = PERF_HES_STOPPED | PERF_HES_UPTODATE;
+	if (flags & PERF_EF_START)
+		cci_pmu_start(event, PERF_EF_RELOAD);
+
+	/* Propagate our changes to the userspace mapping. */
+	perf_event_update_userpage(event);
+
+out:
+	perf_pmu_enable(event->pmu);
+	return err;
 }
 
-static void pmu_stop(struct arm_pmu *cci_pmu)
+static void cci_pmu_del(struct perf_event *event, int flags)
 {
-	u32 val;
-	unsigned long flags;
-	struct pmu_hw_events *events = cci_pmu->get_hw_events();
+	struct cci_pmu *cci_pmu = to_cci_pmu(event->pmu);
+	struct cci_pmu_hw_events *hw_events = &cci_pmu->hw_events;
+	struct hw_perf_event *hwc = &event->hw;
+	int idx = hwc->idx;
 
-	raw_spin_lock_irqsave(&events->pmu_lock, flags);
+	cci_pmu_stop(event, PERF_EF_UPDATE);
+	hw_events->events[idx] = NULL;
+	clear_bit(idx, hw_events->used_mask);
 
-	/* Disable all the PMU counters. */
-	val = readl_relaxed(cci_ctrl_base + CCI_PMCR) & ~CCI_PMCR_CEN;
-	writel(val, cci_ctrl_base + CCI_PMCR);
+	perf_event_update_userpage(event);
+}
 
-	raw_spin_unlock_irqrestore(&events->pmu_lock, flags);
+static int
+validate_event(struct cci_pmu_hw_events *hw_events,
+	       struct perf_event *event)
+{
+	if (is_software_event(event))
+		return 1;
+
+	if (event->state < PERF_EVENT_STATE_OFF)
+		return 1;
+
+	if (event->state == PERF_EVENT_STATE_OFF && !event->attr.enable_on_exec)
+		return 1;
+
+	return pmu_get_event_idx(hw_events, event) >= 0;
 }
 
-static u32 pmu_read_counter(struct perf_event *event)
+static int
+validate_group(struct perf_event *event)
 {
-	struct arm_pmu *cci_pmu = to_arm_pmu(event->pmu);
-	struct hw_perf_event *hw_counter = &event->hw;
-	int idx = hw_counter->idx;
-	u32 value;
+	struct perf_event *sibling, *leader = event->group_leader;
+	struct cci_pmu_hw_events fake_pmu = {
+		/*
+		 * Initialise the fake PMU. We only need to populate the
+		 * used_mask for the purposes of validation.
+		 */
+		.used_mask = CPU_BITS_NONE,
+	};
 
-	if (unlikely(!pmu_is_valid_counter(cci_pmu, idx))) {
-		dev_err(&cci_pmu->plat_device->dev, "Invalid CCI PMU counter %d\n", idx);
-		return 0;
+	if (!validate_event(&fake_pmu, leader))
+		return -EINVAL;
+
+	list_for_each_entry(sibling, &leader->sibling_list, group_entry) {
+		if (!validate_event(&fake_pmu, sibling))
+			return -EINVAL;
 	}
-	value = pmu_read_register(idx, CCI_PMU_CNTR);
 
-	return value;
+	if (!validate_event(&fake_pmu, event))
+		return -EINVAL;
+
+	return 0;
 }
 
-static void pmu_write_counter(struct perf_event *event, u32 value)
+static int
+__hw_perf_event_init(struct perf_event *event)
 {
-	struct arm_pmu *cci_pmu = to_arm_pmu(event->pmu);
-	struct hw_perf_event *hw_counter = &event->hw;
-	int idx = hw_counter->idx;
+	struct hw_perf_event *hwc = &event->hw;
+	int mapping;
 
-	if (unlikely(!pmu_is_valid_counter(cci_pmu, idx)))
-		dev_err(&cci_pmu->plat_device->dev, "Invalid CCI PMU counter %d\n", idx);
-	else
-		pmu_write_register(value, idx, CCI_PMU_CNTR);
+	mapping = pmu_map_event(event);
+
+	if (mapping < 0) {
+		pr_debug("event %x:%llx not supported\n", event->attr.type,
+			 event->attr.config);
+		return mapping;
+	}
+
+	/*
+	 * We don't assign an index until we actually place the event onto
+	 * hardware. Use -1 to signify that we haven't decided where to put it
+	 * yet.
+	 */
+	hwc->idx		= -1;
+	hwc->config_base	= 0;
+	hwc->config		= 0;
+	hwc->event_base		= 0;
+
+	/*
+	 * Store the event encoding into the config_base field.
+	 */
+	hwc->config_base	    |= (unsigned long)mapping;
+
+	/*
+	 * Limit the sample_period to half of the counter width. That way, the
+	 * new counter value is far less likely to overtake the previous one
+	 * unless you have some serious IRQ latency issues.
+	 */
+	hwc->sample_period  = CCI_PMU_CNTR_MASK >> 1;
+	hwc->last_period    = hwc->sample_period;
+	local64_set(&hwc->period_left, hwc->sample_period);
+
+	if (event->group_leader != event) {
+		if (validate_group(event) != 0)
+			return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int cci_pmu_event_init(struct perf_event *event)
+{
+	struct cci_pmu *cci_pmu = to_cci_pmu(event->pmu);
+	atomic_t *active_events = &cci_pmu->active_events;
+	int err = 0;
+	int cpu;
+
+	if (event->attr.type != event->pmu->type)
+		return -ENOENT;
+
+	/* Shared by all CPUs, no meaningful state to sample */
+	if (is_sampling_event(event) || event->attach_state & PERF_ATTACH_TASK)
+		return -EOPNOTSUPP;
+
+	/* We have no filtering of any kind */
+	if (event->attr.exclude_user	||
+	    event->attr.exclude_kernel	||
+	    event->attr.exclude_hv	||
+	    event->attr.exclude_idle	||
+	    event->attr.exclude_host	||
+	    event->attr.exclude_guest)
+		return -EINVAL;
+
+	/*
+	 * Following the example set by other "uncore" PMUs, we accept any CPU
+	 * and rewrite its affinity dynamically rather than having perf core
+	 * handle cpu == -1 and pid == -1 for this case.
+	 *
+	 * The perf core will pin online CPUs for the duration of this call and
+	 * the event being installed into its context, so the PMU's CPU can't
+	 * change under our feet.
+	 */
+	cpu = cpumask_first(&cci_pmu->cpus);
+	if (event->cpu < 0 || cpu < 0)
+		return -EINVAL;
+	event->cpu = cpu;
+
+	event->destroy = hw_perf_event_destroy;
+	if (!atomic_inc_not_zero(active_events)) {
+		mutex_lock(&cci_pmu->reserve_mutex);
+		if (atomic_read(active_events) == 0)
+			err = cci_pmu_get_hw(cci_pmu);
+		if (!err)
+			atomic_inc(active_events);
+		mutex_unlock(&cci_pmu->reserve_mutex);
+	}
+	if (err)
+		return err;
+
+	err = __hw_perf_event_init(event);
+	if (err)
+		hw_perf_event_destroy(event);
+
+	return err;
+}
+
+static ssize_t pmu_attr_cpumask_show(struct device *dev,
+				     struct device_attribute *attr, char *buf)
+{
+	int n = cpulist_scnprintf(buf, PAGE_SIZE - 2, &pmu->cpus);
+
+	buf[n++] = '\n';
+	buf[n] = '\0';
+	return n;
 }
 
-static int cci_pmu_init(struct arm_pmu *cci_pmu, struct platform_device *pdev)
+static DEVICE_ATTR(cpumask, S_IRUGO, pmu_attr_cpumask_show, NULL);
+
+static struct attribute *pmu_attrs[] = {
+	&dev_attr_cpumask.attr,
+	NULL,
+};
+
+static struct attribute_group pmu_attr_group = {
+	.attrs = pmu_attrs,
+};
+
+static const struct attribute_group *pmu_attr_groups[] = {
+	&pmu_attr_group,
+	NULL
+};
+
+static int cci_pmu_init(struct cci_pmu *cci_pmu, struct platform_device *pdev)
 {
-	*cci_pmu = (struct arm_pmu){
-		.name		  = pmu_names[probe_cci_revision()],
-		.max_period       = (1LLU << 32) - 1,
-		.get_hw_events    = pmu_get_hw_events,
-		.get_event_idx    = pmu_get_event_idx,
-		.map_event        = pmu_map_event,
-		.request_irq      = pmu_request_irq,
-		.handle_irq       = pmu_handle_irq,
-		.free_irq         = pmu_free_irq,
-		.enable           = pmu_enable_event,
-		.disable          = pmu_disable_event,
-		.start            = pmu_start,
-		.stop             = pmu_stop,
-		.read_counter     = pmu_read_counter,
-		.write_counter    = pmu_write_counter,
+	char *name = pmu_names[probe_cci_revision()];
+	cci_pmu->pmu = (struct pmu) {
+		.name		= pmu_names[probe_cci_revision()],
+		.task_ctx_nr	= perf_invalid_context,
+		.pmu_enable	= cci_pmu_enable,
+		.pmu_disable	= cci_pmu_disable,
+		.event_init	= cci_pmu_event_init,
+		.add		= cci_pmu_add,
+		.del		= cci_pmu_del,
+		.start		= cci_pmu_start,
+		.stop		= cci_pmu_stop,
+		.read		= pmu_read,
+		.attr_groups	= pmu_attr_groups,
 	};
 
 	cci_pmu->plat_device = pdev;
 	cci_pmu->num_events = pmu_get_max_counters();
 
-	return armpmu_register(cci_pmu, -1);
+	return perf_pmu_register(&cci_pmu->pmu, name, -1);
+}
+
+static int cci_pmu_cpu_notifier(struct notifier_block *self,
+				unsigned long action, void *hcpu)
+{
+	unsigned int cpu = (long)hcpu;
+	unsigned int target;
+
+	switch (action & ~CPU_TASKS_FROZEN) {
+	case CPU_DOWN_PREPARE:
+		if (!cpumask_test_and_clear_cpu(cpu, &pmu->cpus))
+			break;
+		target = cpumask_any_but(cpu_online_mask, cpu);
+		if (target < 0) // UP, last CPU
+			break;
+		/*
+		 * TODO: migrate context once core races on event->ctx have
+		 * been fixed.
+		 */
+		cpumask_set_cpu(target, &pmu->cpus);
+	default:
+		break;
+	}
+
+	return NOTIFY_OK;
 }
 
+static struct notifier_block cci_pmu_cpu_nb = {
+	.notifier_call	= cci_pmu_cpu_notifier,
+	/*
+	 * to migrate uncore events, our notifier should be executed
+	 * before perf core's notifier.
+	 */
+	.priority	= CPU_PRI_PERF + 1,
+};
+
 static const struct of_device_id arm_cci_pmu_matches[] = {
 	{
 		.compatible = "arm,cci-400-pmu",
@@ -604,15 +939,16 @@ static int cci_pmu_probe(struct platform_device *pdev)
 		return -EINVAL;
 	}
 
-	pmu->cci_pmu = devm_kzalloc(&pdev->dev, sizeof(*(pmu->cci_pmu)), GFP_KERNEL);
-	if (!pmu->cci_pmu)
-		return -ENOMEM;
-
-	pmu->hw_events.events = pmu->events;
-	pmu->hw_events.used_mask = pmu->used_mask;
 	raw_spin_lock_init(&pmu->hw_events.pmu_lock);
+	mutex_init(&pmu->reserve_mutex);
+	atomic_set(&pmu->active_events, 0);
+	cpumask_set_cpu(smp_processor_id(), &pmu->cpus);
 
-	ret = cci_pmu_init(pmu->cci_pmu, pdev);
+	ret = register_cpu_notifier(&cci_pmu_cpu_nb);
+	if (ret)
+		return ret;
+
+	ret = cci_pmu_init(pmu, pdev);
 	if (ret)
 		return ret;
 
@@ -976,6 +1312,9 @@ static int cci_probe(void)
 	if (!np)
 		return -ENODEV;
 
+	if (!of_device_is_available(np))
+		return -ENODEV;
+
 	cci_config = of_match_node(arm_cci_matches, np)->data;
 	if (!cci_config)
 		return -ENODEV;
diff --git a/drivers/bus/arm-ccn.c b/drivers/bus/arm-ccn.c
index a60f26400705..aaa0f2a87118 100644
--- a/drivers/bus/arm-ccn.c
+++ b/drivers/bus/arm-ccn.c
@@ -57,6 +57,7 @@
 #define CCN_DT_PMCCNTRSR		0x0190
 #define CCN_DT_PMOVSR			0x0198
 #define CCN_DT_PMOVSR_CLR		0x01a0
+#define CCN_DT_PMOVSR_CLR__MASK				0x1f
 #define CCN_DT_PMCR			0x01a8
 #define CCN_DT_PMCR__OVFL_INTR_EN			(1 << 6)
 #define CCN_DT_PMCR__PMU_EN				(1 << 0)
@@ -1051,7 +1052,8 @@ static irqreturn_t arm_ccn_pmu_overflow_handler(struct arm_ccn_dt *dt)
 		struct perf_event *event = dt->pmu_counters[idx].event;
 		int overflowed = pmovsr & BIT(idx);
 
-		WARN_ON_ONCE(overflowed && !event);
+		WARN_ON_ONCE(overflowed && !event &&
+				idx != CCN_IDX_PMU_CYCLE_COUNTER);
 
 		if (!event || !overflowed)
 			continue;
@@ -1087,6 +1089,7 @@ static int arm_ccn_pmu_init(struct arm_ccn *ccn)
 	/* Initialize DT subsystem */
 	ccn->dt.base = ccn->base + CCN_REGION_SIZE;
 	spin_lock_init(&ccn->dt.config_lock);
+	writel(CCN_DT_PMOVSR_CLR__MASK, ccn->dt.base + CCN_DT_PMOVSR_CLR);
 	writel(CCN_DT_CTL__DT_EN, ccn->dt.base + CCN_DT_CTL);
 	writel(CCN_DT_PMCR__OVFL_INTR_EN | CCN_DT_PMCR__PMU_EN,
 			ccn->dt.base + CCN_DT_PMCR);
diff --git a/drivers/bus/brcmstb_gisb.c b/drivers/bus/brcmstb_gisb.c
index f2cd6a2d40b4..738612c45266 100644
--- a/drivers/bus/brcmstb_gisb.c
+++ b/drivers/bus/brcmstb_gisb.c
@@ -23,35 +23,103 @@
 #include <linux/list.h>
 #include <linux/of.h>
 #include <linux/bitops.h>
+#include <linux/pm.h>
 
+#ifdef CONFIG_ARM
 #include <asm/bug.h>
 #include <asm/signal.h>
+#endif
 
-#define ARB_TIMER			0x008
-#define ARB_ERR_CAP_CLR			0x7e4
 #define  ARB_ERR_CAP_CLEAR		(1 << 0)
-#define ARB_ERR_CAP_HI_ADDR		0x7e8
-#define ARB_ERR_CAP_ADDR		0x7ec
-#define ARB_ERR_CAP_DATA		0x7f0
-#define ARB_ERR_CAP_STATUS		0x7f4
 #define  ARB_ERR_CAP_STATUS_TIMEOUT	(1 << 12)
 #define  ARB_ERR_CAP_STATUS_TEA		(1 << 11)
 #define  ARB_ERR_CAP_STATUS_BS_SHIFT	(1 << 2)
 #define  ARB_ERR_CAP_STATUS_BS_MASK	0x3c
 #define  ARB_ERR_CAP_STATUS_WRITE	(1 << 1)
 #define  ARB_ERR_CAP_STATUS_VALID	(1 << 0)
-#define ARB_ERR_CAP_MASTER		0x7f8
+
+enum {
+	ARB_TIMER,
+	ARB_ERR_CAP_CLR,
+	ARB_ERR_CAP_HI_ADDR,
+	ARB_ERR_CAP_ADDR,
+	ARB_ERR_CAP_DATA,
+	ARB_ERR_CAP_STATUS,
+	ARB_ERR_CAP_MASTER,
+};
+
+static const int gisb_offsets_bcm7038[] = {
+	[ARB_TIMER]		= 0x00c,
+	[ARB_ERR_CAP_CLR]	= 0x0c4,
+	[ARB_ERR_CAP_HI_ADDR]	= -1,
+	[ARB_ERR_CAP_ADDR]	= 0x0c8,
+	[ARB_ERR_CAP_DATA]	= 0x0cc,
+	[ARB_ERR_CAP_STATUS]	= 0x0d0,
+	[ARB_ERR_CAP_MASTER]	= -1,
+};
+
+static const int gisb_offsets_bcm7400[] = {
+	[ARB_TIMER]		= 0x00c,
+	[ARB_ERR_CAP_CLR]	= 0x0c8,
+	[ARB_ERR_CAP_HI_ADDR]	= -1,
+	[ARB_ERR_CAP_ADDR]	= 0x0cc,
+	[ARB_ERR_CAP_DATA]	= 0x0d0,
+	[ARB_ERR_CAP_STATUS]	= 0x0d4,
+	[ARB_ERR_CAP_MASTER]	= 0x0d8,
+};
+
+static const int gisb_offsets_bcm7435[] = {
+	[ARB_TIMER]		= 0x00c,
+	[ARB_ERR_CAP_CLR]	= 0x168,
+	[ARB_ERR_CAP_HI_ADDR]	= -1,
+	[ARB_ERR_CAP_ADDR]	= 0x16c,
+	[ARB_ERR_CAP_DATA]	= 0x170,
+	[ARB_ERR_CAP_STATUS]	= 0x174,
+	[ARB_ERR_CAP_MASTER]	= 0x178,
+};
+
+static const int gisb_offsets_bcm7445[] = {
+	[ARB_TIMER]		= 0x008,
+	[ARB_ERR_CAP_CLR]	= 0x7e4,
+	[ARB_ERR_CAP_HI_ADDR]	= 0x7e8,
+	[ARB_ERR_CAP_ADDR]	= 0x7ec,
+	[ARB_ERR_CAP_DATA]	= 0x7f0,
+	[ARB_ERR_CAP_STATUS]	= 0x7f4,
+	[ARB_ERR_CAP_MASTER]	= 0x7f8,
+};
 
 struct brcmstb_gisb_arb_device {
 	void __iomem	*base;
+	const int	*gisb_offsets;
 	struct mutex	lock;
 	struct list_head next;
 	u32 valid_mask;
 	const char *master_names[sizeof(u32) * BITS_PER_BYTE];
+	u32 saved_timeout;
 };
 
 static LIST_HEAD(brcmstb_gisb_arb_device_list);
 
+static u32 gisb_read(struct brcmstb_gisb_arb_device *gdev, int reg)
+{
+	int offset = gdev->gisb_offsets[reg];
+
+	/* return 1 if the hardware doesn't have ARB_ERR_CAP_MASTER */
+	if (offset == -1)
+		return 1;
+
+	return ioread32(gdev->base + offset);
+}
+
+static void gisb_write(struct brcmstb_gisb_arb_device *gdev, u32 val, int reg)
+{
+	int offset = gdev->gisb_offsets[reg];
+
+	if (offset == -1)
+		return;
+	iowrite32(val, gdev->base + reg);
+}
+
 static ssize_t gisb_arb_get_timeout(struct device *dev,
 				    struct device_attribute *attr,
 				    char *buf)
@@ -61,7 +129,7 @@ static ssize_t gisb_arb_get_timeout(struct device *dev,
 	u32 timeout;
 
 	mutex_lock(&gdev->lock);
-	timeout = ioread32(gdev->base + ARB_TIMER);
+	timeout = gisb_read(gdev, ARB_TIMER);
 	mutex_unlock(&gdev->lock);
 
 	return sprintf(buf, "%d", timeout);
@@ -83,7 +151,7 @@ static ssize_t gisb_arb_set_timeout(struct device *dev,
 		return -EINVAL;
 
 	mutex_lock(&gdev->lock);
-	iowrite32(val, gdev->base + ARB_TIMER);
+	gisb_write(gdev, val, ARB_TIMER);
 	mutex_unlock(&gdev->lock);
 
 	return count;
@@ -110,18 +178,18 @@ static int brcmstb_gisb_arb_decode_addr(struct brcmstb_gisb_arb_device *gdev,
 	const char *m_name;
 	char m_fmt[11];
 
-	cap_status = ioread32(gdev->base + ARB_ERR_CAP_STATUS);
+	cap_status = gisb_read(gdev, ARB_ERR_CAP_STATUS);
 
 	/* Invalid captured address, bail out */
 	if (!(cap_status & ARB_ERR_CAP_STATUS_VALID))
 		return 1;
 
 	/* Read the address and master */
-	arb_addr = ioread32(gdev->base + ARB_ERR_CAP_ADDR) & 0xffffffff;
+	arb_addr = gisb_read(gdev, ARB_ERR_CAP_ADDR) & 0xffffffff;
 #if (IS_ENABLED(CONFIG_PHYS_ADDR_T_64BIT))
-	arb_addr |= (u64)ioread32(gdev->base + ARB_ERR_CAP_HI_ADDR) << 32;
+	arb_addr |= (u64)gisb_read(gdev, ARB_ERR_CAP_HI_ADDR) << 32;
 #endif
-	master = ioread32(gdev->base + ARB_ERR_CAP_MASTER);
+	master = gisb_read(gdev, ARB_ERR_CAP_MASTER);
 
 	m_name = brcmstb_gisb_master_to_str(gdev, master);
 	if (!m_name) {
@@ -136,11 +204,12 @@ static int brcmstb_gisb_arb_decode_addr(struct brcmstb_gisb_arb_device *gdev,
 		m_name);
 
 	/* clear the GISB error */
-	iowrite32(ARB_ERR_CAP_CLEAR, gdev->base + ARB_ERR_CAP_CLR);
+	gisb_write(gdev, ARB_ERR_CAP_CLEAR, ARB_ERR_CAP_CLR);
 
 	return 0;
 }
 
+#ifdef CONFIG_ARM
 static int brcmstb_bus_error_handler(unsigned long addr, unsigned int fsr,
 				     struct pt_regs *regs)
 {
@@ -159,12 +228,7 @@ static int brcmstb_bus_error_handler(unsigned long addr, unsigned int fsr,
 
 	return ret;
 }
-
-void __init brcmstb_hook_fault_code(void)
-{
-	hook_fault_code(22, brcmstb_bus_error_handler, SIGBUS, 0,
-			"imprecise external abort");
-}
+#endif
 
 static irqreturn_t brcmstb_gisb_timeout_handler(int irq, void *dev_id)
 {
@@ -192,10 +256,20 @@ static struct attribute_group gisb_arb_sysfs_attr_group = {
 	.attrs = gisb_arb_sysfs_attrs,
 };
 
-static int brcmstb_gisb_arb_probe(struct platform_device *pdev)
+static const struct of_device_id brcmstb_gisb_arb_of_match[] = {
+	{ .compatible = "brcm,gisb-arb",         .data = gisb_offsets_bcm7445 },
+	{ .compatible = "brcm,bcm7445-gisb-arb", .data = gisb_offsets_bcm7445 },
+	{ .compatible = "brcm,bcm7435-gisb-arb", .data = gisb_offsets_bcm7435 },
+	{ .compatible = "brcm,bcm7400-gisb-arb", .data = gisb_offsets_bcm7400 },
+	{ .compatible = "brcm,bcm7038-gisb-arb", .data = gisb_offsets_bcm7038 },
+	{ },
+};
+
+static int __init brcmstb_gisb_arb_probe(struct platform_device *pdev)
 {
 	struct device_node *dn = pdev->dev.of_node;
 	struct brcmstb_gisb_arb_device *gdev;
+	const struct of_device_id *of_id;
 	struct resource *r;
 	int err, timeout_irq, tea_irq;
 	unsigned int num_masters, j = 0;
@@ -216,6 +290,13 @@ static int brcmstb_gisb_arb_probe(struct platform_device *pdev)
 	if (IS_ERR(gdev->base))
 		return PTR_ERR(gdev->base);
 
+	of_id = of_match_node(brcmstb_gisb_arb_of_match, dn);
+	if (!of_id) {
+		pr_err("failed to look up compatible string\n");
+		return -EINVAL;
+	}
+	gdev->gisb_offsets = of_id->data;
+
 	err = devm_request_irq(&pdev->dev, timeout_irq,
 				brcmstb_gisb_timeout_handler, 0, pdev->name,
 				gdev);
@@ -261,29 +342,62 @@ static int brcmstb_gisb_arb_probe(struct platform_device *pdev)
 
 	list_add_tail(&gdev->next, &brcmstb_gisb_arb_device_list);
 
+#ifdef CONFIG_ARM
+	hook_fault_code(22, brcmstb_bus_error_handler, SIGBUS, 0,
+			"imprecise external abort");
+#endif
+
 	dev_info(&pdev->dev, "registered mem: %p, irqs: %d, %d\n",
 			gdev->base, timeout_irq, tea_irq);
 
 	return 0;
 }
 
-static const struct of_device_id brcmstb_gisb_arb_of_match[] = {
-	{ .compatible = "brcm,gisb-arb" },
-	{ },
+#ifdef CONFIG_PM_SLEEP
+static int brcmstb_gisb_arb_suspend(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct brcmstb_gisb_arb_device *gdev = platform_get_drvdata(pdev);
+
+	gdev->saved_timeout = gisb_read(gdev, ARB_TIMER);
+
+	return 0;
+}
+
+/* Make sure we provide the same timeout value that was configured before, and
+ * do this before the GISB timeout interrupt handler has any chance to run.
+ */
+static int brcmstb_gisb_arb_resume_noirq(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct brcmstb_gisb_arb_device *gdev = platform_get_drvdata(pdev);
+
+	gisb_write(gdev, gdev->saved_timeout, ARB_TIMER);
+
+	return 0;
+}
+#else
+#define brcmstb_gisb_arb_suspend       NULL
+#define brcmstb_gisb_arb_resume_noirq  NULL
+#endif
+
+static const struct dev_pm_ops brcmstb_gisb_arb_pm_ops = {
+	.suspend	= brcmstb_gisb_arb_suspend,
+	.resume_noirq	= brcmstb_gisb_arb_resume_noirq,
 };
 
 static struct platform_driver brcmstb_gisb_arb_driver = {
-	.probe	= brcmstb_gisb_arb_probe,
 	.driver = {
 		.name	= "brcm-gisb-arb",
-		.owner	= THIS_MODULE,
 		.of_match_table = brcmstb_gisb_arb_of_match,
+		.pm	= &brcmstb_gisb_arb_pm_ops,
 	},
 };
 
 static int __init brcm_gisb_driver_init(void)
 {
-	return platform_driver_register(&brcmstb_gisb_arb_driver);
+	return platform_driver_probe(&brcmstb_gisb_arb_driver,
+				     brcmstb_gisb_arb_probe);
 }
 
 module_init(brcm_gisb_driver_init);
diff --git a/drivers/bus/imx-weim.c b/drivers/bus/imx-weim.c
index 75c9681f8021..0958b6981773 100644
--- a/drivers/bus/imx-weim.c
+++ b/drivers/bus/imx-weim.c
@@ -206,7 +206,6 @@ static int __init weim_probe(struct platform_device *pdev)
 static struct platform_driver weim_driver = {
 	.driver = {
 		.name		= "imx-weim",
-		.owner		= THIS_MODULE,
 		.of_match_table	= weim_id_table,
 	},
 };
diff --git a/drivers/bus/mvebu-mbus.c b/drivers/bus/mvebu-mbus.c
index 26c3779d871d..81bf297f1034 100644
--- a/drivers/bus/mvebu-mbus.c
+++ b/drivers/bus/mvebu-mbus.c
@@ -57,6 +57,7 @@
 #include <linux/of_address.h>
 #include <linux/debugfs.h>
 #include <linux/log2.h>
+#include <linux/syscore_ops.h>
 
 /*
  * DDR target is the same on all platforms.
@@ -94,20 +95,42 @@
 
 #define DOVE_DDR_BASE_CS_OFF(n) ((n) << 4)
 
+/* Relative to mbusbridge_base */
+#define MBUS_BRIDGE_CTRL_OFF	0x0
+#define MBUS_BRIDGE_BASE_OFF	0x4
+
+/* Maximum number of windows, for all known platforms */
+#define MBUS_WINS_MAX           20
+
 struct mvebu_mbus_state;
 
 struct mvebu_mbus_soc_data {
 	unsigned int num_wins;
 	unsigned int num_remappable_wins;
+	bool has_mbus_bridge;
 	unsigned int (*win_cfg_offset)(const int win);
 	void (*setup_cpu_target)(struct mvebu_mbus_state *s);
+	int (*save_cpu_target)(struct mvebu_mbus_state *s,
+			       u32 *store_addr);
 	int (*show_cpu_target)(struct mvebu_mbus_state *s,
 			       struct seq_file *seq, void *v);
 };
 
+/*
+ * Used to store the state of one MBus window accross suspend/resume.
+ */
+struct mvebu_mbus_win_data {
+	u32 ctrl;
+	u32 base;
+	u32 remap_lo;
+	u32 remap_hi;
+};
+
 struct mvebu_mbus_state {
 	void __iomem *mbuswins_base;
 	void __iomem *sdramwins_base;
+	void __iomem *mbusbridge_base;
+	phys_addr_t sdramwins_phys_base;
 	struct dentry *debugfs_root;
 	struct dentry *debugfs_sdram;
 	struct dentry *debugfs_devs;
@@ -115,6 +138,11 @@ struct mvebu_mbus_state {
 	struct resource pcie_io_aperture;
 	const struct mvebu_mbus_soc_data *soc;
 	int hw_io_coherency;
+
+	/* Used during suspend/resume */
+	u32 mbus_bridge_ctrl;
+	u32 mbus_bridge_base;
+	struct mvebu_mbus_win_data wins[MBUS_WINS_MAX];
 };
 
 static struct mvebu_mbus_state mbus_state;
@@ -182,12 +210,25 @@ static void mvebu_mbus_disable_window(struct mvebu_mbus_state *mbus,
 }
 
 /* Checks whether the given window number is available */
+
+/* On Armada XP, 375 and 38x the MBus window 13 has the remap
+ * capability, like windows 0 to 7. However, the mvebu-mbus driver
+ * isn't currently taking into account this special case, which means
+ * that when window 13 is actually used, the remap registers are left
+ * to 0, making the device using this MBus window unavailable. The
+ * quick fix for stable is to not use window 13. A follow up patch
+ * will correctly handle this window.
+*/
 static int mvebu_mbus_window_is_free(struct mvebu_mbus_state *mbus,
 				     const int win)
 {
 	void __iomem *addr = mbus->mbuswins_base +
 		mbus->soc->win_cfg_offset(win);
 	u32 ctrl = readl(addr + WIN_CTRL_OFF);
+
+	if (win == 13)
+		return false;
+
 	return !(ctrl & WIN_CTRL_ENABLE);
 }
 
@@ -516,6 +557,28 @@ mvebu_mbus_default_setup_cpu_target(struct mvebu_mbus_state *mbus)
 	mvebu_mbus_dram_info.num_cs = cs;
 }
 
+static int
+mvebu_mbus_default_save_cpu_target(struct mvebu_mbus_state *mbus,
+				   u32 *store_addr)
+{
+	int i;
+
+	for (i = 0; i < 4; i++) {
+		u32 base = readl(mbus->sdramwins_base + DDR_BASE_CS_OFF(i));
+		u32 size = readl(mbus->sdramwins_base + DDR_SIZE_CS_OFF(i));
+
+		writel(mbus->sdramwins_phys_base + DDR_BASE_CS_OFF(i),
+		       store_addr++);
+		writel(base, store_addr++);
+		writel(mbus->sdramwins_phys_base + DDR_SIZE_CS_OFF(i),
+		       store_addr++);
+		writel(size, store_addr++);
+	}
+
+	/* We've written 16 words to the store address */
+	return 16;
+}
+
 static void __init
 mvebu_mbus_dove_setup_cpu_target(struct mvebu_mbus_state *mbus)
 {
@@ -546,10 +609,35 @@ mvebu_mbus_dove_setup_cpu_target(struct mvebu_mbus_state *mbus)
 	mvebu_mbus_dram_info.num_cs = cs;
 }
 
+static int
+mvebu_mbus_dove_save_cpu_target(struct mvebu_mbus_state *mbus,
+				u32 *store_addr)
+{
+	int i;
+
+	for (i = 0; i < 2; i++) {
+		u32 map = readl(mbus->sdramwins_base + DOVE_DDR_BASE_CS_OFF(i));
+
+		writel(mbus->sdramwins_phys_base + DOVE_DDR_BASE_CS_OFF(i),
+		       store_addr++);
+		writel(map, store_addr++);
+	}
+
+	/* We've written 4 words to the store address */
+	return 4;
+}
+
+int mvebu_mbus_save_cpu_target(u32 *store_addr)
+{
+	return mbus_state.soc->save_cpu_target(&mbus_state, store_addr);
+}
+
 static const struct mvebu_mbus_soc_data armada_370_xp_mbus_data = {
 	.num_wins            = 20,
 	.num_remappable_wins = 8,
+	.has_mbus_bridge     = true,
 	.win_cfg_offset      = armada_370_xp_mbus_win_offset,
+	.save_cpu_target     = mvebu_mbus_default_save_cpu_target,
 	.setup_cpu_target    = mvebu_mbus_default_setup_cpu_target,
 	.show_cpu_target     = mvebu_sdram_debug_show_orion,
 };
@@ -558,6 +646,7 @@ static const struct mvebu_mbus_soc_data kirkwood_mbus_data = {
 	.num_wins            = 8,
 	.num_remappable_wins = 4,
 	.win_cfg_offset      = orion_mbus_win_offset,
+	.save_cpu_target     = mvebu_mbus_default_save_cpu_target,
 	.setup_cpu_target    = mvebu_mbus_default_setup_cpu_target,
 	.show_cpu_target     = mvebu_sdram_debug_show_orion,
 };
@@ -566,6 +655,7 @@ static const struct mvebu_mbus_soc_data dove_mbus_data = {
 	.num_wins            = 8,
 	.num_remappable_wins = 4,
 	.win_cfg_offset      = orion_mbus_win_offset,
+	.save_cpu_target     = mvebu_mbus_dove_save_cpu_target,
 	.setup_cpu_target    = mvebu_mbus_dove_setup_cpu_target,
 	.show_cpu_target     = mvebu_sdram_debug_show_dove,
 };
@@ -578,6 +668,7 @@ static const struct mvebu_mbus_soc_data orion5x_4win_mbus_data = {
 	.num_wins            = 8,
 	.num_remappable_wins = 4,
 	.win_cfg_offset      = orion_mbus_win_offset,
+	.save_cpu_target     = mvebu_mbus_default_save_cpu_target,
 	.setup_cpu_target    = mvebu_mbus_default_setup_cpu_target,
 	.show_cpu_target     = mvebu_sdram_debug_show_orion,
 };
@@ -586,6 +677,7 @@ static const struct mvebu_mbus_soc_data orion5x_2win_mbus_data = {
 	.num_wins            = 8,
 	.num_remappable_wins = 2,
 	.win_cfg_offset      = orion_mbus_win_offset,
+	.save_cpu_target     = mvebu_mbus_default_save_cpu_target,
 	.setup_cpu_target    = mvebu_mbus_default_setup_cpu_target,
 	.show_cpu_target     = mvebu_sdram_debug_show_orion,
 };
@@ -594,6 +686,7 @@ static const struct mvebu_mbus_soc_data mv78xx0_mbus_data = {
 	.num_wins            = 14,
 	.num_remappable_wins = 8,
 	.win_cfg_offset      = mv78xx0_mbus_win_offset,
+	.save_cpu_target     = mvebu_mbus_default_save_cpu_target,
 	.setup_cpu_target    = mvebu_mbus_default_setup_cpu_target,
 	.show_cpu_target     = mvebu_sdram_debug_show_orion,
 };
@@ -698,11 +791,73 @@ static __init int mvebu_mbus_debugfs_init(void)
 }
 fs_initcall(mvebu_mbus_debugfs_init);
 
+static int mvebu_mbus_suspend(void)
+{
+	struct mvebu_mbus_state *s = &mbus_state;
+	int win;
+
+	if (!s->mbusbridge_base)
+		return -ENODEV;
+
+	for (win = 0; win < s->soc->num_wins; win++) {
+		void __iomem *addr = s->mbuswins_base +
+			s->soc->win_cfg_offset(win);
+
+		s->wins[win].base = readl(addr + WIN_BASE_OFF);
+		s->wins[win].ctrl = readl(addr + WIN_CTRL_OFF);
+
+		if (win >= s->soc->num_remappable_wins)
+			continue;
+
+		s->wins[win].remap_lo = readl(addr + WIN_REMAP_LO_OFF);
+		s->wins[win].remap_hi = readl(addr + WIN_REMAP_HI_OFF);
+	}
+
+	s->mbus_bridge_ctrl = readl(s->mbusbridge_base +
+				    MBUS_BRIDGE_CTRL_OFF);
+	s->mbus_bridge_base = readl(s->mbusbridge_base +
+				    MBUS_BRIDGE_BASE_OFF);
+
+	return 0;
+}
+
+static void mvebu_mbus_resume(void)
+{
+	struct mvebu_mbus_state *s = &mbus_state;
+	int win;
+
+	writel(s->mbus_bridge_ctrl,
+	       s->mbusbridge_base + MBUS_BRIDGE_CTRL_OFF);
+	writel(s->mbus_bridge_base,
+	       s->mbusbridge_base + MBUS_BRIDGE_BASE_OFF);
+
+	for (win = 0; win < s->soc->num_wins; win++) {
+		void __iomem *addr = s->mbuswins_base +
+			s->soc->win_cfg_offset(win);
+
+		writel(s->wins[win].base, addr + WIN_BASE_OFF);
+		writel(s->wins[win].ctrl, addr + WIN_CTRL_OFF);
+
+		if (win >= s->soc->num_remappable_wins)
+			continue;
+
+		writel(s->wins[win].remap_lo, addr + WIN_REMAP_LO_OFF);
+		writel(s->wins[win].remap_hi, addr + WIN_REMAP_HI_OFF);
+	}
+}
+
+struct syscore_ops mvebu_mbus_syscore_ops = {
+	.suspend	= mvebu_mbus_suspend,
+	.resume		= mvebu_mbus_resume,
+};
+
 static int __init mvebu_mbus_common_init(struct mvebu_mbus_state *mbus,
 					 phys_addr_t mbuswins_phys_base,
 					 size_t mbuswins_size,
 					 phys_addr_t sdramwins_phys_base,
-					 size_t sdramwins_size)
+					 size_t sdramwins_size,
+					 phys_addr_t mbusbridge_phys_base,
+					 size_t mbusbridge_size)
 {
 	int win;
 
@@ -716,11 +871,26 @@ static int __init mvebu_mbus_common_init(struct mvebu_mbus_state *mbus,
 		return -ENOMEM;
 	}
 
+	mbus->sdramwins_phys_base = sdramwins_phys_base;
+
+	if (mbusbridge_phys_base) {
+		mbus->mbusbridge_base = ioremap(mbusbridge_phys_base,
+						mbusbridge_size);
+		if (!mbus->mbusbridge_base) {
+			iounmap(mbus->sdramwins_base);
+			iounmap(mbus->mbuswins_base);
+			return -ENOMEM;
+		}
+	} else
+		mbus->mbusbridge_base = NULL;
+
 	for (win = 0; win < mbus->soc->num_wins; win++)
 		mvebu_mbus_disable_window(mbus, win);
 
 	mbus->soc->setup_cpu_target(mbus);
 
+	register_syscore_ops(&mvebu_mbus_syscore_ops);
+
 	return 0;
 }
 
@@ -746,7 +916,7 @@ int __init mvebu_mbus_init(const char *soc, phys_addr_t mbuswins_phys_base,
 			mbuswins_phys_base,
 			mbuswins_size,
 			sdramwins_phys_base,
-			sdramwins_size);
+			sdramwins_size, 0, 0);
 }
 
 #ifdef CONFIG_OF
@@ -887,7 +1057,7 @@ static void __init mvebu_mbus_get_pcie_resources(struct device_node *np,
 
 int __init mvebu_mbus_dt_init(bool is_coherent)
 {
-	struct resource mbuswins_res, sdramwins_res;
+	struct resource mbuswins_res, sdramwins_res, mbusbridge_res;
 	struct device_node *np, *controller;
 	const struct of_device_id *of_id;
 	const __be32 *prop;
@@ -923,6 +1093,19 @@ int __init mvebu_mbus_dt_init(bool is_coherent)
 		return -EINVAL;
 	}
 
+	/*
+	 * Set the resource to 0 so that it can be left unmapped by
+	 * mvebu_mbus_common_init() if the DT doesn't carry the
+	 * necessary information. This is needed to preserve backward
+	 * compatibility.
+	 */
+	memset(&mbusbridge_res, 0, sizeof(mbusbridge_res));
+
+	if (mbus_state.soc->has_mbus_bridge) {
+		if (of_address_to_resource(controller, 2, &mbusbridge_res))
+			pr_warn(FW_WARN "deprecated mbus-mvebu Device Tree, suspend/resume will not work\n");
+	}
+
 	mbus_state.hw_io_coherency = is_coherent;
 
 	/* Get optional pcie-{mem,io}-aperture properties */
@@ -933,7 +1116,9 @@ int __init mvebu_mbus_dt_init(bool is_coherent)
 				     mbuswins_res.start,
 				     resource_size(&mbuswins_res),
 				     sdramwins_res.start,
-				     resource_size(&sdramwins_res));
+				     resource_size(&sdramwins_res),
+				     mbusbridge_res.start,
+				     resource_size(&mbusbridge_res));
 	if (ret)
 		return ret;
 
diff --git a/drivers/bus/omap-ocp2scp.c b/drivers/bus/omap-ocp2scp.c
index 5511f9814ddd..723ec06ad2c8 100644
--- a/drivers/bus/omap-ocp2scp.c
+++ b/drivers/bus/omap-ocp2scp.c
@@ -77,7 +77,6 @@ static struct platform_driver omap_ocp2scp_driver = {
 	.remove		= omap_ocp2scp_remove,
 	.driver		= {
 		.name	= "omap-ocp2scp",
-		.owner	= THIS_MODULE,
 		.of_match_table = of_match_ptr(omap_ocp2scp_id_table),
 	},
 };
diff --git a/drivers/bus/omap_l3_noc.c b/drivers/bus/omap_l3_noc.c
index 531ae591783b..029bc73de001 100644
--- a/drivers/bus/omap_l3_noc.c
+++ b/drivers/bus/omap_l3_noc.c
@@ -222,10 +222,14 @@ static irqreturn_t l3_interrupt_handler(int irq, void *_l3)
 			}
 
 			/* Error found so break the for loop */
-			break;
+			return IRQ_HANDLED;
 		}
 	}
-	return IRQ_HANDLED;
+
+	dev_err(l3->dev, "L3 %s IRQ not handled!!\n",
+		inttype ? "debug" : "application");
+
+	return IRQ_NONE;
 }
 
 static const struct of_device_id l3_noc_match[] = {
@@ -296,11 +300,65 @@ static int omap_l3_probe(struct platform_device *pdev)
 	return ret;
 }
 
+#ifdef	CONFIG_PM
+
+/**
+ * l3_resume_noirq() - resume function for l3_noc
+ * @dev:	pointer to l3_noc device structure
+ *
+ * We only have the resume handler only since we
+ * have already maintained the delta register
+ * configuration as part of configuring the system
+ */
+static int l3_resume_noirq(struct device *dev)
+{
+	struct omap_l3 *l3 = dev_get_drvdata(dev);
+	int i;
+	struct l3_flagmux_data *flag_mux;
+	void __iomem *base, *mask_regx = NULL;
+	u32 mask_val;
+
+	for (i = 0; i < l3->num_modules; i++) {
+		base = l3->l3_base[i];
+		flag_mux = l3->l3_flagmux[i];
+		if (!flag_mux->mask_app_bits && !flag_mux->mask_dbg_bits)
+			continue;
+
+		mask_regx = base + flag_mux->offset + L3_FLAGMUX_MASK0 +
+			   (L3_APPLICATION_ERROR << 3);
+		mask_val = readl_relaxed(mask_regx);
+		mask_val &= ~(flag_mux->mask_app_bits);
+
+		writel_relaxed(mask_val, mask_regx);
+		mask_regx = base + flag_mux->offset + L3_FLAGMUX_MASK0 +
+			   (L3_DEBUG_ERROR << 3);
+		mask_val = readl_relaxed(mask_regx);
+		mask_val &= ~(flag_mux->mask_dbg_bits);
+
+		writel_relaxed(mask_val, mask_regx);
+	}
+
+	/* Dummy read to force OCP barrier */
+	if (mask_regx)
+		(void)readl(mask_regx);
+
+	return 0;
+}
+
+static const struct dev_pm_ops l3_dev_pm_ops = {
+	.resume_noirq		= l3_resume_noirq,
+};
+
+#define L3_DEV_PM_OPS (&l3_dev_pm_ops)
+#else
+#define L3_DEV_PM_OPS NULL
+#endif
+
 static struct platform_driver omap_l3_driver = {
 	.probe		= omap_l3_probe,
 	.driver		= {
 		.name		= "omap_l3_noc",
-		.owner		= THIS_MODULE,
+		.pm		= L3_DEV_PM_OPS,
 		.of_match_table = of_match_ptr(l3_noc_match),
 	},
 };
diff --git a/drivers/bus/omap_l3_smx.c b/drivers/bus/omap_l3_smx.c
index acc216491b8a..597fdaee7315 100644
--- a/drivers/bus/omap_l3_smx.c
+++ b/drivers/bus/omap_l3_smx.c
@@ -27,6 +27,10 @@
 #include <linux/platform_device.h>
 #include <linux/interrupt.h>
 #include <linux/io.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+
 #include "omap_l3_smx.h"
 
 static inline u64 omap3_l3_readll(void __iomem *base, u16 reg)
@@ -211,7 +215,17 @@ static irqreturn_t omap3_l3_app_irq(int irq, void *_l3)
 	return ret;
 }
 
-static int __init omap3_l3_probe(struct platform_device *pdev)
+#if IS_BUILTIN(CONFIG_OF)
+static const struct of_device_id omap3_l3_match[] = {
+	{
+		.compatible = "ti,omap3-l3-smx",
+	},
+	{ },
+};
+MODULE_DEVICE_TABLE(of, omap3_l3_match);
+#endif
+
+static int omap3_l3_probe(struct platform_device *pdev)
 {
 	struct omap3_l3 *l3;
 	struct resource *res;
@@ -265,7 +279,7 @@ err0:
 	return ret;
 }
 
-static int __exit omap3_l3_remove(struct platform_device *pdev)
+static int omap3_l3_remove(struct platform_device *pdev)
 {
 	struct omap3_l3         *l3 = platform_get_drvdata(pdev);
 
@@ -278,15 +292,17 @@ static int __exit omap3_l3_remove(struct platform_device *pdev)
 }
 
 static struct platform_driver omap3_l3_driver = {
-	.remove         = __exit_p(omap3_l3_remove),
+	.probe		= omap3_l3_probe,
+	.remove         = omap3_l3_remove,
 	.driver         = {
-	.name   = "omap_l3_smx",
+		.name   = "omap_l3_smx",
+		.of_match_table = of_match_ptr(omap3_l3_match),
 	},
 };
 
 static int __init omap3_l3_init(void)
 {
-	return platform_driver_probe(&omap3_l3_driver, omap3_l3_probe);
+	return platform_driver_register(&omap3_l3_driver);
 }
 postcore_initcall_sync(omap3_l3_init);
 
diff --git a/drivers/cdrom/cdrom.c b/drivers/cdrom/cdrom.c
index 898b84bba28a..5d28a45d2960 100644
--- a/drivers/cdrom/cdrom.c
+++ b/drivers/cdrom/cdrom.c
@@ -2180,8 +2180,8 @@ static int cdrom_read_cdda_bpc(struct cdrom_device_info *cdi, __u8 __user *ubuf,
 		len = nr * CD_FRAMESIZE_RAW;
 
 		rq = blk_get_request(q, READ, GFP_KERNEL);
-		if (!rq) {
-			ret = -ENOMEM;
+		if (IS_ERR(rq)) {
+			ret = PTR_ERR(rq);
 			break;
 		}
 		blk_rq_set_block_pc(rq);
diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig
index 6e9f74a5c095..efefd12a0f7b 100644
--- a/drivers/char/Kconfig
+++ b/drivers/char/Kconfig
@@ -600,5 +600,7 @@ config TILE_SROM
 	  device appear much like a simple EEPROM, and knows
 	  how to partition a single ROM for multiple purposes.
 
+source "drivers/char/xillybus/Kconfig"
+
 endmenu
 
diff --git a/drivers/char/Makefile b/drivers/char/Makefile
index a324f9303e36..d06cde26031b 100644
--- a/drivers/char/Makefile
+++ b/drivers/char/Makefile
@@ -61,3 +61,4 @@ obj-$(CONFIG_JS_RTC)		+= js-rtc.o
 js-rtc-y = rtc.o
 
 obj-$(CONFIG_TILE_SROM)		+= tile-srom.o
+obj-$(CONFIG_XILLYBUS)		+= xillybus/
diff --git a/drivers/char/agp/ali-agp.c b/drivers/char/agp/ali-agp.c
index 19db03667650..dcbbb4ea3cc1 100644
--- a/drivers/char/agp/ali-agp.c
+++ b/drivers/char/agp/ali-agp.c
@@ -417,6 +417,6 @@ static void __exit agp_ali_cleanup(void)
 module_init(agp_ali_init);
 module_exit(agp_ali_cleanup);
 
-MODULE_AUTHOR("Dave Jones <davej@redhat.com>");
+MODULE_AUTHOR("Dave Jones");
 MODULE_LICENSE("GPL and additional rights");
 
diff --git a/drivers/char/agp/amd64-agp.c b/drivers/char/agp/amd64-agp.c
index 3b47ed0310e1..0ef350010766 100644
--- a/drivers/char/agp/amd64-agp.c
+++ b/drivers/char/agp/amd64-agp.c
@@ -813,6 +813,6 @@ static void __exit agp_amd64_cleanup(void)
 module_init(agp_amd64_mod_init);
 module_exit(agp_amd64_cleanup);
 
-MODULE_AUTHOR("Dave Jones <davej@redhat.com>, Andi Kleen");
+MODULE_AUTHOR("Dave Jones, Andi Kleen");
 module_param(agp_try_unsupported, bool, 0);
 MODULE_LICENSE("GPL");
diff --git a/drivers/char/agp/ati-agp.c b/drivers/char/agp/ati-agp.c
index 18a7a6baa304..75a9786a77e6 100644
--- a/drivers/char/agp/ati-agp.c
+++ b/drivers/char/agp/ati-agp.c
@@ -579,6 +579,6 @@ static void __exit agp_ati_cleanup(void)
 module_init(agp_ati_init);
 module_exit(agp_ati_cleanup);
 
-MODULE_AUTHOR("Dave Jones <davej@redhat.com>");
+MODULE_AUTHOR("Dave Jones");
 MODULE_LICENSE("GPL and additional rights");
 
diff --git a/drivers/char/agp/backend.c b/drivers/char/agp/backend.c
index 317c28ce8328..38ffb281df97 100644
--- a/drivers/char/agp/backend.c
+++ b/drivers/char/agp/backend.c
@@ -356,7 +356,7 @@ static __init int agp_setup(char *s)
 __setup("agp=", agp_setup);
 #endif
 
-MODULE_AUTHOR("Dave Jones <davej@redhat.com>");
+MODULE_AUTHOR("Dave Jones, Jeff Hartmann");
 MODULE_DESCRIPTION("AGP GART driver");
 MODULE_LICENSE("GPL and additional rights");
 MODULE_ALIAS_MISCDEV(AGPGART_MINOR);
diff --git a/drivers/char/agp/intel-agp.c b/drivers/char/agp/intel-agp.c
index f9b9ca5d31b7..0a21daed5b62 100644
--- a/drivers/char/agp/intel-agp.c
+++ b/drivers/char/agp/intel-agp.c
@@ -920,5 +920,5 @@ static void __exit agp_intel_cleanup(void)
 module_init(agp_intel_init);
 module_exit(agp_intel_cleanup);
 
-MODULE_AUTHOR("Dave Jones <davej@redhat.com>");
+MODULE_AUTHOR("Dave Jones, Various @Intel");
 MODULE_LICENSE("GPL and additional rights");
diff --git a/drivers/char/agp/intel-gtt.c b/drivers/char/agp/intel-gtt.c
index 9a024f899dd4..92aa43fa8d70 100644
--- a/drivers/char/agp/intel-gtt.c
+++ b/drivers/char/agp/intel-gtt.c
@@ -153,7 +153,6 @@ static struct page *i8xx_alloc_pages(void)
 		__free_pages(page, 2);
 		return NULL;
 	}
-	get_page(page);
 	atomic_inc(&agp_bridge->current_memory_agp);
 	return page;
 }
@@ -164,7 +163,6 @@ static void i8xx_destroy_pages(struct page *page)
 		return;
 
 	set_pages_wb(page, 4);
-	put_page(page);
 	__free_pages(page, 2);
 	atomic_dec(&agp_bridge->current_memory_agp);
 }
@@ -300,7 +298,6 @@ static int intel_gtt_setup_scratch_page(void)
 	page = alloc_page(GFP_KERNEL | GFP_DMA32 | __GFP_ZERO);
 	if (page == NULL)
 		return -ENOMEM;
-	get_page(page);
 	set_pages_uc(page, 1);
 
 	if (intel_private.needs_dmar) {
@@ -560,7 +557,6 @@ static void intel_gtt_teardown_scratch_page(void)
 	set_pages_wb(intel_private.scratch_page, 1);
 	pci_unmap_page(intel_private.pcidev, intel_private.scratch_page_dma,
 		       PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
-	put_page(intel_private.scratch_page);
 	__free_page(intel_private.scratch_page);
 }
 
@@ -1442,5 +1438,5 @@ void intel_gmch_remove(void)
 }
 EXPORT_SYMBOL(intel_gmch_remove);
 
-MODULE_AUTHOR("Dave Jones <davej@redhat.com>");
+MODULE_AUTHOR("Dave Jones, Various @Intel");
 MODULE_LICENSE("GPL and additional rights");
diff --git a/drivers/char/agp/nvidia-agp.c b/drivers/char/agp/nvidia-agp.c
index a1861b75eb31..6c8d39cb566e 100644
--- a/drivers/char/agp/nvidia-agp.c
+++ b/drivers/char/agp/nvidia-agp.c
@@ -1,7 +1,7 @@
 /*
  * Nvidia AGPGART routines.
  * Based upon a 2.4 agpgart diff by the folks from NVIDIA, and hacked up
- * to work in 2.5 by Dave Jones <davej@redhat.com>
+ * to work in 2.5 by Dave Jones.
  */
 
 #include <linux/module.h>
diff --git a/drivers/char/agp/via-agp.c b/drivers/char/agp/via-agp.c
index 228f20cddc05..a4961d35e940 100644
--- a/drivers/char/agp/via-agp.c
+++ b/drivers/char/agp/via-agp.c
@@ -595,4 +595,4 @@ module_init(agp_via_init);
 module_exit(agp_via_cleanup);
 
 MODULE_LICENSE("GPL");
-MODULE_AUTHOR("Dave Jones <davej@redhat.com>");
+MODULE_AUTHOR("Dave Jones");
diff --git a/drivers/char/hangcheck-timer.c b/drivers/char/hangcheck-timer.c
index ebc4c73d8ca4..a7c5c59675f0 100644
--- a/drivers/char/hangcheck-timer.c
+++ b/drivers/char/hangcheck-timer.c
@@ -168,8 +168,8 @@ static int __init hangcheck_init(void)
 	printk("Hangcheck: starting hangcheck timer %s (tick is %d seconds, margin is %d seconds).\n",
 	       VERSION_STR, hangcheck_tick, hangcheck_margin);
 	hangcheck_tsc_margin =
-		(unsigned long long)(hangcheck_margin + hangcheck_tick);
-	hangcheck_tsc_margin *= (unsigned long long)TIMER_FREQ;
+		(unsigned long long)hangcheck_margin + hangcheck_tick;
+	hangcheck_tsc_margin *= TIMER_FREQ;
 
 	hangcheck_tsc = ktime_get_ns();
 	mod_timer(&hangcheck_ticktock, jiffies + (hangcheck_tick*HZ));
diff --git a/drivers/char/hw_random/Kconfig b/drivers/char/hw_random/Kconfig
index 836b061ced35..de57b38809c7 100644
--- a/drivers/char/hw_random/Kconfig
+++ b/drivers/char/hw_random/Kconfig
@@ -64,7 +64,7 @@ config HW_RANDOM_AMD
 
 config HW_RANDOM_ATMEL
 	tristate "Atmel Random Number Generator support"
-	depends on ARCH_AT91 && HAVE_CLK
+	depends on ARCH_AT91 && HAVE_CLK && OF
 	default HW_RANDOM
 	---help---
 	  This driver provides kernel-side support for the Random Number
@@ -333,6 +333,19 @@ config HW_RANDOM_MSM
 
 	  If unsure, say Y.
 
+config HW_RANDOM_XGENE
+	tristate "APM X-Gene True Random Number Generator (TRNG) support"
+	depends on HW_RANDOM && ARCH_XGENE
+	default HW_RANDOM
+	---help---
+	  This driver provides kernel-side support for the Random Number
+	  Generator hardware found on APM X-Gene SoC.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called xgene_rng.
+
+	  If unsure, say Y.
+
 endif # HW_RANDOM
 
 config UML_RANDOM
diff --git a/drivers/char/hw_random/Makefile b/drivers/char/hw_random/Makefile
index 199ed283e149..0b4cd57f4e24 100644
--- a/drivers/char/hw_random/Makefile
+++ b/drivers/char/hw_random/Makefile
@@ -29,3 +29,4 @@ obj-$(CONFIG_HW_RANDOM_EXYNOS)	+= exynos-rng.o
 obj-$(CONFIG_HW_RANDOM_TPM) += tpm-rng.o
 obj-$(CONFIG_HW_RANDOM_BCM2835) += bcm2835-rng.o
 obj-$(CONFIG_HW_RANDOM_MSM) += msm-rng.o
+obj-$(CONFIG_HW_RANDOM_XGENE) += xgene-rng.o
diff --git a/drivers/char/hw_random/amd-rng.c b/drivers/char/hw_random/amd-rng.c
index c6af038682f1..48f6a83cdd61 100644
--- a/drivers/char/hw_random/amd-rng.c
+++ b/drivers/char/hw_random/amd-rng.c
@@ -142,10 +142,10 @@ found:
 	amd_rng.priv = (unsigned long)pmbase;
 	amd_pdev = pdev;
 
-	printk(KERN_INFO "AMD768 RNG detected\n");
+	pr_info("AMD768 RNG detected\n");
 	err = hwrng_register(&amd_rng);
 	if (err) {
-		printk(KERN_ERR PFX "RNG registering failed (%d)\n",
+		pr_err(PFX "RNG registering failed (%d)\n",
 		       err);
 		release_region(pmbase + 0xF0, 8);
 		goto out;
diff --git a/drivers/char/hw_random/atmel-rng.c b/drivers/char/hw_random/atmel-rng.c
index 851bc7e20ad2..0fcc9e69a346 100644
--- a/drivers/char/hw_random/atmel-rng.c
+++ b/drivers/char/hw_random/atmel-rng.c
@@ -67,7 +67,7 @@ static int atmel_trng_probe(struct platform_device *pdev)
 	if (IS_ERR(trng->clk))
 		return PTR_ERR(trng->clk);
 
-	ret = clk_enable(trng->clk);
+	ret = clk_prepare_enable(trng->clk);
 	if (ret)
 		return ret;
 
@@ -95,7 +95,7 @@ static int atmel_trng_remove(struct platform_device *pdev)
 	hwrng_unregister(&trng->rng);
 
 	writel(TRNG_KEY, trng->base + TRNG_CR);
-	clk_disable(trng->clk);
+	clk_disable_unprepare(trng->clk);
 
 	return 0;
 }
@@ -105,7 +105,7 @@ static int atmel_trng_suspend(struct device *dev)
 {
 	struct atmel_trng *trng = dev_get_drvdata(dev);
 
-	clk_disable(trng->clk);
+	clk_disable_unprepare(trng->clk);
 
 	return 0;
 }
@@ -114,7 +114,7 @@ static int atmel_trng_resume(struct device *dev)
 {
 	struct atmel_trng *trng = dev_get_drvdata(dev);
 
-	return clk_enable(trng->clk);
+	return clk_prepare_enable(trng->clk);
 }
 
 static const struct dev_pm_ops atmel_trng_pm_ops = {
@@ -123,15 +123,21 @@ static const struct dev_pm_ops atmel_trng_pm_ops = {
 };
 #endif /* CONFIG_PM */
 
+static const struct of_device_id atmel_trng_dt_ids[] = {
+	{ .compatible = "atmel,at91sam9g45-trng" },
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, atmel_trng_dt_ids);
+
 static struct platform_driver atmel_trng_driver = {
 	.probe		= atmel_trng_probe,
 	.remove		= atmel_trng_remove,
 	.driver		= {
 		.name	= "atmel-trng",
-		.owner	= THIS_MODULE,
 #ifdef CONFIG_PM
 		.pm	= &atmel_trng_pm_ops,
 #endif /* CONFIG_PM */
+		.of_match_table = atmel_trng_dt_ids,
 	},
 };
 
diff --git a/drivers/char/hw_random/bcm2835-rng.c b/drivers/char/hw_random/bcm2835-rng.c
index e900961cdd2e..7192ec25f667 100644
--- a/drivers/char/hw_random/bcm2835-rng.c
+++ b/drivers/char/hw_random/bcm2835-rng.c
@@ -99,7 +99,6 @@ MODULE_DEVICE_TABLE(of, bcm2835_rng_of_match);
 static struct platform_driver bcm2835_rng_driver = {
 	.driver = {
 		.name = "bcm2835-rng",
-		.owner = THIS_MODULE,
 		.of_match_table = bcm2835_rng_of_match,
 	},
 	.probe		= bcm2835_rng_probe,
diff --git a/drivers/char/hw_random/bcm63xx-rng.c b/drivers/char/hw_random/bcm63xx-rng.c
index 36581ea562cb..ba6a65ac023b 100644
--- a/drivers/char/hw_random/bcm63xx-rng.c
+++ b/drivers/char/hw_random/bcm63xx-rng.c
@@ -162,7 +162,6 @@ static struct platform_driver bcm63xx_rng_driver = {
 	.remove		= bcm63xx_rng_remove,
 	.driver		= {
 		.name	= "bcm63xx-rng",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/char/hw_random/core.c b/drivers/char/hw_random/core.c
index aa30a25c8d49..1500cfd799a7 100644
--- a/drivers/char/hw_random/core.c
+++ b/drivers/char/hw_random/core.c
@@ -281,7 +281,6 @@ static ssize_t hwrng_attr_available_show(struct device *dev,
 					 char *buf)
 {
 	int err;
-	ssize_t ret = 0;
 	struct hwrng *rng;
 
 	err = mutex_lock_interruptible(&rng_mutex);
@@ -289,16 +288,13 @@ static ssize_t hwrng_attr_available_show(struct device *dev,
 		return -ERESTARTSYS;
 	buf[0] = '\0';
 	list_for_each_entry(rng, &rng_list, list) {
-		strncat(buf, rng->name, PAGE_SIZE - ret - 1);
-		ret += strlen(rng->name);
-		strncat(buf, " ", PAGE_SIZE - ret - 1);
-		ret++;
+		strlcat(buf, rng->name, PAGE_SIZE);
+		strlcat(buf, " ", PAGE_SIZE);
 	}
-	strncat(buf, "\n", PAGE_SIZE - ret - 1);
-	ret++;
+	strlcat(buf, "\n", PAGE_SIZE);
 	mutex_unlock(&rng_mutex);
 
-	return ret;
+	return strlen(buf);
 }
 
 static DEVICE_ATTR(rng_current, S_IRUGO | S_IWUSR,
diff --git a/drivers/char/hw_random/exynos-rng.c b/drivers/char/hw_random/exynos-rng.c
index 9f8277cc44b4..fed0830bf724 100644
--- a/drivers/char/hw_random/exynos-rng.c
+++ b/drivers/char/hw_random/exynos-rng.c
@@ -143,7 +143,7 @@ static int exynos_rng_remove(struct platform_device *pdev)
 	return 0;
 }
 
-#if defined(CONFIG_PM_SLEEP) || defined(CONFIG_PM_RUNTIME)
+#ifdef CONFIG_PM
 static int exynos_rng_runtime_suspend(struct device *dev)
 {
 	struct platform_device *pdev = to_platform_device(dev);
@@ -169,7 +169,6 @@ static UNIVERSAL_DEV_PM_OPS(exynos_rng_pm_ops, exynos_rng_runtime_suspend,
 static struct platform_driver exynos_rng_driver = {
 	.driver		= {
 		.name	= "exynos-rng",
-		.owner	= THIS_MODULE,
 		.pm	= &exynos_rng_pm_ops,
 	},
 	.probe		= exynos_rng_probe,
diff --git a/drivers/char/hw_random/geode-rng.c b/drivers/char/hw_random/geode-rng.c
index 4c4d4e140f98..0d0579fe465e 100644
--- a/drivers/char/hw_random/geode-rng.c
+++ b/drivers/char/hw_random/geode-rng.c
@@ -109,10 +109,10 @@ found:
 		goto out;
 	geode_rng.priv = (unsigned long)mem;
 
-	printk(KERN_INFO "AMD Geode RNG detected\n");
+	pr_info("AMD Geode RNG detected\n");
 	err = hwrng_register(&geode_rng);
 	if (err) {
-		printk(KERN_ERR PFX "RNG registering failed (%d)\n",
+		pr_err(PFX "RNG registering failed (%d)\n",
 		       err);
 		goto err_unmap;
 	}
diff --git a/drivers/char/hw_random/intel-rng.c b/drivers/char/hw_random/intel-rng.c
index 86fe45c19968..290c880266bf 100644
--- a/drivers/char/hw_random/intel-rng.c
+++ b/drivers/char/hw_random/intel-rng.c
@@ -199,7 +199,7 @@ static int intel_rng_init(struct hwrng *rng)
 	if ((hw_status & INTEL_RNG_ENABLED) == 0)
 		hw_status = hwstatus_set(mem, hw_status | INTEL_RNG_ENABLED);
 	if ((hw_status & INTEL_RNG_ENABLED) == 0) {
-		printk(KERN_ERR PFX "cannot enable RNG, aborting\n");
+		pr_err(PFX "cannot enable RNG, aborting\n");
 		goto out;
 	}
 	err = 0;
@@ -216,7 +216,7 @@ static void intel_rng_cleanup(struct hwrng *rng)
 	if (hw_status & INTEL_RNG_ENABLED)
 		hwstatus_set(mem, hw_status & ~INTEL_RNG_ENABLED);
 	else
-		printk(KERN_WARNING PFX "unusual: RNG already disabled\n");
+		pr_warn(PFX "unusual: RNG already disabled\n");
 }
 
 
@@ -274,7 +274,7 @@ static int __init intel_rng_hw_init(void *_intel_rng_hw)
 	if (mfc != INTEL_FWH_MANUFACTURER_CODE ||
 	    (dvc != INTEL_FWH_DEVICE_CODE_8M &&
 	     dvc != INTEL_FWH_DEVICE_CODE_4M)) {
-		printk(KERN_NOTICE PFX "FWH not detected\n");
+		pr_notice(PFX "FWH not detected\n");
 		return -ENODEV;
 	}
 
@@ -306,7 +306,6 @@ static int __init intel_init_hw_struct(struct intel_rng_hw *intel_rng_hw,
 	     (BIOS_CNTL_LOCK_ENABLE_MASK|BIOS_CNTL_WRITE_ENABLE_MASK))
 	    == BIOS_CNTL_LOCK_ENABLE_MASK) {
 		static __initdata /*const*/ char warning[] =
-			KERN_WARNING
 PFX "Firmware space is locked read-only. If you can't or\n"
 PFX "don't want to disable this in firmware setup, and if\n"
 PFX "you are certain that your system has a functional\n"
@@ -314,7 +313,7 @@ PFX "RNG, try using the 'no_fwh_detect' option.\n";
 
 		if (no_fwh_detect)
 			return -ENODEV;
-		printk(warning);
+		pr_warn("%s", warning);
 		return -EBUSY;
 	}
 
@@ -392,10 +391,10 @@ fwh_done:
 		goto out;
 	}
 
-	printk(KERN_INFO "Intel 82802 RNG detected\n");
+	pr_info("Intel 82802 RNG detected\n");
 	err = hwrng_register(&intel_rng);
 	if (err) {
-		printk(KERN_ERR PFX "RNG registering failed (%d)\n",
+		pr_err(PFX "RNG registering failed (%d)\n",
 		       err);
 		iounmap(mem);
 	}
diff --git a/drivers/char/hw_random/msm-rng.c b/drivers/char/hw_random/msm-rng.c
index 148521e51dc6..cea1c703d62f 100644
--- a/drivers/char/hw_random/msm-rng.c
+++ b/drivers/char/hw_random/msm-rng.c
@@ -185,7 +185,6 @@ static struct platform_driver msm_rng_driver = {
 	.remove = msm_rng_remove,
 	.driver = {
 		.name = KBUILD_MODNAME,
-		.owner = THIS_MODULE,
 		.of_match_table = of_match_ptr(msm_rng_of_match),
 	}
 };
diff --git a/drivers/char/hw_random/mxc-rnga.c b/drivers/char/hw_random/mxc-rnga.c
index 6a86b6f56af2..6cbb72ec6013 100644
--- a/drivers/char/hw_random/mxc-rnga.c
+++ b/drivers/char/hw_random/mxc-rnga.c
@@ -206,7 +206,6 @@ static int __exit mxc_rnga_remove(struct platform_device *pdev)
 static struct platform_driver mxc_rnga_driver = {
 	.driver = {
 		   .name = "mxc_rnga",
-		   .owner = THIS_MODULE,
 		   },
 	.remove = __exit_p(mxc_rnga_remove),
 };
diff --git a/drivers/char/hw_random/n2-drv.c b/drivers/char/hw_random/n2-drv.c
index 292a5889f675..843d6f6aee7a 100644
--- a/drivers/char/hw_random/n2-drv.c
+++ b/drivers/char/hw_random/n2-drv.c
@@ -750,7 +750,6 @@ MODULE_DEVICE_TABLE(of, n2rng_match);
 static struct platform_driver n2rng_driver = {
 	.driver = {
 		.name = "n2rng",
-		.owner = THIS_MODULE,
 		.of_match_table = n2rng_match,
 	},
 	.probe		= n2rng_probe,
diff --git a/drivers/char/hw_random/octeon-rng.c b/drivers/char/hw_random/octeon-rng.c
index b5cc3420c659..be1c3f607398 100644
--- a/drivers/char/hw_random/octeon-rng.c
+++ b/drivers/char/hw_random/octeon-rng.c
@@ -117,7 +117,6 @@ static int __exit octeon_rng_remove(struct platform_device *pdev)
 static struct platform_driver octeon_rng_driver = {
 	.driver = {
 		.name		= "octeon_rng",
-		.owner		= THIS_MODULE,
 	},
 	.probe		= octeon_rng_probe,
 	.remove		= __exit_p(octeon_rng_remove),
diff --git a/drivers/char/hw_random/omap-rng.c b/drivers/char/hw_random/omap-rng.c
index f66ea258382f..d14dcf788f17 100644
--- a/drivers/char/hw_random/omap-rng.c
+++ b/drivers/char/hw_random/omap-rng.c
@@ -456,7 +456,6 @@ static SIMPLE_DEV_PM_OPS(omap_rng_pm, omap_rng_suspend, omap_rng_resume);
 static struct platform_driver omap_rng_driver = {
 	.driver = {
 		.name		= "omap_rng",
-		.owner		= THIS_MODULE,
 		.pm		= OMAP_RNG_PM,
 		.of_match_table = of_match_ptr(omap_rng_of_match),
 	},
diff --git a/drivers/char/hw_random/omap3-rom-rng.c b/drivers/char/hw_random/omap3-rom-rng.c
index 6f2eaffed623..a405cdcd8dd2 100644
--- a/drivers/char/hw_random/omap3-rom-rng.c
+++ b/drivers/char/hw_random/omap3-rom-rng.c
@@ -126,7 +126,6 @@ static int omap3_rom_rng_remove(struct platform_device *pdev)
 static struct platform_driver omap3_rom_rng_driver = {
 	.driver = {
 		.name		= "omap3-rom-rng",
-		.owner		= THIS_MODULE,
 	},
 	.probe		= omap3_rom_rng_probe,
 	.remove		= omap3_rom_rng_remove,
diff --git a/drivers/char/hw_random/pasemi-rng.c b/drivers/char/hw_random/pasemi-rng.c
index c66279bb6ef3..3eb7bdd7f93b 100644
--- a/drivers/char/hw_random/pasemi-rng.c
+++ b/drivers/char/hw_random/pasemi-rng.c
@@ -113,7 +113,7 @@ static int rng_probe(struct platform_device *ofdev)
 
 	pasemi_rng.priv = (unsigned long)rng_regs;
 
-	printk(KERN_INFO "Registering PA Semi RNG\n");
+	pr_info("Registering PA Semi RNG\n");
 
 	err = hwrng_register(&pasemi_rng);
 
@@ -142,7 +142,6 @@ static struct of_device_id rng_match[] = {
 static struct platform_driver rng_driver = {
 	.driver = {
 		.name = "pasemi-rng",
-		.owner = THIS_MODULE,
 		.of_match_table = rng_match,
 	},
 	.probe		= rng_probe,
diff --git a/drivers/char/hw_random/ppc4xx-rng.c b/drivers/char/hw_random/ppc4xx-rng.c
index 521f76b0934b..c85d31a5f9e3 100644
--- a/drivers/char/hw_random/ppc4xx-rng.c
+++ b/drivers/char/hw_random/ppc4xx-rng.c
@@ -133,7 +133,6 @@ static struct of_device_id ppc4xx_rng_match[] = {
 static struct platform_driver ppc4xx_rng_driver = {
 	.driver = {
 		.name = MODULE_NAME,
-		.owner = THIS_MODULE,
 		.of_match_table = ppc4xx_rng_match,
 	},
 	.probe = ppc4xx_rng_probe,
diff --git a/drivers/char/hw_random/pseries-rng.c b/drivers/char/hw_random/pseries-rng.c
index ab7ffdec0ec3..bcf86f91800a 100644
--- a/drivers/char/hw_random/pseries-rng.c
+++ b/drivers/char/hw_random/pseries-rng.c
@@ -25,18 +25,21 @@
 #include <asm/vio.h>
 
 
-static int pseries_rng_data_read(struct hwrng *rng, u32 *data)
+static int pseries_rng_read(struct hwrng *rng, void *data, size_t max, bool wait)
 {
+	u64 buffer[PLPAR_HCALL_BUFSIZE];
+	size_t size = max < 8 ? max : 8;
 	int rc;
 
-	rc = plpar_hcall(H_RANDOM, (unsigned long *)data);
+	rc = plpar_hcall(H_RANDOM, (unsigned long *)buffer);
 	if (rc != H_SUCCESS) {
 		pr_err_ratelimited("H_RANDOM call failed %d\n", rc);
 		return -EIO;
 	}
+	memcpy(data, buffer, size);
 
 	/* The hypervisor interface returns 64 bits */
-	return 8;
+	return size;
 }
 
 /**
@@ -55,7 +58,7 @@ static unsigned long pseries_rng_get_desired_dma(struct vio_dev *vdev)
 
 static struct hwrng pseries_rng = {
 	.name		= KBUILD_MODNAME,
-	.data_read	= pseries_rng_data_read,
+	.read		= pseries_rng_read,
 };
 
 static int __init pseries_rng_probe(struct vio_dev *dev,
@@ -86,7 +89,7 @@ static struct vio_driver pseries_rng_driver = {
 
 static int __init rng_init(void)
 {
-	printk(KERN_INFO "Registering IBM pSeries RNG driver\n");
+	pr_info("Registering IBM pSeries RNG driver\n");
 	return vio_register_driver(&pseries_rng_driver);
 }
 
diff --git a/drivers/char/hw_random/timeriomem-rng.c b/drivers/char/hw_random/timeriomem-rng.c
index b6ab9ac3f34d..cf37db263ecd 100644
--- a/drivers/char/hw_random/timeriomem-rng.c
+++ b/drivers/char/hw_random/timeriomem-rng.c
@@ -200,7 +200,6 @@ MODULE_DEVICE_TABLE(of, timeriomem_rng_match);
 static struct platform_driver timeriomem_rng_driver = {
 	.driver = {
 		.name		= "timeriomem_rng",
-		.owner		= THIS_MODULE,
 		.of_match_table	= timeriomem_rng_match,
 	},
 	.probe		= timeriomem_rng_probe,
diff --git a/drivers/char/hw_random/tx4939-rng.c b/drivers/char/hw_random/tx4939-rng.c
index 09c5fbea2b93..a7b694913416 100644
--- a/drivers/char/hw_random/tx4939-rng.c
+++ b/drivers/char/hw_random/tx4939-rng.c
@@ -158,7 +158,6 @@ static int __exit tx4939_rng_remove(struct platform_device *dev)
 static struct platform_driver tx4939_rng_driver = {
 	.driver		= {
 		.name	= "tx4939-rng",
-		.owner	= THIS_MODULE,
 	},
 	.remove = tx4939_rng_remove,
 };
diff --git a/drivers/char/hw_random/via-rng.c b/drivers/char/hw_random/via-rng.c
index de5a6dcfb3e2..a3bebef255ad 100644
--- a/drivers/char/hw_random/via-rng.c
+++ b/drivers/char/hw_random/via-rng.c
@@ -141,7 +141,7 @@ static int via_rng_init(struct hwrng *rng)
 	 * register */
 	if ((c->x86 == 6) && (c->x86_model >= 0x0f)) {
 		if (!cpu_has_xstore_enabled) {
-			printk(KERN_ERR PFX "can't enable hardware RNG "
+			pr_err(PFX "can't enable hardware RNG "
 				"if XSTORE is not enabled\n");
 			return -ENODEV;
 		}
@@ -180,7 +180,7 @@ static int via_rng_init(struct hwrng *rng)
 	   unneeded */
 	rdmsr(MSR_VIA_RNG, lo, hi);
 	if ((lo & VIA_RNG_ENABLE) == 0) {
-		printk(KERN_ERR PFX "cannot enable VIA C3 RNG, aborting\n");
+		pr_err(PFX "cannot enable VIA C3 RNG, aborting\n");
 		return -ENODEV;
 	}
 
@@ -202,10 +202,10 @@ static int __init mod_init(void)
 
 	if (!cpu_has_xstore)
 		return -ENODEV;
-	printk(KERN_INFO "VIA RNG detected\n");
+	pr_info("VIA RNG detected\n");
 	err = hwrng_register(&via_rng);
 	if (err) {
-		printk(KERN_ERR PFX "RNG registering failed (%d)\n",
+		pr_err(PFX "RNG registering failed (%d)\n",
 		       err);
 		goto out;
 	}
diff --git a/drivers/char/hw_random/virtio-rng.c b/drivers/char/hw_random/virtio-rng.c
index 132c9ccfdc62..72295ea2fd1c 100644
--- a/drivers/char/hw_random/virtio-rng.c
+++ b/drivers/char/hw_random/virtio-rng.c
@@ -109,8 +109,8 @@ static int probe_common(struct virtio_device *vdev)
 
 	vi->index = index = ida_simple_get(&rng_index_ida, 0, 0, GFP_KERNEL);
 	if (index < 0) {
-		kfree(vi);
-		return index;
+		err = index;
+		goto err_ida;
 	}
 	sprintf(vi->name, "virtio_rng.%d", index);
 	init_completion(&vi->have_data);
@@ -128,13 +128,16 @@ static int probe_common(struct virtio_device *vdev)
 	vi->vq = virtio_find_single_vq(vdev, random_recv_done, "input");
 	if (IS_ERR(vi->vq)) {
 		err = PTR_ERR(vi->vq);
-		vi->vq = NULL;
-		kfree(vi);
-		ida_simple_remove(&rng_index_ida, index);
-		return err;
+		goto err_find;
 	}
 
 	return 0;
+
+err_find:
+	ida_simple_remove(&rng_index_ida, index);
+err_ida:
+	kfree(vi);
+	return err;
 }
 
 static void remove_common(struct virtio_device *vdev)
diff --git a/drivers/char/hw_random/xgene-rng.c b/drivers/char/hw_random/xgene-rng.c
new file mode 100644
index 000000000000..23caa05380a8
--- /dev/null
+++ b/drivers/char/hw_random/xgene-rng.c
@@ -0,0 +1,423 @@
+/*
+ * APM X-Gene SoC RNG Driver
+ *
+ * Copyright (c) 2014, Applied Micro Circuits Corporation
+ * Author: Rameshwar Prasad Sahu <rsahu@apm.com>
+ *	   Shamal Winchurkar <swinchurkar@apm.com>
+ *	   Feng Kan <fkan@apm.com>
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ *
+ */
+
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/hw_random.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/of_platform.h>
+#include <linux/of_irq.h>
+#include <linux/of_address.h>
+#include <linux/timer.h>
+
+#define RNG_MAX_DATUM			4
+#define MAX_TRY				100
+#define XGENE_RNG_RETRY_COUNT		20
+#define XGENE_RNG_RETRY_INTERVAL	10
+
+/* RNG  Registers */
+#define RNG_INOUT_0			0x00
+#define RNG_INTR_STS_ACK		0x10
+#define RNG_CONTROL			0x14
+#define RNG_CONFIG			0x18
+#define RNG_ALARMCNT			0x1c
+#define RNG_FROENABLE			0x20
+#define RNG_FRODETUNE			0x24
+#define RNG_ALARMMASK			0x28
+#define RNG_ALARMSTOP			0x2c
+#define RNG_OPTIONS			0x78
+#define RNG_EIP_REV			0x7c
+
+#define MONOBIT_FAIL_MASK		BIT(7)
+#define POKER_FAIL_MASK			BIT(6)
+#define LONG_RUN_FAIL_MASK		BIT(5)
+#define RUN_FAIL_MASK			BIT(4)
+#define NOISE_FAIL_MASK			BIT(3)
+#define STUCK_OUT_MASK			BIT(2)
+#define SHUTDOWN_OFLO_MASK		BIT(1)
+#define READY_MASK			BIT(0)
+
+#define MAJOR_HW_REV_RD(src)		(((src) & 0x0f000000) >> 24)
+#define MINOR_HW_REV_RD(src)		(((src) & 0x00f00000) >> 20)
+#define HW_PATCH_LEVEL_RD(src)		(((src) & 0x000f0000) >> 16)
+#define MAX_REFILL_CYCLES_SET(dst, src) \
+			((dst & ~0xffff0000) | (((u32)src << 16) & 0xffff0000))
+#define MIN_REFILL_CYCLES_SET(dst, src) \
+			((dst & ~0x000000ff) | (((u32)src) & 0x000000ff))
+#define ALARM_THRESHOLD_SET(dst, src) \
+			((dst & ~0x000000ff) | (((u32)src) & 0x000000ff))
+#define ENABLE_RNG_SET(dst, src) \
+			((dst & ~BIT(10)) | (((u32)src << 10) & BIT(10)))
+#define REGSPEC_TEST_MODE_SET(dst, src) \
+			((dst & ~BIT(8)) | (((u32)src << 8) & BIT(8)))
+#define MONOBIT_FAIL_MASK_SET(dst, src) \
+			((dst & ~BIT(7)) | (((u32)src << 7) & BIT(7)))
+#define POKER_FAIL_MASK_SET(dst, src) \
+			((dst & ~BIT(6)) | (((u32)src << 6) & BIT(6)))
+#define LONG_RUN_FAIL_MASK_SET(dst, src) \
+			((dst & ~BIT(5)) | (((u32)src << 5) & BIT(5)))
+#define RUN_FAIL_MASK_SET(dst, src) \
+			((dst & ~BIT(4)) | (((u32)src << 4) & BIT(4)))
+#define NOISE_FAIL_MASK_SET(dst, src) \
+			((dst & ~BIT(3)) | (((u32)src << 3) & BIT(3)))
+#define STUCK_OUT_MASK_SET(dst, src) \
+			((dst & ~BIT(2)) | (((u32)src << 2) & BIT(2)))
+#define SHUTDOWN_OFLO_MASK_SET(dst, src) \
+			((dst & ~BIT(1)) | (((u32)src << 1) & BIT(1)))
+
+struct xgene_rng_dev {
+	u32 irq;
+	void  __iomem *csr_base;
+	u32 revision;
+	u32 datum_size;
+	u32 failure_cnt;	/* Failure count last minute */
+	unsigned long failure_ts;/* First failure timestamp */
+	struct timer_list failure_timer;
+	struct device *dev;
+	struct clk *clk;
+};
+
+static void xgene_rng_expired_timer(unsigned long arg)
+{
+	struct xgene_rng_dev *ctx = (struct xgene_rng_dev *) arg;
+
+	/* Clear failure counter as timer expired */
+	disable_irq(ctx->irq);
+	ctx->failure_cnt = 0;
+	del_timer(&ctx->failure_timer);
+	enable_irq(ctx->irq);
+}
+
+static void xgene_rng_start_timer(struct xgene_rng_dev *ctx)
+{
+	ctx->failure_timer.data = (unsigned long) ctx;
+	ctx->failure_timer.function = xgene_rng_expired_timer;
+	ctx->failure_timer.expires = jiffies + 120 * HZ;
+	add_timer(&ctx->failure_timer);
+}
+
+/*
+ * Initialize or reinit free running oscillators (FROs)
+ */
+static void xgene_rng_init_fro(struct xgene_rng_dev *ctx, u32 fro_val)
+{
+	writel(fro_val, ctx->csr_base + RNG_FRODETUNE);
+	writel(0x00000000, ctx->csr_base + RNG_ALARMMASK);
+	writel(0x00000000, ctx->csr_base + RNG_ALARMSTOP);
+	writel(0xFFFFFFFF, ctx->csr_base + RNG_FROENABLE);
+}
+
+static void xgene_rng_chk_overflow(struct xgene_rng_dev *ctx)
+{
+	u32 val;
+
+	val = readl(ctx->csr_base + RNG_INTR_STS_ACK);
+	if (val & MONOBIT_FAIL_MASK)
+		/*
+		 * LFSR detected an out-of-bounds number of 1s after
+		 * checking 20,000 bits (test T1 as specified in the
+		 * AIS-31 standard)
+		 */
+		dev_err(ctx->dev, "test monobit failure error 0x%08X\n", val);
+	if (val & POKER_FAIL_MASK)
+		/*
+		 * LFSR detected an out-of-bounds value in at least one
+		 * of the 16 poker_count_X counters or an out of bounds sum
+		 * of squares value after checking 20,000 bits (test T2 as
+		 * specified in the AIS-31 standard)
+		 */
+		dev_err(ctx->dev, "test poker failure error 0x%08X\n", val);
+	if (val & LONG_RUN_FAIL_MASK)
+		/*
+		 * LFSR detected a sequence of 34 identical bits
+		 * (test T4 as specified in the AIS-31 standard)
+		 */
+		dev_err(ctx->dev, "test long run failure error 0x%08X\n", val);
+	if (val & RUN_FAIL_MASK)
+		/*
+		 * LFSR detected an outof-bounds value for at least one
+		 * of the running counters after checking 20,000 bits
+		 * (test T3 as specified in the AIS-31 standard)
+		 */
+		dev_err(ctx->dev, "test run failure error 0x%08X\n", val);
+	if (val & NOISE_FAIL_MASK)
+		/* LFSR detected a sequence of 48 identical bits */
+		dev_err(ctx->dev, "noise failure error 0x%08X\n", val);
+	if (val & STUCK_OUT_MASK)
+		/*
+		 * Detected output data registers generated same value twice
+		 * in a row
+		 */
+		dev_err(ctx->dev, "stuck out failure error 0x%08X\n", val);
+
+	if (val & SHUTDOWN_OFLO_MASK) {
+		u32 frostopped;
+
+		/* FROs shut down after a second error event. Try recover. */
+		if (++ctx->failure_cnt == 1) {
+			/* 1st time, just recover */
+			ctx->failure_ts = jiffies;
+			frostopped = readl(ctx->csr_base + RNG_ALARMSTOP);
+			xgene_rng_init_fro(ctx, frostopped);
+
+			/*
+			 * We must start a timer to clear out this error
+			 * in case the system timer wrap around
+			 */
+			xgene_rng_start_timer(ctx);
+		} else {
+			/* 2nd time failure in lesser than 1 minute? */
+			if (time_after(ctx->failure_ts + 60 * HZ, jiffies)) {
+				dev_err(ctx->dev,
+					"FRO shutdown failure error 0x%08X\n",
+					val);
+			} else {
+				/* 2nd time failure after 1 minutes, recover */
+				ctx->failure_ts = jiffies;
+				ctx->failure_cnt = 1;
+				/*
+				 * We must start a timer to clear out this
+				 * error in case the system timer wrap
+				 * around
+				 */
+				xgene_rng_start_timer(ctx);
+			}
+			frostopped = readl(ctx->csr_base + RNG_ALARMSTOP);
+			xgene_rng_init_fro(ctx, frostopped);
+		}
+	}
+	/* Clear them all */
+	writel(val, ctx->csr_base + RNG_INTR_STS_ACK);
+}
+
+static irqreturn_t xgene_rng_irq_handler(int irq, void *id)
+{
+	struct xgene_rng_dev *ctx = (struct xgene_rng_dev *) id;
+
+	/* RNG Alarm Counter overflow */
+	xgene_rng_chk_overflow(ctx);
+
+	return IRQ_HANDLED;
+}
+
+static int xgene_rng_data_present(struct hwrng *rng, int wait)
+{
+	struct xgene_rng_dev *ctx = (struct xgene_rng_dev *) rng->priv;
+	u32 i, val = 0;
+
+	for (i = 0; i < XGENE_RNG_RETRY_COUNT; i++) {
+		val = readl(ctx->csr_base + RNG_INTR_STS_ACK);
+		if ((val & READY_MASK) || !wait)
+			break;
+		udelay(XGENE_RNG_RETRY_INTERVAL);
+	}
+
+	return (val & READY_MASK);
+}
+
+static int xgene_rng_data_read(struct hwrng *rng, u32 *data)
+{
+	struct xgene_rng_dev *ctx = (struct xgene_rng_dev *) rng->priv;
+	int i;
+
+	for (i = 0; i < ctx->datum_size; i++)
+		data[i] = readl(ctx->csr_base + RNG_INOUT_0 + i * 4);
+
+	/* Clear ready bit to start next transaction */
+	writel(READY_MASK, ctx->csr_base + RNG_INTR_STS_ACK);
+
+	return ctx->datum_size << 2;
+}
+
+static void xgene_rng_init_internal(struct xgene_rng_dev *ctx)
+{
+	u32 val;
+
+	writel(0x00000000, ctx->csr_base + RNG_CONTROL);
+
+	val = MAX_REFILL_CYCLES_SET(0, 10);
+	val = MIN_REFILL_CYCLES_SET(val, 10);
+	writel(val, ctx->csr_base + RNG_CONFIG);
+
+	val = ALARM_THRESHOLD_SET(0, 0xFF);
+	writel(val, ctx->csr_base + RNG_ALARMCNT);
+
+	xgene_rng_init_fro(ctx, 0);
+
+	writel(MONOBIT_FAIL_MASK |
+		POKER_FAIL_MASK	|
+		LONG_RUN_FAIL_MASK |
+		RUN_FAIL_MASK |
+		NOISE_FAIL_MASK |
+		STUCK_OUT_MASK |
+		SHUTDOWN_OFLO_MASK |
+		READY_MASK, ctx->csr_base + RNG_INTR_STS_ACK);
+
+	val = ENABLE_RNG_SET(0, 1);
+	val = MONOBIT_FAIL_MASK_SET(val, 1);
+	val = POKER_FAIL_MASK_SET(val, 1);
+	val = LONG_RUN_FAIL_MASK_SET(val, 1);
+	val = RUN_FAIL_MASK_SET(val, 1);
+	val = NOISE_FAIL_MASK_SET(val, 1);
+	val = STUCK_OUT_MASK_SET(val, 1);
+	val = SHUTDOWN_OFLO_MASK_SET(val, 1);
+	writel(val, ctx->csr_base + RNG_CONTROL);
+}
+
+static int xgene_rng_init(struct hwrng *rng)
+{
+	struct xgene_rng_dev *ctx = (struct xgene_rng_dev *) rng->priv;
+
+	ctx->failure_cnt = 0;
+	init_timer(&ctx->failure_timer);
+
+	ctx->revision = readl(ctx->csr_base + RNG_EIP_REV);
+
+	dev_dbg(ctx->dev, "Rev %d.%d.%d\n",
+		MAJOR_HW_REV_RD(ctx->revision),
+		MINOR_HW_REV_RD(ctx->revision),
+		HW_PATCH_LEVEL_RD(ctx->revision));
+
+	dev_dbg(ctx->dev, "Options 0x%08X",
+		readl(ctx->csr_base + RNG_OPTIONS));
+
+	xgene_rng_init_internal(ctx);
+
+	ctx->datum_size = RNG_MAX_DATUM;
+
+	return 0;
+}
+
+static struct hwrng xgene_rng_func = {
+	.name		= "xgene-rng",
+	.init		= xgene_rng_init,
+	.data_present	= xgene_rng_data_present,
+	.data_read	= xgene_rng_data_read,
+};
+
+static int xgene_rng_probe(struct platform_device *pdev)
+{
+	struct resource *res;
+	struct xgene_rng_dev *ctx;
+	int rc = 0;
+
+	ctx = devm_kzalloc(&pdev->dev, sizeof(*ctx), GFP_KERNEL);
+	if (!ctx)
+		return -ENOMEM;
+
+	ctx->dev = &pdev->dev;
+	platform_set_drvdata(pdev, ctx);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	ctx->csr_base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(ctx->csr_base))
+		return PTR_ERR(ctx->csr_base);
+
+	ctx->irq = platform_get_irq(pdev, 0);
+	if (ctx->irq < 0) {
+		dev_err(&pdev->dev, "No IRQ resource\n");
+		return ctx->irq;
+	}
+
+	dev_dbg(&pdev->dev, "APM X-Gene RNG BASE %p ALARM IRQ %d",
+		ctx->csr_base, ctx->irq);
+
+	rc = devm_request_irq(&pdev->dev, ctx->irq, xgene_rng_irq_handler, 0,
+				dev_name(&pdev->dev), ctx);
+	if (rc) {
+		dev_err(&pdev->dev, "Could not request RNG alarm IRQ\n");
+		return rc;
+	}
+
+	/* Enable IP clock */
+	ctx->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(ctx->clk)) {
+		dev_warn(&pdev->dev, "Couldn't get the clock for RNG\n");
+	} else {
+		rc = clk_prepare_enable(ctx->clk);
+		if (rc) {
+			dev_warn(&pdev->dev,
+				 "clock prepare enable failed for RNG");
+			return rc;
+		}
+	}
+
+	xgene_rng_func.priv = (unsigned long) ctx;
+
+	rc = hwrng_register(&xgene_rng_func);
+	if (rc) {
+		dev_err(&pdev->dev, "RNG registering failed error %d\n", rc);
+		if (!IS_ERR(ctx->clk))
+			clk_disable_unprepare(ctx->clk);
+		return rc;
+	}
+
+	rc = device_init_wakeup(&pdev->dev, 1);
+	if (rc) {
+		dev_err(&pdev->dev, "RNG device_init_wakeup failed error %d\n",
+			rc);
+		if (!IS_ERR(ctx->clk))
+			clk_disable_unprepare(ctx->clk);
+		hwrng_unregister(&xgene_rng_func);
+		return rc;
+	}
+
+	return 0;
+}
+
+static int xgene_rng_remove(struct platform_device *pdev)
+{
+	struct xgene_rng_dev *ctx = platform_get_drvdata(pdev);
+	int rc;
+
+	rc = device_init_wakeup(&pdev->dev, 0);
+	if (rc)
+		dev_err(&pdev->dev, "RNG init wakeup failed error %d\n", rc);
+	if (!IS_ERR(ctx->clk))
+		clk_disable_unprepare(ctx->clk);
+	hwrng_unregister(&xgene_rng_func);
+
+	return rc;
+}
+
+static const struct of_device_id xgene_rng_of_match[] = {
+	{ .compatible = "apm,xgene-rng" },
+	{ }
+};
+
+MODULE_DEVICE_TABLE(of, xgene_rng_of_match);
+
+static struct platform_driver xgene_rng_driver = {
+	.probe = xgene_rng_probe,
+	.remove	= xgene_rng_remove,
+	.driver = {
+		.name		= "xgene-rng",
+		.of_match_table = xgene_rng_of_match,
+	},
+};
+
+module_platform_driver(xgene_rng_driver);
+MODULE_DESCRIPTION("APM X-Gene RNG driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/char/i8k.c b/drivers/char/i8k.c
index 65525c7e903c..e34a019eb930 100644
--- a/drivers/char/i8k.c
+++ b/drivers/char/i8k.c
@@ -298,7 +298,7 @@ static int i8k_get_temp(int sensor)
 	int temp;
 
 #ifdef I8K_TEMPERATURE_BUG
-	static int prev[4];
+	static int prev[4] = { I8K_MAX_TEMP+1, I8K_MAX_TEMP+1, I8K_MAX_TEMP+1, I8K_MAX_TEMP+1 };
 #endif
 	regs.ebx = sensor & 0xff;
 	rc = i8k_smm(&regs);
@@ -317,10 +317,12 @@ static int i8k_get_temp(int sensor)
 	 */
 	if (temp > I8K_MAX_TEMP) {
 		temp = prev[sensor];
-		prev[sensor] = I8K_MAX_TEMP;
+		prev[sensor] = I8K_MAX_TEMP+1;
 	} else {
 		prev[sensor] = temp;
 	}
+	if (temp > I8K_MAX_TEMP)
+		return -ERANGE;
 #endif
 
 	return temp;
@@ -499,6 +501,8 @@ static ssize_t i8k_hwmon_show_temp(struct device *dev,
 	int temp;
 
 	temp = i8k_get_temp(index);
+	if (temp == -ERANGE)
+		return -EINVAL;
 	if (temp < 0)
 		return temp;
 	return sprintf(buf, "%d\n", temp * 1000);
@@ -610,17 +614,17 @@ static int __init i8k_init_hwmon(void)
 
 	/* CPU temperature attributes, if temperature reading is OK */
 	err = i8k_get_temp(0);
-	if (err >= 0)
+	if (err >= 0 || err == -ERANGE)
 		i8k_hwmon_flags |= I8K_HWMON_HAVE_TEMP1;
 	/* check for additional temperature sensors */
 	err = i8k_get_temp(1);
-	if (err >= 0)
+	if (err >= 0 || err == -ERANGE)
 		i8k_hwmon_flags |= I8K_HWMON_HAVE_TEMP2;
 	err = i8k_get_temp(2);
-	if (err >= 0)
+	if (err >= 0 || err == -ERANGE)
 		i8k_hwmon_flags |= I8K_HWMON_HAVE_TEMP3;
 	err = i8k_get_temp(3);
-	if (err >= 0)
+	if (err >= 0 || err == -ERANGE)
 		i8k_hwmon_flags |= I8K_HWMON_HAVE_TEMP4;
 
 	/* Left fan attributes, if left fan is present */
@@ -651,6 +655,7 @@ struct i8k_config_data {
 
 enum i8k_configs {
 	DELL_LATITUDE_D520,
+	DELL_LATITUDE_E6540,
 	DELL_PRECISION_490,
 	DELL_STUDIO,
 	DELL_XPS_M140,
@@ -661,6 +666,10 @@ static const struct i8k_config_data i8k_config_data[] = {
 		.fan_mult = 1,
 		.fan_max = I8K_FAN_TURBO,
 	},
+	[DELL_LATITUDE_E6540] = {
+		.fan_mult = 1,
+		.fan_max = I8K_FAN_HIGH,
+	},
 	[DELL_PRECISION_490] = {
 		.fan_mult = 1,
 		.fan_max = I8K_FAN_TURBO,
@@ -706,6 +715,22 @@ static struct dmi_system_id i8k_dmi_table[] __initdata = {
 		.driver_data = (void *)&i8k_config_data[DELL_LATITUDE_D520],
 	},
 	{
+		.ident = "Dell Latitude E6440",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E6440"),
+		},
+		.driver_data = (void *)&i8k_config_data[DELL_LATITUDE_E6540],
+	},
+	{
+		.ident = "Dell Latitude E6540",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E6540"),
+		},
+		.driver_data = (void *)&i8k_config_data[DELL_LATITUDE_E6540],
+	},
+	{
 		.ident = "Dell Latitude 2",
 		.matches = {
 			DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
@@ -775,6 +800,8 @@ static struct dmi_system_id i8k_dmi_table[] __initdata = {
 	{ }
 };
 
+MODULE_DEVICE_TABLE(dmi, i8k_dmi_table);
+
 /*
  * Probe for the presence of a supported laptop.
  */
diff --git a/drivers/char/ipmi/Kconfig b/drivers/char/ipmi/Kconfig
index db1c9b7adaa6..6ed9e9fe5233 100644
--- a/drivers/char/ipmi/Kconfig
+++ b/drivers/char/ipmi/Kconfig
@@ -62,6 +62,20 @@ config IPMI_SI_PROBE_DEFAULTS
 	 only be available on older systems if the "ipmi_si_intf.trydefaults=1"
 	 boot argument is passed.
 
+config IPMI_SSIF
+       tristate 'IPMI SMBus handler (SSIF)'
+       select I2C
+       help
+         Provides a driver for a SMBus interface to a BMC, meaning that you
+	 have a driver that must be accessed over an I2C bus instead of a
+	 standard interface.  This module requires I2C support.
+
+config IPMI_POWERNV
+       depends on PPC_POWERNV
+       tristate 'POWERNV (OPAL firmware) IPMI interface'
+       help
+         Provides a driver for OPAL firmware-based IPMI interfaces.
+
 config IPMI_WATCHDOG
        tristate 'IPMI Watchdog Timer'
        help
diff --git a/drivers/char/ipmi/Makefile b/drivers/char/ipmi/Makefile
index 16a93648d54e..f3ffde1f5f1f 100644
--- a/drivers/char/ipmi/Makefile
+++ b/drivers/char/ipmi/Makefile
@@ -7,5 +7,7 @@ ipmi_si-y := ipmi_si_intf.o ipmi_kcs_sm.o ipmi_smic_sm.o ipmi_bt_sm.o
 obj-$(CONFIG_IPMI_HANDLER) += ipmi_msghandler.o
 obj-$(CONFIG_IPMI_DEVICE_INTERFACE) += ipmi_devintf.o
 obj-$(CONFIG_IPMI_SI) += ipmi_si.o
+obj-$(CONFIG_IPMI_SSIF) += ipmi_ssif.o
+obj-$(CONFIG_IPMI_POWERNV) += ipmi_powernv.o
 obj-$(CONFIG_IPMI_WATCHDOG) += ipmi_watchdog.o
 obj-$(CONFIG_IPMI_POWEROFF) += ipmi_poweroff.o
diff --git a/drivers/char/ipmi/ipmi_msghandler.c b/drivers/char/ipmi/ipmi_msghandler.c
index e6db9381b2c7..6b65fa4e0c55 100644
--- a/drivers/char/ipmi/ipmi_msghandler.c
+++ b/drivers/char/ipmi/ipmi_msghandler.c
@@ -56,6 +56,8 @@ static int ipmi_init_msghandler(void);
 static void smi_recv_tasklet(unsigned long);
 static void handle_new_recv_msgs(ipmi_smi_t intf);
 static void need_waiter(ipmi_smi_t intf);
+static int handle_one_recv_msg(ipmi_smi_t          intf,
+			       struct ipmi_smi_msg *msg);
 
 static int initialized;
 
@@ -191,25 +193,14 @@ struct ipmi_proc_entry {
 #endif
 
 struct bmc_device {
-	struct platform_device *dev;
+	struct platform_device pdev;
 	struct ipmi_device_id  id;
 	unsigned char          guid[16];
 	int                    guid_set;
-
-	struct kref	       refcount;
-
-	/* bmc device attributes */
-	struct device_attribute device_id_attr;
-	struct device_attribute provides_dev_sdrs_attr;
-	struct device_attribute revision_attr;
-	struct device_attribute firmware_rev_attr;
-	struct device_attribute version_attr;
-	struct device_attribute add_dev_support_attr;
-	struct device_attribute manufacturer_id_attr;
-	struct device_attribute product_id_attr;
-	struct device_attribute guid_attr;
-	struct device_attribute aux_firmware_rev_attr;
+	char                   name[16];
+	struct kref	       usecount;
 };
+#define to_bmc_device(x) container_of((x), struct bmc_device, pdev.dev)
 
 /*
  * Various statistics for IPMI, these index stats[] in the ipmi_smi
@@ -323,6 +314,9 @@ struct ipmi_smi {
 
 	struct kref refcount;
 
+	/* Set when the interface is being unregistered. */
+	bool in_shutdown;
+
 	/* Used for a list of interfaces. */
 	struct list_head link;
 
@@ -341,7 +335,6 @@ struct ipmi_smi {
 
 	struct bmc_device *bmc;
 	char *my_dev_name;
-	char *sysfs_name;
 
 	/*
 	 * This is the lower-layer's sender routine.  Note that you
@@ -377,11 +370,16 @@ struct ipmi_smi {
 	 * periodic timer interrupt.  The tasklet is for handling received
 	 * messages directly from the handler.
 	 */
-	spinlock_t       waiting_msgs_lock;
-	struct list_head waiting_msgs;
+	spinlock_t       waiting_rcv_msgs_lock;
+	struct list_head waiting_rcv_msgs;
 	atomic_t	 watchdog_pretimeouts_to_deliver;
 	struct tasklet_struct recv_tasklet;
 
+	spinlock_t             xmit_msgs_lock;
+	struct list_head       xmit_msgs;
+	struct ipmi_smi_msg    *curr_msg;
+	struct list_head       hp_xmit_msgs;
+
 	/*
 	 * The list of command receivers that are registered for commands
 	 * on this interface.
@@ -474,6 +472,18 @@ static DEFINE_MUTEX(smi_watchers_mutex);
 #define ipmi_get_stat(intf, stat) \
 	((unsigned int) atomic_read(&(intf)->stats[IPMI_STAT_ ## stat]))
 
+static char *addr_src_to_str[] = { "invalid", "hotmod", "hardcoded", "SPMI",
+				   "ACPI", "SMBIOS", "PCI",
+				   "device-tree", "default" };
+
+const char *ipmi_addr_src_to_str(enum ipmi_addr_src src)
+{
+	if (src > SI_DEFAULT)
+		src = 0; /* Invalid */
+	return addr_src_to_str[src];
+}
+EXPORT_SYMBOL(ipmi_addr_src_to_str);
+
 static int is_lan_addr(struct ipmi_addr *addr)
 {
 	return addr->addr_type == IPMI_LAN_ADDR_TYPE;
@@ -517,7 +527,7 @@ static void clean_up_interface_data(ipmi_smi_t intf)
 
 	tasklet_kill(&intf->recv_tasklet);
 
-	free_smi_msg_list(&intf->waiting_msgs);
+	free_smi_msg_list(&intf->waiting_rcv_msgs);
 	free_recv_msg_list(&intf->waiting_events);
 
 	/*
@@ -1473,6 +1483,30 @@ static inline void format_lan_msg(struct ipmi_smi_msg   *smi_msg,
 	smi_msg->msgid = msgid;
 }
 
+static void smi_send(ipmi_smi_t intf, struct ipmi_smi_handlers *handlers,
+		     struct ipmi_smi_msg *smi_msg, int priority)
+{
+	int run_to_completion = intf->run_to_completion;
+	unsigned long flags;
+
+	if (!run_to_completion)
+		spin_lock_irqsave(&intf->xmit_msgs_lock, flags);
+	if (intf->curr_msg) {
+		if (priority > 0)
+			list_add_tail(&smi_msg->link, &intf->hp_xmit_msgs);
+		else
+			list_add_tail(&smi_msg->link, &intf->xmit_msgs);
+		smi_msg = NULL;
+	} else {
+		intf->curr_msg = smi_msg;
+	}
+	if (!run_to_completion)
+		spin_unlock_irqrestore(&intf->xmit_msgs_lock, flags);
+
+	if (smi_msg)
+		handlers->sender(intf->send_info, smi_msg);
+}
+
 /*
  * Separate from ipmi_request so that the user does not have to be
  * supplied in certain circumstances (mainly at panic time).  If
@@ -1497,7 +1531,6 @@ static int i_ipmi_request(ipmi_user_t          user,
 	struct ipmi_smi_msg      *smi_msg;
 	struct ipmi_recv_msg     *recv_msg;
 	unsigned long            flags;
-	struct ipmi_smi_handlers *handlers;
 
 
 	if (supplied_recv)
@@ -1520,8 +1553,7 @@ static int i_ipmi_request(ipmi_user_t          user,
 	}
 
 	rcu_read_lock();
-	handlers = intf->handlers;
-	if (!handlers) {
+	if (intf->in_shutdown) {
 		rv = -ENODEV;
 		goto out_err;
 	}
@@ -1856,7 +1888,7 @@ static int i_ipmi_request(ipmi_user_t          user,
 	}
 #endif
 
-	handlers->sender(intf->send_info, smi_msg, priority);
+	smi_send(intf, intf->handlers, smi_msg, priority);
 	rcu_read_unlock();
 
 	return 0;
@@ -2153,7 +2185,7 @@ static void remove_proc_entries(ipmi_smi_t smi)
 static int __find_bmc_guid(struct device *dev, void *data)
 {
 	unsigned char *id = data;
-	struct bmc_device *bmc = dev_get_drvdata(dev);
+	struct bmc_device *bmc = to_bmc_device(dev);
 	return memcmp(bmc->guid, id, 16) == 0;
 }
 
@@ -2164,7 +2196,7 @@ static struct bmc_device *ipmi_find_bmc_guid(struct device_driver *drv,
 
 	dev = driver_find_device(drv, NULL, guid, __find_bmc_guid);
 	if (dev)
-		return dev_get_drvdata(dev);
+		return to_bmc_device(dev);
 	else
 		return NULL;
 }
@@ -2177,7 +2209,7 @@ struct prod_dev_id {
 static int __find_bmc_prod_dev_id(struct device *dev, void *data)
 {
 	struct prod_dev_id *id = data;
-	struct bmc_device *bmc = dev_get_drvdata(dev);
+	struct bmc_device *bmc = to_bmc_device(dev);
 
 	return (bmc->id.product_id == id->product_id
 		&& bmc->id.device_id == id->device_id);
@@ -2195,7 +2227,7 @@ static struct bmc_device *ipmi_find_bmc_prod_dev_id(
 
 	dev = driver_find_device(drv, NULL, &id, __find_bmc_prod_dev_id);
 	if (dev)
-		return dev_get_drvdata(dev);
+		return to_bmc_device(dev);
 	else
 		return NULL;
 }
@@ -2204,84 +2236,94 @@ static ssize_t device_id_show(struct device *dev,
 			      struct device_attribute *attr,
 			      char *buf)
 {
-	struct bmc_device *bmc = dev_get_drvdata(dev);
+	struct bmc_device *bmc = to_bmc_device(dev);
 
 	return snprintf(buf, 10, "%u\n", bmc->id.device_id);
 }
+static DEVICE_ATTR(device_id, S_IRUGO, device_id_show, NULL);
 
-static ssize_t provides_dev_sdrs_show(struct device *dev,
-				      struct device_attribute *attr,
-				      char *buf)
+static ssize_t provides_device_sdrs_show(struct device *dev,
+					 struct device_attribute *attr,
+					 char *buf)
 {
-	struct bmc_device *bmc = dev_get_drvdata(dev);
+	struct bmc_device *bmc = to_bmc_device(dev);
 
 	return snprintf(buf, 10, "%u\n",
 			(bmc->id.device_revision & 0x80) >> 7);
 }
+static DEVICE_ATTR(provides_device_sdrs, S_IRUGO, provides_device_sdrs_show,
+		   NULL);
 
 static ssize_t revision_show(struct device *dev, struct device_attribute *attr,
 			     char *buf)
 {
-	struct bmc_device *bmc = dev_get_drvdata(dev);
+	struct bmc_device *bmc = to_bmc_device(dev);
 
 	return snprintf(buf, 20, "%u\n",
 			bmc->id.device_revision & 0x0F);
 }
+static DEVICE_ATTR(revision, S_IRUGO, revision_show, NULL);
 
-static ssize_t firmware_rev_show(struct device *dev,
-				 struct device_attribute *attr,
-				 char *buf)
+static ssize_t firmware_revision_show(struct device *dev,
+				      struct device_attribute *attr,
+				      char *buf)
 {
-	struct bmc_device *bmc = dev_get_drvdata(dev);
+	struct bmc_device *bmc = to_bmc_device(dev);
 
 	return snprintf(buf, 20, "%u.%x\n", bmc->id.firmware_revision_1,
 			bmc->id.firmware_revision_2);
 }
+static DEVICE_ATTR(firmware_revision, S_IRUGO, firmware_revision_show, NULL);
 
 static ssize_t ipmi_version_show(struct device *dev,
 				 struct device_attribute *attr,
 				 char *buf)
 {
-	struct bmc_device *bmc = dev_get_drvdata(dev);
+	struct bmc_device *bmc = to_bmc_device(dev);
 
 	return snprintf(buf, 20, "%u.%u\n",
 			ipmi_version_major(&bmc->id),
 			ipmi_version_minor(&bmc->id));
 }
+static DEVICE_ATTR(ipmi_version, S_IRUGO, ipmi_version_show, NULL);
 
 static ssize_t add_dev_support_show(struct device *dev,
 				    struct device_attribute *attr,
 				    char *buf)
 {
-	struct bmc_device *bmc = dev_get_drvdata(dev);
+	struct bmc_device *bmc = to_bmc_device(dev);
 
 	return snprintf(buf, 10, "0x%02x\n",
 			bmc->id.additional_device_support);
 }
+static DEVICE_ATTR(additional_device_support, S_IRUGO, add_dev_support_show,
+		   NULL);
 
 static ssize_t manufacturer_id_show(struct device *dev,
 				    struct device_attribute *attr,
 				    char *buf)
 {
-	struct bmc_device *bmc = dev_get_drvdata(dev);
+	struct bmc_device *bmc = to_bmc_device(dev);
 
 	return snprintf(buf, 20, "0x%6.6x\n", bmc->id.manufacturer_id);
 }
+static DEVICE_ATTR(manufacturer_id, S_IRUGO, manufacturer_id_show, NULL);
 
 static ssize_t product_id_show(struct device *dev,
 			       struct device_attribute *attr,
 			       char *buf)
 {
-	struct bmc_device *bmc = dev_get_drvdata(dev);
+	struct bmc_device *bmc = to_bmc_device(dev);
 
 	return snprintf(buf, 10, "0x%4.4x\n", bmc->id.product_id);
 }
+static DEVICE_ATTR(product_id, S_IRUGO, product_id_show, NULL);
 
 static ssize_t aux_firmware_rev_show(struct device *dev,
 				     struct device_attribute *attr,
 				     char *buf)
 {
-	struct bmc_device *bmc = dev_get_drvdata(dev);
+	struct bmc_device *bmc = to_bmc_device(dev);
 
 	return snprintf(buf, 21, "0x%02x 0x%02x 0x%02x 0x%02x\n",
 			bmc->id.aux_firmware_revision[3],
@@ -2289,175 +2331,95 @@ static ssize_t aux_firmware_rev_show(struct device *dev,
 			bmc->id.aux_firmware_revision[1],
 			bmc->id.aux_firmware_revision[0]);
 }
+static DEVICE_ATTR(aux_firmware_revision, S_IRUGO, aux_firmware_rev_show, NULL);
 
 static ssize_t guid_show(struct device *dev, struct device_attribute *attr,
 			 char *buf)
 {
-	struct bmc_device *bmc = dev_get_drvdata(dev);
+	struct bmc_device *bmc = to_bmc_device(dev);
 
 	return snprintf(buf, 100, "%Lx%Lx\n",
 			(long long) bmc->guid[0],
 			(long long) bmc->guid[8]);
 }
+static DEVICE_ATTR(guid, S_IRUGO, guid_show, NULL);
+
+static struct attribute *bmc_dev_attrs[] = {
+	&dev_attr_device_id.attr,
+	&dev_attr_provides_device_sdrs.attr,
+	&dev_attr_revision.attr,
+	&dev_attr_firmware_revision.attr,
+	&dev_attr_ipmi_version.attr,
+	&dev_attr_additional_device_support.attr,
+	&dev_attr_manufacturer_id.attr,
+	&dev_attr_product_id.attr,
+	NULL
+};
 
-static void remove_files(struct bmc_device *bmc)
-{
-	if (!bmc->dev)
-		return;
+static struct attribute_group bmc_dev_attr_group = {
+	.attrs		= bmc_dev_attrs,
+};
 
-	device_remove_file(&bmc->dev->dev,
-			   &bmc->device_id_attr);
-	device_remove_file(&bmc->dev->dev,
-			   &bmc->provides_dev_sdrs_attr);
-	device_remove_file(&bmc->dev->dev,
-			   &bmc->revision_attr);
-	device_remove_file(&bmc->dev->dev,
-			   &bmc->firmware_rev_attr);
-	device_remove_file(&bmc->dev->dev,
-			   &bmc->version_attr);
-	device_remove_file(&bmc->dev->dev,
-			   &bmc->add_dev_support_attr);
-	device_remove_file(&bmc->dev->dev,
-			   &bmc->manufacturer_id_attr);
-	device_remove_file(&bmc->dev->dev,
-			   &bmc->product_id_attr);
+static const struct attribute_group *bmc_dev_attr_groups[] = {
+	&bmc_dev_attr_group,
+	NULL
+};
 
-	if (bmc->id.aux_firmware_revision_set)
-		device_remove_file(&bmc->dev->dev,
-				   &bmc->aux_firmware_rev_attr);
-	if (bmc->guid_set)
-		device_remove_file(&bmc->dev->dev,
-				   &bmc->guid_attr);
+static struct device_type bmc_device_type = {
+	.groups		= bmc_dev_attr_groups,
+};
+
+static void
+release_bmc_device(struct device *dev)
+{
+	kfree(to_bmc_device(dev));
 }
 
 static void
 cleanup_bmc_device(struct kref *ref)
 {
-	struct bmc_device *bmc;
+	struct bmc_device *bmc = container_of(ref, struct bmc_device, usecount);
 
-	bmc = container_of(ref, struct bmc_device, refcount);
+	if (bmc->id.aux_firmware_revision_set)
+		device_remove_file(&bmc->pdev.dev,
+				   &dev_attr_aux_firmware_revision);
+	if (bmc->guid_set)
+		device_remove_file(&bmc->pdev.dev,
+				   &dev_attr_guid);
 
-	remove_files(bmc);
-	platform_device_unregister(bmc->dev);
-	kfree(bmc);
+	platform_device_unregister(&bmc->pdev);
 }
 
 static void ipmi_bmc_unregister(ipmi_smi_t intf)
 {
 	struct bmc_device *bmc = intf->bmc;
 
-	if (intf->sysfs_name) {
-		sysfs_remove_link(&intf->si_dev->kobj, intf->sysfs_name);
-		kfree(intf->sysfs_name);
-		intf->sysfs_name = NULL;
-	}
+	sysfs_remove_link(&intf->si_dev->kobj, "bmc");
 	if (intf->my_dev_name) {
-		sysfs_remove_link(&bmc->dev->dev.kobj, intf->my_dev_name);
+		sysfs_remove_link(&bmc->pdev.dev.kobj, intf->my_dev_name);
 		kfree(intf->my_dev_name);
 		intf->my_dev_name = NULL;
 	}
 
 	mutex_lock(&ipmidriver_mutex);
-	kref_put(&bmc->refcount, cleanup_bmc_device);
+	kref_put(&bmc->usecount, cleanup_bmc_device);
 	intf->bmc = NULL;
 	mutex_unlock(&ipmidriver_mutex);
 }
 
-static int create_files(struct bmc_device *bmc)
+static int create_bmc_files(struct bmc_device *bmc)
 {
 	int err;
 
-	bmc->device_id_attr.attr.name = "device_id";
-	bmc->device_id_attr.attr.mode = S_IRUGO;
-	bmc->device_id_attr.show = device_id_show;
-	sysfs_attr_init(&bmc->device_id_attr.attr);
-
-	bmc->provides_dev_sdrs_attr.attr.name = "provides_device_sdrs";
-	bmc->provides_dev_sdrs_attr.attr.mode = S_IRUGO;
-	bmc->provides_dev_sdrs_attr.show = provides_dev_sdrs_show;
-	sysfs_attr_init(&bmc->provides_dev_sdrs_attr.attr);
-
-	bmc->revision_attr.attr.name = "revision";
-	bmc->revision_attr.attr.mode = S_IRUGO;
-	bmc->revision_attr.show = revision_show;
-	sysfs_attr_init(&bmc->revision_attr.attr);
-
-	bmc->firmware_rev_attr.attr.name = "firmware_revision";
-	bmc->firmware_rev_attr.attr.mode = S_IRUGO;
-	bmc->firmware_rev_attr.show = firmware_rev_show;
-	sysfs_attr_init(&bmc->firmware_rev_attr.attr);
-
-	bmc->version_attr.attr.name = "ipmi_version";
-	bmc->version_attr.attr.mode = S_IRUGO;
-	bmc->version_attr.show = ipmi_version_show;
-	sysfs_attr_init(&bmc->version_attr.attr);
-
-	bmc->add_dev_support_attr.attr.name = "additional_device_support";
-	bmc->add_dev_support_attr.attr.mode = S_IRUGO;
-	bmc->add_dev_support_attr.show = add_dev_support_show;
-	sysfs_attr_init(&bmc->add_dev_support_attr.attr);
-
-	bmc->manufacturer_id_attr.attr.name = "manufacturer_id";
-	bmc->manufacturer_id_attr.attr.mode = S_IRUGO;
-	bmc->manufacturer_id_attr.show = manufacturer_id_show;
-	sysfs_attr_init(&bmc->manufacturer_id_attr.attr);
-
-	bmc->product_id_attr.attr.name = "product_id";
-	bmc->product_id_attr.attr.mode = S_IRUGO;
-	bmc->product_id_attr.show = product_id_show;
-	sysfs_attr_init(&bmc->product_id_attr.attr);
-
-	bmc->guid_attr.attr.name = "guid";
-	bmc->guid_attr.attr.mode = S_IRUGO;
-	bmc->guid_attr.show = guid_show;
-	sysfs_attr_init(&bmc->guid_attr.attr);
-
-	bmc->aux_firmware_rev_attr.attr.name = "aux_firmware_revision";
-	bmc->aux_firmware_rev_attr.attr.mode = S_IRUGO;
-	bmc->aux_firmware_rev_attr.show = aux_firmware_rev_show;
-	sysfs_attr_init(&bmc->aux_firmware_rev_attr.attr);
-
-	err = device_create_file(&bmc->dev->dev,
-			   &bmc->device_id_attr);
-	if (err)
-		goto out;
-	err = device_create_file(&bmc->dev->dev,
-			   &bmc->provides_dev_sdrs_attr);
-	if (err)
-		goto out_devid;
-	err = device_create_file(&bmc->dev->dev,
-			   &bmc->revision_attr);
-	if (err)
-		goto out_sdrs;
-	err = device_create_file(&bmc->dev->dev,
-			   &bmc->firmware_rev_attr);
-	if (err)
-		goto out_rev;
-	err = device_create_file(&bmc->dev->dev,
-			   &bmc->version_attr);
-	if (err)
-		goto out_firm;
-	err = device_create_file(&bmc->dev->dev,
-			   &bmc->add_dev_support_attr);
-	if (err)
-		goto out_version;
-	err = device_create_file(&bmc->dev->dev,
-			   &bmc->manufacturer_id_attr);
-	if (err)
-		goto out_add_dev;
-	err = device_create_file(&bmc->dev->dev,
-			   &bmc->product_id_attr);
-	if (err)
-		goto out_manu;
 	if (bmc->id.aux_firmware_revision_set) {
-		err = device_create_file(&bmc->dev->dev,
-				   &bmc->aux_firmware_rev_attr);
+		err = device_create_file(&bmc->pdev.dev,
+					 &dev_attr_aux_firmware_revision);
 		if (err)
-			goto out_prod_id;
+			goto out;
 	}
 	if (bmc->guid_set) {
-		err = device_create_file(&bmc->dev->dev,
-				   &bmc->guid_attr);
+		err = device_create_file(&bmc->pdev.dev,
+					 &dev_attr_guid);
 		if (err)
 			goto out_aux_firm;
 	}
@@ -2466,44 +2428,17 @@ static int create_files(struct bmc_device *bmc)
 
 out_aux_firm:
 	if (bmc->id.aux_firmware_revision_set)
-		device_remove_file(&bmc->dev->dev,
-				   &bmc->aux_firmware_rev_attr);
-out_prod_id:
-	device_remove_file(&bmc->dev->dev,
-			   &bmc->product_id_attr);
-out_manu:
-	device_remove_file(&bmc->dev->dev,
-			   &bmc->manufacturer_id_attr);
-out_add_dev:
-	device_remove_file(&bmc->dev->dev,
-			   &bmc->add_dev_support_attr);
-out_version:
-	device_remove_file(&bmc->dev->dev,
-			   &bmc->version_attr);
-out_firm:
-	device_remove_file(&bmc->dev->dev,
-			   &bmc->firmware_rev_attr);
-out_rev:
-	device_remove_file(&bmc->dev->dev,
-			   &bmc->revision_attr);
-out_sdrs:
-	device_remove_file(&bmc->dev->dev,
-			   &bmc->provides_dev_sdrs_attr);
-out_devid:
-	device_remove_file(&bmc->dev->dev,
-			   &bmc->device_id_attr);
+		device_remove_file(&bmc->pdev.dev,
+				   &dev_attr_aux_firmware_revision);
 out:
 	return err;
 }
 
-static int ipmi_bmc_register(ipmi_smi_t intf, int ifnum,
-			     const char *sysfs_name)
+static int ipmi_bmc_register(ipmi_smi_t intf, int ifnum)
 {
 	int               rv;
 	struct bmc_device *bmc = intf->bmc;
 	struct bmc_device *old_bmc;
-	int               size;
-	char              dummy[1];
 
 	mutex_lock(&ipmidriver_mutex);
 
@@ -2527,7 +2462,7 @@ static int ipmi_bmc_register(ipmi_smi_t intf, int ifnum,
 		intf->bmc = old_bmc;
 		bmc = old_bmc;
 
-		kref_get(&bmc->refcount);
+		kref_get(&bmc->usecount);
 		mutex_unlock(&ipmidriver_mutex);
 
 		printk(KERN_INFO
@@ -2537,12 +2472,12 @@ static int ipmi_bmc_register(ipmi_smi_t intf, int ifnum,
 		       bmc->id.product_id,
 		       bmc->id.device_id);
 	} else {
-		char name[14];
 		unsigned char orig_dev_id = bmc->id.device_id;
 		int warn_printed = 0;
 
-		snprintf(name, sizeof(name),
+		snprintf(bmc->name, sizeof(bmc->name),
 			 "ipmi_bmc.%4.4x", bmc->id.product_id);
+		bmc->pdev.name = bmc->name;
 
 		while (ipmi_find_bmc_prod_dev_id(&ipmidriver.driver,
 						 bmc->id.product_id,
@@ -2566,23 +2501,16 @@ static int ipmi_bmc_register(ipmi_smi_t intf, int ifnum,
 			}
 		}
 
-		bmc->dev = platform_device_alloc(name, bmc->id.device_id);
-		if (!bmc->dev) {
-			mutex_unlock(&ipmidriver_mutex);
-			printk(KERN_ERR
-			       "ipmi_msghandler:"
-			       " Unable to allocate platform device\n");
-			return -ENOMEM;
-		}
-		bmc->dev->dev.driver = &ipmidriver.driver;
-		dev_set_drvdata(&bmc->dev->dev, bmc);
-		kref_init(&bmc->refcount);
+		bmc->pdev.dev.driver = &ipmidriver.driver;
+		bmc->pdev.id = bmc->id.device_id;
+		bmc->pdev.dev.release = release_bmc_device;
+		bmc->pdev.dev.type = &bmc_device_type;
+		kref_init(&bmc->usecount);
 
-		rv = platform_device_add(bmc->dev);
+		rv = platform_device_register(&bmc->pdev);
 		mutex_unlock(&ipmidriver_mutex);
 		if (rv) {
-			platform_device_put(bmc->dev);
-			bmc->dev = NULL;
+			put_device(&bmc->pdev.dev);
 			printk(KERN_ERR
 			       "ipmi_msghandler:"
 			       " Unable to register bmc device: %d\n",
@@ -2594,10 +2522,10 @@ static int ipmi_bmc_register(ipmi_smi_t intf, int ifnum,
 			return rv;
 		}
 
-		rv = create_files(bmc);
+		rv = create_bmc_files(bmc);
 		if (rv) {
 			mutex_lock(&ipmidriver_mutex);
-			platform_device_unregister(bmc->dev);
+			platform_device_unregister(&bmc->pdev);
 			mutex_unlock(&ipmidriver_mutex);
 
 			return rv;
@@ -2614,44 +2542,26 @@ static int ipmi_bmc_register(ipmi_smi_t intf, int ifnum,
 	 * create symlink from system interface device to bmc device
 	 * and back.
 	 */
-	intf->sysfs_name = kstrdup(sysfs_name, GFP_KERNEL);
-	if (!intf->sysfs_name) {
-		rv = -ENOMEM;
-		printk(KERN_ERR
-		       "ipmi_msghandler: allocate link to BMC: %d\n",
-		       rv);
-		goto out_err;
-	}
-
-	rv = sysfs_create_link(&intf->si_dev->kobj,
-			       &bmc->dev->dev.kobj, intf->sysfs_name);
+	rv = sysfs_create_link(&intf->si_dev->kobj, &bmc->pdev.dev.kobj, "bmc");
 	if (rv) {
-		kfree(intf->sysfs_name);
-		intf->sysfs_name = NULL;
 		printk(KERN_ERR
 		       "ipmi_msghandler: Unable to create bmc symlink: %d\n",
 		       rv);
 		goto out_err;
 	}
 
-	size = snprintf(dummy, 0, "ipmi%d", ifnum);
-	intf->my_dev_name = kmalloc(size+1, GFP_KERNEL);
+	intf->my_dev_name = kasprintf(GFP_KERNEL, "ipmi%d", ifnum);
 	if (!intf->my_dev_name) {
-		kfree(intf->sysfs_name);
-		intf->sysfs_name = NULL;
 		rv = -ENOMEM;
 		printk(KERN_ERR
 		       "ipmi_msghandler: allocate link from BMC: %d\n",
 		       rv);
 		goto out_err;
 	}
-	snprintf(intf->my_dev_name, size+1, "ipmi%d", ifnum);
 
-	rv = sysfs_create_link(&bmc->dev->dev.kobj, &intf->si_dev->kobj,
+	rv = sysfs_create_link(&bmc->pdev.dev.kobj, &intf->si_dev->kobj,
 			       intf->my_dev_name);
 	if (rv) {
-		kfree(intf->sysfs_name);
-		intf->sysfs_name = NULL;
 		kfree(intf->my_dev_name);
 		intf->my_dev_name = NULL;
 		printk(KERN_ERR
@@ -2796,7 +2706,6 @@ channel_handler(ipmi_smi_t intf, struct ipmi_recv_msg *msg)
 					= IPMI_CHANNEL_MEDIUM_IPMB;
 				intf->channels[0].protocol
 					= IPMI_CHANNEL_PROTOCOL_IPMB;
-				rv = -ENOSYS;
 
 				intf->curr_channel = IPMI_MAX_CHANNELS;
 				wake_up(&intf->waitq);
@@ -2821,12 +2730,12 @@ channel_handler(ipmi_smi_t intf, struct ipmi_recv_msg *msg)
 
 		if (rv) {
 			/* Got an error somehow, just give up. */
+			printk(KERN_WARNING PFX
+			       "Error sending channel information for channel"
+			       " %d: %d\n", intf->curr_channel, rv);
+
 			intf->curr_channel = IPMI_MAX_CHANNELS;
 			wake_up(&intf->waitq);
-
-			printk(KERN_WARNING PFX
-			       "Error sending channel information: %d\n",
-			       rv);
 		}
 	}
  out:
@@ -2851,7 +2760,6 @@ int ipmi_register_smi(struct ipmi_smi_handlers *handlers,
 		      void		       *send_info,
 		      struct ipmi_device_id    *device_id,
 		      struct device            *si_dev,
-		      const char               *sysfs_name,
 		      unsigned char            slave_addr)
 {
 	int              i, j;
@@ -2910,12 +2818,15 @@ int ipmi_register_smi(struct ipmi_smi_handlers *handlers,
 #ifdef CONFIG_PROC_FS
 	mutex_init(&intf->proc_entry_lock);
 #endif
-	spin_lock_init(&intf->waiting_msgs_lock);
-	INIT_LIST_HEAD(&intf->waiting_msgs);
+	spin_lock_init(&intf->waiting_rcv_msgs_lock);
+	INIT_LIST_HEAD(&intf->waiting_rcv_msgs);
 	tasklet_init(&intf->recv_tasklet,
 		     smi_recv_tasklet,
 		     (unsigned long) intf);
 	atomic_set(&intf->watchdog_pretimeouts_to_deliver, 0);
+	spin_lock_init(&intf->xmit_msgs_lock);
+	INIT_LIST_HEAD(&intf->xmit_msgs);
+	INIT_LIST_HEAD(&intf->hp_xmit_msgs);
 	spin_lock_init(&intf->events_lock);
 	atomic_set(&intf->event_waiters, 0);
 	intf->ticks_to_req_ev = IPMI_REQUEST_EV_TIME;
@@ -2964,8 +2875,12 @@ int ipmi_register_smi(struct ipmi_smi_handlers *handlers,
 		intf->null_user_handler = channel_handler;
 		intf->curr_channel = 0;
 		rv = send_channel_info_cmd(intf, 0);
-		if (rv)
+		if (rv) {
+			printk(KERN_WARNING PFX
+			       "Error sending channel information for channel"
+			       " 0, %d\n", rv);
 			goto out;
+		}
 
 		/* Wait for the channel info to be read. */
 		wait_event(intf->waitq,
@@ -2981,7 +2896,7 @@ int ipmi_register_smi(struct ipmi_smi_handlers *handlers,
 	if (rv == 0)
 		rv = add_proc_entries(intf, i);
 
-	rv = ipmi_bmc_register(intf, i, sysfs_name);
+	rv = ipmi_bmc_register(intf, i);
 
  out:
 	if (rv) {
@@ -3011,12 +2926,50 @@ int ipmi_register_smi(struct ipmi_smi_handlers *handlers,
 }
 EXPORT_SYMBOL(ipmi_register_smi);
 
+static void deliver_smi_err_response(ipmi_smi_t intf,
+				     struct ipmi_smi_msg *msg,
+				     unsigned char err)
+{
+	msg->rsp[0] = msg->data[0] | 4;
+	msg->rsp[1] = msg->data[1];
+	msg->rsp[2] = err;
+	msg->rsp_size = 3;
+	/* It's an error, so it will never requeue, no need to check return. */
+	handle_one_recv_msg(intf, msg);
+}
+
 static void cleanup_smi_msgs(ipmi_smi_t intf)
 {
 	int              i;
 	struct seq_table *ent;
+	struct ipmi_smi_msg *msg;
+	struct list_head *entry;
+	struct list_head tmplist;
+
+	/* Clear out our transmit queues and hold the messages. */
+	INIT_LIST_HEAD(&tmplist);
+	list_splice_tail(&intf->hp_xmit_msgs, &tmplist);
+	list_splice_tail(&intf->xmit_msgs, &tmplist);
+
+	/* Current message first, to preserve order */
+	while (intf->curr_msg && !list_empty(&intf->waiting_rcv_msgs)) {
+		/* Wait for the message to clear out. */
+		schedule_timeout(1);
+	}
 
 	/* No need for locks, the interface is down. */
+
+	/*
+	 * Return errors for all pending messages in queue and in the
+	 * tables waiting for remote responses.
+	 */
+	while (!list_empty(&tmplist)) {
+		entry = tmplist.next;
+		list_del(entry);
+		msg = list_entry(entry, struct ipmi_smi_msg, link);
+		deliver_smi_err_response(intf, msg, IPMI_ERR_UNSPECIFIED);
+	}
+
 	for (i = 0; i < IPMI_IPMB_NUM_SEQ; i++) {
 		ent = &(intf->seq_table[i]);
 		if (!ent->inuse)
@@ -3028,20 +2981,33 @@ static void cleanup_smi_msgs(ipmi_smi_t intf)
 int ipmi_unregister_smi(ipmi_smi_t intf)
 {
 	struct ipmi_smi_watcher *w;
-	int    intf_num = intf->intf_num;
+	int intf_num = intf->intf_num;
+	ipmi_user_t user;
 
 	ipmi_bmc_unregister(intf);
 
 	mutex_lock(&smi_watchers_mutex);
 	mutex_lock(&ipmi_interfaces_mutex);
 	intf->intf_num = -1;
-	intf->handlers = NULL;
+	intf->in_shutdown = true;
 	list_del_rcu(&intf->link);
 	mutex_unlock(&ipmi_interfaces_mutex);
 	synchronize_rcu();
 
 	cleanup_smi_msgs(intf);
 
+	/* Clean up the effects of users on the lower-level software. */
+	mutex_lock(&ipmi_interfaces_mutex);
+	rcu_read_lock();
+	list_for_each_entry_rcu(user, &intf->users, link) {
+		module_put(intf->handlers->owner);
+		if (intf->handlers->dec_usecount)
+			intf->handlers->dec_usecount(intf->send_info);
+	}
+	rcu_read_unlock();
+	intf->handlers = NULL;
+	mutex_unlock(&ipmi_interfaces_mutex);
+
 	remove_proc_entries(intf);
 
 	/*
@@ -3131,7 +3097,6 @@ static int handle_ipmb_get_msg_cmd(ipmi_smi_t          intf,
 	ipmi_user_t              user = NULL;
 	struct ipmi_ipmb_addr    *ipmb_addr;
 	struct ipmi_recv_msg     *recv_msg;
-	struct ipmi_smi_handlers *handlers;
 
 	if (msg->rsp_size < 10) {
 		/* Message not big enough, just ignore it. */
@@ -3185,9 +3150,8 @@ static int handle_ipmb_get_msg_cmd(ipmi_smi_t          intf,
 	}
 #endif
 		rcu_read_lock();
-		handlers = intf->handlers;
-		if (handlers) {
-			handlers->sender(intf->send_info, msg, 0);
+		if (!intf->in_shutdown) {
+			smi_send(intf, intf->handlers, msg, 0);
 			/*
 			 * We used the message, so return the value
 			 * that causes it to not be freed or
@@ -3854,32 +3818,32 @@ static void handle_new_recv_msgs(ipmi_smi_t intf)
 
 	/* See if any waiting messages need to be processed. */
 	if (!run_to_completion)
-		spin_lock_irqsave(&intf->waiting_msgs_lock, flags);
-	while (!list_empty(&intf->waiting_msgs)) {
-		smi_msg = list_entry(intf->waiting_msgs.next,
+		spin_lock_irqsave(&intf->waiting_rcv_msgs_lock, flags);
+	while (!list_empty(&intf->waiting_rcv_msgs)) {
+		smi_msg = list_entry(intf->waiting_rcv_msgs.next,
 				     struct ipmi_smi_msg, link);
-		list_del(&smi_msg->link);
 		if (!run_to_completion)
-			spin_unlock_irqrestore(&intf->waiting_msgs_lock, flags);
+			spin_unlock_irqrestore(&intf->waiting_rcv_msgs_lock,
+					       flags);
 		rv = handle_one_recv_msg(intf, smi_msg);
 		if (!run_to_completion)
-			spin_lock_irqsave(&intf->waiting_msgs_lock, flags);
-		if (rv == 0) {
-			/* Message handled */
-			ipmi_free_smi_msg(smi_msg);
-		} else if (rv < 0) {
-			/* Fatal error on the message, del but don't free. */
-		} else {
+			spin_lock_irqsave(&intf->waiting_rcv_msgs_lock, flags);
+		if (rv > 0) {
 			/*
 			 * To preserve message order, quit if we
 			 * can't handle a message.
 			 */
-			list_add(&smi_msg->link, &intf->waiting_msgs);
 			break;
+		} else {
+			list_del(&smi_msg->link);
+			if (rv == 0)
+				/* Message handled */
+				ipmi_free_smi_msg(smi_msg);
+			/* If rv < 0, fatal error, del but don't free. */
 		}
 	}
 	if (!run_to_completion)
-		spin_unlock_irqrestore(&intf->waiting_msgs_lock, flags);
+		spin_unlock_irqrestore(&intf->waiting_rcv_msgs_lock, flags);
 
 	/*
 	 * If the pretimout count is non-zero, decrement one from it and
@@ -3900,7 +3864,41 @@ static void handle_new_recv_msgs(ipmi_smi_t intf)
 
 static void smi_recv_tasklet(unsigned long val)
 {
-	handle_new_recv_msgs((ipmi_smi_t) val);
+	unsigned long flags = 0; /* keep us warning-free. */
+	ipmi_smi_t intf = (ipmi_smi_t) val;
+	int run_to_completion = intf->run_to_completion;
+	struct ipmi_smi_msg *newmsg = NULL;
+
+	/*
+	 * Start the next message if available.
+	 *
+	 * Do this here, not in the actual receiver, because we may deadlock
+	 * because the lower layer is allowed to hold locks while calling
+	 * message delivery.
+	 */
+	if (!run_to_completion)
+		spin_lock_irqsave(&intf->xmit_msgs_lock, flags);
+	if (intf->curr_msg == NULL && !intf->in_shutdown) {
+		struct list_head *entry = NULL;
+
+		/* Pick the high priority queue first. */
+		if (!list_empty(&intf->hp_xmit_msgs))
+			entry = intf->hp_xmit_msgs.next;
+		else if (!list_empty(&intf->xmit_msgs))
+			entry = intf->xmit_msgs.next;
+
+		if (entry) {
+			list_del(entry);
+			newmsg = list_entry(entry, struct ipmi_smi_msg, link);
+			intf->curr_msg = newmsg;
+		}
+	}
+	if (!run_to_completion)
+		spin_unlock_irqrestore(&intf->xmit_msgs_lock, flags);
+	if (newmsg)
+		intf->handlers->sender(intf->send_info, newmsg);
+
+	handle_new_recv_msgs(intf);
 }
 
 /* Handle a new message from the lower layer. */
@@ -3908,13 +3906,16 @@ void ipmi_smi_msg_received(ipmi_smi_t          intf,
 			   struct ipmi_smi_msg *msg)
 {
 	unsigned long flags = 0; /* keep us warning-free. */
-	int           run_to_completion;
-
+	int run_to_completion = intf->run_to_completion;
 
 	if ((msg->data_size >= 2)
 	    && (msg->data[0] == (IPMI_NETFN_APP_REQUEST << 2))
 	    && (msg->data[1] == IPMI_SEND_MSG_CMD)
 	    && (msg->user_data == NULL)) {
+
+		if (intf->in_shutdown)
+			goto free_msg;
+
 		/*
 		 * This is the local response to a command send, start
 		 * the timer for these.  The user_data will not be
@@ -3950,29 +3951,40 @@ void ipmi_smi_msg_received(ipmi_smi_t          intf,
 			/* The message was sent, start the timer. */
 			intf_start_seq_timer(intf, msg->msgid);
 
+free_msg:
 		ipmi_free_smi_msg(msg);
-		goto out;
+	} else {
+		/*
+		 * To preserve message order, we keep a queue and deliver from
+		 * a tasklet.
+		 */
+		if (!run_to_completion)
+			spin_lock_irqsave(&intf->waiting_rcv_msgs_lock, flags);
+		list_add_tail(&msg->link, &intf->waiting_rcv_msgs);
+		if (!run_to_completion)
+			spin_unlock_irqrestore(&intf->waiting_rcv_msgs_lock,
+					       flags);
 	}
 
-	/*
-	 * To preserve message order, if the list is not empty, we
-	 * tack this message onto the end of the list.
-	 */
-	run_to_completion = intf->run_to_completion;
 	if (!run_to_completion)
-		spin_lock_irqsave(&intf->waiting_msgs_lock, flags);
-	list_add_tail(&msg->link, &intf->waiting_msgs);
+		spin_lock_irqsave(&intf->xmit_msgs_lock, flags);
+	if (msg == intf->curr_msg)
+		intf->curr_msg = NULL;
 	if (!run_to_completion)
-		spin_unlock_irqrestore(&intf->waiting_msgs_lock, flags);
+		spin_unlock_irqrestore(&intf->xmit_msgs_lock, flags);
 
-	tasklet_schedule(&intf->recv_tasklet);
- out:
-	return;
+	if (run_to_completion)
+		smi_recv_tasklet((unsigned long) intf);
+	else
+		tasklet_schedule(&intf->recv_tasklet);
 }
 EXPORT_SYMBOL(ipmi_smi_msg_received);
 
 void ipmi_smi_watchdog_pretimeout(ipmi_smi_t intf)
 {
+	if (intf->in_shutdown)
+		return;
+
 	atomic_set(&intf->watchdog_pretimeouts_to_deliver, 1);
 	tasklet_schedule(&intf->recv_tasklet);
 }
@@ -4014,7 +4026,7 @@ static void check_msg_timeout(ipmi_smi_t intf, struct seq_table *ent,
 	struct ipmi_recv_msg     *msg;
 	struct ipmi_smi_handlers *handlers;
 
-	if (intf->intf_num == -1)
+	if (intf->in_shutdown)
 		return;
 
 	if (!ent->inuse)
@@ -4079,8 +4091,7 @@ static void check_msg_timeout(ipmi_smi_t intf, struct seq_table *ent,
 				ipmi_inc_stat(intf,
 					      retransmitted_ipmb_commands);
 
-			intf->handlers->sender(intf->send_info,
-					       smi_msg, 0);
+			smi_send(intf, intf->handlers, smi_msg, 0);
 		} else
 			ipmi_free_smi_msg(smi_msg);
 
@@ -4142,15 +4153,12 @@ static unsigned int ipmi_timeout_handler(ipmi_smi_t intf, long timeout_period)
 
 static void ipmi_request_event(ipmi_smi_t intf)
 {
-	struct ipmi_smi_handlers *handlers;
-
 	/* No event requests when in maintenance mode. */
 	if (intf->maintenance_mode_enable)
 		return;
 
-	handlers = intf->handlers;
-	if (handlers)
-		handlers->request_events(intf->send_info);
+	if (!intf->in_shutdown)
+		intf->handlers->request_events(intf->send_info);
 }
 
 static struct timer_list ipmi_timer;
@@ -4545,6 +4553,7 @@ static int ipmi_init_msghandler(void)
 	proc_ipmi_root = proc_mkdir("ipmi", NULL);
 	if (!proc_ipmi_root) {
 	    printk(KERN_ERR PFX "Unable to create IPMI proc dir");
+	    driver_unregister(&ipmidriver.driver);
 	    return -ENOMEM;
 	}
 
diff --git a/drivers/char/ipmi/ipmi_powernv.c b/drivers/char/ipmi/ipmi_powernv.c
new file mode 100644
index 000000000000..79524ed2a3cb
--- /dev/null
+++ b/drivers/char/ipmi/ipmi_powernv.c
@@ -0,0 +1,310 @@
+/*
+ * PowerNV OPAL IPMI driver
+ *
+ * Copyright 2014 IBM Corp.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the Free
+ * Software Foundation; either version 2 of the License, or (at your option)
+ * any later version.
+ */
+
+#define pr_fmt(fmt)        "ipmi-powernv: " fmt
+
+#include <linux/ipmi_smi.h>
+#include <linux/list.h>
+#include <linux/module.h>
+#include <linux/of.h>
+
+#include <asm/opal.h>
+
+
+struct ipmi_smi_powernv {
+	u64			interface_id;
+	struct ipmi_device_id	ipmi_id;
+	ipmi_smi_t		intf;
+	u64			event;
+	struct notifier_block	event_nb;
+
+	/**
+	 * We assume that there can only be one outstanding request, so
+	 * keep the pending message in cur_msg. We protect this from concurrent
+	 * updates through send & recv calls, (and consequently opal_msg, which
+	 * is in-use when cur_msg is set) with msg_lock
+	 */
+	spinlock_t		msg_lock;
+	struct ipmi_smi_msg	*cur_msg;
+	struct opal_ipmi_msg	*opal_msg;
+};
+
+static int ipmi_powernv_start_processing(void *send_info, ipmi_smi_t intf)
+{
+	struct ipmi_smi_powernv *smi = send_info;
+
+	smi->intf = intf;
+	return 0;
+}
+
+static void send_error_reply(struct ipmi_smi_powernv *smi,
+		struct ipmi_smi_msg *msg, u8 completion_code)
+{
+	msg->rsp[0] = msg->data[0] | 0x4;
+	msg->rsp[1] = msg->data[1];
+	msg->rsp[2] = completion_code;
+	msg->rsp_size = 3;
+	ipmi_smi_msg_received(smi->intf, msg);
+}
+
+static void ipmi_powernv_send(void *send_info, struct ipmi_smi_msg *msg)
+{
+	struct ipmi_smi_powernv *smi = send_info;
+	struct opal_ipmi_msg *opal_msg;
+	unsigned long flags;
+	int comp, rc;
+	size_t size;
+
+	/* ensure data_len will fit in the opal_ipmi_msg buffer... */
+	if (msg->data_size > IPMI_MAX_MSG_LENGTH) {
+		comp = IPMI_REQ_LEN_EXCEEDED_ERR;
+		goto err;
+	}
+
+	/* ... and that we at least have netfn and cmd bytes */
+	if (msg->data_size < 2) {
+		comp = IPMI_REQ_LEN_INVALID_ERR;
+		goto err;
+	}
+
+	spin_lock_irqsave(&smi->msg_lock, flags);
+
+	if (smi->cur_msg) {
+		comp = IPMI_NODE_BUSY_ERR;
+		goto err_unlock;
+	}
+
+	/* format our data for the OPAL API */
+	opal_msg = smi->opal_msg;
+	opal_msg->version = OPAL_IPMI_MSG_FORMAT_VERSION_1;
+	opal_msg->netfn = msg->data[0];
+	opal_msg->cmd = msg->data[1];
+	if (msg->data_size > 2)
+		memcpy(opal_msg->data, msg->data + 2, msg->data_size - 2);
+
+	/* data_size already includes the netfn and cmd bytes */
+	size = sizeof(*opal_msg) + msg->data_size - 2;
+
+	pr_devel("%s: opal_ipmi_send(0x%llx, %p, %ld)\n", __func__,
+			smi->interface_id, opal_msg, size);
+	rc = opal_ipmi_send(smi->interface_id, opal_msg, size);
+	pr_devel("%s:  -> %d\n", __func__, rc);
+
+	if (!rc) {
+		smi->cur_msg = msg;
+		spin_unlock_irqrestore(&smi->msg_lock, flags);
+		return;
+	}
+
+	comp = IPMI_ERR_UNSPECIFIED;
+err_unlock:
+	spin_unlock_irqrestore(&smi->msg_lock, flags);
+err:
+	send_error_reply(smi, msg, comp);
+}
+
+static int ipmi_powernv_recv(struct ipmi_smi_powernv *smi)
+{
+	struct opal_ipmi_msg *opal_msg;
+	struct ipmi_smi_msg *msg;
+	unsigned long flags;
+	uint64_t size;
+	int rc;
+
+	pr_devel("%s: opal_ipmi_recv(%llx, msg, sz)\n", __func__,
+			smi->interface_id);
+
+	spin_lock_irqsave(&smi->msg_lock, flags);
+
+	if (!smi->cur_msg) {
+		pr_warn("no current message?\n");
+		return 0;
+	}
+
+	msg = smi->cur_msg;
+	opal_msg = smi->opal_msg;
+
+	size = cpu_to_be64(sizeof(*opal_msg) + IPMI_MAX_MSG_LENGTH);
+
+	rc = opal_ipmi_recv(smi->interface_id,
+			opal_msg,
+			&size);
+	size = be64_to_cpu(size);
+	pr_devel("%s:   -> %d (size %lld)\n", __func__,
+			rc, rc == 0 ? size : 0);
+	if (rc) {
+		spin_unlock_irqrestore(&smi->msg_lock, flags);
+		ipmi_free_smi_msg(msg);
+		return 0;
+	}
+
+	if (size < sizeof(*opal_msg)) {
+		spin_unlock_irqrestore(&smi->msg_lock, flags);
+		pr_warn("unexpected IPMI message size %lld\n", size);
+		return 0;
+	}
+
+	if (opal_msg->version != OPAL_IPMI_MSG_FORMAT_VERSION_1) {
+		spin_unlock_irqrestore(&smi->msg_lock, flags);
+		pr_warn("unexpected IPMI message format (version %d)\n",
+				opal_msg->version);
+		return 0;
+	}
+
+	msg->rsp[0] = opal_msg->netfn;
+	msg->rsp[1] = opal_msg->cmd;
+	if (size > sizeof(*opal_msg))
+		memcpy(&msg->rsp[2], opal_msg->data, size - sizeof(*opal_msg));
+	msg->rsp_size = 2 + size - sizeof(*opal_msg);
+
+	smi->cur_msg = NULL;
+	spin_unlock_irqrestore(&smi->msg_lock, flags);
+	ipmi_smi_msg_received(smi->intf, msg);
+	return 0;
+}
+
+static void ipmi_powernv_request_events(void *send_info)
+{
+}
+
+static void ipmi_powernv_set_run_to_completion(void *send_info,
+		bool run_to_completion)
+{
+}
+
+static void ipmi_powernv_poll(void *send_info)
+{
+	struct ipmi_smi_powernv *smi = send_info;
+
+	ipmi_powernv_recv(smi);
+}
+
+static struct ipmi_smi_handlers ipmi_powernv_smi_handlers = {
+	.owner			= THIS_MODULE,
+	.start_processing	= ipmi_powernv_start_processing,
+	.sender			= ipmi_powernv_send,
+	.request_events		= ipmi_powernv_request_events,
+	.set_run_to_completion	= ipmi_powernv_set_run_to_completion,
+	.poll			= ipmi_powernv_poll,
+};
+
+static int ipmi_opal_event(struct notifier_block *nb,
+			  unsigned long events, void *change)
+{
+	struct ipmi_smi_powernv *smi = container_of(nb,
+					struct ipmi_smi_powernv, event_nb);
+
+	if (events & smi->event)
+		ipmi_powernv_recv(smi);
+	return 0;
+}
+
+static int ipmi_powernv_probe(struct platform_device *pdev)
+{
+	struct ipmi_smi_powernv *ipmi;
+	struct device *dev;
+	u32 prop;
+	int rc;
+
+	if (!pdev || !pdev->dev.of_node)
+		return -ENODEV;
+
+	dev = &pdev->dev;
+
+	ipmi = devm_kzalloc(dev, sizeof(*ipmi), GFP_KERNEL);
+	if (!ipmi)
+		return -ENOMEM;
+
+	spin_lock_init(&ipmi->msg_lock);
+
+	rc = of_property_read_u32(dev->of_node, "ibm,ipmi-interface-id",
+			&prop);
+	if (rc) {
+		dev_warn(dev, "No interface ID property\n");
+		goto err_free;
+	}
+	ipmi->interface_id = prop;
+
+	rc = of_property_read_u32(dev->of_node, "interrupts", &prop);
+	if (rc) {
+		dev_warn(dev, "No interrupts property\n");
+		goto err_free;
+	}
+
+	ipmi->event = 1ull << prop;
+	ipmi->event_nb.notifier_call = ipmi_opal_event;
+
+	rc = opal_notifier_register(&ipmi->event_nb);
+	if (rc) {
+		dev_warn(dev, "OPAL notifier registration failed (%d)\n", rc);
+		goto err_free;
+	}
+
+	ipmi->opal_msg = devm_kmalloc(dev,
+			sizeof(*ipmi->opal_msg) + IPMI_MAX_MSG_LENGTH,
+			GFP_KERNEL);
+	if (!ipmi->opal_msg) {
+		rc = -ENOMEM;
+		goto err_unregister;
+	}
+
+	/* todo: query actual ipmi_device_id */
+	rc = ipmi_register_smi(&ipmi_powernv_smi_handlers, ipmi,
+			&ipmi->ipmi_id, dev, 0);
+	if (rc) {
+		dev_warn(dev, "IPMI SMI registration failed (%d)\n", rc);
+		goto err_free_msg;
+	}
+
+	dev_set_drvdata(dev, ipmi);
+	return 0;
+
+err_free_msg:
+	devm_kfree(dev, ipmi->opal_msg);
+err_unregister:
+	opal_notifier_unregister(&ipmi->event_nb);
+err_free:
+	devm_kfree(dev, ipmi);
+	return rc;
+}
+
+static int ipmi_powernv_remove(struct platform_device *pdev)
+{
+	struct ipmi_smi_powernv *smi = dev_get_drvdata(&pdev->dev);
+
+	ipmi_unregister_smi(smi->intf);
+	opal_notifier_unregister(&smi->event_nb);
+	return 0;
+}
+
+static const struct of_device_id ipmi_powernv_match[] = {
+	{ .compatible = "ibm,opal-ipmi" },
+	{ },
+};
+
+
+static struct platform_driver powernv_ipmi_driver = {
+	.driver = {
+		.name		= "ipmi-powernv",
+		.owner		= THIS_MODULE,
+		.of_match_table	= ipmi_powernv_match,
+	},
+	.probe	= ipmi_powernv_probe,
+	.remove	= ipmi_powernv_remove,
+};
+
+
+module_platform_driver(powernv_ipmi_driver);
+
+MODULE_DEVICE_TABLE(of, ipmi_powernv_match);
+MODULE_DESCRIPTION("powernv IPMI driver");
+MODULE_AUTHOR("Jeremy Kerr <jk@ozlabs.org>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/char/ipmi/ipmi_si_intf.c b/drivers/char/ipmi/ipmi_si_intf.c
index 5d665680ae33..967b73aa4e66 100644
--- a/drivers/char/ipmi/ipmi_si_intf.c
+++ b/drivers/char/ipmi/ipmi_si_intf.c
@@ -92,12 +92,9 @@ enum si_intf_state {
 	SI_GETTING_FLAGS,
 	SI_GETTING_EVENTS,
 	SI_CLEARING_FLAGS,
-	SI_CLEARING_FLAGS_THEN_SET_IRQ,
 	SI_GETTING_MESSAGES,
-	SI_ENABLE_INTERRUPTS1,
-	SI_ENABLE_INTERRUPTS2,
-	SI_DISABLE_INTERRUPTS1,
-	SI_DISABLE_INTERRUPTS2
+	SI_CHECKING_ENABLES,
+	SI_SETTING_ENABLES
 	/* FIXME - add watchdog stuff. */
 };
 
@@ -111,10 +108,6 @@ enum si_type {
 };
 static char *si_to_str[] = { "kcs", "smic", "bt" };
 
-static char *ipmi_addr_src_to_str[] = { NULL, "hotmod", "hardcoded", "SPMI",
-					"ACPI", "SMBIOS", "PCI",
-					"device-tree", "default" };
-
 #define DEVICE_NAME "ipmi_si"
 
 static struct platform_driver ipmi_driver;
@@ -174,8 +167,7 @@ struct smi_info {
 	struct si_sm_handlers  *handlers;
 	enum si_type           si_type;
 	spinlock_t             si_lock;
-	struct list_head       xmit_msgs;
-	struct list_head       hp_xmit_msgs;
+	struct ipmi_smi_msg    *waiting_msg;
 	struct ipmi_smi_msg    *curr_msg;
 	enum si_intf_state     si_state;
 
@@ -254,9 +246,6 @@ struct smi_info {
 	/* The time (in jiffies) the last timeout occurred at. */
 	unsigned long       last_timeout_jiffies;
 
-	/* Used to gracefully stop the timer without race conditions. */
-	atomic_t            stop_operation;
-
 	/* Are we waiting for the events, pretimeouts, received msgs? */
 	atomic_t            need_watch;
 
@@ -268,6 +257,16 @@ struct smi_info {
 	 */
 	bool interrupt_disabled;
 
+	/*
+	 * Does the BMC support events?
+	 */
+	bool supports_event_msg_buff;
+
+	/*
+	 * Did we get an attention that we did not handle?
+	 */
+	bool got_attn;
+
 	/* From the get device id response... */
 	struct ipmi_device_id device_id;
 
@@ -332,7 +331,10 @@ static void deliver_recv_msg(struct smi_info *smi_info,
 			     struct ipmi_smi_msg *msg)
 {
 	/* Deliver the message to the upper layer. */
-	ipmi_smi_msg_received(smi_info->intf, msg);
+	if (smi_info->intf)
+		ipmi_smi_msg_received(smi_info->intf, msg);
+	else
+		ipmi_free_smi_msg(msg);
 }
 
 static void return_hosed_msg(struct smi_info *smi_info, int cCode)
@@ -356,28 +358,18 @@ static void return_hosed_msg(struct smi_info *smi_info, int cCode)
 static enum si_sm_result start_next_msg(struct smi_info *smi_info)
 {
 	int              rv;
-	struct list_head *entry = NULL;
 #ifdef DEBUG_TIMING
 	struct timeval t;
 #endif
 
-	/* Pick the high priority queue first. */
-	if (!list_empty(&(smi_info->hp_xmit_msgs))) {
-		entry = smi_info->hp_xmit_msgs.next;
-	} else if (!list_empty(&(smi_info->xmit_msgs))) {
-		entry = smi_info->xmit_msgs.next;
-	}
-
-	if (!entry) {
+	if (!smi_info->waiting_msg) {
 		smi_info->curr_msg = NULL;
 		rv = SI_SM_IDLE;
 	} else {
 		int err;
 
-		list_del(entry);
-		smi_info->curr_msg = list_entry(entry,
-						struct ipmi_smi_msg,
-						link);
+		smi_info->curr_msg = smi_info->waiting_msg;
+		smi_info->waiting_msg = NULL;
 #ifdef DEBUG_TIMING
 		do_gettimeofday(&t);
 		printk(KERN_DEBUG "**Start2: %d.%9.9d\n", t.tv_sec, t.tv_usec);
@@ -401,22 +393,7 @@ static enum si_sm_result start_next_msg(struct smi_info *smi_info)
 	return rv;
 }
 
-static void start_enable_irq(struct smi_info *smi_info)
-{
-	unsigned char msg[2];
-
-	/*
-	 * If we are enabling interrupts, we have to tell the
-	 * BMC to use them.
-	 */
-	msg[0] = (IPMI_NETFN_APP_REQUEST << 2);
-	msg[1] = IPMI_GET_BMC_GLOBAL_ENABLES_CMD;
-
-	smi_info->handlers->start_transaction(smi_info->si_sm, msg, 2);
-	smi_info->si_state = SI_ENABLE_INTERRUPTS1;
-}
-
-static void start_disable_irq(struct smi_info *smi_info)
+static void start_check_enables(struct smi_info *smi_info)
 {
 	unsigned char msg[2];
 
@@ -424,7 +401,7 @@ static void start_disable_irq(struct smi_info *smi_info)
 	msg[1] = IPMI_GET_BMC_GLOBAL_ENABLES_CMD;
 
 	smi_info->handlers->start_transaction(smi_info->si_sm, msg, 2);
-	smi_info->si_state = SI_DISABLE_INTERRUPTS1;
+	smi_info->si_state = SI_CHECKING_ENABLES;
 }
 
 static void start_clear_flags(struct smi_info *smi_info)
@@ -440,6 +417,32 @@ static void start_clear_flags(struct smi_info *smi_info)
 	smi_info->si_state = SI_CLEARING_FLAGS;
 }
 
+static void start_getting_msg_queue(struct smi_info *smi_info)
+{
+	smi_info->curr_msg->data[0] = (IPMI_NETFN_APP_REQUEST << 2);
+	smi_info->curr_msg->data[1] = IPMI_GET_MSG_CMD;
+	smi_info->curr_msg->data_size = 2;
+
+	smi_info->handlers->start_transaction(
+		smi_info->si_sm,
+		smi_info->curr_msg->data,
+		smi_info->curr_msg->data_size);
+	smi_info->si_state = SI_GETTING_MESSAGES;
+}
+
+static void start_getting_events(struct smi_info *smi_info)
+{
+	smi_info->curr_msg->data[0] = (IPMI_NETFN_APP_REQUEST << 2);
+	smi_info->curr_msg->data[1] = IPMI_READ_EVENT_MSG_BUFFER_CMD;
+	smi_info->curr_msg->data_size = 2;
+
+	smi_info->handlers->start_transaction(
+		smi_info->si_sm,
+		smi_info->curr_msg->data,
+		smi_info->curr_msg->data_size);
+	smi_info->si_state = SI_GETTING_EVENTS;
+}
+
 static void smi_mod_timer(struct smi_info *smi_info, unsigned long new_val)
 {
 	smi_info->last_timeout_jiffies = jiffies;
@@ -453,22 +456,45 @@ static void smi_mod_timer(struct smi_info *smi_info, unsigned long new_val)
  * polled until we can allocate some memory.  Once we have some
  * memory, we will re-enable the interrupt.
  */
-static inline void disable_si_irq(struct smi_info *smi_info)
+static inline bool disable_si_irq(struct smi_info *smi_info)
 {
 	if ((smi_info->irq) && (!smi_info->interrupt_disabled)) {
-		start_disable_irq(smi_info);
 		smi_info->interrupt_disabled = true;
-		if (!atomic_read(&smi_info->stop_operation))
-			smi_mod_timer(smi_info, jiffies + SI_TIMEOUT_JIFFIES);
+		start_check_enables(smi_info);
+		return true;
 	}
+	return false;
 }
 
-static inline void enable_si_irq(struct smi_info *smi_info)
+static inline bool enable_si_irq(struct smi_info *smi_info)
 {
 	if ((smi_info->irq) && (smi_info->interrupt_disabled)) {
-		start_enable_irq(smi_info);
 		smi_info->interrupt_disabled = false;
+		start_check_enables(smi_info);
+		return true;
 	}
+	return false;
+}
+
+/*
+ * Allocate a message.  If unable to allocate, start the interrupt
+ * disable process and return NULL.  If able to allocate but
+ * interrupts are disabled, free the message and return NULL after
+ * starting the interrupt enable process.
+ */
+static struct ipmi_smi_msg *alloc_msg_handle_irq(struct smi_info *smi_info)
+{
+	struct ipmi_smi_msg *msg;
+
+	msg = ipmi_alloc_smi_msg();
+	if (!msg) {
+		if (!disable_si_irq(smi_info))
+			smi_info->si_state = SI_NORMAL;
+	} else if (enable_si_irq(smi_info)) {
+		ipmi_free_smi_msg(msg);
+		msg = NULL;
+	}
+	return msg;
 }
 
 static void handle_flags(struct smi_info *smi_info)
@@ -480,45 +506,22 @@ static void handle_flags(struct smi_info *smi_info)
 
 		start_clear_flags(smi_info);
 		smi_info->msg_flags &= ~WDT_PRE_TIMEOUT_INT;
-		ipmi_smi_watchdog_pretimeout(smi_info->intf);
+		if (smi_info->intf)
+			ipmi_smi_watchdog_pretimeout(smi_info->intf);
 	} else if (smi_info->msg_flags & RECEIVE_MSG_AVAIL) {
 		/* Messages available. */
-		smi_info->curr_msg = ipmi_alloc_smi_msg();
-		if (!smi_info->curr_msg) {
-			disable_si_irq(smi_info);
-			smi_info->si_state = SI_NORMAL;
+		smi_info->curr_msg = alloc_msg_handle_irq(smi_info);
+		if (!smi_info->curr_msg)
 			return;
-		}
-		enable_si_irq(smi_info);
-
-		smi_info->curr_msg->data[0] = (IPMI_NETFN_APP_REQUEST << 2);
-		smi_info->curr_msg->data[1] = IPMI_GET_MSG_CMD;
-		smi_info->curr_msg->data_size = 2;
 
-		smi_info->handlers->start_transaction(
-			smi_info->si_sm,
-			smi_info->curr_msg->data,
-			smi_info->curr_msg->data_size);
-		smi_info->si_state = SI_GETTING_MESSAGES;
+		start_getting_msg_queue(smi_info);
 	} else if (smi_info->msg_flags & EVENT_MSG_BUFFER_FULL) {
 		/* Events available. */
-		smi_info->curr_msg = ipmi_alloc_smi_msg();
-		if (!smi_info->curr_msg) {
-			disable_si_irq(smi_info);
-			smi_info->si_state = SI_NORMAL;
+		smi_info->curr_msg = alloc_msg_handle_irq(smi_info);
+		if (!smi_info->curr_msg)
 			return;
-		}
-		enable_si_irq(smi_info);
 
-		smi_info->curr_msg->data[0] = (IPMI_NETFN_APP_REQUEST << 2);
-		smi_info->curr_msg->data[1] = IPMI_READ_EVENT_MSG_BUFFER_CMD;
-		smi_info->curr_msg->data_size = 2;
-
-		smi_info->handlers->start_transaction(
-			smi_info->si_sm,
-			smi_info->curr_msg->data,
-			smi_info->curr_msg->data_size);
-		smi_info->si_state = SI_GETTING_EVENTS;
+		start_getting_events(smi_info);
 	} else if (smi_info->msg_flags & OEM_DATA_AVAIL &&
 		   smi_info->oem_data_avail_handler) {
 		if (smi_info->oem_data_avail_handler(smi_info))
@@ -527,6 +530,55 @@ static void handle_flags(struct smi_info *smi_info)
 		smi_info->si_state = SI_NORMAL;
 }
 
+/*
+ * Global enables we care about.
+ */
+#define GLOBAL_ENABLES_MASK (IPMI_BMC_EVT_MSG_BUFF | IPMI_BMC_RCV_MSG_INTR | \
+			     IPMI_BMC_EVT_MSG_INTR)
+
+static u8 current_global_enables(struct smi_info *smi_info, u8 base,
+				 bool *irq_on)
+{
+	u8 enables = 0;
+
+	if (smi_info->supports_event_msg_buff)
+		enables |= IPMI_BMC_EVT_MSG_BUFF;
+	else
+		enables &= ~IPMI_BMC_EVT_MSG_BUFF;
+
+	if (smi_info->irq && !smi_info->interrupt_disabled)
+		enables |= IPMI_BMC_RCV_MSG_INTR;
+	else
+		enables &= ~IPMI_BMC_RCV_MSG_INTR;
+
+	if (smi_info->supports_event_msg_buff &&
+	    smi_info->irq && !smi_info->interrupt_disabled)
+
+		enables |= IPMI_BMC_EVT_MSG_INTR;
+	else
+		enables &= ~IPMI_BMC_EVT_MSG_INTR;
+
+	*irq_on = enables & (IPMI_BMC_EVT_MSG_INTR | IPMI_BMC_RCV_MSG_INTR);
+
+	return enables;
+}
+
+static void check_bt_irq(struct smi_info *smi_info, bool irq_on)
+{
+	u8 irqstate = smi_info->io.inputb(&smi_info->io, IPMI_BT_INTMASK_REG);
+
+	irqstate &= IPMI_BT_INTMASK_ENABLE_IRQ_BIT;
+
+	if ((bool)irqstate == irq_on)
+		return;
+
+	if (irq_on)
+		smi_info->io.outputb(&smi_info->io, IPMI_BT_INTMASK_REG,
+				     IPMI_BT_INTMASK_ENABLE_IRQ_BIT);
+	else
+		smi_info->io.outputb(&smi_info->io, IPMI_BT_INTMASK_REG, 0);
+}
+
 static void handle_transaction_done(struct smi_info *smi_info)
 {
 	struct ipmi_smi_msg *msg;
@@ -581,7 +633,6 @@ static void handle_transaction_done(struct smi_info *smi_info)
 	}
 
 	case SI_CLEARING_FLAGS:
-	case SI_CLEARING_FLAGS_THEN_SET_IRQ:
 	{
 		unsigned char msg[3];
 
@@ -592,10 +643,7 @@ static void handle_transaction_done(struct smi_info *smi_info)
 			dev_warn(smi_info->dev,
 				 "Error clearing flags: %2.2x\n", msg[2]);
 		}
-		if (smi_info->si_state == SI_CLEARING_FLAGS_THEN_SET_IRQ)
-			start_enable_irq(smi_info);
-		else
-			smi_info->si_state = SI_NORMAL;
+		smi_info->si_state = SI_NORMAL;
 		break;
 	}
 
@@ -675,9 +723,11 @@ static void handle_transaction_done(struct smi_info *smi_info)
 		break;
 	}
 
-	case SI_ENABLE_INTERRUPTS1:
+	case SI_CHECKING_ENABLES:
 	{
 		unsigned char msg[4];
+		u8 enables;
+		bool irq_on;
 
 		/* We got the flags from the SMI, now handle them. */
 		smi_info->handlers->get_result(smi_info->si_sm, msg, 4);
@@ -687,70 +737,53 @@ static void handle_transaction_done(struct smi_info *smi_info)
 			dev_warn(smi_info->dev,
 				 "Maybe ok, but ipmi might run very slowly.\n");
 			smi_info->si_state = SI_NORMAL;
-		} else {
+			break;
+		}
+		enables = current_global_enables(smi_info, 0, &irq_on);
+		if (smi_info->si_type == SI_BT)
+			/* BT has its own interrupt enable bit. */
+			check_bt_irq(smi_info, irq_on);
+		if (enables != (msg[3] & GLOBAL_ENABLES_MASK)) {
+			/* Enables are not correct, fix them. */
 			msg[0] = (IPMI_NETFN_APP_REQUEST << 2);
 			msg[1] = IPMI_SET_BMC_GLOBAL_ENABLES_CMD;
-			msg[2] = (msg[3] |
-				  IPMI_BMC_RCV_MSG_INTR |
-				  IPMI_BMC_EVT_MSG_INTR);
+			msg[2] = enables | (msg[3] & ~GLOBAL_ENABLES_MASK);
 			smi_info->handlers->start_transaction(
 				smi_info->si_sm, msg, 3);
-			smi_info->si_state = SI_ENABLE_INTERRUPTS2;
+			smi_info->si_state = SI_SETTING_ENABLES;
+		} else if (smi_info->supports_event_msg_buff) {
+			smi_info->curr_msg = ipmi_alloc_smi_msg();
+			if (!smi_info->curr_msg) {
+				smi_info->si_state = SI_NORMAL;
+				break;
+			}
+			start_getting_msg_queue(smi_info);
+		} else {
+			smi_info->si_state = SI_NORMAL;
 		}
 		break;
 	}
 
-	case SI_ENABLE_INTERRUPTS2:
+	case SI_SETTING_ENABLES:
 	{
 		unsigned char msg[4];
 
-		/* We got the flags from the SMI, now handle them. */
 		smi_info->handlers->get_result(smi_info->si_sm, msg, 4);
-		if (msg[2] != 0) {
-			dev_warn(smi_info->dev,
-				 "Couldn't set irq info: %x.\n", msg[2]);
+		if (msg[2] != 0)
 			dev_warn(smi_info->dev,
-				 "Maybe ok, but ipmi might run very slowly.\n");
-		} else
-			smi_info->interrupt_disabled = false;
-		smi_info->si_state = SI_NORMAL;
-		break;
-	}
-
-	case SI_DISABLE_INTERRUPTS1:
-	{
-		unsigned char msg[4];
+				 "Could not set the global enables: 0x%x.\n",
+				 msg[2]);
 
-		/* We got the flags from the SMI, now handle them. */
-		smi_info->handlers->get_result(smi_info->si_sm, msg, 4);
-		if (msg[2] != 0) {
-			dev_warn(smi_info->dev, "Could not disable interrupts"
-				 ", failed get.\n");
-			smi_info->si_state = SI_NORMAL;
+		if (smi_info->supports_event_msg_buff) {
+			smi_info->curr_msg = ipmi_alloc_smi_msg();
+			if (!smi_info->curr_msg) {
+				smi_info->si_state = SI_NORMAL;
+				break;
+			}
+			start_getting_msg_queue(smi_info);
 		} else {
-			msg[0] = (IPMI_NETFN_APP_REQUEST << 2);
-			msg[1] = IPMI_SET_BMC_GLOBAL_ENABLES_CMD;
-			msg[2] = (msg[3] &
-				  ~(IPMI_BMC_RCV_MSG_INTR |
-				    IPMI_BMC_EVT_MSG_INTR));
-			smi_info->handlers->start_transaction(
-				smi_info->si_sm, msg, 3);
-			smi_info->si_state = SI_DISABLE_INTERRUPTS2;
-		}
-		break;
-	}
-
-	case SI_DISABLE_INTERRUPTS2:
-	{
-		unsigned char msg[4];
-
-		/* We got the flags from the SMI, now handle them. */
-		smi_info->handlers->get_result(smi_info->si_sm, msg, 4);
-		if (msg[2] != 0) {
-			dev_warn(smi_info->dev, "Could not disable interrupts"
-				 ", failed set.\n");
+			smi_info->si_state = SI_NORMAL;
 		}
-		smi_info->si_state = SI_NORMAL;
 		break;
 	}
 	}
@@ -808,25 +841,35 @@ static enum si_sm_result smi_event_handler(struct smi_info *smi_info,
 	 * We prefer handling attn over new messages.  But don't do
 	 * this if there is not yet an upper layer to handle anything.
 	 */
-	if (likely(smi_info->intf) && si_sm_result == SI_SM_ATTN) {
+	if (likely(smi_info->intf) &&
+	    (si_sm_result == SI_SM_ATTN || smi_info->got_attn)) {
 		unsigned char msg[2];
 
-		smi_inc_stat(smi_info, attentions);
+		if (smi_info->si_state != SI_NORMAL) {
+			/*
+			 * We got an ATTN, but we are doing something else.
+			 * Handle the ATTN later.
+			 */
+			smi_info->got_attn = true;
+		} else {
+			smi_info->got_attn = false;
+			smi_inc_stat(smi_info, attentions);
 
-		/*
-		 * Got a attn, send down a get message flags to see
-		 * what's causing it.  It would be better to handle
-		 * this in the upper layer, but due to the way
-		 * interrupts work with the SMI, that's not really
-		 * possible.
-		 */
-		msg[0] = (IPMI_NETFN_APP_REQUEST << 2);
-		msg[1] = IPMI_GET_MSG_FLAGS_CMD;
+			/*
+			 * Got a attn, send down a get message flags to see
+			 * what's causing it.  It would be better to handle
+			 * this in the upper layer, but due to the way
+			 * interrupts work with the SMI, that's not really
+			 * possible.
+			 */
+			msg[0] = (IPMI_NETFN_APP_REQUEST << 2);
+			msg[1] = IPMI_GET_MSG_FLAGS_CMD;
 
-		smi_info->handlers->start_transaction(
-			smi_info->si_sm, msg, 2);
-		smi_info->si_state = SI_GETTING_FLAGS;
-		goto restart;
+			smi_info->handlers->start_transaction(
+				smi_info->si_sm, msg, 2);
+			smi_info->si_state = SI_GETTING_FLAGS;
+			goto restart;
+		}
 	}
 
 	/* If we are currently idle, try to start the next message. */
@@ -846,19 +889,21 @@ static enum si_sm_result smi_event_handler(struct smi_info *smi_info,
 		 */
 		atomic_set(&smi_info->req_events, 0);
 
-		smi_info->curr_msg = ipmi_alloc_smi_msg();
-		if (!smi_info->curr_msg)
-			goto out;
-
-		smi_info->curr_msg->data[0] = (IPMI_NETFN_APP_REQUEST << 2);
-		smi_info->curr_msg->data[1] = IPMI_READ_EVENT_MSG_BUFFER_CMD;
-		smi_info->curr_msg->data_size = 2;
+		/*
+		 * Take this opportunity to check the interrupt and
+		 * message enable state for the BMC.  The BMC can be
+		 * asynchronously reset, and may thus get interrupts
+		 * disable and messages disabled.
+		 */
+		if (smi_info->supports_event_msg_buff || smi_info->irq) {
+			start_check_enables(smi_info);
+		} else {
+			smi_info->curr_msg = alloc_msg_handle_irq(smi_info);
+			if (!smi_info->curr_msg)
+				goto out;
 
-		smi_info->handlers->start_transaction(
-			smi_info->si_sm,
-			smi_info->curr_msg->data,
-			smi_info->curr_msg->data_size);
-		smi_info->si_state = SI_GETTING_EVENTS;
+			start_getting_events(smi_info);
+		}
 		goto restart;
 	}
  out:
@@ -879,8 +924,7 @@ static void check_start_timer_thread(struct smi_info *smi_info)
 }
 
 static void sender(void                *send_info,
-		   struct ipmi_smi_msg *msg,
-		   int                 priority)
+		   struct ipmi_smi_msg *msg)
 {
 	struct smi_info   *smi_info = send_info;
 	enum si_sm_result result;
@@ -889,14 +933,8 @@ static void sender(void                *send_info,
 	struct timeval    t;
 #endif
 
-	if (atomic_read(&smi_info->stop_operation)) {
-		msg->rsp[0] = msg->data[0] | 4;
-		msg->rsp[1] = msg->data[1];
-		msg->rsp[2] = IPMI_ERR_UNSPECIFIED;
-		msg->rsp_size = 3;
-		deliver_recv_msg(smi_info, msg);
-		return;
-	}
+	BUG_ON(smi_info->waiting_msg);
+	smi_info->waiting_msg = msg;
 
 #ifdef DEBUG_TIMING
 	do_gettimeofday(&t);
@@ -905,16 +943,16 @@ static void sender(void                *send_info,
 
 	if (smi_info->run_to_completion) {
 		/*
-		 * If we are running to completion, then throw it in
-		 * the list and run transactions until everything is
-		 * clear.  Priority doesn't matter here.
+		 * If we are running to completion, start it and run
+		 * transactions until everything is clear.
 		 */
+		smi_info->curr_msg = smi_info->waiting_msg;
+		smi_info->waiting_msg = NULL;
 
 		/*
 		 * Run to completion means we are single-threaded, no
 		 * need for locks.
 		 */
-		list_add_tail(&(msg->link), &(smi_info->xmit_msgs));
 
 		result = smi_event_handler(smi_info, 0);
 		while (result != SI_SM_IDLE) {
@@ -926,11 +964,6 @@ static void sender(void                *send_info,
 	}
 
 	spin_lock_irqsave(&smi_info->si_lock, flags);
-	if (priority > 0)
-		list_add_tail(&msg->link, &smi_info->hp_xmit_msgs);
-	else
-		list_add_tail(&msg->link, &smi_info->xmit_msgs);
-
 	check_start_timer_thread(smi_info);
 	spin_unlock_irqrestore(&smi_info->si_lock, flags);
 }
@@ -965,9 +998,9 @@ static inline int ipmi_si_is_busy(struct timespec *ts)
 	return ts->tv_nsec != -1;
 }
 
-static int ipmi_thread_busy_wait(enum si_sm_result smi_result,
-				 const struct smi_info *smi_info,
-				 struct timespec *busy_until)
+static inline int ipmi_thread_busy_wait(enum si_sm_result smi_result,
+					const struct smi_info *smi_info,
+					struct timespec *busy_until)
 {
 	unsigned int max_busy_us = 0;
 
@@ -1068,8 +1101,7 @@ static void request_events(void *send_info)
 {
 	struct smi_info *smi_info = send_info;
 
-	if (atomic_read(&smi_info->stop_operation) ||
-				!smi_info->has_event_buffer)
+	if (!smi_info->has_event_buffer)
 		return;
 
 	atomic_set(&smi_info->req_events, 1);
@@ -1697,7 +1729,7 @@ static int parse_str(struct hotmod_vals *v, int *val, char *name, char **curr)
 	}
 	*s = '\0';
 	s++;
-	for (i = 0; hotmod_ops[i].name; i++) {
+	for (i = 0; v[i].name; i++) {
 		if (strcmp(*curr, v[i].name) == 0) {
 			*val = v[i].val;
 			*curr = s;
@@ -2133,6 +2165,9 @@ static int try_init_spmi(struct SPMITable *spmi)
 	case 3:	/* BT */
 		info->si_type = SI_BT;
 		break;
+	case 4: /* SSIF, just ignore */
+		kfree(info);
+		return -EIO;
 	default:
 		printk(KERN_INFO PFX "Unknown ACPI/SPMI SI type %d\n",
 		       spmi->InterfaceType);
@@ -2250,6 +2285,8 @@ static int ipmi_pnp_probe(struct pnp_dev *dev,
 	case 3:
 		info->si_type = SI_BT;
 		break;
+	case 4: /* SSIF, just ignore */
+		goto err_free;
 	default:
 		dev_info(&dev->dev, "unknown IPMI type %lld\n", tmp);
 		goto err_free;
@@ -2658,6 +2695,9 @@ static int ipmi_probe(struct platform_device *dev)
 	if (!match)
 		return -EINVAL;
 
+	if (!of_device_is_available(np))
+		return -EINVAL;
+
 	ret = of_address_to_resource(np, 0, &resource);
 	if (ret) {
 		dev_warn(&dev->dev, PFX "invalid address from OF\n");
@@ -2748,7 +2788,6 @@ static struct of_device_id ipmi_match[] =
 static struct platform_driver ipmi_driver = {
 	.driver = {
 		.name = DEVICE_NAME,
-		.owner = THIS_MODULE,
 		.of_match_table = ipmi_match,
 	},
 	.probe		= ipmi_probe,
@@ -2910,9 +2949,11 @@ static int try_enable_event_buffer(struct smi_info *smi_info)
 		goto out;
 	}
 
-	if (resp[3] & IPMI_BMC_EVT_MSG_BUFF)
+	if (resp[3] & IPMI_BMC_EVT_MSG_BUFF) {
 		/* buffer is already enabled, nothing to do. */
+		smi_info->supports_event_msg_buff = true;
 		goto out;
+	}
 
 	msg[0] = IPMI_NETFN_APP_REQUEST << 2;
 	msg[1] = IPMI_SET_BMC_GLOBAL_ENABLES_CMD;
@@ -2945,6 +2986,9 @@ static int try_enable_event_buffer(struct smi_info *smi_info)
 		 * that the event buffer is not supported.
 		 */
 		rv = -ENOENT;
+	else
+		smi_info->supports_event_msg_buff = true;
+
  out:
 	kfree(resp);
 	return rv;
@@ -3185,15 +3229,10 @@ static void setup_xaction_handlers(struct smi_info *smi_info)
 
 static inline void wait_for_timer_and_thread(struct smi_info *smi_info)
 {
-	if (smi_info->intf) {
-		/*
-		 * The timer and thread are only running if the
-		 * interface has been started up and registered.
-		 */
-		if (smi_info->thread != NULL)
-			kthread_stop(smi_info->thread);
+	if (smi_info->thread != NULL)
+		kthread_stop(smi_info->thread);
+	if (smi_info->timer_running)
 		del_timer_sync(&smi_info->si_timer);
-	}
 }
 
 static struct ipmi_default_vals
@@ -3271,8 +3310,8 @@ static int add_smi(struct smi_info *new_smi)
 	int rv = 0;
 
 	printk(KERN_INFO PFX "Adding %s-specified %s state machine",
-			ipmi_addr_src_to_str[new_smi->addr_source],
-			si_to_str[new_smi->si_type]);
+	       ipmi_addr_src_to_str(new_smi->addr_source),
+	       si_to_str[new_smi->si_type]);
 	mutex_lock(&smi_infos_lock);
 	if (!is_new_interface(new_smi)) {
 		printk(KERN_CONT " duplicate interface\n");
@@ -3302,7 +3341,7 @@ static int try_smi_init(struct smi_info *new_smi)
 	printk(KERN_INFO PFX "Trying %s-specified %s state"
 	       " machine at %s address 0x%lx, slave address 0x%x,"
 	       " irq %d\n",
-	       ipmi_addr_src_to_str[new_smi->addr_source],
+	       ipmi_addr_src_to_str(new_smi->addr_source),
 	       si_to_str[new_smi->si_type],
 	       addr_space_to_str[new_smi->io.addr_type],
 	       new_smi->io.addr_data,
@@ -3368,8 +3407,7 @@ static int try_smi_init(struct smi_info *new_smi)
 	setup_oem_data_handler(new_smi);
 	setup_xaction_handlers(new_smi);
 
-	INIT_LIST_HEAD(&(new_smi->xmit_msgs));
-	INIT_LIST_HEAD(&(new_smi->hp_xmit_msgs));
+	new_smi->waiting_msg = NULL;
 	new_smi->curr_msg = NULL;
 	atomic_set(&new_smi->req_events, 0);
 	new_smi->run_to_completion = false;
@@ -3377,7 +3415,6 @@ static int try_smi_init(struct smi_info *new_smi)
 		atomic_set(&new_smi->stats[i], 0);
 
 	new_smi->interrupt_disabled = true;
-	atomic_set(&new_smi->stop_operation, 0);
 	atomic_set(&new_smi->need_watch, 0);
 	new_smi->intf_num = smi_num;
 	smi_num++;
@@ -3391,9 +3428,15 @@ static int try_smi_init(struct smi_info *new_smi)
 	 * timer to avoid racing with the timer.
 	 */
 	start_clear_flags(new_smi);
-	/* IRQ is defined to be set when non-zero. */
-	if (new_smi->irq)
-		new_smi->si_state = SI_CLEARING_FLAGS_THEN_SET_IRQ;
+
+	/*
+	 * IRQ is defined to be set when non-zero.  req_events will
+	 * cause a global flags check that will enable interrupts.
+	 */
+	if (new_smi->irq) {
+		new_smi->interrupt_disabled = false;
+		atomic_set(&new_smi->req_events, 1);
+	}
 
 	if (!new_smi->dev) {
 		/*
@@ -3425,7 +3468,6 @@ static int try_smi_init(struct smi_info *new_smi)
 			       new_smi,
 			       &new_smi->device_id,
 			       new_smi->dev,
-			       "bmc",
 			       new_smi->slave_addr);
 	if (rv) {
 		dev_err(new_smi->dev, "Unable to register device: error %d\n",
@@ -3463,15 +3505,15 @@ static int try_smi_init(struct smi_info *new_smi)
 	return 0;
 
  out_err_stop_timer:
-	atomic_inc(&new_smi->stop_operation);
 	wait_for_timer_and_thread(new_smi);
 
  out_err:
 	new_smi->interrupt_disabled = true;
 
 	if (new_smi->intf) {
-		ipmi_unregister_smi(new_smi->intf);
+		ipmi_smi_t intf = new_smi->intf;
 		new_smi->intf = NULL;
+		ipmi_unregister_smi(intf);
 	}
 
 	if (new_smi->irq_cleanup) {
@@ -3650,57 +3692,49 @@ module_init(init_ipmi_si);
 static void cleanup_one_si(struct smi_info *to_clean)
 {
 	int           rv = 0;
-	unsigned long flags;
 
 	if (!to_clean)
 		return;
 
-	list_del(&to_clean->link);
+	if (to_clean->intf) {
+		ipmi_smi_t intf = to_clean->intf;
+
+		to_clean->intf = NULL;
+		rv = ipmi_unregister_smi(intf);
+		if (rv) {
+			pr_err(PFX "Unable to unregister device: errno=%d\n",
+			       rv);
+		}
+	}
+
+	if (to_clean->dev)
+		dev_set_drvdata(to_clean->dev, NULL);
 
-	/* Tell the driver that we are shutting down. */
-	atomic_inc(&to_clean->stop_operation);
+	list_del(&to_clean->link);
 
 	/*
-	 * Make sure the timer and thread are stopped and will not run
-	 * again.
+	 * Make sure that interrupts, the timer and the thread are
+	 * stopped and will not run again.
 	 */
+	if (to_clean->irq_cleanup)
+		to_clean->irq_cleanup(to_clean);
 	wait_for_timer_and_thread(to_clean);
 
 	/*
 	 * Timeouts are stopped, now make sure the interrupts are off
-	 * for the device.  A little tricky with locks to make sure
-	 * there are no races.
+	 * in the BMC.  Note that timers and CPU interrupts are off,
+	 * so no need for locks.
 	 */
-	spin_lock_irqsave(&to_clean->si_lock, flags);
 	while (to_clean->curr_msg || (to_clean->si_state != SI_NORMAL)) {
-		spin_unlock_irqrestore(&to_clean->si_lock, flags);
 		poll(to_clean);
 		schedule_timeout_uninterruptible(1);
-		spin_lock_irqsave(&to_clean->si_lock, flags);
 	}
 	disable_si_irq(to_clean);
-	spin_unlock_irqrestore(&to_clean->si_lock, flags);
 	while (to_clean->curr_msg || (to_clean->si_state != SI_NORMAL)) {
 		poll(to_clean);
 		schedule_timeout_uninterruptible(1);
 	}
 
-	/* Clean up interrupts and make sure that everything is done. */
-	if (to_clean->irq_cleanup)
-		to_clean->irq_cleanup(to_clean);
-	while (to_clean->curr_msg || (to_clean->si_state != SI_NORMAL)) {
-		poll(to_clean);
-		schedule_timeout_uninterruptible(1);
-	}
-
-	if (to_clean->intf)
-		rv = ipmi_unregister_smi(to_clean->intf);
-
-	if (rv) {
-		printk(KERN_ERR PFX "Unable to unregister device: errno=%d\n",
-		       rv);
-	}
-
 	if (to_clean->handlers)
 		to_clean->handlers->cleanup(to_clean->si_sm);
 
diff --git a/drivers/char/ipmi/ipmi_ssif.c b/drivers/char/ipmi/ipmi_ssif.c
new file mode 100644
index 000000000000..982b96323f82
--- /dev/null
+++ b/drivers/char/ipmi/ipmi_ssif.c
@@ -0,0 +1,1872 @@
+/*
+ * ipmi_ssif.c
+ *
+ * The interface to the IPMI driver for SMBus access to a SMBus
+ * compliant device.  Called SSIF by the IPMI spec.
+ *
+ * Author: Intel Corporation
+ *         Todd Davis <todd.c.davis@intel.com>
+ *
+ * Rewritten by Corey Minyard <minyard@acm.org> to support the
+ * non-blocking I2C interface, add support for multi-part
+ * transactions, add PEC support, and general clenaup.
+ *
+ * Copyright 2003 Intel Corporation
+ * Copyright 2005 MontaVista Software
+ *
+ *  This program is free software; you can redistribute it and/or modify it
+ *  under the terms of the GNU General Public License as published by the
+ *  Free Software Foundation; either version 2 of the License, or (at your
+ *  option) any later version.
+ */
+
+/*
+ * This file holds the "policy" for the interface to the SSIF state
+ * machine.  It does the configuration, handles timers and interrupts,
+ * and drives the real SSIF state machine.
+ */
+
+/*
+ * TODO: Figure out how to use SMB alerts.  This will require a new
+ * interface into the I2C driver, I believe.
+ */
+
+#include <linux/version.h>
+#if defined(MODVERSIONS)
+#include <linux/modversions.h>
+#endif
+
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/sched.h>
+#include <linux/seq_file.h>
+#include <linux/timer.h>
+#include <linux/delay.h>
+#include <linux/errno.h>
+#include <linux/spinlock.h>
+#include <linux/slab.h>
+#include <linux/list.h>
+#include <linux/i2c.h>
+#include <linux/ipmi_smi.h>
+#include <linux/init.h>
+#include <linux/dmi.h>
+#include <linux/kthread.h>
+#include <linux/acpi.h>
+#include <linux/ctype.h>
+
+#define PFX "ipmi_ssif: "
+#define DEVICE_NAME "ipmi_ssif"
+
+#define IPMI_GET_SYSTEM_INTERFACE_CAPABILITIES_CMD	0x57
+
+#define	SSIF_IPMI_REQUEST			2
+#define	SSIF_IPMI_MULTI_PART_REQUEST_START	6
+#define	SSIF_IPMI_MULTI_PART_REQUEST_MIDDLE	7
+#define	SSIF_IPMI_RESPONSE			3
+#define	SSIF_IPMI_MULTI_PART_RESPONSE_MIDDLE	9
+
+/* ssif_debug is a bit-field
+ *	SSIF_DEBUG_MSG -	commands and their responses
+ *	SSIF_DEBUG_STATES -	message states
+ *	SSIF_DEBUG_TIMING -	 Measure times between events in the driver
+ */
+#define SSIF_DEBUG_TIMING	4
+#define SSIF_DEBUG_STATE	2
+#define SSIF_DEBUG_MSG		1
+#define SSIF_NODEBUG		0
+#define SSIF_DEFAULT_DEBUG	(SSIF_NODEBUG)
+
+/*
+ * Timer values
+ */
+#define SSIF_MSG_USEC		20000	/* 20ms between message tries. */
+#define SSIF_MSG_PART_USEC	5000	/* 5ms for a message part */
+
+/* How many times to we retry sending/receiving the message. */
+#define	SSIF_SEND_RETRIES	5
+#define	SSIF_RECV_RETRIES	250
+
+#define SSIF_MSG_MSEC		(SSIF_MSG_USEC / 1000)
+#define SSIF_MSG_JIFFIES	((SSIF_MSG_USEC * 1000) / TICK_NSEC)
+#define SSIF_MSG_PART_JIFFIES	((SSIF_MSG_PART_USEC * 1000) / TICK_NSEC)
+
+enum ssif_intf_state {
+	SSIF_NORMAL,
+	SSIF_GETTING_FLAGS,
+	SSIF_GETTING_EVENTS,
+	SSIF_CLEARING_FLAGS,
+	SSIF_GETTING_MESSAGES,
+	/* FIXME - add watchdog stuff. */
+};
+
+#define SSIF_IDLE(ssif)	 ((ssif)->ssif_state == SSIF_NORMAL \
+			  && (ssif)->curr_msg == NULL)
+
+/*
+ * Indexes into stats[] in ssif_info below.
+ */
+enum ssif_stat_indexes {
+	/* Number of total messages sent. */
+	SSIF_STAT_sent_messages = 0,
+
+	/*
+	 * Number of message parts sent.  Messages may be broken into
+	 * parts if they are long.
+	 */
+	SSIF_STAT_sent_messages_parts,
+
+	/*
+	 * Number of time a message was retried.
+	 */
+	SSIF_STAT_send_retries,
+
+	/*
+	 * Number of times the send of a message failed.
+	 */
+	SSIF_STAT_send_errors,
+
+	/*
+	 * Number of message responses received.
+	 */
+	SSIF_STAT_received_messages,
+
+	/*
+	 * Number of message fragments received.
+	 */
+	SSIF_STAT_received_message_parts,
+
+	/*
+	 * Number of times the receive of a message was retried.
+	 */
+	SSIF_STAT_receive_retries,
+
+	/*
+	 * Number of errors receiving messages.
+	 */
+	SSIF_STAT_receive_errors,
+
+	/*
+	 * Number of times a flag fetch was requested.
+	 */
+	SSIF_STAT_flag_fetches,
+
+	/*
+	 * Number of times the hardware didn't follow the state machine.
+	 */
+	SSIF_STAT_hosed,
+
+	/*
+	 * Number of received events.
+	 */
+	SSIF_STAT_events,
+
+	/* Number of asyncronous messages received. */
+	SSIF_STAT_incoming_messages,
+
+	/* Number of watchdog pretimeouts. */
+	SSIF_STAT_watchdog_pretimeouts,
+
+	/* Always add statistics before this value, it must be last. */
+	SSIF_NUM_STATS
+};
+
+struct ssif_addr_info {
+	unsigned short addr;
+	struct i2c_board_info binfo;
+	char *adapter_name;
+	int debug;
+	int slave_addr;
+	enum ipmi_addr_src addr_src;
+	union ipmi_smi_info_union addr_info;
+
+	struct mutex clients_mutex;
+	struct list_head clients;
+
+	struct list_head link;
+};
+
+struct ssif_info;
+
+typedef void (*ssif_i2c_done)(struct ssif_info *ssif_info, int result,
+			     unsigned char *data, unsigned int len);
+
+struct ssif_info {
+	ipmi_smi_t          intf;
+	int                 intf_num;
+	spinlock_t	    lock;
+	struct ipmi_smi_msg *waiting_msg;
+	struct ipmi_smi_msg *curr_msg;
+	enum ssif_intf_state ssif_state;
+	unsigned long       ssif_debug;
+
+	struct ipmi_smi_handlers handlers;
+
+	enum ipmi_addr_src addr_source; /* ACPI, PCI, SMBIOS, hardcode, etc. */
+	union ipmi_smi_info_union addr_info;
+
+	/*
+	 * Flags from the last GET_MSG_FLAGS command, used when an ATTN
+	 * is set to hold the flags until we are done handling everything
+	 * from the flags.
+	 */
+#define RECEIVE_MSG_AVAIL	0x01
+#define EVENT_MSG_BUFFER_FULL	0x02
+#define WDT_PRE_TIMEOUT_INT	0x08
+	unsigned char       msg_flags;
+
+	bool		    has_event_buffer;
+
+	/*
+	 * If set to true, this will request events the next time the
+	 * state machine is idle.
+	 */
+	bool                req_events;
+
+	/*
+	 * If set to true, this will request flags the next time the
+	 * state machine is idle.
+	 */
+	bool                req_flags;
+
+	/*
+	 * Used to perform timer operations when run-to-completion
+	 * mode is on.  This is a countdown timer.
+	 */
+	int                 rtc_us_timer;
+
+	/* Used for sending/receiving data.  +1 for the length. */
+	unsigned char data[IPMI_MAX_MSG_LENGTH + 1];
+	unsigned int  data_len;
+
+	/* Temp receive buffer, gets copied into data. */
+	unsigned char recv[I2C_SMBUS_BLOCK_MAX];
+
+	struct i2c_client *client;
+	ssif_i2c_done done_handler;
+
+	/* Thread interface handling */
+	struct task_struct *thread;
+	struct completion wake_thread;
+	bool stopping;
+	int i2c_read_write;
+	int i2c_command;
+	unsigned char *i2c_data;
+	unsigned int i2c_size;
+
+	/* From the device id response. */
+	struct ipmi_device_id device_id;
+
+	struct timer_list retry_timer;
+	int retries_left;
+
+	/* Info from SSIF cmd */
+	unsigned char max_xmit_msg_size;
+	unsigned char max_recv_msg_size;
+	unsigned int  multi_support;
+	int           supports_pec;
+
+#define SSIF_NO_MULTI		0
+#define SSIF_MULTI_2_PART	1
+#define SSIF_MULTI_n_PART	2
+	unsigned char *multi_data;
+	unsigned int  multi_len;
+	unsigned int  multi_pos;
+
+	atomic_t stats[SSIF_NUM_STATS];
+};
+
+#define ssif_inc_stat(ssif, stat) \
+	atomic_inc(&(ssif)->stats[SSIF_STAT_ ## stat])
+#define ssif_get_stat(ssif, stat) \
+	((unsigned int) atomic_read(&(ssif)->stats[SSIF_STAT_ ## stat]))
+
+static bool initialized;
+
+static atomic_t next_intf = ATOMIC_INIT(0);
+
+static void return_hosed_msg(struct ssif_info *ssif_info,
+			     struct ipmi_smi_msg *msg);
+static void start_next_msg(struct ssif_info *ssif_info, unsigned long *flags);
+static int start_send(struct ssif_info *ssif_info,
+		      unsigned char   *data,
+		      unsigned int    len);
+
+static unsigned long *ipmi_ssif_lock_cond(struct ssif_info *ssif_info,
+					  unsigned long *flags)
+{
+	spin_lock_irqsave(&ssif_info->lock, *flags);
+	return flags;
+}
+
+static void ipmi_ssif_unlock_cond(struct ssif_info *ssif_info,
+				  unsigned long *flags)
+{
+	spin_unlock_irqrestore(&ssif_info->lock, *flags);
+}
+
+static void deliver_recv_msg(struct ssif_info *ssif_info,
+			     struct ipmi_smi_msg *msg)
+{
+	ipmi_smi_t    intf = ssif_info->intf;
+
+	if (!intf) {
+		ipmi_free_smi_msg(msg);
+	} else if (msg->rsp_size < 0) {
+		return_hosed_msg(ssif_info, msg);
+		pr_err(PFX
+		       "Malformed message in deliver_recv_msg: rsp_size = %d\n",
+		       msg->rsp_size);
+	} else {
+		ipmi_smi_msg_received(intf, msg);
+	}
+}
+
+static void return_hosed_msg(struct ssif_info *ssif_info,
+			     struct ipmi_smi_msg *msg)
+{
+	ssif_inc_stat(ssif_info, hosed);
+
+	/* Make it a response */
+	msg->rsp[0] = msg->data[0] | 4;
+	msg->rsp[1] = msg->data[1];
+	msg->rsp[2] = 0xFF; /* Unknown error. */
+	msg->rsp_size = 3;
+
+	deliver_recv_msg(ssif_info, msg);
+}
+
+/*
+ * Must be called with the message lock held.  This will release the
+ * message lock.  Note that the caller will check SSIF_IDLE and start a
+ * new operation, so there is no need to check for new messages to
+ * start in here.
+ */
+static void start_clear_flags(struct ssif_info *ssif_info, unsigned long *flags)
+{
+	unsigned char msg[3];
+
+	ssif_info->msg_flags &= ~WDT_PRE_TIMEOUT_INT;
+	ssif_info->ssif_state = SSIF_CLEARING_FLAGS;
+	ipmi_ssif_unlock_cond(ssif_info, flags);
+
+	/* Make sure the watchdog pre-timeout flag is not set at startup. */
+	msg[0] = (IPMI_NETFN_APP_REQUEST << 2);
+	msg[1] = IPMI_CLEAR_MSG_FLAGS_CMD;
+	msg[2] = WDT_PRE_TIMEOUT_INT;
+
+	if (start_send(ssif_info, msg, 3) != 0) {
+		/* Error, just go to normal state. */
+		ssif_info->ssif_state = SSIF_NORMAL;
+	}
+}
+
+static void start_flag_fetch(struct ssif_info *ssif_info, unsigned long *flags)
+{
+	unsigned char mb[2];
+
+	ssif_info->req_flags = false;
+	ssif_info->ssif_state = SSIF_GETTING_FLAGS;
+	ipmi_ssif_unlock_cond(ssif_info, flags);
+
+	mb[0] = (IPMI_NETFN_APP_REQUEST << 2);
+	mb[1] = IPMI_GET_MSG_FLAGS_CMD;
+	if (start_send(ssif_info, mb, 2) != 0)
+		ssif_info->ssif_state = SSIF_NORMAL;
+}
+
+static void check_start_send(struct ssif_info *ssif_info, unsigned long *flags,
+			     struct ipmi_smi_msg *msg)
+{
+	if (start_send(ssif_info, msg->data, msg->data_size) != 0) {
+		unsigned long oflags;
+
+		flags = ipmi_ssif_lock_cond(ssif_info, &oflags);
+		ssif_info->curr_msg = NULL;
+		ssif_info->ssif_state = SSIF_NORMAL;
+		ipmi_ssif_unlock_cond(ssif_info, flags);
+		ipmi_free_smi_msg(msg);
+	}
+}
+
+static void start_event_fetch(struct ssif_info *ssif_info, unsigned long *flags)
+{
+	struct ipmi_smi_msg *msg;
+
+	ssif_info->req_events = false;
+
+	msg = ipmi_alloc_smi_msg();
+	if (!msg) {
+		ssif_info->ssif_state = SSIF_NORMAL;
+		return;
+	}
+
+	ssif_info->curr_msg = msg;
+	ssif_info->ssif_state = SSIF_GETTING_EVENTS;
+	ipmi_ssif_unlock_cond(ssif_info, flags);
+
+	msg->data[0] = (IPMI_NETFN_APP_REQUEST << 2);
+	msg->data[1] = IPMI_READ_EVENT_MSG_BUFFER_CMD;
+	msg->data_size = 2;
+
+	check_start_send(ssif_info, flags, msg);
+}
+
+static void start_recv_msg_fetch(struct ssif_info *ssif_info,
+				 unsigned long *flags)
+{
+	struct ipmi_smi_msg *msg;
+
+	msg = ipmi_alloc_smi_msg();
+	if (!msg) {
+		ssif_info->ssif_state = SSIF_NORMAL;
+		return;
+	}
+
+	ssif_info->curr_msg = msg;
+	ssif_info->ssif_state = SSIF_GETTING_MESSAGES;
+	ipmi_ssif_unlock_cond(ssif_info, flags);
+
+	msg->data[0] = (IPMI_NETFN_APP_REQUEST << 2);
+	msg->data[1] = IPMI_GET_MSG_CMD;
+	msg->data_size = 2;
+
+	check_start_send(ssif_info, flags, msg);
+}
+
+/*
+ * Must be called with the message lock held.  This will release the
+ * message lock.  Note that the caller will check SSIF_IDLE and start a
+ * new operation, so there is no need to check for new messages to
+ * start in here.
+ */
+static void handle_flags(struct ssif_info *ssif_info, unsigned long *flags)
+{
+	if (ssif_info->msg_flags & WDT_PRE_TIMEOUT_INT) {
+		ipmi_smi_t intf = ssif_info->intf;
+		/* Watchdog pre-timeout */
+		ssif_inc_stat(ssif_info, watchdog_pretimeouts);
+		start_clear_flags(ssif_info, flags);
+		if (intf)
+			ipmi_smi_watchdog_pretimeout(intf);
+	} else if (ssif_info->msg_flags & RECEIVE_MSG_AVAIL)
+		/* Messages available. */
+		start_recv_msg_fetch(ssif_info, flags);
+	else if (ssif_info->msg_flags & EVENT_MSG_BUFFER_FULL)
+		/* Events available. */
+		start_event_fetch(ssif_info, flags);
+	else {
+		ssif_info->ssif_state = SSIF_NORMAL;
+		ipmi_ssif_unlock_cond(ssif_info, flags);
+	}
+}
+
+static int ipmi_ssif_thread(void *data)
+{
+	struct ssif_info *ssif_info = data;
+
+	while (!kthread_should_stop()) {
+		int result;
+
+		/* Wait for something to do */
+		wait_for_completion(&ssif_info->wake_thread);
+		init_completion(&ssif_info->wake_thread);
+
+		if (ssif_info->stopping)
+			break;
+
+		if (ssif_info->i2c_read_write == I2C_SMBUS_WRITE) {
+			result = i2c_smbus_write_block_data(
+				ssif_info->client, SSIF_IPMI_REQUEST,
+				ssif_info->i2c_data[0],
+				ssif_info->i2c_data + 1);
+			ssif_info->done_handler(ssif_info, result, NULL, 0);
+		} else {
+			result = i2c_smbus_read_block_data(
+				ssif_info->client, SSIF_IPMI_RESPONSE,
+				ssif_info->i2c_data);
+			if (result < 0)
+				ssif_info->done_handler(ssif_info, result,
+							NULL, 0);
+			else
+				ssif_info->done_handler(ssif_info, 0,
+							ssif_info->i2c_data,
+							result);
+		}
+	}
+
+	return 0;
+}
+
+static int ssif_i2c_send(struct ssif_info *ssif_info,
+			ssif_i2c_done handler,
+			int read_write, int command,
+			unsigned char *data, unsigned int size)
+{
+	ssif_info->done_handler = handler;
+
+	ssif_info->i2c_read_write = read_write;
+	ssif_info->i2c_command = command;
+	ssif_info->i2c_data = data;
+	ssif_info->i2c_size = size;
+	complete(&ssif_info->wake_thread);
+	return 0;
+}
+
+
+static void msg_done_handler(struct ssif_info *ssif_info, int result,
+			     unsigned char *data, unsigned int len);
+
+static void retry_timeout(unsigned long data)
+{
+	struct ssif_info *ssif_info = (void *) data;
+	int rv;
+
+	if (ssif_info->stopping)
+		return;
+
+	ssif_info->rtc_us_timer = 0;
+
+	rv = ssif_i2c_send(ssif_info, msg_done_handler, I2C_SMBUS_READ,
+			  SSIF_IPMI_RESPONSE,
+			  ssif_info->recv, I2C_SMBUS_BLOCK_DATA);
+	if (rv < 0) {
+		/* request failed, just return the error. */
+		if (ssif_info->ssif_debug & SSIF_DEBUG_MSG)
+			pr_info("Error from i2c_non_blocking_op(5)\n");
+
+		msg_done_handler(ssif_info, -EIO, NULL, 0);
+	}
+}
+
+static int start_resend(struct ssif_info *ssif_info);
+
+static void msg_done_handler(struct ssif_info *ssif_info, int result,
+			     unsigned char *data, unsigned int len)
+{
+	struct ipmi_smi_msg *msg;
+	unsigned long oflags, *flags;
+	int rv;
+
+	/*
+	 * We are single-threaded here, so no need for a lock until we
+	 * start messing with driver states or the queues.
+	 */
+
+	if (result < 0) {
+		ssif_info->retries_left--;
+		if (ssif_info->retries_left > 0) {
+			ssif_inc_stat(ssif_info, receive_retries);
+
+			mod_timer(&ssif_info->retry_timer,
+				  jiffies + SSIF_MSG_JIFFIES);
+			ssif_info->rtc_us_timer = SSIF_MSG_USEC;
+			return;
+		}
+
+		ssif_inc_stat(ssif_info, receive_errors);
+
+		if  (ssif_info->ssif_debug & SSIF_DEBUG_MSG)
+			pr_info("Error in msg_done_handler: %d\n", result);
+		len = 0;
+		goto continue_op;
+	}
+
+	if ((len > 1) && (ssif_info->multi_pos == 0)
+				&& (data[0] == 0x00) && (data[1] == 0x01)) {
+		/* Start of multi-part read.  Start the next transaction. */
+		int i;
+
+		ssif_inc_stat(ssif_info, received_message_parts);
+
+		/* Remove the multi-part read marker. */
+		for (i = 0; i < (len-2); i++)
+			ssif_info->data[i] = data[i+2];
+		len -= 2;
+		ssif_info->multi_len = len;
+		ssif_info->multi_pos = 1;
+
+		rv = ssif_i2c_send(ssif_info, msg_done_handler, I2C_SMBUS_READ,
+				  SSIF_IPMI_MULTI_PART_RESPONSE_MIDDLE,
+				  ssif_info->recv, I2C_SMBUS_BLOCK_DATA);
+		if (rv < 0) {
+			if (ssif_info->ssif_debug & SSIF_DEBUG_MSG)
+				pr_info("Error from i2c_non_blocking_op(1)\n");
+
+			result = -EIO;
+		} else
+			return;
+	} else if (ssif_info->multi_pos) {
+		/* Middle of multi-part read.  Start the next transaction. */
+		int i;
+		unsigned char blocknum;
+
+		if (len == 0) {
+			result = -EIO;
+			if (ssif_info->ssif_debug & SSIF_DEBUG_MSG)
+				pr_info(PFX "Middle message with no data\n");
+
+			goto continue_op;
+		}
+
+		blocknum = data[ssif_info->multi_len];
+
+		if (ssif_info->multi_len+len-1 > IPMI_MAX_MSG_LENGTH) {
+			/* Received message too big, abort the operation. */
+			result = -E2BIG;
+			if (ssif_info->ssif_debug & SSIF_DEBUG_MSG)
+				pr_info("Received message too big\n");
+
+			goto continue_op;
+		}
+
+		/* Remove the blocknum from the data. */
+		for (i = 0; i < (len-1); i++)
+			ssif_info->data[i+ssif_info->multi_len] = data[i+1];
+		len--;
+		ssif_info->multi_len += len;
+		if (blocknum == 0xff) {
+			/* End of read */
+			len = ssif_info->multi_len;
+			data = ssif_info->data;
+		} else if ((blocknum+1) != ssif_info->multi_pos) {
+			/*
+			 * Out of sequence block, just abort.  Block
+			 * numbers start at zero for the second block,
+			 * but multi_pos starts at one, so the +1.
+			 */
+			result = -EIO;
+		} else {
+			ssif_inc_stat(ssif_info, received_message_parts);
+
+			ssif_info->multi_pos++;
+
+			rv = ssif_i2c_send(ssif_info, msg_done_handler,
+					   I2C_SMBUS_READ,
+					   SSIF_IPMI_MULTI_PART_RESPONSE_MIDDLE,
+					   ssif_info->recv,
+					   I2C_SMBUS_BLOCK_DATA);
+			if (rv < 0) {
+				if (ssif_info->ssif_debug & SSIF_DEBUG_MSG)
+					pr_info(PFX
+						"Error from i2c_non_blocking_op(2)\n");
+
+				result = -EIO;
+			} else
+				return;
+		}
+	}
+
+	if (result < 0) {
+		ssif_inc_stat(ssif_info, receive_errors);
+	} else {
+		ssif_inc_stat(ssif_info, received_messages);
+		ssif_inc_stat(ssif_info, received_message_parts);
+	}
+
+
+ continue_op:
+	if (ssif_info->ssif_debug & SSIF_DEBUG_STATE)
+		pr_info(PFX "DONE 1: state = %d, result=%d.\n",
+			ssif_info->ssif_state, result);
+
+	flags = ipmi_ssif_lock_cond(ssif_info, &oflags);
+	msg = ssif_info->curr_msg;
+	if (msg) {
+		msg->rsp_size = len;
+		if (msg->rsp_size > IPMI_MAX_MSG_LENGTH)
+			msg->rsp_size = IPMI_MAX_MSG_LENGTH;
+		memcpy(msg->rsp, data, msg->rsp_size);
+		ssif_info->curr_msg = NULL;
+	}
+
+	switch (ssif_info->ssif_state) {
+	case SSIF_NORMAL:
+		ipmi_ssif_unlock_cond(ssif_info, flags);
+		if (!msg)
+			break;
+
+		if (result < 0)
+			return_hosed_msg(ssif_info, msg);
+		else
+			deliver_recv_msg(ssif_info, msg);
+		break;
+
+	case SSIF_GETTING_FLAGS:
+		/* We got the flags from the SSIF, now handle them. */
+		if ((result < 0) || (len < 4) || (data[2] != 0)) {
+			/*
+			 * Error fetching flags, or invalid length,
+			 * just give up for now.
+			 */
+			ssif_info->ssif_state = SSIF_NORMAL;
+			ipmi_ssif_unlock_cond(ssif_info, flags);
+			pr_warn(PFX "Error getting flags: %d %d, %x\n",
+			       result, len, data[2]);
+		} else if (data[0] != (IPMI_NETFN_APP_REQUEST | 1) << 2
+			   || data[1] != IPMI_GET_MSG_FLAGS_CMD) {
+			pr_warn(PFX "Invalid response getting flags: %x %x\n",
+				data[0], data[1]);
+		} else {
+			ssif_inc_stat(ssif_info, flag_fetches);
+			ssif_info->msg_flags = data[3];
+			handle_flags(ssif_info, flags);
+		}
+		break;
+
+	case SSIF_CLEARING_FLAGS:
+		/* We cleared the flags. */
+		if ((result < 0) || (len < 3) || (data[2] != 0)) {
+			/* Error clearing flags */
+			pr_warn(PFX "Error clearing flags: %d %d, %x\n",
+			       result, len, data[2]);
+		} else if (data[0] != (IPMI_NETFN_APP_REQUEST | 1) << 2
+			   || data[1] != IPMI_CLEAR_MSG_FLAGS_CMD) {
+			pr_warn(PFX "Invalid response clearing flags: %x %x\n",
+				data[0], data[1]);
+		}
+		ssif_info->ssif_state = SSIF_NORMAL;
+		ipmi_ssif_unlock_cond(ssif_info, flags);
+		break;
+
+	case SSIF_GETTING_EVENTS:
+		if ((result < 0) || (len < 3) || (msg->rsp[2] != 0)) {
+			/* Error getting event, probably done. */
+			msg->done(msg);
+
+			/* Take off the event flag. */
+			ssif_info->msg_flags &= ~EVENT_MSG_BUFFER_FULL;
+			handle_flags(ssif_info, flags);
+		} else if (msg->rsp[0] != (IPMI_NETFN_APP_REQUEST | 1) << 2
+			   || msg->rsp[1] != IPMI_READ_EVENT_MSG_BUFFER_CMD) {
+			pr_warn(PFX "Invalid response getting events: %x %x\n",
+				msg->rsp[0], msg->rsp[1]);
+			msg->done(msg);
+			/* Take off the event flag. */
+			ssif_info->msg_flags &= ~EVENT_MSG_BUFFER_FULL;
+			handle_flags(ssif_info, flags);
+		} else {
+			handle_flags(ssif_info, flags);
+			ssif_inc_stat(ssif_info, events);
+			deliver_recv_msg(ssif_info, msg);
+		}
+		break;
+
+	case SSIF_GETTING_MESSAGES:
+		if ((result < 0) || (len < 3) || (msg->rsp[2] != 0)) {
+			/* Error getting event, probably done. */
+			msg->done(msg);
+
+			/* Take off the msg flag. */
+			ssif_info->msg_flags &= ~RECEIVE_MSG_AVAIL;
+			handle_flags(ssif_info, flags);
+		} else if (msg->rsp[0] != (IPMI_NETFN_APP_REQUEST | 1) << 2
+			   || msg->rsp[1] != IPMI_GET_MSG_CMD) {
+			pr_warn(PFX "Invalid response clearing flags: %x %x\n",
+				msg->rsp[0], msg->rsp[1]);
+			msg->done(msg);
+
+			/* Take off the msg flag. */
+			ssif_info->msg_flags &= ~RECEIVE_MSG_AVAIL;
+			handle_flags(ssif_info, flags);
+		} else {
+			ssif_inc_stat(ssif_info, incoming_messages);
+			handle_flags(ssif_info, flags);
+			deliver_recv_msg(ssif_info, msg);
+		}
+		break;
+	}
+
+	flags = ipmi_ssif_lock_cond(ssif_info, &oflags);
+	if (SSIF_IDLE(ssif_info) && !ssif_info->stopping) {
+		if (ssif_info->req_events)
+			start_event_fetch(ssif_info, flags);
+		else if (ssif_info->req_flags)
+			start_flag_fetch(ssif_info, flags);
+		else
+			start_next_msg(ssif_info, flags);
+	} else
+		ipmi_ssif_unlock_cond(ssif_info, flags);
+
+	if (ssif_info->ssif_debug & SSIF_DEBUG_STATE)
+		pr_info(PFX "DONE 2: state = %d.\n", ssif_info->ssif_state);
+}
+
+static void msg_written_handler(struct ssif_info *ssif_info, int result,
+				unsigned char *data, unsigned int len)
+{
+	int rv;
+
+	/* We are single-threaded here, so no need for a lock. */
+	if (result < 0) {
+		ssif_info->retries_left--;
+		if (ssif_info->retries_left > 0) {
+			if (!start_resend(ssif_info)) {
+				ssif_inc_stat(ssif_info, send_retries);
+				return;
+			}
+			/* request failed, just return the error. */
+			ssif_inc_stat(ssif_info, send_errors);
+
+			if (ssif_info->ssif_debug & SSIF_DEBUG_MSG)
+				pr_info(PFX
+					"Out of retries in msg_written_handler\n");
+			msg_done_handler(ssif_info, -EIO, NULL, 0);
+			return;
+		}
+
+		ssif_inc_stat(ssif_info, send_errors);
+
+		/*
+		 * Got an error on transmit, let the done routine
+		 * handle it.
+		 */
+		if (ssif_info->ssif_debug & SSIF_DEBUG_MSG)
+			pr_info("Error in msg_written_handler: %d\n", result);
+
+		msg_done_handler(ssif_info, result, NULL, 0);
+		return;
+	}
+
+	if (ssif_info->multi_data) {
+		/* In the middle of a multi-data write. */
+		int left;
+
+		ssif_inc_stat(ssif_info, sent_messages_parts);
+
+		left = ssif_info->multi_len - ssif_info->multi_pos;
+		if (left > 32)
+			left = 32;
+		/* Length byte. */
+		ssif_info->multi_data[ssif_info->multi_pos] = left;
+		ssif_info->multi_pos += left;
+		if (left < 32)
+			/*
+			 * Write is finished.  Note that we must end
+			 * with a write of less than 32 bytes to
+			 * complete the transaction, even if it is
+			 * zero bytes.
+			 */
+			ssif_info->multi_data = NULL;
+
+		rv = ssif_i2c_send(ssif_info, msg_written_handler,
+				  I2C_SMBUS_WRITE,
+				  SSIF_IPMI_MULTI_PART_REQUEST_MIDDLE,
+				  ssif_info->multi_data + ssif_info->multi_pos,
+				  I2C_SMBUS_BLOCK_DATA);
+		if (rv < 0) {
+			/* request failed, just return the error. */
+			ssif_inc_stat(ssif_info, send_errors);
+
+			if (ssif_info->ssif_debug & SSIF_DEBUG_MSG)
+				pr_info("Error from i2c_non_blocking_op(3)\n");
+			msg_done_handler(ssif_info, -EIO, NULL, 0);
+		}
+	} else {
+		ssif_inc_stat(ssif_info, sent_messages);
+		ssif_inc_stat(ssif_info, sent_messages_parts);
+
+		/* Wait a jiffie then request the next message */
+		ssif_info->retries_left = SSIF_RECV_RETRIES;
+		ssif_info->rtc_us_timer = SSIF_MSG_PART_USEC;
+		mod_timer(&ssif_info->retry_timer,
+			  jiffies + SSIF_MSG_PART_JIFFIES);
+		return;
+	}
+}
+
+static int start_resend(struct ssif_info *ssif_info)
+{
+	int rv;
+	int command;
+
+	if (ssif_info->data_len > 32) {
+		command = SSIF_IPMI_MULTI_PART_REQUEST_START;
+		ssif_info->multi_data = ssif_info->data;
+		ssif_info->multi_len = ssif_info->data_len;
+		/*
+		 * Subtle thing, this is 32, not 33, because we will
+		 * overwrite the thing at position 32 (which was just
+		 * transmitted) with the new length.
+		 */
+		ssif_info->multi_pos = 32;
+		ssif_info->data[0] = 32;
+	} else {
+		ssif_info->multi_data = NULL;
+		command = SSIF_IPMI_REQUEST;
+		ssif_info->data[0] = ssif_info->data_len;
+	}
+
+	rv = ssif_i2c_send(ssif_info, msg_written_handler, I2C_SMBUS_WRITE,
+			  command, ssif_info->data, I2C_SMBUS_BLOCK_DATA);
+	if (rv && (ssif_info->ssif_debug & SSIF_DEBUG_MSG))
+		pr_info("Error from i2c_non_blocking_op(4)\n");
+	return rv;
+}
+
+static int start_send(struct ssif_info *ssif_info,
+		      unsigned char   *data,
+		      unsigned int    len)
+{
+	if (len > IPMI_MAX_MSG_LENGTH)
+		return -E2BIG;
+	if (len > ssif_info->max_xmit_msg_size)
+		return -E2BIG;
+
+	ssif_info->retries_left = SSIF_SEND_RETRIES;
+	memcpy(ssif_info->data+1, data, len);
+	ssif_info->data_len = len;
+	return start_resend(ssif_info);
+}
+
+/* Must be called with the message lock held. */
+static void start_next_msg(struct ssif_info *ssif_info, unsigned long *flags)
+{
+	struct ipmi_smi_msg *msg;
+	unsigned long oflags;
+
+ restart:
+	if (!SSIF_IDLE(ssif_info)) {
+		ipmi_ssif_unlock_cond(ssif_info, flags);
+		return;
+	}
+
+	if (!ssif_info->waiting_msg) {
+		ssif_info->curr_msg = NULL;
+		ipmi_ssif_unlock_cond(ssif_info, flags);
+	} else {
+		int rv;
+
+		ssif_info->curr_msg = ssif_info->waiting_msg;
+		ssif_info->waiting_msg = NULL;
+		ipmi_ssif_unlock_cond(ssif_info, flags);
+		rv = start_send(ssif_info,
+				ssif_info->curr_msg->data,
+				ssif_info->curr_msg->data_size);
+		if (rv) {
+			msg = ssif_info->curr_msg;
+			ssif_info->curr_msg = NULL;
+			return_hosed_msg(ssif_info, msg);
+			flags = ipmi_ssif_lock_cond(ssif_info, &oflags);
+			goto restart;
+		}
+	}
+}
+
+static void sender(void                *send_info,
+		   struct ipmi_smi_msg *msg)
+{
+	struct ssif_info *ssif_info = (struct ssif_info *) send_info;
+	unsigned long oflags, *flags;
+
+	BUG_ON(ssif_info->waiting_msg);
+	ssif_info->waiting_msg = msg;
+
+	flags = ipmi_ssif_lock_cond(ssif_info, &oflags);
+	start_next_msg(ssif_info, flags);
+
+	if (ssif_info->ssif_debug & SSIF_DEBUG_TIMING) {
+		struct timeval t;
+
+		do_gettimeofday(&t);
+		pr_info("**Enqueue %02x %02x: %ld.%6.6ld\n",
+		       msg->data[0], msg->data[1],
+		       (long) t.tv_sec, (long) t.tv_usec);
+	}
+}
+
+static int get_smi_info(void *send_info, struct ipmi_smi_info *data)
+{
+	struct ssif_info *ssif_info = send_info;
+
+	data->addr_src = ssif_info->addr_source;
+	data->dev = &ssif_info->client->dev;
+	data->addr_info = ssif_info->addr_info;
+	get_device(data->dev);
+
+	return 0;
+}
+
+/*
+ * Instead of having our own timer to periodically check the message
+ * flags, we let the message handler drive us.
+ */
+static void request_events(void *send_info)
+{
+	struct ssif_info *ssif_info = (struct ssif_info *) send_info;
+	unsigned long oflags, *flags;
+
+	if (!ssif_info->has_event_buffer)
+		return;
+
+	flags = ipmi_ssif_lock_cond(ssif_info, &oflags);
+	/*
+	 * Request flags first, not events, because the lower layer
+	 * doesn't have a way to send an attention.  But make sure
+	 * event checking still happens.
+	 */
+	ssif_info->req_events = true;
+	if (SSIF_IDLE(ssif_info))
+		start_flag_fetch(ssif_info, flags);
+	else {
+		ssif_info->req_flags = true;
+		ipmi_ssif_unlock_cond(ssif_info, flags);
+	}
+}
+
+static int inc_usecount(void *send_info)
+{
+	struct ssif_info *ssif_info = send_info;
+
+	if (!i2c_get_adapter(ssif_info->client->adapter->nr))
+		return -ENODEV;
+
+	i2c_use_client(ssif_info->client);
+	return 0;
+}
+
+static void dec_usecount(void *send_info)
+{
+	struct ssif_info *ssif_info = send_info;
+
+	i2c_release_client(ssif_info->client);
+	i2c_put_adapter(ssif_info->client->adapter);
+}
+
+static int ssif_start_processing(void *send_info,
+				 ipmi_smi_t intf)
+{
+	struct ssif_info *ssif_info = send_info;
+
+	ssif_info->intf = intf;
+
+	return 0;
+}
+
+#define MAX_SSIF_BMCS 4
+
+static unsigned short addr[MAX_SSIF_BMCS];
+static int num_addrs;
+module_param_array(addr, ushort, &num_addrs, 0);
+MODULE_PARM_DESC(addr, "The addresses to scan for IPMI BMCs on the SSIFs.");
+
+static char *adapter_name[MAX_SSIF_BMCS];
+static int num_adapter_names;
+module_param_array(adapter_name, charp, &num_adapter_names, 0);
+MODULE_PARM_DESC(adapter_name, "The string name of the I2C device that has the BMC.  By default all devices are scanned.");
+
+static int slave_addrs[MAX_SSIF_BMCS];
+static int num_slave_addrs;
+module_param_array(slave_addrs, int, &num_slave_addrs, 0);
+MODULE_PARM_DESC(slave_addrs,
+		 "The default IPMB slave address for the controller.");
+
+/*
+ * Bit 0 enables message debugging, bit 1 enables state debugging, and
+ * bit 2 enables timing debugging.  This is an array indexed by
+ * interface number"
+ */
+static int dbg[MAX_SSIF_BMCS];
+static int num_dbg;
+module_param_array(dbg, int, &num_dbg, 0);
+MODULE_PARM_DESC(dbg, "Turn on debugging.");
+
+static bool ssif_dbg_probe;
+module_param_named(dbg_probe, ssif_dbg_probe, bool, 0);
+MODULE_PARM_DESC(dbg_probe, "Enable debugging of probing of adapters.");
+
+static int use_thread;
+module_param(use_thread, int, 0);
+MODULE_PARM_DESC(use_thread, "Use the thread interface.");
+
+static bool ssif_tryacpi = 1;
+module_param_named(tryacpi, ssif_tryacpi, bool, 0);
+MODULE_PARM_DESC(tryacpi, "Setting this to zero will disable the default scan of the interfaces identified via ACPI");
+
+static bool ssif_trydmi = 1;
+module_param_named(trydmi, ssif_trydmi, bool, 0);
+MODULE_PARM_DESC(trydmi, "Setting this to zero will disable the default scan of the interfaces identified via DMI (SMBIOS)");
+
+static DEFINE_MUTEX(ssif_infos_mutex);
+static LIST_HEAD(ssif_infos);
+
+static int ssif_remove(struct i2c_client *client)
+{
+	struct ssif_info *ssif_info = i2c_get_clientdata(client);
+	int rv;
+
+	if (!ssif_info)
+		return 0;
+
+	i2c_set_clientdata(client, NULL);
+
+	/*
+	 * After this point, we won't deliver anything asychronously
+	 * to the message handler.  We can unregister ourself.
+	 */
+	rv = ipmi_unregister_smi(ssif_info->intf);
+	if (rv) {
+		pr_err(PFX "Unable to unregister device: errno=%d\n", rv);
+		return rv;
+	}
+	ssif_info->intf = NULL;
+
+	/* make sure the driver is not looking for flags any more. */
+	while (ssif_info->ssif_state != SSIF_NORMAL)
+		schedule_timeout(1);
+
+	ssif_info->stopping = true;
+	del_timer_sync(&ssif_info->retry_timer);
+	if (ssif_info->thread) {
+		complete(&ssif_info->wake_thread);
+		kthread_stop(ssif_info->thread);
+	}
+
+	/*
+	 * No message can be outstanding now, we have removed the
+	 * upper layer and it permitted us to do so.
+	 */
+	kfree(ssif_info);
+	return 0;
+}
+
+static int do_cmd(struct i2c_client *client, int len, unsigned char *msg,
+		  int *resp_len, unsigned char *resp)
+{
+	int retry_cnt;
+	int ret;
+
+	retry_cnt = SSIF_SEND_RETRIES;
+ retry1:
+	ret = i2c_smbus_write_block_data(client, SSIF_IPMI_REQUEST, len, msg);
+	if (ret) {
+		retry_cnt--;
+		if (retry_cnt > 0)
+			goto retry1;
+		return -ENODEV;
+	}
+
+	ret = -ENODEV;
+	retry_cnt = SSIF_RECV_RETRIES;
+	while (retry_cnt > 0) {
+		ret = i2c_smbus_read_block_data(client, SSIF_IPMI_RESPONSE,
+						resp);
+		if (ret > 0)
+			break;
+		msleep(SSIF_MSG_MSEC);
+		retry_cnt--;
+		if (retry_cnt <= 0)
+			break;
+	}
+
+	if (ret > 0) {
+		/* Validate that the response is correct. */
+		if (ret < 3 ||
+		    (resp[0] != (msg[0] | (1 << 2))) ||
+		    (resp[1] != msg[1]))
+			ret = -EINVAL;
+		else {
+			*resp_len = ret;
+			ret = 0;
+		}
+	}
+
+	return ret;
+}
+
+static int ssif_detect(struct i2c_client *client, struct i2c_board_info *info)
+{
+	unsigned char *resp;
+	unsigned char msg[3];
+	int           rv;
+	int           len;
+
+	resp = kmalloc(IPMI_MAX_MSG_LENGTH, GFP_KERNEL);
+	if (!resp)
+		return -ENOMEM;
+
+	/* Do a Get Device ID command, since it is required. */
+	msg[0] = IPMI_NETFN_APP_REQUEST << 2;
+	msg[1] = IPMI_GET_DEVICE_ID_CMD;
+	rv = do_cmd(client, 2, msg, &len, resp);
+	if (rv)
+		rv = -ENODEV;
+	else
+		strlcpy(info->type, DEVICE_NAME, I2C_NAME_SIZE);
+	kfree(resp);
+	return rv;
+}
+
+static int smi_type_proc_show(struct seq_file *m, void *v)
+{
+	return seq_puts(m, "ssif\n");
+}
+
+static int smi_type_proc_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, smi_type_proc_show, inode->i_private);
+}
+
+static const struct file_operations smi_type_proc_ops = {
+	.open		= smi_type_proc_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+};
+
+static int smi_stats_proc_show(struct seq_file *m, void *v)
+{
+	struct ssif_info *ssif_info = m->private;
+
+	seq_printf(m, "sent_messages:          %u\n",
+		   ssif_get_stat(ssif_info, sent_messages));
+	seq_printf(m, "sent_messages_parts:    %u\n",
+		   ssif_get_stat(ssif_info, sent_messages_parts));
+	seq_printf(m, "send_retries:           %u\n",
+		   ssif_get_stat(ssif_info, send_retries));
+	seq_printf(m, "send_errors:            %u\n",
+		   ssif_get_stat(ssif_info, send_errors));
+	seq_printf(m, "received_messages:      %u\n",
+		   ssif_get_stat(ssif_info, received_messages));
+	seq_printf(m, "received_message_parts: %u\n",
+		   ssif_get_stat(ssif_info, received_message_parts));
+	seq_printf(m, "receive_retries:        %u\n",
+		   ssif_get_stat(ssif_info, receive_retries));
+	seq_printf(m, "receive_errors:         %u\n",
+		   ssif_get_stat(ssif_info, receive_errors));
+	seq_printf(m, "flag_fetches:           %u\n",
+		   ssif_get_stat(ssif_info, flag_fetches));
+	seq_printf(m, "hosed:                  %u\n",
+		   ssif_get_stat(ssif_info, hosed));
+	seq_printf(m, "events:                 %u\n",
+		   ssif_get_stat(ssif_info, events));
+	seq_printf(m, "watchdog_pretimeouts:   %u\n",
+		   ssif_get_stat(ssif_info, watchdog_pretimeouts));
+	return 0;
+}
+
+static int smi_stats_proc_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, smi_stats_proc_show, PDE_DATA(inode));
+}
+
+static const struct file_operations smi_stats_proc_ops = {
+	.open		= smi_stats_proc_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+};
+
+static struct ssif_addr_info *ssif_info_find(unsigned short addr,
+					     char *adapter_name,
+					     bool match_null_name)
+{
+	struct ssif_addr_info *info, *found = NULL;
+
+restart:
+	list_for_each_entry(info, &ssif_infos, link) {
+		if (info->binfo.addr == addr) {
+			if (info->adapter_name || adapter_name) {
+				if (!info->adapter_name != !adapter_name) {
+					/* One is NULL and one is not */
+					continue;
+				}
+				if (strcmp(info->adapter_name, adapter_name))
+					/* Names to not match */
+					continue;
+			}
+			found = info;
+			break;
+		}
+	}
+
+	if (!found && match_null_name) {
+		/* Try to get an exact match first, then try with a NULL name */
+		adapter_name = NULL;
+		match_null_name = false;
+		goto restart;
+	}
+
+	return found;
+}
+
+static bool check_acpi(struct ssif_info *ssif_info, struct device *dev)
+{
+#ifdef CONFIG_ACPI
+	acpi_handle acpi_handle;
+
+	acpi_handle = ACPI_HANDLE(dev);
+	if (acpi_handle) {
+		ssif_info->addr_source = SI_ACPI;
+		ssif_info->addr_info.acpi_info.acpi_handle = acpi_handle;
+		return true;
+	}
+#endif
+	return false;
+}
+
+static int ssif_probe(struct i2c_client *client, const struct i2c_device_id *id)
+{
+	unsigned char     msg[3];
+	unsigned char     *resp;
+	struct ssif_info   *ssif_info;
+	int               rv = 0;
+	int               len;
+	int               i;
+	u8		  slave_addr = 0;
+	struct ssif_addr_info *addr_info = NULL;
+
+
+	resp = kmalloc(IPMI_MAX_MSG_LENGTH, GFP_KERNEL);
+	if (!resp)
+		return -ENOMEM;
+
+	ssif_info = kzalloc(sizeof(*ssif_info), GFP_KERNEL);
+	if (!ssif_info) {
+		kfree(resp);
+		return -ENOMEM;
+	}
+
+	if (!check_acpi(ssif_info, &client->dev)) {
+		addr_info = ssif_info_find(client->addr, client->adapter->name,
+					   true);
+		if (!addr_info) {
+			/* Must have come in through sysfs. */
+			ssif_info->addr_source = SI_HOTMOD;
+		} else {
+			ssif_info->addr_source = addr_info->addr_src;
+			ssif_info->ssif_debug = addr_info->debug;
+			ssif_info->addr_info = addr_info->addr_info;
+			slave_addr = addr_info->slave_addr;
+		}
+	}
+
+	pr_info(PFX "Trying %s-specified SSIF interface at i2c address 0x%x, adapter %s, slave address 0x%x\n",
+	       ipmi_addr_src_to_str(ssif_info->addr_source),
+	       client->addr, client->adapter->name, slave_addr);
+
+	/*
+	 * Do a Get Device ID command, since it comes back with some
+	 * useful info.
+	 */
+	msg[0] = IPMI_NETFN_APP_REQUEST << 2;
+	msg[1] = IPMI_GET_DEVICE_ID_CMD;
+	rv = do_cmd(client, 2, msg, &len, resp);
+	if (rv)
+		goto out;
+
+	rv = ipmi_demangle_device_id(resp, len, &ssif_info->device_id);
+	if (rv)
+		goto out;
+
+	ssif_info->client = client;
+	i2c_set_clientdata(client, ssif_info);
+
+	/* Now check for system interface capabilities */
+	msg[0] = IPMI_NETFN_APP_REQUEST << 2;
+	msg[1] = IPMI_GET_SYSTEM_INTERFACE_CAPABILITIES_CMD;
+	msg[2] = 0; /* SSIF */
+	rv = do_cmd(client, 3, msg, &len, resp);
+	if (!rv && (len >= 3) && (resp[2] == 0)) {
+		if (len < 7) {
+			if (ssif_dbg_probe)
+				pr_info(PFX "SSIF info too short: %d\n", len);
+			goto no_support;
+		}
+
+		/* Got a good SSIF response, handle it. */
+		ssif_info->max_xmit_msg_size = resp[5];
+		ssif_info->max_recv_msg_size = resp[6];
+		ssif_info->multi_support = (resp[4] >> 6) & 0x3;
+		ssif_info->supports_pec = (resp[4] >> 3) & 0x1;
+
+		/* Sanitize the data */
+		switch (ssif_info->multi_support) {
+		case SSIF_NO_MULTI:
+			if (ssif_info->max_xmit_msg_size > 32)
+				ssif_info->max_xmit_msg_size = 32;
+			if (ssif_info->max_recv_msg_size > 32)
+				ssif_info->max_recv_msg_size = 32;
+			break;
+
+		case SSIF_MULTI_2_PART:
+			if (ssif_info->max_xmit_msg_size > 64)
+				ssif_info->max_xmit_msg_size = 64;
+			if (ssif_info->max_recv_msg_size > 62)
+				ssif_info->max_recv_msg_size = 62;
+			break;
+
+		case SSIF_MULTI_n_PART:
+			break;
+
+		default:
+			/* Data is not sane, just give up. */
+			goto no_support;
+		}
+	} else {
+ no_support:
+		/* Assume no multi-part or PEC support */
+		pr_info(PFX "Error fetching SSIF: %d %d %2.2x, your system probably doesn't support this command so  using defaults\n",
+		       rv, len, resp[2]);
+
+		ssif_info->max_xmit_msg_size = 32;
+		ssif_info->max_recv_msg_size = 32;
+		ssif_info->multi_support = SSIF_NO_MULTI;
+		ssif_info->supports_pec = 0;
+	}
+
+	/* Make sure the NMI timeout is cleared. */
+	msg[0] = IPMI_NETFN_APP_REQUEST << 2;
+	msg[1] = IPMI_CLEAR_MSG_FLAGS_CMD;
+	msg[2] = WDT_PRE_TIMEOUT_INT;
+	rv = do_cmd(client, 3, msg, &len, resp);
+	if (rv || (len < 3) || (resp[2] != 0))
+		pr_warn(PFX "Unable to clear message flags: %d %d %2.2x\n",
+			rv, len, resp[2]);
+
+	/* Attempt to enable the event buffer. */
+	msg[0] = IPMI_NETFN_APP_REQUEST << 2;
+	msg[1] = IPMI_GET_BMC_GLOBAL_ENABLES_CMD;
+	rv = do_cmd(client, 2, msg, &len, resp);
+	if (rv || (len < 4) || (resp[2] != 0)) {
+		pr_warn(PFX "Error getting global enables: %d %d %2.2x\n",
+			rv, len, resp[2]);
+		rv = 0; /* Not fatal */
+		goto found;
+	}
+
+	if (resp[3] & IPMI_BMC_EVT_MSG_BUFF) {
+		ssif_info->has_event_buffer = true;
+		/* buffer is already enabled, nothing to do. */
+		goto found;
+	}
+
+	msg[0] = IPMI_NETFN_APP_REQUEST << 2;
+	msg[1] = IPMI_SET_BMC_GLOBAL_ENABLES_CMD;
+	msg[2] = resp[3] | IPMI_BMC_EVT_MSG_BUFF;
+	rv = do_cmd(client, 3, msg, &len, resp);
+	if (rv || (len < 2)) {
+		pr_warn(PFX "Error getting global enables: %d %d %2.2x\n",
+			rv, len, resp[2]);
+		rv = 0; /* Not fatal */
+		goto found;
+	}
+
+	if (resp[2] == 0)
+		/* A successful return means the event buffer is supported. */
+		ssif_info->has_event_buffer = true;
+
+ found:
+	ssif_info->intf_num = atomic_inc_return(&next_intf);
+
+	if (ssif_dbg_probe) {
+		pr_info("ssif_probe: i2c_probe found device at i2c address %x\n",
+			client->addr);
+	}
+
+	spin_lock_init(&ssif_info->lock);
+	ssif_info->ssif_state = SSIF_NORMAL;
+	init_timer(&ssif_info->retry_timer);
+	ssif_info->retry_timer.data = (unsigned long) ssif_info;
+	ssif_info->retry_timer.function = retry_timeout;
+
+	for (i = 0; i < SSIF_NUM_STATS; i++)
+		atomic_set(&ssif_info->stats[i], 0);
+
+	if (ssif_info->supports_pec)
+		ssif_info->client->flags |= I2C_CLIENT_PEC;
+
+	ssif_info->handlers.owner = THIS_MODULE;
+	ssif_info->handlers.start_processing = ssif_start_processing;
+	ssif_info->handlers.get_smi_info = get_smi_info;
+	ssif_info->handlers.sender = sender;
+	ssif_info->handlers.request_events = request_events;
+	ssif_info->handlers.inc_usecount = inc_usecount;
+	ssif_info->handlers.dec_usecount = dec_usecount;
+
+	{
+		unsigned int thread_num;
+
+		thread_num = ((ssif_info->client->adapter->nr << 8) |
+			      ssif_info->client->addr);
+		init_completion(&ssif_info->wake_thread);
+		ssif_info->thread = kthread_run(ipmi_ssif_thread, ssif_info,
+					       "kssif%4.4x", thread_num);
+		if (IS_ERR(ssif_info->thread)) {
+			rv = PTR_ERR(ssif_info->thread);
+			dev_notice(&ssif_info->client->dev,
+				   "Could not start kernel thread: error %d\n",
+				   rv);
+			goto out;
+		}
+	}
+
+	rv = ipmi_register_smi(&ssif_info->handlers,
+			       ssif_info,
+			       &ssif_info->device_id,
+			       &ssif_info->client->dev,
+			       slave_addr);
+	 if (rv) {
+		pr_err(PFX "Unable to register device: error %d\n", rv);
+		goto out;
+	}
+
+	rv = ipmi_smi_add_proc_entry(ssif_info->intf, "type",
+				     &smi_type_proc_ops,
+				     ssif_info);
+	if (rv) {
+		pr_err(PFX "Unable to create proc entry: %d\n", rv);
+		goto out_err_unreg;
+	}
+
+	rv = ipmi_smi_add_proc_entry(ssif_info->intf, "ssif_stats",
+				     &smi_stats_proc_ops,
+				     ssif_info);
+	if (rv) {
+		pr_err(PFX "Unable to create proc entry: %d\n", rv);
+		goto out_err_unreg;
+	}
+
+ out:
+	if (rv)
+		kfree(ssif_info);
+	kfree(resp);
+	return rv;
+
+ out_err_unreg:
+	ipmi_unregister_smi(ssif_info->intf);
+	goto out;
+}
+
+static int ssif_adapter_handler(struct device *adev, void *opaque)
+{
+	struct ssif_addr_info *addr_info = opaque;
+
+	if (adev->type != &i2c_adapter_type)
+		return 0;
+
+	i2c_new_device(to_i2c_adapter(adev), &addr_info->binfo);
+
+	if (!addr_info->adapter_name)
+		return 1; /* Only try the first I2C adapter by default. */
+	return 0;
+}
+
+static int new_ssif_client(int addr, char *adapter_name,
+			   int debug, int slave_addr,
+			   enum ipmi_addr_src addr_src)
+{
+	struct ssif_addr_info *addr_info;
+	int rv = 0;
+
+	mutex_lock(&ssif_infos_mutex);
+	if (ssif_info_find(addr, adapter_name, false)) {
+		rv = -EEXIST;
+		goto out_unlock;
+	}
+
+	addr_info = kzalloc(sizeof(*addr_info), GFP_KERNEL);
+	if (!addr_info) {
+		rv = -ENOMEM;
+		goto out_unlock;
+	}
+
+	if (adapter_name) {
+		addr_info->adapter_name = kstrdup(adapter_name, GFP_KERNEL);
+		if (!addr_info->adapter_name) {
+			kfree(addr_info);
+			rv = -ENOMEM;
+			goto out_unlock;
+		}
+	}
+
+	strncpy(addr_info->binfo.type, DEVICE_NAME,
+		sizeof(addr_info->binfo.type));
+	addr_info->binfo.addr = addr;
+	addr_info->binfo.platform_data = addr_info;
+	addr_info->debug = debug;
+	addr_info->slave_addr = slave_addr;
+	addr_info->addr_src = addr_src;
+
+	list_add_tail(&addr_info->link, &ssif_infos);
+
+	if (initialized)
+		i2c_for_each_dev(addr_info, ssif_adapter_handler);
+	/* Otherwise address list will get it */
+
+out_unlock:
+	mutex_unlock(&ssif_infos_mutex);
+	return rv;
+}
+
+static void free_ssif_clients(void)
+{
+	struct ssif_addr_info *info, *tmp;
+
+	mutex_lock(&ssif_infos_mutex);
+	list_for_each_entry_safe(info, tmp, &ssif_infos, link) {
+		list_del(&info->link);
+		kfree(info->adapter_name);
+		kfree(info);
+	}
+	mutex_unlock(&ssif_infos_mutex);
+}
+
+static unsigned short *ssif_address_list(void)
+{
+	struct ssif_addr_info *info;
+	unsigned int count = 0, i;
+	unsigned short *address_list;
+
+	list_for_each_entry(info, &ssif_infos, link)
+		count++;
+
+	address_list = kzalloc(sizeof(*address_list) * (count + 1), GFP_KERNEL);
+	if (!address_list)
+		return NULL;
+
+	i = 0;
+	list_for_each_entry(info, &ssif_infos, link) {
+		unsigned short addr = info->binfo.addr;
+		int j;
+
+		for (j = 0; j < i; j++) {
+			if (address_list[j] == addr)
+				goto skip_addr;
+		}
+		address_list[i] = addr;
+skip_addr:
+		i++;
+	}
+	address_list[i] = I2C_CLIENT_END;
+
+	return address_list;
+}
+
+#ifdef CONFIG_ACPI
+static struct acpi_device_id ssif_acpi_match[] = {
+	{ "IPI0001", 0 },
+	{ },
+};
+MODULE_DEVICE_TABLE(acpi, ssif_acpi_match);
+
+/*
+ * Once we get an ACPI failure, we don't try any more, because we go
+ * through the tables sequentially.  Once we don't find a table, there
+ * are no more.
+ */
+static int acpi_failure;
+
+/*
+ * Defined in the IPMI 2.0 spec.
+ */
+struct SPMITable {
+	s8	Signature[4];
+	u32	Length;
+	u8	Revision;
+	u8	Checksum;
+	s8	OEMID[6];
+	s8	OEMTableID[8];
+	s8	OEMRevision[4];
+	s8	CreatorID[4];
+	s8	CreatorRevision[4];
+	u8	InterfaceType;
+	u8	IPMIlegacy;
+	s16	SpecificationRevision;
+
+	/*
+	 * Bit 0 - SCI interrupt supported
+	 * Bit 1 - I/O APIC/SAPIC
+	 */
+	u8	InterruptType;
+
+	/*
+	 * If bit 0 of InterruptType is set, then this is the SCI
+	 * interrupt in the GPEx_STS register.
+	 */
+	u8	GPE;
+
+	s16	Reserved;
+
+	/*
+	 * If bit 1 of InterruptType is set, then this is the I/O
+	 * APIC/SAPIC interrupt.
+	 */
+	u32	GlobalSystemInterrupt;
+
+	/* The actual register address. */
+	struct acpi_generic_address addr;
+
+	u8	UID[4];
+
+	s8      spmi_id[1]; /* A '\0' terminated array starts here. */
+};
+
+static int try_init_spmi(struct SPMITable *spmi)
+{
+	unsigned short myaddr;
+
+	if (num_addrs >= MAX_SSIF_BMCS)
+		return -1;
+
+	if (spmi->IPMIlegacy != 1) {
+		pr_warn("IPMI: Bad SPMI legacy: %d\n", spmi->IPMIlegacy);
+		return -ENODEV;
+	}
+
+	if (spmi->InterfaceType != 4)
+		return -ENODEV;
+
+	if (spmi->addr.space_id != ACPI_ADR_SPACE_SMBUS) {
+		pr_warn(PFX "Invalid ACPI SSIF I/O Address type: %d\n",
+			spmi->addr.space_id);
+		return -EIO;
+	}
+
+	myaddr = spmi->addr.address >> 1;
+
+	return new_ssif_client(myaddr, NULL, 0, 0, SI_SPMI);
+}
+
+static void spmi_find_bmc(void)
+{
+	acpi_status      status;
+	struct SPMITable *spmi;
+	int              i;
+
+	if (acpi_disabled)
+		return;
+
+	if (acpi_failure)
+		return;
+
+	for (i = 0; ; i++) {
+		status = acpi_get_table(ACPI_SIG_SPMI, i+1,
+					(struct acpi_table_header **)&spmi);
+		if (status != AE_OK)
+			return;
+
+		try_init_spmi(spmi);
+	}
+}
+#else
+static void spmi_find_bmc(void) { }
+#endif
+
+#ifdef CONFIG_DMI
+static int decode_dmi(const struct dmi_device *dmi_dev)
+{
+	struct dmi_header *dm = dmi_dev->device_data;
+	u8             *data = (u8 *) dm;
+	u8             len = dm->length;
+	unsigned short myaddr;
+	int            slave_addr;
+
+	if (num_addrs >= MAX_SSIF_BMCS)
+		return -1;
+
+	if (len < 9)
+		return -1;
+
+	if (data[0x04] != 4) /* Not SSIF */
+		return -1;
+
+	if ((data[8] >> 1) == 0) {
+		/*
+		 * Some broken systems put the I2C address in
+		 * the slave address field.  We try to
+		 * accommodate them here.
+		 */
+		myaddr = data[6] >> 1;
+		slave_addr = 0;
+	} else {
+		myaddr = data[8] >> 1;
+		slave_addr = data[6];
+	}
+
+	return new_ssif_client(myaddr, NULL, 0, 0, SI_SMBIOS);
+}
+
+static void dmi_iterator(void)
+{
+	const struct dmi_device *dev = NULL;
+
+	while ((dev = dmi_find_device(DMI_DEV_TYPE_IPMI, NULL, dev)))
+		decode_dmi(dev);
+}
+#else
+static void dmi_iterator(void) { }
+#endif
+
+static const struct i2c_device_id ssif_id[] = {
+	{ DEVICE_NAME, 0 },
+	{ }
+};
+MODULE_DEVICE_TABLE(i2c, ssif_id);
+
+static struct i2c_driver ssif_i2c_driver = {
+	.class		= I2C_CLASS_HWMON,
+	.driver		= {
+		.owner			= THIS_MODULE,
+		.name			= DEVICE_NAME
+	},
+	.probe		= ssif_probe,
+	.remove		= ssif_remove,
+	.id_table	= ssif_id,
+	.detect		= ssif_detect
+};
+
+static int init_ipmi_ssif(void)
+{
+	int i;
+	int rv;
+
+	if (initialized)
+		return 0;
+
+	pr_info("IPMI SSIF Interface driver\n");
+
+	/* build list for i2c from addr list */
+	for (i = 0; i < num_addrs; i++) {
+		rv = new_ssif_client(addr[i], adapter_name[i],
+				     dbg[i], slave_addrs[i],
+				     SI_HARDCODED);
+		if (!rv)
+			pr_err(PFX
+			       "Couldn't add hardcoded device at addr 0x%x\n",
+			       addr[i]);
+	}
+
+	if (ssif_tryacpi)
+		ssif_i2c_driver.driver.acpi_match_table	=
+			ACPI_PTR(ssif_acpi_match);
+	if (ssif_trydmi)
+		dmi_iterator();
+	if (ssif_tryacpi)
+		spmi_find_bmc();
+
+	ssif_i2c_driver.address_list = ssif_address_list();
+
+	rv = i2c_add_driver(&ssif_i2c_driver);
+	if (!rv)
+		initialized = true;
+
+	return rv;
+}
+module_init(init_ipmi_ssif);
+
+static void cleanup_ipmi_ssif(void)
+{
+	if (!initialized)
+		return;
+
+	initialized = false;
+
+	i2c_del_driver(&ssif_i2c_driver);
+
+	free_ssif_clients();
+}
+module_exit(cleanup_ipmi_ssif);
+
+MODULE_AUTHOR("Todd C Davis <todd.c.davis@intel.com>, Corey Minyard <minyard@acm.org>");
+MODULE_DESCRIPTION("IPMI driver for management controllers on a SMBus");
+MODULE_LICENSE("GPL");
diff --git a/drivers/char/mem.c b/drivers/char/mem.c
index 917403fe10da..4c58333b4257 100644
--- a/drivers/char/mem.c
+++ b/drivers/char/mem.c
@@ -84,9 +84,12 @@ static inline int range_is_allowed(unsigned long pfn, unsigned long size)
 }
 #endif
 
-void __weak unxlate_dev_mem_ptr(unsigned long phys, void *addr)
+#ifndef unxlate_dev_mem_ptr
+#define unxlate_dev_mem_ptr unxlate_dev_mem_ptr
+void __weak unxlate_dev_mem_ptr(phys_addr_t phys, void *addr)
 {
 }
+#endif
 
 /*
  * This funcion reads the *physical* memory. The f_pos points directly to the
@@ -97,7 +100,7 @@ static ssize_t read_mem(struct file *file, char __user *buf,
 {
 	phys_addr_t p = *ppos;
 	ssize_t read, sz;
-	char *ptr;
+	void *ptr;
 
 	if (p != *ppos)
 		return 0;
@@ -400,7 +403,7 @@ static ssize_t read_kmem(struct file *file, char __user *buf,
 			 * uncached, then it must also be accessed uncached
 			 * by the kernel or data corruption may occur
 			 */
-			kbuf = xlate_dev_kmem_ptr((char *)p);
+			kbuf = xlate_dev_kmem_ptr((void *)p);
 
 			if (copy_to_user(buf, kbuf, sz))
 				return -EFAULT;
@@ -461,7 +464,7 @@ static ssize_t do_write_kmem(unsigned long p, const char __user *buf,
 #endif
 
 	while (count > 0) {
-		char *ptr;
+		void *ptr;
 
 		sz = size_inside_page(p, count);
 
@@ -470,7 +473,7 @@ static ssize_t do_write_kmem(unsigned long p, const char __user *buf,
 		 * it must also be accessed uncached by the kernel or data
 		 * corruption may occur.
 		 */
-		ptr = xlate_dev_kmem_ptr((char *)p);
+		ptr = xlate_dev_kmem_ptr((void *)p);
 
 		copied = copy_from_user(ptr, buf, sz);
 		if (copied) {
@@ -622,53 +625,23 @@ static ssize_t splice_write_null(struct pipe_inode_info *pipe, struct file *out,
 	return splice_from_pipe(pipe, out, ppos, len, flags, pipe_to_null);
 }
 
-static ssize_t read_zero(struct file *file, char __user *buf,
-			 size_t count, loff_t *ppos)
+static ssize_t read_iter_zero(struct kiocb *iocb, struct iov_iter *iter)
 {
-	size_t written;
-
-	if (!count)
-		return 0;
-
-	if (!access_ok(VERIFY_WRITE, buf, count))
-		return -EFAULT;
-
-	written = 0;
-	while (count) {
-		unsigned long unwritten;
-		size_t chunk = count;
+	size_t written = 0;
 
+	while (iov_iter_count(iter)) {
+		size_t chunk = iov_iter_count(iter), n;
 		if (chunk > PAGE_SIZE)
 			chunk = PAGE_SIZE;	/* Just for latency reasons */
-		unwritten = __clear_user(buf, chunk);
-		written += chunk - unwritten;
-		if (unwritten)
-			break;
+		n = iov_iter_zero(chunk, iter);
+		if (!n && iov_iter_count(iter))
+			return written ? written : -EFAULT;
+		written += n;
 		if (signal_pending(current))
 			return written ? written : -ERESTARTSYS;
-		buf += chunk;
-		count -= chunk;
 		cond_resched();
 	}
-	return written ? written : -EFAULT;
-}
-
-static ssize_t aio_read_zero(struct kiocb *iocb, const struct iovec *iov,
-			     unsigned long nr_segs, loff_t pos)
-{
-	size_t written = 0;
-	unsigned long i;
-	ssize_t ret;
-
-	for (i = 0; i < nr_segs; i++) {
-		ret = read_zero(iocb->ki_filp, iov[i].iov_base, iov[i].iov_len,
-				&pos);
-		if (ret < 0)
-			break;
-		written += ret;
-	}
-
-	return written ? written : -EFAULT;
+	return written;
 }
 
 static int mmap_zero(struct file *file, struct vm_area_struct *vma)
@@ -738,7 +711,6 @@ static int open_port(struct inode *inode, struct file *filp)
 #define zero_lseek	null_lseek
 #define full_lseek      null_lseek
 #define write_zero	write_null
-#define read_full       read_zero
 #define aio_write_zero	aio_write_null
 #define open_mem	open_port
 #define open_kmem	open_mem
@@ -783,9 +755,9 @@ static const struct file_operations port_fops = {
 
 static const struct file_operations zero_fops = {
 	.llseek		= zero_lseek,
-	.read		= read_zero,
+	.read		= new_sync_read,
 	.write		= write_zero,
-	.aio_read	= aio_read_zero,
+	.read_iter	= read_iter_zero,
 	.aio_write	= aio_write_zero,
 	.mmap		= mmap_zero,
 };
@@ -802,7 +774,8 @@ static struct backing_dev_info zero_bdi = {
 
 static const struct file_operations full_fops = {
 	.llseek		= full_lseek,
-	.read		= read_full,
+	.read		= new_sync_read,
+	.read_iter	= read_iter_zero,
 	.write		= write_full,
 };
 
diff --git a/drivers/char/random.c b/drivers/char/random.c
index c18d41db83d8..04645c09fe5e 100644
--- a/drivers/char/random.c
+++ b/drivers/char/random.c
@@ -874,7 +874,7 @@ static __u32 get_reg(struct fast_pool *f, struct pt_regs *regs)
 void add_interrupt_randomness(int irq, int irq_flags)
 {
 	struct entropy_store	*r;
-	struct fast_pool	*fast_pool = &__get_cpu_var(irq_randomness);
+	struct fast_pool	*fast_pool = this_cpu_ptr(&irq_randomness);
 	struct pt_regs		*regs = get_irq_regs();
 	unsigned long		now = jiffies;
 	cycles_t		cycles = random_get_entropy();
@@ -1106,7 +1106,7 @@ static void extract_buf(struct entropy_store *r, __u8 *out)
 	__mix_pool_bytes(r, hash.w, sizeof(hash.w));
 	spin_unlock_irqrestore(&r->lock, flags);
 
-	memset(workspace, 0, sizeof(workspace));
+	memzero_explicit(workspace, sizeof(workspace));
 
 	/*
 	 * In case the hash function has some recognizable output
@@ -1118,7 +1118,7 @@ static void extract_buf(struct entropy_store *r, __u8 *out)
 	hash.w[2] ^= rol32(hash.w[2], 16);
 
 	memcpy(out, &hash, EXTRACT_SIZE);
-	memset(&hash, 0, sizeof(hash));
+	memzero_explicit(&hash, sizeof(hash));
 }
 
 /*
@@ -1175,7 +1175,7 @@ static ssize_t extract_entropy(struct entropy_store *r, void *buf,
 	}
 
 	/* Wipe data just returned from memory */
-	memset(tmp, 0, sizeof(tmp));
+	memzero_explicit(tmp, sizeof(tmp));
 
 	return ret;
 }
@@ -1218,7 +1218,7 @@ static ssize_t extract_entropy_user(struct entropy_store *r, void __user *buf,
 	}
 
 	/* Wipe data just returned from memory */
-	memset(tmp, 0, sizeof(tmp));
+	memzero_explicit(tmp, sizeof(tmp));
 
 	return ret;
 }
diff --git a/drivers/char/raw.c b/drivers/char/raw.c
index 0102dc788608..a24891b97547 100644
--- a/drivers/char/raw.c
+++ b/drivers/char/raw.c
@@ -285,7 +285,7 @@ static long raw_ctl_compat_ioctl(struct file *file, unsigned int cmd,
 
 static const struct file_operations raw_fops = {
 	.read		= new_sync_read,
-	.read_iter	= generic_file_read_iter,
+	.read_iter	= blkdev_read_iter,
 	.write		= new_sync_write,
 	.write_iter	= blkdev_write_iter,
 	.fsync		= blkdev_fsync,
diff --git a/drivers/char/sonypi.c b/drivers/char/sonypi.c
index 7cc1fe2241fd..e496daefe9e0 100644
--- a/drivers/char/sonypi.c
+++ b/drivers/char/sonypi.c
@@ -1482,7 +1482,6 @@ static void sonypi_shutdown(struct platform_device *dev)
 static struct platform_driver sonypi_driver = {
 	.driver		= {
 		.name	= "sonypi",
-		.owner	= THIS_MODULE,
 		.pm	= SONYPI_PM,
 	},
 	.probe		= sonypi_probe,
diff --git a/drivers/char/tb0219.c b/drivers/char/tb0219.c
index 47b9fdfcf083..480a777db577 100644
--- a/drivers/char/tb0219.c
+++ b/drivers/char/tb0219.c
@@ -337,7 +337,6 @@ static struct platform_driver tb0219_device_driver = {
 	.remove		= tb0219_remove,
 	.driver		= {
 		.name	= "TB0219",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/char/tile-srom.c b/drivers/char/tile-srom.c
index bd377472dcfb..02e76ac6d282 100644
--- a/drivers/char/tile-srom.c
+++ b/drivers/char/tile-srom.c
@@ -76,6 +76,7 @@ MODULE_LICENSE("GPL");
 
 static int srom_devs;			/* Number of SROM partitions */
 static struct cdev srom_cdev;
+static struct platform_device *srom_parent;
 static struct class *srom_class;
 static struct srom_dev *srom_devices;
 
@@ -350,7 +351,7 @@ static int srom_setup_minor(struct srom_dev *srom, int index)
 		       SROM_PAGE_SIZE_OFF, sizeof(srom->page_size)) < 0)
 		return -EIO;
 
-	dev = device_create(srom_class, &platform_bus,
+	dev = device_create(srom_class, &srom_parent->dev,
 			    MKDEV(srom_major, index), srom, "%d", index);
 	return PTR_ERR_OR_ZERO(dev);
 }
@@ -415,6 +416,13 @@ static int srom_init(void)
 	if (result < 0)
 		goto fail_chrdev;
 
+	/* Create a parent device */
+	srom_parent = platform_device_register_simple("srom", -1, NULL, 0);
+	if (IS_ERR(srom_parent)) {
+		result = PTR_ERR(srom_parent);
+		goto fail_pdev;
+	}
+
 	/* Create a sysfs class. */
 	srom_class = class_create(THIS_MODULE, "srom");
 	if (IS_ERR(srom_class)) {
@@ -438,6 +446,8 @@ fail_class:
 		device_destroy(srom_class, MKDEV(srom_major, i));
 	class_destroy(srom_class);
 fail_cdev:
+	platform_device_unregister(srom_parent);
+fail_pdev:
 	cdev_del(&srom_cdev);
 fail_chrdev:
 	unregister_chrdev_region(dev, srom_devs);
@@ -454,6 +464,7 @@ static void srom_cleanup(void)
 		device_destroy(srom_class, MKDEV(srom_major, i));
 	class_destroy(srom_class);
 	cdev_del(&srom_cdev);
+	platform_device_unregister(srom_parent);
 	unregister_chrdev_region(MKDEV(srom_major, 0), srom_devs);
 	kfree(srom_devices);
 }
diff --git a/drivers/char/tpm/tpm_atmel.c b/drivers/char/tpm/tpm_atmel.c
index 6069d13ae4ac..435c8b9dd2f8 100644
--- a/drivers/char/tpm/tpm_atmel.c
+++ b/drivers/char/tpm/tpm_atmel.c
@@ -152,7 +152,6 @@ static SIMPLE_DEV_PM_OPS(tpm_atml_pm, tpm_pm_suspend, tpm_pm_resume);
 static struct platform_driver atml_drv = {
 	.driver = {
 		.name = "tpm_atmel",
-		.owner		= THIS_MODULE,
 		.pm		= &tpm_atml_pm,
 	},
 };
diff --git a/drivers/char/tpm/tpm_nsc.c b/drivers/char/tpm/tpm_nsc.c
index 3179ec9cffdc..4d0a17ea8cde 100644
--- a/drivers/char/tpm/tpm_nsc.c
+++ b/drivers/char/tpm/tpm_nsc.c
@@ -258,7 +258,6 @@ static SIMPLE_DEV_PM_OPS(tpm_nsc_pm, tpm_pm_suspend, tpm_pm_resume);
 static struct platform_driver nsc_drv = {
 	.driver          = {
 		.name    = "tpm_nsc",
-		.owner   = THIS_MODULE,
 		.pm      = &tpm_nsc_pm,
 	},
 };
diff --git a/drivers/char/tpm/tpm_tis.c b/drivers/char/tpm/tpm_tis.c
index 2c46734b266d..6f1985496112 100644
--- a/drivers/char/tpm/tpm_tis.c
+++ b/drivers/char/tpm/tpm_tis.c
@@ -837,7 +837,6 @@ MODULE_PARM_DESC(hid, "Set additional specific HID for this driver to probe");
 static struct platform_driver tis_drv = {
 	.driver = {
 		.name = "tpm_tis",
-		.owner		= THIS_MODULE,
 		.pm		= &tpm_tis_pm,
 	},
 };
diff --git a/drivers/char/tpm/xen-tpmfront.c b/drivers/char/tpm/xen-tpmfront.c
index 2064b4527040..441b44e54226 100644
--- a/drivers/char/tpm/xen-tpmfront.c
+++ b/drivers/char/tpm/xen-tpmfront.c
@@ -367,12 +367,13 @@ static const struct xenbus_device_id tpmfront_ids[] = {
 };
 MODULE_ALIAS("xen:vtpm");
 
-static DEFINE_XENBUS_DRIVER(tpmfront, ,
-		.probe = tpmfront_probe,
-		.remove = tpmfront_remove,
-		.resume = tpmfront_resume,
-		.otherend_changed = backend_changed,
-	);
+static struct xenbus_driver tpmfront_driver = {
+	.ids = tpmfront_ids,
+	.probe = tpmfront_probe,
+	.remove = tpmfront_remove,
+	.resume = tpmfront_resume,
+	.otherend_changed = backend_changed,
+};
 
 static int __init xen_tpmfront_init(void)
 {
diff --git a/drivers/char/virtio_console.c b/drivers/char/virtio_console.c
index b585b4789822..de03df9dd7c9 100644
--- a/drivers/char/virtio_console.c
+++ b/drivers/char/virtio_console.c
@@ -355,7 +355,7 @@ static inline bool use_multiport(struct ports_device *portdev)
 	 */
 	if (!portdev->vdev)
 		return 0;
-	return portdev->vdev->features[0] & (1 << VIRTIO_CONSOLE_F_MULTIPORT);
+	return __virtio_test_bit(portdev->vdev, VIRTIO_CONSOLE_F_MULTIPORT);
 }
 
 static DEFINE_SPINLOCK(dma_bufs_lock);
@@ -566,9 +566,9 @@ static ssize_t __send_control_msg(struct ports_device *portdev, u32 port_id,
 	if (!use_multiport(portdev))
 		return 0;
 
-	cpkt.id = port_id;
-	cpkt.event = event;
-	cpkt.value = value;
+	cpkt.id = cpu_to_virtio32(portdev->vdev, port_id);
+	cpkt.event = cpu_to_virtio16(portdev->vdev, event);
+	cpkt.value = cpu_to_virtio16(portdev->vdev, value);
 
 	vq = portdev->c_ovq;
 
@@ -669,8 +669,8 @@ done:
  * Give out the data that's requested from the buffer that we have
  * queued up.
  */
-static ssize_t fill_readbuf(struct port *port, char *out_buf, size_t out_count,
-			    bool to_user)
+static ssize_t fill_readbuf(struct port *port, char __user *out_buf,
+			    size_t out_count, bool to_user)
 {
 	struct port_buffer *buf;
 	unsigned long flags;
@@ -688,7 +688,8 @@ static ssize_t fill_readbuf(struct port *port, char *out_buf, size_t out_count,
 		if (ret)
 			return -EFAULT;
 	} else {
-		memcpy(out_buf, buf->buf + buf->offset, out_count);
+		memcpy((__force char *)out_buf, buf->buf + buf->offset,
+		       out_count);
 	}
 
 	buf->offset += out_count;
@@ -1162,7 +1163,7 @@ static int get_chars(u32 vtermno, char *buf, int count)
 	/* If we don't have an input queue yet, we can't get input. */
 	BUG_ON(!port->in_vq);
 
-	return fill_readbuf(port, buf, count, false);
+	return fill_readbuf(port, (__force char __user *)buf, count, false);
 }
 
 static void resize_console(struct port *port)
@@ -1602,7 +1603,8 @@ static void unplug_port(struct port *port)
 }
 
 /* Any private messages that the Host and Guest want to share */
-static void handle_control_message(struct ports_device *portdev,
+static void handle_control_message(struct virtio_device *vdev,
+				   struct ports_device *portdev,
 				   struct port_buffer *buf)
 {
 	struct virtio_console_control *cpkt;
@@ -1612,15 +1614,16 @@ static void handle_control_message(struct ports_device *portdev,
 
 	cpkt = (struct virtio_console_control *)(buf->buf + buf->offset);
 
-	port = find_port_by_id(portdev, cpkt->id);
-	if (!port && cpkt->event != VIRTIO_CONSOLE_PORT_ADD) {
+	port = find_port_by_id(portdev, virtio32_to_cpu(vdev, cpkt->id));
+	if (!port &&
+	    cpkt->event != cpu_to_virtio16(vdev, VIRTIO_CONSOLE_PORT_ADD)) {
 		/* No valid header at start of buffer.  Drop it. */
 		dev_dbg(&portdev->vdev->dev,
 			"Invalid index %u in control packet\n", cpkt->id);
 		return;
 	}
 
-	switch (cpkt->event) {
+	switch (virtio16_to_cpu(vdev, cpkt->event)) {
 	case VIRTIO_CONSOLE_PORT_ADD:
 		if (port) {
 			dev_dbg(&portdev->vdev->dev,
@@ -1628,13 +1631,15 @@ static void handle_control_message(struct ports_device *portdev,
 			send_control_msg(port, VIRTIO_CONSOLE_PORT_READY, 1);
 			break;
 		}
-		if (cpkt->id >= portdev->config.max_nr_ports) {
+		if (virtio32_to_cpu(vdev, cpkt->id) >=
+		    portdev->config.max_nr_ports) {
 			dev_warn(&portdev->vdev->dev,
-				"Request for adding port with out-of-bound id %u, max. supported id: %u\n",
+				"Request for adding port with "
+				"out-of-bound id %u, max. supported id: %u\n",
 				cpkt->id, portdev->config.max_nr_ports - 1);
 			break;
 		}
-		add_port(portdev, cpkt->id);
+		add_port(portdev, virtio32_to_cpu(vdev, cpkt->id));
 		break;
 	case VIRTIO_CONSOLE_PORT_REMOVE:
 		unplug_port(port);
@@ -1670,7 +1675,7 @@ static void handle_control_message(struct ports_device *portdev,
 		break;
 	}
 	case VIRTIO_CONSOLE_PORT_OPEN:
-		port->host_connected = cpkt->value;
+		port->host_connected = virtio16_to_cpu(vdev, cpkt->value);
 		wake_up_interruptible(&port->waitqueue);
 		/*
 		 * If the host port got closed and the host had any
@@ -1752,7 +1757,7 @@ static void control_work_handler(struct work_struct *work)
 		buf->len = len;
 		buf->offset = 0;
 
-		handle_control_message(portdev, buf);
+		handle_control_message(vq->vdev, portdev, buf);
 
 		spin_lock(&portdev->c_ivq_lock);
 		if (add_inbuf(portdev->c_ivq, buf) < 0) {
@@ -2024,6 +2029,8 @@ static int virtcons_probe(struct virtio_device *vdev)
 	spin_lock_init(&portdev->ports_lock);
 	INIT_LIST_HEAD(&portdev->ports);
 
+	virtio_device_ready(portdev->vdev);
+
 	if (multiport) {
 		unsigned int nr_added_bufs;
 
@@ -2182,6 +2189,8 @@ static int virtcons_restore(struct virtio_device *vdev)
 	if (ret)
 		return ret;
 
+	virtio_device_ready(portdev->vdev);
+
 	if (use_multiport(portdev))
 		fill_queue(portdev->c_ivq, &portdev->c_ivq_lock);
 
diff --git a/drivers/char/xilinx_hwicap/xilinx_hwicap.c b/drivers/char/xilinx_hwicap/xilinx_hwicap.c
index 9b1a5ac4881d..c07dfe5c4da3 100644
--- a/drivers/char/xilinx_hwicap/xilinx_hwicap.c
+++ b/drivers/char/xilinx_hwicap/xilinx_hwicap.c
@@ -843,7 +843,6 @@ static struct platform_driver hwicap_platform_driver = {
 	.probe = hwicap_drv_probe,
 	.remove = hwicap_drv_remove,
 	.driver = {
-		.owner = THIS_MODULE,
 		.name = DRIVER_NAME,
 		.of_match_table = hwicap_of_match,
 	},
diff --git a/drivers/staging/xillybus/Kconfig b/drivers/char/xillybus/Kconfig
index b53bdf12da0d..b53bdf12da0d 100644
--- a/drivers/staging/xillybus/Kconfig
+++ b/drivers/char/xillybus/Kconfig
diff --git a/drivers/staging/xillybus/Makefile b/drivers/char/xillybus/Makefile
index b68b7ebfd381..b68b7ebfd381 100644
--- a/drivers/staging/xillybus/Makefile
+++ b/drivers/char/xillybus/Makefile
diff --git a/drivers/staging/xillybus/xillybus.h b/drivers/char/xillybus/xillybus.h
index a0806b5ee2cb..b9a9eb6d4f72 100644
--- a/drivers/staging/xillybus/xillybus.h
+++ b/drivers/char/xillybus/xillybus.h
@@ -146,7 +146,6 @@ struct xilly_mapping {
 	int direction;
 };
 
-
 irqreturn_t xillybus_isr(int irq, void *data);
 
 struct xilly_endpoint *xillybus_init_endpoint(struct pci_dev *pdev,
diff --git a/drivers/staging/xillybus/xillybus_core.c b/drivers/char/xillybus/xillybus_core.c
index 0214009f7513..b827fa095f1b 100644
--- a/drivers/staging/xillybus/xillybus_core.c
+++ b/drivers/char/xillybus/xillybus_core.c
@@ -133,30 +133,22 @@ irqreturn_t xillybus_isr(int irq, void *data)
 	unsigned int msg_channel, msg_bufno, msg_data, msg_dir;
 	struct xilly_channel *channel;
 
-	/*
-	 * The endpoint structure is altered during periods when it's
-	 * guaranteed no interrupt will occur, but in theory, the cache
-	 * lines may not be updated. So a memory barrier is issued.
-	 */
-	smp_rmb();
-
 	buf = ep->msgbuf_addr;
 	buf_size = ep->msg_buf_size/sizeof(u32);
 
-
 	ep->ephw->hw_sync_sgl_for_cpu(ep,
 				      ep->msgbuf_dma_addr,
 				      ep->msg_buf_size,
 				      DMA_FROM_DEVICE);
 
-	for (i = 0; i < buf_size; i += 2)
+	for (i = 0; i < buf_size; i += 2) {
 		if (((buf[i+1] >> 28) & 0xf) != ep->msg_counter) {
 			malformed_message(ep, &buf[i]);
 			dev_warn(ep->dev,
 				 "Sending a NACK on counter %x (instead of %x) on entry %d\n",
-				((buf[i+1] >> 28) & 0xf),
-				ep->msg_counter,
-				i/2);
+				 ((buf[i+1] >> 28) & 0xf),
+				 ep->msg_counter,
+				 i/2);
 
 			if (++ep->failed_messages > 10) {
 				dev_err(ep->dev,
@@ -174,15 +166,16 @@ irqreturn_t xillybus_isr(int irq, void *data)
 			return IRQ_HANDLED;
 		} else if (buf[i] & (1 << 22)) /* Last message */
 			break;
+	}
 
 	if (i >= buf_size) {
 		dev_err(ep->dev, "Bad interrupt message. Stopping.\n");
 		return IRQ_HANDLED;
 	}
 
-	buf_size = i;
+	buf_size = i + 2;
 
-	for (i = 0; i <= buf_size; i += 2) { /* Scan through messages */
+	for (i = 0; i < buf_size; i += 2) { /* Scan through messages */
 		opcode = (buf[i] >> 24) & 0xff;
 
 		msg_dir = buf[i] & 1;
@@ -192,7 +185,6 @@ irqreturn_t xillybus_isr(int irq, void *data)
 
 		switch (opcode) {
 		case XILLYMSG_OPCODE_RELEASEBUF:
-
 			if ((msg_channel > ep->num_channels) ||
 			    (msg_channel == 0)) {
 				malformed_message(ep, &buf[i]);
@@ -337,10 +329,9 @@ static int xilly_get_dma_buffers(struct xilly_endpoint *ep,
 	struct xilly_buffer *this_buffer = NULL; /* Init to silence warning */
 
 	if (buffers) { /* Not the message buffer */
-		this_buffer = devm_kzalloc(
-			dev, bufnum * sizeof(struct xilly_buffer),
-			GFP_KERNEL);
-
+		this_buffer = devm_kcalloc(dev, bufnum,
+					   sizeof(struct xilly_buffer),
+					   GFP_KERNEL);
 		if (!this_buffer)
 			return -ENOMEM;
 	}
@@ -372,16 +363,15 @@ static int xilly_get_dma_buffers(struct xilly_endpoint *ep,
 				dev,
 				GFP_KERNEL | __GFP_DMA32 | __GFP_ZERO,
 				allocorder);
-
 			if (!s->salami)
 				return -ENOMEM;
+
 			s->left_of_salami = allocsize;
 		}
 
 		rc = ep->ephw->map_single(ep, s->salami,
 					  bytebufsize, s->direction,
 					  &dma_addr);
-
 		if (rc)
 			return rc;
 
@@ -389,7 +379,6 @@ static int xilly_get_dma_buffers(struct xilly_endpoint *ep,
 			  ep->registers + fpga_dma_bufaddr_lowaddr_reg);
 		iowrite32(((u32) ((((u64) dma_addr) >> 32) & 0xffffffff)),
 			  ep->registers + fpga_dma_bufaddr_highaddr_reg);
-		mmiowb();
 
 		if (buffers) { /* Not the message buffer */
 			this_buffer->addr = s->salami;
@@ -410,13 +399,12 @@ static int xilly_get_dma_buffers(struct xilly_endpoint *ep,
 		s->left_of_salami -= bytebufsize;
 		s->salami += bytebufsize;
 	}
-	return 0; /* Success */
+	return 0;
 }
 
 static int xilly_setupchannels(struct xilly_endpoint *ep,
 			       unsigned char *chandesc,
-			       int entries
-	)
+			       int entries)
 {
 	struct device *dev = ep->dev;
 	int i, entry, rc;
@@ -443,18 +431,16 @@ static int xilly_setupchannels(struct xilly_endpoint *ep,
 		.regdirection = 0x80000000,
 	};
 
-	channel = devm_kzalloc(dev, ep->num_channels *
+	channel = devm_kcalloc(dev, ep->num_channels,
 			       sizeof(struct xilly_channel), GFP_KERNEL);
-
 	if (!channel)
-		goto memfail;
+		return -ENOMEM;
 
-	ep->channels = devm_kzalloc(dev, (ep->num_channels + 1) *
+	ep->channels = devm_kcalloc(dev, ep->num_channels + 1,
 				    sizeof(struct xilly_channel *),
 				    GFP_KERNEL);
-
 	if (!ep->channels)
-		goto memfail;
+		return -ENOMEM;
 
 	ep->channels[0] = NULL; /* Channel 0 is message buf. */
 
@@ -526,12 +512,11 @@ static int xilly_setupchannels(struct xilly_endpoint *ep,
 			bytebufsize = channel->rd_buf_size = bufsize *
 				(1 << channel->log2_element_size);
 
-			buffers = devm_kzalloc(dev,
-				bufnum * sizeof(struct xilly_buffer *),
-				GFP_KERNEL);
-
+			buffers = devm_kcalloc(dev, bufnum,
+					       sizeof(struct xilly_buffer *),
+					       GFP_KERNEL);
 			if (!buffers)
-				goto memfail;
+				return -ENOMEM;
 		} else {
 			bytebufsize = bufsize << 2;
 		}
@@ -566,7 +551,7 @@ static int xilly_setupchannels(struct xilly_endpoint *ep,
 		}
 
 		if (rc)
-			goto memfail;
+			return -ENOMEM;
 	}
 
 	if (!msg_buf_done) {
@@ -575,15 +560,10 @@ static int xilly_setupchannels(struct xilly_endpoint *ep,
 		return -ENODEV;
 	}
 	return 0;
-
-memfail:
-	dev_err(ep->dev,
-		"Failed to assign DMA buffer memory. Aborting.\n");
-	return -ENOMEM;
 }
 
-static void xilly_scan_idt(struct xilly_endpoint *endpoint,
-			   struct xilly_idt_handle *idt_handle)
+static int xilly_scan_idt(struct xilly_endpoint *endpoint,
+			  struct xilly_idt_handle *idt_handle)
 {
 	int count = 0;
 	unsigned char *idt = endpoint->channels[1]->wr_buffers[0]->addr;
@@ -607,53 +587,49 @@ static void xilly_scan_idt(struct xilly_endpoint *endpoint,
 	if (scan > end_of_idt) {
 		dev_err(endpoint->dev,
 			"IDT device name list overflow. Aborting.\n");
-		idt_handle->chandesc = NULL;
-		return;
+		return -ENODEV;
 	}
 	idt_handle->chandesc = scan;
 
 	len = endpoint->idtlen - (3 + ((int) (scan - idt)));
 
 	if (len & 0x03) {
-		idt_handle->chandesc = NULL;
-
 		dev_err(endpoint->dev,
 			"Corrupt IDT device name list. Aborting.\n");
+		return -ENODEV;
 	}
 
 	idt_handle->entries = len >> 2;
-
 	endpoint->num_channels = count;
+
+	return 0;
 }
 
 static int xilly_obtain_idt(struct xilly_endpoint *endpoint)
 {
-	int rc = 0;
 	struct xilly_channel *channel;
 	unsigned char *version;
+	long t;
 
 	channel = endpoint->channels[1]; /* This should be generated ad-hoc */
 
 	channel->wr_sleepy = 1;
-	wmb(); /* Setting wr_sleepy must come before the command */
 
 	iowrite32(1 |
-		   (3 << 24), /* Opcode 3 for channel 0 = Send IDT */
-		   endpoint->registers + fpga_buf_ctrl_reg);
-	mmiowb(); /* Just to appear safe */
+		  (3 << 24), /* Opcode 3 for channel 0 = Send IDT */
+		  endpoint->registers + fpga_buf_ctrl_reg);
 
-	wait_event_interruptible_timeout(channel->wr_wait,
-					 (!channel->wr_sleepy),
-					 XILLY_TIMEOUT);
+	t = wait_event_interruptible_timeout(channel->wr_wait,
+					     (!channel->wr_sleepy),
+					     XILLY_TIMEOUT);
 
-	if (channel->wr_sleepy) {
+	if (t <= 0) {
 		dev_err(endpoint->dev, "Failed to obtain IDT. Aborting.\n");
 
 		if (endpoint->fatal_error)
 			return -EIO;
 
-		rc = -ENODEV;
-		return rc;
+		return -ENODEV;
 	}
 
 	endpoint->ephw->hw_sync_sgl_for_cpu(
@@ -665,16 +641,14 @@ static int xilly_obtain_idt(struct xilly_endpoint *endpoint)
 	if (channel->wr_buffers[0]->end_offset != endpoint->idtlen) {
 		dev_err(endpoint->dev,
 			"IDT length mismatch (%d != %d). Aborting.\n",
-		       channel->wr_buffers[0]->end_offset, endpoint->idtlen);
-		rc = -ENODEV;
-		return rc;
+			channel->wr_buffers[0]->end_offset, endpoint->idtlen);
+		return -ENODEV;
 	}
 
 	if (crc32_le(~0, channel->wr_buffers[0]->addr,
 		     endpoint->idtlen+1) != 0) {
 		dev_err(endpoint->dev, "IDT failed CRC check. Aborting.\n");
-		rc = -ENODEV;
-		return rc;
+		return -ENODEV;
 	}
 
 	version = channel->wr_buffers[0]->addr;
@@ -683,12 +657,11 @@ static int xilly_obtain_idt(struct xilly_endpoint *endpoint)
 	if (*version > 0x82) {
 		dev_err(endpoint->dev,
 			"No support for IDT version 0x%02x. Maybe the xillybus driver needs an upgarde. Aborting.\n",
-		       (int) *version);
-		rc = -ENODEV;
-		return rc;
+			*version);
+		return -ENODEV;
 	}
 
-	return 0; /* Success */
+	return 0;
 }
 
 static ssize_t xillybus_read(struct file *filp, char __user *userbuf,
@@ -713,12 +686,9 @@ static ssize_t xillybus_read(struct file *filp, char __user *userbuf,
 	deadline = jiffies + 1 + XILLY_RX_TIMEOUT;
 
 	rc = mutex_lock_interruptible(&channel->wr_mutex);
-
 	if (rc)
 		return rc;
 
-	rc = 0; /* Just to be clear about it. Compiler optimizes this out */
-
 	while (1) { /* Note that we may drop mutex within this loop */
 		int bytes_to_do = count - bytes_done;
 
@@ -793,14 +763,11 @@ static ssize_t xillybus_read(struct file *filp, char __user *userbuf,
 			bytes_done += howmany;
 
 			if (bufferdone) {
-				channel->endpoint->ephw->
-					hw_sync_sgl_for_device
-					(
-						channel->endpoint,
-						channel->wr_buffers[bufidx]->
-						dma_addr,
-						channel->wr_buf_size,
-						DMA_FROM_DEVICE);
+				channel->endpoint->ephw->hw_sync_sgl_for_device(
+					channel->endpoint,
+					channel->wr_buffers[bufidx]->dma_addr,
+					channel->wr_buf_size,
+					DMA_FROM_DEVICE);
 
 				/*
 				 * Tell FPGA the buffer is done with. It's an
@@ -810,11 +777,10 @@ static ssize_t xillybus_read(struct file *filp, char __user *userbuf,
 				 * the channel-specific mutex.
 				 */
 
-				iowrite32(1 | (channel->chan_num << 1)
-					   | (bufidx << 12),
-					   channel->endpoint->registers +
-					   fpga_buf_ctrl_reg);
-				mmiowb(); /* Just to appear safe */
+				iowrite32(1 | (channel->chan_num << 1) |
+					  (bufidx << 12),
+					  channel->endpoint->registers +
+					  fpga_buf_ctrl_reg);
 			}
 
 			if (rc) {
@@ -851,7 +817,7 @@ static ssize_t xillybus_read(struct file *filp, char __user *userbuf,
 			if (ready)
 				goto desperate;
 
-			bytes_done = -EAGAIN;
+			rc = -EAGAIN;
 			break;
 		}
 
@@ -895,26 +861,21 @@ static ssize_t xillybus_read(struct file *filp, char __user *userbuf,
 
 			if (channel->wr_synchronous ||
 			    (offsetlimit < (buf_elements - 1))) {
-
 				mutex_lock(&channel->endpoint->register_mutex);
 
 				iowrite32(offsetlimit,
 					  channel->endpoint->registers +
 					  fpga_buf_offset_reg);
-				mmiowb();
 
 				iowrite32(1 | (channel->chan_num << 1) |
-					   (2 << 24) |  /* 2 = offset limit */
-					   (waiting_bufidx << 12),
-					   channel->endpoint->registers +
-					   fpga_buf_ctrl_reg);
-
-				mmiowb(); /* Just to appear safe */
+					  (2 << 24) |  /* 2 = offset limit */
+					  (waiting_bufidx << 12),
+					  channel->endpoint->registers +
+					  fpga_buf_ctrl_reg);
 
 				mutex_unlock(&channel->endpoint->
 					     register_mutex);
 			}
-
 		}
 
 		/*
@@ -925,7 +886,6 @@ static ssize_t xillybus_read(struct file *filp, char __user *userbuf,
 
 		if (!channel->wr_allow_partial ||
 		    (no_time_left && (bytes_done == 0))) {
-
 			/*
 			 * This do-loop will run more than once if another
 			 * thread reasserted wr_sleepy before we got the mutex
@@ -972,7 +932,7 @@ interrupted: /* Mutex is not held if got here */
 					(!channel->wr_sleepy),
 					left_to_sleep);
 
-			if (!channel->wr_sleepy)
+			if (left_to_sleep > 0) /* wr_sleepy deasserted */
 				continue;
 
 			if (left_to_sleep < 0) { /* Interrupt */
@@ -997,18 +957,18 @@ desperate:
 			 */
 
 			iowrite32(1 | (channel->chan_num << 1) |
-				   (3 << 24) |  /* Opcode 3, flush it all! */
-				   (waiting_bufidx << 12),
-				   channel->endpoint->registers +
-				   fpga_buf_ctrl_reg);
-			mmiowb(); /* Just to appear safe */
+				  (3 << 24) |  /* Opcode 3, flush it all! */
+				  (waiting_bufidx << 12),
+				  channel->endpoint->registers +
+				  fpga_buf_ctrl_reg);
 		}
 
 		/*
-		 * Formally speaking, we should block for data at this point.
-		 * But to keep the code cleaner, we'll just finish the loop,
-		 * make the unlikely check for data, and then block at the
-		 * usual place.
+		 * Reaching here means that we *do* have data in the buffer,
+		 * but the "partial" flag disallows returning less than
+		 * required. And we don't have as much. So loop again,
+		 * which is likely to end up blocking indefinitely until
+		 * enough data has arrived.
 		 */
 	}
 
@@ -1017,6 +977,9 @@ desperate:
 	if (channel->endpoint->fatal_error)
 		return -EIO;
 
+	if (rc)
+		return rc;
+
 	return bytes_done;
 }
 
@@ -1029,7 +992,7 @@ desperate:
 
 static int xillybus_myflush(struct xilly_channel *channel, long timeout)
 {
-	int rc = 0;
+	int rc;
 	unsigned long flags;
 
 	int end_offset_plus1;
@@ -1041,7 +1004,6 @@ static int xillybus_myflush(struct xilly_channel *channel, long timeout)
 	if (channel->endpoint->fatal_error)
 		return -EIO;
 	rc = mutex_lock_interruptible(&channel->rd_mutex);
-
 	if (rc)
 		return rc;
 
@@ -1056,7 +1018,9 @@ static int xillybus_myflush(struct xilly_channel *channel, long timeout)
 
 	bufidx = channel->rd_host_buf_idx;
 
-	bufidx_minus1 = (bufidx == 0) ? channel->num_rd_buffers - 1 : bufidx-1;
+	bufidx_minus1 = (bufidx == 0) ?
+		channel->num_rd_buffers - 1 :
+		bufidx - 1;
 
 	end_offset_plus1 = channel->rd_host_buf_pos >>
 		channel->log2_element_size;
@@ -1113,26 +1077,24 @@ static int xillybus_myflush(struct xilly_channel *channel, long timeout)
 
 		iowrite32(end_offset_plus1 - 1,
 			  channel->endpoint->registers + fpga_buf_offset_reg);
-		mmiowb();
 
 		iowrite32((channel->chan_num << 1) | /* Channel ID */
-			   (2 << 24) |  /* Opcode 2, submit buffer */
-			   (bufidx << 12),
-			   channel->endpoint->registers + fpga_buf_ctrl_reg);
-		mmiowb(); /* Just to appear safe */
+			  (2 << 24) |  /* Opcode 2, submit buffer */
+			  (bufidx << 12),
+			  channel->endpoint->registers + fpga_buf_ctrl_reg);
 
 		mutex_unlock(&channel->endpoint->register_mutex);
-	} else if (bufidx == 0)
+	} else if (bufidx == 0) {
 		bufidx = channel->num_rd_buffers - 1;
-	else
+	} else {
 		bufidx--;
+	}
 
 	channel->rd_host_buf_pos = new_rd_host_buf_pos;
 
 	if (timeout < 0)
 		goto done; /* Autoflush */
 
-
 	/*
 	 * bufidx is now the last buffer written to (or equal to
 	 * rd_fpga_buf_idx if buffer was never written to), and
@@ -1141,8 +1103,6 @@ static int xillybus_myflush(struct xilly_channel *channel, long timeout)
 	 * If bufidx == channel->rd_fpga_buf_idx we're either empty or full.
 	 */
 
-	rc = 0;
-
 	while (1) { /* Loop waiting for draining of buffers */
 		spin_lock_irqsave(&channel->rd_spinlock, flags);
 
@@ -1173,7 +1133,7 @@ static int xillybus_myflush(struct xilly_channel *channel, long timeout)
 				 (!channel->rd_full),
 				 timeout) == 0) {
 			dev_warn(channel->endpoint->dev,
-				"Timed out while flushing. Output data may be lost.\n");
+				 "Timed out while flushing. Output data may be lost.\n");
 
 			rc = -ETIMEDOUT;
 			break;
@@ -1211,7 +1171,6 @@ static void xillybus_autoflush(struct work_struct *work)
 	int rc;
 
 	rc = xillybus_myflush(channel, -1);
-
 	if (rc == -EINTR)
 		dev_warn(channel->endpoint->dev,
 			 "Autoflush failed because work queue thread got a signal.\n");
@@ -1238,12 +1197,9 @@ static ssize_t xillybus_write(struct file *filp, const char __user *userbuf,
 		return -EIO;
 
 	rc = mutex_lock_interruptible(&channel->rd_mutex);
-
 	if (rc)
 		return rc;
 
-	rc = 0; /* Just to be clear about it. Compiler optimizes this out */
-
 	while (1) {
 		int bytes_to_do = count - bytes_done;
 
@@ -1351,26 +1307,23 @@ static ssize_t xillybus_write(struct file *filp, const char __user *userbuf,
 			bytes_done += howmany;
 
 			if (bufferdone) {
-				channel->endpoint->ephw->
-					hw_sync_sgl_for_device(
-						channel->endpoint,
-						channel->rd_buffers[bufidx]->
-						dma_addr,
-						channel->rd_buf_size,
-						DMA_TO_DEVICE);
+				channel->endpoint->ephw->hw_sync_sgl_for_device(
+					channel->endpoint,
+					channel->rd_buffers[bufidx]->dma_addr,
+					channel->rd_buf_size,
+					DMA_TO_DEVICE);
 
 				mutex_lock(&channel->endpoint->register_mutex);
 
 				iowrite32(end_offset_plus1 - 1,
 					  channel->endpoint->registers +
 					  fpga_buf_offset_reg);
-				mmiowb();
+
 				iowrite32((channel->chan_num << 1) |
-					   (2 << 24) |  /* 2 = submit buffer */
-					   (bufidx << 12),
-					   channel->endpoint->registers +
-					   fpga_buf_ctrl_reg);
-				mmiowb(); /* Just to appear safe */
+					  (2 << 24) |  /* 2 = submit buffer */
+					  (bufidx << 12),
+					  channel->endpoint->registers +
+					  fpga_buf_ctrl_reg);
 
 				mutex_unlock(&channel->endpoint->
 					     register_mutex);
@@ -1411,14 +1364,12 @@ static ssize_t xillybus_write(struct file *filp, const char __user *userbuf,
 		 */
 
 		if (filp->f_flags & O_NONBLOCK) {
-			bytes_done = -EAGAIN;
+			rc = -EAGAIN;
 			break;
 		}
 
-		wait_event_interruptible(channel->rd_wait,
-					 (!channel->rd_full));
-
-		if (channel->rd_full) {
+		if (wait_event_interruptible(channel->rd_wait,
+					     (!channel->rd_full))) {
 			mutex_unlock(&channel->rd_mutex);
 
 			if (channel->endpoint->fatal_error)
@@ -1437,6 +1388,12 @@ static ssize_t xillybus_write(struct file *filp, const char __user *userbuf,
 				   &channel->rd_workitem,
 				   XILLY_RX_TIMEOUT);
 
+	if (channel->endpoint->fatal_error)
+		return -EIO;
+
+	if (rc)
+		return rc;
+
 	if ((channel->rd_synchronous) && (bytes_done > 0)) {
 		rc = xillybus_myflush(filp->private_data, 0); /* No timeout */
 
@@ -1444,9 +1401,6 @@ static ssize_t xillybus_write(struct file *filp, const char __user *userbuf,
 			return rc;
 	}
 
-	if (channel->endpoint->fatal_error)
-		return -EIO;
-
 	return bytes_done;
 }
 
@@ -1484,7 +1438,6 @@ static int xillybus_open(struct inode *inode, struct file *filp)
 	channel = endpoint->channels[1 + minor - endpoint->lowest_minor];
 	filp->private_data = channel;
 
-
 	/*
 	 * It gets complicated because:
 	 * 1. We don't want to take a mutex we don't have to
@@ -1545,7 +1498,6 @@ static int xillybus_open(struct inode *inode, struct file *filp)
 		goto unlock;
 	}
 
-
 	if (filp->f_mode & FMODE_READ) {
 		if (channel->wr_ref_count == 0) { /* First open of file */
 			/* Move the host to first buffer */
@@ -1566,7 +1518,6 @@ static int xillybus_open(struct inode *inode, struct file *filp)
 				  ((channel->wr_synchronous & 1) << 23),
 				  channel->endpoint->registers +
 				  fpga_buf_ctrl_reg);
-			mmiowb(); /* Just to appear safe */
 		}
 
 		channel->wr_ref_count++;
@@ -1588,7 +1539,6 @@ static int xillybus_open(struct inode *inode, struct file *filp)
 				  (4 << 24),   /* Opcode 4, open channel */
 				  channel->endpoint->registers +
 				  fpga_buf_ctrl_reg);
-			mmiowb(); /* Just to appear safe */
 		}
 
 		channel->rd_ref_count++;
@@ -1609,7 +1559,6 @@ unlock_wr:
 
 static int xillybus_release(struct inode *inode, struct file *filp)
 {
-	int rc;
 	unsigned long flags;
 	struct xilly_channel *channel = filp->private_data;
 
@@ -1620,18 +1569,11 @@ static int xillybus_release(struct inode *inode, struct file *filp)
 		return -EIO;
 
 	if (filp->f_mode & FMODE_WRITE) {
-		rc = mutex_lock_interruptible(&channel->rd_mutex);
-
-		if (rc) {
-			dev_warn(channel->endpoint->dev,
-				 "Failed to close file. Hardware left in messy state.\n");
-			return rc;
-		}
+		mutex_lock(&channel->rd_mutex);
 
 		channel->rd_ref_count--;
 
 		if (channel->rd_ref_count == 0) {
-
 			/*
 			 * We rely on the kernel calling flush()
 			 * before we get here.
@@ -1641,28 +1583,20 @@ static int xillybus_release(struct inode *inode, struct file *filp)
 				  (5 << 24),  /* Opcode 5, close channel */
 				  channel->endpoint->registers +
 				  fpga_buf_ctrl_reg);
-			mmiowb(); /* Just to appear safe */
 		}
 		mutex_unlock(&channel->rd_mutex);
 	}
 
 	if (filp->f_mode & FMODE_READ) {
-		rc = mutex_lock_interruptible(&channel->wr_mutex);
-		if (rc) {
-			dev_warn(channel->endpoint->dev,
-				 "Failed to close file. Hardware left in messy state.\n");
-			return rc;
-		}
+		mutex_lock(&channel->wr_mutex);
 
 		channel->wr_ref_count--;
 
 		if (channel->wr_ref_count == 0) {
-
 			iowrite32(1 | (channel->chan_num << 1) |
-				   (5 << 24),  /* Opcode 5, close channel */
-				   channel->endpoint->registers +
-				   fpga_buf_ctrl_reg);
-			mmiowb(); /* Just to appear safe */
+				  (5 << 24),  /* Opcode 5, close channel */
+				  channel->endpoint->registers +
+				  fpga_buf_ctrl_reg);
 
 			/*
 			 * This is crazily cautious: We make sure that not
@@ -1723,6 +1657,7 @@ static int xillybus_release(struct inode *inode, struct file *filp)
 
 	return 0;
 }
+
 static loff_t xillybus_llseek(struct file *filp, loff_t offset, int whence)
 {
 	struct xilly_channel *channel = filp->private_data;
@@ -1743,13 +1678,13 @@ static loff_t xillybus_llseek(struct file *filp, loff_t offset, int whence)
 	mutex_lock(&channel->rd_mutex);
 
 	switch (whence) {
-	case 0:
+	case SEEK_SET:
 		pos = offset;
 		break;
-	case 1:
+	case SEEK_CUR:
 		pos += offset;
 		break;
-	case 2:
+	case SEEK_END:
 		pos = offset; /* Going to the end => to the beginning */
 		break;
 	default:
@@ -1767,11 +1702,10 @@ static loff_t xillybus_llseek(struct file *filp, loff_t offset, int whence)
 
 	iowrite32(pos >> channel->log2_element_size,
 		  channel->endpoint->registers + fpga_buf_offset_reg);
-	mmiowb();
+
 	iowrite32((channel->chan_num << 1) |
 		  (6 << 24),  /* Opcode 6, set address */
 		  channel->endpoint->registers + fpga_buf_ctrl_reg);
-	mmiowb(); /* Just to appear safe */
 
 	mutex_unlock(&channel->endpoint->register_mutex);
 
@@ -1876,10 +1810,9 @@ static int xillybus_init_chrdev(struct xilly_endpoint *endpoint,
 	rc = alloc_chrdev_region(&dev, 0, /* minor start */
 				 endpoint->num_channels,
 				 xillyname);
-
 	if (rc) {
 		dev_warn(endpoint->dev, "Failed to obtain major/minors");
-		goto error1;
+		return rc;
 	}
 
 	endpoint->major = major = MAJOR(dev);
@@ -1891,7 +1824,7 @@ static int xillybus_init_chrdev(struct xilly_endpoint *endpoint,
 		      endpoint->num_channels);
 	if (rc) {
 		dev_warn(endpoint->dev, "Failed to add cdev. Aborting.\n");
-		goto error2;
+		goto unregister_chrdev;
 	}
 
 	idt++;
@@ -1916,7 +1849,8 @@ static int xillybus_init_chrdev(struct xilly_endpoint *endpoint,
 			dev_warn(endpoint->dev,
 				 "Failed to create %s device. Aborting.\n",
 				 devname);
-			goto error3;
+			rc = -ENODEV;
+			goto unroll_device_create;
 		}
 	}
 
@@ -1924,15 +1858,14 @@ static int xillybus_init_chrdev(struct xilly_endpoint *endpoint,
 		 endpoint->num_channels);
 	return 0; /* succeed */
 
-error3:
+unroll_device_create:
 	devnum--; i--;
 	for (; devnum >= 0; devnum--, i--)
 		device_destroy(xillybus_class, MKDEV(major, i));
 
 	cdev_del(&endpoint->cdev);
-error2:
+unregister_chrdev:
 	unregister_chrdev_region(MKDEV(major, minor), endpoint->num_channels);
-error1:
 
 	return rc;
 }
@@ -1954,7 +1887,6 @@ static void xillybus_cleanup_chrdev(struct xilly_endpoint *endpoint)
 		 endpoint->num_channels);
 }
 
-
 struct xilly_endpoint *xillybus_init_endpoint(struct pci_dev *pdev,
 					      struct device *dev,
 					      struct xilly_endpoint_hardware
@@ -1982,27 +1914,28 @@ EXPORT_SYMBOL(xillybus_init_endpoint);
 
 static int xilly_quiesce(struct xilly_endpoint *endpoint)
 {
+	long t;
+
 	endpoint->idtlen = -1;
-	wmb(); /* Make sure idtlen is set before sending command */
+
 	iowrite32((u32) (endpoint->dma_using_dac & 0x0001),
 		  endpoint->registers + fpga_dma_control_reg);
-	mmiowb();
 
-	wait_event_interruptible_timeout(endpoint->ep_wait,
-					 (endpoint->idtlen >= 0),
-					 XILLY_TIMEOUT);
-
-	if (endpoint->idtlen < 0) {
+	t = wait_event_interruptible_timeout(endpoint->ep_wait,
+					     (endpoint->idtlen >= 0),
+					     XILLY_TIMEOUT);
+	if (t <= 0) {
 		dev_err(endpoint->dev,
 			"Failed to quiesce the device on exit.\n");
 		return -ENODEV;
 	}
-	return 0; /* Success */
+	return 0;
 }
 
 int xillybus_endpoint_discovery(struct xilly_endpoint *endpoint)
 {
-	int rc = 0;
+	int rc;
+	long t;
 
 	void *bootstrap_resources;
 	int idtbuffersize = (1 << PAGE_SHIFT);
@@ -2026,7 +1959,6 @@ int xillybus_endpoint_discovery(struct xilly_endpoint *endpoint)
 	 */
 
 	iowrite32(1, endpoint->registers + fpga_endian_reg);
-	mmiowb(); /* Writes below are affected by the one above. */
 
 	/* Bootstrap phase I: Allocate temporary message buffer */
 
@@ -2037,39 +1969,32 @@ int xillybus_endpoint_discovery(struct xilly_endpoint *endpoint)
 	endpoint->num_channels = 0;
 
 	rc = xilly_setupchannels(endpoint, bogus_idt, 1);
-
 	if (rc)
 		return rc;
 
 	/* Clear the message subsystem (and counter in particular) */
 	iowrite32(0x04, endpoint->registers + fpga_msg_ctrl_reg);
-	mmiowb();
 
 	endpoint->idtlen = -1;
 
-	smp_wmb();
-
 	/*
 	 * Set DMA 32/64 bit mode, quiesce the device (?!) and get IDT
 	 * buffer size.
 	 */
 	iowrite32((u32) (endpoint->dma_using_dac & 0x0001),
-		   endpoint->registers + fpga_dma_control_reg);
-	mmiowb();
-
-	wait_event_interruptible_timeout(endpoint->ep_wait,
-					 (endpoint->idtlen >= 0),
-					 XILLY_TIMEOUT);
+		  endpoint->registers + fpga_dma_control_reg);
 
-	if (endpoint->idtlen < 0) {
+	t = wait_event_interruptible_timeout(endpoint->ep_wait,
+					     (endpoint->idtlen >= 0),
+					     XILLY_TIMEOUT);
+	if (t <= 0) {
 		dev_err(endpoint->dev, "No response from FPGA. Aborting.\n");
 		return -ENODEV;
 	}
 
 	/* Enable DMA */
 	iowrite32((u32) (0x0002 | (endpoint->dma_using_dac & 0x0001)),
-		   endpoint->registers + fpga_dma_control_reg);
-	mmiowb();
+		  endpoint->registers + fpga_dma_control_reg);
 
 	/* Bootstrap phase II: Allocate buffer for IDT and obtain it */
 	while (endpoint->idtlen >= idtbuffersize) {
@@ -2080,23 +2005,16 @@ int xillybus_endpoint_discovery(struct xilly_endpoint *endpoint)
 	endpoint->num_channels = 1;
 
 	rc = xilly_setupchannels(endpoint, bogus_idt, 2);
-
 	if (rc)
 		goto failed_idt;
 
-	smp_wmb();
-
 	rc = xilly_obtain_idt(endpoint);
-
 	if (rc)
 		goto failed_idt;
 
-	xilly_scan_idt(endpoint, &idt_handle);
-
-	if (!idt_handle.chandesc) {
-		rc = -ENODEV;
+	rc = xilly_scan_idt(endpoint, &idt_handle);
+	if (rc)
 		goto failed_idt;
-	}
 
 	devres_close_group(dev, bootstrap_resources);
 
@@ -2105,12 +2023,9 @@ int xillybus_endpoint_discovery(struct xilly_endpoint *endpoint)
 	rc = xilly_setupchannels(endpoint,
 				 idt_handle.chandesc,
 				 idt_handle.entries);
-
 	if (rc)
 		goto failed_idt;
 
-	smp_wmb(); /* mutex_lock below should suffice, but won't hurt.*/
-
 	/*
 	 * endpoint is now completely configured. We put it on the list
 	 * available to open() before registering the char device(s)
@@ -2121,7 +2036,6 @@ int xillybus_endpoint_discovery(struct xilly_endpoint *endpoint)
 	mutex_unlock(&ep_list_lock);
 
 	rc = xillybus_init_chrdev(endpoint, idt_handle.idt);
-
 	if (rc)
 		goto failed_chrdevs;
 
@@ -2162,25 +2076,19 @@ EXPORT_SYMBOL(xillybus_endpoint_remove);
 
 static int __init xillybus_init(void)
 {
-	int rc = 0;
-
 	mutex_init(&ep_list_lock);
 
 	xillybus_class = class_create(THIS_MODULE, xillyname);
-	if (IS_ERR(xillybus_class)) {
-		rc = PTR_ERR(xillybus_class);
-		pr_warn("Failed to register class xillybus\n");
-
-		return rc;
-	}
+	if (IS_ERR(xillybus_class))
+		return PTR_ERR(xillybus_class);
 
 	xillybus_wq = alloc_workqueue(xillyname, 0, 0);
 	if (!xillybus_wq) {
 		class_destroy(xillybus_class);
-		rc = -ENOMEM;
+		return -ENOMEM;
 	}
 
-	return rc;
+	return 0;
 }
 
 static void __exit xillybus_exit(void)
diff --git a/drivers/staging/xillybus/xillybus_of.c b/drivers/char/xillybus/xillybus_of.c
index e0ae2346b820..2002a3a28146 100644
--- a/drivers/staging/xillybus/xillybus_of.c
+++ b/drivers/char/xillybus/xillybus_of.c
@@ -106,9 +106,10 @@ static int xilly_map_single_of(struct xilly_endpoint *ep,
 	if (rc) {
 		dma_unmap_single(ep->dev, addr, size, direction);
 		kfree(this);
+		return rc;
 	}
 
-	return rc;
+	return 0;
 }
 
 static struct xilly_endpoint_hardware of_hw = {
@@ -129,7 +130,7 @@ static int xilly_drv_probe(struct platform_device *op)
 {
 	struct device *dev = &op->dev;
 	struct xilly_endpoint *endpoint;
-	int rc = 0;
+	int rc;
 	int irq;
 	struct resource res;
 	struct xilly_endpoint_hardware *ephw = &of_hw;
@@ -178,7 +179,6 @@ static struct platform_driver xillybus_platform_driver = {
 	.remove = xilly_drv_remove,
 	.driver = {
 		.name = xillyname,
-		.owner = THIS_MODULE,
 		.of_match_table = xillybus_of_match,
 	},
 };
diff --git a/drivers/staging/xillybus/xillybus_pcie.c b/drivers/char/xillybus/xillybus_pcie.c
index 96c2c9f67e0c..d8266bc2ae35 100644
--- a/drivers/staging/xillybus/xillybus_pcie.c
+++ b/drivers/char/xillybus/xillybus_pcie.c
@@ -98,7 +98,7 @@ static int xilly_map_single_pci(struct xilly_endpoint *ep,
 	int pci_direction;
 	dma_addr_t addr;
 	struct xilly_mapping *this;
-	int rc = 0;
+	int rc;
 
 	this = kzalloc(sizeof(*this), GFP_KERNEL);
 	if (!this)
@@ -121,13 +121,13 @@ static int xilly_map_single_pci(struct xilly_endpoint *ep,
 	*ret_dma_handle = addr;
 
 	rc = devm_add_action(ep->dev, xilly_pci_unmap, this);
-
 	if (rc) {
 		pci_unmap_single(ep->pdev, addr, size, pci_direction);
 		kfree(this);
+		return rc;
 	}
 
-	return rc;
+	return 0;
 }
 
 static struct xilly_endpoint_hardware pci_hw = {
@@ -138,10 +138,10 @@ static struct xilly_endpoint_hardware pci_hw = {
 };
 
 static int xilly_probe(struct pci_dev *pdev,
-				 const struct pci_device_id *ent)
+		       const struct pci_device_id *ent)
 {
 	struct xilly_endpoint *endpoint;
-	int rc = 0;
+	int rc;
 
 	endpoint = xillybus_init_endpoint(pdev, &pdev->dev, &pci_hw);
 
@@ -151,7 +151,6 @@ static int xilly_probe(struct pci_dev *pdev,
 	pci_set_drvdata(pdev, endpoint);
 
 	rc = pcim_enable_device(pdev);
-
 	if (rc) {
 		dev_err(endpoint->dev,
 			"pcim_enable_device() failed. Aborting.\n");
@@ -187,7 +186,6 @@ static int xilly_probe(struct pci_dev *pdev,
 	}
 	rc = devm_request_irq(&pdev->dev, pdev->irq, xillybus_isr, 0,
 			      xillyname, endpoint);
-
 	if (rc) {
 		dev_err(endpoint->dev,
 			"Failed to register MSI handler. Aborting.\n");
@@ -201,9 +199,9 @@ static int xilly_probe(struct pci_dev *pdev,
 	 * nobody and use 32 bits DMA addressing in any case.
 	 */
 
-	if (!pci_set_dma_mask(pdev, DMA_BIT_MASK(32)))
+	if (!pci_set_dma_mask(pdev, DMA_BIT_MASK(32))) {
 		endpoint->dma_using_dac = 0;
-	else {
+	} else {
 		dev_err(endpoint->dev, "Failed to set DMA mask. Aborting.\n");
 		return -ENODEV;
 	}
diff --git a/drivers/clk/Kconfig b/drivers/clk/Kconfig
index cfd3af7b2cbd..3f44f292d066 100644
--- a/drivers/clk/Kconfig
+++ b/drivers/clk/Kconfig
@@ -28,16 +28,36 @@ config COMMON_CLK_WM831X
 	depends on MFD_WM831X
 	---help---
           Supports the clocking subsystem of the WM831x/2x series of
-	  PMICs from Wolfson Microlectronics.
+	  PMICs from Wolfson Microelectronics.
 
 source "drivers/clk/versatile/Kconfig"
 
+config COMMON_CLK_MAX_GEN
+        bool
+
 config COMMON_CLK_MAX77686
 	tristate "Clock driver for Maxim 77686 MFD"
 	depends on MFD_MAX77686
+	select COMMON_CLK_MAX_GEN
 	---help---
 	  This driver supports Maxim 77686 crystal oscillator clock. 
 
+config COMMON_CLK_MAX77802
+	tristate "Clock driver for Maxim 77802 PMIC"
+	depends on MFD_MAX77686
+	select COMMON_CLK_MAX_GEN
+	---help---
+	  This driver supports Maxim 77802 crystal oscillator clock.
+
+config COMMON_CLK_RK808
+	tristate "Clock driver for RK808"
+	depends on MFD_RK808
+	---help---
+	  This driver supports RK808 crystal oscillator clock. These
+	  multi-function devices have two fixed-rate oscillators,
+	  clocked at 32KHz each. Clkout1 is always on, Clkout2 can off
+	  by control register.
+
 config COMMON_CLK_SI5351
 	tristate "Clock driver for SiLabs 5351A/B/C"
 	depends on I2C
@@ -109,6 +129,11 @@ config COMMON_CLK_PALMAS
 	  This driver supports TI Palmas devices 32KHz output KG and KG_AUDIO
 	  using common clock framework.
 
+config COMMON_CLK_PXA
+	def_bool COMMON_CLK && ARCH_PXA
+	---help---
+	  Sypport for the Marvell PXA SoC.
+
 source "drivers/clk/qcom/Kconfig"
 
 endmenu
diff --git a/drivers/clk/Makefile b/drivers/clk/Makefile
index f537a0b1f798..d5fba5bc6e1b 100644
--- a/drivers/clk/Makefile
+++ b/drivers/clk/Makefile
@@ -9,6 +9,7 @@ obj-$(CONFIG_COMMON_CLK)	+= clk-gate.o
 obj-$(CONFIG_COMMON_CLK)	+= clk-mux.o
 obj-$(CONFIG_COMMON_CLK)	+= clk-composite.o
 obj-$(CONFIG_COMMON_CLK)	+= clk-fractional-divider.o
+obj-$(CONFIG_COMMON_CLK)	+= clk-gpio-gate.o
 ifeq ($(CONFIG_OF), y)
 obj-$(CONFIG_COMMON_CLK)	+= clk-conf.o
 endif
@@ -22,12 +23,15 @@ obj-$(CONFIG_ARCH_CLPS711X)		+= clk-clps711x.o
 obj-$(CONFIG_ARCH_EFM32)		+= clk-efm32gg.o
 obj-$(CONFIG_ARCH_HIGHBANK)		+= clk-highbank.o
 obj-$(CONFIG_MACH_LOONGSON1)		+= clk-ls1x.o
+obj-$(CONFIG_COMMON_CLK_MAX_GEN)	+= clk-max-gen.o
 obj-$(CONFIG_COMMON_CLK_MAX77686)	+= clk-max77686.o
+obj-$(CONFIG_COMMON_CLK_MAX77802)	+= clk-max77802.o
 obj-$(CONFIG_ARCH_MOXART)		+= clk-moxart.o
 obj-$(CONFIG_ARCH_NOMADIK)		+= clk-nomadik.o
 obj-$(CONFIG_ARCH_NSPIRE)		+= clk-nspire.o
 obj-$(CONFIG_COMMON_CLK_PALMAS)		+= clk-palmas.o
 obj-$(CONFIG_CLK_PPC_CORENET)		+= clk-ppc-corenet.o
+obj-$(CONFIG_COMMON_CLK_RK808)		+= clk-rk808.o
 obj-$(CONFIG_COMMON_CLK_S2MPS11)	+= clk-s2mps11.o
 obj-$(CONFIG_COMMON_CLK_SI5351)		+= clk-si5351.o
 obj-$(CONFIG_COMMON_CLK_SI570)		+= clk-si570.o
@@ -48,6 +52,7 @@ obj-$(CONFIG_ARCH_MMP)			+= mmp/
 endif
 obj-$(CONFIG_PLAT_ORION)		+= mvebu/
 obj-$(CONFIG_ARCH_MXS)			+= mxs/
+obj-$(CONFIG_COMMON_CLK_PXA)		+= pxa/
 obj-$(CONFIG_COMMON_CLK_QCOM)		+= qcom/
 obj-$(CONFIG_ARCH_ROCKCHIP)		+= rockchip/
 obj-$(CONFIG_COMMON_CLK_SAMSUNG)	+= samsung/
diff --git a/drivers/clk/at91/Makefile b/drivers/clk/at91/Makefile
index 4998aee59267..89a48a7bd5df 100644
--- a/drivers/clk/at91/Makefile
+++ b/drivers/clk/at91/Makefile
@@ -9,3 +9,4 @@ obj-y += clk-system.o clk-peripheral.o clk-programmable.o
 obj-$(CONFIG_HAVE_AT91_UTMI)		+= clk-utmi.o
 obj-$(CONFIG_HAVE_AT91_USB_CLK)		+= clk-usb.o
 obj-$(CONFIG_HAVE_AT91_SMD)		+= clk-smd.o
+obj-$(CONFIG_HAVE_AT91_H32MX)		+= clk-h32mx.o
diff --git a/drivers/clk/at91/clk-h32mx.c b/drivers/clk/at91/clk-h32mx.c
new file mode 100644
index 000000000000..152dcb3f7b5f
--- /dev/null
+++ b/drivers/clk/at91/clk-h32mx.c
@@ -0,0 +1,123 @@
+/*
+ * clk-h32mx.c
+ *
+ *  Copyright (C) 2014 Atmel
+ *
+ * Alexandre Belloni <alexandre.belloni@free-electrons.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ */
+
+#include <linux/clk-provider.h>
+#include <linux/clkdev.h>
+#include <linux/clk/at91_pmc.h>
+#include <linux/delay.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/io.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/sched.h>
+#include <linux/wait.h>
+
+#include "pmc.h"
+
+#define H32MX_MAX_FREQ	90000000
+
+struct clk_sama5d4_h32mx {
+	struct clk_hw hw;
+	struct at91_pmc *pmc;
+};
+
+#define to_clk_sama5d4_h32mx(hw) container_of(hw, struct clk_sama5d4_h32mx, hw)
+
+static unsigned long clk_sama5d4_h32mx_recalc_rate(struct clk_hw *hw,
+						 unsigned long parent_rate)
+{
+	struct clk_sama5d4_h32mx *h32mxclk = to_clk_sama5d4_h32mx(hw);
+
+	if (pmc_read(h32mxclk->pmc, AT91_PMC_MCKR) & AT91_PMC_H32MXDIV)
+		return parent_rate / 2;
+
+	if (parent_rate > H32MX_MAX_FREQ)
+		pr_warn("H32MX clock is too fast\n");
+	return parent_rate;
+}
+
+static long clk_sama5d4_h32mx_round_rate(struct clk_hw *hw, unsigned long rate,
+				       unsigned long *parent_rate)
+{
+	unsigned long div;
+
+	if (rate > *parent_rate)
+		return *parent_rate;
+	div = *parent_rate / 2;
+	if (rate < div)
+		return div;
+
+	if (rate - div < *parent_rate - rate)
+		return div;
+
+	return *parent_rate;
+}
+
+static int clk_sama5d4_h32mx_set_rate(struct clk_hw *hw, unsigned long rate,
+				    unsigned long parent_rate)
+{
+	struct clk_sama5d4_h32mx *h32mxclk = to_clk_sama5d4_h32mx(hw);
+	struct at91_pmc *pmc = h32mxclk->pmc;
+	u32 tmp;
+
+	if (parent_rate != rate && (parent_rate / 2) != rate)
+		return -EINVAL;
+
+	pmc_lock(pmc);
+	tmp = pmc_read(pmc, AT91_PMC_MCKR) & ~AT91_PMC_H32MXDIV;
+	if ((parent_rate / 2) == rate)
+		tmp |= AT91_PMC_H32MXDIV;
+	pmc_write(pmc, AT91_PMC_MCKR, tmp);
+	pmc_unlock(pmc);
+
+	return 0;
+}
+
+static const struct clk_ops h32mx_ops = {
+	.recalc_rate = clk_sama5d4_h32mx_recalc_rate,
+	.round_rate = clk_sama5d4_h32mx_round_rate,
+	.set_rate = clk_sama5d4_h32mx_set_rate,
+};
+
+void __init of_sama5d4_clk_h32mx_setup(struct device_node *np,
+				     struct at91_pmc *pmc)
+{
+	struct clk_sama5d4_h32mx *h32mxclk;
+	struct clk_init_data init;
+	const char *parent_name;
+	struct clk *clk;
+
+	h32mxclk = kzalloc(sizeof(*h32mxclk), GFP_KERNEL);
+	if (!h32mxclk)
+		return;
+
+	parent_name = of_clk_get_parent_name(np, 0);
+
+	init.name = np->name;
+	init.ops = &h32mx_ops;
+	init.parent_names = parent_name ? &parent_name : NULL;
+	init.num_parents = parent_name ? 1 : 0;
+	init.flags = CLK_SET_RATE_GATE;
+
+	h32mxclk->hw.init = &init;
+	h32mxclk->pmc = pmc;
+
+	clk = clk_register(NULL, &h32mxclk->hw);
+	if (!clk)
+		return;
+
+	of_clk_add_provider(np, of_clk_src_simple_get, clk);
+}
diff --git a/drivers/clk/at91/clk-pll.c b/drivers/clk/at91/clk-pll.c
index cf6ed023504c..6ec79dbc0840 100644
--- a/drivers/clk/at91/clk-pll.c
+++ b/drivers/clk/at91/clk-pll.c
@@ -15,6 +15,7 @@
 #include <linux/of_address.h>
 #include <linux/of_irq.h>
 #include <linux/io.h>
+#include <linux/kernel.h>
 #include <linux/wait.h>
 #include <linux/sched.h>
 #include <linux/interrupt.h>
@@ -29,9 +30,12 @@
 #define PLL_DIV(reg)		((reg) & PLL_DIV_MASK)
 #define PLL_MUL(reg, layout)	(((reg) >> (layout)->mul_shift) & \
 				 (layout)->mul_mask)
+#define PLL_MUL_MIN		2
+#define PLL_MUL_MASK(layout)	((layout)->mul_mask)
+#define PLL_MUL_MAX(layout)	(PLL_MUL_MASK(layout) + 1)
 #define PLL_ICPR_SHIFT(id)	((id) * 16)
 #define PLL_ICPR_MASK(id)	(0xffff << PLL_ICPR_SHIFT(id))
-#define PLL_MAX_COUNT		0x3ff
+#define PLL_MAX_COUNT		0x3f
 #define PLL_COUNT_SHIFT		8
 #define PLL_OUT_SHIFT		14
 #define PLL_MAX_ID		1
@@ -147,115 +151,113 @@ static unsigned long clk_pll_recalc_rate(struct clk_hw *hw,
 					 unsigned long parent_rate)
 {
 	struct clk_pll *pll = to_clk_pll(hw);
-	const struct clk_pll_layout *layout = pll->layout;
-	struct at91_pmc *pmc = pll->pmc;
-	int offset = PLL_REG(pll->id);
-	u32 tmp = pmc_read(pmc, offset) & layout->pllr_mask;
-	u8 div = PLL_DIV(tmp);
-	u16 mul = PLL_MUL(tmp, layout);
-	if (!div || !mul)
+
+	if (!pll->div || !pll->mul)
 		return 0;
 
-	return (parent_rate * (mul + 1)) / div;
+	return (parent_rate / pll->div) * (pll->mul + 1);
 }
 
 static long clk_pll_get_best_div_mul(struct clk_pll *pll, unsigned long rate,
 				     unsigned long parent_rate,
 				     u32 *div, u32 *mul,
 				     u32 *index) {
-	unsigned long maxrate;
-	unsigned long minrate;
-	unsigned long divrate;
-	unsigned long bestdiv = 1;
-	unsigned long bestmul;
-	unsigned long tmpdiv;
-	unsigned long roundup;
-	unsigned long rounddown;
-	unsigned long remainder;
-	unsigned long bestremainder;
-	unsigned long maxmul;
-	unsigned long maxdiv;
-	unsigned long mindiv;
-	int i = 0;
 	const struct clk_pll_layout *layout = pll->layout;
 	const struct clk_pll_characteristics *characteristics =
 							pll->characteristics;
+	unsigned long bestremainder = ULONG_MAX;
+	unsigned long maxdiv, mindiv, tmpdiv;
+	long bestrate = -ERANGE;
+	unsigned long bestdiv;
+	unsigned long bestmul;
+	int i = 0;
 
-	/* Minimum divider = 1 */
-	/* Maximum multiplier = max_mul */
-	maxmul = layout->mul_mask + 1;
-	maxrate = (parent_rate * maxmul) / 1;
-
-	/* Maximum divider = max_div */
-	/* Minimum multiplier = 2 */
-	maxdiv = PLL_DIV_MAX;
-	minrate = (parent_rate * 2) / maxdiv;
-
+	/* Check if parent_rate is a valid input rate */
 	if (parent_rate < characteristics->input.min ||
-	    parent_rate < characteristics->input.max)
-		return -ERANGE;
-
-	if (parent_rate < minrate || parent_rate > maxrate)
+	    parent_rate > characteristics->input.max)
 		return -ERANGE;
 
-	for (i = 0; i < characteristics->num_output; i++) {
-		if (parent_rate >= characteristics->output[i].min &&
-		    parent_rate <= characteristics->output[i].max)
-			break;
-	}
-
-	if (i >= characteristics->num_output)
-		return -ERANGE;
-
-	bestmul = rate / parent_rate;
-	rounddown = parent_rate % rate;
-	roundup = rate - rounddown;
-	bestremainder = roundup < rounddown ? roundup : rounddown;
-
-	if (!bestremainder) {
-		if (div)
-			*div = bestdiv;
-		if (mul)
-			*mul = bestmul;
-		if (index)
-			*index = i;
-		return rate;
-	}
-
-	maxdiv = 255 / (bestmul + 1);
-	if (parent_rate / maxdiv < characteristics->input.min)
-		maxdiv = parent_rate / characteristics->input.min;
-	mindiv = parent_rate / characteristics->input.max;
-	if (parent_rate % characteristics->input.max)
-		mindiv++;
-
-	for (tmpdiv = mindiv; tmpdiv < maxdiv; tmpdiv++) {
-		divrate = parent_rate / tmpdiv;
-
-		rounddown = rate % divrate;
-		roundup = divrate - rounddown;
-		remainder = roundup < rounddown ? roundup : rounddown;
-
+	/*
+	 * Calculate minimum divider based on the minimum multiplier, the
+	 * parent_rate and the requested rate.
+	 * Should always be 2 according to the input and output characteristics
+	 * of the PLL blocks.
+	 */
+	mindiv = (parent_rate * PLL_MUL_MIN) / rate;
+	if (!mindiv)
+		mindiv = 1;
+
+	/*
+	 * Calculate the maximum divider which is limited by PLL register
+	 * layout (limited by the MUL or DIV field size).
+	 */
+	maxdiv = DIV_ROUND_UP(parent_rate * PLL_MUL_MAX(layout), rate);
+	if (maxdiv > PLL_DIV_MAX)
+		maxdiv = PLL_DIV_MAX;
+
+	/*
+	 * Iterate over the acceptable divider values to find the best
+	 * divider/multiplier pair (the one that generates the closest
+	 * rate to the requested one).
+	 */
+	for (tmpdiv = mindiv; tmpdiv <= maxdiv; tmpdiv++) {
+		unsigned long remainder;
+		unsigned long tmprate;
+		unsigned long tmpmul;
+
+		/*
+		 * Calculate the multiplier associated with the current
+		 * divider that provide the closest rate to the requested one.
+		 */
+		tmpmul = DIV_ROUND_CLOSEST(rate, parent_rate / tmpdiv);
+		tmprate = (parent_rate / tmpdiv) * tmpmul;
+		if (tmprate > rate)
+			remainder = tmprate - rate;
+		else
+			remainder = rate - tmprate;
+
+		/*
+		 * Compare the remainder with the best remainder found until
+		 * now and elect a new best multiplier/divider pair if the
+		 * current remainder is smaller than the best one.
+		 */
 		if (remainder < bestremainder) {
 			bestremainder = remainder;
-			bestmul = rate / divrate;
 			bestdiv = tmpdiv;
+			bestmul = tmpmul;
+			bestrate = tmprate;
 		}
 
+		/*
+		 * We've found a perfect match!
+		 * Stop searching now and use this multiplier/divider pair.
+		 */
 		if (!remainder)
 			break;
 	}
 
-	rate = (parent_rate / bestdiv) * bestmul;
+	/* We haven't found any multiplier/divider pair => return -ERANGE */
+	if (bestrate < 0)
+		return bestrate;
+
+	/* Check if bestrate is a valid output rate  */
+	for (i = 0; i < characteristics->num_output; i++) {
+		if (bestrate >= characteristics->output[i].min &&
+		    bestrate <= characteristics->output[i].max)
+			break;
+	}
+
+	if (i >= characteristics->num_output)
+		return -ERANGE;
 
 	if (div)
 		*div = bestdiv;
 	if (mul)
-		*mul = bestmul;
+		*mul = bestmul - 1;
 	if (index)
 		*index = i;
 
-	return rate;
+	return bestrate;
 }
 
 static long clk_pll_round_rate(struct clk_hw *hw, unsigned long rate,
diff --git a/drivers/clk/at91/clk-programmable.c b/drivers/clk/at91/clk-programmable.c
index 62e2509f9df1..bbdb1b985c91 100644
--- a/drivers/clk/at91/clk-programmable.c
+++ b/drivers/clk/at91/clk-programmable.c
@@ -57,7 +57,7 @@ static unsigned long clk_programmable_recalc_rate(struct clk_hw *hw,
 static long clk_programmable_determine_rate(struct clk_hw *hw,
 					    unsigned long rate,
 					    unsigned long *best_parent_rate,
-					    struct clk **best_parent_clk)
+					    struct clk_hw **best_parent_hw)
 {
 	struct clk *parent = NULL;
 	long best_rate = -EINVAL;
@@ -84,7 +84,7 @@ static long clk_programmable_determine_rate(struct clk_hw *hw,
 		if (best_rate < 0 || (rate - tmp_rate) < (rate - best_rate)) {
 			best_rate = tmp_rate;
 			*best_parent_rate = parent_rate;
-			*best_parent_clk = parent;
+			*best_parent_hw = __clk_get_hw(parent);
 		}
 
 		if (!best_rate)
diff --git a/drivers/clk/at91/clk-slow.c b/drivers/clk/at91/clk-slow.c
index 32f7c1b36204..2f13bd5246b5 100644
--- a/drivers/clk/at91/clk-slow.c
+++ b/drivers/clk/at91/clk-slow.c
@@ -70,6 +70,7 @@ struct clk_sam9x5_slow {
 
 #define to_clk_sam9x5_slow(hw) container_of(hw, struct clk_sam9x5_slow, hw)
 
+static struct clk *slow_clk;
 
 static int clk_slow_osc_prepare(struct clk_hw *hw)
 {
@@ -357,6 +358,8 @@ at91_clk_register_sam9x5_slow(void __iomem *sckcr,
 	clk = clk_register(NULL, &slowck->hw);
 	if (IS_ERR(clk))
 		kfree(slowck);
+	else
+		slow_clk = clk;
 
 	return clk;
 }
@@ -433,6 +436,8 @@ at91_clk_register_sam9260_slow(struct at91_pmc *pmc,
 	clk = clk_register(NULL, &slowck->hw);
 	if (IS_ERR(clk))
 		kfree(slowck);
+	else
+		slow_clk = clk;
 
 	return clk;
 }
@@ -465,3 +470,25 @@ void __init of_at91sam9260_clk_slow_setup(struct device_node *np,
 
 	of_clk_add_provider(np, of_clk_src_simple_get, clk);
 }
+
+/*
+ * FIXME: All slow clk users are not properly claiming it (get + prepare +
+ * enable) before using it.
+ * If all users properly claiming this clock decide that they don't need it
+ * anymore (or are removed), it is disabled while faulty users are still
+ * requiring it, and the system hangs.
+ * Prevent this clock from being disabled until all users are properly
+ * requesting it.
+ * Once this is done we should remove this function and the slow_clk variable.
+ */
+static int __init of_at91_clk_slow_retain(void)
+{
+	if (!slow_clk)
+		return 0;
+
+	__clk_get(slow_clk);
+	clk_prepare_enable(slow_clk);
+
+	return 0;
+}
+arch_initcall(of_at91_clk_slow_retain);
diff --git a/drivers/clk/at91/clk-system.c b/drivers/clk/at91/clk-system.c
index 8c96307d7363..a76d03fd577b 100644
--- a/drivers/clk/at91/clk-system.c
+++ b/drivers/clk/at91/clk-system.c
@@ -119,13 +119,7 @@ at91_clk_register_system(struct at91_pmc *pmc, const char *name,
 	init.ops = &system_ops;
 	init.parent_names = &parent_name;
 	init.num_parents = 1;
-	/*
-	 * CLK_IGNORE_UNUSED is used to avoid ddrck switch off.
-	 * TODO : we should implement a driver supporting at91 ddr controller
-	 * (see drivers/memory) which would request and enable the ddrck clock.
-	 * When this is done we will be able to remove CLK_IGNORE_UNUSED flag.
-	 */
-	init.flags = CLK_SET_RATE_PARENT | CLK_IGNORE_UNUSED;
+	init.flags = CLK_SET_RATE_PARENT;
 
 	sys->id = id;
 	sys->hw.init = &init;
diff --git a/drivers/clk/at91/clk-usb.c b/drivers/clk/at91/clk-usb.c
index 7d1d26a4bd04..a23ac0c724f0 100644
--- a/drivers/clk/at91/clk-usb.c
+++ b/drivers/clk/at91/clk-usb.c
@@ -52,29 +52,26 @@ static unsigned long at91sam9x5_clk_usb_recalc_rate(struct clk_hw *hw,
 
 	tmp = pmc_read(pmc, AT91_PMC_USB);
 	usbdiv = (tmp & AT91_PMC_OHCIUSBDIV) >> SAM9X5_USB_DIV_SHIFT;
-	return parent_rate / (usbdiv + 1);
+
+	return DIV_ROUND_CLOSEST(parent_rate, (usbdiv + 1));
 }
 
 static long at91sam9x5_clk_usb_round_rate(struct clk_hw *hw, unsigned long rate,
 					  unsigned long *parent_rate)
 {
 	unsigned long div;
-	unsigned long bestrate;
-	unsigned long tmp;
+
+	if (!rate)
+		return -EINVAL;
 
 	if (rate >= *parent_rate)
 		return *parent_rate;
 
-	div = *parent_rate / rate;
-	if (div >= SAM9X5_USB_MAX_DIV)
-		return *parent_rate / (SAM9X5_USB_MAX_DIV + 1);
-
-	bestrate = *parent_rate / div;
-	tmp = *parent_rate / (div + 1);
-	if (bestrate - rate > rate - tmp)
-		bestrate = tmp;
+	div = DIV_ROUND_CLOSEST(*parent_rate, rate);
+	if (div > SAM9X5_USB_MAX_DIV + 1)
+		div = SAM9X5_USB_MAX_DIV + 1;
 
-	return bestrate;
+	return DIV_ROUND_CLOSEST(*parent_rate, div);
 }
 
 static int at91sam9x5_clk_usb_set_parent(struct clk_hw *hw, u8 index)
@@ -106,9 +103,13 @@ static int at91sam9x5_clk_usb_set_rate(struct clk_hw *hw, unsigned long rate,
 	u32 tmp;
 	struct at91sam9x5_clk_usb *usb = to_at91sam9x5_clk_usb(hw);
 	struct at91_pmc *pmc = usb->pmc;
-	unsigned long div = parent_rate / rate;
+	unsigned long div;
+
+	if (!rate)
+		return -EINVAL;
 
-	if (parent_rate % rate || div < 1 || div >= SAM9X5_USB_MAX_DIV)
+	div = DIV_ROUND_CLOSEST(parent_rate, rate);
+	if (div > SAM9X5_USB_MAX_DIV + 1 || !div)
 		return -EINVAL;
 
 	tmp = pmc_read(pmc, AT91_PMC_USB) & ~AT91_PMC_OHCIUSBDIV;
@@ -238,16 +239,22 @@ static long at91rm9200_clk_usb_round_rate(struct clk_hw *hw, unsigned long rate,
 					  unsigned long *parent_rate)
 {
 	struct at91rm9200_clk_usb *usb = to_at91rm9200_clk_usb(hw);
+	struct clk *parent = __clk_get_parent(hw->clk);
 	unsigned long bestrate = 0;
 	int bestdiff = -1;
 	unsigned long tmprate;
 	int tmpdiff;
 	int i = 0;
 
-	for (i = 0; i < 4; i++) {
+	for (i = 0; i < RM9200_USB_DIV_TAB_SIZE; i++) {
+		unsigned long tmp_parent_rate;
+
 		if (!usb->divisors[i])
 			continue;
-		tmprate = *parent_rate / usb->divisors[i];
+
+		tmp_parent_rate = rate * usb->divisors[i];
+		tmp_parent_rate = __clk_round_rate(parent, tmp_parent_rate);
+		tmprate = DIV_ROUND_CLOSEST(tmp_parent_rate, usb->divisors[i]);
 		if (tmprate < rate)
 			tmpdiff = rate - tmprate;
 		else
@@ -256,6 +263,7 @@ static long at91rm9200_clk_usb_round_rate(struct clk_hw *hw, unsigned long rate,
 		if (bestdiff < 0 || bestdiff > tmpdiff) {
 			bestrate = tmprate;
 			bestdiff = tmpdiff;
+			*parent_rate = tmp_parent_rate;
 		}
 
 		if (!bestdiff)
@@ -272,10 +280,13 @@ static int at91rm9200_clk_usb_set_rate(struct clk_hw *hw, unsigned long rate,
 	int i;
 	struct at91rm9200_clk_usb *usb = to_at91rm9200_clk_usb(hw);
 	struct at91_pmc *pmc = usb->pmc;
-	unsigned long div = parent_rate / rate;
+	unsigned long div;
 
-	if (parent_rate % rate)
+	if (!rate)
 		return -EINVAL;
+
+	div = DIV_ROUND_CLOSEST(parent_rate, rate);
+
 	for (i = 0; i < RM9200_USB_DIV_TAB_SIZE; i++) {
 		if (usb->divisors[i] == div) {
 			tmp = pmc_read(pmc, AT91_CKGR_PLLBR) &
@@ -311,7 +322,7 @@ at91rm9200_clk_register_usb(struct at91_pmc *pmc, const char *name,
 	init.ops = &at91rm9200_usb_ops;
 	init.parent_names = &parent_name;
 	init.num_parents = 1;
-	init.flags = 0;
+	init.flags = CLK_SET_RATE_PARENT;
 
 	usb->hw.init = &init;
 	usb->pmc = pmc;
diff --git a/drivers/clk/at91/pmc.c b/drivers/clk/at91/pmc.c
index 524196bb35a5..386999b4f8eb 100644
--- a/drivers/clk/at91/pmc.c
+++ b/drivers/clk/at91/pmc.c
@@ -337,6 +337,12 @@ static const struct of_device_id pmc_clk_ids[] __initconst = {
 		.data = of_at91sam9x5_clk_smd_setup,
 	},
 #endif
+#if defined(CONFIG_HAVE_AT91_H32MX)
+	{
+		.compatible = "atmel,sama5d4-clk-h32mx",
+		.data = of_sama5d4_clk_h32mx_setup,
+	},
+#endif
 	{ /*sentinel*/ }
 };
 
diff --git a/drivers/clk/at91/pmc.h b/drivers/clk/at91/pmc.h
index 6c7625976113..52d2041fa3f6 100644
--- a/drivers/clk/at91/pmc.h
+++ b/drivers/clk/at91/pmc.h
@@ -120,4 +120,9 @@ extern void __init of_at91sam9x5_clk_smd_setup(struct device_node *np,
 					       struct at91_pmc *pmc);
 #endif
 
+#if defined(CONFIG_HAVE_AT91_SMD)
+extern void __init of_sama5d4_clk_h32mx_setup(struct device_node *np,
+					      struct at91_pmc *pmc);
+#endif
+
 #endif /* __PMC_H_ */
diff --git a/drivers/clk/bcm/clk-kona.c b/drivers/clk/bcm/clk-kona.c
index 95af2e665dd3..1c06f6f3a8c5 100644
--- a/drivers/clk/bcm/clk-kona.c
+++ b/drivers/clk/bcm/clk-kona.c
@@ -1032,7 +1032,7 @@ static long kona_peri_clk_round_rate(struct clk_hw *hw, unsigned long rate,
 }
 
 static long kona_peri_clk_determine_rate(struct clk_hw *hw, unsigned long rate,
-		unsigned long *best_parent_rate, struct clk **best_parent)
+		unsigned long *best_parent_rate, struct clk_hw **best_parent)
 {
 	struct kona_clk *bcm_clk = to_kona_clk(hw);
 	struct clk *clk = hw->clk;
@@ -1075,7 +1075,7 @@ static long kona_peri_clk_determine_rate(struct clk_hw *hw, unsigned long rate,
 		if (delta < best_delta) {
 			best_delta = delta;
 			best_rate = other_rate;
-			*best_parent = parent;
+			*best_parent = __clk_get_hw(parent);
 			*best_parent_rate = parent_rate;
 		}
 	}
diff --git a/drivers/clk/berlin/bg2q.c b/drivers/clk/berlin/bg2q.c
index 21784e4eb3f0..440ef81ab15c 100644
--- a/drivers/clk/berlin/bg2q.c
+++ b/drivers/clk/berlin/bg2q.c
@@ -285,7 +285,6 @@ static const struct berlin2_gate_data bg2q_gates[] __initconst = {
 	{ "pbridge",	"perif",	15, CLK_IGNORE_UNUSED },
 	{ "sdio",	"perif",	16, CLK_IGNORE_UNUSED },
 	{ "nfc",	"perif",	18 },
-	{ "smemc",	"perif",	19 },
 	{ "pcie",	"perif",	22 },
 };
 
diff --git a/drivers/clk/clk-axi-clkgen.c b/drivers/clk/clk-axi-clkgen.c
index 1127ee46b802..e619285c6def 100644
--- a/drivers/clk/clk-axi-clkgen.c
+++ b/drivers/clk/clk-axi-clkgen.c
@@ -544,7 +544,6 @@ static int axi_clkgen_remove(struct platform_device *pdev)
 static struct platform_driver axi_clkgen_driver = {
 	.driver = {
 		.name = "adi-axi-clkgen",
-		.owner = THIS_MODULE,
 		.of_match_table = axi_clkgen_ids,
 	},
 	.probe = axi_clkgen_probe,
diff --git a/drivers/clk/clk-axm5516.c b/drivers/clk/clk-axm5516.c
index d2f1e119b450..0f6368ceec4c 100644
--- a/drivers/clk/clk-axm5516.c
+++ b/drivers/clk/clk-axm5516.c
@@ -593,7 +593,6 @@ static struct platform_driver axmclk_driver = {
 	.remove		= axmclk_remove,
 	.driver		= {
 		.name	= "clk-axm5516",
-		.owner	= THIS_MODULE,
 		.of_match_table = axmclk_match_table,
 	},
 };
diff --git a/drivers/clk/clk-composite.c b/drivers/clk/clk-composite.c
index b9355daf8065..4386697236a7 100644
--- a/drivers/clk/clk-composite.c
+++ b/drivers/clk/clk-composite.c
@@ -57,7 +57,7 @@ static unsigned long clk_composite_recalc_rate(struct clk_hw *hw,
 
 static long clk_composite_determine_rate(struct clk_hw *hw, unsigned long rate,
 					unsigned long *best_parent_rate,
-					struct clk **best_parent_p)
+					struct clk_hw **best_parent_p)
 {
 	struct clk_composite *composite = to_clk_composite(hw);
 	const struct clk_ops *rate_ops = composite->rate_ops;
@@ -80,8 +80,9 @@ static long clk_composite_determine_rate(struct clk_hw *hw, unsigned long rate,
 		*best_parent_p = NULL;
 
 		if (__clk_get_flags(hw->clk) & CLK_SET_RATE_NO_REPARENT) {
-			*best_parent_p = clk_get_parent(mux_hw->clk);
-			*best_parent_rate = __clk_get_rate(*best_parent_p);
+			parent = clk_get_parent(mux_hw->clk);
+			*best_parent_p = __clk_get_hw(parent);
+			*best_parent_rate = __clk_get_rate(parent);
 
 			return rate_ops->round_rate(rate_hw, rate,
 						    best_parent_rate);
@@ -103,7 +104,7 @@ static long clk_composite_determine_rate(struct clk_hw *hw, unsigned long rate,
 
 			if (!rate_diff || !*best_parent_p
 				       || best_rate_diff > rate_diff) {
-				*best_parent_p = parent;
+				*best_parent_p = __clk_get_hw(parent);
 				*best_parent_rate = parent_rate;
 				best_rate_diff = rate_diff;
 				best_rate = tmp_rate;
diff --git a/drivers/clk/clk-divider.c b/drivers/clk/clk-divider.c
index 18a9de29df0e..c0a842b335c5 100644
--- a/drivers/clk/clk-divider.c
+++ b/drivers/clk/clk-divider.c
@@ -263,6 +263,14 @@ static int clk_divider_bestdiv(struct clk_hw *hw, unsigned long rate,
 	if (!rate)
 		rate = 1;
 
+	/* if read only, just return current value */
+	if (divider->flags & CLK_DIVIDER_READ_ONLY) {
+		bestdiv = readl(divider->reg) >> divider->shift;
+		bestdiv &= div_mask(divider);
+		bestdiv = _get_div(divider, bestdiv);
+		return bestdiv;
+	}
+
 	maxdiv = _get_maxdiv(divider);
 
 	if (!(__clk_get_flags(hw->clk) & CLK_SET_RATE_PARENT)) {
@@ -361,11 +369,6 @@ const struct clk_ops clk_divider_ops = {
 };
 EXPORT_SYMBOL_GPL(clk_divider_ops);
 
-const struct clk_ops clk_divider_ro_ops = {
-	.recalc_rate = clk_divider_recalc_rate,
-};
-EXPORT_SYMBOL_GPL(clk_divider_ro_ops);
-
 static struct clk *_register_divider(struct device *dev, const char *name,
 		const char *parent_name, unsigned long flags,
 		void __iomem *reg, u8 shift, u8 width,
@@ -391,10 +394,7 @@ static struct clk *_register_divider(struct device *dev, const char *name,
 	}
 
 	init.name = name;
-	if (clk_divider_flags & CLK_DIVIDER_READ_ONLY)
-		init.ops = &clk_divider_ro_ops;
-	else
-		init.ops = &clk_divider_ops;
+	init.ops = &clk_divider_ops;
 	init.flags = flags | CLK_IS_BASIC;
 	init.parent_names = (parent_name ? &parent_name: NULL);
 	init.num_parents = (parent_name ? 1 : 0);
diff --git a/drivers/clk/clk-fractional-divider.c b/drivers/clk/clk-fractional-divider.c
index ede685ca0d20..82a59d0086cc 100644
--- a/drivers/clk/clk-fractional-divider.c
+++ b/drivers/clk/clk-fractional-divider.c
@@ -36,7 +36,7 @@ static unsigned long clk_fd_recalc_rate(struct clk_hw *hw,
 	m = (val & fd->mmask) >> fd->mshift;
 	n = (val & fd->nmask) >> fd->nshift;
 
-	ret = parent_rate * m;
+	ret = (u64)parent_rate * m;
 	do_div(ret, n);
 
 	return ret;
diff --git a/drivers/clk/clk-gate.c b/drivers/clk/clk-gate.c
index 4a58c55255bd..51fd87fb7ba6 100644
--- a/drivers/clk/clk-gate.c
+++ b/drivers/clk/clk-gate.c
@@ -45,7 +45,7 @@ static void clk_gate_endisable(struct clk_hw *hw, int enable)
 {
 	struct clk_gate *gate = to_clk_gate(hw);
 	int set = gate->flags & CLK_GATE_SET_TO_DISABLE ? 1 : 0;
-	unsigned long flags = 0;
+	unsigned long uninitialized_var(flags);
 	u32 reg;
 
 	set ^= enable;
diff --git a/drivers/clk/clk-gpio-gate.c b/drivers/clk/clk-gpio-gate.c
new file mode 100644
index 000000000000..08e43224fd52
--- /dev/null
+++ b/drivers/clk/clk-gpio-gate.c
@@ -0,0 +1,205 @@
+/*
+ * Copyright (C) 2013 - 2014 Texas Instruments Incorporated - http://www.ti.com
+ * Author: Jyri Sarha <jsarha@ti.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * Gpio gated clock implementation
+ */
+
+#include <linux/clk-provider.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/gpio.h>
+#include <linux/gpio/consumer.h>
+#include <linux/of_gpio.h>
+#include <linux/err.h>
+#include <linux/device.h>
+
+/**
+ * DOC: basic gpio gated clock which can be enabled and disabled
+ *      with gpio output
+ * Traits of this clock:
+ * prepare - clk_(un)prepare only ensures parent is (un)prepared
+ * enable - clk_enable and clk_disable are functional & control gpio
+ * rate - inherits rate from parent.  No clk_set_rate support
+ * parent - fixed parent.  No clk_set_parent support
+ */
+
+#define to_clk_gpio(_hw) container_of(_hw, struct clk_gpio, hw)
+
+static int clk_gpio_gate_enable(struct clk_hw *hw)
+{
+	struct clk_gpio *clk = to_clk_gpio(hw);
+
+	gpiod_set_value(clk->gpiod, 1);
+
+	return 0;
+}
+
+static void clk_gpio_gate_disable(struct clk_hw *hw)
+{
+	struct clk_gpio *clk = to_clk_gpio(hw);
+
+	gpiod_set_value(clk->gpiod, 0);
+}
+
+static int clk_gpio_gate_is_enabled(struct clk_hw *hw)
+{
+	struct clk_gpio *clk = to_clk_gpio(hw);
+
+	return gpiod_get_value(clk->gpiod);
+}
+
+const struct clk_ops clk_gpio_gate_ops = {
+	.enable = clk_gpio_gate_enable,
+	.disable = clk_gpio_gate_disable,
+	.is_enabled = clk_gpio_gate_is_enabled,
+};
+EXPORT_SYMBOL_GPL(clk_gpio_gate_ops);
+
+/**
+ * clk_register_gpio - register a gpip clock with the clock framework
+ * @dev: device that is registering this clock
+ * @name: name of this clock
+ * @parent_name: name of this clock's parent
+ * @gpiod: gpio descriptor to gate this clock
+ */
+struct clk *clk_register_gpio_gate(struct device *dev, const char *name,
+		const char *parent_name, struct gpio_desc *gpiod,
+		unsigned long flags)
+{
+	struct clk_gpio *clk_gpio = NULL;
+	struct clk *clk = ERR_PTR(-EINVAL);
+	struct clk_init_data init = { NULL };
+	unsigned long gpio_flags;
+	int err;
+
+	if (gpiod_is_active_low(gpiod))
+		gpio_flags = GPIOF_OUT_INIT_HIGH;
+	else
+		gpio_flags = GPIOF_OUT_INIT_LOW;
+
+	if (dev)
+		err = devm_gpio_request_one(dev, desc_to_gpio(gpiod),
+					    gpio_flags, name);
+	else
+		err = gpio_request_one(desc_to_gpio(gpiod), gpio_flags, name);
+
+	if (err) {
+		pr_err("%s: %s: Error requesting clock control gpio %u\n",
+		       __func__, name, desc_to_gpio(gpiod));
+		return ERR_PTR(err);
+	}
+
+	if (dev)
+		clk_gpio = devm_kzalloc(dev, sizeof(struct clk_gpio),
+					GFP_KERNEL);
+	else
+		clk_gpio = kzalloc(sizeof(struct clk_gpio), GFP_KERNEL);
+
+	if (!clk_gpio) {
+		clk = ERR_PTR(-ENOMEM);
+		goto clk_register_gpio_gate_err;
+	}
+
+	init.name = name;
+	init.ops = &clk_gpio_gate_ops;
+	init.flags = flags | CLK_IS_BASIC;
+	init.parent_names = (parent_name ? &parent_name : NULL);
+	init.num_parents = (parent_name ? 1 : 0);
+
+	clk_gpio->gpiod = gpiod;
+	clk_gpio->hw.init = &init;
+
+	clk = clk_register(dev, &clk_gpio->hw);
+
+	if (!IS_ERR(clk))
+		return clk;
+
+	if (!dev)
+		kfree(clk_gpio);
+
+clk_register_gpio_gate_err:
+	gpiod_put(gpiod);
+
+	return clk;
+}
+EXPORT_SYMBOL_GPL(clk_register_gpio_gate);
+
+#ifdef CONFIG_OF
+/**
+ * The clk_register_gpio_gate has to be delayed, because the EPROBE_DEFER
+ * can not be handled properly at of_clk_init() call time.
+ */
+
+struct clk_gpio_gate_delayed_register_data {
+	struct device_node *node;
+	struct mutex lock;
+	struct clk *clk;
+};
+
+static struct clk *of_clk_gpio_gate_delayed_register_get(
+		struct of_phandle_args *clkspec,
+		void *_data)
+{
+	struct clk_gpio_gate_delayed_register_data *data = _data;
+	struct clk *clk;
+	const char *clk_name = data->node->name;
+	const char *parent_name;
+	struct gpio_desc *gpiod;
+	int gpio;
+
+	mutex_lock(&data->lock);
+
+	if (data->clk) {
+		mutex_unlock(&data->lock);
+		return data->clk;
+	}
+
+	gpio = of_get_named_gpio_flags(data->node, "enable-gpios", 0, NULL);
+	if (gpio < 0) {
+		mutex_unlock(&data->lock);
+		if (gpio != -EPROBE_DEFER)
+			pr_err("%s: %s: Can't get 'enable-gpios' DT property\n",
+			       __func__, clk_name);
+		return ERR_PTR(gpio);
+	}
+	gpiod = gpio_to_desc(gpio);
+
+	parent_name = of_clk_get_parent_name(data->node, 0);
+
+	clk = clk_register_gpio_gate(NULL, clk_name, parent_name, gpiod, 0);
+	if (IS_ERR(clk)) {
+		mutex_unlock(&data->lock);
+		return clk;
+	}
+
+	data->clk = clk;
+	mutex_unlock(&data->lock);
+
+	return clk;
+}
+
+/**
+ * of_gpio_gate_clk_setup() - Setup function for gpio controlled clock
+ */
+void __init of_gpio_gate_clk_setup(struct device_node *node)
+{
+	struct clk_gpio_gate_delayed_register_data *data;
+
+	data = kzalloc(sizeof(struct clk_gpio_gate_delayed_register_data),
+		       GFP_KERNEL);
+	if (!data)
+		return;
+
+	data->node = node;
+	mutex_init(&data->lock);
+
+	of_clk_add_provider(node, of_clk_gpio_gate_delayed_register_get, data);
+}
+EXPORT_SYMBOL_GPL(of_gpio_gate_clk_setup);
+CLK_OF_DECLARE(gpio_gate_clk, "gpio-gate-clock", of_gpio_gate_clk_setup);
+#endif
diff --git a/drivers/clk/clk-ls1x.c b/drivers/clk/clk-ls1x.c
index f20b750235f6..ca80103ac188 100644
--- a/drivers/clk/clk-ls1x.c
+++ b/drivers/clk/clk-ls1x.c
@@ -15,7 +15,8 @@
 
 #include <loongson1.h>
 
-#define OSC	33
+#define OSC		(33 * 1000000)
+#define DIV_APB		2
 
 static DEFINE_SPINLOCK(_lock);
 
@@ -29,13 +30,12 @@ static void ls1x_pll_clk_disable(struct clk_hw *hw)
 }
 
 static unsigned long ls1x_pll_recalc_rate(struct clk_hw *hw,
-					     unsigned long parent_rate)
+					  unsigned long parent_rate)
 {
 	u32 pll, rate;
 
 	pll = __raw_readl(LS1X_CLK_PLL_FREQ);
-	rate = ((12 + (pll & 0x3f)) * 1000000) +
-		((((pll >> 8) & 0x3ff) * 1000000) >> 10);
+	rate = 12 + (pll & 0x3f) + (((pll >> 8) & 0x3ff) >> 10);
 	rate *= OSC;
 	rate >>= 1;
 
@@ -48,8 +48,10 @@ static const struct clk_ops ls1x_pll_clk_ops = {
 	.recalc_rate = ls1x_pll_recalc_rate,
 };
 
-static struct clk * __init clk_register_pll(struct device *dev,
-	 const char *name, const char *parent_name, unsigned long flags)
+static struct clk *__init clk_register_pll(struct device *dev,
+					   const char *name,
+					   const char *parent_name,
+					   unsigned long flags)
 {
 	struct clk_hw *hw;
 	struct clk *clk;
@@ -78,34 +80,83 @@ static struct clk * __init clk_register_pll(struct device *dev,
 	return clk;
 }
 
+static const char const *cpu_parents[] = { "cpu_clk_div", "osc_33m_clk", };
+static const char const *ahb_parents[] = { "ahb_clk_div", "osc_33m_clk", };
+static const char const *dc_parents[] = { "dc_clk_div", "osc_33m_clk", };
+
 void __init ls1x_clk_init(void)
 {
 	struct clk *clk;
 
-	clk = clk_register_pll(NULL, "pll_clk", NULL, CLK_IS_ROOT);
-	clk_prepare_enable(clk);
-
-	clk = clk_register_divider(NULL, "cpu_clk", "pll_clk",
-			CLK_SET_RATE_PARENT, LS1X_CLK_PLL_DIV, DIV_CPU_SHIFT,
-			DIV_CPU_WIDTH, CLK_DIVIDER_ONE_BASED, &_lock);
-	clk_prepare_enable(clk);
-	clk_register_clkdev(clk, "cpu", NULL);
-
-	clk = clk_register_divider(NULL, "dc_clk", "pll_clk",
-			CLK_SET_RATE_PARENT, LS1X_CLK_PLL_DIV, DIV_DC_SHIFT,
-			DIV_DC_WIDTH, CLK_DIVIDER_ONE_BASED, &_lock);
-	clk_prepare_enable(clk);
-	clk_register_clkdev(clk, "dc", NULL);
-
-	clk = clk_register_divider(NULL, "ahb_clk", "pll_clk",
-			CLK_SET_RATE_PARENT, LS1X_CLK_PLL_DIV, DIV_DDR_SHIFT,
-			DIV_DDR_WIDTH, CLK_DIVIDER_ONE_BASED, &_lock);
-	clk_prepare_enable(clk);
-	clk_register_clkdev(clk, "ahb", NULL);
+	clk = clk_register_fixed_rate(NULL, "osc_33m_clk", NULL, CLK_IS_ROOT,
+				      OSC);
+	clk_register_clkdev(clk, "osc_33m_clk", NULL);
+
+	/* clock derived from 33 MHz OSC clk */
+	clk = clk_register_pll(NULL, "pll_clk", "osc_33m_clk", 0);
+	clk_register_clkdev(clk, "pll_clk", NULL);
+
+	/* clock derived from PLL clk */
+	/*                                 _____
+	 *         _______________________|     |
+	 * OSC ___/                       | MUX |___ CPU CLK
+	 *        \___ PLL ___ CPU DIV ___|     |
+	 *                                |_____|
+	 */
+	clk = clk_register_divider(NULL, "cpu_clk_div", "pll_clk",
+				   CLK_GET_RATE_NOCACHE, LS1X_CLK_PLL_DIV,
+				   DIV_CPU_SHIFT, DIV_CPU_WIDTH,
+				   CLK_DIVIDER_ONE_BASED |
+				   CLK_DIVIDER_ROUND_CLOSEST, &_lock);
+	clk_register_clkdev(clk, "cpu_clk_div", NULL);
+	clk = clk_register_mux(NULL, "cpu_clk", cpu_parents,
+			       ARRAY_SIZE(cpu_parents),
+			       CLK_SET_RATE_NO_REPARENT, LS1X_CLK_PLL_DIV,
+			       BYPASS_CPU_SHIFT, BYPASS_CPU_WIDTH, 0, &_lock);
+	clk_register_clkdev(clk, "cpu_clk", NULL);
+
+	/*                                 _____
+	 *         _______________________|     |
+	 * OSC ___/                       | MUX |___ DC  CLK
+	 *        \___ PLL ___ DC  DIV ___|     |
+	 *                                |_____|
+	 */
+	clk = clk_register_divider(NULL, "dc_clk_div", "pll_clk",
+				   0, LS1X_CLK_PLL_DIV, DIV_DC_SHIFT,
+				   DIV_DC_WIDTH, CLK_DIVIDER_ONE_BASED, &_lock);
+	clk_register_clkdev(clk, "dc_clk_div", NULL);
+	clk = clk_register_mux(NULL, "dc_clk", dc_parents,
+			       ARRAY_SIZE(dc_parents),
+			       CLK_SET_RATE_NO_REPARENT, LS1X_CLK_PLL_DIV,
+			       BYPASS_DC_SHIFT, BYPASS_DC_WIDTH, 0, &_lock);
+	clk_register_clkdev(clk, "dc_clk", NULL);
+
+	/*                                 _____
+	 *         _______________________|     |
+	 * OSC ___/                       | MUX |___ DDR CLK
+	 *        \___ PLL ___ DDR DIV ___|     |
+	 *                                |_____|
+	 */
+	clk = clk_register_divider(NULL, "ahb_clk_div", "pll_clk",
+				   0, LS1X_CLK_PLL_DIV, DIV_DDR_SHIFT,
+				   DIV_DDR_WIDTH, CLK_DIVIDER_ONE_BASED,
+				   &_lock);
+	clk_register_clkdev(clk, "ahb_clk_div", NULL);
+	clk = clk_register_mux(NULL, "ahb_clk", ahb_parents,
+			       ARRAY_SIZE(ahb_parents),
+			       CLK_SET_RATE_NO_REPARENT, LS1X_CLK_PLL_DIV,
+			       BYPASS_DDR_SHIFT, BYPASS_DDR_WIDTH, 0, &_lock);
+	clk_register_clkdev(clk, "ahb_clk", NULL);
 	clk_register_clkdev(clk, "stmmaceth", NULL);
 
-	clk = clk_register_fixed_factor(NULL, "apb_clk", "ahb_clk", 0, 1, 2);
-	clk_prepare_enable(clk);
-	clk_register_clkdev(clk, "apb", NULL);
+	/* clock derived from AHB clk */
+	/* APB clk is always half of the AHB clk */
+	clk = clk_register_fixed_factor(NULL, "apb_clk", "ahb_clk", 0, 1,
+					DIV_APB);
+	clk_register_clkdev(clk, "apb_clk", NULL);
+	clk_register_clkdev(clk, "ls1x_i2c", NULL);
+	clk_register_clkdev(clk, "ls1x_pwmtimer", NULL);
+	clk_register_clkdev(clk, "ls1x_spi", NULL);
+	clk_register_clkdev(clk, "ls1x_wdt", NULL);
 	clk_register_clkdev(clk, "serial8250", NULL);
 }
diff --git a/drivers/clk/clk-max-gen.c b/drivers/clk/clk-max-gen.c
new file mode 100644
index 000000000000..6505049d50f1
--- /dev/null
+++ b/drivers/clk/clk-max-gen.c
@@ -0,0 +1,192 @@
+/*
+ * clk-max-gen.c - Generic clock driver for Maxim PMICs clocks
+ *
+ * Copyright (C) 2014 Google, Inc
+ *
+ * Copyright (C) 2012 Samsung Electornics
+ * Jonghwa Lee <jonghwa3.lee@samsung.com>
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * This driver is based on clk-max77686.c
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/err.h>
+#include <linux/regmap.h>
+#include <linux/platform_device.h>
+#include <linux/clk-provider.h>
+#include <linux/mutex.h>
+#include <linux/clkdev.h>
+#include <linux/of.h>
+#include <linux/export.h>
+
+struct max_gen_clk {
+	struct regmap *regmap;
+	u32 mask;
+	u32 reg;
+	struct clk_hw hw;
+};
+
+static struct max_gen_clk *to_max_gen_clk(struct clk_hw *hw)
+{
+	return container_of(hw, struct max_gen_clk, hw);
+}
+
+static int max_gen_clk_prepare(struct clk_hw *hw)
+{
+	struct max_gen_clk *max_gen = to_max_gen_clk(hw);
+
+	return regmap_update_bits(max_gen->regmap, max_gen->reg,
+				  max_gen->mask, max_gen->mask);
+}
+
+static void max_gen_clk_unprepare(struct clk_hw *hw)
+{
+	struct max_gen_clk *max_gen = to_max_gen_clk(hw);
+
+	regmap_update_bits(max_gen->regmap, max_gen->reg,
+			   max_gen->mask, ~max_gen->mask);
+}
+
+static int max_gen_clk_is_prepared(struct clk_hw *hw)
+{
+	struct max_gen_clk *max_gen = to_max_gen_clk(hw);
+	int ret;
+	u32 val;
+
+	ret = regmap_read(max_gen->regmap, max_gen->reg, &val);
+
+	if (ret < 0)
+		return -EINVAL;
+
+	return val & max_gen->mask;
+}
+
+static unsigned long max_gen_recalc_rate(struct clk_hw *hw,
+					 unsigned long parent_rate)
+{
+	return 32768;
+}
+
+struct clk_ops max_gen_clk_ops = {
+	.prepare	= max_gen_clk_prepare,
+	.unprepare	= max_gen_clk_unprepare,
+	.is_prepared	= max_gen_clk_is_prepared,
+	.recalc_rate	= max_gen_recalc_rate,
+};
+EXPORT_SYMBOL_GPL(max_gen_clk_ops);
+
+static struct clk *max_gen_clk_register(struct device *dev,
+					struct max_gen_clk *max_gen)
+{
+	struct clk *clk;
+	struct clk_hw *hw = &max_gen->hw;
+	int ret;
+
+	clk = devm_clk_register(dev, hw);
+	if (IS_ERR(clk))
+		return clk;
+
+	ret = clk_register_clkdev(clk, hw->init->name, NULL);
+
+	if (ret)
+		return ERR_PTR(ret);
+
+	return clk;
+}
+
+int max_gen_clk_probe(struct platform_device *pdev, struct regmap *regmap,
+		      u32 reg, struct clk_init_data *clks_init, int num_init)
+{
+	int i, ret;
+	struct max_gen_clk *max_gen_clks;
+	struct clk **clocks;
+	struct device *dev = pdev->dev.parent;
+	const char *clk_name;
+	struct clk_init_data *init;
+
+	clocks = devm_kzalloc(dev, sizeof(struct clk *) * num_init, GFP_KERNEL);
+	if (!clocks)
+		return -ENOMEM;
+
+	max_gen_clks = devm_kzalloc(dev, sizeof(struct max_gen_clk)
+				    * num_init, GFP_KERNEL);
+	if (!max_gen_clks)
+		return -ENOMEM;
+
+	for (i = 0; i < num_init; i++) {
+		max_gen_clks[i].regmap = regmap;
+		max_gen_clks[i].mask = 1 << i;
+		max_gen_clks[i].reg = reg;
+
+		init = devm_kzalloc(dev, sizeof(*init), GFP_KERNEL);
+		if (!init)
+			return -ENOMEM;
+
+		if (dev->of_node &&
+		    !of_property_read_string_index(dev->of_node,
+						   "clock-output-names",
+						   i, &clk_name))
+			init->name = clk_name;
+		else
+			init->name = clks_init[i].name;
+
+		init->ops = clks_init[i].ops;
+		init->flags = clks_init[i].flags;
+
+		max_gen_clks[i].hw.init = init;
+
+		clocks[i] = max_gen_clk_register(dev, &max_gen_clks[i]);
+		if (IS_ERR(clocks[i])) {
+			ret = PTR_ERR(clocks[i]);
+			dev_err(dev, "failed to register %s\n",
+				max_gen_clks[i].hw.init->name);
+			return ret;
+		}
+	}
+
+	platform_set_drvdata(pdev, clocks);
+
+	if (dev->of_node) {
+		struct clk_onecell_data *of_data;
+
+		of_data = devm_kzalloc(dev, sizeof(*of_data), GFP_KERNEL);
+		if (!of_data)
+			return -ENOMEM;
+
+		of_data->clks = clocks;
+		of_data->clk_num = num_init;
+		ret = of_clk_add_provider(dev->of_node, of_clk_src_onecell_get,
+					  of_data);
+
+		if (ret) {
+			dev_err(dev, "failed to register OF clock provider\n");
+			return ret;
+		}
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(max_gen_clk_probe);
+
+int max_gen_clk_remove(struct platform_device *pdev, int num_init)
+{
+	struct device *dev = pdev->dev.parent;
+
+	if (dev->of_node)
+		of_clk_del_provider(dev->of_node);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(max_gen_clk_remove);
diff --git a/drivers/clk/clk-max-gen.h b/drivers/clk/clk-max-gen.h
new file mode 100644
index 000000000000..997e86fc3f4d
--- /dev/null
+++ b/drivers/clk/clk-max-gen.h
@@ -0,0 +1,32 @@
+/*
+ * clk-max-gen.h - Generic clock driver for Maxim PMICs clocks
+ *
+ * Copyright (C) 2014 Google, Inc
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ */
+
+#ifndef __CLK_MAX_GEN_H__
+#define __CLK_MAX_GEN_H__
+
+#include <linux/types.h>
+#include <linux/device.h>
+#include <linux/clkdev.h>
+#include <linux/regmap.h>
+#include <linux/platform_device.h>
+
+int max_gen_clk_probe(struct platform_device *pdev, struct regmap *regmap,
+		      u32 reg, struct clk_init_data *clks_init, int num_init);
+int max_gen_clk_remove(struct platform_device *pdev, int num_init);
+extern struct clk_ops max_gen_clk_ops;
+
+#endif /* __CLK_MAX_GEN_H__ */
diff --git a/drivers/clk/clk-max77686.c b/drivers/clk/clk-max77686.c
index 3d7e8dd8fd58..86cdb3a28629 100644
--- a/drivers/clk/clk-max77686.c
+++ b/drivers/clk/clk-max77686.c
@@ -30,193 +30,38 @@
 #include <linux/mutex.h>
 #include <linux/clkdev.h>
 
-enum {
-	MAX77686_CLK_AP = 0,
-	MAX77686_CLK_CP,
-	MAX77686_CLK_PMIC,
-	MAX77686_CLKS_NUM,
-};
-
-struct max77686_clk {
-	struct max77686_dev *iodev;
-	u32 mask;
-	struct clk_hw hw;
-	struct clk_lookup *lookup;
-};
-
-static struct max77686_clk *to_max77686_clk(struct clk_hw *hw)
-{
-	return container_of(hw, struct max77686_clk, hw);
-}
-
-static int max77686_clk_prepare(struct clk_hw *hw)
-{
-	struct max77686_clk *max77686 = to_max77686_clk(hw);
-
-	return regmap_update_bits(max77686->iodev->regmap,
-				  MAX77686_REG_32KHZ, max77686->mask,
-				  max77686->mask);
-}
-
-static void max77686_clk_unprepare(struct clk_hw *hw)
-{
-	struct max77686_clk *max77686 = to_max77686_clk(hw);
-
-	regmap_update_bits(max77686->iodev->regmap,
-		MAX77686_REG_32KHZ, max77686->mask, ~max77686->mask);
-}
-
-static int max77686_clk_is_prepared(struct clk_hw *hw)
-{
-	struct max77686_clk *max77686 = to_max77686_clk(hw);
-	int ret;
-	u32 val;
-
-	ret = regmap_read(max77686->iodev->regmap,
-				MAX77686_REG_32KHZ, &val);
-
-	if (ret < 0)
-		return -EINVAL;
-
-	return val & max77686->mask;
-}
-
-static unsigned long max77686_recalc_rate(struct clk_hw *hw,
-					  unsigned long parent_rate)
-{
-	return 32768;
-}
-
-static struct clk_ops max77686_clk_ops = {
-	.prepare	= max77686_clk_prepare,
-	.unprepare	= max77686_clk_unprepare,
-	.is_prepared	= max77686_clk_is_prepared,
-	.recalc_rate	= max77686_recalc_rate,
-};
+#include <dt-bindings/clock/maxim,max77686.h>
+#include "clk-max-gen.h"
 
 static struct clk_init_data max77686_clks_init[MAX77686_CLKS_NUM] = {
 	[MAX77686_CLK_AP] = {
 		.name = "32khz_ap",
-		.ops = &max77686_clk_ops,
+		.ops = &max_gen_clk_ops,
 		.flags = CLK_IS_ROOT,
 	},
 	[MAX77686_CLK_CP] = {
 		.name = "32khz_cp",
-		.ops = &max77686_clk_ops,
+		.ops = &max_gen_clk_ops,
 		.flags = CLK_IS_ROOT,
 	},
 	[MAX77686_CLK_PMIC] = {
 		.name = "32khz_pmic",
-		.ops = &max77686_clk_ops,
+		.ops = &max_gen_clk_ops,
 		.flags = CLK_IS_ROOT,
 	},
 };
 
-static struct clk *max77686_clk_register(struct device *dev,
-				struct max77686_clk *max77686)
-{
-	struct clk *clk;
-	struct clk_hw *hw = &max77686->hw;
-
-	clk = clk_register(dev, hw);
-	if (IS_ERR(clk))
-		return clk;
-
-	max77686->lookup = kzalloc(sizeof(struct clk_lookup), GFP_KERNEL);
-	if (!max77686->lookup)
-		return ERR_PTR(-ENOMEM);
-
-	max77686->lookup->con_id = hw->init->name;
-	max77686->lookup->clk = clk;
-
-	clkdev_add(max77686->lookup);
-
-	return clk;
-}
-
 static int max77686_clk_probe(struct platform_device *pdev)
 {
 	struct max77686_dev *iodev = dev_get_drvdata(pdev->dev.parent);
-	struct max77686_clk *max77686_clks[MAX77686_CLKS_NUM];
-	struct clk **clocks;
-	int i, ret;
-
-	clocks = devm_kzalloc(&pdev->dev, sizeof(struct clk *)
-					* MAX77686_CLKS_NUM, GFP_KERNEL);
-	if (!clocks)
-		return -ENOMEM;
-
-	for (i = 0; i < MAX77686_CLKS_NUM; i++) {
-		max77686_clks[i] = devm_kzalloc(&pdev->dev,
-					sizeof(struct max77686_clk), GFP_KERNEL);
-		if (!max77686_clks[i])
-			return -ENOMEM;
-	}
-
-	for (i = 0; i < MAX77686_CLKS_NUM; i++) {
-		max77686_clks[i]->iodev = iodev;
-		max77686_clks[i]->mask = 1 << i;
-		max77686_clks[i]->hw.init = &max77686_clks_init[i];
-
-		clocks[i] = max77686_clk_register(&pdev->dev, max77686_clks[i]);
-		if (IS_ERR(clocks[i])) {
-			ret = PTR_ERR(clocks[i]);
-			dev_err(&pdev->dev, "failed to register %s\n",
-				max77686_clks[i]->hw.init->name);
-			goto err_clocks;
-		}
-	}
-
-	platform_set_drvdata(pdev, clocks);
-
-	if (iodev->dev->of_node) {
-		struct clk_onecell_data *of_data;
 
-		of_data = devm_kzalloc(&pdev->dev,
-					sizeof(*of_data), GFP_KERNEL);
-		if (!of_data) {
-			ret = -ENOMEM;
-			goto err_clocks;
-		}
-
-		of_data->clks = clocks;
-		of_data->clk_num = MAX77686_CLKS_NUM;
-		ret = of_clk_add_provider(iodev->dev->of_node,
-					of_clk_src_onecell_get, of_data);
-		if (ret) {
-			dev_err(&pdev->dev, "failed to register OF clock provider\n");
-			goto err_clocks;
-		}
-	}
-
-	return 0;
-
-err_clocks:
-	for (--i; i >= 0; --i) {
-		clkdev_drop(max77686_clks[i]->lookup);
-		clk_unregister(max77686_clks[i]->hw.clk);
-	}
-
-	return ret;
+	return max_gen_clk_probe(pdev, iodev->regmap, MAX77686_REG_32KHZ,
+				 max77686_clks_init, MAX77686_CLKS_NUM);
 }
 
 static int max77686_clk_remove(struct platform_device *pdev)
 {
-	struct max77686_dev *iodev = dev_get_drvdata(pdev->dev.parent);
-	struct clk **clocks = platform_get_drvdata(pdev);
-	int i;
-
-	if (iodev->dev->of_node)
-		of_clk_del_provider(iodev->dev->of_node);
-
-	for (i = 0; i < MAX77686_CLKS_NUM; i++) {
-		struct clk_hw *hw = __clk_get_hw(clocks[i]);
-		struct max77686_clk *max77686 = to_max77686_clk(hw);
-
-		clkdev_drop(max77686->lookup);
-		clk_unregister(clocks[i]);
-	}
-	return 0;
+	return max_gen_clk_remove(pdev, MAX77686_CLKS_NUM);
 }
 
 static const struct platform_device_id max77686_clk_id[] = {
@@ -228,24 +73,13 @@ MODULE_DEVICE_TABLE(platform, max77686_clk_id);
 static struct platform_driver max77686_clk_driver = {
 	.driver = {
 		.name  = "max77686-clk",
-		.owner = THIS_MODULE,
 	},
 	.probe = max77686_clk_probe,
 	.remove = max77686_clk_remove,
 	.id_table = max77686_clk_id,
 };
 
-static int __init max77686_clk_init(void)
-{
-	return platform_driver_register(&max77686_clk_driver);
-}
-subsys_initcall(max77686_clk_init);
-
-static void __init max77686_clk_cleanup(void)
-{
-	platform_driver_unregister(&max77686_clk_driver);
-}
-module_exit(max77686_clk_cleanup);
+module_platform_driver(max77686_clk_driver);
 
 MODULE_DESCRIPTION("MAXIM 77686 Clock Driver");
 MODULE_AUTHOR("Jonghwa Lee <jonghwa3.lee@samsung.com>");
diff --git a/drivers/clk/clk-max77802.c b/drivers/clk/clk-max77802.c
new file mode 100644
index 000000000000..0729dc723a8f
--- /dev/null
+++ b/drivers/clk/clk-max77802.c
@@ -0,0 +1,97 @@
+/*
+ * clk-max77802.c - Clock driver for Maxim 77802
+ *
+ * Copyright (C) 2014 Google, Inc
+ *
+ * Copyright (C) 2012 Samsung Electornics
+ * Jonghwa Lee <jonghwa3.lee@samsung.com>
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * This driver is based on clk-max77686.c
+ */
+
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/err.h>
+#include <linux/platform_device.h>
+#include <linux/mfd/max77686-private.h>
+#include <linux/clk-provider.h>
+#include <linux/mutex.h>
+#include <linux/clkdev.h>
+
+#include <dt-bindings/clock/maxim,max77802.h>
+#include "clk-max-gen.h"
+
+#define MAX77802_CLOCK_OPMODE_MASK	0x1
+#define MAX77802_CLOCK_LOW_JITTER_SHIFT 0x3
+
+static struct clk_init_data max77802_clks_init[MAX77802_CLKS_NUM] = {
+	[MAX77802_CLK_32K_AP] = {
+		.name = "32khz_ap",
+		.ops = &max_gen_clk_ops,
+		.flags = CLK_IS_ROOT,
+	},
+	[MAX77802_CLK_32K_CP] = {
+		.name = "32khz_cp",
+		.ops = &max_gen_clk_ops,
+		.flags = CLK_IS_ROOT,
+	},
+};
+
+static int max77802_clk_probe(struct platform_device *pdev)
+{
+	struct max77686_dev *iodev = dev_get_drvdata(pdev->dev.parent);
+	int ret;
+
+	ret = max_gen_clk_probe(pdev, iodev->regmap, MAX77802_REG_32KHZ,
+				max77802_clks_init, MAX77802_CLKS_NUM);
+
+	if (ret) {
+		dev_err(&pdev->dev, "generic probe failed %d\n", ret);
+		return ret;
+	}
+
+	/* Enable low-jitter mode on the 32khz clocks. */
+	ret = regmap_update_bits(iodev->regmap, MAX77802_REG_32KHZ,
+				 1 << MAX77802_CLOCK_LOW_JITTER_SHIFT,
+				 1 << MAX77802_CLOCK_LOW_JITTER_SHIFT);
+	if (ret < 0)
+		dev_err(&pdev->dev, "failed to enable low-jitter mode\n");
+
+	return ret;
+}
+
+static int max77802_clk_remove(struct platform_device *pdev)
+{
+	return max_gen_clk_remove(pdev, MAX77802_CLKS_NUM);
+}
+
+static const struct platform_device_id max77802_clk_id[] = {
+	{ "max77802-clk", 0},
+	{ },
+};
+MODULE_DEVICE_TABLE(platform, max77802_clk_id);
+
+static struct platform_driver max77802_clk_driver = {
+	.driver = {
+		.name  = "max77802-clk",
+	},
+	.probe = max77802_clk_probe,
+	.remove = max77802_clk_remove,
+	.id_table = max77802_clk_id,
+};
+
+module_platform_driver(max77802_clk_driver);
+
+MODULE_DESCRIPTION("MAXIM 77802 Clock Driver");
+MODULE_AUTHOR("Javier Martinez Canillas <javier.martinez@collabora.co.uk>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/clk/clk-mux.c b/drivers/clk/clk-mux.c
index 4f96ff3ba728..6e1ecf94bf58 100644
--- a/drivers/clk/clk-mux.c
+++ b/drivers/clk/clk-mux.c
@@ -77,7 +77,7 @@ static int clk_mux_set_parent(struct clk_hw *hw, u8 index)
 
 	else {
 		if (mux->flags & CLK_MUX_INDEX_BIT)
-			index = (1 << ffs(index));
+			index = 1 << index;
 
 		if (mux->flags & CLK_MUX_INDEX_ONE)
 			index++;
diff --git a/drivers/clk/clk-palmas.c b/drivers/clk/clk-palmas.c
index 781630e1372b..8d459923a15f 100644
--- a/drivers/clk/clk-palmas.c
+++ b/drivers/clk/clk-palmas.c
@@ -292,7 +292,6 @@ static int palmas_clks_remove(struct platform_device *pdev)
 static struct platform_driver palmas_clks_driver = {
 	.driver = {
 		.name = "palmas-clk",
-		.owner = THIS_MODULE,
 		.of_match_table = palmas_clks_of_match,
 	},
 	.probe = palmas_clks_probe,
diff --git a/drivers/clk/clk-ppc-corenet.c b/drivers/clk/clk-ppc-corenet.c
index 8e58edfeeb37..0a47d6f49cd6 100644
--- a/drivers/clk/clk-ppc-corenet.c
+++ b/drivers/clk/clk-ppc-corenet.c
@@ -291,10 +291,9 @@ static const struct of_device_id ppc_clk_ids[] __initconst = {
 	{}
 };
 
-static struct platform_driver ppc_corenet_clk_driver __initdata = {
+static struct platform_driver ppc_corenet_clk_driver = {
 	.driver = {
 		.name = "ppc_corenet_clock",
-		.owner = THIS_MODULE,
 		.of_match_table = ppc_clk_ids,
 	},
 	.probe = ppc_corenet_clk_probe,
diff --git a/drivers/clk/clk-rk808.c b/drivers/clk/clk-rk808.c
new file mode 100644
index 000000000000..83902b9cd49e
--- /dev/null
+++ b/drivers/clk/clk-rk808.c
@@ -0,0 +1,170 @@
+/*
+ * Clkout driver for Rockchip RK808
+ *
+ * Copyright (c) 2014, Fuzhou Rockchip Electronics Co., Ltd
+ *
+ * Author:Chris Zhong <zyw@rock-chips.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ */
+
+#include <linux/clk.h>
+#include <linux/clk-provider.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/platform_device.h>
+#include <linux/mfd/rk808.h>
+#include <linux/i2c.h>
+
+#define RK808_NR_OUTPUT 2
+
+struct rk808_clkout {
+	struct rk808 *rk808;
+	struct clk_onecell_data clk_data;
+	struct clk_hw		clkout1_hw;
+	struct clk_hw		clkout2_hw;
+};
+
+static unsigned long rk808_clkout_recalc_rate(struct clk_hw *hw,
+					      unsigned long parent_rate)
+{
+	return 32768;
+}
+
+static int rk808_clkout2_enable(struct clk_hw *hw, bool enable)
+{
+	struct rk808_clkout *rk808_clkout = container_of(hw,
+							 struct rk808_clkout,
+							 clkout2_hw);
+	struct rk808 *rk808 = rk808_clkout->rk808;
+
+	return regmap_update_bits(rk808->regmap, RK808_CLK32OUT_REG,
+				  CLK32KOUT2_EN, enable ? CLK32KOUT2_EN : 0);
+}
+
+static int rk808_clkout2_prepare(struct clk_hw *hw)
+{
+	return rk808_clkout2_enable(hw, true);
+}
+
+static void rk808_clkout2_unprepare(struct clk_hw *hw)
+{
+	rk808_clkout2_enable(hw, false);
+}
+
+static int rk808_clkout2_is_prepared(struct clk_hw *hw)
+{
+	struct rk808_clkout *rk808_clkout = container_of(hw,
+							 struct rk808_clkout,
+							 clkout2_hw);
+	struct rk808 *rk808 = rk808_clkout->rk808;
+	uint32_t val;
+
+	int ret = regmap_read(rk808->regmap, RK808_CLK32OUT_REG, &val);
+
+	if (ret < 0)
+		return ret;
+
+	return (val & CLK32KOUT2_EN) ? 1 : 0;
+}
+
+static const struct clk_ops rk808_clkout1_ops = {
+	.recalc_rate = rk808_clkout_recalc_rate,
+};
+
+static const struct clk_ops rk808_clkout2_ops = {
+	.prepare = rk808_clkout2_prepare,
+	.unprepare = rk808_clkout2_unprepare,
+	.is_prepared = rk808_clkout2_is_prepared,
+	.recalc_rate = rk808_clkout_recalc_rate,
+};
+
+static int rk808_clkout_probe(struct platform_device *pdev)
+{
+	struct rk808 *rk808 = dev_get_drvdata(pdev->dev.parent);
+	struct i2c_client *client = rk808->i2c;
+	struct device_node *node = client->dev.of_node;
+	struct clk_init_data init = {};
+	struct clk **clk_table;
+	struct rk808_clkout *rk808_clkout;
+
+	rk808_clkout = devm_kzalloc(&client->dev,
+				    sizeof(*rk808_clkout), GFP_KERNEL);
+	if (!rk808_clkout)
+		return -ENOMEM;
+
+	rk808_clkout->rk808 = rk808;
+
+	clk_table = devm_kcalloc(&client->dev, RK808_NR_OUTPUT,
+				 sizeof(struct clk *), GFP_KERNEL);
+	if (!clk_table)
+		return -ENOMEM;
+
+	init.flags = CLK_IS_ROOT;
+	init.parent_names = NULL;
+	init.num_parents = 0;
+	init.name = "rk808-clkout1";
+	init.ops = &rk808_clkout1_ops;
+	rk808_clkout->clkout1_hw.init = &init;
+
+	/* optional override of the clockname */
+	of_property_read_string_index(node, "clock-output-names",
+				      0, &init.name);
+
+	clk_table[0] = devm_clk_register(&client->dev,
+					 &rk808_clkout->clkout1_hw);
+	if (IS_ERR(clk_table[0]))
+		return PTR_ERR(clk_table[0]);
+
+	init.name = "rk808-clkout2";
+	init.ops = &rk808_clkout2_ops;
+	rk808_clkout->clkout2_hw.init = &init;
+
+	/* optional override of the clockname */
+	of_property_read_string_index(node, "clock-output-names",
+				      1, &init.name);
+
+	clk_table[1] = devm_clk_register(&client->dev,
+					 &rk808_clkout->clkout2_hw);
+	if (IS_ERR(clk_table[1]))
+		return PTR_ERR(clk_table[1]);
+
+	rk808_clkout->clk_data.clks = clk_table;
+	rk808_clkout->clk_data.clk_num = RK808_NR_OUTPUT;
+
+	return of_clk_add_provider(node, of_clk_src_onecell_get,
+				   &rk808_clkout->clk_data);
+}
+
+static int rk808_clkout_remove(struct platform_device *pdev)
+{
+	struct rk808 *rk808 = dev_get_drvdata(pdev->dev.parent);
+	struct i2c_client *client = rk808->i2c;
+	struct device_node *node = client->dev.of_node;
+
+	of_clk_del_provider(node);
+
+	return 0;
+}
+
+static struct platform_driver rk808_clkout_driver = {
+	.probe = rk808_clkout_probe,
+	.remove = rk808_clkout_remove,
+	.driver		= {
+		.name	= "rk808-clkout",
+	},
+};
+
+module_platform_driver(rk808_clkout_driver);
+
+MODULE_DESCRIPTION("Clkout driver for the rk808 series PMICs");
+MODULE_AUTHOR("Chris Zhong <zyw@rock-chips.com>");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:rk808-clkout");
diff --git a/drivers/clk/clk-s2mps11.c b/drivers/clk/clk-s2mps11.c
index b7797fb12e12..bfa1e64e267d 100644
--- a/drivers/clk/clk-s2mps11.c
+++ b/drivers/clk/clk-s2mps11.c
@@ -23,6 +23,7 @@
 #include <linux/clk-provider.h>
 #include <linux/platform_device.h>
 #include <linux/mfd/samsung/s2mps11.h>
+#include <linux/mfd/samsung/s2mps13.h>
 #include <linux/mfd/samsung/s2mps14.h>
 #include <linux/mfd/samsung/s5m8767.h>
 #include <linux/mfd/samsung/core.h>
@@ -120,6 +121,24 @@ static struct clk_init_data s2mps11_clks_init[S2MPS11_CLKS_NUM] = {
 	},
 };
 
+static struct clk_init_data s2mps13_clks_init[S2MPS11_CLKS_NUM] = {
+	[S2MPS11_CLK_AP] = {
+		.name = "s2mps13_ap",
+		.ops = &s2mps11_clk_ops,
+		.flags = CLK_IS_ROOT,
+	},
+	[S2MPS11_CLK_CP] = {
+		.name = "s2mps13_cp",
+		.ops = &s2mps11_clk_ops,
+		.flags = CLK_IS_ROOT,
+	},
+	[S2MPS11_CLK_BT] = {
+		.name = "s2mps13_bt",
+		.ops = &s2mps11_clk_ops,
+		.flags = CLK_IS_ROOT,
+	},
+};
+
 static struct clk_init_data s2mps14_clks_init[S2MPS11_CLKS_NUM] = {
 	[S2MPS11_CLK_AP] = {
 		.name = "s2mps14_ap",
@@ -184,6 +203,10 @@ static int s2mps11_clk_probe(struct platform_device *pdev)
 		s2mps11_reg = S2MPS11_REG_RTC_CTRL;
 		clks_init = s2mps11_clks_init;
 		break;
+	case S2MPS13X:
+		s2mps11_reg = S2MPS13_REG_RTCCTRL;
+		clks_init = s2mps13_clks_init;
+		break;
 	case S2MPS14X:
 		s2mps11_reg = S2MPS14_REG_RTCCTRL;
 		clks_init = s2mps14_clks_init;
@@ -195,7 +218,7 @@ static int s2mps11_clk_probe(struct platform_device *pdev)
 	default:
 		dev_err(&pdev->dev, "Invalid device type\n");
 		return -EINVAL;
-	};
+	}
 
 	/* Store clocks of_node in first element of s2mps11_clks array */
 	s2mps11_clks->clk_np = s2mps11_clk_parse_dt(pdev, clks_init);
@@ -279,6 +302,7 @@ static int s2mps11_clk_remove(struct platform_device *pdev)
 
 static const struct platform_device_id s2mps11_clk_id[] = {
 	{ "s2mps11-clk", S2MPS11X},
+	{ "s2mps13-clk", S2MPS13X},
 	{ "s2mps14-clk", S2MPS14X},
 	{ "s5m8767-clk", S5M8767X},
 	{ },
@@ -288,7 +312,6 @@ MODULE_DEVICE_TABLE(platform, s2mps11_clk_id);
 static struct platform_driver s2mps11_clk_driver = {
 	.driver = {
 		.name  = "s2mps11-clk",
-		.owner = THIS_MODULE,
 	},
 	.probe = s2mps11_clk_probe,
 	.remove = s2mps11_clk_remove,
diff --git a/drivers/clk/clk-twl6040.c b/drivers/clk/clk-twl6040.c
index 1ada79a28052..4a755135bcd3 100644
--- a/drivers/clk/clk-twl6040.c
+++ b/drivers/clk/clk-twl6040.c
@@ -112,7 +112,6 @@ static int twl6040_clk_remove(struct platform_device *pdev)
 static struct platform_driver twl6040_clk_driver = {
 	.driver = {
 		.name = "twl6040-clk",
-		.owner = THIS_MODULE,
 	},
 	.probe = twl6040_clk_probe,
 	.remove = twl6040_clk_remove,
diff --git a/drivers/clk/clk-wm831x.c b/drivers/clk/clk-wm831x.c
index b131041c8f48..ef67719f4e52 100644
--- a/drivers/clk/clk-wm831x.c
+++ b/drivers/clk/clk-wm831x.c
@@ -395,7 +395,6 @@ static struct platform_driver wm831x_clk_driver = {
 	.probe = wm831x_clk_probe,
 	.driver		= {
 		.name	= "wm831x-clk",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c
index bacc06ff939b..d48ac71c6c8b 100644
--- a/drivers/clk/clk.c
+++ b/drivers/clk/clk.c
@@ -100,6 +100,8 @@ static void clk_enable_unlock(unsigned long flags)
 
 static struct dentry *rootdir;
 static int inited = 0;
+static DEFINE_MUTEX(clk_debug_lock);
+static HLIST_HEAD(clk_debug_list);
 
 static struct hlist_head *all_lists[] = {
 	&clk_root_list,
@@ -117,11 +119,11 @@ static void clk_summary_show_one(struct seq_file *s, struct clk *c, int level)
 	if (!c)
 		return;
 
-	seq_printf(s, "%*s%-*s %11d %12d %11lu %10lu\n",
+	seq_printf(s, "%*s%-*s %11d %12d %11lu %10lu %-3d\n",
 		   level * 3 + 1, "",
 		   30 - level * 3, c->name,
 		   c->enable_count, c->prepare_count, clk_get_rate(c),
-		   clk_get_accuracy(c));
+		   clk_get_accuracy(c), clk_get_phase(c));
 }
 
 static void clk_summary_show_subtree(struct seq_file *s, struct clk *c,
@@ -143,8 +145,8 @@ static int clk_summary_show(struct seq_file *s, void *data)
 	struct clk *c;
 	struct hlist_head **lists = (struct hlist_head **)s->private;
 
-	seq_puts(s, "   clock                         enable_cnt  prepare_cnt        rate   accuracy\n");
-	seq_puts(s, "--------------------------------------------------------------------------------\n");
+	seq_puts(s, "   clock                         enable_cnt  prepare_cnt        rate   accuracy   phase\n");
+	seq_puts(s, "----------------------------------------------------------------------------------------\n");
 
 	clk_prepare_lock();
 
@@ -180,6 +182,7 @@ static void clk_dump_one(struct seq_file *s, struct clk *c, int level)
 	seq_printf(s, "\"prepare_count\": %d,", c->prepare_count);
 	seq_printf(s, "\"rate\": %lu", clk_get_rate(c));
 	seq_printf(s, "\"accuracy\": %lu", clk_get_accuracy(c));
+	seq_printf(s, "\"phase\": %d", clk_get_phase(c));
 }
 
 static void clk_dump_subtree(struct seq_file *s, struct clk *c, int level)
@@ -237,7 +240,6 @@ static const struct file_operations clk_dump_fops = {
 	.release	= single_release,
 };
 
-/* caller must hold prepare_lock */
 static int clk_debug_create_one(struct clk *clk, struct dentry *pdentry)
 {
 	struct dentry *d;
@@ -264,6 +266,11 @@ static int clk_debug_create_one(struct clk *clk, struct dentry *pdentry)
 	if (!d)
 		goto err_out;
 
+	d = debugfs_create_u32("clk_phase", S_IRUGO, clk->dentry,
+			(u32 *)&clk->phase);
+	if (!d)
+		goto err_out;
+
 	d = debugfs_create_x32("clk_flags", S_IRUGO, clk->dentry,
 			(u32 *)&clk->flags);
 	if (!d)
@@ -300,28 +307,6 @@ out:
 	return ret;
 }
 
-/* caller must hold prepare_lock */
-static int clk_debug_create_subtree(struct clk *clk, struct dentry *pdentry)
-{
-	struct clk *child;
-	int ret = -EINVAL;;
-
-	if (!clk || !pdentry)
-		goto out;
-
-	ret = clk_debug_create_one(clk, pdentry);
-
-	if (ret)
-		goto out;
-
-	hlist_for_each_entry(child, &clk->children, child_node)
-		clk_debug_create_subtree(child, pdentry);
-
-	ret = 0;
-out:
-	return ret;
-}
-
 /**
  * clk_debug_register - add a clk node to the debugfs clk tree
  * @clk: the clk being added to the debugfs clk tree
@@ -329,20 +314,21 @@ out:
  * Dynamically adds a clk to the debugfs clk tree if debugfs has been
  * initialized.  Otherwise it bails out early since the debugfs clk tree
  * will be created lazily by clk_debug_init as part of a late_initcall.
- *
- * Caller must hold prepare_lock.  Only clk_init calls this function (so
- * far) so this is taken care.
  */
 static int clk_debug_register(struct clk *clk)
 {
 	int ret = 0;
 
+	mutex_lock(&clk_debug_lock);
+	hlist_add_head(&clk->debug_node, &clk_debug_list);
+
 	if (!inited)
-		goto out;
+		goto unlock;
 
-	ret = clk_debug_create_subtree(clk, rootdir);
+	ret = clk_debug_create_one(clk, rootdir);
+unlock:
+	mutex_unlock(&clk_debug_lock);
 
-out:
 	return ret;
 }
 
@@ -353,21 +339,27 @@ out:
  * Dynamically removes a clk and all it's children clk nodes from the
  * debugfs clk tree if clk->dentry points to debugfs created by
  * clk_debug_register in __clk_init.
- *
- * Caller must hold prepare_lock.
  */
 static void clk_debug_unregister(struct clk *clk)
 {
+	mutex_lock(&clk_debug_lock);
+	if (!clk->dentry)
+		goto out;
+
+	hlist_del_init(&clk->debug_node);
 	debugfs_remove_recursive(clk->dentry);
+	clk->dentry = NULL;
+out:
+	mutex_unlock(&clk_debug_lock);
 }
 
-struct dentry *clk_debugfs_add_file(struct clk *clk, char *name, umode_t mode,
+struct dentry *clk_debugfs_add_file(struct clk_hw *hw, char *name, umode_t mode,
 				void *data, const struct file_operations *fops)
 {
 	struct dentry *d = NULL;
 
-	if (clk->dentry)
-		d = debugfs_create_file(name, mode, clk->dentry, data, fops);
+	if (hw->clk->dentry)
+		d = debugfs_create_file(name, mode, hw->clk->dentry, data, fops);
 
 	return d;
 }
@@ -415,17 +407,12 @@ static int __init clk_debug_init(void)
 	if (!d)
 		return -ENOMEM;
 
-	clk_prepare_lock();
-
-	hlist_for_each_entry(clk, &clk_root_list, child_node)
-		clk_debug_create_subtree(clk, rootdir);
-
-	hlist_for_each_entry(clk, &clk_orphan_list, child_node)
-		clk_debug_create_subtree(clk, rootdir);
+	mutex_lock(&clk_debug_lock);
+	hlist_for_each_entry(clk, &clk_debug_list, debug_node)
+		clk_debug_create_one(clk, rootdir);
 
 	inited = 1;
-
-	clk_prepare_unlock();
+	mutex_unlock(&clk_debug_lock);
 
 	return 0;
 }
@@ -586,11 +573,6 @@ unsigned int __clk_get_enable_count(struct clk *clk)
 	return !clk ? 0 : clk->enable_count;
 }
 
-unsigned int __clk_get_prepare_count(struct clk *clk)
-{
-	return !clk ? 0 : clk->prepare_count;
-}
-
 unsigned long __clk_get_rate(struct clk *clk)
 {
 	unsigned long ret;
@@ -613,7 +595,7 @@ out:
 }
 EXPORT_SYMBOL_GPL(__clk_get_rate);
 
-unsigned long __clk_get_accuracy(struct clk *clk)
+static unsigned long __clk_get_accuracy(struct clk *clk)
 {
 	if (!clk)
 		return 0;
@@ -719,7 +701,7 @@ struct clk *__clk_lookup(const char *name)
  */
 long __clk_mux_determine_rate(struct clk_hw *hw, unsigned long rate,
 			      unsigned long *best_parent_rate,
-			      struct clk **best_parent_p)
+			      struct clk_hw **best_parent_p)
 {
 	struct clk *clk = hw->clk, *parent, *best_parent = NULL;
 	int i, num_parents;
@@ -755,7 +737,7 @@ long __clk_mux_determine_rate(struct clk_hw *hw, unsigned long rate,
 
 out:
 	if (best_parent)
-		*best_parent_p = best_parent;
+		*best_parent_p = best_parent->hw;
 	*best_parent_rate = best;
 
 	return best;
@@ -963,6 +945,7 @@ unsigned long __clk_round_rate(struct clk *clk, unsigned long rate)
 {
 	unsigned long parent_rate = 0;
 	struct clk *parent;
+	struct clk_hw *parent_hw;
 
 	if (!clk)
 		return 0;
@@ -971,10 +954,11 @@ unsigned long __clk_round_rate(struct clk *clk, unsigned long rate)
 	if (parent)
 		parent_rate = parent->rate;
 
-	if (clk->ops->determine_rate)
+	if (clk->ops->determine_rate) {
+		parent_hw = parent ? parent->hw : NULL;
 		return clk->ops->determine_rate(clk->hw, rate, &parent_rate,
-						&parent);
-	else if (clk->ops->round_rate)
+						&parent_hw);
+	} else if (clk->ops->round_rate)
 		return clk->ops->round_rate(clk->hw, rate, &parent_rate);
 	else if (clk->flags & CLK_SET_RATE_PARENT)
 		return __clk_round_rate(clk->parent, rate);
@@ -1362,6 +1346,7 @@ static struct clk *clk_calc_new_rates(struct clk *clk, unsigned long rate)
 {
 	struct clk *top = clk;
 	struct clk *old_parent, *parent;
+	struct clk_hw *parent_hw;
 	unsigned long best_parent_rate = 0;
 	unsigned long new_rate;
 	int p_index = 0;
@@ -1377,9 +1362,11 @@ static struct clk *clk_calc_new_rates(struct clk *clk, unsigned long rate)
 
 	/* find the closest rate and parent clk/rate */
 	if (clk->ops->determine_rate) {
+		parent_hw = parent ? parent->hw : NULL;
 		new_rate = clk->ops->determine_rate(clk->hw, rate,
 						    &best_parent_rate,
-						    &parent);
+						    &parent_hw);
+		parent = parent_hw ? parent_hw->clk : NULL;
 	} else if (clk->ops->round_rate) {
 		new_rate = clk->ops->round_rate(clk->hw, rate,
 						&best_parent_rate);
@@ -1626,7 +1613,7 @@ static struct clk *__clk_init_parent(struct clk *clk)
 
 	if (clk->num_parents == 1) {
 		if (IS_ERR_OR_NULL(clk->parent))
-			ret = clk->parent = __clk_lookup(clk->parent_names[0]);
+			clk->parent = __clk_lookup(clk->parent_names[0]);
 		ret = clk->parent;
 		goto out;
 	}
@@ -1744,6 +1731,77 @@ out:
 EXPORT_SYMBOL_GPL(clk_set_parent);
 
 /**
+ * clk_set_phase - adjust the phase shift of a clock signal
+ * @clk: clock signal source
+ * @degrees: number of degrees the signal is shifted
+ *
+ * Shifts the phase of a clock signal by the specified
+ * degrees. Returns 0 on success, -EERROR otherwise.
+ *
+ * This function makes no distinction about the input or reference
+ * signal that we adjust the clock signal phase against. For example
+ * phase locked-loop clock signal generators we may shift phase with
+ * respect to feedback clock signal input, but for other cases the
+ * clock phase may be shifted with respect to some other, unspecified
+ * signal.
+ *
+ * Additionally the concept of phase shift does not propagate through
+ * the clock tree hierarchy, which sets it apart from clock rates and
+ * clock accuracy. A parent clock phase attribute does not have an
+ * impact on the phase attribute of a child clock.
+ */
+int clk_set_phase(struct clk *clk, int degrees)
+{
+	int ret = 0;
+
+	if (!clk)
+		goto out;
+
+	/* sanity check degrees */
+	degrees %= 360;
+	if (degrees < 0)
+		degrees += 360;
+
+	clk_prepare_lock();
+
+	if (!clk->ops->set_phase)
+		goto out_unlock;
+
+	ret = clk->ops->set_phase(clk->hw, degrees);
+
+	if (!ret)
+		clk->phase = degrees;
+
+out_unlock:
+	clk_prepare_unlock();
+
+out:
+	return ret;
+}
+
+/**
+ * clk_get_phase - return the phase shift of a clock signal
+ * @clk: clock signal source
+ *
+ * Returns the phase shift of a clock node in degrees, otherwise returns
+ * -EERROR.
+ */
+int clk_get_phase(struct clk *clk)
+{
+	int ret = 0;
+
+	if (!clk)
+		goto out;
+
+	clk_prepare_lock();
+	ret = clk->phase;
+	clk_prepare_unlock();
+
+out:
+	return ret;
+}
+
+/**
  * __clk_init - initialize the data structures in a struct clk
  * @dev:	device initializing this clk, placeholder for now
  * @clk:	clk being initialized
@@ -1862,6 +1920,16 @@ int __clk_init(struct device *dev, struct clk *clk)
 		clk->accuracy = 0;
 
 	/*
+	 * Set clk's phase.
+	 * Since a phase is by definition relative to its parent, just
+	 * query the current clock phase, or just assume it's in phase.
+	 */
+	if (clk->ops->get_phase)
+		clk->phase = clk->ops->get_phase(clk->hw);
+	else
+		clk->phase = 0;
+
+	/*
 	 * Set clk's rate.  The preferred method is to use .recalc_rate.  For
 	 * simple clocks and lazy developers the default fallback is to use the
 	 * parent's rate.  If a clock doesn't have a parent (or is orphaned)
@@ -1875,7 +1943,6 @@ int __clk_init(struct device *dev, struct clk *clk)
 	else
 		clk->rate = 0;
 
-	clk_debug_register(clk);
 	/*
 	 * walk the list of orphan clocks and reparent any that are children of
 	 * this clock
@@ -1910,6 +1977,9 @@ int __clk_init(struct device *dev, struct clk *clk)
 out:
 	clk_prepare_unlock();
 
+	if (!ret)
+		clk_debug_register(clk);
+
 	return ret;
 }
 
@@ -2092,14 +2162,16 @@ void clk_unregister(struct clk *clk)
 {
 	unsigned long flags;
 
-       if (!clk || WARN_ON_ONCE(IS_ERR(clk)))
-               return;
+	if (!clk || WARN_ON_ONCE(IS_ERR(clk)))
+		return;
+
+	clk_debug_unregister(clk);
 
 	clk_prepare_lock();
 
 	if (clk->ops == &clk_nodrv_ops) {
 		pr_err("%s: unregistered clock: %s\n", __func__, clk->name);
-		goto out;
+		return;
 	}
 	/*
 	 * Assign empty clock ops for consumers that might still hold
@@ -2118,16 +2190,13 @@ void clk_unregister(struct clk *clk)
 			clk_set_parent(child, NULL);
 	}
 
-	clk_debug_unregister(clk);
-
 	hlist_del_init(&clk->child_node);
 
 	if (clk->prepare_count)
 		pr_warn("%s: unregistering prepared clock: %s\n",
 					__func__, clk->name);
-
 	kref_put(&clk->ref, __clk_release);
-out:
+
 	clk_prepare_unlock();
 }
 EXPORT_SYMBOL_GPL(clk_unregister);
@@ -2205,14 +2274,17 @@ int __clk_get(struct clk *clk)
 
 void __clk_put(struct clk *clk)
 {
+	struct module *owner;
+
 	if (!clk || WARN_ON_ONCE(IS_ERR(clk)))
 		return;
 
 	clk_prepare_lock();
+	owner = clk->owner;
 	kref_put(&clk->ref, __clk_release);
 	clk_prepare_unlock();
 
-	module_put(clk->owner);
+	module_put(owner);
 }
 
 /***        clk rate change notifiers        ***/
diff --git a/drivers/clk/hisilicon/clk-hi3620.c b/drivers/clk/hisilicon/clk-hi3620.c
index 339945d2503b..007144f81f50 100644
--- a/drivers/clk/hisilicon/clk-hi3620.c
+++ b/drivers/clk/hisilicon/clk-hi3620.c
@@ -38,44 +38,44 @@
 #include "clk.h"
 
 /* clock parent list */
-static const char *timer0_mux_p[] __initdata = { "osc32k", "timerclk01", };
-static const char *timer1_mux_p[] __initdata = { "osc32k", "timerclk01", };
-static const char *timer2_mux_p[] __initdata = { "osc32k", "timerclk23", };
-static const char *timer3_mux_p[] __initdata = { "osc32k", "timerclk23", };
-static const char *timer4_mux_p[] __initdata = { "osc32k", "timerclk45", };
-static const char *timer5_mux_p[] __initdata = { "osc32k", "timerclk45", };
-static const char *timer6_mux_p[] __initdata = { "osc32k", "timerclk67", };
-static const char *timer7_mux_p[] __initdata = { "osc32k", "timerclk67", };
-static const char *timer8_mux_p[] __initdata = { "osc32k", "timerclk89", };
-static const char *timer9_mux_p[] __initdata = { "osc32k", "timerclk89", };
-static const char *uart0_mux_p[] __initdata = { "osc26m", "pclk", };
-static const char *uart1_mux_p[] __initdata = { "osc26m", "pclk", };
-static const char *uart2_mux_p[] __initdata = { "osc26m", "pclk", };
-static const char *uart3_mux_p[] __initdata = { "osc26m", "pclk", };
-static const char *uart4_mux_p[] __initdata = { "osc26m", "pclk", };
-static const char *spi0_mux_p[] __initdata = { "osc26m", "rclk_cfgaxi", };
-static const char *spi1_mux_p[] __initdata = { "osc26m", "rclk_cfgaxi", };
-static const char *spi2_mux_p[] __initdata = { "osc26m", "rclk_cfgaxi", };
+static const char *timer0_mux_p[] __initconst = { "osc32k", "timerclk01", };
+static const char *timer1_mux_p[] __initconst = { "osc32k", "timerclk01", };
+static const char *timer2_mux_p[] __initconst = { "osc32k", "timerclk23", };
+static const char *timer3_mux_p[] __initconst = { "osc32k", "timerclk23", };
+static const char *timer4_mux_p[] __initconst = { "osc32k", "timerclk45", };
+static const char *timer5_mux_p[] __initconst = { "osc32k", "timerclk45", };
+static const char *timer6_mux_p[] __initconst = { "osc32k", "timerclk67", };
+static const char *timer7_mux_p[] __initconst = { "osc32k", "timerclk67", };
+static const char *timer8_mux_p[] __initconst = { "osc32k", "timerclk89", };
+static const char *timer9_mux_p[] __initconst = { "osc32k", "timerclk89", };
+static const char *uart0_mux_p[] __initconst = { "osc26m", "pclk", };
+static const char *uart1_mux_p[] __initconst = { "osc26m", "pclk", };
+static const char *uart2_mux_p[] __initconst = { "osc26m", "pclk", };
+static const char *uart3_mux_p[] __initconst = { "osc26m", "pclk", };
+static const char *uart4_mux_p[] __initconst = { "osc26m", "pclk", };
+static const char *spi0_mux_p[] __initconst = { "osc26m", "rclk_cfgaxi", };
+static const char *spi1_mux_p[] __initconst = { "osc26m", "rclk_cfgaxi", };
+static const char *spi2_mux_p[] __initconst = { "osc26m", "rclk_cfgaxi", };
 /* share axi parent */
-static const char *saxi_mux_p[] __initdata = { "armpll3", "armpll2", };
-static const char *pwm0_mux_p[] __initdata = { "osc32k", "osc26m", };
-static const char *pwm1_mux_p[] __initdata = { "osc32k", "osc26m", };
-static const char *sd_mux_p[] __initdata = { "armpll2", "armpll3", };
-static const char *mmc1_mux_p[] __initdata = { "armpll2", "armpll3", };
-static const char *mmc1_mux2_p[] __initdata = { "osc26m", "mmc1_div", };
-static const char *g2d_mux_p[] __initdata = { "armpll2", "armpll3", };
-static const char *venc_mux_p[] __initdata = { "armpll2", "armpll3", };
-static const char *vdec_mux_p[] __initdata = { "armpll2", "armpll3", };
-static const char *vpp_mux_p[] __initdata = { "armpll2", "armpll3", };
-static const char *edc0_mux_p[] __initdata = { "armpll2", "armpll3", };
-static const char *ldi0_mux_p[] __initdata = { "armpll2", "armpll4",
+static const char *saxi_mux_p[] __initconst = { "armpll3", "armpll2", };
+static const char *pwm0_mux_p[] __initconst = { "osc32k", "osc26m", };
+static const char *pwm1_mux_p[] __initconst = { "osc32k", "osc26m", };
+static const char *sd_mux_p[] __initconst = { "armpll2", "armpll3", };
+static const char *mmc1_mux_p[] __initconst = { "armpll2", "armpll3", };
+static const char *mmc1_mux2_p[] __initconst = { "osc26m", "mmc1_div", };
+static const char *g2d_mux_p[] __initconst = { "armpll2", "armpll3", };
+static const char *venc_mux_p[] __initconst = { "armpll2", "armpll3", };
+static const char *vdec_mux_p[] __initconst = { "armpll2", "armpll3", };
+static const char *vpp_mux_p[] __initconst = { "armpll2", "armpll3", };
+static const char *edc0_mux_p[] __initconst = { "armpll2", "armpll3", };
+static const char *ldi0_mux_p[] __initconst = { "armpll2", "armpll4",
 					     "armpll3", "armpll5", };
-static const char *edc1_mux_p[] __initdata = { "armpll2", "armpll3", };
-static const char *ldi1_mux_p[] __initdata = { "armpll2", "armpll4",
+static const char *edc1_mux_p[] __initconst = { "armpll2", "armpll3", };
+static const char *ldi1_mux_p[] __initconst = { "armpll2", "armpll4",
 					     "armpll3", "armpll5", };
-static const char *rclk_hsic_p[] __initdata = { "armpll3", "armpll2", };
-static const char *mmc2_mux_p[] __initdata = { "armpll2", "armpll3", };
-static const char *mmc3_mux_p[] __initdata = { "armpll2", "armpll3", };
+static const char *rclk_hsic_p[] __initconst = { "armpll3", "armpll2", };
+static const char *mmc2_mux_p[] __initconst = { "armpll2", "armpll3", };
+static const char *mmc3_mux_p[] __initconst = { "armpll2", "armpll3", };
 
 
 /* fixed rate clocks */
@@ -296,7 +296,7 @@ static unsigned long mmc_clk_recalc_rate(struct clk_hw *hw,
 
 static long mmc_clk_determine_rate(struct clk_hw *hw, unsigned long rate,
 			      unsigned long *best_parent_rate,
-			      struct clk **best_parent_p)
+			      struct clk_hw **best_parent_p)
 {
 	struct clk_mmc *mclk = to_mmc(hw);
 	unsigned long best = 0;
diff --git a/drivers/clk/hisilicon/clk-hix5hd2.c b/drivers/clk/hisilicon/clk-hix5hd2.c
index e5fcfb4e32ef..3f369c60fe56 100644
--- a/drivers/clk/hisilicon/clk-hix5hd2.c
+++ b/drivers/clk/hisilicon/clk-hix5hd2.c
@@ -9,6 +9,8 @@
 
 #include <linux/of_address.h>
 #include <dt-bindings/clock/hix5hd2-clock.h>
+#include <linux/slab.h>
+#include <linux/delay.h>
 #include "clk.h"
 
 static struct hisi_fixed_rate_clock hix5hd2_fixed_rate_clks[] __initdata = {
@@ -48,9 +50,9 @@ static const char *sfc_mux_p[] __initconst = {
 		"24m", "150m", "200m", "100m", "75m", };
 static u32 sfc_mux_table[] = {0, 4, 5, 6, 7};
 
-static const char *sdio1_mux_p[] __initconst = {
+static const char *sdio_mux_p[] __initconst = {
 		"75m", "100m", "50m", "15m", };
-static u32 sdio1_mux_table[] = {0, 1, 2, 3};
+static u32 sdio_mux_table[] = {0, 1, 2, 3};
 
 static const char *fephy_mux_p[] __initconst = { "25m", "125m"};
 static u32 fephy_mux_table[] = {0, 1};
@@ -59,28 +61,243 @@ static u32 fephy_mux_table[] = {0, 1};
 static struct hisi_mux_clock hix5hd2_mux_clks[] __initdata = {
 	{ HIX5HD2_SFC_MUX, "sfc_mux", sfc_mux_p, ARRAY_SIZE(sfc_mux_p),
 		CLK_SET_RATE_PARENT, 0x5c, 8, 3, 0, sfc_mux_table, },
-	{ HIX5HD2_MMC_MUX, "mmc_mux", sdio1_mux_p, ARRAY_SIZE(sdio1_mux_p),
-		CLK_SET_RATE_PARENT, 0xa0, 8, 2, 0, sdio1_mux_table, },
+	{ HIX5HD2_MMC_MUX, "mmc_mux", sdio_mux_p, ARRAY_SIZE(sdio_mux_p),
+		CLK_SET_RATE_PARENT, 0xa0, 8, 2, 0, sdio_mux_table, },
+	{ HIX5HD2_SD_MUX, "sd_mux", sdio_mux_p, ARRAY_SIZE(sdio_mux_p),
+		CLK_SET_RATE_PARENT, 0x9c, 8, 2, 0, sdio_mux_table, },
 	{ HIX5HD2_FEPHY_MUX, "fephy_mux",
 		fephy_mux_p, ARRAY_SIZE(fephy_mux_p),
 		CLK_SET_RATE_PARENT, 0x120, 8, 2, 0, fephy_mux_table, },
 };
 
 static struct hisi_gate_clock hix5hd2_gate_clks[] __initdata = {
-	/*sfc*/
+	/* sfc */
 	{ HIX5HD2_SFC_CLK, "clk_sfc", "sfc_mux",
 		CLK_SET_RATE_PARENT, 0x5c, 0, 0, },
 	{ HIX5HD2_SFC_RST, "rst_sfc", "clk_sfc",
 		CLK_SET_RATE_PARENT, 0x5c, 4, CLK_GATE_SET_TO_DISABLE, },
-	/*sdio1*/
+	/* sdio0 */
+	{ HIX5HD2_SD_BIU_CLK, "clk_sd_biu", "200m",
+		CLK_SET_RATE_PARENT, 0x9c, 0, 0, },
+	{ HIX5HD2_SD_CIU_CLK, "clk_sd_ciu", "sd_mux",
+		CLK_SET_RATE_PARENT, 0x9c, 1, 0, },
+	{ HIX5HD2_SD_CIU_RST, "rst_sd_ciu", "clk_sd_ciu",
+		CLK_SET_RATE_PARENT, 0x9c, 4, CLK_GATE_SET_TO_DISABLE, },
+	/* sdio1 */
 	{ HIX5HD2_MMC_BIU_CLK, "clk_mmc_biu", "200m",
 		CLK_SET_RATE_PARENT, 0xa0, 0, 0, },
 	{ HIX5HD2_MMC_CIU_CLK, "clk_mmc_ciu", "mmc_mux",
 		CLK_SET_RATE_PARENT, 0xa0, 1, 0, },
 	{ HIX5HD2_MMC_CIU_RST, "rst_mmc_ciu", "clk_mmc_ciu",
 		CLK_SET_RATE_PARENT, 0xa0, 4, CLK_GATE_SET_TO_DISABLE, },
+	/* gsf */
+	{ HIX5HD2_FWD_BUS_CLK, "clk_fwd_bus", NULL, 0, 0xcc, 0, 0, },
+	{ HIX5HD2_FWD_SYS_CLK, "clk_fwd_sys", "clk_fwd_bus", 0, 0xcc, 5, 0, },
+	{ HIX5HD2_MAC0_PHY_CLK, "clk_fephy", "clk_fwd_sys",
+		 CLK_SET_RATE_PARENT, 0x120, 0, 0, },
+	/* wdg0 */
+	{ HIX5HD2_WDG0_CLK, "clk_wdg0", "24m",
+		CLK_SET_RATE_PARENT, 0x178, 0, 0, },
+	{ HIX5HD2_WDG0_RST, "rst_wdg0", "clk_wdg0",
+		CLK_SET_RATE_PARENT, 0x178, 4, CLK_GATE_SET_TO_DISABLE, },
+	/* I2C */
+	{HIX5HD2_I2C0_CLK, "clk_i2c0", "100m",
+		 CLK_SET_RATE_PARENT, 0x06c, 4, 0, },
+	{HIX5HD2_I2C0_RST, "rst_i2c0", "clk_i2c0",
+		 CLK_SET_RATE_PARENT, 0x06c, 5, CLK_GATE_SET_TO_DISABLE, },
+	{HIX5HD2_I2C1_CLK, "clk_i2c1", "100m",
+		 CLK_SET_RATE_PARENT, 0x06c, 8, 0, },
+	{HIX5HD2_I2C1_RST, "rst_i2c1", "clk_i2c1",
+		 CLK_SET_RATE_PARENT, 0x06c, 9, CLK_GATE_SET_TO_DISABLE, },
+	{HIX5HD2_I2C2_CLK, "clk_i2c2", "100m",
+		 CLK_SET_RATE_PARENT, 0x06c, 12, 0, },
+	{HIX5HD2_I2C2_RST, "rst_i2c2", "clk_i2c2",
+		 CLK_SET_RATE_PARENT, 0x06c, 13, CLK_GATE_SET_TO_DISABLE, },
+	{HIX5HD2_I2C3_CLK, "clk_i2c3", "100m",
+		 CLK_SET_RATE_PARENT, 0x06c, 16, 0, },
+	{HIX5HD2_I2C3_RST, "rst_i2c3", "clk_i2c3",
+		 CLK_SET_RATE_PARENT, 0x06c, 17, CLK_GATE_SET_TO_DISABLE, },
+	{HIX5HD2_I2C4_CLK, "clk_i2c4", "100m",
+		 CLK_SET_RATE_PARENT, 0x06c, 20, 0, },
+	{HIX5HD2_I2C4_RST, "rst_i2c4", "clk_i2c4",
+		 CLK_SET_RATE_PARENT, 0x06c, 21, CLK_GATE_SET_TO_DISABLE, },
+	{HIX5HD2_I2C5_CLK, "clk_i2c5", "100m",
+		 CLK_SET_RATE_PARENT, 0x06c, 0, 0, },
+	{HIX5HD2_I2C5_RST, "rst_i2c5", "clk_i2c5",
+		 CLK_SET_RATE_PARENT, 0x06c, 1, CLK_GATE_SET_TO_DISABLE, },
 };
 
+enum hix5hd2_clk_type {
+	TYPE_COMPLEX,
+	TYPE_ETHER,
+};
+
+struct hix5hd2_complex_clock {
+	const char	*name;
+	const char	*parent_name;
+	u32		id;
+	u32		ctrl_reg;
+	u32		ctrl_clk_mask;
+	u32		ctrl_rst_mask;
+	u32		phy_reg;
+	u32		phy_clk_mask;
+	u32		phy_rst_mask;
+	enum hix5hd2_clk_type type;
+};
+
+struct hix5hd2_clk_complex {
+	struct clk_hw	hw;
+	u32		id;
+	void __iomem	*ctrl_reg;
+	u32		ctrl_clk_mask;
+	u32		ctrl_rst_mask;
+	void __iomem	*phy_reg;
+	u32		phy_clk_mask;
+	u32		phy_rst_mask;
+};
+
+static struct hix5hd2_complex_clock hix5hd2_complex_clks[] __initdata = {
+	{"clk_mac0", "clk_fephy", HIX5HD2_MAC0_CLK,
+		0xcc, 0xa, 0x500, 0x120, 0, 0x10, TYPE_ETHER},
+	{"clk_mac1", "clk_fwd_sys", HIX5HD2_MAC1_CLK,
+		0xcc, 0x14, 0xa00, 0x168, 0x2, 0, TYPE_ETHER},
+	{"clk_sata", NULL, HIX5HD2_SATA_CLK,
+		0xa8, 0x1f, 0x300, 0xac, 0x1, 0x0, TYPE_COMPLEX},
+	{"clk_usb", NULL, HIX5HD2_USB_CLK,
+		0xb8, 0xff, 0x3f000, 0xbc, 0x7, 0x3f00, TYPE_COMPLEX},
+};
+
+#define to_complex_clk(_hw) container_of(_hw, struct hix5hd2_clk_complex, hw)
+
+static int clk_ether_prepare(struct clk_hw *hw)
+{
+	struct hix5hd2_clk_complex *clk = to_complex_clk(hw);
+	u32 val;
+
+	val = readl_relaxed(clk->ctrl_reg);
+	val |= clk->ctrl_clk_mask | clk->ctrl_rst_mask;
+	writel_relaxed(val, clk->ctrl_reg);
+	val &= ~(clk->ctrl_rst_mask);
+	writel_relaxed(val, clk->ctrl_reg);
+
+	val = readl_relaxed(clk->phy_reg);
+	val |= clk->phy_clk_mask;
+	val &= ~(clk->phy_rst_mask);
+	writel_relaxed(val, clk->phy_reg);
+	mdelay(10);
+
+	val &= ~(clk->phy_clk_mask);
+	val |= clk->phy_rst_mask;
+	writel_relaxed(val, clk->phy_reg);
+	mdelay(10);
+
+	val |= clk->phy_clk_mask;
+	val &= ~(clk->phy_rst_mask);
+	writel_relaxed(val, clk->phy_reg);
+	mdelay(30);
+	return 0;
+}
+
+static void clk_ether_unprepare(struct clk_hw *hw)
+{
+	struct hix5hd2_clk_complex *clk = to_complex_clk(hw);
+	u32 val;
+
+	val = readl_relaxed(clk->ctrl_reg);
+	val &= ~(clk->ctrl_clk_mask);
+	writel_relaxed(val, clk->ctrl_reg);
+}
+
+static struct clk_ops clk_ether_ops = {
+	.prepare = clk_ether_prepare,
+	.unprepare = clk_ether_unprepare,
+};
+
+static int clk_complex_enable(struct clk_hw *hw)
+{
+	struct hix5hd2_clk_complex *clk = to_complex_clk(hw);
+	u32 val;
+
+	val = readl_relaxed(clk->ctrl_reg);
+	val |= clk->ctrl_clk_mask;
+	val &= ~(clk->ctrl_rst_mask);
+	writel_relaxed(val, clk->ctrl_reg);
+
+	val = readl_relaxed(clk->phy_reg);
+	val |= clk->phy_clk_mask;
+	val &= ~(clk->phy_rst_mask);
+	writel_relaxed(val, clk->phy_reg);
+
+	return 0;
+}
+
+static void clk_complex_disable(struct clk_hw *hw)
+{
+	struct hix5hd2_clk_complex *clk = to_complex_clk(hw);
+	u32 val;
+
+	val = readl_relaxed(clk->ctrl_reg);
+	val |= clk->ctrl_rst_mask;
+	val &= ~(clk->ctrl_clk_mask);
+	writel_relaxed(val, clk->ctrl_reg);
+
+	val = readl_relaxed(clk->phy_reg);
+	val |= clk->phy_rst_mask;
+	val &= ~(clk->phy_clk_mask);
+	writel_relaxed(val, clk->phy_reg);
+}
+
+static struct clk_ops clk_complex_ops = {
+	.enable = clk_complex_enable,
+	.disable = clk_complex_disable,
+};
+
+void __init hix5hd2_clk_register_complex(struct hix5hd2_complex_clock *clks,
+					 int nums, struct hisi_clock_data *data)
+{
+	void __iomem *base = data->base;
+	int i;
+
+	for (i = 0; i < nums; i++) {
+		struct hix5hd2_clk_complex *p_clk;
+		struct clk *clk;
+		struct clk_init_data init;
+
+		p_clk = kzalloc(sizeof(*p_clk), GFP_KERNEL);
+		if (!p_clk)
+			return;
+
+		init.name = clks[i].name;
+		if (clks[i].type == TYPE_ETHER)
+			init.ops = &clk_ether_ops;
+		else
+			init.ops = &clk_complex_ops;
+
+		init.flags = CLK_IS_BASIC;
+		init.parent_names =
+			(clks[i].parent_name ? &clks[i].parent_name : NULL);
+		init.num_parents = (clks[i].parent_name ? 1 : 0);
+
+		p_clk->ctrl_reg = base + clks[i].ctrl_reg;
+		p_clk->ctrl_clk_mask = clks[i].ctrl_clk_mask;
+		p_clk->ctrl_rst_mask = clks[i].ctrl_rst_mask;
+		p_clk->phy_reg = base + clks[i].phy_reg;
+		p_clk->phy_clk_mask = clks[i].phy_clk_mask;
+		p_clk->phy_rst_mask = clks[i].phy_rst_mask;
+		p_clk->hw.init = &init;
+
+		clk = clk_register(NULL, &p_clk->hw);
+		if (IS_ERR(clk)) {
+			kfree(p_clk);
+			pr_err("%s: failed to register clock %s\n",
+			       __func__, clks[i].name);
+			continue;
+		}
+
+		data->clk_data.clks[clks[i].id] = clk;
+	}
+}
+
 static void __init hix5hd2_clk_init(struct device_node *np)
 {
 	struct hisi_clock_data *clk_data;
@@ -96,6 +313,9 @@ static void __init hix5hd2_clk_init(struct device_node *np)
 					clk_data);
 	hisi_clk_register_gate(hix5hd2_gate_clks,
 			ARRAY_SIZE(hix5hd2_gate_clks), clk_data);
+	hix5hd2_clk_register_complex(hix5hd2_complex_clks,
+				     ARRAY_SIZE(hix5hd2_complex_clks),
+				     clk_data);
 }
 
 CLK_OF_DECLARE(hix5hd2_clk, "hisilicon,hix5hd2-clock", hix5hd2_clk_init);
diff --git a/drivers/clk/mmp/Makefile b/drivers/clk/mmp/Makefile
index 392d78044ce3..3caaf7cc169c 100644
--- a/drivers/clk/mmp/Makefile
+++ b/drivers/clk/mmp/Makefile
@@ -2,7 +2,12 @@
 # Makefile for mmp specific clk
 #
 
-obj-y += clk-apbc.o clk-apmu.o clk-frac.o
+obj-y += clk-apbc.o clk-apmu.o clk-frac.o clk-mix.o clk-gate.o clk.o
+
+obj-$(CONFIG_RESET_CONTROLLER) += reset.o
+
+obj-$(CONFIG_MACH_MMP_DT) += clk-of-pxa168.o clk-of-pxa910.o
+obj-$(CONFIG_MACH_MMP2_DT) += clk-of-mmp2.o
 
 obj-$(CONFIG_CPU_PXA168) += clk-pxa168.o
 obj-$(CONFIG_CPU_PXA910) += clk-pxa910.o
diff --git a/drivers/clk/mmp/clk-frac.c b/drivers/clk/mmp/clk-frac.c
index 23a56f561812..584a9927993b 100644
--- a/drivers/clk/mmp/clk-frac.c
+++ b/drivers/clk/mmp/clk-frac.c
@@ -22,19 +22,12 @@
  * numerator/denominator = Fin / (Fout * factor)
  */
 
-#define to_clk_factor(hw) container_of(hw, struct clk_factor, hw)
-struct clk_factor {
-	struct clk_hw		hw;
-	void __iomem		*base;
-	struct clk_factor_masks	*masks;
-	struct clk_factor_tbl	*ftbl;
-	unsigned int		ftbl_cnt;
-};
+#define to_clk_factor(hw) container_of(hw, struct mmp_clk_factor, hw)
 
 static long clk_factor_round_rate(struct clk_hw *hw, unsigned long drate,
 		unsigned long *prate)
 {
-	struct clk_factor *factor = to_clk_factor(hw);
+	struct mmp_clk_factor *factor = to_clk_factor(hw);
 	unsigned long rate = 0, prev_rate;
 	int i;
 
@@ -58,8 +51,8 @@ static long clk_factor_round_rate(struct clk_hw *hw, unsigned long drate,
 static unsigned long clk_factor_recalc_rate(struct clk_hw *hw,
 		unsigned long parent_rate)
 {
-	struct clk_factor *factor = to_clk_factor(hw);
-	struct clk_factor_masks *masks = factor->masks;
+	struct mmp_clk_factor *factor = to_clk_factor(hw);
+	struct mmp_clk_factor_masks *masks = factor->masks;
 	unsigned int val, num, den;
 
 	val = readl_relaxed(factor->base);
@@ -81,11 +74,12 @@ static unsigned long clk_factor_recalc_rate(struct clk_hw *hw,
 static int clk_factor_set_rate(struct clk_hw *hw, unsigned long drate,
 				unsigned long prate)
 {
-	struct clk_factor *factor = to_clk_factor(hw);
-	struct clk_factor_masks *masks = factor->masks;
+	struct mmp_clk_factor *factor = to_clk_factor(hw);
+	struct mmp_clk_factor_masks *masks = factor->masks;
 	int i;
 	unsigned long val;
 	unsigned long prev_rate, rate = 0;
+	unsigned long flags = 0;
 
 	for (i = 0; i < factor->ftbl_cnt; i++) {
 		prev_rate = rate;
@@ -97,6 +91,9 @@ static int clk_factor_set_rate(struct clk_hw *hw, unsigned long drate,
 	if (i > 0)
 		i--;
 
+	if (factor->lock)
+		spin_lock_irqsave(factor->lock, flags);
+
 	val = readl_relaxed(factor->base);
 
 	val &= ~(masks->num_mask << masks->num_shift);
@@ -107,21 +104,65 @@ static int clk_factor_set_rate(struct clk_hw *hw, unsigned long drate,
 
 	writel_relaxed(val, factor->base);
 
+	if (factor->lock)
+		spin_unlock_irqrestore(factor->lock, flags);
+
 	return 0;
 }
 
+static void clk_factor_init(struct clk_hw *hw)
+{
+	struct mmp_clk_factor *factor = to_clk_factor(hw);
+	struct mmp_clk_factor_masks *masks = factor->masks;
+	u32 val, num, den;
+	int i;
+	unsigned long flags = 0;
+
+	if (factor->lock)
+		spin_lock_irqsave(factor->lock, flags);
+
+	val = readl(factor->base);
+
+	/* calculate numerator */
+	num = (val >> masks->num_shift) & masks->num_mask;
+
+	/* calculate denominator */
+	den = (val >> masks->den_shift) & masks->den_mask;
+
+	for (i = 0; i < factor->ftbl_cnt; i++)
+		if (den == factor->ftbl[i].den && num == factor->ftbl[i].num)
+			break;
+
+	if (i >= factor->ftbl_cnt) {
+		val &= ~(masks->num_mask << masks->num_shift);
+		val |= (factor->ftbl[0].num & masks->num_mask) <<
+			masks->num_shift;
+
+		val &= ~(masks->den_mask << masks->den_shift);
+		val |= (factor->ftbl[0].den & masks->den_mask) <<
+			masks->den_shift;
+
+		writel(val, factor->base);
+	}
+
+	if (factor->lock)
+		spin_unlock_irqrestore(factor->lock, flags);
+}
+
 static struct clk_ops clk_factor_ops = {
 	.recalc_rate = clk_factor_recalc_rate,
 	.round_rate = clk_factor_round_rate,
 	.set_rate = clk_factor_set_rate,
+	.init = clk_factor_init,
 };
 
 struct clk *mmp_clk_register_factor(const char *name, const char *parent_name,
 		unsigned long flags, void __iomem *base,
-		struct clk_factor_masks *masks, struct clk_factor_tbl *ftbl,
-		unsigned int ftbl_cnt)
+		struct mmp_clk_factor_masks *masks,
+		struct mmp_clk_factor_tbl *ftbl,
+		unsigned int ftbl_cnt, spinlock_t *lock)
 {
-	struct clk_factor *factor;
+	struct mmp_clk_factor *factor;
 	struct clk_init_data init;
 	struct clk *clk;
 
@@ -142,6 +183,7 @@ struct clk *mmp_clk_register_factor(const char *name, const char *parent_name,
 	factor->ftbl = ftbl;
 	factor->ftbl_cnt = ftbl_cnt;
 	factor->hw.init = &init;
+	factor->lock = lock;
 
 	init.name = name;
 	init.ops = &clk_factor_ops;
diff --git a/drivers/clk/mmp/clk-gate.c b/drivers/clk/mmp/clk-gate.c
new file mode 100644
index 000000000000..adbd9d64ded2
--- /dev/null
+++ b/drivers/clk/mmp/clk-gate.c
@@ -0,0 +1,133 @@
+/*
+ * mmp gate clock operation source file
+ *
+ * Copyright (C) 2014 Marvell
+ * Chao Xie <chao.xie@marvell.com>
+ *
+ * This file is licensed under the terms of the GNU General Public
+ * License version 2. This program is licensed "as is" without any
+ * warranty of any kind, whether express or implied.
+ */
+
+#include <linux/clk-provider.h>
+#include <linux/slab.h>
+#include <linux/io.h>
+#include <linux/err.h>
+#include <linux/delay.h>
+
+#include "clk.h"
+
+/*
+ * Some clocks will have mutiple bits to enable the clocks, and
+ * the bits to disable the clock is not same as enabling bits.
+ */
+
+#define to_clk_mmp_gate(hw)	container_of(hw, struct mmp_clk_gate, hw)
+
+static int mmp_clk_gate_enable(struct clk_hw *hw)
+{
+	struct mmp_clk_gate *gate = to_clk_mmp_gate(hw);
+	struct clk *clk = hw->clk;
+	unsigned long flags = 0;
+	unsigned long rate;
+	u32 tmp;
+
+	if (gate->lock)
+		spin_lock_irqsave(gate->lock, flags);
+
+	tmp = readl(gate->reg);
+	tmp &= ~gate->mask;
+	tmp |= gate->val_enable;
+	writel(tmp, gate->reg);
+
+	if (gate->lock)
+		spin_unlock_irqrestore(gate->lock, flags);
+
+	if (gate->flags & MMP_CLK_GATE_NEED_DELAY) {
+		rate = __clk_get_rate(clk);
+		/* Need delay 2 cycles. */
+		udelay(2000000/rate);
+	}
+
+	return 0;
+}
+
+static void mmp_clk_gate_disable(struct clk_hw *hw)
+{
+	struct mmp_clk_gate *gate = to_clk_mmp_gate(hw);
+	unsigned long flags = 0;
+	u32 tmp;
+
+	if (gate->lock)
+		spin_lock_irqsave(gate->lock, flags);
+
+	tmp = readl(gate->reg);
+	tmp &= ~gate->mask;
+	tmp |= gate->val_disable;
+	writel(tmp, gate->reg);
+
+	if (gate->lock)
+		spin_unlock_irqrestore(gate->lock, flags);
+}
+
+static int mmp_clk_gate_is_enabled(struct clk_hw *hw)
+{
+	struct mmp_clk_gate *gate = to_clk_mmp_gate(hw);
+	unsigned long flags = 0;
+	u32 tmp;
+
+	if (gate->lock)
+		spin_lock_irqsave(gate->lock, flags);
+
+	tmp = readl(gate->reg);
+
+	if (gate->lock)
+		spin_unlock_irqrestore(gate->lock, flags);
+
+	return (tmp & gate->mask) == gate->val_enable;
+}
+
+const struct clk_ops mmp_clk_gate_ops = {
+	.enable = mmp_clk_gate_enable,
+	.disable = mmp_clk_gate_disable,
+	.is_enabled = mmp_clk_gate_is_enabled,
+};
+
+struct clk *mmp_clk_register_gate(struct device *dev, const char *name,
+		const char *parent_name, unsigned long flags,
+		void __iomem *reg, u32 mask, u32 val_enable, u32 val_disable,
+		unsigned int gate_flags, spinlock_t *lock)
+{
+	struct mmp_clk_gate *gate;
+	struct clk *clk;
+	struct clk_init_data init;
+
+	/* allocate the gate */
+	gate = kzalloc(sizeof(*gate), GFP_KERNEL);
+	if (!gate) {
+		pr_err("%s:%s could not allocate gate clk\n", __func__, name);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	init.name = name;
+	init.ops = &mmp_clk_gate_ops;
+	init.flags = flags | CLK_IS_BASIC;
+	init.parent_names = (parent_name ? &parent_name : NULL);
+	init.num_parents = (parent_name ? 1 : 0);
+
+	/* struct clk_gate assignments */
+	gate->reg = reg;
+	gate->mask = mask;
+	gate->val_enable = val_enable;
+	gate->val_disable = val_disable;
+	gate->flags = gate_flags;
+	gate->lock = lock;
+	gate->hw.init = &init;
+
+	clk = clk_register(dev, &gate->hw);
+
+	if (IS_ERR(clk))
+		kfree(gate);
+
+	return clk;
+}
diff --git a/drivers/clk/mmp/clk-mix.c b/drivers/clk/mmp/clk-mix.c
new file mode 100644
index 000000000000..48fa53c7ce5e
--- /dev/null
+++ b/drivers/clk/mmp/clk-mix.c
@@ -0,0 +1,513 @@
+/*
+ * mmp mix(div and mux) clock operation source file
+ *
+ * Copyright (C) 2014 Marvell
+ * Chao Xie <chao.xie@marvell.com>
+ *
+ * This file is licensed under the terms of the GNU General Public
+ * License version 2. This program is licensed "as is" without any
+ * warranty of any kind, whether express or implied.
+ */
+
+#include <linux/clk-provider.h>
+#include <linux/slab.h>
+#include <linux/io.h>
+#include <linux/err.h>
+
+#include "clk.h"
+
+/*
+ * The mix clock is a clock combined mux and div type clock.
+ * Because the div field and mux field need to be set at same
+ * time, we can not divide it into 2 types of clock
+ */
+
+#define to_clk_mix(hw)	container_of(hw, struct mmp_clk_mix, hw)
+
+static unsigned int _get_maxdiv(struct mmp_clk_mix *mix)
+{
+	unsigned int div_mask = (1 << mix->reg_info.width_div) - 1;
+	unsigned int maxdiv = 0;
+	struct clk_div_table *clkt;
+
+	if (mix->div_flags & CLK_DIVIDER_ONE_BASED)
+		return div_mask;
+	if (mix->div_flags & CLK_DIVIDER_POWER_OF_TWO)
+		return 1 << div_mask;
+	if (mix->div_table) {
+		for (clkt = mix->div_table; clkt->div; clkt++)
+			if (clkt->div > maxdiv)
+				maxdiv = clkt->div;
+		return maxdiv;
+	}
+	return div_mask + 1;
+}
+
+static unsigned int _get_div(struct mmp_clk_mix *mix, unsigned int val)
+{
+	struct clk_div_table *clkt;
+
+	if (mix->div_flags & CLK_DIVIDER_ONE_BASED)
+		return val;
+	if (mix->div_flags & CLK_DIVIDER_POWER_OF_TWO)
+		return 1 << val;
+	if (mix->div_table) {
+		for (clkt = mix->div_table; clkt->div; clkt++)
+			if (clkt->val == val)
+				return clkt->div;
+		if (clkt->div == 0)
+			return 0;
+	}
+	return val + 1;
+}
+
+static unsigned int _get_mux(struct mmp_clk_mix *mix, unsigned int val)
+{
+	int num_parents = __clk_get_num_parents(mix->hw.clk);
+	int i;
+
+	if (mix->mux_flags & CLK_MUX_INDEX_BIT)
+		return ffs(val) - 1;
+	if (mix->mux_flags & CLK_MUX_INDEX_ONE)
+		return val - 1;
+	if (mix->mux_table) {
+		for (i = 0; i < num_parents; i++)
+			if (mix->mux_table[i] == val)
+				return i;
+		if (i == num_parents)
+			return 0;
+	}
+
+	return val;
+}
+static unsigned int _get_div_val(struct mmp_clk_mix *mix, unsigned int div)
+{
+	struct clk_div_table *clkt;
+
+	if (mix->div_flags & CLK_DIVIDER_ONE_BASED)
+		return div;
+	if (mix->div_flags & CLK_DIVIDER_POWER_OF_TWO)
+		return __ffs(div);
+	if (mix->div_table) {
+		for (clkt = mix->div_table; clkt->div; clkt++)
+			if (clkt->div == div)
+				return clkt->val;
+		if (clkt->div == 0)
+			return 0;
+	}
+
+	return div - 1;
+}
+
+static unsigned int _get_mux_val(struct mmp_clk_mix *mix, unsigned int mux)
+{
+	if (mix->mux_table)
+		return mix->mux_table[mux];
+
+	return mux;
+}
+
+static void _filter_clk_table(struct mmp_clk_mix *mix,
+				struct mmp_clk_mix_clk_table *table,
+				unsigned int table_size)
+{
+	int i;
+	struct mmp_clk_mix_clk_table *item;
+	struct clk *parent, *clk;
+	unsigned long parent_rate;
+
+	clk = mix->hw.clk;
+
+	for (i = 0; i < table_size; i++) {
+		item = &table[i];
+		parent = clk_get_parent_by_index(clk, item->parent_index);
+		parent_rate = __clk_get_rate(parent);
+		if (parent_rate % item->rate) {
+			item->valid = 0;
+		} else {
+			item->divisor = parent_rate / item->rate;
+			item->valid = 1;
+		}
+	}
+}
+
+static int _set_rate(struct mmp_clk_mix *mix, u32 mux_val, u32 div_val,
+			unsigned int change_mux, unsigned int change_div)
+{
+	struct mmp_clk_mix_reg_info *ri = &mix->reg_info;
+	u8 width, shift;
+	u32 mux_div, fc_req;
+	int ret, timeout = 50;
+	unsigned long flags = 0;
+
+	if (!change_mux && !change_div)
+		return -EINVAL;
+
+	if (mix->lock)
+		spin_lock_irqsave(mix->lock, flags);
+
+	if (mix->type == MMP_CLK_MIX_TYPE_V1
+		|| mix->type == MMP_CLK_MIX_TYPE_V2)
+		mux_div = readl(ri->reg_clk_ctrl);
+	else
+		mux_div = readl(ri->reg_clk_sel);
+
+	if (change_div) {
+		width = ri->width_div;
+		shift = ri->shift_div;
+		mux_div &= ~MMP_CLK_BITS_MASK(width, shift);
+		mux_div |= MMP_CLK_BITS_SET_VAL(div_val, width, shift);
+	}
+
+	if (change_mux) {
+		width = ri->width_mux;
+		shift = ri->shift_mux;
+		mux_div &= ~MMP_CLK_BITS_MASK(width, shift);
+		mux_div |= MMP_CLK_BITS_SET_VAL(mux_val, width, shift);
+	}
+
+	if (mix->type == MMP_CLK_MIX_TYPE_V1) {
+		writel(mux_div, ri->reg_clk_ctrl);
+	} else if (mix->type == MMP_CLK_MIX_TYPE_V2) {
+		mux_div |= (1 << ri->bit_fc);
+		writel(mux_div, ri->reg_clk_ctrl);
+
+		do {
+			fc_req = readl(ri->reg_clk_ctrl);
+			timeout--;
+			if (!(fc_req & (1 << ri->bit_fc)))
+				break;
+		} while (timeout);
+
+		if (timeout == 0) {
+			pr_err("%s:%s cannot do frequency change\n",
+				__func__, __clk_get_name(mix->hw.clk));
+			ret = -EBUSY;
+			goto error;
+		}
+	} else {
+		fc_req = readl(ri->reg_clk_ctrl);
+		fc_req |= 1 << ri->bit_fc;
+		writel(fc_req, ri->reg_clk_ctrl);
+		writel(mux_div, ri->reg_clk_sel);
+		fc_req &= ~(1 << ri->bit_fc);
+	}
+
+	ret = 0;
+error:
+	if (mix->lock)
+		spin_unlock_irqrestore(mix->lock, flags);
+
+	return ret;
+}
+
+static long mmp_clk_mix_determine_rate(struct clk_hw *hw, unsigned long rate,
+					unsigned long *best_parent_rate,
+					struct clk_hw **best_parent_clk)
+{
+	struct mmp_clk_mix *mix = to_clk_mix(hw);
+	struct mmp_clk_mix_clk_table *item;
+	struct clk *parent, *parent_best, *mix_clk;
+	unsigned long parent_rate, mix_rate, mix_rate_best, parent_rate_best;
+	unsigned long gap, gap_best;
+	u32 div_val_max;
+	unsigned int div;
+	int i, j;
+
+	mix_clk = hw->clk;
+
+	parent = NULL;
+	mix_rate_best = 0;
+	parent_rate_best = 0;
+	gap_best = rate;
+	parent_best = NULL;
+
+	if (mix->table) {
+		for (i = 0; i < mix->table_size; i++) {
+			item = &mix->table[i];
+			if (item->valid == 0)
+				continue;
+			parent = clk_get_parent_by_index(mix_clk,
+							item->parent_index);
+			parent_rate = __clk_get_rate(parent);
+			mix_rate = parent_rate / item->divisor;
+			gap = abs(mix_rate - rate);
+			if (parent_best == NULL || gap < gap_best) {
+				parent_best = parent;
+				parent_rate_best = parent_rate;
+				mix_rate_best = mix_rate;
+				gap_best = gap;
+				if (gap_best == 0)
+					goto found;
+			}
+		}
+	} else {
+		for (i = 0; i < __clk_get_num_parents(mix_clk); i++) {
+			parent = clk_get_parent_by_index(mix_clk, i);
+			parent_rate = __clk_get_rate(parent);
+			div_val_max = _get_maxdiv(mix);
+			for (j = 0; j < div_val_max; j++) {
+				div = _get_div(mix, j);
+				mix_rate = parent_rate / div;
+				gap = abs(mix_rate - rate);
+				if (parent_best == NULL || gap < gap_best) {
+					parent_best = parent;
+					parent_rate_best = parent_rate;
+					mix_rate_best = mix_rate;
+					gap_best = gap;
+					if (gap_best == 0)
+						goto found;
+				}
+			}
+		}
+	}
+
+found:
+	*best_parent_rate = parent_rate_best;
+	*best_parent_clk = __clk_get_hw(parent_best);
+
+	return mix_rate_best;
+}
+
+static int mmp_clk_mix_set_rate_and_parent(struct clk_hw *hw,
+						unsigned long rate,
+						unsigned long parent_rate,
+						u8 index)
+{
+	struct mmp_clk_mix *mix = to_clk_mix(hw);
+	unsigned int div;
+	u32 div_val, mux_val;
+
+	div = parent_rate / rate;
+	div_val = _get_div_val(mix, div);
+	mux_val = _get_mux_val(mix, index);
+
+	return _set_rate(mix, mux_val, div_val, 1, 1);
+}
+
+static u8 mmp_clk_mix_get_parent(struct clk_hw *hw)
+{
+	struct mmp_clk_mix *mix = to_clk_mix(hw);
+	struct mmp_clk_mix_reg_info *ri = &mix->reg_info;
+	unsigned long flags = 0;
+	u32 mux_div = 0;
+	u8 width, shift;
+	u32 mux_val;
+
+	if (mix->lock)
+		spin_lock_irqsave(mix->lock, flags);
+
+	if (mix->type == MMP_CLK_MIX_TYPE_V1
+		|| mix->type == MMP_CLK_MIX_TYPE_V2)
+		mux_div = readl(ri->reg_clk_ctrl);
+	else
+		mux_div = readl(ri->reg_clk_sel);
+
+	if (mix->lock)
+		spin_unlock_irqrestore(mix->lock, flags);
+
+	width = mix->reg_info.width_mux;
+	shift = mix->reg_info.shift_mux;
+
+	mux_val = MMP_CLK_BITS_GET_VAL(mux_div, width, shift);
+
+	return _get_mux(mix, mux_val);
+}
+
+static unsigned long mmp_clk_mix_recalc_rate(struct clk_hw *hw,
+					unsigned long parent_rate)
+{
+	struct mmp_clk_mix *mix = to_clk_mix(hw);
+	struct mmp_clk_mix_reg_info *ri = &mix->reg_info;
+	unsigned long flags = 0;
+	u32 mux_div = 0;
+	u8 width, shift;
+	unsigned int div;
+
+	if (mix->lock)
+		spin_lock_irqsave(mix->lock, flags);
+
+	if (mix->type == MMP_CLK_MIX_TYPE_V1
+		|| mix->type == MMP_CLK_MIX_TYPE_V2)
+		mux_div = readl(ri->reg_clk_ctrl);
+	else
+		mux_div = readl(ri->reg_clk_sel);
+
+	if (mix->lock)
+		spin_unlock_irqrestore(mix->lock, flags);
+
+	width = mix->reg_info.width_div;
+	shift = mix->reg_info.shift_div;
+
+	div = _get_div(mix, MMP_CLK_BITS_GET_VAL(mux_div, width, shift));
+
+	return parent_rate / div;
+}
+
+static int mmp_clk_set_parent(struct clk_hw *hw, u8 index)
+{
+	struct mmp_clk_mix *mix = to_clk_mix(hw);
+	struct mmp_clk_mix_clk_table *item;
+	int i;
+	u32 div_val, mux_val;
+
+	if (mix->table) {
+		for (i = 0; i < mix->table_size; i++) {
+			item = &mix->table[i];
+			if (item->valid == 0)
+				continue;
+			if (item->parent_index == index)
+				break;
+		}
+		if (i < mix->table_size) {
+			div_val = _get_div_val(mix, item->divisor);
+			mux_val = _get_mux_val(mix, item->parent_index);
+		} else
+			return -EINVAL;
+	} else {
+		mux_val = _get_mux_val(mix, index);
+		div_val = 0;
+	}
+
+	return _set_rate(mix, mux_val, div_val, 1, div_val ? 1 : 0);
+}
+
+static int mmp_clk_set_rate(struct clk_hw *hw, unsigned long rate,
+				unsigned long best_parent_rate)
+{
+	struct mmp_clk_mix *mix = to_clk_mix(hw);
+	struct mmp_clk_mix_clk_table *item;
+	unsigned long parent_rate;
+	unsigned int best_divisor;
+	struct clk *mix_clk, *parent;
+	int i;
+
+	best_divisor = best_parent_rate / rate;
+
+	mix_clk = hw->clk;
+	if (mix->table) {
+		for (i = 0; i < mix->table_size; i++) {
+			item = &mix->table[i];
+			if (item->valid == 0)
+				continue;
+			parent = clk_get_parent_by_index(mix_clk,
+							item->parent_index);
+			parent_rate = __clk_get_rate(parent);
+			if (parent_rate == best_parent_rate
+				&& item->divisor == best_divisor)
+				break;
+		}
+		if (i < mix->table_size)
+			return _set_rate(mix,
+					_get_mux_val(mix, item->parent_index),
+					_get_div_val(mix, item->divisor),
+					1, 1);
+		else
+			return -EINVAL;
+	} else {
+		for (i = 0; i < __clk_get_num_parents(mix_clk); i++) {
+			parent = clk_get_parent_by_index(mix_clk, i);
+			parent_rate = __clk_get_rate(parent);
+			if (parent_rate == best_parent_rate)
+				break;
+		}
+		if (i < __clk_get_num_parents(mix_clk))
+			return _set_rate(mix, _get_mux_val(mix, i),
+					_get_div_val(mix, best_divisor), 1, 1);
+		else
+			return -EINVAL;
+	}
+}
+
+static void mmp_clk_mix_init(struct clk_hw *hw)
+{
+	struct mmp_clk_mix *mix = to_clk_mix(hw);
+
+	if (mix->table)
+		_filter_clk_table(mix, mix->table, mix->table_size);
+}
+
+const struct clk_ops mmp_clk_mix_ops = {
+	.determine_rate = mmp_clk_mix_determine_rate,
+	.set_rate_and_parent = mmp_clk_mix_set_rate_and_parent,
+	.set_rate = mmp_clk_set_rate,
+	.set_parent = mmp_clk_set_parent,
+	.get_parent = mmp_clk_mix_get_parent,
+	.recalc_rate = mmp_clk_mix_recalc_rate,
+	.init = mmp_clk_mix_init,
+};
+
+struct clk *mmp_clk_register_mix(struct device *dev,
+					const char *name,
+					const char **parent_names,
+					u8 num_parents,
+					unsigned long flags,
+					struct mmp_clk_mix_config *config,
+					spinlock_t *lock)
+{
+	struct mmp_clk_mix *mix;
+	struct clk *clk;
+	struct clk_init_data init;
+	size_t table_bytes;
+
+	mix = kzalloc(sizeof(*mix), GFP_KERNEL);
+	if (!mix) {
+		pr_err("%s:%s: could not allocate mmp mix clk\n",
+			__func__, name);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	init.name = name;
+	init.flags = flags | CLK_GET_RATE_NOCACHE;
+	init.parent_names = parent_names;
+	init.num_parents = num_parents;
+	init.ops = &mmp_clk_mix_ops;
+
+	memcpy(&mix->reg_info, &config->reg_info, sizeof(config->reg_info));
+	if (config->table) {
+		table_bytes = sizeof(*config->table) * config->table_size;
+		mix->table = kzalloc(table_bytes, GFP_KERNEL);
+		if (!mix->table) {
+			pr_err("%s:%s: could not allocate mmp mix table\n",
+				__func__, name);
+			kfree(mix);
+			return ERR_PTR(-ENOMEM);
+		}
+		memcpy(mix->table, config->table, table_bytes);
+		mix->table_size = config->table_size;
+	}
+
+	if (config->mux_table) {
+		table_bytes = sizeof(u32) * num_parents;
+		mix->mux_table = kzalloc(table_bytes, GFP_KERNEL);
+		if (!mix->mux_table) {
+			pr_err("%s:%s: could not allocate mmp mix mux-table\n",
+				__func__, name);
+			kfree(mix->table);
+			kfree(mix);
+			return ERR_PTR(-ENOMEM);
+		}
+		memcpy(mix->mux_table, config->mux_table, table_bytes);
+	}
+
+	mix->div_flags = config->div_flags;
+	mix->mux_flags = config->mux_flags;
+	mix->lock = lock;
+	mix->hw.init = &init;
+
+	if (config->reg_info.bit_fc >= 32)
+		mix->type = MMP_CLK_MIX_TYPE_V1;
+	else if (config->reg_info.reg_clk_sel)
+		mix->type = MMP_CLK_MIX_TYPE_V3;
+	else
+		mix->type = MMP_CLK_MIX_TYPE_V2;
+	clk = clk_register(dev, &mix->hw);
+
+	if (IS_ERR(clk)) {
+		kfree(mix->mux_table);
+		kfree(mix->table);
+		kfree(mix);
+	}
+
+	return clk;
+}
diff --git a/drivers/clk/mmp/clk-mmp2.c b/drivers/clk/mmp/clk-mmp2.c
index b2721cae257a..5c90a4230fa3 100644
--- a/drivers/clk/mmp/clk-mmp2.c
+++ b/drivers/clk/mmp/clk-mmp2.c
@@ -54,7 +54,7 @@
 
 static DEFINE_SPINLOCK(clk_lock);
 
-static struct clk_factor_masks uart_factor_masks = {
+static struct mmp_clk_factor_masks uart_factor_masks = {
 	.factor = 2,
 	.num_mask = 0x1fff,
 	.den_mask = 0x1fff,
@@ -62,7 +62,7 @@ static struct clk_factor_masks uart_factor_masks = {
 	.den_shift = 0,
 };
 
-static struct clk_factor_tbl uart_factor_tbl[] = {
+static struct mmp_clk_factor_tbl uart_factor_tbl[] = {
 	{.num = 14634, .den = 2165},	/*14.745MHZ */
 	{.num = 3521, .den = 689},	/*19.23MHZ */
 	{.num = 9679, .den = 5728},	/*58.9824MHZ */
@@ -191,7 +191,7 @@ void __init mmp2_clk_init(void)
 	clk = mmp_clk_register_factor("uart_pll", "pll1_4", 0,
 				mpmu_base + MPMU_UART_PLL,
 				&uart_factor_masks, uart_factor_tbl,
-				ARRAY_SIZE(uart_factor_tbl));
+				ARRAY_SIZE(uart_factor_tbl), &clk_lock);
 	clk_set_rate(clk, 14745600);
 	clk_register_clkdev(clk, "uart_pll", NULL);
 
diff --git a/drivers/clk/mmp/clk-of-mmp2.c b/drivers/clk/mmp/clk-of-mmp2.c
new file mode 100644
index 000000000000..2cbc2b43ae52
--- /dev/null
+++ b/drivers/clk/mmp/clk-of-mmp2.c
@@ -0,0 +1,334 @@
+/*
+ * mmp2 clock framework source file
+ *
+ * Copyright (C) 2012 Marvell
+ * Chao Xie <xiechao.mail@gmail.com>
+ *
+ * This file is licensed under the terms of the GNU General Public
+ * License version 2. This program is licensed "as is" without any
+ * warranty of any kind, whether express or implied.
+ */
+
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/spinlock.h>
+#include <linux/io.h>
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/of_address.h>
+
+#include <dt-bindings/clock/marvell,mmp2.h>
+
+#include "clk.h"
+#include "reset.h"
+
+#define APBC_RTC	0x0
+#define APBC_TWSI0	0x4
+#define APBC_TWSI1	0x8
+#define APBC_TWSI2	0xc
+#define APBC_TWSI3	0x10
+#define APBC_TWSI4	0x7c
+#define APBC_TWSI5	0x80
+#define APBC_KPC	0x18
+#define APBC_UART0	0x2c
+#define APBC_UART1	0x30
+#define APBC_UART2	0x34
+#define APBC_UART3	0x88
+#define APBC_GPIO	0x38
+#define APBC_PWM0	0x3c
+#define APBC_PWM1	0x40
+#define APBC_PWM2	0x44
+#define APBC_PWM3	0x48
+#define APBC_SSP0	0x50
+#define APBC_SSP1	0x54
+#define APBC_SSP2	0x58
+#define APBC_SSP3	0x5c
+#define APMU_SDH0	0x54
+#define APMU_SDH1	0x58
+#define APMU_SDH2	0xe8
+#define APMU_SDH3	0xec
+#define APMU_USB	0x5c
+#define APMU_DISP0	0x4c
+#define APMU_DISP1	0x110
+#define APMU_CCIC0	0x50
+#define APMU_CCIC1	0xf4
+#define MPMU_UART_PLL	0x14
+
+struct mmp2_clk_unit {
+	struct mmp_clk_unit unit;
+	void __iomem *mpmu_base;
+	void __iomem *apmu_base;
+	void __iomem *apbc_base;
+};
+
+static struct mmp_param_fixed_rate_clk fixed_rate_clks[] = {
+	{MMP2_CLK_CLK32, "clk32", NULL, CLK_IS_ROOT, 32768},
+	{MMP2_CLK_VCTCXO, "vctcxo", NULL, CLK_IS_ROOT, 26000000},
+	{MMP2_CLK_PLL1, "pll1", NULL, CLK_IS_ROOT, 800000000},
+	{MMP2_CLK_PLL2, "pll2", NULL, CLK_IS_ROOT, 960000000},
+	{MMP2_CLK_USB_PLL, "usb_pll", NULL, CLK_IS_ROOT, 480000000},
+};
+
+static struct mmp_param_fixed_factor_clk fixed_factor_clks[] = {
+	{MMP2_CLK_PLL1_2, "pll1_2", "pll1", 1, 2, 0},
+	{MMP2_CLK_PLL1_4, "pll1_4", "pll1_2", 1, 2, 0},
+	{MMP2_CLK_PLL1_8, "pll1_8", "pll1_4", 1, 2, 0},
+	{MMP2_CLK_PLL1_16, "pll1_16", "pll1_8", 1, 2, 0},
+	{MMP2_CLK_PLL1_20, "pll1_20", "pll1_4", 1, 5, 0},
+	{MMP2_CLK_PLL1_3, "pll1_3", "pll1", 1, 3, 0},
+	{MMP2_CLK_PLL1_6, "pll1_6", "pll1_3", 1, 2, 0},
+	{MMP2_CLK_PLL1_12, "pll1_12", "pll1_6", 1, 2, 0},
+	{MMP2_CLK_PLL2_2, "pll2_2", "pll2", 1, 2, 0},
+	{MMP2_CLK_PLL2_4, "pll2_4", "pll2_2", 1, 2, 0},
+	{MMP2_CLK_PLL2_8, "pll2_8", "pll2_4", 1, 2, 0},
+	{MMP2_CLK_PLL2_16, "pll2_16", "pll2_8", 1, 2, 0},
+	{MMP2_CLK_PLL2_3, "pll2_3", "pll2", 1, 3, 0},
+	{MMP2_CLK_PLL2_6, "pll2_6", "pll2_3", 1, 2, 0},
+	{MMP2_CLK_PLL2_12, "pll2_12", "pll2_6", 1, 2, 0},
+	{MMP2_CLK_VCTCXO_2, "vctcxo_2", "vctcxo", 1, 2, 0},
+	{MMP2_CLK_VCTCXO_4, "vctcxo_4", "vctcxo_2", 1, 2, 0},
+};
+
+static struct mmp_clk_factor_masks uart_factor_masks = {
+	.factor = 2,
+	.num_mask = 0x1fff,
+	.den_mask = 0x1fff,
+	.num_shift = 16,
+	.den_shift = 0,
+};
+
+static struct mmp_clk_factor_tbl uart_factor_tbl[] = {
+	{.num = 14634, .den = 2165},	/*14.745MHZ */
+	{.num = 3521, .den = 689},	/*19.23MHZ */
+	{.num = 9679, .den = 5728},	/*58.9824MHZ */
+	{.num = 15850, .den = 9451},	/*59.429MHZ */
+};
+
+static void mmp2_pll_init(struct mmp2_clk_unit *pxa_unit)
+{
+	struct clk *clk;
+	struct mmp_clk_unit *unit = &pxa_unit->unit;
+
+	mmp_register_fixed_rate_clks(unit, fixed_rate_clks,
+					ARRAY_SIZE(fixed_rate_clks));
+
+	mmp_register_fixed_factor_clks(unit, fixed_factor_clks,
+					ARRAY_SIZE(fixed_factor_clks));
+
+	clk = mmp_clk_register_factor("uart_pll", "pll1_4",
+				CLK_SET_RATE_PARENT,
+				pxa_unit->mpmu_base + MPMU_UART_PLL,
+				&uart_factor_masks, uart_factor_tbl,
+				ARRAY_SIZE(uart_factor_tbl), NULL);
+	mmp_clk_add(unit, MMP2_CLK_UART_PLL, clk);
+}
+
+static DEFINE_SPINLOCK(uart0_lock);
+static DEFINE_SPINLOCK(uart1_lock);
+static DEFINE_SPINLOCK(uart2_lock);
+static const char *uart_parent_names[] = {"uart_pll", "vctcxo"};
+
+static DEFINE_SPINLOCK(ssp0_lock);
+static DEFINE_SPINLOCK(ssp1_lock);
+static DEFINE_SPINLOCK(ssp2_lock);
+static DEFINE_SPINLOCK(ssp3_lock);
+static const char *ssp_parent_names[] = {"vctcxo_4", "vctcxo_2", "vctcxo", "pll1_16"};
+
+static DEFINE_SPINLOCK(reset_lock);
+
+static struct mmp_param_mux_clk apbc_mux_clks[] = {
+	{0, "uart0_mux", uart_parent_names, ARRAY_SIZE(uart_parent_names), CLK_SET_RATE_PARENT, APBC_UART0, 4, 3, 0, &uart0_lock},
+	{0, "uart1_mux", uart_parent_names, ARRAY_SIZE(uart_parent_names), CLK_SET_RATE_PARENT, APBC_UART1, 4, 3, 0, &uart1_lock},
+	{0, "uart2_mux", uart_parent_names, ARRAY_SIZE(uart_parent_names), CLK_SET_RATE_PARENT, APBC_UART2, 4, 3, 0, &uart2_lock},
+	{0, "uart3_mux", uart_parent_names, ARRAY_SIZE(uart_parent_names), CLK_SET_RATE_PARENT, APBC_UART3, 4, 3, 0, &uart2_lock},
+	{0, "ssp0_mux", ssp_parent_names, ARRAY_SIZE(ssp_parent_names), CLK_SET_RATE_PARENT, APBC_SSP0, 4, 3, 0, &ssp0_lock},
+	{0, "ssp1_mux", ssp_parent_names, ARRAY_SIZE(ssp_parent_names), CLK_SET_RATE_PARENT, APBC_SSP1, 4, 3, 0, &ssp1_lock},
+	{0, "ssp2_mux", ssp_parent_names, ARRAY_SIZE(ssp_parent_names), CLK_SET_RATE_PARENT, APBC_SSP2, 4, 3, 0, &ssp2_lock},
+	{0, "ssp3_mux", ssp_parent_names, ARRAY_SIZE(ssp_parent_names), CLK_SET_RATE_PARENT, APBC_SSP3, 4, 3, 0, &ssp3_lock},
+};
+
+static struct mmp_param_gate_clk apbc_gate_clks[] = {
+	{MMP2_CLK_TWSI0, "twsi0_clk", "vctcxo", CLK_SET_RATE_PARENT, APBC_TWSI0, 0x7, 0x3, 0x0, 0, &reset_lock},
+	{MMP2_CLK_TWSI1, "twsi1_clk", "vctcxo", CLK_SET_RATE_PARENT, APBC_TWSI1, 0x7, 0x3, 0x0, 0, &reset_lock},
+	{MMP2_CLK_TWSI2, "twsi2_clk", "vctcxo", CLK_SET_RATE_PARENT, APBC_TWSI2, 0x7, 0x3, 0x0, 0, &reset_lock},
+	{MMP2_CLK_TWSI3, "twsi3_clk", "vctcxo", CLK_SET_RATE_PARENT, APBC_TWSI3, 0x7, 0x3, 0x0, 0, &reset_lock},
+	{MMP2_CLK_TWSI4, "twsi4_clk", "vctcxo", CLK_SET_RATE_PARENT, APBC_TWSI4, 0x7, 0x3, 0x0, 0, &reset_lock},
+	{MMP2_CLK_TWSI5, "twsi5_clk", "vctcxo", CLK_SET_RATE_PARENT, APBC_TWSI5, 0x7, 0x3, 0x0, 0, &reset_lock},
+	{MMP2_CLK_GPIO, "gpio_clk", "vctcxo", CLK_SET_RATE_PARENT, APBC_GPIO, 0x7, 0x3, 0x0, 0, &reset_lock},
+	{MMP2_CLK_KPC, "kpc_clk", "clk32", CLK_SET_RATE_PARENT, APBC_KPC, 0x7, 0x3, 0x0, MMP_CLK_GATE_NEED_DELAY, &reset_lock},
+	{MMP2_CLK_RTC, "rtc_clk", "clk32", CLK_SET_RATE_PARENT, APBC_RTC, 0x87, 0x83, 0x0, MMP_CLK_GATE_NEED_DELAY, &reset_lock},
+	{MMP2_CLK_PWM0, "pwm0_clk", "pll1_48", CLK_SET_RATE_PARENT, APBC_PWM0, 0x7, 0x3, 0x0, 0, &reset_lock},
+	{MMP2_CLK_PWM1, "pwm1_clk", "pll1_48", CLK_SET_RATE_PARENT, APBC_PWM1, 0x7, 0x3, 0x0, 0, &reset_lock},
+	{MMP2_CLK_PWM2, "pwm2_clk", "pll1_48", CLK_SET_RATE_PARENT, APBC_PWM2, 0x7, 0x3, 0x0, 0, &reset_lock},
+	{MMP2_CLK_PWM3, "pwm3_clk", "pll1_48", CLK_SET_RATE_PARENT, APBC_PWM3, 0x7, 0x3, 0x0, 0, &reset_lock},
+	/* The gate clocks has mux parent. */
+	{MMP2_CLK_UART0, "uart0_clk", "uart0_mux", CLK_SET_RATE_PARENT, APBC_UART0, 0x7, 0x3, 0x0, 0, &uart0_lock},
+	{MMP2_CLK_UART1, "uart1_clk", "uart1_mux", CLK_SET_RATE_PARENT, APBC_UART1, 0x7, 0x3, 0x0, 0, &uart1_lock},
+	{MMP2_CLK_UART2, "uart2_clk", "uart2_mux", CLK_SET_RATE_PARENT, APBC_UART2, 0x7, 0x3, 0x0, 0, &uart2_lock},
+	{MMP2_CLK_UART3, "uart3_clk", "uart3_mux", CLK_SET_RATE_PARENT, APBC_UART3, 0x7, 0x3, 0x0, 0, &uart2_lock},
+	{MMP2_CLK_SSP0, "ssp0_clk", "ssp0_mux", CLK_SET_RATE_PARENT, APBC_SSP0, 0x7, 0x3, 0x0, 0, &ssp0_lock},
+	{MMP2_CLK_SSP1, "ssp1_clk", "ssp1_mux", CLK_SET_RATE_PARENT, APBC_SSP1, 0x7, 0x3, 0x0, 0, &ssp1_lock},
+	{MMP2_CLK_SSP2, "ssp2_clk", "ssp2_mux", CLK_SET_RATE_PARENT, APBC_SSP2, 0x7, 0x3, 0x0, 0, &ssp2_lock},
+	{MMP2_CLK_SSP3, "ssp3_clk", "ssp3_mux", CLK_SET_RATE_PARENT, APBC_SSP3, 0x7, 0x3, 0x0, 0, &ssp3_lock},
+};
+
+static void mmp2_apb_periph_clk_init(struct mmp2_clk_unit *pxa_unit)
+{
+	struct mmp_clk_unit *unit = &pxa_unit->unit;
+
+	mmp_register_mux_clks(unit, apbc_mux_clks, pxa_unit->apbc_base,
+				ARRAY_SIZE(apbc_mux_clks));
+
+	mmp_register_gate_clks(unit, apbc_gate_clks, pxa_unit->apbc_base,
+				ARRAY_SIZE(apbc_gate_clks));
+}
+
+static DEFINE_SPINLOCK(sdh_lock);
+static const char *sdh_parent_names[] = {"pll1_4", "pll2", "usb_pll", "pll1"};
+static struct mmp_clk_mix_config sdh_mix_config = {
+	.reg_info = DEFINE_MIX_REG_INFO(4, 10, 2, 8, 32),
+};
+
+static DEFINE_SPINLOCK(usb_lock);
+
+static DEFINE_SPINLOCK(disp0_lock);
+static DEFINE_SPINLOCK(disp1_lock);
+static const char *disp_parent_names[] = {"pll1", "pll1_16", "pll2", "vctcxo"};
+
+static DEFINE_SPINLOCK(ccic0_lock);
+static DEFINE_SPINLOCK(ccic1_lock);
+static const char *ccic_parent_names[] = {"pll1_2", "pll1_16", "vctcxo"};
+static struct mmp_clk_mix_config ccic0_mix_config = {
+	.reg_info = DEFINE_MIX_REG_INFO(4, 17, 2, 6, 32),
+};
+static struct mmp_clk_mix_config ccic1_mix_config = {
+	.reg_info = DEFINE_MIX_REG_INFO(4, 16, 2, 6, 32),
+};
+
+static struct mmp_param_mux_clk apmu_mux_clks[] = {
+	{MMP2_CLK_DISP0_MUX, "disp0_mux", disp_parent_names, ARRAY_SIZE(disp_parent_names), CLK_SET_RATE_PARENT, APMU_DISP0, 6, 2, 0, &disp0_lock},
+	{MMP2_CLK_DISP1_MUX, "disp1_mux", disp_parent_names, ARRAY_SIZE(disp_parent_names), CLK_SET_RATE_PARENT, APMU_DISP1, 6, 2, 0, &disp1_lock},
+};
+
+static struct mmp_param_div_clk apmu_div_clks[] = {
+	{0, "disp0_div", "disp0_mux", CLK_SET_RATE_PARENT, APMU_DISP0, 8, 4, 0, &disp0_lock},
+	{0, "disp0_sphy_div", "disp0_mux", CLK_SET_RATE_PARENT, APMU_DISP0, 15, 5, 0, &disp0_lock},
+	{0, "disp1_div", "disp1_mux", CLK_SET_RATE_PARENT, APMU_DISP1, 8, 4, 0, &disp1_lock},
+	{0, "ccic0_sphy_div", "ccic0_mix_clk", CLK_SET_RATE_PARENT, APMU_CCIC0, 10, 5, 0, &ccic0_lock},
+	{0, "ccic1_sphy_div", "ccic1_mix_clk", CLK_SET_RATE_PARENT, APMU_CCIC1, 10, 5, 0, &ccic1_lock},
+};
+
+static struct mmp_param_gate_clk apmu_gate_clks[] = {
+	{MMP2_CLK_USB, "usb_clk", "usb_pll", 0, APMU_USB, 0x9, 0x9, 0x0, 0, &usb_lock},
+	/* The gate clocks has mux parent. */
+	{MMP2_CLK_SDH0, "sdh0_clk", "sdh_mix_clk", CLK_SET_RATE_PARENT, APMU_SDH0, 0x1b, 0x1b, 0x0, 0, &sdh_lock},
+	{MMP2_CLK_SDH1, "sdh1_clk", "sdh_mix_clk", CLK_SET_RATE_PARENT, APMU_SDH1, 0x1b, 0x1b, 0x0, 0, &sdh_lock},
+	{MMP2_CLK_SDH1, "sdh2_clk", "sdh_mix_clk", CLK_SET_RATE_PARENT, APMU_SDH2, 0x1b, 0x1b, 0x0, 0, &sdh_lock},
+	{MMP2_CLK_SDH1, "sdh3_clk", "sdh_mix_clk", CLK_SET_RATE_PARENT, APMU_SDH3, 0x1b, 0x1b, 0x0, 0, &sdh_lock},
+	{MMP2_CLK_DISP0, "disp0_clk", "disp0_div", CLK_SET_RATE_PARENT, APMU_DISP0, 0x1b, 0x1b, 0x0, 0, &disp0_lock},
+	{MMP2_CLK_DISP0_SPHY, "disp0_sphy_clk", "disp0_sphy_div", CLK_SET_RATE_PARENT, APMU_DISP0, 0x1024, 0x1024, 0x0, 0, &disp0_lock},
+	{MMP2_CLK_DISP1, "disp1_clk", "disp1_div", CLK_SET_RATE_PARENT, APMU_DISP1, 0x1b, 0x1b, 0x0, 0, &disp1_lock},
+	{MMP2_CLK_CCIC_ARBITER, "ccic_arbiter", "vctcxo", CLK_SET_RATE_PARENT, APMU_CCIC0, 0x1800, 0x1800, 0x0, 0, &ccic0_lock},
+	{MMP2_CLK_CCIC0, "ccic0_clk", "ccic0_mix_clk", CLK_SET_RATE_PARENT, APMU_CCIC0, 0x1b, 0x1b, 0x0, 0, &ccic0_lock},
+	{MMP2_CLK_CCIC0_PHY, "ccic0_phy_clk", "ccic0_mix_clk", CLK_SET_RATE_PARENT, APMU_CCIC0, 0x24, 0x24, 0x0, 0, &ccic0_lock},
+	{MMP2_CLK_CCIC0_SPHY, "ccic0_sphy_clk", "ccic0_sphy_div", CLK_SET_RATE_PARENT, APMU_CCIC0, 0x300, 0x300, 0x0, 0, &ccic0_lock},
+	{MMP2_CLK_CCIC1, "ccic1_clk", "ccic1_mix_clk", CLK_SET_RATE_PARENT, APMU_CCIC1, 0x1b, 0x1b, 0x0, 0, &ccic1_lock},
+	{MMP2_CLK_CCIC1_PHY, "ccic1_phy_clk", "ccic1_mix_clk", CLK_SET_RATE_PARENT, APMU_CCIC1, 0x24, 0x24, 0x0, 0, &ccic1_lock},
+	{MMP2_CLK_CCIC1_SPHY, "ccic1_sphy_clk", "ccic1_sphy_div", CLK_SET_RATE_PARENT, APMU_CCIC1, 0x300, 0x300, 0x0, 0, &ccic1_lock},
+};
+
+static void mmp2_axi_periph_clk_init(struct mmp2_clk_unit *pxa_unit)
+{
+	struct clk *clk;
+	struct mmp_clk_unit *unit = &pxa_unit->unit;
+
+	sdh_mix_config.reg_info.reg_clk_ctrl = pxa_unit->apmu_base + APMU_SDH0;
+	clk = mmp_clk_register_mix(NULL, "sdh_mix_clk", sdh_parent_names,
+					ARRAY_SIZE(sdh_parent_names),
+					CLK_SET_RATE_PARENT,
+					&sdh_mix_config, &sdh_lock);
+
+	ccic0_mix_config.reg_info.reg_clk_ctrl = pxa_unit->apmu_base + APMU_CCIC0;
+	clk = mmp_clk_register_mix(NULL, "ccic0_mix_clk", ccic_parent_names,
+					ARRAY_SIZE(ccic_parent_names),
+					CLK_SET_RATE_PARENT,
+					&ccic0_mix_config, &ccic0_lock);
+	mmp_clk_add(unit, MMP2_CLK_CCIC0_MIX, clk);
+
+	ccic1_mix_config.reg_info.reg_clk_ctrl = pxa_unit->apmu_base + APMU_CCIC1;
+	clk = mmp_clk_register_mix(NULL, "ccic1_mix_clk", ccic_parent_names,
+					ARRAY_SIZE(ccic_parent_names),
+					CLK_SET_RATE_PARENT,
+					&ccic1_mix_config, &ccic1_lock);
+	mmp_clk_add(unit, MMP2_CLK_CCIC1_MIX, clk);
+
+	mmp_register_mux_clks(unit, apmu_mux_clks, pxa_unit->apmu_base,
+				ARRAY_SIZE(apmu_mux_clks));
+
+	mmp_register_div_clks(unit, apmu_div_clks, pxa_unit->apmu_base,
+				ARRAY_SIZE(apmu_div_clks));
+
+	mmp_register_gate_clks(unit, apmu_gate_clks, pxa_unit->apmu_base,
+				ARRAY_SIZE(apmu_gate_clks));
+}
+
+static void mmp2_clk_reset_init(struct device_node *np,
+				struct mmp2_clk_unit *pxa_unit)
+{
+	struct mmp_clk_reset_cell *cells;
+	int i, nr_resets;
+
+	nr_resets = ARRAY_SIZE(apbc_gate_clks);
+	cells = kcalloc(nr_resets, sizeof(*cells), GFP_KERNEL);
+	if (!cells)
+		return;
+
+	for (i = 0; i < nr_resets; i++) {
+		cells[i].clk_id = apbc_gate_clks[i].id;
+		cells[i].reg = pxa_unit->apbc_base + apbc_gate_clks[i].offset;
+		cells[i].flags = 0;
+		cells[i].lock = apbc_gate_clks[i].lock;
+		cells[i].bits = 0x4;
+	}
+
+	mmp_clk_reset_register(np, cells, nr_resets);
+}
+
+static void __init mmp2_clk_init(struct device_node *np)
+{
+	struct mmp2_clk_unit *pxa_unit;
+
+	pxa_unit = kzalloc(sizeof(*pxa_unit), GFP_KERNEL);
+	if (!pxa_unit)
+		return;
+
+	pxa_unit->mpmu_base = of_iomap(np, 0);
+	if (!pxa_unit->mpmu_base) {
+		pr_err("failed to map mpmu registers\n");
+		return;
+	}
+
+	pxa_unit->apmu_base = of_iomap(np, 1);
+	if (!pxa_unit->mpmu_base) {
+		pr_err("failed to map apmu registers\n");
+		return;
+	}
+
+	pxa_unit->apbc_base = of_iomap(np, 2);
+	if (!pxa_unit->apbc_base) {
+		pr_err("failed to map apbc registers\n");
+		return;
+	}
+
+	mmp_clk_init(np, &pxa_unit->unit, MMP2_NR_CLKS);
+
+	mmp2_pll_init(pxa_unit);
+
+	mmp2_apb_periph_clk_init(pxa_unit);
+
+	mmp2_axi_periph_clk_init(pxa_unit);
+
+	mmp2_clk_reset_init(np, pxa_unit);
+}
+
+CLK_OF_DECLARE(mmp2_clk, "marvell,mmp2-clock", mmp2_clk_init);
diff --git a/drivers/clk/mmp/clk-of-pxa168.c b/drivers/clk/mmp/clk-of-pxa168.c
new file mode 100644
index 000000000000..5b1810dc4bd2
--- /dev/null
+++ b/drivers/clk/mmp/clk-of-pxa168.c
@@ -0,0 +1,279 @@
+/*
+ * pxa168 clock framework source file
+ *
+ * Copyright (C) 2012 Marvell
+ * Chao Xie <xiechao.mail@gmail.com>
+ *
+ * This file is licensed under the terms of the GNU General Public
+ * License version 2. This program is licensed "as is" without any
+ * warranty of any kind, whether express or implied.
+ */
+
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/spinlock.h>
+#include <linux/io.h>
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/of_address.h>
+
+#include <dt-bindings/clock/marvell,pxa168.h>
+
+#include "clk.h"
+#include "reset.h"
+
+#define APBC_RTC	0x28
+#define APBC_TWSI0	0x2c
+#define APBC_KPC	0x30
+#define APBC_UART0	0x0
+#define APBC_UART1	0x4
+#define APBC_GPIO	0x8
+#define APBC_PWM0	0xc
+#define APBC_PWM1	0x10
+#define APBC_PWM2	0x14
+#define APBC_PWM3	0x18
+#define APBC_SSP0	0x81c
+#define APBC_SSP1	0x820
+#define APBC_SSP2	0x84c
+#define APBC_SSP3	0x858
+#define APBC_SSP4	0x85c
+#define APBC_TWSI1	0x6c
+#define APBC_UART2	0x70
+#define APMU_SDH0	0x54
+#define APMU_SDH1	0x58
+#define APMU_USB	0x5c
+#define APMU_DISP0	0x4c
+#define APMU_CCIC0	0x50
+#define APMU_DFC	0x60
+#define MPMU_UART_PLL	0x14
+
+struct pxa168_clk_unit {
+	struct mmp_clk_unit unit;
+	void __iomem *mpmu_base;
+	void __iomem *apmu_base;
+	void __iomem *apbc_base;
+};
+
+static struct mmp_param_fixed_rate_clk fixed_rate_clks[] = {
+	{PXA168_CLK_CLK32, "clk32", NULL, CLK_IS_ROOT, 32768},
+	{PXA168_CLK_VCTCXO, "vctcxo", NULL, CLK_IS_ROOT, 26000000},
+	{PXA168_CLK_PLL1, "pll1", NULL, CLK_IS_ROOT, 624000000},
+};
+
+static struct mmp_param_fixed_factor_clk fixed_factor_clks[] = {
+	{PXA168_CLK_PLL1_2, "pll1_2", "pll1", 1, 2, 0},
+	{PXA168_CLK_PLL1_4, "pll1_4", "pll1_2", 1, 2, 0},
+	{PXA168_CLK_PLL1_8, "pll1_8", "pll1_4", 1, 2, 0},
+	{PXA168_CLK_PLL1_16, "pll1_16", "pll1_8", 1, 2, 0},
+	{PXA168_CLK_PLL1_6, "pll1_6", "pll1_2", 1, 3, 0},
+	{PXA168_CLK_PLL1_12, "pll1_12", "pll1_6", 1, 2, 0},
+	{PXA168_CLK_PLL1_24, "pll1_24", "pll1_12", 1, 2, 0},
+	{PXA168_CLK_PLL1_48, "pll1_48", "pll1_24", 1, 2, 0},
+	{PXA168_CLK_PLL1_96, "pll1_96", "pll1_48", 1, 2, 0},
+	{PXA168_CLK_PLL1_13, "pll1_13", "pll1", 1, 13, 0},
+	{PXA168_CLK_PLL1_13_1_5, "pll1_13_1_5", "pll1_13", 2, 3, 0},
+	{PXA168_CLK_PLL1_2_1_5, "pll1_2_1_5", "pll1_2", 2, 3, 0},
+	{PXA168_CLK_PLL1_3_16, "pll1_3_16", "pll1", 3, 16, 0},
+};
+
+static struct mmp_clk_factor_masks uart_factor_masks = {
+	.factor = 2,
+	.num_mask = 0x1fff,
+	.den_mask = 0x1fff,
+	.num_shift = 16,
+	.den_shift = 0,
+};
+
+static struct mmp_clk_factor_tbl uart_factor_tbl[] = {
+	{.num = 8125, .den = 1536},	/*14.745MHZ */
+};
+
+static void pxa168_pll_init(struct pxa168_clk_unit *pxa_unit)
+{
+	struct clk *clk;
+	struct mmp_clk_unit *unit = &pxa_unit->unit;
+
+	mmp_register_fixed_rate_clks(unit, fixed_rate_clks,
+					ARRAY_SIZE(fixed_rate_clks));
+
+	mmp_register_fixed_factor_clks(unit, fixed_factor_clks,
+					ARRAY_SIZE(fixed_factor_clks));
+
+	clk = mmp_clk_register_factor("uart_pll", "pll1_4",
+				CLK_SET_RATE_PARENT,
+				pxa_unit->mpmu_base + MPMU_UART_PLL,
+				&uart_factor_masks, uart_factor_tbl,
+				ARRAY_SIZE(uart_factor_tbl), NULL);
+	mmp_clk_add(unit, PXA168_CLK_UART_PLL, clk);
+}
+
+static DEFINE_SPINLOCK(uart0_lock);
+static DEFINE_SPINLOCK(uart1_lock);
+static DEFINE_SPINLOCK(uart2_lock);
+static const char *uart_parent_names[] = {"pll1_3_16", "uart_pll"};
+
+static DEFINE_SPINLOCK(ssp0_lock);
+static DEFINE_SPINLOCK(ssp1_lock);
+static DEFINE_SPINLOCK(ssp2_lock);
+static DEFINE_SPINLOCK(ssp3_lock);
+static DEFINE_SPINLOCK(ssp4_lock);
+static const char *ssp_parent_names[] = {"pll1_96", "pll1_48", "pll1_24", "pll1_12"};
+
+static DEFINE_SPINLOCK(reset_lock);
+
+static struct mmp_param_mux_clk apbc_mux_clks[] = {
+	{0, "uart0_mux", uart_parent_names, ARRAY_SIZE(uart_parent_names), CLK_SET_RATE_PARENT, APBC_UART0, 4, 3, 0, &uart0_lock},
+	{0, "uart1_mux", uart_parent_names, ARRAY_SIZE(uart_parent_names), CLK_SET_RATE_PARENT, APBC_UART1, 4, 3, 0, &uart1_lock},
+	{0, "uart2_mux", uart_parent_names, ARRAY_SIZE(uart_parent_names), CLK_SET_RATE_PARENT, APBC_UART2, 4, 3, 0, &uart2_lock},
+	{0, "ssp0_mux", ssp_parent_names, ARRAY_SIZE(ssp_parent_names), CLK_SET_RATE_PARENT, APBC_SSP0, 4, 3, 0, &ssp0_lock},
+	{0, "ssp1_mux", ssp_parent_names, ARRAY_SIZE(ssp_parent_names), CLK_SET_RATE_PARENT, APBC_SSP1, 4, 3, 0, &ssp1_lock},
+	{0, "ssp2_mux", ssp_parent_names, ARRAY_SIZE(ssp_parent_names), CLK_SET_RATE_PARENT, APBC_SSP2, 4, 3, 0, &ssp2_lock},
+	{0, "ssp3_mux", ssp_parent_names, ARRAY_SIZE(ssp_parent_names), CLK_SET_RATE_PARENT, APBC_SSP3, 4, 3, 0, &ssp3_lock},
+	{0, "ssp4_mux", ssp_parent_names, ARRAY_SIZE(ssp_parent_names), CLK_SET_RATE_PARENT, APBC_SSP4, 4, 3, 0, &ssp4_lock},
+};
+
+static struct mmp_param_gate_clk apbc_gate_clks[] = {
+	{PXA168_CLK_TWSI0, "twsi0_clk", "pll1_13_1_5", CLK_SET_RATE_PARENT, APBC_TWSI0, 0x3, 0x3, 0x0, 0, &reset_lock},
+	{PXA168_CLK_TWSI1, "twsi1_clk", "pll1_13_1_5", CLK_SET_RATE_PARENT, APBC_TWSI1, 0x3, 0x3, 0x0, 0, &reset_lock},
+	{PXA168_CLK_GPIO, "gpio_clk", "vctcxo", CLK_SET_RATE_PARENT, APBC_GPIO, 0x3, 0x3, 0x0, 0, &reset_lock},
+	{PXA168_CLK_KPC, "kpc_clk", "clk32", CLK_SET_RATE_PARENT, APBC_KPC, 0x3, 0x3, 0x0, MMP_CLK_GATE_NEED_DELAY, NULL},
+	{PXA168_CLK_RTC, "rtc_clk", "clk32", CLK_SET_RATE_PARENT, APBC_RTC, 0x83, 0x83, 0x0, MMP_CLK_GATE_NEED_DELAY, NULL},
+	{PXA168_CLK_PWM0, "pwm0_clk", "pll1_48", CLK_SET_RATE_PARENT, APBC_PWM0, 0x3, 0x3, 0x0, 0, &reset_lock},
+	{PXA168_CLK_PWM1, "pwm1_clk", "pll1_48", CLK_SET_RATE_PARENT, APBC_PWM1, 0x3, 0x3, 0x0, 0, &reset_lock},
+	{PXA168_CLK_PWM2, "pwm2_clk", "pll1_48", CLK_SET_RATE_PARENT, APBC_PWM2, 0x3, 0x3, 0x0, 0, &reset_lock},
+	{PXA168_CLK_PWM3, "pwm3_clk", "pll1_48", CLK_SET_RATE_PARENT, APBC_PWM3, 0x3, 0x3, 0x0, 0, &reset_lock},
+	/* The gate clocks has mux parent. */
+	{PXA168_CLK_UART0, "uart0_clk", "uart0_mux", CLK_SET_RATE_PARENT, APBC_UART0, 0x3, 0x3, 0x0, 0, &uart0_lock},
+	{PXA168_CLK_UART1, "uart1_clk", "uart1_mux", CLK_SET_RATE_PARENT, APBC_UART1, 0x3, 0x3, 0x0, 0, &uart1_lock},
+	{PXA168_CLK_UART2, "uart2_clk", "uart2_mux", CLK_SET_RATE_PARENT, APBC_UART2, 0x3, 0x3, 0x0, 0, &uart2_lock},
+	{PXA168_CLK_SSP0, "ssp0_clk", "ssp0_mux", CLK_SET_RATE_PARENT, APBC_SSP0, 0x3, 0x3, 0x0, 0, &ssp0_lock},
+	{PXA168_CLK_SSP1, "ssp1_clk", "ssp1_mux", CLK_SET_RATE_PARENT, APBC_SSP1, 0x3, 0x3, 0x0, 0, &ssp1_lock},
+	{PXA168_CLK_SSP2, "ssp2_clk", "ssp2_mux", CLK_SET_RATE_PARENT, APBC_SSP2, 0x3, 0x3, 0x0, 0, &ssp2_lock},
+	{PXA168_CLK_SSP3, "ssp3_clk", "ssp3_mux", CLK_SET_RATE_PARENT, APBC_SSP3, 0x3, 0x3, 0x0, 0, &ssp3_lock},
+	{PXA168_CLK_SSP4, "ssp4_clk", "ssp4_mux", CLK_SET_RATE_PARENT, APBC_SSP4, 0x3, 0x3, 0x0, 0, &ssp4_lock},
+};
+
+static void pxa168_apb_periph_clk_init(struct pxa168_clk_unit *pxa_unit)
+{
+	struct mmp_clk_unit *unit = &pxa_unit->unit;
+
+	mmp_register_mux_clks(unit, apbc_mux_clks, pxa_unit->apbc_base,
+				ARRAY_SIZE(apbc_mux_clks));
+
+	mmp_register_gate_clks(unit, apbc_gate_clks, pxa_unit->apbc_base,
+				ARRAY_SIZE(apbc_gate_clks));
+
+}
+
+static DEFINE_SPINLOCK(sdh0_lock);
+static DEFINE_SPINLOCK(sdh1_lock);
+static const char *sdh_parent_names[] = {"pll1_12", "pll1_13"};
+
+static DEFINE_SPINLOCK(usb_lock);
+
+static DEFINE_SPINLOCK(disp0_lock);
+static const char *disp_parent_names[] = {"pll1_2", "pll1_12"};
+
+static DEFINE_SPINLOCK(ccic0_lock);
+static const char *ccic_parent_names[] = {"pll1_2", "pll1_12"};
+static const char *ccic_phy_parent_names[] = {"pll1_6", "pll1_12"};
+
+static struct mmp_param_mux_clk apmu_mux_clks[] = {
+	{0, "sdh0_mux", sdh_parent_names, ARRAY_SIZE(sdh_parent_names), CLK_SET_RATE_PARENT, APMU_SDH0, 6, 1, 0, &sdh0_lock},
+	{0, "sdh1_mux", sdh_parent_names, ARRAY_SIZE(sdh_parent_names), CLK_SET_RATE_PARENT, APMU_SDH1, 6, 1, 0, &sdh1_lock},
+	{0, "disp0_mux", disp_parent_names, ARRAY_SIZE(disp_parent_names), CLK_SET_RATE_PARENT, APMU_DISP0, 6, 1, 0, &disp0_lock},
+	{0, "ccic0_mux", ccic_parent_names, ARRAY_SIZE(ccic_parent_names), CLK_SET_RATE_PARENT, APMU_CCIC0, 6, 1, 0, &ccic0_lock},
+	{0, "ccic0_phy_mux", ccic_phy_parent_names, ARRAY_SIZE(ccic_phy_parent_names), CLK_SET_RATE_PARENT, APMU_CCIC0, 7, 1, 0, &ccic0_lock},
+};
+
+static struct mmp_param_div_clk apmu_div_clks[] = {
+	{0, "ccic0_sphy_div", "ccic0_mux", CLK_SET_RATE_PARENT, APMU_CCIC0, 10, 5, 0, &ccic0_lock},
+};
+
+static struct mmp_param_gate_clk apmu_gate_clks[] = {
+	{PXA168_CLK_DFC, "dfc_clk", "pll1_4", CLK_SET_RATE_PARENT, APMU_DFC, 0x19b, 0x19b, 0x0, 0, NULL},
+	{PXA168_CLK_USB, "usb_clk", "usb_pll", 0, APMU_USB, 0x9, 0x9, 0x0, 0, &usb_lock},
+	{PXA168_CLK_SPH, "sph_clk", "usb_pll", 0, APMU_USB, 0x12, 0x12, 0x0, 0, &usb_lock},
+	/* The gate clocks has mux parent. */
+	{PXA168_CLK_SDH0, "sdh0_clk", "sdh0_mux", CLK_SET_RATE_PARENT, APMU_SDH0, 0x1b, 0x1b, 0x0, 0, &sdh0_lock},
+	{PXA168_CLK_SDH1, "sdh1_clk", "sdh1_mux", CLK_SET_RATE_PARENT, APMU_SDH1, 0x1b, 0x1b, 0x0, 0, &sdh1_lock},
+	{PXA168_CLK_DISP0, "disp0_clk", "disp0_mux", CLK_SET_RATE_PARENT, APMU_DISP0, 0x1b, 0x1b, 0x0, 0, &disp0_lock},
+	{PXA168_CLK_CCIC0, "ccic0_clk", "ccic0_mux", CLK_SET_RATE_PARENT, APMU_CCIC0, 0x1b, 0x1b, 0x0, 0, &ccic0_lock},
+	{PXA168_CLK_CCIC0_PHY, "ccic0_phy_clk", "ccic0_phy_mux", CLK_SET_RATE_PARENT, APMU_CCIC0, 0x24, 0x24, 0x0, 0, &ccic0_lock},
+	{PXA168_CLK_CCIC0_SPHY, "ccic0_sphy_clk", "ccic0_sphy_div", CLK_SET_RATE_PARENT, APMU_CCIC0, 0x300, 0x300, 0x0, 0, &ccic0_lock},
+};
+
+static void pxa168_axi_periph_clk_init(struct pxa168_clk_unit *pxa_unit)
+{
+	struct mmp_clk_unit *unit = &pxa_unit->unit;
+
+	mmp_register_mux_clks(unit, apmu_mux_clks, pxa_unit->apmu_base,
+				ARRAY_SIZE(apmu_mux_clks));
+
+	mmp_register_div_clks(unit, apmu_div_clks, pxa_unit->apmu_base,
+				ARRAY_SIZE(apmu_div_clks));
+
+	mmp_register_gate_clks(unit, apmu_gate_clks, pxa_unit->apmu_base,
+				ARRAY_SIZE(apmu_gate_clks));
+}
+
+static void pxa168_clk_reset_init(struct device_node *np,
+				struct pxa168_clk_unit *pxa_unit)
+{
+	struct mmp_clk_reset_cell *cells;
+	int i, nr_resets;
+
+	nr_resets = ARRAY_SIZE(apbc_gate_clks);
+	cells = kcalloc(nr_resets, sizeof(*cells), GFP_KERNEL);
+	if (!cells)
+		return;
+
+	for (i = 0; i < nr_resets; i++) {
+		cells[i].clk_id = apbc_gate_clks[i].id;
+		cells[i].reg = pxa_unit->apbc_base + apbc_gate_clks[i].offset;
+		cells[i].flags = 0;
+		cells[i].lock = apbc_gate_clks[i].lock;
+		cells[i].bits = 0x4;
+	}
+
+	mmp_clk_reset_register(np, cells, nr_resets);
+}
+
+static void __init pxa168_clk_init(struct device_node *np)
+{
+	struct pxa168_clk_unit *pxa_unit;
+
+	pxa_unit = kzalloc(sizeof(*pxa_unit), GFP_KERNEL);
+	if (!pxa_unit)
+		return;
+
+	pxa_unit->mpmu_base = of_iomap(np, 0);
+	if (!pxa_unit->mpmu_base) {
+		pr_err("failed to map mpmu registers\n");
+		return;
+	}
+
+	pxa_unit->apmu_base = of_iomap(np, 1);
+	if (!pxa_unit->mpmu_base) {
+		pr_err("failed to map apmu registers\n");
+		return;
+	}
+
+	pxa_unit->apbc_base = of_iomap(np, 2);
+	if (!pxa_unit->apbc_base) {
+		pr_err("failed to map apbc registers\n");
+		return;
+	}
+
+	mmp_clk_init(np, &pxa_unit->unit, PXA168_NR_CLKS);
+
+	pxa168_pll_init(pxa_unit);
+
+	pxa168_apb_periph_clk_init(pxa_unit);
+
+	pxa168_axi_periph_clk_init(pxa_unit);
+
+	pxa168_clk_reset_init(np, pxa_unit);
+}
+
+CLK_OF_DECLARE(pxa168_clk, "marvell,pxa168-clock", pxa168_clk_init);
diff --git a/drivers/clk/mmp/clk-of-pxa910.c b/drivers/clk/mmp/clk-of-pxa910.c
new file mode 100644
index 000000000000..5e3c80dad336
--- /dev/null
+++ b/drivers/clk/mmp/clk-of-pxa910.c
@@ -0,0 +1,301 @@
+/*
+ * pxa910 clock framework source file
+ *
+ * Copyright (C) 2012 Marvell
+ * Chao Xie <xiechao.mail@gmail.com>
+ *
+ * This file is licensed under the terms of the GNU General Public
+ * License version 2. This program is licensed "as is" without any
+ * warranty of any kind, whether express or implied.
+ */
+
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/spinlock.h>
+#include <linux/io.h>
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/of_address.h>
+
+#include <dt-bindings/clock/marvell,pxa910.h>
+
+#include "clk.h"
+#include "reset.h"
+
+#define APBC_RTC	0x28
+#define APBC_TWSI0	0x2c
+#define APBC_KPC	0x18
+#define APBC_UART0	0x0
+#define APBC_UART1	0x4
+#define APBC_GPIO	0x8
+#define APBC_PWM0	0xc
+#define APBC_PWM1	0x10
+#define APBC_PWM2	0x14
+#define APBC_PWM3	0x18
+#define APBC_SSP0	0x1c
+#define APBC_SSP1	0x20
+#define APBC_SSP2	0x4c
+#define APBCP_TWSI1	0x28
+#define APBCP_UART2	0x1c
+#define APMU_SDH0	0x54
+#define APMU_SDH1	0x58
+#define APMU_USB	0x5c
+#define APMU_DISP0	0x4c
+#define APMU_CCIC0	0x50
+#define APMU_DFC	0x60
+#define MPMU_UART_PLL	0x14
+
+struct pxa910_clk_unit {
+	struct mmp_clk_unit unit;
+	void __iomem *mpmu_base;
+	void __iomem *apmu_base;
+	void __iomem *apbc_base;
+	void __iomem *apbcp_base;
+};
+
+static struct mmp_param_fixed_rate_clk fixed_rate_clks[] = {
+	{PXA910_CLK_CLK32, "clk32", NULL, CLK_IS_ROOT, 32768},
+	{PXA910_CLK_VCTCXO, "vctcxo", NULL, CLK_IS_ROOT, 26000000},
+	{PXA910_CLK_PLL1, "pll1", NULL, CLK_IS_ROOT, 624000000},
+};
+
+static struct mmp_param_fixed_factor_clk fixed_factor_clks[] = {
+	{PXA910_CLK_PLL1_2, "pll1_2", "pll1", 1, 2, 0},
+	{PXA910_CLK_PLL1_4, "pll1_4", "pll1_2", 1, 2, 0},
+	{PXA910_CLK_PLL1_8, "pll1_8", "pll1_4", 1, 2, 0},
+	{PXA910_CLK_PLL1_16, "pll1_16", "pll1_8", 1, 2, 0},
+	{PXA910_CLK_PLL1_6, "pll1_6", "pll1_2", 1, 3, 0},
+	{PXA910_CLK_PLL1_12, "pll1_12", "pll1_6", 1, 2, 0},
+	{PXA910_CLK_PLL1_24, "pll1_24", "pll1_12", 1, 2, 0},
+	{PXA910_CLK_PLL1_48, "pll1_48", "pll1_24", 1, 2, 0},
+	{PXA910_CLK_PLL1_96, "pll1_96", "pll1_48", 1, 2, 0},
+	{PXA910_CLK_PLL1_13, "pll1_13", "pll1", 1, 13, 0},
+	{PXA910_CLK_PLL1_13_1_5, "pll1_13_1_5", "pll1_13", 2, 3, 0},
+	{PXA910_CLK_PLL1_2_1_5, "pll1_2_1_5", "pll1_2", 2, 3, 0},
+	{PXA910_CLK_PLL1_3_16, "pll1_3_16", "pll1", 3, 16, 0},
+};
+
+static struct mmp_clk_factor_masks uart_factor_masks = {
+	.factor = 2,
+	.num_mask = 0x1fff,
+	.den_mask = 0x1fff,
+	.num_shift = 16,
+	.den_shift = 0,
+};
+
+static struct mmp_clk_factor_tbl uart_factor_tbl[] = {
+	{.num = 8125, .den = 1536},	/*14.745MHZ */
+};
+
+static void pxa910_pll_init(struct pxa910_clk_unit *pxa_unit)
+{
+	struct clk *clk;
+	struct mmp_clk_unit *unit = &pxa_unit->unit;
+
+	mmp_register_fixed_rate_clks(unit, fixed_rate_clks,
+					ARRAY_SIZE(fixed_rate_clks));
+
+	mmp_register_fixed_factor_clks(unit, fixed_factor_clks,
+					ARRAY_SIZE(fixed_factor_clks));
+
+	clk = mmp_clk_register_factor("uart_pll", "pll1_4",
+				CLK_SET_RATE_PARENT,
+				pxa_unit->mpmu_base + MPMU_UART_PLL,
+				&uart_factor_masks, uart_factor_tbl,
+				ARRAY_SIZE(uart_factor_tbl), NULL);
+	mmp_clk_add(unit, PXA910_CLK_UART_PLL, clk);
+}
+
+static DEFINE_SPINLOCK(uart0_lock);
+static DEFINE_SPINLOCK(uart1_lock);
+static DEFINE_SPINLOCK(uart2_lock);
+static const char *uart_parent_names[] = {"pll1_3_16", "uart_pll"};
+
+static DEFINE_SPINLOCK(ssp0_lock);
+static DEFINE_SPINLOCK(ssp1_lock);
+static const char *ssp_parent_names[] = {"pll1_96", "pll1_48", "pll1_24", "pll1_12"};
+
+static DEFINE_SPINLOCK(reset_lock);
+
+static struct mmp_param_mux_clk apbc_mux_clks[] = {
+	{0, "uart0_mux", uart_parent_names, ARRAY_SIZE(uart_parent_names), CLK_SET_RATE_PARENT, APBC_UART0, 4, 3, 0, &uart0_lock},
+	{0, "uart1_mux", uart_parent_names, ARRAY_SIZE(uart_parent_names), CLK_SET_RATE_PARENT, APBC_UART1, 4, 3, 0, &uart1_lock},
+	{0, "ssp0_mux", ssp_parent_names, ARRAY_SIZE(ssp_parent_names), CLK_SET_RATE_PARENT, APBC_SSP0, 4, 3, 0, &ssp0_lock},
+	{0, "ssp1_mux", ssp_parent_names, ARRAY_SIZE(ssp_parent_names), CLK_SET_RATE_PARENT, APBC_SSP1, 4, 3, 0, &ssp1_lock},
+};
+
+static struct mmp_param_mux_clk apbcp_mux_clks[] = {
+	{0, "uart2_mux", uart_parent_names, ARRAY_SIZE(uart_parent_names), CLK_SET_RATE_PARENT, APBCP_UART2, 4, 3, 0, &uart2_lock},
+};
+
+static struct mmp_param_gate_clk apbc_gate_clks[] = {
+	{PXA910_CLK_TWSI0, "twsi0_clk", "pll1_13_1_5", CLK_SET_RATE_PARENT, APBC_TWSI0, 0x3, 0x3, 0x0, 0, &reset_lock},
+	{PXA910_CLK_GPIO, "gpio_clk", "vctcxo", CLK_SET_RATE_PARENT, APBC_GPIO, 0x3, 0x3, 0x0, 0, &reset_lock},
+	{PXA910_CLK_KPC, "kpc_clk", "clk32", CLK_SET_RATE_PARENT, APBC_KPC, 0x3, 0x3, 0x0, MMP_CLK_GATE_NEED_DELAY, NULL},
+	{PXA910_CLK_RTC, "rtc_clk", "clk32", CLK_SET_RATE_PARENT, APBC_RTC, 0x83, 0x83, 0x0, MMP_CLK_GATE_NEED_DELAY, NULL},
+	{PXA910_CLK_PWM0, "pwm0_clk", "pll1_48", CLK_SET_RATE_PARENT, APBC_PWM0, 0x3, 0x3, 0x0, 0, &reset_lock},
+	{PXA910_CLK_PWM1, "pwm1_clk", "pll1_48", CLK_SET_RATE_PARENT, APBC_PWM1, 0x3, 0x3, 0x0, 0, &reset_lock},
+	{PXA910_CLK_PWM2, "pwm2_clk", "pll1_48", CLK_SET_RATE_PARENT, APBC_PWM2, 0x3, 0x3, 0x0, 0, &reset_lock},
+	{PXA910_CLK_PWM3, "pwm3_clk", "pll1_48", CLK_SET_RATE_PARENT, APBC_PWM3, 0x3, 0x3, 0x0, 0, &reset_lock},
+	/* The gate clocks has mux parent. */
+	{PXA910_CLK_UART0, "uart0_clk", "uart0_mux", CLK_SET_RATE_PARENT, APBC_UART0, 0x3, 0x3, 0x0, 0, &uart0_lock},
+	{PXA910_CLK_UART1, "uart1_clk", "uart1_mux", CLK_SET_RATE_PARENT, APBC_UART1, 0x3, 0x3, 0x0, 0, &uart1_lock},
+	{PXA910_CLK_SSP0, "ssp0_clk", "ssp0_mux", CLK_SET_RATE_PARENT, APBC_SSP0, 0x3, 0x3, 0x0, 0, &ssp0_lock},
+	{PXA910_CLK_SSP1, "ssp1_clk", "ssp1_mux", CLK_SET_RATE_PARENT, APBC_SSP1, 0x3, 0x3, 0x0, 0, &ssp1_lock},
+};
+
+static struct mmp_param_gate_clk apbcp_gate_clks[] = {
+	{PXA910_CLK_TWSI1, "twsi1_clk", "pll1_13_1_5", CLK_SET_RATE_PARENT, APBCP_TWSI1, 0x3, 0x3, 0x0, 0, &reset_lock},
+	/* The gate clocks has mux parent. */
+	{PXA910_CLK_UART2, "uart2_clk", "uart2_mux", CLK_SET_RATE_PARENT, APBCP_UART2, 0x3, 0x3, 0x0, 0, &uart2_lock},
+};
+
+static void pxa910_apb_periph_clk_init(struct pxa910_clk_unit *pxa_unit)
+{
+	struct mmp_clk_unit *unit = &pxa_unit->unit;
+
+	mmp_register_mux_clks(unit, apbc_mux_clks, pxa_unit->apbc_base,
+				ARRAY_SIZE(apbc_mux_clks));
+
+	mmp_register_mux_clks(unit, apbcp_mux_clks, pxa_unit->apbcp_base,
+				ARRAY_SIZE(apbcp_mux_clks));
+
+	mmp_register_gate_clks(unit, apbc_gate_clks, pxa_unit->apbc_base,
+				ARRAY_SIZE(apbc_gate_clks));
+
+	mmp_register_gate_clks(unit, apbcp_gate_clks, pxa_unit->apbcp_base,
+				ARRAY_SIZE(apbcp_gate_clks));
+}
+
+static DEFINE_SPINLOCK(sdh0_lock);
+static DEFINE_SPINLOCK(sdh1_lock);
+static const char *sdh_parent_names[] = {"pll1_12", "pll1_13"};
+
+static DEFINE_SPINLOCK(usb_lock);
+
+static DEFINE_SPINLOCK(disp0_lock);
+static const char *disp_parent_names[] = {"pll1_2", "pll1_12"};
+
+static DEFINE_SPINLOCK(ccic0_lock);
+static const char *ccic_parent_names[] = {"pll1_2", "pll1_12"};
+static const char *ccic_phy_parent_names[] = {"pll1_6", "pll1_12"};
+
+static struct mmp_param_mux_clk apmu_mux_clks[] = {
+	{0, "sdh0_mux", sdh_parent_names, ARRAY_SIZE(sdh_parent_names), CLK_SET_RATE_PARENT, APMU_SDH0, 6, 1, 0, &sdh0_lock},
+	{0, "sdh1_mux", sdh_parent_names, ARRAY_SIZE(sdh_parent_names), CLK_SET_RATE_PARENT, APMU_SDH1, 6, 1, 0, &sdh1_lock},
+	{0, "disp0_mux", disp_parent_names, ARRAY_SIZE(disp_parent_names), CLK_SET_RATE_PARENT, APMU_DISP0, 6, 1, 0, &disp0_lock},
+	{0, "ccic0_mux", ccic_parent_names, ARRAY_SIZE(ccic_parent_names), CLK_SET_RATE_PARENT, APMU_CCIC0, 6, 1, 0, &ccic0_lock},
+	{0, "ccic0_phy_mux", ccic_phy_parent_names, ARRAY_SIZE(ccic_phy_parent_names), CLK_SET_RATE_PARENT, APMU_CCIC0, 7, 1, 0, &ccic0_lock},
+};
+
+static struct mmp_param_div_clk apmu_div_clks[] = {
+	{0, "ccic0_sphy_div", "ccic0_mux", CLK_SET_RATE_PARENT, APMU_CCIC0, 10, 5, 0, &ccic0_lock},
+};
+
+static struct mmp_param_gate_clk apmu_gate_clks[] = {
+	{PXA910_CLK_DFC, "dfc_clk", "pll1_4", CLK_SET_RATE_PARENT, APMU_DFC, 0x19b, 0x19b, 0x0, 0, NULL},
+	{PXA910_CLK_USB, "usb_clk", "usb_pll", 0, APMU_USB, 0x9, 0x9, 0x0, 0, &usb_lock},
+	{PXA910_CLK_SPH, "sph_clk", "usb_pll", 0, APMU_USB, 0x12, 0x12, 0x0, 0, &usb_lock},
+	/* The gate clocks has mux parent. */
+	{PXA910_CLK_SDH0, "sdh0_clk", "sdh0_mux", CLK_SET_RATE_PARENT, APMU_SDH0, 0x1b, 0x1b, 0x0, 0, &sdh0_lock},
+	{PXA910_CLK_SDH1, "sdh1_clk", "sdh1_mux", CLK_SET_RATE_PARENT, APMU_SDH1, 0x1b, 0x1b, 0x0, 0, &sdh1_lock},
+	{PXA910_CLK_DISP0, "disp0_clk", "disp0_mux", CLK_SET_RATE_PARENT, APMU_DISP0, 0x1b, 0x1b, 0x0, 0, &disp0_lock},
+	{PXA910_CLK_CCIC0, "ccic0_clk", "ccic0_mux", CLK_SET_RATE_PARENT, APMU_CCIC0, 0x1b, 0x1b, 0x0, 0, &ccic0_lock},
+	{PXA910_CLK_CCIC0_PHY, "ccic0_phy_clk", "ccic0_phy_mux", CLK_SET_RATE_PARENT, APMU_CCIC0, 0x24, 0x24, 0x0, 0, &ccic0_lock},
+	{PXA910_CLK_CCIC0_SPHY, "ccic0_sphy_clk", "ccic0_sphy_div", CLK_SET_RATE_PARENT, APMU_CCIC0, 0x300, 0x300, 0x0, 0, &ccic0_lock},
+};
+
+static void pxa910_axi_periph_clk_init(struct pxa910_clk_unit *pxa_unit)
+{
+	struct mmp_clk_unit *unit = &pxa_unit->unit;
+
+	mmp_register_mux_clks(unit, apmu_mux_clks, pxa_unit->apmu_base,
+				ARRAY_SIZE(apmu_mux_clks));
+
+	mmp_register_div_clks(unit, apmu_div_clks, pxa_unit->apmu_base,
+				ARRAY_SIZE(apmu_div_clks));
+
+	mmp_register_gate_clks(unit, apmu_gate_clks, pxa_unit->apmu_base,
+				ARRAY_SIZE(apmu_gate_clks));
+}
+
+static void pxa910_clk_reset_init(struct device_node *np,
+				struct pxa910_clk_unit *pxa_unit)
+{
+	struct mmp_clk_reset_cell *cells;
+	int i, base, nr_resets_apbc, nr_resets_apbcp, nr_resets;
+
+	nr_resets_apbc = ARRAY_SIZE(apbc_gate_clks);
+	nr_resets_apbcp = ARRAY_SIZE(apbcp_gate_clks);
+	nr_resets = nr_resets_apbc + nr_resets_apbcp;
+	cells = kcalloc(nr_resets, sizeof(*cells), GFP_KERNEL);
+	if (!cells)
+		return;
+
+	base = 0;
+	for (i = 0; i < nr_resets_apbc; i++) {
+		cells[base + i].clk_id = apbc_gate_clks[i].id;
+		cells[base + i].reg =
+			pxa_unit->apbc_base + apbc_gate_clks[i].offset;
+		cells[base + i].flags = 0;
+		cells[base + i].lock = apbc_gate_clks[i].lock;
+		cells[base + i].bits = 0x4;
+	}
+
+	base = nr_resets_apbc;
+	for (i = 0; i < nr_resets_apbcp; i++) {
+		cells[base + i].clk_id = apbcp_gate_clks[i].id;
+		cells[base + i].reg =
+			pxa_unit->apbc_base + apbc_gate_clks[i].offset;
+		cells[base + i].flags = 0;
+		cells[base + i].lock = apbc_gate_clks[i].lock;
+		cells[base + i].bits = 0x4;
+	}
+
+	mmp_clk_reset_register(np, cells, nr_resets);
+}
+
+static void __init pxa910_clk_init(struct device_node *np)
+{
+	struct pxa910_clk_unit *pxa_unit;
+
+	pxa_unit = kzalloc(sizeof(*pxa_unit), GFP_KERNEL);
+	if (!pxa_unit)
+		return;
+
+	pxa_unit->mpmu_base = of_iomap(np, 0);
+	if (!pxa_unit->mpmu_base) {
+		pr_err("failed to map mpmu registers\n");
+		return;
+	}
+
+	pxa_unit->apmu_base = of_iomap(np, 1);
+	if (!pxa_unit->mpmu_base) {
+		pr_err("failed to map apmu registers\n");
+		return;
+	}
+
+	pxa_unit->apbc_base = of_iomap(np, 2);
+	if (!pxa_unit->apbc_base) {
+		pr_err("failed to map apbc registers\n");
+		return;
+	}
+
+	pxa_unit->apbcp_base = of_iomap(np, 3);
+	if (!pxa_unit->mpmu_base) {
+		pr_err("failed to map apbcp registers\n");
+		return;
+	}
+
+	mmp_clk_init(np, &pxa_unit->unit, PXA910_NR_CLKS);
+
+	pxa910_pll_init(pxa_unit);
+
+	pxa910_apb_periph_clk_init(pxa_unit);
+
+	pxa910_axi_periph_clk_init(pxa_unit);
+
+	pxa910_clk_reset_init(np, pxa_unit);
+}
+
+CLK_OF_DECLARE(pxa910_clk, "marvell,pxa910-clock", pxa910_clk_init);
diff --git a/drivers/clk/mmp/clk-pxa168.c b/drivers/clk/mmp/clk-pxa168.c
index 014396b028a2..93e967c0f972 100644
--- a/drivers/clk/mmp/clk-pxa168.c
+++ b/drivers/clk/mmp/clk-pxa168.c
@@ -47,7 +47,7 @@
 
 static DEFINE_SPINLOCK(clk_lock);
 
-static struct clk_factor_masks uart_factor_masks = {
+static struct mmp_clk_factor_masks uart_factor_masks = {
 	.factor = 2,
 	.num_mask = 0x1fff,
 	.den_mask = 0x1fff,
@@ -55,7 +55,7 @@ static struct clk_factor_masks uart_factor_masks = {
 	.den_shift = 0,
 };
 
-static struct clk_factor_tbl uart_factor_tbl[] = {
+static struct mmp_clk_factor_tbl uart_factor_tbl[] = {
 	{.num = 8125, .den = 1536},	/*14.745MHZ */
 };
 
@@ -158,7 +158,7 @@ void __init pxa168_clk_init(void)
 	uart_pll = mmp_clk_register_factor("uart_pll", "pll1_4", 0,
 				mpmu_base + MPMU_UART_PLL,
 				&uart_factor_masks, uart_factor_tbl,
-				ARRAY_SIZE(uart_factor_tbl));
+				ARRAY_SIZE(uart_factor_tbl), &clk_lock);
 	clk_set_rate(uart_pll, 14745600);
 	clk_register_clkdev(uart_pll, "uart_pll", NULL);
 
diff --git a/drivers/clk/mmp/clk-pxa910.c b/drivers/clk/mmp/clk-pxa910.c
index 9efc6a47535d..993abcdb32cc 100644
--- a/drivers/clk/mmp/clk-pxa910.c
+++ b/drivers/clk/mmp/clk-pxa910.c
@@ -45,7 +45,7 @@
 
 static DEFINE_SPINLOCK(clk_lock);
 
-static struct clk_factor_masks uart_factor_masks = {
+static struct mmp_clk_factor_masks uart_factor_masks = {
 	.factor = 2,
 	.num_mask = 0x1fff,
 	.den_mask = 0x1fff,
@@ -53,7 +53,7 @@ static struct clk_factor_masks uart_factor_masks = {
 	.den_shift = 0,
 };
 
-static struct clk_factor_tbl uart_factor_tbl[] = {
+static struct mmp_clk_factor_tbl uart_factor_tbl[] = {
 	{.num = 8125, .den = 1536},	/*14.745MHZ */
 };
 
@@ -163,7 +163,7 @@ void __init pxa910_clk_init(void)
 	uart_pll =  mmp_clk_register_factor("uart_pll", "pll1_4", 0,
 				mpmu_base + MPMU_UART_PLL,
 				&uart_factor_masks, uart_factor_tbl,
-				ARRAY_SIZE(uart_factor_tbl));
+				ARRAY_SIZE(uart_factor_tbl), &clk_lock);
 	clk_set_rate(uart_pll, 14745600);
 	clk_register_clkdev(uart_pll, "uart_pll", NULL);
 
diff --git a/drivers/clk/mmp/clk.c b/drivers/clk/mmp/clk.c
new file mode 100644
index 000000000000..cf038ef54c59
--- /dev/null
+++ b/drivers/clk/mmp/clk.c
@@ -0,0 +1,192 @@
+#include <linux/io.h>
+#include <linux/clk.h>
+#include <linux/clk-provider.h>
+#include <linux/clkdev.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+
+#include "clk.h"
+
+void mmp_clk_init(struct device_node *np, struct mmp_clk_unit *unit,
+		int nr_clks)
+{
+	static struct clk **clk_table;
+
+	clk_table = kcalloc(nr_clks, sizeof(struct clk *), GFP_KERNEL);
+	if (!clk_table)
+		return;
+
+	unit->clk_table = clk_table;
+	unit->nr_clks = nr_clks;
+	unit->clk_data.clks = clk_table;
+	unit->clk_data.clk_num = nr_clks;
+	of_clk_add_provider(np, of_clk_src_onecell_get, &unit->clk_data);
+}
+
+void mmp_register_fixed_rate_clks(struct mmp_clk_unit *unit,
+				struct mmp_param_fixed_rate_clk *clks,
+				int size)
+{
+	int i;
+	struct clk *clk;
+
+	for (i = 0; i < size; i++) {
+		clk = clk_register_fixed_rate(NULL, clks[i].name,
+					clks[i].parent_name,
+					clks[i].flags,
+					clks[i].fixed_rate);
+		if (IS_ERR(clk)) {
+			pr_err("%s: failed to register clock %s\n",
+			       __func__, clks[i].name);
+			continue;
+		}
+		if (clks[i].id)
+			unit->clk_table[clks[i].id] = clk;
+	}
+}
+
+void mmp_register_fixed_factor_clks(struct mmp_clk_unit *unit,
+				struct mmp_param_fixed_factor_clk *clks,
+				int size)
+{
+	struct clk *clk;
+	int i;
+
+	for (i = 0; i < size; i++) {
+		clk = clk_register_fixed_factor(NULL, clks[i].name,
+						clks[i].parent_name,
+						clks[i].flags, clks[i].mult,
+						clks[i].div);
+		if (IS_ERR(clk)) {
+			pr_err("%s: failed to register clock %s\n",
+			       __func__, clks[i].name);
+			continue;
+		}
+		if (clks[i].id)
+			unit->clk_table[clks[i].id] = clk;
+	}
+}
+
+void mmp_register_general_gate_clks(struct mmp_clk_unit *unit,
+				struct mmp_param_general_gate_clk *clks,
+				void __iomem *base, int size)
+{
+	struct clk *clk;
+	int i;
+
+	for (i = 0; i < size; i++) {
+		clk = clk_register_gate(NULL, clks[i].name,
+					clks[i].parent_name,
+					clks[i].flags,
+					base + clks[i].offset,
+					clks[i].bit_idx,
+					clks[i].gate_flags,
+					clks[i].lock);
+
+		if (IS_ERR(clk)) {
+			pr_err("%s: failed to register clock %s\n",
+			       __func__, clks[i].name);
+			continue;
+		}
+		if (clks[i].id)
+			unit->clk_table[clks[i].id] = clk;
+	}
+}
+
+void mmp_register_gate_clks(struct mmp_clk_unit *unit,
+			struct mmp_param_gate_clk *clks,
+			void __iomem *base, int size)
+{
+	struct clk *clk;
+	int i;
+
+	for (i = 0; i < size; i++) {
+		clk = mmp_clk_register_gate(NULL, clks[i].name,
+					clks[i].parent_name,
+					clks[i].flags,
+					base + clks[i].offset,
+					clks[i].mask,
+					clks[i].val_enable,
+					clks[i].val_disable,
+					clks[i].gate_flags,
+					clks[i].lock);
+
+		if (IS_ERR(clk)) {
+			pr_err("%s: failed to register clock %s\n",
+			       __func__, clks[i].name);
+			continue;
+		}
+		if (clks[i].id)
+			unit->clk_table[clks[i].id] = clk;
+	}
+}
+
+void mmp_register_mux_clks(struct mmp_clk_unit *unit,
+			struct mmp_param_mux_clk *clks,
+			void __iomem *base, int size)
+{
+	struct clk *clk;
+	int i;
+
+	for (i = 0; i < size; i++) {
+		clk = clk_register_mux(NULL, clks[i].name,
+					clks[i].parent_name,
+					clks[i].num_parents,
+					clks[i].flags,
+					base + clks[i].offset,
+					clks[i].shift,
+					clks[i].width,
+					clks[i].mux_flags,
+					clks[i].lock);
+
+		if (IS_ERR(clk)) {
+			pr_err("%s: failed to register clock %s\n",
+			       __func__, clks[i].name);
+			continue;
+		}
+		if (clks[i].id)
+			unit->clk_table[clks[i].id] = clk;
+	}
+}
+
+void mmp_register_div_clks(struct mmp_clk_unit *unit,
+			struct mmp_param_div_clk *clks,
+			void __iomem *base, int size)
+{
+	struct clk *clk;
+	int i;
+
+	for (i = 0; i < size; i++) {
+		clk = clk_register_divider(NULL, clks[i].name,
+					clks[i].parent_name,
+					clks[i].flags,
+					base + clks[i].offset,
+					clks[i].shift,
+					clks[i].width,
+					clks[i].div_flags,
+					clks[i].lock);
+
+		if (IS_ERR(clk)) {
+			pr_err("%s: failed to register clock %s\n",
+			       __func__, clks[i].name);
+			continue;
+		}
+		if (clks[i].id)
+			unit->clk_table[clks[i].id] = clk;
+	}
+}
+
+void mmp_clk_add(struct mmp_clk_unit *unit, unsigned int id,
+			struct clk *clk)
+{
+	if (IS_ERR_OR_NULL(clk)) {
+		pr_err("CLK %d has invalid pointer %p\n", id, clk);
+		return;
+	}
+	if (id > unit->nr_clks) {
+		pr_err("CLK %d is invalid\n", id);
+		return;
+	}
+
+	unit->clk_table[id] = clk;
+}
diff --git a/drivers/clk/mmp/clk.h b/drivers/clk/mmp/clk.h
index ab86dd4a416a..adf9b711b037 100644
--- a/drivers/clk/mmp/clk.h
+++ b/drivers/clk/mmp/clk.h
@@ -7,19 +7,123 @@
 #define APBC_NO_BUS_CTRL	BIT(0)
 #define APBC_POWER_CTRL		BIT(1)
 
-struct clk_factor_masks {
-	unsigned int	factor;
-	unsigned int	num_mask;
-	unsigned int	den_mask;
-	unsigned int	num_shift;
-	unsigned int	den_shift;
+
+/* Clock type "factor" */
+struct mmp_clk_factor_masks {
+	unsigned int factor;
+	unsigned int num_mask;
+	unsigned int den_mask;
+	unsigned int num_shift;
+	unsigned int den_shift;
 };
 
-struct clk_factor_tbl {
+struct mmp_clk_factor_tbl {
 	unsigned int num;
 	unsigned int den;
 };
 
+struct mmp_clk_factor {
+	struct clk_hw hw;
+	void __iomem *base;
+	struct mmp_clk_factor_masks *masks;
+	struct mmp_clk_factor_tbl *ftbl;
+	unsigned int ftbl_cnt;
+	spinlock_t *lock;
+};
+
+extern struct clk *mmp_clk_register_factor(const char *name,
+		const char *parent_name, unsigned long flags,
+		void __iomem *base, struct mmp_clk_factor_masks *masks,
+		struct mmp_clk_factor_tbl *ftbl, unsigned int ftbl_cnt,
+		spinlock_t *lock);
+
+/* Clock type "mix" */
+#define MMP_CLK_BITS_MASK(width, shift)			\
+		(((1 << (width)) - 1) << (shift))
+#define MMP_CLK_BITS_GET_VAL(data, width, shift)	\
+		((data & MMP_CLK_BITS_MASK(width, shift)) >> (shift))
+#define MMP_CLK_BITS_SET_VAL(val, width, shift)		\
+		(((val) << (shift)) & MMP_CLK_BITS_MASK(width, shift))
+
+enum {
+	MMP_CLK_MIX_TYPE_V1,
+	MMP_CLK_MIX_TYPE_V2,
+	MMP_CLK_MIX_TYPE_V3,
+};
+
+/* The register layout */
+struct mmp_clk_mix_reg_info {
+	void __iomem *reg_clk_ctrl;
+	void __iomem *reg_clk_sel;
+	u8 width_div;
+	u8 shift_div;
+	u8 width_mux;
+	u8 shift_mux;
+	u8 bit_fc;
+};
+
+/* The suggested clock table from user. */
+struct mmp_clk_mix_clk_table {
+	unsigned long rate;
+	u8 parent_index;
+	unsigned int divisor;
+	unsigned int valid;
+};
+
+struct mmp_clk_mix_config {
+	struct mmp_clk_mix_reg_info reg_info;
+	struct mmp_clk_mix_clk_table *table;
+	unsigned int table_size;
+	u32 *mux_table;
+	struct clk_div_table *div_table;
+	u8 div_flags;
+	u8 mux_flags;
+};
+
+struct mmp_clk_mix {
+	struct clk_hw hw;
+	struct mmp_clk_mix_reg_info reg_info;
+	struct mmp_clk_mix_clk_table *table;
+	u32 *mux_table;
+	struct clk_div_table *div_table;
+	unsigned int table_size;
+	u8 div_flags;
+	u8 mux_flags;
+	unsigned int type;
+	spinlock_t *lock;
+};
+
+extern const struct clk_ops mmp_clk_mix_ops;
+extern struct clk *mmp_clk_register_mix(struct device *dev,
+					const char *name,
+					const char **parent_names,
+					u8 num_parents,
+					unsigned long flags,
+					struct mmp_clk_mix_config *config,
+					spinlock_t *lock);
+
+
+/* Clock type "gate". MMP private gate */
+#define MMP_CLK_GATE_NEED_DELAY		BIT(0)
+
+struct mmp_clk_gate {
+	struct clk_hw hw;
+	void __iomem *reg;
+	u32 mask;
+	u32 val_enable;
+	u32 val_disable;
+	unsigned int flags;
+	spinlock_t *lock;
+};
+
+extern const struct clk_ops mmp_clk_gate_ops;
+extern struct clk *mmp_clk_register_gate(struct device *dev, const char *name,
+			const char *parent_name, unsigned long flags,
+			void __iomem *reg, u32 mask, u32 val_enable,
+			u32 val_disable, unsigned int gate_flags,
+			spinlock_t *lock);
+
+
 extern struct clk *mmp_clk_register_pll2(const char *name,
 		const char *parent_name, unsigned long flags);
 extern struct clk *mmp_clk_register_apbc(const char *name,
@@ -28,8 +132,108 @@ extern struct clk *mmp_clk_register_apbc(const char *name,
 extern struct clk *mmp_clk_register_apmu(const char *name,
 		const char *parent_name, void __iomem *base, u32 enable_mask,
 		spinlock_t *lock);
-extern struct clk *mmp_clk_register_factor(const char *name,
-		const char *parent_name, unsigned long flags,
-		void __iomem *base, struct clk_factor_masks *masks,
-		struct clk_factor_tbl *ftbl, unsigned int ftbl_cnt);
+
+struct mmp_clk_unit {
+	unsigned int nr_clks;
+	struct clk **clk_table;
+	struct clk_onecell_data clk_data;
+};
+
+struct mmp_param_fixed_rate_clk {
+	unsigned int id;
+	char *name;
+	const char *parent_name;
+	unsigned long flags;
+	unsigned long fixed_rate;
+};
+void mmp_register_fixed_rate_clks(struct mmp_clk_unit *unit,
+				struct mmp_param_fixed_rate_clk *clks,
+				int size);
+
+struct mmp_param_fixed_factor_clk {
+	unsigned int id;
+	char *name;
+	const char *parent_name;
+	unsigned long mult;
+	unsigned long div;
+	unsigned long flags;
+};
+void mmp_register_fixed_factor_clks(struct mmp_clk_unit *unit,
+				struct mmp_param_fixed_factor_clk *clks,
+				int size);
+
+struct mmp_param_general_gate_clk {
+	unsigned int id;
+	const char *name;
+	const char *parent_name;
+	unsigned long flags;
+	unsigned long offset;
+	u8 bit_idx;
+	u8 gate_flags;
+	spinlock_t *lock;
+};
+void mmp_register_general_gate_clks(struct mmp_clk_unit *unit,
+				struct mmp_param_general_gate_clk *clks,
+				void __iomem *base, int size);
+
+struct mmp_param_gate_clk {
+	unsigned int id;
+	char *name;
+	const char *parent_name;
+	unsigned long flags;
+	unsigned long offset;
+	u32 mask;
+	u32 val_enable;
+	u32 val_disable;
+	unsigned int gate_flags;
+	spinlock_t *lock;
+};
+void mmp_register_gate_clks(struct mmp_clk_unit *unit,
+			struct mmp_param_gate_clk *clks,
+			void __iomem *base, int size);
+
+struct mmp_param_mux_clk {
+	unsigned int id;
+	char *name;
+	const char **parent_name;
+	u8 num_parents;
+	unsigned long flags;
+	unsigned long offset;
+	u8 shift;
+	u8 width;
+	u8 mux_flags;
+	spinlock_t *lock;
+};
+void mmp_register_mux_clks(struct mmp_clk_unit *unit,
+			struct mmp_param_mux_clk *clks,
+			void __iomem *base, int size);
+
+struct mmp_param_div_clk {
+	unsigned int id;
+	char *name;
+	const char *parent_name;
+	unsigned long flags;
+	unsigned long offset;
+	u8 shift;
+	u8 width;
+	u8 div_flags;
+	spinlock_t *lock;
+};
+void mmp_register_div_clks(struct mmp_clk_unit *unit,
+			struct mmp_param_div_clk *clks,
+			void __iomem *base, int size);
+
+#define DEFINE_MIX_REG_INFO(w_d, s_d, w_m, s_m, fc)	\
+{							\
+	.width_div = (w_d),				\
+	.shift_div = (s_d),				\
+	.width_mux = (w_m),				\
+	.shift_mux = (s_m),				\
+	.bit_fc = (fc),					\
+}
+
+void mmp_clk_init(struct device_node *np, struct mmp_clk_unit *unit,
+		int nr_clks);
+void mmp_clk_add(struct mmp_clk_unit *unit, unsigned int id,
+		struct clk *clk);
 #endif
diff --git a/drivers/clk/mmp/reset.c b/drivers/clk/mmp/reset.c
new file mode 100644
index 000000000000..b54da1fe73f0
--- /dev/null
+++ b/drivers/clk/mmp/reset.c
@@ -0,0 +1,99 @@
+#include <linux/slab.h>
+#include <linux/io.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/reset-controller.h>
+
+#include "reset.h"
+
+#define rcdev_to_unit(rcdev) container_of(rcdev, struct mmp_clk_reset_unit, rcdev)
+
+static int mmp_of_reset_xlate(struct reset_controller_dev *rcdev,
+			  const struct of_phandle_args *reset_spec)
+{
+	struct mmp_clk_reset_unit *unit = rcdev_to_unit(rcdev);
+	struct mmp_clk_reset_cell *cell;
+	int i;
+
+	if (WARN_ON(reset_spec->args_count != rcdev->of_reset_n_cells))
+		return -EINVAL;
+
+	for (i = 0; i < rcdev->nr_resets; i++) {
+		cell = &unit->cells[i];
+		if (cell->clk_id == reset_spec->args[0])
+			break;
+	}
+
+	if (i == rcdev->nr_resets)
+		return -EINVAL;
+
+	return i;
+}
+
+static int mmp_clk_reset_assert(struct reset_controller_dev *rcdev,
+				unsigned long id)
+{
+	struct mmp_clk_reset_unit *unit = rcdev_to_unit(rcdev);
+	struct mmp_clk_reset_cell *cell;
+	unsigned long flags = 0;
+	u32 val;
+
+	cell = &unit->cells[id];
+	if (cell->lock)
+		spin_lock_irqsave(cell->lock, flags);
+
+	val = readl(cell->reg);
+	val |= cell->bits;
+	writel(val, cell->reg);
+
+	if (cell->lock)
+		spin_unlock_irqrestore(cell->lock, flags);
+
+	return 0;
+}
+
+static int mmp_clk_reset_deassert(struct reset_controller_dev *rcdev,
+				unsigned long id)
+{
+	struct mmp_clk_reset_unit *unit = rcdev_to_unit(rcdev);
+	struct mmp_clk_reset_cell *cell;
+	unsigned long flags = 0;
+	u32 val;
+
+	cell = &unit->cells[id];
+	if (cell->lock)
+		spin_lock_irqsave(cell->lock, flags);
+
+	val = readl(cell->reg);
+	val &= ~cell->bits;
+	writel(val, cell->reg);
+
+	if (cell->lock)
+		spin_unlock_irqrestore(cell->lock, flags);
+
+	return 0;
+}
+
+static struct reset_control_ops mmp_clk_reset_ops = {
+	.assert		= mmp_clk_reset_assert,
+	.deassert	= mmp_clk_reset_deassert,
+};
+
+void mmp_clk_reset_register(struct device_node *np,
+			struct mmp_clk_reset_cell *cells, int nr_resets)
+{
+	struct mmp_clk_reset_unit *unit;
+
+	unit = kzalloc(sizeof(*unit), GFP_KERNEL);
+	if (!unit)
+		return;
+
+	unit->cells = cells;
+	unit->rcdev.of_reset_n_cells = 1;
+	unit->rcdev.nr_resets = nr_resets;
+	unit->rcdev.ops = &mmp_clk_reset_ops;
+	unit->rcdev.of_node = np;
+	unit->rcdev.of_xlate = mmp_of_reset_xlate;
+
+	reset_controller_register(&unit->rcdev);
+}
diff --git a/drivers/clk/mmp/reset.h b/drivers/clk/mmp/reset.h
new file mode 100644
index 000000000000..be8b1a7000f7
--- /dev/null
+++ b/drivers/clk/mmp/reset.h
@@ -0,0 +1,31 @@
+#ifndef __MACH_MMP_CLK_RESET_H
+#define __MACH_MMP_CLK_RESET_H
+
+#include <linux/reset-controller.h>
+
+#define MMP_RESET_INVERT	1
+
+struct mmp_clk_reset_cell {
+	unsigned int clk_id;
+	void __iomem *reg;
+	u32 bits;
+	unsigned int flags;
+	spinlock_t *lock;
+};
+
+struct mmp_clk_reset_unit {
+	struct reset_controller_dev rcdev;
+	struct mmp_clk_reset_cell *cells;
+};
+
+#ifdef CONFIG_RESET_CONTROLLER
+void mmp_clk_reset_register(struct device_node *np,
+			struct mmp_clk_reset_cell *cells, int nr_resets);
+#else
+static inline void mmp_clk_reset_register(struct device_node *np,
+			struct mmp_clk_reset_cell *cells, int nr_resets)
+{
+}
+#endif
+
+#endif
diff --git a/drivers/clk/mvebu/armada-370.c b/drivers/clk/mvebu/armada-370.c
index bef198a83863..756f0f39d6a3 100644
--- a/drivers/clk/mvebu/armada-370.c
+++ b/drivers/clk/mvebu/armada-370.c
@@ -23,6 +23,7 @@
  */
 
 #define SARL				0	/* Low part [0:31] */
+#define	 SARL_A370_SSCG_ENABLE		BIT(10)
 #define	 SARL_A370_PCLK_FREQ_OPT	11
 #define	 SARL_A370_PCLK_FREQ_OPT_MASK	0xF
 #define	 SARL_A370_FAB_FREQ_OPT		15
@@ -133,10 +134,17 @@ static void __init a370_get_clk_ratio(
 	}
 }
 
+static bool a370_is_sscg_enabled(void __iomem *sar)
+{
+	return !(readl(sar) & SARL_A370_SSCG_ENABLE);
+}
+
 static const struct coreclk_soc_desc a370_coreclks = {
 	.get_tclk_freq = a370_get_tclk_freq,
 	.get_cpu_freq = a370_get_cpu_freq,
 	.get_clk_ratio = a370_get_clk_ratio,
+	.is_sscg_enabled = a370_is_sscg_enabled,
+	.fix_sscg_deviation = kirkwood_fix_sscg_deviation,
 	.ratios = a370_coreclk_ratios,
 	.num_ratios = ARRAY_SIZE(a370_coreclk_ratios),
 };
diff --git a/drivers/clk/mvebu/armada-375.c b/drivers/clk/mvebu/armada-375.c
index c991a4d95e10..c7af2242b796 100644
--- a/drivers/clk/mvebu/armada-375.c
+++ b/drivers/clk/mvebu/armada-375.c
@@ -27,14 +27,14 @@
  * all modified at the same time, and not separately as for the Armada
  * 370 or the Armada XP SoCs.
  *
- * SAR0[21:17]   : CPU frequency    DDR frequency   L2 frequency
+ * SAR1[21:17]   : CPU frequency    DDR frequency   L2 frequency
  *		 6   =  400 MHz	    400 MHz	    200 MHz
  *		 15  =  600 MHz	    600 MHz	    300 MHz
  *		 21  =  800 MHz	    534 MHz	    400 MHz
  *		 25  = 1000 MHz	    500 MHz	    500 MHz
  *		 others reserved.
  *
- * SAR0[22]   : TCLK frequency
+ * SAR1[22]   : TCLK frequency
  *		 0 = 166 MHz
  *		 1 = 200 MHz
  */
diff --git a/drivers/clk/mvebu/common.c b/drivers/clk/mvebu/common.c
index 25ceccf939ad..0d4d1216f2dd 100644
--- a/drivers/clk/mvebu/common.c
+++ b/drivers/clk/mvebu/common.c
@@ -19,6 +19,7 @@
 #include <linux/io.h>
 #include <linux/of.h>
 #include <linux/of_address.h>
+#include <linux/syscore_ops.h>
 
 #include "common.h"
 
@@ -26,8 +27,85 @@
  * Core Clocks
  */
 
+#define SSCG_CONF_MODE(reg)	(((reg) >> 16) & 0x3)
+#define SSCG_SPREAD_DOWN	0x0
+#define SSCG_SPREAD_UP		0x1
+#define SSCG_SPREAD_CENTRAL	0x2
+#define SSCG_CONF_LOW(reg)	(((reg) >> 8) & 0xFF)
+#define SSCG_CONF_HIGH(reg)	((reg) & 0xFF)
+
 static struct clk_onecell_data clk_data;
 
+/*
+ * This function can be used by the Kirkwood, the Armada 370, the
+ * Armada XP and the Armada 375 SoC. The name of the function was
+ * chosen following the dt convention: using the first known SoC
+ * compatible with it.
+ */
+u32 kirkwood_fix_sscg_deviation(u32 system_clk)
+{
+	struct device_node *sscg_np = NULL;
+	void __iomem *sscg_map;
+	u32 sscg_reg;
+	s32 low_bound, high_bound;
+	u64 freq_swing_half;
+
+	sscg_np = of_find_node_by_name(NULL, "sscg");
+	if (sscg_np == NULL) {
+		pr_err("cannot get SSCG register node\n");
+		return system_clk;
+	}
+
+	sscg_map = of_iomap(sscg_np, 0);
+	if (sscg_map == NULL) {
+		pr_err("cannot map SSCG register\n");
+		goto out;
+	}
+
+	sscg_reg = readl(sscg_map);
+	high_bound = SSCG_CONF_HIGH(sscg_reg);
+	low_bound = SSCG_CONF_LOW(sscg_reg);
+
+	if ((high_bound - low_bound) <= 0)
+		goto out;
+	/*
+	 * From Marvell engineer we got the following formula (when
+	 * this code was written, the datasheet was erroneous)
+	 * Spread percentage = 1/96 * (H - L) / H
+	 * H = SSCG_High_Boundary
+	 * L = SSCG_Low_Boundary
+	 *
+	 * As the deviation is half of spread then it lead to the
+	 * following formula in the code.
+	 *
+	 * To avoid an overflow and not lose any significant digit in
+	 * the same time we have to use a 64 bit integer.
+	 */
+
+	freq_swing_half = (((u64)high_bound - (u64)low_bound)
+			* (u64)system_clk);
+	do_div(freq_swing_half, (2 * 96 * high_bound));
+
+	switch (SSCG_CONF_MODE(sscg_reg)) {
+	case SSCG_SPREAD_DOWN:
+		system_clk -= freq_swing_half;
+		break;
+	case SSCG_SPREAD_UP:
+		system_clk += freq_swing_half;
+		break;
+	case SSCG_SPREAD_CENTRAL:
+	default:
+		break;
+	}
+
+	iounmap(sscg_map);
+
+out:
+	of_node_put(sscg_np);
+
+	return system_clk;
+}
+
 void __init mvebu_coreclk_setup(struct device_node *np,
 				const struct coreclk_soc_desc *desc)
 {
@@ -62,6 +140,11 @@ void __init mvebu_coreclk_setup(struct device_node *np,
 	of_property_read_string_index(np, "clock-output-names", 1,
 				      &cpuclk_name);
 	rate = desc->get_cpu_freq(base);
+
+	if (desc->is_sscg_enabled && desc->fix_sscg_deviation
+		&& desc->is_sscg_enabled(base))
+		rate = desc->fix_sscg_deviation(rate);
+
 	clk_data.clks[1] = clk_register_fixed_rate(NULL, cpuclk_name, NULL,
 						   CLK_IS_ROOT, rate);
 	WARN_ON(IS_ERR(clk_data.clks[1]));
@@ -89,18 +172,23 @@ void __init mvebu_coreclk_setup(struct device_node *np,
  * Clock Gating Control
  */
 
+DEFINE_SPINLOCK(ctrl_gating_lock);
+
 struct clk_gating_ctrl {
-	spinlock_t lock;
+	spinlock_t *lock;
 	struct clk **gates;
 	int num_gates;
+	void __iomem *base;
+	u32 saved_reg;
 };
 
 #define to_clk_gate(_hw) container_of(_hw, struct clk_gate, hw)
 
+static struct clk_gating_ctrl *ctrl;
+
 static struct clk *clk_gating_get_src(
 	struct of_phandle_args *clkspec, void *data)
 {
-	struct clk_gating_ctrl *ctrl = (struct clk_gating_ctrl *)data;
 	int n;
 
 	if (clkspec->args_count < 1)
@@ -115,15 +203,35 @@ static struct clk *clk_gating_get_src(
 	return ERR_PTR(-ENODEV);
 }
 
+static int mvebu_clk_gating_suspend(void)
+{
+	ctrl->saved_reg = readl(ctrl->base);
+	return 0;
+}
+
+static void mvebu_clk_gating_resume(void)
+{
+	writel(ctrl->saved_reg, ctrl->base);
+}
+
+static struct syscore_ops clk_gate_syscore_ops = {
+	.suspend = mvebu_clk_gating_suspend,
+	.resume = mvebu_clk_gating_resume,
+};
+
 void __init mvebu_clk_gating_setup(struct device_node *np,
 				   const struct clk_gating_soc_desc *desc)
 {
-	struct clk_gating_ctrl *ctrl;
 	struct clk *clk;
 	void __iomem *base;
 	const char *default_parent = NULL;
 	int n;
 
+	if (ctrl) {
+		pr_err("mvebu-clk-gating: cannot instantiate more than one gatable clock device\n");
+		return;
+	}
+
 	base = of_iomap(np, 0);
 	if (WARN_ON(!base))
 		return;
@@ -138,7 +246,10 @@ void __init mvebu_clk_gating_setup(struct device_node *np,
 	if (WARN_ON(!ctrl))
 		goto ctrl_out;
 
-	spin_lock_init(&ctrl->lock);
+	/* lock must already be initialized */
+	ctrl->lock = &ctrl_gating_lock;
+
+	ctrl->base = base;
 
 	/* Count, allocate, and register clock gates */
 	for (n = 0; desc[n].name;)
@@ -155,12 +266,14 @@ void __init mvebu_clk_gating_setup(struct device_node *np,
 			(desc[n].parent) ? desc[n].parent : default_parent;
 		ctrl->gates[n] = clk_register_gate(NULL, desc[n].name, parent,
 					desc[n].flags, base, desc[n].bit_idx,
-					0, &ctrl->lock);
+					0, ctrl->lock);
 		WARN_ON(IS_ERR(ctrl->gates[n]));
 	}
 
 	of_clk_add_provider(np, clk_gating_get_src, ctrl);
 
+	register_syscore_ops(&clk_gate_syscore_ops);
+
 	return;
 gates_out:
 	kfree(ctrl);
diff --git a/drivers/clk/mvebu/common.h b/drivers/clk/mvebu/common.h
index f968b4d9df92..783b5631a453 100644
--- a/drivers/clk/mvebu/common.h
+++ b/drivers/clk/mvebu/common.h
@@ -17,6 +17,8 @@
 
 #include <linux/kernel.h>
 
+extern spinlock_t ctrl_gating_lock;
+
 struct device_node;
 
 struct coreclk_ratio {
@@ -28,6 +30,8 @@ struct coreclk_soc_desc {
 	u32 (*get_tclk_freq)(void __iomem *sar);
 	u32 (*get_cpu_freq)(void __iomem *sar);
 	void (*get_clk_ratio)(void __iomem *sar, int id, int *mult, int *div);
+	bool (*is_sscg_enabled)(void __iomem *sar);
+	u32 (*fix_sscg_deviation)(u32 system_clk);
 	const struct coreclk_ratio *ratios;
 	int num_ratios;
 };
@@ -45,4 +49,9 @@ void __init mvebu_coreclk_setup(struct device_node *np,
 void __init mvebu_clk_gating_setup(struct device_node *np,
 				   const struct clk_gating_soc_desc *desc);
 
+/*
+ * This function is shared among the Kirkwood, Armada 370, Armada XP
+ * and Armada 375 SoC
+ */
+u32 kirkwood_fix_sscg_deviation(u32 system_clk);
 #endif
diff --git a/drivers/clk/mvebu/kirkwood.c b/drivers/clk/mvebu/kirkwood.c
index ddb666a86500..99550f25975e 100644
--- a/drivers/clk/mvebu/kirkwood.c
+++ b/drivers/clk/mvebu/kirkwood.c
@@ -13,9 +13,11 @@
  */
 
 #include <linux/kernel.h>
+#include <linux/slab.h>
 #include <linux/clk-provider.h>
 #include <linux/io.h>
 #include <linux/of.h>
+#include <linux/of_address.h>
 #include "common.h"
 
 /*
@@ -214,7 +216,6 @@ static const struct clk_gating_soc_desc kirkwood_gating_desc[] __initconst = {
 	{ "runit", NULL, 7, 0 },
 	{ "xor0", NULL, 8, 0 },
 	{ "audio", NULL, 9, 0 },
-	{ "powersave", "cpuclk", 11, 0 },
 	{ "sata0", NULL, 14, 0 },
 	{ "sata1", NULL, 15, 0 },
 	{ "xor1", NULL, 16, 0 },
@@ -225,6 +226,101 @@ static const struct clk_gating_soc_desc kirkwood_gating_desc[] __initconst = {
 	{ }
 };
 
+
+/*
+ * Clock Muxing Control
+ */
+
+struct clk_muxing_soc_desc {
+	const char *name;
+	const char **parents;
+	int num_parents;
+	int shift;
+	int width;
+	unsigned long flags;
+};
+
+struct clk_muxing_ctrl {
+	spinlock_t *lock;
+	struct clk **muxes;
+	int num_muxes;
+};
+
+static const char *powersave_parents[] = {
+	"cpuclk",
+	"ddrclk",
+};
+
+static const struct clk_muxing_soc_desc kirkwood_mux_desc[] __initconst = {
+	{ "powersave", powersave_parents, ARRAY_SIZE(powersave_parents),
+		11, 1, 0 },
+};
+
+#define to_clk_mux(_hw) container_of(_hw, struct clk_mux, hw)
+
+static struct clk *clk_muxing_get_src(
+	struct of_phandle_args *clkspec, void *data)
+{
+	struct clk_muxing_ctrl *ctrl = (struct clk_muxing_ctrl *)data;
+	int n;
+
+	if (clkspec->args_count < 1)
+		return ERR_PTR(-EINVAL);
+
+	for (n = 0; n < ctrl->num_muxes; n++) {
+		struct clk_mux *mux =
+			to_clk_mux(__clk_get_hw(ctrl->muxes[n]));
+		if (clkspec->args[0] == mux->shift)
+			return ctrl->muxes[n];
+	}
+	return ERR_PTR(-ENODEV);
+}
+
+static void __init kirkwood_clk_muxing_setup(struct device_node *np,
+				   const struct clk_muxing_soc_desc *desc)
+{
+	struct clk_muxing_ctrl *ctrl;
+	void __iomem *base;
+	int n;
+
+	base = of_iomap(np, 0);
+	if (WARN_ON(!base))
+		return;
+
+	ctrl = kzalloc(sizeof(*ctrl), GFP_KERNEL);
+	if (WARN_ON(!ctrl))
+		goto ctrl_out;
+
+	/* lock must already be initialized */
+	ctrl->lock = &ctrl_gating_lock;
+
+	/* Count, allocate, and register clock muxes */
+	for (n = 0; desc[n].name;)
+		n++;
+
+	ctrl->num_muxes = n;
+	ctrl->muxes = kcalloc(ctrl->num_muxes, sizeof(struct clk *),
+			GFP_KERNEL);
+	if (WARN_ON(!ctrl->muxes))
+		goto muxes_out;
+
+	for (n = 0; n < ctrl->num_muxes; n++) {
+		ctrl->muxes[n] = clk_register_mux(NULL, desc[n].name,
+				desc[n].parents, desc[n].num_parents,
+				desc[n].flags, base, desc[n].shift,
+				desc[n].width, desc[n].flags, ctrl->lock);
+		WARN_ON(IS_ERR(ctrl->muxes[n]));
+	}
+
+	of_clk_add_provider(np, clk_muxing_get_src, ctrl);
+
+	return;
+muxes_out:
+	kfree(ctrl);
+ctrl_out:
+	iounmap(base);
+}
+
 static void __init kirkwood_clk_init(struct device_node *np)
 {
 	struct device_node *cgnp =
@@ -236,8 +332,10 @@ static void __init kirkwood_clk_init(struct device_node *np)
 	else
 		mvebu_coreclk_setup(np, &kirkwood_coreclks);
 
-	if (cgnp)
+	if (cgnp) {
 		mvebu_clk_gating_setup(cgnp, kirkwood_gating_desc);
+		kirkwood_clk_muxing_setup(cgnp, kirkwood_mux_desc);
+	}
 }
 CLK_OF_DECLARE(kirkwood_clk, "marvell,kirkwood-core-clock",
 	       kirkwood_clk_init);
diff --git a/drivers/clk/pxa/Makefile b/drivers/clk/pxa/Makefile
new file mode 100644
index 000000000000..38e915344605
--- /dev/null
+++ b/drivers/clk/pxa/Makefile
@@ -0,0 +1,3 @@
+obj-y				+= clk-pxa.o
+obj-$(CONFIG_PXA25x)		+= clk-pxa25x.o
+obj-$(CONFIG_PXA27x)		+= clk-pxa27x.o
diff --git a/drivers/clk/pxa/clk-pxa.c b/drivers/clk/pxa/clk-pxa.c
new file mode 100644
index 000000000000..4e834753ab09
--- /dev/null
+++ b/drivers/clk/pxa/clk-pxa.c
@@ -0,0 +1,108 @@
+/*
+ * Marvell PXA family clocks
+ *
+ * Copyright (C) 2014 Robert Jarzmik
+ *
+ * Common clock code for PXA clocks ("CKEN" type clocks + DT)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ */
+#include <linux/clk.h>
+#include <linux/clk-provider.h>
+#include <linux/clkdev.h>
+#include <linux/of.h>
+
+#include <dt-bindings/clock/pxa-clock.h>
+#include "clk-pxa.h"
+
+DEFINE_SPINLOCK(lock);
+
+static struct clk *pxa_clocks[CLK_MAX];
+static struct clk_onecell_data onecell_data = {
+	.clks = pxa_clocks,
+	.clk_num = CLK_MAX,
+};
+
+struct pxa_clk {
+	struct clk_hw hw;
+	struct clk_fixed_factor lp;
+	struct clk_fixed_factor hp;
+	struct clk_gate gate;
+	bool (*is_in_low_power)(void);
+};
+
+#define to_pxa_clk(_hw) container_of(_hw, struct pxa_clk, hw)
+
+static unsigned long cken_recalc_rate(struct clk_hw *hw,
+				      unsigned long parent_rate)
+{
+	struct pxa_clk *pclk = to_pxa_clk(hw);
+	struct clk_fixed_factor *fix;
+
+	if (!pclk->is_in_low_power || pclk->is_in_low_power())
+		fix = &pclk->lp;
+	else
+		fix = &pclk->hp;
+	fix->hw.clk = hw->clk;
+	return clk_fixed_factor_ops.recalc_rate(&fix->hw, parent_rate);
+}
+
+static struct clk_ops cken_rate_ops = {
+	.recalc_rate = cken_recalc_rate,
+};
+
+static u8 cken_get_parent(struct clk_hw *hw)
+{
+	struct pxa_clk *pclk = to_pxa_clk(hw);
+
+	if (!pclk->is_in_low_power)
+		return 0;
+	return pclk->is_in_low_power() ? 0 : 1;
+}
+
+static struct clk_ops cken_mux_ops = {
+	.get_parent = cken_get_parent,
+	.set_parent = dummy_clk_set_parent,
+};
+
+void __init clkdev_pxa_register(int ckid, const char *con_id,
+				const char *dev_id, struct clk *clk)
+{
+	if (!IS_ERR(clk) && (ckid != CLK_NONE))
+		pxa_clocks[ckid] = clk;
+	if (!IS_ERR(clk))
+		clk_register_clkdev(clk, con_id, dev_id);
+}
+
+int __init clk_pxa_cken_init(const struct desc_clk_cken *clks, int nb_clks)
+{
+	int i;
+	struct pxa_clk *pxa_clk;
+	struct clk *clk;
+
+	for (i = 0; i < nb_clks; i++) {
+		pxa_clk = kzalloc(sizeof(*pxa_clk), GFP_KERNEL);
+		pxa_clk->is_in_low_power = clks[i].is_in_low_power;
+		pxa_clk->lp = clks[i].lp;
+		pxa_clk->hp = clks[i].hp;
+		pxa_clk->gate = clks[i].gate;
+		pxa_clk->gate.lock = &lock;
+		clk = clk_register_composite(NULL, clks[i].name,
+					     clks[i].parent_names, 2,
+					     &pxa_clk->hw, &cken_mux_ops,
+					     &pxa_clk->hw, &cken_rate_ops,
+					     &pxa_clk->gate.hw, &clk_gate_ops,
+					     clks[i].flags);
+		clkdev_pxa_register(clks[i].ckid, clks[i].con_id,
+				    clks[i].dev_id, clk);
+	}
+	return 0;
+}
+
+void __init clk_pxa_dt_common_init(struct device_node *np)
+{
+	of_clk_add_provider(np, of_clk_src_onecell_get, &onecell_data);
+}
diff --git a/drivers/clk/pxa/clk-pxa.h b/drivers/clk/pxa/clk-pxa.h
new file mode 100644
index 000000000000..323965430111
--- /dev/null
+++ b/drivers/clk/pxa/clk-pxa.h
@@ -0,0 +1,108 @@
+/*
+ * Marvell PXA family clocks
+ *
+ * Copyright (C) 2014 Robert Jarzmik
+ *
+ * Common clock code for PXA clocks ("CKEN" type clocks + DT)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ */
+#ifndef _CLK_PXA_
+#define _CLK_PXA_
+
+#define PARENTS(name) \
+	static const char *name ## _parents[] __initconst
+#define MUX_RO_RATE_RO_OPS(name, clk_name)			\
+	static struct clk_hw name ## _mux_hw;			\
+	static struct clk_hw name ## _rate_hw;			\
+	static struct clk_ops name ## _mux_ops = {		\
+		.get_parent = name ## _get_parent,		\
+		.set_parent = dummy_clk_set_parent,		\
+	};							\
+	static struct clk_ops name ## _rate_ops = {		\
+		.recalc_rate = name ## _get_rate,		\
+	};							\
+	static struct clk * __init clk_register_ ## name(void)	\
+	{							\
+		return clk_register_composite(NULL, clk_name,	\
+			name ## _parents,			\
+			ARRAY_SIZE(name ## _parents),		\
+			&name ## _mux_hw, &name ## _mux_ops,	\
+			&name ## _rate_hw, &name ## _rate_ops,	\
+			NULL, NULL, CLK_GET_RATE_NOCACHE);	\
+	}
+
+#define RATE_RO_OPS(name, clk_name)			\
+	static struct clk_hw name ## _rate_hw;			\
+	static struct clk_ops name ## _rate_ops = {		\
+		.recalc_rate = name ## _get_rate,		\
+	};							\
+	static struct clk * __init clk_register_ ## name(void)	\
+	{							\
+		return clk_register_composite(NULL, clk_name,	\
+			name ## _parents,			\
+			ARRAY_SIZE(name ## _parents),		\
+			NULL, NULL,				\
+			&name ## _rate_hw, &name ## _rate_ops,	\
+			NULL, NULL, CLK_GET_RATE_NOCACHE);	\
+	}
+
+/*
+ * CKEN clock type
+ * This clock takes it source from 2 possible parents :
+ *  - a low power parent
+ *  - a normal parent
+ *
+ *  +------------+     +-----------+
+ *  |  Low Power | --- | x mult_lp |
+ *  |    Clock   |     | / div_lp  |\
+ *  +------------+     +-----------+ \+-----+   +-----------+
+ *                                    | Mux |---| CKEN gate |
+ *  +------------+     +-----------+ /+-----+   +-----------+
+ *  | High Power |     | x mult_hp |/
+ *  |    Clock   | --- | / div_hp  |
+ *  +------------+     +-----------+
+ */
+struct desc_clk_cken {
+	struct clk_hw hw;
+	int ckid;
+	const char *name;
+	const char *dev_id;
+	const char *con_id;
+	const char **parent_names;
+	struct clk_fixed_factor lp;
+	struct clk_fixed_factor hp;
+	struct clk_gate gate;
+	bool (*is_in_low_power)(void);
+	const unsigned long flags;
+};
+
+#define PXA_CKEN(_dev_id, _con_id, _name, parents, _mult_lp, _div_lp,	\
+		 _mult_hp, _div_hp, is_lp, _cken_reg, _cken_bit, flag)	\
+	{ .ckid = CLK_ ## _name, .name = #_name,			\
+	  .dev_id = _dev_id, .con_id = _con_id,	.parent_names = parents,\
+	  .lp = { .mult = _mult_lp, .div = _div_lp },			\
+	  .hp = { .mult = _mult_hp, .div = _div_hp },			\
+	  .is_in_low_power = is_lp,					\
+	  .gate = { .reg = (void __iomem *)_cken_reg, .bit_idx = _cken_bit }, \
+	  .flags = flag,						\
+	}
+#define PXA_CKEN_1RATE(dev_id, con_id, name, parents, cken_reg,		\
+			    cken_bit, flag)				\
+	PXA_CKEN(dev_id, con_id, name, parents, 1, 1, 1, 1,		\
+		 NULL, cken_reg, cken_bit, flag)
+
+static int dummy_clk_set_parent(struct clk_hw *hw, u8 index)
+{
+	return 0;
+}
+
+extern void clkdev_pxa_register(int ckid, const char *con_id,
+				const char *dev_id, struct clk *clk);
+extern int clk_pxa_cken_init(const struct desc_clk_cken *clks, int nb_clks);
+void clk_pxa_dt_common_init(struct device_node *np);
+
+#endif
diff --git a/drivers/clk/pxa/clk-pxa25x.c b/drivers/clk/pxa/clk-pxa25x.c
new file mode 100644
index 000000000000..6cd88d963a7f
--- /dev/null
+++ b/drivers/clk/pxa/clk-pxa25x.c
@@ -0,0 +1,273 @@
+/*
+ * Marvell PXA25x family clocks
+ *
+ * Copyright (C) 2014 Robert Jarzmik
+ *
+ * Heavily inspired from former arch/arm/mach-pxa/pxa25x.c.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * For non-devicetree platforms. Once pxa is fully converted to devicetree, this
+ * should go away.
+ */
+#include <linux/clk-provider.h>
+#include <linux/clk.h>
+#include <linux/clkdev.h>
+#include <linux/io.h>
+#include <linux/of.h>
+#include <mach/pxa25x.h>
+#include <mach/pxa2xx-regs.h>
+
+#include <dt-bindings/clock/pxa-clock.h>
+#include "clk-pxa.h"
+
+#define KHz 1000
+#define MHz (1000 * 1000)
+
+enum {
+	PXA_CORE_RUN = 0,
+	PXA_CORE_TURBO,
+};
+
+/*
+ * Various clock factors driven by the CCCR register.
+ */
+
+/* Crystal Frequency to Memory Frequency Multiplier (L) */
+static unsigned char L_clk_mult[32] = { 0, 27, 32, 36, 40, 45, 0, };
+
+/* Memory Frequency to Run Mode Frequency Multiplier (M) */
+static unsigned char M_clk_mult[4] = { 0, 1, 2, 4 };
+
+/* Run Mode Frequency to Turbo Mode Frequency Multiplier (N) */
+/* Note: we store the value N * 2 here. */
+static unsigned char N2_clk_mult[8] = { 0, 0, 2, 3, 4, 0, 6, 0 };
+
+static const char * const get_freq_khz[] = {
+	"core", "run", "cpll", "memory"
+};
+
+/*
+ * Get the clock frequency as reflected by CCCR and the turbo flag.
+ * We assume these values have been applied via a fcs.
+ * If info is not 0 we also display the current settings.
+ */
+unsigned int pxa25x_get_clk_frequency_khz(int info)
+{
+	struct clk *clk;
+	unsigned long clks[5];
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(get_freq_khz); i++) {
+		clk = clk_get(NULL, get_freq_khz[i]);
+		if (IS_ERR(clk)) {
+			clks[i] = 0;
+		} else {
+			clks[i] = clk_get_rate(clk);
+			clk_put(clk);
+		}
+	}
+
+	if (info) {
+		pr_info("Run Mode clock: %ld.%02ldMHz\n",
+			clks[1] / 1000000, (clks[1] % 1000000) / 10000);
+		pr_info("Turbo Mode clock: %ld.%02ldMHz\n",
+			clks[2] / 1000000, (clks[2] % 1000000) / 10000);
+		pr_info("Memory clock: %ld.%02ldMHz\n",
+			clks[3] / 1000000, (clks[3] % 1000000) / 10000);
+	}
+
+	return (unsigned int)clks[0];
+}
+
+static unsigned long clk_pxa25x_memory_get_rate(struct clk_hw *hw,
+						unsigned long parent_rate)
+{
+	unsigned long cccr = CCCR;
+	unsigned int m = M_clk_mult[(cccr >> 5) & 0x03];
+
+	return parent_rate / m;
+}
+PARENTS(clk_pxa25x_memory) = { "run" };
+RATE_RO_OPS(clk_pxa25x_memory, "memory");
+
+PARENTS(pxa25x_pbus95) = { "ppll_95_85mhz", "ppll_95_85mhz" };
+PARENTS(pxa25x_pbus147) = { "ppll_147_46mhz", "ppll_147_46mhz" };
+PARENTS(pxa25x_osc3) = { "osc_3_6864mhz", "osc_3_6864mhz" };
+
+#define PXA25X_CKEN(dev_id, con_id, parents, mult, div,			\
+		    bit, is_lp, flags)					\
+	PXA_CKEN(dev_id, con_id, bit, parents, mult, div, mult, div,	\
+		 is_lp,  &CKEN, CKEN_ ## bit, flags)
+#define PXA25X_PBUS95_CKEN(dev_id, con_id, bit, mult_hp, div_hp, delay)	\
+	PXA25X_CKEN(dev_id, con_id, pxa25x_pbus95_parents, mult_hp,	\
+		    div_hp, bit, NULL, 0)
+#define PXA25X_PBUS147_CKEN(dev_id, con_id, bit, mult_hp, div_hp, delay)\
+	PXA25X_CKEN(dev_id, con_id, pxa25x_pbus147_parents, mult_hp,	\
+		    div_hp, bit, NULL, 0)
+#define PXA25X_OSC3_CKEN(dev_id, con_id, bit, mult_hp, div_hp, delay)	\
+	PXA25X_CKEN(dev_id, con_id, pxa25x_osc3_parents, mult_hp,	\
+		    div_hp, bit, NULL, 0)
+
+#define PXA25X_CKEN_1RATE(dev_id, con_id, bit, parents, delay)		\
+	PXA_CKEN_1RATE(dev_id, con_id, bit, parents,			\
+		       &CKEN, CKEN_ ## bit, 0)
+#define PXA25X_CKEN_1RATE_AO(dev_id, con_id, bit, parents, delay)	\
+	PXA_CKEN_1RATE(dev_id, con_id, bit, parents,			\
+		       &CKEN, CKEN_ ## bit, CLK_IGNORE_UNUSED)
+
+static struct desc_clk_cken pxa25x_clocks[] __initdata = {
+	PXA25X_PBUS95_CKEN("pxa2xx-mci.0", NULL, MMC, 1, 5, 0),
+	PXA25X_PBUS95_CKEN("pxa2xx-i2c.0", NULL, I2C, 1, 3, 0),
+	PXA25X_PBUS95_CKEN("pxa2xx-ir", "FICPCLK", FICP, 1, 2, 0),
+	PXA25X_PBUS95_CKEN("pxa25x-udc", NULL, USB, 1, 2, 5),
+	PXA25X_PBUS147_CKEN("pxa2xx-uart.0", NULL, FFUART, 1, 10, 1),
+	PXA25X_PBUS147_CKEN("pxa2xx-uart.1", NULL, BTUART, 1, 10, 1),
+	PXA25X_PBUS147_CKEN("pxa2xx-uart.2", NULL, STUART, 1, 10, 1),
+	PXA25X_PBUS147_CKEN("pxa2xx-uart.3", NULL, HWUART, 1, 10, 1),
+	PXA25X_PBUS147_CKEN("pxa2xx-i2s", NULL, I2S, 1, 10, 0),
+	PXA25X_PBUS147_CKEN(NULL, "AC97CLK", AC97, 1, 12, 0),
+	PXA25X_OSC3_CKEN("pxa25x-ssp.0", NULL, SSP, 1, 1, 0),
+	PXA25X_OSC3_CKEN("pxa25x-nssp.1", NULL, NSSP, 1, 1, 0),
+	PXA25X_OSC3_CKEN("pxa25x-nssp.2", NULL, ASSP, 1, 1, 0),
+	PXA25X_OSC3_CKEN("pxa25x-pwm.0", NULL, PWM0, 1, 1, 0),
+	PXA25X_OSC3_CKEN("pxa25x-pwm.1", NULL, PWM1, 1, 1, 0),
+
+	PXA25X_CKEN_1RATE("pxa2xx-fb", NULL, LCD, clk_pxa25x_memory_parents, 0),
+	PXA25X_CKEN_1RATE_AO("pxa2xx-pcmcia", NULL, MEMC,
+			     clk_pxa25x_memory_parents, 0),
+};
+
+static u8 clk_pxa25x_core_get_parent(struct clk_hw *hw)
+{
+	unsigned long clkcfg;
+	unsigned int t;
+
+	asm("mrc\tp14, 0, %0, c6, c0, 0" : "=r" (clkcfg));
+	t  = clkcfg & (1 << 0);
+	if (t)
+		return PXA_CORE_TURBO;
+	return PXA_CORE_RUN;
+}
+
+static unsigned long clk_pxa25x_core_get_rate(struct clk_hw *hw,
+					      unsigned long parent_rate)
+{
+	return parent_rate;
+}
+PARENTS(clk_pxa25x_core) = { "run", "cpll" };
+MUX_RO_RATE_RO_OPS(clk_pxa25x_core, "core");
+
+static unsigned long clk_pxa25x_run_get_rate(struct clk_hw *hw,
+					     unsigned long parent_rate)
+{
+	unsigned long cccr = CCCR;
+	unsigned int n2 = N2_clk_mult[(cccr >> 7) & 0x07];
+
+	return (parent_rate / n2) * 2;
+}
+PARENTS(clk_pxa25x_run) = { "cpll" };
+RATE_RO_OPS(clk_pxa25x_run, "run");
+
+static unsigned long clk_pxa25x_cpll_get_rate(struct clk_hw *hw,
+	unsigned long parent_rate)
+{
+	unsigned long clkcfg, cccr = CCCR;
+	unsigned int l, m, n2, t;
+
+	asm("mrc\tp14, 0, %0, c6, c0, 0" : "=r" (clkcfg));
+	t = clkcfg & (1 << 0);
+	l  =  L_clk_mult[(cccr >> 0) & 0x1f];
+	m = M_clk_mult[(cccr >> 5) & 0x03];
+	n2 = N2_clk_mult[(cccr >> 7) & 0x07];
+
+	if (t)
+		return m * l * n2 * parent_rate / 2;
+	return m * l * parent_rate;
+}
+PARENTS(clk_pxa25x_cpll) = { "osc_3_6864mhz" };
+RATE_RO_OPS(clk_pxa25x_cpll, "cpll");
+
+static void __init pxa25x_register_core(void)
+{
+	clk_register_clk_pxa25x_cpll();
+	clk_register_clk_pxa25x_run();
+	clkdev_pxa_register(CLK_CORE, "core", NULL,
+			    clk_register_clk_pxa25x_core());
+}
+
+static void __init pxa25x_register_plls(void)
+{
+	clk_register_fixed_rate(NULL, "osc_3_6864mhz", NULL,
+				CLK_GET_RATE_NOCACHE | CLK_IS_ROOT,
+				3686400);
+	clk_register_fixed_rate(NULL, "osc_32_768khz", NULL,
+				CLK_GET_RATE_NOCACHE | CLK_IS_ROOT,
+				32768);
+	clk_register_fixed_rate(NULL, "clk_dummy", NULL, CLK_IS_ROOT, 0);
+	clk_register_fixed_factor(NULL, "ppll_95_85mhz", "osc_3_6864mhz",
+				  0, 26, 1);
+	clk_register_fixed_factor(NULL, "ppll_147_46mhz", "osc_3_6864mhz",
+				  0, 40, 1);
+}
+
+static void __init pxa25x_base_clocks_init(void)
+{
+	pxa25x_register_plls();
+	pxa25x_register_core();
+	clk_register_clk_pxa25x_memory();
+}
+
+#define DUMMY_CLK(_con_id, _dev_id, _parent) \
+	{ .con_id = _con_id, .dev_id = _dev_id, .parent = _parent }
+struct dummy_clk {
+	const char *con_id;
+	const char *dev_id;
+	const char *parent;
+};
+static struct dummy_clk dummy_clks[] __initdata = {
+	DUMMY_CLK(NULL, "pxa25x-gpio", "osc_32_768khz"),
+	DUMMY_CLK(NULL, "pxa26x-gpio", "osc_32_768khz"),
+	DUMMY_CLK("GPIO11_CLK", NULL, "osc_3_6864mhz"),
+	DUMMY_CLK("GPIO12_CLK", NULL, "osc_32_768khz"),
+	DUMMY_CLK(NULL, "sa1100-rtc", "osc_32_768khz"),
+	DUMMY_CLK("OSTIMER0", NULL, "osc_32_768khz"),
+	DUMMY_CLK("UARTCLK", "pxa2xx-ir", "STUART"),
+};
+
+static void __init pxa25x_dummy_clocks_init(void)
+{
+	struct clk *clk;
+	struct dummy_clk *d;
+	const char *name;
+	int i;
+
+	/*
+	 * All pinctrl logic has been wiped out of the clock driver, especially
+	 * for gpio11 and gpio12 outputs. Machine code should ensure proper pin
+	 * control (ie. pxa2xx_mfp_config() invocation).
+	 */
+	for (i = 0; i < ARRAY_SIZE(dummy_clks); i++) {
+		d = &dummy_clks[i];
+		name = d->dev_id ? d->dev_id : d->con_id;
+		clk = clk_register_fixed_factor(NULL, name, d->parent, 0, 1, 1);
+		clk_register_clkdev(clk, d->con_id, d->dev_id);
+	}
+}
+
+int __init pxa25x_clocks_init(void)
+{
+	pxa25x_base_clocks_init();
+	pxa25x_dummy_clocks_init();
+	return clk_pxa_cken_init(pxa25x_clocks, ARRAY_SIZE(pxa25x_clocks));
+}
+
+static void __init pxa25x_dt_clocks_init(struct device_node *np)
+{
+	pxa25x_clocks_init();
+	clk_pxa_dt_common_init(np);
+}
+CLK_OF_DECLARE(pxa25x_clks, "marvell,pxa250-core-clocks",
+	       pxa25x_dt_clocks_init);
diff --git a/drivers/clk/pxa/clk-pxa27x.c b/drivers/clk/pxa/clk-pxa27x.c
new file mode 100644
index 000000000000..5f9b54b024b9
--- /dev/null
+++ b/drivers/clk/pxa/clk-pxa27x.c
@@ -0,0 +1,377 @@
+/*
+ * Marvell PXA27x family clocks
+ *
+ * Copyright (C) 2014 Robert Jarzmik
+ *
+ * Heavily inspired from former arch/arm/mach-pxa/clock.c.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ */
+#include <linux/clk-provider.h>
+#include <mach/pxa2xx-regs.h>
+#include <linux/io.h>
+#include <linux/clk.h>
+#include <linux/clkdev.h>
+#include <linux/of.h>
+
+#include <dt-bindings/clock/pxa-clock.h>
+#include "clk-pxa.h"
+
+#define KHz 1000
+#define MHz (1000 * 1000)
+
+enum {
+	PXA_CORE_13Mhz = 0,
+	PXA_CORE_RUN,
+	PXA_CORE_TURBO,
+};
+
+enum {
+	PXA_BUS_13Mhz = 0,
+	PXA_BUS_RUN,
+};
+
+enum {
+	PXA_LCD_13Mhz = 0,
+	PXA_LCD_RUN,
+};
+
+enum {
+	PXA_MEM_13Mhz = 0,
+	PXA_MEM_SYSTEM_BUS,
+	PXA_MEM_RUN,
+};
+
+static const char * const get_freq_khz[] = {
+	"core", "run", "cpll", "memory",
+	"system_bus"
+};
+
+/*
+ * Get the clock frequency as reflected by CCSR and the turbo flag.
+ * We assume these values have been applied via a fcs.
+ * If info is not 0 we also display the current settings.
+ */
+unsigned int pxa27x_get_clk_frequency_khz(int info)
+{
+	struct clk *clk;
+	unsigned long clks[5];
+	int i;
+
+	for (i = 0; i < 5; i++) {
+		clk = clk_get(NULL, get_freq_khz[i]);
+		if (IS_ERR(clk)) {
+			clks[i] = 0;
+		} else {
+			clks[i] = clk_get_rate(clk);
+			clk_put(clk);
+		}
+	}
+	if (info) {
+		pr_info("Run Mode clock: %ld.%02ldMHz\n",
+			clks[1] / 1000000, (clks[1] % 1000000) / 10000);
+		pr_info("Turbo Mode clock: %ld.%02ldMHz\n",
+			clks[2] / 1000000, (clks[2] % 1000000) / 10000);
+		pr_info("Memory clock: %ld.%02ldMHz\n",
+			clks[3] / 1000000, (clks[3] % 1000000) / 10000);
+		pr_info("System bus clock: %ld.%02ldMHz\n",
+			clks[4] / 1000000, (clks[4] % 1000000) / 10000);
+	}
+	return (unsigned int)clks[0];
+}
+
+bool pxa27x_is_ppll_disabled(void)
+{
+	unsigned long ccsr = CCSR;
+
+	return ccsr & (1 << CCCR_PPDIS_BIT);
+}
+
+#define PXA27X_CKEN(dev_id, con_id, parents, mult_hp, div_hp,		\
+		    bit, is_lp, flags)					\
+	PXA_CKEN(dev_id, con_id, bit, parents, 1, 1, mult_hp, div_hp,	\
+		 is_lp,  &CKEN, CKEN_ ## bit, flags)
+#define PXA27X_PBUS_CKEN(dev_id, con_id, bit, mult_hp, div_hp, delay)	\
+	PXA27X_CKEN(dev_id, con_id, pxa27x_pbus_parents, mult_hp,	\
+		    div_hp, bit, pxa27x_is_ppll_disabled, 0)
+
+PARENTS(pxa27x_pbus) = { "osc_13mhz", "ppll_312mhz" };
+PARENTS(pxa27x_sbus) = { "system_bus", "system_bus" };
+PARENTS(pxa27x_32Mhz_bus) = { "osc_32_768khz", "osc_32_768khz" };
+PARENTS(pxa27x_lcd_bus) = { "lcd_base", "lcd_base" };
+PARENTS(pxa27x_membus) = { "lcd_base", "lcd_base" };
+
+#define PXA27X_CKEN_1RATE(dev_id, con_id, bit, parents, delay)		\
+	PXA_CKEN_1RATE(dev_id, con_id, bit, parents,			\
+		       &CKEN, CKEN_ ## bit, 0)
+#define PXA27X_CKEN_1RATE_AO(dev_id, con_id, bit, parents, delay)	\
+	PXA_CKEN_1RATE(dev_id, con_id, bit, parents,			\
+		       &CKEN, CKEN_ ## bit, CLK_IGNORE_UNUSED)
+
+static struct desc_clk_cken pxa27x_clocks[] __initdata = {
+	PXA27X_PBUS_CKEN("pxa2xx-uart.0", NULL, FFUART, 2, 42, 1),
+	PXA27X_PBUS_CKEN("pxa2xx-uart.1", NULL, BTUART, 2, 42, 1),
+	PXA27X_PBUS_CKEN("pxa2xx-uart.2", NULL, STUART, 2, 42, 1),
+	PXA27X_PBUS_CKEN("pxa2xx-i2s", NULL, I2S, 2, 51, 0),
+	PXA27X_PBUS_CKEN("pxa2xx-i2c.0", NULL, I2C, 2, 19, 0),
+	PXA27X_PBUS_CKEN("pxa27x-udc", NULL, USB, 2, 13, 5),
+	PXA27X_PBUS_CKEN("pxa2xx-mci.0", NULL, MMC, 2, 32, 0),
+	PXA27X_PBUS_CKEN("pxa2xx-ir", "FICPCLK", FICP, 2, 13, 0),
+	PXA27X_PBUS_CKEN("pxa27x-ohci", NULL, USBHOST, 2, 13, 0),
+	PXA27X_PBUS_CKEN("pxa2xx-i2c.1", NULL, PWRI2C, 1, 24, 0),
+	PXA27X_PBUS_CKEN("pxa27x-ssp.0", NULL, SSP1, 1, 24, 0),
+	PXA27X_PBUS_CKEN("pxa27x-ssp.1", NULL, SSP2, 1, 24, 0),
+	PXA27X_PBUS_CKEN("pxa27x-ssp.2", NULL, SSP3, 1, 24, 0),
+	PXA27X_PBUS_CKEN("pxa27x-pwm.0", NULL, PWM0, 1, 24, 0),
+	PXA27X_PBUS_CKEN("pxa27x-pwm.1", NULL, PWM1, 1, 24, 0),
+	PXA27X_PBUS_CKEN(NULL, "MSLCLK", MSL, 2, 13, 0),
+	PXA27X_PBUS_CKEN(NULL, "USIMCLK", USIM, 2, 13, 0),
+	PXA27X_PBUS_CKEN(NULL, "MSTKCLK", MEMSTK, 2, 32, 0),
+	PXA27X_PBUS_CKEN(NULL, "AC97CLK", AC97, 1, 1, 0),
+	PXA27X_PBUS_CKEN(NULL, "AC97CONFCLK", AC97CONF, 1, 1, 0),
+	PXA27X_PBUS_CKEN(NULL, "OSTIMER0", OSTIMER, 1, 96, 0),
+
+	PXA27X_CKEN_1RATE("pxa27x-keypad", NULL, KEYPAD,
+			  pxa27x_32Mhz_bus_parents, 0),
+	PXA27X_CKEN_1RATE(NULL, "IMCLK", IM, pxa27x_sbus_parents, 0),
+	PXA27X_CKEN_1RATE("pxa2xx-fb", NULL, LCD, pxa27x_lcd_bus_parents, 0),
+	PXA27X_CKEN_1RATE("pxa27x-camera.0", NULL, CAMERA,
+			  pxa27x_lcd_bus_parents, 0),
+	PXA27X_CKEN_1RATE_AO("pxa2xx-pcmcia", NULL, MEMC,
+			     pxa27x_membus_parents, 0),
+
+};
+
+static unsigned long clk_pxa27x_cpll_get_rate(struct clk_hw *hw,
+	unsigned long parent_rate)
+{
+	unsigned long clkcfg;
+	unsigned int t, ht;
+	unsigned int l, L, n2, N;
+	unsigned long ccsr = CCSR;
+
+	asm("mrc\tp14, 0, %0, c6, c0, 0" : "=r" (clkcfg));
+	t  = clkcfg & (1 << 0);
+	ht = clkcfg & (1 << 2);
+
+	l  = ccsr & CCSR_L_MASK;
+	n2 = (ccsr & CCSR_N2_MASK) >> CCSR_N2_SHIFT;
+	L  = l * parent_rate;
+	N  = (L * n2) / 2;
+
+	return t ? N : L;
+}
+PARENTS(clk_pxa27x_cpll) = { "osc_13mhz" };
+RATE_RO_OPS(clk_pxa27x_cpll, "cpll");
+
+static unsigned long clk_pxa27x_lcd_base_get_rate(struct clk_hw *hw,
+						  unsigned long parent_rate)
+{
+	unsigned int l, osc_forced;
+	unsigned long ccsr = CCSR;
+	unsigned long cccr = CCCR;
+
+	l  = ccsr & CCSR_L_MASK;
+	osc_forced = ccsr & (1 << CCCR_CPDIS_BIT);
+	if (osc_forced) {
+		if (cccr & (1 << CCCR_LCD_26_BIT))
+			return parent_rate * 2;
+		else
+			return parent_rate;
+	}
+
+	if (l <= 7)
+		return parent_rate;
+	if (l <= 16)
+		return parent_rate / 2;
+	return parent_rate / 4;
+}
+
+static u8 clk_pxa27x_lcd_base_get_parent(struct clk_hw *hw)
+{
+	unsigned int osc_forced;
+	unsigned long ccsr = CCSR;
+
+	osc_forced = ccsr & (1 << CCCR_CPDIS_BIT);
+	if (osc_forced)
+		return PXA_LCD_13Mhz;
+	else
+		return PXA_LCD_RUN;
+}
+
+PARENTS(clk_pxa27x_lcd_base) = { "osc_13mhz", "run" };
+MUX_RO_RATE_RO_OPS(clk_pxa27x_lcd_base, "lcd_base");
+
+static void __init pxa27x_register_plls(void)
+{
+	clk_register_fixed_rate(NULL, "osc_13mhz", NULL,
+				CLK_GET_RATE_NOCACHE | CLK_IS_ROOT,
+				13 * MHz);
+	clk_register_fixed_rate(NULL, "osc_32_768khz", NULL,
+				CLK_GET_RATE_NOCACHE | CLK_IS_ROOT,
+				32768 * KHz);
+	clk_register_fixed_rate(NULL, "clk_dummy", NULL, CLK_IS_ROOT, 0);
+	clk_register_fixed_factor(NULL, "ppll_312mhz", "osc_13mhz", 0, 24, 1);
+}
+
+static unsigned long clk_pxa27x_core_get_rate(struct clk_hw *hw,
+					      unsigned long parent_rate)
+{
+	unsigned long clkcfg;
+	unsigned int t, ht, b, osc_forced;
+	unsigned long ccsr = CCSR;
+
+	osc_forced = ccsr & (1 << CCCR_CPDIS_BIT);
+	asm("mrc\tp14, 0, %0, c6, c0, 0" : "=r" (clkcfg));
+	t  = clkcfg & (1 << 0);
+	ht = clkcfg & (1 << 2);
+	b  = clkcfg & (1 << 3);
+
+	if (osc_forced)
+		return parent_rate;
+	if (ht)
+		return parent_rate / 2;
+	else
+		return parent_rate;
+}
+
+static u8 clk_pxa27x_core_get_parent(struct clk_hw *hw)
+{
+	unsigned long clkcfg;
+	unsigned int t, ht, b, osc_forced;
+	unsigned long ccsr = CCSR;
+
+	osc_forced = ccsr & (1 << CCCR_CPDIS_BIT);
+	if (osc_forced)
+		return PXA_CORE_13Mhz;
+
+	asm("mrc\tp14, 0, %0, c6, c0, 0" : "=r" (clkcfg));
+	t  = clkcfg & (1 << 0);
+	ht = clkcfg & (1 << 2);
+	b  = clkcfg & (1 << 3);
+
+	if (ht || t)
+		return PXA_CORE_TURBO;
+	return PXA_CORE_RUN;
+}
+PARENTS(clk_pxa27x_core) = { "osc_13mhz", "run", "cpll" };
+MUX_RO_RATE_RO_OPS(clk_pxa27x_core, "core");
+
+static unsigned long clk_pxa27x_run_get_rate(struct clk_hw *hw,
+					     unsigned long parent_rate)
+{
+	unsigned long ccsr = CCSR;
+	unsigned int n2 = (ccsr & CCSR_N2_MASK) >> CCSR_N2_SHIFT;
+
+	return (parent_rate / n2) * 2;
+}
+PARENTS(clk_pxa27x_run) = { "cpll" };
+RATE_RO_OPS(clk_pxa27x_run, "run");
+
+static void __init pxa27x_register_core(void)
+{
+	clk_register_clk_pxa27x_cpll();
+	clk_register_clk_pxa27x_run();
+
+	clkdev_pxa_register(CLK_CORE, "core", NULL,
+			    clk_register_clk_pxa27x_core());
+}
+
+static unsigned long clk_pxa27x_system_bus_get_rate(struct clk_hw *hw,
+						    unsigned long parent_rate)
+{
+	unsigned long clkcfg;
+	unsigned int b, osc_forced;
+	unsigned long ccsr = CCSR;
+
+	osc_forced = ccsr & (1 << CCCR_CPDIS_BIT);
+	asm("mrc\tp14, 0, %0, c6, c0, 0" : "=r" (clkcfg));
+	b  = clkcfg & (1 << 3);
+
+	if (osc_forced)
+		return parent_rate;
+	if (b)
+		return parent_rate / 2;
+	else
+		return parent_rate;
+}
+
+static u8 clk_pxa27x_system_bus_get_parent(struct clk_hw *hw)
+{
+	unsigned int osc_forced;
+	unsigned long ccsr = CCSR;
+
+	osc_forced = ccsr & (1 << CCCR_CPDIS_BIT);
+	if (osc_forced)
+		return PXA_BUS_13Mhz;
+	else
+		return PXA_BUS_RUN;
+}
+
+PARENTS(clk_pxa27x_system_bus) = { "osc_13mhz", "run" };
+MUX_RO_RATE_RO_OPS(clk_pxa27x_system_bus, "system_bus");
+
+static unsigned long clk_pxa27x_memory_get_rate(struct clk_hw *hw,
+						unsigned long parent_rate)
+{
+	unsigned int a, l, osc_forced;
+	unsigned long cccr = CCCR;
+	unsigned long ccsr = CCSR;
+
+	osc_forced = ccsr & (1 << CCCR_CPDIS_BIT);
+	a = cccr & (1 << CCCR_A_BIT);
+	l  = ccsr & CCSR_L_MASK;
+
+	if (osc_forced || a)
+		return parent_rate;
+	if (l <= 10)
+		return parent_rate;
+	if (l <= 20)
+		return parent_rate / 2;
+	return parent_rate / 4;
+}
+
+static u8 clk_pxa27x_memory_get_parent(struct clk_hw *hw)
+{
+	unsigned int osc_forced, a;
+	unsigned long cccr = CCCR;
+	unsigned long ccsr = CCSR;
+
+	osc_forced = ccsr & (1 << CCCR_CPDIS_BIT);
+	a = cccr & (1 << CCCR_A_BIT);
+	if (osc_forced)
+		return PXA_MEM_13Mhz;
+	if (a)
+		return PXA_MEM_SYSTEM_BUS;
+	else
+		return PXA_MEM_RUN;
+}
+
+PARENTS(clk_pxa27x_memory) = { "osc_13mhz", "system_bus", "run" };
+MUX_RO_RATE_RO_OPS(clk_pxa27x_memory, "memory");
+
+static void __init pxa27x_base_clocks_init(void)
+{
+	pxa27x_register_plls();
+	pxa27x_register_core();
+	clk_register_clk_pxa27x_system_bus();
+	clk_register_clk_pxa27x_memory();
+	clk_register_clk_pxa27x_lcd_base();
+}
+
+static int __init pxa27x_clocks_init(void)
+{
+	pxa27x_base_clocks_init();
+	return clk_pxa_cken_init(pxa27x_clocks, ARRAY_SIZE(pxa27x_clocks));
+}
+postcore_initcall(pxa27x_clocks_init);
+
+static void __init pxa27x_dt_clocks_init(struct device_node *np)
+{
+	pxa27x_clocks_init();
+	clk_pxa_dt_common_init(np);
+}
+CLK_OF_DECLARE(pxa_clks, "marvell,pxa270-clocks", pxa27x_dt_clocks_init);
diff --git a/drivers/clk/qcom/clk-pll.c b/drivers/clk/qcom/clk-pll.c
index 9db03d3b1657..60873a7f45d9 100644
--- a/drivers/clk/qcom/clk-pll.c
+++ b/drivers/clk/qcom/clk-pll.c
@@ -97,7 +97,7 @@ static unsigned long
 clk_pll_recalc_rate(struct clk_hw *hw, unsigned long parent_rate)
 {
 	struct clk_pll *pll = to_clk_pll(hw);
-	u32 l, m, n;
+	u32 l, m, n, config;
 	unsigned long rate;
 	u64 tmp;
 
@@ -116,13 +116,79 @@ clk_pll_recalc_rate(struct clk_hw *hw, unsigned long parent_rate)
 		do_div(tmp, n);
 		rate += tmp;
 	}
+	if (pll->post_div_width) {
+		regmap_read(pll->clkr.regmap, pll->config_reg, &config);
+		config >>= pll->post_div_shift;
+		config &= BIT(pll->post_div_width) - 1;
+		rate /= config + 1;
+	}
+
 	return rate;
 }
 
+static const
+struct pll_freq_tbl *find_freq(const struct pll_freq_tbl *f, unsigned long rate)
+{
+	if (!f)
+		return NULL;
+
+	for (; f->freq; f++)
+		if (rate <= f->freq)
+			return f;
+
+	return NULL;
+}
+
+static long
+clk_pll_determine_rate(struct clk_hw *hw, unsigned long rate,
+		       unsigned long *p_rate, struct clk_hw **p)
+{
+	struct clk_pll *pll = to_clk_pll(hw);
+	const struct pll_freq_tbl *f;
+
+	f = find_freq(pll->freq_tbl, rate);
+	if (!f)
+		return clk_pll_recalc_rate(hw, *p_rate);
+
+	return f->freq;
+}
+
+static int
+clk_pll_set_rate(struct clk_hw *hw, unsigned long rate, unsigned long p_rate)
+{
+	struct clk_pll *pll = to_clk_pll(hw);
+	const struct pll_freq_tbl *f;
+	bool enabled;
+	u32 mode;
+	u32 enable_mask = PLL_OUTCTRL | PLL_BYPASSNL | PLL_RESET_N;
+
+	f = find_freq(pll->freq_tbl, rate);
+	if (!f)
+		return -EINVAL;
+
+	regmap_read(pll->clkr.regmap, pll->mode_reg, &mode);
+	enabled = (mode & enable_mask) == enable_mask;
+
+	if (enabled)
+		clk_pll_disable(hw);
+
+	regmap_update_bits(pll->clkr.regmap, pll->l_reg, 0x3ff, f->l);
+	regmap_update_bits(pll->clkr.regmap, pll->m_reg, 0x7ffff, f->m);
+	regmap_update_bits(pll->clkr.regmap, pll->n_reg, 0x7ffff, f->n);
+	regmap_write(pll->clkr.regmap, pll->config_reg, f->ibits);
+
+	if (enabled)
+		clk_pll_enable(hw);
+
+	return 0;
+}
+
 const struct clk_ops clk_pll_ops = {
 	.enable = clk_pll_enable,
 	.disable = clk_pll_disable,
 	.recalc_rate = clk_pll_recalc_rate,
+	.determine_rate = clk_pll_determine_rate,
+	.set_rate = clk_pll_set_rate,
 };
 EXPORT_SYMBOL_GPL(clk_pll_ops);
 
diff --git a/drivers/clk/qcom/clk-pll.h b/drivers/clk/qcom/clk-pll.h
index 3003e9962472..c9c0cda306d0 100644
--- a/drivers/clk/qcom/clk-pll.h
+++ b/drivers/clk/qcom/clk-pll.h
@@ -18,6 +18,21 @@
 #include "clk-regmap.h"
 
 /**
+ * struct pll_freq_tbl - PLL frequency table
+ * @l: L value
+ * @m: M value
+ * @n: N value
+ * @ibits: internal values
+ */
+struct pll_freq_tbl {
+	unsigned long freq;
+	u16 l;
+	u16 m;
+	u16 n;
+	u32 ibits;
+};
+
+/**
  * struct clk_pll - phase locked loop (PLL)
  * @l_reg: L register
  * @m_reg: M register
@@ -26,6 +41,7 @@
  * @mode_reg: mode register
  * @status_reg: status register
  * @status_bit: ANDed with @status_reg to determine if PLL is enabled
+ * @freq_tbl: PLL frequency table
  * @hw: handle between common and hardware-specific interfaces
  */
 struct clk_pll {
@@ -36,6 +52,10 @@ struct clk_pll {
 	u32	mode_reg;
 	u32	status_reg;
 	u8	status_bit;
+	u8	post_div_width;
+	u8	post_div_shift;
+
+	const struct pll_freq_tbl *freq_tbl;
 
 	struct clk_regmap clkr;
 };
diff --git a/drivers/clk/qcom/clk-rcg.c b/drivers/clk/qcom/clk-rcg.c
index b638c5846dbf..0b93972c8807 100644
--- a/drivers/clk/qcom/clk-rcg.c
+++ b/drivers/clk/qcom/clk-rcg.c
@@ -21,6 +21,7 @@
 #include <asm/div64.h>
 
 #include "clk-rcg.h"
+#include "common.h"
 
 static u32 ns_to_src(struct src_sel *s, u32 ns)
 {
@@ -67,16 +68,16 @@ static u8 clk_dyn_rcg_get_parent(struct clk_hw *hw)
 {
 	struct clk_dyn_rcg *rcg = to_clk_dyn_rcg(hw);
 	int num_parents = __clk_get_num_parents(hw->clk);
-	u32 ns, ctl;
+	u32 ns, reg;
 	int bank;
 	int i;
 	struct src_sel *s;
 
-	regmap_read(rcg->clkr.regmap, rcg->clkr.enable_reg, &ctl);
-	bank = reg_to_bank(rcg, ctl);
+	regmap_read(rcg->clkr.regmap, rcg->bank_reg, &reg);
+	bank = reg_to_bank(rcg, reg);
 	s = &rcg->s[bank];
 
-	regmap_read(rcg->clkr.regmap, rcg->ns_reg, &ns);
+	regmap_read(rcg->clkr.regmap, rcg->ns_reg[bank], &ns);
 	ns = ns_to_src(s, ns);
 
 	for (i = 0; i < num_parents; i++)
@@ -192,90 +193,93 @@ static u32 mn_to_reg(struct mn *mn, u32 m, u32 n, u32 val)
 
 static void configure_bank(struct clk_dyn_rcg *rcg, const struct freq_tbl *f)
 {
-	u32 ns, md, ctl, *regp;
+	u32 ns, md, reg;
 	int bank, new_bank;
 	struct mn *mn;
 	struct pre_div *p;
 	struct src_sel *s;
 	bool enabled;
-	u32 md_reg;
-	u32 bank_reg;
+	u32 md_reg, ns_reg;
 	bool banked_mn = !!rcg->mn[1].width;
+	bool banked_p = !!rcg->p[1].pre_div_width;
 	struct clk_hw *hw = &rcg->clkr.hw;
 
 	enabled = __clk_is_enabled(hw->clk);
 
-	regmap_read(rcg->clkr.regmap, rcg->ns_reg, &ns);
-	regmap_read(rcg->clkr.regmap, rcg->clkr.enable_reg, &ctl);
-
-	if (banked_mn) {
-		regp = &ctl;
-		bank_reg = rcg->clkr.enable_reg;
-	} else {
-		regp = &ns;
-		bank_reg = rcg->ns_reg;
-	}
-
-	bank = reg_to_bank(rcg, *regp);
+	regmap_read(rcg->clkr.regmap, rcg->bank_reg, &reg);
+	bank = reg_to_bank(rcg, reg);
 	new_bank = enabled ? !bank : bank;
 
+	ns_reg = rcg->ns_reg[new_bank];
+	regmap_read(rcg->clkr.regmap, ns_reg, &ns);
+
 	if (banked_mn) {
 		mn = &rcg->mn[new_bank];
 		md_reg = rcg->md_reg[new_bank];
 
 		ns |= BIT(mn->mnctr_reset_bit);
-		regmap_write(rcg->clkr.regmap, rcg->ns_reg, ns);
+		regmap_write(rcg->clkr.regmap, ns_reg, ns);
 
 		regmap_read(rcg->clkr.regmap, md_reg, &md);
 		md = mn_to_md(mn, f->m, f->n, md);
 		regmap_write(rcg->clkr.regmap, md_reg, md);
 
 		ns = mn_to_ns(mn, f->m, f->n, ns);
-		regmap_write(rcg->clkr.regmap, rcg->ns_reg, ns);
+		regmap_write(rcg->clkr.regmap, ns_reg, ns);
 
-		ctl = mn_to_reg(mn, f->m, f->n, ctl);
-		regmap_write(rcg->clkr.regmap, rcg->clkr.enable_reg, ctl);
+		/* Two NS registers means mode control is in NS register */
+		if (rcg->ns_reg[0] != rcg->ns_reg[1]) {
+			ns = mn_to_reg(mn, f->m, f->n, ns);
+			regmap_write(rcg->clkr.regmap, ns_reg, ns);
+		} else {
+			reg = mn_to_reg(mn, f->m, f->n, reg);
+			regmap_write(rcg->clkr.regmap, rcg->bank_reg, reg);
+		}
 
 		ns &= ~BIT(mn->mnctr_reset_bit);
-		regmap_write(rcg->clkr.regmap, rcg->ns_reg, ns);
-	} else {
+		regmap_write(rcg->clkr.regmap, ns_reg, ns);
+	}
+
+	if (banked_p) {
 		p = &rcg->p[new_bank];
 		ns = pre_div_to_ns(p, f->pre_div - 1, ns);
 	}
 
 	s = &rcg->s[new_bank];
 	ns = src_to_ns(s, s->parent_map[f->src], ns);
-	regmap_write(rcg->clkr.regmap, rcg->ns_reg, ns);
+	regmap_write(rcg->clkr.regmap, ns_reg, ns);
 
 	if (enabled) {
-		*regp ^= BIT(rcg->mux_sel_bit);
-		regmap_write(rcg->clkr.regmap, bank_reg, *regp);
+		regmap_read(rcg->clkr.regmap, rcg->bank_reg, &reg);
+		reg ^= BIT(rcg->mux_sel_bit);
+		regmap_write(rcg->clkr.regmap, rcg->bank_reg, reg);
 	}
 }
 
 static int clk_dyn_rcg_set_parent(struct clk_hw *hw, u8 index)
 {
 	struct clk_dyn_rcg *rcg = to_clk_dyn_rcg(hw);
-	u32 ns, ctl, md, reg;
+	u32 ns, md, reg;
 	int bank;
 	struct freq_tbl f = { 0 };
 	bool banked_mn = !!rcg->mn[1].width;
+	bool banked_p = !!rcg->p[1].pre_div_width;
 
-	regmap_read(rcg->clkr.regmap, rcg->ns_reg, &ns);
-	regmap_read(rcg->clkr.regmap, rcg->clkr.enable_reg, &ctl);
-	reg = banked_mn ? ctl : ns;
-
+	regmap_read(rcg->clkr.regmap, rcg->bank_reg, &reg);
 	bank = reg_to_bank(rcg, reg);
 
+	regmap_read(rcg->clkr.regmap, rcg->ns_reg[bank], &ns);
+
 	if (banked_mn) {
 		regmap_read(rcg->clkr.regmap, rcg->md_reg[bank], &md);
 		f.m = md_to_m(&rcg->mn[bank], md);
 		f.n = ns_m_to_n(&rcg->mn[bank], ns, f.m);
-	} else {
-		f.pre_div = ns_to_pre_div(&rcg->p[bank], ns) + 1;
 	}
-	f.src = index;
 
+	if (banked_p)
+		f.pre_div = ns_to_pre_div(&rcg->p[bank], ns) + 1;
+
+	f.src = index;
 	configure_bank(rcg, &f);
 
 	return 0;
@@ -336,55 +340,45 @@ clk_dyn_rcg_recalc_rate(struct clk_hw *hw, unsigned long parent_rate)
 	u32 m, n, pre_div, ns, md, mode, reg;
 	int bank;
 	struct mn *mn;
+	bool banked_p = !!rcg->p[1].pre_div_width;
 	bool banked_mn = !!rcg->mn[1].width;
 
-	regmap_read(rcg->clkr.regmap, rcg->ns_reg, &ns);
-
-	if (banked_mn)
-		regmap_read(rcg->clkr.regmap, rcg->clkr.enable_reg, &reg);
-	else
-		reg = ns;
-
+	regmap_read(rcg->clkr.regmap, rcg->bank_reg, &reg);
 	bank = reg_to_bank(rcg, reg);
 
+	regmap_read(rcg->clkr.regmap, rcg->ns_reg[bank], &ns);
+	m = n = pre_div = mode = 0;
+
 	if (banked_mn) {
 		mn = &rcg->mn[bank];
 		regmap_read(rcg->clkr.regmap, rcg->md_reg[bank], &md);
 		m = md_to_m(mn, md);
 		n = ns_m_to_n(mn, ns, m);
+		/* Two NS registers means mode control is in NS register */
+		if (rcg->ns_reg[0] != rcg->ns_reg[1])
+			reg = ns;
 		mode = reg_to_mnctr_mode(mn, reg);
-		return calc_rate(parent_rate, m, n, mode, 0);
-	} else {
-		pre_div = ns_to_pre_div(&rcg->p[bank], ns);
-		return calc_rate(parent_rate, 0, 0, 0, pre_div);
 	}
-}
 
-static const
-struct freq_tbl *find_freq(const struct freq_tbl *f, unsigned long rate)
-{
-	if (!f)
-		return NULL;
-
-	for (; f->freq; f++)
-		if (rate <= f->freq)
-			return f;
+	if (banked_p)
+		pre_div = ns_to_pre_div(&rcg->p[bank], ns);
 
-	return NULL;
+	return calc_rate(parent_rate, m, n, mode, pre_div);
 }
 
 static long _freq_tbl_determine_rate(struct clk_hw *hw,
 		const struct freq_tbl *f, unsigned long rate,
-		unsigned long *p_rate, struct clk **p)
+		unsigned long *p_rate, struct clk_hw **p_hw)
 {
 	unsigned long clk_flags;
+	struct clk *p;
 
-	f = find_freq(f, rate);
+	f = qcom_find_freq(f, rate);
 	if (!f)
 		return -EINVAL;
 
 	clk_flags = __clk_get_flags(hw->clk);
-	*p = clk_get_parent_by_index(hw->clk, f->src);
+	p = clk_get_parent_by_index(hw->clk, f->src);
 	if (clk_flags & CLK_SET_RATE_PARENT) {
 		rate = rate * f->pre_div;
 		if (f->n) {
@@ -394,15 +388,16 @@ static long _freq_tbl_determine_rate(struct clk_hw *hw,
 			rate = tmp;
 		}
 	} else {
-		rate =  __clk_get_rate(*p);
+		rate =  __clk_get_rate(p);
 	}
+	*p_hw = __clk_get_hw(p);
 	*p_rate = rate;
 
 	return f->freq;
 }
 
 static long clk_rcg_determine_rate(struct clk_hw *hw, unsigned long rate,
-		unsigned long *p_rate, struct clk **p)
+		unsigned long *p_rate, struct clk_hw **p)
 {
 	struct clk_rcg *rcg = to_clk_rcg(hw);
 
@@ -410,7 +405,7 @@ static long clk_rcg_determine_rate(struct clk_hw *hw, unsigned long rate,
 }
 
 static long clk_dyn_rcg_determine_rate(struct clk_hw *hw, unsigned long rate,
-		unsigned long *p_rate, struct clk **p)
+		unsigned long *p_rate, struct clk_hw **p)
 {
 	struct clk_dyn_rcg *rcg = to_clk_dyn_rcg(hw);
 
@@ -418,13 +413,15 @@ static long clk_dyn_rcg_determine_rate(struct clk_hw *hw, unsigned long rate,
 }
 
 static long clk_rcg_bypass_determine_rate(struct clk_hw *hw, unsigned long rate,
-		unsigned long *p_rate, struct clk **p)
+		unsigned long *p_rate, struct clk_hw **p_hw)
 {
 	struct clk_rcg *rcg = to_clk_rcg(hw);
 	const struct freq_tbl *f = rcg->freq_tbl;
+	struct clk *p;
 
-	*p = clk_get_parent_by_index(hw->clk, f->src);
-	*p_rate = __clk_round_rate(*p, rate);
+	p = clk_get_parent_by_index(hw->clk, f->src);
+	*p_hw = __clk_get_hw(p);
+	*p_rate = __clk_round_rate(p, rate);
 
 	return *p_rate;
 }
@@ -477,7 +474,7 @@ static int clk_rcg_set_rate(struct clk_hw *hw, unsigned long rate,
 	struct clk_rcg *rcg = to_clk_rcg(hw);
 	const struct freq_tbl *f;
 
-	f = find_freq(rcg->freq_tbl, rate);
+	f = qcom_find_freq(rcg->freq_tbl, rate);
 	if (!f)
 		return -EINVAL;
 
@@ -497,7 +494,7 @@ static int __clk_dyn_rcg_set_rate(struct clk_hw *hw, unsigned long rate)
 	struct clk_dyn_rcg *rcg = to_clk_dyn_rcg(hw);
 	const struct freq_tbl *f;
 
-	f = find_freq(rcg->freq_tbl, rate);
+	f = qcom_find_freq(rcg->freq_tbl, rate);
 	if (!f)
 		return -EINVAL;
 
diff --git a/drivers/clk/qcom/clk-rcg.h b/drivers/clk/qcom/clk-rcg.h
index ba0523cefd2e..687e41f91d7c 100644
--- a/drivers/clk/qcom/clk-rcg.h
+++ b/drivers/clk/qcom/clk-rcg.h
@@ -103,8 +103,9 @@ extern const struct clk_ops clk_rcg_bypass_ops;
  * struct clk_dyn_rcg - root clock generator with glitch free mux
  *
  * @mux_sel_bit: bit to switch glitch free mux
- * @ns_reg: NS register
+ * @ns_reg: NS0 and NS1 register
  * @md_reg: MD0 and MD1 register
+ * @bank_reg: register to XOR @mux_sel_bit into to switch glitch free mux
  * @mn: mn counter (banked)
  * @s: source selector (banked)
  * @freq_tbl: frequency table
@@ -113,8 +114,9 @@ extern const struct clk_ops clk_rcg_bypass_ops;
  *
  */
 struct clk_dyn_rcg {
-	u32	ns_reg;
+	u32	ns_reg[2];
 	u32	md_reg[2];
+	u32	bank_reg;
 
 	u8	mux_sel_bit;
 
diff --git a/drivers/clk/qcom/clk-rcg2.c b/drivers/clk/qcom/clk-rcg2.c
index cd185d5cc67a..08b8b3729f53 100644
--- a/drivers/clk/qcom/clk-rcg2.c
+++ b/drivers/clk/qcom/clk-rcg2.c
@@ -24,6 +24,7 @@
 #include <asm/div64.h>
 
 #include "clk-rcg.h"
+#include "common.h"
 
 #define CMD_REG			0x0
 #define CMD_UPDATE		BIT(0)
@@ -172,32 +173,19 @@ clk_rcg2_recalc_rate(struct clk_hw *hw, unsigned long parent_rate)
 	return calc_rate(parent_rate, m, n, mode, hid_div);
 }
 
-static const
-struct freq_tbl *find_freq(const struct freq_tbl *f, unsigned long rate)
-{
-	if (!f)
-		return NULL;
-
-	for (; f->freq; f++)
-		if (rate <= f->freq)
-			return f;
-
-	/* Default to our fastest rate */
-	return f - 1;
-}
-
 static long _freq_tbl_determine_rate(struct clk_hw *hw,
 		const struct freq_tbl *f, unsigned long rate,
-		unsigned long *p_rate, struct clk **p)
+		unsigned long *p_rate, struct clk_hw **p_hw)
 {
 	unsigned long clk_flags;
+	struct clk *p;
 
-	f = find_freq(f, rate);
+	f = qcom_find_freq(f, rate);
 	if (!f)
 		return -EINVAL;
 
 	clk_flags = __clk_get_flags(hw->clk);
-	*p = clk_get_parent_by_index(hw->clk, f->src);
+	p = clk_get_parent_by_index(hw->clk, f->src);
 	if (clk_flags & CLK_SET_RATE_PARENT) {
 		if (f->pre_div) {
 			rate /= 2;
@@ -211,15 +199,16 @@ static long _freq_tbl_determine_rate(struct clk_hw *hw,
 			rate = tmp;
 		}
 	} else {
-		rate =  __clk_get_rate(*p);
+		rate =  __clk_get_rate(p);
 	}
+	*p_hw = __clk_get_hw(p);
 	*p_rate = rate;
 
 	return f->freq;
 }
 
 static long clk_rcg2_determine_rate(struct clk_hw *hw, unsigned long rate,
-		unsigned long *p_rate, struct clk **p)
+		unsigned long *p_rate, struct clk_hw **p)
 {
 	struct clk_rcg2 *rcg = to_clk_rcg2(hw);
 
@@ -268,7 +257,7 @@ static int __clk_rcg2_set_rate(struct clk_hw *hw, unsigned long rate)
 	struct clk_rcg2 *rcg = to_clk_rcg2(hw);
 	const struct freq_tbl *f;
 
-	f = find_freq(rcg->freq_tbl, rate);
+	f = qcom_find_freq(rcg->freq_tbl, rate);
 	if (!f)
 		return -EINVAL;
 
@@ -372,7 +361,7 @@ static int clk_edp_pixel_set_rate_and_parent(struct clk_hw *hw,
 }
 
 static long clk_edp_pixel_determine_rate(struct clk_hw *hw, unsigned long rate,
-				 unsigned long *p_rate, struct clk **p)
+				 unsigned long *p_rate, struct clk_hw **p)
 {
 	struct clk_rcg2 *rcg = to_clk_rcg2(hw);
 	const struct freq_tbl *f = rcg->freq_tbl;
@@ -384,7 +373,7 @@ static long clk_edp_pixel_determine_rate(struct clk_hw *hw, unsigned long rate,
 	u32 hid_div;
 
 	/* Force the correct parent */
-	*p = clk_get_parent_by_index(hw->clk, f->src);
+	*p = __clk_get_hw(clk_get_parent_by_index(hw->clk, f->src));
 
 	if (src_rate == 810000000)
 		frac = frac_table_810m;
@@ -423,18 +412,20 @@ const struct clk_ops clk_edp_pixel_ops = {
 EXPORT_SYMBOL_GPL(clk_edp_pixel_ops);
 
 static long clk_byte_determine_rate(struct clk_hw *hw, unsigned long rate,
-			 unsigned long *p_rate, struct clk **p)
+			 unsigned long *p_rate, struct clk_hw **p_hw)
 {
 	struct clk_rcg2 *rcg = to_clk_rcg2(hw);
 	const struct freq_tbl *f = rcg->freq_tbl;
 	unsigned long parent_rate, div;
 	u32 mask = BIT(rcg->hid_width) - 1;
+	struct clk *p;
 
 	if (rate == 0)
 		return -EINVAL;
 
-	*p = clk_get_parent_by_index(hw->clk, f->src);
-	*p_rate = parent_rate = __clk_round_rate(*p, rate);
+	p = clk_get_parent_by_index(hw->clk, f->src);
+	*p_hw = __clk_get_hw(p);
+	*p_rate = parent_rate = __clk_round_rate(p, rate);
 
 	div = DIV_ROUND_UP((2 * parent_rate), rate) - 1;
 	div = min_t(u32, div, mask);
@@ -485,14 +476,16 @@ static const struct frac_entry frac_table_pixel[] = {
 };
 
 static long clk_pixel_determine_rate(struct clk_hw *hw, unsigned long rate,
-				 unsigned long *p_rate, struct clk **p)
+				 unsigned long *p_rate, struct clk_hw **p)
 {
 	struct clk_rcg2 *rcg = to_clk_rcg2(hw);
 	unsigned long request, src_rate;
 	int delta = 100000;
 	const struct freq_tbl *f = rcg->freq_tbl;
 	const struct frac_entry *frac = frac_table_pixel;
-	struct clk *parent = *p = clk_get_parent_by_index(hw->clk, f->src);
+	struct clk *parent = clk_get_parent_by_index(hw->clk, f->src);
+
+	*p = __clk_get_hw(parent);
 
 	for (; frac->num; frac++) {
 		request = (rate * frac->den) / frac->num;
diff --git a/drivers/clk/qcom/common.c b/drivers/clk/qcom/common.c
index eeb3eea01f4c..e20d947db3e5 100644
--- a/drivers/clk/qcom/common.c
+++ b/drivers/clk/qcom/common.c
@@ -18,6 +18,7 @@
 #include <linux/reset-controller.h>
 
 #include "common.h"
+#include "clk-rcg.h"
 #include "clk-regmap.h"
 #include "reset.h"
 
@@ -27,6 +28,21 @@ struct qcom_cc {
 	struct clk *clks[];
 };
 
+const
+struct freq_tbl *qcom_find_freq(const struct freq_tbl *f, unsigned long rate)
+{
+	if (!f)
+		return NULL;
+
+	for (; f->freq; f++)
+		if (rate <= f->freq)
+			return f;
+
+	/* Default to our fastest rate */
+	return f - 1;
+}
+EXPORT_SYMBOL_GPL(qcom_find_freq);
+
 struct regmap *
 qcom_cc_map(struct platform_device *pdev, const struct qcom_cc_desc *desc)
 {
diff --git a/drivers/clk/qcom/common.h b/drivers/clk/qcom/common.h
index 2765e9d3da97..f519322acdf3 100644
--- a/drivers/clk/qcom/common.h
+++ b/drivers/clk/qcom/common.h
@@ -18,6 +18,7 @@ struct regmap_config;
 struct clk_regmap;
 struct qcom_reset_map;
 struct regmap;
+struct freq_tbl;
 
 struct qcom_cc_desc {
 	const struct regmap_config *config;
@@ -27,6 +28,9 @@ struct qcom_cc_desc {
 	size_t num_resets;
 };
 
+extern const struct freq_tbl *qcom_find_freq(const struct freq_tbl *f,
+					     unsigned long rate);
+
 extern struct regmap *qcom_cc_map(struct platform_device *pdev,
 				  const struct qcom_cc_desc *desc);
 extern int qcom_cc_really_probe(struct platform_device *pdev,
diff --git a/drivers/clk/qcom/gcc-apq8084.c b/drivers/clk/qcom/gcc-apq8084.c
index ee52eb1c838a..e3ef90264214 100644
--- a/drivers/clk/qcom/gcc-apq8084.c
+++ b/drivers/clk/qcom/gcc-apq8084.c
@@ -3589,7 +3589,6 @@ static struct platform_driver gcc_apq8084_driver = {
 	.remove		= gcc_apq8084_remove,
 	.driver		= {
 		.name	= "gcc-apq8084",
-		.owner	= THIS_MODULE,
 		.of_match_table = gcc_apq8084_match_table,
 	},
 };
diff --git a/drivers/clk/qcom/gcc-ipq806x.c b/drivers/clk/qcom/gcc-ipq806x.c
index 3b83b7dd78c7..afed5eb0691e 100644
--- a/drivers/clk/qcom/gcc-ipq806x.c
+++ b/drivers/clk/qcom/gcc-ipq806x.c
@@ -32,6 +32,33 @@
 #include "clk-branch.h"
 #include "reset.h"
 
+static struct clk_pll pll0 = {
+	.l_reg = 0x30c4,
+	.m_reg = 0x30c8,
+	.n_reg = 0x30cc,
+	.config_reg = 0x30d4,
+	.mode_reg = 0x30c0,
+	.status_reg = 0x30d8,
+	.status_bit = 16,
+	.clkr.hw.init = &(struct clk_init_data){
+		.name = "pll0",
+		.parent_names = (const char *[]){ "pxo" },
+		.num_parents = 1,
+		.ops = &clk_pll_ops,
+	},
+};
+
+static struct clk_regmap pll0_vote = {
+	.enable_reg = 0x34c0,
+	.enable_mask = BIT(0),
+	.hw.init = &(struct clk_init_data){
+		.name = "pll0_vote",
+		.parent_names = (const char *[]){ "pll0" },
+		.num_parents = 1,
+		.ops = &clk_pll_vote_ops,
+	},
+};
+
 static struct clk_pll pll3 = {
 	.l_reg = 0x3164,
 	.m_reg = 0x3168,
@@ -154,7 +181,7 @@ static const u8 gcc_pxo_pll8_pll0[] = {
 static const char *gcc_pxo_pll8_pll0_map[] = {
 	"pxo",
 	"pll8_vote",
-	"pll0",
+	"pll0_vote",
 };
 
 static struct freq_tbl clk_tbl_gsbi_uart[] = {
@@ -2133,6 +2160,8 @@ static struct clk_branch usb_fs1_h_clk = {
 };
 
 static struct clk_regmap *gcc_ipq806x_clks[] = {
+	[PLL0] = &pll0.clkr,
+	[PLL0_VOTE] = &pll0_vote,
 	[PLL3] = &pll3.clkr,
 	[PLL8] = &pll8.clkr,
 	[PLL8_VOTE] = &pll8_vote,
@@ -2402,7 +2431,6 @@ static struct platform_driver gcc_ipq806x_driver = {
 	.remove		= gcc_ipq806x_remove,
 	.driver		= {
 		.name	= "gcc-ipq806x",
-		.owner	= THIS_MODULE,
 		.of_match_table = gcc_ipq806x_match_table,
 	},
 };
diff --git a/drivers/clk/qcom/gcc-msm8660.c b/drivers/clk/qcom/gcc-msm8660.c
index 0c4b727ae429..f366e68f7316 100644
--- a/drivers/clk/qcom/gcc-msm8660.c
+++ b/drivers/clk/qcom/gcc-msm8660.c
@@ -2744,7 +2744,6 @@ static struct platform_driver gcc_msm8660_driver = {
 	.remove		= gcc_msm8660_remove,
 	.driver		= {
 		.name	= "gcc-msm8660",
-		.owner	= THIS_MODULE,
 		.of_match_table = gcc_msm8660_match_table,
 	},
 };
diff --git a/drivers/clk/qcom/gcc-msm8960.c b/drivers/clk/qcom/gcc-msm8960.c
index 007534f7a2d7..b0b562b9ce0e 100644
--- a/drivers/clk/qcom/gcc-msm8960.c
+++ b/drivers/clk/qcom/gcc-msm8960.c
@@ -3519,7 +3519,6 @@ static struct platform_driver gcc_msm8960_driver = {
 	.remove		= gcc_msm8960_remove,
 	.driver		= {
 		.name	= "gcc-msm8960",
-		.owner	= THIS_MODULE,
 		.of_match_table = gcc_msm8960_match_table,
 	},
 };
diff --git a/drivers/clk/qcom/gcc-msm8974.c b/drivers/clk/qcom/gcc-msm8974.c
index 7af7c18d2144..a6937fe78d8a 100644
--- a/drivers/clk/qcom/gcc-msm8974.c
+++ b/drivers/clk/qcom/gcc-msm8974.c
@@ -2737,7 +2737,6 @@ static struct platform_driver gcc_msm8974_driver = {
 	.remove		= gcc_msm8974_remove,
 	.driver		= {
 		.name	= "gcc-msm8974",
-		.owner	= THIS_MODULE,
 		.of_match_table = gcc_msm8974_match_table,
 	},
 };
diff --git a/drivers/clk/qcom/mmcc-apq8084.c b/drivers/clk/qcom/mmcc-apq8084.c
index 751eea376a2b..157139a5c1ca 100644
--- a/drivers/clk/qcom/mmcc-apq8084.c
+++ b/drivers/clk/qcom/mmcc-apq8084.c
@@ -3122,7 +3122,7 @@ static struct clk_regmap *mmcc_apq8084_clocks[] = {
 	[ESC1_CLK_SRC] = &esc1_clk_src.clkr,
 	[HDMI_CLK_SRC] = &hdmi_clk_src.clkr,
 	[VSYNC_CLK_SRC] = &vsync_clk_src.clkr,
-	[RBCPR_CLK_SRC] = &rbcpr_clk_src.clkr,
+	[MMSS_RBCPR_CLK_SRC] = &rbcpr_clk_src.clkr,
 	[RBBMTIMER_CLK_SRC] = &rbbmtimer_clk_src.clkr,
 	[MAPLE_CLK_SRC] = &maple_clk_src.clkr,
 	[VDP_CLK_SRC] = &vdp_clk_src.clkr,
@@ -3341,7 +3341,6 @@ static struct platform_driver mmcc_apq8084_driver = {
 	.remove		= mmcc_apq8084_remove,
 	.driver		= {
 		.name	= "mmcc-apq8084",
-		.owner	= THIS_MODULE,
 		.of_match_table = mmcc_apq8084_match_table,
 	},
 };
diff --git a/drivers/clk/qcom/mmcc-msm8960.c b/drivers/clk/qcom/mmcc-msm8960.c
index 2e80a219b8ea..e8b33bbc362f 100644
--- a/drivers/clk/qcom/mmcc-msm8960.c
+++ b/drivers/clk/qcom/mmcc-msm8960.c
@@ -773,9 +773,11 @@ static struct freq_tbl clk_tbl_gfx2d[] = {
 };
 
 static struct clk_dyn_rcg gfx2d0_src = {
-	.ns_reg = 0x0070,
+	.ns_reg[0] = 0x0070,
+	.ns_reg[1] = 0x0070,
 	.md_reg[0] = 0x0064,
 	.md_reg[1] = 0x0068,
+	.bank_reg = 0x0060,
 	.mn[0] = {
 		.mnctr_en_bit = 8,
 		.mnctr_reset_bit = 25,
@@ -831,9 +833,11 @@ static struct clk_branch gfx2d0_clk = {
 };
 
 static struct clk_dyn_rcg gfx2d1_src = {
-	.ns_reg = 0x007c,
+	.ns_reg[0] = 0x007c,
+	.ns_reg[1] = 0x007c,
 	.md_reg[0] = 0x0078,
 	.md_reg[1] = 0x006c,
+	.bank_reg = 0x0074,
 	.mn[0] = {
 		.mnctr_en_bit = 8,
 		.mnctr_reset_bit = 25,
@@ -930,9 +934,11 @@ static struct freq_tbl clk_tbl_gfx3d_8064[] = {
 };
 
 static struct clk_dyn_rcg gfx3d_src = {
-	.ns_reg = 0x008c,
+	.ns_reg[0] = 0x008c,
+	.ns_reg[1] = 0x008c,
 	.md_reg[0] = 0x0084,
 	.md_reg[1] = 0x0088,
+	.bank_reg = 0x0080,
 	.mn[0] = {
 		.mnctr_en_bit = 8,
 		.mnctr_reset_bit = 25,
@@ -1006,9 +1012,11 @@ static struct freq_tbl clk_tbl_vcap[] = {
 };
 
 static struct clk_dyn_rcg vcap_src = {
-	.ns_reg = 0x021c,
+	.ns_reg[0] = 0x021c,
+	.ns_reg[1] = 0x021c,
 	.md_reg[0] = 0x01ec,
 	.md_reg[1] = 0x0218,
+	.bank_reg = 0x0178,
 	.mn[0] = {
 		.mnctr_en_bit = 8,
 		.mnctr_reset_bit = 23,
@@ -1211,9 +1219,11 @@ static struct freq_tbl clk_tbl_mdp[] = {
 };
 
 static struct clk_dyn_rcg mdp_src = {
-	.ns_reg = 0x00d0,
+	.ns_reg[0] = 0x00d0,
+	.ns_reg[1] = 0x00d0,
 	.md_reg[0] = 0x00c4,
 	.md_reg[1] = 0x00c8,
+	.bank_reg = 0x00c0,
 	.mn[0] = {
 		.mnctr_en_bit = 8,
 		.mnctr_reset_bit = 31,
@@ -1318,7 +1328,9 @@ static struct freq_tbl clk_tbl_rot[] = {
 };
 
 static struct clk_dyn_rcg rot_src = {
-	.ns_reg = 0x00e8,
+	.ns_reg[0] = 0x00e8,
+	.ns_reg[1] = 0x00e8,
+	.bank_reg = 0x00e8,
 	.p[0] = {
 		.pre_div_shift = 22,
 		.pre_div_width = 4,
@@ -1542,9 +1554,11 @@ static struct freq_tbl clk_tbl_vcodec[] = {
 };
 
 static struct clk_dyn_rcg vcodec_src = {
-	.ns_reg = 0x0100,
+	.ns_reg[0] = 0x0100,
+	.ns_reg[1] = 0x0100,
 	.md_reg[0] = 0x00fc,
 	.md_reg[1] = 0x0128,
+	.bank_reg = 0x00f8,
 	.mn[0] = {
 		.mnctr_en_bit = 5,
 		.mnctr_reset_bit = 31,
@@ -2679,7 +2693,6 @@ static struct platform_driver mmcc_msm8960_driver = {
 	.remove		= mmcc_msm8960_remove,
 	.driver		= {
 		.name	= "mmcc-msm8960",
-		.owner	= THIS_MODULE,
 		.of_match_table = mmcc_msm8960_match_table,
 	},
 };
diff --git a/drivers/clk/qcom/mmcc-msm8974.c b/drivers/clk/qcom/mmcc-msm8974.c
index bc8f519c47aa..be94c54a9a4f 100644
--- a/drivers/clk/qcom/mmcc-msm8974.c
+++ b/drivers/clk/qcom/mmcc-msm8974.c
@@ -2570,7 +2570,6 @@ static struct platform_driver mmcc_msm8974_driver = {
 	.remove		= mmcc_msm8974_remove,
 	.driver		= {
 		.name	= "mmcc-msm8974",
-		.owner	= THIS_MODULE,
 		.of_match_table = mmcc_msm8974_match_table,
 	},
 };
diff --git a/drivers/clk/rockchip/Makefile b/drivers/clk/rockchip/Makefile
index ee6b077381e1..2714097f90db 100644
--- a/drivers/clk/rockchip/Makefile
+++ b/drivers/clk/rockchip/Makefile
@@ -5,6 +5,8 @@
 obj-y	+= clk-rockchip.o
 obj-y	+= clk.o
 obj-y	+= clk-pll.o
+obj-y	+= clk-cpu.o
+obj-y	+= clk-mmc-phase.o
 obj-$(CONFIG_RESET_CONTROLLER)	+= softrst.o
 
 obj-y	+= clk-rk3188.o
diff --git a/drivers/clk/rockchip/clk-cpu.c b/drivers/clk/rockchip/clk-cpu.c
new file mode 100644
index 000000000000..8539c4fd34cc
--- /dev/null
+++ b/drivers/clk/rockchip/clk-cpu.c
@@ -0,0 +1,331 @@
+/*
+ * Copyright (c) 2014 MundoReader S.L.
+ * Author: Heiko Stuebner <heiko@sntech.de>
+ *
+ * based on clk/samsung/clk-cpu.c
+ * Copyright (c) 2014 Samsung Electronics Co., Ltd.
+ * Author: Thomas Abraham <thomas.ab@samsung.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * A CPU clock is defined as a clock supplied to a CPU or a group of CPUs.
+ * The CPU clock is typically derived from a hierarchy of clock
+ * blocks which includes mux and divider blocks. There are a number of other
+ * auxiliary clocks supplied to the CPU domain such as the debug blocks and AXI
+ * clock for CPU domain. The rates of these auxiliary clocks are related to the
+ * CPU clock rate and this relation is usually specified in the hardware manual
+ * of the SoC or supplied after the SoC characterization.
+ *
+ * The below implementation of the CPU clock allows the rate changes of the CPU
+ * clock and the corresponding rate changes of the auxillary clocks of the CPU
+ * domain. The platform clock driver provides a clock register configuration
+ * for each configurable rate which is then used to program the clock hardware
+ * registers to acheive a fast co-oridinated rate change for all the CPU domain
+ * clocks.
+ *
+ * On a rate change request for the CPU clock, the rate change is propagated
+ * upto the PLL supplying the clock to the CPU domain clock blocks. While the
+ * CPU domain PLL is reconfigured, the CPU domain clocks are driven using an
+ * alternate clock source. If required, the alternate clock source is divided
+ * down in order to keep the output clock rate within the previous OPP limits.
+ */
+
+#include <linux/of.h>
+#include <linux/slab.h>
+#include <linux/io.h>
+#include <linux/clk-provider.h>
+#include "clk.h"
+
+/**
+ * struct rockchip_cpuclk: information about clock supplied to a CPU core.
+ * @hw:		handle between ccf and cpu clock.
+ * @alt_parent:	alternate parent clock to use when switching the speed
+ *		of the primary parent clock.
+ * @reg_base:	base register for cpu-clock values.
+ * @clk_nb:	clock notifier registered for changes in clock speed of the
+ *		primary parent clock.
+ * @rate_count:	number of rates in the rate_table
+ * @rate_table:	pll-rates and their associated dividers
+ * @reg_data:	cpu-specific register settings
+ * @lock:	clock lock
+ */
+struct rockchip_cpuclk {
+	struct clk_hw				hw;
+
+	struct clk_mux				cpu_mux;
+	const struct clk_ops			*cpu_mux_ops;
+
+	struct clk				*alt_parent;
+	void __iomem				*reg_base;
+	struct notifier_block			clk_nb;
+	unsigned int				rate_count;
+	struct rockchip_cpuclk_rate_table	*rate_table;
+	const struct rockchip_cpuclk_reg_data	*reg_data;
+	spinlock_t				*lock;
+};
+
+#define to_rockchip_cpuclk_hw(hw) container_of(hw, struct rockchip_cpuclk, hw)
+#define to_rockchip_cpuclk_nb(nb) \
+			container_of(nb, struct rockchip_cpuclk, clk_nb)
+
+static const struct rockchip_cpuclk_rate_table *rockchip_get_cpuclk_settings(
+			    struct rockchip_cpuclk *cpuclk, unsigned long rate)
+{
+	const struct rockchip_cpuclk_rate_table *rate_table =
+							cpuclk->rate_table;
+	int i;
+
+	for (i = 0; i < cpuclk->rate_count; i++) {
+		if (rate == rate_table[i].prate)
+			return &rate_table[i];
+	}
+
+	return NULL;
+}
+
+static unsigned long rockchip_cpuclk_recalc_rate(struct clk_hw *hw,
+					unsigned long parent_rate)
+{
+	struct rockchip_cpuclk *cpuclk = to_rockchip_cpuclk_hw(hw);
+	const struct rockchip_cpuclk_reg_data *reg_data = cpuclk->reg_data;
+	u32 clksel0 = readl_relaxed(cpuclk->reg_base + reg_data->core_reg);
+
+	clksel0 >>= reg_data->div_core_shift;
+	clksel0 &= reg_data->div_core_mask;
+	return parent_rate / (clksel0 + 1);
+}
+
+static const struct clk_ops rockchip_cpuclk_ops = {
+	.recalc_rate = rockchip_cpuclk_recalc_rate,
+};
+
+static void rockchip_cpuclk_set_dividers(struct rockchip_cpuclk *cpuclk,
+				const struct rockchip_cpuclk_rate_table *rate)
+{
+	int i;
+
+	/* alternate parent is active now. set the dividers */
+	for (i = 0; i < ARRAY_SIZE(rate->divs); i++) {
+		const struct rockchip_cpuclk_clksel *clksel = &rate->divs[i];
+
+		if (!clksel->reg)
+			continue;
+
+		pr_debug("%s: setting reg 0x%x to 0x%x\n",
+			 __func__, clksel->reg, clksel->val);
+		writel(clksel->val , cpuclk->reg_base + clksel->reg);
+	}
+}
+
+static int rockchip_cpuclk_pre_rate_change(struct rockchip_cpuclk *cpuclk,
+					   struct clk_notifier_data *ndata)
+{
+	const struct rockchip_cpuclk_reg_data *reg_data = cpuclk->reg_data;
+	unsigned long alt_prate, alt_div;
+	unsigned long flags;
+
+	alt_prate = clk_get_rate(cpuclk->alt_parent);
+
+	spin_lock_irqsave(cpuclk->lock, flags);
+
+	/*
+	 * If the old parent clock speed is less than the clock speed
+	 * of the alternate parent, then it should be ensured that at no point
+	 * the armclk speed is more than the old_rate until the dividers are
+	 * set.
+	 */
+	if (alt_prate > ndata->old_rate) {
+		/* calculate dividers */
+		alt_div =  DIV_ROUND_UP(alt_prate, ndata->old_rate) - 1;
+		if (alt_div > reg_data->div_core_mask) {
+			pr_warn("%s: limiting alt-divider %lu to %d\n",
+				__func__, alt_div, reg_data->div_core_mask);
+			alt_div = reg_data->div_core_mask;
+		}
+
+		/*
+		 * Change parents and add dividers in a single transaction.
+		 *
+		 * NOTE: we do this in a single transaction so we're never
+		 * dividing the primary parent by the extra dividers that were
+		 * needed for the alt.
+		 */
+		pr_debug("%s: setting div %lu as alt-rate %lu > old-rate %lu\n",
+			 __func__, alt_div, alt_prate, ndata->old_rate);
+
+		writel(HIWORD_UPDATE(alt_div, reg_data->div_core_mask,
+					      reg_data->div_core_shift) |
+		       HIWORD_UPDATE(1, 1, reg_data->mux_core_shift),
+		       cpuclk->reg_base + reg_data->core_reg);
+	} else {
+		/* select alternate parent */
+		writel(HIWORD_UPDATE(1, 1, reg_data->mux_core_shift),
+			cpuclk->reg_base + reg_data->core_reg);
+	}
+
+	spin_unlock_irqrestore(cpuclk->lock, flags);
+	return 0;
+}
+
+static int rockchip_cpuclk_post_rate_change(struct rockchip_cpuclk *cpuclk,
+					    struct clk_notifier_data *ndata)
+{
+	const struct rockchip_cpuclk_reg_data *reg_data = cpuclk->reg_data;
+	const struct rockchip_cpuclk_rate_table *rate;
+	unsigned long flags;
+
+	rate = rockchip_get_cpuclk_settings(cpuclk, ndata->new_rate);
+	if (!rate) {
+		pr_err("%s: Invalid rate : %lu for cpuclk\n",
+		       __func__, ndata->new_rate);
+		return -EINVAL;
+	}
+
+	spin_lock_irqsave(cpuclk->lock, flags);
+
+	if (ndata->old_rate < ndata->new_rate)
+		rockchip_cpuclk_set_dividers(cpuclk, rate);
+
+	/*
+	 * post-rate change event, re-mux to primary parent and remove dividers.
+	 *
+	 * NOTE: we do this in a single transaction so we're never dividing the
+	 * primary parent by the extra dividers that were needed for the alt.
+	 */
+
+	writel(HIWORD_UPDATE(0, reg_data->div_core_mask,
+				reg_data->div_core_shift) |
+	       HIWORD_UPDATE(0, 1, reg_data->mux_core_shift),
+	       cpuclk->reg_base + reg_data->core_reg);
+
+	if (ndata->old_rate > ndata->new_rate)
+		rockchip_cpuclk_set_dividers(cpuclk, rate);
+
+	spin_unlock_irqrestore(cpuclk->lock, flags);
+	return 0;
+}
+
+/*
+ * This clock notifier is called when the frequency of the parent clock
+ * of cpuclk is to be changed. This notifier handles the setting up all
+ * the divider clocks, remux to temporary parent and handling the safe
+ * frequency levels when using temporary parent.
+ */
+static int rockchip_cpuclk_notifier_cb(struct notifier_block *nb,
+					unsigned long event, void *data)
+{
+	struct clk_notifier_data *ndata = data;
+	struct rockchip_cpuclk *cpuclk = to_rockchip_cpuclk_nb(nb);
+	int ret = 0;
+
+	pr_debug("%s: event %lu, old_rate %lu, new_rate: %lu\n",
+		 __func__, event, ndata->old_rate, ndata->new_rate);
+	if (event == PRE_RATE_CHANGE)
+		ret = rockchip_cpuclk_pre_rate_change(cpuclk, ndata);
+	else if (event == POST_RATE_CHANGE)
+		ret = rockchip_cpuclk_post_rate_change(cpuclk, ndata);
+
+	return notifier_from_errno(ret);
+}
+
+struct clk *rockchip_clk_register_cpuclk(const char *name,
+			const char **parent_names, u8 num_parents,
+			const struct rockchip_cpuclk_reg_data *reg_data,
+			const struct rockchip_cpuclk_rate_table *rates,
+			int nrates, void __iomem *reg_base, spinlock_t *lock)
+{
+	struct rockchip_cpuclk *cpuclk;
+	struct clk_init_data init;
+	struct clk *clk, *cclk;
+	int ret;
+
+	if (num_parents != 2) {
+		pr_err("%s: needs two parent clocks\n", __func__);
+		return ERR_PTR(-EINVAL);
+	}
+
+	cpuclk = kzalloc(sizeof(*cpuclk), GFP_KERNEL);
+	if (!cpuclk)
+		return ERR_PTR(-ENOMEM);
+
+	init.name = name;
+	init.parent_names = &parent_names[0];
+	init.num_parents = 1;
+	init.ops = &rockchip_cpuclk_ops;
+
+	/* only allow rate changes when we have a rate table */
+	init.flags = (nrates > 0) ? CLK_SET_RATE_PARENT : 0;
+
+	/* disallow automatic parent changes by ccf */
+	init.flags |= CLK_SET_RATE_NO_REPARENT;
+
+	init.flags |= CLK_GET_RATE_NOCACHE;
+
+	cpuclk->reg_base = reg_base;
+	cpuclk->lock = lock;
+	cpuclk->reg_data = reg_data;
+	cpuclk->clk_nb.notifier_call = rockchip_cpuclk_notifier_cb;
+	cpuclk->hw.init = &init;
+
+	cpuclk->alt_parent = __clk_lookup(parent_names[1]);
+	if (!cpuclk->alt_parent) {
+		pr_err("%s: could not lookup alternate parent\n",
+		       __func__);
+		ret = -EINVAL;
+		goto free_cpuclk;
+	}
+
+	ret = clk_prepare_enable(cpuclk->alt_parent);
+	if (ret) {
+		pr_err("%s: could not enable alternate parent\n",
+		       __func__);
+		goto free_cpuclk;
+	}
+
+	clk = __clk_lookup(parent_names[0]);
+	if (!clk) {
+		pr_err("%s: could not lookup parent clock %s\n",
+		       __func__, parent_names[0]);
+		ret = -EINVAL;
+		goto free_cpuclk;
+	}
+
+	ret = clk_notifier_register(clk, &cpuclk->clk_nb);
+	if (ret) {
+		pr_err("%s: failed to register clock notifier for %s\n",
+				__func__, name);
+		goto free_cpuclk;
+	}
+
+	if (nrates > 0) {
+		cpuclk->rate_count = nrates;
+		cpuclk->rate_table = kmemdup(rates,
+					     sizeof(*rates) * nrates,
+					     GFP_KERNEL);
+		if (!cpuclk->rate_table) {
+			pr_err("%s: could not allocate memory for cpuclk rates\n",
+			       __func__);
+			ret = -ENOMEM;
+			goto unregister_notifier;
+		}
+	}
+
+	cclk = clk_register(NULL, &cpuclk->hw);
+	if (IS_ERR(clk)) {
+		pr_err("%s: could not register cpuclk %s\n", __func__,	name);
+		ret = PTR_ERR(clk);
+		goto free_rate_table;
+	}
+
+	return cclk;
+
+free_rate_table:
+	kfree(cpuclk->rate_table);
+unregister_notifier:
+	clk_notifier_unregister(clk, &cpuclk->clk_nb);
+free_cpuclk:
+	kfree(cpuclk);
+	return ERR_PTR(ret);
+}
diff --git a/drivers/clk/rockchip/clk-mmc-phase.c b/drivers/clk/rockchip/clk-mmc-phase.c
new file mode 100644
index 000000000000..c842e3b60f21
--- /dev/null
+++ b/drivers/clk/rockchip/clk-mmc-phase.c
@@ -0,0 +1,154 @@
+/*
+ * Copyright 2014 Google, Inc
+ * Author: Alexandru M Stan <amstan@chromium.org>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/slab.h>
+#include <linux/clk-provider.h>
+#include "clk.h"
+
+struct rockchip_mmc_clock {
+	struct clk_hw	hw;
+	void __iomem	*reg;
+	int		id;
+	int		shift;
+};
+
+#define to_mmc_clock(_hw) container_of(_hw, struct rockchip_mmc_clock, hw)
+
+#define RK3288_MMC_CLKGEN_DIV 2
+
+static unsigned long rockchip_mmc_recalc(struct clk_hw *hw,
+					 unsigned long parent_rate)
+{
+	return parent_rate / RK3288_MMC_CLKGEN_DIV;
+}
+
+#define ROCKCHIP_MMC_DELAY_SEL BIT(10)
+#define ROCKCHIP_MMC_DEGREE_MASK 0x3
+#define ROCKCHIP_MMC_DELAYNUM_OFFSET 2
+#define ROCKCHIP_MMC_DELAYNUM_MASK (0xff << ROCKCHIP_MMC_DELAYNUM_OFFSET)
+
+#define PSECS_PER_SEC 1000000000000LL
+
+/*
+ * Each fine delay is between 40ps-80ps. Assume each fine delay is 60ps to
+ * simplify calculations. So 45degs could be anywhere between 33deg and 66deg.
+ */
+#define ROCKCHIP_MMC_DELAY_ELEMENT_PSEC 60
+
+static int rockchip_mmc_get_phase(struct clk_hw *hw)
+{
+	struct rockchip_mmc_clock *mmc_clock = to_mmc_clock(hw);
+	unsigned long rate = clk_get_rate(hw->clk);
+	u32 raw_value;
+	u16 degrees;
+	u32 delay_num = 0;
+
+	raw_value = readl(mmc_clock->reg) >> (mmc_clock->shift);
+
+	degrees = (raw_value & ROCKCHIP_MMC_DEGREE_MASK) * 90;
+
+	if (raw_value & ROCKCHIP_MMC_DELAY_SEL) {
+		/* degrees/delaynum * 10000 */
+		unsigned long factor = (ROCKCHIP_MMC_DELAY_ELEMENT_PSEC / 10) *
+					36 * (rate / 1000000);
+
+		delay_num = (raw_value & ROCKCHIP_MMC_DELAYNUM_MASK);
+		delay_num >>= ROCKCHIP_MMC_DELAYNUM_OFFSET;
+		degrees += delay_num * factor / 10000;
+	}
+
+	return degrees % 360;
+}
+
+static int rockchip_mmc_set_phase(struct clk_hw *hw, int degrees)
+{
+	struct rockchip_mmc_clock *mmc_clock = to_mmc_clock(hw);
+	unsigned long rate = clk_get_rate(hw->clk);
+	u8 nineties, remainder;
+	u8 delay_num;
+	u32 raw_value;
+	u64 delay;
+
+	/* allow 22 to be 22.5 */
+	degrees++;
+	/* floor to 22.5 increment */
+	degrees -= ((degrees) * 10 % 225) / 10;
+
+	nineties = degrees / 90;
+	/* 22.5 multiples */
+	remainder = (degrees % 90) / 22;
+
+	delay = PSECS_PER_SEC;
+	do_div(delay, rate);
+	/* / 360 / 22.5 */
+	do_div(delay, 16);
+	do_div(delay, ROCKCHIP_MMC_DELAY_ELEMENT_PSEC);
+
+	delay *= remainder;
+	delay_num = (u8) min(delay, 255ULL);
+
+	raw_value = delay_num ? ROCKCHIP_MMC_DELAY_SEL : 0;
+	raw_value |= delay_num << ROCKCHIP_MMC_DELAYNUM_OFFSET;
+	raw_value |= nineties;
+	writel(HIWORD_UPDATE(raw_value, 0x07ff, mmc_clock->shift), mmc_clock->reg);
+
+	pr_debug("%s->set_phase(%d) delay_nums=%u reg[0x%p]=0x%03x actual_degrees=%d\n",
+		__clk_get_name(hw->clk), degrees, delay_num,
+		mmc_clock->reg, raw_value>>(mmc_clock->shift),
+		rockchip_mmc_get_phase(hw)
+	);
+
+	return 0;
+}
+
+static const struct clk_ops rockchip_mmc_clk_ops = {
+	.recalc_rate	= rockchip_mmc_recalc,
+	.get_phase	= rockchip_mmc_get_phase,
+	.set_phase	= rockchip_mmc_set_phase,
+};
+
+struct clk *rockchip_clk_register_mmc(const char *name,
+				const char **parent_names, u8 num_parents,
+				void __iomem *reg, int shift)
+{
+	struct clk_init_data init;
+	struct rockchip_mmc_clock *mmc_clock;
+	struct clk *clk;
+
+	mmc_clock = kmalloc(sizeof(*mmc_clock), GFP_KERNEL);
+	if (!mmc_clock)
+		return NULL;
+
+	init.num_parents = num_parents;
+	init.parent_names = parent_names;
+	init.ops = &rockchip_mmc_clk_ops;
+
+	mmc_clock->hw.init = &init;
+	mmc_clock->reg = reg;
+	mmc_clock->shift = shift;
+
+	if (name)
+		init.name = name;
+
+	clk = clk_register(NULL, &mmc_clock->hw);
+	if (IS_ERR(clk))
+		goto err_free;
+
+	return clk;
+
+err_free:
+	kfree(mmc_clock);
+	return NULL;
+}
diff --git a/drivers/clk/rockchip/clk-pll.c b/drivers/clk/rockchip/clk-pll.c
index f2a1c7abf4d9..f8d3baf275b2 100644
--- a/drivers/clk/rockchip/clk-pll.c
+++ b/drivers/clk/rockchip/clk-pll.c
@@ -34,12 +34,12 @@ struct rockchip_clk_pll {
 	const struct clk_ops	*pll_mux_ops;
 
 	struct notifier_block	clk_nb;
-	bool			rate_change_remuxed;
 
 	void __iomem		*reg_base;
 	int			lock_offset;
 	unsigned int		lock_shift;
 	enum rockchip_pll_type	type;
+	u8			flags;
 	const struct rockchip_pll_rate_table *rate_table;
 	unsigned int		rate_count;
 	spinlock_t		*lock;
@@ -109,38 +109,6 @@ static int rockchip_pll_wait_lock(struct rockchip_clk_pll *pll)
 }
 
 /**
- * Set pll mux when changing the pll rate.
- * This makes sure to move the pll mux away from the actual pll before
- * changing its rate and back to the original parent after the change.
- */
-static int rockchip_pll_notifier_cb(struct notifier_block *nb,
-					unsigned long event, void *data)
-{
-	struct rockchip_clk_pll *pll = to_rockchip_clk_pll_nb(nb);
-	struct clk_mux *pll_mux = &pll->pll_mux;
-	const struct clk_ops *pll_mux_ops = pll->pll_mux_ops;
-	int cur_parent;
-
-	switch (event) {
-	case PRE_RATE_CHANGE:
-		cur_parent = pll_mux_ops->get_parent(&pll_mux->hw);
-		if (cur_parent == PLL_MODE_NORM) {
-			pll_mux_ops->set_parent(&pll_mux->hw, PLL_MODE_SLOW);
-			pll->rate_change_remuxed = 1;
-		}
-		break;
-	case POST_RATE_CHANGE:
-		if (pll->rate_change_remuxed) {
-			pll_mux_ops->set_parent(&pll_mux->hw, PLL_MODE_NORM);
-			pll->rate_change_remuxed = 0;
-		}
-		break;
-	}
-
-	return NOTIFY_OK;
-}
-
-/**
  * PLL used in RK3066, RK3188 and RK3288
  */
 
@@ -194,6 +162,10 @@ static int rockchip_rk3066_pll_set_rate(struct clk_hw *hw, unsigned long drate,
 	const struct rockchip_pll_rate_table *rate;
 	unsigned long old_rate = rockchip_rk3066_pll_recalc_rate(hw, prate);
 	struct regmap *grf = rockchip_clk_get_grf();
+	struct clk_mux *pll_mux = &pll->pll_mux;
+	const struct clk_ops *pll_mux_ops = pll->pll_mux_ops;
+	int rate_change_remuxed = 0;
+	int cur_parent;
 	int ret;
 
 	if (IS_ERR(grf)) {
@@ -216,6 +188,12 @@ static int rockchip_rk3066_pll_set_rate(struct clk_hw *hw, unsigned long drate,
 	pr_debug("%s: rate settings for %lu (nr, no, nf): (%d, %d, %d)\n",
 		 __func__, rate->rate, rate->nr, rate->no, rate->nf);
 
+	cur_parent = pll_mux_ops->get_parent(&pll_mux->hw);
+	if (cur_parent == PLL_MODE_NORM) {
+		pll_mux_ops->set_parent(&pll_mux->hw, PLL_MODE_SLOW);
+		rate_change_remuxed = 1;
+	}
+
 	/* enter reset mode */
 	writel(HIWORD_UPDATE(RK3066_PLLCON3_RESET, RK3066_PLLCON3_RESET, 0),
 	       pll->reg_base + RK3066_PLLCON(3));
@@ -247,6 +225,9 @@ static int rockchip_rk3066_pll_set_rate(struct clk_hw *hw, unsigned long drate,
 		rockchip_rk3066_pll_set_rate(hw, old_rate, prate);
 	}
 
+	if (rate_change_remuxed)
+		pll_mux_ops->set_parent(&pll_mux->hw, PLL_MODE_NORM);
+
 	return ret;
 }
 
@@ -277,6 +258,55 @@ static int rockchip_rk3066_pll_is_enabled(struct clk_hw *hw)
 	return !(pllcon & RK3066_PLLCON3_PWRDOWN);
 }
 
+static void rockchip_rk3066_pll_init(struct clk_hw *hw)
+{
+	struct rockchip_clk_pll *pll = to_rockchip_clk_pll(hw);
+	const struct rockchip_pll_rate_table *rate;
+	unsigned int nf, nr, no, bwadj;
+	unsigned long drate;
+	u32 pllcon;
+
+	if (!(pll->flags & ROCKCHIP_PLL_SYNC_RATE))
+		return;
+
+	drate = __clk_get_rate(hw->clk);
+	rate = rockchip_get_pll_settings(pll, drate);
+
+	/* when no rate setting for the current rate, rely on clk_set_rate */
+	if (!rate)
+		return;
+
+	pllcon = readl_relaxed(pll->reg_base + RK3066_PLLCON(0));
+	nr = ((pllcon >> RK3066_PLLCON0_NR_SHIFT) & RK3066_PLLCON0_NR_MASK) + 1;
+	no = ((pllcon >> RK3066_PLLCON0_OD_SHIFT) & RK3066_PLLCON0_OD_MASK) + 1;
+
+	pllcon = readl_relaxed(pll->reg_base + RK3066_PLLCON(1));
+	nf = ((pllcon >> RK3066_PLLCON1_NF_SHIFT) & RK3066_PLLCON1_NF_MASK) + 1;
+
+	pllcon = readl_relaxed(pll->reg_base + RK3066_PLLCON(2));
+	bwadj = (pllcon >> RK3066_PLLCON2_BWADJ_SHIFT) & RK3066_PLLCON2_BWADJ_MASK;
+
+	pr_debug("%s: pll %s@%lu: nr (%d:%d); no (%d:%d); nf(%d:%d), bwadj(%d:%d)\n",
+		 __func__, __clk_get_name(hw->clk), drate, rate->nr, nr,
+		rate->no, no, rate->nf, nf, rate->bwadj, bwadj);
+	if (rate->nr != nr || rate->no != no || rate->nf != nf
+					     || rate->bwadj != bwadj) {
+		struct clk *parent = __clk_get_parent(hw->clk);
+		unsigned long prate;
+
+		if (!parent) {
+			pr_warn("%s: parent of %s not available\n",
+				__func__, __clk_get_name(hw->clk));
+			return;
+		}
+
+		pr_debug("%s: pll %s: rate params do not match rate table, adjusting\n",
+			 __func__, __clk_get_name(hw->clk));
+		prate = __clk_get_rate(parent);
+		rockchip_rk3066_pll_set_rate(hw, drate, prate);
+	}
+}
+
 static const struct clk_ops rockchip_rk3066_pll_clk_norate_ops = {
 	.recalc_rate = rockchip_rk3066_pll_recalc_rate,
 	.enable = rockchip_rk3066_pll_enable,
@@ -291,6 +321,7 @@ static const struct clk_ops rockchip_rk3066_pll_clk_ops = {
 	.enable = rockchip_rk3066_pll_enable,
 	.disable = rockchip_rk3066_pll_disable,
 	.is_enabled = rockchip_rk3066_pll_is_enabled,
+	.init = rockchip_rk3066_pll_init,
 };
 
 /*
@@ -302,7 +333,7 @@ struct clk *rockchip_clk_register_pll(enum rockchip_pll_type pll_type,
 		void __iomem *base, int con_offset, int grf_lock_offset,
 		int lock_shift, int mode_offset, int mode_shift,
 		struct rockchip_pll_rate_table *rate_table,
-		spinlock_t *lock)
+		u8 clk_pll_flags, spinlock_t *lock)
 {
 	const char *pll_parents[3];
 	struct clk_init_data init;
@@ -310,7 +341,6 @@ struct clk *rockchip_clk_register_pll(enum rockchip_pll_type pll_type,
 	struct clk_mux *pll_mux;
 	struct clk *pll_clk, *mux_clk;
 	char pll_name[20];
-	int ret;
 
 	if (num_parents != 2) {
 		pr_err("%s: needs two parent clocks\n", __func__);
@@ -366,8 +396,21 @@ struct clk *rockchip_clk_register_pll(enum rockchip_pll_type pll_type,
 	pll->reg_base = base + con_offset;
 	pll->lock_offset = grf_lock_offset;
 	pll->lock_shift = lock_shift;
+	pll->flags = clk_pll_flags;
 	pll->lock = lock;
-	pll->clk_nb.notifier_call = rockchip_pll_notifier_cb;
+
+	/* create the mux on top of the real pll */
+	pll->pll_mux_ops = &clk_mux_ops;
+	pll_mux = &pll->pll_mux;
+	pll_mux->reg = base + mode_offset;
+	pll_mux->shift = mode_shift;
+	pll_mux->mask = PLL_MODE_MASK;
+	pll_mux->flags = 0;
+	pll_mux->lock = lock;
+	pll_mux->hw.init = &init;
+
+	if (pll_type == pll_rk3066)
+		pll_mux->flags |= CLK_MUX_HIWORD_MASK;
 
 	pll_clk = clk_register(NULL, &pll->hw);
 	if (IS_ERR(pll_clk)) {
@@ -377,18 +420,6 @@ struct clk *rockchip_clk_register_pll(enum rockchip_pll_type pll_type,
 		goto err_pll;
 	}
 
-	ret = clk_notifier_register(pll_clk, &pll->clk_nb);
-	if (ret) {
-		pr_err("%s: failed to register clock notifier for %s : %d\n",
-				__func__, name, ret);
-		mux_clk = ERR_PTR(ret);
-		goto err_pll_notifier;
-	}
-
-	/* create the mux on top of the real pll */
-	pll->pll_mux_ops = &clk_mux_ops;
-	pll_mux = &pll->pll_mux;
-
 	/* the actual muxing is xin24m, pll-output, xin32k */
 	pll_parents[0] = parent_names[0];
 	pll_parents[1] = pll_name;
@@ -400,16 +431,6 @@ struct clk *rockchip_clk_register_pll(enum rockchip_pll_type pll_type,
 	init.parent_names = pll_parents;
 	init.num_parents = ARRAY_SIZE(pll_parents);
 
-	pll_mux->reg = base + mode_offset;
-	pll_mux->shift = mode_shift;
-	pll_mux->mask = PLL_MODE_MASK;
-	pll_mux->flags = 0;
-	pll_mux->lock = lock;
-	pll_mux->hw.init = &init;
-
-	if (pll_type == pll_rk3066)
-		pll_mux->flags |= CLK_MUX_HIWORD_MASK;
-
 	mux_clk = clk_register(NULL, &pll_mux->hw);
 	if (IS_ERR(mux_clk))
 		goto err_mux;
@@ -417,13 +438,6 @@ struct clk *rockchip_clk_register_pll(enum rockchip_pll_type pll_type,
 	return mux_clk;
 
 err_mux:
-	ret = clk_notifier_unregister(pll_clk, &pll->clk_nb);
-	if (ret) {
-		pr_err("%s: could not unregister clock notifier in error path : %d\n",
-		       __func__, ret);
-		return mux_clk;
-	}
-err_pll_notifier:
 	clk_unregister(pll_clk);
 err_pll:
 	kfree(pll);
diff --git a/drivers/clk/rockchip/clk-rk3188.c b/drivers/clk/rockchip/clk-rk3188.c
index a83a6d8d0fb6..7eb684c50d42 100644
--- a/drivers/clk/rockchip/clk-rk3188.c
+++ b/drivers/clk/rockchip/clk-rk3188.c
@@ -19,6 +19,7 @@
 #include <dt-bindings/clock/rk3188-cru-common.h>
 #include "clk.h"
 
+#define RK3066_GRF_SOC_STATUS	0x15c
 #define RK3188_GRF_SOC_STATUS	0xac
 
 enum rk3188_plls {
@@ -100,6 +101,98 @@ struct rockchip_pll_rate_table rk3188_pll_rates[] = {
 	{ /* sentinel */ },
 };
 
+#define RK3066_DIV_CORE_PERIPH_MASK	0x3
+#define RK3066_DIV_CORE_PERIPH_SHIFT	6
+#define RK3066_DIV_ACLK_CORE_MASK	0x7
+#define RK3066_DIV_ACLK_CORE_SHIFT	0
+#define RK3066_DIV_ACLK_HCLK_MASK	0x3
+#define RK3066_DIV_ACLK_HCLK_SHIFT	8
+#define RK3066_DIV_ACLK_PCLK_MASK	0x3
+#define RK3066_DIV_ACLK_PCLK_SHIFT	12
+#define RK3066_DIV_AHB2APB_MASK		0x3
+#define RK3066_DIV_AHB2APB_SHIFT	14
+
+#define RK3066_CLKSEL0(_core_peri)					\
+	{								\
+		.reg = RK2928_CLKSEL_CON(0),				\
+		.val = HIWORD_UPDATE(_core_peri, RK3066_DIV_CORE_PERIPH_MASK, \
+				RK3066_DIV_CORE_PERIPH_SHIFT)		\
+	}
+#define RK3066_CLKSEL1(_aclk_core, _aclk_hclk, _aclk_pclk, _ahb2apb)	\
+	{								\
+		.reg = RK2928_CLKSEL_CON(1),				\
+		.val = HIWORD_UPDATE(_aclk_core, RK3066_DIV_ACLK_CORE_MASK, \
+				RK3066_DIV_ACLK_CORE_SHIFT) |		\
+		       HIWORD_UPDATE(_aclk_hclk, RK3066_DIV_ACLK_HCLK_MASK, \
+				RK3066_DIV_ACLK_HCLK_SHIFT) |		\
+		       HIWORD_UPDATE(_aclk_pclk, RK3066_DIV_ACLK_PCLK_MASK, \
+				RK3066_DIV_ACLK_PCLK_SHIFT) |		\
+		       HIWORD_UPDATE(_ahb2apb, RK3066_DIV_AHB2APB_MASK,	\
+				RK3066_DIV_AHB2APB_SHIFT),		\
+	}
+
+#define RK3066_CPUCLK_RATE(_prate, _core_peri, _acore, _ahclk, _apclk, _h2p) \
+	{								\
+		.prate = _prate,					\
+		.divs = {						\
+			RK3066_CLKSEL0(_core_peri),			\
+			RK3066_CLKSEL1(_acore, _ahclk, _apclk, _h2p),	\
+		},							\
+	}
+
+static struct rockchip_cpuclk_rate_table rk3066_cpuclk_rates[] __initdata = {
+	RK3066_CPUCLK_RATE(1416000000, 2, 3, 1, 2, 1),
+	RK3066_CPUCLK_RATE(1200000000, 2, 3, 1, 2, 1),
+	RK3066_CPUCLK_RATE(1008000000, 2, 2, 1, 2, 1),
+	RK3066_CPUCLK_RATE( 816000000, 2, 2, 1, 2, 1),
+	RK3066_CPUCLK_RATE( 600000000, 1, 2, 1, 2, 1),
+	RK3066_CPUCLK_RATE( 504000000, 1, 1, 1, 2, 1),
+	RK3066_CPUCLK_RATE( 312000000, 0, 1, 1, 1, 0),
+};
+
+static const struct rockchip_cpuclk_reg_data rk3066_cpuclk_data = {
+	.core_reg = RK2928_CLKSEL_CON(0),
+	.div_core_shift = 0,
+	.div_core_mask = 0x1f,
+	.mux_core_shift = 8,
+};
+
+#define RK3188_DIV_ACLK_CORE_MASK	0x7
+#define RK3188_DIV_ACLK_CORE_SHIFT	3
+
+#define RK3188_CLKSEL1(_aclk_core)		\
+	{					\
+		.reg = RK2928_CLKSEL_CON(1),	\
+		.val = HIWORD_UPDATE(_aclk_core, RK3188_DIV_ACLK_CORE_MASK,\
+				 RK3188_DIV_ACLK_CORE_SHIFT) \
+	}
+#define RK3188_CPUCLK_RATE(_prate, _core_peri, _aclk_core)	\
+	{							\
+		.prate = _prate,				\
+		.divs = {					\
+			RK3066_CLKSEL0(_core_peri),		\
+			RK3188_CLKSEL1(_aclk_core),		\
+		},						\
+	}
+
+static struct rockchip_cpuclk_rate_table rk3188_cpuclk_rates[] __initdata = {
+	RK3188_CPUCLK_RATE(1608000000, 2, 3),
+	RK3188_CPUCLK_RATE(1416000000, 2, 3),
+	RK3188_CPUCLK_RATE(1200000000, 2, 3),
+	RK3188_CPUCLK_RATE(1008000000, 2, 3),
+	RK3188_CPUCLK_RATE( 816000000, 2, 3),
+	RK3188_CPUCLK_RATE( 600000000, 1, 3),
+	RK3188_CPUCLK_RATE( 504000000, 1, 3),
+	RK3188_CPUCLK_RATE( 312000000, 0, 1),
+};
+
+static const struct rockchip_cpuclk_reg_data rk3188_cpuclk_data = {
+	.core_reg = RK2928_CLKSEL_CON(0),
+	.div_core_shift = 9,
+	.div_core_mask = 0x1f,
+	.mux_core_shift = 8,
+};
+
 PNAME(mux_pll_p)		= { "xin24m", "xin32k" };
 PNAME(mux_armclk_p)		= { "apll", "gpll_armclk" };
 PNAME(mux_ddrphy_p)		= { "dpll", "gpll_ddr" };
@@ -117,15 +210,26 @@ PNAME(mux_sclk_hsadc_p)		= { "hsadc_src", "hsadc_frac", "ext_hsadc" };
 PNAME(mux_mac_p)		= { "gpll", "dpll" };
 PNAME(mux_sclk_macref_p)	= { "mac_src", "ext_rmii" };
 
+static struct rockchip_pll_clock rk3066_pll_clks[] __initdata = {
+	[apll] = PLL(pll_rk3066, PLL_APLL, "apll", mux_pll_p, 0, RK2928_PLL_CON(0),
+		     RK2928_MODE_CON, 0, 5, 0, rk3188_pll_rates),
+	[dpll] = PLL(pll_rk3066, PLL_DPLL, "dpll", mux_pll_p, 0, RK2928_PLL_CON(4),
+		     RK2928_MODE_CON, 4, 4, 0, NULL),
+	[cpll] = PLL(pll_rk3066, PLL_CPLL, "cpll", mux_pll_p, 0, RK2928_PLL_CON(8),
+		     RK2928_MODE_CON, 8, 6, ROCKCHIP_PLL_SYNC_RATE, rk3188_pll_rates),
+	[gpll] = PLL(pll_rk3066, PLL_GPLL, "gpll", mux_pll_p, 0, RK2928_PLL_CON(12),
+		     RK2928_MODE_CON, 12, 7, ROCKCHIP_PLL_SYNC_RATE, rk3188_pll_rates),
+};
+
 static struct rockchip_pll_clock rk3188_pll_clks[] __initdata = {
 	[apll] = PLL(pll_rk3066, PLL_APLL, "apll", mux_pll_p, 0, RK2928_PLL_CON(0),
-		     RK2928_MODE_CON, 0, 6, rk3188_pll_rates),
+		     RK2928_MODE_CON, 0, 6, 0, rk3188_pll_rates),
 	[dpll] = PLL(pll_rk3066, PLL_DPLL, "dpll", mux_pll_p, 0, RK2928_PLL_CON(4),
-		     RK2928_MODE_CON, 4, 5, NULL),
+		     RK2928_MODE_CON, 4, 5, 0, NULL),
 	[cpll] = PLL(pll_rk3066, PLL_CPLL, "cpll", mux_pll_p, 0, RK2928_PLL_CON(8),
-		     RK2928_MODE_CON, 8, 7, rk3188_pll_rates),
+		     RK2928_MODE_CON, 8, 7, ROCKCHIP_PLL_SYNC_RATE, rk3188_pll_rates),
 	[gpll] = PLL(pll_rk3066, PLL_GPLL, "gpll", mux_pll_p, 0, RK2928_PLL_CON(12),
-		     RK2928_MODE_CON, 12, 8, rk3188_pll_rates),
+		     RK2928_MODE_CON, 12, 8, ROCKCHIP_PLL_SYNC_RATE, rk3188_pll_rates),
 };
 
 #define MFLAGS CLK_MUX_HIWORD_MASK
@@ -164,30 +268,23 @@ static struct rockchip_clk_branch common_clk_branches[] __initdata = {
 	GATE(0, "hclk_vdpu", "aclk_vdpu", 0,
 			RK2928_CLKGATE_CON(3), 12, GFLAGS),
 
-	GATE(0, "gpll_ddr", "gpll", 0,
+	GATE(0, "gpll_ddr", "gpll", CLK_IGNORE_UNUSED,
 			RK2928_CLKGATE_CON(1), 7, GFLAGS),
-	COMPOSITE(0, "ddrphy", mux_ddrphy_p, 0,
+	COMPOSITE(0, "ddrphy", mux_ddrphy_p, CLK_IGNORE_UNUSED,
 			RK2928_CLKSEL_CON(26), 8, 1, MFLAGS, 0, 2, DFLAGS | CLK_DIVIDER_POWER_OF_TWO,
 			RK2928_CLKGATE_CON(0), 2, GFLAGS),
 
 	GATE(0, "aclk_cpu", "aclk_cpu_pre", 0,
 			RK2928_CLKGATE_CON(0), 3, GFLAGS),
 
-	DIV(0, "pclk_cpu_pre", "aclk_cpu_pre", 0,
-			RK2928_CLKSEL_CON(1), 12, 2, DFLAGS | CLK_DIVIDER_POWER_OF_TWO),
 	GATE(0, "atclk_cpu", "pclk_cpu_pre", 0,
 			RK2928_CLKGATE_CON(0), 6, GFLAGS),
 	GATE(0, "pclk_cpu", "pclk_cpu_pre", 0,
 			RK2928_CLKGATE_CON(0), 5, GFLAGS),
-	DIV(0, "hclk_cpu_pre", "aclk_cpu_pre", 0,
-			RK2928_CLKSEL_CON(1), 8, 2, DFLAGS | CLK_DIVIDER_POWER_OF_TWO),
-	COMPOSITE_NOMUX(0, "hclk_ahb2apb", "hclk_cpu_pre", 0,
-			RK2928_CLKSEL_CON(1), 14, 2, DFLAGS | CLK_DIVIDER_POWER_OF_TWO,
-			RK2928_CLKGATE_CON(4), 9, GFLAGS),
-	GATE(0, "hclk_cpu", "hclk_cpu_pre", 0,
+	GATE(0, "hclk_cpu", "hclk_cpu_pre", CLK_IGNORE_UNUSED,
 			RK2928_CLKGATE_CON(0), 4, GFLAGS),
 
-	COMPOSITE(0, "aclk_lcdc0_pre", mux_pll_src_cpll_gpll_p, 0,
+	COMPOSITE(0, "aclk_lcdc0_pre", mux_pll_src_cpll_gpll_p, CLK_IGNORE_UNUSED,
 			RK2928_CLKSEL_CON(31), 7, 1, MFLAGS, 0, 5, DFLAGS,
 			RK2928_CLKGATE_CON(3), 0, GFLAGS),
 	COMPOSITE(0, "aclk_lcdc1_pre", mux_pll_src_cpll_gpll_p, 0,
@@ -218,9 +315,9 @@ static struct rockchip_clk_branch common_clk_branches[] __initdata = {
 	 * the 480m are generated inside the usb block from these clocks,
 	 * but they are also a source for the hsicphy clock.
 	 */
-	GATE(SCLK_OTGPHY0, "sclk_otgphy0", "usb480m", 0,
+	GATE(SCLK_OTGPHY0, "sclk_otgphy0", "usb480m", CLK_IGNORE_UNUSED,
 			RK2928_CLKGATE_CON(1), 5, GFLAGS),
-	GATE(SCLK_OTGPHY1, "sclk_otgphy1", "usb480m", 0,
+	GATE(SCLK_OTGPHY1, "sclk_otgphy1", "usb480m", CLK_IGNORE_UNUSED,
 			RK2928_CLKGATE_CON(1), 6, GFLAGS),
 
 	COMPOSITE(0, "mac_src", mux_mac_p, 0,
@@ -234,9 +331,9 @@ static struct rockchip_clk_branch common_clk_branches[] __initdata = {
 	COMPOSITE(0, "hsadc_src", mux_pll_src_gpll_cpll_p, 0,
 			RK2928_CLKSEL_CON(22), 0, 1, MFLAGS, 8, 8, DFLAGS,
 			RK2928_CLKGATE_CON(2), 6, GFLAGS),
-	COMPOSITE_FRAC(0, "hsadc_frac", "hsadc_src",
+	COMPOSITE_FRAC(0, "hsadc_frac", "hsadc_src", 0,
 			RK2928_CLKSEL_CON(23), 0,
-			RK2928_CLKGATE_CON(2), 7, 0, GFLAGS),
+			RK2928_CLKGATE_CON(2), 7, GFLAGS),
 	MUX(SCLK_HSADC, "sclk_hsadc", mux_sclk_hsadc_p, 0,
 			RK2928_CLKSEL_CON(22), 4, 2, MFLAGS),
 
@@ -244,6 +341,15 @@ static struct rockchip_clk_branch common_clk_branches[] __initdata = {
 			RK2928_CLKSEL_CON(24), 8, 8, DFLAGS,
 			RK2928_CLKGATE_CON(2), 8, GFLAGS),
 
+	COMPOSITE_NOMUX(0, "spdif_pre", "i2s_src", 0,
+			RK2928_CLKSEL_CON(5), 0, 7, DFLAGS,
+			RK2928_CLKGATE_CON(0), 13, GFLAGS),
+	COMPOSITE_FRAC(0, "spdif_frac", "spdif_pll", 0,
+			RK2928_CLKSEL_CON(9), 0,
+			RK2928_CLKGATE_CON(0), 14, GFLAGS),
+	MUX(SCLK_SPDIF, "sclk_spdif", mux_sclk_spdif_p, 0,
+			RK2928_CLKSEL_CON(5), 8, 2, MFLAGS),
+
 	/*
 	 * Clock-Architecture Diagram 4
 	 */
@@ -313,8 +419,8 @@ static struct rockchip_clk_branch common_clk_branches[] __initdata = {
 
 	/* aclk_cpu gates */
 	GATE(ACLK_DMA1, "aclk_dma1", "aclk_cpu", 0, RK2928_CLKGATE_CON(5), 0, GFLAGS),
-	GATE(0, "aclk_intmem", "aclk_cpu", 0, RK2928_CLKGATE_CON(4), 12, GFLAGS),
-	GATE(0, "aclk_strc_sys", "aclk_cpu", 0, RK2928_CLKGATE_CON(4), 10, GFLAGS),
+	GATE(0, "aclk_intmem", "aclk_cpu", CLK_IGNORE_UNUSED, RK2928_CLKGATE_CON(4), 12, GFLAGS),
+	GATE(0, "aclk_strc_sys", "aclk_cpu", CLK_IGNORE_UNUSED, RK2928_CLKGATE_CON(4), 10, GFLAGS),
 
 	/* hclk_cpu gates */
 	GATE(HCLK_ROM, "hclk_rom", "hclk_cpu", 0, RK2928_CLKGATE_CON(5), 6, GFLAGS),
@@ -324,19 +430,19 @@ static struct rockchip_clk_branch common_clk_branches[] __initdata = {
 	/* hclk_ahb2apb is part of a clk branch */
 	GATE(0, "hclk_vio_bus", "hclk_cpu", 0, RK2928_CLKGATE_CON(6), 12, GFLAGS),
 	GATE(HCLK_LCDC0, "hclk_lcdc0", "hclk_cpu", 0, RK2928_CLKGATE_CON(6), 1, GFLAGS),
-	GATE(HCLK_LCDC1, "hclk_lcdc1", "aclk_cpu", 0, RK2928_CLKGATE_CON(6), 2, GFLAGS),
+	GATE(HCLK_LCDC1, "hclk_lcdc1", "hclk_cpu", 0, RK2928_CLKGATE_CON(6), 2, GFLAGS),
 	GATE(HCLK_CIF0, "hclk_cif0", "hclk_cpu", 0, RK2928_CLKGATE_CON(6), 4, GFLAGS),
 	GATE(HCLK_IPP, "hclk_ipp", "hclk_cpu", 0, RK2928_CLKGATE_CON(6), 9, GFLAGS),
 	GATE(HCLK_RGA, "hclk_rga", "hclk_cpu", 0, RK2928_CLKGATE_CON(6), 10, GFLAGS),
 
 	/* hclk_peri gates */
-	GATE(0, "hclk_peri_axi_matrix", "hclk_peri", 0, RK2928_CLKGATE_CON(4), 0, GFLAGS),
-	GATE(0, "hclk_peri_ahb_arbi", "hclk_peri", 0, RK2928_CLKGATE_CON(4), 6, GFLAGS),
-	GATE(0, "hclk_emem_peri", "hclk_peri", 0, RK2928_CLKGATE_CON(4), 7, GFLAGS),
+	GATE(0, "hclk_peri_axi_matrix", "hclk_peri", CLK_IGNORE_UNUSED, RK2928_CLKGATE_CON(4), 0, GFLAGS),
+	GATE(0, "hclk_peri_ahb_arbi", "hclk_peri", CLK_IGNORE_UNUSED, RK2928_CLKGATE_CON(4), 6, GFLAGS),
+	GATE(0, "hclk_emem_peri", "hclk_peri", CLK_IGNORE_UNUSED, RK2928_CLKGATE_CON(4), 7, GFLAGS),
 	GATE(HCLK_EMAC, "hclk_emac", "hclk_peri", 0, RK2928_CLKGATE_CON(7), 0, GFLAGS),
 	GATE(HCLK_NANDC0, "hclk_nandc0", "hclk_peri", 0, RK2928_CLKGATE_CON(5), 9, GFLAGS),
-	GATE(0, "hclk_usb_peri", "hclk_peri", 0, RK2928_CLKGATE_CON(4), 5, GFLAGS),
-	GATE(HCLK_OTG0, "hclk_usbotg0", "hclk_peri", 0, RK2928_CLKGATE_CON(5), 13, GFLAGS),
+	GATE(0, "hclk_usb_peri", "hclk_peri", CLK_IGNORE_UNUSED, RK2928_CLKGATE_CON(4), 5, GFLAGS),
+	GATE(HCLK_OTG0, "hclk_usbotg0", "hclk_peri", CLK_IGNORE_UNUSED, RK2928_CLKGATE_CON(5), 13, GFLAGS),
 	GATE(HCLK_HSADC, "hclk_hsadc", "hclk_peri", 0, RK2928_CLKGATE_CON(7), 5, GFLAGS),
 	GATE(HCLK_PIDF, "hclk_pidfilter", "hclk_peri", 0, RK2928_CLKGATE_CON(7), 6, GFLAGS),
 	GATE(HCLK_SDMMC, "hclk_sdmmc", "hclk_peri", 0, RK2928_CLKGATE_CON(5), 10, GFLAGS),
@@ -371,18 +477,18 @@ static struct rockchip_clk_branch common_clk_branches[] __initdata = {
 	GATE(0, "pclk_ddrupctl", "pclk_cpu", 0, RK2928_CLKGATE_CON(5), 7, GFLAGS),
 	GATE(0, "pclk_ddrpubl", "pclk_cpu", 0, RK2928_CLKGATE_CON(9), 6, GFLAGS),
 	GATE(0, "pclk_dbg", "pclk_cpu", 0, RK2928_CLKGATE_CON(9), 1, GFLAGS),
-	GATE(PCLK_GRF, "pclk_grf", "pclk_cpu", 0, RK2928_CLKGATE_CON(5), 4, GFLAGS),
-	GATE(PCLK_PMU, "pclk_pmu", "pclk_cpu", 0, RK2928_CLKGATE_CON(5), 5, GFLAGS),
+	GATE(PCLK_GRF, "pclk_grf", "pclk_cpu", CLK_IGNORE_UNUSED, RK2928_CLKGATE_CON(5), 4, GFLAGS),
+	GATE(PCLK_PMU, "pclk_pmu", "pclk_cpu", CLK_IGNORE_UNUSED, RK2928_CLKGATE_CON(5), 5, GFLAGS),
 
 	/* aclk_peri */
 	GATE(ACLK_DMA2, "aclk_dma2", "aclk_peri", 0, RK2928_CLKGATE_CON(5), 1, GFLAGS),
 	GATE(ACLK_SMC, "aclk_smc", "aclk_peri", 0, RK2928_CLKGATE_CON(5), 8, GFLAGS),
-	GATE(0, "aclk_peri_niu", "aclk_peri", 0, RK2928_CLKGATE_CON(4), 4, GFLAGS),
-	GATE(0, "aclk_cpu_peri", "aclk_peri", 0, RK2928_CLKGATE_CON(4), 2, GFLAGS),
-	GATE(0, "aclk_peri_axi_matrix", "aclk_peri", 0, RK2928_CLKGATE_CON(4), 3, GFLAGS),
+	GATE(0, "aclk_peri_niu", "aclk_peri", CLK_IGNORE_UNUSED, RK2928_CLKGATE_CON(4), 4, GFLAGS),
+	GATE(0, "aclk_cpu_peri", "aclk_peri", CLK_IGNORE_UNUSED, RK2928_CLKGATE_CON(4), 2, GFLAGS),
+	GATE(0, "aclk_peri_axi_matrix", "aclk_peri", CLK_IGNORE_UNUSED, RK2928_CLKGATE_CON(4), 3, GFLAGS),
 
 	/* pclk_peri gates */
-	GATE(0, "pclk_peri_axi_matrix", "pclk_peri", 0, RK2928_CLKGATE_CON(4), 1, GFLAGS),
+	GATE(0, "pclk_peri_axi_matrix", "pclk_peri", CLK_IGNORE_UNUSED, RK2928_CLKGATE_CON(4), 1, GFLAGS),
 	GATE(PCLK_PWM23, "pclk_pwm23", "pclk_peri", 0, RK2928_CLKGATE_CON(7), 11, GFLAGS),
 	GATE(PCLK_WDT, "pclk_wdt", "pclk_peri", 0, RK2928_CLKGATE_CON(7), 15, GFLAGS),
 	GATE(PCLK_SPI0, "pclk_spi0", "pclk_peri", 0, RK2928_CLKGATE_CON(7), 12, GFLAGS),
@@ -412,12 +518,20 @@ static struct clk_div_table div_aclk_cpu_t[] = {
 };
 
 static struct rockchip_clk_branch rk3066a_clk_branches[] __initdata = {
-	COMPOSITE_NOGATE(0, "armclk", mux_armclk_p, 0,
-			RK2928_CLKSEL_CON(0), 8, 1, MFLAGS, 0, 5, DFLAGS),
 	DIVTBL(0, "aclk_cpu_pre", "armclk", 0,
-			RK2928_CLKSEL_CON(1), 0, 3, DFLAGS, div_aclk_cpu_t),
+			RK2928_CLKSEL_CON(1), 0, 3, DFLAGS | CLK_DIVIDER_READ_ONLY, div_aclk_cpu_t),
+	DIV(0, "pclk_cpu_pre", "aclk_cpu_pre", 0,
+			RK2928_CLKSEL_CON(1), 12, 2, DFLAGS | CLK_DIVIDER_POWER_OF_TWO
+							    | CLK_DIVIDER_READ_ONLY),
+	DIV(0, "hclk_cpu_pre", "aclk_cpu_pre", 0,
+			RK2928_CLKSEL_CON(1), 8, 2, DFLAGS | CLK_DIVIDER_POWER_OF_TWO
+							   | CLK_DIVIDER_READ_ONLY),
+	COMPOSITE_NOMUX(0, "hclk_ahb2apb", "hclk_cpu_pre", 0,
+			RK2928_CLKSEL_CON(1), 14, 2, DFLAGS | CLK_DIVIDER_POWER_OF_TWO
+							    | CLK_DIVIDER_READ_ONLY,
+			RK2928_CLKGATE_CON(4), 9, GFLAGS),
 
-	GATE(CORE_L2C, "core_l2c", "aclk_cpu", 0,
+	GATE(CORE_L2C, "core_l2c", "aclk_cpu", CLK_IGNORE_UNUSED,
 			RK2928_CLKGATE_CON(9), 4, GFLAGS),
 
 	COMPOSITE(0, "aclk_peri_pre", mux_pll_src_gpll_cpll_p, 0,
@@ -483,21 +597,14 @@ static struct rockchip_clk_branch rk3066a_clk_branches[] __initdata = {
 			RK2928_CLKGATE_CON(0), 12, GFLAGS),
 	MUX(SCLK_I2S2, "sclk_i2s2", mux_sclk_i2s2_p, 0,
 			RK2928_CLKSEL_CON(4), 8, 2, MFLAGS),
-	COMPOSITE_NOMUX(0, "spdif_pre", "i2s_src", 0,
-			RK2928_CLKSEL_CON(5), 0, 7, DFLAGS,
-			RK2928_CLKGATE_CON(0), 13, GFLAGS),
-	COMPOSITE_FRAC(0, "spdif_frac", "spdif_pll", 0,
-			RK2928_CLKSEL_CON(9), 0,
-			RK2928_CLKGATE_CON(0), 14, GFLAGS),
-	MUX(SCLK_SPDIF, "sclk_spdif", mux_sclk_spdif_p, 0,
-			RK2928_CLKSEL_CON(5), 8, 2, MFLAGS),
 
 	GATE(HCLK_I2S1, "hclk_i2s1", "hclk_cpu", 0, RK2928_CLKGATE_CON(7), 3, GFLAGS),
 	GATE(HCLK_I2S2, "hclk_i2s2", "hclk_cpu", 0, RK2928_CLKGATE_CON(7), 4, GFLAGS),
 	GATE(0, "hclk_cif1", "hclk_cpu", 0, RK2928_CLKGATE_CON(6), 6, GFLAGS),
 	GATE(0, "hclk_hdmi", "hclk_cpu", 0, RK2928_CLKGATE_CON(4), 14, GFLAGS),
 
-	GATE(HCLK_OTG1, "hclk_usbotg1", "hclk_peri", 0, RK2928_CLKGATE_CON(5), 14, GFLAGS),
+	GATE(HCLK_OTG1, "hclk_usbotg1", "hclk_peri", CLK_IGNORE_UNUSED,
+			RK2928_CLKGATE_CON(5), 14, GFLAGS),
 
 	GATE(0, "aclk_cif1", "aclk_vio1", 0, RK2928_CLKGATE_CON(6), 7, GFLAGS),
 
@@ -524,17 +631,22 @@ PNAME(mux_hsicphy_p)		= { "sclk_otgphy0", "sclk_otgphy1",
 				    "gpll", "cpll" };
 
 static struct rockchip_clk_branch rk3188_clk_branches[] __initdata = {
-	COMPOSITE_NOGATE(0, "armclk", mux_armclk_p, 0,
-			RK2928_CLKSEL_CON(0), 8, 1, MFLAGS, 9, 5, DFLAGS),
-	COMPOSITE_NOMUX_DIVTBL(0, "aclk_core", "armclk", 0,
+	COMPOSITE_NOMUX_DIVTBL(0, "aclk_core", "armclk", CLK_IGNORE_UNUSED,
 			RK2928_CLKSEL_CON(1), 3, 3, DFLAGS | CLK_DIVIDER_READ_ONLY,
 			div_rk3188_aclk_core_t, RK2928_CLKGATE_CON(0), 7, GFLAGS),
 
 	/* do not source aclk_cpu_pre from the apll, to keep complexity down */
 	COMPOSITE_NOGATE(0, "aclk_cpu_pre", mux_aclk_cpu_p, CLK_SET_RATE_NO_REPARENT,
 			RK2928_CLKSEL_CON(0), 5, 1, MFLAGS, 0, 5, DFLAGS),
+	DIV(0, "pclk_cpu_pre", "aclk_cpu_pre", 0,
+			RK2928_CLKSEL_CON(1), 12, 2, DFLAGS | CLK_DIVIDER_POWER_OF_TWO),
+	DIV(0, "hclk_cpu_pre", "aclk_cpu_pre", 0,
+			RK2928_CLKSEL_CON(1), 8, 2, DFLAGS | CLK_DIVIDER_POWER_OF_TWO),
+	COMPOSITE_NOMUX(0, "hclk_ahb2apb", "hclk_cpu_pre", 0,
+			RK2928_CLKSEL_CON(1), 14, 2, DFLAGS | CLK_DIVIDER_POWER_OF_TWO,
+			RK2928_CLKGATE_CON(4), 9, GFLAGS),
 
-	GATE(CORE_L2C, "core_l2c", "armclk", 0,
+	GATE(CORE_L2C, "core_l2c", "armclk", CLK_IGNORE_UNUSED,
 			RK2928_CLKGATE_CON(9), 4, GFLAGS),
 
 	COMPOSITE(0, "aclk_peri_pre", mux_pll_src_cpll_gpll_p, 0,
@@ -564,7 +676,7 @@ static struct rockchip_clk_branch rk3188_clk_branches[] __initdata = {
 			RK2928_CLKSEL_CON(30), 0, 2, DFLAGS,
 			RK2928_CLKGATE_CON(3), 6, GFLAGS),
 	DIV(0, "sclk_hsicphy_12m", "sclk_hsicphy_480m", 0,
-			RK2928_CLKGATE_CON(11), 8, 6, DFLAGS),
+			RK2928_CLKSEL_CON(11), 8, 6, DFLAGS),
 
 	MUX(0, "i2s_src", mux_pll_src_gpll_cpll_p, 0,
 			RK2928_CLKSEL_CON(2), 15, 1, MFLAGS),
@@ -576,19 +688,12 @@ static struct rockchip_clk_branch rk3188_clk_branches[] __initdata = {
 			RK2928_CLKGATE_CON(0), 10, GFLAGS),
 	MUX(SCLK_I2S0, "sclk_i2s0", mux_sclk_i2s0_p, 0,
 			RK2928_CLKSEL_CON(3), 8, 2, MFLAGS),
-	COMPOSITE_NOMUX(0, "spdif_pre", "i2s_src", 0,
-			RK2928_CLKSEL_CON(5), 0, 7, DFLAGS,
-			RK2928_CLKGATE_CON(13), 13, GFLAGS),
-	COMPOSITE_FRAC(0, "spdif_frac", "spdif_pll", 0,
-			RK2928_CLKSEL_CON(9), 0,
-			RK2928_CLKGATE_CON(0), 14, GFLAGS),
-	MUX(SCLK_SPDIF, "sclk_spdif", mux_sclk_spdif_p, 0,
-			RK2928_CLKSEL_CON(5), 8, 2, MFLAGS),
 
 	GATE(0, "hclk_imem0", "hclk_cpu", 0, RK2928_CLKGATE_CON(4), 14, GFLAGS),
 	GATE(0, "hclk_imem1", "hclk_cpu", 0, RK2928_CLKGATE_CON(4), 15, GFLAGS),
 
-	GATE(HCLK_OTG1, "hclk_usbotg1", "hclk_peri", 0, RK2928_CLKGATE_CON(7), 3, GFLAGS),
+	GATE(HCLK_OTG1, "hclk_usbotg1", "hclk_peri", CLK_IGNORE_UNUSED,
+			RK2928_CLKGATE_CON(7), 3, GFLAGS),
 	GATE(HCLK_HSIC, "hclk_hsic", "hclk_peri", 0, RK2928_CLKGATE_CON(7), 4, GFLAGS),
 
 	GATE(PCLK_TIMER3, "pclk_timer3", "pclk_cpu", 0, RK2928_CLKGATE_CON(7), 9, GFLAGS),
@@ -599,6 +704,12 @@ static struct rockchip_clk_branch rk3188_clk_branches[] __initdata = {
 	GATE(ACLK_GPS, "aclk_gps", "aclk_peri", 0, RK2928_CLKGATE_CON(8), 13, GFLAGS),
 };
 
+static const char *rk3188_critical_clocks[] __initconst = {
+	"aclk_cpu",
+	"aclk_peri",
+	"hclk_peri",
+};
+
 static void __init rk3188_common_clk_init(struct device_node *np)
 {
 	void __iomem *reg_base;
@@ -623,29 +734,65 @@ static void __init rk3188_common_clk_init(struct device_node *np)
 		pr_warn("%s: could not register clock usb480m: %ld\n",
 			__func__, PTR_ERR(clk));
 
-	rockchip_clk_register_plls(rk3188_pll_clks,
-				   ARRAY_SIZE(rk3188_pll_clks),
-				   RK3188_GRF_SOC_STATUS);
 	rockchip_clk_register_branches(common_clk_branches,
 				  ARRAY_SIZE(common_clk_branches));
+	rockchip_clk_protect_critical(rk3188_critical_clocks,
+				      ARRAY_SIZE(rk3188_critical_clocks));
 
 	rockchip_register_softrst(np, 9, reg_base + RK2928_SOFTRST_CON(0),
 				  ROCKCHIP_SOFTRST_HIWORD_MASK);
+
+	rockchip_register_restart_notifier(RK2928_GLB_SRST_FST);
 }
 
 static void __init rk3066a_clk_init(struct device_node *np)
 {
 	rk3188_common_clk_init(np);
+	rockchip_clk_register_plls(rk3066_pll_clks,
+				   ARRAY_SIZE(rk3066_pll_clks),
+				   RK3066_GRF_SOC_STATUS);
 	rockchip_clk_register_branches(rk3066a_clk_branches,
 				  ARRAY_SIZE(rk3066a_clk_branches));
+	rockchip_clk_register_armclk(ARMCLK, "armclk",
+			mux_armclk_p, ARRAY_SIZE(mux_armclk_p),
+			&rk3066_cpuclk_data, rk3066_cpuclk_rates,
+			ARRAY_SIZE(rk3066_cpuclk_rates));
 }
 CLK_OF_DECLARE(rk3066a_cru, "rockchip,rk3066a-cru", rk3066a_clk_init);
 
 static void __init rk3188a_clk_init(struct device_node *np)
 {
+	struct clk *clk1, *clk2;
+	unsigned long rate;
+	int ret;
+
 	rk3188_common_clk_init(np);
+	rockchip_clk_register_plls(rk3188_pll_clks,
+				   ARRAY_SIZE(rk3188_pll_clks),
+				   RK3188_GRF_SOC_STATUS);
 	rockchip_clk_register_branches(rk3188_clk_branches,
 				  ARRAY_SIZE(rk3188_clk_branches));
+	rockchip_clk_register_armclk(ARMCLK, "armclk",
+				  mux_armclk_p, ARRAY_SIZE(mux_armclk_p),
+				  &rk3188_cpuclk_data, rk3188_cpuclk_rates,
+				  ARRAY_SIZE(rk3188_cpuclk_rates));
+
+	/* reparent aclk_cpu_pre from apll */
+	clk1 = __clk_lookup("aclk_cpu_pre");
+	clk2 = __clk_lookup("gpll");
+	if (clk1 && clk2) {
+		rate = clk_get_rate(clk1);
+
+		ret = clk_set_parent(clk1, clk2);
+		if (ret < 0)
+			pr_warn("%s: could not reparent aclk_cpu_pre to gpll\n",
+				__func__);
+
+		clk_set_rate(clk1, rate);
+	} else {
+		pr_warn("%s: missing clocks to reparent aclk_cpu_pre to gpll\n",
+			__func__);
+	}
 }
 CLK_OF_DECLARE(rk3188a_cru, "rockchip,rk3188a-cru", rk3188a_clk_init);
 
diff --git a/drivers/clk/rockchip/clk-rk3288.c b/drivers/clk/rockchip/clk-rk3288.c
index b22a2d2f21e9..11194b8329fe 100644
--- a/drivers/clk/rockchip/clk-rk3288.c
+++ b/drivers/clk/rockchip/clk-rk3288.c
@@ -16,11 +16,12 @@
 #include <linux/clk-provider.h>
 #include <linux/of.h>
 #include <linux/of_address.h>
+#include <linux/syscore_ops.h>
 #include <dt-bindings/clock/rk3288-cru.h>
 #include "clk.h"
 
 #define RK3288_GRF_SOC_CON(x)	(0x244 + x * 4)
-#define RK3288_GRF_SOC_STATUS	0x280
+#define RK3288_GRF_SOC_STATUS1	0x284
 
 enum rk3288_plls {
 	apll, dpll, cpll, gpll, npll,
@@ -83,11 +84,13 @@ struct rockchip_pll_rate_table rk3288_pll_rates[] = {
 	RK3066_PLL_RATE( 742500000, 8, 495, 2),
 	RK3066_PLL_RATE( 696000000, 1, 58, 2),
 	RK3066_PLL_RATE( 600000000, 1, 50, 2),
-	RK3066_PLL_RATE( 594000000, 2, 198, 4),
+	RK3066_PLL_RATE_BWADJ(594000000, 1, 198, 8, 1),
 	RK3066_PLL_RATE( 552000000, 1, 46, 2),
 	RK3066_PLL_RATE( 504000000, 1, 84, 4),
+	RK3066_PLL_RATE( 500000000, 3, 125, 2),
 	RK3066_PLL_RATE( 456000000, 1, 76, 4),
 	RK3066_PLL_RATE( 408000000, 1, 68, 4),
+	RK3066_PLL_RATE( 400000000, 3, 100, 2),
 	RK3066_PLL_RATE( 384000000, 2, 128, 4),
 	RK3066_PLL_RATE( 360000000, 1, 60, 4),
 	RK3066_PLL_RATE( 312000000, 1, 52, 4),
@@ -101,6 +104,70 @@ struct rockchip_pll_rate_table rk3288_pll_rates[] = {
 	{ /* sentinel */ },
 };
 
+#define RK3288_DIV_ACLK_CORE_M0_MASK	0xf
+#define RK3288_DIV_ACLK_CORE_M0_SHIFT	0
+#define RK3288_DIV_ACLK_CORE_MP_MASK	0xf
+#define RK3288_DIV_ACLK_CORE_MP_SHIFT	4
+#define RK3288_DIV_L2RAM_MASK		0x7
+#define RK3288_DIV_L2RAM_SHIFT		0
+#define RK3288_DIV_ATCLK_MASK		0x1f
+#define RK3288_DIV_ATCLK_SHIFT		4
+#define RK3288_DIV_PCLK_DBGPRE_MASK	0x1f
+#define RK3288_DIV_PCLK_DBGPRE_SHIFT	9
+
+#define RK3288_CLKSEL0(_core_m0, _core_mp)				\
+	{								\
+		.reg = RK3288_CLKSEL_CON(0),				\
+		.val = HIWORD_UPDATE(_core_m0, RK3288_DIV_ACLK_CORE_M0_MASK, \
+				RK3288_DIV_ACLK_CORE_M0_SHIFT) |	\
+		       HIWORD_UPDATE(_core_mp, RK3288_DIV_ACLK_CORE_MP_MASK, \
+				RK3288_DIV_ACLK_CORE_MP_SHIFT),		\
+	}
+#define RK3288_CLKSEL37(_l2ram, _atclk, _pclk_dbg_pre)			\
+	{								\
+		.reg = RK3288_CLKSEL_CON(37),				\
+		.val = HIWORD_UPDATE(_l2ram, RK3288_DIV_L2RAM_MASK,	\
+				RK3288_DIV_L2RAM_SHIFT) |		\
+		       HIWORD_UPDATE(_atclk, RK3288_DIV_ATCLK_MASK,	\
+				RK3288_DIV_ATCLK_SHIFT) |		\
+		       HIWORD_UPDATE(_pclk_dbg_pre,			\
+				RK3288_DIV_PCLK_DBGPRE_MASK,		\
+				RK3288_DIV_PCLK_DBGPRE_SHIFT),		\
+	}
+
+#define RK3288_CPUCLK_RATE(_prate, _core_m0, _core_mp, _l2ram, _atclk, _pdbg) \
+	{								\
+		.prate = _prate,					\
+		.divs = {						\
+			RK3288_CLKSEL0(_core_m0, _core_mp),		\
+			RK3288_CLKSEL37(_l2ram, _atclk, _pdbg),		\
+		},							\
+	}
+
+static struct rockchip_cpuclk_rate_table rk3288_cpuclk_rates[] __initdata = {
+	RK3288_CPUCLK_RATE(1800000000, 1, 3, 1, 3, 3),
+	RK3288_CPUCLK_RATE(1704000000, 1, 3, 1, 3, 3),
+	RK3288_CPUCLK_RATE(1608000000, 1, 3, 1, 3, 3),
+	RK3288_CPUCLK_RATE(1512000000, 1, 3, 1, 3, 3),
+	RK3288_CPUCLK_RATE(1416000000, 1, 3, 1, 3, 3),
+	RK3288_CPUCLK_RATE(1200000000, 1, 3, 1, 3, 3),
+	RK3288_CPUCLK_RATE(1008000000, 1, 3, 1, 3, 3),
+	RK3288_CPUCLK_RATE( 816000000, 1, 3, 1, 3, 3),
+	RK3288_CPUCLK_RATE( 696000000, 1, 3, 1, 3, 3),
+	RK3288_CPUCLK_RATE( 600000000, 1, 3, 1, 3, 3),
+	RK3288_CPUCLK_RATE( 408000000, 1, 3, 1, 3, 3),
+	RK3288_CPUCLK_RATE( 312000000, 1, 3, 1, 3, 3),
+	RK3288_CPUCLK_RATE( 216000000, 1, 3, 1, 3, 3),
+	RK3288_CPUCLK_RATE( 126000000, 1, 3, 1, 3, 3),
+};
+
+static const struct rockchip_cpuclk_reg_data rk3288_cpuclk_data = {
+	.core_reg = RK3288_CLKSEL_CON(0),
+	.div_core_shift = 8,
+	.div_core_mask = 0x1f,
+	.mux_core_shift = 15,
+};
+
 PNAME(mux_pll_p)		= { "xin24m", "xin32k" };
 PNAME(mux_armclk_p)		= { "apll_core", "gpll_core" };
 PNAME(mux_ddrphy_p)		= { "dpll_ddr", "gpll_ddr" };
@@ -109,14 +176,14 @@ PNAME(mux_aclk_cpu_src_p)	= { "cpll_aclk_cpu", "gpll_aclk_cpu" };
 PNAME(mux_pll_src_cpll_gpll_p)		= { "cpll", "gpll" };
 PNAME(mux_pll_src_npll_cpll_gpll_p)	= { "npll", "cpll", "gpll" };
 PNAME(mux_pll_src_cpll_gpll_npll_p)	= { "cpll", "gpll", "npll" };
-PNAME(mux_pll_src_cpll_gpll_usb480m_p)	= { "cpll", "gpll", "usb480m" };
+PNAME(mux_pll_src_cpll_gpll_usb480m_p)	= { "cpll", "gpll", "usbphy480m_src" };
+PNAME(mux_pll_src_cpll_gll_usb_npll_p)	= { "cpll", "gpll", "usbphy480m_src", "npll" };
 
 PNAME(mux_mmc_src_p)	= { "cpll", "gpll", "xin24m", "xin24m" };
 PNAME(mux_i2s_pre_p)	= { "i2s_src", "i2s_frac", "ext_i2s", "xin12m" };
 PNAME(mux_i2s_clkout_p)	= { "i2s_pre", "xin12m" };
 PNAME(mux_spdif_p)	= { "spdif_pre", "spdif_frac", "xin12m" };
 PNAME(mux_spdif_8ch_p)	= { "spdif_8ch_pre", "spdif_8ch_frac", "xin12m" };
-PNAME(mux_uart0_pll_p)	= { "cpll", "gpll", "usbphy_480m_src", "npll" };
 PNAME(mux_uart0_p)	= { "uart0_src", "uart0_frac", "xin24m" };
 PNAME(mux_uart1_p)	= { "uart1_src", "uart1_frac", "xin24m" };
 PNAME(mux_uart2_p)	= { "uart2_src", "uart2_frac", "xin24m" };
@@ -128,22 +195,22 @@ PNAME(mux_hsadcout_p)	= { "hsadc_src", "ext_hsadc" };
 PNAME(mux_edp_24m_p)	= { "ext_edp_24m", "xin24m" };
 PNAME(mux_tspout_p)	= { "cpll", "gpll", "npll", "xin27m" };
 
-PNAME(mux_usbphy480m_p)		= { "sclk_otgphy0", "sclk_otgphy1",
-				    "sclk_otgphy2" };
+PNAME(mux_usbphy480m_p)		= { "sclk_otgphy1", "sclk_otgphy2",
+				    "sclk_otgphy0" };
 PNAME(mux_hsicphy480m_p)	= { "cpll", "gpll", "usbphy480m_src" };
 PNAME(mux_hsicphy12m_p)		= { "hsicphy12m_xin12m", "hsicphy12m_usbphy" };
 
 static struct rockchip_pll_clock rk3288_pll_clks[] __initdata = {
 	[apll] = PLL(pll_rk3066, PLL_APLL, "apll", mux_pll_p, 0, RK3288_PLL_CON(0),
-		     RK3288_MODE_CON, 0, 6, rk3288_pll_rates),
+		     RK3288_MODE_CON, 0, 6, 0, rk3288_pll_rates),
 	[dpll] = PLL(pll_rk3066, PLL_DPLL, "dpll", mux_pll_p, 0, RK3288_PLL_CON(4),
-		     RK3288_MODE_CON, 4, 5, NULL),
+		     RK3288_MODE_CON, 4, 5, 0, NULL),
 	[cpll] = PLL(pll_rk3066, PLL_CPLL, "cpll", mux_pll_p, 0, RK3288_PLL_CON(8),
-		     RK3288_MODE_CON, 8, 7, rk3288_pll_rates),
+		     RK3288_MODE_CON, 8, 7, ROCKCHIP_PLL_SYNC_RATE, rk3288_pll_rates),
 	[gpll] = PLL(pll_rk3066, PLL_GPLL, "gpll", mux_pll_p, 0, RK3288_PLL_CON(12),
-		     RK3288_MODE_CON, 12, 8, rk3288_pll_rates),
+		     RK3288_MODE_CON, 12, 8, ROCKCHIP_PLL_SYNC_RATE, rk3288_pll_rates),
 	[npll] = PLL(pll_rk3066, PLL_NPLL, "npll",  mux_pll_p, 0, RK3288_PLL_CON(16),
-		     RK3288_MODE_CON, 14, 9, NULL),
+		     RK3288_MODE_CON, 14, 9, ROCKCHIP_PLL_SYNC_RATE, rk3288_pll_rates),
 };
 
 static struct clk_div_table div_hclk_cpu_t[] = {
@@ -162,69 +229,67 @@ static struct rockchip_clk_branch rk3288_clk_branches[] __initdata = {
 	 * Clock-Architecture Diagram 1
 	 */
 
-	GATE(0, "apll_core", "apll", 0,
+	GATE(0, "apll_core", "apll", CLK_IGNORE_UNUSED,
 			RK3288_CLKGATE_CON(0), 1, GFLAGS),
-	GATE(0, "gpll_core", "gpll", 0,
+	GATE(0, "gpll_core", "gpll", CLK_IGNORE_UNUSED,
 			RK3288_CLKGATE_CON(0), 2, GFLAGS),
-	COMPOSITE_NOGATE(0, "armclk", mux_armclk_p, 0,
-			RK3288_CLKSEL_CON(0), 15, 1, MFLAGS, 8, 5, DFLAGS),
 
-	COMPOSITE_NOMUX(0, "armcore0", "armclk", 0,
-			RK3288_CLKSEL_CON(36), 0, 3, DFLAGS,
+	COMPOSITE_NOMUX(0, "armcore0", "armclk", CLK_IGNORE_UNUSED,
+			RK3288_CLKSEL_CON(36), 0, 3, DFLAGS | CLK_DIVIDER_READ_ONLY,
 			RK3288_CLKGATE_CON(12), 0, GFLAGS),
-	COMPOSITE_NOMUX(0, "armcore1", "armclk", 0,
-			RK3288_CLKSEL_CON(36), 4, 3, DFLAGS,
+	COMPOSITE_NOMUX(0, "armcore1", "armclk", CLK_IGNORE_UNUSED,
+			RK3288_CLKSEL_CON(36), 4, 3, DFLAGS | CLK_DIVIDER_READ_ONLY,
 			RK3288_CLKGATE_CON(12), 1, GFLAGS),
-	COMPOSITE_NOMUX(0, "armcore2", "armclk", 0,
-			RK3288_CLKSEL_CON(36), 8, 3, DFLAGS,
+	COMPOSITE_NOMUX(0, "armcore2", "armclk", CLK_IGNORE_UNUSED,
+			RK3288_CLKSEL_CON(36), 8, 3, DFLAGS | CLK_DIVIDER_READ_ONLY,
 			RK3288_CLKGATE_CON(12), 2, GFLAGS),
-	COMPOSITE_NOMUX(0, "armcore3", "armclk", 0,
-			RK3288_CLKSEL_CON(36), 12, 3, DFLAGS,
+	COMPOSITE_NOMUX(0, "armcore3", "armclk", CLK_IGNORE_UNUSED,
+			RK3288_CLKSEL_CON(36), 12, 3, DFLAGS | CLK_DIVIDER_READ_ONLY,
 			RK3288_CLKGATE_CON(12), 3, GFLAGS),
-	COMPOSITE_NOMUX(0, "l2ram", "armclk", 0,
-			RK3288_CLKSEL_CON(37), 0, 3, DFLAGS,
+	COMPOSITE_NOMUX(0, "l2ram", "armclk", CLK_IGNORE_UNUSED,
+			RK3288_CLKSEL_CON(37), 0, 3, DFLAGS | CLK_DIVIDER_READ_ONLY,
 			RK3288_CLKGATE_CON(12), 4, GFLAGS),
-	COMPOSITE_NOMUX(0, "aclk_core_m0", "armclk", 0,
-			RK3288_CLKSEL_CON(0), 0, 4, DFLAGS,
+	COMPOSITE_NOMUX(0, "aclk_core_m0", "armclk", CLK_IGNORE_UNUSED,
+			RK3288_CLKSEL_CON(0), 0, 4, DFLAGS | CLK_DIVIDER_READ_ONLY,
 			RK3288_CLKGATE_CON(12), 5, GFLAGS),
-	COMPOSITE_NOMUX(0, "aclk_core_mp", "armclk", 0,
-			RK3288_CLKSEL_CON(0), 4, 4, DFLAGS,
+	COMPOSITE_NOMUX(0, "aclk_core_mp", "armclk", CLK_IGNORE_UNUSED,
+			RK3288_CLKSEL_CON(0), 4, 4, DFLAGS | CLK_DIVIDER_READ_ONLY,
 			RK3288_CLKGATE_CON(12), 6, GFLAGS),
 	COMPOSITE_NOMUX(0, "atclk", "armclk", 0,
-			RK3288_CLKSEL_CON(37), 4, 5, DFLAGS,
+			RK3288_CLKSEL_CON(37), 4, 5, DFLAGS | CLK_DIVIDER_READ_ONLY,
 			RK3288_CLKGATE_CON(12), 7, GFLAGS),
-	COMPOSITE_NOMUX(0, "pclk_dbg_pre", "armclk", 0,
-			RK3288_CLKSEL_CON(37), 9, 5, DFLAGS,
+	COMPOSITE_NOMUX(0, "pclk_dbg_pre", "armclk", CLK_IGNORE_UNUSED,
+			RK3288_CLKSEL_CON(37), 9, 5, DFLAGS | CLK_DIVIDER_READ_ONLY,
 			RK3288_CLKGATE_CON(12), 8, GFLAGS),
 	GATE(0, "pclk_dbg", "pclk_dbg_pre", 0,
 			RK3288_CLKGATE_CON(12), 9, GFLAGS),
-	GATE(0, "cs_dbg", "pclk_dbg_pre", 0,
+	GATE(0, "cs_dbg", "pclk_dbg_pre", CLK_IGNORE_UNUSED,
 			RK3288_CLKGATE_CON(12), 10, GFLAGS),
 	GATE(0, "pclk_core_niu", "pclk_dbg_pre", 0,
 			RK3288_CLKGATE_CON(12), 11, GFLAGS),
 
-	GATE(0, "dpll_ddr", "dpll", 0,
+	GATE(0, "dpll_ddr", "dpll", CLK_IGNORE_UNUSED,
 			RK3288_CLKGATE_CON(0), 8, GFLAGS),
 	GATE(0, "gpll_ddr", "gpll", 0,
 			RK3288_CLKGATE_CON(0), 9, GFLAGS),
-	COMPOSITE_NOGATE(0, "ddrphy", mux_ddrphy_p, 0,
+	COMPOSITE_NOGATE(0, "ddrphy", mux_ddrphy_p, CLK_IGNORE_UNUSED,
 			RK3288_CLKSEL_CON(26), 2, 1, MFLAGS, 0, 2,
 					DFLAGS | CLK_DIVIDER_POWER_OF_TWO),
 
-	GATE(0, "gpll_aclk_cpu", "gpll", 0,
+	GATE(0, "gpll_aclk_cpu", "gpll", CLK_IGNORE_UNUSED,
 			RK3288_CLKGATE_CON(0), 10, GFLAGS),
-	GATE(0, "cpll_aclk_cpu", "cpll", 0,
+	GATE(0, "cpll_aclk_cpu", "cpll", CLK_IGNORE_UNUSED,
 			RK3288_CLKGATE_CON(0), 11, GFLAGS),
-	COMPOSITE_NOGATE(0, "aclk_cpu_src", mux_aclk_cpu_src_p, 0,
+	COMPOSITE_NOGATE(0, "aclk_cpu_src", mux_aclk_cpu_src_p, CLK_IGNORE_UNUSED,
 			RK3288_CLKSEL_CON(1), 15, 1, MFLAGS, 3, 5, DFLAGS),
-	DIV(0, "aclk_cpu_pre", "aclk_cpu_src", 0,
+	DIV(0, "aclk_cpu_pre", "aclk_cpu_src", CLK_SET_RATE_PARENT,
 			RK3288_CLKSEL_CON(1), 0, 3, DFLAGS),
-	GATE(0, "aclk_cpu", "aclk_cpu_pre", 0,
+	GATE(ACLK_CPU, "aclk_cpu", "aclk_cpu_pre", CLK_IGNORE_UNUSED,
 			RK3288_CLKGATE_CON(0), 3, GFLAGS),
-	COMPOSITE_NOMUX(0, "pclk_cpu", "aclk_cpu_pre", 0,
+	COMPOSITE_NOMUX(PCLK_CPU, "pclk_cpu", "aclk_cpu_pre", CLK_IGNORE_UNUSED,
 			RK3288_CLKSEL_CON(1), 12, 3, DFLAGS,
 			RK3288_CLKGATE_CON(0), 5, GFLAGS),
-	COMPOSITE_NOMUX_DIVTBL(0, "hclk_cpu", "aclk_cpu_pre", 0,
+	COMPOSITE_NOMUX_DIVTBL(HCLK_CPU, "hclk_cpu", "aclk_cpu_pre", CLK_IGNORE_UNUSED,
 			RK3288_CLKSEL_CON(1), 8, 2, DFLAGS, div_hclk_cpu_t,
 			RK3288_CLKGATE_CON(0), 4, GFLAGS),
 	GATE(0, "c2c_host", "aclk_cpu_src", 0,
@@ -232,21 +297,21 @@ static struct rockchip_clk_branch rk3288_clk_branches[] __initdata = {
 	COMPOSITE_NOMUX(0, "crypto", "aclk_cpu_pre", 0,
 			RK3288_CLKSEL_CON(26), 6, 2, DFLAGS,
 			RK3288_CLKGATE_CON(5), 4, GFLAGS),
-	GATE(0, "aclk_bus_2pmu", "aclk_cpu_pre", 0,
+	GATE(0, "aclk_bus_2pmu", "aclk_cpu_pre", CLK_IGNORE_UNUSED,
 			RK3288_CLKGATE_CON(0), 7, GFLAGS),
 
 	COMPOSITE(0, "i2s_src", mux_pll_src_cpll_gpll_p, 0,
 			RK3288_CLKSEL_CON(4), 15, 1, MFLAGS, 0, 7, DFLAGS,
 			RK3288_CLKGATE_CON(4), 1, GFLAGS),
-	COMPOSITE_FRAC(0, "i2s_frac", "i2s_src", 0,
+	COMPOSITE_FRAC(0, "i2s_frac", "i2s_src", CLK_SET_RATE_PARENT,
 			RK3288_CLKSEL_CON(8), 0,
 			RK3288_CLKGATE_CON(4), 2, GFLAGS),
-	MUX(0, "i2s_pre", mux_i2s_pre_p, 0,
+	MUX(0, "i2s_pre", mux_i2s_pre_p, CLK_SET_RATE_PARENT,
 			RK3288_CLKSEL_CON(4), 8, 2, MFLAGS),
-	COMPOSITE_NODIV(0, "i2s0_clkout", mux_i2s_clkout_p, 0,
+	COMPOSITE_NODIV(SCLK_I2S0_OUT, "i2s0_clkout", mux_i2s_clkout_p, 0,
 			RK3288_CLKSEL_CON(4), 12, 1, MFLAGS,
 			RK3288_CLKGATE_CON(4), 0, GFLAGS),
-	GATE(SCLK_I2S0, "sclk_i2s0", "i2s_pre", 0,
+	GATE(SCLK_I2S0, "sclk_i2s0", "i2s_pre", CLK_SET_RATE_PARENT,
 			RK3288_CLKGATE_CON(4), 3, GFLAGS),
 
 	MUX(0, "spdif_src", mux_pll_src_cpll_gpll_p, 0,
@@ -263,7 +328,7 @@ static struct rockchip_clk_branch rk3288_clk_branches[] __initdata = {
 	COMPOSITE_NOMUX(0, "spdif_8ch_pre", "spdif_src", 0,
 			RK3288_CLKSEL_CON(40), 0, 7, DFLAGS,
 			RK3288_CLKGATE_CON(4), 7, GFLAGS),
-	COMPOSITE_FRAC(0, "spdif_8ch_frac", "spdif_8ch_src", 0,
+	COMPOSITE_FRAC(0, "spdif_8ch_frac", "spdif_8ch_pre", 0,
 			RK3288_CLKSEL_CON(41), 0,
 			RK3288_CLKGATE_CON(4), 8, GFLAGS),
 	COMPOSITE_NODIV(SCLK_SPDIF8CH, "sclk_spdif_8ch", mux_spdif_8ch_p, 0,
@@ -296,20 +361,34 @@ static struct rockchip_clk_branch rk3288_clk_branches[] __initdata = {
 	COMPOSITE(0, "aclk_vdpu", mux_pll_src_cpll_gpll_usb480m_p, 0,
 			RK3288_CLKSEL_CON(32), 14, 2, MFLAGS, 8, 5, DFLAGS,
 			RK3288_CLKGATE_CON(3), 11, GFLAGS),
+	/*
+	 * We use aclk_vdpu by default GRF_SOC_CON0[7] setting in system,
+	 * so we ignore the mux and make clocks nodes as following,
+	 */
+	GATE(ACLK_VCODEC, "aclk_vcodec", "aclk_vdpu", 0,
+		RK3288_CLKGATE_CON(9), 0, GFLAGS),
+	/*
+	 * We introduce a virtul node of hclk_vodec_pre_v to split one clock
+	 * struct with a gate and a fix divider into two node in software.
+	 */
+	GATE(0, "hclk_vcodec_pre_v", "aclk_vdpu", 0,
+		RK3288_CLKGATE_CON(3), 10, GFLAGS),
+	GATE(HCLK_VCODEC, "hclk_vcodec", "hclk_vcodec_pre", 0,
+		RK3288_CLKGATE_CON(9), 1, GFLAGS),
 
-	COMPOSITE(0, "aclk_vio0", mux_pll_src_cpll_gpll_usb480m_p, 0,
+	COMPOSITE(0, "aclk_vio0", mux_pll_src_cpll_gpll_usb480m_p, CLK_IGNORE_UNUSED,
 			RK3288_CLKSEL_CON(31), 6, 2, MFLAGS, 0, 5, DFLAGS,
 			RK3288_CLKGATE_CON(3), 0, GFLAGS),
 	DIV(0, "hclk_vio", "aclk_vio0", 0,
 			RK3288_CLKSEL_CON(28), 8, 5, DFLAGS),
-	COMPOSITE(0, "aclk_vio1", mux_pll_src_cpll_gpll_usb480m_p, 0,
+	COMPOSITE(0, "aclk_vio1", mux_pll_src_cpll_gpll_usb480m_p, CLK_IGNORE_UNUSED,
 			RK3288_CLKSEL_CON(31), 14, 2, MFLAGS, 8, 5, DFLAGS,
 			RK3288_CLKGATE_CON(3), 2, GFLAGS),
 
 	COMPOSITE(0, "aclk_rga_pre", mux_pll_src_cpll_gpll_usb480m_p, 0,
 			RK3288_CLKSEL_CON(30), 6, 2, MFLAGS, 0, 5, DFLAGS,
 			RK3288_CLKGATE_CON(3), 5, GFLAGS),
-	COMPOSITE(0, "sclk_rga", mux_pll_src_cpll_gpll_usb480m_p, 0,
+	COMPOSITE(SCLK_RGA, "sclk_rga", mux_pll_src_cpll_gpll_usb480m_p, 0,
 			RK3288_CLKSEL_CON(30), 14, 2, MFLAGS, 8, 5, DFLAGS,
 			RK3288_CLKGATE_CON(3), 4, GFLAGS),
 
@@ -320,35 +399,35 @@ static struct rockchip_clk_branch rk3288_clk_branches[] __initdata = {
 			RK3288_CLKSEL_CON(29), 6, 2, MFLAGS, 8, 8, DFLAGS,
 			RK3288_CLKGATE_CON(3), 3, GFLAGS),
 
-	COMPOSITE_NODIV(0, "sclk_edp_24m", mux_edp_24m_p, 0,
+	COMPOSITE_NODIV(SCLK_EDP_24M, "sclk_edp_24m", mux_edp_24m_p, 0,
 			RK3288_CLKSEL_CON(28), 15, 1, MFLAGS,
 			RK3288_CLKGATE_CON(3), 12, GFLAGS),
-	COMPOSITE(0, "sclk_edp", mux_pll_src_cpll_gpll_npll_p, 0,
+	COMPOSITE(SCLK_EDP, "sclk_edp", mux_pll_src_cpll_gpll_npll_p, 0,
 			RK3288_CLKSEL_CON(28), 6, 2, MFLAGS, 0, 6, DFLAGS,
 			RK3288_CLKGATE_CON(3), 13, GFLAGS),
 
-	COMPOSITE(0, "sclk_isp", mux_pll_src_cpll_gpll_npll_p, 0,
+	COMPOSITE(SCLK_ISP, "sclk_isp", mux_pll_src_cpll_gpll_npll_p, 0,
 			RK3288_CLKSEL_CON(6), 6, 2, MFLAGS, 0, 6, DFLAGS,
 			RK3288_CLKGATE_CON(3), 14, GFLAGS),
-	COMPOSITE(0, "sclk_isp_jpe", mux_pll_src_cpll_gpll_npll_p, 0,
+	COMPOSITE(SCLK_ISP_JPE, "sclk_isp_jpe", mux_pll_src_cpll_gpll_npll_p, 0,
 			RK3288_CLKSEL_CON(6), 14, 2, MFLAGS, 8, 6, DFLAGS,
 			RK3288_CLKGATE_CON(3), 15, GFLAGS),
 
-	GATE(0, "sclk_hdmi_hdcp", "xin24m", 0,
+	GATE(SCLK_HDMI_HDCP, "sclk_hdmi_hdcp", "xin24m", 0,
 			RK3288_CLKGATE_CON(5), 12, GFLAGS),
-	GATE(0, "sclk_hdmi_cec", "xin32k", 0,
+	GATE(SCLK_HDMI_CEC, "sclk_hdmi_cec", "xin32k", 0,
 			RK3288_CLKGATE_CON(5), 11, GFLAGS),
 
-	COMPOSITE(0, "aclk_hevc", mux_pll_src_cpll_gpll_npll_p, 0,
+	COMPOSITE(ACLK_HEVC, "aclk_hevc", mux_pll_src_cpll_gpll_npll_p, 0,
 			RK3288_CLKSEL_CON(39), 14, 2, MFLAGS, 8, 5, DFLAGS,
 			RK3288_CLKGATE_CON(13), 13, GFLAGS),
-	DIV(0, "hclk_hevc", "aclk_hevc", 0,
+	DIV(HCLK_HEVC, "hclk_hevc", "aclk_hevc", 0,
 			RK3288_CLKSEL_CON(40), 12, 2, DFLAGS),
 
-	COMPOSITE(0, "sclk_hevc_cabac", mux_pll_src_cpll_gpll_npll_p, 0,
+	COMPOSITE(SCLK_HEVC_CABAC, "sclk_hevc_cabac", mux_pll_src_cpll_gpll_npll_p, 0,
 			RK3288_CLKSEL_CON(42), 6, 2, MFLAGS, 0, 5, DFLAGS,
 			RK3288_CLKGATE_CON(13), 14, GFLAGS),
-	COMPOSITE(0, "sclk_hevc_core", mux_pll_src_cpll_gpll_npll_p, 0,
+	COMPOSITE(SCLK_HEVC_CORE, "sclk_hevc_core", mux_pll_src_cpll_gpll_npll_p, 0,
 			RK3288_CLKSEL_CON(42), 14, 2, MFLAGS, 8, 5, DFLAGS,
 			RK3288_CLKGATE_CON(13), 15, GFLAGS),
 
@@ -360,24 +439,24 @@ static struct rockchip_clk_branch rk3288_clk_branches[] __initdata = {
 
 	DIV(0, "pclk_pd_alive", "gpll", 0,
 			RK3288_CLKSEL_CON(33), 8, 5, DFLAGS),
-	COMPOSITE_NOMUX(0, "pclk_pd_pmu", "gpll", 0,
+	COMPOSITE_NOMUX(0, "pclk_pd_pmu", "gpll", CLK_IGNORE_UNUSED,
 			RK3288_CLKSEL_CON(33), 0, 5, DFLAGS,
 			RK3288_CLKGATE_CON(5), 8, GFLAGS),
 
-	COMPOSITE(SCLK_GPU, "sclk_gpu", mux_pll_src_cpll_gpll_usb480m_p, 0,
+	COMPOSITE(SCLK_GPU, "sclk_gpu", mux_pll_src_cpll_gll_usb_npll_p, 0,
 			RK3288_CLKSEL_CON(34), 6, 2, MFLAGS, 0, 5, DFLAGS,
 			RK3288_CLKGATE_CON(5), 7, GFLAGS),
 
-	COMPOSITE(0, "aclk_peri_src", mux_pll_src_cpll_gpll_p, 0,
+	COMPOSITE(0, "aclk_peri_src", mux_pll_src_cpll_gpll_p, CLK_IGNORE_UNUSED,
 			RK3288_CLKSEL_CON(10), 15, 1, MFLAGS, 0, 5, DFLAGS,
 			RK3288_CLKGATE_CON(2), 0, GFLAGS),
-	COMPOSITE_NOMUX(0, "pclk_peri", "aclk_peri_src", 0,
+	COMPOSITE_NOMUX(PCLK_PERI, "pclk_peri", "aclk_peri_src", 0,
 			RK3288_CLKSEL_CON(10), 12, 2, DFLAGS | CLK_DIVIDER_POWER_OF_TWO,
 			RK3288_CLKGATE_CON(2), 3, GFLAGS),
-	COMPOSITE_NOMUX(0, "hclk_peri", "aclk_peri_src", 0,
+	COMPOSITE_NOMUX(HCLK_PERI, "hclk_peri", "aclk_peri_src", CLK_IGNORE_UNUSED,
 			RK3288_CLKSEL_CON(10), 8, 2, DFLAGS | CLK_DIVIDER_POWER_OF_TWO,
 			RK3288_CLKGATE_CON(2), 2, GFLAGS),
-	GATE(0, "aclk_peri", "aclk_peri_src", 0,
+	GATE(ACLK_PERI, "aclk_peri", "aclk_peri_src", CLK_IGNORE_UNUSED,
 			RK3288_CLKGATE_CON(2), 1, GFLAGS),
 
 	/*
@@ -407,6 +486,18 @@ static struct rockchip_clk_branch rk3288_clk_branches[] __initdata = {
 			RK3288_CLKSEL_CON(12), 14, 2, MFLAGS, 8, 6, DFLAGS,
 			RK3288_CLKGATE_CON(13), 3, GFLAGS),
 
+	MMC(SCLK_SDMMC_DRV,    "sdmmc_drv",    "sclk_sdmmc", RK3288_SDMMC_CON0, 1),
+	MMC(SCLK_SDMMC_SAMPLE, "sdmmc_sample", "sclk_sdmmc", RK3288_SDMMC_CON1, 0),
+
+	MMC(SCLK_SDIO0_DRV,    "sdio0_drv",    "sclk_sdio0", RK3288_SDIO0_CON0, 1),
+	MMC(SCLK_SDIO0_SAMPLE, "sdio0_sample", "sclk_sdio0", RK3288_SDIO0_CON1, 0),
+
+	MMC(SCLK_SDIO1_DRV,    "sdio1_drv",    "sclk_sdio1", RK3288_SDIO1_CON0, 1),
+	MMC(SCLK_SDIO1_SAMPLE, "sdio1_sample", "sclk_sdio1", RK3288_SDIO1_CON1, 0),
+
+	MMC(SCLK_EMMC_DRV,     "emmc_drv",     "sclk_emmc",  RK3288_EMMC_CON0,  1),
+	MMC(SCLK_EMMC_SAMPLE,  "emmc_sample",  "sclk_emmc",  RK3288_EMMC_CON1,  0),
+
 	COMPOSITE(0, "sclk_tspout", mux_tspout_p, 0,
 			RK3288_CLKSEL_CON(35), 14, 2, MFLAGS, 8, 5, DFLAGS,
 			RK3288_CLKGATE_CON(4), 11, GFLAGS),
@@ -414,13 +505,13 @@ static struct rockchip_clk_branch rk3288_clk_branches[] __initdata = {
 			RK3288_CLKSEL_CON(35), 6, 2, MFLAGS, 0, 5, DFLAGS,
 			RK3288_CLKGATE_CON(4), 10, GFLAGS),
 
-	GATE(SCLK_OTGPHY0, "sclk_otgphy0", "usb480m", 0,
+	GATE(SCLK_OTGPHY0, "sclk_otgphy0", "usb480m", CLK_IGNORE_UNUSED,
 			RK3288_CLKGATE_CON(13), 4, GFLAGS),
-	GATE(SCLK_OTGPHY1, "sclk_otgphy1", "usb480m", 0,
+	GATE(SCLK_OTGPHY1, "sclk_otgphy1", "usb480m", CLK_IGNORE_UNUSED,
 			RK3288_CLKGATE_CON(13), 5, GFLAGS),
-	GATE(SCLK_OTGPHY2, "sclk_otgphy2", "usb480m", 0,
+	GATE(SCLK_OTGPHY2, "sclk_otgphy2", "usb480m", CLK_IGNORE_UNUSED,
 			RK3288_CLKGATE_CON(13), 6, GFLAGS),
-	GATE(SCLK_OTG_ADP, "sclk_otg_adp", "xin32k", 0,
+	GATE(SCLK_OTG_ADP, "sclk_otg_adp", "xin32k", CLK_IGNORE_UNUSED,
 			RK3288_CLKGATE_CON(13), 7, GFLAGS),
 
 	COMPOSITE_NOMUX(SCLK_TSADC, "sclk_tsadc", "xin32k", 0,
@@ -441,7 +532,7 @@ static struct rockchip_clk_branch rk3288_clk_branches[] __initdata = {
 			RK3288_CLKSEL_CON(38), 15, 1, MFLAGS, 8, 5, DFLAGS,
 			RK3288_CLKGATE_CON(5), 6, GFLAGS),
 
-	COMPOSITE(0, "uart0_src", mux_uart0_pll_p, 0,
+	COMPOSITE(0, "uart0_src", mux_pll_src_cpll_gll_usb_npll_p, 0,
 			RK3288_CLKSEL_CON(13), 13, 2, MFLAGS, 0, 7, DFLAGS,
 			RK3288_CLKGATE_CON(1), 8, GFLAGS),
 	COMPOSITE_FRAC(0, "uart0_frac", "uart0_src", 0,
@@ -509,7 +600,7 @@ static struct rockchip_clk_branch rk3288_clk_branches[] __initdata = {
 
 	COMPOSITE_NODIV(0, "usbphy480m_src", mux_usbphy480m_p, 0,
 			RK3288_CLKSEL_CON(13), 11, 2, MFLAGS,
-			RK3288_CLKGATE_CON(5), 15, GFLAGS),
+			RK3288_CLKGATE_CON(5), 14, GFLAGS),
 	COMPOSITE_NODIV(SCLK_HSICPHY480M, "sclk_hsicphy480m", mux_hsicphy480m_p, 0,
 			RK3288_CLKSEL_CON(29), 0, 2, MFLAGS,
 			RK3288_CLKGATE_CON(3), 6, GFLAGS),
@@ -525,19 +616,19 @@ static struct rockchip_clk_branch rk3288_clk_branches[] __initdata = {
 	 */
 
 	/* aclk_cpu gates */
-	GATE(0, "sclk_intmem0", "aclk_cpu", 0, RK3288_CLKGATE_CON(10), 5, GFLAGS),
-	GATE(0, "sclk_intmem1", "aclk_cpu", 0, RK3288_CLKGATE_CON(10), 6, GFLAGS),
-	GATE(0, "sclk_intmem2", "aclk_cpu", 0, RK3288_CLKGATE_CON(10), 7, GFLAGS),
+	GATE(0, "sclk_intmem0", "aclk_cpu", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(10), 5, GFLAGS),
+	GATE(0, "sclk_intmem1", "aclk_cpu", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(10), 6, GFLAGS),
+	GATE(0, "sclk_intmem2", "aclk_cpu", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(10), 7, GFLAGS),
 	GATE(ACLK_DMAC1, "aclk_dmac1", "aclk_cpu", 0, RK3288_CLKGATE_CON(10), 12, GFLAGS),
-	GATE(0, "aclk_strc_sys", "aclk_cpu", 0, RK3288_CLKGATE_CON(10), 13, GFLAGS),
-	GATE(0, "aclk_intmem", "aclk_cpu", 0, RK3288_CLKGATE_CON(10), 4, GFLAGS),
+	GATE(0, "aclk_strc_sys", "aclk_cpu", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(10), 13, GFLAGS),
+	GATE(0, "aclk_intmem", "aclk_cpu", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(10), 4, GFLAGS),
 	GATE(ACLK_CRYPTO, "aclk_crypto", "aclk_cpu", 0, RK3288_CLKGATE_CON(11), 6, GFLAGS),
 	GATE(0, "aclk_ccp", "aclk_cpu", 0, RK3288_CLKGATE_CON(11), 8, GFLAGS),
 
 	/* hclk_cpu gates */
 	GATE(HCLK_CRYPTO, "hclk_crypto", "hclk_cpu", 0, RK3288_CLKGATE_CON(11), 7, GFLAGS),
 	GATE(HCLK_I2S0, "hclk_i2s0", "hclk_cpu", 0, RK3288_CLKGATE_CON(10), 8, GFLAGS),
-	GATE(HCLK_ROM, "hclk_rom", "hclk_cpu", 0, RK3288_CLKGATE_CON(10), 9, GFLAGS),
+	GATE(HCLK_ROM, "hclk_rom", "hclk_cpu", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(10), 9, GFLAGS),
 	GATE(HCLK_SPDIF, "hclk_spdif", "hclk_cpu", 0, RK3288_CLKGATE_CON(10), 10, GFLAGS),
 	GATE(HCLK_SPDIF8CH, "hclk_spdif_8ch", "hclk_cpu", 0, RK3288_CLKGATE_CON(10), 11, GFLAGS),
 
@@ -546,42 +637,42 @@ static struct rockchip_clk_branch rk3288_clk_branches[] __initdata = {
 	GATE(PCLK_TIMER, "pclk_timer", "pclk_cpu", 0, RK3288_CLKGATE_CON(10), 1, GFLAGS),
 	GATE(PCLK_I2C0, "pclk_i2c0", "pclk_cpu", 0, RK3288_CLKGATE_CON(10), 2, GFLAGS),
 	GATE(PCLK_I2C2, "pclk_i2c2", "pclk_cpu", 0, RK3288_CLKGATE_CON(10), 3, GFLAGS),
-	GATE(0, "pclk_ddrupctl0", "pclk_cpu", 0, RK3288_CLKGATE_CON(10), 14, GFLAGS),
-	GATE(0, "pclk_publ0", "pclk_cpu", 0, RK3288_CLKGATE_CON(10), 15, GFLAGS),
-	GATE(0, "pclk_ddrupctl1", "pclk_cpu", 0, RK3288_CLKGATE_CON(11), 0, GFLAGS),
-	GATE(0, "pclk_publ1", "pclk_cpu", 0, RK3288_CLKGATE_CON(11), 1, GFLAGS),
+	GATE(PCLK_DDRUPCTL0, "pclk_ddrupctl0", "pclk_cpu", 0, RK3288_CLKGATE_CON(10), 14, GFLAGS),
+	GATE(PCLK_PUBL0, "pclk_publ0", "pclk_cpu", 0, RK3288_CLKGATE_CON(10), 15, GFLAGS),
+	GATE(PCLK_DDRUPCTL1, "pclk_ddrupctl1", "pclk_cpu", 0, RK3288_CLKGATE_CON(11), 0, GFLAGS),
+	GATE(PCLK_PUBL1, "pclk_publ1", "pclk_cpu", 0, RK3288_CLKGATE_CON(11), 1, GFLAGS),
 	GATE(0, "pclk_efuse_1024", "pclk_cpu", 0, RK3288_CLKGATE_CON(11), 2, GFLAGS),
 	GATE(PCLK_TZPC, "pclk_tzpc", "pclk_cpu", 0, RK3288_CLKGATE_CON(11), 3, GFLAGS),
 	GATE(PCLK_UART2, "pclk_uart2", "pclk_cpu", 0, RK3288_CLKGATE_CON(11), 9, GFLAGS),
 	GATE(0, "pclk_efuse_256", "pclk_cpu", 0, RK3288_CLKGATE_CON(11), 10, GFLAGS),
-	GATE(PCLK_RKPWM, "pclk_rkpwm", "pclk_cpu", 0, RK3288_CLKGATE_CON(11), 11, GFLAGS),
+	GATE(PCLK_RKPWM, "pclk_rkpwm", "pclk_cpu", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(11), 11, GFLAGS),
 
 	/* ddrctrl [DDR Controller PHY clock] gates */
-	GATE(0, "nclk_ddrupctl0", "ddrphy", 0, RK3288_CLKGATE_CON(11), 4, GFLAGS),
-	GATE(0, "nclk_ddrupctl1", "ddrphy", 0, RK3288_CLKGATE_CON(11), 5, GFLAGS),
+	GATE(0, "nclk_ddrupctl0", "ddrphy", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(11), 4, GFLAGS),
+	GATE(0, "nclk_ddrupctl1", "ddrphy", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(11), 5, GFLAGS),
 
 	/* ddrphy gates */
-	GATE(0, "sclk_ddrphy0", "ddrphy", 0, RK3288_CLKGATE_CON(4), 12, GFLAGS),
-	GATE(0, "sclk_ddrphy1", "ddrphy", 0, RK3288_CLKGATE_CON(4), 13, GFLAGS),
+	GATE(0, "sclk_ddrphy0", "ddrphy", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(4), 12, GFLAGS),
+	GATE(0, "sclk_ddrphy1", "ddrphy", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(4), 13, GFLAGS),
 
 	/* aclk_peri gates */
-	GATE(0, "aclk_peri_axi_matrix", "aclk_peri", 0, RK3288_CLKGATE_CON(6), 2, GFLAGS),
+	GATE(0, "aclk_peri_axi_matrix", "aclk_peri", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(6), 2, GFLAGS),
 	GATE(ACLK_DMAC2, "aclk_dmac2", "aclk_peri", 0, RK3288_CLKGATE_CON(6), 3, GFLAGS),
-	GATE(0, "aclk_peri_niu", "aclk_peri", 0, RK3288_CLKGATE_CON(7), 11, GFLAGS),
-	GATE(ACLK_MMU, "aclk_mmu", "aclk_peri", 0, RK3288_CLKGATE_CON(8), 12, GFLAGS),
+	GATE(0, "aclk_peri_niu", "aclk_peri", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(7), 11, GFLAGS),
+	GATE(ACLK_MMU, "aclk_mmu", "aclk_peri", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(8), 12, GFLAGS),
 	GATE(ACLK_GMAC, "aclk_gmac", "aclk_peri", 0, RK3288_CLKGATE_CON(8), 0, GFLAGS),
 	GATE(HCLK_GPS, "hclk_gps", "aclk_peri", 0, RK3288_CLKGATE_CON(8), 2, GFLAGS),
 
 	/* hclk_peri gates */
-	GATE(0, "hclk_peri_matrix", "hclk_peri", 0, RK3288_CLKGATE_CON(6), 0, GFLAGS),
-	GATE(HCLK_OTG0, "hclk_otg0", "hclk_peri", 0, RK3288_CLKGATE_CON(7), 4, GFLAGS),
+	GATE(0, "hclk_peri_matrix", "hclk_peri", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(6), 0, GFLAGS),
+	GATE(HCLK_OTG0, "hclk_otg0", "hclk_peri", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(7), 4, GFLAGS),
 	GATE(HCLK_USBHOST0, "hclk_host0", "hclk_peri", 0, RK3288_CLKGATE_CON(7), 6, GFLAGS),
-	GATE(HCLK_USBHOST1, "hclk_host1", "hclk_peri", 0, RK3288_CLKGATE_CON(7), 7, GFLAGS),
+	GATE(HCLK_USBHOST1, "hclk_host1", "hclk_peri", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(7), 7, GFLAGS),
 	GATE(HCLK_HSIC, "hclk_hsic", "hclk_peri", 0, RK3288_CLKGATE_CON(7), 8, GFLAGS),
-	GATE(0, "hclk_usb_peri", "hclk_peri", 0, RK3288_CLKGATE_CON(7), 9, GFLAGS),
-	GATE(0, "hclk_peri_ahb_arbi", "hclk_peri", 0, RK3288_CLKGATE_CON(7), 10, GFLAGS),
-	GATE(0, "hclk_emem", "hclk_peri", 0, RK3288_CLKGATE_CON(7), 12, GFLAGS),
-	GATE(0, "hclk_mem", "hclk_peri", 0, RK3288_CLKGATE_CON(7), 13, GFLAGS),
+	GATE(0, "hclk_usb_peri", "hclk_peri", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(7), 9, GFLAGS),
+	GATE(0, "hclk_peri_ahb_arbi", "hclk_peri", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(7), 10, GFLAGS),
+	GATE(0, "hclk_emem", "hclk_peri", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(7), 12, GFLAGS),
+	GATE(0, "hclk_mem", "hclk_peri", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(7), 13, GFLAGS),
 	GATE(HCLK_NANDC0, "hclk_nandc0", "hclk_peri", 0, RK3288_CLKGATE_CON(7), 14, GFLAGS),
 	GATE(HCLK_NANDC1, "hclk_nandc1", "hclk_peri", 0, RK3288_CLKGATE_CON(7), 15, GFLAGS),
 	GATE(HCLK_TSP, "hclk_tsp", "hclk_peri", 0, RK3288_CLKGATE_CON(8), 8, GFLAGS),
@@ -593,7 +684,7 @@ static struct rockchip_clk_branch rk3288_clk_branches[] __initdata = {
 	GATE(0, "pmu_hclk_otg0", "hclk_peri", 0, RK3288_CLKGATE_CON(7), 5, GFLAGS),
 
 	/* pclk_peri gates */
-	GATE(0, "pclk_peri_matrix", "pclk_peri", 0, RK3288_CLKGATE_CON(6), 1, GFLAGS),
+	GATE(0, "pclk_peri_matrix", "pclk_peri", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(6), 1, GFLAGS),
 	GATE(PCLK_SPI0, "pclk_spi0", "pclk_peri", 0, RK3288_CLKGATE_CON(6), 4, GFLAGS),
 	GATE(PCLK_SPI1, "pclk_spi1", "pclk_peri", 0, RK3288_CLKGATE_CON(6), 5, GFLAGS),
 	GATE(PCLK_SPI2, "pclk_spi2", "pclk_peri", 0, RK3288_CLKGATE_CON(6), 6, GFLAGS),
@@ -629,48 +720,48 @@ static struct rockchip_clk_branch rk3288_clk_branches[] __initdata = {
 	GATE(PCLK_GPIO4, "pclk_gpio4", "pclk_pd_alive", 0, RK3288_CLKGATE_CON(14), 4, GFLAGS),
 	GATE(PCLK_GPIO5, "pclk_gpio5", "pclk_pd_alive", 0, RK3288_CLKGATE_CON(14), 5, GFLAGS),
 	GATE(PCLK_GPIO6, "pclk_gpio6", "pclk_pd_alive", 0, RK3288_CLKGATE_CON(14), 6, GFLAGS),
-	GATE(PCLK_GRF, "pclk_grf", "pclk_pd_alive", 0, RK3288_CLKGATE_CON(14), 11, GFLAGS),
-	GATE(0, "pclk_alive_niu", "pclk_pd_alive", 0, RK3288_CLKGATE_CON(14), 12, GFLAGS),
+	GATE(PCLK_GRF, "pclk_grf", "pclk_pd_alive", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(14), 11, GFLAGS),
+	GATE(0, "pclk_alive_niu", "pclk_pd_alive", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(14), 12, GFLAGS),
 
 	/* pclk_pd_pmu gates */
-	GATE(PCLK_PMU, "pclk_pmu", "pclk_pd_pmu", 0, RK3288_CLKGATE_CON(17), 0, GFLAGS),
-	GATE(0, "pclk_intmem1", "pclk_pd_pmu", 0, RK3288_CLKGATE_CON(17), 1, GFLAGS),
-	GATE(0, "pclk_pmu_niu", "pclk_pd_pmu", 0, RK3288_CLKGATE_CON(17), 2, GFLAGS),
-	GATE(PCLK_SGRF, "pclk_sgrf", "pclk_pd_pmu", 0, RK3288_CLKGATE_CON(17), 3, GFLAGS),
+	GATE(PCLK_PMU, "pclk_pmu", "pclk_pd_pmu", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(17), 0, GFLAGS),
+	GATE(0, "pclk_intmem1", "pclk_pd_pmu", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(17), 1, GFLAGS),
+	GATE(0, "pclk_pmu_niu", "pclk_pd_pmu", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(17), 2, GFLAGS),
+	GATE(PCLK_SGRF, "pclk_sgrf", "pclk_pd_pmu", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(17), 3, GFLAGS),
 	GATE(PCLK_GPIO0, "pclk_gpio0", "pclk_pd_pmu", 0, RK3288_CLKGATE_CON(17), 4, GFLAGS),
 
 	/* hclk_vio gates */
 	GATE(HCLK_RGA, "hclk_rga", "hclk_vio", 0, RK3288_CLKGATE_CON(15), 1, GFLAGS),
 	GATE(HCLK_VOP0, "hclk_vop0", "hclk_vio", 0, RK3288_CLKGATE_CON(15), 6, GFLAGS),
 	GATE(HCLK_VOP1, "hclk_vop1", "hclk_vio", 0, RK3288_CLKGATE_CON(15), 8, GFLAGS),
-	GATE(0, "hclk_vio_ahb_arbi", "hclk_vio", 0, RK3288_CLKGATE_CON(15), 9, GFLAGS),
-	GATE(0, "hclk_vio_niu", "hclk_vio", 0, RK3288_CLKGATE_CON(15), 10, GFLAGS),
-	GATE(0, "hclk_vip", "hclk_vio", 0, RK3288_CLKGATE_CON(15), 15, GFLAGS),
+	GATE(HCLK_VIO_AHB_ARBI, "hclk_vio_ahb_arbi", "hclk_vio", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(15), 9, GFLAGS),
+	GATE(HCLK_VIO_NIU, "hclk_vio_niu", "hclk_vio", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(15), 10, GFLAGS),
+	GATE(HCLK_VIP, "hclk_vip", "hclk_vio", 0, RK3288_CLKGATE_CON(15), 15, GFLAGS),
 	GATE(HCLK_IEP, "hclk_iep", "hclk_vio", 0, RK3288_CLKGATE_CON(15), 3, GFLAGS),
 	GATE(HCLK_ISP, "hclk_isp", "hclk_vio", 0, RK3288_CLKGATE_CON(16), 1, GFLAGS),
-	GATE(0, "hclk_vio2_h2p", "hclk_vio", 0, RK3288_CLKGATE_CON(16), 10, GFLAGS),
-	GATE(0, "pclk_mipi_dsi0", "hclk_vio", 0, RK3288_CLKGATE_CON(16), 4, GFLAGS),
-	GATE(0, "pclk_mipi_dsi1", "hclk_vio", 0, RK3288_CLKGATE_CON(16), 5, GFLAGS),
-	GATE(0, "pclk_mipi_csi", "hclk_vio", 0, RK3288_CLKGATE_CON(16), 6, GFLAGS),
-	GATE(0, "pclk_lvds_phy", "hclk_vio", 0, RK3288_CLKGATE_CON(16), 7, GFLAGS),
-	GATE(0, "pclk_edp_ctrl", "hclk_vio", 0, RK3288_CLKGATE_CON(16), 8, GFLAGS),
-	GATE(0, "pclk_hdmi_ctrl", "hclk_vio", 0, RK3288_CLKGATE_CON(16), 9, GFLAGS),
-	GATE(0, "pclk_vio2_h2p", "hclk_vio", 0, RK3288_CLKGATE_CON(16), 11, GFLAGS),
+	GATE(HCLK_VIO2_H2P, "hclk_vio2_h2p", "hclk_vio", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(16), 10, GFLAGS),
+	GATE(PCLK_MIPI_DSI0, "pclk_mipi_dsi0", "hclk_vio", 0, RK3288_CLKGATE_CON(16), 4, GFLAGS),
+	GATE(PCLK_MIPI_DSI1, "pclk_mipi_dsi1", "hclk_vio", 0, RK3288_CLKGATE_CON(16), 5, GFLAGS),
+	GATE(PCLK_MIPI_CSI, "pclk_mipi_csi", "hclk_vio", 0, RK3288_CLKGATE_CON(16), 6, GFLAGS),
+	GATE(PCLK_LVDS_PHY, "pclk_lvds_phy", "hclk_vio", 0, RK3288_CLKGATE_CON(16), 7, GFLAGS),
+	GATE(PCLK_EDP_CTRL, "pclk_edp_ctrl", "hclk_vio", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(16), 8, GFLAGS),
+	GATE(PCLK_HDMI_CTRL, "pclk_hdmi_ctrl", "hclk_vio", 0, RK3288_CLKGATE_CON(16), 9, GFLAGS),
+	GATE(PCLK_VIO2_H2P, "pclk_vio2_h2p", "hclk_vio", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(16), 11, GFLAGS),
 
 	/* aclk_vio0 gates */
 	GATE(ACLK_VOP0, "aclk_vop0", "aclk_vio0", 0, RK3288_CLKGATE_CON(15), 5, GFLAGS),
-	GATE(0, "aclk_iep", "aclk_vio0", 0, RK3288_CLKGATE_CON(15), 2, GFLAGS),
-	GATE(0, "aclk_vio0_niu", "aclk_vio0", 0, RK3288_CLKGATE_CON(15), 11, GFLAGS),
-	GATE(0, "aclk_vip", "aclk_vio0", 0, RK3288_CLKGATE_CON(15), 14, GFLAGS),
+	GATE(ACLK_IEP, "aclk_iep", "aclk_vio0", 0, RK3288_CLKGATE_CON(15), 2, GFLAGS),
+	GATE(ACLK_VIO0_NIU, "aclk_vio0_niu", "aclk_vio0", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(15), 11, GFLAGS),
+	GATE(ACLK_VIP, "aclk_vip", "aclk_vio0", 0, RK3288_CLKGATE_CON(15), 14, GFLAGS),
 
 	/* aclk_vio1 gates */
 	GATE(ACLK_VOP1, "aclk_vop1", "aclk_vio1", 0, RK3288_CLKGATE_CON(15), 7, GFLAGS),
-	GATE(0, "aclk_isp", "aclk_vio1", 0, RK3288_CLKGATE_CON(16), 2, GFLAGS),
-	GATE(0, "aclk_vio1_niu", "aclk_vio1", 0, RK3288_CLKGATE_CON(15), 12, GFLAGS),
+	GATE(ACLK_ISP, "aclk_isp", "aclk_vio1", 0, RK3288_CLKGATE_CON(16), 2, GFLAGS),
+	GATE(ACLK_VIO1_NIU, "aclk_vio1_niu", "aclk_vio1", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(15), 12, GFLAGS),
 
 	/* aclk_rga_pre gates */
 	GATE(ACLK_RGA, "aclk_rga", "aclk_rga_pre", 0, RK3288_CLKGATE_CON(15), 0, GFLAGS),
-	GATE(0, "aclk_rga_niu", "aclk_rga_pre", 0, RK3288_CLKGATE_CON(15), 13, GFLAGS),
+	GATE(ACLK_RGA_NIU, "aclk_rga_niu", "aclk_rga_pre", CLK_IGNORE_UNUSED, RK3288_CLKGATE_CON(15), 13, GFLAGS),
 
 	/*
 	 * Other ungrouped clocks.
@@ -680,6 +771,70 @@ static struct rockchip_clk_branch rk3288_clk_branches[] __initdata = {
 	GATE(0, "pclk_isp_in", "ext_isp", 0, RK3288_CLKGATE_CON(16), 3, GFLAGS),
 };
 
+static const char *rk3288_critical_clocks[] __initconst = {
+	"aclk_cpu",
+	"aclk_peri",
+	"hclk_peri",
+};
+
+#ifdef CONFIG_PM_SLEEP
+static void __iomem *rk3288_cru_base;
+
+/* Some CRU registers will be reset in maskrom when the system
+ * wakes up from fastboot.
+ * So save them before suspend, restore them after resume.
+ */
+static const int rk3288_saved_cru_reg_ids[] = {
+	RK3288_MODE_CON,
+	RK3288_CLKSEL_CON(0),
+	RK3288_CLKSEL_CON(1),
+	RK3288_CLKSEL_CON(10),
+	RK3288_CLKSEL_CON(33),
+	RK3288_CLKSEL_CON(37),
+};
+
+static u32 rk3288_saved_cru_regs[ARRAY_SIZE(rk3288_saved_cru_reg_ids)];
+
+static int rk3288_clk_suspend(void)
+{
+	int i, reg_id;
+
+	for (i = 0; i < ARRAY_SIZE(rk3288_saved_cru_reg_ids); i++) {
+		reg_id = rk3288_saved_cru_reg_ids[i];
+
+		rk3288_saved_cru_regs[i] =
+				readl_relaxed(rk3288_cru_base + reg_id);
+	}
+	return 0;
+}
+
+static void rk3288_clk_resume(void)
+{
+	int i, reg_id;
+
+	for (i = ARRAY_SIZE(rk3288_saved_cru_reg_ids) - 1; i >= 0; i--) {
+		reg_id = rk3288_saved_cru_reg_ids[i];
+
+		writel_relaxed(rk3288_saved_cru_regs[i] | 0xffff0000,
+			       rk3288_cru_base + reg_id);
+	}
+}
+
+static struct syscore_ops rk3288_clk_syscore_ops = {
+	.suspend = rk3288_clk_suspend,
+	.resume = rk3288_clk_resume,
+};
+
+static void rk3288_clk_sleep_init(void __iomem *reg_base)
+{
+	rk3288_cru_base = reg_base;
+	register_syscore_ops(&rk3288_clk_syscore_ops);
+}
+
+#else /* CONFIG_PM_SLEEP */
+static void rk3288_clk_sleep_init(void __iomem *reg_base) {}
+#endif
+
 static void __init rk3288_clk_init(struct device_node *np)
 {
 	void __iomem *reg_base;
@@ -705,13 +860,29 @@ static void __init rk3288_clk_init(struct device_node *np)
 		pr_warn("%s: could not register clock usb480m: %ld\n",
 			__func__, PTR_ERR(clk));
 
+	clk = clk_register_fixed_factor(NULL, "hclk_vcodec_pre",
+					"hclk_vcodec_pre_v", 0, 1, 4);
+	if (IS_ERR(clk))
+		pr_warn("%s: could not register clock hclk_vcodec_pre: %ld\n",
+			__func__, PTR_ERR(clk));
+
 	rockchip_clk_register_plls(rk3288_pll_clks,
 				   ARRAY_SIZE(rk3288_pll_clks),
-				   RK3288_GRF_SOC_STATUS);
+				   RK3288_GRF_SOC_STATUS1);
 	rockchip_clk_register_branches(rk3288_clk_branches,
 				  ARRAY_SIZE(rk3288_clk_branches));
+	rockchip_clk_protect_critical(rk3288_critical_clocks,
+				      ARRAY_SIZE(rk3288_critical_clocks));
 
-	rockchip_register_softrst(np, 9, reg_base + RK3288_SOFTRST_CON(0),
+	rockchip_clk_register_armclk(ARMCLK, "armclk",
+			mux_armclk_p, ARRAY_SIZE(mux_armclk_p),
+			&rk3288_cpuclk_data, rk3288_cpuclk_rates,
+			ARRAY_SIZE(rk3288_cpuclk_rates));
+
+	rockchip_register_softrst(np, 12, reg_base + RK3288_SOFTRST_CON(0),
 				  ROCKCHIP_SOFTRST_HIWORD_MASK);
+
+	rockchip_register_restart_notifier(RK3288_GLB_SRST_FST);
+	rk3288_clk_sleep_init(reg_base);
 }
 CLK_OF_DECLARE(rk3288_cru, "rockchip,rk3288-cru", rk3288_clk_init);
diff --git a/drivers/clk/rockchip/clk.c b/drivers/clk/rockchip/clk.c
index 278cf9dd1e23..20e05bbb3a67 100644
--- a/drivers/clk/rockchip/clk.c
+++ b/drivers/clk/rockchip/clk.c
@@ -25,6 +25,7 @@
 #include <linux/clk-provider.h>
 #include <linux/mfd/syscon.h>
 #include <linux/regmap.h>
+#include <linux/reboot.h>
 #include "clk.h"
 
 /**
@@ -37,7 +38,7 @@
  *
  * sometimes without one of those components.
  */
-struct clk *rockchip_clk_register_branch(const char *name,
+static struct clk *rockchip_clk_register_branch(const char *name,
 		const char **parent_names, u8 num_parents, void __iomem *base,
 		int muxdiv_offset, u8 mux_shift, u8 mux_width, u8 mux_flags,
 		u8 div_shift, u8 div_width, u8 div_flags,
@@ -89,9 +90,7 @@ struct clk *rockchip_clk_register_branch(const char *name,
 		div->width = div_width;
 		div->lock = lock;
 		div->table = div_table;
-		div_ops = (div_flags & CLK_DIVIDER_READ_ONLY)
-						? &clk_divider_ro_ops
-						: &clk_divider_ops;
+		div_ops = &clk_divider_ops;
 	}
 
 	clk = clk_register_composite(NULL, name, parent_names, num_parents,
@@ -103,6 +102,54 @@ struct clk *rockchip_clk_register_branch(const char *name,
 	return clk;
 }
 
+static struct clk *rockchip_clk_register_frac_branch(const char *name,
+		const char **parent_names, u8 num_parents, void __iomem *base,
+		int muxdiv_offset, u8 div_flags,
+		int gate_offset, u8 gate_shift, u8 gate_flags,
+		unsigned long flags, spinlock_t *lock)
+{
+	struct clk *clk;
+	struct clk_gate *gate = NULL;
+	struct clk_fractional_divider *div = NULL;
+	const struct clk_ops *div_ops = NULL, *gate_ops = NULL;
+
+	if (gate_offset >= 0) {
+		gate = kzalloc(sizeof(*gate), GFP_KERNEL);
+		if (!gate)
+			return ERR_PTR(-ENOMEM);
+
+		gate->flags = gate_flags;
+		gate->reg = base + gate_offset;
+		gate->bit_idx = gate_shift;
+		gate->lock = lock;
+		gate_ops = &clk_gate_ops;
+	}
+
+	if (muxdiv_offset < 0)
+		return ERR_PTR(-EINVAL);
+
+	div = kzalloc(sizeof(*div), GFP_KERNEL);
+	if (!div)
+		return ERR_PTR(-ENOMEM);
+
+	div->flags = div_flags;
+	div->reg = base + muxdiv_offset;
+	div->mshift = 16;
+	div->mmask = 0xffff0000;
+	div->nshift = 0;
+	div->nmask = 0xffff;
+	div->lock = lock;
+	div_ops = &clk_fractional_divider_ops;
+
+	clk = clk_register_composite(NULL, name, parent_names, num_parents,
+				     NULL, NULL,
+				     &div->hw, div_ops,
+				     gate ? &gate->hw : NULL, gate_ops,
+				     flags);
+
+	return clk;
+}
+
 static DEFINE_SPINLOCK(clk_lock);
 static struct clk **clk_table;
 static void __iomem *reg_base;
@@ -150,7 +197,8 @@ void __init rockchip_clk_register_plls(struct rockchip_pll_clock *list,
 				list->parent_names, list->num_parents,
 				reg_base, list->con_offset, grf_lock_offset,
 				list->lock_shift, list->mode_offset,
-				list->mode_shift, list->rate_table, &clk_lock);
+				list->mode_shift, list->rate_table,
+				list->pll_flags, &clk_lock);
 		if (IS_ERR(clk)) {
 			pr_err("%s: failed to register clock %s\n", __func__,
 				list->name);
@@ -197,24 +245,21 @@ void __init rockchip_clk_register_branches(
 					list->div_flags, &clk_lock);
 			break;
 		case branch_fraction_divider:
-			/* unimplemented */
-			continue;
+			clk = rockchip_clk_register_frac_branch(list->name,
+				list->parent_names, list->num_parents,
+				reg_base, list->muxdiv_offset, list->div_flags,
+				list->gate_offset, list->gate_shift,
+				list->gate_flags, flags, &clk_lock);
 			break;
 		case branch_gate:
 			flags |= CLK_SET_RATE_PARENT;
 
-			/* keep all gates untouched for now */
-			flags |= CLK_IGNORE_UNUSED;
-
 			clk = clk_register_gate(NULL, list->name,
 				list->parent_names[0], flags,
 				reg_base + list->gate_offset,
 				list->gate_shift, list->gate_flags, &clk_lock);
 			break;
 		case branch_composite:
-			/* keep all gates untouched for now */
-			flags |= CLK_IGNORE_UNUSED;
-
 			clk = rockchip_clk_register_branch(list->name,
 				list->parent_names, list->num_parents,
 				reg_base, list->muxdiv_offset, list->mux_shift,
@@ -224,6 +269,14 @@ void __init rockchip_clk_register_branches(
 				list->gate_offset, list->gate_shift,
 				list->gate_flags, flags, &clk_lock);
 			break;
+		case branch_mmc:
+			clk = rockchip_clk_register_mmc(
+				list->name,
+				list->parent_names, list->num_parents,
+				reg_base + list->muxdiv_offset,
+				list->div_shift
+			);
+			break;
 		}
 
 		/* none of the cases above matched */
@@ -242,3 +295,61 @@ void __init rockchip_clk_register_branches(
 		rockchip_clk_add_lookup(clk, list->id);
 	}
 }
+
+void __init rockchip_clk_register_armclk(unsigned int lookup_id,
+			const char *name, const char **parent_names,
+			u8 num_parents,
+			const struct rockchip_cpuclk_reg_data *reg_data,
+			const struct rockchip_cpuclk_rate_table *rates,
+			int nrates)
+{
+	struct clk *clk;
+
+	clk = rockchip_clk_register_cpuclk(name, parent_names, num_parents,
+					   reg_data, rates, nrates, reg_base,
+					   &clk_lock);
+	if (IS_ERR(clk)) {
+		pr_err("%s: failed to register clock %s: %ld\n",
+		       __func__, name, PTR_ERR(clk));
+		return;
+	}
+
+	rockchip_clk_add_lookup(clk, lookup_id);
+}
+
+void __init rockchip_clk_protect_critical(const char *clocks[], int nclocks)
+{
+	int i;
+
+	/* Protect the clocks that needs to stay on */
+	for (i = 0; i < nclocks; i++) {
+		struct clk *clk = __clk_lookup(clocks[i]);
+
+		if (clk)
+			clk_prepare_enable(clk);
+	}
+}
+
+static unsigned int reg_restart;
+static int rockchip_restart_notify(struct notifier_block *this,
+				   unsigned long mode, void *cmd)
+{
+	writel(0xfdb9, reg_base + reg_restart);
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block rockchip_restart_handler = {
+	.notifier_call = rockchip_restart_notify,
+	.priority = 128,
+};
+
+void __init rockchip_register_restart_notifier(unsigned int reg)
+{
+	int ret;
+
+	reg_restart = reg;
+	ret = register_restart_handler(&rockchip_restart_handler);
+	if (ret)
+		pr_err("%s: cannot register restart handler, %d\n",
+		       __func__, ret);
+}
diff --git a/drivers/clk/rockchip/clk.h b/drivers/clk/rockchip/clk.h
index 887cbdeca2aa..58d2e3bdf22f 100644
--- a/drivers/clk/rockchip/clk.h
+++ b/drivers/clk/rockchip/clk.h
@@ -48,6 +48,14 @@
 #define RK3288_GLB_SRST_SND		0x1b4
 #define RK3288_SOFTRST_CON(x)		(x * 0x4 + 0x1b8)
 #define RK3288_MISC_CON			0x1e8
+#define RK3288_SDMMC_CON0		0x200
+#define RK3288_SDMMC_CON1		0x204
+#define RK3288_SDIO0_CON0		0x208
+#define RK3288_SDIO0_CON1		0x20c
+#define RK3288_SDIO1_CON0		0x210
+#define RK3288_SDIO1_CON1		0x214
+#define RK3288_EMMC_CON0		0x218
+#define RK3288_EMMC_CON1		0x21c
 
 enum rockchip_pll_type {
 	pll_rk3066,
@@ -62,6 +70,15 @@ enum rockchip_pll_type {
 	.bwadj = (_nf >> 1),			\
 }
 
+#define RK3066_PLL_RATE_BWADJ(_rate, _nr, _nf, _no, _bw)	\
+{								\
+	.rate	= _rate##U,					\
+	.nr = _nr,						\
+	.nf = _nf,						\
+	.no = _no,						\
+	.bwadj = _bw,						\
+}
+
 struct rockchip_pll_rate_table {
 	unsigned long rate;
 	unsigned int nr;
@@ -81,7 +98,12 @@ struct rockchip_pll_rate_table {
  * @mode_shift: offset inside the mode-register for the mode of this pll.
  * @lock_shift: offset inside the lock register for the lock status.
  * @type: Type of PLL to be registered.
+ * @pll_flags: hardware-specific flags
  * @rate_table: Table of usable pll rates
+ *
+ * Flags:
+ * ROCKCHIP_PLL_SYNC_RATE - check rate parameters to match against the
+ *	rate_table parameters and ajust them if necessary.
  */
 struct rockchip_pll_clock {
 	unsigned int		id;
@@ -94,11 +116,14 @@ struct rockchip_pll_clock {
 	int			mode_shift;
 	int			lock_shift;
 	enum rockchip_pll_type	type;
+	u8			pll_flags;
 	struct rockchip_pll_rate_table *rate_table;
 };
 
+#define ROCKCHIP_PLL_SYNC_RATE		BIT(0)
+
 #define PLL(_type, _id, _name, _pnames, _flags, _con, _mode, _mshift,	\
-		_lshift, _rtable)					\
+		_lshift, _pflags, _rtable)				\
 	{								\
 		.id		= _id,					\
 		.type		= _type,				\
@@ -110,6 +135,7 @@ struct rockchip_pll_clock {
 		.mode_offset	= _mode,				\
 		.mode_shift	= _mshift,				\
 		.lock_shift	= _lshift,				\
+		.pll_flags	= _pflags,				\
 		.rate_table	= _rtable,				\
 	}
 
@@ -118,7 +144,43 @@ struct clk *rockchip_clk_register_pll(enum rockchip_pll_type pll_type,
 		void __iomem *base, int con_offset, int grf_lock_offset,
 		int lock_shift, int reg_mode, int mode_shift,
 		struct rockchip_pll_rate_table *rate_table,
-		spinlock_t *lock);
+		u8 clk_pll_flags, spinlock_t *lock);
+
+struct rockchip_cpuclk_clksel {
+	int reg;
+	u32 val;
+};
+
+#define ROCKCHIP_CPUCLK_NUM_DIVIDERS	2
+struct rockchip_cpuclk_rate_table {
+	unsigned long prate;
+	struct rockchip_cpuclk_clksel divs[ROCKCHIP_CPUCLK_NUM_DIVIDERS];
+};
+
+/**
+ * struct rockchip_cpuclk_reg_data: describes register offsets and masks of the cpuclock
+ * @core_reg:		register offset of the core settings register
+ * @div_core_shift:	core divider offset used to divide the pll value
+ * @div_core_mask:	core divider mask
+ * @mux_core_shift:	offset of the core multiplexer
+ */
+struct rockchip_cpuclk_reg_data {
+	int		core_reg;
+	u8		div_core_shift;
+	u32		div_core_mask;
+	int		mux_core_reg;
+	u8		mux_core_shift;
+};
+
+struct clk *rockchip_clk_register_cpuclk(const char *name,
+			const char **parent_names, u8 num_parents,
+			const struct rockchip_cpuclk_reg_data *reg_data,
+			const struct rockchip_cpuclk_rate_table *rates,
+			int nrates, void __iomem *reg_base, spinlock_t *lock);
+
+struct clk *rockchip_clk_register_mmc(const char *name,
+				const char **parent_names, u8 num_parents,
+				void __iomem *reg, int shift);
 
 #define PNAME(x) static const char *x[] __initconst
 
@@ -128,6 +190,7 @@ enum rockchip_clk_branch_type {
 	branch_divider,
 	branch_fraction_divider,
 	branch_gate,
+	branch_mmc,
 };
 
 struct rockchip_clk_branch {
@@ -320,6 +383,16 @@ struct rockchip_clk_branch {
 		.gate_flags	= gf,				\
 	}
 
+#define MMC(_id, cname, pname, offset, shift)			\
+	{							\
+		.id		= _id,				\
+		.branch_type	= branch_mmc,			\
+		.name		= cname,			\
+		.parent_names	= (const char *[]){ pname },	\
+		.num_parents	= 1,				\
+		.muxdiv_offset	= offset,			\
+		.div_shift	= shift,			\
+	}
 
 void rockchip_clk_init(struct device_node *np, void __iomem *base,
 		       unsigned long nr_clks);
@@ -329,6 +402,13 @@ void rockchip_clk_register_branches(struct rockchip_clk_branch *clk_list,
 				    unsigned int nr_clk);
 void rockchip_clk_register_plls(struct rockchip_pll_clock *pll_list,
 				unsigned int nr_pll, int grf_lock_offset);
+void rockchip_clk_register_armclk(unsigned int lookup_id, const char *name,
+			const char **parent_names, u8 num_parents,
+			const struct rockchip_cpuclk_reg_data *reg_data,
+			const struct rockchip_cpuclk_rate_table *rates,
+			int nrates);
+void rockchip_clk_protect_critical(const char *clocks[], int nclocks);
+void rockchip_register_restart_notifier(unsigned int reg);
 
 #define ROCKCHIP_SOFTRST_HIWORD_MASK	BIT(0)
 
diff --git a/drivers/clk/samsung/Makefile b/drivers/clk/samsung/Makefile
index 6fb4bc602e8a..006c6f294310 100644
--- a/drivers/clk/samsung/Makefile
+++ b/drivers/clk/samsung/Makefile
@@ -5,6 +5,7 @@
 obj-$(CONFIG_COMMON_CLK)	+= clk.o clk-pll.o
 obj-$(CONFIG_SOC_EXYNOS3250)	+= clk-exynos3250.o
 obj-$(CONFIG_ARCH_EXYNOS4)	+= clk-exynos4.o
+obj-$(CONFIG_SOC_EXYNOS4415)	+= clk-exynos4415.o
 obj-$(CONFIG_SOC_EXYNOS5250)	+= clk-exynos5250.o
 obj-$(CONFIG_SOC_EXYNOS5260)	+= clk-exynos5260.o
 obj-$(CONFIG_SOC_EXYNOS5410)	+= clk-exynos5410.o
@@ -12,6 +13,7 @@ obj-$(CONFIG_SOC_EXYNOS5420)	+= clk-exynos5420.o
 obj-$(CONFIG_SOC_EXYNOS5440)	+= clk-exynos5440.o
 obj-$(CONFIG_ARCH_EXYNOS)	+= clk-exynos-audss.o
 obj-$(CONFIG_ARCH_EXYNOS)	+= clk-exynos-clkout.o
+obj-$(CONFIG_ARCH_EXYNOS7)	+= clk-exynos7.o
 obj-$(CONFIG_S3C2410_COMMON_CLK)+= clk-s3c2410.o
 obj-$(CONFIG_S3C2410_COMMON_DCLK)+= clk-s3c2410-dclk.o
 obj-$(CONFIG_S3C2412_COMMON_CLK)+= clk-s3c2412.o
diff --git a/drivers/clk/samsung/clk-exynos-audss.c b/drivers/clk/samsung/clk-exynos-audss.c
index 13eae14c2cc2..f2c2ccce49bb 100644
--- a/drivers/clk/samsung/clk-exynos-audss.c
+++ b/drivers/clk/samsung/clk-exynos-audss.c
@@ -29,6 +29,13 @@ static DEFINE_SPINLOCK(lock);
 static struct clk **clk_table;
 static void __iomem *reg_base;
 static struct clk_onecell_data clk_data;
+/*
+ * On Exynos5420 this will be a clock which has to be enabled before any
+ * access to audss registers. Typically a child of EPLL.
+ *
+ * On other platforms this will be -ENODEV.
+ */
+static struct clk *epll;
 
 #define ASS_CLK_SRC 0x0
 #define ASS_CLK_DIV 0x4
@@ -98,6 +105,8 @@ static int exynos_audss_clk_probe(struct platform_device *pdev)
 		dev_err(&pdev->dev, "failed to map audss registers\n");
 		return PTR_ERR(reg_base);
 	}
+	/* EPLL don't have to be enabled for boards other than Exynos5420 */
+	epll = ERR_PTR(-ENODEV);
 
 	clk_table = devm_kzalloc(&pdev->dev,
 				sizeof(struct clk *) * EXYNOS_AUDSS_MAX_CLKS,
@@ -115,8 +124,20 @@ static int exynos_audss_clk_probe(struct platform_device *pdev)
 	pll_in = devm_clk_get(&pdev->dev, "pll_in");
 	if (!IS_ERR(pll_ref))
 		mout_audss_p[0] = __clk_get_name(pll_ref);
-	if (!IS_ERR(pll_in))
+	if (!IS_ERR(pll_in)) {
 		mout_audss_p[1] = __clk_get_name(pll_in);
+
+		if (variant == TYPE_EXYNOS5420) {
+			epll = pll_in;
+
+			ret = clk_prepare_enable(epll);
+			if (ret) {
+				dev_err(&pdev->dev,
+						"failed to prepare the epll clock\n");
+				return ret;
+			}
+		}
+	}
 	clk_table[EXYNOS_MOUT_AUDSS] = clk_register_mux(NULL, "mout_audss",
 				mout_audss_p, ARRAY_SIZE(mout_audss_p),
 				CLK_SET_RATE_NO_REPARENT,
@@ -203,6 +224,9 @@ unregister:
 			clk_unregister(clk_table[i]);
 	}
 
+	if (!IS_ERR(epll))
+		clk_disable_unprepare(epll);
+
 	return ret;
 }
 
@@ -210,6 +234,10 @@ static int exynos_audss_clk_remove(struct platform_device *pdev)
 {
 	int i;
 
+#ifdef CONFIG_PM_SLEEP
+	unregister_syscore_ops(&exynos_audss_clk_syscore_ops);
+#endif
+
 	of_clk_del_provider(pdev->dev.of_node);
 
 	for (i = 0; i < clk_data.clk_num; i++) {
@@ -217,13 +245,15 @@ static int exynos_audss_clk_remove(struct platform_device *pdev)
 			clk_unregister(clk_table[i]);
 	}
 
+	if (!IS_ERR(epll))
+		clk_disable_unprepare(epll);
+
 	return 0;
 }
 
 static struct platform_driver exynos_audss_clk_driver = {
 	.driver	= {
 		.name = "exynos-audss-clk",
-		.owner = THIS_MODULE,
 		.of_match_table = exynos_audss_clk_of_match,
 	},
 	.probe = exynos_audss_clk_probe,
diff --git a/drivers/clk/samsung/clk-exynos3250.c b/drivers/clk/samsung/clk-exynos3250.c
index dc85f8e7a2d7..6e6cca392082 100644
--- a/drivers/clk/samsung/clk-exynos3250.c
+++ b/drivers/clk/samsung/clk-exynos3250.c
@@ -110,7 +110,14 @@ enum exynos3250_plls {
 	nr_plls
 };
 
+/* list of PLLs in DMC block to be registered */
+enum exynos3250_dmc_plls {
+	bpll, epll,
+	nr_dmc_plls
+};
+
 static void __iomem *reg_base;
+static void __iomem *dmc_reg_base;
 
 /*
  * Support for CMU save/restore across system suspends
@@ -266,6 +273,7 @@ PNAME(group_sclk_cam_blk_p)	= { "xxti", "xusbxti",
 				    "none", "none", "none",
 				    "none", "div_mpll_pre",
 				    "mout_epll_user", "mout_vpll",
+				    "none", "none", "none",
 				    "div_cam_blk_320", };
 PNAME(group_sclk_fimd0_p)	= { "xxti", "xusbxti",
 				    "m_bitclkhsdiv4_2l", "none",
@@ -353,8 +361,8 @@ static struct samsung_mux_clock mux_clks[] __initdata = {
 
 	/* SRC_FSYS */
 	MUX(CLK_MOUT_TSADC, "mout_tsadc", group_sclk_p, SRC_FSYS, 28, 4),
-	MUX(CLK_MOUT_MMC1, "mout_mmc1", group_sclk_p, SRC_FSYS, 4, 3),
-	MUX(CLK_MOUT_MMC0, "mout_mmc0", group_sclk_p, SRC_FSYS, 0, 3),
+	MUX(CLK_MOUT_MMC1, "mout_mmc1", group_sclk_p, SRC_FSYS, 4, 4),
+	MUX(CLK_MOUT_MMC0, "mout_mmc0", group_sclk_p, SRC_FSYS, 0, 4),
 
 	/* SRC_PERIL0 */
 	MUX(CLK_MOUT_UART1, "mout_uart1", group_sclk_p, SRC_PERIL0, 4, 4),
@@ -423,7 +431,7 @@ static struct samsung_div_clock div_clks[] __initdata = {
 	DIV(CLK_DIV_SPI1_ISP, "div_spi1_isp", "mout_spi1_isp", DIV_ISP, 16, 4),
 	DIV_F(CLK_DIV_SPI0_ISP_PRE, "div_spi0_isp_pre", "div_spi0_isp",
 		DIV_ISP, 8, 8, CLK_SET_RATE_PARENT, 0),
-	DIV(CLK_DIV_SPI0_ISP, "div_spi0_isp", "mout_spi0_isp", DIV_ISP, 0, 4),
+	DIV(CLK_DIV_SPI0_ISP, "div_spi0_isp", "mout_spi0_isp", DIV_ISP, 4, 4),
 
 	/* DIV_FSYS0 */
 	DIV_F(CLK_DIV_TSADC_PRE, "div_tsadc_pre", "div_tsadc", DIV_FSYS0, 8, 8,
@@ -724,6 +732,25 @@ static struct samsung_pll_rate_table exynos3250_pll_rates[] = {
 	{ /* sentinel */ }
 };
 
+/* EPLL */
+static struct samsung_pll_rate_table exynos3250_epll_rates[] = {
+	PLL_36XX_RATE(800000000, 200, 3, 1,     0),
+	PLL_36XX_RATE(288000000,  96, 2, 2,     0),
+	PLL_36XX_RATE(192000000, 128, 2, 3,     0),
+	PLL_36XX_RATE(144000000,  96, 2, 3,     0),
+	PLL_36XX_RATE( 96000000, 128, 2, 4,     0),
+	PLL_36XX_RATE( 84000000, 112, 2, 4,     0),
+	PLL_36XX_RATE( 80000004, 106, 2, 4, 43691),
+	PLL_36XX_RATE( 73728000,  98, 2, 4, 19923),
+	PLL_36XX_RATE( 67737598, 270, 3, 5, 62285),
+	PLL_36XX_RATE( 65535999, 174, 2, 5, 49982),
+	PLL_36XX_RATE( 50000000, 200, 3, 5,     0),
+	PLL_36XX_RATE( 49152002, 131, 2, 5,  4719),
+	PLL_36XX_RATE( 48000000, 128, 2, 5,     0),
+	PLL_36XX_RATE( 45158401, 180, 3, 5, 41524),
+	{ /* sentinel */ }
+};
+
 /* VPLL */
 static struct samsung_pll_rate_table exynos3250_vpll_rates[] = {
 	PLL_36XX_RATE(600000000, 100, 2, 1,     0),
@@ -821,3 +848,172 @@ static void __init exynos3250_cmu_init(struct device_node *np)
 	samsung_clk_of_add_provider(np, ctx);
 }
 CLK_OF_DECLARE(exynos3250_cmu, "samsung,exynos3250-cmu", exynos3250_cmu_init);
+
+/*
+ * CMU DMC
+ */
+
+#define BPLL_LOCK		0x0118
+#define BPLL_CON0		0x0218
+#define BPLL_CON1		0x021c
+#define BPLL_CON2		0x0220
+#define SRC_DMC			0x0300
+#define DIV_DMC1		0x0504
+#define GATE_BUS_DMC0		0x0700
+#define GATE_BUS_DMC1		0x0704
+#define GATE_BUS_DMC2		0x0708
+#define GATE_BUS_DMC3		0x070c
+#define GATE_SCLK_DMC		0x0800
+#define GATE_IP_DMC0		0x0900
+#define GATE_IP_DMC1		0x0904
+#define EPLL_LOCK		0x1110
+#define EPLL_CON0		0x1114
+#define EPLL_CON1		0x1118
+#define EPLL_CON2		0x111c
+#define SRC_EPLL		0x1120
+
+/*
+ * Support for CMU save/restore across system suspends
+ */
+#ifdef CONFIG_PM_SLEEP
+static struct samsung_clk_reg_dump *exynos3250_dmc_clk_regs;
+
+static unsigned long exynos3250_cmu_dmc_clk_regs[] __initdata = {
+	BPLL_LOCK,
+	BPLL_CON0,
+	BPLL_CON1,
+	BPLL_CON2,
+	SRC_DMC,
+	DIV_DMC1,
+	GATE_BUS_DMC0,
+	GATE_BUS_DMC1,
+	GATE_BUS_DMC2,
+	GATE_BUS_DMC3,
+	GATE_SCLK_DMC,
+	GATE_IP_DMC0,
+	GATE_IP_DMC1,
+	EPLL_LOCK,
+	EPLL_CON0,
+	EPLL_CON1,
+	EPLL_CON2,
+	SRC_EPLL,
+};
+
+static int exynos3250_dmc_clk_suspend(void)
+{
+	samsung_clk_save(dmc_reg_base, exynos3250_dmc_clk_regs,
+				ARRAY_SIZE(exynos3250_cmu_dmc_clk_regs));
+	return 0;
+}
+
+static void exynos3250_dmc_clk_resume(void)
+{
+	samsung_clk_restore(dmc_reg_base, exynos3250_dmc_clk_regs,
+				ARRAY_SIZE(exynos3250_cmu_dmc_clk_regs));
+}
+
+static struct syscore_ops exynos3250_dmc_clk_syscore_ops = {
+	.suspend = exynos3250_dmc_clk_suspend,
+	.resume = exynos3250_dmc_clk_resume,
+};
+
+static void exynos3250_dmc_clk_sleep_init(void)
+{
+	exynos3250_dmc_clk_regs =
+		samsung_clk_alloc_reg_dump(exynos3250_cmu_dmc_clk_regs,
+				   ARRAY_SIZE(exynos3250_cmu_dmc_clk_regs));
+	if (!exynos3250_dmc_clk_regs) {
+		pr_warn("%s: Failed to allocate sleep save data\n", __func__);
+		goto err;
+	}
+
+	register_syscore_ops(&exynos3250_dmc_clk_syscore_ops);
+	return;
+err:
+	kfree(exynos3250_dmc_clk_regs);
+}
+#else
+static inline void exynos3250_dmc_clk_sleep_init(void) { }
+#endif
+
+PNAME(mout_epll_p)	= { "fin_pll", "fout_epll", };
+PNAME(mout_bpll_p)	= { "fin_pll", "fout_bpll", };
+PNAME(mout_mpll_mif_p)	= { "fin_pll", "sclk_mpll_mif", };
+PNAME(mout_dphy_p)	= { "mout_mpll_mif", "mout_bpll", };
+
+static struct samsung_mux_clock dmc_mux_clks[] __initdata = {
+	/*
+	 * NOTE: Following table is sorted by register address in ascending
+	 * order and then bitfield shift in descending order, as it is done
+	 * in the User's Manual. When adding new entries, please make sure
+	 * that the order is preserved, to avoid merge conflicts and make
+	 * further work with defined data easier.
+	 */
+
+	/* SRC_DMC */
+	MUX(CLK_MOUT_MPLL_MIF, "mout_mpll_mif", mout_mpll_mif_p, SRC_DMC, 12, 1),
+	MUX(CLK_MOUT_BPLL, "mout_bpll", mout_bpll_p, SRC_DMC, 10, 1),
+	MUX(CLK_MOUT_DPHY, "mout_dphy", mout_dphy_p, SRC_DMC, 8, 1),
+	MUX(CLK_MOUT_DMC_BUS, "mout_dmc_bus", mout_dphy_p, SRC_DMC,  4, 1),
+
+	/* SRC_EPLL */
+	MUX(CLK_MOUT_EPLL, "mout_epll", mout_epll_p, SRC_EPLL, 4, 1),
+};
+
+static struct samsung_div_clock dmc_div_clks[] __initdata = {
+	/*
+	 * NOTE: Following table is sorted by register address in ascending
+	 * order and then bitfield shift in descending order, as it is done
+	 * in the User's Manual. When adding new entries, please make sure
+	 * that the order is preserved, to avoid merge conflicts and make
+	 * further work with defined data easier.
+	 */
+
+	/* DIV_DMC1 */
+	DIV(CLK_DIV_DMC, "div_dmc", "div_dmc_pre", DIV_DMC1, 27, 3),
+	DIV(CLK_DIV_DPHY, "div_dphy", "mout_dphy", DIV_DMC1, 23, 3),
+	DIV(CLK_DIV_DMC_PRE, "div_dmc_pre", "mout_dmc_bus", DIV_DMC1, 19, 2),
+	DIV(CLK_DIV_DMCP, "div_dmcp", "div_dmcd", DIV_DMC1, 15, 3),
+	DIV(CLK_DIV_DMCD, "div_dmcd", "div_dmc", DIV_DMC1, 11, 3),
+};
+
+static struct samsung_pll_clock exynos3250_dmc_plls[nr_dmc_plls] __initdata = {
+	[bpll] = PLL(pll_35xx, CLK_FOUT_BPLL, "fout_bpll", "fin_pll",
+			BPLL_LOCK, BPLL_CON0, NULL),
+	[epll] = PLL(pll_36xx, CLK_FOUT_EPLL, "fout_epll", "fin_pll",
+			EPLL_LOCK, EPLL_CON0, NULL),
+};
+
+static void __init exynos3250_cmu_dmc_init(struct device_node *np)
+{
+	struct samsung_clk_provider *ctx;
+
+	dmc_reg_base = of_iomap(np, 0);
+	if (!dmc_reg_base)
+		panic("%s: failed to map registers\n", __func__);
+
+	ctx = samsung_clk_init(np, dmc_reg_base, NR_CLKS_DMC);
+	if (!ctx)
+		panic("%s: unable to allocate context.\n", __func__);
+
+	exynos3250_dmc_plls[bpll].rate_table = exynos3250_pll_rates;
+	exynos3250_dmc_plls[epll].rate_table = exynos3250_epll_rates;
+
+	pr_err("CLK registering epll bpll: %d, %d, %d, %d\n",
+			exynos3250_dmc_plls[bpll].rate_table[0].rate,
+			exynos3250_dmc_plls[bpll].rate_table[0].mdiv,
+			exynos3250_dmc_plls[bpll].rate_table[0].pdiv,
+			exynos3250_dmc_plls[bpll].rate_table[0].sdiv
+	      );
+	samsung_clk_register_pll(ctx, exynos3250_dmc_plls,
+				ARRAY_SIZE(exynos3250_dmc_plls), dmc_reg_base);
+
+	samsung_clk_register_mux(ctx, dmc_mux_clks, ARRAY_SIZE(dmc_mux_clks));
+	samsung_clk_register_div(ctx, dmc_div_clks, ARRAY_SIZE(dmc_div_clks));
+
+	exynos3250_dmc_clk_sleep_init();
+
+	samsung_clk_of_add_provider(np, ctx);
+}
+CLK_OF_DECLARE(exynos3250_cmu_dmc, "samsung,exynos3250-cmu-dmc",
+		exynos3250_cmu_dmc_init);
diff --git a/drivers/clk/samsung/clk-exynos4.c b/drivers/clk/samsung/clk-exynos4.c
index ac163d7f5bc3..88e8c6bbd77f 100644
--- a/drivers/clk/samsung/clk-exynos4.c
+++ b/drivers/clk/samsung/clk-exynos4.c
@@ -505,7 +505,7 @@ static struct samsung_fixed_rate_clock exynos4_fixed_rate_ext_clks[] __initdata
 /* fixed rate clocks generated inside the soc */
 static struct samsung_fixed_rate_clock exynos4_fixed_rate_clks[] __initdata = {
 	FRATE(0, "sclk_hdmi24m", NULL, CLK_IS_ROOT, 24000000),
-	FRATE(CLK_SCLK_HDMIPHY, "sclk_hdmiphy", NULL, CLK_IS_ROOT, 27000000),
+	FRATE(CLK_SCLK_HDMIPHY, "sclk_hdmiphy", "hdmi", 0, 27000000),
 	FRATE(0, "sclk_usbphy0", NULL, CLK_IS_ROOT, 48000000),
 };
 
@@ -517,7 +517,7 @@ static struct samsung_fixed_factor_clock exynos4_fixed_factor_clks[] __initdata
 	FFACTOR(0, "sclk_apll_div_2", "sclk_apll", 1, 2, 0),
 	FFACTOR(0, "fout_mpll_div_2", "fout_mpll", 1, 2, 0),
 	FFACTOR(0, "fout_apll_div_2", "fout_apll", 1, 2, 0),
-	FFACTOR(0, "arm_clk_div_2", "arm_clk", 1, 2, 0),
+	FFACTOR(0, "arm_clk_div_2", "div_core2", 1, 2, 0),
 };
 
 static struct samsung_fixed_factor_clock exynos4210_fixed_factor_clks[] __initdata = {
@@ -535,7 +535,7 @@ static struct samsung_fixed_factor_clock exynos4x12_fixed_factor_clks[] __initda
 static struct samsung_mux_clock exynos4_mux_clks[] __initdata = {
 	MUX_FA(CLK_MOUT_APLL, "mout_apll", mout_apll_p, SRC_CPU, 0, 1,
 			CLK_SET_RATE_PARENT, 0, "mout_apll"),
-	MUX(0, "mout_hdmi", mout_hdmi_p, SRC_TV, 0, 1),
+	MUX(CLK_MOUT_HDMI, "mout_hdmi", mout_hdmi_p, SRC_TV, 0, 1),
 	MUX(0, "mout_mfc1", sclk_evpll_p, SRC_MFC, 4, 1),
 	MUX(0, "mout_mfc", mout_mfc_p, SRC_MFC, 8, 1),
 	MUX_F(CLK_MOUT_G3D1, "mout_g3d1", sclk_evpll_p, SRC_G3D, 4, 1,
@@ -569,7 +569,7 @@ static struct samsung_mux_clock exynos4210_mux_clks[] __initdata = {
 	MUX(0, "mout_aclk100", sclk_ampll_p4210, SRC_TOP0, 16, 1),
 	MUX(0, "mout_aclk160", sclk_ampll_p4210, SRC_TOP0, 20, 1),
 	MUX(0, "mout_aclk133", sclk_ampll_p4210, SRC_TOP0, 24, 1),
-	MUX(0, "mout_mixer", mout_mixer_p4210, SRC_TV, 4, 1),
+	MUX(CLK_MOUT_MIXER, "mout_mixer", mout_mixer_p4210, SRC_TV, 4, 1),
 	MUX(0, "mout_dac", mout_dac_p4210, SRC_TV, 8, 1),
 	MUX(0, "mout_g2d0", sclk_ampll_p4210, E4210_SRC_IMAGE, 0, 1),
 	MUX(0, "mout_g2d1", sclk_evpll_p, E4210_SRC_IMAGE, 4, 1),
@@ -719,7 +719,7 @@ static struct samsung_div_clock exynos4_div_clks[] __initdata = {
 	DIV(0, "div_periph", "div_core2", DIV_CPU0, 12, 3),
 	DIV(0, "div_atb", "mout_core", DIV_CPU0, 16, 3),
 	DIV(0, "div_pclk_dbg", "div_atb", DIV_CPU0, 20, 3),
-	DIV(0, "div_core2", "div_core", DIV_CPU0, 28, 3),
+	DIV(CLK_ARM_CLK, "div_core2", "div_core", DIV_CPU0, 28, 3),
 	DIV(0, "div_copy", "mout_hpm", DIV_CPU1, 0, 3),
 	DIV(0, "div_hpm", "div_copy", DIV_CPU1, 4, 3),
 	DIV(0, "div_clkout_cpu", "mout_clkout_cpu", CLKOUT_CMU_CPU, 8, 6),
@@ -733,8 +733,7 @@ static struct samsung_div_clock exynos4_div_clks[] __initdata = {
 	DIV(0, "div_csis0", "mout_csis0", DIV_CAM, 24, 4),
 	DIV(0, "div_csis1", "mout_csis1", DIV_CAM, 28, 4),
 	DIV(CLK_SCLK_MFC, "sclk_mfc", "mout_mfc", DIV_MFC, 0, 4),
-	DIV_F(0, "div_g3d", "mout_g3d", DIV_G3D, 0, 4,
-			CLK_SET_RATE_PARENT, 0),
+	DIV(CLK_SCLK_G3D, "sclk_g3d", "mout_g3d", DIV_G3D, 0, 4),
 	DIV(0, "div_fimd0", "mout_fimd0", DIV_LCD0, 0, 4),
 	DIV(0, "div_mipi0", "mout_mipi0", DIV_LCD0, 16, 4),
 	DIV(0, "div_audio0", "mout_audio0", DIV_MAUDIO, 0, 4),
@@ -769,7 +768,6 @@ static struct samsung_div_clock exynos4_div_clks[] __initdata = {
 	DIV(0, "div_spi_pre2", "div_spi2", DIV_PERIL2, 8, 8),
 	DIV(0, "div_audio1", "mout_audio1", DIV_PERIL4, 0, 4),
 	DIV(0, "div_audio2", "mout_audio2", DIV_PERIL4, 16, 4),
-	DIV(CLK_ARM_CLK, "arm_clk", "div_core2", DIV_CPU0, 28, 3),
 	DIV(CLK_SCLK_APLL, "sclk_apll", "mout_apll", DIV_CPU0, 24, 3),
 	DIV_F(0, "div_mipi_pre0", "div_mipi0", DIV_LCD0, 20, 4,
 			CLK_SET_RATE_PARENT, 0),
@@ -857,8 +855,7 @@ static struct samsung_gate_clock exynos4_gate_clks[] __initdata = {
 		0),
 	GATE(CLK_TSI, "tsi", "aclk133", GATE_IP_FSYS, 4, 0, 0),
 	GATE(CLK_SROMC, "sromc", "aclk133", GATE_IP_FSYS, 11, 0, 0),
-	GATE(CLK_SCLK_G3D, "sclk_g3d", "div_g3d", GATE_IP_G3D, 0,
-			CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_G3D, "g3d", "aclk200", GATE_IP_G3D, 0, 0, 0),
 	GATE(CLK_PPMUG3D, "ppmug3d", "aclk200", GATE_IP_G3D, 1, 0, 0),
 	GATE(CLK_USB_DEVICE, "usb_device", "aclk133", GATE_IP_FSYS, 13, 0, 0),
 	GATE(CLK_ONENAND, "onenand", "aclk133", GATE_IP_FSYS, 15, 0, 0),
@@ -1183,6 +1180,7 @@ static struct samsung_gate_clock exynos4x12_gate_clks[] __initdata = {
 	GATE(CLK_SPI1_ISP, "spi1_isp", "aclk200", E4X12_GATE_ISP1, 13,
 			CLK_IGNORE_UNUSED | CLK_GET_RATE_NOCACHE, 0),
 	GATE(CLK_G2D, "g2d", "aclk200", GATE_IP_DMC, 23, 0, 0),
+	GATE(CLK_SMMU_G2D, "smmu_g2d", "aclk200", GATE_IP_DMC, 24, 0, 0),
 	GATE(CLK_TMU_APBIF, "tmu_apbif", "aclk100", E4X12_GATE_IP_PERIR, 17, 0,
 		0),
 };
@@ -1486,7 +1484,7 @@ static void __init exynos4_clk_init(struct device_node *np,
 		exynos4_soc == EXYNOS4210 ? "Exynos4210" : "Exynos4x12",
 		_get_rate("sclk_apll"),	_get_rate("sclk_mpll"),
 		_get_rate("sclk_epll"), _get_rate("sclk_vpll"),
-		_get_rate("arm_clk"));
+		_get_rate("div_core2"));
 }
 
 
diff --git a/drivers/clk/samsung/clk-exynos4415.c b/drivers/clk/samsung/clk-exynos4415.c
new file mode 100644
index 000000000000..2123fc251e0f
--- /dev/null
+++ b/drivers/clk/samsung/clk-exynos4415.c
@@ -0,0 +1,1144 @@
+/*
+ * Copyright (c) 2014 Samsung Electronics Co., Ltd.
+ * Author: Chanwoo Choi <cw00.choi@samsung.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * Common Clock Framework support for Exynos4415 SoC.
+ */
+
+#include <linux/clk.h>
+#include <linux/clkdev.h>
+#include <linux/clk-provider.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/platform_device.h>
+#include <linux/syscore_ops.h>
+
+#include <dt-bindings/clock/exynos4415.h>
+
+#include "clk.h"
+#include "clk-pll.h"
+
+#define SRC_LEFTBUS		0x4200
+#define DIV_LEFTBUS		0x4500
+#define GATE_IP_LEFTBUS		0x4800
+#define GATE_IP_IMAGE		0x4930
+#define SRC_RIGHTBUS		0x8200
+#define DIV_RIGHTBUS		0x8500
+#define GATE_IP_RIGHTBUS	0x8800
+#define GATE_IP_PERIR		0x8960
+#define EPLL_LOCK		0xc010
+#define G3D_PLL_LOCK		0xc020
+#define DISP_PLL_LOCK		0xc030
+#define ISP_PLL_LOCK		0xc040
+#define EPLL_CON0		0xc110
+#define EPLL_CON1		0xc114
+#define EPLL_CON2		0xc118
+#define G3D_PLL_CON0		0xc120
+#define G3D_PLL_CON1		0xc124
+#define G3D_PLL_CON2		0xc128
+#define ISP_PLL_CON0		0xc130
+#define ISP_PLL_CON1		0xc134
+#define ISP_PLL_CON2		0xc138
+#define DISP_PLL_CON0		0xc140
+#define DISP_PLL_CON1		0xc144
+#define DISP_PLL_CON2		0xc148
+#define SRC_TOP0		0xc210
+#define SRC_TOP1		0xc214
+#define SRC_CAM			0xc220
+#define SRC_TV			0xc224
+#define SRC_MFC			0xc228
+#define SRC_G3D			0xc22c
+#define SRC_LCD			0xc234
+#define SRC_ISP			0xc238
+#define SRC_MAUDIO		0xc23c
+#define SRC_FSYS		0xc240
+#define SRC_PERIL0		0xc250
+#define SRC_PERIL1		0xc254
+#define SRC_CAM1		0xc258
+#define SRC_TOP_ISP0		0xc25c
+#define SRC_TOP_ISP1		0xc260
+#define SRC_MASK_TOP		0xc310
+#define SRC_MASK_CAM		0xc320
+#define SRC_MASK_TV		0xc324
+#define SRC_MASK_LCD		0xc334
+#define SRC_MASK_ISP		0xc338
+#define SRC_MASK_MAUDIO		0xc33c
+#define SRC_MASK_FSYS		0xc340
+#define SRC_MASK_PERIL0		0xc350
+#define SRC_MASK_PERIL1		0xc354
+#define DIV_TOP			0xc510
+#define DIV_CAM			0xc520
+#define DIV_TV			0xc524
+#define DIV_MFC			0xc528
+#define DIV_G3D			0xc52c
+#define DIV_LCD			0xc534
+#define DIV_ISP			0xc538
+#define DIV_MAUDIO		0xc53c
+#define DIV_FSYS0		0xc540
+#define DIV_FSYS1		0xc544
+#define DIV_FSYS2		0xc548
+#define DIV_PERIL0		0xc550
+#define DIV_PERIL1		0xc554
+#define DIV_PERIL2		0xc558
+#define DIV_PERIL3		0xc55c
+#define DIV_PERIL4		0xc560
+#define DIV_PERIL5		0xc564
+#define DIV_CAM1		0xc568
+#define DIV_TOP_ISP1		0xc56c
+#define DIV_TOP_ISP0		0xc570
+#define CLKDIV2_RATIO		0xc580
+#define GATE_SCLK_CAM		0xc820
+#define GATE_SCLK_TV		0xc824
+#define GATE_SCLK_MFC		0xc828
+#define GATE_SCLK_G3D		0xc82c
+#define GATE_SCLK_LCD		0xc834
+#define GATE_SCLK_MAUDIO	0xc83c
+#define GATE_SCLK_FSYS		0xc840
+#define GATE_SCLK_PERIL		0xc850
+#define GATE_IP_CAM		0xc920
+#define GATE_IP_TV		0xc924
+#define GATE_IP_MFC		0xc928
+#define GATE_IP_G3D		0xc92c
+#define GATE_IP_LCD		0xc934
+#define GATE_IP_FSYS		0xc940
+#define GATE_IP_PERIL		0xc950
+#define GATE_BLOCK		0xc970
+#define APLL_LOCK		0x14000
+#define APLL_CON0		0x14100
+#define SRC_CPU			0x14200
+#define DIV_CPU0		0x14500
+#define DIV_CPU1		0x14504
+
+enum exynos4415_plls {
+	apll, epll, g3d_pll, isp_pll, disp_pll,
+	nr_plls,
+};
+
+static struct samsung_clk_provider *exynos4415_ctx;
+
+/*
+ * Support for CMU save/restore across system suspends
+ */
+#ifdef CONFIG_PM_SLEEP
+static struct samsung_clk_reg_dump *exynos4415_clk_regs;
+
+static unsigned long exynos4415_cmu_clk_regs[] __initdata = {
+	SRC_LEFTBUS,
+	DIV_LEFTBUS,
+	GATE_IP_LEFTBUS,
+	GATE_IP_IMAGE,
+	SRC_RIGHTBUS,
+	DIV_RIGHTBUS,
+	GATE_IP_RIGHTBUS,
+	GATE_IP_PERIR,
+	EPLL_LOCK,
+	G3D_PLL_LOCK,
+	DISP_PLL_LOCK,
+	ISP_PLL_LOCK,
+	EPLL_CON0,
+	EPLL_CON1,
+	EPLL_CON2,
+	G3D_PLL_CON0,
+	G3D_PLL_CON1,
+	G3D_PLL_CON2,
+	ISP_PLL_CON0,
+	ISP_PLL_CON1,
+	ISP_PLL_CON2,
+	DISP_PLL_CON0,
+	DISP_PLL_CON1,
+	DISP_PLL_CON2,
+	SRC_TOP0,
+	SRC_TOP1,
+	SRC_CAM,
+	SRC_TV,
+	SRC_MFC,
+	SRC_G3D,
+	SRC_LCD,
+	SRC_ISP,
+	SRC_MAUDIO,
+	SRC_FSYS,
+	SRC_PERIL0,
+	SRC_PERIL1,
+	SRC_CAM1,
+	SRC_TOP_ISP0,
+	SRC_TOP_ISP1,
+	SRC_MASK_TOP,
+	SRC_MASK_CAM,
+	SRC_MASK_TV,
+	SRC_MASK_LCD,
+	SRC_MASK_ISP,
+	SRC_MASK_MAUDIO,
+	SRC_MASK_FSYS,
+	SRC_MASK_PERIL0,
+	SRC_MASK_PERIL1,
+	DIV_TOP,
+	DIV_CAM,
+	DIV_TV,
+	DIV_MFC,
+	DIV_G3D,
+	DIV_LCD,
+	DIV_ISP,
+	DIV_MAUDIO,
+	DIV_FSYS0,
+	DIV_FSYS1,
+	DIV_FSYS2,
+	DIV_PERIL0,
+	DIV_PERIL1,
+	DIV_PERIL2,
+	DIV_PERIL3,
+	DIV_PERIL4,
+	DIV_PERIL5,
+	DIV_CAM1,
+	DIV_TOP_ISP1,
+	DIV_TOP_ISP0,
+	CLKDIV2_RATIO,
+	GATE_SCLK_CAM,
+	GATE_SCLK_TV,
+	GATE_SCLK_MFC,
+	GATE_SCLK_G3D,
+	GATE_SCLK_LCD,
+	GATE_SCLK_MAUDIO,
+	GATE_SCLK_FSYS,
+	GATE_SCLK_PERIL,
+	GATE_IP_CAM,
+	GATE_IP_TV,
+	GATE_IP_MFC,
+	GATE_IP_G3D,
+	GATE_IP_LCD,
+	GATE_IP_FSYS,
+	GATE_IP_PERIL,
+	GATE_BLOCK,
+	APLL_LOCK,
+	APLL_CON0,
+	SRC_CPU,
+	DIV_CPU0,
+	DIV_CPU1,
+};
+
+static int exynos4415_clk_suspend(void)
+{
+	samsung_clk_save(exynos4415_ctx->reg_base, exynos4415_clk_regs,
+				ARRAY_SIZE(exynos4415_cmu_clk_regs));
+
+	return 0;
+}
+
+static void exynos4415_clk_resume(void)
+{
+	samsung_clk_restore(exynos4415_ctx->reg_base, exynos4415_clk_regs,
+				ARRAY_SIZE(exynos4415_cmu_clk_regs));
+}
+
+static struct syscore_ops exynos4415_clk_syscore_ops = {
+	.suspend = exynos4415_clk_suspend,
+	.resume = exynos4415_clk_resume,
+};
+
+static void exynos4415_clk_sleep_init(void)
+{
+	exynos4415_clk_regs =
+		samsung_clk_alloc_reg_dump(exynos4415_cmu_clk_regs,
+					ARRAY_SIZE(exynos4415_cmu_clk_regs));
+	if (!exynos4415_clk_regs) {
+		pr_warn("%s: Failed to allocate sleep save data\n", __func__);
+		return;
+	}
+
+	register_syscore_ops(&exynos4415_clk_syscore_ops);
+}
+#else
+static inline void exynos4415_clk_sleep_init(void) { }
+#endif
+
+/* list of all parent clock list */
+PNAME(mout_g3d_pllsrc_p)	= { "fin_pll", };
+
+PNAME(mout_apll_p)		= { "fin_pll", "fout_apll", };
+PNAME(mout_g3d_pll_p)		= { "fin_pll", "fout_g3d_pll", };
+PNAME(mout_isp_pll_p)		= { "fin_pll", "fout_isp_pll", };
+PNAME(mout_disp_pll_p)		= { "fin_pll", "fout_disp_pll", };
+
+PNAME(mout_mpll_user_p)		= { "fin_pll", "div_mpll_pre", };
+PNAME(mout_epll_p)		= { "fin_pll", "fout_epll", };
+PNAME(mout_core_p)		= { "mout_apll", "mout_mpll_user_c", };
+PNAME(mout_hpm_p)		= { "mout_apll", "mout_mpll_user_c", };
+
+PNAME(mout_ebi_p)		= { "div_aclk_200", "div_aclk_160", };
+PNAME(mout_ebi_1_p)		= { "mout_ebi", "mout_g3d_pll", };
+
+PNAME(mout_gdl_p)		= { "mout_mpll_user_l", };
+PNAME(mout_gdr_p)		= { "mout_mpll_user_r", };
+
+PNAME(mout_aclk_266_p)		= { "mout_mpll_user_t", "mout_g3d_pll", };
+
+PNAME(group_epll_g3dpll_p)	= { "mout_epll", "mout_g3d_pll" };
+PNAME(group_sclk_p)		= { "xxti", "xusbxti",
+				    "none", "mout_isp_pll",
+				    "none", "none", "div_mpll_pre",
+				    "mout_epll", "mout_g3d_pll", };
+PNAME(group_spdif_p)		= { "mout_audio0", "mout_audio1",
+				    "mout_audio2", "spdif_extclk", };
+PNAME(group_sclk_audio2_p)	= { "audiocdclk2", "none",
+				    "none", "mout_isp_pll",
+				    "mout_disp_pll", "xusbxti",
+				    "div_mpll_pre", "mout_epll",
+				    "mout_g3d_pll", };
+PNAME(group_sclk_audio1_p)	= { "audiocdclk1", "none",
+				    "none", "mout_isp_pll",
+				    "mout_disp_pll", "xusbxti",
+				    "div_mpll_pre", "mout_epll",
+				    "mout_g3d_pll", };
+PNAME(group_sclk_audio0_p)	= { "audiocdclk0", "none",
+				    "none", "mout_isp_pll",
+				    "mout_disp_pll", "xusbxti",
+				    "div_mpll_pre", "mout_epll",
+				    "mout_g3d_pll", };
+PNAME(group_fimc_lclk_p)	= { "xxti", "xusbxti",
+				    "none", "mout_isp_pll",
+				    "none", "mout_disp_pll",
+				    "mout_mpll_user_t", "mout_epll",
+				    "mout_g3d_pll", };
+PNAME(group_sclk_fimd0_p)	= { "xxti", "xusbxti",
+				    "m_bitclkhsdiv4_4l", "mout_isp_pll",
+				    "mout_disp_pll", "sclk_hdmiphy",
+				    "div_mpll_pre", "mout_epll",
+				    "mout_g3d_pll", };
+PNAME(mout_hdmi_p)		= { "sclk_pixel", "sclk_hdmiphy" };
+PNAME(mout_mfc_p)		= { "mout_mfc_0", "mout_mfc_1" };
+PNAME(mout_g3d_p)		= { "mout_g3d_0", "mout_g3d_1" };
+PNAME(mout_jpeg_p)		= { "mout_jpeg_0", "mout_jpeg_1" };
+PNAME(mout_jpeg1_p)		= { "mout_epll", "mout_g3d_pll" };
+PNAME(group_aclk_isp0_300_p)	= { "mout_isp_pll", "div_mpll_pre" };
+PNAME(group_aclk_isp0_400_user_p) = { "fin_pll", "div_aclk_400_mcuisp" };
+PNAME(group_aclk_isp0_300_user_p) = { "fin_pll", "mout_aclk_isp0_300" };
+PNAME(group_aclk_isp1_300_user_p) = { "fin_pll", "mout_aclk_isp1_300" };
+PNAME(group_mout_mpll_user_t_p)	= { "mout_mpll_user_t" };
+
+static struct samsung_fixed_factor_clock exynos4415_fixed_factor_clks[] __initdata = {
+	/* HACK: fin_pll hardcoded to xusbxti until detection is implemented. */
+	FFACTOR(CLK_FIN_PLL, "fin_pll", "xusbxti", 1, 1, 0),
+};
+
+static struct samsung_fixed_rate_clock exynos4415_fixed_rate_clks[] __initdata = {
+	FRATE(CLK_SCLK_HDMIPHY, "sclk_hdmiphy", NULL, CLK_IS_ROOT, 27000000),
+};
+
+static struct samsung_mux_clock exynos4415_mux_clks[] __initdata = {
+	/*
+	 * NOTE: Following table is sorted by register address in ascending
+	 * order and then bitfield shift in descending order, as it is done
+	 * in the User's Manual. When adding new entries, please make sure
+	 * that the order is preserved, to avoid merge conflicts and make
+	 * further work with defined data easier.
+	 */
+
+	/* SRC_LEFTBUS */
+	MUX(CLK_MOUT_MPLL_USER_L, "mout_mpll_user_l", mout_mpll_user_p,
+		SRC_LEFTBUS, 4, 1),
+	MUX(CLK_MOUT_GDL, "mout_gdl", mout_gdl_p, SRC_LEFTBUS, 0, 1),
+
+	/* SRC_RIGHTBUS */
+	MUX(CLK_MOUT_MPLL_USER_R, "mout_mpll_user_r", mout_mpll_user_p,
+		SRC_RIGHTBUS, 4, 1),
+	MUX(CLK_MOUT_GDR, "mout_gdr", mout_gdr_p, SRC_RIGHTBUS, 0, 1),
+
+	/* SRC_TOP0 */
+	MUX(CLK_MOUT_EBI, "mout_ebi", mout_ebi_p, SRC_TOP0, 28, 1),
+	MUX(CLK_MOUT_ACLK_200, "mout_aclk_200", group_mout_mpll_user_t_p,
+		SRC_TOP0, 24, 1),
+	MUX(CLK_MOUT_ACLK_160, "mout_aclk_160", group_mout_mpll_user_t_p,
+		SRC_TOP0, 20, 1),
+	MUX(CLK_MOUT_ACLK_100, "mout_aclk_100", group_mout_mpll_user_t_p,
+		SRC_TOP0, 16, 1),
+	MUX(CLK_MOUT_ACLK_266, "mout_aclk_266", mout_aclk_266_p,
+		SRC_TOP0, 12, 1),
+	MUX(CLK_MOUT_G3D_PLL, "mout_g3d_pll", mout_g3d_pll_p,
+		SRC_TOP0, 8, 1),
+	MUX(CLK_MOUT_EPLL, "mout_epll", mout_epll_p, SRC_TOP0, 4, 1),
+	MUX(CLK_MOUT_EBI_1, "mout_ebi_1", mout_ebi_1_p, SRC_TOP0, 0, 1),
+
+	/* SRC_TOP1 */
+	MUX(CLK_MOUT_ISP_PLL, "mout_isp_pll", mout_isp_pll_p,
+		SRC_TOP1, 28, 1),
+	MUX(CLK_MOUT_DISP_PLL, "mout_disp_pll", mout_disp_pll_p,
+		SRC_TOP1, 16, 1),
+	MUX(CLK_MOUT_MPLL_USER_T, "mout_mpll_user_t", mout_mpll_user_p,
+		SRC_TOP1, 12, 1),
+	MUX(CLK_MOUT_ACLK_400_MCUISP, "mout_aclk_400_mcuisp",
+		group_mout_mpll_user_t_p, SRC_TOP1, 8, 1),
+	MUX(CLK_MOUT_G3D_PLLSRC, "mout_g3d_pllsrc", mout_g3d_pllsrc_p,
+		SRC_TOP1, 0, 1),
+
+	/* SRC_CAM */
+	MUX(CLK_MOUT_CSIS1, "mout_csis1", group_fimc_lclk_p, SRC_CAM, 28, 4),
+	MUX(CLK_MOUT_CSIS0, "mout_csis0", group_fimc_lclk_p, SRC_CAM, 24, 4),
+	MUX(CLK_MOUT_CAM1, "mout_cam1", group_fimc_lclk_p, SRC_CAM, 20, 4),
+	MUX(CLK_MOUT_FIMC3_LCLK, "mout_fimc3_lclk", group_fimc_lclk_p, SRC_CAM,
+		12, 4),
+	MUX(CLK_MOUT_FIMC2_LCLK, "mout_fimc2_lclk", group_fimc_lclk_p, SRC_CAM,
+		8, 4),
+	MUX(CLK_MOUT_FIMC1_LCLK, "mout_fimc1_lclk", group_fimc_lclk_p, SRC_CAM,
+		4, 4),
+	MUX(CLK_MOUT_FIMC0_LCLK, "mout_fimc0_lclk", group_fimc_lclk_p, SRC_CAM,
+		0, 4),
+
+	/* SRC_TV */
+	MUX(CLK_MOUT_HDMI, "mout_hdmi", mout_hdmi_p, SRC_TV, 0, 1),
+
+	/* SRC_MFC */
+	MUX(CLK_MOUT_MFC, "mout_mfc", mout_mfc_p, SRC_MFC, 8, 1),
+	MUX(CLK_MOUT_MFC_1, "mout_mfc_1", group_epll_g3dpll_p, SRC_MFC, 4, 1),
+	MUX(CLK_MOUT_MFC_0, "mout_mfc_0", group_mout_mpll_user_t_p, SRC_MFC, 0,
+		1),
+
+	/* SRC_G3D */
+	MUX(CLK_MOUT_G3D, "mout_g3d", mout_g3d_p, SRC_G3D, 8, 1),
+	MUX(CLK_MOUT_G3D_1, "mout_g3d_1", group_epll_g3dpll_p, SRC_G3D, 4, 1),
+	MUX(CLK_MOUT_G3D_0, "mout_g3d_0", group_mout_mpll_user_t_p, SRC_G3D, 0,
+		1),
+
+	/* SRC_LCD */
+	MUX(CLK_MOUT_MIPI0, "mout_mipi0", group_fimc_lclk_p, SRC_LCD, 12, 4),
+	MUX(CLK_MOUT_FIMD0, "mout_fimd0", group_sclk_fimd0_p, SRC_LCD, 0, 4),
+
+	/* SRC_ISP */
+	MUX(CLK_MOUT_TSADC_ISP, "mout_tsadc_isp", group_fimc_lclk_p, SRC_ISP,
+		16, 4),
+	MUX(CLK_MOUT_UART_ISP, "mout_uart_isp", group_fimc_lclk_p, SRC_ISP,
+		12, 4),
+	MUX(CLK_MOUT_SPI1_ISP, "mout_spi1_isp", group_fimc_lclk_p, SRC_ISP,
+		8, 4),
+	MUX(CLK_MOUT_SPI0_ISP, "mout_spi0_isp", group_fimc_lclk_p, SRC_ISP,
+		4, 4),
+	MUX(CLK_MOUT_PWM_ISP, "mout_pwm_isp", group_fimc_lclk_p, SRC_ISP,
+		0, 4),
+
+	/* SRC_MAUDIO */
+	MUX(CLK_MOUT_AUDIO0, "mout_audio0", group_sclk_audio0_p, SRC_MAUDIO,
+		0, 4),
+
+	/* SRC_FSYS */
+	MUX(CLK_MOUT_TSADC, "mout_tsadc", group_sclk_p, SRC_FSYS, 28, 4),
+	MUX(CLK_MOUT_MMC2, "mout_mmc2", group_sclk_p, SRC_FSYS, 8, 4),
+	MUX(CLK_MOUT_MMC1, "mout_mmc1", group_sclk_p, SRC_FSYS, 4, 4),
+	MUX(CLK_MOUT_MMC0, "mout_mmc0", group_sclk_p, SRC_FSYS, 0, 4),
+
+	/* SRC_PERIL0 */
+	MUX(CLK_MOUT_UART3, "mout_uart3", group_sclk_p, SRC_PERIL0, 12, 4),
+	MUX(CLK_MOUT_UART2, "mout_uart2", group_sclk_p, SRC_PERIL0, 8, 4),
+	MUX(CLK_MOUT_UART1, "mout_uart1", group_sclk_p, SRC_PERIL0, 4, 4),
+	MUX(CLK_MOUT_UART0, "mout_uart0", group_sclk_p, SRC_PERIL0, 0, 4),
+
+	/* SRC_PERIL1 */
+	MUX(CLK_MOUT_SPI2, "mout_spi2", group_sclk_p, SRC_PERIL1, 24, 4),
+	MUX(CLK_MOUT_SPI1, "mout_spi1", group_sclk_p, SRC_PERIL1, 20, 4),
+	MUX(CLK_MOUT_SPI0, "mout_spi0", group_sclk_p, SRC_PERIL1, 16, 4),
+	MUX(CLK_MOUT_SPDIF, "mout_spdif", group_spdif_p, SRC_PERIL1, 8, 4),
+	MUX(CLK_MOUT_AUDIO2, "mout_audio2", group_sclk_audio2_p, SRC_PERIL1,
+		4, 4),
+	MUX(CLK_MOUT_AUDIO1, "mout_audio1", group_sclk_audio1_p, SRC_PERIL1,
+		0, 4),
+
+	/* SRC_CPU */
+	MUX(CLK_MOUT_MPLL_USER_C, "mout_mpll_user_c", mout_mpll_user_p,
+		SRC_CPU, 24, 1),
+	MUX(CLK_MOUT_HPM, "mout_hpm", mout_hpm_p, SRC_CPU, 20, 1),
+	MUX_F(CLK_MOUT_CORE, "mout_core", mout_core_p, SRC_CPU, 16, 1, 0,
+		CLK_MUX_READ_ONLY),
+	MUX_F(CLK_MOUT_APLL, "mout_apll", mout_apll_p, SRC_CPU, 0, 1,
+		CLK_SET_RATE_PARENT, 0),
+
+	/* SRC_CAM1 */
+	MUX(CLK_MOUT_PXLASYNC_CSIS1_FIMC, "mout_pxlasync_csis1",
+		group_fimc_lclk_p, SRC_CAM1, 20, 1),
+	MUX(CLK_MOUT_PXLASYNC_CSIS0_FIMC, "mout_pxlasync_csis0",
+		group_fimc_lclk_p, SRC_CAM1, 16, 1),
+	MUX(CLK_MOUT_JPEG, "mout_jpeg", mout_jpeg_p, SRC_CAM1, 8, 1),
+	MUX(CLK_MOUT_JPEG1, "mout_jpeg_1", mout_jpeg1_p, SRC_CAM1, 4, 1),
+	MUX(CLK_MOUT_JPEG0, "mout_jpeg_0", group_mout_mpll_user_t_p, SRC_CAM1,
+		0, 1),
+
+	/* SRC_TOP_ISP0 */
+	MUX(CLK_MOUT_ACLK_ISP0_300, "mout_aclk_isp0_300",
+		group_aclk_isp0_300_p, SRC_TOP_ISP0, 8, 1),
+	MUX(CLK_MOUT_ACLK_ISP0_400, "mout_aclk_isp0_400_user",
+		group_aclk_isp0_400_user_p, SRC_TOP_ISP0, 4, 1),
+	MUX(CLK_MOUT_ACLK_ISP0_300_USER, "mout_aclk_isp0_300_user",
+		group_aclk_isp0_300_user_p, SRC_TOP_ISP0, 0, 1),
+
+	/* SRC_TOP_ISP1 */
+	MUX(CLK_MOUT_ACLK_ISP1_300, "mout_aclk_isp1_300",
+		group_aclk_isp0_300_p, SRC_TOP_ISP1, 4, 1),
+	MUX(CLK_MOUT_ACLK_ISP1_300_USER, "mout_aclk_isp1_300_user",
+		group_aclk_isp1_300_user_p, SRC_TOP_ISP1, 0, 1),
+};
+
+static struct samsung_div_clock exynos4415_div_clks[] __initdata = {
+	/*
+	 * NOTE: Following table is sorted by register address in ascending
+	 * order and then bitfield shift in descending order, as it is done
+	 * in the User's Manual. When adding new entries, please make sure
+	 * that the order is preserved, to avoid merge conflicts and make
+	 * further work with defined data easier.
+	 */
+
+	/* DIV_LEFTBUS */
+	DIV(CLK_DIV_GPL, "div_gpl", "div_gdl", DIV_LEFTBUS, 4, 3),
+	DIV(CLK_DIV_GDL, "div_gdl", "mout_gdl", DIV_LEFTBUS, 0, 4),
+
+	/* DIV_RIGHTBUS */
+	DIV(CLK_DIV_GPR, "div_gpr", "div_gdr", DIV_RIGHTBUS, 4, 3),
+	DIV(CLK_DIV_GDR, "div_gdr", "mout_gdr", DIV_RIGHTBUS, 0, 4),
+
+	/* DIV_TOP */
+	DIV(CLK_DIV_ACLK_400_MCUISP, "div_aclk_400_mcuisp",
+		"mout_aclk_400_mcuisp", DIV_TOP, 24, 3),
+	DIV(CLK_DIV_EBI, "div_ebi", "mout_ebi_1", DIV_TOP, 16, 3),
+	DIV(CLK_DIV_ACLK_200, "div_aclk_200", "mout_aclk_200", DIV_TOP, 12, 3),
+	DIV(CLK_DIV_ACLK_160, "div_aclk_160", "mout_aclk_160", DIV_TOP, 8, 3),
+	DIV(CLK_DIV_ACLK_100, "div_aclk_100", "mout_aclk_100", DIV_TOP, 4, 4),
+	DIV(CLK_DIV_ACLK_266, "div_aclk_266", "mout_aclk_266", DIV_TOP, 0, 3),
+
+	/* DIV_CAM */
+	DIV(CLK_DIV_CSIS1, "div_csis1", "mout_csis1", DIV_CAM, 28, 4),
+	DIV(CLK_DIV_CSIS0, "div_csis0", "mout_csis0", DIV_CAM, 24, 4),
+	DIV(CLK_DIV_CAM1, "div_cam1", "mout_cam1", DIV_CAM, 20, 4),
+	DIV(CLK_DIV_FIMC3_LCLK, "div_fimc3_lclk", "mout_fimc3_lclk", DIV_CAM,
+		12, 4),
+	DIV(CLK_DIV_FIMC2_LCLK, "div_fimc2_lclk", "mout_fimc2_lclk", DIV_CAM,
+		8, 4),
+	DIV(CLK_DIV_FIMC1_LCLK, "div_fimc1_lclk", "mout_fimc1_lclk", DIV_CAM,
+		4, 4),
+	DIV(CLK_DIV_FIMC0_LCLK, "div_fimc0_lclk", "mout_fimc0_lclk", DIV_CAM,
+		0, 4),
+
+	/* DIV_TV */
+	DIV(CLK_DIV_TV_BLK, "div_tv_blk", "mout_g3d_pll", DIV_TV, 0, 4),
+
+	/* DIV_MFC */
+	DIV(CLK_DIV_MFC, "div_mfc", "mout_mfc", DIV_MFC, 0, 4),
+
+	/* DIV_G3D */
+	DIV(CLK_DIV_G3D, "div_g3d", "mout_g3d", DIV_G3D, 0, 4),
+
+	/* DIV_LCD */
+	DIV_F(CLK_DIV_MIPI0_PRE, "div_mipi0_pre", "div_mipi0", DIV_LCD, 20, 4,
+		CLK_SET_RATE_PARENT, 0),
+	DIV(CLK_DIV_MIPI0, "div_mipi0", "mout_mipi0", DIV_LCD, 16, 4),
+	DIV(CLK_DIV_FIMD0, "div_fimd0", "mout_fimd0", DIV_LCD, 0, 4),
+
+	/* DIV_ISP */
+	DIV(CLK_DIV_UART_ISP, "div_uart_isp", "mout_uart_isp", DIV_ISP, 28, 4),
+	DIV_F(CLK_DIV_SPI1_ISP_PRE, "div_spi1_isp_pre", "div_spi1_isp",
+		DIV_ISP, 20, 8, CLK_SET_RATE_PARENT, 0),
+	DIV(CLK_DIV_SPI1_ISP, "div_spi1_isp", "mout_spi1_isp", DIV_ISP, 16, 4),
+	DIV_F(CLK_DIV_SPI0_ISP_PRE, "div_spi0_isp_pre", "div_spi0_isp",
+		DIV_ISP, 8, 8, CLK_SET_RATE_PARENT, 0),
+	DIV(CLK_DIV_SPI0_ISP, "div_spi0_isp", "mout_spi0_isp", DIV_ISP, 4, 4),
+	DIV(CLK_DIV_PWM_ISP, "div_pwm_isp", "mout_pwm_isp", DIV_ISP, 0, 4),
+
+	/* DIV_MAUDIO */
+	DIV(CLK_DIV_PCM0, "div_pcm0", "div_audio0", DIV_MAUDIO, 4, 8),
+	DIV(CLK_DIV_AUDIO0, "div_audio0", "mout_audio0", DIV_MAUDIO, 0, 4),
+
+	/* DIV_FSYS0 */
+	DIV_F(CLK_DIV_TSADC_PRE, "div_tsadc_pre", "div_tsadc", DIV_FSYS0, 8, 8,
+		CLK_SET_RATE_PARENT, 0),
+	DIV(CLK_DIV_TSADC, "div_tsadc", "mout_tsadc", DIV_FSYS0, 0, 4),
+
+	/* DIV_FSYS1 */
+	DIV_F(CLK_DIV_MMC1_PRE, "div_mmc1_pre", "div_mmc1", DIV_FSYS1, 24, 8,
+		CLK_SET_RATE_PARENT, 0),
+	DIV(CLK_DIV_MMC1, "div_mmc1", "mout_mmc1", DIV_FSYS1, 16, 4),
+	DIV_F(CLK_DIV_MMC0_PRE, "div_mmc0_pre", "div_mmc0", DIV_FSYS1, 8, 8,
+		CLK_SET_RATE_PARENT, 0),
+	DIV(CLK_DIV_MMC0, "div_mmc0", "mout_mmc0", DIV_FSYS1, 0, 4),
+
+	/* DIV_FSYS2 */
+	DIV_F(CLK_DIV_MMC2_PRE, "div_mmc2_pre", "div_mmc2", DIV_FSYS2, 8, 8,
+		CLK_SET_RATE_PARENT, 0),
+	DIV_F(CLK_DIV_MMC2_PRE, "div_mmc2", "mout_mmc2", DIV_FSYS2, 0, 4,
+		CLK_SET_RATE_PARENT, 0),
+
+	/* DIV_PERIL0 */
+	DIV(CLK_DIV_UART3, "div_uart3", "mout_uart3", DIV_PERIL0, 12, 4),
+	DIV(CLK_DIV_UART2, "div_uart2", "mout_uart2", DIV_PERIL0, 8, 4),
+	DIV(CLK_DIV_UART1, "div_uart1", "mout_uart1", DIV_PERIL0, 4, 4),
+	DIV(CLK_DIV_UART0, "div_uart0", "mout_uart0", DIV_PERIL0, 0, 4),
+
+	/* DIV_PERIL1 */
+	DIV_F(CLK_DIV_SPI1_PRE, "div_spi1_pre", "div_spi1", DIV_PERIL1, 24, 8,
+		CLK_SET_RATE_PARENT, 0),
+	DIV(CLK_DIV_SPI1, "div_spi1", "mout_spi1", DIV_PERIL1, 16, 4),
+	DIV_F(CLK_DIV_SPI0_PRE, "div_spi0_pre", "div_spi0", DIV_PERIL1, 8, 8,
+		CLK_SET_RATE_PARENT, 0),
+	DIV(CLK_DIV_SPI0, "div_spi0", "mout_spi0", DIV_PERIL1, 0, 4),
+
+	/* DIV_PERIL2 */
+	DIV_F(CLK_DIV_SPI2_PRE, "div_spi2_pre", "div_spi2", DIV_PERIL2, 8, 8,
+		CLK_SET_RATE_PARENT, 0),
+	DIV(CLK_DIV_SPI2, "div_spi2", "mout_spi2", DIV_PERIL2, 0, 4),
+
+	/* DIV_PERIL4 */
+	DIV(CLK_DIV_PCM2, "div_pcm2", "div_audio2", DIV_PERIL4, 20, 8),
+	DIV(CLK_DIV_AUDIO2, "div_audio2", "mout_audio2", DIV_PERIL4, 16, 4),
+	DIV(CLK_DIV_PCM1, "div_pcm1", "div_audio1", DIV_PERIL4, 20, 8),
+	DIV(CLK_DIV_AUDIO1, "div_audio1", "mout_audio1", DIV_PERIL4, 0, 4),
+
+	/* DIV_PERIL5 */
+	DIV(CLK_DIV_I2S1, "div_i2s1", "div_audio1", DIV_PERIL5, 0, 6),
+
+	/* DIV_CAM1 */
+	DIV(CLK_DIV_PXLASYNC_CSIS1_FIMC, "div_pxlasync_csis1_fimc",
+		"mout_pxlasync_csis1", DIV_CAM1, 24, 4),
+	DIV(CLK_DIV_PXLASYNC_CSIS0_FIMC, "div_pxlasync_csis0_fimc",
+		"mout_pxlasync_csis0", DIV_CAM1, 20, 4),
+	DIV(CLK_DIV_JPEG, "div_jpeg", "mout_jpeg", DIV_CAM1, 0, 4),
+
+	/* DIV_CPU0 */
+	DIV(CLK_DIV_CORE2, "div_core2", "div_core", DIV_CPU0, 28, 3),
+	DIV_F(CLK_DIV_APLL, "div_apll", "mout_apll", DIV_CPU0, 24, 3,
+			CLK_GET_RATE_NOCACHE, CLK_DIVIDER_READ_ONLY),
+	DIV(CLK_DIV_PCLK_DBG, "div_pclk_dbg", "div_core2", DIV_CPU0, 20, 3),
+	DIV(CLK_DIV_ATB, "div_atb", "div_core2", DIV_CPU0, 16, 3),
+	DIV(CLK_DIV_PERIPH, "div_periph", "div_core2", DIV_CPU0, 12, 3),
+	DIV(CLK_DIV_COREM1, "div_corem1", "div_core2", DIV_CPU0, 8, 3),
+	DIV(CLK_DIV_COREM0, "div_corem0", "div_core2", DIV_CPU0, 4, 3),
+	DIV_F(CLK_DIV_CORE, "div_core", "mout_core", DIV_CPU0, 0, 3,
+		CLK_GET_RATE_NOCACHE, CLK_DIVIDER_READ_ONLY),
+
+	/* DIV_CPU1 */
+	DIV(CLK_DIV_HPM, "div_hpm", "div_copy", DIV_CPU1, 4, 3),
+	DIV(CLK_DIV_COPY, "div_copy", "mout_hpm", DIV_CPU1, 0, 3),
+};
+
+static struct samsung_gate_clock exynos4415_gate_clks[] __initdata = {
+	/*
+	 * NOTE: Following table is sorted by register address in ascending
+	 * order and then bitfield shift in descending order, as it is done
+	 * in the User's Manual. When adding new entries, please make sure
+	 * that the order is preserved, to avoid merge conflicts and make
+	 * further work with defined data easier.
+	 */
+
+	/* GATE_IP_LEFTBUS */
+	GATE(CLK_ASYNC_G3D, "async_g3d", "div_aclk_100", GATE_IP_LEFTBUS, 6,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_ASYNC_MFCL, "async_mfcl", "div_aclk_100", GATE_IP_LEFTBUS, 4,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_ASYNC_TVX, "async_tvx", "div_aclk_100", GATE_IP_LEFTBUS, 3,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_PPMULEFT, "ppmuleft", "div_aclk_100", GATE_IP_LEFTBUS, 1,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_GPIO_LEFT, "gpio_left", "div_aclk_100", GATE_IP_LEFTBUS, 0,
+		CLK_IGNORE_UNUSED, 0),
+
+	/* GATE_IP_IMAGE */
+	GATE(CLK_PPMUIMAGE, "ppmuimage", "div_aclk_100", GATE_IP_IMAGE,
+		9, 0, 0),
+	GATE(CLK_QEMDMA2, "qe_mdma2", "div_aclk_100", GATE_IP_IMAGE,
+		8, 0, 0),
+	GATE(CLK_QEROTATOR, "qe_rotator", "div_aclk_100", GATE_IP_IMAGE,
+		7, 0, 0),
+	GATE(CLK_SMMUMDMA2, "smmu_mdam2", "div_aclk_100", GATE_IP_IMAGE,
+		5, 0, 0),
+	GATE(CLK_SMMUROTATOR, "smmu_rotator", "div_aclk_100", GATE_IP_IMAGE,
+		4, 0, 0),
+	GATE(CLK_MDMA2, "mdma2", "div_aclk_100", GATE_IP_IMAGE, 2, 0, 0),
+	GATE(CLK_ROTATOR, "rotator", "div_aclk_100", GATE_IP_IMAGE, 1, 0, 0),
+
+	/* GATE_IP_RIGHTBUS */
+	GATE(CLK_ASYNC_ISPMX, "async_ispmx", "div_aclk_100",
+		GATE_IP_RIGHTBUS, 9, CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_ASYNC_MAUDIOX, "async_maudiox", "div_aclk_100",
+		GATE_IP_RIGHTBUS, 7, CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_ASYNC_MFCR, "async_mfcr", "div_aclk_100",
+		GATE_IP_RIGHTBUS, 6, CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_ASYNC_FSYSD, "async_fsysd", "div_aclk_100",
+		GATE_IP_RIGHTBUS, 5, CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_ASYNC_LCD0X, "async_lcd0x", "div_aclk_100",
+		GATE_IP_RIGHTBUS, 3, CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_ASYNC_CAMX, "async_camx", "div_aclk_100",
+		GATE_IP_RIGHTBUS, 2, CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_PPMURIGHT, "ppmuright", "div_aclk_100",
+		GATE_IP_RIGHTBUS, 1, CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_GPIO_RIGHT, "gpio_right", "div_aclk_100",
+		GATE_IP_RIGHTBUS, 0, CLK_IGNORE_UNUSED, 0),
+
+	/* GATE_IP_PERIR */
+	GATE(CLK_ANTIRBK_APBIF, "antirbk_apbif", "div_aclk_100",
+		GATE_IP_PERIR, 24, CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_EFUSE_WRITER_APBIF, "efuse_writer_apbif", "div_aclk_100",
+		GATE_IP_PERIR, 23, CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_MONOCNT, "monocnt", "div_aclk_100", GATE_IP_PERIR, 22,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_TZPC6, "tzpc6", "div_aclk_100", GATE_IP_PERIR, 21,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_PROVISIONKEY1, "provisionkey1", "div_aclk_100",
+		GATE_IP_PERIR, 20, CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_PROVISIONKEY0, "provisionkey0", "div_aclk_100",
+		GATE_IP_PERIR, 19, CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_CMU_ISPPART, "cmu_isppart", "div_aclk_100", GATE_IP_PERIR, 18,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_TMU_APBIF, "tmu_apbif", "div_aclk_100",
+		GATE_IP_PERIR, 17, 0, 0),
+	GATE(CLK_KEYIF, "keyif", "div_aclk_100", GATE_IP_PERIR, 16, 0, 0),
+	GATE(CLK_RTC, "rtc", "div_aclk_100", GATE_IP_PERIR, 15, 0, 0),
+	GATE(CLK_WDT, "wdt", "div_aclk_100", GATE_IP_PERIR, 14, 0, 0),
+	GATE(CLK_MCT, "mct", "div_aclk_100", GATE_IP_PERIR, 13, 0, 0),
+	GATE(CLK_SECKEY, "seckey", "div_aclk_100", GATE_IP_PERIR, 12,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_HDMI_CEC, "hdmi_cec", "div_aclk_100", GATE_IP_PERIR, 11,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_TZPC5, "tzpc5", "div_aclk_100", GATE_IP_PERIR, 10,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_TZPC4, "tzpc4", "div_aclk_100", GATE_IP_PERIR, 9,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_TZPC3, "tzpc3", "div_aclk_100", GATE_IP_PERIR, 8,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_TZPC2, "tzpc2", "div_aclk_100", GATE_IP_PERIR, 7,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_TZPC1, "tzpc1", "div_aclk_100", GATE_IP_PERIR, 6,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_TZPC0, "tzpc0", "div_aclk_100", GATE_IP_PERIR, 5,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_CMU_COREPART, "cmu_corepart", "div_aclk_100", GATE_IP_PERIR, 4,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_CMU_TOPPART, "cmu_toppart", "div_aclk_100", GATE_IP_PERIR, 3,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_PMU_APBIF, "pmu_apbif", "div_aclk_100", GATE_IP_PERIR, 2,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_SYSREG, "sysreg", "div_aclk_100", GATE_IP_PERIR, 1,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_CHIP_ID, "chip_id", "div_aclk_100", GATE_IP_PERIR, 0,
+		CLK_IGNORE_UNUSED, 0),
+
+	/* GATE_SCLK_CAM - non-completed */
+	GATE(CLK_SCLK_PXLAYSNC_CSIS1_FIMC, "sclk_pxlasync_csis1_fimc",
+		"div_pxlasync_csis1_fimc", GATE_SCLK_CAM, 11,
+		CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_PXLAYSNC_CSIS0_FIMC, "sclk_pxlasync_csis0_fimc",
+		"div_pxlasync_csis0_fimc", GATE_SCLK_CAM,
+		10, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_JPEG, "sclk_jpeg", "div_jpeg",
+		GATE_SCLK_CAM, 8, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_CSIS1, "sclk_csis1", "div_csis1",
+		GATE_SCLK_CAM, 7, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_CSIS0, "sclk_csis0", "div_csis0",
+		GATE_SCLK_CAM, 6, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_CAM1, "sclk_cam1", "div_cam1",
+		GATE_SCLK_CAM, 5, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_FIMC3_LCLK, "sclk_fimc3_lclk", "div_fimc3_lclk",
+		GATE_SCLK_CAM, 3, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_FIMC2_LCLK, "sclk_fimc2_lclk", "div_fimc2_lclk",
+		GATE_SCLK_CAM, 2, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_FIMC1_LCLK, "sclk_fimc1_lclk", "div_fimc1_lclk",
+		GATE_SCLK_CAM, 1, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_FIMC0_LCLK, "sclk_fimc0_lclk", "div_fimc0_lclk",
+		GATE_SCLK_CAM, 0, CLK_SET_RATE_PARENT, 0),
+
+	/* GATE_SCLK_TV */
+	GATE(CLK_SCLK_PIXEL, "sclk_pixel", "div_tv_blk",
+		GATE_SCLK_TV, 3, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_HDMI, "sclk_hdmi", "mout_hdmi",
+		GATE_SCLK_TV, 2, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_MIXER, "sclk_mixer", "div_tv_blk",
+		GATE_SCLK_TV, 0, CLK_SET_RATE_PARENT, 0),
+
+	/* GATE_SCLK_MFC */
+	GATE(CLK_SCLK_MFC, "sclk_mfc", "div_mfc",
+		GATE_SCLK_MFC, 0, CLK_SET_RATE_PARENT, 0),
+
+	/* GATE_SCLK_G3D */
+	GATE(CLK_SCLK_G3D, "sclk_g3d", "div_g3d",
+		GATE_SCLK_G3D, 0, CLK_SET_RATE_PARENT, 0),
+
+	/* GATE_SCLK_LCD */
+	GATE(CLK_SCLK_MIPIDPHY4L, "sclk_mipidphy4l", "div_mipi0",
+		GATE_SCLK_LCD, 4, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_MIPI0, "sclk_mipi0", "div_mipi0_pre",
+		GATE_SCLK_LCD, 3, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_MDNIE0, "sclk_mdnie0", "div_fimd0",
+		GATE_SCLK_LCD, 1, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_FIMD0, "sclk_fimd0", "div_fimd0",
+		GATE_SCLK_LCD, 0, CLK_SET_RATE_PARENT, 0),
+
+	/* GATE_SCLK_MAUDIO */
+	GATE(CLK_SCLK_PCM0, "sclk_pcm0", "div_pcm0",
+		GATE_SCLK_MAUDIO, 1, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_AUDIO0, "sclk_audio0", "div_audio0",
+		GATE_SCLK_MAUDIO, 0, CLK_SET_RATE_PARENT, 0),
+
+	/* GATE_SCLK_FSYS */
+	GATE(CLK_SCLK_TSADC, "sclk_tsadc", "div_tsadc_pre",
+		GATE_SCLK_FSYS, 9, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_EBI, "sclk_ebi", "div_ebi",
+		GATE_SCLK_FSYS, 6, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_MMC2, "sclk_mmc2", "div_mmc2_pre",
+		GATE_SCLK_FSYS, 2, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_MMC1, "sclk_mmc1", "div_mmc1_pre",
+		GATE_SCLK_FSYS, 1, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_MMC0, "sclk_mmc0", "div_mmc0_pre",
+		GATE_SCLK_FSYS, 0, CLK_SET_RATE_PARENT, 0),
+
+	/* GATE_SCLK_PERIL */
+	GATE(CLK_SCLK_I2S, "sclk_i2s1", "div_i2s1",
+		GATE_SCLK_PERIL, 18, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_PCM2, "sclk_pcm2", "div_pcm2",
+		GATE_SCLK_PERIL, 16, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_PCM1, "sclk_pcm1", "div_pcm1",
+		GATE_SCLK_PERIL, 15, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_AUDIO2, "sclk_audio2", "div_audio2",
+		GATE_SCLK_PERIL, 14, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_AUDIO1, "sclk_audio1", "div_audio1",
+		GATE_SCLK_PERIL, 13, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_SPDIF, "sclk_spdif", "mout_spdif",
+		GATE_SCLK_PERIL, 10, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_SPI2, "sclk_spi2", "div_spi2_pre",
+		GATE_SCLK_PERIL, 8, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_SPI1, "sclk_spi1", "div_spi1_pre",
+		GATE_SCLK_PERIL, 7, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_SPI0, "sclk_spi0", "div_spi0_pre",
+		GATE_SCLK_PERIL, 6, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_UART3, "sclk_uart3", "div_uart3",
+		GATE_SCLK_PERIL, 3, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_UART2, "sclk_uart2", "div_uart2",
+		GATE_SCLK_PERIL, 2, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_UART1, "sclk_uart1", "div_uart1",
+		GATE_SCLK_PERIL, 1, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_UART0, "sclk_uart0", "div_uart0",
+		GATE_SCLK_PERIL, 0, CLK_SET_RATE_PARENT, 0),
+
+	/* GATE_IP_CAM */
+	GATE(CLK_SMMUFIMC_LITE2, "smmufimc_lite2", "div_aclk_160", GATE_IP_CAM,
+		22, CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_FIMC_LITE2, "fimc_lite2", "div_aclk_160", GATE_IP_CAM,
+		20, CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_PIXELASYNCM1, "pixelasyncm1", "div_aclk_160", GATE_IP_CAM,
+		18, CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_PIXELASYNCM0, "pixelasyncm0", "div_aclk_160", GATE_IP_CAM,
+		17, CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_PPMUCAMIF, "ppmucamif", "div_aclk_160", GATE_IP_CAM,
+		16, CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_SMMUJPEG, "smmujpeg", "div_aclk_160", GATE_IP_CAM, 11, 0, 0),
+	GATE(CLK_SMMUFIMC3, "smmufimc3", "div_aclk_160", GATE_IP_CAM, 10, 0, 0),
+	GATE(CLK_SMMUFIMC2, "smmufimc2", "div_aclk_160", GATE_IP_CAM, 9, 0, 0),
+	GATE(CLK_SMMUFIMC1, "smmufimc1", "div_aclk_160", GATE_IP_CAM, 8, 0, 0),
+	GATE(CLK_SMMUFIMC0, "smmufimc0", "div_aclk_160", GATE_IP_CAM, 7, 0, 0),
+	GATE(CLK_JPEG, "jpeg", "div_aclk_160", GATE_IP_CAM, 6, 0, 0),
+	GATE(CLK_CSIS1, "csis1", "div_aclk_160", GATE_IP_CAM, 5, 0, 0),
+	GATE(CLK_CSIS0, "csis0", "div_aclk_160", GATE_IP_CAM, 4, 0, 0),
+	GATE(CLK_FIMC3, "fimc3", "div_aclk_160", GATE_IP_CAM, 3, 0, 0),
+	GATE(CLK_FIMC2, "fimc2", "div_aclk_160", GATE_IP_CAM, 2, 0, 0),
+	GATE(CLK_FIMC1, "fimc1", "div_aclk_160", GATE_IP_CAM, 1, 0, 0),
+	GATE(CLK_FIMC0, "fimc0", "div_aclk_160", GATE_IP_CAM, 0, 0, 0),
+
+	/* GATE_IP_TV */
+	GATE(CLK_PPMUTV, "ppmutv", "div_aclk_100", GATE_IP_TV, 5, 0, 0),
+	GATE(CLK_SMMUTV, "smmutv", "div_aclk_100", GATE_IP_TV, 4, 0, 0),
+	GATE(CLK_HDMI, "hdmi", "div_aclk_100", GATE_IP_TV, 3, 0, 0),
+	GATE(CLK_MIXER, "mixer", "div_aclk_100", GATE_IP_TV, 1, 0, 0),
+	GATE(CLK_VP, "vp", "div_aclk_100", GATE_IP_TV, 0, 0, 0),
+
+	/* GATE_IP_MFC */
+	GATE(CLK_PPMUMFC_R, "ppmumfc_r", "div_aclk_200", GATE_IP_MFC, 4,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_PPMUMFC_L, "ppmumfc_l", "div_aclk_200", GATE_IP_MFC, 3,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_SMMUMFC_R, "smmumfc_r", "div_aclk_200", GATE_IP_MFC, 2, 0, 0),
+	GATE(CLK_SMMUMFC_L, "smmumfc_l", "div_aclk_200", GATE_IP_MFC, 1, 0, 0),
+	GATE(CLK_MFC, "mfc", "div_aclk_200", GATE_IP_MFC, 0, 0, 0),
+
+	/* GATE_IP_G3D */
+	GATE(CLK_PPMUG3D, "ppmug3d", "div_aclk_200", GATE_IP_G3D, 1,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_G3D, "g3d", "div_aclk_200", GATE_IP_G3D, 0, 0, 0),
+
+	/* GATE_IP_LCD */
+	GATE(CLK_PPMULCD0, "ppmulcd0", "div_aclk_160", GATE_IP_LCD, 5,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_SMMUFIMD0, "smmufimd0", "div_aclk_160", GATE_IP_LCD, 4, 0, 0),
+	GATE(CLK_DSIM0, "dsim0", "div_aclk_160", GATE_IP_LCD, 3, 0, 0),
+	GATE(CLK_SMIES, "smies", "div_aclk_160", GATE_IP_LCD, 2, 0, 0),
+	GATE(CLK_MIE0, "mie0", "div_aclk_160", GATE_IP_LCD, 1, 0, 0),
+	GATE(CLK_FIMD0, "fimd0", "div_aclk_160", GATE_IP_LCD, 0, 0, 0),
+
+	/* GATE_IP_FSYS */
+	GATE(CLK_TSADC, "tsadc", "div_aclk_200", GATE_IP_FSYS, 20, 0, 0),
+	GATE(CLK_PPMUFILE, "ppmufile", "div_aclk_200", GATE_IP_FSYS, 17,
+		CLK_IGNORE_UNUSED, 0),
+	GATE(CLK_NFCON, "nfcon", "div_aclk_200", GATE_IP_FSYS, 16, 0, 0),
+	GATE(CLK_USBDEVICE, "usbdevice", "div_aclk_200", GATE_IP_FSYS, 13,
+		0, 0),
+	GATE(CLK_USBHOST, "usbhost", "div_aclk_200", GATE_IP_FSYS, 12, 0, 0),
+	GATE(CLK_SROMC, "sromc", "div_aclk_200", GATE_IP_FSYS, 11, 0, 0),
+	GATE(CLK_SDMMC2, "sdmmc2", "div_aclk_200", GATE_IP_FSYS, 7, 0, 0),
+	GATE(CLK_SDMMC1, "sdmmc1", "div_aclk_200", GATE_IP_FSYS, 6, 0, 0),
+	GATE(CLK_SDMMC0, "sdmmc0", "div_aclk_200", GATE_IP_FSYS, 5, 0, 0),
+	GATE(CLK_PDMA1, "pdma1", "div_aclk_200", GATE_IP_FSYS, 1, 0, 0),
+	GATE(CLK_PDMA0, "pdma0", "div_aclk_200", GATE_IP_FSYS, 0, 0, 0),
+
+	/* GATE_IP_PERIL */
+	GATE(CLK_SPDIF, "spdif", "div_aclk_100", GATE_IP_PERIL, 26, 0, 0),
+	GATE(CLK_PWM, "pwm", "div_aclk_100", GATE_IP_PERIL, 24, 0, 0),
+	GATE(CLK_PCM2, "pcm2", "div_aclk_100", GATE_IP_PERIL, 23, 0, 0),
+	GATE(CLK_PCM1, "pcm1", "div_aclk_100", GATE_IP_PERIL, 22, 0, 0),
+	GATE(CLK_I2S1, "i2s1", "div_aclk_100", GATE_IP_PERIL, 20, 0, 0),
+	GATE(CLK_SPI2, "spi2", "div_aclk_100", GATE_IP_PERIL, 18, 0, 0),
+	GATE(CLK_SPI1, "spi1", "div_aclk_100", GATE_IP_PERIL, 17, 0, 0),
+	GATE(CLK_SPI0, "spi0", "div_aclk_100", GATE_IP_PERIL, 16, 0, 0),
+	GATE(CLK_I2CHDMI, "i2chdmi", "div_aclk_100", GATE_IP_PERIL, 14, 0, 0),
+	GATE(CLK_I2C7, "i2c7", "div_aclk_100", GATE_IP_PERIL, 13, 0, 0),
+	GATE(CLK_I2C6, "i2c6", "div_aclk_100", GATE_IP_PERIL, 12, 0, 0),
+	GATE(CLK_I2C5, "i2c5", "div_aclk_100", GATE_IP_PERIL, 11, 0, 0),
+	GATE(CLK_I2C4, "i2c4", "div_aclk_100", GATE_IP_PERIL, 10, 0, 0),
+	GATE(CLK_I2C3, "i2c3", "div_aclk_100", GATE_IP_PERIL, 9, 0, 0),
+	GATE(CLK_I2C2, "i2c2", "div_aclk_100", GATE_IP_PERIL, 8, 0, 0),
+	GATE(CLK_I2C1, "i2c1", "div_aclk_100", GATE_IP_PERIL, 7, 0, 0),
+	GATE(CLK_I2C0, "i2c0", "div_aclk_100", GATE_IP_PERIL, 6, 0, 0),
+	GATE(CLK_UART3, "uart3", "div_aclk_100", GATE_IP_PERIL, 3, 0, 0),
+	GATE(CLK_UART2, "uart2", "div_aclk_100", GATE_IP_PERIL, 2, 0, 0),
+	GATE(CLK_UART1, "uart1", "div_aclk_100", GATE_IP_PERIL, 1, 0, 0),
+	GATE(CLK_UART0, "uart0", "div_aclk_100", GATE_IP_PERIL, 0, 0, 0),
+};
+
+/*
+ * APLL & MPLL & BPLL & ISP_PLL & DISP_PLL & G3D_PLL
+ */
+static struct samsung_pll_rate_table exynos4415_pll_rates[] = {
+	PLL_35XX_RATE(1600000000, 400, 3,  1),
+	PLL_35XX_RATE(1500000000, 250, 2,  1),
+	PLL_35XX_RATE(1400000000, 175, 3,  0),
+	PLL_35XX_RATE(1300000000, 325, 3,  1),
+	PLL_35XX_RATE(1200000000, 400, 4,  1),
+	PLL_35XX_RATE(1100000000, 275, 3,  1),
+	PLL_35XX_RATE(1066000000, 533, 6,  1),
+	PLL_35XX_RATE(1000000000, 250, 3,  1),
+	PLL_35XX_RATE(960000000,  320, 4,  1),
+	PLL_35XX_RATE(900000000,  300, 4,  1),
+	PLL_35XX_RATE(850000000,  425, 6,  1),
+	PLL_35XX_RATE(800000000,  200, 3,  1),
+	PLL_35XX_RATE(700000000,  175, 3,  1),
+	PLL_35XX_RATE(667000000,  667, 12, 1),
+	PLL_35XX_RATE(600000000,  400, 4,  2),
+	PLL_35XX_RATE(550000000,  275, 3,  2),
+	PLL_35XX_RATE(533000000,  533, 6,  2),
+	PLL_35XX_RATE(520000000,  260, 3,  2),
+	PLL_35XX_RATE(500000000,  250, 3,  2),
+	PLL_35XX_RATE(440000000,  220, 3,  2),
+	PLL_35XX_RATE(400000000,  200, 3,  2),
+	PLL_35XX_RATE(350000000,  175, 3,  2),
+	PLL_35XX_RATE(300000000,  300, 3,  3),
+	PLL_35XX_RATE(266000000,  266, 3,  3),
+	PLL_35XX_RATE(200000000,  200, 3,  3),
+	PLL_35XX_RATE(160000000,  160, 3,  3),
+	PLL_35XX_RATE(100000000,  200, 3,  4),
+	{ /* sentinel */ }
+};
+
+/* EPLL */
+static struct samsung_pll_rate_table exynos4415_epll_rates[] = {
+	PLL_36XX_RATE(800000000, 200, 3, 1,     0),
+	PLL_36XX_RATE(288000000,  96, 2, 2,     0),
+	PLL_36XX_RATE(192000000, 128, 2, 3,     0),
+	PLL_36XX_RATE(144000000,  96, 2, 3,     0),
+	PLL_36XX_RATE(96000000,  128, 2, 4,     0),
+	PLL_36XX_RATE(84000000,  112, 2, 4,     0),
+	PLL_36XX_RATE(80750011,  107, 2, 4, 43691),
+	PLL_36XX_RATE(73728004,   98, 2, 4, 19923),
+	PLL_36XX_RATE(67987602,  271, 3, 5, 62285),
+	PLL_36XX_RATE(65911004,  175, 2, 5, 49982),
+	PLL_36XX_RATE(50000000,  200, 3, 5,     0),
+	PLL_36XX_RATE(49152003,  131, 2, 5,  4719),
+	PLL_36XX_RATE(48000000,  128, 2, 5,     0),
+	PLL_36XX_RATE(45250000,  181, 3, 5,     0),
+	{ /* sentinel */ }
+};
+
+static struct samsung_pll_clock exynos4415_plls[nr_plls] __initdata = {
+	[apll] = PLL(pll_35xx, CLK_FOUT_APLL, "fout_apll", "fin_pll",
+			APLL_LOCK, APLL_CON0, NULL),
+	[epll] = PLL(pll_36xx, CLK_FOUT_EPLL, "fout_epll", "fin_pll",
+			EPLL_LOCK, EPLL_CON0, NULL),
+	[g3d_pll] = PLL(pll_35xx, CLK_FOUT_G3D_PLL, "fout_g3d_pll",
+			"mout_g3d_pllsrc", G3D_PLL_LOCK, G3D_PLL_CON0, NULL),
+	[isp_pll] = PLL(pll_35xx, CLK_FOUT_ISP_PLL, "fout_isp_pll", "fin_pll",
+			ISP_PLL_LOCK, ISP_PLL_CON0, NULL),
+	[disp_pll] = PLL(pll_35xx, CLK_FOUT_DISP_PLL, "fout_disp_pll",
+			"fin_pll", DISP_PLL_LOCK, DISP_PLL_CON0, NULL),
+};
+
+static void __init exynos4415_cmu_init(struct device_node *np)
+{
+	void __iomem *reg_base;
+
+	reg_base = of_iomap(np, 0);
+	if (!reg_base)
+		panic("%s: failed to map registers\n", __func__);
+
+	exynos4415_ctx = samsung_clk_init(np, reg_base, CLK_NR_CLKS);
+	if (!exynos4415_ctx)
+		panic("%s: unable to allocate context.\n", __func__);
+
+	exynos4415_plls[apll].rate_table = exynos4415_pll_rates;
+	exynos4415_plls[epll].rate_table = exynos4415_epll_rates;
+	exynos4415_plls[g3d_pll].rate_table = exynos4415_pll_rates;
+	exynos4415_plls[isp_pll].rate_table = exynos4415_pll_rates;
+	exynos4415_plls[disp_pll].rate_table = exynos4415_pll_rates;
+
+	samsung_clk_register_fixed_factor(exynos4415_ctx,
+				exynos4415_fixed_factor_clks,
+				ARRAY_SIZE(exynos4415_fixed_factor_clks));
+	samsung_clk_register_fixed_rate(exynos4415_ctx,
+				exynos4415_fixed_rate_clks,
+				ARRAY_SIZE(exynos4415_fixed_rate_clks));
+
+	samsung_clk_register_pll(exynos4415_ctx, exynos4415_plls,
+				ARRAY_SIZE(exynos4415_plls), reg_base);
+	samsung_clk_register_mux(exynos4415_ctx, exynos4415_mux_clks,
+				ARRAY_SIZE(exynos4415_mux_clks));
+	samsung_clk_register_div(exynos4415_ctx, exynos4415_div_clks,
+				ARRAY_SIZE(exynos4415_div_clks));
+	samsung_clk_register_gate(exynos4415_ctx, exynos4415_gate_clks,
+				ARRAY_SIZE(exynos4415_gate_clks));
+
+	exynos4415_clk_sleep_init();
+
+	samsung_clk_of_add_provider(np, exynos4415_ctx);
+}
+CLK_OF_DECLARE(exynos4415_cmu, "samsung,exynos4415-cmu", exynos4415_cmu_init);
+
+/*
+ * CMU DMC
+ */
+
+#define MPLL_LOCK		0x008
+#define MPLL_CON0		0x108
+#define MPLL_CON1		0x10c
+#define MPLL_CON2		0x110
+#define BPLL_LOCK		0x118
+#define BPLL_CON0		0x218
+#define BPLL_CON1		0x21c
+#define BPLL_CON2		0x220
+#define SRC_DMC			0x300
+#define DIV_DMC1		0x504
+
+enum exynos4415_dmc_plls {
+	mpll, bpll,
+	nr_dmc_plls,
+};
+
+static struct samsung_clk_provider *exynos4415_dmc_ctx;
+
+#ifdef CONFIG_PM_SLEEP
+static struct samsung_clk_reg_dump *exynos4415_dmc_clk_regs;
+
+static unsigned long exynos4415_cmu_dmc_clk_regs[] __initdata = {
+	MPLL_LOCK,
+	MPLL_CON0,
+	MPLL_CON1,
+	MPLL_CON2,
+	BPLL_LOCK,
+	BPLL_CON0,
+	BPLL_CON1,
+	BPLL_CON2,
+	SRC_DMC,
+	DIV_DMC1,
+};
+
+static int exynos4415_dmc_clk_suspend(void)
+{
+	samsung_clk_save(exynos4415_dmc_ctx->reg_base,
+				exynos4415_dmc_clk_regs,
+				ARRAY_SIZE(exynos4415_cmu_dmc_clk_regs));
+	return 0;
+}
+
+static void exynos4415_dmc_clk_resume(void)
+{
+	samsung_clk_restore(exynos4415_dmc_ctx->reg_base,
+				exynos4415_dmc_clk_regs,
+				ARRAY_SIZE(exynos4415_cmu_dmc_clk_regs));
+}
+
+static struct syscore_ops exynos4415_dmc_clk_syscore_ops = {
+	.suspend = exynos4415_dmc_clk_suspend,
+	.resume = exynos4415_dmc_clk_resume,
+};
+
+static void exynos4415_dmc_clk_sleep_init(void)
+{
+	exynos4415_dmc_clk_regs =
+		samsung_clk_alloc_reg_dump(exynos4415_cmu_dmc_clk_regs,
+				ARRAY_SIZE(exynos4415_cmu_dmc_clk_regs));
+	if (!exynos4415_dmc_clk_regs) {
+		pr_warn("%s: Failed to allocate sleep save data\n", __func__);
+		return;
+	}
+
+	register_syscore_ops(&exynos4415_dmc_clk_syscore_ops);
+}
+#else
+static inline void exynos4415_dmc_clk_sleep_init(void) { }
+#endif /* CONFIG_PM_SLEEP */
+
+PNAME(mout_mpll_p)		= { "fin_pll", "fout_mpll", };
+PNAME(mout_bpll_p)		= { "fin_pll", "fout_bpll", };
+PNAME(mbpll_p)			= { "mout_mpll", "mout_bpll", };
+
+static struct samsung_mux_clock exynos4415_dmc_mux_clks[] __initdata = {
+	MUX(CLK_DMC_MOUT_MPLL, "mout_mpll", mout_mpll_p, SRC_DMC, 12, 1),
+	MUX(CLK_DMC_MOUT_BPLL, "mout_bpll", mout_bpll_p, SRC_DMC, 10, 1),
+	MUX(CLK_DMC_MOUT_DPHY, "mout_dphy", mbpll_p, SRC_DMC, 8, 1),
+	MUX(CLK_DMC_MOUT_DMC_BUS, "mout_dmc_bus", mbpll_p, SRC_DMC, 4, 1),
+};
+
+static struct samsung_div_clock exynos4415_dmc_div_clks[] __initdata = {
+	DIV(CLK_DMC_DIV_DMC, "div_dmc", "div_dmc_pre", DIV_DMC1, 27, 3),
+	DIV(CLK_DMC_DIV_DPHY, "div_dphy", "mout_dphy", DIV_DMC1, 23, 3),
+	DIV(CLK_DMC_DIV_DMC_PRE, "div_dmc_pre", "mout_dmc_bus",
+		DIV_DMC1, 19, 2),
+	DIV(CLK_DMC_DIV_DMCP, "div_dmcp", "div_dmcd", DIV_DMC1, 15, 3),
+	DIV(CLK_DMC_DIV_DMCD, "div_dmcd", "div_dmc", DIV_DMC1, 11, 3),
+	DIV(CLK_DMC_DIV_MPLL_PRE, "div_mpll_pre", "mout_mpll", DIV_DMC1, 8, 2),
+};
+
+static struct samsung_pll_clock exynos4415_dmc_plls[nr_dmc_plls] __initdata = {
+	[mpll] = PLL(pll_35xx, CLK_DMC_FOUT_MPLL, "fout_mpll", "fin_pll",
+		MPLL_LOCK, MPLL_CON0, NULL),
+	[bpll] = PLL(pll_35xx, CLK_DMC_FOUT_BPLL, "fout_bpll", "fin_pll",
+		BPLL_LOCK, BPLL_CON0, NULL),
+};
+
+static void __init exynos4415_cmu_dmc_init(struct device_node *np)
+{
+	void __iomem *reg_base;
+
+	reg_base = of_iomap(np, 0);
+	if (!reg_base)
+		panic("%s: failed to map registers\n", __func__);
+
+	exynos4415_dmc_ctx = samsung_clk_init(np, reg_base, NR_CLKS_DMC);
+	if (!exynos4415_dmc_ctx)
+		panic("%s: unable to allocate context.\n", __func__);
+
+	exynos4415_dmc_plls[mpll].rate_table = exynos4415_pll_rates;
+	exynos4415_dmc_plls[bpll].rate_table = exynos4415_pll_rates;
+
+	samsung_clk_register_pll(exynos4415_dmc_ctx, exynos4415_dmc_plls,
+				ARRAY_SIZE(exynos4415_dmc_plls), reg_base);
+	samsung_clk_register_mux(exynos4415_dmc_ctx, exynos4415_dmc_mux_clks,
+				ARRAY_SIZE(exynos4415_dmc_mux_clks));
+	samsung_clk_register_div(exynos4415_dmc_ctx, exynos4415_dmc_div_clks,
+				ARRAY_SIZE(exynos4415_dmc_div_clks));
+
+	exynos4415_dmc_clk_sleep_init();
+
+	samsung_clk_of_add_provider(np, exynos4415_dmc_ctx);
+}
+CLK_OF_DECLARE(exynos4415_cmu_dmc, "samsung,exynos4415-cmu-dmc",
+		exynos4415_cmu_dmc_init);
diff --git a/drivers/clk/samsung/clk-exynos5260.c b/drivers/clk/samsung/clk-exynos5260.c
index ce3de97e5f11..e2e5193d1049 100644
--- a/drivers/clk/samsung/clk-exynos5260.c
+++ b/drivers/clk/samsung/clk-exynos5260.c
@@ -11,10 +11,8 @@
 
 #include <linux/clk.h>
 #include <linux/clkdev.h>
-#include <linux/clk-provider.h>
 #include <linux/of.h>
 #include <linux/of_address.h>
-#include <linux/syscore_ops.h>
 
 #include "clk-exynos5260.h"
 #include "clk.h"
@@ -22,39 +20,6 @@
 
 #include <dt-bindings/clock/exynos5260-clk.h>
 
-static LIST_HEAD(clock_reg_cache_list);
-
-struct exynos5260_clock_reg_cache {
-	struct list_head node;
-	void __iomem *reg_base;
-	struct samsung_clk_reg_dump *rdump;
-	unsigned int rd_num;
-};
-
-struct exynos5260_cmu_info {
-	/* list of pll clocks and respective count */
-	struct samsung_pll_clock *pll_clks;
-	unsigned int nr_pll_clks;
-	/* list of mux clocks and respective count */
-	struct samsung_mux_clock *mux_clks;
-	unsigned int nr_mux_clks;
-	/* list of div clocks and respective count */
-	struct samsung_div_clock *div_clks;
-	unsigned int nr_div_clks;
-	/* list of gate clocks and respective count */
-	struct samsung_gate_clock *gate_clks;
-	unsigned int nr_gate_clks;
-	/* list of fixed clocks and respective count */
-	struct samsung_fixed_rate_clock *fixed_clks;
-	unsigned int nr_fixed_clks;
-	/* total number of clocks with IDs assigned*/
-	unsigned int nr_clk_ids;
-
-	/* list and number of clocks registers */
-	unsigned long *clk_regs;
-	unsigned int nr_clk_regs;
-};
-
 /*
  * Applicable for all 2550 Type PLLS for Exynos5260, listed below
  * DISP_PLL, EGL_PLL, KFC_PLL, MEM_PLL, BUS_PLL, MEDIA_PLL, G3D_PLL.
@@ -113,104 +78,6 @@ static struct samsung_pll_rate_table pll2650_24mhz_tbl[] __initdata = {
 	PLL_36XX_RATE(66000000, 176, 2, 5, 0),
 };
 
-#ifdef CONFIG_PM_SLEEP
-
-static int exynos5260_clk_suspend(void)
-{
-	struct exynos5260_clock_reg_cache *cache;
-
-	list_for_each_entry(cache, &clock_reg_cache_list, node)
-		samsung_clk_save(cache->reg_base, cache->rdump,
-				cache->rd_num);
-
-	return 0;
-}
-
-static void exynos5260_clk_resume(void)
-{
-	struct exynos5260_clock_reg_cache *cache;
-
-	list_for_each_entry(cache, &clock_reg_cache_list, node)
-		samsung_clk_restore(cache->reg_base, cache->rdump,
-				cache->rd_num);
-}
-
-static struct syscore_ops exynos5260_clk_syscore_ops = {
-	.suspend = exynos5260_clk_suspend,
-	.resume = exynos5260_clk_resume,
-};
-
-static void exynos5260_clk_sleep_init(void __iomem *reg_base,
-			unsigned long *rdump,
-			unsigned long nr_rdump)
-{
-	struct exynos5260_clock_reg_cache *reg_cache;
-
-	reg_cache = kzalloc(sizeof(struct exynos5260_clock_reg_cache),
-			GFP_KERNEL);
-	if (!reg_cache)
-		panic("could not allocate register cache.\n");
-
-	reg_cache->rdump = samsung_clk_alloc_reg_dump(rdump, nr_rdump);
-
-	if (!reg_cache->rdump)
-		panic("could not allocate register dump storage.\n");
-
-	if (list_empty(&clock_reg_cache_list))
-		register_syscore_ops(&exynos5260_clk_syscore_ops);
-
-	reg_cache->rd_num = nr_rdump;
-	reg_cache->reg_base = reg_base;
-	list_add_tail(&reg_cache->node, &clock_reg_cache_list);
-}
-
-#else
-static void exynos5260_clk_sleep_init(void __iomem *reg_base,
-			unsigned long *rdump,
-			unsigned long nr_rdump){}
-#endif
-
-/*
- * Common function which registers plls, muxes, dividers and gates
- * for each CMU. It also add CMU register list to register cache.
- */
-
-void __init exynos5260_cmu_register_one(struct device_node *np,
-			struct exynos5260_cmu_info *cmu)
-{
-	void __iomem *reg_base;
-	struct samsung_clk_provider *ctx;
-
-	reg_base = of_iomap(np, 0);
-	if (!reg_base)
-		panic("%s: failed to map registers\n", __func__);
-
-	ctx = samsung_clk_init(np, reg_base, cmu->nr_clk_ids);
-	if (!ctx)
-		panic("%s: unable to alllocate ctx\n", __func__);
-
-	if (cmu->pll_clks)
-		samsung_clk_register_pll(ctx, cmu->pll_clks, cmu->nr_pll_clks,
-			reg_base);
-	if (cmu->mux_clks)
-		samsung_clk_register_mux(ctx,  cmu->mux_clks,
-			cmu->nr_mux_clks);
-	if (cmu->div_clks)
-		samsung_clk_register_div(ctx, cmu->div_clks, cmu->nr_div_clks);
-	if (cmu->gate_clks)
-		samsung_clk_register_gate(ctx, cmu->gate_clks,
-			cmu->nr_gate_clks);
-	if (cmu->fixed_clks)
-		samsung_clk_register_fixed_rate(ctx, cmu->fixed_clks,
-			cmu->nr_fixed_clks);
-	if (cmu->clk_regs)
-		exynos5260_clk_sleep_init(reg_base, cmu->clk_regs,
-			cmu->nr_clk_regs);
-
-	samsung_clk_of_add_provider(np, ctx);
-}
-
-
 /* CMU_AUD */
 
 static unsigned long aud_clk_regs[] __initdata = {
@@ -268,7 +135,7 @@ struct samsung_gate_clock aud_gate_clks[] __initdata = {
 
 static void __init exynos5260_clk_aud_init(struct device_node *np)
 {
-	struct exynos5260_cmu_info cmu = {0};
+	struct samsung_cmu_info cmu = {0};
 
 	cmu.mux_clks = aud_mux_clks;
 	cmu.nr_mux_clks = ARRAY_SIZE(aud_mux_clks);
@@ -280,7 +147,7 @@ static void __init exynos5260_clk_aud_init(struct device_node *np)
 	cmu.clk_regs = aud_clk_regs;
 	cmu.nr_clk_regs = ARRAY_SIZE(aud_clk_regs);
 
-	exynos5260_cmu_register_one(np, &cmu);
+	samsung_cmu_register_one(np, &cmu);
 }
 
 CLK_OF_DECLARE(exynos5260_clk_aud, "samsung,exynos5260-clock-aud",
@@ -458,7 +325,7 @@ struct samsung_gate_clock disp_gate_clks[] __initdata = {
 
 static void __init exynos5260_clk_disp_init(struct device_node *np)
 {
-	struct exynos5260_cmu_info cmu = {0};
+	struct samsung_cmu_info cmu = {0};
 
 	cmu.mux_clks = disp_mux_clks;
 	cmu.nr_mux_clks = ARRAY_SIZE(disp_mux_clks);
@@ -470,7 +337,7 @@ static void __init exynos5260_clk_disp_init(struct device_node *np)
 	cmu.clk_regs = disp_clk_regs;
 	cmu.nr_clk_regs = ARRAY_SIZE(disp_clk_regs);
 
-	exynos5260_cmu_register_one(np, &cmu);
+	samsung_cmu_register_one(np, &cmu);
 }
 
 CLK_OF_DECLARE(exynos5260_clk_disp, "samsung,exynos5260-clock-disp",
@@ -522,7 +389,7 @@ static struct samsung_pll_clock egl_pll_clks[] __initdata = {
 
 static void __init exynos5260_clk_egl_init(struct device_node *np)
 {
-	struct exynos5260_cmu_info cmu = {0};
+	struct samsung_cmu_info cmu = {0};
 
 	cmu.pll_clks = egl_pll_clks;
 	cmu.nr_pll_clks =  ARRAY_SIZE(egl_pll_clks);
@@ -534,7 +401,7 @@ static void __init exynos5260_clk_egl_init(struct device_node *np)
 	cmu.clk_regs = egl_clk_regs;
 	cmu.nr_clk_regs = ARRAY_SIZE(egl_clk_regs);
 
-	exynos5260_cmu_register_one(np, &cmu);
+	samsung_cmu_register_one(np, &cmu);
 }
 
 CLK_OF_DECLARE(exynos5260_clk_egl, "samsung,exynos5260-clock-egl",
@@ -624,7 +491,7 @@ struct samsung_gate_clock fsys_gate_clks[] __initdata = {
 
 static void __init exynos5260_clk_fsys_init(struct device_node *np)
 {
-	struct exynos5260_cmu_info cmu = {0};
+	struct samsung_cmu_info cmu = {0};
 
 	cmu.mux_clks = fsys_mux_clks;
 	cmu.nr_mux_clks = ARRAY_SIZE(fsys_mux_clks);
@@ -634,7 +501,7 @@ static void __init exynos5260_clk_fsys_init(struct device_node *np)
 	cmu.clk_regs = fsys_clk_regs;
 	cmu.nr_clk_regs = ARRAY_SIZE(fsys_clk_regs);
 
-	exynos5260_cmu_register_one(np, &cmu);
+	samsung_cmu_register_one(np, &cmu);
 }
 
 CLK_OF_DECLARE(exynos5260_clk_fsys, "samsung,exynos5260-clock-fsys",
@@ -713,7 +580,7 @@ struct samsung_gate_clock g2d_gate_clks[] __initdata = {
 
 static void __init exynos5260_clk_g2d_init(struct device_node *np)
 {
-	struct exynos5260_cmu_info cmu = {0};
+	struct samsung_cmu_info cmu = {0};
 
 	cmu.mux_clks = g2d_mux_clks;
 	cmu.nr_mux_clks = ARRAY_SIZE(g2d_mux_clks);
@@ -725,7 +592,7 @@ static void __init exynos5260_clk_g2d_init(struct device_node *np)
 	cmu.clk_regs = g2d_clk_regs;
 	cmu.nr_clk_regs = ARRAY_SIZE(g2d_clk_regs);
 
-	exynos5260_cmu_register_one(np, &cmu);
+	samsung_cmu_register_one(np, &cmu);
 }
 
 CLK_OF_DECLARE(exynos5260_clk_g2d, "samsung,exynos5260-clock-g2d",
@@ -774,7 +641,7 @@ static struct samsung_pll_clock g3d_pll_clks[] __initdata = {
 
 static void __init exynos5260_clk_g3d_init(struct device_node *np)
 {
-	struct exynos5260_cmu_info cmu = {0};
+	struct samsung_cmu_info cmu = {0};
 
 	cmu.pll_clks = g3d_pll_clks;
 	cmu.nr_pll_clks =  ARRAY_SIZE(g3d_pll_clks);
@@ -788,7 +655,7 @@ static void __init exynos5260_clk_g3d_init(struct device_node *np)
 	cmu.clk_regs = g3d_clk_regs;
 	cmu.nr_clk_regs = ARRAY_SIZE(g3d_clk_regs);
 
-	exynos5260_cmu_register_one(np, &cmu);
+	samsung_cmu_register_one(np, &cmu);
 }
 
 CLK_OF_DECLARE(exynos5260_clk_g3d, "samsung,exynos5260-clock-g3d",
@@ -909,7 +776,7 @@ struct samsung_gate_clock gscl_gate_clks[] __initdata = {
 
 static void __init exynos5260_clk_gscl_init(struct device_node *np)
 {
-	struct exynos5260_cmu_info cmu = {0};
+	struct samsung_cmu_info cmu = {0};
 
 	cmu.mux_clks = gscl_mux_clks;
 	cmu.nr_mux_clks = ARRAY_SIZE(gscl_mux_clks);
@@ -921,7 +788,7 @@ static void __init exynos5260_clk_gscl_init(struct device_node *np)
 	cmu.clk_regs = gscl_clk_regs;
 	cmu.nr_clk_regs = ARRAY_SIZE(gscl_clk_regs);
 
-	exynos5260_cmu_register_one(np, &cmu);
+	samsung_cmu_register_one(np, &cmu);
 }
 
 CLK_OF_DECLARE(exynos5260_clk_gscl, "samsung,exynos5260-clock-gscl",
@@ -1028,7 +895,7 @@ struct samsung_gate_clock isp_gate_clks[] __initdata = {
 
 static void __init exynos5260_clk_isp_init(struct device_node *np)
 {
-	struct exynos5260_cmu_info cmu = {0};
+	struct samsung_cmu_info cmu = {0};
 
 	cmu.mux_clks = isp_mux_clks;
 	cmu.nr_mux_clks = ARRAY_SIZE(isp_mux_clks);
@@ -1040,7 +907,7 @@ static void __init exynos5260_clk_isp_init(struct device_node *np)
 	cmu.clk_regs = isp_clk_regs;
 	cmu.nr_clk_regs = ARRAY_SIZE(isp_clk_regs);
 
-	exynos5260_cmu_register_one(np, &cmu);
+	samsung_cmu_register_one(np, &cmu);
 }
 
 CLK_OF_DECLARE(exynos5260_clk_isp, "samsung,exynos5260-clock-isp",
@@ -1092,7 +959,7 @@ static struct samsung_pll_clock kfc_pll_clks[] __initdata = {
 
 static void __init exynos5260_clk_kfc_init(struct device_node *np)
 {
-	struct exynos5260_cmu_info cmu = {0};
+	struct samsung_cmu_info cmu = {0};
 
 	cmu.pll_clks = kfc_pll_clks;
 	cmu.nr_pll_clks =  ARRAY_SIZE(kfc_pll_clks);
@@ -1104,7 +971,7 @@ static void __init exynos5260_clk_kfc_init(struct device_node *np)
 	cmu.clk_regs = kfc_clk_regs;
 	cmu.nr_clk_regs = ARRAY_SIZE(kfc_clk_regs);
 
-	exynos5260_cmu_register_one(np, &cmu);
+	samsung_cmu_register_one(np, &cmu);
 }
 
 CLK_OF_DECLARE(exynos5260_clk_kfc, "samsung,exynos5260-clock-kfc",
@@ -1148,7 +1015,7 @@ struct samsung_gate_clock mfc_gate_clks[] __initdata = {
 
 static void __init exynos5260_clk_mfc_init(struct device_node *np)
 {
-	struct exynos5260_cmu_info cmu = {0};
+	struct samsung_cmu_info cmu = {0};
 
 	cmu.mux_clks = mfc_mux_clks;
 	cmu.nr_mux_clks = ARRAY_SIZE(mfc_mux_clks);
@@ -1160,7 +1027,7 @@ static void __init exynos5260_clk_mfc_init(struct device_node *np)
 	cmu.clk_regs = mfc_clk_regs;
 	cmu.nr_clk_regs = ARRAY_SIZE(mfc_clk_regs);
 
-	exynos5260_cmu_register_one(np, &cmu);
+	samsung_cmu_register_one(np, &cmu);
 }
 
 CLK_OF_DECLARE(exynos5260_clk_mfc, "samsung,exynos5260-clock-mfc",
@@ -1295,7 +1162,7 @@ static struct samsung_pll_clock mif_pll_clks[] __initdata = {
 
 static void __init exynos5260_clk_mif_init(struct device_node *np)
 {
-	struct exynos5260_cmu_info cmu = {0};
+	struct samsung_cmu_info cmu = {0};
 
 	cmu.pll_clks = mif_pll_clks;
 	cmu.nr_pll_clks =  ARRAY_SIZE(mif_pll_clks);
@@ -1309,7 +1176,7 @@ static void __init exynos5260_clk_mif_init(struct device_node *np)
 	cmu.clk_regs = mif_clk_regs;
 	cmu.nr_clk_regs = ARRAY_SIZE(mif_clk_regs);
 
-	exynos5260_cmu_register_one(np, &cmu);
+	samsung_cmu_register_one(np, &cmu);
 }
 
 CLK_OF_DECLARE(exynos5260_clk_mif, "samsung,exynos5260-clock-mif",
@@ -1503,7 +1370,7 @@ struct samsung_gate_clock peri_gate_clks[] __initdata = {
 
 static void __init exynos5260_clk_peri_init(struct device_node *np)
 {
-	struct exynos5260_cmu_info cmu = {0};
+	struct samsung_cmu_info cmu = {0};
 
 	cmu.mux_clks = peri_mux_clks;
 	cmu.nr_mux_clks = ARRAY_SIZE(peri_mux_clks);
@@ -1515,7 +1382,7 @@ static void __init exynos5260_clk_peri_init(struct device_node *np)
 	cmu.clk_regs = peri_clk_regs;
 	cmu.nr_clk_regs = ARRAY_SIZE(peri_clk_regs);
 
-	exynos5260_cmu_register_one(np, &cmu);
+	samsung_cmu_register_one(np, &cmu);
 }
 
 CLK_OF_DECLARE(exynos5260_clk_peri, "samsung,exynos5260-clock-peri",
@@ -1581,7 +1448,7 @@ struct samsung_fixed_rate_clock fixed_rate_clks[] __initdata = {
 	FRATE(PHYCLK_HDMI_LINK_O_TMDS_CLKHI, "phyclk_hdmi_link_o_tmds_clkhi",
 			NULL, CLK_IS_ROOT, 125000000),
 	FRATE(PHYCLK_MIPI_DPHY_4L_M_TXBYTECLKHS,
-			"phyclk_mipi_dphy_4l_m_txbyteclkhs" , NULL,
+			"phyclk_mipi_dphy_4l_m_txbyte_clkhs" , NULL,
 			CLK_IS_ROOT, 187500000),
 	FRATE(PHYCLK_DPTX_PHY_O_REF_CLK_24M, "phyclk_dptx_phy_o_ref_clk_24m",
 			NULL, CLK_IS_ROOT, 24000000),
@@ -1959,7 +1826,7 @@ static struct samsung_pll_clock top_pll_clks[] __initdata = {
 
 static void __init exynos5260_clk_top_init(struct device_node *np)
 {
-	struct exynos5260_cmu_info cmu = {0};
+	struct samsung_cmu_info cmu = {0};
 
 	cmu.pll_clks = top_pll_clks;
 	cmu.nr_pll_clks =  ARRAY_SIZE(top_pll_clks);
@@ -1975,7 +1842,7 @@ static void __init exynos5260_clk_top_init(struct device_node *np)
 	cmu.clk_regs = top_clk_regs;
 	cmu.nr_clk_regs = ARRAY_SIZE(top_clk_regs);
 
-	exynos5260_cmu_register_one(np, &cmu);
+	samsung_cmu_register_one(np, &cmu);
 }
 
 CLK_OF_DECLARE(exynos5260_clk_top, "samsung,exynos5260-clock-top",
diff --git a/drivers/clk/samsung/clk-exynos5440.c b/drivers/clk/samsung/clk-exynos5440.c
index 00d1d00a41de..979e81389cdd 100644
--- a/drivers/clk/samsung/clk-exynos5440.c
+++ b/drivers/clk/samsung/clk-exynos5440.c
@@ -15,6 +15,8 @@
 #include <linux/clk-provider.h>
 #include <linux/of.h>
 #include <linux/of_address.h>
+#include <linux/notifier.h>
+#include <linux/reboot.h>
 
 #include "clk.h"
 #include "clk-pll.h"
@@ -23,6 +25,8 @@
 #define CPU_CLK_STATUS		0xfc
 #define MISC_DOUT1		0x558
 
+static void __iomem *reg_base;
+
 /* parent clock name list */
 PNAME(mout_armclk_p)	= { "cplla", "cpllb" };
 PNAME(mout_spi_p)	= { "div125", "div200" };
@@ -89,10 +93,30 @@ static const struct of_device_id ext_clk_match[] __initconst = {
 	{},
 };
 
+static int exynos5440_clk_restart_notify(struct notifier_block *this,
+		unsigned long code, void *unused)
+{
+	u32 val, status;
+
+	status = readl_relaxed(reg_base + 0xbc);
+	val = readl_relaxed(reg_base + 0xcc);
+	val = (val & 0xffff0000) | (status & 0xffff);
+	writel_relaxed(val, reg_base + 0xcc);
+
+	return NOTIFY_DONE;
+}
+
+/*
+ * Exynos5440 Clock restart notifier, handles restart functionality
+ */
+static struct notifier_block exynos5440_clk_restart_handler = {
+	.notifier_call = exynos5440_clk_restart_notify,
+	.priority = 128,
+};
+
 /* register exynos5440 clocks */
 static void __init exynos5440_clk_init(struct device_node *np)
 {
-	void __iomem *reg_base;
 	struct samsung_clk_provider *ctx;
 
 	reg_base = of_iomap(np, 0);
@@ -125,6 +149,9 @@ static void __init exynos5440_clk_init(struct device_node *np)
 
 	samsung_clk_of_add_provider(np, ctx);
 
+	if (register_restart_handler(&exynos5440_clk_restart_handler))
+		pr_warn("exynos5440 clock can't register restart handler\n");
+
 	pr_info("Exynos5440: arm_clk = %ldHz\n", _get_rate("arm_clk"));
 	pr_info("exynos5440 clock initialization complete\n");
 }
diff --git a/drivers/clk/samsung/clk-exynos7.c b/drivers/clk/samsung/clk-exynos7.c
new file mode 100644
index 000000000000..ea4483b8d62e
--- /dev/null
+++ b/drivers/clk/samsung/clk-exynos7.c
@@ -0,0 +1,743 @@
+/*
+ * Copyright (c) 2014 Samsung Electronics Co., Ltd.
+ * Author: Naveen Krishna Ch <naveenkrishna.ch@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+*/
+
+#include <linux/clk.h>
+#include <linux/clkdev.h>
+#include <linux/clk-provider.h>
+#include <linux/of.h>
+
+#include "clk.h"
+#include <dt-bindings/clock/exynos7-clk.h>
+
+/* Register Offset definitions for CMU_TOPC (0x10570000) */
+#define CC_PLL_LOCK		0x0000
+#define BUS0_PLL_LOCK		0x0004
+#define BUS1_DPLL_LOCK		0x0008
+#define MFC_PLL_LOCK		0x000C
+#define AUD_PLL_LOCK		0x0010
+#define CC_PLL_CON0		0x0100
+#define BUS0_PLL_CON0		0x0110
+#define BUS1_DPLL_CON0		0x0120
+#define MFC_PLL_CON0		0x0130
+#define AUD_PLL_CON0		0x0140
+#define MUX_SEL_TOPC0		0x0200
+#define MUX_SEL_TOPC1		0x0204
+#define MUX_SEL_TOPC2		0x0208
+#define MUX_SEL_TOPC3		0x020C
+#define DIV_TOPC0		0x0600
+#define DIV_TOPC1		0x0604
+#define DIV_TOPC3		0x060C
+
+static struct samsung_fixed_factor_clock topc_fixed_factor_clks[] __initdata = {
+	FFACTOR(0, "ffac_topc_bus0_pll_div2", "mout_bus0_pll_ctrl", 1, 2, 0),
+	FFACTOR(0, "ffac_topc_bus0_pll_div4",
+		"ffac_topc_bus0_pll_div2", 1, 2, 0),
+	FFACTOR(0, "ffac_topc_bus1_pll_div2", "mout_bus1_pll_ctrl", 1, 2, 0),
+	FFACTOR(0, "ffac_topc_cc_pll_div2", "mout_cc_pll_ctrl", 1, 2, 0),
+	FFACTOR(0, "ffac_topc_mfc_pll_div2", "mout_mfc_pll_ctrl", 1, 2, 0),
+};
+
+/* List of parent clocks for Muxes in CMU_TOPC */
+PNAME(mout_bus0_pll_ctrl_p)	= { "fin_pll", "fout_bus0_pll" };
+PNAME(mout_bus1_pll_ctrl_p)	= { "fin_pll", "fout_bus1_pll" };
+PNAME(mout_cc_pll_ctrl_p)	= { "fin_pll", "fout_cc_pll" };
+PNAME(mout_mfc_pll_ctrl_p)	= { "fin_pll", "fout_mfc_pll" };
+
+PNAME(mout_topc_group2) = { "mout_sclk_bus0_pll_cmuc",
+	"mout_sclk_bus1_pll_cmuc", "mout_sclk_cc_pll_cmuc",
+	"mout_sclk_mfc_pll_cmuc" };
+
+PNAME(mout_sclk_bus0_pll_cmuc_p) = { "mout_bus0_pll_ctrl",
+	"ffac_topc_bus0_pll_div2", "ffac_topc_bus0_pll_div4"};
+PNAME(mout_sclk_bus1_pll_cmuc_p) = { "mout_bus1_pll_ctrl",
+	"ffac_topc_bus1_pll_div2"};
+PNAME(mout_sclk_cc_pll_cmuc_p) = { "mout_cc_pll_ctrl",
+	"ffac_topc_cc_pll_div2"};
+PNAME(mout_sclk_mfc_pll_cmuc_p) = { "mout_mfc_pll_ctrl",
+	"ffac_topc_mfc_pll_div2"};
+
+
+PNAME(mout_sclk_bus0_pll_out_p) = {"mout_bus0_pll_ctrl",
+	"ffac_topc_bus0_pll_div2"};
+
+static unsigned long topc_clk_regs[] __initdata = {
+	CC_PLL_LOCK,
+	BUS0_PLL_LOCK,
+	BUS1_DPLL_LOCK,
+	MFC_PLL_LOCK,
+	AUD_PLL_LOCK,
+	CC_PLL_CON0,
+	BUS0_PLL_CON0,
+	BUS1_DPLL_CON0,
+	MFC_PLL_CON0,
+	AUD_PLL_CON0,
+	MUX_SEL_TOPC0,
+	MUX_SEL_TOPC1,
+	MUX_SEL_TOPC2,
+	MUX_SEL_TOPC3,
+	DIV_TOPC0,
+	DIV_TOPC1,
+	DIV_TOPC3,
+};
+
+static struct samsung_mux_clock topc_mux_clks[] __initdata = {
+	MUX(0, "mout_bus0_pll_ctrl", mout_bus0_pll_ctrl_p, MUX_SEL_TOPC0, 0, 1),
+	MUX(0, "mout_bus1_pll_ctrl", mout_bus1_pll_ctrl_p, MUX_SEL_TOPC0, 4, 1),
+	MUX(0, "mout_cc_pll_ctrl", mout_cc_pll_ctrl_p, MUX_SEL_TOPC0, 8, 1),
+	MUX(0, "mout_mfc_pll_ctrl", mout_mfc_pll_ctrl_p, MUX_SEL_TOPC0, 12, 1),
+
+	MUX(0, "mout_sclk_bus0_pll_cmuc", mout_sclk_bus0_pll_cmuc_p,
+		MUX_SEL_TOPC0, 16, 2),
+	MUX(0, "mout_sclk_bus1_pll_cmuc", mout_sclk_bus1_pll_cmuc_p,
+		MUX_SEL_TOPC0, 20, 1),
+	MUX(0, "mout_sclk_cc_pll_cmuc", mout_sclk_cc_pll_cmuc_p,
+		MUX_SEL_TOPC0, 24, 1),
+	MUX(0, "mout_sclk_mfc_pll_cmuc", mout_sclk_mfc_pll_cmuc_p,
+		MUX_SEL_TOPC0, 28, 1),
+
+	MUX(0, "mout_sclk_bus0_pll_out", mout_sclk_bus0_pll_out_p,
+		MUX_SEL_TOPC1, 16, 1),
+
+	MUX(0, "mout_aclk_ccore_133", mout_topc_group2,	MUX_SEL_TOPC2, 4, 2),
+
+	MUX(0, "mout_aclk_peris_66", mout_topc_group2, MUX_SEL_TOPC3, 24, 2),
+};
+
+static struct samsung_div_clock topc_div_clks[] __initdata = {
+	DIV(DOUT_ACLK_CCORE_133, "dout_aclk_ccore_133", "mout_aclk_ccore_133",
+		DIV_TOPC0, 4, 4),
+
+	DIV(DOUT_ACLK_PERIS, "dout_aclk_peris_66", "mout_aclk_peris_66",
+		DIV_TOPC1, 24, 4),
+
+	DIV(DOUT_SCLK_BUS0_PLL, "dout_sclk_bus0_pll", "mout_sclk_bus0_pll_out",
+		DIV_TOPC3, 0, 3),
+	DIV(DOUT_SCLK_BUS1_PLL, "dout_sclk_bus1_pll", "mout_bus1_pll_ctrl",
+		DIV_TOPC3, 8, 3),
+	DIV(DOUT_SCLK_CC_PLL, "dout_sclk_cc_pll", "mout_cc_pll_ctrl",
+		DIV_TOPC3, 12, 3),
+	DIV(DOUT_SCLK_MFC_PLL, "dout_sclk_mfc_pll", "mout_mfc_pll_ctrl",
+		DIV_TOPC3, 16, 3),
+};
+
+static struct samsung_pll_clock topc_pll_clks[] __initdata = {
+	PLL(pll_1451x, 0, "fout_bus0_pll", "fin_pll", BUS0_PLL_LOCK,
+		BUS0_PLL_CON0, NULL),
+	PLL(pll_1452x, 0, "fout_cc_pll", "fin_pll", CC_PLL_LOCK,
+		CC_PLL_CON0, NULL),
+	PLL(pll_1452x, 0, "fout_bus1_pll", "fin_pll", BUS1_DPLL_LOCK,
+		BUS1_DPLL_CON0, NULL),
+	PLL(pll_1452x, 0, "fout_mfc_pll", "fin_pll", MFC_PLL_LOCK,
+		MFC_PLL_CON0, NULL),
+	PLL(pll_1460x, 0, "fout_aud_pll", "fin_pll", AUD_PLL_LOCK,
+		AUD_PLL_CON0, NULL),
+};
+
+static struct samsung_cmu_info topc_cmu_info __initdata = {
+	.pll_clks		= topc_pll_clks,
+	.nr_pll_clks		= ARRAY_SIZE(topc_pll_clks),
+	.mux_clks		= topc_mux_clks,
+	.nr_mux_clks		= ARRAY_SIZE(topc_mux_clks),
+	.div_clks		= topc_div_clks,
+	.nr_div_clks		= ARRAY_SIZE(topc_div_clks),
+	.fixed_factor_clks	= topc_fixed_factor_clks,
+	.nr_fixed_factor_clks	= ARRAY_SIZE(topc_fixed_factor_clks),
+	.nr_clk_ids		= TOPC_NR_CLK,
+	.clk_regs		= topc_clk_regs,
+	.nr_clk_regs		= ARRAY_SIZE(topc_clk_regs),
+};
+
+static void __init exynos7_clk_topc_init(struct device_node *np)
+{
+	samsung_cmu_register_one(np, &topc_cmu_info);
+}
+
+CLK_OF_DECLARE(exynos7_clk_topc, "samsung,exynos7-clock-topc",
+	exynos7_clk_topc_init);
+
+/* Register Offset definitions for CMU_TOP0 (0x105D0000) */
+#define MUX_SEL_TOP00			0x0200
+#define MUX_SEL_TOP01			0x0204
+#define MUX_SEL_TOP03			0x020C
+#define MUX_SEL_TOP0_PERIC3		0x023C
+#define DIV_TOP03			0x060C
+#define DIV_TOP0_PERIC3			0x063C
+#define ENABLE_SCLK_TOP0_PERIC3		0x0A3C
+
+/* List of parent clocks for Muxes in CMU_TOP0 */
+PNAME(mout_bus0_pll_p)	= { "fin_pll", "dout_sclk_bus0_pll" };
+PNAME(mout_bus1_pll_p)	= { "fin_pll", "dout_sclk_bus1_pll" };
+PNAME(mout_cc_pll_p)	= { "fin_pll", "dout_sclk_cc_pll" };
+PNAME(mout_mfc_pll_p)	= { "fin_pll", "dout_sclk_mfc_pll" };
+
+PNAME(mout_top0_half_bus0_pll_p) = {"mout_top0_bus0_pll",
+	"ffac_top0_bus0_pll_div2"};
+PNAME(mout_top0_half_bus1_pll_p) = {"mout_top0_bus1_pll",
+	"ffac_top0_bus1_pll_div2"};
+PNAME(mout_top0_half_cc_pll_p) = {"mout_top0_cc_pll",
+	"ffac_top0_cc_pll_div2"};
+PNAME(mout_top0_half_mfc_pll_p) = {"mout_top0_mfc_pll",
+	"ffac_top0_mfc_pll_div2"};
+
+PNAME(mout_top0_group1) = {"mout_top0_half_bus0_pll",
+	"mout_top0_half_bus1_pll", "mout_top0_half_cc_pll",
+	"mout_top0_half_mfc_pll"};
+
+static unsigned long top0_clk_regs[] __initdata = {
+	MUX_SEL_TOP00,
+	MUX_SEL_TOP01,
+	MUX_SEL_TOP03,
+	MUX_SEL_TOP0_PERIC3,
+	DIV_TOP03,
+	DIV_TOP0_PERIC3,
+	ENABLE_SCLK_TOP0_PERIC3,
+};
+
+static struct samsung_mux_clock top0_mux_clks[] __initdata = {
+	MUX(0, "mout_top0_mfc_pll", mout_mfc_pll_p, MUX_SEL_TOP00, 4, 1),
+	MUX(0, "mout_top0_cc_pll", mout_cc_pll_p, MUX_SEL_TOP00, 8, 1),
+	MUX(0, "mout_top0_bus1_pll", mout_bus1_pll_p, MUX_SEL_TOP00, 12, 1),
+	MUX(0, "mout_top0_bus0_pll", mout_bus0_pll_p, MUX_SEL_TOP00, 16, 1),
+
+	MUX(0, "mout_top0_half_mfc_pll", mout_top0_half_mfc_pll_p,
+		MUX_SEL_TOP01, 4, 1),
+	MUX(0, "mout_top0_half_cc_pll", mout_top0_half_cc_pll_p,
+		MUX_SEL_TOP01, 8, 1),
+	MUX(0, "mout_top0_half_bus1_pll", mout_top0_half_bus1_pll_p,
+		MUX_SEL_TOP01, 12, 1),
+	MUX(0, "mout_top0_half_bus0_pll", mout_top0_half_bus0_pll_p,
+		MUX_SEL_TOP01, 16, 1),
+
+	MUX(0, "mout_aclk_peric1_66", mout_top0_group1, MUX_SEL_TOP03, 12, 2),
+	MUX(0, "mout_aclk_peric0_66", mout_top0_group1, MUX_SEL_TOP03, 20, 2),
+
+	MUX(0, "mout_sclk_uart3", mout_top0_group1, MUX_SEL_TOP0_PERIC3, 4, 2),
+	MUX(0, "mout_sclk_uart2", mout_top0_group1, MUX_SEL_TOP0_PERIC3, 8, 2),
+	MUX(0, "mout_sclk_uart1", mout_top0_group1, MUX_SEL_TOP0_PERIC3, 12, 2),
+	MUX(0, "mout_sclk_uart0", mout_top0_group1, MUX_SEL_TOP0_PERIC3, 16, 2),
+};
+
+static struct samsung_div_clock top0_div_clks[] __initdata = {
+	DIV(DOUT_ACLK_PERIC1, "dout_aclk_peric1_66", "mout_aclk_peric1_66",
+		DIV_TOP03, 12, 6),
+	DIV(DOUT_ACLK_PERIC0, "dout_aclk_peric0_66", "mout_aclk_peric0_66",
+		DIV_TOP03, 20, 6),
+
+	DIV(0, "dout_sclk_uart3", "mout_sclk_uart3", DIV_TOP0_PERIC3, 4, 4),
+	DIV(0, "dout_sclk_uart2", "mout_sclk_uart2", DIV_TOP0_PERIC3, 8, 4),
+	DIV(0, "dout_sclk_uart1", "mout_sclk_uart1", DIV_TOP0_PERIC3, 12, 4),
+	DIV(0, "dout_sclk_uart0", "mout_sclk_uart0", DIV_TOP0_PERIC3, 16, 4),
+};
+
+static struct samsung_gate_clock top0_gate_clks[] __initdata = {
+	GATE(CLK_SCLK_UART3, "sclk_uart3", "dout_sclk_uart3",
+		ENABLE_SCLK_TOP0_PERIC3, 4, 0, 0),
+	GATE(CLK_SCLK_UART2, "sclk_uart2", "dout_sclk_uart2",
+		ENABLE_SCLK_TOP0_PERIC3, 8, 0, 0),
+	GATE(CLK_SCLK_UART1, "sclk_uart1", "dout_sclk_uart1",
+		ENABLE_SCLK_TOP0_PERIC3, 12, 0, 0),
+	GATE(CLK_SCLK_UART0, "sclk_uart0", "dout_sclk_uart0",
+		ENABLE_SCLK_TOP0_PERIC3, 16, 0, 0),
+};
+
+static struct samsung_fixed_factor_clock top0_fixed_factor_clks[] __initdata = {
+	FFACTOR(0, "ffac_top0_bus0_pll_div2", "mout_top0_bus0_pll", 1, 2, 0),
+	FFACTOR(0, "ffac_top0_bus1_pll_div2", "mout_top0_bus1_pll", 1, 2, 0),
+	FFACTOR(0, "ffac_top0_cc_pll_div2", "mout_top0_cc_pll", 1, 2, 0),
+	FFACTOR(0, "ffac_top0_mfc_pll_div2", "mout_top0_mfc_pll", 1, 2, 0),
+};
+
+static struct samsung_cmu_info top0_cmu_info __initdata = {
+	.mux_clks		= top0_mux_clks,
+	.nr_mux_clks		= ARRAY_SIZE(top0_mux_clks),
+	.div_clks		= top0_div_clks,
+	.nr_div_clks		= ARRAY_SIZE(top0_div_clks),
+	.gate_clks		= top0_gate_clks,
+	.nr_gate_clks		= ARRAY_SIZE(top0_gate_clks),
+	.fixed_factor_clks	= top0_fixed_factor_clks,
+	.nr_fixed_factor_clks	= ARRAY_SIZE(top0_fixed_factor_clks),
+	.nr_clk_ids		= TOP0_NR_CLK,
+	.clk_regs		= top0_clk_regs,
+	.nr_clk_regs		= ARRAY_SIZE(top0_clk_regs),
+};
+
+static void __init exynos7_clk_top0_init(struct device_node *np)
+{
+	samsung_cmu_register_one(np, &top0_cmu_info);
+}
+
+CLK_OF_DECLARE(exynos7_clk_top0, "samsung,exynos7-clock-top0",
+	exynos7_clk_top0_init);
+
+/* Register Offset definitions for CMU_TOP1 (0x105E0000) */
+#define MUX_SEL_TOP10			0x0200
+#define MUX_SEL_TOP11			0x0204
+#define MUX_SEL_TOP13			0x020C
+#define MUX_SEL_TOP1_FSYS0		0x0224
+#define MUX_SEL_TOP1_FSYS1		0x0228
+#define DIV_TOP13			0x060C
+#define DIV_TOP1_FSYS0			0x0624
+#define DIV_TOP1_FSYS1			0x0628
+#define ENABLE_ACLK_TOP13		0x080C
+#define ENABLE_SCLK_TOP1_FSYS0		0x0A24
+#define ENABLE_SCLK_TOP1_FSYS1		0x0A28
+
+/* List of parent clocks for Muxes in CMU_TOP1 */
+PNAME(mout_top1_bus0_pll_p)	= { "fin_pll", "dout_sclk_bus0_pll" };
+PNAME(mout_top1_bus1_pll_p)	= { "fin_pll", "dout_sclk_bus1_pll_b" };
+PNAME(mout_top1_cc_pll_p)	= { "fin_pll", "dout_sclk_cc_pll_b" };
+PNAME(mout_top1_mfc_pll_p)	= { "fin_pll", "dout_sclk_mfc_pll_b" };
+
+PNAME(mout_top1_half_bus0_pll_p) = {"mout_top1_bus0_pll",
+	"ffac_top1_bus0_pll_div2"};
+PNAME(mout_top1_half_bus1_pll_p) = {"mout_top1_bus1_pll",
+	"ffac_top1_bus1_pll_div2"};
+PNAME(mout_top1_half_cc_pll_p) = {"mout_top1_cc_pll",
+	"ffac_top1_cc_pll_div2"};
+PNAME(mout_top1_half_mfc_pll_p) = {"mout_top1_mfc_pll",
+	"ffac_top1_mfc_pll_div2"};
+
+PNAME(mout_top1_group1) = {"mout_top1_half_bus0_pll",
+	"mout_top1_half_bus1_pll", "mout_top1_half_cc_pll",
+	"mout_top1_half_mfc_pll"};
+
+static unsigned long top1_clk_regs[] __initdata = {
+	MUX_SEL_TOP10,
+	MUX_SEL_TOP11,
+	MUX_SEL_TOP13,
+	MUX_SEL_TOP1_FSYS0,
+	MUX_SEL_TOP1_FSYS1,
+	DIV_TOP13,
+	DIV_TOP1_FSYS0,
+	DIV_TOP1_FSYS1,
+	ENABLE_ACLK_TOP13,
+	ENABLE_SCLK_TOP1_FSYS0,
+	ENABLE_SCLK_TOP1_FSYS1,
+};
+
+static struct samsung_mux_clock top1_mux_clks[] __initdata = {
+	MUX(0, "mout_top1_mfc_pll", mout_top1_mfc_pll_p, MUX_SEL_TOP10, 4, 1),
+	MUX(0, "mout_top1_cc_pll", mout_top1_cc_pll_p, MUX_SEL_TOP10, 8, 1),
+	MUX(0, "mout_top1_bus1_pll", mout_top1_bus1_pll_p,
+		MUX_SEL_TOP10, 12, 1),
+	MUX(0, "mout_top1_bus0_pll", mout_top1_bus0_pll_p,
+		MUX_SEL_TOP10, 16, 1),
+
+	MUX(0, "mout_top1_half_mfc_pll", mout_top1_half_mfc_pll_p,
+		MUX_SEL_TOP11, 4, 1),
+	MUX(0, "mout_top1_half_cc_pll", mout_top1_half_cc_pll_p,
+		MUX_SEL_TOP11, 8, 1),
+	MUX(0, "mout_top1_half_bus1_pll", mout_top1_half_bus1_pll_p,
+		MUX_SEL_TOP11, 12, 1),
+	MUX(0, "mout_top1_half_bus0_pll", mout_top1_half_bus0_pll_p,
+		MUX_SEL_TOP11, 16, 1),
+
+	MUX(0, "mout_aclk_fsys1_200", mout_top1_group1, MUX_SEL_TOP13, 24, 2),
+	MUX(0, "mout_aclk_fsys0_200", mout_top1_group1, MUX_SEL_TOP13, 28, 2),
+
+	MUX(0, "mout_sclk_mmc2", mout_top1_group1, MUX_SEL_TOP1_FSYS0, 24, 2),
+
+	MUX(0, "mout_sclk_mmc1", mout_top1_group1, MUX_SEL_TOP1_FSYS1, 24, 2),
+	MUX(0, "mout_sclk_mmc0", mout_top1_group1, MUX_SEL_TOP1_FSYS1, 28, 2),
+};
+
+static struct samsung_div_clock top1_div_clks[] __initdata = {
+	DIV(DOUT_ACLK_FSYS1_200, "dout_aclk_fsys1_200", "mout_aclk_fsys1_200",
+		DIV_TOP13, 24, 4),
+	DIV(DOUT_ACLK_FSYS0_200, "dout_aclk_fsys0_200", "mout_aclk_fsys0_200",
+		DIV_TOP13, 28, 4),
+
+	DIV(DOUT_SCLK_MMC2, "dout_sclk_mmc2", "mout_sclk_mmc2",
+		DIV_TOP1_FSYS0, 24, 4),
+
+	DIV(DOUT_SCLK_MMC1, "dout_sclk_mmc1", "mout_sclk_mmc1",
+		DIV_TOP1_FSYS1, 24, 4),
+	DIV(DOUT_SCLK_MMC0, "dout_sclk_mmc0", "mout_sclk_mmc0",
+		DIV_TOP1_FSYS1, 28, 4),
+};
+
+static struct samsung_gate_clock top1_gate_clks[] __initdata = {
+	GATE(CLK_SCLK_MMC2, "sclk_mmc2", "dout_sclk_mmc2",
+		ENABLE_SCLK_TOP1_FSYS0, 24, CLK_SET_RATE_PARENT, 0),
+
+	GATE(CLK_SCLK_MMC1, "sclk_mmc1", "dout_sclk_mmc1",
+		ENABLE_SCLK_TOP1_FSYS1, 24, CLK_SET_RATE_PARENT, 0),
+	GATE(CLK_SCLK_MMC0, "sclk_mmc0", "dout_sclk_mmc0",
+		ENABLE_SCLK_TOP1_FSYS1, 28, CLK_SET_RATE_PARENT, 0),
+};
+
+static struct samsung_fixed_factor_clock top1_fixed_factor_clks[] __initdata = {
+	FFACTOR(0, "ffac_top1_bus0_pll_div2", "mout_top1_bus0_pll", 1, 2, 0),
+	FFACTOR(0, "ffac_top1_bus1_pll_div2", "mout_top1_bus1_pll", 1, 2, 0),
+	FFACTOR(0, "ffac_top1_cc_pll_div2", "mout_top1_cc_pll", 1, 2, 0),
+	FFACTOR(0, "ffac_top1_mfc_pll_div2", "mout_top1_mfc_pll", 1, 2, 0),
+};
+
+static struct samsung_cmu_info top1_cmu_info __initdata = {
+	.mux_clks		= top1_mux_clks,
+	.nr_mux_clks		= ARRAY_SIZE(top1_mux_clks),
+	.div_clks		= top1_div_clks,
+	.nr_div_clks		= ARRAY_SIZE(top1_div_clks),
+	.gate_clks		= top1_gate_clks,
+	.nr_gate_clks		= ARRAY_SIZE(top1_gate_clks),
+	.fixed_factor_clks	= top1_fixed_factor_clks,
+	.nr_fixed_factor_clks	= ARRAY_SIZE(top1_fixed_factor_clks),
+	.nr_clk_ids		= TOP1_NR_CLK,
+	.clk_regs		= top1_clk_regs,
+	.nr_clk_regs		= ARRAY_SIZE(top1_clk_regs),
+};
+
+static void __init exynos7_clk_top1_init(struct device_node *np)
+{
+	samsung_cmu_register_one(np, &top1_cmu_info);
+}
+
+CLK_OF_DECLARE(exynos7_clk_top1, "samsung,exynos7-clock-top1",
+	exynos7_clk_top1_init);
+
+/* Register Offset definitions for CMU_CCORE (0x105B0000) */
+#define MUX_SEL_CCORE			0x0200
+#define DIV_CCORE			0x0600
+#define ENABLE_ACLK_CCORE0		0x0800
+#define ENABLE_ACLK_CCORE1		0x0804
+#define ENABLE_PCLK_CCORE		0x0900
+
+/*
+ * List of parent clocks for Muxes in CMU_CCORE
+ */
+PNAME(mout_aclk_ccore_133_p)	= { "fin_pll", "dout_aclk_ccore_133" };
+
+static unsigned long ccore_clk_regs[] __initdata = {
+	MUX_SEL_CCORE,
+	ENABLE_PCLK_CCORE,
+};
+
+static struct samsung_mux_clock ccore_mux_clks[] __initdata = {
+	MUX(0, "mout_aclk_ccore_133_user", mout_aclk_ccore_133_p,
+		MUX_SEL_CCORE, 1, 1),
+};
+
+static struct samsung_gate_clock ccore_gate_clks[] __initdata = {
+	GATE(PCLK_RTC, "pclk_rtc", "mout_aclk_ccore_133_user",
+		ENABLE_PCLK_CCORE, 8, 0, 0),
+};
+
+static struct samsung_cmu_info ccore_cmu_info __initdata = {
+	.mux_clks		= ccore_mux_clks,
+	.nr_mux_clks		= ARRAY_SIZE(ccore_mux_clks),
+	.gate_clks		= ccore_gate_clks,
+	.nr_gate_clks		= ARRAY_SIZE(ccore_gate_clks),
+	.nr_clk_ids		= CCORE_NR_CLK,
+	.clk_regs		= ccore_clk_regs,
+	.nr_clk_regs		= ARRAY_SIZE(ccore_clk_regs),
+};
+
+static void __init exynos7_clk_ccore_init(struct device_node *np)
+{
+	samsung_cmu_register_one(np, &ccore_cmu_info);
+}
+
+CLK_OF_DECLARE(exynos7_clk_ccore, "samsung,exynos7-clock-ccore",
+	exynos7_clk_ccore_init);
+
+/* Register Offset definitions for CMU_PERIC0 (0x13610000) */
+#define MUX_SEL_PERIC0			0x0200
+#define ENABLE_PCLK_PERIC0		0x0900
+#define ENABLE_SCLK_PERIC0		0x0A00
+
+/* List of parent clocks for Muxes in CMU_PERIC0 */
+PNAME(mout_aclk_peric0_66_p)	= { "fin_pll", "dout_aclk_peric0_66" };
+PNAME(mout_sclk_uart0_p)	= { "fin_pll", "sclk_uart0" };
+
+static unsigned long peric0_clk_regs[] __initdata = {
+	MUX_SEL_PERIC0,
+	ENABLE_PCLK_PERIC0,
+	ENABLE_SCLK_PERIC0,
+};
+
+static struct samsung_mux_clock peric0_mux_clks[] __initdata = {
+	MUX(0, "mout_aclk_peric0_66_user", mout_aclk_peric0_66_p,
+		MUX_SEL_PERIC0, 0, 1),
+	MUX(0, "mout_sclk_uart0_user", mout_sclk_uart0_p,
+		MUX_SEL_PERIC0, 16, 1),
+};
+
+static struct samsung_gate_clock peric0_gate_clks[] __initdata = {
+	GATE(PCLK_HSI2C0, "pclk_hsi2c0", "mout_aclk_peric0_66_user",
+		ENABLE_PCLK_PERIC0, 8, 0, 0),
+	GATE(PCLK_HSI2C1, "pclk_hsi2c1", "mout_aclk_peric0_66_user",
+		ENABLE_PCLK_PERIC0, 9, 0, 0),
+	GATE(PCLK_HSI2C4, "pclk_hsi2c4", "mout_aclk_peric0_66_user",
+		ENABLE_PCLK_PERIC0, 10, 0, 0),
+	GATE(PCLK_HSI2C5, "pclk_hsi2c5", "mout_aclk_peric0_66_user",
+		ENABLE_PCLK_PERIC0, 11, 0, 0),
+	GATE(PCLK_HSI2C9, "pclk_hsi2c9", "mout_aclk_peric0_66_user",
+		ENABLE_PCLK_PERIC0, 12, 0, 0),
+	GATE(PCLK_HSI2C10, "pclk_hsi2c10", "mout_aclk_peric0_66_user",
+		ENABLE_PCLK_PERIC0, 13, 0, 0),
+	GATE(PCLK_HSI2C11, "pclk_hsi2c11", "mout_aclk_peric0_66_user",
+		ENABLE_PCLK_PERIC0, 14, 0, 0),
+	GATE(PCLK_UART0, "pclk_uart0", "mout_aclk_peric0_66_user",
+		ENABLE_PCLK_PERIC0, 16, 0, 0),
+	GATE(PCLK_ADCIF, "pclk_adcif", "mout_aclk_peric0_66_user",
+		ENABLE_PCLK_PERIC0, 20, 0, 0),
+	GATE(PCLK_PWM, "pclk_pwm", "mout_aclk_peric0_66_user",
+		ENABLE_PCLK_PERIC0, 21, 0, 0),
+
+	GATE(SCLK_UART0, "sclk_uart0_user", "mout_sclk_uart0_user",
+		ENABLE_SCLK_PERIC0, 16, 0, 0),
+	GATE(SCLK_PWM, "sclk_pwm", "fin_pll", ENABLE_SCLK_PERIC0, 21, 0, 0),
+};
+
+static struct samsung_cmu_info peric0_cmu_info __initdata = {
+	.mux_clks		= peric0_mux_clks,
+	.nr_mux_clks		= ARRAY_SIZE(peric0_mux_clks),
+	.gate_clks		= peric0_gate_clks,
+	.nr_gate_clks		= ARRAY_SIZE(peric0_gate_clks),
+	.nr_clk_ids		= PERIC0_NR_CLK,
+	.clk_regs		= peric0_clk_regs,
+	.nr_clk_regs		= ARRAY_SIZE(peric0_clk_regs),
+};
+
+static void __init exynos7_clk_peric0_init(struct device_node *np)
+{
+	samsung_cmu_register_one(np, &peric0_cmu_info);
+}
+
+/* Register Offset definitions for CMU_PERIC1 (0x14C80000) */
+#define MUX_SEL_PERIC10			0x0200
+#define MUX_SEL_PERIC11			0x0204
+#define ENABLE_PCLK_PERIC1		0x0900
+#define ENABLE_SCLK_PERIC10		0x0A00
+
+CLK_OF_DECLARE(exynos7_clk_peric0, "samsung,exynos7-clock-peric0",
+	exynos7_clk_peric0_init);
+
+/* List of parent clocks for Muxes in CMU_PERIC1 */
+PNAME(mout_aclk_peric1_66_p)	= { "fin_pll", "dout_aclk_peric1_66" };
+PNAME(mout_sclk_uart1_p)	= { "fin_pll", "sclk_uart1" };
+PNAME(mout_sclk_uart2_p)	= { "fin_pll", "sclk_uart2" };
+PNAME(mout_sclk_uart3_p)	= { "fin_pll", "sclk_uart3" };
+
+static unsigned long peric1_clk_regs[] __initdata = {
+	MUX_SEL_PERIC10,
+	MUX_SEL_PERIC11,
+	ENABLE_PCLK_PERIC1,
+	ENABLE_SCLK_PERIC10,
+};
+
+static struct samsung_mux_clock peric1_mux_clks[] __initdata = {
+	MUX(0, "mout_aclk_peric1_66_user", mout_aclk_peric1_66_p,
+		MUX_SEL_PERIC10, 0, 1),
+
+	MUX(0, "mout_sclk_uart1_user", mout_sclk_uart1_p,
+		MUX_SEL_PERIC11, 20, 1),
+	MUX(0, "mout_sclk_uart2_user", mout_sclk_uart2_p,
+		MUX_SEL_PERIC11, 24, 1),
+	MUX(0, "mout_sclk_uart3_user", mout_sclk_uart3_p,
+		MUX_SEL_PERIC11, 28, 1),
+};
+
+static struct samsung_gate_clock peric1_gate_clks[] __initdata = {
+	GATE(PCLK_HSI2C2, "pclk_hsi2c2", "mout_aclk_peric1_66_user",
+		ENABLE_PCLK_PERIC1, 4, 0, 0),
+	GATE(PCLK_HSI2C3, "pclk_hsi2c3", "mout_aclk_peric1_66_user",
+		ENABLE_PCLK_PERIC1, 5, 0, 0),
+	GATE(PCLK_HSI2C6, "pclk_hsi2c6", "mout_aclk_peric1_66_user",
+		ENABLE_PCLK_PERIC1, 6, 0, 0),
+	GATE(PCLK_HSI2C7, "pclk_hsi2c7", "mout_aclk_peric1_66_user",
+		ENABLE_PCLK_PERIC1, 7, 0, 0),
+	GATE(PCLK_HSI2C8, "pclk_hsi2c8", "mout_aclk_peric1_66_user",
+		ENABLE_PCLK_PERIC1, 8, 0, 0),
+	GATE(PCLK_UART1, "pclk_uart1", "mout_aclk_peric1_66_user",
+		ENABLE_PCLK_PERIC1, 9, 0, 0),
+	GATE(PCLK_UART2, "pclk_uart2", "mout_aclk_peric1_66_user",
+		ENABLE_PCLK_PERIC1, 10, 0, 0),
+	GATE(PCLK_UART3, "pclk_uart3", "mout_aclk_peric1_66_user",
+		ENABLE_PCLK_PERIC1, 11, 0, 0),
+
+	GATE(SCLK_UART1, "sclk_uart1_user", "mout_sclk_uart1_user",
+		ENABLE_SCLK_PERIC10, 9, 0, 0),
+	GATE(SCLK_UART2, "sclk_uart2_user", "mout_sclk_uart2_user",
+		ENABLE_SCLK_PERIC10, 10, 0, 0),
+	GATE(SCLK_UART3, "sclk_uart3_user", "mout_sclk_uart3_user",
+		ENABLE_SCLK_PERIC10, 11, 0, 0),
+};
+
+static struct samsung_cmu_info peric1_cmu_info __initdata = {
+	.mux_clks		= peric1_mux_clks,
+	.nr_mux_clks		= ARRAY_SIZE(peric1_mux_clks),
+	.gate_clks		= peric1_gate_clks,
+	.nr_gate_clks		= ARRAY_SIZE(peric1_gate_clks),
+	.nr_clk_ids		= PERIC1_NR_CLK,
+	.clk_regs		= peric1_clk_regs,
+	.nr_clk_regs		= ARRAY_SIZE(peric1_clk_regs),
+};
+
+static void __init exynos7_clk_peric1_init(struct device_node *np)
+{
+	samsung_cmu_register_one(np, &peric1_cmu_info);
+}
+
+CLK_OF_DECLARE(exynos7_clk_peric1, "samsung,exynos7-clock-peric1",
+	exynos7_clk_peric1_init);
+
+/* Register Offset definitions for CMU_PERIS (0x10040000) */
+#define MUX_SEL_PERIS			0x0200
+#define ENABLE_PCLK_PERIS		0x0900
+#define ENABLE_PCLK_PERIS_SECURE_CHIPID	0x0910
+#define ENABLE_SCLK_PERIS		0x0A00
+#define ENABLE_SCLK_PERIS_SECURE_CHIPID	0x0A10
+
+/* List of parent clocks for Muxes in CMU_PERIS */
+PNAME(mout_aclk_peris_66_p) = { "fin_pll", "dout_aclk_peris_66" };
+
+static unsigned long peris_clk_regs[] __initdata = {
+	MUX_SEL_PERIS,
+	ENABLE_PCLK_PERIS,
+	ENABLE_PCLK_PERIS_SECURE_CHIPID,
+	ENABLE_SCLK_PERIS,
+	ENABLE_SCLK_PERIS_SECURE_CHIPID,
+};
+
+static struct samsung_mux_clock peris_mux_clks[] __initdata = {
+	MUX(0, "mout_aclk_peris_66_user",
+		mout_aclk_peris_66_p, MUX_SEL_PERIS, 0, 1),
+};
+
+static struct samsung_gate_clock peris_gate_clks[] __initdata = {
+	GATE(PCLK_WDT, "pclk_wdt", "mout_aclk_peris_66_user",
+		ENABLE_PCLK_PERIS, 6, 0, 0),
+	GATE(PCLK_TMU, "pclk_tmu_apbif", "mout_aclk_peris_66_user",
+		ENABLE_PCLK_PERIS, 10, 0, 0),
+
+	GATE(PCLK_CHIPID, "pclk_chipid", "mout_aclk_peris_66_user",
+		ENABLE_PCLK_PERIS_SECURE_CHIPID, 0, 0, 0),
+	GATE(SCLK_CHIPID, "sclk_chipid", "fin_pll",
+		ENABLE_SCLK_PERIS_SECURE_CHIPID, 0, 0, 0),
+
+	GATE(SCLK_TMU, "sclk_tmu", "fin_pll", ENABLE_SCLK_PERIS, 10, 0, 0),
+};
+
+static struct samsung_cmu_info peris_cmu_info __initdata = {
+	.mux_clks		= peris_mux_clks,
+	.nr_mux_clks		= ARRAY_SIZE(peris_mux_clks),
+	.gate_clks		= peris_gate_clks,
+	.nr_gate_clks		= ARRAY_SIZE(peris_gate_clks),
+	.nr_clk_ids		= PERIS_NR_CLK,
+	.clk_regs		= peris_clk_regs,
+	.nr_clk_regs		= ARRAY_SIZE(peris_clk_regs),
+};
+
+static void __init exynos7_clk_peris_init(struct device_node *np)
+{
+	samsung_cmu_register_one(np, &peris_cmu_info);
+}
+
+CLK_OF_DECLARE(exynos7_clk_peris, "samsung,exynos7-clock-peris",
+	exynos7_clk_peris_init);
+
+/* Register Offset definitions for CMU_FSYS0 (0x10E90000) */
+#define MUX_SEL_FSYS00			0x0200
+#define MUX_SEL_FSYS01			0x0204
+#define ENABLE_ACLK_FSYS01		0x0804
+
+/*
+ * List of parent clocks for Muxes in CMU_FSYS0
+ */
+PNAME(mout_aclk_fsys0_200_p)	= { "fin_pll", "dout_aclk_fsys0_200" };
+PNAME(mout_sclk_mmc2_p)		= { "fin_pll", "sclk_mmc2" };
+
+static unsigned long fsys0_clk_regs[] __initdata = {
+	MUX_SEL_FSYS00,
+	MUX_SEL_FSYS01,
+	ENABLE_ACLK_FSYS01,
+};
+
+static struct samsung_mux_clock fsys0_mux_clks[] __initdata = {
+	MUX(0, "mout_aclk_fsys0_200_user", mout_aclk_fsys0_200_p,
+		MUX_SEL_FSYS00, 24, 1),
+
+	MUX(0, "mout_sclk_mmc2_user", mout_sclk_mmc2_p, MUX_SEL_FSYS01, 24, 1),
+};
+
+static struct samsung_gate_clock fsys0_gate_clks[] __initdata = {
+	GATE(ACLK_MMC2, "aclk_mmc2", "mout_aclk_fsys0_200_user",
+		ENABLE_ACLK_FSYS01, 31, 0, 0),
+};
+
+static struct samsung_cmu_info fsys0_cmu_info __initdata = {
+	.mux_clks		= fsys0_mux_clks,
+	.nr_mux_clks		= ARRAY_SIZE(fsys0_mux_clks),
+	.gate_clks		= fsys0_gate_clks,
+	.nr_gate_clks		= ARRAY_SIZE(fsys0_gate_clks),
+	.nr_clk_ids		= TOP1_NR_CLK,
+	.clk_regs		= fsys0_clk_regs,
+	.nr_clk_regs		= ARRAY_SIZE(fsys0_clk_regs),
+};
+
+static void __init exynos7_clk_fsys0_init(struct device_node *np)
+{
+	samsung_cmu_register_one(np, &fsys0_cmu_info);
+}
+
+CLK_OF_DECLARE(exynos7_clk_fsys0, "samsung,exynos7-clock-fsys0",
+	exynos7_clk_fsys0_init);
+
+/* Register Offset definitions for CMU_FSYS1 (0x156E0000) */
+#define MUX_SEL_FSYS10			0x0200
+#define MUX_SEL_FSYS11			0x0204
+#define ENABLE_ACLK_FSYS1		0x0800
+
+/*
+ * List of parent clocks for Muxes in CMU_FSYS1
+ */
+PNAME(mout_aclk_fsys1_200_p)	= { "fin_pll",  "dout_aclk_fsys1_200" };
+PNAME(mout_sclk_mmc0_p)		= { "fin_pll", "sclk_mmc0" };
+PNAME(mout_sclk_mmc1_p)		= { "fin_pll", "sclk_mmc1" };
+
+static unsigned long fsys1_clk_regs[] __initdata = {
+	MUX_SEL_FSYS10,
+	MUX_SEL_FSYS11,
+	ENABLE_ACLK_FSYS1,
+};
+
+static struct samsung_mux_clock fsys1_mux_clks[] __initdata = {
+	MUX(0, "mout_aclk_fsys1_200_user", mout_aclk_fsys1_200_p,
+		MUX_SEL_FSYS10, 28, 1),
+
+	MUX(0, "mout_sclk_mmc1_user", mout_sclk_mmc1_p, MUX_SEL_FSYS11, 24, 1),
+	MUX(0, "mout_sclk_mmc0_user", mout_sclk_mmc0_p, MUX_SEL_FSYS11, 28, 1),
+};
+
+static struct samsung_gate_clock fsys1_gate_clks[] __initdata = {
+	GATE(ACLK_MMC1, "aclk_mmc1", "mout_aclk_fsys1_200_user",
+		ENABLE_ACLK_FSYS1, 29, 0, 0),
+	GATE(ACLK_MMC0, "aclk_mmc0", "mout_aclk_fsys1_200_user",
+		ENABLE_ACLK_FSYS1, 30, 0, 0),
+};
+
+static struct samsung_cmu_info fsys1_cmu_info __initdata = {
+	.mux_clks		= fsys1_mux_clks,
+	.nr_mux_clks		= ARRAY_SIZE(fsys1_mux_clks),
+	.gate_clks		= fsys1_gate_clks,
+	.nr_gate_clks		= ARRAY_SIZE(fsys1_gate_clks),
+	.nr_clk_ids		= TOP1_NR_CLK,
+	.clk_regs		= fsys1_clk_regs,
+	.nr_clk_regs		= ARRAY_SIZE(fsys1_clk_regs),
+};
+
+static void __init exynos7_clk_fsys1_init(struct device_node *np)
+{
+	samsung_cmu_register_one(np, &fsys1_cmu_info);
+}
+
+CLK_OF_DECLARE(exynos7_clk_fsys1, "samsung,exynos7-clock-fsys1",
+	exynos7_clk_fsys1_init);
diff --git a/drivers/clk/samsung/clk-pll.c b/drivers/clk/samsung/clk-pll.c
index b07fad2a9167..9d70e5c03804 100644
--- a/drivers/clk/samsung/clk-pll.c
+++ b/drivers/clk/samsung/clk-pll.c
@@ -482,6 +482,8 @@ static const struct clk_ops samsung_pll45xx_clk_min_ops = {
 
 #define PLL46XX_VSEL_MASK	(1)
 #define PLL46XX_MDIV_MASK	(0x1FF)
+#define PLL1460X_MDIV_MASK	(0x3FF)
+
 #define PLL46XX_PDIV_MASK	(0x3F)
 #define PLL46XX_SDIV_MASK	(0x7)
 #define PLL46XX_VSEL_SHIFT	(27)
@@ -511,13 +513,15 @@ static unsigned long samsung_pll46xx_recalc_rate(struct clk_hw *hw,
 
 	pll_con0 = __raw_readl(pll->con_reg);
 	pll_con1 = __raw_readl(pll->con_reg + 4);
-	mdiv = (pll_con0 >> PLL46XX_MDIV_SHIFT) & PLL46XX_MDIV_MASK;
+	mdiv = (pll_con0 >> PLL46XX_MDIV_SHIFT) & ((pll->type == pll_1460x) ?
+				PLL1460X_MDIV_MASK : PLL46XX_MDIV_MASK);
 	pdiv = (pll_con0 >> PLL46XX_PDIV_SHIFT) & PLL46XX_PDIV_MASK;
 	sdiv = (pll_con0 >> PLL46XX_SDIV_SHIFT) & PLL46XX_SDIV_MASK;
 	kdiv = pll->type == pll_4650c ? pll_con1 & PLL4650C_KDIV_MASK :
 					pll_con1 & PLL46XX_KDIV_MASK;
 
-	shift = pll->type == pll_4600 ? 16 : 10;
+	shift = ((pll->type == pll_4600) || (pll->type == pll_1460x)) ? 16 : 10;
+
 	fvco *= (mdiv << shift) + kdiv;
 	do_div(fvco, (pdiv << sdiv));
 	fvco >>= shift;
@@ -573,14 +577,21 @@ static int samsung_pll46xx_set_rate(struct clk_hw *hw, unsigned long drate,
 		lock = 0xffff;
 
 	/* Set PLL PMS and VSEL values. */
-	con0 &= ~((PLL46XX_MDIV_MASK << PLL46XX_MDIV_SHIFT) |
+	if (pll->type == pll_1460x) {
+		con0 &= ~((PLL1460X_MDIV_MASK << PLL46XX_MDIV_SHIFT) |
+			(PLL46XX_PDIV_MASK << PLL46XX_PDIV_SHIFT) |
+			(PLL46XX_SDIV_MASK << PLL46XX_SDIV_SHIFT));
+	} else {
+		con0 &= ~((PLL46XX_MDIV_MASK << PLL46XX_MDIV_SHIFT) |
 			(PLL46XX_PDIV_MASK << PLL46XX_PDIV_SHIFT) |
 			(PLL46XX_SDIV_MASK << PLL46XX_SDIV_SHIFT) |
 			(PLL46XX_VSEL_MASK << PLL46XX_VSEL_SHIFT));
+		con0 |=	rate->vsel << PLL46XX_VSEL_SHIFT;
+	}
+
 	con0 |= (rate->mdiv << PLL46XX_MDIV_SHIFT) |
 			(rate->pdiv << PLL46XX_PDIV_SHIFT) |
-			(rate->sdiv << PLL46XX_SDIV_SHIFT) |
-			(rate->vsel << PLL46XX_VSEL_SHIFT);
+			(rate->sdiv << PLL46XX_SDIV_SHIFT);
 
 	/* Set PLL K, MFR and MRR values. */
 	con1 = __raw_readl(pll->con_reg + 0x4);
@@ -1190,6 +1201,9 @@ static void __init _samsung_clk_register_pll(struct samsung_clk_provider *ctx,
 	/* clk_ops for 35xx and 2550 are similar */
 	case pll_35xx:
 	case pll_2550:
+	case pll_1450x:
+	case pll_1451x:
+	case pll_1452x:
 		if (!pll->rate_table)
 			init.ops = &samsung_pll35xx_clk_min_ops;
 		else
@@ -1223,6 +1237,7 @@ static void __init _samsung_clk_register_pll(struct samsung_clk_provider *ctx,
 	case pll_4600:
 	case pll_4650:
 	case pll_4650c:
+	case pll_1460x:
 		if (!pll->rate_table)
 			init.ops = &samsung_pll46xx_clk_min_ops;
 		else
diff --git a/drivers/clk/samsung/clk-pll.h b/drivers/clk/samsung/clk-pll.h
index c0ed4d41fd90..213de9af8b4f 100644
--- a/drivers/clk/samsung/clk-pll.h
+++ b/drivers/clk/samsung/clk-pll.h
@@ -33,6 +33,10 @@ enum samsung_pll_type {
 	pll_s3c2440_mpll,
 	pll_2550xx,
 	pll_2650xx,
+	pll_1450x,
+	pll_1451x,
+	pll_1452x,
+	pll_1460x,
 };
 
 #define PLL_35XX_RATE(_rate, _m, _p, _s)			\
diff --git a/drivers/clk/samsung/clk-s3c2410-dclk.c b/drivers/clk/samsung/clk-s3c2410-dclk.c
index 0449cc0458ed..f4f29ed6bd25 100644
--- a/drivers/clk/samsung/clk-s3c2410-dclk.c
+++ b/drivers/clk/samsung/clk-s3c2410-dclk.c
@@ -426,7 +426,6 @@ MODULE_DEVICE_TABLE(platform, s3c24xx_dclk_driver_ids);
 static struct platform_driver s3c24xx_dclk_driver = {
 	.driver = {
 		.name		= "s3c24xx-dclk",
-		.owner		= THIS_MODULE,
 		.pm		= &s3c24xx_dclk_pm_ops,
 	},
 	.probe = s3c24xx_dclk_probe,
diff --git a/drivers/clk/samsung/clk-s3c2412.c b/drivers/clk/samsung/clk-s3c2412.c
index 34af09f6a155..2ceedaf8ce18 100644
--- a/drivers/clk/samsung/clk-s3c2412.c
+++ b/drivers/clk/samsung/clk-s3c2412.c
@@ -14,6 +14,7 @@
 #include <linux/of.h>
 #include <linux/of_address.h>
 #include <linux/syscore_ops.h>
+#include <linux/reboot.h>
 
 #include <dt-bindings/clock/s3c2412.h>
 
@@ -26,6 +27,7 @@
 #define CLKCON		0x0c
 #define CLKDIVN		0x14
 #define CLKSRC		0x1c
+#define SWRST		0x30
 
 /* list of PLLs to be registered */
 enum s3c2412_plls {
@@ -204,6 +206,28 @@ struct samsung_clock_alias s3c2412_aliases[] __initdata = {
 	ALIAS(MSYSCLK, NULL, "fclk"),
 };
 
+static int s3c2412_restart(struct notifier_block *this,
+			   unsigned long mode, void *cmd)
+{
+	/* errata "Watch-dog/Software Reset Problem" specifies that
+	 * this reset must be done with the SYSCLK sourced from
+	 * EXTCLK instead of FOUT to avoid a glitch in the reset
+	 * mechanism.
+	 *
+	 * See the watchdog section of the S3C2412 manual for more
+	 * information on this fix.
+	 */
+
+	__raw_writel(0x00, reg_base + CLKSRC);
+	__raw_writel(0x533C2412, reg_base + SWRST);
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block s3c2412_restart_handler = {
+	.notifier_call = s3c2412_restart,
+	.priority = 129,
+};
+
 /*
  * fixed rate clocks generated outside the soc
  * Only necessary until the devicetree-move is complete
@@ -233,6 +257,7 @@ void __init s3c2412_common_clk_init(struct device_node *np, unsigned long xti_f,
 				    unsigned long ext_f, void __iomem *base)
 {
 	struct samsung_clk_provider *ctx;
+	int ret;
 	reg_base = base;
 
 	if (np) {
@@ -267,6 +292,10 @@ void __init s3c2412_common_clk_init(struct device_node *np, unsigned long xti_f,
 	s3c2412_clk_sleep_init();
 
 	samsung_clk_of_add_provider(np, ctx);
+
+	ret = register_restart_handler(&s3c2412_restart_handler);
+	if (ret)
+		pr_warn("cannot register restart handler, %d\n", ret);
 }
 
 static void __init s3c2412_clk_init(struct device_node *np)
diff --git a/drivers/clk/samsung/clk-s3c2443.c b/drivers/clk/samsung/clk-s3c2443.c
index c92f853fca9f..0c3c182b902a 100644
--- a/drivers/clk/samsung/clk-s3c2443.c
+++ b/drivers/clk/samsung/clk-s3c2443.c
@@ -14,6 +14,7 @@
 #include <linux/of.h>
 #include <linux/of_address.h>
 #include <linux/syscore_ops.h>
+#include <linux/reboot.h>
 
 #include <dt-bindings/clock/s3c2443.h>
 
@@ -33,6 +34,7 @@
 #define HCLKCON		0x30
 #define PCLKCON		0x34
 #define SCLKCON		0x38
+#define SWRST		0x44
 
 /* the soc types */
 enum supported_socs {
@@ -354,6 +356,18 @@ struct samsung_clock_alias s3c2450_aliases[] __initdata = {
 	ALIAS(PCLK_I2C1, "s3c2410-i2c.1", "i2c"),
 };
 
+static int s3c2443_restart(struct notifier_block *this,
+			   unsigned long mode, void *cmd)
+{
+	__raw_writel(0x533c2443, reg_base + SWRST);
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block s3c2443_restart_handler = {
+	.notifier_call = s3c2443_restart,
+	.priority = 129,
+};
+
 /*
  * fixed rate clocks generated outside the soc
  * Only necessary until the devicetree-move is complete
@@ -378,6 +392,7 @@ void __init s3c2443_common_clk_init(struct device_node *np, unsigned long xti_f,
 				    void __iomem *base)
 {
 	struct samsung_clk_provider *ctx;
+	int ret;
 	reg_base = base;
 
 	if (np) {
@@ -447,6 +462,10 @@ void __init s3c2443_common_clk_init(struct device_node *np, unsigned long xti_f,
 	s3c2443_clk_sleep_init();
 
 	samsung_clk_of_add_provider(np, ctx);
+
+	ret = register_restart_handler(&s3c2443_restart_handler);
+	if (ret)
+		pr_warn("cannot register restart handler, %d\n", ret);
 }
 
 static void __init s3c2416_clk_init(struct device_node *np)
diff --git a/drivers/clk/samsung/clk-s5pv210-audss.c b/drivers/clk/samsung/clk-s5pv210-audss.c
index a8053b4aca56..de4455b75e8a 100644
--- a/drivers/clk/samsung/clk-s5pv210-audss.c
+++ b/drivers/clk/samsung/clk-s5pv210-audss.c
@@ -216,7 +216,6 @@ static const struct of_device_id s5pv210_audss_clk_of_match[] = {
 static struct platform_driver s5pv210_audss_clk_driver = {
 	.driver	= {
 		.name = "s5pv210-audss-clk",
-		.owner = THIS_MODULE,
 		.of_match_table = s5pv210_audss_clk_of_match,
 	},
 	.probe = s5pv210_audss_clk_probe,
diff --git a/drivers/clk/samsung/clk.c b/drivers/clk/samsung/clk.c
index deab84d9f37d..4bda54095a16 100644
--- a/drivers/clk/samsung/clk.c
+++ b/drivers/clk/samsung/clk.c
@@ -11,9 +11,13 @@
  * clock framework for Samsung platforms.
 */
 
+#include <linux/of_address.h>
 #include <linux/syscore_ops.h>
+
 #include "clk.h"
 
+static LIST_HEAD(clock_reg_cache_list);
+
 void samsung_clk_save(void __iomem *base,
 				    struct samsung_clk_reg_dump *rd,
 				    unsigned int num_regs)
@@ -281,7 +285,6 @@ void __init samsung_clk_register_gate(struct samsung_clk_provider *ctx,
  * obtain the clock speed of all external fixed clock sources from device
  * tree and register it
  */
-#ifdef CONFIG_OF
 void __init samsung_clk_of_register_fixed_ext(struct samsung_clk_provider *ctx,
 			struct samsung_fixed_rate_clock *fixed_rate_clk,
 			unsigned int nr_fixed_rate_clk,
@@ -298,7 +301,6 @@ void __init samsung_clk_of_register_fixed_ext(struct samsung_clk_provider *ctx,
 	}
 	samsung_clk_register_fixed_rate(ctx, fixed_rate_clk, nr_fixed_rate_clk);
 }
-#endif
 
 /* utility function to get the rate of a specified clock */
 unsigned long _get_rate(const char *clk_name)
@@ -313,3 +315,99 @@ unsigned long _get_rate(const char *clk_name)
 
 	return clk_get_rate(clk);
 }
+
+#ifdef CONFIG_PM_SLEEP
+static int samsung_clk_suspend(void)
+{
+	struct samsung_clock_reg_cache *reg_cache;
+
+	list_for_each_entry(reg_cache, &clock_reg_cache_list, node)
+		samsung_clk_save(reg_cache->reg_base, reg_cache->rdump,
+				reg_cache->rd_num);
+	return 0;
+}
+
+static void samsung_clk_resume(void)
+{
+	struct samsung_clock_reg_cache *reg_cache;
+
+	list_for_each_entry(reg_cache, &clock_reg_cache_list, node)
+		samsung_clk_restore(reg_cache->reg_base, reg_cache->rdump,
+				reg_cache->rd_num);
+}
+
+static struct syscore_ops samsung_clk_syscore_ops = {
+	.suspend = samsung_clk_suspend,
+	.resume = samsung_clk_resume,
+};
+
+static void samsung_clk_sleep_init(void __iomem *reg_base,
+		const unsigned long *rdump,
+		unsigned long nr_rdump)
+{
+	struct samsung_clock_reg_cache *reg_cache;
+
+	reg_cache = kzalloc(sizeof(struct samsung_clock_reg_cache),
+			GFP_KERNEL);
+	if (!reg_cache)
+		panic("could not allocate register reg_cache.\n");
+	reg_cache->rdump = samsung_clk_alloc_reg_dump(rdump, nr_rdump);
+
+	if (!reg_cache->rdump)
+		panic("could not allocate register dump storage.\n");
+
+	if (list_empty(&clock_reg_cache_list))
+		register_syscore_ops(&samsung_clk_syscore_ops);
+
+	reg_cache->reg_base = reg_base;
+	reg_cache->rd_num = nr_rdump;
+	list_add_tail(&reg_cache->node, &clock_reg_cache_list);
+}
+
+#else
+static void samsung_clk_sleep_init(void __iomem *reg_base,
+		const unsigned long *rdump,
+		unsigned long nr_rdump) {}
+#endif
+
+/*
+ * Common function which registers plls, muxes, dividers and gates
+ * for each CMU. It also add CMU register list to register cache.
+ */
+void __init samsung_cmu_register_one(struct device_node *np,
+			struct samsung_cmu_info *cmu)
+{
+	void __iomem *reg_base;
+	struct samsung_clk_provider *ctx;
+
+	reg_base = of_iomap(np, 0);
+	if (!reg_base)
+		panic("%s: failed to map registers\n", __func__);
+
+	ctx = samsung_clk_init(np, reg_base, cmu->nr_clk_ids);
+	if (!ctx)
+		panic("%s: unable to alllocate ctx\n", __func__);
+
+	if (cmu->pll_clks)
+		samsung_clk_register_pll(ctx, cmu->pll_clks, cmu->nr_pll_clks,
+			reg_base);
+	if (cmu->mux_clks)
+		samsung_clk_register_mux(ctx, cmu->mux_clks,
+			cmu->nr_mux_clks);
+	if (cmu->div_clks)
+		samsung_clk_register_div(ctx, cmu->div_clks, cmu->nr_div_clks);
+	if (cmu->gate_clks)
+		samsung_clk_register_gate(ctx, cmu->gate_clks,
+			cmu->nr_gate_clks);
+	if (cmu->fixed_clks)
+		samsung_clk_register_fixed_rate(ctx, cmu->fixed_clks,
+			cmu->nr_fixed_clks);
+	if (cmu->fixed_factor_clks)
+		samsung_clk_register_fixed_factor(ctx, cmu->fixed_factor_clks,
+			cmu->nr_fixed_factor_clks);
+	if (cmu->clk_regs)
+		samsung_clk_sleep_init(reg_base, cmu->clk_regs,
+			cmu->nr_clk_regs);
+
+	samsung_clk_of_add_provider(np, ctx);
+}
diff --git a/drivers/clk/samsung/clk.h b/drivers/clk/samsung/clk.h
index 66ab36b5cef1..8acabe1f32c4 100644
--- a/drivers/clk/samsung/clk.h
+++ b/drivers/clk/samsung/clk.h
@@ -13,19 +13,15 @@
 #ifndef __SAMSUNG_CLK_H
 #define __SAMSUNG_CLK_H
 
-#include <linux/clk.h>
 #include <linux/clkdev.h>
-#include <linux/io.h>
 #include <linux/clk-provider.h>
-#include <linux/of.h>
-#include <linux/of_address.h>
 #include "clk-pll.h"
 
 /**
  * struct samsung_clk_provider: information about clock provider
  * @reg_base: virtual address for the register base.
  * @clk_data: holds clock related data like clk* and number of clocks.
- * @lock: maintains exclusion bwtween callbacks for a given clock-provider.
+ * @lock: maintains exclusion between callbacks for a given clock-provider.
  */
 struct samsung_clk_provider {
 	void __iomem *reg_base;
@@ -324,6 +320,40 @@ struct samsung_pll_clock {
 	__PLL(_typ, _id, NULL, _name, _pname, CLK_GET_RATE_NOCACHE,	\
 		_lock, _con, _rtable, _alias)
 
+struct samsung_clock_reg_cache {
+	struct list_head node;
+	void __iomem *reg_base;
+	struct samsung_clk_reg_dump *rdump;
+	unsigned int rd_num;
+};
+
+struct samsung_cmu_info {
+	/* list of pll clocks and respective count */
+	struct samsung_pll_clock *pll_clks;
+	unsigned int nr_pll_clks;
+	/* list of mux clocks and respective count */
+	struct samsung_mux_clock *mux_clks;
+	unsigned int nr_mux_clks;
+	/* list of div clocks and respective count */
+	struct samsung_div_clock *div_clks;
+	unsigned int nr_div_clks;
+	/* list of gate clocks and respective count */
+	struct samsung_gate_clock *gate_clks;
+	unsigned int nr_gate_clks;
+	/* list of fixed clocks and respective count */
+	struct samsung_fixed_rate_clock *fixed_clks;
+	unsigned int nr_fixed_clks;
+	/* list of fixed factor clocks and respective count */
+	struct samsung_fixed_factor_clock *fixed_factor_clks;
+	unsigned int nr_fixed_factor_clks;
+	/* total number of clocks with IDs assigned*/
+	unsigned int nr_clk_ids;
+
+	/* list and number of clocks registers */
+	unsigned long *clk_regs;
+	unsigned int nr_clk_regs;
+};
+
 extern struct samsung_clk_provider *__init samsung_clk_init(
 			struct device_node *np, void __iomem *base,
 			unsigned long nr_clks);
@@ -362,6 +392,9 @@ extern void __init samsung_clk_register_pll(struct samsung_clk_provider *ctx,
 			struct samsung_pll_clock *pll_list,
 			unsigned int nr_clk, void __iomem *base);
 
+extern void __init samsung_cmu_register_one(struct device_node *,
+			struct samsung_cmu_info *);
+
 extern unsigned long _get_rate(const char *clk_name);
 
 extern void samsung_clk_save(void __iomem *base,
diff --git a/drivers/clk/shmobile/Makefile b/drivers/clk/shmobile/Makefile
index e0029237827a..960bf22d42ae 100644
--- a/drivers/clk/shmobile/Makefile
+++ b/drivers/clk/shmobile/Makefile
@@ -4,7 +4,6 @@ obj-$(CONFIG_ARCH_R8A7740)		+= clk-r8a7740.o
 obj-$(CONFIG_ARCH_R8A7779)		+= clk-r8a7779.o
 obj-$(CONFIG_ARCH_R8A7790)		+= clk-rcar-gen2.o
 obj-$(CONFIG_ARCH_R8A7791)		+= clk-rcar-gen2.o
+obj-$(CONFIG_ARCH_R8A7794)		+= clk-rcar-gen2.o
 obj-$(CONFIG_ARCH_SHMOBILE_MULTI)	+= clk-div6.o
 obj-$(CONFIG_ARCH_SHMOBILE_MULTI)	+= clk-mstp.o
-# for emply built-in.o
-obj-n	:= dummy
diff --git a/drivers/clk/shmobile/clk-div6.c b/drivers/clk/shmobile/clk-div6.c
index f065f694cb65..639241e31e03 100644
--- a/drivers/clk/shmobile/clk-div6.c
+++ b/drivers/clk/shmobile/clk-div6.c
@@ -32,6 +32,9 @@ struct div6_clock {
 	struct clk_hw hw;
 	void __iomem *reg;
 	unsigned int div;
+	u32 src_shift;
+	u32 src_width;
+	u8 *parents;
 };
 
 #define to_div6_clock(_hw) container_of(_hw, struct div6_clock, hw)
@@ -39,8 +42,11 @@ struct div6_clock {
 static int cpg_div6_clock_enable(struct clk_hw *hw)
 {
 	struct div6_clock *clock = to_div6_clock(hw);
+	u32 val;
 
-	clk_writel(CPG_DIV6_DIV(clock->div - 1), clock->reg);
+	val = (clk_readl(clock->reg) & ~(CPG_DIV6_DIV_MASK | CPG_DIV6_CKSTP))
+	    | CPG_DIV6_DIV(clock->div - 1);
+	clk_writel(val, clock->reg);
 
 	return 0;
 }
@@ -52,7 +58,7 @@ static void cpg_div6_clock_disable(struct clk_hw *hw)
 	/* DIV6 clocks require the divisor field to be non-zero when stopping
 	 * the clock.
 	 */
-	clk_writel(CPG_DIV6_CKSTP | CPG_DIV6_DIV(CPG_DIV6_DIV_MASK),
+	clk_writel(clk_readl(clock->reg) | CPG_DIV6_CKSTP | CPG_DIV6_DIV_MASK,
 		   clock->reg);
 }
 
@@ -94,12 +100,53 @@ static int cpg_div6_clock_set_rate(struct clk_hw *hw, unsigned long rate,
 {
 	struct div6_clock *clock = to_div6_clock(hw);
 	unsigned int div = cpg_div6_clock_calc_div(rate, parent_rate);
+	u32 val;
 
 	clock->div = div;
 
+	val = clk_readl(clock->reg) & ~CPG_DIV6_DIV_MASK;
 	/* Only program the new divisor if the clock isn't stopped. */
-	if (!(clk_readl(clock->reg) & CPG_DIV6_CKSTP))
-		clk_writel(CPG_DIV6_DIV(clock->div - 1), clock->reg);
+	if (!(val & CPG_DIV6_CKSTP))
+		clk_writel(val | CPG_DIV6_DIV(clock->div - 1), clock->reg);
+
+	return 0;
+}
+
+static u8 cpg_div6_clock_get_parent(struct clk_hw *hw)
+{
+	struct div6_clock *clock = to_div6_clock(hw);
+	unsigned int i;
+	u8 hw_index;
+
+	if (clock->src_width == 0)
+		return 0;
+
+	hw_index = (clk_readl(clock->reg) >> clock->src_shift) &
+		   (BIT(clock->src_width) - 1);
+	for (i = 0; i < __clk_get_num_parents(hw->clk); i++) {
+		if (clock->parents[i] == hw_index)
+			return i;
+	}
+
+	pr_err("%s: %s DIV6 clock set to invalid parent %u\n",
+	       __func__, __clk_get_name(hw->clk), hw_index);
+	return 0;
+}
+
+static int cpg_div6_clock_set_parent(struct clk_hw *hw, u8 index)
+{
+	struct div6_clock *clock = to_div6_clock(hw);
+	u8 hw_index;
+	u32 mask;
+
+	if (index >= __clk_get_num_parents(hw->clk))
+		return -EINVAL;
+
+	mask = ~((BIT(clock->src_width) - 1) << clock->src_shift);
+	hw_index = clock->parents[index];
+
+	clk_writel((clk_readl(clock->reg) & mask) |
+		(hw_index << clock->src_shift), clock->reg);
 
 	return 0;
 }
@@ -108,6 +155,8 @@ static const struct clk_ops cpg_div6_clock_ops = {
 	.enable = cpg_div6_clock_enable,
 	.disable = cpg_div6_clock_disable,
 	.is_enabled = cpg_div6_clock_is_enabled,
+	.get_parent = cpg_div6_clock_get_parent,
+	.set_parent = cpg_div6_clock_set_parent,
 	.recalc_rate = cpg_div6_clock_recalc_rate,
 	.round_rate = cpg_div6_clock_round_rate,
 	.set_rate = cpg_div6_clock_set_rate,
@@ -115,20 +164,33 @@ static const struct clk_ops cpg_div6_clock_ops = {
 
 static void __init cpg_div6_clock_init(struct device_node *np)
 {
+	unsigned int num_parents, valid_parents;
+	const char **parent_names;
 	struct clk_init_data init;
 	struct div6_clock *clock;
-	const char *parent_name;
 	const char *name;
 	struct clk *clk;
+	unsigned int i;
 	int ret;
 
 	clock = kzalloc(sizeof(*clock), GFP_KERNEL);
-	if (!clock) {
-		pr_err("%s: failed to allocate %s DIV6 clock\n",
+	if (!clock)
+		return;
+
+	num_parents = of_clk_get_parent_count(np);
+	if (num_parents < 1) {
+		pr_err("%s: no parent found for %s DIV6 clock\n",
 		       __func__, np->name);
 		return;
 	}
 
+	clock->parents = kmalloc_array(num_parents, sizeof(*clock->parents),
+		GFP_KERNEL);
+	parent_names = kmalloc_array(num_parents, sizeof(*parent_names),
+				GFP_KERNEL);
+	if (!parent_names)
+		return;
+
 	/* Remap the clock register and read the divisor. Disabling the
 	 * clock overwrites the divisor, so we need to cache its value for the
 	 * enable operation.
@@ -150,9 +212,34 @@ static void __init cpg_div6_clock_init(struct device_node *np)
 		goto error;
 	}
 
-	parent_name = of_clk_get_parent_name(np, 0);
-	if (parent_name == NULL) {
-		pr_err("%s: failed to get %s DIV6 clock parent name\n",
+
+	for (i = 0, valid_parents = 0; i < num_parents; i++) {
+		const char *name = of_clk_get_parent_name(np, i);
+
+		if (name) {
+			parent_names[valid_parents] = name;
+			clock->parents[valid_parents] = i;
+			valid_parents++;
+		}
+	}
+
+	switch (num_parents) {
+	case 1:
+		/* fixed parent clock */
+		clock->src_shift = clock->src_width = 0;
+		break;
+	case 4:
+		/* clock with EXSRC bits 6-7 */
+		clock->src_shift = 6;
+		clock->src_width = 2;
+		break;
+	case 8:
+		/* VCLK with EXSRC bits 12-14 */
+		clock->src_shift = 12;
+		clock->src_width = 3;
+		break;
+	default:
+		pr_err("%s: invalid number of parents for DIV6 clock %s\n",
 		       __func__, np->name);
 		goto error;
 	}
@@ -161,8 +248,8 @@ static void __init cpg_div6_clock_init(struct device_node *np)
 	init.name = name;
 	init.ops = &cpg_div6_clock_ops;
 	init.flags = CLK_IS_BASIC;
-	init.parent_names = &parent_name;
-	init.num_parents = 1;
+	init.parent_names = parent_names;
+	init.num_parents = valid_parents;
 
 	clock->hw.init = &init;
 
@@ -175,11 +262,13 @@ static void __init cpg_div6_clock_init(struct device_node *np)
 
 	of_clk_add_provider(np, of_clk_src_simple_get, clk);
 
+	kfree(parent_names);
 	return;
 
 error:
 	if (clock->reg)
 		iounmap(clock->reg);
+	kfree(parent_names);
 	kfree(clock);
 }
 CLK_OF_DECLARE(cpg_div6_clk, "renesas,cpg-div6-clock", cpg_div6_clock_init);
diff --git a/drivers/clk/shmobile/clk-rcar-gen2.c b/drivers/clk/shmobile/clk-rcar-gen2.c
index dff7f79a19b9..e996425d06a9 100644
--- a/drivers/clk/shmobile/clk-rcar-gen2.c
+++ b/drivers/clk/shmobile/clk-rcar-gen2.c
@@ -202,6 +202,7 @@ static const struct clk_div_table cpg_sdh_div_table[] = {
 };
 
 static const struct clk_div_table cpg_sd01_div_table[] = {
+	{  4,  8 },
 	{  5, 12 }, {  6, 16 }, {  7, 18 }, {  8, 24 },
 	{ 10, 36 }, { 11, 48 }, { 12, 10 }, {  0,  0 },
 };
diff --git a/drivers/clk/sunxi/Makefile b/drivers/clk/sunxi/Makefile
index 6850cba35871..a66953c0f430 100644
--- a/drivers/clk/sunxi/Makefile
+++ b/drivers/clk/sunxi/Makefile
@@ -5,6 +5,9 @@
 obj-y += clk-sunxi.o clk-factors.o
 obj-y += clk-a10-hosc.o
 obj-y += clk-a20-gmac.o
+obj-y += clk-mod0.o
+obj-y += clk-sun8i-mbus.o
+obj-y += clk-sun9i-core.o
 
 obj-$(CONFIG_MFD_SUN6I_PRCM) += \
 	clk-sun6i-ar100.o clk-sun6i-apb0.o clk-sun6i-apb0-gates.o \
diff --git a/drivers/clk/sunxi/clk-a20-gmac.c b/drivers/clk/sunxi/clk-a20-gmac.c
index 5296fd6dd7b3..0dcf4f205fb8 100644
--- a/drivers/clk/sunxi/clk-a20-gmac.c
+++ b/drivers/clk/sunxi/clk-a20-gmac.c
@@ -53,6 +53,11 @@ static DEFINE_SPINLOCK(gmac_lock);
 #define SUN7I_A20_GMAC_MASK	0x3
 #define SUN7I_A20_GMAC_PARENTS	2
 
+static u32 sun7i_a20_gmac_mux_table[SUN7I_A20_GMAC_PARENTS] = {
+	0x00, /* Select mii_phy_tx_clk */
+	0x02, /* Select gmac_int_tx_clk */
+};
+
 static void __init sun7i_a20_gmac_clk_setup(struct device_node *node)
 {
 	struct clk *clk;
@@ -90,7 +95,7 @@ static void __init sun7i_a20_gmac_clk_setup(struct device_node *node)
 	gate->lock = &gmac_lock;
 	mux->reg = reg;
 	mux->mask = SUN7I_A20_GMAC_MASK;
-	mux->flags = CLK_MUX_INDEX_BIT;
+	mux->table = sun7i_a20_gmac_mux_table;
 	mux->lock = &gmac_lock;
 
 	clk = clk_register_composite(NULL, clk_name,
diff --git a/drivers/clk/sunxi/clk-factors.c b/drivers/clk/sunxi/clk-factors.c
index 2057c8ac648f..62e08fb58554 100644
--- a/drivers/clk/sunxi/clk-factors.c
+++ b/drivers/clk/sunxi/clk-factors.c
@@ -9,18 +9,18 @@
  */
 
 #include <linux/clk-provider.h>
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/io.h>
 #include <linux/module.h>
+#include <linux/of_address.h>
 #include <linux/slab.h>
-#include <linux/io.h>
-#include <linux/err.h>
 #include <linux/string.h>
 
-#include <linux/delay.h>
-
 #include "clk-factors.h"
 
 /*
- * DOC: basic adjustable factor-based clock that cannot gate
+ * DOC: basic adjustable factor-based clock
  *
  * Traits of this clock:
  * prepare - clk_prepare only ensures that parents are prepared
@@ -32,6 +32,8 @@
 
 #define to_clk_factors(_hw) container_of(_hw, struct clk_factors, hw)
 
+#define FACTORS_MAX_PARENTS		5
+
 #define SETMASK(len, pos)		(((1U << (len)) - 1) << (pos))
 #define CLRMASK(len, pos)		(~(SETMASK(len, pos)))
 #define FACTOR_GET(bit, len, reg)	(((reg) & SETMASK(len, bit)) >> (bit))
@@ -79,7 +81,7 @@ static long clk_factors_round_rate(struct clk_hw *hw, unsigned long rate,
 
 static long clk_factors_determine_rate(struct clk_hw *hw, unsigned long rate,
 				       unsigned long *best_parent_rate,
-				       struct clk **best_parent_p)
+				       struct clk_hw **best_parent_p)
 {
 	struct clk *clk = hw->clk, *parent, *best_parent = NULL;
 	int i, num_parents;
@@ -106,7 +108,7 @@ static long clk_factors_determine_rate(struct clk_hw *hw, unsigned long rate,
 	}
 
 	if (best_parent)
-		*best_parent_p = best_parent;
+		*best_parent_p = __clk_get_hw(best_parent);
 	*best_parent_rate = best;
 
 	return best_child_rate;
@@ -147,9 +149,96 @@ static int clk_factors_set_rate(struct clk_hw *hw, unsigned long rate,
 	return 0;
 }
 
-const struct clk_ops clk_factors_ops = {
+static const struct clk_ops clk_factors_ops = {
 	.determine_rate = clk_factors_determine_rate,
 	.recalc_rate = clk_factors_recalc_rate,
 	.round_rate = clk_factors_round_rate,
 	.set_rate = clk_factors_set_rate,
 };
+
+struct clk * __init sunxi_factors_register(struct device_node *node,
+					   const struct factors_data *data,
+					   spinlock_t *lock)
+{
+	struct clk *clk;
+	struct clk_factors *factors;
+	struct clk_gate *gate = NULL;
+	struct clk_mux *mux = NULL;
+	struct clk_hw *gate_hw = NULL;
+	struct clk_hw *mux_hw = NULL;
+	const char *clk_name = node->name;
+	const char *parents[FACTORS_MAX_PARENTS];
+	void __iomem *reg;
+	int i = 0;
+
+	reg = of_iomap(node, 0);
+
+	/* if we have a mux, we will have >1 parents */
+	while (i < FACTORS_MAX_PARENTS &&
+	       (parents[i] = of_clk_get_parent_name(node, i)) != NULL)
+		i++;
+
+	/*
+	 * some factor clocks, such as pll5 and pll6, may have multiple
+	 * outputs, and have their name designated in factors_data
+	 */
+	if (data->name)
+		clk_name = data->name;
+	else
+		of_property_read_string(node, "clock-output-names", &clk_name);
+
+	factors = kzalloc(sizeof(struct clk_factors), GFP_KERNEL);
+	if (!factors)
+		return NULL;
+
+	/* set up factors properties */
+	factors->reg = reg;
+	factors->config = data->table;
+	factors->get_factors = data->getter;
+	factors->lock = lock;
+
+	/* Add a gate if this factor clock can be gated */
+	if (data->enable) {
+		gate = kzalloc(sizeof(struct clk_gate), GFP_KERNEL);
+		if (!gate) {
+			kfree(factors);
+			return NULL;
+		}
+
+		/* set up gate properties */
+		gate->reg = reg;
+		gate->bit_idx = data->enable;
+		gate->lock = factors->lock;
+		gate_hw = &gate->hw;
+	}
+
+	/* Add a mux if this factor clock can be muxed */
+	if (data->mux) {
+		mux = kzalloc(sizeof(struct clk_mux), GFP_KERNEL);
+		if (!mux) {
+			kfree(factors);
+			kfree(gate);
+			return NULL;
+		}
+
+		/* set up gate properties */
+		mux->reg = reg;
+		mux->shift = data->mux;
+		mux->mask = data->muxmask;
+		mux->lock = factors->lock;
+		mux_hw = &mux->hw;
+	}
+
+	clk = clk_register_composite(NULL, clk_name,
+			parents, i,
+			mux_hw, &clk_mux_ops,
+			&factors->hw, &clk_factors_ops,
+			gate_hw, &clk_gate_ops, 0);
+
+	if (!IS_ERR(clk)) {
+		of_clk_add_provider(node, of_clk_src_simple_get, clk);
+		clk_register_clkdev(clk, clk_name, NULL);
+	}
+
+	return clk;
+}
diff --git a/drivers/clk/sunxi/clk-factors.h b/drivers/clk/sunxi/clk-factors.h
index d2d0efa39379..912238fde132 100644
--- a/drivers/clk/sunxi/clk-factors.h
+++ b/drivers/clk/sunxi/clk-factors.h
@@ -3,6 +3,7 @@
 
 #include <linux/clk-provider.h>
 #include <linux/clkdev.h>
+#include <linux/spinlock.h>
 
 #define SUNXI_FACTORS_NOT_APPLICABLE	(0)
 
@@ -18,6 +19,15 @@ struct clk_factors_config {
 	u8 n_start;
 };
 
+struct factors_data {
+	int enable;
+	int mux;
+	int muxmask;
+	struct clk_factors_config *table;
+	void (*getter) (u32 *rate, u32 parent_rate, u8 *n, u8 *k, u8 *m, u8 *p);
+	const char *name;
+};
+
 struct clk_factors {
 	struct clk_hw hw;
 	void __iomem *reg;
@@ -26,5 +36,8 @@ struct clk_factors {
 	spinlock_t *lock;
 };
 
-extern const struct clk_ops clk_factors_ops;
+struct clk * __init sunxi_factors_register(struct device_node *node,
+					   const struct factors_data *data,
+					   spinlock_t *lock);
+
 #endif
diff --git a/drivers/clk/sunxi/clk-mod0.c b/drivers/clk/sunxi/clk-mod0.c
new file mode 100644
index 000000000000..da0524eaee94
--- /dev/null
+++ b/drivers/clk/sunxi/clk-mod0.c
@@ -0,0 +1,284 @@
+/*
+ * Copyright 2013 Emilio López
+ *
+ * Emilio López <emilio@elopez.com.ar>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/clk-provider.h>
+#include <linux/clkdev.h>
+#include <linux/of_address.h>
+
+#include "clk-factors.h"
+
+/**
+ * sun4i_get_mod0_factors() - calculates m, n factors for MOD0-style clocks
+ * MOD0 rate is calculated as follows
+ * rate = (parent_rate >> p) / (m + 1);
+ */
+
+static void sun4i_a10_get_mod0_factors(u32 *freq, u32 parent_rate,
+				       u8 *n, u8 *k, u8 *m, u8 *p)
+{
+	u8 div, calcm, calcp;
+
+	/* These clocks can only divide, so we will never be able to achieve
+	 * frequencies higher than the parent frequency */
+	if (*freq > parent_rate)
+		*freq = parent_rate;
+
+	div = DIV_ROUND_UP(parent_rate, *freq);
+
+	if (div < 16)
+		calcp = 0;
+	else if (div / 2 < 16)
+		calcp = 1;
+	else if (div / 4 < 16)
+		calcp = 2;
+	else
+		calcp = 3;
+
+	calcm = DIV_ROUND_UP(div, 1 << calcp);
+
+	*freq = (parent_rate >> calcp) / calcm;
+
+	/* we were called to round the frequency, we can now return */
+	if (n == NULL)
+		return;
+
+	*m = calcm - 1;
+	*p = calcp;
+}
+
+/* user manual says "n" but it's really "p" */
+static struct clk_factors_config sun4i_a10_mod0_config = {
+	.mshift = 0,
+	.mwidth = 4,
+	.pshift = 16,
+	.pwidth = 2,
+};
+
+static const struct factors_data sun4i_a10_mod0_data __initconst = {
+	.enable = 31,
+	.mux = 24,
+	.muxmask = BIT(1) | BIT(0),
+	.table = &sun4i_a10_mod0_config,
+	.getter = sun4i_a10_get_mod0_factors,
+};
+
+static DEFINE_SPINLOCK(sun4i_a10_mod0_lock);
+
+static void __init sun4i_a10_mod0_setup(struct device_node *node)
+{
+	sunxi_factors_register(node, &sun4i_a10_mod0_data, &sun4i_a10_mod0_lock);
+}
+CLK_OF_DECLARE(sun4i_a10_mod0, "allwinner,sun4i-a10-mod0-clk", sun4i_a10_mod0_setup);
+
+static DEFINE_SPINLOCK(sun5i_a13_mbus_lock);
+
+static void __init sun5i_a13_mbus_setup(struct device_node *node)
+{
+	struct clk *mbus = sunxi_factors_register(node, &sun4i_a10_mod0_data, &sun5i_a13_mbus_lock);
+
+	/* The MBUS clocks needs to be always enabled */
+	__clk_get(mbus);
+	clk_prepare_enable(mbus);
+}
+CLK_OF_DECLARE(sun5i_a13_mbus, "allwinner,sun5i-a13-mbus-clk", sun5i_a13_mbus_setup);
+
+struct mmc_phase_data {
+	u8	offset;
+};
+
+struct mmc_phase {
+	struct clk_hw		hw;
+	void __iomem		*reg;
+	struct mmc_phase_data	*data;
+	spinlock_t		*lock;
+};
+
+#define to_mmc_phase(_hw) container_of(_hw, struct mmc_phase, hw)
+
+static int mmc_get_phase(struct clk_hw *hw)
+{
+	struct clk *mmc, *mmc_parent, *clk = hw->clk;
+	struct mmc_phase *phase = to_mmc_phase(hw);
+	unsigned int mmc_rate, mmc_parent_rate;
+	u16 step, mmc_div;
+	u32 value;
+	u8 delay;
+
+	value = readl(phase->reg);
+	delay = (value >> phase->data->offset) & 0x3;
+
+	if (!delay)
+		return 180;
+
+	/* Get the main MMC clock */
+	mmc = clk_get_parent(clk);
+	if (!mmc)
+		return -EINVAL;
+
+	/* And its rate */
+	mmc_rate = clk_get_rate(mmc);
+	if (!mmc_rate)
+		return -EINVAL;
+
+	/* Now, get the MMC parent (most likely some PLL) */
+	mmc_parent = clk_get_parent(mmc);
+	if (!mmc_parent)
+		return -EINVAL;
+
+	/* And its rate */
+	mmc_parent_rate = clk_get_rate(mmc_parent);
+	if (!mmc_parent_rate)
+		return -EINVAL;
+
+	/* Get MMC clock divider */
+	mmc_div = mmc_parent_rate / mmc_rate;
+
+	step = DIV_ROUND_CLOSEST(360, mmc_div);
+	return delay * step;
+}
+
+static int mmc_set_phase(struct clk_hw *hw, int degrees)
+{
+	struct clk *mmc, *mmc_parent, *clk = hw->clk;
+	struct mmc_phase *phase = to_mmc_phase(hw);
+	unsigned int mmc_rate, mmc_parent_rate;
+	unsigned long flags;
+	u32 value;
+	u8 delay;
+
+	/* Get the main MMC clock */
+	mmc = clk_get_parent(clk);
+	if (!mmc)
+		return -EINVAL;
+
+	/* And its rate */
+	mmc_rate = clk_get_rate(mmc);
+	if (!mmc_rate)
+		return -EINVAL;
+
+	/* Now, get the MMC parent (most likely some PLL) */
+	mmc_parent = clk_get_parent(mmc);
+	if (!mmc_parent)
+		return -EINVAL;
+
+	/* And its rate */
+	mmc_parent_rate = clk_get_rate(mmc_parent);
+	if (!mmc_parent_rate)
+		return -EINVAL;
+
+	if (degrees != 180) {
+		u16 step, mmc_div;
+
+		/* Get MMC clock divider */
+		mmc_div = mmc_parent_rate / mmc_rate;
+
+		/*
+		 * We can only outphase the clocks by multiple of the
+		 * PLL's period.
+		 *
+		 * Since the MMC clock in only a divider, and the
+		 * formula to get the outphasing in degrees is deg =
+		 * 360 * delta / period
+		 *
+		 * If we simplify this formula, we can see that the
+		 * only thing that we're concerned about is the number
+		 * of period we want to outphase our clock from, and
+		 * the divider set by the MMC clock.
+		 */
+		step = DIV_ROUND_CLOSEST(360, mmc_div);
+		delay = DIV_ROUND_CLOSEST(degrees, step);
+	} else {
+		delay = 0;
+	}
+
+	spin_lock_irqsave(phase->lock, flags);
+	value = readl(phase->reg);
+	value &= ~GENMASK(phase->data->offset + 3, phase->data->offset);
+	value |= delay << phase->data->offset;
+	writel(value, phase->reg);
+	spin_unlock_irqrestore(phase->lock, flags);
+
+	return 0;
+}
+
+static const struct clk_ops mmc_clk_ops = {
+	.get_phase	= mmc_get_phase,
+	.set_phase	= mmc_set_phase,
+};
+
+static void __init sun4i_a10_mmc_phase_setup(struct device_node *node,
+					     struct mmc_phase_data *data)
+{
+	const char *parent_names[1] = { of_clk_get_parent_name(node, 0) };
+	struct clk_init_data init = {
+		.num_parents	= 1,
+		.parent_names	= parent_names,
+		.ops		= &mmc_clk_ops,
+	};
+
+	struct mmc_phase *phase;
+	struct clk *clk;
+
+	phase = kmalloc(sizeof(*phase), GFP_KERNEL);
+	if (!phase)
+		return;
+
+	phase->hw.init = &init;
+
+	phase->reg = of_iomap(node, 0);
+	if (!phase->reg)
+		goto err_free;
+
+	phase->data = data;
+	phase->lock = &sun4i_a10_mod0_lock;
+
+	if (of_property_read_string(node, "clock-output-names", &init.name))
+		init.name = node->name;
+
+	clk = clk_register(NULL, &phase->hw);
+	if (IS_ERR(clk))
+		goto err_unmap;
+
+	of_clk_add_provider(node, of_clk_src_simple_get, clk);
+
+	return;
+
+err_unmap:
+	iounmap(phase->reg);
+err_free:
+	kfree(phase);
+}
+
+
+static struct mmc_phase_data mmc_output_clk = {
+	.offset	= 8,
+};
+
+static struct mmc_phase_data mmc_sample_clk = {
+	.offset	= 20,
+};
+
+static void __init sun4i_a10_mmc_output_setup(struct device_node *node)
+{
+	sun4i_a10_mmc_phase_setup(node, &mmc_output_clk);
+}
+CLK_OF_DECLARE(sun4i_a10_mmc_output, "allwinner,sun4i-a10-mmc-output-clk", sun4i_a10_mmc_output_setup);
+
+static void __init sun4i_a10_mmc_sample_setup(struct device_node *node)
+{
+	sun4i_a10_mmc_phase_setup(node, &mmc_sample_clk);
+}
+CLK_OF_DECLARE(sun4i_a10_mmc_sample, "allwinner,sun4i-a10-mmc-sample-clk", sun4i_a10_mmc_sample_setup);
diff --git a/drivers/clk/sunxi/clk-sun6i-apb0-gates.c b/drivers/clk/sunxi/clk-sun6i-apb0-gates.c
index e10d0521ec76..64f3e46d383c 100644
--- a/drivers/clk/sunxi/clk-sun6i-apb0-gates.c
+++ b/drivers/clk/sunxi/clk-sun6i-apb0-gates.c
@@ -99,7 +99,6 @@ static int sun6i_a31_apb0_gates_clk_probe(struct platform_device *pdev)
 static struct platform_driver sun6i_a31_apb0_gates_clk_driver = {
 	.driver = {
 		.name = "sun6i-a31-apb0-gates-clk",
-		.owner = THIS_MODULE,
 		.of_match_table = sun6i_a31_apb0_gates_clk_dt_ids,
 	},
 	.probe = sun6i_a31_apb0_gates_clk_probe,
diff --git a/drivers/clk/sunxi/clk-sun6i-apb0.c b/drivers/clk/sunxi/clk-sun6i-apb0.c
index 1fa23371c8c6..70763600aeae 100644
--- a/drivers/clk/sunxi/clk-sun6i-apb0.c
+++ b/drivers/clk/sunxi/clk-sun6i-apb0.c
@@ -65,7 +65,6 @@ static const struct of_device_id sun6i_a31_apb0_clk_dt_ids[] = {
 static struct platform_driver sun6i_a31_apb0_clk_driver = {
 	.driver = {
 		.name = "sun6i-a31-apb0-clk",
-		.owner = THIS_MODULE,
 		.of_match_table = sun6i_a31_apb0_clk_dt_ids,
 	},
 	.probe = sun6i_a31_apb0_clk_probe,
diff --git a/drivers/clk/sunxi/clk-sun6i-ar100.c b/drivers/clk/sunxi/clk-sun6i-ar100.c
index eca8ca025b6a..3d282fb8f85c 100644
--- a/drivers/clk/sunxi/clk-sun6i-ar100.c
+++ b/drivers/clk/sunxi/clk-sun6i-ar100.c
@@ -46,7 +46,7 @@ static unsigned long ar100_recalc_rate(struct clk_hw *hw,
 
 static long ar100_determine_rate(struct clk_hw *hw, unsigned long rate,
 				 unsigned long *best_parent_rate,
-				 struct clk **best_parent_clk)
+				 struct clk_hw **best_parent_clk)
 {
 	int nparents = __clk_get_num_parents(hw->clk);
 	long best_rate = -EINVAL;
@@ -100,7 +100,7 @@ static long ar100_determine_rate(struct clk_hw *hw, unsigned long rate,
 
 		tmp_rate = (parent_rate >> shift) / div;
 		if (!*best_parent_clk || tmp_rate > best_rate) {
-			*best_parent_clk = parent;
+			*best_parent_clk = __clk_get_hw(parent);
 			*best_parent_rate = parent_rate;
 			best_rate = tmp_rate;
 		}
@@ -221,7 +221,6 @@ static const struct of_device_id sun6i_a31_ar100_clk_dt_ids[] = {
 static struct platform_driver sun6i_a31_ar100_clk_driver = {
 	.driver = {
 		.name = "sun6i-a31-ar100-clk",
-		.owner = THIS_MODULE,
 		.of_match_table = sun6i_a31_ar100_clk_dt_ids,
 	},
 	.probe = sun6i_a31_ar100_clk_probe,
diff --git a/drivers/clk/sunxi/clk-sun8i-apb0.c b/drivers/clk/sunxi/clk-sun8i-apb0.c
index 1f5ba9b4b8cd..155d0022194f 100644
--- a/drivers/clk/sunxi/clk-sun8i-apb0.c
+++ b/drivers/clk/sunxi/clk-sun8i-apb0.c
@@ -56,7 +56,6 @@ static const struct of_device_id sun8i_a23_apb0_clk_dt_ids[] = {
 static struct platform_driver sun8i_a23_apb0_clk_driver = {
 	.driver = {
 		.name = "sun8i-a23-apb0-clk",
-		.owner = THIS_MODULE,
 		.of_match_table = sun8i_a23_apb0_clk_dt_ids,
 	},
 	.probe = sun8i_a23_apb0_clk_probe,
diff --git a/drivers/clk/sunxi/clk-sun8i-mbus.c b/drivers/clk/sunxi/clk-sun8i-mbus.c
new file mode 100644
index 000000000000..ef49786eefd3
--- /dev/null
+++ b/drivers/clk/sunxi/clk-sun8i-mbus.c
@@ -0,0 +1,79 @@
+/*
+ * Copyright 2014 Chen-Yu Tsai
+ *
+ * Chen-Yu Tsai <wens@csie.org>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/clk-provider.h>
+#include <linux/clkdev.h>
+#include <linux/of_address.h>
+
+#include "clk-factors.h"
+
+/**
+ * sun8i_a23_get_mbus_factors() - calculates m factor for MBUS clocks
+ * MBUS rate is calculated as follows
+ * rate = parent_rate / (m + 1);
+ */
+
+static void sun8i_a23_get_mbus_factors(u32 *freq, u32 parent_rate,
+				       u8 *n, u8 *k, u8 *m, u8 *p)
+{
+	u8 div;
+
+	/*
+	 * These clocks can only divide, so we will never be able to
+	 * achieve frequencies higher than the parent frequency
+	 */
+	if (*freq > parent_rate)
+		*freq = parent_rate;
+
+	div = DIV_ROUND_UP(parent_rate, *freq);
+
+	if (div > 8)
+		div = 8;
+
+	*freq = parent_rate / div;
+
+	/* we were called to round the frequency, we can now return */
+	if (m == NULL)
+		return;
+
+	*m = div - 1;
+}
+
+static struct clk_factors_config sun8i_a23_mbus_config = {
+	.mshift = 0,
+	.mwidth = 3,
+};
+
+static const struct factors_data sun8i_a23_mbus_data __initconst = {
+	.enable = 31,
+	.mux = 24,
+	.muxmask = BIT(1) | BIT(0),
+	.table = &sun8i_a23_mbus_config,
+	.getter = sun8i_a23_get_mbus_factors,
+};
+
+static DEFINE_SPINLOCK(sun8i_a23_mbus_lock);
+
+static void __init sun8i_a23_mbus_setup(struct device_node *node)
+{
+	struct clk *mbus = sunxi_factors_register(node, &sun8i_a23_mbus_data,
+						  &sun8i_a23_mbus_lock);
+
+	/* The MBUS clocks needs to be always enabled */
+	__clk_get(mbus);
+	clk_prepare_enable(mbus);
+}
+CLK_OF_DECLARE(sun8i_a23_mbus, "allwinner,sun8i-a23-mbus-clk", sun8i_a23_mbus_setup);
diff --git a/drivers/clk/sunxi/clk-sun9i-core.c b/drivers/clk/sunxi/clk-sun9i-core.c
new file mode 100644
index 000000000000..3cb9036d91bb
--- /dev/null
+++ b/drivers/clk/sunxi/clk-sun9i-core.c
@@ -0,0 +1,271 @@
+/*
+ * Copyright 2014 Chen-Yu Tsai
+ *
+ * Chen-Yu Tsai <wens@csie.org>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/clk-provider.h>
+#include <linux/clkdev.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/log2.h>
+
+#include "clk-factors.h"
+
+
+/**
+ * sun9i_a80_get_pll4_factors() - calculates n, p, m factors for PLL1
+ * PLL4 rate is calculated as follows
+ * rate = (parent_rate * n >> p) / (m + 1);
+ * parent_rate is always 24Mhz
+ *
+ * p and m are named div1 and div2 in Allwinner's SDK
+ */
+
+static void sun9i_a80_get_pll4_factors(u32 *freq, u32 parent_rate,
+				       u8 *n, u8 *k, u8 *m, u8 *p)
+{
+	int div;
+
+	/* Normalize value to a 6M multiple */
+	div = DIV_ROUND_UP(*freq, 6000000);
+
+	/* divs above 256 cannot be odd */
+	if (div > 256)
+		div = round_up(div, 2);
+
+	/* divs above 512 must be a multiple of 4 */
+	if (div > 512)
+		div = round_up(div, 4);
+
+	*freq = 6000000 * div;
+
+	/* we were called to round the frequency, we can now return */
+	if (n == NULL)
+		return;
+
+	/* p will be 1 for divs under 512 */
+	if (div < 512)
+		*p = 1;
+	else
+		*p = 0;
+
+	/* m will be 1 if div is odd */
+	if (div & 1)
+		*m = 1;
+	else
+		*m = 0;
+
+	/* calculate a suitable n based on m and p */
+	*n = div / (*p + 1) / (*m + 1);
+}
+
+static struct clk_factors_config sun9i_a80_pll4_config = {
+	.mshift = 18,
+	.mwidth = 1,
+	.nshift = 8,
+	.nwidth = 8,
+	.pshift = 16,
+	.pwidth = 1,
+};
+
+static const struct factors_data sun9i_a80_pll4_data __initconst = {
+	.enable = 31,
+	.table = &sun9i_a80_pll4_config,
+	.getter = sun9i_a80_get_pll4_factors,
+};
+
+static DEFINE_SPINLOCK(sun9i_a80_pll4_lock);
+
+static void __init sun9i_a80_pll4_setup(struct device_node *node)
+{
+	sunxi_factors_register(node, &sun9i_a80_pll4_data, &sun9i_a80_pll4_lock);
+}
+CLK_OF_DECLARE(sun9i_a80_pll4, "allwinner,sun9i-a80-pll4-clk", sun9i_a80_pll4_setup);
+
+
+/**
+ * sun9i_a80_get_gt_factors() - calculates m factor for GT
+ * GT rate is calculated as follows
+ * rate = parent_rate / (m + 1);
+ */
+
+static void sun9i_a80_get_gt_factors(u32 *freq, u32 parent_rate,
+				     u8 *n, u8 *k, u8 *m, u8 *p)
+{
+	u32 div;
+
+	if (parent_rate < *freq)
+		*freq = parent_rate;
+
+	div = DIV_ROUND_UP(parent_rate, *freq);
+
+	/* maximum divider is 4 */
+	if (div > 4)
+		div = 4;
+
+	*freq = parent_rate / div;
+
+	/* we were called to round the frequency, we can now return */
+	if (!m)
+		return;
+
+	*m = div;
+}
+
+static struct clk_factors_config sun9i_a80_gt_config = {
+	.mshift = 0,
+	.mwidth = 2,
+};
+
+static const struct factors_data sun9i_a80_gt_data __initconst = {
+	.mux = 24,
+	.muxmask = BIT(1) | BIT(0),
+	.table = &sun9i_a80_gt_config,
+	.getter = sun9i_a80_get_gt_factors,
+};
+
+static DEFINE_SPINLOCK(sun9i_a80_gt_lock);
+
+static void __init sun9i_a80_gt_setup(struct device_node *node)
+{
+	struct clk *gt = sunxi_factors_register(node, &sun9i_a80_gt_data,
+						&sun9i_a80_gt_lock);
+
+	/* The GT bus clock needs to be always enabled */
+	__clk_get(gt);
+	clk_prepare_enable(gt);
+}
+CLK_OF_DECLARE(sun9i_a80_gt, "allwinner,sun9i-a80-gt-clk", sun9i_a80_gt_setup);
+
+
+/**
+ * sun9i_a80_get_ahb_factors() - calculates p factor for AHB0/1/2
+ * AHB rate is calculated as follows
+ * rate = parent_rate >> p;
+ */
+
+static void sun9i_a80_get_ahb_factors(u32 *freq, u32 parent_rate,
+				      u8 *n, u8 *k, u8 *m, u8 *p)
+{
+	u32 _p;
+
+	if (parent_rate < *freq)
+		*freq = parent_rate;
+
+	_p = order_base_2(DIV_ROUND_UP(parent_rate, *freq));
+
+	/* maximum p is 3 */
+	if (_p > 3)
+		_p = 3;
+
+	*freq = parent_rate >> _p;
+
+	/* we were called to round the frequency, we can now return */
+	if (!p)
+		return;
+
+	*p = _p;
+}
+
+static struct clk_factors_config sun9i_a80_ahb_config = {
+	.pshift = 0,
+	.pwidth = 2,
+};
+
+static const struct factors_data sun9i_a80_ahb_data __initconst = {
+	.mux = 24,
+	.muxmask = BIT(1) | BIT(0),
+	.table = &sun9i_a80_ahb_config,
+	.getter = sun9i_a80_get_ahb_factors,
+};
+
+static DEFINE_SPINLOCK(sun9i_a80_ahb_lock);
+
+static void __init sun9i_a80_ahb_setup(struct device_node *node)
+{
+	sunxi_factors_register(node, &sun9i_a80_ahb_data, &sun9i_a80_ahb_lock);
+}
+CLK_OF_DECLARE(sun9i_a80_ahb, "allwinner,sun9i-a80-ahb-clk", sun9i_a80_ahb_setup);
+
+
+static const struct factors_data sun9i_a80_apb0_data __initconst = {
+	.mux = 24,
+	.muxmask = BIT(0),
+	.table = &sun9i_a80_ahb_config,
+	.getter = sun9i_a80_get_ahb_factors,
+};
+
+static DEFINE_SPINLOCK(sun9i_a80_apb0_lock);
+
+static void __init sun9i_a80_apb0_setup(struct device_node *node)
+{
+	sunxi_factors_register(node, &sun9i_a80_apb0_data, &sun9i_a80_apb0_lock);
+}
+CLK_OF_DECLARE(sun9i_a80_apb0, "allwinner,sun9i-a80-apb0-clk", sun9i_a80_apb0_setup);
+
+
+/**
+ * sun9i_a80_get_apb1_factors() - calculates m, p factors for APB1
+ * APB1 rate is calculated as follows
+ * rate = (parent_rate >> p) / (m + 1);
+ */
+
+static void sun9i_a80_get_apb1_factors(u32 *freq, u32 parent_rate,
+				       u8 *n, u8 *k, u8 *m, u8 *p)
+{
+	u32 div;
+	u8 calcm, calcp;
+
+	if (parent_rate < *freq)
+		*freq = parent_rate;
+
+	div = DIV_ROUND_UP(parent_rate, *freq);
+
+	/* Highest possible divider is 256 (p = 3, m = 31) */
+	if (div > 256)
+		div = 256;
+
+	calcp = order_base_2(div);
+	calcm = (parent_rate >> calcp) - 1;
+	*freq = (parent_rate >> calcp) / (calcm + 1);
+
+	/* we were called to round the frequency, we can now return */
+	if (n == NULL)
+		return;
+
+	*m = calcm;
+	*p = calcp;
+}
+
+static struct clk_factors_config sun9i_a80_apb1_config = {
+	.mshift = 0,
+	.mwidth = 5,
+	.pshift = 16,
+	.pwidth = 2,
+};
+
+static const struct factors_data sun9i_a80_apb1_data __initconst = {
+	.mux = 24,
+	.muxmask = BIT(0),
+	.table = &sun9i_a80_apb1_config,
+	.getter = sun9i_a80_get_apb1_factors,
+};
+
+static DEFINE_SPINLOCK(sun9i_a80_apb1_lock);
+
+static void __init sun9i_a80_apb1_setup(struct device_node *node)
+{
+	sunxi_factors_register(node, &sun9i_a80_apb1_data, &sun9i_a80_apb1_lock);
+}
+CLK_OF_DECLARE(sun9i_a80_apb1, "allwinner,sun9i-a80-apb1-clk", sun9i_a80_apb1_setup);
diff --git a/drivers/clk/sunxi/clk-sunxi.c b/drivers/clk/sunxi/clk-sunxi.c
index b654b7b1d137..570202582dcf 100644
--- a/drivers/clk/sunxi/clk-sunxi.c
+++ b/drivers/clk/sunxi/clk-sunxi.c
@@ -19,6 +19,7 @@
 #include <linux/of.h>
 #include <linux/of_address.h>
 #include <linux/reset-controller.h>
+#include <linux/spinlock.h>
 
 #include "clk-factors.h"
 
@@ -244,9 +245,9 @@ static void sun4i_get_pll5_factors(u32 *freq, u32 parent_rate,
 }
 
 /**
- * sun6i_a31_get_pll6_factors() - calculates n, k factors for A31 PLL6
- * PLL6 rate is calculated as follows
- * rate = parent_rate * n * (k + 1) / 2
+ * sun6i_a31_get_pll6_factors() - calculates n, k factors for A31 PLL6x2
+ * PLL6x2 rate is calculated as follows
+ * rate = parent_rate * (n + 1) * (k + 1)
  * parent_rate is always 24Mhz
  */
 
@@ -255,13 +256,7 @@ static void sun6i_a31_get_pll6_factors(u32 *freq, u32 parent_rate,
 {
 	u8 div;
 
-	/*
-	 * We always have 24MHz / 2, so we can just say that our
-	 * parent clock is 12MHz.
-	 */
-	parent_rate = parent_rate / 2;
-
-	/* Normalize value to a parent_rate multiple (24M / 2) */
+	/* Normalize value to a parent_rate multiple (24M) */
 	div = *freq / parent_rate;
 	*freq = parent_rate * div;
 
@@ -273,7 +268,7 @@ static void sun6i_a31_get_pll6_factors(u32 *freq, u32 parent_rate,
 	if (*k > 3)
 		*k = 3;
 
-	*n = DIV_ROUND_UP(div, (*k+1));
+	*n = DIV_ROUND_UP(div, (*k+1)) - 1;
 }
 
 /**
@@ -319,46 +314,6 @@ static void sun4i_get_apb1_factors(u32 *freq, u32 parent_rate,
 
 
 
-/**
- * sun4i_get_mod0_factors() - calculates m, n factors for MOD0-style clocks
- * MOD0 rate is calculated as follows
- * rate = (parent_rate >> p) / (m + 1);
- */
-
-static void sun4i_get_mod0_factors(u32 *freq, u32 parent_rate,
-				   u8 *n, u8 *k, u8 *m, u8 *p)
-{
-	u8 div, calcm, calcp;
-
-	/* These clocks can only divide, so we will never be able to achieve
-	 * frequencies higher than the parent frequency */
-	if (*freq > parent_rate)
-		*freq = parent_rate;
-
-	div = DIV_ROUND_UP(parent_rate, *freq);
-
-	if (div < 16)
-		calcp = 0;
-	else if (div / 2 < 16)
-		calcp = 1;
-	else if (div / 4 < 16)
-		calcp = 2;
-	else
-		calcp = 3;
-
-	calcm = DIV_ROUND_UP(div, 1 << calcp);
-
-	*freq = (parent_rate >> calcp) / calcm;
-
-	/* we were called to round the frequency, we can now return */
-	if (n == NULL)
-		return;
-
-	*m = calcm - 1;
-	*p = calcp;
-}
-
-
 
 /**
  * sun7i_a20_get_out_factors() - calculates m, p factors for CLK_OUT_A/B
@@ -440,16 +395,6 @@ EXPORT_SYMBOL(clk_sunxi_mmc_phase_control);
  * sunxi_factors_clk_setup() - Setup function for factor clocks
  */
 
-#define SUNXI_FACTORS_MUX_MASK 0x3
-
-struct factors_data {
-	int enable;
-	int mux;
-	struct clk_factors_config *table;
-	void (*getter) (u32 *rate, u32 parent_rate, u8 *n, u8 *k, u8 *m, u8 *p);
-	const char *name;
-};
-
 static struct clk_factors_config sun4i_pll1_config = {
 	.nshift = 8,
 	.nwidth = 5,
@@ -494,6 +439,7 @@ static struct clk_factors_config sun6i_a31_pll6_config = {
 	.nwidth = 5,
 	.kshift = 4,
 	.kwidth = 2,
+	.n_start = 1,
 };
 
 static struct clk_factors_config sun4i_apb1_config = {
@@ -504,14 +450,6 @@ static struct clk_factors_config sun4i_apb1_config = {
 };
 
 /* user manual says "n" but it's really "p" */
-static struct clk_factors_config sun4i_mod0_config = {
-	.mshift = 0,
-	.mwidth = 4,
-	.pshift = 16,
-	.pwidth = 2,
-};
-
-/* user manual says "n" but it's really "p" */
 static struct clk_factors_config sun7i_a20_out_config = {
 	.mshift = 8,
 	.mwidth = 5,
@@ -561,111 +499,28 @@ static const struct factors_data sun6i_a31_pll6_data __initconst = {
 	.enable = 31,
 	.table = &sun6i_a31_pll6_config,
 	.getter = sun6i_a31_get_pll6_factors,
+	.name = "pll6x2",
 };
 
 static const struct factors_data sun4i_apb1_data __initconst = {
+	.mux = 24,
+	.muxmask = BIT(1) | BIT(0),
 	.table = &sun4i_apb1_config,
 	.getter = sun4i_get_apb1_factors,
 };
 
-static const struct factors_data sun4i_mod0_data __initconst = {
-	.enable = 31,
-	.mux = 24,
-	.table = &sun4i_mod0_config,
-	.getter = sun4i_get_mod0_factors,
-};
-
 static const struct factors_data sun7i_a20_out_data __initconst = {
 	.enable = 31,
 	.mux = 24,
+	.muxmask = BIT(1) | BIT(0),
 	.table = &sun7i_a20_out_config,
 	.getter = sun7i_a20_get_out_factors,
 };
 
 static struct clk * __init sunxi_factors_clk_setup(struct device_node *node,
-						const struct factors_data *data)
+						   const struct factors_data *data)
 {
-	struct clk *clk;
-	struct clk_factors *factors;
-	struct clk_gate *gate = NULL;
-	struct clk_mux *mux = NULL;
-	struct clk_hw *gate_hw = NULL;
-	struct clk_hw *mux_hw = NULL;
-	const char *clk_name = node->name;
-	const char *parents[SUNXI_MAX_PARENTS];
-	void __iomem *reg;
-	int i = 0;
-
-	reg = of_iomap(node, 0);
-
-	/* if we have a mux, we will have >1 parents */
-	while (i < SUNXI_MAX_PARENTS &&
-	       (parents[i] = of_clk_get_parent_name(node, i)) != NULL)
-		i++;
-
-	/*
-	 * some factor clocks, such as pll5 and pll6, may have multiple
-	 * outputs, and have their name designated in factors_data
-	 */
-	if (data->name)
-		clk_name = data->name;
-	else
-		of_property_read_string(node, "clock-output-names", &clk_name);
-
-	factors = kzalloc(sizeof(struct clk_factors), GFP_KERNEL);
-	if (!factors)
-		return NULL;
-
-	/* Add a gate if this factor clock can be gated */
-	if (data->enable) {
-		gate = kzalloc(sizeof(struct clk_gate), GFP_KERNEL);
-		if (!gate) {
-			kfree(factors);
-			return NULL;
-		}
-
-		/* set up gate properties */
-		gate->reg = reg;
-		gate->bit_idx = data->enable;
-		gate->lock = &clk_lock;
-		gate_hw = &gate->hw;
-	}
-
-	/* Add a mux if this factor clock can be muxed */
-	if (data->mux) {
-		mux = kzalloc(sizeof(struct clk_mux), GFP_KERNEL);
-		if (!mux) {
-			kfree(factors);
-			kfree(gate);
-			return NULL;
-		}
-
-		/* set up gate properties */
-		mux->reg = reg;
-		mux->shift = data->mux;
-		mux->mask = SUNXI_FACTORS_MUX_MASK;
-		mux->lock = &clk_lock;
-		mux_hw = &mux->hw;
-	}
-
-	/* set up factors properties */
-	factors->reg = reg;
-	factors->config = data->table;
-	factors->get_factors = data->getter;
-	factors->lock = &clk_lock;
-
-	clk = clk_register_composite(NULL, clk_name,
-			parents, i,
-			mux_hw, &clk_mux_ops,
-			&factors->hw, &clk_factors_ops,
-			gate_hw, &clk_gate_ops, 0);
-
-	if (!IS_ERR(clk)) {
-		of_clk_add_provider(node, of_clk_src_simple_get, clk);
-		clk_register_clkdev(clk, clk_name, NULL);
-	}
-
-	return clk;
+	return sunxi_factors_register(node, data, &clk_lock);
 }
 
 
@@ -688,10 +543,6 @@ static const struct mux_data sun6i_a31_ahb1_mux_data __initconst = {
 	.shift = 12,
 };
 
-static const struct mux_data sun4i_apb1_mux_data __initconst = {
-	.shift = 24,
-};
-
 static void __init sunxi_mux_clk_setup(struct device_node *node,
 				       struct mux_data *data)
 {
@@ -762,16 +613,19 @@ static const struct div_data sun4i_ahb_data __initconst = {
 	.width	= 2,
 };
 
+static const struct clk_div_table sun4i_apb0_table[] __initconst = {
+	{ .val = 0, .div = 2 },
+	{ .val = 1, .div = 2 },
+	{ .val = 2, .div = 4 },
+	{ .val = 3, .div = 8 },
+	{ } /* sentinel */
+};
+
 static const struct div_data sun4i_apb0_data __initconst = {
 	.shift	= 8,
 	.pow	= 1,
 	.width	= 2,
-};
-
-static const struct div_data sun6i_a31_apb2_div_data __initconst = {
-	.shift	= 0,
-	.pow	= 0,
-	.width	= 4,
+	.table	= sun4i_apb0_table,
 };
 
 static void __init sunxi_divider_clk_setup(struct device_node *node,
@@ -892,6 +746,18 @@ static const struct gates_data sun8i_a23_ahb1_gates_data __initconst = {
 	.mask = {0x25386742, 0x2505111},
 };
 
+static const struct gates_data sun9i_a80_ahb0_gates_data __initconst = {
+	.mask = {0xF5F12B},
+};
+
+static const struct gates_data sun9i_a80_ahb1_gates_data __initconst = {
+	.mask = {0x1E20003},
+};
+
+static const struct gates_data sun9i_a80_ahb2_gates_data __initconst = {
+	.mask = {0x9B7},
+};
+
 static const struct gates_data sun4i_apb0_gates_data __initconst = {
 	.mask = {0x4EF},
 };
@@ -908,6 +774,10 @@ static const struct gates_data sun7i_a20_apb0_gates_data __initconst = {
 	.mask = { 0x4ff },
 };
 
+static const struct gates_data sun9i_a80_apb0_gates_data __initconst = {
+	.mask = {0xEB822},
+};
+
 static const struct gates_data sun4i_apb1_gates_data __initconst = {
 	.mask = {0xFF00F7},
 };
@@ -936,6 +806,10 @@ static const struct gates_data sun7i_a20_apb1_gates_data __initconst = {
 	.mask = { 0xff80ff },
 };
 
+static const struct gates_data sun9i_a80_apb1_gates_data __initconst = {
+	.mask = {0x3F001F},
+};
+
 static const struct gates_data sun8i_a23_apb2_gates_data __initconst = {
 	.mask = {0x1F0007},
 };
@@ -1028,6 +902,7 @@ static void __init sunxi_gates_clk_setup(struct device_node *node,
 
 struct divs_data {
 	const struct factors_data *factors; /* data for the factor clock */
+	int ndivs; /* number of children */
 	struct {
 		u8 fixed; /* is it a fixed divisor? if not... */
 		struct clk_div_table *table; /* is it a table based divisor? */
@@ -1047,6 +922,7 @@ static struct clk_div_table pll6_sata_tbl[] = {
 
 static const struct divs_data pll5_divs_data __initconst = {
 	.factors = &sun4i_pll5_data,
+	.ndivs = 2,
 	.div = {
 		{ .shift = 0, .pow = 0, }, /* M, DDR */
 		{ .shift = 16, .pow = 1, }, /* P, other */
@@ -1055,12 +931,21 @@ static const struct divs_data pll5_divs_data __initconst = {
 
 static const struct divs_data pll6_divs_data __initconst = {
 	.factors = &sun4i_pll6_data,
+	.ndivs = 2,
 	.div = {
 		{ .shift = 0, .table = pll6_sata_tbl, .gate = 14 }, /* M, SATA */
 		{ .fixed = 2 }, /* P, other */
 	}
 };
 
+static const struct divs_data sun6i_a31_pll6_divs_data __initconst = {
+	.factors = &sun6i_a31_pll6_data,
+	.ndivs = 1,
+	.div = {
+		{ .fixed = 2 }, /* normal output */
+	}
+};
+
 /**
  * sunxi_divs_clk_setup() - Setup function for leaf divisors on clocks
  *
@@ -1085,7 +970,7 @@ static void __init sunxi_divs_clk_setup(struct device_node *node,
 	struct clk_fixed_factor *fix_factor;
 	struct clk_divider *divider;
 	void __iomem *reg;
-	int i = 0;
+	int ndivs = SUNXI_DIVS_MAX_QTY, i = 0;
 	int flags, clkflags;
 
 	/* Set up factor clock that we will be dividing */
@@ -1108,7 +993,11 @@ static void __init sunxi_divs_clk_setup(struct device_node *node,
 	 * our RAM clock! */
 	clkflags = !strcmp("pll5", parent) ? 0 : CLK_SET_RATE_PARENT;
 
-	for (i = 0; i < SUNXI_DIVS_MAX_QTY; i++) {
+	/* if number of children known, use it */
+	if (data->ndivs)
+		ndivs = data->ndivs;
+
+	for (i = 0; i < ndivs; i++) {
 		if (of_property_read_string_index(node, "clock-output-names",
 						  i, &clk_name) != 0)
 			break;
@@ -1197,9 +1086,7 @@ static const struct of_device_id clk_factors_match[] __initconst = {
 	{.compatible = "allwinner,sun6i-a31-pll1-clk", .data = &sun6i_a31_pll1_data,},
 	{.compatible = "allwinner,sun8i-a23-pll1-clk", .data = &sun8i_a23_pll1_data,},
 	{.compatible = "allwinner,sun7i-a20-pll4-clk", .data = &sun7i_a20_pll4_data,},
-	{.compatible = "allwinner,sun6i-a31-pll6-clk", .data = &sun6i_a31_pll6_data,},
 	{.compatible = "allwinner,sun4i-a10-apb1-clk", .data = &sun4i_apb1_data,},
-	{.compatible = "allwinner,sun4i-a10-mod0-clk", .data = &sun4i_mod0_data,},
 	{.compatible = "allwinner,sun7i-a20-out-clk", .data = &sun7i_a20_out_data,},
 	{}
 };
@@ -1210,7 +1097,6 @@ static const struct of_device_id clk_div_match[] __initconst = {
 	{.compatible = "allwinner,sun8i-a23-axi-clk", .data = &sun8i_a23_axi_data,},
 	{.compatible = "allwinner,sun4i-a10-ahb-clk", .data = &sun4i_ahb_data,},
 	{.compatible = "allwinner,sun4i-a10-apb0-clk", .data = &sun4i_apb0_data,},
-	{.compatible = "allwinner,sun6i-a31-apb2-div-clk", .data = &sun6i_a31_apb2_div_data,},
 	{}
 };
 
@@ -1218,13 +1104,13 @@ static const struct of_device_id clk_div_match[] __initconst = {
 static const struct of_device_id clk_divs_match[] __initconst = {
 	{.compatible = "allwinner,sun4i-a10-pll5-clk", .data = &pll5_divs_data,},
 	{.compatible = "allwinner,sun4i-a10-pll6-clk", .data = &pll6_divs_data,},
+	{.compatible = "allwinner,sun6i-a31-pll6-clk", .data = &sun6i_a31_pll6_divs_data,},
 	{}
 };
 
 /* Matches for mux clocks */
 static const struct of_device_id clk_mux_match[] __initconst = {
 	{.compatible = "allwinner,sun4i-a10-cpu-clk", .data = &sun4i_cpu_mux_data,},
-	{.compatible = "allwinner,sun4i-a10-apb1-mux-clk", .data = &sun4i_apb1_mux_data,},
 	{.compatible = "allwinner,sun6i-a31-ahb1-mux-clk", .data = &sun6i_a31_ahb1_mux_data,},
 	{}
 };
@@ -1238,16 +1124,21 @@ static const struct of_device_id clk_gates_match[] __initconst = {
 	{.compatible = "allwinner,sun6i-a31-ahb1-gates-clk", .data = &sun6i_a31_ahb1_gates_data,},
 	{.compatible = "allwinner,sun7i-a20-ahb-gates-clk", .data = &sun7i_a20_ahb_gates_data,},
 	{.compatible = "allwinner,sun8i-a23-ahb1-gates-clk", .data = &sun8i_a23_ahb1_gates_data,},
+	{.compatible = "allwinner,sun9i-a80-ahb0-gates-clk", .data = &sun9i_a80_ahb0_gates_data,},
+	{.compatible = "allwinner,sun9i-a80-ahb1-gates-clk", .data = &sun9i_a80_ahb1_gates_data,},
+	{.compatible = "allwinner,sun9i-a80-ahb2-gates-clk", .data = &sun9i_a80_ahb2_gates_data,},
 	{.compatible = "allwinner,sun4i-a10-apb0-gates-clk", .data = &sun4i_apb0_gates_data,},
 	{.compatible = "allwinner,sun5i-a10s-apb0-gates-clk", .data = &sun5i_a10s_apb0_gates_data,},
 	{.compatible = "allwinner,sun5i-a13-apb0-gates-clk", .data = &sun5i_a13_apb0_gates_data,},
 	{.compatible = "allwinner,sun7i-a20-apb0-gates-clk", .data = &sun7i_a20_apb0_gates_data,},
+	{.compatible = "allwinner,sun9i-a80-apb0-gates-clk", .data = &sun9i_a80_apb0_gates_data,},
 	{.compatible = "allwinner,sun4i-a10-apb1-gates-clk", .data = &sun4i_apb1_gates_data,},
 	{.compatible = "allwinner,sun5i-a10s-apb1-gates-clk", .data = &sun5i_a10s_apb1_gates_data,},
 	{.compatible = "allwinner,sun5i-a13-apb1-gates-clk", .data = &sun5i_a13_apb1_gates_data,},
 	{.compatible = "allwinner,sun6i-a31-apb1-gates-clk", .data = &sun6i_a31_apb1_gates_data,},
 	{.compatible = "allwinner,sun7i-a20-apb1-gates-clk", .data = &sun7i_a20_apb1_gates_data,},
 	{.compatible = "allwinner,sun8i-a23-apb1-gates-clk", .data = &sun8i_a23_apb1_gates_data,},
+	{.compatible = "allwinner,sun9i-a80-apb1-gates-clk", .data = &sun9i_a80_apb1_gates_data,},
 	{.compatible = "allwinner,sun6i-a31-apb2-gates-clk", .data = &sun6i_a31_apb2_gates_data,},
 	{.compatible = "allwinner,sun8i-a23-apb2-gates-clk", .data = &sun8i_a23_apb2_gates_data,},
 	{.compatible = "allwinner,sun4i-a10-usb-clk", .data = &sun4i_a10_usb_gates_data,},
@@ -1311,7 +1202,6 @@ static void __init sun4i_a10_init_clocks(struct device_node *node)
 CLK_OF_DECLARE(sun4i_a10_clk_init, "allwinner,sun4i-a10", sun4i_a10_init_clocks);
 
 static const char *sun5i_critical_clocks[] __initdata = {
-	"mbus",
 	"pll5_ddr",
 	"ahb_sdram",
 };
@@ -1337,3 +1227,9 @@ static void __init sun6i_init_clocks(struct device_node *node)
 }
 CLK_OF_DECLARE(sun6i_a31_clk_init, "allwinner,sun6i-a31", sun6i_init_clocks);
 CLK_OF_DECLARE(sun8i_a23_clk_init, "allwinner,sun8i-a23", sun6i_init_clocks);
+
+static void __init sun9i_init_clocks(struct device_node *node)
+{
+	sunxi_init_clocks(NULL, 0);
+}
+CLK_OF_DECLARE(sun9i_a80_clk_init, "allwinner,sun9i-a80", sun9i_init_clocks);
diff --git a/drivers/clk/tegra/clk-divider.c b/drivers/clk/tegra/clk-divider.c
index 290f9c1a3749..59a5714dfe18 100644
--- a/drivers/clk/tegra/clk-divider.c
+++ b/drivers/clk/tegra/clk-divider.c
@@ -185,3 +185,16 @@ struct clk *tegra_clk_register_divider(const char *name,
 
 	return clk;
 }
+
+static const struct clk_div_table mc_div_table[] = {
+	{ .val = 0, .div = 2 },
+	{ .val = 1, .div = 1 },
+	{ .val = 0, .div = 0 },
+};
+
+struct clk *tegra_clk_register_mc(const char *name, const char *parent_name,
+				  void __iomem *reg, spinlock_t *lock)
+{
+	return clk_register_divider_table(NULL, name, parent_name, 0, reg,
+					  16, 1, 0, mc_div_table, lock);
+}
diff --git a/drivers/clk/tegra/clk-tegra114.c b/drivers/clk/tegra/clk-tegra114.c
index f760f31d05c4..0b03d2cf7264 100644
--- a/drivers/clk/tegra/clk-tegra114.c
+++ b/drivers/clk/tegra/clk-tegra114.c
@@ -173,6 +173,7 @@ static DEFINE_SPINLOCK(pll_d_lock);
 static DEFINE_SPINLOCK(pll_d2_lock);
 static DEFINE_SPINLOCK(pll_u_lock);
 static DEFINE_SPINLOCK(pll_re_lock);
+static DEFINE_SPINLOCK(emc_lock);
 
 static struct div_nmp pllxc_nmp = {
 	.divm_shift = 0,
@@ -1228,7 +1229,11 @@ static __init void tegra114_periph_clk_init(void __iomem *clk_base,
 			       ARRAY_SIZE(mux_pllmcp_clkm),
 			       CLK_SET_RATE_NO_REPARENT,
 			       clk_base + CLK_SOURCE_EMC,
-			       29, 3, 0, NULL);
+			       29, 3, 0, &emc_lock);
+
+	clk = tegra_clk_register_mc("mc", "emc_mux", clk_base + CLK_SOURCE_EMC,
+				    &emc_lock);
+	clks[TEGRA114_CLK_MC] = clk;
 
 	for (i = 0; i < ARRAY_SIZE(tegra_periph_clk_list); i++) {
 		data = &tegra_periph_clk_list[i];
diff --git a/drivers/clk/tegra/clk-tegra124.c b/drivers/clk/tegra/clk-tegra124.c
index 9525c684d149..f5f9baca7bb6 100644
--- a/drivers/clk/tegra/clk-tegra124.c
+++ b/drivers/clk/tegra/clk-tegra124.c
@@ -132,6 +132,7 @@ static DEFINE_SPINLOCK(pll_d2_lock);
 static DEFINE_SPINLOCK(pll_e_lock);
 static DEFINE_SPINLOCK(pll_re_lock);
 static DEFINE_SPINLOCK(pll_u_lock);
+static DEFINE_SPINLOCK(emc_lock);
 
 /* possible OSC frequencies in Hz */
 static unsigned long tegra124_input_freq[] = {
@@ -1127,7 +1128,11 @@ static __init void tegra124_periph_clk_init(void __iomem *clk_base,
 	clk = clk_register_mux(NULL, "emc_mux", mux_pllmcp_clkm,
 			       ARRAY_SIZE(mux_pllmcp_clkm), 0,
 			       clk_base + CLK_SOURCE_EMC,
-			       29, 3, 0, NULL);
+			       29, 3, 0, &emc_lock);
+
+	clk = tegra_clk_register_mc("mc", "emc_mux", clk_base + CLK_SOURCE_EMC,
+				    &emc_lock);
+	clks[TEGRA124_CLK_MC] = clk;
 
 	/* cml0 */
 	clk = clk_register_gate(NULL, "cml0", "pll_e", 0, clk_base + PLLE_AUX,
@@ -1166,6 +1171,12 @@ static void __init tegra124_pll_init(void __iomem *clk_base,
 	clk_register_clkdev(clk, "pll_c_out1", NULL);
 	clks[TEGRA124_CLK_PLL_C_OUT1] = clk;
 
+	/* PLLC_UD */
+	clk = clk_register_fixed_factor(NULL, "pll_c_ud", "pll_c",
+					CLK_SET_RATE_PARENT, 1, 1);
+	clk_register_clkdev(clk, "pll_c_ud", NULL);
+	clks[TEGRA124_CLK_PLL_C_UD] = clk;
+
 	/* PLLC2 */
 	clk = tegra_clk_register_pllc("pll_c2", "pll_ref", clk_base, pmc, 0,
 			     &pll_c2_params, NULL);
@@ -1198,6 +1209,8 @@ static void __init tegra124_pll_init(void __iomem *clk_base,
 	/* PLLM_UD */
 	clk = clk_register_fixed_factor(NULL, "pll_m_ud", "pll_m",
 					CLK_SET_RATE_PARENT, 1, 1);
+	clk_register_clkdev(clk, "pll_m_ud", NULL);
+	clks[TEGRA124_CLK_PLL_M_UD] = clk;
 
 	/* PLLU */
 	val = readl(clk_base + pll_u_params.base_reg);
diff --git a/drivers/clk/tegra/clk-tegra20.c b/drivers/clk/tegra/clk-tegra20.c
index dace2b1b5ae6..41272dcc9e22 100644
--- a/drivers/clk/tegra/clk-tegra20.c
+++ b/drivers/clk/tegra/clk-tegra20.c
@@ -140,6 +140,8 @@ static struct cpu_clk_suspend_context {
 static void __iomem *clk_base;
 static void __iomem *pmc_base;
 
+static DEFINE_SPINLOCK(emc_lock);
+
 #define TEGRA_INIT_DATA_MUX(_name, _parents, _offset,	\
 			    _clk_num, _gate_flags, _clk_id)	\
 	TEGRA_INIT_DATA(_name, NULL, NULL, _parents, _offset,	\
@@ -819,11 +821,15 @@ static void __init tegra20_periph_clk_init(void)
 			       ARRAY_SIZE(mux_pllmcp_clkm),
 			       CLK_SET_RATE_NO_REPARENT,
 			       clk_base + CLK_SOURCE_EMC,
-			       30, 2, 0, NULL);
+			       30, 2, 0, &emc_lock);
 	clk = tegra_clk_register_periph_gate("emc", "emc_mux", 0, clk_base, 0,
 				    57, periph_clk_enb_refcnt);
 	clks[TEGRA20_CLK_EMC] = clk;
 
+	clk = tegra_clk_register_mc("mc", "emc_mux", clk_base + CLK_SOURCE_EMC,
+				    &emc_lock);
+	clks[TEGRA20_CLK_MC] = clk;
+
 	/* dsi */
 	clk = tegra_clk_register_periph_gate("dsi", "pll_d", 0, clk_base, 0,
 				    48, periph_clk_enb_refcnt);
diff --git a/drivers/clk/tegra/clk-tegra30.c b/drivers/clk/tegra/clk-tegra30.c
index 5bbacd01094f..4b9d8bd3d0bf 100644
--- a/drivers/clk/tegra/clk-tegra30.c
+++ b/drivers/clk/tegra/clk-tegra30.c
@@ -177,6 +177,7 @@ static unsigned long input_freq;
 
 static DEFINE_SPINLOCK(cml_lock);
 static DEFINE_SPINLOCK(pll_d_lock);
+static DEFINE_SPINLOCK(emc_lock);
 
 #define TEGRA_INIT_DATA_MUX(_name, _parents, _offset,	\
 			    _clk_num, _gate_flags, _clk_id)	\
@@ -1157,11 +1158,15 @@ static void __init tegra30_periph_clk_init(void)
 			       ARRAY_SIZE(mux_pllmcp_clkm),
 			       CLK_SET_RATE_NO_REPARENT,
 			       clk_base + CLK_SOURCE_EMC,
-			       30, 2, 0, NULL);
+			       30, 2, 0, &emc_lock);
 	clk = tegra_clk_register_periph_gate("emc", "emc_mux", 0, clk_base, 0,
 				    57, periph_clk_enb_refcnt);
 	clks[TEGRA30_CLK_EMC] = clk;
 
+	clk = tegra_clk_register_mc("mc", "emc_mux", clk_base + CLK_SOURCE_EMC,
+				    &emc_lock);
+	clks[TEGRA30_CLK_MC] = clk;
+
 	/* cml0 */
 	clk = clk_register_gate(NULL, "cml0", "pll_e", 0, clk_base + PLLE_AUX,
 				0, 0, &cml_lock);
diff --git a/drivers/clk/tegra/clk.c b/drivers/clk/tegra/clk.c
index f87c609e8f72..97dc8595c3cd 100644
--- a/drivers/clk/tegra/clk.c
+++ b/drivers/clk/tegra/clk.c
@@ -207,8 +207,13 @@ void __init tegra_init_from_table(struct tegra_clk_init_table *tbl,
 
 	for (; tbl->clk_id < clk_max; tbl++) {
 		clk = clks[tbl->clk_id];
-		if (IS_ERR_OR_NULL(clk))
-			return;
+		if (IS_ERR_OR_NULL(clk)) {
+			pr_err("%s: invalid entry %ld in clks array for id %d\n",
+			       __func__, PTR_ERR(clk), tbl->clk_id);
+			WARN_ON(1);
+
+			continue;
+		}
 
 		if (tbl->parent_id < clk_max) {
 			struct clk *parent = clks[tbl->parent_id];
diff --git a/drivers/clk/tegra/clk.h b/drivers/clk/tegra/clk.h
index 16ec8d6bb87f..4e458aa8d45c 100644
--- a/drivers/clk/tegra/clk.h
+++ b/drivers/clk/tegra/clk.h
@@ -86,6 +86,8 @@ struct clk *tegra_clk_register_divider(const char *name,
 		const char *parent_name, void __iomem *reg,
 		unsigned long flags, u8 clk_divider_flags, u8 shift, u8 width,
 		u8 frac_width, spinlock_t *lock);
+struct clk *tegra_clk_register_mc(const char *name, const char *parent_name,
+				  void __iomem *reg, spinlock_t *lock);
 
 /*
  * Tegra PLL:
diff --git a/drivers/clk/ti/clk-dra7-atl.c b/drivers/clk/ti/clk-dra7-atl.c
index af29359677da..59bb4b39d12e 100644
--- a/drivers/clk/ti/clk-dra7-atl.c
+++ b/drivers/clk/ti/clk-dra7-atl.c
@@ -203,6 +203,7 @@ static void __init of_dra7_atl_clock_setup(struct device_node *node)
 
 	if (!IS_ERR(clk)) {
 		of_clk_add_provider(node, of_clk_src_simple_get, clk);
+		kfree(parent_names);
 		return;
 	}
 cleanup:
@@ -228,6 +229,7 @@ static int of_dra7_atl_clk_probe(struct platform_device *pdev)
 	cinfo->iobase = of_iomap(node, 0);
 	cinfo->dev = &pdev->dev;
 	pm_runtime_enable(cinfo->dev);
+	pm_runtime_irq_safe(cinfo->dev);
 
 	pm_runtime_get_sync(cinfo->dev);
 	atl_write(cinfo, DRA7_ATL_PCLKMUX_REG(0), DRA7_ATL_PCLKMUX);
@@ -301,7 +303,6 @@ MODULE_DEVICE_TABLE(of, of_dra7_atl_clk_match_tbl);
 static struct platform_driver dra7_atl_clk_driver = {
 	.driver = {
 		.name = "dra7-atl",
-		.owner = THIS_MODULE,
 		.of_match_table = of_dra7_atl_clk_match_tbl,
 	},
 	.probe = of_dra7_atl_clk_probe,
diff --git a/drivers/clk/ti/clk.c b/drivers/clk/ti/clk.c
index b1a6f7144f3f..337abe5909e1 100644
--- a/drivers/clk/ti/clk.c
+++ b/drivers/clk/ti/clk.c
@@ -25,8 +25,8 @@
 #undef pr_fmt
 #define pr_fmt(fmt) "%s: " fmt, __func__
 
-static int ti_dt_clk_memmap_index;
 struct ti_clk_ll_ops *ti_clk_ll_ops;
+static struct device_node *clocks_node_ptr[CLK_MAX_MEMMAPS];
 
 /**
  * ti_dt_clocks_register - register DT alias clocks during boot
@@ -108,9 +108,21 @@ void __iomem *ti_clk_get_reg_addr(struct device_node *node, int index)
 	struct clk_omap_reg *reg;
 	u32 val;
 	u32 tmp;
+	int i;
 
 	reg = (struct clk_omap_reg *)&tmp;
-	reg->index = ti_dt_clk_memmap_index;
+
+	for (i = 0; i < CLK_MAX_MEMMAPS; i++) {
+		if (clocks_node_ptr[i] == node->parent)
+			break;
+	}
+
+	if (i == CLK_MAX_MEMMAPS) {
+		pr_err("clk-provider not found for %s!\n", node->name);
+		return NULL;
+	}
+
+	reg->index = i;
 
 	if (of_property_read_u32_index(node, "reg", index, &val)) {
 		pr_err("%s must have reg[%d]!\n", node->name, index);
@@ -127,20 +139,14 @@ void __iomem *ti_clk_get_reg_addr(struct device_node *node, int index)
  * @parent: master node
  * @index: internal index for clk_reg_ops
  *
- * Initializes a master clock IP block and its child clock nodes.
- * Regmap is provided for accessing the register space for the
- * IP block and all the clocks under it.
+ * Initializes a master clock IP block. This basically sets up the
+ * mapping from clocks node to the memory map index. All the clocks
+ * are then initialized through the common of_clk_init call, and the
+ * clocks will access their memory maps based on the node layout.
  */
 void ti_dt_clk_init_provider(struct device_node *parent, int index)
 {
-	const struct of_device_id *match;
-	struct device_node *np;
 	struct device_node *clocks;
-	of_clk_init_cb_t clk_init_cb;
-	struct clk_init_item *retry;
-	struct clk_init_item *tmp;
-
-	ti_dt_clk_memmap_index = index;
 
 	/* get clocks for this parent */
 	clocks = of_get_child_by_name(parent, "clocks");
@@ -149,19 +155,31 @@ void ti_dt_clk_init_provider(struct device_node *parent, int index)
 		return;
 	}
 
-	for_each_child_of_node(clocks, np) {
-		match = of_match_node(&__clk_of_table, np);
-		if (!match)
-			continue;
-		clk_init_cb = (of_clk_init_cb_t)match->data;
-		pr_debug("%s: initializing: %s\n", __func__, np->name);
-		clk_init_cb(np);
-	}
+	/* add clocks node info */
+	clocks_node_ptr[index] = clocks;
+}
 
-	list_for_each_entry_safe(retry, tmp, &retry_list, link) {
-		pr_debug("retry-init: %s\n", retry->node->name);
-		retry->func(retry->hw, retry->node);
-		list_del(&retry->link);
-		kfree(retry);
+/**
+ * ti_dt_clk_init_retry_clks - init clocks from the retry list
+ *
+ * Initializes any clocks that have failed to initialize before,
+ * reasons being missing parent node(s) during earlier init. This
+ * typically happens only for DPLLs which need to have both of their
+ * parent clocks ready during init.
+ */
+void ti_dt_clk_init_retry_clks(void)
+{
+	struct clk_init_item *retry;
+	struct clk_init_item *tmp;
+	int retries = 5;
+
+	while (!list_empty(&retry_list) && retries) {
+		list_for_each_entry_safe(retry, tmp, &retry_list, link) {
+			pr_debug("retry-init: %s\n", retry->node->name);
+			retry->func(retry->hw, retry->node);
+			list_del(&retry->link);
+			kfree(retry);
+		}
+		retries--;
 	}
 }
diff --git a/drivers/clk/ti/clockdomain.c b/drivers/clk/ti/clockdomain.c
index f1e0038d76ac..b4c5faccaece 100644
--- a/drivers/clk/ti/clockdomain.c
+++ b/drivers/clk/ti/clockdomain.c
@@ -36,6 +36,11 @@ static void __init of_ti_clockdomain_setup(struct device_node *node)
 
 	for (i = 0; i < num_clks; i++) {
 		clk = of_clk_get(node, i);
+		if (IS_ERR(clk)) {
+			pr_err("%s: Failed get %s' clock nr %d (%ld)\n",
+			       __func__, node->full_name, i, PTR_ERR(clk));
+			continue;
+		}
 		if (__clk_get_flags(clk) & CLK_IS_BASIC) {
 			pr_warn("can't setup clkdm for basic clk %s\n",
 				__clk_get_name(clk));
diff --git a/drivers/clk/ti/divider.c b/drivers/clk/ti/divider.c
index a837f703be65..bff2b5b8ff59 100644
--- a/drivers/clk/ti/divider.c
+++ b/drivers/clk/ti/divider.c
@@ -300,8 +300,8 @@ static struct clk *_register_divider(struct device *dev, const char *name,
 	return clk;
 }
 
-static struct clk_div_table
-__init *ti_clk_get_div_table(struct device_node *node)
+static struct clk_div_table *
+__init ti_clk_get_div_table(struct device_node *node)
 {
 	struct clk_div_table *table;
 	const __be32 *divspec;
diff --git a/drivers/clk/ti/dpll.c b/drivers/clk/ti/dpll.c
index 79791e1bf282..85ac0dd501de 100644
--- a/drivers/clk/ti/dpll.c
+++ b/drivers/clk/ti/dpll.c
@@ -33,6 +33,9 @@ static const struct clk_ops dpll_m4xen_ck_ops = {
 	.recalc_rate	= &omap4_dpll_regm4xen_recalc,
 	.round_rate	= &omap4_dpll_regm4xen_round_rate,
 	.set_rate	= &omap3_noncore_dpll_set_rate,
+	.set_parent	= &omap3_noncore_dpll_set_parent,
+	.set_rate_and_parent	= &omap3_noncore_dpll_set_rate_and_parent,
+	.determine_rate	= &omap4_dpll_regm4xen_determine_rate,
 	.get_parent	= &omap2_init_dpll_parent,
 };
 #else
@@ -53,6 +56,9 @@ static const struct clk_ops dpll_ck_ops = {
 	.recalc_rate	= &omap3_dpll_recalc,
 	.round_rate	= &omap2_dpll_round_rate,
 	.set_rate	= &omap3_noncore_dpll_set_rate,
+	.set_parent	= &omap3_noncore_dpll_set_parent,
+	.set_rate_and_parent	= &omap3_noncore_dpll_set_rate_and_parent,
+	.determine_rate	= &omap3_noncore_dpll_determine_rate,
 	.get_parent	= &omap2_init_dpll_parent,
 };
 
@@ -61,6 +67,9 @@ static const struct clk_ops dpll_no_gate_ck_ops = {
 	.get_parent	= &omap2_init_dpll_parent,
 	.round_rate	= &omap2_dpll_round_rate,
 	.set_rate	= &omap3_noncore_dpll_set_rate,
+	.set_parent	= &omap3_noncore_dpll_set_parent,
+	.set_rate_and_parent	= &omap3_noncore_dpll_set_rate_and_parent,
+	.determine_rate	= &omap3_noncore_dpll_determine_rate,
 };
 #else
 static const struct clk_ops dpll_core_ck_ops = {};
@@ -97,6 +106,9 @@ static const struct clk_ops omap3_dpll_ck_ops = {
 	.get_parent	= &omap2_init_dpll_parent,
 	.recalc_rate	= &omap3_dpll_recalc,
 	.set_rate	= &omap3_noncore_dpll_set_rate,
+	.set_parent	= &omap3_noncore_dpll_set_parent,
+	.set_rate_and_parent	= &omap3_noncore_dpll_set_rate_and_parent,
+	.determine_rate	= &omap3_noncore_dpll_determine_rate,
 	.round_rate	= &omap2_dpll_round_rate,
 };
 
@@ -106,6 +118,9 @@ static const struct clk_ops omap3_dpll_per_ck_ops = {
 	.get_parent	= &omap2_init_dpll_parent,
 	.recalc_rate	= &omap3_dpll_recalc,
 	.set_rate	= &omap3_dpll4_set_rate,
+	.set_parent	= &omap3_noncore_dpll_set_parent,
+	.set_rate_and_parent	= &omap3_dpll4_set_rate_and_parent,
+	.determine_rate	= &omap3_noncore_dpll_determine_rate,
 	.round_rate	= &omap2_dpll_round_rate,
 };
 #endif
diff --git a/drivers/clk/ux500/abx500-clk.c b/drivers/clk/ux500/abx500-clk.c
index e7bd62cf60b3..3e5e05101302 100644
--- a/drivers/clk/ux500/abx500-clk.c
+++ b/drivers/clk/ux500/abx500-clk.c
@@ -121,7 +121,6 @@ static int abx500_clk_probe(struct platform_device *pdev)
 static struct platform_driver abx500_clk_driver = {
 	.driver = {
 		.name = "abx500-clk",
-		.owner = THIS_MODULE,
 	},
 	.probe	= abx500_clk_probe,
 };
diff --git a/drivers/clk/versatile/Makefile b/drivers/clk/versatile/Makefile
index 162e519cb0f9..8ff03744fe98 100644
--- a/drivers/clk/versatile/Makefile
+++ b/drivers/clk/versatile/Makefile
@@ -2,6 +2,5 @@
 obj-$(CONFIG_ICST)		+= clk-icst.o clk-versatile.o
 obj-$(CONFIG_INTEGRATOR_IMPD1)	+= clk-impd1.o
 obj-$(CONFIG_ARCH_REALVIEW)	+= clk-realview.o
-obj-$(CONFIG_ARCH_VEXPRESS)	+= clk-vexpress.o
 obj-$(CONFIG_CLK_SP810)		+= clk-sp810.o
 obj-$(CONFIG_CLK_VEXPRESS_OSC)	+= clk-vexpress-osc.o
diff --git a/drivers/clk/versatile/clk-vexpress-osc.c b/drivers/clk/versatile/clk-vexpress-osc.c
index 529a59c0fbfa..765f1e0eeeb2 100644
--- a/drivers/clk/versatile/clk-vexpress-osc.c
+++ b/drivers/clk/versatile/clk-vexpress-osc.c
@@ -70,7 +70,6 @@ static struct clk_ops vexpress_osc_ops = {
 
 static int vexpress_osc_probe(struct platform_device *pdev)
 {
-	struct clk_lookup *cl = pdev->dev.platform_data; /* Non-DT lookup */
 	struct clk_init_data init;
 	struct vexpress_osc *osc;
 	struct clk *clk;
@@ -106,12 +105,6 @@ static int vexpress_osc_probe(struct platform_device *pdev)
 
 	of_clk_add_provider(pdev->dev.of_node, of_clk_src_simple_get, clk);
 
-	/* Only happens for non-DT cases */
-	if (cl) {
-		cl->clk = clk;
-		clkdev_add(cl);
-	}
-
 	dev_dbg(&pdev->dev, "Registered clock '%s'\n", init.name);
 
 	return 0;
diff --git a/drivers/clk/versatile/clk-vexpress.c b/drivers/clk/versatile/clk-vexpress.c
deleted file mode 100644
index 2d5e1b4820e0..000000000000
--- a/drivers/clk/versatile/clk-vexpress.c
+++ /dev/null
@@ -1,86 +0,0 @@
-/*
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * Copyright (C) 2012 ARM Limited
- */
-
-#include <linux/amba/sp810.h>
-#include <linux/clkdev.h>
-#include <linux/clk-provider.h>
-#include <linux/err.h>
-#include <linux/vexpress.h>
-
-static struct clk *vexpress_sp810_timerclken[4];
-static DEFINE_SPINLOCK(vexpress_sp810_lock);
-
-static void __init vexpress_sp810_init(void __iomem *base)
-{
-	int i;
-
-	if (WARN_ON(!base))
-		return;
-
-	for (i = 0; i < ARRAY_SIZE(vexpress_sp810_timerclken); i++) {
-		char name[12];
-		const char *parents[] = {
-			"v2m:refclk32khz", /* REFCLK */
-			"v2m:refclk1mhz" /* TIMCLK */
-		};
-
-		snprintf(name, ARRAY_SIZE(name), "timerclken%d", i);
-
-		vexpress_sp810_timerclken[i] = clk_register_mux(NULL, name,
-				parents, 2, CLK_SET_RATE_NO_REPARENT,
-				base + SCCTRL, SCCTRL_TIMERENnSEL_SHIFT(i), 1,
-				0, &vexpress_sp810_lock);
-
-		if (WARN_ON(IS_ERR(vexpress_sp810_timerclken[i])))
-			break;
-	}
-}
-
-
-static const char * const vexpress_clk_24mhz_periphs[] __initconst = {
-	"mb:uart0", "mb:uart1", "mb:uart2", "mb:uart3",
-	"mb:mmci", "mb:kmi0", "mb:kmi1"
-};
-
-void __init vexpress_clk_init(void __iomem *sp810_base)
-{
-	struct clk *clk;
-	int i;
-
-	clk = clk_register_fixed_rate(NULL, "dummy_apb_pclk", NULL,
-			CLK_IS_ROOT, 0);
-	WARN_ON(clk_register_clkdev(clk, "apb_pclk", NULL));
-
-	clk = clk_register_fixed_rate(NULL, "v2m:clk_24mhz", NULL,
-			CLK_IS_ROOT, 24000000);
-	for (i = 0; i < ARRAY_SIZE(vexpress_clk_24mhz_periphs); i++)
-		WARN_ON(clk_register_clkdev(clk, NULL,
-				vexpress_clk_24mhz_periphs[i]));
-
-	clk = clk_register_fixed_rate(NULL, "v2m:refclk32khz", NULL,
-			CLK_IS_ROOT, 32768);
-	WARN_ON(clk_register_clkdev(clk, NULL, "v2m:wdt"));
-
-	clk = clk_register_fixed_rate(NULL, "v2m:refclk1mhz", NULL,
-			CLK_IS_ROOT, 1000000);
-
-	vexpress_sp810_init(sp810_base);
-
-	for (i = 0; i < ARRAY_SIZE(vexpress_sp810_timerclken); i++)
-		WARN_ON(clk_set_parent(vexpress_sp810_timerclken[i], clk));
-
-	WARN_ON(clk_register_clkdev(vexpress_sp810_timerclken[0],
-				"v2m-timer0", "sp804"));
-	WARN_ON(clk_register_clkdev(vexpress_sp810_timerclken[1],
-				"v2m-timer1", "sp804"));
-}
diff --git a/drivers/clk/x86/clk-lpt.c b/drivers/clk/x86/clk-lpt.c
index 812f83f8b0c6..f827083defc4 100644
--- a/drivers/clk/x86/clk-lpt.c
+++ b/drivers/clk/x86/clk-lpt.c
@@ -42,7 +42,6 @@ static int lpt_clk_probe(struct platform_device *pdev)
 static struct platform_driver lpt_clk_driver = {
 	.driver = {
 		.name = "clk-lpt",
-		.owner = THIS_MODULE,
 	},
 	.probe = lpt_clk_probe,
 };
diff --git a/drivers/clk/zynq/clkc.c b/drivers/clk/zynq/clkc.c
index 246cf1226eaa..9037bebd69f7 100644
--- a/drivers/clk/zynq/clkc.c
+++ b/drivers/clk/zynq/clkc.c
@@ -85,24 +85,22 @@ static DEFINE_SPINLOCK(canmioclk_lock);
 static DEFINE_SPINLOCK(dbgclk_lock);
 static DEFINE_SPINLOCK(aperclk_lock);
 
-static const char dummy_nm[] __initconst = "dummy_name";
-
-static const char *armpll_parents[] __initdata = {"armpll_int", "ps_clk"};
-static const char *ddrpll_parents[] __initdata = {"ddrpll_int", "ps_clk"};
-static const char *iopll_parents[] __initdata = {"iopll_int", "ps_clk"};
-static const char *gem0_mux_parents[] __initdata = {"gem0_div1", dummy_nm};
-static const char *gem1_mux_parents[] __initdata = {"gem1_div1", dummy_nm};
-static const char *can0_mio_mux2_parents[] __initdata = {"can0_gate",
+static const char *armpll_parents[] __initconst = {"armpll_int", "ps_clk"};
+static const char *ddrpll_parents[] __initconst = {"ddrpll_int", "ps_clk"};
+static const char *iopll_parents[] __initconst = {"iopll_int", "ps_clk"};
+static const char *gem0_mux_parents[] __initconst = {"gem0_div1", "dummy_name"};
+static const char *gem1_mux_parents[] __initconst = {"gem1_div1", "dummy_name"};
+static const char *can0_mio_mux2_parents[] __initconst = {"can0_gate",
 	"can0_mio_mux"};
-static const char *can1_mio_mux2_parents[] __initdata = {"can1_gate",
+static const char *can1_mio_mux2_parents[] __initconst = {"can1_gate",
 	"can1_mio_mux"};
-static const char *dbg_emio_mux_parents[] __initdata = {"dbg_div",
-	dummy_nm};
+static const char *dbg_emio_mux_parents[] __initconst = {"dbg_div",
+	"dummy_name"};
 
-static const char *dbgtrc_emio_input_names[] __initdata = {"trace_emio_clk"};
-static const char *gem0_emio_input_names[] __initdata = {"gem0_emio_clk"};
-static const char *gem1_emio_input_names[] __initdata = {"gem1_emio_clk"};
-static const char *swdt_ext_clk_input_names[] __initdata = {"swdt_ext_clk"};
+static const char *dbgtrc_emio_input_names[] __initconst = {"trace_emio_clk"};
+static const char *gem0_emio_input_names[] __initconst = {"gem0_emio_clk"};
+static const char *gem1_emio_input_names[] __initconst = {"gem1_emio_clk"};
+static const char *swdt_ext_clk_input_names[] __initconst = {"swdt_ext_clk"};
 
 static void __init zynq_clk_register_fclk(enum zynq_clk fclk,
 		const char *clk_name, void __iomem *fclk_ctrl_reg,
@@ -230,6 +228,7 @@ static void __init zynq_clk_setup(struct device_node *np)
 	const char *periph_parents[4];
 	const char *swdt_ext_clk_mux_parents[2];
 	const char *can_mio_mux_parents[NUM_MIO_PINS];
+	const char *dummy_nm = "dummy_name";
 
 	pr_info("Zynq clock init\n");
 
@@ -619,5 +618,4 @@ void __init zynq_clock_init(void)
 np_err:
 	of_node_put(np);
 	BUG();
-	return;
 }
diff --git a/drivers/clk/zynq/pll.c b/drivers/clk/zynq/pll.c
index cec97596fe65..00d72fb5c036 100644
--- a/drivers/clk/zynq/pll.c
+++ b/drivers/clk/zynq/pll.c
@@ -211,10 +211,8 @@ struct clk *clk_register_zynq_pll(const char *name, const char *parent,
 	};
 
 	pll = kmalloc(sizeof(*pll), GFP_KERNEL);
-	if (!pll) {
-		pr_err("%s: Could not allocate Zynq PLL clk.\n", __func__);
+	if (!pll)
 		return ERR_PTR(-ENOMEM);
-	}
 
 	/* Populate the struct */
 	pll->hw.init = &initd;
diff --git a/drivers/clocksource/Kconfig b/drivers/clocksource/Kconfig
index cfd6519df661..fc01ec27d3c8 100644
--- a/drivers/clocksource/Kconfig
+++ b/drivers/clocksource/Kconfig
@@ -30,6 +30,10 @@ config ARMADA_370_XP_TIMER
 	bool
 	select CLKSRC_OF
 
+config MESON6_TIMER
+	bool
+	select CLKSRC_MMIO
+
 config ORION_TIMER
 	select CLKSRC_OF
 	select CLKSRC_MMIO
@@ -120,6 +124,10 @@ config CLKSRC_ARM_GLOBAL_TIMER_SCHED_CLOCK
 	help
 	 Use ARM global timer clock source as sched_clock
 
+config ATMEL_PIT
+	select CLKSRC_OF if OF
+	def_bool SOC_AT91SAM9 || SOC_SAMA5
+
 config CLKSRC_METAG_GENERIC
 	def_bool y if METAG
 	help
@@ -216,4 +224,9 @@ config CLKSRC_VERSATILE
 	  ARM Versatile, RealView and Versatile Express reference
 	  platforms.
 
+config CLKSRC_MIPS_GIC
+	bool
+	depends on MIPS_GIC
+	select CLKSRC_OF
+
 endmenu
diff --git a/drivers/clocksource/Makefile b/drivers/clocksource/Makefile
index 7fd9fd1dff42..94d90b24b56b 100644
--- a/drivers/clocksource/Makefile
+++ b/drivers/clocksource/Makefile
@@ -1,4 +1,5 @@
 obj-$(CONFIG_CLKSRC_OF)	+= clksrc-of.o
+obj-$(CONFIG_ATMEL_PIT)		+= timer-atmel-pit.o
 obj-$(CONFIG_ATMEL_TCB_CLKSRC)	+= tcb_clksrc.o
 obj-$(CONFIG_X86_PM_TIMER)	+= acpi_pm.o
 obj-$(CONFIG_SCx200HR_TIMER)	+= scx200_hrt.o
@@ -25,6 +26,7 @@ obj-$(CONFIG_ARCH_PRIMA2)	+= timer-prima2.o
 obj-$(CONFIG_ARCH_U300)		+= timer-u300.o
 obj-$(CONFIG_SUN4I_TIMER)	+= sun4i_timer.o
 obj-$(CONFIG_SUN5I_HSTIMER)	+= timer-sun5i.o
+obj-$(CONFIG_MESON6_TIMER)	+= meson6_timer.o
 obj-$(CONFIG_ARCH_TEGRA)	+= tegra20_timer.o
 obj-$(CONFIG_VT8500_TIMER)	+= vt8500_timer.o
 obj-$(CONFIG_ARCH_NSPIRE)	+= zevio-timer.o
@@ -43,4 +45,6 @@ obj-$(CONFIG_ARM_GLOBAL_TIMER)		+= arm_global_timer.o
 obj-$(CONFIG_CLKSRC_METAG_GENERIC)	+= metag_generic.o
 obj-$(CONFIG_ARCH_HAS_TICK_BROADCAST)	+= dummy_timer.o
 obj-$(CONFIG_ARCH_KEYSTONE)		+= timer-keystone.o
+obj-$(CONFIG_ARCH_INTEGRATOR_AP)	+= timer-integrator-ap.o
 obj-$(CONFIG_CLKSRC_VERSATILE)		+= versatile.o
+obj-$(CONFIG_CLKSRC_MIPS_GIC)		+= mips-gic-timer.o
diff --git a/drivers/clocksource/arm_arch_timer.c b/drivers/clocksource/arm_arch_timer.c
index 5163ec13429d..095c1774592c 100644
--- a/drivers/clocksource/arm_arch_timer.c
+++ b/drivers/clocksource/arm_arch_timer.c
@@ -299,6 +299,21 @@ static void __arch_timer_setup(unsigned type,
 	clockevents_config_and_register(clk, arch_timer_rate, 0xf, 0x7fffffff);
 }
 
+static void arch_timer_evtstrm_enable(int divider)
+{
+	u32 cntkctl = arch_timer_get_cntkctl();
+
+	cntkctl &= ~ARCH_TIMER_EVT_TRIGGER_MASK;
+	/* Set the divider and enable virtual event stream */
+	cntkctl |= (divider << ARCH_TIMER_EVT_TRIGGER_SHIFT)
+			| ARCH_TIMER_VIRT_EVT_EN;
+	arch_timer_set_cntkctl(cntkctl);
+	elf_hwcap |= HWCAP_EVTSTRM;
+#ifdef CONFIG_COMPAT
+	compat_elf_hwcap |= COMPAT_HWCAP_EVTSTRM;
+#endif
+}
+
 static void arch_timer_configure_evtstream(void)
 {
 	int evt_stream_div, pos;
@@ -312,6 +327,23 @@ static void arch_timer_configure_evtstream(void)
 	arch_timer_evtstrm_enable(min(pos, 15));
 }
 
+static void arch_counter_set_user_access(void)
+{
+	u32 cntkctl = arch_timer_get_cntkctl();
+
+	/* Disable user access to the timers and the physical counter */
+	/* Also disable virtual event stream */
+	cntkctl &= ~(ARCH_TIMER_USR_PT_ACCESS_EN
+			| ARCH_TIMER_USR_VT_ACCESS_EN
+			| ARCH_TIMER_VIRT_EVT_EN
+			| ARCH_TIMER_USR_PCT_ACCESS_EN);
+
+	/* Enable user access to the virtual counter */
+	cntkctl |= ARCH_TIMER_USR_VCT_ACCESS_EN;
+
+	arch_timer_set_cntkctl(cntkctl);
+}
+
 static int arch_timer_setup(struct clock_event_device *clk)
 {
 	__arch_timer_setup(ARCH_CP15_TIMER, clk);
@@ -429,11 +461,22 @@ static void __init arch_counter_register(unsigned type)
 	u64 start_count;
 
 	/* Register the CP15 based counter if we have one */
-	if (type & ARCH_CP15_TIMER)
-		arch_timer_read_counter = arch_counter_get_cntvct;
-	else
+	if (type & ARCH_CP15_TIMER) {
+		if (IS_ENABLED(CONFIG_ARM64) || arch_timer_use_virtual)
+			arch_timer_read_counter = arch_counter_get_cntvct;
+		else
+			arch_timer_read_counter = arch_counter_get_cntpct;
+	} else {
 		arch_timer_read_counter = arch_counter_get_cntvct_mem;
 
+		/* If the clocksource name is "arch_sys_counter" the
+		 * VDSO will attempt to read the CP15-based counter.
+		 * Ensure this does not happen when CP15-based
+		 * counter is not available.
+		 */
+		clocksource_counter.name = "arch_mem_counter";
+	}
+
 	start_count = arch_timer_read_counter();
 	clocksource_register_hz(&clocksource_counter, arch_timer_rate);
 	cyclecounter.mult = clocksource_counter.mult;
@@ -616,17 +659,29 @@ static const struct of_device_id arch_timer_mem_of_match[] __initconst = {
 	{},
 };
 
+static bool __init
+arch_timer_probed(int type, const struct of_device_id *matches)
+{
+	struct device_node *dn;
+	bool probed = true;
+
+	dn = of_find_matching_node(NULL, matches);
+	if (dn && of_device_is_available(dn) && !(arch_timers_present & type))
+		probed = false;
+	of_node_put(dn);
+
+	return probed;
+}
+
 static void __init arch_timer_common_init(void)
 {
 	unsigned mask = ARCH_CP15_TIMER | ARCH_MEM_TIMER;
 
 	/* Wait until both nodes are probed if we have two timers */
 	if ((arch_timers_present & mask) != mask) {
-		if (of_find_matching_node(NULL, arch_timer_mem_of_match) &&
-				!(arch_timers_present & ARCH_MEM_TIMER))
+		if (!arch_timer_probed(ARCH_MEM_TIMER, arch_timer_mem_of_match))
 			return;
-		if (of_find_matching_node(NULL, arch_timer_of_match) &&
-				!(arch_timers_present & ARCH_CP15_TIMER))
+		if (!arch_timer_probed(ARCH_CP15_TIMER, arch_timer_of_match))
 			return;
 	}
 
@@ -650,6 +705,14 @@ static void __init arch_timer_init(struct device_node *np)
 	arch_timer_detect_rate(NULL, np);
 
 	/*
+	 * If we cannot rely on firmware initializing the timer registers then
+	 * we should use the physical timers instead.
+	 */
+	if (IS_ENABLED(CONFIG_ARM) &&
+	    of_property_read_bool(np, "arm,cpu-registers-not-fw-configured"))
+			arch_timer_use_virtual = false;
+
+	/*
 	 * If HYP mode is available, we know that the physical timer
 	 * has been configured to be accessible from PL1. Use it, so
 	 * that a guest can use the virtual timer instead.
diff --git a/drivers/clocksource/bcm_kona_timer.c b/drivers/clocksource/bcm_kona_timer.c
index 0595dc6c453e..f1e33d08dd83 100644
--- a/drivers/clocksource/bcm_kona_timer.c
+++ b/drivers/clocksource/bcm_kona_timer.c
@@ -68,9 +68,8 @@ static void kona_timer_disable_and_clear(void __iomem *base)
 }
 
 static void
-kona_timer_get_counter(void *timer_base, uint32_t *msw, uint32_t *lsw)
+kona_timer_get_counter(void __iomem *timer_base, uint32_t *msw, uint32_t *lsw)
 {
-	void __iomem *base = IOMEM(timer_base);
 	int loop_limit = 4;
 
 	/*
@@ -86,9 +85,9 @@ kona_timer_get_counter(void *timer_base, uint32_t *msw, uint32_t *lsw)
 	 */
 
 	while (--loop_limit) {
-		*msw = readl(base + KONA_GPTIMER_STCHI_OFFSET);
-		*lsw = readl(base + KONA_GPTIMER_STCLO_OFFSET);
-		if (*msw == readl(base + KONA_GPTIMER_STCHI_OFFSET))
+		*msw = readl(timer_base + KONA_GPTIMER_STCHI_OFFSET);
+		*lsw = readl(timer_base + KONA_GPTIMER_STCLO_OFFSET);
+		if (*msw == readl(timer_base + KONA_GPTIMER_STCHI_OFFSET))
 			break;
 	}
 	if (!loop_limit) {
diff --git a/drivers/clocksource/cadence_ttc_timer.c b/drivers/clocksource/cadence_ttc_timer.c
index 7a08811df9aa..510c8a1d37b3 100644
--- a/drivers/clocksource/cadence_ttc_timer.c
+++ b/drivers/clocksource/cadence_ttc_timer.c
@@ -25,7 +25,7 @@
 #include <linux/sched_clock.h>
 
 /*
- * This driver configures the 2 16-bit count-up timers as follows:
+ * This driver configures the 2 16/32-bit count-up timers as follows:
  *
  * T1: Timer 1, clocksource for generic timekeeping
  * T2: Timer 2, clockevent source for hrtimers
@@ -321,7 +321,8 @@ static int ttc_rate_change_clocksource_cb(struct notifier_block *nb,
 	return NOTIFY_DONE;
 }
 
-static void __init ttc_setup_clocksource(struct clk *clk, void __iomem *base)
+static void __init ttc_setup_clocksource(struct clk *clk, void __iomem *base,
+					 u32 timer_width)
 {
 	struct ttc_timer_clocksource *ttccs;
 	int err;
@@ -351,7 +352,7 @@ static void __init ttc_setup_clocksource(struct clk *clk, void __iomem *base)
 	ttccs->cs.name = "ttc_clocksource";
 	ttccs->cs.rating = 200;
 	ttccs->cs.read = __ttc_clocksource_read;
-	ttccs->cs.mask = CLOCKSOURCE_MASK(16);
+	ttccs->cs.mask = CLOCKSOURCE_MASK(timer_width);
 	ttccs->cs.flags = CLOCK_SOURCE_IS_CONTINUOUS;
 
 	/*
@@ -372,7 +373,8 @@ static void __init ttc_setup_clocksource(struct clk *clk, void __iomem *base)
 	}
 
 	ttc_sched_clock_val_reg = base + TTC_COUNT_VAL_OFFSET;
-	sched_clock_register(ttc_sched_clock_read, 16, ttccs->ttc.freq / PRESCALE);
+	sched_clock_register(ttc_sched_clock_read, timer_width,
+			     ttccs->ttc.freq / PRESCALE);
 }
 
 static int ttc_rate_change_clockevent_cb(struct notifier_block *nb,
@@ -467,6 +469,7 @@ static void __init ttc_timer_init(struct device_node *timer)
 	struct clk *clk_cs, *clk_ce;
 	static int initialized;
 	int clksel;
+	u32 timer_width = 16;
 
 	if (initialized)
 		return;
@@ -490,6 +493,8 @@ static void __init ttc_timer_init(struct device_node *timer)
 		BUG();
 	}
 
+	of_property_read_u32(timer, "timer-width", &timer_width);
+
 	clksel = readl_relaxed(timer_baseaddr + TTC_CLK_CNTRL_OFFSET);
 	clksel = !!(clksel & TTC_CLK_CNTRL_CSRC_MASK);
 	clk_cs = of_clk_get(timer, clksel);
@@ -506,7 +511,7 @@ static void __init ttc_timer_init(struct device_node *timer)
 		BUG();
 	}
 
-	ttc_setup_clocksource(clk_cs, timer_baseaddr);
+	ttc_setup_clocksource(clk_cs, timer_baseaddr, timer_width);
 	ttc_setup_clockevent(clk_ce, timer_baseaddr + 4, irq);
 
 	pr_info("%s #0 at %p, irq=%d\n", timer->name, timer_baseaddr, irq);
diff --git a/drivers/clocksource/dummy_timer.c b/drivers/clocksource/dummy_timer.c
index ad3572541728..31990600fcff 100644
--- a/drivers/clocksource/dummy_timer.c
+++ b/drivers/clocksource/dummy_timer.c
@@ -28,7 +28,7 @@ static void dummy_timer_set_mode(enum clock_event_mode mode,
 static void dummy_timer_setup(void)
 {
 	int cpu = smp_processor_id();
-	struct clock_event_device *evt = __this_cpu_ptr(&dummy_timer_evt);
+	struct clock_event_device *evt = raw_cpu_ptr(&dummy_timer_evt);
 
 	evt->name	= "dummy_timer";
 	evt->features	= CLOCK_EVT_FEAT_PERIODIC |
diff --git a/drivers/clocksource/exynos_mct.c b/drivers/clocksource/exynos_mct.c
index 9403061a2acc..83564c9cfdbe 100644
--- a/drivers/clocksource/exynos_mct.c
+++ b/drivers/clocksource/exynos_mct.c
@@ -97,8 +97,8 @@ static void exynos4_mct_write(unsigned int value, unsigned long offset)
 	writel_relaxed(value, reg_base + offset);
 
 	if (likely(offset >= EXYNOS4_MCT_L_BASE(0))) {
-		stat_addr = (offset & ~EXYNOS4_MCT_L_MASK) + MCT_L_WSTAT_OFFSET;
-		switch (offset & EXYNOS4_MCT_L_MASK) {
+		stat_addr = (offset & EXYNOS4_MCT_L_MASK) + MCT_L_WSTAT_OFFSET;
+		switch (offset & ~EXYNOS4_MCT_L_MASK) {
 		case MCT_L_TCON_OFFSET:
 			mask = 1 << 3;		/* L_TCON write status */
 			break;
diff --git a/drivers/clocksource/meson6_timer.c b/drivers/clocksource/meson6_timer.c
new file mode 100644
index 000000000000..5c15cba41dca
--- /dev/null
+++ b/drivers/clocksource/meson6_timer.c
@@ -0,0 +1,167 @@
+/*
+ * Amlogic Meson6 SoCs timer handling.
+ *
+ * Copyright (C) 2014 Carlo Caione <carlo@caione.org>
+ *
+ * Based on code from Amlogic, Inc
+ *
+ * This file is licensed under the terms of the GNU General Public
+ * License version 2.  This program is licensed "as is" without any
+ * warranty of any kind, whether express or implied.
+ */
+
+#include <linux/clk.h>
+#include <linux/clockchips.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/irqreturn.h>
+#include <linux/sched_clock.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+
+#define CED_ID			0
+#define CSD_ID			4
+
+#define TIMER_ISA_MUX		0
+#define TIMER_ISA_VAL(t)	(((t) + 1) << 2)
+
+#define TIMER_INPUT_BIT(t)	(2 * (t))
+#define TIMER_ENABLE_BIT(t)	(16 + (t))
+#define TIMER_PERIODIC_BIT(t)	(12 + (t))
+
+#define TIMER_CED_INPUT_MASK	(3UL << TIMER_INPUT_BIT(CED_ID))
+#define TIMER_CSD_INPUT_MASK	(7UL << TIMER_INPUT_BIT(CSD_ID))
+
+#define TIMER_CED_UNIT_1US	0
+#define TIMER_CSD_UNIT_1US	1
+
+static void __iomem *timer_base;
+
+static u64 notrace meson6_timer_sched_read(void)
+{
+	return (u64)readl(timer_base + TIMER_ISA_VAL(CSD_ID));
+}
+
+static void meson6_clkevt_time_stop(unsigned char timer)
+{
+	u32 val = readl(timer_base + TIMER_ISA_MUX);
+
+	writel(val & ~TIMER_ENABLE_BIT(timer), timer_base + TIMER_ISA_MUX);
+}
+
+static void meson6_clkevt_time_setup(unsigned char timer, unsigned long delay)
+{
+	writel(delay, timer_base + TIMER_ISA_VAL(timer));
+}
+
+static void meson6_clkevt_time_start(unsigned char timer, bool periodic)
+{
+	u32 val = readl(timer_base + TIMER_ISA_MUX);
+
+	if (periodic)
+		val |= TIMER_PERIODIC_BIT(timer);
+	else
+		val &= ~TIMER_PERIODIC_BIT(timer);
+
+	writel(val | TIMER_ENABLE_BIT(timer), timer_base + TIMER_ISA_MUX);
+}
+
+static void meson6_clkevt_mode(enum clock_event_mode mode,
+			       struct clock_event_device *clk)
+{
+	switch (mode) {
+	case CLOCK_EVT_MODE_PERIODIC:
+		meson6_clkevt_time_stop(CED_ID);
+		meson6_clkevt_time_setup(CED_ID, USEC_PER_SEC/HZ - 1);
+		meson6_clkevt_time_start(CED_ID, true);
+		break;
+	case CLOCK_EVT_MODE_ONESHOT:
+		meson6_clkevt_time_stop(CED_ID);
+		meson6_clkevt_time_start(CED_ID, false);
+		break;
+	case CLOCK_EVT_MODE_UNUSED:
+	case CLOCK_EVT_MODE_SHUTDOWN:
+	default:
+		meson6_clkevt_time_stop(CED_ID);
+		break;
+	}
+}
+
+static int meson6_clkevt_next_event(unsigned long evt,
+				    struct clock_event_device *unused)
+{
+	meson6_clkevt_time_stop(CED_ID);
+	meson6_clkevt_time_setup(CED_ID, evt);
+	meson6_clkevt_time_start(CED_ID, false);
+
+	return 0;
+}
+
+static struct clock_event_device meson6_clockevent = {
+	.name		= "meson6_tick",
+	.rating		= 400,
+	.features	= CLOCK_EVT_FEAT_PERIODIC | CLOCK_EVT_FEAT_ONESHOT,
+	.set_mode	= meson6_clkevt_mode,
+	.set_next_event	= meson6_clkevt_next_event,
+};
+
+static irqreturn_t meson6_timer_interrupt(int irq, void *dev_id)
+{
+	struct clock_event_device *evt = (struct clock_event_device *)dev_id;
+
+	evt->event_handler(evt);
+
+	return IRQ_HANDLED;
+}
+
+static struct irqaction meson6_timer_irq = {
+	.name		= "meson6_timer",
+	.flags		= IRQF_TIMER | IRQF_IRQPOLL,
+	.handler	= meson6_timer_interrupt,
+	.dev_id		= &meson6_clockevent,
+};
+
+static void __init meson6_timer_init(struct device_node *node)
+{
+	u32 val;
+	int ret, irq;
+
+	timer_base = of_io_request_and_map(node, 0, "meson6-timer");
+	if (IS_ERR(timer_base))
+		panic("Can't map registers");
+
+	irq = irq_of_parse_and_map(node, 0);
+	if (irq <= 0)
+		panic("Can't parse IRQ");
+
+	/* Set 1us for timer E */
+	val = readl(timer_base + TIMER_ISA_MUX);
+	val &= ~TIMER_CSD_INPUT_MASK;
+	val |= TIMER_CSD_UNIT_1US << TIMER_INPUT_BIT(CSD_ID);
+	writel(val, timer_base + TIMER_ISA_MUX);
+
+	sched_clock_register(meson6_timer_sched_read, 32, USEC_PER_SEC);
+	clocksource_mmio_init(timer_base + TIMER_ISA_VAL(CSD_ID), node->name,
+			      1000 * 1000, 300, 32, clocksource_mmio_readl_up);
+
+	/* Timer A base 1us */
+	val &= ~TIMER_CED_INPUT_MASK;
+	val |= TIMER_CED_UNIT_1US << TIMER_INPUT_BIT(CED_ID);
+	writel(val, timer_base + TIMER_ISA_MUX);
+
+	/* Stop the timer A */
+	meson6_clkevt_time_stop(CED_ID);
+
+	ret = setup_irq(irq, &meson6_timer_irq);
+	if (ret)
+		pr_warn("failed to setup irq %d\n", irq);
+
+	meson6_clockevent.cpumask = cpu_possible_mask;
+	meson6_clockevent.irq = irq;
+
+	clockevents_config_and_register(&meson6_clockevent, USEC_PER_SEC,
+					1, 0xfffe);
+}
+CLOCKSOURCE_OF_DECLARE(meson6, "amlogic,meson6-timer",
+		       meson6_timer_init);
diff --git a/drivers/clocksource/metag_generic.c b/drivers/clocksource/metag_generic.c
index 9e4db41abe3c..b7384b853e5a 100644
--- a/drivers/clocksource/metag_generic.c
+++ b/drivers/clocksource/metag_generic.c
@@ -90,7 +90,7 @@ static struct clocksource clocksource_metag = {
 
 static irqreturn_t metag_timer_interrupt(int irq, void *dummy)
 {
-	struct clock_event_device *evt = &__get_cpu_var(local_clockevent);
+	struct clock_event_device *evt = this_cpu_ptr(&local_clockevent);
 
 	evt->event_handler(evt);
 
diff --git a/drivers/clocksource/mips-gic-timer.c b/drivers/clocksource/mips-gic-timer.c
new file mode 100644
index 000000000000..3bd31b1321f6
--- /dev/null
+++ b/drivers/clocksource/mips-gic-timer.c
@@ -0,0 +1,166 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2012 MIPS Technologies, Inc.  All rights reserved.
+ */
+#include <linux/clockchips.h>
+#include <linux/cpu.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/irqchip/mips-gic.h>
+#include <linux/notifier.h>
+#include <linux/of_irq.h>
+#include <linux/percpu.h>
+#include <linux/smp.h>
+#include <linux/time.h>
+
+static DEFINE_PER_CPU(struct clock_event_device, gic_clockevent_device);
+static int gic_timer_irq;
+static unsigned int gic_frequency;
+
+static int gic_next_event(unsigned long delta, struct clock_event_device *evt)
+{
+	u64 cnt;
+	int res;
+
+	cnt = gic_read_count();
+	cnt += (u64)delta;
+	gic_write_cpu_compare(cnt, cpumask_first(evt->cpumask));
+	res = ((int)(gic_read_count() - cnt) >= 0) ? -ETIME : 0;
+	return res;
+}
+
+static void gic_set_clock_mode(enum clock_event_mode mode,
+				struct clock_event_device *evt)
+{
+	/* Nothing to do ...  */
+}
+
+static irqreturn_t gic_compare_interrupt(int irq, void *dev_id)
+{
+	struct clock_event_device *cd = dev_id;
+
+	gic_write_compare(gic_read_compare());
+	cd->event_handler(cd);
+	return IRQ_HANDLED;
+}
+
+struct irqaction gic_compare_irqaction = {
+	.handler = gic_compare_interrupt,
+	.percpu_dev_id = &gic_clockevent_device,
+	.flags = IRQF_PERCPU | IRQF_TIMER,
+	.name = "timer",
+};
+
+static void gic_clockevent_cpu_init(struct clock_event_device *cd)
+{
+	unsigned int cpu = smp_processor_id();
+
+	cd->name		= "MIPS GIC";
+	cd->features		= CLOCK_EVT_FEAT_ONESHOT |
+				  CLOCK_EVT_FEAT_C3STOP;
+
+	cd->rating		= 350;
+	cd->irq			= gic_timer_irq;
+	cd->cpumask		= cpumask_of(cpu);
+	cd->set_next_event	= gic_next_event;
+	cd->set_mode		= gic_set_clock_mode;
+
+	clockevents_config_and_register(cd, gic_frequency, 0x300, 0x7fffffff);
+
+	enable_percpu_irq(gic_timer_irq, IRQ_TYPE_NONE);
+}
+
+static void gic_clockevent_cpu_exit(struct clock_event_device *cd)
+{
+	disable_percpu_irq(gic_timer_irq);
+}
+
+static int gic_cpu_notifier(struct notifier_block *nb, unsigned long action,
+				void *data)
+{
+	switch (action & ~CPU_TASKS_FROZEN) {
+	case CPU_STARTING:
+		gic_clockevent_cpu_init(this_cpu_ptr(&gic_clockevent_device));
+		break;
+	case CPU_DYING:
+		gic_clockevent_cpu_exit(this_cpu_ptr(&gic_clockevent_device));
+		break;
+	}
+
+	return NOTIFY_OK;
+}
+
+static struct notifier_block gic_cpu_nb = {
+	.notifier_call = gic_cpu_notifier,
+};
+
+static int gic_clockevent_init(void)
+{
+	if (!cpu_has_counter || !gic_frequency)
+		return -ENXIO;
+
+	setup_percpu_irq(gic_timer_irq, &gic_compare_irqaction);
+
+	register_cpu_notifier(&gic_cpu_nb);
+
+	gic_clockevent_cpu_init(this_cpu_ptr(&gic_clockevent_device));
+
+	return 0;
+}
+
+static cycle_t gic_hpt_read(struct clocksource *cs)
+{
+	return gic_read_count();
+}
+
+static struct clocksource gic_clocksource = {
+	.name	= "GIC",
+	.read	= gic_hpt_read,
+	.flags	= CLOCK_SOURCE_IS_CONTINUOUS,
+};
+
+static void __init __gic_clocksource_init(void)
+{
+	/* Set clocksource mask. */
+	gic_clocksource.mask = CLOCKSOURCE_MASK(gic_get_count_width());
+
+	/* Calculate a somewhat reasonable rating value. */
+	gic_clocksource.rating = 200 + gic_frequency / 10000000;
+
+	clocksource_register_hz(&gic_clocksource, gic_frequency);
+
+	gic_clockevent_init();
+}
+
+void __init gic_clocksource_init(unsigned int frequency)
+{
+	gic_frequency = frequency;
+	gic_timer_irq = MIPS_GIC_IRQ_BASE +
+		GIC_LOCAL_TO_HWIRQ(GIC_LOCAL_INT_COMPARE);
+
+	__gic_clocksource_init();
+}
+
+static void __init gic_clocksource_of_init(struct device_node *node)
+{
+	if (WARN_ON(!gic_present || !node->parent ||
+		    !of_device_is_compatible(node->parent, "mti,gic")))
+		return;
+
+	if (of_property_read_u32(node, "clock-frequency", &gic_frequency)) {
+		pr_err("GIC frequency not specified.\n");
+		return;
+	}
+	gic_timer_irq = irq_of_parse_and_map(node, 0);
+	if (!gic_timer_irq) {
+		pr_err("GIC timer IRQ not specified.\n");
+		return;
+	}
+
+	__gic_clocksource_init();
+}
+CLOCKSOURCE_OF_DECLARE(mips_gic_timer, "mti,gic-timer",
+		       gic_clocksource_of_init);
diff --git a/drivers/clocksource/qcom-timer.c b/drivers/clocksource/qcom-timer.c
index 8d115db1e651..098c542e5c53 100644
--- a/drivers/clocksource/qcom-timer.c
+++ b/drivers/clocksource/qcom-timer.c
@@ -219,7 +219,7 @@ static void __init msm_timer_init(u32 dgt_hz, int sched_bits, int irq,
 		}
 
 		/* Immediately configure the timer on the boot CPU */
-		msm_local_timer_setup(__this_cpu_ptr(msm_evt));
+		msm_local_timer_setup(raw_cpu_ptr(msm_evt));
 	}
 
 err:
diff --git a/drivers/clocksource/sh_tmu.c b/drivers/clocksource/sh_tmu.c
index 0f665b8f2461..f150ca82bfaf 100644
--- a/drivers/clocksource/sh_tmu.c
+++ b/drivers/clocksource/sh_tmu.c
@@ -428,7 +428,7 @@ static void sh_tmu_register_clockevent(struct sh_tmu_channel *ch,
 	ced->features = CLOCK_EVT_FEAT_PERIODIC;
 	ced->features |= CLOCK_EVT_FEAT_ONESHOT;
 	ced->rating = 200;
-	ced->cpumask = cpumask_of(0);
+	ced->cpumask = cpu_possible_mask;
 	ced->set_next_event = sh_tmu_clock_event_next;
 	ced->set_mode = sh_tmu_clock_event_mode;
 	ced->suspend = sh_tmu_clock_event_suspend;
diff --git a/drivers/clocksource/sun4i_timer.c b/drivers/clocksource/sun4i_timer.c
index efb17c3ee120..f4a9c0058b4d 100644
--- a/drivers/clocksource/sun4i_timer.c
+++ b/drivers/clocksource/sun4i_timer.c
@@ -182,6 +182,12 @@ static void __init sun4i_timer_init(struct device_node *node)
 	/* Make sure timer is stopped before playing with interrupts */
 	sun4i_clkevt_time_stop(0);
 
+	sun4i_clockevent.cpumask = cpu_possible_mask;
+	sun4i_clockevent.irq = irq;
+
+	clockevents_config_and_register(&sun4i_clockevent, rate,
+					TIMER_SYNC_TICKS, 0xffffffff);
+
 	ret = setup_irq(irq, &sun4i_timer_irq);
 	if (ret)
 		pr_warn("failed to setup irq %d\n", irq);
@@ -189,12 +195,6 @@ static void __init sun4i_timer_init(struct device_node *node)
 	/* Enable timer0 interrupt */
 	val = readl(timer_base + TIMER_IRQ_EN_REG);
 	writel(val | TIMER_IRQ_EN(0), timer_base + TIMER_IRQ_EN_REG);
-
-	sun4i_clockevent.cpumask = cpu_possible_mask;
-	sun4i_clockevent.irq = irq;
-
-	clockevents_config_and_register(&sun4i_clockevent, rate,
-					TIMER_SYNC_TICKS, 0xffffffff);
 }
 CLOCKSOURCE_OF_DECLARE(sun4i, "allwinner,sun4i-a10-timer",
 		       sun4i_timer_init);
diff --git a/drivers/clocksource/tcb_clksrc.c b/drivers/clocksource/tcb_clksrc.c
index a8d7ea14f183..8bdbc45c6dad 100644
--- a/drivers/clocksource/tcb_clksrc.c
+++ b/drivers/clocksource/tcb_clksrc.c
@@ -178,12 +178,6 @@ static irqreturn_t ch2_irq(int irq, void *handle)
 	return IRQ_NONE;
 }
 
-static struct irqaction tc_irqaction = {
-	.name		= "tc_clkevt",
-	.flags		= IRQF_TIMER,
-	.handler	= ch2_irq,
-};
-
 static int __init setup_clkevents(struct atmel_tc *tc, int clk32k_divisor_idx)
 {
 	int ret;
@@ -198,15 +192,16 @@ static int __init setup_clkevents(struct atmel_tc *tc, int clk32k_divisor_idx)
 
 	clkevt.regs = tc->regs;
 	clkevt.clk = t2_clk;
-	tc_irqaction.dev_id = &clkevt;
 
 	timer_clock = clk32k_divisor_idx;
 
 	clkevt.clkevt.cpumask = cpumask_of(0);
 
-	ret = setup_irq(irq, &tc_irqaction);
-	if (ret)
+	ret = request_irq(irq, ch2_irq, IRQF_TIMER, "tc_clkevt", &clkevt);
+	if (ret) {
+		clk_disable_unprepare(t2_clk);
 		return ret;
+	}
 
 	clockevents_config_and_register(&clkevt.clkevt, 32768, 1, 0xffff);
 
@@ -279,7 +274,7 @@ static int __init tcb_clksrc_init(void)
 	int i;
 	int ret;
 
-	tc = atmel_tc_alloc(CONFIG_ATMEL_TCB_CLKSRC_BLOCK, clksrc.name);
+	tc = atmel_tc_alloc(CONFIG_ATMEL_TCB_CLKSRC_BLOCK);
 	if (!tc) {
 		pr_debug("can't alloc TC for clocksource\n");
 		return -ENODEV;
diff --git a/drivers/clocksource/time-armada-370-xp.c b/drivers/clocksource/time-armada-370-xp.c
index 0451e62fac7a..0c8c5e337540 100644
--- a/drivers/clocksource/time-armada-370-xp.c
+++ b/drivers/clocksource/time-armada-370-xp.c
@@ -43,6 +43,7 @@
 #include <linux/module.h>
 #include <linux/sched_clock.h>
 #include <linux/percpu.h>
+#include <linux/syscore_ops.h>
 
 /*
  * Timer block registers.
@@ -223,6 +224,28 @@ static struct notifier_block armada_370_xp_timer_cpu_nb = {
 	.notifier_call = armada_370_xp_timer_cpu_notify,
 };
 
+static u32 timer0_ctrl_reg, timer0_local_ctrl_reg;
+
+static int armada_370_xp_timer_suspend(void)
+{
+	timer0_ctrl_reg = readl(timer_base + TIMER_CTRL_OFF);
+	timer0_local_ctrl_reg = readl(local_base + TIMER_CTRL_OFF);
+	return 0;
+}
+
+static void armada_370_xp_timer_resume(void)
+{
+	writel(0xffffffff, timer_base + TIMER0_VAL_OFF);
+	writel(0xffffffff, timer_base + TIMER0_RELOAD_OFF);
+	writel(timer0_ctrl_reg, timer_base + TIMER_CTRL_OFF);
+	writel(timer0_local_ctrl_reg, local_base + TIMER_CTRL_OFF);
+}
+
+struct syscore_ops armada_370_xp_timer_syscore_ops = {
+	.suspend	= armada_370_xp_timer_suspend,
+	.resume		= armada_370_xp_timer_resume,
+};
+
 static void __init armada_370_xp_timer_common_init(struct device_node *np)
 {
 	u32 clr = 0, set = 0;
@@ -285,6 +308,8 @@ static void __init armada_370_xp_timer_common_init(struct device_node *np)
 	/* Immediately configure the timer on the boot CPU */
 	if (!res)
 		armada_370_xp_timer_setup(this_cpu_ptr(armada_370_xp_evt));
+
+	register_syscore_ops(&armada_370_xp_timer_syscore_ops);
 }
 
 static void __init armada_xp_timer_init(struct device_node *np)
@@ -293,6 +318,7 @@ static void __init armada_xp_timer_init(struct device_node *np)
 
 	/* The 25Mhz fixed clock is mandatory, and must always be available */
 	BUG_ON(IS_ERR(clk));
+	clk_prepare_enable(clk);
 	timer_clk = clk_get_rate(clk);
 
 	armada_370_xp_timer_common_init(np);
@@ -300,11 +326,40 @@ static void __init armada_xp_timer_init(struct device_node *np)
 CLOCKSOURCE_OF_DECLARE(armada_xp, "marvell,armada-xp-timer",
 		       armada_xp_timer_init);
 
+static void __init armada_375_timer_init(struct device_node *np)
+{
+	struct clk *clk;
+
+	clk = of_clk_get_by_name(np, "fixed");
+	if (!IS_ERR(clk)) {
+		clk_prepare_enable(clk);
+		timer_clk = clk_get_rate(clk);
+	} else {
+
+		/*
+		 * This fallback is required in order to retain proper
+		 * devicetree backwards compatibility.
+		 */
+		clk = of_clk_get(np, 0);
+
+		/* Must have at least a clock */
+		BUG_ON(IS_ERR(clk));
+		clk_prepare_enable(clk);
+		timer_clk = clk_get_rate(clk) / TIMER_DIVIDER;
+		timer25Mhz = false;
+	}
+
+	armada_370_xp_timer_common_init(np);
+}
+CLOCKSOURCE_OF_DECLARE(armada_375, "marvell,armada-375-timer",
+		       armada_375_timer_init);
+
 static void __init armada_370_timer_init(struct device_node *np)
 {
 	struct clk *clk = of_clk_get(np, 0);
 
 	BUG_ON(IS_ERR(clk));
+	clk_prepare_enable(clk);
 	timer_clk = clk_get_rate(clk) / TIMER_DIVIDER;
 	timer25Mhz = false;
 
diff --git a/drivers/clocksource/timer-atmel-pit.c b/drivers/clocksource/timer-atmel-pit.c
new file mode 100644
index 000000000000..b5b4d4585c9a
--- /dev/null
+++ b/drivers/clocksource/timer-atmel-pit.c
@@ -0,0 +1,264 @@
+/*
+ * at91sam926x_time.c - Periodic Interval Timer (PIT) for at91sam926x
+ *
+ * Copyright (C) 2005-2006 M. Amine SAYA, ATMEL Rousset, France
+ * Revision	 2005 M. Nicolas Diremdjian, ATMEL Rousset, France
+ * Converted to ClockSource/ClockEvents by David Brownell.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#define pr_fmt(fmt)	"AT91: PIT: " fmt
+
+#include <linux/clk.h>
+#include <linux/clockchips.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/kernel.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/slab.h>
+
+#define AT91_PIT_MR		0x00			/* Mode Register */
+#define AT91_PIT_PITIEN			BIT(25)			/* Timer Interrupt Enable */
+#define AT91_PIT_PITEN			BIT(24)			/* Timer Enabled */
+#define AT91_PIT_PIV			GENMASK(19, 0)		/* Periodic Interval Value */
+
+#define AT91_PIT_SR		0x04			/* Status Register */
+#define AT91_PIT_PITS			BIT(0)			/* Timer Status */
+
+#define AT91_PIT_PIVR		0x08			/* Periodic Interval Value Register */
+#define AT91_PIT_PIIR		0x0c			/* Periodic Interval Image Register */
+#define AT91_PIT_PICNT			GENMASK(31, 20)		/* Interval Counter */
+#define AT91_PIT_CPIV			GENMASK(19, 0)		/* Inverval Value */
+
+#define PIT_CPIV(x)	((x) & AT91_PIT_CPIV)
+#define PIT_PICNT(x)	(((x) & AT91_PIT_PICNT) >> 20)
+
+struct pit_data {
+	struct clock_event_device	clkevt;
+	struct clocksource		clksrc;
+
+	void __iomem	*base;
+	u32		cycle;
+	u32		cnt;
+	unsigned int	irq;
+	struct clk	*mck;
+};
+
+static inline struct pit_data *clksrc_to_pit_data(struct clocksource *clksrc)
+{
+	return container_of(clksrc, struct pit_data, clksrc);
+}
+
+static inline struct pit_data *clkevt_to_pit_data(struct clock_event_device *clkevt)
+{
+	return container_of(clkevt, struct pit_data, clkevt);
+}
+
+static inline unsigned int pit_read(void __iomem *base, unsigned int reg_offset)
+{
+	return __raw_readl(base + reg_offset);
+}
+
+static inline void pit_write(void __iomem *base, unsigned int reg_offset, unsigned long value)
+{
+	__raw_writel(value, base + reg_offset);
+}
+
+/*
+ * Clocksource:  just a monotonic counter of MCK/16 cycles.
+ * We don't care whether or not PIT irqs are enabled.
+ */
+static cycle_t read_pit_clk(struct clocksource *cs)
+{
+	struct pit_data *data = clksrc_to_pit_data(cs);
+	unsigned long flags;
+	u32 elapsed;
+	u32 t;
+
+	raw_local_irq_save(flags);
+	elapsed = data->cnt;
+	t = pit_read(data->base, AT91_PIT_PIIR);
+	raw_local_irq_restore(flags);
+
+	elapsed += PIT_PICNT(t) * data->cycle;
+	elapsed += PIT_CPIV(t);
+	return elapsed;
+}
+
+/*
+ * Clockevent device:  interrupts every 1/HZ (== pit_cycles * MCK/16)
+ */
+static void
+pit_clkevt_mode(enum clock_event_mode mode, struct clock_event_device *dev)
+{
+	struct pit_data *data = clkevt_to_pit_data(dev);
+
+	switch (mode) {
+	case CLOCK_EVT_MODE_PERIODIC:
+		/* update clocksource counter */
+		data->cnt += data->cycle * PIT_PICNT(pit_read(data->base, AT91_PIT_PIVR));
+		pit_write(data->base, AT91_PIT_MR,
+			  (data->cycle - 1) | AT91_PIT_PITEN | AT91_PIT_PITIEN);
+		break;
+	case CLOCK_EVT_MODE_ONESHOT:
+		BUG();
+		/* FALLTHROUGH */
+	case CLOCK_EVT_MODE_SHUTDOWN:
+	case CLOCK_EVT_MODE_UNUSED:
+		/* disable irq, leaving the clocksource active */
+		pit_write(data->base, AT91_PIT_MR,
+			  (data->cycle - 1) | AT91_PIT_PITEN);
+		break;
+	case CLOCK_EVT_MODE_RESUME:
+		break;
+	}
+}
+
+static void at91sam926x_pit_suspend(struct clock_event_device *cedev)
+{
+	struct pit_data *data = clkevt_to_pit_data(cedev);
+
+	/* Disable timer */
+	pit_write(data->base, AT91_PIT_MR, 0);
+}
+
+static void at91sam926x_pit_reset(struct pit_data *data)
+{
+	/* Disable timer and irqs */
+	pit_write(data->base, AT91_PIT_MR, 0);
+
+	/* Clear any pending interrupts, wait for PIT to stop counting */
+	while (PIT_CPIV(pit_read(data->base, AT91_PIT_PIVR)) != 0)
+		cpu_relax();
+
+	/* Start PIT but don't enable IRQ */
+	pit_write(data->base, AT91_PIT_MR,
+		  (data->cycle - 1) | AT91_PIT_PITEN);
+}
+
+static void at91sam926x_pit_resume(struct clock_event_device *cedev)
+{
+	struct pit_data *data = clkevt_to_pit_data(cedev);
+
+	at91sam926x_pit_reset(data);
+}
+
+/*
+ * IRQ handler for the timer.
+ */
+static irqreturn_t at91sam926x_pit_interrupt(int irq, void *dev_id)
+{
+	struct pit_data *data = dev_id;
+
+	/*
+	 * irqs should be disabled here, but as the irq is shared they are only
+	 * guaranteed to be off if the timer irq is registered first.
+	 */
+	WARN_ON_ONCE(!irqs_disabled());
+
+	/* The PIT interrupt may be disabled, and is shared */
+	if ((data->clkevt.mode == CLOCK_EVT_MODE_PERIODIC) &&
+	    (pit_read(data->base, AT91_PIT_SR) & AT91_PIT_PITS)) {
+		unsigned nr_ticks;
+
+		/* Get number of ticks performed before irq, and ack it */
+		nr_ticks = PIT_PICNT(pit_read(data->base, AT91_PIT_PIVR));
+		do {
+			data->cnt += data->cycle;
+			data->clkevt.event_handler(&data->clkevt);
+			nr_ticks--;
+		} while (nr_ticks);
+
+		return IRQ_HANDLED;
+	}
+
+	return IRQ_NONE;
+}
+
+/*
+ * Set up both clocksource and clockevent support.
+ */
+static void __init at91sam926x_pit_common_init(struct pit_data *data)
+{
+	unsigned long	pit_rate;
+	unsigned	bits;
+	int		ret;
+
+	/*
+	 * Use our actual MCK to figure out how many MCK/16 ticks per
+	 * 1/HZ period (instead of a compile-time constant LATCH).
+	 */
+	pit_rate = clk_get_rate(data->mck) / 16;
+	data->cycle = DIV_ROUND_CLOSEST(pit_rate, HZ);
+	WARN_ON(((data->cycle - 1) & ~AT91_PIT_PIV) != 0);
+
+	/* Initialize and enable the timer */
+	at91sam926x_pit_reset(data);
+
+	/*
+	 * Register clocksource.  The high order bits of PIV are unused,
+	 * so this isn't a 32-bit counter unless we get clockevent irqs.
+	 */
+	bits = 12 /* PICNT */ + ilog2(data->cycle) /* PIV */;
+	data->clksrc.mask = CLOCKSOURCE_MASK(bits);
+	data->clksrc.name = "pit";
+	data->clksrc.rating = 175;
+	data->clksrc.read = read_pit_clk,
+	data->clksrc.flags = CLOCK_SOURCE_IS_CONTINUOUS,
+	clocksource_register_hz(&data->clksrc, pit_rate);
+
+	/* Set up irq handler */
+	ret = request_irq(data->irq, at91sam926x_pit_interrupt,
+			  IRQF_SHARED | IRQF_TIMER | IRQF_IRQPOLL,
+			  "at91_tick", data);
+	if (ret)
+		panic(pr_fmt("Unable to setup IRQ\n"));
+
+	/* Set up and register clockevents */
+	data->clkevt.name = "pit";
+	data->clkevt.features = CLOCK_EVT_FEAT_PERIODIC;
+	data->clkevt.shift = 32;
+	data->clkevt.mult = div_sc(pit_rate, NSEC_PER_SEC, data->clkevt.shift);
+	data->clkevt.rating = 100;
+	data->clkevt.cpumask = cpumask_of(0);
+
+	data->clkevt.set_mode = pit_clkevt_mode;
+	data->clkevt.resume = at91sam926x_pit_resume;
+	data->clkevt.suspend = at91sam926x_pit_suspend;
+	clockevents_register_device(&data->clkevt);
+}
+
+static void __init at91sam926x_pit_dt_init(struct device_node *node)
+{
+	struct pit_data *data;
+
+	data = kzalloc(sizeof(*data), GFP_KERNEL);
+	if (!data)
+		panic(pr_fmt("Unable to allocate memory\n"));
+
+	data->base = of_iomap(node, 0);
+	if (!data->base)
+		panic(pr_fmt("Could not map PIT address\n"));
+
+	data->mck = of_clk_get(node, 0);
+	if (IS_ERR(data->mck))
+		/* Fallback on clkdev for !CCF-based boards */
+		data->mck = clk_get(NULL, "mck");
+
+	if (IS_ERR(data->mck))
+		panic(pr_fmt("Unable to get mck clk\n"));
+
+	/* Get the interrupts property */
+	data->irq = irq_of_parse_and_map(node, 0);
+	if (!data->irq)
+		panic(pr_fmt("Unable to get IRQ from DT\n"));
+
+	at91sam926x_pit_common_init(data);
+}
+CLOCKSOURCE_OF_DECLARE(at91sam926x_pit, "atmel,at91sam9260-pit",
+		       at91sam926x_pit_dt_init);
diff --git a/drivers/clocksource/timer-integrator-ap.c b/drivers/clocksource/timer-integrator-ap.c
new file mode 100644
index 000000000000..b9efd30513d5
--- /dev/null
+++ b/drivers/clocksource/timer-integrator-ap.c
@@ -0,0 +1,210 @@
+/*
+ * Integrator/AP timer driver
+ * Copyright (C) 2000-2003 Deep Blue Solutions Ltd
+ * Copyright (c) 2014, Linaro Limited
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/clk.h>
+#include <linux/clocksource.h>
+#include <linux/of_irq.h>
+#include <linux/of_address.h>
+#include <linux/of_platform.h>
+#include <linux/clockchips.h>
+#include <linux/interrupt.h>
+#include <linux/sched_clock.h>
+#include <asm/hardware/arm_timer.h>
+
+static void __iomem * sched_clk_base;
+
+static u64 notrace integrator_read_sched_clock(void)
+{
+	return -readl(sched_clk_base + TIMER_VALUE);
+}
+
+static void integrator_clocksource_init(unsigned long inrate,
+					void __iomem *base)
+{
+	u32 ctrl = TIMER_CTRL_ENABLE | TIMER_CTRL_PERIODIC;
+	unsigned long rate = inrate;
+
+	if (rate >= 1500000) {
+		rate /= 16;
+		ctrl |= TIMER_CTRL_DIV16;
+	}
+
+	writel(0xffff, base + TIMER_LOAD);
+	writel(ctrl, base + TIMER_CTRL);
+
+	clocksource_mmio_init(base + TIMER_VALUE, "timer2",
+			rate, 200, 16, clocksource_mmio_readl_down);
+
+	sched_clk_base = base;
+	sched_clock_register(integrator_read_sched_clock, 16, rate);
+}
+
+static unsigned long timer_reload;
+static void __iomem * clkevt_base;
+
+/*
+ * IRQ handler for the timer
+ */
+static irqreturn_t integrator_timer_interrupt(int irq, void *dev_id)
+{
+	struct clock_event_device *evt = dev_id;
+
+	/* clear the interrupt */
+	writel(1, clkevt_base + TIMER_INTCLR);
+
+	evt->event_handler(evt);
+
+	return IRQ_HANDLED;
+}
+
+static void clkevt_set_mode(enum clock_event_mode mode, struct clock_event_device *evt)
+{
+	u32 ctrl = readl(clkevt_base + TIMER_CTRL) & ~TIMER_CTRL_ENABLE;
+
+	/* Disable timer */
+	writel(ctrl, clkevt_base + TIMER_CTRL);
+
+	switch (mode) {
+	case CLOCK_EVT_MODE_PERIODIC:
+		/* Enable the timer and start the periodic tick */
+		writel(timer_reload, clkevt_base + TIMER_LOAD);
+		ctrl |= TIMER_CTRL_PERIODIC | TIMER_CTRL_ENABLE;
+		writel(ctrl, clkevt_base + TIMER_CTRL);
+		break;
+	case CLOCK_EVT_MODE_ONESHOT:
+		/* Leave the timer disabled, .set_next_event will enable it */
+		ctrl &= ~TIMER_CTRL_PERIODIC;
+		writel(ctrl, clkevt_base + TIMER_CTRL);
+		break;
+	case CLOCK_EVT_MODE_UNUSED:
+	case CLOCK_EVT_MODE_SHUTDOWN:
+	case CLOCK_EVT_MODE_RESUME:
+	default:
+		/* Just leave in disabled state */
+		break;
+	}
+
+}
+
+static int clkevt_set_next_event(unsigned long next, struct clock_event_device *evt)
+{
+	unsigned long ctrl = readl(clkevt_base + TIMER_CTRL);
+
+	writel(ctrl & ~TIMER_CTRL_ENABLE, clkevt_base + TIMER_CTRL);
+	writel(next, clkevt_base + TIMER_LOAD);
+	writel(ctrl | TIMER_CTRL_ENABLE, clkevt_base + TIMER_CTRL);
+
+	return 0;
+}
+
+static struct clock_event_device integrator_clockevent = {
+	.name		= "timer1",
+	.features	= CLOCK_EVT_FEAT_PERIODIC | CLOCK_EVT_FEAT_ONESHOT,
+	.set_mode	= clkevt_set_mode,
+	.set_next_event	= clkevt_set_next_event,
+	.rating		= 300,
+};
+
+static struct irqaction integrator_timer_irq = {
+	.name		= "timer",
+	.flags		= IRQF_TIMER | IRQF_IRQPOLL,
+	.handler	= integrator_timer_interrupt,
+	.dev_id		= &integrator_clockevent,
+};
+
+static void integrator_clockevent_init(unsigned long inrate,
+				void __iomem *base, int irq)
+{
+	unsigned long rate = inrate;
+	unsigned int ctrl = 0;
+
+	clkevt_base = base;
+	/* Calculate and program a divisor */
+	if (rate > 0x100000 * HZ) {
+		rate /= 256;
+		ctrl |= TIMER_CTRL_DIV256;
+	} else if (rate > 0x10000 * HZ) {
+		rate /= 16;
+		ctrl |= TIMER_CTRL_DIV16;
+	}
+	timer_reload = rate / HZ;
+	writel(ctrl, clkevt_base + TIMER_CTRL);
+
+	setup_irq(irq, &integrator_timer_irq);
+	clockevents_config_and_register(&integrator_clockevent,
+					rate,
+					1,
+					0xffffU);
+}
+
+static void __init integrator_ap_timer_init_of(struct device_node *node)
+{
+	const char *path;
+	void __iomem *base;
+	int err;
+	int irq;
+	struct clk *clk;
+	unsigned long rate;
+	struct device_node *pri_node;
+	struct device_node *sec_node;
+
+	base = of_io_request_and_map(node, 0, "integrator-timer");
+	if (!base)
+		return;
+
+	clk = of_clk_get(node, 0);
+	if (IS_ERR(clk)) {
+		pr_err("No clock for %s\n", node->name);
+		return;
+	}
+	clk_prepare_enable(clk);
+	rate = clk_get_rate(clk);
+	writel(0, base + TIMER_CTRL);
+
+	err = of_property_read_string(of_aliases,
+				"arm,timer-primary", &path);
+	if (WARN_ON(err))
+		return;
+	pri_node = of_find_node_by_path(path);
+	err = of_property_read_string(of_aliases,
+				"arm,timer-secondary", &path);
+	if (WARN_ON(err))
+		return;
+	sec_node = of_find_node_by_path(path);
+
+	if (node == pri_node) {
+		/* The primary timer lacks IRQ, use as clocksource */
+		integrator_clocksource_init(rate, base);
+		return;
+	}
+
+	if (node == sec_node) {
+		/* The secondary timer will drive the clock event */
+		irq = irq_of_parse_and_map(node, 0);
+		integrator_clockevent_init(rate, base, irq);
+		return;
+	}
+
+	pr_info("Timer @%p unused\n", base);
+	clk_disable_unprepare(clk);
+}
+
+CLOCKSOURCE_OF_DECLARE(integrator_ap_timer, "arm,integrator-timer",
+		       integrator_ap_timer_init_of);
diff --git a/drivers/clocksource/timer-marco.c b/drivers/clocksource/timer-marco.c
index 330e93064692..361a789d4bee 100644
--- a/drivers/clocksource/timer-marco.c
+++ b/drivers/clocksource/timer-marco.c
@@ -20,8 +20,6 @@
 #include <linux/of_address.h>
 #include <linux/sched_clock.h>
 
-#define MARCO_CLOCK_FREQ 1000000
-
 #define SIRFSOC_TIMER_32COUNTER_0_CTRL			0x0000
 #define SIRFSOC_TIMER_32COUNTER_1_CTRL			0x0004
 #define SIRFSOC_TIMER_MATCH_0				0x0018
@@ -40,6 +38,8 @@
 
 #define SIRFSOC_TIMER_REG_CNT 6
 
+static unsigned long marco_timer_rate;
+
 static const u32 sirfsoc_timer_reg_list[SIRFSOC_TIMER_REG_CNT] = {
 	SIRFSOC_TIMER_WATCHDOG_EN,
 	SIRFSOC_TIMER_32COUNTER_0_CTRL,
@@ -63,7 +63,7 @@ static inline void sirfsoc_timer_count_disable(int idx)
 /* enable count and interrupt */
 static inline void sirfsoc_timer_count_enable(int idx)
 {
-	writel_relaxed(readl_relaxed(sirfsoc_timer_base + SIRFSOC_TIMER_32COUNTER_0_CTRL + 4 * idx) | 0x7,
+	writel_relaxed(readl_relaxed(sirfsoc_timer_base + SIRFSOC_TIMER_32COUNTER_0_CTRL + 4 * idx) | 0x3,
 		sirfsoc_timer_base + SIRFSOC_TIMER_32COUNTER_0_CTRL + 4 * idx);
 }
 
@@ -103,6 +103,9 @@ static int sirfsoc_timer_set_next_event(unsigned long delta,
 {
 	int cpu = smp_processor_id();
 
+	/* disable timer first, then modify the related registers */
+	sirfsoc_timer_count_disable(cpu);
+
 	writel_relaxed(0, sirfsoc_timer_base + SIRFSOC_TIMER_COUNTER_0 +
 		4 * cpu);
 	writel_relaxed(delta, sirfsoc_timer_base + SIRFSOC_TIMER_MATCH_0 +
@@ -192,7 +195,7 @@ static int sirfsoc_local_timer_setup(struct clock_event_device *ce)
 	ce->rating = 200;
 	ce->set_mode = sirfsoc_timer_set_mode;
 	ce->set_next_event = sirfsoc_timer_set_next_event;
-	clockevents_calc_mult_shift(ce, MARCO_CLOCK_FREQ, 60);
+	clockevents_calc_mult_shift(ce, marco_timer_rate, 60);
 	ce->max_delta_ns = clockevent_delta2ns(-2, ce);
 	ce->min_delta_ns = clockevent_delta2ns(2, ce);
 	ce->cpumask = cpumask_of(cpu);
@@ -254,7 +257,6 @@ static void __init sirfsoc_clockevent_init(void)
 /* initialize the kernel jiffy timer source */
 static void __init sirfsoc_marco_timer_init(struct device_node *np)
 {
-	unsigned long rate;
 	u32 timer_div;
 	struct clk *clk;
 
@@ -263,16 +265,12 @@ static void __init sirfsoc_marco_timer_init(struct device_node *np)
 
 	BUG_ON(clk_prepare_enable(clk));
 
-	rate = clk_get_rate(clk);
-
-	BUG_ON(rate < MARCO_CLOCK_FREQ);
-	BUG_ON(rate % MARCO_CLOCK_FREQ);
+	marco_timer_rate = clk_get_rate(clk);
 
-	/* Initialize the timer dividers */
-	timer_div = rate / MARCO_CLOCK_FREQ - 1;
-	writel_relaxed(timer_div << 16, sirfsoc_timer_base + SIRFSOC_TIMER_64COUNTER_CTRL);
-	writel_relaxed(timer_div << 16, sirfsoc_timer_base + SIRFSOC_TIMER_32COUNTER_0_CTRL);
-	writel_relaxed(timer_div << 16, sirfsoc_timer_base + SIRFSOC_TIMER_32COUNTER_1_CTRL);
+	/* timer dividers: 0, not divided */
+	writel_relaxed(0, sirfsoc_timer_base + SIRFSOC_TIMER_64COUNTER_CTRL);
+	writel_relaxed(0, sirfsoc_timer_base + SIRFSOC_TIMER_32COUNTER_0_CTRL);
+	writel_relaxed(0, sirfsoc_timer_base + SIRFSOC_TIMER_32COUNTER_1_CTRL);
 
 	/* Initialize timer counters to 0 */
 	writel_relaxed(0, sirfsoc_timer_base + SIRFSOC_TIMER_64COUNTER_LOAD_LO);
@@ -285,7 +283,7 @@ static void __init sirfsoc_marco_timer_init(struct device_node *np)
 	/* Clear all interrupts */
 	writel_relaxed(0xFFFF, sirfsoc_timer_base + SIRFSOC_TIMER_INTR_STATUS);
 
-	BUG_ON(clocksource_register_hz(&sirfsoc_clocksource, MARCO_CLOCK_FREQ));
+	BUG_ON(clocksource_register_hz(&sirfsoc_clocksource, marco_timer_rate));
 
 	sirfsoc_clockevent_init();
 }
diff --git a/drivers/clocksource/vf_pit_timer.c b/drivers/clocksource/vf_pit_timer.c
index a918bc481c52..b45ac6229b57 100644
--- a/drivers/clocksource/vf_pit_timer.c
+++ b/drivers/clocksource/vf_pit_timer.c
@@ -93,6 +93,10 @@ static void pit_set_mode(enum clock_event_mode mode,
 	case CLOCK_EVT_MODE_PERIODIC:
 		pit_set_next_event(cycle_per_jiffy, evt);
 		break;
+	case CLOCK_EVT_MODE_SHUTDOWN:
+	case CLOCK_EVT_MODE_UNUSED:
+		pit_timer_disable();
+		break;
 	default:
 		break;
 	}
diff --git a/drivers/connector/connector.c b/drivers/connector/connector.c
index f612d68629dc..30f522848c73 100644
--- a/drivers/connector/connector.c
+++ b/drivers/connector/connector.c
@@ -141,12 +141,18 @@ EXPORT_SYMBOL_GPL(cn_netlink_send);
  */
 static int cn_call_callback(struct sk_buff *skb)
 {
+	struct nlmsghdr *nlh;
 	struct cn_callback_entry *i, *cbq = NULL;
 	struct cn_dev *dev = &cdev;
 	struct cn_msg *msg = nlmsg_data(nlmsg_hdr(skb));
 	struct netlink_skb_parms *nsp = &NETLINK_CB(skb);
 	int err = -ENODEV;
 
+	/* verify msg->len is within skb */
+	nlh = nlmsg_hdr(skb);
+	if (nlh->nlmsg_len < NLMSG_HDRLEN + sizeof(struct cn_msg) + msg->len)
+		return -EINVAL;
+
 	spin_lock_bh(&dev->cbdev->queue_lock);
 	list_for_each_entry(i, &dev->cbdev->queue_list, callback_entry) {
 		if (cn_cb_equal(&i->id.id, &msg->id)) {
diff --git a/drivers/coresight/Makefile b/drivers/coresight/Makefile
new file mode 100644
index 000000000000..4b4bec890ef5
--- /dev/null
+++ b/drivers/coresight/Makefile
@@ -0,0 +1,11 @@
+#
+# Makefile for CoreSight drivers.
+#
+obj-$(CONFIG_CORESIGHT) += coresight.o
+obj-$(CONFIG_OF) += of_coresight.o
+obj-$(CONFIG_CORESIGHT_LINK_AND_SINK_TMC) += coresight-tmc.o
+obj-$(CONFIG_CORESIGHT_SINK_TPIU) += coresight-tpiu.o
+obj-$(CONFIG_CORESIGHT_SINK_ETBV10) += coresight-etb10.o
+obj-$(CONFIG_CORESIGHT_LINKS_AND_SINKS) += coresight-funnel.o \
+					   coresight-replicator.o
+obj-$(CONFIG_CORESIGHT_SOURCE_ETM3X) += coresight-etm3x.o coresight-etm-cp14.o
diff --git a/drivers/coresight/coresight-etb10.c b/drivers/coresight/coresight-etb10.c
new file mode 100644
index 000000000000..c922d4aded8a
--- /dev/null
+++ b/drivers/coresight/coresight-etb10.c
@@ -0,0 +1,537 @@
+/* Copyright (c) 2011-2012, The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/device.h>
+#include <linux/io.h>
+#include <linux/err.h>
+#include <linux/fs.h>
+#include <linux/miscdevice.h>
+#include <linux/uaccess.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+#include <linux/clk.h>
+#include <linux/seq_file.h>
+#include <linux/coresight.h>
+#include <linux/amba/bus.h>
+
+#include "coresight-priv.h"
+
+#define ETB_RAM_DEPTH_REG	0x004
+#define ETB_STATUS_REG		0x00c
+#define ETB_RAM_READ_DATA_REG	0x010
+#define ETB_RAM_READ_POINTER	0x014
+#define ETB_RAM_WRITE_POINTER	0x018
+#define ETB_TRG			0x01c
+#define ETB_CTL_REG		0x020
+#define ETB_RWD_REG		0x024
+#define ETB_FFSR		0x300
+#define ETB_FFCR		0x304
+#define ETB_ITMISCOP0		0xee0
+#define ETB_ITTRFLINACK		0xee4
+#define ETB_ITTRFLIN		0xee8
+#define ETB_ITATBDATA0		0xeeC
+#define ETB_ITATBCTR2		0xef0
+#define ETB_ITATBCTR1		0xef4
+#define ETB_ITATBCTR0		0xef8
+
+/* register description */
+/* STS - 0x00C */
+#define ETB_STATUS_RAM_FULL	BIT(0)
+/* CTL - 0x020 */
+#define ETB_CTL_CAPT_EN		BIT(0)
+/* FFCR - 0x304 */
+#define ETB_FFCR_EN_FTC		BIT(0)
+#define ETB_FFCR_FON_MAN	BIT(6)
+#define ETB_FFCR_STOP_FI	BIT(12)
+#define ETB_FFCR_STOP_TRIGGER	BIT(13)
+
+#define ETB_FFCR_BIT		6
+#define ETB_FFSR_BIT		1
+#define ETB_FRAME_SIZE_WORDS	4
+
+/**
+ * struct etb_drvdata - specifics associated to an ETB component
+ * @base:	memory mapped base address for this component.
+ * @dev:	the device entity associated to this component.
+ * @csdev:	component vitals needed by the framework.
+ * @miscdev:	specifics to handle "/dev/xyz.etb" entry.
+ * @clk:	the clock this component is associated to.
+ * @spinlock:	only one at a time pls.
+ * @in_use:	synchronise user space access to etb buffer.
+ * @buf:	area of memory where ETB buffer content gets sent.
+ * @buffer_depth: size of @buf.
+ * @enable:	this ETB is being used.
+ * @trigger_cntr: amount of words to store after a trigger.
+ */
+struct etb_drvdata {
+	void __iomem		*base;
+	struct device		*dev;
+	struct coresight_device	*csdev;
+	struct miscdevice	miscdev;
+	struct clk		*clk;
+	spinlock_t		spinlock;
+	atomic_t		in_use;
+	u8			*buf;
+	u32			buffer_depth;
+	bool			enable;
+	u32			trigger_cntr;
+};
+
+static unsigned int etb_get_buffer_depth(struct etb_drvdata *drvdata)
+{
+	int ret;
+	u32 depth = 0;
+
+	ret = clk_prepare_enable(drvdata->clk);
+	if (ret)
+		return ret;
+
+	/* RO registers don't need locking */
+	depth = readl_relaxed(drvdata->base + ETB_RAM_DEPTH_REG);
+
+	clk_disable_unprepare(drvdata->clk);
+	return depth;
+}
+
+static void etb_enable_hw(struct etb_drvdata *drvdata)
+{
+	int i;
+	u32 depth;
+
+	CS_UNLOCK(drvdata->base);
+
+	depth = drvdata->buffer_depth;
+	/* reset write RAM pointer address */
+	writel_relaxed(0x0, drvdata->base + ETB_RAM_WRITE_POINTER);
+	/* clear entire RAM buffer */
+	for (i = 0; i < depth; i++)
+		writel_relaxed(0x0, drvdata->base + ETB_RWD_REG);
+
+	/* reset write RAM pointer address */
+	writel_relaxed(0x0, drvdata->base + ETB_RAM_WRITE_POINTER);
+	/* reset read RAM pointer address */
+	writel_relaxed(0x0, drvdata->base + ETB_RAM_READ_POINTER);
+
+	writel_relaxed(drvdata->trigger_cntr, drvdata->base + ETB_TRG);
+	writel_relaxed(ETB_FFCR_EN_FTC | ETB_FFCR_STOP_TRIGGER,
+		       drvdata->base + ETB_FFCR);
+	/* ETB trace capture enable */
+	writel_relaxed(ETB_CTL_CAPT_EN, drvdata->base + ETB_CTL_REG);
+
+	CS_LOCK(drvdata->base);
+}
+
+static int etb_enable(struct coresight_device *csdev)
+{
+	struct etb_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+	int ret;
+	unsigned long flags;
+
+	ret = clk_prepare_enable(drvdata->clk);
+	if (ret)
+		return ret;
+
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+	etb_enable_hw(drvdata);
+	drvdata->enable = true;
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+
+	dev_info(drvdata->dev, "ETB enabled\n");
+	return 0;
+}
+
+static void etb_disable_hw(struct etb_drvdata *drvdata)
+{
+	u32 ffcr;
+
+	CS_UNLOCK(drvdata->base);
+
+	ffcr = readl_relaxed(drvdata->base + ETB_FFCR);
+	/* stop formatter when a stop has completed */
+	ffcr |= ETB_FFCR_STOP_FI;
+	writel_relaxed(ffcr, drvdata->base + ETB_FFCR);
+	/* manually generate a flush of the system */
+	ffcr |= ETB_FFCR_FON_MAN;
+	writel_relaxed(ffcr, drvdata->base + ETB_FFCR);
+
+	if (coresight_timeout(drvdata->base, ETB_FFCR, ETB_FFCR_BIT, 0)) {
+		dev_err(drvdata->dev,
+			"timeout observed when probing at offset %#x\n",
+			ETB_FFCR);
+	}
+
+	/* disable trace capture */
+	writel_relaxed(0x0, drvdata->base + ETB_CTL_REG);
+
+	if (coresight_timeout(drvdata->base, ETB_FFSR, ETB_FFSR_BIT, 1)) {
+		dev_err(drvdata->dev,
+			"timeout observed when probing at offset %#x\n",
+			ETB_FFCR);
+	}
+
+	CS_LOCK(drvdata->base);
+}
+
+static void etb_dump_hw(struct etb_drvdata *drvdata)
+{
+	int i;
+	u8 *buf_ptr;
+	u32 read_data, depth;
+	u32 read_ptr, write_ptr;
+	u32 frame_off, frame_endoff;
+
+	CS_UNLOCK(drvdata->base);
+
+	read_ptr = readl_relaxed(drvdata->base + ETB_RAM_READ_POINTER);
+	write_ptr = readl_relaxed(drvdata->base + ETB_RAM_WRITE_POINTER);
+
+	frame_off = write_ptr % ETB_FRAME_SIZE_WORDS;
+	frame_endoff = ETB_FRAME_SIZE_WORDS - frame_off;
+	if (frame_off) {
+		dev_err(drvdata->dev,
+			"write_ptr: %lu not aligned to formatter frame size\n",
+			(unsigned long)write_ptr);
+		dev_err(drvdata->dev, "frameoff: %lu, frame_endoff: %lu\n",
+			(unsigned long)frame_off, (unsigned long)frame_endoff);
+		write_ptr += frame_endoff;
+	}
+
+	if ((readl_relaxed(drvdata->base + ETB_STATUS_REG)
+		      & ETB_STATUS_RAM_FULL) == 0)
+		writel_relaxed(0x0, drvdata->base + ETB_RAM_READ_POINTER);
+	else
+		writel_relaxed(write_ptr, drvdata->base + ETB_RAM_READ_POINTER);
+
+	depth = drvdata->buffer_depth;
+	buf_ptr = drvdata->buf;
+	for (i = 0; i < depth; i++) {
+		read_data = readl_relaxed(drvdata->base +
+					  ETB_RAM_READ_DATA_REG);
+		*buf_ptr++ = read_data >> 0;
+		*buf_ptr++ = read_data >> 8;
+		*buf_ptr++ = read_data >> 16;
+		*buf_ptr++ = read_data >> 24;
+	}
+
+	if (frame_off) {
+		buf_ptr -= (frame_endoff * 4);
+		for (i = 0; i < frame_endoff; i++) {
+			*buf_ptr++ = 0x0;
+			*buf_ptr++ = 0x0;
+			*buf_ptr++ = 0x0;
+			*buf_ptr++ = 0x0;
+		}
+	}
+
+	writel_relaxed(read_ptr, drvdata->base + ETB_RAM_READ_POINTER);
+
+	CS_LOCK(drvdata->base);
+}
+
+static void etb_disable(struct coresight_device *csdev)
+{
+	struct etb_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+	unsigned long flags;
+
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+	etb_disable_hw(drvdata);
+	etb_dump_hw(drvdata);
+	drvdata->enable = false;
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+
+	clk_disable_unprepare(drvdata->clk);
+
+	dev_info(drvdata->dev, "ETB disabled\n");
+}
+
+static const struct coresight_ops_sink etb_sink_ops = {
+	.enable		= etb_enable,
+	.disable	= etb_disable,
+};
+
+static const struct coresight_ops etb_cs_ops = {
+	.sink_ops	= &etb_sink_ops,
+};
+
+static void etb_dump(struct etb_drvdata *drvdata)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+	if (drvdata->enable) {
+		etb_disable_hw(drvdata);
+		etb_dump_hw(drvdata);
+		etb_enable_hw(drvdata);
+	}
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+
+	dev_info(drvdata->dev, "ETB dumped\n");
+}
+
+static int etb_open(struct inode *inode, struct file *file)
+{
+	struct etb_drvdata *drvdata = container_of(file->private_data,
+						   struct etb_drvdata, miscdev);
+
+	if (atomic_cmpxchg(&drvdata->in_use, 0, 1))
+		return -EBUSY;
+
+	dev_dbg(drvdata->dev, "%s: successfully opened\n", __func__);
+	return 0;
+}
+
+static ssize_t etb_read(struct file *file, char __user *data,
+				size_t len, loff_t *ppos)
+{
+	u32 depth;
+	struct etb_drvdata *drvdata = container_of(file->private_data,
+						   struct etb_drvdata, miscdev);
+
+	etb_dump(drvdata);
+
+	depth = drvdata->buffer_depth;
+	if (*ppos + len > depth * 4)
+		len = depth * 4 - *ppos;
+
+	if (copy_to_user(data, drvdata->buf + *ppos, len)) {
+		dev_dbg(drvdata->dev, "%s: copy_to_user failed\n", __func__);
+		return -EFAULT;
+	}
+
+	*ppos += len;
+
+	dev_dbg(drvdata->dev, "%s: %d bytes copied, %d bytes left\n",
+		__func__, len, (int) (depth * 4 - *ppos));
+	return len;
+}
+
+static int etb_release(struct inode *inode, struct file *file)
+{
+	struct etb_drvdata *drvdata = container_of(file->private_data,
+						   struct etb_drvdata, miscdev);
+	atomic_set(&drvdata->in_use, 0);
+
+	dev_dbg(drvdata->dev, "%s: released\n", __func__);
+	return 0;
+}
+
+static const struct file_operations etb_fops = {
+	.owner		= THIS_MODULE,
+	.open		= etb_open,
+	.read		= etb_read,
+	.release	= etb_release,
+	.llseek		= no_llseek,
+};
+
+static ssize_t status_show(struct device *dev,
+			   struct device_attribute *attr, char *buf)
+{
+	int ret;
+	unsigned long flags;
+	u32 etb_rdr, etb_sr, etb_rrp, etb_rwp;
+	u32 etb_trg, etb_cr, etb_ffsr, etb_ffcr;
+	struct etb_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = clk_prepare_enable(drvdata->clk);
+	if (ret)
+		goto out;
+
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+	CS_UNLOCK(drvdata->base);
+
+	etb_rdr = readl_relaxed(drvdata->base + ETB_RAM_DEPTH_REG);
+	etb_sr = readl_relaxed(drvdata->base + ETB_STATUS_REG);
+	etb_rrp = readl_relaxed(drvdata->base + ETB_RAM_READ_POINTER);
+	etb_rwp = readl_relaxed(drvdata->base + ETB_RAM_WRITE_POINTER);
+	etb_trg = readl_relaxed(drvdata->base + ETB_TRG);
+	etb_cr = readl_relaxed(drvdata->base + ETB_CTL_REG);
+	etb_ffsr = readl_relaxed(drvdata->base + ETB_FFSR);
+	etb_ffcr = readl_relaxed(drvdata->base + ETB_FFCR);
+
+	CS_LOCK(drvdata->base);
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+
+	clk_disable_unprepare(drvdata->clk);
+
+	return sprintf(buf,
+		       "Depth:\t\t0x%x\n"
+		       "Status:\t\t0x%x\n"
+		       "RAM read ptr:\t0x%x\n"
+		       "RAM wrt ptr:\t0x%x\n"
+		       "Trigger cnt:\t0x%x\n"
+		       "Control:\t0x%x\n"
+		       "Flush status:\t0x%x\n"
+		       "Flush ctrl:\t0x%x\n",
+		       etb_rdr, etb_sr, etb_rrp, etb_rwp,
+		       etb_trg, etb_cr, etb_ffsr, etb_ffcr);
+out:
+	return -EINVAL;
+}
+static DEVICE_ATTR_RO(status);
+
+static ssize_t trigger_cntr_show(struct device *dev,
+			    struct device_attribute *attr, char *buf)
+{
+	struct etb_drvdata *drvdata = dev_get_drvdata(dev->parent);
+	unsigned long val = drvdata->trigger_cntr;
+
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t trigger_cntr_store(struct device *dev,
+			     struct device_attribute *attr,
+			     const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etb_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	drvdata->trigger_cntr = val;
+	return size;
+}
+static DEVICE_ATTR_RW(trigger_cntr);
+
+static struct attribute *coresight_etb_attrs[] = {
+	&dev_attr_trigger_cntr.attr,
+	&dev_attr_status.attr,
+	NULL,
+};
+ATTRIBUTE_GROUPS(coresight_etb);
+
+static int etb_probe(struct amba_device *adev, const struct amba_id *id)
+{
+	int ret;
+	void __iomem *base;
+	struct device *dev = &adev->dev;
+	struct coresight_platform_data *pdata = NULL;
+	struct etb_drvdata *drvdata;
+	struct resource *res = &adev->res;
+	struct coresight_desc *desc;
+	struct device_node *np = adev->dev.of_node;
+
+	if (np) {
+		pdata = of_get_coresight_platform_data(dev, np);
+		if (IS_ERR(pdata))
+			return PTR_ERR(pdata);
+		adev->dev.platform_data = pdata;
+	}
+
+	drvdata = devm_kzalloc(dev, sizeof(*drvdata), GFP_KERNEL);
+	if (!drvdata)
+		return -ENOMEM;
+
+	drvdata->dev = &adev->dev;
+	dev_set_drvdata(dev, drvdata);
+
+	/* validity for the resource is already checked by the AMBA core */
+	base = devm_ioremap_resource(dev, res);
+	if (IS_ERR(base))
+		return PTR_ERR(base);
+
+	drvdata->base = base;
+
+	spin_lock_init(&drvdata->spinlock);
+
+	drvdata->clk = adev->pclk;
+	ret = clk_prepare_enable(drvdata->clk);
+	if (ret)
+		return ret;
+
+	drvdata->buffer_depth =  etb_get_buffer_depth(drvdata);
+	clk_disable_unprepare(drvdata->clk);
+
+	if (drvdata->buffer_depth < 0)
+		return -EINVAL;
+
+	drvdata->buf = devm_kzalloc(dev,
+				    drvdata->buffer_depth * 4, GFP_KERNEL);
+	if (!drvdata->buf)
+		return -ENOMEM;
+
+	desc = devm_kzalloc(dev, sizeof(*desc), GFP_KERNEL);
+	if (!desc)
+		return -ENOMEM;
+
+	desc->type = CORESIGHT_DEV_TYPE_SINK;
+	desc->subtype.sink_subtype = CORESIGHT_DEV_SUBTYPE_SINK_BUFFER;
+	desc->ops = &etb_cs_ops;
+	desc->pdata = pdata;
+	desc->dev = dev;
+	desc->groups = coresight_etb_groups;
+	drvdata->csdev = coresight_register(desc);
+	if (IS_ERR(drvdata->csdev))
+		return PTR_ERR(drvdata->csdev);
+
+	drvdata->miscdev.name = pdata->name;
+	drvdata->miscdev.minor = MISC_DYNAMIC_MINOR;
+	drvdata->miscdev.fops = &etb_fops;
+	ret = misc_register(&drvdata->miscdev);
+	if (ret)
+		goto err_misc_register;
+
+	dev_info(dev, "ETB initialized\n");
+	return 0;
+
+err_misc_register:
+	coresight_unregister(drvdata->csdev);
+	return ret;
+}
+
+static int etb_remove(struct amba_device *adev)
+{
+	struct etb_drvdata *drvdata = amba_get_drvdata(adev);
+
+	misc_deregister(&drvdata->miscdev);
+	coresight_unregister(drvdata->csdev);
+	return 0;
+}
+
+static struct amba_id etb_ids[] = {
+	{
+		.id	= 0x0003b907,
+		.mask	= 0x0003ffff,
+	},
+	{ 0, 0},
+};
+
+static struct amba_driver etb_driver = {
+	.drv = {
+		.name	= "coresight-etb10",
+		.owner	= THIS_MODULE,
+	},
+	.probe		= etb_probe,
+	.remove		= etb_remove,
+	.id_table	= etb_ids,
+};
+
+static int __init etb_init(void)
+{
+	return amba_driver_register(&etb_driver);
+}
+module_init(etb_init);
+
+static void __exit etb_exit(void)
+{
+	amba_driver_unregister(&etb_driver);
+}
+module_exit(etb_exit);
+
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("CoreSight Embedded Trace Buffer driver");
diff --git a/drivers/coresight/coresight-etm-cp14.c b/drivers/coresight/coresight-etm-cp14.c
new file mode 100644
index 000000000000..12a220682117
--- /dev/null
+++ b/drivers/coresight/coresight-etm-cp14.c
@@ -0,0 +1,591 @@
+/* Copyright (c) 2012, The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/bug.h>
+#include <asm/hardware/cp14.h>
+
+#include "coresight-etm.h"
+
+int etm_readl_cp14(u32 reg, unsigned int *val)
+{
+	switch (reg) {
+	case ETMCR:
+		*val = etm_read(ETMCR);
+		return 0;
+	case ETMCCR:
+		*val = etm_read(ETMCCR);
+		return 0;
+	case ETMTRIGGER:
+		*val = etm_read(ETMTRIGGER);
+		return 0;
+	case ETMSR:
+		*val = etm_read(ETMSR);
+		return 0;
+	case ETMSCR:
+		*val = etm_read(ETMSCR);
+		return 0;
+	case ETMTSSCR:
+		*val = etm_read(ETMTSSCR);
+		return 0;
+	case ETMTEEVR:
+		*val = etm_read(ETMTEEVR);
+		return 0;
+	case ETMTECR1:
+		*val = etm_read(ETMTECR1);
+		return 0;
+	case ETMFFLR:
+		*val = etm_read(ETMFFLR);
+		return 0;
+	case ETMACVRn(0):
+		*val = etm_read(ETMACVR0);
+		return 0;
+	case ETMACVRn(1):
+		*val = etm_read(ETMACVR1);
+		return 0;
+	case ETMACVRn(2):
+		*val = etm_read(ETMACVR2);
+		return 0;
+	case ETMACVRn(3):
+		*val = etm_read(ETMACVR3);
+		return 0;
+	case ETMACVRn(4):
+		*val = etm_read(ETMACVR4);
+		return 0;
+	case ETMACVRn(5):
+		*val = etm_read(ETMACVR5);
+		return 0;
+	case ETMACVRn(6):
+		*val = etm_read(ETMACVR6);
+		return 0;
+	case ETMACVRn(7):
+		*val = etm_read(ETMACVR7);
+		return 0;
+	case ETMACVRn(8):
+		*val = etm_read(ETMACVR8);
+		return 0;
+	case ETMACVRn(9):
+		*val = etm_read(ETMACVR9);
+		return 0;
+	case ETMACVRn(10):
+		*val = etm_read(ETMACVR10);
+		return 0;
+	case ETMACVRn(11):
+		*val = etm_read(ETMACVR11);
+		return 0;
+	case ETMACVRn(12):
+		*val = etm_read(ETMACVR12);
+		return 0;
+	case ETMACVRn(13):
+		*val = etm_read(ETMACVR13);
+		return 0;
+	case ETMACVRn(14):
+		*val = etm_read(ETMACVR14);
+		return 0;
+	case ETMACVRn(15):
+		*val = etm_read(ETMACVR15);
+		return 0;
+	case ETMACTRn(0):
+		*val = etm_read(ETMACTR0);
+		return 0;
+	case ETMACTRn(1):
+		*val = etm_read(ETMACTR1);
+		return 0;
+	case ETMACTRn(2):
+		*val = etm_read(ETMACTR2);
+		return 0;
+	case ETMACTRn(3):
+		*val = etm_read(ETMACTR3);
+		return 0;
+	case ETMACTRn(4):
+		*val = etm_read(ETMACTR4);
+		return 0;
+	case ETMACTRn(5):
+		*val = etm_read(ETMACTR5);
+		return 0;
+	case ETMACTRn(6):
+		*val = etm_read(ETMACTR6);
+		return 0;
+	case ETMACTRn(7):
+		*val = etm_read(ETMACTR7);
+		return 0;
+	case ETMACTRn(8):
+		*val = etm_read(ETMACTR8);
+		return 0;
+	case ETMACTRn(9):
+		*val = etm_read(ETMACTR9);
+		return 0;
+	case ETMACTRn(10):
+		*val = etm_read(ETMACTR10);
+		return 0;
+	case ETMACTRn(11):
+		*val = etm_read(ETMACTR11);
+		return 0;
+	case ETMACTRn(12):
+		*val = etm_read(ETMACTR12);
+		return 0;
+	case ETMACTRn(13):
+		*val = etm_read(ETMACTR13);
+		return 0;
+	case ETMACTRn(14):
+		*val = etm_read(ETMACTR14);
+		return 0;
+	case ETMACTRn(15):
+		*val = etm_read(ETMACTR15);
+		return 0;
+	case ETMCNTRLDVRn(0):
+		*val = etm_read(ETMCNTRLDVR0);
+		return 0;
+	case ETMCNTRLDVRn(1):
+		*val = etm_read(ETMCNTRLDVR1);
+		return 0;
+	case ETMCNTRLDVRn(2):
+		*val = etm_read(ETMCNTRLDVR2);
+		return 0;
+	case ETMCNTRLDVRn(3):
+		*val = etm_read(ETMCNTRLDVR3);
+		return 0;
+	case ETMCNTENRn(0):
+		*val = etm_read(ETMCNTENR0);
+		return 0;
+	case ETMCNTENRn(1):
+		*val = etm_read(ETMCNTENR1);
+		return 0;
+	case ETMCNTENRn(2):
+		*val = etm_read(ETMCNTENR2);
+		return 0;
+	case ETMCNTENRn(3):
+		*val = etm_read(ETMCNTENR3);
+		return 0;
+	case ETMCNTRLDEVRn(0):
+		*val = etm_read(ETMCNTRLDEVR0);
+		return 0;
+	case ETMCNTRLDEVRn(1):
+		*val = etm_read(ETMCNTRLDEVR1);
+		return 0;
+	case ETMCNTRLDEVRn(2):
+		*val = etm_read(ETMCNTRLDEVR2);
+		return 0;
+	case ETMCNTRLDEVRn(3):
+		*val = etm_read(ETMCNTRLDEVR3);
+		return 0;
+	case ETMCNTVRn(0):
+		*val = etm_read(ETMCNTVR0);
+		return 0;
+	case ETMCNTVRn(1):
+		*val = etm_read(ETMCNTVR1);
+		return 0;
+	case ETMCNTVRn(2):
+		*val = etm_read(ETMCNTVR2);
+		return 0;
+	case ETMCNTVRn(3):
+		*val = etm_read(ETMCNTVR3);
+		return 0;
+	case ETMSQ12EVR:
+		*val = etm_read(ETMSQ12EVR);
+		return 0;
+	case ETMSQ21EVR:
+		*val = etm_read(ETMSQ21EVR);
+		return 0;
+	case ETMSQ23EVR:
+		*val = etm_read(ETMSQ23EVR);
+		return 0;
+	case ETMSQ31EVR:
+		*val = etm_read(ETMSQ31EVR);
+		return 0;
+	case ETMSQ32EVR:
+		*val = etm_read(ETMSQ32EVR);
+		return 0;
+	case ETMSQ13EVR:
+		*val = etm_read(ETMSQ13EVR);
+		return 0;
+	case ETMSQR:
+		*val = etm_read(ETMSQR);
+		return 0;
+	case ETMEXTOUTEVRn(0):
+		*val = etm_read(ETMEXTOUTEVR0);
+		return 0;
+	case ETMEXTOUTEVRn(1):
+		*val = etm_read(ETMEXTOUTEVR1);
+		return 0;
+	case ETMEXTOUTEVRn(2):
+		*val = etm_read(ETMEXTOUTEVR2);
+		return 0;
+	case ETMEXTOUTEVRn(3):
+		*val = etm_read(ETMEXTOUTEVR3);
+		return 0;
+	case ETMCIDCVRn(0):
+		*val = etm_read(ETMCIDCVR0);
+		return 0;
+	case ETMCIDCVRn(1):
+		*val = etm_read(ETMCIDCVR1);
+		return 0;
+	case ETMCIDCVRn(2):
+		*val = etm_read(ETMCIDCVR2);
+		return 0;
+	case ETMCIDCMR:
+		*val = etm_read(ETMCIDCMR);
+		return 0;
+	case ETMIMPSPEC0:
+		*val = etm_read(ETMIMPSPEC0);
+		return 0;
+	case ETMIMPSPEC1:
+		*val = etm_read(ETMIMPSPEC1);
+		return 0;
+	case ETMIMPSPEC2:
+		*val = etm_read(ETMIMPSPEC2);
+		return 0;
+	case ETMIMPSPEC3:
+		*val = etm_read(ETMIMPSPEC3);
+		return 0;
+	case ETMIMPSPEC4:
+		*val = etm_read(ETMIMPSPEC4);
+		return 0;
+	case ETMIMPSPEC5:
+		*val = etm_read(ETMIMPSPEC5);
+		return 0;
+	case ETMIMPSPEC6:
+		*val = etm_read(ETMIMPSPEC6);
+		return 0;
+	case ETMIMPSPEC7:
+		*val = etm_read(ETMIMPSPEC7);
+		return 0;
+	case ETMSYNCFR:
+		*val = etm_read(ETMSYNCFR);
+		return 0;
+	case ETMIDR:
+		*val = etm_read(ETMIDR);
+		return 0;
+	case ETMCCER:
+		*val = etm_read(ETMCCER);
+		return 0;
+	case ETMEXTINSELR:
+		*val = etm_read(ETMEXTINSELR);
+		return 0;
+	case ETMTESSEICR:
+		*val = etm_read(ETMTESSEICR);
+		return 0;
+	case ETMEIBCR:
+		*val = etm_read(ETMEIBCR);
+		return 0;
+	case ETMTSEVR:
+		*val = etm_read(ETMTSEVR);
+		return 0;
+	case ETMAUXCR:
+		*val = etm_read(ETMAUXCR);
+		return 0;
+	case ETMTRACEIDR:
+		*val = etm_read(ETMTRACEIDR);
+		return 0;
+	case ETMVMIDCVR:
+		*val = etm_read(ETMVMIDCVR);
+		return 0;
+	case ETMOSLSR:
+		*val = etm_read(ETMOSLSR);
+		return 0;
+	case ETMOSSRR:
+		*val = etm_read(ETMOSSRR);
+		return 0;
+	case ETMPDCR:
+		*val = etm_read(ETMPDCR);
+		return 0;
+	case ETMPDSR:
+		*val = etm_read(ETMPDSR);
+		return 0;
+	default:
+		*val = 0;
+		return -EINVAL;
+	}
+}
+
+int etm_writel_cp14(u32 reg, u32 val)
+{
+	switch (reg) {
+	case ETMCR:
+		etm_write(val, ETMCR);
+		break;
+	case ETMTRIGGER:
+		etm_write(val, ETMTRIGGER);
+		break;
+	case ETMSR:
+		etm_write(val, ETMSR);
+		break;
+	case ETMTSSCR:
+		etm_write(val, ETMTSSCR);
+		break;
+	case ETMTEEVR:
+		etm_write(val, ETMTEEVR);
+		break;
+	case ETMTECR1:
+		etm_write(val, ETMTECR1);
+		break;
+	case ETMFFLR:
+		etm_write(val, ETMFFLR);
+		break;
+	case ETMACVRn(0):
+		etm_write(val, ETMACVR0);
+		break;
+	case ETMACVRn(1):
+		etm_write(val, ETMACVR1);
+		break;
+	case ETMACVRn(2):
+		etm_write(val, ETMACVR2);
+		break;
+	case ETMACVRn(3):
+		etm_write(val, ETMACVR3);
+		break;
+	case ETMACVRn(4):
+		etm_write(val, ETMACVR4);
+		break;
+	case ETMACVRn(5):
+		etm_write(val, ETMACVR5);
+		break;
+	case ETMACVRn(6):
+		etm_write(val, ETMACVR6);
+		break;
+	case ETMACVRn(7):
+		etm_write(val, ETMACVR7);
+		break;
+	case ETMACVRn(8):
+		etm_write(val, ETMACVR8);
+		break;
+	case ETMACVRn(9):
+		etm_write(val, ETMACVR9);
+		break;
+	case ETMACVRn(10):
+		etm_write(val, ETMACVR10);
+		break;
+	case ETMACVRn(11):
+		etm_write(val, ETMACVR11);
+		break;
+	case ETMACVRn(12):
+		etm_write(val, ETMACVR12);
+		break;
+	case ETMACVRn(13):
+		etm_write(val, ETMACVR13);
+		break;
+	case ETMACVRn(14):
+		etm_write(val, ETMACVR14);
+		break;
+	case ETMACVRn(15):
+		etm_write(val, ETMACVR15);
+		break;
+	case ETMACTRn(0):
+		etm_write(val, ETMACTR0);
+		break;
+	case ETMACTRn(1):
+		etm_write(val, ETMACTR1);
+		break;
+	case ETMACTRn(2):
+		etm_write(val, ETMACTR2);
+		break;
+	case ETMACTRn(3):
+		etm_write(val, ETMACTR3);
+		break;
+	case ETMACTRn(4):
+		etm_write(val, ETMACTR4);
+		break;
+	case ETMACTRn(5):
+		etm_write(val, ETMACTR5);
+		break;
+	case ETMACTRn(6):
+		etm_write(val, ETMACTR6);
+		break;
+	case ETMACTRn(7):
+		etm_write(val, ETMACTR7);
+		break;
+	case ETMACTRn(8):
+		etm_write(val, ETMACTR8);
+		break;
+	case ETMACTRn(9):
+		etm_write(val, ETMACTR9);
+		break;
+	case ETMACTRn(10):
+		etm_write(val, ETMACTR10);
+		break;
+	case ETMACTRn(11):
+		etm_write(val, ETMACTR11);
+		break;
+	case ETMACTRn(12):
+		etm_write(val, ETMACTR12);
+		break;
+	case ETMACTRn(13):
+		etm_write(val, ETMACTR13);
+		break;
+	case ETMACTRn(14):
+		etm_write(val, ETMACTR14);
+		break;
+	case ETMACTRn(15):
+		etm_write(val, ETMACTR15);
+		break;
+	case ETMCNTRLDVRn(0):
+		etm_write(val, ETMCNTRLDVR0);
+		break;
+	case ETMCNTRLDVRn(1):
+		etm_write(val, ETMCNTRLDVR1);
+		break;
+	case ETMCNTRLDVRn(2):
+		etm_write(val, ETMCNTRLDVR2);
+		break;
+	case ETMCNTRLDVRn(3):
+		etm_write(val, ETMCNTRLDVR3);
+		break;
+	case ETMCNTENRn(0):
+		etm_write(val, ETMCNTENR0);
+		break;
+	case ETMCNTENRn(1):
+		etm_write(val, ETMCNTENR1);
+		break;
+	case ETMCNTENRn(2):
+		etm_write(val, ETMCNTENR2);
+		break;
+	case ETMCNTENRn(3):
+		etm_write(val, ETMCNTENR3);
+		break;
+	case ETMCNTRLDEVRn(0):
+		etm_write(val, ETMCNTRLDEVR0);
+		break;
+	case ETMCNTRLDEVRn(1):
+		etm_write(val, ETMCNTRLDEVR1);
+		break;
+	case ETMCNTRLDEVRn(2):
+		etm_write(val, ETMCNTRLDEVR2);
+		break;
+	case ETMCNTRLDEVRn(3):
+		etm_write(val, ETMCNTRLDEVR3);
+		break;
+	case ETMCNTVRn(0):
+		etm_write(val, ETMCNTVR0);
+		break;
+	case ETMCNTVRn(1):
+		etm_write(val, ETMCNTVR1);
+		break;
+	case ETMCNTVRn(2):
+		etm_write(val, ETMCNTVR2);
+		break;
+	case ETMCNTVRn(3):
+		etm_write(val, ETMCNTVR3);
+		break;
+	case ETMSQ12EVR:
+		etm_write(val, ETMSQ12EVR);
+		break;
+	case ETMSQ21EVR:
+		etm_write(val, ETMSQ21EVR);
+		break;
+	case ETMSQ23EVR:
+		etm_write(val, ETMSQ23EVR);
+		break;
+	case ETMSQ31EVR:
+		etm_write(val, ETMSQ31EVR);
+		break;
+	case ETMSQ32EVR:
+		etm_write(val, ETMSQ32EVR);
+		break;
+	case ETMSQ13EVR:
+		etm_write(val, ETMSQ13EVR);
+		break;
+	case ETMSQR:
+		etm_write(val, ETMSQR);
+		break;
+	case ETMEXTOUTEVRn(0):
+		etm_write(val, ETMEXTOUTEVR0);
+		break;
+	case ETMEXTOUTEVRn(1):
+		etm_write(val, ETMEXTOUTEVR1);
+		break;
+	case ETMEXTOUTEVRn(2):
+		etm_write(val, ETMEXTOUTEVR2);
+		break;
+	case ETMEXTOUTEVRn(3):
+		etm_write(val, ETMEXTOUTEVR3);
+		break;
+	case ETMCIDCVRn(0):
+		etm_write(val, ETMCIDCVR0);
+		break;
+	case ETMCIDCVRn(1):
+		etm_write(val, ETMCIDCVR1);
+		break;
+	case ETMCIDCVRn(2):
+		etm_write(val, ETMCIDCVR2);
+		break;
+	case ETMCIDCMR:
+		etm_write(val, ETMCIDCMR);
+		break;
+	case ETMIMPSPEC0:
+		etm_write(val, ETMIMPSPEC0);
+		break;
+	case ETMIMPSPEC1:
+		etm_write(val, ETMIMPSPEC1);
+		break;
+	case ETMIMPSPEC2:
+		etm_write(val, ETMIMPSPEC2);
+		break;
+	case ETMIMPSPEC3:
+		etm_write(val, ETMIMPSPEC3);
+		break;
+	case ETMIMPSPEC4:
+		etm_write(val, ETMIMPSPEC4);
+		break;
+	case ETMIMPSPEC5:
+		etm_write(val, ETMIMPSPEC5);
+		break;
+	case ETMIMPSPEC6:
+		etm_write(val, ETMIMPSPEC6);
+		break;
+	case ETMIMPSPEC7:
+		etm_write(val, ETMIMPSPEC7);
+		break;
+	case ETMSYNCFR:
+		etm_write(val, ETMSYNCFR);
+		break;
+	case ETMEXTINSELR:
+		etm_write(val, ETMEXTINSELR);
+		break;
+	case ETMTESSEICR:
+		etm_write(val, ETMTESSEICR);
+		break;
+	case ETMEIBCR:
+		etm_write(val, ETMEIBCR);
+		break;
+	case ETMTSEVR:
+		etm_write(val, ETMTSEVR);
+		break;
+	case ETMAUXCR:
+		etm_write(val, ETMAUXCR);
+		break;
+	case ETMTRACEIDR:
+		etm_write(val, ETMTRACEIDR);
+		break;
+	case ETMVMIDCVR:
+		etm_write(val, ETMVMIDCVR);
+		break;
+	case ETMOSLAR:
+		etm_write(val, ETMOSLAR);
+		break;
+	case ETMOSSRR:
+		etm_write(val, ETMOSSRR);
+		break;
+	case ETMPDCR:
+		etm_write(val, ETMPDCR);
+		break;
+	case ETMPDSR:
+		etm_write(val, ETMPDSR);
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
diff --git a/drivers/coresight/coresight-etm.h b/drivers/coresight/coresight-etm.h
new file mode 100644
index 000000000000..501c5fac8a45
--- /dev/null
+++ b/drivers/coresight/coresight-etm.h
@@ -0,0 +1,251 @@
+/* Copyright (c) 2014-2015, The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#ifndef _CORESIGHT_CORESIGHT_ETM_H
+#define _CORESIGHT_CORESIGHT_ETM_H
+
+#include <linux/spinlock.h>
+#include "coresight-priv.h"
+
+/*
+ * Device registers:
+ * 0x000 - 0x2FC: Trace         registers
+ * 0x300 - 0x314: Management    registers
+ * 0x318 - 0xEFC: Trace         registers
+ *
+ * Coresight registers
+ * 0xF00 - 0xF9C: Management    registers
+ * 0xFA0 - 0xFA4: Management    registers in PFTv1.0
+ *                Trace         registers in PFTv1.1
+ * 0xFA8 - 0xFFC: Management    registers
+ */
+
+/* Trace registers (0x000-0x2FC) */
+#define ETMCR			0x000
+#define ETMCCR			0x004
+#define ETMTRIGGER		0x008
+#define ETMSR			0x010
+#define ETMSCR			0x014
+#define ETMTSSCR		0x018
+#define ETMTECR2		0x01c
+#define ETMTEEVR		0x020
+#define ETMTECR1		0x024
+#define ETMFFLR			0x02c
+#define ETMACVRn(n)		(0x040 + (n * 4))
+#define ETMACTRn(n)		(0x080 + (n * 4))
+#define ETMCNTRLDVRn(n)		(0x140 + (n * 4))
+#define ETMCNTENRn(n)		(0x150 + (n * 4))
+#define ETMCNTRLDEVRn(n)	(0x160 + (n * 4))
+#define ETMCNTVRn(n)		(0x170 + (n * 4))
+#define ETMSQ12EVR		0x180
+#define ETMSQ21EVR		0x184
+#define ETMSQ23EVR		0x188
+#define ETMSQ31EVR		0x18c
+#define ETMSQ32EVR		0x190
+#define ETMSQ13EVR		0x194
+#define ETMSQR			0x19c
+#define ETMEXTOUTEVRn(n)	(0x1a0 + (n * 4))
+#define ETMCIDCVRn(n)		(0x1b0 + (n * 4))
+#define ETMCIDCMR		0x1bc
+#define ETMIMPSPEC0		0x1c0
+#define ETMIMPSPEC1		0x1c4
+#define ETMIMPSPEC2		0x1c8
+#define ETMIMPSPEC3		0x1cc
+#define ETMIMPSPEC4		0x1d0
+#define ETMIMPSPEC5		0x1d4
+#define ETMIMPSPEC6		0x1d8
+#define ETMIMPSPEC7		0x1dc
+#define ETMSYNCFR		0x1e0
+#define ETMIDR			0x1e4
+#define ETMCCER			0x1e8
+#define ETMEXTINSELR		0x1ec
+#define ETMTESSEICR		0x1f0
+#define ETMEIBCR		0x1f4
+#define ETMTSEVR		0x1f8
+#define ETMAUXCR		0x1fc
+#define ETMTRACEIDR		0x200
+#define ETMVMIDCVR		0x240
+/* Management registers (0x300-0x314) */
+#define ETMOSLAR		0x300
+#define ETMOSLSR		0x304
+#define ETMOSSRR		0x308
+#define ETMPDCR			0x310
+#define ETMPDSR			0x314
+#define ETM_MAX_ADDR_CMP	16
+#define ETM_MAX_CNTR		4
+#define ETM_MAX_CTXID_CMP	3
+
+/* Register definition */
+/* ETMCR - 0x00 */
+#define ETMCR_PWD_DWN		BIT(0)
+#define ETMCR_STALL_MODE	BIT(7)
+#define ETMCR_ETM_PRG		BIT(10)
+#define ETMCR_ETM_EN		BIT(11)
+#define ETMCR_CYC_ACC		BIT(12)
+#define ETMCR_CTXID_SIZE	(BIT(14)|BIT(15))
+#define ETMCR_TIMESTAMP_EN	BIT(28)
+/* ETMCCR - 0x04 */
+#define ETMCCR_FIFOFULL		BIT(23)
+/* ETMPDCR - 0x310 */
+#define ETMPDCR_PWD_UP		BIT(3)
+/* ETMTECR1 - 0x024 */
+#define ETMTECR1_ADDR_COMP_1	BIT(0)
+#define ETMTECR1_INC_EXC	BIT(24)
+#define ETMTECR1_START_STOP	BIT(25)
+/* ETMCCER - 0x1E8 */
+#define ETMCCER_TIMESTAMP	BIT(22)
+
+#define ETM_MODE_EXCLUDE	BIT(0)
+#define ETM_MODE_CYCACC		BIT(1)
+#define ETM_MODE_STALL		BIT(2)
+#define ETM_MODE_TIMESTAMP	BIT(3)
+#define ETM_MODE_CTXID		BIT(4)
+#define ETM_MODE_ALL		0x1f
+
+#define ETM_SQR_MASK		0x3
+#define ETM_TRACEID_MASK	0x3f
+#define ETM_EVENT_MASK		0x1ffff
+#define ETM_SYNC_MASK		0xfff
+#define ETM_ALL_MASK		0xffffffff
+
+#define ETMSR_PROG_BIT		1
+#define ETM_SEQ_STATE_MAX_VAL	(0x2)
+#define PORT_SIZE_MASK		(GENMASK(21, 21) | GENMASK(6, 4))
+
+#define ETM_HARD_WIRE_RES_A	/* Hard wired, always true */	\
+				((0x0f << 0)	|		\
+				/* Resource index A */		\
+				(0x06 << 4))
+
+#define ETM_ADD_COMP_0		/* Single addr comparator 1 */	\
+				((0x00 << 7)	|		\
+				/* Resource index B */		\
+				(0x00 << 11))
+
+#define ETM_EVENT_NOT_A		BIT(14) /* NOT(A) */
+
+#define ETM_DEFAULT_EVENT_VAL	(ETM_HARD_WIRE_RES_A	|	\
+				 ETM_ADD_COMP_0		|	\
+				 ETM_EVENT_NOT_A)
+/**
+ * struct etm_drvdata - specifics associated to an ETM component
+ * @base:	memory mapped base address for this component.
+ * @dev:	the device entity associated to this component.
+ * @csdev:	component vitals needed by the framework.
+ * @clk:	the clock this component is associated to.
+ * @spinlock:	only one at a time pls.
+ * @cpu:	the cpu this component is affined to.
+ * @port_size:	port size as reported by ETMCR bit 4-6 and 21.
+ * @arch:	ETM/PTM version number.
+ * @use_cpu14:	true if management registers need to be accessed via CP14.
+ * @enable:	is this ETM/PTM currently tracing.
+ * @sticky_enable: true if ETM base configuration has been done.
+ * @boot_enable:true if we should start tracing at boot time.
+ * @os_unlock:	true if access to management registers is allowed.
+ * @nr_addr_cmp:Number of pairs of address comparators as found in ETMCCR.
+ * @nr_cntr:	Number of counters as found in ETMCCR bit 13-15.
+ * @nr_ext_inp:	Number of external input as found in ETMCCR bit 17-19.
+ * @nr_ext_out:	Number of external output as found in ETMCCR bit 20-22.
+ * @nr_ctxid_cmp: Number of contextID comparators as found in ETMCCR bit 24-25.
+ * @etmccr:	value of register ETMCCR.
+ * @etmccer:	value of register ETMCCER.
+ * @traceid:	value of the current ID for this component.
+ * @mode:	controls various modes supported by this ETM/PTM.
+ * @ctrl:	used in conjunction with @mode.
+ * @trigger_event: setting for register ETMTRIGGER.
+ * @startstop_ctrl: setting for register ETMTSSCR.
+ * @enable_event: setting for register ETMTEEVR.
+ * @enable_ctrl1: setting for register ETMTECR1.
+ * @fifofull_level: setting for register ETMFFLR.
+ * @addr_idx:	index for the address comparator selection.
+ * @addr_val:	value for address comparator register.
+ * @addr_acctype: access type for address comparator register.
+ * @addr_type:	current status of the comparator register.
+ * @cntr_idx:	index for the counter register selection.
+ * @cntr_rld_val: reload value of a counter register.
+ * @cntr_event:	control for counter enable register.
+ * @cntr_rld_event: value for counter reload event register.
+ * @cntr_val:	counter value register.
+ * @seq_12_event: event causing the transition from 1 to 2.
+ * @seq_21_event: event causing the transition from 2 to 1.
+ * @seq_23_event: event causing the transition from 2 to 3.
+ * @seq_31_event: event causing the transition from 3 to 1.
+ * @seq_32_event: event causing the transition from 3 to 2.
+ * @seq_13_event: event causing the transition from 1 to 3.
+ * @seq_curr_state: current value of the sequencer register.
+ * @ctxid_idx: index for the context ID registers.
+ * @ctxid_val: value for the context ID to trigger on.
+ * @ctxid_mask: mask applicable to all the context IDs.
+ * @sync_freq:	Synchronisation frequency.
+ * @timestamp_event: Defines an event that requests the insertion
+		     of a timestamp into the trace stream.
+ */
+struct etm_drvdata {
+	void __iomem			*base;
+	struct device			*dev;
+	struct coresight_device		*csdev;
+	struct clk			*clk;
+	spinlock_t			spinlock;
+	int				cpu;
+	int				port_size;
+	u8				arch;
+	bool				use_cp14;
+	bool				enable;
+	bool				sticky_enable;
+	bool				boot_enable;
+	bool				os_unlock;
+	u8				nr_addr_cmp;
+	u8				nr_cntr;
+	u8				nr_ext_inp;
+	u8				nr_ext_out;
+	u8				nr_ctxid_cmp;
+	u32				etmccr;
+	u32				etmccer;
+	u32				traceid;
+	u32				mode;
+	u32				ctrl;
+	u32				trigger_event;
+	u32				startstop_ctrl;
+	u32				enable_event;
+	u32				enable_ctrl1;
+	u32				fifofull_level;
+	u8				addr_idx;
+	u32				addr_val[ETM_MAX_ADDR_CMP];
+	u32				addr_acctype[ETM_MAX_ADDR_CMP];
+	u32				addr_type[ETM_MAX_ADDR_CMP];
+	u8				cntr_idx;
+	u32				cntr_rld_val[ETM_MAX_CNTR];
+	u32				cntr_event[ETM_MAX_CNTR];
+	u32				cntr_rld_event[ETM_MAX_CNTR];
+	u32				cntr_val[ETM_MAX_CNTR];
+	u32				seq_12_event;
+	u32				seq_21_event;
+	u32				seq_23_event;
+	u32				seq_31_event;
+	u32				seq_32_event;
+	u32				seq_13_event;
+	u32				seq_curr_state;
+	u8				ctxid_idx;
+	u32				ctxid_val[ETM_MAX_CTXID_CMP];
+	u32				ctxid_mask;
+	u32				sync_freq;
+	u32				timestamp_event;
+};
+
+enum etm_addr_type {
+	ETM_ADDR_TYPE_NONE,
+	ETM_ADDR_TYPE_SINGLE,
+	ETM_ADDR_TYPE_RANGE,
+	ETM_ADDR_TYPE_START,
+	ETM_ADDR_TYPE_STOP,
+};
+#endif
diff --git a/drivers/coresight/coresight-etm3x.c b/drivers/coresight/coresight-etm3x.c
new file mode 100644
index 000000000000..d9e3ed6aa857
--- /dev/null
+++ b/drivers/coresight/coresight-etm3x.c
@@ -0,0 +1,1928 @@
+/* Copyright (c) 2011-2012, The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/device.h>
+#include <linux/io.h>
+#include <linux/err.h>
+#include <linux/fs.h>
+#include <linux/slab.h>
+#include <linux/delay.h>
+#include <linux/smp.h>
+#include <linux/sysfs.h>
+#include <linux/stat.h>
+#include <linux/clk.h>
+#include <linux/cpu.h>
+#include <linux/of.h>
+#include <linux/coresight.h>
+#include <linux/amba/bus.h>
+#include <linux/seq_file.h>
+#include <linux/uaccess.h>
+#include <asm/sections.h>
+
+#include "coresight-etm.h"
+
+#ifdef CONFIG_CORESIGHT_SOURCE_ETM_DEFAULT_ENABLE
+static int boot_enable = 1;
+#else
+static int boot_enable;
+#endif
+module_param_named(
+	boot_enable, boot_enable, int, S_IRUGO
+);
+
+/* The number of ETM/PTM currently registered */
+static int etm_count;
+static struct etm_drvdata *etmdrvdata[NR_CPUS];
+
+static inline void etm_writel(struct etm_drvdata *drvdata,
+			      u32 val, u32 off)
+{
+	if (drvdata->use_cp14) {
+		if (etm_writel_cp14(off, val)) {
+			dev_err(drvdata->dev,
+				"invalid CP14 access to ETM reg: %#x", off);
+		}
+	} else {
+		writel_relaxed(val, drvdata->base + off);
+	}
+}
+
+static inline unsigned int etm_readl(struct etm_drvdata *drvdata, u32 off)
+{
+	u32 val;
+
+	if (drvdata->use_cp14) {
+		if (etm_readl_cp14(off, &val)) {
+			dev_err(drvdata->dev,
+				"invalid CP14 access to ETM reg: %#x", off);
+		}
+	} else {
+		val = readl_relaxed(drvdata->base + off);
+	}
+
+	return val;
+}
+
+/*
+ * Memory mapped writes to clear os lock are not supported on some processors
+ * and OS lock must be unlocked before any memory mapped access on such
+ * processors, otherwise memory mapped reads/writes will be invalid.
+ */
+static void etm_os_unlock(void *info)
+{
+	struct etm_drvdata *drvdata = (struct etm_drvdata *)info;
+	/* Writing any value to ETMOSLAR unlocks the trace registers */
+	etm_writel(drvdata, 0x0, ETMOSLAR);
+	isb();
+}
+
+static void etm_set_pwrdwn(struct etm_drvdata *drvdata)
+{
+	u32 etmcr;
+
+	/* Ensure pending cp14 accesses complete before setting pwrdwn */
+	mb();
+	isb();
+	etmcr = etm_readl(drvdata, ETMCR);
+	etmcr |= ETMCR_PWD_DWN;
+	etm_writel(drvdata, etmcr, ETMCR);
+}
+
+static void etm_clr_pwrdwn(struct etm_drvdata *drvdata)
+{
+	u32 etmcr;
+
+	etmcr = etm_readl(drvdata, ETMCR);
+	etmcr &= ~ETMCR_PWD_DWN;
+	etm_writel(drvdata, etmcr, ETMCR);
+	/* Ensure pwrup completes before subsequent cp14 accesses */
+	mb();
+	isb();
+}
+
+static void etm_set_pwrup(struct etm_drvdata *drvdata)
+{
+	u32 etmpdcr;
+
+	etmpdcr = readl_relaxed(drvdata->base + ETMPDCR);
+	etmpdcr |= ETMPDCR_PWD_UP;
+	writel_relaxed(etmpdcr, drvdata->base + ETMPDCR);
+	/* Ensure pwrup completes before subsequent cp14 accesses */
+	mb();
+	isb();
+}
+
+static void etm_clr_pwrup(struct etm_drvdata *drvdata)
+{
+	u32 etmpdcr;
+
+	/* Ensure pending cp14 accesses complete before clearing pwrup */
+	mb();
+	isb();
+	etmpdcr = readl_relaxed(drvdata->base + ETMPDCR);
+	etmpdcr &= ~ETMPDCR_PWD_UP;
+	writel_relaxed(etmpdcr, drvdata->base + ETMPDCR);
+}
+
+/**
+ * coresight_timeout_etm - loop until a bit has changed to a specific state.
+ * @drvdata: etm's private data structure.
+ * @offset: address of a register, starting from @addr.
+ * @position: the position of the bit of interest.
+ * @value: the value the bit should have.
+ *
+ * Basically the same as @coresight_timeout except for the register access
+ * method where we have to account for CP14 configurations.
+
+ * Return: 0 as soon as the bit has taken the desired state or -EAGAIN if
+ * TIMEOUT_US has elapsed, which ever happens first.
+ */
+
+static int coresight_timeout_etm(struct etm_drvdata *drvdata, u32 offset,
+				  int position, int value)
+{
+	int i;
+	u32 val;
+
+	for (i = TIMEOUT_US; i > 0; i--) {
+		val = etm_readl(drvdata, offset);
+		/* Waiting on the bit to go from 0 to 1 */
+		if (value) {
+			if (val & BIT(position))
+				return 0;
+		/* Waiting on the bit to go from 1 to 0 */
+		} else {
+			if (!(val & BIT(position)))
+				return 0;
+		}
+
+		/*
+		 * Delay is arbitrary - the specification doesn't say how long
+		 * we are expected to wait.  Extra check required to make sure
+		 * we don't wait needlessly on the last iteration.
+		 */
+		if (i - 1)
+			udelay(1);
+	}
+
+	return -EAGAIN;
+}
+
+
+static void etm_set_prog(struct etm_drvdata *drvdata)
+{
+	u32 etmcr;
+
+	etmcr = etm_readl(drvdata, ETMCR);
+	etmcr |= ETMCR_ETM_PRG;
+	etm_writel(drvdata, etmcr, ETMCR);
+	/*
+	 * Recommended by spec for cp14 accesses to ensure etmcr write is
+	 * complete before polling etmsr
+	 */
+	isb();
+	if (coresight_timeout_etm(drvdata, ETMSR, ETMSR_PROG_BIT, 1)) {
+		dev_err(drvdata->dev,
+			"timeout observed when probing at offset %#x\n", ETMSR);
+	}
+}
+
+static void etm_clr_prog(struct etm_drvdata *drvdata)
+{
+	u32 etmcr;
+
+	etmcr = etm_readl(drvdata, ETMCR);
+	etmcr &= ~ETMCR_ETM_PRG;
+	etm_writel(drvdata, etmcr, ETMCR);
+	/*
+	 * Recommended by spec for cp14 accesses to ensure etmcr write is
+	 * complete before polling etmsr
+	 */
+	isb();
+	if (coresight_timeout_etm(drvdata, ETMSR, ETMSR_PROG_BIT, 0)) {
+		dev_err(drvdata->dev,
+			"timeout observed when probing at offset %#x\n", ETMSR);
+	}
+}
+
+static void etm_set_default(struct etm_drvdata *drvdata)
+{
+	int i;
+
+	drvdata->trigger_event = ETM_DEFAULT_EVENT_VAL;
+	drvdata->enable_event = ETM_HARD_WIRE_RES_A;
+
+	drvdata->seq_12_event = ETM_DEFAULT_EVENT_VAL;
+	drvdata->seq_21_event = ETM_DEFAULT_EVENT_VAL;
+	drvdata->seq_23_event = ETM_DEFAULT_EVENT_VAL;
+	drvdata->seq_31_event = ETM_DEFAULT_EVENT_VAL;
+	drvdata->seq_32_event = ETM_DEFAULT_EVENT_VAL;
+	drvdata->seq_13_event = ETM_DEFAULT_EVENT_VAL;
+	drvdata->timestamp_event = ETM_DEFAULT_EVENT_VAL;
+
+	for (i = 0; i < drvdata->nr_cntr; i++) {
+		drvdata->cntr_rld_val[i] = 0x0;
+		drvdata->cntr_event[i] = ETM_DEFAULT_EVENT_VAL;
+		drvdata->cntr_rld_event[i] = ETM_DEFAULT_EVENT_VAL;
+		drvdata->cntr_val[i] = 0x0;
+	}
+
+	drvdata->seq_curr_state = 0x0;
+	drvdata->ctxid_idx = 0x0;
+	for (i = 0; i < drvdata->nr_ctxid_cmp; i++)
+		drvdata->ctxid_val[i] = 0x0;
+	drvdata->ctxid_mask = 0x0;
+}
+
+static void etm_enable_hw(void *info)
+{
+	int i;
+	u32 etmcr;
+	struct etm_drvdata *drvdata = info;
+
+	CS_UNLOCK(drvdata->base);
+
+	/* Turn engine on */
+	etm_clr_pwrdwn(drvdata);
+	/* Apply power to trace registers */
+	etm_set_pwrup(drvdata);
+	/* Make sure all registers are accessible */
+	etm_os_unlock(drvdata);
+
+	etm_set_prog(drvdata);
+
+	etmcr = etm_readl(drvdata, ETMCR);
+	etmcr &= (ETMCR_PWD_DWN | ETMCR_ETM_PRG);
+	etmcr |= drvdata->port_size;
+	etm_writel(drvdata, drvdata->ctrl | etmcr, ETMCR);
+	etm_writel(drvdata, drvdata->trigger_event, ETMTRIGGER);
+	etm_writel(drvdata, drvdata->startstop_ctrl, ETMTSSCR);
+	etm_writel(drvdata, drvdata->enable_event, ETMTEEVR);
+	etm_writel(drvdata, drvdata->enable_ctrl1, ETMTECR1);
+	etm_writel(drvdata, drvdata->fifofull_level, ETMFFLR);
+	for (i = 0; i < drvdata->nr_addr_cmp; i++) {
+		etm_writel(drvdata, drvdata->addr_val[i], ETMACVRn(i));
+		etm_writel(drvdata, drvdata->addr_acctype[i], ETMACTRn(i));
+	}
+	for (i = 0; i < drvdata->nr_cntr; i++) {
+		etm_writel(drvdata, drvdata->cntr_rld_val[i], ETMCNTRLDVRn(i));
+		etm_writel(drvdata, drvdata->cntr_event[i], ETMCNTENRn(i));
+		etm_writel(drvdata, drvdata->cntr_rld_event[i],
+			   ETMCNTRLDEVRn(i));
+		etm_writel(drvdata, drvdata->cntr_val[i], ETMCNTVRn(i));
+	}
+	etm_writel(drvdata, drvdata->seq_12_event, ETMSQ12EVR);
+	etm_writel(drvdata, drvdata->seq_21_event, ETMSQ21EVR);
+	etm_writel(drvdata, drvdata->seq_23_event, ETMSQ23EVR);
+	etm_writel(drvdata, drvdata->seq_31_event, ETMSQ31EVR);
+	etm_writel(drvdata, drvdata->seq_32_event, ETMSQ32EVR);
+	etm_writel(drvdata, drvdata->seq_13_event, ETMSQ13EVR);
+	etm_writel(drvdata, drvdata->seq_curr_state, ETMSQR);
+	for (i = 0; i < drvdata->nr_ext_out; i++)
+		etm_writel(drvdata, ETM_DEFAULT_EVENT_VAL, ETMEXTOUTEVRn(i));
+	for (i = 0; i < drvdata->nr_ctxid_cmp; i++)
+		etm_writel(drvdata, drvdata->ctxid_val[i], ETMCIDCVRn(i));
+	etm_writel(drvdata, drvdata->ctxid_mask, ETMCIDCMR);
+	etm_writel(drvdata, drvdata->sync_freq, ETMSYNCFR);
+	/* No external input selected */
+	etm_writel(drvdata, 0x0, ETMEXTINSELR);
+	etm_writel(drvdata, drvdata->timestamp_event, ETMTSEVR);
+	/* No auxiliary control selected */
+	etm_writel(drvdata, 0x0, ETMAUXCR);
+	etm_writel(drvdata, drvdata->traceid, ETMTRACEIDR);
+	/* No VMID comparator value selected */
+	etm_writel(drvdata, 0x0, ETMVMIDCVR);
+
+	/* Ensures trace output is enabled from this ETM */
+	etm_writel(drvdata, drvdata->ctrl | ETMCR_ETM_EN | etmcr, ETMCR);
+
+	etm_clr_prog(drvdata);
+	CS_LOCK(drvdata->base);
+
+	dev_dbg(drvdata->dev, "cpu: %d enable smp call done\n", drvdata->cpu);
+}
+
+static int etm_trace_id_simple(struct etm_drvdata *drvdata)
+{
+	if (!drvdata->enable)
+		return drvdata->traceid;
+
+	return (etm_readl(drvdata, ETMTRACEIDR) & ETM_TRACEID_MASK);
+}
+
+static int etm_trace_id(struct coresight_device *csdev)
+{
+	struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+	unsigned long flags;
+	int trace_id = -1;
+
+	if (!drvdata->enable)
+		return drvdata->traceid;
+
+	if (clk_prepare_enable(drvdata->clk))
+		goto out;
+
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+
+	CS_UNLOCK(drvdata->base);
+	trace_id = (etm_readl(drvdata, ETMTRACEIDR) & ETM_TRACEID_MASK);
+	CS_LOCK(drvdata->base);
+
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+	clk_disable_unprepare(drvdata->clk);
+out:
+	return trace_id;
+}
+
+static int etm_enable(struct coresight_device *csdev)
+{
+	struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+	int ret;
+
+	ret = clk_prepare_enable(drvdata->clk);
+	if (ret)
+		goto err_clk;
+
+	spin_lock(&drvdata->spinlock);
+
+	/*
+	 * Configure the ETM only if the CPU is online.  If it isn't online
+	 * hw configuration will take place when 'CPU_STARTING' is received
+	 * in @etm_cpu_callback.
+	 */
+	if (cpu_online(drvdata->cpu)) {
+		ret = smp_call_function_single(drvdata->cpu,
+					       etm_enable_hw, drvdata, 1);
+		if (ret)
+			goto err;
+	}
+
+	drvdata->enable = true;
+	drvdata->sticky_enable = true;
+
+	spin_unlock(&drvdata->spinlock);
+
+	dev_info(drvdata->dev, "ETM tracing enabled\n");
+	return 0;
+err:
+	spin_unlock(&drvdata->spinlock);
+	clk_disable_unprepare(drvdata->clk);
+err_clk:
+	return ret;
+}
+
+static void etm_disable_hw(void *info)
+{
+	int i;
+	struct etm_drvdata *drvdata = info;
+
+	CS_UNLOCK(drvdata->base);
+	etm_set_prog(drvdata);
+
+	/* Program trace enable to low by using always false event */
+	etm_writel(drvdata, ETM_HARD_WIRE_RES_A | ETM_EVENT_NOT_A, ETMTEEVR);
+
+	/* Read back sequencer and counters for post trace analysis */
+	drvdata->seq_curr_state = (etm_readl(drvdata, ETMSQR) & ETM_SQR_MASK);
+
+	for (i = 0; i < drvdata->nr_cntr; i++)
+		drvdata->cntr_val[i] = etm_readl(drvdata, ETMCNTVRn(i));
+
+	etm_set_pwrdwn(drvdata);
+	CS_LOCK(drvdata->base);
+
+	dev_dbg(drvdata->dev, "cpu: %d disable smp call done\n", drvdata->cpu);
+}
+
+static void etm_disable(struct coresight_device *csdev)
+{
+	struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	/*
+	 * Taking hotplug lock here protects from clocks getting disabled
+	 * with tracing being left on (crash scenario) if user disable occurs
+	 * after cpu online mask indicates the cpu is offline but before the
+	 * DYING hotplug callback is serviced by the ETM driver.
+	 */
+	get_online_cpus();
+	spin_lock(&drvdata->spinlock);
+
+	/*
+	 * Executing etm_disable_hw on the cpu whose ETM is being disabled
+	 * ensures that register writes occur when cpu is powered.
+	 */
+	smp_call_function_single(drvdata->cpu, etm_disable_hw, drvdata, 1);
+	drvdata->enable = false;
+
+	spin_unlock(&drvdata->spinlock);
+	put_online_cpus();
+
+	clk_disable_unprepare(drvdata->clk);
+
+	dev_info(drvdata->dev, "ETM tracing disabled\n");
+}
+
+static const struct coresight_ops_source etm_source_ops = {
+	.trace_id	= etm_trace_id,
+	.enable		= etm_enable,
+	.disable	= etm_disable,
+};
+
+static const struct coresight_ops etm_cs_ops = {
+	.source_ops	= &etm_source_ops,
+};
+
+static ssize_t nr_addr_cmp_show(struct device *dev,
+				struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->nr_addr_cmp;
+	return sprintf(buf, "%#lx\n", val);
+}
+static DEVICE_ATTR_RO(nr_addr_cmp);
+
+static ssize_t nr_cntr_show(struct device *dev,
+			    struct device_attribute *attr, char *buf)
+{	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->nr_cntr;
+	return sprintf(buf, "%#lx\n", val);
+}
+static DEVICE_ATTR_RO(nr_cntr);
+
+static ssize_t nr_ctxid_cmp_show(struct device *dev,
+				 struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->nr_ctxid_cmp;
+	return sprintf(buf, "%#lx\n", val);
+}
+static DEVICE_ATTR_RO(nr_ctxid_cmp);
+
+static ssize_t etmsr_show(struct device *dev,
+			  struct device_attribute *attr, char *buf)
+{
+	int ret;
+	unsigned long flags, val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = clk_prepare_enable(drvdata->clk);
+	if (ret)
+		return ret;
+
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+	CS_UNLOCK(drvdata->base);
+
+	val = etm_readl(drvdata, ETMSR);
+
+	CS_LOCK(drvdata->base);
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+	clk_disable_unprepare(drvdata->clk);
+
+	return sprintf(buf, "%#lx\n", val);
+}
+static DEVICE_ATTR_RO(etmsr);
+
+static ssize_t reset_store(struct device *dev,
+			   struct device_attribute *attr,
+			   const char *buf, size_t size)
+{
+	int i, ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	if (val) {
+		spin_lock(&drvdata->spinlock);
+		drvdata->mode = ETM_MODE_EXCLUDE;
+		drvdata->ctrl = 0x0;
+		drvdata->trigger_event = ETM_DEFAULT_EVENT_VAL;
+		drvdata->startstop_ctrl = 0x0;
+		drvdata->addr_idx = 0x0;
+		for (i = 0; i < drvdata->nr_addr_cmp; i++) {
+			drvdata->addr_val[i] = 0x0;
+			drvdata->addr_acctype[i] = 0x0;
+			drvdata->addr_type[i] = ETM_ADDR_TYPE_NONE;
+		}
+		drvdata->cntr_idx = 0x0;
+
+		etm_set_default(drvdata);
+		spin_unlock(&drvdata->spinlock);
+	}
+
+	return size;
+}
+static DEVICE_ATTR_WO(reset);
+
+static ssize_t mode_show(struct device *dev,
+			 struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->mode;
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t mode_store(struct device *dev,
+			  struct device_attribute *attr,
+			  const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	spin_lock(&drvdata->spinlock);
+	drvdata->mode = val & ETM_MODE_ALL;
+
+	if (drvdata->mode & ETM_MODE_EXCLUDE)
+		drvdata->enable_ctrl1 |= ETMTECR1_INC_EXC;
+	else
+		drvdata->enable_ctrl1 &= ~ETMTECR1_INC_EXC;
+
+	if (drvdata->mode & ETM_MODE_CYCACC)
+		drvdata->ctrl |= ETMCR_CYC_ACC;
+	else
+		drvdata->ctrl &= ~ETMCR_CYC_ACC;
+
+	if (drvdata->mode & ETM_MODE_STALL) {
+		if (!(drvdata->etmccr & ETMCCR_FIFOFULL)) {
+			dev_warn(drvdata->dev, "stall mode not supported\n");
+			return -EINVAL;
+		}
+		drvdata->ctrl |= ETMCR_STALL_MODE;
+	 } else
+		drvdata->ctrl &= ~ETMCR_STALL_MODE;
+
+	if (drvdata->mode & ETM_MODE_TIMESTAMP) {
+		if (!(drvdata->etmccer & ETMCCER_TIMESTAMP)) {
+			dev_warn(drvdata->dev, "timestamp not supported\n");
+			return -EINVAL;
+		}
+		drvdata->ctrl |= ETMCR_TIMESTAMP_EN;
+	} else
+		drvdata->ctrl &= ~ETMCR_TIMESTAMP_EN;
+
+	if (drvdata->mode & ETM_MODE_CTXID)
+		drvdata->ctrl |= ETMCR_CTXID_SIZE;
+	else
+		drvdata->ctrl &= ~ETMCR_CTXID_SIZE;
+	spin_unlock(&drvdata->spinlock);
+
+	return size;
+}
+static DEVICE_ATTR_RW(mode);
+
+static ssize_t trigger_event_show(struct device *dev,
+				  struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->trigger_event;
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t trigger_event_store(struct device *dev,
+				   struct device_attribute *attr,
+				   const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	drvdata->trigger_event = val & ETM_EVENT_MASK;
+
+	return size;
+}
+static DEVICE_ATTR_RW(trigger_event);
+
+static ssize_t enable_event_show(struct device *dev,
+				 struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->enable_event;
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t enable_event_store(struct device *dev,
+				  struct device_attribute *attr,
+				  const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	drvdata->enable_event = val & ETM_EVENT_MASK;
+
+	return size;
+}
+static DEVICE_ATTR_RW(enable_event);
+
+static ssize_t fifofull_level_show(struct device *dev,
+				   struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->fifofull_level;
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t fifofull_level_store(struct device *dev,
+				    struct device_attribute *attr,
+				    const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	drvdata->fifofull_level = val;
+
+	return size;
+}
+static DEVICE_ATTR_RW(fifofull_level);
+
+static ssize_t addr_idx_show(struct device *dev,
+			     struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->addr_idx;
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t addr_idx_store(struct device *dev,
+			      struct device_attribute *attr,
+			      const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	if (val >= drvdata->nr_addr_cmp)
+		return -EINVAL;
+
+	/*
+	 * Use spinlock to ensure index doesn't change while it gets
+	 * dereferenced multiple times within a spinlock block elsewhere.
+	 */
+	spin_lock(&drvdata->spinlock);
+	drvdata->addr_idx = val;
+	spin_unlock(&drvdata->spinlock);
+
+	return size;
+}
+static DEVICE_ATTR_RW(addr_idx);
+
+static ssize_t addr_single_show(struct device *dev,
+				struct device_attribute *attr, char *buf)
+{
+	u8 idx;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	spin_lock(&drvdata->spinlock);
+	idx = drvdata->addr_idx;
+	if (!(drvdata->addr_type[idx] == ETM_ADDR_TYPE_NONE ||
+	      drvdata->addr_type[idx] == ETM_ADDR_TYPE_SINGLE)) {
+		spin_unlock(&drvdata->spinlock);
+		return -EINVAL;
+	}
+
+	val = drvdata->addr_val[idx];
+	spin_unlock(&drvdata->spinlock);
+
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t addr_single_store(struct device *dev,
+				 struct device_attribute *attr,
+				 const char *buf, size_t size)
+{
+	u8 idx;
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	spin_lock(&drvdata->spinlock);
+	idx = drvdata->addr_idx;
+	if (!(drvdata->addr_type[idx] == ETM_ADDR_TYPE_NONE ||
+	      drvdata->addr_type[idx] == ETM_ADDR_TYPE_SINGLE)) {
+		spin_unlock(&drvdata->spinlock);
+		return -EINVAL;
+	}
+
+	drvdata->addr_val[idx] = val;
+	drvdata->addr_type[idx] = ETM_ADDR_TYPE_SINGLE;
+	spin_unlock(&drvdata->spinlock);
+
+	return size;
+}
+static DEVICE_ATTR_RW(addr_single);
+
+static ssize_t addr_range_show(struct device *dev,
+			       struct device_attribute *attr, char *buf)
+{
+	u8 idx;
+	unsigned long val1, val2;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	spin_lock(&drvdata->spinlock);
+	idx = drvdata->addr_idx;
+	if (idx % 2 != 0) {
+		spin_unlock(&drvdata->spinlock);
+		return -EPERM;
+	}
+	if (!((drvdata->addr_type[idx] == ETM_ADDR_TYPE_NONE &&
+	       drvdata->addr_type[idx + 1] == ETM_ADDR_TYPE_NONE) ||
+	      (drvdata->addr_type[idx] == ETM_ADDR_TYPE_RANGE &&
+	       drvdata->addr_type[idx + 1] == ETM_ADDR_TYPE_RANGE))) {
+		spin_unlock(&drvdata->spinlock);
+		return -EPERM;
+	}
+
+	val1 = drvdata->addr_val[idx];
+	val2 = drvdata->addr_val[idx + 1];
+	spin_unlock(&drvdata->spinlock);
+
+	return sprintf(buf, "%#lx %#lx\n", val1, val2);
+}
+
+static ssize_t addr_range_store(struct device *dev,
+			      struct device_attribute *attr,
+			      const char *buf, size_t size)
+{
+	u8 idx;
+	unsigned long val1, val2;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	if (sscanf(buf, "%lx %lx", &val1, &val2) != 2)
+		return -EINVAL;
+	/* Lower address comparator cannot have a higher address value */
+	if (val1 > val2)
+		return -EINVAL;
+
+	spin_lock(&drvdata->spinlock);
+	idx = drvdata->addr_idx;
+	if (idx % 2 != 0) {
+		spin_unlock(&drvdata->spinlock);
+		return -EPERM;
+	}
+	if (!((drvdata->addr_type[idx] == ETM_ADDR_TYPE_NONE &&
+	       drvdata->addr_type[idx + 1] == ETM_ADDR_TYPE_NONE) ||
+	      (drvdata->addr_type[idx] == ETM_ADDR_TYPE_RANGE &&
+	       drvdata->addr_type[idx + 1] == ETM_ADDR_TYPE_RANGE))) {
+		spin_unlock(&drvdata->spinlock);
+		return -EPERM;
+	}
+
+	drvdata->addr_val[idx] = val1;
+	drvdata->addr_type[idx] = ETM_ADDR_TYPE_RANGE;
+	drvdata->addr_val[idx + 1] = val2;
+	drvdata->addr_type[idx + 1] = ETM_ADDR_TYPE_RANGE;
+	drvdata->enable_ctrl1 |= (1 << (idx/2));
+	spin_unlock(&drvdata->spinlock);
+
+	return size;
+}
+static DEVICE_ATTR_RW(addr_range);
+
+static ssize_t addr_start_show(struct device *dev,
+			       struct device_attribute *attr, char *buf)
+{
+	u8 idx;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	spin_lock(&drvdata->spinlock);
+	idx = drvdata->addr_idx;
+	if (!(drvdata->addr_type[idx] == ETM_ADDR_TYPE_NONE ||
+	      drvdata->addr_type[idx] == ETM_ADDR_TYPE_START)) {
+		spin_unlock(&drvdata->spinlock);
+		return -EPERM;
+	}
+
+	val = drvdata->addr_val[idx];
+	spin_unlock(&drvdata->spinlock);
+
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t addr_start_store(struct device *dev,
+				struct device_attribute *attr,
+				const char *buf, size_t size)
+{
+	u8 idx;
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	spin_lock(&drvdata->spinlock);
+	idx = drvdata->addr_idx;
+	if (!(drvdata->addr_type[idx] == ETM_ADDR_TYPE_NONE ||
+	      drvdata->addr_type[idx] == ETM_ADDR_TYPE_START)) {
+		spin_unlock(&drvdata->spinlock);
+		return -EPERM;
+	}
+
+	drvdata->addr_val[idx] = val;
+	drvdata->addr_type[idx] = ETM_ADDR_TYPE_START;
+	drvdata->startstop_ctrl |= (1 << idx);
+	drvdata->enable_ctrl1 |= BIT(25);
+	spin_unlock(&drvdata->spinlock);
+
+	return size;
+}
+static DEVICE_ATTR_RW(addr_start);
+
+static ssize_t addr_stop_show(struct device *dev,
+			      struct device_attribute *attr, char *buf)
+{
+	u8 idx;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	spin_lock(&drvdata->spinlock);
+	idx = drvdata->addr_idx;
+	if (!(drvdata->addr_type[idx] == ETM_ADDR_TYPE_NONE ||
+	      drvdata->addr_type[idx] == ETM_ADDR_TYPE_STOP)) {
+		spin_unlock(&drvdata->spinlock);
+		return -EPERM;
+	}
+
+	val = drvdata->addr_val[idx];
+	spin_unlock(&drvdata->spinlock);
+
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t addr_stop_store(struct device *dev,
+			       struct device_attribute *attr,
+			       const char *buf, size_t size)
+{
+	u8 idx;
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	spin_lock(&drvdata->spinlock);
+	idx = drvdata->addr_idx;
+	if (!(drvdata->addr_type[idx] == ETM_ADDR_TYPE_NONE ||
+	      drvdata->addr_type[idx] == ETM_ADDR_TYPE_STOP)) {
+		spin_unlock(&drvdata->spinlock);
+		return -EPERM;
+	}
+
+	drvdata->addr_val[idx] = val;
+	drvdata->addr_type[idx] = ETM_ADDR_TYPE_STOP;
+	drvdata->startstop_ctrl |= (1 << (idx + 16));
+	drvdata->enable_ctrl1 |= ETMTECR1_START_STOP;
+	spin_unlock(&drvdata->spinlock);
+
+	return size;
+}
+static DEVICE_ATTR_RW(addr_stop);
+
+static ssize_t addr_acctype_show(struct device *dev,
+				 struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	spin_lock(&drvdata->spinlock);
+	val = drvdata->addr_acctype[drvdata->addr_idx];
+	spin_unlock(&drvdata->spinlock);
+
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t addr_acctype_store(struct device *dev,
+				  struct device_attribute *attr,
+				  const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	spin_lock(&drvdata->spinlock);
+	drvdata->addr_acctype[drvdata->addr_idx] = val;
+	spin_unlock(&drvdata->spinlock);
+
+	return size;
+}
+static DEVICE_ATTR_RW(addr_acctype);
+
+static ssize_t cntr_idx_show(struct device *dev,
+			     struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->cntr_idx;
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t cntr_idx_store(struct device *dev,
+			      struct device_attribute *attr,
+			      const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	if (val >= drvdata->nr_cntr)
+		return -EINVAL;
+	/*
+	 * Use spinlock to ensure index doesn't change while it gets
+	 * dereferenced multiple times within a spinlock block elsewhere.
+	 */
+	spin_lock(&drvdata->spinlock);
+	drvdata->cntr_idx = val;
+	spin_unlock(&drvdata->spinlock);
+
+	return size;
+}
+static DEVICE_ATTR_RW(cntr_idx);
+
+static ssize_t cntr_rld_val_show(struct device *dev,
+				 struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	spin_lock(&drvdata->spinlock);
+	val = drvdata->cntr_rld_val[drvdata->cntr_idx];
+	spin_unlock(&drvdata->spinlock);
+
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t cntr_rld_val_store(struct device *dev,
+				  struct device_attribute *attr,
+				  const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	spin_lock(&drvdata->spinlock);
+	drvdata->cntr_rld_val[drvdata->cntr_idx] = val;
+	spin_unlock(&drvdata->spinlock);
+
+	return size;
+}
+static DEVICE_ATTR_RW(cntr_rld_val);
+
+static ssize_t cntr_event_show(struct device *dev,
+			       struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	spin_lock(&drvdata->spinlock);
+	val = drvdata->cntr_event[drvdata->cntr_idx];
+	spin_unlock(&drvdata->spinlock);
+
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t cntr_event_store(struct device *dev,
+				struct device_attribute *attr,
+				const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	spin_lock(&drvdata->spinlock);
+	drvdata->cntr_event[drvdata->cntr_idx] = val & ETM_EVENT_MASK;
+	spin_unlock(&drvdata->spinlock);
+
+	return size;
+}
+static DEVICE_ATTR_RW(cntr_event);
+
+static ssize_t cntr_rld_event_show(struct device *dev,
+				   struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	spin_lock(&drvdata->spinlock);
+	val = drvdata->cntr_rld_event[drvdata->cntr_idx];
+	spin_unlock(&drvdata->spinlock);
+
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t cntr_rld_event_store(struct device *dev,
+				    struct device_attribute *attr,
+				    const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	spin_lock(&drvdata->spinlock);
+	drvdata->cntr_rld_event[drvdata->cntr_idx] = val & ETM_EVENT_MASK;
+	spin_unlock(&drvdata->spinlock);
+
+	return size;
+}
+static DEVICE_ATTR_RW(cntr_rld_event);
+
+static ssize_t cntr_val_show(struct device *dev,
+			     struct device_attribute *attr, char *buf)
+{
+	int i, ret = 0;
+	u32 val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	if (!drvdata->enable) {
+		spin_lock(&drvdata->spinlock);
+		for (i = 0; i < drvdata->nr_cntr; i++)
+			ret += sprintf(buf, "counter %d: %x\n",
+				       i, drvdata->cntr_val[i]);
+		spin_unlock(&drvdata->spinlock);
+		return ret;
+	}
+
+	for (i = 0; i < drvdata->nr_cntr; i++) {
+		val = etm_readl(drvdata, ETMCNTVRn(i));
+		ret += sprintf(buf, "counter %d: %x\n", i, val);
+	}
+
+	return ret;
+}
+
+static ssize_t cntr_val_store(struct device *dev,
+			      struct device_attribute *attr,
+			      const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	spin_lock(&drvdata->spinlock);
+	drvdata->cntr_val[drvdata->cntr_idx] = val;
+	spin_unlock(&drvdata->spinlock);
+
+	return size;
+}
+static DEVICE_ATTR_RW(cntr_val);
+
+static ssize_t seq_12_event_show(struct device *dev,
+				 struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->seq_12_event;
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t seq_12_event_store(struct device *dev,
+				  struct device_attribute *attr,
+				  const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	drvdata->seq_12_event = val & ETM_EVENT_MASK;
+	return size;
+}
+static DEVICE_ATTR_RW(seq_12_event);
+
+static ssize_t seq_21_event_show(struct device *dev,
+				 struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->seq_21_event;
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t seq_21_event_store(struct device *dev,
+				  struct device_attribute *attr,
+				  const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	drvdata->seq_21_event = val & ETM_EVENT_MASK;
+	return size;
+}
+static DEVICE_ATTR_RW(seq_21_event);
+
+static ssize_t seq_23_event_show(struct device *dev,
+				 struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->seq_23_event;
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t seq_23_event_store(struct device *dev,
+				  struct device_attribute *attr,
+				  const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	drvdata->seq_23_event = val & ETM_EVENT_MASK;
+	return size;
+}
+static DEVICE_ATTR_RW(seq_23_event);
+
+static ssize_t seq_31_event_show(struct device *dev,
+				 struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->seq_31_event;
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t seq_31_event_store(struct device *dev,
+				  struct device_attribute *attr,
+				  const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	drvdata->seq_31_event = val & ETM_EVENT_MASK;
+	return size;
+}
+static DEVICE_ATTR_RW(seq_31_event);
+
+static ssize_t seq_32_event_show(struct device *dev,
+				 struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->seq_32_event;
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t seq_32_event_store(struct device *dev,
+				  struct device_attribute *attr,
+				  const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	drvdata->seq_32_event = val & ETM_EVENT_MASK;
+	return size;
+}
+static DEVICE_ATTR_RW(seq_32_event);
+
+static ssize_t seq_13_event_show(struct device *dev,
+				 struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->seq_13_event;
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t seq_13_event_store(struct device *dev,
+				  struct device_attribute *attr,
+				  const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	drvdata->seq_13_event = val & ETM_EVENT_MASK;
+	return size;
+}
+static DEVICE_ATTR_RW(seq_13_event);
+
+static ssize_t seq_curr_state_show(struct device *dev,
+				   struct device_attribute *attr, char *buf)
+{
+	int ret;
+	unsigned long val, flags;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	if (!drvdata->enable) {
+		val = drvdata->seq_curr_state;
+		goto out;
+	}
+
+	ret = clk_prepare_enable(drvdata->clk);
+	if (ret)
+		return ret;
+
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+
+	CS_UNLOCK(drvdata->base);
+	val = (etm_readl(drvdata, ETMSQR) & ETM_SQR_MASK);
+	CS_LOCK(drvdata->base);
+
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+	clk_disable_unprepare(drvdata->clk);
+out:
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t seq_curr_state_store(struct device *dev,
+				    struct device_attribute *attr,
+				    const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	if (val > ETM_SEQ_STATE_MAX_VAL)
+		return -EINVAL;
+
+	drvdata->seq_curr_state = val;
+
+	return size;
+}
+static DEVICE_ATTR_RW(seq_curr_state);
+
+static ssize_t ctxid_idx_show(struct device *dev,
+			      struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->ctxid_idx;
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t ctxid_idx_store(struct device *dev,
+				struct device_attribute *attr,
+				const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	if (val >= drvdata->nr_ctxid_cmp)
+		return -EINVAL;
+
+	/*
+	 * Use spinlock to ensure index doesn't change while it gets
+	 * dereferenced multiple times within a spinlock block elsewhere.
+	 */
+	spin_lock(&drvdata->spinlock);
+	drvdata->ctxid_idx = val;
+	spin_unlock(&drvdata->spinlock);
+
+	return size;
+}
+static DEVICE_ATTR_RW(ctxid_idx);
+
+static ssize_t ctxid_val_show(struct device *dev,
+			      struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	spin_lock(&drvdata->spinlock);
+	val = drvdata->ctxid_val[drvdata->ctxid_idx];
+	spin_unlock(&drvdata->spinlock);
+
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t ctxid_val_store(struct device *dev,
+			       struct device_attribute *attr,
+			       const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	spin_lock(&drvdata->spinlock);
+	drvdata->ctxid_val[drvdata->ctxid_idx] = val;
+	spin_unlock(&drvdata->spinlock);
+
+	return size;
+}
+static DEVICE_ATTR_RW(ctxid_val);
+
+static ssize_t ctxid_mask_show(struct device *dev,
+			       struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->ctxid_mask;
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t ctxid_mask_store(struct device *dev,
+				struct device_attribute *attr,
+				const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	drvdata->ctxid_mask = val;
+	return size;
+}
+static DEVICE_ATTR_RW(ctxid_mask);
+
+static ssize_t sync_freq_show(struct device *dev,
+			      struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->sync_freq;
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t sync_freq_store(struct device *dev,
+			       struct device_attribute *attr,
+			       const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	drvdata->sync_freq = val & ETM_SYNC_MASK;
+	return size;
+}
+static DEVICE_ATTR_RW(sync_freq);
+
+static ssize_t timestamp_event_show(struct device *dev,
+				    struct device_attribute *attr, char *buf)
+{
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	val = drvdata->timestamp_event;
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t timestamp_event_store(struct device *dev,
+				     struct device_attribute *attr,
+				     const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	drvdata->timestamp_event = val & ETM_EVENT_MASK;
+	return size;
+}
+static DEVICE_ATTR_RW(timestamp_event);
+
+static ssize_t status_show(struct device *dev,
+			   struct device_attribute *attr, char *buf)
+{
+	int ret;
+	unsigned long flags;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = clk_prepare_enable(drvdata->clk);
+	if (ret)
+		return ret;
+
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+
+	CS_UNLOCK(drvdata->base);
+	ret = sprintf(buf,
+		      "ETMCCR: 0x%08x\n"
+		      "ETMCCER: 0x%08x\n"
+		      "ETMSCR: 0x%08x\n"
+		      "ETMIDR: 0x%08x\n"
+		      "ETMCR: 0x%08x\n"
+		      "ETMTRACEIDR: 0x%08x\n"
+		      "Enable event: 0x%08x\n"
+		      "Enable start/stop: 0x%08x\n"
+		      "Enable control: CR1 0x%08x CR2 0x%08x\n"
+		      "CPU affinity: %d\n",
+		      drvdata->etmccr, drvdata->etmccer,
+		      etm_readl(drvdata, ETMSCR), etm_readl(drvdata, ETMIDR),
+		      etm_readl(drvdata, ETMCR), etm_trace_id_simple(drvdata),
+		      etm_readl(drvdata, ETMTEEVR),
+		      etm_readl(drvdata, ETMTSSCR),
+		      etm_readl(drvdata, ETMTECR1),
+		      etm_readl(drvdata, ETMTECR2),
+		      drvdata->cpu);
+	CS_LOCK(drvdata->base);
+
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+	clk_disable_unprepare(drvdata->clk);
+
+	return ret;
+}
+static DEVICE_ATTR_RO(status);
+
+static ssize_t traceid_show(struct device *dev,
+			    struct device_attribute *attr, char *buf)
+{
+	int ret;
+	unsigned long val, flags;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	if (!drvdata->enable) {
+		val = drvdata->traceid;
+		goto out;
+	}
+
+	ret = clk_prepare_enable(drvdata->clk);
+	if (ret)
+		return ret;
+
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+	CS_UNLOCK(drvdata->base);
+
+	val = (etm_readl(drvdata, ETMTRACEIDR) & ETM_TRACEID_MASK);
+
+	CS_LOCK(drvdata->base);
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+	clk_disable_unprepare(drvdata->clk);
+out:
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t traceid_store(struct device *dev,
+			     struct device_attribute *attr,
+			     const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	drvdata->traceid = val & ETM_TRACEID_MASK;
+	return size;
+}
+static DEVICE_ATTR_RW(traceid);
+
+static struct attribute *coresight_etm_attrs[] = {
+	&dev_attr_nr_addr_cmp.attr,
+	&dev_attr_nr_cntr.attr,
+	&dev_attr_nr_ctxid_cmp.attr,
+	&dev_attr_etmsr.attr,
+	&dev_attr_reset.attr,
+	&dev_attr_mode.attr,
+	&dev_attr_trigger_event.attr,
+	&dev_attr_enable_event.attr,
+	&dev_attr_fifofull_level.attr,
+	&dev_attr_addr_idx.attr,
+	&dev_attr_addr_single.attr,
+	&dev_attr_addr_range.attr,
+	&dev_attr_addr_start.attr,
+	&dev_attr_addr_stop.attr,
+	&dev_attr_addr_acctype.attr,
+	&dev_attr_cntr_idx.attr,
+	&dev_attr_cntr_rld_val.attr,
+	&dev_attr_cntr_event.attr,
+	&dev_attr_cntr_rld_event.attr,
+	&dev_attr_cntr_val.attr,
+	&dev_attr_seq_12_event.attr,
+	&dev_attr_seq_21_event.attr,
+	&dev_attr_seq_23_event.attr,
+	&dev_attr_seq_31_event.attr,
+	&dev_attr_seq_32_event.attr,
+	&dev_attr_seq_13_event.attr,
+	&dev_attr_seq_curr_state.attr,
+	&dev_attr_ctxid_idx.attr,
+	&dev_attr_ctxid_val.attr,
+	&dev_attr_ctxid_mask.attr,
+	&dev_attr_sync_freq.attr,
+	&dev_attr_timestamp_event.attr,
+	&dev_attr_status.attr,
+	&dev_attr_traceid.attr,
+	NULL,
+};
+ATTRIBUTE_GROUPS(coresight_etm);
+
+static int etm_cpu_callback(struct notifier_block *nfb, unsigned long action,
+			    void *hcpu)
+{
+	unsigned int cpu = (unsigned long)hcpu;
+
+	if (!etmdrvdata[cpu])
+		goto out;
+
+	switch (action & (~CPU_TASKS_FROZEN)) {
+	case CPU_STARTING:
+		spin_lock(&etmdrvdata[cpu]->spinlock);
+		if (!etmdrvdata[cpu]->os_unlock) {
+			etm_os_unlock(etmdrvdata[cpu]);
+			etmdrvdata[cpu]->os_unlock = true;
+		}
+
+		if (etmdrvdata[cpu]->enable)
+			etm_enable_hw(etmdrvdata[cpu]);
+		spin_unlock(&etmdrvdata[cpu]->spinlock);
+		break;
+
+	case CPU_ONLINE:
+		if (etmdrvdata[cpu]->boot_enable &&
+		    !etmdrvdata[cpu]->sticky_enable)
+			coresight_enable(etmdrvdata[cpu]->csdev);
+		break;
+
+	case CPU_DYING:
+		spin_lock(&etmdrvdata[cpu]->spinlock);
+		if (etmdrvdata[cpu]->enable)
+			etm_disable_hw(etmdrvdata[cpu]);
+		spin_unlock(&etmdrvdata[cpu]->spinlock);
+		break;
+	}
+out:
+	return NOTIFY_OK;
+}
+
+static struct notifier_block etm_cpu_notifier = {
+	.notifier_call = etm_cpu_callback,
+};
+
+static bool etm_arch_supported(u8 arch)
+{
+	switch (arch) {
+	case ETM_ARCH_V3_3:
+		break;
+	case ETM_ARCH_V3_5:
+		break;
+	case PFT_ARCH_V1_0:
+		break;
+	case PFT_ARCH_V1_1:
+		break;
+	default:
+		return false;
+	}
+	return true;
+}
+
+static void etm_init_arch_data(void *info)
+{
+	u32 etmidr;
+	u32 etmccr;
+	struct etm_drvdata *drvdata = info;
+
+	CS_UNLOCK(drvdata->base);
+
+	/* First dummy read */
+	(void)etm_readl(drvdata, ETMPDSR);
+	/* Provide power to ETM: ETMPDCR[3] == 1 */
+	etm_set_pwrup(drvdata);
+	/*
+	 * Clear power down bit since when this bit is set writes to
+	 * certain registers might be ignored.
+	 */
+	etm_clr_pwrdwn(drvdata);
+	/*
+	 * Set prog bit. It will be set from reset but this is included to
+	 * ensure it is set
+	 */
+	etm_set_prog(drvdata);
+
+	/* Find all capabilities */
+	etmidr = etm_readl(drvdata, ETMIDR);
+	drvdata->arch = BMVAL(etmidr, 4, 11);
+	drvdata->port_size = etm_readl(drvdata, ETMCR) & PORT_SIZE_MASK;
+
+	drvdata->etmccer = etm_readl(drvdata, ETMCCER);
+	etmccr = etm_readl(drvdata, ETMCCR);
+	drvdata->etmccr = etmccr;
+	drvdata->nr_addr_cmp = BMVAL(etmccr, 0, 3) * 2;
+	drvdata->nr_cntr = BMVAL(etmccr, 13, 15);
+	drvdata->nr_ext_inp = BMVAL(etmccr, 17, 19);
+	drvdata->nr_ext_out = BMVAL(etmccr, 20, 22);
+	drvdata->nr_ctxid_cmp = BMVAL(etmccr, 24, 25);
+
+	etm_set_pwrdwn(drvdata);
+	etm_clr_pwrup(drvdata);
+	CS_LOCK(drvdata->base);
+}
+
+static void etm_init_default_data(struct etm_drvdata *drvdata)
+{
+	static int etm3x_traceid;
+
+	u32 flags = (1 << 0 | /* instruction execute*/
+		     3 << 3 | /* ARM instruction */
+		     0 << 5 | /* No data value comparison */
+		     0 << 7 | /* No exact mach */
+		     0 << 8 | /* Ignore context ID */
+		     0 << 10); /* Security ignored */
+
+	/*
+	 * Initial configuration only - guarantees sources handled by
+	 * this driver have a unique ID at startup time but not between
+	 * all other types of sources.  For that we lean on the core
+	 * framework.
+	 */
+	drvdata->traceid = etm3x_traceid++;
+	drvdata->ctrl = (ETMCR_CYC_ACC | ETMCR_TIMESTAMP_EN);
+	drvdata->enable_ctrl1 = ETMTECR1_ADDR_COMP_1;
+	if (drvdata->nr_addr_cmp >= 2) {
+		drvdata->addr_val[0] = (u32) _stext;
+		drvdata->addr_val[1] = (u32) _etext;
+		drvdata->addr_acctype[0] = flags;
+		drvdata->addr_acctype[1] = flags;
+		drvdata->addr_type[0] = ETM_ADDR_TYPE_RANGE;
+		drvdata->addr_type[1] = ETM_ADDR_TYPE_RANGE;
+	}
+
+	etm_set_default(drvdata);
+}
+
+static int etm_probe(struct amba_device *adev, const struct amba_id *id)
+{
+	int ret;
+	void __iomem *base;
+	struct device *dev = &adev->dev;
+	struct coresight_platform_data *pdata = NULL;
+	struct etm_drvdata *drvdata;
+	struct resource *res = &adev->res;
+	struct coresight_desc *desc;
+	struct device_node *np = adev->dev.of_node;
+
+	desc = devm_kzalloc(dev, sizeof(*desc), GFP_KERNEL);
+	if (!desc)
+		return -ENOMEM;
+
+	drvdata = devm_kzalloc(dev, sizeof(*drvdata), GFP_KERNEL);
+	if (!drvdata)
+		return -ENOMEM;
+
+	if (np) {
+		pdata = of_get_coresight_platform_data(dev, np);
+		if (IS_ERR(pdata))
+			return PTR_ERR(pdata);
+
+		adev->dev.platform_data = pdata;
+		drvdata->use_cp14 = of_property_read_bool(np, "arm,cp14");
+	}
+
+	drvdata->dev = &adev->dev;
+	dev_set_drvdata(dev, drvdata);
+
+	/* Validity for the resource is already checked by the AMBA core */
+	base = devm_ioremap_resource(dev, res);
+	if (IS_ERR(base))
+		return PTR_ERR(base);
+
+	drvdata->base = base;
+
+	spin_lock_init(&drvdata->spinlock);
+
+	drvdata->clk = adev->pclk;
+	ret = clk_prepare_enable(drvdata->clk);
+	if (ret)
+		return ret;
+
+	drvdata->cpu = pdata ? pdata->cpu : 0;
+
+	get_online_cpus();
+	etmdrvdata[drvdata->cpu] = drvdata;
+
+	if (!smp_call_function_single(drvdata->cpu, etm_os_unlock, drvdata, 1))
+		drvdata->os_unlock = true;
+
+	if (smp_call_function_single(drvdata->cpu,
+				     etm_init_arch_data,  drvdata, 1))
+		dev_err(dev, "ETM arch init failed\n");
+
+	if (!etm_count++)
+		register_hotcpu_notifier(&etm_cpu_notifier);
+
+	put_online_cpus();
+
+	if (etm_arch_supported(drvdata->arch) == false) {
+		ret = -EINVAL;
+		goto err_arch_supported;
+	}
+	etm_init_default_data(drvdata);
+
+	clk_disable_unprepare(drvdata->clk);
+
+	desc->type = CORESIGHT_DEV_TYPE_SOURCE;
+	desc->subtype.source_subtype = CORESIGHT_DEV_SUBTYPE_SOURCE_PROC;
+	desc->ops = &etm_cs_ops;
+	desc->pdata = pdata;
+	desc->dev = dev;
+	desc->groups = coresight_etm_groups;
+	drvdata->csdev = coresight_register(desc);
+	if (IS_ERR(drvdata->csdev)) {
+		ret = PTR_ERR(drvdata->csdev);
+		goto err_arch_supported;
+	}
+
+	dev_info(dev, "ETM initialized\n");
+
+	if (boot_enable) {
+		coresight_enable(drvdata->csdev);
+		drvdata->boot_enable = true;
+	}
+
+	return 0;
+
+err_arch_supported:
+	clk_disable_unprepare(drvdata->clk);
+	if (--etm_count == 0)
+		unregister_hotcpu_notifier(&etm_cpu_notifier);
+	return ret;
+}
+
+static int etm_remove(struct amba_device *adev)
+{
+	struct etm_drvdata *drvdata = amba_get_drvdata(adev);
+
+	coresight_unregister(drvdata->csdev);
+	if (--etm_count == 0)
+		unregister_hotcpu_notifier(&etm_cpu_notifier);
+
+	return 0;
+}
+
+static struct amba_id etm_ids[] = {
+	{	/* ETM 3.3 */
+		.id	= 0x0003b921,
+		.mask	= 0x0003ffff,
+	},
+	{	/* ETM 3.5 */
+		.id	= 0x0003b956,
+		.mask	= 0x0003ffff,
+	},
+	{	/* PTM 1.0 */
+		.id	= 0x0003b950,
+		.mask	= 0x0003ffff,
+	},
+	{	/* PTM 1.1 */
+		.id	= 0x0003b95f,
+		.mask	= 0x0003ffff,
+	},
+	{ 0, 0},
+};
+
+static struct amba_driver etm_driver = {
+	.drv = {
+		.name	= "coresight-etm3x",
+		.owner	= THIS_MODULE,
+	},
+	.probe		= etm_probe,
+	.remove		= etm_remove,
+	.id_table	= etm_ids,
+};
+
+int __init etm_init(void)
+{
+	return amba_driver_register(&etm_driver);
+}
+module_init(etm_init);
+
+void __exit etm_exit(void)
+{
+	amba_driver_unregister(&etm_driver);
+}
+module_exit(etm_exit);
+
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("CoreSight Program Flow Trace driver");
diff --git a/drivers/coresight/coresight-funnel.c b/drivers/coresight/coresight-funnel.c
new file mode 100644
index 000000000000..2108edffe1f4
--- /dev/null
+++ b/drivers/coresight/coresight-funnel.c
@@ -0,0 +1,268 @@
+/* Copyright (c) 2011-2012, The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/device.h>
+#include <linux/err.h>
+#include <linux/fs.h>
+#include <linux/slab.h>
+#include <linux/clk.h>
+#include <linux/coresight.h>
+#include <linux/amba/bus.h>
+
+#include "coresight-priv.h"
+
+#define FUNNEL_FUNCTL		0x000
+#define FUNNEL_PRICTL		0x004
+
+#define FUNNEL_HOLDTIME_MASK	0xf00
+#define FUNNEL_HOLDTIME_SHFT	0x8
+#define FUNNEL_HOLDTIME		(0x7 << FUNNEL_HOLDTIME_SHFT)
+
+/**
+ * struct funnel_drvdata - specifics associated to a funnel component
+ * @base:	memory mapped base address for this component.
+ * @dev:	the device entity associated to this component.
+ * @csdev:	component vitals needed by the framework.
+ * @clk:	the clock this component is associated to.
+ * @priority:	port selection order.
+ */
+struct funnel_drvdata {
+	void __iomem		*base;
+	struct device		*dev;
+	struct coresight_device	*csdev;
+	struct clk		*clk;
+	unsigned long		priority;
+};
+
+static void funnel_enable_hw(struct funnel_drvdata *drvdata, int port)
+{
+	u32 functl;
+
+	CS_UNLOCK(drvdata->base);
+
+	functl = readl_relaxed(drvdata->base + FUNNEL_FUNCTL);
+	functl &= ~FUNNEL_HOLDTIME_MASK;
+	functl |= FUNNEL_HOLDTIME;
+	functl |= (1 << port);
+	writel_relaxed(functl, drvdata->base + FUNNEL_FUNCTL);
+	writel_relaxed(drvdata->priority, drvdata->base + FUNNEL_PRICTL);
+
+	CS_LOCK(drvdata->base);
+}
+
+static int funnel_enable(struct coresight_device *csdev, int inport,
+			 int outport)
+{
+	struct funnel_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+	int ret;
+
+	ret = clk_prepare_enable(drvdata->clk);
+	if (ret)
+		return ret;
+
+	funnel_enable_hw(drvdata, inport);
+
+	dev_info(drvdata->dev, "FUNNEL inport %d enabled\n", inport);
+	return 0;
+}
+
+static void funnel_disable_hw(struct funnel_drvdata *drvdata, int inport)
+{
+	u32 functl;
+
+	CS_UNLOCK(drvdata->base);
+
+	functl = readl_relaxed(drvdata->base + FUNNEL_FUNCTL);
+	functl &= ~(1 << inport);
+	writel_relaxed(functl, drvdata->base + FUNNEL_FUNCTL);
+
+	CS_LOCK(drvdata->base);
+}
+
+static void funnel_disable(struct coresight_device *csdev, int inport,
+			   int outport)
+{
+	struct funnel_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	funnel_disable_hw(drvdata, inport);
+
+	clk_disable_unprepare(drvdata->clk);
+
+	dev_info(drvdata->dev, "FUNNEL inport %d disabled\n", inport);
+}
+
+static const struct coresight_ops_link funnel_link_ops = {
+	.enable		= funnel_enable,
+	.disable	= funnel_disable,
+};
+
+static const struct coresight_ops funnel_cs_ops = {
+	.link_ops	= &funnel_link_ops,
+};
+
+static ssize_t priority_show(struct device *dev,
+			     struct device_attribute *attr, char *buf)
+{
+	struct funnel_drvdata *drvdata = dev_get_drvdata(dev->parent);
+	unsigned long val = drvdata->priority;
+
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t priority_store(struct device *dev,
+			      struct device_attribute *attr,
+			      const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct funnel_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	drvdata->priority = val;
+	return size;
+}
+static DEVICE_ATTR_RW(priority);
+
+static u32 get_funnel_ctrl_hw(struct funnel_drvdata *drvdata)
+{
+	u32 functl;
+
+	CS_UNLOCK(drvdata->base);
+	functl = readl_relaxed(drvdata->base + FUNNEL_FUNCTL);
+	CS_LOCK(drvdata->base);
+
+	return functl;
+}
+
+static ssize_t funnel_ctrl_show(struct device *dev,
+			     struct device_attribute *attr, char *buf)
+{
+	int ret;
+	u32 val;
+	struct funnel_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = clk_prepare_enable(drvdata->clk);
+	if (ret)
+		return ret;
+
+	val = get_funnel_ctrl_hw(drvdata);
+	clk_disable_unprepare(drvdata->clk);
+
+	return sprintf(buf, "%#x\n", val);
+}
+static DEVICE_ATTR_RO(funnel_ctrl);
+
+static struct attribute *coresight_funnel_attrs[] = {
+	&dev_attr_funnel_ctrl.attr,
+	&dev_attr_priority.attr,
+	NULL,
+};
+ATTRIBUTE_GROUPS(coresight_funnel);
+
+static int funnel_probe(struct amba_device *adev, const struct amba_id *id)
+{
+	void __iomem *base;
+	struct device *dev = &adev->dev;
+	struct coresight_platform_data *pdata = NULL;
+	struct funnel_drvdata *drvdata;
+	struct resource *res = &adev->res;
+	struct coresight_desc *desc;
+	struct device_node *np = adev->dev.of_node;
+
+	if (np) {
+		pdata = of_get_coresight_platform_data(dev, np);
+		if (IS_ERR(pdata))
+			return PTR_ERR(pdata);
+		adev->dev.platform_data = pdata;
+	}
+
+	drvdata = devm_kzalloc(dev, sizeof(*drvdata), GFP_KERNEL);
+	if (!drvdata)
+		return -ENOMEM;
+
+	drvdata->dev = &adev->dev;
+	dev_set_drvdata(dev, drvdata);
+
+	/* Validity for the resource is already checked by the AMBA core */
+	base = devm_ioremap_resource(dev, res);
+	if (IS_ERR(base))
+		return PTR_ERR(base);
+
+	drvdata->base = base;
+
+	drvdata->clk = adev->pclk;
+
+	desc = devm_kzalloc(dev, sizeof(*desc), GFP_KERNEL);
+	if (!desc)
+		return -ENOMEM;
+
+	desc->type = CORESIGHT_DEV_TYPE_LINK;
+	desc->subtype.link_subtype = CORESIGHT_DEV_SUBTYPE_LINK_MERG;
+	desc->ops = &funnel_cs_ops;
+	desc->pdata = pdata;
+	desc->dev = dev;
+	desc->groups = coresight_funnel_groups;
+	drvdata->csdev = coresight_register(desc);
+	if (IS_ERR(drvdata->csdev))
+		return PTR_ERR(drvdata->csdev);
+
+	dev_info(dev, "FUNNEL initialized\n");
+	return 0;
+}
+
+static int funnel_remove(struct amba_device *adev)
+{
+	struct funnel_drvdata *drvdata = amba_get_drvdata(adev);
+
+	coresight_unregister(drvdata->csdev);
+	return 0;
+}
+
+static struct amba_id funnel_ids[] = {
+	{
+		.id     = 0x0003b908,
+		.mask   = 0x0003ffff,
+	},
+	{ 0, 0},
+};
+
+static struct amba_driver funnel_driver = {
+	.drv = {
+		.name	= "coresight-funnel",
+		.owner	= THIS_MODULE,
+	},
+	.probe		= funnel_probe,
+	.remove		= funnel_remove,
+	.id_table	= funnel_ids,
+};
+
+static int __init funnel_init(void)
+{
+	return amba_driver_register(&funnel_driver);
+}
+module_init(funnel_init);
+
+static void __exit funnel_exit(void)
+{
+	amba_driver_unregister(&funnel_driver);
+}
+module_exit(funnel_exit);
+
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("CoreSight Funnel driver");
diff --git a/drivers/coresight/coresight-priv.h b/drivers/coresight/coresight-priv.h
new file mode 100644
index 000000000000..7b3372fca4f6
--- /dev/null
+++ b/drivers/coresight/coresight-priv.h
@@ -0,0 +1,63 @@
+/* Copyright (c) 2011-2012, The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#ifndef _CORESIGHT_PRIV_H
+#define _CORESIGHT_PRIV_H
+
+#include <linux/bitops.h>
+#include <linux/io.h>
+#include <linux/coresight.h>
+
+/*
+ * Coresight management registers (0xf00-0xfcc)
+ * 0xfa0 - 0xfa4: Management	registers in PFTv1.0
+ *		  Trace		registers in PFTv1.1
+ */
+#define CORESIGHT_ITCTRL	0xf00
+#define CORESIGHT_CLAIMSET	0xfa0
+#define CORESIGHT_CLAIMCLR	0xfa4
+#define CORESIGHT_LAR		0xfb0
+#define CORESIGHT_LSR		0xfb4
+#define CORESIGHT_AUTHSTATUS	0xfb8
+#define CORESIGHT_DEVID		0xfc8
+#define CORESIGHT_DEVTYPE	0xfcc
+
+#define TIMEOUT_US		100
+#define BMVAL(val, lsb, msb)	((val & GENMASK(msb, lsb)) >> lsb)
+
+static inline void CS_LOCK(void __iomem *addr)
+{
+	do {
+		/* Wait for things to settle */
+		mb();
+		writel_relaxed(0x0, addr + CORESIGHT_LAR);
+	} while (0);
+}
+
+static inline void CS_UNLOCK(void __iomem *addr)
+{
+	do {
+		writel_relaxed(CORESIGHT_UNLOCK, addr + CORESIGHT_LAR);
+		/* Make sure everyone has seen this */
+		mb();
+	} while (0);
+}
+
+#ifdef CONFIG_CORESIGHT_SOURCE_ETM3X
+extern int etm_readl_cp14(u32 off, unsigned int *val);
+extern int etm_writel_cp14(u32 off, u32 val);
+#else
+static inline int etm_readl_cp14(u32 off, unsigned int *val) { return 0; }
+static inline int etm_writel_cp14(u32 val, u32 off) { return 0; }
+#endif
+
+#endif
diff --git a/drivers/coresight/coresight-replicator.c b/drivers/coresight/coresight-replicator.c
new file mode 100644
index 000000000000..a2dfcf903551
--- /dev/null
+++ b/drivers/coresight/coresight-replicator.c
@@ -0,0 +1,137 @@
+/* Copyright (c) 2011-2012, The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/device.h>
+#include <linux/platform_device.h>
+#include <linux/io.h>
+#include <linux/err.h>
+#include <linux/slab.h>
+#include <linux/clk.h>
+#include <linux/of.h>
+#include <linux/coresight.h>
+
+#include "coresight-priv.h"
+
+/**
+ * struct replicator_drvdata - specifics associated to a replicator component
+ * @dev:	the device entity associated with this component
+ * @csdev:	component vitals needed by the framework
+ */
+struct replicator_drvdata {
+	struct device		*dev;
+	struct coresight_device	*csdev;
+};
+
+static int replicator_enable(struct coresight_device *csdev, int inport,
+			     int outport)
+{
+	struct replicator_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	dev_info(drvdata->dev, "REPLICATOR enabled\n");
+	return 0;
+}
+
+static void replicator_disable(struct coresight_device *csdev, int inport,
+			       int outport)
+{
+	struct replicator_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	dev_info(drvdata->dev, "REPLICATOR disabled\n");
+}
+
+static const struct coresight_ops_link replicator_link_ops = {
+	.enable		= replicator_enable,
+	.disable	= replicator_disable,
+};
+
+static const struct coresight_ops replicator_cs_ops = {
+	.link_ops	= &replicator_link_ops,
+};
+
+static int replicator_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct coresight_platform_data *pdata = NULL;
+	struct replicator_drvdata *drvdata;
+	struct coresight_desc *desc;
+	struct device_node *np = pdev->dev.of_node;
+
+	if (np) {
+		pdata = of_get_coresight_platform_data(dev, np);
+		if (IS_ERR(pdata))
+			return PTR_ERR(pdata);
+		pdev->dev.platform_data = pdata;
+	}
+
+	drvdata = devm_kzalloc(dev, sizeof(*drvdata), GFP_KERNEL);
+	if (!drvdata)
+		return -ENOMEM;
+
+	drvdata->dev = &pdev->dev;
+	platform_set_drvdata(pdev, drvdata);
+
+	desc = devm_kzalloc(dev, sizeof(*desc), GFP_KERNEL);
+	if (!desc)
+		return -ENOMEM;
+
+	desc->type = CORESIGHT_DEV_TYPE_LINK;
+	desc->subtype.sink_subtype = CORESIGHT_DEV_SUBTYPE_LINK_SPLIT;
+	desc->ops = &replicator_cs_ops;
+	desc->pdata = pdev->dev.platform_data;
+	desc->dev = &pdev->dev;
+	drvdata->csdev = coresight_register(desc);
+	if (IS_ERR(drvdata->csdev))
+		return PTR_ERR(drvdata->csdev);
+
+	dev_info(dev, "REPLICATOR initialized\n");
+	return 0;
+}
+
+static int replicator_remove(struct platform_device *pdev)
+{
+	struct replicator_drvdata *drvdata = platform_get_drvdata(pdev);
+
+	coresight_unregister(drvdata->csdev);
+	return 0;
+}
+
+static struct of_device_id replicator_match[] = {
+	{.compatible = "arm,coresight-replicator"},
+	{}
+};
+
+static struct platform_driver replicator_driver = {
+	.probe          = replicator_probe,
+	.remove         = replicator_remove,
+	.driver         = {
+		.name   = "coresight-replicator",
+		.of_match_table = replicator_match,
+	},
+};
+
+static int __init replicator_init(void)
+{
+	return platform_driver_register(&replicator_driver);
+}
+module_init(replicator_init);
+
+static void __exit replicator_exit(void)
+{
+	platform_driver_unregister(&replicator_driver);
+}
+module_exit(replicator_exit);
+
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("CoreSight Replicator driver");
diff --git a/drivers/coresight/coresight-tmc.c b/drivers/coresight/coresight-tmc.c
new file mode 100644
index 000000000000..ce2c293f1707
--- /dev/null
+++ b/drivers/coresight/coresight-tmc.c
@@ -0,0 +1,776 @@
+/* Copyright (c) 2012, The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/device.h>
+#include <linux/io.h>
+#include <linux/err.h>
+#include <linux/fs.h>
+#include <linux/miscdevice.h>
+#include <linux/uaccess.h>
+#include <linux/slab.h>
+#include <linux/dma-mapping.h>
+#include <linux/spinlock.h>
+#include <linux/clk.h>
+#include <linux/of.h>
+#include <linux/coresight.h>
+#include <linux/amba/bus.h>
+
+#include "coresight-priv.h"
+
+#define TMC_RSZ			0x004
+#define TMC_STS			0x00c
+#define TMC_RRD			0x010
+#define TMC_RRP			0x014
+#define TMC_RWP			0x018
+#define TMC_TRG			0x01c
+#define TMC_CTL			0x020
+#define TMC_RWD			0x024
+#define TMC_MODE		0x028
+#define TMC_LBUFLEVEL		0x02c
+#define TMC_CBUFLEVEL		0x030
+#define TMC_BUFWM		0x034
+#define TMC_RRPHI		0x038
+#define TMC_RWPHI		0x03c
+#define TMC_AXICTL		0x110
+#define TMC_DBALO		0x118
+#define TMC_DBAHI		0x11c
+#define TMC_FFSR		0x300
+#define TMC_FFCR		0x304
+#define TMC_PSCR		0x308
+#define TMC_ITMISCOP0		0xee0
+#define TMC_ITTRFLIN		0xee8
+#define TMC_ITATBDATA0		0xeec
+#define TMC_ITATBCTR2		0xef0
+#define TMC_ITATBCTR1		0xef4
+#define TMC_ITATBCTR0		0xef8
+
+/* register description */
+/* TMC_CTL - 0x020 */
+#define TMC_CTL_CAPT_EN		BIT(0)
+/* TMC_STS - 0x00C */
+#define TMC_STS_TRIGGERED	BIT(1)
+/* TMC_AXICTL - 0x110 */
+#define TMC_AXICTL_PROT_CTL_B0	BIT(0)
+#define TMC_AXICTL_PROT_CTL_B1	BIT(1)
+#define TMC_AXICTL_SCT_GAT_MODE	BIT(7)
+#define TMC_AXICTL_WR_BURST_LEN 0xF00
+/* TMC_FFCR - 0x304 */
+#define TMC_FFCR_EN_FMT		BIT(0)
+#define TMC_FFCR_EN_TI		BIT(1)
+#define TMC_FFCR_FON_FLIN	BIT(4)
+#define TMC_FFCR_FON_TRIG_EVT	BIT(5)
+#define TMC_FFCR_FLUSHMAN	BIT(6)
+#define TMC_FFCR_TRIGON_TRIGIN	BIT(8)
+#define TMC_FFCR_STOP_ON_FLUSH	BIT(12)
+
+#define TMC_STS_TRIGGERED_BIT	2
+#define TMC_FFCR_FLUSHMAN_BIT	6
+
+enum tmc_config_type {
+	TMC_CONFIG_TYPE_ETB,
+	TMC_CONFIG_TYPE_ETR,
+	TMC_CONFIG_TYPE_ETF,
+};
+
+enum tmc_mode {
+	TMC_MODE_CIRCULAR_BUFFER,
+	TMC_MODE_SOFTWARE_FIFO,
+	TMC_MODE_HARDWARE_FIFO,
+};
+
+enum tmc_mem_intf_width {
+	TMC_MEM_INTF_WIDTH_32BITS	= 0x2,
+	TMC_MEM_INTF_WIDTH_64BITS	= 0x3,
+	TMC_MEM_INTF_WIDTH_128BITS	= 0x4,
+	TMC_MEM_INTF_WIDTH_256BITS	= 0x5,
+};
+
+/**
+ * struct tmc_drvdata - specifics associated to an TMC component
+ * @base:	memory mapped base address for this component.
+ * @dev:	the device entity associated to this component.
+ * @csdev:	component vitals needed by the framework.
+ * @miscdev:	specifics to handle "/dev/xyz.tmc" entry.
+ * @clk:	the clock this component is associated to.
+ * @spinlock:	only one at a time pls.
+ * @read_count:	manages preparation of buffer for reading.
+ * @buf:	area of memory where trace data get sent.
+ * @paddr:	DMA start location in RAM.
+ * @vaddr:	virtual representation of @paddr.
+ * @size:	@buf size.
+ * @enable:	this TMC is being used.
+ * @config_type: TMC variant, must be of type @tmc_config_type.
+ * @trigger_cntr: amount of words to store after a trigger.
+ */
+struct tmc_drvdata {
+	void __iomem		*base;
+	struct device		*dev;
+	struct coresight_device	*csdev;
+	struct miscdevice	miscdev;
+	struct clk		*clk;
+	spinlock_t		spinlock;
+	int			read_count;
+	bool			reading;
+	char			*buf;
+	dma_addr_t		paddr;
+	void __iomem		*vaddr;
+	u32			size;
+	bool			enable;
+	enum tmc_config_type	config_type;
+	u32			trigger_cntr;
+};
+
+static void tmc_wait_for_ready(struct tmc_drvdata *drvdata)
+{
+	/* Ensure formatter, unformatter and hardware fifo are empty */
+	if (coresight_timeout(drvdata->base,
+			      TMC_STS, TMC_STS_TRIGGERED_BIT, 1)) {
+		dev_err(drvdata->dev,
+			"timeout observed when probing at offset %#x\n",
+			TMC_STS);
+	}
+}
+
+static void tmc_flush_and_stop(struct tmc_drvdata *drvdata)
+{
+	u32 ffcr;
+
+	ffcr = readl_relaxed(drvdata->base + TMC_FFCR);
+	ffcr |= TMC_FFCR_STOP_ON_FLUSH;
+	writel_relaxed(ffcr, drvdata->base + TMC_FFCR);
+	ffcr |= TMC_FFCR_FLUSHMAN;
+	writel_relaxed(ffcr, drvdata->base + TMC_FFCR);
+	/* Ensure flush completes */
+	if (coresight_timeout(drvdata->base,
+			      TMC_FFCR, TMC_FFCR_FLUSHMAN_BIT, 0)) {
+		dev_err(drvdata->dev,
+			"timeout observed when probing at offset %#x\n",
+			TMC_FFCR);
+	}
+
+	tmc_wait_for_ready(drvdata);
+}
+
+static void tmc_enable_hw(struct tmc_drvdata *drvdata)
+{
+	writel_relaxed(TMC_CTL_CAPT_EN, drvdata->base + TMC_CTL);
+}
+
+static void tmc_disable_hw(struct tmc_drvdata *drvdata)
+{
+	writel_relaxed(0x0, drvdata->base + TMC_CTL);
+}
+
+static void tmc_etb_enable_hw(struct tmc_drvdata *drvdata)
+{
+	/* Zero out the memory to help with debug */
+	memset(drvdata->buf, 0, drvdata->size);
+
+	CS_UNLOCK(drvdata->base);
+
+	writel_relaxed(TMC_MODE_CIRCULAR_BUFFER, drvdata->base + TMC_MODE);
+	writel_relaxed(TMC_FFCR_EN_FMT | TMC_FFCR_EN_TI |
+		       TMC_FFCR_FON_FLIN | TMC_FFCR_FON_TRIG_EVT |
+		       TMC_FFCR_TRIGON_TRIGIN,
+		       drvdata->base + TMC_FFCR);
+
+	writel_relaxed(drvdata->trigger_cntr, drvdata->base + TMC_TRG);
+	tmc_enable_hw(drvdata);
+
+	CS_LOCK(drvdata->base);
+}
+
+static void tmc_etr_enable_hw(struct tmc_drvdata *drvdata)
+{
+	u32 axictl;
+
+	/* Zero out the memory to help with debug */
+	memset(drvdata->vaddr, 0, drvdata->size);
+
+	CS_UNLOCK(drvdata->base);
+
+	writel_relaxed(drvdata->size / 4, drvdata->base + TMC_RSZ);
+	writel_relaxed(TMC_MODE_CIRCULAR_BUFFER, drvdata->base + TMC_MODE);
+
+	axictl = readl_relaxed(drvdata->base + TMC_AXICTL);
+	axictl |= TMC_AXICTL_WR_BURST_LEN;
+	writel_relaxed(axictl, drvdata->base + TMC_AXICTL);
+	axictl &= ~TMC_AXICTL_SCT_GAT_MODE;
+	writel_relaxed(axictl, drvdata->base + TMC_AXICTL);
+	axictl = (axictl &
+		  ~(TMC_AXICTL_PROT_CTL_B0 | TMC_AXICTL_PROT_CTL_B1)) |
+		  TMC_AXICTL_PROT_CTL_B1;
+	writel_relaxed(axictl, drvdata->base + TMC_AXICTL);
+
+	writel_relaxed(drvdata->paddr, drvdata->base + TMC_DBALO);
+	writel_relaxed(0x0, drvdata->base + TMC_DBAHI);
+	writel_relaxed(TMC_FFCR_EN_FMT | TMC_FFCR_EN_TI |
+		       TMC_FFCR_FON_FLIN | TMC_FFCR_FON_TRIG_EVT |
+		       TMC_FFCR_TRIGON_TRIGIN,
+		       drvdata->base + TMC_FFCR);
+	writel_relaxed(drvdata->trigger_cntr, drvdata->base + TMC_TRG);
+	tmc_enable_hw(drvdata);
+
+	CS_LOCK(drvdata->base);
+}
+
+static void tmc_etf_enable_hw(struct tmc_drvdata *drvdata)
+{
+	CS_UNLOCK(drvdata->base);
+
+	writel_relaxed(TMC_MODE_HARDWARE_FIFO, drvdata->base + TMC_MODE);
+	writel_relaxed(TMC_FFCR_EN_FMT | TMC_FFCR_EN_TI,
+		       drvdata->base + TMC_FFCR);
+	writel_relaxed(0x0, drvdata->base + TMC_BUFWM);
+	tmc_enable_hw(drvdata);
+
+	CS_LOCK(drvdata->base);
+}
+
+static int tmc_enable(struct tmc_drvdata *drvdata, enum tmc_mode mode)
+{
+	int ret;
+	unsigned long flags;
+
+	ret = clk_prepare_enable(drvdata->clk);
+	if (ret)
+		return ret;
+
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+	if (drvdata->reading) {
+		spin_unlock_irqrestore(&drvdata->spinlock, flags);
+		clk_disable_unprepare(drvdata->clk);
+		return -EBUSY;
+	}
+
+	if (drvdata->config_type == TMC_CONFIG_TYPE_ETB) {
+		tmc_etb_enable_hw(drvdata);
+	} else if (drvdata->config_type == TMC_CONFIG_TYPE_ETR) {
+		tmc_etr_enable_hw(drvdata);
+	} else {
+		if (mode == TMC_MODE_CIRCULAR_BUFFER)
+			tmc_etb_enable_hw(drvdata);
+		else
+			tmc_etf_enable_hw(drvdata);
+	}
+	drvdata->enable = true;
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+
+	dev_info(drvdata->dev, "TMC enabled\n");
+	return 0;
+}
+
+static int tmc_enable_sink(struct coresight_device *csdev)
+{
+	struct tmc_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	return tmc_enable(drvdata, TMC_MODE_CIRCULAR_BUFFER);
+}
+
+static int tmc_enable_link(struct coresight_device *csdev, int inport,
+			   int outport)
+{
+	struct tmc_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	return tmc_enable(drvdata, TMC_MODE_HARDWARE_FIFO);
+}
+
+static void tmc_etb_dump_hw(struct tmc_drvdata *drvdata)
+{
+	enum tmc_mem_intf_width memwidth;
+	u8 memwords;
+	char *bufp;
+	u32 read_data;
+	int i;
+
+	memwidth = BMVAL(readl_relaxed(drvdata->base + CORESIGHT_DEVID), 8, 10);
+	if (memwidth == TMC_MEM_INTF_WIDTH_32BITS)
+		memwords = 1;
+	else if (memwidth == TMC_MEM_INTF_WIDTH_64BITS)
+		memwords = 2;
+	else if (memwidth == TMC_MEM_INTF_WIDTH_128BITS)
+		memwords = 4;
+	else
+		memwords = 8;
+
+	bufp = drvdata->buf;
+	while (1) {
+		for (i = 0; i < memwords; i++) {
+			read_data = readl_relaxed(drvdata->base + TMC_RRD);
+			if (read_data == 0xFFFFFFFF)
+				return;
+			memcpy(bufp, &read_data, 4);
+			bufp += 4;
+		}
+	}
+}
+
+static void tmc_etb_disable_hw(struct tmc_drvdata *drvdata)
+{
+	CS_UNLOCK(drvdata->base);
+
+	tmc_flush_and_stop(drvdata);
+	tmc_etb_dump_hw(drvdata);
+	tmc_disable_hw(drvdata);
+
+	CS_LOCK(drvdata->base);
+}
+
+static void tmc_etr_dump_hw(struct tmc_drvdata *drvdata)
+{
+	u32 rwp, val;
+
+	rwp = readl_relaxed(drvdata->base + TMC_RWP);
+	val = readl_relaxed(drvdata->base + TMC_STS);
+
+	/* How much memory do we still have */
+	if (val & BIT(0))
+		drvdata->buf = drvdata->vaddr + rwp - drvdata->paddr;
+	else
+		drvdata->buf = drvdata->vaddr;
+}
+
+static void tmc_etr_disable_hw(struct tmc_drvdata *drvdata)
+{
+	CS_UNLOCK(drvdata->base);
+
+	tmc_flush_and_stop(drvdata);
+	tmc_etr_dump_hw(drvdata);
+	tmc_disable_hw(drvdata);
+
+	CS_LOCK(drvdata->base);
+}
+
+static void tmc_etf_disable_hw(struct tmc_drvdata *drvdata)
+{
+	CS_UNLOCK(drvdata->base);
+
+	tmc_flush_and_stop(drvdata);
+	tmc_disable_hw(drvdata);
+
+	CS_LOCK(drvdata->base);
+}
+
+static void tmc_disable(struct tmc_drvdata *drvdata, enum tmc_mode mode)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+	if (drvdata->reading)
+		goto out;
+
+	if (drvdata->config_type == TMC_CONFIG_TYPE_ETB) {
+		tmc_etb_disable_hw(drvdata);
+	} else if (drvdata->config_type == TMC_CONFIG_TYPE_ETR) {
+		tmc_etr_disable_hw(drvdata);
+	} else {
+		if (mode == TMC_MODE_CIRCULAR_BUFFER)
+			tmc_etb_disable_hw(drvdata);
+		else
+			tmc_etf_disable_hw(drvdata);
+	}
+out:
+	drvdata->enable = false;
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+
+	clk_disable_unprepare(drvdata->clk);
+
+	dev_info(drvdata->dev, "TMC disabled\n");
+}
+
+static void tmc_disable_sink(struct coresight_device *csdev)
+{
+	struct tmc_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	tmc_disable(drvdata, TMC_MODE_CIRCULAR_BUFFER);
+}
+
+static void tmc_disable_link(struct coresight_device *csdev, int inport,
+			     int outport)
+{
+	struct tmc_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	tmc_disable(drvdata, TMC_MODE_HARDWARE_FIFO);
+}
+
+static const struct coresight_ops_sink tmc_sink_ops = {
+	.enable		= tmc_enable_sink,
+	.disable	= tmc_disable_sink,
+};
+
+static const struct coresight_ops_link tmc_link_ops = {
+	.enable		= tmc_enable_link,
+	.disable	= tmc_disable_link,
+};
+
+static const struct coresight_ops tmc_etb_cs_ops = {
+	.sink_ops	= &tmc_sink_ops,
+};
+
+static const struct coresight_ops tmc_etr_cs_ops = {
+	.sink_ops	= &tmc_sink_ops,
+};
+
+static const struct coresight_ops tmc_etf_cs_ops = {
+	.sink_ops	= &tmc_sink_ops,
+	.link_ops	= &tmc_link_ops,
+};
+
+static int tmc_read_prepare(struct tmc_drvdata *drvdata)
+{
+	int ret;
+	unsigned long flags;
+	enum tmc_mode mode;
+
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+	if (!drvdata->enable)
+		goto out;
+
+	if (drvdata->config_type == TMC_CONFIG_TYPE_ETB) {
+		tmc_etb_disable_hw(drvdata);
+	} else if (drvdata->config_type == TMC_CONFIG_TYPE_ETR) {
+		tmc_etr_disable_hw(drvdata);
+	} else {
+		mode = readl_relaxed(drvdata->base + TMC_MODE);
+		if (mode == TMC_MODE_CIRCULAR_BUFFER) {
+			tmc_etb_disable_hw(drvdata);
+		} else {
+			ret = -ENODEV;
+			goto err;
+		}
+	}
+out:
+	drvdata->reading = true;
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+
+	dev_info(drvdata->dev, "TMC read start\n");
+	return 0;
+err:
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+	return ret;
+}
+
+static void tmc_read_unprepare(struct tmc_drvdata *drvdata)
+{
+	unsigned long flags;
+	enum tmc_mode mode;
+
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+	if (!drvdata->enable)
+		goto out;
+
+	if (drvdata->config_type == TMC_CONFIG_TYPE_ETB) {
+		tmc_etb_enable_hw(drvdata);
+	} else if (drvdata->config_type == TMC_CONFIG_TYPE_ETR) {
+		tmc_etr_enable_hw(drvdata);
+	} else {
+		mode = readl_relaxed(drvdata->base + TMC_MODE);
+		if (mode == TMC_MODE_CIRCULAR_BUFFER)
+			tmc_etb_enable_hw(drvdata);
+	}
+out:
+	drvdata->reading = false;
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+
+	dev_info(drvdata->dev, "TMC read end\n");
+}
+
+static int tmc_open(struct inode *inode, struct file *file)
+{
+	struct tmc_drvdata *drvdata = container_of(file->private_data,
+						   struct tmc_drvdata, miscdev);
+	int ret = 0;
+
+	if (drvdata->read_count++)
+		goto out;
+
+	ret = tmc_read_prepare(drvdata);
+	if (ret)
+		return ret;
+out:
+	nonseekable_open(inode, file);
+
+	dev_dbg(drvdata->dev, "%s: successfully opened\n", __func__);
+	return 0;
+}
+
+static ssize_t tmc_read(struct file *file, char __user *data, size_t len,
+			loff_t *ppos)
+{
+	struct tmc_drvdata *drvdata = container_of(file->private_data,
+						   struct tmc_drvdata, miscdev);
+	char *bufp = drvdata->buf + *ppos;
+
+	if (*ppos + len > drvdata->size)
+		len = drvdata->size - *ppos;
+
+	if (drvdata->config_type == TMC_CONFIG_TYPE_ETR) {
+		if (bufp == (char *)(drvdata->vaddr + drvdata->size))
+			bufp = drvdata->vaddr;
+		else if (bufp > (char *)(drvdata->vaddr + drvdata->size))
+			bufp -= drvdata->size;
+		if ((bufp + len) > (char *)(drvdata->vaddr + drvdata->size))
+			len = (char *)(drvdata->vaddr + drvdata->size) - bufp;
+	}
+
+	if (copy_to_user(data, bufp, len)) {
+		dev_dbg(drvdata->dev, "%s: copy_to_user failed\n", __func__);
+		return -EFAULT;
+	}
+
+	*ppos += len;
+
+	dev_dbg(drvdata->dev, "%s: %d bytes copied, %d bytes left\n",
+		__func__, len, (int) (drvdata->size - *ppos));
+	return len;
+}
+
+static int tmc_release(struct inode *inode, struct file *file)
+{
+	struct tmc_drvdata *drvdata = container_of(file->private_data,
+						   struct tmc_drvdata, miscdev);
+
+	if (--drvdata->read_count) {
+		if (drvdata->read_count < 0) {
+			dev_err(drvdata->dev, "mismatched close\n");
+			drvdata->read_count = 0;
+		}
+		goto out;
+	}
+
+	tmc_read_unprepare(drvdata);
+out:
+	dev_dbg(drvdata->dev, "%s: released\n", __func__);
+	return 0;
+}
+
+static const struct file_operations tmc_fops = {
+	.owner		= THIS_MODULE,
+	.open		= tmc_open,
+	.read		= tmc_read,
+	.release	= tmc_release,
+	.llseek		= no_llseek,
+};
+
+static ssize_t trigger_cntr_show(struct device *dev,
+			    struct device_attribute *attr, char *buf)
+{
+	struct tmc_drvdata *drvdata = dev_get_drvdata(dev->parent);
+	unsigned long val = drvdata->trigger_cntr;
+
+	return sprintf(buf, "%#lx\n", val);
+}
+
+static ssize_t trigger_cntr_store(struct device *dev,
+			     struct device_attribute *attr,
+			     const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct tmc_drvdata *drvdata = dev_get_drvdata(dev->parent);
+
+	ret = kstrtoul(buf, 16, &val);
+	if (ret)
+		return ret;
+
+	drvdata->trigger_cntr = val;
+	return size;
+}
+static DEVICE_ATTR_RW(trigger_cntr);
+
+static struct attribute *coresight_etb_attrs[] = {
+	&dev_attr_trigger_cntr.attr,
+	NULL,
+};
+ATTRIBUTE_GROUPS(coresight_etb);
+
+static struct attribute *coresight_etr_attrs[] = {
+	&dev_attr_trigger_cntr.attr,
+	NULL,
+};
+ATTRIBUTE_GROUPS(coresight_etr);
+
+static struct attribute *coresight_etf_attrs[] = {
+	&dev_attr_trigger_cntr.attr,
+	NULL,
+};
+ATTRIBUTE_GROUPS(coresight_etf);
+
+static int tmc_probe(struct amba_device *adev, const struct amba_id *id)
+{
+	int ret = 0;
+	u32 devid;
+	void __iomem *base;
+	struct device *dev = &adev->dev;
+	struct coresight_platform_data *pdata = NULL;
+	struct tmc_drvdata *drvdata;
+	struct resource *res = &adev->res;
+	struct coresight_desc *desc;
+	struct device_node *np = adev->dev.of_node;
+
+	if (np) {
+		pdata = of_get_coresight_platform_data(dev, np);
+		if (IS_ERR(pdata))
+			return PTR_ERR(pdata);
+		adev->dev.platform_data = pdata;
+	}
+
+	drvdata = devm_kzalloc(dev, sizeof(*drvdata), GFP_KERNEL);
+	if (!drvdata)
+		return -ENOMEM;
+
+	drvdata->dev = &adev->dev;
+	dev_set_drvdata(dev, drvdata);
+
+	/* Validity for the resource is already checked by the AMBA core */
+	base = devm_ioremap_resource(dev, res);
+	if (IS_ERR(base))
+		return PTR_ERR(base);
+
+	drvdata->base = base;
+
+	spin_lock_init(&drvdata->spinlock);
+
+	drvdata->clk = adev->pclk;
+	ret = clk_prepare_enable(drvdata->clk);
+	if (ret)
+		return ret;
+
+	devid = readl_relaxed(drvdata->base + CORESIGHT_DEVID);
+	drvdata->config_type = BMVAL(devid, 6, 7);
+
+	if (drvdata->config_type == TMC_CONFIG_TYPE_ETR) {
+		if (np)
+			ret = of_property_read_u32(np,
+						   "arm,buffer-size",
+						   &drvdata->size);
+		if (ret)
+			drvdata->size = SZ_1M;
+	} else {
+		drvdata->size = readl_relaxed(drvdata->base + TMC_RSZ) * 4;
+	}
+
+	clk_disable_unprepare(drvdata->clk);
+
+	if (drvdata->config_type == TMC_CONFIG_TYPE_ETR) {
+		drvdata->vaddr = dma_alloc_coherent(dev, drvdata->size,
+						&drvdata->paddr, GFP_KERNEL);
+		if (!drvdata->vaddr)
+			return -ENOMEM;
+
+		memset(drvdata->vaddr, 0, drvdata->size);
+		drvdata->buf = drvdata->vaddr;
+	} else {
+		drvdata->buf = devm_kzalloc(dev, drvdata->size, GFP_KERNEL);
+		if (!drvdata->buf)
+			return -ENOMEM;
+	}
+
+	desc = devm_kzalloc(dev, sizeof(*desc), GFP_KERNEL);
+	if (!desc) {
+		ret = -ENOMEM;
+		goto err_devm_kzalloc;
+	}
+
+	desc->pdata = pdata;
+	desc->dev = dev;
+	desc->subtype.sink_subtype = CORESIGHT_DEV_SUBTYPE_SINK_BUFFER;
+
+	if (drvdata->config_type == TMC_CONFIG_TYPE_ETB) {
+		desc->type = CORESIGHT_DEV_TYPE_SINK;
+		desc->ops = &tmc_etb_cs_ops;
+		desc->groups = coresight_etb_groups;
+	} else if (drvdata->config_type == TMC_CONFIG_TYPE_ETR) {
+		desc->type = CORESIGHT_DEV_TYPE_SINK;
+		desc->ops = &tmc_etr_cs_ops;
+		desc->groups = coresight_etr_groups;
+	} else {
+		desc->type = CORESIGHT_DEV_TYPE_LINKSINK;
+		desc->subtype.link_subtype = CORESIGHT_DEV_SUBTYPE_LINK_FIFO;
+		desc->ops = &tmc_etf_cs_ops;
+		desc->groups = coresight_etf_groups;
+	}
+
+	drvdata->csdev = coresight_register(desc);
+	if (IS_ERR(drvdata->csdev)) {
+		ret = PTR_ERR(drvdata->csdev);
+		goto err_devm_kzalloc;
+	}
+
+	drvdata->miscdev.name = pdata->name;
+	drvdata->miscdev.minor = MISC_DYNAMIC_MINOR;
+	drvdata->miscdev.fops = &tmc_fops;
+	ret = misc_register(&drvdata->miscdev);
+	if (ret)
+		goto err_misc_register;
+
+	dev_info(dev, "TMC initialized\n");
+	return 0;
+
+err_misc_register:
+	coresight_unregister(drvdata->csdev);
+err_devm_kzalloc:
+	if (drvdata->config_type == TMC_CONFIG_TYPE_ETR)
+		dma_free_coherent(dev, drvdata->size,
+				&drvdata->paddr, GFP_KERNEL);
+	return ret;
+}
+
+static int tmc_remove(struct amba_device *adev)
+{
+	struct tmc_drvdata *drvdata = amba_get_drvdata(adev);
+
+	misc_deregister(&drvdata->miscdev);
+	coresight_unregister(drvdata->csdev);
+	if (drvdata->config_type == TMC_CONFIG_TYPE_ETR)
+		dma_free_coherent(drvdata->dev, drvdata->size,
+				  &drvdata->paddr, GFP_KERNEL);
+
+	return 0;
+}
+
+static struct amba_id tmc_ids[] = {
+	{
+		.id     = 0x0003b961,
+		.mask   = 0x0003ffff,
+	},
+	{ 0, 0},
+};
+
+static struct amba_driver tmc_driver = {
+	.drv = {
+		.name   = "coresight-tmc",
+		.owner  = THIS_MODULE,
+	},
+	.probe		= tmc_probe,
+	.remove		= tmc_remove,
+	.id_table	= tmc_ids,
+};
+
+static int __init tmc_init(void)
+{
+	return amba_driver_register(&tmc_driver);
+}
+module_init(tmc_init);
+
+static void __exit tmc_exit(void)
+{
+	amba_driver_unregister(&tmc_driver);
+}
+module_exit(tmc_exit);
+
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("CoreSight Trace Memory Controller driver");
diff --git a/drivers/coresight/coresight-tpiu.c b/drivers/coresight/coresight-tpiu.c
new file mode 100644
index 000000000000..ae101082791a
--- /dev/null
+++ b/drivers/coresight/coresight-tpiu.c
@@ -0,0 +1,217 @@
+/* Copyright (c) 2011-2012, The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/device.h>
+#include <linux/io.h>
+#include <linux/err.h>
+#include <linux/slab.h>
+#include <linux/clk.h>
+#include <linux/coresight.h>
+#include <linux/amba/bus.h>
+
+#include "coresight-priv.h"
+
+#define TPIU_SUPP_PORTSZ	0x000
+#define TPIU_CURR_PORTSZ	0x004
+#define TPIU_SUPP_TRIGMODES	0x100
+#define TPIU_TRIG_CNTRVAL	0x104
+#define TPIU_TRIG_MULT		0x108
+#define TPIU_SUPP_TESTPATM	0x200
+#define TPIU_CURR_TESTPATM	0x204
+#define TPIU_TEST_PATREPCNTR	0x208
+#define TPIU_FFSR		0x300
+#define TPIU_FFCR		0x304
+#define TPIU_FSYNC_CNTR		0x308
+#define TPIU_EXTCTL_INPORT	0x400
+#define TPIU_EXTCTL_OUTPORT	0x404
+#define TPIU_ITTRFLINACK	0xee4
+#define TPIU_ITTRFLIN		0xee8
+#define TPIU_ITATBDATA0		0xeec
+#define TPIU_ITATBCTR2		0xef0
+#define TPIU_ITATBCTR1		0xef4
+#define TPIU_ITATBCTR0		0xef8
+
+/** register definition **/
+/* FFCR - 0x304 */
+#define FFCR_FON_MAN		BIT(6)
+
+/**
+ * @base:	memory mapped base address for this component.
+ * @dev:	the device entity associated to this component.
+ * @csdev:	component vitals needed by the framework.
+ * @clk:	the clock this component is associated to.
+ */
+struct tpiu_drvdata {
+	void __iomem		*base;
+	struct device		*dev;
+	struct coresight_device	*csdev;
+	struct clk		*clk;
+};
+
+static void tpiu_enable_hw(struct tpiu_drvdata *drvdata)
+{
+	CS_UNLOCK(drvdata->base);
+
+	/* TODO: fill this up */
+
+	CS_LOCK(drvdata->base);
+}
+
+static int tpiu_enable(struct coresight_device *csdev)
+{
+	struct tpiu_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+	int ret;
+
+	ret = clk_prepare_enable(drvdata->clk);
+	if (ret)
+		return ret;
+
+	tpiu_enable_hw(drvdata);
+
+	dev_info(drvdata->dev, "TPIU enabled\n");
+	return 0;
+}
+
+static void tpiu_disable_hw(struct tpiu_drvdata *drvdata)
+{
+	CS_UNLOCK(drvdata->base);
+
+	/* Clear formatter controle reg. */
+	writel_relaxed(0x0, drvdata->base + TPIU_FFCR);
+	/* Generate manual flush */
+	writel_relaxed(FFCR_FON_MAN, drvdata->base + TPIU_FFCR);
+
+	CS_LOCK(drvdata->base);
+}
+
+static void tpiu_disable(struct coresight_device *csdev)
+{
+	struct tpiu_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	tpiu_disable_hw(drvdata);
+
+	clk_disable_unprepare(drvdata->clk);
+
+	dev_info(drvdata->dev, "TPIU disabled\n");
+}
+
+static const struct coresight_ops_sink tpiu_sink_ops = {
+	.enable		= tpiu_enable,
+	.disable	= tpiu_disable,
+};
+
+static const struct coresight_ops tpiu_cs_ops = {
+	.sink_ops	= &tpiu_sink_ops,
+};
+
+static int tpiu_probe(struct amba_device *adev, const struct amba_id *id)
+{
+	int ret;
+	void __iomem *base;
+	struct device *dev = &adev->dev;
+	struct coresight_platform_data *pdata = NULL;
+	struct tpiu_drvdata *drvdata;
+	struct resource *res = &adev->res;
+	struct coresight_desc *desc;
+	struct device_node *np = adev->dev.of_node;
+
+	if (np) {
+		pdata = of_get_coresight_platform_data(dev, np);
+		if (IS_ERR(pdata))
+			return PTR_ERR(pdata);
+		adev->dev.platform_data = pdata;
+	}
+
+	drvdata = devm_kzalloc(dev, sizeof(*drvdata), GFP_KERNEL);
+	if (!drvdata)
+		return -ENOMEM;
+
+	drvdata->dev = &adev->dev;
+	dev_set_drvdata(dev, drvdata);
+
+	/* Validity for the resource is already checked by the AMBA core */
+	base = devm_ioremap_resource(dev, res);
+	if (IS_ERR(base))
+		return PTR_ERR(base);
+
+	drvdata->base = base;
+
+	drvdata->clk = adev->pclk;
+	ret = clk_prepare_enable(drvdata->clk);
+	if (ret)
+		return ret;
+
+	/* Disable tpiu to support older devices */
+	tpiu_disable_hw(drvdata);
+
+	clk_disable_unprepare(drvdata->clk);
+
+	desc = devm_kzalloc(dev, sizeof(*desc), GFP_KERNEL);
+	if (!desc)
+		return -ENOMEM;
+
+	desc->type = CORESIGHT_DEV_TYPE_SINK;
+	desc->subtype.sink_subtype = CORESIGHT_DEV_SUBTYPE_SINK_PORT;
+	desc->ops = &tpiu_cs_ops;
+	desc->pdata = pdata;
+	desc->dev = dev;
+	drvdata->csdev = coresight_register(desc);
+	if (IS_ERR(drvdata->csdev))
+		return PTR_ERR(drvdata->csdev);
+
+	dev_info(dev, "TPIU initialized\n");
+	return 0;
+}
+
+static int tpiu_remove(struct amba_device *adev)
+{
+	struct tpiu_drvdata *drvdata = amba_get_drvdata(adev);
+
+	coresight_unregister(drvdata->csdev);
+	return 0;
+}
+
+static struct amba_id tpiu_ids[] = {
+	{
+		.id	= 0x0003b912,
+		.mask	= 0x0003ffff,
+	},
+	{ 0, 0},
+};
+
+static struct amba_driver tpiu_driver = {
+	.drv = {
+		.name	= "coresight-tpiu",
+		.owner	= THIS_MODULE,
+	},
+	.probe		= tpiu_probe,
+	.remove		= tpiu_remove,
+	.id_table	= tpiu_ids,
+};
+
+static int __init tpiu_init(void)
+{
+	return amba_driver_register(&tpiu_driver);
+}
+module_init(tpiu_init);
+
+static void __exit tpiu_exit(void)
+{
+	amba_driver_unregister(&tpiu_driver);
+}
+module_exit(tpiu_exit);
+
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("CoreSight Trace Port Interface Unit driver");
diff --git a/drivers/coresight/coresight.c b/drivers/coresight/coresight.c
new file mode 100644
index 000000000000..6e0181f84425
--- /dev/null
+++ b/drivers/coresight/coresight.c
@@ -0,0 +1,717 @@
+/* Copyright (c) 2012, The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/device.h>
+#include <linux/io.h>
+#include <linux/err.h>
+#include <linux/export.h>
+#include <linux/slab.h>
+#include <linux/mutex.h>
+#include <linux/clk.h>
+#include <linux/coresight.h>
+#include <linux/of_platform.h>
+#include <linux/delay.h>
+
+#include "coresight-priv.h"
+
+static DEFINE_MUTEX(coresight_mutex);
+
+static int coresight_id_match(struct device *dev, void *data)
+{
+	int trace_id, i_trace_id;
+	struct coresight_device *csdev, *i_csdev;
+
+	csdev = data;
+	i_csdev = to_coresight_device(dev);
+
+	/*
+	 * No need to care about oneself and components that are not
+	 * sources or not enabled
+	 */
+	if (i_csdev == csdev || !i_csdev->enable ||
+	    i_csdev->type != CORESIGHT_DEV_TYPE_SOURCE)
+		return 0;
+
+	/* Get the source ID for both compoment */
+	trace_id = source_ops(csdev)->trace_id(csdev);
+	i_trace_id = source_ops(i_csdev)->trace_id(i_csdev);
+
+	/* All you need is one */
+	if (trace_id == i_trace_id)
+		return 1;
+
+	return 0;
+}
+
+static int coresight_source_is_unique(struct coresight_device *csdev)
+{
+	int trace_id = source_ops(csdev)->trace_id(csdev);
+
+	/* this shouldn't happen */
+	if (trace_id < 0)
+		return 0;
+
+	return !bus_for_each_dev(&coresight_bustype, NULL,
+				 csdev, coresight_id_match);
+}
+
+static int coresight_find_link_inport(struct coresight_device *csdev)
+{
+	int i;
+	struct coresight_device *parent;
+	struct coresight_connection *conn;
+
+	parent = container_of(csdev->path_link.next,
+			      struct coresight_device, path_link);
+
+	for (i = 0; i < parent->nr_outport; i++) {
+		conn = &parent->conns[i];
+		if (conn->child_dev == csdev)
+			return conn->child_port;
+	}
+
+	dev_err(&csdev->dev, "couldn't find inport, parent: %s, child: %s\n",
+		dev_name(&parent->dev), dev_name(&csdev->dev));
+
+	return 0;
+}
+
+static int coresight_find_link_outport(struct coresight_device *csdev)
+{
+	int i;
+	struct coresight_device *child;
+	struct coresight_connection *conn;
+
+	child = container_of(csdev->path_link.prev,
+			     struct coresight_device, path_link);
+
+	for (i = 0; i < csdev->nr_outport; i++) {
+		conn = &csdev->conns[i];
+		if (conn->child_dev == child)
+			return conn->outport;
+	}
+
+	dev_err(&csdev->dev, "couldn't find outport, parent: %s, child: %s\n",
+		dev_name(&csdev->dev), dev_name(&child->dev));
+
+	return 0;
+}
+
+static int coresight_enable_sink(struct coresight_device *csdev)
+{
+	int ret;
+
+	if (!csdev->enable) {
+		if (sink_ops(csdev)->enable) {
+			ret = sink_ops(csdev)->enable(csdev);
+			if (ret)
+				return ret;
+		}
+		csdev->enable = true;
+	}
+
+	atomic_inc(csdev->refcnt);
+
+	return 0;
+}
+
+static void coresight_disable_sink(struct coresight_device *csdev)
+{
+	if (atomic_dec_return(csdev->refcnt) == 0) {
+		if (sink_ops(csdev)->disable) {
+			sink_ops(csdev)->disable(csdev);
+			csdev->enable = false;
+		}
+	}
+}
+
+static int coresight_enable_link(struct coresight_device *csdev)
+{
+	int ret;
+	int link_subtype;
+	int refport, inport, outport;
+
+	inport = coresight_find_link_inport(csdev);
+	outport = coresight_find_link_outport(csdev);
+	link_subtype = csdev->subtype.link_subtype;
+
+	if (link_subtype == CORESIGHT_DEV_SUBTYPE_LINK_MERG)
+		refport = inport;
+	else if (link_subtype == CORESIGHT_DEV_SUBTYPE_LINK_SPLIT)
+		refport = outport;
+	else
+		refport = 0;
+
+	if (atomic_inc_return(&csdev->refcnt[refport]) == 1) {
+		if (link_ops(csdev)->enable) {
+			ret = link_ops(csdev)->enable(csdev, inport, outport);
+			if (ret)
+				return ret;
+		}
+	}
+
+	csdev->enable = true;
+
+	return 0;
+}
+
+static void coresight_disable_link(struct coresight_device *csdev)
+{
+	int i, nr_conns;
+	int link_subtype;
+	int refport, inport, outport;
+
+	inport = coresight_find_link_inport(csdev);
+	outport = coresight_find_link_outport(csdev);
+	link_subtype = csdev->subtype.link_subtype;
+
+	if (link_subtype == CORESIGHT_DEV_SUBTYPE_LINK_MERG) {
+		refport = inport;
+		nr_conns = csdev->nr_inport;
+	} else if (link_subtype == CORESIGHT_DEV_SUBTYPE_LINK_SPLIT) {
+		refport = outport;
+		nr_conns = csdev->nr_outport;
+	} else {
+		refport = 0;
+		nr_conns = 1;
+	}
+
+	if (atomic_dec_return(&csdev->refcnt[refport]) == 0) {
+		if (link_ops(csdev)->disable)
+			link_ops(csdev)->disable(csdev, inport, outport);
+	}
+
+	for (i = 0; i < nr_conns; i++)
+		if (atomic_read(&csdev->refcnt[i]) != 0)
+			return;
+
+	csdev->enable = false;
+}
+
+static int coresight_enable_source(struct coresight_device *csdev)
+{
+	int ret;
+
+	if (!coresight_source_is_unique(csdev)) {
+		dev_warn(&csdev->dev, "traceID %d not unique\n",
+			 source_ops(csdev)->trace_id(csdev));
+		return -EINVAL;
+	}
+
+	if (!csdev->enable) {
+		if (source_ops(csdev)->enable) {
+			ret = source_ops(csdev)->enable(csdev);
+			if (ret)
+				return ret;
+		}
+		csdev->enable = true;
+	}
+
+	atomic_inc(csdev->refcnt);
+
+	return 0;
+}
+
+static void coresight_disable_source(struct coresight_device *csdev)
+{
+	if (atomic_dec_return(csdev->refcnt) == 0) {
+		if (source_ops(csdev)->disable) {
+			source_ops(csdev)->disable(csdev);
+			csdev->enable = false;
+		}
+	}
+}
+
+static int coresight_enable_path(struct list_head *path)
+{
+	int ret = 0;
+	struct coresight_device *cd;
+
+	list_for_each_entry(cd, path, path_link) {
+		if (cd == list_first_entry(path, struct coresight_device,
+					   path_link)) {
+			ret = coresight_enable_sink(cd);
+		} else if (list_is_last(&cd->path_link, path)) {
+			/*
+			 * Don't enable the source just yet - this needs to
+			 * happen at the very end when all links and sink
+			 * along the path have been configured properly.
+			 */
+			;
+		} else {
+			ret = coresight_enable_link(cd);
+		}
+		if (ret)
+			goto err;
+	}
+
+	return 0;
+err:
+	list_for_each_entry_continue_reverse(cd, path, path_link) {
+		if (cd == list_first_entry(path, struct coresight_device,
+					   path_link)) {
+			coresight_disable_sink(cd);
+		} else if (list_is_last(&cd->path_link, path)) {
+			;
+		} else {
+			coresight_disable_link(cd);
+		}
+	}
+
+	return ret;
+}
+
+static int coresight_disable_path(struct list_head *path)
+{
+	struct coresight_device *cd;
+
+	list_for_each_entry_reverse(cd, path, path_link) {
+		if (cd == list_first_entry(path, struct coresight_device,
+					   path_link)) {
+			coresight_disable_sink(cd);
+		} else if (list_is_last(&cd->path_link, path)) {
+			/*
+			 * The source has already been stopped, no need
+			 * to do it again here.
+			 */
+			;
+		} else {
+			coresight_disable_link(cd);
+		}
+	}
+
+	return 0;
+}
+
+static int coresight_build_paths(struct coresight_device *csdev,
+				 struct list_head *path,
+				 bool enable)
+{
+	int i, ret = -EINVAL;
+	struct coresight_connection *conn;
+
+	list_add(&csdev->path_link, path);
+
+	if (csdev->type == CORESIGHT_DEV_TYPE_SINK && csdev->activated) {
+		if (enable)
+			ret = coresight_enable_path(path);
+		else
+			ret = coresight_disable_path(path);
+	} else {
+		for (i = 0; i < csdev->nr_outport; i++) {
+			conn = &csdev->conns[i];
+			if (coresight_build_paths(conn->child_dev,
+						    path, enable) == 0)
+				ret = 0;
+		}
+	}
+
+	if (list_first_entry(path, struct coresight_device, path_link) != csdev)
+		dev_err(&csdev->dev, "wrong device in %s\n", __func__);
+
+	list_del(&csdev->path_link);
+
+	return ret;
+}
+
+int coresight_enable(struct coresight_device *csdev)
+{
+	int ret = 0;
+	LIST_HEAD(path);
+
+	mutex_lock(&coresight_mutex);
+	if (csdev->type != CORESIGHT_DEV_TYPE_SOURCE) {
+		ret = -EINVAL;
+		dev_err(&csdev->dev, "wrong device type in %s\n", __func__);
+		goto out;
+	}
+	if (csdev->enable)
+		goto out;
+
+	if (coresight_build_paths(csdev, &path, true)) {
+		dev_err(&csdev->dev, "building path(s) failed\n");
+		goto out;
+	}
+
+	if (coresight_enable_source(csdev))
+		dev_err(&csdev->dev, "source enable failed\n");
+out:
+	mutex_unlock(&coresight_mutex);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(coresight_enable);
+
+void coresight_disable(struct coresight_device *csdev)
+{
+	LIST_HEAD(path);
+
+	mutex_lock(&coresight_mutex);
+	if (csdev->type != CORESIGHT_DEV_TYPE_SOURCE) {
+		dev_err(&csdev->dev, "wrong device type in %s\n", __func__);
+		goto out;
+	}
+	if (!csdev->enable)
+		goto out;
+
+	coresight_disable_source(csdev);
+	if (coresight_build_paths(csdev, &path, false))
+		dev_err(&csdev->dev, "releasing path(s) failed\n");
+
+out:
+	mutex_unlock(&coresight_mutex);
+}
+EXPORT_SYMBOL_GPL(coresight_disable);
+
+static ssize_t enable_sink_show(struct device *dev,
+				struct device_attribute *attr, char *buf)
+{
+	struct coresight_device *csdev = to_coresight_device(dev);
+
+	return scnprintf(buf, PAGE_SIZE, "%u\n", (unsigned)csdev->activated);
+}
+
+static ssize_t enable_sink_store(struct device *dev,
+				 struct device_attribute *attr,
+				 const char *buf, size_t size)
+{
+	int ret;
+	unsigned long val;
+	struct coresight_device *csdev = to_coresight_device(dev);
+
+	ret = kstrtoul(buf, 10, &val);
+	if (ret)
+		return ret;
+
+	if (val)
+		csdev->activated = true;
+	else
+		csdev->activated = false;
+
+	return size;
+
+}
+static DEVICE_ATTR_RW(enable_sink);
+
+static ssize_t enable_source_show(struct device *dev,
+				  struct device_attribute *attr, char *buf)
+{
+	struct coresight_device *csdev = to_coresight_device(dev);
+
+	return scnprintf(buf, PAGE_SIZE, "%u\n", (unsigned)csdev->enable);
+}
+
+static ssize_t enable_source_store(struct device *dev,
+				   struct device_attribute *attr,
+				   const char *buf, size_t size)
+{
+	int ret = 0;
+	unsigned long val;
+	struct coresight_device *csdev = to_coresight_device(dev);
+
+	ret = kstrtoul(buf, 10, &val);
+	if (ret)
+		return ret;
+
+	if (val) {
+		ret = coresight_enable(csdev);
+		if (ret)
+			return ret;
+	} else {
+		coresight_disable(csdev);
+	}
+
+	return size;
+}
+static DEVICE_ATTR_RW(enable_source);
+
+static struct attribute *coresight_sink_attrs[] = {
+	&dev_attr_enable_sink.attr,
+	NULL,
+};
+ATTRIBUTE_GROUPS(coresight_sink);
+
+static struct attribute *coresight_source_attrs[] = {
+	&dev_attr_enable_source.attr,
+	NULL,
+};
+ATTRIBUTE_GROUPS(coresight_source);
+
+static struct device_type coresight_dev_type[] = {
+	{
+		.name = "none",
+	},
+	{
+		.name = "sink",
+		.groups = coresight_sink_groups,
+	},
+	{
+		.name = "link",
+	},
+	{
+		.name = "linksink",
+		.groups = coresight_sink_groups,
+	},
+	{
+		.name = "source",
+		.groups = coresight_source_groups,
+	},
+};
+
+static void coresight_device_release(struct device *dev)
+{
+	struct coresight_device *csdev = to_coresight_device(dev);
+
+	kfree(csdev);
+}
+
+static int coresight_orphan_match(struct device *dev, void *data)
+{
+	int i;
+	bool still_orphan = false;
+	struct coresight_device *csdev, *i_csdev;
+	struct coresight_connection *conn;
+
+	csdev = data;
+	i_csdev = to_coresight_device(dev);
+
+	/* No need to check oneself */
+	if (csdev == i_csdev)
+		return 0;
+
+	/* Move on to another component if no connection is orphan */
+	if (!i_csdev->orphan)
+		return 0;
+	/*
+	 * Circle throuch all the connection of that component.  If we find
+	 * an orphan connection whose name matches @csdev, link it.
+	 */
+	for (i = 0; i < i_csdev->nr_outport; i++)	{
+		conn = &i_csdev->conns[i];
+
+		/* We have found at least one orphan connection */
+		if (conn->child_dev == NULL) {
+			/* Does it match this newly added device? */
+			if (!strcmp(dev_name(&csdev->dev), conn->child_name))
+				conn->child_dev = csdev;
+		} else {
+			/* Too bad, this component still has an orphan */
+			still_orphan = true;
+		}
+	}
+
+	i_csdev->orphan = still_orphan;
+
+	/*
+	 * Returning '0' ensures that all known component on the
+	 * bus will be checked.
+	 */
+	return 0;
+}
+
+static void coresight_fixup_orphan_conns(struct coresight_device *csdev)
+{
+	/*
+	 * No need to check for a return value as orphan connection(s)
+	 * are hooked-up with each newly added component.
+	 */
+	bus_for_each_dev(&coresight_bustype, NULL,
+				 csdev, coresight_orphan_match);
+}
+
+
+static int coresight_name_match(struct device *dev, void *data)
+{
+	char *to_match;
+	struct coresight_device *i_csdev;
+
+	to_match = data;
+	i_csdev = to_coresight_device(dev);
+
+	if (!strcmp(to_match, dev_name(&i_csdev->dev)))
+		return 1;
+
+	return 0;
+}
+
+static void coresight_fixup_device_conns(struct coresight_device *csdev)
+{
+	int i;
+	struct device *dev = NULL;
+	struct coresight_connection *conn;
+
+	for (i = 0; i < csdev->nr_outport; i++) {
+		conn = &csdev->conns[i];
+		dev = bus_find_device(&coresight_bustype, NULL,
+				      (void *)conn->child_name,
+				      coresight_name_match);
+
+		if (dev) {
+			conn->child_dev = to_coresight_device(dev);
+		} else {
+			csdev->orphan = true;
+			conn->child_dev = NULL;
+		}
+	}
+}
+
+/**
+ * coresight_timeout - loop until a bit has changed to a specific state.
+ * @addr: base address of the area of interest.
+ * @offset: address of a register, starting from @addr.
+ * @position: the position of the bit of interest.
+ * @value: the value the bit should have.
+ *
+ * Return: 0 as soon as the bit has taken the desired state or -EAGAIN if
+ * TIMEOUT_US has elapsed, which ever happens first.
+ */
+
+int coresight_timeout(void __iomem *addr, u32 offset, int position, int value)
+{
+	int i;
+	u32 val;
+
+	for (i = TIMEOUT_US; i > 0; i--) {
+		val = __raw_readl(addr + offset);
+		/* waiting on the bit to go from 0 to 1 */
+		if (value) {
+			if (val & BIT(position))
+				return 0;
+		/* waiting on the bit to go from 1 to 0 */
+		} else {
+			if (!(val & BIT(position)))
+				return 0;
+		}
+
+		/*
+		 * Delay is arbitrary - the specification doesn't say how long
+		 * we are expected to wait.  Extra check required to make sure
+		 * we don't wait needlessly on the last iteration.
+		 */
+		if (i - 1)
+			udelay(1);
+	}
+
+	return -EAGAIN;
+}
+
+struct bus_type coresight_bustype = {
+	.name	= "coresight",
+};
+
+static int __init coresight_init(void)
+{
+	return bus_register(&coresight_bustype);
+}
+postcore_initcall(coresight_init);
+
+struct coresight_device *coresight_register(struct coresight_desc *desc)
+{
+	int i;
+	int ret;
+	int link_subtype;
+	int nr_refcnts = 1;
+	atomic_t *refcnts = NULL;
+	struct coresight_device *csdev;
+	struct coresight_connection *conns;
+
+	csdev = kzalloc(sizeof(*csdev), GFP_KERNEL);
+	if (!csdev) {
+		ret = -ENOMEM;
+		goto err_kzalloc_csdev;
+	}
+
+	if (desc->type == CORESIGHT_DEV_TYPE_LINK ||
+	    desc->type == CORESIGHT_DEV_TYPE_LINKSINK) {
+		link_subtype = desc->subtype.link_subtype;
+
+		if (link_subtype == CORESIGHT_DEV_SUBTYPE_LINK_MERG)
+			nr_refcnts = desc->pdata->nr_inport;
+		else if (link_subtype == CORESIGHT_DEV_SUBTYPE_LINK_SPLIT)
+			nr_refcnts = desc->pdata->nr_outport;
+	}
+
+	refcnts = kcalloc(nr_refcnts, sizeof(*refcnts), GFP_KERNEL);
+	if (!refcnts) {
+		ret = -ENOMEM;
+		goto err_kzalloc_refcnts;
+	}
+
+	csdev->refcnt = refcnts;
+
+	csdev->nr_inport = desc->pdata->nr_inport;
+	csdev->nr_outport = desc->pdata->nr_outport;
+	conns = kcalloc(csdev->nr_outport, sizeof(*conns), GFP_KERNEL);
+	if (!conns) {
+		ret = -ENOMEM;
+		goto err_kzalloc_conns;
+	}
+
+	for (i = 0; i < csdev->nr_outport; i++) {
+		conns[i].outport = desc->pdata->outports[i];
+		conns[i].child_name = desc->pdata->child_names[i];
+		conns[i].child_port = desc->pdata->child_ports[i];
+	}
+
+	csdev->conns = conns;
+
+	csdev->type = desc->type;
+	csdev->subtype = desc->subtype;
+	csdev->ops = desc->ops;
+	csdev->orphan = false;
+
+	csdev->dev.type = &coresight_dev_type[desc->type];
+	csdev->dev.groups = desc->groups;
+	csdev->dev.parent = desc->dev;
+	csdev->dev.release = coresight_device_release;
+	csdev->dev.bus = &coresight_bustype;
+	dev_set_name(&csdev->dev, "%s", desc->pdata->name);
+
+	ret = device_register(&csdev->dev);
+	if (ret)
+		goto err_device_register;
+
+	mutex_lock(&coresight_mutex);
+
+	coresight_fixup_device_conns(csdev);
+	coresight_fixup_orphan_conns(csdev);
+
+	mutex_unlock(&coresight_mutex);
+
+	return csdev;
+
+err_device_register:
+	kfree(conns);
+err_kzalloc_conns:
+	kfree(refcnts);
+err_kzalloc_refcnts:
+	kfree(csdev);
+err_kzalloc_csdev:
+	return ERR_PTR(ret);
+}
+EXPORT_SYMBOL_GPL(coresight_register);
+
+void coresight_unregister(struct coresight_device *csdev)
+{
+	mutex_lock(&coresight_mutex);
+
+	kfree(csdev->conns);
+	device_unregister(&csdev->dev);
+
+	mutex_unlock(&coresight_mutex);
+}
+EXPORT_SYMBOL_GPL(coresight_unregister);
+
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/coresight/of_coresight.c b/drivers/coresight/of_coresight.c
new file mode 100644
index 000000000000..5030c0734508
--- /dev/null
+++ b/drivers/coresight/of_coresight.c
@@ -0,0 +1,204 @@
+/* Copyright (c) 2012, The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/err.h>
+#include <linux/slab.h>
+#include <linux/clk.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_graph.h>
+#include <linux/of_platform.h>
+#include <linux/platform_device.h>
+#include <linux/amba/bus.h>
+#include <linux/coresight.h>
+#include <asm/smp_plat.h>
+
+
+static int of_dev_node_match(struct device *dev, void *data)
+{
+	return dev->of_node == data;
+}
+
+static struct device *
+of_coresight_get_endpoint_device(struct device_node *endpoint)
+{
+	struct device *dev = NULL;
+
+	/*
+	 * If we have a non-configuable replicator, it will be found on the
+	 * platform bus.
+	 */
+	dev = bus_find_device(&platform_bus_type, NULL,
+			      endpoint, of_dev_node_match);
+	if (dev)
+		return dev;
+
+	/*
+	 * We have a configurable component - circle through the AMBA bus
+	 * looking for the device that matches the endpoint node.
+	 */
+	return bus_find_device(&amba_bustype, NULL,
+			       endpoint, of_dev_node_match);
+}
+
+static struct device_node *of_get_coresight_endpoint(
+		const struct device_node *parent, struct device_node *prev)
+{
+	struct device_node *node = of_graph_get_next_endpoint(parent, prev);
+
+	of_node_put(prev);
+	return node;
+}
+
+static void of_coresight_get_ports(struct device_node *node,
+				   int *nr_inport, int *nr_outport)
+{
+	struct device_node *ep = NULL;
+	int in = 0, out = 0;
+
+	do {
+		ep = of_get_coresight_endpoint(node, ep);
+		if (!ep)
+			break;
+
+		if (of_property_read_bool(ep, "slave-mode"))
+			in++;
+		else
+			out++;
+
+	} while (ep);
+
+	*nr_inport = in;
+	*nr_outport = out;
+}
+
+static int of_coresight_alloc_memory(struct device *dev,
+			struct coresight_platform_data *pdata)
+{
+	/* List of output port on this component */
+	pdata->outports = devm_kzalloc(dev, pdata->nr_outport *
+				       sizeof(*pdata->outports),
+				       GFP_KERNEL);
+	if (!pdata->outports)
+		return -ENOMEM;
+
+	/* Children connected to this component via @outport */
+	 pdata->child_names = devm_kzalloc(dev, pdata->nr_outport *
+					  sizeof(*pdata->child_names),
+					  GFP_KERNEL);
+	if (!pdata->child_names)
+		return -ENOMEM;
+
+	/* Port number on the child this component is connected to */
+	pdata->child_ports = devm_kzalloc(dev, pdata->nr_outport *
+					  sizeof(*pdata->child_ports),
+					  GFP_KERNEL);
+	if (!pdata->child_ports)
+		return -ENOMEM;
+
+	return 0;
+}
+
+struct coresight_platform_data *of_get_coresight_platform_data(
+				struct device *dev, struct device_node *node)
+{
+	int i = 0, ret = 0;
+	struct coresight_platform_data *pdata;
+	struct of_endpoint endpoint, rendpoint;
+	struct device *rdev;
+	struct device_node *cpu;
+	struct device_node *ep = NULL;
+	struct device_node *rparent = NULL;
+	struct device_node *rport = NULL;
+
+	pdata = devm_kzalloc(dev, sizeof(*pdata), GFP_KERNEL);
+	if (!pdata)
+		return ERR_PTR(-ENOMEM);
+
+	/* Use device name as debugfs handle */
+	pdata->name = dev_name(dev);
+
+	/* Get the number of input and output port for this component */
+	of_coresight_get_ports(node, &pdata->nr_inport, &pdata->nr_outport);
+
+	if (pdata->nr_outport) {
+		ret = of_coresight_alloc_memory(dev, pdata);
+		if (ret)
+			return ERR_PTR(ret);
+
+		/* Iterate through each port to discover topology */
+		do {
+			/* Get a handle on a port */
+			ep = of_get_coresight_endpoint(node, ep);
+			if (!ep)
+				break;
+
+			/*
+			 * No need to deal with input ports, processing for as
+			 * processing for output ports will deal with them.
+			 */
+			if (of_find_property(ep, "slave-mode", NULL))
+				continue;
+
+			/* Get a handle on the local endpoint */
+			ret = of_graph_parse_endpoint(ep, &endpoint);
+
+			if (ret)
+				continue;
+
+			/* The local out port number */
+			pdata->outports[i] = endpoint.id;
+
+			/*
+			 * Get a handle on the remote port and parent
+			 * attached to it.
+			 */
+			rparent = of_graph_get_remote_port_parent(ep);
+			rport = of_graph_get_remote_port(ep);
+
+			if (!rparent || !rport)
+				continue;
+
+			if (of_graph_parse_endpoint(rport, &rendpoint))
+				continue;
+
+			rdev = of_coresight_get_endpoint_device(rparent);
+			if (!dev)
+				continue;
+
+			pdata->child_names[i] = dev_name(rdev);
+			pdata->child_ports[i] = rendpoint.id;
+
+			i++;
+		} while (ep);
+	}
+
+	/* Affinity defaults to CPU0 */
+	pdata->cpu = 0;
+	cpu = of_parse_phandle(node, "cpu", 0);
+	if (cpu) {
+		const u32 *mpidr;
+		int len, index;
+
+		mpidr = of_get_property(cpu, "reg", &len);
+		if (mpidr && len == 4) {
+			index = get_logical_index(be32_to_cpup(mpidr));
+			if (index != -EINVAL)
+				pdata->cpu = index;
+		}
+	}
+
+	return pdata;
+}
+EXPORT_SYMBOL_GPL(of_get_coresight_platform_data);
diff --git a/drivers/cpufreq/Kconfig b/drivers/cpufreq/Kconfig
index ffe350f86bca..29b2ef5a68b9 100644
--- a/drivers/cpufreq/Kconfig
+++ b/drivers/cpufreq/Kconfig
@@ -63,7 +63,6 @@ config CPU_FREQ_DEFAULT_GOV_PERFORMANCE
 
 config CPU_FREQ_DEFAULT_GOV_POWERSAVE
 	bool "powersave"
-	depends on EXPERT
 	select CPU_FREQ_GOV_POWERSAVE
 	help
 	  Use the CPUFreq governor 'powersave' as default. This sets
@@ -183,32 +182,34 @@ config CPU_FREQ_GOV_CONSERVATIVE
 
 	  If in doubt, say N.
 
-config GENERIC_CPUFREQ_CPU0
-	tristate "Generic CPU0 cpufreq driver"
+comment "CPU frequency scaling drivers"
+
+config CPUFREQ_DT
+	tristate "Generic DT based cpufreq driver"
 	depends on HAVE_CLK && OF
-	# if CPU_THERMAL is on and THERMAL=m, CPU0 cannot be =y:
+	# if CPU_THERMAL is on and THERMAL=m, CPUFREQ_DT cannot be =y:
 	depends on !CPU_THERMAL || THERMAL
 	select PM_OPP
 	help
-	  This adds a generic cpufreq driver for CPU0 frequency management.
+	  This adds a generic DT based cpufreq driver for frequency management.
 	  It supports both uniprocessor (UP) and symmetric multiprocessor (SMP)
 	  systems which share clock and voltage across all CPUs.
 
 	  If in doubt, say N.
 
-menu "x86 CPU frequency scaling drivers"
-depends on X86
+if X86
 source "drivers/cpufreq/Kconfig.x86"
-endmenu
+endif
 
-menu "ARM CPU frequency scaling drivers"
-depends on ARM || ARM64
+if ARM || ARM64
 source "drivers/cpufreq/Kconfig.arm"
-endmenu
+endif
 
-menu "AVR32 CPU frequency scaling drivers"
-depends on AVR32
+if PPC32 || PPC64
+source "drivers/cpufreq/Kconfig.powerpc"
+endif
 
+if AVR32
 config AVR32_AT32AP_CPUFREQ
 	bool "CPU frequency driver for AT32AP"
 	depends on PLATFORM_AT32AP
@@ -216,12 +217,9 @@ config AVR32_AT32AP_CPUFREQ
 	help
 	  This enables the CPU frequency driver for AT32AP processors.
 	  If in doubt, say N.
+endif
 
-endmenu
-
-menu "CPUFreq processor drivers"
-depends on IA64
-
+if IA64
 config IA64_ACPI_CPUFREQ
 	tristate "ACPI Processor P-States driver"
 	depends on ACPI_PROCESSOR
@@ -232,12 +230,9 @@ config IA64_ACPI_CPUFREQ
 	For details, take a look at <file:Documentation/cpu-freq/>.
 
 	If in doubt, say N.
+endif
 
-endmenu
-
-menu "MIPS CPUFreq processor drivers"
-depends on MIPS
-
+if MIPS
 config LOONGSON2_CPUFREQ
 	tristate "Loongson2 CPUFreq Driver"
 	help
@@ -250,15 +245,18 @@ config LOONGSON2_CPUFREQ
 
 	  If in doubt, say N.
 
-endmenu
+config LOONGSON1_CPUFREQ
+	tristate "Loongson1 CPUFreq Driver"
+	help
+	  This option adds a CPUFreq driver for loongson1 processors which
+	  support software configurable cpu frequency.
 
-menu "PowerPC CPU frequency scaling drivers"
-depends on PPC32 || PPC64
-source "drivers/cpufreq/Kconfig.powerpc"
-endmenu
+	  For details, take a look at <file:Documentation/cpu-freq/>.
 
-menu "SPARC CPU frequency scaling drivers"
-depends on SPARC64
+	  If in doubt, say N.
+endif
+
+if SPARC64
 config SPARC_US3_CPUFREQ
 	tristate "UltraSPARC-III CPU Frequency driver"
 	help
@@ -276,10 +274,9 @@ config SPARC_US2E_CPUFREQ
 	  For details, take a look at <file:Documentation/cpu-freq>.
 
 	  If in doubt, say N.
-endmenu
+endif
 
-menu "SH CPU Frequency scaling"
-depends on SUPERH
+if SUPERH
 config SH_CPU_FREQ
 	tristate "SuperH CPU Frequency driver"
 	help
@@ -293,7 +290,7 @@ config SH_CPU_FREQ
 	  For details, take a look at <file:Documentation/cpu-freq>.
 
 	  If unsure, say N.
-endmenu
+endif
 
 endif
 endmenu
diff --git a/drivers/cpufreq/Kconfig.arm b/drivers/cpufreq/Kconfig.arm
index 7364a538e056..0f9a2c3c0e0d 100644
--- a/drivers/cpufreq/Kconfig.arm
+++ b/drivers/cpufreq/Kconfig.arm
@@ -77,7 +77,7 @@ config ARM_EXYNOS5440_CPUFREQ
 
 config ARM_EXYNOS_CPU_FREQ_BOOST_SW
 	bool "EXYNOS Frequency Overclocking - Software"
-	depends on ARM_EXYNOS_CPUFREQ
+	depends on ARM_EXYNOS_CPUFREQ && THERMAL
 	select CPU_FREQ_BOOST_SW
 	select EXYNOS_THERMAL
 	help
@@ -92,7 +92,7 @@ config ARM_EXYNOS_CPU_FREQ_BOOST_SW
 
 config ARM_HIGHBANK_CPUFREQ
 	tristate "Calxeda Highbank-based"
-	depends on ARCH_HIGHBANK && GENERIC_CPUFREQ_CPU0 && REGULATOR
+	depends on ARCH_HIGHBANK && CPUFREQ_DT && REGULATOR
 	default m
 	help
 	  This adds the CPUFreq driver for Calxeda Highbank SoC
@@ -119,7 +119,7 @@ config ARM_INTEGRATOR
 	  If in doubt, say Y.
 
 config ARM_KIRKWOOD_CPUFREQ
-	def_bool ARCH_KIRKWOOD || MACH_KIRKWOOD
+	def_bool MACH_KIRKWOOD
 	help
 	  This adds the CPUFreq driver for Marvell Kirkwood
 	  SoCs.
@@ -247,3 +247,11 @@ config ARM_TEGRA_CPUFREQ
 	default y
 	help
 	  This adds the CPUFreq driver support for TEGRA SOCs.
+
+config ARM_PXA2xx_CPUFREQ
+	tristate "Intel PXA2xx CPUfreq driver"
+	depends on PXA27x || PXA25x
+	help
+	  This add the CPUFreq driver support for Intel PXA2xx SOCs.
+
+	  If in doubt, say N.
diff --git a/drivers/cpufreq/Makefile b/drivers/cpufreq/Makefile
index db6d9a2fea4d..b3ca7b0b2c33 100644
--- a/drivers/cpufreq/Makefile
+++ b/drivers/cpufreq/Makefile
@@ -13,7 +13,7 @@ obj-$(CONFIG_CPU_FREQ_GOV_ONDEMAND)	+= cpufreq_ondemand.o
 obj-$(CONFIG_CPU_FREQ_GOV_CONSERVATIVE)	+= cpufreq_conservative.o
 obj-$(CONFIG_CPU_FREQ_GOV_COMMON)		+= cpufreq_governor.o
 
-obj-$(CONFIG_GENERIC_CPUFREQ_CPU0)	+= cpufreq-cpu0.o
+obj-$(CONFIG_CPUFREQ_DT)		+= cpufreq-dt.o
 
 ##################################################################################
 # x86 drivers.
@@ -61,8 +61,7 @@ obj-$(CONFIG_ARM_IMX6Q_CPUFREQ)		+= imx6q-cpufreq.o
 obj-$(CONFIG_ARM_INTEGRATOR)		+= integrator-cpufreq.o
 obj-$(CONFIG_ARM_KIRKWOOD_CPUFREQ)	+= kirkwood-cpufreq.o
 obj-$(CONFIG_ARM_OMAP2PLUS_CPUFREQ)	+= omap-cpufreq.o
-obj-$(CONFIG_PXA25x)			+= pxa2xx-cpufreq.o
-obj-$(CONFIG_PXA27x)			+= pxa2xx-cpufreq.o
+obj-$(CONFIG_ARM_PXA2xx_CPUFREQ)	+= pxa2xx-cpufreq.o
 obj-$(CONFIG_PXA3xx)			+= pxa3xx-cpufreq.o
 obj-$(CONFIG_ARM_S3C24XX_CPUFREQ)	+= s3c24xx-cpufreq.o
 obj-$(CONFIG_ARM_S3C24XX_CPUFREQ_DEBUGFS) += s3c24xx-cpufreq-debugfs.o
@@ -98,6 +97,7 @@ obj-$(CONFIG_CRIS_MACH_ARTPEC3)		+= cris-artpec3-cpufreq.o
 obj-$(CONFIG_ETRAXFS)			+= cris-etraxfs-cpufreq.o
 obj-$(CONFIG_IA64_ACPI_CPUFREQ)		+= ia64-acpi-cpufreq.o
 obj-$(CONFIG_LOONGSON2_CPUFREQ)		+= loongson2_cpufreq.o
+obj-$(CONFIG_LOONGSON1_CPUFREQ)		+= ls1x-cpufreq.o
 obj-$(CONFIG_SH_CPU_FREQ)		+= sh-cpufreq.o
 obj-$(CONFIG_SPARC_US2E_CPUFREQ)	+= sparc-us2e-cpufreq.o
 obj-$(CONFIG_SPARC_US3_CPUFREQ)		+= sparc-us3-cpufreq.o
diff --git a/drivers/cpufreq/arm_big_little.c b/drivers/cpufreq/arm_big_little.c
index a46c223c2506..e1a6ba66a7f5 100644
--- a/drivers/cpufreq/arm_big_little.c
+++ b/drivers/cpufreq/arm_big_little.c
@@ -289,6 +289,8 @@ static void _put_cluster_clk_and_freq_table(struct device *cpu_dev)
 
 	clk_put(clk[cluster]);
 	dev_pm_opp_free_cpufreq_table(cpu_dev, &freq_table[cluster]);
+	if (arm_bL_ops->free_opp_table)
+		arm_bL_ops->free_opp_table(cpu_dev);
 	dev_dbg(cpu_dev, "%s: cluster: %d\n", __func__, cluster);
 }
 
@@ -337,7 +339,7 @@ static int _get_cluster_clk_and_freq_table(struct device *cpu_dev)
 	if (ret) {
 		dev_err(cpu_dev, "%s: failed to init cpufreq table, cpu: %d, err: %d\n",
 				__func__, cpu_dev->id, ret);
-		goto out;
+		goto free_opp_table;
 	}
 
 	name[12] = cluster + '0';
@@ -354,6 +356,9 @@ static int _get_cluster_clk_and_freq_table(struct device *cpu_dev)
 	ret = PTR_ERR(clk[cluster]);
 	dev_pm_opp_free_cpufreq_table(cpu_dev, &freq_table[cluster]);
 
+free_opp_table:
+	if (arm_bL_ops->free_opp_table)
+		arm_bL_ops->free_opp_table(cpu_dev);
 out:
 	dev_err(cpu_dev, "%s: Failed to get data for cluster: %d\n", __func__,
 			cluster);
diff --git a/drivers/cpufreq/arm_big_little.h b/drivers/cpufreq/arm_big_little.h
index 70f18fc12d4a..a211f7db9d32 100644
--- a/drivers/cpufreq/arm_big_little.h
+++ b/drivers/cpufreq/arm_big_little.h
@@ -25,13 +25,16 @@
 
 struct cpufreq_arm_bL_ops {
 	char name[CPUFREQ_NAME_LEN];
-	int (*get_transition_latency)(struct device *cpu_dev);
 
 	/*
 	 * This must set opp table for cpu_dev in a similar way as done by
 	 * of_init_opp_table().
 	 */
 	int (*init_opp_table)(struct device *cpu_dev);
+
+	/* Optional */
+	int (*get_transition_latency)(struct device *cpu_dev);
+	void (*free_opp_table)(struct device *cpu_dev);
 };
 
 int bL_cpufreq_register(struct cpufreq_arm_bL_ops *ops);
diff --git a/drivers/cpufreq/arm_big_little_dt.c b/drivers/cpufreq/arm_big_little_dt.c
index 4550f6976768..36d91dba2965 100644
--- a/drivers/cpufreq/arm_big_little_dt.c
+++ b/drivers/cpufreq/arm_big_little_dt.c
@@ -82,6 +82,7 @@ static struct cpufreq_arm_bL_ops dt_bL_ops = {
 	.name	= "dt-bl",
 	.get_transition_latency = dt_get_transition_latency,
 	.init_opp_table = dt_init_opp_table,
+	.free_opp_table = of_free_opp_table,
 };
 
 static int generic_bL_probe(struct platform_device *pdev)
@@ -105,7 +106,6 @@ static int generic_bL_remove(struct platform_device *pdev)
 static struct platform_driver generic_bL_platdrv = {
 	.driver = {
 		.name	= "arm-bL-cpufreq-dt",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= generic_bL_probe,
 	.remove		= generic_bL_remove,
diff --git a/drivers/cpufreq/cpufreq-cpu0.c b/drivers/cpufreq/cpufreq-cpu0.c
deleted file mode 100644
index 0d2172b07765..000000000000
--- a/drivers/cpufreq/cpufreq-cpu0.c
+++ /dev/null
@@ -1,248 +0,0 @@
-/*
- * Copyright (C) 2012 Freescale Semiconductor, Inc.
- *
- * The OPP code in function cpu0_set_target() is reused from
- * drivers/cpufreq/omap-cpufreq.c
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- */
-
-#define pr_fmt(fmt)	KBUILD_MODNAME ": " fmt
-
-#include <linux/clk.h>
-#include <linux/cpu.h>
-#include <linux/cpu_cooling.h>
-#include <linux/cpufreq.h>
-#include <linux/cpumask.h>
-#include <linux/err.h>
-#include <linux/module.h>
-#include <linux/of.h>
-#include <linux/pm_opp.h>
-#include <linux/platform_device.h>
-#include <linux/regulator/consumer.h>
-#include <linux/slab.h>
-#include <linux/thermal.h>
-
-static unsigned int transition_latency;
-static unsigned int voltage_tolerance; /* in percentage */
-
-static struct device *cpu_dev;
-static struct clk *cpu_clk;
-static struct regulator *cpu_reg;
-static struct cpufreq_frequency_table *freq_table;
-static struct thermal_cooling_device *cdev;
-
-static int cpu0_set_target(struct cpufreq_policy *policy, unsigned int index)
-{
-	struct dev_pm_opp *opp;
-	unsigned long volt = 0, volt_old = 0, tol = 0;
-	unsigned int old_freq, new_freq;
-	long freq_Hz, freq_exact;
-	int ret;
-
-	freq_Hz = clk_round_rate(cpu_clk, freq_table[index].frequency * 1000);
-	if (freq_Hz <= 0)
-		freq_Hz = freq_table[index].frequency * 1000;
-
-	freq_exact = freq_Hz;
-	new_freq = freq_Hz / 1000;
-	old_freq = clk_get_rate(cpu_clk) / 1000;
-
-	if (!IS_ERR(cpu_reg)) {
-		rcu_read_lock();
-		opp = dev_pm_opp_find_freq_ceil(cpu_dev, &freq_Hz);
-		if (IS_ERR(opp)) {
-			rcu_read_unlock();
-			pr_err("failed to find OPP for %ld\n", freq_Hz);
-			return PTR_ERR(opp);
-		}
-		volt = dev_pm_opp_get_voltage(opp);
-		rcu_read_unlock();
-		tol = volt * voltage_tolerance / 100;
-		volt_old = regulator_get_voltage(cpu_reg);
-	}
-
-	pr_debug("%u MHz, %ld mV --> %u MHz, %ld mV\n",
-		 old_freq / 1000, volt_old ? volt_old / 1000 : -1,
-		 new_freq / 1000, volt ? volt / 1000 : -1);
-
-	/* scaling up?  scale voltage before frequency */
-	if (!IS_ERR(cpu_reg) && new_freq > old_freq) {
-		ret = regulator_set_voltage_tol(cpu_reg, volt, tol);
-		if (ret) {
-			pr_err("failed to scale voltage up: %d\n", ret);
-			return ret;
-		}
-	}
-
-	ret = clk_set_rate(cpu_clk, freq_exact);
-	if (ret) {
-		pr_err("failed to set clock rate: %d\n", ret);
-		if (!IS_ERR(cpu_reg))
-			regulator_set_voltage_tol(cpu_reg, volt_old, tol);
-		return ret;
-	}
-
-	/* scaling down?  scale voltage after frequency */
-	if (!IS_ERR(cpu_reg) && new_freq < old_freq) {
-		ret = regulator_set_voltage_tol(cpu_reg, volt, tol);
-		if (ret) {
-			pr_err("failed to scale voltage down: %d\n", ret);
-			clk_set_rate(cpu_clk, old_freq * 1000);
-		}
-	}
-
-	return ret;
-}
-
-static int cpu0_cpufreq_init(struct cpufreq_policy *policy)
-{
-	policy->clk = cpu_clk;
-	return cpufreq_generic_init(policy, freq_table, transition_latency);
-}
-
-static struct cpufreq_driver cpu0_cpufreq_driver = {
-	.flags = CPUFREQ_STICKY | CPUFREQ_NEED_INITIAL_FREQ_CHECK,
-	.verify = cpufreq_generic_frequency_table_verify,
-	.target_index = cpu0_set_target,
-	.get = cpufreq_generic_get,
-	.init = cpu0_cpufreq_init,
-	.name = "generic_cpu0",
-	.attr = cpufreq_generic_attr,
-};
-
-static int cpu0_cpufreq_probe(struct platform_device *pdev)
-{
-	struct device_node *np;
-	int ret;
-
-	cpu_dev = get_cpu_device(0);
-	if (!cpu_dev) {
-		pr_err("failed to get cpu0 device\n");
-		return -ENODEV;
-	}
-
-	np = of_node_get(cpu_dev->of_node);
-	if (!np) {
-		pr_err("failed to find cpu0 node\n");
-		return -ENOENT;
-	}
-
-	cpu_reg = regulator_get_optional(cpu_dev, "cpu0");
-	if (IS_ERR(cpu_reg)) {
-		/*
-		 * If cpu0 regulator supply node is present, but regulator is
-		 * not yet registered, we should try defering probe.
-		 */
-		if (PTR_ERR(cpu_reg) == -EPROBE_DEFER) {
-			dev_dbg(cpu_dev, "cpu0 regulator not ready, retry\n");
-			ret = -EPROBE_DEFER;
-			goto out_put_node;
-		}
-		pr_warn("failed to get cpu0 regulator: %ld\n",
-			PTR_ERR(cpu_reg));
-	}
-
-	cpu_clk = clk_get(cpu_dev, NULL);
-	if (IS_ERR(cpu_clk)) {
-		ret = PTR_ERR(cpu_clk);
-		pr_err("failed to get cpu0 clock: %d\n", ret);
-		goto out_put_reg;
-	}
-
-	/* OPPs might be populated at runtime, don't check for error here */
-	of_init_opp_table(cpu_dev);
-
-	ret = dev_pm_opp_init_cpufreq_table(cpu_dev, &freq_table);
-	if (ret) {
-		pr_err("failed to init cpufreq table: %d\n", ret);
-		goto out_put_clk;
-	}
-
-	of_property_read_u32(np, "voltage-tolerance", &voltage_tolerance);
-
-	if (of_property_read_u32(np, "clock-latency", &transition_latency))
-		transition_latency = CPUFREQ_ETERNAL;
-
-	if (!IS_ERR(cpu_reg)) {
-		struct dev_pm_opp *opp;
-		unsigned long min_uV, max_uV;
-		int i;
-
-		/*
-		 * OPP is maintained in order of increasing frequency, and
-		 * freq_table initialised from OPP is therefore sorted in the
-		 * same order.
-		 */
-		for (i = 0; freq_table[i].frequency != CPUFREQ_TABLE_END; i++)
-			;
-		rcu_read_lock();
-		opp = dev_pm_opp_find_freq_exact(cpu_dev,
-				freq_table[0].frequency * 1000, true);
-		min_uV = dev_pm_opp_get_voltage(opp);
-		opp = dev_pm_opp_find_freq_exact(cpu_dev,
-				freq_table[i-1].frequency * 1000, true);
-		max_uV = dev_pm_opp_get_voltage(opp);
-		rcu_read_unlock();
-		ret = regulator_set_voltage_time(cpu_reg, min_uV, max_uV);
-		if (ret > 0)
-			transition_latency += ret * 1000;
-	}
-
-	ret = cpufreq_register_driver(&cpu0_cpufreq_driver);
-	if (ret) {
-		pr_err("failed register driver: %d\n", ret);
-		goto out_free_table;
-	}
-
-	/*
-	 * For now, just loading the cooling device;
-	 * thermal DT code takes care of matching them.
-	 */
-	if (of_find_property(np, "#cooling-cells", NULL)) {
-		cdev = of_cpufreq_cooling_register(np, cpu_present_mask);
-		if (IS_ERR(cdev))
-			pr_err("running cpufreq without cooling device: %ld\n",
-			       PTR_ERR(cdev));
-	}
-
-	of_node_put(np);
-	return 0;
-
-out_free_table:
-	dev_pm_opp_free_cpufreq_table(cpu_dev, &freq_table);
-out_put_clk:
-	if (!IS_ERR(cpu_clk))
-		clk_put(cpu_clk);
-out_put_reg:
-	if (!IS_ERR(cpu_reg))
-		regulator_put(cpu_reg);
-out_put_node:
-	of_node_put(np);
-	return ret;
-}
-
-static int cpu0_cpufreq_remove(struct platform_device *pdev)
-{
-	cpufreq_cooling_unregister(cdev);
-	cpufreq_unregister_driver(&cpu0_cpufreq_driver);
-	dev_pm_opp_free_cpufreq_table(cpu_dev, &freq_table);
-
-	return 0;
-}
-
-static struct platform_driver cpu0_cpufreq_platdrv = {
-	.driver = {
-		.name	= "cpufreq-cpu0",
-		.owner	= THIS_MODULE,
-	},
-	.probe		= cpu0_cpufreq_probe,
-	.remove		= cpu0_cpufreq_remove,
-};
-module_platform_driver(cpu0_cpufreq_platdrv);
-
-MODULE_AUTHOR("Shawn Guo <shawn.guo@linaro.org>");
-MODULE_DESCRIPTION("Generic CPU0 cpufreq driver");
-MODULE_LICENSE("GPL");
diff --git a/drivers/cpufreq/cpufreq-dt.c b/drivers/cpufreq/cpufreq-dt.c
new file mode 100644
index 000000000000..fde97d6e31d6
--- /dev/null
+++ b/drivers/cpufreq/cpufreq-dt.c
@@ -0,0 +1,423 @@
+/*
+ * Copyright (C) 2012 Freescale Semiconductor, Inc.
+ *
+ * Copyright (C) 2014 Linaro.
+ * Viresh Kumar <viresh.kumar@linaro.org>
+ *
+ * The OPP code in function set_target() is reused from
+ * drivers/cpufreq/omap-cpufreq.c
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#define pr_fmt(fmt)	KBUILD_MODNAME ": " fmt
+
+#include <linux/clk.h>
+#include <linux/cpu.h>
+#include <linux/cpu_cooling.h>
+#include <linux/cpufreq.h>
+#include <linux/cpufreq-dt.h>
+#include <linux/cpumask.h>
+#include <linux/err.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/pm_opp.h>
+#include <linux/platform_device.h>
+#include <linux/regulator/consumer.h>
+#include <linux/slab.h>
+#include <linux/thermal.h>
+
+struct private_data {
+	struct device *cpu_dev;
+	struct regulator *cpu_reg;
+	struct thermal_cooling_device *cdev;
+	unsigned int voltage_tolerance; /* in percentage */
+};
+
+static int set_target(struct cpufreq_policy *policy, unsigned int index)
+{
+	struct dev_pm_opp *opp;
+	struct cpufreq_frequency_table *freq_table = policy->freq_table;
+	struct clk *cpu_clk = policy->clk;
+	struct private_data *priv = policy->driver_data;
+	struct device *cpu_dev = priv->cpu_dev;
+	struct regulator *cpu_reg = priv->cpu_reg;
+	unsigned long volt = 0, volt_old = 0, tol = 0;
+	unsigned int old_freq, new_freq;
+	long freq_Hz, freq_exact;
+	int ret;
+
+	freq_Hz = clk_round_rate(cpu_clk, freq_table[index].frequency * 1000);
+	if (freq_Hz <= 0)
+		freq_Hz = freq_table[index].frequency * 1000;
+
+	freq_exact = freq_Hz;
+	new_freq = freq_Hz / 1000;
+	old_freq = clk_get_rate(cpu_clk) / 1000;
+
+	if (!IS_ERR(cpu_reg)) {
+		unsigned long opp_freq;
+
+		rcu_read_lock();
+		opp = dev_pm_opp_find_freq_ceil(cpu_dev, &freq_Hz);
+		if (IS_ERR(opp)) {
+			rcu_read_unlock();
+			dev_err(cpu_dev, "failed to find OPP for %ld\n",
+				freq_Hz);
+			return PTR_ERR(opp);
+		}
+		volt = dev_pm_opp_get_voltage(opp);
+		opp_freq = dev_pm_opp_get_freq(opp);
+		rcu_read_unlock();
+		tol = volt * priv->voltage_tolerance / 100;
+		volt_old = regulator_get_voltage(cpu_reg);
+		dev_dbg(cpu_dev, "Found OPP: %ld kHz, %ld uV\n",
+			opp_freq / 1000, volt);
+	}
+
+	dev_dbg(cpu_dev, "%u MHz, %ld mV --> %u MHz, %ld mV\n",
+		old_freq / 1000, (volt_old > 0) ? volt_old / 1000 : -1,
+		new_freq / 1000, volt ? volt / 1000 : -1);
+
+	/* scaling up?  scale voltage before frequency */
+	if (!IS_ERR(cpu_reg) && new_freq > old_freq) {
+		ret = regulator_set_voltage_tol(cpu_reg, volt, tol);
+		if (ret) {
+			dev_err(cpu_dev, "failed to scale voltage up: %d\n",
+				ret);
+			return ret;
+		}
+	}
+
+	ret = clk_set_rate(cpu_clk, freq_exact);
+	if (ret) {
+		dev_err(cpu_dev, "failed to set clock rate: %d\n", ret);
+		if (!IS_ERR(cpu_reg) && volt_old > 0)
+			regulator_set_voltage_tol(cpu_reg, volt_old, tol);
+		return ret;
+	}
+
+	/* scaling down?  scale voltage after frequency */
+	if (!IS_ERR(cpu_reg) && new_freq < old_freq) {
+		ret = regulator_set_voltage_tol(cpu_reg, volt, tol);
+		if (ret) {
+			dev_err(cpu_dev, "failed to scale voltage down: %d\n",
+				ret);
+			clk_set_rate(cpu_clk, old_freq * 1000);
+		}
+	}
+
+	return ret;
+}
+
+static int allocate_resources(int cpu, struct device **cdev,
+			      struct regulator **creg, struct clk **cclk)
+{
+	struct device *cpu_dev;
+	struct regulator *cpu_reg;
+	struct clk *cpu_clk;
+	int ret = 0;
+	char *reg_cpu0 = "cpu0", *reg_cpu = "cpu", *reg;
+
+	cpu_dev = get_cpu_device(cpu);
+	if (!cpu_dev) {
+		pr_err("failed to get cpu%d device\n", cpu);
+		return -ENODEV;
+	}
+
+	/* Try "cpu0" for older DTs */
+	if (!cpu)
+		reg = reg_cpu0;
+	else
+		reg = reg_cpu;
+
+try_again:
+	cpu_reg = regulator_get_optional(cpu_dev, reg);
+	if (IS_ERR(cpu_reg)) {
+		/*
+		 * If cpu's regulator supply node is present, but regulator is
+		 * not yet registered, we should try defering probe.
+		 */
+		if (PTR_ERR(cpu_reg) == -EPROBE_DEFER) {
+			dev_dbg(cpu_dev, "cpu%d regulator not ready, retry\n",
+				cpu);
+			return -EPROBE_DEFER;
+		}
+
+		/* Try with "cpu-supply" */
+		if (reg == reg_cpu0) {
+			reg = reg_cpu;
+			goto try_again;
+		}
+
+		dev_dbg(cpu_dev, "no regulator for cpu%d: %ld\n",
+			cpu, PTR_ERR(cpu_reg));
+	}
+
+	cpu_clk = clk_get(cpu_dev, NULL);
+	if (IS_ERR(cpu_clk)) {
+		/* put regulator */
+		if (!IS_ERR(cpu_reg))
+			regulator_put(cpu_reg);
+
+		ret = PTR_ERR(cpu_clk);
+
+		/*
+		 * If cpu's clk node is present, but clock is not yet
+		 * registered, we should try defering probe.
+		 */
+		if (ret == -EPROBE_DEFER)
+			dev_dbg(cpu_dev, "cpu%d clock not ready, retry\n", cpu);
+		else
+			dev_err(cpu_dev, "failed to get cpu%d clock: %d\n", cpu,
+				ret);
+	} else {
+		*cdev = cpu_dev;
+		*creg = cpu_reg;
+		*cclk = cpu_clk;
+	}
+
+	return ret;
+}
+
+static int cpufreq_init(struct cpufreq_policy *policy)
+{
+	struct cpufreq_dt_platform_data *pd;
+	struct cpufreq_frequency_table *freq_table;
+	struct device_node *np;
+	struct private_data *priv;
+	struct device *cpu_dev;
+	struct regulator *cpu_reg;
+	struct clk *cpu_clk;
+	unsigned long min_uV = ~0, max_uV = 0;
+	unsigned int transition_latency;
+	int ret;
+
+	ret = allocate_resources(policy->cpu, &cpu_dev, &cpu_reg, &cpu_clk);
+	if (ret) {
+		pr_err("%s: Failed to allocate resources: %d\n", __func__, ret);
+		return ret;
+	}
+
+	np = of_node_get(cpu_dev->of_node);
+	if (!np) {
+		dev_err(cpu_dev, "failed to find cpu%d node\n", policy->cpu);
+		ret = -ENOENT;
+		goto out_put_reg_clk;
+	}
+
+	/* OPPs might be populated at runtime, don't check for error here */
+	of_init_opp_table(cpu_dev);
+
+	/*
+	 * But we need OPP table to function so if it is not there let's
+	 * give platform code chance to provide it for us.
+	 */
+	ret = dev_pm_opp_get_opp_count(cpu_dev);
+	if (ret <= 0) {
+		pr_debug("OPP table is not ready, deferring probe\n");
+		ret = -EPROBE_DEFER;
+		goto out_free_opp;
+	}
+
+	priv = kzalloc(sizeof(*priv), GFP_KERNEL);
+	if (!priv) {
+		ret = -ENOMEM;
+		goto out_free_opp;
+	}
+
+	of_property_read_u32(np, "voltage-tolerance", &priv->voltage_tolerance);
+
+	if (of_property_read_u32(np, "clock-latency", &transition_latency))
+		transition_latency = CPUFREQ_ETERNAL;
+
+	if (!IS_ERR(cpu_reg)) {
+		unsigned long opp_freq = 0;
+
+		/*
+		 * Disable any OPPs where the connected regulator isn't able to
+		 * provide the specified voltage and record minimum and maximum
+		 * voltage levels.
+		 */
+		while (1) {
+			struct dev_pm_opp *opp;
+			unsigned long opp_uV, tol_uV;
+
+			rcu_read_lock();
+			opp = dev_pm_opp_find_freq_ceil(cpu_dev, &opp_freq);
+			if (IS_ERR(opp)) {
+				rcu_read_unlock();
+				break;
+			}
+			opp_uV = dev_pm_opp_get_voltage(opp);
+			rcu_read_unlock();
+
+			tol_uV = opp_uV * priv->voltage_tolerance / 100;
+			if (regulator_is_supported_voltage(cpu_reg, opp_uV,
+							   opp_uV + tol_uV)) {
+				if (opp_uV < min_uV)
+					min_uV = opp_uV;
+				if (opp_uV > max_uV)
+					max_uV = opp_uV;
+			} else {
+				dev_pm_opp_disable(cpu_dev, opp_freq);
+			}
+
+			opp_freq++;
+		}
+
+		ret = regulator_set_voltage_time(cpu_reg, min_uV, max_uV);
+		if (ret > 0)
+			transition_latency += ret * 1000;
+	}
+
+	ret = dev_pm_opp_init_cpufreq_table(cpu_dev, &freq_table);
+	if (ret) {
+		pr_err("failed to init cpufreq table: %d\n", ret);
+		goto out_free_priv;
+	}
+
+	priv->cpu_dev = cpu_dev;
+	priv->cpu_reg = cpu_reg;
+	policy->driver_data = priv;
+
+	policy->clk = cpu_clk;
+	ret = cpufreq_table_validate_and_show(policy, freq_table);
+	if (ret) {
+		dev_err(cpu_dev, "%s: invalid frequency table: %d\n", __func__,
+			ret);
+		goto out_free_cpufreq_table;
+	}
+
+	policy->cpuinfo.transition_latency = transition_latency;
+
+	pd = cpufreq_get_driver_data();
+	if (!pd || !pd->independent_clocks)
+		cpumask_setall(policy->cpus);
+
+	of_node_put(np);
+
+	return 0;
+
+out_free_cpufreq_table:
+	dev_pm_opp_free_cpufreq_table(cpu_dev, &freq_table);
+out_free_priv:
+	kfree(priv);
+out_free_opp:
+	of_free_opp_table(cpu_dev);
+	of_node_put(np);
+out_put_reg_clk:
+	clk_put(cpu_clk);
+	if (!IS_ERR(cpu_reg))
+		regulator_put(cpu_reg);
+
+	return ret;
+}
+
+static int cpufreq_exit(struct cpufreq_policy *policy)
+{
+	struct private_data *priv = policy->driver_data;
+
+	if (priv->cdev)
+		cpufreq_cooling_unregister(priv->cdev);
+	dev_pm_opp_free_cpufreq_table(priv->cpu_dev, &policy->freq_table);
+	of_free_opp_table(priv->cpu_dev);
+	clk_put(policy->clk);
+	if (!IS_ERR(priv->cpu_reg))
+		regulator_put(priv->cpu_reg);
+	kfree(priv);
+
+	return 0;
+}
+
+static void cpufreq_ready(struct cpufreq_policy *policy)
+{
+	struct private_data *priv = policy->driver_data;
+	struct device_node *np = of_node_get(priv->cpu_dev->of_node);
+
+	if (WARN_ON(!np))
+		return;
+
+	/*
+	 * For now, just loading the cooling device;
+	 * thermal DT code takes care of matching them.
+	 */
+	if (of_find_property(np, "#cooling-cells", NULL)) {
+		priv->cdev = of_cpufreq_cooling_register(np,
+							 policy->related_cpus);
+		if (IS_ERR(priv->cdev)) {
+			dev_err(priv->cpu_dev,
+				"running cpufreq without cooling device: %ld\n",
+				PTR_ERR(priv->cdev));
+
+			priv->cdev = NULL;
+		}
+	}
+
+	of_node_put(np);
+}
+
+static struct cpufreq_driver dt_cpufreq_driver = {
+	.flags = CPUFREQ_STICKY | CPUFREQ_NEED_INITIAL_FREQ_CHECK,
+	.verify = cpufreq_generic_frequency_table_verify,
+	.target_index = set_target,
+	.get = cpufreq_generic_get,
+	.init = cpufreq_init,
+	.exit = cpufreq_exit,
+	.ready = cpufreq_ready,
+	.name = "cpufreq-dt",
+	.attr = cpufreq_generic_attr,
+};
+
+static int dt_cpufreq_probe(struct platform_device *pdev)
+{
+	struct device *cpu_dev;
+	struct regulator *cpu_reg;
+	struct clk *cpu_clk;
+	int ret;
+
+	/*
+	 * All per-cluster (CPUs sharing clock/voltages) initialization is done
+	 * from ->init(). In probe(), we just need to make sure that clk and
+	 * regulators are available. Else defer probe and retry.
+	 *
+	 * FIXME: Is checking this only for CPU0 sufficient ?
+	 */
+	ret = allocate_resources(0, &cpu_dev, &cpu_reg, &cpu_clk);
+	if (ret)
+		return ret;
+
+	clk_put(cpu_clk);
+	if (!IS_ERR(cpu_reg))
+		regulator_put(cpu_reg);
+
+	dt_cpufreq_driver.driver_data = dev_get_platdata(&pdev->dev);
+
+	ret = cpufreq_register_driver(&dt_cpufreq_driver);
+	if (ret)
+		dev_err(cpu_dev, "failed register driver: %d\n", ret);
+
+	return ret;
+}
+
+static int dt_cpufreq_remove(struct platform_device *pdev)
+{
+	cpufreq_unregister_driver(&dt_cpufreq_driver);
+	return 0;
+}
+
+static struct platform_driver dt_cpufreq_platdrv = {
+	.driver = {
+		.name	= "cpufreq-dt",
+	},
+	.probe		= dt_cpufreq_probe,
+	.remove		= dt_cpufreq_remove,
+};
+module_platform_driver(dt_cpufreq_platdrv);
+
+MODULE_AUTHOR("Viresh Kumar <viresh.kumar@linaro.org>");
+MODULE_AUTHOR("Shawn Guo <shawn.guo@linaro.org>");
+MODULE_DESCRIPTION("Generic cpufreq driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
index 61190f6b4829..46bed4f81cde 100644
--- a/drivers/cpufreq/cpufreq.c
+++ b/drivers/cpufreq/cpufreq.c
@@ -437,7 +437,7 @@ static struct cpufreq_governor *__find_governor(const char *str_governor)
 	struct cpufreq_governor *t;
 
 	list_for_each_entry(t, &cpufreq_governor_list, governor_list)
-		if (!strnicmp(str_governor, t->name, CPUFREQ_NAME_LEN))
+		if (!strncasecmp(str_governor, t->name, CPUFREQ_NAME_LEN))
 			return t;
 
 	return NULL;
@@ -455,10 +455,10 @@ static int cpufreq_parse_governor(char *str_governor, unsigned int *policy,
 		goto out;
 
 	if (cpufreq_driver->setpolicy) {
-		if (!strnicmp(str_governor, "performance", CPUFREQ_NAME_LEN)) {
+		if (!strncasecmp(str_governor, "performance", CPUFREQ_NAME_LEN)) {
 			*policy = CPUFREQ_POLICY_PERFORMANCE;
 			err = 0;
-		} else if (!strnicmp(str_governor, "powersave",
+		} else if (!strncasecmp(str_governor, "powersave",
 						CPUFREQ_NAME_LEN)) {
 			*policy = CPUFREQ_POLICY_POWERSAVE;
 			err = 0;
@@ -512,7 +512,18 @@ show_one(cpuinfo_max_freq, cpuinfo.max_freq);
 show_one(cpuinfo_transition_latency, cpuinfo.transition_latency);
 show_one(scaling_min_freq, min);
 show_one(scaling_max_freq, max);
-show_one(scaling_cur_freq, cur);
+
+static ssize_t show_scaling_cur_freq(
+	struct cpufreq_policy *policy, char *buf)
+{
+	ssize_t ret;
+
+	if (cpufreq_driver && cpufreq_driver->setpolicy && cpufreq_driver->get)
+		ret = sprintf(buf, "%u\n", cpufreq_driver->get(policy->cpu));
+	else
+		ret = sprintf(buf, "%u\n", policy->cur);
+	return ret;
+}
 
 static int cpufreq_set_policy(struct cpufreq_policy *policy,
 				struct cpufreq_policy *new_policy);
@@ -524,7 +535,7 @@ static int cpufreq_set_policy(struct cpufreq_policy *policy,
 static ssize_t store_##file_name					\
 (struct cpufreq_policy *policy, const char *buf, size_t count)		\
 {									\
-	int ret;							\
+	int ret, temp;							\
 	struct cpufreq_policy new_policy;				\
 									\
 	ret = cpufreq_get_policy(&new_policy, policy->cpu);		\
@@ -535,8 +546,10 @@ static ssize_t store_##file_name					\
 	if (ret != 1)							\
 		return -EINVAL;						\
 									\
+	temp = new_policy.object;					\
 	ret = cpufreq_set_policy(policy, &new_policy);		\
-	policy->user_policy.object = policy->object;			\
+	if (!ret)							\
+		policy->user_policy.object = temp;			\
 									\
 	return ret ? ret : count;					\
 }
@@ -887,46 +900,31 @@ static int cpufreq_add_dev_interface(struct cpufreq_policy *policy,
 	struct freq_attr **drv_attr;
 	int ret = 0;
 
-	/* prepare interface data */
-	ret = kobject_init_and_add(&policy->kobj, &ktype_cpufreq,
-				   &dev->kobj, "cpufreq");
-	if (ret)
-		return ret;
-
 	/* set up files for this cpu device */
 	drv_attr = cpufreq_driver->attr;
 	while ((drv_attr) && (*drv_attr)) {
 		ret = sysfs_create_file(&policy->kobj, &((*drv_attr)->attr));
 		if (ret)
-			goto err_out_kobj_put;
+			return ret;
 		drv_attr++;
 	}
 	if (cpufreq_driver->get) {
 		ret = sysfs_create_file(&policy->kobj, &cpuinfo_cur_freq.attr);
 		if (ret)
-			goto err_out_kobj_put;
-	}
-	if (has_target()) {
-		ret = sysfs_create_file(&policy->kobj, &scaling_cur_freq.attr);
-		if (ret)
-			goto err_out_kobj_put;
+			return ret;
 	}
+
+	ret = sysfs_create_file(&policy->kobj, &scaling_cur_freq.attr);
+	if (ret)
+		return ret;
+
 	if (cpufreq_driver->bios_limit) {
 		ret = sysfs_create_file(&policy->kobj, &bios_limit.attr);
 		if (ret)
-			goto err_out_kobj_put;
+			return ret;
 	}
 
-	ret = cpufreq_add_dev_symlink(policy);
-	if (ret)
-		goto err_out_kobj_put;
-
-	return ret;
-
-err_out_kobj_put:
-	kobject_put(&policy->kobj);
-	wait_for_completion(&policy->kobj_unregister);
-	return ret;
+	return cpufreq_add_dev_symlink(policy);
 }
 
 static void cpufreq_init_policy(struct cpufreq_policy *policy)
@@ -1011,7 +1009,8 @@ static struct cpufreq_policy *cpufreq_policy_restore(unsigned int cpu)
 
 	read_unlock_irqrestore(&cpufreq_driver_lock, flags);
 
-	policy->governor = NULL;
+	if (policy)
+		policy->governor = NULL;
 
 	return policy;
 }
@@ -1184,6 +1183,8 @@ static int __cpufreq_add_dev(struct device *dev, struct subsys_interface *sif)
 		goto err_set_policy_cpu;
 	}
 
+	down_write(&policy->rwsem);
+
 	/* related cpus should atleast have policy->cpus */
 	cpumask_or(policy->related_cpus, policy->related_cpus, policy->cpus);
 
@@ -1196,9 +1197,17 @@ static int __cpufreq_add_dev(struct device *dev, struct subsys_interface *sif)
 	if (!recover_policy) {
 		policy->user_policy.min = policy->min;
 		policy->user_policy.max = policy->max;
+
+		/* prepare interface data */
+		ret = kobject_init_and_add(&policy->kobj, &ktype_cpufreq,
+					   &dev->kobj, "cpufreq");
+		if (ret) {
+			pr_err("%s: failed to init policy->kobj: %d\n",
+			       __func__, ret);
+			goto err_init_policy_kobj;
+		}
 	}
 
-	down_write(&policy->rwsem);
 	write_lock_irqsave(&cpufreq_driver_lock, flags);
 	for_each_cpu(j, policy->cpus)
 		per_cpu(cpufreq_cpu_data, j) = policy;
@@ -1276,8 +1285,13 @@ static int __cpufreq_add_dev(struct device *dev, struct subsys_interface *sif)
 	up_write(&policy->rwsem);
 
 	kobject_uevent(&policy->kobj, KOBJ_ADD);
+
 	up_read(&cpufreq_rwsem);
 
+	/* Callback for handling stuff after policy is ready */
+	if (cpufreq_driver->ready)
+		cpufreq_driver->ready(policy);
+
 	pr_debug("initialization complete\n");
 
 	return 0;
@@ -1289,6 +1303,11 @@ err_get_freq:
 		per_cpu(cpufreq_cpu_data, j) = NULL;
 	write_unlock_irqrestore(&cpufreq_driver_lock, flags);
 
+	if (!recover_policy) {
+		kobject_put(&policy->kobj);
+		wait_for_completion(&policy->kobj_unregister);
+	}
+err_init_policy_kobj:
 	up_write(&policy->rwsem);
 
 	if (cpufreq_driver->exit)
@@ -1382,7 +1401,7 @@ static int __cpufreq_remove_dev_prepare(struct device *dev,
 		if (!cpufreq_suspended)
 			pr_debug("%s: policy Kobject moved to cpu: %d from: %d\n",
 				 __func__, new_cpu, cpu);
-	} else if (cpufreq_driver->stop_cpu && cpufreq_driver->setpolicy) {
+	} else if (cpufreq_driver->stop_cpu) {
 		cpufreq_driver->stop_cpu(policy);
 	}
 
@@ -1731,6 +1750,21 @@ const char *cpufreq_get_current_driver(void)
 }
 EXPORT_SYMBOL_GPL(cpufreq_get_current_driver);
 
+/**
+ *	cpufreq_get_driver_data - return current driver data
+ *
+ *	Return the private data of the currently loaded cpufreq
+ *	driver, or NULL if no cpufreq driver is loaded.
+ */
+void *cpufreq_get_driver_data(void)
+{
+	if (cpufreq_driver)
+		return cpufreq_driver->driver_data;
+
+	return NULL;
+}
+EXPORT_SYMBOL_GPL(cpufreq_get_driver_data);
+
 /*********************************************************************
  *                     NOTIFIER LISTS INTERFACE                      *
  *********************************************************************/
@@ -1994,6 +2028,12 @@ static int __cpufreq_governor(struct cpufreq_policy *policy,
 	/* Don't start any governor operations if we are entering suspend */
 	if (cpufreq_suspended)
 		return 0;
+	/*
+	 * Governor might not be initiated here if ACPI _PPC changed
+	 * notification happened, so check it.
+	 */
+	if (!policy->governor)
+		return -EINVAL;
 
 	if (policy->governor->max_transition_latency &&
 	    policy->cpuinfo.transition_latency >
diff --git a/drivers/cpufreq/davinci-cpufreq.c b/drivers/cpufreq/davinci-cpufreq.c
index 28a16dc6e02e..7e336d20c184 100644
--- a/drivers/cpufreq/davinci-cpufreq.c
+++ b/drivers/cpufreq/davinci-cpufreq.c
@@ -169,7 +169,6 @@ static int __exit davinci_cpufreq_remove(struct platform_device *pdev)
 static struct platform_driver davinci_cpufreq_driver = {
 	.driver = {
 		.name	 = "cpufreq-davinci",
-		.owner	 = THIS_MODULE,
 	},
 	.remove = __exit_p(davinci_cpufreq_remove),
 };
diff --git a/drivers/cpufreq/dbx500-cpufreq.c b/drivers/cpufreq/dbx500-cpufreq.c
index 4bebc1b5db48..5c3ec1dd4921 100644
--- a/drivers/cpufreq/dbx500-cpufreq.c
+++ b/drivers/cpufreq/dbx500-cpufreq.c
@@ -69,7 +69,6 @@ static int dbx500_cpufreq_probe(struct platform_device *pdev)
 static struct platform_driver dbx500_cpufreq_plat_driver = {
 	.driver = {
 		.name = "cpufreq-ux500",
-		.owner = THIS_MODULE,
 	},
 	.probe = dbx500_cpufreq_probe,
 };
diff --git a/drivers/cpufreq/exynos-cpufreq.c b/drivers/cpufreq/exynos-cpufreq.c
index 1e0ec57bf6e3..f99a0b0b7c06 100644
--- a/drivers/cpufreq/exynos-cpufreq.c
+++ b/drivers/cpufreq/exynos-cpufreq.c
@@ -211,7 +211,6 @@ err_vdd_arm:
 static struct platform_driver exynos_cpufreq_platdrv = {
 	.driver = {
 		.name	= "exynos-cpufreq",
-		.owner	= THIS_MODULE,
 	},
 	.probe = exynos_cpufreq_probe,
 };
diff --git a/drivers/cpufreq/exynos4210-cpufreq.c b/drivers/cpufreq/exynos4210-cpufreq.c
index 61a54310a1b9..843ec824fd91 100644
--- a/drivers/cpufreq/exynos4210-cpufreq.c
+++ b/drivers/cpufreq/exynos4210-cpufreq.c
@@ -127,7 +127,7 @@ int exynos4210_cpufreq_init(struct exynos_dvfs_info *info)
 	 * dependencies on platform headers. It is necessary to enable
 	 * Exynos multi-platform support and will be removed together with
 	 * this whole driver as soon as Exynos gets migrated to use
-	 * cpufreq-cpu0 driver.
+	 * cpufreq-dt driver.
 	 */
 	np = of_find_compatible_node(NULL, NULL, "samsung,exynos4210-clock");
 	if (!np) {
diff --git a/drivers/cpufreq/exynos4x12-cpufreq.c b/drivers/cpufreq/exynos4x12-cpufreq.c
index 351a2074cfea..9e78a850e29f 100644
--- a/drivers/cpufreq/exynos4x12-cpufreq.c
+++ b/drivers/cpufreq/exynos4x12-cpufreq.c
@@ -174,7 +174,7 @@ int exynos4x12_cpufreq_init(struct exynos_dvfs_info *info)
 	 * dependencies on platform headers. It is necessary to enable
 	 * Exynos multi-platform support and will be removed together with
 	 * this whole driver as soon as Exynos gets migrated to use
-	 * cpufreq-cpu0 driver.
+	 * cpufreq-dt driver.
 	 */
 	np = of_find_compatible_node(NULL, NULL, "samsung,exynos4412-clock");
 	if (!np) {
diff --git a/drivers/cpufreq/exynos5250-cpufreq.c b/drivers/cpufreq/exynos5250-cpufreq.c
index c91ce69dc631..3eafdc7ba787 100644
--- a/drivers/cpufreq/exynos5250-cpufreq.c
+++ b/drivers/cpufreq/exynos5250-cpufreq.c
@@ -153,7 +153,7 @@ int exynos5250_cpufreq_init(struct exynos_dvfs_info *info)
 	 * dependencies on platform headers. It is necessary to enable
 	 * Exynos multi-platform support and will be removed together with
 	 * this whole driver as soon as Exynos gets migrated to use
-	 * cpufreq-cpu0 driver.
+	 * cpufreq-dt driver.
 	 */
 	np = of_find_compatible_node(NULL, NULL, "samsung,exynos5250-clock");
 	if (!np) {
diff --git a/drivers/cpufreq/exynos5440-cpufreq.c b/drivers/cpufreq/exynos5440-cpufreq.c
index f33f25b483ca..21a90ed7f3d8 100644
--- a/drivers/cpufreq/exynos5440-cpufreq.c
+++ b/drivers/cpufreq/exynos5440-cpufreq.c
@@ -371,7 +371,7 @@ static int exynos_cpufreq_probe(struct platform_device *pdev)
 	if (ret) {
 		dev_err(dvfs_info->dev,
 			"failed to init cpufreq table: %d\n", ret);
-		goto err_put_node;
+		goto err_free_opp;
 	}
 	dvfs_info->freq_count = dev_pm_opp_get_opp_count(dvfs_info->dev);
 	exynos_sort_descend_freq_table();
@@ -423,6 +423,8 @@ static int exynos_cpufreq_probe(struct platform_device *pdev)
 
 err_free_table:
 	dev_pm_opp_free_cpufreq_table(dvfs_info->dev, &dvfs_info->freq_table);
+err_free_opp:
+	of_free_opp_table(dvfs_info->dev);
 err_put_node:
 	of_node_put(np);
 	dev_err(&pdev->dev, "%s: failed initialization\n", __func__);
@@ -433,13 +435,13 @@ static int exynos_cpufreq_remove(struct platform_device *pdev)
 {
 	cpufreq_unregister_driver(&exynos_driver);
 	dev_pm_opp_free_cpufreq_table(dvfs_info->dev, &dvfs_info->freq_table);
+	of_free_opp_table(dvfs_info->dev);
 	return 0;
 }
 
 static struct platform_driver exynos_cpufreq_platdrv = {
 	.driver = {
 		.name	= "exynos5440-cpufreq",
-		.owner	= THIS_MODULE,
 		.of_match_table = exynos_cpufreq_match,
 	},
 	.probe		= exynos_cpufreq_probe,
diff --git a/drivers/cpufreq/highbank-cpufreq.c b/drivers/cpufreq/highbank-cpufreq.c
index bf8902a0866d..1608f7105c9f 100644
--- a/drivers/cpufreq/highbank-cpufreq.c
+++ b/drivers/cpufreq/highbank-cpufreq.c
@@ -6,7 +6,7 @@
  * published by the Free Software Foundation.
  *
  * This driver provides the clk notifier callbacks that are used when
- * the cpufreq-cpu0 driver changes to frequency to alert the highbank
+ * the cpufreq-dt driver changes to frequency to alert the highbank
  * EnergyCore Management Engine (ECME) about the need to change
  * voltage. The ECME interfaces with the actual voltage regulators.
  */
@@ -19,7 +19,7 @@
 #include <linux/cpu.h>
 #include <linux/err.h>
 #include <linux/of.h>
-#include <linux/mailbox.h>
+#include <linux/pl320-ipc.h>
 #include <linux/platform_device.h>
 
 #define HB_CPUFREQ_CHANGE_NOTE	0x80000001
@@ -60,7 +60,7 @@ static struct notifier_block hb_cpufreq_clk_nb = {
 
 static int hb_cpufreq_driver_init(void)
 {
-	struct platform_device_info devinfo = { .name = "cpufreq-cpu0", };
+	struct platform_device_info devinfo = { .name = "cpufreq-dt", };
 	struct device *cpu_dev;
 	struct clk *cpu_clk;
 	struct device_node *np;
@@ -95,7 +95,7 @@ static int hb_cpufreq_driver_init(void)
 		goto out_put_node;
 	}
 
-	/* Instantiate cpufreq-cpu0 */
+	/* Instantiate cpufreq-dt */
 	platform_device_register_full(&devinfo);
 
 out_put_node:
diff --git a/drivers/cpufreq/imx6q-cpufreq.c b/drivers/cpufreq/imx6q-cpufreq.c
index c2d30765bf3d..380a90d3c57e 100644
--- a/drivers/cpufreq/imx6q-cpufreq.c
+++ b/drivers/cpufreq/imx6q-cpufreq.c
@@ -31,6 +31,7 @@ static struct clk *step_clk;
 static struct clk *pll2_pfd2_396m_clk;
 
 static struct device *cpu_dev;
+static bool free_opp;
 static struct cpufreq_frequency_table *freq_table;
 static unsigned int transition_latency;
 
@@ -207,11 +208,14 @@ static int imx6q_cpufreq_probe(struct platform_device *pdev)
 			goto put_reg;
 		}
 
+		/* Because we have added the OPPs here, we must free them */
+		free_opp = true;
+
 		num = dev_pm_opp_get_opp_count(cpu_dev);
 		if (num < 0) {
 			ret = num;
 			dev_err(cpu_dev, "no OPP table is found: %d\n", ret);
-			goto put_reg;
+			goto out_free_opp;
 		}
 	}
 
@@ -306,6 +310,9 @@ soc_opp_out:
 
 free_freq_table:
 	dev_pm_opp_free_cpufreq_table(cpu_dev, &freq_table);
+out_free_opp:
+	if (free_opp)
+		of_free_opp_table(cpu_dev);
 put_reg:
 	if (!IS_ERR(arm_reg))
 		regulator_put(arm_reg);
@@ -332,6 +339,8 @@ static int imx6q_cpufreq_remove(struct platform_device *pdev)
 {
 	cpufreq_unregister_driver(&imx6q_cpufreq_driver);
 	dev_pm_opp_free_cpufreq_table(cpu_dev, &freq_table);
+	if (free_opp)
+		of_free_opp_table(cpu_dev);
 	regulator_put(arm_reg);
 	if (!IS_ERR(pu_reg))
 		regulator_put(pu_reg);
@@ -348,7 +357,6 @@ static int imx6q_cpufreq_remove(struct platform_device *pdev)
 static struct platform_driver imx6q_cpufreq_platdrv = {
 	.driver = {
 		.name	= "imx6q-cpufreq",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= imx6q_cpufreq_probe,
 	.remove		= imx6q_cpufreq_remove,
diff --git a/drivers/cpufreq/integrator-cpufreq.c b/drivers/cpufreq/integrator-cpufreq.c
index 6bd69adc3c5e..129e266f7621 100644
--- a/drivers/cpufreq/integrator-cpufreq.c
+++ b/drivers/cpufreq/integrator-cpufreq.c
@@ -226,7 +226,6 @@ static const struct of_device_id integrator_cpufreq_match[] = {
 static struct platform_driver integrator_cpufreq_driver = {
 	.driver = {
 		.name = "integrator-cpufreq",
-		.owner = THIS_MODULE,
 		.of_match_table = integrator_cpufreq_match,
 	},
 	.remove = __exit_p(integrator_cpufreq_remove),
diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c
index 0668b389c516..742eefba12c2 100644
--- a/drivers/cpufreq/intel_pstate.c
+++ b/drivers/cpufreq/intel_pstate.c
@@ -52,6 +52,17 @@ static inline int32_t div_fp(int32_t x, int32_t y)
 	return div_s64((int64_t)x << FRAC_BITS, y);
 }
 
+static inline int ceiling_fp(int32_t x)
+{
+	int mask, ret;
+
+	ret = fp_toint(x);
+	mask = (1 << FRAC_BITS) - 1;
+	if (x & mask)
+		ret += 1;
+	return ret;
+}
+
 struct sample {
 	int32_t core_pct_busy;
 	u64 aperf;
@@ -64,6 +75,7 @@ struct pstate_data {
 	int	current_pstate;
 	int	min_pstate;
 	int	max_pstate;
+	int	scaling;
 	int	turbo_pstate;
 };
 
@@ -113,6 +125,7 @@ struct pstate_funcs {
 	int (*get_max)(void);
 	int (*get_min)(void);
 	int (*get_turbo)(void);
+	int (*get_scaling)(void);
 	void (*set)(struct cpudata*, int pstate);
 	void (*get_vid)(struct cpudata *);
 };
@@ -124,6 +137,7 @@ struct cpu_defaults {
 
 static struct pstate_adjust_policy pid_params;
 static struct pstate_funcs pstate_funcs;
+static int hwp_active;
 
 struct perf_limits {
 	int no_turbo;
@@ -138,6 +152,7 @@ struct perf_limits {
 
 static struct perf_limits limits = {
 	.no_turbo = 0,
+	.turbo_disabled = 0,
 	.max_perf_pct = 100,
 	.max_perf = int_tofp(1),
 	.min_perf_pct = 0,
@@ -184,7 +199,14 @@ static signed int pid_calc(struct _pid *pid, int32_t busy)
 
 	pid->integral += fp_error;
 
-	/* limit the integral term */
+	/*
+	 * We limit the integral here so that it will never
+	 * get higher than 30.  This prevents it from becoming
+	 * too large an input over long periods of time and allows
+	 * it to get factored out sooner.
+	 *
+	 * The value of 30 was chosen through experimentation.
+	 */
 	integral_limit = int_tofp(30);
 	if (pid->integral > integral_limit)
 		pid->integral = integral_limit;
@@ -218,6 +240,46 @@ static inline void intel_pstate_reset_all_pid(void)
 	}
 }
 
+static inline void update_turbo_state(void)
+{
+	u64 misc_en;
+	struct cpudata *cpu;
+
+	cpu = all_cpu_data[0];
+	rdmsrl(MSR_IA32_MISC_ENABLE, misc_en);
+	limits.turbo_disabled =
+		(misc_en & MSR_IA32_MISC_ENABLE_TURBO_DISABLE ||
+		 cpu->pstate.max_pstate == cpu->pstate.turbo_pstate);
+}
+
+#define PCT_TO_HWP(x) (x * 255 / 100)
+static void intel_pstate_hwp_set(void)
+{
+	int min, max, cpu;
+	u64 value, freq;
+
+	get_online_cpus();
+
+	for_each_online_cpu(cpu) {
+		rdmsrl_on_cpu(cpu, MSR_HWP_REQUEST, &value);
+		min = PCT_TO_HWP(limits.min_perf_pct);
+		value &= ~HWP_MIN_PERF(~0L);
+		value |= HWP_MIN_PERF(min);
+
+		max = PCT_TO_HWP(limits.max_perf_pct);
+		if (limits.no_turbo) {
+			rdmsrl( MSR_HWP_CAPABILITIES, freq);
+			max = HWP_GUARANTEED_PERF(freq);
+		}
+
+		value &= ~HWP_MAX_PERF(~0L);
+		value |= HWP_MAX_PERF(max);
+		wrmsrl_on_cpu(cpu, MSR_HWP_REQUEST, value);
+	}
+
+	put_online_cpus();
+}
+
 /************************** debugfs begin ************************/
 static int pid_param_set(void *data, u64 val)
 {
@@ -253,6 +315,8 @@ static void __init intel_pstate_debug_expose_params(void)
 	struct dentry *debugfs_parent;
 	int i = 0;
 
+	if (hwp_active)
+		return;
 	debugfs_parent = debugfs_create_dir("pstate_snb", NULL);
 	if (IS_ERR_OR_NULL(debugfs_parent))
 		return;
@@ -274,6 +338,20 @@ static void __init intel_pstate_debug_expose_params(void)
 		return sprintf(buf, "%u\n", limits.object);		\
 	}
 
+static ssize_t show_no_turbo(struct kobject *kobj,
+			     struct attribute *attr, char *buf)
+{
+	ssize_t ret;
+
+	update_turbo_state();
+	if (limits.turbo_disabled)
+		ret = sprintf(buf, "%u\n", limits.turbo_disabled);
+	else
+		ret = sprintf(buf, "%u\n", limits.no_turbo);
+
+	return ret;
+}
+
 static ssize_t store_no_turbo(struct kobject *a, struct attribute *b,
 			      const char *buf, size_t count)
 {
@@ -283,11 +361,18 @@ static ssize_t store_no_turbo(struct kobject *a, struct attribute *b,
 	ret = sscanf(buf, "%u", &input);
 	if (ret != 1)
 		return -EINVAL;
-	limits.no_turbo = clamp_t(int, input, 0 , 1);
+
+	update_turbo_state();
 	if (limits.turbo_disabled) {
 		pr_warn("Turbo disabled by BIOS or unavailable on processor\n");
-		limits.no_turbo = limits.turbo_disabled;
+		return -EPERM;
 	}
+
+	limits.no_turbo = clamp_t(int, input, 0, 1);
+
+	if (hwp_active)
+		intel_pstate_hwp_set();
+
 	return count;
 }
 
@@ -305,6 +390,8 @@ static ssize_t store_max_perf_pct(struct kobject *a, struct attribute *b,
 	limits.max_perf_pct = min(limits.max_policy_pct, limits.max_sysfs_pct);
 	limits.max_perf = div_fp(int_tofp(limits.max_perf_pct), int_tofp(100));
 
+	if (hwp_active)
+		intel_pstate_hwp_set();
 	return count;
 }
 
@@ -320,10 +407,11 @@ static ssize_t store_min_perf_pct(struct kobject *a, struct attribute *b,
 	limits.min_perf_pct = clamp_t(int, input, 0 , 100);
 	limits.min_perf = div_fp(int_tofp(limits.min_perf_pct), int_tofp(100));
 
+	if (hwp_active)
+		intel_pstate_hwp_set();
 	return count;
 }
 
-show_one(no_turbo, no_turbo);
 show_one(max_perf_pct, max_perf_pct);
 show_one(min_perf_pct, min_perf_pct);
 
@@ -353,8 +441,16 @@ static void __init intel_pstate_sysfs_expose_params(void)
 	rc = sysfs_create_group(intel_pstate_kobject, &intel_pstate_attr_group);
 	BUG_ON(rc);
 }
-
 /************************** sysfs end ************************/
+
+static void intel_pstate_hwp_enable(void)
+{
+	hwp_active++;
+	pr_info("intel_pstate HWP enabled\n");
+
+	wrmsrl( MSR_PM_ENABLE, 0x1);
+}
+
 static int byt_get_min_pstate(void)
 {
 	u64 value;
@@ -394,7 +490,7 @@ static void byt_set_pstate(struct cpudata *cpudata, int pstate)
 		cpudata->vid.ratio);
 
 	vid_fp = clamp_t(int32_t, vid_fp, cpudata->vid.min, cpudata->vid.max);
-	vid = fp_toint(vid_fp);
+	vid = ceiling_fp(vid_fp);
 
 	if (pstate > cpudata->pstate.max_pstate)
 		vid = cpudata->vid.turbo;
@@ -404,6 +500,22 @@ static void byt_set_pstate(struct cpudata *cpudata, int pstate)
 	wrmsrl(MSR_IA32_PERF_CTL, val);
 }
 
+#define BYT_BCLK_FREQS 5
+static int byt_freq_table[BYT_BCLK_FREQS] = { 833, 1000, 1333, 1167, 800};
+
+static int byt_get_scaling(void)
+{
+	u64 value;
+	int i;
+
+	rdmsrl(MSR_FSB_FREQ, value);
+	i = value & 0x3;
+
+	BUG_ON(i > BYT_BCLK_FREQS);
+
+	return byt_freq_table[i] * 100;
+}
+
 static void byt_get_vid(struct cpudata *cpudata)
 {
 	u64 value;
@@ -449,6 +561,11 @@ static int core_get_turbo_pstate(void)
 	return ret;
 }
 
+static inline int core_get_scaling(void)
+{
+	return 100000;
+}
+
 static void core_set_pstate(struct cpudata *cpudata, int pstate)
 {
 	u64 val;
@@ -473,6 +590,7 @@ static struct cpu_defaults core_params = {
 		.get_max = core_get_max_pstate,
 		.get_min = core_get_min_pstate,
 		.get_turbo = core_get_turbo_pstate,
+		.get_scaling = core_get_scaling,
 		.set = core_set_pstate,
 	},
 };
@@ -491,6 +609,7 @@ static struct cpu_defaults byt_params = {
 		.get_min = byt_get_min_pstate,
 		.get_turbo = byt_get_turbo_pstate,
 		.set = byt_set_pstate,
+		.get_scaling = byt_get_scaling,
 		.get_vid = byt_get_vid,
 	},
 };
@@ -501,9 +620,14 @@ static void intel_pstate_get_min_max(struct cpudata *cpu, int *min, int *max)
 	int max_perf_adj;
 	int min_perf;
 
-	if (limits.no_turbo)
+	if (limits.no_turbo || limits.turbo_disabled)
 		max_perf = cpu->pstate.max_pstate;
 
+	/*
+	 * performance can be limited by user through sysfs, by cpufreq
+	 * policy, or by cpu specific default values determined through
+	 * experimentation.
+	 */
 	max_perf_adj = fp_toint(mul_fp(int_tofp(max_perf), limits.max_perf));
 	*max = clamp_t(int, max_perf_adj,
 			cpu->pstate.min_pstate, cpu->pstate.turbo_pstate);
@@ -516,6 +640,8 @@ static void intel_pstate_set_pstate(struct cpudata *cpu, int pstate)
 {
 	int max_perf, min_perf;
 
+	update_turbo_state();
+
 	intel_pstate_get_min_max(cpu, &min_perf, &max_perf);
 
 	pstate = clamp_t(int, pstate, min_perf, max_perf);
@@ -523,7 +649,7 @@ static void intel_pstate_set_pstate(struct cpudata *cpu, int pstate)
 	if (pstate == cpu->pstate.current_pstate)
 		return;
 
-	trace_cpu_frequency(pstate * 100000, cpu->cpu);
+	trace_cpu_frequency(pstate * cpu->pstate.scaling, cpu->cpu);
 
 	cpu->pstate.current_pstate = pstate;
 
@@ -535,6 +661,7 @@ static void intel_pstate_get_cpu_pstates(struct cpudata *cpu)
 	cpu->pstate.min_pstate = pstate_funcs.get_min();
 	cpu->pstate.max_pstate = pstate_funcs.get_max();
 	cpu->pstate.turbo_pstate = pstate_funcs.get_turbo();
+	cpu->pstate.scaling = pstate_funcs.get_scaling();
 
 	if (pstate_funcs.get_vid)
 		pstate_funcs.get_vid(cpu);
@@ -550,7 +677,9 @@ static inline void intel_pstate_calc_busy(struct cpudata *cpu)
 	core_pct = div64_u64(core_pct, int_tofp(sample->mperf));
 
 	sample->freq = fp_toint(
-		mul_fp(int_tofp(cpu->pstate.max_pstate * 1000), core_pct));
+		mul_fp(int_tofp(
+			cpu->pstate.max_pstate * cpu->pstate.scaling / 100),
+			core_pct));
 
 	sample->core_pct_busy = (int32_t)core_pct;
 }
@@ -578,6 +707,14 @@ static inline void intel_pstate_sample(struct cpudata *cpu)
 	cpu->prev_mperf = mperf;
 }
 
+static inline void intel_hwp_set_sample_time(struct cpudata *cpu)
+{
+	int delay;
+
+	delay = msecs_to_jiffies(50);
+	mod_timer_pinned(&cpu->timer, jiffies + delay);
+}
+
 static inline void intel_pstate_set_sample_time(struct cpudata *cpu)
 {
 	int delay;
@@ -592,11 +729,29 @@ static inline int32_t intel_pstate_get_scaled_busy(struct cpudata *cpu)
 	u32 duration_us;
 	u32 sample_time;
 
+	/*
+	 * core_busy is the ratio of actual performance to max
+	 * max_pstate is the max non turbo pstate available
+	 * current_pstate was the pstate that was requested during
+	 * 	the last sample period.
+	 *
+	 * We normalize core_busy, which was our actual percent
+	 * performance to what we requested during the last sample
+	 * period. The result will be a percentage of busy at a
+	 * specified pstate.
+	 */
 	core_busy = cpu->sample.core_pct_busy;
 	max_pstate = int_tofp(cpu->pstate.max_pstate);
 	current_pstate = int_tofp(cpu->pstate.current_pstate);
 	core_busy = mul_fp(core_busy, div_fp(max_pstate, current_pstate));
 
+	/*
+	 * Since we have a deferred timer, it will not fire unless
+	 * we are in C0.  So, determine if the actual elapsed time
+	 * is significantly greater (3x) than our sample interval.  If it
+	 * is, then we were idle for a long enough period of time
+	 * to adjust our busyness.
+	 */
 	sample_time = pid_params.sample_rate_ms  * USEC_PER_MSEC;
 	duration_us = (u32) ktime_us_delta(cpu->sample.time,
 					   cpu->last_sample_time);
@@ -624,6 +779,14 @@ static inline void intel_pstate_adjust_busy_pstate(struct cpudata *cpu)
 	intel_pstate_set_pstate(cpu, cpu->pstate.current_pstate - ctl);
 }
 
+static void intel_hwp_timer_func(unsigned long __data)
+{
+	struct cpudata *cpu = (struct cpudata *) __data;
+
+	intel_pstate_sample(cpu);
+	intel_hwp_set_sample_time(cpu);
+}
+
 static void intel_pstate_timer_func(unsigned long __data)
 {
 	struct cpudata *cpu = (struct cpudata *) __data;
@@ -660,6 +823,7 @@ static const struct x86_cpu_id intel_pstate_cpu_ids[] = {
 	ICPU(0x3f, core_params),
 	ICPU(0x45, core_params),
 	ICPU(0x46, core_params),
+	ICPU(0x47, core_params),
 	ICPU(0x4c, byt_params),
 	ICPU(0x4f, core_params),
 	ICPU(0x56, core_params),
@@ -667,11 +831,18 @@ static const struct x86_cpu_id intel_pstate_cpu_ids[] = {
 };
 MODULE_DEVICE_TABLE(x86cpu, intel_pstate_cpu_ids);
 
+static const struct x86_cpu_id intel_pstate_cpu_oob_ids[] = {
+	ICPU(0x56, core_params),
+	{}
+};
+
 static int intel_pstate_init_cpu(unsigned int cpunum)
 {
 	struct cpudata *cpu;
 
-	all_cpu_data[cpunum] = kzalloc(sizeof(struct cpudata), GFP_KERNEL);
+	if (!all_cpu_data[cpunum])
+		all_cpu_data[cpunum] = kzalloc(sizeof(struct cpudata),
+					       GFP_KERNEL);
 	if (!all_cpu_data[cpunum])
 		return -ENOMEM;
 
@@ -681,9 +852,14 @@ static int intel_pstate_init_cpu(unsigned int cpunum)
 	intel_pstate_get_cpu_pstates(cpu);
 
 	init_timer_deferrable(&cpu->timer);
-	cpu->timer.function = intel_pstate_timer_func;
 	cpu->timer.data = (unsigned long)cpu;
 	cpu->timer.expires = jiffies + HZ/100;
+
+	if (!hwp_active)
+		cpu->timer.function = intel_pstate_timer_func;
+	else
+		cpu->timer.function = intel_hwp_timer_func;
+
 	intel_pstate_busy_pid_reset(cpu);
 	intel_pstate_sample(cpu);
 
@@ -714,11 +890,13 @@ static int intel_pstate_set_policy(struct cpufreq_policy *policy)
 	if (policy->policy == CPUFREQ_POLICY_PERFORMANCE) {
 		limits.min_perf_pct = 100;
 		limits.min_perf = int_tofp(1);
+		limits.max_policy_pct = 100;
 		limits.max_perf_pct = 100;
 		limits.max_perf = int_tofp(1);
-		limits.no_turbo = limits.turbo_disabled;
+		limits.no_turbo = 0;
 		return 0;
 	}
+
 	limits.min_perf_pct = (policy->min * 100) / policy->cpuinfo.max_freq;
 	limits.min_perf_pct = clamp_t(int, limits.min_perf_pct, 0 , 100);
 	limits.min_perf = div_fp(int_tofp(limits.min_perf_pct), int_tofp(100));
@@ -728,6 +906,9 @@ static int intel_pstate_set_policy(struct cpufreq_policy *policy)
 	limits.max_perf_pct = min(limits.max_policy_pct, limits.max_sysfs_pct);
 	limits.max_perf = div_fp(int_tofp(limits.max_perf_pct), int_tofp(100));
 
+	if (hwp_active)
+		intel_pstate_hwp_set();
+
 	return 0;
 }
 
@@ -750,16 +931,16 @@ static void intel_pstate_stop_cpu(struct cpufreq_policy *policy)
 	pr_info("intel_pstate CPU %d exiting\n", cpu_num);
 
 	del_timer_sync(&all_cpu_data[cpu_num]->timer);
+	if (hwp_active)
+		return;
+
 	intel_pstate_set_pstate(cpu, cpu->pstate.min_pstate);
-	kfree(all_cpu_data[cpu_num]);
-	all_cpu_data[cpu_num] = NULL;
 }
 
 static int intel_pstate_cpu_init(struct cpufreq_policy *policy)
 {
 	struct cpudata *cpu;
 	int rc;
-	u64 misc_en;
 
 	rc = intel_pstate_init_cpu(policy->cpu);
 	if (rc)
@@ -767,23 +948,18 @@ static int intel_pstate_cpu_init(struct cpufreq_policy *policy)
 
 	cpu = all_cpu_data[policy->cpu];
 
-	rdmsrl(MSR_IA32_MISC_ENABLE, misc_en);
-	if (misc_en & MSR_IA32_MISC_ENABLE_TURBO_DISABLE ||
-	    cpu->pstate.max_pstate == cpu->pstate.turbo_pstate) {
-		limits.turbo_disabled = 1;
-		limits.no_turbo = 1;
-	}
 	if (limits.min_perf_pct == 100 && limits.max_perf_pct == 100)
 		policy->policy = CPUFREQ_POLICY_PERFORMANCE;
 	else
 		policy->policy = CPUFREQ_POLICY_POWERSAVE;
 
-	policy->min = cpu->pstate.min_pstate * 100000;
-	policy->max = cpu->pstate.turbo_pstate * 100000;
+	policy->min = cpu->pstate.min_pstate * cpu->pstate.scaling;
+	policy->max = cpu->pstate.turbo_pstate * cpu->pstate.scaling;
 
 	/* cpuinfo and default policy values */
-	policy->cpuinfo.min_freq = cpu->pstate.min_pstate * 100000;
-	policy->cpuinfo.max_freq = cpu->pstate.turbo_pstate * 100000;
+	policy->cpuinfo.min_freq = cpu->pstate.min_pstate * cpu->pstate.scaling;
+	policy->cpuinfo.max_freq =
+		cpu->pstate.turbo_pstate * cpu->pstate.scaling;
 	policy->cpuinfo.transition_latency = CPUFREQ_ETERNAL;
 	cpumask_set_cpu(policy->cpu, policy->cpus);
 
@@ -801,6 +977,8 @@ static struct cpufreq_driver intel_pstate_driver = {
 };
 
 static int __initdata no_load;
+static int __initdata no_hwp;
+static unsigned int force_load;
 
 static int intel_pstate_msrs_not_valid(void)
 {
@@ -841,6 +1019,7 @@ static void copy_cpu_funcs(struct pstate_funcs *funcs)
 	pstate_funcs.get_max   = funcs->get_max;
 	pstate_funcs.get_min   = funcs->get_min;
 	pstate_funcs.get_turbo = funcs->get_turbo;
+	pstate_funcs.get_scaling = funcs->get_scaling;
 	pstate_funcs.set       = funcs->set;
 	pstate_funcs.get_vid   = funcs->get_vid;
 }
@@ -877,15 +1056,46 @@ static bool intel_pstate_no_acpi_pss(void)
 	return true;
 }
 
+static bool intel_pstate_has_acpi_ppc(void)
+{
+	int i;
+
+	for_each_possible_cpu(i) {
+		struct acpi_processor *pr = per_cpu(processors, i);
+
+		if (!pr)
+			continue;
+		if (acpi_has_method(pr->handle, "_PPC"))
+			return true;
+	}
+	return false;
+}
+
+enum {
+	PSS,
+	PPC,
+};
+
 struct hw_vendor_info {
 	u16  valid;
 	char oem_id[ACPI_OEM_ID_SIZE];
 	char oem_table_id[ACPI_OEM_TABLE_ID_SIZE];
+	int  oem_pwr_table;
 };
 
 /* Hardware vendor-specific info that has its own power management modes */
 static struct hw_vendor_info vendor_info[] = {
-	{1, "HP    ", "ProLiant"},
+	{1, "HP    ", "ProLiant", PSS},
+	{1, "ORACLE", "X4-2    ", PPC},
+	{1, "ORACLE", "X4-2L   ", PPC},
+	{1, "ORACLE", "X4-2B   ", PPC},
+	{1, "ORACLE", "X3-2    ", PPC},
+	{1, "ORACLE", "X3-2L   ", PPC},
+	{1, "ORACLE", "X3-2B   ", PPC},
+	{1, "ORACLE", "X4470M2 ", PPC},
+	{1, "ORACLE", "X4270M3 ", PPC},
+	{1, "ORACLE", "X4270M2 ", PPC},
+	{1, "ORACLE", "X4170M2 ", PPC},
 	{0, "", ""},
 };
 
@@ -893,6 +1103,15 @@ static bool intel_pstate_platform_pwr_mgmt_exists(void)
 {
 	struct acpi_table_header hdr;
 	struct hw_vendor_info *v_info;
+	const struct x86_cpu_id *id;
+	u64 misc_pwr;
+
+	id = x86_match_cpu(intel_pstate_cpu_oob_ids);
+	if (id) {
+		rdmsrl(MSR_MISC_PWR_MGMT, misc_pwr);
+		if ( misc_pwr & (1 << 8))
+			return true;
+	}
 
 	if (acpi_disabled ||
 	    ACPI_FAILURE(acpi_get_table_header(ACPI_SIG_FADT, 0, &hdr)))
@@ -900,15 +1119,22 @@ static bool intel_pstate_platform_pwr_mgmt_exists(void)
 
 	for (v_info = vendor_info; v_info->valid; v_info++) {
 		if (!strncmp(hdr.oem_id, v_info->oem_id, ACPI_OEM_ID_SIZE) &&
-		    !strncmp(hdr.oem_table_id, v_info->oem_table_id, ACPI_OEM_TABLE_ID_SIZE) &&
-		    intel_pstate_no_acpi_pss())
-			return true;
+			!strncmp(hdr.oem_table_id, v_info->oem_table_id,
+						ACPI_OEM_TABLE_ID_SIZE))
+			switch (v_info->oem_pwr_table) {
+			case PSS:
+				return intel_pstate_no_acpi_pss();
+			case PPC:
+				return intel_pstate_has_acpi_ppc() &&
+					(!force_load);
+			}
 	}
 
 	return false;
 }
 #else /* CONFIG_ACPI not enabled */
 static inline bool intel_pstate_platform_pwr_mgmt_exists(void) { return false; }
+static inline bool intel_pstate_has_acpi_ppc(void) { return false; }
 #endif /* CONFIG_ACPI */
 
 static int __init intel_pstate_init(void)
@@ -916,6 +1142,7 @@ static int __init intel_pstate_init(void)
 	int cpu, rc = 0;
 	const struct x86_cpu_id *id;
 	struct cpu_defaults *cpu_info;
+	struct cpuinfo_x86 *c = &boot_cpu_data;
 
 	if (no_load)
 		return -ENODEV;
@@ -945,6 +1172,9 @@ static int __init intel_pstate_init(void)
 	if (!all_cpu_data)
 		return -ENOMEM;
 
+	if (cpu_has(c,X86_FEATURE_HWP) && !no_hwp)
+		intel_pstate_hwp_enable();
+
 	rc = cpufreq_register_driver(&intel_pstate_driver);
 	if (rc)
 		goto out;
@@ -975,6 +1205,10 @@ static int __init intel_pstate_setup(char *str)
 
 	if (!strcmp(str, "disable"))
 		no_load = 1;
+	if (!strcmp(str, "no_hwp"))
+		no_hwp = 1;
+	if (!strcmp(str, "force"))
+		force_load = 1;
 	return 0;
 }
 early_param("intel_pstate", intel_pstate_setup);
diff --git a/drivers/cpufreq/kirkwood-cpufreq.c b/drivers/cpufreq/kirkwood-cpufreq.c
index 37a480680cd0..be42f103db60 100644
--- a/drivers/cpufreq/kirkwood-cpufreq.c
+++ b/drivers/cpufreq/kirkwood-cpufreq.c
@@ -12,7 +12,6 @@
 #include <linux/kernel.h>
 #include <linux/module.h>
 #include <linux/clk.h>
-#include <linux/clk-provider.h>
 #include <linux/cpufreq.h>
 #include <linux/of_device.h>
 #include <linux/platform_device.h>
@@ -39,8 +38,7 @@ static struct priv
  * - cpu clk
  * - ddr clk
  *
- * The frequencies are set at runtime before registering this *
- * table.
+ * The frequencies are set at runtime before registering this table.
  */
 static struct cpufreq_frequency_table kirkwood_freq_table[] = {
 	{0, STATE_CPU_FREQ,	0}, /* CPU uses cpuclk */
@@ -50,9 +48,7 @@ static struct cpufreq_frequency_table kirkwood_freq_table[] = {
 
 static unsigned int kirkwood_cpufreq_get_cpu_frequency(unsigned int cpu)
 {
-	if (__clk_is_enabled(priv.powersave_clk))
-		return kirkwood_freq_table[1].frequency;
-	return kirkwood_freq_table[0].frequency;
+	return clk_get_rate(priv.powersave_clk) / 1000;
 }
 
 static int kirkwood_cpufreq_target(struct cpufreq_policy *policy,
@@ -70,10 +66,10 @@ static int kirkwood_cpufreq_target(struct cpufreq_policy *policy,
 
 	switch (state) {
 	case STATE_CPU_FREQ:
-		clk_disable(priv.powersave_clk);
+		clk_set_parent(priv.powersave_clk, priv.cpu_clk);
 		break;
 	case STATE_DDR_FREQ:
-		clk_enable(priv.powersave_clk);
+		clk_set_parent(priv.powersave_clk, priv.ddr_clk);
 		break;
 	}
 
@@ -150,7 +146,7 @@ static int kirkwood_cpufreq_probe(struct platform_device *pdev)
 		err = PTR_ERR(priv.powersave_clk);
 		goto out_ddr;
 	}
-	clk_prepare(priv.powersave_clk);
+	clk_prepare_enable(priv.powersave_clk);
 
 	of_node_put(np);
 	np = NULL;
@@ -187,7 +183,6 @@ static struct platform_driver kirkwood_cpufreq_platform_driver = {
 	.remove = kirkwood_cpufreq_remove,
 	.driver = {
 		.name = "kirkwood-cpufreq",
-		.owner = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/cpufreq/longhaul.c b/drivers/cpufreq/longhaul.c
index c913906a719e..0f6b229afcb9 100644
--- a/drivers/cpufreq/longhaul.c
+++ b/drivers/cpufreq/longhaul.c
@@ -1,5 +1,5 @@
 /*
- *  (C) 2001-2004  Dave Jones. <davej@redhat.com>
+ *  (C) 2001-2004  Dave Jones.
  *  (C) 2002  Padraig Brady. <padraig@antefacto.com>
  *
  *  Licensed under the terms of the GNU GPL License version 2.
@@ -1008,7 +1008,7 @@ MODULE_PARM_DESC(revid_errata, "Ignore CPU Revision ID");
 module_param(enable, int, 0644);
 MODULE_PARM_DESC(enable, "Enable driver");
 
-MODULE_AUTHOR("Dave Jones <davej@redhat.com>");
+MODULE_AUTHOR("Dave Jones");
 MODULE_DESCRIPTION("Longhaul driver for VIA Cyrix processors.");
 MODULE_LICENSE("GPL");
 
diff --git a/drivers/cpufreq/loongson2_cpufreq.c b/drivers/cpufreq/loongson2_cpufreq.c
index 9fa177206032..fc897babab55 100644
--- a/drivers/cpufreq/loongson2_cpufreq.c
+++ b/drivers/cpufreq/loongson2_cpufreq.c
@@ -130,7 +130,6 @@ MODULE_DEVICE_TABLE(platform, platform_device_ids);
 static struct platform_driver platform_driver = {
 	.driver = {
 		.name = "loongson2_cpufreq",
-		.owner = THIS_MODULE,
 	},
 	.id_table = platform_device_ids,
 };
diff --git a/drivers/cpufreq/ls1x-cpufreq.c b/drivers/cpufreq/ls1x-cpufreq.c
new file mode 100644
index 000000000000..25fbd6a1374f
--- /dev/null
+++ b/drivers/cpufreq/ls1x-cpufreq.c
@@ -0,0 +1,223 @@
+/*
+ * CPU Frequency Scaling for Loongson 1 SoC
+ *
+ * Copyright (C) 2014 Zhang, Keguang <keguang.zhang@gmail.com>
+ *
+ * This file is licensed under the terms of the GNU General Public
+ * License version 2. This program is licensed "as is" without any
+ * warranty of any kind, whether express or implied.
+ */
+
+#include <linux/clk.h>
+#include <linux/clk-provider.h>
+#include <linux/cpu.h>
+#include <linux/cpufreq.h>
+#include <linux/delay.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+
+#include <asm/mach-loongson1/cpufreq.h>
+#include <asm/mach-loongson1/loongson1.h>
+
+static struct {
+	struct device *dev;
+	struct clk *clk;	/* CPU clk */
+	struct clk *mux_clk;	/* MUX of CPU clk */
+	struct clk *pll_clk;	/* PLL clk */
+	struct clk *osc_clk;	/* OSC clk */
+	unsigned int max_freq;
+	unsigned int min_freq;
+} ls1x_cpufreq;
+
+static int ls1x_cpufreq_notifier(struct notifier_block *nb,
+				 unsigned long val, void *data)
+{
+	if (val == CPUFREQ_POSTCHANGE)
+		current_cpu_data.udelay_val = loops_per_jiffy;
+
+	return NOTIFY_OK;
+}
+
+static struct notifier_block ls1x_cpufreq_notifier_block = {
+	.notifier_call = ls1x_cpufreq_notifier
+};
+
+static int ls1x_cpufreq_target(struct cpufreq_policy *policy,
+			       unsigned int index)
+{
+	unsigned int old_freq, new_freq;
+
+	old_freq = policy->cur;
+	new_freq = policy->freq_table[index].frequency;
+
+	/*
+	 * The procedure of reconfiguring CPU clk is as below.
+	 *
+	 *  - Reparent CPU clk to OSC clk
+	 *  - Reset CPU clock (very important)
+	 *  - Reconfigure CPU DIV
+	 *  - Reparent CPU clk back to CPU DIV clk
+	 */
+
+	dev_dbg(ls1x_cpufreq.dev, "%u KHz --> %u KHz\n", old_freq, new_freq);
+	clk_set_parent(policy->clk, ls1x_cpufreq.osc_clk);
+	__raw_writel(__raw_readl(LS1X_CLK_PLL_DIV) | RST_CPU_EN | RST_CPU,
+		     LS1X_CLK_PLL_DIV);
+	__raw_writel(__raw_readl(LS1X_CLK_PLL_DIV) & ~(RST_CPU_EN | RST_CPU),
+		     LS1X_CLK_PLL_DIV);
+	clk_set_rate(ls1x_cpufreq.mux_clk, new_freq * 1000);
+	clk_set_parent(policy->clk, ls1x_cpufreq.mux_clk);
+
+	return 0;
+}
+
+static int ls1x_cpufreq_init(struct cpufreq_policy *policy)
+{
+	struct cpufreq_frequency_table *freq_tbl;
+	unsigned int pll_freq, freq;
+	int steps, i, ret;
+
+	pll_freq = clk_get_rate(ls1x_cpufreq.pll_clk) / 1000;
+
+	steps = 1 << DIV_CPU_WIDTH;
+	freq_tbl = kzalloc(sizeof(*freq_tbl) * steps, GFP_KERNEL);
+	if (!freq_tbl) {
+		dev_err(ls1x_cpufreq.dev,
+			"failed to alloc cpufreq_frequency_table\n");
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	for (i = 0; i < (steps - 1); i++) {
+		freq = pll_freq / (i + 1);
+		if ((freq < ls1x_cpufreq.min_freq) ||
+		    (freq > ls1x_cpufreq.max_freq))
+			freq_tbl[i].frequency = CPUFREQ_ENTRY_INVALID;
+		else
+			freq_tbl[i].frequency = freq;
+		dev_dbg(ls1x_cpufreq.dev,
+			"cpufreq table: index %d: frequency %d\n", i,
+			freq_tbl[i].frequency);
+	}
+	freq_tbl[i].frequency = CPUFREQ_TABLE_END;
+
+	policy->clk = ls1x_cpufreq.clk;
+	ret = cpufreq_generic_init(policy, freq_tbl, 0);
+	if (ret)
+		kfree(freq_tbl);
+out:
+	return ret;
+}
+
+static int ls1x_cpufreq_exit(struct cpufreq_policy *policy)
+{
+	kfree(policy->freq_table);
+	return 0;
+}
+
+static struct cpufreq_driver ls1x_cpufreq_driver = {
+	.name		= "cpufreq-ls1x",
+	.flags		= CPUFREQ_STICKY | CPUFREQ_NEED_INITIAL_FREQ_CHECK,
+	.verify		= cpufreq_generic_frequency_table_verify,
+	.target_index	= ls1x_cpufreq_target,
+	.get		= cpufreq_generic_get,
+	.init		= ls1x_cpufreq_init,
+	.exit		= ls1x_cpufreq_exit,
+	.attr		= cpufreq_generic_attr,
+};
+
+static int ls1x_cpufreq_remove(struct platform_device *pdev)
+{
+	cpufreq_unregister_notifier(&ls1x_cpufreq_notifier_block,
+				    CPUFREQ_TRANSITION_NOTIFIER);
+	cpufreq_unregister_driver(&ls1x_cpufreq_driver);
+
+	return 0;
+}
+
+static int ls1x_cpufreq_probe(struct platform_device *pdev)
+{
+	struct plat_ls1x_cpufreq *pdata = pdev->dev.platform_data;
+	struct clk *clk;
+	int ret;
+
+	if (!pdata || !pdata->clk_name || !pdata->osc_clk_name)
+		return -EINVAL;
+
+	ls1x_cpufreq.dev = &pdev->dev;
+
+	clk = devm_clk_get(&pdev->dev, pdata->clk_name);
+	if (IS_ERR(clk)) {
+		dev_err(ls1x_cpufreq.dev, "unable to get %s clock\n",
+			pdata->clk_name);
+		ret = PTR_ERR(clk);
+		goto out;
+	}
+	ls1x_cpufreq.clk = clk;
+
+	clk = clk_get_parent(clk);
+	if (IS_ERR(clk)) {
+		dev_err(ls1x_cpufreq.dev, "unable to get parent of %s clock\n",
+			__clk_get_name(ls1x_cpufreq.clk));
+		ret = PTR_ERR(clk);
+		goto out;
+	}
+	ls1x_cpufreq.mux_clk = clk;
+
+	clk = clk_get_parent(clk);
+	if (IS_ERR(clk)) {
+		dev_err(ls1x_cpufreq.dev, "unable to get parent of %s clock\n",
+			__clk_get_name(ls1x_cpufreq.mux_clk));
+		ret = PTR_ERR(clk);
+		goto out;
+	}
+	ls1x_cpufreq.pll_clk = clk;
+
+	clk = devm_clk_get(&pdev->dev, pdata->osc_clk_name);
+	if (IS_ERR(clk)) {
+		dev_err(ls1x_cpufreq.dev, "unable to get %s clock\n",
+			pdata->osc_clk_name);
+		ret = PTR_ERR(clk);
+		goto out;
+	}
+	ls1x_cpufreq.osc_clk = clk;
+
+	ls1x_cpufreq.max_freq = pdata->max_freq;
+	ls1x_cpufreq.min_freq = pdata->min_freq;
+
+	ret = cpufreq_register_driver(&ls1x_cpufreq_driver);
+	if (ret) {
+		dev_err(ls1x_cpufreq.dev,
+			"failed to register cpufreq driver: %d\n", ret);
+		goto out;
+	}
+
+	ret = cpufreq_register_notifier(&ls1x_cpufreq_notifier_block,
+					CPUFREQ_TRANSITION_NOTIFIER);
+
+	if (!ret)
+		goto out;
+
+	dev_err(ls1x_cpufreq.dev, "failed to register cpufreq notifier: %d\n",
+		ret);
+
+	cpufreq_unregister_driver(&ls1x_cpufreq_driver);
+out:
+	return ret;
+}
+
+static struct platform_driver ls1x_cpufreq_platdrv = {
+	.driver = {
+		.name	= "ls1x-cpufreq",
+		.owner	= THIS_MODULE,
+	},
+	.probe		= ls1x_cpufreq_probe,
+	.remove		= ls1x_cpufreq_remove,
+};
+
+module_platform_driver(ls1x_cpufreq_platdrv);
+
+MODULE_AUTHOR("Kelvin Cheung <keguang.zhang@gmail.com>");
+MODULE_DESCRIPTION("Loongson 1 CPUFreq driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/cpufreq/omap-cpufreq.c b/drivers/cpufreq/omap-cpufreq.c
index 5f69c9aa703c..e3866e0d5bf8 100644
--- a/drivers/cpufreq/omap-cpufreq.c
+++ b/drivers/cpufreq/omap-cpufreq.c
@@ -195,7 +195,6 @@ static int omap_cpufreq_remove(struct platform_device *pdev)
 static struct platform_driver omap_cpufreq_platdrv = {
 	.driver = {
 		.name	= "omap-cpufreq",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= omap_cpufreq_probe,
 	.remove		= omap_cpufreq_remove,
diff --git a/drivers/cpufreq/pcc-cpufreq.c b/drivers/cpufreq/pcc-cpufreq.c
index 4d2c8e861089..2a0d58959acf 100644
--- a/drivers/cpufreq/pcc-cpufreq.c
+++ b/drivers/cpufreq/pcc-cpufreq.c
@@ -603,6 +603,13 @@ static void __exit pcc_cpufreq_exit(void)
 	free_percpu(pcc_cpu_info);
 }
 
+static const struct acpi_device_id processor_device_ids[] = {
+	{ACPI_PROCESSOR_OBJECT_HID, },
+	{ACPI_PROCESSOR_DEVICE_HID, },
+	{},
+};
+MODULE_DEVICE_TABLE(acpi, processor_device_ids);
+
 MODULE_AUTHOR("Matthew Garrett, Naga Chumbalkar");
 MODULE_VERSION(PCC_VERSION);
 MODULE_DESCRIPTION("Processor Clocking Control interface driver");
diff --git a/drivers/cpufreq/pmac32-cpufreq.c b/drivers/cpufreq/pmac32-cpufreq.c
index 7615180d7ee3..1f49d97a70ea 100644
--- a/drivers/cpufreq/pmac32-cpufreq.c
+++ b/drivers/cpufreq/pmac32-cpufreq.c
@@ -611,7 +611,7 @@ static int __init pmac_cpufreq_setup(void)
 	struct device_node	*cpunode;
 	const u32		*value;
 
-	if (strstr(cmd_line, "nocpufreq"))
+	if (strstr(boot_command_line, "nocpufreq"))
 		return 0;
 
 	/* Get first CPU node */
diff --git a/drivers/cpufreq/powernow-k6.c b/drivers/cpufreq/powernow-k6.c
index f91027259c3c..e6f24b281e3e 100644
--- a/drivers/cpufreq/powernow-k6.c
+++ b/drivers/cpufreq/powernow-k6.c
@@ -300,7 +300,7 @@ static void __exit powernow_k6_exit(void)
 }
 
 
-MODULE_AUTHOR("Arjan van de Ven, Dave Jones <davej@redhat.com>, "
+MODULE_AUTHOR("Arjan van de Ven, Dave Jones, "
 		"Dominik Brodowski <linux@brodo.de>");
 MODULE_DESCRIPTION("PowerNow! driver for AMD K6-2+ / K6-3+ processors.");
 MODULE_LICENSE("GPL");
diff --git a/drivers/cpufreq/powernow-k7.c b/drivers/cpufreq/powernow-k7.c
index e61e224475ad..37c5742482d8 100644
--- a/drivers/cpufreq/powernow-k7.c
+++ b/drivers/cpufreq/powernow-k7.c
@@ -1,7 +1,6 @@
 /*
  *  AMD K7 Powernow driver.
  *  (C) 2003 Dave Jones on behalf of SuSE Labs.
- *  (C) 2003-2004 Dave Jones <davej@redhat.com>
  *
  *  Licensed under the terms of the GNU GPL License version 2.
  *  Based upon datasheets & sample CPUs kindly provided by AMD.
@@ -701,7 +700,7 @@ static void __exit powernow_exit(void)
 module_param(acpi_force,  int, 0444);
 MODULE_PARM_DESC(acpi_force, "Force ACPI to be used.");
 
-MODULE_AUTHOR("Dave Jones <davej@redhat.com>");
+MODULE_AUTHOR("Dave Jones");
 MODULE_DESCRIPTION("Powernow driver for AMD K7 processors.");
 MODULE_LICENSE("GPL");
 
diff --git a/drivers/cpufreq/powernv-cpufreq.c b/drivers/cpufreq/powernv-cpufreq.c
index 379c0837f5a9..2dfd4fdb5a52 100644
--- a/drivers/cpufreq/powernv-cpufreq.c
+++ b/drivers/cpufreq/powernv-cpufreq.c
@@ -26,6 +26,7 @@
 #include <linux/cpufreq.h>
 #include <linux/smp.h>
 #include <linux/of.h>
+#include <linux/reboot.h>
 
 #include <asm/cputhreads.h>
 #include <asm/firmware.h>
@@ -35,6 +36,7 @@
 #define POWERNV_MAX_PSTATES	256
 
 static struct cpufreq_frequency_table powernv_freqs[POWERNV_MAX_PSTATES+1];
+static bool rebooting;
 
 /*
  * Note: The set of pstates consists of contiguous integers, the
@@ -284,6 +286,15 @@ static void set_pstate(void *freq_data)
 }
 
 /*
+ * get_nominal_index: Returns the index corresponding to the nominal
+ * pstate in the cpufreq table
+ */
+static inline unsigned int get_nominal_index(void)
+{
+	return powernv_pstate_info.max - powernv_pstate_info.nominal;
+}
+
+/*
  * powernv_cpufreq_target_index: Sets the frequency corresponding to
  * the cpufreq table entry indexed by new_index on the cpus in the
  * mask policy->cpus
@@ -293,6 +304,9 @@ static int powernv_cpufreq_target_index(struct cpufreq_policy *policy,
 {
 	struct powernv_smp_call_data freq_data;
 
+	if (unlikely(rebooting) && new_index != get_nominal_index())
+		return 0;
+
 	freq_data.pstate_id = powernv_freqs[new_index].driver_data;
 
 	/*
@@ -317,6 +331,33 @@ static int powernv_cpufreq_cpu_init(struct cpufreq_policy *policy)
 	return cpufreq_table_validate_and_show(policy, powernv_freqs);
 }
 
+static int powernv_cpufreq_reboot_notifier(struct notifier_block *nb,
+				unsigned long action, void *unused)
+{
+	int cpu;
+	struct cpufreq_policy cpu_policy;
+
+	rebooting = true;
+	for_each_online_cpu(cpu) {
+		cpufreq_get_policy(&cpu_policy, cpu);
+		powernv_cpufreq_target_index(&cpu_policy, get_nominal_index());
+	}
+
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block powernv_cpufreq_reboot_nb = {
+	.notifier_call = powernv_cpufreq_reboot_notifier,
+};
+
+static void powernv_cpufreq_stop_cpu(struct cpufreq_policy *policy)
+{
+	struct powernv_smp_call_data freq_data;
+
+	freq_data.pstate_id = powernv_pstate_info.min;
+	smp_call_function_single(policy->cpu, set_pstate, &freq_data, 1);
+}
+
 static struct cpufreq_driver powernv_cpufreq_driver = {
 	.name		= "powernv-cpufreq",
 	.flags		= CPUFREQ_CONST_LOOPS,
@@ -324,6 +365,7 @@ static struct cpufreq_driver powernv_cpufreq_driver = {
 	.verify		= cpufreq_generic_frequency_table_verify,
 	.target_index	= powernv_cpufreq_target_index,
 	.get		= powernv_cpufreq_get,
+	.stop_cpu	= powernv_cpufreq_stop_cpu,
 	.attr		= powernv_cpu_freq_attr,
 };
 
@@ -342,12 +384,14 @@ static int __init powernv_cpufreq_init(void)
 		return rc;
 	}
 
+	register_reboot_notifier(&powernv_cpufreq_reboot_nb);
 	return cpufreq_register_driver(&powernv_cpufreq_driver);
 }
 module_init(powernv_cpufreq_init);
 
 static void __exit powernv_cpufreq_exit(void)
 {
+	unregister_reboot_notifier(&powernv_cpufreq_reboot_nb);
 	cpufreq_unregister_driver(&powernv_cpufreq_driver);
 }
 module_exit(powernv_cpufreq_exit);
diff --git a/drivers/cpufreq/ppc-corenet-cpufreq.c b/drivers/cpufreq/ppc-corenet-cpufreq.c
index 3607070797af..bee5df7794d3 100644
--- a/drivers/cpufreq/ppc-corenet-cpufreq.c
+++ b/drivers/cpufreq/ppc-corenet-cpufreq.c
@@ -199,7 +199,6 @@ static int corenet_cpufreq_cpu_init(struct cpufreq_policy *policy)
 	}
 
 	data->table = table;
-	per_cpu(cpu_data, cpu) = data;
 
 	/* update ->cpus if we have cluster, no harm if not */
 	cpumask_copy(policy->cpus, per_cpu(cpu_mask, cpu));
diff --git a/drivers/cpufreq/s5pv210-cpufreq.c b/drivers/cpufreq/s5pv210-cpufreq.c
index 3f9791f07b8e..b0dac7d6ba31 100644
--- a/drivers/cpufreq/s5pv210-cpufreq.c
+++ b/drivers/cpufreq/s5pv210-cpufreq.c
@@ -597,7 +597,7 @@ static int s5pv210_cpufreq_probe(struct platform_device *pdev)
 	 * and dependencies on platform headers. It is necessary to enable
 	 * S5PV210 multi-platform support and will be removed together with
 	 * this whole driver as soon as S5PV210 gets migrated to use
-	 * cpufreq-cpu0 driver.
+	 * cpufreq-dt driver.
 	 */
 	np = of_find_compatible_node(NULL, NULL, "samsung,s5pv210-clock");
 	if (!np) {
@@ -656,7 +656,6 @@ static int s5pv210_cpufreq_probe(struct platform_device *pdev)
 static struct platform_driver s5pv210_cpufreq_platdrv = {
 	.driver = {
 		.name	= "s5pv210-cpufreq",
-		.owner	= THIS_MODULE,
 	},
 	.probe = s5pv210_cpufreq_probe,
 };
diff --git a/drivers/cpufreq/spear-cpufreq.c b/drivers/cpufreq/spear-cpufreq.c
index 38678396636d..4894924a3ca2 100644
--- a/drivers/cpufreq/spear-cpufreq.c
+++ b/drivers/cpufreq/spear-cpufreq.c
@@ -236,7 +236,6 @@ out_put_node:
 static struct platform_driver spear_cpufreq_platdrv = {
 	.driver = {
 		.name	= "spear-cpufreq",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= spear_cpufreq_probe,
 };
diff --git a/drivers/cpufreq/speedstep-ich.c b/drivers/cpufreq/speedstep-ich.c
index 1a07b5904ed5..e56d632a8b21 100644
--- a/drivers/cpufreq/speedstep-ich.c
+++ b/drivers/cpufreq/speedstep-ich.c
@@ -378,8 +378,7 @@ static void __exit speedstep_exit(void)
 }
 
 
-MODULE_AUTHOR("Dave Jones <davej@redhat.com>, "
-		"Dominik Brodowski <linux@brodo.de>");
+MODULE_AUTHOR("Dave Jones, Dominik Brodowski <linux@brodo.de>");
 MODULE_DESCRIPTION("Speedstep driver for Intel mobile processors on chipsets "
 		"with ICH-M southbridges.");
 MODULE_LICENSE("GPL");
diff --git a/drivers/cpufreq/vexpress-spc-cpufreq.c b/drivers/cpufreq/vexpress-spc-cpufreq.c
index 7f7c9c01b44e..433e93fd4900 100644
--- a/drivers/cpufreq/vexpress-spc-cpufreq.c
+++ b/drivers/cpufreq/vexpress-spc-cpufreq.c
@@ -60,7 +60,6 @@ static int ve_spc_cpufreq_remove(struct platform_device *pdev)
 static struct platform_driver ve_spc_cpufreq_platdrv = {
 	.driver = {
 		.name	= "vexpress-spc-cpufreq",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= ve_spc_cpufreq_probe,
 	.remove		= ve_spc_cpufreq_remove,
diff --git a/drivers/cpuidle/Kconfig b/drivers/cpuidle/Kconfig
index 32748c36c477..c5029c1209b4 100644
--- a/drivers/cpuidle/Kconfig
+++ b/drivers/cpuidle/Kconfig
@@ -25,11 +25,19 @@ config CPU_IDLE_GOV_MENU
 	bool "Menu governor (for tickless system)"
 	default y
 
+config DT_IDLE_STATES
+	bool
+
 menu "ARM CPU Idle Drivers"
 depends on ARM
 source "drivers/cpuidle/Kconfig.arm"
 endmenu
 
+menu "ARM64 CPU Idle Drivers"
+depends on ARM64
+source "drivers/cpuidle/Kconfig.arm64"
+endmenu
+
 menu "MIPS CPU Idle Drivers"
 depends on MIPS
 source "drivers/cpuidle/Kconfig.mips"
diff --git a/drivers/cpuidle/Kconfig.arm b/drivers/cpuidle/Kconfig.arm
index 38cff69ffe06..8c16ab20fb15 100644
--- a/drivers/cpuidle/Kconfig.arm
+++ b/drivers/cpuidle/Kconfig.arm
@@ -7,6 +7,7 @@ config ARM_BIG_LITTLE_CPUIDLE
 	depends on MCPM
 	select ARM_CPU_SUSPEND
 	select CPU_IDLE_MULTIPLE_DRIVERS
+	select DT_IDLE_STATES
 	help
 	  Select this option to enable CPU idle driver for big.LITTLE based
 	  ARM systems. Driver manages CPUs coordination through MCPM and
@@ -28,7 +29,7 @@ config ARM_HIGHBANK_CPUIDLE
 
 config ARM_KIRKWOOD_CPUIDLE
 	bool "CPU Idle Driver for Marvell Kirkwood SoCs"
-	depends on ARCH_KIRKWOOD || MACH_KIRKWOOD
+	depends on MACH_KIRKWOOD
 	help
 	  This adds the CPU Idle driver for Marvell Kirkwood SoCs.
 
diff --git a/drivers/cpuidle/Kconfig.arm64 b/drivers/cpuidle/Kconfig.arm64
new file mode 100644
index 000000000000..d0a08ed1b2ee
--- /dev/null
+++ b/drivers/cpuidle/Kconfig.arm64
@@ -0,0 +1,14 @@
+#
+# ARM64 CPU Idle drivers
+#
+
+config ARM64_CPUIDLE
+	bool "Generic ARM64 CPU idle Driver"
+	select ARM64_CPU_SUSPEND
+	select DT_IDLE_STATES
+	help
+	  Select this to enable generic cpuidle driver for ARM64.
+	  It provides a generic idle driver whose idle states are configured
+	  at run-time through DT nodes. The CPUidle suspend backend is
+	  initialized by calling the CPU operations init idle hook
+	  provided by architecture code.
diff --git a/drivers/cpuidle/Kconfig.mips b/drivers/cpuidle/Kconfig.mips
index 0e70ee28a5ca..4102be01d06a 100644
--- a/drivers/cpuidle/Kconfig.mips
+++ b/drivers/cpuidle/Kconfig.mips
@@ -3,7 +3,7 @@
 #
 config MIPS_CPS_CPUIDLE
 	bool "CPU Idle driver for MIPS CPS platforms"
-	depends on CPU_IDLE
+	depends on CPU_IDLE && MIPS_CPS
 	depends on SYS_SUPPORTS_MIPS_CPS
 	select ARCH_NEEDS_CPU_IDLE_COUPLED if MIPS_MT
 	select GENERIC_CLOCKEVENTS_BROADCAST if SMP
diff --git a/drivers/cpuidle/Makefile b/drivers/cpuidle/Makefile
index 11edb31c55e9..4d177b916f75 100644
--- a/drivers/cpuidle/Makefile
+++ b/drivers/cpuidle/Makefile
@@ -4,6 +4,7 @@
 
 obj-y += cpuidle.o driver.o governor.o sysfs.o governors/
 obj-$(CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED) += coupled.o
+obj-$(CONFIG_DT_IDLE_STATES)		  += dt_idle_states.o
 
 ##################################################################################
 # ARM SoC drivers
@@ -22,6 +23,10 @@ obj-$(CONFIG_ARM_EXYNOS_CPUIDLE)        += cpuidle-exynos.o
 obj-$(CONFIG_MIPS_CPS_CPUIDLE)		+= cpuidle-cps.o
 
 ###############################################################################
+# ARM64 drivers
+obj-$(CONFIG_ARM64_CPUIDLE)		+= cpuidle-arm64.o
+
+###############################################################################
 # POWERPC drivers
 obj-$(CONFIG_PSERIES_CPUIDLE)		+= cpuidle-pseries.o
 obj-$(CONFIG_POWERNV_CPUIDLE)		+= cpuidle-powernv.o
diff --git a/drivers/cpuidle/cpuidle-arm64.c b/drivers/cpuidle/cpuidle-arm64.c
new file mode 100644
index 000000000000..80704b931ba4
--- /dev/null
+++ b/drivers/cpuidle/cpuidle-arm64.c
@@ -0,0 +1,123 @@
+/*
+ * ARM64 generic CPU idle driver.
+ *
+ * Copyright (C) 2014 ARM Ltd.
+ * Author: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#define pr_fmt(fmt) "CPUidle arm64: " fmt
+
+#include <linux/cpuidle.h>
+#include <linux/cpumask.h>
+#include <linux/cpu_pm.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+
+#include <asm/cpuidle.h>
+#include <asm/suspend.h>
+
+#include "dt_idle_states.h"
+
+/*
+ * arm64_enter_idle_state - Programs CPU to enter the specified state
+ *
+ * dev: cpuidle device
+ * drv: cpuidle driver
+ * idx: state index
+ *
+ * Called from the CPUidle framework to program the device to the
+ * specified target state selected by the governor.
+ */
+static int arm64_enter_idle_state(struct cpuidle_device *dev,
+				  struct cpuidle_driver *drv, int idx)
+{
+	int ret;
+
+	if (!idx) {
+		cpu_do_idle();
+		return idx;
+	}
+
+	ret = cpu_pm_enter();
+	if (!ret) {
+		/*
+		 * Pass idle state index to cpu_suspend which in turn will
+		 * call the CPU ops suspend protocol with idle index as a
+		 * parameter.
+		 */
+		ret = cpu_suspend(idx);
+
+		cpu_pm_exit();
+	}
+
+	return ret ? -1 : idx;
+}
+
+static struct cpuidle_driver arm64_idle_driver = {
+	.name = "arm64_idle",
+	.owner = THIS_MODULE,
+	/*
+	 * State at index 0 is standby wfi and considered standard
+	 * on all ARM platforms. If in some platforms simple wfi
+	 * can't be used as "state 0", DT bindings must be implemented
+	 * to work around this issue and allow installing a special
+	 * handler for idle state index 0.
+	 */
+	.states[0] = {
+		.enter                  = arm64_enter_idle_state,
+		.exit_latency           = 1,
+		.target_residency       = 1,
+		.power_usage		= UINT_MAX,
+		.name                   = "WFI",
+		.desc                   = "ARM64 WFI",
+	}
+};
+
+static const struct of_device_id arm64_idle_state_match[] __initconst = {
+	{ .compatible = "arm,idle-state",
+	  .data = arm64_enter_idle_state },
+	{ },
+};
+
+/*
+ * arm64_idle_init
+ *
+ * Registers the arm64 specific cpuidle driver with the cpuidle
+ * framework. It relies on core code to parse the idle states
+ * and initialize them using driver data structures accordingly.
+ */
+static int __init arm64_idle_init(void)
+{
+	int cpu, ret;
+	struct cpuidle_driver *drv = &arm64_idle_driver;
+
+	/*
+	 * Initialize idle states data, starting at index 1.
+	 * This driver is DT only, if no DT idle states are detected (ret == 0)
+	 * let the driver initialization fail accordingly since there is no
+	 * reason to initialize the idle driver if only wfi is supported.
+	 */
+	ret = dt_init_idle_driver(drv, arm64_idle_state_match, 1);
+	if (ret <= 0)
+		return ret ? : -ENODEV;
+
+	/*
+	 * Call arch CPU operations in order to initialize
+	 * idle states suspend back-end specific data
+	 */
+	for_each_possible_cpu(cpu) {
+		ret = cpu_init_idle(cpu);
+		if (ret) {
+			pr_err("CPU %d failed to init idle CPU ops\n", cpu);
+			return ret;
+		}
+	}
+
+	return cpuidle_register(drv, NULL);
+}
+device_initcall(arm64_idle_init);
diff --git a/drivers/cpuidle/cpuidle-at91.c b/drivers/cpuidle/cpuidle-at91.c
index a0774370c6bc..aae7bfc1ea36 100644
--- a/drivers/cpuidle/cpuidle-at91.c
+++ b/drivers/cpuidle/cpuidle-at91.c
@@ -43,7 +43,6 @@ static struct cpuidle_driver at91_idle_driver = {
 		.enter			= at91_enter_idle,
 		.exit_latency		= 10,
 		.target_residency	= 10000,
-		.flags			= CPUIDLE_FLAG_TIME_VALID,
 		.name			= "RAM_SR",
 		.desc			= "WFI and DDR Self Refresh",
 	},
@@ -61,7 +60,6 @@ static int at91_cpuidle_probe(struct platform_device *dev)
 static struct platform_driver at91_cpuidle_driver = {
 	.driver = {
 		.name = "cpuidle-at91",
-		.owner = THIS_MODULE,
 	},
 	.probe = at91_cpuidle_probe,
 };
diff --git a/drivers/cpuidle/cpuidle-big_little.c b/drivers/cpuidle/cpuidle-big_little.c
index ef94c3b81f18..e3e225fe6b45 100644
--- a/drivers/cpuidle/cpuidle-big_little.c
+++ b/drivers/cpuidle/cpuidle-big_little.c
@@ -24,6 +24,8 @@
 #include <asm/smp_plat.h>
 #include <asm/suspend.h>
 
+#include "dt_idle_states.h"
+
 static int bl_enter_powerdown(struct cpuidle_device *dev,
 			      struct cpuidle_driver *drv, int idx);
 
@@ -65,14 +67,19 @@ static struct cpuidle_driver bl_idle_little_driver = {
 		.enter			= bl_enter_powerdown,
 		.exit_latency		= 700,
 		.target_residency	= 2500,
-		.flags			= CPUIDLE_FLAG_TIME_VALID |
-					  CPUIDLE_FLAG_TIMER_STOP,
+		.flags			= CPUIDLE_FLAG_TIMER_STOP,
 		.name			= "C1",
 		.desc			= "ARM little-cluster power down",
 	},
 	.state_count = 2,
 };
 
+static const struct of_device_id bl_idle_state_match[] __initconst = {
+	{ .compatible = "arm,idle-state",
+	  .data = bl_enter_powerdown },
+	{ },
+};
+
 static struct cpuidle_driver bl_idle_big_driver = {
 	.name = "big_idle",
 	.owner = THIS_MODULE,
@@ -81,8 +88,7 @@ static struct cpuidle_driver bl_idle_big_driver = {
 		.enter			= bl_enter_powerdown,
 		.exit_latency		= 500,
 		.target_residency	= 2000,
-		.flags			= CPUIDLE_FLAG_TIME_VALID |
-					  CPUIDLE_FLAG_TIMER_STOP,
+		.flags			= CPUIDLE_FLAG_TIMER_STOP,
 		.name			= "C1",
 		.desc			= "ARM big-cluster power down",
 	},
@@ -159,6 +165,7 @@ static int __init bl_idle_driver_init(struct cpuidle_driver *drv, int part_id)
 static const struct of_device_id compatible_machine_match[] = {
 	{ .compatible = "arm,vexpress,v2p-ca15_a7" },
 	{ .compatible = "samsung,exynos5420" },
+	{ .compatible = "samsung,exynos5800" },
 	{},
 };
 
@@ -190,6 +197,17 @@ static int __init bl_idle_init(void)
 	if (ret)
 		goto out_uninit_little;
 
+	/* Start at index 1, index 0 standard WFI */
+	ret = dt_init_idle_driver(&bl_idle_big_driver, bl_idle_state_match, 1);
+	if (ret < 0)
+		goto out_uninit_big;
+
+	/* Start at index 1, index 0 standard WFI */
+	ret = dt_init_idle_driver(&bl_idle_little_driver,
+				  bl_idle_state_match, 1);
+	if (ret < 0)
+		goto out_uninit_big;
+
 	ret = cpuidle_register(&bl_idle_little_driver, NULL);
 	if (ret)
 		goto out_uninit_big;
diff --git a/drivers/cpuidle/cpuidle-calxeda.c b/drivers/cpuidle/cpuidle-calxeda.c
index 6e51114057d0..9445e6cc02be 100644
--- a/drivers/cpuidle/cpuidle-calxeda.c
+++ b/drivers/cpuidle/cpuidle-calxeda.c
@@ -55,7 +55,6 @@ static struct cpuidle_driver calxeda_idle_driver = {
 		{
 			.name = "PG",
 			.desc = "Power Gate",
-			.flags = CPUIDLE_FLAG_TIME_VALID,
 			.exit_latency = 30,
 			.power_usage = 50,
 			.target_residency = 200,
@@ -73,7 +72,6 @@ static int calxeda_cpuidle_probe(struct platform_device *pdev)
 static struct platform_driver calxeda_cpuidle_plat_driver = {
         .driver = {
                 .name = "cpuidle-calxeda",
-                .owner = THIS_MODULE,
         },
         .probe = calxeda_cpuidle_probe,
 };
diff --git a/drivers/cpuidle/cpuidle-clps711x.c b/drivers/cpuidle/cpuidle-clps711x.c
index 5243811daa6e..18a7f7380508 100644
--- a/drivers/cpuidle/cpuidle-clps711x.c
+++ b/drivers/cpuidle/cpuidle-clps711x.c
@@ -54,7 +54,6 @@ static int __init clps711x_cpuidle_probe(struct platform_device *pdev)
 static struct platform_driver clps711x_cpuidle_driver = {
 	.driver	= {
 		.name	= CLPS711X_CPUIDLE_NAME,
-		.owner	= THIS_MODULE,
 	},
 };
 module_platform_driver_probe(clps711x_cpuidle_driver, clps711x_cpuidle_probe);
diff --git a/drivers/cpuidle/cpuidle-cps.c b/drivers/cpuidle/cpuidle-cps.c
index fc7b62720deb..1adb6980b707 100644
--- a/drivers/cpuidle/cpuidle-cps.c
+++ b/drivers/cpuidle/cpuidle-cps.c
@@ -79,7 +79,6 @@ static struct cpuidle_driver cps_driver = {
 			.enter	= cps_nc_enter,
 			.exit_latency		= 200,
 			.target_residency	= 450,
-			.flags	= CPUIDLE_FLAG_TIME_VALID,
 			.name	= "nc-wait",
 			.desc	= "non-coherent MIPS wait",
 		},
@@ -87,8 +86,7 @@ static struct cpuidle_driver cps_driver = {
 			.enter	= cps_nc_enter,
 			.exit_latency		= 300,
 			.target_residency	= 700,
-			.flags	= CPUIDLE_FLAG_TIME_VALID |
-				  CPUIDLE_FLAG_TIMER_STOP,
+			.flags	= CPUIDLE_FLAG_TIMER_STOP,
 			.name	= "clock-gated",
 			.desc	= "core clock gated",
 		},
@@ -96,8 +94,7 @@ static struct cpuidle_driver cps_driver = {
 			.enter	= cps_nc_enter,
 			.exit_latency		= 600,
 			.target_residency	= 1000,
-			.flags	= CPUIDLE_FLAG_TIME_VALID |
-				  CPUIDLE_FLAG_TIMER_STOP,
+			.flags	= CPUIDLE_FLAG_TIMER_STOP,
 			.name	= "power-gated",
 			.desc	= "core power gated",
 		},
diff --git a/drivers/cpuidle/cpuidle-exynos.c b/drivers/cpuidle/cpuidle-exynos.c
index ba9b34b579f3..4003a3160865 100644
--- a/drivers/cpuidle/cpuidle-exynos.c
+++ b/drivers/cpuidle/cpuidle-exynos.c
@@ -47,7 +47,6 @@ static struct cpuidle_driver exynos_idle_driver = {
 			.enter			= exynos_enter_lowpower,
 			.exit_latency		= 300,
 			.target_residency	= 100000,
-			.flags			= CPUIDLE_FLAG_TIME_VALID,
 			.name			= "C1",
 			.desc			= "ARM power down",
 		},
@@ -75,7 +74,6 @@ static struct platform_driver exynos_cpuidle_driver = {
 	.probe	= exynos_cpuidle_probe,
 	.driver = {
 		.name = "exynos_cpuidle",
-		.owner = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/cpuidle/cpuidle-kirkwood.c b/drivers/cpuidle/cpuidle-kirkwood.c
index 41ba843251b8..cea0a6c4b1db 100644
--- a/drivers/cpuidle/cpuidle-kirkwood.c
+++ b/drivers/cpuidle/cpuidle-kirkwood.c
@@ -47,7 +47,6 @@ static struct cpuidle_driver kirkwood_idle_driver = {
 		.enter			= kirkwood_enter_idle,
 		.exit_latency		= 10,
 		.target_residency	= 100000,
-		.flags			= CPUIDLE_FLAG_TIME_VALID,
 		.name			= "DDR SR",
 		.desc			= "WFI and DDR Self Refresh",
 	},
@@ -78,7 +77,6 @@ static struct platform_driver kirkwood_cpuidle_driver = {
 	.remove = kirkwood_cpuidle_remove,
 	.driver = {
 		   .name = "kirkwood_cpuidle",
-		   .owner = THIS_MODULE,
 		   },
 };
 
diff --git a/drivers/cpuidle/cpuidle-mvebu-v7.c b/drivers/cpuidle/cpuidle-mvebu-v7.c
index 45371bb16214..38e68618513a 100644
--- a/drivers/cpuidle/cpuidle-mvebu-v7.c
+++ b/drivers/cpuidle/cpuidle-mvebu-v7.c
@@ -53,7 +53,6 @@ static struct cpuidle_driver armadaxp_idle_driver = {
 		.exit_latency		= 10,
 		.power_usage		= 50,
 		.target_residency	= 100,
-		.flags			= CPUIDLE_FLAG_TIME_VALID,
 		.name			= "MV CPU IDLE",
 		.desc			= "CPU power down",
 	},
@@ -62,8 +61,7 @@ static struct cpuidle_driver armadaxp_idle_driver = {
 		.exit_latency		= 100,
 		.power_usage		= 5,
 		.target_residency	= 1000,
-		.flags			= CPUIDLE_FLAG_TIME_VALID |
-						MVEBU_V7_FLAG_DEEP_IDLE,
+		.flags			= MVEBU_V7_FLAG_DEEP_IDLE,
 		.name			= "MV CPU DEEP IDLE",
 		.desc			= "CPU and L2 Fabric power down",
 	},
@@ -78,8 +76,7 @@ static struct cpuidle_driver armada370_idle_driver = {
 		.exit_latency		= 100,
 		.power_usage		= 5,
 		.target_residency	= 1000,
-		.flags			= (CPUIDLE_FLAG_TIME_VALID |
-					   MVEBU_V7_FLAG_DEEP_IDLE),
+		.flags			= MVEBU_V7_FLAG_DEEP_IDLE,
 		.name			= "Deep Idle",
 		.desc			= "CPU and L2 Fabric power down",
 	},
@@ -94,7 +91,6 @@ static struct cpuidle_driver armada38x_idle_driver = {
 		.exit_latency		= 10,
 		.power_usage		= 5,
 		.target_residency	= 100,
-		.flags			= CPUIDLE_FLAG_TIME_VALID,
 		.name			= "Idle",
 		.desc			= "CPU and SCU power down",
 	},
@@ -118,7 +114,6 @@ static int mvebu_v7_cpuidle_probe(struct platform_device *pdev)
 static struct platform_driver armadaxp_cpuidle_plat_driver = {
 	.driver = {
 		.name = "cpuidle-armada-xp",
-		.owner = THIS_MODULE,
 	},
 	.probe = mvebu_v7_cpuidle_probe,
 };
@@ -128,7 +123,6 @@ module_platform_driver(armadaxp_cpuidle_plat_driver);
 static struct platform_driver armada370_cpuidle_plat_driver = {
 	.driver = {
 		.name = "cpuidle-armada-370",
-		.owner = THIS_MODULE,
 	},
 	.probe = mvebu_v7_cpuidle_probe,
 };
@@ -138,7 +132,6 @@ module_platform_driver(armada370_cpuidle_plat_driver);
 static struct platform_driver armada38x_cpuidle_plat_driver = {
 	.driver = {
 		.name = "cpuidle-armada-38x",
-		.owner = THIS_MODULE,
 	},
 	.probe = mvebu_v7_cpuidle_probe,
 };
diff --git a/drivers/cpuidle/cpuidle-powernv.c b/drivers/cpuidle/cpuidle-powernv.c
index a64be578dab2..aedec0957934 100644
--- a/drivers/cpuidle/cpuidle-powernv.c
+++ b/drivers/cpuidle/cpuidle-powernv.c
@@ -16,13 +16,10 @@
 
 #include <asm/machdep.h>
 #include <asm/firmware.h>
+#include <asm/opal.h>
 #include <asm/runlatch.h>
 
-/* Flags and constants used in PowerNV platform */
-
 #define MAX_POWERNV_IDLE_STATES	8
-#define IDLE_USE_INST_NAP	0x00010000 /* Use nap instruction */
-#define IDLE_USE_INST_SLEEP	0x00020000 /* Use sleep instruction */
 
 struct cpuidle_driver powernv_idle_driver = {
 	.name             = "powernv_idle",
@@ -93,7 +90,6 @@ static struct cpuidle_state powernv_states[MAX_POWERNV_IDLE_STATES] = {
 	{ /* Snooze */
 		.name = "snooze",
 		.desc = "snooze",
-		.flags = CPUIDLE_FLAG_TIME_VALID,
 		.exit_latency = 0,
 		.target_residency = 0,
 		.enter = &snooze_loop },
@@ -163,7 +159,8 @@ static int powernv_add_idle_states(void)
 	int nr_idle_states = 1; /* Snooze */
 	int dt_idle_states;
 	const __be32 *idle_state_flags;
-	u32 len_flags, flags;
+	const __be32 *idle_state_latency;
+	u32 len_flags, flags, latency_ns;
 	int i;
 
 	/* Currently we have snooze statically defined */
@@ -180,30 +177,46 @@ static int powernv_add_idle_states(void)
 		return nr_idle_states;
 	}
 
+	idle_state_latency = of_get_property(power_mgt,
+			"ibm,cpu-idle-state-latencies-ns", NULL);
+	if (!idle_state_latency) {
+		pr_warn("DT-PowerMgmt: missing ibm,cpu-idle-state-latencies-ns\n");
+		return nr_idle_states;
+	}
+
 	dt_idle_states = len_flags / sizeof(u32);
 
 	for (i = 0; i < dt_idle_states; i++) {
 
 		flags = be32_to_cpu(idle_state_flags[i]);
-		if (flags & IDLE_USE_INST_NAP) {
+
+		/* Cpuidle accepts exit_latency in us and we estimate
+		 * target residency to be 10x exit_latency
+		 */
+		latency_ns = be32_to_cpu(idle_state_latency[i]);
+		if (flags & OPAL_PM_NAP_ENABLED) {
 			/* Add NAP state */
 			strcpy(powernv_states[nr_idle_states].name, "Nap");
 			strcpy(powernv_states[nr_idle_states].desc, "Nap");
-			powernv_states[nr_idle_states].flags = CPUIDLE_FLAG_TIME_VALID;
-			powernv_states[nr_idle_states].exit_latency = 10;
-			powernv_states[nr_idle_states].target_residency = 100;
+			powernv_states[nr_idle_states].flags = 0;
+			powernv_states[nr_idle_states].exit_latency =
+					((unsigned int)latency_ns) / 1000;
+			powernv_states[nr_idle_states].target_residency =
+					((unsigned int)latency_ns / 100);
 			powernv_states[nr_idle_states].enter = &nap_loop;
 			nr_idle_states++;
 		}
 
-		if (flags & IDLE_USE_INST_SLEEP) {
+		if (flags & OPAL_PM_SLEEP_ENABLED ||
+			flags & OPAL_PM_SLEEP_ENABLED_ER1) {
 			/* Add FASTSLEEP state */
 			strcpy(powernv_states[nr_idle_states].name, "FastSleep");
 			strcpy(powernv_states[nr_idle_states].desc, "FastSleep");
-			powernv_states[nr_idle_states].flags =
-				CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TIMER_STOP;
-			powernv_states[nr_idle_states].exit_latency = 300;
-			powernv_states[nr_idle_states].target_residency = 1000000;
+			powernv_states[nr_idle_states].flags = CPUIDLE_FLAG_TIMER_STOP;
+			powernv_states[nr_idle_states].exit_latency =
+					((unsigned int)latency_ns) / 1000;
+			powernv_states[nr_idle_states].target_residency =
+					((unsigned int)latency_ns / 100);
 			powernv_states[nr_idle_states].enter = &fastsleep_loop;
 			nr_idle_states++;
 		}
diff --git a/drivers/cpuidle/cpuidle-pseries.c b/drivers/cpuidle/cpuidle-pseries.c
index 6f7b01956885..bb9e2b6f3ecc 100644
--- a/drivers/cpuidle/cpuidle-pseries.c
+++ b/drivers/cpuidle/cpuidle-pseries.c
@@ -142,14 +142,12 @@ static struct cpuidle_state dedicated_states[] = {
 	{ /* Snooze */
 		.name = "snooze",
 		.desc = "snooze",
-		.flags = CPUIDLE_FLAG_TIME_VALID,
 		.exit_latency = 0,
 		.target_residency = 0,
 		.enter = &snooze_loop },
 	{ /* CEDE */
 		.name = "CEDE",
 		.desc = "CEDE",
-		.flags = CPUIDLE_FLAG_TIME_VALID,
 		.exit_latency = 10,
 		.target_residency = 100,
 		.enter = &dedicated_cede_loop },
@@ -162,7 +160,6 @@ static struct cpuidle_state shared_states[] = {
 	{ /* Shared Cede */
 		.name = "Shared Cede",
 		.desc = "Shared Cede",
-		.flags = CPUIDLE_FLAG_TIME_VALID,
 		.exit_latency = 0,
 		.target_residency = 0,
 		.enter = &shared_cede_loop },
diff --git a/drivers/cpuidle/cpuidle-ux500.c b/drivers/cpuidle/cpuidle-ux500.c
index 5e35804b1a95..66f81e410f0d 100644
--- a/drivers/cpuidle/cpuidle-ux500.c
+++ b/drivers/cpuidle/cpuidle-ux500.c
@@ -101,8 +101,7 @@ static struct cpuidle_driver ux500_idle_driver = {
 			.enter		  = ux500_enter_idle,
 			.exit_latency	  = 70,
 			.target_residency = 260,
-			.flags		  = CPUIDLE_FLAG_TIME_VALID |
-			                    CPUIDLE_FLAG_TIMER_STOP,
+			.flags		  = CPUIDLE_FLAG_TIMER_STOP,
 			.name		  = "ApIdle",
 			.desc		  = "ARM Retention",
 		},
@@ -123,7 +122,6 @@ static int dbx500_cpuidle_probe(struct platform_device *pdev)
 static struct platform_driver dbx500_cpuidle_plat_driver = {
 	.driver = {
 		.name = "cpuidle-dbx500",
-		.owner = THIS_MODULE,
 	},
 	.probe = dbx500_cpuidle_probe,
 };
diff --git a/drivers/cpuidle/cpuidle-zynq.c b/drivers/cpuidle/cpuidle-zynq.c
index aded75928028..002b8c9f98f5 100644
--- a/drivers/cpuidle/cpuidle-zynq.c
+++ b/drivers/cpuidle/cpuidle-zynq.c
@@ -26,7 +26,6 @@
  */
 
 #include <linux/init.h>
-#include <linux/cpu_pm.h>
 #include <linux/cpuidle.h>
 #include <linux/platform_device.h>
 #include <asm/proc-fns.h>
@@ -38,15 +37,9 @@
 static int zynq_enter_idle(struct cpuidle_device *dev,
 			   struct cpuidle_driver *drv, int index)
 {
-	/* Devices must be stopped here */
-	cpu_pm_enter();
-
 	/* Add code for DDR self refresh start */
 	cpu_do_idle();
 
-	/* Add code for DDR self refresh stop */
-	cpu_pm_exit();
-
 	return index;
 }
 
@@ -59,8 +52,6 @@ static struct cpuidle_driver zynq_idle_driver = {
 			.enter			= zynq_enter_idle,
 			.exit_latency		= 10,
 			.target_residency	= 10000,
-			.flags			= CPUIDLE_FLAG_TIME_VALID |
-						  CPUIDLE_FLAG_TIMER_STOP,
 			.name			= "RAM_SR",
 			.desc			= "WFI and RAM Self Refresh",
 		},
@@ -80,7 +71,6 @@ static int zynq_cpuidle_probe(struct platform_device *pdev)
 static struct platform_driver zynq_cpuidle_driver = {
 	.driver = {
 		.name = "cpuidle-zynq",
-		.owner = THIS_MODULE,
 	},
 	.probe = zynq_cpuidle_probe,
 };
diff --git a/drivers/cpuidle/cpuidle.c b/drivers/cpuidle/cpuidle.c
index ee9df5e3f5eb..125150dc6e81 100644
--- a/drivers/cpuidle/cpuidle.c
+++ b/drivers/cpuidle/cpuidle.c
@@ -223,8 +223,14 @@ void cpuidle_uninstall_idle_handler(void)
 {
 	if (enabled_devices) {
 		initialized = 0;
-		kick_all_cpus_sync();
+		wake_up_all_idle_cpus();
 	}
+
+	/*
+	 * Make sure external observers (such as the scheduler)
+	 * are done looking at pointed idle states.
+	 */
+	synchronize_rcu();
 }
 
 /**
@@ -530,11 +536,6 @@ EXPORT_SYMBOL_GPL(cpuidle_register);
 
 #ifdef CONFIG_SMP
 
-static void smp_callback(void *v)
-{
-	/* we already woke the CPU up, nothing more to do */
-}
-
 /*
  * This function gets called when a part of the kernel has a new latency
  * requirement.  This means we need to get all processors out of their C-state,
@@ -544,7 +545,7 @@ static void smp_callback(void *v)
 static int cpuidle_latency_notify(struct notifier_block *b,
 		unsigned long l, void *v)
 {
-	smp_call_function(smp_callback, NULL, 1);
+	wake_up_all_idle_cpus();
 	return NOTIFY_OK;
 }
 
diff --git a/drivers/cpuidle/driver.c b/drivers/cpuidle/driver.c
index e431d11abf8d..2697e87d5b34 100644
--- a/drivers/cpuidle/driver.c
+++ b/drivers/cpuidle/driver.c
@@ -201,7 +201,6 @@ static void poll_idle_init(struct cpuidle_driver *drv)
 	state->exit_latency = 0;
 	state->target_residency = 0;
 	state->power_usage = -1;
-	state->flags = CPUIDLE_FLAG_TIME_VALID;
 	state->enter = poll_idle;
 	state->disabled = false;
 }
diff --git a/drivers/cpuidle/dt_idle_states.c b/drivers/cpuidle/dt_idle_states.c
new file mode 100644
index 000000000000..a5c111b67f37
--- /dev/null
+++ b/drivers/cpuidle/dt_idle_states.c
@@ -0,0 +1,221 @@
+/*
+ * DT idle states parsing code.
+ *
+ * Copyright (C) 2014 ARM Ltd.
+ * Author: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#define pr_fmt(fmt) "DT idle-states: " fmt
+
+#include <linux/cpuidle.h>
+#include <linux/cpumask.h>
+#include <linux/errno.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+
+#include "dt_idle_states.h"
+
+static int init_state_node(struct cpuidle_state *idle_state,
+			   const struct of_device_id *matches,
+			   struct device_node *state_node)
+{
+	int err;
+	const struct of_device_id *match_id;
+	const char *desc;
+
+	match_id = of_match_node(matches, state_node);
+	if (!match_id)
+		return -ENODEV;
+	/*
+	 * CPUidle drivers are expected to initialize the const void *data
+	 * pointer of the passed in struct of_device_id array to the idle
+	 * state enter function.
+	 */
+	idle_state->enter = match_id->data;
+
+	err = of_property_read_u32(state_node, "wakeup-latency-us",
+				   &idle_state->exit_latency);
+	if (err) {
+		u32 entry_latency, exit_latency;
+
+		err = of_property_read_u32(state_node, "entry-latency-us",
+					   &entry_latency);
+		if (err) {
+			pr_debug(" * %s missing entry-latency-us property\n",
+				 state_node->full_name);
+			return -EINVAL;
+		}
+
+		err = of_property_read_u32(state_node, "exit-latency-us",
+					   &exit_latency);
+		if (err) {
+			pr_debug(" * %s missing exit-latency-us property\n",
+				 state_node->full_name);
+			return -EINVAL;
+		}
+		/*
+		 * If wakeup-latency-us is missing, default to entry+exit
+		 * latencies as defined in idle states bindings
+		 */
+		idle_state->exit_latency = entry_latency + exit_latency;
+	}
+
+	err = of_property_read_u32(state_node, "min-residency-us",
+				   &idle_state->target_residency);
+	if (err) {
+		pr_debug(" * %s missing min-residency-us property\n",
+			     state_node->full_name);
+		return -EINVAL;
+	}
+
+	err = of_property_read_string(state_node, "idle-state-name", &desc);
+	if (err)
+		desc = state_node->name;
+
+	idle_state->flags = 0;
+	if (of_property_read_bool(state_node, "local-timer-stop"))
+		idle_state->flags |= CPUIDLE_FLAG_TIMER_STOP;
+	/*
+	 * TODO:
+	 *	replace with kstrdup and pointer assignment when name
+	 *	and desc become string pointers
+	 */
+	strncpy(idle_state->name, state_node->name, CPUIDLE_NAME_LEN - 1);
+	strncpy(idle_state->desc, desc, CPUIDLE_DESC_LEN - 1);
+	return 0;
+}
+
+/*
+ * Check that the idle state is uniform across all CPUs in the CPUidle driver
+ * cpumask
+ */
+static bool idle_state_valid(struct device_node *state_node, unsigned int idx,
+			     const cpumask_t *cpumask)
+{
+	int cpu;
+	struct device_node *cpu_node, *curr_state_node;
+	bool valid = true;
+
+	/*
+	 * Compare idle state phandles for index idx on all CPUs in the
+	 * CPUidle driver cpumask. Start from next logical cpu following
+	 * cpumask_first(cpumask) since that's the CPU state_node was
+	 * retrieved from. If a mismatch is found bail out straight
+	 * away since we certainly hit a firmware misconfiguration.
+	 */
+	for (cpu = cpumask_next(cpumask_first(cpumask), cpumask);
+	     cpu < nr_cpu_ids; cpu = cpumask_next(cpu, cpumask)) {
+		cpu_node = of_cpu_device_node_get(cpu);
+		curr_state_node = of_parse_phandle(cpu_node, "cpu-idle-states",
+						   idx);
+		if (state_node != curr_state_node)
+			valid = false;
+
+		of_node_put(curr_state_node);
+		of_node_put(cpu_node);
+		if (!valid)
+			break;
+	}
+
+	return valid;
+}
+
+/**
+ * dt_init_idle_driver() - Parse the DT idle states and initialize the
+ *			   idle driver states array
+ * @drv:	  Pointer to CPU idle driver to be initialized
+ * @matches:	  Array of of_device_id match structures to search in for
+ *		  compatible idle state nodes. The data pointer for each valid
+ *		  struct of_device_id entry in the matches array must point to
+ *		  a function with the following signature, that corresponds to
+ *		  the CPUidle state enter function signature:
+ *
+ *		  int (*)(struct cpuidle_device *dev,
+ *			  struct cpuidle_driver *drv,
+ *			  int index);
+ *
+ * @start_idx:    First idle state index to be initialized
+ *
+ * If DT idle states are detected and are valid the state count and states
+ * array entries in the cpuidle driver are initialized accordingly starting
+ * from index start_idx.
+ *
+ * Return: number of valid DT idle states parsed, <0 on failure
+ */
+int dt_init_idle_driver(struct cpuidle_driver *drv,
+			const struct of_device_id *matches,
+			unsigned int start_idx)
+{
+	struct cpuidle_state *idle_state;
+	struct device_node *state_node, *cpu_node;
+	int i, err = 0;
+	const cpumask_t *cpumask;
+	unsigned int state_idx = start_idx;
+
+	if (state_idx >= CPUIDLE_STATE_MAX)
+		return -EINVAL;
+	/*
+	 * We get the idle states for the first logical cpu in the
+	 * driver mask (or cpu_possible_mask if the driver cpumask is not set)
+	 * and we check through idle_state_valid() if they are uniform
+	 * across CPUs, otherwise we hit a firmware misconfiguration.
+	 */
+	cpumask = drv->cpumask ? : cpu_possible_mask;
+	cpu_node = of_cpu_device_node_get(cpumask_first(cpumask));
+
+	for (i = 0; ; i++) {
+		state_node = of_parse_phandle(cpu_node, "cpu-idle-states", i);
+		if (!state_node)
+			break;
+
+		if (!of_device_is_available(state_node))
+			continue;
+
+		if (!idle_state_valid(state_node, i, cpumask)) {
+			pr_warn("%s idle state not valid, bailing out\n",
+				state_node->full_name);
+			err = -EINVAL;
+			break;
+		}
+
+		if (state_idx == CPUIDLE_STATE_MAX) {
+			pr_warn("State index reached static CPU idle driver states array size\n");
+			break;
+		}
+
+		idle_state = &drv->states[state_idx++];
+		err = init_state_node(idle_state, matches, state_node);
+		if (err) {
+			pr_err("Parsing idle state node %s failed with err %d\n",
+			       state_node->full_name, err);
+			err = -EINVAL;
+			break;
+		}
+		of_node_put(state_node);
+	}
+
+	of_node_put(state_node);
+	of_node_put(cpu_node);
+	if (err)
+		return err;
+	/*
+	 * Update the driver state count only if some valid DT idle states
+	 * were detected
+	 */
+	if (i)
+		drv->state_count = state_idx;
+
+	/*
+	 * Return the number of present and valid DT idle states, which can
+	 * also be 0 on platforms with missing DT idle states or legacy DT
+	 * configuration predating the DT idle states bindings.
+	 */
+	return i;
+}
+EXPORT_SYMBOL_GPL(dt_init_idle_driver);
diff --git a/drivers/cpuidle/dt_idle_states.h b/drivers/cpuidle/dt_idle_states.h
new file mode 100644
index 000000000000..4818134bc65b
--- /dev/null
+++ b/drivers/cpuidle/dt_idle_states.h
@@ -0,0 +1,7 @@
+#ifndef __DT_IDLE_STATES
+#define __DT_IDLE_STATES
+
+int dt_init_idle_driver(struct cpuidle_driver *drv,
+			const struct of_device_id *matches,
+			unsigned int start_idx);
+#endif
diff --git a/drivers/cpuidle/governor.c b/drivers/cpuidle/governor.c
index ca89412f5122..fb9f511cca23 100644
--- a/drivers/cpuidle/governor.c
+++ b/drivers/cpuidle/governor.c
@@ -28,7 +28,7 @@ static struct cpuidle_governor * __cpuidle_find_governor(const char *str)
 	struct cpuidle_governor *gov;
 
 	list_for_each_entry(gov, &cpuidle_governors, governor_list)
-		if (!strnicmp(str, gov->name, CPUIDLE_NAME_LEN))
+		if (!strncasecmp(str, gov->name, CPUIDLE_NAME_LEN))
 			return gov;
 
 	return NULL;
diff --git a/drivers/cpuidle/governors/ladder.c b/drivers/cpuidle/governors/ladder.c
index 044ee0df5871..401c0106ed34 100644
--- a/drivers/cpuidle/governors/ladder.c
+++ b/drivers/cpuidle/governors/ladder.c
@@ -66,7 +66,7 @@ static inline void ladder_do_selection(struct ladder_device *ldev,
 static int ladder_select_state(struct cpuidle_driver *drv,
 				struct cpuidle_device *dev)
 {
-	struct ladder_device *ldev = &__get_cpu_var(ladder_devices);
+	struct ladder_device *ldev = this_cpu_ptr(&ladder_devices);
 	struct ladder_device_state *last_state;
 	int last_residency, last_idx = ldev->last_state_idx;
 	int latency_req = pm_qos_request(PM_QOS_CPU_DMA_LATENCY);
@@ -79,12 +79,7 @@ static int ladder_select_state(struct cpuidle_driver *drv,
 
 	last_state = &ldev->states[last_idx];
 
-	if (drv->states[last_idx].flags & CPUIDLE_FLAG_TIME_VALID) {
-		last_residency = cpuidle_get_last_residency(dev) - \
-					 drv->states[last_idx].exit_latency;
-	}
-	else
-		last_residency = last_state->threshold.promotion_time + 1;
+	last_residency = cpuidle_get_last_residency(dev) - drv->states[last_idx].exit_latency;
 
 	/* consider promotion */
 	if (last_idx < drv->state_count - 1 &&
@@ -170,7 +165,7 @@ static int ladder_enable_device(struct cpuidle_driver *drv,
  */
 static void ladder_reflect(struct cpuidle_device *dev, int index)
 {
-	struct ladder_device *ldev = &__get_cpu_var(ladder_devices);
+	struct ladder_device *ldev = this_cpu_ptr(&ladder_devices);
 	if (index > 0)
 		ldev->last_state_idx = index;
 }
diff --git a/drivers/cpuidle/governors/menu.c b/drivers/cpuidle/governors/menu.c
index 34db2fb3ef1e..40580794e23d 100644
--- a/drivers/cpuidle/governors/menu.c
+++ b/drivers/cpuidle/governors/menu.c
@@ -289,7 +289,7 @@ again:
  */
 static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev)
 {
-	struct menu_device *data = &__get_cpu_var(menu_devices);
+	struct menu_device *data = this_cpu_ptr(&menu_devices);
 	int latency_req = pm_qos_request(PM_QOS_CPU_DMA_LATENCY);
 	int i;
 	unsigned int interactivity_req;
@@ -372,7 +372,7 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev)
  */
 static void menu_reflect(struct cpuidle_device *dev, int index)
 {
-	struct menu_device *data = &__get_cpu_var(menu_devices);
+	struct menu_device *data = this_cpu_ptr(&menu_devices);
 	data->last_state_idx = index;
 	if (index >= 0)
 		data->needs_update = 1;
@@ -385,7 +385,7 @@ static void menu_reflect(struct cpuidle_device *dev, int index)
  */
 static void menu_update(struct cpuidle_driver *drv, struct cpuidle_device *dev)
 {
-	struct menu_device *data = &__get_cpu_var(menu_devices);
+	struct menu_device *data = this_cpu_ptr(&menu_devices);
 	int last_idx = data->last_state_idx;
 	struct cpuidle_state *target = &drv->states[last_idx];
 	unsigned int measured_us;
@@ -396,8 +396,8 @@ static void menu_update(struct cpuidle_driver *drv, struct cpuidle_device *dev)
 	 * power state and occurrence of the wakeup event.
 	 *
 	 * If the entered idle state didn't support residency measurements,
-	 * we are basically lost in the dark how much time passed.
-	 * As a compromise, assume we slept for the whole expected time.
+	 * we use them anyway if they are short, and if long,
+	 * truncate to the whole expected time.
 	 *
 	 * Any measured amount of time will include the exit latency.
 	 * Since we are interested in when the wakeup begun, not when it
@@ -405,22 +405,17 @@ static void menu_update(struct cpuidle_driver *drv, struct cpuidle_device *dev)
 	 * the measured amount of time is less than the exit latency,
 	 * assume the state was never reached and the exit latency is 0.
 	 */
-	if (unlikely(!(target->flags & CPUIDLE_FLAG_TIME_VALID))) {
-		/* Use timer value as is */
-		measured_us = data->next_timer_us;
 
-	} else {
-		/* Use measured value */
-		measured_us = cpuidle_get_last_residency(dev);
+	/* measured value */
+	measured_us = cpuidle_get_last_residency(dev);
 
-		/* Deduct exit latency */
-		if (measured_us > target->exit_latency)
-			measured_us -= target->exit_latency;
+	/* Deduct exit latency */
+	if (measured_us > target->exit_latency)
+		measured_us -= target->exit_latency;
 
-		/* Make sure our coefficients do not exceed unity */
-		if (measured_us > data->next_timer_us)
-			measured_us = data->next_timer_us;
-	}
+	/* Make sure our coefficients do not exceed unity */
+	if (measured_us > data->next_timer_us)
+		measured_us = data->next_timer_us;
 
 	/* Update our correction ratio */
 	new_factor = data->correction_factor[data->bucket];
diff --git a/drivers/crypto/amcc/crypto4xx_core.c b/drivers/crypto/amcc/crypto4xx_core.c
index e4c6c58fbb03..d02b77150070 100644
--- a/drivers/crypto/amcc/crypto4xx_core.c
+++ b/drivers/crypto/amcc/crypto4xx_core.c
@@ -1288,7 +1288,6 @@ static const struct of_device_id crypto4xx_match[] = {
 static struct platform_driver crypto4xx_driver = {
 	.driver = {
 		.name = "crypto4xx",
-		.owner = THIS_MODULE,
 		.of_match_table = crypto4xx_match,
 	},
 	.probe		= crypto4xx_probe,
diff --git a/drivers/crypto/atmel-aes.c b/drivers/crypto/atmel-aes.c
index a083474991ab..53d1c330f8a8 100644
--- a/drivers/crypto/atmel-aes.c
+++ b/drivers/crypto/atmel-aes.c
@@ -1473,7 +1473,6 @@ static struct platform_driver atmel_aes_driver = {
 	.remove		= atmel_aes_remove,
 	.driver		= {
 		.name	= "atmel_aes",
-		.owner	= THIS_MODULE,
 		.of_match_table = of_match_ptr(atmel_aes_dt_ids),
 	},
 };
diff --git a/drivers/crypto/atmel-sha.c b/drivers/crypto/atmel-sha.c
index 9a4f69eaa5e0..d94f07c78e19 100644
--- a/drivers/crypto/atmel-sha.c
+++ b/drivers/crypto/atmel-sha.c
@@ -1529,7 +1529,6 @@ static struct platform_driver atmel_sha_driver = {
 	.remove		= atmel_sha_remove,
 	.driver		= {
 		.name	= "atmel_sha",
-		.owner	= THIS_MODULE,
 		.of_match_table	= of_match_ptr(atmel_sha_dt_ids),
 	},
 };
diff --git a/drivers/crypto/atmel-tdes.c b/drivers/crypto/atmel-tdes.c
index d3a9041938ea..5e7c896cde30 100644
--- a/drivers/crypto/atmel-tdes.c
+++ b/drivers/crypto/atmel-tdes.c
@@ -1524,7 +1524,6 @@ static struct platform_driver atmel_tdes_driver = {
 	.remove		= atmel_tdes_remove,
 	.driver		= {
 		.name	= "atmel_tdes",
-		.owner	= THIS_MODULE,
 		.of_match_table = of_match_ptr(atmel_tdes_dt_ids),
 	},
 };
diff --git a/drivers/crypto/bfin_crc.c b/drivers/crypto/bfin_crc.c
index b099e33cb073..9ae149bddb6e 100644
--- a/drivers/crypto/bfin_crc.c
+++ b/drivers/crypto/bfin_crc.c
@@ -21,13 +21,13 @@
 #include <linux/scatterlist.h>
 #include <linux/dma-mapping.h>
 #include <linux/delay.h>
-#include <linux/unaligned/access_ok.h>
 #include <linux/crypto.h>
 #include <linux/cryptohash.h>
 #include <crypto/scatterwalk.h>
 #include <crypto/algapi.h>
 #include <crypto/hash.h>
 #include <crypto/internal/hash.h>
+#include <asm/unaligned.h>
 
 #include <asm/dma.h>
 #include <asm/portmux.h>
@@ -724,7 +724,6 @@ static struct platform_driver bfin_crypto_crc_driver = {
 	.resume    = bfin_crypto_crc_resume,
 	.driver    = {
 		.name  = DRIVER_NAME,
-		.owner = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/crypto/caam/caamalg.c b/drivers/crypto/caam/caamalg.c
index a80ea853701d..3187400daf31 100644
--- a/drivers/crypto/caam/caamalg.c
+++ b/drivers/crypto/caam/caamalg.c
@@ -60,6 +60,7 @@
 #define CAAM_CRA_PRIORITY		3000
 /* max key is sum of AES_MAX_KEY_SIZE, max split key size */
 #define CAAM_MAX_KEY_SIZE		(AES_MAX_KEY_SIZE + \
+					 CTR_RFC3686_NONCE_SIZE + \
 					 SHA512_DIGEST_SIZE * 2)
 /* max IV is max of AES_BLOCK_SIZE, DES3_EDE_BLOCK_SIZE */
 #define CAAM_MAX_IV_LENGTH		16
@@ -70,17 +71,34 @@
 #define DESC_AEAD_DEC_LEN		(DESC_AEAD_BASE + 18 * CAAM_CMD_SZ)
 #define DESC_AEAD_GIVENC_LEN		(DESC_AEAD_ENC_LEN + 7 * CAAM_CMD_SZ)
 
+/* Note: Nonce is counted in enckeylen */
+#define DESC_AEAD_CTR_RFC3686_LEN	(6 * CAAM_CMD_SZ)
+
 #define DESC_AEAD_NULL_BASE		(3 * CAAM_CMD_SZ)
 #define DESC_AEAD_NULL_ENC_LEN		(DESC_AEAD_NULL_BASE + 14 * CAAM_CMD_SZ)
 #define DESC_AEAD_NULL_DEC_LEN		(DESC_AEAD_NULL_BASE + 17 * CAAM_CMD_SZ)
 
+#define DESC_GCM_BASE			(3 * CAAM_CMD_SZ)
+#define DESC_GCM_ENC_LEN		(DESC_GCM_BASE + 23 * CAAM_CMD_SZ)
+#define DESC_GCM_DEC_LEN		(DESC_GCM_BASE + 19 * CAAM_CMD_SZ)
+
+#define DESC_RFC4106_BASE		(3 * CAAM_CMD_SZ)
+#define DESC_RFC4106_ENC_LEN		(DESC_RFC4106_BASE + 15 * CAAM_CMD_SZ)
+#define DESC_RFC4106_DEC_LEN		(DESC_RFC4106_BASE + 14 * CAAM_CMD_SZ)
+#define DESC_RFC4106_GIVENC_LEN		(DESC_RFC4106_BASE + 21 * CAAM_CMD_SZ)
+
+#define DESC_RFC4543_BASE		(3 * CAAM_CMD_SZ)
+#define DESC_RFC4543_ENC_LEN		(DESC_RFC4543_BASE + 25 * CAAM_CMD_SZ)
+#define DESC_RFC4543_DEC_LEN		(DESC_RFC4543_BASE + 27 * CAAM_CMD_SZ)
+#define DESC_RFC4543_GIVENC_LEN		(DESC_RFC4543_BASE + 30 * CAAM_CMD_SZ)
+
 #define DESC_ABLKCIPHER_BASE		(3 * CAAM_CMD_SZ)
 #define DESC_ABLKCIPHER_ENC_LEN		(DESC_ABLKCIPHER_BASE + \
 					 20 * CAAM_CMD_SZ)
 #define DESC_ABLKCIPHER_DEC_LEN		(DESC_ABLKCIPHER_BASE + \
 					 15 * CAAM_CMD_SZ)
 
-#define DESC_MAX_USED_BYTES		(DESC_AEAD_GIVENC_LEN + \
+#define DESC_MAX_USED_BYTES		(DESC_RFC4543_GIVENC_LEN + \
 					 CAAM_MAX_KEY_SIZE)
 #define DESC_MAX_USED_LEN		(DESC_MAX_USED_BYTES / CAAM_CMD_SZ)
 
@@ -128,11 +146,13 @@ static inline void aead_append_src_dst(u32 *desc, u32 msg_type)
 /*
  * For aead encrypt and decrypt, read iv for both classes
  */
-static inline void aead_append_ld_iv(u32 *desc, int ivsize)
+static inline void aead_append_ld_iv(u32 *desc, int ivsize, int ivoffset)
 {
-	append_cmd(desc, CMD_SEQ_LOAD | LDST_SRCDST_BYTE_CONTEXT |
-		   LDST_CLASS_1_CCB | ivsize);
-	append_move(desc, MOVE_SRC_CLASS1CTX | MOVE_DEST_CLASS2INFIFO | ivsize);
+	append_seq_load(desc, ivsize, LDST_CLASS_1_CCB |
+			LDST_SRCDST_BYTE_CONTEXT |
+			(ivoffset << LDST_OFFSET_SHIFT));
+	append_move(desc, MOVE_SRC_CLASS1CTX | MOVE_DEST_CLASS2INFIFO |
+		    (ivoffset << MOVE_OFFSET_SHIFT) | ivsize);
 }
 
 /*
@@ -178,35 +198,60 @@ struct caam_ctx {
 };
 
 static void append_key_aead(u32 *desc, struct caam_ctx *ctx,
-			    int keys_fit_inline)
+			    int keys_fit_inline, bool is_rfc3686)
 {
+	u32 *nonce;
+	unsigned int enckeylen = ctx->enckeylen;
+
+	/*
+	 * RFC3686 specific:
+	 *	| ctx->key = {AUTH_KEY, ENC_KEY, NONCE}
+	 *	| enckeylen = encryption key size + nonce size
+	 */
+	if (is_rfc3686)
+		enckeylen -= CTR_RFC3686_NONCE_SIZE;
+
 	if (keys_fit_inline) {
 		append_key_as_imm(desc, ctx->key, ctx->split_key_pad_len,
 				  ctx->split_key_len, CLASS_2 |
 				  KEY_DEST_MDHA_SPLIT | KEY_ENC);
 		append_key_as_imm(desc, (void *)ctx->key +
-				  ctx->split_key_pad_len, ctx->enckeylen,
-				  ctx->enckeylen, CLASS_1 | KEY_DEST_CLASS_REG);
+				  ctx->split_key_pad_len, enckeylen,
+				  enckeylen, CLASS_1 | KEY_DEST_CLASS_REG);
 	} else {
 		append_key(desc, ctx->key_dma, ctx->split_key_len, CLASS_2 |
 			   KEY_DEST_MDHA_SPLIT | KEY_ENC);
 		append_key(desc, ctx->key_dma + ctx->split_key_pad_len,
-			   ctx->enckeylen, CLASS_1 | KEY_DEST_CLASS_REG);
+			   enckeylen, CLASS_1 | KEY_DEST_CLASS_REG);
+	}
+
+	/* Load Counter into CONTEXT1 reg */
+	if (is_rfc3686) {
+		nonce = (u32 *)((void *)ctx->key + ctx->split_key_pad_len +
+			       enckeylen);
+		append_load_imm_u32(desc, *nonce, LDST_CLASS_IND_CCB |
+				    LDST_SRCDST_BYTE_OUTFIFO | LDST_IMM);
+		append_move(desc,
+			    MOVE_SRC_OUTFIFO |
+			    MOVE_DEST_CLASS1CTX |
+			    (16 << MOVE_OFFSET_SHIFT) |
+			    (CTR_RFC3686_NONCE_SIZE << MOVE_LEN_SHIFT));
 	}
 }
 
 static void init_sh_desc_key_aead(u32 *desc, struct caam_ctx *ctx,
-				  int keys_fit_inline)
+				  int keys_fit_inline, bool is_rfc3686)
 {
 	u32 *key_jump_cmd;
 
-	init_sh_desc(desc, HDR_SHARE_SERIAL);
+	/* Note: Context registers are saved. */
+	init_sh_desc(desc, HDR_SHARE_SERIAL | HDR_SAVECTX);
 
 	/* Skip if already shared */
 	key_jump_cmd = append_jump(desc, JUMP_JSL | JUMP_TEST_ALL |
 				   JUMP_COND_SHRD);
 
-	append_key_aead(desc, ctx, keys_fit_inline);
+	append_key_aead(desc, ctx, keys_fit_inline, is_rfc3686);
 
 	set_jump_tgt_here(desc, key_jump_cmd);
 }
@@ -406,10 +451,17 @@ static int aead_set_sh_desc(struct crypto_aead *aead)
 {
 	struct aead_tfm *tfm = &aead->base.crt_aead;
 	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	struct crypto_tfm *ctfm = crypto_aead_tfm(aead);
+	const char *alg_name = crypto_tfm_alg_name(ctfm);
 	struct device *jrdev = ctx->jrdev;
-	bool keys_fit_inline = false;
+	bool keys_fit_inline;
 	u32 geniv, moveiv;
+	u32 ctx1_iv_off = 0;
 	u32 *desc;
+	const bool ctr_mode = ((ctx->class1_alg_type & OP_ALG_AAI_MASK) ==
+			       OP_ALG_AAI_CTR_MOD128);
+	const bool is_rfc3686 = (ctr_mode &&
+				 (strstr(alg_name, "rfc3686") != NULL));
 
 	if (!ctx->authsize)
 		return 0;
@@ -419,18 +471,36 @@ static int aead_set_sh_desc(struct crypto_aead *aead)
 		return aead_null_set_sh_desc(aead);
 
 	/*
+	 * AES-CTR needs to load IV in CONTEXT1 reg
+	 * at an offset of 128bits (16bytes)
+	 * CONTEXT1[255:128] = IV
+	 */
+	if (ctr_mode)
+		ctx1_iv_off = 16;
+
+	/*
+	 * RFC3686 specific:
+	 *	CONTEXT1[255:128] = {NONCE, IV, COUNTER}
+	 */
+	if (is_rfc3686)
+		ctx1_iv_off = 16 + CTR_RFC3686_NONCE_SIZE;
+
+	/*
 	 * Job Descriptor and Shared Descriptors
 	 * must all fit into the 64-word Descriptor h/w Buffer
 	 */
+	keys_fit_inline = false;
 	if (DESC_AEAD_ENC_LEN + DESC_JOB_IO_LEN +
-	    ctx->split_key_pad_len + ctx->enckeylen <=
+	    ctx->split_key_pad_len + ctx->enckeylen +
+	    (is_rfc3686 ? DESC_AEAD_CTR_RFC3686_LEN : 0) <=
 	    CAAM_DESC_BYTES_MAX)
 		keys_fit_inline = true;
 
 	/* aead_encrypt shared descriptor */
 	desc = ctx->sh_desc_enc;
 
-	init_sh_desc_key_aead(desc, ctx, keys_fit_inline);
+	/* Note: Context registers are saved. */
+	init_sh_desc_key_aead(desc, ctx, keys_fit_inline, is_rfc3686);
 
 	/* Class 2 operation */
 	append_operation(desc, ctx->class2_alg_type |
@@ -448,7 +518,15 @@ static int aead_set_sh_desc(struct crypto_aead *aead)
 	/* read assoc before reading payload */
 	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS2 | FIFOLD_TYPE_MSG |
 			     KEY_VLF);
-	aead_append_ld_iv(desc, tfm->ivsize);
+	aead_append_ld_iv(desc, tfm->ivsize, ctx1_iv_off);
+
+	/* Load Counter into CONTEXT1 reg */
+	if (is_rfc3686)
+		append_load_imm_u32(desc, be32_to_cpu(1), LDST_IMM |
+				    LDST_CLASS_1_CCB |
+				    LDST_SRCDST_BYTE_CONTEXT |
+				    ((ctx1_iv_off + CTR_RFC3686_IV_SIZE) <<
+				     LDST_OFFSET_SHIFT));
 
 	/* Class 1 operation */
 	append_operation(desc, ctx->class1_alg_type |
@@ -482,14 +560,16 @@ static int aead_set_sh_desc(struct crypto_aead *aead)
 	 */
 	keys_fit_inline = false;
 	if (DESC_AEAD_DEC_LEN + DESC_JOB_IO_LEN +
-	    ctx->split_key_pad_len + ctx->enckeylen <=
+	    ctx->split_key_pad_len + ctx->enckeylen +
+	    (is_rfc3686 ? DESC_AEAD_CTR_RFC3686_LEN : 0) <=
 	    CAAM_DESC_BYTES_MAX)
 		keys_fit_inline = true;
 
 	/* aead_decrypt shared descriptor */
 	desc = ctx->sh_desc_dec;
 
-	init_sh_desc_key_aead(desc, ctx, keys_fit_inline);
+	/* Note: Context registers are saved. */
+	init_sh_desc_key_aead(desc, ctx, keys_fit_inline, is_rfc3686);
 
 	/* Class 2 operation */
 	append_operation(desc, ctx->class2_alg_type |
@@ -506,9 +586,22 @@ static int aead_set_sh_desc(struct crypto_aead *aead)
 	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS2 | FIFOLD_TYPE_MSG |
 			     KEY_VLF);
 
-	aead_append_ld_iv(desc, tfm->ivsize);
+	aead_append_ld_iv(desc, tfm->ivsize, ctx1_iv_off);
 
-	append_dec_op1(desc, ctx->class1_alg_type);
+	/* Load Counter into CONTEXT1 reg */
+	if (is_rfc3686)
+		append_load_imm_u32(desc, be32_to_cpu(1), LDST_IMM |
+				    LDST_CLASS_1_CCB |
+				    LDST_SRCDST_BYTE_CONTEXT |
+				    ((ctx1_iv_off + CTR_RFC3686_IV_SIZE) <<
+				     LDST_OFFSET_SHIFT));
+
+	/* Choose operation */
+	if (ctr_mode)
+		append_operation(desc, ctx->class1_alg_type |
+				 OP_ALG_AS_INITFINAL | OP_ALG_DECRYPT);
+	else
+		append_dec_op1(desc, ctx->class1_alg_type);
 
 	/* Read and write cryptlen bytes */
 	append_math_add(desc, VARSEQINLEN, ZERO, REG2, CAAM_CMD_SZ);
@@ -538,14 +631,16 @@ static int aead_set_sh_desc(struct crypto_aead *aead)
 	 */
 	keys_fit_inline = false;
 	if (DESC_AEAD_GIVENC_LEN + DESC_JOB_IO_LEN +
-	    ctx->split_key_pad_len + ctx->enckeylen <=
+	    ctx->split_key_pad_len + ctx->enckeylen +
+	    (is_rfc3686 ? DESC_AEAD_CTR_RFC3686_LEN : 0) <=
 	    CAAM_DESC_BYTES_MAX)
 		keys_fit_inline = true;
 
 	/* aead_givencrypt shared descriptor */
 	desc = ctx->sh_desc_givenc;
 
-	init_sh_desc_key_aead(desc, ctx, keys_fit_inline);
+	/* Note: Context registers are saved. */
+	init_sh_desc_key_aead(desc, ctx, keys_fit_inline, is_rfc3686);
 
 	/* Generate IV */
 	geniv = NFIFOENTRY_STYPE_PAD | NFIFOENTRY_DEST_DECO |
@@ -554,13 +649,16 @@ static int aead_set_sh_desc(struct crypto_aead *aead)
 	append_load_imm_u32(desc, geniv, LDST_CLASS_IND_CCB |
 			    LDST_SRCDST_WORD_INFO_FIFO | LDST_IMM);
 	append_cmd(desc, CMD_LOAD | DISABLE_AUTO_INFO_FIFO);
-	append_move(desc, MOVE_SRC_INFIFO |
-		    MOVE_DEST_CLASS1CTX | (tfm->ivsize << MOVE_LEN_SHIFT));
+	append_move(desc, MOVE_WAITCOMP |
+		    MOVE_SRC_INFIFO | MOVE_DEST_CLASS1CTX |
+		    (ctx1_iv_off << MOVE_OFFSET_SHIFT) |
+		    (tfm->ivsize << MOVE_LEN_SHIFT));
 	append_cmd(desc, CMD_LOAD | ENABLE_AUTO_INFO_FIFO);
 
 	/* Copy IV to class 1 context */
-	append_move(desc, MOVE_SRC_CLASS1CTX |
-		    MOVE_DEST_OUTFIFO | (tfm->ivsize << MOVE_LEN_SHIFT));
+	append_move(desc, MOVE_SRC_CLASS1CTX | MOVE_DEST_OUTFIFO |
+		    (ctx1_iv_off << MOVE_OFFSET_SHIFT) |
+		    (tfm->ivsize << MOVE_LEN_SHIFT));
 
 	/* Return to encryption */
 	append_operation(desc, ctx->class2_alg_type |
@@ -576,7 +674,7 @@ static int aead_set_sh_desc(struct crypto_aead *aead)
 	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS2 | FIFOLD_TYPE_MSG |
 			     KEY_VLF);
 
-	/* Copy iv from class 1 ctx to class 2 fifo*/
+	/* Copy iv from outfifo to class 2 fifo */
 	moveiv = NFIFOENTRY_STYPE_OFIFO | NFIFOENTRY_DEST_CLASS2 |
 		 NFIFOENTRY_DTYPE_MSG | (tfm->ivsize << NFIFOENTRY_DLEN_SHIFT);
 	append_load_imm_u32(desc, moveiv, LDST_CLASS_IND_CCB |
@@ -584,6 +682,14 @@ static int aead_set_sh_desc(struct crypto_aead *aead)
 	append_load_imm_u32(desc, tfm->ivsize, LDST_CLASS_2_CCB |
 			    LDST_SRCDST_WORD_DATASZ_REG | LDST_IMM);
 
+	/* Load Counter into CONTEXT1 reg */
+	if (is_rfc3686)
+		append_load_imm_u32(desc, be32_to_cpu(1), LDST_IMM |
+				    LDST_CLASS_1_CCB |
+				    LDST_SRCDST_BYTE_CONTEXT |
+				    ((ctx1_iv_off + CTR_RFC3686_IV_SIZE) <<
+				     LDST_OFFSET_SHIFT));
+
 	/* Class 1 operation */
 	append_operation(desc, ctx->class1_alg_type |
 			 OP_ALG_AS_INITFINAL | OP_ALG_ENCRYPT);
@@ -630,6 +736,912 @@ static int aead_setauthsize(struct crypto_aead *authenc,
 	return 0;
 }
 
+static int gcm_set_sh_desc(struct crypto_aead *aead)
+{
+	struct aead_tfm *tfm = &aead->base.crt_aead;
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	struct device *jrdev = ctx->jrdev;
+	bool keys_fit_inline = false;
+	u32 *key_jump_cmd, *zero_payload_jump_cmd,
+	    *zero_assoc_jump_cmd1, *zero_assoc_jump_cmd2;
+	u32 *desc;
+
+	if (!ctx->enckeylen || !ctx->authsize)
+		return 0;
+
+	/*
+	 * AES GCM encrypt shared descriptor
+	 * Job Descriptor and Shared Descriptor
+	 * must fit into the 64-word Descriptor h/w Buffer
+	 */
+	if (DESC_GCM_ENC_LEN + DESC_JOB_IO_LEN +
+	    ctx->enckeylen <= CAAM_DESC_BYTES_MAX)
+		keys_fit_inline = true;
+
+	desc = ctx->sh_desc_enc;
+
+	init_sh_desc(desc, HDR_SHARE_SERIAL);
+
+	/* skip key loading if they are loaded due to sharing */
+	key_jump_cmd = append_jump(desc, JUMP_JSL | JUMP_TEST_ALL |
+				   JUMP_COND_SHRD | JUMP_COND_SELF);
+	if (keys_fit_inline)
+		append_key_as_imm(desc, (void *)ctx->key, ctx->enckeylen,
+				  ctx->enckeylen, CLASS_1 | KEY_DEST_CLASS_REG);
+	else
+		append_key(desc, ctx->key_dma, ctx->enckeylen,
+			   CLASS_1 | KEY_DEST_CLASS_REG);
+	set_jump_tgt_here(desc, key_jump_cmd);
+
+	/* class 1 operation */
+	append_operation(desc, ctx->class1_alg_type |
+			 OP_ALG_AS_INITFINAL | OP_ALG_ENCRYPT);
+
+	/* cryptlen = seqoutlen - authsize */
+	append_math_sub_imm_u32(desc, REG3, SEQOUTLEN, IMM, ctx->authsize);
+
+	/* assoclen + cryptlen = seqinlen - ivsize */
+	append_math_sub_imm_u32(desc, REG2, SEQINLEN, IMM, tfm->ivsize);
+
+	/* assoclen = (assoclen + cryptlen) - cryptlen */
+	append_math_sub(desc, REG1, REG2, REG3, CAAM_CMD_SZ);
+
+	/* if cryptlen is ZERO jump to zero-payload commands */
+	append_math_add(desc, VARSEQOUTLEN, ZERO, REG3, CAAM_CMD_SZ);
+	zero_payload_jump_cmd = append_jump(desc, JUMP_TEST_ALL |
+					    JUMP_COND_MATH_Z);
+	/* read IV */
+	append_seq_fifo_load(desc, tfm->ivsize, FIFOLD_CLASS_CLASS1 |
+			     FIFOLD_TYPE_IV | FIFOLD_TYPE_FLUSH1);
+
+	/* if assoclen is ZERO, skip reading the assoc data */
+	append_math_add(desc, VARSEQINLEN, ZERO, REG1, CAAM_CMD_SZ);
+	zero_assoc_jump_cmd1 = append_jump(desc, JUMP_TEST_ALL |
+					   JUMP_COND_MATH_Z);
+
+	/* read assoc data */
+	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS1 | FIFOLDST_VLF |
+			     FIFOLD_TYPE_AAD | FIFOLD_TYPE_FLUSH1);
+	set_jump_tgt_here(desc, zero_assoc_jump_cmd1);
+
+	append_math_add(desc, VARSEQINLEN, ZERO, REG3, CAAM_CMD_SZ);
+
+	/* write encrypted data */
+	append_seq_fifo_store(desc, 0, FIFOST_TYPE_MESSAGE_DATA | FIFOLDST_VLF);
+
+	/* read payload data */
+	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS1 | FIFOLDST_VLF |
+			     FIFOLD_TYPE_MSG | FIFOLD_TYPE_LAST1);
+
+	/* jump the zero-payload commands */
+	append_jump(desc, JUMP_TEST_ALL | 7);
+
+	/* zero-payload commands */
+	set_jump_tgt_here(desc, zero_payload_jump_cmd);
+
+	/* if assoclen is ZERO, jump to IV reading - is the only input data */
+	append_math_add(desc, VARSEQINLEN, ZERO, REG1, CAAM_CMD_SZ);
+	zero_assoc_jump_cmd2 = append_jump(desc, JUMP_TEST_ALL |
+					   JUMP_COND_MATH_Z);
+	/* read IV */
+	append_seq_fifo_load(desc, tfm->ivsize, FIFOLD_CLASS_CLASS1 |
+			     FIFOLD_TYPE_IV | FIFOLD_TYPE_FLUSH1);
+
+	/* read assoc data */
+	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS1 | FIFOLDST_VLF |
+			     FIFOLD_TYPE_AAD | FIFOLD_TYPE_LAST1);
+
+	/* jump to ICV writing */
+	append_jump(desc, JUMP_TEST_ALL | 2);
+
+	/* read IV - is the only input data */
+	set_jump_tgt_here(desc, zero_assoc_jump_cmd2);
+	append_seq_fifo_load(desc, tfm->ivsize, FIFOLD_CLASS_CLASS1 |
+			     FIFOLD_TYPE_IV | FIFOLD_TYPE_FLUSH1 |
+			     FIFOLD_TYPE_LAST1);
+
+	/* write ICV */
+	append_seq_store(desc, ctx->authsize, LDST_CLASS_1_CCB |
+			 LDST_SRCDST_BYTE_CONTEXT);
+
+	ctx->sh_desc_enc_dma = dma_map_single(jrdev, desc,
+					      desc_bytes(desc),
+					      DMA_TO_DEVICE);
+	if (dma_mapping_error(jrdev, ctx->sh_desc_enc_dma)) {
+		dev_err(jrdev, "unable to map shared descriptor\n");
+		return -ENOMEM;
+	}
+#ifdef DEBUG
+	print_hex_dump(KERN_ERR, "gcm enc shdesc@"__stringify(__LINE__)": ",
+		       DUMP_PREFIX_ADDRESS, 16, 4, desc,
+		       desc_bytes(desc), 1);
+#endif
+
+	/*
+	 * Job Descriptor and Shared Descriptors
+	 * must all fit into the 64-word Descriptor h/w Buffer
+	 */
+	keys_fit_inline = false;
+	if (DESC_GCM_DEC_LEN + DESC_JOB_IO_LEN +
+	    ctx->enckeylen <= CAAM_DESC_BYTES_MAX)
+		keys_fit_inline = true;
+
+	desc = ctx->sh_desc_dec;
+
+	init_sh_desc(desc, HDR_SHARE_SERIAL);
+
+	/* skip key loading if they are loaded due to sharing */
+	key_jump_cmd = append_jump(desc, JUMP_JSL |
+				   JUMP_TEST_ALL | JUMP_COND_SHRD |
+				   JUMP_COND_SELF);
+	if (keys_fit_inline)
+		append_key_as_imm(desc, (void *)ctx->key, ctx->enckeylen,
+				  ctx->enckeylen, CLASS_1 | KEY_DEST_CLASS_REG);
+	else
+		append_key(desc, ctx->key_dma, ctx->enckeylen,
+			   CLASS_1 | KEY_DEST_CLASS_REG);
+	set_jump_tgt_here(desc, key_jump_cmd);
+
+	/* class 1 operation */
+	append_operation(desc, ctx->class1_alg_type |
+			 OP_ALG_AS_INITFINAL | OP_ALG_DECRYPT | OP_ALG_ICV_ON);
+
+	/* assoclen + cryptlen = seqinlen - ivsize - icvsize */
+	append_math_sub_imm_u32(desc, REG3, SEQINLEN, IMM,
+				ctx->authsize + tfm->ivsize);
+
+	/* assoclen = (assoclen + cryptlen) - cryptlen */
+	append_math_sub(desc, REG2, SEQOUTLEN, REG0, CAAM_CMD_SZ);
+	append_math_sub(desc, REG1, REG3, REG2, CAAM_CMD_SZ);
+
+	/* read IV */
+	append_seq_fifo_load(desc, tfm->ivsize, FIFOLD_CLASS_CLASS1 |
+			     FIFOLD_TYPE_IV | FIFOLD_TYPE_FLUSH1);
+
+	/* jump to zero-payload command if cryptlen is zero */
+	append_math_add(desc, VARSEQOUTLEN, ZERO, REG2, CAAM_CMD_SZ);
+	zero_payload_jump_cmd = append_jump(desc, JUMP_TEST_ALL |
+					    JUMP_COND_MATH_Z);
+
+	append_math_add(desc, VARSEQINLEN, ZERO, REG1, CAAM_CMD_SZ);
+	/* if asoclen is ZERO, skip reading assoc data */
+	zero_assoc_jump_cmd1 = append_jump(desc, JUMP_TEST_ALL |
+					   JUMP_COND_MATH_Z);
+	/* read assoc data */
+	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS1 | FIFOLDST_VLF |
+			     FIFOLD_TYPE_AAD | FIFOLD_TYPE_FLUSH1);
+	set_jump_tgt_here(desc, zero_assoc_jump_cmd1);
+
+	append_math_add(desc, VARSEQINLEN, ZERO, REG2, CAAM_CMD_SZ);
+
+	/* store encrypted data */
+	append_seq_fifo_store(desc, 0, FIFOST_TYPE_MESSAGE_DATA | FIFOLDST_VLF);
+
+	/* read payload data */
+	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS1 | FIFOLDST_VLF |
+			     FIFOLD_TYPE_MSG | FIFOLD_TYPE_FLUSH1);
+
+	/* jump the zero-payload commands */
+	append_jump(desc, JUMP_TEST_ALL | 4);
+
+	/* zero-payload command */
+	set_jump_tgt_here(desc, zero_payload_jump_cmd);
+
+	/* if assoclen is ZERO, jump to ICV reading */
+	append_math_add(desc, VARSEQINLEN, ZERO, REG1, CAAM_CMD_SZ);
+	zero_assoc_jump_cmd2 = append_jump(desc, JUMP_TEST_ALL |
+					   JUMP_COND_MATH_Z);
+	/* read assoc data */
+	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS1 | FIFOLDST_VLF |
+			     FIFOLD_TYPE_AAD | FIFOLD_TYPE_FLUSH1);
+	set_jump_tgt_here(desc, zero_assoc_jump_cmd2);
+
+	/* read ICV */
+	append_seq_fifo_load(desc, ctx->authsize, FIFOLD_CLASS_CLASS1 |
+			     FIFOLD_TYPE_ICV | FIFOLD_TYPE_LAST1);
+
+	ctx->sh_desc_dec_dma = dma_map_single(jrdev, desc,
+					      desc_bytes(desc),
+					      DMA_TO_DEVICE);
+	if (dma_mapping_error(jrdev, ctx->sh_desc_dec_dma)) {
+		dev_err(jrdev, "unable to map shared descriptor\n");
+		return -ENOMEM;
+	}
+#ifdef DEBUG
+	print_hex_dump(KERN_ERR, "gcm dec shdesc@"__stringify(__LINE__)": ",
+		       DUMP_PREFIX_ADDRESS, 16, 4, desc,
+		       desc_bytes(desc), 1);
+#endif
+
+	return 0;
+}
+
+static int gcm_setauthsize(struct crypto_aead *authenc, unsigned int authsize)
+{
+	struct caam_ctx *ctx = crypto_aead_ctx(authenc);
+
+	ctx->authsize = authsize;
+	gcm_set_sh_desc(authenc);
+
+	return 0;
+}
+
+static int rfc4106_set_sh_desc(struct crypto_aead *aead)
+{
+	struct aead_tfm *tfm = &aead->base.crt_aead;
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	struct device *jrdev = ctx->jrdev;
+	bool keys_fit_inline = false;
+	u32 *key_jump_cmd, *move_cmd, *write_iv_cmd;
+	u32 *desc;
+	u32 geniv;
+
+	if (!ctx->enckeylen || !ctx->authsize)
+		return 0;
+
+	/*
+	 * RFC4106 encrypt shared descriptor
+	 * Job Descriptor and Shared Descriptor
+	 * must fit into the 64-word Descriptor h/w Buffer
+	 */
+	if (DESC_RFC4106_ENC_LEN + DESC_JOB_IO_LEN +
+	    ctx->enckeylen <= CAAM_DESC_BYTES_MAX)
+		keys_fit_inline = true;
+
+	desc = ctx->sh_desc_enc;
+
+	init_sh_desc(desc, HDR_SHARE_SERIAL);
+
+	/* Skip key loading if it is loaded due to sharing */
+	key_jump_cmd = append_jump(desc, JUMP_JSL | JUMP_TEST_ALL |
+				   JUMP_COND_SHRD);
+	if (keys_fit_inline)
+		append_key_as_imm(desc, (void *)ctx->key, ctx->enckeylen,
+				  ctx->enckeylen, CLASS_1 | KEY_DEST_CLASS_REG);
+	else
+		append_key(desc, ctx->key_dma, ctx->enckeylen,
+			   CLASS_1 | KEY_DEST_CLASS_REG);
+	set_jump_tgt_here(desc, key_jump_cmd);
+
+	/* Class 1 operation */
+	append_operation(desc, ctx->class1_alg_type |
+			 OP_ALG_AS_INITFINAL | OP_ALG_ENCRYPT);
+
+	/* cryptlen = seqoutlen - authsize */
+	append_math_sub_imm_u32(desc, REG3, SEQOUTLEN, IMM, ctx->authsize);
+	append_math_add(desc, VARSEQOUTLEN, ZERO, REG3, CAAM_CMD_SZ);
+
+	/* assoclen + cryptlen = seqinlen - ivsize */
+	append_math_sub_imm_u32(desc, REG2, SEQINLEN, IMM, tfm->ivsize);
+
+	/* assoclen = (assoclen + cryptlen) - cryptlen */
+	append_math_sub(desc, VARSEQINLEN, REG2, REG3, CAAM_CMD_SZ);
+
+	/* Read Salt */
+	append_fifo_load_as_imm(desc, (void *)(ctx->key + ctx->enckeylen),
+				4, FIFOLD_CLASS_CLASS1 | FIFOLD_TYPE_IV);
+	/* Read AES-GCM-ESP IV */
+	append_seq_fifo_load(desc, tfm->ivsize, FIFOLD_CLASS_CLASS1 |
+			     FIFOLD_TYPE_IV | FIFOLD_TYPE_FLUSH1);
+
+	/* Read assoc data */
+	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS1 | FIFOLDST_VLF |
+			     FIFOLD_TYPE_AAD | FIFOLD_TYPE_FLUSH1);
+
+	/* Will read cryptlen bytes */
+	append_math_add(desc, VARSEQINLEN, ZERO, REG3, CAAM_CMD_SZ);
+
+	/* Write encrypted data */
+	append_seq_fifo_store(desc, 0, FIFOST_TYPE_MESSAGE_DATA | FIFOLDST_VLF);
+
+	/* Read payload data */
+	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS1 | FIFOLDST_VLF |
+			     FIFOLD_TYPE_MSG | FIFOLD_TYPE_LAST1);
+
+	/* Write ICV */
+	append_seq_store(desc, ctx->authsize, LDST_CLASS_1_CCB |
+			 LDST_SRCDST_BYTE_CONTEXT);
+
+	ctx->sh_desc_enc_dma = dma_map_single(jrdev, desc,
+					      desc_bytes(desc),
+					      DMA_TO_DEVICE);
+	if (dma_mapping_error(jrdev, ctx->sh_desc_enc_dma)) {
+		dev_err(jrdev, "unable to map shared descriptor\n");
+		return -ENOMEM;
+	}
+#ifdef DEBUG
+	print_hex_dump(KERN_ERR, "rfc4106 enc shdesc@"__stringify(__LINE__)": ",
+		       DUMP_PREFIX_ADDRESS, 16, 4, desc,
+		       desc_bytes(desc), 1);
+#endif
+
+	/*
+	 * Job Descriptor and Shared Descriptors
+	 * must all fit into the 64-word Descriptor h/w Buffer
+	 */
+	keys_fit_inline = false;
+	if (DESC_RFC4106_DEC_LEN + DESC_JOB_IO_LEN +
+	    ctx->enckeylen <= CAAM_DESC_BYTES_MAX)
+		keys_fit_inline = true;
+
+	desc = ctx->sh_desc_dec;
+
+	init_sh_desc(desc, HDR_SHARE_SERIAL);
+
+	/* Skip key loading if it is loaded due to sharing */
+	key_jump_cmd = append_jump(desc, JUMP_JSL |
+				   JUMP_TEST_ALL | JUMP_COND_SHRD);
+	if (keys_fit_inline)
+		append_key_as_imm(desc, (void *)ctx->key, ctx->enckeylen,
+				  ctx->enckeylen, CLASS_1 | KEY_DEST_CLASS_REG);
+	else
+		append_key(desc, ctx->key_dma, ctx->enckeylen,
+			   CLASS_1 | KEY_DEST_CLASS_REG);
+	set_jump_tgt_here(desc, key_jump_cmd);
+
+	/* Class 1 operation */
+	append_operation(desc, ctx->class1_alg_type |
+			 OP_ALG_AS_INITFINAL | OP_ALG_DECRYPT | OP_ALG_ICV_ON);
+
+	/* assoclen + cryptlen = seqinlen - ivsize - icvsize */
+	append_math_sub_imm_u32(desc, REG3, SEQINLEN, IMM,
+				ctx->authsize + tfm->ivsize);
+
+	/* assoclen = (assoclen + cryptlen) - cryptlen */
+	append_math_sub(desc, REG2, SEQOUTLEN, REG0, CAAM_CMD_SZ);
+	append_math_sub(desc, VARSEQINLEN, REG3, REG2, CAAM_CMD_SZ);
+
+	/* Will write cryptlen bytes */
+	append_math_sub(desc, VARSEQOUTLEN, SEQOUTLEN, REG0, CAAM_CMD_SZ);
+
+	/* Read Salt */
+	append_fifo_load_as_imm(desc, (void *)(ctx->key + ctx->enckeylen),
+				4, FIFOLD_CLASS_CLASS1 | FIFOLD_TYPE_IV);
+	/* Read AES-GCM-ESP IV */
+	append_seq_fifo_load(desc, tfm->ivsize, FIFOLD_CLASS_CLASS1 |
+			     FIFOLD_TYPE_IV | FIFOLD_TYPE_FLUSH1);
+
+	/* Read assoc data */
+	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS1 | FIFOLDST_VLF |
+			     FIFOLD_TYPE_AAD | FIFOLD_TYPE_FLUSH1);
+
+	/* Will read cryptlen bytes */
+	append_math_add(desc, VARSEQINLEN, ZERO, REG2, CAAM_CMD_SZ);
+
+	/* Store payload data */
+	append_seq_fifo_store(desc, 0, FIFOST_TYPE_MESSAGE_DATA | FIFOLDST_VLF);
+
+	/* Read encrypted data */
+	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS1 | FIFOLDST_VLF |
+			     FIFOLD_TYPE_MSG | FIFOLD_TYPE_FLUSH1);
+
+	/* Read ICV */
+	append_seq_fifo_load(desc, ctx->authsize, FIFOLD_CLASS_CLASS1 |
+			     FIFOLD_TYPE_ICV | FIFOLD_TYPE_LAST1);
+
+	ctx->sh_desc_dec_dma = dma_map_single(jrdev, desc,
+					      desc_bytes(desc),
+					      DMA_TO_DEVICE);
+	if (dma_mapping_error(jrdev, ctx->sh_desc_dec_dma)) {
+		dev_err(jrdev, "unable to map shared descriptor\n");
+		return -ENOMEM;
+	}
+#ifdef DEBUG
+	print_hex_dump(KERN_ERR, "rfc4106 dec shdesc@"__stringify(__LINE__)": ",
+		       DUMP_PREFIX_ADDRESS, 16, 4, desc,
+		       desc_bytes(desc), 1);
+#endif
+
+	/*
+	 * Job Descriptor and Shared Descriptors
+	 * must all fit into the 64-word Descriptor h/w Buffer
+	 */
+	keys_fit_inline = false;
+	if (DESC_RFC4106_GIVENC_LEN + DESC_JOB_IO_LEN +
+	    ctx->split_key_pad_len + ctx->enckeylen <=
+	    CAAM_DESC_BYTES_MAX)
+		keys_fit_inline = true;
+
+	/* rfc4106_givencrypt shared descriptor */
+	desc = ctx->sh_desc_givenc;
+
+	init_sh_desc(desc, HDR_SHARE_SERIAL);
+
+	/* Skip key loading if it is loaded due to sharing */
+	key_jump_cmd = append_jump(desc, JUMP_JSL | JUMP_TEST_ALL |
+				   JUMP_COND_SHRD);
+	if (keys_fit_inline)
+		append_key_as_imm(desc, (void *)ctx->key, ctx->enckeylen,
+				  ctx->enckeylen, CLASS_1 | KEY_DEST_CLASS_REG);
+	else
+		append_key(desc, ctx->key_dma, ctx->enckeylen,
+			   CLASS_1 | KEY_DEST_CLASS_REG);
+	set_jump_tgt_here(desc, key_jump_cmd);
+
+	/* Generate IV */
+	geniv = NFIFOENTRY_STYPE_PAD | NFIFOENTRY_DEST_DECO |
+		NFIFOENTRY_DTYPE_MSG | NFIFOENTRY_LC1 |
+		NFIFOENTRY_PTYPE_RND | (tfm->ivsize << NFIFOENTRY_DLEN_SHIFT);
+	append_load_imm_u32(desc, geniv, LDST_CLASS_IND_CCB |
+			    LDST_SRCDST_WORD_INFO_FIFO | LDST_IMM);
+	append_cmd(desc, CMD_LOAD | DISABLE_AUTO_INFO_FIFO);
+	move_cmd = append_move(desc, MOVE_SRC_INFIFO | MOVE_DEST_DESCBUF |
+			       (tfm->ivsize << MOVE_LEN_SHIFT));
+	append_cmd(desc, CMD_LOAD | ENABLE_AUTO_INFO_FIFO);
+
+	/* Copy generated IV to OFIFO */
+	write_iv_cmd = append_move(desc, MOVE_SRC_DESCBUF | MOVE_DEST_OUTFIFO |
+				   (tfm->ivsize << MOVE_LEN_SHIFT));
+
+	/* Class 1 operation */
+	append_operation(desc, ctx->class1_alg_type |
+			 OP_ALG_AS_INITFINAL | OP_ALG_ENCRYPT);
+
+	/* ivsize + cryptlen = seqoutlen - authsize */
+	append_math_sub_imm_u32(desc, REG3, SEQOUTLEN, IMM, ctx->authsize);
+
+	/* assoclen = seqinlen - (ivsize + cryptlen) */
+	append_math_sub(desc, VARSEQINLEN, SEQINLEN, REG3, CAAM_CMD_SZ);
+
+	/* Will write ivsize + cryptlen */
+	append_math_add(desc, VARSEQOUTLEN, REG3, REG0, CAAM_CMD_SZ);
+
+	/* Read Salt and generated IV */
+	append_cmd(desc, CMD_FIFO_LOAD | FIFOLD_CLASS_CLASS1 | FIFOLD_TYPE_IV |
+		   FIFOLD_TYPE_FLUSH1 | IMMEDIATE | 12);
+	/* Append Salt */
+	append_data(desc, (void *)(ctx->key + ctx->enckeylen), 4);
+	set_move_tgt_here(desc, move_cmd);
+	set_move_tgt_here(desc, write_iv_cmd);
+	/* Blank commands. Will be overwritten by generated IV. */
+	append_cmd(desc, 0x00000000);
+	append_cmd(desc, 0x00000000);
+	/* End of blank commands */
+
+	/* No need to reload iv */
+	append_seq_fifo_load(desc, tfm->ivsize, FIFOLD_CLASS_SKIP);
+
+	/* Read assoc data */
+	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS1 | FIFOLDST_VLF |
+			     FIFOLD_TYPE_AAD | FIFOLD_TYPE_FLUSH1);
+
+	/* Will read cryptlen */
+	append_math_add(desc, VARSEQINLEN, SEQINLEN, REG0, CAAM_CMD_SZ);
+
+	/* Store generated IV and encrypted data */
+	append_seq_fifo_store(desc, 0, FIFOST_TYPE_MESSAGE_DATA | FIFOLDST_VLF);
+
+	/* Read payload data */
+	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS1 | FIFOLDST_VLF |
+			     FIFOLD_TYPE_MSG | FIFOLD_TYPE_LAST1);
+
+	/* Write ICV */
+	append_seq_store(desc, ctx->authsize, LDST_CLASS_1_CCB |
+			 LDST_SRCDST_BYTE_CONTEXT);
+
+	ctx->sh_desc_givenc_dma = dma_map_single(jrdev, desc,
+						 desc_bytes(desc),
+						 DMA_TO_DEVICE);
+	if (dma_mapping_error(jrdev, ctx->sh_desc_givenc_dma)) {
+		dev_err(jrdev, "unable to map shared descriptor\n");
+		return -ENOMEM;
+	}
+#ifdef DEBUG
+	print_hex_dump(KERN_ERR,
+		       "rfc4106 givenc shdesc@"__stringify(__LINE__)": ",
+		       DUMP_PREFIX_ADDRESS, 16, 4, desc,
+		       desc_bytes(desc), 1);
+#endif
+
+	return 0;
+}
+
+static int rfc4106_setauthsize(struct crypto_aead *authenc,
+			       unsigned int authsize)
+{
+	struct caam_ctx *ctx = crypto_aead_ctx(authenc);
+
+	ctx->authsize = authsize;
+	rfc4106_set_sh_desc(authenc);
+
+	return 0;
+}
+
+static int rfc4543_set_sh_desc(struct crypto_aead *aead)
+{
+	struct aead_tfm *tfm = &aead->base.crt_aead;
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	struct device *jrdev = ctx->jrdev;
+	bool keys_fit_inline = false;
+	u32 *key_jump_cmd, *write_iv_cmd, *write_aad_cmd;
+	u32 *read_move_cmd, *write_move_cmd;
+	u32 *desc;
+	u32 geniv;
+
+	if (!ctx->enckeylen || !ctx->authsize)
+		return 0;
+
+	/*
+	 * RFC4543 encrypt shared descriptor
+	 * Job Descriptor and Shared Descriptor
+	 * must fit into the 64-word Descriptor h/w Buffer
+	 */
+	if (DESC_RFC4543_ENC_LEN + DESC_JOB_IO_LEN +
+	    ctx->enckeylen <= CAAM_DESC_BYTES_MAX)
+		keys_fit_inline = true;
+
+	desc = ctx->sh_desc_enc;
+
+	init_sh_desc(desc, HDR_SHARE_SERIAL);
+
+	/* Skip key loading if it is loaded due to sharing */
+	key_jump_cmd = append_jump(desc, JUMP_JSL | JUMP_TEST_ALL |
+				   JUMP_COND_SHRD);
+	if (keys_fit_inline)
+		append_key_as_imm(desc, (void *)ctx->key, ctx->enckeylen,
+				  ctx->enckeylen, CLASS_1 | KEY_DEST_CLASS_REG);
+	else
+		append_key(desc, ctx->key_dma, ctx->enckeylen,
+			   CLASS_1 | KEY_DEST_CLASS_REG);
+	set_jump_tgt_here(desc, key_jump_cmd);
+
+	/* Class 1 operation */
+	append_operation(desc, ctx->class1_alg_type |
+			 OP_ALG_AS_INITFINAL | OP_ALG_ENCRYPT);
+
+	/* Load AES-GMAC ESP IV into Math1 register */
+	append_cmd(desc, CMD_SEQ_LOAD | LDST_SRCDST_WORD_DECO_MATH1 |
+		   LDST_CLASS_DECO | tfm->ivsize);
+
+	/* Wait the DMA transaction to finish */
+	append_jump(desc, JUMP_TEST_ALL | JUMP_COND_CALM |
+		    (1 << JUMP_OFFSET_SHIFT));
+
+	/* Overwrite blank immediate AES-GMAC ESP IV data */
+	write_iv_cmd = append_move(desc, MOVE_SRC_MATH1 | MOVE_DEST_DESCBUF |
+				   (tfm->ivsize << MOVE_LEN_SHIFT));
+
+	/* Overwrite blank immediate AAD data */
+	write_aad_cmd = append_move(desc, MOVE_SRC_MATH1 | MOVE_DEST_DESCBUF |
+				    (tfm->ivsize << MOVE_LEN_SHIFT));
+
+	/* cryptlen = seqoutlen - authsize */
+	append_math_sub_imm_u32(desc, REG3, SEQOUTLEN, IMM, ctx->authsize);
+
+	/* assoclen = (seqinlen - ivsize) - cryptlen */
+	append_math_sub(desc, VARSEQINLEN, SEQINLEN, REG3, CAAM_CMD_SZ);
+
+	/* Read Salt and AES-GMAC ESP IV */
+	append_cmd(desc, CMD_FIFO_LOAD | FIFOLD_CLASS_CLASS1 | IMMEDIATE |
+		   FIFOLD_TYPE_IV | FIFOLD_TYPE_FLUSH1 | (4 + tfm->ivsize));
+	/* Append Salt */
+	append_data(desc, (void *)(ctx->key + ctx->enckeylen), 4);
+	set_move_tgt_here(desc, write_iv_cmd);
+	/* Blank commands. Will be overwritten by AES-GMAC ESP IV. */
+	append_cmd(desc, 0x00000000);
+	append_cmd(desc, 0x00000000);
+	/* End of blank commands */
+
+	/* Read assoc data */
+	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS1 | FIFOLDST_VLF |
+			     FIFOLD_TYPE_AAD);
+
+	/* Will read cryptlen bytes */
+	append_math_add(desc, VARSEQINLEN, ZERO, REG3, CAAM_CMD_SZ);
+
+	/* Will write cryptlen bytes */
+	append_math_add(desc, VARSEQOUTLEN, ZERO, REG3, CAAM_CMD_SZ);
+
+	/*
+	 * MOVE_LEN opcode is not available in all SEC HW revisions,
+	 * thus need to do some magic, i.e. self-patch the descriptor
+	 * buffer.
+	 */
+	read_move_cmd = append_move(desc, MOVE_SRC_DESCBUF | MOVE_DEST_MATH3 |
+				    (0x6 << MOVE_LEN_SHIFT));
+	write_move_cmd = append_move(desc, MOVE_SRC_MATH3 | MOVE_DEST_DESCBUF |
+				     (0x8 << MOVE_LEN_SHIFT));
+
+	/* Authenticate AES-GMAC ESP IV  */
+	append_cmd(desc, CMD_FIFO_LOAD | FIFOLD_CLASS_CLASS1 | IMMEDIATE |
+		   FIFOLD_TYPE_AAD | tfm->ivsize);
+	set_move_tgt_here(desc, write_aad_cmd);
+	/* Blank commands. Will be overwritten by AES-GMAC ESP IV. */
+	append_cmd(desc, 0x00000000);
+	append_cmd(desc, 0x00000000);
+	/* End of blank commands */
+
+	/* Read and write cryptlen bytes */
+	aead_append_src_dst(desc, FIFOLD_TYPE_AAD);
+
+	set_move_tgt_here(desc, read_move_cmd);
+	set_move_tgt_here(desc, write_move_cmd);
+	append_cmd(desc, CMD_LOAD | DISABLE_AUTO_INFO_FIFO);
+	/* Move payload data to OFIFO */
+	append_move(desc, MOVE_SRC_INFIFO_CL | MOVE_DEST_OUTFIFO);
+
+	/* Write ICV */
+	append_seq_store(desc, ctx->authsize, LDST_CLASS_1_CCB |
+			 LDST_SRCDST_BYTE_CONTEXT);
+
+	ctx->sh_desc_enc_dma = dma_map_single(jrdev, desc,
+					      desc_bytes(desc),
+					      DMA_TO_DEVICE);
+	if (dma_mapping_error(jrdev, ctx->sh_desc_enc_dma)) {
+		dev_err(jrdev, "unable to map shared descriptor\n");
+		return -ENOMEM;
+	}
+#ifdef DEBUG
+	print_hex_dump(KERN_ERR, "rfc4543 enc shdesc@"__stringify(__LINE__)": ",
+		       DUMP_PREFIX_ADDRESS, 16, 4, desc,
+		       desc_bytes(desc), 1);
+#endif
+
+	/*
+	 * Job Descriptor and Shared Descriptors
+	 * must all fit into the 64-word Descriptor h/w Buffer
+	 */
+	keys_fit_inline = false;
+	if (DESC_RFC4543_DEC_LEN + DESC_JOB_IO_LEN +
+	    ctx->enckeylen <= CAAM_DESC_BYTES_MAX)
+		keys_fit_inline = true;
+
+	desc = ctx->sh_desc_dec;
+
+	init_sh_desc(desc, HDR_SHARE_SERIAL);
+
+	/* Skip key loading if it is loaded due to sharing */
+	key_jump_cmd = append_jump(desc, JUMP_JSL |
+				   JUMP_TEST_ALL | JUMP_COND_SHRD);
+	if (keys_fit_inline)
+		append_key_as_imm(desc, (void *)ctx->key, ctx->enckeylen,
+				  ctx->enckeylen, CLASS_1 | KEY_DEST_CLASS_REG);
+	else
+		append_key(desc, ctx->key_dma, ctx->enckeylen,
+			   CLASS_1 | KEY_DEST_CLASS_REG);
+	set_jump_tgt_here(desc, key_jump_cmd);
+
+	/* Class 1 operation */
+	append_operation(desc, ctx->class1_alg_type |
+			 OP_ALG_AS_INITFINAL | OP_ALG_DECRYPT | OP_ALG_ICV_ON);
+
+	/* Load AES-GMAC ESP IV into Math1 register */
+	append_cmd(desc, CMD_SEQ_LOAD | LDST_SRCDST_WORD_DECO_MATH1 |
+		   LDST_CLASS_DECO | tfm->ivsize);
+
+	/* Wait the DMA transaction to finish */
+	append_jump(desc, JUMP_TEST_ALL | JUMP_COND_CALM |
+		    (1 << JUMP_OFFSET_SHIFT));
+
+	/* assoclen + cryptlen = (seqinlen - ivsize) - icvsize */
+	append_math_sub_imm_u32(desc, REG3, SEQINLEN, IMM, ctx->authsize);
+
+	/* Overwrite blank immediate AES-GMAC ESP IV data */
+	write_iv_cmd = append_move(desc, MOVE_SRC_MATH1 | MOVE_DEST_DESCBUF |
+				   (tfm->ivsize << MOVE_LEN_SHIFT));
+
+	/* Overwrite blank immediate AAD data */
+	write_aad_cmd = append_move(desc, MOVE_SRC_MATH1 | MOVE_DEST_DESCBUF |
+				    (tfm->ivsize << MOVE_LEN_SHIFT));
+
+	/* assoclen = (assoclen + cryptlen) - cryptlen */
+	append_math_sub(desc, REG2, SEQOUTLEN, REG0, CAAM_CMD_SZ);
+	append_math_sub(desc, VARSEQINLEN, REG3, REG2, CAAM_CMD_SZ);
+
+	/*
+	 * MOVE_LEN opcode is not available in all SEC HW revisions,
+	 * thus need to do some magic, i.e. self-patch the descriptor
+	 * buffer.
+	 */
+	read_move_cmd = append_move(desc, MOVE_SRC_DESCBUF | MOVE_DEST_MATH3 |
+				    (0x6 << MOVE_LEN_SHIFT));
+	write_move_cmd = append_move(desc, MOVE_SRC_MATH3 | MOVE_DEST_DESCBUF |
+				     (0x8 << MOVE_LEN_SHIFT));
+
+	/* Read Salt and AES-GMAC ESP IV */
+	append_cmd(desc, CMD_FIFO_LOAD | FIFOLD_CLASS_CLASS1 | IMMEDIATE |
+		   FIFOLD_TYPE_IV | FIFOLD_TYPE_FLUSH1 | (4 + tfm->ivsize));
+	/* Append Salt */
+	append_data(desc, (void *)(ctx->key + ctx->enckeylen), 4);
+	set_move_tgt_here(desc, write_iv_cmd);
+	/* Blank commands. Will be overwritten by AES-GMAC ESP IV. */
+	append_cmd(desc, 0x00000000);
+	append_cmd(desc, 0x00000000);
+	/* End of blank commands */
+
+	/* Read assoc data */
+	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS1 | FIFOLDST_VLF |
+			     FIFOLD_TYPE_AAD);
+
+	/* Will read cryptlen bytes */
+	append_math_add(desc, VARSEQINLEN, ZERO, REG2, CAAM_CMD_SZ);
+
+	/* Will write cryptlen bytes */
+	append_math_add(desc, VARSEQOUTLEN, ZERO, REG2, CAAM_CMD_SZ);
+
+	/* Authenticate AES-GMAC ESP IV  */
+	append_cmd(desc, CMD_FIFO_LOAD | FIFOLD_CLASS_CLASS1 | IMMEDIATE |
+		   FIFOLD_TYPE_AAD | tfm->ivsize);
+	set_move_tgt_here(desc, write_aad_cmd);
+	/* Blank commands. Will be overwritten by AES-GMAC ESP IV. */
+	append_cmd(desc, 0x00000000);
+	append_cmd(desc, 0x00000000);
+	/* End of blank commands */
+
+	/* Store payload data */
+	append_seq_fifo_store(desc, 0, FIFOST_TYPE_MESSAGE_DATA | FIFOLDST_VLF);
+
+	/* In-snoop cryptlen data */
+	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_BOTH | FIFOLDST_VLF |
+			     FIFOLD_TYPE_AAD | FIFOLD_TYPE_LAST2FLUSH1);
+
+	set_move_tgt_here(desc, read_move_cmd);
+	set_move_tgt_here(desc, write_move_cmd);
+	append_cmd(desc, CMD_LOAD | DISABLE_AUTO_INFO_FIFO);
+	/* Move payload data to OFIFO */
+	append_move(desc, MOVE_SRC_INFIFO_CL | MOVE_DEST_OUTFIFO);
+	append_cmd(desc, CMD_LOAD | ENABLE_AUTO_INFO_FIFO);
+
+	/* Read ICV */
+	append_seq_fifo_load(desc, ctx->authsize, FIFOLD_CLASS_CLASS1 |
+			     FIFOLD_TYPE_ICV | FIFOLD_TYPE_LAST1);
+
+	ctx->sh_desc_dec_dma = dma_map_single(jrdev, desc,
+					      desc_bytes(desc),
+					      DMA_TO_DEVICE);
+	if (dma_mapping_error(jrdev, ctx->sh_desc_dec_dma)) {
+		dev_err(jrdev, "unable to map shared descriptor\n");
+		return -ENOMEM;
+	}
+#ifdef DEBUG
+	print_hex_dump(KERN_ERR, "rfc4543 dec shdesc@"__stringify(__LINE__)": ",
+		       DUMP_PREFIX_ADDRESS, 16, 4, desc,
+		       desc_bytes(desc), 1);
+#endif
+
+	/*
+	 * Job Descriptor and Shared Descriptors
+	 * must all fit into the 64-word Descriptor h/w Buffer
+	 */
+	keys_fit_inline = false;
+	if (DESC_RFC4543_GIVENC_LEN + DESC_JOB_IO_LEN +
+	    ctx->enckeylen <= CAAM_DESC_BYTES_MAX)
+		keys_fit_inline = true;
+
+	/* rfc4543_givencrypt shared descriptor */
+	desc = ctx->sh_desc_givenc;
+
+	init_sh_desc(desc, HDR_SHARE_SERIAL);
+
+	/* Skip key loading if it is loaded due to sharing */
+	key_jump_cmd = append_jump(desc, JUMP_JSL | JUMP_TEST_ALL |
+				   JUMP_COND_SHRD);
+	if (keys_fit_inline)
+		append_key_as_imm(desc, (void *)ctx->key, ctx->enckeylen,
+				  ctx->enckeylen, CLASS_1 | KEY_DEST_CLASS_REG);
+	else
+		append_key(desc, ctx->key_dma, ctx->enckeylen,
+			   CLASS_1 | KEY_DEST_CLASS_REG);
+	set_jump_tgt_here(desc, key_jump_cmd);
+
+	/* Generate IV */
+	geniv = NFIFOENTRY_STYPE_PAD | NFIFOENTRY_DEST_DECO |
+		NFIFOENTRY_DTYPE_MSG | NFIFOENTRY_LC1 |
+		NFIFOENTRY_PTYPE_RND | (tfm->ivsize << NFIFOENTRY_DLEN_SHIFT);
+	append_load_imm_u32(desc, geniv, LDST_CLASS_IND_CCB |
+			    LDST_SRCDST_WORD_INFO_FIFO | LDST_IMM);
+	append_cmd(desc, CMD_LOAD | DISABLE_AUTO_INFO_FIFO);
+	/* Move generated IV to Math1 register */
+	append_move(desc, MOVE_SRC_INFIFO | MOVE_DEST_MATH1 |
+		    (tfm->ivsize << MOVE_LEN_SHIFT));
+	append_cmd(desc, CMD_LOAD | ENABLE_AUTO_INFO_FIFO);
+
+	/* Overwrite blank immediate AES-GMAC IV data */
+	write_iv_cmd = append_move(desc, MOVE_SRC_MATH1 | MOVE_DEST_DESCBUF |
+				   (tfm->ivsize << MOVE_LEN_SHIFT));
+
+	/* Overwrite blank immediate AAD data */
+	write_aad_cmd = append_move(desc, MOVE_SRC_MATH1 | MOVE_DEST_DESCBUF |
+				    (tfm->ivsize << MOVE_LEN_SHIFT));
+
+	/* Copy generated IV to OFIFO */
+	append_move(desc, MOVE_SRC_MATH1 | MOVE_DEST_OUTFIFO |
+		    (tfm->ivsize << MOVE_LEN_SHIFT));
+
+	/* Class 1 operation */
+	append_operation(desc, ctx->class1_alg_type |
+			 OP_ALG_AS_INITFINAL | OP_ALG_ENCRYPT);
+
+	/* ivsize + cryptlen = seqoutlen - authsize */
+	append_math_sub_imm_u32(desc, REG3, SEQOUTLEN, IMM, ctx->authsize);
+
+	/* assoclen = seqinlen - (ivsize + cryptlen) */
+	append_math_sub(desc, VARSEQINLEN, SEQINLEN, REG3, CAAM_CMD_SZ);
+
+	/* Will write ivsize + cryptlen */
+	append_math_add(desc, VARSEQOUTLEN, REG3, REG0, CAAM_CMD_SZ);
+
+	/*
+	 * MOVE_LEN opcode is not available in all SEC HW revisions,
+	 * thus need to do some magic, i.e. self-patch the descriptor
+	 * buffer.
+	 */
+	read_move_cmd = append_move(desc, MOVE_SRC_DESCBUF | MOVE_DEST_MATH3 |
+				    (0x6 << MOVE_LEN_SHIFT));
+	write_move_cmd = append_move(desc, MOVE_SRC_MATH3 | MOVE_DEST_DESCBUF |
+				     (0x8 << MOVE_LEN_SHIFT));
+
+	/* Read Salt and AES-GMAC generated IV */
+	append_cmd(desc, CMD_FIFO_LOAD | FIFOLD_CLASS_CLASS1 | IMMEDIATE |
+		   FIFOLD_TYPE_IV | FIFOLD_TYPE_FLUSH1 | (4 + tfm->ivsize));
+	/* Append Salt */
+	append_data(desc, (void *)(ctx->key + ctx->enckeylen), 4);
+	set_move_tgt_here(desc, write_iv_cmd);
+	/* Blank commands. Will be overwritten by AES-GMAC generated IV. */
+	append_cmd(desc, 0x00000000);
+	append_cmd(desc, 0x00000000);
+	/* End of blank commands */
+
+	/* No need to reload iv */
+	append_seq_fifo_load(desc, tfm->ivsize, FIFOLD_CLASS_SKIP);
+
+	/* Read assoc data */
+	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_CLASS1 | FIFOLDST_VLF |
+			     FIFOLD_TYPE_AAD);
+
+	/* Will read cryptlen */
+	append_math_add(desc, VARSEQINLEN, SEQINLEN, REG0, CAAM_CMD_SZ);
+
+	/* Authenticate AES-GMAC IV  */
+	append_cmd(desc, CMD_FIFO_LOAD | FIFOLD_CLASS_CLASS1 | IMMEDIATE |
+		   FIFOLD_TYPE_AAD | tfm->ivsize);
+	set_move_tgt_here(desc, write_aad_cmd);
+	/* Blank commands. Will be overwritten by AES-GMAC IV. */
+	append_cmd(desc, 0x00000000);
+	append_cmd(desc, 0x00000000);
+	/* End of blank commands */
+
+	/* Read and write cryptlen bytes */
+	aead_append_src_dst(desc, FIFOLD_TYPE_AAD);
+
+	set_move_tgt_here(desc, read_move_cmd);
+	set_move_tgt_here(desc, write_move_cmd);
+	append_cmd(desc, CMD_LOAD | DISABLE_AUTO_INFO_FIFO);
+	/* Move payload data to OFIFO */
+	append_move(desc, MOVE_SRC_INFIFO_CL | MOVE_DEST_OUTFIFO);
+
+	/* Write ICV */
+	append_seq_store(desc, ctx->authsize, LDST_CLASS_1_CCB |
+			 LDST_SRCDST_BYTE_CONTEXT);
+
+	ctx->sh_desc_givenc_dma = dma_map_single(jrdev, desc,
+						 desc_bytes(desc),
+						 DMA_TO_DEVICE);
+	if (dma_mapping_error(jrdev, ctx->sh_desc_givenc_dma)) {
+		dev_err(jrdev, "unable to map shared descriptor\n");
+		return -ENOMEM;
+	}
+#ifdef DEBUG
+	print_hex_dump(KERN_ERR,
+		       "rfc4543 givenc shdesc@"__stringify(__LINE__)": ",
+		       DUMP_PREFIX_ADDRESS, 16, 4, desc,
+		       desc_bytes(desc), 1);
+#endif
+
+	return 0;
+}
+
+static int rfc4543_setauthsize(struct crypto_aead *authenc,
+			       unsigned int authsize)
+{
+	struct caam_ctx *ctx = crypto_aead_ctx(authenc);
+
+	ctx->authsize = authsize;
+	rfc4543_set_sh_desc(authenc);
+
+	return 0;
+}
+
 static u32 gen_split_aead_key(struct caam_ctx *ctx, const u8 *key_in,
 			      u32 authkeylen)
 {
@@ -703,20 +1715,154 @@ badkey:
 	return -EINVAL;
 }
 
+static int gcm_setkey(struct crypto_aead *aead,
+		      const u8 *key, unsigned int keylen)
+{
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	struct device *jrdev = ctx->jrdev;
+	int ret = 0;
+
+#ifdef DEBUG
+	print_hex_dump(KERN_ERR, "key in @"__stringify(__LINE__)": ",
+		       DUMP_PREFIX_ADDRESS, 16, 4, key, keylen, 1);
+#endif
+
+	memcpy(ctx->key, key, keylen);
+	ctx->key_dma = dma_map_single(jrdev, ctx->key, keylen,
+				      DMA_TO_DEVICE);
+	if (dma_mapping_error(jrdev, ctx->key_dma)) {
+		dev_err(jrdev, "unable to map key i/o memory\n");
+		return -ENOMEM;
+	}
+	ctx->enckeylen = keylen;
+
+	ret = gcm_set_sh_desc(aead);
+	if (ret) {
+		dma_unmap_single(jrdev, ctx->key_dma, ctx->enckeylen,
+				 DMA_TO_DEVICE);
+	}
+
+	return ret;
+}
+
+static int rfc4106_setkey(struct crypto_aead *aead,
+			  const u8 *key, unsigned int keylen)
+{
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	struct device *jrdev = ctx->jrdev;
+	int ret = 0;
+
+	if (keylen < 4)
+		return -EINVAL;
+
+#ifdef DEBUG
+	print_hex_dump(KERN_ERR, "key in @"__stringify(__LINE__)": ",
+		       DUMP_PREFIX_ADDRESS, 16, 4, key, keylen, 1);
+#endif
+
+	memcpy(ctx->key, key, keylen);
+
+	/*
+	 * The last four bytes of the key material are used as the salt value
+	 * in the nonce. Update the AES key length.
+	 */
+	ctx->enckeylen = keylen - 4;
+
+	ctx->key_dma = dma_map_single(jrdev, ctx->key, ctx->enckeylen,
+				      DMA_TO_DEVICE);
+	if (dma_mapping_error(jrdev, ctx->key_dma)) {
+		dev_err(jrdev, "unable to map key i/o memory\n");
+		return -ENOMEM;
+	}
+
+	ret = rfc4106_set_sh_desc(aead);
+	if (ret) {
+		dma_unmap_single(jrdev, ctx->key_dma, ctx->enckeylen,
+				 DMA_TO_DEVICE);
+	}
+
+	return ret;
+}
+
+static int rfc4543_setkey(struct crypto_aead *aead,
+			  const u8 *key, unsigned int keylen)
+{
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	struct device *jrdev = ctx->jrdev;
+	int ret = 0;
+
+	if (keylen < 4)
+		return -EINVAL;
+
+#ifdef DEBUG
+	print_hex_dump(KERN_ERR, "key in @"__stringify(__LINE__)": ",
+		       DUMP_PREFIX_ADDRESS, 16, 4, key, keylen, 1);
+#endif
+
+	memcpy(ctx->key, key, keylen);
+
+	/*
+	 * The last four bytes of the key material are used as the salt value
+	 * in the nonce. Update the AES key length.
+	 */
+	ctx->enckeylen = keylen - 4;
+
+	ctx->key_dma = dma_map_single(jrdev, ctx->key, ctx->enckeylen,
+				      DMA_TO_DEVICE);
+	if (dma_mapping_error(jrdev, ctx->key_dma)) {
+		dev_err(jrdev, "unable to map key i/o memory\n");
+		return -ENOMEM;
+	}
+
+	ret = rfc4543_set_sh_desc(aead);
+	if (ret) {
+		dma_unmap_single(jrdev, ctx->key_dma, ctx->enckeylen,
+				 DMA_TO_DEVICE);
+	}
+
+	return ret;
+}
+
 static int ablkcipher_setkey(struct crypto_ablkcipher *ablkcipher,
 			     const u8 *key, unsigned int keylen)
 {
 	struct caam_ctx *ctx = crypto_ablkcipher_ctx(ablkcipher);
-	struct ablkcipher_tfm *tfm = &ablkcipher->base.crt_ablkcipher;
+	struct ablkcipher_tfm *crt = &ablkcipher->base.crt_ablkcipher;
+	struct crypto_tfm *tfm = crypto_ablkcipher_tfm(ablkcipher);
+	const char *alg_name = crypto_tfm_alg_name(tfm);
 	struct device *jrdev = ctx->jrdev;
 	int ret = 0;
 	u32 *key_jump_cmd;
 	u32 *desc;
+	u32 *nonce;
+	u32 geniv;
+	u32 ctx1_iv_off = 0;
+	const bool ctr_mode = ((ctx->class1_alg_type & OP_ALG_AAI_MASK) ==
+			       OP_ALG_AAI_CTR_MOD128);
+	const bool is_rfc3686 = (ctr_mode &&
+				 (strstr(alg_name, "rfc3686") != NULL));
 
 #ifdef DEBUG
 	print_hex_dump(KERN_ERR, "key in @"__stringify(__LINE__)": ",
 		       DUMP_PREFIX_ADDRESS, 16, 4, key, keylen, 1);
 #endif
+	/*
+	 * AES-CTR needs to load IV in CONTEXT1 reg
+	 * at an offset of 128bits (16bytes)
+	 * CONTEXT1[255:128] = IV
+	 */
+	if (ctr_mode)
+		ctx1_iv_off = 16;
+
+	/*
+	 * RFC3686 specific:
+	 *	| CONTEXT1[255:128] = {NONCE, IV, COUNTER}
+	 *	| *key = {KEY, NONCE}
+	 */
+	if (is_rfc3686) {
+		ctx1_iv_off = 16 + CTR_RFC3686_NONCE_SIZE;
+		keylen -= CTR_RFC3686_NONCE_SIZE;
+	}
 
 	memcpy(ctx->key, key, keylen);
 	ctx->key_dma = dma_map_single(jrdev, ctx->key, keylen,
@@ -729,7 +1875,7 @@ static int ablkcipher_setkey(struct crypto_ablkcipher *ablkcipher,
 
 	/* ablkcipher_encrypt shared descriptor */
 	desc = ctx->sh_desc_enc;
-	init_sh_desc(desc, HDR_SHARE_SERIAL);
+	init_sh_desc(desc, HDR_SHARE_SERIAL | HDR_SAVECTX);
 	/* Skip if already shared */
 	key_jump_cmd = append_jump(desc, JUMP_JSL | JUMP_TEST_ALL |
 				   JUMP_COND_SHRD);
@@ -739,11 +1885,31 @@ static int ablkcipher_setkey(struct crypto_ablkcipher *ablkcipher,
 			  ctx->enckeylen, CLASS_1 |
 			  KEY_DEST_CLASS_REG);
 
+	/* Load nonce into CONTEXT1 reg */
+	if (is_rfc3686) {
+		nonce = (u32 *)(key + keylen);
+		append_load_imm_u32(desc, *nonce, LDST_CLASS_IND_CCB |
+				    LDST_SRCDST_BYTE_OUTFIFO | LDST_IMM);
+		append_move(desc, MOVE_WAITCOMP |
+			    MOVE_SRC_OUTFIFO |
+			    MOVE_DEST_CLASS1CTX |
+			    (16 << MOVE_OFFSET_SHIFT) |
+			    (CTR_RFC3686_NONCE_SIZE << MOVE_LEN_SHIFT));
+	}
+
 	set_jump_tgt_here(desc, key_jump_cmd);
 
 	/* Load iv */
-	append_cmd(desc, CMD_SEQ_LOAD | LDST_SRCDST_BYTE_CONTEXT |
-		   LDST_CLASS_1_CCB | tfm->ivsize);
+	append_seq_load(desc, crt->ivsize, LDST_SRCDST_BYTE_CONTEXT |
+			LDST_CLASS_1_CCB | (ctx1_iv_off << LDST_OFFSET_SHIFT));
+
+	/* Load counter into CONTEXT1 reg */
+	if (is_rfc3686)
+		append_load_imm_u32(desc, be32_to_cpu(1), LDST_IMM |
+				    LDST_CLASS_1_CCB |
+				    LDST_SRCDST_BYTE_CONTEXT |
+				    ((ctx1_iv_off + CTR_RFC3686_IV_SIZE) <<
+				     LDST_OFFSET_SHIFT));
 
 	/* Load operation */
 	append_operation(desc, ctx->class1_alg_type |
@@ -768,7 +1934,7 @@ static int ablkcipher_setkey(struct crypto_ablkcipher *ablkcipher,
 	/* ablkcipher_decrypt shared descriptor */
 	desc = ctx->sh_desc_dec;
 
-	init_sh_desc(desc, HDR_SHARE_SERIAL);
+	init_sh_desc(desc, HDR_SHARE_SERIAL | HDR_SAVECTX);
 	/* Skip if already shared */
 	key_jump_cmd = append_jump(desc, JUMP_JSL | JUMP_TEST_ALL |
 				   JUMP_COND_SHRD);
@@ -778,14 +1944,38 @@ static int ablkcipher_setkey(struct crypto_ablkcipher *ablkcipher,
 			  ctx->enckeylen, CLASS_1 |
 			  KEY_DEST_CLASS_REG);
 
+	/* Load nonce into CONTEXT1 reg */
+	if (is_rfc3686) {
+		nonce = (u32 *)(key + keylen);
+		append_load_imm_u32(desc, *nonce, LDST_CLASS_IND_CCB |
+				    LDST_SRCDST_BYTE_OUTFIFO | LDST_IMM);
+		append_move(desc, MOVE_WAITCOMP |
+			    MOVE_SRC_OUTFIFO |
+			    MOVE_DEST_CLASS1CTX |
+			    (16 << MOVE_OFFSET_SHIFT) |
+			    (CTR_RFC3686_NONCE_SIZE << MOVE_LEN_SHIFT));
+	}
+
 	set_jump_tgt_here(desc, key_jump_cmd);
 
 	/* load IV */
-	append_cmd(desc, CMD_SEQ_LOAD | LDST_SRCDST_BYTE_CONTEXT |
-		   LDST_CLASS_1_CCB | tfm->ivsize);
+	append_seq_load(desc, crt->ivsize, LDST_SRCDST_BYTE_CONTEXT |
+			LDST_CLASS_1_CCB | (ctx1_iv_off << LDST_OFFSET_SHIFT));
+
+	/* Load counter into CONTEXT1 reg */
+	if (is_rfc3686)
+		append_load_imm_u32(desc, be32_to_cpu(1), LDST_IMM |
+				    LDST_CLASS_1_CCB |
+				    LDST_SRCDST_BYTE_CONTEXT |
+				    ((ctx1_iv_off + CTR_RFC3686_IV_SIZE) <<
+				     LDST_OFFSET_SHIFT));
 
 	/* Choose operation */
-	append_dec_op1(desc, ctx->class1_alg_type);
+	if (ctr_mode)
+		append_operation(desc, ctx->class1_alg_type |
+				 OP_ALG_AS_INITFINAL | OP_ALG_DECRYPT);
+	else
+		append_dec_op1(desc, ctx->class1_alg_type);
 
 	/* Perform operation */
 	ablkcipher_append_src_dst(desc);
@@ -804,6 +1994,83 @@ static int ablkcipher_setkey(struct crypto_ablkcipher *ablkcipher,
 		       DUMP_PREFIX_ADDRESS, 16, 4, desc,
 		       desc_bytes(desc), 1);
 #endif
+	/* ablkcipher_givencrypt shared descriptor */
+	desc = ctx->sh_desc_givenc;
+
+	init_sh_desc(desc, HDR_SHARE_SERIAL | HDR_SAVECTX);
+	/* Skip if already shared */
+	key_jump_cmd = append_jump(desc, JUMP_JSL | JUMP_TEST_ALL |
+				   JUMP_COND_SHRD);
+
+	/* Load class1 key only */
+	append_key_as_imm(desc, (void *)ctx->key, ctx->enckeylen,
+			  ctx->enckeylen, CLASS_1 |
+			  KEY_DEST_CLASS_REG);
+
+	/* Load Nonce into CONTEXT1 reg */
+	if (is_rfc3686) {
+		nonce = (u32 *)(key + keylen);
+		append_load_imm_u32(desc, *nonce, LDST_CLASS_IND_CCB |
+				    LDST_SRCDST_BYTE_OUTFIFO | LDST_IMM);
+		append_move(desc, MOVE_WAITCOMP |
+			    MOVE_SRC_OUTFIFO |
+			    MOVE_DEST_CLASS1CTX |
+			    (16 << MOVE_OFFSET_SHIFT) |
+			    (CTR_RFC3686_NONCE_SIZE << MOVE_LEN_SHIFT));
+	}
+	set_jump_tgt_here(desc, key_jump_cmd);
+
+	/* Generate IV */
+	geniv = NFIFOENTRY_STYPE_PAD | NFIFOENTRY_DEST_DECO |
+		NFIFOENTRY_DTYPE_MSG | NFIFOENTRY_LC1 |
+		NFIFOENTRY_PTYPE_RND | (crt->ivsize << NFIFOENTRY_DLEN_SHIFT);
+	append_load_imm_u32(desc, geniv, LDST_CLASS_IND_CCB |
+			    LDST_SRCDST_WORD_INFO_FIFO | LDST_IMM);
+	append_cmd(desc, CMD_LOAD | DISABLE_AUTO_INFO_FIFO);
+	append_move(desc, MOVE_WAITCOMP |
+		    MOVE_SRC_INFIFO |
+		    MOVE_DEST_CLASS1CTX |
+		    (crt->ivsize << MOVE_LEN_SHIFT) |
+		    (ctx1_iv_off << MOVE_OFFSET_SHIFT));
+	append_cmd(desc, CMD_LOAD | ENABLE_AUTO_INFO_FIFO);
+
+	/* Copy generated IV to memory */
+	append_seq_store(desc, crt->ivsize,
+			 LDST_SRCDST_BYTE_CONTEXT | LDST_CLASS_1_CCB |
+			 (ctx1_iv_off << LDST_OFFSET_SHIFT));
+
+	/* Load Counter into CONTEXT1 reg */
+	if (is_rfc3686)
+		append_load_imm_u32(desc, (u32)1, LDST_IMM |
+				    LDST_CLASS_1_CCB |
+				    LDST_SRCDST_BYTE_CONTEXT |
+				    ((ctx1_iv_off + CTR_RFC3686_IV_SIZE) <<
+				     LDST_OFFSET_SHIFT));
+
+	if (ctx1_iv_off)
+		append_jump(desc, JUMP_JSL | JUMP_TEST_ALL | JUMP_COND_NCP |
+			    (1 << JUMP_OFFSET_SHIFT));
+
+	/* Load operation */
+	append_operation(desc, ctx->class1_alg_type |
+			 OP_ALG_AS_INITFINAL | OP_ALG_ENCRYPT);
+
+	/* Perform operation */
+	ablkcipher_append_src_dst(desc);
+
+	ctx->sh_desc_givenc_dma = dma_map_single(jrdev, desc,
+						 desc_bytes(desc),
+						 DMA_TO_DEVICE);
+	if (dma_mapping_error(jrdev, ctx->sh_desc_givenc_dma)) {
+		dev_err(jrdev, "unable to map shared descriptor\n");
+		return -ENOMEM;
+	}
+#ifdef DEBUG
+	print_hex_dump(KERN_ERR,
+		       "ablkcipher givenc shdesc@" __stringify(__LINE__) ": ",
+		       DUMP_PREFIX_ADDRESS, 16, 4, desc,
+		       desc_bytes(desc), 1);
+#endif
 
 	return ret;
 }
@@ -1088,6 +2355,7 @@ static void init_aead_job(u32 *sh_desc, dma_addr_t ptr,
 	u32 out_options = 0, in_options;
 	dma_addr_t dst_dma, src_dma;
 	int len, sec4_sg_index = 0;
+	bool is_gcm = false;
 
 #ifdef DEBUG
 	debug("assoclen %d cryptlen %d authsize %d\n",
@@ -1106,11 +2374,19 @@ static void init_aead_job(u32 *sh_desc, dma_addr_t ptr,
 		       desc_bytes(sh_desc), 1);
 #endif
 
+	if (((ctx->class1_alg_type & OP_ALG_ALGSEL_MASK) ==
+	      OP_ALG_ALGSEL_AES) &&
+	    ((ctx->class1_alg_type & OP_ALG_AAI_MASK) == OP_ALG_AAI_GCM))
+		is_gcm = true;
+
 	len = desc_len(sh_desc);
 	init_job_desc_shared(desc, ptr, len, HDR_SHARE_DEFER | HDR_REVERSE);
 
 	if (all_contig) {
-		src_dma = sg_dma_address(req->assoc);
+		if (is_gcm)
+			src_dma = edesc->iv_dma;
+		else
+			src_dma = sg_dma_address(req->assoc);
 		in_options = 0;
 	} else {
 		src_dma = edesc->sec4_sg_dma;
@@ -1164,6 +2440,7 @@ static void init_aead_giv_job(u32 *sh_desc, dma_addr_t ptr,
 	u32 out_options = 0, in_options;
 	dma_addr_t dst_dma, src_dma;
 	int len, sec4_sg_index = 0;
+	bool is_gcm = false;
 
 #ifdef DEBUG
 	debug("assoclen %d cryptlen %d authsize %d\n",
@@ -1181,11 +2458,19 @@ static void init_aead_giv_job(u32 *sh_desc, dma_addr_t ptr,
 		       desc_bytes(sh_desc), 1);
 #endif
 
+	if (((ctx->class1_alg_type & OP_ALG_ALGSEL_MASK) ==
+	      OP_ALG_ALGSEL_AES) &&
+	    ((ctx->class1_alg_type & OP_ALG_AAI_MASK) == OP_ALG_AAI_GCM))
+		is_gcm = true;
+
 	len = desc_len(sh_desc);
 	init_job_desc_shared(desc, ptr, len, HDR_SHARE_DEFER | HDR_REVERSE);
 
 	if (contig & GIV_SRC_CONTIG) {
-		src_dma = sg_dma_address(req->assoc);
+		if (is_gcm)
+			src_dma = edesc->iv_dma;
+		else
+			src_dma = sg_dma_address(req->assoc);
 		in_options = 0;
 	} else {
 		src_dma = edesc->sec4_sg_dma;
@@ -1200,7 +2485,8 @@ static void init_aead_giv_job(u32 *sh_desc, dma_addr_t ptr,
 	} else {
 		if (likely(req->src == req->dst)) {
 			dst_dma = src_dma + sizeof(struct sec4_sg_entry) *
-				  edesc->assoc_nents;
+				  (edesc->assoc_nents +
+				   (is_gcm ? 1 + edesc->src_nents : 0));
 			out_options = LDST_SGF;
 		} else {
 			dst_dma = edesc->sec4_sg_dma +
@@ -1272,6 +2558,54 @@ static void init_ablkcipher_job(u32 *sh_desc, dma_addr_t ptr,
 }
 
 /*
+ * Fill in ablkcipher givencrypt job descriptor
+ */
+static void init_ablkcipher_giv_job(u32 *sh_desc, dma_addr_t ptr,
+				    struct ablkcipher_edesc *edesc,
+				    struct ablkcipher_request *req,
+				    bool iv_contig)
+{
+	struct crypto_ablkcipher *ablkcipher = crypto_ablkcipher_reqtfm(req);
+	int ivsize = crypto_ablkcipher_ivsize(ablkcipher);
+	u32 *desc = edesc->hw_desc;
+	u32 out_options, in_options;
+	dma_addr_t dst_dma, src_dma;
+	int len, sec4_sg_index = 0;
+
+#ifdef DEBUG
+	print_hex_dump(KERN_ERR, "presciv@" __stringify(__LINE__) ": ",
+		       DUMP_PREFIX_ADDRESS, 16, 4, req->info,
+		       ivsize, 1);
+	print_hex_dump(KERN_ERR, "src    @" __stringify(__LINE__) ": ",
+		       DUMP_PREFIX_ADDRESS, 16, 4, sg_virt(req->src),
+		       edesc->src_nents ? 100 : req->nbytes, 1);
+#endif
+
+	len = desc_len(sh_desc);
+	init_job_desc_shared(desc, ptr, len, HDR_SHARE_DEFER | HDR_REVERSE);
+
+	if (!edesc->src_nents) {
+		src_dma = sg_dma_address(req->src);
+		in_options = 0;
+	} else {
+		src_dma = edesc->sec4_sg_dma;
+		sec4_sg_index += edesc->src_nents;
+		in_options = LDST_SGF;
+	}
+	append_seq_in_ptr(desc, src_dma, req->nbytes, in_options);
+
+	if (iv_contig) {
+		dst_dma = edesc->iv_dma;
+		out_options = 0;
+	} else {
+		dst_dma = edesc->sec4_sg_dma +
+			  sec4_sg_index * sizeof(struct sec4_sg_entry);
+		out_options = LDST_SGF;
+	}
+	append_seq_out_ptr(desc, dst_dma, req->nbytes + ivsize, out_options);
+}
+
+/*
  * allocate and map the aead extended descriptor
  */
 static struct aead_edesc *aead_edesc_alloc(struct aead_request *req,
@@ -1292,6 +2626,7 @@ static struct aead_edesc *aead_edesc_alloc(struct aead_request *req,
 	int ivsize = crypto_aead_ivsize(aead);
 	int sec4_sg_index, sec4_sg_len = 0, sec4_sg_bytes;
 	unsigned int authsize = ctx->authsize;
+	bool is_gcm = false;
 
 	assoc_nents = sg_count(req->assoc, req->assoclen, &assoc_chained);
 
@@ -1326,15 +2661,31 @@ static struct aead_edesc *aead_edesc_alloc(struct aead_request *req,
 		return ERR_PTR(-ENOMEM);
 	}
 
-	/* Check if data are contiguous */
-	if (assoc_nents || sg_dma_address(req->assoc) + req->assoclen !=
-	    iv_dma || src_nents || iv_dma + ivsize !=
-	    sg_dma_address(req->src)) {
-		all_contig = false;
+	if (((ctx->class1_alg_type & OP_ALG_ALGSEL_MASK) ==
+	      OP_ALG_ALGSEL_AES) &&
+	    ((ctx->class1_alg_type & OP_ALG_AAI_MASK) == OP_ALG_AAI_GCM))
+		is_gcm = true;
+
+	/*
+	 * Check if data are contiguous.
+	 * GCM expected input sequence: IV, AAD, text
+	 * All other - expected input sequence: AAD, IV, text
+	 */
+	if (is_gcm)
+		all_contig = (!assoc_nents &&
+			      iv_dma + ivsize == sg_dma_address(req->assoc) &&
+			      !src_nents && sg_dma_address(req->assoc) +
+			      req->assoclen == sg_dma_address(req->src));
+	else
+		all_contig = (!assoc_nents && sg_dma_address(req->assoc) +
+			      req->assoclen == iv_dma && !src_nents &&
+			      iv_dma + ivsize == sg_dma_address(req->src));
+	if (!all_contig) {
 		assoc_nents = assoc_nents ? : 1;
 		src_nents = src_nents ? : 1;
 		sec4_sg_len = assoc_nents + 1 + src_nents;
 	}
+
 	sec4_sg_len += dst_nents;
 
 	sec4_sg_bytes = sec4_sg_len * sizeof(struct sec4_sg_entry);
@@ -1361,14 +2712,26 @@ static struct aead_edesc *aead_edesc_alloc(struct aead_request *req,
 
 	sec4_sg_index = 0;
 	if (!all_contig) {
-		sg_to_sec4_sg(req->assoc,
-			      (assoc_nents ? : 1),
-			      edesc->sec4_sg +
-			      sec4_sg_index, 0);
-		sec4_sg_index += assoc_nents ? : 1;
+		if (!is_gcm) {
+			sg_to_sec4_sg(req->assoc,
+				      (assoc_nents ? : 1),
+				      edesc->sec4_sg +
+				      sec4_sg_index, 0);
+			sec4_sg_index += assoc_nents ? : 1;
+		}
+
 		dma_to_sec4_sg_one(edesc->sec4_sg + sec4_sg_index,
 				   iv_dma, ivsize, 0);
 		sec4_sg_index += 1;
+
+		if (is_gcm) {
+			sg_to_sec4_sg(req->assoc,
+				      (assoc_nents ? : 1),
+				      edesc->sec4_sg +
+				      sec4_sg_index, 0);
+			sec4_sg_index += assoc_nents ? : 1;
+		}
+
 		sg_to_sec4_sg_last(req->src,
 				   (src_nents ? : 1),
 				   edesc->sec4_sg +
@@ -1490,6 +2853,7 @@ static struct aead_edesc *aead_giv_edesc_alloc(struct aead_givcrypt_request
 	int ivsize = crypto_aead_ivsize(aead);
 	bool assoc_chained = false, src_chained = false, dst_chained = false;
 	int sec4_sg_index, sec4_sg_len = 0, sec4_sg_bytes;
+	bool is_gcm = false;
 
 	assoc_nents = sg_count(req->assoc, req->assoclen, &assoc_chained);
 	src_nents = sg_count(req->src, req->cryptlen, &src_chained);
@@ -1516,24 +2880,53 @@ static struct aead_edesc *aead_giv_edesc_alloc(struct aead_givcrypt_request
 		return ERR_PTR(-ENOMEM);
 	}
 
-	/* Check if data are contiguous */
-	if (assoc_nents || sg_dma_address(req->assoc) + req->assoclen !=
-	    iv_dma || src_nents || iv_dma + ivsize != sg_dma_address(req->src))
-		contig &= ~GIV_SRC_CONTIG;
+	if (((ctx->class1_alg_type & OP_ALG_ALGSEL_MASK) ==
+	      OP_ALG_ALGSEL_AES) &&
+	    ((ctx->class1_alg_type & OP_ALG_AAI_MASK) == OP_ALG_AAI_GCM))
+		is_gcm = true;
+
+	/*
+	 * Check if data are contiguous.
+	 * GCM expected input sequence: IV, AAD, text
+	 * All other - expected input sequence: AAD, IV, text
+	 */
+
+	if (is_gcm) {
+		if (assoc_nents || iv_dma + ivsize !=
+		    sg_dma_address(req->assoc) || src_nents ||
+		    sg_dma_address(req->assoc) + req->assoclen !=
+		    sg_dma_address(req->src))
+			contig &= ~GIV_SRC_CONTIG;
+	} else {
+		if (assoc_nents ||
+		    sg_dma_address(req->assoc) + req->assoclen != iv_dma ||
+		    src_nents || iv_dma + ivsize != sg_dma_address(req->src))
+			contig &= ~GIV_SRC_CONTIG;
+	}
+
 	if (dst_nents || iv_dma + ivsize != sg_dma_address(req->dst))
 		contig &= ~GIV_DST_CONTIG;
-	if (unlikely(req->src != req->dst)) {
-		dst_nents = dst_nents ? : 1;
-		sec4_sg_len += 1;
-	}
+
 	if (!(contig & GIV_SRC_CONTIG)) {
 		assoc_nents = assoc_nents ? : 1;
 		src_nents = src_nents ? : 1;
 		sec4_sg_len += assoc_nents + 1 + src_nents;
-		if (likely(req->src == req->dst))
+		if (req->src == req->dst &&
+		    (src_nents || iv_dma + ivsize != sg_dma_address(req->src)))
 			contig &= ~GIV_DST_CONTIG;
 	}
-	sec4_sg_len += dst_nents;
+
+	/*
+	 * Add new sg entries for GCM output sequence.
+	 * Expected output sequence: IV, encrypted text.
+	 */
+	if (is_gcm && req->src == req->dst && !(contig & GIV_DST_CONTIG))
+		sec4_sg_len += 1 + src_nents;
+
+	if (unlikely(req->src != req->dst)) {
+		dst_nents = dst_nents ? : 1;
+		sec4_sg_len += 1 + dst_nents;
+	}
 
 	sec4_sg_bytes = sec4_sg_len * sizeof(struct sec4_sg_entry);
 
@@ -1559,18 +2952,36 @@ static struct aead_edesc *aead_giv_edesc_alloc(struct aead_givcrypt_request
 
 	sec4_sg_index = 0;
 	if (!(contig & GIV_SRC_CONTIG)) {
-		sg_to_sec4_sg(req->assoc, assoc_nents,
-			      edesc->sec4_sg +
-			      sec4_sg_index, 0);
-		sec4_sg_index += assoc_nents;
+		if (!is_gcm) {
+			sg_to_sec4_sg(req->assoc, assoc_nents,
+				      edesc->sec4_sg + sec4_sg_index, 0);
+			sec4_sg_index += assoc_nents;
+		}
+
 		dma_to_sec4_sg_one(edesc->sec4_sg + sec4_sg_index,
 				   iv_dma, ivsize, 0);
 		sec4_sg_index += 1;
+
+		if (is_gcm) {
+			sg_to_sec4_sg(req->assoc, assoc_nents,
+				      edesc->sec4_sg + sec4_sg_index, 0);
+			sec4_sg_index += assoc_nents;
+		}
+
 		sg_to_sec4_sg_last(req->src, src_nents,
 				   edesc->sec4_sg +
 				   sec4_sg_index, 0);
 		sec4_sg_index += src_nents;
 	}
+
+	if (is_gcm && req->src == req->dst && !(contig & GIV_DST_CONTIG)) {
+		dma_to_sec4_sg_one(edesc->sec4_sg + sec4_sg_index,
+				   iv_dma, ivsize, 0);
+		sec4_sg_index += 1;
+		sg_to_sec4_sg_last(req->src, src_nents,
+				   edesc->sec4_sg + sec4_sg_index, 0);
+	}
+
 	if (unlikely(req->src != req->dst && !(contig & GIV_DST_CONTIG))) {
 		dma_to_sec4_sg_one(edesc->sec4_sg + sec4_sg_index,
 				   iv_dma, ivsize, 0);
@@ -1814,6 +3225,151 @@ static int ablkcipher_decrypt(struct ablkcipher_request *req)
 	return ret;
 }
 
+/*
+ * allocate and map the ablkcipher extended descriptor
+ * for ablkcipher givencrypt
+ */
+static struct ablkcipher_edesc *ablkcipher_giv_edesc_alloc(
+				struct skcipher_givcrypt_request *greq,
+				int desc_bytes,
+				bool *iv_contig_out)
+{
+	struct ablkcipher_request *req = &greq->creq;
+	struct crypto_ablkcipher *ablkcipher = crypto_ablkcipher_reqtfm(req);
+	struct caam_ctx *ctx = crypto_ablkcipher_ctx(ablkcipher);
+	struct device *jrdev = ctx->jrdev;
+	gfp_t flags = (req->base.flags & (CRYPTO_TFM_REQ_MAY_BACKLOG |
+					  CRYPTO_TFM_REQ_MAY_SLEEP)) ?
+		       GFP_KERNEL : GFP_ATOMIC;
+	int src_nents, dst_nents = 0, sec4_sg_bytes;
+	struct ablkcipher_edesc *edesc;
+	dma_addr_t iv_dma = 0;
+	bool iv_contig = false;
+	int sgc;
+	int ivsize = crypto_ablkcipher_ivsize(ablkcipher);
+	bool src_chained = false, dst_chained = false;
+	int sec4_sg_index;
+
+	src_nents = sg_count(req->src, req->nbytes, &src_chained);
+
+	if (unlikely(req->dst != req->src))
+		dst_nents = sg_count(req->dst, req->nbytes, &dst_chained);
+
+	if (likely(req->src == req->dst)) {
+		sgc = dma_map_sg_chained(jrdev, req->src, src_nents ? : 1,
+					 DMA_BIDIRECTIONAL, src_chained);
+	} else {
+		sgc = dma_map_sg_chained(jrdev, req->src, src_nents ? : 1,
+					 DMA_TO_DEVICE, src_chained);
+		sgc = dma_map_sg_chained(jrdev, req->dst, dst_nents ? : 1,
+					 DMA_FROM_DEVICE, dst_chained);
+	}
+
+	/*
+	 * Check if iv can be contiguous with source and destination.
+	 * If so, include it. If not, create scatterlist.
+	 */
+	iv_dma = dma_map_single(jrdev, greq->giv, ivsize, DMA_TO_DEVICE);
+	if (dma_mapping_error(jrdev, iv_dma)) {
+		dev_err(jrdev, "unable to map IV\n");
+		return ERR_PTR(-ENOMEM);
+	}
+
+	if (!dst_nents && iv_dma + ivsize == sg_dma_address(req->dst))
+		iv_contig = true;
+	else
+		dst_nents = dst_nents ? : 1;
+	sec4_sg_bytes = ((iv_contig ? 0 : 1) + src_nents + dst_nents) *
+			sizeof(struct sec4_sg_entry);
+
+	/* allocate space for base edesc and hw desc commands, link tables */
+	edesc = kmalloc(sizeof(*edesc) + desc_bytes +
+			sec4_sg_bytes, GFP_DMA | flags);
+	if (!edesc) {
+		dev_err(jrdev, "could not allocate extended descriptor\n");
+		return ERR_PTR(-ENOMEM);
+	}
+
+	edesc->src_nents = src_nents;
+	edesc->src_chained = src_chained;
+	edesc->dst_nents = dst_nents;
+	edesc->dst_chained = dst_chained;
+	edesc->sec4_sg_bytes = sec4_sg_bytes;
+	edesc->sec4_sg = (void *)edesc + sizeof(struct ablkcipher_edesc) +
+			 desc_bytes;
+
+	sec4_sg_index = 0;
+	if (src_nents) {
+		sg_to_sec4_sg_last(req->src, src_nents, edesc->sec4_sg, 0);
+		sec4_sg_index += src_nents;
+	}
+
+	if (!iv_contig) {
+		dma_to_sec4_sg_one(edesc->sec4_sg + sec4_sg_index,
+				   iv_dma, ivsize, 0);
+		sec4_sg_index += 1;
+		sg_to_sec4_sg_last(req->dst, dst_nents,
+				   edesc->sec4_sg + sec4_sg_index, 0);
+	}
+
+	edesc->sec4_sg_dma = dma_map_single(jrdev, edesc->sec4_sg,
+					    sec4_sg_bytes, DMA_TO_DEVICE);
+	if (dma_mapping_error(jrdev, edesc->sec4_sg_dma)) {
+		dev_err(jrdev, "unable to map S/G table\n");
+		return ERR_PTR(-ENOMEM);
+	}
+	edesc->iv_dma = iv_dma;
+
+#ifdef DEBUG
+	print_hex_dump(KERN_ERR,
+		       "ablkcipher sec4_sg@" __stringify(__LINE__) ": ",
+		       DUMP_PREFIX_ADDRESS, 16, 4, edesc->sec4_sg,
+		       sec4_sg_bytes, 1);
+#endif
+
+	*iv_contig_out = iv_contig;
+	return edesc;
+}
+
+static int ablkcipher_givencrypt(struct skcipher_givcrypt_request *creq)
+{
+	struct ablkcipher_request *req = &creq->creq;
+	struct ablkcipher_edesc *edesc;
+	struct crypto_ablkcipher *ablkcipher = crypto_ablkcipher_reqtfm(req);
+	struct caam_ctx *ctx = crypto_ablkcipher_ctx(ablkcipher);
+	struct device *jrdev = ctx->jrdev;
+	bool iv_contig;
+	u32 *desc;
+	int ret = 0;
+
+	/* allocate extended descriptor */
+	edesc = ablkcipher_giv_edesc_alloc(creq, DESC_JOB_IO_LEN *
+				       CAAM_CMD_SZ, &iv_contig);
+	if (IS_ERR(edesc))
+		return PTR_ERR(edesc);
+
+	/* Create and submit job descriptor*/
+	init_ablkcipher_giv_job(ctx->sh_desc_givenc, ctx->sh_desc_givenc_dma,
+				edesc, req, iv_contig);
+#ifdef DEBUG
+	print_hex_dump(KERN_ERR,
+		       "ablkcipher jobdesc@" __stringify(__LINE__) ": ",
+		       DUMP_PREFIX_ADDRESS, 16, 4, edesc->hw_desc,
+		       desc_bytes(edesc->hw_desc), 1);
+#endif
+	desc = edesc->hw_desc;
+	ret = caam_jr_enqueue(jrdev, desc, ablkcipher_encrypt_done, req);
+
+	if (!ret) {
+		ret = -EINPROGRESS;
+	} else {
+		ablkcipher_unmap(jrdev, edesc, req);
+		kfree(edesc);
+	}
+
+	return ret;
+}
+
 #define template_aead		template_u.aead
 #define template_ablkcipher	template_u.ablkcipher
 struct caam_alg_template {
@@ -2309,17 +3865,188 @@ static struct caam_alg_template driver_algs[] = {
 				   OP_ALG_AAI_HMAC_PRECOMP,
 		.alg_op = OP_ALG_ALGSEL_SHA512 | OP_ALG_AAI_HMAC,
 	},
+	{
+		.name = "authenc(hmac(md5),rfc3686(ctr(aes)))",
+		.driver_name = "authenc-hmac-md5-rfc3686-ctr-aes-caam",
+		.blocksize = 1,
+		.type = CRYPTO_ALG_TYPE_AEAD,
+		.template_aead = {
+			.setkey = aead_setkey,
+			.setauthsize = aead_setauthsize,
+			.encrypt = aead_encrypt,
+			.decrypt = aead_decrypt,
+			.givencrypt = aead_givencrypt,
+			.geniv = "<built-in>",
+			.ivsize = CTR_RFC3686_IV_SIZE,
+			.maxauthsize = MD5_DIGEST_SIZE,
+			},
+		.class1_alg_type = OP_ALG_ALGSEL_AES | OP_ALG_AAI_CTR_MOD128,
+		.class2_alg_type = OP_ALG_ALGSEL_MD5 | OP_ALG_AAI_HMAC_PRECOMP,
+		.alg_op = OP_ALG_ALGSEL_MD5 | OP_ALG_AAI_HMAC,
+	},
+	{
+		.name = "authenc(hmac(sha1),rfc3686(ctr(aes)))",
+		.driver_name = "authenc-hmac-sha1-rfc3686-ctr-aes-caam",
+		.blocksize = 1,
+		.type = CRYPTO_ALG_TYPE_AEAD,
+		.template_aead = {
+			.setkey = aead_setkey,
+			.setauthsize = aead_setauthsize,
+			.encrypt = aead_encrypt,
+			.decrypt = aead_decrypt,
+			.givencrypt = aead_givencrypt,
+			.geniv = "<built-in>",
+			.ivsize = CTR_RFC3686_IV_SIZE,
+			.maxauthsize = SHA1_DIGEST_SIZE,
+			},
+		.class1_alg_type = OP_ALG_ALGSEL_AES | OP_ALG_AAI_CTR_MOD128,
+		.class2_alg_type = OP_ALG_ALGSEL_SHA1 | OP_ALG_AAI_HMAC_PRECOMP,
+		.alg_op = OP_ALG_ALGSEL_SHA1 | OP_ALG_AAI_HMAC,
+	},
+	{
+		.name = "authenc(hmac(sha224),rfc3686(ctr(aes)))",
+		.driver_name = "authenc-hmac-sha224-rfc3686-ctr-aes-caam",
+		.blocksize = 1,
+		.type = CRYPTO_ALG_TYPE_AEAD,
+		.template_aead = {
+			.setkey = aead_setkey,
+			.setauthsize = aead_setauthsize,
+			.encrypt = aead_encrypt,
+			.decrypt = aead_decrypt,
+			.givencrypt = aead_givencrypt,
+			.geniv = "<built-in>",
+			.ivsize = CTR_RFC3686_IV_SIZE,
+			.maxauthsize = SHA224_DIGEST_SIZE,
+			},
+		.class1_alg_type = OP_ALG_ALGSEL_AES | OP_ALG_AAI_CTR_MOD128,
+		.class2_alg_type = OP_ALG_ALGSEL_SHA224 |
+				   OP_ALG_AAI_HMAC_PRECOMP,
+		.alg_op = OP_ALG_ALGSEL_SHA224 | OP_ALG_AAI_HMAC,
+	},
+	{
+		.name = "authenc(hmac(sha256),rfc3686(ctr(aes)))",
+		.driver_name = "authenc-hmac-sha256-rfc3686-ctr-aes-caam",
+		.blocksize = 1,
+		.type = CRYPTO_ALG_TYPE_AEAD,
+		.template_aead = {
+			.setkey = aead_setkey,
+			.setauthsize = aead_setauthsize,
+			.encrypt = aead_encrypt,
+			.decrypt = aead_decrypt,
+			.givencrypt = aead_givencrypt,
+			.geniv = "<built-in>",
+			.ivsize = CTR_RFC3686_IV_SIZE,
+			.maxauthsize = SHA256_DIGEST_SIZE,
+			},
+		.class1_alg_type = OP_ALG_ALGSEL_AES | OP_ALG_AAI_CTR_MOD128,
+		.class2_alg_type = OP_ALG_ALGSEL_SHA256 |
+				   OP_ALG_AAI_HMAC_PRECOMP,
+		.alg_op = OP_ALG_ALGSEL_SHA256 | OP_ALG_AAI_HMAC,
+	},
+	{
+		.name = "authenc(hmac(sha384),rfc3686(ctr(aes)))",
+		.driver_name = "authenc-hmac-sha384-rfc3686-ctr-aes-caam",
+		.blocksize = 1,
+		.type = CRYPTO_ALG_TYPE_AEAD,
+		.template_aead = {
+			.setkey = aead_setkey,
+			.setauthsize = aead_setauthsize,
+			.encrypt = aead_encrypt,
+			.decrypt = aead_decrypt,
+			.givencrypt = aead_givencrypt,
+			.geniv = "<built-in>",
+			.ivsize = CTR_RFC3686_IV_SIZE,
+			.maxauthsize = SHA384_DIGEST_SIZE,
+			},
+		.class1_alg_type = OP_ALG_ALGSEL_AES | OP_ALG_AAI_CTR_MOD128,
+		.class2_alg_type = OP_ALG_ALGSEL_SHA384 |
+				   OP_ALG_AAI_HMAC_PRECOMP,
+		.alg_op = OP_ALG_ALGSEL_SHA384 | OP_ALG_AAI_HMAC,
+	},
+	{
+		.name = "authenc(hmac(sha512),rfc3686(ctr(aes)))",
+		.driver_name = "authenc-hmac-sha512-rfc3686-ctr-aes-caam",
+		.blocksize = 1,
+		.type = CRYPTO_ALG_TYPE_AEAD,
+		.template_aead = {
+			.setkey = aead_setkey,
+			.setauthsize = aead_setauthsize,
+			.encrypt = aead_encrypt,
+			.decrypt = aead_decrypt,
+			.givencrypt = aead_givencrypt,
+			.geniv = "<built-in>",
+			.ivsize = CTR_RFC3686_IV_SIZE,
+			.maxauthsize = SHA512_DIGEST_SIZE,
+			},
+		.class1_alg_type = OP_ALG_ALGSEL_AES | OP_ALG_AAI_CTR_MOD128,
+		.class2_alg_type = OP_ALG_ALGSEL_SHA512 |
+				   OP_ALG_AAI_HMAC_PRECOMP,
+		.alg_op = OP_ALG_ALGSEL_SHA512 | OP_ALG_AAI_HMAC,
+	},
+	{
+		.name = "rfc4106(gcm(aes))",
+		.driver_name = "rfc4106-gcm-aes-caam",
+		.blocksize = 1,
+		.type = CRYPTO_ALG_TYPE_AEAD,
+		.template_aead = {
+			.setkey = rfc4106_setkey,
+			.setauthsize = rfc4106_setauthsize,
+			.encrypt = aead_encrypt,
+			.decrypt = aead_decrypt,
+			.givencrypt = aead_givencrypt,
+			.geniv = "<built-in>",
+			.ivsize = 8,
+			.maxauthsize = AES_BLOCK_SIZE,
+			},
+		.class1_alg_type = OP_ALG_ALGSEL_AES | OP_ALG_AAI_GCM,
+	},
+	{
+		.name = "rfc4543(gcm(aes))",
+		.driver_name = "rfc4543-gcm-aes-caam",
+		.blocksize = 1,
+		.type = CRYPTO_ALG_TYPE_AEAD,
+		.template_aead = {
+			.setkey = rfc4543_setkey,
+			.setauthsize = rfc4543_setauthsize,
+			.encrypt = aead_encrypt,
+			.decrypt = aead_decrypt,
+			.givencrypt = aead_givencrypt,
+			.geniv = "<built-in>",
+			.ivsize = 8,
+			.maxauthsize = AES_BLOCK_SIZE,
+			},
+		.class1_alg_type = OP_ALG_ALGSEL_AES | OP_ALG_AAI_GCM,
+	},
+	/* Galois Counter Mode */
+	{
+		.name = "gcm(aes)",
+		.driver_name = "gcm-aes-caam",
+		.blocksize = 1,
+		.type = CRYPTO_ALG_TYPE_AEAD,
+		.template_aead = {
+			.setkey = gcm_setkey,
+			.setauthsize = gcm_setauthsize,
+			.encrypt = aead_encrypt,
+			.decrypt = aead_decrypt,
+			.givencrypt = NULL,
+			.geniv = "<built-in>",
+			.ivsize = 12,
+			.maxauthsize = AES_BLOCK_SIZE,
+			},
+		.class1_alg_type = OP_ALG_ALGSEL_AES | OP_ALG_AAI_GCM,
+	},
 	/* ablkcipher descriptor */
 	{
 		.name = "cbc(aes)",
 		.driver_name = "cbc-aes-caam",
 		.blocksize = AES_BLOCK_SIZE,
-		.type = CRYPTO_ALG_TYPE_ABLKCIPHER,
+		.type = CRYPTO_ALG_TYPE_GIVCIPHER,
 		.template_ablkcipher = {
 			.setkey = ablkcipher_setkey,
 			.encrypt = ablkcipher_encrypt,
 			.decrypt = ablkcipher_decrypt,
-			.geniv = "eseqiv",
+			.givencrypt = ablkcipher_givencrypt,
+			.geniv = "<built-in>",
 			.min_keysize = AES_MIN_KEY_SIZE,
 			.max_keysize = AES_MAX_KEY_SIZE,
 			.ivsize = AES_BLOCK_SIZE,
@@ -2330,12 +4057,13 @@ static struct caam_alg_template driver_algs[] = {
 		.name = "cbc(des3_ede)",
 		.driver_name = "cbc-3des-caam",
 		.blocksize = DES3_EDE_BLOCK_SIZE,
-		.type = CRYPTO_ALG_TYPE_ABLKCIPHER,
+		.type = CRYPTO_ALG_TYPE_GIVCIPHER,
 		.template_ablkcipher = {
 			.setkey = ablkcipher_setkey,
 			.encrypt = ablkcipher_encrypt,
 			.decrypt = ablkcipher_decrypt,
-			.geniv = "eseqiv",
+			.givencrypt = ablkcipher_givencrypt,
+			.geniv = "<built-in>",
 			.min_keysize = DES3_EDE_KEY_SIZE,
 			.max_keysize = DES3_EDE_KEY_SIZE,
 			.ivsize = DES3_EDE_BLOCK_SIZE,
@@ -2346,17 +4074,53 @@ static struct caam_alg_template driver_algs[] = {
 		.name = "cbc(des)",
 		.driver_name = "cbc-des-caam",
 		.blocksize = DES_BLOCK_SIZE,
-		.type = CRYPTO_ALG_TYPE_ABLKCIPHER,
+		.type = CRYPTO_ALG_TYPE_GIVCIPHER,
 		.template_ablkcipher = {
 			.setkey = ablkcipher_setkey,
 			.encrypt = ablkcipher_encrypt,
 			.decrypt = ablkcipher_decrypt,
-			.geniv = "eseqiv",
+			.givencrypt = ablkcipher_givencrypt,
+			.geniv = "<built-in>",
 			.min_keysize = DES_KEY_SIZE,
 			.max_keysize = DES_KEY_SIZE,
 			.ivsize = DES_BLOCK_SIZE,
 			},
 		.class1_alg_type = OP_ALG_ALGSEL_DES | OP_ALG_AAI_CBC,
+	},
+	{
+		.name = "ctr(aes)",
+		.driver_name = "ctr-aes-caam",
+		.blocksize = 1,
+		.type = CRYPTO_ALG_TYPE_ABLKCIPHER,
+		.template_ablkcipher = {
+			.setkey = ablkcipher_setkey,
+			.encrypt = ablkcipher_encrypt,
+			.decrypt = ablkcipher_decrypt,
+			.geniv = "chainiv",
+			.min_keysize = AES_MIN_KEY_SIZE,
+			.max_keysize = AES_MAX_KEY_SIZE,
+			.ivsize = AES_BLOCK_SIZE,
+			},
+		.class1_alg_type = OP_ALG_ALGSEL_AES | OP_ALG_AAI_CTR_MOD128,
+	},
+	{
+		.name = "rfc3686(ctr(aes))",
+		.driver_name = "rfc3686-ctr-aes-caam",
+		.blocksize = 1,
+		.type = CRYPTO_ALG_TYPE_GIVCIPHER,
+		.template_ablkcipher = {
+			.setkey = ablkcipher_setkey,
+			.encrypt = ablkcipher_encrypt,
+			.decrypt = ablkcipher_decrypt,
+			.givencrypt = ablkcipher_givencrypt,
+			.geniv = "<built-in>",
+			.min_keysize = AES_MIN_KEY_SIZE +
+				       CTR_RFC3686_NONCE_SIZE,
+			.max_keysize = AES_MAX_KEY_SIZE +
+				       CTR_RFC3686_NONCE_SIZE,
+			.ivsize = CTR_RFC3686_IV_SIZE,
+			},
+		.class1_alg_type = OP_ALG_ALGSEL_AES | OP_ALG_AAI_CTR_MOD128,
 	}
 };
 
@@ -2457,6 +4221,10 @@ static struct caam_crypto_alg *caam_alg_alloc(struct caam_alg_template
 	alg->cra_flags = CRYPTO_ALG_ASYNC | CRYPTO_ALG_KERN_DRIVER_ONLY |
 			 template->type;
 	switch (template->type) {
+	case CRYPTO_ALG_TYPE_GIVCIPHER:
+		alg->cra_type = &crypto_givcipher_type;
+		alg->cra_ablkcipher = template->template_ablkcipher;
+		break;
 	case CRYPTO_ALG_TYPE_ABLKCIPHER:
 		alg->cra_type = &crypto_ablkcipher_type;
 		alg->cra_ablkcipher = template->template_ablkcipher;
diff --git a/drivers/crypto/caam/caamhash.c b/drivers/crypto/caam/caamhash.c
index b464d03ebf40..f347ab7eea95 100644
--- a/drivers/crypto/caam/caamhash.c
+++ b/drivers/crypto/caam/caamhash.c
@@ -836,8 +836,9 @@ static int ahash_update_ctx(struct ahash_request *req)
 					   edesc->sec4_sg + sec4_sg_src_index,
 					   chained);
 			if (*next_buflen) {
-				sg_copy_part(next_buf, req->src, to_hash -
-					     *buflen, req->nbytes);
+				scatterwalk_map_and_copy(next_buf, req->src,
+							 to_hash - *buflen,
+							 *next_buflen, 0);
 				state->current_buf = !state->current_buf;
 			}
 		} else {
@@ -878,7 +879,8 @@ static int ahash_update_ctx(struct ahash_request *req)
 			kfree(edesc);
 		}
 	} else if (*next_buflen) {
-		sg_copy(buf + *buflen, req->src, req->nbytes);
+		scatterwalk_map_and_copy(buf + *buflen, req->src, 0,
+					 req->nbytes, 0);
 		*buflen = *next_buflen;
 		*next_buflen = last_buflen;
 	}
@@ -1262,8 +1264,9 @@ static int ahash_update_no_ctx(struct ahash_request *req)
 		src_map_to_sec4_sg(jrdev, req->src, src_nents,
 				   edesc->sec4_sg + 1, chained);
 		if (*next_buflen) {
-			sg_copy_part(next_buf, req->src, to_hash - *buflen,
-				    req->nbytes);
+			scatterwalk_map_and_copy(next_buf, req->src,
+						 to_hash - *buflen,
+						 *next_buflen, 0);
 			state->current_buf = !state->current_buf;
 		}
 
@@ -1304,7 +1307,8 @@ static int ahash_update_no_ctx(struct ahash_request *req)
 			kfree(edesc);
 		}
 	} else if (*next_buflen) {
-		sg_copy(buf + *buflen, req->src, req->nbytes);
+		scatterwalk_map_and_copy(buf + *buflen, req->src, 0,
+					 req->nbytes, 0);
 		*buflen = *next_buflen;
 		*next_buflen = 0;
 	}
@@ -1413,9 +1417,9 @@ static int ahash_update_first(struct ahash_request *req)
 	struct device *jrdev = ctx->jrdev;
 	gfp_t flags = (req->base.flags & (CRYPTO_TFM_REQ_MAY_BACKLOG |
 		       CRYPTO_TFM_REQ_MAY_SLEEP)) ? GFP_KERNEL : GFP_ATOMIC;
-	u8 *next_buf = state->buf_0 + state->current_buf *
-		       CAAM_MAX_HASH_BLOCK_SIZE;
-	int *next_buflen = &state->buflen_0 + state->current_buf;
+	u8 *next_buf = state->current_buf ? state->buf_1 : state->buf_0;
+	int *next_buflen = state->current_buf ?
+		&state->buflen_1 : &state->buflen_0;
 	int to_hash;
 	u32 *sh_desc = ctx->sh_desc_update_first, *desc;
 	dma_addr_t ptr = ctx->sh_desc_update_first_dma;
@@ -1476,7 +1480,8 @@ static int ahash_update_first(struct ahash_request *req)
 		}
 
 		if (*next_buflen)
-			sg_copy_part(next_buf, req->src, to_hash, req->nbytes);
+			scatterwalk_map_and_copy(next_buf, req->src, to_hash,
+						 *next_buflen, 0);
 
 		sh_len = desc_len(sh_desc);
 		desc = edesc->hw_desc;
@@ -1511,7 +1516,8 @@ static int ahash_update_first(struct ahash_request *req)
 		state->update = ahash_update_no_ctx;
 		state->finup = ahash_finup_no_ctx;
 		state->final = ahash_final_no_ctx;
-		sg_copy(next_buf, req->src, req->nbytes);
+		scatterwalk_map_and_copy(next_buf, req->src, 0,
+					 req->nbytes, 0);
 	}
 #ifdef DEBUG
 	print_hex_dump(KERN_ERR, "next buf@"__stringify(__LINE__)": ",
diff --git a/drivers/crypto/caam/compat.h b/drivers/crypto/caam/compat.h
index f227922cea38..acd7743e2603 100644
--- a/drivers/crypto/caam/compat.h
+++ b/drivers/crypto/caam/compat.h
@@ -28,6 +28,7 @@
 #include <crypto/algapi.h>
 #include <crypto/null.h>
 #include <crypto/aes.h>
+#include <crypto/ctr.h>
 #include <crypto/des.h>
 #include <crypto/sha.h>
 #include <crypto/md5.h>
diff --git a/drivers/crypto/caam/ctrl.c b/drivers/crypto/caam/ctrl.c
index 3cade79ea41e..70f1e6f37336 100644
--- a/drivers/crypto/caam/ctrl.c
+++ b/drivers/crypto/caam/ctrl.c
@@ -1,5 +1,4 @@
-/*
- * CAAM control-plane driver backend
+/* * CAAM control-plane driver backend
  * Controller-level driver, kernel property detection, initialization
  *
  * Copyright 2008-2012 Freescale Semiconductor, Inc.
@@ -81,38 +80,37 @@ static inline int run_descriptor_deco0(struct device *ctrldev, u32 *desc,
 					u32 *status)
 {
 	struct caam_drv_private *ctrlpriv = dev_get_drvdata(ctrldev);
-	struct caam_full __iomem *topregs;
+	struct caam_ctrl __iomem *ctrl = ctrlpriv->ctrl;
+	struct caam_deco __iomem *deco = ctrlpriv->deco;
 	unsigned int timeout = 100000;
 	u32 deco_dbg_reg, flags;
 	int i;
 
-	/* Set the bit to request direct access to DECO0 */
-	topregs = (struct caam_full __iomem *)ctrlpriv->ctrl;
 
 	if (ctrlpriv->virt_en == 1) {
-		setbits32(&topregs->ctrl.deco_rsr, DECORSR_JR0);
+		setbits32(&ctrl->deco_rsr, DECORSR_JR0);
 
-		while (!(rd_reg32(&topregs->ctrl.deco_rsr) & DECORSR_VALID) &&
+		while (!(rd_reg32(&ctrl->deco_rsr) & DECORSR_VALID) &&
 		       --timeout)
 			cpu_relax();
 
 		timeout = 100000;
 	}
 
-	setbits32(&topregs->ctrl.deco_rq, DECORR_RQD0ENABLE);
+	setbits32(&ctrl->deco_rq, DECORR_RQD0ENABLE);
 
-	while (!(rd_reg32(&topregs->ctrl.deco_rq) & DECORR_DEN0) &&
+	while (!(rd_reg32(&ctrl->deco_rq) & DECORR_DEN0) &&
 								 --timeout)
 		cpu_relax();
 
 	if (!timeout) {
 		dev_err(ctrldev, "failed to acquire DECO 0\n");
-		clrbits32(&topregs->ctrl.deco_rq, DECORR_RQD0ENABLE);
+		clrbits32(&ctrl->deco_rq, DECORR_RQD0ENABLE);
 		return -ENODEV;
 	}
 
 	for (i = 0; i < desc_len(desc); i++)
-		wr_reg32(&topregs->deco.descbuf[i], *(desc + i));
+		wr_reg32(&deco->descbuf[i], *(desc + i));
 
 	flags = DECO_JQCR_WHL;
 	/*
@@ -123,11 +121,11 @@ static inline int run_descriptor_deco0(struct device *ctrldev, u32 *desc,
 		flags |= DECO_JQCR_FOUR;
 
 	/* Instruct the DECO to execute it */
-	wr_reg32(&topregs->deco.jr_ctl_hi, flags);
+	wr_reg32(&deco->jr_ctl_hi, flags);
 
 	timeout = 10000000;
 	do {
-		deco_dbg_reg = rd_reg32(&topregs->deco.desc_dbg);
+		deco_dbg_reg = rd_reg32(&deco->desc_dbg);
 		/*
 		 * If an error occured in the descriptor, then
 		 * the DECO status field will be set to 0x0D
@@ -138,14 +136,14 @@ static inline int run_descriptor_deco0(struct device *ctrldev, u32 *desc,
 		cpu_relax();
 	} while ((deco_dbg_reg & DESC_DBG_DECO_STAT_VALID) && --timeout);
 
-	*status = rd_reg32(&topregs->deco.op_status_hi) &
+	*status = rd_reg32(&deco->op_status_hi) &
 		  DECO_OP_STATUS_HI_ERR_MASK;
 
 	if (ctrlpriv->virt_en == 1)
-		clrbits32(&topregs->ctrl.deco_rsr, DECORSR_JR0);
+		clrbits32(&ctrl->deco_rsr, DECORSR_JR0);
 
 	/* Mark the DECO as free */
-	clrbits32(&topregs->ctrl.deco_rq, DECORR_RQD0ENABLE);
+	clrbits32(&ctrl->deco_rq, DECORR_RQD0ENABLE);
 
 	if (!timeout)
 		return -EAGAIN;
@@ -176,13 +174,13 @@ static int instantiate_rng(struct device *ctrldev, int state_handle_mask,
 			   int gen_sk)
 {
 	struct caam_drv_private *ctrlpriv = dev_get_drvdata(ctrldev);
-	struct caam_full __iomem *topregs;
+	struct caam_ctrl __iomem *ctrl;
 	struct rng4tst __iomem *r4tst;
 	u32 *desc, status, rdsta_val;
 	int ret = 0, sh_idx;
 
-	topregs = (struct caam_full __iomem *)ctrlpriv->ctrl;
-	r4tst = &topregs->ctrl.r4tst[0];
+	ctrl = (struct caam_ctrl __iomem *)ctrlpriv->ctrl;
+	r4tst = &ctrl->r4tst[0];
 
 	desc = kmalloc(CAAM_CMD_SZ * 7, GFP_KERNEL);
 	if (!desc)
@@ -212,12 +210,11 @@ static int instantiate_rng(struct device *ctrldev, int state_handle_mask,
 		 * CAAM eras), then try again.
 		 */
 		rdsta_val =
-			rd_reg32(&topregs->ctrl.r4tst[0].rdsta) & RDSTA_IFMASK;
+			rd_reg32(&ctrl->r4tst[0].rdsta) & RDSTA_IFMASK;
 		if (status || !(rdsta_val & (1 << sh_idx)))
 			ret = -EAGAIN;
 		if (ret)
 			break;
-
 		dev_info(ctrldev, "Instantiated RNG4 SH%d\n", sh_idx);
 		/* Clear the contents before recreating the descriptor */
 		memset(desc, 0x00, CAAM_CMD_SZ * 7);
@@ -285,12 +282,12 @@ static int caam_remove(struct platform_device *pdev)
 {
 	struct device *ctrldev;
 	struct caam_drv_private *ctrlpriv;
-	struct caam_full __iomem *topregs;
+	struct caam_ctrl __iomem *ctrl;
 	int ring, ret = 0;
 
 	ctrldev = &pdev->dev;
 	ctrlpriv = dev_get_drvdata(ctrldev);
-	topregs = (struct caam_full __iomem *)ctrlpriv->ctrl;
+	ctrl = (struct caam_ctrl __iomem *)ctrlpriv->ctrl;
 
 	/* Remove platform devices for JobRs */
 	for (ring = 0; ring < ctrlpriv->total_jobrs; ring++) {
@@ -308,7 +305,7 @@ static int caam_remove(struct platform_device *pdev)
 #endif
 
 	/* Unmap controller region */
-	iounmap(&topregs->ctrl);
+	iounmap(&ctrl);
 
 	return ret;
 }
@@ -323,12 +320,12 @@ static void kick_trng(struct platform_device *pdev, int ent_delay)
 {
 	struct device *ctrldev = &pdev->dev;
 	struct caam_drv_private *ctrlpriv = dev_get_drvdata(ctrldev);
-	struct caam_full __iomem *topregs;
+	struct caam_ctrl __iomem *ctrl;
 	struct rng4tst __iomem *r4tst;
 	u32 val;
 
-	topregs = (struct caam_full __iomem *)ctrlpriv->ctrl;
-	r4tst = &topregs->ctrl.r4tst[0];
+	ctrl = (struct caam_ctrl __iomem *)ctrlpriv->ctrl;
+	r4tst = &ctrl->r4tst[0];
 
 	/* put RNG4 into program mode */
 	setbits32(&r4tst->rtmctl, RTMCTL_PRGM);
@@ -355,10 +352,19 @@ static void kick_trng(struct platform_device *pdev, int ent_delay)
 	wr_reg32(&r4tst->rtsdctl, val);
 	/* min. freq. count, equal to 1/4 of the entropy sample length */
 	wr_reg32(&r4tst->rtfrqmin, ent_delay >> 2);
-	/* max. freq. count, equal to 8 times the entropy sample length */
-	wr_reg32(&r4tst->rtfrqmax, ent_delay << 3);
+	/* disable maximum frequency count */
+	wr_reg32(&r4tst->rtfrqmax, RTFRQMAX_DISABLE);
+	/* read the control register */
+	val = rd_reg32(&r4tst->rtmctl);
+	/*
+	 * select raw sampling in both entropy shifter
+	 * and statistical checker
+	 */
+	setbits32(&val, RTMCTL_SAMP_MODE_RAW_ES_SC);
 	/* put RNG4 into run mode */
-	clrbits32(&r4tst->rtmctl, RTMCTL_PRGM);
+	clrbits32(&val, RTMCTL_PRGM);
+	/* write back the control register */
+	wr_reg32(&r4tst->rtmctl, val);
 }
 
 /**
@@ -387,13 +393,14 @@ static int caam_probe(struct platform_device *pdev)
 	struct device *dev;
 	struct device_node *nprop, *np;
 	struct caam_ctrl __iomem *ctrl;
-	struct caam_full __iomem *topregs;
 	struct caam_drv_private *ctrlpriv;
 #ifdef CONFIG_DEBUG_FS
 	struct caam_perfmon *perfmon;
 #endif
 	u32 scfgr, comp_params;
 	u32 cha_vid_ls;
+	int pg_size;
+	int BLOCK_OFFSET = 0;
 
 	ctrlpriv = devm_kzalloc(&pdev->dev, sizeof(struct caam_drv_private),
 				GFP_KERNEL);
@@ -412,10 +419,27 @@ static int caam_probe(struct platform_device *pdev)
 		dev_err(dev, "caam: of_iomap() failed\n");
 		return -ENOMEM;
 	}
-	ctrlpriv->ctrl = (struct caam_ctrl __force *)ctrl;
+	/* Finding the page size for using the CTPR_MS register */
+	comp_params = rd_reg32(&ctrl->perfmon.comp_parms_ms);
+	pg_size = (comp_params & CTPR_MS_PG_SZ_MASK) >> CTPR_MS_PG_SZ_SHIFT;
 
-	/* topregs used to derive pointers to CAAM sub-blocks only */
-	topregs = (struct caam_full __iomem *)ctrl;
+	/* Allocating the BLOCK_OFFSET based on the supported page size on
+	 * the platform
+	 */
+	if (pg_size == 0)
+		BLOCK_OFFSET = PG_SIZE_4K;
+	else
+		BLOCK_OFFSET = PG_SIZE_64K;
+
+	ctrlpriv->ctrl = (struct caam_ctrl __force *)ctrl;
+	ctrlpriv->assure = (struct caam_assurance __force *)
+			   ((uint8_t *)ctrl +
+			    BLOCK_OFFSET * ASSURE_BLOCK_NUMBER
+			   );
+	ctrlpriv->deco = (struct caam_deco __force *)
+			 ((uint8_t *)ctrl +
+			 BLOCK_OFFSET * DECO_BLOCK_NUMBER
+			 );
 
 	/* Get the IRQ of the controller (for security violations only) */
 	ctrlpriv->secvio_irq = irq_of_parse_and_map(nprop, 0);
@@ -424,15 +448,14 @@ static int caam_probe(struct platform_device *pdev)
 	 * Enable DECO watchdogs and, if this is a PHYS_ADDR_T_64BIT kernel,
 	 * long pointers in master configuration register
 	 */
-	setbits32(&topregs->ctrl.mcr, MCFGR_WDENABLE |
+	setbits32(&ctrl->mcr, MCFGR_WDENABLE |
 		  (sizeof(dma_addr_t) == sizeof(u64) ? MCFGR_LONG_PTR : 0));
 
 	/*
 	 *  Read the Compile Time paramters and SCFGR to determine
 	 * if Virtualization is enabled for this platform
 	 */
-	comp_params = rd_reg32(&topregs->ctrl.perfmon.comp_parms_ms);
-	scfgr = rd_reg32(&topregs->ctrl.scfgr);
+	scfgr = rd_reg32(&ctrl->scfgr);
 
 	ctrlpriv->virt_en = 0;
 	if (comp_params & CTPR_MS_VIRT_EN_INCL) {
@@ -450,7 +473,7 @@ static int caam_probe(struct platform_device *pdev)
 	}
 
 	if (ctrlpriv->virt_en == 1)
-		setbits32(&topregs->ctrl.jrstart, JRSTART_JR0_START |
+		setbits32(&ctrl->jrstart, JRSTART_JR0_START |
 			  JRSTART_JR1_START | JRSTART_JR2_START |
 			  JRSTART_JR3_START);
 
@@ -477,7 +500,7 @@ static int caam_probe(struct platform_device *pdev)
 					sizeof(struct platform_device *) * rspec,
 					GFP_KERNEL);
 	if (ctrlpriv->jrpdev == NULL) {
-		iounmap(&topregs->ctrl);
+		iounmap(&ctrl);
 		return -ENOMEM;
 	}
 
@@ -493,18 +516,26 @@ static int caam_probe(struct platform_device *pdev)
 					ring);
 				continue;
 			}
+			ctrlpriv->jr[ring] = (struct caam_job_ring __force *)
+					     ((uint8_t *)ctrl +
+					     (ring + JR_BLOCK_NUMBER) *
+					      BLOCK_OFFSET
+					     );
 			ctrlpriv->total_jobrs++;
 			ring++;
-		}
+	}
 
 	/* Check to see if QI present. If so, enable */
 	ctrlpriv->qi_present =
-			!!(rd_reg32(&topregs->ctrl.perfmon.comp_parms_ms) &
+			!!(rd_reg32(&ctrl->perfmon.comp_parms_ms) &
 			   CTPR_MS_QI_MASK);
 	if (ctrlpriv->qi_present) {
-		ctrlpriv->qi = (struct caam_queue_if __force *)&topregs->qi;
+		ctrlpriv->qi = (struct caam_queue_if __force *)
+			       ((uint8_t *)ctrl +
+				 BLOCK_OFFSET * QI_BLOCK_NUMBER
+			       );
 		/* This is all that's required to physically enable QI */
-		wr_reg32(&topregs->qi.qi_control_lo, QICTL_DQEN);
+		wr_reg32(&ctrlpriv->qi->qi_control_lo, QICTL_DQEN);
 	}
 
 	/* If no QI and no rings specified, quit and go home */
@@ -514,7 +545,7 @@ static int caam_probe(struct platform_device *pdev)
 		return -ENOMEM;
 	}
 
-	cha_vid_ls = rd_reg32(&topregs->ctrl.perfmon.cha_id_ls);
+	cha_vid_ls = rd_reg32(&ctrl->perfmon.cha_id_ls);
 
 	/*
 	 * If SEC has RNG version >= 4 and RNG state handle has not been
@@ -522,7 +553,7 @@ static int caam_probe(struct platform_device *pdev)
 	 */
 	if ((cha_vid_ls & CHA_ID_LS_RNG_MASK) >> CHA_ID_LS_RNG_SHIFT >= 4) {
 		ctrlpriv->rng4_sh_init =
-			rd_reg32(&topregs->ctrl.r4tst[0].rdsta);
+			rd_reg32(&ctrl->r4tst[0].rdsta);
 		/*
 		 * If the secure keys (TDKEK, JDKEK, TDSK), were already
 		 * generated, signal this to the function that is instantiating
@@ -533,7 +564,7 @@ static int caam_probe(struct platform_device *pdev)
 		ctrlpriv->rng4_sh_init &= RDSTA_IFMASK;
 		do {
 			int inst_handles =
-				rd_reg32(&topregs->ctrl.r4tst[0].rdsta) &
+				rd_reg32(&ctrl->r4tst[0].rdsta) &
 								RDSTA_IFMASK;
 			/*
 			 * If either SH were instantiated by somebody else
@@ -544,6 +575,9 @@ static int caam_probe(struct platform_device *pdev)
 			 * the TRNG parameters.
 			 */
 			if (!(ctrlpriv->rng4_sh_init || inst_handles)) {
+				dev_info(dev,
+					 "Entropy delay = %u\n",
+					 ent_delay);
 				kick_trng(pdev, ent_delay);
 				ent_delay += 400;
 			}
@@ -556,6 +590,12 @@ static int caam_probe(struct platform_device *pdev)
 			 */
 			ret = instantiate_rng(dev, inst_handles,
 					      gen_sk);
+			if (ret == -EAGAIN)
+				/*
+				 * if here, the loop will rerun,
+				 * so don't hog the CPU
+				 */
+				cpu_relax();
 		} while ((ret == -EAGAIN) && (ent_delay < RTSDCTL_ENT_DLY_MAX));
 		if (ret) {
 			dev_err(dev, "failed to instantiate RNG");
@@ -569,13 +609,13 @@ static int caam_probe(struct platform_device *pdev)
 		ctrlpriv->rng4_sh_init = ~ctrlpriv->rng4_sh_init & RDSTA_IFMASK;
 
 		/* Enable RDB bit so that RNG works faster */
-		setbits32(&topregs->ctrl.scfgr, SCFGR_RDBENABLE);
+		setbits32(&ctrl->scfgr, SCFGR_RDBENABLE);
 	}
 
 	/* NOTE: RTIC detection ought to go here, around Si time */
 
-	caam_id = (u64)rd_reg32(&topregs->ctrl.perfmon.caam_id_ms) << 32 |
-		  (u64)rd_reg32(&topregs->ctrl.perfmon.caam_id_ls);
+	caam_id = (u64)rd_reg32(&ctrl->perfmon.caam_id_ms) << 32 |
+		  (u64)rd_reg32(&ctrl->perfmon.caam_id_ls);
 
 	/* Report "alive" for developer to see */
 	dev_info(dev, "device ID = 0x%016llx (Era %d)\n", caam_id,
@@ -680,7 +720,6 @@ MODULE_DEVICE_TABLE(of, caam_match);
 static struct platform_driver caam_driver = {
 	.driver = {
 		.name = "caam",
-		.owner = THIS_MODULE,
 		.of_match_table = caam_match,
 	},
 	.probe       = caam_probe,
diff --git a/drivers/crypto/caam/desc_constr.h b/drivers/crypto/caam/desc_constr.h
index 7eec20bb3849..9f79fd7bd4d7 100644
--- a/drivers/crypto/caam/desc_constr.h
+++ b/drivers/crypto/caam/desc_constr.h
@@ -192,6 +192,8 @@ static inline void append_##cmd(u32 *desc, unsigned int len, u32 options) \
 	PRINT_POS; \
 	append_cmd(desc, CMD_##op | len | options); \
 }
+
+APPEND_CMD_LEN(seq_load, SEQ_LOAD)
 APPEND_CMD_LEN(seq_store, SEQ_STORE)
 APPEND_CMD_LEN(seq_fifo_load, SEQ_FIFO_LOAD)
 APPEND_CMD_LEN(seq_fifo_store, SEQ_FIFO_STORE)
diff --git a/drivers/crypto/caam/error.c b/drivers/crypto/caam/error.c
index 6531054a44c8..66d73bf54166 100644
--- a/drivers/crypto/caam/error.c
+++ b/drivers/crypto/caam/error.c
@@ -213,27 +213,36 @@ void caam_jr_strstatus(struct device *jrdev, u32 status)
 		void (*report_ssed)(struct device *jrdev, const u32 status,
 				    const char *error);
 		const char *error;
-	} status_src[] = {
+	} status_src[16] = {
 		{ NULL, "No error" },
 		{ NULL, NULL },
 		{ report_ccb_status, "CCB" },
 		{ report_jump_status, "Jump" },
 		{ report_deco_status, "DECO" },
-		{ NULL, NULL },
+		{ NULL, "Queue Manager Interface" },
 		{ report_jr_status, "Job Ring" },
 		{ report_cond_code_status, "Condition Code" },
+		{ NULL, NULL },
+		{ NULL, NULL },
+		{ NULL, NULL },
+		{ NULL, NULL },
+		{ NULL, NULL },
+		{ NULL, NULL },
+		{ NULL, NULL },
+		{ NULL, NULL },
 	};
 	u32 ssrc = status >> JRSTA_SSRC_SHIFT;
 	const char *error = status_src[ssrc].error;
 
 	/*
-	 * If there is no further error handling function, just
-	 * print the error code, error string and exit. Otherwise
-	 * call the handler function.
+	 * If there is an error handling function, call it to report the error.
+	 * Otherwise print the error source name.
 	 */
-	if (!status_src[ssrc].report_ssed)
-		dev_err(jrdev, "%08x: %s: \n", status, status_src[ssrc].error);
-	else
+	if (status_src[ssrc].report_ssed)
 		status_src[ssrc].report_ssed(jrdev, status, error);
+	else if (error)
+		dev_err(jrdev, "%d: %s\n", ssrc, error);
+	else
+		dev_err(jrdev, "%d: unknown error source\n", ssrc);
 }
 EXPORT_SYMBOL(caam_jr_strstatus);
diff --git a/drivers/crypto/caam/intern.h b/drivers/crypto/caam/intern.h
index 97363db4e56e..89b94cc9e7a2 100644
--- a/drivers/crypto/caam/intern.h
+++ b/drivers/crypto/caam/intern.h
@@ -70,10 +70,11 @@ struct caam_drv_private {
 	struct platform_device *pdev;
 
 	/* Physical-presence section */
-	struct caam_ctrl *ctrl; /* controller region */
-	struct caam_deco **deco; /* DECO/CCB views */
-	struct caam_assurance *ac;
-	struct caam_queue_if *qi; /* QI control region */
+	struct caam_ctrl __iomem *ctrl; /* controller region */
+	struct caam_deco __iomem *deco; /* DECO/CCB views */
+	struct caam_assurance __iomem *assure;
+	struct caam_queue_if __iomem *qi; /* QI control region */
+	struct caam_job_ring __iomem *jr[4];	/* JobR's register space */
 
 	/*
 	 * Detected geometry block. Filled in from device tree if powerpc,
diff --git a/drivers/crypto/caam/jr.c b/drivers/crypto/caam/jr.c
index 4d18e27ffa9e..9b3ef1bc9bd7 100644
--- a/drivers/crypto/caam/jr.c
+++ b/drivers/crypto/caam/jr.c
@@ -181,8 +181,6 @@ static void caam_jr_dequeue(unsigned long devarg)
 		for (i = 0; CIRC_CNT(head, tail + i, JOBR_DEPTH) >= 1; i++) {
 			sw_idx = (tail + i) & (JOBR_DEPTH - 1);
 
-			smp_read_barrier_depends();
-
 			if (jrp->outring[hw_idx].desc ==
 			    jrp->entinfo[sw_idx].desc_addr_dma)
 				break; /* found */
@@ -218,7 +216,6 @@ static void caam_jr_dequeue(unsigned long devarg)
 		if (sw_idx == tail) {
 			do {
 				tail = (tail + 1) & (JOBR_DEPTH - 1);
-				smp_read_barrier_depends();
 			} while (CIRC_CNT(head, tail, JOBR_DEPTH) >= 1 &&
 				 jrp->entinfo[tail].desc_addr_dma == 0);
 
@@ -514,7 +511,6 @@ MODULE_DEVICE_TABLE(of, caam_jr_match);
 static struct platform_driver caam_jr_driver = {
 	.driver = {
 		.name = "caam_jr",
-		.owner = THIS_MODULE,
 		.of_match_table = caam_jr_match,
 	},
 	.probe       = caam_jr_probe,
diff --git a/drivers/crypto/caam/key_gen.c b/drivers/crypto/caam/key_gen.c
index 871703c49d2c..e1eaf4ff9762 100644
--- a/drivers/crypto/caam/key_gen.c
+++ b/drivers/crypto/caam/key_gen.c
@@ -48,23 +48,29 @@ int gen_split_key(struct device *jrdev, u8 *key_out, int split_key_len,
 	u32 *desc;
 	struct split_key_result result;
 	dma_addr_t dma_addr_in, dma_addr_out;
-	int ret = 0;
+	int ret = -ENOMEM;
 
 	desc = kmalloc(CAAM_CMD_SZ * 6 + CAAM_PTR_SZ * 2, GFP_KERNEL | GFP_DMA);
 	if (!desc) {
 		dev_err(jrdev, "unable to allocate key input memory\n");
-		return -ENOMEM;
+		return ret;
 	}
 
-	init_job_desc(desc, 0);
-
 	dma_addr_in = dma_map_single(jrdev, (void *)key_in, keylen,
 				     DMA_TO_DEVICE);
 	if (dma_mapping_error(jrdev, dma_addr_in)) {
 		dev_err(jrdev, "unable to map key input memory\n");
-		kfree(desc);
-		return -ENOMEM;
+		goto out_free;
 	}
+
+	dma_addr_out = dma_map_single(jrdev, key_out, split_key_pad_len,
+				      DMA_FROM_DEVICE);
+	if (dma_mapping_error(jrdev, dma_addr_out)) {
+		dev_err(jrdev, "unable to map key output memory\n");
+		goto out_unmap_in;
+	}
+
+	init_job_desc(desc, 0);
 	append_key(desc, dma_addr_in, keylen, CLASS_2 | KEY_DEST_CLASS_REG);
 
 	/* Sets MDHA up into an HMAC-INIT */
@@ -81,13 +87,6 @@ int gen_split_key(struct device *jrdev, u8 *key_out, int split_key_len,
 	 * FIFO_STORE with the explicit split-key content store
 	 * (0x26 output type)
 	 */
-	dma_addr_out = dma_map_single(jrdev, key_out, split_key_pad_len,
-				      DMA_FROM_DEVICE);
-	if (dma_mapping_error(jrdev, dma_addr_out)) {
-		dev_err(jrdev, "unable to map key output memory\n");
-		kfree(desc);
-		return -ENOMEM;
-	}
 	append_fifo_store(desc, dma_addr_out, split_key_len,
 			  LDST_CLASS_2_CCB | FIFOST_TYPE_SPLIT_KEK);
 
@@ -115,10 +114,10 @@ int gen_split_key(struct device *jrdev, u8 *key_out, int split_key_len,
 
 	dma_unmap_single(jrdev, dma_addr_out, split_key_pad_len,
 			 DMA_FROM_DEVICE);
+out_unmap_in:
 	dma_unmap_single(jrdev, dma_addr_in, keylen, DMA_TO_DEVICE);
-
+out_free:
 	kfree(desc);
-
 	return ret;
 }
 EXPORT_SYMBOL(gen_split_key);
diff --git a/drivers/crypto/caam/regs.h b/drivers/crypto/caam/regs.h
index f48e344ffc39..378ddc17f60e 100644
--- a/drivers/crypto/caam/regs.h
+++ b/drivers/crypto/caam/regs.h
@@ -194,6 +194,8 @@ struct caam_perfmon {
 #define CTPR_MS_QI_MASK		(0x1ull << CTPR_MS_QI_SHIFT)
 #define CTPR_MS_VIRT_EN_INCL	0x00000001
 #define CTPR_MS_VIRT_EN_POR	0x00000002
+#define CTPR_MS_PG_SZ_MASK	0x10
+#define CTPR_MS_PG_SZ_SHIFT	4
 	u32 comp_parms_ms;	/* CTPR - Compile Parameters Register	*/
 	u32 comp_parms_ls;	/* CTPR - Compile Parameters Register	*/
 	u64 rsvd1[2];
@@ -269,6 +271,16 @@ struct rngtst {
 /* RNG4 TRNG test registers */
 struct rng4tst {
 #define RTMCTL_PRGM	0x00010000	/* 1 -> program mode, 0 -> run mode */
+#define RTMCTL_SAMP_MODE_VON_NEUMANN_ES_SC	0 /* use von Neumann data in
+						     both entropy shifter and
+						     statistical checker */
+#define RTMCTL_SAMP_MODE_RAW_ES_SC		1 /* use raw data in both
+						     entropy shifter and
+						     statistical checker */
+#define RTMCTL_SAMP_MODE_VON_NEUMANN_ES_RAW_SC	2 /* use von Neumann data in
+						     entropy shifter, raw data
+						     in statistical checker */
+#define RTMCTL_SAMP_MODE_INVALID		3 /* invalid combination */
 	u32 rtmctl;		/* misc. control register */
 	u32 rtscmisc;		/* statistical check misc. register */
 	u32 rtpkrrng;		/* poker range register */
@@ -278,7 +290,7 @@ struct rng4tst {
 	};
 #define RTSDCTL_ENT_DLY_SHIFT 16
 #define RTSDCTL_ENT_DLY_MASK (0xffff << RTSDCTL_ENT_DLY_SHIFT)
-#define RTSDCTL_ENT_DLY_MIN 1200
+#define RTSDCTL_ENT_DLY_MIN 3200
 #define RTSDCTL_ENT_DLY_MAX 12800
 	u32 rtsdctl;		/* seed control register */
 	union {
@@ -286,6 +298,7 @@ struct rng4tst {
 		u32 rttotsam;	/* PRGM=0: total samples register */
 	};
 	u32 rtfrqmin;		/* frequency count min. limit register */
+#define RTFRQMAX_DISABLE	(1 << 20)
 	union {
 		u32 rtfrqmax;	/* PRGM=1: freq. count max. limit register */
 		u32 rtfrqcnt;	/* PRGM=0: freq. count register */
@@ -758,34 +771,10 @@ struct caam_deco {
 #define DECO_JQCR_WHL		0x20000000
 #define DECO_JQCR_FOUR		0x10000000
 
-/*
- * Current top-level view of memory map is:
- *
- * 0x0000 - 0x0fff - CAAM Top-Level Control
- * 0x1000 - 0x1fff - Job Ring 0
- * 0x2000 - 0x2fff - Job Ring 1
- * 0x3000 - 0x3fff - Job Ring 2
- * 0x4000 - 0x4fff - Job Ring 3
- * 0x5000 - 0x5fff - (unused)
- * 0x6000 - 0x6fff - Assurance Controller
- * 0x7000 - 0x7fff - Queue Interface
- * 0x8000 - 0x8fff - DECO-CCB 0
- * 0x9000 - 0x9fff - DECO-CCB 1
- * 0xa000 - 0xafff - DECO-CCB 2
- * 0xb000 - 0xbfff - DECO-CCB 3
- * 0xc000 - 0xcfff - DECO-CCB 4
- *
- * caam_full describes the full register view of CAAM if useful,
- * although many configurations may choose to implement parts of
- * the register map separately, in differing privilege regions
- */
-struct caam_full {
-	struct caam_ctrl __iomem ctrl;
-	struct caam_job_ring jr[4];
-	u64 rsvd[512];
-	struct caam_assurance assure;
-	struct caam_queue_if qi;
-	struct caam_deco deco;
-};
-
+#define JR_BLOCK_NUMBER		1
+#define ASSURE_BLOCK_NUMBER	6
+#define QI_BLOCK_NUMBER		7
+#define DECO_BLOCK_NUMBER	8
+#define PG_SIZE_4K		0x1000
+#define PG_SIZE_64K		0x10000
 #endif /* REGS_H */
diff --git a/drivers/crypto/caam/sg_sw_sec4.h b/drivers/crypto/caam/sg_sw_sec4.h
index b12ff85f4241..ce28a563effc 100644
--- a/drivers/crypto/caam/sg_sw_sec4.h
+++ b/drivers/crypto/caam/sg_sw_sec4.h
@@ -116,57 +116,3 @@ static int dma_unmap_sg_chained(struct device *dev, struct scatterlist *sg,
 	}
 	return nents;
 }
-
-/* Map SG page in kernel virtual address space and copy */
-static inline void sg_map_copy(u8 *dest, struct scatterlist *sg,
-			       int len, int offset)
-{
-	u8 *mapped_addr;
-
-	/*
-	 * Page here can be user-space pinned using get_user_pages
-	 * Same must be kmapped before use and kunmapped subsequently
-	 */
-	mapped_addr = kmap_atomic(sg_page(sg));
-	memcpy(dest, mapped_addr + offset, len);
-	kunmap_atomic(mapped_addr);
-}
-
-/* Copy from len bytes of sg to dest, starting from beginning */
-static inline void sg_copy(u8 *dest, struct scatterlist *sg, unsigned int len)
-{
-	struct scatterlist *current_sg = sg;
-	int cpy_index = 0, next_cpy_index = current_sg->length;
-
-	while (next_cpy_index < len) {
-		sg_map_copy(dest + cpy_index, current_sg, current_sg->length,
-			    current_sg->offset);
-		current_sg = scatterwalk_sg_next(current_sg);
-		cpy_index = next_cpy_index;
-		next_cpy_index += current_sg->length;
-	}
-	if (cpy_index < len)
-		sg_map_copy(dest + cpy_index, current_sg, len-cpy_index,
-			    current_sg->offset);
-}
-
-/* Copy sg data, from to_skip to end, to dest */
-static inline void sg_copy_part(u8 *dest, struct scatterlist *sg,
-				      int to_skip, unsigned int end)
-{
-	struct scatterlist *current_sg = sg;
-	int sg_index, cpy_index, offset;
-
-	sg_index = current_sg->length;
-	while (sg_index <= to_skip) {
-		current_sg = scatterwalk_sg_next(current_sg);
-		sg_index += current_sg->length;
-	}
-	cpy_index = sg_index - to_skip;
-	offset = current_sg->offset + current_sg->length - cpy_index;
-	sg_map_copy(dest, current_sg, cpy_index, offset);
-	if (end - sg_index) {
-		current_sg = scatterwalk_sg_next(current_sg);
-		sg_copy(dest + cpy_index, current_sg, end - sg_index);
-	}
-}
diff --git a/drivers/crypto/ccp/ccp-crypto-sha.c b/drivers/crypto/ccp/ccp-crypto-sha.c
index 873f23425245..96531571f7cf 100644
--- a/drivers/crypto/ccp/ccp-crypto-sha.c
+++ b/drivers/crypto/ccp/ccp-crypto-sha.c
@@ -198,10 +198,9 @@ static int ccp_sha_setkey(struct crypto_ahash *tfm, const u8 *key,
 {
 	struct ccp_ctx *ctx = crypto_tfm_ctx(crypto_ahash_tfm(tfm));
 	struct crypto_shash *shash = ctx->u.sha.hmac_tfm;
-	struct {
-		struct shash_desc sdesc;
-		char ctx[crypto_shash_descsize(shash)];
-	} desc;
+
+	SHASH_DESC_ON_STACK(sdesc, shash);
+
 	unsigned int block_size = crypto_shash_blocksize(shash);
 	unsigned int digest_size = crypto_shash_digestsize(shash);
 	int i, ret;
@@ -216,11 +215,11 @@ static int ccp_sha_setkey(struct crypto_ahash *tfm, const u8 *key,
 
 	if (key_len > block_size) {
 		/* Must hash the input key */
-		desc.sdesc.tfm = shash;
-		desc.sdesc.flags = crypto_ahash_get_flags(tfm) &
+		sdesc->tfm = shash;
+		sdesc->flags = crypto_ahash_get_flags(tfm) &
 			CRYPTO_TFM_REQ_MAY_SLEEP;
 
-		ret = crypto_shash_digest(&desc.sdesc, key, key_len,
+		ret = crypto_shash_digest(sdesc, key, key_len,
 					  ctx->u.sha.key);
 		if (ret) {
 			crypto_ahash_set_flags(tfm, CRYPTO_TFM_RES_BAD_KEY_LEN);
diff --git a/drivers/crypto/ccp/ccp-platform.c b/drivers/crypto/ccp/ccp-platform.c
index b0a2806908f1..8c50bad25f7e 100644
--- a/drivers/crypto/ccp/ccp-platform.c
+++ b/drivers/crypto/ccp/ccp-platform.c
@@ -208,7 +208,6 @@ static const struct of_device_id ccp_platform_ids[] = {
 static struct platform_driver ccp_platform_driver = {
 	.driver = {
 		.name = "AMD Cryptographic Coprocessor",
-		.owner = THIS_MODULE,
 		.of_match_table = ccp_platform_ids,
 	},
 	.probe = ccp_platform_probe,
diff --git a/drivers/crypto/mv_cesa.c b/drivers/crypto/mv_cesa.c
index 29d0ee504907..f91f15ddee92 100644
--- a/drivers/crypto/mv_cesa.c
+++ b/drivers/crypto/mv_cesa.c
@@ -402,26 +402,23 @@ static int mv_hash_final_fallback(struct ahash_request *req)
 {
 	const struct mv_tfm_hash_ctx *tfm_ctx = crypto_tfm_ctx(req->base.tfm);
 	struct mv_req_hash_ctx *req_ctx = ahash_request_ctx(req);
-	struct {
-		struct shash_desc shash;
-		char ctx[crypto_shash_descsize(tfm_ctx->fallback)];
-	} desc;
+	SHASH_DESC_ON_STACK(shash, tfm_ctx->fallback);
 	int rc;
 
-	desc.shash.tfm = tfm_ctx->fallback;
-	desc.shash.flags = CRYPTO_TFM_REQ_MAY_SLEEP;
+	shash->tfm = tfm_ctx->fallback;
+	shash->flags = CRYPTO_TFM_REQ_MAY_SLEEP;
 	if (unlikely(req_ctx->first_hash)) {
-		crypto_shash_init(&desc.shash);
-		crypto_shash_update(&desc.shash, req_ctx->buffer,
+		crypto_shash_init(shash);
+		crypto_shash_update(shash, req_ctx->buffer,
 				    req_ctx->extra_bytes);
 	} else {
 		/* only SHA1 for now....
 		 */
-		rc = mv_hash_import_sha1_ctx(req_ctx, &desc.shash);
+		rc = mv_hash_import_sha1_ctx(req_ctx, shash);
 		if (rc)
 			goto out;
 	}
-	rc = crypto_shash_final(&desc.shash, req->result);
+	rc = crypto_shash_final(shash, req->result);
 out:
 	return rc;
 }
@@ -794,23 +791,21 @@ static int mv_hash_setkey(struct crypto_ahash *tfm, const u8 * key,
 	ss = crypto_shash_statesize(ctx->base_hash);
 
 	{
-		struct {
-			struct shash_desc shash;
-			char ctx[crypto_shash_descsize(ctx->base_hash)];
-		} desc;
+		SHASH_DESC_ON_STACK(shash, ctx->base_hash);
+
 		unsigned int i;
 		char ipad[ss];
 		char opad[ss];
 
-		desc.shash.tfm = ctx->base_hash;
-		desc.shash.flags = crypto_shash_get_flags(ctx->base_hash) &
+		shash->tfm = ctx->base_hash;
+		shash->flags = crypto_shash_get_flags(ctx->base_hash) &
 		    CRYPTO_TFM_REQ_MAY_SLEEP;
 
 		if (keylen > bs) {
 			int err;
 
 			err =
-			    crypto_shash_digest(&desc.shash, key, keylen, ipad);
+			    crypto_shash_digest(shash, key, keylen, ipad);
 			if (err)
 				return err;
 
@@ -826,12 +821,12 @@ static int mv_hash_setkey(struct crypto_ahash *tfm, const u8 * key,
 			opad[i] ^= 0x5c;
 		}
 
-		rc = crypto_shash_init(&desc.shash) ? :
-		    crypto_shash_update(&desc.shash, ipad, bs) ? :
-		    crypto_shash_export(&desc.shash, ipad) ? :
-		    crypto_shash_init(&desc.shash) ? :
-		    crypto_shash_update(&desc.shash, opad, bs) ? :
-		    crypto_shash_export(&desc.shash, opad);
+		rc = crypto_shash_init(shash) ? :
+		    crypto_shash_update(shash, ipad, bs) ? :
+		    crypto_shash_export(shash, ipad) ? :
+		    crypto_shash_init(shash) ? :
+		    crypto_shash_update(shash, opad, bs) ? :
+		    crypto_shash_export(shash, opad);
 
 		if (rc == 0)
 			mv_hash_init_ivs(ctx, ipad, opad);
@@ -1185,7 +1180,6 @@ static struct platform_driver marvell_crypto = {
 	.probe		= mv_probe,
 	.remove		= mv_remove,
 	.driver		= {
-		.owner	= THIS_MODULE,
 		.name	= "mv_crypto",
 		.of_match_table = mv_cesa_of_match_table,
 	},
diff --git a/drivers/crypto/mv_cesa.h b/drivers/crypto/mv_cesa.h
index 08fcb1116d90..9249d3ed184b 100644
--- a/drivers/crypto/mv_cesa.h
+++ b/drivers/crypto/mv_cesa.h
@@ -1,4 +1,5 @@
 #ifndef __MV_CRYPTO_H__
+#define __MV_CRYPTO_H__
 
 #define DIGEST_INITIAL_VAL_A	0xdd00
 #define DIGEST_INITIAL_VAL_B	0xdd04
diff --git a/drivers/crypto/mxs-dcp.c b/drivers/crypto/mxs-dcp.c
index b5f7e6db24d4..829d6394fb33 100644
--- a/drivers/crypto/mxs-dcp.c
+++ b/drivers/crypto/mxs-dcp.c
@@ -1090,7 +1090,6 @@ static struct platform_driver mxs_dcp_driver = {
 	.remove	= mxs_dcp_remove,
 	.driver	= {
 		.name		= "mxs-dcp",
-		.owner		= THIS_MODULE,
 		.of_match_table	= mxs_dcp_dt_ids,
 	},
 };
diff --git a/drivers/crypto/n2_core.c b/drivers/crypto/n2_core.c
index 7263c10a56ee..afd136b45f49 100644
--- a/drivers/crypto/n2_core.c
+++ b/drivers/crypto/n2_core.c
@@ -445,10 +445,7 @@ static int n2_hmac_async_setkey(struct crypto_ahash *tfm, const u8 *key,
 	struct n2_hmac_ctx *ctx = crypto_ahash_ctx(tfm);
 	struct crypto_shash *child_shash = ctx->child_shash;
 	struct crypto_ahash *fallback_tfm;
-	struct {
-		struct shash_desc shash;
-		char ctx[crypto_shash_descsize(child_shash)];
-	} desc;
+	SHASH_DESC_ON_STACK(shash, child_shash);
 	int err, bs, ds;
 
 	fallback_tfm = ctx->base.fallback_tfm;
@@ -456,15 +453,15 @@ static int n2_hmac_async_setkey(struct crypto_ahash *tfm, const u8 *key,
 	if (err)
 		return err;
 
-	desc.shash.tfm = child_shash;
-	desc.shash.flags = crypto_ahash_get_flags(tfm) &
+	shash->tfm = child_shash;
+	shash->flags = crypto_ahash_get_flags(tfm) &
 		CRYPTO_TFM_REQ_MAY_SLEEP;
 
 	bs = crypto_shash_blocksize(child_shash);
 	ds = crypto_shash_digestsize(child_shash);
 	BUG_ON(ds > N2_HASH_KEY_MAX);
 	if (keylen > bs) {
-		err = crypto_shash_digest(&desc.shash, key, keylen,
+		err = crypto_shash_digest(shash, key, keylen,
 					  ctx->hash_key);
 		if (err)
 			return err;
@@ -2213,7 +2210,6 @@ MODULE_DEVICE_TABLE(of, n2_crypto_match);
 static struct platform_driver n2_crypto_driver = {
 	.driver = {
 		.name		=	"n2cp",
-		.owner		=	THIS_MODULE,
 		.of_match_table	=	n2_crypto_match,
 	},
 	.probe		=	n2_crypto_probe,
@@ -2241,7 +2237,6 @@ MODULE_DEVICE_TABLE(of, n2_mau_match);
 static struct platform_driver n2_mau_driver = {
 	.driver = {
 		.name		=	"ncp",
-		.owner		=	THIS_MODULE,
 		.of_match_table	=	n2_mau_match,
 	},
 	.probe		=	n2_mau_probe,
diff --git a/drivers/crypto/nx/nx-842.c b/drivers/crypto/nx/nx-842.c
index 061407d59520..887196e9b50c 100644
--- a/drivers/crypto/nx/nx-842.c
+++ b/drivers/crypto/nx/nx-842.c
@@ -1009,9 +1009,9 @@ error_out:
  *		notifier_to_errno() to decode this value
  */
 static int nx842_OF_notifier(struct notifier_block *np, unsigned long action,
-			     void *update)
+			     void *data)
 {
-	struct of_prop_reconfig *upd = update;
+	struct of_reconfig_data *upd = data;
 	struct nx842_devdata *local_devdata;
 	struct device_node *node = NULL;
 
diff --git a/drivers/crypto/nx/nx-aes-cbc.c b/drivers/crypto/nx/nx-aes-cbc.c
index cc00b52306ba..a066cc3450ae 100644
--- a/drivers/crypto/nx/nx-aes-cbc.c
+++ b/drivers/crypto/nx/nx-aes-cbc.c
@@ -72,27 +72,19 @@ static int cbc_aes_nx_crypt(struct blkcipher_desc *desc,
 	struct nx_csbcpb *csbcpb = nx_ctx->csbcpb;
 	unsigned long irq_flags;
 	unsigned int processed = 0, to_process;
-	u32 max_sg_len;
 	int rc;
 
 	spin_lock_irqsave(&nx_ctx->lock, irq_flags);
 
-	max_sg_len = min_t(u32, nx_driver.of.max_sg_len/sizeof(struct nx_sg),
-			   nx_ctx->ap->sglen);
-
 	if (enc)
 		NX_CPB_FDM(csbcpb) |= NX_FDM_ENDE_ENCRYPT;
 	else
 		NX_CPB_FDM(csbcpb) &= ~NX_FDM_ENDE_ENCRYPT;
 
 	do {
-		to_process = min_t(u64, nbytes - processed,
-				   nx_ctx->ap->databytelen);
-		to_process = min_t(u64, to_process,
-				   NX_PAGE_SIZE * (max_sg_len - 1));
-		to_process = to_process & ~(AES_BLOCK_SIZE - 1);
+		to_process = nbytes - processed;
 
-		rc = nx_build_sg_lists(nx_ctx, desc, dst, src, to_process,
+		rc = nx_build_sg_lists(nx_ctx, desc, dst, src, &to_process,
 				       processed, csbcpb->cpb.aes_cbc.iv);
 		if (rc)
 			goto out;
diff --git a/drivers/crypto/nx/nx-aes-ccm.c b/drivers/crypto/nx/nx-aes-ccm.c
index 5ecd4c2414aa..67f80813a06f 100644
--- a/drivers/crypto/nx/nx-aes-ccm.c
+++ b/drivers/crypto/nx/nx-aes-ccm.c
@@ -181,6 +181,7 @@ static int generate_pat(u8                   *iv,
 	unsigned int iauth_len = 0;
 	u8 tmp[16], *b1 = NULL, *b0 = NULL, *result = NULL;
 	int rc;
+	unsigned int max_sg_len;
 
 	/* zero the ctr value */
 	memset(iv + 15 - iv[0], 0, iv[0] + 1);
@@ -248,10 +249,19 @@ static int generate_pat(u8                   *iv,
 	if (!req->assoclen) {
 		return rc;
 	} else if (req->assoclen <= 14) {
-		nx_insg = nx_build_sg_list(nx_insg, b1, 16, nx_ctx->ap->sglen);
-		nx_outsg = nx_build_sg_list(nx_outsg, tmp, 16,
+		unsigned int len = 16;
+
+		nx_insg = nx_build_sg_list(nx_insg, b1, &len, nx_ctx->ap->sglen);
+
+		if (len != 16)
+			return -EINVAL;
+
+		nx_outsg = nx_build_sg_list(nx_outsg, tmp, &len,
 					    nx_ctx->ap->sglen);
 
+		if (len != 16)
+			return -EINVAL;
+
 		/* inlen should be negative, indicating to phyp that its a
 		 * pointer to an sg list */
 		nx_ctx->op.inlen = (nx_ctx->in_sg - nx_insg) *
@@ -273,21 +283,24 @@ static int generate_pat(u8                   *iv,
 		atomic64_add(req->assoclen, &(nx_ctx->stats->aes_bytes));
 
 	} else {
-		u32 max_sg_len;
 		unsigned int processed = 0, to_process;
 
-		/* page_limit: number of sg entries that fit on one page */
-		max_sg_len = min_t(u32,
-				   nx_driver.of.max_sg_len/sizeof(struct nx_sg),
-				   nx_ctx->ap->sglen);
-
 		processed += iauth_len;
 
+		/* page_limit: number of sg entries that fit on one page */
+		max_sg_len = min_t(u64, nx_ctx->ap->sglen,
+				nx_driver.of.max_sg_len/sizeof(struct nx_sg));
+		max_sg_len = min_t(u64, max_sg_len,
+				nx_ctx->ap->databytelen/NX_PAGE_SIZE);
+
 		do {
 			to_process = min_t(u32, req->assoclen - processed,
 					   nx_ctx->ap->databytelen);
-			to_process = min_t(u64, to_process,
-					   NX_PAGE_SIZE * (max_sg_len - 1));
+
+			nx_insg = nx_walk_and_build(nx_ctx->in_sg,
+						    nx_ctx->ap->sglen,
+						    req->assoc, processed,
+						    &to_process);
 
 			if ((to_process + processed) < req->assoclen) {
 				NX_CPB_FDM(nx_ctx->csbcpb_aead) |=
@@ -297,10 +310,6 @@ static int generate_pat(u8                   *iv,
 					~NX_FDM_INTERMEDIATE;
 			}
 
-			nx_insg = nx_walk_and_build(nx_ctx->in_sg,
-						    nx_ctx->ap->sglen,
-						    req->assoc, processed,
-						    to_process);
 
 			nx_ctx->op_aead.inlen = (nx_ctx->in_sg - nx_insg) *
 						sizeof(struct nx_sg);
@@ -343,7 +352,6 @@ static int ccm_nx_decrypt(struct aead_request   *req,
 	struct nx_ccm_priv *priv = &nx_ctx->priv.ccm;
 	unsigned long irq_flags;
 	unsigned int processed = 0, to_process;
-	u32 max_sg_len;
 	int rc = -1;
 
 	spin_lock_irqsave(&nx_ctx->lock, irq_flags);
@@ -360,19 +368,12 @@ static int ccm_nx_decrypt(struct aead_request   *req,
 	if (rc)
 		goto out;
 
-	/* page_limit: number of sg entries that fit on one page */
-	max_sg_len = min_t(u32, nx_driver.of.max_sg_len/sizeof(struct nx_sg),
-			   nx_ctx->ap->sglen);
-
 	do {
 
 		/* to_process: the AES_BLOCK_SIZE data chunk to process in this
 		 * update. This value is bound by sg list limits.
 		 */
-		to_process = min_t(u64, nbytes - processed,
-				   nx_ctx->ap->databytelen);
-		to_process = min_t(u64, to_process,
-				   NX_PAGE_SIZE * (max_sg_len - 1));
+		to_process = nbytes - processed;
 
 		if ((to_process + processed) < nbytes)
 			NX_CPB_FDM(csbcpb) |= NX_FDM_INTERMEDIATE;
@@ -382,7 +383,7 @@ static int ccm_nx_decrypt(struct aead_request   *req,
 		NX_CPB_FDM(nx_ctx->csbcpb) &= ~NX_FDM_ENDE_ENCRYPT;
 
 		rc = nx_build_sg_lists(nx_ctx, desc, req->dst, req->src,
-					to_process, processed,
+					&to_process, processed,
 					csbcpb->cpb.aes_ccm.iv_or_ctr);
 		if (rc)
 			goto out;
@@ -427,7 +428,6 @@ static int ccm_nx_encrypt(struct aead_request   *req,
 	unsigned int authsize = crypto_aead_authsize(crypto_aead_reqtfm(req));
 	unsigned long irq_flags;
 	unsigned int processed = 0, to_process;
-	u32 max_sg_len;
 	int rc = -1;
 
 	spin_lock_irqsave(&nx_ctx->lock, irq_flags);
@@ -437,18 +437,11 @@ static int ccm_nx_encrypt(struct aead_request   *req,
 	if (rc)
 		goto out;
 
-	/* page_limit: number of sg entries that fit on one page */
-	max_sg_len = min_t(u32, nx_driver.of.max_sg_len/sizeof(struct nx_sg),
-			   nx_ctx->ap->sglen);
-
 	do {
 		/* to process: the AES_BLOCK_SIZE data chunk to process in this
 		 * update. This value is bound by sg list limits.
 		 */
-		to_process = min_t(u64, nbytes - processed,
-				   nx_ctx->ap->databytelen);
-		to_process = min_t(u64, to_process,
-				   NX_PAGE_SIZE * (max_sg_len - 1));
+		to_process = nbytes - processed;
 
 		if ((to_process + processed) < nbytes)
 			NX_CPB_FDM(csbcpb) |= NX_FDM_INTERMEDIATE;
@@ -458,7 +451,7 @@ static int ccm_nx_encrypt(struct aead_request   *req,
 		NX_CPB_FDM(csbcpb) |= NX_FDM_ENDE_ENCRYPT;
 
 		rc = nx_build_sg_lists(nx_ctx, desc, req->dst, req->src,
-					to_process, processed,
+					&to_process, processed,
 				       csbcpb->cpb.aes_ccm.iv_or_ctr);
 		if (rc)
 			goto out;
diff --git a/drivers/crypto/nx/nx-aes-ctr.c b/drivers/crypto/nx/nx-aes-ctr.c
index a37d009dc75c..2617cd4d54dd 100644
--- a/drivers/crypto/nx/nx-aes-ctr.c
+++ b/drivers/crypto/nx/nx-aes-ctr.c
@@ -90,22 +90,14 @@ static int ctr_aes_nx_crypt(struct blkcipher_desc *desc,
 	struct nx_csbcpb *csbcpb = nx_ctx->csbcpb;
 	unsigned long irq_flags;
 	unsigned int processed = 0, to_process;
-	u32 max_sg_len;
 	int rc;
 
 	spin_lock_irqsave(&nx_ctx->lock, irq_flags);
 
-	max_sg_len = min_t(u32, nx_driver.of.max_sg_len/sizeof(struct nx_sg),
-			   nx_ctx->ap->sglen);
-
 	do {
-		to_process = min_t(u64, nbytes - processed,
-				   nx_ctx->ap->databytelen);
-		to_process = min_t(u64, to_process,
-				   NX_PAGE_SIZE * (max_sg_len - 1));
-		to_process = to_process & ~(AES_BLOCK_SIZE - 1);
+		to_process = nbytes - processed;
 
-		rc = nx_build_sg_lists(nx_ctx, desc, dst, src, to_process,
+		rc = nx_build_sg_lists(nx_ctx, desc, dst, src, &to_process,
 				       processed, csbcpb->cpb.aes_ctr.iv);
 		if (rc)
 			goto out;
@@ -143,6 +135,7 @@ static int ctr3686_aes_nx_crypt(struct blkcipher_desc *desc,
 
 	memcpy(iv + CTR_RFC3686_NONCE_SIZE,
 	       desc->info, CTR_RFC3686_IV_SIZE);
+	iv[12] = iv[13] = iv[14] = 0;
 	iv[15] = 1;
 
 	desc->info = nx_ctx->priv.ctr.iv;
diff --git a/drivers/crypto/nx/nx-aes-ecb.c b/drivers/crypto/nx/nx-aes-ecb.c
index 85a8d23cf29d..cfdde8b8bc76 100644
--- a/drivers/crypto/nx/nx-aes-ecb.c
+++ b/drivers/crypto/nx/nx-aes-ecb.c
@@ -72,27 +72,19 @@ static int ecb_aes_nx_crypt(struct blkcipher_desc *desc,
 	struct nx_csbcpb *csbcpb = nx_ctx->csbcpb;
 	unsigned long irq_flags;
 	unsigned int processed = 0, to_process;
-	u32 max_sg_len;
 	int rc;
 
 	spin_lock_irqsave(&nx_ctx->lock, irq_flags);
 
-	max_sg_len = min_t(u32, nx_driver.of.max_sg_len/sizeof(struct nx_sg),
-			   nx_ctx->ap->sglen);
-
 	if (enc)
 		NX_CPB_FDM(csbcpb) |= NX_FDM_ENDE_ENCRYPT;
 	else
 		NX_CPB_FDM(csbcpb) &= ~NX_FDM_ENDE_ENCRYPT;
 
 	do {
-		to_process = min_t(u64, nbytes - processed,
-				   nx_ctx->ap->databytelen);
-		to_process = min_t(u64, to_process,
-				   NX_PAGE_SIZE * (max_sg_len - 1));
-		to_process = to_process & ~(AES_BLOCK_SIZE - 1);
+		to_process = nbytes - processed;
 
-		rc = nx_build_sg_lists(nx_ctx, desc, dst, src, to_process,
+		rc = nx_build_sg_lists(nx_ctx, desc, dst, src, &to_process,
 				processed, NULL);
 		if (rc)
 			goto out;
diff --git a/drivers/crypto/nx/nx-aes-gcm.c b/drivers/crypto/nx/nx-aes-gcm.c
index 025d9a8d5b19..88c562434bc0 100644
--- a/drivers/crypto/nx/nx-aes-gcm.c
+++ b/drivers/crypto/nx/nx-aes-gcm.c
@@ -131,7 +131,7 @@ static int nx_gca(struct nx_crypto_ctx  *nx_ctx,
 	struct nx_sg *nx_sg = nx_ctx->in_sg;
 	unsigned int nbytes = req->assoclen;
 	unsigned int processed = 0, to_process;
-	u32 max_sg_len;
+	unsigned int max_sg_len;
 
 	if (nbytes <= AES_BLOCK_SIZE) {
 		scatterwalk_start(&walk, req->assoc);
@@ -143,8 +143,10 @@ static int nx_gca(struct nx_crypto_ctx  *nx_ctx,
 	NX_CPB_FDM(csbcpb_aead) &= ~NX_FDM_CONTINUATION;
 
 	/* page_limit: number of sg entries that fit on one page */
-	max_sg_len = min_t(u32, nx_driver.of.max_sg_len/sizeof(struct nx_sg),
+	max_sg_len = min_t(u64, nx_driver.of.max_sg_len/sizeof(struct nx_sg),
 			   nx_ctx->ap->sglen);
+	max_sg_len = min_t(u64, max_sg_len,
+			   nx_ctx->ap->databytelen/NX_PAGE_SIZE);
 
 	do {
 		/*
@@ -156,13 +158,14 @@ static int nx_gca(struct nx_crypto_ctx  *nx_ctx,
 		to_process = min_t(u64, to_process,
 				   NX_PAGE_SIZE * (max_sg_len - 1));
 
+		nx_sg = nx_walk_and_build(nx_ctx->in_sg, max_sg_len,
+					  req->assoc, processed, &to_process);
+
 		if ((to_process + processed) < nbytes)
 			NX_CPB_FDM(csbcpb_aead) |= NX_FDM_INTERMEDIATE;
 		else
 			NX_CPB_FDM(csbcpb_aead) &= ~NX_FDM_INTERMEDIATE;
 
-		nx_sg = nx_walk_and_build(nx_ctx->in_sg, nx_ctx->ap->sglen,
-					  req->assoc, processed, to_process);
 		nx_ctx->op_aead.inlen = (nx_ctx->in_sg - nx_sg)
 					* sizeof(struct nx_sg);
 
@@ -195,7 +198,7 @@ static int gmac(struct aead_request *req, struct blkcipher_desc *desc)
 	struct nx_sg *nx_sg;
 	unsigned int nbytes = req->assoclen;
 	unsigned int processed = 0, to_process;
-	u32 max_sg_len;
+	unsigned int max_sg_len;
 
 	/* Set GMAC mode */
 	csbcpb->cpb.hdr.mode = NX_MODE_AES_GMAC;
@@ -203,8 +206,10 @@ static int gmac(struct aead_request *req, struct blkcipher_desc *desc)
 	NX_CPB_FDM(csbcpb) &= ~NX_FDM_CONTINUATION;
 
 	/* page_limit: number of sg entries that fit on one page */
-	max_sg_len = min_t(u32, nx_driver.of.max_sg_len/sizeof(struct nx_sg),
+	max_sg_len = min_t(u64, nx_driver.of.max_sg_len/sizeof(struct nx_sg),
 			   nx_ctx->ap->sglen);
+	max_sg_len = min_t(u64, max_sg_len,
+			   nx_ctx->ap->databytelen/NX_PAGE_SIZE);
 
 	/* Copy IV */
 	memcpy(csbcpb->cpb.aes_gcm.iv_or_cnt, desc->info, AES_BLOCK_SIZE);
@@ -219,13 +224,14 @@ static int gmac(struct aead_request *req, struct blkcipher_desc *desc)
 		to_process = min_t(u64, to_process,
 				   NX_PAGE_SIZE * (max_sg_len - 1));
 
+		nx_sg = nx_walk_and_build(nx_ctx->in_sg, max_sg_len,
+					  req->assoc, processed, &to_process);
+
 		if ((to_process + processed) < nbytes)
 			NX_CPB_FDM(csbcpb) |= NX_FDM_INTERMEDIATE;
 		else
 			NX_CPB_FDM(csbcpb) &= ~NX_FDM_INTERMEDIATE;
 
-		nx_sg = nx_walk_and_build(nx_ctx->in_sg, nx_ctx->ap->sglen,
-					  req->assoc, processed, to_process);
 		nx_ctx->op.inlen = (nx_ctx->in_sg - nx_sg)
 					* sizeof(struct nx_sg);
 
@@ -264,6 +270,7 @@ static int gcm_empty(struct aead_request *req, struct blkcipher_desc *desc,
 	struct nx_csbcpb *csbcpb = nx_ctx->csbcpb;
 	char out[AES_BLOCK_SIZE];
 	struct nx_sg *in_sg, *out_sg;
+	int len;
 
 	/* For scenarios where the input message is zero length, AES CTR mode
 	 * may be used. Set the source data to be a single block (16B) of all
@@ -279,11 +286,22 @@ static int gcm_empty(struct aead_request *req, struct blkcipher_desc *desc,
 	else
 		NX_CPB_FDM(csbcpb) &= ~NX_FDM_ENDE_ENCRYPT;
 
+	len = AES_BLOCK_SIZE;
+
 	/* Encrypt the counter/IV */
 	in_sg = nx_build_sg_list(nx_ctx->in_sg, (u8 *) desc->info,
-				 AES_BLOCK_SIZE, nx_ctx->ap->sglen);
-	out_sg = nx_build_sg_list(nx_ctx->out_sg, (u8 *) out, sizeof(out),
+				 &len, nx_ctx->ap->sglen);
+
+	if (len != AES_BLOCK_SIZE)
+		return -EINVAL;
+
+	len = sizeof(out);
+	out_sg = nx_build_sg_list(nx_ctx->out_sg, (u8 *) out, &len,
 				  nx_ctx->ap->sglen);
+
+	if (len != sizeof(out))
+		return -EINVAL;
+
 	nx_ctx->op.inlen = (nx_ctx->in_sg - in_sg) * sizeof(struct nx_sg);
 	nx_ctx->op.outlen = (nx_ctx->out_sg - out_sg) * sizeof(struct nx_sg);
 
@@ -317,7 +335,6 @@ static int gcm_aes_nx_crypt(struct aead_request *req, int enc)
 	unsigned int nbytes = req->cryptlen;
 	unsigned int processed = 0, to_process;
 	unsigned long irq_flags;
-	u32 max_sg_len;
 	int rc = -EINVAL;
 
 	spin_lock_irqsave(&nx_ctx->lock, irq_flags);
@@ -354,33 +371,24 @@ static int gcm_aes_nx_crypt(struct aead_request *req, int enc)
 		nbytes -= crypto_aead_authsize(crypto_aead_reqtfm(req));
 	}
 
-	/* page_limit: number of sg entries that fit on one page */
-	max_sg_len = min_t(u32, nx_driver.of.max_sg_len/sizeof(struct nx_sg),
-			   nx_ctx->ap->sglen);
-
 	do {
-		/*
-		 * to_process: the data chunk to process in this update.
-		 * This value is bound by sg list limits.
-		 */
-		to_process = min_t(u64, nbytes - processed,
-				   nx_ctx->ap->databytelen);
-		to_process = min_t(u64, to_process,
-				   NX_PAGE_SIZE * (max_sg_len - 1));
-
-		if ((to_process + processed) < nbytes)
-			NX_CPB_FDM(csbcpb) |= NX_FDM_INTERMEDIATE;
-		else
-			NX_CPB_FDM(csbcpb) &= ~NX_FDM_INTERMEDIATE;
+		to_process = nbytes - processed;
 
 		csbcpb->cpb.aes_gcm.bit_length_data = nbytes * 8;
 		desc.tfm = (struct crypto_blkcipher *) req->base.tfm;
 		rc = nx_build_sg_lists(nx_ctx, &desc, req->dst,
-				       req->src, to_process, processed,
+				       req->src, &to_process, processed,
 				       csbcpb->cpb.aes_gcm.iv_or_cnt);
+
 		if (rc)
 			goto out;
 
+		if ((to_process + processed) < nbytes)
+			NX_CPB_FDM(csbcpb) |= NX_FDM_INTERMEDIATE;
+		else
+			NX_CPB_FDM(csbcpb) &= ~NX_FDM_INTERMEDIATE;
+
+
 		rc = nx_hcall_sync(nx_ctx, &nx_ctx->op,
 				   req->base.flags & CRYPTO_TFM_REQ_MAY_SLEEP);
 		if (rc)
diff --git a/drivers/crypto/nx/nx-aes-xcbc.c b/drivers/crypto/nx/nx-aes-xcbc.c
index 03c4bf57d066..8c2faffab4a3 100644
--- a/drivers/crypto/nx/nx-aes-xcbc.c
+++ b/drivers/crypto/nx/nx-aes-xcbc.c
@@ -75,6 +75,7 @@ static int nx_xcbc_empty(struct shash_desc *desc, u8 *out)
 	u8 keys[2][AES_BLOCK_SIZE];
 	u8 key[32];
 	int rc = 0;
+	int len;
 
 	/* Change to ECB mode */
 	csbcpb->cpb.hdr.mode = NX_MODE_AES_ECB;
@@ -86,11 +87,20 @@ static int nx_xcbc_empty(struct shash_desc *desc, u8 *out)
 	memset(keys[0], 0x01, sizeof(keys[0]));
 	memset(keys[1], 0x03, sizeof(keys[1]));
 
+	len = sizeof(keys);
 	/* Generate K1 and K3 encrypting the patterns */
-	in_sg = nx_build_sg_list(nx_ctx->in_sg, (u8 *) keys, sizeof(keys),
+	in_sg = nx_build_sg_list(nx_ctx->in_sg, (u8 *) keys, &len,
 				 nx_ctx->ap->sglen);
-	out_sg = nx_build_sg_list(nx_ctx->out_sg, (u8 *) keys, sizeof(keys),
+
+	if (len != sizeof(keys))
+		return -EINVAL;
+
+	out_sg = nx_build_sg_list(nx_ctx->out_sg, (u8 *) keys, &len,
 				  nx_ctx->ap->sglen);
+
+	if (len != sizeof(keys))
+		return -EINVAL;
+
 	nx_ctx->op.inlen = (nx_ctx->in_sg - in_sg) * sizeof(struct nx_sg);
 	nx_ctx->op.outlen = (nx_ctx->out_sg - out_sg) * sizeof(struct nx_sg);
 
@@ -103,12 +113,23 @@ static int nx_xcbc_empty(struct shash_desc *desc, u8 *out)
 	/* XOr K3 with the padding for a 0 length message */
 	keys[1][0] ^= 0x80;
 
+	len = sizeof(keys[1]);
+
 	/* Encrypt the final result */
 	memcpy(csbcpb->cpb.aes_ecb.key, keys[0], AES_BLOCK_SIZE);
-	in_sg = nx_build_sg_list(nx_ctx->in_sg, (u8 *) keys[1], sizeof(keys[1]),
+	in_sg = nx_build_sg_list(nx_ctx->in_sg, (u8 *) keys[1], &len,
 				 nx_ctx->ap->sglen);
-	out_sg = nx_build_sg_list(nx_ctx->out_sg, out, AES_BLOCK_SIZE,
+
+	if (len != sizeof(keys[1]))
+		return -EINVAL;
+
+	len = AES_BLOCK_SIZE;
+	out_sg = nx_build_sg_list(nx_ctx->out_sg, out, &len,
 				  nx_ctx->ap->sglen);
+
+	if (len != AES_BLOCK_SIZE)
+		return -EINVAL;
+
 	nx_ctx->op.inlen = (nx_ctx->in_sg - in_sg) * sizeof(struct nx_sg);
 	nx_ctx->op.outlen = (nx_ctx->out_sg - out_sg) * sizeof(struct nx_sg);
 
@@ -133,6 +154,7 @@ static int nx_xcbc_init(struct shash_desc *desc)
 	struct nx_crypto_ctx *nx_ctx = crypto_tfm_ctx(&desc->tfm->base);
 	struct nx_csbcpb *csbcpb = nx_ctx->csbcpb;
 	struct nx_sg *out_sg;
+	int len;
 
 	nx_ctx_init(nx_ctx, HCOP_FC_AES);
 
@@ -144,8 +166,13 @@ static int nx_xcbc_init(struct shash_desc *desc)
 	memcpy(csbcpb->cpb.aes_xcbc.key, nx_ctx->priv.xcbc.key, AES_BLOCK_SIZE);
 	memset(nx_ctx->priv.xcbc.key, 0, sizeof *nx_ctx->priv.xcbc.key);
 
+	len = AES_BLOCK_SIZE;
 	out_sg = nx_build_sg_list(nx_ctx->out_sg, (u8 *)sctx->state,
-				  AES_BLOCK_SIZE, nx_ctx->ap->sglen);
+				  &len, nx_ctx->ap->sglen);
+
+	if (len != AES_BLOCK_SIZE)
+		return -EINVAL;
+
 	nx_ctx->op.outlen = (nx_ctx->out_sg - out_sg) * sizeof(struct nx_sg);
 
 	return 0;
@@ -159,10 +186,11 @@ static int nx_xcbc_update(struct shash_desc *desc,
 	struct nx_crypto_ctx *nx_ctx = crypto_tfm_ctx(&desc->tfm->base);
 	struct nx_csbcpb *csbcpb = nx_ctx->csbcpb;
 	struct nx_sg *in_sg;
-	u32 to_process, leftover, total;
-	u32 max_sg_len;
+	u32 to_process = 0, leftover, total;
+	unsigned int max_sg_len;
 	unsigned long irq_flags;
 	int rc = 0;
+	int data_len;
 
 	spin_lock_irqsave(&nx_ctx->lock, irq_flags);
 
@@ -180,17 +208,15 @@ static int nx_xcbc_update(struct shash_desc *desc,
 	}
 
 	in_sg = nx_ctx->in_sg;
-	max_sg_len = min_t(u32, nx_driver.of.max_sg_len/sizeof(struct nx_sg),
+	max_sg_len = min_t(u64, nx_driver.of.max_sg_len/sizeof(struct nx_sg),
 				nx_ctx->ap->sglen);
+	max_sg_len = min_t(u64, max_sg_len,
+				nx_ctx->ap->databytelen/NX_PAGE_SIZE);
 
 	do {
-
-		/* to_process: the AES_BLOCK_SIZE data chunk to process in this
-		 * update */
-		to_process = min_t(u64, total, nx_ctx->ap->databytelen);
-		to_process = min_t(u64, to_process,
-					NX_PAGE_SIZE * (max_sg_len - 1));
+		to_process = total - to_process;
 		to_process = to_process & ~(AES_BLOCK_SIZE - 1);
+
 		leftover = total - to_process;
 
 		/* the hardware will not accept a 0 byte operation for this
@@ -204,15 +230,24 @@ static int nx_xcbc_update(struct shash_desc *desc,
 		}
 
 		if (sctx->count) {
+			data_len = sctx->count;
 			in_sg = nx_build_sg_list(nx_ctx->in_sg,
 						(u8 *) sctx->buffer,
-						sctx->count,
+						&data_len,
 						max_sg_len);
+			if (data_len != sctx->count)
+				return -EINVAL;
 		}
+
+		data_len = to_process - sctx->count;
 		in_sg = nx_build_sg_list(in_sg,
 					(u8 *) data,
-					to_process - sctx->count,
+					&data_len,
 					max_sg_len);
+
+		if (data_len != to_process - sctx->count)
+			return -EINVAL;
+
 		nx_ctx->op.inlen = (nx_ctx->in_sg - in_sg) *
 					sizeof(struct nx_sg);
 
@@ -263,6 +298,7 @@ static int nx_xcbc_final(struct shash_desc *desc, u8 *out)
 	struct nx_sg *in_sg, *out_sg;
 	unsigned long irq_flags;
 	int rc = 0;
+	int len;
 
 	spin_lock_irqsave(&nx_ctx->lock, irq_flags);
 
@@ -285,11 +321,20 @@ static int nx_xcbc_final(struct shash_desc *desc, u8 *out)
 	 * this is not an intermediate operation */
 	NX_CPB_FDM(csbcpb) &= ~NX_FDM_INTERMEDIATE;
 
+	len = sctx->count;
 	in_sg = nx_build_sg_list(nx_ctx->in_sg, (u8 *)sctx->buffer,
-				 sctx->count, nx_ctx->ap->sglen);
-	out_sg = nx_build_sg_list(nx_ctx->out_sg, out, AES_BLOCK_SIZE,
+				 &len, nx_ctx->ap->sglen);
+
+	if (len != sctx->count)
+		return -EINVAL;
+
+	len = AES_BLOCK_SIZE;
+	out_sg = nx_build_sg_list(nx_ctx->out_sg, out, &len,
 				  nx_ctx->ap->sglen);
 
+	if (len != AES_BLOCK_SIZE)
+		return -EINVAL;
+
 	nx_ctx->op.inlen = (nx_ctx->in_sg - in_sg) * sizeof(struct nx_sg);
 	nx_ctx->op.outlen = (nx_ctx->out_sg - out_sg) * sizeof(struct nx_sg);
 
diff --git a/drivers/crypto/nx/nx-sha256.c b/drivers/crypto/nx/nx-sha256.c
index da0b24a7633f..23621da624c3 100644
--- a/drivers/crypto/nx/nx-sha256.c
+++ b/drivers/crypto/nx/nx-sha256.c
@@ -23,6 +23,7 @@
 #include <crypto/sha.h>
 #include <linux/module.h>
 #include <asm/vio.h>
+#include <asm/byteorder.h>
 
 #include "nx_csbcpb.h"
 #include "nx.h"
@@ -32,7 +33,8 @@ static int nx_sha256_init(struct shash_desc *desc)
 {
 	struct sha256_state *sctx = shash_desc_ctx(desc);
 	struct nx_crypto_ctx *nx_ctx = crypto_tfm_ctx(&desc->tfm->base);
-	struct nx_sg *out_sg;
+	int len;
+	int rc;
 
 	nx_ctx_init(nx_ctx, HCOP_FC_SHA);
 
@@ -41,10 +43,28 @@ static int nx_sha256_init(struct shash_desc *desc)
 	nx_ctx->ap = &nx_ctx->props[NX_PROPS_SHA256];
 
 	NX_CPB_SET_DIGEST_SIZE(nx_ctx->csbcpb, NX_DS_SHA256);
-	out_sg = nx_build_sg_list(nx_ctx->out_sg, (u8 *)sctx->state,
-				  SHA256_DIGEST_SIZE, nx_ctx->ap->sglen);
-	nx_ctx->op.outlen = (nx_ctx->out_sg - out_sg) * sizeof(struct nx_sg);
 
+	len = SHA256_DIGEST_SIZE;
+	rc = nx_sha_build_sg_list(nx_ctx, nx_ctx->out_sg,
+				  &nx_ctx->op.outlen,
+				  &len,
+				  (u8 *) sctx->state,
+				  NX_DS_SHA256);
+
+	if (rc)
+		goto out;
+
+	sctx->state[0] = __cpu_to_be32(SHA256_H0);
+	sctx->state[1] = __cpu_to_be32(SHA256_H1);
+	sctx->state[2] = __cpu_to_be32(SHA256_H2);
+	sctx->state[3] = __cpu_to_be32(SHA256_H3);
+	sctx->state[4] = __cpu_to_be32(SHA256_H4);
+	sctx->state[5] = __cpu_to_be32(SHA256_H5);
+	sctx->state[6] = __cpu_to_be32(SHA256_H6);
+	sctx->state[7] = __cpu_to_be32(SHA256_H7);
+	sctx->count = 0;
+
+out:
 	return 0;
 }
 
@@ -54,11 +74,11 @@ static int nx_sha256_update(struct shash_desc *desc, const u8 *data,
 	struct sha256_state *sctx = shash_desc_ctx(desc);
 	struct nx_crypto_ctx *nx_ctx = crypto_tfm_ctx(&desc->tfm->base);
 	struct nx_csbcpb *csbcpb = (struct nx_csbcpb *)nx_ctx->csbcpb;
-	struct nx_sg *in_sg;
-	u64 to_process, leftover, total;
-	u32 max_sg_len;
+	u64 to_process = 0, leftover, total;
 	unsigned long irq_flags;
 	int rc = 0;
+	int data_len;
+	u64 buf_len = (sctx->count % SHA256_BLOCK_SIZE);
 
 	spin_lock_irqsave(&nx_ctx->lock, irq_flags);
 
@@ -66,16 +86,16 @@ static int nx_sha256_update(struct shash_desc *desc, const u8 *data,
 	 *  1: < SHA256_BLOCK_SIZE: copy into state, return 0
 	 *  2: >= SHA256_BLOCK_SIZE: process X blocks, copy in leftover
 	 */
-	total = sctx->count + len;
+	total = (sctx->count % SHA256_BLOCK_SIZE) + len;
 	if (total < SHA256_BLOCK_SIZE) {
-		memcpy(sctx->buf + sctx->count, data, len);
+		memcpy(sctx->buf + buf_len, data, len);
 		sctx->count += len;
 		goto out;
 	}
 
-	in_sg = nx_ctx->in_sg;
-	max_sg_len = min_t(u32, nx_driver.of.max_sg_len/sizeof(struct nx_sg),
-			   nx_ctx->ap->sglen);
+	memcpy(csbcpb->cpb.sha256.message_digest, sctx->state, SHA256_DIGEST_SIZE);
+	NX_CPB_FDM(csbcpb) |= NX_FDM_INTERMEDIATE;
+	NX_CPB_FDM(csbcpb) |= NX_FDM_CONTINUATION;
 
 	do {
 		/*
@@ -83,34 +103,42 @@ static int nx_sha256_update(struct shash_desc *desc, const u8 *data,
 		 * this update. This value is also restricted by the sg list
 		 * limits.
 		 */
-		to_process = min_t(u64, total, nx_ctx->ap->databytelen);
-		to_process = min_t(u64, to_process,
-				   NX_PAGE_SIZE * (max_sg_len - 1));
+		to_process = total - to_process;
 		to_process = to_process & ~(SHA256_BLOCK_SIZE - 1);
-		leftover = total - to_process;
 
-		if (sctx->count) {
-			in_sg = nx_build_sg_list(nx_ctx->in_sg,
-						 (u8 *) sctx->buf,
-						 sctx->count, max_sg_len);
+		if (buf_len) {
+			data_len = buf_len;
+			rc = nx_sha_build_sg_list(nx_ctx, nx_ctx->in_sg,
+						  &nx_ctx->op.inlen,
+						  &data_len,
+						  (u8 *) sctx->buf,
+						  NX_DS_SHA256);
+
+			if (rc || data_len != buf_len)
+				goto out;
 		}
-		in_sg = nx_build_sg_list(in_sg, (u8 *) data,
-					 to_process - sctx->count,
-					 max_sg_len);
-		nx_ctx->op.inlen = (nx_ctx->in_sg - in_sg) *
-					sizeof(struct nx_sg);
-
-		if (NX_CPB_FDM(csbcpb) & NX_FDM_CONTINUATION) {
-			/*
-			 * we've hit the nx chip previously and we're updating
-			 * again, so copy over the partial digest.
-			 */
-			memcpy(csbcpb->cpb.sha256.input_partial_digest,
+
+		data_len = to_process - buf_len;
+		rc = nx_sha_build_sg_list(nx_ctx, nx_ctx->in_sg,
+					  &nx_ctx->op.inlen,
+					  &data_len,
+					  (u8 *) data,
+					  NX_DS_SHA256);
+
+		if (rc)
+			goto out;
+
+		to_process = (data_len + buf_len);
+		leftover = total - to_process;
+
+		/*
+		 * we've hit the nx chip previously and we're updating
+		 * again, so copy over the partial digest.
+		 */
+		memcpy(csbcpb->cpb.sha256.input_partial_digest,
 			       csbcpb->cpb.sha256.message_digest,
 			       SHA256_DIGEST_SIZE);
-		}
 
-		NX_CPB_FDM(csbcpb) |= NX_FDM_INTERMEDIATE;
 		if (!nx_ctx->op.inlen || !nx_ctx->op.outlen) {
 			rc = -EINVAL;
 			goto out;
@@ -122,22 +150,19 @@ static int nx_sha256_update(struct shash_desc *desc, const u8 *data,
 			goto out;
 
 		atomic_inc(&(nx_ctx->stats->sha256_ops));
-		csbcpb->cpb.sha256.message_bit_length += (u64)
-			(csbcpb->cpb.sha256.spbc * 8);
-
-		/* everything after the first update is continuation */
-		NX_CPB_FDM(csbcpb) |= NX_FDM_CONTINUATION;
 
 		total -= to_process;
-		data += to_process - sctx->count;
-		sctx->count = 0;
-		in_sg = nx_ctx->in_sg;
+		data += to_process - buf_len;
+		buf_len = 0;
+
 	} while (leftover >= SHA256_BLOCK_SIZE);
 
 	/* copy the leftover back into the state struct */
 	if (leftover)
 		memcpy(sctx->buf, data, leftover);
-	sctx->count = leftover;
+
+	sctx->count += len;
+	memcpy(sctx->state, csbcpb->cpb.sha256.message_digest, SHA256_DIGEST_SIZE);
 out:
 	spin_unlock_irqrestore(&nx_ctx->lock, irq_flags);
 	return rc;
@@ -148,34 +173,46 @@ static int nx_sha256_final(struct shash_desc *desc, u8 *out)
 	struct sha256_state *sctx = shash_desc_ctx(desc);
 	struct nx_crypto_ctx *nx_ctx = crypto_tfm_ctx(&desc->tfm->base);
 	struct nx_csbcpb *csbcpb = (struct nx_csbcpb *)nx_ctx->csbcpb;
-	struct nx_sg *in_sg, *out_sg;
-	u32 max_sg_len;
 	unsigned long irq_flags;
 	int rc;
+	int len;
 
 	spin_lock_irqsave(&nx_ctx->lock, irq_flags);
 
-	max_sg_len = min_t(u32, nx_driver.of.max_sg_len, nx_ctx->ap->sglen);
-
-	if (NX_CPB_FDM(csbcpb) & NX_FDM_CONTINUATION) {
+	/* final is represented by continuing the operation and indicating that
+	 * this is not an intermediate operation */
+	if (sctx->count >= SHA256_BLOCK_SIZE) {
 		/* we've hit the nx chip previously, now we're finalizing,
 		 * so copy over the partial digest */
-		memcpy(csbcpb->cpb.sha256.input_partial_digest,
-		       csbcpb->cpb.sha256.message_digest, SHA256_DIGEST_SIZE);
+		memcpy(csbcpb->cpb.sha256.input_partial_digest, sctx->state, SHA256_DIGEST_SIZE);
+		NX_CPB_FDM(csbcpb) &= ~NX_FDM_INTERMEDIATE;
+		NX_CPB_FDM(csbcpb) |= NX_FDM_CONTINUATION;
+	} else {
+		NX_CPB_FDM(csbcpb) &= ~NX_FDM_INTERMEDIATE;
+		NX_CPB_FDM(csbcpb) &= ~NX_FDM_CONTINUATION;
 	}
 
-	/* final is represented by continuing the operation and indicating that
-	 * this is not an intermediate operation */
-	NX_CPB_FDM(csbcpb) &= ~NX_FDM_INTERMEDIATE;
+	csbcpb->cpb.sha256.message_bit_length = (u64) (sctx->count * 8);
 
-	csbcpb->cpb.sha256.message_bit_length += (u64)(sctx->count * 8);
+	len = sctx->count & (SHA256_BLOCK_SIZE - 1);
+	rc = nx_sha_build_sg_list(nx_ctx, nx_ctx->in_sg,
+				  &nx_ctx->op.inlen,
+				  &len,
+				  (u8 *) sctx->buf,
+				  NX_DS_SHA256);
 
-	in_sg = nx_build_sg_list(nx_ctx->in_sg, (u8 *)sctx->buf,
-				 sctx->count, max_sg_len);
-	out_sg = nx_build_sg_list(nx_ctx->out_sg, out, SHA256_DIGEST_SIZE,
-				  max_sg_len);
-	nx_ctx->op.inlen = (nx_ctx->in_sg - in_sg) * sizeof(struct nx_sg);
-	nx_ctx->op.outlen = (nx_ctx->out_sg - out_sg) * sizeof(struct nx_sg);
+	if (rc || len != (sctx->count & (SHA256_BLOCK_SIZE - 1)))
+		goto out;
+
+	len = SHA256_DIGEST_SIZE;
+	rc = nx_sha_build_sg_list(nx_ctx, nx_ctx->out_sg,
+				  &nx_ctx->op.outlen,
+				  &len,
+				  out,
+				  NX_DS_SHA256);
+
+	if (rc || len != SHA256_DIGEST_SIZE)
+		goto out;
 
 	if (!nx_ctx->op.outlen) {
 		rc = -EINVAL;
@@ -189,8 +226,7 @@ static int nx_sha256_final(struct shash_desc *desc, u8 *out)
 
 	atomic_inc(&(nx_ctx->stats->sha256_ops));
 
-	atomic64_add(csbcpb->cpb.sha256.message_bit_length / 8,
-		     &(nx_ctx->stats->sha256_bytes));
+	atomic64_add(sctx->count, &(nx_ctx->stats->sha256_bytes));
 	memcpy(out, csbcpb->cpb.sha256.message_digest, SHA256_DIGEST_SIZE);
 out:
 	spin_unlock_irqrestore(&nx_ctx->lock, irq_flags);
@@ -200,62 +236,18 @@ out:
 static int nx_sha256_export(struct shash_desc *desc, void *out)
 {
 	struct sha256_state *sctx = shash_desc_ctx(desc);
-	struct nx_crypto_ctx *nx_ctx = crypto_tfm_ctx(&desc->tfm->base);
-	struct nx_csbcpb *csbcpb = (struct nx_csbcpb *)nx_ctx->csbcpb;
-	struct sha256_state *octx = out;
-	unsigned long irq_flags;
-
-	spin_lock_irqsave(&nx_ctx->lock, irq_flags);
 
-	octx->count = sctx->count +
-		      (csbcpb->cpb.sha256.message_bit_length / 8);
-	memcpy(octx->buf, sctx->buf, sizeof(octx->buf));
-
-	/* if no data has been processed yet, we need to export SHA256's
-	 * initial data, in case this context gets imported into a software
-	 * context */
-	if (csbcpb->cpb.sha256.message_bit_length)
-		memcpy(octx->state, csbcpb->cpb.sha256.message_digest,
-		       SHA256_DIGEST_SIZE);
-	else {
-		octx->state[0] = SHA256_H0;
-		octx->state[1] = SHA256_H1;
-		octx->state[2] = SHA256_H2;
-		octx->state[3] = SHA256_H3;
-		octx->state[4] = SHA256_H4;
-		octx->state[5] = SHA256_H5;
-		octx->state[6] = SHA256_H6;
-		octx->state[7] = SHA256_H7;
-	}
+	memcpy(out, sctx, sizeof(*sctx));
 
-	spin_unlock_irqrestore(&nx_ctx->lock, irq_flags);
 	return 0;
 }
 
 static int nx_sha256_import(struct shash_desc *desc, const void *in)
 {
 	struct sha256_state *sctx = shash_desc_ctx(desc);
-	struct nx_crypto_ctx *nx_ctx = crypto_tfm_ctx(&desc->tfm->base);
-	struct nx_csbcpb *csbcpb = (struct nx_csbcpb *)nx_ctx->csbcpb;
-	const struct sha256_state *ictx = in;
-	unsigned long irq_flags;
-
-	spin_lock_irqsave(&nx_ctx->lock, irq_flags);
 
-	memcpy(sctx->buf, ictx->buf, sizeof(ictx->buf));
+	memcpy(sctx, in, sizeof(*sctx));
 
-	sctx->count = ictx->count & 0x3f;
-	csbcpb->cpb.sha256.message_bit_length = (ictx->count & ~0x3f) * 8;
-
-	if (csbcpb->cpb.sha256.message_bit_length) {
-		memcpy(csbcpb->cpb.sha256.message_digest, ictx->state,
-		       SHA256_DIGEST_SIZE);
-
-		NX_CPB_FDM(csbcpb) |= NX_FDM_CONTINUATION;
-		NX_CPB_FDM(csbcpb) |= NX_FDM_INTERMEDIATE;
-	}
-
-	spin_unlock_irqrestore(&nx_ctx->lock, irq_flags);
 	return 0;
 }
 
diff --git a/drivers/crypto/nx/nx-sha512.c b/drivers/crypto/nx/nx-sha512.c
index 4ae5b0f221d5..b3adf1022673 100644
--- a/drivers/crypto/nx/nx-sha512.c
+++ b/drivers/crypto/nx/nx-sha512.c
@@ -32,7 +32,8 @@ static int nx_sha512_init(struct shash_desc *desc)
 {
 	struct sha512_state *sctx = shash_desc_ctx(desc);
 	struct nx_crypto_ctx *nx_ctx = crypto_tfm_ctx(&desc->tfm->base);
-	struct nx_sg *out_sg;
+	int len;
+	int rc;
 
 	nx_ctx_init(nx_ctx, HCOP_FC_SHA);
 
@@ -41,10 +42,28 @@ static int nx_sha512_init(struct shash_desc *desc)
 	nx_ctx->ap = &nx_ctx->props[NX_PROPS_SHA512];
 
 	NX_CPB_SET_DIGEST_SIZE(nx_ctx->csbcpb, NX_DS_SHA512);
-	out_sg = nx_build_sg_list(nx_ctx->out_sg, (u8 *)sctx->state,
-				  SHA512_DIGEST_SIZE, nx_ctx->ap->sglen);
-	nx_ctx->op.outlen = (nx_ctx->out_sg - out_sg) * sizeof(struct nx_sg);
 
+	len = SHA512_DIGEST_SIZE;
+	rc = nx_sha_build_sg_list(nx_ctx, nx_ctx->out_sg,
+				  &nx_ctx->op.outlen,
+				  &len,
+				  (u8 *)sctx->state,
+				  NX_DS_SHA512);
+
+	if (rc || len != SHA512_DIGEST_SIZE)
+		goto out;
+
+	sctx->state[0] = __cpu_to_be64(SHA512_H0);
+	sctx->state[1] = __cpu_to_be64(SHA512_H1);
+	sctx->state[2] = __cpu_to_be64(SHA512_H2);
+	sctx->state[3] = __cpu_to_be64(SHA512_H3);
+	sctx->state[4] = __cpu_to_be64(SHA512_H4);
+	sctx->state[5] = __cpu_to_be64(SHA512_H5);
+	sctx->state[6] = __cpu_to_be64(SHA512_H6);
+	sctx->state[7] = __cpu_to_be64(SHA512_H7);
+	sctx->count[0] = 0;
+
+out:
 	return 0;
 }
 
@@ -54,11 +73,11 @@ static int nx_sha512_update(struct shash_desc *desc, const u8 *data,
 	struct sha512_state *sctx = shash_desc_ctx(desc);
 	struct nx_crypto_ctx *nx_ctx = crypto_tfm_ctx(&desc->tfm->base);
 	struct nx_csbcpb *csbcpb = (struct nx_csbcpb *)nx_ctx->csbcpb;
-	struct nx_sg *in_sg;
-	u64 to_process, leftover, total, spbc_bits;
-	u32 max_sg_len;
+	u64 to_process, leftover = 0, total;
 	unsigned long irq_flags;
 	int rc = 0;
+	int data_len;
+	u64 buf_len = (sctx->count[0] % SHA512_BLOCK_SIZE);
 
 	spin_lock_irqsave(&nx_ctx->lock, irq_flags);
 
@@ -66,16 +85,16 @@ static int nx_sha512_update(struct shash_desc *desc, const u8 *data,
 	 *  1: < SHA512_BLOCK_SIZE: copy into state, return 0
 	 *  2: >= SHA512_BLOCK_SIZE: process X blocks, copy in leftover
 	 */
-	total = sctx->count[0] + len;
+	total = (sctx->count[0] % SHA512_BLOCK_SIZE) + len;
 	if (total < SHA512_BLOCK_SIZE) {
-		memcpy(sctx->buf + sctx->count[0], data, len);
+		memcpy(sctx->buf + buf_len, data, len);
 		sctx->count[0] += len;
 		goto out;
 	}
 
-	in_sg = nx_ctx->in_sg;
-	max_sg_len = min_t(u32, nx_driver.of.max_sg_len/sizeof(struct nx_sg),
-			   nx_ctx->ap->sglen);
+	memcpy(csbcpb->cpb.sha512.message_digest, sctx->state, SHA512_DIGEST_SIZE);
+	NX_CPB_FDM(csbcpb) |= NX_FDM_INTERMEDIATE;
+	NX_CPB_FDM(csbcpb) |= NX_FDM_CONTINUATION;
 
 	do {
 		/*
@@ -83,34 +102,43 @@ static int nx_sha512_update(struct shash_desc *desc, const u8 *data,
 		 * this update. This value is also restricted by the sg list
 		 * limits.
 		 */
-		to_process = min_t(u64, total, nx_ctx->ap->databytelen);
-		to_process = min_t(u64, to_process,
-				   NX_PAGE_SIZE * (max_sg_len - 1));
+		to_process = total - leftover;
 		to_process = to_process & ~(SHA512_BLOCK_SIZE - 1);
 		leftover = total - to_process;
 
-		if (sctx->count[0]) {
-			in_sg = nx_build_sg_list(nx_ctx->in_sg,
-						 (u8 *) sctx->buf,
-						 sctx->count[0], max_sg_len);
+		if (buf_len) {
+			data_len = buf_len;
+			rc = nx_sha_build_sg_list(nx_ctx, nx_ctx->in_sg,
+						  &nx_ctx->op.inlen,
+						  &data_len,
+						  (u8 *) sctx->buf,
+						  NX_DS_SHA512);
+
+			if (rc || data_len != buf_len)
+				goto out;
 		}
-		in_sg = nx_build_sg_list(in_sg, (u8 *) data,
-					 to_process - sctx->count[0],
-					 max_sg_len);
-		nx_ctx->op.inlen = (nx_ctx->in_sg - in_sg) *
-					sizeof(struct nx_sg);
-
-		if (NX_CPB_FDM(csbcpb) & NX_FDM_CONTINUATION) {
-			/*
-			 * we've hit the nx chip previously and we're updating
-			 * again, so copy over the partial digest.
-			 */
-			memcpy(csbcpb->cpb.sha512.input_partial_digest,
+
+		data_len = to_process - buf_len;
+		rc = nx_sha_build_sg_list(nx_ctx, nx_ctx->in_sg,
+					  &nx_ctx->op.inlen,
+					  &data_len,
+					  (u8 *) data,
+					  NX_DS_SHA512);
+
+		if (rc || data_len != (to_process - buf_len))
+			goto out;
+
+		to_process = (data_len + buf_len);
+		leftover = total - to_process;
+
+		/*
+		 * we've hit the nx chip previously and we're updating
+		 * again, so copy over the partial digest.
+		 */
+		memcpy(csbcpb->cpb.sha512.input_partial_digest,
 			       csbcpb->cpb.sha512.message_digest,
 			       SHA512_DIGEST_SIZE);
-		}
 
-		NX_CPB_FDM(csbcpb) |= NX_FDM_INTERMEDIATE;
 		if (!nx_ctx->op.inlen || !nx_ctx->op.outlen) {
 			rc = -EINVAL;
 			goto out;
@@ -122,24 +150,18 @@ static int nx_sha512_update(struct shash_desc *desc, const u8 *data,
 			goto out;
 
 		atomic_inc(&(nx_ctx->stats->sha512_ops));
-		spbc_bits = csbcpb->cpb.sha512.spbc * 8;
-		csbcpb->cpb.sha512.message_bit_length_lo += spbc_bits;
-		if (csbcpb->cpb.sha512.message_bit_length_lo < spbc_bits)
-			csbcpb->cpb.sha512.message_bit_length_hi++;
-
-		/* everything after the first update is continuation */
-		NX_CPB_FDM(csbcpb) |= NX_FDM_CONTINUATION;
 
 		total -= to_process;
-		data += to_process - sctx->count[0];
-		sctx->count[0] = 0;
-		in_sg = nx_ctx->in_sg;
+		data += to_process - buf_len;
+		buf_len = 0;
+
 	} while (leftover >= SHA512_BLOCK_SIZE);
 
 	/* copy the leftover back into the state struct */
 	if (leftover)
 		memcpy(sctx->buf, data, leftover);
-	sctx->count[0] = leftover;
+	sctx->count[0] += len;
+	memcpy(sctx->state, csbcpb->cpb.sha512.message_digest, SHA512_DIGEST_SIZE);
 out:
 	spin_unlock_irqrestore(&nx_ctx->lock, irq_flags);
 	return rc;
@@ -150,39 +172,52 @@ static int nx_sha512_final(struct shash_desc *desc, u8 *out)
 	struct sha512_state *sctx = shash_desc_ctx(desc);
 	struct nx_crypto_ctx *nx_ctx = crypto_tfm_ctx(&desc->tfm->base);
 	struct nx_csbcpb *csbcpb = (struct nx_csbcpb *)nx_ctx->csbcpb;
-	struct nx_sg *in_sg, *out_sg;
-	u32 max_sg_len;
 	u64 count0;
 	unsigned long irq_flags;
 	int rc;
+	int len;
 
 	spin_lock_irqsave(&nx_ctx->lock, irq_flags);
 
-	max_sg_len = min_t(u32, nx_driver.of.max_sg_len, nx_ctx->ap->sglen);
-
-	if (NX_CPB_FDM(csbcpb) & NX_FDM_CONTINUATION) {
+	/* final is represented by continuing the operation and indicating that
+	 * this is not an intermediate operation */
+	if (sctx->count[0] >= SHA512_BLOCK_SIZE) {
 		/* we've hit the nx chip previously, now we're finalizing,
 		 * so copy over the partial digest */
-		memcpy(csbcpb->cpb.sha512.input_partial_digest,
-		       csbcpb->cpb.sha512.message_digest, SHA512_DIGEST_SIZE);
+		memcpy(csbcpb->cpb.sha512.input_partial_digest, sctx->state,
+							SHA512_DIGEST_SIZE);
+		NX_CPB_FDM(csbcpb) &= ~NX_FDM_INTERMEDIATE;
+		NX_CPB_FDM(csbcpb) |= NX_FDM_CONTINUATION;
+	} else {
+		NX_CPB_FDM(csbcpb) &= ~NX_FDM_INTERMEDIATE;
+		NX_CPB_FDM(csbcpb) &= ~NX_FDM_CONTINUATION;
 	}
 
-	/* final is represented by continuing the operation and indicating that
-	 * this is not an intermediate operation */
 	NX_CPB_FDM(csbcpb) &= ~NX_FDM_INTERMEDIATE;
 
 	count0 = sctx->count[0] * 8;
 
-	csbcpb->cpb.sha512.message_bit_length_lo += count0;
-	if (csbcpb->cpb.sha512.message_bit_length_lo < count0)
-		csbcpb->cpb.sha512.message_bit_length_hi++;
+	csbcpb->cpb.sha512.message_bit_length_lo = count0;
 
-	in_sg = nx_build_sg_list(nx_ctx->in_sg, sctx->buf, sctx->count[0],
-				 max_sg_len);
-	out_sg = nx_build_sg_list(nx_ctx->out_sg, out, SHA512_DIGEST_SIZE,
-				  max_sg_len);
-	nx_ctx->op.inlen = (nx_ctx->in_sg - in_sg) * sizeof(struct nx_sg);
-	nx_ctx->op.outlen = (nx_ctx->out_sg - out_sg) * sizeof(struct nx_sg);
+	len = sctx->count[0] & (SHA512_BLOCK_SIZE - 1);
+	rc = nx_sha_build_sg_list(nx_ctx, nx_ctx->in_sg,
+				  &nx_ctx->op.inlen,
+				  &len,
+				  (u8 *)sctx->buf,
+				  NX_DS_SHA512);
+
+	if (rc || len != (sctx->count[0] & (SHA512_BLOCK_SIZE - 1)))
+		goto out;
+
+	len = SHA512_DIGEST_SIZE;
+	rc = nx_sha_build_sg_list(nx_ctx, nx_ctx->out_sg,
+				  &nx_ctx->op.outlen,
+				  &len,
+				  out,
+				  NX_DS_SHA512);
+
+	if (rc)
+		goto out;
 
 	if (!nx_ctx->op.outlen) {
 		rc = -EINVAL;
@@ -195,8 +230,7 @@ static int nx_sha512_final(struct shash_desc *desc, u8 *out)
 		goto out;
 
 	atomic_inc(&(nx_ctx->stats->sha512_ops));
-	atomic64_add(csbcpb->cpb.sha512.message_bit_length_lo / 8,
-		     &(nx_ctx->stats->sha512_bytes));
+	atomic64_add(sctx->count[0], &(nx_ctx->stats->sha512_bytes));
 
 	memcpy(out, csbcpb->cpb.sha512.message_digest, SHA512_DIGEST_SIZE);
 out:
@@ -207,74 +241,18 @@ out:
 static int nx_sha512_export(struct shash_desc *desc, void *out)
 {
 	struct sha512_state *sctx = shash_desc_ctx(desc);
-	struct nx_crypto_ctx *nx_ctx = crypto_tfm_ctx(&desc->tfm->base);
-	struct nx_csbcpb *csbcpb = (struct nx_csbcpb *)nx_ctx->csbcpb;
-	struct sha512_state *octx = out;
-	unsigned long irq_flags;
 
-	spin_lock_irqsave(&nx_ctx->lock, irq_flags);
+	memcpy(out, sctx, sizeof(*sctx));
 
-	/* move message_bit_length (128 bits) into count and convert its value
-	 * to bytes */
-	octx->count[0] = csbcpb->cpb.sha512.message_bit_length_lo >> 3 |
-			 ((csbcpb->cpb.sha512.message_bit_length_hi & 7) << 61);
-	octx->count[1] = csbcpb->cpb.sha512.message_bit_length_hi >> 3;
-
-	octx->count[0] += sctx->count[0];
-	if (octx->count[0] < sctx->count[0])
-		octx->count[1]++;
-
-	memcpy(octx->buf, sctx->buf, sizeof(octx->buf));
-
-	/* if no data has been processed yet, we need to export SHA512's
-	 * initial data, in case this context gets imported into a software
-	 * context */
-	if (csbcpb->cpb.sha512.message_bit_length_hi ||
-	    csbcpb->cpb.sha512.message_bit_length_lo)
-		memcpy(octx->state, csbcpb->cpb.sha512.message_digest,
-		       SHA512_DIGEST_SIZE);
-	else {
-		octx->state[0] = SHA512_H0;
-		octx->state[1] = SHA512_H1;
-		octx->state[2] = SHA512_H2;
-		octx->state[3] = SHA512_H3;
-		octx->state[4] = SHA512_H4;
-		octx->state[5] = SHA512_H5;
-		octx->state[6] = SHA512_H6;
-		octx->state[7] = SHA512_H7;
-	}
-
-	spin_unlock_irqrestore(&nx_ctx->lock, irq_flags);
 	return 0;
 }
 
 static int nx_sha512_import(struct shash_desc *desc, const void *in)
 {
 	struct sha512_state *sctx = shash_desc_ctx(desc);
-	struct nx_crypto_ctx *nx_ctx = crypto_tfm_ctx(&desc->tfm->base);
-	struct nx_csbcpb *csbcpb = (struct nx_csbcpb *)nx_ctx->csbcpb;
-	const struct sha512_state *ictx = in;
-	unsigned long irq_flags;
-
-	spin_lock_irqsave(&nx_ctx->lock, irq_flags);
-
-	memcpy(sctx->buf, ictx->buf, sizeof(ictx->buf));
-	sctx->count[0] = ictx->count[0] & 0x3f;
-	csbcpb->cpb.sha512.message_bit_length_lo = (ictx->count[0] & ~0x3f)
-							<< 3;
-	csbcpb->cpb.sha512.message_bit_length_hi = ictx->count[1] << 3 |
-						   ictx->count[0] >> 61;
-
-	if (csbcpb->cpb.sha512.message_bit_length_hi ||
-	    csbcpb->cpb.sha512.message_bit_length_lo) {
-		memcpy(csbcpb->cpb.sha512.message_digest, ictx->state,
-		       SHA512_DIGEST_SIZE);
 
-		NX_CPB_FDM(csbcpb) |= NX_FDM_CONTINUATION;
-		NX_CPB_FDM(csbcpb) |= NX_FDM_INTERMEDIATE;
-	}
+	memcpy(sctx, in, sizeof(*sctx));
 
-	spin_unlock_irqrestore(&nx_ctx->lock, irq_flags);
 	return 0;
 }
 
diff --git a/drivers/crypto/nx/nx.c b/drivers/crypto/nx/nx.c
index 5533fe31c90d..a392465d3e3f 100644
--- a/drivers/crypto/nx/nx.c
+++ b/drivers/crypto/nx/nx.c
@@ -90,7 +90,7 @@ int nx_hcall_sync(struct nx_crypto_ctx *nx_ctx,
  */
 struct nx_sg *nx_build_sg_list(struct nx_sg *sg_head,
 			       u8           *start_addr,
-			       unsigned int  len,
+			       unsigned int *len,
 			       u32           sgmax)
 {
 	unsigned int sg_len = 0;
@@ -106,7 +106,7 @@ struct nx_sg *nx_build_sg_list(struct nx_sg *sg_head,
 	else
 		sg_addr = __pa(sg_addr);
 
-	end_addr = sg_addr + len;
+	end_addr = sg_addr + *len;
 
 	/* each iteration will write one struct nx_sg element and add the
 	 * length of data described by that element to sg_len. Once @len bytes
@@ -118,7 +118,7 @@ struct nx_sg *nx_build_sg_list(struct nx_sg *sg_head,
 	 * Also when using vmalloc'ed data, every time that a system page
 	 * boundary is crossed the physical address needs to be re-calculated.
 	 */
-	for (sg = sg_head; sg_len < len; sg++) {
+	for (sg = sg_head; sg_len < *len; sg++) {
 		u64 next_page;
 
 		sg->addr = sg_addr;
@@ -133,15 +133,17 @@ struct nx_sg *nx_build_sg_list(struct nx_sg *sg_head,
 				is_vmalloc_addr(start_addr + sg_len)) {
 			sg_addr = page_to_phys(vmalloc_to_page(
 						start_addr + sg_len));
-			end_addr = sg_addr + len - sg_len;
+			end_addr = sg_addr + *len - sg_len;
 		}
 
 		if ((sg - sg_head) == sgmax) {
 			pr_err("nx: scatter/gather list overflow, pid: %d\n",
 			       current->pid);
-			return NULL;
+			sg++;
+			break;
 		}
 	}
+	*len = sg_len;
 
 	/* return the moved sg_head pointer */
 	return sg;
@@ -160,11 +162,11 @@ struct nx_sg *nx_walk_and_build(struct nx_sg       *nx_dst,
 				unsigned int        sglen,
 				struct scatterlist *sg_src,
 				unsigned int        start,
-				unsigned int        src_len)
+				unsigned int       *src_len)
 {
 	struct scatter_walk walk;
 	struct nx_sg *nx_sg = nx_dst;
-	unsigned int n, offset = 0, len = src_len;
+	unsigned int n, offset = 0, len = *src_len;
 	char *dst;
 
 	/* we need to fast forward through @start bytes first */
@@ -182,27 +184,101 @@ struct nx_sg *nx_walk_and_build(struct nx_sg       *nx_dst,
 	 * element we're currently looking at */
 	scatterwalk_advance(&walk, start - offset);
 
-	while (len && nx_sg) {
+	while (len && (nx_sg - nx_dst) < sglen) {
 		n = scatterwalk_clamp(&walk, len);
 		if (!n) {
-			scatterwalk_start(&walk, sg_next(walk.sg));
+			/* In cases where we have scatterlist chain scatterwalk_sg_next
+			 * handles with it properly */
+			scatterwalk_start(&walk, scatterwalk_sg_next(walk.sg));
 			n = scatterwalk_clamp(&walk, len);
 		}
 		dst = scatterwalk_map(&walk);
 
-		nx_sg = nx_build_sg_list(nx_sg, dst, n, sglen);
+		nx_sg = nx_build_sg_list(nx_sg, dst, &n, sglen - (nx_sg - nx_dst));
 		len -= n;
 
 		scatterwalk_unmap(dst);
 		scatterwalk_advance(&walk, n);
 		scatterwalk_done(&walk, SCATTERWALK_FROM_SG, len);
 	}
+	/* update to_process */
+	*src_len -= len;
 
 	/* return the moved destination pointer */
 	return nx_sg;
 }
 
 /**
+ * trim_sg_list - ensures the bound in sg list.
+ * @sg: sg list head
+ * @end: sg lisg end
+ * @delta:  is the amount we need to crop in order to bound the list.
+ *
+ */
+static long int trim_sg_list(struct nx_sg *sg, struct nx_sg *end, unsigned int delta)
+{
+	while (delta && end > sg) {
+		struct nx_sg *last = end - 1;
+
+		if (last->len > delta) {
+			last->len -= delta;
+			delta = 0;
+		} else {
+			end--;
+			delta -= last->len;
+		}
+	}
+	return (sg - end) * sizeof(struct nx_sg);
+}
+
+/**
+ * nx_sha_build_sg_list - walk and build sg list to sha modes
+ *			  using right bounds and limits.
+ * @nx_ctx: NX crypto context for the lists we're building
+ * @nx_sg: current sg list in or out list
+ * @op_len: current op_len to be used in order to build a sg list
+ * @nbytes:  number or bytes to be processed
+ * @offset: buf offset
+ * @mode: SHA256 or SHA512
+ */
+int nx_sha_build_sg_list(struct nx_crypto_ctx *nx_ctx,
+			  struct nx_sg 	      *nx_in_outsg,
+			  s64		      *op_len,
+			  unsigned int        *nbytes,
+			  u8 		      *offset,
+			  u32		      mode)
+{
+	unsigned int delta = 0;
+	unsigned int total = *nbytes;
+	struct nx_sg *nx_insg = nx_in_outsg;
+	unsigned int max_sg_len;
+
+	max_sg_len = min_t(u64, nx_ctx->ap->sglen,
+			nx_driver.of.max_sg_len/sizeof(struct nx_sg));
+	max_sg_len = min_t(u64, max_sg_len,
+			nx_ctx->ap->databytelen/NX_PAGE_SIZE);
+
+	*nbytes = min_t(u64, *nbytes, nx_ctx->ap->databytelen);
+	nx_insg = nx_build_sg_list(nx_insg, offset, nbytes, max_sg_len);
+
+	switch (mode) {
+	case NX_DS_SHA256:
+		if (*nbytes < total)
+			delta = *nbytes - (*nbytes & ~(SHA256_BLOCK_SIZE - 1));
+		break;
+	case NX_DS_SHA512:
+		if (*nbytes < total)
+			delta = *nbytes - (*nbytes & ~(SHA512_BLOCK_SIZE - 1));
+		break;
+	default:
+		return -EINVAL;
+	}
+	*op_len = trim_sg_list(nx_in_outsg, nx_insg, delta);
+
+	return 0;
+}
+
+/**
  * nx_build_sg_lists - walk the input scatterlists and build arrays of NX
  *                     scatterlists based on them.
  *
@@ -223,26 +299,39 @@ int nx_build_sg_lists(struct nx_crypto_ctx  *nx_ctx,
 		      struct blkcipher_desc *desc,
 		      struct scatterlist    *dst,
 		      struct scatterlist    *src,
-		      unsigned int           nbytes,
+		      unsigned int          *nbytes,
 		      unsigned int           offset,
 		      u8                    *iv)
 {
+	unsigned int delta = 0;
+	unsigned int total = *nbytes;
 	struct nx_sg *nx_insg = nx_ctx->in_sg;
 	struct nx_sg *nx_outsg = nx_ctx->out_sg;
+	unsigned int max_sg_len;
+
+	max_sg_len = min_t(u64, nx_ctx->ap->sglen,
+			nx_driver.of.max_sg_len/sizeof(struct nx_sg));
+	max_sg_len = min_t(u64, max_sg_len,
+			nx_ctx->ap->databytelen/NX_PAGE_SIZE);
 
 	if (iv)
 		memcpy(iv, desc->info, AES_BLOCK_SIZE);
 
-	nx_insg = nx_walk_and_build(nx_insg, nx_ctx->ap->sglen, src,
-				    offset, nbytes);
-	nx_outsg = nx_walk_and_build(nx_outsg, nx_ctx->ap->sglen, dst,
-				    offset, nbytes);
+	*nbytes = min_t(u64, *nbytes, nx_ctx->ap->databytelen);
+
+	nx_outsg = nx_walk_and_build(nx_outsg, max_sg_len, dst,
+					offset, nbytes);
+	nx_insg = nx_walk_and_build(nx_insg, max_sg_len, src,
+					offset, nbytes);
+
+	if (*nbytes < total)
+		delta = *nbytes - (*nbytes & ~(AES_BLOCK_SIZE - 1));
 
 	/* these lengths should be negative, which will indicate to phyp that
 	 * the input and output parameters are scatterlists, not linear
 	 * buffers */
-	nx_ctx->op.inlen = (nx_ctx->in_sg - nx_insg) * sizeof(struct nx_sg);
-	nx_ctx->op.outlen = (nx_ctx->out_sg - nx_outsg) * sizeof(struct nx_sg);
+	nx_ctx->op.inlen = trim_sg_list(nx_ctx->in_sg, nx_insg, delta);
+	nx_ctx->op.outlen = trim_sg_list(nx_ctx->out_sg, nx_outsg, delta);
 
 	return 0;
 }
@@ -540,10 +629,10 @@ static int nx_crypto_ctx_init(struct nx_crypto_ctx *nx_ctx, u32 fc, u32 mode)
 
 	/* we need an extra page for csbcpb_aead for these modes */
 	if (mode == NX_MODE_AES_GCM || mode == NX_MODE_AES_CCM)
-		nx_ctx->kmem_len = (4 * NX_PAGE_SIZE) +
+		nx_ctx->kmem_len = (5 * NX_PAGE_SIZE) +
 				   sizeof(struct nx_csbcpb);
 	else
-		nx_ctx->kmem_len = (3 * NX_PAGE_SIZE) +
+		nx_ctx->kmem_len = (4 * NX_PAGE_SIZE) +
 				   sizeof(struct nx_csbcpb);
 
 	nx_ctx->kmem = kmalloc(nx_ctx->kmem_len, GFP_KERNEL);
diff --git a/drivers/crypto/nx/nx.h b/drivers/crypto/nx/nx.h
index befda07ca1da..6c9ecaaead52 100644
--- a/drivers/crypto/nx/nx.h
+++ b/drivers/crypto/nx/nx.h
@@ -153,13 +153,15 @@ void nx_crypto_ctx_exit(struct crypto_tfm *tfm);
 void nx_ctx_init(struct nx_crypto_ctx *nx_ctx, unsigned int function);
 int nx_hcall_sync(struct nx_crypto_ctx *ctx, struct vio_pfo_op *op,
 		  u32 may_sleep);
-struct nx_sg *nx_build_sg_list(struct nx_sg *, u8 *, unsigned int, u32);
+int nx_sha_build_sg_list(struct nx_crypto_ctx *, struct nx_sg *,
+			 s64 *, unsigned int *, u8 *, u32);
+struct nx_sg *nx_build_sg_list(struct nx_sg *, u8 *, unsigned int *, u32);
 int nx_build_sg_lists(struct nx_crypto_ctx *, struct blkcipher_desc *,
-		      struct scatterlist *, struct scatterlist *, unsigned int,
+		      struct scatterlist *, struct scatterlist *, unsigned int *,
 		      unsigned int, u8 *);
 struct nx_sg *nx_walk_and_build(struct nx_sg *, unsigned int,
 				struct scatterlist *, unsigned int,
-				unsigned int);
+				unsigned int *);
 
 #ifdef CONFIG_DEBUG_FS
 #define NX_DEBUGFS_INIT(drv)	nx_debugfs_init(drv)
diff --git a/drivers/crypto/omap-aes.c b/drivers/crypto/omap-aes.c
index cb98fa54573d..f79dd410dede 100644
--- a/drivers/crypto/omap-aes.c
+++ b/drivers/crypto/omap-aes.c
@@ -1314,7 +1314,6 @@ static struct platform_driver omap_aes_driver = {
 	.remove	= omap_aes_remove,
 	.driver	= {
 		.name	= "omap-aes",
-		.owner	= THIS_MODULE,
 		.pm	= &omap_aes_pm_ops,
 		.of_match_table	= omap_aes_of_match,
 	},
diff --git a/drivers/crypto/omap-des.c b/drivers/crypto/omap-des.c
index b8bc84be8741..e350f5be4d2e 100644
--- a/drivers/crypto/omap-des.c
+++ b/drivers/crypto/omap-des.c
@@ -1222,7 +1222,6 @@ static struct platform_driver omap_des_driver = {
 	.remove	= omap_des_remove,
 	.driver	= {
 		.name	= "omap-des",
-		.owner	= THIS_MODULE,
 		.pm	= &omap_des_pm_ops,
 		.of_match_table	= of_match_ptr(omap_des_of_match),
 	},
diff --git a/drivers/crypto/omap-sham.c b/drivers/crypto/omap-sham.c
index 710d86386965..3c76696ee578 100644
--- a/drivers/crypto/omap-sham.c
+++ b/drivers/crypto/omap-sham.c
@@ -949,17 +949,14 @@ static int omap_sham_finish_hmac(struct ahash_request *req)
 	struct omap_sham_hmac_ctx *bctx = tctx->base;
 	int bs = crypto_shash_blocksize(bctx->shash);
 	int ds = crypto_shash_digestsize(bctx->shash);
-	struct {
-		struct shash_desc shash;
-		char ctx[crypto_shash_descsize(bctx->shash)];
-	} desc;
+	SHASH_DESC_ON_STACK(shash, bctx->shash);
 
-	desc.shash.tfm = bctx->shash;
-	desc.shash.flags = 0; /* not CRYPTO_TFM_REQ_MAY_SLEEP */
+	shash->tfm = bctx->shash;
+	shash->flags = 0; /* not CRYPTO_TFM_REQ_MAY_SLEEP */
 
-	return crypto_shash_init(&desc.shash) ?:
-	       crypto_shash_update(&desc.shash, bctx->opad, bs) ?:
-	       crypto_shash_finup(&desc.shash, req->result, ds, req->result);
+	return crypto_shash_init(shash) ?:
+	       crypto_shash_update(shash, bctx->opad, bs) ?:
+	       crypto_shash_finup(shash, req->result, ds, req->result);
 }
 
 static int omap_sham_finish(struct ahash_request *req)
@@ -1118,18 +1115,15 @@ static int omap_sham_update(struct ahash_request *req)
 	return omap_sham_enqueue(req, OP_UPDATE);
 }
 
-static int omap_sham_shash_digest(struct crypto_shash *shash, u32 flags,
+static int omap_sham_shash_digest(struct crypto_shash *tfm, u32 flags,
 				  const u8 *data, unsigned int len, u8 *out)
 {
-	struct {
-		struct shash_desc shash;
-		char ctx[crypto_shash_descsize(shash)];
-	} desc;
+	SHASH_DESC_ON_STACK(shash, tfm);
 
-	desc.shash.tfm = shash;
-	desc.shash.flags = flags & CRYPTO_TFM_REQ_MAY_SLEEP;
+	shash->tfm = tfm;
+	shash->flags = flags & CRYPTO_TFM_REQ_MAY_SLEEP;
 
-	return crypto_shash_digest(&desc.shash, data, len, out);
+	return crypto_shash_digest(shash, data, len, out);
 }
 
 static int omap_sham_final_shash(struct ahash_request *req)
@@ -2035,7 +2029,6 @@ static struct platform_driver omap_sham_driver = {
 	.remove	= omap_sham_remove,
 	.driver	= {
 		.name	= "omap-sham",
-		.owner	= THIS_MODULE,
 		.pm	= &omap_sham_pm_ops,
 		.of_match_table	= omap_sham_of_match,
 	},
diff --git a/drivers/crypto/padlock-aes.c b/drivers/crypto/padlock-aes.c
index 633ba945e153..c178ed8c3908 100644
--- a/drivers/crypto/padlock-aes.c
+++ b/drivers/crypto/padlock-aes.c
@@ -563,4 +563,4 @@ MODULE_DESCRIPTION("VIA PadLock AES algorithm support");
 MODULE_LICENSE("GPL");
 MODULE_AUTHOR("Michal Ludvig");
 
-MODULE_ALIAS("aes");
+MODULE_ALIAS_CRYPTO("aes");
diff --git a/drivers/crypto/padlock-sha.c b/drivers/crypto/padlock-sha.c
index bace885634f2..95f7d27ce491 100644
--- a/drivers/crypto/padlock-sha.c
+++ b/drivers/crypto/padlock-sha.c
@@ -593,7 +593,7 @@ MODULE_DESCRIPTION("VIA PadLock SHA1/SHA256 algorithms support.");
 MODULE_LICENSE("GPL");
 MODULE_AUTHOR("Michal Ludvig");
 
-MODULE_ALIAS("sha1-all");
-MODULE_ALIAS("sha256-all");
-MODULE_ALIAS("sha1-padlock");
-MODULE_ALIAS("sha256-padlock");
+MODULE_ALIAS_CRYPTO("sha1-all");
+MODULE_ALIAS_CRYPTO("sha256-all");
+MODULE_ALIAS_CRYPTO("sha1-padlock");
+MODULE_ALIAS_CRYPTO("sha256-padlock");
diff --git a/drivers/crypto/qat/qat_common/adf_accel_devices.h b/drivers/crypto/qat/qat_common/adf_accel_devices.h
index 9282381b03ce..2ed425664a16 100644
--- a/drivers/crypto/qat/qat_common/adf_accel_devices.h
+++ b/drivers/crypto/qat/qat_common/adf_accel_devices.h
@@ -56,8 +56,6 @@
 #define PCI_VENDOR_ID_INTEL 0x8086
 #define ADF_DH895XCC_DEVICE_NAME "dh895xcc"
 #define ADF_DH895XCC_PCI_DEVICE_ID 0x435
-#define ADF_DH895XCC_PMISC_BAR 1
-#define ADF_DH895XCC_ETR_BAR 2
 #define ADF_PCI_MAX_BARS 3
 #define ADF_DEVICE_NAME_LENGTH 32
 #define ADF_ETR_MAX_RINGS_PER_BANK 16
@@ -198,8 +196,7 @@ struct adf_accel_dev {
 	struct dentry *debugfs_dir;
 	struct list_head list;
 	struct module *owner;
-	uint8_t accel_id;
-	uint8_t numa_node;
 	struct adf_accel_pci accel_pci_dev;
+	uint8_t accel_id;
 } __packed;
 #endif
diff --git a/drivers/crypto/qat/qat_common/adf_aer.c b/drivers/crypto/qat/qat_common/adf_aer.c
index c29d4c3926bf..10ce4a2854ab 100644
--- a/drivers/crypto/qat/qat_common/adf_aer.c
+++ b/drivers/crypto/qat/qat_common/adf_aer.c
@@ -90,7 +90,7 @@ static void adf_dev_restore(struct adf_accel_dev *accel_dev)
 	uint16_t ppdstat = 0, bridge_ctl = 0;
 	int pending = 0;
 
-	pr_info("QAT: Reseting device qat_dev%d\n", accel_dev->accel_id);
+	pr_info("QAT: Resetting device qat_dev%d\n", accel_dev->accel_id);
 	pci_read_config_word(pdev, PPDSTAT_OFFSET, &ppdstat);
 	pending = ppdstat & PCI_EXP_DEVSTA_TRPND;
 	if (pending) {
diff --git a/drivers/crypto/qat/qat_common/adf_ctl_drv.c b/drivers/crypto/qat/qat_common/adf_ctl_drv.c
index d97069b8a8e4..7ee93f881db6 100644
--- a/drivers/crypto/qat/qat_common/adf_ctl_drv.c
+++ b/drivers/crypto/qat/qat_common/adf_ctl_drv.c
@@ -52,6 +52,7 @@
 #include <linux/pci.h>
 #include <linux/cdev.h>
 #include <linux/uaccess.h>
+#include <linux/crypto.h>
 
 #include "adf_accel_devices.h"
 #include "adf_common_drv.h"
@@ -111,7 +112,7 @@ static int adf_chr_drv_create(void)
 	drv_device = device_create(adt_ctl_drv.drv_class, NULL,
 				   MKDEV(adt_ctl_drv.major, 0),
 				   NULL, DEVICE_NAME);
-	if (!drv_device) {
+	if (IS_ERR(drv_device)) {
 		pr_err("QAT: failed to create device\n");
 		goto err_cdev_del;
 	}
@@ -436,7 +437,7 @@ static long adf_ctl_ioctl(struct file *fp, unsigned int cmd, unsigned long arg)
 		ret = adf_ctl_ioctl_get_status(fp, cmd, arg);
 		break;
 	default:
-		pr_err("QAT: Invalid ioclt\n");
+		pr_err("QAT: Invalid ioctl\n");
 		ret = -EFAULT;
 		break;
 	}
@@ -487,4 +488,4 @@ module_exit(adf_unregister_ctl_device_driver);
 MODULE_LICENSE("Dual BSD/GPL");
 MODULE_AUTHOR("Intel");
 MODULE_DESCRIPTION("Intel(R) QuickAssist Technology");
-MODULE_ALIAS("intel_qat");
+MODULE_ALIAS_CRYPTO("intel_qat");
diff --git a/drivers/crypto/qat/qat_common/adf_dev_mgr.c b/drivers/crypto/qat/qat_common/adf_dev_mgr.c
index ae71555c0868..4a0a829d4500 100644
--- a/drivers/crypto/qat/qat_common/adf_dev_mgr.c
+++ b/drivers/crypto/qat/qat_common/adf_dev_mgr.c
@@ -129,12 +129,13 @@ struct adf_accel_dev *adf_devmgr_get_first(void)
  * Function returns acceleration device associated with the given pci device.
  * To be used by QAT device specific drivers.
  *
- * Return: pinter to accel_dev or NULL if not found.
+ * Return: pointer to accel_dev or NULL if not found.
  */
 struct adf_accel_dev *adf_devmgr_pci_to_accel_dev(struct pci_dev *pci_dev)
 {
 	struct list_head *itr;
 
+	mutex_lock(&table_lock);
 	list_for_each(itr, &accel_table) {
 		struct adf_accel_dev *ptr =
 				list_entry(itr, struct adf_accel_dev, list);
@@ -144,6 +145,7 @@ struct adf_accel_dev *adf_devmgr_pci_to_accel_dev(struct pci_dev *pci_dev)
 			return ptr;
 		}
 	}
+	mutex_unlock(&table_lock);
 	return NULL;
 }
 EXPORT_SYMBOL_GPL(adf_devmgr_pci_to_accel_dev);
@@ -152,6 +154,7 @@ struct adf_accel_dev *adf_devmgr_get_dev_by_id(uint32_t id)
 {
 	struct list_head *itr;
 
+	mutex_lock(&table_lock);
 	list_for_each(itr, &accel_table) {
 		struct adf_accel_dev *ptr =
 				list_entry(itr, struct adf_accel_dev, list);
@@ -161,6 +164,7 @@ struct adf_accel_dev *adf_devmgr_get_dev_by_id(uint32_t id)
 			return ptr;
 		}
 	}
+	mutex_unlock(&table_lock);
 	return NULL;
 }
 
diff --git a/drivers/crypto/qat/qat_common/adf_transport.c b/drivers/crypto/qat/qat_common/adf_transport.c
index 5f3fa45348b4..7dd54aaee9fa 100644
--- a/drivers/crypto/qat/qat_common/adf_transport.c
+++ b/drivers/crypto/qat/qat_common/adf_transport.c
@@ -376,8 +376,9 @@ static inline int adf_get_cfg_int(struct adf_accel_dev *accel_dev,
 	return 0;
 }
 
-static void adf_enable_coalesc(struct adf_etr_bank_data *bank,
-			       const char *section, uint32_t bank_num_in_accel)
+static void adf_get_coalesc_timer(struct adf_etr_bank_data *bank,
+				  const char *section,
+				  uint32_t bank_num_in_accel)
 {
 	if (adf_get_cfg_int(bank->accel_dev, section,
 			    ADF_ETRMGR_COALESCE_TIMER_FORMAT,
@@ -396,7 +397,7 @@ static int adf_init_bank(struct adf_accel_dev *accel_dev,
 	struct adf_hw_device_data *hw_data = accel_dev->hw_device;
 	struct adf_etr_ring_data *ring;
 	struct adf_etr_ring_data *tx_ring;
-	uint32_t i, coalesc_enabled;
+	uint32_t i, coalesc_enabled = 0;
 
 	memset(bank, 0, sizeof(*bank));
 	bank->bank_number = bank_num;
@@ -407,10 +408,10 @@ static int adf_init_bank(struct adf_accel_dev *accel_dev,
 	/* Enable IRQ coalescing always. This will allow to use
 	 * the optimised flag and coalesc register.
 	 * If it is disabled in the config file just use min time value */
-	if (adf_get_cfg_int(accel_dev, "Accelerator0",
-			    ADF_ETRMGR_COALESCING_ENABLED_FORMAT,
-			    bank_num, &coalesc_enabled) && coalesc_enabled)
-		adf_enable_coalesc(bank, "Accelerator0", bank_num);
+	if ((adf_get_cfg_int(accel_dev, "Accelerator0",
+			     ADF_ETRMGR_COALESCING_ENABLED_FORMAT, bank_num,
+			     &coalesc_enabled) == 0) && coalesc_enabled)
+		adf_get_coalesc_timer(bank, "Accelerator0", bank_num);
 	else
 		bank->irq_coalesc_timer = ADF_COALESCING_MIN_TIME;
 
@@ -419,9 +420,10 @@ static int adf_init_bank(struct adf_accel_dev *accel_dev,
 		WRITE_CSR_RING_BASE(csr_addr, bank_num, i, 0);
 		ring = &bank->rings[i];
 		if (hw_data->tx_rings_mask & (1 << i)) {
-			ring->inflights = kzalloc_node(sizeof(atomic_t),
-						       GFP_KERNEL,
-						       accel_dev->numa_node);
+			ring->inflights =
+				kzalloc_node(sizeof(atomic_t),
+					     GFP_KERNEL,
+					     dev_to_node(&GET_DEV(accel_dev)));
 			if (!ring->inflights)
 				goto err;
 		} else {
@@ -469,13 +471,14 @@ int adf_init_etr_data(struct adf_accel_dev *accel_dev)
 	int i, ret;
 
 	etr_data = kzalloc_node(sizeof(*etr_data), GFP_KERNEL,
-				accel_dev->numa_node);
+				dev_to_node(&GET_DEV(accel_dev)));
 	if (!etr_data)
 		return -ENOMEM;
 
 	num_banks = GET_MAX_BANKS(accel_dev);
 	size = num_banks * sizeof(struct adf_etr_bank_data);
-	etr_data->banks = kzalloc_node(size, GFP_KERNEL, accel_dev->numa_node);
+	etr_data->banks = kzalloc_node(size, GFP_KERNEL,
+				       dev_to_node(&GET_DEV(accel_dev)));
 	if (!etr_data->banks) {
 		ret = -ENOMEM;
 		goto err_bank;
diff --git a/drivers/crypto/qat/qat_common/adf_transport_access_macros.h b/drivers/crypto/qat/qat_common/adf_transport_access_macros.h
index 91d88d676580..160c9a36c919 100644
--- a/drivers/crypto/qat/qat_common/adf_transport_access_macros.h
+++ b/drivers/crypto/qat/qat_common/adf_transport_access_macros.h
@@ -83,14 +83,14 @@
 #define ADF_MAX_RING_SIZE ADF_RING_SIZE_4M
 #define ADF_DEFAULT_RING_SIZE ADF_RING_SIZE_16K
 
-/* Valid internal msg size values internal */
+/* Valid internal msg size values */
 #define ADF_MSG_SIZE_32 0x01
 #define ADF_MSG_SIZE_64 0x02
 #define ADF_MSG_SIZE_128 0x04
 #define ADF_MIN_MSG_SIZE ADF_MSG_SIZE_32
 #define ADF_MAX_MSG_SIZE ADF_MSG_SIZE_128
 
-/* Size to bytes conversion macros for ring and msg values */
+/* Size to bytes conversion macros for ring and msg size values */
 #define ADF_MSG_SIZE_TO_BYTES(SIZE) (SIZE << 5)
 #define ADF_BYTES_TO_MSG_SIZE(SIZE) (SIZE >> 5)
 #define ADF_SIZE_TO_RING_SIZE_IN_BYTES(SIZE) ((1 << (SIZE - 1)) << 7)
@@ -100,8 +100,11 @@
 #define ADF_RING_SIZE_BYTES_MIN(SIZE) ((SIZE < ADF_RING_SIZE_4K) ? \
 				ADF_RING_SIZE_4K : SIZE)
 #define ADF_RING_SIZE_MODULO(SIZE) (SIZE + 0x6)
+#define ADF_SIZE_TO_POW(SIZE) ((((SIZE & 0x4) >> 1) | ((SIZE & 0x4) >> 2) | \
+				SIZE) & ~0x4)
+/* Max outstanding requests */
 #define ADF_MAX_INFLIGHTS(RING_SIZE, MSG_SIZE) \
-	((((1 << (RING_SIZE - 1)) << 4) >> MSG_SIZE) - 1)
+	((((1 << (RING_SIZE - 1)) << 3) >> ADF_SIZE_TO_POW(MSG_SIZE)) - 1)
 #define BUILD_RING_CONFIG(size)	\
 	((ADF_RING_NEAR_WATERMARK_0 << ADF_RING_CONFIG_NEAR_FULL_WM) \
 	| (ADF_RING_NEAR_WATERMARK_0 << ADF_RING_CONFIG_NEAR_EMPTY_WM) \
diff --git a/drivers/crypto/qat/qat_common/adf_transport_internal.h b/drivers/crypto/qat/qat_common/adf_transport_internal.h
index f854bac276b0..c40546079981 100644
--- a/drivers/crypto/qat/qat_common/adf_transport_internal.h
+++ b/drivers/crypto/qat/qat_common/adf_transport_internal.h
@@ -75,7 +75,7 @@ struct adf_etr_ring_data {
 
 struct adf_etr_bank_data {
 	struct adf_etr_ring_data rings[ADF_ETR_MAX_RINGS_PER_BANK];
-	struct tasklet_struct resp_hanlder;
+	struct tasklet_struct resp_handler;
 	void __iomem *csr_addr;
 	struct adf_accel_dev *accel_dev;
 	uint32_t irq_coalesc_timer;
diff --git a/drivers/crypto/qat/qat_common/qat_algs.c b/drivers/crypto/qat/qat_common/qat_algs.c
index 59df48872955..19eea1c832ac 100644
--- a/drivers/crypto/qat/qat_common/qat_algs.c
+++ b/drivers/crypto/qat/qat_common/qat_algs.c
@@ -105,7 +105,7 @@ struct qat_alg_cd {
 #define MAX_AUTH_STATE_SIZE sizeof(struct icp_qat_hw_auth_algo_blk)
 
 struct qat_auth_state {
-	uint8_t data[MAX_AUTH_STATE_SIZE];
+	uint8_t data[MAX_AUTH_STATE_SIZE + 64];
 } __aligned(64);
 
 struct qat_alg_session_ctx {
@@ -113,10 +113,6 @@ struct qat_alg_session_ctx {
 	dma_addr_t enc_cd_paddr;
 	struct qat_alg_cd *dec_cd;
 	dma_addr_t dec_cd_paddr;
-	struct qat_auth_state *auth_hw_state_enc;
-	dma_addr_t auth_state_enc_paddr;
-	struct qat_auth_state *auth_hw_state_dec;
-	dma_addr_t auth_state_dec_paddr;
 	struct icp_qat_fw_la_bulk_req enc_fw_req_tmpl;
 	struct icp_qat_fw_la_bulk_req dec_fw_req_tmpl;
 	struct qat_crypto_instance *inst;
@@ -150,42 +146,41 @@ static int qat_get_inter_state_size(enum icp_qat_hw_auth_algo qat_hash_alg)
 static int qat_alg_do_precomputes(struct icp_qat_hw_auth_algo_blk *hash,
 				  struct qat_alg_session_ctx *ctx,
 				  const uint8_t *auth_key,
-				  unsigned int auth_keylen, uint8_t *auth_state)
+				  unsigned int auth_keylen)
 {
-	struct {
-		struct shash_desc shash;
-		char ctx[crypto_shash_descsize(ctx->hash_tfm)];
-	} desc;
+	struct qat_auth_state auth_state;
+	SHASH_DESC_ON_STACK(shash, ctx->hash_tfm);
 	struct sha1_state sha1;
 	struct sha256_state sha256;
 	struct sha512_state sha512;
 	int block_size = crypto_shash_blocksize(ctx->hash_tfm);
 	int digest_size = crypto_shash_digestsize(ctx->hash_tfm);
-	uint8_t *ipad = auth_state;
+	uint8_t *ipad = auth_state.data;
 	uint8_t *opad = ipad + block_size;
 	__be32 *hash_state_out;
 	__be64 *hash512_state_out;
 	int i, offset;
 
-	desc.shash.tfm = ctx->hash_tfm;
-	desc.shash.flags = 0x0;
+	memzero_explicit(auth_state.data, MAX_AUTH_STATE_SIZE + 64);
+	shash->tfm = ctx->hash_tfm;
+	shash->flags = 0x0;
 
 	if (auth_keylen > block_size) {
 		char buff[SHA512_BLOCK_SIZE];
-		int ret = crypto_shash_digest(&desc.shash, auth_key,
+		int ret = crypto_shash_digest(shash, auth_key,
 					      auth_keylen, buff);
 		if (ret)
 			return ret;
 
 		memcpy(ipad, buff, digest_size);
 		memcpy(opad, buff, digest_size);
-		memset(ipad + digest_size, 0, block_size - digest_size);
-		memset(opad + digest_size, 0, block_size - digest_size);
+		memzero_explicit(ipad + digest_size, block_size - digest_size);
+		memzero_explicit(opad + digest_size, block_size - digest_size);
 	} else {
 		memcpy(ipad, auth_key, auth_keylen);
 		memcpy(opad, auth_key, auth_keylen);
-		memset(ipad + auth_keylen, 0, block_size - auth_keylen);
-		memset(opad + auth_keylen, 0, block_size - auth_keylen);
+		memzero_explicit(ipad + auth_keylen, block_size - auth_keylen);
+		memzero_explicit(opad + auth_keylen, block_size - auth_keylen);
 	}
 
 	for (i = 0; i < block_size; i++) {
@@ -195,10 +190,10 @@ static int qat_alg_do_precomputes(struct icp_qat_hw_auth_algo_blk *hash,
 		*opad_ptr ^= 0x5C;
 	}
 
-	if (crypto_shash_init(&desc.shash))
+	if (crypto_shash_init(shash))
 		return -EFAULT;
 
-	if (crypto_shash_update(&desc.shash, ipad, block_size))
+	if (crypto_shash_update(shash, ipad, block_size))
 		return -EFAULT;
 
 	hash_state_out = (__be32 *)hash->sha.state1;
@@ -206,19 +201,19 @@ static int qat_alg_do_precomputes(struct icp_qat_hw_auth_algo_blk *hash,
 
 	switch (ctx->qat_hash_alg) {
 	case ICP_QAT_HW_AUTH_ALGO_SHA1:
-		if (crypto_shash_export(&desc.shash, &sha1))
+		if (crypto_shash_export(shash, &sha1))
 			return -EFAULT;
 		for (i = 0; i < digest_size >> 2; i++, hash_state_out++)
 			*hash_state_out = cpu_to_be32(*(sha1.state + i));
 		break;
 	case ICP_QAT_HW_AUTH_ALGO_SHA256:
-		if (crypto_shash_export(&desc.shash, &sha256))
+		if (crypto_shash_export(shash, &sha256))
 			return -EFAULT;
 		for (i = 0; i < digest_size >> 2; i++, hash_state_out++)
 			*hash_state_out = cpu_to_be32(*(sha256.state + i));
 		break;
 	case ICP_QAT_HW_AUTH_ALGO_SHA512:
-		if (crypto_shash_export(&desc.shash, &sha512))
+		if (crypto_shash_export(shash, &sha512))
 			return -EFAULT;
 		for (i = 0; i < digest_size >> 3; i++, hash512_state_out++)
 			*hash512_state_out = cpu_to_be64(*(sha512.state + i));
@@ -227,10 +222,10 @@ static int qat_alg_do_precomputes(struct icp_qat_hw_auth_algo_blk *hash,
 		return -EFAULT;
 	}
 
-	if (crypto_shash_init(&desc.shash))
+	if (crypto_shash_init(shash))
 		return -EFAULT;
 
-	if (crypto_shash_update(&desc.shash, opad, block_size))
+	if (crypto_shash_update(shash, opad, block_size))
 		return -EFAULT;
 
 	offset = round_up(qat_get_inter_state_size(ctx->qat_hash_alg), 8);
@@ -239,19 +234,19 @@ static int qat_alg_do_precomputes(struct icp_qat_hw_auth_algo_blk *hash,
 
 	switch (ctx->qat_hash_alg) {
 	case ICP_QAT_HW_AUTH_ALGO_SHA1:
-		if (crypto_shash_export(&desc.shash, &sha1))
+		if (crypto_shash_export(shash, &sha1))
 			return -EFAULT;
 		for (i = 0; i < digest_size >> 2; i++, hash_state_out++)
 			*hash_state_out = cpu_to_be32(*(sha1.state + i));
 		break;
 	case ICP_QAT_HW_AUTH_ALGO_SHA256:
-		if (crypto_shash_export(&desc.shash, &sha256))
+		if (crypto_shash_export(shash, &sha256))
 			return -EFAULT;
 		for (i = 0; i < digest_size >> 2; i++, hash_state_out++)
 			*hash_state_out = cpu_to_be32(*(sha256.state + i));
 		break;
 	case ICP_QAT_HW_AUTH_ALGO_SHA512:
-		if (crypto_shash_export(&desc.shash, &sha512))
+		if (crypto_shash_export(shash, &sha512))
 			return -EFAULT;
 		for (i = 0; i < digest_size >> 3; i++, hash512_state_out++)
 			*hash512_state_out = cpu_to_be64(*(sha512.state + i));
@@ -259,6 +254,8 @@ static int qat_alg_do_precomputes(struct icp_qat_hw_auth_algo_blk *hash,
 	default:
 		return -EFAULT;
 	}
+	memzero_explicit(ipad, block_size);
+	memzero_explicit(opad, block_size);
 	return 0;
 }
 
@@ -298,10 +295,6 @@ static int qat_alg_init_enc_session(struct qat_alg_session_ctx *ctx,
 	void *ptr = &req_tmpl->cd_ctrl;
 	struct icp_qat_fw_cipher_cd_ctrl_hdr *cipher_cd_ctrl = ptr;
 	struct icp_qat_fw_auth_cd_ctrl_hdr *hash_cd_ctrl = ptr;
-	struct icp_qat_fw_la_auth_req_params *auth_param =
-		(struct icp_qat_fw_la_auth_req_params *)
-		((char *)&req_tmpl->serv_specif_rqpars +
-		 sizeof(struct icp_qat_fw_la_cipher_req_params));
 
 	/* CD setup */
 	cipher->aes.cipher_config.val = QAT_AES_HW_CONFIG_ENC(alg);
@@ -312,8 +305,7 @@ static int qat_alg_init_enc_session(struct qat_alg_session_ctx *ctx,
 	hash->sha.inner_setup.auth_counter.counter =
 		cpu_to_be32(crypto_shash_blocksize(ctx->hash_tfm));
 
-	if (qat_alg_do_precomputes(hash, ctx, keys->authkey, keys->authkeylen,
-				   (uint8_t *)ctx->auth_hw_state_enc))
+	if (qat_alg_do_precomputes(hash, ctx, keys->authkey, keys->authkeylen))
 		return -EFAULT;
 
 	/* Request setup */
@@ -359,9 +351,6 @@ static int qat_alg_init_enc_session(struct qat_alg_session_ctx *ctx,
 	hash_cd_ctrl->inner_state2_offset = hash_cd_ctrl->hash_cfg_offset +
 			((sizeof(struct icp_qat_hw_auth_setup) +
 			 round_up(hash_cd_ctrl->inner_state1_sz, 8)) >> 3);
-	auth_param->u1.auth_partial_st_prefix = ctx->auth_state_enc_paddr +
-			sizeof(struct icp_qat_hw_auth_counter) +
-			round_up(hash_cd_ctrl->inner_state1_sz, 8);
 	ICP_QAT_FW_COMN_CURR_ID_SET(hash_cd_ctrl, ICP_QAT_FW_SLICE_AUTH);
 	ICP_QAT_FW_COMN_NEXT_ID_SET(hash_cd_ctrl, ICP_QAT_FW_SLICE_DRAM_WR);
 	return 0;
@@ -399,8 +388,7 @@ static int qat_alg_init_dec_session(struct qat_alg_session_ctx *ctx,
 	hash->sha.inner_setup.auth_counter.counter =
 		cpu_to_be32(crypto_shash_blocksize(ctx->hash_tfm));
 
-	if (qat_alg_do_precomputes(hash, ctx, keys->authkey, keys->authkeylen,
-				   (uint8_t *)ctx->auth_hw_state_dec))
+	if (qat_alg_do_precomputes(hash, ctx, keys->authkey, keys->authkeylen))
 		return -EFAULT;
 
 	/* Request setup */
@@ -450,9 +438,6 @@ static int qat_alg_init_dec_session(struct qat_alg_session_ctx *ctx,
 	hash_cd_ctrl->inner_state2_offset = hash_cd_ctrl->hash_cfg_offset +
 			((sizeof(struct icp_qat_hw_auth_setup) +
 			 round_up(hash_cd_ctrl->inner_state1_sz, 8)) >> 3);
-	auth_param->u1.auth_partial_st_prefix = ctx->auth_state_enc_paddr +
-			sizeof(struct icp_qat_hw_auth_counter) +
-			round_up(hash_cd_ctrl->inner_state1_sz, 8);
 	auth_param->auth_res_sz = digestsize;
 	ICP_QAT_FW_COMN_CURR_ID_SET(hash_cd_ctrl, ICP_QAT_FW_SLICE_AUTH);
 	ICP_QAT_FW_COMN_NEXT_ID_SET(hash_cd_ctrl, ICP_QAT_FW_SLICE_CIPHER);
@@ -483,7 +468,6 @@ static int qat_alg_init_sessions(struct qat_alg_session_ctx *ctx,
 		break;
 	default:
 		goto bad_key;
-		break;
 	}
 
 	if (qat_alg_init_enc_session(ctx, alg, &keys))
@@ -510,16 +494,12 @@ static int qat_alg_setkey(struct crypto_aead *tfm, const uint8_t *key,
 	if (ctx->enc_cd) {
 		/* rekeying */
 		dev = &GET_DEV(ctx->inst->accel_dev);
-		memset(ctx->enc_cd, 0, sizeof(struct qat_alg_cd));
-		memset(ctx->dec_cd, 0, sizeof(struct qat_alg_cd));
-		memset(ctx->auth_hw_state_enc, 0,
-		       sizeof(struct qat_auth_state));
-		memset(ctx->auth_hw_state_dec, 0,
-		       sizeof(struct qat_auth_state));
-		memset(&ctx->enc_fw_req_tmpl, 0,
-		       sizeof(struct icp_qat_fw_la_bulk_req));
-		memset(&ctx->dec_fw_req_tmpl, 0,
-		       sizeof(struct icp_qat_fw_la_bulk_req));
+		memzero_explicit(ctx->enc_cd, sizeof(struct qat_alg_cd));
+		memzero_explicit(ctx->dec_cd, sizeof(struct qat_alg_cd));
+		memzero_explicit(&ctx->enc_fw_req_tmpl,
+				 sizeof(struct icp_qat_fw_la_bulk_req));
+		memzero_explicit(&ctx->dec_fw_req_tmpl,
+				 sizeof(struct icp_qat_fw_la_bulk_req));
 	} else {
 		/* new key */
 		int node = get_current_node();
@@ -548,22 +528,6 @@ static int qat_alg_setkey(struct crypto_aead *tfm, const uint8_t *key,
 			spin_unlock(&ctx->lock);
 			goto out_free_enc;
 		}
-		ctx->auth_hw_state_enc =
-			dma_zalloc_coherent(dev, sizeof(struct qat_auth_state),
-					    &ctx->auth_state_enc_paddr,
-					    GFP_ATOMIC);
-		if (!ctx->auth_hw_state_enc) {
-			spin_unlock(&ctx->lock);
-			goto out_free_dec;
-		}
-		ctx->auth_hw_state_dec =
-			dma_zalloc_coherent(dev, sizeof(struct qat_auth_state),
-					    &ctx->auth_state_dec_paddr,
-					    GFP_ATOMIC);
-		if (!ctx->auth_hw_state_dec) {
-			spin_unlock(&ctx->lock);
-			goto out_free_auth_enc;
-		}
 	}
 	spin_unlock(&ctx->lock);
 	if (qat_alg_init_sessions(ctx, key, keylen))
@@ -572,18 +536,12 @@ static int qat_alg_setkey(struct crypto_aead *tfm, const uint8_t *key,
 	return 0;
 
 out_free_all:
-	dma_free_coherent(dev, sizeof(struct qat_auth_state),
-			  ctx->auth_hw_state_dec, ctx->auth_state_dec_paddr);
-	ctx->auth_hw_state_dec = NULL;
-out_free_auth_enc:
-	dma_free_coherent(dev, sizeof(struct qat_auth_state),
-			  ctx->auth_hw_state_enc, ctx->auth_state_enc_paddr);
-	ctx->auth_hw_state_enc = NULL;
-out_free_dec:
+	memzero_explicit(ctx->dec_cd, sizeof(struct qat_alg_cd));
 	dma_free_coherent(dev, sizeof(struct qat_alg_cd),
 			  ctx->dec_cd, ctx->dec_cd_paddr);
 	ctx->dec_cd = NULL;
 out_free_enc:
+	memzero_explicit(ctx->enc_cd, sizeof(struct qat_alg_cd));
 	dma_free_coherent(dev, sizeof(struct qat_alg_cd),
 			  ctx->enc_cd, ctx->enc_cd_paddr);
 	ctx->enc_cd = NULL;
@@ -641,7 +599,8 @@ static int qat_alg_sgl_to_bufl(struct qat_crypto_instance *inst,
 	if (unlikely(!n))
 		return -EINVAL;
 
-	bufl = kmalloc_node(sz, GFP_ATOMIC, inst->accel_dev->numa_node);
+	bufl = kmalloc_node(sz, GFP_ATOMIC,
+			    dev_to_node(&GET_DEV(inst->accel_dev)));
 	if (unlikely(!bufl))
 		return -ENOMEM;
 
@@ -650,6 +609,8 @@ static int qat_alg_sgl_to_bufl(struct qat_crypto_instance *inst,
 		goto err;
 
 	for_each_sg(assoc, sg, assoc_n, i) {
+		if (!sg->length)
+			continue;
 		bufl->bufers[bufs].addr = dma_map_single(dev,
 							 sg_virt(sg),
 							 sg->length,
@@ -685,7 +646,7 @@ static int qat_alg_sgl_to_bufl(struct qat_crypto_instance *inst,
 		struct qat_alg_buf *bufers;
 
 		buflout = kmalloc_node(sz, GFP_ATOMIC,
-				       inst->accel_dev->numa_node);
+				       dev_to_node(&GET_DEV(inst->accel_dev)));
 		if (unlikely(!buflout))
 			goto err;
 		bloutp = dma_map_single(dev, buflout, sz, DMA_TO_DEVICE);
@@ -878,7 +839,7 @@ static int qat_alg_init(struct crypto_tfm *tfm,
 {
 	struct qat_alg_session_ctx *ctx = crypto_tfm_ctx(tfm);
 
-	memset(ctx, '\0', sizeof(*ctx));
+	memzero_explicit(ctx, sizeof(*ctx));
 	ctx->hash_tfm = crypto_alloc_shash(hash_name, 0, 0);
 	if (IS_ERR(ctx->hash_tfm))
 		return -EFAULT;
@@ -918,22 +879,16 @@ static void qat_alg_exit(struct crypto_tfm *tfm)
 		return;
 
 	dev = &GET_DEV(inst->accel_dev);
-	if (ctx->enc_cd)
+	if (ctx->enc_cd) {
+		memzero_explicit(ctx->enc_cd, sizeof(struct qat_alg_cd));
 		dma_free_coherent(dev, sizeof(struct qat_alg_cd),
 				  ctx->enc_cd, ctx->enc_cd_paddr);
-	if (ctx->dec_cd)
+	}
+	if (ctx->dec_cd) {
+		memzero_explicit(ctx->dec_cd, sizeof(struct qat_alg_cd));
 		dma_free_coherent(dev, sizeof(struct qat_alg_cd),
 				  ctx->dec_cd, ctx->dec_cd_paddr);
-	if (ctx->auth_hw_state_enc)
-		dma_free_coherent(dev, sizeof(struct qat_auth_state),
-				  ctx->auth_hw_state_enc,
-				  ctx->auth_state_enc_paddr);
-
-	if (ctx->auth_hw_state_dec)
-		dma_free_coherent(dev, sizeof(struct qat_auth_state),
-				  ctx->auth_hw_state_dec,
-				  ctx->auth_state_dec_paddr);
-
+	}
 	qat_crypto_put_instance(inst);
 }
 
diff --git a/drivers/crypto/qat/qat_common/qat_crypto.c b/drivers/crypto/qat/qat_common/qat_crypto.c
index 0d59bcb50de1..828f2a686aab 100644
--- a/drivers/crypto/qat/qat_common/qat_crypto.c
+++ b/drivers/crypto/qat/qat_common/qat_crypto.c
@@ -109,12 +109,14 @@ struct qat_crypto_instance *qat_crypto_get_instance_node(int node)
 
 	list_for_each(itr, adf_devmgr_get_head()) {
 		accel_dev = list_entry(itr, struct adf_accel_dev, list);
-		if (accel_dev->numa_node == node && adf_dev_started(accel_dev))
+		if ((node == dev_to_node(&GET_DEV(accel_dev)) ||
+			dev_to_node(&GET_DEV(accel_dev)) < 0)
+				&& adf_dev_started(accel_dev))
 			break;
 		accel_dev = NULL;
 	}
 	if (!accel_dev) {
-		pr_err("QAT: Could not find device on give node\n");
+		pr_err("QAT: Could not find device on node %d\n", node);
 		accel_dev = adf_devmgr_get_first();
 	}
 	if (!accel_dev || !adf_dev_started(accel_dev))
@@ -164,7 +166,7 @@ static int qat_crypto_create_instances(struct adf_accel_dev *accel_dev)
 
 	for (i = 0; i < num_inst; i++) {
 		inst = kzalloc_node(sizeof(*inst), GFP_KERNEL,
-				    accel_dev->numa_node);
+				    dev_to_node(&GET_DEV(accel_dev)));
 		if (!inst)
 			goto err;
 
diff --git a/drivers/crypto/qat/qat_common/qat_hal.c b/drivers/crypto/qat/qat_common/qat_hal.c
index 9b8a31521ff3..b818c19713bf 100644
--- a/drivers/crypto/qat/qat_common/qat_hal.c
+++ b/drivers/crypto/qat/qat_common/qat_hal.c
@@ -679,7 +679,8 @@ int qat_hal_init(struct adf_accel_dev *accel_dev)
 	struct icp_qat_fw_loader_handle *handle;
 	struct adf_accel_pci *pci_info = &accel_dev->accel_pci_dev;
 	struct adf_hw_device_data *hw_data = accel_dev->hw_device;
-	struct adf_bar *bar = &pci_info->pci_bars[ADF_DH895XCC_PMISC_BAR];
+	struct adf_bar *bar =
+			&pci_info->pci_bars[hw_data->get_misc_bar_id(hw_data)];
 
 	handle = kzalloc(sizeof(*handle), GFP_KERNEL);
 	if (!handle)
diff --git a/drivers/crypto/qat/qat_dh895xcc/adf_admin.c b/drivers/crypto/qat/qat_dh895xcc/adf_admin.c
index 978d6c56639d..53c491b59f07 100644
--- a/drivers/crypto/qat/qat_dh895xcc/adf_admin.c
+++ b/drivers/crypto/qat/qat_dh895xcc/adf_admin.c
@@ -108,7 +108,7 @@ int adf_init_admin_comms(struct adf_accel_dev *accel_dev)
 	uint64_t reg_val;
 
 	admin = kzalloc_node(sizeof(*accel_dev->admin), GFP_KERNEL,
-			     accel_dev->numa_node);
+			     dev_to_node(&GET_DEV(accel_dev)));
 	if (!admin)
 		return -ENOMEM;
 	admin->virt_addr = dma_zalloc_coherent(&GET_DEV(accel_dev), PAGE_SIZE,
diff --git a/drivers/crypto/qat/qat_dh895xcc/adf_dh895xcc_hw_data.h b/drivers/crypto/qat/qat_dh895xcc/adf_dh895xcc_hw_data.h
index 65dd1ff93d3b..01e0be21e93a 100644
--- a/drivers/crypto/qat/qat_dh895xcc/adf_dh895xcc_hw_data.h
+++ b/drivers/crypto/qat/qat_dh895xcc/adf_dh895xcc_hw_data.h
@@ -48,6 +48,8 @@
 #define ADF_DH895x_HW_DATA_H_
 
 /* PCIe configuration space */
+#define ADF_DH895XCC_PMISC_BAR 1
+#define ADF_DH895XCC_ETR_BAR 2
 #define ADF_DH895XCC_RX_RINGS_OFFSET 8
 #define ADF_DH895XCC_TX_RINGS_MASK 0xFF
 #define ADF_DH895XCC_FUSECTL_OFFSET 0x40
diff --git a/drivers/crypto/qat/qat_dh895xcc/adf_drv.c b/drivers/crypto/qat/qat_dh895xcc/adf_drv.c
index 0d0435a41be9..948f66be262b 100644
--- a/drivers/crypto/qat/qat_dh895xcc/adf_drv.c
+++ b/drivers/crypto/qat/qat_dh895xcc/adf_drv.c
@@ -119,21 +119,6 @@ static void adf_cleanup_accel(struct adf_accel_dev *accel_dev)
 	kfree(accel_dev);
 }
 
-static uint8_t adf_get_dev_node_id(struct pci_dev *pdev)
-{
-	unsigned int bus_per_cpu = 0;
-	struct cpuinfo_x86 *c = &cpu_data(num_online_cpus() - 1);
-
-	if (!c->phys_proc_id)
-		return 0;
-
-	bus_per_cpu = 256 / (c->phys_proc_id + 1);
-
-	if (bus_per_cpu != 0)
-		return pdev->bus->number / bus_per_cpu;
-	return 0;
-}
-
 static int qat_dev_start(struct adf_accel_dev *accel_dev)
 {
 	int cpus = num_online_cpus();
@@ -235,7 +220,6 @@ static int adf_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	void __iomem *pmisc_bar_addr = NULL;
 	char name[ADF_DEVICE_NAME_LENGTH];
 	unsigned int i, bar_nr;
-	uint8_t node;
 	int ret;
 
 	switch (ent->device) {
@@ -246,12 +230,19 @@ static int adf_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 		return -ENODEV;
 	}
 
-	node = adf_get_dev_node_id(pdev);
-	accel_dev = kzalloc_node(sizeof(*accel_dev), GFP_KERNEL, node);
+	if (num_possible_nodes() > 1 && dev_to_node(&pdev->dev) < 0) {
+		/* If the accelerator is connected to a node with no memory
+		 * there is no point in using the accelerator since the remote
+		 * memory transaction will be very slow. */
+		dev_err(&pdev->dev, "Invalid NUMA configuration.\n");
+		return -EINVAL;
+	}
+
+	accel_dev = kzalloc_node(sizeof(*accel_dev), GFP_KERNEL,
+			         dev_to_node(&pdev->dev));
 	if (!accel_dev)
 		return -ENOMEM;
 
-	accel_dev->numa_node = node;
 	INIT_LIST_HEAD(&accel_dev->crypto_list);
 
 	/* Add accel device to accel table.
@@ -264,7 +255,8 @@ static int adf_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 
 	accel_dev->owner = THIS_MODULE;
 	/* Allocate and configure device configuration structure */
-	hw_data = kzalloc_node(sizeof(*hw_data), GFP_KERNEL, node);
+	hw_data = kzalloc_node(sizeof(*hw_data), GFP_KERNEL,
+			       dev_to_node(&pdev->dev));
 	if (!hw_data) {
 		ret = -ENOMEM;
 		goto out_err;
diff --git a/drivers/crypto/qat/qat_dh895xcc/adf_isr.c b/drivers/crypto/qat/qat_dh895xcc/adf_isr.c
index d4172dedf775..fe8f89697ad8 100644
--- a/drivers/crypto/qat/qat_dh895xcc/adf_isr.c
+++ b/drivers/crypto/qat/qat_dh895xcc/adf_isr.c
@@ -70,9 +70,9 @@ static int adf_enable_msix(struct adf_accel_dev *accel_dev)
 	for (i = 0; i < msix_num_entries; i++)
 		pci_dev_info->msix_entries.entries[i].entry = i;
 
-	if (pci_enable_msix(pci_dev_info->pci_dev,
-			    pci_dev_info->msix_entries.entries,
-			    msix_num_entries)) {
+	if (pci_enable_msix_exact(pci_dev_info->pci_dev,
+				  pci_dev_info->msix_entries.entries,
+				  msix_num_entries)) {
 		pr_err("QAT: Failed to enable MSIX IRQ\n");
 		return -EFAULT;
 	}
@@ -89,7 +89,7 @@ static irqreturn_t adf_msix_isr_bundle(int irq, void *bank_ptr)
 	struct adf_etr_bank_data *bank = bank_ptr;
 
 	WRITE_CSR_INT_FLAG_AND_COL(bank->csr_addr, bank->bank_number, 0);
-	tasklet_hi_schedule(&bank->resp_hanlder);
+	tasklet_hi_schedule(&bank->resp_handler);
 	return IRQ_HANDLED;
 }
 
@@ -168,7 +168,7 @@ static int adf_isr_alloc_msix_entry_table(struct adf_accel_dev *accel_dev)
 	uint32_t msix_num_entries = hw_data->num_banks + 1;
 
 	entries = kzalloc_node(msix_num_entries * sizeof(*entries),
-			       GFP_KERNEL, accel_dev->numa_node);
+			       GFP_KERNEL, dev_to_node(&GET_DEV(accel_dev)));
 	if (!entries)
 		return -ENOMEM;
 
@@ -186,10 +186,8 @@ static int adf_isr_alloc_msix_entry_table(struct adf_accel_dev *accel_dev)
 	accel_dev->accel_pci_dev.msix_entries.names = names;
 	return 0;
 err:
-	for (i = 0; i < msix_num_entries; i++) {
-		if (*(names + i))
-			kfree(*(names + i));
-	}
+	for (i = 0; i < msix_num_entries; i++)
+		kfree(*(names + i));
 	kfree(entries);
 	kfree(names);
 	return -ENOMEM;
@@ -203,10 +201,8 @@ static void adf_isr_free_msix_entry_table(struct adf_accel_dev *accel_dev)
 	int i;
 
 	kfree(accel_dev->accel_pci_dev.msix_entries.entries);
-	for (i = 0; i < msix_num_entries; i++) {
-		if (*(names + i))
-			kfree(*(names + i));
-	}
+	for (i = 0; i < msix_num_entries; i++)
+		kfree(*(names + i));
 	kfree(names);
 }
 
@@ -217,7 +213,7 @@ static int adf_setup_bh(struct adf_accel_dev *accel_dev)
 	int i;
 
 	for (i = 0; i < hw_data->num_banks; i++)
-		tasklet_init(&priv_data->banks[i].resp_hanlder,
+		tasklet_init(&priv_data->banks[i].resp_handler,
 			     adf_response_handler,
 			     (unsigned long)&priv_data->banks[i]);
 	return 0;
@@ -230,8 +226,8 @@ static void adf_cleanup_bh(struct adf_accel_dev *accel_dev)
 	int i;
 
 	for (i = 0; i < hw_data->num_banks; i++) {
-		tasklet_disable(&priv_data->banks[i].resp_hanlder);
-		tasklet_kill(&priv_data->banks[i].resp_hanlder);
+		tasklet_disable(&priv_data->banks[i].resp_handler);
+		tasklet_kill(&priv_data->banks[i].resp_handler);
 	}
 }
 
diff --git a/drivers/crypto/qce/core.c b/drivers/crypto/qce/core.c
index 33ae3545dc48..718b32a3112e 100644
--- a/drivers/crypto/qce/core.c
+++ b/drivers/crypto/qce/core.c
@@ -273,7 +273,6 @@ static struct platform_driver qce_crypto_driver = {
 	.probe = qce_crypto_probe,
 	.remove = qce_crypto_remove,
 	.driver = {
-		.owner = THIS_MODULE,
 		.name = KBUILD_MODNAME,
 		.of_match_table = qce_crypto_of_match,
 	},
diff --git a/drivers/crypto/qce/dma.h b/drivers/crypto/qce/dma.h
index 805e378d59e9..65bedb81de0b 100644
--- a/drivers/crypto/qce/dma.h
+++ b/drivers/crypto/qce/dma.h
@@ -14,6 +14,8 @@
 #ifndef _DMA_H_
 #define _DMA_H_
 
+#include <linux/dmaengine.h>
+
 /* maximum data transfer block size between BAM and CE */
 #define QCE_BAM_BURST_SIZE		64
 
diff --git a/drivers/crypto/s5p-sss.c b/drivers/crypto/s5p-sss.c
index 4197ad9a711b..f214a8755827 100644
--- a/drivers/crypto/s5p-sss.c
+++ b/drivers/crypto/s5p-sss.c
@@ -745,7 +745,6 @@ static struct platform_driver s5p_aes_crypto = {
 	.probe	= s5p_aes_probe,
 	.remove	= s5p_aes_remove,
 	.driver	= {
-		.owner	= THIS_MODULE,
 		.name	= "s5p-secss",
 		.of_match_table = s5p_sss_dt_match,
 	},
diff --git a/drivers/crypto/sahara.c b/drivers/crypto/sahara.c
index 164e1ec624e3..220b92f7eabc 100644
--- a/drivers/crypto/sahara.c
+++ b/drivers/crypto/sahara.c
@@ -3,6 +3,7 @@
  *
  * Support for SAHARA cryptographic accelerator.
  *
+ * Copyright (c) 2014 Steffen Trumtrar <s.trumtrar@pengutronix.de>
  * Copyright (c) 2013 Vista Silicon S.L.
  * Author: Javier Martin <javier.martin@vista-silicon.com>
  *
@@ -15,6 +16,10 @@
 
 #include <crypto/algapi.h>
 #include <crypto/aes.h>
+#include <crypto/hash.h>
+#include <crypto/internal/hash.h>
+#include <crypto/scatterwalk.h>
+#include <crypto/sha.h>
 
 #include <linux/clk.h>
 #include <linux/crypto.h>
@@ -22,12 +27,19 @@
 #include <linux/io.h>
 #include <linux/irq.h>
 #include <linux/kernel.h>
+#include <linux/kthread.h>
 #include <linux/module.h>
+#include <linux/mutex.h>
 #include <linux/of.h>
+#include <linux/of_device.h>
 #include <linux/platform_device.h>
 
+#define SHA_BUFFER_LEN		PAGE_SIZE
+#define SAHARA_MAX_SHA_BLOCK_SIZE	SHA256_BLOCK_SIZE
+
 #define SAHARA_NAME "sahara"
 #define SAHARA_VERSION_3	3
+#define SAHARA_VERSION_4	4
 #define SAHARA_TIMEOUT_MS	1000
 #define SAHARA_MAX_HW_DESC	2
 #define SAHARA_MAX_HW_LINK	20
@@ -36,7 +48,6 @@
 #define FLAGS_ENCRYPT		BIT(0)
 #define FLAGS_CBC		BIT(1)
 #define FLAGS_NEW_KEY		BIT(3)
-#define FLAGS_BUSY		4
 
 #define SAHARA_HDR_BASE			0x00800000
 #define SAHARA_HDR_SKHA_ALG_AES	0
@@ -50,6 +61,23 @@
 #define SAHARA_HDR_CHA_MDHA		(2 << 28)
 #define SAHARA_HDR_PARITY_BIT		(1 << 31)
 
+#define SAHARA_HDR_MDHA_SET_MODE_MD_KEY	0x20880000
+#define SAHARA_HDR_MDHA_SET_MODE_HASH	0x208D0000
+#define SAHARA_HDR_MDHA_HASH		0xA0850000
+#define SAHARA_HDR_MDHA_STORE_DIGEST	0x20820000
+#define SAHARA_HDR_MDHA_ALG_SHA1	0
+#define SAHARA_HDR_MDHA_ALG_MD5		1
+#define SAHARA_HDR_MDHA_ALG_SHA256	2
+#define SAHARA_HDR_MDHA_ALG_SHA224	3
+#define SAHARA_HDR_MDHA_PDATA		(1 << 2)
+#define SAHARA_HDR_MDHA_HMAC		(1 << 3)
+#define SAHARA_HDR_MDHA_INIT		(1 << 5)
+#define SAHARA_HDR_MDHA_IPAD		(1 << 6)
+#define SAHARA_HDR_MDHA_OPAD		(1 << 7)
+#define SAHARA_HDR_MDHA_SWAP		(1 << 8)
+#define SAHARA_HDR_MDHA_MAC_FULL	(1 << 9)
+#define SAHARA_HDR_MDHA_SSL		(1 << 10)
+
 /* SAHARA can only process one request at a time */
 #define SAHARA_QUEUE_LENGTH	1
 
@@ -117,31 +145,74 @@ struct sahara_hw_link {
 };
 
 struct sahara_ctx {
-	struct sahara_dev *dev;
 	unsigned long flags;
+
+	/* AES-specific context */
 	int keylen;
 	u8 key[AES_KEYSIZE_128];
 	struct crypto_ablkcipher *fallback;
+
+	/* SHA-specific context */
+	struct crypto_shash *shash_fallback;
 };
 
 struct sahara_aes_reqctx {
 	unsigned long mode;
 };
 
+/*
+ * struct sahara_sha_reqctx - private data per request
+ * @buf: holds data for requests smaller than block_size
+ * @rembuf: used to prepare one block_size-aligned request
+ * @context: hw-specific context for request. Digest is extracted from this
+ * @mode: specifies what type of hw-descriptor needs to be built
+ * @digest_size: length of digest for this request
+ * @context_size: length of hw-context for this request.
+ *                Always digest_size + 4
+ * @buf_cnt: number of bytes saved in buf
+ * @sg_in_idx: number of hw links
+ * @in_sg: scatterlist for input data
+ * @in_sg_chain: scatterlists for chained input data
+ * @in_sg_chained: specifies if chained scatterlists are used or not
+ * @total: total number of bytes for transfer
+ * @last: is this the last block
+ * @first: is this the first block
+ * @active: inside a transfer
+ */
+struct sahara_sha_reqctx {
+	u8			buf[SAHARA_MAX_SHA_BLOCK_SIZE];
+	u8			rembuf[SAHARA_MAX_SHA_BLOCK_SIZE];
+	u8			context[SHA256_DIGEST_SIZE + 4];
+	struct mutex		mutex;
+	unsigned int		mode;
+	unsigned int		digest_size;
+	unsigned int		context_size;
+	unsigned int		buf_cnt;
+	unsigned int		sg_in_idx;
+	struct scatterlist	*in_sg;
+	struct scatterlist	in_sg_chain[2];
+	bool			in_sg_chained;
+	size_t			total;
+	unsigned int		last;
+	unsigned int		first;
+	unsigned int		active;
+};
+
 struct sahara_dev {
 	struct device		*device;
+	unsigned int		version;
 	void __iomem		*regs_base;
 	struct clk		*clk_ipg;
 	struct clk		*clk_ahb;
+	struct mutex		queue_mutex;
+	struct task_struct	*kthread;
+	struct completion	dma_completion;
 
 	struct sahara_ctx	*ctx;
 	spinlock_t		lock;
 	struct crypto_queue	queue;
 	unsigned long		flags;
 
-	struct tasklet_struct	done_task;
-	struct tasklet_struct	queue_task;
-
 	struct sahara_hw_desc	*hw_desc[SAHARA_MAX_HW_DESC];
 	dma_addr_t		hw_phys_desc[SAHARA_MAX_HW_DESC];
 
@@ -151,10 +222,12 @@ struct sahara_dev {
 	u8			*iv_base;
 	dma_addr_t		iv_phys_base;
 
+	u8			*context_base;
+	dma_addr_t		context_phys_base;
+
 	struct sahara_hw_link	*hw_link[SAHARA_MAX_HW_LINK];
 	dma_addr_t		hw_phys_link[SAHARA_MAX_HW_LINK];
 
-	struct ablkcipher_request *req;
 	size_t			total;
 	struct scatterlist	*in_sg;
 	unsigned int		nb_in_sg;
@@ -162,7 +235,6 @@ struct sahara_dev {
 	unsigned int		nb_out_sg;
 
 	u32			error;
-	struct timer_list	watchdog;
 };
 
 static struct sahara_dev *dev_ptr;
@@ -401,34 +473,6 @@ static void sahara_dump_links(struct sahara_dev *dev)
 	dev_dbg(dev->device, "\n");
 }
 
-static void sahara_aes_done_task(unsigned long data)
-{
-	struct sahara_dev *dev = (struct sahara_dev *)data;
-
-	dma_unmap_sg(dev->device, dev->out_sg, dev->nb_out_sg,
-		DMA_TO_DEVICE);
-	dma_unmap_sg(dev->device, dev->in_sg, dev->nb_in_sg,
-		DMA_FROM_DEVICE);
-
-	spin_lock(&dev->lock);
-	clear_bit(FLAGS_BUSY, &dev->flags);
-	spin_unlock(&dev->lock);
-
-	dev->req->base.complete(&dev->req->base, dev->error);
-}
-
-static void sahara_watchdog(unsigned long data)
-{
-	struct sahara_dev *dev = (struct sahara_dev *)data;
-	unsigned int err = sahara_read(dev, SAHARA_REG_ERRSTATUS);
-	unsigned int stat = sahara_read(dev, SAHARA_REG_STATUS);
-
-	sahara_decode_status(dev, stat);
-	sahara_decode_error(dev, err);
-	dev->error = -ETIMEDOUT;
-	sahara_aes_done_task(data);
-}
-
 static int sahara_hw_descriptor_create(struct sahara_dev *dev)
 {
 	struct sahara_ctx *ctx = dev->ctx;
@@ -512,9 +556,6 @@ static int sahara_hw_descriptor_create(struct sahara_dev *dev)
 	sahara_dump_descriptors(dev);
 	sahara_dump_links(dev);
 
-	/* Start processing descriptor chain. */
-	mod_timer(&dev->watchdog,
-		  jiffies + msecs_to_jiffies(SAHARA_TIMEOUT_MS));
 	sahara_write(dev, dev->hw_phys_desc[0], SAHARA_REG_DAR);
 
 	return 0;
@@ -529,37 +570,19 @@ unmap_in:
 	return -EINVAL;
 }
 
-static void sahara_aes_queue_task(unsigned long data)
+static int sahara_aes_process(struct ablkcipher_request *req)
 {
-	struct sahara_dev *dev = (struct sahara_dev *)data;
-	struct crypto_async_request *async_req, *backlog;
+	struct sahara_dev *dev = dev_ptr;
 	struct sahara_ctx *ctx;
 	struct sahara_aes_reqctx *rctx;
-	struct ablkcipher_request *req;
 	int ret;
 
-	spin_lock(&dev->lock);
-	backlog = crypto_get_backlog(&dev->queue);
-	async_req = crypto_dequeue_request(&dev->queue);
-	if (!async_req)
-		clear_bit(FLAGS_BUSY, &dev->flags);
-	spin_unlock(&dev->lock);
-
-	if (!async_req)
-		return;
-
-	if (backlog)
-		backlog->complete(backlog, -EINPROGRESS);
-
-	req = ablkcipher_request_cast(async_req);
-
 	/* Request is ready to be dispatched by the device */
 	dev_dbg(dev->device,
 		"dispatch request (nbytes=%d, src=%p, dst=%p)\n",
 		req->nbytes, req->src, req->dst);
 
 	/* assign new request to device */
-	dev->req = req;
 	dev->total = req->nbytes;
 	dev->in_sg = req->src;
 	dev->out_sg = req->dst;
@@ -573,16 +596,25 @@ static void sahara_aes_queue_task(unsigned long data)
 		memcpy(dev->iv_base, req->info, AES_KEYSIZE_128);
 
 	/* assign new context to device */
-	ctx->dev = dev;
 	dev->ctx = ctx;
 
+	reinit_completion(&dev->dma_completion);
+
 	ret = sahara_hw_descriptor_create(dev);
-	if (ret < 0) {
-		spin_lock(&dev->lock);
-		clear_bit(FLAGS_BUSY, &dev->flags);
-		spin_unlock(&dev->lock);
-		dev->req->base.complete(&dev->req->base, ret);
+
+	ret = wait_for_completion_timeout(&dev->dma_completion,
+				msecs_to_jiffies(SAHARA_TIMEOUT_MS));
+	if (!ret) {
+		dev_err(dev->device, "AES timeout\n");
+		return -ETIMEDOUT;
 	}
+
+	dma_unmap_sg(dev->device, dev->out_sg, dev->nb_out_sg,
+		DMA_TO_DEVICE);
+	dma_unmap_sg(dev->device, dev->in_sg, dev->nb_in_sg,
+		DMA_FROM_DEVICE);
+
+	return 0;
 }
 
 static int sahara_aes_setkey(struct crypto_ablkcipher *tfm, const u8 *key,
@@ -624,12 +656,9 @@ static int sahara_aes_setkey(struct crypto_ablkcipher *tfm, const u8 *key,
 
 static int sahara_aes_crypt(struct ablkcipher_request *req, unsigned long mode)
 {
-	struct sahara_ctx *ctx = crypto_ablkcipher_ctx(
-		crypto_ablkcipher_reqtfm(req));
 	struct sahara_aes_reqctx *rctx = ablkcipher_request_ctx(req);
 	struct sahara_dev *dev = dev_ptr;
 	int err = 0;
-	int busy;
 
 	dev_dbg(dev->device, "nbytes: %d, enc: %d, cbc: %d\n",
 		req->nbytes, !!(mode & FLAGS_ENCRYPT), !!(mode & FLAGS_CBC));
@@ -640,16 +669,13 @@ static int sahara_aes_crypt(struct ablkcipher_request *req, unsigned long mode)
 		return -EINVAL;
 	}
 
-	ctx->dev = dev;
-
 	rctx->mode = mode;
-	spin_lock_bh(&dev->lock);
+
+	mutex_lock(&dev->queue_mutex);
 	err = ablkcipher_enqueue_request(&dev->queue, req);
-	busy = test_and_set_bit(FLAGS_BUSY, &dev->flags);
-	spin_unlock_bh(&dev->lock);
+	mutex_unlock(&dev->queue_mutex);
 
-	if (!busy)
-		tasklet_schedule(&dev->queue_task);
+	wake_up_process(dev->kthread);
 
 	return err;
 }
@@ -752,6 +778,484 @@ static void sahara_aes_cra_exit(struct crypto_tfm *tfm)
 	ctx->fallback = NULL;
 }
 
+static u32 sahara_sha_init_hdr(struct sahara_dev *dev,
+			      struct sahara_sha_reqctx *rctx)
+{
+	u32 hdr = 0;
+
+	hdr = rctx->mode;
+
+	if (rctx->first) {
+		hdr |= SAHARA_HDR_MDHA_SET_MODE_HASH;
+		hdr |= SAHARA_HDR_MDHA_INIT;
+	} else {
+		hdr |= SAHARA_HDR_MDHA_SET_MODE_MD_KEY;
+	}
+
+	if (rctx->last)
+		hdr |= SAHARA_HDR_MDHA_PDATA;
+
+	if (hweight_long(hdr) % 2 == 0)
+		hdr |= SAHARA_HDR_PARITY_BIT;
+
+	return hdr;
+}
+
+static int sahara_sha_hw_links_create(struct sahara_dev *dev,
+				       struct sahara_sha_reqctx *rctx,
+				       int start)
+{
+	struct scatterlist *sg;
+	unsigned int i;
+	int ret;
+
+	dev->in_sg = rctx->in_sg;
+
+	dev->nb_in_sg = sahara_sg_length(dev->in_sg, rctx->total);
+	if ((dev->nb_in_sg) > SAHARA_MAX_HW_LINK) {
+		dev_err(dev->device, "not enough hw links (%d)\n",
+			dev->nb_in_sg + dev->nb_out_sg);
+		return -EINVAL;
+	}
+
+	if (rctx->in_sg_chained) {
+		i = start;
+		sg = dev->in_sg;
+		while (sg) {
+			ret = dma_map_sg(dev->device, sg, 1,
+					 DMA_TO_DEVICE);
+			if (!ret)
+				return -EFAULT;
+
+			dev->hw_link[i]->len = sg->length;
+			dev->hw_link[i]->p = sg->dma_address;
+			dev->hw_link[i]->next = dev->hw_phys_link[i + 1];
+			sg = sg_next(sg);
+			i += 1;
+		}
+		dev->hw_link[i-1]->next = 0;
+	} else {
+		sg = dev->in_sg;
+		ret = dma_map_sg(dev->device, dev->in_sg, dev->nb_in_sg,
+				 DMA_TO_DEVICE);
+		if (!ret)
+			return -EFAULT;
+
+		for (i = start; i < dev->nb_in_sg + start; i++) {
+			dev->hw_link[i]->len = sg->length;
+			dev->hw_link[i]->p = sg->dma_address;
+			if (i == (dev->nb_in_sg + start - 1)) {
+				dev->hw_link[i]->next = 0;
+			} else {
+				dev->hw_link[i]->next = dev->hw_phys_link[i + 1];
+				sg = sg_next(sg);
+			}
+		}
+	}
+
+	return i;
+}
+
+static int sahara_sha_hw_data_descriptor_create(struct sahara_dev *dev,
+						struct sahara_sha_reqctx *rctx,
+						struct ahash_request *req,
+						int index)
+{
+	unsigned result_len;
+	int i = index;
+
+	if (rctx->first)
+		/* Create initial descriptor: #8*/
+		dev->hw_desc[index]->hdr = sahara_sha_init_hdr(dev, rctx);
+	else
+		/* Create hash descriptor: #10. Must follow #6. */
+		dev->hw_desc[index]->hdr = SAHARA_HDR_MDHA_HASH;
+
+	dev->hw_desc[index]->len1 = rctx->total;
+	if (dev->hw_desc[index]->len1 == 0) {
+		/* if len1 is 0, p1 must be 0, too */
+		dev->hw_desc[index]->p1 = 0;
+		rctx->sg_in_idx = 0;
+	} else {
+		/* Create input links */
+		dev->hw_desc[index]->p1 = dev->hw_phys_link[index];
+		i = sahara_sha_hw_links_create(dev, rctx, index);
+
+		rctx->sg_in_idx = index;
+		if (i < 0)
+			return i;
+	}
+
+	dev->hw_desc[index]->p2 = dev->hw_phys_link[i];
+
+	/* Save the context for the next operation */
+	result_len = rctx->context_size;
+	dev->hw_link[i]->p = dev->context_phys_base;
+
+	dev->hw_link[i]->len = result_len;
+	dev->hw_desc[index]->len2 = result_len;
+
+	dev->hw_link[i]->next = 0;
+
+	return 0;
+}
+
+/*
+ * Load descriptor aka #6
+ *
+ * To load a previously saved context back to the MDHA unit
+ *
+ * p1: Saved Context
+ * p2: NULL
+ *
+ */
+static int sahara_sha_hw_context_descriptor_create(struct sahara_dev *dev,
+						struct sahara_sha_reqctx *rctx,
+						struct ahash_request *req,
+						int index)
+{
+	dev->hw_desc[index]->hdr = sahara_sha_init_hdr(dev, rctx);
+
+	dev->hw_desc[index]->len1 = rctx->context_size;
+	dev->hw_desc[index]->p1 = dev->hw_phys_link[index];
+	dev->hw_desc[index]->len2 = 0;
+	dev->hw_desc[index]->p2 = 0;
+
+	dev->hw_link[index]->len = rctx->context_size;
+	dev->hw_link[index]->p = dev->context_phys_base;
+	dev->hw_link[index]->next = 0;
+
+	return 0;
+}
+
+static int sahara_walk_and_recalc(struct scatterlist *sg, unsigned int nbytes)
+{
+	if (!sg || !sg->length)
+		return nbytes;
+
+	while (nbytes && sg) {
+		if (nbytes <= sg->length) {
+			sg->length = nbytes;
+			sg_mark_end(sg);
+			break;
+		}
+		nbytes -= sg->length;
+		sg = scatterwalk_sg_next(sg);
+	}
+
+	return nbytes;
+}
+
+static int sahara_sha_prepare_request(struct ahash_request *req)
+{
+	struct crypto_ahash *tfm = crypto_ahash_reqtfm(req);
+	struct sahara_sha_reqctx *rctx = ahash_request_ctx(req);
+	unsigned int hash_later;
+	unsigned int block_size;
+	unsigned int len;
+
+	block_size = crypto_tfm_alg_blocksize(crypto_ahash_tfm(tfm));
+
+	/* append bytes from previous operation */
+	len = rctx->buf_cnt + req->nbytes;
+
+	/* only the last transfer can be padded in hardware */
+	if (!rctx->last && (len < block_size)) {
+		/* to few data, save for next operation */
+		scatterwalk_map_and_copy(rctx->buf + rctx->buf_cnt, req->src,
+					 0, req->nbytes, 0);
+		rctx->buf_cnt += req->nbytes;
+
+		return 0;
+	}
+
+	/* add data from previous operation first */
+	if (rctx->buf_cnt)
+		memcpy(rctx->rembuf, rctx->buf, rctx->buf_cnt);
+
+	/* data must always be a multiple of block_size */
+	hash_later = rctx->last ? 0 : len & (block_size - 1);
+	if (hash_later) {
+		unsigned int offset = req->nbytes - hash_later;
+		/* Save remaining bytes for later use */
+		scatterwalk_map_and_copy(rctx->buf, req->src, offset,
+					hash_later, 0);
+	}
+
+	/* nbytes should now be multiple of blocksize */
+	req->nbytes = req->nbytes - hash_later;
+
+	sahara_walk_and_recalc(req->src, req->nbytes);
+
+	/* have data from previous operation and current */
+	if (rctx->buf_cnt && req->nbytes) {
+		sg_init_table(rctx->in_sg_chain, 2);
+		sg_set_buf(rctx->in_sg_chain, rctx->rembuf, rctx->buf_cnt);
+
+		scatterwalk_sg_chain(rctx->in_sg_chain, 2, req->src);
+
+		rctx->total = req->nbytes + rctx->buf_cnt;
+		rctx->in_sg = rctx->in_sg_chain;
+
+		rctx->in_sg_chained = true;
+		req->src = rctx->in_sg_chain;
+	/* only data from previous operation */
+	} else if (rctx->buf_cnt) {
+		if (req->src)
+			rctx->in_sg = req->src;
+		else
+			rctx->in_sg = rctx->in_sg_chain;
+		/* buf was copied into rembuf above */
+		sg_init_one(rctx->in_sg, rctx->rembuf, rctx->buf_cnt);
+		rctx->total = rctx->buf_cnt;
+		rctx->in_sg_chained = false;
+	/* no data from previous operation */
+	} else {
+		rctx->in_sg = req->src;
+		rctx->total = req->nbytes;
+		req->src = rctx->in_sg;
+		rctx->in_sg_chained = false;
+	}
+
+	/* on next call, we only have the remaining data in the buffer */
+	rctx->buf_cnt = hash_later;
+
+	return -EINPROGRESS;
+}
+
+static void sahara_sha_unmap_sg(struct sahara_dev *dev,
+				struct sahara_sha_reqctx *rctx)
+{
+	struct scatterlist *sg;
+
+	if (rctx->in_sg_chained) {
+		sg = dev->in_sg;
+		while (sg) {
+			dma_unmap_sg(dev->device, sg, 1, DMA_TO_DEVICE);
+			sg = sg_next(sg);
+		}
+	} else {
+		dma_unmap_sg(dev->device, dev->in_sg, dev->nb_in_sg,
+			DMA_TO_DEVICE);
+	}
+}
+
+static int sahara_sha_process(struct ahash_request *req)
+{
+	struct sahara_dev *dev = dev_ptr;
+	struct sahara_sha_reqctx *rctx = ahash_request_ctx(req);
+	int ret = -EINPROGRESS;
+
+	ret = sahara_sha_prepare_request(req);
+	if (!ret)
+		return ret;
+
+	if (rctx->first) {
+		sahara_sha_hw_data_descriptor_create(dev, rctx, req, 0);
+		dev->hw_desc[0]->next = 0;
+		rctx->first = 0;
+	} else {
+		memcpy(dev->context_base, rctx->context, rctx->context_size);
+
+		sahara_sha_hw_context_descriptor_create(dev, rctx, req, 0);
+		dev->hw_desc[0]->next = dev->hw_phys_desc[1];
+		sahara_sha_hw_data_descriptor_create(dev, rctx, req, 1);
+		dev->hw_desc[1]->next = 0;
+	}
+
+	sahara_dump_descriptors(dev);
+	sahara_dump_links(dev);
+
+	reinit_completion(&dev->dma_completion);
+
+	sahara_write(dev, dev->hw_phys_desc[0], SAHARA_REG_DAR);
+
+	ret = wait_for_completion_timeout(&dev->dma_completion,
+				msecs_to_jiffies(SAHARA_TIMEOUT_MS));
+	if (!ret) {
+		dev_err(dev->device, "SHA timeout\n");
+		return -ETIMEDOUT;
+	}
+
+	if (rctx->sg_in_idx)
+		sahara_sha_unmap_sg(dev, rctx);
+
+	memcpy(rctx->context, dev->context_base, rctx->context_size);
+
+	if (req->result)
+		memcpy(req->result, rctx->context, rctx->digest_size);
+
+	return 0;
+}
+
+static int sahara_queue_manage(void *data)
+{
+	struct sahara_dev *dev = (struct sahara_dev *)data;
+	struct crypto_async_request *async_req;
+	int ret = 0;
+
+	do {
+		__set_current_state(TASK_INTERRUPTIBLE);
+
+		mutex_lock(&dev->queue_mutex);
+		async_req = crypto_dequeue_request(&dev->queue);
+		mutex_unlock(&dev->queue_mutex);
+
+		if (async_req) {
+			if (crypto_tfm_alg_type(async_req->tfm) ==
+			    CRYPTO_ALG_TYPE_AHASH) {
+				struct ahash_request *req =
+					ahash_request_cast(async_req);
+
+				ret = sahara_sha_process(req);
+			} else {
+				struct ablkcipher_request *req =
+					ablkcipher_request_cast(async_req);
+
+				ret = sahara_aes_process(req);
+			}
+
+			async_req->complete(async_req, ret);
+
+			continue;
+		}
+
+		schedule();
+	} while (!kthread_should_stop());
+
+	return 0;
+}
+
+static int sahara_sha_enqueue(struct ahash_request *req, int last)
+{
+	struct sahara_sha_reqctx *rctx = ahash_request_ctx(req);
+	struct sahara_dev *dev = dev_ptr;
+	int ret;
+
+	if (!req->nbytes && !last)
+		return 0;
+
+	mutex_lock(&rctx->mutex);
+	rctx->last = last;
+
+	if (!rctx->active) {
+		rctx->active = 1;
+		rctx->first = 1;
+	}
+
+	mutex_lock(&dev->queue_mutex);
+	ret = crypto_enqueue_request(&dev->queue, &req->base);
+	mutex_unlock(&dev->queue_mutex);
+
+	wake_up_process(dev->kthread);
+	mutex_unlock(&rctx->mutex);
+
+	return ret;
+}
+
+static int sahara_sha_init(struct ahash_request *req)
+{
+	struct crypto_ahash *tfm = crypto_ahash_reqtfm(req);
+	struct sahara_sha_reqctx *rctx = ahash_request_ctx(req);
+
+	memset(rctx, 0, sizeof(*rctx));
+
+	switch (crypto_ahash_digestsize(tfm)) {
+	case SHA1_DIGEST_SIZE:
+		rctx->mode |= SAHARA_HDR_MDHA_ALG_SHA1;
+		rctx->digest_size = SHA1_DIGEST_SIZE;
+		break;
+	case SHA256_DIGEST_SIZE:
+		rctx->mode |= SAHARA_HDR_MDHA_ALG_SHA256;
+		rctx->digest_size = SHA256_DIGEST_SIZE;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	rctx->context_size = rctx->digest_size + 4;
+	rctx->active = 0;
+
+	mutex_init(&rctx->mutex);
+
+	return 0;
+}
+
+static int sahara_sha_update(struct ahash_request *req)
+{
+	return sahara_sha_enqueue(req, 0);
+}
+
+static int sahara_sha_final(struct ahash_request *req)
+{
+	req->nbytes = 0;
+	return sahara_sha_enqueue(req, 1);
+}
+
+static int sahara_sha_finup(struct ahash_request *req)
+{
+	return sahara_sha_enqueue(req, 1);
+}
+
+static int sahara_sha_digest(struct ahash_request *req)
+{
+	sahara_sha_init(req);
+
+	return sahara_sha_finup(req);
+}
+
+static int sahara_sha_export(struct ahash_request *req, void *out)
+{
+	struct crypto_ahash *ahash = crypto_ahash_reqtfm(req);
+	struct sahara_ctx *ctx = crypto_ahash_ctx(ahash);
+	struct sahara_sha_reqctx *rctx = ahash_request_ctx(req);
+
+	memcpy(out, ctx, sizeof(struct sahara_ctx));
+	memcpy(out + sizeof(struct sahara_sha_reqctx), rctx,
+	       sizeof(struct sahara_sha_reqctx));
+
+	return 0;
+}
+
+static int sahara_sha_import(struct ahash_request *req, const void *in)
+{
+	struct crypto_ahash *ahash = crypto_ahash_reqtfm(req);
+	struct sahara_ctx *ctx = crypto_ahash_ctx(ahash);
+	struct sahara_sha_reqctx *rctx = ahash_request_ctx(req);
+
+	memcpy(ctx, in, sizeof(struct sahara_ctx));
+	memcpy(rctx, in + sizeof(struct sahara_sha_reqctx),
+	       sizeof(struct sahara_sha_reqctx));
+
+	return 0;
+}
+
+static int sahara_sha_cra_init(struct crypto_tfm *tfm)
+{
+	const char *name = crypto_tfm_alg_name(tfm);
+	struct sahara_ctx *ctx = crypto_tfm_ctx(tfm);
+
+	ctx->shash_fallback = crypto_alloc_shash(name, 0,
+					CRYPTO_ALG_NEED_FALLBACK);
+	if (IS_ERR(ctx->shash_fallback)) {
+		pr_err("Error allocating fallback algo %s\n", name);
+		return PTR_ERR(ctx->shash_fallback);
+	}
+	crypto_ahash_set_reqsize(__crypto_ahash_cast(tfm),
+				 sizeof(struct sahara_sha_reqctx) +
+				 SHA_BUFFER_LEN + SHA256_BLOCK_SIZE);
+
+	return 0;
+}
+
+static void sahara_sha_cra_exit(struct crypto_tfm *tfm)
+{
+	struct sahara_ctx *ctx = crypto_tfm_ctx(tfm);
+
+	crypto_free_shash(ctx->shash_fallback);
+	ctx->shash_fallback = NULL;
+}
+
 static struct crypto_alg aes_algs[] = {
 {
 	.cra_name		= "ecb(aes)",
@@ -797,14 +1301,66 @@ static struct crypto_alg aes_algs[] = {
 }
 };
 
+static struct ahash_alg sha_v3_algs[] = {
+{
+	.init		= sahara_sha_init,
+	.update		= sahara_sha_update,
+	.final		= sahara_sha_final,
+	.finup		= sahara_sha_finup,
+	.digest		= sahara_sha_digest,
+	.export		= sahara_sha_export,
+	.import		= sahara_sha_import,
+	.halg.digestsize	= SHA1_DIGEST_SIZE,
+	.halg.base	= {
+		.cra_name		= "sha1",
+		.cra_driver_name	= "sahara-sha1",
+		.cra_priority		= 300,
+		.cra_flags		= CRYPTO_ALG_TYPE_AHASH |
+						CRYPTO_ALG_ASYNC |
+						CRYPTO_ALG_NEED_FALLBACK,
+		.cra_blocksize		= SHA1_BLOCK_SIZE,
+		.cra_ctxsize		= sizeof(struct sahara_ctx),
+		.cra_alignmask		= 0,
+		.cra_module		= THIS_MODULE,
+		.cra_init		= sahara_sha_cra_init,
+		.cra_exit		= sahara_sha_cra_exit,
+	}
+},
+};
+
+static struct ahash_alg sha_v4_algs[] = {
+{
+	.init		= sahara_sha_init,
+	.update		= sahara_sha_update,
+	.final		= sahara_sha_final,
+	.finup		= sahara_sha_finup,
+	.digest		= sahara_sha_digest,
+	.export		= sahara_sha_export,
+	.import		= sahara_sha_import,
+	.halg.digestsize	= SHA256_DIGEST_SIZE,
+	.halg.base	= {
+		.cra_name		= "sha256",
+		.cra_driver_name	= "sahara-sha256",
+		.cra_priority		= 300,
+		.cra_flags		= CRYPTO_ALG_TYPE_AHASH |
+						CRYPTO_ALG_ASYNC |
+						CRYPTO_ALG_NEED_FALLBACK,
+		.cra_blocksize		= SHA256_BLOCK_SIZE,
+		.cra_ctxsize		= sizeof(struct sahara_ctx),
+		.cra_alignmask		= 0,
+		.cra_module		= THIS_MODULE,
+		.cra_init		= sahara_sha_cra_init,
+		.cra_exit		= sahara_sha_cra_exit,
+	}
+},
+};
+
 static irqreturn_t sahara_irq_handler(int irq, void *data)
 {
 	struct sahara_dev *dev = (struct sahara_dev *)data;
 	unsigned int stat = sahara_read(dev, SAHARA_REG_STATUS);
 	unsigned int err = sahara_read(dev, SAHARA_REG_ERRSTATUS);
 
-	del_timer(&dev->watchdog);
-
 	sahara_write(dev, SAHARA_CMD_CLEAR_INT | SAHARA_CMD_CLEAR_ERR,
 		     SAHARA_REG_CMD);
 
@@ -819,7 +1375,7 @@ static irqreturn_t sahara_irq_handler(int irq, void *data)
 		dev->error = -EINVAL;
 	}
 
-	tasklet_schedule(&dev->done_task);
+	complete(&dev->dma_completion);
 
 	return IRQ_HANDLED;
 }
@@ -827,7 +1383,8 @@ static irqreturn_t sahara_irq_handler(int irq, void *data)
 
 static int sahara_register_algs(struct sahara_dev *dev)
 {
-	int err, i, j;
+	int err;
+	unsigned int i, j, k, l;
 
 	for (i = 0; i < ARRAY_SIZE(aes_algs); i++) {
 		INIT_LIST_HEAD(&aes_algs[i].cra_list);
@@ -836,8 +1393,29 @@ static int sahara_register_algs(struct sahara_dev *dev)
 			goto err_aes_algs;
 	}
 
+	for (k = 0; k < ARRAY_SIZE(sha_v3_algs); k++) {
+		err = crypto_register_ahash(&sha_v3_algs[k]);
+		if (err)
+			goto err_sha_v3_algs;
+	}
+
+	if (dev->version > SAHARA_VERSION_3)
+		for (l = 0; l < ARRAY_SIZE(sha_v4_algs); l++) {
+			err = crypto_register_ahash(&sha_v4_algs[l]);
+			if (err)
+				goto err_sha_v4_algs;
+		}
+
 	return 0;
 
+err_sha_v4_algs:
+	for (j = 0; j < l; j++)
+		crypto_unregister_ahash(&sha_v4_algs[j]);
+
+err_sha_v3_algs:
+	for (j = 0; j < k; j++)
+		crypto_unregister_ahash(&sha_v4_algs[j]);
+
 err_aes_algs:
 	for (j = 0; j < i; j++)
 		crypto_unregister_alg(&aes_algs[j]);
@@ -847,10 +1425,17 @@ err_aes_algs:
 
 static void sahara_unregister_algs(struct sahara_dev *dev)
 {
-	int i;
+	unsigned int i;
 
 	for (i = 0; i < ARRAY_SIZE(aes_algs); i++)
 		crypto_unregister_alg(&aes_algs[i]);
+
+	for (i = 0; i < ARRAY_SIZE(sha_v4_algs); i++)
+		crypto_unregister_ahash(&sha_v3_algs[i]);
+
+	if (dev->version > SAHARA_VERSION_3)
+		for (i = 0; i < ARRAY_SIZE(sha_v4_algs); i++)
+			crypto_unregister_ahash(&sha_v4_algs[i]);
 }
 
 static struct platform_device_id sahara_platform_ids[] = {
@@ -860,6 +1445,7 @@ static struct platform_device_id sahara_platform_ids[] = {
 MODULE_DEVICE_TABLE(platform, sahara_platform_ids);
 
 static struct of_device_id sahara_dt_ids[] = {
+	{ .compatible = "fsl,imx53-sahara" },
 	{ .compatible = "fsl,imx27-sahara" },
 	{ /* sentinel */ }
 };
@@ -939,6 +1525,16 @@ static int sahara_probe(struct platform_device *pdev)
 	dev->iv_base = dev->key_base + AES_KEYSIZE_128;
 	dev->iv_phys_base = dev->key_phys_base + AES_KEYSIZE_128;
 
+	/* Allocate space for context: largest digest + message length field */
+	dev->context_base = dma_alloc_coherent(&pdev->dev,
+					SHA256_DIGEST_SIZE + 4,
+					&dev->context_phys_base, GFP_KERNEL);
+	if (!dev->context_base) {
+		dev_err(&pdev->dev, "Could not allocate memory for MDHA context\n");
+		err = -ENOMEM;
+		goto err_key;
+	}
+
 	/* Allocate space for HW links */
 	dev->hw_link[0] = dma_alloc_coherent(&pdev->dev,
 			SAHARA_MAX_HW_LINK * sizeof(struct sahara_hw_link),
@@ -956,28 +1552,40 @@ static int sahara_probe(struct platform_device *pdev)
 
 	crypto_init_queue(&dev->queue, SAHARA_QUEUE_LENGTH);
 
+	spin_lock_init(&dev->lock);
+	mutex_init(&dev->queue_mutex);
+
 	dev_ptr = dev;
 
-	tasklet_init(&dev->queue_task, sahara_aes_queue_task,
-		     (unsigned long)dev);
-	tasklet_init(&dev->done_task, sahara_aes_done_task,
-		     (unsigned long)dev);
+	dev->kthread = kthread_run(sahara_queue_manage, dev, "sahara_crypto");
+	if (IS_ERR(dev->kthread)) {
+		err = PTR_ERR(dev->kthread);
+		goto err_link;
+	}
 
-	init_timer(&dev->watchdog);
-	dev->watchdog.function = &sahara_watchdog;
-	dev->watchdog.data = (unsigned long)dev;
+	init_completion(&dev->dma_completion);
 
 	clk_prepare_enable(dev->clk_ipg);
 	clk_prepare_enable(dev->clk_ahb);
 
 	version = sahara_read(dev, SAHARA_REG_VERSION);
-	if (version != SAHARA_VERSION_3) {
+	if (of_device_is_compatible(pdev->dev.of_node, "fsl,imx27-sahara")) {
+		if (version != SAHARA_VERSION_3)
+			err = -ENODEV;
+	} else if (of_device_is_compatible(pdev->dev.of_node,
+			"fsl,imx53-sahara")) {
+		if (((version >> 8) & 0xff) != SAHARA_VERSION_4)
+			err = -ENODEV;
+		version = (version >> 8) & 0xff;
+	}
+	if (err == -ENODEV) {
 		dev_err(&pdev->dev, "SAHARA version %d not supported\n",
-			version);
-		err = -ENODEV;
+				version);
 		goto err_algs;
 	}
 
+	dev->version = version;
+
 	sahara_write(dev, SAHARA_CMD_RESET | SAHARA_CMD_MODE_BATCH,
 		     SAHARA_REG_CMD);
 	sahara_write(dev, SAHARA_CONTROL_SET_THROTTLE(0) |
@@ -1000,11 +1608,15 @@ err_algs:
 			  dev->hw_link[0], dev->hw_phys_link[0]);
 	clk_disable_unprepare(dev->clk_ipg);
 	clk_disable_unprepare(dev->clk_ahb);
+	kthread_stop(dev->kthread);
 	dev_ptr = NULL;
 err_link:
 	dma_free_coherent(&pdev->dev,
 			  2 * AES_KEYSIZE_128,
 			  dev->key_base, dev->key_phys_base);
+	dma_free_coherent(&pdev->dev,
+			  SHA256_DIGEST_SIZE,
+			  dev->context_base, dev->context_phys_base);
 err_key:
 	dma_free_coherent(&pdev->dev,
 			  SAHARA_MAX_HW_DESC * sizeof(struct sahara_hw_desc),
@@ -1027,8 +1639,7 @@ static int sahara_remove(struct platform_device *pdev)
 			  SAHARA_MAX_HW_DESC * sizeof(struct sahara_hw_desc),
 			  dev->hw_desc[0], dev->hw_phys_desc[0]);
 
-	tasklet_kill(&dev->done_task);
-	tasklet_kill(&dev->queue_task);
+	kthread_stop(dev->kthread);
 
 	sahara_unregister_algs(dev);
 
@@ -1045,7 +1656,6 @@ static struct platform_driver sahara_driver = {
 	.remove		= sahara_remove,
 	.driver		= {
 		.name	= SAHARA_NAME,
-		.owner	= THIS_MODULE,
 		.of_match_table = sahara_dt_ids,
 	},
 	.id_table = sahara_platform_ids,
@@ -1055,4 +1665,5 @@ module_platform_driver(sahara_driver);
 
 MODULE_LICENSE("GPL");
 MODULE_AUTHOR("Javier Martin <javier.martin@vista-silicon.com>");
+MODULE_AUTHOR("Steffen Trumtrar <s.trumtrar@pengutronix.de>");
 MODULE_DESCRIPTION("SAHARA2 HW crypto accelerator");
diff --git a/drivers/crypto/talitos.c b/drivers/crypto/talitos.c
index 624b8be0c365..067ec2193d71 100644
--- a/drivers/crypto/talitos.c
+++ b/drivers/crypto/talitos.c
@@ -2811,7 +2811,6 @@ MODULE_DEVICE_TABLE(of, talitos_match);
 static struct platform_driver talitos_driver = {
 	.driver = {
 		.name = "talitos",
-		.owner = THIS_MODULE,
 		.of_match_table = talitos_match,
 	},
 	.probe = talitos_probe,
diff --git a/drivers/crypto/ux500/cryp/cryp_core.c b/drivers/crypto/ux500/cryp/cryp_core.c
index 92105f3dc8e0..f831bb952b2f 100644
--- a/drivers/crypto/ux500/cryp/cryp_core.c
+++ b/drivers/crypto/ux500/cryp/cryp_core.c
@@ -1688,6 +1688,7 @@ static void ux500_cryp_shutdown(struct platform_device *pdev)
 
 }
 
+#ifdef CONFIG_PM_SLEEP
 static int ux500_cryp_suspend(struct device *dev)
 {
 	int ret;
@@ -1768,6 +1769,7 @@ static int ux500_cryp_resume(struct device *dev)
 
 	return ret;
 }
+#endif
 
 static SIMPLE_DEV_PM_OPS(ux500_cryp_pm, ux500_cryp_suspend, ux500_cryp_resume);
 
@@ -1781,7 +1783,6 @@ static struct platform_driver cryp_driver = {
 	.remove = ux500_cryp_remove,
 	.shutdown = ux500_cryp_shutdown,
 	.driver = {
-		.owner = THIS_MODULE,
 		.name  = "cryp1",
 		.of_match_table = ux500_cryp_match,
 		.pm    = &ux500_cryp_pm,
@@ -1810,7 +1811,7 @@ module_exit(ux500_cryp_mod_fini);
 module_param(cryp_mode, int, 0);
 
 MODULE_DESCRIPTION("Driver for ST-Ericsson UX500 CRYP crypto engine.");
-MODULE_ALIAS("aes-all");
-MODULE_ALIAS("des-all");
+MODULE_ALIAS_CRYPTO("aes-all");
+MODULE_ALIAS_CRYPTO("des-all");
 
 MODULE_LICENSE("GPL");
diff --git a/drivers/crypto/ux500/hash/hash_core.c b/drivers/crypto/ux500/hash/hash_core.c
index 1c73f4fbc252..70a20871e998 100644
--- a/drivers/crypto/ux500/hash/hash_core.c
+++ b/drivers/crypto/ux500/hash/hash_core.c
@@ -1881,6 +1881,7 @@ static void ux500_hash_shutdown(struct platform_device *pdev)
 			__func__);
 }
 
+#ifdef CONFIG_PM_SLEEP
 /**
  * ux500_hash_suspend - Function that suspends the hash device.
  * @dev:	Device to suspend.
@@ -1949,6 +1950,7 @@ static int ux500_hash_resume(struct device *dev)
 
 	return ret;
 }
+#endif
 
 static SIMPLE_DEV_PM_OPS(ux500_hash_pm, ux500_hash_suspend, ux500_hash_resume);
 
@@ -1962,7 +1964,6 @@ static struct platform_driver hash_driver = {
 	.remove = ux500_hash_remove,
 	.shutdown = ux500_hash_shutdown,
 	.driver = {
-		.owner = THIS_MODULE,
 		.name  = "hash1",
 		.of_match_table = ux500_hash_match,
 		.pm    = &ux500_hash_pm,
@@ -1995,7 +1996,7 @@ module_exit(ux500_hash_mod_fini);
 MODULE_DESCRIPTION("Driver for ST-Ericsson UX500 HASH engine.");
 MODULE_LICENSE("GPL");
 
-MODULE_ALIAS("sha1-all");
-MODULE_ALIAS("sha256-all");
-MODULE_ALIAS("hmac-sha1-all");
-MODULE_ALIAS("hmac-sha256-all");
+MODULE_ALIAS_CRYPTO("sha1-all");
+MODULE_ALIAS_CRYPTO("sha256-all");
+MODULE_ALIAS_CRYPTO("hmac-sha1-all");
+MODULE_ALIAS_CRYPTO("hmac-sha256-all");
diff --git a/drivers/devfreq/Kconfig b/drivers/devfreq/Kconfig
index 3dced0a9eae3..faf4e70c42e0 100644
--- a/drivers/devfreq/Kconfig
+++ b/drivers/devfreq/Kconfig
@@ -78,9 +78,8 @@ config ARM_EXYNOS4_BUS_DEVFREQ
 	  This does not yet operate with optimal voltages.
 
 config ARM_EXYNOS5_BUS_DEVFREQ
-	bool "ARM Exynos5250 Bus DEVFREQ Driver"
+	tristate "ARM Exynos5250 Bus DEVFREQ Driver"
 	depends on SOC_EXYNOS5250
-	select ARCH_HAS_OPP
 	select DEVFREQ_GOV_SIMPLE_ONDEMAND
 	select PM_OPP
 	help
diff --git a/drivers/devfreq/devfreq.c b/drivers/devfreq/devfreq.c
index 9f90369dd6bd..30b538d8cc90 100644
--- a/drivers/devfreq/devfreq.c
+++ b/drivers/devfreq/devfreq.c
@@ -1119,6 +1119,7 @@ struct dev_pm_opp *devfreq_recommended_opp(struct device *dev,
 
 	return opp;
 }
+EXPORT_SYMBOL(devfreq_recommended_opp);
 
 /**
  * devfreq_register_opp_notifier() - Helper function to get devfreq notified
@@ -1142,6 +1143,7 @@ int devfreq_register_opp_notifier(struct device *dev, struct devfreq *devfreq)
 
 	return ret;
 }
+EXPORT_SYMBOL(devfreq_register_opp_notifier);
 
 /**
  * devfreq_unregister_opp_notifier() - Helper function to stop getting devfreq
@@ -1168,6 +1170,7 @@ int devfreq_unregister_opp_notifier(struct device *dev, struct devfreq *devfreq)
 
 	return ret;
 }
+EXPORT_SYMBOL(devfreq_unregister_opp_notifier);
 
 static void devm_devfreq_opp_release(struct device *dev, void *res)
 {
diff --git a/drivers/devfreq/exynos/exynos4_bus.c b/drivers/devfreq/exynos/exynos4_bus.c
index d9b08d3b6830..da9509205169 100644
--- a/drivers/devfreq/exynos/exynos4_bus.c
+++ b/drivers/devfreq/exynos/exynos4_bus.c
@@ -1034,7 +1034,6 @@ static struct platform_driver exynos4_busfreq_driver = {
 	.id_table = exynos4_busfreq_id,
 	.driver = {
 		.name	= "exynos4-busfreq",
-		.owner	= THIS_MODULE,
 		.pm	= &exynos4_busfreq_pm_ops,
 	},
 };
diff --git a/drivers/devfreq/exynos/exynos5_bus.c b/drivers/devfreq/exynos/exynos5_bus.c
index 6cd0392e2798..297ea30d4159 100644
--- a/drivers/devfreq/exynos/exynos5_bus.c
+++ b/drivers/devfreq/exynos/exynos5_bus.c
@@ -393,7 +393,6 @@ static struct platform_driver exynos5_busfreq_int_driver = {
 	.remove		= exynos5_busfreq_int_remove,
 	.driver		= {
 		.name		= "exynos5-bus-int",
-		.owner		= THIS_MODULE,
 		.pm		= &exynos5_busfreq_int_pm_ops,
 	},
 };
diff --git a/drivers/devfreq/exynos/exynos_ppmu.c b/drivers/devfreq/exynos/exynos_ppmu.c
index 75fcc5140ffb..97b75e513d29 100644
--- a/drivers/devfreq/exynos/exynos_ppmu.c
+++ b/drivers/devfreq/exynos/exynos_ppmu.c
@@ -73,6 +73,7 @@ void busfreq_mon_reset(struct busfreq_ppmu_data *ppmu_data)
 		exynos_ppmu_start(ppmu_base);
 	}
 }
+EXPORT_SYMBOL(busfreq_mon_reset);
 
 void exynos_read_ppmu(struct busfreq_ppmu_data *ppmu_data)
 {
@@ -97,6 +98,7 @@ void exynos_read_ppmu(struct busfreq_ppmu_data *ppmu_data)
 
 	busfreq_mon_reset(ppmu_data);
 }
+EXPORT_SYMBOL(exynos_read_ppmu);
 
 int exynos_get_busier_ppmu(struct busfreq_ppmu_data *ppmu_data)
 {
@@ -114,3 +116,4 @@ int exynos_get_busier_ppmu(struct busfreq_ppmu_data *ppmu_data)
 
 	return busy;
 }
+EXPORT_SYMBOL(exynos_get_busier_ppmu);
diff --git a/drivers/dma-buf/dma-buf.c b/drivers/dma-buf/dma-buf.c
index f3014c448e1e..5be225c2ba98 100644
--- a/drivers/dma-buf/dma-buf.c
+++ b/drivers/dma-buf/dma-buf.c
@@ -799,7 +799,7 @@ static int dma_buf_describe(struct seq_file *s)
 		seq_printf(s, "%08zu\t%08x\t%08x\t%08ld\t%s\n",
 				buf_obj->size,
 				buf_obj->file->f_flags, buf_obj->file->f_mode,
-				(long)(buf_obj->file->f_count.counter),
+				file_count(buf_obj->file),
 				buf_obj->exp_name);
 
 		seq_puts(s, "\tAttached Devices:\n");
diff --git a/drivers/dma-buf/fence.c b/drivers/dma-buf/fence.c
index 7bb9d65d9a2c..e5541117b3e9 100644
--- a/drivers/dma-buf/fence.c
+++ b/drivers/dma-buf/fence.c
@@ -283,7 +283,7 @@ EXPORT_SYMBOL(fence_add_callback);
  * @cb:		[in]	the callback to remove
  *
  * Remove a previously queued callback from the fence. This function returns
- * true if the callback is succesfully removed, or false if the fence has
+ * true if the callback is successfully removed, or false if the fence has
  * already been signaled.
  *
  * *WARNING*:
diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig
index 9b1ea0ef59af..f2b2c4e87aef 100644
--- a/drivers/dma/Kconfig
+++ b/drivers/dma/Kconfig
@@ -107,6 +107,13 @@ config AT_HDMAC
 	help
 	  Support the Atmel AHB DMA controller.
 
+config AT_XDMAC
+	tristate "Atmel XDMA support"
+	depends on ARCH_AT91
+	select DMA_ENGINE
+	help
+	  Support the Atmel XDMA controller.
+
 config FSL_DMA
 	tristate "Freescale Elo series DMA support"
 	depends on FSL_SOC
@@ -270,7 +277,7 @@ config IMX_SDMA
 	select DMA_ENGINE
 	help
 	  Support the i.MX SDMA engine. This engine is integrated into
-	  Freescale i.MX25/31/35/51/53 chips.
+	  Freescale i.MX25/31/35/51/53/6 chips.
 
 config IMX_DMA
 	tristate "i.MX DMA support"
@@ -395,12 +402,12 @@ config XILINX_VDMA
 
 config DMA_SUN6I
 	tristate "Allwinner A31 SoCs DMA support"
-	depends on MACH_SUN6I || COMPILE_TEST
+	depends on MACH_SUN6I || MACH_SUN8I || COMPILE_TEST
 	depends on RESET_CONTROLLER
 	select DMA_ENGINE
 	select DMA_VIRTUAL_CHANNELS
 	help
-	  Support for the DMA engine for Allwinner A31 SoCs.
+	  Support for the DMA engine first found in Allwinner A31 SoCs.
 
 config NBPFAXI_DMA
 	tristate "Renesas Type-AXI NBPF DMA support"
@@ -427,18 +434,6 @@ config DMA_OF
 comment "DMA Clients"
 	depends on DMA_ENGINE
 
-config NET_DMA
-	bool "Network: TCP receive copy offload"
-	depends on DMA_ENGINE && NET
-	default (INTEL_IOATDMA || FSL_DMA)
-	depends on BROKEN
-	help
-	  This enables the use of DMA engines in the network stack to
-	  offload receive copy-to-user operations, freeing CPU cycles.
-
-	  Say Y here if you enabled INTEL_IOATDMA or FSL_DMA, otherwise
-	  say N.
-
 config ASYNC_TX_DMA
 	bool "Async_tx: Offload support for the async_tx api"
 	depends on DMA_ENGINE
diff --git a/drivers/dma/Makefile b/drivers/dma/Makefile
index c6adb925f0b9..2022b5451377 100644
--- a/drivers/dma/Makefile
+++ b/drivers/dma/Makefile
@@ -6,7 +6,6 @@ obj-$(CONFIG_DMA_VIRTUAL_CHANNELS) += virt-dma.o
 obj-$(CONFIG_DMA_ACPI) += acpi-dma.o
 obj-$(CONFIG_DMA_OF) += of-dma.o
 
-obj-$(CONFIG_NET_DMA) += iovlock.o
 obj-$(CONFIG_INTEL_MID_DMAC) += intel_mid_dma.o
 obj-$(CONFIG_DMATEST) += dmatest.o
 obj-$(CONFIG_INTEL_IOATDMA) += ioat/
@@ -17,6 +16,7 @@ obj-$(CONFIG_PPC_BESTCOMM) += bestcomm/
 obj-$(CONFIG_MV_XOR) += mv_xor.o
 obj-$(CONFIG_DW_DMAC_CORE) += dw/
 obj-$(CONFIG_AT_HDMAC) += at_hdmac.o
+obj-$(CONFIG_AT_XDMAC) += at_xdmac.o
 obj-$(CONFIG_MX3_IPU) += ipu/
 obj-$(CONFIG_TXX9_DMAC) += txx9dmac.o
 obj-$(CONFIG_SH_DMAE_BASE) += sh/
diff --git a/drivers/dma/amba-pl08x.c b/drivers/dma/amba-pl08x.c
index e34024b000a4..1364d00881dd 100644
--- a/drivers/dma/amba-pl08x.c
+++ b/drivers/dma/amba-pl08x.c
@@ -2164,7 +2164,6 @@ static int pl08x_probe(struct amba_device *adev, const struct amba_id *id)
 			 __func__, ret);
 		goto out_no_memcpy;
 	}
-	pl08x->memcpy.chancnt = ret;
 
 	/* Register slave channels */
 	ret = pl08x_dma_init_virtual_channels(pl08x, &pl08x->slave,
@@ -2175,7 +2174,6 @@ static int pl08x_probe(struct amba_device *adev, const struct amba_id *id)
 				__func__, ret);
 		goto out_no_slave;
 	}
-	pl08x->slave.chancnt = ret;
 
 	ret = dma_async_device_register(&pl08x->memcpy);
 	if (ret) {
diff --git a/drivers/dma/at_xdmac.c b/drivers/dma/at_xdmac.c
new file mode 100644
index 000000000000..b60d77a22df6
--- /dev/null
+++ b/drivers/dma/at_xdmac.c
@@ -0,0 +1,1524 @@
+/*
+ * Driver for the Atmel Extensible DMA Controller (aka XDMAC on AT91 systems)
+ *
+ * Copyright (C) 2014 Atmel Corporation
+ *
+ * Author: Ludovic Desroches <ludovic.desroches@atmel.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <asm/barrier.h>
+#include <dt-bindings/dma/at91.h>
+#include <linux/clk.h>
+#include <linux/dmaengine.h>
+#include <linux/dmapool.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/list.h>
+#include <linux/module.h>
+#include <linux/of_dma.h>
+#include <linux/of_platform.h>
+#include <linux/platform_device.h>
+#include <linux/pm.h>
+
+#include "dmaengine.h"
+
+/* Global registers */
+#define AT_XDMAC_GTYPE		0x00	/* Global Type Register */
+#define		AT_XDMAC_NB_CH(i)	(((i) & 0x1F) + 1)		/* Number of Channels Minus One */
+#define		AT_XDMAC_FIFO_SZ(i)	(((i) >> 5) & 0x7FF)		/* Number of Bytes */
+#define		AT_XDMAC_NB_REQ(i)	((((i) >> 16) & 0x3F) + 1)	/* Number of Peripheral Requests Minus One */
+#define AT_XDMAC_GCFG		0x04	/* Global Configuration Register */
+#define AT_XDMAC_GWAC		0x08	/* Global Weighted Arbiter Configuration Register */
+#define AT_XDMAC_GIE		0x0C	/* Global Interrupt Enable Register */
+#define AT_XDMAC_GID		0x10	/* Global Interrupt Disable Register */
+#define AT_XDMAC_GIM		0x14	/* Global Interrupt Mask Register */
+#define AT_XDMAC_GIS		0x18	/* Global Interrupt Status Register */
+#define AT_XDMAC_GE		0x1C	/* Global Channel Enable Register */
+#define AT_XDMAC_GD		0x20	/* Global Channel Disable Register */
+#define AT_XDMAC_GS		0x24	/* Global Channel Status Register */
+#define AT_XDMAC_GRS		0x28	/* Global Channel Read Suspend Register */
+#define AT_XDMAC_GWS		0x2C	/* Global Write Suspend Register */
+#define AT_XDMAC_GRWS		0x30	/* Global Channel Read Write Suspend Register */
+#define AT_XDMAC_GRWR		0x34	/* Global Channel Read Write Resume Register */
+#define AT_XDMAC_GSWR		0x38	/* Global Channel Software Request Register */
+#define AT_XDMAC_GSWS		0x3C	/* Global channel Software Request Status Register */
+#define AT_XDMAC_GSWF		0x40	/* Global Channel Software Flush Request Register */
+#define AT_XDMAC_VERSION	0xFFC	/* XDMAC Version Register */
+
+/* Channel relative registers offsets */
+#define AT_XDMAC_CIE		0x00	/* Channel Interrupt Enable Register */
+#define		AT_XDMAC_CIE_BIE	BIT(0)	/* End of Block Interrupt Enable Bit */
+#define		AT_XDMAC_CIE_LIE	BIT(1)	/* End of Linked List Interrupt Enable Bit */
+#define		AT_XDMAC_CIE_DIE	BIT(2)	/* End of Disable Interrupt Enable Bit */
+#define		AT_XDMAC_CIE_FIE	BIT(3)	/* End of Flush Interrupt Enable Bit */
+#define		AT_XDMAC_CIE_RBEIE	BIT(4)	/* Read Bus Error Interrupt Enable Bit */
+#define		AT_XDMAC_CIE_WBEIE	BIT(5)	/* Write Bus Error Interrupt Enable Bit */
+#define		AT_XDMAC_CIE_ROIE	BIT(6)	/* Request Overflow Interrupt Enable Bit */
+#define AT_XDMAC_CID		0x04	/* Channel Interrupt Disable Register */
+#define		AT_XDMAC_CID_BID	BIT(0)	/* End of Block Interrupt Disable Bit */
+#define		AT_XDMAC_CID_LID	BIT(1)	/* End of Linked List Interrupt Disable Bit */
+#define		AT_XDMAC_CID_DID	BIT(2)	/* End of Disable Interrupt Disable Bit */
+#define		AT_XDMAC_CID_FID	BIT(3)	/* End of Flush Interrupt Disable Bit */
+#define		AT_XDMAC_CID_RBEID	BIT(4)	/* Read Bus Error Interrupt Disable Bit */
+#define		AT_XDMAC_CID_WBEID	BIT(5)	/* Write Bus Error Interrupt Disable Bit */
+#define		AT_XDMAC_CID_ROID	BIT(6)	/* Request Overflow Interrupt Disable Bit */
+#define AT_XDMAC_CIM		0x08	/* Channel Interrupt Mask Register */
+#define		AT_XDMAC_CIM_BIM	BIT(0)	/* End of Block Interrupt Mask Bit */
+#define		AT_XDMAC_CIM_LIM	BIT(1)	/* End of Linked List Interrupt Mask Bit */
+#define		AT_XDMAC_CIM_DIM	BIT(2)	/* End of Disable Interrupt Mask Bit */
+#define		AT_XDMAC_CIM_FIM	BIT(3)	/* End of Flush Interrupt Mask Bit */
+#define		AT_XDMAC_CIM_RBEIM	BIT(4)	/* Read Bus Error Interrupt Mask Bit */
+#define		AT_XDMAC_CIM_WBEIM	BIT(5)	/* Write Bus Error Interrupt Mask Bit */
+#define		AT_XDMAC_CIM_ROIM	BIT(6)	/* Request Overflow Interrupt Mask Bit */
+#define AT_XDMAC_CIS		0x0C	/* Channel Interrupt Status Register */
+#define		AT_XDMAC_CIS_BIS	BIT(0)	/* End of Block Interrupt Status Bit */
+#define		AT_XDMAC_CIS_LIS	BIT(1)	/* End of Linked List Interrupt Status Bit */
+#define		AT_XDMAC_CIS_DIS	BIT(2)	/* End of Disable Interrupt Status Bit */
+#define		AT_XDMAC_CIS_FIS	BIT(3)	/* End of Flush Interrupt Status Bit */
+#define		AT_XDMAC_CIS_RBEIS	BIT(4)	/* Read Bus Error Interrupt Status Bit */
+#define		AT_XDMAC_CIS_WBEIS	BIT(5)	/* Write Bus Error Interrupt Status Bit */
+#define		AT_XDMAC_CIS_ROIS	BIT(6)	/* Request Overflow Interrupt Status Bit */
+#define AT_XDMAC_CSA		0x10	/* Channel Source Address Register */
+#define AT_XDMAC_CDA		0x14	/* Channel Destination Address Register */
+#define AT_XDMAC_CNDA		0x18	/* Channel Next Descriptor Address Register */
+#define		AT_XDMAC_CNDA_NDAIF(i)	((i) & 0x1)			/* Channel x Next Descriptor Interface */
+#define		AT_XDMAC_CNDA_NDA(i)	((i) & 0xfffffffc)		/* Channel x Next Descriptor Address */
+#define AT_XDMAC_CNDC		0x1C	/* Channel Next Descriptor Control Register */
+#define		AT_XDMAC_CNDC_NDE		(0x1 << 0)		/* Channel x Next Descriptor Enable */
+#define		AT_XDMAC_CNDC_NDSUP		(0x1 << 1)		/* Channel x Next Descriptor Source Update */
+#define		AT_XDMAC_CNDC_NDDUP		(0x1 << 2)		/* Channel x Next Descriptor Destination Update */
+#define		AT_XDMAC_CNDC_NDVIEW_NDV0	(0x0 << 3)		/* Channel x Next Descriptor View 0 */
+#define		AT_XDMAC_CNDC_NDVIEW_NDV1	(0x1 << 3)		/* Channel x Next Descriptor View 1 */
+#define		AT_XDMAC_CNDC_NDVIEW_NDV2	(0x2 << 3)		/* Channel x Next Descriptor View 2 */
+#define		AT_XDMAC_CNDC_NDVIEW_NDV3	(0x3 << 3)		/* Channel x Next Descriptor View 3 */
+#define AT_XDMAC_CUBC		0x20	/* Channel Microblock Control Register */
+#define AT_XDMAC_CBC		0x24	/* Channel Block Control Register */
+#define AT_XDMAC_CC		0x28	/* Channel Configuration Register */
+#define		AT_XDMAC_CC_TYPE	(0x1 << 0)	/* Channel Transfer Type */
+#define			AT_XDMAC_CC_TYPE_MEM_TRAN	(0x0 << 0)	/* Memory to Memory Transfer */
+#define			AT_XDMAC_CC_TYPE_PER_TRAN	(0x1 << 0)	/* Peripheral to Memory or Memory to Peripheral Transfer */
+#define		AT_XDMAC_CC_MBSIZE_MASK	(0x3 << 1)
+#define			AT_XDMAC_CC_MBSIZE_SINGLE	(0x0 << 1)
+#define			AT_XDMAC_CC_MBSIZE_FOUR		(0x1 << 1)
+#define			AT_XDMAC_CC_MBSIZE_EIGHT	(0x2 << 1)
+#define			AT_XDMAC_CC_MBSIZE_SIXTEEN	(0x3 << 1)
+#define		AT_XDMAC_CC_DSYNC	(0x1 << 4)	/* Channel Synchronization */
+#define			AT_XDMAC_CC_DSYNC_PER2MEM	(0x0 << 4)
+#define			AT_XDMAC_CC_DSYNC_MEM2PER	(0x1 << 4)
+#define		AT_XDMAC_CC_PROT	(0x1 << 5)	/* Channel Protection */
+#define			AT_XDMAC_CC_PROT_SEC		(0x0 << 5)
+#define			AT_XDMAC_CC_PROT_UNSEC		(0x1 << 5)
+#define		AT_XDMAC_CC_SWREQ	(0x1 << 6)	/* Channel Software Request Trigger */
+#define			AT_XDMAC_CC_SWREQ_HWR_CONNECTED	(0x0 << 6)
+#define			AT_XDMAC_CC_SWREQ_SWR_CONNECTED	(0x1 << 6)
+#define		AT_XDMAC_CC_MEMSET	(0x1 << 7)	/* Channel Fill Block of memory */
+#define			AT_XDMAC_CC_MEMSET_NORMAL_MODE	(0x0 << 7)
+#define			AT_XDMAC_CC_MEMSET_HW_MODE	(0x1 << 7)
+#define		AT_XDMAC_CC_CSIZE(i)	((0x7 & (i)) << 8)	/* Channel Chunk Size */
+#define		AT_XDMAC_CC_DWIDTH_OFFSET	11
+#define		AT_XDMAC_CC_DWIDTH_MASK	(0x3 << AT_XDMAC_CC_DWIDTH_OFFSET)
+#define		AT_XDMAC_CC_DWIDTH(i)	((0x3 & (i)) << AT_XDMAC_CC_DWIDTH_OFFSET)	/* Channel Data Width */
+#define			AT_XDMAC_CC_DWIDTH_BYTE		0x0
+#define			AT_XDMAC_CC_DWIDTH_HALFWORD	0x1
+#define			AT_XDMAC_CC_DWIDTH_WORD		0x2
+#define			AT_XDMAC_CC_DWIDTH_DWORD	0x3
+#define		AT_XDMAC_CC_SIF(i)	((0x1 & (i)) << 13)	/* Channel Source Interface Identifier */
+#define		AT_XDMAC_CC_DIF(i)	((0x1 & (i)) << 14)	/* Channel Destination Interface Identifier */
+#define		AT_XDMAC_CC_SAM_MASK	(0x3 << 16)	/* Channel Source Addressing Mode */
+#define			AT_XDMAC_CC_SAM_FIXED_AM	(0x0 << 16)
+#define			AT_XDMAC_CC_SAM_INCREMENTED_AM	(0x1 << 16)
+#define			AT_XDMAC_CC_SAM_UBS_AM		(0x2 << 16)
+#define			AT_XDMAC_CC_SAM_UBS_DS_AM	(0x3 << 16)
+#define		AT_XDMAC_CC_DAM_MASK	(0x3 << 18)	/* Channel Source Addressing Mode */
+#define			AT_XDMAC_CC_DAM_FIXED_AM	(0x0 << 18)
+#define			AT_XDMAC_CC_DAM_INCREMENTED_AM	(0x1 << 18)
+#define			AT_XDMAC_CC_DAM_UBS_AM		(0x2 << 18)
+#define			AT_XDMAC_CC_DAM_UBS_DS_AM	(0x3 << 18)
+#define		AT_XDMAC_CC_INITD	(0x1 << 21)	/* Channel Initialization Terminated (read only) */
+#define			AT_XDMAC_CC_INITD_TERMINATED	(0x0 << 21)
+#define			AT_XDMAC_CC_INITD_IN_PROGRESS	(0x1 << 21)
+#define		AT_XDMAC_CC_RDIP	(0x1 << 22)	/* Read in Progress (read only) */
+#define			AT_XDMAC_CC_RDIP_DONE		(0x0 << 22)
+#define			AT_XDMAC_CC_RDIP_IN_PROGRESS	(0x1 << 22)
+#define		AT_XDMAC_CC_WRIP	(0x1 << 23)	/* Write in Progress (read only) */
+#define			AT_XDMAC_CC_WRIP_DONE		(0x0 << 23)
+#define			AT_XDMAC_CC_WRIP_IN_PROGRESS	(0x1 << 23)
+#define		AT_XDMAC_CC_PERID(i)	(0x7f & (h) << 24)	/* Channel Peripheral Identifier */
+#define AT_XDMAC_CDS_MSP	0x2C	/* Channel Data Stride Memory Set Pattern */
+#define AT_XDMAC_CSUS		0x30	/* Channel Source Microblock Stride */
+#define AT_XDMAC_CDUS		0x34	/* Channel Destination Microblock Stride */
+
+#define AT_XDMAC_CHAN_REG_BASE	0x50	/* Channel registers base address */
+
+/* Microblock control members */
+#define AT_XDMAC_MBR_UBC_UBLEN_MAX	0xFFFFFFUL	/* Maximum Microblock Length */
+#define AT_XDMAC_MBR_UBC_NDE		(0x1 << 24)	/* Next Descriptor Enable */
+#define AT_XDMAC_MBR_UBC_NSEN		(0x1 << 25)	/* Next Descriptor Source Update */
+#define AT_XDMAC_MBR_UBC_NDEN		(0x1 << 26)	/* Next Descriptor Destination Update */
+#define AT_XDMAC_MBR_UBC_NDV0		(0x0 << 27)	/* Next Descriptor View 0 */
+#define AT_XDMAC_MBR_UBC_NDV1		(0x1 << 27)	/* Next Descriptor View 1 */
+#define AT_XDMAC_MBR_UBC_NDV2		(0x2 << 27)	/* Next Descriptor View 2 */
+#define AT_XDMAC_MBR_UBC_NDV3		(0x3 << 27)	/* Next Descriptor View 3 */
+
+#define AT_XDMAC_MAX_CHAN	0x20
+
+enum atc_status {
+	AT_XDMAC_CHAN_IS_CYCLIC = 0,
+	AT_XDMAC_CHAN_IS_PAUSED,
+};
+
+/* ----- Channels ----- */
+struct at_xdmac_chan {
+	struct dma_chan			chan;
+	void __iomem			*ch_regs;
+	u32				mask;		/* Channel Mask */
+	u32				cfg[3];		/* Channel Configuration Register */
+	#define	AT_XDMAC_CUR_CFG	0		/* Current channel conf */
+	#define	AT_XDMAC_DEV_TO_MEM_CFG	1		/* Predifined dev to mem channel conf */
+	#define	AT_XDMAC_MEM_TO_DEV_CFG	2		/* Predifined mem to dev channel conf */
+	u8				perid;		/* Peripheral ID */
+	u8				perif;		/* Peripheral Interface */
+	u8				memif;		/* Memory Interface */
+	u32				per_src_addr;
+	u32				per_dst_addr;
+	u32				save_cim;
+	u32				save_cnda;
+	u32				save_cndc;
+	unsigned long			status;
+	struct tasklet_struct		tasklet;
+
+	spinlock_t			lock;
+
+	struct list_head		xfers_list;
+	struct list_head		free_descs_list;
+};
+
+
+/* ----- Controller ----- */
+struct at_xdmac {
+	struct dma_device	dma;
+	void __iomem		*regs;
+	int			irq;
+	struct clk		*clk;
+	u32			save_gim;
+	u32			save_gs;
+	struct dma_pool		*at_xdmac_desc_pool;
+	struct at_xdmac_chan	chan[0];
+};
+
+
+/* ----- Descriptors ----- */
+
+/* Linked List Descriptor */
+struct at_xdmac_lld {
+	dma_addr_t	mbr_nda;	/* Next Descriptor Member */
+	u32		mbr_ubc;	/* Microblock Control Member */
+	dma_addr_t	mbr_sa;		/* Source Address Member */
+	dma_addr_t	mbr_da;		/* Destination Address Member */
+	u32		mbr_cfg;	/* Configuration Register */
+};
+
+
+struct at_xdmac_desc {
+	struct at_xdmac_lld		lld;
+	enum dma_transfer_direction	direction;
+	struct dma_async_tx_descriptor	tx_dma_desc;
+	struct list_head		desc_node;
+	/* Following members are only used by the first descriptor */
+	bool				active_xfer;
+	unsigned int			xfer_size;
+	struct list_head		descs_list;
+	struct list_head		xfer_node;
+};
+
+static inline void __iomem *at_xdmac_chan_reg_base(struct at_xdmac *atxdmac, unsigned int chan_nb)
+{
+	return atxdmac->regs + (AT_XDMAC_CHAN_REG_BASE + chan_nb * 0x40);
+}
+
+#define at_xdmac_read(atxdmac, reg) readl_relaxed((atxdmac)->regs + (reg))
+#define at_xdmac_write(atxdmac, reg, value) \
+	writel_relaxed((value), (atxdmac)->regs + (reg))
+
+#define at_xdmac_chan_read(atchan, reg) readl_relaxed((atchan)->ch_regs + (reg))
+#define at_xdmac_chan_write(atchan, reg, value) writel_relaxed((value), (atchan)->ch_regs + (reg))
+
+static inline struct at_xdmac_chan *to_at_xdmac_chan(struct dma_chan *dchan)
+{
+	return container_of(dchan, struct at_xdmac_chan, chan);
+}
+
+static struct device *chan2dev(struct dma_chan *chan)
+{
+	return &chan->dev->device;
+}
+
+static inline struct at_xdmac *to_at_xdmac(struct dma_device *ddev)
+{
+	return container_of(ddev, struct at_xdmac, dma);
+}
+
+static inline struct at_xdmac_desc *txd_to_at_desc(struct dma_async_tx_descriptor *txd)
+{
+	return container_of(txd, struct at_xdmac_desc, tx_dma_desc);
+}
+
+static inline int at_xdmac_chan_is_cyclic(struct at_xdmac_chan *atchan)
+{
+	return test_bit(AT_XDMAC_CHAN_IS_CYCLIC, &atchan->status);
+}
+
+static inline int at_xdmac_chan_is_paused(struct at_xdmac_chan *atchan)
+{
+	return test_bit(AT_XDMAC_CHAN_IS_PAUSED, &atchan->status);
+}
+
+static inline int at_xdmac_csize(u32 maxburst)
+{
+	int csize;
+
+	csize = ffs(maxburst) - 1;
+	if (csize > 4)
+		csize = -EINVAL;
+
+	return csize;
+};
+
+static inline u8 at_xdmac_get_dwidth(u32 cfg)
+{
+	return (cfg & AT_XDMAC_CC_DWIDTH_MASK) >> AT_XDMAC_CC_DWIDTH_OFFSET;
+};
+
+static unsigned int init_nr_desc_per_channel = 64;
+module_param(init_nr_desc_per_channel, uint, 0644);
+MODULE_PARM_DESC(init_nr_desc_per_channel,
+		 "initial descriptors per channel (default: 64)");
+
+
+static bool at_xdmac_chan_is_enabled(struct at_xdmac_chan *atchan)
+{
+	return at_xdmac_chan_read(atchan, AT_XDMAC_GS) & atchan->mask;
+}
+
+static void at_xdmac_off(struct at_xdmac *atxdmac)
+{
+	at_xdmac_write(atxdmac, AT_XDMAC_GD, -1L);
+
+	/* Wait that all chans are disabled. */
+	while (at_xdmac_read(atxdmac, AT_XDMAC_GS))
+		cpu_relax();
+
+	at_xdmac_write(atxdmac, AT_XDMAC_GID, -1L);
+}
+
+/* Call with lock hold. */
+static void at_xdmac_start_xfer(struct at_xdmac_chan *atchan,
+				struct at_xdmac_desc *first)
+{
+	struct at_xdmac	*atxdmac = to_at_xdmac(atchan->chan.device);
+	u32		reg;
+
+	dev_vdbg(chan2dev(&atchan->chan), "%s: desc 0x%p\n", __func__, first);
+
+	if (at_xdmac_chan_is_enabled(atchan))
+		return;
+
+	/* Set transfer as active to not try to start it again. */
+	first->active_xfer = true;
+
+	/* Tell xdmac where to get the first descriptor. */
+	reg = AT_XDMAC_CNDA_NDA(first->tx_dma_desc.phys)
+	      | AT_XDMAC_CNDA_NDAIF(atchan->memif);
+	at_xdmac_chan_write(atchan, AT_XDMAC_CNDA, reg);
+
+	/*
+	 * When doing memory to memory transfer we need to use the next
+	 * descriptor view 2 since some fields of the configuration register
+	 * depend on transfer size and src/dest addresses.
+	 */
+	if (is_slave_direction(first->direction)) {
+		reg = AT_XDMAC_CNDC_NDVIEW_NDV1;
+		if (first->direction == DMA_MEM_TO_DEV)
+			atchan->cfg[AT_XDMAC_CUR_CFG] =
+				atchan->cfg[AT_XDMAC_MEM_TO_DEV_CFG];
+		else
+			atchan->cfg[AT_XDMAC_CUR_CFG] =
+				atchan->cfg[AT_XDMAC_DEV_TO_MEM_CFG];
+		at_xdmac_chan_write(atchan, AT_XDMAC_CC,
+				    atchan->cfg[AT_XDMAC_CUR_CFG]);
+	} else {
+		/*
+		 * No need to write AT_XDMAC_CC reg, it will be done when the
+		 * descriptor is fecthed.
+		 */
+		reg = AT_XDMAC_CNDC_NDVIEW_NDV2;
+	}
+
+	reg |= AT_XDMAC_CNDC_NDDUP
+	       | AT_XDMAC_CNDC_NDSUP
+	       | AT_XDMAC_CNDC_NDE;
+	at_xdmac_chan_write(atchan, AT_XDMAC_CNDC, reg);
+
+	dev_vdbg(chan2dev(&atchan->chan),
+		 "%s: CC=0x%08x CNDA=0x%08x, CNDC=0x%08x, CSA=0x%08x, CDA=0x%08x, CUBC=0x%08x\n",
+		 __func__, at_xdmac_chan_read(atchan, AT_XDMAC_CC),
+		 at_xdmac_chan_read(atchan, AT_XDMAC_CNDA),
+		 at_xdmac_chan_read(atchan, AT_XDMAC_CNDC),
+		 at_xdmac_chan_read(atchan, AT_XDMAC_CSA),
+		 at_xdmac_chan_read(atchan, AT_XDMAC_CDA),
+		 at_xdmac_chan_read(atchan, AT_XDMAC_CUBC));
+
+	at_xdmac_chan_write(atchan, AT_XDMAC_CID, 0xffffffff);
+	reg = AT_XDMAC_CIE_RBEIE | AT_XDMAC_CIE_WBEIE | AT_XDMAC_CIE_ROIE;
+	/*
+	 * There is no end of list when doing cyclic dma, we need to get
+	 * an interrupt after each periods.
+	 */
+	if (at_xdmac_chan_is_cyclic(atchan))
+		at_xdmac_chan_write(atchan, AT_XDMAC_CIE,
+				    reg | AT_XDMAC_CIE_BIE);
+	else
+		at_xdmac_chan_write(atchan, AT_XDMAC_CIE,
+				    reg | AT_XDMAC_CIE_LIE);
+	at_xdmac_write(atxdmac, AT_XDMAC_GIE, atchan->mask);
+	dev_vdbg(chan2dev(&atchan->chan),
+		 "%s: enable channel (0x%08x)\n", __func__, atchan->mask);
+	wmb();
+	at_xdmac_write(atxdmac, AT_XDMAC_GE, atchan->mask);
+
+	dev_vdbg(chan2dev(&atchan->chan),
+		 "%s: CC=0x%08x CNDA=0x%08x, CNDC=0x%08x, CSA=0x%08x, CDA=0x%08x, CUBC=0x%08x\n",
+		 __func__, at_xdmac_chan_read(atchan, AT_XDMAC_CC),
+		 at_xdmac_chan_read(atchan, AT_XDMAC_CNDA),
+		 at_xdmac_chan_read(atchan, AT_XDMAC_CNDC),
+		 at_xdmac_chan_read(atchan, AT_XDMAC_CSA),
+		 at_xdmac_chan_read(atchan, AT_XDMAC_CDA),
+		 at_xdmac_chan_read(atchan, AT_XDMAC_CUBC));
+
+}
+
+static dma_cookie_t at_xdmac_tx_submit(struct dma_async_tx_descriptor *tx)
+{
+	struct at_xdmac_desc	*desc = txd_to_at_desc(tx);
+	struct at_xdmac_chan	*atchan = to_at_xdmac_chan(tx->chan);
+	dma_cookie_t		cookie;
+
+	spin_lock_bh(&atchan->lock);
+	cookie = dma_cookie_assign(tx);
+
+	dev_vdbg(chan2dev(tx->chan), "%s: atchan 0x%p, add desc 0x%p to xfers_list\n",
+		 __func__, atchan, desc);
+	list_add_tail(&desc->xfer_node, &atchan->xfers_list);
+	if (list_is_singular(&atchan->xfers_list))
+		at_xdmac_start_xfer(atchan, desc);
+
+	spin_unlock_bh(&atchan->lock);
+	return cookie;
+}
+
+static struct at_xdmac_desc *at_xdmac_alloc_desc(struct dma_chan *chan,
+						 gfp_t gfp_flags)
+{
+	struct at_xdmac_desc	*desc;
+	struct at_xdmac		*atxdmac = to_at_xdmac(chan->device);
+	dma_addr_t		phys;
+
+	desc = dma_pool_alloc(atxdmac->at_xdmac_desc_pool, gfp_flags, &phys);
+	if (desc) {
+		memset(desc, 0, sizeof(*desc));
+		INIT_LIST_HEAD(&desc->descs_list);
+		dma_async_tx_descriptor_init(&desc->tx_dma_desc, chan);
+		desc->tx_dma_desc.tx_submit = at_xdmac_tx_submit;
+		desc->tx_dma_desc.phys = phys;
+	}
+
+	return desc;
+}
+
+/* Call must be protected by lock. */
+static struct at_xdmac_desc *at_xdmac_get_desc(struct at_xdmac_chan *atchan)
+{
+	struct at_xdmac_desc *desc;
+
+	if (list_empty(&atchan->free_descs_list)) {
+		desc = at_xdmac_alloc_desc(&atchan->chan, GFP_NOWAIT);
+	} else {
+		desc = list_first_entry(&atchan->free_descs_list,
+					struct at_xdmac_desc, desc_node);
+		list_del(&desc->desc_node);
+		desc->active_xfer = false;
+	}
+
+	return desc;
+}
+
+static struct dma_chan *at_xdmac_xlate(struct of_phandle_args *dma_spec,
+				       struct of_dma *of_dma)
+{
+	struct at_xdmac		*atxdmac = of_dma->of_dma_data;
+	struct at_xdmac_chan	*atchan;
+	struct dma_chan		*chan;
+	struct device		*dev = atxdmac->dma.dev;
+
+	if (dma_spec->args_count != 1) {
+		dev_err(dev, "dma phandler args: bad number of args\n");
+		return NULL;
+	}
+
+	chan = dma_get_any_slave_channel(&atxdmac->dma);
+	if (!chan) {
+		dev_err(dev, "can't get a dma channel\n");
+		return NULL;
+	}
+
+	atchan = to_at_xdmac_chan(chan);
+	atchan->memif = AT91_XDMAC_DT_GET_MEM_IF(dma_spec->args[0]);
+	atchan->perif = AT91_XDMAC_DT_GET_PER_IF(dma_spec->args[0]);
+	atchan->perid = AT91_XDMAC_DT_GET_PERID(dma_spec->args[0]);
+	dev_dbg(dev, "chan dt cfg: memif=%u perif=%u perid=%u\n",
+		 atchan->memif, atchan->perif, atchan->perid);
+
+	return chan;
+}
+
+static int at_xdmac_set_slave_config(struct dma_chan *chan,
+				      struct dma_slave_config *sconfig)
+{
+	struct at_xdmac_chan	*atchan = to_at_xdmac_chan(chan);
+	u8 dwidth;
+	int csize;
+
+	atchan->cfg[AT_XDMAC_DEV_TO_MEM_CFG] =
+		AT91_XDMAC_DT_PERID(atchan->perid)
+		| AT_XDMAC_CC_DAM_INCREMENTED_AM
+		| AT_XDMAC_CC_SAM_FIXED_AM
+		| AT_XDMAC_CC_DIF(atchan->memif)
+		| AT_XDMAC_CC_SIF(atchan->perif)
+		| AT_XDMAC_CC_SWREQ_HWR_CONNECTED
+		| AT_XDMAC_CC_DSYNC_PER2MEM
+		| AT_XDMAC_CC_MBSIZE_SIXTEEN
+		| AT_XDMAC_CC_TYPE_PER_TRAN;
+	csize = at_xdmac_csize(sconfig->src_maxburst);
+	if (csize < 0) {
+		dev_err(chan2dev(chan), "invalid src maxburst value\n");
+		return -EINVAL;
+	}
+	atchan->cfg[AT_XDMAC_DEV_TO_MEM_CFG] |= AT_XDMAC_CC_CSIZE(csize);
+	dwidth = ffs(sconfig->src_addr_width) - 1;
+	atchan->cfg[AT_XDMAC_DEV_TO_MEM_CFG] |= AT_XDMAC_CC_DWIDTH(dwidth);
+
+
+	atchan->cfg[AT_XDMAC_MEM_TO_DEV_CFG] =
+		AT91_XDMAC_DT_PERID(atchan->perid)
+		| AT_XDMAC_CC_DAM_FIXED_AM
+		| AT_XDMAC_CC_SAM_INCREMENTED_AM
+		| AT_XDMAC_CC_DIF(atchan->perif)
+		| AT_XDMAC_CC_SIF(atchan->memif)
+		| AT_XDMAC_CC_SWREQ_HWR_CONNECTED
+		| AT_XDMAC_CC_DSYNC_MEM2PER
+		| AT_XDMAC_CC_MBSIZE_SIXTEEN
+		| AT_XDMAC_CC_TYPE_PER_TRAN;
+	csize = at_xdmac_csize(sconfig->dst_maxburst);
+	if (csize < 0) {
+		dev_err(chan2dev(chan), "invalid src maxburst value\n");
+		return -EINVAL;
+	}
+	atchan->cfg[AT_XDMAC_MEM_TO_DEV_CFG] |= AT_XDMAC_CC_CSIZE(csize);
+	dwidth = ffs(sconfig->dst_addr_width) - 1;
+	atchan->cfg[AT_XDMAC_MEM_TO_DEV_CFG] |= AT_XDMAC_CC_DWIDTH(dwidth);
+
+	/* Src and dst addr are needed to configure the link list descriptor. */
+	atchan->per_src_addr = sconfig->src_addr;
+	atchan->per_dst_addr = sconfig->dst_addr;
+
+	dev_dbg(chan2dev(chan),
+		"%s: cfg[dev2mem]=0x%08x, cfg[mem2dev]=0x%08x, per_src_addr=0x%08x, per_dst_addr=0x%08x\n",
+		__func__, atchan->cfg[AT_XDMAC_DEV_TO_MEM_CFG],
+		atchan->cfg[AT_XDMAC_MEM_TO_DEV_CFG],
+		atchan->per_src_addr, atchan->per_dst_addr);
+
+	return 0;
+}
+
+static struct dma_async_tx_descriptor *
+at_xdmac_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
+		       unsigned int sg_len, enum dma_transfer_direction direction,
+		       unsigned long flags, void *context)
+{
+	struct at_xdmac_chan	*atchan = to_at_xdmac_chan(chan);
+	struct at_xdmac_desc	*first = NULL, *prev = NULL;
+	struct scatterlist	*sg;
+	int			i;
+	u32			cfg;
+	unsigned int		xfer_size = 0;
+
+	if (!sgl)
+		return NULL;
+
+	if (!is_slave_direction(direction)) {
+		dev_err(chan2dev(chan), "invalid DMA direction\n");
+		return NULL;
+	}
+
+	dev_dbg(chan2dev(chan), "%s: sg_len=%d, dir=%s, flags=0x%lx\n",
+		 __func__, sg_len,
+		 direction == DMA_MEM_TO_DEV ? "to device" : "from device",
+		 flags);
+
+	/* Protect dma_sconfig field that can be modified by set_slave_conf. */
+	spin_lock_bh(&atchan->lock);
+
+	/* Prepare descriptors. */
+	for_each_sg(sgl, sg, sg_len, i) {
+		struct at_xdmac_desc	*desc = NULL;
+		u32			len, mem;
+
+		len = sg_dma_len(sg);
+		mem = sg_dma_address(sg);
+		if (unlikely(!len)) {
+			dev_err(chan2dev(chan), "sg data length is zero\n");
+			spin_unlock_bh(&atchan->lock);
+			return NULL;
+		}
+		dev_dbg(chan2dev(chan), "%s: * sg%d len=%u, mem=0x%08x\n",
+			 __func__, i, len, mem);
+
+		desc = at_xdmac_get_desc(atchan);
+		if (!desc) {
+			dev_err(chan2dev(chan), "can't get descriptor\n");
+			if (first)
+				list_splice_init(&first->descs_list, &atchan->free_descs_list);
+			spin_unlock_bh(&atchan->lock);
+			return NULL;
+		}
+
+		/* Linked list descriptor setup. */
+		if (direction == DMA_DEV_TO_MEM) {
+			desc->lld.mbr_sa = atchan->per_src_addr;
+			desc->lld.mbr_da = mem;
+			cfg = atchan->cfg[AT_XDMAC_DEV_TO_MEM_CFG];
+		} else {
+			desc->lld.mbr_sa = mem;
+			desc->lld.mbr_da = atchan->per_dst_addr;
+			cfg = atchan->cfg[AT_XDMAC_MEM_TO_DEV_CFG];
+		}
+		desc->lld.mbr_ubc = AT_XDMAC_MBR_UBC_NDV1		/* next descriptor view */
+			| AT_XDMAC_MBR_UBC_NDEN				/* next descriptor dst parameter update */
+			| AT_XDMAC_MBR_UBC_NSEN				/* next descriptor src parameter update */
+			| (i == sg_len - 1 ? 0 : AT_XDMAC_MBR_UBC_NDE)	/* descriptor fetch */
+			| len / (1 << at_xdmac_get_dwidth(cfg));	/* microblock length */
+		dev_dbg(chan2dev(chan),
+			 "%s: lld: mbr_sa=%pad, mbr_da=%pad, mbr_ubc=0x%08x\n",
+			 __func__, &desc->lld.mbr_sa, &desc->lld.mbr_da, desc->lld.mbr_ubc);
+
+		/* Chain lld. */
+		if (prev) {
+			prev->lld.mbr_nda = desc->tx_dma_desc.phys;
+			dev_dbg(chan2dev(chan),
+				 "%s: chain lld: prev=0x%p, mbr_nda=%pad\n",
+				 __func__, prev, &prev->lld.mbr_nda);
+		}
+
+		prev = desc;
+		if (!first)
+			first = desc;
+
+		dev_dbg(chan2dev(chan), "%s: add desc 0x%p to descs_list 0x%p\n",
+			 __func__, desc, first);
+		list_add_tail(&desc->desc_node, &first->descs_list);
+		xfer_size += len;
+	}
+
+	spin_unlock_bh(&atchan->lock);
+
+	first->tx_dma_desc.flags = flags;
+	first->xfer_size = xfer_size;
+	first->direction = direction;
+
+	return &first->tx_dma_desc;
+}
+
+static struct dma_async_tx_descriptor *
+at_xdmac_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t buf_addr,
+			 size_t buf_len, size_t period_len,
+			 enum dma_transfer_direction direction,
+			 unsigned long flags)
+{
+	struct at_xdmac_chan	*atchan = to_at_xdmac_chan(chan);
+	struct at_xdmac_desc	*first = NULL, *prev = NULL;
+	unsigned int		periods = buf_len / period_len;
+	int			i;
+	u32			cfg;
+
+	dev_dbg(chan2dev(chan), "%s: buf_addr=%pad, buf_len=%zd, period_len=%zd, dir=%s, flags=0x%lx\n",
+		__func__, &buf_addr, buf_len, period_len,
+		direction == DMA_MEM_TO_DEV ? "mem2per" : "per2mem", flags);
+
+	if (!is_slave_direction(direction)) {
+		dev_err(chan2dev(chan), "invalid DMA direction\n");
+		return NULL;
+	}
+
+	if (test_and_set_bit(AT_XDMAC_CHAN_IS_CYCLIC, &atchan->status)) {
+		dev_err(chan2dev(chan), "channel currently used\n");
+		return NULL;
+	}
+
+	for (i = 0; i < periods; i++) {
+		struct at_xdmac_desc	*desc = NULL;
+
+		spin_lock_bh(&atchan->lock);
+		desc = at_xdmac_get_desc(atchan);
+		if (!desc) {
+			dev_err(chan2dev(chan), "can't get descriptor\n");
+			if (first)
+				list_splice_init(&first->descs_list, &atchan->free_descs_list);
+			spin_unlock_bh(&atchan->lock);
+			return NULL;
+		}
+		spin_unlock_bh(&atchan->lock);
+		dev_dbg(chan2dev(chan),
+			"%s: desc=0x%p, tx_dma_desc.phys=%pad\n",
+			__func__, desc, &desc->tx_dma_desc.phys);
+
+		if (direction == DMA_DEV_TO_MEM) {
+			desc->lld.mbr_sa = atchan->per_src_addr;
+			desc->lld.mbr_da = buf_addr + i * period_len;
+			cfg = atchan->cfg[AT_XDMAC_DEV_TO_MEM_CFG];
+		} else {
+			desc->lld.mbr_sa = buf_addr + i * period_len;
+			desc->lld.mbr_da = atchan->per_dst_addr;
+			cfg = atchan->cfg[AT_XDMAC_MEM_TO_DEV_CFG];
+		}
+		desc->lld.mbr_ubc = AT_XDMAC_MBR_UBC_NDV1
+			| AT_XDMAC_MBR_UBC_NDEN
+			| AT_XDMAC_MBR_UBC_NSEN
+			| AT_XDMAC_MBR_UBC_NDE
+			| period_len >> at_xdmac_get_dwidth(cfg);
+
+		dev_dbg(chan2dev(chan),
+			 "%s: lld: mbr_sa=%pad, mbr_da=%pad, mbr_ubc=0x%08x\n",
+			 __func__, &desc->lld.mbr_sa, &desc->lld.mbr_da, desc->lld.mbr_ubc);
+
+		/* Chain lld. */
+		if (prev) {
+			prev->lld.mbr_nda = desc->tx_dma_desc.phys;
+			dev_dbg(chan2dev(chan),
+				 "%s: chain lld: prev=0x%p, mbr_nda=%pad\n",
+				 __func__, prev, &prev->lld.mbr_nda);
+		}
+
+		prev = desc;
+		if (!first)
+			first = desc;
+
+		dev_dbg(chan2dev(chan), "%s: add desc 0x%p to descs_list 0x%p\n",
+			 __func__, desc, first);
+		list_add_tail(&desc->desc_node, &first->descs_list);
+	}
+
+	prev->lld.mbr_nda = first->tx_dma_desc.phys;
+	dev_dbg(chan2dev(chan),
+		"%s: chain lld: prev=0x%p, mbr_nda=%pad\n",
+		__func__, prev, &prev->lld.mbr_nda);
+	first->tx_dma_desc.flags = flags;
+	first->xfer_size = buf_len;
+	first->direction = direction;
+
+	return &first->tx_dma_desc;
+}
+
+static struct dma_async_tx_descriptor *
+at_xdmac_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
+			 size_t len, unsigned long flags)
+{
+	struct at_xdmac_chan	*atchan = to_at_xdmac_chan(chan);
+	struct at_xdmac_desc	*first = NULL, *prev = NULL;
+	size_t			remaining_size = len, xfer_size = 0, ublen;
+	dma_addr_t		src_addr = src, dst_addr = dest;
+	u32			dwidth;
+	/*
+	 * WARNING: We don't know the direction, it involves we can't
+	 * dynamically set the source and dest interface so we have to use the
+	 * same one. Only interface 0 allows EBI access. Hopefully we can
+	 * access DDR through both ports (at least on SAMA5D4x), so we can use
+	 * the same interface for source and dest, that solves the fact we
+	 * don't know the direction.
+	 */
+	u32			chan_cc = AT_XDMAC_CC_DAM_INCREMENTED_AM
+					| AT_XDMAC_CC_SAM_INCREMENTED_AM
+					| AT_XDMAC_CC_DIF(0)
+					| AT_XDMAC_CC_SIF(0)
+					| AT_XDMAC_CC_MBSIZE_SIXTEEN
+					| AT_XDMAC_CC_TYPE_MEM_TRAN;
+
+	dev_dbg(chan2dev(chan), "%s: src=%pad, dest=%pad, len=%zd, flags=0x%lx\n",
+		__func__, &src, &dest, len, flags);
+
+	if (unlikely(!len))
+		return NULL;
+
+	/*
+	 * Check address alignment to select the greater data width we can use.
+	 * Some XDMAC implementations don't provide dword transfer, in this
+	 * case selecting dword has the same behavior as selecting word transfers.
+	 */
+	if (!((src_addr | dst_addr) & 7)) {
+		dwidth = AT_XDMAC_CC_DWIDTH_DWORD;
+		dev_dbg(chan2dev(chan), "%s: dwidth: double word\n", __func__);
+	} else if (!((src_addr | dst_addr)  & 3)) {
+		dwidth = AT_XDMAC_CC_DWIDTH_WORD;
+		dev_dbg(chan2dev(chan), "%s: dwidth: word\n", __func__);
+	} else if (!((src_addr | dst_addr) & 1)) {
+		dwidth = AT_XDMAC_CC_DWIDTH_HALFWORD;
+		dev_dbg(chan2dev(chan), "%s: dwidth: half word\n", __func__);
+	} else {
+		dwidth = AT_XDMAC_CC_DWIDTH_BYTE;
+		dev_dbg(chan2dev(chan), "%s: dwidth: byte\n", __func__);
+	}
+
+	/* Prepare descriptors. */
+	while (remaining_size) {
+		struct at_xdmac_desc	*desc = NULL;
+
+		dev_dbg(chan2dev(chan), "%s: remaining_size=%zu\n", __func__, remaining_size);
+
+		spin_lock_bh(&atchan->lock);
+		desc = at_xdmac_get_desc(atchan);
+		spin_unlock_bh(&atchan->lock);
+		if (!desc) {
+			dev_err(chan2dev(chan), "can't get descriptor\n");
+			if (first)
+				list_splice_init(&first->descs_list, &atchan->free_descs_list);
+			return NULL;
+		}
+
+		/* Update src and dest addresses. */
+		src_addr += xfer_size;
+		dst_addr += xfer_size;
+
+		if (remaining_size >= AT_XDMAC_MBR_UBC_UBLEN_MAX << dwidth)
+			xfer_size = AT_XDMAC_MBR_UBC_UBLEN_MAX << dwidth;
+		else
+			xfer_size = remaining_size;
+
+		dev_dbg(chan2dev(chan), "%s: xfer_size=%zu\n", __func__, xfer_size);
+
+		/* Check remaining length and change data width if needed. */
+		if (!((src_addr | dst_addr | xfer_size) & 7)) {
+			dwidth = AT_XDMAC_CC_DWIDTH_DWORD;
+			dev_dbg(chan2dev(chan), "%s: dwidth: double word\n", __func__);
+		} else if (!((src_addr | dst_addr | xfer_size)  & 3)) {
+			dwidth = AT_XDMAC_CC_DWIDTH_WORD;
+			dev_dbg(chan2dev(chan), "%s: dwidth: word\n", __func__);
+		} else if (!((src_addr | dst_addr | xfer_size) & 1)) {
+			dwidth = AT_XDMAC_CC_DWIDTH_HALFWORD;
+			dev_dbg(chan2dev(chan), "%s: dwidth: half word\n", __func__);
+		} else if ((src_addr | dst_addr | xfer_size) & 1) {
+			dwidth = AT_XDMAC_CC_DWIDTH_BYTE;
+			dev_dbg(chan2dev(chan), "%s: dwidth: byte\n", __func__);
+		}
+		chan_cc |= AT_XDMAC_CC_DWIDTH(dwidth);
+
+		ublen = xfer_size >> dwidth;
+		remaining_size -= xfer_size;
+
+		desc->lld.mbr_sa = src_addr;
+		desc->lld.mbr_da = dst_addr;
+		desc->lld.mbr_ubc = AT_XDMAC_MBR_UBC_NDV2
+			| AT_XDMAC_MBR_UBC_NDEN
+			| AT_XDMAC_MBR_UBC_NSEN
+			| (remaining_size ? AT_XDMAC_MBR_UBC_NDE : 0)
+			| ublen;
+		desc->lld.mbr_cfg = chan_cc;
+
+		dev_dbg(chan2dev(chan),
+			 "%s: lld: mbr_sa=%pad, mbr_da=%pad, mbr_ubc=0x%08x, mbr_cfg=0x%08x\n",
+			 __func__, &desc->lld.mbr_sa, &desc->lld.mbr_da, desc->lld.mbr_ubc, desc->lld.mbr_cfg);
+
+		/* Chain lld. */
+		if (prev) {
+			prev->lld.mbr_nda = desc->tx_dma_desc.phys;
+			dev_dbg(chan2dev(chan),
+				 "%s: chain lld: prev=0x%p, mbr_nda=0x%08x\n",
+				 __func__, prev, prev->lld.mbr_nda);
+		}
+
+		prev = desc;
+		if (!first)
+			first = desc;
+
+		dev_dbg(chan2dev(chan), "%s: add desc 0x%p to descs_list 0x%p\n",
+			 __func__, desc, first);
+		list_add_tail(&desc->desc_node, &first->descs_list);
+	}
+
+	first->tx_dma_desc.flags = flags;
+	first->xfer_size = len;
+
+	return &first->tx_dma_desc;
+}
+
+static enum dma_status
+at_xdmac_tx_status(struct dma_chan *chan, dma_cookie_t cookie,
+		struct dma_tx_state *txstate)
+{
+	struct at_xdmac_chan	*atchan = to_at_xdmac_chan(chan);
+	struct at_xdmac		*atxdmac = to_at_xdmac(atchan->chan.device);
+	struct at_xdmac_desc	*desc, *_desc;
+	struct list_head	*descs_list;
+	enum dma_status		ret;
+	int			residue;
+	u32			cur_nda, mask, value;
+	u8			dwidth = at_xdmac_get_dwidth(atchan->cfg[AT_XDMAC_CUR_CFG]);
+
+	ret = dma_cookie_status(chan, cookie, txstate);
+	if (ret == DMA_COMPLETE)
+		return ret;
+
+	if (!txstate)
+		return ret;
+
+	spin_lock_bh(&atchan->lock);
+
+	desc = list_first_entry(&atchan->xfers_list, struct at_xdmac_desc, xfer_node);
+
+	/*
+	 * If the transfer has not been started yet, don't need to compute the
+	 * residue, it's the transfer length.
+	 */
+	if (!desc->active_xfer) {
+		dma_set_residue(txstate, desc->xfer_size);
+		spin_unlock_bh(&atchan->lock);
+		return ret;
+	}
+
+	residue = desc->xfer_size;
+	/*
+	 * Flush FIFO: only relevant when the transfer is source peripheral
+	 * synchronized.
+	 */
+	mask = AT_XDMAC_CC_TYPE | AT_XDMAC_CC_DSYNC;
+	value = AT_XDMAC_CC_TYPE_PER_TRAN | AT_XDMAC_CC_DSYNC_PER2MEM;
+	if ((atchan->cfg[AT_XDMAC_CUR_CFG] & mask) == value) {
+		at_xdmac_write(atxdmac, AT_XDMAC_GSWF, atchan->mask);
+		while (!(at_xdmac_chan_read(atchan, AT_XDMAC_CIS) & AT_XDMAC_CIS_FIS))
+			cpu_relax();
+	}
+
+	cur_nda = at_xdmac_chan_read(atchan, AT_XDMAC_CNDA) & 0xfffffffc;
+	/*
+	 * Remove size of all microblocks already transferred and the current
+	 * one. Then add the remaining size to transfer of the current
+	 * microblock.
+	 */
+	descs_list = &desc->descs_list;
+	list_for_each_entry_safe(desc, _desc, descs_list, desc_node) {
+		residue -= (desc->lld.mbr_ubc & 0xffffff) << dwidth;
+		if ((desc->lld.mbr_nda & 0xfffffffc) == cur_nda)
+			break;
+	}
+	residue += at_xdmac_chan_read(atchan, AT_XDMAC_CUBC) << dwidth;
+
+	spin_unlock_bh(&atchan->lock);
+
+	dma_set_residue(txstate, residue);
+
+	dev_dbg(chan2dev(chan),
+		 "%s: desc=0x%p, tx_dma_desc.phys=%pad, tx_status=%d, cookie=%d, residue=%d\n",
+		 __func__, desc, &desc->tx_dma_desc.phys, ret, cookie, residue);
+
+	return ret;
+}
+
+/* Call must be protected by lock. */
+static void at_xdmac_remove_xfer(struct at_xdmac_chan *atchan,
+				    struct at_xdmac_desc *desc)
+{
+	dev_dbg(chan2dev(&atchan->chan), "%s: desc 0x%p\n", __func__, desc);
+
+	/*
+	 * Remove the transfer from the transfer list then move the transfer
+	 * descriptors into the free descriptors list.
+	 */
+	list_del(&desc->xfer_node);
+	list_splice_init(&desc->descs_list, &atchan->free_descs_list);
+}
+
+static void at_xdmac_advance_work(struct at_xdmac_chan *atchan)
+{
+	struct at_xdmac_desc	*desc;
+
+	spin_lock_bh(&atchan->lock);
+
+	/*
+	 * If channel is enabled, do nothing, advance_work will be triggered
+	 * after the interruption.
+	 */
+	if (!at_xdmac_chan_is_enabled(atchan) && !list_empty(&atchan->xfers_list)) {
+		desc = list_first_entry(&atchan->xfers_list,
+					struct at_xdmac_desc,
+					xfer_node);
+		dev_vdbg(chan2dev(&atchan->chan), "%s: desc 0x%p\n", __func__, desc);
+		if (!desc->active_xfer)
+			at_xdmac_start_xfer(atchan, desc);
+	}
+
+	spin_unlock_bh(&atchan->lock);
+}
+
+static void at_xdmac_handle_cyclic(struct at_xdmac_chan *atchan)
+{
+	struct at_xdmac_desc		*desc;
+	struct dma_async_tx_descriptor	*txd;
+
+	desc = list_first_entry(&atchan->xfers_list, struct at_xdmac_desc, xfer_node);
+	txd = &desc->tx_dma_desc;
+
+	if (txd->callback && (txd->flags & DMA_PREP_INTERRUPT))
+		txd->callback(txd->callback_param);
+}
+
+static void at_xdmac_tasklet(unsigned long data)
+{
+	struct at_xdmac_chan	*atchan = (struct at_xdmac_chan *)data;
+	struct at_xdmac_desc	*desc;
+	u32			error_mask;
+
+	dev_dbg(chan2dev(&atchan->chan), "%s: status=0x%08lx\n",
+		 __func__, atchan->status);
+
+	error_mask = AT_XDMAC_CIS_RBEIS
+		     | AT_XDMAC_CIS_WBEIS
+		     | AT_XDMAC_CIS_ROIS;
+
+	if (at_xdmac_chan_is_cyclic(atchan)) {
+		at_xdmac_handle_cyclic(atchan);
+	} else if ((atchan->status & AT_XDMAC_CIS_LIS)
+		   || (atchan->status & error_mask)) {
+		struct dma_async_tx_descriptor  *txd;
+
+		if (atchan->status & AT_XDMAC_CIS_RBEIS)
+			dev_err(chan2dev(&atchan->chan), "read bus error!!!");
+		if (atchan->status & AT_XDMAC_CIS_WBEIS)
+			dev_err(chan2dev(&atchan->chan), "write bus error!!!");
+		if (atchan->status & AT_XDMAC_CIS_ROIS)
+			dev_err(chan2dev(&atchan->chan), "request overflow error!!!");
+
+		spin_lock_bh(&atchan->lock);
+		desc = list_first_entry(&atchan->xfers_list,
+					struct at_xdmac_desc,
+					xfer_node);
+		dev_vdbg(chan2dev(&atchan->chan), "%s: desc 0x%p\n", __func__, desc);
+		BUG_ON(!desc->active_xfer);
+
+		txd = &desc->tx_dma_desc;
+
+		at_xdmac_remove_xfer(atchan, desc);
+		spin_unlock_bh(&atchan->lock);
+
+		if (!at_xdmac_chan_is_cyclic(atchan)) {
+			dma_cookie_complete(txd);
+			if (txd->callback && (txd->flags & DMA_PREP_INTERRUPT))
+				txd->callback(txd->callback_param);
+		}
+
+		dma_run_dependencies(txd);
+
+		at_xdmac_advance_work(atchan);
+	}
+}
+
+static irqreturn_t at_xdmac_interrupt(int irq, void *dev_id)
+{
+	struct at_xdmac		*atxdmac = (struct at_xdmac *)dev_id;
+	struct at_xdmac_chan	*atchan;
+	u32			imr, status, pending;
+	u32			chan_imr, chan_status;
+	int			i, ret = IRQ_NONE;
+
+	do {
+		imr = at_xdmac_read(atxdmac, AT_XDMAC_GIM);
+		status = at_xdmac_read(atxdmac, AT_XDMAC_GIS);
+		pending = status & imr;
+
+		dev_vdbg(atxdmac->dma.dev,
+			 "%s: status=0x%08x, imr=0x%08x, pending=0x%08x\n",
+			 __func__, status, imr, pending);
+
+		if (!pending)
+			break;
+
+		/* We have to find which channel has generated the interrupt. */
+		for (i = 0; i < atxdmac->dma.chancnt; i++) {
+			if (!((1 << i) & pending))
+				continue;
+
+			atchan = &atxdmac->chan[i];
+			chan_imr = at_xdmac_chan_read(atchan, AT_XDMAC_CIM);
+			chan_status = at_xdmac_chan_read(atchan, AT_XDMAC_CIS);
+			atchan->status = chan_status & chan_imr;
+			dev_vdbg(atxdmac->dma.dev,
+				 "%s: chan%d: imr=0x%x, status=0x%x\n",
+				 __func__, i, chan_imr, chan_status);
+			dev_vdbg(chan2dev(&atchan->chan),
+				 "%s: CC=0x%08x CNDA=0x%08x, CNDC=0x%08x, CSA=0x%08x, CDA=0x%08x, CUBC=0x%08x\n",
+				 __func__,
+				 at_xdmac_chan_read(atchan, AT_XDMAC_CC),
+				 at_xdmac_chan_read(atchan, AT_XDMAC_CNDA),
+				 at_xdmac_chan_read(atchan, AT_XDMAC_CNDC),
+				 at_xdmac_chan_read(atchan, AT_XDMAC_CSA),
+				 at_xdmac_chan_read(atchan, AT_XDMAC_CDA),
+				 at_xdmac_chan_read(atchan, AT_XDMAC_CUBC));
+
+			if (atchan->status & (AT_XDMAC_CIS_RBEIS | AT_XDMAC_CIS_WBEIS))
+				at_xdmac_write(atxdmac, AT_XDMAC_GD, atchan->mask);
+
+			tasklet_schedule(&atchan->tasklet);
+			ret = IRQ_HANDLED;
+		}
+
+	} while (pending);
+
+	return ret;
+}
+
+static void at_xdmac_issue_pending(struct dma_chan *chan)
+{
+	struct at_xdmac_chan *atchan = to_at_xdmac_chan(chan);
+
+	dev_dbg(chan2dev(&atchan->chan), "%s\n", __func__);
+
+	if (!at_xdmac_chan_is_cyclic(atchan))
+		at_xdmac_advance_work(atchan);
+
+	return;
+}
+
+static int at_xdmac_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
+			    unsigned long arg)
+{
+	struct at_xdmac_desc	*desc, *_desc;
+	struct at_xdmac_chan	*atchan = to_at_xdmac_chan(chan);
+	struct at_xdmac		*atxdmac = to_at_xdmac(atchan->chan.device);
+	int			ret = 0;
+
+	dev_dbg(chan2dev(chan), "%s: cmd=%d\n", __func__, cmd);
+
+	spin_lock_bh(&atchan->lock);
+
+	switch (cmd) {
+	case DMA_PAUSE:
+		at_xdmac_write(atxdmac, AT_XDMAC_GRWS, atchan->mask);
+		set_bit(AT_XDMAC_CHAN_IS_PAUSED, &atchan->status);
+		break;
+
+	case DMA_RESUME:
+		if (!at_xdmac_chan_is_paused(atchan))
+			break;
+
+		at_xdmac_write(atxdmac, AT_XDMAC_GRWR, atchan->mask);
+		clear_bit(AT_XDMAC_CHAN_IS_PAUSED, &atchan->status);
+		break;
+
+	case DMA_TERMINATE_ALL:
+		at_xdmac_write(atxdmac, AT_XDMAC_GD, atchan->mask);
+		while (at_xdmac_read(atxdmac, AT_XDMAC_GS) & atchan->mask)
+			cpu_relax();
+
+		/* Cancel all pending transfers. */
+		list_for_each_entry_safe(desc, _desc, &atchan->xfers_list, xfer_node)
+			at_xdmac_remove_xfer(atchan, desc);
+
+		clear_bit(AT_XDMAC_CHAN_IS_CYCLIC, &atchan->status);
+		break;
+
+	case DMA_SLAVE_CONFIG:
+		ret = at_xdmac_set_slave_config(chan,
+				(struct dma_slave_config *)arg);
+		break;
+
+	default:
+		dev_err(chan2dev(chan),
+			"unmanaged or unknown dma control cmd: %d\n", cmd);
+		ret = -ENXIO;
+	}
+
+	spin_unlock_bh(&atchan->lock);
+
+	return ret;
+}
+
+static int at_xdmac_alloc_chan_resources(struct dma_chan *chan)
+{
+	struct at_xdmac_chan	*atchan = to_at_xdmac_chan(chan);
+	struct at_xdmac_desc	*desc;
+	int			i;
+
+	spin_lock_bh(&atchan->lock);
+
+	if (at_xdmac_chan_is_enabled(atchan)) {
+		dev_err(chan2dev(chan),
+			"can't allocate channel resources (channel enabled)\n");
+		i = -EIO;
+		goto spin_unlock;
+	}
+
+	if (!list_empty(&atchan->free_descs_list)) {
+		dev_err(chan2dev(chan),
+			"can't allocate channel resources (channel not free from a previous use)\n");
+		i = -EIO;
+		goto spin_unlock;
+	}
+
+	for (i = 0; i < init_nr_desc_per_channel; i++) {
+		desc = at_xdmac_alloc_desc(chan, GFP_ATOMIC);
+		if (!desc) {
+			dev_warn(chan2dev(chan),
+				"only %d descriptors have been allocated\n", i);
+			break;
+		}
+		list_add_tail(&desc->desc_node, &atchan->free_descs_list);
+	}
+
+	dma_cookie_init(chan);
+
+	dev_dbg(chan2dev(chan), "%s: allocated %d descriptors\n", __func__, i);
+
+spin_unlock:
+	spin_unlock_bh(&atchan->lock);
+	return i;
+}
+
+static void at_xdmac_free_chan_resources(struct dma_chan *chan)
+{
+	struct at_xdmac_chan	*atchan = to_at_xdmac_chan(chan);
+	struct at_xdmac		*atxdmac = to_at_xdmac(chan->device);
+	struct at_xdmac_desc	*desc, *_desc;
+
+	list_for_each_entry_safe(desc, _desc, &atchan->free_descs_list, desc_node) {
+		dev_dbg(chan2dev(chan), "%s: freeing descriptor %p\n", __func__, desc);
+		list_del(&desc->desc_node);
+		dma_pool_free(atxdmac->at_xdmac_desc_pool, desc, desc->tx_dma_desc.phys);
+	}
+
+	return;
+}
+
+#define AT_XDMAC_DMA_BUSWIDTHS\
+	(BIT(DMA_SLAVE_BUSWIDTH_UNDEFINED) |\
+	BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) |\
+	BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) |\
+	BIT(DMA_SLAVE_BUSWIDTH_4_BYTES) |\
+	BIT(DMA_SLAVE_BUSWIDTH_8_BYTES))
+
+static int at_xdmac_device_slave_caps(struct dma_chan *dchan,
+				      struct dma_slave_caps *caps)
+{
+
+	caps->src_addr_widths = AT_XDMAC_DMA_BUSWIDTHS;
+	caps->dstn_addr_widths = AT_XDMAC_DMA_BUSWIDTHS;
+	caps->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
+	caps->cmd_pause = true;
+	caps->cmd_terminate = true;
+	caps->residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
+
+	return 0;
+}
+
+#ifdef CONFIG_PM
+static int atmel_xdmac_prepare(struct device *dev)
+{
+	struct platform_device	*pdev = to_platform_device(dev);
+	struct at_xdmac		*atxdmac = platform_get_drvdata(pdev);
+	struct dma_chan		*chan, *_chan;
+
+	list_for_each_entry_safe(chan, _chan, &atxdmac->dma.channels, device_node) {
+		struct at_xdmac_chan	*atchan = to_at_xdmac_chan(chan);
+
+		/* Wait for transfer completion, except in cyclic case. */
+		if (at_xdmac_chan_is_enabled(atchan) && !at_xdmac_chan_is_cyclic(atchan))
+			return -EAGAIN;
+	}
+	return 0;
+}
+#else
+#	define atmel_xdmac_prepare NULL
+#endif
+
+#ifdef CONFIG_PM_SLEEP
+static int atmel_xdmac_suspend(struct device *dev)
+{
+	struct platform_device	*pdev = to_platform_device(dev);
+	struct at_xdmac		*atxdmac = platform_get_drvdata(pdev);
+	struct dma_chan		*chan, *_chan;
+
+	list_for_each_entry_safe(chan, _chan, &atxdmac->dma.channels, device_node) {
+		struct at_xdmac_chan	*atchan = to_at_xdmac_chan(chan);
+
+		if (at_xdmac_chan_is_cyclic(atchan)) {
+			if (!at_xdmac_chan_is_paused(atchan))
+				at_xdmac_control(chan, DMA_PAUSE, 0);
+			atchan->save_cim = at_xdmac_chan_read(atchan, AT_XDMAC_CIM);
+			atchan->save_cnda = at_xdmac_chan_read(atchan, AT_XDMAC_CNDA);
+			atchan->save_cndc = at_xdmac_chan_read(atchan, AT_XDMAC_CNDC);
+		}
+	}
+	atxdmac->save_gim = at_xdmac_read(atxdmac, AT_XDMAC_GIM);
+
+	at_xdmac_off(atxdmac);
+	clk_disable_unprepare(atxdmac->clk);
+	return 0;
+}
+
+static int atmel_xdmac_resume(struct device *dev)
+{
+	struct platform_device	*pdev = to_platform_device(dev);
+	struct at_xdmac		*atxdmac = platform_get_drvdata(pdev);
+	struct at_xdmac_chan	*atchan;
+	struct dma_chan		*chan, *_chan;
+	int			i;
+	u32			cfg;
+
+	clk_prepare_enable(atxdmac->clk);
+
+	/* Clear pending interrupts. */
+	for (i = 0; i < atxdmac->dma.chancnt; i++) {
+		atchan = &atxdmac->chan[i];
+		while (at_xdmac_chan_read(atchan, AT_XDMAC_CIS))
+			cpu_relax();
+	}
+
+	at_xdmac_write(atxdmac, AT_XDMAC_GIE, atxdmac->save_gim);
+	at_xdmac_write(atxdmac, AT_XDMAC_GE, atxdmac->save_gs);
+	list_for_each_entry_safe(chan, _chan, &atxdmac->dma.channels, device_node) {
+		atchan = to_at_xdmac_chan(chan);
+		cfg = atchan->cfg[AT_XDMAC_CUR_CFG];
+		at_xdmac_chan_write(atchan, AT_XDMAC_CC, cfg);
+		if (at_xdmac_chan_is_cyclic(atchan)) {
+			at_xdmac_chan_write(atchan, AT_XDMAC_CNDA, atchan->save_cnda);
+			at_xdmac_chan_write(atchan, AT_XDMAC_CNDC, atchan->save_cndc);
+			at_xdmac_chan_write(atchan, AT_XDMAC_CIE, atchan->save_cim);
+			wmb();
+			at_xdmac_write(atxdmac, AT_XDMAC_GE, atchan->mask);
+		}
+	}
+	return 0;
+}
+#endif /* CONFIG_PM_SLEEP */
+
+static int at_xdmac_probe(struct platform_device *pdev)
+{
+	struct resource	*res;
+	struct at_xdmac	*atxdmac;
+	int		irq, size, nr_channels, i, ret;
+	void __iomem	*base;
+	u32		reg;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res)
+		return -EINVAL;
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0)
+		return irq;
+
+	base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(base))
+		return PTR_ERR(base);
+
+	/*
+	 * Read number of xdmac channels, read helper function can't be used
+	 * since atxdmac is not yet allocated and we need to know the number
+	 * of channels to do the allocation.
+	 */
+	reg = readl_relaxed(base + AT_XDMAC_GTYPE);
+	nr_channels = AT_XDMAC_NB_CH(reg);
+	if (nr_channels > AT_XDMAC_MAX_CHAN) {
+		dev_err(&pdev->dev, "invalid number of channels (%u)\n",
+			nr_channels);
+		return -EINVAL;
+	}
+
+	size = sizeof(*atxdmac);
+	size += nr_channels * sizeof(struct at_xdmac_chan);
+	atxdmac = devm_kzalloc(&pdev->dev, size, GFP_KERNEL);
+	if (!atxdmac) {
+		dev_err(&pdev->dev, "can't allocate at_xdmac structure\n");
+		return -ENOMEM;
+	}
+
+	atxdmac->regs = base;
+	atxdmac->irq = irq;
+
+	atxdmac->clk = devm_clk_get(&pdev->dev, "dma_clk");
+	if (IS_ERR(atxdmac->clk)) {
+		dev_err(&pdev->dev, "can't get dma_clk\n");
+		return PTR_ERR(atxdmac->clk);
+	}
+
+	/* Do not use dev res to prevent races with tasklet */
+	ret = request_irq(atxdmac->irq, at_xdmac_interrupt, 0, "at_xdmac", atxdmac);
+	if (ret) {
+		dev_err(&pdev->dev, "can't request irq\n");
+		return ret;
+	}
+
+	ret = clk_prepare_enable(atxdmac->clk);
+	if (ret) {
+		dev_err(&pdev->dev, "can't prepare or enable clock\n");
+		goto err_free_irq;
+	}
+
+	atxdmac->at_xdmac_desc_pool =
+		dmam_pool_create(dev_name(&pdev->dev), &pdev->dev,
+				sizeof(struct at_xdmac_desc), 4, 0);
+	if (!atxdmac->at_xdmac_desc_pool) {
+		dev_err(&pdev->dev, "no memory for descriptors dma pool\n");
+		ret = -ENOMEM;
+		goto err_clk_disable;
+	}
+
+	dma_cap_set(DMA_CYCLIC, atxdmac->dma.cap_mask);
+	dma_cap_set(DMA_MEMCPY, atxdmac->dma.cap_mask);
+	dma_cap_set(DMA_SLAVE, atxdmac->dma.cap_mask);
+	/*
+	 * Without DMA_PRIVATE the driver is not able to allocate more than
+	 * one channel, second allocation fails in private_candidate.
+	 */
+	dma_cap_set(DMA_PRIVATE, atxdmac->dma.cap_mask);
+	atxdmac->dma.dev				= &pdev->dev;
+	atxdmac->dma.device_alloc_chan_resources	= at_xdmac_alloc_chan_resources;
+	atxdmac->dma.device_free_chan_resources		= at_xdmac_free_chan_resources;
+	atxdmac->dma.device_tx_status			= at_xdmac_tx_status;
+	atxdmac->dma.device_issue_pending		= at_xdmac_issue_pending;
+	atxdmac->dma.device_prep_dma_cyclic		= at_xdmac_prep_dma_cyclic;
+	atxdmac->dma.device_prep_dma_memcpy		= at_xdmac_prep_dma_memcpy;
+	atxdmac->dma.device_prep_slave_sg		= at_xdmac_prep_slave_sg;
+	atxdmac->dma.device_control			= at_xdmac_control;
+	atxdmac->dma.device_slave_caps			= at_xdmac_device_slave_caps;
+
+	/* Disable all chans and interrupts. */
+	at_xdmac_off(atxdmac);
+
+	/* Init channels. */
+	INIT_LIST_HEAD(&atxdmac->dma.channels);
+	for (i = 0; i < nr_channels; i++) {
+		struct at_xdmac_chan *atchan = &atxdmac->chan[i];
+
+		atchan->chan.device = &atxdmac->dma;
+		list_add_tail(&atchan->chan.device_node,
+			      &atxdmac->dma.channels);
+
+		atchan->ch_regs = at_xdmac_chan_reg_base(atxdmac, i);
+		atchan->mask = 1 << i;
+
+		spin_lock_init(&atchan->lock);
+		INIT_LIST_HEAD(&atchan->xfers_list);
+		INIT_LIST_HEAD(&atchan->free_descs_list);
+		tasklet_init(&atchan->tasklet, at_xdmac_tasklet,
+			     (unsigned long)atchan);
+
+		/* Clear pending interrupts. */
+		while (at_xdmac_chan_read(atchan, AT_XDMAC_CIS))
+			cpu_relax();
+	}
+	platform_set_drvdata(pdev, atxdmac);
+
+	ret = dma_async_device_register(&atxdmac->dma);
+	if (ret) {
+		dev_err(&pdev->dev, "fail to register DMA engine device\n");
+		goto err_clk_disable;
+	}
+
+	ret = of_dma_controller_register(pdev->dev.of_node,
+					 at_xdmac_xlate, atxdmac);
+	if (ret) {
+		dev_err(&pdev->dev, "could not register of dma controller\n");
+		goto err_dma_unregister;
+	}
+
+	dev_info(&pdev->dev, "%d channels, mapped at 0x%p\n",
+		 nr_channels, atxdmac->regs);
+
+	return 0;
+
+err_dma_unregister:
+	dma_async_device_unregister(&atxdmac->dma);
+err_clk_disable:
+	clk_disable_unprepare(atxdmac->clk);
+err_free_irq:
+	free_irq(atxdmac->irq, atxdmac->dma.dev);
+	return ret;
+}
+
+static int at_xdmac_remove(struct platform_device *pdev)
+{
+	struct at_xdmac	*atxdmac = (struct at_xdmac *)platform_get_drvdata(pdev);
+	int		i;
+
+	at_xdmac_off(atxdmac);
+	of_dma_controller_free(pdev->dev.of_node);
+	dma_async_device_unregister(&atxdmac->dma);
+	clk_disable_unprepare(atxdmac->clk);
+
+	synchronize_irq(atxdmac->irq);
+
+	free_irq(atxdmac->irq, atxdmac->dma.dev);
+
+	for (i = 0; i < atxdmac->dma.chancnt; i++) {
+		struct at_xdmac_chan *atchan = &atxdmac->chan[i];
+
+		tasklet_kill(&atchan->tasklet);
+		at_xdmac_free_chan_resources(&atchan->chan);
+	}
+
+	return 0;
+}
+
+static const struct dev_pm_ops atmel_xdmac_dev_pm_ops = {
+	.prepare	= atmel_xdmac_prepare,
+	SET_LATE_SYSTEM_SLEEP_PM_OPS(atmel_xdmac_suspend, atmel_xdmac_resume)
+};
+
+static const struct of_device_id atmel_xdmac_dt_ids[] = {
+	{
+		.compatible = "atmel,sama5d4-dma",
+	}, {
+		/* sentinel */
+	}
+};
+MODULE_DEVICE_TABLE(of, atmel_xdmac_dt_ids);
+
+static struct platform_driver at_xdmac_driver = {
+	.probe		= at_xdmac_probe,
+	.remove		= at_xdmac_remove,
+	.driver = {
+		.name		= "at_xdmac",
+		.owner		= THIS_MODULE,
+		.of_match_table	= of_match_ptr(atmel_xdmac_dt_ids),
+		.pm		= &atmel_xdmac_dev_pm_ops,
+	}
+};
+
+static int __init at_xdmac_init(void)
+{
+	return platform_driver_probe(&at_xdmac_driver, at_xdmac_probe);
+}
+subsys_initcall(at_xdmac_init);
+
+MODULE_DESCRIPTION("Atmel Extended DMA Controller driver");
+MODULE_AUTHOR("Ludovic Desroches <ludovic.desroches@atmel.com>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/dma/bcm2835-dma.c b/drivers/dma/bcm2835-dma.c
index 68007974961a..918b7b3f766f 100644
--- a/drivers/dma/bcm2835-dma.c
+++ b/drivers/dma/bcm2835-dma.c
@@ -525,8 +525,6 @@ static int bcm2835_dma_chan_init(struct bcm2835_dmadev *d, int chan_id, int irq)
 	vchan_init(&c->vc, &d->ddev);
 	INIT_LIST_HEAD(&c->node);
 
-	d->ddev.chancnt++;
-
 	c->chan_base = BCM2835_DMA_CHANIO(d->base, chan_id);
 	c->ch = chan_id;
 	c->irq_number = irq;
@@ -694,7 +692,6 @@ static struct platform_driver bcm2835_dma_driver = {
 	.remove	= bcm2835_dma_remove,
 	.driver = {
 		.name = "bcm2835-dma",
-		.owner = THIS_MODULE,
 		.of_match_table = of_match_ptr(bcm2835_dma_of_match),
 	},
 };
diff --git a/drivers/dma/bestcomm/bestcomm.c b/drivers/dma/bestcomm/bestcomm.c
index a8c2e2994d2e..fa378d88f6c8 100644
--- a/drivers/dma/bestcomm/bestcomm.c
+++ b/drivers/dma/bestcomm/bestcomm.c
@@ -495,7 +495,6 @@ static struct platform_driver mpc52xx_bcom_of_platform_driver = {
 	.remove		= mpc52xx_bcom_remove,
 	.driver = {
 		.name = DRIVER_NAME,
-		.owner = THIS_MODULE,
 		.of_match_table = mpc52xx_bcom_of_match,
 	},
 };
diff --git a/drivers/dma/coh901318.c b/drivers/dma/coh901318.c
index 3c6716e0b78e..e88588d8ecd3 100644
--- a/drivers/dma/coh901318.c
+++ b/drivers/dma/coh901318.c
@@ -2156,7 +2156,7 @@ coh901318_free_chan_resources(struct dma_chan *chan)
 
 	spin_unlock_irqrestore(&cohc->lock, flags);
 
-	chan->device->device_control(chan, DMA_TERMINATE_ALL, 0);
+	dmaengine_terminate_all(chan);
 }
 
 
diff --git a/drivers/dma/cppi41.c b/drivers/dma/cppi41.c
index 8f8b0b608875..b743adf56465 100644
--- a/drivers/dma/cppi41.c
+++ b/drivers/dma/cppi41.c
@@ -1,3 +1,4 @@
+#include <linux/delay.h>
 #include <linux/dmaengine.h>
 #include <linux/dma-mapping.h>
 #include <linux/platform_device.h>
@@ -567,7 +568,7 @@ static int cppi41_tear_down_chan(struct cppi41_channel *c)
 		reg |= GCR_TEARDOWN;
 		cppi_writel(reg, c->gcr_reg);
 		c->td_queued = 1;
-		c->td_retry = 100;
+		c->td_retry = 500;
 	}
 
 	if (!c->td_seen || !c->td_desc_seen) {
@@ -603,12 +604,16 @@ static int cppi41_tear_down_chan(struct cppi41_channel *c)
 	 * descriptor before the TD we fetch it from enqueue, it has to be
 	 * there waiting for us.
 	 */
-	if (!c->td_seen && c->td_retry)
+	if (!c->td_seen && c->td_retry) {
+		udelay(1);
 		return -EAGAIN;
-
+	}
 	WARN_ON(!c->td_retry);
+
 	if (!c->td_desc_seen) {
 		desc_phys = cppi41_pop_desc(cdd, c->q_num);
+		if (!desc_phys)
+			desc_phys = cppi41_pop_desc(cdd, c->q_comp_num);
 		WARN_ON(!desc_phys);
 	}
 
@@ -938,7 +943,7 @@ static int cppi41_dma_probe(struct platform_device *pdev)
 	if (!glue_info)
 		return -EINVAL;
 
-	cdd = kzalloc(sizeof(*cdd), GFP_KERNEL);
+	cdd = devm_kzalloc(&pdev->dev, sizeof(*cdd), GFP_KERNEL);
 	if (!cdd)
 		return -ENOMEM;
 
@@ -959,10 +964,8 @@ static int cppi41_dma_probe(struct platform_device *pdev)
 	cdd->qmgr_mem = of_iomap(dev->of_node, 3);
 
 	if (!cdd->usbss_mem || !cdd->ctrl_mem || !cdd->sched_mem ||
-			!cdd->qmgr_mem) {
-		ret = -ENXIO;
-		goto err_remap;
-	}
+			!cdd->qmgr_mem)
+		return -ENXIO;
 
 	pm_runtime_enable(dev);
 	ret = pm_runtime_get_sync(dev);
@@ -989,7 +992,7 @@ static int cppi41_dma_probe(struct platform_device *pdev)
 
 	cppi_writel(USBSS_IRQ_PD_COMP, cdd->usbss_mem + USBSS_IRQ_ENABLER);
 
-	ret = request_irq(irq, glue_info->isr, IRQF_SHARED,
+	ret = devm_request_irq(&pdev->dev, irq, glue_info->isr, IRQF_SHARED,
 			dev_name(dev), cdd);
 	if (ret)
 		goto err_irq;
@@ -1009,7 +1012,6 @@ static int cppi41_dma_probe(struct platform_device *pdev)
 err_of:
 	dma_async_device_unregister(&cdd->ddev);
 err_dma_reg:
-	free_irq(irq, cdd);
 err_irq:
 	cppi_writel(0, cdd->usbss_mem + USBSS_IRQ_CLEARR);
 	cleanup_chans(cdd);
@@ -1023,8 +1025,6 @@ err_get_sync:
 	iounmap(cdd->ctrl_mem);
 	iounmap(cdd->sched_mem);
 	iounmap(cdd->qmgr_mem);
-err_remap:
-	kfree(cdd);
 	return ret;
 }
 
@@ -1036,7 +1036,7 @@ static int cppi41_dma_remove(struct platform_device *pdev)
 	dma_async_device_unregister(&cdd->ddev);
 
 	cppi_writel(0, cdd->usbss_mem + USBSS_IRQ_CLEARR);
-	free_irq(cdd->irq, cdd);
+	devm_free_irq(&pdev->dev, cdd->irq, cdd);
 	cleanup_chans(cdd);
 	deinit_cppi41(&pdev->dev, cdd);
 	iounmap(cdd->usbss_mem);
@@ -1045,7 +1045,6 @@ static int cppi41_dma_remove(struct platform_device *pdev)
 	iounmap(cdd->qmgr_mem);
 	pm_runtime_put(&pdev->dev);
 	pm_runtime_disable(&pdev->dev);
-	kfree(cdd);
 	return 0;
 }
 
@@ -1094,7 +1093,6 @@ static struct platform_driver cpp41_dma_driver = {
 	.remove = cppi41_dma_remove,
 	.driver = {
 		.name = "cppi41-dma-engine",
-		.owner = THIS_MODULE,
 		.pm = &cppi41_pm_ops,
 		.of_match_table = of_match_ptr(cppi41_dma_ids),
 	},
diff --git a/drivers/dma/dma-jz4740.c b/drivers/dma/dma-jz4740.c
index ae2ab14e64b3..bdeafeefa5f6 100644
--- a/drivers/dma/dma-jz4740.c
+++ b/drivers/dma/dma-jz4740.c
@@ -563,10 +563,9 @@ static int jz4740_dma_probe(struct platform_device *pdev)
 	dd->device_prep_dma_cyclic = jz4740_dma_prep_dma_cyclic;
 	dd->device_control = jz4740_dma_control;
 	dd->dev = &pdev->dev;
-	dd->chancnt = JZ_DMA_NR_CHANS;
 	INIT_LIST_HEAD(&dd->channels);
 
-	for (i = 0; i < dd->chancnt; i++) {
+	for (i = 0; i < JZ_DMA_NR_CHANS; i++) {
 		chan = &dmadev->chan[i];
 		chan->id = i;
 		chan->vchan.desc_free = jz4740_dma_desc_free;
@@ -608,7 +607,6 @@ static struct platform_driver jz4740_dma_driver = {
 	.remove = jz4740_dma_remove,
 	.driver = {
 		.name = "jz4740-dma",
-		.owner = THIS_MODULE,
 	},
 };
 module_platform_driver(jz4740_dma_driver);
diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
index d5d30ed863ce..e057935e3023 100644
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -330,8 +330,7 @@ static int __init dma_channel_table_init(void)
 	if (err) {
 		pr_err("initialization failure\n");
 		for_each_dma_cap_mask(cap, dma_cap_mask_all)
-			if (channel_table[cap])
-				free_percpu(channel_table[cap]);
+			free_percpu(channel_table[cap]);
 	}
 
 	return err;
@@ -1081,110 +1080,6 @@ dmaengine_get_unmap_data(struct device *dev, int nr, gfp_t flags)
 }
 EXPORT_SYMBOL(dmaengine_get_unmap_data);
 
-/**
- * dma_async_memcpy_pg_to_pg - offloaded copy from page to page
- * @chan: DMA channel to offload copy to
- * @dest_pg: destination page
- * @dest_off: offset in page to copy to
- * @src_pg: source page
- * @src_off: offset in page to copy from
- * @len: length
- *
- * Both @dest_page/@dest_off and @src_page/@src_off must be mappable to a bus
- * address according to the DMA mapping API rules for streaming mappings.
- * Both @dest_page/@dest_off and @src_page/@src_off must stay memory resident
- * (kernel memory or locked user space pages).
- */
-dma_cookie_t
-dma_async_memcpy_pg_to_pg(struct dma_chan *chan, struct page *dest_pg,
-	unsigned int dest_off, struct page *src_pg, unsigned int src_off,
-	size_t len)
-{
-	struct dma_device *dev = chan->device;
-	struct dma_async_tx_descriptor *tx;
-	struct dmaengine_unmap_data *unmap;
-	dma_cookie_t cookie;
-	unsigned long flags;
-
-	unmap = dmaengine_get_unmap_data(dev->dev, 2, GFP_NOWAIT);
-	if (!unmap)
-		return -ENOMEM;
-
-	unmap->to_cnt = 1;
-	unmap->from_cnt = 1;
-	unmap->addr[0] = dma_map_page(dev->dev, src_pg, src_off, len,
-				      DMA_TO_DEVICE);
-	unmap->addr[1] = dma_map_page(dev->dev, dest_pg, dest_off, len,
-				      DMA_FROM_DEVICE);
-	unmap->len = len;
-	flags = DMA_CTRL_ACK;
-	tx = dev->device_prep_dma_memcpy(chan, unmap->addr[1], unmap->addr[0],
-					 len, flags);
-
-	if (!tx) {
-		dmaengine_unmap_put(unmap);
-		return -ENOMEM;
-	}
-
-	dma_set_unmap(tx, unmap);
-	cookie = tx->tx_submit(tx);
-	dmaengine_unmap_put(unmap);
-
-	preempt_disable();
-	__this_cpu_add(chan->local->bytes_transferred, len);
-	__this_cpu_inc(chan->local->memcpy_count);
-	preempt_enable();
-
-	return cookie;
-}
-EXPORT_SYMBOL(dma_async_memcpy_pg_to_pg);
-
-/**
- * dma_async_memcpy_buf_to_buf - offloaded copy between virtual addresses
- * @chan: DMA channel to offload copy to
- * @dest: destination address (virtual)
- * @src: source address (virtual)
- * @len: length
- *
- * Both @dest and @src must be mappable to a bus address according to the
- * DMA mapping API rules for streaming mappings.
- * Both @dest and @src must stay memory resident (kernel memory or locked
- * user space pages).
- */
-dma_cookie_t
-dma_async_memcpy_buf_to_buf(struct dma_chan *chan, void *dest,
-			    void *src, size_t len)
-{
-	return dma_async_memcpy_pg_to_pg(chan, virt_to_page(dest),
-					 (unsigned long) dest & ~PAGE_MASK,
-					 virt_to_page(src),
-					 (unsigned long) src & ~PAGE_MASK, len);
-}
-EXPORT_SYMBOL(dma_async_memcpy_buf_to_buf);
-
-/**
- * dma_async_memcpy_buf_to_pg - offloaded copy from address to page
- * @chan: DMA channel to offload copy to
- * @page: destination page
- * @offset: offset in page to copy to
- * @kdata: source address (virtual)
- * @len: length
- *
- * Both @page/@offset and @kdata must be mappable to a bus address according
- * to the DMA mapping API rules for streaming mappings.
- * Both @page/@offset and @kdata must stay memory resident (kernel memory or
- * locked user space pages)
- */
-dma_cookie_t
-dma_async_memcpy_buf_to_pg(struct dma_chan *chan, struct page *page,
-			   unsigned int offset, void *kdata, size_t len)
-{
-	return dma_async_memcpy_pg_to_pg(chan, page, offset,
-					 virt_to_page(kdata),
-					 (unsigned long) kdata & ~PAGE_MASK, len);
-}
-EXPORT_SYMBOL(dma_async_memcpy_buf_to_pg);
-
 void dma_async_tx_descriptor_init(struct dma_async_tx_descriptor *tx,
 	struct dma_chan *chan)
 {
diff --git a/drivers/dma/dmatest.c b/drivers/dma/dmatest.c
index e27cec25c59e..a8d7809e2f4c 100644
--- a/drivers/dma/dmatest.c
+++ b/drivers/dma/dmatest.c
@@ -688,14 +688,14 @@ static int dmatest_func(void *data)
 	runtime = ktime_us_delta(ktime_get(), ktime);
 
 	ret = 0;
+err_dstbuf:
 	for (i = 0; thread->dsts[i]; i++)
 		kfree(thread->dsts[i]);
-err_dstbuf:
 	kfree(thread->dsts);
 err_dsts:
+err_srcbuf:
 	for (i = 0; thread->srcs[i]; i++)
 		kfree(thread->srcs[i]);
-err_srcbuf:
 	kfree(thread->srcs);
 err_srcs:
 	kfree(pq_coefs);
diff --git a/drivers/dma/dw/core.c b/drivers/dma/dw/core.c
index 1af731b83b3f..5c062548957c 100644
--- a/drivers/dma/dw/core.c
+++ b/drivers/dma/dw/core.c
@@ -11,7 +11,6 @@
  */
 
 #include <linux/bitops.h>
-#include <linux/clk.h>
 #include <linux/delay.h>
 #include <linux/dmaengine.h>
 #include <linux/dma-mapping.h>
@@ -23,6 +22,7 @@
 #include <linux/mm.h>
 #include <linux/module.h>
 #include <linux/slab.h>
+#include <linux/pm_runtime.h>
 
 #include "../dmaengine.h"
 #include "internal.h"
@@ -37,24 +37,6 @@
  * support descriptor writeback.
  */
 
-static inline bool is_request_line_unset(struct dw_dma_chan *dwc)
-{
-	return dwc->request_line == (typeof(dwc->request_line))~0;
-}
-
-static inline void dwc_set_masters(struct dw_dma_chan *dwc)
-{
-	struct dw_dma *dw = to_dw_dma(dwc->chan.device);
-	struct dw_dma_slave *dws = dwc->chan.private;
-	unsigned char mmax = dw->nr_masters - 1;
-
-	if (!is_request_line_unset(dwc))
-		return;
-
-	dwc->src_master = min_t(unsigned char, mmax, dwc_get_sms(dws));
-	dwc->dst_master = min_t(unsigned char, mmax, dwc_get_dms(dws));
-}
-
 #define DWC_DEFAULT_CTLLO(_chan) ({				\
 		struct dw_dma_chan *_dwc = to_dw_dma_chan(_chan);	\
 		struct dma_slave_config	*_sconfig = &_dwc->dma_sconfig;	\
@@ -155,13 +137,11 @@ static void dwc_initialize(struct dw_dma_chan *dwc)
 		 */
 		BUG_ON(!dws->dma_dev || dws->dma_dev != dw->dma.dev);
 
-		cfghi = dws->cfg_hi;
-		cfglo |= dws->cfg_lo & ~DWC_CFGL_CH_PRIOR_MASK;
+		cfghi |= DWC_CFGH_DST_PER(dws->dst_id);
+		cfghi |= DWC_CFGH_SRC_PER(dws->src_id);
 	} else {
-		if (dwc->direction == DMA_MEM_TO_DEV)
-			cfghi = DWC_CFGH_DST_PER(dwc->request_line);
-		else if (dwc->direction == DMA_DEV_TO_MEM)
-			cfghi = DWC_CFGH_SRC_PER(dwc->request_line);
+		cfghi |= DWC_CFGH_DST_PER(dwc->dst_id);
+		cfghi |= DWC_CFGH_SRC_PER(dwc->src_id);
 	}
 
 	channel_writel(dwc, CFG_LO, cfglo);
@@ -939,6 +919,26 @@ err_desc_get:
 	return NULL;
 }
 
+bool dw_dma_filter(struct dma_chan *chan, void *param)
+{
+	struct dw_dma_chan *dwc = to_dw_dma_chan(chan);
+	struct dw_dma_slave *dws = param;
+
+	if (!dws || dws->dma_dev != chan->device->dev)
+		return false;
+
+	/* We have to copy data since dws can be temporary storage */
+
+	dwc->src_id = dws->src_id;
+	dwc->dst_id = dws->dst_id;
+
+	dwc->src_master = dws->src_master;
+	dwc->dst_master = dws->dst_master;
+
+	return true;
+}
+EXPORT_SYMBOL_GPL(dw_dma_filter);
+
 /*
  * Fix sconfig's burst size according to dw_dmac. We need to convert them as:
  * 1 -> 0, 4 -> 1, 8 -> 2, 16 -> 3.
@@ -967,10 +967,6 @@ set_runtime_config(struct dma_chan *chan, struct dma_slave_config *sconfig)
 	memcpy(&dwc->dma_sconfig, sconfig, sizeof(*sconfig));
 	dwc->direction = sconfig->direction;
 
-	/* Take the request line from slave_id member */
-	if (is_request_line_unset(dwc))
-		dwc->request_line = sconfig->slave_id;
-
 	convert_burst(&dwc->dma_sconfig.src_maxburst);
 	convert_burst(&dwc->dma_sconfig.dst_maxburst);
 
@@ -1099,6 +1095,31 @@ static void dwc_issue_pending(struct dma_chan *chan)
 	spin_unlock_irqrestore(&dwc->lock, flags);
 }
 
+/*----------------------------------------------------------------------*/
+
+static void dw_dma_off(struct dw_dma *dw)
+{
+	int i;
+
+	dma_writel(dw, CFG, 0);
+
+	channel_clear_bit(dw, MASK.XFER, dw->all_chan_mask);
+	channel_clear_bit(dw, MASK.SRC_TRAN, dw->all_chan_mask);
+	channel_clear_bit(dw, MASK.DST_TRAN, dw->all_chan_mask);
+	channel_clear_bit(dw, MASK.ERROR, dw->all_chan_mask);
+
+	while (dma_readl(dw, CFG) & DW_CFG_DMA_EN)
+		cpu_relax();
+
+	for (i = 0; i < dw->dma.chancnt; i++)
+		dw->chan[i].initialized = false;
+}
+
+static void dw_dma_on(struct dw_dma *dw)
+{
+	dma_writel(dw, CFG, DW_CFG_DMA_EN);
+}
+
 static int dwc_alloc_chan_resources(struct dma_chan *chan)
 {
 	struct dw_dma_chan	*dwc = to_dw_dma_chan(chan);
@@ -1123,7 +1144,10 @@ static int dwc_alloc_chan_resources(struct dma_chan *chan)
 	 * doesn't mean what you think it means), and status writeback.
 	 */
 
-	dwc_set_masters(dwc);
+	/* Enable controller here if needed */
+	if (!dw->in_use)
+		dw_dma_on(dw);
+	dw->in_use |= dwc->mask;
 
 	spin_lock_irqsave(&dwc->lock, flags);
 	i = dwc->descs_allocated;
@@ -1182,7 +1206,6 @@ static void dwc_free_chan_resources(struct dma_chan *chan)
 	list_splice_init(&dwc->free_list, &list);
 	dwc->descs_allocated = 0;
 	dwc->initialized = false;
-	dwc->request_line = ~0;
 
 	/* Disable interrupts */
 	channel_clear_bit(dw, MASK.XFER, dwc->mask);
@@ -1190,6 +1213,11 @@ static void dwc_free_chan_resources(struct dma_chan *chan)
 
 	spin_unlock_irqrestore(&dwc->lock, flags);
 
+	/* Disable controller in case it was a last user */
+	dw->in_use &= ~dwc->mask;
+	if (!dw->in_use)
+		dw_dma_off(dw);
+
 	list_for_each_entry_safe(desc, _desc, &list, desc_node) {
 		dev_vdbg(chan2dev(chan), "  freeing descriptor %p\n", desc);
 		dma_pool_free(dw->desc_pool, desc, desc->txd.phys);
@@ -1460,24 +1488,6 @@ EXPORT_SYMBOL(dw_dma_cyclic_free);
 
 /*----------------------------------------------------------------------*/
 
-static void dw_dma_off(struct dw_dma *dw)
-{
-	int i;
-
-	dma_writel(dw, CFG, 0);
-
-	channel_clear_bit(dw, MASK.XFER, dw->all_chan_mask);
-	channel_clear_bit(dw, MASK.SRC_TRAN, dw->all_chan_mask);
-	channel_clear_bit(dw, MASK.DST_TRAN, dw->all_chan_mask);
-	channel_clear_bit(dw, MASK.ERROR, dw->all_chan_mask);
-
-	while (dma_readl(dw, CFG) & DW_CFG_DMA_EN)
-		cpu_relax();
-
-	for (i = 0; i < dw->dma.chancnt; i++)
-		dw->chan[i].initialized = false;
-}
-
 int dw_dma_probe(struct dw_dma_chip *chip, struct dw_dma_platform_data *pdata)
 {
 	struct dw_dma		*dw;
@@ -1495,12 +1505,7 @@ int dw_dma_probe(struct dw_dma_chip *chip, struct dw_dma_platform_data *pdata)
 	dw->regs = chip->regs;
 	chip->dw = dw;
 
-	dw->clk = devm_clk_get(chip->dev, "hclk");
-	if (IS_ERR(dw->clk))
-		return PTR_ERR(dw->clk);
-	err = clk_prepare_enable(dw->clk);
-	if (err)
-		return err;
+	pm_runtime_get_sync(chip->dev);
 
 	dw_params = dma_read_byaddr(chip->regs, DW_PARAMS);
 	autocfg = dw_params >> DW_PARAMS_EN & 0x1;
@@ -1604,7 +1609,6 @@ int dw_dma_probe(struct dw_dma_chip *chip, struct dw_dma_platform_data *pdata)
 		channel_clear_bit(dw, CH_EN, dwc->mask);
 
 		dwc->direction = DMA_TRANS_NONE;
-		dwc->request_line = ~0;
 
 		/* Hardware configuration */
 		if (autocfg) {
@@ -1659,8 +1663,6 @@ int dw_dma_probe(struct dw_dma_chip *chip, struct dw_dma_platform_data *pdata)
 	dw->dma.device_tx_status = dwc_tx_status;
 	dw->dma.device_issue_pending = dwc_issue_pending;
 
-	dma_writel(dw, CFG, DW_CFG_DMA_EN);
-
 	err = dma_async_device_register(&dw->dma);
 	if (err)
 		goto err_dma_register;
@@ -1668,12 +1670,14 @@ int dw_dma_probe(struct dw_dma_chip *chip, struct dw_dma_platform_data *pdata)
 	dev_info(chip->dev, "DesignWare DMA Controller, %d channels\n",
 		 nr_channels);
 
+	pm_runtime_put_sync_suspend(chip->dev);
+
 	return 0;
 
 err_dma_register:
 	free_irq(chip->irq, dw);
 err_pdata:
-	clk_disable_unprepare(dw->clk);
+	pm_runtime_put_sync_suspend(chip->dev);
 	return err;
 }
 EXPORT_SYMBOL_GPL(dw_dma_probe);
@@ -1683,6 +1687,8 @@ int dw_dma_remove(struct dw_dma_chip *chip)
 	struct dw_dma		*dw = chip->dw;
 	struct dw_dma_chan	*dwc, *_dwc;
 
+	pm_runtime_get_sync(chip->dev);
+
 	dw_dma_off(dw);
 	dma_async_device_unregister(&dw->dma);
 
@@ -1695,46 +1701,28 @@ int dw_dma_remove(struct dw_dma_chip *chip)
 		channel_clear_bit(dw, CH_EN, dwc->mask);
 	}
 
-	clk_disable_unprepare(dw->clk);
-
+	pm_runtime_put_sync_suspend(chip->dev);
 	return 0;
 }
 EXPORT_SYMBOL_GPL(dw_dma_remove);
 
-void dw_dma_shutdown(struct dw_dma_chip *chip)
-{
-	struct dw_dma *dw = chip->dw;
-
-	dw_dma_off(dw);
-	clk_disable_unprepare(dw->clk);
-}
-EXPORT_SYMBOL_GPL(dw_dma_shutdown);
-
-#ifdef CONFIG_PM_SLEEP
-
-int dw_dma_suspend(struct dw_dma_chip *chip)
+int dw_dma_disable(struct dw_dma_chip *chip)
 {
 	struct dw_dma *dw = chip->dw;
 
 	dw_dma_off(dw);
-	clk_disable_unprepare(dw->clk);
-
 	return 0;
 }
-EXPORT_SYMBOL_GPL(dw_dma_suspend);
+EXPORT_SYMBOL_GPL(dw_dma_disable);
 
-int dw_dma_resume(struct dw_dma_chip *chip)
+int dw_dma_enable(struct dw_dma_chip *chip)
 {
 	struct dw_dma *dw = chip->dw;
 
-	clk_prepare_enable(dw->clk);
-	dma_writel(dw, CFG, DW_CFG_DMA_EN);
-
+	dw_dma_on(dw);
 	return 0;
 }
-EXPORT_SYMBOL_GPL(dw_dma_resume);
-
-#endif /* CONFIG_PM_SLEEP */
+EXPORT_SYMBOL_GPL(dw_dma_enable);
 
 MODULE_LICENSE("GPL v2");
 MODULE_DESCRIPTION("Synopsys DesignWare DMA Controller core driver");
diff --git a/drivers/dma/dw/internal.h b/drivers/dma/dw/internal.h
index 32667f9e0dda..41439732ff6b 100644
--- a/drivers/dma/dw/internal.h
+++ b/drivers/dma/dw/internal.h
@@ -8,63 +8,16 @@
  * published by the Free Software Foundation.
  */
 
-#ifndef _DW_DMAC_INTERNAL_H
-#define _DW_DMAC_INTERNAL_H
+#ifndef _DMA_DW_INTERNAL_H
+#define _DMA_DW_INTERNAL_H
 
-#include <linux/device.h>
-#include <linux/dw_dmac.h>
+#include <linux/dma/dw.h>
 
 #include "regs.h"
 
-/**
- * struct dw_dma_chip - representation of DesignWare DMA controller hardware
- * @dev:		struct device of the DMA controller
- * @irq:		irq line
- * @regs:		memory mapped I/O space
- * @dw:			struct dw_dma that is filed by dw_dma_probe()
- */
-struct dw_dma_chip {
-	struct device	*dev;
-	int		irq;
-	void __iomem	*regs;
-	struct dw_dma	*dw;
-};
-
-/* Export to the platform drivers */
-int dw_dma_probe(struct dw_dma_chip *chip, struct dw_dma_platform_data *pdata);
-int dw_dma_remove(struct dw_dma_chip *chip);
-
-void dw_dma_shutdown(struct dw_dma_chip *chip);
-
-#ifdef CONFIG_PM_SLEEP
-
-int dw_dma_suspend(struct dw_dma_chip *chip);
-int dw_dma_resume(struct dw_dma_chip *chip);
-
-#endif /* CONFIG_PM_SLEEP */
+int dw_dma_disable(struct dw_dma_chip *chip);
+int dw_dma_enable(struct dw_dma_chip *chip);
 
-/**
- * dwc_get_dms - get destination master
- * @slave:	pointer to the custom slave configuration
- *
- * Returns destination master in the custom slave configuration if defined, or
- * default value otherwise.
- */
-static inline unsigned int dwc_get_dms(struct dw_dma_slave *slave)
-{
-	return slave ? slave->dst_master : 0;
-}
-
-/**
- * dwc_get_sms - get source master
- * @slave:	pointer to the custom slave configuration
- *
- * Returns source master in the custom slave configuration if defined, or
- * default value otherwise.
- */
-static inline unsigned int dwc_get_sms(struct dw_dma_slave *slave)
-{
-	return slave ? slave->src_master : 1;
-}
+extern bool dw_dma_filter(struct dma_chan *chan, void *param);
 
-#endif /* _DW_DMAC_INTERNAL_H */
+#endif /* _DMA_DW_INTERNAL_H */
diff --git a/drivers/dma/dw/pci.c b/drivers/dma/dw/pci.c
index 39e30c3c7a9d..b144706b3d85 100644
--- a/drivers/dma/dw/pci.c
+++ b/drivers/dma/dw/pci.c
@@ -82,7 +82,7 @@ static int dw_pci_suspend_late(struct device *dev)
 	struct pci_dev *pci = to_pci_dev(dev);
 	struct dw_dma_chip *chip = pci_get_drvdata(pci);
 
-	return dw_dma_suspend(chip);
+	return dw_dma_disable(chip);
 };
 
 static int dw_pci_resume_early(struct device *dev)
@@ -90,7 +90,7 @@ static int dw_pci_resume_early(struct device *dev)
 	struct pci_dev *pci = to_pci_dev(dev);
 	struct dw_dma_chip *chip = pci_get_drvdata(pci);
 
-	return dw_dma_resume(chip);
+	return dw_dma_enable(chip);
 };
 
 #endif /* CONFIG_PM_SLEEP */
@@ -108,6 +108,10 @@ static const struct pci_device_id dw_pci_id_table[] = {
 	{ PCI_VDEVICE(INTEL, 0x0f06), (kernel_ulong_t)&dw_pci_pdata },
 	{ PCI_VDEVICE(INTEL, 0x0f40), (kernel_ulong_t)&dw_pci_pdata },
 
+	/* Braswell */
+	{ PCI_VDEVICE(INTEL, 0x2286), (kernel_ulong_t)&dw_pci_pdata },
+	{ PCI_VDEVICE(INTEL, 0x22c0), (kernel_ulong_t)&dw_pci_pdata },
+
 	/* Haswell */
 	{ PCI_VDEVICE(INTEL, 0x9c60), (kernel_ulong_t)&dw_pci_pdata },
 	{ }
diff --git a/drivers/dma/dw/platform.c b/drivers/dma/dw/platform.c
index c5b339af6be5..32ea1aca7a0e 100644
--- a/drivers/dma/dw/platform.c
+++ b/drivers/dma/dw/platform.c
@@ -15,6 +15,7 @@
 #include <linux/module.h>
 #include <linux/device.h>
 #include <linux/clk.h>
+#include <linux/pm_runtime.h>
 #include <linux/platform_device.h>
 #include <linux/dmaengine.h>
 #include <linux/dma-mapping.h>
@@ -25,72 +26,49 @@
 
 #include "internal.h"
 
-struct dw_dma_of_filter_args {
-	struct dw_dma *dw;
-	unsigned int req;
-	unsigned int src;
-	unsigned int dst;
-};
-
-static bool dw_dma_of_filter(struct dma_chan *chan, void *param)
-{
-	struct dw_dma_chan *dwc = to_dw_dma_chan(chan);
-	struct dw_dma_of_filter_args *fargs = param;
-
-	/* Ensure the device matches our channel */
-	if (chan->device != &fargs->dw->dma)
-		return false;
-
-	dwc->request_line = fargs->req;
-	dwc->src_master	= fargs->src;
-	dwc->dst_master	= fargs->dst;
-
-	return true;
-}
-
 static struct dma_chan *dw_dma_of_xlate(struct of_phandle_args *dma_spec,
 					struct of_dma *ofdma)
 {
 	struct dw_dma *dw = ofdma->of_dma_data;
-	struct dw_dma_of_filter_args fargs = {
-		.dw = dw,
+	struct dw_dma_slave slave = {
+		.dma_dev = dw->dma.dev,
 	};
 	dma_cap_mask_t cap;
 
 	if (dma_spec->args_count != 3)
 		return NULL;
 
-	fargs.req = dma_spec->args[0];
-	fargs.src = dma_spec->args[1];
-	fargs.dst = dma_spec->args[2];
+	slave.src_id = dma_spec->args[0];
+	slave.dst_id = dma_spec->args[0];
+	slave.src_master = dma_spec->args[1];
+	slave.dst_master = dma_spec->args[2];
 
-	if (WARN_ON(fargs.req >= DW_DMA_MAX_NR_REQUESTS ||
-		    fargs.src >= dw->nr_masters ||
-		    fargs.dst >= dw->nr_masters))
+	if (WARN_ON(slave.src_id >= DW_DMA_MAX_NR_REQUESTS ||
+		    slave.dst_id >= DW_DMA_MAX_NR_REQUESTS ||
+		    slave.src_master >= dw->nr_masters ||
+		    slave.dst_master >= dw->nr_masters))
 		return NULL;
 
 	dma_cap_zero(cap);
 	dma_cap_set(DMA_SLAVE, cap);
 
 	/* TODO: there should be a simpler way to do this */
-	return dma_request_channel(cap, dw_dma_of_filter, &fargs);
+	return dma_request_channel(cap, dw_dma_filter, &slave);
 }
 
 #ifdef CONFIG_ACPI
 static bool dw_dma_acpi_filter(struct dma_chan *chan, void *param)
 {
-	struct dw_dma_chan *dwc = to_dw_dma_chan(chan);
 	struct acpi_dma_spec *dma_spec = param;
+	struct dw_dma_slave slave = {
+		.dma_dev = dma_spec->dev,
+		.src_id = dma_spec->slave_id,
+		.dst_id = dma_spec->slave_id,
+		.src_master = 1,
+		.dst_master = 0,
+	};
 
-	if (chan->device->dev != dma_spec->dev ||
-	    chan->chan_id != dma_spec->chan_id)
-		return false;
-
-	dwc->request_line = dma_spec->slave_id;
-	dwc->src_master = dwc_get_sms(NULL);
-	dwc->dst_master = dwc_get_dms(NULL);
-
-	return true;
+	return dw_dma_filter(chan, &slave);
 }
 
 static void dw_dma_acpi_controller_register(struct dw_dma *dw)
@@ -201,10 +179,19 @@ static int dw_probe(struct platform_device *pdev)
 
 	chip->dev = dev;
 
-	err = dw_dma_probe(chip, pdata);
+	chip->clk = devm_clk_get(chip->dev, "hclk");
+	if (IS_ERR(chip->clk))
+		return PTR_ERR(chip->clk);
+	err = clk_prepare_enable(chip->clk);
 	if (err)
 		return err;
 
+	pm_runtime_enable(&pdev->dev);
+
+	err = dw_dma_probe(chip, pdata);
+	if (err)
+		goto err_dw_dma_probe;
+
 	platform_set_drvdata(pdev, chip);
 
 	if (pdev->dev.of_node) {
@@ -219,6 +206,11 @@ static int dw_probe(struct platform_device *pdev)
 		dw_dma_acpi_controller_register(chip->dw);
 
 	return 0;
+
+err_dw_dma_probe:
+	pm_runtime_disable(&pdev->dev);
+	clk_disable_unprepare(chip->clk);
+	return err;
 }
 
 static int dw_remove(struct platform_device *pdev)
@@ -228,14 +220,19 @@ static int dw_remove(struct platform_device *pdev)
 	if (pdev->dev.of_node)
 		of_dma_controller_free(pdev->dev.of_node);
 
-	return dw_dma_remove(chip);
+	dw_dma_remove(chip);
+	pm_runtime_disable(&pdev->dev);
+	clk_disable_unprepare(chip->clk);
+
+	return 0;
 }
 
 static void dw_shutdown(struct platform_device *pdev)
 {
 	struct dw_dma_chip *chip = platform_get_drvdata(pdev);
 
-	dw_dma_shutdown(chip);
+	dw_dma_disable(chip);
+	clk_disable_unprepare(chip->clk);
 }
 
 #ifdef CONFIG_OF
@@ -261,7 +258,10 @@ static int dw_suspend_late(struct device *dev)
 	struct platform_device *pdev = to_platform_device(dev);
 	struct dw_dma_chip *chip = platform_get_drvdata(pdev);
 
-	return dw_dma_suspend(chip);
+	dw_dma_disable(chip);
+	clk_disable_unprepare(chip->clk);
+
+	return 0;
 }
 
 static int dw_resume_early(struct device *dev)
@@ -269,7 +269,8 @@ static int dw_resume_early(struct device *dev)
 	struct platform_device *pdev = to_platform_device(dev);
 	struct dw_dma_chip *chip = platform_get_drvdata(pdev);
 
-	return dw_dma_resume(chip);
+	clk_prepare_enable(chip->clk);
+	return dw_dma_enable(chip);
 }
 
 #endif /* CONFIG_PM_SLEEP */
@@ -281,7 +282,7 @@ static const struct dev_pm_ops dw_dev_pm_ops = {
 static struct platform_driver dw_driver = {
 	.probe		= dw_probe,
 	.remove		= dw_remove,
-	.shutdown	= dw_shutdown,
+	.shutdown       = dw_shutdown,
 	.driver = {
 		.name	= "dw_dmac",
 		.pm	= &dw_dev_pm_ops,
diff --git a/drivers/dma/dw/regs.h b/drivers/dma/dw/regs.h
index bb98d3e91e8b..848e232f7cc7 100644
--- a/drivers/dma/dw/regs.h
+++ b/drivers/dma/dw/regs.h
@@ -11,7 +11,6 @@
 
 #include <linux/interrupt.h>
 #include <linux/dmaengine.h>
-#include <linux/dw_dmac.h>
 
 #define DW_DMA_MAX_NR_CHANNELS	8
 #define DW_DMA_MAX_NR_REQUESTS	16
@@ -132,6 +131,18 @@ struct dw_dma_regs {
 /* Bitfields in DWC_PARAMS */
 #define DWC_PARAMS_MBLK_EN	11		/* multi block transfer */
 
+/* bursts size */
+enum dw_dma_msize {
+	DW_DMA_MSIZE_1,
+	DW_DMA_MSIZE_4,
+	DW_DMA_MSIZE_8,
+	DW_DMA_MSIZE_16,
+	DW_DMA_MSIZE_32,
+	DW_DMA_MSIZE_64,
+	DW_DMA_MSIZE_128,
+	DW_DMA_MSIZE_256,
+};
+
 /* Bitfields in CTL_LO */
 #define DWC_CTLL_INT_EN		(1 << 0)	/* irqs enabled? */
 #define DWC_CTLL_DST_WIDTH(n)	((n)<<1)	/* bytes per element */
@@ -161,20 +172,35 @@ struct dw_dma_regs {
 #define DWC_CTLH_DONE		0x00001000
 #define DWC_CTLH_BLOCK_TS_MASK	0x00000fff
 
-/* Bitfields in CFG_LO. Platform-configurable bits are in <linux/dw_dmac.h> */
+/* Bitfields in CFG_LO */
 #define DWC_CFGL_CH_PRIOR_MASK	(0x7 << 5)	/* priority mask */
 #define DWC_CFGL_CH_PRIOR(x)	((x) << 5)	/* priority */
 #define DWC_CFGL_CH_SUSP	(1 << 8)	/* pause xfer */
 #define DWC_CFGL_FIFO_EMPTY	(1 << 9)	/* pause xfer */
 #define DWC_CFGL_HS_DST		(1 << 10)	/* handshake w/dst */
 #define DWC_CFGL_HS_SRC		(1 << 11)	/* handshake w/src */
+#define DWC_CFGL_LOCK_CH_XFER	(0 << 12)	/* scope of LOCK_CH */
+#define DWC_CFGL_LOCK_CH_BLOCK	(1 << 12)
+#define DWC_CFGL_LOCK_CH_XACT	(2 << 12)
+#define DWC_CFGL_LOCK_BUS_XFER	(0 << 14)	/* scope of LOCK_BUS */
+#define DWC_CFGL_LOCK_BUS_BLOCK	(1 << 14)
+#define DWC_CFGL_LOCK_BUS_XACT	(2 << 14)
+#define DWC_CFGL_LOCK_CH	(1 << 15)	/* channel lockout */
+#define DWC_CFGL_LOCK_BUS	(1 << 16)	/* busmaster lockout */
+#define DWC_CFGL_HS_DST_POL	(1 << 18)	/* dst handshake active low */
+#define DWC_CFGL_HS_SRC_POL	(1 << 19)	/* src handshake active low */
 #define DWC_CFGL_MAX_BURST(x)	((x) << 20)
 #define DWC_CFGL_RELOAD_SAR	(1 << 30)
 #define DWC_CFGL_RELOAD_DAR	(1 << 31)
 
-/* Bitfields in CFG_HI. Platform-configurable bits are in <linux/dw_dmac.h> */
+/* Bitfields in CFG_HI */
+#define DWC_CFGH_FCMODE		(1 << 0)
+#define DWC_CFGH_FIFO_MODE	(1 << 1)
+#define DWC_CFGH_PROTCTL(x)	((x) << 2)
 #define DWC_CFGH_DS_UPD_EN	(1 << 5)
 #define DWC_CFGH_SS_UPD_EN	(1 << 6)
+#define DWC_CFGH_SRC_PER(x)	((x) << 7)
+#define DWC_CFGH_DST_PER(x)	((x) << 11)
 
 /* Bitfields in SGR */
 #define DWC_SGR_SGI(x)		((x) << 0)
@@ -221,9 +247,10 @@ struct dw_dma_chan {
 	bool			nollp;
 
 	/* custom slave configuration */
-	unsigned int		request_line;
-	unsigned char		src_master;
-	unsigned char		dst_master;
+	u8			src_id;
+	u8			dst_id;
+	u8			src_master;
+	u8			dst_master;
 
 	/* configuration passed via DMA_SLAVE_CONFIG */
 	struct dma_slave_config dma_sconfig;
@@ -250,11 +277,11 @@ struct dw_dma {
 	void __iomem		*regs;
 	struct dma_pool		*desc_pool;
 	struct tasklet_struct	tasklet;
-	struct clk		*clk;
 
 	/* channels */
 	struct dw_dma_chan	*chan;
 	u8			all_chan_mask;
+	u8			in_use;
 
 	/* hardware configuration */
 	unsigned char		nr_masters;
diff --git a/drivers/dma/edma.c b/drivers/dma/edma.c
index 7b65633f495e..b969206439b7 100644
--- a/drivers/dma/edma.c
+++ b/drivers/dma/edma.c
@@ -288,7 +288,7 @@ static int edma_slave_config(struct edma_chan *echan,
 static int edma_dma_pause(struct edma_chan *echan)
 {
 	/* Pause/Resume only allowed with cyclic mode */
-	if (!echan->edesc->cyclic)
+	if (!echan->edesc || !echan->edesc->cyclic)
 		return -EINVAL;
 
 	edma_pause(echan->ch_num);
@@ -1092,7 +1092,6 @@ static struct platform_driver edma_driver = {
 	.remove		= edma_remove,
 	.driver = {
 		.name = "edma-dma-engine",
-		.owner = THIS_MODULE,
 	},
 };
 
@@ -1107,52 +1106,14 @@ bool edma_filter_fn(struct dma_chan *chan, void *param)
 }
 EXPORT_SYMBOL(edma_filter_fn);
 
-static struct platform_device *pdev0, *pdev1;
-
-static const struct platform_device_info edma_dev_info0 = {
-	.name = "edma-dma-engine",
-	.id = 0,
-	.dma_mask = DMA_BIT_MASK(32),
-};
-
-static const struct platform_device_info edma_dev_info1 = {
-	.name = "edma-dma-engine",
-	.id = 1,
-	.dma_mask = DMA_BIT_MASK(32),
-};
-
 static int edma_init(void)
 {
-	int ret = platform_driver_register(&edma_driver);
-
-	if (ret == 0) {
-		pdev0 = platform_device_register_full(&edma_dev_info0);
-		if (IS_ERR(pdev0)) {
-			platform_driver_unregister(&edma_driver);
-			ret = PTR_ERR(pdev0);
-			goto out;
-		}
-	}
-
-	if (!of_have_populated_dt() && EDMA_CTLRS == 2) {
-		pdev1 = platform_device_register_full(&edma_dev_info1);
-		if (IS_ERR(pdev1)) {
-			platform_driver_unregister(&edma_driver);
-			platform_device_unregister(pdev0);
-			ret = PTR_ERR(pdev1);
-		}
-	}
-
-out:
-	return ret;
+	return platform_driver_register(&edma_driver);
 }
 subsys_initcall(edma_init);
 
 static void __exit edma_exit(void)
 {
-	platform_device_unregister(pdev0);
-	if (pdev1)
-		platform_device_unregister(pdev1);
 	platform_driver_unregister(&edma_driver);
 }
 module_exit(edma_exit);
diff --git a/drivers/dma/fsl-edma.c b/drivers/dma/fsl-edma.c
index 3c5711d5fe97..e9ebb89e1711 100644
--- a/drivers/dma/fsl-edma.c
+++ b/drivers/dma/fsl-edma.c
@@ -118,17 +118,17 @@
 				BIT(DMA_SLAVE_BUSWIDTH_8_BYTES)
 
 struct fsl_edma_hw_tcd {
-	u32	saddr;
-	u16	soff;
-	u16	attr;
-	u32	nbytes;
-	u32	slast;
-	u32	daddr;
-	u16	doff;
-	u16	citer;
-	u32	dlast_sga;
-	u16	csr;
-	u16	biter;
+	__le32	saddr;
+	__le16	soff;
+	__le16	attr;
+	__le32	nbytes;
+	__le32	slast;
+	__le32	daddr;
+	__le16	doff;
+	__le16	citer;
+	__le32	dlast_sga;
+	__le16	csr;
+	__le16	biter;
 };
 
 struct fsl_edma_sw_tcd {
@@ -175,18 +175,12 @@ struct fsl_edma_engine {
 };
 
 /*
- * R/W functions for big- or little-endian registers
- * the eDMA controller's endian is independent of the CPU core's endian.
+ * R/W functions for big- or little-endian registers:
+ * The eDMA controller's endian is independent of the CPU core's endian.
+ * For the big-endian IP module, the offset for 8-bit or 16-bit registers
+ * should also be swapped opposite to that in little-endian IP.
  */
 
-static u16 edma_readw(struct fsl_edma_engine *edma, void __iomem *addr)
-{
-	if (edma->big_endian)
-		return ioread16be(addr);
-	else
-		return ioread16(addr);
-}
-
 static u32 edma_readl(struct fsl_edma_engine *edma, void __iomem *addr)
 {
 	if (edma->big_endian)
@@ -197,13 +191,18 @@ static u32 edma_readl(struct fsl_edma_engine *edma, void __iomem *addr)
 
 static void edma_writeb(struct fsl_edma_engine *edma, u8 val, void __iomem *addr)
 {
-	iowrite8(val, addr);
+	/* swap the reg offset for these in big-endian mode */
+	if (edma->big_endian)
+		iowrite8(val, (void __iomem *)((unsigned long)addr ^ 0x3));
+	else
+		iowrite8(val, addr);
 }
 
 static void edma_writew(struct fsl_edma_engine *edma, u16 val, void __iomem *addr)
 {
+	/* swap the reg offset for these in big-endian mode */
 	if (edma->big_endian)
-		iowrite16be(val, addr);
+		iowrite16be(val, (void __iomem *)((unsigned long)addr ^ 0x2));
 	else
 		iowrite16(val, addr);
 }
@@ -254,13 +253,12 @@ static void fsl_edma_chan_mux(struct fsl_edma_chan *fsl_chan,
 	chans_per_mux = fsl_chan->edma->n_chans / DMAMUX_NR;
 	ch_off = fsl_chan->vchan.chan.chan_id % chans_per_mux;
 	muxaddr = fsl_chan->edma->muxbase[ch / chans_per_mux];
+	slot = EDMAMUX_CHCFG_SOURCE(slot);
 
 	if (enable)
-		edma_writeb(fsl_chan->edma,
-				EDMAMUX_CHCFG_ENBL | EDMAMUX_CHCFG_SOURCE(slot),
-				muxaddr + ch_off);
+		iowrite8(EDMAMUX_CHCFG_ENBL | slot, muxaddr + ch_off);
 	else
-		edma_writeb(fsl_chan->edma, EDMAMUX_CHCFG_DIS, muxaddr + ch_off);
+		iowrite8(EDMAMUX_CHCFG_DIS, muxaddr + ch_off);
 }
 
 static unsigned int fsl_edma_get_tcd_attr(enum dma_slave_buswidth addr_width)
@@ -286,9 +284,8 @@ static void fsl_edma_free_desc(struct virt_dma_desc *vdesc)
 
 	fsl_desc = to_fsl_edma_desc(vdesc);
 	for (i = 0; i < fsl_desc->n_tcds; i++)
-			dma_pool_free(fsl_desc->echan->tcd_pool,
-					fsl_desc->tcd[i].vtcd,
-					fsl_desc->tcd[i].ptcd);
+		dma_pool_free(fsl_desc->echan->tcd_pool, fsl_desc->tcd[i].vtcd,
+			      fsl_desc->tcd[i].ptcd);
 	kfree(fsl_desc);
 }
 
@@ -363,8 +360,8 @@ static size_t fsl_edma_desc_residue(struct fsl_edma_chan *fsl_chan,
 
 	/* calculate the total size in this desc */
 	for (len = i = 0; i < fsl_chan->edesc->n_tcds; i++)
-		len += edma_readl(fsl_chan->edma, &(edesc->tcd[i].vtcd->nbytes))
-			* edma_readw(fsl_chan->edma, &(edesc->tcd[i].vtcd->biter));
+		len += le32_to_cpu(edesc->tcd[i].vtcd->nbytes)
+			* le16_to_cpu(edesc->tcd[i].vtcd->biter);
 
 	if (!in_progress)
 		return len;
@@ -376,17 +373,15 @@ static size_t fsl_edma_desc_residue(struct fsl_edma_chan *fsl_chan,
 
 	/* figure out the finished and calculate the residue */
 	for (i = 0; i < fsl_chan->edesc->n_tcds; i++) {
-		size = edma_readl(fsl_chan->edma, &(edesc->tcd[i].vtcd->nbytes))
-			* edma_readw(fsl_chan->edma, &(edesc->tcd[i].vtcd->biter));
+		size = le32_to_cpu(edesc->tcd[i].vtcd->nbytes)
+			* le16_to_cpu(edesc->tcd[i].vtcd->biter);
 		if (dir == DMA_MEM_TO_DEV)
-			dma_addr = edma_readl(fsl_chan->edma,
-					&(edesc->tcd[i].vtcd->saddr));
+			dma_addr = le32_to_cpu(edesc->tcd[i].vtcd->saddr);
 		else
-			dma_addr = edma_readl(fsl_chan->edma,
-					&(edesc->tcd[i].vtcd->daddr));
+			dma_addr = le32_to_cpu(edesc->tcd[i].vtcd->daddr);
 
 		len -= size;
-		if (cur_addr > dma_addr && cur_addr < dma_addr + size) {
+		if (cur_addr >= dma_addr && cur_addr < dma_addr + size) {
 			len += dma_addr + size - cur_addr;
 			break;
 		}
@@ -424,55 +419,67 @@ static enum dma_status fsl_edma_tx_status(struct dma_chan *chan,
 	return fsl_chan->status;
 }
 
-static void fsl_edma_set_tcd_params(struct fsl_edma_chan *fsl_chan,
-		u32 src, u32 dst, u16 attr, u16 soff, u32 nbytes,
-		u32 slast, u16 citer, u16 biter, u32 doff, u32 dlast_sga,
-		u16 csr)
+static void fsl_edma_set_tcd_regs(struct fsl_edma_chan *fsl_chan,
+				  struct fsl_edma_hw_tcd *tcd)
 {
+	struct fsl_edma_engine *edma = fsl_chan->edma;
 	void __iomem *addr = fsl_chan->edma->membase;
 	u32 ch = fsl_chan->vchan.chan.chan_id;
 
 	/*
-	 * TCD parameters have been swapped in fill_tcd_params(),
-	 * so just write them to registers in the cpu endian here
+	 * TCD parameters are stored in struct fsl_edma_hw_tcd in little
+	 * endian format. However, we need to load the TCD registers in
+	 * big- or little-endian obeying the eDMA engine model endian.
 	 */
-	writew(0, addr + EDMA_TCD_CSR(ch));
-	writel(src, addr + EDMA_TCD_SADDR(ch));
-	writel(dst, addr + EDMA_TCD_DADDR(ch));
-	writew(attr, addr + EDMA_TCD_ATTR(ch));
-	writew(soff, addr + EDMA_TCD_SOFF(ch));
-	writel(nbytes, addr + EDMA_TCD_NBYTES(ch));
-	writel(slast, addr + EDMA_TCD_SLAST(ch));
-	writew(citer, addr + EDMA_TCD_CITER(ch));
-	writew(biter, addr + EDMA_TCD_BITER(ch));
-	writew(doff, addr + EDMA_TCD_DOFF(ch));
-	writel(dlast_sga, addr + EDMA_TCD_DLAST_SGA(ch));
-	writew(csr, addr + EDMA_TCD_CSR(ch));
-}
-
-static void fill_tcd_params(struct fsl_edma_engine *edma,
-		struct fsl_edma_hw_tcd *tcd, u32 src, u32 dst,
-		u16 attr, u16 soff, u32 nbytes, u32 slast, u16 citer,
-		u16 biter, u16 doff, u32 dlast_sga, bool major_int,
-		bool disable_req, bool enable_sg)
+	edma_writew(edma, 0, addr + EDMA_TCD_CSR(ch));
+	edma_writel(edma, le32_to_cpu(tcd->saddr), addr + EDMA_TCD_SADDR(ch));
+	edma_writel(edma, le32_to_cpu(tcd->daddr), addr + EDMA_TCD_DADDR(ch));
+
+	edma_writew(edma, le16_to_cpu(tcd->attr), addr + EDMA_TCD_ATTR(ch));
+	edma_writew(edma, le16_to_cpu(tcd->soff), addr + EDMA_TCD_SOFF(ch));
+
+	edma_writel(edma, le32_to_cpu(tcd->nbytes), addr + EDMA_TCD_NBYTES(ch));
+	edma_writel(edma, le32_to_cpu(tcd->slast), addr + EDMA_TCD_SLAST(ch));
+
+	edma_writew(edma, le16_to_cpu(tcd->citer), addr + EDMA_TCD_CITER(ch));
+	edma_writew(edma, le16_to_cpu(tcd->biter), addr + EDMA_TCD_BITER(ch));
+	edma_writew(edma, le16_to_cpu(tcd->doff), addr + EDMA_TCD_DOFF(ch));
+
+	edma_writel(edma, le32_to_cpu(tcd->dlast_sga), addr + EDMA_TCD_DLAST_SGA(ch));
+
+	edma_writew(edma, le16_to_cpu(tcd->csr), addr + EDMA_TCD_CSR(ch));
+}
+
+static inline
+void fsl_edma_fill_tcd(struct fsl_edma_hw_tcd *tcd, u32 src, u32 dst,
+		       u16 attr, u16 soff, u32 nbytes, u32 slast, u16 citer,
+		       u16 biter, u16 doff, u32 dlast_sga, bool major_int,
+		       bool disable_req, bool enable_sg)
 {
 	u16 csr = 0;
 
 	/*
-	 * eDMA hardware SGs require the TCD parameters stored in memory
-	 * the same endian as the eDMA module so that they can be loaded
-	 * automatically by the engine
+	 * eDMA hardware SGs require the TCDs to be stored in little
+	 * endian format irrespective of the register endian model.
+	 * So we put the value in little endian in memory, waiting
+	 * for fsl_edma_set_tcd_regs doing the swap.
 	 */
-	edma_writel(edma, src, &(tcd->saddr));
-	edma_writel(edma, dst, &(tcd->daddr));
-	edma_writew(edma, attr, &(tcd->attr));
-	edma_writew(edma, EDMA_TCD_SOFF_SOFF(soff), &(tcd->soff));
-	edma_writel(edma, EDMA_TCD_NBYTES_NBYTES(nbytes), &(tcd->nbytes));
-	edma_writel(edma, EDMA_TCD_SLAST_SLAST(slast), &(tcd->slast));
-	edma_writew(edma, EDMA_TCD_CITER_CITER(citer), &(tcd->citer));
-	edma_writew(edma, EDMA_TCD_DOFF_DOFF(doff), &(tcd->doff));
-	edma_writel(edma, EDMA_TCD_DLAST_SGA_DLAST_SGA(dlast_sga), &(tcd->dlast_sga));
-	edma_writew(edma, EDMA_TCD_BITER_BITER(biter), &(tcd->biter));
+	tcd->saddr = cpu_to_le32(src);
+	tcd->daddr = cpu_to_le32(dst);
+
+	tcd->attr = cpu_to_le16(attr);
+
+	tcd->soff = cpu_to_le16(EDMA_TCD_SOFF_SOFF(soff));
+
+	tcd->nbytes = cpu_to_le32(EDMA_TCD_NBYTES_NBYTES(nbytes));
+	tcd->slast = cpu_to_le32(EDMA_TCD_SLAST_SLAST(slast));
+
+	tcd->citer = cpu_to_le16(EDMA_TCD_CITER_CITER(citer));
+	tcd->doff = cpu_to_le16(EDMA_TCD_DOFF_DOFF(doff));
+
+	tcd->dlast_sga = cpu_to_le32(EDMA_TCD_DLAST_SGA_DLAST_SGA(dlast_sga));
+
+	tcd->biter = cpu_to_le16(EDMA_TCD_BITER_BITER(biter));
 	if (major_int)
 		csr |= EDMA_TCD_CSR_INT_MAJOR;
 
@@ -482,7 +489,7 @@ static void fill_tcd_params(struct fsl_edma_engine *edma,
 	if (enable_sg)
 		csr |= EDMA_TCD_CSR_E_SG;
 
-	edma_writew(edma, csr, &(tcd->csr));
+	tcd->csr = cpu_to_le16(csr);
 }
 
 static struct fsl_edma_desc *fsl_edma_alloc_desc(struct fsl_edma_chan *fsl_chan,
@@ -558,9 +565,9 @@ static struct dma_async_tx_descriptor *fsl_edma_prep_dma_cyclic(
 			doff = fsl_chan->fsc.addr_width;
 		}
 
-		fill_tcd_params(fsl_chan->edma, fsl_desc->tcd[i].vtcd, src_addr,
-				dst_addr, fsl_chan->fsc.attr, soff, nbytes, 0,
-				iter, iter, doff, last_sg, true, false, true);
+		fsl_edma_fill_tcd(fsl_desc->tcd[i].vtcd, src_addr, dst_addr,
+				  fsl_chan->fsc.attr, soff, nbytes, 0, iter,
+				  iter, doff, last_sg, true, false, true);
 		dma_buf_next += period_len;
 	}
 
@@ -607,16 +614,16 @@ static struct dma_async_tx_descriptor *fsl_edma_prep_slave_sg(
 		iter = sg_dma_len(sg) / nbytes;
 		if (i < sg_len - 1) {
 			last_sg = fsl_desc->tcd[(i + 1)].ptcd;
-			fill_tcd_params(fsl_chan->edma, fsl_desc->tcd[i].vtcd,
-					src_addr, dst_addr, fsl_chan->fsc.attr,
-					soff, nbytes, 0, iter, iter, doff, last_sg,
-					false, false, true);
+			fsl_edma_fill_tcd(fsl_desc->tcd[i].vtcd, src_addr,
+					  dst_addr, fsl_chan->fsc.attr, soff,
+					  nbytes, 0, iter, iter, doff, last_sg,
+					  false, false, true);
 		} else {
 			last_sg = 0;
-			fill_tcd_params(fsl_chan->edma, fsl_desc->tcd[i].vtcd,
-					src_addr, dst_addr, fsl_chan->fsc.attr,
-					soff, nbytes, 0, iter, iter, doff, last_sg,
-					true, true, false);
+			fsl_edma_fill_tcd(fsl_desc->tcd[i].vtcd, src_addr,
+					  dst_addr, fsl_chan->fsc.attr, soff,
+					  nbytes, 0, iter, iter, doff, last_sg,
+					  true, true, false);
 		}
 	}
 
@@ -625,17 +632,13 @@ static struct dma_async_tx_descriptor *fsl_edma_prep_slave_sg(
 
 static void fsl_edma_xfer_desc(struct fsl_edma_chan *fsl_chan)
 {
-	struct fsl_edma_hw_tcd *tcd;
 	struct virt_dma_desc *vdesc;
 
 	vdesc = vchan_next_desc(&fsl_chan->vchan);
 	if (!vdesc)
 		return;
 	fsl_chan->edesc = to_fsl_edma_desc(vdesc);
-	tcd = fsl_chan->edesc->tcd[0].vtcd;
-	fsl_edma_set_tcd_params(fsl_chan, tcd->saddr, tcd->daddr, tcd->attr,
-			tcd->soff, tcd->nbytes, tcd->slast, tcd->citer,
-			tcd->biter, tcd->doff, tcd->dlast_sga, tcd->csr);
+	fsl_edma_set_tcd_regs(fsl_chan, fsl_chan->edesc->tcd[0].vtcd);
 	fsl_edma_enable_request(fsl_chan);
 	fsl_chan->status = DMA_IN_PROGRESS;
 }
@@ -963,7 +966,6 @@ MODULE_DEVICE_TABLE(of, fsl_edma_dt_ids);
 static struct platform_driver fsl_edma_driver = {
 	.driver		= {
 		.name	= "fsl-edma",
-		.owner  = THIS_MODULE,
 		.of_match_table = fsl_edma_dt_ids,
 	},
 	.probe          = fsl_edma_probe,
diff --git a/drivers/dma/fsldma.c b/drivers/dma/fsldma.c
index d5d6885ab341..38821cdf862b 100644
--- a/drivers/dma/fsldma.c
+++ b/drivers/dma/fsldma.c
@@ -36,7 +36,7 @@
 #include <linux/of_address.h>
 #include <linux/of_irq.h>
 #include <linux/of_platform.h>
-
+#include <linux/fsldma.h>
 #include "dmaengine.h"
 #include "fsldma.h"
 
@@ -367,6 +367,20 @@ static void fsl_chan_toggle_ext_start(struct fsldma_chan *chan, int enable)
 		chan->feature &= ~FSL_DMA_CHAN_START_EXT;
 }
 
+int fsl_dma_external_start(struct dma_chan *dchan, int enable)
+{
+	struct fsldma_chan *chan;
+
+	if (!dchan)
+		return -EINVAL;
+
+	chan = to_fsl_chan(dchan);
+
+	fsl_chan_toggle_ext_start(chan, enable);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(fsl_dma_external_start);
+
 static void append_ld_queue(struct fsldma_chan *chan, struct fsl_desc_sw *desc)
 {
 	struct fsl_desc_sw *tail = to_fsl_desc(chan->ld_pending.prev);
@@ -998,15 +1012,6 @@ static int fsl_dma_device_control(struct dma_chan *dchan,
 		chan->set_request_count(chan, size);
 		return 0;
 
-	case FSLDMA_EXTERNAL_START:
-
-		/* make sure the channel supports external start */
-		if (!chan->toggle_ext_start)
-			return -ENXIO;
-
-		chan->toggle_ext_start(chan, arg);
-		return 0;
-
 	default:
 		return -ENXIO;
 	}
@@ -1332,7 +1337,6 @@ static int fsl_dma_chan_probe(struct fsldma_device *fdev,
 
 	/* Add the channel to DMA device channel list */
 	list_add_tail(&chan->common.device_node, &fdev->common.channels);
-	fdev->common.chancnt++;
 
 	dev_info(fdev->dev, "#%d (%s), irq %d\n", chan->id, compatible,
 		 chan->irq != NO_IRQ ? chan->irq : fdev->irq);
@@ -1535,7 +1539,6 @@ static const struct of_device_id fsldma_of_ids[] = {
 static struct platform_driver fsldma_of_driver = {
 	.driver = {
 		.name = "fsl-elo-dma",
-		.owner = THIS_MODULE,
 		.of_match_table = fsldma_of_ids,
 #ifdef CONFIG_PM
 		.pm = &fsldma_pm_ops,
diff --git a/drivers/dma/imx-dma.c b/drivers/dma/imx-dma.c
index 9d2c9e7374dc..10bbc0a675b0 100644
--- a/drivers/dma/imx-dma.c
+++ b/drivers/dma/imx-dma.c
@@ -1236,7 +1236,6 @@ static int imxdma_remove(struct platform_device *pdev)
 static struct platform_driver imxdma_driver = {
 	.driver		= {
 		.name	= "imx-dma",
-		.owner	= THIS_MODULE,
 		.of_match_table = imx_dma_of_dev_id,
 	},
 	.id_table	= imx_dma_devtype,
diff --git a/drivers/dma/imx-sdma.c b/drivers/dma/imx-sdma.c
index f7626e37d0b8..d0df198f62e9 100644
--- a/drivers/dma/imx-sdma.c
+++ b/drivers/dma/imx-sdma.c
@@ -729,6 +729,7 @@ static void sdma_get_pc(struct sdma_channel *sdmac,
 	case IMX_DMATYPE_CSPI:
 	case IMX_DMATYPE_EXT:
 	case IMX_DMATYPE_SSI:
+	case IMX_DMATYPE_SAI:
 		per_2_emi = sdma->script_addrs->app_2_mcu_addr;
 		emi_2_per = sdma->script_addrs->mcu_2_app_addr;
 		break;
@@ -1287,7 +1288,8 @@ static void sdma_load_firmware(const struct firmware *fw, void *context)
 	unsigned short *ram_code;
 
 	if (!fw) {
-		dev_err(sdma->dev, "firmware not found\n");
+		dev_info(sdma->dev, "external firmware not found, using ROM firmware\n");
+		/* In this case we just use the ROM firmware. */
 		return;
 	}
 
@@ -1334,7 +1336,7 @@ err_firmware:
 	release_firmware(fw);
 }
 
-static int __init sdma_get_firmware(struct sdma_engine *sdma,
+static int sdma_get_firmware(struct sdma_engine *sdma,
 		const char *fw_name)
 {
 	int ret;
@@ -1346,7 +1348,7 @@ static int __init sdma_get_firmware(struct sdma_engine *sdma,
 	return ret;
 }
 
-static int __init sdma_init(struct sdma_engine *sdma)
+static int sdma_init(struct sdma_engine *sdma)
 {
 	int i, ret;
 	dma_addr_t ccb_phys;
@@ -1448,7 +1450,7 @@ static struct dma_chan *sdma_xlate(struct of_phandle_args *dma_spec,
 	return dma_request_channel(mask, sdma_filter_fn, &data);
 }
 
-static int __init sdma_probe(struct platform_device *pdev)
+static int sdma_probe(struct platform_device *pdev)
 {
 	const struct of_device_id *of_id =
 			of_match_device(sdma_dt_ids, &pdev->dev);
@@ -1603,6 +1605,8 @@ static int __init sdma_probe(struct platform_device *pdev)
 	sdma->dma_device.dev->dma_parms = &sdma->dma_parms;
 	dma_set_max_seg_size(sdma->dma_device.dev, 65535);
 
+	platform_set_drvdata(pdev, sdma);
+
 	ret = dma_async_device_register(&sdma->dma_device);
 	if (ret) {
 		dev_err(&pdev->dev, "unable to register\n");
@@ -1640,7 +1644,27 @@ err_irq:
 
 static int sdma_remove(struct platform_device *pdev)
 {
-	return -EBUSY;
+	struct sdma_engine *sdma = platform_get_drvdata(pdev);
+	struct resource *iores = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	int irq = platform_get_irq(pdev, 0);
+	int i;
+
+	dma_async_device_unregister(&sdma->dma_device);
+	kfree(sdma->script_addrs);
+	free_irq(irq, sdma);
+	iounmap(sdma->regs);
+	release_mem_region(iores->start, resource_size(iores));
+	/* Kill the tasklet */
+	for (i = 0; i < MAX_DMA_CHANNELS; i++) {
+		struct sdma_channel *sdmac = &sdma->channel[i];
+
+		tasklet_kill(&sdmac->tasklet);
+	}
+	kfree(sdma);
+
+	platform_set_drvdata(pdev, NULL);
+	dev_info(&pdev->dev, "Removed...\n");
+	return 0;
 }
 
 static struct platform_driver sdma_driver = {
@@ -1650,13 +1674,10 @@ static struct platform_driver sdma_driver = {
 	},
 	.id_table	= sdma_devtypes,
 	.remove		= sdma_remove,
+	.probe		= sdma_probe,
 };
 
-static int __init sdma_module_init(void)
-{
-	return platform_driver_probe(&sdma_driver, sdma_probe);
-}
-module_init(sdma_module_init);
+module_platform_driver(sdma_driver);
 
 MODULE_AUTHOR("Sascha Hauer, Pengutronix <s.hauer@pengutronix.de>");
 MODULE_DESCRIPTION("i.MX SDMA driver");
diff --git a/drivers/dma/ioat/dca.c b/drivers/dma/ioat/dca.c
index 9e84d5bc9307..3b55bb8d969a 100644
--- a/drivers/dma/ioat/dca.c
+++ b/drivers/dma/ioat/dca.c
@@ -35,6 +35,7 @@
 
 #include "dma.h"
 #include "registers.h"
+#include "dma_v2.h"
 
 /*
  * Bit 7 of a tag map entry is the "valid" bit, if it is set then bits 0:6
@@ -147,7 +148,7 @@ static int ioat_dca_add_requester(struct dca_provider *dca, struct device *dev)
 	u16 id;
 
 	/* This implementation only supports PCI-Express */
-	if (dev->bus != &pci_bus_type)
+	if (!dev_is_pci(dev))
 		return -ENODEV;
 	pdev = to_pci_dev(dev);
 	id = dcaid_from_pcidev(pdev);
@@ -179,7 +180,7 @@ static int ioat_dca_remove_requester(struct dca_provider *dca,
 	int i;
 
 	/* This implementation only supports PCI-Express */
-	if (dev->bus != &pci_bus_type)
+	if (!dev_is_pci(dev))
 		return -ENODEV;
 	pdev = to_pci_dev(dev);
 
@@ -320,7 +321,7 @@ static int ioat2_dca_add_requester(struct dca_provider *dca, struct device *dev)
 	u16 global_req_table;
 
 	/* This implementation only supports PCI-Express */
-	if (dev->bus != &pci_bus_type)
+	if (!dev_is_pci(dev))
 		return -ENODEV;
 	pdev = to_pci_dev(dev);
 	id = dcaid_from_pcidev(pdev);
@@ -354,7 +355,7 @@ static int ioat2_dca_remove_requester(struct dca_provider *dca,
 	u16 global_req_table;
 
 	/* This implementation only supports PCI-Express */
-	if (dev->bus != &pci_bus_type)
+	if (!dev_is_pci(dev))
 		return -ENODEV;
 	pdev = to_pci_dev(dev);
 
@@ -496,7 +497,7 @@ static int ioat3_dca_add_requester(struct dca_provider *dca, struct device *dev)
 	u16 global_req_table;
 
 	/* This implementation only supports PCI-Express */
-	if (dev->bus != &pci_bus_type)
+	if (!dev_is_pci(dev))
 		return -ENODEV;
 	pdev = to_pci_dev(dev);
 	id = dcaid_from_pcidev(pdev);
@@ -530,7 +531,7 @@ static int ioat3_dca_remove_requester(struct dca_provider *dca,
 	u16 global_req_table;
 
 	/* This implementation only supports PCI-Express */
-	if (dev->bus != &pci_bus_type)
+	if (!dev_is_pci(dev))
 		return -ENODEV;
 	pdev = to_pci_dev(dev);
 
diff --git a/drivers/dma/ioat/dma.c b/drivers/dma/ioat/dma.c
index 4e3549a16132..940c1502a8b5 100644
--- a/drivers/dma/ioat/dma.c
+++ b/drivers/dma/ioat/dma.c
@@ -947,7 +947,7 @@ msix:
 	for (i = 0; i < msixcnt; i++)
 		device->msix_entries[i].entry = i;
 
-	err = pci_enable_msix(pdev, device->msix_entries, msixcnt);
+	err = pci_enable_msix_exact(pdev, device->msix_entries, msixcnt);
 	if (err)
 		goto msi;
 
@@ -1222,7 +1222,6 @@ int ioat1_dma_probe(struct ioatdma_device *device, int dca)
 	err = ioat_probe(device);
 	if (err)
 		return err;
-	ioat_set_tcp_copy_break(4096);
 	err = ioat_register(device);
 	if (err)
 		return err;
diff --git a/drivers/dma/ioat/dma.h b/drivers/dma/ioat/dma.h
index e982f00a9843..d63f68b1aa35 100644
--- a/drivers/dma/ioat/dma.h
+++ b/drivers/dma/ioat/dma.h
@@ -214,13 +214,6 @@ __dump_desc_dbg(struct ioat_chan_common *chan, struct ioat_dma_descriptor *hw,
 #define dump_desc_dbg(c, d) \
 	({ if (d) __dump_desc_dbg(&c->base, d->hw, &d->txd, desc_id(d)); 0; })
 
-static inline void ioat_set_tcp_copy_break(unsigned long copybreak)
-{
-	#ifdef CONFIG_NET_DMA
-	sysctl_tcp_dma_copybreak = copybreak;
-	#endif
-}
-
 static inline struct ioat_chan_common *
 ioat_chan_by_index(struct ioatdma_device *device, int index)
 {
diff --git a/drivers/dma/ioat/dma_v2.c b/drivers/dma/ioat/dma_v2.c
index 8d1058085eeb..695483e6be32 100644
--- a/drivers/dma/ioat/dma_v2.c
+++ b/drivers/dma/ioat/dma_v2.c
@@ -735,7 +735,8 @@ int ioat2_check_space_lock(struct ioat2_dma_chan *ioat, int num_descs)
 	 * called under bh_disabled so we need to trigger the timer
 	 * event directly
 	 */
-	if (jiffies > chan->timer.expires && timer_pending(&chan->timer)) {
+	if (time_is_before_jiffies(chan->timer.expires)
+	    && timer_pending(&chan->timer)) {
 		struct ioatdma_device *device = chan->device;
 
 		mod_timer(&chan->timer, jiffies + COMPLETION_TIMEOUT);
@@ -899,7 +900,6 @@ int ioat2_dma_probe(struct ioatdma_device *device, int dca)
 	err = ioat_probe(device);
 	if (err)
 		return err;
-	ioat_set_tcp_copy_break(2048);
 
 	list_for_each_entry(c, &dma->channels, device_node) {
 		chan = to_chan_common(c);
diff --git a/drivers/dma/ioat/dma_v3.c b/drivers/dma/ioat/dma_v3.c
index b9b38a1cf92f..32eae38291e5 100644
--- a/drivers/dma/ioat/dma_v3.c
+++ b/drivers/dma/ioat/dma_v3.c
@@ -740,7 +740,7 @@ ioat3_prep_xor(struct dma_chan *chan, dma_addr_t dest, dma_addr_t *src,
 	return __ioat3_prep_xor_lock(chan, NULL, dest, src, src_cnt, len, flags);
 }
 
-struct dma_async_tx_descriptor *
+static struct dma_async_tx_descriptor *
 ioat3_prep_xor_val(struct dma_chan *chan, dma_addr_t *src,
 		    unsigned int src_cnt, size_t len,
 		    enum sum_check_flags *result, unsigned long flags)
@@ -1091,7 +1091,7 @@ ioat3_prep_pq(struct dma_chan *chan, dma_addr_t *dst, dma_addr_t *src,
 	}
 }
 
-struct dma_async_tx_descriptor *
+static struct dma_async_tx_descriptor *
 ioat3_prep_pq_val(struct dma_chan *chan, dma_addr_t *pq, dma_addr_t *src,
 		  unsigned int src_cnt, const unsigned char *scf, size_t len,
 		  enum sum_check_flags *pqres, unsigned long flags)
@@ -1133,7 +1133,7 @@ ioat3_prep_pqxor(struct dma_chan *chan, dma_addr_t dst, dma_addr_t *src,
 				     flags);
 }
 
-struct dma_async_tx_descriptor *
+static struct dma_async_tx_descriptor *
 ioat3_prep_pqxor_val(struct dma_chan *chan, dma_addr_t *src,
 		     unsigned int src_cnt, size_t len,
 		     enum sum_check_flags *result, unsigned long flags)
@@ -1265,9 +1265,17 @@ static int ioat_xor_val_self_test(struct ioatdma_device *device)
 	op = IOAT_OP_XOR;
 
 	dest_dma = dma_map_page(dev, dest, 0, PAGE_SIZE, DMA_FROM_DEVICE);
+	if (dma_mapping_error(dev, dest_dma))
+		goto dma_unmap;
+
 	for (i = 0; i < IOAT_NUM_SRC_TEST; i++)
+		dma_srcs[i] = DMA_ERROR_CODE;
+	for (i = 0; i < IOAT_NUM_SRC_TEST; i++) {
 		dma_srcs[i] = dma_map_page(dev, xor_srcs[i], 0, PAGE_SIZE,
 					   DMA_TO_DEVICE);
+		if (dma_mapping_error(dev, dma_srcs[i]))
+			goto dma_unmap;
+	}
 	tx = dma->device_prep_dma_xor(dma_chan, dest_dma, dma_srcs,
 				      IOAT_NUM_SRC_TEST, PAGE_SIZE,
 				      DMA_PREP_INTERRUPT);
@@ -1298,7 +1306,6 @@ static int ioat_xor_val_self_test(struct ioatdma_device *device)
 		goto dma_unmap;
 	}
 
-	dma_unmap_page(dev, dest_dma, PAGE_SIZE, DMA_FROM_DEVICE);
 	for (i = 0; i < IOAT_NUM_SRC_TEST; i++)
 		dma_unmap_page(dev, dma_srcs[i], PAGE_SIZE, DMA_TO_DEVICE);
 
@@ -1313,6 +1320,8 @@ static int ioat_xor_val_self_test(struct ioatdma_device *device)
 	}
 	dma_sync_single_for_device(dev, dest_dma, PAGE_SIZE, DMA_FROM_DEVICE);
 
+	dma_unmap_page(dev, dest_dma, PAGE_SIZE, DMA_FROM_DEVICE);
+
 	/* skip validate if the capability is not present */
 	if (!dma_has_cap(DMA_XOR_VAL, dma_chan->device->cap_mask))
 		goto free_resources;
@@ -1327,8 +1336,13 @@ static int ioat_xor_val_self_test(struct ioatdma_device *device)
 	xor_val_result = 1;
 
 	for (i = 0; i < IOAT_NUM_SRC_TEST + 1; i++)
+		dma_srcs[i] = DMA_ERROR_CODE;
+	for (i = 0; i < IOAT_NUM_SRC_TEST + 1; i++) {
 		dma_srcs[i] = dma_map_page(dev, xor_val_srcs[i], 0, PAGE_SIZE,
 					   DMA_TO_DEVICE);
+		if (dma_mapping_error(dev, dma_srcs[i]))
+			goto dma_unmap;
+	}
 	tx = dma->device_prep_dma_xor_val(dma_chan, dma_srcs,
 					  IOAT_NUM_SRC_TEST + 1, PAGE_SIZE,
 					  &xor_val_result, DMA_PREP_INTERRUPT);
@@ -1374,8 +1388,13 @@ static int ioat_xor_val_self_test(struct ioatdma_device *device)
 
 	xor_val_result = 0;
 	for (i = 0; i < IOAT_NUM_SRC_TEST + 1; i++)
+		dma_srcs[i] = DMA_ERROR_CODE;
+	for (i = 0; i < IOAT_NUM_SRC_TEST + 1; i++) {
 		dma_srcs[i] = dma_map_page(dev, xor_val_srcs[i], 0, PAGE_SIZE,
 					   DMA_TO_DEVICE);
+		if (dma_mapping_error(dev, dma_srcs[i]))
+			goto dma_unmap;
+	}
 	tx = dma->device_prep_dma_xor_val(dma_chan, dma_srcs,
 					  IOAT_NUM_SRC_TEST + 1, PAGE_SIZE,
 					  &xor_val_result, DMA_PREP_INTERRUPT);
@@ -1417,14 +1436,18 @@ static int ioat_xor_val_self_test(struct ioatdma_device *device)
 	goto free_resources;
 dma_unmap:
 	if (op == IOAT_OP_XOR) {
-		dma_unmap_page(dev, dest_dma, PAGE_SIZE, DMA_FROM_DEVICE);
+		if (dest_dma != DMA_ERROR_CODE)
+			dma_unmap_page(dev, dest_dma, PAGE_SIZE,
+				       DMA_FROM_DEVICE);
 		for (i = 0; i < IOAT_NUM_SRC_TEST; i++)
-			dma_unmap_page(dev, dma_srcs[i], PAGE_SIZE,
-				       DMA_TO_DEVICE);
+			if (dma_srcs[i] != DMA_ERROR_CODE)
+				dma_unmap_page(dev, dma_srcs[i], PAGE_SIZE,
+					       DMA_TO_DEVICE);
 	} else if (op == IOAT_OP_XOR_VAL) {
 		for (i = 0; i < IOAT_NUM_SRC_TEST + 1; i++)
-			dma_unmap_page(dev, dma_srcs[i], PAGE_SIZE,
-				       DMA_TO_DEVICE);
+			if (dma_srcs[i] != DMA_ERROR_CODE)
+				dma_unmap_page(dev, dma_srcs[i], PAGE_SIZE,
+					       DMA_TO_DEVICE);
 	}
 free_resources:
 	dma->device_free_chan_resources(dma_chan);
@@ -1655,7 +1678,6 @@ int ioat3_dma_probe(struct ioatdma_device *device, int dca)
 	err = ioat_probe(device);
 	if (err)
 		return err;
-	ioat_set_tcp_copy_break(262144);
 
 	list_for_each_entry(c, &dma->channels, device_node) {
 		chan = to_chan_common(c);
diff --git a/drivers/dma/iop-adma.c b/drivers/dma/iop-adma.c
index c56137bc3868..263d9f6a207e 100644
--- a/drivers/dma/iop-adma.c
+++ b/drivers/dma/iop-adma.c
@@ -1557,7 +1557,6 @@ static struct platform_driver iop_adma_driver = {
 	.probe		= iop_adma_probe,
 	.remove		= iop_adma_remove,
 	.driver		= {
-		.owner	= THIS_MODULE,
 		.name	= "iop-adma",
 	},
 };
diff --git a/drivers/dma/iovlock.c b/drivers/dma/iovlock.c
deleted file mode 100644
index bb48a57c2fc1..000000000000
--- a/drivers/dma/iovlock.c
+++ /dev/null
@@ -1,280 +0,0 @@
-/*
- * Copyright(c) 2004 - 2006 Intel Corporation. All rights reserved.
- * Portions based on net/core/datagram.c and copyrighted by their authors.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License as published by the Free
- * Software Foundation; either version 2 of the License, or (at your option)
- * any later version.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc., 59
- * Temple Place - Suite 330, Boston, MA  02111-1307, USA.
- *
- * The full GNU General Public License is included in this distribution in the
- * file called COPYING.
- */
-
-/*
- * This code allows the net stack to make use of a DMA engine for
- * skb to iovec copies.
- */
-
-#include <linux/dmaengine.h>
-#include <linux/pagemap.h>
-#include <linux/slab.h>
-#include <net/tcp.h> /* for memcpy_toiovec */
-#include <asm/io.h>
-#include <asm/uaccess.h>
-
-static int num_pages_spanned(struct iovec *iov)
-{
-	return
-	((PAGE_ALIGN((unsigned long)iov->iov_base + iov->iov_len) -
-	((unsigned long)iov->iov_base & PAGE_MASK)) >> PAGE_SHIFT);
-}
-
-/*
- * Pin down all the iovec pages needed for len bytes.
- * Return a struct dma_pinned_list to keep track of pages pinned down.
- *
- * We are allocating a single chunk of memory, and then carving it up into
- * 3 sections, the latter 2 whose size depends on the number of iovecs and the
- * total number of pages, respectively.
- */
-struct dma_pinned_list *dma_pin_iovec_pages(struct iovec *iov, size_t len)
-{
-	struct dma_pinned_list *local_list;
-	struct page **pages;
-	int i;
-	int ret;
-	int nr_iovecs = 0;
-	int iovec_len_used = 0;
-	int iovec_pages_used = 0;
-
-	/* don't pin down non-user-based iovecs */
-	if (segment_eq(get_fs(), KERNEL_DS))
-		return NULL;
-
-	/* determine how many iovecs/pages there are, up front */
-	do {
-		iovec_len_used += iov[nr_iovecs].iov_len;
-		iovec_pages_used += num_pages_spanned(&iov[nr_iovecs]);
-		nr_iovecs++;
-	} while (iovec_len_used < len);
-
-	/* single kmalloc for pinned list, page_list[], and the page arrays */
-	local_list = kmalloc(sizeof(*local_list)
-		+ (nr_iovecs * sizeof (struct dma_page_list))
-		+ (iovec_pages_used * sizeof (struct page*)), GFP_KERNEL);
-	if (!local_list)
-		goto out;
-
-	/* list of pages starts right after the page list array */
-	pages = (struct page **) &local_list->page_list[nr_iovecs];
-
-	local_list->nr_iovecs = 0;
-
-	for (i = 0; i < nr_iovecs; i++) {
-		struct dma_page_list *page_list = &local_list->page_list[i];
-
-		len -= iov[i].iov_len;
-
-		if (!access_ok(VERIFY_WRITE, iov[i].iov_base, iov[i].iov_len))
-			goto unpin;
-
-		page_list->nr_pages = num_pages_spanned(&iov[i]);
-		page_list->base_address = iov[i].iov_base;
-
-		page_list->pages = pages;
-		pages += page_list->nr_pages;
-
-		/* pin pages down */
-		down_read(&current->mm->mmap_sem);
-		ret = get_user_pages(
-			current,
-			current->mm,
-			(unsigned long) iov[i].iov_base,
-			page_list->nr_pages,
-			1,	/* write */
-			0,	/* force */
-			page_list->pages,
-			NULL);
-		up_read(&current->mm->mmap_sem);
-
-		if (ret != page_list->nr_pages)
-			goto unpin;
-
-		local_list->nr_iovecs = i + 1;
-	}
-
-	return local_list;
-
-unpin:
-	dma_unpin_iovec_pages(local_list);
-out:
-	return NULL;
-}
-
-void dma_unpin_iovec_pages(struct dma_pinned_list *pinned_list)
-{
-	int i, j;
-
-	if (!pinned_list)
-		return;
-
-	for (i = 0; i < pinned_list->nr_iovecs; i++) {
-		struct dma_page_list *page_list = &pinned_list->page_list[i];
-		for (j = 0; j < page_list->nr_pages; j++) {
-			set_page_dirty_lock(page_list->pages[j]);
-			page_cache_release(page_list->pages[j]);
-		}
-	}
-
-	kfree(pinned_list);
-}
-
-
-/*
- * We have already pinned down the pages we will be using in the iovecs.
- * Each entry in iov array has corresponding entry in pinned_list->page_list.
- * Using array indexing to keep iov[] and page_list[] in sync.
- * Initial elements in iov array's iov->iov_len will be 0 if already copied into
- *   by another call.
- * iov array length remaining guaranteed to be bigger than len.
- */
-dma_cookie_t dma_memcpy_to_iovec(struct dma_chan *chan, struct iovec *iov,
-	struct dma_pinned_list *pinned_list, unsigned char *kdata, size_t len)
-{
-	int iov_byte_offset;
-	int copy;
-	dma_cookie_t dma_cookie = 0;
-	int iovec_idx;
-	int page_idx;
-
-	if (!chan)
-		return memcpy_toiovec(iov, kdata, len);
-
-	iovec_idx = 0;
-	while (iovec_idx < pinned_list->nr_iovecs) {
-		struct dma_page_list *page_list;
-
-		/* skip already used-up iovecs */
-		while (!iov[iovec_idx].iov_len)
-			iovec_idx++;
-
-		page_list = &pinned_list->page_list[iovec_idx];
-
-		iov_byte_offset = ((unsigned long)iov[iovec_idx].iov_base & ~PAGE_MASK);
-		page_idx = (((unsigned long)iov[iovec_idx].iov_base & PAGE_MASK)
-			 - ((unsigned long)page_list->base_address & PAGE_MASK)) >> PAGE_SHIFT;
-
-		/* break up copies to not cross page boundary */
-		while (iov[iovec_idx].iov_len) {
-			copy = min_t(int, PAGE_SIZE - iov_byte_offset, len);
-			copy = min_t(int, copy, iov[iovec_idx].iov_len);
-
-			dma_cookie = dma_async_memcpy_buf_to_pg(chan,
-					page_list->pages[page_idx],
-					iov_byte_offset,
-					kdata,
-					copy);
-			/* poll for a descriptor slot */
-			if (unlikely(dma_cookie < 0)) {
-				dma_async_issue_pending(chan);
-				continue;
-			}
-
-			len -= copy;
-			iov[iovec_idx].iov_len -= copy;
-			iov[iovec_idx].iov_base += copy;
-
-			if (!len)
-				return dma_cookie;
-
-			kdata += copy;
-			iov_byte_offset = 0;
-			page_idx++;
-		}
-		iovec_idx++;
-	}
-
-	/* really bad if we ever run out of iovecs */
-	BUG();
-	return -EFAULT;
-}
-
-dma_cookie_t dma_memcpy_pg_to_iovec(struct dma_chan *chan, struct iovec *iov,
-	struct dma_pinned_list *pinned_list, struct page *page,
-	unsigned int offset, size_t len)
-{
-	int iov_byte_offset;
-	int copy;
-	dma_cookie_t dma_cookie = 0;
-	int iovec_idx;
-	int page_idx;
-	int err;
-
-	/* this needs as-yet-unimplemented buf-to-buff, so punt. */
-	/* TODO: use dma for this */
-	if (!chan || !pinned_list) {
-		u8 *vaddr = kmap(page);
-		err = memcpy_toiovec(iov, vaddr + offset, len);
-		kunmap(page);
-		return err;
-	}
-
-	iovec_idx = 0;
-	while (iovec_idx < pinned_list->nr_iovecs) {
-		struct dma_page_list *page_list;
-
-		/* skip already used-up iovecs */
-		while (!iov[iovec_idx].iov_len)
-			iovec_idx++;
-
-		page_list = &pinned_list->page_list[iovec_idx];
-
-		iov_byte_offset = ((unsigned long)iov[iovec_idx].iov_base & ~PAGE_MASK);
-		page_idx = (((unsigned long)iov[iovec_idx].iov_base & PAGE_MASK)
-			 - ((unsigned long)page_list->base_address & PAGE_MASK)) >> PAGE_SHIFT;
-
-		/* break up copies to not cross page boundary */
-		while (iov[iovec_idx].iov_len) {
-			copy = min_t(int, PAGE_SIZE - iov_byte_offset, len);
-			copy = min_t(int, copy, iov[iovec_idx].iov_len);
-
-			dma_cookie = dma_async_memcpy_pg_to_pg(chan,
-					page_list->pages[page_idx],
-					iov_byte_offset,
-					page,
-					offset,
-					copy);
-			/* poll for a descriptor slot */
-			if (unlikely(dma_cookie < 0)) {
-				dma_async_issue_pending(chan);
-				continue;
-			}
-
-			len -= copy;
-			iov[iovec_idx].iov_len -= copy;
-			iov[iovec_idx].iov_base += copy;
-
-			if (!len)
-				return dma_cookie;
-
-			offset += copy;
-			iov_byte_offset = 0;
-			page_idx++;
-		}
-		iovec_idx++;
-	}
-
-	/* really bad if we ever run out of iovecs */
-	BUG();
-	return -EFAULT;
-}
diff --git a/drivers/dma/ipu/ipu_idmac.c b/drivers/dma/ipu/ipu_idmac.c
index bbf62927bd72..c2b017ad139d 100644
--- a/drivers/dma/ipu/ipu_idmac.c
+++ b/drivers/dma/ipu/ipu_idmac.c
@@ -1783,7 +1783,6 @@ static int ipu_remove(struct platform_device *pdev)
 static struct platform_driver ipu_platform_driver = {
 	.driver = {
 		.name	= "ipu-core",
-		.owner	= THIS_MODULE,
 	},
 	.remove		= ipu_remove,
 };
diff --git a/drivers/dma/k3dma.c b/drivers/dma/k3dma.c
index a1f911aaf220..a1de14ab2c51 100644
--- a/drivers/dma/k3dma.c
+++ b/drivers/dma/k3dma.c
@@ -722,7 +722,6 @@ static int k3_dma_probe(struct platform_device *op)
 	d->slave.device_issue_pending = k3_dma_issue_pending;
 	d->slave.device_control = k3_dma_control;
 	d->slave.copy_align = DMA_ALIGN;
-	d->slave.chancnt = d->dma_requests;
 
 	/* init virtual channel */
 	d->chans = devm_kzalloc(&op->dev,
@@ -787,6 +786,7 @@ static int k3_dma_remove(struct platform_device *op)
 	return 0;
 }
 
+#ifdef CONFIG_PM_SLEEP
 static int k3_dma_suspend(struct device *dev)
 {
 	struct k3_dma_dev *d = dev_get_drvdata(dev);
@@ -816,13 +816,13 @@ static int k3_dma_resume(struct device *dev)
 	k3_dma_enable_dma(d, true);
 	return 0;
 }
+#endif
 
 static SIMPLE_DEV_PM_OPS(k3_dma_pmops, k3_dma_suspend, k3_dma_resume);
 
 static struct platform_driver k3_pdma_driver = {
 	.driver		= {
 		.name	= DRIVER_NAME,
-		.owner  = THIS_MODULE,
 		.pm	= &k3_dma_pmops,
 		.of_match_table = k3_pdma_dt_ids,
 	},
diff --git a/drivers/dma/mmp_pdma.c b/drivers/dma/mmp_pdma.c
index a1a4db5721b8..8b8952f35e6c 100644
--- a/drivers/dma/mmp_pdma.c
+++ b/drivers/dma/mmp_pdma.c
@@ -1098,7 +1098,6 @@ static const struct platform_device_id mmp_pdma_id_table[] = {
 static struct platform_driver mmp_pdma_driver = {
 	.driver		= {
 		.name	= "mmp-pdma",
-		.owner  = THIS_MODULE,
 		.of_match_table = mmp_pdma_dt_ids,
 	},
 	.id_table	= mmp_pdma_id_table,
diff --git a/drivers/dma/mmp_tdma.c b/drivers/dma/mmp_tdma.c
index 6ad30e2c5038..bfb46957c3dc 100644
--- a/drivers/dma/mmp_tdma.c
+++ b/drivers/dma/mmp_tdma.c
@@ -148,10 +148,16 @@ static void mmp_tdma_chan_set_desc(struct mmp_tdma_chan *tdmac, dma_addr_t phys)
 					tdmac->reg_base + TDCR);
 }
 
+static void mmp_tdma_enable_irq(struct mmp_tdma_chan *tdmac, bool enable)
+{
+	if (enable)
+		writel(TDIMR_COMP, tdmac->reg_base + TDIMR);
+	else
+		writel(0, tdmac->reg_base + TDIMR);
+}
+
 static void mmp_tdma_enable_chan(struct mmp_tdma_chan *tdmac)
 {
-	/* enable irq */
-	writel(TDIMR_COMP, tdmac->reg_base + TDIMR);
 	/* enable dma chan */
 	writel(readl(tdmac->reg_base + TDCR) | TDCR_CHANEN,
 					tdmac->reg_base + TDCR);
@@ -163,9 +169,6 @@ static void mmp_tdma_disable_chan(struct mmp_tdma_chan *tdmac)
 	writel(readl(tdmac->reg_base + TDCR) & ~TDCR_CHANEN,
 					tdmac->reg_base + TDCR);
 
-	/* disable irq */
-	writel(0, tdmac->reg_base + TDIMR);
-
 	tdmac->status = DMA_COMPLETE;
 }
 
@@ -434,6 +437,10 @@ static struct dma_async_tx_descriptor *mmp_tdma_prep_dma_cyclic(
 		i++;
 	}
 
+	/* enable interrupt */
+	if (flags & DMA_PREP_INTERRUPT)
+		mmp_tdma_enable_irq(tdmac, true);
+
 	tdmac->buf_len = buf_len;
 	tdmac->period_len = period_len;
 	tdmac->pos = 0;
@@ -455,6 +462,8 @@ static int mmp_tdma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
 	switch (cmd) {
 	case DMA_TERMINATE_ALL:
 		mmp_tdma_disable_chan(tdmac);
+		/* disable interrupt */
+		mmp_tdma_enable_irq(tdmac, false);
 		break;
 	case DMA_PAUSE:
 		mmp_tdma_pause_chan(tdmac);
@@ -694,7 +703,6 @@ static const struct platform_device_id mmp_tdma_id_table[] = {
 static struct platform_driver mmp_tdma_driver = {
 	.driver		= {
 		.name	= "mmp-tdma",
-		.owner  = THIS_MODULE,
 		.of_match_table = mmp_tdma_dt_ids,
 	},
 	.id_table	= mmp_tdma_id_table,
diff --git a/drivers/dma/moxart-dma.c b/drivers/dma/moxart-dma.c
index 3258e484e4f6..53032bac06e0 100644
--- a/drivers/dma/moxart-dma.c
+++ b/drivers/dma/moxart-dma.c
@@ -677,7 +677,6 @@ static struct platform_driver moxart_driver = {
 	.remove	= moxart_remove,
 	.driver = {
 		.name		= "moxart-dma-engine",
-		.owner		= THIS_MODULE,
 		.of_match_table	= moxart_dma_match,
 	},
 };
diff --git a/drivers/dma/mpc512x_dma.c b/drivers/dma/mpc512x_dma.c
index 881db2bcb48b..01bec4023de2 100644
--- a/drivers/dma/mpc512x_dma.c
+++ b/drivers/dma/mpc512x_dma.c
@@ -885,6 +885,7 @@ static int mpc_dma_probe(struct platform_device *op)
 	struct resource res;
 	ulong regs_start, regs_size;
 	int retval, i;
+	u8 chancnt;
 
 	mdma = devm_kzalloc(dev, sizeof(struct mpc_dma), GFP_KERNEL);
 	if (!mdma) {
@@ -956,10 +957,6 @@ static int mpc_dma_probe(struct platform_device *op)
 
 	dma = &mdma->dma;
 	dma->dev = dev;
-	if (mdma->is_mpc8308)
-		dma->chancnt = MPC8308_DMACHAN_MAX;
-	else
-		dma->chancnt = MPC512x_DMACHAN_MAX;
 	dma->device_alloc_chan_resources = mpc_dma_alloc_chan_resources;
 	dma->device_free_chan_resources = mpc_dma_free_chan_resources;
 	dma->device_issue_pending = mpc_dma_issue_pending;
@@ -972,7 +969,12 @@ static int mpc_dma_probe(struct platform_device *op)
 	dma_cap_set(DMA_MEMCPY, dma->cap_mask);
 	dma_cap_set(DMA_SLAVE, dma->cap_mask);
 
-	for (i = 0; i < dma->chancnt; i++) {
+	if (mdma->is_mpc8308)
+		chancnt = MPC8308_DMACHAN_MAX;
+	else
+		chancnt = MPC512x_DMACHAN_MAX;
+
+	for (i = 0; i < chancnt; i++) {
 		mchan = &mdma->channels[i];
 
 		mchan->chan.device = dma;
@@ -1090,7 +1092,6 @@ static struct platform_driver mpc_dma_driver = {
 	.remove		= mpc_dma_remove,
 	.driver = {
 		.name = DRV_NAME,
-		.owner = THIS_MODULE,
 		.of_match_table	= mpc_dma_match,
 	},
 };
diff --git a/drivers/dma/mv_xor.c b/drivers/dma/mv_xor.c
index 394cbc5c93e3..d7ac558c2c1c 100644
--- a/drivers/dma/mv_xor.c
+++ b/drivers/dma/mv_xor.c
@@ -45,19 +45,18 @@ static void mv_xor_issue_pending(struct dma_chan *chan);
 #define mv_chan_to_devp(chan)           \
 	((chan)->dmadev.dev)
 
-static void mv_desc_init(struct mv_xor_desc_slot *desc, unsigned long flags)
+static void mv_desc_init(struct mv_xor_desc_slot *desc,
+			 dma_addr_t addr, u32 byte_count,
+			 enum dma_ctrl_flags flags)
 {
 	struct mv_xor_desc *hw_desc = desc->hw_desc;
 
-	hw_desc->status = (1 << 31);
+	hw_desc->status = XOR_DESC_DMA_OWNED;
 	hw_desc->phy_next_desc = 0;
-	hw_desc->desc_command = (1 << 31);
-}
-
-static void mv_desc_set_byte_count(struct mv_xor_desc_slot *desc,
-				   u32 byte_count)
-{
-	struct mv_xor_desc *hw_desc = desc->hw_desc;
+	/* Enable end-of-descriptor interrupts only for DMA_PREP_INTERRUPT */
+	hw_desc->desc_command = (flags & DMA_PREP_INTERRUPT) ?
+				XOR_DESC_EOD_INT_EN : 0;
+	hw_desc->phy_dest_addr = addr;
 	hw_desc->byte_count = byte_count;
 }
 
@@ -75,20 +74,6 @@ static void mv_desc_clear_next_desc(struct mv_xor_desc_slot *desc)
 	hw_desc->phy_next_desc = 0;
 }
 
-static void mv_desc_set_dest_addr(struct mv_xor_desc_slot *desc,
-				  dma_addr_t addr)
-{
-	struct mv_xor_desc *hw_desc = desc->hw_desc;
-	hw_desc->phy_dest_addr = addr;
-}
-
-static int mv_chan_memset_slot_count(size_t len)
-{
-	return 1;
-}
-
-#define mv_chan_memcpy_slot_count(c) mv_chan_memset_slot_count(c)
-
 static void mv_desc_set_src_addr(struct mv_xor_desc_slot *desc,
 				 int index, dma_addr_t addr)
 {
@@ -123,17 +108,12 @@ static u32 mv_chan_get_intr_cause(struct mv_xor_chan *chan)
 	return intr_cause;
 }
 
-static int mv_is_err_intr(u32 intr_cause)
-{
-	if (intr_cause & ((1<<4)|(1<<5)|(1<<6)|(1<<7)|(1<<8)|(1<<9)))
-		return 1;
-
-	return 0;
-}
-
 static void mv_xor_device_clear_eoc_cause(struct mv_xor_chan *chan)
 {
-	u32 val = ~(1 << (chan->idx * 16));
+	u32 val;
+
+	val = XOR_INT_END_OF_DESC | XOR_INT_END_OF_CHAIN | XOR_INT_STOPPED;
+	val = ~(val << (chan->idx * 16));
 	dev_dbg(mv_chan_to_devp(chan), "%s, val 0x%08x\n", __func__, val);
 	writel_relaxed(val, XOR_INTR_CAUSE(chan));
 }
@@ -144,17 +124,6 @@ static void mv_xor_device_clear_err_status(struct mv_xor_chan *chan)
 	writel_relaxed(val, XOR_INTR_CAUSE(chan));
 }
 
-static int mv_can_chain(struct mv_xor_desc_slot *desc)
-{
-	struct mv_xor_desc_slot *chain_old_tail = list_entry(
-		desc->chain_node.prev, struct mv_xor_desc_slot, chain_node);
-
-	if (chain_old_tail->type != desc->type)
-		return 0;
-
-	return 1;
-}
-
 static void mv_set_mode(struct mv_xor_chan *chan,
 			       enum dma_transaction_type type)
 {
@@ -206,11 +175,6 @@ static char mv_chan_is_busy(struct mv_xor_chan *chan)
 	return (state == 1) ? 1 : 0;
 }
 
-static int mv_chan_xor_slot_count(size_t len, int src_cnt)
-{
-	return 1;
-}
-
 /**
  * mv_xor_free_slots - flags descriptor slots for reuse
  * @slot: Slot to free
@@ -222,7 +186,7 @@ static void mv_xor_free_slots(struct mv_xor_chan *mv_chan,
 	dev_dbg(mv_chan_to_devp(mv_chan), "%s %d slot %p\n",
 		__func__, __LINE__, slot);
 
-	slot->slots_per_op = 0;
+	slot->slot_used = 0;
 
 }
 
@@ -236,13 +200,11 @@ static void mv_xor_start_new_chain(struct mv_xor_chan *mv_chan,
 {
 	dev_dbg(mv_chan_to_devp(mv_chan), "%s %d: sw_desc %p\n",
 		__func__, __LINE__, sw_desc);
-	if (sw_desc->type != mv_chan->current_type)
-		mv_set_mode(mv_chan, sw_desc->type);
 
 	/* set the hardware chain */
 	mv_chan_set_next_descriptor(mv_chan, sw_desc->async_tx.phys);
 
-	mv_chan->pending += sw_desc->slot_cnt;
+	mv_chan->pending++;
 	mv_xor_issue_pending(&mv_chan->dmachan);
 }
 
@@ -263,8 +225,6 @@ mv_xor_run_tx_complete_actions(struct mv_xor_desc_slot *desc,
 				desc->async_tx.callback_param);
 
 		dma_descriptor_unmap(&desc->async_tx);
-		if (desc->group_head)
-			desc->group_head = NULL;
 	}
 
 	/* run dependent operations */
@@ -310,7 +270,8 @@ mv_xor_clean_slot(struct mv_xor_desc_slot *desc,
 	return 0;
 }
 
-static void __mv_xor_slot_cleanup(struct mv_xor_chan *mv_chan)
+/* This function must be called with the mv_xor_chan spinlock held */
+static void mv_xor_slot_cleanup(struct mv_xor_chan *mv_chan)
 {
 	struct mv_xor_desc_slot *iter, *_iter;
 	dma_cookie_t cookie = 0;
@@ -366,34 +327,26 @@ static void __mv_xor_slot_cleanup(struct mv_xor_chan *mv_chan)
 		mv_chan->dmachan.completed_cookie = cookie;
 }
 
-static void
-mv_xor_slot_cleanup(struct mv_xor_chan *mv_chan)
-{
-	spin_lock_bh(&mv_chan->lock);
-	__mv_xor_slot_cleanup(mv_chan);
-	spin_unlock_bh(&mv_chan->lock);
-}
-
 static void mv_xor_tasklet(unsigned long data)
 {
 	struct mv_xor_chan *chan = (struct mv_xor_chan *) data;
+
+	spin_lock_bh(&chan->lock);
 	mv_xor_slot_cleanup(chan);
+	spin_unlock_bh(&chan->lock);
 }
 
 static struct mv_xor_desc_slot *
-mv_xor_alloc_slots(struct mv_xor_chan *mv_chan, int num_slots,
-		    int slots_per_op)
+mv_xor_alloc_slot(struct mv_xor_chan *mv_chan)
 {
-	struct mv_xor_desc_slot *iter, *_iter, *alloc_start = NULL;
-	LIST_HEAD(chain);
-	int slots_found, retry = 0;
+	struct mv_xor_desc_slot *iter, *_iter;
+	int retry = 0;
 
 	/* start search from the last allocated descrtiptor
 	 * if a contiguous allocation can not be found start searching
 	 * from the beginning of the list
 	 */
 retry:
-	slots_found = 0;
 	if (retry == 0)
 		iter = mv_chan->last_used;
 	else
@@ -403,55 +356,29 @@ retry:
 
 	list_for_each_entry_safe_continue(
 		iter, _iter, &mv_chan->all_slots, slot_node) {
+
 		prefetch(_iter);
 		prefetch(&_iter->async_tx);
-		if (iter->slots_per_op) {
+		if (iter->slot_used) {
 			/* give up after finding the first busy slot
 			 * on the second pass through the list
 			 */
 			if (retry)
 				break;
-
-			slots_found = 0;
 			continue;
 		}
 
-		/* start the allocation if the slot is correctly aligned */
-		if (!slots_found++)
-			alloc_start = iter;
-
-		if (slots_found == num_slots) {
-			struct mv_xor_desc_slot *alloc_tail = NULL;
-			struct mv_xor_desc_slot *last_used = NULL;
-			iter = alloc_start;
-			while (num_slots) {
-				int i;
-
-				/* pre-ack all but the last descriptor */
-				async_tx_ack(&iter->async_tx);
-
-				list_add_tail(&iter->chain_node, &chain);
-				alloc_tail = iter;
-				iter->async_tx.cookie = 0;
-				iter->slot_cnt = num_slots;
-				iter->xor_check_result = NULL;
-				for (i = 0; i < slots_per_op; i++) {
-					iter->slots_per_op = slots_per_op - i;
-					last_used = iter;
-					iter = list_entry(iter->slot_node.next,
-						struct mv_xor_desc_slot,
-						slot_node);
-				}
-				num_slots -= slots_per_op;
-			}
-			alloc_tail->group_head = alloc_start;
-			alloc_tail->async_tx.cookie = -EBUSY;
-			list_splice(&chain, &alloc_tail->tx_list);
-			mv_chan->last_used = last_used;
-			mv_desc_clear_next_desc(alloc_start);
-			mv_desc_clear_next_desc(alloc_tail);
-			return alloc_tail;
-		}
+		/* pre-ack descriptor */
+		async_tx_ack(&iter->async_tx);
+
+		iter->slot_used = 1;
+		INIT_LIST_HEAD(&iter->chain_node);
+		iter->async_tx.cookie = -EBUSY;
+		mv_chan->last_used = iter;
+		mv_desc_clear_next_desc(iter);
+
+		return iter;
+
 	}
 	if (!retry++)
 		goto retry;
@@ -468,7 +395,7 @@ mv_xor_tx_submit(struct dma_async_tx_descriptor *tx)
 {
 	struct mv_xor_desc_slot *sw_desc = to_mv_xor_slot(tx);
 	struct mv_xor_chan *mv_chan = to_mv_xor_chan(tx->chan);
-	struct mv_xor_desc_slot *grp_start, *old_chain_tail;
+	struct mv_xor_desc_slot *old_chain_tail;
 	dma_cookie_t cookie;
 	int new_hw_chain = 1;
 
@@ -476,30 +403,24 @@ mv_xor_tx_submit(struct dma_async_tx_descriptor *tx)
 		"%s sw_desc %p: async_tx %p\n",
 		__func__, sw_desc, &sw_desc->async_tx);
 
-	grp_start = sw_desc->group_head;
-
 	spin_lock_bh(&mv_chan->lock);
 	cookie = dma_cookie_assign(tx);
 
 	if (list_empty(&mv_chan->chain))
-		list_splice_init(&sw_desc->tx_list, &mv_chan->chain);
+		list_add_tail(&sw_desc->chain_node, &mv_chan->chain);
 	else {
 		new_hw_chain = 0;
 
 		old_chain_tail = list_entry(mv_chan->chain.prev,
 					    struct mv_xor_desc_slot,
 					    chain_node);
-		list_splice_init(&grp_start->tx_list,
-				 &old_chain_tail->chain_node);
-
-		if (!mv_can_chain(grp_start))
-			goto submit_done;
+		list_add_tail(&sw_desc->chain_node, &mv_chan->chain);
 
 		dev_dbg(mv_chan_to_devp(mv_chan), "Append to last desc %pa\n",
 			&old_chain_tail->async_tx.phys);
 
 		/* fix up the hardware chain */
-		mv_desc_set_next_desc(old_chain_tail, grp_start->async_tx.phys);
+		mv_desc_set_next_desc(old_chain_tail, sw_desc->async_tx.phys);
 
 		/* if the channel is not busy */
 		if (!mv_chan_is_busy(mv_chan)) {
@@ -514,9 +435,8 @@ mv_xor_tx_submit(struct dma_async_tx_descriptor *tx)
 	}
 
 	if (new_hw_chain)
-		mv_xor_start_new_chain(mv_chan, grp_start);
+		mv_xor_start_new_chain(mv_chan, sw_desc);
 
-submit_done:
 	spin_unlock_bh(&mv_chan->lock);
 
 	return cookie;
@@ -537,8 +457,9 @@ static int mv_xor_alloc_chan_resources(struct dma_chan *chan)
 	while (idx < num_descs_in_pool) {
 		slot = kzalloc(sizeof(*slot), GFP_KERNEL);
 		if (!slot) {
-			printk(KERN_INFO "MV XOR Channel only initialized"
-				" %d descriptor slots", idx);
+			dev_info(mv_chan_to_devp(mv_chan),
+				 "channel only initialized %d descriptor slots",
+				 idx);
 			break;
 		}
 		virt_desc = mv_chan->dma_desc_pool_virt;
@@ -548,7 +469,6 @@ static int mv_xor_alloc_chan_resources(struct dma_chan *chan)
 		slot->async_tx.tx_submit = mv_xor_tx_submit;
 		INIT_LIST_HEAD(&slot->chain_node);
 		INIT_LIST_HEAD(&slot->slot_node);
-		INIT_LIST_HEAD(&slot->tx_list);
 		dma_desc = mv_chan->dma_desc_pool;
 		slot->async_tx.phys = dma_desc + idx * MV_XOR_SLOT_SIZE;
 		slot->idx = idx++;
@@ -572,51 +492,11 @@ static int mv_xor_alloc_chan_resources(struct dma_chan *chan)
 }
 
 static struct dma_async_tx_descriptor *
-mv_xor_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
-		size_t len, unsigned long flags)
-{
-	struct mv_xor_chan *mv_chan = to_mv_xor_chan(chan);
-	struct mv_xor_desc_slot *sw_desc, *grp_start;
-	int slot_cnt;
-
-	dev_dbg(mv_chan_to_devp(mv_chan),
-		"%s dest: %pad src %pad len: %u flags: %ld\n",
-		__func__, &dest, &src, len, flags);
-	if (unlikely(len < MV_XOR_MIN_BYTE_COUNT))
-		return NULL;
-
-	BUG_ON(len > MV_XOR_MAX_BYTE_COUNT);
-
-	spin_lock_bh(&mv_chan->lock);
-	slot_cnt = mv_chan_memcpy_slot_count(len);
-	sw_desc = mv_xor_alloc_slots(mv_chan, slot_cnt, 1);
-	if (sw_desc) {
-		sw_desc->type = DMA_MEMCPY;
-		sw_desc->async_tx.flags = flags;
-		grp_start = sw_desc->group_head;
-		mv_desc_init(grp_start, flags);
-		mv_desc_set_byte_count(grp_start, len);
-		mv_desc_set_dest_addr(sw_desc->group_head, dest);
-		mv_desc_set_src_addr(grp_start, 0, src);
-		sw_desc->unmap_src_cnt = 1;
-		sw_desc->unmap_len = len;
-	}
-	spin_unlock_bh(&mv_chan->lock);
-
-	dev_dbg(mv_chan_to_devp(mv_chan),
-		"%s sw_desc %p async_tx %p\n",
-		__func__, sw_desc, sw_desc ? &sw_desc->async_tx : NULL);
-
-	return sw_desc ? &sw_desc->async_tx : NULL;
-}
-
-static struct dma_async_tx_descriptor *
 mv_xor_prep_dma_xor(struct dma_chan *chan, dma_addr_t dest, dma_addr_t *src,
 		    unsigned int src_cnt, size_t len, unsigned long flags)
 {
 	struct mv_xor_chan *mv_chan = to_mv_xor_chan(chan);
-	struct mv_xor_desc_slot *sw_desc, *grp_start;
-	int slot_cnt;
+	struct mv_xor_desc_slot *sw_desc;
 
 	if (unlikely(len < MV_XOR_MIN_BYTE_COUNT))
 		return NULL;
@@ -628,20 +508,13 @@ mv_xor_prep_dma_xor(struct dma_chan *chan, dma_addr_t dest, dma_addr_t *src,
 		__func__, src_cnt, len, &dest, flags);
 
 	spin_lock_bh(&mv_chan->lock);
-	slot_cnt = mv_chan_xor_slot_count(len, src_cnt);
-	sw_desc = mv_xor_alloc_slots(mv_chan, slot_cnt, 1);
+	sw_desc = mv_xor_alloc_slot(mv_chan);
 	if (sw_desc) {
 		sw_desc->type = DMA_XOR;
 		sw_desc->async_tx.flags = flags;
-		grp_start = sw_desc->group_head;
-		mv_desc_init(grp_start, flags);
-		/* the byte count field is the same as in memcpy desc*/
-		mv_desc_set_byte_count(grp_start, len);
-		mv_desc_set_dest_addr(sw_desc->group_head, dest);
-		sw_desc->unmap_src_cnt = src_cnt;
-		sw_desc->unmap_len = len;
+		mv_desc_init(sw_desc, dest, len, flags);
 		while (src_cnt--)
-			mv_desc_set_src_addr(grp_start, src_cnt, src[src_cnt]);
+			mv_desc_set_src_addr(sw_desc, src_cnt, src[src_cnt]);
 	}
 	spin_unlock_bh(&mv_chan->lock);
 	dev_dbg(mv_chan_to_devp(mv_chan),
@@ -650,15 +523,45 @@ mv_xor_prep_dma_xor(struct dma_chan *chan, dma_addr_t dest, dma_addr_t *src,
 	return sw_desc ? &sw_desc->async_tx : NULL;
 }
 
+static struct dma_async_tx_descriptor *
+mv_xor_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
+		size_t len, unsigned long flags)
+{
+	/*
+	 * A MEMCPY operation is identical to an XOR operation with only
+	 * a single source address.
+	 */
+	return mv_xor_prep_dma_xor(chan, dest, &src, 1, len, flags);
+}
+
+static struct dma_async_tx_descriptor *
+mv_xor_prep_dma_interrupt(struct dma_chan *chan, unsigned long flags)
+{
+	struct mv_xor_chan *mv_chan = to_mv_xor_chan(chan);
+	dma_addr_t src, dest;
+	size_t len;
+
+	src = mv_chan->dummy_src_addr;
+	dest = mv_chan->dummy_dst_addr;
+	len = MV_XOR_MIN_BYTE_COUNT;
+
+	/*
+	 * We implement the DMA_INTERRUPT operation as a minimum sized
+	 * XOR operation with a single dummy source address.
+	 */
+	return mv_xor_prep_dma_xor(chan, dest, &src, 1, len, flags);
+}
+
 static void mv_xor_free_chan_resources(struct dma_chan *chan)
 {
 	struct mv_xor_chan *mv_chan = to_mv_xor_chan(chan);
 	struct mv_xor_desc_slot *iter, *_iter;
 	int in_use_descs = 0;
 
+	spin_lock_bh(&mv_chan->lock);
+
 	mv_xor_slot_cleanup(mv_chan);
 
-	spin_lock_bh(&mv_chan->lock);
 	list_for_each_entry_safe(iter, _iter, &mv_chan->chain,
 					chain_node) {
 		in_use_descs++;
@@ -700,11 +603,12 @@ static enum dma_status mv_xor_status(struct dma_chan *chan,
 	enum dma_status ret;
 
 	ret = dma_cookie_status(chan, cookie, txstate);
-	if (ret == DMA_COMPLETE) {
-		mv_xor_clean_completed_slots(mv_chan);
+	if (ret == DMA_COMPLETE)
 		return ret;
-	}
+
+	spin_lock_bh(&mv_chan->lock);
 	mv_xor_slot_cleanup(mv_chan);
+	spin_unlock_bh(&mv_chan->lock);
 
 	return dma_cookie_status(chan, cookie, txstate);
 }
@@ -735,18 +639,16 @@ static void mv_dump_xor_regs(struct mv_xor_chan *chan)
 static void mv_xor_err_interrupt_handler(struct mv_xor_chan *chan,
 					 u32 intr_cause)
 {
-	if (intr_cause & (1 << 4)) {
-	     dev_dbg(mv_chan_to_devp(chan),
-		     "ignore this error\n");
-	     return;
+	if (intr_cause & XOR_INT_ERR_DECODE) {
+		dev_dbg(mv_chan_to_devp(chan), "ignoring address decode error\n");
+		return;
 	}
 
-	dev_err(mv_chan_to_devp(chan),
-		"error on chan %d. intr cause 0x%08x\n",
+	dev_err(mv_chan_to_devp(chan), "error on chan %d. intr cause 0x%08x\n",
 		chan->idx, intr_cause);
 
 	mv_dump_xor_regs(chan);
-	BUG();
+	WARN_ON(1);
 }
 
 static irqreturn_t mv_xor_interrupt_handler(int irq, void *data)
@@ -756,7 +658,7 @@ static irqreturn_t mv_xor_interrupt_handler(int irq, void *data)
 
 	dev_dbg(mv_chan_to_devp(chan), "intr cause %x\n", intr_cause);
 
-	if (mv_is_err_intr(intr_cause))
+	if (intr_cause & XOR_INTR_ERRORS)
 		mv_xor_err_interrupt_handler(chan, intr_cause);
 
 	tasklet_schedule(&chan->irq_tasklet);
@@ -782,7 +684,7 @@ static void mv_xor_issue_pending(struct dma_chan *chan)
 
 static int mv_xor_memcpy_self_test(struct mv_xor_chan *mv_chan)
 {
-	int i;
+	int i, ret;
 	void *src, *dest;
 	dma_addr_t src_dma, dest_dma;
 	struct dma_chan *dma_chan;
@@ -819,19 +721,44 @@ static int mv_xor_memcpy_self_test(struct mv_xor_chan *mv_chan)
 
 	src_dma = dma_map_page(dma_chan->device->dev, virt_to_page(src), 0,
 				 PAGE_SIZE, DMA_TO_DEVICE);
-	unmap->to_cnt = 1;
 	unmap->addr[0] = src_dma;
 
+	ret = dma_mapping_error(dma_chan->device->dev, src_dma);
+	if (ret) {
+		err = -ENOMEM;
+		goto free_resources;
+	}
+	unmap->to_cnt = 1;
+
 	dest_dma = dma_map_page(dma_chan->device->dev, virt_to_page(dest), 0,
 				  PAGE_SIZE, DMA_FROM_DEVICE);
-	unmap->from_cnt = 1;
 	unmap->addr[1] = dest_dma;
 
+	ret = dma_mapping_error(dma_chan->device->dev, dest_dma);
+	if (ret) {
+		err = -ENOMEM;
+		goto free_resources;
+	}
+	unmap->from_cnt = 1;
 	unmap->len = PAGE_SIZE;
 
 	tx = mv_xor_prep_dma_memcpy(dma_chan, dest_dma, src_dma,
 				    PAGE_SIZE, 0);
+	if (!tx) {
+		dev_err(dma_chan->device->dev,
+			"Self-test cannot prepare operation, disabling\n");
+		err = -ENODEV;
+		goto free_resources;
+	}
+
 	cookie = mv_xor_tx_submit(tx);
+	if (dma_submit_error(cookie)) {
+		dev_err(dma_chan->device->dev,
+			"Self-test submit error, disabling\n");
+		err = -ENODEV;
+		goto free_resources;
+	}
+
 	mv_xor_issue_pending(dma_chan);
 	async_tx_ack(tx);
 	msleep(1);
@@ -866,7 +793,7 @@ out:
 static int
 mv_xor_xor_self_test(struct mv_xor_chan *mv_chan)
 {
-	int i, src_idx;
+	int i, src_idx, ret;
 	struct page *dest;
 	struct page *xor_srcs[MV_XOR_NUM_SRC_TEST];
 	dma_addr_t dma_srcs[MV_XOR_NUM_SRC_TEST];
@@ -929,19 +856,42 @@ mv_xor_xor_self_test(struct mv_xor_chan *mv_chan)
 		unmap->addr[i] = dma_map_page(dma_chan->device->dev, xor_srcs[i],
 					      0, PAGE_SIZE, DMA_TO_DEVICE);
 		dma_srcs[i] = unmap->addr[i];
+		ret = dma_mapping_error(dma_chan->device->dev, unmap->addr[i]);
+		if (ret) {
+			err = -ENOMEM;
+			goto free_resources;
+		}
 		unmap->to_cnt++;
 	}
 
 	unmap->addr[src_count] = dma_map_page(dma_chan->device->dev, dest, 0, PAGE_SIZE,
 				      DMA_FROM_DEVICE);
 	dest_dma = unmap->addr[src_count];
+	ret = dma_mapping_error(dma_chan->device->dev, unmap->addr[src_count]);
+	if (ret) {
+		err = -ENOMEM;
+		goto free_resources;
+	}
 	unmap->from_cnt = 1;
 	unmap->len = PAGE_SIZE;
 
 	tx = mv_xor_prep_dma_xor(dma_chan, dest_dma, dma_srcs,
 				 src_count, PAGE_SIZE, 0);
+	if (!tx) {
+		dev_err(dma_chan->device->dev,
+			"Self-test cannot prepare operation, disabling\n");
+		err = -ENODEV;
+		goto free_resources;
+	}
 
 	cookie = mv_xor_tx_submit(tx);
+	if (dma_submit_error(cookie)) {
+		dev_err(dma_chan->device->dev,
+			"Self-test submit error, disabling\n");
+		err = -ENODEV;
+		goto free_resources;
+	}
+
 	mv_xor_issue_pending(dma_chan);
 	async_tx_ack(tx);
 	msleep(8);
@@ -995,6 +945,10 @@ static int mv_xor_channel_remove(struct mv_xor_chan *mv_chan)
 
 	dma_free_coherent(dev, MV_XOR_POOL_SIZE,
 			  mv_chan->dma_desc_pool_virt, mv_chan->dma_desc_pool);
+	dma_unmap_single(dev, mv_chan->dummy_src_addr,
+			 MV_XOR_MIN_BYTE_COUNT, DMA_FROM_DEVICE);
+	dma_unmap_single(dev, mv_chan->dummy_dst_addr,
+			 MV_XOR_MIN_BYTE_COUNT, DMA_TO_DEVICE);
 
 	list_for_each_entry_safe(chan, _chan, &mv_chan->dmadev.channels,
 				 device_node) {
@@ -1024,6 +978,16 @@ mv_xor_channel_add(struct mv_xor_device *xordev,
 
 	dma_dev = &mv_chan->dmadev;
 
+	/*
+	 * These source and destination dummy buffers are used to implement
+	 * a DMA_INTERRUPT operation as a minimum-sized XOR operation.
+	 * Hence, we only need to map the buffers at initialization-time.
+	 */
+	mv_chan->dummy_src_addr = dma_map_single(dma_dev->dev,
+		mv_chan->dummy_src, MV_XOR_MIN_BYTE_COUNT, DMA_FROM_DEVICE);
+	mv_chan->dummy_dst_addr = dma_map_single(dma_dev->dev,
+		mv_chan->dummy_dst, MV_XOR_MIN_BYTE_COUNT, DMA_TO_DEVICE);
+
 	/* allocate coherent memory for hardware descriptors
 	 * note: writecombine gives slightly better performance, but
 	 * requires that we explicitly flush the writes
@@ -1048,6 +1012,8 @@ mv_xor_channel_add(struct mv_xor_device *xordev,
 	dma_dev->dev = &pdev->dev;
 
 	/* set prep routines based on capability */
+	if (dma_has_cap(DMA_INTERRUPT, dma_dev->cap_mask))
+		dma_dev->device_prep_dma_interrupt = mv_xor_prep_dma_interrupt;
 	if (dma_has_cap(DMA_MEMCPY, dma_dev->cap_mask))
 		dma_dev->device_prep_dma_memcpy = mv_xor_prep_dma_memcpy;
 	if (dma_has_cap(DMA_XOR, dma_dev->cap_mask)) {
@@ -1070,7 +1036,7 @@ mv_xor_channel_add(struct mv_xor_device *xordev,
 
 	mv_chan_unmask_interrupts(mv_chan);
 
-	mv_set_mode(mv_chan, DMA_MEMCPY);
+	mv_set_mode(mv_chan, DMA_XOR);
 
 	spin_lock_init(&mv_chan->lock);
 	INIT_LIST_HEAD(&mv_chan->chain);
@@ -1303,7 +1269,6 @@ static struct platform_driver mv_xor_driver = {
 	.probe		= mv_xor_probe,
 	.remove		= mv_xor_remove,
 	.driver		= {
-		.owner	        = THIS_MODULE,
 		.name	        = MV_XOR_NAME,
 		.of_match_table = of_match_ptr(mv_xor_dt_ids),
 	},
diff --git a/drivers/dma/mv_xor.h b/drivers/dma/mv_xor.h
index d0749229c875..78edc7e44569 100644
--- a/drivers/dma/mv_xor.h
+++ b/drivers/dma/mv_xor.h
@@ -23,17 +23,22 @@
 #include <linux/dmaengine.h>
 #include <linux/interrupt.h>
 
-#define USE_TIMER
 #define MV_XOR_POOL_SIZE		PAGE_SIZE
 #define MV_XOR_SLOT_SIZE		64
 #define MV_XOR_THRESHOLD		1
 #define MV_XOR_MAX_CHANNELS             2
 
+#define MV_XOR_MIN_BYTE_COUNT		SZ_128
+#define MV_XOR_MAX_BYTE_COUNT		(SZ_16M - 1)
+
 /* Values for the XOR_CONFIG register */
 #define XOR_OPERATION_MODE_XOR		0
 #define XOR_OPERATION_MODE_MEMCPY	2
 #define XOR_DESCRIPTOR_SWAP		BIT(14)
 
+#define XOR_DESC_DMA_OWNED		BIT(31)
+#define XOR_DESC_EOD_INT_EN		BIT(31)
+
 #define XOR_CURR_DESC(chan)	(chan->mmr_high_base + 0x10 + (chan->idx * 4))
 #define XOR_NEXT_DESC(chan)	(chan->mmr_high_base + 0x00 + (chan->idx * 4))
 #define XOR_BYTE_COUNT(chan)	(chan->mmr_high_base + 0x20 + (chan->idx * 4))
@@ -48,7 +53,24 @@
 #define XOR_INTR_MASK(chan)	(chan->mmr_base + 0x40)
 #define XOR_ERROR_CAUSE(chan)	(chan->mmr_base + 0x50)
 #define XOR_ERROR_ADDR(chan)	(chan->mmr_base + 0x60)
-#define XOR_INTR_MASK_VALUE	0x3F5
+
+#define XOR_INT_END_OF_DESC	BIT(0)
+#define XOR_INT_END_OF_CHAIN	BIT(1)
+#define XOR_INT_STOPPED		BIT(2)
+#define XOR_INT_PAUSED		BIT(3)
+#define XOR_INT_ERR_DECODE	BIT(4)
+#define XOR_INT_ERR_RDPROT	BIT(5)
+#define XOR_INT_ERR_WRPROT	BIT(6)
+#define XOR_INT_ERR_OWN		BIT(7)
+#define XOR_INT_ERR_PAR		BIT(8)
+#define XOR_INT_ERR_MBUS	BIT(9)
+
+#define XOR_INTR_ERRORS		(XOR_INT_ERR_DECODE | XOR_INT_ERR_RDPROT | \
+				 XOR_INT_ERR_WRPROT | XOR_INT_ERR_OWN    | \
+				 XOR_INT_ERR_PAR    | XOR_INT_ERR_MBUS)
+
+#define XOR_INTR_MASK_VALUE	(XOR_INT_END_OF_DESC | XOR_INT_END_OF_CHAIN | \
+				 XOR_INT_STOPPED     | XOR_INTR_ERRORS)
 
 #define WINDOW_BASE(w)		(0x50 + ((w) << 2))
 #define WINDOW_SIZE(w)		(0x70 + ((w) << 2))
@@ -97,10 +119,9 @@ struct mv_xor_chan {
 	struct list_head	all_slots;
 	int			slots_allocated;
 	struct tasklet_struct	irq_tasklet;
-#ifdef USE_TIMER
-	unsigned long		cleanup_time;
-	u32			current_on_last_cleanup;
-#endif
+	char			dummy_src[MV_XOR_MIN_BYTE_COUNT];
+	char			dummy_dst[MV_XOR_MIN_BYTE_COUNT];
+	dma_addr_t		dummy_src_addr, dummy_dst_addr;
 };
 
 /**
@@ -110,16 +131,10 @@ struct mv_xor_chan {
  * @completed_node: node on the mv_xor_chan.completed_slots list
  * @hw_desc: virtual address of the hardware descriptor chain
  * @phys: hardware address of the hardware descriptor chain
- * @group_head: first operation in a transaction
- * @slot_cnt: total slots used in an transaction (group of operations)
- * @slots_per_op: number of slots per operation
+ * @slot_used: slot in use or not
  * @idx: pool index
- * @unmap_src_cnt: number of xor sources
- * @unmap_len: transaction bytecount
  * @tx_list: list of slots that make up a multi-descriptor transaction
  * @async_tx: support for the async_tx api
- * @xor_check_result: result of zero sum
- * @crc32_result: result crc calculation
  */
 struct mv_xor_desc_slot {
 	struct list_head	slot_node;
@@ -127,23 +142,9 @@ struct mv_xor_desc_slot {
 	struct list_head	completed_node;
 	enum dma_transaction_type	type;
 	void			*hw_desc;
-	struct mv_xor_desc_slot	*group_head;
-	u16			slot_cnt;
-	u16			slots_per_op;
+	u16			slot_used;
 	u16			idx;
-	u16			unmap_src_cnt;
-	u32			value;
-	size_t			unmap_len;
-	struct list_head	tx_list;
 	struct dma_async_tx_descriptor	async_tx;
-	union {
-		u32		*xor_check_result;
-		u32		*crc32_result;
-	};
-#ifdef USE_TIMER
-	unsigned long		arrival_time;
-	struct timer_list	timeout;
-#endif
 };
 
 /*
@@ -189,9 +190,4 @@ struct mv_xor_desc {
 #define mv_hw_desc_slot_idx(hw_desc, idx)	\
 	((void *)(((unsigned long)hw_desc) + ((idx) << 5)))
 
-#define MV_XOR_MIN_BYTE_COUNT	(128)
-#define XOR_MAX_BYTE_COUNT	((16 * 1024 * 1024) - 1)
-#define MV_XOR_MAX_BYTE_COUNT	XOR_MAX_BYTE_COUNT
-
-
 #endif
diff --git a/drivers/dma/nbpfaxi.c b/drivers/dma/nbpfaxi.c
index 5aeada56a442..d7d61e1a01c3 100644
--- a/drivers/dma/nbpfaxi.c
+++ b/drivers/dma/nbpfaxi.c
@@ -1479,7 +1479,7 @@ static struct platform_device_id nbpf_ids[] = {
 };
 MODULE_DEVICE_TABLE(platform, nbpf_ids);
 
-#ifdef CONFIG_PM_RUNTIME
+#ifdef CONFIG_PM
 static int nbpf_runtime_suspend(struct device *dev)
 {
 	struct nbpf_device *nbpf = platform_get_drvdata(to_platform_device(dev));
@@ -1500,7 +1500,6 @@ static const struct dev_pm_ops nbpf_pm_ops = {
 
 static struct platform_driver nbpf_driver = {
 	.driver = {
-		.owner = THIS_MODULE,
 		.name = "dma-nbpf",
 		.of_match_table = nbpf_match,
 		.pm = &nbpf_pm_ops,
diff --git a/drivers/dma/omap-dma.c b/drivers/dma/omap-dma.c
index bbea8243f9e8..c0016a68b446 100644
--- a/drivers/dma/omap-dma.c
+++ b/drivers/dma/omap-dma.c
@@ -1074,8 +1074,6 @@ static int omap_dma_chan_init(struct omap_dmadev *od, int dma_sig)
 	vchan_init(&c->vc, &od->ddev);
 	INIT_LIST_HEAD(&c->node);
 
-	od->ddev.chancnt++;
-
 	return 0;
 }
 
@@ -1233,7 +1231,6 @@ static struct platform_driver omap_dma_driver = {
 	.remove	= omap_dma_remove,
 	.driver = {
 		.name = "omap-dma-engine",
-		.owner = THIS_MODULE,
 		.of_match_table = of_match_ptr(omap_dma_match),
 	},
 };
diff --git a/drivers/dma/pch_dma.c b/drivers/dma/pch_dma.c
index 9f9ca9fe5ce6..6e0e47d76b23 100644
--- a/drivers/dma/pch_dma.c
+++ b/drivers/dma/pch_dma.c
@@ -997,7 +997,7 @@ static void pch_dma_remove(struct pci_dev *pdev)
 #define PCI_DEVICE_ID_ML7831_DMA1_8CH	0x8810
 #define PCI_DEVICE_ID_ML7831_DMA2_4CH	0x8815
 
-const struct pci_device_id pch_dma_id_table[] = {
+static const struct pci_device_id pch_dma_id_table[] = {
 	{ PCI_VDEVICE(INTEL, PCI_DEVICE_ID_EG20T_PCH_DMA_8CH), 8 },
 	{ PCI_VDEVICE(INTEL, PCI_DEVICE_ID_EG20T_PCH_DMA_4CH), 4 },
 	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7213_DMA1_8CH), 8}, /* UART Video */
diff --git a/drivers/dma/pl330.c b/drivers/dma/pl330.c
index d5149aacd2fe..bdf40b530032 100644
--- a/drivers/dma/pl330.c
+++ b/drivers/dma/pl330.c
@@ -27,6 +27,7 @@
 #include <linux/of.h>
 #include <linux/of_dma.h>
 #include <linux/err.h>
+#include <linux/pm_runtime.h>
 
 #include "dmaengine.h"
 #define PL330_MAX_CHAN		8
@@ -265,13 +266,16 @@ static unsigned cmd_line;
 
 #define NR_DEFAULT_DESC	16
 
+/* Delay for runtime PM autosuspend, ms */
+#define PL330_AUTOSUSPEND_DELAY 20
+
 /* Populated by the PL330 core driver for DMA API driver's info */
 struct pl330_config {
 	u32	periph_id;
 #define DMAC_MODE_NS	(1 << 0)
 	unsigned int	mode;
 	unsigned int	data_bus_width:10; /* In number of bits */
-	unsigned int	data_buf_dep:10;
+	unsigned int	data_buf_dep:11;
 	unsigned int	num_chan:4;
 	unsigned int	num_peri:6;
 	u32		peri_ns;
@@ -1367,17 +1371,10 @@ static int pl330_submit_req(struct pl330_thread *thrd,
 	struct pl330_dmac *pl330 = thrd->dmac;
 	struct _xfer_spec xs;
 	unsigned long flags;
-	void __iomem *regs;
 	unsigned idx;
 	u32 ccr;
 	int ret = 0;
 
-	/* No Req or Unacquired Channel or DMAC */
-	if (!desc || !thrd || thrd->free)
-		return -EINVAL;
-
-	regs = thrd->dmac->base;
-
 	if (pl330->state == DYING
 		|| pl330->dmac_tbd.reset_chan & (1 << thrd->id)) {
 		dev_info(thrd->dmac->ddma.dev, "%s:%d\n",
@@ -1965,6 +1962,7 @@ static void pl330_tasklet(unsigned long data)
 	struct dma_pl330_chan *pch = (struct dma_pl330_chan *)data;
 	struct dma_pl330_desc *desc, *_dt;
 	unsigned long flags;
+	bool power_down = false;
 
 	spin_lock_irqsave(&pch->lock, flags);
 
@@ -1979,10 +1977,17 @@ static void pl330_tasklet(unsigned long data)
 	/* Try to submit a req imm. next to the last completed cookie */
 	fill_queue(pch);
 
-	/* Make sure the PL330 Channel thread is active */
-	spin_lock(&pch->thread->dmac->lock);
-	_start(pch->thread);
-	spin_unlock(&pch->thread->dmac->lock);
+	if (list_empty(&pch->work_list)) {
+		spin_lock(&pch->thread->dmac->lock);
+		_stop(pch->thread);
+		spin_unlock(&pch->thread->dmac->lock);
+		power_down = true;
+	} else {
+		/* Make sure the PL330 Channel thread is active */
+		spin_lock(&pch->thread->dmac->lock);
+		_start(pch->thread);
+		spin_unlock(&pch->thread->dmac->lock);
+	}
 
 	while (!list_empty(&pch->completed_list)) {
 		dma_async_tx_callback callback;
@@ -1997,6 +2002,12 @@ static void pl330_tasklet(unsigned long data)
 		if (pch->cyclic) {
 			desc->status = PREP;
 			list_move_tail(&desc->node, &pch->work_list);
+			if (power_down) {
+				spin_lock(&pch->thread->dmac->lock);
+				_start(pch->thread);
+				spin_unlock(&pch->thread->dmac->lock);
+				power_down = false;
+			}
 		} else {
 			desc->status = FREE;
 			list_move_tail(&desc->node, &pch->dmac->desc_pool);
@@ -2011,6 +2022,12 @@ static void pl330_tasklet(unsigned long data)
 		}
 	}
 	spin_unlock_irqrestore(&pch->lock, flags);
+
+	/* If work list empty, power down */
+	if (power_down) {
+		pm_runtime_mark_last_busy(pch->dmac->ddma.dev);
+		pm_runtime_put_autosuspend(pch->dmac->ddma.dev);
+	}
 }
 
 bool pl330_filter(struct dma_chan *chan, void *param)
@@ -2080,6 +2097,7 @@ static int pl330_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd, unsigned
 
 	switch (cmd) {
 	case DMA_TERMINATE_ALL:
+		pm_runtime_get_sync(pl330->ddma.dev);
 		spin_lock_irqsave(&pch->lock, flags);
 
 		spin_lock(&pl330->lock);
@@ -2106,10 +2124,15 @@ static int pl330_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd, unsigned
 			dma_cookie_complete(&desc->txd);
 		}
 
+		if (!list_empty(&pch->work_list))
+			pm_runtime_put(pl330->ddma.dev);
+
 		list_splice_tail_init(&pch->submitted_list, &pl330->desc_pool);
 		list_splice_tail_init(&pch->work_list, &pl330->desc_pool);
 		list_splice_tail_init(&pch->completed_list, &pl330->desc_pool);
 		spin_unlock_irqrestore(&pch->lock, flags);
+		pm_runtime_mark_last_busy(pl330->ddma.dev);
+		pm_runtime_put_autosuspend(pl330->ddma.dev);
 		break;
 	case DMA_SLAVE_CONFIG:
 		slave_config = (struct dma_slave_config *)arg;
@@ -2145,6 +2168,7 @@ static void pl330_free_chan_resources(struct dma_chan *chan)
 
 	tasklet_kill(&pch->task);
 
+	pm_runtime_get_sync(pch->dmac->ddma.dev);
 	spin_lock_irqsave(&pch->lock, flags);
 
 	pl330_release_channel(pch->thread);
@@ -2154,6 +2178,8 @@ static void pl330_free_chan_resources(struct dma_chan *chan)
 		list_splice_tail_init(&pch->work_list, &pch->dmac->desc_pool);
 
 	spin_unlock_irqrestore(&pch->lock, flags);
+	pm_runtime_mark_last_busy(pch->dmac->ddma.dev);
+	pm_runtime_put_autosuspend(pch->dmac->ddma.dev);
 }
 
 static enum dma_status
@@ -2169,6 +2195,15 @@ static void pl330_issue_pending(struct dma_chan *chan)
 	unsigned long flags;
 
 	spin_lock_irqsave(&pch->lock, flags);
+	if (list_empty(&pch->work_list)) {
+		/*
+		 * Warn on nothing pending. Empty submitted_list may
+		 * break our pm_runtime usage counter as it is
+		 * updated on work_list emptiness status.
+		 */
+		WARN_ON(list_empty(&pch->submitted_list));
+		pm_runtime_get_sync(pch->dmac->ddma.dev);
+	}
 	list_splice_tail_init(&pch->submitted_list, &pch->work_list);
 	spin_unlock_irqrestore(&pch->lock, flags);
 
@@ -2343,7 +2378,7 @@ static inline int get_burst_len(struct dma_pl330_desc *desc, size_t len)
 	int burst_len;
 
 	burst_len = pl330->pcfg.data_bus_width / 8;
-	burst_len *= pl330->pcfg.data_buf_dep;
+	burst_len *= pl330->pcfg.data_buf_dep / pl330->pcfg.num_chan;
 	burst_len >>= desc->rqcfg.brst_size;
 
 	/* src/dst_burst_len can't be more than 16 */
@@ -2466,16 +2501,25 @@ pl330_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dst,
 	/* Select max possible burst size */
 	burst = pl330->pcfg.data_bus_width / 8;
 
-	while (burst > 1) {
-		if (!(len % burst))
-			break;
+	/*
+	 * Make sure we use a burst size that aligns with all the memcpy
+	 * parameters because our DMA programming algorithm doesn't cope with
+	 * transfers which straddle an entry in the DMA device's MFIFO.
+	 */
+	while ((src | dst | len) & (burst - 1))
 		burst /= 2;
-	}
 
 	desc->rqcfg.brst_size = 0;
 	while (burst != (1 << desc->rqcfg.brst_size))
 		desc->rqcfg.brst_size++;
 
+	/*
+	 * If burst size is smaller than bus width then make sure we only
+	 * transfer one at a time to avoid a burst stradling an MFIFO entry.
+	 */
+	if (desc->rqcfg.brst_size * 8 < pl330->pcfg.data_bus_width)
+		desc->rqcfg.brst_len = 1;
+
 	desc->rqcfg.brst_len = get_burst_len(desc, len);
 
 	desc->txd.flags = flags;
@@ -2592,6 +2636,46 @@ static int pl330_dma_device_slave_caps(struct dma_chan *dchan,
 	return 0;
 }
 
+/*
+ * Runtime PM callbacks are provided by amba/bus.c driver.
+ *
+ * It is assumed here that IRQ safe runtime PM is chosen in probe and amba
+ * bus driver will only disable/enable the clock in runtime PM callbacks.
+ */
+static int __maybe_unused pl330_suspend(struct device *dev)
+{
+	struct amba_device *pcdev = to_amba_device(dev);
+
+	pm_runtime_disable(dev);
+
+	if (!pm_runtime_status_suspended(dev)) {
+		/* amba did not disable the clock */
+		amba_pclk_disable(pcdev);
+	}
+	amba_pclk_unprepare(pcdev);
+
+	return 0;
+}
+
+static int __maybe_unused pl330_resume(struct device *dev)
+{
+	struct amba_device *pcdev = to_amba_device(dev);
+	int ret;
+
+	ret = amba_pclk_prepare(pcdev);
+	if (ret)
+		return ret;
+
+	if (!pm_runtime_status_suspended(dev))
+		ret = amba_pclk_enable(pcdev);
+
+	pm_runtime_enable(dev);
+
+	return ret;
+}
+
+static SIMPLE_DEV_PM_OPS(pl330_pm, pl330_suspend, pl330_resume);
+
 static int
 pl330_probe(struct amba_device *adev, const struct amba_id *id)
 {
@@ -2617,6 +2701,9 @@ pl330_probe(struct amba_device *adev, const struct amba_id *id)
 		return -ENOMEM;
 	}
 
+	pd = &pl330->ddma;
+	pd->dev = &adev->dev;
+
 	pl330->mcbufsz = pdat ? pdat->mcbuf_sz : 0;
 
 	res = &adev->res;
@@ -2653,7 +2740,6 @@ pl330_probe(struct amba_device *adev, const struct amba_id *id)
 	if (!add_desc(pl330, GFP_KERNEL, NR_DEFAULT_DESC))
 		dev_warn(&adev->dev, "unable to allocate desc\n");
 
-	pd = &pl330->ddma;
 	INIT_LIST_HEAD(&pd->channels);
 
 	/* Initialize channel parameters */
@@ -2690,7 +2776,6 @@ pl330_probe(struct amba_device *adev, const struct amba_id *id)
 		list_add_tail(&pch->chan.device_node, &pd->channels);
 	}
 
-	pd->dev = &adev->dev;
 	if (pdat) {
 		pd->cap_mask = pdat->cap_mask;
 	} else {
@@ -2739,12 +2824,18 @@ pl330_probe(struct amba_device *adev, const struct amba_id *id)
 
 
 	dev_info(&adev->dev,
-		"Loaded driver for PL330 DMAC-%d\n", adev->periphid);
+		"Loaded driver for PL330 DMAC-%x\n", adev->periphid);
 	dev_info(&adev->dev,
 		"\tDBUFF-%ux%ubytes Num_Chans-%u Num_Peri-%u Num_Events-%u\n",
 		pcfg->data_buf_dep, pcfg->data_bus_width / 8, pcfg->num_chan,
 		pcfg->num_peri, pcfg->num_events);
 
+	pm_runtime_irq_safe(&adev->dev);
+	pm_runtime_use_autosuspend(&adev->dev);
+	pm_runtime_set_autosuspend_delay(&adev->dev, PL330_AUTOSUSPEND_DELAY);
+	pm_runtime_mark_last_busy(&adev->dev);
+	pm_runtime_put_autosuspend(&adev->dev);
+
 	return 0;
 probe_err3:
 	/* Idle the DMAC */
@@ -2755,8 +2846,10 @@ probe_err3:
 		list_del(&pch->chan.device_node);
 
 		/* Flush the channel */
-		pl330_control(&pch->chan, DMA_TERMINATE_ALL, 0);
-		pl330_free_chan_resources(&pch->chan);
+		if (pch->thread) {
+			pl330_control(&pch->chan, DMA_TERMINATE_ALL, 0);
+			pl330_free_chan_resources(&pch->chan);
+		}
 	}
 probe_err2:
 	pl330_del(pl330);
@@ -2769,6 +2862,8 @@ static int pl330_remove(struct amba_device *adev)
 	struct pl330_dmac *pl330 = amba_get_drvdata(adev);
 	struct dma_pl330_chan *pch, *_p;
 
+	pm_runtime_get_noresume(pl330->ddma.dev);
+
 	if (adev->dev.of_node)
 		of_dma_controller_free(adev->dev.of_node);
 
@@ -2782,8 +2877,10 @@ static int pl330_remove(struct amba_device *adev)
 		list_del(&pch->chan.device_node);
 
 		/* Flush the channel */
-		pl330_control(&pch->chan, DMA_TERMINATE_ALL, 0);
-		pl330_free_chan_resources(&pch->chan);
+		if (pch->thread) {
+			pl330_control(&pch->chan, DMA_TERMINATE_ALL, 0);
+			pl330_free_chan_resources(&pch->chan);
+		}
 	}
 
 	pl330_del(pl330);
@@ -2805,6 +2902,7 @@ static struct amba_driver pl330_driver = {
 	.drv = {
 		.owner = THIS_MODULE,
 		.name = "dma-pl330",
+		.pm = &pl330_pm,
 	},
 	.id_table = pl330_ids,
 	.probe = pl330_probe,
@@ -2813,6 +2911,6 @@ static struct amba_driver pl330_driver = {
 
 module_amba_driver(pl330_driver);
 
-MODULE_AUTHOR("Jaswinder Singh <jassi.brar@samsung.com>");
+MODULE_AUTHOR("Jaswinder Singh <jassisinghbrar@gmail.com>");
 MODULE_DESCRIPTION("API Driver for PL330 DMAC");
 MODULE_LICENSE("GPL");
diff --git a/drivers/dma/ppc4xx/adma.c b/drivers/dma/ppc4xx/adma.c
index ce7a8d7564ba..fa764a39cd36 100644
--- a/drivers/dma/ppc4xx/adma.c
+++ b/drivers/dma/ppc4xx/adma.c
@@ -4579,7 +4579,6 @@ static struct platform_driver ppc440spe_adma_driver = {
 	.remove = ppc440spe_adma_remove,
 	.driver = {
 		.name = "PPC440SP(E)-ADMA",
-		.owner = THIS_MODULE,
 		.of_match_table = ppc440spe_adma_of_match,
 	},
 };
diff --git a/drivers/dma/qcom_bam_dma.c b/drivers/dma/qcom_bam_dma.c
index 7a4bbb0f80a5..3122a99ec06b 100644
--- a/drivers/dma/qcom_bam_dma.c
+++ b/drivers/dma/qcom_bam_dma.c
@@ -79,35 +79,97 @@ struct bam_async_desc {
 	struct bam_desc_hw desc[0];
 };
 
-#define BAM_CTRL			0x0000
-#define BAM_REVISION			0x0004
-#define BAM_SW_REVISION			0x0080
-#define BAM_NUM_PIPES			0x003C
-#define BAM_TIMER			0x0040
-#define BAM_TIMER_CTRL			0x0044
-#define BAM_DESC_CNT_TRSHLD		0x0008
-#define BAM_IRQ_SRCS			0x000C
-#define BAM_IRQ_SRCS_MSK		0x0010
-#define BAM_IRQ_SRCS_UNMASKED		0x0030
-#define BAM_IRQ_STTS			0x0014
-#define BAM_IRQ_CLR			0x0018
-#define BAM_IRQ_EN			0x001C
-#define BAM_CNFG_BITS			0x007C
-#define BAM_IRQ_SRCS_EE(ee)		(0x0800 + ((ee) * 0x80))
-#define BAM_IRQ_SRCS_MSK_EE(ee)		(0x0804 + ((ee) * 0x80))
-#define BAM_P_CTRL(pipe)		(0x1000 + ((pipe) * 0x1000))
-#define BAM_P_RST(pipe)			(0x1004 + ((pipe) * 0x1000))
-#define BAM_P_HALT(pipe)		(0x1008 + ((pipe) * 0x1000))
-#define BAM_P_IRQ_STTS(pipe)		(0x1010 + ((pipe) * 0x1000))
-#define BAM_P_IRQ_CLR(pipe)		(0x1014 + ((pipe) * 0x1000))
-#define BAM_P_IRQ_EN(pipe)		(0x1018 + ((pipe) * 0x1000))
-#define BAM_P_EVNT_DEST_ADDR(pipe)	(0x182C + ((pipe) * 0x1000))
-#define BAM_P_EVNT_REG(pipe)		(0x1818 + ((pipe) * 0x1000))
-#define BAM_P_SW_OFSTS(pipe)		(0x1800 + ((pipe) * 0x1000))
-#define BAM_P_DATA_FIFO_ADDR(pipe)	(0x1824 + ((pipe) * 0x1000))
-#define BAM_P_DESC_FIFO_ADDR(pipe)	(0x181C + ((pipe) * 0x1000))
-#define BAM_P_EVNT_TRSHLD(pipe)		(0x1828 + ((pipe) * 0x1000))
-#define BAM_P_FIFO_SIZES(pipe)		(0x1820 + ((pipe) * 0x1000))
+enum bam_reg {
+	BAM_CTRL,
+	BAM_REVISION,
+	BAM_NUM_PIPES,
+	BAM_DESC_CNT_TRSHLD,
+	BAM_IRQ_SRCS,
+	BAM_IRQ_SRCS_MSK,
+	BAM_IRQ_SRCS_UNMASKED,
+	BAM_IRQ_STTS,
+	BAM_IRQ_CLR,
+	BAM_IRQ_EN,
+	BAM_CNFG_BITS,
+	BAM_IRQ_SRCS_EE,
+	BAM_IRQ_SRCS_MSK_EE,
+	BAM_P_CTRL,
+	BAM_P_RST,
+	BAM_P_HALT,
+	BAM_P_IRQ_STTS,
+	BAM_P_IRQ_CLR,
+	BAM_P_IRQ_EN,
+	BAM_P_EVNT_DEST_ADDR,
+	BAM_P_EVNT_REG,
+	BAM_P_SW_OFSTS,
+	BAM_P_DATA_FIFO_ADDR,
+	BAM_P_DESC_FIFO_ADDR,
+	BAM_P_EVNT_GEN_TRSHLD,
+	BAM_P_FIFO_SIZES,
+};
+
+struct reg_offset_data {
+	u32 base_offset;
+	unsigned int pipe_mult, evnt_mult, ee_mult;
+};
+
+static const struct reg_offset_data bam_v1_3_reg_info[] = {
+	[BAM_CTRL]		= { 0x0F80, 0x00, 0x00, 0x00 },
+	[BAM_REVISION]		= { 0x0F84, 0x00, 0x00, 0x00 },
+	[BAM_NUM_PIPES]		= { 0x0FBC, 0x00, 0x00, 0x00 },
+	[BAM_DESC_CNT_TRSHLD]	= { 0x0F88, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_SRCS]		= { 0x0F8C, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_SRCS_MSK]	= { 0x0F90, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_SRCS_UNMASKED]	= { 0x0FB0, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_STTS]		= { 0x0F94, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_CLR]		= { 0x0F98, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_EN]		= { 0x0F9C, 0x00, 0x00, 0x00 },
+	[BAM_CNFG_BITS]		= { 0x0FFC, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_SRCS_EE]	= { 0x1800, 0x00, 0x00, 0x80 },
+	[BAM_IRQ_SRCS_MSK_EE]	= { 0x1804, 0x00, 0x00, 0x80 },
+	[BAM_P_CTRL]		= { 0x0000, 0x80, 0x00, 0x00 },
+	[BAM_P_RST]		= { 0x0004, 0x80, 0x00, 0x00 },
+	[BAM_P_HALT]		= { 0x0008, 0x80, 0x00, 0x00 },
+	[BAM_P_IRQ_STTS]	= { 0x0010, 0x80, 0x00, 0x00 },
+	[BAM_P_IRQ_CLR]		= { 0x0014, 0x80, 0x00, 0x00 },
+	[BAM_P_IRQ_EN]		= { 0x0018, 0x80, 0x00, 0x00 },
+	[BAM_P_EVNT_DEST_ADDR]	= { 0x102C, 0x00, 0x40, 0x00 },
+	[BAM_P_EVNT_REG]	= { 0x1018, 0x00, 0x40, 0x00 },
+	[BAM_P_SW_OFSTS]	= { 0x1000, 0x00, 0x40, 0x00 },
+	[BAM_P_DATA_FIFO_ADDR]	= { 0x1024, 0x00, 0x40, 0x00 },
+	[BAM_P_DESC_FIFO_ADDR]	= { 0x101C, 0x00, 0x40, 0x00 },
+	[BAM_P_EVNT_GEN_TRSHLD]	= { 0x1028, 0x00, 0x40, 0x00 },
+	[BAM_P_FIFO_SIZES]	= { 0x1020, 0x00, 0x40, 0x00 },
+};
+
+static const struct reg_offset_data bam_v1_4_reg_info[] = {
+	[BAM_CTRL]		= { 0x0000, 0x00, 0x00, 0x00 },
+	[BAM_REVISION]		= { 0x0004, 0x00, 0x00, 0x00 },
+	[BAM_NUM_PIPES]		= { 0x003C, 0x00, 0x00, 0x00 },
+	[BAM_DESC_CNT_TRSHLD]	= { 0x0008, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_SRCS]		= { 0x000C, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_SRCS_MSK]	= { 0x0010, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_SRCS_UNMASKED]	= { 0x0030, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_STTS]		= { 0x0014, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_CLR]		= { 0x0018, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_EN]		= { 0x001C, 0x00, 0x00, 0x00 },
+	[BAM_CNFG_BITS]		= { 0x007C, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_SRCS_EE]	= { 0x0800, 0x00, 0x00, 0x80 },
+	[BAM_IRQ_SRCS_MSK_EE]	= { 0x0804, 0x00, 0x00, 0x80 },
+	[BAM_P_CTRL]		= { 0x1000, 0x1000, 0x00, 0x00 },
+	[BAM_P_RST]		= { 0x1004, 0x1000, 0x00, 0x00 },
+	[BAM_P_HALT]		= { 0x1008, 0x1000, 0x00, 0x00 },
+	[BAM_P_IRQ_STTS]	= { 0x1010, 0x1000, 0x00, 0x00 },
+	[BAM_P_IRQ_CLR]		= { 0x1014, 0x1000, 0x00, 0x00 },
+	[BAM_P_IRQ_EN]		= { 0x1018, 0x1000, 0x00, 0x00 },
+	[BAM_P_EVNT_DEST_ADDR]	= { 0x102C, 0x00, 0x1000, 0x00 },
+	[BAM_P_EVNT_REG]	= { 0x1018, 0x00, 0x1000, 0x00 },
+	[BAM_P_SW_OFSTS]	= { 0x1000, 0x00, 0x1000, 0x00 },
+	[BAM_P_DATA_FIFO_ADDR]	= { 0x1824, 0x00, 0x1000, 0x00 },
+	[BAM_P_DESC_FIFO_ADDR]	= { 0x181C, 0x00, 0x1000, 0x00 },
+	[BAM_P_EVNT_GEN_TRSHLD]	= { 0x1828, 0x00, 0x1000, 0x00 },
+	[BAM_P_FIFO_SIZES]	= { 0x1820, 0x00, 0x1000, 0x00 },
+};
 
 /* BAM CTRL */
 #define BAM_SW_RST			BIT(0)
@@ -297,6 +359,8 @@ struct bam_device {
 	/* execution environment ID, from DT */
 	u32 ee;
 
+	const struct reg_offset_data *layout;
+
 	struct clk *bamclk;
 	int irq;
 
@@ -305,6 +369,23 @@ struct bam_device {
 };
 
 /**
+ * bam_addr - returns BAM register address
+ * @bdev: bam device
+ * @pipe: pipe instance (ignored when register doesn't have multiple instances)
+ * @reg:  register enum
+ */
+static inline void __iomem *bam_addr(struct bam_device *bdev, u32 pipe,
+		enum bam_reg reg)
+{
+	const struct reg_offset_data r = bdev->layout[reg];
+
+	return bdev->regs + r.base_offset +
+		r.pipe_mult * pipe +
+		r.evnt_mult * pipe +
+		r.ee_mult * bdev->ee;
+}
+
+/**
  * bam_reset_channel - Reset individual BAM DMA channel
  * @bchan: bam channel
  *
@@ -317,8 +398,8 @@ static void bam_reset_channel(struct bam_chan *bchan)
 	lockdep_assert_held(&bchan->vc.lock);
 
 	/* reset channel */
-	writel_relaxed(1, bdev->regs + BAM_P_RST(bchan->id));
-	writel_relaxed(0, bdev->regs + BAM_P_RST(bchan->id));
+	writel_relaxed(1, bam_addr(bdev, bchan->id, BAM_P_RST));
+	writel_relaxed(0, bam_addr(bdev, bchan->id, BAM_P_RST));
 
 	/* don't allow cpu to reorder BAM register accesses done after this */
 	wmb();
@@ -347,17 +428,18 @@ static void bam_chan_init_hw(struct bam_chan *bchan,
 	 * because we allocated 1 more descriptor (8 bytes) than we can use
 	 */
 	writel_relaxed(ALIGN(bchan->fifo_phys, sizeof(struct bam_desc_hw)),
-			bdev->regs + BAM_P_DESC_FIFO_ADDR(bchan->id));
-	writel_relaxed(BAM_DESC_FIFO_SIZE, bdev->regs +
-			BAM_P_FIFO_SIZES(bchan->id));
+			bam_addr(bdev, bchan->id, BAM_P_DESC_FIFO_ADDR));
+	writel_relaxed(BAM_DESC_FIFO_SIZE,
+			bam_addr(bdev, bchan->id, BAM_P_FIFO_SIZES));
 
 	/* enable the per pipe interrupts, enable EOT, ERR, and INT irqs */
-	writel_relaxed(P_DEFAULT_IRQS_EN, bdev->regs + BAM_P_IRQ_EN(bchan->id));
+	writel_relaxed(P_DEFAULT_IRQS_EN,
+			bam_addr(bdev, bchan->id, BAM_P_IRQ_EN));
 
 	/* unmask the specific pipe and EE combo */
-	val = readl_relaxed(bdev->regs + BAM_IRQ_SRCS_MSK_EE(bdev->ee));
+	val = readl_relaxed(bam_addr(bdev, 0, BAM_IRQ_SRCS_MSK_EE));
 	val |= BIT(bchan->id);
-	writel_relaxed(val, bdev->regs + BAM_IRQ_SRCS_MSK_EE(bdev->ee));
+	writel_relaxed(val, bam_addr(bdev, 0, BAM_IRQ_SRCS_MSK_EE));
 
 	/* don't allow cpu to reorder the channel enable done below */
 	wmb();
@@ -367,7 +449,7 @@ static void bam_chan_init_hw(struct bam_chan *bchan,
 	if (dir == DMA_DEV_TO_MEM)
 		val |= P_DIRECTION;
 
-	writel_relaxed(val, bdev->regs + BAM_P_CTRL(bchan->id));
+	writel_relaxed(val, bam_addr(bdev, bchan->id, BAM_P_CTRL));
 
 	bchan->initialized = 1;
 
@@ -432,12 +514,12 @@ static void bam_free_chan(struct dma_chan *chan)
 	bchan->fifo_virt = NULL;
 
 	/* mask irq for pipe/channel */
-	val = readl_relaxed(bdev->regs + BAM_IRQ_SRCS_MSK_EE(bdev->ee));
+	val = readl_relaxed(bam_addr(bdev, 0, BAM_IRQ_SRCS_MSK_EE));
 	val &= ~BIT(bchan->id);
-	writel_relaxed(val, bdev->regs + BAM_IRQ_SRCS_MSK_EE(bdev->ee));
+	writel_relaxed(val, bam_addr(bdev, 0, BAM_IRQ_SRCS_MSK_EE));
 
 	/* disable irq */
-	writel_relaxed(0, bdev->regs + BAM_P_IRQ_EN(bchan->id));
+	writel_relaxed(0, bam_addr(bdev, bchan->id, BAM_P_IRQ_EN));
 }
 
 /**
@@ -583,14 +665,14 @@ static int bam_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
 	switch (cmd) {
 	case DMA_PAUSE:
 		spin_lock_irqsave(&bchan->vc.lock, flag);
-		writel_relaxed(1, bdev->regs + BAM_P_HALT(bchan->id));
+		writel_relaxed(1, bam_addr(bdev, bchan->id, BAM_P_HALT));
 		bchan->paused = 1;
 		spin_unlock_irqrestore(&bchan->vc.lock, flag);
 		break;
 
 	case DMA_RESUME:
 		spin_lock_irqsave(&bchan->vc.lock, flag);
-		writel_relaxed(0, bdev->regs + BAM_P_HALT(bchan->id));
+		writel_relaxed(0, bam_addr(bdev, bchan->id, BAM_P_HALT));
 		bchan->paused = 0;
 		spin_unlock_irqrestore(&bchan->vc.lock, flag);
 		break;
@@ -626,7 +708,7 @@ static u32 process_channel_irqs(struct bam_device *bdev)
 	unsigned long flags;
 	struct bam_async_desc *async_desc;
 
-	srcs = readl_relaxed(bdev->regs + BAM_IRQ_SRCS_EE(bdev->ee));
+	srcs = readl_relaxed(bam_addr(bdev, 0, BAM_IRQ_SRCS_EE));
 
 	/* return early if no pipe/channel interrupts are present */
 	if (!(srcs & P_IRQ))
@@ -639,11 +721,9 @@ static u32 process_channel_irqs(struct bam_device *bdev)
 			continue;
 
 		/* clear pipe irq */
-		pipe_stts = readl_relaxed(bdev->regs +
-			BAM_P_IRQ_STTS(i));
+		pipe_stts = readl_relaxed(bam_addr(bdev, i, BAM_P_IRQ_STTS));
 
-		writel_relaxed(pipe_stts, bdev->regs +
-				BAM_P_IRQ_CLR(i));
+		writel_relaxed(pipe_stts, bam_addr(bdev, i, BAM_P_IRQ_CLR));
 
 		spin_lock_irqsave(&bchan->vc.lock, flags);
 		async_desc = bchan->curr_txd;
@@ -694,12 +774,12 @@ static irqreturn_t bam_dma_irq(int irq, void *data)
 		tasklet_schedule(&bdev->task);
 
 	if (srcs & BAM_IRQ)
-		clr_mask = readl_relaxed(bdev->regs + BAM_IRQ_STTS);
+		clr_mask = readl_relaxed(bam_addr(bdev, 0, BAM_IRQ_STTS));
 
 	/* don't allow reorder of the various accesses to the BAM registers */
 	mb();
 
-	writel_relaxed(clr_mask, bdev->regs + BAM_IRQ_CLR);
+	writel_relaxed(clr_mask, bam_addr(bdev, 0, BAM_IRQ_CLR));
 
 	return IRQ_HANDLED;
 }
@@ -763,7 +843,7 @@ static void bam_apply_new_config(struct bam_chan *bchan,
 	else
 		maxburst = bchan->slave.dst_maxburst;
 
-	writel_relaxed(maxburst, bdev->regs + BAM_DESC_CNT_TRSHLD);
+	writel_relaxed(maxburst, bam_addr(bdev, 0, BAM_DESC_CNT_TRSHLD));
 
 	bchan->reconfigure = 0;
 }
@@ -830,7 +910,7 @@ static void bam_start_dma(struct bam_chan *bchan)
 	/* ensure descriptor writes and dma start not reordered */
 	wmb();
 	writel_relaxed(bchan->tail * sizeof(struct bam_desc_hw),
-			bdev->regs + BAM_P_EVNT_REG(bchan->id));
+			bam_addr(bdev, bchan->id, BAM_P_EVNT_REG));
 }
 
 /**
@@ -918,43 +998,44 @@ static int bam_init(struct bam_device *bdev)
 	u32 val;
 
 	/* read revision and configuration information */
-	val = readl_relaxed(bdev->regs + BAM_REVISION) >> NUM_EES_SHIFT;
+	val = readl_relaxed(bam_addr(bdev, 0, BAM_REVISION)) >> NUM_EES_SHIFT;
 	val &= NUM_EES_MASK;
 
 	/* check that configured EE is within range */
 	if (bdev->ee >= val)
 		return -EINVAL;
 
-	val = readl_relaxed(bdev->regs + BAM_NUM_PIPES);
+	val = readl_relaxed(bam_addr(bdev, 0, BAM_NUM_PIPES));
 	bdev->num_channels = val & BAM_NUM_PIPES_MASK;
 
 	/* s/w reset bam */
 	/* after reset all pipes are disabled and idle */
-	val = readl_relaxed(bdev->regs + BAM_CTRL);
+	val = readl_relaxed(bam_addr(bdev, 0, BAM_CTRL));
 	val |= BAM_SW_RST;
-	writel_relaxed(val, bdev->regs + BAM_CTRL);
+	writel_relaxed(val, bam_addr(bdev, 0, BAM_CTRL));
 	val &= ~BAM_SW_RST;
-	writel_relaxed(val, bdev->regs + BAM_CTRL);
+	writel_relaxed(val, bam_addr(bdev, 0, BAM_CTRL));
 
 	/* make sure previous stores are visible before enabling BAM */
 	wmb();
 
 	/* enable bam */
 	val |= BAM_EN;
-	writel_relaxed(val, bdev->regs + BAM_CTRL);
+	writel_relaxed(val, bam_addr(bdev, 0, BAM_CTRL));
 
 	/* set descriptor threshhold, start with 4 bytes */
-	writel_relaxed(DEFAULT_CNT_THRSHLD, bdev->regs + BAM_DESC_CNT_TRSHLD);
+	writel_relaxed(DEFAULT_CNT_THRSHLD,
+			bam_addr(bdev, 0, BAM_DESC_CNT_TRSHLD));
 
 	/* Enable default set of h/w workarounds, ie all except BAM_FULL_PIPE */
-	writel_relaxed(BAM_CNFG_BITS_DEFAULT, bdev->regs + BAM_CNFG_BITS);
+	writel_relaxed(BAM_CNFG_BITS_DEFAULT, bam_addr(bdev, 0, BAM_CNFG_BITS));
 
 	/* enable irqs for errors */
 	writel_relaxed(BAM_ERROR_EN | BAM_HRESP_ERR_EN,
-				bdev->regs + BAM_IRQ_EN);
+			bam_addr(bdev, 0, BAM_IRQ_EN));
 
 	/* unmask global bam interrupt */
-	writel_relaxed(BAM_IRQ_MSK, bdev->regs + BAM_IRQ_SRCS_MSK_EE(bdev->ee));
+	writel_relaxed(BAM_IRQ_MSK, bam_addr(bdev, 0, BAM_IRQ_SRCS_MSK_EE));
 
 	return 0;
 }
@@ -969,9 +1050,18 @@ static void bam_channel_init(struct bam_device *bdev, struct bam_chan *bchan,
 	bchan->vc.desc_free = bam_dma_free_desc;
 }
 
+static const struct of_device_id bam_of_match[] = {
+	{ .compatible = "qcom,bam-v1.3.0", .data = &bam_v1_3_reg_info },
+	{ .compatible = "qcom,bam-v1.4.0", .data = &bam_v1_4_reg_info },
+	{}
+};
+
+MODULE_DEVICE_TABLE(of, bam_of_match);
+
 static int bam_dma_probe(struct platform_device *pdev)
 {
 	struct bam_device *bdev;
+	const struct of_device_id *match;
 	struct resource *iores;
 	int ret, i;
 
@@ -981,6 +1071,14 @@ static int bam_dma_probe(struct platform_device *pdev)
 
 	bdev->dev = &pdev->dev;
 
+	match = of_match_node(bam_of_match, pdev->dev.of_node);
+	if (!match) {
+		dev_err(&pdev->dev, "Unsupported BAM module\n");
+		return -ENODEV;
+	}
+
+	bdev->layout = match->data;
+
 	iores = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	bdev->regs = devm_ioremap_resource(&pdev->dev, iores);
 	if (IS_ERR(bdev->regs))
@@ -1084,7 +1182,7 @@ static int bam_dma_remove(struct platform_device *pdev)
 	dma_async_device_unregister(&bdev->common);
 
 	/* mask all interrupts for this execution environment */
-	writel_relaxed(0, bdev->regs + BAM_IRQ_SRCS_MSK_EE(bdev->ee));
+	writel_relaxed(0, bam_addr(bdev, 0,  BAM_IRQ_SRCS_MSK_EE));
 
 	devm_free_irq(bdev->dev, bdev->irq, bdev);
 
@@ -1104,18 +1202,11 @@ static int bam_dma_remove(struct platform_device *pdev)
 	return 0;
 }
 
-static const struct of_device_id bam_of_match[] = {
-	{ .compatible = "qcom,bam-v1.4.0", },
-	{}
-};
-MODULE_DEVICE_TABLE(of, bam_of_match);
-
 static struct platform_driver bam_dma_driver = {
 	.probe = bam_dma_probe,
 	.remove = bam_dma_remove,
 	.driver = {
 		.name = "bam-dma-engine",
-		.owner = THIS_MODULE,
 		.of_match_table = bam_of_match,
 	},
 };
diff --git a/drivers/dma/s3c24xx-dma.c b/drivers/dma/s3c24xx-dma.c
index 7416572d1e40..6941a77521c3 100644
--- a/drivers/dma/s3c24xx-dma.c
+++ b/drivers/dma/s3c24xx-dma.c
@@ -1402,7 +1402,6 @@ static int s3c24xx_dma_remove(struct platform_device *pdev)
 static struct platform_driver s3c24xx_dma_driver = {
 	.driver		= {
 		.name	= "s3c24xx-dma",
-		.owner	= THIS_MODULE,
 	},
 	.id_table	= s3c24xx_dma_driver_ids,
 	.probe		= s3c24xx_dma_probe,
diff --git a/drivers/dma/sa11x0-dma.c b/drivers/dma/sa11x0-dma.c
index 4b0ef043729a..96bb62c39c41 100644
--- a/drivers/dma/sa11x0-dma.c
+++ b/drivers/dma/sa11x0-dma.c
@@ -829,7 +829,6 @@ static int sa11x0_dma_init_dmadev(struct dma_device *dmadev,
 {
 	unsigned i;
 
-	dmadev->chancnt = ARRAY_SIZE(chan_desc);
 	INIT_LIST_HEAD(&dmadev->channels);
 	dmadev->dev = dev;
 	dmadev->device_alloc_chan_resources = sa11x0_dma_alloc_chan_resources;
@@ -838,7 +837,7 @@ static int sa11x0_dma_init_dmadev(struct dma_device *dmadev,
 	dmadev->device_tx_status = sa11x0_dma_tx_status;
 	dmadev->device_issue_pending = sa11x0_dma_issue_pending;
 
-	for (i = 0; i < dmadev->chancnt; i++) {
+	for (i = 0; i < ARRAY_SIZE(chan_desc); i++) {
 		struct sa11x0_dma_chan *c;
 
 		c = kzalloc(sizeof(*c), GFP_KERNEL);
@@ -1064,7 +1063,6 @@ static const struct dev_pm_ops sa11x0_dma_pm_ops = {
 static struct platform_driver sa11x0_dma_driver = {
 	.driver = {
 		.name	= "sa11x0-dma",
-		.owner	= THIS_MODULE,
 		.pm	= &sa11x0_dma_pm_ops,
 	},
 	.probe		= sa11x0_dma_probe,
diff --git a/drivers/dma/sh/rcar-audmapp.c b/drivers/dma/sh/rcar-audmapp.c
index dabbf0aba2e9..d95bbdd721f4 100644
--- a/drivers/dma/sh/rcar-audmapp.c
+++ b/drivers/dma/sh/rcar-audmapp.c
@@ -117,7 +117,7 @@ static void audmapp_start_xfer(struct shdma_chan *schan,
 	audmapp_write(auchan, chcr,	PDMACHCR);
 }
 
-static void audmapp_get_config(struct audmapp_chan *auchan, int slave_id,
+static int audmapp_get_config(struct audmapp_chan *auchan, int slave_id,
 			      u32 *chcr, dma_addr_t *dst)
 {
 	struct audmapp_device *audev = to_dev(auchan);
@@ -131,20 +131,22 @@ static void audmapp_get_config(struct audmapp_chan *auchan, int slave_id,
 	if (!pdata) { /* DT */
 		*chcr = ((u32)slave_id) << 16;
 		auchan->shdma_chan.slave_id = (slave_id) >> 8;
-		return;
+		return 0;
 	}
 
 	/* non-DT */
 
 	if (slave_id >= AUDMAPP_SLAVE_NUMBER)
-		return;
+		return -ENXIO;
 
 	for (i = 0, cfg = pdata->slave; i < pdata->slave_num; i++, cfg++)
 		if (cfg->slave_id == slave_id) {
 			*chcr	= cfg->chcr;
 			*dst	= cfg->dst;
-			break;
+			return 0;
 		}
+
+	return -ENXIO;
 }
 
 static int audmapp_set_slave(struct shdma_chan *schan, int slave_id,
@@ -153,8 +155,11 @@ static int audmapp_set_slave(struct shdma_chan *schan, int slave_id,
 	struct audmapp_chan *auchan = to_chan(schan);
 	u32 chcr;
 	dma_addr_t dst;
+	int ret;
 
-	audmapp_get_config(auchan, slave_id, &chcr, &dst);
+	ret = audmapp_get_config(auchan, slave_id, &chcr, &dst);
+	if (ret < 0)
+		return ret;
 
 	if (try)
 		return 0;
@@ -248,7 +253,6 @@ static int audmapp_chan_probe(struct platform_device *pdev,
 
 static void audmapp_chan_remove(struct audmapp_device *audev)
 {
-	struct dma_device *dma_dev = &audev->shdma_dev.dma_dev;
 	struct shdma_chan *schan;
 	int i;
 
@@ -256,7 +260,6 @@ static void audmapp_chan_remove(struct audmapp_device *audev)
 		BUG_ON(!schan);
 		shdma_chan_remove(schan);
 	}
-	dma_dev->chancnt = 0;
 }
 
 static struct dma_chan *audmapp_of_xlate(struct of_phandle_args *dma_spec,
@@ -362,7 +365,6 @@ static struct platform_driver audmapp_driver = {
 	.probe		= audmapp_probe,
 	.remove		= audmapp_remove,
 	.driver		= {
-		.owner	= THIS_MODULE,
 		.name	= "rcar-audmapp-engine",
 		.of_match_table = audmapp_of_match,
 	},
diff --git a/drivers/dma/sh/rcar-hpbdma.c b/drivers/dma/sh/rcar-hpbdma.c
index b212d9471ab5..20a6f6f2a018 100644
--- a/drivers/dma/sh/rcar-hpbdma.c
+++ b/drivers/dma/sh/rcar-hpbdma.c
@@ -619,7 +619,6 @@ error:
 
 static void hpb_dmae_chan_remove(struct hpb_dmae_device *hpbdev)
 {
-	struct dma_device *dma_dev = &hpbdev->shdma_dev.dma_dev;
 	struct shdma_chan *schan;
 	int i;
 
@@ -628,7 +627,6 @@ static void hpb_dmae_chan_remove(struct hpb_dmae_device *hpbdev)
 
 		shdma_chan_remove(schan);
 	}
-	dma_dev->chancnt = 0;
 }
 
 static int hpb_dmae_remove(struct platform_device *pdev)
@@ -655,7 +653,6 @@ static struct platform_driver hpb_dmae_driver = {
 	.remove		= hpb_dmae_remove,
 	.shutdown	= hpb_dmae_shutdown,
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= "hpb-dma-engine",
 	},
 };
diff --git a/drivers/dma/sh/shdma-base.c b/drivers/dma/sh/shdma-base.c
index 42d497416196..3a2adb131d46 100644
--- a/drivers/dma/sh/shdma-base.c
+++ b/drivers/dma/sh/shdma-base.c
@@ -391,6 +391,8 @@ static dma_async_tx_callback __ld_cleanup(struct shdma_chan *schan, bool all)
 				dev_dbg(schan->dev, "Bring down channel %d\n", schan->id);
 				pm_runtime_put(schan->dev);
 				schan->pm_state = SHDMA_PM_ESTABLISHED;
+			} else if (schan->pm_state == SHDMA_PM_PENDING) {
+				shdma_chan_xfer_ld_queue(schan);
 			}
 		}
 	}
@@ -951,7 +953,7 @@ void shdma_chan_probe(struct shdma_dev *sdev,
 	/* Add the channel to DMA device channel list */
 	list_add_tail(&schan->dma_chan.device_node,
 			&sdev->dma_dev.channels);
-	sdev->schan[sdev->dma_dev.chancnt++] = schan;
+	sdev->schan[id] = schan;
 }
 EXPORT_SYMBOL(shdma_chan_probe);
 
diff --git a/drivers/dma/sh/shdma-of.c b/drivers/dma/sh/shdma-of.c
index b4ff9d3e56d1..f999f9b0d314 100644
--- a/drivers/dma/sh/shdma-of.c
+++ b/drivers/dma/sh/shdma-of.c
@@ -66,7 +66,6 @@ MODULE_DEVICE_TABLE(of, sh_dmae_of_match);
 
 static struct platform_driver shdma_of = {
 	.driver		= {
-		.owner	= THIS_MODULE,
 		.name	= "shdma-of",
 		.of_match_table = shdma_of_match,
 	},
diff --git a/drivers/dma/sh/shdmac.c b/drivers/dma/sh/shdmac.c
index 58eb85770eba..aec8a84784a4 100644
--- a/drivers/dma/sh/shdmac.c
+++ b/drivers/dma/sh/shdmac.c
@@ -572,7 +572,6 @@ err_no_irq:
 
 static void sh_dmae_chan_remove(struct sh_dmae_device *shdev)
 {
-	struct dma_device *dma_dev = &shdev->shdma_dev.dma_dev;
 	struct shdma_chan *schan;
 	int i;
 
@@ -581,7 +580,6 @@ static void sh_dmae_chan_remove(struct sh_dmae_device *shdev)
 
 		shdma_chan_remove(schan);
 	}
-	dma_dev->chancnt = 0;
 }
 
 static void sh_dmae_shutdown(struct platform_device *pdev)
@@ -925,7 +923,6 @@ static int sh_dmae_remove(struct platform_device *pdev)
 
 static struct platform_driver sh_dmae_driver = {
 	.driver 	= {
-		.owner	= THIS_MODULE,
 		.pm	= &sh_dmae_pm,
 		.name	= SH_DMAE_DRV_NAME,
 		.of_match_table = sh_dmae_of_match,
diff --git a/drivers/dma/sh/sudmac.c b/drivers/dma/sh/sudmac.c
index 3ce103909896..6da2eaa6c294 100644
--- a/drivers/dma/sh/sudmac.c
+++ b/drivers/dma/sh/sudmac.c
@@ -295,7 +295,6 @@ err_no_irq:
 
 static void sudmac_chan_remove(struct sudmac_device *su_dev)
 {
-	struct dma_device *dma_dev = &su_dev->shdma_dev.dma_dev;
 	struct shdma_chan *schan;
 	int i;
 
@@ -304,7 +303,6 @@ static void sudmac_chan_remove(struct sudmac_device *su_dev)
 
 		shdma_chan_remove(schan);
 	}
-	dma_dev->chancnt = 0;
 }
 
 static dma_addr_t sudmac_slave_addr(struct shdma_chan *schan)
@@ -411,7 +409,6 @@ static int sudmac_remove(struct platform_device *pdev)
 
 static struct platform_driver sudmac_driver = {
 	.driver		= {
-		.owner	= THIS_MODULE,
 		.name	= SUDMAC_DRV_NAME,
 	},
 	.probe		= sudmac_probe,
diff --git a/drivers/dma/sirf-dma.c b/drivers/dma/sirf-dma.c
index aac03ab10c54..3492a5f91d31 100644
--- a/drivers/dma/sirf-dma.c
+++ b/drivers/dma/sirf-dma.c
@@ -735,7 +735,6 @@ static int sirfsoc_dma_probe(struct platform_device *op)
 
 	dma = &sdma->dma;
 	dma->dev = dev;
-	dma->chancnt = SIRFSOC_DMA_CHANNELS;
 
 	dma->device_alloc_chan_resources = sirfsoc_dma_alloc_chan_resources;
 	dma->device_free_chan_resources = sirfsoc_dma_free_chan_resources;
@@ -752,7 +751,7 @@ static int sirfsoc_dma_probe(struct platform_device *op)
 	dma_cap_set(DMA_INTERLEAVE, dma->cap_mask);
 	dma_cap_set(DMA_PRIVATE, dma->cap_mask);
 
-	for (i = 0; i < dma->chancnt; i++) {
+	for (i = 0; i < SIRFSOC_DMA_CHANNELS; i++) {
 		schan = &sdma->channels[i];
 
 		schan->chan.device = dma;
@@ -835,6 +834,7 @@ static int sirfsoc_dma_runtime_resume(struct device *dev)
 	return 0;
 }
 
+#ifdef CONFIG_PM_SLEEP
 static int sirfsoc_dma_pm_suspend(struct device *dev)
 {
 	struct sirfsoc_dma *sdma = dev_get_drvdata(dev);
@@ -916,6 +916,7 @@ static int sirfsoc_dma_pm_resume(struct device *dev)
 
 	return 0;
 }
+#endif
 
 static const struct dev_pm_ops sirfsoc_dma_pm_ops = {
 	SET_RUNTIME_PM_OPS(sirfsoc_dma_runtime_suspend, sirfsoc_dma_runtime_resume, NULL)
@@ -933,7 +934,6 @@ static struct platform_driver sirfsoc_dma_driver = {
 	.remove		= sirfsoc_dma_remove,
 	.driver = {
 		.name = DRV_NAME,
-		.owner = THIS_MODULE,
 		.pm = &sirfsoc_dma_pm_ops,
 		.of_match_table	= sirfsoc_dma_match,
 	},
diff --git a/drivers/dma/ste_dma40.c b/drivers/dma/ste_dma40.c
index 5fe59335e247..15d49461c0d2 100644
--- a/drivers/dma/ste_dma40.c
+++ b/drivers/dma/ste_dma40.c
@@ -3051,7 +3051,7 @@ static int dma40_runtime_resume(struct device *dev)
 
 static const struct dev_pm_ops dma40_pm_ops = {
 	SET_LATE_SYSTEM_SLEEP_PM_OPS(dma40_suspend, dma40_resume)
-	SET_PM_RUNTIME_PM_OPS(dma40_runtime_suspend,
+	SET_RUNTIME_PM_OPS(dma40_runtime_suspend,
 				dma40_runtime_resume,
 				NULL)
 };
@@ -3432,6 +3432,7 @@ static int __init d40_lcla_allocate(struct d40_base *base)
 
 			d40_err(base->dev, "Failed to allocate %d pages.\n",
 				base->lcla_pool.pages);
+			ret = -ENOMEM;
 
 			for (j = 0; j < i; j++)
 				free_pages(page_list[j], base->lcla_pool.pages);
@@ -3749,7 +3750,6 @@ static const struct of_device_id d40_match[] = {
 
 static struct platform_driver d40_driver = {
 	.driver = {
-		.owner = THIS_MODULE,
 		.name  = D40_NAME,
 		.pm = &dma40_pm_ops,
 		.of_match_table = d40_match,
diff --git a/drivers/dma/sun6i-dma.c b/drivers/dma/sun6i-dma.c
index 1f92a56fd2b6..159f1736a16f 100644
--- a/drivers/dma/sun6i-dma.c
+++ b/drivers/dma/sun6i-dma.c
@@ -18,6 +18,7 @@
 #include <linux/interrupt.h>
 #include <linux/module.h>
 #include <linux/of_dma.h>
+#include <linux/of_device.h>
 #include <linux/platform_device.h>
 #include <linux/reset.h>
 #include <linux/slab.h>
@@ -26,24 +27,6 @@
 #include "virt-dma.h"
 
 /*
- * There's 16 physical channels that can work in parallel.
- *
- * However we have 30 different endpoints for our requests.
- *
- * Since the channels are able to handle only an unidirectional
- * transfer, we need to allocate more virtual channels so that
- * everyone can grab one channel.
- *
- * Some devices can't work in both direction (mostly because it
- * wouldn't make sense), so we have a bit fewer virtual channels than
- * 2 channels per endpoints.
- */
-
-#define NR_MAX_CHANNELS		16
-#define NR_MAX_REQUESTS		30
-#define NR_MAX_VCHANS		53
-
-/*
  * Common registers
  */
 #define DMA_IRQ_EN(x)		((x) * 0x04)
@@ -60,6 +43,12 @@
 #define DMA_STAT		0x30
 
 /*
+ * sun8i specific registers
+ */
+#define SUN8I_DMA_GATE		0x20
+#define SUN8I_DMA_GATE_ENABLE	0x4
+
+/*
  * Channels specific registers
  */
 #define DMA_CHAN_ENABLE		0x00
@@ -102,6 +91,19 @@
 #define DRQ_SDRAM	1
 
 /*
+ * Hardware channels / ports representation
+ *
+ * The hardware is used in several SoCs, with differing numbers
+ * of channels and endpoints. This structure ties those numbers
+ * to a certain compatible string.
+ */
+struct sun6i_dma_config {
+	u32 nr_max_channels;
+	u32 nr_max_requests;
+	u32 nr_max_vchans;
+};
+
+/*
  * Hardware representation of the LLI
  *
  * The hardware will be fed the physical address of this structure,
@@ -159,6 +161,7 @@ struct sun6i_dma_dev {
 	struct dma_pool		*pool;
 	struct sun6i_pchan	*pchans;
 	struct sun6i_vchan	*vchans;
+	const struct sun6i_dma_config *cfg;
 };
 
 static struct device *chan2dev(struct dma_chan *chan)
@@ -230,30 +233,25 @@ static inline void sun6i_dma_dump_chan_regs(struct sun6i_dma_dev *sdev,
 		readl(pchan->base + DMA_CHAN_CUR_PARA));
 }
 
-static inline int convert_burst(u32 maxburst, u8 *burst)
+static inline s8 convert_burst(u32 maxburst)
 {
 	switch (maxburst) {
 	case 1:
-		*burst = 0;
-		break;
+		return 0;
 	case 8:
-		*burst = 2;
-		break;
+		return 2;
 	default:
 		return -EINVAL;
 	}
-
-	return 0;
 }
 
-static inline int convert_buswidth(enum dma_slave_buswidth addr_width, u8 *width)
+static inline s8 convert_buswidth(enum dma_slave_buswidth addr_width)
 {
 	if ((addr_width < DMA_SLAVE_BUSWIDTH_1_BYTE) ||
 	    (addr_width > DMA_SLAVE_BUSWIDTH_4_BYTES))
 		return -EINVAL;
 
-	*width = addr_width >> 1;
-	return 0;
+	return addr_width >> 1;
 }
 
 static void *sun6i_dma_lli_add(struct sun6i_dma_lli *prev,
@@ -284,26 +282,25 @@ static inline int sun6i_dma_cfg_lli(struct sun6i_dma_lli *lli,
 				    struct dma_slave_config *config)
 {
 	u8 src_width, dst_width, src_burst, dst_burst;
-	int ret;
 
 	if (!config)
 		return -EINVAL;
 
-	ret = convert_burst(config->src_maxburst, &src_burst);
-	if (ret)
-		return ret;
+	src_burst = convert_burst(config->src_maxburst);
+	if (src_burst)
+		return src_burst;
 
-	ret = convert_burst(config->dst_maxburst, &dst_burst);
-	if (ret)
-		return ret;
+	dst_burst = convert_burst(config->dst_maxburst);
+	if (dst_burst)
+		return dst_burst;
 
-	ret = convert_buswidth(config->src_addr_width, &src_width);
-	if (ret)
-		return ret;
+	src_width = convert_buswidth(config->src_addr_width);
+	if (src_width)
+		return src_width;
 
-	ret = convert_buswidth(config->dst_addr_width, &dst_width);
-	if (ret)
-		return ret;
+	dst_width = convert_buswidth(config->dst_addr_width);
+	if (dst_width)
+		return dst_width;
 
 	lli->cfg = DMA_CHAN_CFG_SRC_BURST(src_burst) |
 		DMA_CHAN_CFG_SRC_WIDTH(src_width) |
@@ -432,6 +429,7 @@ static int sun6i_dma_start_desc(struct sun6i_vchan *vchan)
 static void sun6i_dma_tasklet(unsigned long data)
 {
 	struct sun6i_dma_dev *sdev = (struct sun6i_dma_dev *)data;
+	const struct sun6i_dma_config *cfg = sdev->cfg;
 	struct sun6i_vchan *vchan;
 	struct sun6i_pchan *pchan;
 	unsigned int pchan_alloc = 0;
@@ -459,7 +457,7 @@ static void sun6i_dma_tasklet(unsigned long data)
 	}
 
 	spin_lock_irq(&sdev->lock);
-	for (pchan_idx = 0; pchan_idx < NR_MAX_CHANNELS; pchan_idx++) {
+	for (pchan_idx = 0; pchan_idx < cfg->nr_max_channels; pchan_idx++) {
 		pchan = &sdev->pchans[pchan_idx];
 
 		if (pchan->vchan || list_empty(&sdev->pending))
@@ -480,7 +478,7 @@ static void sun6i_dma_tasklet(unsigned long data)
 	}
 	spin_unlock_irq(&sdev->lock);
 
-	for (pchan_idx = 0; pchan_idx < NR_MAX_CHANNELS; pchan_idx++) {
+	for (pchan_idx = 0; pchan_idx < cfg->nr_max_channels; pchan_idx++) {
 		if (!(pchan_alloc & BIT(pchan_idx)))
 			continue;
 
@@ -502,7 +500,7 @@ static irqreturn_t sun6i_dma_interrupt(int irq, void *dev_id)
 	int i, j, ret = IRQ_NONE;
 	u32 status;
 
-	for (i = 0; i < 2; i++) {
+	for (i = 0; i < sdev->cfg->nr_max_channels / DMA_IRQ_CHAN_NR; i++) {
 		status = readl(sdev->base + DMA_IRQ_STAT(i));
 		if (!status)
 			continue;
@@ -512,7 +510,7 @@ static irqreturn_t sun6i_dma_interrupt(int irq, void *dev_id)
 
 		writel(status, sdev->base + DMA_IRQ_STAT(i));
 
-		for (j = 0; (j < 8) && status; j++) {
+		for (j = 0; (j < DMA_IRQ_CHAN_NR) && status; j++) {
 			if (status & DMA_IRQ_QUEUE) {
 				pchan = sdev->pchans + j;
 				vchan = pchan->vchan;
@@ -525,7 +523,7 @@ static irqreturn_t sun6i_dma_interrupt(int irq, void *dev_id)
 				}
 			}
 
-			status = status >> 4;
+			status = status >> DMA_IRQ_CHAN_WIDTH;
 		}
 
 		if (!atomic_read(&sdev->tasklet_shutdown))
@@ -542,11 +540,10 @@ static struct dma_async_tx_descriptor *sun6i_dma_prep_dma_memcpy(
 {
 	struct sun6i_dma_dev *sdev = to_sun6i_dma_dev(chan->device);
 	struct sun6i_vchan *vchan = to_sun6i_vchan(chan);
-	struct dma_slave_config *sconfig = &vchan->cfg;
 	struct sun6i_dma_lli *v_lli;
 	struct sun6i_desc *txd;
 	dma_addr_t p_lli;
-	int ret;
+	s8 burst, width;
 
 	dev_dbg(chan2dev(chan),
 		"%s; chan: %d, dest: %pad, src: %pad, len: %zu. flags: 0x%08lx\n",
@@ -565,14 +562,21 @@ static struct dma_async_tx_descriptor *sun6i_dma_prep_dma_memcpy(
 		goto err_txd_free;
 	}
 
-	ret = sun6i_dma_cfg_lli(v_lli, src, dest, len, sconfig);
-	if (ret)
-		goto err_dma_free;
+	v_lli->src = src;
+	v_lli->dst = dest;
+	v_lli->len = len;
+	v_lli->para = NORMAL_WAIT;
 
+	burst = convert_burst(8);
+	width = convert_buswidth(DMA_SLAVE_BUSWIDTH_4_BYTES);
 	v_lli->cfg |= DMA_CHAN_CFG_SRC_DRQ(DRQ_SDRAM) |
 		DMA_CHAN_CFG_DST_DRQ(DRQ_SDRAM) |
 		DMA_CHAN_CFG_DST_LINEAR_MODE |
-		DMA_CHAN_CFG_SRC_LINEAR_MODE;
+		DMA_CHAN_CFG_SRC_LINEAR_MODE |
+		DMA_CHAN_CFG_SRC_BURST(burst) |
+		DMA_CHAN_CFG_SRC_WIDTH(width) |
+		DMA_CHAN_CFG_DST_BURST(burst) |
+		DMA_CHAN_CFG_DST_WIDTH(width);
 
 	sun6i_dma_lli_add(NULL, v_lli, p_lli, txd);
 
@@ -580,8 +584,6 @@ static struct dma_async_tx_descriptor *sun6i_dma_prep_dma_memcpy(
 
 	return vchan_tx_prep(&vchan->vc, &txd->vd, flags);
 
-err_dma_free:
-	dma_pool_free(sdev->pool, v_lli, p_lli);
 err_txd_free:
 	kfree(txd);
 	return NULL;
@@ -817,7 +819,7 @@ static struct dma_chan *sun6i_dma_of_xlate(struct of_phandle_args *dma_spec,
 	struct dma_chan *chan;
 	u8 port = dma_spec->args[0];
 
-	if (port > NR_MAX_REQUESTS)
+	if (port > sdev->cfg->nr_max_requests)
 		return NULL;
 
 	chan = dma_get_any_slave_channel(&sdev->slave);
@@ -850,7 +852,7 @@ static inline void sun6i_dma_free(struct sun6i_dma_dev *sdev)
 {
 	int i;
 
-	for (i = 0; i < NR_MAX_VCHANS; i++) {
+	for (i = 0; i < sdev->cfg->nr_max_vchans; i++) {
 		struct sun6i_vchan *vchan = &sdev->vchans[i];
 
 		list_del(&vchan->vc.chan.device_node);
@@ -858,17 +860,61 @@ static inline void sun6i_dma_free(struct sun6i_dma_dev *sdev)
 	}
 }
 
+/*
+ * For A31:
+ *
+ * There's 16 physical channels that can work in parallel.
+ *
+ * However we have 30 different endpoints for our requests.
+ *
+ * Since the channels are able to handle only an unidirectional
+ * transfer, we need to allocate more virtual channels so that
+ * everyone can grab one channel.
+ *
+ * Some devices can't work in both direction (mostly because it
+ * wouldn't make sense), so we have a bit fewer virtual channels than
+ * 2 channels per endpoints.
+ */
+
+static struct sun6i_dma_config sun6i_a31_dma_cfg = {
+	.nr_max_channels = 16,
+	.nr_max_requests = 30,
+	.nr_max_vchans   = 53,
+};
+
+/*
+ * The A23 only has 8 physical channels, a maximum DRQ port id of 24,
+ * and a total of 37 usable source and destination endpoints.
+ */
+
+static struct sun6i_dma_config sun8i_a23_dma_cfg = {
+	.nr_max_channels = 8,
+	.nr_max_requests = 24,
+	.nr_max_vchans   = 37,
+};
+
+static struct of_device_id sun6i_dma_match[] = {
+	{ .compatible = "allwinner,sun6i-a31-dma", .data = &sun6i_a31_dma_cfg },
+	{ .compatible = "allwinner,sun8i-a23-dma", .data = &sun8i_a23_dma_cfg },
+	{ /* sentinel */ }
+};
+
 static int sun6i_dma_probe(struct platform_device *pdev)
 {
+	const struct of_device_id *device;
 	struct sun6i_dma_dev *sdc;
 	struct resource *res;
-	struct clk *mux, *pll6;
 	int ret, i;
 
 	sdc = devm_kzalloc(&pdev->dev, sizeof(*sdc), GFP_KERNEL);
 	if (!sdc)
 		return -ENOMEM;
 
+	device = of_match_device(sun6i_dma_match, &pdev->dev);
+	if (!device)
+		return -ENODEV;
+	sdc->cfg = device->data;
+
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	sdc->base = devm_ioremap_resource(&pdev->dev, res);
 	if (IS_ERR(sdc->base))
@@ -886,28 +932,6 @@ static int sun6i_dma_probe(struct platform_device *pdev)
 		return PTR_ERR(sdc->clk);
 	}
 
-	mux = clk_get(NULL, "ahb1_mux");
-	if (IS_ERR(mux)) {
-		dev_err(&pdev->dev, "Couldn't get AHB1 Mux\n");
-		return PTR_ERR(mux);
-	}
-
-	pll6 = clk_get(NULL, "pll6");
-	if (IS_ERR(pll6)) {
-		dev_err(&pdev->dev, "Couldn't get PLL6\n");
-		clk_put(mux);
-		return PTR_ERR(pll6);
-	}
-
-	ret = clk_set_parent(mux, pll6);
-	clk_put(pll6);
-	clk_put(mux);
-
-	if (ret) {
-		dev_err(&pdev->dev, "Couldn't reparent AHB1 on PLL6\n");
-		return ret;
-	}
-
 	sdc->rstc = devm_reset_control_get(&pdev->dev, NULL);
 	if (IS_ERR(sdc->rstc)) {
 		dev_err(&pdev->dev, "No reset controller specified\n");
@@ -937,30 +961,30 @@ static int sun6i_dma_probe(struct platform_device *pdev)
 	sdc->slave.device_prep_slave_sg		= sun6i_dma_prep_slave_sg;
 	sdc->slave.device_prep_dma_memcpy	= sun6i_dma_prep_dma_memcpy;
 	sdc->slave.device_control		= sun6i_dma_control;
-	sdc->slave.chancnt			= NR_MAX_VCHANS;
+	sdc->slave.copy_align			= 4;
 
 	sdc->slave.dev = &pdev->dev;
 
-	sdc->pchans = devm_kcalloc(&pdev->dev, NR_MAX_CHANNELS,
+	sdc->pchans = devm_kcalloc(&pdev->dev, sdc->cfg->nr_max_channels,
 				   sizeof(struct sun6i_pchan), GFP_KERNEL);
 	if (!sdc->pchans)
 		return -ENOMEM;
 
-	sdc->vchans = devm_kcalloc(&pdev->dev, NR_MAX_VCHANS,
+	sdc->vchans = devm_kcalloc(&pdev->dev, sdc->cfg->nr_max_vchans,
 				   sizeof(struct sun6i_vchan), GFP_KERNEL);
 	if (!sdc->vchans)
 		return -ENOMEM;
 
 	tasklet_init(&sdc->task, sun6i_dma_tasklet, (unsigned long)sdc);
 
-	for (i = 0; i < NR_MAX_CHANNELS; i++) {
+	for (i = 0; i < sdc->cfg->nr_max_channels; i++) {
 		struct sun6i_pchan *pchan = &sdc->pchans[i];
 
 		pchan->idx = i;
 		pchan->base = sdc->base + 0x100 + i * 0x40;
 	}
 
-	for (i = 0; i < NR_MAX_VCHANS; i++) {
+	for (i = 0; i < sdc->cfg->nr_max_vchans; i++) {
 		struct sun6i_vchan *vchan = &sdc->vchans[i];
 
 		INIT_LIST_HEAD(&vchan->node);
@@ -1000,6 +1024,15 @@ static int sun6i_dma_probe(struct platform_device *pdev)
 		goto err_dma_unregister;
 	}
 
+	/*
+	 * sun8i variant requires us to toggle a dma gating register,
+	 * as seen in Allwinner's SDK. This register is not documented
+	 * in the A23 user manual.
+	 */
+	if (of_device_is_compatible(pdev->dev.of_node,
+				    "allwinner,sun8i-a23-dma"))
+		writel(SUN8I_DMA_GATE_ENABLE, sdc->base + SUN8I_DMA_GATE);
+
 	return 0;
 
 err_dma_unregister:
@@ -1032,11 +1065,6 @@ static int sun6i_dma_remove(struct platform_device *pdev)
 	return 0;
 }
 
-static struct of_device_id sun6i_dma_match[] = {
-	{ .compatible = "allwinner,sun6i-a31-dma" },
-	{ /* sentinel */ }
-};
-
 static struct platform_driver sun6i_dma_driver = {
 	.probe		= sun6i_dma_probe,
 	.remove		= sun6i_dma_remove,
diff --git a/drivers/dma/tegra20-apb-dma.c b/drivers/dma/tegra20-apb-dma.c
index 16efa603ff65..d8450c3f35f0 100644
--- a/drivers/dma/tegra20-apb-dma.c
+++ b/drivers/dma/tegra20-apb-dma.c
@@ -1587,7 +1587,7 @@ static int tegra_dma_pm_resume(struct device *dev)
 #endif
 
 static const struct dev_pm_ops tegra_dma_dev_pm_ops = {
-#ifdef CONFIG_PM_RUNTIME
+#ifdef CONFIG_PM
 	.runtime_suspend = tegra_dma_runtime_suspend,
 	.runtime_resume = tegra_dma_runtime_resume,
 #endif
@@ -1597,7 +1597,6 @@ static const struct dev_pm_ops tegra_dma_dev_pm_ops = {
 static struct platform_driver tegra_dmac_driver = {
 	.driver = {
 		.name	= "tegra-apbdma",
-		.owner = THIS_MODULE,
 		.pm	= &tegra_dma_dev_pm_ops,
 		.of_match_table = tegra_dma_of_match,
 	},
diff --git a/drivers/dma/timb_dma.c b/drivers/dma/timb_dma.c
index 4506a7b4f972..2407ccf1a64b 100644
--- a/drivers/dma/timb_dma.c
+++ b/drivers/dma/timb_dma.c
@@ -783,7 +783,6 @@ static int td_remove(struct platform_device *pdev)
 static struct platform_driver td_driver = {
 	.driver = {
 		.name	= DRIVER_NAME,
-		.owner  = THIS_MODULE,
 	},
 	.probe	= td_probe,
 	.remove	= td_remove,
diff --git a/drivers/dma/txx9dmac.c b/drivers/dma/txx9dmac.c
index 17686caf64d5..0659ec9c4488 100644
--- a/drivers/dma/txx9dmac.c
+++ b/drivers/dma/txx9dmac.c
@@ -76,7 +76,7 @@ static void channel64_write_CHAR(const struct txx9dmac_chan *dc, dma_addr_t val)
 
 static void channel64_clear_CHAR(const struct txx9dmac_chan *dc)
 {
-#if defined(CONFIG_32BIT) && !defined(CONFIG_64BIT_PHYS_ADDR)
+#if defined(CONFIG_32BIT) && !defined(CONFIG_PHYS_ADDR_T_64BIT)
 	channel64_writel(dc, CHAR, 0);
 	channel64_writel(dc, __pad_CHAR, 0);
 #else
diff --git a/drivers/dma/txx9dmac.h b/drivers/dma/txx9dmac.h
index f5a760598882..f6517b928bab 100644
--- a/drivers/dma/txx9dmac.h
+++ b/drivers/dma/txx9dmac.h
@@ -67,7 +67,7 @@ static inline bool txx9_dma_have_SMPCHN(void)
 
 /* Hardware register definitions. */
 struct txx9dmac_cregs {
-#if defined(CONFIG_32BIT) && !defined(CONFIG_64BIT_PHYS_ADDR)
+#if defined(CONFIG_32BIT) && !defined(CONFIG_PHYS_ADDR_T_64BIT)
 	TXX9_DMA_REG32(CHAR);	/* Chain Address Register */
 #else
 	u64 CHAR;		/* Chain Address Register */
@@ -201,7 +201,7 @@ static inline bool is_dmac64(const struct txx9dmac_chan *dc)
 #ifdef TXX9_DMA_USE_SIMPLE_CHAIN
 /* Hardware descriptor definition. (for simple-chain) */
 struct txx9dmac_hwdesc {
-#if defined(CONFIG_32BIT) && !defined(CONFIG_64BIT_PHYS_ADDR)
+#if defined(CONFIG_32BIT) && !defined(CONFIG_PHYS_ADDR_T_64BIT)
 	TXX9_DMA_REG32(CHAR);
 #else
 	u64 CHAR;
diff --git a/drivers/dma/xilinx/xilinx_vdma.c b/drivers/dma/xilinx/xilinx_vdma.c
index 42a13e8d4607..4a3a8f3137b3 100644
--- a/drivers/dma/xilinx/xilinx_vdma.c
+++ b/drivers/dma/xilinx/xilinx_vdma.c
@@ -942,6 +942,9 @@ xilinx_vdma_dma_prep_interleaved(struct dma_chan *dchan,
 	if (!xt->numf || !xt->sgl[0].size)
 		return NULL;
 
+	if (xt->frame_size != 1)
+		return NULL;
+
 	/* Allocate a transaction descriptor. */
 	desc = xilinx_vdma_alloc_tx_descriptor(chan);
 	if (!desc)
@@ -960,7 +963,7 @@ xilinx_vdma_dma_prep_interleaved(struct dma_chan *dchan,
 	hw = &segment->hw;
 	hw->vsize = xt->numf;
 	hw->hsize = xt->sgl[0].size;
-	hw->stride = xt->sgl[0].icg <<
+	hw->stride = (xt->sgl[0].icg + xt->sgl[0].size) <<
 			XILINX_VDMA_FRMDLY_STRIDE_STRIDE_SHIFT;
 	hw->stride |= chan->config.frm_dly <<
 			XILINX_VDMA_FRMDLY_STRIDE_FRMDLY_SHIFT;
@@ -971,9 +974,11 @@ xilinx_vdma_dma_prep_interleaved(struct dma_chan *dchan,
 		hw->buf_addr = xt->src_start;
 
 	/* Link the previous next descriptor to current */
-	prev = list_last_entry(&desc->segments,
-				struct xilinx_vdma_tx_segment, node);
-	prev->hw.next_desc = segment->phys;
+	if (!list_empty(&desc->segments)) {
+		prev = list_last_entry(&desc->segments,
+				       struct xilinx_vdma_tx_segment, node);
+		prev->hw.next_desc = segment->phys;
+	}
 
 	/* Insert the segment into the descriptor segments list. */
 	list_add_tail(&segment->node, &desc->segments);
@@ -1365,7 +1370,6 @@ static const struct of_device_id xilinx_vdma_of_ids[] = {
 static struct platform_driver xilinx_vdma_driver = {
 	.driver = {
 		.name = "xilinx-vdma",
-		.owner = THIS_MODULE,
 		.of_match_table = xilinx_vdma_of_ids,
 	},
 	.probe = xilinx_vdma_probe,
diff --git a/drivers/edac/Kconfig b/drivers/edac/Kconfig
index fd89ca982748..49c265255a07 100644
--- a/drivers/edac/Kconfig
+++ b/drivers/edac/Kconfig
@@ -61,14 +61,14 @@ config EDAC_DECODE_MCE
 	  has been initialized.
 
 config EDAC_MCE_INJ
-	tristate "Simple MCE injection interface over /sysfs"
-	depends on EDAC_DECODE_MCE
+	tristate "Simple MCE injection interface"
+	depends on EDAC_DECODE_MCE && DEBUG_FS
 	default n
 	help
-	  This is a simple interface to inject MCEs over /sysfs and test
-	  the MCE decoding code in EDAC.
+	  This is a simple debugfs interface to inject MCEs and test different
+	  aspects of the MCE handling code.
 
-	  This is currently AMD-only.
+	  WARNING: Do not even assume this interface is staying stable!
 
 config EDAC_MM_EDAC
 	tristate "Main Memory EDAC (Error Detection And Correction) reporting"
@@ -105,11 +105,11 @@ config EDAC_GHES
 	  In doubt, say 'Y'.
 
 config EDAC_AMD64
-	tristate "AMD64 (Opteron, Athlon64) K8, F10h"
-	depends on EDAC_MM_EDAC && AMD_NB && X86_64 && EDAC_DECODE_MCE
+	tristate "AMD64 (Opteron, Athlon64)"
+	depends on EDAC_MM_EDAC && AMD_NB && EDAC_DECODE_MCE
 	help
 	  Support for error detection and correction of DRAM ECC errors on
-	  the AMD64 families of memory controllers (K8 and F10h)
+	  the AMD64 families (>= K8) of memory controllers.
 
 config EDAC_AMD64_ERROR_INJECTION
 	bool "Sysfs HW Error injection facilities"
@@ -376,4 +376,13 @@ config EDAC_OCTEON_PCI
 	  Support for error detection and correction on the
 	  Cavium Octeon family of SOCs.
 
+config EDAC_ALTERA_MC
+	tristate "Altera SDRAM Memory Controller EDAC"
+	depends on EDAC_MM_EDAC && ARCH_SOCFPGA
+	help
+	  Support for error detection and correction on the
+	  Altera SDRAM memory controller. Note that the
+	  preloader must initialize the SDRAM before loading
+	  the kernel.
+
 endif # EDAC
diff --git a/drivers/edac/Makefile b/drivers/edac/Makefile
index c479a24d8f77..d40c69a04df7 100644
--- a/drivers/edac/Makefile
+++ b/drivers/edac/Makefile
@@ -9,7 +9,7 @@
 obj-$(CONFIG_EDAC)			:= edac_stub.o
 obj-$(CONFIG_EDAC_MM_EDAC)		+= edac_core.o
 
-edac_core-y	:= edac_mc.o edac_device.o edac_mc_sysfs.o edac_pci_sysfs.o
+edac_core-y	:= edac_mc.o edac_device.o edac_mc_sysfs.o
 edac_core-y	+= edac_module.o edac_device_sysfs.o
 
 ifdef CONFIG_PCI
@@ -65,3 +65,5 @@ obj-$(CONFIG_EDAC_OCTEON_PC)		+= octeon_edac-pc.o
 obj-$(CONFIG_EDAC_OCTEON_L2C)		+= octeon_edac-l2c.o
 obj-$(CONFIG_EDAC_OCTEON_LMC)		+= octeon_edac-lmc.o
 obj-$(CONFIG_EDAC_OCTEON_PCI)		+= octeon_edac-pci.o
+
+obj-$(CONFIG_EDAC_ALTERA_MC)		+= altera_edac.o
diff --git a/drivers/edac/altera_edac.c b/drivers/edac/altera_edac.c
new file mode 100644
index 000000000000..3c4929fda9d5
--- /dev/null
+++ b/drivers/edac/altera_edac.c
@@ -0,0 +1,410 @@
+/*
+ *  Copyright Altera Corporation (C) 2014. All rights reserved.
+ *  Copyright 2011-2012 Calxeda, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ *
+ * Adapted from the highbank_mc_edac driver.
+ */
+
+#include <linux/ctype.h>
+#include <linux/edac.h>
+#include <linux/interrupt.h>
+#include <linux/kernel.h>
+#include <linux/mfd/syscon.h>
+#include <linux/of_platform.h>
+#include <linux/platform_device.h>
+#include <linux/regmap.h>
+#include <linux/types.h>
+#include <linux/uaccess.h>
+
+#include "edac_core.h"
+#include "edac_module.h"
+
+#define EDAC_MOD_STR		"altera_edac"
+#define EDAC_VERSION		"1"
+
+/* SDRAM Controller CtrlCfg Register */
+#define CTLCFG_OFST             0x00
+
+/* SDRAM Controller CtrlCfg Register Bit Masks */
+#define CTLCFG_ECC_EN           0x400
+#define CTLCFG_ECC_CORR_EN      0x800
+#define CTLCFG_GEN_SB_ERR       0x2000
+#define CTLCFG_GEN_DB_ERR       0x4000
+
+#define CTLCFG_ECC_AUTO_EN	(CTLCFG_ECC_EN | \
+				 CTLCFG_ECC_CORR_EN)
+
+/* SDRAM Controller Address Width Register */
+#define DRAMADDRW_OFST          0x2C
+
+/* SDRAM Controller Address Widths Field Register */
+#define DRAMADDRW_COLBIT_MASK   0x001F
+#define DRAMADDRW_COLBIT_SHIFT  0
+#define DRAMADDRW_ROWBIT_MASK   0x03E0
+#define DRAMADDRW_ROWBIT_SHIFT  5
+#define DRAMADDRW_BANKBIT_MASK	0x1C00
+#define DRAMADDRW_BANKBIT_SHIFT 10
+#define DRAMADDRW_CSBIT_MASK	0xE000
+#define DRAMADDRW_CSBIT_SHIFT   13
+
+/* SDRAM Controller Interface Data Width Register */
+#define DRAMIFWIDTH_OFST        0x30
+
+/* SDRAM Controller Interface Data Width Defines */
+#define DRAMIFWIDTH_16B_ECC     24
+#define DRAMIFWIDTH_32B_ECC     40
+
+/* SDRAM Controller DRAM Status Register */
+#define DRAMSTS_OFST            0x38
+
+/* SDRAM Controller DRAM Status Register Bit Masks */
+#define DRAMSTS_SBEERR          0x04
+#define DRAMSTS_DBEERR          0x08
+#define DRAMSTS_CORR_DROP       0x10
+
+/* SDRAM Controller DRAM IRQ Register */
+#define DRAMINTR_OFST           0x3C
+
+/* SDRAM Controller DRAM IRQ Register Bit Masks */
+#define DRAMINTR_INTREN         0x01
+#define DRAMINTR_SBEMASK        0x02
+#define DRAMINTR_DBEMASK        0x04
+#define DRAMINTR_CORRDROPMASK   0x08
+#define DRAMINTR_INTRCLR        0x10
+
+/* SDRAM Controller Single Bit Error Count Register */
+#define SBECOUNT_OFST           0x40
+
+/* SDRAM Controller Single Bit Error Count Register Bit Masks */
+#define SBECOUNT_MASK           0x0F
+
+/* SDRAM Controller Double Bit Error Count Register */
+#define DBECOUNT_OFST           0x44
+
+/* SDRAM Controller Double Bit Error Count Register Bit Masks */
+#define DBECOUNT_MASK           0x0F
+
+/* SDRAM Controller ECC Error Address Register */
+#define ERRADDR_OFST            0x48
+
+/* SDRAM Controller ECC Error Address Register Bit Masks */
+#define ERRADDR_MASK            0xFFFFFFFF
+
+/* Altera SDRAM Memory Controller data */
+struct altr_sdram_mc_data {
+	struct regmap *mc_vbase;
+};
+
+static irqreturn_t altr_sdram_mc_err_handler(int irq, void *dev_id)
+{
+	struct mem_ctl_info *mci = dev_id;
+	struct altr_sdram_mc_data *drvdata = mci->pvt_info;
+	u32 status, err_count, err_addr;
+
+	/* Error Address is shared by both SBE & DBE */
+	regmap_read(drvdata->mc_vbase, ERRADDR_OFST, &err_addr);
+
+	regmap_read(drvdata->mc_vbase, DRAMSTS_OFST, &status);
+
+	if (status & DRAMSTS_DBEERR) {
+		regmap_read(drvdata->mc_vbase, DBECOUNT_OFST, &err_count);
+		panic("\nEDAC: [%d Uncorrectable errors @ 0x%08X]\n",
+		      err_count, err_addr);
+	}
+	if (status & DRAMSTS_SBEERR) {
+		regmap_read(drvdata->mc_vbase, SBECOUNT_OFST, &err_count);
+		edac_mc_handle_error(HW_EVENT_ERR_CORRECTED, mci, err_count,
+				     err_addr >> PAGE_SHIFT,
+				     err_addr & ~PAGE_MASK, 0,
+				     0, 0, -1, mci->ctl_name, "");
+	}
+
+	regmap_write(drvdata->mc_vbase,	DRAMINTR_OFST,
+		     (DRAMINTR_INTRCLR | DRAMINTR_INTREN));
+
+	return IRQ_HANDLED;
+}
+
+#ifdef CONFIG_EDAC_DEBUG
+static ssize_t altr_sdr_mc_err_inject_write(struct file *file,
+					    const char __user *data,
+					    size_t count, loff_t *ppos)
+{
+	struct mem_ctl_info *mci = file->private_data;
+	struct altr_sdram_mc_data *drvdata = mci->pvt_info;
+	u32 *ptemp;
+	dma_addr_t dma_handle;
+	u32 reg, read_reg;
+
+	ptemp = dma_alloc_coherent(mci->pdev, 16, &dma_handle, GFP_KERNEL);
+	if (!ptemp) {
+		dma_free_coherent(mci->pdev, 16, ptemp, dma_handle);
+		edac_printk(KERN_ERR, EDAC_MC,
+			    "Inject: Buffer Allocation error\n");
+		return -ENOMEM;
+	}
+
+	regmap_read(drvdata->mc_vbase, CTLCFG_OFST, &read_reg);
+	read_reg &= ~(CTLCFG_GEN_SB_ERR | CTLCFG_GEN_DB_ERR);
+
+	/* Error are injected by writing a word while the SBE or DBE
+	 * bit in the CTLCFG register is set. Reading the word will
+	 * trigger the SBE or DBE error and the corresponding IRQ.
+	 */
+	if (count == 3) {
+		edac_printk(KERN_ALERT, EDAC_MC,
+			    "Inject Double bit error\n");
+		regmap_write(drvdata->mc_vbase, CTLCFG_OFST,
+			     (read_reg | CTLCFG_GEN_DB_ERR));
+	} else {
+		edac_printk(KERN_ALERT, EDAC_MC,
+			    "Inject Single bit error\n");
+		regmap_write(drvdata->mc_vbase,	CTLCFG_OFST,
+			     (read_reg | CTLCFG_GEN_SB_ERR));
+	}
+
+	ptemp[0] = 0x5A5A5A5A;
+	ptemp[1] = 0xA5A5A5A5;
+
+	/* Clear the error injection bits */
+	regmap_write(drvdata->mc_vbase,	CTLCFG_OFST, read_reg);
+	/* Ensure it has been written out */
+	wmb();
+
+	/*
+	 * To trigger the error, we need to read the data back
+	 * (the data was written with errors above).
+	 * The ACCESS_ONCE macros and printk are used to prevent the
+	 * the compiler optimizing these reads out.
+	 */
+	reg = ACCESS_ONCE(ptemp[0]);
+	read_reg = ACCESS_ONCE(ptemp[1]);
+	/* Force Read */
+	rmb();
+
+	edac_printk(KERN_ALERT, EDAC_MC, "Read Data [0x%X, 0x%X]\n",
+		    reg, read_reg);
+
+	dma_free_coherent(mci->pdev, 16, ptemp, dma_handle);
+
+	return count;
+}
+
+static const struct file_operations altr_sdr_mc_debug_inject_fops = {
+	.open = simple_open,
+	.write = altr_sdr_mc_err_inject_write,
+	.llseek = generic_file_llseek,
+};
+
+static void altr_sdr_mc_create_debugfs_nodes(struct mem_ctl_info *mci)
+{
+	if (mci->debugfs)
+		debugfs_create_file("inject_ctrl", S_IWUSR, mci->debugfs, mci,
+				    &altr_sdr_mc_debug_inject_fops);
+}
+#else
+static void altr_sdr_mc_create_debugfs_nodes(struct mem_ctl_info *mci)
+{}
+#endif
+
+/* Get total memory size in bytes */
+static u32 altr_sdram_get_total_mem_size(struct regmap *mc_vbase)
+{
+	u32 size, read_reg, row, bank, col, cs, width;
+
+	if (regmap_read(mc_vbase, DRAMADDRW_OFST, &read_reg) < 0)
+		return 0;
+
+	if (regmap_read(mc_vbase, DRAMIFWIDTH_OFST, &width) < 0)
+		return 0;
+
+	col = (read_reg & DRAMADDRW_COLBIT_MASK) >>
+		DRAMADDRW_COLBIT_SHIFT;
+	row = (read_reg & DRAMADDRW_ROWBIT_MASK) >>
+		DRAMADDRW_ROWBIT_SHIFT;
+	bank = (read_reg & DRAMADDRW_BANKBIT_MASK) >>
+		DRAMADDRW_BANKBIT_SHIFT;
+	cs = (read_reg & DRAMADDRW_CSBIT_MASK) >>
+		DRAMADDRW_CSBIT_SHIFT;
+
+	/* Correct for ECC as its not addressible */
+	if (width == DRAMIFWIDTH_32B_ECC)
+		width = 32;
+	if (width == DRAMIFWIDTH_16B_ECC)
+		width = 16;
+
+	/* calculate the SDRAM size base on this info */
+	size = 1 << (row + bank + col);
+	size = size * cs * (width / 8);
+	return size;
+}
+
+static int altr_sdram_probe(struct platform_device *pdev)
+{
+	struct edac_mc_layer layers[2];
+	struct mem_ctl_info *mci;
+	struct altr_sdram_mc_data *drvdata;
+	struct regmap *mc_vbase;
+	struct dimm_info *dimm;
+	u32 read_reg, mem_size;
+	int irq;
+	int res = 0;
+
+	/* Validate the SDRAM controller has ECC enabled */
+	/* Grab the register range from the sdr controller in device tree */
+	mc_vbase = syscon_regmap_lookup_by_phandle(pdev->dev.of_node,
+						   "altr,sdr-syscon");
+	if (IS_ERR(mc_vbase)) {
+		edac_printk(KERN_ERR, EDAC_MC,
+			    "regmap for altr,sdr-syscon lookup failed.\n");
+		return -ENODEV;
+	}
+
+	if (regmap_read(mc_vbase, CTLCFG_OFST, &read_reg) ||
+	    ((read_reg & CTLCFG_ECC_AUTO_EN) !=	CTLCFG_ECC_AUTO_EN)) {
+		edac_printk(KERN_ERR, EDAC_MC,
+			    "No ECC/ECC disabled [0x%08X]\n", read_reg);
+		return -ENODEV;
+	}
+
+	/* Grab memory size from device tree. */
+	mem_size = altr_sdram_get_total_mem_size(mc_vbase);
+	if (!mem_size) {
+		edac_printk(KERN_ERR, EDAC_MC,
+			    "Unable to calculate memory size\n");
+		return -ENODEV;
+	}
+
+	/* Ensure the SDRAM Interrupt is disabled and cleared */
+	if (regmap_write(mc_vbase, DRAMINTR_OFST, DRAMINTR_INTRCLR)) {
+		edac_printk(KERN_ERR, EDAC_MC,
+			    "Error clearing SDRAM ECC IRQ\n");
+		return -ENODEV;
+	}
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		edac_printk(KERN_ERR, EDAC_MC,
+			    "No irq %d in DT\n", irq);
+		return -ENODEV;
+	}
+
+	layers[0].type = EDAC_MC_LAYER_CHIP_SELECT;
+	layers[0].size = 1;
+	layers[0].is_virt_csrow = true;
+	layers[1].type = EDAC_MC_LAYER_CHANNEL;
+	layers[1].size = 1;
+	layers[1].is_virt_csrow = false;
+	mci = edac_mc_alloc(0, ARRAY_SIZE(layers), layers,
+			    sizeof(struct altr_sdram_mc_data));
+	if (!mci)
+		return -ENOMEM;
+
+	mci->pdev = &pdev->dev;
+	drvdata = mci->pvt_info;
+	drvdata->mc_vbase = mc_vbase;
+	platform_set_drvdata(pdev, mci);
+
+	if (!devres_open_group(&pdev->dev, NULL, GFP_KERNEL)) {
+		res = -ENOMEM;
+		goto free;
+	}
+
+	mci->mtype_cap = MEM_FLAG_DDR3;
+	mci->edac_ctl_cap = EDAC_FLAG_NONE | EDAC_FLAG_SECDED;
+	mci->edac_cap = EDAC_FLAG_SECDED;
+	mci->mod_name = EDAC_MOD_STR;
+	mci->mod_ver = EDAC_VERSION;
+	mci->ctl_name = dev_name(&pdev->dev);
+	mci->scrub_mode = SCRUB_SW_SRC;
+	mci->dev_name = dev_name(&pdev->dev);
+
+	dimm = *mci->dimms;
+	dimm->nr_pages = ((mem_size - 1) >> PAGE_SHIFT) + 1;
+	dimm->grain = 8;
+	dimm->dtype = DEV_X8;
+	dimm->mtype = MEM_DDR3;
+	dimm->edac_mode = EDAC_SECDED;
+
+	res = edac_mc_add_mc(mci);
+	if (res < 0)
+		goto err;
+
+	res = devm_request_irq(&pdev->dev, irq, altr_sdram_mc_err_handler,
+			       0, dev_name(&pdev->dev), mci);
+	if (res < 0) {
+		edac_mc_printk(mci, KERN_ERR,
+			       "Unable to request irq %d\n", irq);
+		res = -ENODEV;
+		goto err2;
+	}
+
+	if (regmap_write(drvdata->mc_vbase, DRAMINTR_OFST,
+			 (DRAMINTR_INTRCLR | DRAMINTR_INTREN))) {
+		edac_mc_printk(mci, KERN_ERR,
+			       "Error enabling SDRAM ECC IRQ\n");
+		res = -ENODEV;
+		goto err2;
+	}
+
+	altr_sdr_mc_create_debugfs_nodes(mci);
+
+	devres_close_group(&pdev->dev, NULL);
+
+	return 0;
+
+err2:
+	edac_mc_del_mc(&pdev->dev);
+err:
+	devres_release_group(&pdev->dev, NULL);
+free:
+	edac_mc_free(mci);
+	edac_printk(KERN_ERR, EDAC_MC,
+		    "EDAC Probe Failed; Error %d\n", res);
+
+	return res;
+}
+
+static int altr_sdram_remove(struct platform_device *pdev)
+{
+	struct mem_ctl_info *mci = platform_get_drvdata(pdev);
+
+	edac_mc_del_mc(&pdev->dev);
+	edac_mc_free(mci);
+	platform_set_drvdata(pdev, NULL);
+
+	return 0;
+}
+
+static const struct of_device_id altr_sdram_ctrl_of_match[] = {
+	{ .compatible = "altr,sdram-edac", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, altr_sdram_ctrl_of_match);
+
+static struct platform_driver altr_sdram_edac_driver = {
+	.probe = altr_sdram_probe,
+	.remove = altr_sdram_remove,
+	.driver = {
+		.name = "altr_sdram_edac",
+		.of_match_table = altr_sdram_ctrl_of_match,
+	},
+};
+
+module_platform_driver(altr_sdram_edac_driver);
+
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR("Thor Thayer");
+MODULE_DESCRIPTION("EDAC Driver for Altera SDRAM Controller");
diff --git a/drivers/edac/amd64_edac.c b/drivers/edac/amd64_edac.c
index f8bf00010d45..17638d7cf5c2 100644
--- a/drivers/edac/amd64_edac.c
+++ b/drivers/edac/amd64_edac.c
@@ -87,61 +87,73 @@ int __amd64_write_pci_cfg_dword(struct pci_dev *pdev, int offset,
 }
 
 /*
+ * Select DCT to which PCI cfg accesses are routed
+ */
+static void f15h_select_dct(struct amd64_pvt *pvt, u8 dct)
+{
+	u32 reg = 0;
+
+	amd64_read_pci_cfg(pvt->F1, DCT_CFG_SEL, &reg);
+	reg &= (pvt->model == 0x30) ? ~3 : ~1;
+	reg |= dct;
+	amd64_write_pci_cfg(pvt->F1, DCT_CFG_SEL, reg);
+}
+
+/*
  *
  * Depending on the family, F2 DCT reads need special handling:
  *
- * K8: has a single DCT only
+ * K8: has a single DCT only and no address offsets >= 0x100
  *
  * F10h: each DCT has its own set of regs
  *	DCT0 -> F2x040..
  *	DCT1 -> F2x140..
  *
- * F15h: we select which DCT we access using F1x10C[DctCfgSel]
- *
  * F16h: has only 1 DCT
+ *
+ * F15h: we select which DCT we access using F1x10C[DctCfgSel]
  */
-static int k8_read_dct_pci_cfg(struct amd64_pvt *pvt, int addr, u32 *val,
-			       const char *func)
+static inline int amd64_read_dct_pci_cfg(struct amd64_pvt *pvt, u8 dct,
+					 int offset, u32 *val)
 {
-	if (addr >= 0x100)
-		return -EINVAL;
-
-	return __amd64_read_pci_cfg_dword(pvt->F2, addr, val, func);
-}
+	switch (pvt->fam) {
+	case 0xf:
+		if (dct || offset >= 0x100)
+			return -EINVAL;
+		break;
 
-static int f10_read_dct_pci_cfg(struct amd64_pvt *pvt, int addr, u32 *val,
-				 const char *func)
-{
-	return __amd64_read_pci_cfg_dword(pvt->F2, addr, val, func);
-}
+	case 0x10:
+		if (dct) {
+			/*
+			 * Note: If ganging is enabled, barring the regs
+			 * F2x[1,0]98 and F2x[1,0]9C; reads reads to F2x1xx
+			 * return 0. (cf. Section 2.8.1 F10h BKDG)
+			 */
+			if (dct_ganging_enabled(pvt))
+				return 0;
 
-/*
- * Select DCT to which PCI cfg accesses are routed
- */
-static void f15h_select_dct(struct amd64_pvt *pvt, u8 dct)
-{
-	u32 reg = 0;
+			offset += 0x100;
+		}
+		break;
 
-	amd64_read_pci_cfg(pvt->F1, DCT_CFG_SEL, &reg);
-	reg &= (pvt->model >= 0x30) ? ~3 : ~1;
-	reg |= dct;
-	amd64_write_pci_cfg(pvt->F1, DCT_CFG_SEL, reg);
-}
+	case 0x15:
+		/*
+		 * F15h: F2x1xx addresses do not map explicitly to DCT1.
+		 * We should select which DCT we access using F1x10C[DctCfgSel]
+		 */
+		dct = (dct && pvt->model == 0x30) ? 3 : dct;
+		f15h_select_dct(pvt, dct);
+		break;
 
-static int f15_read_dct_pci_cfg(struct amd64_pvt *pvt, int addr, u32 *val,
-				 const char *func)
-{
-	u8 dct  = 0;
+	case 0x16:
+		if (dct)
+			return -EINVAL;
+		break;
 
-	/* For F15 M30h, the second dct is DCT 3, refer to BKDG Section 2.10 */
-	if (addr >= 0x140 && addr <= 0x1a0) {
-		dct   = (pvt->model >= 0x30) ? 3 : 1;
-		addr -= 0x100;
+	default:
+		break;
 	}
-
-	f15h_select_dct(pvt, dct);
-
-	return __amd64_read_pci_cfg_dword(pvt->F2, addr, val, func);
+	return amd64_read_pci_cfg(pvt->F2, offset, val);
 }
 
 /*
@@ -680,9 +692,19 @@ static void debug_dump_dramcfg_low(struct amd64_pvt *pvt, u32 dclr, int chan)
 {
 	edac_dbg(1, "F2x%d90 (DRAM Cfg Low): 0x%08x\n", chan, dclr);
 
-	edac_dbg(1, "  DIMM type: %sbuffered; all DIMMs support ECC: %s\n",
-		 (dclr & BIT(16)) ?  "un" : "",
-		 (dclr & BIT(19)) ? "yes" : "no");
+	if (pvt->dram_type == MEM_LRDDR3) {
+		u32 dcsm = pvt->csels[chan].csmasks[0];
+		/*
+		 * It's assumed all LRDIMMs in a DCT are going to be of
+		 * same 'type' until proven otherwise. So, use a cs
+		 * value of '0' here to get dcsm value.
+		 */
+		edac_dbg(1, " LRDIMM %dx rank multiply\n", (dcsm & 0x3));
+	}
+
+	edac_dbg(1, "All DIMMs support ECC:%s\n",
+		    (dclr & BIT(19)) ? "yes" : "no");
+
 
 	edac_dbg(1, "  PAR/ERR parity: %s\n",
 		 (dclr & BIT(8)) ?  "enabled" : "disabled");
@@ -744,7 +766,7 @@ static void prep_chip_selects(struct amd64_pvt *pvt)
 	if (pvt->fam == 0xf && pvt->ext_model < K8_REV_F) {
 		pvt->csels[0].b_cnt = pvt->csels[1].b_cnt = 8;
 		pvt->csels[0].m_cnt = pvt->csels[1].m_cnt = 8;
-	} else if (pvt->fam == 0x15 && pvt->model >= 0x30) {
+	} else if (pvt->fam == 0x15 && pvt->model == 0x30) {
 		pvt->csels[0].b_cnt = pvt->csels[1].b_cnt = 4;
 		pvt->csels[0].m_cnt = pvt->csels[1].m_cnt = 2;
 	} else {
@@ -768,16 +790,17 @@ static void read_dct_base_mask(struct amd64_pvt *pvt)
 		u32 *base0 = &pvt->csels[0].csbases[cs];
 		u32 *base1 = &pvt->csels[1].csbases[cs];
 
-		if (!amd64_read_dct_pci_cfg(pvt, reg0, base0))
+		if (!amd64_read_dct_pci_cfg(pvt, 0, reg0, base0))
 			edac_dbg(0, "  DCSB0[%d]=0x%08x reg: F2x%x\n",
 				 cs, *base0, reg0);
 
-		if (pvt->fam == 0xf || dct_ganging_enabled(pvt))
+		if (pvt->fam == 0xf)
 			continue;
 
-		if (!amd64_read_dct_pci_cfg(pvt, reg1, base1))
+		if (!amd64_read_dct_pci_cfg(pvt, 1, reg0, base1))
 			edac_dbg(0, "  DCSB1[%d]=0x%08x reg: F2x%x\n",
-				 cs, *base1, reg1);
+				 cs, *base1, (pvt->fam == 0x10) ? reg1
+								: reg0);
 	}
 
 	for_each_chip_select_mask(cs, 0, pvt) {
@@ -786,38 +809,77 @@ static void read_dct_base_mask(struct amd64_pvt *pvt)
 		u32 *mask0 = &pvt->csels[0].csmasks[cs];
 		u32 *mask1 = &pvt->csels[1].csmasks[cs];
 
-		if (!amd64_read_dct_pci_cfg(pvt, reg0, mask0))
+		if (!amd64_read_dct_pci_cfg(pvt, 0, reg0, mask0))
 			edac_dbg(0, "    DCSM0[%d]=0x%08x reg: F2x%x\n",
 				 cs, *mask0, reg0);
 
-		if (pvt->fam == 0xf || dct_ganging_enabled(pvt))
+		if (pvt->fam == 0xf)
 			continue;
 
-		if (!amd64_read_dct_pci_cfg(pvt, reg1, mask1))
+		if (!amd64_read_dct_pci_cfg(pvt, 1, reg0, mask1))
 			edac_dbg(0, "    DCSM1[%d]=0x%08x reg: F2x%x\n",
-				 cs, *mask1, reg1);
+				 cs, *mask1, (pvt->fam == 0x10) ? reg1
+								: reg0);
 	}
 }
 
-static enum mem_type determine_memory_type(struct amd64_pvt *pvt, int cs)
+static void determine_memory_type(struct amd64_pvt *pvt)
 {
-	enum mem_type type;
+	u32 dram_ctrl, dcsm;
 
-	/* F15h supports only DDR3 */
-	if (pvt->fam >= 0x15)
-		type = (pvt->dclr0 & BIT(16)) ?	MEM_DDR3 : MEM_RDDR3;
-	else if (pvt->fam == 0x10 || pvt->ext_model >= K8_REV_F) {
+	switch (pvt->fam) {
+	case 0xf:
+		if (pvt->ext_model >= K8_REV_F)
+			goto ddr3;
+
+		pvt->dram_type = (pvt->dclr0 & BIT(18)) ? MEM_DDR : MEM_RDDR;
+		return;
+
+	case 0x10:
 		if (pvt->dchr0 & DDR3_MODE)
-			type = (pvt->dclr0 & BIT(16)) ?	MEM_DDR3 : MEM_RDDR3;
+			goto ddr3;
+
+		pvt->dram_type = (pvt->dclr0 & BIT(16)) ? MEM_DDR2 : MEM_RDDR2;
+		return;
+
+	case 0x15:
+		if (pvt->model < 0x60)
+			goto ddr3;
+
+		/*
+		 * Model 0x60h needs special handling:
+		 *
+		 * We use a Chip Select value of '0' to obtain dcsm.
+		 * Theoretically, it is possible to populate LRDIMMs of different
+		 * 'Rank' value on a DCT. But this is not the common case. So,
+		 * it's reasonable to assume all DIMMs are going to be of same
+		 * 'type' until proven otherwise.
+		 */
+		amd64_read_dct_pci_cfg(pvt, 0, DRAM_CONTROL, &dram_ctrl);
+		dcsm = pvt->csels[0].csmasks[0];
+
+		if (((dram_ctrl >> 8) & 0x7) == 0x2)
+			pvt->dram_type = MEM_DDR4;
+		else if (pvt->dclr0 & BIT(16))
+			pvt->dram_type = MEM_DDR3;
+		else if (dcsm & 0x3)
+			pvt->dram_type = MEM_LRDDR3;
 		else
-			type = (pvt->dclr0 & BIT(16)) ? MEM_DDR2 : MEM_RDDR2;
-	} else {
-		type = (pvt->dclr0 & BIT(18)) ? MEM_DDR : MEM_RDDR;
-	}
+			pvt->dram_type = MEM_RDDR3;
 
-	amd64_info("CS%d: %s\n", cs, edac_mem_types[type]);
+		return;
+
+	case 0x16:
+		goto ddr3;
 
-	return type;
+	default:
+		WARN(1, KERN_ERR "%s: Family??? 0x%x\n", __func__, pvt->fam);
+		pvt->dram_type = MEM_EMPTY;
+	}
+	return;
+
+ddr3:
+	pvt->dram_type = (pvt->dclr0 & BIT(16)) ? MEM_DDR3 : MEM_RDDR3;
 }
 
 /* Get the number of DCT channels the memory controller is using. */
@@ -944,8 +1006,12 @@ static void read_dram_base_limit_regs(struct amd64_pvt *pvt, unsigned range)
 	if (WARN_ON(!nb))
 		return;
 
-	pci_func = (pvt->model == 0x30) ? PCI_DEVICE_ID_AMD_15H_M30H_NB_F1
-					: PCI_DEVICE_ID_AMD_15H_NB_F1;
+	if (pvt->model == 0x60)
+		pci_func = PCI_DEVICE_ID_AMD_15H_M60H_NB_F1;
+	else if (pvt->model == 0x30)
+		pci_func = PCI_DEVICE_ID_AMD_15H_M30H_NB_F1;
+	else
+		pci_func = PCI_DEVICE_ID_AMD_15H_NB_F1;
 
 	f1 = pci_get_related_function(nb->misc->vendor, pci_func, nb->misc);
 	if (WARN_ON(!f1))
@@ -1035,7 +1101,7 @@ static int ddr2_cs_size(unsigned i, bool dct_width)
 }
 
 static int k8_dbam_to_chip_select(struct amd64_pvt *pvt, u8 dct,
-				  unsigned cs_mode)
+				  unsigned cs_mode, int cs_mask_nr)
 {
 	u32 dclr = dct ? pvt->dclr1 : pvt->dclr0;
 
@@ -1153,8 +1219,43 @@ static int ddr3_cs_size(unsigned i, bool dct_width)
 	return cs_size;
 }
 
+static int ddr3_lrdimm_cs_size(unsigned i, unsigned rank_multiply)
+{
+	unsigned shift = 0;
+	int cs_size = 0;
+
+	if (i < 4 || i == 6)
+		cs_size = -1;
+	else if (i == 12)
+		shift = 7;
+	else if (!(i & 0x1))
+		shift = i >> 1;
+	else
+		shift = (i + 1) >> 1;
+
+	if (cs_size != -1)
+		cs_size = rank_multiply * (128 << shift);
+
+	return cs_size;
+}
+
+static int ddr4_cs_size(unsigned i)
+{
+	int cs_size = 0;
+
+	if (i == 0)
+		cs_size = -1;
+	else if (i == 1)
+		cs_size = 1024;
+	else
+		/* Min cs_size = 1G */
+		cs_size = 1024 * (1 << (i >> 1));
+
+	return cs_size;
+}
+
 static int f10_dbam_to_chip_select(struct amd64_pvt *pvt, u8 dct,
-				   unsigned cs_mode)
+				   unsigned cs_mode, int cs_mask_nr)
 {
 	u32 dclr = dct ? pvt->dclr1 : pvt->dclr0;
 
@@ -1170,18 +1271,49 @@ static int f10_dbam_to_chip_select(struct amd64_pvt *pvt, u8 dct,
  * F15h supports only 64bit DCT interfaces
  */
 static int f15_dbam_to_chip_select(struct amd64_pvt *pvt, u8 dct,
-				   unsigned cs_mode)
+				   unsigned cs_mode, int cs_mask_nr)
 {
 	WARN_ON(cs_mode > 12);
 
 	return ddr3_cs_size(cs_mode, false);
 }
 
+/* F15h M60h supports DDR4 mapping as well.. */
+static int f15_m60h_dbam_to_chip_select(struct amd64_pvt *pvt, u8 dct,
+					unsigned cs_mode, int cs_mask_nr)
+{
+	int cs_size;
+	u32 dcsm = pvt->csels[dct].csmasks[cs_mask_nr];
+
+	WARN_ON(cs_mode > 12);
+
+	if (pvt->dram_type == MEM_DDR4) {
+		if (cs_mode > 9)
+			return -1;
+
+		cs_size = ddr4_cs_size(cs_mode);
+	} else if (pvt->dram_type == MEM_LRDDR3) {
+		unsigned rank_multiply = dcsm & 0xf;
+
+		if (rank_multiply == 3)
+			rank_multiply = 4;
+		cs_size = ddr3_lrdimm_cs_size(cs_mode, rank_multiply);
+	} else {
+		/* Minimum cs size is 512mb for F15hM60h*/
+		if (cs_mode == 0x1)
+			return -1;
+
+		cs_size = ddr3_cs_size(cs_mode, false);
+	}
+
+	return cs_size;
+}
+
 /*
  * F16h and F15h model 30h have only limited cs_modes.
  */
 static int f16_dbam_to_chip_select(struct amd64_pvt *pvt, u8 dct,
-				unsigned cs_mode)
+				unsigned cs_mode, int cs_mask_nr)
 {
 	WARN_ON(cs_mode > 12);
 
@@ -1198,7 +1330,7 @@ static void read_dram_ctl_register(struct amd64_pvt *pvt)
 	if (pvt->fam == 0xf)
 		return;
 
-	if (!amd64_read_dct_pci_cfg(pvt, DCT_SEL_LO, &pvt->dct_sel_lo)) {
+	if (!amd64_read_pci_cfg(pvt->F2, DCT_SEL_LO, &pvt->dct_sel_lo)) {
 		edac_dbg(0, "F2x110 (DCTSelLow): 0x%08x, High range addrs at: 0x%x\n",
 			 pvt->dct_sel_lo, dct_sel_baseaddr(pvt));
 
@@ -1219,7 +1351,7 @@ static void read_dram_ctl_register(struct amd64_pvt *pvt)
 			 dct_sel_interleave_addr(pvt));
 	}
 
-	amd64_read_dct_pci_cfg(pvt, DCT_SEL_HI, &pvt->dct_sel_hi);
+	amd64_read_pci_cfg(pvt->F2, DCT_SEL_HI, &pvt->dct_sel_hi);
 }
 
 /*
@@ -1430,7 +1562,7 @@ static u64 f1x_swap_interleaved_region(struct amd64_pvt *pvt, u64 sys_addr)
 			return sys_addr;
 	}
 
-	amd64_read_dct_pci_cfg(pvt, SWAP_INTLV_REG, &swap_reg);
+	amd64_read_pci_cfg(pvt->F2, SWAP_INTLV_REG, &swap_reg);
 
 	if (!(swap_reg & 0x1))
 		return sys_addr;
@@ -1723,10 +1855,16 @@ static void debug_display_dimm_sizes(struct amd64_pvt *pvt, u8 ctrl)
 		       WARN_ON(ctrl != 0);
 	}
 
-	dbam = (ctrl && !dct_ganging_enabled(pvt)) ? pvt->dbam1 : pvt->dbam0;
-	dcsb = (ctrl && !dct_ganging_enabled(pvt)) ? pvt->csels[1].csbases
-						   : pvt->csels[0].csbases;
-
+	if (pvt->fam == 0x10) {
+		dbam = (ctrl && !dct_ganging_enabled(pvt)) ? pvt->dbam1
+							   : pvt->dbam0;
+		dcsb = (ctrl && !dct_ganging_enabled(pvt)) ?
+				 pvt->csels[1].csbases :
+				 pvt->csels[0].csbases;
+	} else if (ctrl) {
+		dbam = pvt->dbam0;
+		dcsb = pvt->csels[1].csbases;
+	}
 	edac_dbg(1, "F2x%d80 (DRAM Bank Address Mapping): 0x%08x\n",
 		 ctrl, dbam);
 
@@ -1737,13 +1875,20 @@ static void debug_display_dimm_sizes(struct amd64_pvt *pvt, u8 ctrl)
 
 		size0 = 0;
 		if (dcsb[dimm*2] & DCSB_CS_ENABLE)
+			/* For f15m60h, need multiplier for LRDIMM cs_size
+			 * calculation. We pass 'dimm' value to the dbam_to_cs
+			 * mapper so we can find the multiplier from the
+			 * corresponding DCSM.
+			 */
 			size0 = pvt->ops->dbam_to_cs(pvt, ctrl,
-						     DBAM_DIMM(dimm, dbam));
+						     DBAM_DIMM(dimm, dbam),
+						     dimm);
 
 		size1 = 0;
 		if (dcsb[dimm*2 + 1] & DCSB_CS_ENABLE)
 			size1 = pvt->ops->dbam_to_cs(pvt, ctrl,
-						     DBAM_DIMM(dimm, dbam));
+						     DBAM_DIMM(dimm, dbam),
+						     dimm);
 
 		amd64_info(EDAC_MC ": %d: %5dMB %d: %5dMB\n",
 				dimm * 2,     size0,
@@ -1760,7 +1905,6 @@ static struct amd64_family_type family_types[] = {
 			.early_channel_count	= k8_early_channel_count,
 			.map_sysaddr_to_csrow	= k8_map_sysaddr_to_csrow,
 			.dbam_to_cs		= k8_dbam_to_chip_select,
-			.read_dct_pci_cfg	= k8_read_dct_pci_cfg,
 		}
 	},
 	[F10_CPUS] = {
@@ -1771,7 +1915,6 @@ static struct amd64_family_type family_types[] = {
 			.early_channel_count	= f1x_early_channel_count,
 			.map_sysaddr_to_csrow	= f1x_map_sysaddr_to_csrow,
 			.dbam_to_cs		= f10_dbam_to_chip_select,
-			.read_dct_pci_cfg	= f10_read_dct_pci_cfg,
 		}
 	},
 	[F15_CPUS] = {
@@ -1782,7 +1925,6 @@ static struct amd64_family_type family_types[] = {
 			.early_channel_count	= f1x_early_channel_count,
 			.map_sysaddr_to_csrow	= f1x_map_sysaddr_to_csrow,
 			.dbam_to_cs		= f15_dbam_to_chip_select,
-			.read_dct_pci_cfg	= f15_read_dct_pci_cfg,
 		}
 	},
 	[F15_M30H_CPUS] = {
@@ -1793,7 +1935,16 @@ static struct amd64_family_type family_types[] = {
 			.early_channel_count	= f1x_early_channel_count,
 			.map_sysaddr_to_csrow	= f1x_map_sysaddr_to_csrow,
 			.dbam_to_cs		= f16_dbam_to_chip_select,
-			.read_dct_pci_cfg	= f15_read_dct_pci_cfg,
+		}
+	},
+	[F15_M60H_CPUS] = {
+		.ctl_name = "F15h_M60h",
+		.f1_id = PCI_DEVICE_ID_AMD_15H_M60H_NB_F1,
+		.f3_id = PCI_DEVICE_ID_AMD_15H_M60H_NB_F3,
+		.ops = {
+			.early_channel_count	= f1x_early_channel_count,
+			.map_sysaddr_to_csrow	= f1x_map_sysaddr_to_csrow,
+			.dbam_to_cs		= f15_m60h_dbam_to_chip_select,
 		}
 	},
 	[F16_CPUS] = {
@@ -1804,7 +1955,6 @@ static struct amd64_family_type family_types[] = {
 			.early_channel_count	= f1x_early_channel_count,
 			.map_sysaddr_to_csrow	= f1x_map_sysaddr_to_csrow,
 			.dbam_to_cs		= f16_dbam_to_chip_select,
-			.read_dct_pci_cfg	= f10_read_dct_pci_cfg,
 		}
 	},
 	[F16_M30H_CPUS] = {
@@ -1815,7 +1965,6 @@ static struct amd64_family_type family_types[] = {
 			.early_channel_count	= f1x_early_channel_count,
 			.map_sysaddr_to_csrow	= f1x_map_sysaddr_to_csrow,
 			.dbam_to_cs		= f16_dbam_to_chip_select,
-			.read_dct_pci_cfg	= f10_read_dct_pci_cfg,
 		}
 	},
 };
@@ -2148,25 +2297,27 @@ static void read_mc_regs(struct amd64_pvt *pvt)
 	read_dct_base_mask(pvt);
 
 	amd64_read_pci_cfg(pvt->F1, DHAR, &pvt->dhar);
-	amd64_read_dct_pci_cfg(pvt, DBAM0, &pvt->dbam0);
+	amd64_read_dct_pci_cfg(pvt, 0, DBAM0, &pvt->dbam0);
 
 	amd64_read_pci_cfg(pvt->F3, F10_ONLINE_SPARE, &pvt->online_spare);
 
-	amd64_read_dct_pci_cfg(pvt, DCLR0, &pvt->dclr0);
-	amd64_read_dct_pci_cfg(pvt, DCHR0, &pvt->dchr0);
+	amd64_read_dct_pci_cfg(pvt, 0, DCLR0, &pvt->dclr0);
+	amd64_read_dct_pci_cfg(pvt, 0, DCHR0, &pvt->dchr0);
 
 	if (!dct_ganging_enabled(pvt)) {
-		amd64_read_dct_pci_cfg(pvt, DCLR1, &pvt->dclr1);
-		amd64_read_dct_pci_cfg(pvt, DCHR1, &pvt->dchr1);
+		amd64_read_dct_pci_cfg(pvt, 1, DCLR0, &pvt->dclr1);
+		amd64_read_dct_pci_cfg(pvt, 1, DCHR0, &pvt->dchr1);
 	}
 
 	pvt->ecc_sym_sz = 4;
+	determine_memory_type(pvt);
+	edac_dbg(1, "  DIMM type: %s\n", edac_mem_types[pvt->dram_type]);
 
 	if (pvt->fam >= 0x10) {
 		amd64_read_pci_cfg(pvt->F3, EXT_NB_MCA_CFG, &tmp);
+		/* F16h has only DCT0, so no need to read dbam1 */
 		if (pvt->fam != 0x16)
-			/* F16h has only DCT0 */
-			amd64_read_dct_pci_cfg(pvt, DBAM1, &pvt->dbam1);
+			amd64_read_dct_pci_cfg(pvt, 1, DBAM0, &pvt->dbam1);
 
 		/* F10h, revD and later can do x8 ECC too */
 		if ((pvt->fam > 0x10 || pvt->model > 7) && tmp & BIT(25))
@@ -2224,7 +2375,8 @@ static u32 get_csrow_nr_pages(struct amd64_pvt *pvt, u8 dct, int csrow_nr)
 	 */
 	cs_mode = DBAM_DIMM(csrow_nr / 2, dbam);
 
-	nr_pages = pvt->ops->dbam_to_cs(pvt, dct, cs_mode) << (20 - PAGE_SHIFT);
+	nr_pages = pvt->ops->dbam_to_cs(pvt, dct, cs_mode, (csrow_nr / 2))
+							   << (20 - PAGE_SHIFT);
 
 	edac_dbg(0, "csrow: %d, channel: %d, DBAM idx: %d\n",
 		    csrow_nr, dct,  cs_mode);
@@ -2243,7 +2395,6 @@ static int init_csrows(struct mem_ctl_info *mci)
 	struct csrow_info *csrow;
 	struct dimm_info *dimm;
 	enum edac_type edac_mode;
-	enum mem_type mtype;
 	int i, j, empty = 1;
 	int nr_pages = 0;
 	u32 val;
@@ -2288,8 +2439,6 @@ static int init_csrows(struct mem_ctl_info *mci)
 			nr_pages += row_dct1_pages;
 		}
 
-		mtype = determine_memory_type(pvt, i);
-
 		edac_dbg(1, "Total csrow%d pages: %u\n", i, nr_pages);
 
 		/*
@@ -2303,7 +2452,7 @@ static int init_csrows(struct mem_ctl_info *mci)
 
 		for (j = 0; j < pvt->channel_count; j++) {
 			dimm = csrow->channels[j]->dimm;
-			dimm->mtype = mtype;
+			dimm->mtype = pvt->dram_type;
 			dimm->edac_mode = edac_mode;
 		}
 	}
@@ -2590,6 +2739,10 @@ static struct amd64_family_type *per_family_init(struct amd64_pvt *pvt)
 			fam_type = &family_types[F15_M30H_CPUS];
 			pvt->ops = &family_types[F15_M30H_CPUS].ops;
 			break;
+		} else if (pvt->model == 0x60) {
+			fam_type = &family_types[F15_M60H_CPUS];
+			pvt->ops = &family_types[F15_M60H_CPUS].ops;
+			break;
 		}
 
 		fam_type	= &family_types[F15_CPUS];
@@ -2814,55 +2967,13 @@ static void remove_one_instance(struct pci_dev *pdev)
  * inquiry this table to see if this driver is for a given device found.
  */
 static const struct pci_device_id amd64_pci_table[] = {
-	{
-		.vendor		= PCI_VENDOR_ID_AMD,
-		.device		= PCI_DEVICE_ID_AMD_K8_NB_MEMCTL,
-		.subvendor	= PCI_ANY_ID,
-		.subdevice	= PCI_ANY_ID,
-		.class		= 0,
-		.class_mask	= 0,
-	},
-	{
-		.vendor		= PCI_VENDOR_ID_AMD,
-		.device		= PCI_DEVICE_ID_AMD_10H_NB_DRAM,
-		.subvendor	= PCI_ANY_ID,
-		.subdevice	= PCI_ANY_ID,
-		.class		= 0,
-		.class_mask	= 0,
-	},
-	{
-		.vendor		= PCI_VENDOR_ID_AMD,
-		.device		= PCI_DEVICE_ID_AMD_15H_NB_F2,
-		.subvendor	= PCI_ANY_ID,
-		.subdevice	= PCI_ANY_ID,
-		.class		= 0,
-		.class_mask	= 0,
-	},
-	{
-		.vendor		= PCI_VENDOR_ID_AMD,
-		.device		= PCI_DEVICE_ID_AMD_15H_M30H_NB_F2,
-		.subvendor	= PCI_ANY_ID,
-		.subdevice	= PCI_ANY_ID,
-		.class		= 0,
-		.class_mask	= 0,
-	},
-	{
-		.vendor		= PCI_VENDOR_ID_AMD,
-		.device		= PCI_DEVICE_ID_AMD_16H_NB_F2,
-		.subvendor	= PCI_ANY_ID,
-		.subdevice	= PCI_ANY_ID,
-		.class		= 0,
-		.class_mask	= 0,
-	},
-	{
-		.vendor		= PCI_VENDOR_ID_AMD,
-		.device		= PCI_DEVICE_ID_AMD_16H_M30H_NB_F2,
-		.subvendor	= PCI_ANY_ID,
-		.subdevice	= PCI_ANY_ID,
-		.class		= 0,
-		.class_mask	= 0,
-	},
-
+	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_K8_NB_MEMCTL) },
+	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_10H_NB_DRAM) },
+	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_NB_F2) },
+	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_M30H_NB_F2) },
+	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_M60H_NB_F2) },
+	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_16H_NB_F2) },
+	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_16H_M30H_NB_F2) },
 	{0, }
 };
 MODULE_DEVICE_TABLE(pci, amd64_pci_table);
@@ -2924,6 +3035,11 @@ static int __init amd64_edac_init(void)
 		goto err_no_instances;
 
 	setup_pci_device();
+
+#ifdef CONFIG_X86_32
+	amd64_err("%s on 32-bit is unsupported. USE AT YOUR OWN RISK!\n", EDAC_MOD_STR);
+#endif
+
 	return 0;
 
 err_no_instances:
diff --git a/drivers/edac/amd64_edac.h b/drivers/edac/amd64_edac.h
index d903e0c21144..d8468c667925 100644
--- a/drivers/edac/amd64_edac.h
+++ b/drivers/edac/amd64_edac.h
@@ -162,10 +162,12 @@
 /*
  * PCI-defined configuration space registers
  */
-#define PCI_DEVICE_ID_AMD_15H_M30H_NB_F1 0x141b
-#define PCI_DEVICE_ID_AMD_15H_M30H_NB_F2 0x141c
 #define PCI_DEVICE_ID_AMD_15H_NB_F1	0x1601
 #define PCI_DEVICE_ID_AMD_15H_NB_F2	0x1602
+#define PCI_DEVICE_ID_AMD_15H_M30H_NB_F1 0x141b
+#define PCI_DEVICE_ID_AMD_15H_M30H_NB_F2 0x141c
+#define PCI_DEVICE_ID_AMD_15H_M60H_NB_F1 0x1571
+#define PCI_DEVICE_ID_AMD_15H_M60H_NB_F2 0x1572
 #define PCI_DEVICE_ID_AMD_16H_NB_F1	0x1531
 #define PCI_DEVICE_ID_AMD_16H_NB_F2	0x1532
 #define PCI_DEVICE_ID_AMD_16H_M30H_NB_F1 0x1581
@@ -221,6 +223,8 @@
 
 #define csrow_enabled(i, dct, pvt)	((pvt)->csels[(dct)].csbases[(i)] & DCSB_CS_ENABLE)
 
+#define DRAM_CONTROL			0x78
+
 #define DBAM0				0x80
 #define DBAM1				0x180
 
@@ -301,6 +305,7 @@ enum amd_families {
 	F10_CPUS,
 	F15_CPUS,
 	F15_M30H_CPUS,
+	F15_M60H_CPUS,
 	F16_CPUS,
 	F16_M30H_CPUS,
 	NUM_FAMILIES,
@@ -379,6 +384,9 @@ struct amd64_pvt {
 
 	/* place to store error injection parameters prior to issue */
 	struct error_injection injection;
+
+	/* cache the dram_type */
+	enum mem_type dram_type;
 };
 
 enum err_codes {
@@ -480,9 +488,8 @@ struct low_ops {
 	int (*early_channel_count)	(struct amd64_pvt *pvt);
 	void (*map_sysaddr_to_csrow)	(struct mem_ctl_info *mci, u64 sys_addr,
 					 struct err_info *);
-	int (*dbam_to_cs)		(struct amd64_pvt *pvt, u8 dct, unsigned cs_mode);
-	int (*read_dct_pci_cfg)		(struct amd64_pvt *pvt, int offset,
-					 u32 *val, const char *func);
+	int (*dbam_to_cs)		(struct amd64_pvt *pvt, u8 dct,
+					 unsigned cs_mode, int cs_mask_nr);
 };
 
 struct amd64_family_type {
@@ -502,9 +509,6 @@ int __amd64_write_pci_cfg_dword(struct pci_dev *pdev, int offset,
 #define amd64_write_pci_cfg(pdev, offset, val)	\
 	__amd64_write_pci_cfg_dword(pdev, offset, val, __func__)
 
-#define amd64_read_dct_pci_cfg(pvt, offset, val) \
-	pvt->ops->read_dct_pci_cfg(pvt, offset, val, __func__)
-
 int amd64_get_dram_hole_info(struct mem_ctl_info *mci, u64 *hole_base,
 			     u64 *hole_offset, u64 *hole_size);
 
diff --git a/drivers/edac/cell_edac.c b/drivers/edac/cell_edac.c
index a12c8552f6a6..a9259b069dcd 100644
--- a/drivers/edac/cell_edac.c
+++ b/drivers/edac/cell_edac.c
@@ -245,7 +245,6 @@ static int cell_edac_remove(struct platform_device *pdev)
 static struct platform_driver cell_edac_driver = {
 	.driver		= {
 		.name	= "cbe-mic",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= cell_edac_probe,
 	.remove		= cell_edac_remove,
diff --git a/drivers/edac/cpc925_edac.c b/drivers/edac/cpc925_edac.c
index df6575f1430d..682288ced4ac 100644
--- a/drivers/edac/cpc925_edac.c
+++ b/drivers/edac/cpc925_edac.c
@@ -562,7 +562,7 @@ static void cpc925_mc_check(struct mem_ctl_info *mci)
 
 	if (apiexcp & UECC_EXCP_DETECTED) {
 		cpc925_mc_printk(mci, KERN_INFO, "DRAM UECC Fault\n");
-		edac_mc_handle_error(HW_EVENT_ERR_CORRECTED, mci, 1,
+		edac_mc_handle_error(HW_EVENT_ERR_UNCORRECTED, mci, 1,
 				     pfn, offset, 0,
 				     csrow, -1, -1,
 				     mci->ctl_name, "");
diff --git a/drivers/edac/e7xxx_edac.c b/drivers/edac/e7xxx_edac.c
index 3cda79bc8b00..ece3aef16bb1 100644
--- a/drivers/edac/e7xxx_edac.c
+++ b/drivers/edac/e7xxx_edac.c
@@ -226,7 +226,7 @@ static void process_ce(struct mem_ctl_info *mci, struct e7xxx_error_info *info)
 static void process_ce_no_info(struct mem_ctl_info *mci)
 {
 	edac_dbg(3, "\n");
-	edac_mc_handle_error(HW_EVENT_ERR_UNCORRECTED, mci, 1, 0, 0, 0, -1, -1, -1,
+	edac_mc_handle_error(HW_EVENT_ERR_CORRECTED, mci, 1, 0, 0, 0, -1, -1, -1,
 			     "e7xxx CE log register overflow", "");
 }
 
diff --git a/drivers/edac/edac_core.h b/drivers/edac/edac_core.h
index 3c2625e7980d..6c9f381e8fe6 100644
--- a/drivers/edac/edac_core.h
+++ b/drivers/edac/edac_core.h
@@ -66,7 +66,7 @@
 #define EDAC_PCI "PCI"
 #define EDAC_DEBUG "DEBUG"
 
-extern const char *edac_mem_types[];
+extern const char * const edac_mem_types[];
 
 #ifdef CONFIG_EDAC_DEBUG
 extern int edac_debug_level;
diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
index 9f134823fa75..1747906f10ce 100644
--- a/drivers/edac/edac_mc.c
+++ b/drivers/edac/edac_mc.c
@@ -125,27 +125,27 @@ static void edac_mc_dump_mci(struct mem_ctl_info *mci)
 
 #endif				/* CONFIG_EDAC_DEBUG */
 
-/*
- * keep those in sync with the enum mem_type
- */
-const char *edac_mem_types[] = {
-	"Empty csrow",
-	"Reserved csrow type",
-	"Unknown csrow type",
-	"Fast page mode RAM",
-	"Extended data out RAM",
-	"Burst Extended data out RAM",
-	"Single data rate SDRAM",
-	"Registered single data rate SDRAM",
-	"Double data rate SDRAM",
-	"Registered Double data rate SDRAM",
-	"Rambus DRAM",
-	"Unbuffered DDR2 RAM",
-	"Fully buffered DDR2",
-	"Registered DDR2 RAM",
-	"Rambus XDR",
-	"Unbuffered DDR3 RAM",
-	"Registered DDR3 RAM",
+const char * const edac_mem_types[] = {
+	[MEM_EMPTY]	= "Empty csrow",
+	[MEM_RESERVED]	= "Reserved csrow type",
+	[MEM_UNKNOWN]	= "Unknown csrow type",
+	[MEM_FPM]	= "Fast page mode RAM",
+	[MEM_EDO]	= "Extended data out RAM",
+	[MEM_BEDO]	= "Burst Extended data out RAM",
+	[MEM_SDR]	= "Single data rate SDRAM",
+	[MEM_RDR]	= "Registered single data rate SDRAM",
+	[MEM_DDR]	= "Double data rate SDRAM",
+	[MEM_RDDR]	= "Registered Double data rate SDRAM",
+	[MEM_RMBS]	= "Rambus DRAM",
+	[MEM_DDR2]	= "Unbuffered DDR2 RAM",
+	[MEM_FB_DDR2]	= "Fully buffered DDR2",
+	[MEM_RDDR2]	= "Registered DDR2 RAM",
+	[MEM_XDR]	= "Rambus XDR",
+	[MEM_DDR3]	= "Unbuffered DDR3 RAM",
+	[MEM_RDDR3]	= "Registered DDR3 RAM",
+	[MEM_LRDDR3]	= "Load-Reduced DDR3 RAM",
+	[MEM_DDR4]	= "Unbuffered DDR4 RAM",
+	[MEM_RDDR4]	= "Registered DDR4 RAM",
 };
 EXPORT_SYMBOL_GPL(edac_mem_types);
 
diff --git a/drivers/edac/edac_mc_sysfs.c b/drivers/edac/edac_mc_sysfs.c
index a6cd36100663..670d2829c547 100644
--- a/drivers/edac/edac_mc_sysfs.c
+++ b/drivers/edac/edac_mc_sysfs.c
@@ -372,7 +372,7 @@ static int edac_create_csrow_object(struct mem_ctl_info *mci,
 {
 	int err, chan;
 
-	if (csrow->nr_channels >= EDAC_NR_CHANNELS)
+	if (csrow->nr_channels > EDAC_NR_CHANNELS)
 		return -ENODEV;
 
 	csrow->dev.type = &csrow_attr_type;
diff --git a/drivers/edac/edac_pci_sysfs.c b/drivers/edac/edac_pci_sysfs.c
index e8658e451762..24d877f6e577 100644
--- a/drivers/edac/edac_pci_sysfs.c
+++ b/drivers/edac/edac_pci_sysfs.c
@@ -14,9 +14,6 @@
 #include "edac_core.h"
 #include "edac_module.h"
 
-/* Turn off this whole feature if PCI is not configured */
-#ifdef CONFIG_PCI
-
 #define EDAC_PCI_SYMLINK	"device"
 
 /* data variables exported via sysfs */
@@ -761,5 +758,3 @@ MODULE_PARM_DESC(check_pci_errors,
 module_param(edac_pci_panic_on_pe, int, 0644);
 MODULE_PARM_DESC(edac_pci_panic_on_pe,
 		 "Panic on PCI Bus Parity error: 0=off 1=on");
-
-#endif				/* CONFIG_PCI */
diff --git a/drivers/edac/ghes_edac.c b/drivers/edac/ghes_edac.c
index 8399b4e16fe0..b24681998740 100644
--- a/drivers/edac/ghes_edac.c
+++ b/drivers/edac/ghes_edac.c
@@ -413,8 +413,8 @@ void ghes_edac_report_mem_error(struct ghes *ghes, int sev,
 
 	/* Generate the trace event */
 	grain_bits = fls_long(e->grain);
-	sprintf(pvt->detail_location, "APEI location: %s %s",
-		e->location, e->other_detail);
+	snprintf(pvt->detail_location, sizeof(pvt->detail_location),
+		 "APEI location: %s %s", e->location, e->other_detail);
 	trace_mc_event(type, e->msg, e->label, e->error_count,
 		       mci->mc_idx, e->top_layer, e->mid_layer, e->low_layer,
 		       PAGES_TO_MiB(e->page_frame_number) | e->offset_in_page,
diff --git a/drivers/edac/i3000_edac.c b/drivers/edac/i3000_edac.c
index cd28b968e5c7..5cb36a6022cc 100644
--- a/drivers/edac/i3000_edac.c
+++ b/drivers/edac/i3000_edac.c
@@ -542,8 +542,7 @@ fail1:
 	pci_unregister_driver(&i3000_driver);
 
 fail0:
-	if (mci_pdev)
-		pci_dev_put(mci_pdev);
+	pci_dev_put(mci_pdev);
 
 	return pci_rc;
 }
diff --git a/drivers/edac/i3200_edac.c b/drivers/edac/i3200_edac.c
index 022a70273ada..4ad062b0ef26 100644
--- a/drivers/edac/i3200_edac.c
+++ b/drivers/edac/i3200_edac.c
@@ -242,11 +242,11 @@ static void i3200_process_error_info(struct mem_ctl_info *mci,
 					     -1, -1,
 					     "i3000 UE", "");
 		} else if (log & I3200_ECCERRLOG_CE) {
-			edac_mc_handle_error(HW_EVENT_ERR_UNCORRECTED, mci, 1,
+			edac_mc_handle_error(HW_EVENT_ERR_CORRECTED, mci, 1,
 					     0, 0, eccerrlog_syndrome(log),
 					     eccerrlog_row(channel, log),
 					     -1, -1,
-					     "i3000 UE", "");
+					     "i3000 CE", "");
 		}
 	}
 }
@@ -523,8 +523,7 @@ fail1:
 	pci_unregister_driver(&i3200_driver);
 
 fail0:
-	if (mci_pdev)
-		pci_dev_put(mci_pdev);
+	pci_dev_put(mci_pdev);
 
 	return pci_rc;
 }
diff --git a/drivers/edac/i82443bxgx_edac.c b/drivers/edac/i82443bxgx_edac.c
index d730e276d1a8..b4705d9366bf 100644
--- a/drivers/edac/i82443bxgx_edac.c
+++ b/drivers/edac/i82443bxgx_edac.c
@@ -458,8 +458,7 @@ static void __exit i82443bxgx_edacmc_exit(void)
 	if (!i82443bxgx_registered)
 		i82443bxgx_edacmc_remove_one(mci_pdev);
 
-	if (mci_pdev)
-		pci_dev_put(mci_pdev);
+	pci_dev_put(mci_pdev);
 }
 
 module_init(i82443bxgx_edacmc_init);
diff --git a/drivers/edac/i82860_edac.c b/drivers/edac/i82860_edac.c
index 3382f6344e42..4382343a7c60 100644
--- a/drivers/edac/i82860_edac.c
+++ b/drivers/edac/i82860_edac.c
@@ -124,7 +124,7 @@ static int i82860_process_error_info(struct mem_ctl_info *mci,
 				     dimm->location[0], dimm->location[1], -1,
 				     "i82860 UE", "");
 	else
-		edac_mc_handle_error(HW_EVENT_ERR_UNCORRECTED, mci, 1,
+		edac_mc_handle_error(HW_EVENT_ERR_CORRECTED, mci, 1,
 				     info->eap, 0, info->derrsyn,
 				     dimm->location[0], dimm->location[1], -1,
 				     "i82860 CE", "");
diff --git a/drivers/edac/mce_amd.c b/drivers/edac/mce_amd.c
index f78c1c54dbd5..58586d59bf8e 100644
--- a/drivers/edac/mce_amd.c
+++ b/drivers/edac/mce_amd.c
@@ -138,6 +138,15 @@ static const char * const mc5_mce_desc[] = {
 	"Retire status queue"
 };
 
+static const char * const mc6_mce_desc[] = {
+	"Hardware Assertion",
+	"Free List",
+	"Physical Register File",
+	"Retire Queue",
+	"Scheduler table",
+	"Status Register File",
+};
+
 static bool f12h_mc0_mce(u16 ec, u8 xec)
 {
 	bool ret = false;
@@ -432,8 +441,8 @@ static bool k8_mc2_mce(u16 ec, u8 xec)
 		pr_cont(": %s error in the L2 cache tags.\n", R4_MSG(ec));
 	else if (xec == 0x0) {
 		if (TLB_ERROR(ec))
-			pr_cont(": %s error in a Page Descriptor Cache or "
-				"Guest TLB.\n", TT_MSG(ec));
+			pr_cont("%s error in a Page Descriptor Cache or Guest TLB.\n",
+				TT_MSG(ec));
 		else if (BUS_ERROR(ec))
 			pr_cont(": %s/ECC error in data read from NB: %s.\n",
 				R4_MSG(ec), PP_MSG(ec));
@@ -672,38 +681,10 @@ static void decode_mc6_mce(struct mce *m)
 
 	pr_emerg(HW_ERR "MC6 Error: ");
 
-	switch (xec) {
-	case 0x0:
-		pr_cont("Hardware Assertion");
-		break;
-
-	case 0x1:
-		pr_cont("Free List");
-		break;
-
-	case 0x2:
-		pr_cont("Physical Register File");
-		break;
-
-	case 0x3:
-		pr_cont("Retire Queue");
-		break;
-
-	case 0x4:
-		pr_cont("Scheduler table");
-		break;
-
-	case 0x5:
-		pr_cont("Status Register File");
-		break;
-
-	default:
+	if (xec > 0x5)
 		goto wrong_mc6_mce;
-		break;
-	}
-
-	pr_cont(" parity error.\n");
 
+	pr_cont("%s parity error.\n", mc6_mce_desc[xec]);
 	return;
 
  wrong_mc6_mce:
@@ -800,7 +781,7 @@ int amd_decode_mce(struct notifier_block *nb, unsigned long val, void *data)
 	pr_cont("]: 0x%016llx\n", m->status);
 
 	if (m->status & MCI_STATUS_ADDRV)
-		pr_emerg(HW_ERR "MC%d_ADDR: 0x%016llx\n", m->bank, m->addr);
+		pr_emerg(HW_ERR "MC%d Error Address: 0x%016llx\n", m->bank, m->addr);
 
 	if (!fam_ops)
 		goto err_code;
diff --git a/drivers/edac/mce_amd.h b/drivers/edac/mce_amd.h
index 51b7e3a36e37..c2359a1ea6b3 100644
--- a/drivers/edac/mce_amd.h
+++ b/drivers/edac/mce_amd.h
@@ -32,9 +32,6 @@
 #define R4(x)				(((x) >> 4) & 0xf)
 #define R4_MSG(x)			((R4(x) < 9) ?  rrrr_msgs[R4(x)] : "Wrong R4!")
 
-#define MCI_STATUS_DEFERRED		BIT_64(44)
-#define MCI_STATUS_POISON		BIT_64(43)
-
 extern const char * const pp_msgs[];
 
 enum tt_ids {
diff --git a/drivers/edac/mce_amd_inj.c b/drivers/edac/mce_amd_inj.c
index 5e46a9fea31b..0bd91a802c67 100644
--- a/drivers/edac/mce_amd_inj.c
+++ b/drivers/edac/mce_amd_inj.c
@@ -1,173 +1,262 @@
 /*
- * A simple MCE injection facility for testing the MCE decoding code. This
- * driver should be built as module so that it can be loaded on production
- * kernels for testing purposes.
+ * A simple MCE injection facility for testing different aspects of the RAS
+ * code. This driver should be built as module so that it can be loaded
+ * on production kernels for testing purposes.
  *
  * This file may be distributed under the terms of the GNU General Public
  * License version 2.
  *
- * Copyright (c) 2010:  Borislav Petkov <bp@alien8.de>
+ * Copyright (c) 2010-14:  Borislav Petkov <bp@alien8.de>
  *			Advanced Micro Devices Inc.
  */
 
 #include <linux/kobject.h>
+#include <linux/debugfs.h>
 #include <linux/device.h>
-#include <linux/edac.h>
 #include <linux/module.h>
+#include <linux/cpu.h>
 #include <asm/mce.h>
 
 #include "mce_amd.h"
 
-struct edac_mce_attr {
-	struct attribute attr;
-	ssize_t (*show) (struct kobject *kobj, struct edac_mce_attr *attr, char *buf);
-	ssize_t (*store)(struct kobject *kobj, struct edac_mce_attr *attr,
-			 const char *buf, size_t count);
-};
-
-#define EDAC_MCE_ATTR(_name, _mode, _show, _store)			\
-static struct edac_mce_attr mce_attr_##_name = __ATTR(_name, _mode, _show, _store)
-
-static struct kobject *mce_kobj;
-
 /*
  * Collect all the MCi_XXX settings
  */
 static struct mce i_mce;
+static struct dentry *dfs_inj;
 
-#define MCE_INJECT_STORE(reg)						\
-static ssize_t edac_inject_##reg##_store(struct kobject *kobj,		\
-					 struct edac_mce_attr *attr,	\
-					 const char *data, size_t count)\
+#define MCE_INJECT_SET(reg)						\
+static int inj_##reg##_set(void *data, u64 val)				\
 {									\
-	int ret = 0;							\
-	unsigned long value;						\
-									\
-	ret = kstrtoul(data, 16, &value);				\
-	if (ret < 0)							\
-		printk(KERN_ERR "Error writing MCE " #reg " field.\n");	\
+	struct mce *m = (struct mce *)data;				\
 									\
-	i_mce.reg = value;						\
-									\
-	return count;							\
+	m->reg = val;							\
+	return 0;							\
 }
 
-MCE_INJECT_STORE(status);
-MCE_INJECT_STORE(misc);
-MCE_INJECT_STORE(addr);
+MCE_INJECT_SET(status);
+MCE_INJECT_SET(misc);
+MCE_INJECT_SET(addr);
 
-#define MCE_INJECT_SHOW(reg)						\
-static ssize_t edac_inject_##reg##_show(struct kobject *kobj,		\
-					struct edac_mce_attr *attr,	\
-					char *buf)			\
+#define MCE_INJECT_GET(reg)						\
+static int inj_##reg##_get(void *data, u64 *val)			\
 {									\
-	return sprintf(buf, "0x%016llx\n", i_mce.reg);			\
+	struct mce *m = (struct mce *)data;				\
+									\
+	*val = m->reg;							\
+	return 0;							\
 }
 
-MCE_INJECT_SHOW(status);
-MCE_INJECT_SHOW(misc);
-MCE_INJECT_SHOW(addr);
+MCE_INJECT_GET(status);
+MCE_INJECT_GET(misc);
+MCE_INJECT_GET(addr);
 
-EDAC_MCE_ATTR(status, 0644, edac_inject_status_show, edac_inject_status_store);
-EDAC_MCE_ATTR(misc, 0644, edac_inject_misc_show, edac_inject_misc_store);
-EDAC_MCE_ATTR(addr, 0644, edac_inject_addr_show, edac_inject_addr_store);
+DEFINE_SIMPLE_ATTRIBUTE(status_fops, inj_status_get, inj_status_set, "%llx\n");
+DEFINE_SIMPLE_ATTRIBUTE(misc_fops, inj_misc_get, inj_misc_set, "%llx\n");
+DEFINE_SIMPLE_ATTRIBUTE(addr_fops, inj_addr_get, inj_addr_set, "%llx\n");
 
 /*
- * This denotes into which bank we're injecting and triggers
- * the injection, at the same time.
+ * Caller needs to be make sure this cpu doesn't disappear
+ * from under us, i.e.: get_cpu/put_cpu.
  */
-static ssize_t edac_inject_bank_store(struct kobject *kobj,
-				      struct edac_mce_attr *attr,
-				      const char *data, size_t count)
+static int toggle_hw_mce_inject(unsigned int cpu, bool enable)
 {
-	int ret = 0;
-	unsigned long value;
+	u32 l, h;
+	int err;
 
-	ret = kstrtoul(data, 10, &value);
-	if (ret < 0) {
-		printk(KERN_ERR "Invalid bank value!\n");
-		return -EINVAL;
+	err = rdmsr_on_cpu(cpu, MSR_K7_HWCR, &l, &h);
+	if (err) {
+		pr_err("%s: error reading HWCR\n", __func__);
+		return err;
 	}
 
-	if (value > 5)
-		if (boot_cpu_data.x86 != 0x15 || value > 6) {
-			printk(KERN_ERR "Non-existent MCE bank: %lu\n", value);
-			return -EINVAL;
-		}
+	enable ? (l |= BIT(18)) : (l &= ~BIT(18));
 
-	i_mce.bank = value;
+	err = wrmsr_on_cpu(cpu, MSR_K7_HWCR, l, h);
+	if (err)
+		pr_err("%s: error writing HWCR\n", __func__);
 
-	amd_decode_mce(NULL, 0, &i_mce);
+	return err;
+}
 
-	return count;
+static int flags_get(void *data, u64 *val)
+{
+	struct mce *m = (struct mce *)data;
+
+	*val = m->inject_flags;
+
+	return 0;
 }
 
-static ssize_t edac_inject_bank_show(struct kobject *kobj,
-				     struct edac_mce_attr *attr, char *buf)
+static int flags_set(void *data, u64 val)
 {
-	return sprintf(buf, "%d\n", i_mce.bank);
+	struct mce *m = (struct mce *)data;
+
+	m->inject_flags = (u8)val;
+	return 0;
 }
 
-EDAC_MCE_ATTR(bank, 0644, edac_inject_bank_show, edac_inject_bank_store);
+DEFINE_SIMPLE_ATTRIBUTE(flags_fops, flags_get, flags_set, "%llu\n");
 
-static struct edac_mce_attr *sysfs_attrs[] = { &mce_attr_status, &mce_attr_misc,
-					       &mce_attr_addr, &mce_attr_bank
-};
+/*
+ * On which CPU to inject?
+ */
+MCE_INJECT_GET(extcpu);
 
-static int __init edac_init_mce_inject(void)
+static int inj_extcpu_set(void *data, u64 val)
 {
-	struct bus_type *edac_subsys = NULL;
-	int i, err = 0;
+	struct mce *m = (struct mce *)data;
 
-	edac_subsys = edac_get_sysfs_subsys();
-	if (!edac_subsys)
+	if (val >= nr_cpu_ids || !cpu_online(val)) {
+		pr_err("%s: Invalid CPU: %llu\n", __func__, val);
 		return -EINVAL;
+	}
+	m->extcpu = val;
+	return 0;
+}
+
+DEFINE_SIMPLE_ATTRIBUTE(extcpu_fops, inj_extcpu_get, inj_extcpu_set, "%llu\n");
 
-	mce_kobj = kobject_create_and_add("mce", &edac_subsys->dev_root->kobj);
-	if (!mce_kobj) {
-		printk(KERN_ERR "Error creating a mce kset.\n");
-		err = -ENOMEM;
-		goto err_mce_kobj;
+static void trigger_mce(void *info)
+{
+	asm volatile("int $18");
+}
+
+static void do_inject(void)
+{
+	u64 mcg_status = 0;
+	unsigned int cpu = i_mce.extcpu;
+	u8 b = i_mce.bank;
+
+	if (!(i_mce.inject_flags & MCJ_EXCEPTION)) {
+		amd_decode_mce(NULL, 0, &i_mce);
+		return;
 	}
 
-	for (i = 0; i < ARRAY_SIZE(sysfs_attrs); i++) {
-		err = sysfs_create_file(mce_kobj, &sysfs_attrs[i]->attr);
-		if (err) {
-			printk(KERN_ERR "Error creating %s in sysfs.\n",
-					sysfs_attrs[i]->attr.name);
-			goto err_sysfs_create;
+	get_online_cpus();
+	if (!cpu_online(cpu))
+		goto err;
+
+	/* prep MCE global settings for the injection */
+	mcg_status = MCG_STATUS_MCIP | MCG_STATUS_EIPV;
+
+	if (!(i_mce.status & MCI_STATUS_PCC))
+		mcg_status |= MCG_STATUS_RIPV;
+
+	toggle_hw_mce_inject(cpu, true);
+
+	wrmsr_on_cpu(cpu, MSR_IA32_MCG_STATUS,
+		     (u32)mcg_status, (u32)(mcg_status >> 32));
+
+	wrmsr_on_cpu(cpu, MSR_IA32_MCx_STATUS(b),
+		     (u32)i_mce.status, (u32)(i_mce.status >> 32));
+
+	wrmsr_on_cpu(cpu, MSR_IA32_MCx_ADDR(b),
+		     (u32)i_mce.addr, (u32)(i_mce.addr >> 32));
+
+	wrmsr_on_cpu(cpu, MSR_IA32_MCx_MISC(b),
+		     (u32)i_mce.misc, (u32)(i_mce.misc >> 32));
+
+	toggle_hw_mce_inject(cpu, false);
+
+	smp_call_function_single(cpu, trigger_mce, NULL, 0);
+
+err:
+	put_online_cpus();
+
+}
+
+/*
+ * This denotes into which bank we're injecting and triggers
+ * the injection, at the same time.
+ */
+static int inj_bank_set(void *data, u64 val)
+{
+	struct mce *m = (struct mce *)data;
+
+	if (val > 5) {
+		if (boot_cpu_data.x86 != 0x15 || val > 6) {
+			pr_err("Non-existent MCE bank: %llu\n", val);
+			return -EINVAL;
 		}
 	}
-	return 0;
 
-err_sysfs_create:
-	while (--i >= 0)
-		sysfs_remove_file(mce_kobj, &sysfs_attrs[i]->attr);
+	m->bank = val;
+	do_inject();
 
-	kobject_del(mce_kobj);
+	return 0;
+}
 
-err_mce_kobj:
-	edac_put_sysfs_subsys();
+static int inj_bank_get(void *data, u64 *val)
+{
+	struct mce *m = (struct mce *)data;
 
-	return err;
+	*val = m->bank;
+	return 0;
 }
 
-static void __exit edac_exit_mce_inject(void)
+DEFINE_SIMPLE_ATTRIBUTE(bank_fops, inj_bank_get, inj_bank_set, "%llu\n");
+
+struct dfs_node {
+	char *name;
+	struct dentry *d;
+	const struct file_operations *fops;
+} dfs_fls[] = {
+	{ .name = "status",	.fops = &status_fops },
+	{ .name = "misc",	.fops = &misc_fops },
+	{ .name = "addr",	.fops = &addr_fops },
+	{ .name = "bank",	.fops = &bank_fops },
+	{ .name = "flags",	.fops = &flags_fops },
+	{ .name = "cpu",	.fops = &extcpu_fops },
+};
+
+static int __init init_mce_inject(void)
 {
 	int i;
 
-	for (i = 0; i < ARRAY_SIZE(sysfs_attrs); i++)
-		sysfs_remove_file(mce_kobj, &sysfs_attrs[i]->attr);
+	dfs_inj = debugfs_create_dir("mce-inject", NULL);
+	if (!dfs_inj)
+		return -EINVAL;
+
+	for (i = 0; i < ARRAY_SIZE(dfs_fls); i++) {
+		dfs_fls[i].d = debugfs_create_file(dfs_fls[i].name,
+						    S_IRUSR | S_IWUSR,
+						    dfs_inj,
+						    &i_mce,
+						    dfs_fls[i].fops);
+
+		if (!dfs_fls[i].d)
+			goto err_dfs_add;
+	}
+
+	return 0;
+
+err_dfs_add:
+	while (--i >= 0)
+		debugfs_remove(dfs_fls[i].d);
 
-	kobject_del(mce_kobj);
+	debugfs_remove(dfs_inj);
+	dfs_inj = NULL;
 
-	edac_put_sysfs_subsys();
+	return -ENOMEM;
 }
 
-module_init(edac_init_mce_inject);
-module_exit(edac_exit_mce_inject);
+static void __exit exit_mce_inject(void)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(dfs_fls); i++)
+		debugfs_remove(dfs_fls[i].d);
+
+	memset(&dfs_fls, 0, sizeof(dfs_fls));
+
+	debugfs_remove(dfs_inj);
+	dfs_inj = NULL;
+}
+module_init(init_mce_inject);
+module_exit(exit_mce_inject);
 
 MODULE_LICENSE("GPL");
 MODULE_AUTHOR("Borislav Petkov <bp@alien8.de>");
 MODULE_AUTHOR("AMD Inc.");
-MODULE_DESCRIPTION("MCE injection facility for testing MCE decoding");
+MODULE_DESCRIPTION("MCE injection facility for RAS testing");
diff --git a/drivers/edac/mpc85xx_edac.c b/drivers/edac/mpc85xx_edac.c
index f4aec2e6ef56..ffb1a9a15ccd 100644
--- a/drivers/edac/mpc85xx_edac.c
+++ b/drivers/edac/mpc85xx_edac.c
@@ -633,7 +633,7 @@ static int mpc85xx_l2_err_probe(struct platform_device *op)
 	if (edac_op_state == EDAC_OPSTATE_INT) {
 		pdata->irq = irq_of_parse_and_map(op->dev.of_node, 0);
 		res = devm_request_irq(&op->dev, pdata->irq,
-				       mpc85xx_l2_isr, 0,
+				       mpc85xx_l2_isr, IRQF_SHARED,
 				       "[EDAC] L2 err", edac_dev);
 		if (res < 0) {
 			printk(KERN_ERR
@@ -715,7 +715,6 @@ static struct platform_driver mpc85xx_l2_err_driver = {
 	.remove = mpc85xx_l2_err_remove,
 	.driver = {
 		.name = "mpc85xx_l2_err",
-		.owner = THIS_MODULE,
 		.of_match_table = mpc85xx_l2_err_of_match,
 	},
 };
@@ -1215,7 +1214,6 @@ static struct platform_driver mpc85xx_mc_err_driver = {
 	.remove = mpc85xx_mc_err_remove,
 	.driver = {
 		.name = "mpc85xx_mc_err",
-		.owner = THIS_MODULE,
 		.of_match_table = mpc85xx_mc_err_of_match,
 	},
 };
diff --git a/drivers/edac/mv64x60_edac.c b/drivers/edac/mv64x60_edac.c
index 542fad70e360..6366e880f978 100644
--- a/drivers/edac/mv64x60_edac.c
+++ b/drivers/edac/mv64x60_edac.c
@@ -178,7 +178,7 @@ static int mv64x60_pci_err_probe(struct platform_device *pdev)
 		res = devm_request_irq(&pdev->dev,
 				       pdata->irq,
 				       mv64x60_pci_isr,
-				       IRQF_DISABLED,
+				       0,
 				       "[EDAC] PCI err",
 				       pci);
 		if (res < 0) {
@@ -345,7 +345,7 @@ static int mv64x60_sram_err_probe(struct platform_device *pdev)
 		res = devm_request_irq(&pdev->dev,
 				       pdata->irq,
 				       mv64x60_sram_isr,
-				       IRQF_DISABLED,
+				       0,
 				       "[EDAC] SRAM err",
 				       edac_dev);
 		if (res < 0) {
@@ -540,7 +540,7 @@ static int mv64x60_cpu_err_probe(struct platform_device *pdev)
 		res = devm_request_irq(&pdev->dev,
 				       pdata->irq,
 				       mv64x60_cpu_isr,
-				       IRQF_DISABLED,
+				       0,
 				       "[EDAC] CPU err",
 				       edac_dev);
 		if (res < 0) {
@@ -800,7 +800,7 @@ static int mv64x60_mc_err_probe(struct platform_device *pdev)
 		res = devm_request_irq(&pdev->dev,
 				       pdata->irq,
 				       mv64x60_mc_isr,
-				       IRQF_DISABLED,
+				       0,
 				       "[EDAC] MC err",
 				       mci);
 		if (res < 0) {
diff --git a/drivers/edac/ppc4xx_edac.c b/drivers/edac/ppc4xx_edac.c
index ef6b7e08f485..1b64fd060821 100644
--- a/drivers/edac/ppc4xx_edac.c
+++ b/drivers/edac/ppc4xx_edac.c
@@ -204,7 +204,6 @@ static struct platform_driver ppc4xx_edac_driver = {
 	.probe			= ppc4xx_edac_probe,
 	.remove			= ppc4xx_edac_remove,
 	.driver = {
-		.owner = THIS_MODULE,
 		.name = PPC4XX_EDAC_MODULE_NAME,
 		.of_match_table = ppc4xx_edac_match,
 	},
@@ -974,7 +973,7 @@ static int ppc4xx_edac_init_csrows(struct mem_ctl_info *mci, u32 mcopt1)
 		 * page size (PAGE_SIZE) or the memory width (2 or 4).
 		 */
 		for (j = 0; j < csi->nr_channels; j++) {
-			struct dimm_info *dimm = csi->channels[j].dimm;
+			struct dimm_info *dimm = csi->channels[j]->dimm;
 
 			dimm->nr_pages  = nr_pages / csi->nr_channels;
 			dimm->grain	= 1;
@@ -1120,7 +1119,7 @@ static int ppc4xx_edac_register_irq(struct platform_device *op,
 
 	status = request_irq(ded_irq,
 			     ppc4xx_edac_isr,
-			     IRQF_DISABLED,
+			     0,
 			     "[EDAC] MC ECCDED",
 			     mci);
 
@@ -1134,7 +1133,7 @@ static int ppc4xx_edac_register_irq(struct platform_device *op,
 
 	status = request_irq(sec_irq,
 			     ppc4xx_edac_isr,
-			     IRQF_DISABLED,
+			     0,
 			     "[EDAC] MC ECCSEC",
 			     mci);
 
diff --git a/drivers/edac/sb_edac.c b/drivers/edac/sb_edac.c
index 0034c4844428..63aa6730e89e 100644
--- a/drivers/edac/sb_edac.c
+++ b/drivers/edac/sb_edac.c
@@ -52,36 +52,6 @@ static int probed;
 #define GET_BITFIELD(v, lo, hi)	\
 	(((v) & GENMASK_ULL(hi, lo)) >> (lo))
 
-/*
- * sbridge Memory Controller Registers
- */
-
-/*
- * FIXME: For now, let's order by device function, as it makes
- * easier for driver's development process. This table should be
- * moved to pci_id.h when submitted upstream
- */
-#define PCI_DEVICE_ID_INTEL_SBRIDGE_SAD0	0x3cf4	/* 12.6 */
-#define PCI_DEVICE_ID_INTEL_SBRIDGE_SAD1	0x3cf6	/* 12.7 */
-#define PCI_DEVICE_ID_INTEL_SBRIDGE_BR		0x3cf5	/* 13.6 */
-#define PCI_DEVICE_ID_INTEL_SBRIDGE_IMC_HA0	0x3ca0	/* 14.0 */
-#define PCI_DEVICE_ID_INTEL_SBRIDGE_IMC_TA	0x3ca8	/* 15.0 */
-#define PCI_DEVICE_ID_INTEL_SBRIDGE_IMC_RAS	0x3c71	/* 15.1 */
-#define PCI_DEVICE_ID_INTEL_SBRIDGE_IMC_TAD0	0x3caa	/* 15.2 */
-#define PCI_DEVICE_ID_INTEL_SBRIDGE_IMC_TAD1	0x3cab	/* 15.3 */
-#define PCI_DEVICE_ID_INTEL_SBRIDGE_IMC_TAD2	0x3cac	/* 15.4 */
-#define PCI_DEVICE_ID_INTEL_SBRIDGE_IMC_TAD3	0x3cad	/* 15.5 */
-#define PCI_DEVICE_ID_INTEL_SBRIDGE_IMC_DDRIO	0x3cb8	/* 17.0 */
-
-	/*
-	 * Currently, unused, but will be needed in the future
-	 * implementations, as they hold the error counters
-	 */
-#define PCI_DEVICE_ID_INTEL_SBRIDGE_IMC_ERR0	0x3c72	/* 16.2 */
-#define PCI_DEVICE_ID_INTEL_SBRIDGE_IMC_ERR1	0x3c73	/* 16.3 */
-#define PCI_DEVICE_ID_INTEL_SBRIDGE_IMC_ERR2	0x3c76	/* 16.6 */
-#define PCI_DEVICE_ID_INTEL_SBRIDGE_IMC_ERR3	0x3c77	/* 16.7 */
-
 /* Devices 12 Function 6, Offsets 0x80 to 0xcc */
 static const u32 sbridge_dram_rule[] = {
 	0x80, 0x88, 0x90, 0x98, 0xa0,
@@ -165,6 +135,7 @@ static inline int sad_pkg(const struct interleave_pkg *table, u32 reg,
 
 #define TOLM		0x80
 #define	TOHM		0x84
+#define HASWELL_TOLM	0xd0
 #define HASWELL_TOHM_0	0xd4
 #define HASWELL_TOHM_1	0xd8
 
@@ -283,13 +254,15 @@ static const u32 correrrthrsld[] = {
  * sbridge structs
  */
 
-#define NUM_CHANNELS	4
-#define MAX_DIMMS	3		/* Max DIMMS per channel */
+#define NUM_CHANNELS		4
+#define MAX_DIMMS		3	/* Max DIMMS per channel */
+#define CHANNEL_UNSPECIFIED	0xf	/* Intel IA32 SDM 15-14 */
 
 enum type {
 	SANDY_BRIDGE,
 	IVY_BRIDGE,
 	HASWELL,
+	BROADWELL,
 };
 
 struct sbridge_pvt;
@@ -474,7 +447,7 @@ static const struct pci_id_table pci_dev_descr_ibridge_table[] = {
  *	- each SMI channel interfaces with a scalable memory buffer
  *	- each scalable memory buffer supports 4 DDR3/DDR4 channels, 3 DPC
  */
-#define HASWELL_DDRCRCLKCONTROLS 0xa10
+#define HASWELL_DDRCRCLKCONTROLS 0xa10 /* Ditto on Broadwell */
 #define HASWELL_HASYSDEFEATURE2 0x84
 #define PCI_DEVICE_ID_INTEL_HASWELL_IMC_VTD_MISC 0x2f28
 #define PCI_DEVICE_ID_INTEL_HASWELL_IMC_HA0	0x2fa0
@@ -526,12 +499,53 @@ static const struct pci_id_table pci_dev_descr_haswell_table[] = {
 };
 
 /*
+ * Broadwell support
+ *
+ * DE processor:
+ *	- 1 IMC
+ *	- 2 DDR3 channels, 2 DPC per channel
+ */
+#define PCI_DEVICE_ID_INTEL_BROADWELL_IMC_VTD_MISC 0x6f28
+#define PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0	0x6fa0
+#define PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_TA	0x6fa8
+#define PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_THERMAL 0x6f71
+#define PCI_DEVICE_ID_INTEL_BROADWELL_IMC_CBO_SAD0 0x6ffc
+#define PCI_DEVICE_ID_INTEL_BROADWELL_IMC_CBO_SAD1 0x6ffd
+#define PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_TAD0 0x6faa
+#define PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_TAD1 0x6fab
+#define PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_TAD2 0x6fac
+#define PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_TAD3 0x6fad
+#define PCI_DEVICE_ID_INTEL_BROADWELL_IMC_DDRIO0 0x6faf
+
+static const struct pci_id_descr pci_dev_descr_broadwell[] = {
+	/* first item must be the HA */
+	{ PCI_DESCR(PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0, 0)		},
+
+	{ PCI_DESCR(PCI_DEVICE_ID_INTEL_BROADWELL_IMC_CBO_SAD0, 0)	},
+	{ PCI_DESCR(PCI_DEVICE_ID_INTEL_BROADWELL_IMC_CBO_SAD1, 0)	},
+
+	{ PCI_DESCR(PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_TA, 0)	},
+	{ PCI_DESCR(PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_THERMAL, 0)	},
+	{ PCI_DESCR(PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_TAD0, 0)	},
+	{ PCI_DESCR(PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_TAD1, 0)	},
+	{ PCI_DESCR(PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_TAD2, 0)	},
+	{ PCI_DESCR(PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_TAD3, 0)	},
+	{ PCI_DESCR(PCI_DEVICE_ID_INTEL_BROADWELL_IMC_DDRIO0, 1)	},
+};
+
+static const struct pci_id_table pci_dev_descr_broadwell_table[] = {
+	PCI_ID_TABLE_ENTRY(pci_dev_descr_broadwell),
+	{0,}			/* 0 terminated list. */
+};
+
+/*
  *	pci_device_id	table for which devices we are looking for
  */
 static const struct pci_device_id sbridge_pci_tbl[] = {
-	{PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_SBRIDGE_IMC_TA)},
+	{PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_SBRIDGE_IMC_HA0)},
 	{PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_IBRIDGE_IMC_HA0_TA)},
 	{PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_HASWELL_IMC_HA0)},
+	{PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0)},
 	{0,}			/* 0 terminated list. */
 };
 
@@ -735,8 +749,8 @@ static u64 haswell_get_tolm(struct sbridge_pvt *pvt)
 {
 	u32 reg;
 
-	pci_read_config_dword(pvt->info.pci_vtd, TOLM, &reg);
-	return (GET_BITFIELD(reg, 26, 31) << 26) | 0x1ffffff;
+	pci_read_config_dword(pvt->info.pci_vtd, HASWELL_TOLM, &reg);
+	return (GET_BITFIELD(reg, 26, 31) << 26) | 0x3ffffff;
 }
 
 static u64 haswell_get_tohm(struct sbridge_pvt *pvt)
@@ -796,12 +810,22 @@ static int check_if_ecc_is_active(const u8 bus, enum type type)
 	struct pci_dev *pdev = NULL;
 	u32 mcmtr, id;
 
-	if (type == IVY_BRIDGE)
+	switch (type) {
+	case IVY_BRIDGE:
 		id = PCI_DEVICE_ID_INTEL_IBRIDGE_IMC_HA0_TA;
-	else if (type == HASWELL)
+		break;
+	case HASWELL:
 		id = PCI_DEVICE_ID_INTEL_HASWELL_IMC_HA0_TA;
-	else
+		break;
+	case SANDY_BRIDGE:
 		id = PCI_DEVICE_ID_INTEL_SBRIDGE_IMC_TA;
+		break;
+	case BROADWELL:
+		id = PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_TA;
+		break;
+	default:
+		return -ENODEV;
+	}
 
 	pdev = get_pdev_same_bus(bus, id);
 	if (!pdev) {
@@ -829,7 +853,7 @@ static int get_dimm_config(struct mem_ctl_info *mci)
 	enum edac_type mode;
 	enum mem_type mtype;
 
-	if (pvt->info.type == HASWELL)
+	if (pvt->info.type == HASWELL || pvt->info.type == BROADWELL)
 		pci_read_config_dword(pvt->pci_sad1, SAD_TARGET, &reg);
 	else
 		pci_read_config_dword(pvt->pci_br0, SAD_TARGET, &reg);
@@ -877,7 +901,7 @@ static int get_dimm_config(struct mem_ctl_info *mci)
 	else
 		edac_dbg(0, "Memory is unregistered\n");
 
-	if (mtype == MEM_DDR4 || MEM_RDDR4)
+	if (mtype == MEM_DDR4 || mtype == MEM_RDDR4)
 		banks = 16;
 	else
 		banks = 8;
@@ -938,7 +962,7 @@ static void get_memory_layout(const struct mem_ctl_info *mci)
 	u32 reg;
 	u64 limit, prv = 0;
 	u64 tmp_mb;
-	u32 mb, kb;
+	u32 gb, mb;
 	u32 rir_way;
 
 	/*
@@ -948,15 +972,17 @@ static void get_memory_layout(const struct mem_ctl_info *mci)
 	pvt->tolm = pvt->info.get_tolm(pvt);
 	tmp_mb = (1 + pvt->tolm) >> 20;
 
-	mb = div_u64_rem(tmp_mb, 1000, &kb);
-	edac_dbg(0, "TOLM: %u.%03u GB (0x%016Lx)\n", mb, kb, (u64)pvt->tolm);
+	gb = div_u64_rem(tmp_mb, 1024, &mb);
+	edac_dbg(0, "TOLM: %u.%03u GB (0x%016Lx)\n",
+		gb, (mb*1000)/1024, (u64)pvt->tolm);
 
 	/* Address range is already 45:25 */
 	pvt->tohm = pvt->info.get_tohm(pvt);
 	tmp_mb = (1 + pvt->tohm) >> 20;
 
-	mb = div_u64_rem(tmp_mb, 1000, &kb);
-	edac_dbg(0, "TOHM: %u.%03u GB (0x%016Lx)\n", mb, kb, (u64)pvt->tohm);
+	gb = div_u64_rem(tmp_mb, 1024, &mb);
+	edac_dbg(0, "TOHM: %u.%03u GB (0x%016Lx)\n",
+		gb, (mb*1000)/1024, (u64)pvt->tohm);
 
 	/*
 	 * Step 2) Get SAD range and SAD Interleave list
@@ -978,11 +1004,11 @@ static void get_memory_layout(const struct mem_ctl_info *mci)
 			break;
 
 		tmp_mb = (limit + 1) >> 20;
-		mb = div_u64_rem(tmp_mb, 1000, &kb);
+		gb = div_u64_rem(tmp_mb, 1024, &mb);
 		edac_dbg(0, "SAD#%d %s up to %u.%03u GB (0x%016Lx) Interleave: %s reg=0x%08x\n",
 			 n_sads,
 			 get_dram_attr(reg),
-			 mb, kb,
+			 gb, (mb*1000)/1024,
 			 ((u64)tmp_mb) << 20L,
 			 INTERLEAVE_MODE(reg) ? "8:6" : "[8:6]XOR[18:16]",
 			 reg);
@@ -1013,9 +1039,9 @@ static void get_memory_layout(const struct mem_ctl_info *mci)
 			break;
 		tmp_mb = (limit + 1) >> 20;
 
-		mb = div_u64_rem(tmp_mb, 1000, &kb);
+		gb = div_u64_rem(tmp_mb, 1024, &mb);
 		edac_dbg(0, "TAD#%d: up to %u.%03u GB (0x%016Lx), socket interleave %d, memory interleave %d, TGT: %d, %d, %d, %d, reg=0x%08x\n",
-			 n_tads, mb, kb,
+			 n_tads, gb, (mb*1000)/1024,
 			 ((u64)tmp_mb) << 20L,
 			 (u32)TAD_SOCK(reg),
 			 (u32)TAD_CH(reg),
@@ -1038,10 +1064,10 @@ static void get_memory_layout(const struct mem_ctl_info *mci)
 					      tad_ch_nilv_offset[j],
 					      &reg);
 			tmp_mb = TAD_OFFSET(reg) >> 20;
-			mb = div_u64_rem(tmp_mb, 1000, &kb);
+			gb = div_u64_rem(tmp_mb, 1024, &mb);
 			edac_dbg(0, "TAD CH#%d, offset #%d: %u.%03u GB (0x%016Lx), reg=0x%08x\n",
 				 i, j,
-				 mb, kb,
+				 gb, (mb*1000)/1024,
 				 ((u64)tmp_mb) << 20L,
 				 reg);
 		}
@@ -1063,10 +1089,10 @@ static void get_memory_layout(const struct mem_ctl_info *mci)
 
 			tmp_mb = pvt->info.rir_limit(reg) >> 20;
 			rir_way = 1 << RIR_WAY(reg);
-			mb = div_u64_rem(tmp_mb, 1000, &kb);
+			gb = div_u64_rem(tmp_mb, 1024, &mb);
 			edac_dbg(0, "CH#%d RIR#%d, limit: %u.%03u GB (0x%016Lx), way: %d, reg=0x%08x\n",
 				 i, j,
-				 mb, kb,
+				 gb, (mb*1000)/1024,
 				 ((u64)tmp_mb) << 20L,
 				 rir_way,
 				 reg);
@@ -1077,10 +1103,10 @@ static void get_memory_layout(const struct mem_ctl_info *mci)
 						      &reg);
 				tmp_mb = RIR_OFFSET(reg) << 6;
 
-				mb = div_u64_rem(tmp_mb, 1000, &kb);
+				gb = div_u64_rem(tmp_mb, 1024, &mb);
 				edac_dbg(0, "CH#%d RIR#%d INTL#%d, offset %u.%03u GB (0x%016Lx), tgt: %d, reg=0x%08x\n",
 					 i, j, k,
-					 mb, kb,
+					 gb, (mb*1000)/1024,
 					 ((u64)tmp_mb) << 20L,
 					 (u32)RIR_RNK_TGT(reg),
 					 reg);
@@ -1118,7 +1144,7 @@ static int get_memory_error_data(struct mem_ctl_info *mci,
 	u8			ch_way, sck_way, pkg, sad_ha = 0;
 	u32			tad_offset;
 	u32			rir_way;
-	u32			mb, kb;
+	u32			mb, gb;
 	u64			ch_addr, offset, limit = 0, prv = 0;
 
 
@@ -1208,7 +1234,7 @@ static int get_memory_error_data(struct mem_ctl_info *mci,
 		*socket = sad_interleave[idx];
 		edac_dbg(0, "SAD interleave index: %d (wayness %d) = CPU socket %d\n",
 			 idx, sad_way, *socket);
-	} else if (pvt->info.type == HASWELL) {
+	} else if (pvt->info.type == HASWELL || pvt->info.type == BROADWELL) {
 		int bits, a7mode = A7MODE(dram_rule);
 
 		if (a7mode) {
@@ -1387,10 +1413,10 @@ static int get_memory_error_data(struct mem_ctl_info *mci,
 			continue;
 
 		limit = pvt->info.rir_limit(reg);
-		mb = div_u64_rem(limit >> 20, 1000, &kb);
+		gb = div_u64_rem(limit >> 20, 1024, &mb);
 		edac_dbg(0, "RIR#%d, limit: %u.%03u GB (0x%016Lx), way: %d\n",
 			 n_rir,
-			 mb, kb,
+			 gb, (mb*1000)/1024,
 			 limit,
 			 1 << RIR_WAY(reg));
 		if  (ch_addr <= limit)
@@ -1857,6 +1883,82 @@ enodev:
 	return -ENODEV;
 }
 
+static int broadwell_mci_bind_devs(struct mem_ctl_info *mci,
+				 struct sbridge_dev *sbridge_dev)
+{
+	struct sbridge_pvt *pvt = mci->pvt_info;
+	struct pci_dev *pdev;
+	int i;
+
+	/* there's only one device per system; not tied to any bus */
+	if (pvt->info.pci_vtd == NULL)
+		/* result will be checked later */
+		pvt->info.pci_vtd = pci_get_device(PCI_VENDOR_ID_INTEL,
+						   PCI_DEVICE_ID_INTEL_BROADWELL_IMC_VTD_MISC,
+						   NULL);
+
+	for (i = 0; i < sbridge_dev->n_devs; i++) {
+		pdev = sbridge_dev->pdev[i];
+		if (!pdev)
+			continue;
+
+		switch (pdev->device) {
+		case PCI_DEVICE_ID_INTEL_BROADWELL_IMC_CBO_SAD0:
+			pvt->pci_sad0 = pdev;
+			break;
+		case PCI_DEVICE_ID_INTEL_BROADWELL_IMC_CBO_SAD1:
+			pvt->pci_sad1 = pdev;
+			break;
+		case PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0:
+			pvt->pci_ha0 = pdev;
+			break;
+		case PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_TA:
+			pvt->pci_ta = pdev;
+			break;
+		case PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_THERMAL:
+			pvt->pci_ras = pdev;
+			break;
+		case PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_TAD0:
+			pvt->pci_tad[0] = pdev;
+			break;
+		case PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_TAD1:
+			pvt->pci_tad[1] = pdev;
+			break;
+		case PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_TAD2:
+			pvt->pci_tad[2] = pdev;
+			break;
+		case PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0_TAD3:
+			pvt->pci_tad[3] = pdev;
+			break;
+		case PCI_DEVICE_ID_INTEL_BROADWELL_IMC_DDRIO0:
+			pvt->pci_ddrio = pdev;
+			break;
+		default:
+			break;
+		}
+
+		edac_dbg(0, "Associated PCI %02x.%02d.%d with dev = %p\n",
+			 sbridge_dev->bus,
+			 PCI_SLOT(pdev->devfn), PCI_FUNC(pdev->devfn),
+			 pdev);
+	}
+
+	/* Check if everything were registered */
+	if (!pvt->pci_sad0 || !pvt->pci_ha0 || !pvt->pci_sad1 ||
+	    !pvt->pci_ras  || !pvt->pci_ta || !pvt->info.pci_vtd)
+		goto enodev;
+
+	for (i = 0; i < NUM_CHANNELS; i++) {
+		if (!pvt->pci_tad[i])
+			goto enodev;
+	}
+	return 0;
+
+enodev:
+	sbridge_printk(KERN_ERR, "Some needed devices are missing\n");
+	return -ENODEV;
+}
+
 /****************************************************************************
 			Error check routines
  ****************************************************************************/
@@ -1991,6 +2093,9 @@ static void sbridge_mce_output_error(struct mem_ctl_info *mci,
 
 	/* FIXME: need support for channel mask */
 
+	if (channel == CHANNEL_UNSPECIFIED)
+		channel = -1;
+
 	/* Call the helper to output message */
 	edac_mc_handle_error(tp_event, mci, core_err_cnt,
 			     m->addr >> PAGE_SHIFT, m->addr & ~PAGE_MASK, 0,
@@ -2266,6 +2371,25 @@ static int sbridge_register_mci(struct sbridge_dev *sbridge_dev, enum type type)
 		if (unlikely(rc < 0))
 			goto fail0;
 		break;
+	case BROADWELL:
+		/* rankcfgr isn't used */
+		pvt->info.get_tolm = haswell_get_tolm;
+		pvt->info.get_tohm = haswell_get_tohm;
+		pvt->info.dram_rule = ibridge_dram_rule;
+		pvt->info.get_memory_type = haswell_get_memory_type;
+		pvt->info.get_node_id = haswell_get_node_id;
+		pvt->info.rir_limit = haswell_rir_limit;
+		pvt->info.max_sad = ARRAY_SIZE(ibridge_dram_rule);
+		pvt->info.interleave_list = ibridge_interleave_list;
+		pvt->info.max_interleave = ARRAY_SIZE(ibridge_interleave_list);
+		pvt->info.interleave_pkg = ibridge_interleave_pkg;
+		mci->ctl_name = kasprintf(GFP_KERNEL, "Broadwell Socket#%d", mci->mc_idx);
+
+		/* Store pci devices at mci for faster access */
+		rc = broadwell_mci_bind_devs(mci, sbridge_dev);
+		if (unlikely(rc < 0))
+			goto fail0;
+		break;
 	}
 
 	/* Get dimm basic config and the memory layout */
@@ -2331,6 +2455,10 @@ static int sbridge_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 		rc = sbridge_get_all_devices(&num_mc, pci_dev_descr_haswell_table);
 		type = HASWELL;
 		break;
+	case PCI_DEVICE_ID_INTEL_BROADWELL_IMC_HA0:
+		rc = sbridge_get_all_devices(&num_mc, pci_dev_descr_broadwell_table);
+		type = BROADWELL;
+		break;
 	}
 	if (unlikely(rc < 0))
 		goto fail0;
diff --git a/drivers/edac/tile_edac.c b/drivers/edac/tile_edac.c
index 578f915ee195..71381642ce2a 100644
--- a/drivers/edac/tile_edac.c
+++ b/drivers/edac/tile_edac.c
@@ -199,7 +199,6 @@ static int tile_edac_mc_remove(struct platform_device *pdev)
 static struct platform_driver tile_edac_mc_driver = {
 	.driver		= {
 		.name	= DRV_NAME,
-		.owner	= THIS_MODULE,
 	},
 	.probe		= tile_edac_mc_probe,
 	.remove		= tile_edac_mc_remove,
diff --git a/drivers/edac/x38_edac.c b/drivers/edac/x38_edac.c
index e644b52c287c..7c5cdc62f31c 100644
--- a/drivers/edac/x38_edac.c
+++ b/drivers/edac/x38_edac.c
@@ -500,8 +500,7 @@ fail1:
 	pci_unregister_driver(&x38_driver);
 
 fail0:
-	if (mci_pdev)
-		pci_dev_put(mci_pdev);
+	pci_dev_put(mci_pdev);
 
 	return pci_rc;
 }
diff --git a/drivers/extcon/Kconfig b/drivers/extcon/Kconfig
index 6f2f4727de2c..6a1f7de6fa54 100644
--- a/drivers/extcon/Kconfig
+++ b/drivers/extcon/Kconfig
@@ -70,8 +70,21 @@ config EXTCON_PALMAS
 	  Say Y here to enable support for USB peripheral and USB host
 	  detection by palmas usb.
 
+config EXTCON_RT8973A
+	tristate "RT8973A EXTCON support"
+	depends on I2C
+	select IRQ_DOMAIN
+	select REGMAP_I2C
+	select REGMAP_IRQ
+	help
+	  If you say yes here you get support for the MUIC device of
+	  Richtek RT8973A. The RT8973A is a USB port accessory detector
+	  and switch that is optimized to protect low voltage system
+	  from abnormal high input voltage (up to 28V).
+
 config EXTCON_SM5502
 	tristate "SM5502 EXTCON support"
+	depends on I2C
 	select IRQ_DOMAIN
 	select REGMAP_I2C
 	select REGMAP_IRQ
diff --git a/drivers/extcon/Makefile b/drivers/extcon/Makefile
index b38546eb522a..0370b42e5a27 100644
--- a/drivers/extcon/Makefile
+++ b/drivers/extcon/Makefile
@@ -10,4 +10,5 @@ obj-$(CONFIG_EXTCON_MAX14577)	+= extcon-max14577.o
 obj-$(CONFIG_EXTCON_MAX77693)	+= extcon-max77693.o
 obj-$(CONFIG_EXTCON_MAX8997)	+= extcon-max8997.o
 obj-$(CONFIG_EXTCON_PALMAS)	+= extcon-palmas.o
+obj-$(CONFIG_EXTCON_RT8973A)	+= extcon-rt8973a.o
 obj-$(CONFIG_EXTCON_SM5502)	+= extcon-sm5502.o
diff --git a/drivers/extcon/extcon-adc-jack.c b/drivers/extcon/extcon-adc-jack.c
index d860229e4de1..5d7ab577fba9 100644
--- a/drivers/extcon/extcon-adc-jack.c
+++ b/drivers/extcon/extcon-adc-jack.c
@@ -182,7 +182,6 @@ static struct platform_driver adc_jack_driver = {
 	.remove         = adc_jack_remove,
 	.driver         = {
 		.name   = "adc-jack",
-		.owner  = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/extcon/extcon-arizona.c b/drivers/extcon/extcon-arizona.c
index ba51588cc000..63f01c42aed4 100644
--- a/drivers/extcon/extcon-arizona.c
+++ b/drivers/extcon/extcon-arizona.c
@@ -1469,7 +1469,6 @@ static int arizona_extcon_remove(struct platform_device *pdev)
 static struct platform_driver arizona_extcon_driver = {
 	.driver		= {
 		.name	= "arizona-extcon",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= arizona_extcon_probe,
 	.remove		= arizona_extcon_remove,
diff --git a/drivers/extcon/extcon-class.c b/drivers/extcon/extcon-class.c
index 4c2f2c543bb7..043dcd9946c9 100644
--- a/drivers/extcon/extcon-class.c
+++ b/drivers/extcon/extcon-class.c
@@ -29,6 +29,7 @@
 #include <linux/fs.h>
 #include <linux/err.h>
 #include <linux/extcon.h>
+#include <linux/of.h>
 #include <linux/slab.h>
 #include <linux/sysfs.h>
 #include <linux/of.h>
@@ -997,13 +998,16 @@ struct extcon_dev *extcon_get_edev_by_phandle(struct device *dev, int index)
 		return ERR_PTR(-ENODEV);
 	}
 
-	edev = extcon_get_extcon_dev(node->name);
-	if (!edev) {
-		dev_err(dev, "unable to get extcon device : %s\n", node->name);
-		return ERR_PTR(-ENODEV);
+	mutex_lock(&extcon_dev_list_lock);
+	list_for_each_entry(edev, &extcon_dev_list, entry) {
+		if (edev->dev.parent && edev->dev.parent->of_node == node) {
+			mutex_unlock(&extcon_dev_list_lock);
+			return edev;
+		}
 	}
+	mutex_unlock(&extcon_dev_list_lock);
 
-	return edev;
+	return ERR_PTR(-EPROBE_DEFER);
 }
 #else
 struct extcon_dev *extcon_get_edev_by_phandle(struct device *dev, int index)
diff --git a/drivers/extcon/extcon-gpio.c b/drivers/extcon/extcon-gpio.c
index 5b7ec274cb63..7af33fc433cd 100644
--- a/drivers/extcon/extcon-gpio.c
+++ b/drivers/extcon/extcon-gpio.c
@@ -20,16 +20,16 @@
  *
 */
 
-#include <linux/module.h>
-#include <linux/kernel.h>
+#include <linux/extcon.h>
+#include <linux/extcon/extcon-gpio.h>
+#include <linux/gpio.h>
 #include <linux/init.h>
 #include <linux/interrupt.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
 #include <linux/platform_device.h>
 #include <linux/slab.h>
 #include <linux/workqueue.h>
-#include <linux/gpio.h>
-#include <linux/extcon.h>
-#include <linux/extcon/extcon-gpio.h>
 
 struct gpio_extcon_data {
 	struct extcon_dev *edev;
@@ -181,7 +181,6 @@ static struct platform_driver gpio_extcon_driver = {
 	.remove		= gpio_extcon_remove,
 	.driver		= {
 		.name	= "extcon-gpio",
-		.owner	= THIS_MODULE,
 		.pm	= &gpio_extcon_pm_ops,
 	},
 };
diff --git a/drivers/extcon/extcon-max14577.c b/drivers/extcon/extcon-max14577.c
index 7309743d0da1..c1bf0cf747b0 100644
--- a/drivers/extcon/extcon-max14577.c
+++ b/drivers/extcon/extcon-max14577.c
@@ -1,7 +1,7 @@
 /*
  * extcon-max14577.c - MAX14577/77836 extcon driver to support MUIC
  *
- * Copyright (C) 2013,2014 Samsung Electrnoics
+ * Copyright (C) 2013,2014 Samsung Electronics
  * Chanwoo Choi <cw00.choi@samsung.com>
  * Krzysztof Kozlowski <k.kozlowski@samsung.com>
  *
@@ -807,7 +807,6 @@ MODULE_DEVICE_TABLE(platform, max14577_muic_id);
 static struct platform_driver max14577_muic_driver = {
 	.driver		= {
 		.name	= "max14577-muic",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= max14577_muic_probe,
 	.remove		= max14577_muic_remove,
diff --git a/drivers/extcon/extcon-max77693.c b/drivers/extcon/extcon-max77693.c
index 77460f2c1ca1..740a14d35072 100644
--- a/drivers/extcon/extcon-max77693.c
+++ b/drivers/extcon/extcon-max77693.c
@@ -232,7 +232,7 @@ static const char *max77693_extcon_cable[] = {
 	[EXTCON_CABLE_JIG_USB_ON]		= "JIG-USB-ON",
 	[EXTCON_CABLE_JIG_USB_OFF]		= "JIG-USB-OFF",
 	[EXTCON_CABLE_JIG_UART_OFF]		= "JIG-UART-OFF",
-	[EXTCON_CABLE_JIG_UART_ON]		= "Dock-Car",
+	[EXTCON_CABLE_JIG_UART_ON]		= "JIG-UART-ON",
 	[EXTCON_CABLE_DOCK_SMART]		= "Dock-Smart",
 	[EXTCON_CABLE_DOCK_DESK]		= "Dock-Desk",
 	[EXTCON_CABLE_DOCK_AUDIO]		= "Dock-Audio",
@@ -255,10 +255,14 @@ static int max77693_muic_set_debounce_time(struct max77693_muic_info *info,
 	case ADC_DEBOUNCE_TIME_10MS:
 	case ADC_DEBOUNCE_TIME_25MS:
 	case ADC_DEBOUNCE_TIME_38_62MS:
-		ret = regmap_update_bits(info->max77693->regmap_muic,
-					  MAX77693_MUIC_REG_CTRL3,
-					  CONTROL3_ADCDBSET_MASK,
-					  time << CONTROL3_ADCDBSET_SHIFT);
+		/*
+		 * Don't touch BTLDset, JIGset when you want to change adc
+		 * debounce time. If it writes other than 0 to BTLDset, JIGset
+		 * muic device will be reset and loose current state.
+		 */
+		ret = regmap_write(info->max77693->regmap_muic,
+				  MAX77693_MUIC_REG_CTRL3,
+				  time << CONTROL3_ADCDBSET_SHIFT);
 		if (ret) {
 			dev_err(info->dev, "failed to set ADC debounce time\n");
 			return ret;
@@ -528,9 +532,6 @@ static int max77693_muic_dock_handler(struct max77693_muic_info *info,
 		extcon_set_cable_state(info->edev, "Dock-Smart", attached);
 		extcon_set_cable_state(info->edev, "MHL", attached);
 		goto out;
-	case MAX77693_MUIC_ADC_FACTORY_MODE_UART_ON:	/* Dock-Car */
-		strcpy(dock_name, "Dock-Car");
-		break;
 	case MAX77693_MUIC_ADC_AUDIO_MODE_REMOTE:	/* Dock-Desk */
 		strcpy(dock_name, "Dock-Desk");
 		break;
@@ -665,6 +666,11 @@ static int max77693_muic_jig_handler(struct max77693_muic_info *info,
 		strcpy(cable_name, "JIG-UART-OFF");
 		path = CONTROL1_SW_UART;
 		break;
+	case MAX77693_MUIC_ADC_FACTORY_MODE_UART_ON:	/* ADC_JIG_UART_ON */
+		/* PATH:AP_UART */
+		strcpy(cable_name, "JIG-UART-ON");
+		path = CONTROL1_SW_UART;
+		break;
 	default:
 		dev_err(info->dev, "failed to detect %s jig cable\n",
 			attached ? "attached" : "detached");
@@ -704,13 +710,13 @@ static int max77693_muic_adc_handler(struct max77693_muic_info *info)
 	case MAX77693_MUIC_ADC_FACTORY_MODE_USB_OFF:
 	case MAX77693_MUIC_ADC_FACTORY_MODE_USB_ON:
 	case MAX77693_MUIC_ADC_FACTORY_MODE_UART_OFF:
+	case MAX77693_MUIC_ADC_FACTORY_MODE_UART_ON:
 		/* JIG */
 		ret = max77693_muic_jig_handler(info, cable_type, attached);
 		if (ret < 0)
 			return ret;
 		break;
 	case MAX77693_MUIC_ADC_RESERVED_ACC_3:		/* Dock-Smart */
-	case MAX77693_MUIC_ADC_FACTORY_MODE_UART_ON:	/* Dock-Car */
 	case MAX77693_MUIC_ADC_AUDIO_MODE_REMOTE:	/* Dock-Desk */
 	case MAX77693_MUIC_ADC_AV_CABLE_NOLOAD:		/* Dock-Audio */
 		/*
@@ -1155,13 +1161,11 @@ static int max77693_muic_probe(struct platform_device *pdev)
 
 		virq = regmap_irq_get_virq(max77693->irq_data_muic,
 					muic_irq->irq);
-		if (!virq) {
-			ret = -EINVAL;
-			goto err_irq;
-		}
+		if (!virq)
+			return -EINVAL;
 		muic_irq->virq = virq;
 
-		ret = request_threaded_irq(virq, NULL,
+		ret = devm_request_threaded_irq(&pdev->dev, virq, NULL,
 				max77693_muic_irq_handler,
 				IRQF_NO_SUSPEND,
 				muic_irq->name, info);
@@ -1170,7 +1174,7 @@ static int max77693_muic_probe(struct platform_device *pdev)
 				"failed: irq request (IRQ: %d,"
 				" error :%d)\n",
 				muic_irq->irq, ret);
-			goto err_irq;
+			return ret;
 		}
 	}
 
@@ -1179,15 +1183,14 @@ static int max77693_muic_probe(struct platform_device *pdev)
 					      max77693_extcon_cable);
 	if (IS_ERR(info->edev)) {
 		dev_err(&pdev->dev, "failed to allocate memory for extcon\n");
-		ret = -ENOMEM;
-		goto err_irq;
+		return -ENOMEM;
 	}
 	info->edev->name = DEV_NAME;
 
 	ret = devm_extcon_dev_register(&pdev->dev, info->edev);
 	if (ret) {
 		dev_err(&pdev->dev, "failed to register extcon device\n");
-		goto err_irq;
+		return ret;
 	}
 
 	/* Initialize MUIC register by using platform data or default data */
@@ -1265,7 +1268,7 @@ static int max77693_muic_probe(struct platform_device *pdev)
 			MAX77693_MUIC_REG_ID, &id);
 	if (ret < 0) {
 		dev_err(&pdev->dev, "failed to read revision number\n");
-		goto err_irq;
+		return ret;
 	}
 	dev_info(info->dev, "device ID : 0x%x\n", id);
 
@@ -1285,20 +1288,12 @@ static int max77693_muic_probe(struct platform_device *pdev)
 			delay_jiffies);
 
 	return ret;
-
-err_irq:
-	while (--i >= 0)
-		free_irq(muic_irqs[i].virq, info);
-	return ret;
 }
 
 static int max77693_muic_remove(struct platform_device *pdev)
 {
 	struct max77693_muic_info *info = platform_get_drvdata(pdev);
-	int i;
 
-	for (i = 0; i < ARRAY_SIZE(muic_irqs); i++)
-		free_irq(muic_irqs[i].virq, info);
 	cancel_work_sync(&info->irq_work);
 	input_unregister_device(info->dock);
 
@@ -1308,7 +1303,6 @@ static int max77693_muic_remove(struct platform_device *pdev)
 static struct platform_driver max77693_muic_driver = {
 	.driver		= {
 		.name	= DEV_NAME,
-		.owner	= THIS_MODULE,
 	},
 	.probe		= max77693_muic_probe,
 	.remove		= max77693_muic_remove,
diff --git a/drivers/extcon/extcon-max8997.c b/drivers/extcon/extcon-max8997.c
index 75e501c98005..fc1678fa95c4 100644
--- a/drivers/extcon/extcon-max8997.c
+++ b/drivers/extcon/extcon-max8997.c
@@ -792,7 +792,6 @@ static int max8997_muic_remove(struct platform_device *pdev)
 static struct platform_driver max8997_muic_driver = {
 	.driver		= {
 		.name	= DEV_NAME,
-		.owner	= THIS_MODULE,
 	},
 	.probe		= max8997_muic_probe,
 	.remove		= max8997_muic_remove,
diff --git a/drivers/extcon/extcon-palmas.c b/drivers/extcon/extcon-palmas.c
index 230e1220ce48..11c6757b6c40 100644
--- a/drivers/extcon/extcon-palmas.c
+++ b/drivers/extcon/extcon-palmas.c
@@ -291,7 +291,6 @@ static struct platform_driver palmas_usb_driver = {
 	.driver = {
 		.name = "palmas-usb",
 		.of_match_table = of_palmas_match_tbl,
-		.owner = THIS_MODULE,
 		.pm = &palmas_pm_ops,
 	},
 };
diff --git a/drivers/extcon/extcon-rt8973a.c b/drivers/extcon/extcon-rt8973a.c
new file mode 100644
index 000000000000..a784b2d5ee72
--- /dev/null
+++ b/drivers/extcon/extcon-rt8973a.c
@@ -0,0 +1,740 @@
+/*
+ * extcon-rt8973a.c - Richtek RT8973A extcon driver to support USB switches
+ *
+ * Copyright (c) 2014 Samsung Electronics Co., Ltd
+ * Author: Chanwoo Choi <cw00.choi@samsung.com>
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+
+#include <linux/err.h>
+#include <linux/i2c.h>
+#include <linux/input.h>
+#include <linux/interrupt.h>
+#include <linux/irqdomain.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/regmap.h>
+#include <linux/slab.h>
+#include <linux/extcon.h>
+
+#include "extcon-rt8973a.h"
+
+#define	DELAY_MS_DEFAULT		20000	/* unit: millisecond */
+
+struct muic_irq {
+	unsigned int irq;
+	const char *name;
+	unsigned int virq;
+};
+
+struct reg_data {
+	u8 reg;
+	u8 mask;
+	u8 val;
+	bool invert;
+};
+
+struct rt8973a_muic_info {
+	struct device *dev;
+	struct extcon_dev *edev;
+
+	struct i2c_client *i2c;
+	struct regmap *regmap;
+
+	struct regmap_irq_chip_data *irq_data;
+	struct muic_irq *muic_irqs;
+	unsigned int num_muic_irqs;
+	int irq;
+	bool irq_attach;
+	bool irq_detach;
+	bool irq_ovp;
+	bool irq_otp;
+	struct work_struct irq_work;
+
+	struct reg_data *reg_data;
+	unsigned int num_reg_data;
+	bool auto_config;
+
+	struct mutex mutex;
+
+	/*
+	 * Use delayed workqueue to detect cable state and then
+	 * notify cable state to notifiee/platform through uevent.
+	 * After completing the booting of platform, the extcon provider
+	 * driver should notify cable state to upper layer.
+	 */
+	struct delayed_work wq_detcable;
+};
+
+/* Default value of RT8973A register to bring up MUIC device. */
+static struct reg_data rt8973a_reg_data[] = {
+	{
+		.reg = RT8973A_REG_CONTROL1,
+		.mask = RT8973A_REG_CONTROL1_ADC_EN_MASK
+			| RT8973A_REG_CONTROL1_USB_CHD_EN_MASK
+			| RT8973A_REG_CONTROL1_CHGTYP_MASK
+			| RT8973A_REG_CONTROL1_SWITCH_OPEN_MASK
+			| RT8973A_REG_CONTROL1_AUTO_CONFIG_MASK
+			| RT8973A_REG_CONTROL1_INTM_MASK,
+		.val = RT8973A_REG_CONTROL1_ADC_EN_MASK
+			| RT8973A_REG_CONTROL1_USB_CHD_EN_MASK
+			| RT8973A_REG_CONTROL1_CHGTYP_MASK,
+		.invert = false,
+	},
+	{ /* sentinel */ }
+};
+
+/* List of detectable cables */
+enum {
+	EXTCON_CABLE_USB = 0,
+	EXTCON_CABLE_USB_HOST,
+	EXTCON_CABLE_TA,
+	EXTCON_CABLE_JIG_OFF_USB,
+	EXTCON_CABLE_JIG_ON_USB,
+	EXTCON_CABLE_JIG_OFF_UART,
+	EXTCON_CABLE_JIG_ON_UART,
+
+	EXTCON_CABLE_END,
+};
+
+static const char *rt8973a_extcon_cable[] = {
+	[EXTCON_CABLE_USB]		= "USB",
+	[EXTCON_CABLE_USB_HOST]		= "USB-Host",
+	[EXTCON_CABLE_TA]		= "TA",
+	[EXTCON_CABLE_JIG_OFF_USB]	= "JIG-USB-OFF",
+	[EXTCON_CABLE_JIG_ON_USB]	= "JIG-USB-ON",
+	[EXTCON_CABLE_JIG_OFF_UART]	= "JIG-UART-OFF",
+	[EXTCON_CABLE_JIG_ON_UART]	= "JIG-UART-ON",
+	NULL,
+};
+
+/* Define OVP (Over Voltage Protection), OTP (Over Temperature Protection) */
+enum rt8973a_event_type {
+	RT8973A_EVENT_ATTACH = 1,
+	RT8973A_EVENT_DETACH,
+	RT8973A_EVENT_OVP,
+	RT8973A_EVENT_OTP,
+};
+
+/* Define supported accessory type */
+enum rt8973a_muic_acc_type {
+	RT8973A_MUIC_ADC_OTG = 0x0,
+	RT8973A_MUIC_ADC_AUDIO_SEND_END_BUTTON,
+	RT8973A_MUIC_ADC_AUDIO_REMOTE_S1_BUTTON,
+	RT8973A_MUIC_ADC_AUDIO_REMOTE_S2_BUTTON,
+	RT8973A_MUIC_ADC_AUDIO_REMOTE_S3_BUTTON,
+	RT8973A_MUIC_ADC_AUDIO_REMOTE_S4_BUTTON,
+	RT8973A_MUIC_ADC_AUDIO_REMOTE_S5_BUTTON,
+	RT8973A_MUIC_ADC_AUDIO_REMOTE_S6_BUTTON,
+	RT8973A_MUIC_ADC_AUDIO_REMOTE_S7_BUTTON,
+	RT8973A_MUIC_ADC_AUDIO_REMOTE_S8_BUTTON,
+	RT8973A_MUIC_ADC_AUDIO_REMOTE_S9_BUTTON,
+	RT8973A_MUIC_ADC_AUDIO_REMOTE_S10_BUTTON,
+	RT8973A_MUIC_ADC_AUDIO_REMOTE_S11_BUTTON,
+	RT8973A_MUIC_ADC_AUDIO_REMOTE_S12_BUTTON,
+	RT8973A_MUIC_ADC_RESERVED_ACC_1,
+	RT8973A_MUIC_ADC_RESERVED_ACC_2,
+	RT8973A_MUIC_ADC_RESERVED_ACC_3,
+	RT8973A_MUIC_ADC_RESERVED_ACC_4,
+	RT8973A_MUIC_ADC_RESERVED_ACC_5,
+	RT8973A_MUIC_ADC_AUDIO_TYPE2,
+	RT8973A_MUIC_ADC_PHONE_POWERED_DEV,
+	RT8973A_MUIC_ADC_UNKNOWN_ACC_1,
+	RT8973A_MUIC_ADC_UNKNOWN_ACC_2,
+	RT8973A_MUIC_ADC_TA,
+	RT8973A_MUIC_ADC_FACTORY_MODE_BOOT_OFF_USB,
+	RT8973A_MUIC_ADC_FACTORY_MODE_BOOT_ON_USB,
+	RT8973A_MUIC_ADC_UNKNOWN_ACC_3,
+	RT8973A_MUIC_ADC_UNKNOWN_ACC_4,
+	RT8973A_MUIC_ADC_FACTORY_MODE_BOOT_OFF_UART,
+	RT8973A_MUIC_ADC_FACTORY_MODE_BOOT_ON_UART,
+	RT8973A_MUIC_ADC_UNKNOWN_ACC_5,
+	RT8973A_MUIC_ADC_OPEN = 0x1f,
+
+	/* The below accessories has same ADC value (0x1f).
+	   So, Device type1 is used to separate specific accessory. */
+					/* |---------|--ADC| */
+					/* |    [7:5]|[4:0]| */
+	RT8973A_MUIC_ADC_USB = 0x3f,	/* |      001|11111| */
+};
+
+/* List of supported interrupt for RT8973A */
+static struct muic_irq rt8973a_muic_irqs[] = {
+	{ RT8973A_INT1_ATTACH,		"muic-attach" },
+	{ RT8973A_INT1_DETACH,		"muic-detach" },
+	{ RT8973A_INT1_CHGDET,		"muic-chgdet" },
+	{ RT8973A_INT1_DCD_T,		"muic-dcd-t" },
+	{ RT8973A_INT1_OVP,		"muic-ovp" },
+	{ RT8973A_INT1_CONNECT,		"muic-connect" },
+	{ RT8973A_INT1_ADC_CHG,		"muic-adc-chg" },
+	{ RT8973A_INT1_OTP,		"muic-otp" },
+	{ RT8973A_INT2_UVLO,		"muic-uvlo" },
+	{ RT8973A_INT2_POR,		"muic-por" },
+	{ RT8973A_INT2_OTP_FET,		"muic-otp-fet" },
+	{ RT8973A_INT2_OVP_FET,		"muic-ovp-fet" },
+	{ RT8973A_INT2_OCP_LATCH,	"muic-ocp-latch" },
+	{ RT8973A_INT2_OCP,		"muic-ocp" },
+	{ RT8973A_INT2_OVP_OCP,		"muic-ovp-ocp" },
+};
+
+/* Define interrupt list of RT8973A to register regmap_irq */
+static const struct regmap_irq rt8973a_irqs[] = {
+	/* INT1 interrupts */
+	{ .reg_offset = 0, .mask = RT8973A_INT1_ATTACH_MASK, },
+	{ .reg_offset = 0, .mask = RT8973A_INT1_DETACH_MASK, },
+	{ .reg_offset = 0, .mask = RT8973A_INT1_CHGDET_MASK, },
+	{ .reg_offset = 0, .mask = RT8973A_INT1_DCD_T_MASK, },
+	{ .reg_offset = 0, .mask = RT8973A_INT1_OVP_MASK, },
+	{ .reg_offset = 0, .mask = RT8973A_INT1_CONNECT_MASK, },
+	{ .reg_offset = 0, .mask = RT8973A_INT1_ADC_CHG_MASK, },
+	{ .reg_offset = 0, .mask = RT8973A_INT1_OTP_MASK, },
+
+	/* INT2 interrupts */
+	{ .reg_offset = 1, .mask = RT8973A_INT2_UVLOT_MASK,},
+	{ .reg_offset = 1, .mask = RT8973A_INT2_POR_MASK, },
+	{ .reg_offset = 1, .mask = RT8973A_INT2_OTP_FET_MASK, },
+	{ .reg_offset = 1, .mask = RT8973A_INT2_OVP_FET_MASK, },
+	{ .reg_offset = 1, .mask = RT8973A_INT2_OCP_LATCH_MASK, },
+	{ .reg_offset = 1, .mask = RT8973A_INT2_OCP_MASK, },
+	{ .reg_offset = 1, .mask = RT8973A_INT2_OVP_OCP_MASK, },
+};
+
+static const struct regmap_irq_chip rt8973a_muic_irq_chip = {
+	.name			= "rt8973a",
+	.status_base		= RT8973A_REG_INT1,
+	.mask_base		= RT8973A_REG_INTM1,
+	.mask_invert		= false,
+	.num_regs		= 2,
+	.irqs			= rt8973a_irqs,
+	.num_irqs		= ARRAY_SIZE(rt8973a_irqs),
+};
+
+/* Define regmap configuration of RT8973A for I2C communication  */
+static bool rt8973a_muic_volatile_reg(struct device *dev, unsigned int reg)
+{
+	switch (reg) {
+	case RT8973A_REG_INTM1:
+	case RT8973A_REG_INTM2:
+		return true;
+	default:
+		break;
+	}
+	return false;
+}
+
+static const struct regmap_config rt8973a_muic_regmap_config = {
+	.reg_bits	= 8,
+	.val_bits	= 8,
+	.volatile_reg	= rt8973a_muic_volatile_reg,
+	.max_register	= RT8973A_REG_END,
+};
+
+/* Change DM_CON/DP_CON/VBUSIN switch according to cable type */
+static int rt8973a_muic_set_path(struct rt8973a_muic_info *info,
+				unsigned int con_sw, bool attached)
+{
+	int ret;
+
+	/*
+	 * Don't need to set h/w path according to cable type
+	 * if Auto-configuration mode of CONTROL1 register is true.
+	 */
+	if (info->auto_config)
+		return 0;
+
+	if (!attached)
+		con_sw	= DM_DP_SWITCH_UART;
+
+	switch (con_sw) {
+	case DM_DP_SWITCH_OPEN:
+	case DM_DP_SWITCH_USB:
+	case DM_DP_SWITCH_UART:
+		ret = regmap_update_bits(info->regmap, RT8973A_REG_MANUAL_SW1,
+					RT8973A_REG_MANUAL_SW1_DP_MASK |
+					RT8973A_REG_MANUAL_SW1_DM_MASK,
+					con_sw);
+		if (ret < 0) {
+			dev_err(info->dev,
+				"cannot update DM_CON/DP_CON switch\n");
+			return ret;
+		}
+		break;
+	default:
+		dev_err(info->dev, "Unknown DM_CON/DP_CON switch type (%d)\n",
+				con_sw);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int rt8973a_muic_get_cable_type(struct rt8973a_muic_info *info)
+{
+	unsigned int adc, dev1;
+	int ret, cable_type;
+
+	/* Read ADC value according to external cable or button */
+	ret = regmap_read(info->regmap, RT8973A_REG_ADC, &adc);
+	if (ret) {
+		dev_err(info->dev, "failed to read ADC register\n");
+		return ret;
+	}
+	cable_type = adc & RT8973A_REG_ADC_MASK;
+
+	/* Read Device 1 reigster to identify correct cable type */
+	ret = regmap_read(info->regmap, RT8973A_REG_DEV1, &dev1);
+	if (ret) {
+		dev_err(info->dev, "failed to read DEV1 register\n");
+		return ret;
+	}
+
+	switch (adc) {
+	case RT8973A_MUIC_ADC_OPEN:
+		if (dev1 & RT8973A_REG_DEV1_USB_MASK)
+			cable_type = RT8973A_MUIC_ADC_USB;
+		else if (dev1 & RT8973A_REG_DEV1_DCPORT_MASK)
+			cable_type = RT8973A_MUIC_ADC_TA;
+		else
+			cable_type = RT8973A_MUIC_ADC_OPEN;
+		break;
+	default:
+		break;
+	}
+
+	return cable_type;
+}
+
+static int rt8973a_muic_cable_handler(struct rt8973a_muic_info *info,
+					enum rt8973a_event_type event)
+{
+	static unsigned int prev_cable_type;
+	const char **cable_names = info->edev->supported_cable;
+	unsigned int con_sw = DM_DP_SWITCH_UART;
+	int ret, idx = 0, cable_type;
+	bool attached = false;
+
+	if (!cable_names)
+		return 0;
+
+	switch (event) {
+	case RT8973A_EVENT_ATTACH:
+		cable_type = rt8973a_muic_get_cable_type(info);
+		attached = true;
+		break;
+	case RT8973A_EVENT_DETACH:
+		cable_type = prev_cable_type;
+		attached = false;
+		break;
+	case RT8973A_EVENT_OVP:
+	case RT8973A_EVENT_OTP:
+		dev_warn(info->dev,
+			"happen Over %s issue. Need to disconnect all cables\n",
+			event == RT8973A_EVENT_OVP ? "Voltage" : "Temperature");
+		cable_type = prev_cable_type;
+		attached = false;
+		break;
+	default:
+		dev_err(info->dev,
+			"Cannot handle this event (event:%d)\n", event);
+		return -EINVAL;
+	}
+	prev_cable_type = cable_type;
+
+	switch (cable_type) {
+	case RT8973A_MUIC_ADC_OTG:
+		idx = EXTCON_CABLE_USB_HOST;
+		con_sw = DM_DP_SWITCH_USB;
+		break;
+	case RT8973A_MUIC_ADC_TA:
+		idx = EXTCON_CABLE_TA;
+		con_sw = DM_DP_SWITCH_OPEN;
+		break;
+	case RT8973A_MUIC_ADC_FACTORY_MODE_BOOT_OFF_USB:
+		idx = EXTCON_CABLE_JIG_OFF_USB;
+		con_sw = DM_DP_SWITCH_UART;
+		break;
+	case RT8973A_MUIC_ADC_FACTORY_MODE_BOOT_ON_USB:
+		idx = EXTCON_CABLE_JIG_ON_USB;
+		con_sw = DM_DP_SWITCH_UART;
+		break;
+	case RT8973A_MUIC_ADC_FACTORY_MODE_BOOT_OFF_UART:
+		idx = EXTCON_CABLE_JIG_OFF_UART;
+		con_sw = DM_DP_SWITCH_UART;
+		break;
+	case RT8973A_MUIC_ADC_FACTORY_MODE_BOOT_ON_UART:
+		idx = EXTCON_CABLE_JIG_ON_UART;
+		con_sw = DM_DP_SWITCH_UART;
+		break;
+	case RT8973A_MUIC_ADC_USB:
+		idx = EXTCON_CABLE_USB;
+		con_sw = DM_DP_SWITCH_USB;
+		break;
+	case RT8973A_MUIC_ADC_OPEN:
+		return 0;
+	case RT8973A_MUIC_ADC_UNKNOWN_ACC_1:
+	case RT8973A_MUIC_ADC_UNKNOWN_ACC_2:
+	case RT8973A_MUIC_ADC_UNKNOWN_ACC_3:
+	case RT8973A_MUIC_ADC_UNKNOWN_ACC_4:
+	case RT8973A_MUIC_ADC_UNKNOWN_ACC_5:
+		dev_warn(info->dev,
+			"Unknown accessory type (adc:0x%x)\n", cable_type);
+		return 0;
+	case RT8973A_MUIC_ADC_AUDIO_SEND_END_BUTTON:
+	case RT8973A_MUIC_ADC_AUDIO_REMOTE_S1_BUTTON:
+	case RT8973A_MUIC_ADC_AUDIO_REMOTE_S2_BUTTON:
+	case RT8973A_MUIC_ADC_AUDIO_REMOTE_S3_BUTTON:
+	case RT8973A_MUIC_ADC_AUDIO_REMOTE_S4_BUTTON:
+	case RT8973A_MUIC_ADC_AUDIO_REMOTE_S5_BUTTON:
+	case RT8973A_MUIC_ADC_AUDIO_REMOTE_S6_BUTTON:
+	case RT8973A_MUIC_ADC_AUDIO_REMOTE_S7_BUTTON:
+	case RT8973A_MUIC_ADC_AUDIO_REMOTE_S8_BUTTON:
+	case RT8973A_MUIC_ADC_AUDIO_REMOTE_S9_BUTTON:
+	case RT8973A_MUIC_ADC_AUDIO_REMOTE_S10_BUTTON:
+	case RT8973A_MUIC_ADC_AUDIO_REMOTE_S11_BUTTON:
+	case RT8973A_MUIC_ADC_AUDIO_REMOTE_S12_BUTTON:
+	case RT8973A_MUIC_ADC_AUDIO_TYPE2:
+		dev_warn(info->dev,
+			"Audio device/button type (adc:0x%x)\n", cable_type);
+		return 0;
+	case RT8973A_MUIC_ADC_RESERVED_ACC_1:
+	case RT8973A_MUIC_ADC_RESERVED_ACC_2:
+	case RT8973A_MUIC_ADC_RESERVED_ACC_3:
+	case RT8973A_MUIC_ADC_RESERVED_ACC_4:
+	case RT8973A_MUIC_ADC_RESERVED_ACC_5:
+	case RT8973A_MUIC_ADC_PHONE_POWERED_DEV:
+		return 0;
+	default:
+		dev_err(info->dev,
+			"Cannot handle this cable_type (adc:0x%x)\n",
+			cable_type);
+		return -EINVAL;
+	}
+
+	/* Change internal hardware path(DM_CON/DP_CON) */
+	ret = rt8973a_muic_set_path(info, con_sw, attached);
+	if (ret < 0)
+		return ret;
+
+	/* Change the state of external accessory */
+	extcon_set_cable_state(info->edev, cable_names[idx], attached);
+
+	return 0;
+}
+
+static void rt8973a_muic_irq_work(struct work_struct *work)
+{
+	struct rt8973a_muic_info *info = container_of(work,
+			struct rt8973a_muic_info, irq_work);
+	int ret = 0;
+
+	if (!info->edev)
+		return;
+
+	mutex_lock(&info->mutex);
+
+	/* Detect attached or detached cables */
+	if (info->irq_attach) {
+		ret = rt8973a_muic_cable_handler(info, RT8973A_EVENT_ATTACH);
+		info->irq_attach = false;
+	}
+
+	if (info->irq_detach) {
+		ret = rt8973a_muic_cable_handler(info, RT8973A_EVENT_DETACH);
+		info->irq_detach = false;
+	}
+
+	if (info->irq_ovp) {
+		ret = rt8973a_muic_cable_handler(info, RT8973A_EVENT_OVP);
+		info->irq_ovp = false;
+	}
+
+	if (info->irq_otp) {
+		ret = rt8973a_muic_cable_handler(info, RT8973A_EVENT_OTP);
+		info->irq_otp = false;
+	}
+
+	if (ret < 0)
+		dev_err(info->dev, "failed to handle MUIC interrupt\n");
+
+	mutex_unlock(&info->mutex);
+}
+
+static irqreturn_t rt8973a_muic_irq_handler(int irq, void *data)
+{
+	struct rt8973a_muic_info *info = data;
+	int i, irq_type = -1;
+
+	for (i = 0; i < info->num_muic_irqs; i++)
+		if (irq == info->muic_irqs[i].virq)
+			irq_type = info->muic_irqs[i].irq;
+
+	switch (irq_type) {
+	case RT8973A_INT1_ATTACH:
+		info->irq_attach = true;
+		break;
+	case RT8973A_INT1_DETACH:
+		info->irq_detach = true;
+		break;
+	case RT8973A_INT1_OVP:
+		info->irq_ovp = true;
+		break;
+	case RT8973A_INT1_OTP:
+		info->irq_otp = true;
+		break;
+	case RT8973A_INT1_CHGDET:
+	case RT8973A_INT1_DCD_T:
+	case RT8973A_INT1_CONNECT:
+	case RT8973A_INT1_ADC_CHG:
+	case RT8973A_INT2_UVLO:
+	case RT8973A_INT2_POR:
+	case RT8973A_INT2_OTP_FET:
+	case RT8973A_INT2_OVP_FET:
+	case RT8973A_INT2_OCP_LATCH:
+	case RT8973A_INT2_OCP:
+	case RT8973A_INT2_OVP_OCP:
+	default:
+		dev_dbg(info->dev,
+			"Cannot handle this interrupt (%d)\n", irq_type);
+		break;
+	}
+
+	schedule_work(&info->irq_work);
+
+	return IRQ_HANDLED;
+}
+
+static void rt8973a_muic_detect_cable_wq(struct work_struct *work)
+{
+	struct rt8973a_muic_info *info = container_of(to_delayed_work(work),
+				struct rt8973a_muic_info, wq_detcable);
+	int ret;
+
+	/* Notify the state of connector cable or not  */
+	ret = rt8973a_muic_cable_handler(info, RT8973A_EVENT_ATTACH);
+	if (ret < 0)
+		dev_warn(info->dev, "failed to detect cable state\n");
+}
+
+static void rt8973a_init_dev_type(struct rt8973a_muic_info *info)
+{
+	unsigned int data, vendor_id, version_id;
+	int i, ret;
+
+	/* To test I2C, Print version_id and vendor_id of RT8973A */
+	ret = regmap_read(info->regmap, RT8973A_REG_DEVICE_ID, &data);
+	if (ret) {
+		dev_err(info->dev,
+			"failed to read DEVICE_ID register: %d\n", ret);
+		return;
+	}
+
+	vendor_id = ((data & RT8973A_REG_DEVICE_ID_VENDOR_MASK) >>
+				RT8973A_REG_DEVICE_ID_VENDOR_SHIFT);
+	version_id = ((data & RT8973A_REG_DEVICE_ID_VERSION_MASK) >>
+				RT8973A_REG_DEVICE_ID_VERSION_SHIFT);
+
+	dev_info(info->dev, "Device type: version: 0x%x, vendor: 0x%x\n",
+			    version_id, vendor_id);
+
+	/* Initiazle the register of RT8973A device to bring-up */
+	for (i = 0; i < info->num_reg_data; i++) {
+		u8 reg = info->reg_data[i].reg;
+		u8 mask = info->reg_data[i].mask;
+		u8 val = 0;
+
+		if (info->reg_data[i].invert)
+			val = ~info->reg_data[i].val;
+		else
+			val = info->reg_data[i].val;
+
+		regmap_update_bits(info->regmap, reg, mask, val);
+	}
+
+	/* Check whether RT8973A is auto swithcing mode or not */
+	ret = regmap_read(info->regmap, RT8973A_REG_CONTROL1, &data);
+	if (ret) {
+		dev_err(info->dev,
+			"failed to read CONTROL1 register: %d\n", ret);
+		return;
+	}
+
+	data &= RT8973A_REG_CONTROL1_AUTO_CONFIG_MASK;
+	if (data) {
+		info->auto_config = true;
+		dev_info(info->dev,
+			"Enable Auto-configuration for internal path\n");
+	}
+}
+
+static int rt8973a_muic_i2c_probe(struct i2c_client *i2c,
+				 const struct i2c_device_id *id)
+{
+	struct device_node *np = i2c->dev.of_node;
+	struct rt8973a_muic_info *info;
+	int i, ret, irq_flags;
+
+	if (!np)
+		return -EINVAL;
+
+	info = devm_kzalloc(&i2c->dev, sizeof(*info), GFP_KERNEL);
+	if (!info) {
+		dev_err(&i2c->dev, "failed to allocate memory\n");
+		return -ENOMEM;
+	}
+	i2c_set_clientdata(i2c, info);
+
+	info->dev = &i2c->dev;
+	info->i2c = i2c;
+	info->irq = i2c->irq;
+	info->muic_irqs = rt8973a_muic_irqs;
+	info->num_muic_irqs = ARRAY_SIZE(rt8973a_muic_irqs);
+	info->reg_data = rt8973a_reg_data;
+	info->num_reg_data = ARRAY_SIZE(rt8973a_reg_data);
+
+	mutex_init(&info->mutex);
+
+	INIT_WORK(&info->irq_work, rt8973a_muic_irq_work);
+
+	info->regmap = devm_regmap_init_i2c(i2c, &rt8973a_muic_regmap_config);
+	if (IS_ERR(info->regmap)) {
+		ret = PTR_ERR(info->regmap);
+		dev_err(info->dev, "failed to allocate register map: %d\n",
+				   ret);
+		return ret;
+	}
+
+	/* Support irq domain for RT8973A MUIC device */
+	irq_flags = IRQF_TRIGGER_FALLING | IRQF_ONESHOT | IRQF_SHARED;
+	ret = regmap_add_irq_chip(info->regmap, info->irq, irq_flags, 0,
+				  &rt8973a_muic_irq_chip, &info->irq_data);
+	if (ret != 0) {
+		dev_err(info->dev, "failed to add irq_chip (irq:%d, err:%d)\n",
+				    info->irq, ret);
+		return ret;
+	}
+
+	for (i = 0; i < info->num_muic_irqs; i++) {
+		struct muic_irq *muic_irq = &info->muic_irqs[i];
+		unsigned int virq = 0;
+
+		virq = regmap_irq_get_virq(info->irq_data, muic_irq->irq);
+		if (virq <= 0)
+			return -EINVAL;
+		muic_irq->virq = virq;
+
+		ret = devm_request_threaded_irq(info->dev, virq, NULL,
+						rt8973a_muic_irq_handler,
+						IRQF_NO_SUSPEND,
+						muic_irq->name, info);
+		if (ret) {
+			dev_err(info->dev,
+				"failed: irq request (IRQ: %d, error :%d)\n",
+				muic_irq->irq, ret);
+			return ret;
+		}
+	}
+
+	/* Allocate extcon device */
+	info->edev = devm_extcon_dev_allocate(info->dev, rt8973a_extcon_cable);
+	if (IS_ERR(info->edev)) {
+		dev_err(info->dev, "failed to allocate memory for extcon\n");
+		return -ENOMEM;
+	}
+	info->edev->name = np->name;
+
+	/* Register extcon device */
+	ret = devm_extcon_dev_register(info->dev, info->edev);
+	if (ret) {
+		dev_err(info->dev, "failed to register extcon device\n");
+		return ret;
+	}
+
+	/*
+	 * Detect accessory after completing the initialization of platform
+	 *
+	 * - Use delayed workqueue to detect cable state and then
+	 * notify cable state to notifiee/platform through uevent.
+	 * After completing the booting of platform, the extcon provider
+	 * driver should notify cable state to upper layer.
+	 */
+	INIT_DELAYED_WORK(&info->wq_detcable, rt8973a_muic_detect_cable_wq);
+	queue_delayed_work(system_power_efficient_wq, &info->wq_detcable,
+			msecs_to_jiffies(DELAY_MS_DEFAULT));
+
+	/* Initialize RT8973A device and print vendor id and version id */
+	rt8973a_init_dev_type(info);
+
+	return 0;
+}
+
+static int rt8973a_muic_i2c_remove(struct i2c_client *i2c)
+{
+	struct rt8973a_muic_info *info = i2c_get_clientdata(i2c);
+
+	regmap_del_irq_chip(info->irq, info->irq_data);
+
+	return 0;
+}
+
+static struct of_device_id rt8973a_dt_match[] = {
+	{ .compatible = "richtek,rt8973a-muic" },
+	{ },
+};
+
+#ifdef CONFIG_PM_SLEEP
+static int rt8973a_muic_suspend(struct device *dev)
+{
+	struct i2c_client *i2c = container_of(dev, struct i2c_client, dev);
+	struct rt8973a_muic_info *info = i2c_get_clientdata(i2c);
+
+	enable_irq_wake(info->irq);
+
+	return 0;
+}
+
+static int rt8973a_muic_resume(struct device *dev)
+{
+	struct i2c_client *i2c = container_of(dev, struct i2c_client, dev);
+	struct rt8973a_muic_info *info = i2c_get_clientdata(i2c);
+
+	disable_irq_wake(info->irq);
+
+	return 0;
+}
+#endif
+
+static SIMPLE_DEV_PM_OPS(rt8973a_muic_pm_ops,
+			 rt8973a_muic_suspend, rt8973a_muic_resume);
+
+static const struct i2c_device_id rt8973a_i2c_id[] = {
+	{ "rt8973a", TYPE_RT8973A },
+	{ }
+};
+MODULE_DEVICE_TABLE(i2c, rt8973a_i2c_id);
+
+static struct i2c_driver rt8973a_muic_i2c_driver = {
+	.driver		= {
+		.name	= "rt8973a",
+		.owner	= THIS_MODULE,
+		.pm	= &rt8973a_muic_pm_ops,
+		.of_match_table = rt8973a_dt_match,
+	},
+	.probe	= rt8973a_muic_i2c_probe,
+	.remove	= rt8973a_muic_i2c_remove,
+	.id_table = rt8973a_i2c_id,
+};
+
+static int __init rt8973a_muic_i2c_init(void)
+{
+	return i2c_add_driver(&rt8973a_muic_i2c_driver);
+}
+subsys_initcall(rt8973a_muic_i2c_init);
+
+MODULE_DESCRIPTION("Richtek RT8973A Extcon driver");
+MODULE_AUTHOR("Chanwoo Choi <cw00.choi@samsung.com>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/extcon/extcon-rt8973a.h b/drivers/extcon/extcon-rt8973a.h
new file mode 100644
index 000000000000..9dc3e0227eb7
--- /dev/null
+++ b/drivers/extcon/extcon-rt8973a.h
@@ -0,0 +1,203 @@
+/*
+ * rt8973a.h
+ *
+ * Copyright (c) 2014 Samsung Electronics Co., Ltd
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+
+#ifndef __LINUX_EXTCON_RT8973A_H
+#define __LINUX_EXTCON_RT8973A_H
+
+enum rt8973a_types {
+	TYPE_RT8973A,
+};
+
+/* RT8973A registers */
+enum rt8973A_reg {
+	RT8973A_REG_DEVICE_ID = 0x1,
+	RT8973A_REG_CONTROL1,
+	RT8973A_REG_INT1,
+	RT8973A_REG_INT2,
+	RT8973A_REG_INTM1,
+	RT8973A_REG_INTM2,
+	RT8973A_REG_ADC,
+	RT8973A_REG_RSVD_1,
+	RT8973A_REG_RSVD_2,
+	RT8973A_REG_DEV1,
+	RT8973A_REG_DEV2,
+	RT8973A_REG_RSVD_3,
+	RT8973A_REG_RSVD_4,
+	RT8973A_REG_RSVD_5,
+	RT8973A_REG_RSVD_6,
+	RT8973A_REG_RSVD_7,
+	RT8973A_REG_RSVD_8,
+	RT8973A_REG_RSVD_9,
+	RT8973A_REG_MANUAL_SW1,
+	RT8973A_REG_MANUAL_SW2,
+	RT8973A_REG_RSVD_10,
+	RT8973A_REG_RSVD_11,
+	RT8973A_REG_RSVD_12,
+	RT8973A_REG_RSVD_13,
+	RT8973A_REG_RSVD_14,
+	RT8973A_REG_RSVD_15,
+	RT8973A_REG_RESET,
+
+	RT8973A_REG_END,
+};
+
+/* Define RT8973A MASK/SHIFT constant */
+#define RT8973A_REG_DEVICE_ID_VENDOR_SHIFT	0
+#define RT8973A_REG_DEVICE_ID_VERSION_SHIFT	3
+#define RT8973A_REG_DEVICE_ID_VENDOR_MASK	(0x7 << RT8973A_REG_DEVICE_ID_VENDOR_SHIFT)
+#define RT8973A_REG_DEVICE_ID_VERSION_MASK	(0x1f << RT8973A_REG_DEVICE_ID_VERSION_SHIFT)
+
+#define RT8973A_REG_CONTROL1_INTM_SHIFT	0
+#define RT8973A_REG_CONTROL1_AUTO_CONFIG_SHIFT	2
+#define RT8973A_REG_CONTROL1_I2C_RST_EN_SHIFT	3
+#define RT8973A_REG_CONTROL1_SWITCH_OPEN_SHIFT	4
+#define RT8973A_REG_CONTROL1_CHGTYP_SHIFT	5
+#define RT8973A_REG_CONTROL1_USB_CHD_EN_SHIFT	6
+#define RT8973A_REG_CONTROL1_ADC_EN_SHIFT	7
+#define RT8973A_REG_CONTROL1_INTM_MASK		(0x1 << RT8973A_REG_CONTROL1_INTM_SHIFT)
+#define RT8973A_REG_CONTROL1_AUTO_CONFIG_MASK	(0x1 << RT8973A_REG_CONTROL1_AUTO_CONFIG_SHIFT)
+#define RT8973A_REG_CONTROL1_I2C_RST_EN_MASK	(0x1 << RT8973A_REG_CONTROL1_I2C_RST_EN_SHIFT)
+#define RT8973A_REG_CONTROL1_SWITCH_OPEN_MASK	(0x1 << RT8973A_REG_CONTROL1_SWITCH_OPEN_SHIFT)
+#define RT8973A_REG_CONTROL1_CHGTYP_MASK	(0x1 << RT8973A_REG_CONTROL1_CHGTYP_SHIFT)
+#define RT8973A_REG_CONTROL1_USB_CHD_EN_MASK	(0x1 << RT8973A_REG_CONTROL1_USB_CHD_EN_SHIFT)
+#define RT8973A_REG_CONTROL1_ADC_EN_MASK	(0x1 << RT8973A_REG_CONTROL1_ADC_EN_SHIFT)
+
+#define RT9873A_REG_INTM1_ATTACH_SHIFT		0
+#define RT9873A_REG_INTM1_DETACH_SHIFT		1
+#define RT9873A_REG_INTM1_CHGDET_SHIFT		2
+#define RT9873A_REG_INTM1_DCD_T_SHIFT		3
+#define RT9873A_REG_INTM1_OVP_SHIFT		4
+#define RT9873A_REG_INTM1_CONNECT_SHIFT		5
+#define RT9873A_REG_INTM1_ADC_CHG_SHIFT		6
+#define RT9873A_REG_INTM1_OTP_SHIFT		7
+#define RT9873A_REG_INTM1_ATTACH_MASK		(0x1 << RT9873A_REG_INTM1_ATTACH_SHIFT)
+#define RT9873A_REG_INTM1_DETACH_MASK		(0x1 <<  RT9873A_REG_INTM1_DETACH_SHIFT)
+#define RT9873A_REG_INTM1_CHGDET_MASK		(0x1 <<  RT9873A_REG_INTM1_CHGDET_SHIFT)
+#define RT9873A_REG_INTM1_DCD_T_MASK		(0x1 <<  RT9873A_REG_INTM1_DCD_T_SHIFT)
+#define RT9873A_REG_INTM1_OVP_MASK		(0x1 <<  RT9873A_REG_INTM1_OVP_SHIFT)
+#define RT9873A_REG_INTM1_CONNECT_MASK		(0x1 <<  RT9873A_REG_INTM1_CONNECT_SHIFT)
+#define RT9873A_REG_INTM1_ADC_CHG_MASK		(0x1 <<  RT9873A_REG_INTM1_ADC_CHG_SHIFT)
+#define RT9873A_REG_INTM1_OTP_MASK		(0x1 <<  RT9873A_REG_INTM1_OTP_SHIFT)
+
+#define RT9873A_REG_INTM2_UVLO_SHIFT		1
+#define RT9873A_REG_INTM2_POR_SHIFT		2
+#define RT9873A_REG_INTM2_OTP_FET_SHIFT		3
+#define RT9873A_REG_INTM2_OVP_FET_SHIFT		4
+#define RT9873A_REG_INTM2_OCP_LATCH_SHIFT	5
+#define RT9873A_REG_INTM2_OCP_SHIFT		6
+#define RT9873A_REG_INTM2_OVP_OCP_SHIFT		7
+#define RT9873A_REG_INTM2_UVLO_MASK		(0x1 << RT9873A_REG_INTM2_UVLO_SHIFT)
+#define RT9873A_REG_INTM2_POR_MASK		(0x1 <<  RT9873A_REG_INTM2_POR_SHIFT)
+#define RT9873A_REG_INTM2_OTP_FET_MASK		(0x1 <<  RT9873A_REG_INTM2_OTP_FET_SHIFT)
+#define RT9873A_REG_INTM2_OVP_FET_MASK		(0x1 <<  RT9873A_REG_INTM2_OVP_FET_SHIFT)
+#define RT9873A_REG_INTM2_OCP_LATCH_MASK	(0x1 <<  RT9873A_REG_INTM2_OCP_LATCH_SHIFT)
+#define RT9873A_REG_INTM2_OCP_MASK		(0x1 <<  RT9873A_REG_INTM2_OCP_SHIFT)
+#define RT9873A_REG_INTM2_OVP_OCP_MASK		(0x1 <<  RT9873A_REG_INTM2_OVP_OCP_SHIFT)
+
+#define RT8973A_REG_ADC_SHIFT			0
+#define RT8973A_REG_ADC_MASK			(0x1f << RT8973A_REG_ADC_SHIFT)
+
+#define RT8973A_REG_DEV1_OTG_SHIFT		0
+#define RT8973A_REG_DEV1_SDP_SHIFT		2
+#define RT8973A_REG_DEV1_UART_SHIFT		3
+#define RT8973A_REG_DEV1_CAR_KIT_TYPE1_SHIFT	4
+#define RT8973A_REG_DEV1_CDPORT_SHIFT		5
+#define RT8973A_REG_DEV1_DCPORT_SHIFT		6
+#define RT8973A_REG_DEV1_OTG_MASK		(0x1 << RT8973A_REG_DEV1_OTG_SHIFT)
+#define RT8973A_REG_DEV1_SDP_MASK		(0x1 << RT8973A_REG_DEV1_SDP_SHIFT)
+#define RT8973A_REG_DEV1_UART_MASK		(0x1 << RT8973A_REG_DEV1_UART_SHIFT)
+#define RT8973A_REG_DEV1_CAR_KIT_TYPE1_MASK	(0x1 << RT8973A_REG_DEV1_CAR_KIT_TYPE1_SHIFT)
+#define RT8973A_REG_DEV1_CDPORT_MASK		(0x1 << RT8973A_REG_DEV1_CDPORT_SHIFT)
+#define RT8973A_REG_DEV1_DCPORT_MASK		(0x1 << RT8973A_REG_DEV1_DCPORT_SHIFT)
+#define RT8973A_REG_DEV1_USB_MASK		(RT8973A_REG_DEV1_SDP_MASK \
+						| RT8973A_REG_DEV1_CDPORT_MASK)
+
+#define RT8973A_REG_DEV2_JIG_USB_ON_SHIFT	0
+#define RT8973A_REG_DEV2_JIG_USB_OFF_SHIFT	1
+#define RT8973A_REG_DEV2_JIG_UART_ON_SHIFT	2
+#define RT8973A_REG_DEV2_JIG_UART_OFF_SHIFT	3
+#define RT8973A_REG_DEV2_JIG_USB_ON_MASK	(0x1 << RT8973A_REG_DEV2_JIG_USB_ON_SHIFT)
+#define RT8973A_REG_DEV2_JIG_USB_OFF_MASK	(0x1 << RT8973A_REG_DEV2_JIG_USB_OFF_SHIFT)
+#define RT8973A_REG_DEV2_JIG_UART_ON_MASK	(0x1 << RT8973A_REG_DEV2_JIG_UART_ON_SHIFT)
+#define RT8973A_REG_DEV2_JIG_UART_OFF_MASK	(0x1 << RT8973A_REG_DEV2_JIG_UART_OFF_SHIFT)
+
+#define RT8973A_REG_MANUAL_SW1_DP_SHIFT		2
+#define RT8973A_REG_MANUAL_SW1_DM_SHIFT		5
+#define RT8973A_REG_MANUAL_SW1_DP_MASK		(0x7 << RT8973A_REG_MANUAL_SW1_DP_SHIFT)
+#define RT8973A_REG_MANUAL_SW1_DM_MASK		(0x7 << RT8973A_REG_MANUAL_SW1_DM_SHIFT)
+#define DM_DP_CON_SWITCH_OPEN			0x0
+#define DM_DP_CON_SWITCH_USB			0x1
+#define DM_DP_CON_SWITCH_UART			0x3
+#define DM_DP_SWITCH_OPEN			((DM_DP_CON_SWITCH_OPEN << RT8973A_REG_MANUAL_SW1_DP_SHIFT) \
+						| (DM_DP_CON_SWITCH_OPEN << RT8973A_REG_MANUAL_SW1_DM_SHIFT))
+#define DM_DP_SWITCH_USB			((DM_DP_CON_SWITCH_USB << RT8973A_REG_MANUAL_SW1_DP_SHIFT) \
+						| (DM_DP_CON_SWITCH_USB << RT8973A_REG_MANUAL_SW1_DM_SHIFT))
+#define DM_DP_SWITCH_UART			((DM_DP_CON_SWITCH_UART << RT8973A_REG_MANUAL_SW1_DP_SHIFT) \
+						| (DM_DP_CON_SWITCH_UART << RT8973A_REG_MANUAL_SW1_DM_SHIFT))
+
+#define RT8973A_REG_MANUAL_SW2_FET_ON_SHIFT	0
+#define RT8973A_REG_MANUAL_SW2_JIG_ON_SHIFT	2
+#define RT8973A_REG_MANUAL_SW2_BOOT_SW_SHIFT	3
+#define RT8973A_REG_MANUAL_SW2_FET_ON_MASK	(0x1 << RT8973A_REG_MANUAL_SW2_FET_ON_SHIFT)
+#define RT8973A_REG_MANUAL_SW2_JIG_ON_MASK	(0x1 << RT8973A_REG_MANUAL_SW2_JIG_ON_SHIFT)
+#define RT8973A_REG_MANUAL_SW2_BOOT_SW_MASK	(0x1 << RT8973A_REG_MANUAL_SW2_BOOT_SW_SHIFT)
+#define RT8973A_REG_MANUAL_SW2_FET_ON		0
+#define RT8973A_REG_MANUAL_SW2_FET_OFF		0x1
+#define RT8973A_REG_MANUAL_SW2_JIG_OFF		0
+#define RT8973A_REG_MANUAL_SW2_JIG_ON		0x1
+#define RT8973A_REG_MANUAL_SW2_BOOT_SW_ON	0
+#define RT8973A_REG_MANUAL_SW2_BOOT_SW_OFF	0x1
+
+#define RT8973A_REG_RESET_SHIFT			0
+#define RT8973A_REG_RESET_MASK			(0x1 << RT8973A_REG_RESET_SHIFT)
+#define RT8973A_REG_RESET			0x1
+
+/* RT8973A Interrupts */
+enum rt8973a_irq {
+	/* Interrupt1*/
+	RT8973A_INT1_ATTACH,
+	RT8973A_INT1_DETACH,
+	RT8973A_INT1_CHGDET,
+	RT8973A_INT1_DCD_T,
+	RT8973A_INT1_OVP,
+	RT8973A_INT1_CONNECT,
+	RT8973A_INT1_ADC_CHG,
+	RT8973A_INT1_OTP,
+
+	/* Interrupt2*/
+	RT8973A_INT2_UVLO,
+	RT8973A_INT2_POR,
+	RT8973A_INT2_OTP_FET,
+	RT8973A_INT2_OVP_FET,
+	RT8973A_INT2_OCP_LATCH,
+	RT8973A_INT2_OCP,
+	RT8973A_INT2_OVP_OCP,
+
+	RT8973A_NUM,
+};
+
+#define RT8973A_INT1_ATTACH_MASK		BIT(0)
+#define RT8973A_INT1_DETACH_MASK		BIT(1)
+#define RT8973A_INT1_CHGDET_MASK		BIT(2)
+#define RT8973A_INT1_DCD_T_MASK			BIT(3)
+#define RT8973A_INT1_OVP_MASK			BIT(4)
+#define RT8973A_INT1_CONNECT_MASK		BIT(5)
+#define RT8973A_INT1_ADC_CHG_MASK		BIT(6)
+#define RT8973A_INT1_OTP_MASK			BIT(7)
+#define RT8973A_INT2_UVLOT_MASK			BIT(0)
+#define RT8973A_INT2_POR_MASK			BIT(1)
+#define RT8973A_INT2_OTP_FET_MASK		BIT(2)
+#define RT8973A_INT2_OVP_FET_MASK		BIT(3)
+#define RT8973A_INT2_OCP_LATCH_MASK		BIT(4)
+#define RT8973A_INT2_OCP_MASK			BIT(5)
+#define RT8973A_INT2_OVP_OCP_MASK		BIT(6)
+
+#endif /*  __LINUX_EXTCON_RT8973A_H */
diff --git a/drivers/extcon/extcon-sm5502.c b/drivers/extcon/extcon-sm5502.c
index 560d7dccec7b..b0f7bd82af90 100644
--- a/drivers/extcon/extcon-sm5502.c
+++ b/drivers/extcon/extcon-sm5502.c
@@ -8,16 +8,10 @@
  * under  the terms of  the GNU General  Public License as published by the
  * Free Software Foundation;  either version 2 of the  License, or (at your
  * option) any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
  */
 
 #include <linux/err.h>
 #include <linux/i2c.h>
-#include <linux/input.h>
 #include <linux/interrupt.h>
 #include <linux/irqdomain.h>
 #include <linux/kernel.h>
@@ -26,7 +20,8 @@
 #include <linux/regmap.h>
 #include <linux/slab.h>
 #include <linux/extcon.h>
-#include <linux/extcon/sm5502.h>
+
+#include "extcon-sm5502.h"
 
 #define	DELAY_MS_DEFAULT		17000	/* unit: millisecond */
 
@@ -300,7 +295,7 @@ static unsigned int sm5502_muic_get_cable_type(struct sm5502_muic_info *info)
 	 * If ADC is SM5502_MUIC_ADC_GROUND(0x0), external cable hasn't
 	 * connected with to MUIC device.
 	 */
-	cable_type &= SM5502_REG_ADC_MASK;
+	cable_type = adc & SM5502_REG_ADC_MASK;
 	if (cable_type == SM5502_MUIC_ADC_GROUND)
 		return SM5502_MUIC_ADC_GROUND;
 
@@ -395,7 +390,7 @@ static int sm5502_muic_cable_handler(struct sm5502_muic_info *info,
 	/* Get the type of attached or detached cable */
 	if (attached)
 		cable_type = sm5502_muic_get_cable_type(info);
-	else if (!attached)
+	else
 		cable_type = prev_cable_type;
 	prev_cable_type = cable_type;
 
@@ -457,8 +452,6 @@ static void sm5502_muic_irq_work(struct work_struct *work)
 		dev_err(info->dev, "failed to handle MUIC interrupt\n");
 
 	mutex_unlock(&info->mutex);
-
-	return;
 }
 
 /*
@@ -617,8 +610,9 @@ static int sm5022_muic_i2c_probe(struct i2c_client *i2c,
 						IRQF_NO_SUSPEND,
 						muic_irq->name, info);
 		if (ret) {
-			dev_err(info->dev, "failed: irq request (IRQ: %d,"
-				" error :%d)\n", muic_irq->irq, ret);
+			dev_err(info->dev,
+				"failed: irq request (IRQ: %d, error :%d)\n",
+				muic_irq->irq, ret);
 			return ret;
 		}
 	}
diff --git a/drivers/extcon/extcon-sm5502.h b/drivers/extcon/extcon-sm5502.h
new file mode 100644
index 000000000000..974b53222f56
--- /dev/null
+++ b/drivers/extcon/extcon-sm5502.h
@@ -0,0 +1,282 @@
+/*
+ * sm5502.h
+ *
+ * Copyright (c) 2014 Samsung Electronics Co., Ltd
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+
+#ifndef __LINUX_EXTCON_SM5502_H
+#define __LINUX_EXTCON_SM5502_H
+
+enum sm5502_types {
+	TYPE_SM5502,
+};
+
+/* SM5502 registers */
+enum sm5502_reg {
+	SM5502_REG_DEVICE_ID = 0x01,
+	SM5502_REG_CONTROL,
+	SM5502_REG_INT1,
+	SM5502_REG_INT2,
+	SM5502_REG_INTMASK1,
+	SM5502_REG_INTMASK2,
+	SM5502_REG_ADC,
+	SM5502_REG_TIMING_SET1,
+	SM5502_REG_TIMING_SET2,
+	SM5502_REG_DEV_TYPE1,
+	SM5502_REG_DEV_TYPE2,
+	SM5502_REG_BUTTON1,
+	SM5502_REG_BUTTON2,
+	SM5502_REG_CAR_KIT_STATUS,
+	SM5502_REG_RSVD1,
+	SM5502_REG_RSVD2,
+	SM5502_REG_RSVD3,
+	SM5502_REG_RSVD4,
+	SM5502_REG_MANUAL_SW1,
+	SM5502_REG_MANUAL_SW2,
+	SM5502_REG_DEV_TYPE3,
+	SM5502_REG_RSVD5,
+	SM5502_REG_RSVD6,
+	SM5502_REG_RSVD7,
+	SM5502_REG_RSVD8,
+	SM5502_REG_RSVD9,
+	SM5502_REG_RESET,
+	SM5502_REG_RSVD10,
+	SM5502_REG_RESERVED_ID1,
+	SM5502_REG_RSVD11,
+	SM5502_REG_RSVD12,
+	SM5502_REG_RESERVED_ID2,
+	SM5502_REG_RSVD13,
+	SM5502_REG_OCP,
+	SM5502_REG_RSVD14,
+	SM5502_REG_RSVD15,
+	SM5502_REG_RSVD16,
+	SM5502_REG_RSVD17,
+	SM5502_REG_RSVD18,
+	SM5502_REG_RSVD19,
+	SM5502_REG_RSVD20,
+	SM5502_REG_RSVD21,
+	SM5502_REG_RSVD22,
+	SM5502_REG_RSVD23,
+	SM5502_REG_RSVD24,
+	SM5502_REG_RSVD25,
+	SM5502_REG_RSVD26,
+	SM5502_REG_RSVD27,
+	SM5502_REG_RSVD28,
+	SM5502_REG_RSVD29,
+	SM5502_REG_RSVD30,
+	SM5502_REG_RSVD31,
+	SM5502_REG_RSVD32,
+	SM5502_REG_RSVD33,
+	SM5502_REG_RSVD34,
+	SM5502_REG_RSVD35,
+	SM5502_REG_RSVD36,
+	SM5502_REG_RESERVED_ID3,
+
+	SM5502_REG_END,
+};
+
+/* Define SM5502 MASK/SHIFT constant */
+#define SM5502_REG_DEVICE_ID_VENDOR_SHIFT	0
+#define SM5502_REG_DEVICE_ID_VERSION_SHIFT	3
+#define SM5502_REG_DEVICE_ID_VENDOR_MASK	(0x3 << SM5502_REG_DEVICE_ID_VENDOR_SHIFT)
+#define SM5502_REG_DEVICE_ID_VERSION_MASK	(0x1f << SM5502_REG_DEVICE_ID_VERSION_SHIFT)
+
+#define SM5502_REG_CONTROL_MASK_INT_SHIFT	0
+#define SM5502_REG_CONTROL_WAIT_SHIFT		1
+#define SM5502_REG_CONTROL_MANUAL_SW_SHIFT	2
+#define SM5502_REG_CONTROL_RAW_DATA_SHIFT	3
+#define SM5502_REG_CONTROL_SW_OPEN_SHIFT	4
+#define SM5502_REG_CONTROL_MASK_INT_MASK	(0x1 << SM5502_REG_CONTROL_MASK_INT_SHIFT)
+#define SM5502_REG_CONTROL_WAIT_MASK		(0x1 << SM5502_REG_CONTROL_WAIT_SHIFT)
+#define SM5502_REG_CONTROL_MANUAL_SW_MASK	(0x1 << SM5502_REG_CONTROL_MANUAL_SW_SHIFT)
+#define SM5502_REG_CONTROL_RAW_DATA_MASK	(0x1 << SM5502_REG_CONTROL_RAW_DATA_SHIFT)
+#define SM5502_REG_CONTROL_SW_OPEN_MASK		(0x1 << SM5502_REG_CONTROL_SW_OPEN_SHIFT)
+
+#define SM5502_REG_INTM1_ATTACH_SHIFT		0
+#define SM5502_REG_INTM1_DETACH_SHIFT		1
+#define SM5502_REG_INTM1_KP_SHIFT		2
+#define SM5502_REG_INTM1_LKP_SHIFT		3
+#define SM5502_REG_INTM1_LKR_SHIFT		4
+#define SM5502_REG_INTM1_OVP_EVENT_SHIFT	5
+#define SM5502_REG_INTM1_OCP_EVENT_SHIFT	6
+#define SM5502_REG_INTM1_OVP_OCP_DIS_SHIFT	7
+#define SM5502_REG_INTM1_ATTACH_MASK		(0x1 << SM5502_REG_INTM1_ATTACH_SHIFT)
+#define SM5502_REG_INTM1_DETACH_MASK		(0x1 << SM5502_REG_INTM1_DETACH_SHIFT)
+#define SM5502_REG_INTM1_KP_MASK		(0x1 << SM5502_REG_INTM1_KP_SHIFT)
+#define SM5502_REG_INTM1_LKP_MASK		(0x1 << SM5502_REG_INTM1_LKP_SHIFT)
+#define SM5502_REG_INTM1_LKR_MASK		(0x1 << SM5502_REG_INTM1_LKR_SHIFT)
+#define SM5502_REG_INTM1_OVP_EVENT_MASK		(0x1 << SM5502_REG_INTM1_OVP_EVENT_SHIFT)
+#define SM5502_REG_INTM1_OCP_EVENT_MASK		(0x1 << SM5502_REG_INTM1_OCP_EVENT_SHIFT)
+#define SM5502_REG_INTM1_OVP_OCP_DIS_MASK	(0x1 << SM5502_REG_INTM1_OVP_OCP_DIS_SHIFT)
+
+#define SM5502_REG_INTM2_VBUS_DET_SHIFT		0
+#define SM5502_REG_INTM2_REV_ACCE_SHIFT		1
+#define SM5502_REG_INTM2_ADC_CHG_SHIFT		2
+#define SM5502_REG_INTM2_STUCK_KEY_SHIFT	3
+#define SM5502_REG_INTM2_STUCK_KEY_RCV_SHIFT	4
+#define SM5502_REG_INTM2_MHL_SHIFT		5
+#define SM5502_REG_INTM2_VBUS_DET_MASK		(0x1 << SM5502_REG_INTM2_VBUS_DET_SHIFT)
+#define SM5502_REG_INTM2_REV_ACCE_MASK		(0x1 << SM5502_REG_INTM2_REV_ACCE_SHIFT)
+#define SM5502_REG_INTM2_ADC_CHG_MASK		(0x1 << SM5502_REG_INTM2_ADC_CHG_SHIFT)
+#define SM5502_REG_INTM2_STUCK_KEY_MASK		(0x1 << SM5502_REG_INTM2_STUCK_KEY_SHIFT)
+#define SM5502_REG_INTM2_STUCK_KEY_RCV_MASK	(0x1 << SM5502_REG_INTM2_STUCK_KEY_RCV_SHIFT)
+#define SM5502_REG_INTM2_MHL_MASK		(0x1 << SM5502_REG_INTM2_MHL_SHIFT)
+
+#define SM5502_REG_ADC_SHIFT			0
+#define SM5502_REG_ADC_MASK			(0x1f << SM5502_REG_ADC_SHIFT)
+
+#define SM5502_REG_TIMING_SET1_KEY_PRESS_SHIFT	4
+#define SM5502_REG_TIMING_SET1_KEY_PRESS_MASK	(0xf << SM5502_REG_TIMING_SET1_KEY_PRESS_SHIFT)
+#define TIMING_KEY_PRESS_100MS			0x0
+#define TIMING_KEY_PRESS_200MS			0x1
+#define TIMING_KEY_PRESS_300MS			0x2
+#define TIMING_KEY_PRESS_400MS			0x3
+#define TIMING_KEY_PRESS_500MS			0x4
+#define TIMING_KEY_PRESS_600MS			0x5
+#define TIMING_KEY_PRESS_700MS			0x6
+#define TIMING_KEY_PRESS_800MS			0x7
+#define TIMING_KEY_PRESS_900MS			0x8
+#define TIMING_KEY_PRESS_1000MS			0x9
+#define SM5502_REG_TIMING_SET1_ADC_DET_SHIFT	0
+#define SM5502_REG_TIMING_SET1_ADC_DET_MASK	(0xf << SM5502_REG_TIMING_SET1_ADC_DET_SHIFT)
+#define TIMING_ADC_DET_50MS			0x0
+#define TIMING_ADC_DET_100MS			0x1
+#define TIMING_ADC_DET_150MS			0x2
+#define TIMING_ADC_DET_200MS			0x3
+#define TIMING_ADC_DET_300MS			0x4
+#define TIMING_ADC_DET_400MS			0x5
+#define TIMING_ADC_DET_500MS			0x6
+#define TIMING_ADC_DET_600MS			0x7
+#define TIMING_ADC_DET_700MS			0x8
+#define TIMING_ADC_DET_800MS			0x9
+#define TIMING_ADC_DET_900MS			0xA
+#define TIMING_ADC_DET_1000MS			0xB
+
+#define SM5502_REG_TIMING_SET2_SW_WAIT_SHIFT	4
+#define SM5502_REG_TIMING_SET2_SW_WAIT_MASK	(0xf << SM5502_REG_TIMING_SET2_SW_WAIT_SHIFT)
+#define TIMING_SW_WAIT_10MS			0x0
+#define TIMING_SW_WAIT_30MS			0x1
+#define TIMING_SW_WAIT_50MS			0x2
+#define TIMING_SW_WAIT_70MS			0x3
+#define TIMING_SW_WAIT_90MS			0x4
+#define TIMING_SW_WAIT_110MS			0x5
+#define TIMING_SW_WAIT_130MS			0x6
+#define TIMING_SW_WAIT_150MS			0x7
+#define TIMING_SW_WAIT_170MS			0x8
+#define TIMING_SW_WAIT_190MS			0x9
+#define TIMING_SW_WAIT_210MS			0xA
+#define SM5502_REG_TIMING_SET2_LONG_KEY_SHIFT	0
+#define SM5502_REG_TIMING_SET2_LONG_KEY_MASK	(0xf << SM5502_REG_TIMING_SET2_LONG_KEY_SHIFT)
+#define TIMING_LONG_KEY_300MS			0x0
+#define TIMING_LONG_KEY_400MS			0x1
+#define TIMING_LONG_KEY_500MS			0x2
+#define TIMING_LONG_KEY_600MS			0x3
+#define TIMING_LONG_KEY_700MS			0x4
+#define TIMING_LONG_KEY_800MS			0x5
+#define TIMING_LONG_KEY_900MS			0x6
+#define TIMING_LONG_KEY_1000MS			0x7
+#define TIMING_LONG_KEY_1100MS			0x8
+#define TIMING_LONG_KEY_1200MS			0x9
+#define TIMING_LONG_KEY_1300MS			0xA
+#define TIMING_LONG_KEY_1400MS			0xB
+#define TIMING_LONG_KEY_1500MS			0xC
+
+#define SM5502_REG_DEV_TYPE1_AUDIO_TYPE1_SHIFT		0
+#define SM5502_REG_DEV_TYPE1_AUDIO_TYPE2_SHIFT		1
+#define SM5502_REG_DEV_TYPE1_USB_SDP_SHIFT		2
+#define SM5502_REG_DEV_TYPE1_UART_SHIFT			3
+#define SM5502_REG_DEV_TYPE1_CAR_KIT_CHARGER_SHIFT	4
+#define SM5502_REG_DEV_TYPE1_USB_CHG_SHIFT		5
+#define SM5502_REG_DEV_TYPE1_DEDICATED_CHG_SHIFT	6
+#define SM5502_REG_DEV_TYPE1_USB_OTG_SHIFT		7
+#define SM5502_REG_DEV_TYPE1_AUDIO_TYPE1_MASK		(0x1 << SM5502_REG_DEV_TYPE1_AUDIO_TYPE1_SHIFT)
+#define SM5502_REG_DEV_TYPE1_AUDIO_TYPE1__MASK		(0x1 << SM5502_REG_DEV_TYPE1_AUDIO_TYPE2_SHIFT)
+#define SM5502_REG_DEV_TYPE1_USB_SDP_MASK		(0x1 << SM5502_REG_DEV_TYPE1_USB_SDP_SHIFT)
+#define SM5502_REG_DEV_TYPE1_UART_MASK			(0x1 << SM5502_REG_DEV_TYPE1_UART_SHIFT)
+#define SM5502_REG_DEV_TYPE1_CAR_KIT_CHARGER_MASK	(0x1 << SM5502_REG_DEV_TYPE1_CAR_KIT_CHARGER_SHIFT)
+#define SM5502_REG_DEV_TYPE1_USB_CHG_MASK		(0x1 << SM5502_REG_DEV_TYPE1_USB_CHG_SHIFT)
+#define SM5502_REG_DEV_TYPE1_DEDICATED_CHG_MASK		(0x1 << SM5502_REG_DEV_TYPE1_DEDICATED_CHG_SHIFT)
+#define SM5502_REG_DEV_TYPE1_USB_OTG_MASK		(0x1 << SM5502_REG_DEV_TYPE1_USB_OTG_SHIFT)
+
+#define SM5502_REG_DEV_TYPE2_JIG_USB_ON_SHIFT		0
+#define SM5502_REG_DEV_TYPE2_JIG_USB_OFF_SHIFT		1
+#define SM5502_REG_DEV_TYPE2_JIG_UART_ON_SHIFT		2
+#define SM5502_REG_DEV_TYPE2_JIG_UART_OFF_SHIFT		3
+#define SM5502_REG_DEV_TYPE2_PPD_SHIFT			4
+#define SM5502_REG_DEV_TYPE2_TTY_SHIFT			5
+#define SM5502_REG_DEV_TYPE2_AV_CABLE_SHIFT		6
+#define SM5502_REG_DEV_TYPE2_JIG_USB_ON_MASK		(0x1 << SM5502_REG_DEV_TYPE2_JIG_USB_ON_SHIFT)
+#define SM5502_REG_DEV_TYPE2_JIG_USB_OFF_MASK		(0x1 << SM5502_REG_DEV_TYPE2_JIG_USB_OFF_SHIFT)
+#define SM5502_REG_DEV_TYPE2_JIG_UART_ON_MASK		(0x1 << SM5502_REG_DEV_TYPE2_JIG_UART_ON_SHIFT)
+#define SM5502_REG_DEV_TYPE2_JIG_UART_OFF_MASK		(0x1 << SM5502_REG_DEV_TYPE2_JIG_UART_OFF_SHIFT)
+#define SM5502_REG_DEV_TYPE2_PPD_MASK			(0x1 << SM5502_REG_DEV_TYPE2_PPD_SHIFT)
+#define SM5502_REG_DEV_TYPE2_TTY_MASK			(0x1 << SM5502_REG_DEV_TYPE2_TTY_SHIFT)
+#define SM5502_REG_DEV_TYPE2_AV_CABLE_MASK		(0x1 << SM5502_REG_DEV_TYPE2_AV_CABLE_SHIFT)
+
+#define SM5502_REG_MANUAL_SW1_VBUSIN_SHIFT	0
+#define SM5502_REG_MANUAL_SW1_DP_SHIFT		2
+#define SM5502_REG_MANUAL_SW1_DM_SHIFT		5
+#define SM5502_REG_MANUAL_SW1_VBUSIN_MASK	(0x3 << SM5502_REG_MANUAL_SW1_VBUSIN_SHIFT)
+#define SM5502_REG_MANUAL_SW1_DP_MASK		(0x7 << SM5502_REG_MANUAL_SW1_DP_SHIFT)
+#define SM5502_REG_MANUAL_SW1_DM_MASK		(0x7 << SM5502_REG_MANUAL_SW1_DM_SHIFT)
+#define VBUSIN_SWITCH_OPEN			0x0
+#define VBUSIN_SWITCH_VBUSOUT			0x1
+#define VBUSIN_SWITCH_MIC			0x2
+#define VBUSIN_SWITCH_VBUSOUT_WITH_USB		0x3
+#define DM_DP_CON_SWITCH_OPEN			0x0
+#define DM_DP_CON_SWITCH_USB			0x1
+#define DM_DP_CON_SWITCH_AUDIO			0x2
+#define DM_DP_CON_SWITCH_UART			0x3
+#define DM_DP_SWITCH_OPEN			((DM_DP_CON_SWITCH_OPEN <<SM5502_REG_MANUAL_SW1_DP_SHIFT) \
+						| (DM_DP_CON_SWITCH_OPEN <<SM5502_REG_MANUAL_SW1_DM_SHIFT))
+#define DM_DP_SWITCH_USB			((DM_DP_CON_SWITCH_USB <<SM5502_REG_MANUAL_SW1_DP_SHIFT) \
+						| (DM_DP_CON_SWITCH_USB <<SM5502_REG_MANUAL_SW1_DM_SHIFT))
+#define DM_DP_SWITCH_AUDIO			((DM_DP_CON_SWITCH_AUDIO <<SM5502_REG_MANUAL_SW1_DP_SHIFT) \
+						| (DM_DP_CON_SWITCH_AUDIO <<SM5502_REG_MANUAL_SW1_DM_SHIFT))
+#define DM_DP_SWITCH_UART			((DM_DP_CON_SWITCH_UART <<SM5502_REG_MANUAL_SW1_DP_SHIFT) \
+						| (DM_DP_CON_SWITCH_UART <<SM5502_REG_MANUAL_SW1_DM_SHIFT))
+
+/* SM5502 Interrupts */
+enum sm5502_irq {
+	/* INT1 */
+	SM5502_IRQ_INT1_ATTACH,
+	SM5502_IRQ_INT1_DETACH,
+	SM5502_IRQ_INT1_KP,
+	SM5502_IRQ_INT1_LKP,
+	SM5502_IRQ_INT1_LKR,
+	SM5502_IRQ_INT1_OVP_EVENT,
+	SM5502_IRQ_INT1_OCP_EVENT,
+	SM5502_IRQ_INT1_OVP_OCP_DIS,
+
+	/* INT2 */
+	SM5502_IRQ_INT2_VBUS_DET,
+	SM5502_IRQ_INT2_REV_ACCE,
+	SM5502_IRQ_INT2_ADC_CHG,
+	SM5502_IRQ_INT2_STUCK_KEY,
+	SM5502_IRQ_INT2_STUCK_KEY_RCV,
+	SM5502_IRQ_INT2_MHL,
+
+	SM5502_IRQ_NUM,
+};
+
+#define SM5502_IRQ_INT1_ATTACH_MASK		BIT(0)
+#define SM5502_IRQ_INT1_DETACH_MASK		BIT(1)
+#define SM5502_IRQ_INT1_KP_MASK			BIT(2)
+#define SM5502_IRQ_INT1_LKP_MASK		BIT(3)
+#define SM5502_IRQ_INT1_LKR_MASK		BIT(4)
+#define SM5502_IRQ_INT1_OVP_EVENT_MASK		BIT(5)
+#define SM5502_IRQ_INT1_OCP_EVENT_MASK		BIT(6)
+#define SM5502_IRQ_INT1_OVP_OCP_DIS_MASK	BIT(7)
+#define SM5502_IRQ_INT2_VBUS_DET_MASK		BIT(0)
+#define SM5502_IRQ_INT2_REV_ACCE_MASK		BIT(1)
+#define SM5502_IRQ_INT2_ADC_CHG_MASK		BIT(2)
+#define SM5502_IRQ_INT2_STUCK_KEY_MASK		BIT(3)
+#define SM5502_IRQ_INT2_STUCK_KEY_RCV_MASK	BIT(4)
+#define SM5502_IRQ_INT2_MHL_MASK		BIT(5)
+
+#endif /*  __LINUX_EXTCON_SM5502_H */
diff --git a/drivers/firewire/core-cdev.c b/drivers/firewire/core-cdev.c
index 5d997a33907e..2a3973a7c441 100644
--- a/drivers/firewire/core-cdev.c
+++ b/drivers/firewire/core-cdev.c
@@ -1637,8 +1637,7 @@ static int dispatch_ioctl(struct client *client,
 	    _IOC_SIZE(cmd) > sizeof(buffer))
 		return -ENOTTY;
 
-	if (_IOC_DIR(cmd) == _IOC_READ)
-		memset(&buffer, 0, _IOC_SIZE(cmd));
+	memset(&buffer, 0, sizeof(buffer));
 
 	if (_IOC_DIR(cmd) & _IOC_WRITE)
 		if (copy_from_user(&buffer, arg, _IOC_SIZE(cmd)))
diff --git a/drivers/firewire/core-device.c b/drivers/firewire/core-device.c
index 2c6d5e118ac1..f9e3aee6a211 100644
--- a/drivers/firewire/core-device.c
+++ b/drivers/firewire/core-device.c
@@ -115,6 +115,9 @@ static int textual_leaf_to_string(const u32 *block, char *buf, size_t size)
  *
  * The string is taken from a minimal ASCII text descriptor leaf after
  * the immediate entry with @key.  The string is zero-terminated.
+ * An overlong string is silently truncated such that it and the
+ * zero byte fit into @size.
+ *
  * Returns strlen(buf) or a negative error code.
  */
 int fw_csr_string(const u32 *directory, int key, char *buf, size_t size)
diff --git a/drivers/firewire/ohci.c b/drivers/firewire/ohci.c
index a66a3217f1d9..aff9018d0658 100644
--- a/drivers/firewire/ohci.c
+++ b/drivers/firewire/ohci.c
@@ -689,8 +689,7 @@ static void ar_context_release(struct ar_context *ctx)
 {
 	unsigned int i;
 
-	if (ctx->buffer)
-		vm_unmap_ram(ctx->buffer, AR_BUFFERS + AR_WRAPAROUND_PAGES);
+	vunmap(ctx->buffer);
 
 	for (i = 0; i < AR_BUFFERS; i++)
 		if (ctx->pages[i]) {
@@ -1018,8 +1017,7 @@ static int ar_context_init(struct ar_context *ctx, struct fw_ohci *ohci,
 		pages[i]              = ctx->pages[i];
 	for (i = 0; i < AR_WRAPAROUND_PAGES; i++)
 		pages[AR_BUFFERS + i] = ctx->pages[i];
-	ctx->buffer = vm_map_ram(pages, AR_BUFFERS + AR_WRAPAROUND_PAGES,
-				 -1, PAGE_KERNEL);
+	ctx->buffer = vmap(pages, ARRAY_SIZE(pages), VM_MAP, PAGE_KERNEL);
 	if (!ctx->buffer)
 		goto out_of_memory;
 
diff --git a/drivers/firewire/sbp2.c b/drivers/firewire/sbp2.c
index 7aef911fdc71..64ac8f8f5098 100644
--- a/drivers/firewire/sbp2.c
+++ b/drivers/firewire/sbp2.c
@@ -174,6 +174,7 @@ struct sbp2_target {
 	unsigned int mgt_orb_timeout;
 	unsigned int max_payload;
 
+	spinlock_t lock;
 	int dont_block;	/* counter for each logical unit */
 	int blocked;	/* ditto */
 };
@@ -270,6 +271,7 @@ struct sbp2_orb {
 	dma_addr_t request_bus;
 	int rcode;
 	void (*callback)(struct sbp2_orb * orb, struct sbp2_status * status);
+	struct sbp2_logical_unit *lu;
 	struct list_head link;
 };
 
@@ -321,7 +323,6 @@ struct sbp2_command_orb {
 		u8 command_block[SBP2_MAX_CDB_SIZE];
 	} request;
 	struct scsi_cmnd *cmd;
-	struct sbp2_logical_unit *lu;
 
 	struct sbp2_pointer page_table[SG_ALL] __attribute__((aligned(8)));
 	dma_addr_t page_table_bus;
@@ -444,7 +445,7 @@ static void sbp2_status_write(struct fw_card *card, struct fw_request *request,
 	}
 
 	/* Lookup the orb corresponding to this status write. */
-	spin_lock_irqsave(&card->lock, flags);
+	spin_lock_irqsave(&lu->tgt->lock, flags);
 	list_for_each_entry(orb, &lu->orb_list, link) {
 		if (STATUS_GET_ORB_HIGH(status) == 0 &&
 		    STATUS_GET_ORB_LOW(status) == orb->request_bus) {
@@ -453,7 +454,7 @@ static void sbp2_status_write(struct fw_card *card, struct fw_request *request,
 			break;
 		}
 	}
-	spin_unlock_irqrestore(&card->lock, flags);
+	spin_unlock_irqrestore(&lu->tgt->lock, flags);
 
 	if (&orb->link != &lu->orb_list) {
 		orb->callback(orb, &status);
@@ -480,18 +481,18 @@ static void complete_transaction(struct fw_card *card, int rcode,
 	 * been set and only does the cleanup if the transaction
 	 * failed and we didn't already get a status write.
 	 */
-	spin_lock_irqsave(&card->lock, flags);
+	spin_lock_irqsave(&orb->lu->tgt->lock, flags);
 
 	if (orb->rcode == -1)
 		orb->rcode = rcode;
 	if (orb->rcode != RCODE_COMPLETE) {
 		list_del(&orb->link);
-		spin_unlock_irqrestore(&card->lock, flags);
+		spin_unlock_irqrestore(&orb->lu->tgt->lock, flags);
 
 		orb->callback(orb, NULL);
 		kref_put(&orb->kref, free_orb); /* orb callback reference */
 	} else {
-		spin_unlock_irqrestore(&card->lock, flags);
+		spin_unlock_irqrestore(&orb->lu->tgt->lock, flags);
 	}
 
 	kref_put(&orb->kref, free_orb); /* transaction callback reference */
@@ -507,9 +508,10 @@ static void sbp2_send_orb(struct sbp2_orb *orb, struct sbp2_logical_unit *lu,
 	orb_pointer.high = 0;
 	orb_pointer.low = cpu_to_be32(orb->request_bus);
 
-	spin_lock_irqsave(&device->card->lock, flags);
+	orb->lu = lu;
+	spin_lock_irqsave(&lu->tgt->lock, flags);
 	list_add_tail(&orb->link, &lu->orb_list);
-	spin_unlock_irqrestore(&device->card->lock, flags);
+	spin_unlock_irqrestore(&lu->tgt->lock, flags);
 
 	kref_get(&orb->kref); /* transaction callback reference */
 	kref_get(&orb->kref); /* orb callback reference */
@@ -524,13 +526,12 @@ static int sbp2_cancel_orbs(struct sbp2_logical_unit *lu)
 	struct fw_device *device = target_parent_device(lu->tgt);
 	struct sbp2_orb *orb, *next;
 	struct list_head list;
-	unsigned long flags;
 	int retval = -ENOENT;
 
 	INIT_LIST_HEAD(&list);
-	spin_lock_irqsave(&device->card->lock, flags);
+	spin_lock_irq(&lu->tgt->lock);
 	list_splice_init(&lu->orb_list, &list);
-	spin_unlock_irqrestore(&device->card->lock, flags);
+	spin_unlock_irq(&lu->tgt->lock);
 
 	list_for_each_entry_safe(orb, next, &list, link) {
 		retval = 0;
@@ -687,16 +688,11 @@ static void sbp2_agent_reset_no_wait(struct sbp2_logical_unit *lu)
 			&d, 4, complete_agent_reset_write_no_wait, t);
 }
 
-static inline void sbp2_allow_block(struct sbp2_logical_unit *lu)
+static inline void sbp2_allow_block(struct sbp2_target *tgt)
 {
-	/*
-	 * We may access dont_block without taking card->lock here:
-	 * All callers of sbp2_allow_block() and all callers of sbp2_unblock()
-	 * are currently serialized against each other.
-	 * And a wrong result in sbp2_conditionally_block()'s access of
-	 * dont_block is rather harmless, it simply misses its first chance.
-	 */
-	--lu->tgt->dont_block;
+	spin_lock_irq(&tgt->lock);
+	--tgt->dont_block;
+	spin_unlock_irq(&tgt->lock);
 }
 
 /*
@@ -705,7 +701,7 @@ static inline void sbp2_allow_block(struct sbp2_logical_unit *lu)
  *     logical units have been finished (indicated by dont_block == 0).
  *   - lu->generation is stale.
  *
- * Note, scsi_block_requests() must be called while holding card->lock,
+ * Note, scsi_block_requests() must be called while holding tgt->lock,
  * otherwise it might foil sbp2_[conditionally_]unblock()'s attempt to
  * unblock the target.
  */
@@ -717,20 +713,20 @@ static void sbp2_conditionally_block(struct sbp2_logical_unit *lu)
 		container_of((void *)tgt, struct Scsi_Host, hostdata[0]);
 	unsigned long flags;
 
-	spin_lock_irqsave(&card->lock, flags);
+	spin_lock_irqsave(&tgt->lock, flags);
 	if (!tgt->dont_block && !lu->blocked &&
 	    lu->generation != card->generation) {
 		lu->blocked = true;
 		if (++tgt->blocked == 1)
 			scsi_block_requests(shost);
 	}
-	spin_unlock_irqrestore(&card->lock, flags);
+	spin_unlock_irqrestore(&tgt->lock, flags);
 }
 
 /*
  * Unblocks lu->tgt as soon as all its logical units can be unblocked.
  * Note, it is harmless to run scsi_unblock_requests() outside the
- * card->lock protected section.  On the other hand, running it inside
+ * tgt->lock protected section.  On the other hand, running it inside
  * the section might clash with shost->host_lock.
  */
 static void sbp2_conditionally_unblock(struct sbp2_logical_unit *lu)
@@ -739,15 +735,14 @@ static void sbp2_conditionally_unblock(struct sbp2_logical_unit *lu)
 	struct fw_card *card = target_parent_device(tgt)->card;
 	struct Scsi_Host *shost =
 		container_of((void *)tgt, struct Scsi_Host, hostdata[0]);
-	unsigned long flags;
 	bool unblock = false;
 
-	spin_lock_irqsave(&card->lock, flags);
+	spin_lock_irq(&tgt->lock);
 	if (lu->blocked && lu->generation == card->generation) {
 		lu->blocked = false;
 		unblock = --tgt->blocked == 0;
 	}
-	spin_unlock_irqrestore(&card->lock, flags);
+	spin_unlock_irq(&tgt->lock);
 
 	if (unblock)
 		scsi_unblock_requests(shost);
@@ -756,19 +751,17 @@ static void sbp2_conditionally_unblock(struct sbp2_logical_unit *lu)
 /*
  * Prevents future blocking of tgt and unblocks it.
  * Note, it is harmless to run scsi_unblock_requests() outside the
- * card->lock protected section.  On the other hand, running it inside
+ * tgt->lock protected section.  On the other hand, running it inside
  * the section might clash with shost->host_lock.
  */
 static void sbp2_unblock(struct sbp2_target *tgt)
 {
-	struct fw_card *card = target_parent_device(tgt)->card;
 	struct Scsi_Host *shost =
 		container_of((void *)tgt, struct Scsi_Host, hostdata[0]);
-	unsigned long flags;
 
-	spin_lock_irqsave(&card->lock, flags);
+	spin_lock_irq(&tgt->lock);
 	++tgt->dont_block;
-	spin_unlock_irqrestore(&card->lock, flags);
+	spin_unlock_irq(&tgt->lock);
 
 	scsi_unblock_requests(shost);
 }
@@ -904,7 +897,7 @@ static void sbp2_login(struct work_struct *work)
 	/* No error during __scsi_add_device() */
 	lu->has_sdev = true;
 	scsi_device_put(sdev);
-	sbp2_allow_block(lu);
+	sbp2_allow_block(tgt);
 
 	return;
 
@@ -1163,6 +1156,7 @@ static int sbp2_probe(struct fw_unit *unit, const struct ieee1394_device_id *id)
 	dev_set_drvdata(&unit->device, tgt);
 	tgt->unit = unit;
 	INIT_LIST_HEAD(&tgt->lu_list);
+	spin_lock_init(&tgt->lock);
 	tgt->guid = (u64)device->config_rom[3] << 32 | device->config_rom[4];
 
 	if (fw_device_enable_phys_dma(device) < 0)
@@ -1359,12 +1353,12 @@ static void complete_command_orb(struct sbp2_orb *base_orb,
 {
 	struct sbp2_command_orb *orb =
 		container_of(base_orb, struct sbp2_command_orb, base);
-	struct fw_device *device = target_parent_device(orb->lu->tgt);
+	struct fw_device *device = target_parent_device(base_orb->lu->tgt);
 	int result;
 
 	if (status != NULL) {
 		if (STATUS_GET_DEAD(*status))
-			sbp2_agent_reset_no_wait(orb->lu);
+			sbp2_agent_reset_no_wait(base_orb->lu);
 
 		switch (STATUS_GET_RESPONSE(*status)) {
 		case SBP2_STATUS_REQUEST_COMPLETE:
@@ -1390,7 +1384,7 @@ static void complete_command_orb(struct sbp2_orb *base_orb,
 		 * or when sending the write (less likely).
 		 */
 		result = DID_BUS_BUSY << 16;
-		sbp2_conditionally_block(orb->lu);
+		sbp2_conditionally_block(base_orb->lu);
 	}
 
 	dma_unmap_single(device->card->device, orb->base.request_bus,
@@ -1487,7 +1481,6 @@ static int sbp2_scsi_queuecommand(struct Scsi_Host *shost,
 	/* Initialize rcode to something not RCODE_COMPLETE. */
 	orb->base.rcode = -1;
 	kref_init(&orb->base.kref);
-	orb->lu = lu;
 	orb->cmd = cmd;
 	orb->request.next.high = cpu_to_be32(SBP2_ORB_NULL);
 	orb->request.misc = cpu_to_be32(
diff --git a/drivers/firmware/dcdbas.c b/drivers/firmware/dcdbas.c
index 7160c43c59fc..829eec8959f2 100644
--- a/drivers/firmware/dcdbas.c
+++ b/drivers/firmware/dcdbas.c
@@ -578,7 +578,6 @@ static int dcdbas_remove(struct platform_device *dev)
 static struct platform_driver dcdbas_driver = {
 	.driver		= {
 		.name	= DRIVER_NAME,
-		.owner	= THIS_MODULE,
 	},
 	.probe		= dcdbas_probe,
 	.remove		= dcdbas_remove,
diff --git a/drivers/firmware/dmi_scan.c b/drivers/firmware/dmi_scan.c
index 17afc51f3054..c5f7b4e9eb6c 100644
--- a/drivers/firmware/dmi_scan.c
+++ b/drivers/firmware/dmi_scan.c
@@ -93,6 +93,12 @@ static void dmi_table(u8 *buf, int len, int num,
 		const struct dmi_header *dm = (const struct dmi_header *)data;
 
 		/*
+		 * 7.45 End-of-Table (Type 127) [SMBIOS reference spec v3.0.0]
+		 */
+		if (dm->type == DMI_ENTRY_END_OF_TABLE)
+			break;
+
+		/*
 		 *  We want to know the total length (formatted area and
 		 *  strings) before decoding to make sure we won't run off the
 		 *  table in dmi_decode or dmi_string
@@ -107,7 +113,7 @@ static void dmi_table(u8 *buf, int len, int num,
 	}
 }
 
-static u32 dmi_base;
+static phys_addr_t dmi_base;
 static u16 dmi_len;
 static u16 dmi_num;
 
@@ -467,7 +473,7 @@ static int __init dmi_present(const u8 *buf)
 
 	if (memcmp(buf, "_SM_", 4) == 0 &&
 	    buf[5] < 32 && dmi_checksum(buf, buf[5])) {
-		smbios_ver = (buf[6] << 8) + buf[7];
+		smbios_ver = get_unaligned_be16(buf + 6);
 
 		/* Some BIOS report weird SMBIOS version, fix that up */
 		switch (smbios_ver) {
@@ -489,10 +495,9 @@ static int __init dmi_present(const u8 *buf)
 	buf += 16;
 
 	if (memcmp(buf, "_DMI_", 5) == 0 && dmi_checksum(buf, 15)) {
-		dmi_num = (buf[13] << 8) | buf[12];
-		dmi_len = (buf[7] << 8) | buf[6];
-		dmi_base = (buf[11] << 24) | (buf[10] << 16) |
-			(buf[9] << 8) | buf[8];
+		dmi_num = get_unaligned_le16(buf + 12);
+		dmi_len = get_unaligned_le16(buf + 6);
+		dmi_base = get_unaligned_le32(buf + 8);
 
 		if (dmi_walk_early(dmi_decode) == 0) {
 			if (smbios_ver) {
@@ -514,12 +519,72 @@ static int __init dmi_present(const u8 *buf)
 	return 1;
 }
 
+/*
+ * Check for the SMBIOS 3.0 64-bit entry point signature. Unlike the legacy
+ * 32-bit entry point, there is no embedded DMI header (_DMI_) in here.
+ */
+static int __init dmi_smbios3_present(const u8 *buf)
+{
+	if (memcmp(buf, "_SM3_", 5) == 0 &&
+	    buf[6] < 32 && dmi_checksum(buf, buf[6])) {
+		dmi_ver = get_unaligned_be16(buf + 7);
+		dmi_len = get_unaligned_le32(buf + 12);
+		dmi_base = get_unaligned_le64(buf + 16);
+
+		/*
+		 * The 64-bit SMBIOS 3.0 entry point no longer has a field
+		 * containing the number of structures present in the table.
+		 * Instead, it defines the table size as a maximum size, and
+		 * relies on the end-of-table structure type (#127) to be used
+		 * to signal the end of the table.
+		 * So let's define dmi_num as an upper bound as well: each
+		 * structure has a 4 byte header, so dmi_len / 4 is an upper
+		 * bound for the number of structures in the table.
+		 */
+		dmi_num = dmi_len / 4;
+
+		if (dmi_walk_early(dmi_decode) == 0) {
+			pr_info("SMBIOS %d.%d present.\n",
+				dmi_ver >> 8, dmi_ver & 0xFF);
+			dmi_format_ids(dmi_ids_string, sizeof(dmi_ids_string));
+			pr_debug("DMI: %s\n", dmi_ids_string);
+			return 0;
+		}
+	}
+	return 1;
+}
+
 void __init dmi_scan_machine(void)
 {
 	char __iomem *p, *q;
 	char buf[32];
 
 	if (efi_enabled(EFI_CONFIG_TABLES)) {
+		/*
+		 * According to the DMTF SMBIOS reference spec v3.0.0, it is
+		 * allowed to define both the 64-bit entry point (smbios3) and
+		 * the 32-bit entry point (smbios), in which case they should
+		 * either both point to the same SMBIOS structure table, or the
+		 * table pointed to by the 64-bit entry point should contain a
+		 * superset of the table contents pointed to by the 32-bit entry
+		 * point (section 5.2)
+		 * This implies that the 64-bit entry point should have
+		 * precedence if it is defined and supported by the OS. If we
+		 * have the 64-bit entry point, but fail to decode it, fall
+		 * back to the legacy one (if available)
+		 */
+		if (efi.smbios3 != EFI_INVALID_TABLE_ADDR) {
+			p = dmi_early_remap(efi.smbios3, 32);
+			if (p == NULL)
+				goto error;
+			memcpy_fromio(buf, p, 32);
+			dmi_early_unmap(p, 32);
+
+			if (!dmi_smbios3_present(buf)) {
+				dmi_available = 1;
+				goto out;
+			}
+		}
 		if (efi.smbios == EFI_INVALID_TABLE_ADDR)
 			goto error;
 
@@ -552,7 +617,7 @@ void __init dmi_scan_machine(void)
 		memset(buf, 0, 16);
 		for (q = p; q < p + 0x10000; q += 16) {
 			memcpy_fromio(buf + 16, q, 16);
-			if (!dmi_present(buf)) {
+			if (!dmi_smbios3_present(buf) || !dmi_present(buf)) {
 				dmi_available = 1;
 				dmi_early_unmap(p, 0x10000);
 				goto out;
diff --git a/drivers/firmware/efi/Makefile b/drivers/firmware/efi/Makefile
index aef6a95adef5..d8be608a9f3b 100644
--- a/drivers/firmware/efi/Makefile
+++ b/drivers/firmware/efi/Makefile
@@ -7,4 +7,4 @@ obj-$(CONFIG_EFI_VARS_PSTORE)		+= efi-pstore.o
 obj-$(CONFIG_UEFI_CPER)			+= cper.o
 obj-$(CONFIG_EFI_RUNTIME_MAP)		+= runtime-map.o
 obj-$(CONFIG_EFI_RUNTIME_WRAPPERS)	+= runtime-wrappers.o
-obj-$(CONFIG_EFI_ARM_STUB)		+= libstub/
+obj-$(CONFIG_EFI_STUB)			+= libstub/
diff --git a/drivers/firmware/efi/cper.c b/drivers/firmware/efi/cper.c
index 5b53d6183b6b..4fd9961d552e 100644
--- a/drivers/firmware/efi/cper.c
+++ b/drivers/firmware/efi/cper.c
@@ -294,7 +294,7 @@ void cper_mem_err_pack(const struct cper_sec_mem_err *mem,
 const char *cper_mem_err_unpack(struct trace_seq *p,
 				struct cper_mem_err_compact *cmem)
 {
-	const char *ret = p->buffer + p->len;
+	const char *ret = trace_seq_buffer_ptr(p);
 
 	if (cper_mem_err_location(cmem, rcd_decode_str))
 		trace_seq_printf(p, "%s", rcd_decode_str);
diff --git a/drivers/firmware/efi/efi.c b/drivers/firmware/efi/efi.c
index 64ecbb501c50..9035c1b74d58 100644
--- a/drivers/firmware/efi/efi.c
+++ b/drivers/firmware/efi/efi.c
@@ -30,6 +30,7 @@ struct efi __read_mostly efi = {
 	.acpi       = EFI_INVALID_TABLE_ADDR,
 	.acpi20     = EFI_INVALID_TABLE_ADDR,
 	.smbios     = EFI_INVALID_TABLE_ADDR,
+	.smbios3    = EFI_INVALID_TABLE_ADDR,
 	.sal_systab = EFI_INVALID_TABLE_ADDR,
 	.boot_info  = EFI_INVALID_TABLE_ADDR,
 	.hcdp       = EFI_INVALID_TABLE_ADDR,
@@ -41,6 +42,28 @@ struct efi __read_mostly efi = {
 };
 EXPORT_SYMBOL(efi);
 
+static bool disable_runtime;
+static int __init setup_noefi(char *arg)
+{
+	disable_runtime = true;
+	return 0;
+}
+early_param("noefi", setup_noefi);
+
+bool efi_runtime_disabled(void)
+{
+	return disable_runtime;
+}
+
+static int __init parse_efi_cmdline(char *str)
+{
+	if (parse_option_str(str, "noruntime"))
+		disable_runtime = true;
+
+	return 0;
+}
+early_param("efi", parse_efi_cmdline);
+
 static struct kobject *efi_kobj;
 static struct kobject *efivars_kobj;
 
@@ -64,6 +87,8 @@ static ssize_t systab_show(struct kobject *kobj,
 		str += sprintf(str, "ACPI=0x%lx\n", efi.acpi);
 	if (efi.smbios != EFI_INVALID_TABLE_ADDR)
 		str += sprintf(str, "SMBIOS=0x%lx\n", efi.smbios);
+	if (efi.smbios3 != EFI_INVALID_TABLE_ADDR)
+		str += sprintf(str, "SMBIOS3=0x%lx\n", efi.smbios3);
 	if (efi.hcdp != EFI_INVALID_TABLE_ADDR)
 		str += sprintf(str, "HCDP=0x%lx\n", efi.hcdp);
 	if (efi.boot_info != EFI_INVALID_TABLE_ADDR)
@@ -238,6 +263,7 @@ static __initdata efi_config_table_type_t common_tables[] = {
 	{MPS_TABLE_GUID, "MPS", &efi.mps},
 	{SAL_SYSTEM_TABLE_GUID, "SALsystab", &efi.sal_systab},
 	{SMBIOS_TABLE_GUID, "SMBIOS", &efi.smbios},
+	{SMBIOS3_TABLE_GUID, "SMBIOS 3.0", &efi.smbios3},
 	{UGA_IO_PROTOCOL_GUID, "UGA", &efi.uga},
 	{NULL_GUID, NULL, NULL},
 };
@@ -423,3 +449,60 @@ int __init efi_get_fdt_params(struct efi_fdt_params *params, int verbose)
 	return ret;
 }
 #endif /* CONFIG_EFI_PARAMS_FROM_FDT */
+
+static __initdata char memory_type_name[][20] = {
+	"Reserved",
+	"Loader Code",
+	"Loader Data",
+	"Boot Code",
+	"Boot Data",
+	"Runtime Code",
+	"Runtime Data",
+	"Conventional Memory",
+	"Unusable Memory",
+	"ACPI Reclaim Memory",
+	"ACPI Memory NVS",
+	"Memory Mapped I/O",
+	"MMIO Port Space",
+	"PAL Code"
+};
+
+char * __init efi_md_typeattr_format(char *buf, size_t size,
+				     const efi_memory_desc_t *md)
+{
+	char *pos;
+	int type_len;
+	u64 attr;
+
+	pos = buf;
+	if (md->type >= ARRAY_SIZE(memory_type_name))
+		type_len = snprintf(pos, size, "[type=%u", md->type);
+	else
+		type_len = snprintf(pos, size, "[%-*s",
+				    (int)(sizeof(memory_type_name[0]) - 1),
+				    memory_type_name[md->type]);
+	if (type_len >= size)
+		return buf;
+
+	pos += type_len;
+	size -= type_len;
+
+	attr = md->attribute;
+	if (attr & ~(EFI_MEMORY_UC | EFI_MEMORY_WC | EFI_MEMORY_WT |
+		     EFI_MEMORY_WB | EFI_MEMORY_UCE | EFI_MEMORY_WP |
+		     EFI_MEMORY_RP | EFI_MEMORY_XP | EFI_MEMORY_RUNTIME))
+		snprintf(pos, size, "|attr=0x%016llx]",
+			 (unsigned long long)attr);
+	else
+		snprintf(pos, size, "|%3s|%2s|%2s|%2s|%3s|%2s|%2s|%2s|%2s]",
+			 attr & EFI_MEMORY_RUNTIME ? "RUN" : "",
+			 attr & EFI_MEMORY_XP      ? "XP"  : "",
+			 attr & EFI_MEMORY_RP      ? "RP"  : "",
+			 attr & EFI_MEMORY_WP      ? "WP"  : "",
+			 attr & EFI_MEMORY_UCE     ? "UCE" : "",
+			 attr & EFI_MEMORY_WB      ? "WB"  : "",
+			 attr & EFI_MEMORY_WT      ? "WT"  : "",
+			 attr & EFI_MEMORY_WC      ? "WC"  : "",
+			 attr & EFI_MEMORY_UC      ? "UC"  : "");
+	return buf;
+}
diff --git a/drivers/firmware/efi/libstub/arm-stub.c b/drivers/firmware/efi/libstub/arm-stub.c
index 480339b6b110..eb48a1a1a576 100644
--- a/drivers/firmware/efi/libstub/arm-stub.c
+++ b/drivers/firmware/efi/libstub/arm-stub.c
@@ -226,6 +226,10 @@ unsigned long __init efi_entry(void *handle, efi_system_table_t *sys_table,
 		goto fail_free_image;
 	}
 
+	status = efi_parse_options(cmdline_ptr);
+	if (status != EFI_SUCCESS)
+		pr_efi_err(sys_table, "Failed to parse EFI cmdline options\n");
+
 	/*
 	 * Unauthenticated device tree data is a security hazard, so
 	 * ignore 'dtb=' unless UEFI Secure Boot is disabled.
@@ -243,9 +247,18 @@ unsigned long __init efi_entry(void *handle, efi_system_table_t *sys_table,
 			goto fail_free_cmdline;
 		}
 	}
-	if (!fdt_addr)
+
+	if (fdt_addr) {
+		pr_efi(sys_table, "Using DTB from command line\n");
+	} else {
 		/* Look for a device tree configuration table entry. */
 		fdt_addr = (uintptr_t)get_fdt(sys_table);
+		if (fdt_addr)
+			pr_efi(sys_table, "Using DTB from configuration table\n");
+	}
+
+	if (!fdt_addr)
+		pr_efi(sys_table, "Generating empty DTB\n");
 
 	status = handle_cmdline_files(sys_table, image, cmdline_ptr,
 				      "initrd=", dram_base + SZ_512M,
diff --git a/drivers/firmware/efi/libstub/efi-stub-helper.c b/drivers/firmware/efi/libstub/efi-stub-helper.c
index 32d5cca30f49..a920fec8fe88 100644
--- a/drivers/firmware/efi/libstub/efi-stub-helper.c
+++ b/drivers/firmware/efi/libstub/efi-stub-helper.c
@@ -15,8 +15,23 @@
 
 #include "efistub.h"
 
+/*
+ * Some firmware implementations have problems reading files in one go.
+ * A read chunk size of 1MB seems to work for most platforms.
+ *
+ * Unfortunately, reading files in chunks triggers *other* bugs on some
+ * platforms, so we provide a way to disable this workaround, which can
+ * be done by passing "efi=nochunk" on the EFI boot stub command line.
+ *
+ * If you experience issues with initrd images being corrupt it's worth
+ * trying efi=nochunk, but chunking is enabled by default because there
+ * are far more machines that require the workaround than those that
+ * break with it enabled.
+ */
 #define EFI_READ_CHUNK_SIZE	(1024 * 1024)
 
+static unsigned long __chunk_size = EFI_READ_CHUNK_SIZE;
+
 struct file_info {
 	efi_file_handle_t *handle;
 	u64 size;
@@ -281,6 +296,49 @@ void efi_free(efi_system_table_t *sys_table_arg, unsigned long size,
 	efi_call_early(free_pages, addr, nr_pages);
 }
 
+/*
+ * Parse the ASCII string 'cmdline' for EFI options, denoted by the efi=
+ * option, e.g. efi=nochunk.
+ *
+ * It should be noted that efi= is parsed in two very different
+ * environments, first in the early boot environment of the EFI boot
+ * stub, and subsequently during the kernel boot.
+ */
+efi_status_t efi_parse_options(char *cmdline)
+{
+	char *str;
+
+	/*
+	 * If no EFI parameters were specified on the cmdline we've got
+	 * nothing to do.
+	 */
+	str = strstr(cmdline, "efi=");
+	if (!str)
+		return EFI_SUCCESS;
+
+	/* Skip ahead to first argument */
+	str += strlen("efi=");
+
+	/*
+	 * Remember, because efi= is also used by the kernel we need to
+	 * skip over arguments we don't understand.
+	 */
+	while (*str) {
+		if (!strncmp(str, "nochunk", 7)) {
+			str += strlen("nochunk");
+			__chunk_size = -1UL;
+		}
+
+		/* Group words together, delimited by "," */
+		while (*str && *str != ',')
+			str++;
+
+		if (*str == ',')
+			str++;
+	}
+
+	return EFI_SUCCESS;
+}
 
 /*
  * Check the cmdline for a LILO-style file= arguments.
@@ -423,8 +481,8 @@ efi_status_t handle_cmdline_files(efi_system_table_t *sys_table_arg,
 			size = files[j].size;
 			while (size) {
 				unsigned long chunksize;
-				if (size > EFI_READ_CHUNK_SIZE)
-					chunksize = EFI_READ_CHUNK_SIZE;
+				if (size > __chunk_size)
+					chunksize = __chunk_size;
 				else
 					chunksize = size;
 
diff --git a/drivers/firmware/efi/runtime-wrappers.c b/drivers/firmware/efi/runtime-wrappers.c
index 10daa4bbb258..228bbf910461 100644
--- a/drivers/firmware/efi/runtime-wrappers.c
+++ b/drivers/firmware/efi/runtime-wrappers.c
@@ -14,11 +14,80 @@
  * This file is released under the GPLv2.
  */
 
+#include <linux/bug.h>
 #include <linux/efi.h>
-#include <linux/spinlock.h>             /* spinlock_t */
+#include <linux/mutex.h>
+#include <linux/spinlock.h>
 #include <asm/efi.h>
 
 /*
+ * According to section 7.1 of the UEFI spec, Runtime Services are not fully
+ * reentrant, and there are particular combinations of calls that need to be
+ * serialized. (source: UEFI Specification v2.4A)
+ *
+ * Table 31. Rules for Reentry Into Runtime Services
+ * +------------------------------------+-------------------------------+
+ * | If previous call is busy in	| Forbidden to call		|
+ * +------------------------------------+-------------------------------+
+ * | Any				| SetVirtualAddressMap()	|
+ * +------------------------------------+-------------------------------+
+ * | ConvertPointer()			| ConvertPointer()		|
+ * +------------------------------------+-------------------------------+
+ * | SetVariable()			| ResetSystem()			|
+ * | UpdateCapsule()			|				|
+ * | SetTime()				|				|
+ * | SetWakeupTime()			|				|
+ * | GetNextHighMonotonicCount()	|				|
+ * +------------------------------------+-------------------------------+
+ * | GetVariable()			| GetVariable()			|
+ * | GetNextVariableName()		| GetNextVariableName()		|
+ * | SetVariable()			| SetVariable()			|
+ * | QueryVariableInfo()		| QueryVariableInfo()		|
+ * | UpdateCapsule()			| UpdateCapsule()		|
+ * | QueryCapsuleCapabilities()		| QueryCapsuleCapabilities()	|
+ * | GetNextHighMonotonicCount()	| GetNextHighMonotonicCount()	|
+ * +------------------------------------+-------------------------------+
+ * | GetTime()				| GetTime()			|
+ * | SetTime()				| SetTime()			|
+ * | GetWakeupTime()			| GetWakeupTime()		|
+ * | SetWakeupTime()			| SetWakeupTime()		|
+ * +------------------------------------+-------------------------------+
+ *
+ * Due to the fact that the EFI pstore may write to the variable store in
+ * interrupt context, we need to use a spinlock for at least the groups that
+ * contain SetVariable() and QueryVariableInfo(). That leaves little else, as
+ * none of the remaining functions are actually ever called at runtime.
+ * So let's just use a single spinlock to serialize all Runtime Services calls.
+ */
+static DEFINE_SPINLOCK(efi_runtime_lock);
+
+/*
+ * Some runtime services calls can be reentrant under NMI, even if the table
+ * above says they are not. (source: UEFI Specification v2.4A)
+ *
+ * Table 32. Functions that may be called after Machine Check, INIT and NMI
+ * +----------------------------+------------------------------------------+
+ * | Function			| Called after Machine Check, INIT and NMI |
+ * +----------------------------+------------------------------------------+
+ * | GetTime()			| Yes, even if previously busy.		   |
+ * | GetVariable()		| Yes, even if previously busy		   |
+ * | GetNextVariableName()	| Yes, even if previously busy		   |
+ * | QueryVariableInfo()	| Yes, even if previously busy		   |
+ * | SetVariable()		| Yes, even if previously busy		   |
+ * | UpdateCapsule()		| Yes, even if previously busy		   |
+ * | QueryCapsuleCapabilities()	| Yes, even if previously busy		   |
+ * | ResetSystem()		| Yes, even if previously busy		   |
+ * +----------------------------+------------------------------------------+
+ *
+ * In order to prevent deadlocks under NMI, the wrappers for these functions
+ * may only grab the efi_runtime_lock or rtc_lock spinlocks if !efi_in_nmi().
+ * However, not all of the services listed are reachable through NMI code paths,
+ * so the the special handling as suggested by the UEFI spec is only implemented
+ * for QueryVariableInfo() and SetVariable(), as these can be reached in NMI
+ * context through efi_pstore_write().
+ */
+
+/*
  * As per commit ef68c8f87ed1 ("x86: Serialize EFI time accesses on rtc_lock"),
  * the EFI specification requires that callers of the time related runtime
  * functions serialize with other CMOS accesses in the kernel, as the EFI time
@@ -32,7 +101,9 @@ static efi_status_t virt_efi_get_time(efi_time_t *tm, efi_time_cap_t *tc)
 	efi_status_t status;
 
 	spin_lock_irqsave(&rtc_lock, flags);
+	spin_lock(&efi_runtime_lock);
 	status = efi_call_virt(get_time, tm, tc);
+	spin_unlock(&efi_runtime_lock);
 	spin_unlock_irqrestore(&rtc_lock, flags);
 	return status;
 }
@@ -43,7 +114,9 @@ static efi_status_t virt_efi_set_time(efi_time_t *tm)
 	efi_status_t status;
 
 	spin_lock_irqsave(&rtc_lock, flags);
+	spin_lock(&efi_runtime_lock);
 	status = efi_call_virt(set_time, tm);
+	spin_unlock(&efi_runtime_lock);
 	spin_unlock_irqrestore(&rtc_lock, flags);
 	return status;
 }
@@ -56,7 +129,9 @@ static efi_status_t virt_efi_get_wakeup_time(efi_bool_t *enabled,
 	efi_status_t status;
 
 	spin_lock_irqsave(&rtc_lock, flags);
+	spin_lock(&efi_runtime_lock);
 	status = efi_call_virt(get_wakeup_time, enabled, pending, tm);
+	spin_unlock(&efi_runtime_lock);
 	spin_unlock_irqrestore(&rtc_lock, flags);
 	return status;
 }
@@ -67,7 +142,9 @@ static efi_status_t virt_efi_set_wakeup_time(efi_bool_t enabled, efi_time_t *tm)
 	efi_status_t status;
 
 	spin_lock_irqsave(&rtc_lock, flags);
+	spin_lock(&efi_runtime_lock);
 	status = efi_call_virt(set_wakeup_time, enabled, tm);
+	spin_unlock(&efi_runtime_lock);
 	spin_unlock_irqrestore(&rtc_lock, flags);
 	return status;
 }
@@ -78,14 +155,27 @@ static efi_status_t virt_efi_get_variable(efi_char16_t *name,
 					  unsigned long *data_size,
 					  void *data)
 {
-	return efi_call_virt(get_variable, name, vendor, attr, data_size, data);
+	unsigned long flags;
+	efi_status_t status;
+
+	spin_lock_irqsave(&efi_runtime_lock, flags);
+	status = efi_call_virt(get_variable, name, vendor, attr, data_size,
+			       data);
+	spin_unlock_irqrestore(&efi_runtime_lock, flags);
+	return status;
 }
 
 static efi_status_t virt_efi_get_next_variable(unsigned long *name_size,
 					       efi_char16_t *name,
 					       efi_guid_t *vendor)
 {
-	return efi_call_virt(get_next_variable, name_size, name, vendor);
+	unsigned long flags;
+	efi_status_t status;
+
+	spin_lock_irqsave(&efi_runtime_lock, flags);
+	status = efi_call_virt(get_next_variable, name_size, name, vendor);
+	spin_unlock_irqrestore(&efi_runtime_lock, flags);
+	return status;
 }
 
 static efi_status_t virt_efi_set_variable(efi_char16_t *name,
@@ -94,24 +184,61 @@ static efi_status_t virt_efi_set_variable(efi_char16_t *name,
 					  unsigned long data_size,
 					  void *data)
 {
-	return efi_call_virt(set_variable, name, vendor, attr, data_size, data);
+	unsigned long flags;
+	efi_status_t status;
+
+	spin_lock_irqsave(&efi_runtime_lock, flags);
+	status = efi_call_virt(set_variable, name, vendor, attr, data_size,
+			       data);
+	spin_unlock_irqrestore(&efi_runtime_lock, flags);
+	return status;
 }
 
+static efi_status_t
+virt_efi_set_variable_nonblocking(efi_char16_t *name, efi_guid_t *vendor,
+				  u32 attr, unsigned long data_size,
+				  void *data)
+{
+	unsigned long flags;
+	efi_status_t status;
+
+	if (!spin_trylock_irqsave(&efi_runtime_lock, flags))
+		return EFI_NOT_READY;
+
+	status = efi_call_virt(set_variable, name, vendor, attr, data_size,
+			       data);
+	spin_unlock_irqrestore(&efi_runtime_lock, flags);
+	return status;
+}
+
+
 static efi_status_t virt_efi_query_variable_info(u32 attr,
 						 u64 *storage_space,
 						 u64 *remaining_space,
 						 u64 *max_variable_size)
 {
+	unsigned long flags;
+	efi_status_t status;
+
 	if (efi.runtime_version < EFI_2_00_SYSTEM_TABLE_REVISION)
 		return EFI_UNSUPPORTED;
 
-	return efi_call_virt(query_variable_info, attr, storage_space,
-			     remaining_space, max_variable_size);
+	spin_lock_irqsave(&efi_runtime_lock, flags);
+	status = efi_call_virt(query_variable_info, attr, storage_space,
+			       remaining_space, max_variable_size);
+	spin_unlock_irqrestore(&efi_runtime_lock, flags);
+	return status;
 }
 
 static efi_status_t virt_efi_get_next_high_mono_count(u32 *count)
 {
-	return efi_call_virt(get_next_high_mono_count, count);
+	unsigned long flags;
+	efi_status_t status;
+
+	spin_lock_irqsave(&efi_runtime_lock, flags);
+	status = efi_call_virt(get_next_high_mono_count, count);
+	spin_unlock_irqrestore(&efi_runtime_lock, flags);
+	return status;
 }
 
 static void virt_efi_reset_system(int reset_type,
@@ -119,17 +246,27 @@ static void virt_efi_reset_system(int reset_type,
 				  unsigned long data_size,
 				  efi_char16_t *data)
 {
+	unsigned long flags;
+
+	spin_lock_irqsave(&efi_runtime_lock, flags);
 	__efi_call_virt(reset_system, reset_type, status, data_size, data);
+	spin_unlock_irqrestore(&efi_runtime_lock, flags);
 }
 
 static efi_status_t virt_efi_update_capsule(efi_capsule_header_t **capsules,
 					    unsigned long count,
 					    unsigned long sg_list)
 {
+	unsigned long flags;
+	efi_status_t status;
+
 	if (efi.runtime_version < EFI_2_00_SYSTEM_TABLE_REVISION)
 		return EFI_UNSUPPORTED;
 
-	return efi_call_virt(update_capsule, capsules, count, sg_list);
+	spin_lock_irqsave(&efi_runtime_lock, flags);
+	status = efi_call_virt(update_capsule, capsules, count, sg_list);
+	spin_unlock_irqrestore(&efi_runtime_lock, flags);
+	return status;
 }
 
 static efi_status_t virt_efi_query_capsule_caps(efi_capsule_header_t **capsules,
@@ -137,11 +274,17 @@ static efi_status_t virt_efi_query_capsule_caps(efi_capsule_header_t **capsules,
 						u64 *max_size,
 						int *reset_type)
 {
+	unsigned long flags;
+	efi_status_t status;
+
 	if (efi.runtime_version < EFI_2_00_SYSTEM_TABLE_REVISION)
 		return EFI_UNSUPPORTED;
 
-	return efi_call_virt(query_capsule_caps, capsules, count, max_size,
-			     reset_type);
+	spin_lock_irqsave(&efi_runtime_lock, flags);
+	status = efi_call_virt(query_capsule_caps, capsules, count, max_size,
+			       reset_type);
+	spin_unlock_irqrestore(&efi_runtime_lock, flags);
+	return status;
 }
 
 void efi_native_runtime_setup(void)
@@ -153,6 +296,7 @@ void efi_native_runtime_setup(void)
 	efi.get_variable = virt_efi_get_variable;
 	efi.get_next_variable = virt_efi_get_next_variable;
 	efi.set_variable = virt_efi_set_variable;
+	efi.set_variable_nonblocking = virt_efi_set_variable_nonblocking;
 	efi.get_next_high_mono_count = virt_efi_get_next_high_mono_count;
 	efi.reset_system = virt_efi_reset_system;
 	efi.query_variable_info = virt_efi_query_variable_info;
diff --git a/drivers/firmware/efi/vars.c b/drivers/firmware/efi/vars.c
index 5abe943e3404..70a0fb10517f 100644
--- a/drivers/firmware/efi/vars.c
+++ b/drivers/firmware/efi/vars.c
@@ -321,11 +321,11 @@ static unsigned long var_name_strnsize(efi_char16_t *variable_name,
  * Print a warning when duplicate EFI variables are encountered and
  * disable the sysfs workqueue since the firmware is buggy.
  */
-static void dup_variable_bug(efi_char16_t *s16, efi_guid_t *vendor_guid,
+static void dup_variable_bug(efi_char16_t *str16, efi_guid_t *vendor_guid,
 			     unsigned long len16)
 {
 	size_t i, len8 = len16 / sizeof(efi_char16_t);
-	char *s8;
+	char *str8;
 
 	/*
 	 * Disable the workqueue since the algorithm it uses for
@@ -334,16 +334,16 @@ static void dup_variable_bug(efi_char16_t *s16, efi_guid_t *vendor_guid,
 	 */
 	efivar_wq_enabled = false;
 
-	s8 = kzalloc(len8, GFP_KERNEL);
-	if (!s8)
+	str8 = kzalloc(len8, GFP_KERNEL);
+	if (!str8)
 		return;
 
 	for (i = 0; i < len8; i++)
-		s8[i] = s16[i];
+		str8[i] = str16[i];
 
 	printk(KERN_WARNING "efivars: duplicate variable: %s-%pUl\n",
-	       s8, vendor_guid);
-	kfree(s8);
+	       str8, vendor_guid);
+	kfree(str8);
 }
 
 /**
@@ -595,6 +595,39 @@ int efivar_entry_set(struct efivar_entry *entry, u32 attributes,
 }
 EXPORT_SYMBOL_GPL(efivar_entry_set);
 
+/*
+ * efivar_entry_set_nonblocking - call set_variable_nonblocking()
+ *
+ * This function is guaranteed to not block and is suitable for calling
+ * from crash/panic handlers.
+ *
+ * Crucially, this function will not block if it cannot acquire
+ * __efivars->lock. Instead, it returns -EBUSY.
+ */
+static int
+efivar_entry_set_nonblocking(efi_char16_t *name, efi_guid_t vendor,
+			     u32 attributes, unsigned long size, void *data)
+{
+	const struct efivar_operations *ops = __efivars->ops;
+	unsigned long flags;
+	efi_status_t status;
+
+	if (!spin_trylock_irqsave(&__efivars->lock, flags))
+		return -EBUSY;
+
+	status = check_var_size(attributes, size + ucs2_strsize(name, 1024));
+	if (status != EFI_SUCCESS) {
+		spin_unlock_irqrestore(&__efivars->lock, flags);
+		return -ENOSPC;
+	}
+
+	status = ops->set_variable_nonblocking(name, &vendor, attributes,
+					       size, data);
+
+	spin_unlock_irqrestore(&__efivars->lock, flags);
+	return efi_status_to_err(status);
+}
+
 /**
  * efivar_entry_set_safe - call set_variable() if enough space in firmware
  * @name: buffer containing the variable name
@@ -622,6 +655,20 @@ int efivar_entry_set_safe(efi_char16_t *name, efi_guid_t vendor, u32 attributes,
 	if (!ops->query_variable_store)
 		return -ENOSYS;
 
+	/*
+	 * If the EFI variable backend provides a non-blocking
+	 * ->set_variable() operation and we're in a context where we
+	 * cannot block, then we need to use it to avoid live-locks,
+	 * since the implication is that the regular ->set_variable()
+	 * will block.
+	 *
+	 * If no ->set_variable_nonblocking() is provided then
+	 * ->set_variable() is assumed to be non-blocking.
+	 */
+	if (!block && ops->set_variable_nonblocking)
+		return efivar_entry_set_nonblocking(name, vendor, attributes,
+						    size, data);
+
 	if (!block) {
 		if (!spin_trylock_irqsave(&__efivars->lock, flags))
 			return -EBUSY;
diff --git a/drivers/firmware/memmap.c b/drivers/firmware/memmap.c
index 79f18e6d9c4f..cc016c615c19 100644
--- a/drivers/firmware/memmap.c
+++ b/drivers/firmware/memmap.c
@@ -184,6 +184,9 @@ static int add_sysfs_fw_map_entry(struct firmware_map_entry *entry)
 	static int map_entries_nr;
 	static struct kset *mmap_kset;
 
+	if (entry->kobj.state_in_sysfs)
+		return -EEXIST;
+
 	if (!mmap_kset) {
 		mmap_kset = kset_create_and_add("memmap", NULL, firmware_kobj);
 		if (!mmap_kset)
diff --git a/drivers/gpio/Kconfig b/drivers/gpio/Kconfig
index 9de1515e5808..633ec216e185 100644
--- a/drivers/gpio/Kconfig
+++ b/drivers/gpio/Kconfig
@@ -112,6 +112,20 @@ config GPIO_MAX730X
 
 comment "Memory mapped GPIO drivers:"
 
+config GPIO_74XX_MMIO
+	tristate "GPIO driver for 74xx-ICs with MMIO access"
+	depends on OF_GPIO
+	select GPIO_GENERIC
+	help
+	  Say yes here to support GPIO functionality for 74xx-compatible ICs
+	  with MMIO access. Compatible models include:
+	    1 bit:	741G125 (Input), 741G74 (Output)
+	    2 bits:	742G125 (Input), 7474 (Output)
+	    4 bits:	74125 (Input), 74175 (Output)
+	    6 bits:	74365 (Input), 74174 (Output)
+	    8 bits:	74244 (Input), 74273 (Output)
+	    16 bits:	741624 (Input), 7416374 (Output)
+
 config GPIO_CLPS711X
 	tristate "CLPS711X GPIO support"
 	depends on ARCH_CLPS711X || COMPILE_TEST
@@ -134,9 +148,10 @@ config GPIO_GENERIC_PLATFORM
 
 config GPIO_DWAPB
 	tristate "Synopsys DesignWare APB GPIO driver"
+	depends on ARM
+	depends on OF_GPIO
 	select GPIO_GENERIC
 	select GENERIC_IRQ_CHIP
-	depends on OF_GPIO
 	help
 	  Say Y or M here to build support for the Synopsys DesignWare APB
 	  GPIO block.
@@ -334,6 +349,22 @@ config GPIO_TZ1090_PDC
 	help
 	  Say yes here to support Toumaz Xenif TZ1090 PDC GPIOs.
 
+config GPIO_VF610
+	def_bool y
+	depends on ARCH_MXC && SOC_VF610
+	select GPIOLIB_IRQCHIP
+	help
+	  Say yes here to support Vybrid vf610 GPIOs.
+
+config GPIO_XGENE
+	bool "APM X-Gene GPIO controller support"
+	depends on ARM64 && OF_GPIO
+	help
+	  This driver is to support the GPIO block within the APM X-Gene SoC
+	  platform's generic flash controller. The GPIO pins are muxed with
+	  the generic flash controller's address and data pins. Say yes
+	  here to enable the GFC GPIO functionality.
+
 config GPIO_XILINX
 	bool "Xilinx GPIO support"
 	depends on PPC_OF || MICROBLAZE || ARCH_ZYNQ
@@ -681,6 +712,7 @@ config GPIO_ADP5588_IRQ
 config GPIO_ADNP
 	tristate "Avionic Design N-bit GPIO expander"
 	depends on I2C && OF_GPIO
+	select GPIOLIB_IRQCHIP
 	help
 	  This option enables support for N GPIOs found on Avionic Design
 	  I2C GPIO expanders. The register space will be extended by powers
@@ -796,7 +828,6 @@ config GPIO_MAX7301
 
 config GPIO_MCP23S08
 	tristate "Microchip MCP23xxx I/O expander"
-	depends on OF_GPIO
 	depends on (SPI_MASTER && !I2C) || I2C
 	help
 	  SPI/I2C driver for Microchip MCP23S08/MCP23S17/MCP23008/MCP23017
@@ -880,7 +911,7 @@ config GPIO_MSIC
 
 config GPIO_BCM_KONA
 	bool "Broadcom Kona GPIO"
-	depends on OF_GPIO
+	depends on OF_GPIO && (ARCH_BCM_MOBILE || COMPILE_TEST)
 	help
 	  Turn on GPIO support for Broadcom "Kona" chips.
 
@@ -897,4 +928,16 @@ config GPIO_VIPERBOARD
           River Tech's viperboard.h for detailed meaning
           of the module parameters.
 
+config GPIO_DLN2
+	tristate "Diolan DLN2 GPIO support"
+	depends on MFD_DLN2
+	select GPIOLIB_IRQCHIP
+
+	help
+	  Select this option to enable GPIO driver for the Diolan DLN2
+	  board.
+
+	  This driver can also be built as a module. If so, the module
+	  will be called gpio-dln2.
+
 endif
diff --git a/drivers/gpio/Makefile b/drivers/gpio/Makefile
index 5d024e396622..81755f1305e6 100644
--- a/drivers/gpio/Makefile
+++ b/drivers/gpio/Makefile
@@ -13,6 +13,7 @@ obj-$(CONFIG_GPIO_ACPI)		+= gpiolib-acpi.o
 obj-$(CONFIG_GPIO_GENERIC)	+= gpio-generic.o
 
 obj-$(CONFIG_GPIO_74X164)	+= gpio-74x164.o
+obj-$(CONFIG_GPIO_74XX_MMIO)	+= gpio-74xx-mmio.o
 obj-$(CONFIG_GPIO_ADNP)		+= gpio-adnp.o
 obj-$(CONFIG_GPIO_ADP5520)	+= gpio-adp5520.o
 obj-$(CONFIG_GPIO_ADP5588)	+= gpio-adp5588.o
@@ -26,6 +27,7 @@ obj-$(CONFIG_GPIO_CRYSTAL_COVE)	+= gpio-crystalcove.o
 obj-$(CONFIG_GPIO_DA9052)	+= gpio-da9052.o
 obj-$(CONFIG_GPIO_DA9055)	+= gpio-da9055.o
 obj-$(CONFIG_GPIO_DAVINCI)	+= gpio-davinci.o
+obj-$(CONFIG_GPIO_DLN2)		+= gpio-dln2.o
 obj-$(CONFIG_GPIO_DWAPB)	+= gpio-dwapb.o
 obj-$(CONFIG_GPIO_EM)		+= gpio-em.o
 obj-$(CONFIG_GPIO_EP93XX)	+= gpio-ep93xx.o
@@ -95,12 +97,14 @@ obj-$(CONFIG_GPIO_TWL6040)	+= gpio-twl6040.o
 obj-$(CONFIG_GPIO_TZ1090)	+= gpio-tz1090.o
 obj-$(CONFIG_GPIO_TZ1090_PDC)	+= gpio-tz1090-pdc.o
 obj-$(CONFIG_GPIO_UCB1400)	+= gpio-ucb1400.o
+obj-$(CONFIG_GPIO_VF610)	+= gpio-vf610.o
 obj-$(CONFIG_GPIO_VIPERBOARD)	+= gpio-viperboard.o
 obj-$(CONFIG_GPIO_VR41XX)	+= gpio-vr41xx.o
 obj-$(CONFIG_GPIO_VX855)	+= gpio-vx855.o
 obj-$(CONFIG_GPIO_WM831X)	+= gpio-wm831x.o
 obj-$(CONFIG_GPIO_WM8350)	+= gpio-wm8350.o
 obj-$(CONFIG_GPIO_WM8994)	+= gpio-wm8994.o
+obj-$(CONFIG_GPIO_XGENE)	+= gpio-xgene.o
 obj-$(CONFIG_GPIO_XILINX)	+= gpio-xilinx.o
 obj-$(CONFIG_GPIO_XTENSA)	+= gpio-xtensa.o
 obj-$(CONFIG_GPIO_ZEVIO)	+= gpio-zevio.o
diff --git a/drivers/gpio/devres.c b/drivers/gpio/devres.c
index 954b9f6b0ef8..13dbd3dfc33a 100644
--- a/drivers/gpio/devres.c
+++ b/drivers/gpio/devres.c
@@ -109,6 +109,38 @@ struct gpio_desc *__must_check __devm_gpiod_get_index(struct device *dev,
 EXPORT_SYMBOL(__devm_gpiod_get_index);
 
 /**
+ * devm_get_gpiod_from_child - get a GPIO descriptor from a device's child node
+ * @dev:	GPIO consumer
+ * @child:	firmware node (child of @dev)
+ *
+ * GPIO descriptors returned from this function are automatically disposed on
+ * driver detach.
+ */
+struct gpio_desc *devm_get_gpiod_from_child(struct device *dev,
+					    struct fwnode_handle *child)
+{
+	struct gpio_desc **dr;
+	struct gpio_desc *desc;
+
+	dr = devres_alloc(devm_gpiod_release, sizeof(struct gpio_desc *),
+			  GFP_KERNEL);
+	if (!dr)
+		return ERR_PTR(-ENOMEM);
+
+	desc = fwnode_get_named_gpiod(child, "gpios");
+	if (IS_ERR(desc)) {
+		devres_free(dr);
+		return desc;
+	}
+
+	*dr = desc;
+	devres_add(dev, dr);
+
+	return desc;
+}
+EXPORT_SYMBOL(devm_get_gpiod_from_child);
+
+/**
  * devm_gpiod_get_index_optional - Resource-managed gpiod_get_index_optional()
  * @dev: GPIO consumer
  * @con_id: function within the GPIO consumer
diff --git a/drivers/gpio/gpio-74xx-mmio.c b/drivers/gpio/gpio-74xx-mmio.c
new file mode 100644
index 000000000000..0763655cca6c
--- /dev/null
+++ b/drivers/gpio/gpio-74xx-mmio.c
@@ -0,0 +1,170 @@
+/*
+ * 74xx MMIO GPIO driver
+ *
+ *  Copyright (C) 2014 Alexander Shiyan <shc_work@mail.ru>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/err.h>
+#include <linux/gpio.h>
+#include <linux/module.h>
+#include <linux/of_device.h>
+#include <linux/basic_mmio_gpio.h>
+#include <linux/platform_device.h>
+
+#define MMIO_74XX_DIR_IN	(0 << 8)
+#define MMIO_74XX_DIR_OUT	(1 << 8)
+#define MMIO_74XX_BIT_CNT(x)	((x) & 0xff)
+
+struct mmio_74xx_gpio_priv {
+	struct bgpio_chip	bgc;
+	unsigned		flags;
+};
+
+static const struct of_device_id mmio_74xx_gpio_ids[] = {
+	{
+		.compatible	= "ti,741g125",
+		.data		= (const void *)(MMIO_74XX_DIR_IN | 1),
+	},
+	{
+		.compatible	= "ti,742g125",
+		.data		= (const void *)(MMIO_74XX_DIR_IN | 2),
+	},
+	{
+		.compatible	= "ti,74125",
+		.data		= (const void *)(MMIO_74XX_DIR_IN | 4),
+	},
+	{
+		.compatible	= "ti,74365",
+		.data		= (const void *)(MMIO_74XX_DIR_IN | 6),
+	},
+	{
+		.compatible	= "ti,74244",
+		.data		= (const void *)(MMIO_74XX_DIR_IN | 8),
+	},
+	{
+		.compatible	= "ti,741624",
+		.data		= (const void *)(MMIO_74XX_DIR_IN | 16),
+	},
+	{
+		.compatible	= "ti,741g74",
+		.data		= (const void *)(MMIO_74XX_DIR_OUT | 1),
+	},
+	{
+		.compatible	= "ti,7474",
+		.data		= (const void *)(MMIO_74XX_DIR_OUT | 2),
+	},
+	{
+		.compatible	= "ti,74175",
+		.data		= (const void *)(MMIO_74XX_DIR_OUT | 4),
+	},
+	{
+		.compatible	= "ti,74174",
+		.data		= (const void *)(MMIO_74XX_DIR_OUT | 6),
+	},
+	{
+		.compatible	= "ti,74273",
+		.data		= (const void *)(MMIO_74XX_DIR_OUT | 8),
+	},
+	{
+		.compatible	= "ti,7416374",
+		.data		= (const void *)(MMIO_74XX_DIR_OUT | 16),
+	},
+	{ }
+};
+MODULE_DEVICE_TABLE(of, mmio_74xx_gpio_ids);
+
+static inline struct mmio_74xx_gpio_priv *to_74xx_gpio(struct gpio_chip *gc)
+{
+	struct bgpio_chip *bgc = to_bgpio_chip(gc);
+
+	return container_of(bgc, struct mmio_74xx_gpio_priv, bgc);
+}
+
+static int mmio_74xx_get_direction(struct gpio_chip *gc, unsigned offset)
+{
+	struct mmio_74xx_gpio_priv *priv = to_74xx_gpio(gc);
+
+	return (priv->flags & MMIO_74XX_DIR_OUT) ? GPIOF_DIR_OUT : GPIOF_DIR_IN;
+}
+
+static int mmio_74xx_dir_in(struct gpio_chip *gc, unsigned int gpio)
+{
+	struct mmio_74xx_gpio_priv *priv = to_74xx_gpio(gc);
+
+	return (priv->flags & MMIO_74XX_DIR_OUT) ? -ENOTSUPP : 0;
+}
+
+static int mmio_74xx_dir_out(struct gpio_chip *gc, unsigned int gpio, int val)
+{
+	struct mmio_74xx_gpio_priv *priv = to_74xx_gpio(gc);
+
+	if (priv->flags & MMIO_74XX_DIR_OUT) {
+		gc->set(gc, gpio, val);
+		return 0;
+	}
+
+	return -ENOTSUPP;
+}
+
+static int mmio_74xx_gpio_probe(struct platform_device *pdev)
+{
+	const struct of_device_id *of_id =
+		of_match_device(mmio_74xx_gpio_ids, &pdev->dev);
+	struct mmio_74xx_gpio_priv *priv;
+	struct resource *res;
+	void __iomem *dat;
+	int err;
+
+	priv = devm_kzalloc(&pdev->dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	dat = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(dat))
+		return PTR_ERR(dat);
+
+	priv->flags = (unsigned)of_id->data;
+
+	err = bgpio_init(&priv->bgc, &pdev->dev,
+			 DIV_ROUND_UP(MMIO_74XX_BIT_CNT(priv->flags), 8),
+			 dat, NULL, NULL, NULL, NULL, 0);
+	if (err)
+		return err;
+
+	priv->bgc.gc.direction_input = mmio_74xx_dir_in;
+	priv->bgc.gc.direction_output = mmio_74xx_dir_out;
+	priv->bgc.gc.get_direction = mmio_74xx_get_direction;
+	priv->bgc.gc.ngpio = MMIO_74XX_BIT_CNT(priv->flags);
+	priv->bgc.gc.owner = THIS_MODULE;
+
+	platform_set_drvdata(pdev, priv);
+
+	return gpiochip_add(&priv->bgc.gc);
+}
+
+static int mmio_74xx_gpio_remove(struct platform_device *pdev)
+{
+	struct mmio_74xx_gpio_priv *priv = platform_get_drvdata(pdev);
+
+	return bgpio_remove(&priv->bgc);
+}
+
+static struct platform_driver mmio_74xx_gpio_driver = {
+	.driver	= {
+		.name		= "74xx-mmio-gpio",
+		.of_match_table	= mmio_74xx_gpio_ids,
+	},
+	.probe	= mmio_74xx_gpio_probe,
+	.remove	= mmio_74xx_gpio_remove,
+};
+module_platform_driver(mmio_74xx_gpio_driver);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Alexander Shiyan <shc_work@mail.ru>");
+MODULE_DESCRIPTION("74xx MMIO GPIO driver");
diff --git a/drivers/gpio/gpio-adnp.c b/drivers/gpio/gpio-adnp.c
index 416b2200d4f1..d3d0a90fe542 100644
--- a/drivers/gpio/gpio-adnp.c
+++ b/drivers/gpio/gpio-adnp.c
@@ -6,10 +6,9 @@
  * published by the Free Software Foundation.
  */
 
-#include <linux/gpio.h>
+#include <linux/gpio/driver.h>
 #include <linux/i2c.h>
 #include <linux/interrupt.h>
-#include <linux/irqdomain.h>
 #include <linux/module.h>
 #include <linux/of_irq.h>
 #include <linux/seq_file.h>
@@ -27,8 +26,6 @@ struct adnp {
 	unsigned int reg_shift;
 
 	struct mutex i2c_lock;
-
-	struct irq_domain *domain;
 	struct mutex irq_lock;
 
 	u8 *irq_enable;
@@ -253,6 +250,7 @@ static void adnp_gpio_dbg_show(struct seq_file *s, struct gpio_chip *chip)
 static int adnp_gpio_setup(struct adnp *adnp, unsigned int num_gpios)
 {
 	struct gpio_chip *chip = &adnp->gpio;
+	int err;
 
 	adnp->reg_shift = get_count_order(num_gpios) - 3;
 
@@ -272,6 +270,10 @@ static int adnp_gpio_setup(struct adnp *adnp, unsigned int num_gpios)
 	chip->of_node = chip->dev->of_node;
 	chip->owner = THIS_MODULE;
 
+	err = gpiochip_add(chip);
+	if (err)
+		return err;
+
 	return 0;
 }
 
@@ -326,7 +328,8 @@ static irqreturn_t adnp_irq(int irq, void *data)
 
 		for_each_set_bit(bit, &pending, 8) {
 			unsigned int child_irq;
-			child_irq = irq_find_mapping(adnp->domain, base + bit);
+			child_irq = irq_find_mapping(adnp->gpio.irqdomain,
+						     base + bit);
 			handle_nested_irq(child_irq);
 		}
 	}
@@ -334,35 +337,32 @@ static irqreturn_t adnp_irq(int irq, void *data)
 	return IRQ_HANDLED;
 }
 
-static int adnp_gpio_to_irq(struct gpio_chip *chip, unsigned offset)
-{
-	struct adnp *adnp = to_adnp(chip);
-	return irq_create_mapping(adnp->domain, offset);
-}
-
-static void adnp_irq_mask(struct irq_data *data)
+static void adnp_irq_mask(struct irq_data *d)
 {
-	struct adnp *adnp = irq_data_get_irq_chip_data(data);
-	unsigned int reg = data->hwirq >> adnp->reg_shift;
-	unsigned int pos = data->hwirq & 7;
+	struct gpio_chip *gc = irq_data_get_irq_chip_data(d);
+	struct adnp *adnp = to_adnp(gc);
+	unsigned int reg = d->hwirq >> adnp->reg_shift;
+	unsigned int pos = d->hwirq & 7;
 
 	adnp->irq_enable[reg] &= ~BIT(pos);
 }
 
-static void adnp_irq_unmask(struct irq_data *data)
+static void adnp_irq_unmask(struct irq_data *d)
 {
-	struct adnp *adnp = irq_data_get_irq_chip_data(data);
-	unsigned int reg = data->hwirq >> adnp->reg_shift;
-	unsigned int pos = data->hwirq & 7;
+	struct gpio_chip *gc = irq_data_get_irq_chip_data(d);
+	struct adnp *adnp = to_adnp(gc);
+	unsigned int reg = d->hwirq >> adnp->reg_shift;
+	unsigned int pos = d->hwirq & 7;
 
 	adnp->irq_enable[reg] |= BIT(pos);
 }
 
-static int adnp_irq_set_type(struct irq_data *data, unsigned int type)
+static int adnp_irq_set_type(struct irq_data *d, unsigned int type)
 {
-	struct adnp *adnp = irq_data_get_irq_chip_data(data);
-	unsigned int reg = data->hwirq >> adnp->reg_shift;
-	unsigned int pos = data->hwirq & 7;
+	struct gpio_chip *gc = irq_data_get_irq_chip_data(d);
+	struct adnp *adnp = to_adnp(gc);
+	unsigned int reg = d->hwirq >> adnp->reg_shift;
+	unsigned int pos = d->hwirq & 7;
 
 	if (type & IRQ_TYPE_EDGE_RISING)
 		adnp->irq_rise[reg] |= BIT(pos);
@@ -387,16 +387,18 @@ static int adnp_irq_set_type(struct irq_data *data, unsigned int type)
 	return 0;
 }
 
-static void adnp_irq_bus_lock(struct irq_data *data)
+static void adnp_irq_bus_lock(struct irq_data *d)
 {
-	struct adnp *adnp = irq_data_get_irq_chip_data(data);
+	struct gpio_chip *gc = irq_data_get_irq_chip_data(d);
+	struct adnp *adnp = to_adnp(gc);
 
 	mutex_lock(&adnp->irq_lock);
 }
 
-static void adnp_irq_bus_unlock(struct irq_data *data)
+static void adnp_irq_bus_unlock(struct irq_data *d)
 {
-	struct adnp *adnp = irq_data_get_irq_chip_data(data);
+	struct gpio_chip *gc = irq_data_get_irq_chip_data(d);
+	struct adnp *adnp = to_adnp(gc);
 	unsigned int num_regs = 1 << adnp->reg_shift, i;
 
 	mutex_lock(&adnp->i2c_lock);
@@ -408,26 +410,6 @@ static void adnp_irq_bus_unlock(struct irq_data *data)
 	mutex_unlock(&adnp->irq_lock);
 }
 
-static int adnp_irq_reqres(struct irq_data *data)
-{
-	struct adnp *adnp = irq_data_get_irq_chip_data(data);
-
-	if (gpio_lock_as_irq(&adnp->gpio, data->hwirq)) {
-		dev_err(adnp->gpio.dev,
-			"unable to lock HW IRQ %lu for IRQ\n",
-			data->hwirq);
-		return -EINVAL;
-	}
-	return 0;
-}
-
-static void adnp_irq_relres(struct irq_data *data)
-{
-	struct adnp *adnp = irq_data_get_irq_chip_data(data);
-
-	gpio_unlock_as_irq(&adnp->gpio, data->hwirq);
-}
-
 static struct irq_chip adnp_irq_chip = {
 	.name = "gpio-adnp",
 	.irq_mask = adnp_irq_mask,
@@ -435,29 +417,6 @@ static struct irq_chip adnp_irq_chip = {
 	.irq_set_type = adnp_irq_set_type,
 	.irq_bus_lock = adnp_irq_bus_lock,
 	.irq_bus_sync_unlock = adnp_irq_bus_unlock,
-	.irq_request_resources = adnp_irq_reqres,
-	.irq_release_resources = adnp_irq_relres,
-};
-
-static int adnp_irq_map(struct irq_domain *domain, unsigned int irq,
-			irq_hw_number_t hwirq)
-{
-	irq_set_chip_data(irq, domain->host_data);
-	irq_set_chip(irq, &adnp_irq_chip);
-	irq_set_nested_thread(irq, true);
-
-#ifdef CONFIG_ARM
-	set_irq_flags(irq, IRQF_VALID);
-#else
-	irq_set_noprobe(irq);
-#endif
-
-	return 0;
-}
-
-static const struct irq_domain_ops adnp_irq_domain_ops = {
-	.map = adnp_irq_map,
-	.xlate = irq_domain_xlate_twocell,
 };
 
 static int adnp_irq_setup(struct adnp *adnp)
@@ -503,35 +462,28 @@ static int adnp_irq_setup(struct adnp *adnp)
 		adnp->irq_enable[i] = 0x00;
 	}
 
-	adnp->domain = irq_domain_add_linear(chip->of_node, chip->ngpio,
-					     &adnp_irq_domain_ops, adnp);
-
-	err = request_threaded_irq(adnp->client->irq, NULL, adnp_irq,
-				   IRQF_TRIGGER_RISING | IRQF_ONESHOT,
-				   dev_name(chip->dev), adnp);
+	err = devm_request_threaded_irq(chip->dev, adnp->client->irq,
+					NULL, adnp_irq,
+					IRQF_TRIGGER_RISING | IRQF_ONESHOT,
+					dev_name(chip->dev), adnp);
 	if (err != 0) {
 		dev_err(chip->dev, "can't request IRQ#%d: %d\n",
 			adnp->client->irq, err);
 		return err;
 	}
 
-	chip->to_irq = adnp_gpio_to_irq;
-	return 0;
-}
-
-static void adnp_irq_teardown(struct adnp *adnp)
-{
-	unsigned int irq, i;
-
-	free_irq(adnp->client->irq, adnp);
-
-	for (i = 0; i < adnp->gpio.ngpio; i++) {
-		irq = irq_find_mapping(adnp->domain, i);
-		if (irq > 0)
-			irq_dispose_mapping(irq);
+	err = gpiochip_irqchip_add(chip,
+				   &adnp_irq_chip,
+				   0,
+				   handle_simple_irq,
+				   IRQ_TYPE_NONE);
+	if (err) {
+		dev_err(chip->dev,
+			"could not connect irqchip to gpiochip\n");
+		return err;
 	}
 
-	irq_domain_remove(adnp->domain);
+	return 0;
 }
 
 static int adnp_i2c_probe(struct i2c_client *client,
@@ -558,38 +510,25 @@ static int adnp_i2c_probe(struct i2c_client *client,
 	adnp->client = client;
 
 	err = adnp_gpio_setup(adnp, num_gpios);
-	if (err < 0)
+	if (err)
 		return err;
 
 	if (of_find_property(np, "interrupt-controller", NULL)) {
 		err = adnp_irq_setup(adnp);
-		if (err < 0)
-			goto teardown;
+		if (err)
+			return err;
 	}
 
-	err = gpiochip_add(&adnp->gpio);
-	if (err < 0)
-		goto teardown;
-
 	i2c_set_clientdata(client, adnp);
-	return 0;
 
-teardown:
-	if (of_find_property(np, "interrupt-controller", NULL))
-		adnp_irq_teardown(adnp);
-
-	return err;
+	return 0;
 }
 
 static int adnp_i2c_remove(struct i2c_client *client)
 {
 	struct adnp *adnp = i2c_get_clientdata(client);
-	struct device_node *np = client->dev.of_node;
 
 	gpiochip_remove(&adnp->gpio);
-	if (of_find_property(np, "interrupt-controller", NULL))
-		adnp_irq_teardown(adnp);
-
 	return 0;
 }
 
diff --git a/drivers/gpio/gpio-adp5520.c b/drivers/gpio/gpio-adp5520.c
index b08bd169e568..caff711ca5a9 100644
--- a/drivers/gpio/gpio-adp5520.c
+++ b/drivers/gpio/gpio-adp5520.c
@@ -177,7 +177,6 @@ static int adp5520_gpio_remove(struct platform_device *pdev)
 static struct platform_driver adp5520_gpio_driver = {
 	.driver	= {
 		.name	= "adp5520-gpio",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= adp5520_gpio_probe,
 	.remove		= adp5520_gpio_remove,
diff --git a/drivers/gpio/gpio-amd8111.c b/drivers/gpio/gpio-amd8111.c
index 3c09f1a6872a..d3d2d1099f64 100644
--- a/drivers/gpio/gpio-amd8111.c
+++ b/drivers/gpio/gpio-amd8111.c
@@ -223,6 +223,7 @@ found:
 	if (err) {
 		printk(KERN_ERR "GPIO registering failed (%d)\n",
 		       err);
+		ioport_unmap(gp.pm);
 		release_region(gp.pmbase + PMBASE_OFFSET, PMBASE_SIZE);
 		goto out;
 	}
diff --git a/drivers/gpio/gpio-bcm-kona.c b/drivers/gpio/gpio-bcm-kona.c
index 3f6b33ce9bd4..b164ce837b43 100644
--- a/drivers/gpio/gpio-bcm-kona.c
+++ b/drivers/gpio/gpio-bcm-kona.c
@@ -470,7 +470,7 @@ static int bcm_kona_gpio_irq_reqres(struct irq_data *d)
 {
 	struct bcm_kona_gpio *kona_gpio = irq_data_get_irq_chip_data(d);
 
-	if (gpio_lock_as_irq(&kona_gpio->gpio_chip, d->hwirq)) {
+	if (gpiochip_lock_as_irq(&kona_gpio->gpio_chip, d->hwirq)) {
 		dev_err(kona_gpio->gpio_chip.dev,
 			"unable to lock HW IRQ %lu for IRQ\n",
 			d->hwirq);
@@ -483,7 +483,7 @@ static void bcm_kona_gpio_irq_relres(struct irq_data *d)
 {
 	struct bcm_kona_gpio *kona_gpio = irq_data_get_irq_chip_data(d);
 
-	gpio_unlock_as_irq(&kona_gpio->gpio_chip, d->hwirq);
+	gpiochip_unlock_as_irq(&kona_gpio->gpio_chip, d->hwirq);
 }
 
 static struct irq_chip bcm_gpio_irq_chip = {
@@ -496,7 +496,7 @@ static struct irq_chip bcm_gpio_irq_chip = {
 	.irq_release_resources = bcm_kona_gpio_irq_relres,
 };
 
-static struct __initconst of_device_id bcm_kona_gpio_of_match[] = {
+static struct of_device_id const bcm_kona_gpio_of_match[] = {
 	{ .compatible = "brcm,kona-gpio" },
 	{}
 };
@@ -668,7 +668,6 @@ err_irq_domain:
 static struct platform_driver bcm_kona_gpio_driver = {
 	.driver = {
 			.name = "bcm-kona-gpio",
-			.owner = THIS_MODULE,
 			.of_match_table = bcm_kona_gpio_of_match,
 	},
 	.probe = bcm_kona_gpio_probe,
diff --git a/drivers/gpio/gpio-clps711x.c b/drivers/gpio/gpio-clps711x.c
index e1e861239e95..b6908f1ff1ab 100644
--- a/drivers/gpio/gpio-clps711x.c
+++ b/drivers/gpio/gpio-clps711x.c
@@ -87,7 +87,6 @@ MODULE_DEVICE_TABLE(of, clps711x_gpio_ids);
 static struct platform_driver clps711x_gpio_driver = {
 	.driver	= {
 		.name		= "clps711x-gpio",
-		.owner		= THIS_MODULE,
 		.of_match_table	= of_match_ptr(clps711x_gpio_ids),
 	},
 	.probe	= clps711x_gpio_probe,
diff --git a/drivers/gpio/gpio-crystalcove.c b/drivers/gpio/gpio-crystalcove.c
index 934462f5bd22..3d9e08f7e823 100644
--- a/drivers/gpio/gpio-crystalcove.c
+++ b/drivers/gpio/gpio-crystalcove.c
@@ -24,6 +24,7 @@
 #include <linux/mfd/intel_soc_pmic.h>
 
 #define CRYSTALCOVE_GPIO_NUM	16
+#define CRYSTALCOVE_VGPIO_NUM	94
 
 #define UPDATE_IRQ_TYPE		BIT(0)
 #define UPDATE_IRQ_MASK		BIT(1)
@@ -130,6 +131,9 @@ static int crystalcove_gpio_dir_in(struct gpio_chip *chip, unsigned gpio)
 {
 	struct crystalcove_gpio *cg = to_cg(chip);
 
+	if (gpio > CRYSTALCOVE_VGPIO_NUM)
+		return 0;
+
 	return regmap_write(cg->regmap, to_reg(gpio, CTRL_OUT),
 			    CTLO_INPUT_SET);
 }
@@ -139,6 +143,9 @@ static int crystalcove_gpio_dir_out(struct gpio_chip *chip, unsigned gpio,
 {
 	struct crystalcove_gpio *cg = to_cg(chip);
 
+	if (gpio > CRYSTALCOVE_VGPIO_NUM)
+		return 0;
+
 	return regmap_write(cg->regmap, to_reg(gpio, CTRL_OUT),
 			    CTLO_OUTPUT_SET | value);
 }
@@ -149,6 +156,9 @@ static int crystalcove_gpio_get(struct gpio_chip *chip, unsigned gpio)
 	int ret;
 	unsigned int val;
 
+	if (gpio > CRYSTALCOVE_VGPIO_NUM)
+		return 0;
+
 	ret = regmap_read(cg->regmap, to_reg(gpio, CTRL_IN), &val);
 	if (ret)
 		return ret;
@@ -161,6 +171,9 @@ static void crystalcove_gpio_set(struct gpio_chip *chip,
 {
 	struct crystalcove_gpio *cg = to_cg(chip);
 
+	if (gpio > CRYSTALCOVE_VGPIO_NUM)
+		return;
+
 	if (value)
 		regmap_update_bits(cg->regmap, to_reg(gpio, CTRL_OUT), 1, 1);
 	else
@@ -256,10 +269,10 @@ static irqreturn_t crystalcove_gpio_irq_handler(int irq, void *data)
 
 	pending = p0 | p1 << 8;
 
-	for (gpio = 0; gpio < cg->chip.ngpio; gpio++) {
+	for (gpio = 0; gpio < CRYSTALCOVE_GPIO_NUM; gpio++) {
 		if (pending & BIT(gpio)) {
 			virq = irq_find_mapping(cg->chip.irqdomain, gpio);
-			generic_handle_irq(virq);
+			handle_nested_irq(virq);
 		}
 	}
 
@@ -273,7 +286,7 @@ static void crystalcove_gpio_dbg_show(struct seq_file *s,
 	int gpio, offset;
 	unsigned int ctlo, ctli, mirqs0, mirqsx, irq;
 
-	for (gpio = 0; gpio < cg->chip.ngpio; gpio++) {
+	for (gpio = 0; gpio < CRYSTALCOVE_GPIO_NUM; gpio++) {
 		regmap_read(cg->regmap, to_reg(gpio, CTRL_OUT), &ctlo);
 		regmap_read(cg->regmap, to_reg(gpio, CTRL_IN), &ctli);
 		regmap_read(cg->regmap, gpio < 8 ? MGPIO0IRQS0 : MGPIO1IRQS0,
@@ -320,7 +333,7 @@ static int crystalcove_gpio_probe(struct platform_device *pdev)
 	cg->chip.get = crystalcove_gpio_get;
 	cg->chip.set = crystalcove_gpio_set;
 	cg->chip.base = -1;
-	cg->chip.ngpio = CRYSTALCOVE_GPIO_NUM;
+	cg->chip.ngpio = CRYSTALCOVE_VGPIO_NUM;
 	cg->chip.can_sleep = true;
 	cg->chip.dev = dev;
 	cg->chip.dbg_show = crystalcove_gpio_dbg_show;
@@ -346,7 +359,7 @@ static int crystalcove_gpio_probe(struct platform_device *pdev)
 	return 0;
 
 out_remove_gpio:
-	WARN_ON(gpiochip_remove(&cg->chip));
+	gpiochip_remove(&cg->chip);
 	return retval;
 }
 
@@ -354,14 +367,11 @@ static int crystalcove_gpio_remove(struct platform_device *pdev)
 {
 	struct crystalcove_gpio *cg = platform_get_drvdata(pdev);
 	int irq = platform_get_irq(pdev, 0);
-	int err;
-
-	err = gpiochip_remove(&cg->chip);
 
+	gpiochip_remove(&cg->chip);
 	if (irq >= 0)
 		free_irq(irq, cg);
-
-	return err;
+	return 0;
 }
 
 static struct platform_driver crystalcove_gpio_driver = {
@@ -369,7 +379,6 @@ static struct platform_driver crystalcove_gpio_driver = {
 	.remove = crystalcove_gpio_remove,
 	.driver = {
 		.name = "crystal_cove_gpio",
-		.owner = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/gpio/gpio-cs5535.c b/drivers/gpio/gpio-cs5535.c
index 92ec58fa9236..7b0b198a563d 100644
--- a/drivers/gpio/gpio-cs5535.c
+++ b/drivers/gpio/gpio-cs5535.c
@@ -201,7 +201,8 @@ EXPORT_SYMBOL_GPL(cs5535_gpio_setup_event);
 
 static int chip_gpio_request(struct gpio_chip *c, unsigned offset)
 {
-	struct cs5535_gpio_chip *chip = (struct cs5535_gpio_chip *) c;
+	struct cs5535_gpio_chip *chip =
+		container_of(c, struct cs5535_gpio_chip, chip);
 	unsigned long flags;
 
 	spin_lock_irqsave(&chip->lock, flags);
@@ -241,7 +242,8 @@ static void chip_gpio_set(struct gpio_chip *chip, unsigned offset, int val)
 
 static int chip_direction_input(struct gpio_chip *c, unsigned offset)
 {
-	struct cs5535_gpio_chip *chip = (struct cs5535_gpio_chip *) c;
+	struct cs5535_gpio_chip *chip =
+		container_of(c, struct cs5535_gpio_chip, chip);
 	unsigned long flags;
 
 	spin_lock_irqsave(&chip->lock, flags);
@@ -254,7 +256,8 @@ static int chip_direction_input(struct gpio_chip *c, unsigned offset)
 
 static int chip_direction_output(struct gpio_chip *c, unsigned offset, int val)
 {
-	struct cs5535_gpio_chip *chip = (struct cs5535_gpio_chip *) c;
+	struct cs5535_gpio_chip *chip =
+		container_of(c, struct cs5535_gpio_chip, chip);
 	unsigned long flags;
 
 	spin_lock_irqsave(&chip->lock, flags);
@@ -319,7 +322,8 @@ static int cs5535_gpio_probe(struct platform_device *pdev)
 		goto done;
 	}
 
-	if (!request_region(res->start, resource_size(res), pdev->name)) {
+	if (!devm_request_region(&pdev->dev, res->start, resource_size(res),
+				 pdev->name)) {
 		dev_err(&pdev->dev, "can't request region\n");
 		goto done;
 	}
@@ -345,31 +349,24 @@ static int cs5535_gpio_probe(struct platform_device *pdev)
 	/* finally, register with the generic GPIO API */
 	err = gpiochip_add(&cs5535_gpio_chip.chip);
 	if (err)
-		goto release_region;
+		goto done;
 
 	return 0;
 
-release_region:
-	release_region(res->start, resource_size(res));
 done:
 	return err;
 }
 
 static int cs5535_gpio_remove(struct platform_device *pdev)
 {
-	struct resource *r;
-
 	gpiochip_remove(&cs5535_gpio_chip.chip);
 
-	r = platform_get_resource(pdev, IORESOURCE_IO, 0);
-	release_region(r->start, resource_size(r));
 	return 0;
 }
 
 static struct platform_driver cs5535_gpio_driver = {
 	.driver = {
 		.name = DRV_NAME,
-		.owner = THIS_MODULE,
 	},
 	.probe = cs5535_gpio_probe,
 	.remove = cs5535_gpio_remove,
diff --git a/drivers/gpio/gpio-da9052.c b/drivers/gpio/gpio-da9052.c
index c5bccd4dec96..389a4d2a4926 100644
--- a/drivers/gpio/gpio-da9052.c
+++ b/drivers/gpio/gpio-da9052.c
@@ -246,7 +246,6 @@ static struct platform_driver da9052_gpio_driver = {
 	.remove = da9052_gpio_remove,
 	.driver = {
 		.name	= "da9052-gpio",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/gpio/gpio-da9055.c b/drivers/gpio/gpio-da9055.c
index 9167c4331081..b8d757036887 100644
--- a/drivers/gpio/gpio-da9055.c
+++ b/drivers/gpio/gpio-da9055.c
@@ -183,7 +183,6 @@ static struct platform_driver da9055_gpio_driver = {
 	.remove = da9055_gpio_remove,
 	.driver = {
 		.name	= "da9055-gpio",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/gpio/gpio-davinci.c b/drivers/gpio/gpio-davinci.c
index 9f0682534e2f..c5e05c82d67c 100644
--- a/drivers/gpio/gpio-davinci.c
+++ b/drivers/gpio/gpio-davinci.c
@@ -234,11 +234,6 @@ static int davinci_gpio_probe(struct platform_device *pdev)
 		return -ENOMEM;
 
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	if (!res) {
-		dev_err(dev, "Invalid memory resource\n");
-		return -EBUSY;
-	}
-
 	gpio_base = devm_ioremap_resource(dev, res);
 	if (IS_ERR(gpio_base))
 		return PTR_ERR(gpio_base);
@@ -619,7 +614,6 @@ static struct platform_driver davinci_gpio_driver = {
 	.probe		= davinci_gpio_probe,
 	.driver		= {
 		.name		= "davinci_gpio",
-		.owner		= THIS_MODULE,
 		.of_match_table	= of_match_ptr(davinci_gpio_ids),
 	},
 };
diff --git a/drivers/gpio/gpio-dln2.c b/drivers/gpio/gpio-dln2.c
new file mode 100644
index 000000000000..ce3c1558cb0a
--- /dev/null
+++ b/drivers/gpio/gpio-dln2.c
@@ -0,0 +1,531 @@
+/*
+ * Driver for the Diolan DLN-2 USB-GPIO adapter
+ *
+ * Copyright (c) 2014 Intel Corporation
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation, version 2.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/types.h>
+#include <linux/irqdomain.h>
+#include <linux/irq.h>
+#include <linux/irqchip/chained_irq.h>
+#include <linux/gpio.h>
+#include <linux/gpio/driver.h>
+#include <linux/platform_device.h>
+#include <linux/mfd/dln2.h>
+
+#define DLN2_GPIO_ID			0x01
+
+#define DLN2_GPIO_GET_PIN_COUNT		DLN2_CMD(0x01, DLN2_GPIO_ID)
+#define DLN2_GPIO_SET_DEBOUNCE		DLN2_CMD(0x04, DLN2_GPIO_ID)
+#define DLN2_GPIO_GET_DEBOUNCE		DLN2_CMD(0x05, DLN2_GPIO_ID)
+#define DLN2_GPIO_PORT_GET_VAL		DLN2_CMD(0x06, DLN2_GPIO_ID)
+#define DLN2_GPIO_PIN_GET_VAL		DLN2_CMD(0x0B, DLN2_GPIO_ID)
+#define DLN2_GPIO_PIN_SET_OUT_VAL	DLN2_CMD(0x0C, DLN2_GPIO_ID)
+#define DLN2_GPIO_PIN_GET_OUT_VAL	DLN2_CMD(0x0D, DLN2_GPIO_ID)
+#define DLN2_GPIO_CONDITION_MET_EV	DLN2_CMD(0x0F, DLN2_GPIO_ID)
+#define DLN2_GPIO_PIN_ENABLE		DLN2_CMD(0x10, DLN2_GPIO_ID)
+#define DLN2_GPIO_PIN_DISABLE		DLN2_CMD(0x11, DLN2_GPIO_ID)
+#define DLN2_GPIO_PIN_SET_DIRECTION	DLN2_CMD(0x13, DLN2_GPIO_ID)
+#define DLN2_GPIO_PIN_GET_DIRECTION	DLN2_CMD(0x14, DLN2_GPIO_ID)
+#define DLN2_GPIO_PIN_SET_EVENT_CFG	DLN2_CMD(0x1E, DLN2_GPIO_ID)
+#define DLN2_GPIO_PIN_GET_EVENT_CFG	DLN2_CMD(0x1F, DLN2_GPIO_ID)
+
+#define DLN2_GPIO_EVENT_NONE		0
+#define DLN2_GPIO_EVENT_CHANGE		1
+#define DLN2_GPIO_EVENT_LVL_HIGH	2
+#define DLN2_GPIO_EVENT_LVL_LOW		3
+#define DLN2_GPIO_EVENT_CHANGE_RISING	0x11
+#define DLN2_GPIO_EVENT_CHANGE_FALLING  0x21
+#define DLN2_GPIO_EVENT_MASK		0x0F
+
+#define DLN2_GPIO_MAX_PINS 32
+
+struct dln2_gpio {
+	struct platform_device *pdev;
+	struct gpio_chip gpio;
+
+	/*
+	 * Cache pin direction to save us one transfer, since the hardware has
+	 * separate commands to read the in and out values.
+	 */
+	DECLARE_BITMAP(output_enabled, DLN2_GPIO_MAX_PINS);
+
+	/* active IRQs - not synced to hardware */
+	DECLARE_BITMAP(unmasked_irqs, DLN2_GPIO_MAX_PINS);
+	/* active IRQS - synced to hardware */
+	DECLARE_BITMAP(enabled_irqs, DLN2_GPIO_MAX_PINS);
+	int irq_type[DLN2_GPIO_MAX_PINS];
+	struct mutex irq_lock;
+};
+
+struct dln2_gpio_pin {
+	__le16 pin;
+};
+
+struct dln2_gpio_pin_val {
+	__le16 pin __packed;
+	u8 value;
+};
+
+static int dln2_gpio_get_pin_count(struct platform_device *pdev)
+{
+	int ret;
+	__le16 count;
+	int len = sizeof(count);
+
+	ret = dln2_transfer_rx(pdev, DLN2_GPIO_GET_PIN_COUNT, &count, &len);
+	if (ret < 0)
+		return ret;
+	if (len < sizeof(count))
+		return -EPROTO;
+
+	return le16_to_cpu(count);
+}
+
+static int dln2_gpio_pin_cmd(struct dln2_gpio *dln2, int cmd, unsigned pin)
+{
+	struct dln2_gpio_pin req = {
+		.pin = cpu_to_le16(pin),
+	};
+
+	return dln2_transfer_tx(dln2->pdev, cmd, &req, sizeof(req));
+}
+
+static int dln2_gpio_pin_val(struct dln2_gpio *dln2, int cmd, unsigned int pin)
+{
+	int ret;
+	struct dln2_gpio_pin req = {
+		.pin = cpu_to_le16(pin),
+	};
+	struct dln2_gpio_pin_val rsp;
+	int len = sizeof(rsp);
+
+	ret = dln2_transfer(dln2->pdev, cmd, &req, sizeof(req), &rsp, &len);
+	if (ret < 0)
+		return ret;
+	if (len < sizeof(rsp) || req.pin != rsp.pin)
+		return -EPROTO;
+
+	return rsp.value;
+}
+
+static int dln2_gpio_pin_get_in_val(struct dln2_gpio *dln2, unsigned int pin)
+{
+	int ret;
+
+	ret = dln2_gpio_pin_val(dln2, DLN2_GPIO_PIN_GET_VAL, pin);
+	if (ret < 0)
+		return ret;
+	return !!ret;
+}
+
+static int dln2_gpio_pin_get_out_val(struct dln2_gpio *dln2, unsigned int pin)
+{
+	int ret;
+
+	ret = dln2_gpio_pin_val(dln2, DLN2_GPIO_PIN_GET_OUT_VAL, pin);
+	if (ret < 0)
+		return ret;
+	return !!ret;
+}
+
+static int dln2_gpio_pin_set_out_val(struct dln2_gpio *dln2,
+				     unsigned int pin, int value)
+{
+	struct dln2_gpio_pin_val req = {
+		.pin = cpu_to_le16(pin),
+		.value = value,
+	};
+
+	return dln2_transfer_tx(dln2->pdev, DLN2_GPIO_PIN_SET_OUT_VAL, &req,
+				sizeof(req));
+}
+
+#define DLN2_GPIO_DIRECTION_IN		0
+#define DLN2_GPIO_DIRECTION_OUT		1
+
+static int dln2_gpio_request(struct gpio_chip *chip, unsigned offset)
+{
+	struct dln2_gpio *dln2 = container_of(chip, struct dln2_gpio, gpio);
+	struct dln2_gpio_pin req = {
+		.pin = cpu_to_le16(offset),
+	};
+	struct dln2_gpio_pin_val rsp;
+	int len = sizeof(rsp);
+	int ret;
+
+	ret = dln2_gpio_pin_cmd(dln2, DLN2_GPIO_PIN_ENABLE, offset);
+	if (ret < 0)
+		return ret;
+
+	/* cache the pin direction */
+	ret = dln2_transfer(dln2->pdev, DLN2_GPIO_PIN_GET_DIRECTION,
+			    &req, sizeof(req), &rsp, &len);
+	if (ret < 0)
+		return ret;
+	if (len < sizeof(rsp) || req.pin != rsp.pin) {
+		ret = -EPROTO;
+		goto out_disable;
+	}
+
+	switch (rsp.value) {
+	case DLN2_GPIO_DIRECTION_IN:
+		clear_bit(offset, dln2->output_enabled);
+		return 0;
+	case DLN2_GPIO_DIRECTION_OUT:
+		set_bit(offset, dln2->output_enabled);
+		return 0;
+	default:
+		ret = -EPROTO;
+		goto out_disable;
+	}
+
+out_disable:
+	dln2_gpio_pin_cmd(dln2, DLN2_GPIO_PIN_DISABLE, offset);
+	return ret;
+}
+
+static void dln2_gpio_free(struct gpio_chip *chip, unsigned offset)
+{
+	struct dln2_gpio *dln2 = container_of(chip, struct dln2_gpio, gpio);
+
+	dln2_gpio_pin_cmd(dln2, DLN2_GPIO_PIN_DISABLE, offset);
+}
+
+static int dln2_gpio_get_direction(struct gpio_chip *chip, unsigned offset)
+{
+	struct dln2_gpio *dln2 = container_of(chip, struct dln2_gpio, gpio);
+
+	if (test_bit(offset, dln2->output_enabled))
+		return GPIOF_DIR_OUT;
+
+	return GPIOF_DIR_IN;
+}
+
+static int dln2_gpio_get(struct gpio_chip *chip, unsigned int offset)
+{
+	struct dln2_gpio *dln2 = container_of(chip, struct dln2_gpio, gpio);
+	int dir;
+
+	dir = dln2_gpio_get_direction(chip, offset);
+	if (dir < 0)
+		return dir;
+
+	if (dir == GPIOF_DIR_IN)
+		return dln2_gpio_pin_get_in_val(dln2, offset);
+
+	return dln2_gpio_pin_get_out_val(dln2, offset);
+}
+
+static void dln2_gpio_set(struct gpio_chip *chip, unsigned offset, int value)
+{
+	struct dln2_gpio *dln2 = container_of(chip, struct dln2_gpio, gpio);
+
+	dln2_gpio_pin_set_out_val(dln2, offset, value);
+}
+
+static int dln2_gpio_set_direction(struct gpio_chip *chip, unsigned offset,
+				   unsigned dir)
+{
+	struct dln2_gpio *dln2 = container_of(chip, struct dln2_gpio, gpio);
+	struct dln2_gpio_pin_val req = {
+		.pin = cpu_to_le16(offset),
+		.value = dir,
+	};
+	int ret;
+
+	ret = dln2_transfer_tx(dln2->pdev, DLN2_GPIO_PIN_SET_DIRECTION,
+			       &req, sizeof(req));
+	if (ret < 0)
+		return ret;
+
+	if (dir == DLN2_GPIO_DIRECTION_OUT)
+		set_bit(offset, dln2->output_enabled);
+	else
+		clear_bit(offset, dln2->output_enabled);
+
+	return ret;
+}
+
+static int dln2_gpio_direction_input(struct gpio_chip *chip, unsigned offset)
+{
+	return dln2_gpio_set_direction(chip, offset, DLN2_GPIO_DIRECTION_IN);
+}
+
+static int dln2_gpio_direction_output(struct gpio_chip *chip, unsigned offset,
+				      int value)
+{
+	struct dln2_gpio *dln2 = container_of(chip, struct dln2_gpio, gpio);
+	int ret;
+
+	ret = dln2_gpio_pin_set_out_val(dln2, offset, value);
+	if (ret < 0)
+		return ret;
+
+	return dln2_gpio_set_direction(chip, offset, DLN2_GPIO_DIRECTION_OUT);
+}
+
+static int dln2_gpio_set_debounce(struct gpio_chip *chip, unsigned offset,
+				  unsigned debounce)
+{
+	struct dln2_gpio *dln2 = container_of(chip, struct dln2_gpio, gpio);
+	__le32 duration = cpu_to_le32(debounce);
+
+	return dln2_transfer_tx(dln2->pdev, DLN2_GPIO_SET_DEBOUNCE,
+				&duration, sizeof(duration));
+}
+
+static int dln2_gpio_set_event_cfg(struct dln2_gpio *dln2, unsigned pin,
+				   unsigned type, unsigned period)
+{
+	struct {
+		__le16 pin;
+		u8 type;
+		__le16 period;
+	} __packed req = {
+		.pin = cpu_to_le16(pin),
+		.type = type,
+		.period = cpu_to_le16(period),
+	};
+
+	return dln2_transfer_tx(dln2->pdev, DLN2_GPIO_PIN_SET_EVENT_CFG,
+				&req, sizeof(req));
+}
+
+static void dln2_irq_unmask(struct irq_data *irqd)
+{
+	struct gpio_chip *gc = irq_data_get_irq_chip_data(irqd);
+	struct dln2_gpio *dln2 = container_of(gc, struct dln2_gpio, gpio);
+	int pin = irqd_to_hwirq(irqd);
+
+	set_bit(pin, dln2->unmasked_irqs);
+}
+
+static void dln2_irq_mask(struct irq_data *irqd)
+{
+	struct gpio_chip *gc = irq_data_get_irq_chip_data(irqd);
+	struct dln2_gpio *dln2 = container_of(gc, struct dln2_gpio, gpio);
+	int pin = irqd_to_hwirq(irqd);
+
+	clear_bit(pin, dln2->unmasked_irqs);
+}
+
+static int dln2_irq_set_type(struct irq_data *irqd, unsigned type)
+{
+	struct gpio_chip *gc = irq_data_get_irq_chip_data(irqd);
+	struct dln2_gpio *dln2 = container_of(gc, struct dln2_gpio, gpio);
+	int pin = irqd_to_hwirq(irqd);
+
+	switch (type) {
+	case IRQ_TYPE_LEVEL_HIGH:
+		dln2->irq_type[pin] = DLN2_GPIO_EVENT_LVL_HIGH;
+		break;
+	case IRQ_TYPE_LEVEL_LOW:
+		dln2->irq_type[pin] = DLN2_GPIO_EVENT_LVL_LOW;
+		break;
+	case IRQ_TYPE_EDGE_BOTH:
+		dln2->irq_type[pin] = DLN2_GPIO_EVENT_CHANGE;
+		break;
+	case IRQ_TYPE_EDGE_RISING:
+		dln2->irq_type[pin] = DLN2_GPIO_EVENT_CHANGE_RISING;
+		break;
+	case IRQ_TYPE_EDGE_FALLING:
+		dln2->irq_type[pin] = DLN2_GPIO_EVENT_CHANGE_FALLING;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static void dln2_irq_bus_lock(struct irq_data *irqd)
+{
+	struct gpio_chip *gc = irq_data_get_irq_chip_data(irqd);
+	struct dln2_gpio *dln2 = container_of(gc, struct dln2_gpio, gpio);
+
+	mutex_lock(&dln2->irq_lock);
+}
+
+static void dln2_irq_bus_unlock(struct irq_data *irqd)
+{
+	struct gpio_chip *gc = irq_data_get_irq_chip_data(irqd);
+	struct dln2_gpio *dln2 = container_of(gc, struct dln2_gpio, gpio);
+	int pin = irqd_to_hwirq(irqd);
+	int enabled, unmasked;
+	unsigned type;
+	int ret;
+
+	enabled = test_bit(pin, dln2->enabled_irqs);
+	unmasked = test_bit(pin, dln2->unmasked_irqs);
+
+	if (enabled != unmasked) {
+		if (unmasked) {
+			type = dln2->irq_type[pin] & DLN2_GPIO_EVENT_MASK;
+			set_bit(pin, dln2->enabled_irqs);
+		} else {
+			type = DLN2_GPIO_EVENT_NONE;
+			clear_bit(pin, dln2->enabled_irqs);
+		}
+
+		ret = dln2_gpio_set_event_cfg(dln2, pin, type, 0);
+		if (ret)
+			dev_err(dln2->gpio.dev, "failed to set event\n");
+	}
+
+	mutex_unlock(&dln2->irq_lock);
+}
+
+static struct irq_chip dln2_gpio_irqchip = {
+	.name = "dln2-irq",
+	.irq_mask = dln2_irq_mask,
+	.irq_unmask = dln2_irq_unmask,
+	.irq_set_type = dln2_irq_set_type,
+	.irq_bus_lock = dln2_irq_bus_lock,
+	.irq_bus_sync_unlock = dln2_irq_bus_unlock,
+};
+
+static void dln2_gpio_event(struct platform_device *pdev, u16 echo,
+			    const void *data, int len)
+{
+	int pin, irq;
+	const struct {
+		__le16 count;
+		__u8 type;
+		__le16 pin;
+		__u8 value;
+	} __packed *event = data;
+	struct dln2_gpio *dln2 = platform_get_drvdata(pdev);
+
+	if (len < sizeof(*event)) {
+		dev_err(dln2->gpio.dev, "short event message\n");
+		return;
+	}
+
+	pin = le16_to_cpu(event->pin);
+	if (pin >= dln2->gpio.ngpio) {
+		dev_err(dln2->gpio.dev, "out of bounds pin %d\n", pin);
+		return;
+	}
+
+	irq = irq_find_mapping(dln2->gpio.irqdomain, pin);
+	if (!irq) {
+		dev_err(dln2->gpio.dev, "pin %d not mapped to IRQ\n", pin);
+		return;
+	}
+
+	switch (dln2->irq_type[pin]) {
+	case DLN2_GPIO_EVENT_CHANGE_RISING:
+		if (event->value)
+			generic_handle_irq(irq);
+		break;
+	case DLN2_GPIO_EVENT_CHANGE_FALLING:
+		if (!event->value)
+			generic_handle_irq(irq);
+		break;
+	default:
+		generic_handle_irq(irq);
+	}
+}
+
+static int dln2_gpio_probe(struct platform_device *pdev)
+{
+	struct dln2_gpio *dln2;
+	struct device *dev = &pdev->dev;
+	int pins;
+	int ret;
+
+	pins = dln2_gpio_get_pin_count(pdev);
+	if (pins < 0) {
+		dev_err(dev, "failed to get pin count: %d\n", pins);
+		return pins;
+	}
+	if (pins > DLN2_GPIO_MAX_PINS) {
+		pins = DLN2_GPIO_MAX_PINS;
+		dev_warn(dev, "clamping pins to %d\n", DLN2_GPIO_MAX_PINS);
+	}
+
+	dln2 = devm_kzalloc(&pdev->dev, sizeof(*dln2), GFP_KERNEL);
+	if (!dln2)
+		return -ENOMEM;
+
+	mutex_init(&dln2->irq_lock);
+
+	dln2->pdev = pdev;
+
+	dln2->gpio.label = "dln2";
+	dln2->gpio.dev = dev;
+	dln2->gpio.owner = THIS_MODULE;
+	dln2->gpio.base = -1;
+	dln2->gpio.ngpio = pins;
+	dln2->gpio.exported = true;
+	dln2->gpio.can_sleep = true;
+	dln2->gpio.irq_not_threaded = true;
+	dln2->gpio.set = dln2_gpio_set;
+	dln2->gpio.get = dln2_gpio_get;
+	dln2->gpio.request = dln2_gpio_request;
+	dln2->gpio.free = dln2_gpio_free;
+	dln2->gpio.get_direction = dln2_gpio_get_direction;
+	dln2->gpio.direction_input = dln2_gpio_direction_input;
+	dln2->gpio.direction_output = dln2_gpio_direction_output;
+	dln2->gpio.set_debounce = dln2_gpio_set_debounce;
+
+	platform_set_drvdata(pdev, dln2);
+
+	ret = gpiochip_add(&dln2->gpio);
+	if (ret < 0) {
+		dev_err(dev, "failed to add gpio chip: %d\n", ret);
+		goto out;
+	}
+
+	ret = gpiochip_irqchip_add(&dln2->gpio, &dln2_gpio_irqchip, 0,
+				   handle_simple_irq, IRQ_TYPE_NONE);
+	if (ret < 0) {
+		dev_err(dev, "failed to add irq chip: %d\n", ret);
+		goto out_gpiochip_remove;
+	}
+
+	ret = dln2_register_event_cb(pdev, DLN2_GPIO_CONDITION_MET_EV,
+				     dln2_gpio_event);
+	if (ret) {
+		dev_err(dev, "failed to register event cb: %d\n", ret);
+		goto out_gpiochip_remove;
+	}
+
+	return 0;
+
+out_gpiochip_remove:
+	gpiochip_remove(&dln2->gpio);
+out:
+	return ret;
+}
+
+static int dln2_gpio_remove(struct platform_device *pdev)
+{
+	struct dln2_gpio *dln2 = platform_get_drvdata(pdev);
+
+	dln2_unregister_event_cb(pdev, DLN2_GPIO_CONDITION_MET_EV);
+	gpiochip_remove(&dln2->gpio);
+
+	return 0;
+}
+
+static struct platform_driver dln2_gpio_driver = {
+	.driver.name	= "dln2-gpio",
+	.probe		= dln2_gpio_probe,
+	.remove		= dln2_gpio_remove,
+};
+
+module_platform_driver(dln2_gpio_driver);
+
+MODULE_AUTHOR("Daniel Baluta <daniel.baluta@intel.com");
+MODULE_DESCRIPTION("Driver for the Diolan DLN2 GPIO interface");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS("platform:dln2-gpio");
diff --git a/drivers/gpio/gpio-dwapb.c b/drivers/gpio/gpio-dwapb.c
index d6618a6e2399..b4eb6a657d34 100644
--- a/drivers/gpio/gpio-dwapb.c
+++ b/drivers/gpio/gpio-dwapb.c
@@ -21,6 +21,8 @@
 #include <linux/of_irq.h>
 #include <linux/platform_device.h>
 #include <linux/spinlock.h>
+#include <linux/platform_data/gpio-dwapb.h>
+#include <linux/slab.h>
 
 #define GPIO_SWPORTA_DR		0x00
 #define GPIO_SWPORTA_DDR	0x04
@@ -35,6 +37,7 @@
 #define GPIO_INTTYPE_LEVEL	0x38
 #define GPIO_INT_POLARITY	0x3c
 #define GPIO_INTSTATUS		0x40
+#define GPIO_PORTA_DEBOUNCE	0x48
 #define GPIO_PORTA_EOI		0x4c
 #define GPIO_EXT_PORTA		0x50
 #define GPIO_EXT_PORTB		0x54
@@ -48,10 +51,28 @@
 
 struct dwapb_gpio;
 
+#ifdef CONFIG_PM_SLEEP
+/* Store GPIO context across system-wide suspend/resume transitions */
+struct dwapb_context {
+	u32 data;
+	u32 dir;
+	u32 ext;
+	u32 int_en;
+	u32 int_mask;
+	u32 int_type;
+	u32 int_pol;
+	u32 int_deb;
+};
+#endif
+
 struct dwapb_gpio_port {
 	struct bgpio_chip	bgc;
 	bool			is_registered;
 	struct dwapb_gpio	*gpio;
+#ifdef CONFIG_PM_SLEEP
+	struct dwapb_context	*ctx;
+#endif
+	unsigned int		idx;
 };
 
 struct dwapb_gpio {
@@ -62,11 +83,33 @@ struct dwapb_gpio {
 	struct irq_domain	*domain;
 };
 
+static inline struct dwapb_gpio_port *
+to_dwapb_gpio_port(struct bgpio_chip *bgc)
+{
+	return container_of(bgc, struct dwapb_gpio_port, bgc);
+}
+
+static inline u32 dwapb_read(struct dwapb_gpio *gpio, unsigned int offset)
+{
+	struct bgpio_chip *bgc	= &gpio->ports[0].bgc;
+	void __iomem *reg_base	= gpio->regs;
+
+	return bgc->read_reg(reg_base + offset);
+}
+
+static inline void dwapb_write(struct dwapb_gpio *gpio, unsigned int offset,
+			       u32 val)
+{
+	struct bgpio_chip *bgc	= &gpio->ports[0].bgc;
+	void __iomem *reg_base	= gpio->regs;
+
+	bgc->write_reg(reg_base + offset, val);
+}
+
 static int dwapb_gpio_to_irq(struct gpio_chip *gc, unsigned offset)
 {
 	struct bgpio_chip *bgc = to_bgpio_chip(gc);
-	struct dwapb_gpio_port *port = container_of(bgc, struct
-						    dwapb_gpio_port, bgc);
+	struct dwapb_gpio_port *port = to_dwapb_gpio_port(bgc);
 	struct dwapb_gpio *gpio = port->gpio;
 
 	return irq_find_mapping(gpio->domain, offset);
@@ -74,21 +117,20 @@ static int dwapb_gpio_to_irq(struct gpio_chip *gc, unsigned offset)
 
 static void dwapb_toggle_trigger(struct dwapb_gpio *gpio, unsigned int offs)
 {
-	u32 v = readl(gpio->regs + GPIO_INT_POLARITY);
+	u32 v = dwapb_read(gpio, GPIO_INT_POLARITY);
 
 	if (gpio_get_value(gpio->ports[0].bgc.gc.base + offs))
 		v &= ~BIT(offs);
 	else
 		v |= BIT(offs);
 
-	writel(v, gpio->regs + GPIO_INT_POLARITY);
+	dwapb_write(gpio, GPIO_INT_POLARITY, v);
 }
 
-static void dwapb_irq_handler(u32 irq, struct irq_desc *desc)
+static u32 dwapb_do_irq(struct dwapb_gpio *gpio)
 {
-	struct dwapb_gpio *gpio = irq_get_handler_data(irq);
-	struct irq_chip *chip = irq_desc_get_chip(desc);
 	u32 irq_status = readl_relaxed(gpio->regs + GPIO_INTSTATUS);
+	u32 ret = irq_status;
 
 	while (irq_status) {
 		int hwirq = fls(irq_status) - 1;
@@ -102,6 +144,16 @@ static void dwapb_irq_handler(u32 irq, struct irq_desc *desc)
 			dwapb_toggle_trigger(gpio, hwirq);
 	}
 
+	return ret;
+}
+
+static void dwapb_irq_handler(u32 irq, struct irq_desc *desc)
+{
+	struct dwapb_gpio *gpio = irq_get_handler_data(irq);
+	struct irq_chip *chip = irq_desc_get_chip(desc);
+
+	dwapb_do_irq(gpio);
+
 	if (chip->irq_eoi)
 		chip->irq_eoi(irq_desc_get_irq_data(desc));
 }
@@ -115,9 +167,9 @@ static void dwapb_irq_enable(struct irq_data *d)
 	u32 val;
 
 	spin_lock_irqsave(&bgc->lock, flags);
-	val = readl(gpio->regs + GPIO_INTEN);
+	val = dwapb_read(gpio, GPIO_INTEN);
 	val |= BIT(d->hwirq);
-	writel(val, gpio->regs + GPIO_INTEN);
+	dwapb_write(gpio, GPIO_INTEN, val);
 	spin_unlock_irqrestore(&bgc->lock, flags);
 }
 
@@ -130,9 +182,9 @@ static void dwapb_irq_disable(struct irq_data *d)
 	u32 val;
 
 	spin_lock_irqsave(&bgc->lock, flags);
-	val = readl(gpio->regs + GPIO_INTEN);
+	val = dwapb_read(gpio, GPIO_INTEN);
 	val &= ~BIT(d->hwirq);
-	writel(val, gpio->regs + GPIO_INTEN);
+	dwapb_write(gpio, GPIO_INTEN, val);
 	spin_unlock_irqrestore(&bgc->lock, flags);
 }
 
@@ -142,7 +194,7 @@ static int dwapb_irq_reqres(struct irq_data *d)
 	struct dwapb_gpio *gpio = igc->private;
 	struct bgpio_chip *bgc = &gpio->ports[0].bgc;
 
-	if (gpio_lock_as_irq(&bgc->gc, irqd_to_hwirq(d))) {
+	if (gpiochip_lock_as_irq(&bgc->gc, irqd_to_hwirq(d))) {
 		dev_err(gpio->dev, "unable to lock HW IRQ %lu for IRQ\n",
 			irqd_to_hwirq(d));
 		return -EINVAL;
@@ -156,7 +208,7 @@ static void dwapb_irq_relres(struct irq_data *d)
 	struct dwapb_gpio *gpio = igc->private;
 	struct bgpio_chip *bgc = &gpio->ports[0].bgc;
 
-	gpio_unlock_as_irq(&bgc->gc, irqd_to_hwirq(d));
+	gpiochip_unlock_as_irq(&bgc->gc, irqd_to_hwirq(d));
 }
 
 static int dwapb_irq_set_type(struct irq_data *d, u32 type)
@@ -172,8 +224,8 @@ static int dwapb_irq_set_type(struct irq_data *d, u32 type)
 		return -EINVAL;
 
 	spin_lock_irqsave(&bgc->lock, flags);
-	level = readl(gpio->regs + GPIO_INTTYPE_LEVEL);
-	polarity = readl(gpio->regs + GPIO_INT_POLARITY);
+	level = dwapb_read(gpio, GPIO_INTTYPE_LEVEL);
+	polarity = dwapb_read(gpio, GPIO_INT_POLARITY);
 
 	switch (type) {
 	case IRQ_TYPE_EDGE_BOTH:
@@ -200,29 +252,55 @@ static int dwapb_irq_set_type(struct irq_data *d, u32 type)
 
 	irq_setup_alt_chip(d, type);
 
-	writel(level, gpio->regs + GPIO_INTTYPE_LEVEL);
-	writel(polarity, gpio->regs + GPIO_INT_POLARITY);
+	dwapb_write(gpio, GPIO_INTTYPE_LEVEL, level);
+	dwapb_write(gpio, GPIO_INT_POLARITY, polarity);
+	spin_unlock_irqrestore(&bgc->lock, flags);
+
+	return 0;
+}
+
+static int dwapb_gpio_set_debounce(struct gpio_chip *gc,
+				   unsigned offset, unsigned debounce)
+{
+	struct bgpio_chip *bgc = to_bgpio_chip(gc);
+	struct dwapb_gpio_port *port = to_dwapb_gpio_port(bgc);
+	struct dwapb_gpio *gpio = port->gpio;
+	unsigned long flags, val_deb;
+	unsigned long mask = bgc->pin2mask(bgc, offset);
+
+	spin_lock_irqsave(&bgc->lock, flags);
+
+	val_deb = dwapb_read(gpio, GPIO_PORTA_DEBOUNCE);
+	if (debounce)
+		dwapb_write(gpio, GPIO_PORTA_DEBOUNCE, val_deb | mask);
+	else
+		dwapb_write(gpio, GPIO_PORTA_DEBOUNCE, val_deb & ~mask);
+
 	spin_unlock_irqrestore(&bgc->lock, flags);
 
 	return 0;
 }
 
+static irqreturn_t dwapb_irq_handler_mfd(int irq, void *dev_id)
+{
+	u32 worked;
+	struct dwapb_gpio *gpio = dev_id;
+
+	worked = dwapb_do_irq(gpio);
+
+	return worked ? IRQ_HANDLED : IRQ_NONE;
+}
+
 static void dwapb_configure_irqs(struct dwapb_gpio *gpio,
-				 struct dwapb_gpio_port *port)
+				 struct dwapb_gpio_port *port,
+				 struct dwapb_port_property *pp)
 {
 	struct gpio_chip *gc = &port->bgc.gc;
-	struct device_node *node =  gc->of_node;
-	struct irq_chip_generic	*irq_gc;
+	struct device_node *node = pp->node;
+	struct irq_chip_generic	*irq_gc = NULL;
 	unsigned int hwirq, ngpio = gc->ngpio;
 	struct irq_chip_type *ct;
-	int err, irq, i;
-
-	irq = irq_of_parse_and_map(node, 0);
-	if (!irq) {
-		dev_warn(gpio->dev, "no irq for bank %s\n",
-			port->bgc.gc.of_node->full_name);
-		return;
-	}
+	int err, i;
 
 	gpio->domain = irq_domain_add_linear(node, ngpio,
 					     &irq_generic_chip_ops, gpio);
@@ -269,8 +347,24 @@ static void dwapb_configure_irqs(struct dwapb_gpio *gpio,
 	irq_gc->chip_types[1].type = IRQ_TYPE_EDGE_BOTH;
 	irq_gc->chip_types[1].handler = handle_edge_irq;
 
-	irq_set_chained_handler(irq, dwapb_irq_handler);
-	irq_set_handler_data(irq, gpio);
+	if (!pp->irq_shared) {
+		irq_set_chained_handler(pp->irq, dwapb_irq_handler);
+		irq_set_handler_data(pp->irq, gpio);
+	} else {
+		/*
+		 * Request a shared IRQ since where MFD would have devices
+		 * using the same irq pin
+		 */
+		err = devm_request_irq(gpio->dev, pp->irq,
+				       dwapb_irq_handler_mfd,
+				       IRQF_SHARED, "gpio-dwapb-mfd", gpio);
+		if (err) {
+			dev_err(gpio->dev, "error requesting IRQ\n");
+			irq_domain_remove(gpio->domain);
+			gpio->domain = NULL;
+			return;
+		}
+	}
 
 	for (hwirq = 0 ; hwirq < ngpio ; hwirq++)
 		irq_create_mapping(gpio->domain, hwirq);
@@ -296,57 +390,53 @@ static void dwapb_irq_teardown(struct dwapb_gpio *gpio)
 }
 
 static int dwapb_gpio_add_port(struct dwapb_gpio *gpio,
-			       struct device_node *port_np,
+			       struct dwapb_port_property *pp,
 			       unsigned int offs)
 {
 	struct dwapb_gpio_port *port;
-	u32 port_idx, ngpio;
 	void __iomem *dat, *set, *dirout;
 	int err;
 
-	if (of_property_read_u32(port_np, "reg", &port_idx) ||
-		port_idx >= DWAPB_MAX_PORTS) {
-		dev_err(gpio->dev, "missing/invalid port index for %s\n",
-			port_np->full_name);
-		return -EINVAL;
-	}
-
 	port = &gpio->ports[offs];
 	port->gpio = gpio;
+	port->idx = pp->idx;
 
-	if (of_property_read_u32(port_np, "snps,nr-gpios", &ngpio)) {
-		dev_info(gpio->dev, "failed to get number of gpios for %s\n",
-			 port_np->full_name);
-		ngpio = 32;
-	}
+#ifdef CONFIG_PM_SLEEP
+	port->ctx = devm_kzalloc(gpio->dev, sizeof(*port->ctx), GFP_KERNEL);
+	if (!port->ctx)
+		return -ENOMEM;
+#endif
 
-	dat = gpio->regs + GPIO_EXT_PORTA + (port_idx * GPIO_EXT_PORT_SIZE);
-	set = gpio->regs + GPIO_SWPORTA_DR + (port_idx * GPIO_SWPORT_DR_SIZE);
+	dat = gpio->regs + GPIO_EXT_PORTA + (pp->idx * GPIO_EXT_PORT_SIZE);
+	set = gpio->regs + GPIO_SWPORTA_DR + (pp->idx * GPIO_SWPORT_DR_SIZE);
 	dirout = gpio->regs + GPIO_SWPORTA_DDR +
-		(port_idx * GPIO_SWPORT_DDR_SIZE);
+		(pp->idx * GPIO_SWPORT_DDR_SIZE);
 
 	err = bgpio_init(&port->bgc, gpio->dev, 4, dat, set, NULL, dirout,
 			 NULL, false);
 	if (err) {
 		dev_err(gpio->dev, "failed to init gpio chip for %s\n",
-			port_np->full_name);
+			pp->name);
 		return err;
 	}
 
-	port->bgc.gc.ngpio = ngpio;
-	port->bgc.gc.of_node = port_np;
+#ifdef CONFIG_OF_GPIO
+	port->bgc.gc.of_node = pp->node;
+#endif
+	port->bgc.gc.ngpio = pp->ngpio;
+	port->bgc.gc.base = pp->gpio_base;
 
-	/*
-	 * Only port A can provide interrupts in all configurations of the IP.
-	 */
-	if (port_idx == 0 &&
-	    of_property_read_bool(port_np, "interrupt-controller"))
-		dwapb_configure_irqs(gpio, port);
+	/* Only port A support debounce */
+	if (pp->idx == 0)
+		port->bgc.gc.set_debounce = dwapb_gpio_set_debounce;
+
+	if (pp->irq)
+		dwapb_configure_irqs(gpio, port, pp);
 
 	err = gpiochip_add(&port->bgc.gc);
 	if (err)
 		dev_err(gpio->dev, "failed to register gpiochip for %s\n",
-			port_np->full_name);
+			pp->name);
 	else
 		port->is_registered = true;
 
@@ -362,25 +452,116 @@ static void dwapb_gpio_unregister(struct dwapb_gpio *gpio)
 			gpiochip_remove(&gpio->ports[m].bgc.gc);
 }
 
+static struct dwapb_platform_data *
+dwapb_gpio_get_pdata_of(struct device *dev)
+{
+	struct device_node *node, *port_np;
+	struct dwapb_platform_data *pdata;
+	struct dwapb_port_property *pp;
+	int nports;
+	int i;
+
+	node = dev->of_node;
+	if (!IS_ENABLED(CONFIG_OF_GPIO) || !node)
+		return ERR_PTR(-ENODEV);
+
+	nports = of_get_child_count(node);
+	if (nports == 0)
+		return ERR_PTR(-ENODEV);
+
+	pdata = kzalloc(sizeof(*pdata), GFP_KERNEL);
+	if (!pdata)
+		return ERR_PTR(-ENOMEM);
+
+	pdata->properties = kcalloc(nports, sizeof(*pp), GFP_KERNEL);
+	if (!pdata->properties) {
+		kfree(pdata);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	pdata->nports = nports;
+
+	i = 0;
+	for_each_child_of_node(node, port_np) {
+		pp = &pdata->properties[i++];
+		pp->node = port_np;
+
+		if (of_property_read_u32(port_np, "reg", &pp->idx) ||
+		    pp->idx >= DWAPB_MAX_PORTS) {
+			dev_err(dev, "missing/invalid port index for %s\n",
+				port_np->full_name);
+			kfree(pdata->properties);
+			kfree(pdata);
+			return ERR_PTR(-EINVAL);
+		}
+
+		if (of_property_read_u32(port_np, "snps,nr-gpios",
+					 &pp->ngpio)) {
+			dev_info(dev, "failed to get number of gpios for %s\n",
+				 port_np->full_name);
+			pp->ngpio = 32;
+		}
+
+		/*
+		 * Only port A can provide interrupts in all configurations of
+		 * the IP.
+		 */
+		if (pp->idx == 0 &&
+		    of_property_read_bool(port_np, "interrupt-controller")) {
+			pp->irq = irq_of_parse_and_map(port_np, 0);
+			if (!pp->irq) {
+				dev_warn(dev, "no irq for bank %s\n",
+					 port_np->full_name);
+			}
+		}
+
+		pp->irq_shared	= false;
+		pp->gpio_base	= -1;
+		pp->name	= port_np->full_name;
+	}
+
+	return pdata;
+}
+
+static inline void dwapb_free_pdata_of(struct dwapb_platform_data *pdata)
+{
+	if (!IS_ENABLED(CONFIG_OF_GPIO) || !pdata)
+		return;
+
+	kfree(pdata->properties);
+	kfree(pdata);
+}
+
 static int dwapb_gpio_probe(struct platform_device *pdev)
 {
+	unsigned int i;
 	struct resource *res;
 	struct dwapb_gpio *gpio;
-	struct device_node *np;
 	int err;
-	unsigned int offs = 0;
+	struct device *dev = &pdev->dev;
+	struct dwapb_platform_data *pdata = dev_get_platdata(dev);
+	bool is_pdata_alloc = !pdata;
+
+	if (is_pdata_alloc) {
+		pdata = dwapb_gpio_get_pdata_of(dev);
+		if (IS_ERR(pdata))
+			return PTR_ERR(pdata);
+	}
 
-	gpio = devm_kzalloc(&pdev->dev, sizeof(*gpio), GFP_KERNEL);
-	if (!gpio)
-		return -ENOMEM;
-	gpio->dev = &pdev->dev;
+	if (!pdata->nports) {
+		err = -ENODEV;
+		goto out_err;
+	}
 
-	gpio->nr_ports = of_get_child_count(pdev->dev.of_node);
-	if (!gpio->nr_ports) {
-		err = -EINVAL;
+	gpio = devm_kzalloc(&pdev->dev, sizeof(*gpio), GFP_KERNEL);
+	if (!gpio) {
+		err = -ENOMEM;
 		goto out_err;
 	}
-	gpio->ports = devm_kzalloc(&pdev->dev, gpio->nr_ports *
+	gpio->dev = &pdev->dev;
+	gpio->nr_ports = pdata->nports;
+
+	gpio->ports = devm_kcalloc(&pdev->dev, gpio->nr_ports,
 				   sizeof(*gpio->ports), GFP_KERNEL);
 	if (!gpio->ports) {
 		err = -ENOMEM;
@@ -394,20 +575,23 @@ static int dwapb_gpio_probe(struct platform_device *pdev)
 		goto out_err;
 	}
 
-	for_each_child_of_node(pdev->dev.of_node, np) {
-		err = dwapb_gpio_add_port(gpio, np, offs++);
+	for (i = 0; i < gpio->nr_ports; i++) {
+		err = dwapb_gpio_add_port(gpio, &pdata->properties[i], i);
 		if (err)
 			goto out_unregister;
 	}
 	platform_set_drvdata(pdev, gpio);
 
-	return 0;
+	goto out_err;
 
 out_unregister:
 	dwapb_gpio_unregister(gpio);
 	dwapb_irq_teardown(gpio);
 
 out_err:
+	if (is_pdata_alloc)
+		dwapb_free_pdata_of(pdata);
+
 	return err;
 }
 
@@ -427,10 +611,99 @@ static const struct of_device_id dwapb_of_match[] = {
 };
 MODULE_DEVICE_TABLE(of, dwapb_of_match);
 
+#ifdef CONFIG_PM_SLEEP
+static int dwapb_gpio_suspend(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct dwapb_gpio *gpio = platform_get_drvdata(pdev);
+	struct bgpio_chip *bgc	= &gpio->ports[0].bgc;
+	unsigned long flags;
+	int i;
+
+	spin_lock_irqsave(&bgc->lock, flags);
+	for (i = 0; i < gpio->nr_ports; i++) {
+		unsigned int offset;
+		unsigned int idx = gpio->ports[i].idx;
+		struct dwapb_context *ctx = gpio->ports[i].ctx;
+
+		BUG_ON(!ctx);
+
+		offset = GPIO_SWPORTA_DDR + idx * GPIO_SWPORT_DDR_SIZE;
+		ctx->dir = dwapb_read(gpio, offset);
+
+		offset = GPIO_SWPORTA_DR + idx * GPIO_SWPORT_DR_SIZE;
+		ctx->data = dwapb_read(gpio, offset);
+
+		offset = GPIO_EXT_PORTA + idx * GPIO_EXT_PORT_SIZE;
+		ctx->ext = dwapb_read(gpio, offset);
+
+		/* Only port A can provide interrupts */
+		if (idx == 0) {
+			ctx->int_mask	= dwapb_read(gpio, GPIO_INTMASK);
+			ctx->int_en	= dwapb_read(gpio, GPIO_INTEN);
+			ctx->int_pol	= dwapb_read(gpio, GPIO_INT_POLARITY);
+			ctx->int_type	= dwapb_read(gpio, GPIO_INTTYPE_LEVEL);
+			ctx->int_deb	= dwapb_read(gpio, GPIO_PORTA_DEBOUNCE);
+
+			/* Mask out interrupts */
+			dwapb_write(gpio, GPIO_INTMASK, 0xffffffff);
+		}
+	}
+	spin_unlock_irqrestore(&bgc->lock, flags);
+
+	return 0;
+}
+
+static int dwapb_gpio_resume(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct dwapb_gpio *gpio = platform_get_drvdata(pdev);
+	struct bgpio_chip *bgc	= &gpio->ports[0].bgc;
+	unsigned long flags;
+	int i;
+
+	spin_lock_irqsave(&bgc->lock, flags);
+	for (i = 0; i < gpio->nr_ports; i++) {
+		unsigned int offset;
+		unsigned int idx = gpio->ports[i].idx;
+		struct dwapb_context *ctx = gpio->ports[i].ctx;
+
+		BUG_ON(!ctx);
+
+		offset = GPIO_SWPORTA_DR + idx * GPIO_SWPORT_DR_SIZE;
+		dwapb_write(gpio, offset, ctx->data);
+
+		offset = GPIO_SWPORTA_DDR + idx * GPIO_SWPORT_DDR_SIZE;
+		dwapb_write(gpio, offset, ctx->dir);
+
+		offset = GPIO_EXT_PORTA + idx * GPIO_EXT_PORT_SIZE;
+		dwapb_write(gpio, offset, ctx->ext);
+
+		/* Only port A can provide interrupts */
+		if (idx == 0) {
+			dwapb_write(gpio, GPIO_INTTYPE_LEVEL, ctx->int_type);
+			dwapb_write(gpio, GPIO_INT_POLARITY, ctx->int_pol);
+			dwapb_write(gpio, GPIO_PORTA_DEBOUNCE, ctx->int_deb);
+			dwapb_write(gpio, GPIO_INTEN, ctx->int_en);
+			dwapb_write(gpio, GPIO_INTMASK, ctx->int_mask);
+
+			/* Clear out spurious interrupts */
+			dwapb_write(gpio, GPIO_PORTA_EOI, 0xffffffff);
+		}
+	}
+	spin_unlock_irqrestore(&bgc->lock, flags);
+
+	return 0;
+}
+#endif
+
+static SIMPLE_DEV_PM_OPS(dwapb_gpio_pm_ops, dwapb_gpio_suspend,
+			 dwapb_gpio_resume);
+
 static struct platform_driver dwapb_gpio_driver = {
 	.driver		= {
 		.name	= "gpio-dwapb",
-		.owner	= THIS_MODULE,
+		.pm	= &dwapb_gpio_pm_ops,
 		.of_match_table = of_match_ptr(dwapb_of_match),
 	},
 	.probe		= dwapb_gpio_probe,
diff --git a/drivers/gpio/gpio-em.c b/drivers/gpio/gpio-em.c
index fe49ec3cdb7d..3cfcfc620c8e 100644
--- a/drivers/gpio/gpio-em.c
+++ b/drivers/gpio/gpio-em.c
@@ -103,7 +103,7 @@ static int em_gio_irq_reqres(struct irq_data *d)
 {
 	struct em_gio_priv *p = irq_data_get_irq_chip_data(d);
 
-	if (gpio_lock_as_irq(&p->gpio_chip, irqd_to_hwirq(d))) {
+	if (gpiochip_lock_as_irq(&p->gpio_chip, irqd_to_hwirq(d))) {
 		dev_err(p->gpio_chip.dev,
 			"unable to lock HW IRQ %lu for IRQ\n",
 			irqd_to_hwirq(d));
@@ -116,7 +116,7 @@ static void em_gio_irq_relres(struct irq_data *d)
 {
 	struct em_gio_priv *p = irq_data_get_irq_chip_data(d);
 
-	gpio_unlock_as_irq(&p->gpio_chip, irqd_to_hwirq(d));
+	gpiochip_unlock_as_irq(&p->gpio_chip, irqd_to_hwirq(d));
 }
 
 
@@ -330,12 +330,7 @@ static int em_gio_probe(struct platform_device *pdev)
 			goto err0;
 		}
 
-		ret = of_alias_get_id(pdev->dev.of_node, "gpio");
-		if (ret < 0) {
-			dev_err(&pdev->dev, "Couldn't get OF id\n");
-			goto err0;
-		}
-		pdata->gpio_base = ret * 32; /* 32 GPIOs per instance */
+		pdata->gpio_base = -1;
 	}
 
 	gpio_chip = &p->gpio_chip;
@@ -428,7 +423,6 @@ static struct platform_driver em_gio_device_driver = {
 	.driver		= {
 		.name	= "em_gio",
 		.of_match_table = em_gio_dt_ids,
-		.owner		= THIS_MODULE,
 	}
 };
 
diff --git a/drivers/gpio/gpio-ep93xx.c b/drivers/gpio/gpio-ep93xx.c
index dcc2bb4074ef..45684f36ddb1 100644
--- a/drivers/gpio/gpio-ep93xx.c
+++ b/drivers/gpio/gpio-ep93xx.c
@@ -374,7 +374,6 @@ static int ep93xx_gpio_probe(struct platform_device *pdev)
 static struct platform_driver ep93xx_gpio_driver = {
 	.driver		= {
 		.name	= "gpio-ep93xx",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= ep93xx_gpio_probe,
 };
diff --git a/drivers/gpio/gpio-f7188x.c b/drivers/gpio/gpio-f7188x.c
index fd3202f968ff..1be291ac6319 100644
--- a/drivers/gpio/gpio-f7188x.c
+++ b/drivers/gpio/gpio-f7188x.c
@@ -417,7 +417,6 @@ err:
 
 static struct platform_driver f7188x_gpio_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= DRVNAME,
 	},
 	.probe		= f7188x_gpio_probe,
diff --git a/drivers/gpio/gpio-ge.c b/drivers/gpio/gpio-ge.c
index 1237a73c3c91..aea5c2a53cc0 100644
--- a/drivers/gpio/gpio-ge.c
+++ b/drivers/gpio/gpio-ge.c
@@ -120,7 +120,6 @@ static int __init gef_gpio_probe(struct platform_device *pdev)
 static struct platform_driver gef_gpio_driver = {
 	.driver = {
 		.name		= "gef-gpio",
-		.owner		= THIS_MODULE,
 		.of_match_table	= gef_gpio_ids,
 	},
 };
diff --git a/drivers/gpio/gpio-grgpio.c b/drivers/gpio/gpio-grgpio.c
index 66ad3df9d9cf..3a5a71050559 100644
--- a/drivers/gpio/gpio-grgpio.c
+++ b/drivers/gpio/gpio-grgpio.c
@@ -441,6 +441,8 @@ static int grgpio_probe(struct platform_device *ofdev)
 	err = gpiochip_add(gc);
 	if (err) {
 		dev_err(&ofdev->dev, "Could not add gpiochip\n");
+		if (priv->domain)
+			irq_domain_remove(priv->domain);
 		return err;
 	}
 
@@ -490,7 +492,6 @@ MODULE_DEVICE_TABLE(of, grgpio_match);
 static struct platform_driver grgpio_driver = {
 	.driver = {
 		.name = "grgpio",
-		.owner = THIS_MODULE,
 		.of_match_table = grgpio_match,
 	},
 	.probe = grgpio_probe,
diff --git a/drivers/gpio/gpio-ich.c b/drivers/gpio/gpio-ich.c
index 3784e81e7762..7818cd1453ae 100644
--- a/drivers/gpio/gpio-ich.c
+++ b/drivers/gpio/gpio-ich.c
@@ -526,7 +526,6 @@ static int ichx_gpio_remove(struct platform_device *pdev)
 
 static struct platform_driver ichx_gpio_driver = {
 	.driver		= {
-		.owner	= THIS_MODULE,
 		.name	= DRV_NAME,
 	},
 	.probe		= ichx_gpio_probe,
diff --git a/drivers/gpio/gpio-iop.c b/drivers/gpio/gpio-iop.c
index 0a5e9d3f308c..2ed0237a8baf 100644
--- a/drivers/gpio/gpio-iop.c
+++ b/drivers/gpio/gpio-iop.c
@@ -120,7 +120,6 @@ static int iop3xx_gpio_probe(struct platform_device *pdev)
 static struct platform_driver iop3xx_gpio_driver = {
 	.driver = {
 		.name = "gpio-iop",
-		.owner = THIS_MODULE,
 	},
 	.probe = iop3xx_gpio_probe,
 };
diff --git a/drivers/gpio/gpio-janz-ttl.c b/drivers/gpio/gpio-janz-ttl.c
index 29ffe22ad97a..3a1664335f5e 100644
--- a/drivers/gpio/gpio-janz-ttl.c
+++ b/drivers/gpio/gpio-janz-ttl.c
@@ -203,7 +203,6 @@ static int ttl_remove(struct platform_device *pdev)
 static struct platform_driver ttl_driver = {
 	.driver		= {
 		.name	= DRV_NAME,
-		.owner	= THIS_MODULE,
 	},
 	.probe		= ttl_probe,
 	.remove		= ttl_remove,
diff --git a/drivers/gpio/gpio-kempld.c b/drivers/gpio/gpio-kempld.c
index fd150adeebf9..443518f63f15 100644
--- a/drivers/gpio/gpio-kempld.c
+++ b/drivers/gpio/gpio-kempld.c
@@ -206,7 +206,6 @@ static int kempld_gpio_remove(struct platform_device *pdev)
 static struct platform_driver kempld_gpio_driver = {
 	.driver = {
 		.name = "kempld-gpio",
-		.owner = THIS_MODULE,
 	},
 	.probe		= kempld_gpio_probe,
 	.remove		= kempld_gpio_remove,
diff --git a/drivers/gpio/gpio-ks8695.c b/drivers/gpio/gpio-ks8695.c
index 464a83de0d6a..cc09b237e88c 100644
--- a/drivers/gpio/gpio-ks8695.c
+++ b/drivers/gpio/gpio-ks8695.c
@@ -265,29 +265,27 @@ static int ks8695_gpio_show(struct seq_file *s, void *unused)
 				seq_printf(s, "EXT%i ", i);
 
 				switch ((ctrl & intmask[i]) >> (4 * i)) {
-					case IOPC_TM_LOW:
-						seq_printf(s, "(Low)");		break;
-					case IOPC_TM_HIGH:
-						seq_printf(s, "(High)");	break;
-					case IOPC_TM_RISING:
-						seq_printf(s, "(Rising)");	break;
-					case IOPC_TM_FALLING:
-						seq_printf(s, "(Falling)");	break;
-					case IOPC_TM_EDGE:
-						seq_printf(s, "(Edges)");	break;
+				case IOPC_TM_LOW:
+					seq_printf(s, "(Low)");		break;
+				case IOPC_TM_HIGH:
+					seq_printf(s, "(High)");	break;
+				case IOPC_TM_RISING:
+					seq_printf(s, "(Rising)");	break;
+				case IOPC_TM_FALLING:
+					seq_printf(s, "(Falling)");	break;
+				case IOPC_TM_EDGE:
+					seq_printf(s, "(Edges)");	break;
 				}
-			}
-			else
+			} else
 				seq_printf(s, "GPIO\t");
-		}
-		else if (i <= KS8695_GPIO_5) {
+		} else if (i <= KS8695_GPIO_5) {
 			if (ctrl & enable[i])
 				seq_printf(s, "TOUT%i\t", i - KS8695_GPIO_4);
 			else
 				seq_printf(s, "GPIO\t");
-		}
-		else
+		} else {
 			seq_printf(s, "GPIO\t");
+		}
 
 		seq_printf(s, "\t");
 
diff --git a/drivers/gpio/gpio-lp3943.c b/drivers/gpio/gpio-lp3943.c
index 6bbdad805b78..cfc5b12b43ad 100644
--- a/drivers/gpio/gpio-lp3943.c
+++ b/drivers/gpio/gpio-lp3943.c
@@ -231,7 +231,6 @@ static struct platform_driver lp3943_gpio_driver = {
 	.remove = lp3943_gpio_remove,
 	.driver = {
 		.name = "lp3943-gpio",
-		.owner = THIS_MODULE,
 		.of_match_table = lp3943_gpio_of_match,
 	},
 };
diff --git a/drivers/gpio/gpio-lpc32xx.c b/drivers/gpio/gpio-lpc32xx.c
index b9b9799b368b..47e2dde63734 100644
--- a/drivers/gpio/gpio-lpc32xx.c
+++ b/drivers/gpio/gpio-lpc32xx.c
@@ -569,7 +569,6 @@ static const struct of_device_id lpc32xx_gpio_of_match[] = {
 static struct platform_driver lpc32xx_gpio_driver = {
 	.driver		= {
 		.name	= "lpc32xx-gpio",
-		.owner	= THIS_MODULE,
 		.of_match_table = of_match_ptr(lpc32xx_gpio_of_match),
 	},
 	.probe		= lpc32xx_gpio_probe,
diff --git a/drivers/gpio/gpio-lynxpoint.c b/drivers/gpio/gpio-lynxpoint.c
index fa945ec9ccff..127c755b38dc 100644
--- a/drivers/gpio/gpio-lynxpoint.c
+++ b/drivers/gpio/gpio-lynxpoint.c
@@ -450,7 +450,6 @@ static struct platform_driver lp_gpio_driver = {
 	.remove         = lp_gpio_remove,
 	.driver         = {
 		.name   = "lp_gpio",
-		.owner  = THIS_MODULE,
 		.pm	= &lp_gpio_pm_ops,
 		.acpi_match_table = ACPI_PTR(lynxpoint_gpio_acpi_match),
 	},
diff --git a/drivers/gpio/gpio-mcp23s08.c b/drivers/gpio/gpio-mcp23s08.c
index 6f183d9b487e..da9c316059bc 100644
--- a/drivers/gpio/gpio-mcp23s08.c
+++ b/drivers/gpio/gpio-mcp23s08.c
@@ -65,6 +65,7 @@ struct mcp23s08_ops {
 
 struct mcp23s08 {
 	u8			addr;
+	bool			irq_active_high;
 
 	u16			cache[11];
 	u16			irq_rise;
@@ -444,7 +445,7 @@ static int mcp23s08_irq_reqres(struct irq_data *data)
 {
 	struct mcp23s08 *mcp = irq_data_get_irq_chip_data(data);
 
-	if (gpio_lock_as_irq(&mcp->chip, data->hwirq)) {
+	if (gpiochip_lock_as_irq(&mcp->chip, data->hwirq)) {
 		dev_err(mcp->chip.dev,
 			"unable to lock HW IRQ %lu for IRQ usage\n",
 			data->hwirq);
@@ -458,7 +459,7 @@ static void mcp23s08_irq_relres(struct irq_data *data)
 {
 	struct mcp23s08 *mcp = irq_data_get_irq_chip_data(data);
 
-	gpio_unlock_as_irq(&mcp->chip, data->hwirq);
+	gpiochip_unlock_as_irq(&mcp->chip, data->hwirq);
 }
 
 static struct irq_chip mcp23s08_irq_chip = {
@@ -476,17 +477,22 @@ static int mcp23s08_irq_setup(struct mcp23s08 *mcp)
 {
 	struct gpio_chip *chip = &mcp->chip;
 	int err, irq, j;
+	unsigned long irqflags = IRQF_ONESHOT | IRQF_SHARED;
 
 	mutex_init(&mcp->irq_lock);
 
-	mcp->irq_domain = irq_domain_add_linear(chip->of_node, chip->ngpio,
+	mcp->irq_domain = irq_domain_add_linear(chip->dev->of_node, chip->ngpio,
 						&irq_domain_simple_ops, mcp);
 	if (!mcp->irq_domain)
 		return -ENODEV;
 
+	if (mcp->irq_active_high)
+		irqflags |= IRQF_TRIGGER_HIGH;
+	else
+		irqflags |= IRQF_TRIGGER_LOW;
+
 	err = devm_request_threaded_irq(chip->dev, mcp->irq, NULL, mcp23s08_irq,
-					IRQF_TRIGGER_LOW | IRQF_ONESHOT,
-					dev_name(chip->dev), mcp);
+					irqflags, dev_name(chip->dev), mcp);
 	if (err != 0) {
 		dev_err(chip->dev, "unable to request IRQ#%d: %d\n",
 			mcp->irq, err);
@@ -514,8 +520,6 @@ static void mcp23s08_irq_teardown(struct mcp23s08 *mcp)
 {
 	unsigned int irq, i;
 
-	free_irq(mcp->irq, mcp);
-
 	for (i = 0; i < mcp->chip.ngpio; i++) {
 		irq = irq_find_mapping(mcp->irq_domain, i);
 		if (irq > 0)
@@ -581,7 +585,7 @@ done:
 
 static int mcp23s08_probe_one(struct mcp23s08 *mcp, struct device *dev,
 			      void *data, unsigned addr, unsigned type,
-			      unsigned base, unsigned pullups)
+			      struct mcp23s08_platform_data *pdata, int cs)
 {
 	int status;
 	bool mirror = false;
@@ -590,6 +594,7 @@ static int mcp23s08_probe_one(struct mcp23s08 *mcp, struct device *dev,
 
 	mcp->data = data;
 	mcp->addr = addr;
+	mcp->irq_active_high = false;
 
 	mcp->chip.direction_input = mcp23s08_direction_input;
 	mcp->chip.get = mcp23s08_get;
@@ -635,7 +640,7 @@ static int mcp23s08_probe_one(struct mcp23s08 *mcp, struct device *dev,
 		return -EINVAL;
 	}
 
-	mcp->chip.base = base;
+	mcp->chip.base = pdata->base;
 	mcp->chip.can_sleep = true;
 	mcp->chip.dev = dev;
 	mcp->chip.owner = THIS_MODULE;
@@ -648,17 +653,26 @@ static int mcp23s08_probe_one(struct mcp23s08 *mcp, struct device *dev,
 	if (status < 0)
 		goto fail;
 
-	mcp->irq_controller = of_property_read_bool(mcp->chip.of_node,
-						"interrupt-controller");
-	if (mcp->irq && mcp->irq_controller && (type == MCP_TYPE_017))
-		mirror = of_property_read_bool(mcp->chip.of_node,
-						"microchip,irq-mirror");
+	mcp->irq_controller = pdata->irq_controller;
+	if (mcp->irq && mcp->irq_controller) {
+		mcp->irq_active_high =
+			of_property_read_bool(mcp->chip.dev->of_node,
+					      "microchip,irq-active-high");
+
+		if (type == MCP_TYPE_017)
+			mirror = pdata->mirror;
+	}
 
-	if ((status & IOCON_SEQOP) || !(status & IOCON_HAEN) || mirror) {
+	if ((status & IOCON_SEQOP) || !(status & IOCON_HAEN) || mirror ||
+	     mcp->irq_active_high) {
 		/* mcp23s17 has IOCON twice, make sure they are in sync */
 		status &= ~(IOCON_SEQOP | (IOCON_SEQOP << 8));
 		status |= IOCON_HAEN | (IOCON_HAEN << 8);
-		status &= ~(IOCON_INTPOL | (IOCON_INTPOL << 8));
+		if (mcp->irq_active_high)
+			status |= IOCON_INTPOL | (IOCON_INTPOL << 8);
+		else
+			status &= ~(IOCON_INTPOL | (IOCON_INTPOL << 8));
+
 		if (mirror)
 			status |= IOCON_MIRROR | (IOCON_MIRROR << 8);
 
@@ -668,7 +682,7 @@ static int mcp23s08_probe_one(struct mcp23s08 *mcp, struct device *dev,
 	}
 
 	/* configure ~100K pullups */
-	status = mcp->ops->write(mcp, MCP_GPPU, pullups);
+	status = mcp->ops->write(mcp, MCP_GPPU, pdata->chip[cs].pullups);
 	if (status < 0)
 		goto fail;
 
@@ -768,25 +782,29 @@ MODULE_DEVICE_TABLE(of, mcp23s08_i2c_of_match);
 static int mcp230xx_probe(struct i2c_client *client,
 				    const struct i2c_device_id *id)
 {
-	struct mcp23s08_platform_data *pdata;
+	struct mcp23s08_platform_data *pdata, local_pdata;
 	struct mcp23s08 *mcp;
-	int status, base, pullups;
+	int status;
 	const struct of_device_id *match;
 
 	match = of_match_device(of_match_ptr(mcp23s08_i2c_of_match),
 					&client->dev);
-	pdata = dev_get_platdata(&client->dev);
-	if (match || !pdata) {
-		base = -1;
-		pullups = 0;
+	if (match) {
+		pdata = &local_pdata;
+		pdata->base = -1;
+		pdata->chip[0].pullups = 0;
+		pdata->irq_controller =	of_property_read_bool(
+					client->dev.of_node,
+					"interrupt-controller");
+		pdata->mirror = of_property_read_bool(client->dev.of_node,
+						      "microchip,irq-mirror");
 		client->irq = irq_of_parse_and_map(client->dev.of_node, 0);
 	} else {
-		if (!gpio_is_valid(pdata->base)) {
+		pdata = dev_get_platdata(&client->dev);
+		if (!pdata || !gpio_is_valid(pdata->base)) {
 			dev_dbg(&client->dev, "invalid platform data\n");
 			return -EINVAL;
 		}
-		base = pdata->base;
-		pullups = pdata->chip[0].pullups;
 	}
 
 	mcp = kzalloc(sizeof(*mcp), GFP_KERNEL);
@@ -795,7 +813,7 @@ static int mcp230xx_probe(struct i2c_client *client,
 
 	mcp->irq = client->irq;
 	status = mcp23s08_probe_one(mcp, &client->dev, client, client->addr,
-				    id->driver_data, base, pullups);
+				    id->driver_data, pdata, 0);
 	if (status)
 		goto fail;
 
@@ -863,14 +881,12 @@ static void mcp23s08_i2c_exit(void) { }
 
 static int mcp23s08_probe(struct spi_device *spi)
 {
-	struct mcp23s08_platform_data	*pdata;
+	struct mcp23s08_platform_data	*pdata, local_pdata;
 	unsigned			addr;
 	int				chips = 0;
 	struct mcp23s08_driver_data	*data;
 	int				status, type;
-	unsigned			base = -1,
-					ngpio = 0,
-					pullups[ARRAY_SIZE(pdata->chip)];
+	unsigned			ngpio = 0;
 	const struct			of_device_id *match;
 	u32				spi_present_mask = 0;
 
@@ -893,11 +909,18 @@ static int mcp23s08_probe(struct spi_device *spi)
 			return -ENODEV;
 		}
 
+		pdata = &local_pdata;
+		pdata->base = -1;
 		for (addr = 0; addr < ARRAY_SIZE(pdata->chip); addr++) {
-			pullups[addr] = 0;
+			pdata->chip[addr].pullups = 0;
 			if (spi_present_mask & (1 << addr))
 				chips++;
 		}
+		pdata->irq_controller =	of_property_read_bool(
+					spi->dev.of_node,
+					"interrupt-controller");
+		pdata->mirror = of_property_read_bool(spi->dev.of_node,
+						      "microchip,irq-mirror");
 	} else {
 		type = spi_get_device_id(spi)->driver_data;
 		pdata = dev_get_platdata(&spi->dev);
@@ -917,10 +940,7 @@ static int mcp23s08_probe(struct spi_device *spi)
 				return -EINVAL;
 			}
 			spi_present_mask |= 1 << addr;
-			pullups[addr] = pdata->chip[addr].pullups;
 		}
-
-		base = pdata->base;
 	}
 
 	if (!chips)
@@ -932,19 +952,22 @@ static int mcp23s08_probe(struct spi_device *spi)
 		return -ENOMEM;
 	spi_set_drvdata(spi, data);
 
+	spi->irq = irq_of_parse_and_map(spi->dev.of_node, 0);
+
 	for (addr = 0; addr < ARRAY_SIZE(pdata->chip); addr++) {
 		if (!(spi_present_mask & (1 << addr)))
 			continue;
 		chips--;
 		data->mcp[addr] = &data->chip[chips];
+		data->mcp[addr]->irq = spi->irq;
 		status = mcp23s08_probe_one(data->mcp[addr], &spi->dev, spi,
-					    0x40 | (addr << 1), type, base,
-					    pullups[addr]);
+					    0x40 | (addr << 1), type, pdata,
+					    addr);
 		if (status < 0)
 			goto fail;
 
-		if (base != -1)
-			base += (type == MCP_TYPE_S17) ? 16 : 8;
+		if (pdata->base != -1)
+			pdata->base += (type == MCP_TYPE_S17) ? 16 : 8;
 		ngpio += (type == MCP_TYPE_S17) ? 16 : 8;
 	}
 	data->ngpio = ngpio;
@@ -977,6 +1000,8 @@ static int mcp23s08_remove(struct spi_device *spi)
 		if (!data->mcp[addr])
 			continue;
 
+		if (spi->irq && data->mcp[addr]->irq_controller)
+			mcp23s08_irq_teardown(data->mcp[addr]);
 		gpiochip_remove(&data->mcp[addr]->chip);
 	}
 	kfree(data);
diff --git a/drivers/gpio/gpio-mm-lantiq.c b/drivers/gpio/gpio-mm-lantiq.c
index 2983dfbd0668..f228b1ce0ce0 100644
--- a/drivers/gpio/gpio-mm-lantiq.c
+++ b/drivers/gpio/gpio-mm-lantiq.c
@@ -145,7 +145,6 @@ static struct platform_driver ltq_mm_driver = {
 	.probe = ltq_mm_probe,
 	.driver = {
 		.name = "gpio-mm-ltq",
-		.owner = THIS_MODULE,
 		.of_match_table = ltq_mm_match,
 	},
 };
diff --git a/drivers/gpio/gpio-moxart.c b/drivers/gpio/gpio-moxart.c
index 4661e181be04..31e2551ed903 100644
--- a/drivers/gpio/gpio-moxart.c
+++ b/drivers/gpio/gpio-moxart.c
@@ -142,7 +142,6 @@ static const struct of_device_id moxart_gpio_match[] = {
 static struct platform_driver moxart_gpio_driver = {
 	.driver	= {
 		.name		= "moxart-gpio",
-		.owner		= THIS_MODULE,
 		.of_match_table	= moxart_gpio_match,
 	},
 	.probe	= moxart_gpio_probe,
diff --git a/drivers/gpio/gpio-mpc5200.c b/drivers/gpio/gpio-mpc5200.c
index 42647f26c9e0..8ce6c9510035 100644
--- a/drivers/gpio/gpio-mpc5200.c
+++ b/drivers/gpio/gpio-mpc5200.c
@@ -192,7 +192,6 @@ static const struct of_device_id mpc52xx_wkup_gpiochip_match[] = {
 static struct platform_driver mpc52xx_wkup_gpiochip_driver = {
 	.driver = {
 		.name = "mpc5200-gpio-wkup",
-		.owner = THIS_MODULE,
 		.of_match_table = mpc52xx_wkup_gpiochip_match,
 	},
 	.probe = mpc52xx_wkup_gpiochip_probe,
@@ -347,7 +346,6 @@ static const struct of_device_id mpc52xx_simple_gpiochip_match[] = {
 static struct platform_driver mpc52xx_simple_gpiochip_driver = {
 	.driver = {
 		.name = "mpc5200-gpio",
-		.owner = THIS_MODULE,
 		.of_match_table = mpc52xx_simple_gpiochip_match,
 	},
 	.probe = mpc52xx_simple_gpiochip_probe,
diff --git a/drivers/gpio/gpio-mpc8xxx.c b/drivers/gpio/gpio-mpc8xxx.c
index d7d6d72eba33..d1ff879e6ff2 100644
--- a/drivers/gpio/gpio-mpc8xxx.c
+++ b/drivers/gpio/gpio-mpc8xxx.c
@@ -105,6 +105,32 @@ static void mpc8xxx_gpio_set(struct gpio_chip *gc, unsigned int gpio, int val)
 	spin_unlock_irqrestore(&mpc8xxx_gc->lock, flags);
 }
 
+static void mpc8xxx_gpio_set_multiple(struct gpio_chip *gc,
+				      unsigned long *mask, unsigned long *bits)
+{
+	struct of_mm_gpio_chip *mm = to_of_mm_gpio_chip(gc);
+	struct mpc8xxx_gpio_chip *mpc8xxx_gc = to_mpc8xxx_gpio_chip(mm);
+	unsigned long flags;
+	int i;
+
+	spin_lock_irqsave(&mpc8xxx_gc->lock, flags);
+
+	for (i = 0; i < gc->ngpio; i++) {
+		if (*mask == 0)
+			break;
+		if (__test_and_clear_bit(i, mask)) {
+			if (test_bit(i, bits))
+				mpc8xxx_gc->data |= mpc8xxx_gpio2mask(i);
+			else
+				mpc8xxx_gc->data &= ~mpc8xxx_gpio2mask(i);
+		}
+	}
+
+	out_be32(mm->regs + GPIO_DAT, mpc8xxx_gc->data);
+
+	spin_unlock_irqrestore(&mpc8xxx_gc->lock, flags);
+}
+
 static int mpc8xxx_gpio_dir_in(struct gpio_chip *gc, unsigned int gpio)
 {
 	struct of_mm_gpio_chip *mm = to_of_mm_gpio_chip(gc);
@@ -344,6 +370,7 @@ static void __init mpc8xxx_add_controller(struct device_node *np)
 	gc->get = of_device_is_compatible(np, "fsl,mpc8572-gpio") ?
 		mpc8572_gpio_get : mpc8xxx_gpio_get;
 	gc->set = mpc8xxx_gpio_set;
+	gc->set_multiple = mpc8xxx_gpio_set_multiple;
 	gc->to_irq = mpc8xxx_gpio_to_irq;
 
 	ret = of_mm_gpiochip_add(np, mm_gc);
diff --git a/drivers/gpio/gpio-msic.c b/drivers/gpio/gpio-msic.c
index 8f70ded82a2b..01acf0a8cdb1 100644
--- a/drivers/gpio/gpio-msic.c
+++ b/drivers/gpio/gpio-msic.c
@@ -321,7 +321,6 @@ err:
 static struct platform_driver platform_msic_gpio_driver = {
 	.driver = {
 		.name		= "msic_gpio",
-		.owner		= THIS_MODULE,
 	},
 	.probe		= platform_msic_gpio_probe,
 };
diff --git a/drivers/gpio/gpio-msm-v1.c b/drivers/gpio/gpio-msm-v1.c
index 73b73969d361..edf285e26667 100644
--- a/drivers/gpio/gpio-msm-v1.c
+++ b/drivers/gpio/gpio-msm-v1.c
@@ -686,7 +686,7 @@ static int gpio_msm_v1_probe(struct platform_device *pdev)
 	irq_set_chained_handler(irq1, msm_gpio_irq_handler);
 	irq_set_chained_handler(irq2, msm_gpio_irq_handler);
 	irq_set_irq_wake(irq1, 1);
-	irq_set_irq_wake(irq2, 2);
+	irq_set_irq_wake(irq2, 1);
 	return 0;
 }
 
@@ -701,7 +701,6 @@ MODULE_DEVICE_TABLE(platform, gpio_msm_v1_device_ids);
 static struct platform_driver gpio_msm_v1_driver = {
 	.driver = {
 		.name = "gpio-msm-v1",
-		.owner = THIS_MODULE,
 	},
 	.probe = gpio_msm_v1_probe,
 	.id_table = gpio_msm_v1_device_ids,
diff --git a/drivers/gpio/gpio-msm-v2.c b/drivers/gpio/gpio-msm-v2.c
index 94f57670df9a..52ff18229fdc 100644
--- a/drivers/gpio/gpio-msm-v2.c
+++ b/drivers/gpio/gpio-msm-v2.c
@@ -450,7 +450,6 @@ static struct platform_driver msm_gpio_driver = {
 	.remove = msm_gpio_remove,
 	.driver = {
 		.name = "msmgpio",
-		.owner = THIS_MODULE,
 		.of_match_table = msm_gpio_of_match,
 	},
 };
diff --git a/drivers/gpio/gpio-mvebu.c b/drivers/gpio/gpio-mvebu.c
index 418e38650363..7bc3e9b288f3 100644
--- a/drivers/gpio/gpio-mvebu.c
+++ b/drivers/gpio/gpio-mvebu.c
@@ -83,6 +83,14 @@ struct mvebu_gpio_chip {
 	int		   irqbase;
 	struct irq_domain *domain;
 	int                soc_variant;
+
+	/* Used to preserve GPIO registers accross suspend/resume */
+	u32                out_reg;
+	u32                io_conf_reg;
+	u32                blink_en_reg;
+	u32                in_pol_reg;
+	u32                edge_mask_regs[4];
+	u32                level_mask_regs[4];
 };
 
 /*
@@ -554,6 +562,93 @@ static const struct of_device_id mvebu_gpio_of_match[] = {
 };
 MODULE_DEVICE_TABLE(of, mvebu_gpio_of_match);
 
+static int mvebu_gpio_suspend(struct platform_device *pdev, pm_message_t state)
+{
+	struct mvebu_gpio_chip *mvchip = platform_get_drvdata(pdev);
+	int i;
+
+	mvchip->out_reg = readl(mvebu_gpioreg_out(mvchip));
+	mvchip->io_conf_reg = readl(mvebu_gpioreg_io_conf(mvchip));
+	mvchip->blink_en_reg = readl(mvebu_gpioreg_blink(mvchip));
+	mvchip->in_pol_reg = readl(mvebu_gpioreg_in_pol(mvchip));
+
+	switch (mvchip->soc_variant) {
+	case MVEBU_GPIO_SOC_VARIANT_ORION:
+		mvchip->edge_mask_regs[0] =
+			readl(mvchip->membase + GPIO_EDGE_MASK_OFF);
+		mvchip->level_mask_regs[0] =
+			readl(mvchip->membase + GPIO_LEVEL_MASK_OFF);
+		break;
+	case MVEBU_GPIO_SOC_VARIANT_MV78200:
+		for (i = 0; i < 2; i++) {
+			mvchip->edge_mask_regs[i] =
+				readl(mvchip->membase +
+				      GPIO_EDGE_MASK_MV78200_OFF(i));
+			mvchip->level_mask_regs[i] =
+				readl(mvchip->membase +
+				      GPIO_LEVEL_MASK_MV78200_OFF(i));
+		}
+		break;
+	case MVEBU_GPIO_SOC_VARIANT_ARMADAXP:
+		for (i = 0; i < 4; i++) {
+			mvchip->edge_mask_regs[i] =
+				readl(mvchip->membase +
+				      GPIO_EDGE_MASK_ARMADAXP_OFF(i));
+			mvchip->level_mask_regs[i] =
+				readl(mvchip->membase +
+				      GPIO_LEVEL_MASK_ARMADAXP_OFF(i));
+		}
+		break;
+	default:
+		BUG();
+	}
+
+	return 0;
+}
+
+static int mvebu_gpio_resume(struct platform_device *pdev)
+{
+	struct mvebu_gpio_chip *mvchip = platform_get_drvdata(pdev);
+	int i;
+
+	writel(mvchip->out_reg, mvebu_gpioreg_out(mvchip));
+	writel(mvchip->io_conf_reg, mvebu_gpioreg_io_conf(mvchip));
+	writel(mvchip->blink_en_reg, mvebu_gpioreg_blink(mvchip));
+	writel(mvchip->in_pol_reg, mvebu_gpioreg_in_pol(mvchip));
+
+	switch (mvchip->soc_variant) {
+	case MVEBU_GPIO_SOC_VARIANT_ORION:
+		writel(mvchip->edge_mask_regs[0],
+		       mvchip->membase + GPIO_EDGE_MASK_OFF);
+		writel(mvchip->level_mask_regs[0],
+		       mvchip->membase + GPIO_LEVEL_MASK_OFF);
+		break;
+	case MVEBU_GPIO_SOC_VARIANT_MV78200:
+		for (i = 0; i < 2; i++) {
+			writel(mvchip->edge_mask_regs[i],
+			       mvchip->membase + GPIO_EDGE_MASK_MV78200_OFF(i));
+			writel(mvchip->level_mask_regs[i],
+			       mvchip->membase +
+			       GPIO_LEVEL_MASK_MV78200_OFF(i));
+		}
+		break;
+	case MVEBU_GPIO_SOC_VARIANT_ARMADAXP:
+		for (i = 0; i < 4; i++) {
+			writel(mvchip->edge_mask_regs[i],
+			       mvchip->membase +
+			       GPIO_EDGE_MASK_ARMADAXP_OFF(i));
+			writel(mvchip->level_mask_regs[i],
+			       mvchip->membase +
+			       GPIO_LEVEL_MASK_ARMADAXP_OFF(i));
+		}
+		break;
+	default:
+		BUG();
+	}
+
+	return 0;
+}
+
 static int mvebu_gpio_probe(struct platform_device *pdev)
 {
 	struct mvebu_gpio_chip *mvchip;
@@ -577,6 +672,8 @@ static int mvebu_gpio_probe(struct platform_device *pdev)
 	if (!mvchip)
 		return -ENOMEM;
 
+	platform_set_drvdata(pdev, mvchip);
+
 	if (of_property_read_u32(pdev->dev.of_node, "ngpios", &ngpios)) {
 		dev_err(&pdev->dev, "Missing ngpios OF property\n");
 		return -ENODEV;
@@ -731,9 +828,10 @@ static int mvebu_gpio_probe(struct platform_device *pdev)
 static struct platform_driver mvebu_gpio_driver = {
 	.driver		= {
 		.name	        = "mvebu-gpio",
-		.owner	        = THIS_MODULE,
 		.of_match_table = mvebu_gpio_of_match,
 	},
 	.probe		= mvebu_gpio_probe,
+	.suspend        = mvebu_gpio_suspend,
+	.resume         = mvebu_gpio_resume,
 };
 module_platform_driver(mvebu_gpio_driver);
diff --git a/drivers/gpio/gpio-mxc.c b/drivers/gpio/gpio-mxc.c
index f4e54a92e04a..9f7446a7ac64 100644
--- a/drivers/gpio/gpio-mxc.c
+++ b/drivers/gpio/gpio-mxc.c
@@ -496,7 +496,6 @@ out_bgio:
 static struct platform_driver mxc_gpio_driver = {
 	.driver		= {
 		.name	= "gpio-mxc",
-		.owner	= THIS_MODULE,
 		.of_match_table = mxc_gpio_dt_ids,
 	},
 	.probe		= mxc_gpio_probe,
diff --git a/drivers/gpio/gpio-mxs.c b/drivers/gpio/gpio-mxs.c
index 8ffdd7d2bade..84cbda6acdda 100644
--- a/drivers/gpio/gpio-mxs.c
+++ b/drivers/gpio/gpio-mxs.c
@@ -227,6 +227,18 @@ static int mxs_gpio_to_irq(struct gpio_chip *gc, unsigned offset)
 	return irq_find_mapping(port->domain, offset);
 }
 
+static int mxs_gpio_get_direction(struct gpio_chip *gc, unsigned offset)
+{
+	struct bgpio_chip *bgc = to_bgpio_chip(gc);
+	struct mxs_gpio_port *port =
+		container_of(bgc, struct mxs_gpio_port, bgc);
+	u32 mask = 1 << offset;
+	u32 dir;
+
+	dir = readl(port->base + PINCTRL_DOE(port));
+	return !(dir & mask);
+}
+
 static struct platform_device_id mxs_gpio_ids[] = {
 	{
 		.name = "imx23-gpio",
@@ -320,6 +332,7 @@ static int mxs_gpio_probe(struct platform_device *pdev)
 		goto out_irqdesc_free;
 
 	port->bgc.gc.to_irq = mxs_gpio_to_irq;
+	port->bgc.gc.get_direction = mxs_gpio_get_direction;
 	port->bgc.gc.base = port->id * 32;
 
 	err = gpiochip_add(&port->bgc.gc);
@@ -338,7 +351,6 @@ out_irqdesc_free:
 static struct platform_driver mxs_gpio_driver = {
 	.driver		= {
 		.name	= "gpio-mxs",
-		.owner	= THIS_MODULE,
 		.of_match_table = mxs_gpio_dt_ids,
 	},
 	.probe		= mxs_gpio_probe,
diff --git a/drivers/gpio/gpio-octeon.c b/drivers/gpio/gpio-octeon.c
index 5c5770c99c80..62ae251d4490 100644
--- a/drivers/gpio/gpio-octeon.c
+++ b/drivers/gpio/gpio-octeon.c
@@ -144,7 +144,6 @@ MODULE_DEVICE_TABLE(of, octeon_gpio_match);
 static struct platform_driver octeon_gpio_driver = {
 	.driver = {
 		.name		= "octeon_gpio",
-		.owner		= THIS_MODULE,
 		.of_match_table = octeon_gpio_match,
 	},
 	.probe		= octeon_gpio_probe,
diff --git a/drivers/gpio/gpio-omap.c b/drivers/gpio/gpio-omap.c
index 174932165fcb..30646cfe0efa 100644
--- a/drivers/gpio/gpio-omap.c
+++ b/drivers/gpio/gpio-omap.c
@@ -800,7 +800,7 @@ static void omap_gpio_irq_shutdown(struct irq_data *d)
 	unsigned offset = GPIO_INDEX(bank, gpio);
 
 	spin_lock_irqsave(&bank->lock, flags);
-	gpio_unlock_as_irq(&bank->chip, offset);
+	gpiochip_unlock_as_irq(&bank->chip, offset);
 	bank->irq_usage &= ~(BIT(offset));
 	omap_disable_gpio_module(bank, offset);
 	omap_reset_gpio(bank, gpio);
@@ -857,16 +857,6 @@ static void omap_gpio_unmask_irq(struct irq_data *d)
 	spin_unlock_irqrestore(&bank->lock, flags);
 }
 
-static struct irq_chip gpio_irq_chip = {
-	.name		= "GPIO",
-	.irq_shutdown	= omap_gpio_irq_shutdown,
-	.irq_ack	= omap_gpio_ack_irq,
-	.irq_mask	= omap_gpio_mask_irq,
-	.irq_unmask	= omap_gpio_unmask_irq,
-	.irq_set_type	= omap_gpio_irq_type,
-	.irq_set_wake	= omap_gpio_wake_enable,
-};
-
 /*---------------------------------------------------------------------*/
 
 static int omap_mpuio_suspend_noirq(struct device *dev)
@@ -1088,7 +1078,7 @@ omap_mpuio_alloc_gc(struct gpio_bank *bank, unsigned int irq_start,
 			       IRQ_NOREQUEST | IRQ_NOPROBE, 0);
 }
 
-static int omap_gpio_chip_init(struct gpio_bank *bank)
+static int omap_gpio_chip_init(struct gpio_bank *bank, struct irq_chip *irqc)
 {
 	int j;
 	static int gpio;
@@ -1137,17 +1127,17 @@ static int omap_gpio_chip_init(struct gpio_bank *bank)
 	}
 #endif
 
-	ret = gpiochip_irqchip_add(&bank->chip, &gpio_irq_chip,
+	ret = gpiochip_irqchip_add(&bank->chip, irqc,
 				   irq_base, omap_gpio_irq_handler,
 				   IRQ_TYPE_NONE);
 
 	if (ret) {
 		dev_err(bank->dev, "Couldn't add irqchip to gpiochip %d\n", ret);
-		ret = gpiochip_remove(&bank->chip);
+		gpiochip_remove(&bank->chip);
 		return -ENODEV;
 	}
 
-	gpiochip_set_chained_irqchip(&bank->chip, &gpio_irq_chip,
+	gpiochip_set_chained_irqchip(&bank->chip, irqc,
 				     bank->irq, omap_gpio_irq_handler);
 
 	for (j = 0; j < bank->width; j++) {
@@ -1172,6 +1162,7 @@ static int omap_gpio_probe(struct platform_device *pdev)
 	const struct omap_gpio_platform_data *pdata;
 	struct resource *res;
 	struct gpio_bank *bank;
+	struct irq_chip *irqc;
 	int ret;
 
 	match = of_match_device(of_match_ptr(omap_gpio_match), dev);
@@ -1186,6 +1177,18 @@ static int omap_gpio_probe(struct platform_device *pdev)
 		return -ENOMEM;
 	}
 
+	irqc = devm_kzalloc(dev, sizeof(*irqc), GFP_KERNEL);
+	if (!irqc)
+		return -ENOMEM;
+
+	irqc->irq_shutdown = omap_gpio_irq_shutdown,
+	irqc->irq_ack = omap_gpio_ack_irq,
+	irqc->irq_mask = omap_gpio_mask_irq,
+	irqc->irq_unmask = omap_gpio_unmask_irq,
+	irqc->irq_set_type = omap_gpio_irq_type,
+	irqc->irq_set_wake = omap_gpio_wake_enable,
+	irqc->name = dev_name(&pdev->dev);
+
 	res = platform_get_resource(pdev, IORESOURCE_IRQ, 0);
 	if (unlikely(!res)) {
 		dev_err(dev, "Invalid IRQ resource\n");
@@ -1241,7 +1244,7 @@ static int omap_gpio_probe(struct platform_device *pdev)
 
 	omap_gpio_mod_init(bank);
 
-	ret = omap_gpio_chip_init(bank);
+	ret = omap_gpio_chip_init(bank, irqc);
 	if (ret)
 		return ret;
 
@@ -1256,7 +1259,7 @@ static int omap_gpio_probe(struct platform_device *pdev)
 
 #ifdef CONFIG_ARCH_OMAP2PLUS
 
-#if defined(CONFIG_PM_RUNTIME)
+#if defined(CONFIG_PM)
 static void omap_gpio_restore_context(struct gpio_bank *bank);
 
 static int omap_gpio_runtime_suspend(struct device *dev)
@@ -1437,7 +1440,7 @@ static int omap_gpio_runtime_resume(struct device *dev)
 
 	return 0;
 }
-#endif /* CONFIG_PM_RUNTIME */
+#endif /* CONFIG_PM */
 
 void omap2_gpio_prepare_for_idle(int pwr_mode)
 {
@@ -1465,7 +1468,7 @@ void omap2_gpio_resume_after_idle(void)
 	}
 }
 
-#if defined(CONFIG_PM_RUNTIME)
+#if defined(CONFIG_PM)
 static void omap_gpio_init_context(struct gpio_bank *p)
 {
 	struct omap_gpio_reg_offs *regs = p->regs;
@@ -1522,7 +1525,7 @@ static void omap_gpio_restore_context(struct gpio_bank *bank)
 	writel_relaxed(bank->context.irqenable2,
 				bank->base + bank->regs->irqenable2);
 }
-#endif /* CONFIG_PM_RUNTIME */
+#endif /* CONFIG_PM */
 #else
 #define omap_gpio_runtime_suspend NULL
 #define omap_gpio_runtime_resume NULL
diff --git a/drivers/gpio/gpio-pca953x.c b/drivers/gpio/gpio-pca953x.c
index f9961eea2120..e2da64abbccd 100644
--- a/drivers/gpio/gpio-pca953x.c
+++ b/drivers/gpio/gpio-pca953x.c
@@ -520,7 +520,7 @@ static int pca953x_irq_setup(struct pca953x_chip *chip,
 	struct i2c_client *client = chip->client;
 	int ret, i, offset = 0;
 
-	if (irq_base != -1
+	if (client->irq && irq_base != -1
 			&& (id->driver_data & PCA_INT)) {
 
 		switch (chip->chip_type) {
@@ -586,50 +586,6 @@ static int pca953x_irq_setup(struct pca953x_chip *chip,
 }
 #endif
 
-/*
- * Handlers for alternative sources of platform_data
- */
-#ifdef CONFIG_OF_GPIO
-/*
- * Translate OpenFirmware node properties into platform_data
- * WARNING: This is DEPRECATED and will be removed eventually!
- */
-static void
-pca953x_get_alt_pdata(struct i2c_client *client, int *gpio_base, u32 *invert)
-{
-	struct device_node *node;
-	const __be32 *val;
-	int size;
-
-	*gpio_base = -1;
-
-	node = client->dev.of_node;
-	if (node == NULL)
-		return;
-
-	val = of_get_property(node, "linux,gpio-base", &size);
-	WARN(val, "%s: device-tree property 'linux,gpio-base' is deprecated!", __func__);
-	if (val) {
-		if (size != sizeof(*val))
-			dev_warn(&client->dev, "%s: wrong linux,gpio-base\n",
-				 node->full_name);
-		else
-			*gpio_base = be32_to_cpup(val);
-	}
-
-	val = of_get_property(node, "polarity", NULL);
-	WARN(val, "%s: device-tree property 'polarity' is deprecated!", __func__);
-	if (val)
-		*invert = *val;
-}
-#else
-static void
-pca953x_get_alt_pdata(struct i2c_client *client, int *gpio_base, u32 *invert)
-{
-	*gpio_base = -1;
-}
-#endif
-
 static int device_pca953x_init(struct pca953x_chip *chip, u32 invert)
 {
 	int ret;
@@ -704,12 +660,8 @@ static int pca953x_probe(struct i2c_client *client,
 		invert = pdata->invert;
 		chip->names = pdata->names;
 	} else {
-		pca953x_get_alt_pdata(client, &chip->gpio_start, &invert);
-#ifdef CONFIG_OF_GPIO
-		/* If I2C node has no interrupts property, disable GPIO interrupts */
-		if (of_find_property(client->dev.of_node, "interrupts", NULL) == NULL)
-			irq_base = -1;
-#endif
+		chip->gpio_start = -1;
+		irq_base = 0;
 	}
 
 	chip->client = client;
diff --git a/drivers/gpio/gpio-pch.c b/drivers/gpio/gpio-pch.c
index e0ac549dccb5..2d9a950ca2d4 100644
--- a/drivers/gpio/gpio-pch.c
+++ b/drivers/gpio/gpio-pch.c
@@ -171,6 +171,7 @@ static int pch_gpio_direction_input(struct gpio_chip *gpio, unsigned nr)
 	return 0;
 }
 
+#ifdef CONFIG_PM
 /*
  * Save register configuration and disable interrupts.
  */
@@ -206,6 +207,7 @@ static void pch_gpio_restore_reg_conf(struct pch_gpio *chip)
 		iowrite32(chip->pch_gpio_reg.gpio_use_sel_reg,
 			  &chip->reg->gpio_use_sel);
 }
+#endif
 
 static int pch_gpio_to_irq(struct gpio_chip *gpio, unsigned offset)
 {
diff --git a/drivers/gpio/gpio-pl061.c b/drivers/gpio/gpio-pl061.c
index 84b49cfb81a8..04756130437f 100644
--- a/drivers/gpio/gpio-pl061.c
+++ b/drivers/gpio/gpio-pl061.c
@@ -52,28 +52,34 @@ struct pl061_gpio {
 
 	void __iomem		*base;
 	struct gpio_chip	gc;
+	bool			uses_pinctrl;
 
 #ifdef CONFIG_PM
 	struct pl061_context_save_regs csave_regs;
 #endif
 };
 
-static int pl061_gpio_request(struct gpio_chip *chip, unsigned offset)
+static int pl061_gpio_request(struct gpio_chip *gc, unsigned offset)
 {
 	/*
 	 * Map back to global GPIO space and request muxing, the direction
 	 * parameter does not matter for this controller.
 	 */
-	int gpio = chip->base + offset;
+	struct pl061_gpio *chip = container_of(gc, struct pl061_gpio, gc);
+	int gpio = gc->base + offset;
 
-	return pinctrl_request_gpio(gpio);
+	if (chip->uses_pinctrl)
+		return pinctrl_request_gpio(gpio);
+	return 0;
 }
 
-static void pl061_gpio_free(struct gpio_chip *chip, unsigned offset)
+static void pl061_gpio_free(struct gpio_chip *gc, unsigned offset)
 {
-	int gpio = chip->base + offset;
+	struct pl061_gpio *chip = container_of(gc, struct pl061_gpio, gc);
+	int gpio = gc->base + offset;
 
-	pinctrl_free_gpio(gpio);
+	if (chip->uses_pinctrl)
+		pinctrl_free_gpio(gpio);
 }
 
 static int pl061_direction_input(struct gpio_chip *gc, unsigned offset)
@@ -263,6 +269,8 @@ static int pl061_probe(struct amba_device *adev, const struct amba_id *id)
 		return PTR_ERR(chip->base);
 
 	spin_lock_init(&chip->lock);
+	if (of_property_read_bool(dev->of_node, "gpio-ranges"))
+		chip->uses_pinctrl = true;
 
 	chip->gc.request = pl061_gpio_request;
 	chip->gc.free = pl061_gpio_free;
diff --git a/drivers/gpio/gpio-rc5t583.c b/drivers/gpio/gpio-rc5t583.c
index 769233d2da6d..6eabf239676b 100644
--- a/drivers/gpio/gpio-rc5t583.c
+++ b/drivers/gpio/gpio-rc5t583.c
@@ -155,7 +155,6 @@ static int rc5t583_gpio_remove(struct platform_device *pdev)
 static struct platform_driver rc5t583_gpio_driver = {
 	.driver = {
 		.name    = "rc5t583-gpio",
-		.owner   = THIS_MODULE,
 	},
 	.probe		= rc5t583_gpio_probe,
 	.remove		= rc5t583_gpio_remove,
diff --git a/drivers/gpio/gpio-rcar.c b/drivers/gpio/gpio-rcar.c
index bf6c09450fee..584484e3f1e3 100644
--- a/drivers/gpio/gpio-rcar.c
+++ b/drivers/gpio/gpio-rcar.c
@@ -1,6 +1,7 @@
 /*
  * Renesas R-Car GPIO Support
  *
+ *  Copyright (C) 2014 Renesas Electronics Corporation
  *  Copyright (C) 2013 Magnus Damm
  *
  * This program is free software; you can redistribute it and/or modify
@@ -291,22 +292,30 @@ struct gpio_rcar_info {
 	bool has_both_edge_trigger;
 };
 
+static const struct gpio_rcar_info gpio_rcar_info_gen1 = {
+	.has_both_edge_trigger = false,
+};
+
+static const struct gpio_rcar_info gpio_rcar_info_gen2 = {
+	.has_both_edge_trigger = true,
+};
+
 static const struct of_device_id gpio_rcar_of_table[] = {
 	{
 		.compatible = "renesas,gpio-r8a7790",
-		.data = (void *)&(const struct gpio_rcar_info) {
-			.has_both_edge_trigger = true,
-		},
+		.data = &gpio_rcar_info_gen2,
 	}, {
 		.compatible = "renesas,gpio-r8a7791",
-		.data = (void *)&(const struct gpio_rcar_info) {
-			.has_both_edge_trigger = true,
-		},
+		.data = &gpio_rcar_info_gen2,
+	}, {
+		.compatible = "renesas,gpio-r8a7793",
+		.data = &gpio_rcar_info_gen2,
+	}, {
+		.compatible = "renesas,gpio-r8a7794",
+		.data = &gpio_rcar_info_gen2,
 	}, {
 		.compatible = "renesas,gpio-rcar",
-		.data = (void *)&(const struct gpio_rcar_info) {
-			.has_both_edge_trigger = false,
-		},
+		.data = &gpio_rcar_info_gen1,
 	}, {
 		/* Terminator */
 	},
diff --git a/drivers/gpio/gpio-samsung.c b/drivers/gpio/gpio-samsung.c
index 3810da47043f..7c288ba4dc87 100644
--- a/drivers/gpio/gpio-samsung.c
+++ b/drivers/gpio/gpio-samsung.c
@@ -1309,56 +1309,6 @@ samsung_gpio_pull_t s3c_gpio_getpull(unsigned int pin)
 }
 EXPORT_SYMBOL(s3c_gpio_getpull);
 
-#ifdef CONFIG_S5P_GPIO_DRVSTR
-s5p_gpio_drvstr_t s5p_gpio_get_drvstr(unsigned int pin)
-{
-	struct samsung_gpio_chip *chip = samsung_gpiolib_getchip(pin);
-	unsigned int off;
-	void __iomem *reg;
-	int shift;
-	u32 drvstr;
-
-	if (!chip)
-		return -EINVAL;
-
-	off = pin - chip->chip.base;
-	shift = off * 2;
-	reg = chip->base + 0x0C;
-
-	drvstr = __raw_readl(reg);
-	drvstr = drvstr >> shift;
-	drvstr &= 0x3;
-
-	return (__force s5p_gpio_drvstr_t)drvstr;
-}
-EXPORT_SYMBOL(s5p_gpio_get_drvstr);
-
-int s5p_gpio_set_drvstr(unsigned int pin, s5p_gpio_drvstr_t drvstr)
-{
-	struct samsung_gpio_chip *chip = samsung_gpiolib_getchip(pin);
-	unsigned int off;
-	void __iomem *reg;
-	int shift;
-	u32 tmp;
-
-	if (!chip)
-		return -EINVAL;
-
-	off = pin - chip->chip.base;
-	shift = off * 2;
-	reg = chip->base + 0x0C;
-
-	tmp = __raw_readl(reg);
-	tmp &= ~(0x3 << shift);
-	tmp |= drvstr << shift;
-
-	__raw_writel(tmp, reg);
-
-	return 0;
-}
-EXPORT_SYMBOL(s5p_gpio_set_drvstr);
-#endif	/* CONFIG_S5P_GPIO_DRVSTR */
-
 #ifdef CONFIG_PLAT_S3C24XX
 unsigned int s3c2410_modify_misccr(unsigned int clear, unsigned int change)
 {
diff --git a/drivers/gpio/gpio-sch.c b/drivers/gpio/gpio-sch.c
index 41e91d70301e..0a0cf1307d2f 100644
--- a/drivers/gpio/gpio-sch.c
+++ b/drivers/gpio/gpio-sch.c
@@ -29,297 +29,227 @@
 
 #include <linux/gpio.h>
 
-static DEFINE_SPINLOCK(gpio_lock);
-
-#define CGEN	(0x00)
-#define CGIO	(0x04)
-#define CGLV	(0x08)
-
-#define RGEN	(0x20)
-#define RGIO	(0x24)
-#define RGLV	(0x28)
-
-static unsigned short gpio_ba;
-
-static int sch_gpio_core_direction_in(struct gpio_chip *gc, unsigned  gpio_num)
-{
-	u8 curr_dirs;
-	unsigned short offset, bit;
-
-	spin_lock(&gpio_lock);
-
-	offset = CGIO + gpio_num / 8;
-	bit = gpio_num % 8;
-
-	curr_dirs = inb(gpio_ba + offset);
-
-	if (!(curr_dirs & (1 << bit)))
-		outb(curr_dirs | (1 << bit), gpio_ba + offset);
+#define GEN	0x00
+#define GIO	0x04
+#define GLV	0x08
+
+struct sch_gpio {
+	struct gpio_chip chip;
+	spinlock_t lock;
+	unsigned short iobase;
+	unsigned short core_base;
+	unsigned short resume_base;
+};
 
-	spin_unlock(&gpio_lock);
-	return 0;
-}
+#define to_sch_gpio(c)	container_of(c, struct sch_gpio, chip)
 
-static int sch_gpio_core_get(struct gpio_chip *gc, unsigned gpio_num)
+static unsigned sch_gpio_offset(struct sch_gpio *sch, unsigned gpio,
+				unsigned reg)
 {
-	int res;
-	unsigned short offset, bit;
+	unsigned base = 0;
 
-	offset = CGLV + gpio_num / 8;
-	bit = gpio_num % 8;
+	if (gpio >= sch->resume_base) {
+		gpio -= sch->resume_base;
+		base += 0x20;
+	}
 
-	res = !!(inb(gpio_ba + offset) & (1 << bit));
-	return res;
+	return base + reg + gpio / 8;
 }
 
-static void sch_gpio_core_set(struct gpio_chip *gc, unsigned gpio_num, int val)
+static unsigned sch_gpio_bit(struct sch_gpio *sch, unsigned gpio)
 {
-	u8 curr_vals;
-	unsigned short offset, bit;
-
-	spin_lock(&gpio_lock);
-
-	offset = CGLV + gpio_num / 8;
-	bit = gpio_num % 8;
-
-	curr_vals = inb(gpio_ba + offset);
-
-	if (val)
-		outb(curr_vals | (1 << bit), gpio_ba + offset);
-	else
-		outb((curr_vals & ~(1 << bit)), gpio_ba + offset);
-	spin_unlock(&gpio_lock);
+	if (gpio >= sch->resume_base)
+		gpio -= sch->resume_base;
+	return gpio % 8;
 }
 
-static int sch_gpio_core_direction_out(struct gpio_chip *gc,
-					unsigned gpio_num, int val)
+static void sch_gpio_enable(struct sch_gpio *sch, unsigned gpio)
 {
-	u8 curr_dirs;
 	unsigned short offset, bit;
+	u8 enable;
 
-	spin_lock(&gpio_lock);
+	spin_lock(&sch->lock);
 
-	offset = CGIO + gpio_num / 8;
-	bit = gpio_num % 8;
-
-	curr_dirs = inb(gpio_ba + offset);
-	if (curr_dirs & (1 << bit))
-		outb(curr_dirs & ~(1 << bit), gpio_ba + offset);
+	offset = sch_gpio_offset(sch, gpio, GEN);
+	bit = sch_gpio_bit(sch, gpio);
 
-	spin_unlock(&gpio_lock);
+	enable = inb(sch->iobase + offset);
+	if (!(enable & (1 << bit)))
+		outb(enable | (1 << bit), sch->iobase + offset);
 
-	/*
-	 * according to the datasheet, writing to the level register has no
-	 * effect when GPIO is programmed as input.
-	 * Actually the the level register is read-only when configured as input.
-	 * Thus presetting the output level before switching to output is _NOT_ possible.
-	 * Hence we set the level after configuring the GPIO as output.
-	 * But we cannot prevent a short low pulse if direction is set to high
-	 * and an external pull-up is connected.
-	 */
-	sch_gpio_core_set(gc, gpio_num, val);
-	return 0;
+	spin_unlock(&sch->lock);
 }
 
-static struct gpio_chip sch_gpio_core = {
-	.label			= "sch_gpio_core",
-	.owner			= THIS_MODULE,
-	.direction_input	= sch_gpio_core_direction_in,
-	.get			= sch_gpio_core_get,
-	.direction_output	= sch_gpio_core_direction_out,
-	.set			= sch_gpio_core_set,
-};
-
-static int sch_gpio_resume_direction_in(struct gpio_chip *gc,
-					unsigned gpio_num)
+static int sch_gpio_direction_in(struct gpio_chip *gc, unsigned  gpio_num)
 {
+	struct sch_gpio *sch = to_sch_gpio(gc);
 	u8 curr_dirs;
 	unsigned short offset, bit;
 
-	spin_lock(&gpio_lock);
+	spin_lock(&sch->lock);
 
-	offset = RGIO + gpio_num / 8;
-	bit = gpio_num % 8;
+	offset = sch_gpio_offset(sch, gpio_num, GIO);
+	bit = sch_gpio_bit(sch, gpio_num);
 
-	curr_dirs = inb(gpio_ba + offset);
+	curr_dirs = inb(sch->iobase + offset);
 
 	if (!(curr_dirs & (1 << bit)))
-		outb(curr_dirs | (1 << bit), gpio_ba + offset);
+		outb(curr_dirs | (1 << bit), sch->iobase + offset);
 
-	spin_unlock(&gpio_lock);
+	spin_unlock(&sch->lock);
 	return 0;
 }
 
-static int sch_gpio_resume_get(struct gpio_chip *gc, unsigned gpio_num)
+static int sch_gpio_get(struct gpio_chip *gc, unsigned gpio_num)
 {
+	struct sch_gpio *sch = to_sch_gpio(gc);
+	int res;
 	unsigned short offset, bit;
 
-	offset = RGLV + gpio_num / 8;
-	bit = gpio_num % 8;
+	offset = sch_gpio_offset(sch, gpio_num, GLV);
+	bit = sch_gpio_bit(sch, gpio_num);
+
+	res = !!(inb(sch->iobase + offset) & (1 << bit));
 
-	return !!(inb(gpio_ba + offset) & (1 << bit));
+	return res;
 }
 
-static void sch_gpio_resume_set(struct gpio_chip *gc,
-				unsigned gpio_num, int val)
+static void sch_gpio_set(struct gpio_chip *gc, unsigned gpio_num, int val)
 {
+	struct sch_gpio *sch = to_sch_gpio(gc);
 	u8 curr_vals;
 	unsigned short offset, bit;
 
-	spin_lock(&gpio_lock);
+	spin_lock(&sch->lock);
 
-	offset = RGLV + gpio_num / 8;
-	bit = gpio_num % 8;
+	offset = sch_gpio_offset(sch, gpio_num, GLV);
+	bit = sch_gpio_bit(sch, gpio_num);
 
-	curr_vals = inb(gpio_ba + offset);
+	curr_vals = inb(sch->iobase + offset);
 
 	if (val)
-		outb(curr_vals | (1 << bit), gpio_ba + offset);
+		outb(curr_vals | (1 << bit), sch->iobase + offset);
 	else
-		outb((curr_vals & ~(1 << bit)), gpio_ba + offset);
+		outb((curr_vals & ~(1 << bit)), sch->iobase + offset);
 
-	spin_unlock(&gpio_lock);
+	spin_unlock(&sch->lock);
 }
 
-static int sch_gpio_resume_direction_out(struct gpio_chip *gc,
-					unsigned gpio_num, int val)
+static int sch_gpio_direction_out(struct gpio_chip *gc, unsigned gpio_num,
+				  int val)
 {
+	struct sch_gpio *sch = to_sch_gpio(gc);
 	u8 curr_dirs;
 	unsigned short offset, bit;
 
-	offset = RGIO + gpio_num / 8;
-	bit = gpio_num % 8;
+	spin_lock(&sch->lock);
 
-	spin_lock(&gpio_lock);
+	offset = sch_gpio_offset(sch, gpio_num, GIO);
+	bit = sch_gpio_bit(sch, gpio_num);
 
-	curr_dirs = inb(gpio_ba + offset);
+	curr_dirs = inb(sch->iobase + offset);
 	if (curr_dirs & (1 << bit))
-		outb(curr_dirs & ~(1 << bit), gpio_ba + offset);
+		outb(curr_dirs & ~(1 << bit), sch->iobase + offset);
 
-	spin_unlock(&gpio_lock);
+	spin_unlock(&sch->lock);
 
 	/*
-	* according to the datasheet, writing to the level register has no
-	* effect when GPIO is programmed as input.
-	* Actually the the level register is read-only when configured as input.
-	* Thus presetting the output level before switching to output is _NOT_ possible.
-	* Hence we set the level after configuring the GPIO as output.
-	* But we cannot prevent a short low pulse if direction is set to high
-	* and an external pull-up is connected.
-	*/
-	sch_gpio_resume_set(gc, gpio_num, val);
+	 * according to the datasheet, writing to the level register has no
+	 * effect when GPIO is programmed as input.
+	 * Actually the the level register is read-only when configured as input.
+	 * Thus presetting the output level before switching to output is _NOT_ possible.
+	 * Hence we set the level after configuring the GPIO as output.
+	 * But we cannot prevent a short low pulse if direction is set to high
+	 * and an external pull-up is connected.
+	 */
+	sch_gpio_set(gc, gpio_num, val);
 	return 0;
 }
 
-static struct gpio_chip sch_gpio_resume = {
-	.label			= "sch_gpio_resume",
+static struct gpio_chip sch_gpio_chip = {
+	.label			= "sch_gpio",
 	.owner			= THIS_MODULE,
-	.direction_input	= sch_gpio_resume_direction_in,
-	.get			= sch_gpio_resume_get,
-	.direction_output	= sch_gpio_resume_direction_out,
-	.set			= sch_gpio_resume_set,
+	.direction_input	= sch_gpio_direction_in,
+	.get			= sch_gpio_get,
+	.direction_output	= sch_gpio_direction_out,
+	.set			= sch_gpio_set,
 };
 
 static int sch_gpio_probe(struct platform_device *pdev)
 {
+	struct sch_gpio *sch;
 	struct resource *res;
-	int err, id;
 
-	id = pdev->id;
-	if (!id)
-		return -ENODEV;
+	sch = devm_kzalloc(&pdev->dev, sizeof(*sch), GFP_KERNEL);
+	if (!sch)
+		return -ENOMEM;
 
 	res = platform_get_resource(pdev, IORESOURCE_IO, 0);
 	if (!res)
 		return -EBUSY;
 
-	if (!request_region(res->start, resource_size(res), pdev->name))
+	if (!devm_request_region(&pdev->dev, res->start, resource_size(res),
+				 pdev->name))
 		return -EBUSY;
 
-	gpio_ba = res->start;
+	spin_lock_init(&sch->lock);
+	sch->iobase = res->start;
+	sch->chip = sch_gpio_chip;
+	sch->chip.label = dev_name(&pdev->dev);
+	sch->chip.dev = &pdev->dev;
 
-	switch (id) {
+	switch (pdev->id) {
 	case PCI_DEVICE_ID_INTEL_SCH_LPC:
-		sch_gpio_core.base = 0;
-		sch_gpio_core.ngpio = 10;
-		sch_gpio_resume.base = 10;
-		sch_gpio_resume.ngpio = 4;
+		sch->core_base = 0;
+		sch->resume_base = 10;
+		sch->chip.ngpio = 14;
+
 		/*
 		 * GPIO[6:0] enabled by default
 		 * GPIO7 is configured by the CMC as SLPIOVR
 		 * Enable GPIO[9:8] core powered gpios explicitly
 		 */
-		outb(0x3, gpio_ba + CGEN + 1);
+		sch_gpio_enable(sch, 8);
+		sch_gpio_enable(sch, 9);
 		/*
 		 * SUS_GPIO[2:0] enabled by default
 		 * Enable SUS_GPIO3 resume powered gpio explicitly
 		 */
-		outb(0x8, gpio_ba + RGEN);
+		sch_gpio_enable(sch, 13);
 		break;
 
 	case PCI_DEVICE_ID_INTEL_ITC_LPC:
-		sch_gpio_core.base = 0;
-		sch_gpio_core.ngpio = 5;
-		sch_gpio_resume.base = 5;
-		sch_gpio_resume.ngpio = 9;
+		sch->core_base = 0;
+		sch->resume_base = 5;
+		sch->chip.ngpio = 14;
 		break;
 
 	case PCI_DEVICE_ID_INTEL_CENTERTON_ILB:
-		sch_gpio_core.base = 0;
-		sch_gpio_core.ngpio = 21;
-		sch_gpio_resume.base = 21;
-		sch_gpio_resume.ngpio = 9;
+		sch->core_base = 0;
+		sch->resume_base = 21;
+		sch->chip.ngpio = 30;
 		break;
 
 	default:
-		err = -ENODEV;
-		goto err_sch_gpio_core;
+		return -ENODEV;
 	}
 
-	sch_gpio_core.dev = &pdev->dev;
-	sch_gpio_resume.dev = &pdev->dev;
-
-	err = gpiochip_add(&sch_gpio_core);
-	if (err < 0)
-		goto err_sch_gpio_core;
+	platform_set_drvdata(pdev, sch);
 
-	err = gpiochip_add(&sch_gpio_resume);
-	if (err < 0)
-		goto err_sch_gpio_resume;
-
-	return 0;
-
-err_sch_gpio_resume:
-	gpiochip_remove(&sch_gpio_core);
-
-err_sch_gpio_core:
-	release_region(res->start, resource_size(res));
-	gpio_ba = 0;
-
-	return err;
+	return gpiochip_add(&sch->chip);
 }
 
 static int sch_gpio_remove(struct platform_device *pdev)
 {
-	struct resource *res;
-	if (gpio_ba) {
-
-		gpiochip_remove(&sch_gpio_core);
-		gpiochip_remove(&sch_gpio_resume);
-
-		res = platform_get_resource(pdev, IORESOURCE_IO, 0);
-
-		release_region(res->start, resource_size(res));
-		gpio_ba = 0;
-	}
+	struct sch_gpio *sch = platform_get_drvdata(pdev);
 
+	gpiochip_remove(&sch->chip);
 	return 0;
 }
 
 static struct platform_driver sch_gpio_driver = {
 	.driver = {
 		.name = "sch_gpio",
-		.owner = THIS_MODULE,
 	},
 	.probe		= sch_gpio_probe,
 	.remove		= sch_gpio_remove,
diff --git a/drivers/gpio/gpio-spear-spics.c b/drivers/gpio/gpio-spear-spics.c
index 353263c85d26..69ffca5b073b 100644
--- a/drivers/gpio/gpio-spear-spics.c
+++ b/drivers/gpio/gpio-spear-spics.c
@@ -191,7 +191,6 @@ MODULE_DEVICE_TABLE(of, spics_gpio_of_match);
 static struct platform_driver spics_gpio_driver = {
 	.probe = spics_gpio_probe,
 	.driver = {
-		.owner = THIS_MODULE,
 		.name = "spear-spics-gpio",
 		.of_match_table = spics_gpio_of_match,
 	},
@@ -204,5 +203,5 @@ static int __init spics_gpio_init(void)
 subsys_initcall(spics_gpio_init);
 
 MODULE_AUTHOR("Shiraz Hashim <shiraz.linux.kernel@gmail.com>");
-MODULE_DESCRIPTION("ST Microlectronics SPEAr SPI Chip Select Abstraction");
+MODULE_DESCRIPTION("STMicroelectronics SPEAr SPI Chip Select Abstraction");
 MODULE_LICENSE("GPL");
diff --git a/drivers/gpio/gpio-sta2x11.c b/drivers/gpio/gpio-sta2x11.c
index 68e3fcb1acea..18579ac65b2b 100644
--- a/drivers/gpio/gpio-sta2x11.c
+++ b/drivers/gpio/gpio-sta2x11.c
@@ -429,7 +429,6 @@ err_free_descs:
 static struct platform_driver sta2x11_gpio_platform_driver = {
 	.driver = {
 		.name	= "sta2x11-gpio",
-		.owner	= THIS_MODULE,
 	},
 	.probe = gsta_probe,
 };
diff --git a/drivers/gpio/gpio-stmpe.c b/drivers/gpio/gpio-stmpe.c
index 845025a57240..85c5b1974294 100644
--- a/drivers/gpio/gpio-stmpe.c
+++ b/drivers/gpio/gpio-stmpe.c
@@ -13,6 +13,7 @@
 #include <linux/interrupt.h>
 #include <linux/of.h>
 #include <linux/mfd/stmpe.h>
+#include <linux/seq_file.h>
 
 /*
  * These registers are modified under the irq bus lock and cached to avoid
@@ -127,19 +128,19 @@ static int stmpe_gpio_irq_set_type(struct irq_data *d, unsigned int type)
 	int regoffset = offset / 8;
 	int mask = 1 << (offset % 8);
 
-	if (type == IRQ_TYPE_LEVEL_LOW || type == IRQ_TYPE_LEVEL_HIGH)
+	if (type & IRQ_TYPE_LEVEL_LOW || type & IRQ_TYPE_LEVEL_HIGH)
 		return -EINVAL;
 
 	/* STMPE801 doesn't have RE and FE registers */
 	if (stmpe_gpio->stmpe->partnum == STMPE801)
 		return 0;
 
-	if (type == IRQ_TYPE_EDGE_RISING)
+	if (type & IRQ_TYPE_EDGE_RISING)
 		stmpe_gpio->regs[REG_RE][regoffset] |= mask;
 	else
 		stmpe_gpio->regs[REG_RE][regoffset] &= ~mask;
 
-	if (type == IRQ_TYPE_EDGE_FALLING)
+	if (type & IRQ_TYPE_EDGE_FALLING)
 		stmpe_gpio->regs[REG_FE][regoffset] |= mask;
 	else
 		stmpe_gpio->regs[REG_FE][regoffset] &= ~mask;
@@ -211,6 +212,77 @@ static void stmpe_gpio_irq_unmask(struct irq_data *d)
 	stmpe_gpio->regs[REG_IE][regoffset] |= mask;
 }
 
+static void stmpe_dbg_show_one(struct seq_file *s,
+			       struct gpio_chip *gc,
+			       unsigned offset, unsigned gpio)
+{
+	struct stmpe_gpio *stmpe_gpio = to_stmpe_gpio(gc);
+	struct stmpe *stmpe = stmpe_gpio->stmpe;
+	const char *label = gpiochip_is_requested(gc, offset);
+	int num_banks = DIV_ROUND_UP(stmpe->num_gpios, 8);
+	bool val = !!stmpe_gpio_get(gc, offset);
+	u8 dir_reg = stmpe->regs[STMPE_IDX_GPDR_LSB] - (offset / 8);
+	u8 mask = 1 << (offset % 8);
+	int ret;
+	u8 dir;
+
+	ret = stmpe_reg_read(stmpe, dir_reg);
+	if (ret < 0)
+		return;
+	dir = !!(ret & mask);
+
+	if (dir) {
+		seq_printf(s, " gpio-%-3d (%-20.20s) out %s",
+			   gpio, label ?: "(none)",
+			   val ? "hi" : "lo");
+	} else {
+		u8 edge_det_reg = stmpe->regs[STMPE_IDX_GPEDR_MSB] + num_banks - 1 - (offset / 8);
+		u8 rise_reg = stmpe->regs[STMPE_IDX_GPRER_LSB] - (offset / 8);
+		u8 fall_reg = stmpe->regs[STMPE_IDX_GPFER_LSB] - (offset / 8);
+		u8 irqen_reg = stmpe->regs[STMPE_IDX_IEGPIOR_LSB] - (offset / 8);
+		bool edge_det;
+		bool rise;
+		bool fall;
+		bool irqen;
+
+		ret = stmpe_reg_read(stmpe, edge_det_reg);
+		if (ret < 0)
+			return;
+		edge_det = !!(ret & mask);
+		ret = stmpe_reg_read(stmpe, rise_reg);
+		if (ret < 0)
+			return;
+		rise = !!(ret & mask);
+		ret = stmpe_reg_read(stmpe, fall_reg);
+		if (ret < 0)
+			return;
+		fall = !!(ret & mask);
+		ret = stmpe_reg_read(stmpe, irqen_reg);
+		if (ret < 0)
+			return;
+		irqen = !!(ret & mask);
+
+		seq_printf(s, " gpio-%-3d (%-20.20s) in  %s %s %s%s%s",
+			   gpio, label ?: "(none)",
+			   val ? "hi" : "lo",
+			   edge_det ? "edge-asserted" : "edge-inactive",
+			   irqen ? "IRQ-enabled" : "",
+			   rise ? " rising-edge-detection" : "",
+			   fall ? " falling-edge-detection" : "");
+	}
+}
+
+static void stmpe_dbg_show(struct seq_file *s, struct gpio_chip *gc)
+{
+	unsigned i;
+	unsigned gpio = gc->base;
+
+	for (i = 0; i < gc->ngpio; i++, gpio++) {
+		stmpe_dbg_show_one(s, gc, i, gpio);
+		seq_printf(s, "\n");
+	}
+}
+
 static struct irq_chip stmpe_gpio_irq_chip = {
 	.name			= "stmpe-gpio",
 	.irq_bus_lock		= stmpe_gpio_irq_lock,
@@ -293,6 +365,9 @@ static int stmpe_gpio_probe(struct platform_device *pdev)
 #endif
 	stmpe_gpio->chip.base = -1;
 
+	if (IS_ENABLED(CONFIG_DEBUG_FS))
+                stmpe_gpio->chip.dbg_show = stmpe_dbg_show;
+
 	if (pdata)
 		stmpe_gpio->norequest_mask = pdata->norequest_mask;
 	else if (np)
@@ -308,6 +383,12 @@ static int stmpe_gpio_probe(struct platform_device *pdev)
 	if (ret)
 		goto out_free;
 
+	ret = gpiochip_add(&stmpe_gpio->chip);
+	if (ret) {
+		dev_err(&pdev->dev, "unable to add gpiochip: %d\n", ret);
+		goto out_disable;
+	}
+
 	if (irq > 0) {
 		ret = devm_request_threaded_irq(&pdev->dev, irq, NULL,
 				stmpe_gpio_irq, IRQF_ONESHOT,
@@ -324,14 +405,13 @@ static int stmpe_gpio_probe(struct platform_device *pdev)
 		if (ret) {
 			dev_err(&pdev->dev,
 				"could not connect irqchip to gpiochip\n");
-			return ret;
+			goto out_disable;
 		}
-	}
 
-	ret = gpiochip_add(&stmpe_gpio->chip);
-	if (ret) {
-		dev_err(&pdev->dev, "unable to add gpiochip: %d\n", ret);
-		goto out_disable;
+		gpiochip_set_chained_irqchip(&stmpe_gpio->chip,
+					     &stmpe_gpio_irq_chip,
+					     irq,
+					     NULL);
 	}
 
 	if (pdata && pdata->setup)
@@ -343,6 +423,7 @@ static int stmpe_gpio_probe(struct platform_device *pdev)
 
 out_disable:
 	stmpe_disable(stmpe, STMPE_BLOCK_GPIO);
+	gpiochip_remove(&stmpe_gpio->chip);
 out_free:
 	kfree(stmpe_gpio);
 	return ret;
diff --git a/drivers/gpio/gpio-stp-xway.c b/drivers/gpio/gpio-stp-xway.c
index 04882a911b65..202361eb7279 100644
--- a/drivers/gpio/gpio-stp-xway.c
+++ b/drivers/gpio/gpio-stp-xway.c
@@ -199,21 +199,17 @@ static int xway_stp_hw_init(struct xway_stp *chip)
 
 static int xway_stp_probe(struct platform_device *pdev)
 {
-	struct resource *res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	struct resource *res;
 	const __be32 *shadow, *groups, *dsl, *phy;
 	struct xway_stp *chip;
 	struct clk *clk;
 	int ret = 0;
 
-	if (!res) {
-		dev_err(&pdev->dev, "failed to request STP resource\n");
-		return -ENOENT;
-	}
-
 	chip = devm_kzalloc(&pdev->dev, sizeof(*chip), GFP_KERNEL);
 	if (!chip)
 		return -ENOMEM;
 
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	chip->virt = devm_ioremap_resource(&pdev->dev, res);
 	if (IS_ERR(chip->virt))
 		return PTR_ERR(chip->virt);
@@ -287,12 +283,11 @@ static struct platform_driver xway_stp_driver = {
 	.probe = xway_stp_probe,
 	.driver = {
 		.name = "gpio-stp-xway",
-		.owner = THIS_MODULE,
 		.of_match_table = xway_stp_match,
 	},
 };
 
-int __init xway_stp_init(void)
+static int __init xway_stp_init(void)
 {
 	return platform_driver_register(&xway_stp_driver);
 }
diff --git a/drivers/gpio/gpio-syscon.c b/drivers/gpio/gpio-syscon.c
index 30884fbc750d..257e2989215c 100644
--- a/drivers/gpio/gpio-syscon.c
+++ b/drivers/gpio/gpio-syscon.c
@@ -37,6 +37,8 @@
  * dat_bit_offset:	Offset (in bits) to the first GPIO bit.
  * dir_bit_offset:	Optional offset (in bits) to the first bit to switch
  *			GPIO direction (Used with GPIO_SYSCON_FEAT_DIR flag).
+ * set:		HW specific callback to assigns output value
+ *			for signal "offset"
  */
 
 struct syscon_gpio_data {
@@ -45,12 +47,16 @@ struct syscon_gpio_data {
 	unsigned int	bit_count;
 	unsigned int	dat_bit_offset;
 	unsigned int	dir_bit_offset;
+	void		(*set)(struct gpio_chip *chip,
+			       unsigned offset, int value);
 };
 
 struct syscon_gpio_priv {
 	struct gpio_chip		chip;
 	struct regmap			*syscon;
 	const struct syscon_gpio_data	*data;
+	u32				dreg_offset;
+	u32				dir_reg_offset;
 };
 
 static inline struct syscon_gpio_priv *to_syscon_gpio(struct gpio_chip *chip)
@@ -61,9 +67,11 @@ static inline struct syscon_gpio_priv *to_syscon_gpio(struct gpio_chip *chip)
 static int syscon_gpio_get(struct gpio_chip *chip, unsigned offset)
 {
 	struct syscon_gpio_priv *priv = to_syscon_gpio(chip);
-	unsigned int val, offs = priv->data->dat_bit_offset + offset;
+	unsigned int val, offs;
 	int ret;
 
+	offs = priv->dreg_offset + priv->data->dat_bit_offset + offset;
+
 	ret = regmap_read(priv->syscon,
 			  (offs / SYSCON_REG_BITS) * SYSCON_REG_SIZE, &val);
 	if (ret)
@@ -75,7 +83,9 @@ static int syscon_gpio_get(struct gpio_chip *chip, unsigned offset)
 static void syscon_gpio_set(struct gpio_chip *chip, unsigned offset, int val)
 {
 	struct syscon_gpio_priv *priv = to_syscon_gpio(chip);
-	unsigned int offs = priv->data->dat_bit_offset + offset;
+	unsigned int offs;
+
+	offs = priv->dreg_offset + priv->data->dat_bit_offset + offset;
 
 	regmap_update_bits(priv->syscon,
 			   (offs / SYSCON_REG_BITS) * SYSCON_REG_SIZE,
@@ -88,7 +98,10 @@ static int syscon_gpio_dir_in(struct gpio_chip *chip, unsigned offset)
 	struct syscon_gpio_priv *priv = to_syscon_gpio(chip);
 
 	if (priv->data->flags & GPIO_SYSCON_FEAT_DIR) {
-		unsigned int offs = priv->data->dir_bit_offset + offset;
+		unsigned int offs;
+
+		offs = priv->dir_reg_offset +
+		       priv->data->dir_bit_offset + offset;
 
 		regmap_update_bits(priv->syscon,
 				   (offs / SYSCON_REG_BITS) * SYSCON_REG_SIZE,
@@ -103,7 +116,10 @@ static int syscon_gpio_dir_out(struct gpio_chip *chip, unsigned offset, int val)
 	struct syscon_gpio_priv *priv = to_syscon_gpio(chip);
 
 	if (priv->data->flags & GPIO_SYSCON_FEAT_DIR) {
-		unsigned int offs = priv->data->dir_bit_offset + offset;
+		unsigned int offs;
+
+		offs = priv->dir_reg_offset +
+		       priv->data->dir_bit_offset + offset;
 
 		regmap_update_bits(priv->syscon,
 				   (offs / SYSCON_REG_BITS) * SYSCON_REG_SIZE,
@@ -111,7 +127,7 @@ static int syscon_gpio_dir_out(struct gpio_chip *chip, unsigned offset, int val)
 				   BIT(offs % SYSCON_REG_BITS));
 	}
 
-	syscon_gpio_set(chip, offset, val);
+	priv->data->set(chip, offset, val);
 
 	return 0;
 }
@@ -124,11 +140,46 @@ static const struct syscon_gpio_data clps711x_mctrl_gpio = {
 	.dat_bit_offset	= 0x40 * 8 + 8,
 };
 
+#define KEYSTONE_LOCK_BIT BIT(0)
+
+static void keystone_gpio_set(struct gpio_chip *chip, unsigned offset, int val)
+{
+	struct syscon_gpio_priv *priv = to_syscon_gpio(chip);
+	unsigned int offs;
+	int ret;
+
+	offs = priv->dreg_offset + priv->data->dat_bit_offset + offset;
+
+	if (!val)
+		return;
+
+	ret = regmap_update_bits(
+			priv->syscon,
+			(offs / SYSCON_REG_BITS) * SYSCON_REG_SIZE,
+			BIT(offs % SYSCON_REG_BITS) | KEYSTONE_LOCK_BIT,
+			BIT(offs % SYSCON_REG_BITS) | KEYSTONE_LOCK_BIT);
+	if (ret < 0)
+		dev_err(chip->dev, "gpio write failed ret(%d)\n", ret);
+}
+
+static const struct syscon_gpio_data keystone_dsp_gpio = {
+	/* ARM Keystone 2 */
+	.compatible	= NULL,
+	.flags		= GPIO_SYSCON_FEAT_OUT,
+	.bit_count	= 28,
+	.dat_bit_offset	= 4,
+	.set		= keystone_gpio_set,
+};
+
 static const struct of_device_id syscon_gpio_ids[] = {
 	{
 		.compatible	= "cirrus,clps711x-mctrl-gpio",
 		.data		= &clps711x_mctrl_gpio,
 	},
+	{
+		.compatible	= "ti,keystone-dsp-gpio",
+		.data		= &keystone_dsp_gpio,
+	},
 	{ }
 };
 MODULE_DEVICE_TABLE(of, syscon_gpio_ids);
@@ -138,6 +189,8 @@ static int syscon_gpio_probe(struct platform_device *pdev)
 	struct device *dev = &pdev->dev;
 	const struct of_device_id *of_id = of_match_device(syscon_gpio_ids, dev);
 	struct syscon_gpio_priv *priv;
+	struct device_node *np = dev->of_node;
+	int ret;
 
 	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
 	if (!priv)
@@ -145,10 +198,31 @@ static int syscon_gpio_probe(struct platform_device *pdev)
 
 	priv->data = of_id->data;
 
-	priv->syscon =
-		syscon_regmap_lookup_by_compatible(priv->data->compatible);
-	if (IS_ERR(priv->syscon))
-		return PTR_ERR(priv->syscon);
+	if (priv->data->compatible) {
+		priv->syscon = syscon_regmap_lookup_by_compatible(
+					priv->data->compatible);
+		if (IS_ERR(priv->syscon))
+			return PTR_ERR(priv->syscon);
+	} else {
+		priv->syscon =
+			syscon_regmap_lookup_by_phandle(np, "gpio,syscon-dev");
+		if (IS_ERR(priv->syscon))
+			return PTR_ERR(priv->syscon);
+
+		ret = of_property_read_u32_index(np, "gpio,syscon-dev", 1,
+						 &priv->dreg_offset);
+		if (ret)
+			dev_err(dev, "can't read the data register offset!\n");
+
+		priv->dreg_offset <<= 3;
+
+		ret = of_property_read_u32_index(np, "gpio,syscon-dev", 2,
+						 &priv->dir_reg_offset);
+		if (ret)
+			dev_err(dev, "can't read the dir register offset!\n");
+
+		priv->dir_reg_offset <<= 3;
+	}
 
 	priv->chip.dev = dev;
 	priv->chip.owner = THIS_MODULE;
@@ -159,7 +233,7 @@ static int syscon_gpio_probe(struct platform_device *pdev)
 	if (priv->data->flags & GPIO_SYSCON_FEAT_IN)
 		priv->chip.direction_input = syscon_gpio_dir_in;
 	if (priv->data->flags & GPIO_SYSCON_FEAT_OUT) {
-		priv->chip.set = syscon_gpio_set;
+		priv->chip.set = priv->data->set ? : syscon_gpio_set;
 		priv->chip.direction_output = syscon_gpio_dir_out;
 	}
 
@@ -179,7 +253,6 @@ static int syscon_gpio_remove(struct platform_device *pdev)
 static struct platform_driver syscon_gpio_driver = {
 	.driver	= {
 		.name		= "gpio-syscon",
-		.owner		= THIS_MODULE,
 		.of_match_table	= syscon_gpio_ids,
 	},
 	.probe	= syscon_gpio_probe,
diff --git a/drivers/gpio/gpio-tb10x.c b/drivers/gpio/gpio-tb10x.c
index 9e615be8032c..62ab9f4b2cd3 100644
--- a/drivers/gpio/gpio-tb10x.c
+++ b/drivers/gpio/gpio-tb10x.c
@@ -195,18 +195,13 @@ static int tb10x_gpio_probe(struct platform_device *pdev)
 	if (of_property_read_u32(dn, "abilis,ngpio", &ngpio))
 		return -EINVAL;
 
-	mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	if (!mem) {
-		dev_err(&pdev->dev, "No memory resource defined.\n");
-		return -EINVAL;
-	}
-
 	tb10x_gpio = devm_kzalloc(&pdev->dev, sizeof(*tb10x_gpio), GFP_KERNEL);
 	if (tb10x_gpio == NULL)
 		return -ENOMEM;
 
 	spin_lock_init(&tb10x_gpio->spinlock);
 
+	mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	tb10x_gpio->base = devm_ioremap_resource(&pdev->dev, mem);
 	if (IS_ERR(tb10x_gpio->base))
 		return PTR_ERR(tb10x_gpio->base);
@@ -316,7 +311,6 @@ static struct platform_driver tb10x_gpio_driver = {
 	.driver = {
 		.name	= "tb10x-gpio",
 		.of_match_table = tb10x_gpio_dt_ids,
-		.owner	= THIS_MODULE,
 	}
 };
 
diff --git a/drivers/gpio/gpio-tc3589x.c b/drivers/gpio/gpio-tc3589x.c
index 7324869c38e0..abdcf58935f5 100644
--- a/drivers/gpio/gpio-tc3589x.c
+++ b/drivers/gpio/gpio-tc3589x.c
@@ -262,7 +262,7 @@ static int tc3589x_gpio_probe(struct platform_device *pdev)
 	tc3589x_gpio->chip = template_chip;
 	tc3589x_gpio->chip.ngpio = tc3589x->num_gpio;
 	tc3589x_gpio->chip.dev = &pdev->dev;
-	tc3589x_gpio->chip.base = (pdata) ? pdata->gpio_base : -1;
+	tc3589x_gpio->chip.base = -1;
 
 #ifdef CONFIG_OF_GPIO
 	tc3589x_gpio->chip.of_node = np;
@@ -300,6 +300,11 @@ static int tc3589x_gpio_probe(struct platform_device *pdev)
 		return ret;
 	}
 
+	gpiochip_set_chained_irqchip(&tc3589x_gpio->chip,
+				     &tc3589x_gpio_irq_chip,
+				     irq,
+				     NULL);
+
 	if (pdata && pdata->setup)
 		pdata->setup(tc3589x, tc3589x_gpio->chip.base);
 
diff --git a/drivers/gpio/gpio-tegra.c b/drivers/gpio/gpio-tegra.c
index 4e8fb8261a87..1741981d53c8 100644
--- a/drivers/gpio/gpio-tegra.c
+++ b/drivers/gpio/gpio-tegra.c
@@ -233,7 +233,7 @@ static int tegra_gpio_irq_set_type(struct irq_data *d, unsigned int type)
 		return -EINVAL;
 	}
 
-	ret = gpio_lock_as_irq(&tegra_gpio_chip, gpio);
+	ret = gpiochip_lock_as_irq(&tegra_gpio_chip, gpio);
 	if (ret) {
 		dev_err(dev, "unable to lock Tegra GPIO %d as IRQ\n", gpio);
 		return ret;
@@ -263,7 +263,7 @@ static void tegra_gpio_irq_shutdown(struct irq_data *d)
 {
 	int gpio = d->hwirq;
 
-	gpio_unlock_as_irq(&tegra_gpio_chip, gpio);
+	gpiochip_unlock_as_irq(&tegra_gpio_chip, gpio);
 }
 
 static void tegra_gpio_irq_handler(unsigned int irq, struct irq_desc *desc)
@@ -528,7 +528,6 @@ static int tegra_gpio_probe(struct platform_device *pdev)
 static struct platform_driver tegra_gpio_driver = {
 	.driver		= {
 		.name	= "tegra-gpio",
-		.owner	= THIS_MODULE,
 		.pm	= &tegra_gpio_pm_ops,
 		.of_match_table = tegra_gpio_of_match,
 	},
diff --git a/drivers/gpio/gpio-timberdale.c b/drivers/gpio/gpio-timberdale.c
index a685a3cbbc81..e8f97e03c9bb 100644
--- a/drivers/gpio/gpio-timberdale.c
+++ b/drivers/gpio/gpio-timberdale.c
@@ -330,7 +330,6 @@ static int timbgpio_remove(struct platform_device *pdev)
 static struct platform_driver timbgpio_platform_driver = {
 	.driver = {
 		.name	= DRIVER_NAME,
-		.owner	= THIS_MODULE,
 	},
 	.probe		= timbgpio_probe,
 	.remove		= timbgpio_remove,
diff --git a/drivers/gpio/gpio-tps65912.c b/drivers/gpio/gpio-tps65912.c
index 22052d84c63b..472fb5b8779f 100644
--- a/drivers/gpio/gpio-tps65912.c
+++ b/drivers/gpio/gpio-tps65912.c
@@ -124,7 +124,6 @@ static int tps65912_gpio_remove(struct platform_device *pdev)
 static struct platform_driver tps65912_gpio_driver = {
 	.driver = {
 		.name = "tps65912-gpio",
-		.owner = THIS_MODULE,
 	},
 	.probe = tps65912_gpio_probe,
 	.remove = tps65912_gpio_remove,
diff --git a/drivers/gpio/gpio-ts5500.c b/drivers/gpio/gpio-ts5500.c
index de18591ff11e..92fbabd82879 100644
--- a/drivers/gpio/gpio-ts5500.c
+++ b/drivers/gpio/gpio-ts5500.c
@@ -452,7 +452,6 @@ MODULE_DEVICE_TABLE(platform, ts5500_dio_ids);
 static struct platform_driver ts5500_dio_driver = {
 	.driver = {
 		.name = "ts5500-dio",
-		.owner = THIS_MODULE,
 	},
 	.probe = ts5500_dio_probe,
 	.remove = ts5500_dio_remove,
diff --git a/drivers/gpio/gpio-twl4030.c b/drivers/gpio/gpio-twl4030.c
index 118828b3736f..9e1dbb9877c1 100644
--- a/drivers/gpio/gpio-twl4030.c
+++ b/drivers/gpio/gpio-twl4030.c
@@ -605,7 +605,6 @@ MODULE_ALIAS("platform:twl4030_gpio");
 static struct platform_driver gpio_twl4030_driver = {
 	.driver = {
 		.name	= "twl4030_gpio",
-		.owner	= THIS_MODULE,
 		.of_match_table = twl_gpio_match,
 	},
 	.probe		= gpio_twl4030_probe,
diff --git a/drivers/gpio/gpio-twl6040.c b/drivers/gpio/gpio-twl6040.c
index f28e04b88aa9..c946e7eef3ee 100644
--- a/drivers/gpio/gpio-twl6040.c
+++ b/drivers/gpio/gpio-twl6040.c
@@ -121,7 +121,6 @@ MODULE_ALIAS("platform:twl6040-gpo");
 static struct platform_driver gpo_twl6040_driver = {
 	.driver = {
 		.name	= "twl6040-gpo",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= gpo_twl6040_probe,
 	.remove		= gpo_twl6040_remove,
diff --git a/drivers/gpio/gpio-tz1090-pdc.c b/drivers/gpio/gpio-tz1090-pdc.c
index f512da299b3d..d7536226b847 100644
--- a/drivers/gpio/gpio-tz1090-pdc.c
+++ b/drivers/gpio/gpio-tz1090-pdc.c
@@ -230,7 +230,6 @@ static struct of_device_id tz1090_pdc_gpio_of_match[] = {
 static struct platform_driver tz1090_pdc_gpio_driver = {
 	.driver = {
 		.name		= "tz1090-pdc-gpio",
-		.owner		= THIS_MODULE,
 		.of_match_table	= tz1090_pdc_gpio_of_match,
 	},
 	.probe		= tz1090_pdc_gpio_probe,
diff --git a/drivers/gpio/gpio-tz1090.c b/drivers/gpio/gpio-tz1090.c
index 5246a60eff6d..e3024bbba447 100644
--- a/drivers/gpio/gpio-tz1090.c
+++ b/drivers/gpio/gpio-tz1090.c
@@ -446,7 +446,7 @@ static int tz1090_gpio_bank_probe(struct tz1090_gpio_bank_info *info)
 	bank->irq = irq_of_parse_and_map(np, 0);
 
 	/* The interrupt is optional (it may be used by another core on chip) */
-	if (bank->irq < 0) {
+	if (!bank->irq) {
 		dev_info(dev, "IRQ not provided for bank %u, IRQs disabled\n",
 			 info->index);
 		return 0;
@@ -593,7 +593,6 @@ static struct of_device_id tz1090_gpio_of_match[] = {
 static struct platform_driver tz1090_gpio_driver = {
 	.driver = {
 		.name		= "tz1090-gpio",
-		.owner		= THIS_MODULE,
 		.of_match_table	= tz1090_gpio_of_match,
 	},
 	.probe		= tz1090_gpio_probe,
diff --git a/drivers/gpio/gpio-vf610.c b/drivers/gpio/gpio-vf610.c
new file mode 100644
index 000000000000..4ee4cee832ec
--- /dev/null
+++ b/drivers/gpio/gpio-vf610.c
@@ -0,0 +1,295 @@
+/*
+ * vf610 GPIO support through PORT and GPIO module
+ *
+ * Copyright (c) 2014 Toradex AG.
+ *
+ * Author: Stefan Agner <stefan@agner.ch>.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/bitops.h>
+#include <linux/err.h>
+#include <linux/gpio.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/ioport.h>
+#include <linux/irq.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/of_irq.h>
+
+#define VF610_GPIO_PER_PORT		32
+
+struct vf610_gpio_port {
+	struct gpio_chip gc;
+	void __iomem *base;
+	void __iomem *gpio_base;
+	u8 irqc[VF610_GPIO_PER_PORT];
+	int irq;
+};
+
+#define GPIO_PDOR		0x00
+#define GPIO_PSOR		0x04
+#define GPIO_PCOR		0x08
+#define GPIO_PTOR		0x0c
+#define GPIO_PDIR		0x10
+
+#define PORT_PCR(n)		((n) * 0x4)
+#define PORT_PCR_IRQC_OFFSET	16
+
+#define PORT_ISFR		0xa0
+#define PORT_DFER		0xc0
+#define PORT_DFCR		0xc4
+#define PORT_DFWR		0xc8
+
+#define PORT_INT_OFF		0x0
+#define PORT_INT_LOGIC_ZERO	0x8
+#define PORT_INT_RISING_EDGE	0x9
+#define PORT_INT_FALLING_EDGE	0xa
+#define PORT_INT_EITHER_EDGE	0xb
+#define PORT_INT_LOGIC_ONE	0xc
+
+static const struct of_device_id vf610_gpio_dt_ids[] = {
+	{ .compatible = "fsl,vf610-gpio" },
+	{ /* sentinel */ }
+};
+
+static inline void vf610_gpio_writel(u32 val, void __iomem *reg)
+{
+	writel_relaxed(val, reg);
+}
+
+static inline u32 vf610_gpio_readl(void __iomem *reg)
+{
+	return readl_relaxed(reg);
+}
+
+static int vf610_gpio_request(struct gpio_chip *chip, unsigned offset)
+{
+	return pinctrl_request_gpio(chip->base + offset);
+}
+
+static void vf610_gpio_free(struct gpio_chip *chip, unsigned offset)
+{
+	pinctrl_free_gpio(chip->base + offset);
+}
+
+static int vf610_gpio_get(struct gpio_chip *gc, unsigned int gpio)
+{
+	struct vf610_gpio_port *port =
+		container_of(gc, struct vf610_gpio_port, gc);
+
+	return !!(vf610_gpio_readl(port->gpio_base + GPIO_PDIR) & BIT(gpio));
+}
+
+static void vf610_gpio_set(struct gpio_chip *gc, unsigned int gpio, int val)
+{
+	struct vf610_gpio_port *port =
+		container_of(gc, struct vf610_gpio_port, gc);
+	unsigned long mask = BIT(gpio);
+
+	if (val)
+		vf610_gpio_writel(mask, port->gpio_base + GPIO_PSOR);
+	else
+		vf610_gpio_writel(mask, port->gpio_base + GPIO_PCOR);
+}
+
+static int vf610_gpio_direction_input(struct gpio_chip *chip, unsigned gpio)
+{
+	return pinctrl_gpio_direction_input(chip->base + gpio);
+}
+
+static int vf610_gpio_direction_output(struct gpio_chip *chip, unsigned gpio,
+				       int value)
+{
+	vf610_gpio_set(chip, gpio, value);
+
+	return pinctrl_gpio_direction_output(chip->base + gpio);
+}
+
+static void vf610_gpio_irq_handler(u32 irq, struct irq_desc *desc)
+{
+	struct vf610_gpio_port *port = irq_get_handler_data(irq);
+	struct irq_chip *chip = irq_desc_get_chip(desc);
+	int pin;
+	unsigned long irq_isfr;
+
+	chained_irq_enter(chip, desc);
+
+	irq_isfr = vf610_gpio_readl(port->base + PORT_ISFR);
+
+	for_each_set_bit(pin, &irq_isfr, VF610_GPIO_PER_PORT) {
+		vf610_gpio_writel(BIT(pin), port->base + PORT_ISFR);
+
+		generic_handle_irq(irq_find_mapping(port->gc.irqdomain, pin));
+	}
+
+	chained_irq_exit(chip, desc);
+}
+
+static void vf610_gpio_irq_ack(struct irq_data *d)
+{
+	struct vf610_gpio_port *port = irq_data_get_irq_chip_data(d);
+	int gpio = d->hwirq;
+
+	vf610_gpio_writel(BIT(gpio), port->base + PORT_ISFR);
+}
+
+static int vf610_gpio_irq_set_type(struct irq_data *d, u32 type)
+{
+	struct vf610_gpio_port *port = irq_data_get_irq_chip_data(d);
+	u8 irqc;
+
+	switch (type) {
+	case IRQ_TYPE_EDGE_RISING:
+		irqc = PORT_INT_RISING_EDGE;
+		break;
+	case IRQ_TYPE_EDGE_FALLING:
+		irqc = PORT_INT_FALLING_EDGE;
+		break;
+	case IRQ_TYPE_EDGE_BOTH:
+		irqc = PORT_INT_EITHER_EDGE;
+		break;
+	case IRQ_TYPE_LEVEL_LOW:
+		irqc = PORT_INT_LOGIC_ZERO;
+		break;
+	case IRQ_TYPE_LEVEL_HIGH:
+		irqc = PORT_INT_LOGIC_ONE;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	port->irqc[d->hwirq] = irqc;
+
+	return 0;
+}
+
+static void vf610_gpio_irq_mask(struct irq_data *d)
+{
+	struct vf610_gpio_port *port = irq_data_get_irq_chip_data(d);
+	void __iomem *pcr_base = port->base + PORT_PCR(d->hwirq);
+
+	vf610_gpio_writel(0, pcr_base);
+}
+
+static void vf610_gpio_irq_unmask(struct irq_data *d)
+{
+	struct vf610_gpio_port *port = irq_data_get_irq_chip_data(d);
+	void __iomem *pcr_base = port->base + PORT_PCR(d->hwirq);
+
+	vf610_gpio_writel(port->irqc[d->hwirq] << PORT_PCR_IRQC_OFFSET,
+			  pcr_base);
+}
+
+static int vf610_gpio_irq_set_wake(struct irq_data *d, u32 enable)
+{
+	struct vf610_gpio_port *port = irq_data_get_irq_chip_data(d);
+
+	if (enable)
+		enable_irq_wake(port->irq);
+	else
+		disable_irq_wake(port->irq);
+
+	return 0;
+}
+
+static struct irq_chip vf610_gpio_irq_chip = {
+	.name		= "gpio-vf610",
+	.irq_ack	= vf610_gpio_irq_ack,
+	.irq_mask	= vf610_gpio_irq_mask,
+	.irq_unmask	= vf610_gpio_irq_unmask,
+	.irq_set_type	= vf610_gpio_irq_set_type,
+	.irq_set_wake	= vf610_gpio_irq_set_wake,
+};
+
+static int vf610_gpio_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct device_node *np = dev->of_node;
+	struct vf610_gpio_port *port;
+	struct resource *iores;
+	struct gpio_chip *gc;
+	int ret;
+
+	port = devm_kzalloc(&pdev->dev, sizeof(*port), GFP_KERNEL);
+	if (!port)
+		return -ENOMEM;
+
+	iores = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	port->base = devm_ioremap_resource(dev, iores);
+	if (IS_ERR(port->base))
+		return PTR_ERR(port->base);
+
+	iores = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+	port->gpio_base = devm_ioremap_resource(dev, iores);
+	if (IS_ERR(port->gpio_base))
+		return PTR_ERR(port->gpio_base);
+
+	port->irq = platform_get_irq(pdev, 0);
+	if (port->irq < 0)
+		return port->irq;
+
+	gc = &port->gc;
+	gc->of_node = np;
+	gc->dev = dev;
+	gc->label = "vf610-gpio",
+	gc->ngpio = VF610_GPIO_PER_PORT,
+	gc->base = of_alias_get_id(np, "gpio") * VF610_GPIO_PER_PORT;
+
+	gc->request = vf610_gpio_request,
+	gc->free = vf610_gpio_free,
+	gc->direction_input = vf610_gpio_direction_input,
+	gc->get = vf610_gpio_get,
+	gc->direction_output = vf610_gpio_direction_output,
+	gc->set = vf610_gpio_set,
+
+	ret = gpiochip_add(gc);
+	if (ret < 0)
+		return ret;
+
+	/* Clear the interrupt status register for all GPIO's */
+	vf610_gpio_writel(~0, port->base + PORT_ISFR);
+
+	ret = gpiochip_irqchip_add(gc, &vf610_gpio_irq_chip, 0,
+				   handle_simple_irq, IRQ_TYPE_NONE);
+	if (ret) {
+		dev_err(dev, "failed to add irqchip\n");
+		gpiochip_remove(gc);
+		return ret;
+	}
+	gpiochip_set_chained_irqchip(gc, &vf610_gpio_irq_chip, port->irq,
+				     vf610_gpio_irq_handler);
+
+	return 0;
+}
+
+static struct platform_driver vf610_gpio_driver = {
+	.driver		= {
+		.name	= "gpio-vf610",
+		.owner	= THIS_MODULE,
+		.of_match_table = vf610_gpio_dt_ids,
+	},
+	.probe		= vf610_gpio_probe,
+};
+
+static int __init gpio_vf610_init(void)
+{
+	return platform_driver_register(&vf610_gpio_driver);
+}
+device_initcall(gpio_vf610_init);
+
+MODULE_AUTHOR("Stefan Agner <stefan@agner.ch>");
+MODULE_DESCRIPTION("Freescale VF610 GPIO");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpio/gpio-vr41xx.c b/drivers/gpio/gpio-vr41xx.c
index dbf28fa03f67..c1caa459c02d 100644
--- a/drivers/gpio/gpio-vr41xx.c
+++ b/drivers/gpio/gpio-vr41xx.c
@@ -138,7 +138,7 @@ static void unmask_giuint_low(struct irq_data *d)
 
 static unsigned int startup_giuint(struct irq_data *data)
 {
-	if (gpio_lock_as_irq(&vr41xx_gpio_chip, data->hwirq))
+	if (gpiochip_lock_as_irq(&vr41xx_gpio_chip, data->hwirq))
 		dev_err(vr41xx_gpio_chip.dev,
 			"unable to lock HW IRQ %lu for IRQ\n",
 			data->hwirq);
@@ -150,7 +150,7 @@ static unsigned int startup_giuint(struct irq_data *data)
 static void shutdown_giuint(struct irq_data *data)
 {
 	mask_giuint_low(data);
-	gpio_unlock_as_irq(&vr41xx_gpio_chip, data->hwirq);
+	gpiochip_unlock_as_irq(&vr41xx_gpio_chip, data->hwirq);
 }
 
 static struct irq_chip giuint_low_irq_chip = {
@@ -591,7 +591,6 @@ static struct platform_driver giu_device_driver = {
 	.remove		= giu_remove,
 	.driver		= {
 		.name	= "GIU",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/gpio/gpio-vx855.c b/drivers/gpio/gpio-vx855.c
index 85971d4e23c1..9d21d2fcc327 100644
--- a/drivers/gpio/gpio-vx855.c
+++ b/drivers/gpio/gpio-vx855.c
@@ -306,7 +306,6 @@ static int vx855gpio_remove(struct platform_device *pdev)
 static struct platform_driver vx855gpio_driver = {
 	.driver = {
 		.name	= MODULE_NAME,
-		.owner	= THIS_MODULE,
 	},
 	.probe		= vx855gpio_probe,
 	.remove		= vx855gpio_remove,
diff --git a/drivers/gpio/gpio-xgene.c b/drivers/gpio/gpio-xgene.c
new file mode 100644
index 000000000000..18a8182d4fec
--- /dev/null
+++ b/drivers/gpio/gpio-xgene.c
@@ -0,0 +1,243 @@
+/*
+ * AppliedMicro X-Gene SoC GPIO Driver
+ *
+ * Copyright (c) 2014, Applied Micro Circuits Corporation
+ * Author: Feng Kan <fkan@apm.com>.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/io.h>
+#include <linux/spinlock.h>
+#include <linux/platform_device.h>
+#include <linux/gpio/driver.h>
+#include <linux/types.h>
+#include <linux/bitops.h>
+
+#define GPIO_SET_DR_OFFSET	0x0C
+#define GPIO_DATA_OFFSET	0x14
+#define GPIO_BANK_STRIDE	0x0C
+
+#define XGENE_GPIOS_PER_BANK	16
+#define XGENE_MAX_GPIO_BANKS	3
+#define XGENE_MAX_GPIOS		(XGENE_GPIOS_PER_BANK * XGENE_MAX_GPIO_BANKS)
+
+#define GPIO_BIT_OFFSET(x)	(x % XGENE_GPIOS_PER_BANK)
+#define GPIO_BANK_OFFSET(x)	((x / XGENE_GPIOS_PER_BANK) * GPIO_BANK_STRIDE)
+
+struct xgene_gpio {
+	struct gpio_chip	chip;
+	void __iomem		*base;
+	spinlock_t		lock;
+#ifdef CONFIG_PM
+	u32			set_dr_val[XGENE_MAX_GPIO_BANKS];
+#endif
+};
+
+static inline struct xgene_gpio *to_xgene_gpio(struct gpio_chip *chip)
+{
+	return container_of(chip, struct xgene_gpio, chip);
+}
+
+static int xgene_gpio_get(struct gpio_chip *gc, unsigned int offset)
+{
+	struct xgene_gpio *chip = to_xgene_gpio(gc);
+	unsigned long bank_offset;
+	u32 bit_offset;
+
+	bank_offset = GPIO_DATA_OFFSET + GPIO_BANK_OFFSET(offset);
+	bit_offset = GPIO_BIT_OFFSET(offset);
+	return !!(ioread32(chip->base + bank_offset) & BIT(bit_offset));
+}
+
+static void __xgene_gpio_set(struct gpio_chip *gc, unsigned int offset, int val)
+{
+	struct xgene_gpio *chip = to_xgene_gpio(gc);
+	unsigned long bank_offset;
+	u32 setval, bit_offset;
+
+	bank_offset = GPIO_SET_DR_OFFSET + GPIO_BANK_OFFSET(offset);
+	bit_offset = GPIO_BIT_OFFSET(offset) + XGENE_GPIOS_PER_BANK;
+
+	setval = ioread32(chip->base + bank_offset);
+	if (val)
+		setval |= BIT(bit_offset);
+	else
+		setval &= ~BIT(bit_offset);
+	iowrite32(setval, chip->base + bank_offset);
+}
+
+static void xgene_gpio_set(struct gpio_chip *gc, unsigned int offset, int val)
+{
+	struct xgene_gpio *chip = to_xgene_gpio(gc);
+	unsigned long flags;
+
+	spin_lock_irqsave(&chip->lock, flags);
+	__xgene_gpio_set(gc, offset, val);
+	spin_unlock_irqrestore(&chip->lock, flags);
+}
+
+static int xgene_gpio_dir_in(struct gpio_chip *gc, unsigned int offset)
+{
+	struct xgene_gpio *chip = to_xgene_gpio(gc);
+	unsigned long flags, bank_offset;
+	u32 dirval, bit_offset;
+
+	bank_offset = GPIO_SET_DR_OFFSET + GPIO_BANK_OFFSET(offset);
+	bit_offset = GPIO_BIT_OFFSET(offset);
+
+	spin_lock_irqsave(&chip->lock, flags);
+
+	dirval = ioread32(chip->base + bank_offset);
+	dirval |= BIT(bit_offset);
+	iowrite32(dirval, chip->base + bank_offset);
+
+	spin_unlock_irqrestore(&chip->lock, flags);
+
+	return 0;
+}
+
+static int xgene_gpio_dir_out(struct gpio_chip *gc,
+					unsigned int offset, int val)
+{
+	struct xgene_gpio *chip = to_xgene_gpio(gc);
+	unsigned long flags, bank_offset;
+	u32 dirval, bit_offset;
+
+	bank_offset = GPIO_SET_DR_OFFSET + GPIO_BANK_OFFSET(offset);
+	bit_offset = GPIO_BIT_OFFSET(offset);
+
+	spin_lock_irqsave(&chip->lock, flags);
+
+	dirval = ioread32(chip->base + bank_offset);
+	dirval &= ~BIT(bit_offset);
+	iowrite32(dirval, chip->base + bank_offset);
+	__xgene_gpio_set(gc, offset, val);
+
+	spin_unlock_irqrestore(&chip->lock, flags);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM
+static int xgene_gpio_suspend(struct device *dev)
+{
+	struct xgene_gpio *gpio = dev_get_drvdata(dev);
+	unsigned long bank_offset;
+	unsigned int bank;
+
+	for (bank = 0; bank < XGENE_MAX_GPIO_BANKS; bank++) {
+		bank_offset = GPIO_SET_DR_OFFSET + bank * GPIO_BANK_STRIDE;
+		gpio->set_dr_val[bank] = ioread32(gpio->base + bank_offset);
+	}
+	return 0;
+}
+
+static int xgene_gpio_resume(struct device *dev)
+{
+	struct xgene_gpio *gpio = dev_get_drvdata(dev);
+	unsigned long bank_offset;
+	unsigned int bank;
+
+	for (bank = 0; bank < XGENE_MAX_GPIO_BANKS; bank++) {
+		bank_offset = GPIO_SET_DR_OFFSET + bank * GPIO_BANK_STRIDE;
+		iowrite32(gpio->set_dr_val[bank], gpio->base + bank_offset);
+	}
+	return 0;
+}
+
+static SIMPLE_DEV_PM_OPS(xgene_gpio_pm, xgene_gpio_suspend, xgene_gpio_resume);
+#define XGENE_GPIO_PM_OPS	(&xgene_gpio_pm)
+#else
+#define XGENE_GPIO_PM_OPS	NULL
+#endif
+
+static int xgene_gpio_probe(struct platform_device *pdev)
+{
+	struct resource *res;
+	struct xgene_gpio *gpio;
+	int err = 0;
+
+	gpio = devm_kzalloc(&pdev->dev, sizeof(*gpio), GFP_KERNEL);
+	if (!gpio) {
+		err = -ENOMEM;
+		goto err;
+	}
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	gpio->base = devm_ioremap_nocache(&pdev->dev, res->start,
+							resource_size(res));
+	if (!gpio->base) {
+		err = -ENOMEM;
+		goto err;
+	}
+
+	gpio->chip.ngpio = XGENE_MAX_GPIOS;
+
+	spin_lock_init(&gpio->lock);
+	gpio->chip.dev = &pdev->dev;
+	gpio->chip.direction_input = xgene_gpio_dir_in;
+	gpio->chip.direction_output = xgene_gpio_dir_out;
+	gpio->chip.get = xgene_gpio_get;
+	gpio->chip.set = xgene_gpio_set;
+	gpio->chip.label = dev_name(&pdev->dev);
+	gpio->chip.base = -1;
+
+	platform_set_drvdata(pdev, gpio);
+
+	err = gpiochip_add(&gpio->chip);
+	if (err) {
+		dev_err(&pdev->dev,
+			"failed to register gpiochip.\n");
+		goto err;
+	}
+
+	dev_info(&pdev->dev, "X-Gene GPIO driver registered.\n");
+	return 0;
+err:
+	dev_err(&pdev->dev, "X-Gene GPIO driver registration failed.\n");
+	return err;
+}
+
+static int xgene_gpio_remove(struct platform_device *pdev)
+{
+	struct xgene_gpio *gpio = platform_get_drvdata(pdev);
+
+	gpiochip_remove(&gpio->chip);
+	return 0;
+}
+
+static const struct of_device_id xgene_gpio_of_match[] = {
+	{ .compatible = "apm,xgene-gpio", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, xgene_gpio_of_match);
+
+static struct platform_driver xgene_gpio_driver = {
+	.driver = {
+		.name = "xgene-gpio",
+		.of_match_table = xgene_gpio_of_match,
+		.pm     = XGENE_GPIO_PM_OPS,
+	},
+	.probe = xgene_gpio_probe,
+	.remove = xgene_gpio_remove,
+};
+
+module_platform_driver(xgene_gpio_driver);
+
+MODULE_AUTHOR("Feng Kan <fkan@apm.com>");
+MODULE_DESCRIPTION("APM X-Gene GPIO driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/gpio/gpio-xilinx.c b/drivers/gpio/gpio-xilinx.c
index 12481867daf1..ba18b06c9a21 100644
--- a/drivers/gpio/gpio-xilinx.c
+++ b/drivers/gpio/gpio-xilinx.c
@@ -197,6 +197,7 @@ static int xgpio_of_probe(struct device_node *np)
 	struct xgpio_instance *chip;
 	int status = 0;
 	const u32 *tree_info;
+	u32 ngpio;
 
 	chip = kzalloc(sizeof(*chip), GFP_KERNEL);
 	if (!chip)
@@ -211,12 +212,13 @@ static int xgpio_of_probe(struct device_node *np)
 	/* Update GPIO direction shadow register with default value */
 	of_property_read_u32(np, "xlnx,tri-default", &chip->gpio_dir);
 
-	/* By default assume full GPIO controller */
-	chip->mmchip.gc.ngpio = 32;
-
-	/* Check device node and parent device node for device width */
-	of_property_read_u32(np, "xlnx,gpio-width",
-			      (u32 *)&chip->mmchip.gc.ngpio);
+	/*
+	 * Check device node and parent device node for device width
+	 * and assume default width of 32
+	 */
+	if (of_property_read_u32(np, "xlnx,gpio-width", &ngpio))
+		ngpio = 32;
+	chip->mmchip.gc.ngpio = (u16)ngpio;
 
 	spin_lock_init(&chip->gpio_lock);
 
@@ -258,12 +260,13 @@ static int xgpio_of_probe(struct device_node *np)
 		/* Update GPIO direction shadow register with default value */
 		of_property_read_u32(np, "xlnx,tri-default-2", &chip->gpio_dir);
 
-		/* By default assume full GPIO controller */
-		chip->mmchip.gc.ngpio = 32;
-
-		/* Check device node and parent device node for device width */
-		of_property_read_u32(np, "xlnx,gpio2-width",
-				     (u32 *)&chip->mmchip.gc.ngpio);
+		/*
+		 * Check device node and parent device node for device width
+		 * and assume default width of 32
+		 */
+		if (of_property_read_u32(np, "xlnx,gpio2-width", &ngpio))
+			ngpio = 32;
+		chip->mmchip.gc.ngpio = (u16)ngpio;
 
 		spin_lock_init(&chip->gpio_lock);
 
diff --git a/drivers/gpio/gpio-xtensa.c b/drivers/gpio/gpio-xtensa.c
index 7081304d6797..93ec95df67a3 100644
--- a/drivers/gpio/gpio-xtensa.c
+++ b/drivers/gpio/gpio-xtensa.c
@@ -157,7 +157,6 @@ static int xtensa_gpio_probe(struct platform_device *pdev)
 static struct platform_driver xtensa_gpio_driver = {
 	.driver		= {
 		.name		= "xtensa-gpio",
-		.owner		= THIS_MODULE,
 	},
 	.probe		= xtensa_gpio_probe,
 };
diff --git a/drivers/gpio/gpio-zevio.c b/drivers/gpio/gpio-zevio.c
index 54e54e4cc6c4..f769cd53f4e4 100644
--- a/drivers/gpio/gpio-zevio.c
+++ b/drivers/gpio/gpio-zevio.c
@@ -212,7 +212,6 @@ MODULE_DEVICE_TABLE(of, zevio_gpio_of_match);
 static struct platform_driver zevio_gpio_driver = {
 	.driver		= {
 		.name	= "gpio-zevio",
-		.owner	= THIS_MODULE,
 		.of_match_table = zevio_gpio_of_match,
 	},
 	.probe		= zevio_gpio_probe,
diff --git a/drivers/gpio/gpio-zynq.c b/drivers/gpio/gpio-zynq.c
index 31ad5df5dbc9..184c4b1b2558 100644
--- a/drivers/gpio/gpio-zynq.c
+++ b/drivers/gpio/gpio-zynq.c
@@ -88,16 +88,17 @@
  * @chip:	instance of the gpio_chip
  * @base_addr:	base address of the GPIO device
  * @clk:	clock resource for this controller
+ * @irq:	interrupt for the GPIO device
  */
 struct zynq_gpio {
 	struct gpio_chip chip;
 	void __iomem *base_addr;
 	struct clk *clk;
+	int irq;
 };
 
 static struct irq_chip zynq_gpio_level_irqchip;
 static struct irq_chip zynq_gpio_edge_irqchip;
-
 /**
  * zynq_gpio_get_bank_pin - Get the bank number and pin number within that bank
  * for a given pin in the GPIO device
@@ -138,6 +139,13 @@ static inline void zynq_gpio_get_bank_pin(unsigned int pin_num,
 	}
 }
 
+static const unsigned int zynq_gpio_bank_offset[] = {
+	ZYNQ_GPIO_BANK0_PIN_MIN,
+	ZYNQ_GPIO_BANK1_PIN_MIN,
+	ZYNQ_GPIO_BANK2_PIN_MIN,
+	ZYNQ_GPIO_BANK3_PIN_MIN,
+};
+
 /**
  * zynq_gpio_get_value - Get the state of the specified pin of GPIO device
  * @chip:	gpio_chip instance to be worked on
@@ -427,10 +435,9 @@ static int zynq_gpio_set_irq_type(struct irq_data *irq_data, unsigned int type)
 
 static int zynq_gpio_set_wake(struct irq_data *data, unsigned int on)
 {
-	if (on)
-		zynq_gpio_irq_unmask(data);
-	else
-		zynq_gpio_irq_mask(data);
+	struct zynq_gpio *gpio = irq_data_get_irq_chip_data(data);
+
+	irq_set_irq_wake(gpio->irq, on);
 
 	return 0;
 }
@@ -444,7 +451,8 @@ static struct irq_chip zynq_gpio_level_irqchip = {
 	.irq_unmask	= zynq_gpio_irq_unmask,
 	.irq_set_type	= zynq_gpio_set_irq_type,
 	.irq_set_wake	= zynq_gpio_set_wake,
-	.flags		= IRQCHIP_EOI_THREADED | IRQCHIP_EOI_IF_HANDLED,
+	.flags		= IRQCHIP_EOI_THREADED | IRQCHIP_EOI_IF_HANDLED |
+			  IRQCHIP_MASK_ON_SUSPEND,
 };
 
 static struct irq_chip zynq_gpio_edge_irqchip = {
@@ -455,8 +463,28 @@ static struct irq_chip zynq_gpio_edge_irqchip = {
 	.irq_unmask	= zynq_gpio_irq_unmask,
 	.irq_set_type	= zynq_gpio_set_irq_type,
 	.irq_set_wake	= zynq_gpio_set_wake,
+	.flags		= IRQCHIP_MASK_ON_SUSPEND,
 };
 
+static void zynq_gpio_handle_bank_irq(struct zynq_gpio *gpio,
+				      unsigned int bank_num,
+				      unsigned long pending)
+{
+	unsigned int bank_offset = zynq_gpio_bank_offset[bank_num];
+	struct irq_domain *irqdomain = gpio->chip.irqdomain;
+	int offset;
+
+	if (!pending)
+		return;
+
+	for_each_set_bit(offset, &pending, 32) {
+		unsigned int gpio_irq;
+
+		gpio_irq = irq_find_mapping(irqdomain, offset + bank_offset);
+		generic_handle_irq(gpio_irq);
+	}
+}
+
 /**
  * zynq_gpio_irqhandler - IRQ handler for the gpio banks of a gpio device
  * @irq:	irq number of the gpio bank where interrupt has occurred
@@ -482,18 +510,7 @@ static void zynq_gpio_irqhandler(unsigned int irq, struct irq_desc *desc)
 					ZYNQ_GPIO_INTSTS_OFFSET(bank_num));
 		int_enb = readl_relaxed(gpio->base_addr +
 					ZYNQ_GPIO_INTMASK_OFFSET(bank_num));
-		int_sts &= ~int_enb;
-		if (int_sts) {
-			int offset;
-			unsigned long pending = int_sts;
-
-			for_each_set_bit(offset, &pending, 32) {
-				unsigned int gpio_irq =
-					irq_find_mapping(gpio->chip.irqdomain,
-							offset);
-				generic_handle_irq(gpio_irq);
-			}
-		}
+		zynq_gpio_handle_bank_irq(gpio, bank_num, int_sts & ~int_enb);
 	}
 
 	chained_irq_exit(irqchip, desc);
@@ -501,7 +518,11 @@ static void zynq_gpio_irqhandler(unsigned int irq, struct irq_desc *desc)
 
 static int __maybe_unused zynq_gpio_suspend(struct device *dev)
 {
-	if (!device_may_wakeup(dev))
+	struct platform_device *pdev = to_platform_device(dev);
+	int irq = platform_get_irq(pdev, 0);
+	struct irq_data *data = irq_get_irq_data(irq);
+
+	if (!irqd_is_wakeup_set(data))
 		return pm_runtime_force_suspend(dev);
 
 	return 0;
@@ -509,7 +530,11 @@ static int __maybe_unused zynq_gpio_suspend(struct device *dev)
 
 static int __maybe_unused zynq_gpio_resume(struct device *dev)
 {
-	if (!device_may_wakeup(dev))
+	struct platform_device *pdev = to_platform_device(dev);
+	int irq = platform_get_irq(pdev, 0);
+	struct irq_data *data = irq_get_irq_data(irq);
+
+	if (!irqd_is_wakeup_set(data))
 		return pm_runtime_force_resume(dev);
 
 	return 0;
@@ -553,7 +578,7 @@ static void zynq_gpio_free(struct gpio_chip *chip, unsigned offset)
 
 static const struct dev_pm_ops zynq_gpio_dev_pm_ops = {
 	SET_SYSTEM_SLEEP_PM_OPS(zynq_gpio_suspend, zynq_gpio_resume)
-	SET_PM_RUNTIME_PM_OPS(zynq_gpio_runtime_suspend,
+	SET_RUNTIME_PM_OPS(zynq_gpio_runtime_suspend,
 			zynq_gpio_runtime_resume, NULL)
 };
 
@@ -570,7 +595,7 @@ static const struct dev_pm_ops zynq_gpio_dev_pm_ops = {
  */
 static int zynq_gpio_probe(struct platform_device *pdev)
 {
-	int ret, bank_num, irq;
+	int ret, bank_num;
 	struct zynq_gpio *gpio;
 	struct gpio_chip *chip;
 	struct resource *res;
@@ -586,10 +611,10 @@ static int zynq_gpio_probe(struct platform_device *pdev)
 	if (IS_ERR(gpio->base_addr))
 		return PTR_ERR(gpio->base_addr);
 
-	irq = platform_get_irq(pdev, 0);
-	if (irq < 0) {
+	gpio->irq = platform_get_irq(pdev, 0);
+	if (gpio->irq < 0) {
 		dev_err(&pdev->dev, "invalid IRQ\n");
-		return irq;
+		return gpio->irq;
 	}
 
 	/* configure the gpio chip */
@@ -637,19 +662,16 @@ static int zynq_gpio_probe(struct platform_device *pdev)
 		goto err_rm_gpiochip;
 	}
 
-	gpiochip_set_chained_irqchip(chip, &zynq_gpio_edge_irqchip, irq,
+	gpiochip_set_chained_irqchip(chip, &zynq_gpio_edge_irqchip, gpio->irq,
 				     zynq_gpio_irqhandler);
 
 	pm_runtime_set_active(&pdev->dev);
 	pm_runtime_enable(&pdev->dev);
 
-	device_set_wakeup_capable(&pdev->dev, 1);
-
 	return 0;
 
 err_rm_gpiochip:
-	if (gpiochip_remove(chip))
-		dev_err(&pdev->dev, "Failed to remove gpio chip\n");
+	gpiochip_remove(chip);
 err_disable_clk:
 	clk_disable_unprepare(gpio->clk);
 
@@ -664,16 +686,10 @@ err_disable_clk:
  */
 static int zynq_gpio_remove(struct platform_device *pdev)
 {
-	int ret;
 	struct zynq_gpio *gpio = platform_get_drvdata(pdev);
 
 	pm_runtime_get_sync(&pdev->dev);
-
-	ret = gpiochip_remove(&gpio->chip);
-	if (ret) {
-		dev_err(&pdev->dev, "Failed to remove gpio chip\n");
-		return ret;
-	}
+	gpiochip_remove(&gpio->chip);
 	clk_disable_unprepare(gpio->clk);
 	device_set_wakeup_capable(&pdev->dev, 0);
 	return 0;
@@ -688,7 +704,6 @@ MODULE_DEVICE_TABLE(of, zynq_gpio_of_match);
 static struct platform_driver zynq_gpio_driver = {
 	.driver	= {
 		.name = DRIVER_NAME,
-		.owner = THIS_MODULE,
 		.pm = &zynq_gpio_dev_pm_ops,
 		.of_match_table = zynq_gpio_of_match,
 	},
diff --git a/drivers/gpio/gpiolib-acpi.c b/drivers/gpio/gpiolib-acpi.c
index 687476fb39e3..c0929d938ced 100644
--- a/drivers/gpio/gpiolib-acpi.c
+++ b/drivers/gpio/gpiolib-acpi.c
@@ -11,12 +11,14 @@
  */
 
 #include <linux/errno.h>
+#include <linux/gpio.h>
 #include <linux/gpio/consumer.h>
 #include <linux/gpio/driver.h>
 #include <linux/export.h>
 #include <linux/acpi.h>
 #include <linux/interrupt.h>
 #include <linux/mutex.h>
+#include <linux/pinctrl/pinctrl.h>
 
 #include "gpiolib.h"
 
@@ -25,10 +27,12 @@ struct acpi_gpio_event {
 	acpi_handle handle;
 	unsigned int pin;
 	unsigned int irq;
+	struct gpio_desc *desc;
 };
 
 struct acpi_gpio_connection {
 	struct list_head node;
+	unsigned int pin;
 	struct gpio_desc *desc;
 };
 
@@ -53,6 +57,58 @@ static int acpi_gpiochip_find(struct gpio_chip *gc, void *data)
 	return ACPI_HANDLE(gc->dev) == data;
 }
 
+#ifdef CONFIG_PINCTRL
+/**
+ * acpi_gpiochip_pin_to_gpio_offset() - translates ACPI GPIO to Linux GPIO
+ * @chip: GPIO chip
+ * @pin: ACPI GPIO pin number from GpioIo/GpioInt resource
+ *
+ * Function takes ACPI GpioIo/GpioInt pin number as a parameter and
+ * translates it to a corresponding offset suitable to be passed to a
+ * GPIO controller driver.
+ *
+ * Typically the returned offset is same as @pin, but if the GPIO
+ * controller uses pin controller and the mapping is not contigous the
+ * offset might be different.
+ */
+static int acpi_gpiochip_pin_to_gpio_offset(struct gpio_chip *chip, int pin)
+{
+	struct gpio_pin_range *pin_range;
+
+	/* If there are no ranges in this chip, use 1:1 mapping */
+	if (list_empty(&chip->pin_ranges))
+		return pin;
+
+	list_for_each_entry(pin_range, &chip->pin_ranges, node) {
+		const struct pinctrl_gpio_range *range = &pin_range->range;
+		int i;
+
+		if (range->pins) {
+			for (i = 0; i < range->npins; i++) {
+				if (range->pins[i] == pin)
+					return range->base + i - chip->base;
+			}
+		} else {
+			if (pin >= range->pin_base &&
+			    pin < range->pin_base + range->npins) {
+				unsigned gpio_base;
+
+				gpio_base = range->base - chip->base;
+				return gpio_base + pin - range->pin_base;
+			}
+		}
+	}
+
+	return -EINVAL;
+}
+#else
+static inline int acpi_gpiochip_pin_to_gpio_offset(struct gpio_chip *chip,
+						   int pin)
+{
+	return pin;
+}
+#endif
+
 /**
  * acpi_get_gpiod() - Translate ACPI GPIO pin to GPIO descriptor usable with GPIO API
  * @path:	ACPI GPIO controller full path name, (e.g. "\\_SB.GPO1")
@@ -67,6 +123,7 @@ static struct gpio_desc *acpi_get_gpiod(char *path, int pin)
 	struct gpio_chip *chip;
 	acpi_handle handle;
 	acpi_status status;
+	int offset;
 
 	status = acpi_get_handle(NULL, path, &handle);
 	if (ACPI_FAILURE(status))
@@ -76,10 +133,11 @@ static struct gpio_desc *acpi_get_gpiod(char *path, int pin)
 	if (!chip)
 		return ERR_PTR(-ENODEV);
 
-	if (pin < 0 || pin > chip->ngpio)
-		return ERR_PTR(-EINVAL);
+	offset = acpi_gpiochip_pin_to_gpio_offset(chip, pin);
+	if (offset < 0)
+		return ERR_PTR(offset);
 
-	return gpiochip_get_desc(chip, pin);
+	return gpiochip_get_desc(chip, offset);
 }
 
 static irqreturn_t acpi_gpio_irq_handler(int irq, void *data)
@@ -143,21 +201,15 @@ static acpi_status acpi_gpiochip_request_interrupt(struct acpi_resource *ares,
 	if (!handler)
 		return AE_BAD_PARAMETER;
 
-	desc = gpiochip_get_desc(chip, pin);
+	desc = gpiochip_request_own_desc(chip, pin, "ACPI:Event");
 	if (IS_ERR(desc)) {
-		dev_err(chip->dev, "Failed to get GPIO descriptor\n");
-		return AE_ERROR;
-	}
-
-	ret = gpiochip_request_own_desc(desc, "ACPI:Event");
-	if (ret) {
 		dev_err(chip->dev, "Failed to request GPIO\n");
 		return AE_ERROR;
 	}
 
 	gpiod_direction_input(desc);
 
-	ret = gpio_lock_as_irq(chip, pin);
+	ret = gpiochip_lock_as_irq(chip, pin);
 	if (ret) {
 		dev_err(chip->dev, "Failed to lock GPIO as interrupt\n");
 		goto fail_free_desc;
@@ -197,6 +249,7 @@ static acpi_status acpi_gpiochip_request_interrupt(struct acpi_resource *ares,
 	event->handle = evt_handle;
 	event->irq = irq;
 	event->pin = pin;
+	event->desc = desc;
 
 	ret = request_threaded_irq(event->irq, NULL, handler, irqflags,
 				   "ACPI:Event", event);
@@ -212,7 +265,7 @@ static acpi_status acpi_gpiochip_request_interrupt(struct acpi_resource *ares,
 fail_free_event:
 	kfree(event);
 fail_unlock_irq:
-	gpio_unlock_as_irq(chip, pin);
+	gpiochip_unlock_as_irq(chip, pin);
 fail_free_desc:
 	gpiochip_free_own_desc(desc);
 
@@ -280,19 +333,55 @@ void acpi_gpiochip_free_interrupts(struct gpio_chip *chip)
 		struct gpio_desc *desc;
 
 		free_irq(event->irq, event);
-		desc = gpiochip_get_desc(chip, event->pin);
+		desc = event->desc;
 		if (WARN_ON(IS_ERR(desc)))
 			continue;
-		gpio_unlock_as_irq(chip, event->pin);
+		gpiochip_unlock_as_irq(chip, event->pin);
 		gpiochip_free_own_desc(desc);
 		list_del(&event->node);
 		kfree(event);
 	}
 }
 
+int acpi_dev_add_driver_gpios(struct acpi_device *adev,
+			      const struct acpi_gpio_mapping *gpios)
+{
+	if (adev && gpios) {
+		adev->driver_gpios = gpios;
+		return 0;
+	}
+	return -EINVAL;
+}
+EXPORT_SYMBOL_GPL(acpi_dev_add_driver_gpios);
+
+static bool acpi_get_driver_gpio_data(struct acpi_device *adev,
+				      const char *name, int index,
+				      struct acpi_reference_args *args)
+{
+	const struct acpi_gpio_mapping *gm;
+
+	if (!adev->driver_gpios)
+		return false;
+
+	for (gm = adev->driver_gpios; gm->name; gm++)
+		if (!strcmp(name, gm->name) && gm->data && index < gm->size) {
+			const struct acpi_gpio_params *par = gm->data + index;
+
+			args->adev = adev;
+			args->args[0] = par->crs_entry_index;
+			args->args[1] = par->line_index;
+			args->args[2] = par->active_low;
+			args->nargs = 3;
+			return true;
+		}
+
+	return false;
+}
+
 struct acpi_gpio_lookup {
 	struct acpi_gpio_info info;
 	int index;
+	int pin_index;
 	struct gpio_desc *desc;
 	int n;
 };
@@ -306,13 +395,24 @@ static int acpi_find_gpio(struct acpi_resource *ares, void *data)
 
 	if (lookup->n++ == lookup->index && !lookup->desc) {
 		const struct acpi_resource_gpio *agpio = &ares->data.gpio;
+		int pin_index = lookup->pin_index;
+
+		if (pin_index >= agpio->pin_table_length)
+			return 1;
 
 		lookup->desc = acpi_get_gpiod(agpio->resource_source.string_ptr,
-					      agpio->pin_table[0]);
+					      agpio->pin_table[pin_index]);
 		lookup->info.gpioint =
 			agpio->connection_type == ACPI_RESOURCE_GPIO_TYPE_INT;
-		lookup->info.active_low =
-			agpio->polarity == ACPI_ACTIVE_LOW;
+
+		/*
+		 * ActiveLow is only specified for GpioInt resource. If
+		 * GpioIo is used then the only way to set the flag is
+		 * to use _DSD "gpios" property.
+		 */
+		if (lookup->info.gpioint)
+			lookup->info.active_low =
+				agpio->polarity == ACPI_ACTIVE_LOW;
 	}
 
 	return 1;
@@ -320,40 +420,79 @@ static int acpi_find_gpio(struct acpi_resource *ares, void *data)
 
 /**
  * acpi_get_gpiod_by_index() - get a GPIO descriptor from device resources
- * @dev: pointer to a device to get GPIO from
+ * @adev: pointer to a ACPI device to get GPIO from
+ * @propname: Property name of the GPIO (optional)
  * @index: index of GpioIo/GpioInt resource (starting from %0)
  * @info: info pointer to fill in (optional)
  *
- * Function goes through ACPI resources for @dev and based on @index looks
+ * Function goes through ACPI resources for @adev and based on @index looks
  * up a GpioIo/GpioInt resource, translates it to the Linux GPIO descriptor,
  * and returns it. @index matches GpioIo/GpioInt resources only so if there
  * are total %3 GPIO resources, the index goes from %0 to %2.
  *
+ * If @propname is specified the GPIO is looked using device property. In
+ * that case @index is used to select the GPIO entry in the property value
+ * (in case of multiple).
+ *
  * If the GPIO cannot be translated or there is an error an ERR_PTR is
  * returned.
  *
  * Note: if the GPIO resource has multiple entries in the pin list, this
  * function only returns the first.
  */
-struct gpio_desc *acpi_get_gpiod_by_index(struct device *dev, int index,
+struct gpio_desc *acpi_get_gpiod_by_index(struct acpi_device *adev,
+					  const char *propname, int index,
 					  struct acpi_gpio_info *info)
 {
 	struct acpi_gpio_lookup lookup;
 	struct list_head resource_list;
-	struct acpi_device *adev;
-	acpi_handle handle;
+	bool active_low = false;
 	int ret;
 
-	if (!dev)
-		return ERR_PTR(-EINVAL);
-
-	handle = ACPI_HANDLE(dev);
-	if (!handle || acpi_bus_get_device(handle, &adev))
+	if (!adev)
 		return ERR_PTR(-ENODEV);
 
 	memset(&lookup, 0, sizeof(lookup));
 	lookup.index = index;
 
+	if (propname) {
+		struct acpi_reference_args args;
+
+		dev_dbg(&adev->dev, "GPIO: looking up %s\n", propname);
+
+		memset(&args, 0, sizeof(args));
+		ret = acpi_dev_get_property_reference(adev, propname,
+						      index, &args);
+		if (ret) {
+			bool found = acpi_get_driver_gpio_data(adev, propname,
+							       index, &args);
+			if (!found)
+				return ERR_PTR(ret);
+		}
+
+		/*
+		 * The property was found and resolved so need to
+		 * lookup the GPIO based on returned args instead.
+		 */
+		adev = args.adev;
+		if (args.nargs >= 2) {
+			lookup.index = args.args[0];
+			lookup.pin_index = args.args[1];
+			/*
+			 * 3rd argument, if present is used to
+			 * specify active_low.
+			 */
+			if (args.nargs >= 3)
+				active_low = !!args.args[2];
+		}
+
+		dev_dbg(&adev->dev, "GPIO: _DSD returned %s %zd %llu %llu %llu\n",
+			dev_name(&adev->dev), args.nargs,
+			args.args[0], args.args[1], args.args[2]);
+	} else {
+		dev_dbg(&adev->dev, "GPIO: looking up %d in _CRS\n", index);
+	}
+
 	INIT_LIST_HEAD(&resource_list);
 	ret = acpi_dev_get_resources(adev, &resource_list, acpi_find_gpio,
 				     &lookup);
@@ -362,8 +501,11 @@ struct gpio_desc *acpi_get_gpiod_by_index(struct device *dev, int index,
 
 	acpi_dev_free_resource_list(&resource_list);
 
-	if (lookup.desc && info)
+	if (lookup.desc && info) {
 		*info = lookup.info;
+		if (active_low)
+			info->active_low = active_low;
+	}
 
 	return lookup.desc ? lookup.desc : ERR_PTR(-ENOENT);
 }
@@ -409,26 +551,20 @@ acpi_gpio_adr_space_handler(u32 function, acpi_physical_address address,
 		struct gpio_desc *desc;
 		bool found;
 
-		desc = gpiochip_get_desc(chip, pin);
-		if (IS_ERR(desc)) {
-			status = AE_ERROR;
-			goto out;
-		}
-
 		mutex_lock(&achip->conn_lock);
 
 		found = false;
 		list_for_each_entry(conn, &achip->conns, node) {
-			if (conn->desc == desc) {
+			if (conn->pin == pin) {
 				found = true;
+				desc = conn->desc;
 				break;
 			}
 		}
 		if (!found) {
-			int ret;
-
-			ret = gpiochip_request_own_desc(desc, "ACPI:OpRegion");
-			if (ret) {
+			desc = gpiochip_request_own_desc(chip, pin,
+							 "ACPI:OpRegion");
+			if (IS_ERR(desc)) {
 				status = AE_ERROR;
 				mutex_unlock(&achip->conn_lock);
 				goto out;
@@ -465,6 +601,7 @@ acpi_gpio_adr_space_handler(u32 function, acpi_physical_address address,
 				goto out;
 			}
 
+			conn->pin = pin;
 			conn->desc = desc;
 			list_add_tail(&conn->node, &achip->conns);
 		}
diff --git a/drivers/gpio/gpiolib-legacy.c b/drivers/gpio/gpiolib-legacy.c
index 078ae6c2df79..8b830996fe02 100644
--- a/drivers/gpio/gpiolib-legacy.c
+++ b/drivers/gpio/gpiolib-legacy.c
@@ -24,6 +24,10 @@ int gpio_request_one(unsigned gpio, unsigned long flags, const char *label)
 
 	desc = gpio_to_desc(gpio);
 
+	/* Compatibility: assume unavailable "valid" GPIOs will appear later */
+	if (!desc && gpio_is_valid(gpio))
+		return -EPROBE_DEFER;
+
 	err = gpiod_request(desc, label);
 	if (err)
 		return err;
@@ -62,7 +66,13 @@ EXPORT_SYMBOL_GPL(gpio_request_one);
 
 int gpio_request(unsigned gpio, const char *label)
 {
-	return gpiod_request(gpio_to_desc(gpio), label);
+	struct gpio_desc *desc = gpio_to_desc(gpio);
+
+	/* Compatibility: assume unavailable "valid" GPIOs will appear later */
+	if (!desc && gpio_is_valid(gpio))
+		return -EPROBE_DEFER;
+
+	return gpiod_request(desc, label);
 }
 EXPORT_SYMBOL_GPL(gpio_request);
 
diff --git a/drivers/gpio/gpiolib-of.c b/drivers/gpio/gpiolib-of.c
index 604dbe60bdee..08261f2b3a82 100644
--- a/drivers/gpio/gpiolib-of.c
+++ b/drivers/gpio/gpiolib-of.c
@@ -45,8 +45,14 @@ static int of_gpiochip_find_and_xlate(struct gpio_chip *gc, void *data)
 		return false;
 
 	ret = gc->of_xlate(gc, &gg_data->gpiospec, gg_data->flags);
-	if (ret < 0)
-		return false;
+	if (ret < 0) {
+		/* We've found the gpio chip, but the translation failed.
+		 * Return true to stop looking and return the translation
+		 * error via out_gpio
+		 */
+		gg_data->out_gpio = ERR_PTR(ret);
+		return true;
+	 }
 
 	gg_data->out_gpio = gpiochip_get_desc(gc, ret);
 	return true;
diff --git a/drivers/gpio/gpiolib-sysfs.c b/drivers/gpio/gpiolib-sysfs.c
index 5f2150b619a7..f62aa115d79a 100644
--- a/drivers/gpio/gpiolib-sysfs.c
+++ b/drivers/gpio/gpiolib-sysfs.c
@@ -41,7 +41,7 @@ static DEFINE_MUTEX(sysfs_lock);
 static ssize_t gpio_direction_show(struct device *dev,
 		struct device_attribute *attr, char *buf)
 {
-	const struct gpio_desc	*desc = dev_get_drvdata(dev);
+	struct gpio_desc	*desc = dev_get_drvdata(dev);
 	ssize_t			status;
 
 	mutex_lock(&sysfs_lock);
@@ -128,7 +128,7 @@ static ssize_t gpio_value_store(struct device *dev,
 	return status;
 }
 
-static const DEVICE_ATTR(value, 0644,
+static DEVICE_ATTR(value, 0644,
 		gpio_value_show, gpio_value_store);
 
 static irqreturn_t gpio_sysfs_irq(int irq, void *priv)
@@ -161,7 +161,7 @@ static int gpio_setup_irq(struct gpio_desc *desc, struct device *dev,
 	desc->flags &= ~GPIO_TRIGGER_MASK;
 
 	if (!gpio_flags) {
-		gpio_unlock_as_irq(desc->chip, gpio_chip_hwgpio(desc));
+		gpiochip_unlock_as_irq(desc->chip, gpio_chip_hwgpio(desc));
 		ret = 0;
 		goto free_id;
 	}
@@ -200,7 +200,7 @@ static int gpio_setup_irq(struct gpio_desc *desc, struct device *dev,
 	if (ret < 0)
 		goto free_id;
 
-	ret = gpio_lock_as_irq(desc->chip, gpio_chip_hwgpio(desc));
+	ret = gpiochip_lock_as_irq(desc->chip, gpio_chip_hwgpio(desc));
 	if (ret < 0) {
 		gpiod_warn(desc, "failed to flag the GPIO for IRQ\n");
 		goto free_id;
@@ -353,17 +353,46 @@ static ssize_t gpio_active_low_store(struct device *dev,
 	return status ? : size;
 }
 
-static const DEVICE_ATTR(active_low, 0644,
+static DEVICE_ATTR(active_low, 0644,
 		gpio_active_low_show, gpio_active_low_store);
 
-static const struct attribute *gpio_attrs[] = {
+static umode_t gpio_is_visible(struct kobject *kobj, struct attribute *attr,
+			       int n)
+{
+	struct device *dev = container_of(kobj, struct device, kobj);
+	struct gpio_desc *desc = dev_get_drvdata(dev);
+	umode_t mode = attr->mode;
+	bool show_direction = test_bit(FLAG_SYSFS_DIR, &desc->flags);
+
+	if (attr == &dev_attr_direction.attr) {
+		if (!show_direction)
+			mode = 0;
+	} else if (attr == &dev_attr_edge.attr) {
+		if (gpiod_to_irq(desc) < 0)
+			mode = 0;
+		if (!show_direction && test_bit(FLAG_IS_OUT, &desc->flags))
+			mode = 0;
+	}
+
+	return mode;
+}
+
+static struct attribute *gpio_attrs[] = {
+	&dev_attr_direction.attr,
+	&dev_attr_edge.attr,
 	&dev_attr_value.attr,
 	&dev_attr_active_low.attr,
 	NULL,
 };
 
-static const struct attribute_group gpio_attr_group = {
-	.attrs = (struct attribute **) gpio_attrs,
+static const struct attribute_group gpio_group = {
+	.attrs = gpio_attrs,
+	.is_visible = gpio_is_visible,
+};
+
+static const struct attribute_group *gpio_groups[] = {
+	&gpio_group,
+	NULL
 };
 
 /*
@@ -400,16 +429,13 @@ static ssize_t chip_ngpio_show(struct device *dev,
 }
 static DEVICE_ATTR(ngpio, 0444, chip_ngpio_show, NULL);
 
-static const struct attribute *gpiochip_attrs[] = {
+static struct attribute *gpiochip_attrs[] = {
 	&dev_attr_base.attr,
 	&dev_attr_label.attr,
 	&dev_attr_ngpio.attr,
 	NULL,
 };
-
-static const struct attribute_group gpiochip_attr_group = {
-	.attrs = (struct attribute **) gpiochip_attrs,
-};
+ATTRIBUTE_GROUPS(gpiochip);
 
 /*
  * /sys/class/gpio/export ... write-only
@@ -556,45 +582,30 @@ int gpiod_export(struct gpio_desc *desc, bool direction_may_change)
 		goto fail_unlock;
 	}
 
-	if (!desc->chip->direction_input || !desc->chip->direction_output)
-		direction_may_change = false;
+	if (desc->chip->direction_input && desc->chip->direction_output &&
+			direction_may_change) {
+		set_bit(FLAG_SYSFS_DIR, &desc->flags);
+	}
+
 	spin_unlock_irqrestore(&gpio_lock, flags);
 
 	offset = gpio_chip_hwgpio(desc);
 	if (desc->chip->names && desc->chip->names[offset])
 		ioname = desc->chip->names[offset];
 
-	dev = device_create(&gpio_class, desc->chip->dev, MKDEV(0, 0),
-			    desc, ioname ? ioname : "gpio%u",
-			    desc_to_gpio(desc));
+	dev = device_create_with_groups(&gpio_class, desc->chip->dev,
+					MKDEV(0, 0), desc, gpio_groups,
+					ioname ? ioname : "gpio%u",
+					desc_to_gpio(desc));
 	if (IS_ERR(dev)) {
 		status = PTR_ERR(dev);
 		goto fail_unlock;
 	}
 
-	status = sysfs_create_group(&dev->kobj, &gpio_attr_group);
-	if (status)
-		goto fail_unregister_device;
-
-	if (direction_may_change) {
-		status = device_create_file(dev, &dev_attr_direction);
-		if (status)
-			goto fail_unregister_device;
-	}
-
-	if (gpiod_to_irq(desc) >= 0 && (direction_may_change ||
-				       !test_bit(FLAG_IS_OUT, &desc->flags))) {
-		status = device_create_file(dev, &dev_attr_edge);
-		if (status)
-			goto fail_unregister_device;
-	}
-
 	set_bit(FLAG_EXPORT, &desc->flags);
 	mutex_unlock(&sysfs_lock);
 	return 0;
 
-fail_unregister_device:
-	device_unregister(dev);
 fail_unlock:
 	mutex_unlock(&sysfs_lock);
 	gpiod_dbg(desc, "%s: status %d\n", __func__, status);
@@ -718,6 +729,7 @@ void gpiod_unexport(struct gpio_desc *desc)
 		dev = class_find_device(&gpio_class, NULL, desc, match_export);
 		if (dev) {
 			gpio_setup_irq(desc, dev, 0);
+			clear_bit(FLAG_SYSFS_DIR, &desc->flags);
 			clear_bit(FLAG_EXPORT, &desc->flags);
 		} else
 			status = -ENODEV;
@@ -750,13 +762,13 @@ int gpiochip_export(struct gpio_chip *chip)
 
 	/* use chip->base for the ID; it's already known to be unique */
 	mutex_lock(&sysfs_lock);
-	dev = device_create(&gpio_class, chip->dev, MKDEV(0, 0), chip,
-				"gpiochip%d", chip->base);
-	if (!IS_ERR(dev)) {
-		status = sysfs_create_group(&dev->kobj,
-				&gpiochip_attr_group);
-	} else
+	dev = device_create_with_groups(&gpio_class, chip->dev, MKDEV(0, 0),
+					chip, gpiochip_groups,
+					"gpiochip%d", chip->base);
+	if (IS_ERR(dev))
 		status = PTR_ERR(dev);
+	else
+		status = 0;
 	chip->exported = (status == 0);
 	mutex_unlock(&sysfs_lock);
 
diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c
index c68d037de656..568aa2b6bdb0 100644
--- a/drivers/gpio/gpiolib.c
+++ b/drivers/gpio/gpiolib.c
@@ -47,8 +47,6 @@
  */
 DEFINE_SPINLOCK(gpio_lock);
 
-static struct gpio_desc gpio_desc[ARCH_NR_GPIOS];
-
 #define GPIO_OFFSET_VALID(chip, offset) (offset >= 0 && offset < chip->ngpio)
 
 static DEFINE_MUTEX(gpio_lookup_lock);
@@ -65,10 +63,24 @@ static inline void desc_set_label(struct gpio_desc *d, const char *label)
  */
 struct gpio_desc *gpio_to_desc(unsigned gpio)
 {
-	if (WARN(!gpio_is_valid(gpio), "invalid GPIO %d\n", gpio))
-		return NULL;
-	else
-		return &gpio_desc[gpio];
+	struct gpio_chip *chip;
+	unsigned long flags;
+
+	spin_lock_irqsave(&gpio_lock, flags);
+
+	list_for_each_entry(chip, &gpio_chips, list) {
+		if (chip->base <= gpio && chip->base + chip->ngpio > gpio) {
+			spin_unlock_irqrestore(&gpio_lock, flags);
+			return &chip->desc[gpio - chip->base];
+		}
+	}
+
+	spin_unlock_irqrestore(&gpio_lock, flags);
+
+	if (!gpio_is_valid(gpio))
+		WARN(1, "invalid GPIO %d\n", gpio);
+
+	return NULL;
 }
 EXPORT_SYMBOL_GPL(gpio_to_desc);
 
@@ -91,7 +103,7 @@ struct gpio_desc *gpiochip_get_desc(struct gpio_chip *chip,
  */
 int desc_to_gpio(const struct gpio_desc *desc)
 {
-	return desc - &gpio_desc[0];
+	return desc->chip->base + (desc - &desc->chip->desc[0]);
 }
 EXPORT_SYMBOL_GPL(desc_to_gpio);
 
@@ -138,7 +150,7 @@ static int gpiochip_find_base(int ngpio)
  *
  * This function may sleep if gpiod_cansleep() is true.
  */
-int gpiod_get_direction(const struct gpio_desc *desc)
+int gpiod_get_direction(struct gpio_desc *desc)
 {
 	struct gpio_chip	*chip;
 	unsigned		offset;
@@ -154,13 +166,11 @@ int gpiod_get_direction(const struct gpio_desc *desc)
 	if (status > 0) {
 		/* GPIOF_DIR_IN, or other positive */
 		status = 1;
-		/* FLAG_IS_OUT is just a cache of the result of get_direction(),
-		 * so it does not affect constness per se */
-		clear_bit(FLAG_IS_OUT, &((struct gpio_desc *)desc)->flags);
+		clear_bit(FLAG_IS_OUT, &desc->flags);
 	}
 	if (status == 0) {
 		/* GPIOF_DIR_OUT */
-		set_bit(FLAG_IS_OUT, &((struct gpio_desc *)desc)->flags);
+		set_bit(FLAG_IS_OUT, &desc->flags);
 	}
 	return status;
 }
@@ -206,7 +216,7 @@ static int gpiochip_add_to_list(struct gpio_chip *chip)
 /**
  * gpiochip_add() - register a gpio_chip
  * @chip: the chip to register, with chip->base initialized
- * Context: potentially before irqs or kmalloc will work
+ * Context: potentially before irqs will work
  *
  * Returns a negative errno if the chip can't be registered, such as
  * because the chip->base is invalid or already associated with a
@@ -226,12 +236,11 @@ int gpiochip_add(struct gpio_chip *chip)
 	int		status = 0;
 	unsigned	id;
 	int		base = chip->base;
+	struct gpio_desc *descs;
 
-	if ((!gpio_is_valid(base) || !gpio_is_valid(base + chip->ngpio - 1))
-			&& base >= 0) {
-		status = -EINVAL;
-		goto fail;
-	}
+	descs = kcalloc(chip->ngpio, sizeof(descs[0]), GFP_KERNEL);
+	if (!descs)
+		return -ENOMEM;
 
 	spin_lock_irqsave(&gpio_lock, flags);
 
@@ -239,33 +248,34 @@ int gpiochip_add(struct gpio_chip *chip)
 		base = gpiochip_find_base(chip->ngpio);
 		if (base < 0) {
 			status = base;
-			goto unlock;
+			spin_unlock_irqrestore(&gpio_lock, flags);
+			goto err_free_descs;
 		}
 		chip->base = base;
 	}
 
 	status = gpiochip_add_to_list(chip);
+	if (status) {
+		spin_unlock_irqrestore(&gpio_lock, flags);
+		goto err_free_descs;
+	}
 
-	if (status == 0) {
-		chip->desc = &gpio_desc[chip->base];
-
-		for (id = 0; id < chip->ngpio; id++) {
-			struct gpio_desc *desc = &chip->desc[id];
-			desc->chip = chip;
-
-			/* REVISIT:  most hardware initializes GPIOs as
-			 * inputs (often with pullups enabled) so power
-			 * usage is minimized.  Linux code should set the
-			 * gpio direction first thing; but until it does,
-			 * and in case chip->get_direction is not set,
-			 * we may expose the wrong direction in sysfs.
-			 */
-			desc->flags = !chip->direction_input
-				? (1 << FLAG_IS_OUT)
-				: 0;
-		}
+	for (id = 0; id < chip->ngpio; id++) {
+		struct gpio_desc *desc = &descs[id];
+
+		desc->chip = chip;
+
+		/* REVISIT: most hardware initializes GPIOs as inputs (often
+		 * with pullups enabled) so power usage is minimized. Linux
+		 * code should set the gpio direction first thing; but until
+		 * it does, and in case chip->get_direction is not set, we may
+		 * expose the wrong direction in sysfs.
+		 */
+		desc->flags = !chip->direction_input ? (1 << FLAG_IS_OUT) : 0;
 	}
 
+	chip->desc = descs;
+
 	spin_unlock_irqrestore(&gpio_lock, flags);
 
 #ifdef CONFIG_PINCTRL
@@ -275,12 +285,9 @@ int gpiochip_add(struct gpio_chip *chip)
 	of_gpiochip_add(chip);
 	acpi_gpiochip_add(chip);
 
-	if (status)
-		goto fail;
-
 	status = gpiochip_export(chip);
 	if (status)
-		goto fail;
+		goto err_remove_chip;
 
 	pr_debug("%s: registered GPIOs %d to %d on device: %s\n", __func__,
 		chip->base, chip->base + chip->ngpio - 1,
@@ -288,9 +295,16 @@ int gpiochip_add(struct gpio_chip *chip)
 
 	return 0;
 
-unlock:
+err_remove_chip:
+	acpi_gpiochip_remove(chip);
+	of_gpiochip_remove(chip);
+	spin_lock_irqsave(&gpio_lock, flags);
+	list_del(&chip->list);
 	spin_unlock_irqrestore(&gpio_lock, flags);
-fail:
+	chip->desc = NULL;
+err_free_descs:
+	kfree(descs);
+
 	/* failures here can mean systems won't boot... */
 	pr_err("%s: GPIOs %d..%d (%s) failed to register\n", __func__,
 		chip->base, chip->base + chip->ngpio - 1,
@@ -308,39 +322,32 @@ static void gpiochip_irqchip_remove(struct gpio_chip *gpiochip);
  *
  * A gpio_chip with any GPIOs still requested may not be removed.
  */
-int gpiochip_remove(struct gpio_chip *chip)
+void gpiochip_remove(struct gpio_chip *chip)
 {
 	unsigned long	flags;
-	int		status = 0;
 	unsigned	id;
 
-	acpi_gpiochip_remove(chip);
-
-	spin_lock_irqsave(&gpio_lock, flags);
+	gpiochip_unexport(chip);
 
 	gpiochip_irqchip_remove(chip);
+
+	acpi_gpiochip_remove(chip);
 	gpiochip_remove_pin_ranges(chip);
 	of_gpiochip_remove(chip);
 
+	spin_lock_irqsave(&gpio_lock, flags);
 	for (id = 0; id < chip->ngpio; id++) {
-		if (test_bit(FLAG_REQUESTED, &chip->desc[id].flags)) {
-			status = -EBUSY;
-			break;
-		}
-	}
-	if (status == 0) {
-		for (id = 0; id < chip->ngpio; id++)
-			chip->desc[id].chip = NULL;
-
-		list_del(&chip->list);
+		if (test_bit(FLAG_REQUESTED, &chip->desc[id].flags))
+			dev_crit(chip->dev, "REMOVING GPIOCHIP WITH GPIOS STILL REQUESTED\n");
 	}
+	for (id = 0; id < chip->ngpio; id++)
+		chip->desc[id].chip = NULL;
 
+	list_del(&chip->list);
 	spin_unlock_irqrestore(&gpio_lock, flags);
 
-	if (status == 0)
-		gpiochip_unexport(chip);
-
-	return status;
+	kfree(chip->desc);
+	chip->desc = NULL;
 }
 EXPORT_SYMBOL_GPL(gpiochip_remove);
 
@@ -395,30 +402,47 @@ static struct gpio_chip *find_chip_by_name(const char *name)
  */
 
 /**
- * gpiochip_add_chained_irqchip() - adds a chained irqchip to a gpiochip
- * @gpiochip: the gpiochip to add the irqchip to
- * @irqchip: the irqchip to add to the gpiochip
+ * gpiochip_set_chained_irqchip() - sets a chained irqchip to a gpiochip
+ * @gpiochip: the gpiochip to set the irqchip chain to
+ * @irqchip: the irqchip to chain to the gpiochip
  * @parent_irq: the irq number corresponding to the parent IRQ for this
  * chained irqchip
  * @parent_handler: the parent interrupt handler for the accumulated IRQ
- * coming out of the gpiochip
+ * coming out of the gpiochip. If the interrupt is nested rather than
+ * cascaded, pass NULL in this handler argument
  */
 void gpiochip_set_chained_irqchip(struct gpio_chip *gpiochip,
 				  struct irq_chip *irqchip,
 				  int parent_irq,
 				  irq_flow_handler_t parent_handler)
 {
-	if (gpiochip->can_sleep) {
-		chip_err(gpiochip, "you cannot have chained interrupts on a chip that may sleep\n");
+	unsigned int offset;
+
+	if (!gpiochip->irqdomain) {
+		chip_err(gpiochip, "called %s before setting up irqchip\n",
+			 __func__);
 		return;
 	}
 
-	/*
-	 * The parent irqchip is already using the chip_data for this
-	 * irqchip, so our callbacks simply use the handler_data.
-	 */
-	irq_set_handler_data(parent_irq, gpiochip);
-	irq_set_chained_handler(parent_irq, parent_handler);
+	if (parent_handler) {
+		if (gpiochip->can_sleep) {
+			chip_err(gpiochip,
+				 "you cannot have chained interrupts on a "
+				 "chip that may sleep\n");
+			return;
+		}
+		/*
+		 * The parent irqchip is already using the chip_data for this
+		 * irqchip, so our callbacks simply use the handler_data.
+		 */
+		irq_set_handler_data(parent_irq, gpiochip);
+		irq_set_chained_handler(parent_irq, parent_handler);
+	}
+
+	/* Set the parent IRQ for all affected IRQs */
+	for (offset = 0; offset < gpiochip->ngpio; offset++)
+		irq_set_parent(irq_find_mapping(gpiochip->irqdomain, offset),
+			       parent_irq);
 }
 EXPORT_SYMBOL_GPL(gpiochip_set_chained_irqchip);
 
@@ -447,7 +471,7 @@ static int gpiochip_irq_map(struct irq_domain *d, unsigned int irq,
 	irq_set_lockdep_class(irq, &gpiochip_irq_lock_class);
 	irq_set_chip_and_handler(irq, chip->irqchip, chip->irq_handler);
 	/* Chips that can sleep need nested thread handlers */
-	if (chip->can_sleep)
+	if (chip->can_sleep && !chip->irq_not_threaded)
 		irq_set_nested_thread(irq, 1);
 #ifdef CONFIG_ARM
 	set_irq_flags(irq, IRQF_VALID);
@@ -488,7 +512,7 @@ static int gpiochip_irq_reqres(struct irq_data *d)
 {
 	struct gpio_chip *chip = irq_data_get_irq_chip_data(d);
 
-	if (gpio_lock_as_irq(chip, d->hwirq)) {
+	if (gpiochip_lock_as_irq(chip, d->hwirq)) {
 		chip_err(chip,
 			"unable to lock HW IRQ %lu for IRQ\n",
 			d->hwirq);
@@ -501,7 +525,7 @@ static void gpiochip_irq_relres(struct irq_data *d)
 {
 	struct gpio_chip *chip = irq_data_get_irq_chip_data(d);
 
-	gpio_unlock_as_irq(chip, d->hwirq);
+	gpiochip_unlock_as_irq(chip, d->hwirq);
 }
 
 static int gpiochip_to_irq(struct gpio_chip *chip, unsigned offset)
@@ -524,7 +548,8 @@ static void gpiochip_irqchip_remove(struct gpio_chip *gpiochip)
 	/* Remove all IRQ mappings and delete the domain */
 	if (gpiochip->irqdomain) {
 		for (offset = 0; offset < gpiochip->ngpio; offset++)
-			irq_dispose_mapping(gpiochip->irq_base + offset);
+			irq_dispose_mapping(
+				irq_find_mapping(gpiochip->irqdomain, offset));
 		irq_domain_remove(gpiochip->irqdomain);
 	}
 
@@ -895,12 +920,22 @@ EXPORT_SYMBOL_GPL(gpiochip_is_requested);
  * allows the GPIO chip module to be unloaded as needed (we assume that the
  * GPIO chip driver handles freeing the GPIOs it has requested).
  */
-int gpiochip_request_own_desc(struct gpio_desc *desc, const char *label)
+struct gpio_desc *gpiochip_request_own_desc(struct gpio_chip *chip, u16 hwnum,
+					    const char *label)
 {
-	if (!desc || !desc->chip)
-		return -EINVAL;
+	struct gpio_desc *desc = gpiochip_get_desc(chip, hwnum);
+	int err;
+
+	if (IS_ERR(desc)) {
+		chip_err(chip, "failed to get GPIO descriptor\n");
+		return desc;
+	}
 
-	return __gpiod_request(desc, label);
+	err = __gpiod_request(desc, label);
+	if (err < 0)
+		return ERR_PTR(err);
+
+	return desc;
 }
 EXPORT_SYMBOL_GPL(gpiochip_request_own_desc);
 
@@ -1236,6 +1271,88 @@ static void _gpiod_set_raw_value(struct gpio_desc *desc, bool value)
 		chip->set(chip, gpio_chip_hwgpio(desc), value);
 }
 
+/*
+ * set multiple outputs on the same chip;
+ * use the chip's set_multiple function if available;
+ * otherwise set the outputs sequentially;
+ * @mask: bit mask array; one bit per output; BITS_PER_LONG bits per word
+ *        defines which outputs are to be changed
+ * @bits: bit value array; one bit per output; BITS_PER_LONG bits per word
+ *        defines the values the outputs specified by mask are to be set to
+ */
+static void gpio_chip_set_multiple(struct gpio_chip *chip,
+				   unsigned long *mask, unsigned long *bits)
+{
+	if (chip->set_multiple) {
+		chip->set_multiple(chip, mask, bits);
+	} else {
+		int i;
+		for (i = 0; i < chip->ngpio; i++) {
+			if (mask[BIT_WORD(i)] == 0) {
+				/* no more set bits in this mask word;
+				 * skip ahead to the next word */
+				i = (BIT_WORD(i) + 1) * BITS_PER_LONG - 1;
+				continue;
+			}
+			/* set outputs if the corresponding mask bit is set */
+			if (__test_and_clear_bit(i, mask)) {
+				chip->set(chip, i, test_bit(i, bits));
+			}
+		}
+	}
+}
+
+static void gpiod_set_array_priv(bool raw, bool can_sleep,
+				 unsigned int array_size,
+				 struct gpio_desc **desc_array,
+				 int *value_array)
+{
+	int i = 0;
+
+	while (i < array_size) {
+		struct gpio_chip *chip = desc_array[i]->chip;
+		unsigned long mask[BITS_TO_LONGS(chip->ngpio)];
+		unsigned long bits[BITS_TO_LONGS(chip->ngpio)];
+		int count = 0;
+
+		if (!can_sleep) {
+			WARN_ON(chip->can_sleep);
+		}
+		memset(mask, 0, sizeof(mask));
+		do {
+			struct gpio_desc *desc = desc_array[i];
+			int hwgpio = gpio_chip_hwgpio(desc);
+			int value = value_array[i];
+
+			if (!raw && test_bit(FLAG_ACTIVE_LOW, &desc->flags))
+				value = !value;
+			trace_gpio_value(desc_to_gpio(desc), 0, value);
+			/*
+			 * collect all normal outputs belonging to the same chip
+			 * open drain and open source outputs are set individually
+			 */
+			if (test_bit(FLAG_OPEN_DRAIN, &desc->flags)) {
+				_gpio_set_open_drain_value(desc,value);
+			} else if (test_bit(FLAG_OPEN_SOURCE, &desc->flags)) {
+				_gpio_set_open_source_value(desc, value);
+			} else {
+				__set_bit(hwgpio, mask);
+				if (value) {
+					__set_bit(hwgpio, bits);
+				} else {
+					__clear_bit(hwgpio, bits);
+				}
+				count++;
+			}
+			i++;
+		} while ((i < array_size) && (desc_array[i]->chip == chip));
+		/* push collected bits to outputs */
+		if (count != 0) {
+			gpio_chip_set_multiple(chip, mask, bits);
+		}
+	}
+}
+
 /**
  * gpiod_set_raw_value() - assign a gpio's raw value
  * @desc: gpio whose value will be assigned
@@ -1281,6 +1398,48 @@ void gpiod_set_value(struct gpio_desc *desc, int value)
 EXPORT_SYMBOL_GPL(gpiod_set_value);
 
 /**
+ * gpiod_set_raw_array() - assign values to an array of GPIOs
+ * @array_size: number of elements in the descriptor / value arrays
+ * @desc_array: array of GPIO descriptors whose values will be assigned
+ * @value_array: array of values to assign
+ *
+ * Set the raw values of the GPIOs, i.e. the values of the physical lines
+ * without regard for their ACTIVE_LOW status.
+ *
+ * This function should be called from contexts where we cannot sleep, and will
+ * complain if the GPIO chip functions potentially sleep.
+ */
+void gpiod_set_raw_array(unsigned int array_size,
+			 struct gpio_desc **desc_array, int *value_array)
+{
+	if (!desc_array)
+		return;
+	gpiod_set_array_priv(true, false, array_size, desc_array, value_array);
+}
+EXPORT_SYMBOL_GPL(gpiod_set_raw_array);
+
+/**
+ * gpiod_set_array() - assign values to an array of GPIOs
+ * @array_size: number of elements in the descriptor / value arrays
+ * @desc_array: array of GPIO descriptors whose values will be assigned
+ * @value_array: array of values to assign
+ *
+ * Set the logical values of the GPIOs, i.e. taking their ACTIVE_LOW status
+ * into account.
+ *
+ * This function should be called from contexts where we cannot sleep, and will
+ * complain if the GPIO chip functions potentially sleep.
+ */
+void gpiod_set_array(unsigned int array_size,
+		     struct gpio_desc **desc_array, int *value_array)
+{
+	if (!desc_array)
+		return;
+	gpiod_set_array_priv(false, false, array_size, desc_array, value_array);
+}
+EXPORT_SYMBOL_GPL(gpiod_set_array);
+
+/**
  * gpiod_cansleep() - report whether gpio value access may sleep
  * @desc: gpio to check
  *
@@ -1314,14 +1473,14 @@ int gpiod_to_irq(const struct gpio_desc *desc)
 EXPORT_SYMBOL_GPL(gpiod_to_irq);
 
 /**
- * gpio_lock_as_irq() - lock a GPIO to be used as IRQ
+ * gpiochip_lock_as_irq() - lock a GPIO to be used as IRQ
  * @chip: the chip the GPIO to lock belongs to
  * @offset: the offset of the GPIO to lock as IRQ
  *
  * This is used directly by GPIO drivers that want to lock down
  * a certain GPIO line to be used for IRQs.
  */
-int gpio_lock_as_irq(struct gpio_chip *chip, unsigned int offset)
+int gpiochip_lock_as_irq(struct gpio_chip *chip, unsigned int offset)
 {
 	if (offset >= chip->ngpio)
 		return -EINVAL;
@@ -1336,24 +1495,24 @@ int gpio_lock_as_irq(struct gpio_chip *chip, unsigned int offset)
 	set_bit(FLAG_USED_AS_IRQ, &chip->desc[offset].flags);
 	return 0;
 }
-EXPORT_SYMBOL_GPL(gpio_lock_as_irq);
+EXPORT_SYMBOL_GPL(gpiochip_lock_as_irq);
 
 /**
- * gpio_unlock_as_irq() - unlock a GPIO used as IRQ
+ * gpiochip_unlock_as_irq() - unlock a GPIO used as IRQ
  * @chip: the chip the GPIO to lock belongs to
  * @offset: the offset of the GPIO to lock as IRQ
  *
  * This is used directly by GPIO drivers that want to indicate
  * that a certain GPIO is no longer used exclusively for IRQ.
  */
-void gpio_unlock_as_irq(struct gpio_chip *chip, unsigned int offset)
+void gpiochip_unlock_as_irq(struct gpio_chip *chip, unsigned int offset)
 {
 	if (offset >= chip->ngpio)
 		return;
 
 	clear_bit(FLAG_USED_AS_IRQ, &chip->desc[offset].flags);
 }
-EXPORT_SYMBOL_GPL(gpio_unlock_as_irq);
+EXPORT_SYMBOL_GPL(gpiochip_unlock_as_irq);
 
 /**
  * gpiod_get_raw_value_cansleep() - return a gpio's raw value
@@ -1440,6 +1599,50 @@ void gpiod_set_value_cansleep(struct gpio_desc *desc, int value)
 EXPORT_SYMBOL_GPL(gpiod_set_value_cansleep);
 
 /**
+ * gpiod_set_raw_array_cansleep() - assign values to an array of GPIOs
+ * @array_size: number of elements in the descriptor / value arrays
+ * @desc_array: array of GPIO descriptors whose values will be assigned
+ * @value_array: array of values to assign
+ *
+ * Set the raw values of the GPIOs, i.e. the values of the physical lines
+ * without regard for their ACTIVE_LOW status.
+ *
+ * This function is to be called from contexts that can sleep.
+ */
+void gpiod_set_raw_array_cansleep(unsigned int array_size,
+				  struct gpio_desc **desc_array,
+				  int *value_array)
+{
+	might_sleep_if(extra_checks);
+	if (!desc_array)
+		return;
+	gpiod_set_array_priv(true, true, array_size, desc_array, value_array);
+}
+EXPORT_SYMBOL_GPL(gpiod_set_raw_array_cansleep);
+
+/**
+ * gpiod_set_array_cansleep() - assign values to an array of GPIOs
+ * @array_size: number of elements in the descriptor / value arrays
+ * @desc_array: array of GPIO descriptors whose values will be assigned
+ * @value_array: array of values to assign
+ *
+ * Set the logical values of the GPIOs, i.e. taking their ACTIVE_LOW status
+ * into account.
+ *
+ * This function is to be called from contexts that can sleep.
+ */
+void gpiod_set_array_cansleep(unsigned int array_size,
+			      struct gpio_desc **desc_array,
+			      int *value_array)
+{
+	might_sleep_if(extra_checks);
+	if (!desc_array)
+		return;
+	gpiod_set_array_priv(false, true, array_size, desc_array, value_array);
+}
+EXPORT_SYMBOL_GPL(gpiod_set_array_cansleep);
+
+/**
  * gpiod_add_lookup_table() - register GPIO device consumers
  * @table: table of consumers to register
  */
@@ -1487,14 +1690,36 @@ static struct gpio_desc *acpi_find_gpio(struct device *dev, const char *con_id,
 					unsigned int idx,
 					enum gpio_lookup_flags *flags)
 {
+	static const char * const suffixes[] = { "gpios", "gpio" };
+	struct acpi_device *adev = ACPI_COMPANION(dev);
 	struct acpi_gpio_info info;
 	struct gpio_desc *desc;
+	char propname[32];
+	int i;
 
-	desc = acpi_get_gpiod_by_index(dev, idx, &info);
-	if (IS_ERR(desc))
-		return desc;
+	/* Try first from _DSD */
+	for (i = 0; i < ARRAY_SIZE(suffixes); i++) {
+		if (con_id && strcmp(con_id, "gpios")) {
+			snprintf(propname, sizeof(propname), "%s-%s",
+				 con_id, suffixes[i]);
+		} else {
+			snprintf(propname, sizeof(propname), "%s",
+				 suffixes[i]);
+		}
 
-	if (info.gpioint && info.active_low)
+		desc = acpi_get_gpiod_by_index(adev, propname, idx, &info);
+		if (!IS_ERR(desc) || (PTR_ERR(desc) == -EPROBE_DEFER))
+			break;
+	}
+
+	/* Then from plain _CRS GPIOs */
+	if (IS_ERR(desc)) {
+		desc = acpi_get_gpiod_by_index(adev, NULL, idx, &info);
+		if (IS_ERR(desc))
+			return desc;
+	}
+
+	if (info.active_low)
 		*flags |= GPIO_ACTIVE_LOW;
 
 	return desc;
@@ -1652,7 +1877,7 @@ struct gpio_desc *__must_check __gpiod_get_index(struct device *dev,
 	 * a result. In that case, use platform lookup as a fallback.
 	 */
 	if (!desc || desc == ERR_PTR(-ENOENT)) {
-		dev_dbg(dev, "using lookup tables for GPIO lookup");
+		dev_dbg(dev, "using lookup tables for GPIO lookup\n");
 		desc = gpiod_find(dev, con_id, idx, &lookupflags);
 	}
 
@@ -1695,6 +1920,61 @@ struct gpio_desc *__must_check __gpiod_get_index(struct device *dev,
 EXPORT_SYMBOL_GPL(__gpiod_get_index);
 
 /**
+ * fwnode_get_named_gpiod - obtain a GPIO from firmware node
+ * @fwnode:	handle of the firmware node
+ * @propname:	name of the firmware property representing the GPIO
+ *
+ * This function can be used for drivers that get their configuration
+ * from firmware.
+ *
+ * Function properly finds the corresponding GPIO using whatever is the
+ * underlying firmware interface and then makes sure that the GPIO
+ * descriptor is requested before it is returned to the caller.
+ *
+ * In case of error an ERR_PTR() is returned.
+ */
+struct gpio_desc *fwnode_get_named_gpiod(struct fwnode_handle *fwnode,
+					 const char *propname)
+{
+	struct gpio_desc *desc = ERR_PTR(-ENODEV);
+	bool active_low = false;
+	int ret;
+
+	if (!fwnode)
+		return ERR_PTR(-EINVAL);
+
+	if (is_of_node(fwnode)) {
+		enum of_gpio_flags flags;
+
+		desc = of_get_named_gpiod_flags(of_node(fwnode), propname, 0,
+						&flags);
+		if (!IS_ERR(desc))
+			active_low = flags & OF_GPIO_ACTIVE_LOW;
+	} else if (is_acpi_node(fwnode)) {
+		struct acpi_gpio_info info;
+
+		desc = acpi_get_gpiod_by_index(acpi_node(fwnode), propname, 0,
+					       &info);
+		if (!IS_ERR(desc))
+			active_low = info.active_low;
+	}
+
+	if (IS_ERR(desc))
+		return desc;
+
+	ret = gpiod_request(desc, NULL);
+	if (ret)
+		return ERR_PTR(ret);
+
+	/* Only value flag can be set from both DT and ACPI is active_low */
+	if (active_low)
+		set_bit(FLAG_ACTIVE_LOW, &desc->flags);
+
+	return desc;
+}
+EXPORT_SYMBOL_GPL(fwnode_get_named_gpiod);
+
+/**
  * gpiod_get_index_optional - obtain an optional GPIO from a multi-index GPIO
  *                            function
  * @dev: GPIO consumer, can be NULL for system-global GPIOs
diff --git a/drivers/gpio/gpiolib.h b/drivers/gpio/gpiolib.h
index 9db2b6a71c5d..550a5eafbd38 100644
--- a/drivers/gpio/gpiolib.h
+++ b/drivers/gpio/gpiolib.h
@@ -34,7 +34,8 @@ void acpi_gpiochip_remove(struct gpio_chip *chip);
 void acpi_gpiochip_request_interrupts(struct gpio_chip *chip);
 void acpi_gpiochip_free_interrupts(struct gpio_chip *chip);
 
-struct gpio_desc *acpi_get_gpiod_by_index(struct device *dev, int index,
+struct gpio_desc *acpi_get_gpiod_by_index(struct acpi_device *adev,
+					  const char *propname, int index,
 					  struct acpi_gpio_info *info);
 #else
 static inline void acpi_gpiochip_add(struct gpio_chip *chip) { }
@@ -47,8 +48,8 @@ static inline void
 acpi_gpiochip_free_interrupts(struct gpio_chip *chip) { }
 
 static inline struct gpio_desc *
-acpi_get_gpiod_by_index(struct device *dev, int index,
-			struct acpi_gpio_info *info)
+acpi_get_gpiod_by_index(struct acpi_device *adev, const char *propname,
+			int index, struct acpi_gpio_info *info)
 {
 	return ERR_PTR(-ENOSYS);
 }
@@ -76,6 +77,7 @@ struct gpio_desc {
 #define FLAG_OPEN_DRAIN	7	/* Gpio is open drain type */
 #define FLAG_OPEN_SOURCE 8	/* Gpio is open source type */
 #define FLAG_USED_AS_IRQ 9	/* GPIO is connected to an IRQ */
+#define FLAG_SYSFS_DIR	10	/* show sysfs direction attribute */
 
 #define ID_SHIFT	16	/* add new flags before this one */
 
diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
index b066bb3ca01a..c3413b6adb17 100644
--- a/drivers/gpu/drm/Kconfig
+++ b/drivers/gpu/drm/Kconfig
@@ -8,6 +8,7 @@ menuconfig DRM
 	tristate "Direct Rendering Manager (XFree86 4.1.0 and higher DRI support)"
 	depends on (AGP || AGP=n) && !EMULATED_CMPXCHG && MMU && HAS_DMA
 	select HDMI
+	select FB_CMDLINE
 	select I2C
 	select I2C_ALGOBIT
 	select DMA_SHARED_BUFFER
@@ -24,12 +25,6 @@ config DRM_MIPI_DSI
 	bool
 	depends on DRM
 
-config DRM_USB
-	tristate
-	depends on DRM
-	depends on USB_SUPPORT && USB_ARCH_HAS_HCD
-	select USB
-
 config DRM_KMS_HELPER
 	tristate
 	depends on DRM
@@ -115,6 +110,7 @@ config DRM_RADEON
 	select HWMON
 	select BACKLIGHT_CLASS_DEVICE
 	select INTERVAL_TREE
+	select MMU_NOTIFIER
 	help
 	  Choose this option if you have an ATI Radeon graphics card.  There
 	  are both PCI and AGP versions.  You don't need to choose this to
@@ -171,6 +167,8 @@ config DRM_SAVAGE
 
 source "drivers/gpu/drm/exynos/Kconfig"
 
+source "drivers/gpu/drm/rockchip/Kconfig"
+
 source "drivers/gpu/drm/vmwgfx/Kconfig"
 
 source "drivers/gpu/drm/gma500/Kconfig"
@@ -204,3 +202,7 @@ source "drivers/gpu/drm/tegra/Kconfig"
 source "drivers/gpu/drm/panel/Kconfig"
 
 source "drivers/gpu/drm/sti/Kconfig"
+
+source "drivers/gpu/drm/amd/amdkfd/Kconfig"
+
+source "drivers/gpu/drm/imx/Kconfig"
diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile
index 4a55d59ccd22..e620807418ea 100644
--- a/drivers/gpu/drm/Makefile
+++ b/drivers/gpu/drm/Makefile
@@ -4,7 +4,7 @@
 
 ccflags-y := -Iinclude/drm
 
-drm-y       :=	drm_auth.o drm_buffer.o drm_bufs.o drm_cache.o \
+drm-y       :=	drm_auth.o drm_bufs.o drm_cache.o \
 		drm_context.o drm_dma.o \
 		drm_fops.o drm_gem.o drm_ioctl.o drm_irq.o \
 		drm_lock.o drm_memory.o drm_drv.o drm_vm.o \
@@ -14,7 +14,7 @@ drm-y       :=	drm_auth.o drm_buffer.o drm_bufs.o drm_cache.o \
 		drm_info.o drm_debugfs.o drm_encoder_slave.o \
 		drm_trace_points.o drm_global.o drm_prime.o \
 		drm_rect.o drm_vma_manager.o drm_flip_work.o \
-		drm_modeset_lock.o
+		drm_modeset_lock.o drm_atomic.o
 
 drm-$(CONFIG_COMPAT) += drm_ioc32.o
 drm-$(CONFIG_DRM_GEM_CMA_HELPER) += drm_gem_cma_helper.o
@@ -22,10 +22,8 @@ drm-$(CONFIG_PCI) += ati_pcigart.o
 drm-$(CONFIG_DRM_PANEL) += drm_panel.o
 drm-$(CONFIG_OF) += drm_of.o
 
-drm-usb-y   := drm_usb.o
-
 drm_kms_helper-y := drm_crtc_helper.o drm_dp_helper.o drm_probe_helper.o \
-		drm_plane_helper.o drm_dp_mst_topology.o
+		drm_plane_helper.o drm_dp_mst_topology.o drm_atomic_helper.o
 drm_kms_helper-$(CONFIG_DRM_LOAD_EDID_FIRMWARE) += drm_edid_load.o
 drm_kms_helper-$(CONFIG_DRM_KMS_FB_HELPER) += drm_fb_helper.o
 drm_kms_helper-$(CONFIG_DRM_KMS_CMA_HELPER) += drm_fb_cma_helper.o
@@ -36,10 +34,10 @@ CFLAGS_drm_trace_points.o := -I$(src)
 
 obj-$(CONFIG_DRM)	+= drm.o
 obj-$(CONFIG_DRM_MIPI_DSI) += drm_mipi_dsi.o
-obj-$(CONFIG_DRM_USB)   += drm_usb.o
 obj-$(CONFIG_DRM_TTM)	+= ttm/
 obj-$(CONFIG_DRM_TDFX)	+= tdfx/
 obj-$(CONFIG_DRM_R128)	+= r128/
+obj-$(CONFIG_HSA_AMD) += amd/amdkfd/
 obj-$(CONFIG_DRM_RADEON)+= radeon/
 obj-$(CONFIG_DRM_MGA)	+= mga/
 obj-$(CONFIG_DRM_I810)	+= i810/
@@ -52,6 +50,7 @@ obj-$(CONFIG_DRM_VMWGFX)+= vmwgfx/
 obj-$(CONFIG_DRM_VIA)	+=via/
 obj-$(CONFIG_DRM_NOUVEAU) +=nouveau/
 obj-$(CONFIG_DRM_EXYNOS) +=exynos/
+obj-$(CONFIG_DRM_ROCKCHIP) +=rockchip/
 obj-$(CONFIG_DRM_GMA500) += gma500/
 obj-$(CONFIG_DRM_UDL) += udl/
 obj-$(CONFIG_DRM_AST) += ast/
@@ -65,6 +64,7 @@ obj-$(CONFIG_DRM_BOCHS) += bochs/
 obj-$(CONFIG_DRM_MSM) += msm/
 obj-$(CONFIG_DRM_TEGRA) += tegra/
 obj-$(CONFIG_DRM_STI) += sti/
+obj-$(CONFIG_DRM_IMX) += imx/
 obj-y			+= i2c/
 obj-y			+= panel/
 obj-y			+= bridge/
diff --git a/drivers/gpu/drm/README.drm b/drivers/gpu/drm/README.drm
deleted file mode 100644
index b5b332722581..000000000000
--- a/drivers/gpu/drm/README.drm
+++ /dev/null
@@ -1,43 +0,0 @@
-************************************************************
-* For the very latest on DRI development, please see:      *
-*     http://dri.freedesktop.org/                          *
-************************************************************
-
-The Direct Rendering Manager (drm) is a device-independent kernel-level
-device driver that provides support for the XFree86 Direct Rendering
-Infrastructure (DRI).
-
-The DRM supports the Direct Rendering Infrastructure (DRI) in four major
-ways:
-
-    1. The DRM provides synchronized access to the graphics hardware via
-       the use of an optimized two-tiered lock.
-
-    2. The DRM enforces the DRI security policy for access to the graphics
-       hardware by only allowing authenticated X11 clients access to
-       restricted regions of memory.
-
-    3. The DRM provides a generic DMA engine, complete with multiple
-       queues and the ability to detect the need for an OpenGL context
-       switch.
-
-    4. The DRM is extensible via the use of small device-specific modules
-       that rely extensively on the API exported by the DRM module.
-
-
-Documentation on the DRI is available from:
-    http://dri.freedesktop.org/wiki/Documentation
-    http://sourceforge.net/project/showfiles.php?group_id=387
-    http://dri.sourceforge.net/doc/
-
-For specific information about kernel-level support, see:
-
-    The Direct Rendering Manager, Kernel Support for the Direct Rendering
-    Infrastructure
-    http://dri.sourceforge.net/doc/drm_low_level.html
-
-    Hardware Locking for the Direct Rendering Infrastructure
-    http://dri.sourceforge.net/doc/hardware_locking_low_level.html
-
-    A Security Analysis of the Direct Rendering Infrastructure
-    http://dri.sourceforge.net/doc/security_low_level.html
diff --git a/drivers/gpu/drm/amd/amdkfd/Kconfig b/drivers/gpu/drm/amd/amdkfd/Kconfig
new file mode 100644
index 000000000000..8dfac37ff327
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/Kconfig
@@ -0,0 +1,9 @@
+#
+# Heterogenous system architecture configuration
+#
+
+config HSA_AMD
+	tristate "HSA kernel driver for AMD GPU devices"
+	depends on DRM_RADEON && AMD_IOMMU_V2 && X86_64
+	help
+	  Enable this if you want to use HSA features on AMD GPU devices.
diff --git a/drivers/gpu/drm/amd/amdkfd/Makefile b/drivers/gpu/drm/amd/amdkfd/Makefile
new file mode 100644
index 000000000000..307a309110e6
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/Makefile
@@ -0,0 +1,13 @@
+#
+# Makefile for Heterogenous System Architecture support for AMD GPU devices
+#
+
+ccflags-y := -Iinclude/drm -Idrivers/gpu/drm/amd/include/
+
+amdkfd-y	:= kfd_module.o kfd_device.o kfd_chardev.o kfd_topology.o \
+		kfd_pasid.o kfd_doorbell.o kfd_flat_memory.o \
+		kfd_process.o kfd_queue.o kfd_mqd_manager.o \
+		kfd_kernel_queue.o kfd_packet_manager.o \
+		kfd_process_queue_manager.o kfd_device_queue_manager.o
+
+obj-$(CONFIG_HSA_AMD)	+= amdkfd.o
diff --git a/drivers/gpu/drm/amd/amdkfd/cik_regs.h b/drivers/gpu/drm/amd/amdkfd/cik_regs.h
new file mode 100644
index 000000000000..607fc5ceadbe
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/cik_regs.h
@@ -0,0 +1,221 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#ifndef CIK_REGS_H
+#define CIK_REGS_H
+
+#define IH_VMID_0_LUT					0x3D40u
+
+#define BIF_DOORBELL_CNTL				0x530Cu
+
+#define	SRBM_GFX_CNTL					0xE44
+#define	PIPEID(x)					((x) << 0)
+#define	MEID(x)						((x) << 2)
+#define	VMID(x)						((x) << 4)
+#define	QUEUEID(x)					((x) << 8)
+
+#define	SQ_CONFIG					0x8C00
+
+#define	SH_MEM_BASES					0x8C28
+/* if PTR32, these are the bases for scratch and lds */
+#define	PRIVATE_BASE(x)					((x) << 0) /* scratch */
+#define	SHARED_BASE(x)					((x) << 16) /* LDS */
+#define	SH_MEM_APE1_BASE				0x8C2C
+/* if PTR32, this is the base location of GPUVM */
+#define	SH_MEM_APE1_LIMIT				0x8C30
+/* if PTR32, this is the upper limit of GPUVM */
+#define	SH_MEM_CONFIG					0x8C34
+#define	PTR32						(1 << 0)
+#define PRIVATE_ATC					(1 << 1)
+#define	ALIGNMENT_MODE(x)				((x) << 2)
+#define	SH_MEM_ALIGNMENT_MODE_DWORD			0
+#define	SH_MEM_ALIGNMENT_MODE_DWORD_STRICT		1
+#define	SH_MEM_ALIGNMENT_MODE_STRICT			2
+#define	SH_MEM_ALIGNMENT_MODE_UNALIGNED			3
+#define	DEFAULT_MTYPE(x)				((x) << 4)
+#define	APE1_MTYPE(x)					((x) << 7)
+
+/* valid for both DEFAULT_MTYPE and APE1_MTYPE */
+#define	MTYPE_CACHED					0
+#define	MTYPE_NONCACHED					3
+
+
+#define SH_STATIC_MEM_CONFIG				0x9604u
+
+#define	TC_CFG_L1_LOAD_POLICY0				0xAC68
+#define	TC_CFG_L1_LOAD_POLICY1				0xAC6C
+#define	TC_CFG_L1_STORE_POLICY				0xAC70
+#define	TC_CFG_L2_LOAD_POLICY0				0xAC74
+#define	TC_CFG_L2_LOAD_POLICY1				0xAC78
+#define	TC_CFG_L2_STORE_POLICY0				0xAC7C
+#define	TC_CFG_L2_STORE_POLICY1				0xAC80
+#define	TC_CFG_L2_ATOMIC_POLICY				0xAC84
+#define	TC_CFG_L1_VOLATILE				0xAC88
+#define	TC_CFG_L2_VOLATILE				0xAC8C
+
+#define CP_PQ_WPTR_POLL_CNTL				0xC20C
+#define	WPTR_POLL_EN					(1 << 31)
+
+#define CPC_INT_CNTL					0xC2D0
+#define CP_ME1_PIPE0_INT_CNTL				0xC214
+#define CP_ME1_PIPE1_INT_CNTL				0xC218
+#define CP_ME1_PIPE2_INT_CNTL				0xC21C
+#define CP_ME1_PIPE3_INT_CNTL				0xC220
+#define CP_ME2_PIPE0_INT_CNTL				0xC224
+#define CP_ME2_PIPE1_INT_CNTL				0xC228
+#define CP_ME2_PIPE2_INT_CNTL				0xC22C
+#define CP_ME2_PIPE3_INT_CNTL				0xC230
+#define DEQUEUE_REQUEST_INT_ENABLE			(1 << 13)
+#define WRM_POLL_TIMEOUT_INT_ENABLE			(1 << 17)
+#define PRIV_REG_INT_ENABLE				(1 << 23)
+#define TIME_STAMP_INT_ENABLE				(1 << 26)
+#define GENERIC2_INT_ENABLE				(1 << 29)
+#define GENERIC1_INT_ENABLE				(1 << 30)
+#define GENERIC0_INT_ENABLE				(1 << 31)
+#define CP_ME1_PIPE0_INT_STATUS				0xC214
+#define CP_ME1_PIPE1_INT_STATUS				0xC218
+#define CP_ME1_PIPE2_INT_STATUS				0xC21C
+#define CP_ME1_PIPE3_INT_STATUS				0xC220
+#define CP_ME2_PIPE0_INT_STATUS				0xC224
+#define CP_ME2_PIPE1_INT_STATUS				0xC228
+#define CP_ME2_PIPE2_INT_STATUS				0xC22C
+#define CP_ME2_PIPE3_INT_STATUS				0xC230
+#define DEQUEUE_REQUEST_INT_STATUS			(1 << 13)
+#define WRM_POLL_TIMEOUT_INT_STATUS			(1 << 17)
+#define PRIV_REG_INT_STATUS				(1 << 23)
+#define TIME_STAMP_INT_STATUS				(1 << 26)
+#define GENERIC2_INT_STATUS				(1 << 29)
+#define GENERIC1_INT_STATUS				(1 << 30)
+#define GENERIC0_INT_STATUS				(1 << 31)
+
+#define CP_HPD_EOP_BASE_ADDR				0xC904
+#define CP_HPD_EOP_BASE_ADDR_HI				0xC908
+#define CP_HPD_EOP_VMID					0xC90C
+#define CP_HPD_EOP_CONTROL				0xC910
+#define	EOP_SIZE(x)					((x) << 0)
+#define	EOP_SIZE_MASK					(0x3f << 0)
+#define CP_MQD_BASE_ADDR				0xC914
+#define CP_MQD_BASE_ADDR_HI				0xC918
+#define CP_HQD_ACTIVE					0xC91C
+#define CP_HQD_VMID					0xC920
+
+#define CP_HQD_PERSISTENT_STATE				0xC924u
+#define	DEFAULT_CP_HQD_PERSISTENT_STATE			(0x33U << 8)
+#define	PRELOAD_REQ					(1 << 0)
+
+#define CP_HQD_PIPE_PRIORITY				0xC928u
+#define CP_HQD_QUEUE_PRIORITY				0xC92Cu
+#define CP_HQD_QUANTUM					0xC930u
+#define	QUANTUM_EN					1U
+#define	QUANTUM_SCALE_1MS				(1U << 4)
+#define	QUANTUM_DURATION(x)				((x) << 8)
+
+#define CP_HQD_PQ_BASE					0xC934
+#define CP_HQD_PQ_BASE_HI				0xC938
+#define CP_HQD_PQ_RPTR					0xC93C
+#define CP_HQD_PQ_RPTR_REPORT_ADDR			0xC940
+#define CP_HQD_PQ_RPTR_REPORT_ADDR_HI			0xC944
+#define CP_HQD_PQ_WPTR_POLL_ADDR			0xC948
+#define CP_HQD_PQ_WPTR_POLL_ADDR_HI			0xC94C
+#define CP_HQD_PQ_DOORBELL_CONTROL			0xC950
+#define	DOORBELL_OFFSET(x)				((x) << 2)
+#define	DOORBELL_OFFSET_MASK				(0x1fffff << 2)
+#define	DOORBELL_SOURCE					(1 << 28)
+#define	DOORBELL_SCHD_HIT				(1 << 29)
+#define	DOORBELL_EN					(1 << 30)
+#define	DOORBELL_HIT					(1 << 31)
+#define CP_HQD_PQ_WPTR					0xC954
+#define CP_HQD_PQ_CONTROL				0xC958
+#define	QUEUE_SIZE(x)					((x) << 0)
+#define	QUEUE_SIZE_MASK					(0x3f << 0)
+#define	RPTR_BLOCK_SIZE(x)				((x) << 8)
+#define	RPTR_BLOCK_SIZE_MASK				(0x3f << 8)
+#define	MIN_AVAIL_SIZE(x)				((x) << 20)
+#define	PQ_ATC_EN					(1 << 23)
+#define	PQ_VOLATILE					(1 << 26)
+#define	NO_UPDATE_RPTR					(1 << 27)
+#define	UNORD_DISPATCH					(1 << 28)
+#define	ROQ_PQ_IB_FLIP					(1 << 29)
+#define	PRIV_STATE					(1 << 30)
+#define	KMD_QUEUE					(1 << 31)
+
+#define	DEFAULT_RPTR_BLOCK_SIZE				RPTR_BLOCK_SIZE(5)
+#define	DEFAULT_MIN_AVAIL_SIZE				MIN_AVAIL_SIZE(3)
+
+#define CP_HQD_IB_BASE_ADDR				0xC95Cu
+#define CP_HQD_IB_BASE_ADDR_HI				0xC960u
+#define CP_HQD_IB_RPTR					0xC964u
+#define CP_HQD_IB_CONTROL				0xC968u
+#define	IB_ATC_EN					(1U << 23)
+#define	DEFAULT_MIN_IB_AVAIL_SIZE			(3U << 20)
+
+#define CP_HQD_DEQUEUE_REQUEST				0xC974
+#define	DEQUEUE_REQUEST_DRAIN				1
+#define DEQUEUE_REQUEST_RESET				2
+#define		DEQUEUE_INT					(1U << 8)
+
+#define CP_HQD_SEMA_CMD					0xC97Cu
+#define CP_HQD_MSG_TYPE					0xC980u
+#define CP_HQD_ATOMIC0_PREOP_LO				0xC984u
+#define CP_HQD_ATOMIC0_PREOP_HI				0xC988u
+#define CP_HQD_ATOMIC1_PREOP_LO				0xC98Cu
+#define CP_HQD_ATOMIC1_PREOP_HI				0xC990u
+#define CP_HQD_HQ_SCHEDULER0				0xC994u
+#define CP_HQD_HQ_SCHEDULER1				0xC998u
+
+
+#define CP_MQD_CONTROL					0xC99C
+#define	MQD_VMID(x)					((x) << 0)
+#define	MQD_VMID_MASK					(0xf << 0)
+#define	MQD_CONTROL_PRIV_STATE_EN			(1U << 8)
+
+#define GRBM_GFX_INDEX					0x30800
+#define	INSTANCE_INDEX(x)				((x) << 0)
+#define	SH_INDEX(x)					((x) << 8)
+#define	SE_INDEX(x)					((x) << 16)
+#define	SH_BROADCAST_WRITES				(1 << 29)
+#define	INSTANCE_BROADCAST_WRITES			(1 << 30)
+#define	SE_BROADCAST_WRITES				(1 << 31)
+
+#define SQC_CACHES					0x30d20
+#define SQC_POLICY					0x8C38u
+#define SQC_VOLATILE					0x8C3Cu
+
+#define CP_PERFMON_CNTL					0x36020
+
+#define ATC_VMID0_PASID_MAPPING				0x339Cu
+#define	ATC_VMID_PASID_MAPPING_UPDATE_STATUS		0x3398u
+#define	ATC_VMID_PASID_MAPPING_VALID			(1U << 31)
+
+#define ATC_VM_APERTURE0_CNTL				0x3310u
+#define	ATS_ACCESS_MODE_NEVER				0
+#define	ATS_ACCESS_MODE_ALWAYS				1
+
+#define ATC_VM_APERTURE0_CNTL2				0x3318u
+#define ATC_VM_APERTURE0_HIGH_ADDR			0x3308u
+#define ATC_VM_APERTURE0_LOW_ADDR			0x3300u
+#define ATC_VM_APERTURE1_CNTL				0x3314u
+#define ATC_VM_APERTURE1_CNTL2				0x331Cu
+#define ATC_VM_APERTURE1_HIGH_ADDR			0x330Cu
+#define ATC_VM_APERTURE1_LOW_ADDR			0x3304u
+
+#endif
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
new file mode 100644
index 000000000000..fcfdf23e1913
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
@@ -0,0 +1,611 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#include <linux/device.h>
+#include <linux/export.h>
+#include <linux/err.h>
+#include <linux/fs.h>
+#include <linux/sched.h>
+#include <linux/slab.h>
+#include <linux/uaccess.h>
+#include <linux/compat.h>
+#include <uapi/linux/kfd_ioctl.h>
+#include <linux/time.h>
+#include <linux/mm.h>
+#include <uapi/asm-generic/mman-common.h>
+#include <asm/processor.h>
+#include "kfd_priv.h"
+#include "kfd_device_queue_manager.h"
+
+static long kfd_ioctl(struct file *, unsigned int, unsigned long);
+static int kfd_open(struct inode *, struct file *);
+static int kfd_mmap(struct file *, struct vm_area_struct *);
+
+static const char kfd_dev_name[] = "kfd";
+
+static const struct file_operations kfd_fops = {
+	.owner = THIS_MODULE,
+	.unlocked_ioctl = kfd_ioctl,
+	.compat_ioctl = kfd_ioctl,
+	.open = kfd_open,
+	.mmap = kfd_mmap,
+};
+
+static int kfd_char_dev_major = -1;
+static struct class *kfd_class;
+struct device *kfd_device;
+
+int kfd_chardev_init(void)
+{
+	int err = 0;
+
+	kfd_char_dev_major = register_chrdev(0, kfd_dev_name, &kfd_fops);
+	err = kfd_char_dev_major;
+	if (err < 0)
+		goto err_register_chrdev;
+
+	kfd_class = class_create(THIS_MODULE, kfd_dev_name);
+	err = PTR_ERR(kfd_class);
+	if (IS_ERR(kfd_class))
+		goto err_class_create;
+
+	kfd_device = device_create(kfd_class, NULL,
+					MKDEV(kfd_char_dev_major, 0),
+					NULL, kfd_dev_name);
+	err = PTR_ERR(kfd_device);
+	if (IS_ERR(kfd_device))
+		goto err_device_create;
+
+	return 0;
+
+err_device_create:
+	class_destroy(kfd_class);
+err_class_create:
+	unregister_chrdev(kfd_char_dev_major, kfd_dev_name);
+err_register_chrdev:
+	return err;
+}
+
+void kfd_chardev_exit(void)
+{
+	device_destroy(kfd_class, MKDEV(kfd_char_dev_major, 0));
+	class_destroy(kfd_class);
+	unregister_chrdev(kfd_char_dev_major, kfd_dev_name);
+}
+
+struct device *kfd_chardev(void)
+{
+	return kfd_device;
+}
+
+
+static int kfd_open(struct inode *inode, struct file *filep)
+{
+	struct kfd_process *process;
+	bool is_32bit_user_mode;
+
+	if (iminor(inode) != 0)
+		return -ENODEV;
+
+	is_32bit_user_mode = is_compat_task();
+
+	if (is_32bit_user_mode == true) {
+		dev_warn(kfd_device,
+			"Process %d (32-bit) failed to open /dev/kfd\n"
+			"32-bit processes are not supported by amdkfd\n",
+			current->pid);
+		return -EPERM;
+	}
+
+	process = kfd_create_process(current);
+	if (IS_ERR(process))
+		return PTR_ERR(process);
+
+	dev_dbg(kfd_device, "process %d opened, compat mode (32 bit) - %d\n",
+		process->pasid, process->is_32bit_user_mode);
+
+	return 0;
+}
+
+static int kfd_ioctl_get_version(struct file *filep, struct kfd_process *p,
+					void *data)
+{
+	struct kfd_ioctl_get_version_args *args = data;
+	int err = 0;
+
+	args->major_version = KFD_IOCTL_MAJOR_VERSION;
+	args->minor_version = KFD_IOCTL_MINOR_VERSION;
+
+	return err;
+}
+
+static int set_queue_properties_from_user(struct queue_properties *q_properties,
+				struct kfd_ioctl_create_queue_args *args)
+{
+	if (args->queue_percentage > KFD_MAX_QUEUE_PERCENTAGE) {
+		pr_err("kfd: queue percentage must be between 0 to KFD_MAX_QUEUE_PERCENTAGE\n");
+		return -EINVAL;
+	}
+
+	if (args->queue_priority > KFD_MAX_QUEUE_PRIORITY) {
+		pr_err("kfd: queue priority must be between 0 to KFD_MAX_QUEUE_PRIORITY\n");
+		return -EINVAL;
+	}
+
+	if ((args->ring_base_address) &&
+		(!access_ok(VERIFY_WRITE,
+			(const void __user *) args->ring_base_address,
+			sizeof(uint64_t)))) {
+		pr_err("kfd: can't access ring base address\n");
+		return -EFAULT;
+	}
+
+	if (!is_power_of_2(args->ring_size) && (args->ring_size != 0)) {
+		pr_err("kfd: ring size must be a power of 2 or 0\n");
+		return -EINVAL;
+	}
+
+	if (!access_ok(VERIFY_WRITE,
+			(const void __user *) args->read_pointer_address,
+			sizeof(uint32_t))) {
+		pr_err("kfd: can't access read pointer\n");
+		return -EFAULT;
+	}
+
+	if (!access_ok(VERIFY_WRITE,
+			(const void __user *) args->write_pointer_address,
+			sizeof(uint32_t))) {
+		pr_err("kfd: can't access write pointer\n");
+		return -EFAULT;
+	}
+
+	q_properties->is_interop = false;
+	q_properties->queue_percent = args->queue_percentage;
+	q_properties->priority = args->queue_priority;
+	q_properties->queue_address = args->ring_base_address;
+	q_properties->queue_size = args->ring_size;
+	q_properties->read_ptr = (uint32_t *) args->read_pointer_address;
+	q_properties->write_ptr = (uint32_t *) args->write_pointer_address;
+	if (args->queue_type == KFD_IOC_QUEUE_TYPE_COMPUTE ||
+		args->queue_type == KFD_IOC_QUEUE_TYPE_COMPUTE_AQL)
+		q_properties->type = KFD_QUEUE_TYPE_COMPUTE;
+	else
+		return -ENOTSUPP;
+
+	if (args->queue_type == KFD_IOC_QUEUE_TYPE_COMPUTE_AQL)
+		q_properties->format = KFD_QUEUE_FORMAT_AQL;
+	else
+		q_properties->format = KFD_QUEUE_FORMAT_PM4;
+
+	pr_debug("Queue Percentage (%d, %d)\n",
+			q_properties->queue_percent, args->queue_percentage);
+
+	pr_debug("Queue Priority (%d, %d)\n",
+			q_properties->priority, args->queue_priority);
+
+	pr_debug("Queue Address (0x%llX, 0x%llX)\n",
+			q_properties->queue_address, args->ring_base_address);
+
+	pr_debug("Queue Size (0x%llX, %u)\n",
+			q_properties->queue_size, args->ring_size);
+
+	pr_debug("Queue r/w Pointers (0x%llX, 0x%llX)\n",
+			(uint64_t) q_properties->read_ptr,
+			(uint64_t) q_properties->write_ptr);
+
+	pr_debug("Queue Format (%d)\n", q_properties->format);
+
+	return 0;
+}
+
+static int kfd_ioctl_create_queue(struct file *filep, struct kfd_process *p,
+					void *data)
+{
+	struct kfd_ioctl_create_queue_args *args = data;
+	struct kfd_dev *dev;
+	int err = 0;
+	unsigned int queue_id;
+	struct kfd_process_device *pdd;
+	struct queue_properties q_properties;
+
+	memset(&q_properties, 0, sizeof(struct queue_properties));
+
+	pr_debug("kfd: creating queue ioctl\n");
+
+	err = set_queue_properties_from_user(&q_properties, args);
+	if (err)
+		return err;
+
+	dev = kfd_device_by_id(args->gpu_id);
+	if (dev == NULL)
+		return -EINVAL;
+
+	mutex_lock(&p->mutex);
+
+	pdd = kfd_bind_process_to_device(dev, p);
+	if (IS_ERR(pdd)) {
+		err = -ESRCH;
+		goto err_bind_process;
+	}
+
+	pr_debug("kfd: creating queue for PASID %d on GPU 0x%x\n",
+			p->pasid,
+			dev->id);
+
+	err = pqm_create_queue(&p->pqm, dev, filep, &q_properties, 0,
+				KFD_QUEUE_TYPE_COMPUTE, &queue_id);
+	if (err != 0)
+		goto err_create_queue;
+
+	args->queue_id = queue_id;
+
+	/* Return gpu_id as doorbell offset for mmap usage */
+	args->doorbell_offset = args->gpu_id << PAGE_SHIFT;
+
+	mutex_unlock(&p->mutex);
+
+	pr_debug("kfd: queue id %d was created successfully\n", args->queue_id);
+
+	pr_debug("ring buffer address == 0x%016llX\n",
+			args->ring_base_address);
+
+	pr_debug("read ptr address    == 0x%016llX\n",
+			args->read_pointer_address);
+
+	pr_debug("write ptr address   == 0x%016llX\n",
+			args->write_pointer_address);
+
+	return 0;
+
+err_create_queue:
+err_bind_process:
+	mutex_unlock(&p->mutex);
+	return err;
+}
+
+static int kfd_ioctl_destroy_queue(struct file *filp, struct kfd_process *p,
+					void *data)
+{
+	int retval;
+	struct kfd_ioctl_destroy_queue_args *args = data;
+
+	pr_debug("kfd: destroying queue id %d for PASID %d\n",
+				args->queue_id,
+				p->pasid);
+
+	mutex_lock(&p->mutex);
+
+	retval = pqm_destroy_queue(&p->pqm, args->queue_id);
+
+	mutex_unlock(&p->mutex);
+	return retval;
+}
+
+static int kfd_ioctl_update_queue(struct file *filp, struct kfd_process *p,
+					void *data)
+{
+	int retval;
+	struct kfd_ioctl_update_queue_args *args = data;
+	struct queue_properties properties;
+
+	if (args->queue_percentage > KFD_MAX_QUEUE_PERCENTAGE) {
+		pr_err("kfd: queue percentage must be between 0 to KFD_MAX_QUEUE_PERCENTAGE\n");
+		return -EINVAL;
+	}
+
+	if (args->queue_priority > KFD_MAX_QUEUE_PRIORITY) {
+		pr_err("kfd: queue priority must be between 0 to KFD_MAX_QUEUE_PRIORITY\n");
+		return -EINVAL;
+	}
+
+	if ((args->ring_base_address) &&
+		(!access_ok(VERIFY_WRITE,
+			(const void __user *) args->ring_base_address,
+			sizeof(uint64_t)))) {
+		pr_err("kfd: can't access ring base address\n");
+		return -EFAULT;
+	}
+
+	if (!is_power_of_2(args->ring_size) && (args->ring_size != 0)) {
+		pr_err("kfd: ring size must be a power of 2 or 0\n");
+		return -EINVAL;
+	}
+
+	properties.queue_address = args->ring_base_address;
+	properties.queue_size = args->ring_size;
+	properties.queue_percent = args->queue_percentage;
+	properties.priority = args->queue_priority;
+
+	pr_debug("kfd: updating queue id %d for PASID %d\n",
+			args->queue_id, p->pasid);
+
+	mutex_lock(&p->mutex);
+
+	retval = pqm_update_queue(&p->pqm, args->queue_id, &properties);
+
+	mutex_unlock(&p->mutex);
+
+	return retval;
+}
+
+static int kfd_ioctl_set_memory_policy(struct file *filep,
+					struct kfd_process *p, void *data)
+{
+	struct kfd_ioctl_set_memory_policy_args *args = data;
+	struct kfd_dev *dev;
+	int err = 0;
+	struct kfd_process_device *pdd;
+	enum cache_policy default_policy, alternate_policy;
+
+	if (args->default_policy != KFD_IOC_CACHE_POLICY_COHERENT
+	    && args->default_policy != KFD_IOC_CACHE_POLICY_NONCOHERENT) {
+		return -EINVAL;
+	}
+
+	if (args->alternate_policy != KFD_IOC_CACHE_POLICY_COHERENT
+	    && args->alternate_policy != KFD_IOC_CACHE_POLICY_NONCOHERENT) {
+		return -EINVAL;
+	}
+
+	dev = kfd_device_by_id(args->gpu_id);
+	if (dev == NULL)
+		return -EINVAL;
+
+	mutex_lock(&p->mutex);
+
+	pdd = kfd_bind_process_to_device(dev, p);
+	if (IS_ERR(pdd)) {
+		err = -ESRCH;
+		goto out;
+	}
+
+	default_policy = (args->default_policy == KFD_IOC_CACHE_POLICY_COHERENT)
+			 ? cache_policy_coherent : cache_policy_noncoherent;
+
+	alternate_policy =
+		(args->alternate_policy == KFD_IOC_CACHE_POLICY_COHERENT)
+		   ? cache_policy_coherent : cache_policy_noncoherent;
+
+	if (!dev->dqm->set_cache_memory_policy(dev->dqm,
+				&pdd->qpd,
+				default_policy,
+				alternate_policy,
+				(void __user *)args->alternate_aperture_base,
+				args->alternate_aperture_size))
+		err = -EINVAL;
+
+out:
+	mutex_unlock(&p->mutex);
+
+	return err;
+}
+
+static int kfd_ioctl_get_clock_counters(struct file *filep,
+				struct kfd_process *p, void *data)
+{
+	struct kfd_ioctl_get_clock_counters_args *args = data;
+	struct kfd_dev *dev;
+	struct timespec time;
+
+	dev = kfd_device_by_id(args->gpu_id);
+	if (dev == NULL)
+		return -EINVAL;
+
+	/* Reading GPU clock counter from KGD */
+	args->gpu_clock_counter = kfd2kgd->get_gpu_clock_counter(dev->kgd);
+
+	/* No access to rdtsc. Using raw monotonic time */
+	getrawmonotonic(&time);
+	args->cpu_clock_counter = (uint64_t)timespec_to_ns(&time);
+
+	get_monotonic_boottime(&time);
+	args->system_clock_counter = (uint64_t)timespec_to_ns(&time);
+
+	/* Since the counter is in nano-seconds we use 1GHz frequency */
+	args->system_clock_freq = 1000000000;
+
+	return 0;
+}
+
+
+static int kfd_ioctl_get_process_apertures(struct file *filp,
+				struct kfd_process *p, void *data)
+{
+	struct kfd_ioctl_get_process_apertures_args *args = data;
+	struct kfd_process_device_apertures *pAperture;
+	struct kfd_process_device *pdd;
+
+	dev_dbg(kfd_device, "get apertures for PASID %d", p->pasid);
+
+	args->num_of_nodes = 0;
+
+	mutex_lock(&p->mutex);
+
+	/*if the process-device list isn't empty*/
+	if (kfd_has_process_device_data(p)) {
+		/* Run over all pdd of the process */
+		pdd = kfd_get_first_process_device_data(p);
+		do {
+			pAperture =
+				&args->process_apertures[args->num_of_nodes];
+			pAperture->gpu_id = pdd->dev->id;
+			pAperture->lds_base = pdd->lds_base;
+			pAperture->lds_limit = pdd->lds_limit;
+			pAperture->gpuvm_base = pdd->gpuvm_base;
+			pAperture->gpuvm_limit = pdd->gpuvm_limit;
+			pAperture->scratch_base = pdd->scratch_base;
+			pAperture->scratch_limit = pdd->scratch_limit;
+
+			dev_dbg(kfd_device,
+				"node id %u\n", args->num_of_nodes);
+			dev_dbg(kfd_device,
+				"gpu id %u\n", pdd->dev->id);
+			dev_dbg(kfd_device,
+				"lds_base %llX\n", pdd->lds_base);
+			dev_dbg(kfd_device,
+				"lds_limit %llX\n", pdd->lds_limit);
+			dev_dbg(kfd_device,
+				"gpuvm_base %llX\n", pdd->gpuvm_base);
+			dev_dbg(kfd_device,
+				"gpuvm_limit %llX\n", pdd->gpuvm_limit);
+			dev_dbg(kfd_device,
+				"scratch_base %llX\n", pdd->scratch_base);
+			dev_dbg(kfd_device,
+				"scratch_limit %llX\n", pdd->scratch_limit);
+
+			args->num_of_nodes++;
+		} while ((pdd = kfd_get_next_process_device_data(p, pdd)) != NULL &&
+				(args->num_of_nodes < NUM_OF_SUPPORTED_GPUS));
+	}
+
+	mutex_unlock(&p->mutex);
+
+	return 0;
+}
+
+#define AMDKFD_IOCTL_DEF(ioctl, _func, _flags) \
+	[_IOC_NR(ioctl)] = {.cmd = ioctl, .func = _func, .flags = _flags, .cmd_drv = 0, .name = #ioctl}
+
+/** Ioctl table */
+static const struct amdkfd_ioctl_desc amdkfd_ioctls[] = {
+	AMDKFD_IOCTL_DEF(AMDKFD_IOC_GET_VERSION,
+			kfd_ioctl_get_version, 0),
+
+	AMDKFD_IOCTL_DEF(AMDKFD_IOC_CREATE_QUEUE,
+			kfd_ioctl_create_queue, 0),
+
+	AMDKFD_IOCTL_DEF(AMDKFD_IOC_DESTROY_QUEUE,
+			kfd_ioctl_destroy_queue, 0),
+
+	AMDKFD_IOCTL_DEF(AMDKFD_IOC_SET_MEMORY_POLICY,
+			kfd_ioctl_set_memory_policy, 0),
+
+	AMDKFD_IOCTL_DEF(AMDKFD_IOC_GET_CLOCK_COUNTERS,
+			kfd_ioctl_get_clock_counters, 0),
+
+	AMDKFD_IOCTL_DEF(AMDKFD_IOC_GET_PROCESS_APERTURES,
+			kfd_ioctl_get_process_apertures, 0),
+
+	AMDKFD_IOCTL_DEF(AMDKFD_IOC_UPDATE_QUEUE,
+			kfd_ioctl_update_queue, 0),
+};
+
+#define AMDKFD_CORE_IOCTL_COUNT	ARRAY_SIZE(amdkfd_ioctls)
+
+static long kfd_ioctl(struct file *filep, unsigned int cmd, unsigned long arg)
+{
+	struct kfd_process *process;
+	amdkfd_ioctl_t *func;
+	const struct amdkfd_ioctl_desc *ioctl = NULL;
+	unsigned int nr = _IOC_NR(cmd);
+	char stack_kdata[128];
+	char *kdata = NULL;
+	unsigned int usize, asize;
+	int retcode = -EINVAL;
+
+	if (nr >= AMDKFD_CORE_IOCTL_COUNT)
+		goto err_i1;
+
+	if ((nr >= AMDKFD_COMMAND_START) && (nr < AMDKFD_COMMAND_END)) {
+		u32 amdkfd_size;
+
+		ioctl = &amdkfd_ioctls[nr];
+
+		amdkfd_size = _IOC_SIZE(ioctl->cmd);
+		usize = asize = _IOC_SIZE(cmd);
+		if (amdkfd_size > asize)
+			asize = amdkfd_size;
+
+		cmd = ioctl->cmd;
+	} else
+		goto err_i1;
+
+	dev_dbg(kfd_device, "ioctl cmd 0x%x (#%d), arg 0x%lx\n", cmd, nr, arg);
+
+	process = kfd_get_process(current);
+	if (IS_ERR(process)) {
+		dev_dbg(kfd_device, "no process\n");
+		goto err_i1;
+	}
+
+	/* Do not trust userspace, use our own definition */
+	func = ioctl->func;
+
+	if (unlikely(!func)) {
+		dev_dbg(kfd_device, "no function\n");
+		retcode = -EINVAL;
+		goto err_i1;
+	}
+
+	if (cmd & (IOC_IN | IOC_OUT)) {
+		if (asize <= sizeof(stack_kdata)) {
+			kdata = stack_kdata;
+		} else {
+			kdata = kmalloc(asize, GFP_KERNEL);
+			if (!kdata) {
+				retcode = -ENOMEM;
+				goto err_i1;
+			}
+		}
+		if (asize > usize)
+			memset(kdata + usize, 0, asize - usize);
+	}
+
+	if (cmd & IOC_IN) {
+		if (copy_from_user(kdata, (void __user *)arg, usize) != 0) {
+			retcode = -EFAULT;
+			goto err_i1;
+		}
+	} else if (cmd & IOC_OUT) {
+		memset(kdata, 0, usize);
+	}
+
+	retcode = func(filep, process, kdata);
+
+	if (cmd & IOC_OUT)
+		if (copy_to_user((void __user *)arg, kdata, usize) != 0)
+			retcode = -EFAULT;
+
+err_i1:
+	if (!ioctl)
+		dev_dbg(kfd_device, "invalid ioctl: pid=%d, cmd=0x%02x, nr=0x%02x\n",
+			  task_pid_nr(current), cmd, nr);
+
+	if (kdata != stack_kdata)
+		kfree(kdata);
+
+	if (retcode)
+		dev_dbg(kfd_device, "ret = %d\n", retcode);
+
+	return retcode;
+}
+
+static int kfd_mmap(struct file *filp, struct vm_area_struct *vma)
+{
+	struct kfd_process *process;
+
+	process = kfd_get_process(current);
+	if (IS_ERR(process))
+		return PTR_ERR(process);
+
+	return kfd_doorbell_mmap(process, vma);
+}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_crat.h b/drivers/gpu/drm/amd/amdkfd/kfd_crat.h
new file mode 100644
index 000000000000..a374fa3d3ee6
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_crat.h
@@ -0,0 +1,294 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#ifndef KFD_CRAT_H_INCLUDED
+#define KFD_CRAT_H_INCLUDED
+
+#include <linux/types.h>
+
+#pragma pack(1)
+
+/*
+ * 4CC signature values for the CRAT and CDIT ACPI tables
+ */
+
+#define CRAT_SIGNATURE	"CRAT"
+#define CDIT_SIGNATURE	"CDIT"
+
+/*
+ * Component Resource Association Table (CRAT)
+ */
+
+#define CRAT_OEMID_LENGTH	6
+#define CRAT_OEMTABLEID_LENGTH	8
+#define CRAT_RESERVED_LENGTH	6
+
+#define CRAT_OEMID_64BIT_MASK ((1ULL << (CRAT_OEMID_LENGTH * 8)) - 1)
+
+struct crat_header {
+	uint32_t	signature;
+	uint32_t	length;
+	uint8_t		revision;
+	uint8_t		checksum;
+	uint8_t		oem_id[CRAT_OEMID_LENGTH];
+	uint8_t		oem_table_id[CRAT_OEMTABLEID_LENGTH];
+	uint32_t	oem_revision;
+	uint32_t	creator_id;
+	uint32_t	creator_revision;
+	uint32_t	total_entries;
+	uint16_t	num_domains;
+	uint8_t		reserved[CRAT_RESERVED_LENGTH];
+};
+
+/*
+ * The header structure is immediately followed by total_entries of the
+ * data definitions
+ */
+
+/*
+ * The currently defined subtype entries in the CRAT
+ */
+#define CRAT_SUBTYPE_COMPUTEUNIT_AFFINITY	0
+#define CRAT_SUBTYPE_MEMORY_AFFINITY		1
+#define CRAT_SUBTYPE_CACHE_AFFINITY		2
+#define CRAT_SUBTYPE_TLB_AFFINITY		3
+#define CRAT_SUBTYPE_CCOMPUTE_AFFINITY		4
+#define CRAT_SUBTYPE_IOLINK_AFFINITY		5
+#define CRAT_SUBTYPE_MAX			6
+
+#define CRAT_SIBLINGMAP_SIZE	32
+
+/*
+ * ComputeUnit Affinity structure and definitions
+ */
+#define CRAT_CU_FLAGS_ENABLED		0x00000001
+#define CRAT_CU_FLAGS_HOT_PLUGGABLE	0x00000002
+#define CRAT_CU_FLAGS_CPU_PRESENT	0x00000004
+#define CRAT_CU_FLAGS_GPU_PRESENT	0x00000008
+#define CRAT_CU_FLAGS_IOMMU_PRESENT	0x00000010
+#define CRAT_CU_FLAGS_RESERVED		0xffffffe0
+
+#define CRAT_COMPUTEUNIT_RESERVED_LENGTH 4
+
+struct crat_subtype_computeunit {
+	uint8_t		type;
+	uint8_t		length;
+	uint16_t	reserved;
+	uint32_t	flags;
+	uint32_t	proximity_domain;
+	uint32_t	processor_id_low;
+	uint16_t	num_cpu_cores;
+	uint16_t	num_simd_cores;
+	uint16_t	max_waves_simd;
+	uint16_t	io_count;
+	uint16_t	hsa_capability;
+	uint16_t	lds_size_in_kb;
+	uint8_t		wave_front_size;
+	uint8_t		num_banks;
+	uint16_t	micro_engine_id;
+	uint8_t		num_arrays;
+	uint8_t		num_cu_per_array;
+	uint8_t		num_simd_per_cu;
+	uint8_t		max_slots_scatch_cu;
+	uint8_t		reserved2[CRAT_COMPUTEUNIT_RESERVED_LENGTH];
+};
+
+/*
+ * HSA Memory Affinity structure and definitions
+ */
+#define CRAT_MEM_FLAGS_ENABLED		0x00000001
+#define CRAT_MEM_FLAGS_HOT_PLUGGABLE	0x00000002
+#define CRAT_MEM_FLAGS_NON_VOLATILE	0x00000004
+#define CRAT_MEM_FLAGS_RESERVED		0xfffffff8
+
+#define CRAT_MEMORY_RESERVED_LENGTH 8
+
+struct crat_subtype_memory {
+	uint8_t		type;
+	uint8_t		length;
+	uint16_t	reserved;
+	uint32_t	flags;
+	uint32_t	promixity_domain;
+	uint32_t	base_addr_low;
+	uint32_t	base_addr_high;
+	uint32_t	length_low;
+	uint32_t	length_high;
+	uint32_t	width;
+	uint8_t		reserved2[CRAT_MEMORY_RESERVED_LENGTH];
+};
+
+/*
+ * HSA Cache Affinity structure and definitions
+ */
+#define CRAT_CACHE_FLAGS_ENABLED	0x00000001
+#define CRAT_CACHE_FLAGS_DATA_CACHE	0x00000002
+#define CRAT_CACHE_FLAGS_INST_CACHE	0x00000004
+#define CRAT_CACHE_FLAGS_CPU_CACHE	0x00000008
+#define CRAT_CACHE_FLAGS_SIMD_CACHE	0x00000010
+#define CRAT_CACHE_FLAGS_RESERVED	0xffffffe0
+
+#define CRAT_CACHE_RESERVED_LENGTH 8
+
+struct crat_subtype_cache {
+	uint8_t		type;
+	uint8_t		length;
+	uint16_t	reserved;
+	uint32_t	flags;
+	uint32_t	processor_id_low;
+	uint8_t		sibling_map[CRAT_SIBLINGMAP_SIZE];
+	uint32_t	cache_size;
+	uint8_t		cache_level;
+	uint8_t		lines_per_tag;
+	uint16_t	cache_line_size;
+	uint8_t		associativity;
+	uint8_t		cache_properties;
+	uint16_t	cache_latency;
+	uint8_t		reserved2[CRAT_CACHE_RESERVED_LENGTH];
+};
+
+/*
+ * HSA TLB Affinity structure and definitions
+ */
+#define CRAT_TLB_FLAGS_ENABLED	0x00000001
+#define CRAT_TLB_FLAGS_DATA_TLB	0x00000002
+#define CRAT_TLB_FLAGS_INST_TLB	0x00000004
+#define CRAT_TLB_FLAGS_CPU_TLB	0x00000008
+#define CRAT_TLB_FLAGS_SIMD_TLB	0x00000010
+#define CRAT_TLB_FLAGS_RESERVED	0xffffffe0
+
+#define CRAT_TLB_RESERVED_LENGTH 4
+
+struct crat_subtype_tlb {
+	uint8_t		type;
+	uint8_t		length;
+	uint16_t	reserved;
+	uint32_t	flags;
+	uint32_t	processor_id_low;
+	uint8_t		sibling_map[CRAT_SIBLINGMAP_SIZE];
+	uint32_t	tlb_level;
+	uint8_t		data_tlb_associativity_2mb;
+	uint8_t		data_tlb_size_2mb;
+	uint8_t		instruction_tlb_associativity_2mb;
+	uint8_t		instruction_tlb_size_2mb;
+	uint8_t		data_tlb_associativity_4k;
+	uint8_t		data_tlb_size_4k;
+	uint8_t		instruction_tlb_associativity_4k;
+	uint8_t		instruction_tlb_size_4k;
+	uint8_t		data_tlb_associativity_1gb;
+	uint8_t		data_tlb_size_1gb;
+	uint8_t		instruction_tlb_associativity_1gb;
+	uint8_t		instruction_tlb_size_1gb;
+	uint8_t		reserved2[CRAT_TLB_RESERVED_LENGTH];
+};
+
+/*
+ * HSA CCompute/APU Affinity structure and definitions
+ */
+#define CRAT_CCOMPUTE_FLAGS_ENABLED	0x00000001
+#define CRAT_CCOMPUTE_FLAGS_RESERVED	0xfffffffe
+
+#define CRAT_CCOMPUTE_RESERVED_LENGTH 16
+
+struct crat_subtype_ccompute {
+	uint8_t		type;
+	uint8_t		length;
+	uint16_t	reserved;
+	uint32_t	flags;
+	uint32_t	processor_id_low;
+	uint8_t		sibling_map[CRAT_SIBLINGMAP_SIZE];
+	uint32_t	apu_size;
+	uint8_t		reserved2[CRAT_CCOMPUTE_RESERVED_LENGTH];
+};
+
+/*
+ * HSA IO Link Affinity structure and definitions
+ */
+#define CRAT_IOLINK_FLAGS_ENABLED	0x00000001
+#define CRAT_IOLINK_FLAGS_COHERENCY	0x00000002
+#define CRAT_IOLINK_FLAGS_RESERVED	0xfffffffc
+
+/*
+ * IO interface types
+ */
+#define CRAT_IOLINK_TYPE_UNDEFINED	0
+#define CRAT_IOLINK_TYPE_HYPERTRANSPORT	1
+#define CRAT_IOLINK_TYPE_PCIEXPRESS	2
+#define CRAT_IOLINK_TYPE_OTHER		3
+#define CRAT_IOLINK_TYPE_MAX		255
+
+#define CRAT_IOLINK_RESERVED_LENGTH 24
+
+struct crat_subtype_iolink {
+	uint8_t		type;
+	uint8_t		length;
+	uint16_t	reserved;
+	uint32_t	flags;
+	uint32_t	proximity_domain_from;
+	uint32_t	proximity_domain_to;
+	uint8_t		io_interface_type;
+	uint8_t		version_major;
+	uint16_t	version_minor;
+	uint32_t	minimum_latency;
+	uint32_t	maximum_latency;
+	uint32_t	minimum_bandwidth_mbs;
+	uint32_t	maximum_bandwidth_mbs;
+	uint32_t	recommended_transfer_size;
+	uint8_t		reserved2[CRAT_IOLINK_RESERVED_LENGTH];
+};
+
+/*
+ * HSA generic sub-type header
+ */
+
+#define CRAT_SUBTYPE_FLAGS_ENABLED 0x00000001
+
+struct crat_subtype_generic {
+	uint8_t		type;
+	uint8_t		length;
+	uint16_t	reserved;
+	uint32_t	flags;
+};
+
+/*
+ * Component Locality Distance Information Table (CDIT)
+ */
+#define CDIT_OEMID_LENGTH	6
+#define CDIT_OEMTABLEID_LENGTH	8
+
+struct cdit_header {
+	uint32_t	signature;
+	uint32_t	length;
+	uint8_t		revision;
+	uint8_t		checksum;
+	uint8_t		oem_id[CDIT_OEMID_LENGTH];
+	uint8_t		oem_table_id[CDIT_OEMTABLEID_LENGTH];
+	uint32_t	oem_revision;
+	uint32_t	creator_id;
+	uint32_t	creator_revision;
+	uint32_t	total_entries;
+	uint16_t	num_domains;
+	uint8_t		entry[1];
+};
+
+#pragma pack()
+
+#endif /* KFD_CRAT_H_INCLUDED */
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
new file mode 100644
index 000000000000..25bc47f3c1cf
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -0,0 +1,290 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#include <linux/amd-iommu.h>
+#include <linux/bsearch.h>
+#include <linux/pci.h>
+#include <linux/slab.h>
+#include "kfd_priv.h"
+#include "kfd_device_queue_manager.h"
+#include "kfd_pm4_headers.h"
+
+#define MQD_SIZE_ALIGNED 768
+
+static const struct kfd_device_info kaveri_device_info = {
+	.max_pasid_bits = 16,
+	.ih_ring_entry_size = 4 * sizeof(uint32_t),
+	.mqd_size_aligned = MQD_SIZE_ALIGNED
+};
+
+struct kfd_deviceid {
+	unsigned short did;
+	const struct kfd_device_info *device_info;
+};
+
+/* Please keep this sorted by increasing device id. */
+static const struct kfd_deviceid supported_devices[] = {
+	{ 0x1304, &kaveri_device_info },	/* Kaveri */
+	{ 0x1305, &kaveri_device_info },	/* Kaveri */
+	{ 0x1306, &kaveri_device_info },	/* Kaveri */
+	{ 0x1307, &kaveri_device_info },	/* Kaveri */
+	{ 0x1309, &kaveri_device_info },	/* Kaveri */
+	{ 0x130A, &kaveri_device_info },	/* Kaveri */
+	{ 0x130B, &kaveri_device_info },	/* Kaveri */
+	{ 0x130C, &kaveri_device_info },	/* Kaveri */
+	{ 0x130D, &kaveri_device_info },	/* Kaveri */
+	{ 0x130E, &kaveri_device_info },	/* Kaveri */
+	{ 0x130F, &kaveri_device_info },	/* Kaveri */
+	{ 0x1310, &kaveri_device_info },	/* Kaveri */
+	{ 0x1311, &kaveri_device_info },	/* Kaveri */
+	{ 0x1312, &kaveri_device_info },	/* Kaveri */
+	{ 0x1313, &kaveri_device_info },	/* Kaveri */
+	{ 0x1315, &kaveri_device_info },	/* Kaveri */
+	{ 0x1316, &kaveri_device_info },	/* Kaveri */
+	{ 0x1317, &kaveri_device_info },	/* Kaveri */
+	{ 0x1318, &kaveri_device_info },	/* Kaveri */
+	{ 0x131B, &kaveri_device_info },	/* Kaveri */
+	{ 0x131C, &kaveri_device_info },	/* Kaveri */
+	{ 0x131D, &kaveri_device_info },	/* Kaveri */
+};
+
+static const struct kfd_device_info *lookup_device_info(unsigned short did)
+{
+	size_t i;
+
+	for (i = 0; i < ARRAY_SIZE(supported_devices); i++) {
+		if (supported_devices[i].did == did) {
+			BUG_ON(supported_devices[i].device_info == NULL);
+			return supported_devices[i].device_info;
+		}
+	}
+
+	return NULL;
+}
+
+struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd, struct pci_dev *pdev)
+{
+	struct kfd_dev *kfd;
+
+	const struct kfd_device_info *device_info =
+					lookup_device_info(pdev->device);
+
+	if (!device_info)
+		return NULL;
+
+	kfd = kzalloc(sizeof(*kfd), GFP_KERNEL);
+	if (!kfd)
+		return NULL;
+
+	kfd->kgd = kgd;
+	kfd->device_info = device_info;
+	kfd->pdev = pdev;
+	kfd->init_complete = false;
+
+	return kfd;
+}
+
+static bool device_iommu_pasid_init(struct kfd_dev *kfd)
+{
+	const u32 required_iommu_flags = AMD_IOMMU_DEVICE_FLAG_ATS_SUP |
+					AMD_IOMMU_DEVICE_FLAG_PRI_SUP |
+					AMD_IOMMU_DEVICE_FLAG_PASID_SUP;
+
+	struct amd_iommu_device_info iommu_info;
+	unsigned int pasid_limit;
+	int err;
+
+	err = amd_iommu_device_info(kfd->pdev, &iommu_info);
+	if (err < 0) {
+		dev_err(kfd_device,
+			"error getting iommu info. is the iommu enabled?\n");
+		return false;
+	}
+
+	if ((iommu_info.flags & required_iommu_flags) != required_iommu_flags) {
+		dev_err(kfd_device, "error required iommu flags ats(%i), pri(%i), pasid(%i)\n",
+		       (iommu_info.flags & AMD_IOMMU_DEVICE_FLAG_ATS_SUP) != 0,
+		       (iommu_info.flags & AMD_IOMMU_DEVICE_FLAG_PRI_SUP) != 0,
+		       (iommu_info.flags & AMD_IOMMU_DEVICE_FLAG_PASID_SUP) != 0);
+		return false;
+	}
+
+	pasid_limit = min_t(unsigned int,
+			(unsigned int)1 << kfd->device_info->max_pasid_bits,
+			iommu_info.max_pasids);
+	/*
+	 * last pasid is used for kernel queues doorbells
+	 * in the future the last pasid might be used for a kernel thread.
+	 */
+	pasid_limit = min_t(unsigned int,
+				pasid_limit,
+				kfd->doorbell_process_limit - 1);
+
+	err = amd_iommu_init_device(kfd->pdev, pasid_limit);
+	if (err < 0) {
+		dev_err(kfd_device, "error initializing iommu device\n");
+		return false;
+	}
+
+	if (!kfd_set_pasid_limit(pasid_limit)) {
+		dev_err(kfd_device, "error setting pasid limit\n");
+		amd_iommu_free_device(kfd->pdev);
+		return false;
+	}
+
+	return true;
+}
+
+static void iommu_pasid_shutdown_callback(struct pci_dev *pdev, int pasid)
+{
+	struct kfd_dev *dev = kfd_device_by_pci_dev(pdev);
+
+	if (dev)
+		kfd_unbind_process_from_device(dev, pasid);
+}
+
+bool kgd2kfd_device_init(struct kfd_dev *kfd,
+			 const struct kgd2kfd_shared_resources *gpu_resources)
+{
+	unsigned int size;
+
+	kfd->shared_resources = *gpu_resources;
+
+	/* calculate max size of mqds needed for queues */
+	size = max_num_of_queues_per_device *
+			kfd->device_info->mqd_size_aligned;
+
+	/* add another 512KB for all other allocations on gart */
+	size += 512 * 1024;
+
+	if (kfd2kgd->init_sa_manager(kfd->kgd, size)) {
+		dev_err(kfd_device,
+			"Error initializing sa manager for device (%x:%x)\n",
+			kfd->pdev->vendor, kfd->pdev->device);
+		goto out;
+	}
+
+	kfd_doorbell_init(kfd);
+
+	if (kfd_topology_add_device(kfd) != 0) {
+		dev_err(kfd_device,
+			"Error adding device (%x:%x) to topology\n",
+			kfd->pdev->vendor, kfd->pdev->device);
+		goto kfd_topology_add_device_error;
+	}
+
+	if (!device_iommu_pasid_init(kfd)) {
+		dev_err(kfd_device,
+			"Error initializing iommuv2 for device (%x:%x)\n",
+			kfd->pdev->vendor, kfd->pdev->device);
+		goto device_iommu_pasid_error;
+	}
+	amd_iommu_set_invalidate_ctx_cb(kfd->pdev,
+						iommu_pasid_shutdown_callback);
+
+	kfd->dqm = device_queue_manager_init(kfd);
+	if (!kfd->dqm) {
+		dev_err(kfd_device,
+			"Error initializing queue manager for device (%x:%x)\n",
+			kfd->pdev->vendor, kfd->pdev->device);
+		goto device_queue_manager_error;
+	}
+
+	if (kfd->dqm->start(kfd->dqm) != 0) {
+		dev_err(kfd_device,
+			"Error starting queuen manager for device (%x:%x)\n",
+			kfd->pdev->vendor, kfd->pdev->device);
+		goto dqm_start_error;
+	}
+
+	kfd->init_complete = true;
+	dev_info(kfd_device, "added device (%x:%x)\n", kfd->pdev->vendor,
+		 kfd->pdev->device);
+
+	pr_debug("kfd: Starting kfd with the following scheduling policy %d\n",
+		sched_policy);
+
+	goto out;
+
+dqm_start_error:
+	device_queue_manager_uninit(kfd->dqm);
+device_queue_manager_error:
+	amd_iommu_free_device(kfd->pdev);
+device_iommu_pasid_error:
+	kfd_topology_remove_device(kfd);
+kfd_topology_add_device_error:
+	kfd2kgd->fini_sa_manager(kfd->kgd);
+	dev_err(kfd_device,
+		"device (%x:%x) NOT added due to errors\n",
+		kfd->pdev->vendor, kfd->pdev->device);
+out:
+	return kfd->init_complete;
+}
+
+void kgd2kfd_device_exit(struct kfd_dev *kfd)
+{
+	if (kfd->init_complete) {
+		device_queue_manager_uninit(kfd->dqm);
+		amd_iommu_free_device(kfd->pdev);
+		kfd_topology_remove_device(kfd);
+	}
+
+	kfree(kfd);
+}
+
+void kgd2kfd_suspend(struct kfd_dev *kfd)
+{
+	BUG_ON(kfd == NULL);
+
+	if (kfd->init_complete) {
+		kfd->dqm->stop(kfd->dqm);
+		amd_iommu_set_invalidate_ctx_cb(kfd->pdev, NULL);
+		amd_iommu_free_device(kfd->pdev);
+	}
+}
+
+int kgd2kfd_resume(struct kfd_dev *kfd)
+{
+	unsigned int pasid_limit;
+	int err;
+
+	BUG_ON(kfd == NULL);
+
+	pasid_limit = kfd_get_pasid_limit();
+
+	if (kfd->init_complete) {
+		err = amd_iommu_init_device(kfd->pdev, pasid_limit);
+		if (err < 0)
+			return -ENXIO;
+		amd_iommu_set_invalidate_ctx_cb(kfd->pdev,
+						iommu_pasid_shutdown_callback);
+		kfd->dqm->start(kfd->dqm);
+	}
+
+	return 0;
+}
+
+/* This is called directly from KGD at ISR. */
+void kgd2kfd_interrupt(struct kfd_dev *kfd, const void *ih_ring_entry)
+{
+	/* Process interrupts / schedule work as necessary */
+}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
new file mode 100644
index 000000000000..0d8694f015c1
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
@@ -0,0 +1,1160 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include <linux/slab.h>
+#include <linux/list.h>
+#include <linux/types.h>
+#include <linux/printk.h>
+#include <linux/bitops.h>
+#include "kfd_priv.h"
+#include "kfd_device_queue_manager.h"
+#include "kfd_mqd_manager.h"
+#include "cik_regs.h"
+#include "kfd_kernel_queue.h"
+#include "../../radeon/cik_reg.h"
+
+/* Size of the per-pipe EOP queue */
+#define CIK_HPD_EOP_BYTES_LOG2 11
+#define CIK_HPD_EOP_BYTES (1U << CIK_HPD_EOP_BYTES_LOG2)
+
+static bool is_mem_initialized;
+
+static int init_memory(struct device_queue_manager *dqm);
+static int set_pasid_vmid_mapping(struct device_queue_manager *dqm,
+					unsigned int pasid, unsigned int vmid);
+
+static int create_compute_queue_nocpsch(struct device_queue_manager *dqm,
+					struct queue *q,
+					struct qcm_process_device *qpd);
+static int execute_queues_cpsch(struct device_queue_manager *dqm, bool lock);
+static int destroy_queues_cpsch(struct device_queue_manager *dqm, bool lock);
+
+
+static inline unsigned int get_pipes_num(struct device_queue_manager *dqm)
+{
+	BUG_ON(!dqm || !dqm->dev);
+	return dqm->dev->shared_resources.compute_pipe_count;
+}
+
+static inline unsigned int get_first_pipe(struct device_queue_manager *dqm)
+{
+	BUG_ON(!dqm);
+	return dqm->dev->shared_resources.first_compute_pipe;
+}
+
+static inline unsigned int get_pipes_num_cpsch(void)
+{
+	return PIPE_PER_ME_CP_SCHEDULING;
+}
+
+static inline unsigned int
+get_sh_mem_bases_nybble_64(struct kfd_process_device *pdd)
+{
+	uint32_t nybble;
+
+	nybble = (pdd->lds_base >> 60) & 0x0E;
+
+	return nybble;
+
+}
+
+static inline unsigned int get_sh_mem_bases_32(struct kfd_process_device *pdd)
+{
+	unsigned int shared_base;
+
+	shared_base = (pdd->lds_base >> 16) & 0xFF;
+
+	return shared_base;
+}
+
+static uint32_t compute_sh_mem_bases_64bit(unsigned int top_address_nybble);
+static void init_process_memory(struct device_queue_manager *dqm,
+				struct qcm_process_device *qpd)
+{
+	struct kfd_process_device *pdd;
+	unsigned int temp;
+
+	BUG_ON(!dqm || !qpd);
+
+	pdd = qpd_to_pdd(qpd);
+
+	/* check if sh_mem_config register already configured */
+	if (qpd->sh_mem_config == 0) {
+		qpd->sh_mem_config =
+			ALIGNMENT_MODE(SH_MEM_ALIGNMENT_MODE_UNALIGNED) |
+			DEFAULT_MTYPE(MTYPE_NONCACHED) |
+			APE1_MTYPE(MTYPE_NONCACHED);
+		qpd->sh_mem_ape1_limit = 0;
+		qpd->sh_mem_ape1_base = 0;
+	}
+
+	if (qpd->pqm->process->is_32bit_user_mode) {
+		temp = get_sh_mem_bases_32(pdd);
+		qpd->sh_mem_bases = SHARED_BASE(temp);
+		qpd->sh_mem_config |= PTR32;
+	} else {
+		temp = get_sh_mem_bases_nybble_64(pdd);
+		qpd->sh_mem_bases = compute_sh_mem_bases_64bit(temp);
+	}
+
+	pr_debug("kfd: is32bit process: %d sh_mem_bases nybble: 0x%X and register 0x%X\n",
+		qpd->pqm->process->is_32bit_user_mode, temp, qpd->sh_mem_bases);
+}
+
+static void program_sh_mem_settings(struct device_queue_manager *dqm,
+					struct qcm_process_device *qpd)
+{
+	return kfd2kgd->program_sh_mem_settings(dqm->dev->kgd, qpd->vmid,
+						qpd->sh_mem_config,
+						qpd->sh_mem_ape1_base,
+						qpd->sh_mem_ape1_limit,
+						qpd->sh_mem_bases);
+}
+
+static int allocate_vmid(struct device_queue_manager *dqm,
+			struct qcm_process_device *qpd,
+			struct queue *q)
+{
+	int bit, allocated_vmid;
+
+	if (dqm->vmid_bitmap == 0)
+		return -ENOMEM;
+
+	bit = find_first_bit((unsigned long *)&dqm->vmid_bitmap, CIK_VMID_NUM);
+	clear_bit(bit, (unsigned long *)&dqm->vmid_bitmap);
+
+	/* Kaveri kfd vmid's starts from vmid 8 */
+	allocated_vmid = bit + KFD_VMID_START_OFFSET;
+	pr_debug("kfd: vmid allocation %d\n", allocated_vmid);
+	qpd->vmid = allocated_vmid;
+	q->properties.vmid = allocated_vmid;
+
+	set_pasid_vmid_mapping(dqm, q->process->pasid, q->properties.vmid);
+	program_sh_mem_settings(dqm, qpd);
+
+	return 0;
+}
+
+static void deallocate_vmid(struct device_queue_manager *dqm,
+				struct qcm_process_device *qpd,
+				struct queue *q)
+{
+	int bit = qpd->vmid - KFD_VMID_START_OFFSET;
+
+	/* Release the vmid mapping */
+	set_pasid_vmid_mapping(dqm, 0, qpd->vmid);
+
+	set_bit(bit, (unsigned long *)&dqm->vmid_bitmap);
+	qpd->vmid = 0;
+	q->properties.vmid = 0;
+}
+
+static int create_queue_nocpsch(struct device_queue_manager *dqm,
+				struct queue *q,
+				struct qcm_process_device *qpd,
+				int *allocated_vmid)
+{
+	int retval;
+
+	BUG_ON(!dqm || !q || !qpd || !allocated_vmid);
+
+	pr_debug("kfd: In func %s\n", __func__);
+	print_queue(q);
+
+	mutex_lock(&dqm->lock);
+
+	if (dqm->total_queue_count >= max_num_of_queues_per_device) {
+		pr_warn("amdkfd: Can't create new usermode queue because %d queues were already created\n",
+				dqm->total_queue_count);
+		mutex_unlock(&dqm->lock);
+		return -EPERM;
+	}
+
+	if (list_empty(&qpd->queues_list)) {
+		retval = allocate_vmid(dqm, qpd, q);
+		if (retval != 0) {
+			mutex_unlock(&dqm->lock);
+			return retval;
+		}
+	}
+	*allocated_vmid = qpd->vmid;
+	q->properties.vmid = qpd->vmid;
+
+	retval = create_compute_queue_nocpsch(dqm, q, qpd);
+
+	if (retval != 0) {
+		if (list_empty(&qpd->queues_list)) {
+			deallocate_vmid(dqm, qpd, q);
+			*allocated_vmid = 0;
+		}
+		mutex_unlock(&dqm->lock);
+		return retval;
+	}
+
+	list_add(&q->list, &qpd->queues_list);
+	dqm->queue_count++;
+
+	/*
+	 * Unconditionally increment this counter, regardless of the queue's
+	 * type or whether the queue is active.
+	 */
+	dqm->total_queue_count++;
+	pr_debug("Total of %d queues are accountable so far\n",
+			dqm->total_queue_count);
+
+	mutex_unlock(&dqm->lock);
+	return 0;
+}
+
+static int allocate_hqd(struct device_queue_manager *dqm, struct queue *q)
+{
+	bool set;
+	int pipe, bit;
+
+	set = false;
+
+	for (pipe = dqm->next_pipe_to_allocate; pipe < get_pipes_num(dqm);
+			pipe = (pipe + 1) % get_pipes_num(dqm)) {
+		if (dqm->allocated_queues[pipe] != 0) {
+			bit = find_first_bit(
+				(unsigned long *)&dqm->allocated_queues[pipe],
+				QUEUES_PER_PIPE);
+
+			clear_bit(bit,
+				(unsigned long *)&dqm->allocated_queues[pipe]);
+			q->pipe = pipe;
+			q->queue = bit;
+			set = true;
+			break;
+		}
+	}
+
+	if (set == false)
+		return -EBUSY;
+
+	pr_debug("kfd: DQM %s hqd slot - pipe (%d) queue(%d)\n",
+				__func__, q->pipe, q->queue);
+	/* horizontal hqd allocation */
+	dqm->next_pipe_to_allocate = (pipe + 1) % get_pipes_num(dqm);
+
+	return 0;
+}
+
+static inline void deallocate_hqd(struct device_queue_manager *dqm,
+				struct queue *q)
+{
+	set_bit(q->queue, (unsigned long *)&dqm->allocated_queues[q->pipe]);
+}
+
+static int create_compute_queue_nocpsch(struct device_queue_manager *dqm,
+					struct queue *q,
+					struct qcm_process_device *qpd)
+{
+	int retval;
+	struct mqd_manager *mqd;
+
+	BUG_ON(!dqm || !q || !qpd);
+
+	mqd = dqm->get_mqd_manager(dqm, KFD_MQD_TYPE_CIK_COMPUTE);
+	if (mqd == NULL)
+		return -ENOMEM;
+
+	retval = allocate_hqd(dqm, q);
+	if (retval != 0)
+		return retval;
+
+	retval = mqd->init_mqd(mqd, &q->mqd, &q->mqd_mem_obj,
+				&q->gart_mqd_addr, &q->properties);
+	if (retval != 0) {
+		deallocate_hqd(dqm, q);
+		return retval;
+	}
+
+	pr_debug("kfd: loading mqd to hqd on pipe (%d) queue (%d)\n",
+			q->pipe,
+			q->queue);
+
+	retval = mqd->load_mqd(mqd, q->mqd, q->pipe,
+			q->queue, (uint32_t __user *) q->properties.write_ptr);
+	if (retval != 0) {
+		deallocate_hqd(dqm, q);
+		mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj);
+		return retval;
+	}
+
+	return 0;
+}
+
+static int destroy_queue_nocpsch(struct device_queue_manager *dqm,
+				struct qcm_process_device *qpd,
+				struct queue *q)
+{
+	int retval;
+	struct mqd_manager *mqd;
+
+	BUG_ON(!dqm || !q || !q->mqd || !qpd);
+
+	retval = 0;
+
+	pr_debug("kfd: In Func %s\n", __func__);
+
+	mutex_lock(&dqm->lock);
+	mqd = dqm->get_mqd_manager(dqm, KFD_MQD_TYPE_CIK_COMPUTE);
+	if (mqd == NULL) {
+		retval = -ENOMEM;
+		goto out;
+	}
+
+	retval = mqd->destroy_mqd(mqd, q->mqd,
+				KFD_PREEMPT_TYPE_WAVEFRONT,
+				QUEUE_PREEMPT_DEFAULT_TIMEOUT_MS,
+				q->pipe, q->queue);
+
+	if (retval != 0)
+		goto out;
+
+	deallocate_hqd(dqm, q);
+
+	mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj);
+
+	list_del(&q->list);
+	if (list_empty(&qpd->queues_list))
+		deallocate_vmid(dqm, qpd, q);
+	dqm->queue_count--;
+
+	/*
+	 * Unconditionally decrement this counter, regardless of the queue's
+	 * type
+	 */
+	dqm->total_queue_count--;
+	pr_debug("Total of %d queues are accountable so far\n",
+			dqm->total_queue_count);
+
+out:
+	mutex_unlock(&dqm->lock);
+	return retval;
+}
+
+static int update_queue(struct device_queue_manager *dqm, struct queue *q)
+{
+	int retval;
+	struct mqd_manager *mqd;
+	bool prev_active = false;
+
+	BUG_ON(!dqm || !q || !q->mqd);
+
+	mutex_lock(&dqm->lock);
+	mqd = dqm->get_mqd_manager(dqm, KFD_MQD_TYPE_CIK_COMPUTE);
+	if (mqd == NULL) {
+		mutex_unlock(&dqm->lock);
+		return -ENOMEM;
+	}
+
+	if (q->properties.is_active == true)
+		prev_active = true;
+
+	/*
+	 *
+	 * check active state vs. the previous state
+	 * and modify counter accordingly
+	 */
+	retval = mqd->update_mqd(mqd, q->mqd, &q->properties);
+	if ((q->properties.is_active == true) && (prev_active == false))
+		dqm->queue_count++;
+	else if ((q->properties.is_active == false) && (prev_active == true))
+		dqm->queue_count--;
+
+	if (sched_policy != KFD_SCHED_POLICY_NO_HWS)
+		retval = execute_queues_cpsch(dqm, false);
+
+	mutex_unlock(&dqm->lock);
+	return retval;
+}
+
+static struct mqd_manager *get_mqd_manager_nocpsch(
+		struct device_queue_manager *dqm, enum KFD_MQD_TYPE type)
+{
+	struct mqd_manager *mqd;
+
+	BUG_ON(!dqm || type >= KFD_MQD_TYPE_MAX);
+
+	pr_debug("kfd: In func %s mqd type %d\n", __func__, type);
+
+	mqd = dqm->mqds[type];
+	if (!mqd) {
+		mqd = mqd_manager_init(type, dqm->dev);
+		if (mqd == NULL)
+			pr_err("kfd: mqd manager is NULL");
+		dqm->mqds[type] = mqd;
+	}
+
+	return mqd;
+}
+
+static int register_process_nocpsch(struct device_queue_manager *dqm,
+					struct qcm_process_device *qpd)
+{
+	struct device_process_node *n;
+
+	BUG_ON(!dqm || !qpd);
+
+	pr_debug("kfd: In func %s\n", __func__);
+
+	n = kzalloc(sizeof(struct device_process_node), GFP_KERNEL);
+	if (!n)
+		return -ENOMEM;
+
+	n->qpd = qpd;
+
+	mutex_lock(&dqm->lock);
+	list_add(&n->list, &dqm->queues);
+
+	init_process_memory(dqm, qpd);
+	dqm->processes_count++;
+
+	mutex_unlock(&dqm->lock);
+
+	return 0;
+}
+
+static int unregister_process_nocpsch(struct device_queue_manager *dqm,
+					struct qcm_process_device *qpd)
+{
+	int retval;
+	struct device_process_node *cur, *next;
+
+	BUG_ON(!dqm || !qpd);
+
+	BUG_ON(!list_empty(&qpd->queues_list));
+
+	pr_debug("kfd: In func %s\n", __func__);
+
+	retval = 0;
+	mutex_lock(&dqm->lock);
+
+	list_for_each_entry_safe(cur, next, &dqm->queues, list) {
+		if (qpd == cur->qpd) {
+			list_del(&cur->list);
+			kfree(cur);
+			dqm->processes_count--;
+			goto out;
+		}
+	}
+	/* qpd not found in dqm list */
+	retval = 1;
+out:
+	mutex_unlock(&dqm->lock);
+	return retval;
+}
+
+static int
+set_pasid_vmid_mapping(struct device_queue_manager *dqm, unsigned int pasid,
+			unsigned int vmid)
+{
+	uint32_t pasid_mapping;
+
+	pasid_mapping = (pasid == 0) ? 0 : (uint32_t)pasid |
+						ATC_VMID_PASID_MAPPING_VALID;
+	return kfd2kgd->set_pasid_vmid_mapping(dqm->dev->kgd, pasid_mapping,
+						vmid);
+}
+
+static uint32_t compute_sh_mem_bases_64bit(unsigned int top_address_nybble)
+{
+	/* In 64-bit mode, we can only control the top 3 bits of the LDS,
+	 * scratch and GPUVM apertures.
+	 * The hardware fills in the remaining 59 bits according to the
+	 * following pattern:
+	 * LDS:		X0000000'00000000 - X0000001'00000000 (4GB)
+	 * Scratch:	X0000001'00000000 - X0000002'00000000 (4GB)
+	 * GPUVM:	Y0010000'00000000 - Y0020000'00000000 (1TB)
+	 *
+	 * (where X/Y is the configurable nybble with the low-bit 0)
+	 *
+	 * LDS and scratch will have the same top nybble programmed in the
+	 * top 3 bits of SH_MEM_BASES.PRIVATE_BASE.
+	 * GPUVM can have a different top nybble programmed in the
+	 * top 3 bits of SH_MEM_BASES.SHARED_BASE.
+	 * We don't bother to support different top nybbles
+	 * for LDS/Scratch and GPUVM.
+	 */
+
+	BUG_ON((top_address_nybble & 1) || top_address_nybble > 0xE ||
+		top_address_nybble == 0);
+
+	return PRIVATE_BASE(top_address_nybble << 12) |
+			SHARED_BASE(top_address_nybble << 12);
+}
+
+static int init_memory(struct device_queue_manager *dqm)
+{
+	int i, retval;
+
+	for (i = 8; i < 16; i++)
+		set_pasid_vmid_mapping(dqm, 0, i);
+
+	retval = kfd2kgd->init_memory(dqm->dev->kgd);
+	if (retval == 0)
+		is_mem_initialized = true;
+	return retval;
+}
+
+
+static int init_pipelines(struct device_queue_manager *dqm,
+			unsigned int pipes_num, unsigned int first_pipe)
+{
+	void *hpdptr;
+	struct mqd_manager *mqd;
+	unsigned int i, err, inx;
+	uint64_t pipe_hpd_addr;
+
+	BUG_ON(!dqm || !dqm->dev);
+
+	pr_debug("kfd: In func %s\n", __func__);
+
+	/*
+	 * Allocate memory for the HPDs. This is hardware-owned per-pipe data.
+	 * The driver never accesses this memory after zeroing it.
+	 * It doesn't even have to be saved/restored on suspend/resume
+	 * because it contains no data when there are no active queues.
+	 */
+
+	err = kfd2kgd->allocate_mem(dqm->dev->kgd,
+				CIK_HPD_EOP_BYTES * pipes_num,
+				PAGE_SIZE,
+				KFD_MEMPOOL_SYSTEM_WRITECOMBINE,
+				(struct kgd_mem **) &dqm->pipeline_mem);
+
+	if (err) {
+		pr_err("kfd: error allocate vidmem num pipes: %d\n",
+			pipes_num);
+		return -ENOMEM;
+	}
+
+	hpdptr = dqm->pipeline_mem->cpu_ptr;
+	dqm->pipelines_addr = dqm->pipeline_mem->gpu_addr;
+
+	memset(hpdptr, 0, CIK_HPD_EOP_BYTES * pipes_num);
+
+	mqd = dqm->get_mqd_manager(dqm, KFD_MQD_TYPE_CIK_COMPUTE);
+	if (mqd == NULL) {
+		kfd2kgd->free_mem(dqm->dev->kgd,
+				(struct kgd_mem *) dqm->pipeline_mem);
+		return -ENOMEM;
+	}
+
+	for (i = 0; i < pipes_num; i++) {
+		inx = i + first_pipe;
+		/*
+		 * HPD buffer on GTT is allocated by amdkfd, no need to waste
+		 * space in GTT for pipelines we don't initialize
+		 */
+		pipe_hpd_addr = dqm->pipelines_addr + i * CIK_HPD_EOP_BYTES;
+		pr_debug("kfd: pipeline address %llX\n", pipe_hpd_addr);
+		/* = log2(bytes/4)-1 */
+		kfd2kgd->init_pipeline(dqm->dev->kgd, inx,
+				CIK_HPD_EOP_BYTES_LOG2 - 3, pipe_hpd_addr);
+	}
+
+	return 0;
+}
+
+
+static int init_scheduler(struct device_queue_manager *dqm)
+{
+	int retval;
+
+	BUG_ON(!dqm);
+
+	pr_debug("kfd: In %s\n", __func__);
+
+	retval = init_pipelines(dqm, get_pipes_num(dqm), get_first_pipe(dqm));
+	if (retval != 0)
+		return retval;
+
+	retval = init_memory(dqm);
+
+	return retval;
+}
+
+static int initialize_nocpsch(struct device_queue_manager *dqm)
+{
+	int i;
+
+	BUG_ON(!dqm);
+
+	pr_debug("kfd: In func %s num of pipes: %d\n",
+			__func__, get_pipes_num(dqm));
+
+	mutex_init(&dqm->lock);
+	INIT_LIST_HEAD(&dqm->queues);
+	dqm->queue_count = dqm->next_pipe_to_allocate = 0;
+	dqm->allocated_queues = kcalloc(get_pipes_num(dqm),
+					sizeof(unsigned int), GFP_KERNEL);
+	if (!dqm->allocated_queues) {
+		mutex_destroy(&dqm->lock);
+		return -ENOMEM;
+	}
+
+	for (i = 0; i < get_pipes_num(dqm); i++)
+		dqm->allocated_queues[i] = (1 << QUEUES_PER_PIPE) - 1;
+
+	dqm->vmid_bitmap = (1 << VMID_PER_DEVICE) - 1;
+
+	init_scheduler(dqm);
+	return 0;
+}
+
+static void uninitialize_nocpsch(struct device_queue_manager *dqm)
+{
+	int i;
+
+	BUG_ON(!dqm);
+
+	BUG_ON(dqm->queue_count > 0 || dqm->processes_count > 0);
+
+	kfree(dqm->allocated_queues);
+	for (i = 0 ; i < KFD_MQD_TYPE_MAX ; i++)
+		kfree(dqm->mqds[i]);
+	mutex_destroy(&dqm->lock);
+	kfd2kgd->free_mem(dqm->dev->kgd,
+			(struct kgd_mem *) dqm->pipeline_mem);
+}
+
+static int start_nocpsch(struct device_queue_manager *dqm)
+{
+	return 0;
+}
+
+static int stop_nocpsch(struct device_queue_manager *dqm)
+{
+	return 0;
+}
+
+/*
+ * Device Queue Manager implementation for cp scheduler
+ */
+
+static int set_sched_resources(struct device_queue_manager *dqm)
+{
+	struct scheduling_resources res;
+	unsigned int queue_num, queue_mask;
+
+	BUG_ON(!dqm);
+
+	pr_debug("kfd: In func %s\n", __func__);
+
+	queue_num = get_pipes_num_cpsch() * QUEUES_PER_PIPE;
+	queue_mask = (1 << queue_num) - 1;
+	res.vmid_mask = (1 << VMID_PER_DEVICE) - 1;
+	res.vmid_mask <<= KFD_VMID_START_OFFSET;
+	res.queue_mask = queue_mask << (get_first_pipe(dqm) * QUEUES_PER_PIPE);
+	res.gws_mask = res.oac_mask = res.gds_heap_base =
+						res.gds_heap_size = 0;
+
+	pr_debug("kfd: scheduling resources:\n"
+			"      vmid mask: 0x%8X\n"
+			"      queue mask: 0x%8llX\n",
+			res.vmid_mask, res.queue_mask);
+
+	return pm_send_set_resources(&dqm->packets, &res);
+}
+
+static int initialize_cpsch(struct device_queue_manager *dqm)
+{
+	int retval;
+
+	BUG_ON(!dqm);
+
+	pr_debug("kfd: In func %s num of pipes: %d\n",
+			__func__, get_pipes_num_cpsch());
+
+	mutex_init(&dqm->lock);
+	INIT_LIST_HEAD(&dqm->queues);
+	dqm->queue_count = dqm->processes_count = 0;
+	dqm->active_runlist = false;
+	retval = init_pipelines(dqm, get_pipes_num(dqm), 0);
+	if (retval != 0)
+		goto fail_init_pipelines;
+
+	return 0;
+
+fail_init_pipelines:
+	mutex_destroy(&dqm->lock);
+	return retval;
+}
+
+static int start_cpsch(struct device_queue_manager *dqm)
+{
+	struct device_process_node *node;
+	int retval;
+
+	BUG_ON(!dqm);
+
+	retval = 0;
+
+	retval = pm_init(&dqm->packets, dqm);
+	if (retval != 0)
+		goto fail_packet_manager_init;
+
+	retval = set_sched_resources(dqm);
+	if (retval != 0)
+		goto fail_set_sched_resources;
+
+	pr_debug("kfd: allocating fence memory\n");
+
+	/* allocate fence memory on the gart */
+	retval = kfd2kgd->allocate_mem(dqm->dev->kgd,
+					sizeof(*dqm->fence_addr),
+					32,
+					KFD_MEMPOOL_SYSTEM_WRITECOMBINE,
+					(struct kgd_mem **) &dqm->fence_mem);
+
+	if (retval != 0)
+		goto fail_allocate_vidmem;
+
+	dqm->fence_addr = dqm->fence_mem->cpu_ptr;
+	dqm->fence_gpu_addr = dqm->fence_mem->gpu_addr;
+
+	list_for_each_entry(node, &dqm->queues, list)
+		if (node->qpd->pqm->process && dqm->dev)
+			kfd_bind_process_to_device(dqm->dev,
+						node->qpd->pqm->process);
+
+	execute_queues_cpsch(dqm, true);
+
+	return 0;
+fail_allocate_vidmem:
+fail_set_sched_resources:
+	pm_uninit(&dqm->packets);
+fail_packet_manager_init:
+	return retval;
+}
+
+static int stop_cpsch(struct device_queue_manager *dqm)
+{
+	struct device_process_node *node;
+	struct kfd_process_device *pdd;
+
+	BUG_ON(!dqm);
+
+	destroy_queues_cpsch(dqm, true);
+
+	list_for_each_entry(node, &dqm->queues, list) {
+		pdd = qpd_to_pdd(node->qpd);
+		pdd->bound = false;
+	}
+	kfd2kgd->free_mem(dqm->dev->kgd,
+			(struct kgd_mem *) dqm->fence_mem);
+	pm_uninit(&dqm->packets);
+
+	return 0;
+}
+
+static int create_kernel_queue_cpsch(struct device_queue_manager *dqm,
+					struct kernel_queue *kq,
+					struct qcm_process_device *qpd)
+{
+	BUG_ON(!dqm || !kq || !qpd);
+
+	pr_debug("kfd: In func %s\n", __func__);
+
+	mutex_lock(&dqm->lock);
+	if (dqm->total_queue_count >= max_num_of_queues_per_device) {
+		pr_warn("amdkfd: Can't create new kernel queue because %d queues were already created\n",
+				dqm->total_queue_count);
+		mutex_unlock(&dqm->lock);
+		return -EPERM;
+	}
+
+	/*
+	 * Unconditionally increment this counter, regardless of the queue's
+	 * type or whether the queue is active.
+	 */
+	dqm->total_queue_count++;
+	pr_debug("Total of %d queues are accountable so far\n",
+			dqm->total_queue_count);
+
+	list_add(&kq->list, &qpd->priv_queue_list);
+	dqm->queue_count++;
+	qpd->is_debug = true;
+	execute_queues_cpsch(dqm, false);
+	mutex_unlock(&dqm->lock);
+
+	return 0;
+}
+
+static void destroy_kernel_queue_cpsch(struct device_queue_manager *dqm,
+					struct kernel_queue *kq,
+					struct qcm_process_device *qpd)
+{
+	BUG_ON(!dqm || !kq);
+
+	pr_debug("kfd: In %s\n", __func__);
+
+	mutex_lock(&dqm->lock);
+	destroy_queues_cpsch(dqm, false);
+	list_del(&kq->list);
+	dqm->queue_count--;
+	qpd->is_debug = false;
+	execute_queues_cpsch(dqm, false);
+	/*
+	 * Unconditionally decrement this counter, regardless of the queue's
+	 * type.
+	 */
+	dqm->total_queue_count++;
+	pr_debug("Total of %d queues are accountable so far\n",
+			dqm->total_queue_count);
+	mutex_unlock(&dqm->lock);
+}
+
+static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q,
+			struct qcm_process_device *qpd, int *allocate_vmid)
+{
+	int retval;
+	struct mqd_manager *mqd;
+
+	BUG_ON(!dqm || !q || !qpd);
+
+	retval = 0;
+
+	if (allocate_vmid)
+		*allocate_vmid = 0;
+
+	mutex_lock(&dqm->lock);
+
+	if (dqm->total_queue_count >= max_num_of_queues_per_device) {
+		pr_warn("amdkfd: Can't create new usermode queue because %d queues were already created\n",
+				dqm->total_queue_count);
+		retval = -EPERM;
+		goto out;
+	}
+
+	mqd = dqm->get_mqd_manager(dqm, KFD_MQD_TYPE_CIK_CP);
+	if (mqd == NULL) {
+		mutex_unlock(&dqm->lock);
+		return -ENOMEM;
+	}
+
+	retval = mqd->init_mqd(mqd, &q->mqd, &q->mqd_mem_obj,
+				&q->gart_mqd_addr, &q->properties);
+	if (retval != 0)
+		goto out;
+
+	list_add(&q->list, &qpd->queues_list);
+	if (q->properties.is_active) {
+		dqm->queue_count++;
+		retval = execute_queues_cpsch(dqm, false);
+	}
+
+	/*
+	 * Unconditionally increment this counter, regardless of the queue's
+	 * type or whether the queue is active.
+	 */
+	dqm->total_queue_count++;
+
+	pr_debug("Total of %d queues are accountable so far\n",
+			dqm->total_queue_count);
+
+out:
+	mutex_unlock(&dqm->lock);
+	return retval;
+}
+
+static int fence_wait_timeout(unsigned int *fence_addr,
+				unsigned int fence_value,
+				unsigned long timeout)
+{
+	BUG_ON(!fence_addr);
+	timeout += jiffies;
+
+	while (*fence_addr != fence_value) {
+		if (time_after(jiffies, timeout)) {
+			pr_err("kfd: qcm fence wait loop timeout expired\n");
+			return -ETIME;
+		}
+		cpu_relax();
+	}
+
+	return 0;
+}
+
+static int destroy_queues_cpsch(struct device_queue_manager *dqm, bool lock)
+{
+	int retval;
+
+	BUG_ON(!dqm);
+
+	retval = 0;
+
+	if (lock)
+		mutex_lock(&dqm->lock);
+	if (dqm->active_runlist == false)
+		goto out;
+	retval = pm_send_unmap_queue(&dqm->packets, KFD_QUEUE_TYPE_COMPUTE,
+			KFD_PREEMPT_TYPE_FILTER_ALL_QUEUES, 0, false, 0);
+	if (retval != 0)
+		goto out;
+
+	*dqm->fence_addr = KFD_FENCE_INIT;
+	pm_send_query_status(&dqm->packets, dqm->fence_gpu_addr,
+				KFD_FENCE_COMPLETED);
+	/* should be timed out */
+	fence_wait_timeout(dqm->fence_addr, KFD_FENCE_COMPLETED,
+				QUEUE_PREEMPT_DEFAULT_TIMEOUT_MS);
+	pm_release_ib(&dqm->packets);
+	dqm->active_runlist = false;
+
+out:
+	if (lock)
+		mutex_unlock(&dqm->lock);
+	return retval;
+}
+
+static int execute_queues_cpsch(struct device_queue_manager *dqm, bool lock)
+{
+	int retval;
+
+	BUG_ON(!dqm);
+
+	if (lock)
+		mutex_lock(&dqm->lock);
+
+	retval = destroy_queues_cpsch(dqm, false);
+	if (retval != 0) {
+		pr_err("kfd: the cp might be in an unrecoverable state due to an unsuccessful queues preemption");
+		goto out;
+	}
+
+	if (dqm->queue_count <= 0 || dqm->processes_count <= 0) {
+		retval = 0;
+		goto out;
+	}
+
+	if (dqm->active_runlist) {
+		retval = 0;
+		goto out;
+	}
+
+	retval = pm_send_runlist(&dqm->packets, &dqm->queues);
+	if (retval != 0) {
+		pr_err("kfd: failed to execute runlist");
+		goto out;
+	}
+	dqm->active_runlist = true;
+
+out:
+	if (lock)
+		mutex_unlock(&dqm->lock);
+	return retval;
+}
+
+static int destroy_queue_cpsch(struct device_queue_manager *dqm,
+				struct qcm_process_device *qpd,
+				struct queue *q)
+{
+	int retval;
+	struct mqd_manager *mqd;
+
+	BUG_ON(!dqm || !qpd || !q);
+
+	retval = 0;
+
+	/* remove queue from list to prevent rescheduling after preemption */
+	mutex_lock(&dqm->lock);
+
+	mqd = dqm->get_mqd_manager(dqm, KFD_MQD_TYPE_CIK_CP);
+	if (!mqd) {
+		retval = -ENOMEM;
+		goto failed;
+	}
+
+	list_del(&q->list);
+	dqm->queue_count--;
+
+	execute_queues_cpsch(dqm, false);
+
+	mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj);
+
+	/*
+	 * Unconditionally decrement this counter, regardless of the queue's
+	 * type
+	 */
+	dqm->total_queue_count--;
+	pr_debug("Total of %d queues are accountable so far\n",
+			dqm->total_queue_count);
+
+	mutex_unlock(&dqm->lock);
+
+	return 0;
+
+failed:
+	mutex_unlock(&dqm->lock);
+	return retval;
+}
+
+/*
+ * Low bits must be 0000/FFFF as required by HW, high bits must be 0 to
+ * stay in user mode.
+ */
+#define APE1_FIXED_BITS_MASK 0xFFFF80000000FFFFULL
+/* APE1 limit is inclusive and 64K aligned. */
+#define APE1_LIMIT_ALIGNMENT 0xFFFF
+
+static bool set_cache_memory_policy(struct device_queue_manager *dqm,
+				   struct qcm_process_device *qpd,
+				   enum cache_policy default_policy,
+				   enum cache_policy alternate_policy,
+				   void __user *alternate_aperture_base,
+				   uint64_t alternate_aperture_size)
+{
+	uint32_t default_mtype;
+	uint32_t ape1_mtype;
+
+	pr_debug("kfd: In func %s\n", __func__);
+
+	mutex_lock(&dqm->lock);
+
+	if (alternate_aperture_size == 0) {
+		/* base > limit disables APE1 */
+		qpd->sh_mem_ape1_base = 1;
+		qpd->sh_mem_ape1_limit = 0;
+	} else {
+		/*
+		 * In FSA64, APE1_Base[63:0] = { 16{SH_MEM_APE1_BASE[31]},
+		 *			SH_MEM_APE1_BASE[31:0], 0x0000 }
+		 * APE1_Limit[63:0] = { 16{SH_MEM_APE1_LIMIT[31]},
+		 *			SH_MEM_APE1_LIMIT[31:0], 0xFFFF }
+		 * Verify that the base and size parameters can be
+		 * represented in this format and convert them.
+		 * Additionally restrict APE1 to user-mode addresses.
+		 */
+
+		uint64_t base = (uintptr_t)alternate_aperture_base;
+		uint64_t limit = base + alternate_aperture_size - 1;
+
+		if (limit <= base)
+			goto out;
+
+		if ((base & APE1_FIXED_BITS_MASK) != 0)
+			goto out;
+
+		if ((limit & APE1_FIXED_BITS_MASK) != APE1_LIMIT_ALIGNMENT)
+			goto out;
+
+		qpd->sh_mem_ape1_base = base >> 16;
+		qpd->sh_mem_ape1_limit = limit >> 16;
+	}
+
+	default_mtype = (default_policy == cache_policy_coherent) ?
+			MTYPE_NONCACHED :
+			MTYPE_CACHED;
+
+	ape1_mtype = (alternate_policy == cache_policy_coherent) ?
+			MTYPE_NONCACHED :
+			MTYPE_CACHED;
+
+	qpd->sh_mem_config = (qpd->sh_mem_config & PTR32)
+			| ALIGNMENT_MODE(SH_MEM_ALIGNMENT_MODE_UNALIGNED)
+			| DEFAULT_MTYPE(default_mtype)
+			| APE1_MTYPE(ape1_mtype);
+
+	if ((sched_policy == KFD_SCHED_POLICY_NO_HWS) && (qpd->vmid != 0))
+		program_sh_mem_settings(dqm, qpd);
+
+	pr_debug("kfd: sh_mem_config: 0x%x, ape1_base: 0x%x, ape1_limit: 0x%x\n",
+		qpd->sh_mem_config, qpd->sh_mem_ape1_base,
+		qpd->sh_mem_ape1_limit);
+
+	mutex_unlock(&dqm->lock);
+	return true;
+
+out:
+	mutex_unlock(&dqm->lock);
+	return false;
+}
+
+struct device_queue_manager *device_queue_manager_init(struct kfd_dev *dev)
+{
+	struct device_queue_manager *dqm;
+
+	BUG_ON(!dev);
+
+	dqm = kzalloc(sizeof(struct device_queue_manager), GFP_KERNEL);
+	if (!dqm)
+		return NULL;
+
+	dqm->dev = dev;
+	switch (sched_policy) {
+	case KFD_SCHED_POLICY_HWS:
+	case KFD_SCHED_POLICY_HWS_NO_OVERSUBSCRIPTION:
+		/* initialize dqm for cp scheduling */
+		dqm->create_queue = create_queue_cpsch;
+		dqm->initialize = initialize_cpsch;
+		dqm->start = start_cpsch;
+		dqm->stop = stop_cpsch;
+		dqm->destroy_queue = destroy_queue_cpsch;
+		dqm->update_queue = update_queue;
+		dqm->get_mqd_manager = get_mqd_manager_nocpsch;
+		dqm->register_process = register_process_nocpsch;
+		dqm->unregister_process = unregister_process_nocpsch;
+		dqm->uninitialize = uninitialize_nocpsch;
+		dqm->create_kernel_queue = create_kernel_queue_cpsch;
+		dqm->destroy_kernel_queue = destroy_kernel_queue_cpsch;
+		dqm->set_cache_memory_policy = set_cache_memory_policy;
+		break;
+	case KFD_SCHED_POLICY_NO_HWS:
+		/* initialize dqm for no cp scheduling */
+		dqm->start = start_nocpsch;
+		dqm->stop = stop_nocpsch;
+		dqm->create_queue = create_queue_nocpsch;
+		dqm->destroy_queue = destroy_queue_nocpsch;
+		dqm->update_queue = update_queue;
+		dqm->get_mqd_manager = get_mqd_manager_nocpsch;
+		dqm->register_process = register_process_nocpsch;
+		dqm->unregister_process = unregister_process_nocpsch;
+		dqm->initialize = initialize_nocpsch;
+		dqm->uninitialize = uninitialize_nocpsch;
+		dqm->set_cache_memory_policy = set_cache_memory_policy;
+		break;
+	default:
+		BUG();
+		break;
+	}
+
+	if (dqm->initialize(dqm) != 0) {
+		kfree(dqm);
+		return NULL;
+	}
+
+	return dqm;
+}
+
+void device_queue_manager_uninit(struct device_queue_manager *dqm)
+{
+	BUG_ON(!dqm);
+
+	dqm->uninitialize(dqm);
+	kfree(dqm);
+}
+
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
new file mode 100644
index 000000000000..52035bf0c1cb
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
@@ -0,0 +1,147 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef KFD_DEVICE_QUEUE_MANAGER_H_
+#define KFD_DEVICE_QUEUE_MANAGER_H_
+
+#include <linux/rwsem.h>
+#include <linux/list.h>
+#include "kfd_priv.h"
+#include "kfd_mqd_manager.h"
+
+#define QUEUE_PREEMPT_DEFAULT_TIMEOUT_MS	(500)
+#define QUEUES_PER_PIPE				(8)
+#define PIPE_PER_ME_CP_SCHEDULING		(3)
+#define CIK_VMID_NUM				(8)
+#define KFD_VMID_START_OFFSET			(8)
+#define VMID_PER_DEVICE				CIK_VMID_NUM
+#define KFD_DQM_FIRST_PIPE			(0)
+
+struct device_process_node {
+	struct qcm_process_device *qpd;
+	struct list_head list;
+};
+
+/**
+ * struct device_queue_manager
+ *
+ * @create_queue: Queue creation routine.
+ *
+ * @destroy_queue: Queue destruction routine.
+ *
+ * @update_queue: Queue update routine.
+ *
+ * @get_mqd_manager: Returns the mqd manager according to the mqd type.
+ *
+ * @exeute_queues: Dispatches the queues list to the H/W.
+ *
+ * @register_process: This routine associates a specific process with device.
+ *
+ * @unregister_process: destroys the associations between process to device.
+ *
+ * @initialize: Initializes the pipelines and memory module for that device.
+ *
+ * @start: Initializes the resources/modules the the device needs for queues
+ * execution. This function is called on device initialization and after the
+ * system woke up after suspension.
+ *
+ * @stop: This routine stops execution of all the active queue running on the
+ * H/W and basically this function called on system suspend.
+ *
+ * @uninitialize: Destroys all the device queue manager resources allocated in
+ * initialize routine.
+ *
+ * @create_kernel_queue: Creates kernel queue. Used for debug queue.
+ *
+ * @destroy_kernel_queue: Destroys kernel queue. Used for debug queue.
+ *
+ * @set_cache_memory_policy: Sets memory policy (cached/ non cached) for the
+ * memory apertures.
+ *
+ * This struct is a base class for the kfd queues scheduler in the
+ * device level. The device base class should expose the basic operations
+ * for queue creation and queue destruction. This base class hides the
+ * scheduling mode of the driver and the specific implementation of the
+ * concrete device. This class is the only class in the queues scheduler
+ * that configures the H/W.
+ */
+
+struct device_queue_manager {
+	int	(*create_queue)(struct device_queue_manager *dqm,
+				struct queue *q,
+				struct qcm_process_device *qpd,
+				int *allocate_vmid);
+	int	(*destroy_queue)(struct device_queue_manager *dqm,
+				struct qcm_process_device *qpd,
+				struct queue *q);
+	int	(*update_queue)(struct device_queue_manager *dqm,
+				struct queue *q);
+
+	struct mqd_manager * (*get_mqd_manager)
+					(struct device_queue_manager *dqm,
+					enum KFD_MQD_TYPE type);
+
+	int	(*register_process)(struct device_queue_manager *dqm,
+					struct qcm_process_device *qpd);
+	int	(*unregister_process)(struct device_queue_manager *dqm,
+					struct qcm_process_device *qpd);
+	int	(*initialize)(struct device_queue_manager *dqm);
+	int	(*start)(struct device_queue_manager *dqm);
+	int	(*stop)(struct device_queue_manager *dqm);
+	void	(*uninitialize)(struct device_queue_manager *dqm);
+	int	(*create_kernel_queue)(struct device_queue_manager *dqm,
+					struct kernel_queue *kq,
+					struct qcm_process_device *qpd);
+	void	(*destroy_kernel_queue)(struct device_queue_manager *dqm,
+					struct kernel_queue *kq,
+					struct qcm_process_device *qpd);
+	bool	(*set_cache_memory_policy)(struct device_queue_manager *dqm,
+					   struct qcm_process_device *qpd,
+					   enum cache_policy default_policy,
+					   enum cache_policy alternate_policy,
+					   void __user *alternate_aperture_base,
+					   uint64_t alternate_aperture_size);
+
+
+	struct mqd_manager	*mqds[KFD_MQD_TYPE_MAX];
+	struct packet_manager	packets;
+	struct kfd_dev		*dev;
+	struct mutex		lock;
+	struct list_head	queues;
+	unsigned int		processes_count;
+	unsigned int		queue_count;
+	unsigned int		total_queue_count;
+	unsigned int		next_pipe_to_allocate;
+	unsigned int		*allocated_queues;
+	unsigned int		vmid_bitmap;
+	uint64_t		pipelines_addr;
+	struct kfd_mem_obj	*pipeline_mem;
+	uint64_t		fence_gpu_addr;
+	unsigned int		*fence_addr;
+	struct kfd_mem_obj	*fence_mem;
+	bool			active_runlist;
+};
+
+
+
+#endif /* KFD_DEVICE_QUEUE_MANAGER_H_ */
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c
new file mode 100644
index 000000000000..b5791a5c7c06
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c
@@ -0,0 +1,256 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "kfd_priv.h"
+#include <linux/mm.h>
+#include <linux/mman.h>
+#include <linux/slab.h>
+#include <linux/io.h>
+
+/*
+ * This extension supports a kernel level doorbells management for
+ * the kernel queues.
+ * Basically the last doorbells page is devoted to kernel queues
+ * and that's assures that any user process won't get access to the
+ * kernel doorbells page
+ */
+static DEFINE_MUTEX(doorbell_mutex);
+static unsigned long doorbell_available_index[
+	DIV_ROUND_UP(KFD_MAX_NUM_OF_QUEUES_PER_PROCESS, BITS_PER_LONG)] = { 0 };
+
+#define KERNEL_DOORBELL_PASID 1
+#define KFD_SIZE_OF_DOORBELL_IN_BYTES 4
+
+/*
+ * Each device exposes a doorbell aperture, a PCI MMIO aperture that
+ * receives 32-bit writes that are passed to queues as wptr values.
+ * The doorbells are intended to be written by applications as part
+ * of queueing work on user-mode queues.
+ * We assign doorbells to applications in PAGE_SIZE-sized and aligned chunks.
+ * We map the doorbell address space into user-mode when a process creates
+ * its first queue on each device.
+ * Although the mapping is done by KFD, it is equivalent to an mmap of
+ * the /dev/kfd with the particular device encoded in the mmap offset.
+ * There will be other uses for mmap of /dev/kfd, so only a range of
+ * offsets (KFD_MMAP_DOORBELL_START-END) is used for doorbells.
+ */
+
+/* # of doorbell bytes allocated for each process. */
+static inline size_t doorbell_process_allocation(void)
+{
+	return roundup(KFD_SIZE_OF_DOORBELL_IN_BYTES *
+			KFD_MAX_NUM_OF_QUEUES_PER_PROCESS,
+			PAGE_SIZE);
+}
+
+/* Doorbell calculations for device init. */
+void kfd_doorbell_init(struct kfd_dev *kfd)
+{
+	size_t doorbell_start_offset;
+	size_t doorbell_aperture_size;
+	size_t doorbell_process_limit;
+
+	/*
+	 * We start with calculations in bytes because the input data might
+	 * only be byte-aligned.
+	 * Only after we have done the rounding can we assume any alignment.
+	 */
+
+	doorbell_start_offset =
+			roundup(kfd->shared_resources.doorbell_start_offset,
+					doorbell_process_allocation());
+
+	doorbell_aperture_size =
+			rounddown(kfd->shared_resources.doorbell_aperture_size,
+					doorbell_process_allocation());
+
+	if (doorbell_aperture_size > doorbell_start_offset)
+		doorbell_process_limit =
+			(doorbell_aperture_size - doorbell_start_offset) /
+						doorbell_process_allocation();
+	else
+		doorbell_process_limit = 0;
+
+	kfd->doorbell_base = kfd->shared_resources.doorbell_physical_address +
+				doorbell_start_offset;
+
+	kfd->doorbell_id_offset = doorbell_start_offset / sizeof(u32);
+	kfd->doorbell_process_limit = doorbell_process_limit - 1;
+
+	kfd->doorbell_kernel_ptr = ioremap(kfd->doorbell_base,
+						doorbell_process_allocation());
+
+	BUG_ON(!kfd->doorbell_kernel_ptr);
+
+	pr_debug("kfd: doorbell initialization:\n");
+	pr_debug("kfd: doorbell base           == 0x%08lX\n",
+			(uintptr_t)kfd->doorbell_base);
+
+	pr_debug("kfd: doorbell_id_offset      == 0x%08lX\n",
+			kfd->doorbell_id_offset);
+
+	pr_debug("kfd: doorbell_process_limit  == 0x%08lX\n",
+			doorbell_process_limit);
+
+	pr_debug("kfd: doorbell_kernel_offset  == 0x%08lX\n",
+			(uintptr_t)kfd->doorbell_base);
+
+	pr_debug("kfd: doorbell aperture size  == 0x%08lX\n",
+			kfd->shared_resources.doorbell_aperture_size);
+
+	pr_debug("kfd: doorbell kernel address == 0x%08lX\n",
+			(uintptr_t)kfd->doorbell_kernel_ptr);
+}
+
+int kfd_doorbell_mmap(struct kfd_process *process, struct vm_area_struct *vma)
+{
+	phys_addr_t address;
+	struct kfd_dev *dev;
+
+	/*
+	 * For simplicitly we only allow mapping of the entire doorbell
+	 * allocation of a single device & process.
+	 */
+	if (vma->vm_end - vma->vm_start != doorbell_process_allocation())
+		return -EINVAL;
+
+	/* Find kfd device according to gpu id */
+	dev = kfd_device_by_id(vma->vm_pgoff);
+	if (dev == NULL)
+		return -EINVAL;
+
+	/* Find if pdd exists for combination of process and gpu id */
+	if (!kfd_get_process_device_data(dev, process, 0))
+		return -EINVAL;
+
+	/* Calculate physical address of doorbell */
+	address = kfd_get_process_doorbells(dev, process);
+
+	vma->vm_flags |= VM_IO | VM_DONTCOPY | VM_DONTEXPAND | VM_NORESERVE |
+				VM_DONTDUMP | VM_PFNMAP;
+
+	vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
+
+	pr_debug("kfd: mapping doorbell page in kfd_doorbell_mmap\n"
+		 "     target user address == 0x%08llX\n"
+		 "     physical address    == 0x%08llX\n"
+		 "     vm_flags            == 0x%04lX\n"
+		 "     size                == 0x%04lX\n",
+		 (unsigned long long) vma->vm_start, address, vma->vm_flags,
+		 doorbell_process_allocation());
+
+
+	return io_remap_pfn_range(vma,
+				vma->vm_start,
+				address >> PAGE_SHIFT,
+				doorbell_process_allocation(),
+				vma->vm_page_prot);
+}
+
+
+/* get kernel iomem pointer for a doorbell */
+u32 __iomem *kfd_get_kernel_doorbell(struct kfd_dev *kfd,
+					unsigned int *doorbell_off)
+{
+	u32 inx;
+
+	BUG_ON(!kfd || !doorbell_off);
+
+	mutex_lock(&doorbell_mutex);
+	inx = find_first_zero_bit(doorbell_available_index,
+					KFD_MAX_NUM_OF_QUEUES_PER_PROCESS);
+
+	__set_bit(inx, doorbell_available_index);
+	mutex_unlock(&doorbell_mutex);
+
+	if (inx >= KFD_MAX_NUM_OF_QUEUES_PER_PROCESS)
+		return NULL;
+
+	/*
+	 * Calculating the kernel doorbell offset using "faked" kernel
+	 * pasid that allocated for kernel queues only
+	 */
+	*doorbell_off = KERNEL_DOORBELL_PASID * (doorbell_process_allocation() /
+							sizeof(u32)) + inx;
+
+	pr_debug("kfd: get kernel queue doorbell\n"
+			 "     doorbell offset   == 0x%08d\n"
+			 "     kernel address    == 0x%08lX\n",
+		*doorbell_off, (uintptr_t)(kfd->doorbell_kernel_ptr + inx));
+
+	return kfd->doorbell_kernel_ptr + inx;
+}
+
+void kfd_release_kernel_doorbell(struct kfd_dev *kfd, u32 __iomem *db_addr)
+{
+	unsigned int inx;
+
+	BUG_ON(!kfd || !db_addr);
+
+	inx = (unsigned int)(db_addr - kfd->doorbell_kernel_ptr);
+
+	mutex_lock(&doorbell_mutex);
+	__clear_bit(inx, doorbell_available_index);
+	mutex_unlock(&doorbell_mutex);
+}
+
+inline void write_kernel_doorbell(u32 __iomem *db, u32 value)
+{
+	if (db) {
+		writel(value, db);
+		pr_debug("writing %d to doorbell address 0x%p\n", value, db);
+	}
+}
+
+/*
+ * queue_ids are in the range [0,MAX_PROCESS_QUEUES) and are mapped 1:1
+ * to doorbells with the process's doorbell page
+ */
+unsigned int kfd_queue_id_to_doorbell(struct kfd_dev *kfd,
+					struct kfd_process *process,
+					unsigned int queue_id)
+{
+	/*
+	 * doorbell_id_offset accounts for doorbells taken by KGD.
+	 * pasid * doorbell_process_allocation/sizeof(u32) adjusts
+	 * to the process's doorbells
+	 */
+	return kfd->doorbell_id_offset +
+		process->pasid * (doorbell_process_allocation()/sizeof(u32)) +
+		queue_id;
+}
+
+uint64_t kfd_get_number_elems(struct kfd_dev *kfd)
+{
+	uint64_t num_of_elems = (kfd->shared_resources.doorbell_aperture_size -
+				kfd->shared_resources.doorbell_start_offset) /
+					doorbell_process_allocation() + 1;
+
+	return num_of_elems;
+
+}
+
+phys_addr_t kfd_get_process_doorbells(struct kfd_dev *dev,
+					struct kfd_process *process)
+{
+	return dev->doorbell_base +
+		process->pasid * doorbell_process_allocation();
+}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c b/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
new file mode 100644
index 000000000000..e64aa99e5e41
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
@@ -0,0 +1,354 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include <linux/device.h>
+#include <linux/export.h>
+#include <linux/err.h>
+#include <linux/fs.h>
+#include <linux/sched.h>
+#include <linux/slab.h>
+#include <linux/uaccess.h>
+#include <linux/compat.h>
+#include <uapi/linux/kfd_ioctl.h>
+#include <linux/time.h>
+#include "kfd_priv.h"
+#include <linux/mm.h>
+#include <uapi/asm-generic/mman-common.h>
+#include <asm/processor.h>
+
+/*
+ * The primary memory I/O features being added for revisions of gfxip
+ * beyond 7.0 (Kaveri) are:
+ *
+ * Access to ATC/IOMMU mapped memory w/ associated extension of VA to 48b
+ *
+ * “Flat” shader memory access – These are new shader vector memory
+ * operations that do not reference a T#/V# so a “pointer” is what is
+ * sourced from the vector gprs for direct access to memory.
+ * This pointer space has the Shared(LDS) and Private(Scratch) memory
+ * mapped into this pointer space as apertures.
+ * The hardware then determines how to direct the memory request
+ * based on what apertures the request falls in.
+ *
+ * Unaligned support and alignment check
+ *
+ *
+ * System Unified Address - SUA
+ *
+ * The standard usage for GPU virtual addresses are that they are mapped by
+ * a set of page tables we call GPUVM and these page tables are managed by
+ * a combination of vidMM/driver software components.  The current virtual
+ * address (VA) range for GPUVM is 40b.
+ *
+ * As of gfxip7.1 and beyond we’re adding the ability for compute memory
+ * clients (CP/RLC, DMA, SHADER(ifetch, scalar, and vector ops)) to access
+ * the same page tables used by host x86 processors and that are managed by
+ * the operating system. This is via a technique and hardware called ATC/IOMMU.
+ * The GPU has the capability of accessing both the GPUVM and ATC address
+ * spaces for a given VMID (process) simultaneously and we call this feature
+ * system unified address (SUA).
+ *
+ * There are three fundamental address modes of operation for a given VMID
+ * (process) on the GPU:
+ *
+ *	HSA64 – 64b pointers and the default address space is ATC
+ *	HSA32 – 32b pointers and the default address space is ATC
+ *	GPUVM – 64b pointers and the default address space is GPUVM (driver
+ *		model mode)
+ *
+ *
+ * HSA64 - ATC/IOMMU 64b
+ *
+ * A 64b pointer in the AMD64/IA64 CPU architecture is not fully utilized
+ * by the CPU so an AMD CPU can only access the high area
+ * (VA[63:47] == 0x1FFFF) and low area (VA[63:47 == 0) of the address space
+ * so the actual VA carried to translation is 48b.  There is a “hole” in
+ * the middle of the 64b VA space.
+ *
+ * The GPU not only has access to all of the CPU accessible address space via
+ * ATC/IOMMU, but it also has access to the GPUVM address space.  The “system
+ * unified address” feature (SUA) is the mapping of GPUVM and ATC address
+ * spaces into a unified pointer space.  The method we take for 64b mode is
+ * to map the full 40b GPUVM address space into the hole of the 64b address
+ * space.
+
+ * The GPUVM_Base/GPUVM_Limit defines the aperture in the 64b space where we
+ * direct requests to be translated via GPUVM page tables instead of the
+ * IOMMU path.
+ *
+ *
+ * 64b to 49b Address conversion
+ *
+ * Note that there are still significant portions of unused regions (holes)
+ * in the 64b address space even for the GPU.  There are several places in
+ * the pipeline (sw and hw), we wish to compress the 64b virtual address
+ * to a 49b address.  This 49b address is constituted of an “ATC” bit
+ * plus a 48b virtual address.  This 49b address is what is passed to the
+ * translation hardware.  ATC==0 means the 48b address is a GPUVM address
+ * (max of 2^40 – 1) intended to be translated via GPUVM page tables.
+ * ATC==1 means the 48b address is intended to be translated via IOMMU
+ * page tables.
+ *
+ * A 64b pointer is compared to the apertures that are defined (Base/Limit), in
+ * this case the GPUVM aperture (red) is defined and if a pointer falls in this
+ * aperture, we subtract the GPUVM_Base address and set the ATC bit to zero
+ * as part of the 64b to 49b conversion.
+ *
+ * Where this 64b to 49b conversion is done is a function of the usage.
+ * Most GPU memory access is via memory objects where the driver builds
+ * a descriptor which consists of a base address and a memory access by
+ * the GPU usually consists of some kind of an offset or Cartesian coordinate
+ * that references this memory descriptor.  This is the case for shader
+ * instructions that reference the T# or V# constants, or for specified
+ * locations of assets (ex. the shader program location).  In these cases
+ * the driver is what handles the 64b to 49b conversion and the base
+ * address in the descriptor (ex. V# or T# or shader program location)
+ * is defined as a 48b address w/ an ATC bit.  For this usage a given
+ * memory object cannot straddle multiple apertures in the 64b address
+ * space. For example a shader program cannot jump in/out between ATC
+ * and GPUVM space.
+ *
+ * In some cases we wish to pass a 64b pointer to the GPU hardware and
+ * the GPU hw does the 64b to 49b conversion before passing memory
+ * requests to the cache/memory system.  This is the case for the
+ * S_LOAD and FLAT_* shader memory instructions where we have 64b pointers
+ * in scalar and vector GPRs respectively.
+ *
+ * In all cases (no matter where the 64b -> 49b conversion is done), the gfxip
+ * hardware sends a 48b address along w/ an ATC bit, to the memory controller
+ * on the memory request interfaces.
+ *
+ *	<client>_MC_rdreq_atc   // read request ATC bit
+ *
+ *		0 : <client>_MC_rdreq_addr is a GPUVM VA
+ *
+ *		1 : <client>_MC_rdreq_addr is a ATC VA
+ *
+ *
+ * “Spare” aperture (APE1)
+ *
+ * We use the GPUVM aperture to differentiate ATC vs. GPUVM, but we also use
+ * apertures to set the Mtype field for S_LOAD/FLAT_* ops which is input to the
+ * config tables for setting cache policies. The “spare” (APE1) aperture is
+ * motivated by getting a different Mtype from the default.
+ * The default aperture isn’t an actual base/limit aperture; it is just the
+ * address space that doesn’t hit any defined base/limit apertures.
+ * The following diagram is a complete picture of the gfxip7.x SUA apertures.
+ * The APE1 can be placed either below or above
+ * the hole (cannot be in the hole).
+ *
+ *
+ * General Aperture definitions and rules
+ *
+ * An aperture register definition consists of a Base, Limit, Mtype, and
+ * usually an ATC bit indicating which translation tables that aperture uses.
+ * In all cases (for SUA and DUA apertures discussed later), aperture base
+ * and limit definitions are 64KB aligned.
+ *
+ *	<ape>_Base[63:0] = { <ape>_Base_register[63:16], 0x0000 }
+ *
+ *	<ape>_Limit[63:0] = { <ape>_Limit_register[63:16], 0xFFFF }
+ *
+ * The base and limit are considered inclusive to an aperture so being
+ * inside an aperture means (address >= Base) AND (address <= Limit).
+ *
+ * In no case is a payload that straddles multiple apertures expected to work.
+ * For example a load_dword_x4 that starts in one aperture and ends in another,
+ * does not work.  For the vector FLAT_* ops we have detection capability in
+ * the shader for reporting a “memory violation” back to the
+ * SQ block for use in traps.
+ * A memory violation results when an op falls into the hole,
+ * or a payload straddles multiple apertures.  The S_LOAD instruction
+ * does not have this detection.
+ *
+ * Apertures cannot overlap.
+ *
+ *
+ *
+ * HSA32 - ATC/IOMMU 32b
+ *
+ * For HSA32 mode, the pointers are interpreted as 32 bits and use a single GPR
+ * instead of two for the S_LOAD and FLAT_* ops. The entire GPUVM space of 40b
+ * will not fit so there is only partial visibility to the GPUVM
+ * space (defined by the aperture) for S_LOAD and FLAT_* ops.
+ * There is no spare (APE1) aperture for HSA32 mode.
+ *
+ *
+ * GPUVM 64b mode (driver model)
+ *
+ * This mode is related to HSA64 in that the difference really is that
+ * the default aperture is GPUVM (ATC==0) and not ATC space.
+ * We have gfxip7.x hardware that has FLAT_* and S_LOAD support for
+ * SUA GPUVM mode, but does not support HSA32/HSA64.
+ *
+ *
+ * Device Unified Address - DUA
+ *
+ * Device unified address (DUA) is the name of the feature that maps the
+ * Shared(LDS) memory and Private(Scratch) memory into the overall address
+ * space for use by the new FLAT_* vector memory ops.  The Shared and
+ * Private memories are mapped as apertures into the address space,
+ * and the hardware detects when a FLAT_* memory request is to be redirected
+ * to the LDS or Scratch memory when it falls into one of these apertures.
+ * Like the SUA apertures, the Shared/Private apertures are 64KB aligned and
+ * the base/limit is “in” the aperture. For both HSA64 and GPUVM SUA modes,
+ * the Shared/Private apertures are always placed in a limited selection of
+ * options in the hole of the 64b address space. For HSA32 mode, the
+ * Shared/Private apertures can be placed anywhere in the 32b space
+ * except at 0.
+ *
+ *
+ * HSA64 Apertures for FLAT_* vector ops
+ *
+ * For HSA64 SUA mode, the Shared and Private apertures are always placed
+ * in the hole w/ a limited selection of possible locations. The requests
+ * that fall in the private aperture are expanded as a function of the
+ * work-item id (tid) and redirected to the location of the
+ * “hidden private memory”. The hidden private can be placed in either GPUVM
+ * or ATC space. The addresses that fall in the shared aperture are
+ * re-directed to the on-chip LDS memory hardware.
+ *
+ *
+ * HSA32 Apertures for FLAT_* vector ops
+ *
+ * In HSA32 mode, the Private and Shared apertures can be placed anywhere
+ * in the 32b space except at 0 (Private or Shared Base at zero disables
+ * the apertures). If the base address of the apertures are non-zero
+ * (ie apertures exists), the size is always 64KB.
+ *
+ *
+ * GPUVM Apertures for FLAT_* vector ops
+ *
+ * In GPUVM mode, the Shared/Private apertures are specified identically
+ * to HSA64 mode where they are always in the hole at a limited selection
+ * of locations.
+ *
+ *
+ * Aperture Definitions for SUA and DUA
+ *
+ * The interpretation of the aperture register definitions for a given
+ * VMID is a function of the “SUA Mode” which is one of HSA64, HSA32, or
+ * GPUVM64 discussed in previous sections. The mode is first decoded, and
+ * then the remaining register decode is a function of the mode.
+ *
+ *
+ * SUA Mode Decode
+ *
+ * For the S_LOAD and FLAT_* shader operations, the SUA mode is decoded from
+ * the COMPUTE_DISPATCH_INITIATOR:DATA_ATC bit and
+ * the SH_MEM_CONFIG:PTR32 bits.
+ *
+ * COMPUTE_DISPATCH_INITIATOR:DATA_ATC    SH_MEM_CONFIG:PTR32        Mode
+ *
+ * 1                                              0                  HSA64
+ *
+ * 1                                              1                  HSA32
+ *
+ * 0                                              X                 GPUVM64
+ *
+ * In general the hardware will ignore the PTR32 bit and treat
+ * as “0” whenever DATA_ATC = “0”, but sw should set PTR32=0
+ * when DATA_ATC=0.
+ *
+ * The DATA_ATC bit is only set for compute dispatches.
+ * All “Draw” dispatches are hardcoded to GPUVM64 mode
+ * for FLAT_* / S_LOAD operations.
+ */
+
+#define MAKE_GPUVM_APP_BASE(gpu_num) \
+	(((uint64_t)(gpu_num) << 61) + 0x1000000000000L)
+
+#define MAKE_GPUVM_APP_LIMIT(base) \
+	(((uint64_t)(base) & \
+		0xFFFFFF0000000000UL) | 0xFFFFFFFFFFL)
+
+#define MAKE_SCRATCH_APP_BASE(gpu_num) \
+	(((uint64_t)(gpu_num) << 61) + 0x100000000L)
+
+#define MAKE_SCRATCH_APP_LIMIT(base) \
+	(((uint64_t)base & 0xFFFFFFFF00000000UL) | 0xFFFFFFFF)
+
+#define MAKE_LDS_APP_BASE(gpu_num) \
+	(((uint64_t)(gpu_num) << 61) + 0x0)
+#define MAKE_LDS_APP_LIMIT(base) \
+	(((uint64_t)(base) & 0xFFFFFFFF00000000UL) | 0xFFFFFFFF)
+
+int kfd_init_apertures(struct kfd_process *process)
+{
+	uint8_t id  = 0;
+	struct kfd_dev *dev;
+	struct kfd_process_device *pdd;
+
+	/*Iterating over all devices*/
+	while ((dev = kfd_topology_enum_kfd_devices(id)) != NULL &&
+		id < NUM_OF_SUPPORTED_GPUS) {
+
+		pdd = kfd_get_process_device_data(dev, process, 1);
+		if (!pdd)
+			return -1;
+
+		/*
+		 * For 64 bit process aperture will be statically reserved in
+		 * the x86_64 non canonical process address space
+		 * amdkfd doesn't currently support apertures for 32 bit process
+		 */
+		if (process->is_32bit_user_mode) {
+			pdd->lds_base = pdd->lds_limit = 0;
+			pdd->gpuvm_base = pdd->gpuvm_limit = 0;
+			pdd->scratch_base = pdd->scratch_limit = 0;
+		} else {
+			/*
+			 * node id couldn't be 0 - the three MSB bits of
+			 * aperture shoudn't be 0
+			 */
+			pdd->lds_base = MAKE_LDS_APP_BASE(id + 1);
+
+			pdd->lds_limit = MAKE_LDS_APP_LIMIT(pdd->lds_base);
+
+			pdd->gpuvm_base = MAKE_GPUVM_APP_BASE(id + 1);
+
+			pdd->gpuvm_limit =
+					MAKE_GPUVM_APP_LIMIT(pdd->gpuvm_base);
+
+			pdd->scratch_base = MAKE_SCRATCH_APP_BASE(id + 1);
+
+			pdd->scratch_limit =
+				MAKE_SCRATCH_APP_LIMIT(pdd->scratch_base);
+		}
+
+		dev_dbg(kfd_device, "node id %u\n", id);
+		dev_dbg(kfd_device, "gpu id %u\n", pdd->dev->id);
+		dev_dbg(kfd_device, "lds_base %llX\n", pdd->lds_base);
+		dev_dbg(kfd_device, "lds_limit %llX\n", pdd->lds_limit);
+		dev_dbg(kfd_device, "gpuvm_base %llX\n", pdd->gpuvm_base);
+		dev_dbg(kfd_device, "gpuvm_limit %llX\n", pdd->gpuvm_limit);
+		dev_dbg(kfd_device, "scratch_base %llX\n", pdd->scratch_base);
+		dev_dbg(kfd_device, "scratch_limit %llX\n", pdd->scratch_limit);
+
+		id++;
+	}
+
+	return 0;
+}
+
+
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
new file mode 100644
index 000000000000..935071410724
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
@@ -0,0 +1,353 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/mutex.h>
+#include <linux/slab.h>
+#include <linux/printk.h>
+#include <linux/sched.h>
+#include "kfd_kernel_queue.h"
+#include "kfd_priv.h"
+#include "kfd_device_queue_manager.h"
+#include "kfd_pm4_headers.h"
+#include "kfd_pm4_opcodes.h"
+
+#define PM4_COUNT_ZERO (((1 << 15) - 1) << 16)
+
+static bool initialize(struct kernel_queue *kq, struct kfd_dev *dev,
+		enum kfd_queue_type type, unsigned int queue_size)
+{
+	struct queue_properties prop;
+	int retval;
+	union PM4_MES_TYPE_3_HEADER nop;
+
+	BUG_ON(!kq || !dev);
+	BUG_ON(type != KFD_QUEUE_TYPE_DIQ && type != KFD_QUEUE_TYPE_HIQ);
+
+	pr_debug("kfd: In func %s initializing queue type %d size %d\n",
+			__func__, KFD_QUEUE_TYPE_HIQ, queue_size);
+
+	nop.opcode = IT_NOP;
+	nop.type = PM4_TYPE_3;
+	nop.u32all |= PM4_COUNT_ZERO;
+
+	kq->dev = dev;
+	kq->nop_packet = nop.u32all;
+	switch (type) {
+	case KFD_QUEUE_TYPE_DIQ:
+	case KFD_QUEUE_TYPE_HIQ:
+		kq->mqd = dev->dqm->get_mqd_manager(dev->dqm,
+						KFD_MQD_TYPE_CIK_HIQ);
+		break;
+	default:
+		BUG();
+		break;
+	}
+
+	if (kq->mqd == NULL)
+		return false;
+
+	prop.doorbell_ptr = kfd_get_kernel_doorbell(dev, &prop.doorbell_off);
+
+	if (prop.doorbell_ptr == NULL)
+		goto err_get_kernel_doorbell;
+
+	retval = kfd2kgd->allocate_mem(dev->kgd,
+					queue_size,
+					PAGE_SIZE,
+					KFD_MEMPOOL_SYSTEM_WRITECOMBINE,
+					(struct kgd_mem **) &kq->pq);
+
+	if (retval != 0)
+		goto err_pq_allocate_vidmem;
+
+	kq->pq_kernel_addr = kq->pq->cpu_ptr;
+	kq->pq_gpu_addr = kq->pq->gpu_addr;
+
+	retval = kfd2kgd->allocate_mem(dev->kgd,
+					sizeof(*kq->rptr_kernel),
+					32,
+					KFD_MEMPOOL_SYSTEM_WRITECOMBINE,
+					(struct kgd_mem **) &kq->rptr_mem);
+
+	if (retval != 0)
+		goto err_rptr_allocate_vidmem;
+
+	kq->rptr_kernel = kq->rptr_mem->cpu_ptr;
+	kq->rptr_gpu_addr = kq->rptr_mem->gpu_addr;
+
+	retval = kfd2kgd->allocate_mem(dev->kgd,
+					sizeof(*kq->wptr_kernel),
+					32,
+					KFD_MEMPOOL_SYSTEM_WRITECOMBINE,
+					(struct kgd_mem **) &kq->wptr_mem);
+
+	if (retval != 0)
+		goto err_wptr_allocate_vidmem;
+
+	kq->wptr_kernel = kq->wptr_mem->cpu_ptr;
+	kq->wptr_gpu_addr = kq->wptr_mem->gpu_addr;
+
+	memset(kq->pq_kernel_addr, 0, queue_size);
+	memset(kq->rptr_kernel, 0, sizeof(*kq->rptr_kernel));
+	memset(kq->wptr_kernel, 0, sizeof(*kq->wptr_kernel));
+
+	prop.queue_size = queue_size;
+	prop.is_interop = false;
+	prop.priority = 1;
+	prop.queue_percent = 100;
+	prop.type = type;
+	prop.vmid = 0;
+	prop.queue_address = kq->pq_gpu_addr;
+	prop.read_ptr = (uint32_t *) kq->rptr_gpu_addr;
+	prop.write_ptr = (uint32_t *) kq->wptr_gpu_addr;
+
+	if (init_queue(&kq->queue, prop) != 0)
+		goto err_init_queue;
+
+	kq->queue->device = dev;
+	kq->queue->process = kfd_get_process(current);
+
+	retval = kq->mqd->init_mqd(kq->mqd, &kq->queue->mqd,
+					&kq->queue->mqd_mem_obj,
+					&kq->queue->gart_mqd_addr,
+					&kq->queue->properties);
+	if (retval != 0)
+		goto err_init_mqd;
+
+	/* assign HIQ to HQD */
+	if (type == KFD_QUEUE_TYPE_HIQ) {
+		pr_debug("assigning hiq to hqd\n");
+		kq->queue->pipe = KFD_CIK_HIQ_PIPE;
+		kq->queue->queue = KFD_CIK_HIQ_QUEUE;
+		kq->mqd->load_mqd(kq->mqd, kq->queue->mqd, kq->queue->pipe,
+					kq->queue->queue, NULL);
+	} else {
+		/* allocate fence for DIQ */
+
+		retval = kfd2kgd->allocate_mem(dev->kgd,
+					sizeof(uint32_t),
+					32,
+					KFD_MEMPOOL_SYSTEM_WRITECOMBINE,
+					(struct kgd_mem **) &kq->fence_mem_obj);
+
+		if (retval != 0)
+			goto err_alloc_fence;
+
+		kq->fence_kernel_address = kq->fence_mem_obj->cpu_ptr;
+		kq->fence_gpu_addr = kq->fence_mem_obj->gpu_addr;
+	}
+
+	print_queue(kq->queue);
+
+	return true;
+err_alloc_fence:
+err_init_mqd:
+	uninit_queue(kq->queue);
+err_init_queue:
+	kfd2kgd->free_mem(dev->kgd, (struct kgd_mem *) kq->wptr_mem);
+err_wptr_allocate_vidmem:
+	kfd2kgd->free_mem(dev->kgd, (struct kgd_mem *) kq->rptr_mem);
+err_rptr_allocate_vidmem:
+	kfd2kgd->free_mem(dev->kgd, (struct kgd_mem *) kq->pq);
+err_pq_allocate_vidmem:
+	pr_err("kfd: error init pq\n");
+	kfd_release_kernel_doorbell(dev, prop.doorbell_ptr);
+err_get_kernel_doorbell:
+	pr_err("kfd: error init doorbell");
+	return false;
+
+}
+
+static void uninitialize(struct kernel_queue *kq)
+{
+	BUG_ON(!kq);
+
+	if (kq->queue->properties.type == KFD_QUEUE_TYPE_HIQ)
+		kq->mqd->destroy_mqd(kq->mqd,
+					NULL,
+					false,
+					QUEUE_PREEMPT_DEFAULT_TIMEOUT_MS,
+					kq->queue->pipe,
+					kq->queue->queue);
+
+	kfd2kgd->free_mem(kq->dev->kgd, (struct kgd_mem *) kq->rptr_mem);
+	kfd2kgd->free_mem(kq->dev->kgd, (struct kgd_mem *) kq->wptr_mem);
+	kfd2kgd->free_mem(kq->dev->kgd, (struct kgd_mem *) kq->pq);
+	kfd_release_kernel_doorbell(kq->dev,
+					kq->queue->properties.doorbell_ptr);
+	uninit_queue(kq->queue);
+}
+
+static int acquire_packet_buffer(struct kernel_queue *kq,
+		size_t packet_size_in_dwords, unsigned int **buffer_ptr)
+{
+	size_t available_size;
+	size_t queue_size_dwords;
+	uint32_t wptr, rptr;
+	unsigned int *queue_address;
+
+	BUG_ON(!kq || !buffer_ptr);
+
+	rptr = *kq->rptr_kernel;
+	wptr = *kq->wptr_kernel;
+	queue_address = (unsigned int *)kq->pq_kernel_addr;
+	queue_size_dwords = kq->queue->properties.queue_size / sizeof(uint32_t);
+
+	pr_debug("kfd: In func %s\nrptr: %d\nwptr: %d\nqueue_address 0x%p\n",
+			__func__, rptr, wptr, queue_address);
+
+	available_size = (rptr - 1 - wptr + queue_size_dwords) %
+							queue_size_dwords;
+
+	if (packet_size_in_dwords >= queue_size_dwords ||
+			packet_size_in_dwords >= available_size) {
+		/*
+		 * make sure calling functions know
+		 * acquire_packet_buffer() failed
+		 */
+		*buffer_ptr = NULL;
+		return -ENOMEM;
+	}
+
+	if (wptr + packet_size_in_dwords >= queue_size_dwords) {
+		while (wptr > 0) {
+			queue_address[wptr] = kq->nop_packet;
+			wptr = (wptr + 1) % queue_size_dwords;
+		}
+	}
+
+	*buffer_ptr = &queue_address[wptr];
+	kq->pending_wptr = wptr + packet_size_in_dwords;
+
+	return 0;
+}
+
+static void submit_packet(struct kernel_queue *kq)
+{
+#ifdef DEBUG
+	int i;
+#endif
+
+	BUG_ON(!kq);
+
+#ifdef DEBUG
+	for (i = *kq->wptr_kernel; i < kq->pending_wptr; i++) {
+		pr_debug("0x%2X ", kq->pq_kernel_addr[i]);
+		if (i % 15 == 0)
+			pr_debug("\n");
+	}
+	pr_debug("\n");
+#endif
+
+	*kq->wptr_kernel = kq->pending_wptr;
+	write_kernel_doorbell(kq->queue->properties.doorbell_ptr,
+				kq->pending_wptr);
+}
+
+static int sync_with_hw(struct kernel_queue *kq, unsigned long timeout_ms)
+{
+	unsigned long org_timeout_ms;
+
+	BUG_ON(!kq);
+
+	org_timeout_ms = timeout_ms;
+	timeout_ms += jiffies * 1000 / HZ;
+	while (*kq->wptr_kernel != *kq->rptr_kernel) {
+		if (time_after(jiffies * 1000 / HZ, timeout_ms)) {
+			pr_err("kfd: kernel_queue %s timeout expired %lu\n",
+				__func__, org_timeout_ms);
+			pr_err("kfd: wptr: %d rptr: %d\n",
+				*kq->wptr_kernel, *kq->rptr_kernel);
+			return -ETIME;
+		}
+		schedule();
+	}
+
+	return 0;
+}
+
+static void rollback_packet(struct kernel_queue *kq)
+{
+	BUG_ON(!kq);
+	kq->pending_wptr = *kq->queue->properties.write_ptr;
+}
+
+struct kernel_queue *kernel_queue_init(struct kfd_dev *dev,
+					enum kfd_queue_type type)
+{
+	struct kernel_queue *kq;
+
+	BUG_ON(!dev);
+
+	kq = kzalloc(sizeof(struct kernel_queue), GFP_KERNEL);
+	if (!kq)
+		return NULL;
+
+	kq->initialize = initialize;
+	kq->uninitialize = uninitialize;
+	kq->acquire_packet_buffer = acquire_packet_buffer;
+	kq->submit_packet = submit_packet;
+	kq->sync_with_hw = sync_with_hw;
+	kq->rollback_packet = rollback_packet;
+
+	if (kq->initialize(kq, dev, type, KFD_KERNEL_QUEUE_SIZE) == false) {
+		pr_err("kfd: failed to init kernel queue\n");
+		kfree(kq);
+		return NULL;
+	}
+	return kq;
+}
+
+void kernel_queue_uninit(struct kernel_queue *kq)
+{
+	BUG_ON(!kq);
+
+	kq->uninitialize(kq);
+	kfree(kq);
+}
+
+static __attribute__((unused)) void test_kq(struct kfd_dev *dev)
+{
+	struct kernel_queue *kq;
+	uint32_t *buffer, i;
+	int retval;
+
+	BUG_ON(!dev);
+
+	pr_debug("kfd: starting kernel queue test\n");
+
+	kq = kernel_queue_init(dev, KFD_QUEUE_TYPE_HIQ);
+	BUG_ON(!kq);
+
+	retval = kq->acquire_packet_buffer(kq, 5, &buffer);
+	BUG_ON(retval != 0);
+	for (i = 0; i < 5; i++)
+		buffer[i] = kq->nop_packet;
+	kq->submit_packet(kq);
+	kq->sync_with_hw(kq, 1000);
+
+	pr_debug("kfd: ending kernel queue test\n");
+}
+
+
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h
new file mode 100644
index 000000000000..dcd2bdb68d44
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h
@@ -0,0 +1,69 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef KFD_KERNEL_QUEUE_H_
+#define KFD_KERNEL_QUEUE_H_
+
+#include <linux/list.h>
+#include <linux/types.h>
+#include "kfd_priv.h"
+
+struct kernel_queue {
+	/* interface */
+	bool	(*initialize)(struct kernel_queue *kq, struct kfd_dev *dev,
+			enum kfd_queue_type type, unsigned int queue_size);
+	void	(*uninitialize)(struct kernel_queue *kq);
+	int	(*acquire_packet_buffer)(struct kernel_queue *kq,
+					size_t packet_size_in_dwords,
+					unsigned int **buffer_ptr);
+
+	void	(*submit_packet)(struct kernel_queue *kq);
+	int	(*sync_with_hw)(struct kernel_queue *kq,
+				unsigned long timeout_ms);
+	void	(*rollback_packet)(struct kernel_queue *kq);
+
+	/* data */
+	struct kfd_dev		*dev;
+	struct mqd_manager	*mqd;
+	struct queue		*queue;
+	uint32_t		pending_wptr;
+	unsigned int		nop_packet;
+
+	struct kfd_mem_obj	*rptr_mem;
+	uint32_t		*rptr_kernel;
+	uint64_t		rptr_gpu_addr;
+	struct kfd_mem_obj	*wptr_mem;
+	uint32_t		*wptr_kernel;
+	uint64_t		wptr_gpu_addr;
+	struct kfd_mem_obj	*pq;
+	uint64_t		pq_gpu_addr;
+	uint32_t		*pq_kernel_addr;
+
+	struct kfd_mem_obj	*fence_mem_obj;
+	uint64_t		fence_gpu_addr;
+	void			*fence_kernel_address;
+
+	struct list_head	list;
+};
+
+#endif /* KFD_KERNEL_QUEUE_H_ */
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_module.c b/drivers/gpu/drm/amd/amdkfd/kfd_module.c
new file mode 100644
index 000000000000..a8be6df85347
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_module.c
@@ -0,0 +1,148 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#include <linux/module.h>
+#include <linux/sched.h>
+#include <linux/moduleparam.h>
+#include <linux/device.h>
+#include "kfd_priv.h"
+
+#define KFD_DRIVER_AUTHOR	"AMD Inc. and others"
+
+#define KFD_DRIVER_DESC		"Standalone HSA driver for AMD's GPUs"
+#define KFD_DRIVER_DATE		"20141113"
+#define KFD_DRIVER_MAJOR	0
+#define KFD_DRIVER_MINOR	7
+#define KFD_DRIVER_PATCHLEVEL	0
+
+const struct kfd2kgd_calls *kfd2kgd;
+static const struct kgd2kfd_calls kgd2kfd = {
+	.exit		= kgd2kfd_exit,
+	.probe		= kgd2kfd_probe,
+	.device_init	= kgd2kfd_device_init,
+	.device_exit	= kgd2kfd_device_exit,
+	.interrupt	= kgd2kfd_interrupt,
+	.suspend	= kgd2kfd_suspend,
+	.resume		= kgd2kfd_resume,
+};
+
+int sched_policy = KFD_SCHED_POLICY_HWS;
+module_param(sched_policy, int, 0444);
+MODULE_PARM_DESC(sched_policy,
+	"Kernel cmdline parameter that defines the amdkfd scheduling policy");
+
+int max_num_of_queues_per_device = KFD_MAX_NUM_OF_QUEUES_PER_DEVICE_DEFAULT;
+module_param(max_num_of_queues_per_device, int, 0444);
+MODULE_PARM_DESC(max_num_of_queues_per_device,
+	"Maximum number of supported queues per device (1 = Minimum, 4096 = default)");
+
+bool kgd2kfd_init(unsigned interface_version,
+		  const struct kfd2kgd_calls *f2g,
+		  const struct kgd2kfd_calls **g2f)
+{
+	/*
+	 * Only one interface version is supported,
+	 * no kfd/kgd version skew allowed.
+	 */
+	if (interface_version != KFD_INTERFACE_VERSION)
+		return false;
+
+	/* Protection against multiple amd kgd loads */
+	if (kfd2kgd)
+		return true;
+
+	kfd2kgd = f2g;
+	*g2f = &kgd2kfd;
+
+	return true;
+}
+EXPORT_SYMBOL(kgd2kfd_init);
+
+void kgd2kfd_exit(void)
+{
+}
+
+static int __init kfd_module_init(void)
+{
+	int err;
+
+	kfd2kgd = NULL;
+
+	/* Verify module parameters */
+	if ((sched_policy < KFD_SCHED_POLICY_HWS) ||
+		(sched_policy > KFD_SCHED_POLICY_NO_HWS)) {
+		pr_err("kfd: sched_policy has invalid value\n");
+		return -1;
+	}
+
+	/* Verify module parameters */
+	if ((max_num_of_queues_per_device < 0) ||
+		(max_num_of_queues_per_device >
+			KFD_MAX_NUM_OF_QUEUES_PER_DEVICE)) {
+		pr_err("kfd: max_num_of_queues_per_device must be between 0 to KFD_MAX_NUM_OF_QUEUES_PER_DEVICE\n");
+		return -1;
+	}
+
+	err = kfd_pasid_init();
+	if (err < 0)
+		goto err_pasid;
+
+	err = kfd_chardev_init();
+	if (err < 0)
+		goto err_ioctl;
+
+	err = kfd_topology_init();
+	if (err < 0)
+		goto err_topology;
+
+	kfd_process_create_wq();
+
+	dev_info(kfd_device, "Initialized module\n");
+
+	return 0;
+
+err_topology:
+	kfd_chardev_exit();
+err_ioctl:
+	kfd_pasid_exit();
+err_pasid:
+	return err;
+}
+
+static void __exit kfd_module_exit(void)
+{
+	kfd_process_destroy_wq();
+	kfd_topology_shutdown();
+	kfd_chardev_exit();
+	kfd_pasid_exit();
+	dev_info(kfd_device, "Removed module\n");
+}
+
+module_init(kfd_module_init);
+module_exit(kfd_module_exit);
+
+MODULE_AUTHOR(KFD_DRIVER_AUTHOR);
+MODULE_DESCRIPTION(KFD_DRIVER_DESC);
+MODULE_LICENSE("GPL and additional rights");
+MODULE_VERSION(__stringify(KFD_DRIVER_MAJOR) "."
+	       __stringify(KFD_DRIVER_MINOR) "."
+	       __stringify(KFD_DRIVER_PATCHLEVEL));
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c
new file mode 100644
index 000000000000..4c3828cf45bf
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c
@@ -0,0 +1,346 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include <linux/printk.h>
+#include <linux/slab.h>
+#include "kfd_priv.h"
+#include "kfd_mqd_manager.h"
+#include "cik_regs.h"
+#include "../../radeon/cik_reg.h"
+
+inline void busy_wait(unsigned long ms)
+{
+	while (time_before(jiffies, ms))
+		cpu_relax();
+}
+
+static inline struct cik_mqd *get_mqd(void *mqd)
+{
+	return (struct cik_mqd *)mqd;
+}
+
+static int init_mqd(struct mqd_manager *mm, void **mqd,
+		struct kfd_mem_obj **mqd_mem_obj, uint64_t *gart_addr,
+		struct queue_properties *q)
+{
+	uint64_t addr;
+	struct cik_mqd *m;
+	int retval;
+
+	BUG_ON(!mm || !q || !mqd);
+
+	pr_debug("kfd: In func %s\n", __func__);
+
+	retval = kfd2kgd->allocate_mem(mm->dev->kgd,
+					sizeof(struct cik_mqd),
+					256,
+					KFD_MEMPOOL_SYSTEM_WRITECOMBINE,
+					(struct kgd_mem **) mqd_mem_obj);
+
+	if (retval != 0)
+		return -ENOMEM;
+
+	m = (struct cik_mqd *) (*mqd_mem_obj)->cpu_ptr;
+	addr = (*mqd_mem_obj)->gpu_addr;
+
+	memset(m, 0, ALIGN(sizeof(struct cik_mqd), 256));
+
+	m->header = 0xC0310800;
+	m->compute_pipelinestat_enable = 1;
+	m->compute_static_thread_mgmt_se0 = 0xFFFFFFFF;
+	m->compute_static_thread_mgmt_se1 = 0xFFFFFFFF;
+	m->compute_static_thread_mgmt_se2 = 0xFFFFFFFF;
+	m->compute_static_thread_mgmt_se3 = 0xFFFFFFFF;
+
+	/*
+	 * Make sure to use the last queue state saved on mqd when the cp
+	 * reassigns the queue, so when queue is switched on/off (e.g over
+	 * subscription or quantum timeout) the context will be consistent
+	 */
+	m->cp_hqd_persistent_state =
+				DEFAULT_CP_HQD_PERSISTENT_STATE | PRELOAD_REQ;
+
+	m->cp_mqd_control             = MQD_CONTROL_PRIV_STATE_EN;
+	m->cp_mqd_base_addr_lo        = lower_32_bits(addr);
+	m->cp_mqd_base_addr_hi        = upper_32_bits(addr);
+
+	m->cp_hqd_ib_control = DEFAULT_MIN_IB_AVAIL_SIZE | IB_ATC_EN;
+	/* Although WinKFD writes this, I suspect it should not be necessary */
+	m->cp_hqd_ib_control = IB_ATC_EN | DEFAULT_MIN_IB_AVAIL_SIZE;
+
+	m->cp_hqd_quantum = QUANTUM_EN | QUANTUM_SCALE_1MS |
+				QUANTUM_DURATION(10);
+
+	/*
+	 * Pipe Priority
+	 * Identifies the pipe relative priority when this queue is connected
+	 * to the pipeline. The pipe priority is against the GFX pipe and HP3D.
+	 * In KFD we are using a fixed pipe priority set to CS_MEDIUM.
+	 * 0 = CS_LOW (typically below GFX)
+	 * 1 = CS_MEDIUM (typically between HP3D and GFX
+	 * 2 = CS_HIGH (typically above HP3D)
+	 */
+	m->cp_hqd_pipe_priority = 1;
+	m->cp_hqd_queue_priority = 15;
+
+	*mqd = m;
+	if (gart_addr != NULL)
+		*gart_addr = addr;
+	retval = mm->update_mqd(mm, m, q);
+
+	return retval;
+}
+
+static void uninit_mqd(struct mqd_manager *mm, void *mqd,
+			struct kfd_mem_obj *mqd_mem_obj)
+{
+	BUG_ON(!mm || !mqd);
+	kfd2kgd->free_mem(mm->dev->kgd, (struct kgd_mem *) mqd_mem_obj);
+}
+
+static int load_mqd(struct mqd_manager *mm, void *mqd, uint32_t pipe_id,
+			uint32_t queue_id, uint32_t __user *wptr)
+{
+	return kfd2kgd->hqd_load(mm->dev->kgd, mqd, pipe_id, queue_id, wptr);
+
+}
+
+static int update_mqd(struct mqd_manager *mm, void *mqd,
+			struct queue_properties *q)
+{
+	struct cik_mqd *m;
+
+	BUG_ON(!mm || !q || !mqd);
+
+	pr_debug("kfd: In func %s\n", __func__);
+
+	m = get_mqd(mqd);
+	m->cp_hqd_pq_control = DEFAULT_RPTR_BLOCK_SIZE |
+				DEFAULT_MIN_AVAIL_SIZE | PQ_ATC_EN;
+
+	/*
+	 * Calculating queue size which is log base 2 of actual queue size -1
+	 * dwords and another -1 for ffs
+	 */
+	m->cp_hqd_pq_control |= ffs(q->queue_size / sizeof(unsigned int))
+								- 1 - 1;
+	m->cp_hqd_pq_base_lo = lower_32_bits((uint64_t)q->queue_address >> 8);
+	m->cp_hqd_pq_base_hi = upper_32_bits((uint64_t)q->queue_address >> 8);
+	m->cp_hqd_pq_rptr_report_addr_lo = lower_32_bits((uint64_t)q->read_ptr);
+	m->cp_hqd_pq_rptr_report_addr_hi = upper_32_bits((uint64_t)q->read_ptr);
+	m->cp_hqd_pq_doorbell_control = DOORBELL_EN |
+					DOORBELL_OFFSET(q->doorbell_off);
+
+	m->cp_hqd_vmid = q->vmid;
+
+	if (q->format == KFD_QUEUE_FORMAT_AQL) {
+		m->cp_hqd_iq_rptr = AQL_ENABLE;
+		m->cp_hqd_pq_control |= NO_UPDATE_RPTR;
+	}
+
+	m->cp_hqd_active = 0;
+	q->is_active = false;
+	if (q->queue_size > 0 &&
+			q->queue_address != 0 &&
+			q->queue_percent > 0) {
+		m->cp_hqd_active = 1;
+		q->is_active = true;
+	}
+
+	return 0;
+}
+
+static int destroy_mqd(struct mqd_manager *mm, void *mqd,
+			enum kfd_preempt_type type,
+			unsigned int timeout, uint32_t pipe_id,
+			uint32_t queue_id)
+{
+	return kfd2kgd->hqd_destroy(mm->dev->kgd, type, timeout,
+					pipe_id, queue_id);
+}
+
+static bool is_occupied(struct mqd_manager *mm, void *mqd,
+			uint64_t queue_address,	uint32_t pipe_id,
+			uint32_t queue_id)
+{
+
+	return kfd2kgd->hqd_is_occupied(mm->dev->kgd, queue_address,
+					pipe_id, queue_id);
+
+}
+
+/*
+ * HIQ MQD Implementation, concrete implementation for HIQ MQD implementation.
+ * The HIQ queue in Kaveri is using the same MQD structure as all the user mode
+ * queues but with different initial values.
+ */
+
+static int init_mqd_hiq(struct mqd_manager *mm, void **mqd,
+		struct kfd_mem_obj **mqd_mem_obj, uint64_t *gart_addr,
+		struct queue_properties *q)
+{
+	uint64_t addr;
+	struct cik_mqd *m;
+	int retval;
+
+	BUG_ON(!mm || !q || !mqd || !mqd_mem_obj);
+
+	pr_debug("kfd: In func %s\n", __func__);
+
+	retval = kfd2kgd->allocate_mem(mm->dev->kgd,
+					sizeof(struct cik_mqd),
+					256,
+					KFD_MEMPOOL_SYSTEM_WRITECOMBINE,
+					(struct kgd_mem **) mqd_mem_obj);
+
+	if (retval != 0)
+		return -ENOMEM;
+
+	m = (struct cik_mqd *) (*mqd_mem_obj)->cpu_ptr;
+	addr = (*mqd_mem_obj)->gpu_addr;
+
+	memset(m, 0, ALIGN(sizeof(struct cik_mqd), 256));
+
+	m->header = 0xC0310800;
+	m->compute_pipelinestat_enable = 1;
+	m->compute_static_thread_mgmt_se0 = 0xFFFFFFFF;
+	m->compute_static_thread_mgmt_se1 = 0xFFFFFFFF;
+	m->compute_static_thread_mgmt_se2 = 0xFFFFFFFF;
+	m->compute_static_thread_mgmt_se3 = 0xFFFFFFFF;
+
+	m->cp_hqd_persistent_state = DEFAULT_CP_HQD_PERSISTENT_STATE |
+					PRELOAD_REQ;
+	m->cp_hqd_quantum = QUANTUM_EN | QUANTUM_SCALE_1MS |
+				QUANTUM_DURATION(10);
+
+	m->cp_mqd_control             = MQD_CONTROL_PRIV_STATE_EN;
+	m->cp_mqd_base_addr_lo        = lower_32_bits(addr);
+	m->cp_mqd_base_addr_hi        = upper_32_bits(addr);
+
+	m->cp_hqd_ib_control = DEFAULT_MIN_IB_AVAIL_SIZE;
+
+	/*
+	 * Pipe Priority
+	 * Identifies the pipe relative priority when this queue is connected
+	 * to the pipeline. The pipe priority is against the GFX pipe and HP3D.
+	 * In KFD we are using a fixed pipe priority set to CS_MEDIUM.
+	 * 0 = CS_LOW (typically below GFX)
+	 * 1 = CS_MEDIUM (typically between HP3D and GFX
+	 * 2 = CS_HIGH (typically above HP3D)
+	 */
+	m->cp_hqd_pipe_priority = 1;
+	m->cp_hqd_queue_priority = 15;
+
+	*mqd = m;
+	if (gart_addr)
+		*gart_addr = addr;
+	retval = mm->update_mqd(mm, m, q);
+
+	return retval;
+}
+
+static int update_mqd_hiq(struct mqd_manager *mm, void *mqd,
+				struct queue_properties *q)
+{
+	struct cik_mqd *m;
+
+	BUG_ON(!mm || !q || !mqd);
+
+	pr_debug("kfd: In func %s\n", __func__);
+
+	m = get_mqd(mqd);
+	m->cp_hqd_pq_control = DEFAULT_RPTR_BLOCK_SIZE |
+				DEFAULT_MIN_AVAIL_SIZE |
+				PRIV_STATE |
+				KMD_QUEUE;
+
+	/*
+	 * Calculating queue size which is log base 2 of actual queue
+	 * size -1 dwords
+	 */
+	m->cp_hqd_pq_control |= ffs(q->queue_size / sizeof(unsigned int))
+								- 1 - 1;
+	m->cp_hqd_pq_base_lo = lower_32_bits((uint64_t)q->queue_address >> 8);
+	m->cp_hqd_pq_base_hi = upper_32_bits((uint64_t)q->queue_address >> 8);
+	m->cp_hqd_pq_rptr_report_addr_lo = lower_32_bits((uint64_t)q->read_ptr);
+	m->cp_hqd_pq_rptr_report_addr_hi = upper_32_bits((uint64_t)q->read_ptr);
+	m->cp_hqd_pq_doorbell_control = DOORBELL_EN |
+					DOORBELL_OFFSET(q->doorbell_off);
+
+	m->cp_hqd_vmid = q->vmid;
+
+	m->cp_hqd_active = 0;
+	q->is_active = false;
+	if (q->queue_size > 0 &&
+			q->queue_address != 0 &&
+			q->queue_percent > 0) {
+		m->cp_hqd_active = 1;
+		q->is_active = true;
+	}
+
+	return 0;
+}
+
+struct mqd_manager *mqd_manager_init(enum KFD_MQD_TYPE type,
+					struct kfd_dev *dev)
+{
+	struct mqd_manager *mqd;
+
+	BUG_ON(!dev);
+	BUG_ON(type >= KFD_MQD_TYPE_MAX);
+
+	pr_debug("kfd: In func %s\n", __func__);
+
+	mqd = kzalloc(sizeof(struct mqd_manager), GFP_KERNEL);
+	if (!mqd)
+		return NULL;
+
+	mqd->dev = dev;
+
+	switch (type) {
+	case KFD_MQD_TYPE_CIK_CP:
+	case KFD_MQD_TYPE_CIK_COMPUTE:
+		mqd->init_mqd = init_mqd;
+		mqd->uninit_mqd = uninit_mqd;
+		mqd->load_mqd = load_mqd;
+		mqd->update_mqd = update_mqd;
+		mqd->destroy_mqd = destroy_mqd;
+		mqd->is_occupied = is_occupied;
+		break;
+	case KFD_MQD_TYPE_CIK_HIQ:
+		mqd->init_mqd = init_mqd_hiq;
+		mqd->uninit_mqd = uninit_mqd;
+		mqd->load_mqd = load_mqd;
+		mqd->update_mqd = update_mqd_hiq;
+		mqd->destroy_mqd = destroy_mqd;
+		mqd->is_occupied = is_occupied;
+		break;
+	default:
+		kfree(mqd);
+		return NULL;
+	}
+
+	return mqd;
+}
+
+/* SDMA queues should be implemented here when the cp will supports them */
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h
new file mode 100644
index 000000000000..213a71e0b6c7
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h
@@ -0,0 +1,91 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef KFD_MQD_MANAGER_H_
+#define KFD_MQD_MANAGER_H_
+
+#include "kfd_priv.h"
+
+/**
+ * struct mqd_manager
+ *
+ * @init_mqd: Allocates the mqd buffer on local gpu memory and initialize it.
+ *
+ * @load_mqd: Loads the mqd to a concrete hqd slot. Used only for no cp
+ * scheduling mode.
+ *
+ * @update_mqd: Handles a update call for the MQD
+ *
+ * @destroy_mqd: Destroys the HQD slot and by that preempt the relevant queue.
+ * Used only for no cp scheduling.
+ *
+ * @uninit_mqd: Releases the mqd buffer from local gpu memory.
+ *
+ * @is_occupied: Checks if the relevant HQD slot is occupied.
+ *
+ * @mqd_mutex: Mqd manager mutex.
+ *
+ * @dev: The kfd device structure coupled with this module.
+ *
+ * MQD stands for Memory Queue Descriptor which represents the current queue
+ * state in the memory and initiate the HQD (Hardware Queue Descriptor) state.
+ * This structure is actually a base class for the different types of MQDs
+ * structures for the variant ASICs that should be supported in the future.
+ * This base class is also contains all the MQD specific operations.
+ * Another important thing to mention is that each queue has a MQD that keeps
+ * his state (or context) after each preemption or reassignment.
+ * Basically there are a instances of the mqd manager class per MQD type per
+ * ASIC. Currently the kfd driver supports only Kaveri so there are instances
+ * per KFD_MQD_TYPE for each device.
+ *
+ */
+
+struct mqd_manager {
+	int	(*init_mqd)(struct mqd_manager *mm, void **mqd,
+			struct kfd_mem_obj **mqd_mem_obj, uint64_t *gart_addr,
+			struct queue_properties *q);
+
+	int	(*load_mqd)(struct mqd_manager *mm, void *mqd,
+				uint32_t pipe_id, uint32_t queue_id,
+				uint32_t __user *wptr);
+
+	int	(*update_mqd)(struct mqd_manager *mm, void *mqd,
+				struct queue_properties *q);
+
+	int	(*destroy_mqd)(struct mqd_manager *mm, void *mqd,
+				enum kfd_preempt_type type,
+				unsigned int timeout, uint32_t pipe_id,
+				uint32_t queue_id);
+
+	void	(*uninit_mqd)(struct mqd_manager *mm, void *mqd,
+				struct kfd_mem_obj *mqd_mem_obj);
+
+	bool	(*is_occupied)(struct mqd_manager *mm, void *mqd,
+				uint64_t queue_address,	uint32_t pipe_id,
+				uint32_t queue_id);
+
+	struct mutex	mqd_mutex;
+	struct kfd_dev	*dev;
+};
+
+#endif /* KFD_MQD_MANAGER_H_ */
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
new file mode 100644
index 000000000000..5ce9233d2004
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
@@ -0,0 +1,565 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include <linux/slab.h>
+#include <linux/mutex.h>
+#include "kfd_device_queue_manager.h"
+#include "kfd_kernel_queue.h"
+#include "kfd_priv.h"
+#include "kfd_pm4_headers.h"
+#include "kfd_pm4_opcodes.h"
+
+static inline void inc_wptr(unsigned int *wptr, unsigned int increment_bytes,
+				unsigned int buffer_size_bytes)
+{
+	unsigned int temp = *wptr + increment_bytes / sizeof(uint32_t);
+
+	BUG_ON((temp * sizeof(uint32_t)) > buffer_size_bytes);
+	*wptr = temp;
+}
+
+static unsigned int build_pm4_header(unsigned int opcode, size_t packet_size)
+{
+	union PM4_MES_TYPE_3_HEADER header;
+
+	header.u32all = 0;
+	header.opcode = opcode;
+	header.count = packet_size/sizeof(uint32_t) - 2;
+	header.type = PM4_TYPE_3;
+
+	return header.u32all;
+}
+
+static void pm_calc_rlib_size(struct packet_manager *pm,
+				unsigned int *rlib_size,
+				bool *over_subscription)
+{
+	unsigned int process_count, queue_count;
+
+	BUG_ON(!pm || !rlib_size || !over_subscription);
+
+	process_count = pm->dqm->processes_count;
+	queue_count = pm->dqm->queue_count;
+
+	/* check if there is over subscription*/
+	*over_subscription = false;
+	if ((process_count > 1) ||
+		queue_count > PIPE_PER_ME_CP_SCHEDULING * QUEUES_PER_PIPE) {
+		*over_subscription = true;
+		pr_debug("kfd: over subscribed runlist\n");
+	}
+
+	/* calculate run list ib allocation size */
+	*rlib_size = process_count * sizeof(struct pm4_map_process) +
+		     queue_count * sizeof(struct pm4_map_queues);
+
+	/*
+	 * Increase the allocation size in case we need a chained run list
+	 * when over subscription
+	 */
+	if (*over_subscription)
+		*rlib_size += sizeof(struct pm4_runlist);
+
+	pr_debug("kfd: runlist ib size %d\n", *rlib_size);
+}
+
+static int pm_allocate_runlist_ib(struct packet_manager *pm,
+				unsigned int **rl_buffer,
+				uint64_t *rl_gpu_buffer,
+				unsigned int *rl_buffer_size,
+				bool *is_over_subscription)
+{
+	int retval;
+
+	BUG_ON(!pm);
+	BUG_ON(pm->allocated == true);
+	BUG_ON(is_over_subscription == NULL);
+
+	pm_calc_rlib_size(pm, rl_buffer_size, is_over_subscription);
+
+	retval = kfd2kgd->allocate_mem(pm->dqm->dev->kgd,
+					*rl_buffer_size,
+					PAGE_SIZE,
+					KFD_MEMPOOL_SYSTEM_WRITECOMBINE,
+					(struct kgd_mem **) &pm->ib_buffer_obj);
+
+	if (retval != 0) {
+		pr_err("kfd: failed to allocate runlist IB\n");
+		return retval;
+	}
+
+	*(void **)rl_buffer = pm->ib_buffer_obj->cpu_ptr;
+	*rl_gpu_buffer = pm->ib_buffer_obj->gpu_addr;
+
+	memset(*rl_buffer, 0, *rl_buffer_size);
+	pm->allocated = true;
+	return retval;
+}
+
+static int pm_create_runlist(struct packet_manager *pm, uint32_t *buffer,
+			uint64_t ib, size_t ib_size_in_dwords, bool chain)
+{
+	struct pm4_runlist *packet;
+
+	BUG_ON(!pm || !buffer || !ib);
+
+	packet = (struct pm4_runlist *)buffer;
+
+	memset(buffer, 0, sizeof(struct pm4_runlist));
+	packet->header.u32all = build_pm4_header(IT_RUN_LIST,
+						sizeof(struct pm4_runlist));
+
+	packet->bitfields4.ib_size = ib_size_in_dwords;
+	packet->bitfields4.chain = chain ? 1 : 0;
+	packet->bitfields4.offload_polling = 0;
+	packet->bitfields4.valid = 1;
+	packet->ordinal2 = lower_32_bits(ib);
+	packet->bitfields3.ib_base_hi = upper_32_bits(ib);
+
+	return 0;
+}
+
+static int pm_create_map_process(struct packet_manager *pm, uint32_t *buffer,
+				struct qcm_process_device *qpd)
+{
+	struct pm4_map_process *packet;
+	struct queue *cur;
+	uint32_t num_queues;
+
+	BUG_ON(!pm || !buffer || !qpd);
+
+	packet = (struct pm4_map_process *)buffer;
+
+	pr_debug("kfd: In func %s\n", __func__);
+
+	memset(buffer, 0, sizeof(struct pm4_map_process));
+
+	packet->header.u32all = build_pm4_header(IT_MAP_PROCESS,
+					sizeof(struct pm4_map_process));
+	packet->bitfields2.diq_enable = (qpd->is_debug) ? 1 : 0;
+	packet->bitfields2.process_quantum = 1;
+	packet->bitfields2.pasid = qpd->pqm->process->pasid;
+	packet->bitfields3.page_table_base = qpd->page_table_base;
+	packet->bitfields10.gds_size = qpd->gds_size;
+	packet->bitfields10.num_gws = qpd->num_gws;
+	packet->bitfields10.num_oac = qpd->num_oac;
+	num_queues = 0;
+	list_for_each_entry(cur, &qpd->queues_list, list)
+		num_queues++;
+	packet->bitfields10.num_queues = num_queues;
+
+	packet->sh_mem_config = qpd->sh_mem_config;
+	packet->sh_mem_bases = qpd->sh_mem_bases;
+	packet->sh_mem_ape1_base = qpd->sh_mem_ape1_base;
+	packet->sh_mem_ape1_limit = qpd->sh_mem_ape1_limit;
+
+	packet->gds_addr_lo = lower_32_bits(qpd->gds_context_area);
+	packet->gds_addr_hi = upper_32_bits(qpd->gds_context_area);
+
+	return 0;
+}
+
+static int pm_create_map_queue(struct packet_manager *pm, uint32_t *buffer,
+				struct queue *q)
+{
+	struct pm4_map_queues *packet;
+
+	BUG_ON(!pm || !buffer || !q);
+
+	pr_debug("kfd: In func %s\n", __func__);
+
+	packet = (struct pm4_map_queues *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_map_queues));
+
+	packet->header.u32all = build_pm4_header(IT_MAP_QUEUES,
+						sizeof(struct pm4_map_queues));
+	packet->bitfields2.alloc_format =
+				alloc_format__mes_map_queues__one_per_pipe;
+	packet->bitfields2.num_queues = 1;
+	packet->bitfields2.queue_sel =
+		queue_sel__mes_map_queues__map_to_hws_determined_queue_slots;
+
+	packet->bitfields2.vidmem = (q->properties.is_interop) ?
+			vidmem__mes_map_queues__uses_video_memory :
+			vidmem__mes_map_queues__uses_no_video_memory;
+
+	switch (q->properties.type) {
+	case KFD_QUEUE_TYPE_COMPUTE:
+	case KFD_QUEUE_TYPE_DIQ:
+		packet->bitfields2.engine_sel =
+				engine_sel__mes_map_queues__compute;
+		break;
+	case KFD_QUEUE_TYPE_SDMA:
+		packet->bitfields2.engine_sel =
+				engine_sel__mes_map_queues__sdma0;
+		break;
+	default:
+		BUG();
+		break;
+	}
+
+	packet->mes_map_queues_ordinals[0].bitfields3.doorbell_offset =
+			q->properties.doorbell_off;
+
+	packet->mes_map_queues_ordinals[0].mqd_addr_lo =
+			lower_32_bits(q->gart_mqd_addr);
+
+	packet->mes_map_queues_ordinals[0].mqd_addr_hi =
+			upper_32_bits(q->gart_mqd_addr);
+
+	packet->mes_map_queues_ordinals[0].wptr_addr_lo =
+			lower_32_bits((uint64_t)q->properties.write_ptr);
+
+	packet->mes_map_queues_ordinals[0].wptr_addr_hi =
+			upper_32_bits((uint64_t)q->properties.write_ptr);
+
+	return 0;
+}
+
+static int pm_create_runlist_ib(struct packet_manager *pm,
+				struct list_head *queues,
+				uint64_t *rl_gpu_addr,
+				size_t *rl_size_bytes)
+{
+	unsigned int alloc_size_bytes;
+	unsigned int *rl_buffer, rl_wptr, i;
+	int retval, proccesses_mapped;
+	struct device_process_node *cur;
+	struct qcm_process_device *qpd;
+	struct queue *q;
+	struct kernel_queue *kq;
+	bool is_over_subscription;
+
+	BUG_ON(!pm || !queues || !rl_size_bytes || !rl_gpu_addr);
+
+	rl_wptr = retval = proccesses_mapped = 0;
+
+	retval = pm_allocate_runlist_ib(pm, &rl_buffer, rl_gpu_addr,
+				&alloc_size_bytes, &is_over_subscription);
+	if (retval != 0)
+		return retval;
+
+	*rl_size_bytes = alloc_size_bytes;
+
+	pr_debug("kfd: In func %s\n", __func__);
+	pr_debug("kfd: building runlist ib process count: %d queues count %d\n",
+		pm->dqm->processes_count, pm->dqm->queue_count);
+
+	/* build the run list ib packet */
+	list_for_each_entry(cur, queues, list) {
+		qpd = cur->qpd;
+		/* build map process packet */
+		if (proccesses_mapped >= pm->dqm->processes_count) {
+			pr_debug("kfd: not enough space left in runlist IB\n");
+			pm_release_ib(pm);
+			return -ENOMEM;
+		}
+		retval = pm_create_map_process(pm, &rl_buffer[rl_wptr], qpd);
+		if (retval != 0)
+			return retval;
+		proccesses_mapped++;
+		inc_wptr(&rl_wptr, sizeof(struct pm4_map_process),
+				alloc_size_bytes);
+
+		list_for_each_entry(kq, &qpd->priv_queue_list, list) {
+			if (kq->queue->properties.is_active != true)
+				continue;
+			retval = pm_create_map_queue(pm, &rl_buffer[rl_wptr],
+							kq->queue);
+			if (retval != 0)
+				return retval;
+			inc_wptr(&rl_wptr, sizeof(struct pm4_map_queues),
+					alloc_size_bytes);
+		}
+
+		list_for_each_entry(q, &qpd->queues_list, list) {
+			if (q->properties.is_active != true)
+				continue;
+			retval = pm_create_map_queue(pm,
+						&rl_buffer[rl_wptr], q);
+			if (retval != 0)
+				return retval;
+			inc_wptr(&rl_wptr, sizeof(struct pm4_map_queues),
+					alloc_size_bytes);
+		}
+	}
+
+	pr_debug("kfd: finished map process and queues to runlist\n");
+
+	if (is_over_subscription)
+		pm_create_runlist(pm, &rl_buffer[rl_wptr], *rl_gpu_addr,
+				alloc_size_bytes / sizeof(uint32_t), true);
+
+	for (i = 0; i < alloc_size_bytes / sizeof(uint32_t); i++)
+		pr_debug("0x%2X ", rl_buffer[i]);
+	pr_debug("\n");
+
+	return 0;
+}
+
+int pm_init(struct packet_manager *pm, struct device_queue_manager *dqm)
+{
+	BUG_ON(!dqm);
+
+	pm->dqm = dqm;
+	mutex_init(&pm->lock);
+	pm->priv_queue = kernel_queue_init(dqm->dev, KFD_QUEUE_TYPE_HIQ);
+	if (pm->priv_queue == NULL) {
+		mutex_destroy(&pm->lock);
+		return -ENOMEM;
+	}
+	pm->allocated = false;
+
+	return 0;
+}
+
+void pm_uninit(struct packet_manager *pm)
+{
+	BUG_ON(!pm);
+
+	mutex_destroy(&pm->lock);
+	kernel_queue_uninit(pm->priv_queue);
+}
+
+int pm_send_set_resources(struct packet_manager *pm,
+				struct scheduling_resources *res)
+{
+	struct pm4_set_resources *packet;
+
+	BUG_ON(!pm || !res);
+
+	pr_debug("kfd: In func %s\n", __func__);
+
+	mutex_lock(&pm->lock);
+	pm->priv_queue->acquire_packet_buffer(pm->priv_queue,
+					sizeof(*packet) / sizeof(uint32_t),
+			(unsigned int **)&packet);
+	if (packet == NULL) {
+		mutex_unlock(&pm->lock);
+		pr_err("kfd: failed to allocate buffer on kernel queue\n");
+		return -ENOMEM;
+	}
+
+	memset(packet, 0, sizeof(struct pm4_set_resources));
+	packet->header.u32all = build_pm4_header(IT_SET_RESOURCES,
+					sizeof(struct pm4_set_resources));
+
+	packet->bitfields2.queue_type =
+			queue_type__mes_set_resources__hsa_interface_queue_hiq;
+	packet->bitfields2.vmid_mask = res->vmid_mask;
+	packet->bitfields2.unmap_latency = KFD_UNMAP_LATENCY;
+	packet->bitfields7.oac_mask = res->oac_mask;
+	packet->bitfields8.gds_heap_base = res->gds_heap_base;
+	packet->bitfields8.gds_heap_size = res->gds_heap_size;
+
+	packet->gws_mask_lo = lower_32_bits(res->gws_mask);
+	packet->gws_mask_hi = upper_32_bits(res->gws_mask);
+
+	packet->queue_mask_lo = lower_32_bits(res->queue_mask);
+	packet->queue_mask_hi = upper_32_bits(res->queue_mask);
+
+	pm->priv_queue->submit_packet(pm->priv_queue);
+	pm->priv_queue->sync_with_hw(pm->priv_queue, KFD_HIQ_TIMEOUT);
+
+	mutex_unlock(&pm->lock);
+
+	return 0;
+}
+
+int pm_send_runlist(struct packet_manager *pm, struct list_head *dqm_queues)
+{
+	uint64_t rl_gpu_ib_addr;
+	uint32_t *rl_buffer;
+	size_t rl_ib_size, packet_size_dwords;
+	int retval;
+
+	BUG_ON(!pm || !dqm_queues);
+
+	retval = pm_create_runlist_ib(pm, dqm_queues, &rl_gpu_ib_addr,
+					&rl_ib_size);
+	if (retval != 0)
+		goto fail_create_runlist_ib;
+
+	pr_debug("kfd: runlist IB address: 0x%llX\n", rl_gpu_ib_addr);
+
+	packet_size_dwords = sizeof(struct pm4_runlist) / sizeof(uint32_t);
+	mutex_lock(&pm->lock);
+
+	retval = pm->priv_queue->acquire_packet_buffer(pm->priv_queue,
+					packet_size_dwords, &rl_buffer);
+	if (retval != 0)
+		goto fail_acquire_packet_buffer;
+
+	retval = pm_create_runlist(pm, rl_buffer, rl_gpu_ib_addr,
+					rl_ib_size / sizeof(uint32_t), false);
+	if (retval != 0)
+		goto fail_create_runlist;
+
+	pm->priv_queue->submit_packet(pm->priv_queue);
+	pm->priv_queue->sync_with_hw(pm->priv_queue, KFD_HIQ_TIMEOUT);
+
+	mutex_unlock(&pm->lock);
+
+	return retval;
+
+fail_create_runlist:
+	pm->priv_queue->rollback_packet(pm->priv_queue);
+fail_acquire_packet_buffer:
+	mutex_unlock(&pm->lock);
+fail_create_runlist_ib:
+	if (pm->allocated == true)
+		pm_release_ib(pm);
+	return retval;
+}
+
+int pm_send_query_status(struct packet_manager *pm, uint64_t fence_address,
+			uint32_t fence_value)
+{
+	int retval;
+	struct pm4_query_status *packet;
+
+	BUG_ON(!pm || !fence_address);
+
+	mutex_lock(&pm->lock);
+	retval = pm->priv_queue->acquire_packet_buffer(
+			pm->priv_queue,
+			sizeof(struct pm4_query_status) / sizeof(uint32_t),
+			(unsigned int **)&packet);
+	if (retval != 0)
+		goto fail_acquire_packet_buffer;
+
+	packet->header.u32all = build_pm4_header(IT_QUERY_STATUS,
+					sizeof(struct pm4_query_status));
+
+	packet->bitfields2.context_id = 0;
+	packet->bitfields2.interrupt_sel =
+			interrupt_sel__mes_query_status__completion_status;
+	packet->bitfields2.command =
+			command__mes_query_status__fence_only_after_write_ack;
+
+	packet->addr_hi = upper_32_bits((uint64_t)fence_address);
+	packet->addr_lo = lower_32_bits((uint64_t)fence_address);
+	packet->data_hi = upper_32_bits((uint64_t)fence_value);
+	packet->data_lo = lower_32_bits((uint64_t)fence_value);
+
+	pm->priv_queue->submit_packet(pm->priv_queue);
+	pm->priv_queue->sync_with_hw(pm->priv_queue, KFD_HIQ_TIMEOUT);
+	mutex_unlock(&pm->lock);
+
+	return 0;
+
+fail_acquire_packet_buffer:
+	mutex_unlock(&pm->lock);
+	return retval;
+}
+
+int pm_send_unmap_queue(struct packet_manager *pm, enum kfd_queue_type type,
+			enum kfd_preempt_type_filter mode,
+			uint32_t filter_param, bool reset,
+			unsigned int sdma_engine)
+{
+	int retval;
+	uint32_t *buffer;
+	struct pm4_unmap_queues *packet;
+
+	BUG_ON(!pm);
+
+	mutex_lock(&pm->lock);
+	retval = pm->priv_queue->acquire_packet_buffer(
+			pm->priv_queue,
+			sizeof(struct pm4_unmap_queues) / sizeof(uint32_t),
+			&buffer);
+	if (retval != 0)
+		goto err_acquire_packet_buffer;
+
+	packet = (struct pm4_unmap_queues *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_unmap_queues));
+
+	packet->header.u32all = build_pm4_header(IT_UNMAP_QUEUES,
+					sizeof(struct pm4_unmap_queues));
+	switch (type) {
+	case KFD_QUEUE_TYPE_COMPUTE:
+	case KFD_QUEUE_TYPE_DIQ:
+		packet->bitfields2.engine_sel =
+			engine_sel__mes_unmap_queues__compute;
+		break;
+	case KFD_QUEUE_TYPE_SDMA:
+		packet->bitfields2.engine_sel =
+			engine_sel__mes_unmap_queues__sdma0 + sdma_engine;
+		break;
+	default:
+		BUG();
+		break;
+	}
+
+	if (reset)
+		packet->bitfields2.action =
+				action__mes_unmap_queues__reset_queues;
+	else
+		packet->bitfields2.action =
+				action__mes_unmap_queues__preempt_queues;
+
+	switch (mode) {
+	case KFD_PREEMPT_TYPE_FILTER_SINGLE_QUEUE:
+		packet->bitfields2.queue_sel =
+				queue_sel__mes_unmap_queues__perform_request_on_specified_queues;
+		packet->bitfields2.num_queues = 1;
+		packet->bitfields3b.doorbell_offset0 = filter_param;
+		break;
+	case KFD_PREEMPT_TYPE_FILTER_BY_PASID:
+		packet->bitfields2.queue_sel =
+				queue_sel__mes_unmap_queues__perform_request_on_pasid_queues;
+		packet->bitfields3a.pasid = filter_param;
+		break;
+	case KFD_PREEMPT_TYPE_FILTER_ALL_QUEUES:
+		packet->bitfields2.queue_sel =
+				queue_sel__mes_unmap_queues__perform_request_on_all_active_queues;
+		break;
+	default:
+		BUG();
+		break;
+	};
+
+	pm->priv_queue->submit_packet(pm->priv_queue);
+	pm->priv_queue->sync_with_hw(pm->priv_queue, KFD_HIQ_TIMEOUT);
+
+	mutex_unlock(&pm->lock);
+	return 0;
+
+err_acquire_packet_buffer:
+	mutex_unlock(&pm->lock);
+	return retval;
+}
+
+void pm_release_ib(struct packet_manager *pm)
+{
+	BUG_ON(!pm);
+
+	mutex_lock(&pm->lock);
+	if (pm->allocated) {
+		kfd2kgd->free_mem(pm->dqm->dev->kgd,
+				(struct kgd_mem *) pm->ib_buffer_obj);
+		pm->allocated = false;
+	}
+	mutex_unlock(&pm->lock);
+}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_pasid.c b/drivers/gpu/drm/amd/amdkfd/kfd_pasid.c
new file mode 100644
index 000000000000..6cfe7f1f18cf
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_pasid.c
@@ -0,0 +1,96 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#include <linux/slab.h>
+#include <linux/types.h>
+#include "kfd_priv.h"
+
+static unsigned long *pasid_bitmap;
+static unsigned int pasid_limit;
+static DEFINE_MUTEX(pasid_mutex);
+
+int kfd_pasid_init(void)
+{
+	pasid_limit = KFD_MAX_NUM_OF_PROCESSES;
+
+	pasid_bitmap = kcalloc(BITS_TO_LONGS(pasid_limit), sizeof(long), GFP_KERNEL);
+	if (!pasid_bitmap)
+		return -ENOMEM;
+
+	set_bit(0, pasid_bitmap); /* PASID 0 is reserved. */
+
+	return 0;
+}
+
+void kfd_pasid_exit(void)
+{
+	kfree(pasid_bitmap);
+}
+
+bool kfd_set_pasid_limit(unsigned int new_limit)
+{
+	if (new_limit < pasid_limit) {
+		bool ok;
+
+		mutex_lock(&pasid_mutex);
+
+		/* ensure that no pasids >= new_limit are in-use */
+		ok = (find_next_bit(pasid_bitmap, pasid_limit, new_limit) ==
+								pasid_limit);
+		if (ok)
+			pasid_limit = new_limit;
+
+		mutex_unlock(&pasid_mutex);
+
+		return ok;
+	}
+
+	return true;
+}
+
+inline unsigned int kfd_get_pasid_limit(void)
+{
+	return pasid_limit;
+}
+
+unsigned int kfd_pasid_alloc(void)
+{
+	unsigned int found;
+
+	mutex_lock(&pasid_mutex);
+
+	found = find_first_zero_bit(pasid_bitmap, pasid_limit);
+	if (found == pasid_limit)
+		found = 0;
+	else
+		set_bit(found, pasid_bitmap);
+
+	mutex_unlock(&pasid_mutex);
+
+	return found;
+}
+
+void kfd_pasid_free(unsigned int pasid)
+{
+	BUG_ON(pasid == 0 || pasid >= pasid_limit);
+	clear_bit(pasid, pasid_bitmap);
+}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers.h b/drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers.h
new file mode 100644
index 000000000000..071ad5724bd2
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers.h
@@ -0,0 +1,405 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef KFD_PM4_HEADERS_H_
+#define KFD_PM4_HEADERS_H_
+
+#ifndef PM4_MES_HEADER_DEFINED
+#define PM4_MES_HEADER_DEFINED
+union PM4_MES_TYPE_3_HEADER {
+	struct {
+		uint32_t reserved1:8;	/* < reserved */
+		uint32_t opcode:8;	/* < IT opcode */
+		uint32_t count:14;	/* < number of DWORDs - 1
+					 * in the information body.
+					 */
+		uint32_t type:2;	/* < packet identifier.
+					 * It should be 3 for type 3 packets
+					 */
+	};
+	uint32_t u32all;
+};
+#endif /* PM4_MES_HEADER_DEFINED */
+
+/* --------------------MES_SET_RESOURCES-------------------- */
+
+#ifndef PM4_MES_SET_RESOURCES_DEFINED
+#define PM4_MES_SET_RESOURCES_DEFINED
+enum set_resources_queue_type_enum {
+	queue_type__mes_set_resources__kernel_interface_queue_kiq = 0,
+	queue_type__mes_set_resources__hsa_interface_queue_hiq = 1,
+	queue_type__mes_set_resources__hsa_debug_interface_queue = 4
+};
+
+struct pm4_set_resources {
+	union {
+		union PM4_MES_TYPE_3_HEADER header;	/* header */
+		uint32_t ordinal1;
+	};
+
+	union {
+		struct {
+			uint32_t vmid_mask:16;
+			uint32_t unmap_latency:8;
+			uint32_t reserved1:5;
+			enum set_resources_queue_type_enum queue_type:3;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	uint32_t queue_mask_lo;
+	uint32_t queue_mask_hi;
+	uint32_t gws_mask_lo;
+	uint32_t gws_mask_hi;
+
+	union {
+		struct {
+			uint32_t oac_mask:16;
+			uint32_t reserved2:16;
+		} bitfields7;
+		uint32_t ordinal7;
+	};
+
+	union {
+		struct {
+			uint32_t gds_heap_base:6;
+			uint32_t reserved3:5;
+			uint32_t gds_heap_size:6;
+			uint32_t reserved4:15;
+		} bitfields8;
+		uint32_t ordinal8;
+	};
+
+};
+#endif
+
+/*--------------------MES_RUN_LIST-------------------- */
+
+#ifndef PM4_MES_RUN_LIST_DEFINED
+#define PM4_MES_RUN_LIST_DEFINED
+
+struct pm4_runlist {
+	union {
+		union PM4_MES_TYPE_3_HEADER header;	/* header */
+		uint32_t ordinal1;
+	};
+
+	union {
+		struct {
+			uint32_t reserved1:2;
+			uint32_t ib_base_lo:30;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	union {
+		struct {
+			uint32_t ib_base_hi:16;
+			uint32_t reserved2:16;
+		} bitfields3;
+		uint32_t ordinal3;
+	};
+
+	union {
+		struct {
+			uint32_t ib_size:20;
+			uint32_t chain:1;
+			uint32_t offload_polling:1;
+			uint32_t reserved3:1;
+			uint32_t valid:1;
+			uint32_t reserved4:8;
+		} bitfields4;
+		uint32_t ordinal4;
+	};
+
+};
+#endif
+
+/*--------------------MES_MAP_PROCESS-------------------- */
+
+#ifndef PM4_MES_MAP_PROCESS_DEFINED
+#define PM4_MES_MAP_PROCESS_DEFINED
+
+struct pm4_map_process {
+	union {
+		union PM4_MES_TYPE_3_HEADER header;	/* header */
+		uint32_t ordinal1;
+	};
+
+	union {
+		struct {
+			uint32_t pasid:16;
+			uint32_t reserved1:8;
+			uint32_t diq_enable:1;
+			uint32_t process_quantum:7;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	union {
+		struct {
+			uint32_t page_table_base:28;
+			uint32_t reserved3:4;
+		} bitfields3;
+		uint32_t ordinal3;
+	};
+
+	uint32_t sh_mem_bases;
+	uint32_t sh_mem_ape1_base;
+	uint32_t sh_mem_ape1_limit;
+	uint32_t sh_mem_config;
+	uint32_t gds_addr_lo;
+	uint32_t gds_addr_hi;
+
+	union {
+		struct {
+			uint32_t num_gws:6;
+			uint32_t reserved4:2;
+			uint32_t num_oac:4;
+			uint32_t reserved5:4;
+			uint32_t gds_size:6;
+			uint32_t num_queues:10;
+		} bitfields10;
+		uint32_t ordinal10;
+	};
+
+};
+#endif
+
+/*--------------------MES_MAP_QUEUES--------------------*/
+
+#ifndef PM4_MES_MAP_QUEUES_DEFINED
+#define PM4_MES_MAP_QUEUES_DEFINED
+enum map_queues_queue_sel_enum {
+	queue_sel__mes_map_queues__map_to_specified_queue_slots = 0,
+	queue_sel__mes_map_queues__map_to_hws_determined_queue_slots = 1,
+	queue_sel__mes_map_queues__enable_process_queues = 2
+};
+
+enum map_queues_vidmem_enum {
+	vidmem__mes_map_queues__uses_no_video_memory = 0,
+	vidmem__mes_map_queues__uses_video_memory = 1
+};
+
+enum map_queues_alloc_format_enum {
+	alloc_format__mes_map_queues__one_per_pipe = 0,
+	alloc_format__mes_map_queues__all_on_one_pipe = 1
+};
+
+enum map_queues_engine_sel_enum {
+	engine_sel__mes_map_queues__compute = 0,
+	engine_sel__mes_map_queues__sdma0 = 2,
+	engine_sel__mes_map_queues__sdma1 = 3
+};
+
+struct pm4_map_queues {
+	union {
+		union PM4_MES_TYPE_3_HEADER header;	/* header */
+		uint32_t ordinal1;
+	};
+
+	union {
+		struct {
+			uint32_t reserved1:4;
+			enum map_queues_queue_sel_enum queue_sel:2;
+			uint32_t reserved2:2;
+			uint32_t vmid:4;
+			uint32_t reserved3:4;
+			enum map_queues_vidmem_enum vidmem:2;
+			uint32_t reserved4:6;
+			enum map_queues_alloc_format_enum alloc_format:2;
+			enum map_queues_engine_sel_enum engine_sel:3;
+			uint32_t num_queues:3;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	struct {
+		union {
+			struct {
+				uint32_t reserved5:2;
+				uint32_t doorbell_offset:21;
+				uint32_t reserved6:3;
+				uint32_t queue:6;
+			} bitfields3;
+			uint32_t ordinal3;
+		};
+
+		uint32_t mqd_addr_lo;
+		uint32_t mqd_addr_hi;
+		uint32_t wptr_addr_lo;
+		uint32_t wptr_addr_hi;
+
+	} mes_map_queues_ordinals[1];	/* 1..N of these ordinal groups */
+
+};
+#endif
+
+/*--------------------MES_QUERY_STATUS--------------------*/
+
+#ifndef PM4_MES_QUERY_STATUS_DEFINED
+#define PM4_MES_QUERY_STATUS_DEFINED
+enum query_status_interrupt_sel_enum {
+	interrupt_sel__mes_query_status__completion_status = 0,
+	interrupt_sel__mes_query_status__process_status = 1,
+	interrupt_sel__mes_query_status__queue_status = 2
+};
+
+enum query_status_command_enum {
+	command__mes_query_status__interrupt_only = 0,
+	command__mes_query_status__fence_only_immediate = 1,
+	command__mes_query_status__fence_only_after_write_ack = 2,
+	command__mes_query_status__fence_wait_for_write_ack_send_interrupt = 3
+};
+
+enum query_status_engine_sel_enum {
+	engine_sel__mes_query_status__compute = 0,
+	engine_sel__mes_query_status__sdma0_queue = 2,
+	engine_sel__mes_query_status__sdma1_queue = 3
+};
+
+struct pm4_query_status {
+	union {
+		union PM4_MES_TYPE_3_HEADER header;	/* header */
+		uint32_t ordinal1;
+	};
+
+	union {
+		struct {
+			uint32_t context_id:28;
+			enum query_status_interrupt_sel_enum interrupt_sel:2;
+			enum query_status_command_enum command:2;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	union {
+		struct {
+			uint32_t pasid:16;
+			uint32_t reserved1:16;
+		} bitfields3a;
+		struct {
+			uint32_t reserved2:2;
+			uint32_t doorbell_offset:21;
+			uint32_t reserved3:3;
+			enum query_status_engine_sel_enum engine_sel:3;
+			uint32_t reserved4:3;
+		} bitfields3b;
+		uint32_t ordinal3;
+	};
+
+	uint32_t addr_lo;
+	uint32_t addr_hi;
+	uint32_t data_lo;
+	uint32_t data_hi;
+};
+#endif
+
+/*--------------------MES_UNMAP_QUEUES--------------------*/
+
+#ifndef PM4_MES_UNMAP_QUEUES_DEFINED
+#define PM4_MES_UNMAP_QUEUES_DEFINED
+enum unmap_queues_action_enum {
+	action__mes_unmap_queues__preempt_queues = 0,
+	action__mes_unmap_queues__reset_queues = 1,
+	action__mes_unmap_queues__disable_process_queues = 2
+};
+
+enum unmap_queues_queue_sel_enum {
+	queue_sel__mes_unmap_queues__perform_request_on_specified_queues = 0,
+	queue_sel__mes_unmap_queues__perform_request_on_pasid_queues = 1,
+	queue_sel__mes_unmap_queues__perform_request_on_all_active_queues = 2
+};
+
+enum unmap_queues_engine_sel_enum {
+	engine_sel__mes_unmap_queues__compute = 0,
+	engine_sel__mes_unmap_queues__sdma0 = 2,
+	engine_sel__mes_unmap_queues__sdma1 = 3
+};
+
+struct pm4_unmap_queues {
+	union {
+		union PM4_MES_TYPE_3_HEADER header;	/* header */
+		uint32_t ordinal1;
+	};
+
+	union {
+		struct {
+			enum unmap_queues_action_enum action:2;
+			uint32_t reserved1:2;
+			enum unmap_queues_queue_sel_enum queue_sel:2;
+			uint32_t reserved2:20;
+			enum unmap_queues_engine_sel_enum engine_sel:3;
+			uint32_t num_queues:3;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	union {
+		struct {
+			uint32_t pasid:16;
+			uint32_t reserved3:16;
+		} bitfields3a;
+		struct {
+			uint32_t reserved4:2;
+			uint32_t doorbell_offset0:21;
+			uint32_t reserved5:9;
+		} bitfields3b;
+		uint32_t ordinal3;
+	};
+
+	union {
+		struct {
+			uint32_t reserved6:2;
+			uint32_t doorbell_offset1:21;
+			uint32_t reserved7:9;
+		} bitfields4;
+		uint32_t ordinal4;
+	};
+
+	union {
+		struct {
+			uint32_t reserved8:2;
+			uint32_t doorbell_offset2:21;
+			uint32_t reserved9:9;
+		} bitfields5;
+		uint32_t ordinal5;
+	};
+
+	union {
+		struct {
+			uint32_t reserved10:2;
+			uint32_t doorbell_offset3:21;
+			uint32_t reserved11:9;
+		} bitfields6;
+		uint32_t ordinal6;
+	};
+
+};
+#endif
+
+enum {
+	CACHE_FLUSH_AND_INV_TS_EVENT = 0x00000014
+};
+
+#endif /* KFD_PM4_HEADERS_H_ */
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_pm4_opcodes.h b/drivers/gpu/drm/amd/amdkfd/kfd_pm4_opcodes.h
new file mode 100644
index 000000000000..b72fa3b8c2d4
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_pm4_opcodes.h
@@ -0,0 +1,107 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+
+#ifndef KFD_PM4_OPCODES_H
+#define KFD_PM4_OPCODES_H
+
+enum it_opcode_type {
+	IT_NOP                               = 0x10,
+	IT_SET_BASE                          = 0x11,
+	IT_CLEAR_STATE                       = 0x12,
+	IT_INDEX_BUFFER_SIZE                 = 0x13,
+	IT_DISPATCH_DIRECT                   = 0x15,
+	IT_DISPATCH_INDIRECT                 = 0x16,
+	IT_ATOMIC_GDS                        = 0x1D,
+	IT_OCCLUSION_QUERY                   = 0x1F,
+	IT_SET_PREDICATION                   = 0x20,
+	IT_REG_RMW                           = 0x21,
+	IT_COND_EXEC                         = 0x22,
+	IT_PRED_EXEC                         = 0x23,
+	IT_DRAW_INDIRECT                     = 0x24,
+	IT_DRAW_INDEX_INDIRECT               = 0x25,
+	IT_INDEX_BASE                        = 0x26,
+	IT_DRAW_INDEX_2                      = 0x27,
+	IT_CONTEXT_CONTROL                   = 0x28,
+	IT_INDEX_TYPE                        = 0x2A,
+	IT_DRAW_INDIRECT_MULTI               = 0x2C,
+	IT_DRAW_INDEX_AUTO                   = 0x2D,
+	IT_NUM_INSTANCES                     = 0x2F,
+	IT_DRAW_INDEX_MULTI_AUTO             = 0x30,
+	IT_INDIRECT_BUFFER_CNST              = 0x33,
+	IT_STRMOUT_BUFFER_UPDATE             = 0x34,
+	IT_DRAW_INDEX_OFFSET_2               = 0x35,
+	IT_DRAW_PREAMBLE                     = 0x36,
+	IT_WRITE_DATA                        = 0x37,
+	IT_DRAW_INDEX_INDIRECT_MULTI         = 0x38,
+	IT_MEM_SEMAPHORE                     = 0x39,
+	IT_COPY_DW                           = 0x3B,
+	IT_WAIT_REG_MEM                      = 0x3C,
+	IT_INDIRECT_BUFFER                   = 0x3F,
+	IT_COPY_DATA                         = 0x40,
+	IT_PFP_SYNC_ME                       = 0x42,
+	IT_SURFACE_SYNC                      = 0x43,
+	IT_COND_WRITE                        = 0x45,
+	IT_EVENT_WRITE                       = 0x46,
+	IT_EVENT_WRITE_EOP                   = 0x47,
+	IT_EVENT_WRITE_EOS                   = 0x48,
+	IT_RELEASE_MEM                       = 0x49,
+	IT_PREAMBLE_CNTL                     = 0x4A,
+	IT_DMA_DATA                          = 0x50,
+	IT_ACQUIRE_MEM                       = 0x58,
+	IT_REWIND                            = 0x59,
+	IT_LOAD_UCONFIG_REG                  = 0x5E,
+	IT_LOAD_SH_REG                       = 0x5F,
+	IT_LOAD_CONFIG_REG                   = 0x60,
+	IT_LOAD_CONTEXT_REG                  = 0x61,
+	IT_SET_CONFIG_REG                    = 0x68,
+	IT_SET_CONTEXT_REG                   = 0x69,
+	IT_SET_CONTEXT_REG_INDIRECT          = 0x73,
+	IT_SET_SH_REG                        = 0x76,
+	IT_SET_SH_REG_OFFSET                 = 0x77,
+	IT_SET_QUEUE_REG                     = 0x78,
+	IT_SET_UCONFIG_REG                   = 0x79,
+	IT_SCRATCH_RAM_WRITE                 = 0x7D,
+	IT_SCRATCH_RAM_READ                  = 0x7E,
+	IT_LOAD_CONST_RAM                    = 0x80,
+	IT_WRITE_CONST_RAM                   = 0x81,
+	IT_DUMP_CONST_RAM                    = 0x83,
+	IT_INCREMENT_CE_COUNTER              = 0x84,
+	IT_INCREMENT_DE_COUNTER              = 0x85,
+	IT_WAIT_ON_CE_COUNTER                = 0x86,
+	IT_WAIT_ON_DE_COUNTER_DIFF           = 0x88,
+	IT_SWITCH_BUFFER                     = 0x8B,
+	IT_SET_RESOURCES                     = 0xA0,
+	IT_MAP_PROCESS                       = 0xA1,
+	IT_MAP_QUEUES                        = 0xA2,
+	IT_UNMAP_QUEUES                      = 0xA3,
+	IT_QUERY_STATUS                      = 0xA4,
+	IT_RUN_LIST                          = 0xA5,
+};
+
+#define PM4_TYPE_0 0
+#define PM4_TYPE_2 2
+#define PM4_TYPE_3 3
+
+#endif /* KFD_PM4_OPCODES_H */
+
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
new file mode 100644
index 000000000000..96dc10e8904a
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -0,0 +1,602 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#ifndef KFD_PRIV_H_INCLUDED
+#define KFD_PRIV_H_INCLUDED
+
+#include <linux/hashtable.h>
+#include <linux/mmu_notifier.h>
+#include <linux/mutex.h>
+#include <linux/types.h>
+#include <linux/atomic.h>
+#include <linux/workqueue.h>
+#include <linux/spinlock.h>
+#include <linux/kfd_ioctl.h>
+#include <kgd_kfd_interface.h>
+
+#define KFD_SYSFS_FILE_MODE 0444
+
+/*
+ * When working with cp scheduler we should assign the HIQ manually or via
+ * the radeon driver to a fixed hqd slot, here are the fixed HIQ hqd slot
+ * definitions for Kaveri. In Kaveri only the first ME queues participates
+ * in the cp scheduling taking that in mind we set the HIQ slot in the
+ * second ME.
+ */
+#define KFD_CIK_HIQ_PIPE 4
+#define KFD_CIK_HIQ_QUEUE 0
+
+/* GPU ID hash width in bits */
+#define KFD_GPU_ID_HASH_WIDTH 16
+
+/* Macro for allocating structures */
+#define kfd_alloc_struct(ptr_to_struct)	\
+	((typeof(ptr_to_struct)) kzalloc(sizeof(*ptr_to_struct), GFP_KERNEL))
+
+#define KFD_MAX_NUM_OF_PROCESSES 512
+#define KFD_MAX_NUM_OF_QUEUES_PER_PROCESS 1024
+
+/*
+ * Kernel module parameter to specify maximum number of supported queues per
+ * device
+ */
+extern int max_num_of_queues_per_device;
+
+#define KFD_MAX_NUM_OF_QUEUES_PER_DEVICE_DEFAULT 4096
+#define KFD_MAX_NUM_OF_QUEUES_PER_DEVICE		\
+	(KFD_MAX_NUM_OF_PROCESSES *			\
+			KFD_MAX_NUM_OF_QUEUES_PER_PROCESS)
+
+#define KFD_KERNEL_QUEUE_SIZE 2048
+
+/* Kernel module parameter to specify the scheduling policy */
+extern int sched_policy;
+
+/**
+ * enum kfd_sched_policy
+ *
+ * @KFD_SCHED_POLICY_HWS: H/W scheduling policy known as command processor (cp)
+ * scheduling. In this scheduling mode we're using the firmware code to
+ * schedule the user mode queues and kernel queues such as HIQ and DIQ.
+ * the HIQ queue is used as a special queue that dispatches the configuration
+ * to the cp and the user mode queues list that are currently running.
+ * the DIQ queue is a debugging queue that dispatches debugging commands to the
+ * firmware.
+ * in this scheduling mode user mode queues over subscription feature is
+ * enabled.
+ *
+ * @KFD_SCHED_POLICY_HWS_NO_OVERSUBSCRIPTION: The same as above but the over
+ * subscription feature disabled.
+ *
+ * @KFD_SCHED_POLICY_NO_HWS: no H/W scheduling policy is a mode which directly
+ * set the command processor registers and sets the queues "manually". This
+ * mode is used *ONLY* for debugging proposes.
+ *
+ */
+enum kfd_sched_policy {
+	KFD_SCHED_POLICY_HWS = 0,
+	KFD_SCHED_POLICY_HWS_NO_OVERSUBSCRIPTION,
+	KFD_SCHED_POLICY_NO_HWS
+};
+
+enum cache_policy {
+	cache_policy_coherent,
+	cache_policy_noncoherent
+};
+
+struct kfd_device_info {
+	unsigned int max_pasid_bits;
+	size_t ih_ring_entry_size;
+	uint16_t mqd_size_aligned;
+};
+
+struct kfd_dev {
+	struct kgd_dev *kgd;
+
+	const struct kfd_device_info *device_info;
+	struct pci_dev *pdev;
+
+	unsigned int id;		/* topology stub index */
+
+	phys_addr_t doorbell_base;	/* Start of actual doorbells used by
+					 * KFD. It is aligned for mapping
+					 * into user mode
+					 */
+	size_t doorbell_id_offset;	/* Doorbell offset (from KFD doorbell
+					 * to HW doorbell, GFX reserved some
+					 * at the start)
+					 */
+	size_t doorbell_process_limit;	/* Number of processes we have doorbell
+					 * space for.
+					 */
+	u32 __iomem *doorbell_kernel_ptr; /* This is a pointer for a doorbells
+					   * page used by kernel queue
+					   */
+
+	struct kgd2kfd_shared_resources shared_resources;
+
+	/* QCM Device instance */
+	struct device_queue_manager *dqm;
+
+	bool init_complete;
+};
+
+/* KGD2KFD callbacks */
+void kgd2kfd_exit(void);
+struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd, struct pci_dev *pdev);
+bool kgd2kfd_device_init(struct kfd_dev *kfd,
+			 const struct kgd2kfd_shared_resources *gpu_resources);
+void kgd2kfd_device_exit(struct kfd_dev *kfd);
+
+extern const struct kfd2kgd_calls *kfd2kgd;
+
+struct kfd_mem_obj {
+	void *bo;
+	uint64_t gpu_addr;
+	uint32_t *cpu_ptr;
+};
+
+enum kfd_mempool {
+	KFD_MEMPOOL_SYSTEM_CACHEABLE = 1,
+	KFD_MEMPOOL_SYSTEM_WRITECOMBINE = 2,
+	KFD_MEMPOOL_FRAMEBUFFER = 3,
+};
+
+/* Character device interface */
+int kfd_chardev_init(void);
+void kfd_chardev_exit(void);
+struct device *kfd_chardev(void);
+
+/**
+ * enum kfd_preempt_type_filter
+ *
+ * @KFD_PREEMPT_TYPE_FILTER_SINGLE_QUEUE: Preempts single queue.
+ *
+ * @KFD_PRERMPT_TYPE_FILTER_ALL_QUEUES: Preempts all queues in the
+ *						running queues list.
+ *
+ * @KFD_PRERMPT_TYPE_FILTER_BY_PASID: Preempts queues that belongs to
+ *						specific process.
+ *
+ */
+enum kfd_preempt_type_filter {
+	KFD_PREEMPT_TYPE_FILTER_SINGLE_QUEUE,
+	KFD_PREEMPT_TYPE_FILTER_ALL_QUEUES,
+	KFD_PREEMPT_TYPE_FILTER_BY_PASID
+};
+
+enum kfd_preempt_type {
+	KFD_PREEMPT_TYPE_WAVEFRONT,
+	KFD_PREEMPT_TYPE_WAVEFRONT_RESET
+};
+
+/**
+ * enum kfd_queue_type
+ *
+ * @KFD_QUEUE_TYPE_COMPUTE: Regular user mode queue type.
+ *
+ * @KFD_QUEUE_TYPE_SDMA: Sdma user mode queue type.
+ *
+ * @KFD_QUEUE_TYPE_HIQ: HIQ queue type.
+ *
+ * @KFD_QUEUE_TYPE_DIQ: DIQ queue type.
+ */
+enum kfd_queue_type  {
+	KFD_QUEUE_TYPE_COMPUTE,
+	KFD_QUEUE_TYPE_SDMA,
+	KFD_QUEUE_TYPE_HIQ,
+	KFD_QUEUE_TYPE_DIQ
+};
+
+enum kfd_queue_format {
+	KFD_QUEUE_FORMAT_PM4,
+	KFD_QUEUE_FORMAT_AQL
+};
+
+/**
+ * struct queue_properties
+ *
+ * @type: The queue type.
+ *
+ * @queue_id: Queue identifier.
+ *
+ * @queue_address: Queue ring buffer address.
+ *
+ * @queue_size: Queue ring buffer size.
+ *
+ * @priority: Defines the queue priority relative to other queues in the
+ * process.
+ * This is just an indication and HW scheduling may override the priority as
+ * necessary while keeping the relative prioritization.
+ * the priority granularity is from 0 to f which f is the highest priority.
+ * currently all queues are initialized with the highest priority.
+ *
+ * @queue_percent: This field is partially implemented and currently a zero in
+ * this field defines that the queue is non active.
+ *
+ * @read_ptr: User space address which points to the number of dwords the
+ * cp read from the ring buffer. This field updates automatically by the H/W.
+ *
+ * @write_ptr: Defines the number of dwords written to the ring buffer.
+ *
+ * @doorbell_ptr: This field aim is to notify the H/W of new packet written to
+ * the queue ring buffer. This field should be similar to write_ptr and the user
+ * should update this field after he updated the write_ptr.
+ *
+ * @doorbell_off: The doorbell offset in the doorbell pci-bar.
+ *
+ * @is_interop: Defines if this is a interop queue. Interop queue means that the
+ * queue can access both graphics and compute resources.
+ *
+ * @is_active: Defines if the queue is active or not.
+ *
+ * @vmid: If the scheduling mode is no cp scheduling the field defines the vmid
+ * of the queue.
+ *
+ * This structure represents the queue properties for each queue no matter if
+ * it's user mode or kernel mode queue.
+ *
+ */
+struct queue_properties {
+	enum kfd_queue_type type;
+	enum kfd_queue_format format;
+	unsigned int queue_id;
+	uint64_t queue_address;
+	uint64_t  queue_size;
+	uint32_t priority;
+	uint32_t queue_percent;
+	uint32_t *read_ptr;
+	uint32_t *write_ptr;
+	uint32_t __iomem *doorbell_ptr;
+	uint32_t doorbell_off;
+	bool is_interop;
+	bool is_active;
+	/* Not relevant for user mode queues in cp scheduling */
+	unsigned int vmid;
+};
+
+/**
+ * struct queue
+ *
+ * @list: Queue linked list.
+ *
+ * @mqd: The queue MQD.
+ *
+ * @mqd_mem_obj: The MQD local gpu memory object.
+ *
+ * @gart_mqd_addr: The MQD gart mc address.
+ *
+ * @properties: The queue properties.
+ *
+ * @mec: Used only in no cp scheduling mode and identifies to micro engine id
+ * that the queue should be execute on.
+ *
+ * @pipe: Used only in no cp scheduling mode and identifies the queue's pipe id.
+ *
+ * @queue: Used only in no cp scheduliong mode and identifies the queue's slot.
+ *
+ * @process: The kfd process that created this queue.
+ *
+ * @device: The kfd device that created this queue.
+ *
+ * This structure represents user mode compute queues.
+ * It contains all the necessary data to handle such queues.
+ *
+ */
+
+struct queue {
+	struct list_head list;
+	void *mqd;
+	struct kfd_mem_obj *mqd_mem_obj;
+	uint64_t gart_mqd_addr;
+	struct queue_properties properties;
+
+	uint32_t mec;
+	uint32_t pipe;
+	uint32_t queue;
+
+	struct kfd_process	*process;
+	struct kfd_dev		*device;
+};
+
+/*
+ * Please read the kfd_mqd_manager.h description.
+ */
+enum KFD_MQD_TYPE {
+	KFD_MQD_TYPE_CIK_COMPUTE = 0, /* for no cp scheduling */
+	KFD_MQD_TYPE_CIK_HIQ, /* for hiq */
+	KFD_MQD_TYPE_CIK_CP, /* for cp queues and diq */
+	KFD_MQD_TYPE_CIK_SDMA, /* for sdma queues */
+	KFD_MQD_TYPE_MAX
+};
+
+struct scheduling_resources {
+	unsigned int vmid_mask;
+	enum kfd_queue_type type;
+	uint64_t queue_mask;
+	uint64_t gws_mask;
+	uint32_t oac_mask;
+	uint32_t gds_heap_base;
+	uint32_t gds_heap_size;
+};
+
+struct process_queue_manager {
+	/* data */
+	struct kfd_process	*process;
+	unsigned int		num_concurrent_processes;
+	struct list_head	queues;
+	unsigned long		*queue_slot_bitmap;
+};
+
+struct qcm_process_device {
+	/* The Device Queue Manager that owns this data */
+	struct device_queue_manager *dqm;
+	struct process_queue_manager *pqm;
+	/* Device Queue Manager lock */
+	struct mutex *lock;
+	/* Queues list */
+	struct list_head queues_list;
+	struct list_head priv_queue_list;
+
+	unsigned int queue_count;
+	unsigned int vmid;
+	bool is_debug;
+	/*
+	 * All the memory management data should be here too
+	 */
+	uint64_t gds_context_area;
+	uint32_t sh_mem_config;
+	uint32_t sh_mem_bases;
+	uint32_t sh_mem_ape1_base;
+	uint32_t sh_mem_ape1_limit;
+	uint32_t page_table_base;
+	uint32_t gds_size;
+	uint32_t num_gws;
+	uint32_t num_oac;
+};
+
+/* Data that is per-process-per device. */
+struct kfd_process_device {
+	/*
+	 * List of all per-device data for a process.
+	 * Starts from kfd_process.per_device_data.
+	 */
+	struct list_head per_device_list;
+
+	/* The device that owns this data. */
+	struct kfd_dev *dev;
+
+
+	/* per-process-per device QCM data structure */
+	struct qcm_process_device qpd;
+
+	/*Apertures*/
+	uint64_t lds_base;
+	uint64_t lds_limit;
+	uint64_t gpuvm_base;
+	uint64_t gpuvm_limit;
+	uint64_t scratch_base;
+	uint64_t scratch_limit;
+
+	/* Is this process/pasid bound to this device? (amd_iommu_bind_pasid) */
+	bool bound;
+};
+
+#define qpd_to_pdd(x) container_of(x, struct kfd_process_device, qpd)
+
+/* Process data */
+struct kfd_process {
+	/*
+	 * kfd_process are stored in an mm_struct*->kfd_process*
+	 * hash table (kfd_processes in kfd_process.c)
+	 */
+	struct hlist_node kfd_processes;
+
+	struct mm_struct *mm;
+
+	struct mutex mutex;
+
+	/*
+	 * In any process, the thread that started main() is the lead
+	 * thread and outlives the rest.
+	 * It is here because amd_iommu_bind_pasid wants a task_struct.
+	 */
+	struct task_struct *lead_thread;
+
+	/* We want to receive a notification when the mm_struct is destroyed */
+	struct mmu_notifier mmu_notifier;
+
+	/* Use for delayed freeing of kfd_process structure */
+	struct rcu_head	rcu;
+
+	unsigned int pasid;
+
+	/*
+	 * List of kfd_process_device structures,
+	 * one for each device the process is using.
+	 */
+	struct list_head per_device_data;
+
+	struct process_queue_manager pqm;
+
+	/* The process's queues. */
+	size_t queue_array_size;
+
+	/* Size is queue_array_size, up to MAX_PROCESS_QUEUES. */
+	struct kfd_queue **queues;
+
+	unsigned long allocated_queue_bitmap[DIV_ROUND_UP(KFD_MAX_NUM_OF_QUEUES_PER_PROCESS, BITS_PER_LONG)];
+
+	/*Is the user space process 32 bit?*/
+	bool is_32bit_user_mode;
+};
+
+/**
+ * Ioctl function type.
+ *
+ * \param filep pointer to file structure.
+ * \param p amdkfd process pointer.
+ * \param data pointer to arg that was copied from user.
+ */
+typedef int amdkfd_ioctl_t(struct file *filep, struct kfd_process *p,
+				void *data);
+
+struct amdkfd_ioctl_desc {
+	unsigned int cmd;
+	int flags;
+	amdkfd_ioctl_t *func;
+	unsigned int cmd_drv;
+	const char *name;
+};
+
+void kfd_process_create_wq(void);
+void kfd_process_destroy_wq(void);
+struct kfd_process *kfd_create_process(const struct task_struct *);
+struct kfd_process *kfd_get_process(const struct task_struct *);
+
+struct kfd_process_device *kfd_bind_process_to_device(struct kfd_dev *dev,
+							struct kfd_process *p);
+void kfd_unbind_process_from_device(struct kfd_dev *dev, unsigned int pasid);
+struct kfd_process_device *kfd_get_process_device_data(struct kfd_dev *dev,
+							struct kfd_process *p,
+							int create_pdd);
+
+/* Process device data iterator */
+struct kfd_process_device *kfd_get_first_process_device_data(struct kfd_process *p);
+struct kfd_process_device *kfd_get_next_process_device_data(struct kfd_process *p,
+						struct kfd_process_device *pdd);
+bool kfd_has_process_device_data(struct kfd_process *p);
+
+/* PASIDs */
+int kfd_pasid_init(void);
+void kfd_pasid_exit(void);
+bool kfd_set_pasid_limit(unsigned int new_limit);
+unsigned int kfd_get_pasid_limit(void);
+unsigned int kfd_pasid_alloc(void);
+void kfd_pasid_free(unsigned int pasid);
+
+/* Doorbells */
+void kfd_doorbell_init(struct kfd_dev *kfd);
+int kfd_doorbell_mmap(struct kfd_process *process, struct vm_area_struct *vma);
+u32 __iomem *kfd_get_kernel_doorbell(struct kfd_dev *kfd,
+					unsigned int *doorbell_off);
+void kfd_release_kernel_doorbell(struct kfd_dev *kfd, u32 __iomem *db_addr);
+u32 read_kernel_doorbell(u32 __iomem *db);
+void write_kernel_doorbell(u32 __iomem *db, u32 value);
+unsigned int kfd_queue_id_to_doorbell(struct kfd_dev *kfd,
+					struct kfd_process *process,
+					unsigned int queue_id);
+
+extern struct device *kfd_device;
+
+/* Topology */
+int kfd_topology_init(void);
+void kfd_topology_shutdown(void);
+int kfd_topology_add_device(struct kfd_dev *gpu);
+int kfd_topology_remove_device(struct kfd_dev *gpu);
+struct kfd_dev *kfd_device_by_id(uint32_t gpu_id);
+struct kfd_dev *kfd_device_by_pci_dev(const struct pci_dev *pdev);
+struct kfd_dev *kfd_topology_enum_kfd_devices(uint8_t idx);
+
+/* Interrupts */
+void kgd2kfd_interrupt(struct kfd_dev *kfd, const void *ih_ring_entry);
+
+/* Power Management */
+void kgd2kfd_suspend(struct kfd_dev *kfd);
+int kgd2kfd_resume(struct kfd_dev *kfd);
+
+/* amdkfd Apertures */
+int kfd_init_apertures(struct kfd_process *process);
+
+/* Queue Context Management */
+inline uint32_t lower_32(uint64_t x);
+inline uint32_t upper_32(uint64_t x);
+
+int init_queue(struct queue **q, struct queue_properties properties);
+void uninit_queue(struct queue *q);
+void print_queue_properties(struct queue_properties *q);
+void print_queue(struct queue *q);
+
+struct mqd_manager *mqd_manager_init(enum KFD_MQD_TYPE type,
+					struct kfd_dev *dev);
+struct device_queue_manager *device_queue_manager_init(struct kfd_dev *dev);
+void device_queue_manager_uninit(struct device_queue_manager *dqm);
+struct kernel_queue *kernel_queue_init(struct kfd_dev *dev,
+					enum kfd_queue_type type);
+void kernel_queue_uninit(struct kernel_queue *kq);
+
+/* Process Queue Manager */
+struct process_queue_node {
+	struct queue *q;
+	struct kernel_queue *kq;
+	struct list_head process_queue_list;
+};
+
+int pqm_init(struct process_queue_manager *pqm, struct kfd_process *p);
+void pqm_uninit(struct process_queue_manager *pqm);
+int pqm_create_queue(struct process_queue_manager *pqm,
+			    struct kfd_dev *dev,
+			    struct file *f,
+			    struct queue_properties *properties,
+			    unsigned int flags,
+			    enum kfd_queue_type type,
+			    unsigned int *qid);
+int pqm_destroy_queue(struct process_queue_manager *pqm, unsigned int qid);
+int pqm_update_queue(struct process_queue_manager *pqm, unsigned int qid,
+			struct queue_properties *p);
+
+/* Packet Manager */
+
+#define KFD_HIQ_TIMEOUT (500)
+
+#define KFD_FENCE_COMPLETED (100)
+#define KFD_FENCE_INIT   (10)
+#define KFD_UNMAP_LATENCY (150)
+
+struct packet_manager {
+	struct device_queue_manager *dqm;
+	struct kernel_queue *priv_queue;
+	struct mutex lock;
+	bool allocated;
+	struct kfd_mem_obj *ib_buffer_obj;
+};
+
+int pm_init(struct packet_manager *pm, struct device_queue_manager *dqm);
+void pm_uninit(struct packet_manager *pm);
+int pm_send_set_resources(struct packet_manager *pm,
+				struct scheduling_resources *res);
+int pm_send_runlist(struct packet_manager *pm, struct list_head *dqm_queues);
+int pm_send_query_status(struct packet_manager *pm, uint64_t fence_address,
+				uint32_t fence_value);
+
+int pm_send_unmap_queue(struct packet_manager *pm, enum kfd_queue_type type,
+			enum kfd_preempt_type_filter mode,
+			uint32_t filter_param, bool reset,
+			unsigned int sdma_engine);
+
+void pm_release_ib(struct packet_manager *pm);
+
+uint64_t kfd_get_number_elems(struct kfd_dev *kfd);
+phys_addr_t kfd_get_process_doorbells(struct kfd_dev *dev,
+					struct kfd_process *process);
+
+#endif
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
new file mode 100644
index 000000000000..3c76ef05cbcf
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
@@ -0,0 +1,419 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#include <linux/mutex.h>
+#include <linux/log2.h>
+#include <linux/sched.h>
+#include <linux/slab.h>
+#include <linux/amd-iommu.h>
+#include <linux/notifier.h>
+#include <linux/compat.h>
+
+struct mm_struct;
+
+#include "kfd_priv.h"
+
+/*
+ * Initial size for the array of queues.
+ * The allocated size is doubled each time
+ * it is exceeded up to MAX_PROCESS_QUEUES.
+ */
+#define INITIAL_QUEUE_ARRAY_SIZE 16
+
+/*
+ * List of struct kfd_process (field kfd_process).
+ * Unique/indexed by mm_struct*
+ */
+#define KFD_PROCESS_TABLE_SIZE 5 /* bits: 32 entries */
+static DEFINE_HASHTABLE(kfd_processes_table, KFD_PROCESS_TABLE_SIZE);
+static DEFINE_MUTEX(kfd_processes_mutex);
+
+DEFINE_STATIC_SRCU(kfd_processes_srcu);
+
+static struct workqueue_struct *kfd_process_wq;
+
+struct kfd_process_release_work {
+	struct work_struct kfd_work;
+	struct kfd_process *p;
+};
+
+static struct kfd_process *find_process(const struct task_struct *thread);
+static struct kfd_process *create_process(const struct task_struct *thread);
+
+void kfd_process_create_wq(void)
+{
+	if (!kfd_process_wq)
+		kfd_process_wq = create_workqueue("kfd_process_wq");
+}
+
+void kfd_process_destroy_wq(void)
+{
+	if (kfd_process_wq) {
+		flush_workqueue(kfd_process_wq);
+		destroy_workqueue(kfd_process_wq);
+		kfd_process_wq = NULL;
+	}
+}
+
+struct kfd_process *kfd_create_process(const struct task_struct *thread)
+{
+	struct kfd_process *process;
+
+	BUG_ON(!kfd_process_wq);
+
+	if (thread->mm == NULL)
+		return ERR_PTR(-EINVAL);
+
+	/* Only the pthreads threading model is supported. */
+	if (thread->group_leader->mm != thread->mm)
+		return ERR_PTR(-EINVAL);
+
+	/* Take mmap_sem because we call __mmu_notifier_register inside */
+	down_write(&thread->mm->mmap_sem);
+
+	/*
+	 * take kfd processes mutex before starting of process creation
+	 * so there won't be a case where two threads of the same process
+	 * create two kfd_process structures
+	 */
+	mutex_lock(&kfd_processes_mutex);
+
+	/* A prior open of /dev/kfd could have already created the process. */
+	process = find_process(thread);
+	if (process)
+		pr_debug("kfd: process already found\n");
+
+	if (!process)
+		process = create_process(thread);
+
+	mutex_unlock(&kfd_processes_mutex);
+
+	up_write(&thread->mm->mmap_sem);
+
+	return process;
+}
+
+struct kfd_process *kfd_get_process(const struct task_struct *thread)
+{
+	struct kfd_process *process;
+
+	if (thread->mm == NULL)
+		return ERR_PTR(-EINVAL);
+
+	/* Only the pthreads threading model is supported. */
+	if (thread->group_leader->mm != thread->mm)
+		return ERR_PTR(-EINVAL);
+
+	process = find_process(thread);
+
+	return process;
+}
+
+static struct kfd_process *find_process_by_mm(const struct mm_struct *mm)
+{
+	struct kfd_process *process;
+
+	hash_for_each_possible_rcu(kfd_processes_table, process,
+					kfd_processes, (uintptr_t)mm)
+		if (process->mm == mm)
+			return process;
+
+	return NULL;
+}
+
+static struct kfd_process *find_process(const struct task_struct *thread)
+{
+	struct kfd_process *p;
+	int idx;
+
+	idx = srcu_read_lock(&kfd_processes_srcu);
+	p = find_process_by_mm(thread->mm);
+	srcu_read_unlock(&kfd_processes_srcu, idx);
+
+	return p;
+}
+
+static void kfd_process_wq_release(struct work_struct *work)
+{
+	struct kfd_process_release_work *my_work;
+	struct kfd_process_device *pdd, *temp;
+	struct kfd_process *p;
+
+	my_work = (struct kfd_process_release_work *) work;
+
+	p = my_work->p;
+
+	mutex_lock(&p->mutex);
+
+	list_for_each_entry_safe(pdd, temp, &p->per_device_data,
+							per_device_list) {
+		amd_iommu_unbind_pasid(pdd->dev->pdev, p->pasid);
+		list_del(&pdd->per_device_list);
+
+		kfree(pdd);
+	}
+
+	kfd_pasid_free(p->pasid);
+
+	mutex_unlock(&p->mutex);
+
+	mutex_destroy(&p->mutex);
+
+	kfree(p->queues);
+
+	kfree(p);
+
+	kfree((void *)work);
+}
+
+static void kfd_process_destroy_delayed(struct rcu_head *rcu)
+{
+	struct kfd_process_release_work *work;
+	struct kfd_process *p;
+
+	BUG_ON(!kfd_process_wq);
+
+	p = container_of(rcu, struct kfd_process, rcu);
+	BUG_ON(atomic_read(&p->mm->mm_count) <= 0);
+
+	mmdrop(p->mm);
+
+	work = (struct kfd_process_release_work *)
+		kmalloc(sizeof(struct kfd_process_release_work), GFP_ATOMIC);
+
+	if (work) {
+		INIT_WORK((struct work_struct *) work, kfd_process_wq_release);
+		work->p = p;
+		queue_work(kfd_process_wq, (struct work_struct *) work);
+	}
+}
+
+static void kfd_process_notifier_release(struct mmu_notifier *mn,
+					struct mm_struct *mm)
+{
+	struct kfd_process *p;
+
+	/*
+	 * The kfd_process structure can not be free because the
+	 * mmu_notifier srcu is read locked
+	 */
+	p = container_of(mn, struct kfd_process, mmu_notifier);
+	BUG_ON(p->mm != mm);
+
+	mutex_lock(&kfd_processes_mutex);
+	hash_del_rcu(&p->kfd_processes);
+	mutex_unlock(&kfd_processes_mutex);
+	synchronize_srcu(&kfd_processes_srcu);
+
+	mutex_lock(&p->mutex);
+
+	/* In case our notifier is called before IOMMU notifier */
+	pqm_uninit(&p->pqm);
+
+	mutex_unlock(&p->mutex);
+
+	/*
+	 * Because we drop mm_count inside kfd_process_destroy_delayed
+	 * and because the mmu_notifier_unregister function also drop
+	 * mm_count we need to take an extra count here.
+	 */
+	atomic_inc(&p->mm->mm_count);
+	mmu_notifier_unregister_no_release(&p->mmu_notifier, p->mm);
+	mmu_notifier_call_srcu(&p->rcu, &kfd_process_destroy_delayed);
+}
+
+static const struct mmu_notifier_ops kfd_process_mmu_notifier_ops = {
+	.release = kfd_process_notifier_release,
+};
+
+static struct kfd_process *create_process(const struct task_struct *thread)
+{
+	struct kfd_process *process;
+	int err = -ENOMEM;
+
+	process = kzalloc(sizeof(*process), GFP_KERNEL);
+
+	if (!process)
+		goto err_alloc_process;
+
+	process->queues = kmalloc_array(INITIAL_QUEUE_ARRAY_SIZE,
+					sizeof(process->queues[0]), GFP_KERNEL);
+	if (!process->queues)
+		goto err_alloc_queues;
+
+	process->pasid = kfd_pasid_alloc();
+	if (process->pasid == 0)
+		goto err_alloc_pasid;
+
+	mutex_init(&process->mutex);
+
+	process->mm = thread->mm;
+
+	/* register notifier */
+	process->mmu_notifier.ops = &kfd_process_mmu_notifier_ops;
+	err = __mmu_notifier_register(&process->mmu_notifier, process->mm);
+	if (err)
+		goto err_mmu_notifier;
+
+	hash_add_rcu(kfd_processes_table, &process->kfd_processes,
+			(uintptr_t)process->mm);
+
+	process->lead_thread = thread->group_leader;
+
+	process->queue_array_size = INITIAL_QUEUE_ARRAY_SIZE;
+
+	INIT_LIST_HEAD(&process->per_device_data);
+
+	err = pqm_init(&process->pqm, process);
+	if (err != 0)
+		goto err_process_pqm_init;
+
+	/* init process apertures*/
+	process->is_32bit_user_mode = is_compat_task();
+	if (kfd_init_apertures(process) != 0)
+		goto err_init_apretures;
+
+	return process;
+
+err_init_apretures:
+	pqm_uninit(&process->pqm);
+err_process_pqm_init:
+	hash_del_rcu(&process->kfd_processes);
+	synchronize_rcu();
+	mmu_notifier_unregister_no_release(&process->mmu_notifier, process->mm);
+err_mmu_notifier:
+	kfd_pasid_free(process->pasid);
+err_alloc_pasid:
+	kfree(process->queues);
+err_alloc_queues:
+	kfree(process);
+err_alloc_process:
+	return ERR_PTR(err);
+}
+
+struct kfd_process_device *kfd_get_process_device_data(struct kfd_dev *dev,
+							struct kfd_process *p,
+							int create_pdd)
+{
+	struct kfd_process_device *pdd = NULL;
+
+	list_for_each_entry(pdd, &p->per_device_data, per_device_list)
+		if (pdd->dev == dev)
+			return pdd;
+
+	if (create_pdd) {
+		pdd = kzalloc(sizeof(*pdd), GFP_KERNEL);
+		if (pdd != NULL) {
+			pdd->dev = dev;
+			INIT_LIST_HEAD(&pdd->qpd.queues_list);
+			INIT_LIST_HEAD(&pdd->qpd.priv_queue_list);
+			pdd->qpd.dqm = dev->dqm;
+			list_add(&pdd->per_device_list, &p->per_device_data);
+		}
+	}
+
+	return pdd;
+}
+
+/*
+ * Direct the IOMMU to bind the process (specifically the pasid->mm)
+ * to the device.
+ * Unbinding occurs when the process dies or the device is removed.
+ *
+ * Assumes that the process lock is held.
+ */
+struct kfd_process_device *kfd_bind_process_to_device(struct kfd_dev *dev,
+							struct kfd_process *p)
+{
+	struct kfd_process_device *pdd = kfd_get_process_device_data(dev, p, 1);
+	int err;
+
+	if (pdd == NULL)
+		return ERR_PTR(-ENOMEM);
+
+	if (pdd->bound)
+		return pdd;
+
+	err = amd_iommu_bind_pasid(dev->pdev, p->pasid, p->lead_thread);
+	if (err < 0)
+		return ERR_PTR(err);
+
+	pdd->bound = true;
+
+	return pdd;
+}
+
+void kfd_unbind_process_from_device(struct kfd_dev *dev, unsigned int pasid)
+{
+	struct kfd_process *p;
+	struct kfd_process_device *pdd;
+	int idx, i;
+
+	BUG_ON(dev == NULL);
+
+	idx = srcu_read_lock(&kfd_processes_srcu);
+
+	hash_for_each_rcu(kfd_processes_table, i, p, kfd_processes)
+		if (p->pasid == pasid)
+			break;
+
+	srcu_read_unlock(&kfd_processes_srcu, idx);
+
+	BUG_ON(p->pasid != pasid);
+
+	mutex_lock(&p->mutex);
+
+	pqm_uninit(&p->pqm);
+
+	pdd = kfd_get_process_device_data(dev, p, 0);
+
+	/*
+	 * Just mark pdd as unbound, because we still need it to call
+	 * amd_iommu_unbind_pasid() in when the process exits.
+	 * We don't call amd_iommu_unbind_pasid() here
+	 * because the IOMMU called us.
+	 */
+	if (pdd)
+		pdd->bound = false;
+
+	mutex_unlock(&p->mutex);
+}
+
+struct kfd_process_device *kfd_get_first_process_device_data(struct kfd_process *p)
+{
+	return list_first_entry(&p->per_device_data,
+				struct kfd_process_device,
+				per_device_list);
+}
+
+struct kfd_process_device *kfd_get_next_process_device_data(struct kfd_process *p,
+						struct kfd_process_device *pdd)
+{
+	if (list_is_last(&pdd->per_device_list, &p->per_device_data))
+		return NULL;
+	return list_next_entry(pdd, per_device_list);
+}
+
+bool kfd_has_process_device_data(struct kfd_process *p)
+{
+	return !(list_empty(&p->per_device_data));
+}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
new file mode 100644
index 000000000000..f37cf5efe642
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
@@ -0,0 +1,347 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include <linux/slab.h>
+#include <linux/list.h>
+#include "kfd_device_queue_manager.h"
+#include "kfd_priv.h"
+#include "kfd_kernel_queue.h"
+
+static inline struct process_queue_node *get_queue_by_qid(
+			struct process_queue_manager *pqm, unsigned int qid)
+{
+	struct process_queue_node *pqn;
+
+	BUG_ON(!pqm);
+
+	list_for_each_entry(pqn, &pqm->queues, process_queue_list) {
+		if (pqn->q && pqn->q->properties.queue_id == qid)
+			return pqn;
+		if (pqn->kq && pqn->kq->queue->properties.queue_id == qid)
+			return pqn;
+	}
+
+	return NULL;
+}
+
+static int find_available_queue_slot(struct process_queue_manager *pqm,
+					unsigned int *qid)
+{
+	unsigned long found;
+
+	BUG_ON(!pqm || !qid);
+
+	pr_debug("kfd: in %s\n", __func__);
+
+	found = find_first_zero_bit(pqm->queue_slot_bitmap,
+			KFD_MAX_NUM_OF_QUEUES_PER_PROCESS);
+
+	pr_debug("kfd: the new slot id %lu\n", found);
+
+	if (found >= KFD_MAX_NUM_OF_QUEUES_PER_PROCESS) {
+		pr_info("amdkfd: Can not open more queues for process with pasid %d\n",
+				pqm->process->pasid);
+		return -ENOMEM;
+	}
+
+	set_bit(found, pqm->queue_slot_bitmap);
+	*qid = found;
+
+	return 0;
+}
+
+int pqm_init(struct process_queue_manager *pqm, struct kfd_process *p)
+{
+	BUG_ON(!pqm);
+
+	INIT_LIST_HEAD(&pqm->queues);
+	pqm->queue_slot_bitmap =
+			kzalloc(DIV_ROUND_UP(KFD_MAX_NUM_OF_QUEUES_PER_PROCESS,
+					BITS_PER_BYTE), GFP_KERNEL);
+	if (pqm->queue_slot_bitmap == NULL)
+		return -ENOMEM;
+	pqm->process = p;
+
+	return 0;
+}
+
+void pqm_uninit(struct process_queue_manager *pqm)
+{
+	int retval;
+	struct process_queue_node *pqn, *next;
+
+	BUG_ON(!pqm);
+
+	pr_debug("In func %s\n", __func__);
+
+	list_for_each_entry_safe(pqn, next, &pqm->queues, process_queue_list) {
+		retval = pqm_destroy_queue(
+				pqm,
+				(pqn->q != NULL) ?
+					pqn->q->properties.queue_id :
+					pqn->kq->queue->properties.queue_id);
+
+		if (retval != 0) {
+			pr_err("kfd: failed to destroy queue\n");
+			return;
+		}
+	}
+	kfree(pqm->queue_slot_bitmap);
+	pqm->queue_slot_bitmap = NULL;
+}
+
+static int create_cp_queue(struct process_queue_manager *pqm,
+				struct kfd_dev *dev, struct queue **q,
+				struct queue_properties *q_properties,
+				struct file *f, unsigned int qid)
+{
+	int retval;
+
+	retval = 0;
+
+	/* Doorbell initialized in user space*/
+	q_properties->doorbell_ptr = NULL;
+
+	q_properties->doorbell_off =
+			kfd_queue_id_to_doorbell(dev, pqm->process, qid);
+
+	/* let DQM handle it*/
+	q_properties->vmid = 0;
+	q_properties->queue_id = qid;
+	q_properties->type = KFD_QUEUE_TYPE_COMPUTE;
+
+	retval = init_queue(q, *q_properties);
+	if (retval != 0)
+		goto err_init_queue;
+
+	(*q)->device = dev;
+	(*q)->process = pqm->process;
+
+	pr_debug("kfd: PQM After init queue");
+
+	return retval;
+
+err_init_queue:
+	return retval;
+}
+
+int pqm_create_queue(struct process_queue_manager *pqm,
+			    struct kfd_dev *dev,
+			    struct file *f,
+			    struct queue_properties *properties,
+			    unsigned int flags,
+			    enum kfd_queue_type type,
+			    unsigned int *qid)
+{
+	int retval;
+	struct kfd_process_device *pdd;
+	struct queue_properties q_properties;
+	struct queue *q;
+	struct process_queue_node *pqn;
+	struct kernel_queue *kq;
+
+	BUG_ON(!pqm || !dev || !properties || !qid);
+
+	memset(&q_properties, 0, sizeof(struct queue_properties));
+	memcpy(&q_properties, properties, sizeof(struct queue_properties));
+	q = NULL;
+	kq = NULL;
+
+	pdd = kfd_get_process_device_data(dev, pqm->process, 1);
+	BUG_ON(!pdd);
+
+	retval = find_available_queue_slot(pqm, qid);
+	if (retval != 0)
+		return retval;
+
+	if (list_empty(&pqm->queues)) {
+		pdd->qpd.pqm = pqm;
+		dev->dqm->register_process(dev->dqm, &pdd->qpd);
+	}
+
+	pqn = kzalloc(sizeof(struct process_queue_node), GFP_KERNEL);
+	if (!pqn) {
+		retval = -ENOMEM;
+		goto err_allocate_pqn;
+	}
+
+	switch (type) {
+	case KFD_QUEUE_TYPE_COMPUTE:
+		/* check if there is over subscription */
+		if ((sched_policy == KFD_SCHED_POLICY_HWS_NO_OVERSUBSCRIPTION) &&
+		((dev->dqm->processes_count >= VMID_PER_DEVICE) ||
+		(dev->dqm->queue_count >= PIPE_PER_ME_CP_SCHEDULING * QUEUES_PER_PIPE))) {
+			pr_err("kfd: over-subscription is not allowed in radeon_kfd.sched_policy == 1\n");
+			retval = -EPERM;
+			goto err_create_queue;
+		}
+
+		retval = create_cp_queue(pqm, dev, &q, &q_properties, f, *qid);
+		if (retval != 0)
+			goto err_create_queue;
+		pqn->q = q;
+		pqn->kq = NULL;
+		retval = dev->dqm->create_queue(dev->dqm, q, &pdd->qpd,
+						&q->properties.vmid);
+		pr_debug("DQM returned %d for create_queue\n", retval);
+		print_queue(q);
+		break;
+	case KFD_QUEUE_TYPE_DIQ:
+		kq = kernel_queue_init(dev, KFD_QUEUE_TYPE_DIQ);
+		if (kq == NULL) {
+			retval = -ENOMEM;
+			goto err_create_queue;
+		}
+		kq->queue->properties.queue_id = *qid;
+		pqn->kq = kq;
+		pqn->q = NULL;
+		retval = dev->dqm->create_kernel_queue(dev->dqm, kq, &pdd->qpd);
+		break;
+	default:
+		BUG();
+		break;
+	}
+
+	if (retval != 0) {
+		pr_debug("Error dqm create queue\n");
+		goto err_create_queue;
+	}
+
+	pr_debug("kfd: PQM After DQM create queue\n");
+
+	list_add(&pqn->process_queue_list, &pqm->queues);
+
+	if (q) {
+		*properties = q->properties;
+		pr_debug("kfd: PQM done creating queue\n");
+		print_queue_properties(properties);
+	}
+
+	return retval;
+
+err_create_queue:
+	kfree(pqn);
+err_allocate_pqn:
+	/* check if queues list is empty unregister process from device */
+	clear_bit(*qid, pqm->queue_slot_bitmap);
+	if (list_empty(&pqm->queues))
+		dev->dqm->unregister_process(dev->dqm, &pdd->qpd);
+	return retval;
+}
+
+int pqm_destroy_queue(struct process_queue_manager *pqm, unsigned int qid)
+{
+	struct process_queue_node *pqn;
+	struct kfd_process_device *pdd;
+	struct device_queue_manager *dqm;
+	struct kfd_dev *dev;
+	int retval;
+
+	dqm = NULL;
+
+	BUG_ON(!pqm);
+	retval = 0;
+
+	pr_debug("kfd: In Func %s\n", __func__);
+
+	pqn = get_queue_by_qid(pqm, qid);
+	if (pqn == NULL) {
+		pr_err("kfd: queue id does not match any known queue\n");
+		return -EINVAL;
+	}
+
+	dev = NULL;
+	if (pqn->kq)
+		dev = pqn->kq->dev;
+	if (pqn->q)
+		dev = pqn->q->device;
+	BUG_ON(!dev);
+
+	pdd = kfd_get_process_device_data(dev, pqm->process, 1);
+	BUG_ON(!pdd);
+
+	if (pqn->kq) {
+		/* destroy kernel queue (DIQ) */
+		dqm = pqn->kq->dev->dqm;
+		dqm->destroy_kernel_queue(dqm, pqn->kq, &pdd->qpd);
+		kernel_queue_uninit(pqn->kq);
+	}
+
+	if (pqn->q) {
+		dqm = pqn->q->device->dqm;
+		retval = dqm->destroy_queue(dqm, &pdd->qpd, pqn->q);
+		if (retval != 0)
+			return retval;
+
+		uninit_queue(pqn->q);
+	}
+
+	list_del(&pqn->process_queue_list);
+	kfree(pqn);
+	clear_bit(qid, pqm->queue_slot_bitmap);
+
+	if (list_empty(&pqm->queues))
+		dqm->unregister_process(dqm, &pdd->qpd);
+
+	return retval;
+}
+
+int pqm_update_queue(struct process_queue_manager *pqm, unsigned int qid,
+			struct queue_properties *p)
+{
+	int retval;
+	struct process_queue_node *pqn;
+
+	BUG_ON(!pqm);
+
+	pqn = get_queue_by_qid(pqm, qid);
+	BUG_ON(!pqn);
+
+	pqn->q->properties.queue_address = p->queue_address;
+	pqn->q->properties.queue_size = p->queue_size;
+	pqn->q->properties.queue_percent = p->queue_percent;
+	pqn->q->properties.priority = p->priority;
+
+	retval = pqn->q->device->dqm->update_queue(pqn->q->device->dqm, pqn->q);
+	if (retval != 0)
+		return retval;
+
+	return 0;
+}
+
+static __attribute__((unused)) struct kernel_queue *pqm_get_kernel_queue(
+					struct process_queue_manager *pqm,
+					unsigned int qid)
+{
+	struct process_queue_node *pqn;
+
+	BUG_ON(!pqm);
+
+	pqn = get_queue_by_qid(pqm, qid);
+	if (pqn && pqn->kq)
+		return pqn->kq;
+
+	return NULL;
+}
+
+
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_queue.c b/drivers/gpu/drm/amd/amdkfd/kfd_queue.c
new file mode 100644
index 000000000000..9a0c90b0702e
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_queue.c
@@ -0,0 +1,85 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include <linux/slab.h>
+#include "kfd_priv.h"
+
+void print_queue_properties(struct queue_properties *q)
+{
+	if (!q)
+		return;
+
+	pr_debug("Printing queue properties:\n");
+	pr_debug("Queue Type: %u\n", q->type);
+	pr_debug("Queue Size: %llu\n", q->queue_size);
+	pr_debug("Queue percent: %u\n", q->queue_percent);
+	pr_debug("Queue Address: 0x%llX\n", q->queue_address);
+	pr_debug("Queue Id: %u\n", q->queue_id);
+	pr_debug("Queue Process Vmid: %u\n", q->vmid);
+	pr_debug("Queue Read Pointer: 0x%p\n", q->read_ptr);
+	pr_debug("Queue Write Pointer: 0x%p\n", q->write_ptr);
+	pr_debug("Queue Doorbell Pointer: 0x%p\n", q->doorbell_ptr);
+	pr_debug("Queue Doorbell Offset: %u\n", q->doorbell_off);
+}
+
+void print_queue(struct queue *q)
+{
+	if (!q)
+		return;
+	pr_debug("Printing queue:\n");
+	pr_debug("Queue Type: %u\n", q->properties.type);
+	pr_debug("Queue Size: %llu\n", q->properties.queue_size);
+	pr_debug("Queue percent: %u\n", q->properties.queue_percent);
+	pr_debug("Queue Address: 0x%llX\n", q->properties.queue_address);
+	pr_debug("Queue Id: %u\n", q->properties.queue_id);
+	pr_debug("Queue Process Vmid: %u\n", q->properties.vmid);
+	pr_debug("Queue Read Pointer: 0x%p\n", q->properties.read_ptr);
+	pr_debug("Queue Write Pointer: 0x%p\n", q->properties.write_ptr);
+	pr_debug("Queue Doorbell Pointer: 0x%p\n", q->properties.doorbell_ptr);
+	pr_debug("Queue Doorbell Offset: %u\n", q->properties.doorbell_off);
+	pr_debug("Queue MQD Address: 0x%p\n", q->mqd);
+	pr_debug("Queue MQD Gart: 0x%llX\n", q->gart_mqd_addr);
+	pr_debug("Queue Process Address: 0x%p\n", q->process);
+	pr_debug("Queue Device Address: 0x%p\n", q->device);
+}
+
+int init_queue(struct queue **q, struct queue_properties properties)
+{
+	struct queue *tmp;
+
+	BUG_ON(!q);
+
+	tmp = kzalloc(sizeof(struct queue), GFP_KERNEL);
+	if (!tmp)
+		return -ENOMEM;
+
+	memcpy(&tmp->properties, &properties, sizeof(struct queue_properties));
+
+	*q = tmp;
+	return 0;
+}
+
+void uninit_queue(struct queue *q)
+{
+	kfree(q);
+}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
new file mode 100644
index 000000000000..cca1708fd811
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
@@ -0,0 +1,1239 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/pci.h>
+#include <linux/errno.h>
+#include <linux/acpi.h>
+#include <linux/hash.h>
+#include <linux/cpufreq.h>
+
+#include "kfd_priv.h"
+#include "kfd_crat.h"
+#include "kfd_topology.h"
+
+static struct list_head topology_device_list;
+static int topology_crat_parsed;
+static struct kfd_system_properties sys_props;
+
+static DECLARE_RWSEM(topology_lock);
+
+struct kfd_dev *kfd_device_by_id(uint32_t gpu_id)
+{
+	struct kfd_topology_device *top_dev;
+	struct kfd_dev *device = NULL;
+
+	down_read(&topology_lock);
+
+	list_for_each_entry(top_dev, &topology_device_list, list)
+		if (top_dev->gpu_id == gpu_id) {
+			device = top_dev->gpu;
+			break;
+		}
+
+	up_read(&topology_lock);
+
+	return device;
+}
+
+struct kfd_dev *kfd_device_by_pci_dev(const struct pci_dev *pdev)
+{
+	struct kfd_topology_device *top_dev;
+	struct kfd_dev *device = NULL;
+
+	down_read(&topology_lock);
+
+	list_for_each_entry(top_dev, &topology_device_list, list)
+		if (top_dev->gpu->pdev == pdev) {
+			device = top_dev->gpu;
+			break;
+		}
+
+	up_read(&topology_lock);
+
+	return device;
+}
+
+static int kfd_topology_get_crat_acpi(void *crat_image, size_t *size)
+{
+	struct acpi_table_header *crat_table;
+	acpi_status status;
+
+	if (!size)
+		return -EINVAL;
+
+	/*
+	 * Fetch the CRAT table from ACPI
+	 */
+	status = acpi_get_table(CRAT_SIGNATURE, 0, &crat_table);
+	if (status == AE_NOT_FOUND) {
+		pr_warn("CRAT table not found\n");
+		return -ENODATA;
+	} else if (ACPI_FAILURE(status)) {
+		const char *err = acpi_format_exception(status);
+
+		pr_err("CRAT table error: %s\n", err);
+		return -EINVAL;
+	}
+
+	if (*size >= crat_table->length && crat_image != NULL)
+		memcpy(crat_image, crat_table, crat_table->length);
+
+	*size = crat_table->length;
+
+	return 0;
+}
+
+static void kfd_populated_cu_info_cpu(struct kfd_topology_device *dev,
+		struct crat_subtype_computeunit *cu)
+{
+	BUG_ON(!dev);
+	BUG_ON(!cu);
+
+	dev->node_props.cpu_cores_count = cu->num_cpu_cores;
+	dev->node_props.cpu_core_id_base = cu->processor_id_low;
+	if (cu->hsa_capability & CRAT_CU_FLAGS_IOMMU_PRESENT)
+		dev->node_props.capability |= HSA_CAP_ATS_PRESENT;
+
+	pr_info("CU CPU: cores=%d id_base=%d\n", cu->num_cpu_cores,
+			cu->processor_id_low);
+}
+
+static void kfd_populated_cu_info_gpu(struct kfd_topology_device *dev,
+		struct crat_subtype_computeunit *cu)
+{
+	BUG_ON(!dev);
+	BUG_ON(!cu);
+
+	dev->node_props.simd_id_base = cu->processor_id_low;
+	dev->node_props.simd_count = cu->num_simd_cores;
+	dev->node_props.lds_size_in_kb = cu->lds_size_in_kb;
+	dev->node_props.max_waves_per_simd = cu->max_waves_simd;
+	dev->node_props.wave_front_size = cu->wave_front_size;
+	dev->node_props.mem_banks_count = cu->num_banks;
+	dev->node_props.array_count = cu->num_arrays;
+	dev->node_props.cu_per_simd_array = cu->num_cu_per_array;
+	dev->node_props.simd_per_cu = cu->num_simd_per_cu;
+	dev->node_props.max_slots_scratch_cu = cu->max_slots_scatch_cu;
+	if (cu->hsa_capability & CRAT_CU_FLAGS_HOT_PLUGGABLE)
+		dev->node_props.capability |= HSA_CAP_HOT_PLUGGABLE;
+	pr_info("CU GPU: simds=%d id_base=%d\n", cu->num_simd_cores,
+				cu->processor_id_low);
+}
+
+/* kfd_parse_subtype_cu is called when the topology mutex is already acquired */
+static int kfd_parse_subtype_cu(struct crat_subtype_computeunit *cu)
+{
+	struct kfd_topology_device *dev;
+	int i = 0;
+
+	BUG_ON(!cu);
+
+	pr_info("Found CU entry in CRAT table with proximity_domain=%d caps=%x\n",
+			cu->proximity_domain, cu->hsa_capability);
+	list_for_each_entry(dev, &topology_device_list, list) {
+		if (cu->proximity_domain == i) {
+			if (cu->flags & CRAT_CU_FLAGS_CPU_PRESENT)
+				kfd_populated_cu_info_cpu(dev, cu);
+
+			if (cu->flags & CRAT_CU_FLAGS_GPU_PRESENT)
+				kfd_populated_cu_info_gpu(dev, cu);
+			break;
+		}
+		i++;
+	}
+
+	return 0;
+}
+
+/*
+ * kfd_parse_subtype_mem is called when the topology mutex is
+ * already acquired
+ */
+static int kfd_parse_subtype_mem(struct crat_subtype_memory *mem)
+{
+	struct kfd_mem_properties *props;
+	struct kfd_topology_device *dev;
+	int i = 0;
+
+	BUG_ON(!mem);
+
+	pr_info("Found memory entry in CRAT table with proximity_domain=%d\n",
+			mem->promixity_domain);
+	list_for_each_entry(dev, &topology_device_list, list) {
+		if (mem->promixity_domain == i) {
+			props = kfd_alloc_struct(props);
+			if (props == NULL)
+				return -ENOMEM;
+
+			if (dev->node_props.cpu_cores_count == 0)
+				props->heap_type = HSA_MEM_HEAP_TYPE_FB_PRIVATE;
+			else
+				props->heap_type = HSA_MEM_HEAP_TYPE_SYSTEM;
+
+			if (mem->flags & CRAT_MEM_FLAGS_HOT_PLUGGABLE)
+				props->flags |= HSA_MEM_FLAGS_HOT_PLUGGABLE;
+			if (mem->flags & CRAT_MEM_FLAGS_NON_VOLATILE)
+				props->flags |= HSA_MEM_FLAGS_NON_VOLATILE;
+
+			props->size_in_bytes =
+				((uint64_t)mem->length_high << 32) +
+							mem->length_low;
+			props->width = mem->width;
+
+			dev->mem_bank_count++;
+			list_add_tail(&props->list, &dev->mem_props);
+
+			break;
+		}
+		i++;
+	}
+
+	return 0;
+}
+
+/*
+ * kfd_parse_subtype_cache is called when the topology mutex
+ * is already acquired
+ */
+static int kfd_parse_subtype_cache(struct crat_subtype_cache *cache)
+{
+	struct kfd_cache_properties *props;
+	struct kfd_topology_device *dev;
+	uint32_t id;
+
+	BUG_ON(!cache);
+
+	id = cache->processor_id_low;
+
+	pr_info("Found cache entry in CRAT table with processor_id=%d\n", id);
+	list_for_each_entry(dev, &topology_device_list, list)
+		if (id == dev->node_props.cpu_core_id_base ||
+		    id == dev->node_props.simd_id_base) {
+			props = kfd_alloc_struct(props);
+			if (props == NULL)
+				return -ENOMEM;
+
+			props->processor_id_low = id;
+			props->cache_level = cache->cache_level;
+			props->cache_size = cache->cache_size;
+			props->cacheline_size = cache->cache_line_size;
+			props->cachelines_per_tag = cache->lines_per_tag;
+			props->cache_assoc = cache->associativity;
+			props->cache_latency = cache->cache_latency;
+
+			if (cache->flags & CRAT_CACHE_FLAGS_DATA_CACHE)
+				props->cache_type |= HSA_CACHE_TYPE_DATA;
+			if (cache->flags & CRAT_CACHE_FLAGS_INST_CACHE)
+				props->cache_type |= HSA_CACHE_TYPE_INSTRUCTION;
+			if (cache->flags & CRAT_CACHE_FLAGS_CPU_CACHE)
+				props->cache_type |= HSA_CACHE_TYPE_CPU;
+			if (cache->flags & CRAT_CACHE_FLAGS_SIMD_CACHE)
+				props->cache_type |= HSA_CACHE_TYPE_HSACU;
+
+			dev->cache_count++;
+			dev->node_props.caches_count++;
+			list_add_tail(&props->list, &dev->cache_props);
+
+			break;
+		}
+
+	return 0;
+}
+
+/*
+ * kfd_parse_subtype_iolink is called when the topology mutex
+ * is already acquired
+ */
+static int kfd_parse_subtype_iolink(struct crat_subtype_iolink *iolink)
+{
+	struct kfd_iolink_properties *props;
+	struct kfd_topology_device *dev;
+	uint32_t i = 0;
+	uint32_t id_from;
+	uint32_t id_to;
+
+	BUG_ON(!iolink);
+
+	id_from = iolink->proximity_domain_from;
+	id_to = iolink->proximity_domain_to;
+
+	pr_info("Found IO link entry in CRAT table with id_from=%d\n", id_from);
+	list_for_each_entry(dev, &topology_device_list, list) {
+		if (id_from == i) {
+			props = kfd_alloc_struct(props);
+			if (props == NULL)
+				return -ENOMEM;
+
+			props->node_from = id_from;
+			props->node_to = id_to;
+			props->ver_maj = iolink->version_major;
+			props->ver_min = iolink->version_minor;
+
+			/*
+			 * weight factor (derived from CDIR), currently always 1
+			 */
+			props->weight = 1;
+
+			props->min_latency = iolink->minimum_latency;
+			props->max_latency = iolink->maximum_latency;
+			props->min_bandwidth = iolink->minimum_bandwidth_mbs;
+			props->max_bandwidth = iolink->maximum_bandwidth_mbs;
+			props->rec_transfer_size =
+					iolink->recommended_transfer_size;
+
+			dev->io_link_count++;
+			dev->node_props.io_links_count++;
+			list_add_tail(&props->list, &dev->io_link_props);
+
+			break;
+		}
+		i++;
+	}
+
+	return 0;
+}
+
+static int kfd_parse_subtype(struct crat_subtype_generic *sub_type_hdr)
+{
+	struct crat_subtype_computeunit *cu;
+	struct crat_subtype_memory *mem;
+	struct crat_subtype_cache *cache;
+	struct crat_subtype_iolink *iolink;
+	int ret = 0;
+
+	BUG_ON(!sub_type_hdr);
+
+	switch (sub_type_hdr->type) {
+	case CRAT_SUBTYPE_COMPUTEUNIT_AFFINITY:
+		cu = (struct crat_subtype_computeunit *)sub_type_hdr;
+		ret = kfd_parse_subtype_cu(cu);
+		break;
+	case CRAT_SUBTYPE_MEMORY_AFFINITY:
+		mem = (struct crat_subtype_memory *)sub_type_hdr;
+		ret = kfd_parse_subtype_mem(mem);
+		break;
+	case CRAT_SUBTYPE_CACHE_AFFINITY:
+		cache = (struct crat_subtype_cache *)sub_type_hdr;
+		ret = kfd_parse_subtype_cache(cache);
+		break;
+	case CRAT_SUBTYPE_TLB_AFFINITY:
+		/*
+		 * For now, nothing to do here
+		 */
+		pr_info("Found TLB entry in CRAT table (not processing)\n");
+		break;
+	case CRAT_SUBTYPE_CCOMPUTE_AFFINITY:
+		/*
+		 * For now, nothing to do here
+		 */
+		pr_info("Found CCOMPUTE entry in CRAT table (not processing)\n");
+		break;
+	case CRAT_SUBTYPE_IOLINK_AFFINITY:
+		iolink = (struct crat_subtype_iolink *)sub_type_hdr;
+		ret = kfd_parse_subtype_iolink(iolink);
+		break;
+	default:
+		pr_warn("Unknown subtype (%d) in CRAT\n",
+				sub_type_hdr->type);
+	}
+
+	return ret;
+}
+
+static void kfd_release_topology_device(struct kfd_topology_device *dev)
+{
+	struct kfd_mem_properties *mem;
+	struct kfd_cache_properties *cache;
+	struct kfd_iolink_properties *iolink;
+
+	BUG_ON(!dev);
+
+	list_del(&dev->list);
+
+	while (dev->mem_props.next != &dev->mem_props) {
+		mem = container_of(dev->mem_props.next,
+				struct kfd_mem_properties, list);
+		list_del(&mem->list);
+		kfree(mem);
+	}
+
+	while (dev->cache_props.next != &dev->cache_props) {
+		cache = container_of(dev->cache_props.next,
+				struct kfd_cache_properties, list);
+		list_del(&cache->list);
+		kfree(cache);
+	}
+
+	while (dev->io_link_props.next != &dev->io_link_props) {
+		iolink = container_of(dev->io_link_props.next,
+				struct kfd_iolink_properties, list);
+		list_del(&iolink->list);
+		kfree(iolink);
+	}
+
+	kfree(dev);
+
+	sys_props.num_devices--;
+}
+
+static void kfd_release_live_view(void)
+{
+	struct kfd_topology_device *dev;
+
+	while (topology_device_list.next != &topology_device_list) {
+		dev = container_of(topology_device_list.next,
+				 struct kfd_topology_device, list);
+		kfd_release_topology_device(dev);
+}
+
+	memset(&sys_props, 0, sizeof(sys_props));
+}
+
+static struct kfd_topology_device *kfd_create_topology_device(void)
+{
+	struct kfd_topology_device *dev;
+
+	dev = kfd_alloc_struct(dev);
+	if (dev == NULL) {
+		pr_err("No memory to allocate a topology device");
+		return NULL;
+	}
+
+	INIT_LIST_HEAD(&dev->mem_props);
+	INIT_LIST_HEAD(&dev->cache_props);
+	INIT_LIST_HEAD(&dev->io_link_props);
+
+	list_add_tail(&dev->list, &topology_device_list);
+	sys_props.num_devices++;
+
+	return dev;
+}
+
+static int kfd_parse_crat_table(void *crat_image)
+{
+	struct kfd_topology_device *top_dev;
+	struct crat_subtype_generic *sub_type_hdr;
+	uint16_t node_id;
+	int ret;
+	struct crat_header *crat_table = (struct crat_header *)crat_image;
+	uint16_t num_nodes;
+	uint32_t image_len;
+
+	if (!crat_image)
+		return -EINVAL;
+
+	num_nodes = crat_table->num_domains;
+	image_len = crat_table->length;
+
+	pr_info("Parsing CRAT table with %d nodes\n", num_nodes);
+
+	for (node_id = 0; node_id < num_nodes; node_id++) {
+		top_dev = kfd_create_topology_device();
+		if (!top_dev) {
+			kfd_release_live_view();
+			return -ENOMEM;
+		}
+	}
+
+	sys_props.platform_id =
+		(*((uint64_t *)crat_table->oem_id)) & CRAT_OEMID_64BIT_MASK;
+	sys_props.platform_oem = *((uint64_t *)crat_table->oem_table_id);
+	sys_props.platform_rev = crat_table->revision;
+
+	sub_type_hdr = (struct crat_subtype_generic *)(crat_table+1);
+	while ((char *)sub_type_hdr + sizeof(struct crat_subtype_generic) <
+			((char *)crat_image) + image_len) {
+		if (sub_type_hdr->flags & CRAT_SUBTYPE_FLAGS_ENABLED) {
+			ret = kfd_parse_subtype(sub_type_hdr);
+			if (ret != 0) {
+				kfd_release_live_view();
+				return ret;
+			}
+		}
+
+		sub_type_hdr = (typeof(sub_type_hdr))((char *)sub_type_hdr +
+				sub_type_hdr->length);
+	}
+
+	sys_props.generation_count++;
+	topology_crat_parsed = 1;
+
+	return 0;
+}
+
+
+#define sysfs_show_gen_prop(buffer, fmt, ...) \
+		snprintf(buffer, PAGE_SIZE, "%s"fmt, buffer, __VA_ARGS__)
+#define sysfs_show_32bit_prop(buffer, name, value) \
+		sysfs_show_gen_prop(buffer, "%s %u\n", name, value)
+#define sysfs_show_64bit_prop(buffer, name, value) \
+		sysfs_show_gen_prop(buffer, "%s %llu\n", name, value)
+#define sysfs_show_32bit_val(buffer, value) \
+		sysfs_show_gen_prop(buffer, "%u\n", value)
+#define sysfs_show_str_val(buffer, value) \
+		sysfs_show_gen_prop(buffer, "%s\n", value)
+
+static ssize_t sysprops_show(struct kobject *kobj, struct attribute *attr,
+		char *buffer)
+{
+	ssize_t ret;
+
+	/* Making sure that the buffer is an empty string */
+	buffer[0] = 0;
+
+	if (attr == &sys_props.attr_genid) {
+		ret = sysfs_show_32bit_val(buffer, sys_props.generation_count);
+	} else if (attr == &sys_props.attr_props) {
+		sysfs_show_64bit_prop(buffer, "platform_oem",
+				sys_props.platform_oem);
+		sysfs_show_64bit_prop(buffer, "platform_id",
+				sys_props.platform_id);
+		ret = sysfs_show_64bit_prop(buffer, "platform_rev",
+				sys_props.platform_rev);
+	} else {
+		ret = -EINVAL;
+	}
+
+	return ret;
+}
+
+static const struct sysfs_ops sysprops_ops = {
+	.show = sysprops_show,
+};
+
+static struct kobj_type sysprops_type = {
+	.sysfs_ops = &sysprops_ops,
+};
+
+static ssize_t iolink_show(struct kobject *kobj, struct attribute *attr,
+		char *buffer)
+{
+	ssize_t ret;
+	struct kfd_iolink_properties *iolink;
+
+	/* Making sure that the buffer is an empty string */
+	buffer[0] = 0;
+
+	iolink = container_of(attr, struct kfd_iolink_properties, attr);
+	sysfs_show_32bit_prop(buffer, "type", iolink->iolink_type);
+	sysfs_show_32bit_prop(buffer, "version_major", iolink->ver_maj);
+	sysfs_show_32bit_prop(buffer, "version_minor", iolink->ver_min);
+	sysfs_show_32bit_prop(buffer, "node_from", iolink->node_from);
+	sysfs_show_32bit_prop(buffer, "node_to", iolink->node_to);
+	sysfs_show_32bit_prop(buffer, "weight", iolink->weight);
+	sysfs_show_32bit_prop(buffer, "min_latency", iolink->min_latency);
+	sysfs_show_32bit_prop(buffer, "max_latency", iolink->max_latency);
+	sysfs_show_32bit_prop(buffer, "min_bandwidth", iolink->min_bandwidth);
+	sysfs_show_32bit_prop(buffer, "max_bandwidth", iolink->max_bandwidth);
+	sysfs_show_32bit_prop(buffer, "recommended_transfer_size",
+			iolink->rec_transfer_size);
+	ret = sysfs_show_32bit_prop(buffer, "flags", iolink->flags);
+
+	return ret;
+}
+
+static const struct sysfs_ops iolink_ops = {
+	.show = iolink_show,
+};
+
+static struct kobj_type iolink_type = {
+	.sysfs_ops = &iolink_ops,
+};
+
+static ssize_t mem_show(struct kobject *kobj, struct attribute *attr,
+		char *buffer)
+{
+	ssize_t ret;
+	struct kfd_mem_properties *mem;
+
+	/* Making sure that the buffer is an empty string */
+	buffer[0] = 0;
+
+	mem = container_of(attr, struct kfd_mem_properties, attr);
+	sysfs_show_32bit_prop(buffer, "heap_type", mem->heap_type);
+	sysfs_show_64bit_prop(buffer, "size_in_bytes", mem->size_in_bytes);
+	sysfs_show_32bit_prop(buffer, "flags", mem->flags);
+	sysfs_show_32bit_prop(buffer, "width", mem->width);
+	ret = sysfs_show_32bit_prop(buffer, "mem_clk_max", mem->mem_clk_max);
+
+	return ret;
+}
+
+static const struct sysfs_ops mem_ops = {
+	.show = mem_show,
+};
+
+static struct kobj_type mem_type = {
+	.sysfs_ops = &mem_ops,
+};
+
+static ssize_t kfd_cache_show(struct kobject *kobj, struct attribute *attr,
+		char *buffer)
+{
+	ssize_t ret;
+	uint32_t i;
+	struct kfd_cache_properties *cache;
+
+	/* Making sure that the buffer is an empty string */
+	buffer[0] = 0;
+
+	cache = container_of(attr, struct kfd_cache_properties, attr);
+	sysfs_show_32bit_prop(buffer, "processor_id_low",
+			cache->processor_id_low);
+	sysfs_show_32bit_prop(buffer, "level", cache->cache_level);
+	sysfs_show_32bit_prop(buffer, "size", cache->cache_size);
+	sysfs_show_32bit_prop(buffer, "cache_line_size", cache->cacheline_size);
+	sysfs_show_32bit_prop(buffer, "cache_lines_per_tag",
+			cache->cachelines_per_tag);
+	sysfs_show_32bit_prop(buffer, "association", cache->cache_assoc);
+	sysfs_show_32bit_prop(buffer, "latency", cache->cache_latency);
+	sysfs_show_32bit_prop(buffer, "type", cache->cache_type);
+	snprintf(buffer, PAGE_SIZE, "%ssibling_map ", buffer);
+	for (i = 0; i < KFD_TOPOLOGY_CPU_SIBLINGS; i++)
+		ret = snprintf(buffer, PAGE_SIZE, "%s%d%s",
+				buffer, cache->sibling_map[i],
+				(i == KFD_TOPOLOGY_CPU_SIBLINGS-1) ?
+						"\n" : ",");
+
+	return ret;
+}
+
+static const struct sysfs_ops cache_ops = {
+	.show = kfd_cache_show,
+};
+
+static struct kobj_type cache_type = {
+	.sysfs_ops = &cache_ops,
+};
+
+static ssize_t node_show(struct kobject *kobj, struct attribute *attr,
+		char *buffer)
+{
+	ssize_t ret;
+	struct kfd_topology_device *dev;
+	char public_name[KFD_TOPOLOGY_PUBLIC_NAME_SIZE];
+	uint32_t i;
+
+	/* Making sure that the buffer is an empty string */
+	buffer[0] = 0;
+
+	if (strcmp(attr->name, "gpu_id") == 0) {
+		dev = container_of(attr, struct kfd_topology_device,
+				attr_gpuid);
+		ret = sysfs_show_32bit_val(buffer, dev->gpu_id);
+	} else if (strcmp(attr->name, "name") == 0) {
+		dev = container_of(attr, struct kfd_topology_device,
+				attr_name);
+		for (i = 0; i < KFD_TOPOLOGY_PUBLIC_NAME_SIZE; i++) {
+			public_name[i] =
+					(char)dev->node_props.marketing_name[i];
+			if (dev->node_props.marketing_name[i] == 0)
+				break;
+		}
+		public_name[KFD_TOPOLOGY_PUBLIC_NAME_SIZE-1] = 0x0;
+		ret = sysfs_show_str_val(buffer, public_name);
+	} else {
+		dev = container_of(attr, struct kfd_topology_device,
+				attr_props);
+		sysfs_show_32bit_prop(buffer, "cpu_cores_count",
+				dev->node_props.cpu_cores_count);
+		sysfs_show_32bit_prop(buffer, "simd_count",
+				dev->node_props.simd_count);
+
+		if (dev->mem_bank_count < dev->node_props.mem_banks_count) {
+			pr_warn("kfd: mem_banks_count truncated from %d to %d\n",
+					dev->node_props.mem_banks_count,
+					dev->mem_bank_count);
+			sysfs_show_32bit_prop(buffer, "mem_banks_count",
+					dev->mem_bank_count);
+		} else {
+			sysfs_show_32bit_prop(buffer, "mem_banks_count",
+					dev->node_props.mem_banks_count);
+		}
+
+		sysfs_show_32bit_prop(buffer, "caches_count",
+				dev->node_props.caches_count);
+		sysfs_show_32bit_prop(buffer, "io_links_count",
+				dev->node_props.io_links_count);
+		sysfs_show_32bit_prop(buffer, "cpu_core_id_base",
+				dev->node_props.cpu_core_id_base);
+		sysfs_show_32bit_prop(buffer, "simd_id_base",
+				dev->node_props.simd_id_base);
+		sysfs_show_32bit_prop(buffer, "capability",
+				dev->node_props.capability);
+		sysfs_show_32bit_prop(buffer, "max_waves_per_simd",
+				dev->node_props.max_waves_per_simd);
+		sysfs_show_32bit_prop(buffer, "lds_size_in_kb",
+				dev->node_props.lds_size_in_kb);
+		sysfs_show_32bit_prop(buffer, "gds_size_in_kb",
+				dev->node_props.gds_size_in_kb);
+		sysfs_show_32bit_prop(buffer, "wave_front_size",
+				dev->node_props.wave_front_size);
+		sysfs_show_32bit_prop(buffer, "array_count",
+				dev->node_props.array_count);
+		sysfs_show_32bit_prop(buffer, "simd_arrays_per_engine",
+				dev->node_props.simd_arrays_per_engine);
+		sysfs_show_32bit_prop(buffer, "cu_per_simd_array",
+				dev->node_props.cu_per_simd_array);
+		sysfs_show_32bit_prop(buffer, "simd_per_cu",
+				dev->node_props.simd_per_cu);
+		sysfs_show_32bit_prop(buffer, "max_slots_scratch_cu",
+				dev->node_props.max_slots_scratch_cu);
+		sysfs_show_32bit_prop(buffer, "vendor_id",
+				dev->node_props.vendor_id);
+		sysfs_show_32bit_prop(buffer, "device_id",
+				dev->node_props.device_id);
+		sysfs_show_32bit_prop(buffer, "location_id",
+				dev->node_props.location_id);
+
+		if (dev->gpu) {
+			sysfs_show_32bit_prop(buffer, "max_engine_clk_fcompute",
+					kfd2kgd->get_max_engine_clock_in_mhz(
+						dev->gpu->kgd));
+			sysfs_show_64bit_prop(buffer, "local_mem_size",
+					kfd2kgd->get_vmem_size(dev->gpu->kgd));
+
+			sysfs_show_32bit_prop(buffer, "fw_version",
+					kfd2kgd->get_fw_version(
+							dev->gpu->kgd,
+							KGD_ENGINE_MEC1));
+
+		}
+
+		ret = sysfs_show_32bit_prop(buffer, "max_engine_clk_ccompute",
+				cpufreq_quick_get_max(0)/1000);
+	}
+
+	return ret;
+}
+
+static const struct sysfs_ops node_ops = {
+	.show = node_show,
+};
+
+static struct kobj_type node_type = {
+	.sysfs_ops = &node_ops,
+};
+
+static void kfd_remove_sysfs_file(struct kobject *kobj, struct attribute *attr)
+{
+	sysfs_remove_file(kobj, attr);
+	kobject_del(kobj);
+	kobject_put(kobj);
+}
+
+static void kfd_remove_sysfs_node_entry(struct kfd_topology_device *dev)
+{
+	struct kfd_iolink_properties *iolink;
+	struct kfd_cache_properties *cache;
+	struct kfd_mem_properties *mem;
+
+	BUG_ON(!dev);
+
+	if (dev->kobj_iolink) {
+		list_for_each_entry(iolink, &dev->io_link_props, list)
+			if (iolink->kobj) {
+				kfd_remove_sysfs_file(iolink->kobj,
+							&iolink->attr);
+				iolink->kobj = NULL;
+			}
+		kobject_del(dev->kobj_iolink);
+		kobject_put(dev->kobj_iolink);
+		dev->kobj_iolink = NULL;
+	}
+
+	if (dev->kobj_cache) {
+		list_for_each_entry(cache, &dev->cache_props, list)
+			if (cache->kobj) {
+				kfd_remove_sysfs_file(cache->kobj,
+							&cache->attr);
+				cache->kobj = NULL;
+			}
+		kobject_del(dev->kobj_cache);
+		kobject_put(dev->kobj_cache);
+		dev->kobj_cache = NULL;
+	}
+
+	if (dev->kobj_mem) {
+		list_for_each_entry(mem, &dev->mem_props, list)
+			if (mem->kobj) {
+				kfd_remove_sysfs_file(mem->kobj, &mem->attr);
+				mem->kobj = NULL;
+			}
+		kobject_del(dev->kobj_mem);
+		kobject_put(dev->kobj_mem);
+		dev->kobj_mem = NULL;
+	}
+
+	if (dev->kobj_node) {
+		sysfs_remove_file(dev->kobj_node, &dev->attr_gpuid);
+		sysfs_remove_file(dev->kobj_node, &dev->attr_name);
+		sysfs_remove_file(dev->kobj_node, &dev->attr_props);
+		kobject_del(dev->kobj_node);
+		kobject_put(dev->kobj_node);
+		dev->kobj_node = NULL;
+	}
+}
+
+static int kfd_build_sysfs_node_entry(struct kfd_topology_device *dev,
+		uint32_t id)
+{
+	struct kfd_iolink_properties *iolink;
+	struct kfd_cache_properties *cache;
+	struct kfd_mem_properties *mem;
+	int ret;
+	uint32_t i;
+
+	BUG_ON(!dev);
+
+	/*
+	 * Creating the sysfs folders
+	 */
+	BUG_ON(dev->kobj_node);
+	dev->kobj_node = kfd_alloc_struct(dev->kobj_node);
+	if (!dev->kobj_node)
+		return -ENOMEM;
+
+	ret = kobject_init_and_add(dev->kobj_node, &node_type,
+			sys_props.kobj_nodes, "%d", id);
+	if (ret < 0)
+		return ret;
+
+	dev->kobj_mem = kobject_create_and_add("mem_banks", dev->kobj_node);
+	if (!dev->kobj_mem)
+		return -ENOMEM;
+
+	dev->kobj_cache = kobject_create_and_add("caches", dev->kobj_node);
+	if (!dev->kobj_cache)
+		return -ENOMEM;
+
+	dev->kobj_iolink = kobject_create_and_add("io_links", dev->kobj_node);
+	if (!dev->kobj_iolink)
+		return -ENOMEM;
+
+	/*
+	 * Creating sysfs files for node properties
+	 */
+	dev->attr_gpuid.name = "gpu_id";
+	dev->attr_gpuid.mode = KFD_SYSFS_FILE_MODE;
+	sysfs_attr_init(&dev->attr_gpuid);
+	dev->attr_name.name = "name";
+	dev->attr_name.mode = KFD_SYSFS_FILE_MODE;
+	sysfs_attr_init(&dev->attr_name);
+	dev->attr_props.name = "properties";
+	dev->attr_props.mode = KFD_SYSFS_FILE_MODE;
+	sysfs_attr_init(&dev->attr_props);
+	ret = sysfs_create_file(dev->kobj_node, &dev->attr_gpuid);
+	if (ret < 0)
+		return ret;
+	ret = sysfs_create_file(dev->kobj_node, &dev->attr_name);
+	if (ret < 0)
+		return ret;
+	ret = sysfs_create_file(dev->kobj_node, &dev->attr_props);
+	if (ret < 0)
+		return ret;
+
+	i = 0;
+	list_for_each_entry(mem, &dev->mem_props, list) {
+		mem->kobj = kzalloc(sizeof(struct kobject), GFP_KERNEL);
+		if (!mem->kobj)
+			return -ENOMEM;
+		ret = kobject_init_and_add(mem->kobj, &mem_type,
+				dev->kobj_mem, "%d", i);
+		if (ret < 0)
+			return ret;
+
+		mem->attr.name = "properties";
+		mem->attr.mode = KFD_SYSFS_FILE_MODE;
+		sysfs_attr_init(&mem->attr);
+		ret = sysfs_create_file(mem->kobj, &mem->attr);
+		if (ret < 0)
+			return ret;
+		i++;
+	}
+
+	i = 0;
+	list_for_each_entry(cache, &dev->cache_props, list) {
+		cache->kobj = kzalloc(sizeof(struct kobject), GFP_KERNEL);
+		if (!cache->kobj)
+			return -ENOMEM;
+		ret = kobject_init_and_add(cache->kobj, &cache_type,
+				dev->kobj_cache, "%d", i);
+		if (ret < 0)
+			return ret;
+
+		cache->attr.name = "properties";
+		cache->attr.mode = KFD_SYSFS_FILE_MODE;
+		sysfs_attr_init(&cache->attr);
+		ret = sysfs_create_file(cache->kobj, &cache->attr);
+		if (ret < 0)
+			return ret;
+		i++;
+	}
+
+	i = 0;
+	list_for_each_entry(iolink, &dev->io_link_props, list) {
+		iolink->kobj = kzalloc(sizeof(struct kobject), GFP_KERNEL);
+		if (!iolink->kobj)
+			return -ENOMEM;
+		ret = kobject_init_and_add(iolink->kobj, &iolink_type,
+				dev->kobj_iolink, "%d", i);
+		if (ret < 0)
+			return ret;
+
+		iolink->attr.name = "properties";
+		iolink->attr.mode = KFD_SYSFS_FILE_MODE;
+		sysfs_attr_init(&iolink->attr);
+		ret = sysfs_create_file(iolink->kobj, &iolink->attr);
+		if (ret < 0)
+			return ret;
+		i++;
+}
+
+	return 0;
+}
+
+static int kfd_build_sysfs_node_tree(void)
+{
+	struct kfd_topology_device *dev;
+	int ret;
+	uint32_t i = 0;
+
+	list_for_each_entry(dev, &topology_device_list, list) {
+		ret = kfd_build_sysfs_node_entry(dev, i);
+		if (ret < 0)
+			return ret;
+		i++;
+	}
+
+	return 0;
+}
+
+static void kfd_remove_sysfs_node_tree(void)
+{
+	struct kfd_topology_device *dev;
+
+	list_for_each_entry(dev, &topology_device_list, list)
+		kfd_remove_sysfs_node_entry(dev);
+}
+
+static int kfd_topology_update_sysfs(void)
+{
+	int ret;
+
+	pr_info("Creating topology SYSFS entries\n");
+	if (sys_props.kobj_topology == NULL) {
+		sys_props.kobj_topology =
+				kfd_alloc_struct(sys_props.kobj_topology);
+		if (!sys_props.kobj_topology)
+			return -ENOMEM;
+
+		ret = kobject_init_and_add(sys_props.kobj_topology,
+				&sysprops_type,  &kfd_device->kobj,
+				"topology");
+		if (ret < 0)
+			return ret;
+
+		sys_props.kobj_nodes = kobject_create_and_add("nodes",
+				sys_props.kobj_topology);
+		if (!sys_props.kobj_nodes)
+			return -ENOMEM;
+
+		sys_props.attr_genid.name = "generation_id";
+		sys_props.attr_genid.mode = KFD_SYSFS_FILE_MODE;
+		sysfs_attr_init(&sys_props.attr_genid);
+		ret = sysfs_create_file(sys_props.kobj_topology,
+				&sys_props.attr_genid);
+		if (ret < 0)
+			return ret;
+
+		sys_props.attr_props.name = "system_properties";
+		sys_props.attr_props.mode = KFD_SYSFS_FILE_MODE;
+		sysfs_attr_init(&sys_props.attr_props);
+		ret = sysfs_create_file(sys_props.kobj_topology,
+				&sys_props.attr_props);
+		if (ret < 0)
+			return ret;
+	}
+
+	kfd_remove_sysfs_node_tree();
+
+	return kfd_build_sysfs_node_tree();
+}
+
+static void kfd_topology_release_sysfs(void)
+{
+	kfd_remove_sysfs_node_tree();
+	if (sys_props.kobj_topology) {
+		sysfs_remove_file(sys_props.kobj_topology,
+				&sys_props.attr_genid);
+		sysfs_remove_file(sys_props.kobj_topology,
+				&sys_props.attr_props);
+		if (sys_props.kobj_nodes) {
+			kobject_del(sys_props.kobj_nodes);
+			kobject_put(sys_props.kobj_nodes);
+			sys_props.kobj_nodes = NULL;
+		}
+		kobject_del(sys_props.kobj_topology);
+		kobject_put(sys_props.kobj_topology);
+		sys_props.kobj_topology = NULL;
+	}
+}
+
+int kfd_topology_init(void)
+{
+	void *crat_image = NULL;
+	size_t image_size = 0;
+	int ret;
+
+	/*
+	 * Initialize the head for the topology device list
+	 */
+	INIT_LIST_HEAD(&topology_device_list);
+	init_rwsem(&topology_lock);
+	topology_crat_parsed = 0;
+
+	memset(&sys_props, 0, sizeof(sys_props));
+
+	/*
+	 * Get the CRAT image from the ACPI
+	 */
+	ret = kfd_topology_get_crat_acpi(crat_image, &image_size);
+	if (ret == 0 && image_size > 0) {
+		pr_info("Found CRAT image with size=%zd\n", image_size);
+		crat_image = kmalloc(image_size, GFP_KERNEL);
+		if (!crat_image) {
+			ret = -ENOMEM;
+			pr_err("No memory for allocating CRAT image\n");
+			goto err;
+		}
+		ret = kfd_topology_get_crat_acpi(crat_image, &image_size);
+
+		if (ret == 0) {
+			down_write(&topology_lock);
+			ret = kfd_parse_crat_table(crat_image);
+			if (ret == 0)
+				ret = kfd_topology_update_sysfs();
+			up_write(&topology_lock);
+		} else {
+			pr_err("Couldn't get CRAT table size from ACPI\n");
+		}
+		kfree(crat_image);
+	} else if (ret == -ENODATA) {
+		ret = 0;
+	} else {
+		pr_err("Couldn't get CRAT table size from ACPI\n");
+	}
+
+err:
+	pr_info("Finished initializing topology ret=%d\n", ret);
+	return ret;
+}
+
+void kfd_topology_shutdown(void)
+{
+	kfd_topology_release_sysfs();
+	kfd_release_live_view();
+}
+
+static void kfd_debug_print_topology(void)
+{
+	struct kfd_topology_device *dev;
+	uint32_t i = 0;
+
+	pr_info("DEBUG PRINT OF TOPOLOGY:");
+	list_for_each_entry(dev, &topology_device_list, list) {
+		pr_info("Node: %d\n", i);
+		pr_info("\tGPU assigned: %s\n", (dev->gpu ? "yes" : "no"));
+		pr_info("\tCPU count: %d\n", dev->node_props.cpu_cores_count);
+		pr_info("\tSIMD count: %d", dev->node_props.simd_count);
+		i++;
+	}
+}
+
+static uint32_t kfd_generate_gpu_id(struct kfd_dev *gpu)
+{
+	uint32_t hashout;
+	uint32_t buf[7];
+	int i;
+
+	if (!gpu)
+		return 0;
+
+	buf[0] = gpu->pdev->devfn;
+	buf[1] = gpu->pdev->subsystem_vendor;
+	buf[2] = gpu->pdev->subsystem_device;
+	buf[3] = gpu->pdev->device;
+	buf[4] = gpu->pdev->bus->number;
+	buf[5] = (uint32_t)(kfd2kgd->get_vmem_size(gpu->kgd) & 0xffffffff);
+	buf[6] = (uint32_t)(kfd2kgd->get_vmem_size(gpu->kgd) >> 32);
+
+	for (i = 0, hashout = 0; i < 7; i++)
+		hashout ^= hash_32(buf[i], KFD_GPU_ID_HASH_WIDTH);
+
+	return hashout;
+}
+
+static struct kfd_topology_device *kfd_assign_gpu(struct kfd_dev *gpu)
+{
+	struct kfd_topology_device *dev;
+	struct kfd_topology_device *out_dev = NULL;
+
+	BUG_ON(!gpu);
+
+	list_for_each_entry(dev, &topology_device_list, list)
+		if (dev->gpu == NULL && dev->node_props.simd_count > 0) {
+			dev->gpu = gpu;
+			out_dev = dev;
+			break;
+		}
+
+	return out_dev;
+}
+
+static void kfd_notify_gpu_change(uint32_t gpu_id, int arrival)
+{
+	/*
+	 * TODO: Generate an event for thunk about the arrival/removal
+	 * of the GPU
+	 */
+}
+
+int kfd_topology_add_device(struct kfd_dev *gpu)
+{
+	uint32_t gpu_id;
+	struct kfd_topology_device *dev;
+	int res;
+
+	BUG_ON(!gpu);
+
+	gpu_id = kfd_generate_gpu_id(gpu);
+
+	pr_debug("kfd: Adding new GPU (ID: 0x%x) to topology\n", gpu_id);
+
+	down_write(&topology_lock);
+	/*
+	 * Try to assign the GPU to existing topology device (generated from
+	 * CRAT table
+	 */
+	dev = kfd_assign_gpu(gpu);
+	if (!dev) {
+		pr_info("GPU was not found in the current topology. Extending.\n");
+		kfd_debug_print_topology();
+		dev = kfd_create_topology_device();
+		if (!dev) {
+			res = -ENOMEM;
+			goto err;
+		}
+		dev->gpu = gpu;
+
+		/*
+		 * TODO: Make a call to retrieve topology information from the
+		 * GPU vBIOS
+		 */
+
+		/*
+		 * Update the SYSFS tree, since we added another topology device
+		 */
+		if (kfd_topology_update_sysfs() < 0)
+			kfd_topology_release_sysfs();
+
+	}
+
+	dev->gpu_id = gpu_id;
+	gpu->id = gpu_id;
+	dev->node_props.vendor_id = gpu->pdev->vendor;
+	dev->node_props.device_id = gpu->pdev->device;
+	dev->node_props.location_id = (gpu->pdev->bus->number << 24) +
+			(gpu->pdev->devfn & 0xffffff);
+	/*
+	 * TODO: Retrieve max engine clock values from KGD
+	 */
+
+	res = 0;
+
+err:
+	up_write(&topology_lock);
+
+	if (res == 0)
+		kfd_notify_gpu_change(gpu_id, 1);
+
+	return res;
+}
+
+int kfd_topology_remove_device(struct kfd_dev *gpu)
+{
+	struct kfd_topology_device *dev;
+	uint32_t gpu_id;
+	int res = -ENODEV;
+
+	BUG_ON(!gpu);
+
+	down_write(&topology_lock);
+
+	list_for_each_entry(dev, &topology_device_list, list)
+		if (dev->gpu == gpu) {
+			gpu_id = dev->gpu_id;
+			kfd_remove_sysfs_node_entry(dev);
+			kfd_release_topology_device(dev);
+			res = 0;
+			if (kfd_topology_update_sysfs() < 0)
+				kfd_topology_release_sysfs();
+			break;
+		}
+
+	up_write(&topology_lock);
+
+	if (res == 0)
+		kfd_notify_gpu_change(gpu_id, 0);
+
+	return res;
+}
+
+/*
+ * When idx is out of bounds, the function will return NULL
+ */
+struct kfd_dev *kfd_topology_enum_kfd_devices(uint8_t idx)
+{
+
+	struct kfd_topology_device *top_dev;
+	struct kfd_dev *device = NULL;
+	uint8_t device_idx = 0;
+
+	down_read(&topology_lock);
+
+	list_for_each_entry(top_dev, &topology_device_list, list) {
+		if (device_idx == idx) {
+			device = top_dev->gpu;
+			break;
+		}
+
+		device_idx++;
+	}
+
+	up_read(&topology_lock);
+
+	return device;
+
+}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.h b/drivers/gpu/drm/amd/amdkfd/kfd_topology.h
new file mode 100644
index 000000000000..989624b3cd14
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.h
@@ -0,0 +1,168 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#ifndef __KFD_TOPOLOGY_H__
+#define __KFD_TOPOLOGY_H__
+
+#include <linux/types.h>
+#include <linux/list.h>
+#include "kfd_priv.h"
+
+#define KFD_TOPOLOGY_PUBLIC_NAME_SIZE 128
+
+#define HSA_CAP_HOT_PLUGGABLE			0x00000001
+#define HSA_CAP_ATS_PRESENT			0x00000002
+#define HSA_CAP_SHARED_WITH_GRAPHICS		0x00000004
+#define HSA_CAP_QUEUE_SIZE_POW2			0x00000008
+#define HSA_CAP_QUEUE_SIZE_32BIT		0x00000010
+#define HSA_CAP_QUEUE_IDLE_EVENT		0x00000020
+#define HSA_CAP_VA_LIMIT			0x00000040
+#define HSA_CAP_WATCH_POINTS_SUPPORTED		0x00000080
+#define HSA_CAP_WATCH_POINTS_TOTALBITS_MASK	0x00000f00
+#define HSA_CAP_WATCH_POINTS_TOTALBITS_SHIFT	8
+#define HSA_CAP_RESERVED			0xfffff000
+
+struct kfd_node_properties {
+	uint32_t cpu_cores_count;
+	uint32_t simd_count;
+	uint32_t mem_banks_count;
+	uint32_t caches_count;
+	uint32_t io_links_count;
+	uint32_t cpu_core_id_base;
+	uint32_t simd_id_base;
+	uint32_t capability;
+	uint32_t max_waves_per_simd;
+	uint32_t lds_size_in_kb;
+	uint32_t gds_size_in_kb;
+	uint32_t wave_front_size;
+	uint32_t array_count;
+	uint32_t simd_arrays_per_engine;
+	uint32_t cu_per_simd_array;
+	uint32_t simd_per_cu;
+	uint32_t max_slots_scratch_cu;
+	uint32_t engine_id;
+	uint32_t vendor_id;
+	uint32_t device_id;
+	uint32_t location_id;
+	uint32_t max_engine_clk_fcompute;
+	uint32_t max_engine_clk_ccompute;
+	uint16_t marketing_name[KFD_TOPOLOGY_PUBLIC_NAME_SIZE];
+};
+
+#define HSA_MEM_HEAP_TYPE_SYSTEM	0
+#define HSA_MEM_HEAP_TYPE_FB_PUBLIC	1
+#define HSA_MEM_HEAP_TYPE_FB_PRIVATE	2
+#define HSA_MEM_HEAP_TYPE_GPU_GDS	3
+#define HSA_MEM_HEAP_TYPE_GPU_LDS	4
+#define HSA_MEM_HEAP_TYPE_GPU_SCRATCH	5
+
+#define HSA_MEM_FLAGS_HOT_PLUGGABLE	0x00000001
+#define HSA_MEM_FLAGS_NON_VOLATILE	0x00000002
+#define HSA_MEM_FLAGS_RESERVED		0xfffffffc
+
+struct kfd_mem_properties {
+	struct list_head	list;
+	uint32_t		heap_type;
+	uint64_t		size_in_bytes;
+	uint32_t		flags;
+	uint32_t		width;
+	uint32_t		mem_clk_max;
+	struct kobject		*kobj;
+	struct attribute	attr;
+};
+
+#define KFD_TOPOLOGY_CPU_SIBLINGS 256
+
+#define HSA_CACHE_TYPE_DATA		0x00000001
+#define HSA_CACHE_TYPE_INSTRUCTION	0x00000002
+#define HSA_CACHE_TYPE_CPU		0x00000004
+#define HSA_CACHE_TYPE_HSACU		0x00000008
+#define HSA_CACHE_TYPE_RESERVED		0xfffffff0
+
+struct kfd_cache_properties {
+	struct list_head	list;
+	uint32_t		processor_id_low;
+	uint32_t		cache_level;
+	uint32_t		cache_size;
+	uint32_t		cacheline_size;
+	uint32_t		cachelines_per_tag;
+	uint32_t		cache_assoc;
+	uint32_t		cache_latency;
+	uint32_t		cache_type;
+	uint8_t			sibling_map[KFD_TOPOLOGY_CPU_SIBLINGS];
+	struct kobject		*kobj;
+	struct attribute	attr;
+};
+
+struct kfd_iolink_properties {
+	struct list_head	list;
+	uint32_t		iolink_type;
+	uint32_t		ver_maj;
+	uint32_t		ver_min;
+	uint32_t		node_from;
+	uint32_t		node_to;
+	uint32_t		weight;
+	uint32_t		min_latency;
+	uint32_t		max_latency;
+	uint32_t		min_bandwidth;
+	uint32_t		max_bandwidth;
+	uint32_t		rec_transfer_size;
+	uint32_t		flags;
+	struct kobject		*kobj;
+	struct attribute	attr;
+};
+
+struct kfd_topology_device {
+	struct list_head		list;
+	uint32_t			gpu_id;
+	struct kfd_node_properties	node_props;
+	uint32_t			mem_bank_count;
+	struct list_head		mem_props;
+	uint32_t			cache_count;
+	struct list_head		cache_props;
+	uint32_t			io_link_count;
+	struct list_head		io_link_props;
+	struct kfd_dev			*gpu;
+	struct kobject			*kobj_node;
+	struct kobject			*kobj_mem;
+	struct kobject			*kobj_cache;
+	struct kobject			*kobj_iolink;
+	struct attribute		attr_gpuid;
+	struct attribute		attr_name;
+	struct attribute		attr_props;
+};
+
+struct kfd_system_properties {
+	uint32_t		num_devices;     /* Number of H-NUMA nodes */
+	uint32_t		generation_count;
+	uint64_t		platform_oem;
+	uint64_t		platform_id;
+	uint64_t		platform_rev;
+	struct kobject		*kobj_topology;
+	struct kobject		*kobj_nodes;
+	struct attribute	attr_genid;
+	struct attribute	attr_props;
+};
+
+
+
+#endif /* __KFD_TOPOLOGY_H__ */
diff --git a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
new file mode 100644
index 000000000000..96a512208fad
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
@@ -0,0 +1,200 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+/*
+ * This file defines the private interface between the
+ * AMD kernel graphics drivers and the AMD KFD.
+ */
+
+#ifndef KGD_KFD_INTERFACE_H_INCLUDED
+#define KGD_KFD_INTERFACE_H_INCLUDED
+
+#include <linux/types.h>
+
+struct pci_dev;
+
+#define KFD_INTERFACE_VERSION 1
+
+struct kfd_dev;
+struct kgd_dev;
+
+struct kgd_mem;
+
+enum kgd_memory_pool {
+	KGD_POOL_SYSTEM_CACHEABLE = 1,
+	KGD_POOL_SYSTEM_WRITECOMBINE = 2,
+	KGD_POOL_FRAMEBUFFER = 3,
+};
+
+enum kgd_engine_type {
+	KGD_ENGINE_PFP = 1,
+	KGD_ENGINE_ME,
+	KGD_ENGINE_CE,
+	KGD_ENGINE_MEC1,
+	KGD_ENGINE_MEC2,
+	KGD_ENGINE_RLC,
+	KGD_ENGINE_SDMA,
+	KGD_ENGINE_MAX
+};
+
+struct kgd2kfd_shared_resources {
+	/* Bit n == 1 means VMID n is available for KFD. */
+	unsigned int compute_vmid_bitmap;
+
+	/* Compute pipes are counted starting from MEC0/pipe0 as 0. */
+	unsigned int first_compute_pipe;
+
+	/* Number of MEC pipes available for KFD. */
+	unsigned int compute_pipe_count;
+
+	/* Base address of doorbell aperture. */
+	phys_addr_t doorbell_physical_address;
+
+	/* Size in bytes of doorbell aperture. */
+	size_t doorbell_aperture_size;
+
+	/* Number of bytes at start of aperture reserved for KGD. */
+	size_t doorbell_start_offset;
+};
+
+/**
+ * struct kgd2kfd_calls
+ *
+ * @exit: Notifies amdkfd that kgd module is unloaded
+ *
+ * @probe: Notifies amdkfd about a probe done on a device in the kgd driver.
+ *
+ * @device_init: Initialize the newly probed device (if it is a device that
+ * amdkfd supports)
+ *
+ * @device_exit: Notifies amdkfd about a removal of a kgd device
+ *
+ * @suspend: Notifies amdkfd about a suspend action done to a kgd device
+ *
+ * @resume: Notifies amdkfd about a resume action done to a kgd device
+ *
+ * This structure contains function callback pointers so the kgd driver
+ * will notify to the amdkfd about certain status changes.
+ *
+ */
+struct kgd2kfd_calls {
+	void (*exit)(void);
+	struct kfd_dev* (*probe)(struct kgd_dev *kgd, struct pci_dev *pdev);
+	bool (*device_init)(struct kfd_dev *kfd,
+			const struct kgd2kfd_shared_resources *gpu_resources);
+	void (*device_exit)(struct kfd_dev *kfd);
+	void (*interrupt)(struct kfd_dev *kfd, const void *ih_ring_entry);
+	void (*suspend)(struct kfd_dev *kfd);
+	int (*resume)(struct kfd_dev *kfd);
+};
+
+/**
+ * struct kfd2kgd_calls
+ *
+ * @init_sa_manager: Initialize an instance of the sa manager, used by
+ * amdkfd for all system memory allocations that are mapped to the GART
+ * address space
+ *
+ * @fini_sa_manager: Releases all memory allocations for amdkfd that are
+ * handled by kgd sa manager
+ *
+ * @allocate_mem: Allocate a buffer from amdkfd's sa manager. The buffer can
+ * be used for mqds, hpds, kernel queue, fence and runlists
+ *
+ * @free_mem: Frees a buffer that was allocated by amdkfd's sa manager
+ *
+ * @get_vmem_size: Retrieves (physical) size of VRAM
+ *
+ * @get_gpu_clock_counter: Retrieves GPU clock counter
+ *
+ * @get_max_engine_clock_in_mhz: Retrieves maximum GPU clock in MHz
+ *
+ * @program_sh_mem_settings: A function that should initiate the memory
+ * properties such as main aperture memory type (cache / non cached) and
+ * secondary aperture base address, size and memory type.
+ * This function is used only for no cp scheduling mode.
+ *
+ * @set_pasid_vmid_mapping: Exposes pasid/vmid pair to the H/W for no cp
+ * scheduling mode. Only used for no cp scheduling mode.
+ *
+ * @init_memory: Initializes memory apertures to fixed base/limit address
+ * and non cached memory types.
+ *
+ * @init_pipeline: Initialized the compute pipelines.
+ *
+ * @hqd_load: Loads the mqd structure to a H/W hqd slot. used only for no cp
+ * sceduling mode.
+ *
+ * @hqd_is_occupies: Checks if a hqd slot is occupied.
+ *
+ * @hqd_destroy: Destructs and preempts the queue assigned to that hqd slot.
+ *
+ * @get_fw_version: Returns FW versions from the header
+ *
+ * This structure contains function pointers to services that the kgd driver
+ * provides to amdkfd driver.
+ *
+ */
+struct kfd2kgd_calls {
+	/* Memory management. */
+	int (*init_sa_manager)(struct kgd_dev *kgd, unsigned int size);
+	void (*fini_sa_manager)(struct kgd_dev *kgd);
+	int (*allocate_mem)(struct kgd_dev *kgd, size_t size, size_t alignment,
+			enum kgd_memory_pool pool, struct kgd_mem **mem);
+
+	void (*free_mem)(struct kgd_dev *kgd, struct kgd_mem *mem);
+
+	uint64_t (*get_vmem_size)(struct kgd_dev *kgd);
+	uint64_t (*get_gpu_clock_counter)(struct kgd_dev *kgd);
+
+	uint32_t (*get_max_engine_clock_in_mhz)(struct kgd_dev *kgd);
+
+	/* Register access functions */
+	void (*program_sh_mem_settings)(struct kgd_dev *kgd, uint32_t vmid,
+			uint32_t sh_mem_config,	uint32_t sh_mem_ape1_base,
+			uint32_t sh_mem_ape1_limit, uint32_t sh_mem_bases);
+
+	int (*set_pasid_vmid_mapping)(struct kgd_dev *kgd, unsigned int pasid,
+					unsigned int vmid);
+
+	int (*init_memory)(struct kgd_dev *kgd);
+	int (*init_pipeline)(struct kgd_dev *kgd, uint32_t pipe_id,
+				uint32_t hpd_size, uint64_t hpd_gpu_addr);
+
+	int (*hqd_load)(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
+			uint32_t queue_id, uint32_t __user *wptr);
+
+	bool (*hqd_is_occupied)(struct kgd_dev *kgd, uint64_t queue_address,
+				uint32_t pipe_id, uint32_t queue_id);
+
+	int (*hqd_destroy)(struct kgd_dev *kgd, uint32_t reset_type,
+				unsigned int timeout, uint32_t pipe_id,
+				uint32_t queue_id);
+	uint16_t (*get_fw_version)(struct kgd_dev *kgd,
+				enum kgd_engine_type type);
+};
+
+bool kgd2kfd_init(unsigned interface_version,
+		  const struct kfd2kgd_calls *f2g,
+		  const struct kgd2kfd_calls **g2f);
+
+#endif /* KGD_KFD_INTERFACE_H_INCLUDED */
diff --git a/drivers/gpu/drm/armada/armada_crtc.c b/drivers/gpu/drm/armada/armada_crtc.c
index 9a0cc09e6653..e3a7a5078e5c 100644
--- a/drivers/gpu/drm/armada/armada_crtc.c
+++ b/drivers/gpu/drm/armada/armada_crtc.c
@@ -12,6 +12,7 @@
 #include <linux/platform_device.h>
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_plane_helper.h>
 #include "armada_crtc.h"
 #include "armada_drm.h"
 #include "armada_fb.h"
@@ -260,7 +261,7 @@ static void armada_drm_vblank_off(struct armada_crtc *dcrtc)
 	 * Tell the DRM core that vblank IRQs aren't going to happen for
 	 * a while.  This cleans up any pending vblank events for us.
 	 */
-	drm_vblank_off(dev, dcrtc->num);
+	drm_crtc_vblank_off(&dcrtc->crtc);
 
 	/* Handle any pending flip event. */
 	spin_lock_irq(&dev->event_lock);
@@ -289,6 +290,8 @@ static void armada_drm_crtc_dpms(struct drm_crtc *crtc, int dpms)
 		armada_drm_crtc_update(dcrtc);
 		if (dpms_blanked(dpms))
 			armada_drm_vblank_off(dcrtc);
+		else
+			drm_crtc_vblank_on(&dcrtc->crtc);
 	}
 }
 
@@ -526,7 +529,7 @@ static int armada_drm_crtc_mode_set(struct drm_crtc *crtc,
 	/* Wait for pending flips to complete */
 	wait_event(dcrtc->frame_wait, !dcrtc->frame_work);
 
-	drm_vblank_pre_modeset(crtc->dev, dcrtc->num);
+	drm_crtc_vblank_off(crtc);
 
 	crtc->mode = *adj;
 
@@ -617,7 +620,7 @@ static int armada_drm_crtc_mode_set(struct drm_crtc *crtc,
 
 	armada_drm_crtc_update(dcrtc);
 
-	drm_vblank_post_modeset(crtc->dev, dcrtc->num);
+	drm_crtc_vblank_on(crtc);
 	armada_drm_crtc_finish_fb(dcrtc, old_fb, dpms_blanked(dcrtc->dpms));
 
 	return 0;
@@ -945,18 +948,15 @@ static int armada_drm_crtc_page_flip(struct drm_crtc *crtc,
 	armada_reg_queue_end(work->regs, i);
 
 	/*
-	 * Hold the old framebuffer for the work - DRM appears to drop our
-	 * reference to the old framebuffer in drm_mode_page_flip_ioctl().
+	 * Ensure that we hold a reference on the new framebuffer.
+	 * This has to match the behaviour in mode_set.
 	 */
-	drm_framebuffer_reference(work->old_fb);
+	drm_framebuffer_reference(fb);
 
 	ret = armada_drm_crtc_queue_frame_work(dcrtc, work);
 	if (ret) {
-		/*
-		 * Undo our reference above; DRM does not drop the reference
-		 * to this object on error, so that's okay.
-		 */
-		drm_framebuffer_unreference(work->old_fb);
+		/* Undo our reference above */
+		drm_framebuffer_unreference(fb);
 		kfree(work);
 		return ret;
 	}
diff --git a/drivers/gpu/drm/armada/armada_drv.c b/drivers/gpu/drm/armada/armada_drv.c
index e2d5792b140f..b01420c84864 100644
--- a/drivers/gpu/drm/armada/armada_drv.c
+++ b/drivers/gpu/drm/armada/armada_drv.c
@@ -190,6 +190,7 @@ static int armada_drm_load(struct drm_device *dev, unsigned long flags)
 	if (ret)
 		goto err_comp;
 
+	dev->irq_enabled = true;
 	dev->vblank_disable_allowed = 1;
 
 	ret = armada_fbdev_init(dev);
@@ -308,6 +309,7 @@ static struct drm_driver armada_drm_driver = {
 	.postclose		= NULL,
 	.lastclose		= armada_drm_lastclose,
 	.unload			= armada_drm_unload,
+	.set_busid		= drm_platform_set_busid,
 	.get_vblank_counter	= drm_vblank_count,
 	.enable_vblank		= armada_drm_enable_vblank,
 	.disable_vblank		= armada_drm_disable_vblank,
@@ -330,7 +332,7 @@ static struct drm_driver armada_drm_driver = {
 	.desc			= "Armada SoC DRM",
 	.date			= "20120730",
 	.driver_features	= DRIVER_GEM | DRIVER_MODESET |
-				  DRIVER_PRIME,
+				  DRIVER_HAVE_IRQ | DRIVER_PRIME,
 	.ioctls			= armada_ioctls,
 	.fops			= &armada_drm_fops,
 };
@@ -484,7 +486,6 @@ static struct platform_driver armada_drm_platform_driver = {
 	.remove	= armada_drm_remove,
 	.driver	= {
 		.name	= "armada-drm",
-		.owner	= THIS_MODULE,
 	},
 	.id_table = armada_drm_platform_ids,
 };
diff --git a/drivers/gpu/drm/armada/armada_gem.c b/drivers/gpu/drm/armada/armada_gem.c
index 7496f55611a5..ef5feeecec84 100644
--- a/drivers/gpu/drm/armada/armada_gem.c
+++ b/drivers/gpu/drm/armada/armada_gem.c
@@ -226,7 +226,7 @@ struct armada_gem_object *armada_gem_alloc_object(struct drm_device *dev,
 
 	obj->dev_addr = DMA_ERROR_CODE;
 
-	mapping = obj->obj.filp->f_path.dentry->d_inode->i_mapping;
+	mapping = file_inode(obj->obj.filp)->i_mapping;
 	mapping_set_gfp_mask(mapping, GFP_HIGHUSER | __GFP_RECLAIMABLE);
 
 	DRM_DEBUG_DRIVER("alloc obj %p size %zu\n", obj, size);
diff --git a/drivers/gpu/drm/armada/armada_gem.h b/drivers/gpu/drm/armada/armada_gem.h
index 00b6cd461a03..b000ea3a829a 100644
--- a/drivers/gpu/drm/armada/armada_gem.h
+++ b/drivers/gpu/drm/armada/armada_gem.h
@@ -8,6 +8,8 @@
 #ifndef ARMADA_GEM_H
 #define ARMADA_GEM_H
 
+#include <drm/drm_gem.h>
+
 /* GEM */
 struct armada_gem_object {
 	struct drm_gem_object	obj;
diff --git a/drivers/gpu/drm/ast/ast_dp501.c b/drivers/gpu/drm/ast/ast_dp501.c
index 5da4b62285fa..76f07f38b941 100644
--- a/drivers/gpu/drm/ast/ast_dp501.c
+++ b/drivers/gpu/drm/ast/ast_dp501.c
@@ -379,11 +379,39 @@ static bool ast_init_dvo(struct drm_device *dev)
 	return true;
 }
 
+
+static void ast_init_analog(struct drm_device *dev)
+{
+	struct ast_private *ast = dev->dev_private;
+	u32 data;
+
+	/*
+	 * Set DAC source to VGA mode in SCU2C via the P2A
+	 * bridge. First configure the P2U to target the SCU
+	 * in case it isn't at this stage.
+	 */
+	ast_write32(ast, 0xf004, 0x1e6e0000);
+	ast_write32(ast, 0xf000, 0x1);
+
+	/* Then unlock the SCU with the magic password */
+	ast_write32(ast, 0x12000, 0x1688a8a8);
+	ast_write32(ast, 0x12000, 0x1688a8a8);
+	ast_write32(ast, 0x12000, 0x1688a8a8);
+
+	/* Finally, clear bits [17:16] of SCU2c */
+	data = ast_read32(ast, 0x1202c);
+	data &= 0xfffcffff;
+	ast_write32(ast, 0, data);
+
+	/* Disable DVO */
+	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xa3, 0xcf, 0x00);
+}
+
 void ast_init_3rdtx(struct drm_device *dev)
 {
 	struct ast_private *ast = dev->dev_private;
 	u8 jreg;
-	u32 data;
+
 	if (ast->chip == AST2300 || ast->chip == AST2400) {
 		jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xd1, 0xff);
 		switch (jreg & 0x0e) {
@@ -399,12 +427,8 @@ void ast_init_3rdtx(struct drm_device *dev)
 		default:
 			if (ast->tx_chip_type == AST_TX_SIL164)
 				ast_init_dvo(dev);
-			else {
-				ast_write32(ast, 0x12000, 0x1688a8a8);
-				data = ast_read32(ast, 0x1202c);
-				data &= 0xfffcffff;
-				ast_write32(ast, 0, data);
-			}
+			else
+				ast_init_analog(dev);
 		}
 	}
 }
diff --git a/drivers/gpu/drm/ast/ast_drv.c b/drivers/gpu/drm/ast/ast_drv.c
index f19682a93c24..9a32d9dfdd26 100644
--- a/drivers/gpu/drm/ast/ast_drv.c
+++ b/drivers/gpu/drm/ast/ast_drv.c
@@ -199,6 +199,7 @@ static struct drm_driver driver = {
 
 	.load = ast_driver_load,
 	.unload = ast_driver_unload,
+	.set_busid = drm_pci_set_busid,
 
 	.fops = &ast_fops,
 	.name = DRIVER_NAME,
diff --git a/drivers/gpu/drm/ast/ast_drv.h b/drivers/gpu/drm/ast/ast_drv.h
index 957d4fabf1e1..86205a28e56b 100644
--- a/drivers/gpu/drm/ast/ast_drv.h
+++ b/drivers/gpu/drm/ast/ast_drv.h
@@ -36,6 +36,8 @@
 #include <drm/ttm/ttm_memory.h>
 #include <drm/ttm/ttm_module.h>
 
+#include <drm/drm_gem.h>
+
 #include <linux/i2c.h>
 #include <linux/i2c-algo-bit.h>
 
@@ -125,8 +127,9 @@ struct ast_gem_object;
 
 #define AST_IO_AR_PORT_WRITE		(0x40)
 #define AST_IO_MISC_PORT_WRITE		(0x42)
+#define AST_IO_VGA_ENABLE_PORT		(0x43)
 #define AST_IO_SEQ_PORT			(0x44)
-#define AST_DAC_INDEX_READ		(0x3c7)
+#define AST_IO_DAC_INDEX_READ		(0x47)
 #define AST_IO_DAC_INDEX_WRITE		(0x48)
 #define AST_IO_DAC_DATA		        (0x49)
 #define AST_IO_GR_PORT			(0x4E)
@@ -134,6 +137,8 @@ struct ast_gem_object;
 #define AST_IO_INPUT_STATUS1_READ	(0x5A)
 #define AST_IO_MISC_PORT_READ		(0x4C)
 
+#define AST_IO_MM_OFFSET		(0x380)
+
 #define __ast_read(x) \
 static inline u##x ast_read##x(struct ast_private *ast, u32 reg) { \
 u##x val = 0;\
@@ -316,7 +321,7 @@ struct ast_bo {
 	struct ttm_placement placement;
 	struct ttm_bo_kmap_obj kmap;
 	struct drm_gem_object gem;
-	u32 placements[3];
+	struct ttm_place placements[3];
 	int pin_count;
 };
 #define gem_to_ast_bo(gobj) container_of((gobj), struct ast_bo, gem)
@@ -381,6 +386,9 @@ int ast_bo_push_sysram(struct ast_bo *bo);
 int ast_mmap(struct file *filp, struct vm_area_struct *vma);
 
 /* ast post */
+void ast_enable_vga(struct drm_device *dev);
+void ast_enable_mmio(struct drm_device *dev);
+bool ast_is_vga_enabled(struct drm_device *dev);
 void ast_post_gpu(struct drm_device *dev);
 u32 ast_mindwm(struct ast_private *ast, u32 r);
 void ast_moutdwm(struct ast_private *ast, u32 r, u32 v);
diff --git a/drivers/gpu/drm/ast/ast_fb.c b/drivers/gpu/drm/ast/ast_fb.c
index cba45c774552..5c60ae524c45 100644
--- a/drivers/gpu/drm/ast/ast_fb.c
+++ b/drivers/gpu/drm/ast/ast_fb.c
@@ -186,7 +186,8 @@ static int astfb_create_object(struct ast_fbdev *afbdev,
 static int astfb_create(struct drm_fb_helper *helper,
 			struct drm_fb_helper_surface_size *sizes)
 {
-	struct ast_fbdev *afbdev = (struct ast_fbdev *)helper;
+	struct ast_fbdev *afbdev =
+		container_of(helper, struct ast_fbdev, helper);
 	struct drm_device *dev = afbdev->helper.dev;
 	struct drm_mode_fb_cmd2 mode_cmd;
 	struct drm_framebuffer *fb;
diff --git a/drivers/gpu/drm/ast/ast_main.c b/drivers/gpu/drm/ast/ast_main.c
index b792194e0d9c..035dacc93382 100644
--- a/drivers/gpu/drm/ast/ast_main.c
+++ b/drivers/gpu/drm/ast/ast_main.c
@@ -63,7 +63,7 @@ uint8_t ast_get_index_reg_mask(struct ast_private *ast,
 }
 
 
-static int ast_detect_chip(struct drm_device *dev)
+static int ast_detect_chip(struct drm_device *dev, bool *need_post)
 {
 	struct ast_private *ast = dev->dev_private;
 	uint32_t data, jreg;
@@ -110,6 +110,21 @@ static int ast_detect_chip(struct drm_device *dev)
 		}
 	}
 
+	/*
+	 * If VGA isn't enabled, we need to enable now or subsequent
+	 * access to the scratch registers will fail. We also inform
+	 * our caller that it needs to POST the chip
+	 * (Assumption: VGA not enabled -> need to POST)
+	 */
+	if (!ast_is_vga_enabled(dev)) {
+		ast_enable_vga(dev);
+		ast_enable_mmio(dev);
+		DRM_INFO("VGA not enabled on entry, requesting chip POST\n");
+		*need_post = true;
+	} else
+		*need_post = false;
+
+	/* Check if we support wide screen */
 	switch (ast->chip) {
 	case AST1180:
 		ast->support_wide_screen = true;
@@ -125,6 +140,7 @@ static int ast_detect_chip(struct drm_device *dev)
 			ast->support_wide_screen = true;
 		else {
 			ast->support_wide_screen = false;
+			/* Read SCU7c (silicon revision register) */
 			ast_write32(ast, 0xf004, 0x1e6e0000);
 			ast_write32(ast, 0xf000, 0x1);
 			data = ast_read32(ast, 0x1207c);
@@ -137,11 +153,29 @@ static int ast_detect_chip(struct drm_device *dev)
 		break;
 	}
 
+	/* Check 3rd Tx option (digital output afaik) */
 	ast->tx_chip_type = AST_TX_NONE;
-	jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xa3, 0xff);
-	if (jreg & 0x80)
-		ast->tx_chip_type = AST_TX_SIL164;
+
+	/*
+	 * VGACRA3 Enhanced Color Mode Register, check if DVO is already
+	 * enabled, in that case, assume we have a SIL164 TMDS transmitter
+	 *
+	 * Don't make that assumption if we the chip wasn't enabled and
+	 * is at power-on reset, otherwise we'll incorrectly "detect" a
+	 * SIL164 when there is none.
+	 */
+	if (!*need_post) {
+		jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xa3, 0xff);
+		if (jreg & 0x80)
+			ast->tx_chip_type = AST_TX_SIL164;
+	}
+
 	if ((ast->chip == AST2300) || (ast->chip == AST2400)) {
+		/*
+		 * On AST2300 and 2400, look the configuration set by the SoC in
+		 * the SOC scratch register #1 bits 11:8 (interestingly marked
+		 * as "reserved" in the spec)
+		 */
 		jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xd1, 0xff);
 		switch (jreg) {
 		case 0x04:
@@ -162,6 +196,17 @@ static int ast_detect_chip(struct drm_device *dev)
 		}
 	}
 
+	/* Print stuff for diagnostic purposes */
+	switch(ast->tx_chip_type) {
+	case AST_TX_SIL164:
+		DRM_INFO("Using Sil164 TMDS transmitter\n");
+		break;
+	case AST_TX_DP501:
+		DRM_INFO("Using DP501 DisplayPort transmitter\n");
+		break;
+	default:
+		DRM_INFO("Analog VGA only\n");
+	}
 	return 0;
 }
 
@@ -346,6 +391,7 @@ static u32 ast_get_vram_info(struct drm_device *dev)
 int ast_driver_load(struct drm_device *dev, unsigned long flags)
 {
 	struct ast_private *ast;
+	bool need_post;
 	int ret = 0;
 
 	ast = kzalloc(sizeof(struct ast_private), GFP_KERNEL);
@@ -360,13 +406,27 @@ int ast_driver_load(struct drm_device *dev, unsigned long flags)
 		ret = -EIO;
 		goto out_free;
 	}
-	ast->ioregs = pci_iomap(dev->pdev, 2, 0);
+
+	/*
+	 * If we don't have IO space at all, use MMIO now and
+	 * assume the chip has MMIO enabled by default (rev 0x20
+	 * and higher).
+	 */
+	if (!(pci_resource_flags(dev->pdev, 2) & IORESOURCE_IO)) {
+		DRM_INFO("platform has no IO space, trying MMIO\n");
+		ast->ioregs = ast->regs + AST_IO_MM_OFFSET;
+	}
+
+	/* "map" IO regs if the above hasn't done so already */
 	if (!ast->ioregs) {
-		ret = -EIO;
-		goto out_free;
+		ast->ioregs = pci_iomap(dev->pdev, 2, 0);
+		if (!ast->ioregs) {
+			ret = -EIO;
+			goto out_free;
+		}
 	}
 
-	ast_detect_chip(dev);
+	ast_detect_chip(dev, &need_post);
 
 	if (ast->chip != AST1180) {
 		ast_get_dram_info(dev);
@@ -374,6 +434,9 @@ int ast_driver_load(struct drm_device *dev, unsigned long flags)
 		DRM_INFO("dram %d %d %d %08x\n", ast->mclk, ast->dram_type, ast->dram_bus_width, ast->vram_size);
 	}
 
+	if (need_post)
+		ast_post_gpu(dev);
+
 	ret = ast_mm_init(ast);
 	if (ret)
 		goto out_free;
diff --git a/drivers/gpu/drm/ast/ast_mode.c b/drivers/gpu/drm/ast/ast_mode.c
index 5389350244f2..b7ee2634e47c 100644
--- a/drivers/gpu/drm/ast/ast_mode.c
+++ b/drivers/gpu/drm/ast/ast_mode.c
@@ -31,6 +31,7 @@
 #include <drm/drmP.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_plane_helper.h>
 #include "ast_drv.h"
 
 #include "ast_tables.h"
@@ -80,6 +81,8 @@ static bool ast_get_vbios_mode_info(struct drm_crtc *crtc, struct drm_display_mo
 	struct ast_private *ast = crtc->dev->dev_private;
 	u32 refresh_rate_index = 0, mode_id, color_index, refresh_rate;
 	u32 hborder, vborder;
+	bool check_sync;
+	struct ast_vbios_enhtable *best = NULL;
 
 	switch (crtc->primary->fb->bits_per_pixel) {
 	case 8:
@@ -141,14 +144,34 @@ static bool ast_get_vbios_mode_info(struct drm_crtc *crtc, struct drm_display_mo
 	}
 
 	refresh_rate = drm_mode_vrefresh(mode);
-	while (vbios_mode->enh_table->refresh_rate < refresh_rate) {
-		vbios_mode->enh_table++;
-		if ((vbios_mode->enh_table->refresh_rate > refresh_rate) ||
-		    (vbios_mode->enh_table->refresh_rate == 0xff)) {
-			vbios_mode->enh_table--;
-			break;
+	check_sync = vbios_mode->enh_table->flags & WideScreenMode;
+	do {
+		struct ast_vbios_enhtable *loop = vbios_mode->enh_table;
+
+		while (loop->refresh_rate != 0xff) {
+			if ((check_sync) &&
+			    (((mode->flags & DRM_MODE_FLAG_NVSYNC)  &&
+			      (loop->flags & PVSync))  ||
+			     ((mode->flags & DRM_MODE_FLAG_PVSYNC)  &&
+			      (loop->flags & NVSync))  ||
+			     ((mode->flags & DRM_MODE_FLAG_NHSYNC)  &&
+			      (loop->flags & PHSync))  ||
+			     ((mode->flags & DRM_MODE_FLAG_PHSYNC)  &&
+			      (loop->flags & NHSync)))) {
+				loop++;
+				continue;
+			}
+			if (loop->refresh_rate <= refresh_rate
+			    && (!best || loop->refresh_rate > best->refresh_rate))
+				best = loop;
+			loop++;
 		}
-	}
+		if (best || !check_sync)
+			break;
+		check_sync = 0;
+	} while (1);
+	if (best)
+		vbios_mode->enh_table = best;
 
 	hborder = (vbios_mode->enh_table->flags & HBorder) ? 8 : 0;
 	vborder = (vbios_mode->enh_table->flags & VBorder) ? 8 : 0;
@@ -419,8 +442,10 @@ static void ast_set_sync_reg(struct drm_device *dev, struct drm_display_mode *mo
 	struct ast_private *ast = dev->dev_private;
 	u8 jreg;
 
-	jreg = ast_io_read8(ast, AST_IO_MISC_PORT_READ);
-	jreg |= (vbios_mode->enh_table->flags & SyncNN);
+	jreg  = ast_io_read8(ast, AST_IO_MISC_PORT_READ);
+	jreg &= ~0xC0;
+	if (vbios_mode->enh_table->flags & NVSync) jreg |= 0x80;
+	if (vbios_mode->enh_table->flags & NHSync) jreg |= 0x40;
 	ast_io_write8(ast, AST_IO_MISC_PORT_WRITE, jreg);
 }
 
@@ -1080,8 +1105,8 @@ static u32 copy_cursor_image(u8 *src, u8 *dst, int width, int height)
 			srcdata32[1].ul = *((u32 *)(srcxor + 4)) & 0xf0f0f0f0;
 			data32.b[0] = srcdata32[0].b[1] | (srcdata32[0].b[0] >> 4);
 			data32.b[1] = srcdata32[0].b[3] | (srcdata32[0].b[2] >> 4);
-			data32.b[2] = srcdata32[0].b[1] | (srcdata32[1].b[0] >> 4);
-			data32.b[3] = srcdata32[0].b[3] | (srcdata32[1].b[2] >> 4);
+			data32.b[2] = srcdata32[1].b[1] | (srcdata32[1].b[0] >> 4);
+			data32.b[3] = srcdata32[1].b[3] | (srcdata32[1].b[2] >> 4);
 
 			writel(data32.ul, dstxor);
 			csum += data32.ul;
diff --git a/drivers/gpu/drm/ast/ast_post.c b/drivers/gpu/drm/ast/ast_post.c
index 38d437f3a267..810c51d92b99 100644
--- a/drivers/gpu/drm/ast/ast_post.c
+++ b/drivers/gpu/drm/ast/ast_post.c
@@ -33,18 +33,23 @@
 
 static void ast_init_dram_2300(struct drm_device *dev);
 
-static void
-ast_enable_vga(struct drm_device *dev)
+void ast_enable_vga(struct drm_device *dev)
+{
+	struct ast_private *ast = dev->dev_private;
+
+	ast_io_write8(ast, AST_IO_VGA_ENABLE_PORT, 0x01);
+	ast_io_write8(ast, AST_IO_MISC_PORT_WRITE, 0x01);
+}
+
+void ast_enable_mmio(struct drm_device *dev)
 {
 	struct ast_private *ast = dev->dev_private;
 
-	ast_io_write8(ast, 0x43, 0x01);
-	ast_io_write8(ast, 0x42, 0x01);
+	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xa1, 0xff, 0x04);
 }
 
-#if 0 /* will use later */
-static bool
-ast_is_vga_enabled(struct drm_device *dev)
+
+bool ast_is_vga_enabled(struct drm_device *dev)
 {
 	struct ast_private *ast = dev->dev_private;
 	u8 ch;
@@ -52,7 +57,7 @@ ast_is_vga_enabled(struct drm_device *dev)
 	if (ast->chip == AST1180) {
 		/* TODO 1180 */
 	} else {
-		ch = ast_io_read8(ast, 0x43);
+		ch = ast_io_read8(ast, AST_IO_VGA_ENABLE_PORT);
 		if (ch) {
 			ast_open_key(ast);
 			ch = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xb6, 0xff);
@@ -61,7 +66,6 @@ ast_is_vga_enabled(struct drm_device *dev)
 	}
 	return 0;
 }
-#endif
 
 static const u8 extreginfo[] = { 0x0f, 0x04, 0x1c, 0xff };
 static const u8 extreginfo_ast2300a0[] = { 0x0f, 0x04, 0x1c, 0xff };
@@ -371,6 +375,7 @@ void ast_post_gpu(struct drm_device *dev)
 	pci_write_config_dword(ast->dev->pdev, 0x04, reg);
 
 	ast_enable_vga(dev);
+	ast_enable_mmio(dev);
 	ast_open_key(ast);
 	ast_set_def_ext_reg(dev);
 
diff --git a/drivers/gpu/drm/ast/ast_tables.h b/drivers/gpu/drm/ast/ast_tables.h
index 05c01ea85294..3608d5aa7451 100644
--- a/drivers/gpu/drm/ast/ast_tables.h
+++ b/drivers/gpu/drm/ast/ast_tables.h
@@ -35,14 +35,18 @@
 #define HalfDCLK                0x00000002
 #define DoubleScanMode          0x00000004
 #define LineCompareOff          0x00000008
-#define SyncPP                  0x00000000
-#define SyncPN                  0x00000040
-#define SyncNP                  0x00000080
-#define SyncNN                  0x000000C0
 #define HBorder                 0x00000020
 #define VBorder                 0x00000010
 #define WideScreenMode		0x00000100
 #define NewModeInfo		0x00000200
+#define NHSync			0x00000400
+#define PHSync			0x00000800
+#define NVSync			0x00001000
+#define PVSync			0x00002000
+#define SyncPP			(PVSync | PHSync)
+#define SyncPN			(PVSync | NHSync)
+#define SyncNP			(NVSync | PHSync)
+#define SyncNN			(NVSync | NHSync)
 
 /* DCLK Index */
 #define VCLK25_175     		0x00
@@ -72,6 +76,7 @@
 #define VCLK119     		0x17
 #define VCLK85_5     		0x18
 #define VCLK97_75     		0x19
+#define VCLK118_25			0x1A
 
 static struct ast_vbios_dclk_info dclk_table[] = {
 	{0x2C, 0xE7, 0x03},					/* 00: VCLK25_175	*/
@@ -100,6 +105,7 @@ static struct ast_vbios_dclk_info dclk_table[] = {
 	{0x77, 0x58, 0x80},					/* 17: VCLK119      */
 	{0x32, 0x67, 0x80},				    /* 18: VCLK85_5     */
 	{0x6a, 0x6d, 0x80},					/* 19: VCLK97_75	*/
+	{0x3b, 0x2c, 0x81},					/* 1A: VCLK118_25	*/
 };
 
 static struct ast_vbios_stdtable vbios_stdtable[] = {
@@ -246,8 +252,10 @@ static struct ast_vbios_enhtable res_1360x768[] = {
 static struct ast_vbios_enhtable res_1600x900[] = {
 	{1760, 1600, 48, 32, 926,  900, 3, 5, VCLK97_75,	/* 60Hz CVT RB */
 	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x3A },
-	{1760, 1600, 48, 32, 926,  900, 3, 5, VCLK97_75,	/* end */
-	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 1, 0x3A }
+	{2112, 1600, 88,168, 934,  900, 3, 5, VCLK118_25,	/* 60Hz CVT */
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 2, 0x3A },
+	{2112, 1600, 88,168, 934,  900, 3, 5, VCLK118_25,	/* 60Hz CVT */
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 2, 0x3A },
 };
 
 static struct ast_vbios_enhtable res_1920x1080[] = {
@@ -261,11 +269,11 @@ static struct ast_vbios_enhtable res_1920x1080[] = {
 /* 16:10 */
 static struct ast_vbios_enhtable res_1280x800[] = {
 	{1440, 1280, 48, 32,  823,  800, 3, 6, VCLK71,	/* 60Hz RB */
-	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 35 },
+	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x35 },
 	{1680, 1280, 72,128,  831,  800, 3, 6, VCLK83_5,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x35 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 2, 0x35 },
 	{1680, 1280, 72,128,  831,  800, 3, 6, VCLK83_5,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 1, 0x35 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 2, 0x35 },
 
 };
 
@@ -273,24 +281,24 @@ static struct ast_vbios_enhtable res_1440x900[] = {
 	{1600, 1440, 48, 32,  926,  900, 3, 6, VCLK88_75,	/* 60Hz RB */
 	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x36 },
 	{1904, 1440, 80,152,  934,  900, 3, 6, VCLK106_5,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x36 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 2, 0x36 },
 	{1904, 1440, 80,152,  934,  900, 3, 6, VCLK106_5,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 1, 0x36 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 2, 0x36 },
 };
 
 static struct ast_vbios_enhtable res_1680x1050[] = {
 	{1840, 1680, 48, 32, 1080, 1050, 3, 6, VCLK119,	/* 60Hz RB */
 	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x37 },
 	{2240, 1680,104,176, 1089, 1050, 3, 6, VCLK146_25,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x37 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 2, 0x37 },
 	{2240, 1680,104,176, 1089, 1050, 3, 6, VCLK146_25,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 1, 0x37 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 2, 0x37 },
 };
 
 static struct ast_vbios_enhtable res_1920x1200[] = {
-	{2080, 1920, 48, 32, 1235, 1200, 3, 6, VCLK154,	/* 60Hz */
+	{2080, 1920, 48, 32, 1235, 1200, 3, 6, VCLK154,	/* 60Hz RB*/
 	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x34 },
-	{2080, 1920, 48, 32, 1235, 1200, 3, 6, VCLK154,	/* 60Hz */
+	{2080, 1920, 48, 32, 1235, 1200, 3, 6, VCLK154,	/* 60Hz RB */
 	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 1, 0x34 },
 };
 
diff --git a/drivers/gpu/drm/ast/ast_ttm.c b/drivers/gpu/drm/ast/ast_ttm.c
index b8246227bab0..08f82eae6939 100644
--- a/drivers/gpu/drm/ast/ast_ttm.c
+++ b/drivers/gpu/drm/ast/ast_ttm.c
@@ -293,18 +293,22 @@ void ast_mm_fini(struct ast_private *ast)
 void ast_ttm_placement(struct ast_bo *bo, int domain)
 {
 	u32 c = 0;
-	bo->placement.fpfn = 0;
-	bo->placement.lpfn = 0;
+	unsigned i;
+
 	bo->placement.placement = bo->placements;
 	bo->placement.busy_placement = bo->placements;
 	if (domain & TTM_PL_FLAG_VRAM)
-		bo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_VRAM;
+		bo->placements[c++].flags = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_VRAM;
 	if (domain & TTM_PL_FLAG_SYSTEM)
-		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		bo->placements[c++].flags = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_SYSTEM;
 	if (!c)
-		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		bo->placements[c++].flags = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_SYSTEM;
 	bo->placement.num_placement = c;
 	bo->placement.num_busy_placement = c;
+	for (i = 0; i < c; ++i) {
+		bo->placements[i].fpfn = 0;
+		bo->placements[i].lpfn = 0;
+	}
 }
 
 int ast_bo_create(struct drm_device *dev, int size, int align,
@@ -335,7 +339,7 @@ int ast_bo_create(struct drm_device *dev, int size, int align,
 	ret = ttm_bo_init(&ast->ttm.bdev, &astbo->bo, size,
 			  ttm_bo_type_device, &astbo->placement,
 			  align >> PAGE_SHIFT, false, NULL, acc_size,
-			  NULL, ast_bo_ttm_destroy);
+			  NULL, NULL, ast_bo_ttm_destroy);
 	if (ret)
 		return ret;
 
@@ -360,7 +364,7 @@ int ast_bo_pin(struct ast_bo *bo, u32 pl_flag, u64 *gpu_addr)
 
 	ast_ttm_placement(bo, pl_flag);
 	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret)
 		return ret;
@@ -383,7 +387,7 @@ int ast_bo_unpin(struct ast_bo *bo)
 		return 0;
 
 	for (i = 0; i < bo->placement.num_placement ; i++)
-		bo->placements[i] &= ~TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags &= ~TTM_PL_FLAG_NO_EVICT;
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret)
 		return ret;
@@ -407,7 +411,7 @@ int ast_bo_push_sysram(struct ast_bo *bo)
 
 	ast_ttm_placement(bo, TTM_PL_FLAG_SYSTEM);
 	for (i = 0; i < bo->placement.num_placement ; i++)
-		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret) {
@@ -423,7 +427,7 @@ int ast_mmap(struct file *filp, struct vm_area_struct *vma)
 	struct ast_private *ast;
 
 	if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET))
-		return drm_mmap(filp, vma);
+		return -EINVAL;
 
 	file_priv = filp->private_data;
 	ast = file_priv->minor->dev->dev_private;
diff --git a/drivers/gpu/drm/ati_pcigart.c b/drivers/gpu/drm/ati_pcigart.c
index c399dea27a3b..6c4d4b6eba80 100644
--- a/drivers/gpu/drm/ati_pcigart.c
+++ b/drivers/gpu/drm/ati_pcigart.c
@@ -34,6 +34,8 @@
 #include <linux/export.h>
 #include <drm/drmP.h>
 
+#include <drm/ati_pcigart.h>
+
 # define ATI_PCIGART_PAGE_SIZE		4096	/**< PCI GART page size */
 
 static int drm_ati_alloc_pcigart_table(struct drm_device *dev,
diff --git a/drivers/gpu/drm/bochs/bochs.h b/drivers/gpu/drm/bochs/bochs.h
index 7eb52dd44b01..71f2687fc3cc 100644
--- a/drivers/gpu/drm/bochs/bochs.h
+++ b/drivers/gpu/drm/bochs/bochs.h
@@ -7,6 +7,8 @@
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_fb_helper.h>
 
+#include <drm/drm_gem.h>
+
 #include <ttm/ttm_bo_driver.h>
 #include <ttm/ttm_page_alloc.h>
 
@@ -99,7 +101,7 @@ struct bochs_bo {
 	struct ttm_placement placement;
 	struct ttm_bo_kmap_obj kmap;
 	struct drm_gem_object gem;
-	u32 placements[3];
+	struct ttm_place placements[3];
 	int pin_count;
 };
 
diff --git a/drivers/gpu/drm/bochs/bochs_drv.c b/drivers/gpu/drm/bochs/bochs_drv.c
index 9738e9b14708..98837bde2d25 100644
--- a/drivers/gpu/drm/bochs/bochs_drv.c
+++ b/drivers/gpu/drm/bochs/bochs_drv.c
@@ -82,6 +82,7 @@ static struct drm_driver bochs_driver = {
 	.driver_features	= DRIVER_GEM | DRIVER_MODESET,
 	.load			= bochs_load,
 	.unload			= bochs_unload,
+	.set_busid		= drm_pci_set_busid,
 	.fops			= &bochs_fops,
 	.name			= "bochs-drm",
 	.desc			= "bochs dispi vga interface (qemu stdvga)",
diff --git a/drivers/gpu/drm/bochs/bochs_fbdev.c b/drivers/gpu/drm/bochs/bochs_fbdev.c
index fe95d31cd110..61dbf09dff5d 100644
--- a/drivers/gpu/drm/bochs/bochs_fbdev.c
+++ b/drivers/gpu/drm/bochs/bochs_fbdev.c
@@ -9,6 +9,17 @@
 
 /* ---------------------------------------------------------------------- */
 
+static int bochsfb_mmap(struct fb_info *info,
+			struct vm_area_struct *vma)
+{
+	struct drm_fb_helper *fb_helper = info->par;
+	struct bochs_device *bochs =
+		container_of(fb_helper, struct bochs_device, fb.helper);
+	struct bochs_bo *bo = gem_to_bochs_bo(bochs->fb.gfb.obj);
+
+	return ttm_fbdev_mmap(vma, &bo->bo);
+}
+
 static struct fb_ops bochsfb_ops = {
 	.owner = THIS_MODULE,
 	.fb_check_var = drm_fb_helper_check_var,
@@ -19,6 +30,7 @@ static struct fb_ops bochsfb_ops = {
 	.fb_pan_display = drm_fb_helper_pan_display,
 	.fb_blank = drm_fb_helper_blank,
 	.fb_setcmap = drm_fb_helper_setcmap,
+	.fb_mmap = bochsfb_mmap,
 };
 
 static int bochsfb_create_object(struct bochs_device *bochs,
@@ -123,11 +135,9 @@ static int bochsfb_create(struct drm_fb_helper *helper,
 	info->screen_base = bo->kmap.virtual;
 	info->screen_size = size;
 
-#if 0
-	/* FIXME: get this right for mmap(/dev/fb0) */
-	info->fix.smem_start = bochs_bo_mmap_offset(bo);
+	drm_vma_offset_remove(&bo->bo.bdev->vma_manager, &bo->bo.vma_node);
+	info->fix.smem_start = 0;
 	info->fix.smem_len = size;
-#endif
 
 	ret = fb_alloc_cmap(&info->cmap, 256, 0);
 	if (ret) {
diff --git a/drivers/gpu/drm/bochs/bochs_hw.c b/drivers/gpu/drm/bochs/bochs_hw.c
index dbe619e6aab4..460389702d31 100644
--- a/drivers/gpu/drm/bochs/bochs_hw.c
+++ b/drivers/gpu/drm/bochs/bochs_hw.c
@@ -51,11 +51,10 @@ int bochs_hw_init(struct drm_device *dev, uint32_t flags)
 {
 	struct bochs_device *bochs = dev->dev_private;
 	struct pci_dev *pdev = dev->pdev;
-	unsigned long addr, size, mem, ioaddr, iosize;
+	unsigned long addr, size, mem, ioaddr, iosize, qext_size;
 	u16 id;
 
-	if (/* (ent->driver_data == BOCHS_QEMU_STDVGA) && */
-	    (pdev->resource[2].flags & IORESOURCE_MEM)) {
+	if (pdev->resource[2].flags & IORESOURCE_MEM) {
 		/* mmio bar with vga and bochs registers present */
 		if (pci_request_region(pdev, 2, "bochs-drm") != 0) {
 			DRM_ERROR("Cannot request mmio region\n");
@@ -116,6 +115,24 @@ int bochs_hw_init(struct drm_device *dev, uint32_t flags)
 		 size / 1024, addr,
 		 bochs->ioports ? "ioports" : "mmio",
 		 ioaddr);
+
+	if (bochs->mmio && pdev->revision >= 2) {
+		qext_size = readl(bochs->mmio + 0x600);
+		if (qext_size < 4 || qext_size > iosize)
+			goto noext;
+		DRM_DEBUG("Found qemu ext regs, size %ld\n", qext_size);
+		if (qext_size >= 8) {
+#ifdef __BIG_ENDIAN
+			writel(0xbebebebe, bochs->mmio + 0x604);
+#else
+			writel(0x1e1e1e1e, bochs->mmio + 0x604);
+#endif
+			DRM_DEBUG("  qext endian: 0x%x\n",
+				  readl(bochs->mmio + 0x604));
+		}
+	}
+
+noext:
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/bochs/bochs_kms.c b/drivers/gpu/drm/bochs/bochs_kms.c
index 6b7efcf363d6..85f0f8cf1fb8 100644
--- a/drivers/gpu/drm/bochs/bochs_kms.c
+++ b/drivers/gpu/drm/bochs/bochs_kms.c
@@ -6,6 +6,7 @@
  */
 
 #include "bochs.h"
+#include <drm/drm_plane_helper.h>
 
 static int defx = 1024;
 static int defy = 768;
@@ -108,11 +109,32 @@ static void bochs_crtc_gamma_set(struct drm_crtc *crtc, u16 *red, u16 *green,
 {
 }
 
+static int bochs_crtc_page_flip(struct drm_crtc *crtc,
+				struct drm_framebuffer *fb,
+				struct drm_pending_vblank_event *event,
+				uint32_t page_flip_flags)
+{
+	struct bochs_device *bochs =
+		container_of(crtc, struct bochs_device, crtc);
+	struct drm_framebuffer *old_fb = crtc->primary->fb;
+	unsigned long irqflags;
+
+	crtc->primary->fb = fb;
+	bochs_crtc_mode_set_base(crtc, 0, 0, old_fb);
+	if (event) {
+		spin_lock_irqsave(&bochs->dev->event_lock, irqflags);
+		drm_send_vblank_event(bochs->dev, -1, event);
+		spin_unlock_irqrestore(&bochs->dev->event_lock, irqflags);
+	}
+	return 0;
+}
+
 /* These provide the minimum set of functions required to handle a CRTC */
 static const struct drm_crtc_funcs bochs_crtc_funcs = {
 	.gamma_set = bochs_crtc_gamma_set,
 	.set_config = drm_crtc_helper_set_config,
 	.destroy = drm_crtc_cleanup,
+	.page_flip = bochs_crtc_page_flip,
 };
 
 static const struct drm_crtc_helper_funcs bochs_helper_funcs = {
diff --git a/drivers/gpu/drm/bochs/bochs_mm.c b/drivers/gpu/drm/bochs/bochs_mm.c
index 1728a1b0b813..66286ff518d4 100644
--- a/drivers/gpu/drm/bochs/bochs_mm.c
+++ b/drivers/gpu/drm/bochs/bochs_mm.c
@@ -257,20 +257,26 @@ void bochs_mm_fini(struct bochs_device *bochs)
 
 static void bochs_ttm_placement(struct bochs_bo *bo, int domain)
 {
+	unsigned i;
 	u32 c = 0;
-	bo->placement.fpfn = 0;
-	bo->placement.lpfn = 0;
 	bo->placement.placement = bo->placements;
 	bo->placement.busy_placement = bo->placements;
 	if (domain & TTM_PL_FLAG_VRAM) {
-		bo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED
+		bo->placements[c++].flags = TTM_PL_FLAG_WC
+			| TTM_PL_FLAG_UNCACHED
 			| TTM_PL_FLAG_VRAM;
 	}
 	if (domain & TTM_PL_FLAG_SYSTEM) {
-		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		bo->placements[c++].flags = TTM_PL_MASK_CACHING
+			| TTM_PL_FLAG_SYSTEM;
 	}
 	if (!c) {
-		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		bo->placements[c++].flags = TTM_PL_MASK_CACHING
+			| TTM_PL_FLAG_SYSTEM;
+	}
+	for (i = 0; i < c; ++i) {
+		bo->placements[i].fpfn = 0;
+		bo->placements[i].lpfn = 0;
 	}
 	bo->placement.num_placement = c;
 	bo->placement.num_busy_placement = c;
@@ -294,7 +300,7 @@ int bochs_bo_pin(struct bochs_bo *bo, u32 pl_flag, u64 *gpu_addr)
 
 	bochs_ttm_placement(bo, pl_flag);
 	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret)
 		return ret;
@@ -319,7 +325,7 @@ int bochs_bo_unpin(struct bochs_bo *bo)
 		return 0;
 
 	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] &= ~TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags &= ~TTM_PL_FLAG_NO_EVICT;
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret)
 		return ret;
@@ -333,7 +339,7 @@ int bochs_mmap(struct file *filp, struct vm_area_struct *vma)
 	struct bochs_device *bochs;
 
 	if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET))
-		return drm_mmap(filp, vma);
+		return -EINVAL;
 
 	file_priv = filp->private_data;
 	bochs = file_priv->minor->dev->dev_private;
@@ -371,7 +377,7 @@ static int bochs_bo_create(struct drm_device *dev, int size, int align,
 	ret = ttm_bo_init(&bochs->ttm.bdev, &bochsbo->bo, size,
 			  ttm_bo_type_device, &bochsbo->placement,
 			  align >> PAGE_SHIFT, false, NULL, acc_size,
-			  NULL, bochs_bo_ttm_destroy);
+			  NULL, NULL, bochs_bo_ttm_destroy);
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/cirrus/cirrus_drv.c b/drivers/gpu/drm/cirrus/cirrus_drv.c
index 919c73b94447..c2a1cba1e984 100644
--- a/drivers/gpu/drm/cirrus/cirrus_drv.c
+++ b/drivers/gpu/drm/cirrus/cirrus_drv.c
@@ -32,6 +32,8 @@ static struct drm_driver driver;
 static const struct pci_device_id pciidlist[] = {
 	{ PCI_VENDOR_ID_CIRRUS, PCI_DEVICE_ID_CIRRUS_5446, 0x1af4, 0x1100, 0,
 	  0, 0 },
+	{ PCI_VENDOR_ID_CIRRUS, PCI_DEVICE_ID_CIRRUS_5446, PCI_VENDOR_ID_XEN,
+	  0x0001, 0, 0, 0 },
 	{0,}
 };
 
@@ -128,6 +130,7 @@ static struct drm_driver driver = {
 	.driver_features = DRIVER_MODESET | DRIVER_GEM,
 	.load = cirrus_driver_load,
 	.unload = cirrus_driver_unload,
+	.set_busid = drm_pci_set_busid,
 	.fops = &cirrus_driver_fops,
 	.name = DRIVER_NAME,
 	.desc = DRIVER_DESC,
diff --git a/drivers/gpu/drm/cirrus/cirrus_drv.h b/drivers/gpu/drm/cirrus/cirrus_drv.h
index 401c890b6c6a..693a4565c4ff 100644
--- a/drivers/gpu/drm/cirrus/cirrus_drv.h
+++ b/drivers/gpu/drm/cirrus/cirrus_drv.h
@@ -21,6 +21,8 @@
 #include <drm/ttm/ttm_memory.h>
 #include <drm/ttm/ttm_module.h>
 
+#include <drm/drm_gem.h>
+
 #define DRIVER_AUTHOR		"Matthew Garrett"
 
 #define DRIVER_NAME		"cirrus"
@@ -163,7 +165,7 @@ struct cirrus_bo {
 	struct ttm_placement placement;
 	struct ttm_bo_kmap_obj kmap;
 	struct drm_gem_object gem;
-	u32 placements[3];
+	struct ttm_place placements[3];
 	int pin_count;
 };
 #define gem_to_cirrus_bo(gobj) container_of((gobj), struct cirrus_bo, gem)
@@ -208,6 +210,9 @@ int cirrus_framebuffer_init(struct drm_device *dev,
 			    struct drm_mode_fb_cmd2 *mode_cmd,
 			    struct drm_gem_object *obj);
 
+bool cirrus_check_framebuffer(struct cirrus_device *cdev, int width, int height,
+			      int bpp, int pitch);
+
 				/* cirrus_display.c */
 int cirrus_modeset_init(struct cirrus_device *cdev);
 void cirrus_modeset_fini(struct cirrus_device *cdev);
diff --git a/drivers/gpu/drm/cirrus/cirrus_fbdev.c b/drivers/gpu/drm/cirrus/cirrus_fbdev.c
index 2a135f253e29..502a89eb54b5 100644
--- a/drivers/gpu/drm/cirrus/cirrus_fbdev.c
+++ b/drivers/gpu/drm/cirrus/cirrus_fbdev.c
@@ -139,6 +139,7 @@ static int cirrusfb_create_object(struct cirrus_fbdev *afbdev,
 			       struct drm_gem_object **gobj_p)
 {
 	struct drm_device *dev = afbdev->helper.dev;
+	struct cirrus_device *cdev = dev->dev_private;
 	u32 bpp, depth;
 	u32 size;
 	struct drm_gem_object *gobj;
@@ -146,8 +147,10 @@ static int cirrusfb_create_object(struct cirrus_fbdev *afbdev,
 	int ret = 0;
 	drm_fb_get_bpp_depth(mode_cmd->pixel_format, &depth, &bpp);
 
-	if (bpp > 24)
+	if (!cirrus_check_framebuffer(cdev, mode_cmd->width, mode_cmd->height,
+				      bpp, mode_cmd->pitches[0]))
 		return -EINVAL;
+
 	size = mode_cmd->pitches[0] * mode_cmd->height;
 	ret = cirrus_gem_create(dev, size, true, &gobj);
 	if (ret)
@@ -160,7 +163,8 @@ static int cirrusfb_create_object(struct cirrus_fbdev *afbdev,
 static int cirrusfb_create(struct drm_fb_helper *helper,
 			   struct drm_fb_helper_surface_size *sizes)
 {
-	struct cirrus_fbdev *gfbdev = (struct cirrus_fbdev *)helper;
+	struct cirrus_fbdev *gfbdev =
+		container_of(helper, struct cirrus_fbdev, helper);
 	struct drm_device *dev = gfbdev->helper.dev;
 	struct cirrus_device *cdev = gfbdev->helper.dev->dev_private;
 	struct fb_info *info;
diff --git a/drivers/gpu/drm/cirrus/cirrus_main.c b/drivers/gpu/drm/cirrus/cirrus_main.c
index 99c1983f99d2..4c2d68e9102d 100644
--- a/drivers/gpu/drm/cirrus/cirrus_main.c
+++ b/drivers/gpu/drm/cirrus/cirrus_main.c
@@ -49,14 +49,16 @@ cirrus_user_framebuffer_create(struct drm_device *dev,
 			       struct drm_file *filp,
 			       struct drm_mode_fb_cmd2 *mode_cmd)
 {
+	struct cirrus_device *cdev = dev->dev_private;
 	struct drm_gem_object *obj;
 	struct cirrus_framebuffer *cirrus_fb;
 	int ret;
 	u32 bpp, depth;
 
 	drm_fb_get_bpp_depth(mode_cmd->pixel_format, &depth, &bpp);
-	/* cirrus can't handle > 24bpp framebuffers at all */
-	if (bpp > 24)
+
+	if (!cirrus_check_framebuffer(cdev, mode_cmd->width, mode_cmd->height,
+				      bpp, mode_cmd->pitches[0]))
 		return ERR_PTR(-EINVAL);
 
 	obj = drm_gem_object_lookup(dev, filp, mode_cmd->handles[0]);
@@ -96,8 +98,7 @@ static int cirrus_vram_init(struct cirrus_device *cdev)
 {
 	/* BAR 0 is VRAM */
 	cdev->mc.vram_base = pci_resource_start(cdev->dev->pdev, 0);
-	/* We have 4MB of VRAM */
-	cdev->mc.vram_size = 4 * 1024 * 1024;
+	cdev->mc.vram_size = pci_resource_len(cdev->dev->pdev, 0);
 
 	if (!request_mem_region(cdev->mc.vram_base, cdev->mc.vram_size,
 				"cirrusdrmfb_vram")) {
@@ -179,17 +180,22 @@ int cirrus_driver_load(struct drm_device *dev, unsigned long flags)
 	}
 
 	r = cirrus_mm_init(cdev);
-	if (r)
+	if (r) {
 		dev_err(&dev->pdev->dev, "fatal err on mm init\n");
+		goto out;
+	}
 
 	r = cirrus_modeset_init(cdev);
-	if (r)
+	if (r) {
 		dev_err(&dev->pdev->dev, "Fatal error during modeset init: %d\n", r);
+		goto out;
+	}
 
 	dev->mode_config.funcs = (void *)&cirrus_mode_funcs;
+
+	return 0;
 out:
-	if (r)
-		cirrus_driver_unload(dev);
+	cirrus_driver_unload(dev);
 	return r;
 }
 
@@ -307,3 +313,21 @@ out_unlock:
 	return ret;
 
 }
+
+bool cirrus_check_framebuffer(struct cirrus_device *cdev, int width, int height,
+			      int bpp, int pitch)
+{
+	const int max_pitch = 0x1FF << 3; /* (4096 - 1) & ~111b bytes */
+	const int max_size = cdev->mc.vram_size;
+
+	if (bpp > 32)
+		return false;
+
+	if (pitch > max_pitch)
+		return false;
+
+	if (pitch * height > max_size)
+		return false;
+
+	return true;
+}
diff --git a/drivers/gpu/drm/cirrus/cirrus_mode.c b/drivers/gpu/drm/cirrus/cirrus_mode.c
index c7c5a9d91fa0..99d4a74ffeaf 100644
--- a/drivers/gpu/drm/cirrus/cirrus_mode.c
+++ b/drivers/gpu/drm/cirrus/cirrus_mode.c
@@ -16,6 +16,7 @@
  */
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_plane_helper.h>
 
 #include <video/cirrus.h>
 
diff --git a/drivers/gpu/drm/cirrus/cirrus_ttm.c b/drivers/gpu/drm/cirrus/cirrus_ttm.c
index 92e6b7786097..dfffd528517a 100644
--- a/drivers/gpu/drm/cirrus/cirrus_ttm.c
+++ b/drivers/gpu/drm/cirrus/cirrus_ttm.c
@@ -298,18 +298,21 @@ void cirrus_mm_fini(struct cirrus_device *cirrus)
 void cirrus_ttm_placement(struct cirrus_bo *bo, int domain)
 {
 	u32 c = 0;
-	bo->placement.fpfn = 0;
-	bo->placement.lpfn = 0;
+	unsigned i;
 	bo->placement.placement = bo->placements;
 	bo->placement.busy_placement = bo->placements;
 	if (domain & TTM_PL_FLAG_VRAM)
-		bo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_VRAM;
+		bo->placements[c++].flags = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_VRAM;
 	if (domain & TTM_PL_FLAG_SYSTEM)
-		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		bo->placements[c++].flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
 	if (!c)
-		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		bo->placements[c++].flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
 	bo->placement.num_placement = c;
 	bo->placement.num_busy_placement = c;
+	for (i = 0; i < c; ++i) {
+		bo->placements[i].fpfn = 0;
+		bo->placements[i].lpfn = 0;
+	}
 }
 
 int cirrus_bo_create(struct drm_device *dev, int size, int align,
@@ -340,7 +343,7 @@ int cirrus_bo_create(struct drm_device *dev, int size, int align,
 	ret = ttm_bo_init(&cirrus->ttm.bdev, &cirrusbo->bo, size,
 			  ttm_bo_type_device, &cirrusbo->placement,
 			  align >> PAGE_SHIFT, false, NULL, acc_size,
-			  NULL, cirrus_bo_ttm_destroy);
+			  NULL, NULL, cirrus_bo_ttm_destroy);
 	if (ret)
 		return ret;
 
@@ -365,7 +368,7 @@ int cirrus_bo_pin(struct cirrus_bo *bo, u32 pl_flag, u64 *gpu_addr)
 
 	cirrus_ttm_placement(bo, pl_flag);
 	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret)
 		return ret;
@@ -392,7 +395,7 @@ int cirrus_bo_push_sysram(struct cirrus_bo *bo)
 
 	cirrus_ttm_placement(bo, TTM_PL_FLAG_SYSTEM);
 	for (i = 0; i < bo->placement.num_placement ; i++)
-		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret) {
@@ -408,7 +411,7 @@ int cirrus_mmap(struct file *filp, struct vm_area_struct *vma)
 	struct cirrus_device *cirrus;
 
 	if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET))
-		return drm_mmap(filp, vma);
+		return -EINVAL;
 
 	file_priv = filp->private_data;
 	cirrus = file_priv->minor->dev->dev_private;
diff --git a/drivers/gpu/drm/drm_agpsupport.c b/drivers/gpu/drm/drm_agpsupport.c
index dde205cef384..4b2b4aa5033b 100644
--- a/drivers/gpu/drm/drm_agpsupport.c
+++ b/drivers/gpu/drm/drm_agpsupport.c
@@ -34,6 +34,7 @@
 #include <drm/drmP.h>
 #include <linux/module.h>
 #include <linux/slab.h>
+#include "drm_legacy.h"
 
 #if __OS_HAS_AGP
 
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
new file mode 100644
index 000000000000..ff5f034cc405
--- /dev/null
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -0,0 +1,657 @@
+/*
+ * Copyright (C) 2014 Red Hat
+ * Copyright (C) 2014 Intel Corp.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors:
+ * Rob Clark <robdclark@gmail.com>
+ * Daniel Vetter <daniel.vetter@ffwll.ch>
+ */
+
+
+#include <drm/drmP.h>
+#include <drm/drm_atomic.h>
+#include <drm/drm_plane_helper.h>
+
+static void kfree_state(struct drm_atomic_state *state)
+{
+	kfree(state->connectors);
+	kfree(state->connector_states);
+	kfree(state->crtcs);
+	kfree(state->crtc_states);
+	kfree(state->planes);
+	kfree(state->plane_states);
+	kfree(state);
+}
+
+/**
+ * drm_atomic_state_alloc - allocate atomic state
+ * @dev: DRM device
+ *
+ * This allocates an empty atomic state to track updates.
+ */
+struct drm_atomic_state *
+drm_atomic_state_alloc(struct drm_device *dev)
+{
+	struct drm_atomic_state *state;
+
+	state = kzalloc(sizeof(*state), GFP_KERNEL);
+	if (!state)
+		return NULL;
+
+	state->num_connector = ACCESS_ONCE(dev->mode_config.num_connector);
+
+	state->crtcs = kcalloc(dev->mode_config.num_crtc,
+			       sizeof(*state->crtcs), GFP_KERNEL);
+	if (!state->crtcs)
+		goto fail;
+	state->crtc_states = kcalloc(dev->mode_config.num_crtc,
+				     sizeof(*state->crtc_states), GFP_KERNEL);
+	if (!state->crtc_states)
+		goto fail;
+	state->planes = kcalloc(dev->mode_config.num_total_plane,
+				sizeof(*state->planes), GFP_KERNEL);
+	if (!state->planes)
+		goto fail;
+	state->plane_states = kcalloc(dev->mode_config.num_total_plane,
+				      sizeof(*state->plane_states), GFP_KERNEL);
+	if (!state->plane_states)
+		goto fail;
+	state->connectors = kcalloc(state->num_connector,
+				    sizeof(*state->connectors),
+				    GFP_KERNEL);
+	if (!state->connectors)
+		goto fail;
+	state->connector_states = kcalloc(state->num_connector,
+					  sizeof(*state->connector_states),
+					  GFP_KERNEL);
+	if (!state->connector_states)
+		goto fail;
+
+	state->dev = dev;
+
+	DRM_DEBUG_KMS("Allocate atomic state %p\n", state);
+
+	return state;
+fail:
+	kfree_state(state);
+
+	return NULL;
+}
+EXPORT_SYMBOL(drm_atomic_state_alloc);
+
+/**
+ * drm_atomic_state_clear - clear state object
+ * @state: atomic state
+ *
+ * When the w/w mutex algorithm detects a deadlock we need to back off and drop
+ * all locks. So someone else could sneak in and change the current modeset
+ * configuration. Which means that all the state assembled in @state is no
+ * longer an atomic update to the current state, but to some arbitrary earlier
+ * state. Which could break assumptions the driver's ->atomic_check likely
+ * relies on.
+ *
+ * Hence we must clear all cached state and completely start over, using this
+ * function.
+ */
+void drm_atomic_state_clear(struct drm_atomic_state *state)
+{
+	struct drm_device *dev = state->dev;
+	struct drm_mode_config *config = &dev->mode_config;
+	int i;
+
+	DRM_DEBUG_KMS("Clearing atomic state %p\n", state);
+
+	for (i = 0; i < state->num_connector; i++) {
+		struct drm_connector *connector = state->connectors[i];
+
+		if (!connector)
+			continue;
+
+		WARN_ON(!drm_modeset_is_locked(&config->connection_mutex));
+
+		connector->funcs->atomic_destroy_state(connector,
+						       state->connector_states[i]);
+	}
+
+	for (i = 0; i < config->num_crtc; i++) {
+		struct drm_crtc *crtc = state->crtcs[i];
+
+		if (!crtc)
+			continue;
+
+		crtc->funcs->atomic_destroy_state(crtc,
+						  state->crtc_states[i]);
+	}
+
+	for (i = 0; i < config->num_total_plane; i++) {
+		struct drm_plane *plane = state->planes[i];
+
+		if (!plane)
+			continue;
+
+		plane->funcs->atomic_destroy_state(plane,
+						   state->plane_states[i]);
+	}
+}
+EXPORT_SYMBOL(drm_atomic_state_clear);
+
+/**
+ * drm_atomic_state_free - free all memory for an atomic state
+ * @state: atomic state to deallocate
+ *
+ * This frees all memory associated with an atomic state, including all the
+ * per-object state for planes, crtcs and connectors.
+ */
+void drm_atomic_state_free(struct drm_atomic_state *state)
+{
+	drm_atomic_state_clear(state);
+
+	DRM_DEBUG_KMS("Freeing atomic state %p\n", state);
+
+	kfree_state(state);
+}
+EXPORT_SYMBOL(drm_atomic_state_free);
+
+/**
+ * drm_atomic_get_crtc_state - get crtc state
+ * @state: global atomic state object
+ * @crtc: crtc to get state object for
+ *
+ * This function returns the crtc state for the given crtc, allocating it if
+ * needed. It will also grab the relevant crtc lock to make sure that the state
+ * is consistent.
+ *
+ * Returns:
+ *
+ * Either the allocated state or the error code encoded into the pointer. When
+ * the error is EDEADLK then the w/w mutex code has detected a deadlock and the
+ * entire atomic sequence must be restarted. All other errors are fatal.
+ */
+struct drm_crtc_state *
+drm_atomic_get_crtc_state(struct drm_atomic_state *state,
+			  struct drm_crtc *crtc)
+{
+	int ret, index;
+	struct drm_crtc_state *crtc_state;
+
+	index = drm_crtc_index(crtc);
+
+	if (state->crtc_states[index])
+		return state->crtc_states[index];
+
+	ret = drm_modeset_lock(&crtc->mutex, state->acquire_ctx);
+	if (ret)
+		return ERR_PTR(ret);
+
+	crtc_state = crtc->funcs->atomic_duplicate_state(crtc);
+	if (!crtc_state)
+		return ERR_PTR(-ENOMEM);
+
+	state->crtc_states[index] = crtc_state;
+	state->crtcs[index] = crtc;
+	crtc_state->state = state;
+
+	DRM_DEBUG_KMS("Added [CRTC:%d] %p state to %p\n",
+		      crtc->base.id, crtc_state, state);
+
+	return crtc_state;
+}
+EXPORT_SYMBOL(drm_atomic_get_crtc_state);
+
+/**
+ * drm_atomic_get_plane_state - get plane state
+ * @state: global atomic state object
+ * @plane: plane to get state object for
+ *
+ * This function returns the plane state for the given plane, allocating it if
+ * needed. It will also grab the relevant plane lock to make sure that the state
+ * is consistent.
+ *
+ * Returns:
+ *
+ * Either the allocated state or the error code encoded into the pointer. When
+ * the error is EDEADLK then the w/w mutex code has detected a deadlock and the
+ * entire atomic sequence must be restarted. All other errors are fatal.
+ */
+struct drm_plane_state *
+drm_atomic_get_plane_state(struct drm_atomic_state *state,
+			  struct drm_plane *plane)
+{
+	int ret, index;
+	struct drm_plane_state *plane_state;
+
+	index = drm_plane_index(plane);
+
+	if (state->plane_states[index])
+		return state->plane_states[index];
+
+	ret = drm_modeset_lock(&plane->mutex, state->acquire_ctx);
+	if (ret)
+		return ERR_PTR(ret);
+
+	plane_state = plane->funcs->atomic_duplicate_state(plane);
+	if (!plane_state)
+		return ERR_PTR(-ENOMEM);
+
+	state->plane_states[index] = plane_state;
+	state->planes[index] = plane;
+	plane_state->state = state;
+
+	DRM_DEBUG_KMS("Added [PLANE:%d] %p state to %p\n",
+		      plane->base.id, plane_state, state);
+
+	if (plane_state->crtc) {
+		struct drm_crtc_state *crtc_state;
+
+		crtc_state = drm_atomic_get_crtc_state(state,
+						       plane_state->crtc);
+		if (IS_ERR(crtc_state))
+			return ERR_CAST(crtc_state);
+	}
+
+	return plane_state;
+}
+EXPORT_SYMBOL(drm_atomic_get_plane_state);
+
+/**
+ * drm_atomic_get_connector_state - get connector state
+ * @state: global atomic state object
+ * @connector: connector to get state object for
+ *
+ * This function returns the connector state for the given connector,
+ * allocating it if needed. It will also grab the relevant connector lock to
+ * make sure that the state is consistent.
+ *
+ * Returns:
+ *
+ * Either the allocated state or the error code encoded into the pointer. When
+ * the error is EDEADLK then the w/w mutex code has detected a deadlock and the
+ * entire atomic sequence must be restarted. All other errors are fatal.
+ */
+struct drm_connector_state *
+drm_atomic_get_connector_state(struct drm_atomic_state *state,
+			  struct drm_connector *connector)
+{
+	int ret, index;
+	struct drm_mode_config *config = &connector->dev->mode_config;
+	struct drm_connector_state *connector_state;
+
+	ret = drm_modeset_lock(&config->connection_mutex, state->acquire_ctx);
+	if (ret)
+		return ERR_PTR(ret);
+
+	index = drm_connector_index(connector);
+
+	/*
+	 * Construction of atomic state updates can race with a connector
+	 * hot-add which might overflow. In this case flip the table and just
+	 * restart the entire ioctl - no one is fast enough to livelock a cpu
+	 * with physical hotplug events anyway.
+	 *
+	 * Note that we only grab the indexes once we have the right lock to
+	 * prevent hotplug/unplugging of connectors. So removal is no problem,
+	 * at most the array is a bit too large.
+	 */
+	if (index >= state->num_connector) {
+		DRM_DEBUG_KMS("Hot-added connector would overflow state array, restarting\n");
+		return ERR_PTR(-EAGAIN);
+	}
+
+	if (state->connector_states[index])
+		return state->connector_states[index];
+
+	connector_state = connector->funcs->atomic_duplicate_state(connector);
+	if (!connector_state)
+		return ERR_PTR(-ENOMEM);
+
+	state->connector_states[index] = connector_state;
+	state->connectors[index] = connector;
+	connector_state->state = state;
+
+	DRM_DEBUG_KMS("Added [CONNECTOR:%d] %p state to %p\n",
+		      connector->base.id, connector_state, state);
+
+	if (connector_state->crtc) {
+		struct drm_crtc_state *crtc_state;
+
+		crtc_state = drm_atomic_get_crtc_state(state,
+						       connector_state->crtc);
+		if (IS_ERR(crtc_state))
+			return ERR_CAST(crtc_state);
+	}
+
+	return connector_state;
+}
+EXPORT_SYMBOL(drm_atomic_get_connector_state);
+
+/**
+ * drm_atomic_set_crtc_for_plane - set crtc for plane
+ * @state: the incoming atomic state
+ * @plane: the plane whose incoming state to update
+ * @crtc: crtc to use for the plane
+ *
+ * Changing the assigned crtc for a plane requires us to grab the lock and state
+ * for the new crtc, as needed. This function takes care of all these details
+ * besides updating the pointer in the state object itself.
+ *
+ * Returns:
+ * 0 on success or can fail with -EDEADLK or -ENOMEM. When the error is EDEADLK
+ * then the w/w mutex code has detected a deadlock and the entire atomic
+ * sequence must be restarted. All other errors are fatal.
+ */
+int
+drm_atomic_set_crtc_for_plane(struct drm_atomic_state *state,
+			      struct drm_plane *plane, struct drm_crtc *crtc)
+{
+	struct drm_plane_state *plane_state =
+			drm_atomic_get_plane_state(state, plane);
+	struct drm_crtc_state *crtc_state;
+
+	if (WARN_ON(IS_ERR(plane_state)))
+		return PTR_ERR(plane_state);
+
+	if (plane_state->crtc) {
+		crtc_state = drm_atomic_get_crtc_state(plane_state->state,
+						       plane_state->crtc);
+		if (WARN_ON(IS_ERR(crtc_state)))
+			return PTR_ERR(crtc_state);
+
+		crtc_state->plane_mask &= ~(1 << drm_plane_index(plane));
+	}
+
+	plane_state->crtc = crtc;
+
+	if (crtc) {
+		crtc_state = drm_atomic_get_crtc_state(plane_state->state,
+						       crtc);
+		if (IS_ERR(crtc_state))
+			return PTR_ERR(crtc_state);
+		crtc_state->plane_mask |= (1 << drm_plane_index(plane));
+	}
+
+	if (crtc)
+		DRM_DEBUG_KMS("Link plane state %p to [CRTC:%d]\n",
+			      plane_state, crtc->base.id);
+	else
+		DRM_DEBUG_KMS("Link plane state %p to [NOCRTC]\n", plane_state);
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_atomic_set_crtc_for_plane);
+
+/**
+ * drm_atomic_set_fb_for_plane - set crtc for plane
+ * @plane_state: atomic state object for the plane
+ * @fb: fb to use for the plane
+ *
+ * Changing the assigned framebuffer for a plane requires us to grab a reference
+ * to the new fb and drop the reference to the old fb, if there is one. This
+ * function takes care of all these details besides updating the pointer in the
+ * state object itself.
+ */
+void
+drm_atomic_set_fb_for_plane(struct drm_plane_state *plane_state,
+			    struct drm_framebuffer *fb)
+{
+	if (plane_state->fb)
+		drm_framebuffer_unreference(plane_state->fb);
+	if (fb)
+		drm_framebuffer_reference(fb);
+	plane_state->fb = fb;
+
+	if (fb)
+		DRM_DEBUG_KMS("Set [FB:%d] for plane state %p\n",
+			      fb->base.id, plane_state);
+	else
+		DRM_DEBUG_KMS("Set [NOFB] for plane state %p\n", plane_state);
+}
+EXPORT_SYMBOL(drm_atomic_set_fb_for_plane);
+
+/**
+ * drm_atomic_set_crtc_for_connector - set crtc for connector
+ * @conn_state: atomic state object for the connector
+ * @crtc: crtc to use for the connector
+ *
+ * Changing the assigned crtc for a connector requires us to grab the lock and
+ * state for the new crtc, as needed. This function takes care of all these
+ * details besides updating the pointer in the state object itself.
+ *
+ * Returns:
+ * 0 on success or can fail with -EDEADLK or -ENOMEM. When the error is EDEADLK
+ * then the w/w mutex code has detected a deadlock and the entire atomic
+ * sequence must be restarted. All other errors are fatal.
+ */
+int
+drm_atomic_set_crtc_for_connector(struct drm_connector_state *conn_state,
+				  struct drm_crtc *crtc)
+{
+	struct drm_crtc_state *crtc_state;
+
+	if (crtc) {
+		crtc_state = drm_atomic_get_crtc_state(conn_state->state, crtc);
+		if (IS_ERR(crtc_state))
+			return PTR_ERR(crtc_state);
+	}
+
+	conn_state->crtc = crtc;
+
+	if (crtc)
+		DRM_DEBUG_KMS("Link connector state %p to [CRTC:%d]\n",
+			      conn_state, crtc->base.id);
+	else
+		DRM_DEBUG_KMS("Link connector state %p to [NOCRTC]\n",
+			      conn_state);
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_atomic_set_crtc_for_connector);
+
+/**
+ * drm_atomic_add_affected_connectors - add connectors for crtc
+ * @state: atomic state
+ * @crtc: DRM crtc
+ *
+ * This function walks the current configuration and adds all connectors
+ * currently using @crtc to the atomic configuration @state. Note that this
+ * function must acquire the connection mutex. This can potentially cause
+ * unneeded seralization if the update is just for the planes on one crtc. Hence
+ * drivers and helpers should only call this when really needed (e.g. when a
+ * full modeset needs to happen due to some change).
+ *
+ * Returns:
+ * 0 on success or can fail with -EDEADLK or -ENOMEM. When the error is EDEADLK
+ * then the w/w mutex code has detected a deadlock and the entire atomic
+ * sequence must be restarted. All other errors are fatal.
+ */
+int
+drm_atomic_add_affected_connectors(struct drm_atomic_state *state,
+				   struct drm_crtc *crtc)
+{
+	struct drm_mode_config *config = &state->dev->mode_config;
+	struct drm_connector *connector;
+	struct drm_connector_state *conn_state;
+	int ret;
+
+	ret = drm_modeset_lock(&config->connection_mutex, state->acquire_ctx);
+	if (ret)
+		return ret;
+
+	DRM_DEBUG_KMS("Adding all current connectors for [CRTC:%d] to %p\n",
+		      crtc->base.id, state);
+
+	/*
+	 * Changed connectors are already in @state, so only need to look at the
+	 * current configuration.
+	 */
+	list_for_each_entry(connector, &config->connector_list, head) {
+		if (connector->state->crtc != crtc)
+			continue;
+
+		conn_state = drm_atomic_get_connector_state(state, connector);
+		if (IS_ERR(conn_state))
+			return PTR_ERR(conn_state);
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_atomic_add_affected_connectors);
+
+/**
+ * drm_atomic_connectors_for_crtc - count number of connected outputs
+ * @state: atomic state
+ * @crtc: DRM crtc
+ *
+ * This function counts all connectors which will be connected to @crtc
+ * according to @state. Useful to recompute the enable state for @crtc.
+ */
+int
+drm_atomic_connectors_for_crtc(struct drm_atomic_state *state,
+			       struct drm_crtc *crtc)
+{
+	int i, num_connected_connectors = 0;
+
+	for (i = 0; i < state->num_connector; i++) {
+		struct drm_connector_state *conn_state;
+
+		conn_state = state->connector_states[i];
+
+		if (conn_state && conn_state->crtc == crtc)
+			num_connected_connectors++;
+	}
+
+	DRM_DEBUG_KMS("State %p has %i connectors for [CRTC:%d]\n",
+		      state, num_connected_connectors, crtc->base.id);
+
+	return num_connected_connectors;
+}
+EXPORT_SYMBOL(drm_atomic_connectors_for_crtc);
+
+/**
+ * drm_atomic_legacy_backoff - locking backoff for legacy ioctls
+ * @state: atomic state
+ *
+ * This function should be used by legacy entry points which don't understand
+ * -EDEADLK semantics. For simplicity this one will grab all modeset locks after
+ *  the slowpath completed.
+ */
+void drm_atomic_legacy_backoff(struct drm_atomic_state *state)
+{
+	int ret;
+
+retry:
+	drm_modeset_backoff(state->acquire_ctx);
+
+	ret = drm_modeset_lock(&state->dev->mode_config.connection_mutex,
+			       state->acquire_ctx);
+	if (ret)
+		goto retry;
+	ret = drm_modeset_lock_all_crtcs(state->dev,
+					 state->acquire_ctx);
+	if (ret)
+		goto retry;
+}
+EXPORT_SYMBOL(drm_atomic_legacy_backoff);
+
+/**
+ * drm_atomic_check_only - check whether a given config would work
+ * @state: atomic configuration to check
+ *
+ * Note that this function can return -EDEADLK if the driver needed to acquire
+ * more locks but encountered a deadlock. The caller must then do the usual w/w
+ * backoff dance and restart. All other errors are fatal.
+ *
+ * Returns:
+ * 0 on success, negative error code on failure.
+ */
+int drm_atomic_check_only(struct drm_atomic_state *state)
+{
+	struct drm_mode_config *config = &state->dev->mode_config;
+
+	DRM_DEBUG_KMS("checking %p\n", state);
+
+	if (config->funcs->atomic_check)
+		return config->funcs->atomic_check(state->dev, state);
+	else
+		return 0;
+}
+EXPORT_SYMBOL(drm_atomic_check_only);
+
+/**
+ * drm_atomic_commit - commit configuration atomically
+ * @state: atomic configuration to check
+ *
+ * Note that this function can return -EDEADLK if the driver needed to acquire
+ * more locks but encountered a deadlock. The caller must then do the usual w/w
+ * backoff dance and restart. All other errors are fatal.
+ *
+ * Also note that on successful execution ownership of @state is transferred
+ * from the caller of this function to the function itself. The caller must not
+ * free or in any other way access @state. If the function fails then the caller
+ * must clean up @state itself.
+ *
+ * Returns:
+ * 0 on success, negative error code on failure.
+ */
+int drm_atomic_commit(struct drm_atomic_state *state)
+{
+	struct drm_mode_config *config = &state->dev->mode_config;
+	int ret;
+
+	ret = drm_atomic_check_only(state);
+	if (ret)
+		return ret;
+
+	DRM_DEBUG_KMS("commiting %p\n", state);
+
+	return config->funcs->atomic_commit(state->dev, state, false);
+}
+EXPORT_SYMBOL(drm_atomic_commit);
+
+/**
+ * drm_atomic_async_commit - atomic&async configuration commit
+ * @state: atomic configuration to check
+ *
+ * Note that this function can return -EDEADLK if the driver needed to acquire
+ * more locks but encountered a deadlock. The caller must then do the usual w/w
+ * backoff dance and restart. All other errors are fatal.
+ *
+ * Also note that on successful execution ownership of @state is transferred
+ * from the caller of this function to the function itself. The caller must not
+ * free or in any other way access @state. If the function fails then the caller
+ * must clean up @state itself.
+ *
+ * Returns:
+ * 0 on success, negative error code on failure.
+ */
+int drm_atomic_async_commit(struct drm_atomic_state *state)
+{
+	struct drm_mode_config *config = &state->dev->mode_config;
+	int ret;
+
+	ret = drm_atomic_check_only(state);
+	if (ret)
+		return ret;
+
+	DRM_DEBUG_KMS("commiting %p asynchronously\n", state);
+
+	return config->funcs->atomic_commit(state->dev, state, true);
+}
+EXPORT_SYMBOL(drm_atomic_async_commit);
diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c
new file mode 100644
index 000000000000..bbdbe4721573
--- /dev/null
+++ b/drivers/gpu/drm/drm_atomic_helper.c
@@ -0,0 +1,1966 @@
+/*
+ * Copyright (C) 2014 Red Hat
+ * Copyright (C) 2014 Intel Corp.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors:
+ * Rob Clark <robdclark@gmail.com>
+ * Daniel Vetter <daniel.vetter@ffwll.ch>
+ */
+
+#include <drm/drmP.h>
+#include <drm/drm_atomic.h>
+#include <drm/drm_plane_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_atomic_helper.h>
+#include <linux/fence.h>
+
+/**
+ * DOC: overview
+ *
+ * This helper library provides implementations of check and commit functions on
+ * top of the CRTC modeset helper callbacks and the plane helper callbacks. It
+ * also provides convenience implementations for the atomic state handling
+ * callbacks for drivers which don't need to subclass the drm core structures to
+ * add their own additional internal state.
+ *
+ * This library also provides default implementations for the check callback in
+ * drm_atomic_helper_check and for the commit callback with
+ * drm_atomic_helper_commit. But the individual stages and callbacks are expose
+ * to allow drivers to mix and match and e.g. use the plane helpers only
+ * together with a driver private modeset implementation.
+ *
+ * This library also provides implementations for all the legacy driver
+ * interfaces on top of the atomic interface. See drm_atomic_helper_set_config,
+ * drm_atomic_helper_disable_plane, drm_atomic_helper_disable_plane and the
+ * various functions to implement set_property callbacks. New drivers must not
+ * implement these functions themselves but must use the provided helpers.
+ */
+static void
+drm_atomic_helper_plane_changed(struct drm_atomic_state *state,
+				struct drm_plane_state *plane_state,
+				struct drm_plane *plane)
+{
+	struct drm_crtc_state *crtc_state;
+
+	if (plane->state->crtc) {
+		crtc_state = state->crtc_states[drm_crtc_index(plane->state->crtc)];
+
+		if (WARN_ON(!crtc_state))
+			return;
+
+		crtc_state->planes_changed = true;
+	}
+
+	if (plane_state->crtc) {
+		crtc_state =
+			state->crtc_states[drm_crtc_index(plane_state->crtc)];
+
+		if (WARN_ON(!crtc_state))
+			return;
+
+		crtc_state->planes_changed = true;
+	}
+}
+
+static struct drm_crtc *
+get_current_crtc_for_encoder(struct drm_device *dev,
+			     struct drm_encoder *encoder)
+{
+	struct drm_mode_config *config = &dev->mode_config;
+	struct drm_connector *connector;
+
+	WARN_ON(!drm_modeset_is_locked(&config->connection_mutex));
+
+	list_for_each_entry(connector, &config->connector_list, head) {
+		if (connector->state->best_encoder != encoder)
+			continue;
+
+		return connector->state->crtc;
+	}
+
+	return NULL;
+}
+
+static int
+steal_encoder(struct drm_atomic_state *state,
+	      struct drm_encoder *encoder,
+	      struct drm_crtc *encoder_crtc)
+{
+	struct drm_mode_config *config = &state->dev->mode_config;
+	struct drm_crtc_state *crtc_state;
+	struct drm_connector *connector;
+	struct drm_connector_state *connector_state;
+	int ret;
+
+	/*
+	 * We can only steal an encoder coming from a connector, which means we
+	 * must already hold the connection_mutex.
+	 */
+	WARN_ON(!drm_modeset_is_locked(&config->connection_mutex));
+
+	DRM_DEBUG_KMS("[ENCODER:%d:%s] in use on [CRTC:%d], stealing it\n",
+		      encoder->base.id, encoder->name,
+		      encoder_crtc->base.id);
+
+	crtc_state = drm_atomic_get_crtc_state(state, encoder_crtc);
+	if (IS_ERR(crtc_state))
+		return PTR_ERR(crtc_state);
+
+	crtc_state->mode_changed = true;
+
+	list_for_each_entry(connector, &config->connector_list, head) {
+		if (connector->state->best_encoder != encoder)
+			continue;
+
+		DRM_DEBUG_KMS("Stealing encoder from [CONNECTOR:%d:%s]\n",
+			      connector->base.id,
+			      connector->name);
+
+		connector_state = drm_atomic_get_connector_state(state,
+								 connector);
+		if (IS_ERR(connector_state))
+			return PTR_ERR(connector_state);
+
+		ret = drm_atomic_set_crtc_for_connector(connector_state, NULL);
+		if (ret)
+			return ret;
+		connector_state->best_encoder = NULL;
+	}
+
+	return 0;
+}
+
+static int
+update_connector_routing(struct drm_atomic_state *state, int conn_idx)
+{
+	struct drm_connector_helper_funcs *funcs;
+	struct drm_encoder *new_encoder;
+	struct drm_crtc *encoder_crtc;
+	struct drm_connector *connector;
+	struct drm_connector_state *connector_state;
+	struct drm_crtc_state *crtc_state;
+	int idx, ret;
+
+	connector = state->connectors[conn_idx];
+	connector_state = state->connector_states[conn_idx];
+
+	if (!connector)
+		return 0;
+
+	DRM_DEBUG_KMS("Updating routing for [CONNECTOR:%d:%s]\n",
+			connector->base.id,
+			connector->name);
+
+	if (connector->state->crtc != connector_state->crtc) {
+		if (connector->state->crtc) {
+			idx = drm_crtc_index(connector->state->crtc);
+
+			crtc_state = state->crtc_states[idx];
+			crtc_state->mode_changed = true;
+		}
+
+		if (connector_state->crtc) {
+			idx = drm_crtc_index(connector_state->crtc);
+
+			crtc_state = state->crtc_states[idx];
+			crtc_state->mode_changed = true;
+		}
+	}
+
+	if (!connector_state->crtc) {
+		DRM_DEBUG_KMS("Disabling [CONNECTOR:%d:%s]\n",
+				connector->base.id,
+				connector->name);
+
+		connector_state->best_encoder = NULL;
+
+		return 0;
+	}
+
+	funcs = connector->helper_private;
+	new_encoder = funcs->best_encoder(connector);
+
+	if (!new_encoder) {
+		DRM_DEBUG_KMS("No suitable encoder found for [CONNECTOR:%d:%s]\n",
+			      connector->base.id,
+			      connector->name);
+		return -EINVAL;
+	}
+
+	if (new_encoder == connector_state->best_encoder) {
+		DRM_DEBUG_KMS("[CONNECTOR:%d:%s] keeps [ENCODER:%d:%s], now on [CRTC:%d]\n",
+			      connector->base.id,
+			      connector->name,
+			      new_encoder->base.id,
+			      new_encoder->name,
+			      connector_state->crtc->base.id);
+
+		return 0;
+	}
+
+	encoder_crtc = get_current_crtc_for_encoder(state->dev,
+						    new_encoder);
+
+	if (encoder_crtc) {
+		ret = steal_encoder(state, new_encoder, encoder_crtc);
+		if (ret) {
+			DRM_DEBUG_KMS("Encoder stealing failed for [CONNECTOR:%d:%s]\n",
+				      connector->base.id,
+				      connector->name);
+			return ret;
+		}
+	}
+
+	connector_state->best_encoder = new_encoder;
+	idx = drm_crtc_index(connector_state->crtc);
+
+	crtc_state = state->crtc_states[idx];
+	crtc_state->mode_changed = true;
+
+	DRM_DEBUG_KMS("[CONNECTOR:%d:%s] using [ENCODER:%d:%s] on [CRTC:%d]\n",
+		      connector->base.id,
+		      connector->name,
+		      new_encoder->base.id,
+		      new_encoder->name,
+		      connector_state->crtc->base.id);
+
+	return 0;
+}
+
+static int
+mode_fixup(struct drm_atomic_state *state)
+{
+	int ncrtcs = state->dev->mode_config.num_crtc;
+	struct drm_crtc_state *crtc_state;
+	struct drm_connector_state *conn_state;
+	int i;
+	bool ret;
+
+	for (i = 0; i < ncrtcs; i++) {
+		crtc_state = state->crtc_states[i];
+
+		if (!crtc_state || !crtc_state->mode_changed)
+			continue;
+
+		drm_mode_copy(&crtc_state->adjusted_mode, &crtc_state->mode);
+	}
+
+	for (i = 0; i < state->num_connector; i++) {
+		struct drm_encoder_helper_funcs *funcs;
+		struct drm_encoder *encoder;
+
+		conn_state = state->connector_states[i];
+
+		if (!conn_state)
+			continue;
+
+		WARN_ON(!!conn_state->best_encoder != !!conn_state->crtc);
+
+		if (!conn_state->crtc || !conn_state->best_encoder)
+			continue;
+
+		crtc_state =
+			state->crtc_states[drm_crtc_index(conn_state->crtc)];
+
+		/*
+		 * Each encoder has at most one connector (since we always steal
+		 * it away), so we won't call ->mode_fixup twice.
+		 */
+		encoder = conn_state->best_encoder;
+		funcs = encoder->helper_private;
+
+		if (encoder->bridge && encoder->bridge->funcs->mode_fixup) {
+			ret = encoder->bridge->funcs->mode_fixup(
+					encoder->bridge, &crtc_state->mode,
+					&crtc_state->adjusted_mode);
+			if (!ret) {
+				DRM_DEBUG_KMS("Bridge fixup failed\n");
+				return -EINVAL;
+			}
+		}
+
+
+		ret = funcs->mode_fixup(encoder, &crtc_state->mode,
+					&crtc_state->adjusted_mode);
+		if (!ret) {
+			DRM_DEBUG_KMS("[ENCODER:%d:%s] fixup failed\n",
+				      encoder->base.id, encoder->name);
+			return -EINVAL;
+		}
+	}
+
+	for (i = 0; i < ncrtcs; i++) {
+		struct drm_crtc_helper_funcs *funcs;
+		struct drm_crtc *crtc;
+
+		crtc_state = state->crtc_states[i];
+		crtc = state->crtcs[i];
+
+		if (!crtc_state || !crtc_state->mode_changed)
+			continue;
+
+		funcs = crtc->helper_private;
+		ret = funcs->mode_fixup(crtc, &crtc_state->mode,
+					&crtc_state->adjusted_mode);
+		if (!ret) {
+			DRM_DEBUG_KMS("[CRTC:%d] fixup failed\n",
+				      crtc->base.id);
+			return -EINVAL;
+		}
+	}
+
+	return 0;
+}
+
+static int
+drm_atomic_helper_check_modeset(struct drm_device *dev,
+				struct drm_atomic_state *state)
+{
+	int ncrtcs = dev->mode_config.num_crtc;
+	struct drm_crtc *crtc;
+	struct drm_crtc_state *crtc_state;
+	int i, ret;
+
+	for (i = 0; i < ncrtcs; i++) {
+		crtc = state->crtcs[i];
+		crtc_state = state->crtc_states[i];
+
+		if (!crtc)
+			continue;
+
+		if (!drm_mode_equal(&crtc->state->mode, &crtc_state->mode)) {
+			DRM_DEBUG_KMS("[CRTC:%d] mode changed\n",
+				      crtc->base.id);
+			crtc_state->mode_changed = true;
+		}
+
+		if (crtc->state->enable != crtc_state->enable) {
+			DRM_DEBUG_KMS("[CRTC:%d] enable changed\n",
+				      crtc->base.id);
+			crtc_state->mode_changed = true;
+		}
+	}
+
+	for (i = 0; i < state->num_connector; i++) {
+		/*
+		 * This only sets crtc->mode_changed for routing changes,
+		 * drivers must set crtc->mode_changed themselves when connector
+		 * properties need to be updated.
+		 */
+		ret = update_connector_routing(state, i);
+		if (ret)
+			return ret;
+	}
+
+	/*
+	 * After all the routing has been prepared we need to add in any
+	 * connector which is itself unchanged, but who's crtc changes it's
+	 * configuration. This must be done before calling mode_fixup in case a
+	 * crtc only changed its mode but has the same set of connectors.
+	 */
+	for (i = 0; i < ncrtcs; i++) {
+		int num_connectors;
+
+		crtc = state->crtcs[i];
+		crtc_state = state->crtc_states[i];
+
+		if (!crtc || !crtc_state->mode_changed)
+			continue;
+
+		DRM_DEBUG_KMS("[CRTC:%d] needs full modeset, enable: %c\n",
+			      crtc->base.id,
+			      crtc_state->enable ? 'y' : 'n');
+
+		ret = drm_atomic_add_affected_connectors(state, crtc);
+		if (ret != 0)
+			return ret;
+
+		num_connectors = drm_atomic_connectors_for_crtc(state,
+								crtc);
+
+		if (crtc_state->enable != !!num_connectors) {
+			DRM_DEBUG_KMS("[CRTC:%d] enabled/connectors mismatch\n",
+				      crtc->base.id);
+
+			return -EINVAL;
+		}
+	}
+
+	return mode_fixup(state);
+}
+
+/**
+ * drm_atomic_helper_check - validate state object
+ * @dev: DRM device
+ * @state: the driver state object
+ *
+ * Check the state object to see if the requested state is physically possible.
+ * Only crtcs and planes have check callbacks, so for any additional (global)
+ * checking that a driver needs it can simply wrap that around this function.
+ * Drivers without such needs can directly use this as their ->atomic_check()
+ * callback.
+ *
+ * RETURNS
+ * Zero for success or -errno
+ */
+int drm_atomic_helper_check(struct drm_device *dev,
+			    struct drm_atomic_state *state)
+{
+	int nplanes = dev->mode_config.num_total_plane;
+	int ncrtcs = dev->mode_config.num_crtc;
+	int i, ret = 0;
+
+	for (i = 0; i < nplanes; i++) {
+		struct drm_plane_helper_funcs *funcs;
+		struct drm_plane *plane = state->planes[i];
+		struct drm_plane_state *plane_state = state->plane_states[i];
+
+		if (!plane)
+			continue;
+
+		funcs = plane->helper_private;
+
+		drm_atomic_helper_plane_changed(state, plane_state, plane);
+
+		if (!funcs || !funcs->atomic_check)
+			continue;
+
+		ret = funcs->atomic_check(plane, plane_state);
+		if (ret) {
+			DRM_DEBUG_KMS("[PLANE:%d] atomic check failed\n",
+				      plane->base.id);
+			return ret;
+		}
+	}
+
+	for (i = 0; i < ncrtcs; i++) {
+		struct drm_crtc_helper_funcs *funcs;
+		struct drm_crtc *crtc = state->crtcs[i];
+
+		if (!crtc)
+			continue;
+
+		funcs = crtc->helper_private;
+
+		if (!funcs || !funcs->atomic_check)
+			continue;
+
+		ret = funcs->atomic_check(crtc, state->crtc_states[i]);
+		if (ret) {
+			DRM_DEBUG_KMS("[CRTC:%d] atomic check failed\n",
+				      crtc->base.id);
+			return ret;
+		}
+	}
+
+	ret = drm_atomic_helper_check_modeset(dev, state);
+	if (ret)
+		return ret;
+
+	return ret;
+}
+EXPORT_SYMBOL(drm_atomic_helper_check);
+
+static void
+disable_outputs(struct drm_device *dev, struct drm_atomic_state *old_state)
+{
+	int ncrtcs = old_state->dev->mode_config.num_crtc;
+	int i;
+
+	for (i = 0; i < old_state->num_connector; i++) {
+		struct drm_connector_state *old_conn_state;
+		struct drm_connector *connector;
+		struct drm_encoder_helper_funcs *funcs;
+		struct drm_encoder *encoder;
+
+		old_conn_state = old_state->connector_states[i];
+		connector = old_state->connectors[i];
+
+		/* Shut down everything that's in the changeset and currently
+		 * still on. So need to check the old, saved state. */
+		if (!old_conn_state || !old_conn_state->crtc)
+			continue;
+
+		encoder = old_conn_state->best_encoder;
+
+		/* We shouldn't get this far if we didn't previously have
+		 * an encoder.. but WARN_ON() rather than explode.
+		 */
+		if (WARN_ON(!encoder))
+			continue;
+
+		funcs = encoder->helper_private;
+
+		/*
+		 * Each encoder has at most one connector (since we always steal
+		 * it away), so we won't call call disable hooks twice.
+		 */
+		if (encoder->bridge)
+			encoder->bridge->funcs->disable(encoder->bridge);
+
+		/* Right function depends upon target state. */
+		if (connector->state->crtc)
+			funcs->prepare(encoder);
+		else if (funcs->disable)
+			funcs->disable(encoder);
+		else
+			funcs->dpms(encoder, DRM_MODE_DPMS_OFF);
+
+		if (encoder->bridge)
+			encoder->bridge->funcs->post_disable(encoder->bridge);
+	}
+
+	for (i = 0; i < ncrtcs; i++) {
+		struct drm_crtc_helper_funcs *funcs;
+		struct drm_crtc *crtc;
+
+		crtc = old_state->crtcs[i];
+
+		/* Shut down everything that needs a full modeset. */
+		if (!crtc || !crtc->state->mode_changed)
+			continue;
+
+		funcs = crtc->helper_private;
+
+		/* Right function depends upon target state. */
+		if (crtc->state->enable)
+			funcs->prepare(crtc);
+		else if (funcs->disable)
+			funcs->disable(crtc);
+		else
+			funcs->dpms(crtc, DRM_MODE_DPMS_OFF);
+	}
+}
+
+static void
+set_routing_links(struct drm_device *dev, struct drm_atomic_state *old_state)
+{
+	int ncrtcs = old_state->dev->mode_config.num_crtc;
+	int i;
+
+	/* clear out existing links */
+	for (i = 0; i < old_state->num_connector; i++) {
+		struct drm_connector *connector;
+
+		connector = old_state->connectors[i];
+
+		if (!connector || !connector->encoder)
+			continue;
+
+		WARN_ON(!connector->encoder->crtc);
+
+		connector->encoder->crtc = NULL;
+		connector->encoder = NULL;
+	}
+
+	/* set new links */
+	for (i = 0; i < old_state->num_connector; i++) {
+		struct drm_connector *connector;
+
+		connector = old_state->connectors[i];
+
+		if (!connector || !connector->state->crtc)
+			continue;
+
+		if (WARN_ON(!connector->state->best_encoder))
+			continue;
+
+		connector->encoder = connector->state->best_encoder;
+		connector->encoder->crtc = connector->state->crtc;
+	}
+
+	/* set legacy state in the crtc structure */
+	for (i = 0; i < ncrtcs; i++) {
+		struct drm_crtc *crtc;
+
+		crtc = old_state->crtcs[i];
+
+		if (!crtc)
+			continue;
+
+		crtc->mode = crtc->state->mode;
+		crtc->enabled = crtc->state->enable;
+		crtc->x = crtc->primary->state->src_x >> 16;
+		crtc->y = crtc->primary->state->src_y >> 16;
+	}
+}
+
+static void
+crtc_set_mode(struct drm_device *dev, struct drm_atomic_state *old_state)
+{
+	int ncrtcs = old_state->dev->mode_config.num_crtc;
+	int i;
+
+	for (i = 0; i < ncrtcs; i++) {
+		struct drm_crtc_helper_funcs *funcs;
+		struct drm_crtc *crtc;
+
+		crtc = old_state->crtcs[i];
+
+		if (!crtc || !crtc->state->mode_changed)
+			continue;
+
+		funcs = crtc->helper_private;
+
+		if (crtc->state->enable)
+			funcs->mode_set_nofb(crtc);
+	}
+
+	for (i = 0; i < old_state->num_connector; i++) {
+		struct drm_connector *connector;
+		struct drm_crtc_state *new_crtc_state;
+		struct drm_encoder_helper_funcs *funcs;
+		struct drm_encoder *encoder;
+		struct drm_display_mode *mode, *adjusted_mode;
+
+		connector = old_state->connectors[i];
+
+		if (!connector || !connector->state->best_encoder)
+			continue;
+
+		encoder = connector->state->best_encoder;
+		funcs = encoder->helper_private;
+		new_crtc_state = connector->state->crtc->state;
+		mode = &new_crtc_state->mode;
+		adjusted_mode = &new_crtc_state->adjusted_mode;
+
+		/*
+		 * Each encoder has at most one connector (since we always steal
+		 * it away), so we won't call call mode_set hooks twice.
+		 */
+		funcs->mode_set(encoder, mode, adjusted_mode);
+
+		if (encoder->bridge && encoder->bridge->funcs->mode_set)
+			encoder->bridge->funcs->mode_set(encoder->bridge,
+							 mode, adjusted_mode);
+	}
+}
+
+/**
+ * drm_atomic_helper_commit_pre_planes - modeset commit before plane updates
+ * @dev: DRM device
+ * @state: atomic state
+ *
+ * This function commits the modeset changes that need to be committed before
+ * updating planes. It shuts down all the outputs that need to be shut down and
+ * prepares them (if required) with the new mode.
+ */
+void drm_atomic_helper_commit_pre_planes(struct drm_device *dev,
+					 struct drm_atomic_state *state)
+{
+	disable_outputs(dev, state);
+	set_routing_links(dev, state);
+	crtc_set_mode(dev, state);
+}
+EXPORT_SYMBOL(drm_atomic_helper_commit_pre_planes);
+
+/**
+ * drm_atomic_helper_commit_post_planes - modeset commit after plane updates
+ * @dev: DRM device
+ * @old_state: atomic state object with old state structures
+ *
+ * This function commits the modeset changes that need to be committed after
+ * updating planes: It enables all the outputs with the new configuration which
+ * had to be turned off for the update.
+ */
+void drm_atomic_helper_commit_post_planes(struct drm_device *dev,
+					  struct drm_atomic_state *old_state)
+{
+	int ncrtcs = old_state->dev->mode_config.num_crtc;
+	int i;
+
+	for (i = 0; i < ncrtcs; i++) {
+		struct drm_crtc_helper_funcs *funcs;
+		struct drm_crtc *crtc;
+
+		crtc = old_state->crtcs[i];
+
+		/* Need to filter out CRTCs where only planes change. */
+		if (!crtc || !crtc->state->mode_changed)
+			continue;
+
+		funcs = crtc->helper_private;
+
+		if (crtc->state->enable)
+			funcs->commit(crtc);
+	}
+
+	for (i = 0; i < old_state->num_connector; i++) {
+		struct drm_connector *connector;
+		struct drm_encoder_helper_funcs *funcs;
+		struct drm_encoder *encoder;
+
+		connector = old_state->connectors[i];
+
+		if (!connector || !connector->state->best_encoder)
+			continue;
+
+		encoder = connector->state->best_encoder;
+		funcs = encoder->helper_private;
+
+		/*
+		 * Each encoder has at most one connector (since we always steal
+		 * it away), so we won't call call enable hooks twice.
+		 */
+		if (encoder->bridge)
+			encoder->bridge->funcs->pre_enable(encoder->bridge);
+
+		funcs->commit(encoder);
+
+		if (encoder->bridge)
+			encoder->bridge->funcs->enable(encoder->bridge);
+	}
+}
+EXPORT_SYMBOL(drm_atomic_helper_commit_post_planes);
+
+static void wait_for_fences(struct drm_device *dev,
+			    struct drm_atomic_state *state)
+{
+	int nplanes = dev->mode_config.num_total_plane;
+	int i;
+
+	for (i = 0; i < nplanes; i++) {
+		struct drm_plane *plane = state->planes[i];
+
+		if (!plane || !plane->state->fence)
+			continue;
+
+		WARN_ON(!plane->state->fb);
+
+		fence_wait(plane->state->fence, false);
+		fence_put(plane->state->fence);
+		plane->state->fence = NULL;
+	}
+}
+
+static bool framebuffer_changed(struct drm_device *dev,
+				struct drm_atomic_state *old_state,
+				struct drm_crtc *crtc)
+{
+	struct drm_plane *plane;
+	struct drm_plane_state *old_plane_state;
+	int nplanes = old_state->dev->mode_config.num_total_plane;
+	int i;
+
+	for (i = 0; i < nplanes; i++) {
+		plane = old_state->planes[i];
+		old_plane_state = old_state->plane_states[i];
+
+		if (!plane)
+			continue;
+
+		if (plane->state->crtc != crtc &&
+		    old_plane_state->crtc != crtc)
+			continue;
+
+		if (plane->state->fb != old_plane_state->fb)
+			return true;
+	}
+
+	return false;
+}
+
+/**
+ * drm_atomic_helper_wait_for_vblanks - wait for vblank on crtcs
+ * @dev: DRM device
+ * @old_state: atomic state object with old state structures
+ *
+ * Helper to, after atomic commit, wait for vblanks on all effected
+ * crtcs (ie. before cleaning up old framebuffers using
+ * drm_atomic_helper_cleanup_planes()). It will only wait on crtcs where the
+ * framebuffers have actually changed to optimize for the legacy cursor and
+ * plane update use-case.
+ */
+void
+drm_atomic_helper_wait_for_vblanks(struct drm_device *dev,
+		struct drm_atomic_state *old_state)
+{
+	struct drm_crtc *crtc;
+	struct drm_crtc_state *old_crtc_state;
+	int ncrtcs = old_state->dev->mode_config.num_crtc;
+	int i, ret;
+
+	for (i = 0; i < ncrtcs; i++) {
+		crtc = old_state->crtcs[i];
+		old_crtc_state = old_state->crtc_states[i];
+
+		if (!crtc)
+			continue;
+
+		/* No one cares about the old state, so abuse it for tracking
+		 * and store whether we hold a vblank reference (and should do a
+		 * vblank wait) in the ->enable boolean. */
+		old_crtc_state->enable = false;
+
+		if (!crtc->state->enable)
+			continue;
+
+		if (!framebuffer_changed(dev, old_state, crtc))
+			continue;
+
+		ret = drm_crtc_vblank_get(crtc);
+		if (ret != 0)
+			continue;
+
+		old_crtc_state->enable = true;
+		old_crtc_state->last_vblank_count = drm_vblank_count(dev, i);
+	}
+
+	for (i = 0; i < ncrtcs; i++) {
+		crtc = old_state->crtcs[i];
+		old_crtc_state = old_state->crtc_states[i];
+
+		if (!crtc || !old_crtc_state->enable)
+			continue;
+
+		ret = wait_event_timeout(dev->vblank[i].queue,
+				old_crtc_state->last_vblank_count !=
+					drm_vblank_count(dev, i),
+				msecs_to_jiffies(50));
+
+		drm_crtc_vblank_put(crtc);
+	}
+}
+EXPORT_SYMBOL(drm_atomic_helper_wait_for_vblanks);
+
+/**
+ * drm_atomic_helper_commit - commit validated state object
+ * @dev: DRM device
+ * @state: the driver state object
+ * @async: asynchronous commit
+ *
+ * This function commits a with drm_atomic_helper_check() pre-validated state
+ * object. This can still fail when e.g. the framebuffer reservation fails. For
+ * now this doesn't implement asynchronous commits.
+ *
+ * RETURNS
+ * Zero for success or -errno.
+ */
+int drm_atomic_helper_commit(struct drm_device *dev,
+			     struct drm_atomic_state *state,
+			     bool async)
+{
+	int ret;
+
+	if (async)
+		return -EBUSY;
+
+	ret = drm_atomic_helper_prepare_planes(dev, state);
+	if (ret)
+		return ret;
+
+	/*
+	 * This is the point of no return - everything below never fails except
+	 * when the hw goes bonghits. Which means we can commit the new state on
+	 * the software side now.
+	 */
+
+	drm_atomic_helper_swap_state(dev, state);
+
+	/*
+	 * Everything below can be run asynchronously without the need to grab
+	 * any modeset locks at all under one conditions: It must be guaranteed
+	 * that the asynchronous work has either been cancelled (if the driver
+	 * supports it, which at least requires that the framebuffers get
+	 * cleaned up with drm_atomic_helper_cleanup_planes()) or completed
+	 * before the new state gets committed on the software side with
+	 * drm_atomic_helper_swap_state().
+	 *
+	 * This scheme allows new atomic state updates to be prepared and
+	 * checked in parallel to the asynchronous completion of the previous
+	 * update. Which is important since compositors need to figure out the
+	 * composition of the next frame right after having submitted the
+	 * current layout.
+	 */
+
+	wait_for_fences(dev, state);
+
+	drm_atomic_helper_commit_pre_planes(dev, state);
+
+	drm_atomic_helper_commit_planes(dev, state);
+
+	drm_atomic_helper_commit_post_planes(dev, state);
+
+	drm_atomic_helper_wait_for_vblanks(dev, state);
+
+	drm_atomic_helper_cleanup_planes(dev, state);
+
+	drm_atomic_state_free(state);
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_atomic_helper_commit);
+
+/**
+ * DOC: implementing async commit
+ *
+ * For now the atomic helpers don't support async commit directly. If there is
+ * real need it could be added though, using the dma-buf fence infrastructure
+ * for generic synchronization with outstanding rendering.
+ *
+ * For now drivers have to implement async commit themselves, with the following
+ * sequence being the recommended one:
+ *
+ * 1. Run drm_atomic_helper_prepare_planes() first. This is the only function
+ * which commit needs to call which can fail, so we want to run it first and
+ * synchronously.
+ *
+ * 2. Synchronize with any outstanding asynchronous commit worker threads which
+ * might be affected the new state update. This can be done by either cancelling
+ * or flushing the work items, depending upon whether the driver can deal with
+ * cancelled updates. Note that it is important to ensure that the framebuffer
+ * cleanup is still done when cancelling.
+ *
+ * For sufficient parallelism it is recommended to have a work item per crtc
+ * (for updates which don't touch global state) and a global one. Then we only
+ * need to synchronize with the crtc work items for changed crtcs and the global
+ * work item, which allows nice concurrent updates on disjoint sets of crtcs.
+ *
+ * 3. The software state is updated synchronously with
+ * drm_atomic_helper_swap_state. Doing this under the protection of all modeset
+ * locks means concurrent callers never see inconsistent state. And doing this
+ * while it's guaranteed that no relevant async worker runs means that async
+ * workers do not need grab any locks. Actually they must not grab locks, for
+ * otherwise the work flushing will deadlock.
+ *
+ * 4. Schedule a work item to do all subsequent steps, using the split-out
+ * commit helpers: a) pre-plane commit b) plane commit c) post-plane commit and
+ * then cleaning up the framebuffers after the old framebuffer is no longer
+ * being displayed.
+ */
+
+/**
+ * drm_atomic_helper_prepare_planes - prepare plane resources after commit
+ * @dev: DRM device
+ * @state: atomic state object with old state structures
+ *
+ * This function prepares plane state, specifically framebuffers, for the new
+ * configuration. If any failure is encountered this function will call
+ * ->cleanup_fb on any already successfully prepared framebuffer.
+ *
+ * Returns:
+ * 0 on success, negative error code on failure.
+ */
+int drm_atomic_helper_prepare_planes(struct drm_device *dev,
+				     struct drm_atomic_state *state)
+{
+	int nplanes = dev->mode_config.num_total_plane;
+	int ret, i;
+
+	for (i = 0; i < nplanes; i++) {
+		struct drm_plane_helper_funcs *funcs;
+		struct drm_plane *plane = state->planes[i];
+		struct drm_framebuffer *fb;
+
+		if (!plane)
+			continue;
+
+		funcs = plane->helper_private;
+
+		fb = state->plane_states[i]->fb;
+
+		if (fb && funcs->prepare_fb) {
+			ret = funcs->prepare_fb(plane, fb);
+			if (ret)
+				goto fail;
+		}
+	}
+
+	return 0;
+
+fail:
+	for (i--; i >= 0; i--) {
+		struct drm_plane_helper_funcs *funcs;
+		struct drm_plane *plane = state->planes[i];
+		struct drm_framebuffer *fb;
+
+		if (!plane)
+			continue;
+
+		funcs = plane->helper_private;
+
+		fb = state->plane_states[i]->fb;
+
+		if (fb && funcs->cleanup_fb)
+			funcs->cleanup_fb(plane, fb);
+
+	}
+
+	return ret;
+}
+EXPORT_SYMBOL(drm_atomic_helper_prepare_planes);
+
+/**
+ * drm_atomic_helper_commit_planes - commit plane state
+ * @dev: DRM device
+ * @old_state: atomic state object with old state structures
+ *
+ * This function commits the new plane state using the plane and atomic helper
+ * functions for planes and crtcs. It assumes that the atomic state has already
+ * been pushed into the relevant object state pointers, since this step can no
+ * longer fail.
+ *
+ * It still requires the global state object @old_state to know which planes and
+ * crtcs need to be updated though.
+ */
+void drm_atomic_helper_commit_planes(struct drm_device *dev,
+				     struct drm_atomic_state *old_state)
+{
+	int nplanes = dev->mode_config.num_total_plane;
+	int ncrtcs = dev->mode_config.num_crtc;
+	int i;
+
+	for (i = 0; i < ncrtcs; i++) {
+		struct drm_crtc_helper_funcs *funcs;
+		struct drm_crtc *crtc = old_state->crtcs[i];
+
+		if (!crtc)
+			continue;
+
+		funcs = crtc->helper_private;
+
+		if (!funcs || !funcs->atomic_begin)
+			continue;
+
+		funcs->atomic_begin(crtc);
+	}
+
+	for (i = 0; i < nplanes; i++) {
+		struct drm_plane_helper_funcs *funcs;
+		struct drm_plane *plane = old_state->planes[i];
+		struct drm_plane_state *old_plane_state;
+
+		if (!plane)
+			continue;
+
+		funcs = plane->helper_private;
+
+		if (!funcs || !funcs->atomic_update)
+			continue;
+
+		old_plane_state = old_state->plane_states[i];
+
+		funcs->atomic_update(plane, old_plane_state);
+	}
+
+	for (i = 0; i < ncrtcs; i++) {
+		struct drm_crtc_helper_funcs *funcs;
+		struct drm_crtc *crtc = old_state->crtcs[i];
+
+		if (!crtc)
+			continue;
+
+		funcs = crtc->helper_private;
+
+		if (!funcs || !funcs->atomic_flush)
+			continue;
+
+		funcs->atomic_flush(crtc);
+	}
+}
+EXPORT_SYMBOL(drm_atomic_helper_commit_planes);
+
+/**
+ * drm_atomic_helper_cleanup_planes - cleanup plane resources after commit
+ * @dev: DRM device
+ * @old_state: atomic state object with old state structures
+ *
+ * This function cleans up plane state, specifically framebuffers, from the old
+ * configuration. Hence the old configuration must be perserved in @old_state to
+ * be able to call this function.
+ *
+ * This function must also be called on the new state when the atomic update
+ * fails at any point after calling drm_atomic_helper_prepare_planes().
+ */
+void drm_atomic_helper_cleanup_planes(struct drm_device *dev,
+				      struct drm_atomic_state *old_state)
+{
+	int nplanes = dev->mode_config.num_total_plane;
+	int i;
+
+	for (i = 0; i < nplanes; i++) {
+		struct drm_plane_helper_funcs *funcs;
+		struct drm_plane *plane = old_state->planes[i];
+		struct drm_framebuffer *old_fb;
+
+		if (!plane)
+			continue;
+
+		funcs = plane->helper_private;
+
+		old_fb = old_state->plane_states[i]->fb;
+
+		if (old_fb && funcs->cleanup_fb)
+			funcs->cleanup_fb(plane, old_fb);
+	}
+}
+EXPORT_SYMBOL(drm_atomic_helper_cleanup_planes);
+
+/**
+ * drm_atomic_helper_swap_state - store atomic state into current sw state
+ * @dev: DRM device
+ * @state: atomic state
+ *
+ * This function stores the atomic state into the current state pointers in all
+ * driver objects. It should be called after all failing steps have been done
+ * and succeeded, but before the actual hardware state is committed.
+ *
+ * For cleanup and error recovery the current state for all changed objects will
+ * be swaped into @state.
+ *
+ * With that sequence it fits perfectly into the plane prepare/cleanup sequence:
+ *
+ * 1. Call drm_atomic_helper_prepare_planes() with the staged atomic state.
+ *
+ * 2. Do any other steps that might fail.
+ *
+ * 3. Put the staged state into the current state pointers with this function.
+ *
+ * 4. Actually commit the hardware state.
+ *
+ * 5. Call drm_atomic_helper_cleanup_planes with @state, which since step 3
+ * contains the old state. Also do any other cleanup required with that state.
+ */
+void drm_atomic_helper_swap_state(struct drm_device *dev,
+				  struct drm_atomic_state *state)
+{
+	int i;
+
+	for (i = 0; i < dev->mode_config.num_connector; i++) {
+		struct drm_connector *connector = state->connectors[i];
+
+		if (!connector)
+			continue;
+
+		connector->state->state = state;
+		swap(state->connector_states[i], connector->state);
+		connector->state->state = NULL;
+	}
+
+	for (i = 0; i < dev->mode_config.num_crtc; i++) {
+		struct drm_crtc *crtc = state->crtcs[i];
+
+		if (!crtc)
+			continue;
+
+		crtc->state->state = state;
+		swap(state->crtc_states[i], crtc->state);
+		crtc->state->state = NULL;
+	}
+
+	for (i = 0; i < dev->mode_config.num_total_plane; i++) {
+		struct drm_plane *plane = state->planes[i];
+
+		if (!plane)
+			continue;
+
+		plane->state->state = state;
+		swap(state->plane_states[i], plane->state);
+		plane->state->state = NULL;
+	}
+}
+EXPORT_SYMBOL(drm_atomic_helper_swap_state);
+
+/**
+ * drm_atomic_helper_update_plane - Helper for primary plane update using atomic
+ * @plane: plane object to update
+ * @crtc: owning CRTC of owning plane
+ * @fb: framebuffer to flip onto plane
+ * @crtc_x: x offset of primary plane on crtc
+ * @crtc_y: y offset of primary plane on crtc
+ * @crtc_w: width of primary plane rectangle on crtc
+ * @crtc_h: height of primary plane rectangle on crtc
+ * @src_x: x offset of @fb for panning
+ * @src_y: y offset of @fb for panning
+ * @src_w: width of source rectangle in @fb
+ * @src_h: height of source rectangle in @fb
+ *
+ * Provides a default plane update handler using the atomic driver interface.
+ *
+ * RETURNS:
+ * Zero on success, error code on failure
+ */
+int drm_atomic_helper_update_plane(struct drm_plane *plane,
+				   struct drm_crtc *crtc,
+				   struct drm_framebuffer *fb,
+				   int crtc_x, int crtc_y,
+				   unsigned int crtc_w, unsigned int crtc_h,
+				   uint32_t src_x, uint32_t src_y,
+				   uint32_t src_w, uint32_t src_h)
+{
+	struct drm_atomic_state *state;
+	struct drm_plane_state *plane_state;
+	int ret = 0;
+
+	state = drm_atomic_state_alloc(plane->dev);
+	if (!state)
+		return -ENOMEM;
+
+	state->acquire_ctx = drm_modeset_legacy_acquire_ctx(crtc);
+retry:
+	plane_state = drm_atomic_get_plane_state(state, plane);
+	if (IS_ERR(plane_state)) {
+		ret = PTR_ERR(plane_state);
+		goto fail;
+	}
+
+	ret = drm_atomic_set_crtc_for_plane(state, plane, crtc);
+	if (ret != 0)
+		goto fail;
+	drm_atomic_set_fb_for_plane(plane_state, fb);
+	plane_state->crtc_x = crtc_x;
+	plane_state->crtc_y = crtc_y;
+	plane_state->crtc_h = crtc_h;
+	plane_state->crtc_w = crtc_w;
+	plane_state->src_x = src_x;
+	plane_state->src_y = src_y;
+	plane_state->src_h = src_h;
+	plane_state->src_w = src_w;
+
+	ret = drm_atomic_commit(state);
+	if (ret != 0)
+		goto fail;
+
+	/* Driver takes ownership of state on successful commit. */
+	return 0;
+fail:
+	if (ret == -EDEADLK)
+		goto backoff;
+
+	drm_atomic_state_free(state);
+
+	return ret;
+backoff:
+	drm_atomic_state_clear(state);
+	drm_atomic_legacy_backoff(state);
+
+	/*
+	 * Someone might have exchanged the framebuffer while we dropped locks
+	 * in the backoff code. We need to fix up the fb refcount tracking the
+	 * core does for us.
+	 */
+	plane->old_fb = plane->fb;
+
+	goto retry;
+}
+EXPORT_SYMBOL(drm_atomic_helper_update_plane);
+
+/**
+ * drm_atomic_helper_disable_plane - Helper for primary plane disable using * atomic
+ * @plane: plane to disable
+ *
+ * Provides a default plane disable handler using the atomic driver interface.
+ *
+ * RETURNS:
+ * Zero on success, error code on failure
+ */
+int drm_atomic_helper_disable_plane(struct drm_plane *plane)
+{
+	struct drm_atomic_state *state;
+	struct drm_plane_state *plane_state;
+	int ret = 0;
+
+	/*
+	 * FIXME: Without plane->crtc set we can't get at the implicit legacy
+	 * acquire context. The real fix will be to wire the acquire ctx through
+	 * everywhere we need it, but meanwhile prevent chaos by just skipping
+	 * this noop. The critical case is the cursor ioctls which a) only grab
+	 * crtc/cursor-plane locks (so we need the crtc to get at the right
+	 * acquire context) and b) can try to disable the plane multiple times.
+	 */
+	if (!plane->crtc)
+		return 0;
+
+	state = drm_atomic_state_alloc(plane->dev);
+	if (!state)
+		return -ENOMEM;
+
+	state->acquire_ctx = drm_modeset_legacy_acquire_ctx(plane->crtc);
+retry:
+	plane_state = drm_atomic_get_plane_state(state, plane);
+	if (IS_ERR(plane_state)) {
+		ret = PTR_ERR(plane_state);
+		goto fail;
+	}
+
+	ret = drm_atomic_set_crtc_for_plane(state, plane, NULL);
+	if (ret != 0)
+		goto fail;
+	drm_atomic_set_fb_for_plane(plane_state, NULL);
+	plane_state->crtc_x = 0;
+	plane_state->crtc_y = 0;
+	plane_state->crtc_h = 0;
+	plane_state->crtc_w = 0;
+	plane_state->src_x = 0;
+	plane_state->src_y = 0;
+	plane_state->src_h = 0;
+	plane_state->src_w = 0;
+
+	ret = drm_atomic_commit(state);
+	if (ret != 0)
+		goto fail;
+
+	/* Driver takes ownership of state on successful commit. */
+	return 0;
+fail:
+	if (ret == -EDEADLK)
+		goto backoff;
+
+	drm_atomic_state_free(state);
+
+	return ret;
+backoff:
+	drm_atomic_state_clear(state);
+	drm_atomic_legacy_backoff(state);
+
+	/*
+	 * Someone might have exchanged the framebuffer while we dropped locks
+	 * in the backoff code. We need to fix up the fb refcount tracking the
+	 * core does for us.
+	 */
+	plane->old_fb = plane->fb;
+
+	goto retry;
+}
+EXPORT_SYMBOL(drm_atomic_helper_disable_plane);
+
+static int update_output_state(struct drm_atomic_state *state,
+			       struct drm_mode_set *set)
+{
+	struct drm_device *dev = set->crtc->dev;
+	struct drm_connector_state *conn_state;
+	int ncrtcs = state->dev->mode_config.num_crtc;
+	int ret, i, j;
+
+	ret = drm_modeset_lock(&dev->mode_config.connection_mutex,
+			       state->acquire_ctx);
+	if (ret)
+		return ret;
+
+	/* First grab all affected connector/crtc states. */
+	for (i = 0; i < set->num_connectors; i++) {
+		conn_state = drm_atomic_get_connector_state(state,
+							    set->connectors[i]);
+		if (IS_ERR(conn_state))
+			return PTR_ERR(conn_state);
+	}
+
+	for (i = 0; i < ncrtcs; i++) {
+		struct drm_crtc *crtc = state->crtcs[i];
+
+		if (!crtc)
+			continue;
+
+		ret = drm_atomic_add_affected_connectors(state, crtc);
+		if (ret)
+			return ret;
+	}
+
+	/* Then recompute connector->crtc links and crtc enabling state. */
+	for (i = 0; i < state->num_connector; i++) {
+		struct drm_connector *connector;
+
+		connector = state->connectors[i];
+		conn_state = state->connector_states[i];
+
+		if (!connector)
+			continue;
+
+		if (conn_state->crtc == set->crtc) {
+			ret = drm_atomic_set_crtc_for_connector(conn_state,
+								NULL);
+			if (ret)
+				return ret;
+		}
+
+		for (j = 0; j < set->num_connectors; j++) {
+			if (set->connectors[j] == connector) {
+				ret = drm_atomic_set_crtc_for_connector(conn_state,
+									set->crtc);
+				if (ret)
+					return ret;
+				break;
+			}
+		}
+	}
+
+	for (i = 0; i < ncrtcs; i++) {
+		struct drm_crtc *crtc = state->crtcs[i];
+		struct drm_crtc_state *crtc_state = state->crtc_states[i];
+
+		if (!crtc)
+			continue;
+
+		/* Don't update ->enable for the CRTC in the set_config request,
+		 * since a mismatch would indicate a bug in the upper layers.
+		 * The actual modeset code later on will catch any
+		 * inconsistencies here. */
+		if (crtc == set->crtc)
+			continue;
+
+		crtc_state->enable =
+			drm_atomic_connectors_for_crtc(state, crtc);
+	}
+
+	return 0;
+}
+
+/**
+ * drm_atomic_helper_set_config - set a new config from userspace
+ * @set: mode set configuration
+ *
+ * Provides a default crtc set_config handler using the atomic driver interface.
+ *
+ * Returns:
+ * Returns 0 on success, negative errno numbers on failure.
+ */
+int drm_atomic_helper_set_config(struct drm_mode_set *set)
+{
+	struct drm_atomic_state *state;
+	struct drm_crtc *crtc = set->crtc;
+	struct drm_crtc_state *crtc_state;
+	struct drm_plane_state *primary_state;
+	int ret = 0;
+
+	state = drm_atomic_state_alloc(crtc->dev);
+	if (!state)
+		return -ENOMEM;
+
+	state->acquire_ctx = drm_modeset_legacy_acquire_ctx(crtc);
+retry:
+	crtc_state = drm_atomic_get_crtc_state(state, crtc);
+	if (IS_ERR(crtc_state)) {
+		ret = PTR_ERR(crtc_state);
+		goto fail;
+	}
+
+	primary_state = drm_atomic_get_plane_state(state, crtc->primary);
+	if (IS_ERR(primary_state)) {
+		ret = PTR_ERR(primary_state);
+		goto fail;
+	}
+
+	if (!set->mode) {
+		WARN_ON(set->fb);
+		WARN_ON(set->num_connectors);
+
+		crtc_state->enable = false;
+
+		ret = drm_atomic_set_crtc_for_plane(state, crtc->primary, NULL);
+		if (ret != 0)
+			goto fail;
+
+		drm_atomic_set_fb_for_plane(primary_state, NULL);
+
+		goto commit;
+	}
+
+	WARN_ON(!set->fb);
+	WARN_ON(!set->num_connectors);
+
+	crtc_state->enable = true;
+	drm_mode_copy(&crtc_state->mode, set->mode);
+
+	ret = drm_atomic_set_crtc_for_plane(state, crtc->primary, crtc);
+	if (ret != 0)
+		goto fail;
+	drm_atomic_set_fb_for_plane(primary_state, set->fb);
+	primary_state->crtc_x = 0;
+	primary_state->crtc_y = 0;
+	primary_state->crtc_h = set->mode->vdisplay;
+	primary_state->crtc_w = set->mode->hdisplay;
+	primary_state->src_x = set->x << 16;
+	primary_state->src_y = set->y << 16;
+	primary_state->src_h = set->mode->vdisplay << 16;
+	primary_state->src_w = set->mode->hdisplay << 16;
+
+commit:
+	ret = update_output_state(state, set);
+	if (ret)
+		goto fail;
+
+	ret = drm_atomic_commit(state);
+	if (ret != 0)
+		goto fail;
+
+	/* Driver takes ownership of state on successful commit. */
+	return 0;
+fail:
+	if (ret == -EDEADLK)
+		goto backoff;
+
+	drm_atomic_state_free(state);
+
+	return ret;
+backoff:
+	drm_atomic_state_clear(state);
+	drm_atomic_legacy_backoff(state);
+
+	/*
+	 * Someone might have exchanged the framebuffer while we dropped locks
+	 * in the backoff code. We need to fix up the fb refcount tracking the
+	 * core does for us.
+	 */
+	crtc->primary->old_fb = crtc->primary->fb;
+
+	goto retry;
+}
+EXPORT_SYMBOL(drm_atomic_helper_set_config);
+
+/**
+ * drm_atomic_helper_crtc_set_property - helper for crtc prorties
+ * @crtc: DRM crtc
+ * @property: DRM property
+ * @val: value of property
+ *
+ * Provides a default plane disablle handler using the atomic driver interface.
+ *
+ * RETURNS:
+ * Zero on success, error code on failure
+ */
+int
+drm_atomic_helper_crtc_set_property(struct drm_crtc *crtc,
+				    struct drm_property *property,
+				    uint64_t val)
+{
+	struct drm_atomic_state *state;
+	struct drm_crtc_state *crtc_state;
+	int ret = 0;
+
+	state = drm_atomic_state_alloc(crtc->dev);
+	if (!state)
+		return -ENOMEM;
+
+	/* ->set_property is always called with all locks held. */
+	state->acquire_ctx = crtc->dev->mode_config.acquire_ctx;
+retry:
+	crtc_state = drm_atomic_get_crtc_state(state, crtc);
+	if (IS_ERR(crtc_state)) {
+		ret = PTR_ERR(crtc_state);
+		goto fail;
+	}
+
+	ret = crtc->funcs->atomic_set_property(crtc, crtc_state,
+					       property, val);
+	if (ret)
+		goto fail;
+
+	ret = drm_atomic_commit(state);
+	if (ret != 0)
+		goto fail;
+
+	/* Driver takes ownership of state on successful commit. */
+	return 0;
+fail:
+	if (ret == -EDEADLK)
+		goto backoff;
+
+	drm_atomic_state_free(state);
+
+	return ret;
+backoff:
+	drm_atomic_state_clear(state);
+	drm_atomic_legacy_backoff(state);
+
+	goto retry;
+}
+EXPORT_SYMBOL(drm_atomic_helper_crtc_set_property);
+
+/**
+ * drm_atomic_helper_plane_set_property - helper for plane prorties
+ * @plane: DRM plane
+ * @property: DRM property
+ * @val: value of property
+ *
+ * Provides a default plane disable handler using the atomic driver interface.
+ *
+ * RETURNS:
+ * Zero on success, error code on failure
+ */
+int
+drm_atomic_helper_plane_set_property(struct drm_plane *plane,
+				    struct drm_property *property,
+				    uint64_t val)
+{
+	struct drm_atomic_state *state;
+	struct drm_plane_state *plane_state;
+	int ret = 0;
+
+	state = drm_atomic_state_alloc(plane->dev);
+	if (!state)
+		return -ENOMEM;
+
+	/* ->set_property is always called with all locks held. */
+	state->acquire_ctx = plane->dev->mode_config.acquire_ctx;
+retry:
+	plane_state = drm_atomic_get_plane_state(state, plane);
+	if (IS_ERR(plane_state)) {
+		ret = PTR_ERR(plane_state);
+		goto fail;
+	}
+
+	ret = plane->funcs->atomic_set_property(plane, plane_state,
+					       property, val);
+	if (ret)
+		goto fail;
+
+	ret = drm_atomic_commit(state);
+	if (ret != 0)
+		goto fail;
+
+	/* Driver takes ownership of state on successful commit. */
+	return 0;
+fail:
+	if (ret == -EDEADLK)
+		goto backoff;
+
+	drm_atomic_state_free(state);
+
+	return ret;
+backoff:
+	drm_atomic_state_clear(state);
+	drm_atomic_legacy_backoff(state);
+
+	goto retry;
+}
+EXPORT_SYMBOL(drm_atomic_helper_plane_set_property);
+
+/**
+ * drm_atomic_helper_connector_set_property - helper for connector prorties
+ * @connector: DRM connector
+ * @property: DRM property
+ * @val: value of property
+ *
+ * Provides a default plane disablle handler using the atomic driver interface.
+ *
+ * RETURNS:
+ * Zero on success, error code on failure
+ */
+int
+drm_atomic_helper_connector_set_property(struct drm_connector *connector,
+				    struct drm_property *property,
+				    uint64_t val)
+{
+	struct drm_atomic_state *state;
+	struct drm_connector_state *connector_state;
+	int ret = 0;
+
+	state = drm_atomic_state_alloc(connector->dev);
+	if (!state)
+		return -ENOMEM;
+
+	/* ->set_property is always called with all locks held. */
+	state->acquire_ctx = connector->dev->mode_config.acquire_ctx;
+retry:
+	connector_state = drm_atomic_get_connector_state(state, connector);
+	if (IS_ERR(connector_state)) {
+		ret = PTR_ERR(connector_state);
+		goto fail;
+	}
+
+	ret = connector->funcs->atomic_set_property(connector, connector_state,
+					       property, val);
+	if (ret)
+		goto fail;
+
+	ret = drm_atomic_commit(state);
+	if (ret != 0)
+		goto fail;
+
+	/* Driver takes ownership of state on successful commit. */
+	return 0;
+fail:
+	if (ret == -EDEADLK)
+		goto backoff;
+
+	drm_atomic_state_free(state);
+
+	return ret;
+backoff:
+	drm_atomic_state_clear(state);
+	drm_atomic_legacy_backoff(state);
+
+	goto retry;
+}
+EXPORT_SYMBOL(drm_atomic_helper_connector_set_property);
+
+/**
+ * drm_atomic_helper_page_flip - execute a legacy page flip
+ * @crtc: DRM crtc
+ * @fb: DRM framebuffer
+ * @event: optional DRM event to signal upon completion
+ * @flags: flip flags for non-vblank sync'ed updates
+ *
+ * Provides a default page flip implementation using the atomic driver interface.
+ *
+ * Note that for now so called async page flips (i.e. updates which are not
+ * synchronized to vblank) are not supported, since the atomic interfaces have
+ * no provisions for this yet.
+ *
+ * Returns:
+ * Returns 0 on success, negative errno numbers on failure.
+ */
+int drm_atomic_helper_page_flip(struct drm_crtc *crtc,
+				struct drm_framebuffer *fb,
+				struct drm_pending_vblank_event *event,
+				uint32_t flags)
+{
+	struct drm_plane *plane = crtc->primary;
+	struct drm_atomic_state *state;
+	struct drm_plane_state *plane_state;
+	struct drm_crtc_state *crtc_state;
+	int ret = 0;
+
+	if (flags & DRM_MODE_PAGE_FLIP_ASYNC)
+		return -EINVAL;
+
+	state = drm_atomic_state_alloc(plane->dev);
+	if (!state)
+		return -ENOMEM;
+
+	state->acquire_ctx = drm_modeset_legacy_acquire_ctx(crtc);
+retry:
+	crtc_state = drm_atomic_get_crtc_state(state, crtc);
+	if (IS_ERR(crtc_state)) {
+		ret = PTR_ERR(crtc_state);
+		goto fail;
+	}
+	crtc_state->event = event;
+
+	plane_state = drm_atomic_get_plane_state(state, plane);
+	if (IS_ERR(plane_state)) {
+		ret = PTR_ERR(plane_state);
+		goto fail;
+	}
+
+	ret = drm_atomic_set_crtc_for_plane(state, plane, crtc);
+	if (ret != 0)
+		goto fail;
+	drm_atomic_set_fb_for_plane(plane_state, fb);
+
+	ret = drm_atomic_async_commit(state);
+	if (ret != 0)
+		goto fail;
+
+	/* TODO: ->page_flip is the only driver callback where the core
+	 * doesn't update plane->fb. For now patch it up here. */
+	plane->fb = plane->state->fb;
+
+	/* Driver takes ownership of state on successful async commit. */
+	return 0;
+fail:
+	if (ret == -EDEADLK)
+		goto backoff;
+
+	drm_atomic_state_free(state);
+
+	return ret;
+backoff:
+	drm_atomic_state_clear(state);
+	drm_atomic_legacy_backoff(state);
+
+	/*
+	 * Someone might have exchanged the framebuffer while we dropped locks
+	 * in the backoff code. We need to fix up the fb refcount tracking the
+	 * core does for us.
+	 */
+	plane->old_fb = plane->fb;
+
+	goto retry;
+}
+EXPORT_SYMBOL(drm_atomic_helper_page_flip);
+
+/**
+ * DOC: atomic state reset and initialization
+ *
+ * Both the drm core and the atomic helpers assume that there is always the full
+ * and correct atomic software state for all connectors, CRTCs and planes
+ * available. Which is a bit a problem on driver load and also after system
+ * suspend. One way to solve this is to have a hardware state read-out
+ * infrastructure which reconstructs the full software state (e.g. the i915
+ * driver).
+ *
+ * The simpler solution is to just reset the software state to everything off,
+ * which is easiest to do by calling drm_mode_config_reset(). To facilitate this
+ * the atomic helpers provide default reset implementations for all hooks.
+ */
+
+/**
+ * drm_atomic_helper_crtc_reset - default ->reset hook for CRTCs
+ * @crtc: drm CRTC
+ *
+ * Resets the atomic state for @crtc by freeing the state pointer (which might
+ * be NULL, e.g. at driver load time) and allocating a new empty state object.
+ */
+void drm_atomic_helper_crtc_reset(struct drm_crtc *crtc)
+{
+	kfree(crtc->state);
+	crtc->state = kzalloc(sizeof(*crtc->state), GFP_KERNEL);
+}
+EXPORT_SYMBOL(drm_atomic_helper_crtc_reset);
+
+/**
+ * drm_atomic_helper_crtc_duplicate_state - default state duplicate hook
+ * @crtc: drm CRTC
+ *
+ * Default CRTC state duplicate hook for drivers which don't have their own
+ * subclassed CRTC state structure.
+ */
+struct drm_crtc_state *
+drm_atomic_helper_crtc_duplicate_state(struct drm_crtc *crtc)
+{
+	struct drm_crtc_state *state;
+
+	if (WARN_ON(!crtc->state))
+		return NULL;
+
+	state = kmemdup(crtc->state, sizeof(*crtc->state), GFP_KERNEL);
+
+	if (state) {
+		state->mode_changed = false;
+		state->planes_changed = false;
+		state->event = NULL;
+	}
+
+	return state;
+}
+EXPORT_SYMBOL(drm_atomic_helper_crtc_duplicate_state);
+
+/**
+ * drm_atomic_helper_crtc_destroy_state - default state destroy hook
+ * @crtc: drm CRTC
+ * @state: CRTC state object to release
+ *
+ * Default CRTC state destroy hook for drivers which don't have their own
+ * subclassed CRTC state structure.
+ */
+void drm_atomic_helper_crtc_destroy_state(struct drm_crtc *crtc,
+					  struct drm_crtc_state *state)
+{
+	kfree(state);
+}
+EXPORT_SYMBOL(drm_atomic_helper_crtc_destroy_state);
+
+/**
+ * drm_atomic_helper_plane_reset - default ->reset hook for planes
+ * @plane: drm plane
+ *
+ * Resets the atomic state for @plane by freeing the state pointer (which might
+ * be NULL, e.g. at driver load time) and allocating a new empty state object.
+ */
+void drm_atomic_helper_plane_reset(struct drm_plane *plane)
+{
+	if (plane->state && plane->state->fb)
+		drm_framebuffer_unreference(plane->state->fb);
+
+	kfree(plane->state);
+	plane->state = kzalloc(sizeof(*plane->state), GFP_KERNEL);
+}
+EXPORT_SYMBOL(drm_atomic_helper_plane_reset);
+
+/**
+ * drm_atomic_helper_plane_duplicate_state - default state duplicate hook
+ * @plane: drm plane
+ *
+ * Default plane state duplicate hook for drivers which don't have their own
+ * subclassed plane state structure.
+ */
+struct drm_plane_state *
+drm_atomic_helper_plane_duplicate_state(struct drm_plane *plane)
+{
+	struct drm_plane_state *state;
+
+	if (WARN_ON(!plane->state))
+		return NULL;
+
+	state = kmemdup(plane->state, sizeof(*plane->state), GFP_KERNEL);
+
+	if (state && state->fb)
+		drm_framebuffer_reference(state->fb);
+
+	return state;
+}
+EXPORT_SYMBOL(drm_atomic_helper_plane_duplicate_state);
+
+/**
+ * drm_atomic_helper_plane_destroy_state - default state destroy hook
+ * @plane: drm plane
+ * @state: plane state object to release
+ *
+ * Default plane state destroy hook for drivers which don't have their own
+ * subclassed plane state structure.
+ */
+void drm_atomic_helper_plane_destroy_state(struct drm_plane *plane,
+					   struct drm_plane_state *state)
+{
+	if (state->fb)
+		drm_framebuffer_unreference(state->fb);
+
+	kfree(state);
+}
+EXPORT_SYMBOL(drm_atomic_helper_plane_destroy_state);
+
+/**
+ * drm_atomic_helper_connector_reset - default ->reset hook for connectors
+ * @connector: drm connector
+ *
+ * Resets the atomic state for @connector by freeing the state pointer (which
+ * might be NULL, e.g. at driver load time) and allocating a new empty state
+ * object.
+ */
+void drm_atomic_helper_connector_reset(struct drm_connector *connector)
+{
+	kfree(connector->state);
+	connector->state = kzalloc(sizeof(*connector->state), GFP_KERNEL);
+}
+EXPORT_SYMBOL(drm_atomic_helper_connector_reset);
+
+/**
+ * drm_atomic_helper_connector_duplicate_state - default state duplicate hook
+ * @connector: drm connector
+ *
+ * Default connector state duplicate hook for drivers which don't have their own
+ * subclassed connector state structure.
+ */
+struct drm_connector_state *
+drm_atomic_helper_connector_duplicate_state(struct drm_connector *connector)
+{
+	if (WARN_ON(!connector->state))
+		return NULL;
+
+	return kmemdup(connector->state, sizeof(*connector->state), GFP_KERNEL);
+}
+EXPORT_SYMBOL(drm_atomic_helper_connector_duplicate_state);
+
+/**
+ * drm_atomic_helper_connector_destroy_state - default state destroy hook
+ * @connector: drm connector
+ * @state: connector state object to release
+ *
+ * Default connector state destroy hook for drivers which don't have their own
+ * subclassed connector state structure.
+ */
+void drm_atomic_helper_connector_destroy_state(struct drm_connector *connector,
+					  struct drm_connector_state *state)
+{
+	kfree(state);
+}
+EXPORT_SYMBOL(drm_atomic_helper_connector_destroy_state);
diff --git a/drivers/gpu/drm/drm_auth.c b/drivers/gpu/drm/drm_auth.c
index 3cedae12b3c1..fc8e8aaa34fb 100644
--- a/drivers/gpu/drm/drm_auth.c
+++ b/drivers/gpu/drm/drm_auth.c
@@ -34,6 +34,13 @@
  */
 
 #include <drm/drmP.h>
+#include "drm_internal.h"
+
+struct drm_magic_entry {
+	struct list_head head;
+	struct drm_hash_item hash_item;
+	struct drm_file *priv;
+};
 
 /**
  * Find the file with the given magic number.
diff --git a/drivers/gpu/drm/drm_bufs.c b/drivers/gpu/drm/drm_bufs.c
index 61acb8f6756d..569064a00693 100644
--- a/drivers/gpu/drm/drm_bufs.c
+++ b/drivers/gpu/drm/drm_bufs.c
@@ -1,18 +1,13 @@
-/**
- * \file drm_bufs.c
- * Generic buffer template
- *
- * \author Rickard E. (Rik) Faith <faith@valinux.com>
- * \author Gareth Hughes <gareth@valinux.com>
- */
-
 /*
- * Created: Thu Nov 23 03:10:50 2000 by gareth@valinux.com
+ * Legacy: Generic DRM Buffer Management
  *
  * Copyright 1999, 2000 Precision Insight, Inc., Cedar Park, Texas.
  * Copyright 2000 VA Linux Systems, Inc., Sunnyvale, California.
  * All Rights Reserved.
  *
+ * Author: Rickard E. (Rik) Faith <faith@valinux.com>
+ * Author: Gareth Hughes <gareth@valinux.com>
+ *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
  * to deal in the Software without restriction, including without limitation
@@ -39,6 +34,7 @@
 #include <linux/export.h>
 #include <asm/shmparam.h>
 #include <drm/drmP.h>
+#include "drm_legacy.h"
 
 static struct drm_map_list *drm_find_matching_map(struct drm_device *dev,
 						  struct drm_local_map *map)
@@ -365,9 +361,9 @@ static int drm_addmap_core(struct drm_device * dev, resource_size_t offset,
 	return 0;
 }
 
-int drm_addmap(struct drm_device * dev, resource_size_t offset,
-	       unsigned int size, enum drm_map_type type,
-	       enum drm_map_flags flags, struct drm_local_map ** map_ptr)
+int drm_legacy_addmap(struct drm_device * dev, resource_size_t offset,
+		      unsigned int size, enum drm_map_type type,
+		      enum drm_map_flags flags, struct drm_local_map **map_ptr)
 {
 	struct drm_map_list *list;
 	int rc;
@@ -377,8 +373,7 @@ int drm_addmap(struct drm_device * dev, resource_size_t offset,
 		*map_ptr = list->map;
 	return rc;
 }
-
-EXPORT_SYMBOL(drm_addmap);
+EXPORT_SYMBOL(drm_legacy_addmap);
 
 /**
  * Ioctl to specify a range of memory that is available for mapping by a
@@ -391,8 +386,8 @@ EXPORT_SYMBOL(drm_addmap);
  * \return zero on success or a negative value on error.
  *
  */
-int drm_addmap_ioctl(struct drm_device *dev, void *data,
-		     struct drm_file *file_priv)
+int drm_legacy_addmap_ioctl(struct drm_device *dev, void *data,
+			    struct drm_file *file_priv)
 {
 	struct drm_map *map = data;
 	struct drm_map_list *maplist;
@@ -429,9 +424,9 @@ int drm_addmap_ioctl(struct drm_device *dev, void *data,
  * its being used, and free any associate resource (such as MTRR's) if it's not
  * being on use.
  *
- * \sa drm_addmap
+ * \sa drm_legacy_addmap
  */
-int drm_rmmap_locked(struct drm_device *dev, struct drm_local_map *map)
+int drm_legacy_rmmap_locked(struct drm_device *dev, struct drm_local_map *map)
 {
 	struct drm_map_list *r_list = NULL, *list_t;
 	drm_dma_handle_t dmah;
@@ -478,26 +473,26 @@ int drm_rmmap_locked(struct drm_device *dev, struct drm_local_map *map)
 		dmah.vaddr = map->handle;
 		dmah.busaddr = map->offset;
 		dmah.size = map->size;
-		__drm_pci_free(dev, &dmah);
+		__drm_legacy_pci_free(dev, &dmah);
 		break;
 	}
 	kfree(map);
 
 	return 0;
 }
-EXPORT_SYMBOL(drm_rmmap_locked);
+EXPORT_SYMBOL(drm_legacy_rmmap_locked);
 
-int drm_rmmap(struct drm_device *dev, struct drm_local_map *map)
+int drm_legacy_rmmap(struct drm_device *dev, struct drm_local_map *map)
 {
 	int ret;
 
 	mutex_lock(&dev->struct_mutex);
-	ret = drm_rmmap_locked(dev, map);
+	ret = drm_legacy_rmmap_locked(dev, map);
 	mutex_unlock(&dev->struct_mutex);
 
 	return ret;
 }
-EXPORT_SYMBOL(drm_rmmap);
+EXPORT_SYMBOL(drm_legacy_rmmap);
 
 /* The rmmap ioctl appears to be unnecessary.  All mappings are torn down on
  * the last close of the device, and this is necessary for cleanup when things
@@ -514,8 +509,8 @@ EXPORT_SYMBOL(drm_rmmap);
  * \param arg pointer to a struct drm_map structure.
  * \return zero on success or a negative value on error.
  */
-int drm_rmmap_ioctl(struct drm_device *dev, void *data,
-		    struct drm_file *file_priv)
+int drm_legacy_rmmap_ioctl(struct drm_device *dev, void *data,
+			   struct drm_file *file_priv)
 {
 	struct drm_map *request = data;
 	struct drm_local_map *map = NULL;
@@ -546,7 +541,7 @@ int drm_rmmap_ioctl(struct drm_device *dev, void *data,
 		return 0;
 	}
 
-	ret = drm_rmmap_locked(dev, map);
+	ret = drm_legacy_rmmap_locked(dev, map);
 
 	mutex_unlock(&dev->struct_mutex);
 
@@ -599,7 +594,8 @@ static void drm_cleanup_buf_error(struct drm_device * dev,
  * reallocates the buffer list of the same size order to accommodate the new
  * buffers.
  */
-int drm_addbufs_agp(struct drm_device * dev, struct drm_buf_desc * request)
+int drm_legacy_addbufs_agp(struct drm_device *dev,
+			   struct drm_buf_desc *request)
 {
 	struct drm_device_dma *dma = dev->dma;
 	struct drm_buf_entry *entry;
@@ -759,10 +755,11 @@ int drm_addbufs_agp(struct drm_device * dev, struct drm_buf_desc * request)
 	atomic_dec(&dev->buf_alloc);
 	return 0;
 }
-EXPORT_SYMBOL(drm_addbufs_agp);
+EXPORT_SYMBOL(drm_legacy_addbufs_agp);
 #endif				/* __OS_HAS_AGP */
 
-int drm_addbufs_pci(struct drm_device * dev, struct drm_buf_desc * request)
+int drm_legacy_addbufs_pci(struct drm_device *dev,
+			   struct drm_buf_desc *request)
 {
 	struct drm_device_dma *dma = dev->dma;
 	int count;
@@ -964,9 +961,10 @@ int drm_addbufs_pci(struct drm_device * dev, struct drm_buf_desc * request)
 	return 0;
 
 }
-EXPORT_SYMBOL(drm_addbufs_pci);
+EXPORT_SYMBOL(drm_legacy_addbufs_pci);
 
-static int drm_addbufs_sg(struct drm_device * dev, struct drm_buf_desc * request)
+static int drm_legacy_addbufs_sg(struct drm_device *dev,
+				 struct drm_buf_desc *request)
 {
 	struct drm_device_dma *dma = dev->dma;
 	struct drm_buf_entry *entry;
@@ -1135,8 +1133,8 @@ static int drm_addbufs_sg(struct drm_device * dev, struct drm_buf_desc * request
  * addbufs_sg() or addbufs_pci() for AGP, scatter-gather or consistent
  * PCI memory respectively.
  */
-int drm_addbufs(struct drm_device *dev, void *data,
-		struct drm_file *file_priv)
+int drm_legacy_addbufs(struct drm_device *dev, void *data,
+		       struct drm_file *file_priv)
 {
 	struct drm_buf_desc *request = data;
 	int ret;
@@ -1149,15 +1147,15 @@ int drm_addbufs(struct drm_device *dev, void *data,
 
 #if __OS_HAS_AGP
 	if (request->flags & _DRM_AGP_BUFFER)
-		ret = drm_addbufs_agp(dev, request);
+		ret = drm_legacy_addbufs_agp(dev, request);
 	else
 #endif
 	if (request->flags & _DRM_SG_BUFFER)
-		ret = drm_addbufs_sg(dev, request);
+		ret = drm_legacy_addbufs_sg(dev, request);
 	else if (request->flags & _DRM_FB_BUFFER)
 		ret = -EINVAL;
 	else
-		ret = drm_addbufs_pci(dev, request);
+		ret = drm_legacy_addbufs_pci(dev, request);
 
 	return ret;
 }
@@ -1179,8 +1177,8 @@ int drm_addbufs(struct drm_device *dev, void *data,
  * lock, preventing of allocating more buffers after this call. Information
  * about each requested buffer is then copied into user space.
  */
-int drm_infobufs(struct drm_device *dev, void *data,
-		 struct drm_file *file_priv)
+int drm_legacy_infobufs(struct drm_device *dev, void *data,
+			struct drm_file *file_priv)
 {
 	struct drm_device_dma *dma = dev->dma;
 	struct drm_buf_info *request = data;
@@ -1260,8 +1258,8 @@ int drm_infobufs(struct drm_device *dev, void *data,
  *
  * \note This ioctl is deprecated and mostly never used.
  */
-int drm_markbufs(struct drm_device *dev, void *data,
-		 struct drm_file *file_priv)
+int drm_legacy_markbufs(struct drm_device *dev, void *data,
+			struct drm_file *file_priv)
 {
 	struct drm_device_dma *dma = dev->dma;
 	struct drm_buf_desc *request = data;
@@ -1307,8 +1305,8 @@ int drm_markbufs(struct drm_device *dev, void *data,
  * Calls free_buffer() for each used buffer.
  * This function is primarily used for debugging.
  */
-int drm_freebufs(struct drm_device *dev, void *data,
-		 struct drm_file *file_priv)
+int drm_legacy_freebufs(struct drm_device *dev, void *data,
+			struct drm_file *file_priv)
 {
 	struct drm_device_dma *dma = dev->dma;
 	struct drm_buf_free *request = data;
@@ -1340,7 +1338,7 @@ int drm_freebufs(struct drm_device *dev, void *data,
 				  task_pid_nr(current));
 			return -EINVAL;
 		}
-		drm_free_buffer(dev, buf);
+		drm_legacy_free_buffer(dev, buf);
 	}
 
 	return 0;
@@ -1360,8 +1358,8 @@ int drm_freebufs(struct drm_device *dev, void *data,
  * offset equal to 0, which drm_mmap() interpretes as PCI buffers and calls
  * drm_mmap_dma().
  */
-int drm_mapbufs(struct drm_device *dev, void *data,
-	        struct drm_file *file_priv)
+int drm_legacy_mapbufs(struct drm_device *dev, void *data,
+		       struct drm_file *file_priv)
 {
 	struct drm_device_dma *dma = dev->dma;
 	int retcode = 0;
@@ -1448,7 +1446,7 @@ int drm_mapbufs(struct drm_device *dev, void *data,
 	return retcode;
 }
 
-int drm_dma_ioctl(struct drm_device *dev, void *data,
+int drm_legacy_dma_ioctl(struct drm_device *dev, void *data,
 		  struct drm_file *file_priv)
 {
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
@@ -1460,7 +1458,7 @@ int drm_dma_ioctl(struct drm_device *dev, void *data,
 		return -EINVAL;
 }
 
-struct drm_local_map *drm_getsarea(struct drm_device *dev)
+struct drm_local_map *drm_legacy_getsarea(struct drm_device *dev)
 {
 	struct drm_map_list *entry;
 
@@ -1472,4 +1470,4 @@ struct drm_local_map *drm_getsarea(struct drm_device *dev)
 	}
 	return NULL;
 }
-EXPORT_SYMBOL(drm_getsarea);
+EXPORT_SYMBOL(drm_legacy_getsarea);
diff --git a/drivers/gpu/drm/drm_crtc.c b/drivers/gpu/drm/drm_crtc.c
index 90e773019eac..5213da499d39 100644
--- a/drivers/gpu/drm/drm_crtc.c
+++ b/drivers/gpu/drm/drm_crtc.c
@@ -40,106 +40,12 @@
 #include <drm/drm_modeset_lock.h>
 
 #include "drm_crtc_internal.h"
+#include "drm_internal.h"
 
 static struct drm_framebuffer *add_framebuffer_internal(struct drm_device *dev,
 							struct drm_mode_fb_cmd2 *r,
 							struct drm_file *file_priv);
 
-/**
- * drm_modeset_lock_all - take all modeset locks
- * @dev: drm device
- *
- * This function takes all modeset locks, suitable where a more fine-grained
- * scheme isn't (yet) implemented. Locks must be dropped with
- * drm_modeset_unlock_all.
- */
-void drm_modeset_lock_all(struct drm_device *dev)
-{
-	struct drm_mode_config *config = &dev->mode_config;
-	struct drm_modeset_acquire_ctx *ctx;
-	int ret;
-
-	ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
-	if (WARN_ON(!ctx))
-		return;
-
-	mutex_lock(&config->mutex);
-
-	drm_modeset_acquire_init(ctx, 0);
-
-retry:
-	ret = drm_modeset_lock(&config->connection_mutex, ctx);
-	if (ret)
-		goto fail;
-	ret = drm_modeset_lock_all_crtcs(dev, ctx);
-	if (ret)
-		goto fail;
-
-	WARN_ON(config->acquire_ctx);
-
-	/* now we hold the locks, so now that it is safe, stash the
-	 * ctx for drm_modeset_unlock_all():
-	 */
-	config->acquire_ctx = ctx;
-
-	drm_warn_on_modeset_not_all_locked(dev);
-
-	return;
-
-fail:
-	if (ret == -EDEADLK) {
-		drm_modeset_backoff(ctx);
-		goto retry;
-	}
-}
-EXPORT_SYMBOL(drm_modeset_lock_all);
-
-/**
- * drm_modeset_unlock_all - drop all modeset locks
- * @dev: device
- *
- * This function drop all modeset locks taken by drm_modeset_lock_all.
- */
-void drm_modeset_unlock_all(struct drm_device *dev)
-{
-	struct drm_mode_config *config = &dev->mode_config;
-	struct drm_modeset_acquire_ctx *ctx = config->acquire_ctx;
-
-	if (WARN_ON(!ctx))
-		return;
-
-	config->acquire_ctx = NULL;
-	drm_modeset_drop_locks(ctx);
-	drm_modeset_acquire_fini(ctx);
-
-	kfree(ctx);
-
-	mutex_unlock(&dev->mode_config.mutex);
-}
-EXPORT_SYMBOL(drm_modeset_unlock_all);
-
-/**
- * drm_warn_on_modeset_not_all_locked - check that all modeset locks are locked
- * @dev: device
- *
- * Useful as a debug assert.
- */
-void drm_warn_on_modeset_not_all_locked(struct drm_device *dev)
-{
-	struct drm_crtc *crtc;
-
-	/* Locking is currently fubar in the panic handler. */
-	if (oops_in_progress)
-		return;
-
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
-		WARN_ON(!drm_modeset_is_locked(&crtc->mutex));
-
-	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
-	WARN_ON(!mutex_is_locked(&dev->mode_config.mutex));
-}
-EXPORT_SYMBOL(drm_warn_on_modeset_not_all_locked);
-
 /* Avoid boilerplate.  I'm tired of typing. */
 #define DRM_ENUM_NAME_FN(fnname, list)				\
 	const char *fnname(int val)				\
@@ -515,9 +421,6 @@ int drm_framebuffer_init(struct drm_device *dev, struct drm_framebuffer *fb,
 	if (ret)
 		goto out;
 
-	/* Grab the idr reference. */
-	drm_framebuffer_reference(fb);
-
 	dev->mode_config.num_fb++;
 	list_add(&fb->head, &dev->mode_config.fb_list);
 out:
@@ -527,10 +430,34 @@ out:
 }
 EXPORT_SYMBOL(drm_framebuffer_init);
 
+/* dev->mode_config.fb_lock must be held! */
+static void __drm_framebuffer_unregister(struct drm_device *dev,
+					 struct drm_framebuffer *fb)
+{
+	mutex_lock(&dev->mode_config.idr_mutex);
+	idr_remove(&dev->mode_config.crtc_idr, fb->base.id);
+	mutex_unlock(&dev->mode_config.idr_mutex);
+
+	fb->base.id = 0;
+}
+
 static void drm_framebuffer_free(struct kref *kref)
 {
 	struct drm_framebuffer *fb =
 			container_of(kref, struct drm_framebuffer, refcount);
+	struct drm_device *dev = fb->dev;
+
+	/*
+	 * The lookup idr holds a weak reference, which has not necessarily been
+	 * removed at this point. Check for that.
+	 */
+	mutex_lock(&dev->mode_config.fb_lock);
+	if (fb->base.id) {
+		/* Mark fb as reaped and drop idr ref. */
+		__drm_framebuffer_unregister(dev, fb);
+	}
+	mutex_unlock(&dev->mode_config.fb_lock);
+
 	fb->funcs->destroy(fb);
 }
 
@@ -567,8 +494,10 @@ struct drm_framebuffer *drm_framebuffer_lookup(struct drm_device *dev,
 
 	mutex_lock(&dev->mode_config.fb_lock);
 	fb = __drm_framebuffer_lookup(dev, id);
-	if (fb)
-		drm_framebuffer_reference(fb);
+	if (fb) {
+		if (!kref_get_unless_zero(&fb->refcount))
+			fb = NULL;
+	}
 	mutex_unlock(&dev->mode_config.fb_lock);
 
 	return fb;
@@ -612,19 +541,6 @@ static void __drm_framebuffer_unreference(struct drm_framebuffer *fb)
 	kref_put(&fb->refcount, drm_framebuffer_free_bug);
 }
 
-/* dev->mode_config.fb_lock must be held! */
-static void __drm_framebuffer_unregister(struct drm_device *dev,
-					 struct drm_framebuffer *fb)
-{
-	mutex_lock(&dev->mode_config.idr_mutex);
-	idr_remove(&dev->mode_config.crtc_idr, fb->base.id);
-	mutex_unlock(&dev->mode_config.idr_mutex);
-
-	fb->base.id = 0;
-
-	__drm_framebuffer_unreference(fb);
-}
-
 /**
  * drm_framebuffer_unregister_private - unregister a private fb from the lookup idr
  * @fb: fb to unregister
@@ -764,14 +680,10 @@ int drm_crtc_init_with_planes(struct drm_device *dev, struct drm_crtc *crtc,
 	crtc->funcs = funcs;
 	crtc->invert_dimensions = false;
 
-	drm_modeset_lock_all(dev);
 	drm_modeset_lock_init(&crtc->mutex);
-	/* dropped by _unlock_all(): */
-	drm_modeset_lock(&crtc->mutex, config->acquire_ctx);
-
 	ret = drm_mode_object_get(dev, &crtc->base, DRM_MODE_OBJECT_CRTC);
 	if (ret)
-		goto out;
+		return ret;
 
 	crtc->base.properties = &crtc->properties;
 
@@ -785,10 +697,7 @@ int drm_crtc_init_with_planes(struct drm_device *dev, struct drm_crtc *crtc,
 	if (cursor)
 		cursor->possible_crtcs = 1 << drm_crtc_index(crtc);
 
- out:
-	drm_modeset_unlock_all(dev);
-
-	return ret;
+	return 0;
 }
 EXPORT_SYMBOL(drm_crtc_init_with_planes);
 
@@ -812,6 +721,12 @@ void drm_crtc_cleanup(struct drm_crtc *crtc)
 	drm_mode_object_put(dev, &crtc->base);
 	list_del(&crtc->head);
 	dev->mode_config.num_crtc--;
+
+	WARN_ON(crtc->state && !crtc->funcs->atomic_destroy_state);
+	if (crtc->state && crtc->funcs->atomic_destroy_state)
+		crtc->funcs->atomic_destroy_state(crtc, crtc->state);
+
+	memset(crtc, 0, sizeof(*crtc));
 }
 EXPORT_SYMBOL(drm_crtc_cleanup);
 
@@ -853,6 +768,58 @@ static void drm_mode_remove(struct drm_connector *connector,
 }
 
 /**
+ * drm_connector_get_cmdline_mode - reads the user's cmdline mode
+ * @connector: connector to quwery
+ *
+ * The kernel supports per-connector configration of its consoles through
+ * use of the video= parameter. This function parses that option and
+ * extracts the user's specified mode (or enable/disable status) for a
+ * particular connector. This is typically only used during the early fbdev
+ * setup.
+ */
+static void drm_connector_get_cmdline_mode(struct drm_connector *connector)
+{
+	struct drm_cmdline_mode *mode = &connector->cmdline_mode;
+	char *option = NULL;
+
+	if (fb_get_options(connector->name, &option))
+		return;
+
+	if (!drm_mode_parse_command_line_for_connector(option,
+						       connector,
+						       mode))
+		return;
+
+	if (mode->force) {
+		const char *s;
+
+		switch (mode->force) {
+		case DRM_FORCE_OFF:
+			s = "OFF";
+			break;
+		case DRM_FORCE_ON_DIGITAL:
+			s = "ON - dig";
+			break;
+		default:
+		case DRM_FORCE_ON:
+			s = "ON";
+			break;
+		}
+
+		DRM_INFO("forcing %s connector %s\n", connector->name, s);
+		connector->force = mode->force;
+	}
+
+	DRM_DEBUG_KMS("cmdline mode for connector %s %dx%d@%dHz%s%s%s\n",
+		      connector->name,
+		      mode->xres, mode->yres,
+		      mode->refresh_specified ? mode->refresh : 60,
+		      mode->rb ? " reduced blanking" : "",
+		      mode->margins ? " with margins" : "",
+		      mode->interlace ?  " interlaced" : "");
+}
+
+/**
  * drm_connector_init - Init a preallocated connector
  * @dev: DRM device
  * @connector: the connector to init
@@ -904,6 +871,10 @@ int drm_connector_init(struct drm_device *dev,
 	connector->edid_blob_ptr = NULL;
 	connector->status = connector_status_unknown;
 
+	drm_connector_get_cmdline_mode(connector);
+
+	/* We should add connectors at the end to avoid upsetting the connector
+	 * index too much. */
 	list_add_tail(&connector->head, &dev->mode_config.connector_list);
 	dev->mode_config.num_connector++;
 
@@ -939,6 +910,11 @@ void drm_connector_cleanup(struct drm_connector *connector)
 	struct drm_device *dev = connector->dev;
 	struct drm_display_mode *mode, *t;
 
+	if (connector->tile_group) {
+		drm_mode_put_tile_group(dev, connector->tile_group);
+		connector->tile_group = NULL;
+	}
+
 	list_for_each_entry_safe(mode, t, &connector->probed_modes, head)
 		drm_mode_remove(connector, mode);
 
@@ -953,10 +929,43 @@ void drm_connector_cleanup(struct drm_connector *connector)
 	connector->name = NULL;
 	list_del(&connector->head);
 	dev->mode_config.num_connector--;
+
+	WARN_ON(connector->state && !connector->funcs->atomic_destroy_state);
+	if (connector->state && connector->funcs->atomic_destroy_state)
+		connector->funcs->atomic_destroy_state(connector,
+						       connector->state);
+
+	memset(connector, 0, sizeof(*connector));
 }
 EXPORT_SYMBOL(drm_connector_cleanup);
 
 /**
+ * drm_connector_index - find the index of a registered connector
+ * @connector: connector to find index for
+ *
+ * Given a registered connector, return the index of that connector within a DRM
+ * device's list of connectors.
+ */
+unsigned int drm_connector_index(struct drm_connector *connector)
+{
+	unsigned int index = 0;
+	struct drm_connector *tmp;
+	struct drm_mode_config *config = &connector->dev->mode_config;
+
+	WARN_ON(!drm_modeset_is_locked(&config->connection_mutex));
+
+	list_for_each_entry(tmp, &connector->dev->mode_config.connector_list, head) {
+		if (tmp == connector)
+			return index;
+
+		index++;
+	}
+
+	BUG();
+}
+EXPORT_SYMBOL(drm_connector_index);
+
+/**
  * drm_connector_register - register a connector
  * @connector: the connector to register
  *
@@ -1068,6 +1077,8 @@ void drm_bridge_cleanup(struct drm_bridge *bridge)
 	list_del(&bridge->head);
 	dev->mode_config.num_bridge--;
 	drm_modeset_unlock_all(dev);
+
+	memset(bridge, 0, sizeof(*bridge));
 }
 EXPORT_SYMBOL(drm_bridge_cleanup);
 
@@ -1134,10 +1145,11 @@ void drm_encoder_cleanup(struct drm_encoder *encoder)
 	drm_modeset_lock_all(dev);
 	drm_mode_object_put(dev, &encoder->base);
 	kfree(encoder->name);
-	encoder->name = NULL;
 	list_del(&encoder->head);
 	dev->mode_config.num_encoder--;
 	drm_modeset_unlock_all(dev);
+
+	memset(encoder, 0, sizeof(*encoder));
 }
 EXPORT_SYMBOL(drm_encoder_cleanup);
 
@@ -1164,11 +1176,11 @@ int drm_universal_plane_init(struct drm_device *dev, struct drm_plane *plane,
 {
 	int ret;
 
-	drm_modeset_lock_all(dev);
-
 	ret = drm_mode_object_get(dev, &plane->base, DRM_MODE_OBJECT_PLANE);
 	if (ret)
-		goto out;
+		return ret;
+
+	drm_modeset_lock_init(&plane->mutex);
 
 	plane->base.properties = &plane->properties;
 	plane->dev = dev;
@@ -1178,8 +1190,7 @@ int drm_universal_plane_init(struct drm_device *dev, struct drm_plane *plane,
 	if (!plane->format_types) {
 		DRM_DEBUG_KMS("out of memory when allocating plane\n");
 		drm_mode_object_put(dev, &plane->base);
-		ret = -ENOMEM;
-		goto out;
+		return -ENOMEM;
 	}
 
 	memcpy(plane->format_types, formats, format_count * sizeof(uint32_t));
@@ -1196,10 +1207,7 @@ int drm_universal_plane_init(struct drm_device *dev, struct drm_plane *plane,
 				   dev->mode_config.plane_type_property,
 				   plane->type);
 
- out:
-	drm_modeset_unlock_all(dev);
-
-	return ret;
+	return 0;
 }
 EXPORT_SYMBOL(drm_universal_plane_init);
 
@@ -1257,10 +1265,39 @@ void drm_plane_cleanup(struct drm_plane *plane)
 	if (plane->type == DRM_PLANE_TYPE_OVERLAY)
 		dev->mode_config.num_overlay_plane--;
 	drm_modeset_unlock_all(dev);
+
+	WARN_ON(plane->state && !plane->funcs->atomic_destroy_state);
+	if (plane->state && plane->funcs->atomic_destroy_state)
+		plane->funcs->atomic_destroy_state(plane, plane->state);
+
+	memset(plane, 0, sizeof(*plane));
 }
 EXPORT_SYMBOL(drm_plane_cleanup);
 
 /**
+ * drm_plane_index - find the index of a registered plane
+ * @plane: plane to find index for
+ *
+ * Given a registered plane, return the index of that CRTC within a DRM
+ * device's list of planes.
+ */
+unsigned int drm_plane_index(struct drm_plane *plane)
+{
+	unsigned int index = 0;
+	struct drm_plane *tmp;
+
+	list_for_each_entry(tmp, &plane->dev->mode_config.plane_list, head) {
+		if (tmp == plane)
+			return index;
+
+		index++;
+	}
+
+	BUG();
+}
+EXPORT_SYMBOL(drm_plane_index);
+
+/**
  * drm_plane_force_disable - Forcibly disable a plane
  * @plane: plane to disable
  *
@@ -1271,19 +1308,21 @@ EXPORT_SYMBOL(drm_plane_cleanup);
  */
 void drm_plane_force_disable(struct drm_plane *plane)
 {
-	struct drm_framebuffer *old_fb = plane->fb;
 	int ret;
 
-	if (!old_fb)
+	if (!plane->fb)
 		return;
 
+	plane->old_fb = plane->fb;
 	ret = plane->funcs->disable_plane(plane);
 	if (ret) {
 		DRM_ERROR("failed to disable plane with busy fb\n");
+		plane->old_fb = NULL;
 		return;
 	}
 	/* disconnect the plane from the fb and crtc: */
-	__drm_framebuffer_unreference(old_fb);
+	__drm_framebuffer_unreference(plane->old_fb);
+	plane->old_fb = NULL;
 	plane->fb = NULL;
 	plane->crtc = NULL;
 }
@@ -1314,6 +1353,11 @@ static int drm_mode_create_standard_connector_properties(struct drm_device *dev)
 				       "PATH", 0);
 	dev->mode_config.path_property = dev_path;
 
+	dev->mode_config.tile_property = drm_property_create(dev,
+							     DRM_MODE_PROP_BLOB |
+							     DRM_MODE_PROP_IMMUTABLE,
+							     "TILE", 0);
+
 	return 0;
 }
 
@@ -1374,12 +1418,13 @@ EXPORT_SYMBOL(drm_mode_create_dvi_i_properties);
  * responsible for allocating a list of format names and passing them to
  * this routine.
  */
-int drm_mode_create_tv_properties(struct drm_device *dev, int num_modes,
+int drm_mode_create_tv_properties(struct drm_device *dev,
+				  unsigned int num_modes,
 				  char *modes[])
 {
 	struct drm_property *tv_selector;
 	struct drm_property *tv_subconnector;
-	int i;
+	unsigned int i;
 
 	if (dev->mode_config.tv_select_subconnector_property)
 		return 0;
@@ -1477,7 +1522,7 @@ EXPORT_SYMBOL(drm_mode_create_scaling_mode_property);
  * connectors.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_create_aspect_ratio_property(struct drm_device *dev)
 {
@@ -1521,6 +1566,30 @@ int drm_mode_create_dirty_info_property(struct drm_device *dev)
 }
 EXPORT_SYMBOL(drm_mode_create_dirty_info_property);
 
+/**
+ * drm_mode_create_suggested_offset_properties - create suggests offset properties
+ * @dev: DRM device
+ *
+ * Create the the suggested x/y offset property for connectors.
+ */
+int drm_mode_create_suggested_offset_properties(struct drm_device *dev)
+{
+	if (dev->mode_config.suggested_x_property && dev->mode_config.suggested_y_property)
+		return 0;
+
+	dev->mode_config.suggested_x_property =
+		drm_property_create_range(dev, DRM_MODE_PROP_IMMUTABLE, "suggested X", 0, 0xffffffff);
+
+	dev->mode_config.suggested_y_property =
+		drm_property_create_range(dev, DRM_MODE_PROP_IMMUTABLE, "suggested Y", 0, 0xffffffff);
+
+	if (dev->mode_config.suggested_x_property == NULL ||
+	    dev->mode_config.suggested_y_property == NULL)
+		return -ENOMEM;
+	return 0;
+}
+EXPORT_SYMBOL(drm_mode_create_suggested_offset_properties);
+
 static int drm_mode_group_init(struct drm_device *dev, struct drm_mode_group *group)
 {
 	uint32_t total_objects = 0;
@@ -1637,7 +1706,7 @@ static void drm_crtc_convert_to_umode(struct drm_mode_modeinfo *out,
  * the caller.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 static int drm_crtc_convert_umode(struct drm_display_mode *out,
 				  const struct drm_mode_modeinfo *in)
@@ -1680,7 +1749,7 @@ static int drm_crtc_convert_umode(struct drm_display_mode *out,
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_getresources(struct drm_device *dev, void *data,
 			  struct drm_file *file_priv)
@@ -1731,7 +1800,9 @@ int drm_mode_getresources(struct drm_device *dev, void *data,
 	card_res->count_fbs = fb_count;
 	mutex_unlock(&file_priv->fbs_lock);
 
-	drm_modeset_lock_all(dev);
+	/* mode_config.mutex protects the connector list against e.g. DP MST
+	 * connector hot-adding. CRTC/Plane lists are invariant. */
+	mutex_lock(&dev->mode_config.mutex);
 	if (!drm_is_primary_client(file_priv)) {
 
 		mode_group = NULL;
@@ -1851,7 +1922,7 @@ int drm_mode_getresources(struct drm_device *dev, void *data,
 		  card_res->count_connectors, card_res->count_encoders);
 
 out:
-	drm_modeset_unlock_all(dev);
+	mutex_unlock(&dev->mode_config.mutex);
 	return ret;
 }
 
@@ -1866,26 +1937,22 @@ out:
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_getcrtc(struct drm_device *dev,
 		     void *data, struct drm_file *file_priv)
 {
 	struct drm_mode_crtc *crtc_resp = data;
 	struct drm_crtc *crtc;
-	int ret = 0;
 
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
 
-	drm_modeset_lock_all(dev);
-
 	crtc = drm_crtc_find(dev, crtc_resp->crtc_id);
-	if (!crtc) {
-		ret = -ENOENT;
-		goto out;
-	}
+	if (!crtc)
+		return -ENOENT;
 
+	drm_modeset_lock_crtc(crtc, crtc->primary);
 	crtc_resp->x = crtc->x;
 	crtc_resp->y = crtc->y;
 	crtc_resp->gamma_size = crtc->gamma_size;
@@ -1902,10 +1969,9 @@ int drm_mode_getcrtc(struct drm_device *dev,
 	} else {
 		crtc_resp->mode_valid = 0;
 	}
+	drm_modeset_unlock_crtc(crtc);
 
-out:
-	drm_modeset_unlock_all(dev);
-	return ret;
+	return 0;
 }
 
 static bool drm_mode_expose_to_userspace(const struct drm_display_mode *mode,
@@ -1921,6 +1987,15 @@ static bool drm_mode_expose_to_userspace(const struct drm_display_mode *mode,
 	return true;
 }
 
+static struct drm_encoder *drm_connector_get_encoder(struct drm_connector *connector)
+{
+	/* For atomic drivers only state objects are synchronously updated and
+	 * protected by modeset locks, so check those first. */
+	if (connector->state)
+		return connector->state->best_encoder;
+	return connector->encoder;
+}
+
 /**
  * drm_mode_getconnector - get connector configuration
  * @dev: drm device for the ioctl
@@ -1932,13 +2007,14 @@ static bool drm_mode_expose_to_userspace(const struct drm_display_mode *mode,
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_getconnector(struct drm_device *dev, void *data,
 			  struct drm_file *file_priv)
 {
 	struct drm_mode_get_connector *out_resp = data;
 	struct drm_connector *connector;
+	struct drm_encoder *encoder;
 	struct drm_display_mode *mode;
 	int mode_count = 0;
 	int props_count = 0;
@@ -1994,8 +2070,10 @@ int drm_mode_getconnector(struct drm_device *dev, void *data,
 	out_resp->subpixel = connector->display_info.subpixel_order;
 	out_resp->connection = connector->status;
 	drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
-	if (connector->encoder)
-		out_resp->encoder_id = connector->encoder->base.id;
+
+	encoder = drm_connector_get_encoder(connector);
+	if (encoder)
+		out_resp->encoder_id = encoder->base.id;
 	else
 		out_resp->encoder_id = 0;
 	drm_modeset_unlock(&dev->mode_config.connection_mutex);
@@ -2065,6 +2143,33 @@ out:
 	return ret;
 }
 
+static struct drm_crtc *drm_encoder_get_crtc(struct drm_encoder *encoder)
+{
+	struct drm_connector *connector;
+	struct drm_device *dev = encoder->dev;
+	bool uses_atomic = false;
+
+	/* For atomic drivers only state objects are synchronously updated and
+	 * protected by modeset locks, so check those first. */
+	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+		if (!connector->state)
+			continue;
+
+		uses_atomic = true;
+
+		if (connector->state->best_encoder != encoder)
+			continue;
+
+		return connector->state->crtc;
+	}
+
+	/* Don't return stale data (e.g. pending async disable). */
+	if (uses_atomic)
+		return NULL;
+
+	return encoder->crtc;
+}
+
 /**
  * drm_mode_getencoder - get encoder configuration
  * @dev: drm device for the ioctl
@@ -2076,37 +2181,38 @@ out:
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_getencoder(struct drm_device *dev, void *data,
 			struct drm_file *file_priv)
 {
 	struct drm_mode_get_encoder *enc_resp = data;
 	struct drm_encoder *encoder;
-	int ret = 0;
+	struct drm_crtc *crtc;
 
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
 
-	drm_modeset_lock_all(dev);
 	encoder = drm_encoder_find(dev, enc_resp->encoder_id);
-	if (!encoder) {
-		ret = -ENOENT;
-		goto out;
-	}
+	if (!encoder)
+		return -ENOENT;
 
-	if (encoder->crtc)
+	drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
+	crtc = drm_encoder_get_crtc(encoder);
+	if (crtc)
+		enc_resp->crtc_id = crtc->base.id;
+	else if (encoder->crtc)
 		enc_resp->crtc_id = encoder->crtc->base.id;
 	else
 		enc_resp->crtc_id = 0;
+	drm_modeset_unlock(&dev->mode_config.connection_mutex);
+
 	enc_resp->encoder_type = encoder->encoder_type;
 	enc_resp->encoder_id = encoder->base.id;
 	enc_resp->possible_crtcs = encoder->possible_crtcs;
 	enc_resp->possible_clones = encoder->possible_clones;
 
-out:
-	drm_modeset_unlock_all(dev);
-	return ret;
+	return 0;
 }
 
 /**
@@ -2120,7 +2226,7 @@ out:
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_getplane_res(struct drm_device *dev, void *data,
 			  struct drm_file *file_priv)
@@ -2129,13 +2235,12 @@ int drm_mode_getplane_res(struct drm_device *dev, void *data,
 	struct drm_mode_config *config;
 	struct drm_plane *plane;
 	uint32_t __user *plane_ptr;
-	int copied = 0, ret = 0;
+	int copied = 0;
 	unsigned num_planes;
 
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
 
-	drm_modeset_lock_all(dev);
 	config = &dev->mode_config;
 
 	if (file_priv->universal_planes)
@@ -2151,6 +2256,7 @@ int drm_mode_getplane_res(struct drm_device *dev, void *data,
 	    (plane_resp->count_planes >= num_planes)) {
 		plane_ptr = (uint32_t __user *)(unsigned long)plane_resp->plane_id_ptr;
 
+		/* Plane lists are invariant, no locking needed. */
 		list_for_each_entry(plane, &config->plane_list, head) {
 			/*
 			 * Unless userspace set the 'universal planes'
@@ -2160,18 +2266,14 @@ int drm_mode_getplane_res(struct drm_device *dev, void *data,
 			    !file_priv->universal_planes)
 				continue;
 
-			if (put_user(plane->base.id, plane_ptr + copied)) {
-				ret = -EFAULT;
-				goto out;
-			}
+			if (put_user(plane->base.id, plane_ptr + copied))
+				return -EFAULT;
 			copied++;
 		}
 	}
 	plane_resp->count_planes = num_planes;
 
-out:
-	drm_modeset_unlock_all(dev);
-	return ret;
+	return 0;
 }
 
 /**
@@ -2185,7 +2287,7 @@ out:
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_getplane(struct drm_device *dev, void *data,
 		      struct drm_file *file_priv)
@@ -2193,18 +2295,15 @@ int drm_mode_getplane(struct drm_device *dev, void *data,
 	struct drm_mode_get_plane *plane_resp = data;
 	struct drm_plane *plane;
 	uint32_t __user *format_ptr;
-	int ret = 0;
 
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
 
-	drm_modeset_lock_all(dev);
 	plane = drm_plane_find(dev, plane_resp->plane_id);
-	if (!plane) {
-		ret = -ENOENT;
-		goto out;
-	}
+	if (!plane)
+		return -ENOENT;
 
+	drm_modeset_lock(&plane->mutex, NULL);
 	if (plane->crtc)
 		plane_resp->crtc_id = plane->crtc->base.id;
 	else
@@ -2214,6 +2313,7 @@ int drm_mode_getplane(struct drm_device *dev, void *data,
 		plane_resp->fb_id = plane->fb->base.id;
 	else
 		plane_resp->fb_id = 0;
+	drm_modeset_unlock(&plane->mutex);
 
 	plane_resp->plane_id = plane->base.id;
 	plane_resp->possible_crtcs = plane->possible_crtcs;
@@ -2229,15 +2329,12 @@ int drm_mode_getplane(struct drm_device *dev, void *data,
 		if (copy_to_user(format_ptr,
 				 plane->format_types,
 				 sizeof(uint32_t) * plane->format_count)) {
-			ret = -EFAULT;
-			goto out;
+			return -EFAULT;
 		}
 	}
 	plane_resp->count_format_types = plane->format_count;
 
-out:
-	drm_modeset_unlock_all(dev);
-	return ret;
+	return 0;
 }
 
 /*
@@ -2249,33 +2346,29 @@ out:
  *
  * src_{x,y,w,h} are provided in 16.16 fixed point format
  */
-static int setplane_internal(struct drm_plane *plane,
-			     struct drm_crtc *crtc,
-			     struct drm_framebuffer *fb,
-			     int32_t crtc_x, int32_t crtc_y,
-			     uint32_t crtc_w, uint32_t crtc_h,
-			     /* src_{x,y,w,h} values are 16.16 fixed point */
-			     uint32_t src_x, uint32_t src_y,
-			     uint32_t src_w, uint32_t src_h)
+static int __setplane_internal(struct drm_plane *plane,
+			       struct drm_crtc *crtc,
+			       struct drm_framebuffer *fb,
+			       int32_t crtc_x, int32_t crtc_y,
+			       uint32_t crtc_w, uint32_t crtc_h,
+			       /* src_{x,y,w,h} values are 16.16 fixed point */
+			       uint32_t src_x, uint32_t src_y,
+			       uint32_t src_w, uint32_t src_h)
 {
-	struct drm_device *dev = plane->dev;
-	struct drm_framebuffer *old_fb = NULL;
 	int ret = 0;
 	unsigned int fb_width, fb_height;
-	int i;
+	unsigned int i;
 
 	/* No fb means shut it down */
 	if (!fb) {
-		drm_modeset_lock_all(dev);
-		old_fb = plane->fb;
+		plane->old_fb = plane->fb;
 		ret = plane->funcs->disable_plane(plane);
 		if (!ret) {
 			plane->crtc = NULL;
 			plane->fb = NULL;
 		} else {
-			old_fb = NULL;
+			plane->old_fb = NULL;
 		}
-		drm_modeset_unlock_all(dev);
 		goto out;
 	}
 
@@ -2315,8 +2408,7 @@ static int setplane_internal(struct drm_plane *plane,
 		goto out;
 	}
 
-	drm_modeset_lock_all(dev);
-	old_fb = plane->fb;
+	plane->old_fb = plane->fb;
 	ret = plane->funcs->update_plane(plane, crtc, fb,
 					 crtc_x, crtc_y, crtc_w, crtc_h,
 					 src_x, src_y, src_w, src_h);
@@ -2325,18 +2417,37 @@ static int setplane_internal(struct drm_plane *plane,
 		plane->fb = fb;
 		fb = NULL;
 	} else {
-		old_fb = NULL;
+		plane->old_fb = NULL;
 	}
-	drm_modeset_unlock_all(dev);
 
 out:
 	if (fb)
 		drm_framebuffer_unreference(fb);
-	if (old_fb)
-		drm_framebuffer_unreference(old_fb);
+	if (plane->old_fb)
+		drm_framebuffer_unreference(plane->old_fb);
+	plane->old_fb = NULL;
 
 	return ret;
+}
+
+static int setplane_internal(struct drm_plane *plane,
+			     struct drm_crtc *crtc,
+			     struct drm_framebuffer *fb,
+			     int32_t crtc_x, int32_t crtc_y,
+			     uint32_t crtc_w, uint32_t crtc_h,
+			     /* src_{x,y,w,h} values are 16.16 fixed point */
+			     uint32_t src_x, uint32_t src_y,
+			     uint32_t src_w, uint32_t src_h)
+{
+	int ret;
 
+	drm_modeset_lock_all(plane->dev);
+	ret = __setplane_internal(plane, crtc, fb,
+				  crtc_x, crtc_y, crtc_w, crtc_h,
+				  src_x, src_y, src_w, src_h);
+	drm_modeset_unlock_all(plane->dev);
+
+	return ret;
 }
 
 /**
@@ -2350,13 +2461,12 @@ out:
  * valid crtc).
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_setplane(struct drm_device *dev, void *data,
 		      struct drm_file *file_priv)
 {
 	struct drm_mode_set_plane *plane_req = data;
-	struct drm_mode_object *obj;
 	struct drm_plane *plane;
 	struct drm_crtc *crtc = NULL;
 	struct drm_framebuffer *fb = NULL;
@@ -2379,14 +2489,12 @@ int drm_mode_setplane(struct drm_device *dev, void *data,
 	 * First, find the plane, crtc, and fb objects.  If not available,
 	 * we don't bother to call the driver.
 	 */
-	obj = drm_mode_object_find(dev, plane_req->plane_id,
-				   DRM_MODE_OBJECT_PLANE);
-	if (!obj) {
+	plane = drm_plane_find(dev, plane_req->plane_id);
+	if (!plane) {
 		DRM_DEBUG_KMS("Unknown plane ID %d\n",
 			      plane_req->plane_id);
 		return -ENOENT;
 	}
-	plane = obj_to_plane(obj);
 
 	if (plane_req->fb_id) {
 		fb = drm_framebuffer_lookup(dev, plane_req->fb_id);
@@ -2396,14 +2504,12 @@ int drm_mode_setplane(struct drm_device *dev, void *data,
 			return -ENOENT;
 		}
 
-		obj = drm_mode_object_find(dev, plane_req->crtc_id,
-					   DRM_MODE_OBJECT_CRTC);
-		if (!obj) {
+		crtc = drm_crtc_find(dev, plane_req->crtc_id);
+		if (!crtc) {
 			DRM_DEBUG_KMS("Unknown crtc ID %d\n",
 				      plane_req->crtc_id);
 			return -ENOENT;
 		}
-		crtc = obj_to_crtc(obj);
 	}
 
 	/*
@@ -2425,7 +2531,7 @@ int drm_mode_setplane(struct drm_device *dev, void *data,
  * interface. The only thing it adds is correct refcounting dance.
  * 
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_set_config_internal(struct drm_mode_set *set)
 {
@@ -2440,7 +2546,7 @@ int drm_mode_set_config_internal(struct drm_mode_set *set)
 	 * crtcs. Atomic modeset will have saner semantics ...
 	 */
 	list_for_each_entry(tmp, &crtc->dev->mode_config.crtc_list, head)
-		tmp->old_fb = tmp->primary->fb;
+		tmp->primary->old_fb = tmp->primary->fb;
 
 	fb = set->fb;
 
@@ -2453,8 +2559,9 @@ int drm_mode_set_config_internal(struct drm_mode_set *set)
 	list_for_each_entry(tmp, &crtc->dev->mode_config.crtc_list, head) {
 		if (tmp->primary->fb)
 			drm_framebuffer_reference(tmp->primary->fb);
-		if (tmp->old_fb)
-			drm_framebuffer_unreference(tmp->old_fb);
+		if (tmp->primary->old_fb)
+			drm_framebuffer_unreference(tmp->primary->old_fb);
+		tmp->primary->old_fb = NULL;
 	}
 
 	return ret;
@@ -2517,7 +2624,7 @@ EXPORT_SYMBOL(drm_crtc_check_viewport);
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_setcrtc(struct drm_device *dev, void *data,
 		     struct drm_file *file_priv)
@@ -2680,7 +2787,7 @@ out:
  * userspace wants to make use of these capabilities.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 static int drm_mode_cursor_universal(struct drm_crtc *crtc,
 				     struct drm_mode_cursor2 *req,
@@ -2701,6 +2808,7 @@ static int drm_mode_cursor_universal(struct drm_crtc *crtc,
 	int ret = 0;
 
 	BUG_ON(!crtc->cursor);
+	WARN_ON(crtc->cursor->crtc != crtc && crtc->cursor->crtc != NULL);
 
 	/*
 	 * Obtain fb we'll be using (either new or existing) and take an extra
@@ -2720,11 +2828,9 @@ static int drm_mode_cursor_universal(struct drm_crtc *crtc,
 			fb = NULL;
 		}
 	} else {
-		mutex_lock(&dev->mode_config.mutex);
 		fb = crtc->cursor->fb;
 		if (fb)
 			drm_framebuffer_reference(fb);
-		mutex_unlock(&dev->mode_config.mutex);
 	}
 
 	if (req->flags & DRM_MODE_CURSOR_MOVE) {
@@ -2746,7 +2852,7 @@ static int drm_mode_cursor_universal(struct drm_crtc *crtc,
 	 * setplane_internal will take care of deref'ing either the old or new
 	 * framebuffer depending on success.
 	 */
-	ret = setplane_internal(crtc->cursor, crtc, fb,
+	ret = __setplane_internal(crtc->cursor, crtc, fb,
 				crtc_x, crtc_y, crtc_w, crtc_h,
 				0, 0, src_w, src_h);
 
@@ -2782,10 +2888,12 @@ static int drm_mode_cursor_common(struct drm_device *dev,
 	 * If this crtc has a universal cursor plane, call that plane's update
 	 * handler rather than using legacy cursor handlers.
 	 */
-	if (crtc->cursor)
-		return drm_mode_cursor_universal(crtc, req, file_priv);
+	drm_modeset_lock_crtc(crtc, crtc->cursor);
+	if (crtc->cursor) {
+		ret = drm_mode_cursor_universal(crtc, req, file_priv);
+		goto out;
+	}
 
-	drm_modeset_lock(&crtc->mutex, NULL);
 	if (req->flags & DRM_MODE_CURSOR_BO) {
 		if (!crtc->funcs->cursor_set && !crtc->funcs->cursor_set2) {
 			ret = -ENXIO;
@@ -2809,7 +2917,7 @@ static int drm_mode_cursor_common(struct drm_device *dev,
 		}
 	}
 out:
-	drm_modeset_unlock(&crtc->mutex);
+	drm_modeset_unlock_crtc(crtc);
 
 	return ret;
 
@@ -2827,7 +2935,7 @@ out:
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_cursor_ioctl(struct drm_device *dev,
 			  void *data, struct drm_file *file_priv)
@@ -2854,7 +2962,7 @@ int drm_mode_cursor_ioctl(struct drm_device *dev,
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_cursor2_ioctl(struct drm_device *dev,
 			   void *data, struct drm_file *file_priv)
@@ -2913,23 +3021,21 @@ EXPORT_SYMBOL(drm_mode_legacy_fb_format);
  * @file_priv: drm file for the ioctl call
  *
  * Add a new FB to the specified CRTC, given a user request. This is the
- * original addfb ioclt which only supported RGB formats.
+ * original addfb ioctl which only supported RGB formats.
  *
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_addfb(struct drm_device *dev,
 		   void *data, struct drm_file *file_priv)
 {
 	struct drm_mode_fb_cmd *or = data;
 	struct drm_mode_fb_cmd2 r = {};
-	struct drm_mode_config *config = &dev->mode_config;
-	struct drm_framebuffer *fb;
-	int ret = 0;
+	int ret;
 
-	/* Use new struct with format internally */
+	/* convert to new format and call new ioctl */
 	r.fb_id = or->fb_id;
 	r.width = or->width;
 	r.height = or->height;
@@ -2937,28 +3043,13 @@ int drm_mode_addfb(struct drm_device *dev,
 	r.pixel_format = drm_mode_legacy_fb_format(or->bpp, or->depth);
 	r.handles[0] = or->handle;
 
-	if (!drm_core_check_feature(dev, DRIVER_MODESET))
-		return -EINVAL;
-
-	if ((config->min_width > r.width) || (r.width > config->max_width))
-		return -EINVAL;
-
-	if ((config->min_height > r.height) || (r.height > config->max_height))
-		return -EINVAL;
-
-	fb = dev->mode_config.funcs->fb_create(dev, file_priv, &r);
-	if (IS_ERR(fb)) {
-		DRM_DEBUG_KMS("could not create framebuffer\n");
-		return PTR_ERR(fb);
-	}
+	ret = drm_mode_addfb2(dev, &r, file_priv);
+	if (ret)
+		return ret;
 
-	mutex_lock(&file_priv->fbs_lock);
-	or->fb_id = fb->base.id;
-	list_add(&fb->filp_head, &file_priv->fbs);
-	DRM_DEBUG_KMS("[FB:%d]\n", fb->base.id);
-	mutex_unlock(&file_priv->fbs_lock);
+	or->fb_id = r.fb_id;
 
-	return ret;
+	return 0;
 }
 
 static int format_check(const struct drm_mode_fb_cmd2 *r)
@@ -3050,7 +3141,7 @@ static int framebuffer_check(const struct drm_mode_fb_cmd2 *r)
 	num_planes = drm_format_num_planes(r->pixel_format);
 
 	if (r->width == 0 || r->width % hsub) {
-		DRM_DEBUG_KMS("bad framebuffer width %u\n", r->height);
+		DRM_DEBUG_KMS("bad framebuffer width %u\n", r->width);
 		return -EINVAL;
 	}
 
@@ -3140,7 +3231,7 @@ static struct drm_framebuffer *add_framebuffer_internal(struct drm_device *dev,
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_addfb2(struct drm_device *dev,
 		    void *data, struct drm_file *file_priv)
@@ -3168,7 +3259,7 @@ int drm_mode_addfb2(struct drm_device *dev,
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_rmfb(struct drm_device *dev,
 		   void *data, struct drm_file *file_priv)
@@ -3222,7 +3313,7 @@ fail_lookup:
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_getfb(struct drm_device *dev,
 		   void *data, struct drm_file *file_priv)
@@ -3283,7 +3374,7 @@ int drm_mode_getfb(struct drm_device *dev,
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_dirtyfb_ioctl(struct drm_device *dev,
 			   void *data, struct drm_file *file_priv)
@@ -3363,14 +3454,23 @@ out_err1:
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 void drm_fb_release(struct drm_file *priv)
 {
 	struct drm_device *dev = priv->minor->dev;
 	struct drm_framebuffer *fb, *tfb;
 
-	mutex_lock(&priv->fbs_lock);
+	/*
+	 * When the file gets released that means no one else can access the fb
+	 * list any more, so no need to grab fpriv->fbs_lock. And we need to
+	 * avoid upsetting lockdep since the universal cursor code adds a
+	 * framebuffer while holding mutex locks.
+	 *
+	 * Note that a real deadlock between fpriv->fbs_lock and the modeset
+	 * locks is impossible here since no one else but this function can get
+	 * at it any more.
+	 */
 	list_for_each_entry_safe(fb, tfb, &priv->fbs, filp_head) {
 
 		mutex_lock(&dev->mode_config.fb_lock);
@@ -3383,7 +3483,6 @@ void drm_fb_release(struct drm_file *priv)
 		/* This will also drop the fpriv->fbs reference. */
 		drm_framebuffer_remove(fb);
 	}
-	mutex_unlock(&priv->fbs_lock);
 }
 
 /**
@@ -3397,6 +3496,10 @@ void drm_fb_release(struct drm_file *priv)
  * object with drm_object_attach_property. The returned property object must be
  * freed with drm_property_destroy.
  *
+ * Note that the DRM core keeps a per-device list of properties and that, if
+ * drm_mode_config_cleanup() is called, it will destroy all properties created
+ * by the driver.
+ *
  * Returns:
  * A pointer to the newly created property on success, NULL on failure.
  */
@@ -3424,7 +3527,7 @@ struct drm_property *drm_property_create(struct drm_device *dev, int flags,
 
 	property->flags = flags;
 	property->num_values = num_values;
-	INIT_LIST_HEAD(&property->enum_blob_list);
+	INIT_LIST_HEAD(&property->enum_list);
 
 	if (name) {
 		strncpy(property->name, name, DRM_PROP_NAME_LEN);
@@ -3495,9 +3598,10 @@ EXPORT_SYMBOL(drm_property_create_enum);
  * @flags: flags specifying the property type
  * @name: name of the property
  * @props: enumeration lists with property bitflags
- * @num_values: number of pre-defined values
+ * @num_props: size of the @props array
+ * @supported_bits: bitmask of all supported enumeration values
  *
- * This creates a new generic drm property which can then be attached to a drm
+ * This creates a new bitmask drm property which can then be attached to a drm
  * object with drm_object_attach_property. The returned property object must be
  * freed with drm_property_destroy.
  *
@@ -3572,7 +3676,7 @@ static struct drm_property *property_create_range(struct drm_device *dev,
  * object with drm_object_attach_property. The returned property object must be
  * freed with drm_property_destroy.
  *
- * Userspace is allowed to set any interger value in the (min, max) range
+ * Userspace is allowed to set any integer value in the (min, max) range
  * inclusive.
  *
  * Returns:
@@ -3645,8 +3749,8 @@ int drm_property_add_enum(struct drm_property *property, int index,
 			(value > 63))
 		return -EINVAL;
 
-	if (!list_empty(&property->enum_blob_list)) {
-		list_for_each_entry(prop_enum, &property->enum_blob_list, head) {
+	if (!list_empty(&property->enum_list)) {
+		list_for_each_entry(prop_enum, &property->enum_list, head) {
 			if (prop_enum->value == value) {
 				strncpy(prop_enum->name, name, DRM_PROP_NAME_LEN);
 				prop_enum->name[DRM_PROP_NAME_LEN-1] = '\0';
@@ -3664,7 +3768,7 @@ int drm_property_add_enum(struct drm_property *property, int index,
 	prop_enum->value = value;
 
 	property->values[index] = value;
-	list_add_tail(&prop_enum->head, &property->enum_blob_list);
+	list_add_tail(&prop_enum->head, &property->enum_list);
 	return 0;
 }
 EXPORT_SYMBOL(drm_property_add_enum);
@@ -3681,7 +3785,7 @@ void drm_property_destroy(struct drm_device *dev, struct drm_property *property)
 {
 	struct drm_property_enum *prop_enum, *pt;
 
-	list_for_each_entry_safe(prop_enum, pt, &property->enum_blob_list, head) {
+	list_for_each_entry_safe(prop_enum, pt, &property->enum_list, head) {
 		list_del(&prop_enum->head);
 		kfree(prop_enum);
 	}
@@ -3784,17 +3888,20 @@ int drm_object_property_get_value(struct drm_mode_object *obj,
 EXPORT_SYMBOL(drm_object_property_get_value);
 
 /**
- * drm_mode_getproperty_ioctl - get the current value of a connector's property
+ * drm_mode_getproperty_ioctl - get the property metadata
  * @dev: DRM device
  * @data: ioctl data
  * @file_priv: DRM file info
  *
- * This function retrieves the current value for an connectors's property.
+ * This function retrieves the metadata for a given property, like the different
+ * possible values for an enum property or the limits for a range property.
+ *
+ * Blob properties are special
  *
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_getproperty_ioctl(struct drm_device *dev,
 			       void *data, struct drm_file *file_priv)
@@ -3802,16 +3909,12 @@ int drm_mode_getproperty_ioctl(struct drm_device *dev,
 	struct drm_mode_get_property *out_resp = data;
 	struct drm_property *property;
 	int enum_count = 0;
-	int blob_count = 0;
 	int value_count = 0;
 	int ret = 0, i;
 	int copied;
 	struct drm_property_enum *prop_enum;
 	struct drm_mode_property_enum __user *enum_ptr;
-	struct drm_property_blob *prop_blob;
-	uint32_t __user *blob_id_ptr;
 	uint64_t __user *values_ptr;
-	uint32_t __user *blob_length_ptr;
 
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
@@ -3825,11 +3928,8 @@ int drm_mode_getproperty_ioctl(struct drm_device *dev,
 
 	if (drm_property_type_is(property, DRM_MODE_PROP_ENUM) ||
 			drm_property_type_is(property, DRM_MODE_PROP_BITMASK)) {
-		list_for_each_entry(prop_enum, &property->enum_blob_list, head)
+		list_for_each_entry(prop_enum, &property->enum_list, head)
 			enum_count++;
-	} else if (drm_property_type_is(property, DRM_MODE_PROP_BLOB)) {
-		list_for_each_entry(prop_blob, &property->enum_blob_list, head)
-			blob_count++;
 	}
 
 	value_count = property->num_values;
@@ -3854,7 +3954,7 @@ int drm_mode_getproperty_ioctl(struct drm_device *dev,
 		if ((out_resp->count_enum_blobs >= enum_count) && enum_count) {
 			copied = 0;
 			enum_ptr = (struct drm_mode_property_enum __user *)(unsigned long)out_resp->enum_blob_ptr;
-			list_for_each_entry(prop_enum, &property->enum_blob_list, head) {
+			list_for_each_entry(prop_enum, &property->enum_list, head) {
 
 				if (copy_to_user(&enum_ptr[copied].value, &prop_enum->value, sizeof(uint64_t))) {
 					ret = -EFAULT;
@@ -3872,35 +3972,24 @@ int drm_mode_getproperty_ioctl(struct drm_device *dev,
 		out_resp->count_enum_blobs = enum_count;
 	}
 
-	if (drm_property_type_is(property, DRM_MODE_PROP_BLOB)) {
-		if ((out_resp->count_enum_blobs >= blob_count) && blob_count) {
-			copied = 0;
-			blob_id_ptr = (uint32_t __user *)(unsigned long)out_resp->enum_blob_ptr;
-			blob_length_ptr = (uint32_t __user *)(unsigned long)out_resp->values_ptr;
-
-			list_for_each_entry(prop_blob, &property->enum_blob_list, head) {
-				if (put_user(prop_blob->base.id, blob_id_ptr + copied)) {
-					ret = -EFAULT;
-					goto done;
-				}
-
-				if (put_user(prop_blob->length, blob_length_ptr + copied)) {
-					ret = -EFAULT;
-					goto done;
-				}
-
-				copied++;
-			}
-		}
-		out_resp->count_enum_blobs = blob_count;
-	}
+	/*
+	 * NOTE: The idea seems to have been to use this to read all the blob
+	 * property values. But nothing ever added them to the corresponding
+	 * list, userspace always used the special-purpose get_blob ioctl to
+	 * read the value for a blob property. It also doesn't make a lot of
+	 * sense to return values here when everything else is just metadata for
+	 * the property itself.
+	 */
+	if (drm_property_type_is(property, DRM_MODE_PROP_BLOB))
+		out_resp->count_enum_blobs = 0;
 done:
 	drm_modeset_unlock_all(dev);
 	return ret;
 }
 
-static struct drm_property_blob *drm_property_create_blob(struct drm_device *dev, int length,
-							  void *data)
+static struct drm_property_blob *
+drm_property_create_blob(struct drm_device *dev, size_t length,
+			 const void *data)
 {
 	struct drm_property_blob *blob;
 	int ret;
@@ -3946,7 +4035,7 @@ static void drm_property_destroy_blob(struct drm_device *dev,
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_getblob_ioctl(struct drm_device *dev,
 			   void *data, struct drm_file *file_priv)
@@ -3980,12 +4069,25 @@ done:
 	return ret;
 }
 
+/**
+ * drm_mode_connector_set_path_property - set tile property on connector
+ * @connector: connector to set property on.
+ * @path: path to use for property.
+ *
+ * This creates a property to expose to userspace to specify a
+ * connector path. This is mainly used for DisplayPort MST where
+ * connectors have a topology and we want to allow userspace to give
+ * them more meaningful names.
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
 int drm_mode_connector_set_path_property(struct drm_connector *connector,
-					 char *path)
+					 const char *path)
 {
 	struct drm_device *dev = connector->dev;
-	int ret, size;
-	size = strlen(path) + 1;
+	size_t size = strlen(path) + 1;
+	int ret;
 
 	connector->path_blob_ptr = drm_property_create_blob(connector->dev,
 							    size, path);
@@ -4000,6 +4102,52 @@ int drm_mode_connector_set_path_property(struct drm_connector *connector,
 EXPORT_SYMBOL(drm_mode_connector_set_path_property);
 
 /**
+ * drm_mode_connector_set_tile_property - set tile property on connector
+ * @connector: connector to set property on.
+ *
+ * This looks up the tile information for a connector, and creates a
+ * property for userspace to parse if it exists. The property is of
+ * the form of 8 integers using ':' as a separator.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
+int drm_mode_connector_set_tile_property(struct drm_connector *connector)
+{
+	struct drm_device *dev = connector->dev;
+	int ret, size;
+	char tile[256];
+
+	if (connector->tile_blob_ptr)
+		drm_property_destroy_blob(dev, connector->tile_blob_ptr);
+
+	if (!connector->has_tile) {
+		connector->tile_blob_ptr = NULL;
+		ret = drm_object_property_set_value(&connector->base,
+						    dev->mode_config.tile_property, 0);
+		return ret;
+	}
+
+	snprintf(tile, 256, "%d:%d:%d:%d:%d:%d:%d:%d",
+		 connector->tile_group->id, connector->tile_is_single_monitor,
+		 connector->num_h_tile, connector->num_v_tile,
+		 connector->tile_h_loc, connector->tile_v_loc,
+		 connector->tile_h_size, connector->tile_v_size);
+	size = strlen(tile) + 1;
+
+	connector->tile_blob_ptr = drm_property_create_blob(connector->dev,
+							    size, tile);
+	if (!connector->tile_blob_ptr)
+		return -EINVAL;
+
+	ret = drm_object_property_set_value(&connector->base,
+					    dev->mode_config.tile_property,
+					    connector->tile_blob_ptr->base.id);
+	return ret;
+}
+EXPORT_SYMBOL(drm_mode_connector_set_tile_property);
+
+/**
  * drm_mode_connector_update_edid_property - update the edid property of a connector
  * @connector: drm connector
  * @edid: new value of the edid property
@@ -4008,13 +4156,14 @@ EXPORT_SYMBOL(drm_mode_connector_set_path_property);
  * connector's edid property.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_connector_update_edid_property(struct drm_connector *connector,
-					    struct edid *edid)
+					    const struct edid *edid)
 {
 	struct drm_device *dev = connector->dev;
-	int ret, size;
+	size_t size;
+	int ret;
 
 	/* ignore requests to set edid when overridden */
 	if (connector->override_edid)
@@ -4104,7 +4253,7 @@ static bool drm_property_change_is_valid(struct drm_property *property,
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_connector_property_set_ioctl(struct drm_device *dev,
 				       void *data, struct drm_file *file_priv)
@@ -4157,12 +4306,25 @@ static int drm_mode_crtc_set_obj_prop(struct drm_mode_object *obj,
 	return ret;
 }
 
-static int drm_mode_plane_set_obj_prop(struct drm_mode_object *obj,
-				      struct drm_property *property,
-				      uint64_t value)
+/**
+ * drm_mode_plane_set_obj_prop - set the value of a property
+ * @plane: drm plane object to set property value for
+ * @property: property to set
+ * @value: value the property should be set to
+ *
+ * This functions sets a given property on a given plane object. This function
+ * calls the driver's ->set_property callback and changes the software state of
+ * the property if the callback succeeds.
+ *
+ * Returns:
+ * Zero on success, error code on failure.
+ */
+int drm_mode_plane_set_obj_prop(struct drm_plane *plane,
+				struct drm_property *property,
+				uint64_t value)
 {
 	int ret = -EINVAL;
-	struct drm_plane *plane = obj_to_plane(obj);
+	struct drm_mode_object *obj = &plane->base;
 
 	if (plane->funcs->set_property)
 		ret = plane->funcs->set_property(plane, property, value);
@@ -4171,9 +4333,10 @@ static int drm_mode_plane_set_obj_prop(struct drm_mode_object *obj,
 
 	return ret;
 }
+EXPORT_SYMBOL(drm_mode_plane_set_obj_prop);
 
 /**
- * drm_mode_getproperty_ioctl - get the current value of a object's property
+ * drm_mode_obj_get_properties_ioctl - get the current value of a object's property
  * @dev: DRM device
  * @data: ioctl data
  * @file_priv: DRM file info
@@ -4185,7 +4348,7 @@ static int drm_mode_plane_set_obj_prop(struct drm_mode_object *obj,
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_obj_get_properties_ioctl(struct drm_device *dev, void *data,
 				      struct drm_file *file_priv)
@@ -4257,7 +4420,7 @@ out:
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_obj_set_property_ioctl(struct drm_device *dev, void *data,
 				    struct drm_file *file_priv)
@@ -4309,7 +4472,8 @@ int drm_mode_obj_set_property_ioctl(struct drm_device *dev, void *data,
 		ret = drm_mode_crtc_set_obj_prop(arg_obj, property, arg->value);
 		break;
 	case DRM_MODE_OBJECT_PLANE:
-		ret = drm_mode_plane_set_obj_prop(arg_obj, property, arg->value);
+		ret = drm_mode_plane_set_obj_prop(obj_to_plane(arg_obj),
+						  property, arg->value);
 		break;
 	}
 
@@ -4328,7 +4492,7 @@ out:
  * possible_clones and possible_crtcs bitmasks.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_connector_attach_encoder(struct drm_connector *connector,
 				      struct drm_encoder *encoder)
@@ -4355,7 +4519,7 @@ EXPORT_SYMBOL(drm_mode_connector_attach_encoder);
  * fixed gamma table size.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_crtc_set_gamma_size(struct drm_crtc *crtc,
 				 int gamma_size)
@@ -4384,7 +4548,7 @@ EXPORT_SYMBOL(drm_mode_crtc_set_gamma_size);
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_gamma_set_ioctl(struct drm_device *dev,
 			     void *data, struct drm_file *file_priv)
@@ -4456,7 +4620,7 @@ out:
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_gamma_get_ioctl(struct drm_device *dev,
 			     void *data, struct drm_file *file_priv)
@@ -4522,14 +4686,14 @@ out:
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_page_flip_ioctl(struct drm_device *dev,
 			     void *data, struct drm_file *file_priv)
 {
 	struct drm_mode_crtc_page_flip *page_flip = data;
 	struct drm_crtc *crtc;
-	struct drm_framebuffer *fb = NULL, *old_fb = NULL;
+	struct drm_framebuffer *fb = NULL;
 	struct drm_pending_vblank_event *e = NULL;
 	unsigned long flags;
 	int ret = -EINVAL;
@@ -4545,7 +4709,7 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
 	if (!crtc)
 		return -ENOENT;
 
-	drm_modeset_lock(&crtc->mutex, NULL);
+	drm_modeset_lock_crtc(crtc, crtc->primary);
 	if (crtc->primary->fb == NULL) {
 		/* The framebuffer is currently unbound, presumably
 		 * due to a hotplug event, that userspace has not
@@ -4601,7 +4765,7 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
 			(void (*) (struct drm_pending_event *)) kfree;
 	}
 
-	old_fb = crtc->primary->fb;
+	crtc->primary->old_fb = crtc->primary->fb;
 	ret = crtc->funcs->page_flip(crtc, fb, e, page_flip->flags);
 	if (ret) {
 		if (page_flip->flags & DRM_MODE_PAGE_FLIP_EVENT) {
@@ -4611,7 +4775,7 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
 			kfree(e);
 		}
 		/* Keep the old fb, don't unref it. */
-		old_fb = NULL;
+		crtc->primary->old_fb = NULL;
 	} else {
 		/*
 		 * Warn if the driver hasn't properly updated the crtc->fb
@@ -4627,9 +4791,10 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
 out:
 	if (fb)
 		drm_framebuffer_unreference(fb);
-	if (old_fb)
-		drm_framebuffer_unreference(old_fb);
-	drm_modeset_unlock(&crtc->mutex);
+	if (crtc->primary->old_fb)
+		drm_framebuffer_unreference(crtc->primary->old_fb);
+	crtc->primary->old_fb = NULL;
+	drm_modeset_unlock_crtc(crtc);
 
 	return ret;
 }
@@ -4645,9 +4810,14 @@ out:
 void drm_mode_config_reset(struct drm_device *dev)
 {
 	struct drm_crtc *crtc;
+	struct drm_plane *plane;
 	struct drm_encoder *encoder;
 	struct drm_connector *connector;
 
+	list_for_each_entry(plane, &dev->mode_config.plane_list, head)
+		if (plane->funcs->reset)
+			plane->funcs->reset(plane);
+
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
 		if (crtc->funcs->reset)
 			crtc->funcs->reset(crtc);
@@ -4682,7 +4852,7 @@ EXPORT_SYMBOL(drm_mode_config_reset);
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_create_dumb_ioctl(struct drm_device *dev,
 			       void *data, struct drm_file *file_priv)
@@ -4709,6 +4879,16 @@ int drm_mode_create_dumb_ioctl(struct drm_device *dev,
 	if (PAGE_ALIGN(size) == 0)
 		return -EINVAL;
 
+	/*
+	 * handle, pitch and size are output parameters. Zero them out to
+	 * prevent drivers from accidentally using uninitialized data. Since
+	 * not all existing userspace is clearing these fields properly we
+	 * cannot reject IOCTL with garbage in them.
+	 */
+	args->handle = 0;
+	args->pitch = 0;
+	args->size = 0;
+
 	return dev->driver->dumb_create(file_priv, dev, args);
 }
 
@@ -4724,7 +4904,7 @@ int drm_mode_create_dumb_ioctl(struct drm_device *dev,
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_mmap_dumb_ioctl(struct drm_device *dev,
 			     void *data, struct drm_file *file_priv)
@@ -4751,7 +4931,7 @@ int drm_mode_mmap_dumb_ioctl(struct drm_device *dev,
  * Called by the user via ioctl.
  *
  * Returns:
- * Zero on success, errno on failure.
+ * Zero on success, negative errno on failure.
  */
 int drm_mode_destroy_dumb_ioctl(struct drm_device *dev,
 				void *data, struct drm_file *file_priv)
@@ -5037,6 +5217,7 @@ void drm_mode_config_init(struct drm_device *dev)
 	INIT_LIST_HEAD(&dev->mode_config.property_blob_list);
 	INIT_LIST_HEAD(&dev->mode_config.plane_list);
 	idr_init(&dev->mode_config.crtc_idr);
+	idr_init(&dev->mode_config.tile_idr);
 
 	drm_modeset_lock_all(dev);
 	drm_mode_create_standard_connector_properties(dev);
@@ -5124,6 +5305,7 @@ void drm_mode_config_cleanup(struct drm_device *dev)
 		crtc->funcs->destroy(crtc);
 	}
 
+	idr_destroy(&dev->mode_config.tile_idr);
 	idr_destroy(&dev->mode_config.crtc_idr);
 	drm_modeset_lock_fini(&dev->mode_config.connection_mutex);
 }
@@ -5146,3 +5328,100 @@ struct drm_property *drm_mode_create_rotation_property(struct drm_device *dev,
 					   supported_rotations);
 }
 EXPORT_SYMBOL(drm_mode_create_rotation_property);
+
+/**
+ * DOC: Tile group
+ *
+ * Tile groups are used to represent tiled monitors with a unique
+ * integer identifier. Tiled monitors using DisplayID v1.3 have
+ * a unique 8-byte handle, we store this in a tile group, so we
+ * have a common identifier for all tiles in a monitor group.
+ */
+static void drm_tile_group_free(struct kref *kref)
+{
+	struct drm_tile_group *tg = container_of(kref, struct drm_tile_group, refcount);
+	struct drm_device *dev = tg->dev;
+	mutex_lock(&dev->mode_config.idr_mutex);
+	idr_remove(&dev->mode_config.tile_idr, tg->id);
+	mutex_unlock(&dev->mode_config.idr_mutex);
+	kfree(tg);
+}
+
+/**
+ * drm_mode_put_tile_group - drop a reference to a tile group.
+ * @dev: DRM device
+ * @tg: tile group to drop reference to.
+ *
+ * drop reference to tile group and free if 0.
+ */
+void drm_mode_put_tile_group(struct drm_device *dev,
+			     struct drm_tile_group *tg)
+{
+	kref_put(&tg->refcount, drm_tile_group_free);
+}
+
+/**
+ * drm_mode_get_tile_group - get a reference to an existing tile group
+ * @dev: DRM device
+ * @topology: 8-bytes unique per monitor.
+ *
+ * Use the unique bytes to get a reference to an existing tile group.
+ *
+ * RETURNS:
+ * tile group or NULL if not found.
+ */
+struct drm_tile_group *drm_mode_get_tile_group(struct drm_device *dev,
+					       char topology[8])
+{
+	struct drm_tile_group *tg;
+	int id;
+	mutex_lock(&dev->mode_config.idr_mutex);
+	idr_for_each_entry(&dev->mode_config.tile_idr, tg, id) {
+		if (!memcmp(tg->group_data, topology, 8)) {
+			if (!kref_get_unless_zero(&tg->refcount))
+				tg = NULL;
+			mutex_unlock(&dev->mode_config.idr_mutex);
+			return tg;
+		}
+	}
+	mutex_unlock(&dev->mode_config.idr_mutex);
+	return NULL;
+}
+
+/**
+ * drm_mode_create_tile_group - create a tile group from a displayid description
+ * @dev: DRM device
+ * @topology: 8-bytes unique per monitor.
+ *
+ * Create a tile group for the unique monitor, and get a unique
+ * identifier for the tile group.
+ *
+ * RETURNS:
+ * new tile group or error.
+ */
+struct drm_tile_group *drm_mode_create_tile_group(struct drm_device *dev,
+						  char topology[8])
+{
+	struct drm_tile_group *tg;
+	int ret;
+
+	tg = kzalloc(sizeof(*tg), GFP_KERNEL);
+	if (!tg)
+		return ERR_PTR(-ENOMEM);
+
+	kref_init(&tg->refcount);
+	memcpy(tg->group_data, topology, 8);
+	tg->dev = dev;
+
+	mutex_lock(&dev->mode_config.idr_mutex);
+	ret = idr_alloc(&dev->mode_config.tile_idr, tg, 1, 0, GFP_KERNEL);
+	if (ret >= 0) {
+		tg->id = ret;
+	} else {
+		kfree(tg);
+		tg = ERR_PTR(ret);
+	}
+
+	mutex_unlock(&dev->mode_config.idr_mutex);
+	return tg;
+}
diff --git a/drivers/gpu/drm/drm_crtc_helper.c b/drivers/gpu/drm/drm_crtc_helper.c
index 6c65a0a28fbd..d552708409de 100644
--- a/drivers/gpu/drm/drm_crtc_helper.c
+++ b/drivers/gpu/drm/drm_crtc_helper.c
@@ -34,12 +34,35 @@
 #include <linux/moduleparam.h>
 
 #include <drm/drmP.h>
+#include <drm/drm_atomic.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_fourcc.h>
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_fb_helper.h>
+#include <drm/drm_plane_helper.h>
+#include <drm/drm_atomic_helper.h>
 #include <drm/drm_edid.h>
 
+/**
+ * DOC: overview
+ *
+ * The CRTC modeset helper library provides a default set_config implementation
+ * in drm_crtc_helper_set_config(). Plus a few other convenience functions using
+ * the same callbacks which drivers can use to e.g. restore the modeset
+ * configuration on resume with drm_helper_resume_force_mode().
+ *
+ * The driver callbacks are mostly compatible with the atomic modeset helpers,
+ * except for the handling of the primary plane: Atomic helpers require that the
+ * primary plane is implemented as a real standalone plane and not directly tied
+ * to the CRTC state. For easier transition this library provides functions to
+ * implement the old semantics required by the CRTC helpers using the new plane
+ * and atomic helper callbacks.
+ *
+ * Drivers are strongly urged to convert to the atomic helpers (by way of first
+ * converting to the plane helpers). New drivers must not use these functions
+ * but need to implement the atomic interface instead, potentially using the
+ * atomic helpers for that.
+ */
 MODULE_AUTHOR("David Airlie, Jesse Barnes");
 MODULE_DESCRIPTION("DRM KMS helper");
 MODULE_LICENSE("GPL and additional rights");
@@ -888,3 +911,112 @@ void drm_helper_resume_force_mode(struct drm_device *dev)
 	drm_modeset_unlock_all(dev);
 }
 EXPORT_SYMBOL(drm_helper_resume_force_mode);
+
+/**
+ * drm_helper_crtc_mode_set - mode_set implementation for atomic plane helpers
+ * @crtc: DRM CRTC
+ * @mode: DRM display mode which userspace requested
+ * @adjusted_mode: DRM display mode adjusted by ->mode_fixup callbacks
+ * @x: x offset of the CRTC scanout area on the underlying framebuffer
+ * @y: y offset of the CRTC scanout area on the underlying framebuffer
+ * @old_fb: previous framebuffer
+ *
+ * This function implements a callback useable as the ->mode_set callback
+ * required by the crtc helpers. Besides the atomic plane helper functions for
+ * the primary plane the driver must also provide the ->mode_set_nofb callback
+ * to set up the crtc.
+ *
+ * This is a transitional helper useful for converting drivers to the atomic
+ * interfaces.
+ */
+int drm_helper_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *mode,
+			     struct drm_display_mode *adjusted_mode, int x, int y,
+			     struct drm_framebuffer *old_fb)
+{
+	struct drm_crtc_state *crtc_state;
+	struct drm_crtc_helper_funcs *crtc_funcs = crtc->helper_private;
+	int ret;
+
+	if (crtc->funcs->atomic_duplicate_state)
+		crtc_state = crtc->funcs->atomic_duplicate_state(crtc);
+	else if (crtc->state)
+		crtc_state = kmemdup(crtc->state, sizeof(*crtc_state),
+				     GFP_KERNEL);
+	else
+		crtc_state = kzalloc(sizeof(*crtc_state), GFP_KERNEL);
+	if (!crtc_state)
+		return -ENOMEM;
+
+	crtc_state->enable = true;
+	crtc_state->planes_changed = true;
+	crtc_state->mode_changed = true;
+	drm_mode_copy(&crtc_state->mode, mode);
+	drm_mode_copy(&crtc_state->adjusted_mode, adjusted_mode);
+
+	if (crtc_funcs->atomic_check) {
+		ret = crtc_funcs->atomic_check(crtc, crtc_state);
+		if (ret) {
+			kfree(crtc_state);
+
+			return ret;
+		}
+	}
+
+	swap(crtc->state, crtc_state);
+
+	crtc_funcs->mode_set_nofb(crtc);
+
+	if (crtc_state) {
+		if (crtc->funcs->atomic_destroy_state)
+			crtc->funcs->atomic_destroy_state(crtc, crtc_state);
+		else
+			kfree(crtc_state);
+	}
+
+	return drm_helper_crtc_mode_set_base(crtc, x, y, old_fb);
+}
+EXPORT_SYMBOL(drm_helper_crtc_mode_set);
+
+/**
+ * drm_helper_crtc_mode_set_base - mode_set_base implementation for atomic plane helpers
+ * @crtc: DRM CRTC
+ * @x: x offset of the CRTC scanout area on the underlying framebuffer
+ * @y: y offset of the CRTC scanout area on the underlying framebuffer
+ * @old_fb: previous framebuffer
+ *
+ * This function implements a callback useable as the ->mode_set_base used
+ * required by the crtc helpers. The driver must provide the atomic plane helper
+ * functions for the primary plane.
+ *
+ * This is a transitional helper useful for converting drivers to the atomic
+ * interfaces.
+ */
+int drm_helper_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
+				  struct drm_framebuffer *old_fb)
+{
+	struct drm_plane_state *plane_state;
+	struct drm_plane *plane = crtc->primary;
+
+	if (plane->funcs->atomic_duplicate_state)
+		plane_state = plane->funcs->atomic_duplicate_state(plane);
+	else if (plane->state)
+		plane_state = drm_atomic_helper_plane_duplicate_state(plane);
+	else
+		plane_state = kzalloc(sizeof(*plane_state), GFP_KERNEL);
+	if (!plane_state)
+		return -ENOMEM;
+
+	plane_state->crtc = crtc;
+	drm_atomic_set_fb_for_plane(plane_state, crtc->primary->fb);
+	plane_state->crtc_x = 0;
+	plane_state->crtc_y = 0;
+	plane_state->crtc_h = crtc->mode.vdisplay;
+	plane_state->crtc_w = crtc->mode.hdisplay;
+	plane_state->src_x = x << 16;
+	plane_state->src_y = y << 16;
+	plane_state->src_h = crtc->mode.vdisplay << 16;
+	plane_state->src_w = crtc->mode.hdisplay << 16;
+
+	return drm_plane_helper_commit(plane, plane_state, old_fb);
+}
+EXPORT_SYMBOL(drm_helper_crtc_mode_set_base);
diff --git a/drivers/gpu/drm/drm_debugfs.c b/drivers/gpu/drm/drm_debugfs.c
index 13bd42923dd4..3bcf8e6a85b3 100644
--- a/drivers/gpu/drm/drm_debugfs.c
+++ b/drivers/gpu/drm/drm_debugfs.c
@@ -36,6 +36,7 @@
 #include <linux/export.h>
 #include <drm/drmP.h>
 #include <drm/drm_edid.h>
+#include "drm_internal.h"
 
 #if defined(CONFIG_DEBUG_FS)
 
@@ -49,9 +50,7 @@ static const struct drm_info_list drm_debugfs_list[] = {
 	{"clients", drm_clients_info, 0},
 	{"bufs", drm_bufs_info, 0},
 	{"gem_names", drm_gem_name_info, DRIVER_GEM},
-#if DRM_DEBUG_CODE
 	{"vma", drm_vma_info, 0},
-#endif
 };
 #define DRM_DEBUGFS_ENTRIES ARRAY_SIZE(drm_debugfs_list)
 
diff --git a/drivers/gpu/drm/drm_dma.c b/drivers/gpu/drm/drm_dma.c
index 8a140a953754..ea481800ef56 100644
--- a/drivers/gpu/drm/drm_dma.c
+++ b/drivers/gpu/drm/drm_dma.c
@@ -35,6 +35,7 @@
 
 #include <linux/export.h>
 #include <drm/drmP.h>
+#include "drm_legacy.h"
 
 /**
  * Initialize the DMA data.
@@ -124,7 +125,7 @@ void drm_legacy_dma_takedown(struct drm_device *dev)
  *
  * Resets the fields of \p buf.
  */
-void drm_free_buffer(struct drm_device *dev, struct drm_buf * buf)
+void drm_legacy_free_buffer(struct drm_device *dev, struct drm_buf * buf)
 {
 	if (!buf)
 		return;
@@ -142,8 +143,8 @@ void drm_free_buffer(struct drm_device *dev, struct drm_buf * buf)
  *
  * Frees each buffer associated with \p file_priv not already on the hardware.
  */
-void drm_core_reclaim_buffers(struct drm_device *dev,
-			      struct drm_file *file_priv)
+void drm_legacy_reclaim_buffers(struct drm_device *dev,
+				struct drm_file *file_priv)
 {
 	struct drm_device_dma *dma = dev->dma;
 	int i;
@@ -154,7 +155,7 @@ void drm_core_reclaim_buffers(struct drm_device *dev,
 		if (dma->buflist[i]->file_priv == file_priv) {
 			switch (dma->buflist[i]->list) {
 			case DRM_LIST_NONE:
-				drm_free_buffer(dev, dma->buflist[i]);
+				drm_legacy_free_buffer(dev, dma->buflist[i]);
 				break;
 			case DRM_LIST_WAIT:
 				dma->buflist[i]->list = DRM_LIST_RECLAIM;
@@ -166,5 +167,3 @@ void drm_core_reclaim_buffers(struct drm_device *dev,
 		}
 	}
 }
-
-EXPORT_SYMBOL(drm_core_reclaim_buffers);
diff --git a/drivers/gpu/drm/drm_dp_helper.c b/drivers/gpu/drm/drm_dp_helper.c
index 08e33b8b13a4..79968e39c8d0 100644
--- a/drivers/gpu/drm/drm_dp_helper.c
+++ b/drivers/gpu/drm/drm_dp_helper.c
@@ -39,198 +39,6 @@
  * blocks, ...
  */
 
-/* Run a single AUX_CH I2C transaction, writing/reading data as necessary */
-static int
-i2c_algo_dp_aux_transaction(struct i2c_adapter *adapter, int mode,
-			    uint8_t write_byte, uint8_t *read_byte)
-{
-	struct i2c_algo_dp_aux_data *algo_data = adapter->algo_data;
-	int ret;
-
-	ret = (*algo_data->aux_ch)(adapter, mode,
-				   write_byte, read_byte);
-	return ret;
-}
-
-/*
- * I2C over AUX CH
- */
-
-/*
- * Send the address. If the I2C link is running, this 'restarts'
- * the connection with the new address, this is used for doing
- * a write followed by a read (as needed for DDC)
- */
-static int
-i2c_algo_dp_aux_address(struct i2c_adapter *adapter, u16 address, bool reading)
-{
-	struct i2c_algo_dp_aux_data *algo_data = adapter->algo_data;
-	int mode = MODE_I2C_START;
-	int ret;
-
-	if (reading)
-		mode |= MODE_I2C_READ;
-	else
-		mode |= MODE_I2C_WRITE;
-	algo_data->address = address;
-	algo_data->running = true;
-	ret = i2c_algo_dp_aux_transaction(adapter, mode, 0, NULL);
-	return ret;
-}
-
-/*
- * Stop the I2C transaction. This closes out the link, sending
- * a bare address packet with the MOT bit turned off
- */
-static void
-i2c_algo_dp_aux_stop(struct i2c_adapter *adapter, bool reading)
-{
-	struct i2c_algo_dp_aux_data *algo_data = adapter->algo_data;
-	int mode = MODE_I2C_STOP;
-
-	if (reading)
-		mode |= MODE_I2C_READ;
-	else
-		mode |= MODE_I2C_WRITE;
-	if (algo_data->running) {
-		(void) i2c_algo_dp_aux_transaction(adapter, mode, 0, NULL);
-		algo_data->running = false;
-	}
-}
-
-/*
- * Write a single byte to the current I2C address, the
- * the I2C link must be running or this returns -EIO
- */
-static int
-i2c_algo_dp_aux_put_byte(struct i2c_adapter *adapter, u8 byte)
-{
-	struct i2c_algo_dp_aux_data *algo_data = adapter->algo_data;
-	int ret;
-
-	if (!algo_data->running)
-		return -EIO;
-
-	ret = i2c_algo_dp_aux_transaction(adapter, MODE_I2C_WRITE, byte, NULL);
-	return ret;
-}
-
-/*
- * Read a single byte from the current I2C address, the
- * I2C link must be running or this returns -EIO
- */
-static int
-i2c_algo_dp_aux_get_byte(struct i2c_adapter *adapter, u8 *byte_ret)
-{
-	struct i2c_algo_dp_aux_data *algo_data = adapter->algo_data;
-	int ret;
-
-	if (!algo_data->running)
-		return -EIO;
-
-	ret = i2c_algo_dp_aux_transaction(adapter, MODE_I2C_READ, 0, byte_ret);
-	return ret;
-}
-
-static int
-i2c_algo_dp_aux_xfer(struct i2c_adapter *adapter,
-		     struct i2c_msg *msgs,
-		     int num)
-{
-	int ret = 0;
-	bool reading = false;
-	int m;
-	int b;
-
-	for (m = 0; m < num; m++) {
-		u16 len = msgs[m].len;
-		u8 *buf = msgs[m].buf;
-		reading = (msgs[m].flags & I2C_M_RD) != 0;
-		ret = i2c_algo_dp_aux_address(adapter, msgs[m].addr, reading);
-		if (ret < 0)
-			break;
-		if (reading) {
-			for (b = 0; b < len; b++) {
-				ret = i2c_algo_dp_aux_get_byte(adapter, &buf[b]);
-				if (ret < 0)
-					break;
-			}
-		} else {
-			for (b = 0; b < len; b++) {
-				ret = i2c_algo_dp_aux_put_byte(adapter, buf[b]);
-				if (ret < 0)
-					break;
-			}
-		}
-		if (ret < 0)
-			break;
-	}
-	if (ret >= 0)
-		ret = num;
-	i2c_algo_dp_aux_stop(adapter, reading);
-	DRM_DEBUG_KMS("dp_aux_xfer return %d\n", ret);
-	return ret;
-}
-
-static u32
-i2c_algo_dp_aux_functionality(struct i2c_adapter *adapter)
-{
-	return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL |
-	       I2C_FUNC_SMBUS_READ_BLOCK_DATA |
-	       I2C_FUNC_SMBUS_BLOCK_PROC_CALL |
-	       I2C_FUNC_10BIT_ADDR;
-}
-
-static const struct i2c_algorithm i2c_dp_aux_algo = {
-	.master_xfer	= i2c_algo_dp_aux_xfer,
-	.functionality	= i2c_algo_dp_aux_functionality,
-};
-
-static void
-i2c_dp_aux_reset_bus(struct i2c_adapter *adapter)
-{
-	(void) i2c_algo_dp_aux_address(adapter, 0, false);
-	(void) i2c_algo_dp_aux_stop(adapter, false);
-}
-
-static int
-i2c_dp_aux_prepare_bus(struct i2c_adapter *adapter)
-{
-	adapter->algo = &i2c_dp_aux_algo;
-	adapter->retries = 3;
-	i2c_dp_aux_reset_bus(adapter);
-	return 0;
-}
-
-/**
- * i2c_dp_aux_add_bus() - register an i2c adapter using the aux ch helper
- * @adapter: i2c adapter to register
- *
- * This registers an i2c adapter that uses dp aux channel as it's underlaying
- * transport. The driver needs to fill out the &i2c_algo_dp_aux_data structure
- * and store it in the algo_data member of the @adapter argument. This will be
- * used by the i2c over dp aux algorithm to drive the hardware.
- *
- * RETURNS:
- * 0 on success, -ERRNO on failure.
- *
- * IMPORTANT:
- * This interface is deprecated, please switch to the new dp aux helpers and
- * drm_dp_aux_register().
- */
-int
-i2c_dp_aux_add_bus(struct i2c_adapter *adapter)
-{
-	int error;
-
-	error = i2c_dp_aux_prepare_bus(adapter);
-	if (error)
-		return error;
-	error = i2c_add_adapter(adapter);
-	return error;
-}
-EXPORT_SYMBOL(i2c_dp_aux_add_bus);
-
 /* Helpers for DP link training */
 static u8 dp_link_status(const u8 link_status[DP_LINK_STATUS_SIZE], int r)
 {
@@ -378,10 +186,11 @@ static int drm_dp_dpcd_access(struct drm_dp_aux *aux, u8 request,
 
 	/*
 	 * The specification doesn't give any recommendation on how often to
-	 * retry native transactions, so retry 7 times like for I2C-over-AUX
-	 * transactions.
+	 * retry native transactions. We used to retry 7 times like for
+	 * aux i2c transactions but real world devices this wasn't
+	 * sufficient, bump to 32 which makes Dell 4k monitors happier.
 	 */
-	for (retry = 0; retry < 7; retry++) {
+	for (retry = 0; retry < 32; retry++) {
 
 		mutex_lock(&aux->hw_mutex);
 		err = aux->transfer(aux, &msg);
@@ -654,10 +463,12 @@ static int drm_dp_i2c_do_msg(struct drm_dp_aux *aux, struct drm_dp_aux_msg *msg)
 
 		case DP_AUX_I2C_REPLY_NACK:
 			DRM_DEBUG_KMS("I2C nack\n");
+			aux->i2c_nack_count++;
 			return -EREMOTEIO;
 
 		case DP_AUX_I2C_REPLY_DEFER:
 			DRM_DEBUG_KMS("I2C defer\n");
+			aux->i2c_defer_count++;
 			usleep_range(400, 500);
 			continue;
 
diff --git a/drivers/gpu/drm/drm_dp_mst_topology.c b/drivers/gpu/drm/drm_dp_mst_topology.c
index ac3c2738db94..9a5b68717ec8 100644
--- a/drivers/gpu/drm/drm_dp_mst_topology.c
+++ b/drivers/gpu/drm/drm_dp_mst_topology.c
@@ -682,7 +682,7 @@ static int build_allocate_payload(struct drm_dp_sideband_msg_tx *msg, int port_n
 static int drm_dp_mst_assign_payload_id(struct drm_dp_mst_topology_mgr *mgr,
 					struct drm_dp_vcpi *vcpi)
 {
-	int ret;
+	int ret, vcpi_ret;
 
 	mutex_lock(&mgr->payload_lock);
 	ret = find_first_zero_bit(&mgr->payload_mask, mgr->max_payloads + 1);
@@ -692,8 +692,16 @@ static int drm_dp_mst_assign_payload_id(struct drm_dp_mst_topology_mgr *mgr,
 		goto out_unlock;
 	}
 
+	vcpi_ret = find_first_zero_bit(&mgr->vcpi_mask, mgr->max_payloads + 1);
+	if (vcpi_ret > mgr->max_payloads) {
+		ret = -EINVAL;
+		DRM_DEBUG_KMS("out of vcpi ids %d\n", ret);
+		goto out_unlock;
+	}
+
 	set_bit(ret, &mgr->payload_mask);
-	vcpi->vcpi = ret;
+	set_bit(vcpi_ret, &mgr->vcpi_mask);
+	vcpi->vcpi = vcpi_ret + 1;
 	mgr->proposed_vcpis[ret - 1] = vcpi;
 out_unlock:
 	mutex_unlock(&mgr->payload_lock);
@@ -701,15 +709,23 @@ out_unlock:
 }
 
 static void drm_dp_mst_put_payload_id(struct drm_dp_mst_topology_mgr *mgr,
-				      int id)
+				      int vcpi)
 {
-	if (id == 0)
+	int i;
+	if (vcpi == 0)
 		return;
 
 	mutex_lock(&mgr->payload_lock);
-	DRM_DEBUG_KMS("putting payload %d\n", id);
-	clear_bit(id, &mgr->payload_mask);
-	mgr->proposed_vcpis[id - 1] = NULL;
+	DRM_DEBUG_KMS("putting payload %d\n", vcpi);
+	clear_bit(vcpi - 1, &mgr->vcpi_mask);
+
+	for (i = 0; i < mgr->max_payloads; i++) {
+		if (mgr->proposed_vcpis[i])
+			if (mgr->proposed_vcpis[i]->vcpi == vcpi) {
+				mgr->proposed_vcpis[i] = NULL;
+				clear_bit(i + 1, &mgr->payload_mask);
+			}
+	}
 	mutex_unlock(&mgr->payload_lock);
 }
 
@@ -823,6 +839,8 @@ static void drm_dp_put_mst_branch_device(struct drm_dp_mst_branch *mstb)
 
 static void drm_dp_port_teardown_pdt(struct drm_dp_mst_port *port, int old_pdt)
 {
+	struct drm_dp_mst_branch *mstb;
+
 	switch (old_pdt) {
 	case DP_PEER_DEVICE_DP_LEGACY_CONV:
 	case DP_PEER_DEVICE_SST_SINK:
@@ -830,8 +848,9 @@ static void drm_dp_port_teardown_pdt(struct drm_dp_mst_port *port, int old_pdt)
 		drm_dp_mst_unregister_i2c_bus(&port->aux);
 		break;
 	case DP_PEER_DEVICE_MST_BRANCHING:
-		drm_dp_put_mst_branch_device(port->mstb);
+		mstb = port->mstb;
 		port->mstb = NULL;
+		drm_dp_put_mst_branch_device(mstb);
 		break;
 	}
 }
@@ -842,6 +861,8 @@ static void drm_dp_destroy_port(struct kref *kref)
 	struct drm_dp_mst_topology_mgr *mgr = port->mgr;
 	if (!port->input) {
 		port->vcpi.num_slots = 0;
+
+		kfree(port->cached_edid);
 		if (port->connector)
 			(*port->mgr->cbs->destroy_connector)(mgr, port->connector);
 		drm_dp_port_teardown_pdt(port, port->pdt);
@@ -995,19 +1016,20 @@ static void drm_dp_check_port_guid(struct drm_dp_mst_branch *mstb,
 
 static void build_mst_prop_path(struct drm_dp_mst_port *port,
 				struct drm_dp_mst_branch *mstb,
-				char *proppath)
+				char *proppath,
+				size_t proppath_size)
 {
 	int i;
 	char temp[8];
-	snprintf(proppath, 255, "mst:%d", mstb->mgr->conn_base_id);
+	snprintf(proppath, proppath_size, "mst:%d", mstb->mgr->conn_base_id);
 	for (i = 0; i < (mstb->lct - 1); i++) {
 		int shift = (i % 2) ? 0 : 4;
 		int port_num = mstb->rad[i / 2] >> shift;
-		snprintf(temp, 8, "-%d", port_num);
-		strncat(proppath, temp, 255);
+		snprintf(temp, sizeof(temp), "-%d", port_num);
+		strlcat(proppath, temp, proppath_size);
 	}
-	snprintf(temp, 8, "-%d", port->port_num);
-	strncat(proppath, temp, 255);
+	snprintf(temp, sizeof(temp), "-%d", port->port_num);
+	strlcat(proppath, temp, proppath_size);
 }
 
 static void drm_dp_add_port(struct drm_dp_mst_branch *mstb,
@@ -1078,8 +1100,12 @@ static void drm_dp_add_port(struct drm_dp_mst_branch *mstb,
 
 	if (created && !port->input) {
 		char proppath[255];
-		build_mst_prop_path(port, mstb, proppath);
+		build_mst_prop_path(port, mstb, proppath, sizeof(proppath));
 		port->connector = (*mstb->mgr->cbs->add_connector)(mstb->mgr, port, proppath);
+
+		if (port->port_num >= 8) {
+			port->cached_edid = drm_get_edid(port->connector, &port->aux.ddc);
+		}
 	}
 
 	/* put reference to this port */
@@ -1563,7 +1589,7 @@ static int drm_dp_destroy_payload_step1(struct drm_dp_mst_topology_mgr *mgr,
 	}
 
 	drm_dp_dpcd_write_payload(mgr, id, payload);
-	payload->payload_state = 0;
+	payload->payload_state = DP_PAYLOAD_DELETE_LOCAL;
 	return 0;
 }
 
@@ -1590,7 +1616,7 @@ static int drm_dp_destroy_payload_step2(struct drm_dp_mst_topology_mgr *mgr,
  */
 int drm_dp_update_payload_part1(struct drm_dp_mst_topology_mgr *mgr)
 {
-	int i;
+	int i, j;
 	int cur_slots = 1;
 	struct drm_dp_payload req_payload;
 	struct drm_dp_mst_port *port;
@@ -1607,26 +1633,46 @@ int drm_dp_update_payload_part1(struct drm_dp_mst_topology_mgr *mgr)
 			port = NULL;
 			req_payload.num_slots = 0;
 		}
+
+		if (mgr->payloads[i].start_slot != req_payload.start_slot) {
+			mgr->payloads[i].start_slot = req_payload.start_slot;
+		}
 		/* work out what is required to happen with this payload */
-		if (mgr->payloads[i].start_slot != req_payload.start_slot ||
-		    mgr->payloads[i].num_slots != req_payload.num_slots) {
+		if (mgr->payloads[i].num_slots != req_payload.num_slots) {
 
 			/* need to push an update for this payload */
 			if (req_payload.num_slots) {
-				drm_dp_create_payload_step1(mgr, i + 1, &req_payload);
+				drm_dp_create_payload_step1(mgr, mgr->proposed_vcpis[i]->vcpi, &req_payload);
 				mgr->payloads[i].num_slots = req_payload.num_slots;
 			} else if (mgr->payloads[i].num_slots) {
 				mgr->payloads[i].num_slots = 0;
-				drm_dp_destroy_payload_step1(mgr, port, i + 1, &mgr->payloads[i]);
+				drm_dp_destroy_payload_step1(mgr, port, port->vcpi.vcpi, &mgr->payloads[i]);
 				req_payload.payload_state = mgr->payloads[i].payload_state;
-			} else
-				req_payload.payload_state = 0;
-
-			mgr->payloads[i].start_slot = req_payload.start_slot;
+				mgr->payloads[i].start_slot = 0;
+			}
 			mgr->payloads[i].payload_state = req_payload.payload_state;
 		}
 		cur_slots += req_payload.num_slots;
 	}
+
+	for (i = 0; i < mgr->max_payloads; i++) {
+		if (mgr->payloads[i].payload_state == DP_PAYLOAD_DELETE_LOCAL) {
+			DRM_DEBUG_KMS("removing payload %d\n", i);
+			for (j = i; j < mgr->max_payloads - 1; j++) {
+				memcpy(&mgr->payloads[j], &mgr->payloads[j + 1], sizeof(struct drm_dp_payload));
+				mgr->proposed_vcpis[j] = mgr->proposed_vcpis[j + 1];
+				if (mgr->proposed_vcpis[j] && mgr->proposed_vcpis[j]->num_slots) {
+					set_bit(j + 1, &mgr->payload_mask);
+				} else {
+					clear_bit(j + 1, &mgr->payload_mask);
+				}
+			}
+			memset(&mgr->payloads[mgr->max_payloads - 1], 0, sizeof(struct drm_dp_payload));
+			mgr->proposed_vcpis[mgr->max_payloads - 1] = NULL;
+			clear_bit(mgr->max_payloads, &mgr->payload_mask);
+
+		}
+	}
 	mutex_unlock(&mgr->payload_lock);
 
 	return 0;
@@ -1657,9 +1703,9 @@ int drm_dp_update_payload_part2(struct drm_dp_mst_topology_mgr *mgr)
 
 		DRM_DEBUG_KMS("payload %d %d\n", i, mgr->payloads[i].payload_state);
 		if (mgr->payloads[i].payload_state == DP_PAYLOAD_LOCAL) {
-			ret = drm_dp_create_payload_step2(mgr, port, i + 1, &mgr->payloads[i]);
+			ret = drm_dp_create_payload_step2(mgr, port, mgr->proposed_vcpis[i]->vcpi, &mgr->payloads[i]);
 		} else if (mgr->payloads[i].payload_state == DP_PAYLOAD_DELETE_LOCAL) {
-			ret = drm_dp_destroy_payload_step2(mgr, i + 1, &mgr->payloads[i]);
+			ret = drm_dp_destroy_payload_step2(mgr, mgr->proposed_vcpis[i]->vcpi, &mgr->payloads[i]);
 		}
 		if (ret) {
 			mutex_unlock(&mgr->payload_lock);
@@ -1762,17 +1808,27 @@ static int drm_dp_send_up_ack_reply(struct drm_dp_mst_topology_mgr *mgr,
 	return 0;
 }
 
-static int drm_dp_get_vc_payload_bw(int dp_link_bw, int dp_link_count)
+static bool drm_dp_get_vc_payload_bw(int dp_link_bw,
+				     int dp_link_count,
+				     int *out)
 {
 	switch (dp_link_bw) {
+	default:
+		DRM_DEBUG_KMS("invalid link bandwidth in DPCD: %x (link count: %d)\n",
+			      dp_link_bw, dp_link_count);
+		return false;
+
 	case DP_LINK_BW_1_62:
-		return 3 * dp_link_count;
+		*out = 3 * dp_link_count;
+		break;
 	case DP_LINK_BW_2_7:
-		return 5 * dp_link_count;
+		*out = 5 * dp_link_count;
+		break;
 	case DP_LINK_BW_5_4:
-		return 10 * dp_link_count;
+		*out = 10 * dp_link_count;
+		break;
 	}
-	return 0;
+	return true;
 }
 
 /**
@@ -1804,7 +1860,13 @@ int drm_dp_mst_topology_mgr_set_mst(struct drm_dp_mst_topology_mgr *mgr, bool ms
 			goto out_unlock;
 		}
 
-		mgr->pbn_div = drm_dp_get_vc_payload_bw(mgr->dpcd[1], mgr->dpcd[2] & DP_MAX_LANE_COUNT_MASK);
+		if (!drm_dp_get_vc_payload_bw(mgr->dpcd[1],
+					      mgr->dpcd[2] & DP_MAX_LANE_COUNT_MASK,
+					      &mgr->pbn_div)) {
+			ret = -EINVAL;
+			goto out_unlock;
+		}
+
 		mgr->total_pbn = 2560;
 		mgr->total_slots = DIV_ROUND_UP(mgr->total_pbn, mgr->pbn_div);
 		mgr->avail_slots = mgr->total_slots;
@@ -1861,6 +1923,7 @@ int drm_dp_mst_topology_mgr_set_mst(struct drm_dp_mst_topology_mgr *mgr, bool ms
 		memset(mgr->payloads, 0, mgr->max_payloads * sizeof(struct drm_dp_payload));
 		mgr->payload_mask = 0;
 		set_bit(0, &mgr->payload_mask);
+		mgr->vcpi_mask = 0;
 	}
 
 out_unlock:
@@ -2071,6 +2134,7 @@ static int drm_dp_mst_handle_up_req(struct drm_dp_mst_topology_mgr *mgr)
  * drm_dp_mst_hpd_irq() - MST hotplug IRQ notify
  * @mgr: manager to notify irq for.
  * @esi: 4 bytes from SINK_COUNT_ESI
+ * @handled: whether the hpd interrupt was consumed or not
  *
  * This should be called from the driver when it detects a short IRQ,
  * along with the value of the DEVICE_SERVICE_IRQ_VECTOR_ESI0. The
@@ -2112,7 +2176,8 @@ EXPORT_SYMBOL(drm_dp_mst_hpd_irq);
  * This returns the current connection state for a port. It validates the
  * port pointer still exists so the caller doesn't require a reference
  */
-enum drm_connector_status drm_dp_mst_detect_port(struct drm_dp_mst_topology_mgr *mgr, struct drm_dp_mst_port *port)
+enum drm_connector_status drm_dp_mst_detect_port(struct drm_connector *connector,
+						 struct drm_dp_mst_topology_mgr *mgr, struct drm_dp_mst_port *port)
 {
 	enum drm_connector_status status = connector_status_disconnected;
 
@@ -2131,6 +2196,10 @@ enum drm_connector_status drm_dp_mst_detect_port(struct drm_dp_mst_topology_mgr
 
 	case DP_PEER_DEVICE_SST_SINK:
 		status = connector_status_connected;
+		/* for logical ports - cache the EDID */
+		if (port->port_num >= 8 && !port->cached_edid) {
+			port->cached_edid = drm_get_edid(connector, &port->aux.ddc);
+		}
 		break;
 	case DP_PEER_DEVICE_DP_LEGACY_CONV:
 		if (port->ldps)
@@ -2162,7 +2231,12 @@ struct edid *drm_dp_mst_get_edid(struct drm_connector *connector, struct drm_dp_
 	if (!port)
 		return NULL;
 
-	edid = drm_get_edid(connector, &port->aux.ddc);
+	if (port->cached_edid)
+		edid = drm_edid_duplicate(port->cached_edid);
+	else
+		edid = drm_get_edid(connector, &port->aux.ddc);
+
+	drm_mode_connector_set_tile_property(connector);
 	drm_dp_put_port(port);
 	return edid;
 }
@@ -2474,7 +2548,7 @@ void drm_dp_mst_dump_topology(struct seq_file *m,
 	mutex_unlock(&mgr->lock);
 
 	mutex_lock(&mgr->payload_lock);
-	seq_printf(m, "vcpi: %lx\n", mgr->payload_mask);
+	seq_printf(m, "vcpi: %lx %lx\n", mgr->payload_mask, mgr->vcpi_mask);
 
 	for (i = 0; i < mgr->max_payloads; i++) {
 		if (mgr->proposed_vcpis[i]) {
diff --git a/drivers/gpu/drm/drm_drv.c b/drivers/gpu/drm/drm_drv.c
index 3242e208c0d0..4f41377b0b80 100644
--- a/drivers/gpu/drm/drm_drv.c
+++ b/drivers/gpu/drm/drm_drv.c
@@ -35,32 +35,20 @@
 #include <drm/drmP.h>
 #include <drm/drm_core.h>
 #include "drm_legacy.h"
+#include "drm_internal.h"
 
 unsigned int drm_debug = 0;	/* 1 to enable debug output */
 EXPORT_SYMBOL(drm_debug);
 
-unsigned int drm_vblank_offdelay = 5000;    /* Default to 5000 msecs. */
-
-unsigned int drm_timestamp_precision = 20;  /* Default to 20 usecs. */
-
-/*
- * Default to use monotonic timestamps for wait-for-vblank and page-flip
- * complete events.
- */
-unsigned int drm_timestamp_monotonic = 1;
-
 MODULE_AUTHOR(CORE_AUTHOR);
 MODULE_DESCRIPTION(CORE_DESC);
 MODULE_LICENSE("GPL and additional rights");
 MODULE_PARM_DESC(debug, "Enable debug output");
-MODULE_PARM_DESC(vblankoffdelay, "Delay until vblank irq auto-disable [msecs]");
+MODULE_PARM_DESC(vblankoffdelay, "Delay until vblank irq auto-disable [msecs] (0: never disable, <0: disable immediately)");
 MODULE_PARM_DESC(timestamp_precision_usec, "Max. error on timestamps [usecs]");
 MODULE_PARM_DESC(timestamp_monotonic, "Use monotonic timestamps");
 
 module_param_named(debug, drm_debug, int, 0600);
-module_param_named(vblankoffdelay, drm_vblank_offdelay, int, 0600);
-module_param_named(timestamp_precision_usec, drm_timestamp_precision, int, 0600);
-module_param_named(timestamp_monotonic, drm_timestamp_monotonic, int, 0600);
 
 static DEFINE_SPINLOCK(drm_minor_lock);
 static struct idr drm_minors_idr;
@@ -68,22 +56,20 @@ static struct idr drm_minors_idr;
 struct class *drm_class;
 static struct dentry *drm_debugfs_root;
 
-int drm_err(const char *func, const char *format, ...)
+void drm_err(const char *format, ...)
 {
 	struct va_format vaf;
 	va_list args;
-	int r;
 
 	va_start(args, format);
 
 	vaf.fmt = format;
 	vaf.va = &args;
 
-	r = printk(KERN_ERR "[" DRM_NAME ":%s] *ERROR* %pV", func, &vaf);
+	printk(KERN_ERR "[" DRM_NAME ":%pf] *ERROR* %pV",
+	       __builtin_return_address(0), &vaf);
 
 	va_end(args);
-
-	return r;
 }
 EXPORT_SYMBOL(drm_err);
 
@@ -102,6 +88,8 @@ void drm_ut_debug_printk(const char *function_name, const char *format, ...)
 }
 EXPORT_SYMBOL(drm_ut_debug_printk);
 
+#define DRM_MAGIC_HASH_ORDER  4  /**< Size of key hash table. Must be power of 2. */
+
 struct drm_master *drm_master_create(struct drm_minor *minor)
 {
 	struct drm_master *master;
@@ -133,7 +121,6 @@ EXPORT_SYMBOL(drm_master_get);
 static void drm_master_destroy(struct kref *kref)
 {
 	struct drm_master *master = container_of(kref, struct drm_master, refcount);
-	struct drm_magic_entry *pt, *next;
 	struct drm_device *dev = master->minor->dev;
 	struct drm_map_list *r_list, *list_temp;
 
@@ -143,7 +130,7 @@ static void drm_master_destroy(struct kref *kref)
 
 	list_for_each_entry_safe(r_list, list_temp, &dev->maplist, head) {
 		if (r_list->master == master) {
-			drm_rmmap_locked(dev, r_list->map);
+			drm_legacy_rmmap_locked(dev, r_list->map);
 			r_list = NULL;
 		}
 	}
@@ -154,12 +141,6 @@ static void drm_master_destroy(struct kref *kref)
 		master->unique_len = 0;
 	}
 
-	list_for_each_entry_safe(pt, next, &master->magicfree, head) {
-		list_del(&pt->head);
-		drm_ht_remove_item(&master->magiclist, &pt->hash_item);
-		kfree(pt);
-	}
-
 	drm_ht_remove(&master->magiclist);
 
 	mutex_unlock(&dev->struct_mutex);
@@ -554,6 +535,8 @@ static void drm_fs_inode_free(struct inode *inode)
  * The initial ref-count of the object is 1. Use drm_dev_ref() and
  * drm_dev_unref() to take and drop further ref-counts.
  *
+ * Note that for purely virtual devices @parent can be NULL.
+ *
  * RETURNS:
  * Pointer to new DRM device, or NULL if out of memory.
  */
@@ -615,7 +598,7 @@ struct drm_device *drm_dev_alloc(struct drm_driver *driver,
 		goto err_ht;
 	}
 
-	if (driver->driver_features & DRIVER_GEM) {
+	if (drm_core_check_feature(dev, DRIVER_GEM)) {
 		ret = drm_gem_init(dev);
 		if (ret) {
 			DRM_ERROR("Cannot initialize graphics execution manager (GEM)\n");
@@ -645,7 +628,7 @@ static void drm_dev_release(struct kref *ref)
 {
 	struct drm_device *dev = container_of(ref, struct drm_device, ref);
 
-	if (dev->driver->driver_features & DRIVER_GEM)
+	if (drm_core_check_feature(dev, DRIVER_GEM))
 		drm_gem_destroy(dev);
 
 	drm_legacy_ctxbitmap_cleanup(dev);
@@ -779,7 +762,7 @@ void drm_dev_unregister(struct drm_device *dev)
 	drm_vblank_cleanup(dev);
 
 	list_for_each_entry_safe(r_list, list_temp, &dev->maplist, head)
-		drm_rmmap(dev, r_list->map);
+		drm_legacy_rmmap(dev, r_list->map);
 
 	drm_minor_unregister(dev, DRM_MINOR_LEGACY);
 	drm_minor_unregister(dev, DRM_MINOR_RENDER);
diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c
index 1dbf3bc4c6a3..53bc7a628909 100644
--- a/drivers/gpu/drm/drm_edid.c
+++ b/drivers/gpu/drm/drm_edid.c
@@ -34,6 +34,7 @@
 #include <linux/module.h>
 #include <drm/drmP.h>
 #include <drm/drm_edid.h>
+#include <drm/drm_displayid.h>
 
 #define version_greater(edid, maj, min) \
 	(((edid)->version > (maj)) || \
@@ -632,27 +633,27 @@ static const struct drm_display_mode edid_cea_modes[] = {
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 6 - 1440x480i@60Hz */
-	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1478,
-		   1602, 1716, 0, 480, 488, 494, 525, 0,
+	/* 6 - 720(1440)x480i@60Hz */
+	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 13500, 720, 739,
+		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 7 - 1440x480i@60Hz */
-	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1478,
-		   1602, 1716, 0, 480, 488, 494, 525, 0,
+	/* 7 - 720(1440)x480i@60Hz */
+	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 13500, 720, 739,
+		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 8 - 1440x240@60Hz */
-	{ DRM_MODE("1440x240", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1478,
-		   1602, 1716, 0, 240, 244, 247, 262, 0,
+	/* 8 - 720(1440)x240@60Hz */
+	{ DRM_MODE("720x240", DRM_MODE_TYPE_DRIVER, 13500, 720, 739,
+		   801, 858, 0, 240, 244, 247, 262, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 9 - 1440x240@60Hz */
-	{ DRM_MODE("1440x240", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1478,
-		   1602, 1716, 0, 240, 244, 247, 262, 0,
+	/* 9 - 720(1440)x240@60Hz */
+	{ DRM_MODE("720x240", DRM_MODE_TYPE_DRIVER, 13500, 720, 739,
+		   801, 858, 0, 240, 244, 247, 262, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
@@ -714,27 +715,27 @@ static const struct drm_display_mode edid_cea_modes[] = {
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 21 - 1440x576i@50Hz */
-	{ DRM_MODE("1440x576i", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1464,
-		   1590, 1728, 0, 576, 580, 586, 625, 0,
+	/* 21 - 720(1440)x576i@50Hz */
+	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 13500, 720, 732,
+		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 22 - 1440x576i@50Hz */
-	{ DRM_MODE("1440x576i", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1464,
-		   1590, 1728, 0, 576, 580, 586, 625, 0,
+	/* 22 - 720(1440)x576i@50Hz */
+	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 13500, 720, 732,
+		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 23 - 1440x288@50Hz */
-	{ DRM_MODE("1440x288", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1464,
-		   1590, 1728, 0, 288, 290, 293, 312, 0,
+	/* 23 - 720(1440)x288@50Hz */
+	{ DRM_MODE("720x288", DRM_MODE_TYPE_DRIVER, 13500, 720, 732,
+		   795, 864, 0, 288, 290, 293, 312, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 24 - 1440x288@50Hz */
-	{ DRM_MODE("1440x288", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1464,
-		   1590, 1728, 0, 288, 290, 293, 312, 0,
+	/* 24 - 720(1440)x288@50Hz */
+	{ DRM_MODE("720x288", DRM_MODE_TYPE_DRIVER, 13500, 720, 732,
+		   795, 864, 0, 288, 290, 293, 312, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
@@ -837,17 +838,17 @@ static const struct drm_display_mode edid_cea_modes[] = {
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 44 - 1440x576i@100Hz */
-	{ DRM_MODE("1440x576", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1464,
-		   1590, 1728, 0, 576, 580, 586, 625, 0,
+	/* 44 - 720(1440)x576i@100Hz */
+	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 27000, 720, 732,
+		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_DBLCLK),
+			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 45 - 1440x576i@100Hz */
-	{ DRM_MODE("1440x576", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1464,
-		   1590, 1728, 0, 576, 580, 586, 625, 0,
+	/* 45 - 720(1440)x576i@100Hz */
+	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 27000, 720, 732,
+		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_DBLCLK),
+			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 46 - 1920x1080i@120Hz */
 	{ DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2008,
@@ -870,15 +871,15 @@ static const struct drm_display_mode edid_cea_modes[] = {
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 50 - 1440x480i@120Hz */
-	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1478,
-		   1602, 1716, 0, 480, 488, 494, 525, 0,
+	/* 50 - 720(1440)x480i@120Hz */
+	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 27000, 720, 739,
+		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 51 - 1440x480i@120Hz */
-	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1478,
-		   1602, 1716, 0, 480, 488, 494, 525, 0,
+	/* 51 - 720(1440)x480i@120Hz */
+	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 27000, 720, 739,
+		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
@@ -892,15 +893,15 @@ static const struct drm_display_mode edid_cea_modes[] = {
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 200, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 54 - 1440x576i@200Hz */
-	{ DRM_MODE("1440x576i", DRM_MODE_TYPE_DRIVER, 108000, 1440, 1464,
-		   1590, 1728, 0, 576, 580, 586, 625, 0,
+	/* 54 - 720(1440)x576i@200Hz */
+	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 54000, 720, 732,
+		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 200, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 55 - 1440x576i@200Hz */
-	{ DRM_MODE("1440x576i", DRM_MODE_TYPE_DRIVER, 108000, 1440, 1464,
-		   1590, 1728, 0, 576, 580, 586, 625, 0,
+	/* 55 - 720(1440)x576i@200Hz */
+	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 54000, 720, 732,
+		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 200, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
@@ -914,15 +915,15 @@ static const struct drm_display_mode edid_cea_modes[] = {
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 240, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 58 - 1440x480i@240 */
-	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 108000, 1440, 1478,
-		   1602, 1716, 0, 480, 488, 494, 525, 0,
+	/* 58 - 720(1440)x480i@240 */
+	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 54000, 720, 739,
+		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 240, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 59 - 1440x480i@240 */
-	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 108000, 1440, 1478,
-		   1602, 1716, 0, 480, 488, 494, 525, 0,
+	/* 59 - 720(1440)x480i@240 */
+	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 54000, 720, 739,
+		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 240, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
@@ -1014,6 +1015,27 @@ module_param_named(edid_fixup, edid_fixup, int, 0400);
 MODULE_PARM_DESC(edid_fixup,
 		 "Minimum number of valid EDID header bytes (0-8, default 6)");
 
+static void drm_get_displayid(struct drm_connector *connector,
+			      struct edid *edid);
+
+static int drm_edid_block_checksum(const u8 *raw_edid)
+{
+	int i;
+	u8 csum = 0;
+	for (i = 0; i < EDID_LENGTH; i++)
+		csum += raw_edid[i];
+
+	return csum;
+}
+
+static bool drm_edid_is_zero(const u8 *in_edid, int length)
+{
+	if (memchr_inv(in_edid, 0, length))
+		return false;
+
+	return true;
+}
+
 /**
  * drm_edid_block_valid - Sanity check the EDID block (base or extension)
  * @raw_edid: pointer to raw EDID block
@@ -1027,8 +1049,7 @@ MODULE_PARM_DESC(edid_fixup,
  */
 bool drm_edid_block_valid(u8 *raw_edid, int block, bool print_bad_edid)
 {
-	int i;
-	u8 csum = 0;
+	u8 csum;
 	struct edid *edid = (struct edid *)raw_edid;
 
 	if (WARN_ON(!raw_edid))
@@ -1048,8 +1069,7 @@ bool drm_edid_block_valid(u8 *raw_edid, int block, bool print_bad_edid)
 		}
 	}
 
-	for (i = 0; i < EDID_LENGTH; i++)
-		csum += raw_edid[i];
+	csum = drm_edid_block_checksum(raw_edid);
 	if (csum) {
 		if (print_bad_edid) {
 			DRM_ERROR("EDID checksum is invalid, remainder is %d\n", csum);
@@ -1080,9 +1100,13 @@ bool drm_edid_block_valid(u8 *raw_edid, int block, bool print_bad_edid)
 
 bad:
 	if (print_bad_edid) {
-		printk(KERN_ERR "Raw EDID:\n");
-		print_hex_dump(KERN_ERR, " \t", DUMP_PREFIX_NONE, 16, 1,
+		if (drm_edid_is_zero(raw_edid, EDID_LENGTH)) {
+			printk(KERN_ERR "EDID block is all zeroes\n");
+		} else {
+			printk(KERN_ERR "Raw EDID:\n");
+			print_hex_dump(KERN_ERR, " \t", DUMP_PREFIX_NONE, 16, 1,
 			       raw_edid, EDID_LENGTH, false);
+		}
 	}
 	return false;
 }
@@ -1115,7 +1139,7 @@ EXPORT_SYMBOL(drm_edid_is_valid);
 #define DDC_SEGMENT_ADDR 0x30
 /**
  * drm_do_probe_ddc_edid() - get EDID information via I2C
- * @adapter: I2C device adaptor
+ * @data: I2C device adapter
  * @buf: EDID data buffer to be filled
  * @block: 128 byte EDID block to start fetching from
  * @len: EDID data buffer length to fetch
@@ -1125,9 +1149,9 @@ EXPORT_SYMBOL(drm_edid_is_valid);
  * Return: 0 on success or -1 on failure.
  */
 static int
-drm_do_probe_ddc_edid(struct i2c_adapter *adapter, unsigned char *buf,
-		      int block, int len)
+drm_do_probe_ddc_edid(void *data, u8 *buf, unsigned int block, size_t len)
 {
+	struct i2c_adapter *adapter = data;
 	unsigned char start = block * EDID_LENGTH;
 	unsigned char segment = block >> 1;
 	unsigned char xfers = segment ? 3 : 2;
@@ -1176,16 +1200,26 @@ drm_do_probe_ddc_edid(struct i2c_adapter *adapter, unsigned char *buf,
 	return ret == xfers ? 0 : -1;
 }
 
-static bool drm_edid_is_zero(u8 *in_edid, int length)
-{
-	if (memchr_inv(in_edid, 0, length))
-		return false;
-
-	return true;
-}
-
-static u8 *
-drm_do_get_edid(struct drm_connector *connector, struct i2c_adapter *adapter)
+/**
+ * drm_do_get_edid - get EDID data using a custom EDID block read function
+ * @connector: connector we're probing
+ * @get_edid_block: EDID block read function
+ * @data: private data passed to the block read function
+ *
+ * When the I2C adapter connected to the DDC bus is hidden behind a device that
+ * exposes a different interface to read EDID blocks this function can be used
+ * to get EDID data using a custom block read function.
+ *
+ * As in the general case the DDC bus is accessible by the kernel at the I2C
+ * level, drivers must make all reasonable efforts to expose it as an I2C
+ * adapter and use drm_get_edid() instead of abusing this function.
+ *
+ * Return: Pointer to valid EDID or NULL if we couldn't find any.
+ */
+struct edid *drm_do_get_edid(struct drm_connector *connector,
+	int (*get_edid_block)(void *data, u8 *buf, unsigned int block,
+			      size_t len),
+	void *data)
 {
 	int i, j = 0, valid_extensions = 0;
 	u8 *block, *new;
@@ -1196,7 +1230,7 @@ drm_do_get_edid(struct drm_connector *connector, struct i2c_adapter *adapter)
 
 	/* base block fetch */
 	for (i = 0; i < 4; i++) {
-		if (drm_do_probe_ddc_edid(adapter, block, 0, EDID_LENGTH))
+		if (get_edid_block(data, block, 0, EDID_LENGTH))
 			goto out;
 		if (drm_edid_block_valid(block, 0, print_bad_edid))
 			break;
@@ -1210,7 +1244,7 @@ drm_do_get_edid(struct drm_connector *connector, struct i2c_adapter *adapter)
 
 	/* if there's no extensions, we're done */
 	if (block[0x7e] == 0)
-		return block;
+		return (struct edid *)block;
 
 	new = krealloc(block, (block[0x7e] + 1) * EDID_LENGTH, GFP_KERNEL);
 	if (!new)
@@ -1219,7 +1253,7 @@ drm_do_get_edid(struct drm_connector *connector, struct i2c_adapter *adapter)
 
 	for (j = 1; j <= block[0x7e]; j++) {
 		for (i = 0; i < 4; i++) {
-			if (drm_do_probe_ddc_edid(adapter,
+			if (get_edid_block(data,
 				  block + (valid_extensions + 1) * EDID_LENGTH,
 				  j, EDID_LENGTH))
 				goto out;
@@ -1247,7 +1281,7 @@ drm_do_get_edid(struct drm_connector *connector, struct i2c_adapter *adapter)
 		block = new;
 	}
 
-	return block;
+	return (struct edid *)block;
 
 carp:
 	if (print_bad_edid) {
@@ -1260,6 +1294,7 @@ out:
 	kfree(block);
 	return NULL;
 }
+EXPORT_SYMBOL_GPL(drm_do_get_edid);
 
 /**
  * drm_probe_ddc() - probe DDC presence
@@ -1289,11 +1324,14 @@ EXPORT_SYMBOL(drm_probe_ddc);
 struct edid *drm_get_edid(struct drm_connector *connector,
 			  struct i2c_adapter *adapter)
 {
-	struct edid *edid = NULL;
+	struct edid *edid;
 
-	if (drm_probe_ddc(adapter))
-		edid = (struct edid *)drm_do_get_edid(connector, adapter);
+	if (!drm_probe_ddc(adapter))
+		return NULL;
 
+	edid = drm_do_get_edid(connector, drm_do_probe_ddc_edid, adapter);
+	if (edid)
+		drm_get_displayid(connector, edid);
 	return edid;
 }
 EXPORT_SYMBOL(drm_get_edid);
@@ -2103,7 +2141,8 @@ static int
 add_inferred_modes(struct drm_connector *connector, struct edid *edid)
 {
 	struct detailed_mode_closure closure = {
-		connector, edid, 0, 0, 0
+		.connector = connector,
+		.edid = edid,
 	};
 
 	if (version_greater(edid, 1, 0))
@@ -2169,7 +2208,8 @@ add_established_modes(struct drm_connector *connector, struct edid *edid)
 		((edid->established_timings.mfg_rsvd & 0x80) << 9);
 	int i, modes = 0;
 	struct detailed_mode_closure closure = {
-		connector, edid, 0, 0, 0
+		.connector = connector,
+		.edid = edid,
 	};
 
 	for (i = 0; i <= EDID_EST_TIMINGS; i++) {
@@ -2227,7 +2267,8 @@ add_standard_modes(struct drm_connector *connector, struct edid *edid)
 {
 	int i, modes = 0;
 	struct detailed_mode_closure closure = {
-		connector, edid, 0, 0, 0
+		.connector = connector,
+		.edid = edid,
 	};
 
 	for (i = 0; i < EDID_STD_TIMINGS; i++) {
@@ -2313,7 +2354,8 @@ static int
 add_cvt_modes(struct drm_connector *connector, struct edid *edid)
 {	
 	struct detailed_mode_closure closure = {
-		connector, edid, 0, 0, 0
+		.connector = connector,
+		.edid = edid,
 	};
 
 	if (version_greater(edid, 1, 2))
@@ -2357,11 +2399,10 @@ add_detailed_modes(struct drm_connector *connector, struct edid *edid,
 		   u32 quirks)
 {
 	struct detailed_mode_closure closure = {
-		connector,
-		edid,
-		1,
-		quirks,
-		0
+		.connector = connector,
+		.edid = edid,
+		.preferred = 1,
+		.quirks = quirks,
 	};
 
 	if (closure.preferred && !version_greater(edid, 1, 3))
@@ -2386,7 +2427,7 @@ add_detailed_modes(struct drm_connector *connector, struct edid *edid,
 /*
  * Search EDID for CEA extension block.
  */
-static u8 *drm_find_cea_extension(struct edid *edid)
+static u8 *drm_find_edid_extension(struct edid *edid, int ext_id)
 {
 	u8 *edid_ext = NULL;
 	int i;
@@ -2398,7 +2439,7 @@ static u8 *drm_find_cea_extension(struct edid *edid)
 	/* Find CEA extension */
 	for (i = 0; i < edid->extensions; i++) {
 		edid_ext = (u8 *)edid + EDID_LENGTH * (i + 1);
-		if (edid_ext[0] == CEA_EXT)
+		if (edid_ext[0] == ext_id)
 			break;
 	}
 
@@ -2408,6 +2449,16 @@ static u8 *drm_find_cea_extension(struct edid *edid)
 	return edid_ext;
 }
 
+static u8 *drm_find_cea_extension(struct edid *edid)
+{
+	return drm_find_edid_extension(edid, CEA_EXT);
+}
+
+static u8 *drm_find_displayid_extension(struct edid *edid)
+{
+	return drm_find_edid_extension(edid, DISPLAYID_EXT);
+}
+
 /*
  * Calculate the alternate clock for the CEA mode
  * (60Hz vs. 59.94Hz etc.)
@@ -3125,9 +3176,12 @@ void drm_edid_to_eld(struct drm_connector *connector, struct edid *edid)
 		}
 	}
 	eld[5] |= sad_count << 4;
-	eld[2] = (20 + mnl + sad_count * 3 + 3) / 4;
 
-	DRM_DEBUG_KMS("ELD size %d, SAD count %d\n", (int)eld[2], sad_count);
+	eld[DRM_ELD_BASELINE_ELD_LEN] =
+		DIV_ROUND_UP(drm_eld_calc_baseline_block_size(eld), 4);
+
+	DRM_DEBUG_KMS("ELD size %d, SAD count %d\n",
+		      drm_eld_size(eld), sad_count);
 }
 EXPORT_SYMBOL(drm_edid_to_eld);
 
@@ -3433,10 +3487,10 @@ EXPORT_SYMBOL(drm_rgb_quant_range_selectable);
 /**
  * drm_assign_hdmi_deep_color_info - detect whether monitor supports
  * hdmi deep color modes and update drm_display_info if so.
- *
  * @edid: monitor EDID information
  * @info: Updated with maximum supported deep color bpc and color format
  *        if deep color supported.
+ * @connector: DRM connector, used only for debug output
  *
  * Parse the CEA extension according to CEA-861-B.
  * Return true if HDMI deep color supported, false if not or unknown.
@@ -3865,3 +3919,123 @@ drm_hdmi_vendor_infoframe_from_display_mode(struct hdmi_vendor_infoframe *frame,
 	return 0;
 }
 EXPORT_SYMBOL(drm_hdmi_vendor_infoframe_from_display_mode);
+
+static int drm_parse_display_id(struct drm_connector *connector,
+				u8 *displayid, int length,
+				bool is_edid_extension)
+{
+	/* if this is an EDID extension the first byte will be 0x70 */
+	int idx = 0;
+	struct displayid_hdr *base;
+	struct displayid_block *block;
+	u8 csum = 0;
+	int i;
+
+	if (is_edid_extension)
+		idx = 1;
+
+	base = (struct displayid_hdr *)&displayid[idx];
+
+	DRM_DEBUG_KMS("base revision 0x%x, length %d, %d %d\n",
+		      base->rev, base->bytes, base->prod_id, base->ext_count);
+
+	if (base->bytes + 5 > length - idx)
+		return -EINVAL;
+
+	for (i = idx; i <= base->bytes + 5; i++) {
+		csum += displayid[i];
+	}
+	if (csum) {
+		DRM_ERROR("DisplayID checksum invalid, remainder is %d\n", csum);
+		return -EINVAL;
+	}
+
+	block = (struct displayid_block *)&displayid[idx + 4];
+	DRM_DEBUG_KMS("block id %d, rev %d, len %d\n",
+		      block->tag, block->rev, block->num_bytes);
+
+	switch (block->tag) {
+	case DATA_BLOCK_TILED_DISPLAY: {
+		struct displayid_tiled_block *tile = (struct displayid_tiled_block *)block;
+
+		u16 w, h;
+		u8 tile_v_loc, tile_h_loc;
+		u8 num_v_tile, num_h_tile;
+		struct drm_tile_group *tg;
+
+		w = tile->tile_size[0] | tile->tile_size[1] << 8;
+		h = tile->tile_size[2] | tile->tile_size[3] << 8;
+
+		num_v_tile = (tile->topo[0] & 0xf) | (tile->topo[2] & 0x30);
+		num_h_tile = (tile->topo[0] >> 4) | ((tile->topo[2] >> 2) & 0x30);
+		tile_v_loc = (tile->topo[1] & 0xf) | ((tile->topo[2] & 0x3) << 4);
+		tile_h_loc = (tile->topo[1] >> 4) | (((tile->topo[2] >> 2) & 0x3) << 4);
+
+		connector->has_tile = true;
+		if (tile->tile_cap & 0x80)
+			connector->tile_is_single_monitor = true;
+
+		connector->num_h_tile = num_h_tile + 1;
+		connector->num_v_tile = num_v_tile + 1;
+		connector->tile_h_loc = tile_h_loc;
+		connector->tile_v_loc = tile_v_loc;
+		connector->tile_h_size = w + 1;
+		connector->tile_v_size = h + 1;
+
+		DRM_DEBUG_KMS("tile cap 0x%x\n", tile->tile_cap);
+		DRM_DEBUG_KMS("tile_size %d x %d\n", w + 1, h + 1);
+		DRM_DEBUG_KMS("topo num tiles %dx%d, location %dx%d\n",
+		       num_h_tile + 1, num_v_tile + 1, tile_h_loc, tile_v_loc);
+		DRM_DEBUG_KMS("vend %c%c%c\n", tile->topology_id[0], tile->topology_id[1], tile->topology_id[2]);
+
+		tg = drm_mode_get_tile_group(connector->dev, tile->topology_id);
+		if (!tg) {
+			tg = drm_mode_create_tile_group(connector->dev, tile->topology_id);
+		}
+		if (!tg)
+			return -ENOMEM;
+
+		if (connector->tile_group != tg) {
+			/* if we haven't got a pointer,
+			   take the reference, drop ref to old tile group */
+			if (connector->tile_group) {
+				drm_mode_put_tile_group(connector->dev, connector->tile_group);
+			}
+			connector->tile_group = tg;
+		} else
+			/* if same tile group, then release the ref we just took. */
+			drm_mode_put_tile_group(connector->dev, tg);
+	}
+		break;
+	default:
+		printk("unknown displayid tag %d\n", block->tag);
+		break;
+	}
+	return 0;
+}
+
+static void drm_get_displayid(struct drm_connector *connector,
+			      struct edid *edid)
+{
+	void *displayid = NULL;
+	int ret;
+	connector->has_tile = false;
+	displayid = drm_find_displayid_extension(edid);
+	if (!displayid) {
+		/* drop reference to any tile group we had */
+		goto out_drop_ref;
+	}
+
+	ret = drm_parse_display_id(connector, displayid, EDID_LENGTH, true);
+	if (ret < 0)
+		goto out_drop_ref;
+	if (!connector->has_tile)
+		goto out_drop_ref;
+	return;
+out_drop_ref:
+	if (connector->tile_group) {
+		drm_mode_put_tile_group(connector->dev, connector->tile_group);
+		connector->tile_group = NULL;
+	}
+	return;
+}
diff --git a/drivers/gpu/drm/drm_edid_load.c b/drivers/gpu/drm/drm_edid_load.c
index 0a235fe61c9b..732cb6f8e653 100644
--- a/drivers/gpu/drm/drm_edid_load.c
+++ b/drivers/gpu/drm/drm_edid_load.c
@@ -254,8 +254,7 @@ static void *edid_load(struct drm_connector *connector, const char *name,
 	    name, connector_name);
 
 out:
-	if (fw)
-		release_firmware(fw);
+	release_firmware(fw);
 	return edid;
 }
 
diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c
index 3144db9dc0f1..dc386ebe5193 100644
--- a/drivers/gpu/drm/drm_fb_helper.c
+++ b/drivers/gpu/drm/drm_fb_helper.c
@@ -126,7 +126,7 @@ int drm_fb_helper_add_one_connector(struct drm_fb_helper *fb_helper, struct drm_
 
 	WARN_ON(!mutex_is_locked(&fb_helper->dev->mode_config.mutex));
 	if (fb_helper->connector_count + 1 > fb_helper->connector_info_alloc_count) {
-		temp = krealloc(fb_helper->connector_info, sizeof(struct drm_fb_helper_connector) * (fb_helper->connector_count + 1), GFP_KERNEL);
+		temp = krealloc(fb_helper->connector_info, sizeof(struct drm_fb_helper_connector *) * (fb_helper->connector_count + 1), GFP_KERNEL);
 		if (!temp)
 			return -ENOMEM;
 
@@ -145,6 +145,31 @@ int drm_fb_helper_add_one_connector(struct drm_fb_helper *fb_helper, struct drm_
 }
 EXPORT_SYMBOL(drm_fb_helper_add_one_connector);
 
+static void remove_from_modeset(struct drm_mode_set *set,
+		struct drm_connector *connector)
+{
+	int i, j;
+
+	for (i = 0; i < set->num_connectors; i++) {
+		if (set->connectors[i] == connector)
+			break;
+	}
+
+	if (i == set->num_connectors)
+		return;
+
+	for (j = i + 1; j < set->num_connectors; j++) {
+		set->connectors[j - 1] = set->connectors[j];
+	}
+	set->num_connectors--;
+
+	/* because i915 is pissy about this..
+	 * TODO maybe need to makes sure we set it back to !=NULL somewhere?
+	 */
+	if (set->num_connectors == 0)
+		set->fb = NULL;
+}
+
 int drm_fb_helper_remove_one_connector(struct drm_fb_helper *fb_helper,
 				       struct drm_connector *connector)
 {
@@ -167,63 +192,14 @@ int drm_fb_helper_remove_one_connector(struct drm_fb_helper *fb_helper,
 	}
 	fb_helper->connector_count--;
 	kfree(fb_helper_connector);
-	return 0;
-}
-EXPORT_SYMBOL(drm_fb_helper_remove_one_connector);
-
-static int drm_fb_helper_parse_command_line(struct drm_fb_helper *fb_helper)
-{
-	struct drm_fb_helper_connector *fb_helper_conn;
-	int i;
-
-	for (i = 0; i < fb_helper->connector_count; i++) {
-		struct drm_cmdline_mode *mode;
-		struct drm_connector *connector;
-		char *option = NULL;
-
-		fb_helper_conn = fb_helper->connector_info[i];
-		connector = fb_helper_conn->connector;
-		mode = &fb_helper_conn->cmdline_mode;
 
-		/* do something on return - turn off connector maybe */
-		if (fb_get_options(connector->name, &option))
-			continue;
-
-		if (drm_mode_parse_command_line_for_connector(option,
-							      connector,
-							      mode)) {
-			if (mode->force) {
-				const char *s;
-				switch (mode->force) {
-				case DRM_FORCE_OFF:
-					s = "OFF";
-					break;
-				case DRM_FORCE_ON_DIGITAL:
-					s = "ON - dig";
-					break;
-				default:
-				case DRM_FORCE_ON:
-					s = "ON";
-					break;
-				}
-
-				DRM_INFO("forcing %s connector %s\n",
-					 connector->name, s);
-				connector->force = mode->force;
-			}
-
-			DRM_DEBUG_KMS("cmdline mode for connector %s %dx%d@%dHz%s%s%s\n",
-				      connector->name,
-				      mode->xres, mode->yres,
-				      mode->refresh_specified ? mode->refresh : 60,
-				      mode->rb ? " reduced blanking" : "",
-				      mode->margins ? " with margins" : "",
-				      mode->interlace ?  " interlaced" : "");
-		}
+	/* also cleanup dangling references to the connector: */
+	for (i = 0; i < fb_helper->crtc_count; i++)
+		remove_from_modeset(&fb_helper->crtc_info[i].mode_set, connector);
 
-	}
 	return 0;
 }
+EXPORT_SYMBOL(drm_fb_helper_remove_one_connector);
 
 static void drm_fb_helper_save_lut_atomic(struct drm_crtc *crtc, struct drm_fb_helper *helper)
 {
@@ -345,10 +321,17 @@ static bool restore_fbdev_mode(struct drm_fb_helper *fb_helper)
 
 	drm_warn_on_modeset_not_all_locked(dev);
 
-	list_for_each_entry(plane, &dev->mode_config.plane_list, head)
+	list_for_each_entry(plane, &dev->mode_config.plane_list, head) {
 		if (plane->type != DRM_PLANE_TYPE_PRIMARY)
 			drm_plane_force_disable(plane);
 
+		if (dev->mode_config.rotation_property) {
+			drm_mode_plane_set_obj_prop(plane,
+						    dev->mode_config.rotation_property,
+						    BIT(DRM_ROTATE_0));
+		}
+	}
+
 	for (i = 0; i < fb_helper->crtc_count; i++) {
 		struct drm_mode_set *mode_set = &fb_helper->crtc_info[i].mode_set;
 		struct drm_crtc *crtc = mode_set->crtc;
@@ -394,9 +377,18 @@ bool drm_fb_helper_restore_fbdev_mode_unlocked(struct drm_fb_helper *fb_helper)
 {
 	struct drm_device *dev = fb_helper->dev;
 	bool ret;
+	bool do_delayed = false;
+
 	drm_modeset_lock_all(dev);
 	ret = restore_fbdev_mode(fb_helper);
+
+	do_delayed = fb_helper->delayed_hotplug;
+	if (do_delayed)
+		fb_helper->delayed_hotplug = false;
 	drm_modeset_unlock_all(dev);
+
+	if (do_delayed)
+		drm_fb_helper_hotplug_event(fb_helper);
 	return ret;
 }
 EXPORT_SYMBOL(drm_fb_helper_restore_fbdev_mode_unlocked);
@@ -419,11 +411,11 @@ static bool drm_fb_helper_force_kernel_mode(void)
 		if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
 			continue;
 
-		/* NOTE: we use lockless flag below to avoid grabbing other
-		 * modeset locks.  So just trylock the underlying mutex
-		 * directly:
+		/*
+		 * NOTE: Use trylock mode to avoid deadlocks and sleeping in
+		 * panic context.
 		 */
-		if (!mutex_trylock(&dev->mode_config.mutex)) {
+		if (__drm_modeset_lock_all(dev, true) != 0) {
 			error = true;
 			continue;
 		}
@@ -432,7 +424,7 @@ static bool drm_fb_helper_force_kernel_mode(void)
 		if (ret)
 			error = true;
 
-		mutex_unlock(&dev->mode_config.mutex);
+		drm_modeset_unlock_all(dev);
 	}
 	return error;
 }
@@ -779,7 +771,9 @@ int drm_fb_helper_setcmap(struct fb_cmap *cmap, struct fb_info *info)
 	int i, j, rc = 0;
 	int start;
 
-	drm_modeset_lock_all(dev);
+	if (__drm_modeset_lock_all(dev, !!oops_in_progress)) {
+		return -EBUSY;
+	}
 	if (!drm_fb_helper_is_bound(fb_helper)) {
 		drm_modeset_unlock_all(dev);
 		return -EBUSY;
@@ -935,10 +929,6 @@ int drm_fb_helper_set_par(struct fb_info *info)
 
 	drm_fb_helper_restore_fbdev_mode_unlocked(fb_helper);
 
-	if (fb_helper->delayed_hotplug) {
-		fb_helper->delayed_hotplug = false;
-		drm_fb_helper_hotplug_event(fb_helper);
-	}
 	return 0;
 }
 EXPORT_SYMBOL(drm_fb_helper_set_par);
@@ -957,7 +947,9 @@ int drm_fb_helper_pan_display(struct fb_var_screeninfo *var,
 	int ret = 0;
 	int i;
 
-	drm_modeset_lock_all(dev);
+	if (__drm_modeset_lock_all(dev, !!oops_in_progress)) {
+		return -EBUSY;
+	}
 	if (!drm_fb_helper_is_bound(fb_helper)) {
 		drm_modeset_unlock_all(dev);
 		return -EBUSY;
@@ -1013,7 +1005,7 @@ static int drm_fb_helper_single_fb_probe(struct drm_fb_helper *fb_helper,
 		struct drm_fb_helper_connector *fb_helper_conn = fb_helper->connector_info[i];
 		struct drm_cmdline_mode *cmdline_mode;
 
-		cmdline_mode = &fb_helper_conn->cmdline_mode;
+		cmdline_mode = &fb_helper_conn->connector->cmdline_mode;
 
 		if (cmdline_mode->bpp_specified) {
 			switch (cmdline_mode->bpp) {
@@ -1042,19 +1034,21 @@ static int drm_fb_helper_single_fb_probe(struct drm_fb_helper *fb_helper,
 	crtc_count = 0;
 	for (i = 0; i < fb_helper->crtc_count; i++) {
 		struct drm_display_mode *desired_mode;
+		int x, y;
 		desired_mode = fb_helper->crtc_info[i].desired_mode;
-
+		x = fb_helper->crtc_info[i].x;
+		y = fb_helper->crtc_info[i].y;
 		if (desired_mode) {
 			if (gamma_size == 0)
 				gamma_size = fb_helper->crtc_info[i].mode_set.crtc->gamma_size;
-			if (desired_mode->hdisplay < sizes.fb_width)
-				sizes.fb_width = desired_mode->hdisplay;
-			if (desired_mode->vdisplay < sizes.fb_height)
-				sizes.fb_height = desired_mode->vdisplay;
-			if (desired_mode->hdisplay > sizes.surface_width)
-				sizes.surface_width = desired_mode->hdisplay;
-			if (desired_mode->vdisplay > sizes.surface_height)
-				sizes.surface_height = desired_mode->vdisplay;
+			if (desired_mode->hdisplay + x < sizes.fb_width)
+				sizes.fb_width = desired_mode->hdisplay + x;
+			if (desired_mode->vdisplay + y < sizes.fb_height)
+				sizes.fb_height = desired_mode->vdisplay + y;
+			if (desired_mode->hdisplay + x > sizes.surface_width)
+				sizes.surface_width = desired_mode->hdisplay + x;
+			if (desired_mode->vdisplay + y > sizes.surface_height)
+				sizes.surface_height = desired_mode->vdisplay + y;
 			crtc_count++;
 		}
 	}
@@ -1260,9 +1254,7 @@ EXPORT_SYMBOL(drm_has_preferred_mode);
 
 static bool drm_has_cmdline_mode(struct drm_fb_helper_connector *fb_connector)
 {
-	struct drm_cmdline_mode *cmdline_mode;
-	cmdline_mode = &fb_connector->cmdline_mode;
-	return cmdline_mode->specified;
+	return fb_connector->connector->cmdline_mode.specified;
 }
 
 struct drm_display_mode *drm_pick_cmdline_mode(struct drm_fb_helper_connector *fb_helper_conn,
@@ -1272,7 +1264,7 @@ struct drm_display_mode *drm_pick_cmdline_mode(struct drm_fb_helper_connector *f
 	struct drm_display_mode *mode = NULL;
 	bool prefer_non_interlace;
 
-	cmdline_mode = &fb_helper_conn->cmdline_mode;
+	cmdline_mode = &fb_helper_conn->connector->cmdline_mode;
 	if (cmdline_mode->specified == false)
 		return mode;
 
@@ -1356,6 +1348,7 @@ static void drm_enable_connectors(struct drm_fb_helper *fb_helper,
 
 static bool drm_target_cloned(struct drm_fb_helper *fb_helper,
 			      struct drm_display_mode **modes,
+			      struct drm_fb_offset *offsets,
 			      bool *enabled, int width, int height)
 {
 	int count, i, j;
@@ -1427,27 +1420,88 @@ static bool drm_target_cloned(struct drm_fb_helper *fb_helper,
 	return false;
 }
 
+static int drm_get_tile_offsets(struct drm_fb_helper *fb_helper,
+				struct drm_display_mode **modes,
+				struct drm_fb_offset *offsets,
+				int idx,
+				int h_idx, int v_idx)
+{
+	struct drm_fb_helper_connector *fb_helper_conn;
+	int i;
+	int hoffset = 0, voffset = 0;
+
+	for (i = 0; i < fb_helper->connector_count; i++) {
+		fb_helper_conn = fb_helper->connector_info[i];
+		if (!fb_helper_conn->connector->has_tile)
+			continue;
+
+		if (!modes[i] && (h_idx || v_idx)) {
+			DRM_DEBUG_KMS("no modes for connector tiled %d %d\n", i,
+				      fb_helper_conn->connector->base.id);
+			continue;
+		}
+		if (fb_helper_conn->connector->tile_h_loc < h_idx)
+			hoffset += modes[i]->hdisplay;
+
+		if (fb_helper_conn->connector->tile_v_loc < v_idx)
+			voffset += modes[i]->vdisplay;
+	}
+	offsets[idx].x = hoffset;
+	offsets[idx].y = voffset;
+	DRM_DEBUG_KMS("returned %d %d for %d %d\n", hoffset, voffset, h_idx, v_idx);
+	return 0;
+}
+
 static bool drm_target_preferred(struct drm_fb_helper *fb_helper,
 				 struct drm_display_mode **modes,
+				 struct drm_fb_offset *offsets,
 				 bool *enabled, int width, int height)
 {
 	struct drm_fb_helper_connector *fb_helper_conn;
 	int i;
-
+	uint64_t conn_configured = 0, mask;
+	int tile_pass = 0;
+	mask = (1 << fb_helper->connector_count) - 1;
+retry:
 	for (i = 0; i < fb_helper->connector_count; i++) {
 		fb_helper_conn = fb_helper->connector_info[i];
 
-		if (enabled[i] == false)
+		if (conn_configured & (1 << i))
+			continue;
+
+		if (enabled[i] == false) {
+			conn_configured |= (1 << i);
+			continue;
+		}
+
+		/* first pass over all the untiled connectors */
+		if (tile_pass == 0 && fb_helper_conn->connector->has_tile)
 			continue;
 
+		if (tile_pass == 1) {
+			if (fb_helper_conn->connector->tile_h_loc != 0 ||
+			    fb_helper_conn->connector->tile_v_loc != 0)
+				continue;
+
+		} else {
+			if (fb_helper_conn->connector->tile_h_loc != tile_pass -1 &&
+			    fb_helper_conn->connector->tile_v_loc != tile_pass - 1)
+			/* if this tile_pass doesn't cover any of the tiles - keep going */
+				continue;
+
+			/* find the tile offsets for this pass - need
+			   to find all tiles left and above */
+			drm_get_tile_offsets(fb_helper, modes, offsets,
+					     i, fb_helper_conn->connector->tile_h_loc, fb_helper_conn->connector->tile_v_loc);
+		}
 		DRM_DEBUG_KMS("looking for cmdline mode on connector %d\n",
 			      fb_helper_conn->connector->base.id);
 
 		/* got for command line mode first */
 		modes[i] = drm_pick_cmdline_mode(fb_helper_conn, width, height);
 		if (!modes[i]) {
-			DRM_DEBUG_KMS("looking for preferred mode on connector %d\n",
-				      fb_helper_conn->connector->base.id);
+			DRM_DEBUG_KMS("looking for preferred mode on connector %d %d\n",
+				      fb_helper_conn->connector->base.id, fb_helper_conn->connector->tile_group ? fb_helper_conn->connector->tile_group->id : 0);
 			modes[i] = drm_has_preferred_mode(fb_helper_conn, width, height);
 		}
 		/* No preferred modes, pick one off the list */
@@ -1457,6 +1511,12 @@ static bool drm_target_preferred(struct drm_fb_helper *fb_helper,
 		}
 		DRM_DEBUG_KMS("found mode %s\n", modes[i] ? modes[i]->name :
 			  "none");
+		conn_configured |= (1 << i);
+	}
+
+	if ((conn_configured & mask) != mask) {
+		tile_pass++;
+		goto retry;
 	}
 	return true;
 }
@@ -1546,6 +1606,7 @@ static void drm_setup_crtcs(struct drm_fb_helper *fb_helper)
 	struct drm_device *dev = fb_helper->dev;
 	struct drm_fb_helper_crtc **crtcs;
 	struct drm_display_mode **modes;
+	struct drm_fb_offset *offsets;
 	struct drm_mode_set *modeset;
 	bool *enabled;
 	int width, height;
@@ -1560,9 +1621,11 @@ static void drm_setup_crtcs(struct drm_fb_helper *fb_helper)
 			sizeof(struct drm_fb_helper_crtc *), GFP_KERNEL);
 	modes = kcalloc(dev->mode_config.num_connector,
 			sizeof(struct drm_display_mode *), GFP_KERNEL);
+	offsets = kcalloc(dev->mode_config.num_connector,
+			  sizeof(struct drm_fb_offset), GFP_KERNEL);
 	enabled = kcalloc(dev->mode_config.num_connector,
 			  sizeof(bool), GFP_KERNEL);
-	if (!crtcs || !modes || !enabled) {
+	if (!crtcs || !modes || !enabled || !offsets) {
 		DRM_ERROR("Memory allocation failed\n");
 		goto out;
 	}
@@ -1572,14 +1635,16 @@ static void drm_setup_crtcs(struct drm_fb_helper *fb_helper)
 
 	if (!(fb_helper->funcs->initial_config &&
 	      fb_helper->funcs->initial_config(fb_helper, crtcs, modes,
+					       offsets,
 					       enabled, width, height))) {
 		memset(modes, 0, dev->mode_config.num_connector*sizeof(modes[0]));
 		memset(crtcs, 0, dev->mode_config.num_connector*sizeof(crtcs[0]));
+		memset(offsets, 0, dev->mode_config.num_connector*sizeof(offsets[0]));
 
-		if (!drm_target_cloned(fb_helper,
-				       modes, enabled, width, height) &&
-		    !drm_target_preferred(fb_helper,
-					  modes, enabled, width, height))
+		if (!drm_target_cloned(fb_helper, modes, offsets,
+				       enabled, width, height) &&
+		    !drm_target_preferred(fb_helper, modes, offsets,
+					  enabled, width, height))
 			DRM_ERROR("Unable to find initial modes\n");
 
 		DRM_DEBUG_KMS("picking CRTCs for %dx%d config\n",
@@ -1599,18 +1664,23 @@ static void drm_setup_crtcs(struct drm_fb_helper *fb_helper)
 	for (i = 0; i < fb_helper->connector_count; i++) {
 		struct drm_display_mode *mode = modes[i];
 		struct drm_fb_helper_crtc *fb_crtc = crtcs[i];
+		struct drm_fb_offset *offset = &offsets[i];
 		modeset = &fb_crtc->mode_set;
 
 		if (mode && fb_crtc) {
-			DRM_DEBUG_KMS("desired mode %s set on crtc %d\n",
-				      mode->name, fb_crtc->mode_set.crtc->base.id);
+			DRM_DEBUG_KMS("desired mode %s set on crtc %d (%d,%d)\n",
+				      mode->name, fb_crtc->mode_set.crtc->base.id, offset->x, offset->y);
 			fb_crtc->desired_mode = mode;
+			fb_crtc->x = offset->x;
+			fb_crtc->y = offset->y;
 			if (modeset->mode)
 				drm_mode_destroy(dev, modeset->mode);
 			modeset->mode = drm_mode_duplicate(dev,
 							   fb_crtc->desired_mode);
 			modeset->connectors[modeset->num_connectors++] = fb_helper->connector_info[i]->connector;
 			modeset->fb = fb_helper->fb;
+			modeset->x = offset->x;
+			modeset->y = offset->y;
 		}
 	}
 
@@ -1619,7 +1689,6 @@ static void drm_setup_crtcs(struct drm_fb_helper *fb_helper)
 		modeset = &fb_helper->crtc_info[i].mode_set;
 		if (modeset->num_connectors == 0) {
 			BUG_ON(modeset->fb);
-			BUG_ON(modeset->num_connectors);
 			if (modeset->mode)
 				drm_mode_destroy(dev, modeset->mode);
 			modeset->mode = NULL;
@@ -1628,6 +1697,7 @@ static void drm_setup_crtcs(struct drm_fb_helper *fb_helper)
 out:
 	kfree(crtcs);
 	kfree(modes);
+	kfree(offsets);
 	kfree(enabled);
 }
 
@@ -1657,8 +1727,6 @@ bool drm_fb_helper_initial_config(struct drm_fb_helper *fb_helper, int bpp_sel)
 	struct drm_device *dev = fb_helper->dev;
 	int count = 0;
 
-	drm_fb_helper_parse_command_line(fb_helper);
-
 	mutex_lock(&dev->mode_config.mutex);
 	count = drm_fb_helper_probe_connector_modes(fb_helper,
 						    dev->mode_config.max_width,
diff --git a/drivers/gpu/drm/drm_flip_work.c b/drivers/gpu/drm/drm_flip_work.c
index f9c7fa3d0012..43d9b950ef9f 100644
--- a/drivers/gpu/drm/drm_flip_work.c
+++ b/drivers/gpu/drm/drm_flip_work.c
@@ -25,6 +25,44 @@
 #include "drm_flip_work.h"
 
 /**
+ * drm_flip_work_allocate_task - allocate a flip-work task
+ * @data: data associated to the task
+ * @flags: allocator flags
+ *
+ * Allocate a drm_flip_task object and attach private data to it.
+ */
+struct drm_flip_task *drm_flip_work_allocate_task(void *data, gfp_t flags)
+{
+	struct drm_flip_task *task;
+
+	task = kzalloc(sizeof(*task), flags);
+	if (task)
+		task->data = data;
+
+	return task;
+}
+EXPORT_SYMBOL(drm_flip_work_allocate_task);
+
+/**
+ * drm_flip_work_queue_task - queue a specific task
+ * @work: the flip-work
+ * @task: the task to handle
+ *
+ * Queues task, that will later be run (passed back to drm_flip_func_t
+ * func) on a work queue after drm_flip_work_commit() is called.
+ */
+void drm_flip_work_queue_task(struct drm_flip_work *work,
+			      struct drm_flip_task *task)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&work->lock, flags);
+	list_add_tail(&task->node, &work->queued);
+	spin_unlock_irqrestore(&work->lock, flags);
+}
+EXPORT_SYMBOL(drm_flip_work_queue_task);
+
+/**
  * drm_flip_work_queue - queue work
  * @work: the flip-work
  * @val: the value to queue
@@ -34,10 +72,14 @@
  */
 void drm_flip_work_queue(struct drm_flip_work *work, void *val)
 {
-	if (kfifo_put(&work->fifo, val)) {
-		atomic_inc(&work->pending);
+	struct drm_flip_task *task;
+
+	task = drm_flip_work_allocate_task(val,
+				drm_can_sleep() ? GFP_KERNEL : GFP_ATOMIC);
+	if (task) {
+		drm_flip_work_queue_task(work, task);
 	} else {
-		DRM_ERROR("%s fifo full!\n", work->name);
+		DRM_ERROR("%s could not allocate task!\n", work->name);
 		work->func(work, val);
 	}
 }
@@ -56,9 +98,12 @@ EXPORT_SYMBOL(drm_flip_work_queue);
 void drm_flip_work_commit(struct drm_flip_work *work,
 		struct workqueue_struct *wq)
 {
-	uint32_t pending = atomic_read(&work->pending);
-	atomic_add(pending, &work->count);
-	atomic_sub(pending, &work->pending);
+	unsigned long flags;
+
+	spin_lock_irqsave(&work->lock, flags);
+	list_splice_tail(&work->queued, &work->commited);
+	INIT_LIST_HEAD(&work->queued);
+	spin_unlock_irqrestore(&work->lock, flags);
 	queue_work(wq, &work->worker);
 }
 EXPORT_SYMBOL(drm_flip_work_commit);
@@ -66,47 +111,46 @@ EXPORT_SYMBOL(drm_flip_work_commit);
 static void flip_worker(struct work_struct *w)
 {
 	struct drm_flip_work *work = container_of(w, struct drm_flip_work, worker);
-	uint32_t count = atomic_read(&work->count);
-	void *val = NULL;
+	struct list_head tasks;
+	unsigned long flags;
+
+	while (1) {
+		struct drm_flip_task *task, *tmp;
+
+		INIT_LIST_HEAD(&tasks);
+		spin_lock_irqsave(&work->lock, flags);
+		list_splice_tail(&work->commited, &tasks);
+		INIT_LIST_HEAD(&work->commited);
+		spin_unlock_irqrestore(&work->lock, flags);
 
-	atomic_sub(count, &work->count);
+		if (list_empty(&tasks))
+			break;
 
-	while(count--)
-		if (!WARN_ON(!kfifo_get(&work->fifo, &val)))
-			work->func(work, val);
+		list_for_each_entry_safe(task, tmp, &tasks, node) {
+			work->func(work, task->data);
+			kfree(task);
+		}
+	}
 }
 
 /**
  * drm_flip_work_init - initialize flip-work
  * @work: the flip-work to initialize
- * @size: the max queue depth
  * @name: debug name
  * @func: the callback work function
  *
  * Initializes/allocates resources for the flip-work
- *
- * RETURNS:
- * Zero on success, error code on failure.
  */
-int drm_flip_work_init(struct drm_flip_work *work, int size,
+void drm_flip_work_init(struct drm_flip_work *work,
 		const char *name, drm_flip_func_t func)
 {
-	int ret;
-
 	work->name = name;
-	atomic_set(&work->count, 0);
-	atomic_set(&work->pending, 0);
+	INIT_LIST_HEAD(&work->queued);
+	INIT_LIST_HEAD(&work->commited);
+	spin_lock_init(&work->lock);
 	work->func = func;
 
-	ret = kfifo_alloc(&work->fifo, size, GFP_KERNEL);
-	if (ret) {
-		DRM_ERROR("could not allocate %s fifo\n", name);
-		return ret;
-	}
-
 	INIT_WORK(&work->worker, flip_worker);
-
-	return 0;
 }
 EXPORT_SYMBOL(drm_flip_work_init);
 
@@ -118,7 +162,6 @@ EXPORT_SYMBOL(drm_flip_work_init);
  */
 void drm_flip_work_cleanup(struct drm_flip_work *work)
 {
-	WARN_ON(!kfifo_is_empty(&work->fifo));
-	kfifo_free(&work->fifo);
+	WARN_ON(!list_empty(&work->queued) || !list_empty(&work->commited));
 }
 EXPORT_SYMBOL(drm_flip_work_cleanup);
diff --git a/drivers/gpu/drm/drm_fops.c b/drivers/gpu/drm/drm_fops.c
index 79d5221c6e41..0b9514b6cd64 100644
--- a/drivers/gpu/drm/drm_fops.c
+++ b/drivers/gpu/drm/drm_fops.c
@@ -39,10 +39,10 @@
 #include <linux/slab.h>
 #include <linux/module.h>
 #include "drm_legacy.h"
+#include "drm_internal.h"
 
 /* from BKL pushdown */
 DEFINE_MUTEX(drm_global_mutex);
-EXPORT_SYMBOL(drm_global_mutex);
 
 static int drm_open_helper(struct file *filp, struct drm_minor *minor);
 
@@ -171,7 +171,7 @@ static int drm_open_helper(struct file *filp, struct drm_minor *minor)
 	init_waitqueue_head(&priv->event_wait);
 	priv->event_space = 4096; /* set aside 4k for event buffer */
 
-	if (dev->driver->driver_features & DRIVER_GEM)
+	if (drm_core_check_feature(dev, DRIVER_GEM))
 		drm_gem_open(dev, priv);
 
 	if (drm_core_check_feature(dev, DRIVER_PRIME))
@@ -256,7 +256,7 @@ out_close:
 out_prime_destroy:
 	if (drm_core_check_feature(dev, DRIVER_PRIME))
 		drm_prime_destroy_file_private(&priv->prime);
-	if (dev->driver->driver_features & DRIVER_GEM)
+	if (drm_core_check_feature(dev, DRIVER_GEM))
 		drm_gem_release(dev, priv);
 	put_pid(priv->pid);
 	kfree(priv);
@@ -268,11 +268,11 @@ static void drm_master_release(struct drm_device *dev, struct file *filp)
 {
 	struct drm_file *file_priv = filp->private_data;
 
-	if (drm_i_have_hw_lock(dev, file_priv)) {
+	if (drm_legacy_i_have_hw_lock(dev, file_priv)) {
 		DRM_DEBUG("File %p released, freeing lock for context %d\n",
 			  filp, _DRM_LOCKING_CONTEXT(file_priv->master->lock.hw_lock->lock));
-		drm_lock_free(&file_priv->master->lock,
-			      _DRM_LOCKING_CONTEXT(file_priv->master->lock.hw_lock->lock));
+		drm_legacy_lock_free(&file_priv->master->lock,
+				     _DRM_LOCKING_CONTEXT(file_priv->master->lock.hw_lock->lock));
 	}
 }
 
@@ -330,8 +330,6 @@ static void drm_legacy_dev_reinit(struct drm_device *dev)
  */
 int drm_lastclose(struct drm_device * dev)
 {
-	struct drm_vma_entry *vma, *vma_temp;
-
 	DRM_DEBUG("\n");
 
 	if (dev->driver->lastclose)
@@ -346,13 +344,7 @@ int drm_lastclose(struct drm_device * dev)
 	drm_agp_clear(dev);
 
 	drm_legacy_sg_cleanup(dev);
-
-	/* Clear vma list (only built for debugging) */
-	list_for_each_entry_safe(vma, vma_temp, &dev->vmalist, head) {
-		list_del(&vma->head);
-		kfree(vma);
-	}
-
+	drm_legacy_vma_flush(dev);
 	drm_legacy_dma_takedown(dev);
 
 	mutex_unlock(&dev->struct_mutex);
@@ -412,14 +404,14 @@ int drm_release(struct inode *inode, struct file *filp)
 		drm_master_release(dev, filp);
 
 	if (drm_core_check_feature(dev, DRIVER_HAVE_DMA))
-		drm_core_reclaim_buffers(dev, file_priv);
+		drm_legacy_reclaim_buffers(dev, file_priv);
 
 	drm_events_release(file_priv);
 
-	if (dev->driver->driver_features & DRIVER_MODESET)
+	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		drm_fb_release(file_priv);
 
-	if (dev->driver->driver_features & DRIVER_GEM)
+	if (drm_core_check_feature(dev, DRIVER_GEM))
 		drm_gem_release(dev, file_priv);
 
 	drm_legacy_ctxbitmap_flush(dev, file_priv);
@@ -464,6 +456,8 @@ int drm_release(struct inode *inode, struct file *filp)
 	if (drm_core_check_feature(dev, DRIVER_PRIME))
 		drm_prime_destroy_file_private(&file_priv->prime);
 
+	WARN_ON(!list_empty(&file_priv->event_list));
+
 	put_pid(file_priv->pid);
 	kfree(file_priv);
 
@@ -521,16 +515,19 @@ ssize_t drm_read(struct file *filp, char __user *buffer,
 	size_t total;
 	ssize_t ret;
 
-	ret = wait_event_interruptible(file_priv->event_wait,
-				       !list_empty(&file_priv->event_list));
-	if (ret < 0)
-		return ret;
+	if ((filp->f_flags & O_NONBLOCK) == 0) {
+		ret = wait_event_interruptible(file_priv->event_wait,
+					       !list_empty(&file_priv->event_list));
+		if (ret < 0)
+			return ret;
+	}
 
 	total = 0;
 	while (drm_dequeue_event(file_priv, total, count, &e)) {
 		if (copy_to_user(buffer + total,
 				 e->event, e->event->length)) {
 			total = -EFAULT;
+			e->destroy(e);
 			break;
 		}
 
@@ -538,7 +535,7 @@ ssize_t drm_read(struct file *filp, char __user *buffer,
 		e->destroy(e);
 	}
 
-	return total;
+	return total ?: -EAGAIN;
 }
 EXPORT_SYMBOL(drm_read);
 
diff --git a/drivers/gpu/drm/drm_gem.c b/drivers/gpu/drm/drm_gem.c
index 6adee4c2afc0..16a164770713 100644
--- a/drivers/gpu/drm/drm_gem.c
+++ b/drivers/gpu/drm/drm_gem.c
@@ -38,6 +38,8 @@
 #include <linux/dma-buf.h>
 #include <drm/drmP.h>
 #include <drm/drm_vma_manager.h>
+#include <drm/drm_gem.h>
+#include "drm_internal.h"
 
 /** @file drm_gem.c
  *
@@ -146,7 +148,7 @@ int drm_gem_object_init(struct drm_device *dev,
 EXPORT_SYMBOL(drm_gem_object_init);
 
 /**
- * drm_gem_object_init - initialize an allocated private GEM object
+ * drm_gem_private_object_init - initialize an allocated private GEM object
  * @dev: drm_device the object should be initialized for
  * @obj: drm_gem_object to initialize
  * @size: object size
@@ -186,7 +188,7 @@ drm_gem_remove_prime_handles(struct drm_gem_object *obj, struct drm_file *filp)
 }
 
 /**
- * drm_gem_object_free - release resources bound to userspace handles
+ * drm_gem_object_handle_free - release resources bound to userspace handles
  * @obj: GEM object to clean up.
  *
  * Called after the last handle to the object has been closed
@@ -307,7 +309,7 @@ EXPORT_SYMBOL(drm_gem_dumb_destroy);
  * drm_gem_handle_create_tail - internal functions to create a handle
  * @file_priv: drm file-private structure to register the handle for
  * @obj: object to register
- * @handlep: pionter to return the created handle to the caller
+ * @handlep: pointer to return the created handle to the caller
  * 
  * This expects the dev->object_name_lock to be held already and will drop it
  * before returning. Used to avoid races in establishing new handles when
@@ -360,7 +362,7 @@ drm_gem_handle_create_tail(struct drm_file *file_priv,
 }
 
 /**
- * gem_handle_create - create a gem handle for an object
+ * drm_gem_handle_create - create a gem handle for an object
  * @file_priv: drm file-private structure to register the handle for
  * @obj: object to register
  * @handlep: pionter to return the created handle to the caller
@@ -369,10 +371,9 @@ drm_gem_handle_create_tail(struct drm_file *file_priv,
  * to the object, which includes a regular reference count. Callers
  * will likely want to dereference the object afterwards.
  */
-int
-drm_gem_handle_create(struct drm_file *file_priv,
-		       struct drm_gem_object *obj,
-		       u32 *handlep)
+int drm_gem_handle_create(struct drm_file *file_priv,
+			  struct drm_gem_object *obj,
+			  u32 *handlep)
 {
 	mutex_lock(&obj->dev->object_name_lock);
 
@@ -579,7 +580,7 @@ drm_gem_close_ioctl(struct drm_device *dev, void *data,
 	struct drm_gem_close *args = data;
 	int ret;
 
-	if (!(dev->driver->driver_features & DRIVER_GEM))
+	if (!drm_core_check_feature(dev, DRIVER_GEM))
 		return -ENODEV;
 
 	ret = drm_gem_handle_delete(file_priv, args->handle);
@@ -606,7 +607,7 @@ drm_gem_flink_ioctl(struct drm_device *dev, void *data,
 	struct drm_gem_object *obj;
 	int ret;
 
-	if (!(dev->driver->driver_features & DRIVER_GEM))
+	if (!drm_core_check_feature(dev, DRIVER_GEM))
 		return -ENODEV;
 
 	obj = drm_gem_object_lookup(dev, file_priv, args->handle);
@@ -659,7 +660,7 @@ drm_gem_open_ioctl(struct drm_device *dev, void *data,
 	int ret;
 	u32 handle;
 
-	if (!(dev->driver->driver_features & DRIVER_GEM))
+	if (!drm_core_check_feature(dev, DRIVER_GEM))
 		return -ENODEV;
 
 	mutex_lock(&dev->object_name_lock);
@@ -887,7 +888,7 @@ int drm_gem_mmap(struct file *filp, struct vm_area_struct *vma)
 					   vma_pages(vma));
 	if (!node) {
 		mutex_unlock(&dev->struct_mutex);
-		return drm_mmap(filp, vma);
+		return -EINVAL;
 	} else if (!drm_vma_node_is_allowed(node, filp)) {
 		mutex_unlock(&dev->struct_mutex);
 		return -EACCES;
diff --git a/drivers/gpu/drm/drm_gem_cma_helper.c b/drivers/gpu/drm/drm_gem_cma_helper.c
index e467e67af6e7..e419eedf751d 100644
--- a/drivers/gpu/drm/drm_gem_cma_helper.c
+++ b/drivers/gpu/drm/drm_gem_cma_helper.c
@@ -29,18 +29,31 @@
 #include <drm/drm_gem_cma_helper.h>
 #include <drm/drm_vma_manager.h>
 
-/*
+/**
+ * DOC: cma helpers
+ *
+ * The Contiguous Memory Allocator reserves a pool of memory at early boot
+ * that is used to service requests for large blocks of contiguous memory.
+ *
+ * The DRM GEM/CMA helpers use this allocator as a means to provide buffer
+ * objects that are physically contiguous in memory. This is useful for
+ * display drivers that are unable to map scattered buffers via an IOMMU.
+ */
+
+/**
  * __drm_gem_cma_create - Create a GEM CMA object without allocating memory
- * @drm: The drm device
- * @size: The GEM object size
+ * @drm: DRM device
+ * @size: size of the object to allocate
  *
- * This function creates and initializes a GEM CMA object of the given size, but
- * doesn't allocate any memory to back the object.
+ * This function creates and initializes a GEM CMA object of the given size,
+ * but doesn't allocate any memory to back the object.
  *
- * Return a struct drm_gem_cma_object* on success or ERR_PTR values on failure.
+ * Returns:
+ * A struct drm_gem_cma_object * on success or an ERR_PTR()-encoded negative
+ * error code on failure.
  */
 static struct drm_gem_cma_object *
-__drm_gem_cma_create(struct drm_device *drm, unsigned int size)
+__drm_gem_cma_create(struct drm_device *drm, size_t size)
 {
 	struct drm_gem_cma_object *cma_obj;
 	struct drm_gem_object *gem_obj;
@@ -69,14 +82,21 @@ error:
 	return ERR_PTR(ret);
 }
 
-/*
+/**
  * drm_gem_cma_create - allocate an object with the given size
+ * @drm: DRM device
+ * @size: size of the object to allocate
+ *
+ * This function creates a CMA GEM object and allocates a contiguous chunk of
+ * memory as backing store. The backing memory has the writecombine attribute
+ * set.
  *
- * returns a struct drm_gem_cma_object* on success or ERR_PTR values
- * on failure.
+ * Returns:
+ * A struct drm_gem_cma_object * on success or an ERR_PTR()-encoded negative
+ * error code on failure.
  */
 struct drm_gem_cma_object *drm_gem_cma_create(struct drm_device *drm,
-		unsigned int size)
+					      size_t size)
 {
 	struct drm_gem_cma_object *cma_obj;
 	int ret;
@@ -104,17 +124,26 @@ error:
 }
 EXPORT_SYMBOL_GPL(drm_gem_cma_create);
 
-/*
- * drm_gem_cma_create_with_handle - allocate an object with the given
- * size and create a gem handle on it
+/**
+ * drm_gem_cma_create_with_handle - allocate an object with the given size and
+ *     return a GEM handle to it
+ * @file_priv: DRM file-private structure to register the handle for
+ * @drm: DRM device
+ * @size: size of the object to allocate
+ * @handle: return location for the GEM handle
+ *
+ * This function creates a CMA GEM object, allocating a physically contiguous
+ * chunk of memory as backing store. The GEM object is then added to the list
+ * of object associated with the given file and a handle to it is returned.
  *
- * returns a struct drm_gem_cma_object* on success or ERR_PTR values
- * on failure.
+ * Returns:
+ * A struct drm_gem_cma_object * on success or an ERR_PTR()-encoded negative
+ * error code on failure.
  */
-static struct drm_gem_cma_object *drm_gem_cma_create_with_handle(
-		struct drm_file *file_priv,
-		struct drm_device *drm, unsigned int size,
-		unsigned int *handle)
+static struct drm_gem_cma_object *
+drm_gem_cma_create_with_handle(struct drm_file *file_priv,
+			       struct drm_device *drm, size_t size,
+			       uint32_t *handle)
 {
 	struct drm_gem_cma_object *cma_obj;
 	struct drm_gem_object *gem_obj;
@@ -145,16 +174,19 @@ err_handle_create:
 	return ERR_PTR(ret);
 }
 
-/*
- * drm_gem_cma_free_object - (struct drm_driver)->gem_free_object callback
- * function
+/**
+ * drm_gem_cma_free_object - free resources associated with a CMA GEM object
+ * @gem_obj: GEM object to free
+ *
+ * This function frees the backing memory of the CMA GEM object, cleans up the
+ * GEM object state and frees the memory used to store the object itself.
+ * Drivers using the CMA helpers should set this as their DRM driver's
+ * ->gem_free_object() callback.
  */
 void drm_gem_cma_free_object(struct drm_gem_object *gem_obj)
 {
 	struct drm_gem_cma_object *cma_obj;
 
-	drm_gem_free_mmap_offset(gem_obj);
-
 	cma_obj = to_drm_gem_cma_obj(gem_obj);
 
 	if (cma_obj->vaddr) {
@@ -170,18 +202,26 @@ void drm_gem_cma_free_object(struct drm_gem_object *gem_obj)
 }
 EXPORT_SYMBOL_GPL(drm_gem_cma_free_object);
 
-/*
- * drm_gem_cma_dumb_create - (struct drm_driver)->dumb_create callback
- * function
+/**
+ * drm_gem_cma_dumb_create_internal - create a dumb buffer object
+ * @file_priv: DRM file-private structure to create the dumb buffer for
+ * @drm: DRM device
+ * @args: IOCTL data
+ *
+ * This aligns the pitch and size arguments to the minimum required. This is
+ * an internal helper that can be wrapped by a driver to account for hardware
+ * with more specific alignment requirements. It should not be used directly
+ * as the ->dumb_create() callback in a DRM driver.
  *
- * This aligns the pitch and size arguments to the minimum required. wrap
- * this into your own function if you need bigger alignment.
+ * Returns:
+ * 0 on success or a negative error code on failure.
  */
-int drm_gem_cma_dumb_create(struct drm_file *file_priv,
-		struct drm_device *dev, struct drm_mode_create_dumb *args)
+int drm_gem_cma_dumb_create_internal(struct drm_file *file_priv,
+				     struct drm_device *drm,
+				     struct drm_mode_create_dumb *args)
 {
+	unsigned int min_pitch = DIV_ROUND_UP(args->width * args->bpp, 8);
 	struct drm_gem_cma_object *cma_obj;
-	int min_pitch = DIV_ROUND_UP(args->width * args->bpp, 8);
 
 	if (args->pitch < min_pitch)
 		args->pitch = min_pitch;
@@ -189,18 +229,63 @@ int drm_gem_cma_dumb_create(struct drm_file *file_priv,
 	if (args->size < args->pitch * args->height)
 		args->size = args->pitch * args->height;
 
-	cma_obj = drm_gem_cma_create_with_handle(file_priv, dev,
-			args->size, &args->handle);
+	cma_obj = drm_gem_cma_create_with_handle(file_priv, drm, args->size,
+						 &args->handle);
+	return PTR_ERR_OR_ZERO(cma_obj);
+}
+EXPORT_SYMBOL_GPL(drm_gem_cma_dumb_create_internal);
+
+/**
+ * drm_gem_cma_dumb_create - create a dumb buffer object
+ * @file_priv: DRM file-private structure to create the dumb buffer for
+ * @drm: DRM device
+ * @args: IOCTL data
+ *
+ * This function computes the pitch of the dumb buffer and rounds it up to an
+ * integer number of bytes per pixel. Drivers for hardware that doesn't have
+ * any additional restrictions on the pitch can directly use this function as
+ * their ->dumb_create() callback.
+ *
+ * For hardware with additional restrictions, drivers can adjust the fields
+ * set up by userspace and pass the IOCTL data along to the
+ * drm_gem_cma_dumb_create_internal() function.
+ *
+ * Returns:
+ * 0 on success or a negative error code on failure.
+ */
+int drm_gem_cma_dumb_create(struct drm_file *file_priv,
+			    struct drm_device *drm,
+			    struct drm_mode_create_dumb *args)
+{
+	struct drm_gem_cma_object *cma_obj;
+
+	args->pitch = DIV_ROUND_UP(args->width * args->bpp, 8);
+	args->size = args->pitch * args->height;
+
+	cma_obj = drm_gem_cma_create_with_handle(file_priv, drm, args->size,
+						 &args->handle);
 	return PTR_ERR_OR_ZERO(cma_obj);
 }
 EXPORT_SYMBOL_GPL(drm_gem_cma_dumb_create);
 
-/*
- * drm_gem_cma_dumb_map_offset - (struct drm_driver)->dumb_map_offset callback
- * function
+/**
+ * drm_gem_cma_dumb_map_offset - return the fake mmap offset for a CMA GEM
+ *     object
+ * @file_priv: DRM file-private structure containing the GEM object
+ * @drm: DRM device
+ * @handle: GEM object handle
+ * @offset: return location for the fake mmap offset
+ *
+ * This function look up an object by its handle and returns the fake mmap
+ * offset associated with it. Drivers using the CMA helpers should set this
+ * as their DRM driver's ->dumb_map_offset() callback.
+ *
+ * Returns:
+ * 0 on success or a negative error code on failure.
  */
 int drm_gem_cma_dumb_map_offset(struct drm_file *file_priv,
-		struct drm_device *drm, uint32_t handle, uint64_t *offset)
+				struct drm_device *drm, u32 handle,
+				u64 *offset)
 {
 	struct drm_gem_object *gem_obj;
 
@@ -208,7 +293,7 @@ int drm_gem_cma_dumb_map_offset(struct drm_file *file_priv,
 
 	gem_obj = drm_gem_object_lookup(drm, file_priv, handle);
 	if (!gem_obj) {
-		dev_err(drm->dev, "failed to lookup gem object\n");
+		dev_err(drm->dev, "failed to lookup GEM object\n");
 		mutex_unlock(&drm->struct_mutex);
 		return -EINVAL;
 	}
@@ -251,8 +336,20 @@ static int drm_gem_cma_mmap_obj(struct drm_gem_cma_object *cma_obj,
 	return ret;
 }
 
-/*
- * drm_gem_cma_mmap - (struct file_operation)->mmap callback function
+/**
+ * drm_gem_cma_mmap - memory-map a CMA GEM object
+ * @filp: file object
+ * @vma: VMA for the area to be mapped
+ *
+ * This function implements an augmented version of the GEM DRM file mmap
+ * operation for CMA objects: In addition to the usual GEM VMA setup it
+ * immediately faults in the entire object instead of using on-demaind
+ * faulting. Drivers which employ the CMA helpers should use this function
+ * as their ->mmap() handler in the DRM device file's file_operations
+ * structure.
+ *
+ * Returns:
+ * 0 on success or a negative error code on failure.
  */
 int drm_gem_cma_mmap(struct file *filp, struct vm_area_struct *vma)
 {
@@ -272,7 +369,16 @@ int drm_gem_cma_mmap(struct file *filp, struct vm_area_struct *vma)
 EXPORT_SYMBOL_GPL(drm_gem_cma_mmap);
 
 #ifdef CONFIG_DEBUG_FS
-void drm_gem_cma_describe(struct drm_gem_cma_object *cma_obj, struct seq_file *m)
+/**
+ * drm_gem_cma_describe - describe a CMA GEM object for debugfs
+ * @cma_obj: CMA GEM object
+ * @m: debugfs file handle
+ *
+ * This function can be used to dump a human-readable representation of the
+ * CMA GEM object into a synthetic file.
+ */
+void drm_gem_cma_describe(struct drm_gem_cma_object *cma_obj,
+			  struct seq_file *m)
 {
 	struct drm_gem_object *obj = &cma_obj->base;
 	struct drm_device *dev = obj->dev;
@@ -291,7 +397,18 @@ void drm_gem_cma_describe(struct drm_gem_cma_object *cma_obj, struct seq_file *m
 EXPORT_SYMBOL_GPL(drm_gem_cma_describe);
 #endif
 
-/* low-level interface prime helpers */
+/**
+ * drm_gem_cma_prime_get_sg_table - provide a scatter/gather table of pinned
+ *     pages for a CMA GEM object
+ * @obj: GEM object
+ *
+ * This function exports a scatter/gather table suitable for PRIME usage by
+ * calling the standard DMA mapping API. Drivers using the CMA helpers should
+ * set this as their DRM driver's ->gem_prime_get_sg_table() callback.
+ *
+ * Returns:
+ * A pointer to the scatter/gather table of pinned pages or NULL on failure.
+ */
 struct sg_table *drm_gem_cma_prime_get_sg_table(struct drm_gem_object *obj)
 {
 	struct drm_gem_cma_object *cma_obj = to_drm_gem_cma_obj(obj);
@@ -315,8 +432,26 @@ out:
 }
 EXPORT_SYMBOL_GPL(drm_gem_cma_prime_get_sg_table);
 
+/**
+ * drm_gem_cma_prime_import_sg_table - produce a CMA GEM object from another
+ *     driver's scatter/gather table of pinned pages
+ * @dev: device to import into
+ * @attach: DMA-BUF attachment
+ * @sgt: scatter/gather table of pinned pages
+ *
+ * This function imports a scatter/gather table exported via DMA-BUF by
+ * another driver. Imported buffers must be physically contiguous in memory
+ * (i.e. the scatter/gather table must contain a single entry). Drivers that
+ * use the CMA helpers should set this as their DRM driver's
+ * ->gem_prime_import_sg_table() callback.
+ *
+ * Returns:
+ * A pointer to a newly created GEM object or an ERR_PTR-encoded negative
+ * error code on failure.
+ */
 struct drm_gem_object *
-drm_gem_cma_prime_import_sg_table(struct drm_device *dev, size_t size,
+drm_gem_cma_prime_import_sg_table(struct drm_device *dev,
+				  struct dma_buf_attachment *attach,
 				  struct sg_table *sgt)
 {
 	struct drm_gem_cma_object *cma_obj;
@@ -325,19 +460,31 @@ drm_gem_cma_prime_import_sg_table(struct drm_device *dev, size_t size,
 		return ERR_PTR(-EINVAL);
 
 	/* Create a CMA GEM buffer. */
-	cma_obj = __drm_gem_cma_create(dev, size);
+	cma_obj = __drm_gem_cma_create(dev, attach->dmabuf->size);
 	if (IS_ERR(cma_obj))
 		return ERR_CAST(cma_obj);
 
 	cma_obj->paddr = sg_dma_address(sgt->sgl);
 	cma_obj->sgt = sgt;
 
-	DRM_DEBUG_PRIME("dma_addr = %pad, size = %zu\n", &cma_obj->paddr, size);
+	DRM_DEBUG_PRIME("dma_addr = %pad, size = %zu\n", &cma_obj->paddr, attach->dmabuf->size);
 
 	return &cma_obj->base;
 }
 EXPORT_SYMBOL_GPL(drm_gem_cma_prime_import_sg_table);
 
+/**
+ * drm_gem_cma_prime_mmap - memory-map an exported CMA GEM object
+ * @obj: GEM object
+ * @vma: VMA for the area to be mapped
+ *
+ * This function maps a buffer imported via DRM PRIME into a userspace
+ * process's address space. Drivers that use the CMA helpers should set this
+ * as their DRM driver's ->gem_prime_mmap() callback.
+ *
+ * Returns:
+ * 0 on success or a negative error code on failure.
+ */
 int drm_gem_cma_prime_mmap(struct drm_gem_object *obj,
 			   struct vm_area_struct *vma)
 {
@@ -356,6 +503,20 @@ int drm_gem_cma_prime_mmap(struct drm_gem_object *obj,
 }
 EXPORT_SYMBOL_GPL(drm_gem_cma_prime_mmap);
 
+/**
+ * drm_gem_cma_prime_vmap - map a CMA GEM object into the kernel's virtual
+ *     address space
+ * @obj: GEM object
+ *
+ * This function maps a buffer exported via DRM PRIME into the kernel's
+ * virtual address space. Since the CMA buffers are already mapped into the
+ * kernel virtual address space this simply returns the cached virtual
+ * address. Drivers using the CMA helpers should set this as their DRM
+ * driver's ->gem_prime_vmap() callback.
+ *
+ * Returns:
+ * The kernel virtual address of the CMA GEM object's backing store.
+ */
 void *drm_gem_cma_prime_vmap(struct drm_gem_object *obj)
 {
 	struct drm_gem_cma_object *cma_obj = to_drm_gem_cma_obj(obj);
@@ -364,6 +525,17 @@ void *drm_gem_cma_prime_vmap(struct drm_gem_object *obj)
 }
 EXPORT_SYMBOL_GPL(drm_gem_cma_prime_vmap);
 
+/**
+ * drm_gem_cma_prime_vunmap - unmap a CMA GEM object from the kernel's virtual
+ *     address space
+ * @obj: GEM object
+ * @vaddr: kernel virtual address where the CMA GEM object was mapped
+ *
+ * This function removes a buffer exported via DRM PRIME from the kernel's
+ * virtual address space. This is a no-op because CMA buffers cannot be
+ * unmapped from kernel space. Drivers using the CMA helpers should set this
+ * as their DRM driver's ->gem_prime_vunmap() callback.
+ */
 void drm_gem_cma_prime_vunmap(struct drm_gem_object *obj, void *vaddr)
 {
 	/* Nothing to do */
diff --git a/drivers/gpu/drm/drm_info.c b/drivers/gpu/drm/drm_info.c
index ecaf0fa2eec8..51efebd434f3 100644
--- a/drivers/gpu/drm/drm_info.c
+++ b/drivers/gpu/drm/drm_info.c
@@ -35,6 +35,9 @@
 
 #include <linux/seq_file.h>
 #include <drm/drmP.h>
+#include <drm/drm_gem.h>
+
+#include "drm_legacy.h"
 
 /**
  * Called when "/proc/dri/.../name" is read.
@@ -183,15 +186,32 @@ int drm_clients_info(struct seq_file *m, void *data)
 	struct drm_device *dev = node->minor->dev;
 	struct drm_file *priv;
 
+	seq_printf(m,
+		   "%20s %5s %3s master a %5s %10s\n",
+		   "command",
+		   "pid",
+		   "dev",
+		   "uid",
+		   "magic");
+
+	/* dev->filelist is sorted youngest first, but we want to present
+	 * oldest first (i.e. kernel, servers, clients), so walk backwardss.
+	 */
 	mutex_lock(&dev->struct_mutex);
-	seq_printf(m, "a dev	pid    uid	magic\n\n");
-	list_for_each_entry(priv, &dev->filelist, lhead) {
-		seq_printf(m, "%c %3d %5d %5d %10u\n",
-			   priv->authenticated ? 'y' : 'n',
-			   priv->minor->index,
+	list_for_each_entry_reverse(priv, &dev->filelist, lhead) {
+		struct task_struct *task;
+
+		rcu_read_lock(); /* locks pid_task()->comm */
+		task = pid_task(priv->pid, PIDTYPE_PID);
+		seq_printf(m, "%20s %5d %3d   %c    %c %5d %10u\n",
+			   task ? task->comm : "<unknown>",
 			   pid_vnr(priv->pid),
+			   priv->minor->index,
+			   priv->is_master ? 'y' : 'n',
+			   priv->authenticated ? 'y' : 'n',
 			   from_kuid_munged(seq_user_ns(m), priv->uid),
 			   priv->magic);
+		rcu_read_unlock();
 	}
 	mutex_unlock(&dev->struct_mutex);
 	return 0;
@@ -223,62 +243,3 @@ int drm_gem_name_info(struct seq_file *m, void *data)
 
 	return 0;
 }
-
-#if DRM_DEBUG_CODE
-
-int drm_vma_info(struct seq_file *m, void *data)
-{
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
-	struct drm_device *dev = node->minor->dev;
-	struct drm_vma_entry *pt;
-	struct vm_area_struct *vma;
-	unsigned long vma_count = 0;
-#if defined(__i386__)
-	unsigned int pgprot;
-#endif
-
-	mutex_lock(&dev->struct_mutex);
-	list_for_each_entry(pt, &dev->vmalist, head)
-		vma_count++;
-
-	seq_printf(m, "vma use count: %lu, high_memory = %pK, 0x%pK\n",
-		   vma_count, high_memory,
-		   (void *)(unsigned long)virt_to_phys(high_memory));
-
-	list_for_each_entry(pt, &dev->vmalist, head) {
-		vma = pt->vma;
-		if (!vma)
-			continue;
-		seq_printf(m,
-			   "\n%5d 0x%pK-0x%pK %c%c%c%c%c%c 0x%08lx000",
-			   pt->pid,
-			   (void *)vma->vm_start, (void *)vma->vm_end,
-			   vma->vm_flags & VM_READ ? 'r' : '-',
-			   vma->vm_flags & VM_WRITE ? 'w' : '-',
-			   vma->vm_flags & VM_EXEC ? 'x' : '-',
-			   vma->vm_flags & VM_MAYSHARE ? 's' : 'p',
-			   vma->vm_flags & VM_LOCKED ? 'l' : '-',
-			   vma->vm_flags & VM_IO ? 'i' : '-',
-			   vma->vm_pgoff);
-
-#if defined(__i386__)
-		pgprot = pgprot_val(vma->vm_page_prot);
-		seq_printf(m, " %c%c%c%c%c%c%c%c%c",
-			   pgprot & _PAGE_PRESENT ? 'p' : '-',
-			   pgprot & _PAGE_RW ? 'w' : 'r',
-			   pgprot & _PAGE_USER ? 'u' : 's',
-			   pgprot & _PAGE_PWT ? 't' : 'b',
-			   pgprot & _PAGE_PCD ? 'u' : 'c',
-			   pgprot & _PAGE_ACCESSED ? 'a' : '-',
-			   pgprot & _PAGE_DIRTY ? 'd' : '-',
-			   pgprot & _PAGE_PSE ? 'm' : 'k',
-			   pgprot & _PAGE_GLOBAL ? 'g' : 'l');
-#endif
-		seq_printf(m, "\n");
-	}
-	mutex_unlock(&dev->struct_mutex);
-	return 0;
-}
-
-#endif
-
diff --git a/drivers/gpu/drm/drm_internal.h b/drivers/gpu/drm/drm_internal.h
new file mode 100644
index 000000000000..7cc0a3516871
--- /dev/null
+++ b/drivers/gpu/drm/drm_internal.h
@@ -0,0 +1,132 @@
+/*
+ * Copyright © 2014 Intel Corporation
+ *   Daniel Vetter <daniel.vetter@ffwll.ch>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+/* drm_irq.c */
+extern unsigned int drm_timestamp_monotonic;
+
+/* drm_fops.c */
+extern struct mutex drm_global_mutex;
+int drm_lastclose(struct drm_device *dev);
+
+/* drm_pci.c */
+int drm_pci_set_unique(struct drm_device *dev,
+		       struct drm_master *master,
+		       struct drm_unique *u);
+int drm_irq_by_busid(struct drm_device *dev, void *data,
+		     struct drm_file *file_priv);
+
+/* drm_vm.c */
+int drm_vma_info(struct seq_file *m, void *data);
+void drm_vm_open_locked(struct drm_device *dev, struct vm_area_struct *vma);
+void drm_vm_close_locked(struct drm_device *dev, struct vm_area_struct *vma);
+
+/* drm_prime.c */
+int drm_prime_handle_to_fd_ioctl(struct drm_device *dev, void *data,
+				 struct drm_file *file_priv);
+int drm_prime_fd_to_handle_ioctl(struct drm_device *dev, void *data,
+				 struct drm_file *file_priv);
+
+void drm_prime_init_file_private(struct drm_prime_file_private *prime_fpriv);
+void drm_prime_destroy_file_private(struct drm_prime_file_private *prime_fpriv);
+void drm_prime_remove_buf_handle_locked(struct drm_prime_file_private *prime_fpriv,
+					struct dma_buf *dma_buf);
+
+/* drm_info.c */
+int drm_name_info(struct seq_file *m, void *data);
+int drm_vm_info(struct seq_file *m, void *data);
+int drm_bufs_info(struct seq_file *m, void *data);
+int drm_vblank_info(struct seq_file *m, void *data);
+int drm_clients_info(struct seq_file *m, void* data);
+int drm_gem_name_info(struct seq_file *m, void *data);
+
+/* drm_irq.c */
+int drm_control(struct drm_device *dev, void *data,
+		struct drm_file *file_priv);
+int drm_modeset_ctl(struct drm_device *dev, void *data,
+		    struct drm_file *file_priv);
+
+/* drm_auth.c */
+int drm_getmagic(struct drm_device *dev, void *data,
+		 struct drm_file *file_priv);
+int drm_authmagic(struct drm_device *dev, void *data,
+		  struct drm_file *file_priv);
+int drm_remove_magic(struct drm_master *master, drm_magic_t magic);
+
+/* drm_sysfs.c */
+extern struct class *drm_class;
+
+struct class *drm_sysfs_create(struct module *owner, char *name);
+void drm_sysfs_destroy(void);
+struct device *drm_sysfs_minor_alloc(struct drm_minor *minor);
+int drm_sysfs_connector_add(struct drm_connector *connector);
+void drm_sysfs_connector_remove(struct drm_connector *connector);
+
+/* drm_gem.c */
+int drm_gem_init(struct drm_device *dev);
+void drm_gem_destroy(struct drm_device *dev);
+int drm_gem_handle_create_tail(struct drm_file *file_priv,
+			       struct drm_gem_object *obj,
+			       u32 *handlep);
+int drm_gem_close_ioctl(struct drm_device *dev, void *data,
+			struct drm_file *file_priv);
+int drm_gem_flink_ioctl(struct drm_device *dev, void *data,
+			struct drm_file *file_priv);
+int drm_gem_open_ioctl(struct drm_device *dev, void *data,
+		       struct drm_file *file_priv);
+void drm_gem_open(struct drm_device *dev, struct drm_file *file_private);
+void drm_gem_release(struct drm_device *dev, struct drm_file *file_private);
+
+/* drm_drv.c */
+int drm_setmaster_ioctl(struct drm_device *dev, void *data,
+			struct drm_file *file_priv);
+int drm_dropmaster_ioctl(struct drm_device *dev, void *data,
+			 struct drm_file *file_priv);
+struct drm_master *drm_master_create(struct drm_minor *minor);
+
+/* drm_debugfs.c */
+#if defined(CONFIG_DEBUG_FS)
+int drm_debugfs_init(struct drm_minor *minor, int minor_id,
+		     struct dentry *root);
+int drm_debugfs_cleanup(struct drm_minor *minor);
+int drm_debugfs_connector_add(struct drm_connector *connector);
+void drm_debugfs_connector_remove(struct drm_connector *connector);
+#else
+static inline int drm_debugfs_init(struct drm_minor *minor, int minor_id,
+				   struct dentry *root)
+{
+	return 0;
+}
+
+static inline int drm_debugfs_cleanup(struct drm_minor *minor)
+{
+	return 0;
+}
+
+static inline int drm_debugfs_connector_add(struct drm_connector *connector)
+{
+	return 0;
+}
+static inline void drm_debugfs_connector_remove(struct drm_connector *connector)
+{
+}
+#endif
diff --git a/drivers/gpu/drm/drm_ioctl.c b/drivers/gpu/drm/drm_ioctl.c
index 40be746b7e68..00587a1e3c83 100644
--- a/drivers/gpu/drm/drm_ioctl.c
+++ b/drivers/gpu/drm/drm_ioctl.c
@@ -31,6 +31,7 @@
 #include <drm/drmP.h>
 #include <drm/drm_core.h>
 #include "drm_legacy.h"
+#include "drm_internal.h"
 
 #include <linux/pci.h>
 #include <linux/export.h>
@@ -41,121 +42,6 @@
 static int drm_version(struct drm_device *dev, void *data,
 		       struct drm_file *file_priv);
 
-#define DRM_IOCTL_DEF(ioctl, _func, _flags) \
-	[DRM_IOCTL_NR(ioctl)] = {.cmd = ioctl, .func = _func, .flags = _flags, .cmd_drv = 0, .name = #ioctl}
-
-/** Ioctl table */
-static const struct drm_ioctl_desc drm_ioctls[] = {
-	DRM_IOCTL_DEF(DRM_IOCTL_VERSION, drm_version, DRM_UNLOCKED|DRM_RENDER_ALLOW),
-	DRM_IOCTL_DEF(DRM_IOCTL_GET_UNIQUE, drm_getunique, 0),
-	DRM_IOCTL_DEF(DRM_IOCTL_GET_MAGIC, drm_getmagic, 0),
-	DRM_IOCTL_DEF(DRM_IOCTL_IRQ_BUSID, drm_irq_by_busid, DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_GET_MAP, drm_getmap, DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_GET_CLIENT, drm_getclient, DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_GET_STATS, drm_getstats, DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_GET_CAP, drm_getcap, DRM_UNLOCKED|DRM_RENDER_ALLOW),
-	DRM_IOCTL_DEF(DRM_IOCTL_SET_CLIENT_CAP, drm_setclientcap, 0),
-	DRM_IOCTL_DEF(DRM_IOCTL_SET_VERSION, drm_setversion, DRM_MASTER),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_SET_UNIQUE, drm_setunique, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_BLOCK, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_UNBLOCK, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_AUTH_MAGIC, drm_authmagic, DRM_AUTH|DRM_MASTER),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_ADD_MAP, drm_addmap_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_RM_MAP, drm_rmmap_ioctl, DRM_AUTH),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_SET_SAREA_CTX, drm_legacy_setsareactx, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_GET_SAREA_CTX, drm_legacy_getsareactx, DRM_AUTH),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_SET_MASTER, drm_setmaster_ioctl, DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_DROP_MASTER, drm_dropmaster_ioctl, DRM_ROOT_ONLY),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_ADD_CTX, drm_legacy_addctx, DRM_AUTH|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_RM_CTX, drm_legacy_rmctx, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_MOD_CTX, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_GET_CTX, drm_legacy_getctx, DRM_AUTH),
-	DRM_IOCTL_DEF(DRM_IOCTL_SWITCH_CTX, drm_legacy_switchctx, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_NEW_CTX, drm_legacy_newctx, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_RES_CTX, drm_legacy_resctx, DRM_AUTH),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_ADD_DRAW, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_RM_DRAW, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_LOCK, drm_lock, DRM_AUTH),
-	DRM_IOCTL_DEF(DRM_IOCTL_UNLOCK, drm_unlock, DRM_AUTH),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_FINISH, drm_noop, DRM_AUTH),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_ADD_BUFS, drm_addbufs, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_MARK_BUFS, drm_markbufs, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_INFO_BUFS, drm_infobufs, DRM_AUTH),
-	DRM_IOCTL_DEF(DRM_IOCTL_MAP_BUFS, drm_mapbufs, DRM_AUTH),
-	DRM_IOCTL_DEF(DRM_IOCTL_FREE_BUFS, drm_freebufs, DRM_AUTH),
-	DRM_IOCTL_DEF(DRM_IOCTL_DMA, drm_dma_ioctl, DRM_AUTH),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_CONTROL, drm_control, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-
-#if __OS_HAS_AGP
-	DRM_IOCTL_DEF(DRM_IOCTL_AGP_ACQUIRE, drm_agp_acquire_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_AGP_RELEASE, drm_agp_release_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_AGP_ENABLE, drm_agp_enable_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_AGP_INFO, drm_agp_info_ioctl, DRM_AUTH),
-	DRM_IOCTL_DEF(DRM_IOCTL_AGP_ALLOC, drm_agp_alloc_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_AGP_FREE, drm_agp_free_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_AGP_BIND, drm_agp_bind_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_AGP_UNBIND, drm_agp_unbind_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-#endif
-
-	DRM_IOCTL_DEF(DRM_IOCTL_SG_ALLOC, drm_sg_alloc, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_SG_FREE, drm_sg_free, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_WAIT_VBLANK, drm_wait_vblank, DRM_UNLOCKED),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_MODESET_CTL, drm_modeset_ctl, 0),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_UPDATE_DRAW, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_GEM_CLOSE, drm_gem_close_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
-	DRM_IOCTL_DEF(DRM_IOCTL_GEM_FLINK, drm_gem_flink_ioctl, DRM_AUTH|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_GEM_OPEN, drm_gem_open_ioctl, DRM_AUTH|DRM_UNLOCKED),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETRESOURCES, drm_mode_getresources, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_PRIME_HANDLE_TO_FD, drm_prime_handle_to_fd_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
-	DRM_IOCTL_DEF(DRM_IOCTL_PRIME_FD_TO_HANDLE, drm_prime_fd_to_handle_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPLANERESOURCES, drm_mode_getplane_res, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETCRTC, drm_mode_getcrtc, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETCRTC, drm_mode_setcrtc, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPLANE, drm_mode_getplane, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETPLANE, drm_mode_setplane, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CURSOR, drm_mode_cursor_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETGAMMA, drm_mode_gamma_get_ioctl, DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETGAMMA, drm_mode_gamma_set_ioctl, DRM_MASTER|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETENCODER, drm_mode_getencoder, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETCONNECTOR, drm_mode_getconnector, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ATTACHMODE, drm_noop, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DETACHMODE, drm_noop, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPROPERTY, drm_mode_getproperty_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETPROPERTY, drm_mode_connector_property_set_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPROPBLOB, drm_mode_getblob_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETFB, drm_mode_getfb, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ADDFB, drm_mode_addfb, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ADDFB2, drm_mode_addfb2, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_RMFB, drm_mode_rmfb, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_PAGE_FLIP, drm_mode_page_flip_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DIRTYFB, drm_mode_dirtyfb_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CREATE_DUMB, drm_mode_create_dumb_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_MAP_DUMB, drm_mode_mmap_dumb_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DESTROY_DUMB, drm_mode_destroy_dumb_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_OBJ_GETPROPERTIES, drm_mode_obj_get_properties_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_OBJ_SETPROPERTY, drm_mode_obj_set_property_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CURSOR2, drm_mode_cursor2_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-};
-
-#define DRM_CORE_IOCTL_COUNT	ARRAY_SIZE( drm_ioctls )
-
 /**
  * Get the bus id.
  *
@@ -167,7 +53,7 @@ static const struct drm_ioctl_desc drm_ioctls[] = {
  *
  * Copies the bus id from drm_device::unique into user space.
  */
-int drm_getunique(struct drm_device *dev, void *data,
+static int drm_getunique(struct drm_device *dev, void *data,
 		  struct drm_file *file_priv)
 {
 	struct drm_unique *u = data;
@@ -189,7 +75,6 @@ drm_unset_busid(struct drm_device *dev,
 	kfree(master->unique);
 	master->unique = NULL;
 	master->unique_len = 0;
-	master->unique_size = 0;
 }
 
 /**
@@ -207,7 +92,7 @@ drm_unset_busid(struct drm_device *dev,
  * version 1.1 or greater. Also note that KMS is all version 1.1 and later and
  * UMS was only ever supported on pci devices.
  */
-int drm_setunique(struct drm_device *dev, void *data,
+static int drm_setunique(struct drm_device *dev, void *data,
 		  struct drm_file *file_priv)
 {
 	struct drm_unique *u = data;
@@ -245,15 +130,15 @@ static int drm_set_busid(struct drm_device *dev, struct drm_file *file_priv)
 	if (master->unique != NULL)
 		drm_unset_busid(dev, master);
 
-	if (dev->driver->bus && dev->driver->bus->set_busid) {
-		ret = dev->driver->bus->set_busid(dev, master);
+	if (dev->driver->set_busid) {
+		ret = dev->driver->set_busid(dev, master);
 		if (ret) {
 			drm_unset_busid(dev, master);
 			return ret;
 		}
 	} else {
 		if (WARN(dev->unique == NULL,
-			 "No drm_bus.set_busid() implementation provided by "
+			 "No drm_driver.set_busid() implementation provided by "
 			 "%ps. Use drm_dev_set_unique() to set the unique "
 			 "name explicitly.", dev->driver))
 			return -EINVAL;
@@ -279,7 +164,7 @@ static int drm_set_busid(struct drm_device *dev, struct drm_file *file_priv)
  * Searches for the mapping with the specified offset and copies its information
  * into userspace
  */
-int drm_getmap(struct drm_device *dev, void *data,
+static int drm_getmap(struct drm_device *dev, void *data,
 	       struct drm_file *file_priv)
 {
 	struct drm_map *map = data;
@@ -340,7 +225,7 @@ int drm_getmap(struct drm_device *dev, void *data,
  * Searches for the client with the specified index and copies its information
  * into userspace
  */
-int drm_getclient(struct drm_device *dev, void *data,
+static int drm_getclient(struct drm_device *dev, void *data,
 		  struct drm_file *file_priv)
 {
 	struct drm_client *client = data;
@@ -380,7 +265,7 @@ int drm_getclient(struct drm_device *dev, void *data,
  *
  * \return zero on success or a negative number on failure.
  */
-int drm_getstats(struct drm_device *dev, void *data,
+static int drm_getstats(struct drm_device *dev, void *data,
 		 struct drm_file *file_priv)
 {
 	struct drm_stats *stats = data;
@@ -394,7 +279,7 @@ int drm_getstats(struct drm_device *dev, void *data,
 /**
  * Get device/driver capabilities
  */
-int drm_getcap(struct drm_device *dev, void *data, struct drm_file *file_priv)
+static int drm_getcap(struct drm_device *dev, void *data, struct drm_file *file_priv)
 {
 	struct drm_get_cap *req = data;
 
@@ -444,7 +329,7 @@ int drm_getcap(struct drm_device *dev, void *data, struct drm_file *file_priv)
 /**
  * Set device/driver capabilities
  */
-int
+static int
 drm_setclientcap(struct drm_device *dev, void *data, struct drm_file *file_priv)
 {
 	struct drm_set_client_cap *req = data;
@@ -478,7 +363,7 @@ drm_setclientcap(struct drm_device *dev, void *data, struct drm_file *file_priv)
  *
  * Sets the requested interface version
  */
-int drm_setversion(struct drm_device *dev, void *data, struct drm_file *file_priv)
+static int drm_setversion(struct drm_device *dev, void *data, struct drm_file *file_priv)
 {
 	struct drm_set_version *sv = data;
 	int if_version, retcode = 0;
@@ -624,6 +509,121 @@ static int drm_ioctl_permit(u32 flags, struct drm_file *file_priv)
 	return 0;
 }
 
+#define DRM_IOCTL_DEF(ioctl, _func, _flags) \
+	[DRM_IOCTL_NR(ioctl)] = {.cmd = ioctl, .func = _func, .flags = _flags, .cmd_drv = 0, .name = #ioctl}
+
+/** Ioctl table */
+static const struct drm_ioctl_desc drm_ioctls[] = {
+	DRM_IOCTL_DEF(DRM_IOCTL_VERSION, drm_version, DRM_UNLOCKED|DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF(DRM_IOCTL_GET_UNIQUE, drm_getunique, 0),
+	DRM_IOCTL_DEF(DRM_IOCTL_GET_MAGIC, drm_getmagic, 0),
+	DRM_IOCTL_DEF(DRM_IOCTL_IRQ_BUSID, drm_irq_by_busid, DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_GET_MAP, drm_getmap, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_GET_CLIENT, drm_getclient, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_GET_STATS, drm_getstats, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_GET_CAP, drm_getcap, DRM_UNLOCKED|DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF(DRM_IOCTL_SET_CLIENT_CAP, drm_setclientcap, 0),
+	DRM_IOCTL_DEF(DRM_IOCTL_SET_VERSION, drm_setversion, DRM_MASTER),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_SET_UNIQUE, drm_setunique, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_BLOCK, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_UNBLOCK, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_AUTH_MAGIC, drm_authmagic, DRM_AUTH|DRM_MASTER),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_ADD_MAP, drm_legacy_addmap_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_RM_MAP, drm_legacy_rmmap_ioctl, DRM_AUTH),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_SET_SAREA_CTX, drm_legacy_setsareactx, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_GET_SAREA_CTX, drm_legacy_getsareactx, DRM_AUTH),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_SET_MASTER, drm_setmaster_ioctl, DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_DROP_MASTER, drm_dropmaster_ioctl, DRM_ROOT_ONLY),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_ADD_CTX, drm_legacy_addctx, DRM_AUTH|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_RM_CTX, drm_legacy_rmctx, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_MOD_CTX, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_GET_CTX, drm_legacy_getctx, DRM_AUTH),
+	DRM_IOCTL_DEF(DRM_IOCTL_SWITCH_CTX, drm_legacy_switchctx, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_NEW_CTX, drm_legacy_newctx, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_RES_CTX, drm_legacy_resctx, DRM_AUTH),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_ADD_DRAW, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_RM_DRAW, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_LOCK, drm_legacy_lock, DRM_AUTH),
+	DRM_IOCTL_DEF(DRM_IOCTL_UNLOCK, drm_legacy_unlock, DRM_AUTH),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_FINISH, drm_noop, DRM_AUTH),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_ADD_BUFS, drm_legacy_addbufs, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_MARK_BUFS, drm_legacy_markbufs, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_INFO_BUFS, drm_legacy_infobufs, DRM_AUTH),
+	DRM_IOCTL_DEF(DRM_IOCTL_MAP_BUFS, drm_legacy_mapbufs, DRM_AUTH),
+	DRM_IOCTL_DEF(DRM_IOCTL_FREE_BUFS, drm_legacy_freebufs, DRM_AUTH),
+	DRM_IOCTL_DEF(DRM_IOCTL_DMA, drm_legacy_dma_ioctl, DRM_AUTH),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_CONTROL, drm_control, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+
+#if __OS_HAS_AGP
+	DRM_IOCTL_DEF(DRM_IOCTL_AGP_ACQUIRE, drm_agp_acquire_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_AGP_RELEASE, drm_agp_release_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_AGP_ENABLE, drm_agp_enable_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_AGP_INFO, drm_agp_info_ioctl, DRM_AUTH),
+	DRM_IOCTL_DEF(DRM_IOCTL_AGP_ALLOC, drm_agp_alloc_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_AGP_FREE, drm_agp_free_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_AGP_BIND, drm_agp_bind_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_AGP_UNBIND, drm_agp_unbind_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+#endif
+
+	DRM_IOCTL_DEF(DRM_IOCTL_SG_ALLOC, drm_legacy_sg_alloc, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_SG_FREE, drm_legacy_sg_free, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_WAIT_VBLANK, drm_wait_vblank, DRM_UNLOCKED),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_MODESET_CTL, drm_modeset_ctl, 0),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_UPDATE_DRAW, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_GEM_CLOSE, drm_gem_close_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF(DRM_IOCTL_GEM_FLINK, drm_gem_flink_ioctl, DRM_AUTH|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_GEM_OPEN, drm_gem_open_ioctl, DRM_AUTH|DRM_UNLOCKED),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETRESOURCES, drm_mode_getresources, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_PRIME_HANDLE_TO_FD, drm_prime_handle_to_fd_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF(DRM_IOCTL_PRIME_FD_TO_HANDLE, drm_prime_fd_to_handle_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPLANERESOURCES, drm_mode_getplane_res, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETCRTC, drm_mode_getcrtc, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETCRTC, drm_mode_setcrtc, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPLANE, drm_mode_getplane, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETPLANE, drm_mode_setplane, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CURSOR, drm_mode_cursor_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETGAMMA, drm_mode_gamma_get_ioctl, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETGAMMA, drm_mode_gamma_set_ioctl, DRM_MASTER|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETENCODER, drm_mode_getencoder, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETCONNECTOR, drm_mode_getconnector, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ATTACHMODE, drm_noop, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DETACHMODE, drm_noop, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPROPERTY, drm_mode_getproperty_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETPROPERTY, drm_mode_connector_property_set_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPROPBLOB, drm_mode_getblob_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETFB, drm_mode_getfb, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ADDFB, drm_mode_addfb, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ADDFB2, drm_mode_addfb2, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_RMFB, drm_mode_rmfb, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_PAGE_FLIP, drm_mode_page_flip_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DIRTYFB, drm_mode_dirtyfb_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CREATE_DUMB, drm_mode_create_dumb_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_MAP_DUMB, drm_mode_mmap_dumb_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DESTROY_DUMB, drm_mode_destroy_dumb_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_OBJ_GETPROPERTIES, drm_mode_obj_get_properties_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_OBJ_SETPROPERTY, drm_mode_obj_set_property_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CURSOR2, drm_mode_cursor2_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+};
+
+#define DRM_CORE_IOCTL_COUNT	ARRAY_SIZE( drm_ioctls )
+
 /**
  * Called whenever a process performs an ioctl on /dev/drm.
  *
diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c
index 08ba1209228e..4d79dad9d44f 100644
--- a/drivers/gpu/drm/drm_irq.c
+++ b/drivers/gpu/drm/drm_irq.c
@@ -34,6 +34,7 @@
 
 #include <drm/drmP.h>
 #include "drm_trace.h"
+#include "drm_internal.h"
 
 #include <linux/interrupt.h>	/* For task queue support */
 #include <linux/slab.h>
@@ -55,12 +56,91 @@
  */
 #define DRM_REDUNDANT_VBLIRQ_THRESH_NS 1000000
 
+static bool
+drm_get_last_vbltimestamp(struct drm_device *dev, int crtc,
+			  struct timeval *tvblank, unsigned flags);
+
+static unsigned int drm_timestamp_precision = 20;  /* Default to 20 usecs. */
+
 /*
- * Clear vblank timestamp buffer for a crtc.
+ * Default to use monotonic timestamps for wait-for-vblank and page-flip
+ * complete events.
  */
-static void clear_vblank_timestamps(struct drm_device *dev, int crtc)
+unsigned int drm_timestamp_monotonic = 1;
+
+static int drm_vblank_offdelay = 5000;    /* Default to 5000 msecs. */
+
+module_param_named(vblankoffdelay, drm_vblank_offdelay, int, 0600);
+module_param_named(timestamp_precision_usec, drm_timestamp_precision, int, 0600);
+module_param_named(timestamp_monotonic, drm_timestamp_monotonic, int, 0600);
+
+/**
+ * drm_update_vblank_count - update the master vblank counter
+ * @dev: DRM device
+ * @crtc: counter to update
+ *
+ * Call back into the driver to update the appropriate vblank counter
+ * (specified by @crtc).  Deal with wraparound, if it occurred, and
+ * update the last read value so we can deal with wraparound on the next
+ * call if necessary.
+ *
+ * Only necessary when going from off->on, to account for frames we
+ * didn't get an interrupt for.
+ *
+ * Note: caller must hold dev->vbl_lock since this reads & writes
+ * device vblank fields.
+ */
+static void drm_update_vblank_count(struct drm_device *dev, int crtc)
 {
-	memset(dev->vblank[crtc].time, 0, sizeof(dev->vblank[crtc].time));
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
+	u32 cur_vblank, diff, tslot;
+	bool rc;
+	struct timeval t_vblank;
+
+	/*
+	 * Interrupts were disabled prior to this call, so deal with counter
+	 * wrap if needed.
+	 * NOTE!  It's possible we lost a full dev->max_vblank_count events
+	 * here if the register is small or we had vblank interrupts off for
+	 * a long time.
+	 *
+	 * We repeat the hardware vblank counter & timestamp query until
+	 * we get consistent results. This to prevent races between gpu
+	 * updating its hardware counter while we are retrieving the
+	 * corresponding vblank timestamp.
+	 */
+	do {
+		cur_vblank = dev->driver->get_vblank_counter(dev, crtc);
+		rc = drm_get_last_vbltimestamp(dev, crtc, &t_vblank, 0);
+	} while (cur_vblank != dev->driver->get_vblank_counter(dev, crtc));
+
+	/* Deal with counter wrap */
+	diff = cur_vblank - vblank->last;
+	if (cur_vblank < vblank->last) {
+		diff += dev->max_vblank_count;
+
+		DRM_DEBUG("last_vblank[%d]=0x%x, cur_vblank=0x%x => diff=0x%x\n",
+			  crtc, vblank->last, cur_vblank, diff);
+	}
+
+	DRM_DEBUG("updating vblank count on crtc %d, missed %d\n",
+		  crtc, diff);
+
+	if (diff == 0)
+		return;
+
+	/* Reinitialize corresponding vblank timestamp if high-precision query
+	 * available. Skip this step if query unsupported or failed. Will
+	 * reinitialize delayed at next vblank interrupt in that case.
+	 */
+	if (rc) {
+		tslot = atomic_read(&vblank->count) + diff;
+		vblanktimestamp(dev, crtc, tslot) = t_vblank;
+	}
+
+	smp_mb__before_atomic();
+	atomic_add(diff, &vblank->count);
+	smp_mb__after_atomic();
 }
 
 /*
@@ -71,10 +151,11 @@ static void clear_vblank_timestamps(struct drm_device *dev, int crtc)
  */
 static void vblank_disable_and_save(struct drm_device *dev, int crtc)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
 	unsigned long irqflags;
 	u32 vblcount;
 	s64 diff_ns;
-	int vblrc;
+	bool vblrc;
 	struct timeval tvblank;
 	int count = DRM_TIMESTAMP_MAXRETRIES;
 
@@ -84,8 +165,28 @@ static void vblank_disable_and_save(struct drm_device *dev, int crtc)
 	 */
 	spin_lock_irqsave(&dev->vblank_time_lock, irqflags);
 
+	/*
+	 * If the vblank interrupt was already disabled update the count
+	 * and timestamp to maintain the appearance that the counter
+	 * has been ticking all along until this time. This makes the
+	 * count account for the entire time between drm_vblank_on() and
+	 * drm_vblank_off().
+	 *
+	 * But only do this if precise vblank timestamps are available.
+	 * Otherwise we might read a totally bogus timestamp since drivers
+	 * lacking precise timestamp support rely upon sampling the system clock
+	 * at vblank interrupt time. Which obviously won't work out well if the
+	 * vblank interrupt is disabled.
+	 */
+	if (!vblank->enabled &&
+	    drm_get_last_vbltimestamp(dev, crtc, &tvblank, 0)) {
+		drm_update_vblank_count(dev, crtc);
+		spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags);
+		return;
+	}
+
 	dev->driver->disable_vblank(dev, crtc);
-	dev->vblank[crtc].enabled = false;
+	vblank->enabled = false;
 
 	/* No further vblank irq's will be processed after
 	 * this point. Get current hardware vblank count and
@@ -100,9 +201,9 @@ static void vblank_disable_and_save(struct drm_device *dev, int crtc)
 	 * delayed gpu counter increment.
 	 */
 	do {
-		dev->vblank[crtc].last = dev->driver->get_vblank_counter(dev, crtc);
+		vblank->last = dev->driver->get_vblank_counter(dev, crtc);
 		vblrc = drm_get_last_vbltimestamp(dev, crtc, &tvblank, 0);
-	} while (dev->vblank[crtc].last != dev->driver->get_vblank_counter(dev, crtc) && (--count) && vblrc);
+	} while (vblank->last != dev->driver->get_vblank_counter(dev, crtc) && (--count) && vblrc);
 
 	if (!count)
 		vblrc = 0;
@@ -110,7 +211,7 @@ static void vblank_disable_and_save(struct drm_device *dev, int crtc)
 	/* Compute time difference to stored timestamp of last vblank
 	 * as updated by last invocation of drm_handle_vblank() in vblank irq.
 	 */
-	vblcount = atomic_read(&dev->vblank[crtc].count);
+	vblcount = atomic_read(&vblank->count);
 	diff_ns = timeval_to_ns(&tvblank) -
 		  timeval_to_ns(&vblanktimestamp(dev, crtc, vblcount));
 
@@ -126,14 +227,18 @@ static void vblank_disable_and_save(struct drm_device *dev, int crtc)
 	 * available. In that case we can't account for this and just
 	 * hope for the best.
 	 */
-	if ((vblrc > 0) && (abs64(diff_ns) > 1000000)) {
-		atomic_inc(&dev->vblank[crtc].count);
+	if (vblrc && (abs64(diff_ns) > 1000000)) {
+		/* Store new timestamp in ringbuffer. */
+		vblanktimestamp(dev, crtc, vblcount + 1) = tvblank;
+
+		/* Increment cooked vblank count. This also atomically commits
+		 * the timestamp computed above.
+		 */
+		smp_mb__before_atomic();
+		atomic_inc(&vblank->count);
 		smp_mb__after_atomic();
 	}
 
-	/* Invalidate all timestamps while vblank irq's are off. */
-	clear_vblank_timestamps(dev, crtc);
-
 	spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags);
 }
 
@@ -164,14 +269,20 @@ static void vblank_disable_fn(unsigned long arg)
 void drm_vblank_cleanup(struct drm_device *dev)
 {
 	int crtc;
+	unsigned long irqflags;
 
 	/* Bail if the driver didn't call drm_vblank_init() */
 	if (dev->num_crtcs == 0)
 		return;
 
 	for (crtc = 0; crtc < dev->num_crtcs; crtc++) {
-		del_timer_sync(&dev->vblank[crtc].disable_timer);
-		vblank_disable_fn((unsigned long)&dev->vblank[crtc]);
+		struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
+
+		del_timer_sync(&vblank->disable_timer);
+
+		spin_lock_irqsave(&dev->vbl_lock, irqflags);
+		vblank_disable_and_save(dev, crtc);
+		spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
 	}
 
 	kfree(dev->vblank);
@@ -204,11 +315,13 @@ int drm_vblank_init(struct drm_device *dev, int num_crtcs)
 		goto err;
 
 	for (i = 0; i < num_crtcs; i++) {
-		dev->vblank[i].dev = dev;
-		dev->vblank[i].crtc = i;
-		init_waitqueue_head(&dev->vblank[i].queue);
-		setup_timer(&dev->vblank[i].disable_timer, vblank_disable_fn,
-			    (unsigned long)&dev->vblank[i]);
+		struct drm_vblank_crtc *vblank = &dev->vblank[i];
+
+		vblank->dev = dev;
+		vblank->crtc = i;
+		init_waitqueue_head(&vblank->queue);
+		setup_timer(&vblank->disable_timer, vblank_disable_fn,
+			    (unsigned long)vblank);
 	}
 
 	DRM_INFO("Supports vblank timestamp caching Rev 2 (21.10.2013).\n");
@@ -224,7 +337,7 @@ int drm_vblank_init(struct drm_device *dev, int num_crtcs)
 	return 0;
 
 err:
-	drm_vblank_cleanup(dev);
+	dev->num_crtcs = 0;
 	return ret;
 }
 EXPORT_SYMBOL(drm_vblank_init);
@@ -360,9 +473,11 @@ int drm_irq_uninstall(struct drm_device *dev)
 	if (dev->num_crtcs) {
 		spin_lock_irqsave(&dev->vbl_lock, irqflags);
 		for (i = 0; i < dev->num_crtcs; i++) {
-			wake_up(&dev->vblank[i].queue);
-			dev->vblank[i].enabled = false;
-			dev->vblank[i].last =
+			struct drm_vblank_crtc *vblank = &dev->vblank[i];
+
+			wake_up(&vblank->queue);
+			vblank->enabled = false;
+			vblank->last =
 				dev->driver->get_vblank_counter(dev, i);
 		}
 		spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
@@ -617,7 +732,7 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 	 * within vblank area, counting down the number of lines until
 	 * start of scanout.
 	 */
-	invbl = vbl_status & DRM_SCANOUTPOS_INVBL;
+	invbl = vbl_status & DRM_SCANOUTPOS_IN_VBLANK;
 
 	/* Convert scanout position into elapsed time at raw_time query
 	 * since start of scanout at first display scanline. delta_ns
@@ -647,7 +762,7 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 
 	vbl_status = DRM_VBLANKTIME_SCANOUTPOS_METHOD;
 	if (invbl)
-		vbl_status |= DRM_VBLANKTIME_INVBL;
+		vbl_status |= DRM_VBLANKTIME_IN_VBLANK;
 
 	return vbl_status;
 }
@@ -679,10 +794,11 @@ static struct timeval get_drm_timestamp(void)
  * call, i.e., it isn't very precisely locked to the true vblank.
  *
  * Returns:
- * Non-zero if timestamp is considered to be very precise, zero otherwise.
+ * True if timestamp is considered to be very precise, false otherwise.
  */
-u32 drm_get_last_vbltimestamp(struct drm_device *dev, int crtc,
-			      struct timeval *tvblank, unsigned flags)
+static bool
+drm_get_last_vbltimestamp(struct drm_device *dev, int crtc,
+			  struct timeval *tvblank, unsigned flags)
 {
 	int ret;
 
@@ -694,7 +810,7 @@ u32 drm_get_last_vbltimestamp(struct drm_device *dev, int crtc,
 		ret = dev->driver->get_vblank_timestamp(dev, crtc, &max_error,
 							tvblank, flags);
 		if (ret > 0)
-			return (u32) ret;
+			return true;
 	}
 
 	/* GPU high precision timestamp query unsupported or failed.
@@ -702,9 +818,8 @@ u32 drm_get_last_vbltimestamp(struct drm_device *dev, int crtc,
 	 */
 	*tvblank = get_drm_timestamp();
 
-	return 0;
+	return false;
 }
-EXPORT_SYMBOL(drm_get_last_vbltimestamp);
 
 /**
  * drm_vblank_count - retrieve "cooked" vblank counter value
@@ -715,16 +830,41 @@ EXPORT_SYMBOL(drm_get_last_vbltimestamp);
  * vblank events since the system was booted, including lost events due to
  * modesetting activity.
  *
+ * This is the legacy version of drm_crtc_vblank_count().
+ *
  * Returns:
  * The software vblank counter.
  */
 u32 drm_vblank_count(struct drm_device *dev, int crtc)
 {
-	return atomic_read(&dev->vblank[crtc].count);
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
+
+	if (WARN_ON(crtc >= dev->num_crtcs))
+		return 0;
+	return atomic_read(&vblank->count);
 }
 EXPORT_SYMBOL(drm_vblank_count);
 
 /**
+ * drm_crtc_vblank_count - retrieve "cooked" vblank counter value
+ * @crtc: which counter to retrieve
+ *
+ * Fetches the "cooked" vblank count value that represents the number of
+ * vblank events since the system was booted, including lost events due to
+ * modesetting activity.
+ *
+ * This is the native KMS version of drm_vblank_count().
+ *
+ * Returns:
+ * The software vblank counter.
+ */
+u32 drm_crtc_vblank_count(struct drm_crtc *crtc)
+{
+	return drm_vblank_count(crtc->dev, drm_crtc_index(crtc));
+}
+EXPORT_SYMBOL(drm_crtc_vblank_count);
+
+/**
  * drm_vblank_count_and_time - retrieve "cooked" vblank counter value
  * and the system timestamp corresponding to that vblank counter value.
  *
@@ -740,18 +880,22 @@ EXPORT_SYMBOL(drm_vblank_count);
 u32 drm_vblank_count_and_time(struct drm_device *dev, int crtc,
 			      struct timeval *vblanktime)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
 	u32 cur_vblank;
 
+	if (WARN_ON(crtc >= dev->num_crtcs))
+		return 0;
+
 	/* Read timestamp from slot of _vblank_time ringbuffer
 	 * that corresponds to current vblank count. Retry if
 	 * count has incremented during readout. This works like
 	 * a seqlock.
 	 */
 	do {
-		cur_vblank = atomic_read(&dev->vblank[crtc].count);
+		cur_vblank = atomic_read(&vblank->count);
 		*vblanktime = vblanktimestamp(dev, crtc, cur_vblank);
 		smp_rmb();
-	} while (cur_vblank != atomic_read(&dev->vblank[crtc].count));
+	} while (cur_vblank != atomic_read(&vblank->count));
 
 	return cur_vblank;
 }
@@ -781,6 +925,8 @@ static void send_vblank_event(struct drm_device *dev,
  *
  * Updates sequence # and timestamp on event, and sends it to userspace.
  * Caller must hold event lock.
+ *
+ * This is the legacy version of drm_crtc_send_vblank_event().
  */
 void drm_send_vblank_event(struct drm_device *dev, int crtc,
 		struct drm_pending_vblank_event *e)
@@ -800,68 +946,21 @@ void drm_send_vblank_event(struct drm_device *dev, int crtc,
 EXPORT_SYMBOL(drm_send_vblank_event);
 
 /**
- * drm_update_vblank_count - update the master vblank counter
- * @dev: DRM device
- * @crtc: counter to update
- *
- * Call back into the driver to update the appropriate vblank counter
- * (specified by @crtc).  Deal with wraparound, if it occurred, and
- * update the last read value so we can deal with wraparound on the next
- * call if necessary.
+ * drm_crtc_send_vblank_event - helper to send vblank event after pageflip
+ * @crtc: the source CRTC of the vblank event
+ * @e: the event to send
  *
- * Only necessary when going from off->on, to account for frames we
- * didn't get an interrupt for.
+ * Updates sequence # and timestamp on event, and sends it to userspace.
+ * Caller must hold event lock.
  *
- * Note: caller must hold dev->vbl_lock since this reads & writes
- * device vblank fields.
+ * This is the native KMS version of drm_send_vblank_event().
  */
-static void drm_update_vblank_count(struct drm_device *dev, int crtc)
+void drm_crtc_send_vblank_event(struct drm_crtc *crtc,
+				struct drm_pending_vblank_event *e)
 {
-	u32 cur_vblank, diff, tslot, rc;
-	struct timeval t_vblank;
-
-	/*
-	 * Interrupts were disabled prior to this call, so deal with counter
-	 * wrap if needed.
-	 * NOTE!  It's possible we lost a full dev->max_vblank_count events
-	 * here if the register is small or we had vblank interrupts off for
-	 * a long time.
-	 *
-	 * We repeat the hardware vblank counter & timestamp query until
-	 * we get consistent results. This to prevent races between gpu
-	 * updating its hardware counter while we are retrieving the
-	 * corresponding vblank timestamp.
-	 */
-	do {
-		cur_vblank = dev->driver->get_vblank_counter(dev, crtc);
-		rc = drm_get_last_vbltimestamp(dev, crtc, &t_vblank, 0);
-	} while (cur_vblank != dev->driver->get_vblank_counter(dev, crtc));
-
-	/* Deal with counter wrap */
-	diff = cur_vblank - dev->vblank[crtc].last;
-	if (cur_vblank < dev->vblank[crtc].last) {
-		diff += dev->max_vblank_count;
-
-		DRM_DEBUG("last_vblank[%d]=0x%x, cur_vblank=0x%x => diff=0x%x\n",
-			  crtc, dev->vblank[crtc].last, cur_vblank, diff);
-	}
-
-	DRM_DEBUG("enabling vblank interrupts on crtc %d, missed %d\n",
-		  crtc, diff);
-
-	/* Reinitialize corresponding vblank timestamp if high-precision query
-	 * available. Skip this step if query unsupported or failed. Will
-	 * reinitialize delayed at next vblank interrupt in that case.
-	 */
-	if (rc) {
-		tslot = atomic_read(&dev->vblank[crtc].count) + diff;
-		vblanktimestamp(dev, crtc, tslot) = t_vblank;
-	}
-
-	smp_mb__before_atomic();
-	atomic_add(diff, &dev->vblank[crtc].count);
-	smp_mb__after_atomic();
+	drm_send_vblank_event(crtc->dev, drm_crtc_index(crtc), e);
 }
+EXPORT_SYMBOL(drm_crtc_send_vblank_event);
 
 /**
  * drm_vblank_enable - enable the vblank interrupt on a CRTC
@@ -870,13 +969,14 @@ static void drm_update_vblank_count(struct drm_device *dev, int crtc)
  */
 static int drm_vblank_enable(struct drm_device *dev, int crtc)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
 	int ret = 0;
 
 	assert_spin_locked(&dev->vbl_lock);
 
 	spin_lock(&dev->vblank_time_lock);
 
-	if (!dev->vblank[crtc].enabled) {
+	if (!vblank->enabled) {
 		/*
 		 * Enable vblank irqs under vblank_time_lock protection.
 		 * All vblank count & timestamp updates are held off
@@ -887,9 +987,9 @@ static int drm_vblank_enable(struct drm_device *dev, int crtc)
 		ret = dev->driver->enable_vblank(dev, crtc);
 		DRM_DEBUG("enabling vblank on crtc %d, ret: %d\n", crtc, ret);
 		if (ret)
-			atomic_dec(&dev->vblank[crtc].refcount);
+			atomic_dec(&vblank->refcount);
 		else {
-			dev->vblank[crtc].enabled = true;
+			vblank->enabled = true;
 			drm_update_vblank_count(dev, crtc);
 		}
 	}
@@ -914,16 +1014,20 @@ static int drm_vblank_enable(struct drm_device *dev, int crtc)
  */
 int drm_vblank_get(struct drm_device *dev, int crtc)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
 	unsigned long irqflags;
 	int ret = 0;
 
+	if (WARN_ON(crtc >= dev->num_crtcs))
+		return -EINVAL;
+
 	spin_lock_irqsave(&dev->vbl_lock, irqflags);
 	/* Going from 0->1 means we have to enable interrupts again */
-	if (atomic_add_return(1, &dev->vblank[crtc].refcount) == 1) {
+	if (atomic_add_return(1, &vblank->refcount) == 1) {
 		ret = drm_vblank_enable(dev, crtc);
 	} else {
-		if (!dev->vblank[crtc].enabled) {
-			atomic_dec(&dev->vblank[crtc].refcount);
+		if (!vblank->enabled) {
+			atomic_dec(&vblank->refcount);
 			ret = -EINVAL;
 		}
 	}
@@ -963,13 +1067,24 @@ EXPORT_SYMBOL(drm_crtc_vblank_get);
  */
 void drm_vblank_put(struct drm_device *dev, int crtc)
 {
-	BUG_ON(atomic_read(&dev->vblank[crtc].refcount) == 0);
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
+
+	if (WARN_ON(atomic_read(&vblank->refcount) == 0))
+		return;
+
+	if (WARN_ON(crtc >= dev->num_crtcs))
+		return;
 
 	/* Last user schedules interrupt disable */
-	if (atomic_dec_and_test(&dev->vblank[crtc].refcount) &&
-	    (drm_vblank_offdelay > 0))
-		mod_timer(&dev->vblank[crtc].disable_timer,
-			  jiffies + ((drm_vblank_offdelay * HZ)/1000));
+	if (atomic_dec_and_test(&vblank->refcount)) {
+		if (drm_vblank_offdelay == 0)
+			return;
+		else if (dev->vblank_disable_immediate || drm_vblank_offdelay < 0)
+			vblank_disable_fn((unsigned long)vblank);
+		else
+			mod_timer(&vblank->disable_timer,
+				  jiffies + ((drm_vblank_offdelay * HZ)/1000));
+	}
 }
 EXPORT_SYMBOL(drm_vblank_put);
 
@@ -989,6 +1104,50 @@ void drm_crtc_vblank_put(struct drm_crtc *crtc)
 EXPORT_SYMBOL(drm_crtc_vblank_put);
 
 /**
+ * drm_wait_one_vblank - wait for one vblank
+ * @dev: DRM device
+ * @crtc: crtc index
+ *
+ * This waits for one vblank to pass on @crtc, using the irq driver interfaces.
+ * It is a failure to call this when the vblank irq for @crtc is disabled, e.g.
+ * due to lack of driver support or because the crtc is off.
+ */
+void drm_wait_one_vblank(struct drm_device *dev, int crtc)
+{
+	int ret;
+	u32 last;
+
+	ret = drm_vblank_get(dev, crtc);
+	if (WARN(ret, "vblank not available on crtc %i, ret=%i\n", crtc, ret))
+		return;
+
+	last = drm_vblank_count(dev, crtc);
+
+	ret = wait_event_timeout(dev->vblank[crtc].queue,
+				 last != drm_vblank_count(dev, crtc),
+				 msecs_to_jiffies(100));
+
+	WARN(ret == 0, "vblank wait timed out on crtc %i\n", crtc);
+
+	drm_vblank_put(dev, crtc);
+}
+EXPORT_SYMBOL(drm_wait_one_vblank);
+
+/**
+ * drm_crtc_wait_one_vblank - wait for one vblank
+ * @crtc: DRM crtc
+ *
+ * This waits for one vblank to pass on @crtc, using the irq driver interfaces.
+ * It is a failure to call this when the vblank irq for @crtc is disabled, e.g.
+ * due to lack of driver support or because the crtc is off.
+ */
+void drm_crtc_wait_one_vblank(struct drm_crtc *crtc)
+{
+	drm_wait_one_vblank(crtc->dev, drm_crtc_index(crtc));
+}
+EXPORT_SYMBOL(drm_crtc_wait_one_vblank);
+
+/**
  * drm_vblank_off - disable vblank events on a CRTC
  * @dev: DRM device
  * @crtc: CRTC in question
@@ -1004,19 +1163,34 @@ EXPORT_SYMBOL(drm_crtc_vblank_put);
  */
 void drm_vblank_off(struct drm_device *dev, int crtc)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
 	struct drm_pending_vblank_event *e, *t;
 	struct timeval now;
 	unsigned long irqflags;
 	unsigned int seq;
 
-	spin_lock_irqsave(&dev->vbl_lock, irqflags);
+	if (WARN_ON(crtc >= dev->num_crtcs))
+		return;
+
+	spin_lock_irqsave(&dev->event_lock, irqflags);
+
+	spin_lock(&dev->vbl_lock);
 	vblank_disable_and_save(dev, crtc);
-	wake_up(&dev->vblank[crtc].queue);
+	wake_up(&vblank->queue);
+
+	/*
+	 * Prevent subsequent drm_vblank_get() from re-enabling
+	 * the vblank interrupt by bumping the refcount.
+	 */
+	if (!vblank->inmodeset) {
+		atomic_inc(&vblank->refcount);
+		vblank->inmodeset = 1;
+	}
+	spin_unlock(&dev->vbl_lock);
 
 	/* Send any queued vblank events, lest the natives grow disquiet */
 	seq = drm_vblank_count_and_time(dev, crtc, &now);
 
-	spin_lock(&dev->event_lock);
 	list_for_each_entry_safe(e, t, &dev->vblank_event_list, base.link) {
 		if (e->pipe != crtc)
 			continue;
@@ -1027,9 +1201,7 @@ void drm_vblank_off(struct drm_device *dev, int crtc)
 		drm_vblank_put(dev, e->pipe);
 		send_vblank_event(dev, e, seq, &now);
 	}
-	spin_unlock(&dev->event_lock);
-
-	spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
+	spin_unlock_irqrestore(&dev->event_lock, irqflags);
 }
 EXPORT_SYMBOL(drm_vblank_off);
 
@@ -1059,18 +1231,42 @@ EXPORT_SYMBOL(drm_crtc_vblank_off);
  *
  * This functions restores the vblank interrupt state captured with
  * drm_vblank_off() again. Note that calls to drm_vblank_on() and
- * drm_vblank_off() can be unbalanced and so can also be unconditionaly called
+ * drm_vblank_off() can be unbalanced and so can also be unconditionally called
  * in driver load code to reflect the current hardware state of the crtc.
  *
  * This is the legacy version of drm_crtc_vblank_on().
  */
 void drm_vblank_on(struct drm_device *dev, int crtc)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
 	unsigned long irqflags;
 
+	if (WARN_ON(crtc >= dev->num_crtcs))
+		return;
+
 	spin_lock_irqsave(&dev->vbl_lock, irqflags);
-	/* re-enable interrupts if there's are users left */
-	if (atomic_read(&dev->vblank[crtc].refcount) != 0)
+	/* Drop our private "prevent drm_vblank_get" refcount */
+	if (vblank->inmodeset) {
+		atomic_dec(&vblank->refcount);
+		vblank->inmodeset = 0;
+	}
+
+	/*
+	 * sample the current counter to avoid random jumps
+	 * when drm_vblank_enable() applies the diff
+	 *
+	 * -1 to make sure user will never see the same
+	 * vblank counter value before and after a modeset
+	 */
+	vblank->last =
+		(dev->driver->get_vblank_counter(dev, crtc) - 1) &
+		dev->max_vblank_count;
+	/*
+	 * re-enable interrupts if there are users left, or the
+	 * user wishes vblank interrupts to be enabled all the time.
+	 */
+	if (atomic_read(&vblank->refcount) != 0 ||
+	    (!dev->vblank_disable_immediate && drm_vblank_offdelay == 0))
 		WARN_ON(drm_vblank_enable(dev, crtc));
 	spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
 }
@@ -1082,7 +1278,7 @@ EXPORT_SYMBOL(drm_vblank_on);
  *
  * This functions restores the vblank interrupt state captured with
  * drm_vblank_off() again. Note that calls to drm_vblank_on() and
- * drm_vblank_off() can be unbalanced and so can also be unconditionaly called
+ * drm_vblank_off() can be unbalanced and so can also be unconditionally called
  * in driver load code to reflect the current hardware state of the crtc.
  *
  * This is the native kms version of drm_vblank_on().
@@ -1118,9 +1314,15 @@ EXPORT_SYMBOL(drm_crtc_vblank_on);
  */
 void drm_vblank_pre_modeset(struct drm_device *dev, int crtc)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
+
 	/* vblank is not initialized (IRQ not installed ?), or has been freed */
 	if (!dev->num_crtcs)
 		return;
+
+	if (WARN_ON(crtc >= dev->num_crtcs))
+		return;
+
 	/*
 	 * To avoid all the problems that might happen if interrupts
 	 * were enabled/disabled around or between these calls, we just
@@ -1128,10 +1330,10 @@ void drm_vblank_pre_modeset(struct drm_device *dev, int crtc)
 	 * to avoid corrupting the count if multiple, mismatch calls occur),
 	 * so that interrupts remain enabled in the interim.
 	 */
-	if (!dev->vblank[crtc].inmodeset) {
-		dev->vblank[crtc].inmodeset = 0x1;
+	if (!vblank->inmodeset) {
+		vblank->inmodeset = 0x1;
 		if (drm_vblank_get(dev, crtc) == 0)
-			dev->vblank[crtc].inmodeset |= 0x2;
+			vblank->inmodeset |= 0x2;
 	}
 }
 EXPORT_SYMBOL(drm_vblank_pre_modeset);
@@ -1146,21 +1348,22 @@ EXPORT_SYMBOL(drm_vblank_pre_modeset);
  */
 void drm_vblank_post_modeset(struct drm_device *dev, int crtc)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
 	unsigned long irqflags;
 
 	/* vblank is not initialized (IRQ not installed ?), or has been freed */
 	if (!dev->num_crtcs)
 		return;
 
-	if (dev->vblank[crtc].inmodeset) {
+	if (vblank->inmodeset) {
 		spin_lock_irqsave(&dev->vbl_lock, irqflags);
 		dev->vblank_disable_allowed = true;
 		spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
 
-		if (dev->vblank[crtc].inmodeset & 0x2)
+		if (vblank->inmodeset & 0x2)
 			drm_vblank_put(dev, crtc);
 
-		dev->vblank[crtc].inmodeset = 0;
+		vblank->inmodeset = 0;
 	}
 }
 EXPORT_SYMBOL(drm_vblank_post_modeset);
@@ -1212,6 +1415,7 @@ static int drm_queue_vblank_event(struct drm_device *dev, int pipe,
 				  union drm_wait_vblank *vblwait,
 				  struct drm_file *file_priv)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[pipe];
 	struct drm_pending_vblank_event *e;
 	struct timeval now;
 	unsigned long flags;
@@ -1235,6 +1439,18 @@ static int drm_queue_vblank_event(struct drm_device *dev, int pipe,
 
 	spin_lock_irqsave(&dev->event_lock, flags);
 
+	/*
+	 * drm_vblank_off() might have been called after we called
+	 * drm_vblank_get(). drm_vblank_off() holds event_lock
+	 * around the vblank disable, so no need for further locking.
+	 * The reference from drm_vblank_get() protects against
+	 * vblank disable from another source.
+	 */
+	if (!vblank->enabled) {
+		ret = -EINVAL;
+		goto err_unlock;
+	}
+
 	if (file_priv->event_space < sizeof e->event) {
 		ret = -EBUSY;
 		goto err_unlock;
@@ -1295,6 +1511,7 @@ err_put:
 int drm_wait_vblank(struct drm_device *dev, void *data,
 		    struct drm_file *file_priv)
 {
+	struct drm_vblank_crtc *vblank;
 	union drm_wait_vblank *vblwait = data;
 	int ret;
 	unsigned int flags, seq, crtc, high_crtc;
@@ -1324,6 +1541,8 @@ int drm_wait_vblank(struct drm_device *dev, void *data,
 	if (crtc >= dev->num_crtcs)
 		return -EINVAL;
 
+	vblank = &dev->vblank[crtc];
+
 	ret = drm_vblank_get(dev, crtc);
 	if (ret) {
 		DRM_DEBUG("failed to acquire vblank counter, %d\n", ret);
@@ -1356,11 +1575,11 @@ int drm_wait_vblank(struct drm_device *dev, void *data,
 
 	DRM_DEBUG("waiting on vblank count %d, crtc %d\n",
 		  vblwait->request.sequence, crtc);
-	dev->vblank[crtc].last_wait = vblwait->request.sequence;
-	DRM_WAIT_ON(ret, dev->vblank[crtc].queue, 3 * HZ,
+	vblank->last_wait = vblwait->request.sequence;
+	DRM_WAIT_ON(ret, vblank->queue, 3 * HZ,
 		    (((drm_vblank_count(dev, crtc) -
 		       vblwait->request.sequence) <= (1 << 23)) ||
-		     !dev->vblank[crtc].enabled ||
+		     !vblank->enabled ||
 		     !dev->irq_enabled));
 
 	if (ret != -EINTR) {
@@ -1385,12 +1604,11 @@ static void drm_handle_vblank_events(struct drm_device *dev, int crtc)
 {
 	struct drm_pending_vblank_event *e, *t;
 	struct timeval now;
-	unsigned long flags;
 	unsigned int seq;
 
-	seq = drm_vblank_count_and_time(dev, crtc, &now);
+	assert_spin_locked(&dev->event_lock);
 
-	spin_lock_irqsave(&dev->event_lock, flags);
+	seq = drm_vblank_count_and_time(dev, crtc, &now);
 
 	list_for_each_entry_safe(e, t, &dev->vblank_event_list, base.link) {
 		if (e->pipe != crtc)
@@ -1406,8 +1624,6 @@ static void drm_handle_vblank_events(struct drm_device *dev, int crtc)
 		send_vblank_event(dev, e, seq, &now);
 	}
 
-	spin_unlock_irqrestore(&dev->event_lock, flags);
-
 	trace_drm_vblank_event(crtc, seq);
 }
 
@@ -1418,9 +1634,12 @@ static void drm_handle_vblank_events(struct drm_device *dev, int crtc)
  *
  * Drivers should call this routine in their vblank interrupt handlers to
  * update the vblank counter and send any signals that may be pending.
+ *
+ * This is the legacy version of drm_crtc_handle_vblank().
  */
 bool drm_handle_vblank(struct drm_device *dev, int crtc)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
 	u32 vblcount;
 	s64 diff_ns;
 	struct timeval tvblank;
@@ -1429,15 +1648,21 @@ bool drm_handle_vblank(struct drm_device *dev, int crtc)
 	if (!dev->num_crtcs)
 		return false;
 
+	if (WARN_ON(crtc >= dev->num_crtcs))
+		return false;
+
+	spin_lock_irqsave(&dev->event_lock, irqflags);
+
 	/* Need timestamp lock to prevent concurrent execution with
 	 * vblank enable/disable, as this would cause inconsistent
 	 * or corrupted timestamps and vblank counts.
 	 */
-	spin_lock_irqsave(&dev->vblank_time_lock, irqflags);
+	spin_lock(&dev->vblank_time_lock);
 
 	/* Vblank irq handling disabled. Nothing to do. */
-	if (!dev->vblank[crtc].enabled) {
-		spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags);
+	if (!vblank->enabled) {
+		spin_unlock(&dev->vblank_time_lock);
+		spin_unlock_irqrestore(&dev->event_lock, irqflags);
 		return false;
 	}
 
@@ -1446,7 +1671,7 @@ bool drm_handle_vblank(struct drm_device *dev, int crtc)
 	 */
 
 	/* Get current timestamp and count. */
-	vblcount = atomic_read(&dev->vblank[crtc].count);
+	vblcount = atomic_read(&vblank->count);
 	drm_get_last_vbltimestamp(dev, crtc, &tvblank, DRM_CALLED_FROM_VBLIRQ);
 
 	/* Compute time difference to timestamp of last vblank */
@@ -1470,17 +1695,38 @@ bool drm_handle_vblank(struct drm_device *dev, int crtc)
 		 * the timestamp computed above.
 		 */
 		smp_mb__before_atomic();
-		atomic_inc(&dev->vblank[crtc].count);
+		atomic_inc(&vblank->count);
 		smp_mb__after_atomic();
 	} else {
 		DRM_DEBUG("crtc %d: Redundant vblirq ignored. diff_ns = %d\n",
 			  crtc, (int) diff_ns);
 	}
 
-	wake_up(&dev->vblank[crtc].queue);
+	spin_unlock(&dev->vblank_time_lock);
+
+	wake_up(&vblank->queue);
 	drm_handle_vblank_events(dev, crtc);
 
-	spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags);
+	spin_unlock_irqrestore(&dev->event_lock, irqflags);
+
 	return true;
 }
 EXPORT_SYMBOL(drm_handle_vblank);
+
+/**
+ * drm_crtc_handle_vblank - handle a vblank event
+ * @crtc: where this event occurred
+ *
+ * Drivers should call this routine in their vblank interrupt handlers to
+ * update the vblank counter and send any signals that may be pending.
+ *
+ * This is the native KMS version of drm_handle_vblank().
+ *
+ * Returns:
+ * True if the event was successfully handled, false on failure.
+ */
+bool drm_crtc_handle_vblank(struct drm_crtc *crtc)
+{
+	return drm_handle_vblank(crtc->dev, drm_crtc_index(crtc));
+}
+EXPORT_SYMBOL(drm_crtc_handle_vblank);
diff --git a/drivers/gpu/drm/drm_legacy.h b/drivers/gpu/drm/drm_legacy.h
index d34f20a79b7c..c1dc61473db5 100644
--- a/drivers/gpu/drm/drm_legacy.h
+++ b/drivers/gpu/drm/drm_legacy.h
@@ -23,6 +23,15 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 
+/*
+ * This file contains legacy interfaces that modern drm drivers
+ * should no longer be using. They cannot be removed as legacy
+ * drivers use them, and removing them are API breaks.
+ */
+#include <linux/list.h>
+#include <drm/drm_legacy.h>
+
+struct agp_memory;
 struct drm_device;
 struct drm_file;
 
@@ -48,4 +57,57 @@ int drm_legacy_rmctx(struct drm_device *d, void *v, struct drm_file *f);
 int drm_legacy_setsareactx(struct drm_device *d, void *v, struct drm_file *f);
 int drm_legacy_getsareactx(struct drm_device *d, void *v, struct drm_file *f);
 
+/*
+ * Generic Buffer Management
+ */
+
+#define DRM_MAP_HASH_OFFSET 0x10000000
+
+int drm_legacy_addmap_ioctl(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_rmmap_ioctl(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_addbufs(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_infobufs(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_markbufs(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_freebufs(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_mapbufs(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_dma_ioctl(struct drm_device *d, void *v, struct drm_file *f);
+
+void drm_legacy_vma_flush(struct drm_device *d);
+
+/*
+ * AGP Support
+ */
+
+struct drm_agp_mem {
+	unsigned long handle;
+	struct agp_memory *memory;
+	unsigned long bound;
+	int pages;
+	struct list_head head;
+};
+
+/*
+ * Generic Userspace Locking-API
+ */
+
+int drm_legacy_i_have_hw_lock(struct drm_device *d, struct drm_file *f);
+int drm_legacy_lock(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_unlock(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_lock_free(struct drm_lock_data *lock, unsigned int ctx);
+
+/* DMA support */
+int drm_legacy_dma_setup(struct drm_device *dev);
+void drm_legacy_dma_takedown(struct drm_device *dev);
+void drm_legacy_free_buffer(struct drm_device *dev,
+			    struct drm_buf * buf);
+void drm_legacy_reclaim_buffers(struct drm_device *dev,
+				struct drm_file *filp);
+
+/* Scatter Gather Support */
+void drm_legacy_sg_cleanup(struct drm_device *dev);
+int drm_legacy_sg_alloc(struct drm_device *dev, void *data,
+			struct drm_file *file_priv);
+int drm_legacy_sg_free(struct drm_device *dev, void *data,
+		       struct drm_file *file_priv);
+
 #endif /* __DRM_LEGACY_H__ */
diff --git a/drivers/gpu/drm/drm_lock.c b/drivers/gpu/drm/drm_lock.c
index e26b59e385ff..f861361a635e 100644
--- a/drivers/gpu/drm/drm_lock.c
+++ b/drivers/gpu/drm/drm_lock.c
@@ -36,6 +36,7 @@
 #include <linux/export.h>
 #include <drm/drmP.h>
 #include "drm_legacy.h"
+#include "drm_internal.h"
 
 static int drm_notifier(void *priv);
 
@@ -52,7 +53,8 @@ static int drm_lock_take(struct drm_lock_data *lock_data, unsigned int context);
  *
  * Add the current task to the lock wait queue, and attempt to take to lock.
  */
-int drm_lock(struct drm_device *dev, void *data, struct drm_file *file_priv)
+int drm_legacy_lock(struct drm_device *dev, void *data,
+		    struct drm_file *file_priv)
 {
 	DECLARE_WAITQUEUE(entry, current);
 	struct drm_lock *lock = data;
@@ -120,7 +122,7 @@ int drm_lock(struct drm_device *dev, void *data, struct drm_file *file_priv)
 		sigaddset(&dev->sigmask, SIGTTOU);
 		dev->sigdata.context = lock->context;
 		dev->sigdata.lock = master->lock.hw_lock;
-		block_all_signals(drm_notifier, &dev->sigdata, &dev->sigmask);
+		block_all_signals(drm_notifier, dev, &dev->sigmask);
 	}
 
 	if (dev->driver->dma_quiescent && (lock->flags & _DRM_LOCK_QUIESCENT))
@@ -146,7 +148,7 @@ int drm_lock(struct drm_device *dev, void *data, struct drm_file *file_priv)
  *
  * Transfer and free the lock.
  */
-int drm_unlock(struct drm_device *dev, void *data, struct drm_file *file_priv)
+int drm_legacy_unlock(struct drm_device *dev, void *data, struct drm_file *file_priv)
 {
 	struct drm_lock *lock = data;
 	struct drm_master *master = file_priv->master;
@@ -157,7 +159,7 @@ int drm_unlock(struct drm_device *dev, void *data, struct drm_file *file_priv)
 		return -EINVAL;
 	}
 
-	if (drm_lock_free(&master->lock, lock->context)) {
+	if (drm_legacy_lock_free(&master->lock, lock->context)) {
 		/* FIXME: Should really bail out here. */
 	}
 
@@ -250,7 +252,7 @@ static int drm_lock_transfer(struct drm_lock_data *lock_data,
  * Marks the lock as not held, via the \p cmpxchg instruction. Wakes any task
  * waiting on the lock queue.
  */
-int drm_lock_free(struct drm_lock_data *lock_data, unsigned int context)
+int drm_legacy_lock_free(struct drm_lock_data *lock_data, unsigned int context)
 {
 	unsigned int old, new, prev;
 	volatile unsigned int *lock = &lock_data->hw_lock->lock;
@@ -286,26 +288,27 @@ int drm_lock_free(struct drm_lock_data *lock_data, unsigned int context)
  * If the lock is not held, then let the signal proceed as usual.  If the lock
  * is held, then set the contended flag and keep the signal blocked.
  *
- * \param priv pointer to a drm_sigdata structure.
+ * \param priv pointer to a drm_device structure.
  * \return one if the signal should be delivered normally, or zero if the
  * signal should be blocked.
  */
 static int drm_notifier(void *priv)
 {
-	struct drm_sigdata *s = (struct drm_sigdata *) priv;
+	struct drm_device *dev = priv;
+	struct drm_hw_lock *lock = dev->sigdata.lock;
 	unsigned int old, new, prev;
 
 	/* Allow signal delivery if lock isn't held */
-	if (!s->lock || !_DRM_LOCK_IS_HELD(s->lock->lock)
-	    || _DRM_LOCKING_CONTEXT(s->lock->lock) != s->context)
+	if (!lock || !_DRM_LOCK_IS_HELD(lock->lock)
+	    || _DRM_LOCKING_CONTEXT(lock->lock) != dev->sigdata.context)
 		return 1;
 
 	/* Otherwise, set flag to force call to
 	   drmUnlock */
 	do {
-		old = s->lock->lock;
+		old = lock->lock;
 		new = old | _DRM_LOCK_CONT;
-		prev = cmpxchg(&s->lock->lock, old, new);
+		prev = cmpxchg(&lock->lock, old, new);
 	} while (prev != old);
 	return 0;
 }
@@ -323,7 +326,7 @@ static int drm_notifier(void *priv)
  * having to worry about starvation.
  */
 
-void drm_idlelock_take(struct drm_lock_data *lock_data)
+void drm_legacy_idlelock_take(struct drm_lock_data *lock_data)
 {
 	int ret;
 
@@ -340,9 +343,9 @@ void drm_idlelock_take(struct drm_lock_data *lock_data)
 	}
 	spin_unlock_bh(&lock_data->spinlock);
 }
-EXPORT_SYMBOL(drm_idlelock_take);
+EXPORT_SYMBOL(drm_legacy_idlelock_take);
 
-void drm_idlelock_release(struct drm_lock_data *lock_data)
+void drm_legacy_idlelock_release(struct drm_lock_data *lock_data)
 {
 	unsigned int old, prev;
 	volatile unsigned int *lock = &lock_data->hw_lock->lock;
@@ -360,9 +363,10 @@ void drm_idlelock_release(struct drm_lock_data *lock_data)
 	}
 	spin_unlock_bh(&lock_data->spinlock);
 }
-EXPORT_SYMBOL(drm_idlelock_release);
+EXPORT_SYMBOL(drm_legacy_idlelock_release);
 
-int drm_i_have_hw_lock(struct drm_device *dev, struct drm_file *file_priv)
+int drm_legacy_i_have_hw_lock(struct drm_device *dev,
+			      struct drm_file *file_priv)
 {
 	struct drm_master *master = file_priv->master;
 	return (file_priv->lock_count && master->lock.hw_lock &&
diff --git a/drivers/gpu/drm/drm_memory.c b/drivers/gpu/drm/drm_memory.c
index 00c67c0f2381..a521ef6ff807 100644
--- a/drivers/gpu/drm/drm_memory.c
+++ b/drivers/gpu/drm/drm_memory.c
@@ -36,8 +36,20 @@
 #include <linux/highmem.h>
 #include <linux/export.h>
 #include <drm/drmP.h>
+#include "drm_legacy.h"
 
 #if __OS_HAS_AGP
+
+#ifdef HAVE_PAGE_AGP
+# include <asm/agp.h>
+#else
+# ifdef __powerpc__
+#  define PAGE_AGP	__pgprot(_PAGE_KERNEL | _PAGE_NO_CACHE)
+# else
+#  define PAGE_AGP	PAGE_KERNEL
+# endif
+#endif
+
 static void *agp_remap(unsigned long offset, unsigned long size,
 		       struct drm_device * dev)
 {
@@ -108,25 +120,25 @@ static inline void *agp_remap(unsigned long offset, unsigned long size,
 
 #endif				/* agp */
 
-void drm_core_ioremap(struct drm_local_map *map, struct drm_device *dev)
+void drm_legacy_ioremap(struct drm_local_map *map, struct drm_device *dev)
 {
 	if (dev->agp && dev->agp->cant_use_aperture && map->type == _DRM_AGP)
 		map->handle = agp_remap(map->offset, map->size, dev);
 	else
 		map->handle = ioremap(map->offset, map->size);
 }
-EXPORT_SYMBOL(drm_core_ioremap);
+EXPORT_SYMBOL(drm_legacy_ioremap);
 
-void drm_core_ioremap_wc(struct drm_local_map *map, struct drm_device *dev)
+void drm_legacy_ioremap_wc(struct drm_local_map *map, struct drm_device *dev)
 {
 	if (dev->agp && dev->agp->cant_use_aperture && map->type == _DRM_AGP)
 		map->handle = agp_remap(map->offset, map->size, dev);
 	else
 		map->handle = ioremap_wc(map->offset, map->size);
 }
-EXPORT_SYMBOL(drm_core_ioremap_wc);
+EXPORT_SYMBOL(drm_legacy_ioremap_wc);
 
-void drm_core_ioremapfree(struct drm_local_map *map, struct drm_device *dev)
+void drm_legacy_ioremapfree(struct drm_local_map *map, struct drm_device *dev)
 {
 	if (!map->handle || !map->size)
 		return;
@@ -136,4 +148,4 @@ void drm_core_ioremapfree(struct drm_local_map *map, struct drm_device *dev)
 	else
 		iounmap(map->handle);
 }
-EXPORT_SYMBOL(drm_core_ioremapfree);
+EXPORT_SYMBOL(drm_legacy_ioremapfree);
diff --git a/drivers/gpu/drm/drm_mipi_dsi.c b/drivers/gpu/drm/drm_mipi_dsi.c
index 6aa6a9e95570..c0644bb865f2 100644
--- a/drivers/gpu/drm/drm_mipi_dsi.c
+++ b/drivers/gpu/drm/drm_mipi_dsi.c
@@ -35,6 +35,16 @@
 
 #include <video/mipi_display.h>
 
+/**
+ * DOC: dsi helpers
+ *
+ * These functions contain some common logic and helpers to deal with MIPI DSI
+ * peripherals.
+ *
+ * Helpers are provided for a number of standard MIPI DSI command as well as a
+ * subset of the MIPI DCS command set.
+ */
+
 static int mipi_dsi_device_match(struct device *dev, struct device_driver *drv)
 {
 	return of_driver_match_device(dev, drv);
@@ -57,6 +67,29 @@ static struct bus_type mipi_dsi_bus_type = {
 	.pm = &mipi_dsi_device_pm_ops,
 };
 
+static int of_device_match(struct device *dev, void *data)
+{
+	return dev->of_node == data;
+}
+
+/**
+ * of_find_mipi_dsi_device_by_node() - find the MIPI DSI device matching a
+ *    device tree node
+ * @np: device tree node
+ *
+ * Return: A pointer to the MIPI DSI device corresponding to @np or NULL if no
+ *    such device exists (or has not been registered yet).
+ */
+struct mipi_dsi_device *of_find_mipi_dsi_device_by_node(struct device_node *np)
+{
+	struct device *dev;
+
+	dev = bus_find_device(&mipi_dsi_bus_type, NULL, np, of_device_match);
+
+	return dev ? to_mipi_dsi_device(dev) : NULL;
+}
+EXPORT_SYMBOL(of_find_mipi_dsi_device_by_node);
+
 static void mipi_dsi_dev_release(struct device *dev)
 {
 	struct mipi_dsi_device *dsi = to_mipi_dsi_device(dev);
@@ -198,56 +231,351 @@ int mipi_dsi_detach(struct mipi_dsi_device *dsi)
 }
 EXPORT_SYMBOL(mipi_dsi_detach);
 
+static ssize_t mipi_dsi_device_transfer(struct mipi_dsi_device *dsi,
+					struct mipi_dsi_msg *msg)
+{
+	const struct mipi_dsi_host_ops *ops = dsi->host->ops;
+
+	if (!ops || !ops->transfer)
+		return -ENOSYS;
+
+	if (dsi->mode_flags & MIPI_DSI_MODE_LPM)
+		msg->flags |= MIPI_DSI_MSG_USE_LPM;
+
+	return ops->transfer(dsi->host, msg);
+}
+
 /**
- * mipi_dsi_dcs_write - send DCS write command
- * @dsi: DSI device
- * @data: pointer to the command followed by parameters
- * @len: length of @data
+ * mipi_dsi_packet_format_is_short - check if a packet is of the short format
+ * @type: MIPI DSI data type of the packet
+ *
+ * Return: true if the packet for the given data type is a short packet, false
+ * otherwise.
  */
-ssize_t mipi_dsi_dcs_write(struct mipi_dsi_device *dsi, const void *data,
-			    size_t len)
+bool mipi_dsi_packet_format_is_short(u8 type)
+{
+	switch (type) {
+	case MIPI_DSI_V_SYNC_START:
+	case MIPI_DSI_V_SYNC_END:
+	case MIPI_DSI_H_SYNC_START:
+	case MIPI_DSI_H_SYNC_END:
+	case MIPI_DSI_END_OF_TRANSMISSION:
+	case MIPI_DSI_COLOR_MODE_OFF:
+	case MIPI_DSI_COLOR_MODE_ON:
+	case MIPI_DSI_SHUTDOWN_PERIPHERAL:
+	case MIPI_DSI_TURN_ON_PERIPHERAL:
+	case MIPI_DSI_GENERIC_SHORT_WRITE_0_PARAM:
+	case MIPI_DSI_GENERIC_SHORT_WRITE_1_PARAM:
+	case MIPI_DSI_GENERIC_SHORT_WRITE_2_PARAM:
+	case MIPI_DSI_GENERIC_READ_REQUEST_0_PARAM:
+	case MIPI_DSI_GENERIC_READ_REQUEST_1_PARAM:
+	case MIPI_DSI_GENERIC_READ_REQUEST_2_PARAM:
+	case MIPI_DSI_DCS_SHORT_WRITE:
+	case MIPI_DSI_DCS_SHORT_WRITE_PARAM:
+	case MIPI_DSI_DCS_READ:
+	case MIPI_DSI_SET_MAXIMUM_RETURN_PACKET_SIZE:
+		return true;
+	}
+
+	return false;
+}
+EXPORT_SYMBOL(mipi_dsi_packet_format_is_short);
+
+/**
+ * mipi_dsi_packet_format_is_long - check if a packet is of the long format
+ * @type: MIPI DSI data type of the packet
+ *
+ * Return: true if the packet for the given data type is a long packet, false
+ * otherwise.
+ */
+bool mipi_dsi_packet_format_is_long(u8 type)
+{
+	switch (type) {
+	case MIPI_DSI_NULL_PACKET:
+	case MIPI_DSI_BLANKING_PACKET:
+	case MIPI_DSI_GENERIC_LONG_WRITE:
+	case MIPI_DSI_DCS_LONG_WRITE:
+	case MIPI_DSI_LOOSELY_PACKED_PIXEL_STREAM_YCBCR20:
+	case MIPI_DSI_PACKED_PIXEL_STREAM_YCBCR24:
+	case MIPI_DSI_PACKED_PIXEL_STREAM_YCBCR16:
+	case MIPI_DSI_PACKED_PIXEL_STREAM_30:
+	case MIPI_DSI_PACKED_PIXEL_STREAM_36:
+	case MIPI_DSI_PACKED_PIXEL_STREAM_YCBCR12:
+	case MIPI_DSI_PACKED_PIXEL_STREAM_16:
+	case MIPI_DSI_PACKED_PIXEL_STREAM_18:
+	case MIPI_DSI_PIXEL_STREAM_3BYTE_18:
+	case MIPI_DSI_PACKED_PIXEL_STREAM_24:
+		return true;
+	}
+
+	return false;
+}
+EXPORT_SYMBOL(mipi_dsi_packet_format_is_long);
+
+/**
+ * mipi_dsi_create_packet - create a packet from a message according to the
+ *     DSI protocol
+ * @packet: pointer to a DSI packet structure
+ * @msg: message to translate into a packet
+ *
+ * Return: 0 on success or a negative error code on failure.
+ */
+int mipi_dsi_create_packet(struct mipi_dsi_packet *packet,
+			   const struct mipi_dsi_msg *msg)
+{
+	const u8 *tx = msg->tx_buf;
+
+	if (!packet || !msg)
+		return -EINVAL;
+
+	/* do some minimum sanity checking */
+	if (!mipi_dsi_packet_format_is_short(msg->type) &&
+	    !mipi_dsi_packet_format_is_long(msg->type))
+		return -EINVAL;
+
+	if (msg->channel > 3)
+		return -EINVAL;
+
+	memset(packet, 0, sizeof(*packet));
+	packet->header[0] = ((msg->channel & 0x3) << 6) | (msg->type & 0x3f);
+
+	/* TODO: compute ECC if hardware support is not available */
+
+	/*
+	 * Long write packets contain the word count in header bytes 1 and 2.
+	 * The payload follows the header and is word count bytes long.
+	 *
+	 * Short write packets encode up to two parameters in header bytes 1
+	 * and 2.
+	 */
+	if (mipi_dsi_packet_format_is_long(msg->type)) {
+		packet->header[1] = (msg->tx_len >> 0) & 0xff;
+		packet->header[2] = (msg->tx_len >> 8) & 0xff;
+
+		packet->payload_length = msg->tx_len;
+		packet->payload = tx;
+	} else {
+		packet->header[1] = (msg->tx_len > 0) ? tx[0] : 0;
+		packet->header[2] = (msg->tx_len > 1) ? tx[1] : 0;
+	}
+
+	packet->size = sizeof(packet->header) + packet->payload_length;
+
+	return 0;
+}
+EXPORT_SYMBOL(mipi_dsi_create_packet);
+
+/*
+ * mipi_dsi_set_maximum_return_packet_size() - specify the maximum size of the
+ *    the payload in a long packet transmitted from the peripheral back to the
+ *    host processor
+ * @dsi: DSI peripheral device
+ * @value: the maximum size of the payload
+ *
+ * Return: 0 on success or a negative error code on failure.
+ */
+int mipi_dsi_set_maximum_return_packet_size(struct mipi_dsi_device *dsi,
+					    u16 value)
+{
+	u8 tx[2] = { value & 0xff, value >> 8 };
+	struct mipi_dsi_msg msg = {
+		.channel = dsi->channel,
+		.type = MIPI_DSI_SET_MAXIMUM_RETURN_PACKET_SIZE,
+		.tx_len = sizeof(tx),
+		.tx_buf = tx,
+	};
+
+	return mipi_dsi_device_transfer(dsi, &msg);
+}
+EXPORT_SYMBOL(mipi_dsi_set_maximum_return_packet_size);
+
+/**
+ * mipi_dsi_generic_write() - transmit data using a generic write packet
+ * @dsi: DSI peripheral device
+ * @payload: buffer containing the payload
+ * @size: size of payload buffer
+ *
+ * This function will automatically choose the right data type depending on
+ * the payload length.
+ *
+ * Return: The number of bytes transmitted on success or a negative error code
+ * on failure.
+ */
+ssize_t mipi_dsi_generic_write(struct mipi_dsi_device *dsi, const void *payload,
+			       size_t size)
+{
+	struct mipi_dsi_msg msg = {
+		.channel = dsi->channel,
+		.tx_buf = payload,
+		.tx_len = size
+	};
+
+	switch (size) {
+	case 0:
+		msg.type = MIPI_DSI_GENERIC_SHORT_WRITE_0_PARAM;
+		break;
+
+	case 1:
+		msg.type = MIPI_DSI_GENERIC_SHORT_WRITE_1_PARAM;
+		break;
+
+	case 2:
+		msg.type = MIPI_DSI_GENERIC_SHORT_WRITE_2_PARAM;
+		break;
+
+	default:
+		msg.type = MIPI_DSI_GENERIC_LONG_WRITE;
+		break;
+	}
+
+	return mipi_dsi_device_transfer(dsi, &msg);
+}
+EXPORT_SYMBOL(mipi_dsi_generic_write);
+
+/**
+ * mipi_dsi_generic_read() - receive data using a generic read packet
+ * @dsi: DSI peripheral device
+ * @params: buffer containing the request parameters
+ * @num_params: number of request parameters
+ * @data: buffer in which to return the received data
+ * @size: size of receive buffer
+ *
+ * This function will automatically choose the right data type depending on
+ * the number of parameters passed in.
+ *
+ * Return: The number of bytes successfully read or a negative error code on
+ * failure.
+ */
+ssize_t mipi_dsi_generic_read(struct mipi_dsi_device *dsi, const void *params,
+			      size_t num_params, void *data, size_t size)
+{
+	struct mipi_dsi_msg msg = {
+		.channel = dsi->channel,
+		.tx_len = num_params,
+		.tx_buf = params,
+		.rx_len = size,
+		.rx_buf = data
+	};
+
+	switch (num_params) {
+	case 0:
+		msg.type = MIPI_DSI_GENERIC_READ_REQUEST_0_PARAM;
+		break;
+
+	case 1:
+		msg.type = MIPI_DSI_GENERIC_READ_REQUEST_1_PARAM;
+		break;
+
+	case 2:
+		msg.type = MIPI_DSI_GENERIC_READ_REQUEST_2_PARAM;
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	return mipi_dsi_device_transfer(dsi, &msg);
+}
+EXPORT_SYMBOL(mipi_dsi_generic_read);
+
+/**
+ * mipi_dsi_dcs_write_buffer() - transmit a DCS command with payload
+ * @dsi: DSI peripheral device
+ * @data: buffer containing data to be transmitted
+ * @len: size of transmission buffer
+ *
+ * This function will automatically choose the right data type depending on
+ * the command payload length.
+ *
+ * Return: The number of bytes successfully transmitted or a negative error
+ * code on failure.
+ */
+ssize_t mipi_dsi_dcs_write_buffer(struct mipi_dsi_device *dsi,
+				  const void *data, size_t len)
 {
-	const struct mipi_dsi_host_ops *ops = dsi->host->ops;
 	struct mipi_dsi_msg msg = {
 		.channel = dsi->channel,
 		.tx_buf = data,
 		.tx_len = len
 	};
 
-	if (!ops || !ops->transfer)
-		return -ENOSYS;
-
 	switch (len) {
 	case 0:
 		return -EINVAL;
+
 	case 1:
 		msg.type = MIPI_DSI_DCS_SHORT_WRITE;
 		break;
+
 	case 2:
 		msg.type = MIPI_DSI_DCS_SHORT_WRITE_PARAM;
 		break;
+
 	default:
 		msg.type = MIPI_DSI_DCS_LONG_WRITE;
 		break;
 	}
 
-	return ops->transfer(dsi->host, &msg);
+	return mipi_dsi_device_transfer(dsi, &msg);
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_write_buffer);
+
+/**
+ * mipi_dsi_dcs_write() - send DCS write command
+ * @dsi: DSI peripheral device
+ * @cmd: DCS command
+ * @data: buffer containing the command payload
+ * @len: command payload length
+ *
+ * This function will automatically choose the right data type depending on
+ * the command payload length.
+ *
+ * Return: The number of bytes successfully transmitted or a negative error
+ * code on failure.
+ */
+ssize_t mipi_dsi_dcs_write(struct mipi_dsi_device *dsi, u8 cmd,
+			   const void *data, size_t len)
+{
+	ssize_t err;
+	size_t size;
+	u8 *tx;
+
+	if (len > 0) {
+		size = 1 + len;
+
+		tx = kmalloc(size, GFP_KERNEL);
+		if (!tx)
+			return -ENOMEM;
+
+		/* concatenate the DCS command byte and the payload */
+		tx[0] = cmd;
+		memcpy(&tx[1], data, len);
+	} else {
+		tx = &cmd;
+		size = 1;
+	}
+
+	err = mipi_dsi_dcs_write_buffer(dsi, tx, size);
+
+	if (len > 0)
+		kfree(tx);
+
+	return err;
 }
 EXPORT_SYMBOL(mipi_dsi_dcs_write);
 
 /**
- * mipi_dsi_dcs_read - send DCS read request command
- * @dsi: DSI device
- * @cmd: DCS read command
- * @data: pointer to read buffer
- * @len: length of @data
+ * mipi_dsi_dcs_read() - send DCS read request command
+ * @dsi: DSI peripheral device
+ * @cmd: DCS command
+ * @data: buffer in which to receive data
+ * @len: size of receive buffer
  *
- * Function returns number of read bytes or error code.
+ * Return: The number of bytes read or a negative error code on failure.
  */
 ssize_t mipi_dsi_dcs_read(struct mipi_dsi_device *dsi, u8 cmd, void *data,
 			  size_t len)
 {
-	const struct mipi_dsi_host_ops *ops = dsi->host->ops;
 	struct mipi_dsi_msg msg = {
 		.channel = dsi->channel,
 		.type = MIPI_DSI_DCS_READ,
@@ -257,13 +585,283 @@ ssize_t mipi_dsi_dcs_read(struct mipi_dsi_device *dsi, u8 cmd, void *data,
 		.rx_len = len
 	};
 
-	if (!ops || !ops->transfer)
-		return -ENOSYS;
-
-	return ops->transfer(dsi->host, &msg);
+	return mipi_dsi_device_transfer(dsi, &msg);
 }
 EXPORT_SYMBOL(mipi_dsi_dcs_read);
 
+/**
+ * mipi_dsi_dcs_nop() - send DCS nop packet
+ * @dsi: DSI peripheral device
+ *
+ * Return: 0 on success or a negative error code on failure.
+ */
+int mipi_dsi_dcs_nop(struct mipi_dsi_device *dsi)
+{
+	ssize_t err;
+
+	err = mipi_dsi_dcs_write(dsi, MIPI_DCS_NOP, NULL, 0);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_nop);
+
+/**
+ * mipi_dsi_dcs_soft_reset() - perform a software reset of the display module
+ * @dsi: DSI peripheral device
+ *
+ * Return: 0 on success or a negative error code on failure.
+ */
+int mipi_dsi_dcs_soft_reset(struct mipi_dsi_device *dsi)
+{
+	ssize_t err;
+
+	err = mipi_dsi_dcs_write(dsi, MIPI_DCS_SOFT_RESET, NULL, 0);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_soft_reset);
+
+/**
+ * mipi_dsi_dcs_get_power_mode() - query the display module's current power
+ *    mode
+ * @dsi: DSI peripheral device
+ * @mode: return location for the current power mode
+ *
+ * Return: 0 on success or a negative error code on failure.
+ */
+int mipi_dsi_dcs_get_power_mode(struct mipi_dsi_device *dsi, u8 *mode)
+{
+	ssize_t err;
+
+	err = mipi_dsi_dcs_read(dsi, MIPI_DCS_GET_POWER_MODE, mode,
+				sizeof(*mode));
+	if (err <= 0) {
+		if (err == 0)
+			err = -ENODATA;
+
+		return err;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_get_power_mode);
+
+/**
+ * mipi_dsi_dcs_get_pixel_format() - gets the pixel format for the RGB image
+ *    data used by the interface
+ * @dsi: DSI peripheral device
+ * @format: return location for the pixel format
+ *
+ * Return: 0 on success or a negative error code on failure.
+ */
+int mipi_dsi_dcs_get_pixel_format(struct mipi_dsi_device *dsi, u8 *format)
+{
+	ssize_t err;
+
+	err = mipi_dsi_dcs_read(dsi, MIPI_DCS_GET_PIXEL_FORMAT, format,
+				sizeof(*format));
+	if (err <= 0) {
+		if (err == 0)
+			err = -ENODATA;
+
+		return err;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_get_pixel_format);
+
+/**
+ * mipi_dsi_dcs_enter_sleep_mode() - disable all unnecessary blocks inside the
+ *    display module except interface communication
+ * @dsi: DSI peripheral device
+ *
+ * Return: 0 on success or a negative error code on failure.
+ */
+int mipi_dsi_dcs_enter_sleep_mode(struct mipi_dsi_device *dsi)
+{
+	ssize_t err;
+
+	err = mipi_dsi_dcs_write(dsi, MIPI_DCS_ENTER_SLEEP_MODE, NULL, 0);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_enter_sleep_mode);
+
+/**
+ * mipi_dsi_dcs_exit_sleep_mode() - enable all blocks inside the display
+ *    module
+ * @dsi: DSI peripheral device
+ *
+ * Return: 0 on success or a negative error code on failure.
+ */
+int mipi_dsi_dcs_exit_sleep_mode(struct mipi_dsi_device *dsi)
+{
+	ssize_t err;
+
+	err = mipi_dsi_dcs_write(dsi, MIPI_DCS_EXIT_SLEEP_MODE, NULL, 0);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_exit_sleep_mode);
+
+/**
+ * mipi_dsi_dcs_set_display_off() - stop displaying the image data on the
+ *    display device
+ * @dsi: DSI peripheral device
+ *
+ * Return: 0 on success or a negative error code on failure.
+ */
+int mipi_dsi_dcs_set_display_off(struct mipi_dsi_device *dsi)
+{
+	ssize_t err;
+
+	err = mipi_dsi_dcs_write(dsi, MIPI_DCS_SET_DISPLAY_OFF, NULL, 0);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_set_display_off);
+
+/**
+ * mipi_dsi_dcs_set_display_on() - start displaying the image data on the
+ *    display device
+ * @dsi: DSI peripheral device
+ *
+ * Return: 0 on success or a negative error code on failure
+ */
+int mipi_dsi_dcs_set_display_on(struct mipi_dsi_device *dsi)
+{
+	ssize_t err;
+
+	err = mipi_dsi_dcs_write(dsi, MIPI_DCS_SET_DISPLAY_ON, NULL, 0);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_set_display_on);
+
+/**
+ * mipi_dsi_dcs_set_column_address() - define the column extent of the frame
+ *    memory accessed by the host processor
+ * @dsi: DSI peripheral device
+ * @start: first column of frame memory
+ * @end: last column of frame memory
+ *
+ * Return: 0 on success or a negative error code on failure.
+ */
+int mipi_dsi_dcs_set_column_address(struct mipi_dsi_device *dsi, u16 start,
+				    u16 end)
+{
+	u8 payload[4] = { start >> 8, start & 0xff, end >> 8, end & 0xff };
+	ssize_t err;
+
+	err = mipi_dsi_dcs_write(dsi, MIPI_DCS_SET_COLUMN_ADDRESS, payload,
+				 sizeof(payload));
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_set_column_address);
+
+/**
+ * mipi_dsi_dcs_set_page_address() - define the page extent of the frame
+ *    memory accessed by the host processor
+ * @dsi: DSI peripheral device
+ * @start: first page of frame memory
+ * @end: last page of frame memory
+ *
+ * Return: 0 on success or a negative error code on failure.
+ */
+int mipi_dsi_dcs_set_page_address(struct mipi_dsi_device *dsi, u16 start,
+				  u16 end)
+{
+	u8 payload[4] = { start >> 8, start & 0xff, end >> 8, end & 0xff };
+	ssize_t err;
+
+	err = mipi_dsi_dcs_write(dsi, MIPI_DCS_SET_PAGE_ADDRESS, payload,
+				 sizeof(payload));
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_set_page_address);
+
+/**
+ * mipi_dsi_dcs_set_tear_off() - turn off the display module's Tearing Effect
+ *    output signal on the TE signal line
+ * @dsi: DSI peripheral device
+ *
+ * Return: 0 on success or a negative error code on failure
+ */
+int mipi_dsi_dcs_set_tear_off(struct mipi_dsi_device *dsi)
+{
+	ssize_t err;
+
+	err = mipi_dsi_dcs_write(dsi, MIPI_DCS_SET_TEAR_OFF, NULL, 0);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_set_tear_off);
+
+/**
+ * mipi_dsi_dcs_set_tear_on() - turn on the display module's Tearing Effect
+ *    output signal on the TE signal line.
+ * @dsi: DSI peripheral device
+ * @mode: the Tearing Effect Output Line mode
+ *
+ * Return: 0 on success or a negative error code on failure
+ */
+int mipi_dsi_dcs_set_tear_on(struct mipi_dsi_device *dsi,
+			     enum mipi_dsi_dcs_tear_mode mode)
+{
+	u8 value = mode;
+	ssize_t err;
+
+	err = mipi_dsi_dcs_write(dsi, MIPI_DCS_SET_TEAR_ON, &value,
+				 sizeof(value));
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_set_tear_on);
+
+/**
+ * mipi_dsi_dcs_set_pixel_format() - sets the pixel format for the RGB image
+ *    data used by the interface
+ * @dsi: DSI peripheral device
+ * @format: pixel format
+ *
+ * Return: 0 on success or a negative error code on failure.
+ */
+int mipi_dsi_dcs_set_pixel_format(struct mipi_dsi_device *dsi, u8 format)
+{
+	ssize_t err;
+
+	err = mipi_dsi_dcs_write(dsi, MIPI_DCS_SET_PIXEL_FORMAT, &format,
+				 sizeof(format));
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_set_pixel_format);
+
 static int mipi_dsi_drv_probe(struct device *dev)
 {
 	struct mipi_dsi_driver *drv = to_mipi_dsi_driver(dev->driver);
@@ -289,12 +887,18 @@ static void mipi_dsi_drv_shutdown(struct device *dev)
 }
 
 /**
- * mipi_dsi_driver_register - register a driver for DSI devices
+ * mipi_dsi_driver_register_full() - register a driver for DSI devices
  * @drv: DSI driver structure
+ * @owner: owner module
+ *
+ * Return: 0 on success or a negative error code on failure.
  */
-int mipi_dsi_driver_register(struct mipi_dsi_driver *drv)
+int mipi_dsi_driver_register_full(struct mipi_dsi_driver *drv,
+				  struct module *owner)
 {
 	drv->driver.bus = &mipi_dsi_bus_type;
+	drv->driver.owner = owner;
+
 	if (drv->probe)
 		drv->driver.probe = mipi_dsi_drv_probe;
 	if (drv->remove)
@@ -304,11 +908,13 @@ int mipi_dsi_driver_register(struct mipi_dsi_driver *drv)
 
 	return driver_register(&drv->driver);
 }
-EXPORT_SYMBOL(mipi_dsi_driver_register);
+EXPORT_SYMBOL(mipi_dsi_driver_register_full);
 
 /**
- * mipi_dsi_driver_unregister - unregister a driver for DSI devices
+ * mipi_dsi_driver_unregister() - unregister a driver for DSI devices
  * @drv: DSI driver structure
+ *
+ * Return: 0 on success or a negative error code on failure.
  */
 void mipi_dsi_driver_unregister(struct mipi_dsi_driver *drv)
 {
diff --git a/drivers/gpu/drm/drm_modes.c b/drivers/gpu/drm/drm_modes.c
index bedf1894e17e..6d8b941c8200 100644
--- a/drivers/gpu/drm/drm_modes.c
+++ b/drivers/gpu/drm/drm_modes.c
@@ -914,7 +914,7 @@ EXPORT_SYMBOL(drm_mode_equal_no_clocks_no_stereo);
  *
  * This function is a helper which can be used to validate modes against size
  * limitations of the DRM device/connector. If a mode is too big its status
- * memeber is updated with the appropriate validation failure code. The list
+ * member is updated with the appropriate validation failure code. The list
  * itself is not changed.
  */
 void drm_mode_validate_size(struct drm_device *dev,
@@ -1259,6 +1259,7 @@ drm_mode_create_from_cmdline_mode(struct drm_device *dev,
 	if (!mode)
 		return NULL;
 
+	mode->type |= DRM_MODE_TYPE_USERDEF;
 	drm_mode_set_crtcinfo(mode, CRTC_INTERLACE_HALVE_V);
 	return mode;
 }
diff --git a/drivers/gpu/drm/drm_modeset_lock.c b/drivers/gpu/drm/drm_modeset_lock.c
index 0dc57d5ecd10..51cc47d827d8 100644
--- a/drivers/gpu/drm/drm_modeset_lock.c
+++ b/drivers/gpu/drm/drm_modeset_lock.c
@@ -35,7 +35,7 @@
  * of extra utility/tracking out of our acquire-ctx.  This is provided
  * by drm_modeset_lock / drm_modeset_acquire_ctx.
  *
- * For basic principles of ww_mutex, see: Documentation/ww-mutex-design.txt
+ * For basic principles of ww_mutex, see: Documentation/locking/ww-mutex-design.txt
  *
  * The basic usage pattern is to:
  *
@@ -57,6 +57,230 @@
 
 
 /**
+ * __drm_modeset_lock_all - internal helper to grab all modeset locks
+ * @dev: DRM device
+ * @trylock: trylock mode for atomic contexts
+ *
+ * This is a special version of drm_modeset_lock_all() which can also be used in
+ * atomic contexts. Then @trylock must be set to true.
+ *
+ * Returns:
+ * 0 on success or negative error code on failure.
+ */
+int __drm_modeset_lock_all(struct drm_device *dev,
+			   bool trylock)
+{
+	struct drm_mode_config *config = &dev->mode_config;
+	struct drm_modeset_acquire_ctx *ctx;
+	int ret;
+
+	ctx = kzalloc(sizeof(*ctx),
+		      trylock ? GFP_ATOMIC : GFP_KERNEL);
+	if (!ctx)
+		return -ENOMEM;
+
+	if (trylock) {
+		if (!mutex_trylock(&config->mutex))
+			return -EBUSY;
+	} else {
+		mutex_lock(&config->mutex);
+	}
+
+	drm_modeset_acquire_init(ctx, 0);
+	ctx->trylock_only = trylock;
+
+retry:
+	ret = drm_modeset_lock(&config->connection_mutex, ctx);
+	if (ret)
+		goto fail;
+	ret = drm_modeset_lock_all_crtcs(dev, ctx);
+	if (ret)
+		goto fail;
+
+	WARN_ON(config->acquire_ctx);
+
+	/* now we hold the locks, so now that it is safe, stash the
+	 * ctx for drm_modeset_unlock_all():
+	 */
+	config->acquire_ctx = ctx;
+
+	drm_warn_on_modeset_not_all_locked(dev);
+
+	return 0;
+
+fail:
+	if (ret == -EDEADLK) {
+		drm_modeset_backoff(ctx);
+		goto retry;
+	}
+
+	return ret;
+}
+EXPORT_SYMBOL(__drm_modeset_lock_all);
+
+/**
+ * drm_modeset_lock_all - take all modeset locks
+ * @dev: drm device
+ *
+ * This function takes all modeset locks, suitable where a more fine-grained
+ * scheme isn't (yet) implemented. Locks must be dropped with
+ * drm_modeset_unlock_all.
+ */
+void drm_modeset_lock_all(struct drm_device *dev)
+{
+	WARN_ON(__drm_modeset_lock_all(dev, false) != 0);
+}
+EXPORT_SYMBOL(drm_modeset_lock_all);
+
+/**
+ * drm_modeset_unlock_all - drop all modeset locks
+ * @dev: device
+ *
+ * This function drop all modeset locks taken by drm_modeset_lock_all.
+ */
+void drm_modeset_unlock_all(struct drm_device *dev)
+{
+	struct drm_mode_config *config = &dev->mode_config;
+	struct drm_modeset_acquire_ctx *ctx = config->acquire_ctx;
+
+	if (WARN_ON(!ctx))
+		return;
+
+	config->acquire_ctx = NULL;
+	drm_modeset_drop_locks(ctx);
+	drm_modeset_acquire_fini(ctx);
+
+	kfree(ctx);
+
+	mutex_unlock(&dev->mode_config.mutex);
+}
+EXPORT_SYMBOL(drm_modeset_unlock_all);
+
+/**
+ * drm_modeset_lock_crtc - lock crtc with hidden acquire ctx for a plane update
+ * @crtc: DRM CRTC
+ * @plane: DRM plane to be updated on @crtc
+ *
+ * This function locks the given crtc and plane (which should be either the
+ * primary or cursor plane) using a hidden acquire context. This is necessary so
+ * that drivers internally using the atomic interfaces can grab further locks
+ * with the lock acquire context.
+ *
+ * Note that @plane can be NULL, e.g. when the cursor support hasn't yet been
+ * converted to universal planes yet.
+ */
+void drm_modeset_lock_crtc(struct drm_crtc *crtc,
+			   struct drm_plane *plane)
+{
+	struct drm_modeset_acquire_ctx *ctx;
+	int ret;
+
+	ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
+	if (WARN_ON(!ctx))
+		return;
+
+	drm_modeset_acquire_init(ctx, 0);
+
+retry:
+	ret = drm_modeset_lock(&crtc->mutex, ctx);
+	if (ret)
+		goto fail;
+
+	if (plane) {
+		ret = drm_modeset_lock(&plane->mutex, ctx);
+		if (ret)
+			goto fail;
+
+		if (plane->crtc) {
+			ret = drm_modeset_lock(&plane->crtc->mutex, ctx);
+			if (ret)
+				goto fail;
+		}
+	}
+
+	WARN_ON(crtc->acquire_ctx);
+
+	/* now we hold the locks, so now that it is safe, stash the
+	 * ctx for drm_modeset_unlock_crtc():
+	 */
+	crtc->acquire_ctx = ctx;
+
+	return;
+
+fail:
+	if (ret == -EDEADLK) {
+		drm_modeset_backoff(ctx);
+		goto retry;
+	}
+}
+EXPORT_SYMBOL(drm_modeset_lock_crtc);
+
+/**
+ * drm_modeset_legacy_acquire_ctx - find acquire ctx for legacy ioctls
+ * @crtc: drm crtc
+ *
+ * Legacy ioctl operations like cursor updates or page flips only have per-crtc
+ * locking, and store the acquire ctx in the corresponding crtc. All other
+ * legacy operations take all locks and use a global acquire context. This
+ * function grabs the right one.
+ */
+struct drm_modeset_acquire_ctx *
+drm_modeset_legacy_acquire_ctx(struct drm_crtc *crtc)
+{
+	if (crtc->acquire_ctx)
+		return crtc->acquire_ctx;
+
+	WARN_ON(!crtc->dev->mode_config.acquire_ctx);
+
+	return crtc->dev->mode_config.acquire_ctx;
+}
+EXPORT_SYMBOL(drm_modeset_legacy_acquire_ctx);
+
+/**
+ * drm_modeset_unlock_crtc - drop crtc lock
+ * @crtc: drm crtc
+ *
+ * This drops the crtc lock acquire with drm_modeset_lock_crtc() and all other
+ * locks acquired through the hidden context.
+ */
+void drm_modeset_unlock_crtc(struct drm_crtc *crtc)
+{
+	struct drm_modeset_acquire_ctx *ctx = crtc->acquire_ctx;
+
+	if (WARN_ON(!ctx))
+		return;
+
+	crtc->acquire_ctx = NULL;
+	drm_modeset_drop_locks(ctx);
+	drm_modeset_acquire_fini(ctx);
+
+	kfree(ctx);
+}
+EXPORT_SYMBOL(drm_modeset_unlock_crtc);
+
+/**
+ * drm_warn_on_modeset_not_all_locked - check that all modeset locks are locked
+ * @dev: device
+ *
+ * Useful as a debug assert.
+ */
+void drm_warn_on_modeset_not_all_locked(struct drm_device *dev)
+{
+	struct drm_crtc *crtc;
+
+	/* Locking is currently fubar in the panic handler. */
+	if (oops_in_progress)
+		return;
+
+	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
+		WARN_ON(!drm_modeset_is_locked(&crtc->mutex));
+
+	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
+	WARN_ON(!mutex_is_locked(&dev->mode_config.mutex));
+}
+EXPORT_SYMBOL(drm_warn_on_modeset_not_all_locked);
+
+/**
  * drm_modeset_acquire_init - initialize acquire context
  * @ctx: the acquire context
  * @flags: for future
@@ -108,7 +332,12 @@ static inline int modeset_lock(struct drm_modeset_lock *lock,
 
 	WARN_ON(ctx->contended);
 
-	if (interruptible && slow) {
+	if (ctx->trylock_only) {
+		if (!ww_mutex_trylock(&lock->mutex))
+			return -EBUSY;
+		else
+			return 0;
+	} else if (interruptible && slow) {
 		ret = ww_mutex_lock_slow_interruptible(&lock->mutex, &ctx->ww_ctx);
 	} else if (interruptible) {
 		ret = ww_mutex_lock_interruptible(&lock->mutex, &ctx->ww_ctx);
@@ -226,15 +455,14 @@ void drm_modeset_unlock(struct drm_modeset_lock *lock)
 }
 EXPORT_SYMBOL(drm_modeset_unlock);
 
-/* Temporary.. until we have sufficiently fine grained locking, there
- * are a couple scenarios where it is convenient to grab all crtc locks.
- * It is planned to remove this:
- */
+/* In some legacy codepaths it's convenient to just grab all the crtc and plane
+ * related locks. */
 int drm_modeset_lock_all_crtcs(struct drm_device *dev,
 		struct drm_modeset_acquire_ctx *ctx)
 {
 	struct drm_mode_config *config = &dev->mode_config;
 	struct drm_crtc *crtc;
+	struct drm_plane *plane;
 	int ret = 0;
 
 	list_for_each_entry(crtc, &config->crtc_list, head) {
@@ -243,6 +471,12 @@ int drm_modeset_lock_all_crtcs(struct drm_device *dev,
 			return ret;
 	}
 
+	list_for_each_entry(plane, &config->plane_list, head) {
+		ret = drm_modeset_lock(&plane->mutex, ctx);
+		if (ret)
+			return ret;
+	}
+
 	return 0;
 }
 EXPORT_SYMBOL(drm_modeset_lock_all_crtcs);
diff --git a/drivers/gpu/drm/drm_pci.c b/drivers/gpu/drm/drm_pci.c
index 020cfd934854..fd29f03645b8 100644
--- a/drivers/gpu/drm/drm_pci.c
+++ b/drivers/gpu/drm/drm_pci.c
@@ -27,6 +27,7 @@
 #include <linux/dma-mapping.h>
 #include <linux/export.h>
 #include <drm/drmP.h>
+#include "drm_legacy.h"
 
 /**
  * drm_pci_alloc - Allocate a PCI consistent memory block, for DMA.
@@ -81,7 +82,7 @@ EXPORT_SYMBOL(drm_pci_alloc);
  *
  * This function is for internal use in the Linux-specific DRM core code.
  */
-void __drm_pci_free(struct drm_device * dev, drm_dma_handle_t * dmah)
+void __drm_legacy_pci_free(struct drm_device * dev, drm_dma_handle_t * dmah)
 {
 	unsigned long addr;
 	size_t sz;
@@ -105,7 +106,7 @@ void __drm_pci_free(struct drm_device * dev, drm_dma_handle_t * dmah)
  */
 void drm_pci_free(struct drm_device * dev, drm_dma_handle_t * dmah)
 {
-	__drm_pci_free(dev, dmah);
+	__drm_legacy_pci_free(dev, dmah);
 	kfree(dmah);
 }
 
@@ -127,34 +128,20 @@ static int drm_get_pci_domain(struct drm_device *dev)
 	return pci_domain_nr(dev->pdev->bus);
 }
 
-static int drm_pci_set_busid(struct drm_device *dev, struct drm_master *master)
+int drm_pci_set_busid(struct drm_device *dev, struct drm_master *master)
 {
-	int len, ret;
-	master->unique_len = 40;
-	master->unique_size = master->unique_len;
-	master->unique = kmalloc(master->unique_size, GFP_KERNEL);
-	if (master->unique == NULL)
+	master->unique = kasprintf(GFP_KERNEL, "pci:%04x:%02x:%02x.%d",
+					drm_get_pci_domain(dev),
+					dev->pdev->bus->number,
+					PCI_SLOT(dev->pdev->devfn),
+					PCI_FUNC(dev->pdev->devfn));
+	if (!master->unique)
 		return -ENOMEM;
 
-
-	len = snprintf(master->unique, master->unique_len,
-		       "pci:%04x:%02x:%02x.%d",
-		       drm_get_pci_domain(dev),
-		       dev->pdev->bus->number,
-		       PCI_SLOT(dev->pdev->devfn),
-		       PCI_FUNC(dev->pdev->devfn));
-
-	if (len >= master->unique_len) {
-		DRM_ERROR("buffer overflow");
-		ret = -EINVAL;
-		goto err;
-	} else
-		master->unique_len = len;
-
+	master->unique_len = strlen(master->unique);
 	return 0;
-err:
-	return ret;
 }
+EXPORT_SYMBOL(drm_pci_set_busid);
 
 int drm_pci_set_unique(struct drm_device *dev,
 		       struct drm_master *master,
@@ -163,8 +150,7 @@ int drm_pci_set_unique(struct drm_device *dev,
 	int domain, bus, slot, func, ret;
 
 	master->unique_len = u->unique_len;
-	master->unique_size = u->unique_len + 1;
-	master->unique = kmalloc(master->unique_size, GFP_KERNEL);
+	master->unique = kmalloc(master->unique_len + 1, GFP_KERNEL);
 	if (!master->unique) {
 		ret = -ENOMEM;
 		goto err;
@@ -269,10 +255,6 @@ void drm_pci_agp_destroy(struct drm_device *dev)
 	}
 }
 
-static struct drm_bus drm_pci_bus = {
-	.set_busid = drm_pci_set_busid,
-};
-
 /**
  * drm_get_pci_dev - Register a PCI device with the DRM subsystem
  * @pdev: PCI device
@@ -353,8 +335,6 @@ int drm_pci_init(struct drm_driver *driver, struct pci_driver *pdriver)
 
 	DRM_DEBUG("\n");
 
-	driver->bus = &drm_pci_bus;
-
 	if (driver->driver_features & DRIVER_MODESET)
 		return pci_register_driver(pdriver);
 
diff --git a/drivers/gpu/drm/drm_plane_helper.c b/drivers/gpu/drm/drm_plane_helper.c
index 827ec1a3040b..18a1ac6ac22f 100644
--- a/drivers/gpu/drm/drm_plane_helper.c
+++ b/drivers/gpu/drm/drm_plane_helper.c
@@ -27,10 +27,38 @@
 #include <drm/drmP.h>
 #include <drm/drm_plane_helper.h>
 #include <drm/drm_rect.h>
-#include <drm/drm_plane_helper.h>
+#include <drm/drm_atomic.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_atomic_helper.h>
 
 #define SUBPIXEL_MASK 0xffff
 
+/**
+ * DOC: overview
+ *
+ * This helper library has two parts. The first part has support to implement
+ * primary plane support on top of the normal CRTC configuration interface.
+ * Since the legacy ->set_config interface ties the primary plane together with
+ * the CRTC state this does not allow userspace to disable the primary plane
+ * itself.  To avoid too much duplicated code use
+ * drm_plane_helper_check_update() which can be used to enforce the same
+ * restrictions as primary planes had thus. The default primary plane only
+ * expose XRBG8888 and ARGB8888 as valid pixel formats for the attached
+ * framebuffer.
+ *
+ * Drivers are highly recommended to implement proper support for primary
+ * planes, and newly merged drivers must not rely upon these transitional
+ * helpers.
+ *
+ * The second part also implements transitional helpers which allow drivers to
+ * gradually switch to the atomic helper infrastructure for plane updates. Once
+ * that switch is complete drivers shouldn't use these any longer, instead using
+ * the proper legacy implementations for update and disable plane hooks provided
+ * by the atomic helpers.
+ *
+ * Again drivers are strongly urged to switch to the new interfaces.
+ */
+
 /*
  * This is the minimal list of formats that seem to be safe for modeset use
  * with all current DRM drivers.  Most hardware can actually support more
@@ -127,6 +155,11 @@ int drm_plane_helper_check_update(struct drm_plane *plane,
 		return -ERANGE;
 	}
 
+	if (!fb) {
+		*visible = false;
+		return 0;
+	}
+
 	*visible = drm_rect_clip_scaled(src, dest, clip, hscale, vscale);
 	if (!*visible)
 		/*
@@ -369,3 +402,171 @@ int drm_crtc_init(struct drm_device *dev, struct drm_crtc *crtc,
 	return drm_crtc_init_with_planes(dev, crtc, primary, NULL, funcs);
 }
 EXPORT_SYMBOL(drm_crtc_init);
+
+int drm_plane_helper_commit(struct drm_plane *plane,
+			    struct drm_plane_state *plane_state,
+			    struct drm_framebuffer *old_fb)
+{
+	struct drm_plane_helper_funcs *plane_funcs;
+	struct drm_crtc *crtc[2];
+	struct drm_crtc_helper_funcs *crtc_funcs[2];
+	int i, ret = 0;
+
+	plane_funcs = plane->helper_private;
+
+	/* Since this is a transitional helper we can't assume that plane->state
+	 * is always valid. Hence we need to use plane->crtc instead of
+	 * plane->state->crtc as the old crtc. */
+	crtc[0] = plane->crtc;
+	crtc[1] = crtc[0] != plane_state->crtc ? plane_state->crtc : NULL;
+
+	for (i = 0; i < 2; i++)
+		crtc_funcs[i] = crtc[i] ? crtc[i]->helper_private : NULL;
+
+	if (plane_funcs->atomic_check) {
+		ret = plane_funcs->atomic_check(plane, plane_state);
+		if (ret)
+			goto out;
+	}
+
+	if (plane_funcs->prepare_fb && plane_state->fb) {
+		ret = plane_funcs->prepare_fb(plane, plane_state->fb);
+		if (ret)
+			goto out;
+	}
+
+	/* Point of no return, commit sw state. */
+	swap(plane->state, plane_state);
+
+	for (i = 0; i < 2; i++) {
+		if (crtc_funcs[i] && crtc_funcs[i]->atomic_begin)
+			crtc_funcs[i]->atomic_begin(crtc[i]);
+	}
+
+	plane_funcs->atomic_update(plane, plane_state);
+
+	for (i = 0; i < 2; i++) {
+		if (crtc_funcs[i] && crtc_funcs[i]->atomic_flush)
+			crtc_funcs[i]->atomic_flush(crtc[i]);
+	}
+
+	for (i = 0; i < 2; i++) {
+		if (!crtc[i])
+			continue;
+
+		/* There's no other way to figure out whether the crtc is running. */
+		ret = drm_crtc_vblank_get(crtc[i]);
+		if (ret == 0) {
+			drm_crtc_wait_one_vblank(crtc[i]);
+			drm_crtc_vblank_put(crtc[i]);
+		}
+
+		ret = 0;
+	}
+
+	if (plane_funcs->cleanup_fb && old_fb)
+		plane_funcs->cleanup_fb(plane, old_fb);
+out:
+	if (plane_state) {
+		if (plane->funcs->atomic_destroy_state)
+			plane->funcs->atomic_destroy_state(plane, plane_state);
+		else
+			drm_atomic_helper_plane_destroy_state(plane, plane_state);
+	}
+
+	return ret;
+}
+
+/**
+ * drm_plane_helper_update() - Helper for primary plane update
+ * @plane: plane object to update
+ * @crtc: owning CRTC of owning plane
+ * @fb: framebuffer to flip onto plane
+ * @crtc_x: x offset of primary plane on crtc
+ * @crtc_y: y offset of primary plane on crtc
+ * @crtc_w: width of primary plane rectangle on crtc
+ * @crtc_h: height of primary plane rectangle on crtc
+ * @src_x: x offset of @fb for panning
+ * @src_y: y offset of @fb for panning
+ * @src_w: width of source rectangle in @fb
+ * @src_h: height of source rectangle in @fb
+ *
+ * Provides a default plane update handler using the atomic plane update
+ * functions. It is fully left to the driver to check plane constraints and
+ * handle corner-cases like a fully occluded or otherwise invisible plane.
+ *
+ * This is useful for piecewise transitioning of a driver to the atomic helpers.
+ *
+ * RETURNS:
+ * Zero on success, error code on failure
+ */
+int drm_plane_helper_update(struct drm_plane *plane, struct drm_crtc *crtc,
+			    struct drm_framebuffer *fb,
+			    int crtc_x, int crtc_y,
+			    unsigned int crtc_w, unsigned int crtc_h,
+			    uint32_t src_x, uint32_t src_y,
+			    uint32_t src_w, uint32_t src_h)
+{
+	struct drm_plane_state *plane_state;
+
+	if (plane->funcs->atomic_duplicate_state)
+		plane_state = plane->funcs->atomic_duplicate_state(plane);
+	else if (plane->state)
+		plane_state = drm_atomic_helper_plane_duplicate_state(plane);
+	else
+		plane_state = kzalloc(sizeof(*plane_state), GFP_KERNEL);
+	if (!plane_state)
+		return -ENOMEM;
+
+	plane_state->crtc = crtc;
+	drm_atomic_set_fb_for_plane(plane_state, fb);
+	plane_state->crtc_x = crtc_x;
+	plane_state->crtc_y = crtc_y;
+	plane_state->crtc_h = crtc_h;
+	plane_state->crtc_w = crtc_w;
+	plane_state->src_x = src_x;
+	plane_state->src_y = src_y;
+	plane_state->src_h = src_h;
+	plane_state->src_w = src_w;
+
+	return drm_plane_helper_commit(plane, plane_state, plane->fb);
+}
+EXPORT_SYMBOL(drm_plane_helper_update);
+
+/**
+ * drm_plane_helper_disable() - Helper for primary plane disable
+ * @plane: plane to disable
+ *
+ * Provides a default plane disable handler using the atomic plane update
+ * functions. It is fully left to the driver to check plane constraints and
+ * handle corner-cases like a fully occluded or otherwise invisible plane.
+ *
+ * This is useful for piecewise transitioning of a driver to the atomic helpers.
+ *
+ * RETURNS:
+ * Zero on success, error code on failure
+ */
+int drm_plane_helper_disable(struct drm_plane *plane)
+{
+	struct drm_plane_state *plane_state;
+
+	/* crtc helpers love to call disable functions for already disabled hw
+	 * functions. So cope with that. */
+	if (!plane->crtc)
+		return 0;
+
+	if (plane->funcs->atomic_duplicate_state)
+		plane_state = plane->funcs->atomic_duplicate_state(plane);
+	else if (plane->state)
+		plane_state = drm_atomic_helper_plane_duplicate_state(plane);
+	else
+		plane_state = kzalloc(sizeof(*plane_state), GFP_KERNEL);
+	if (!plane_state)
+		return -ENOMEM;
+
+	plane_state->crtc = NULL;
+	drm_atomic_set_fb_for_plane(plane_state, NULL);
+
+	return drm_plane_helper_commit(plane, plane_state, plane->fb);
+}
+EXPORT_SYMBOL(drm_plane_helper_disable);
diff --git a/drivers/gpu/drm/drm_platform.c b/drivers/gpu/drm/drm_platform.c
index d5b76f148c12..5314c9d5fef4 100644
--- a/drivers/gpu/drm/drm_platform.c
+++ b/drivers/gpu/drm/drm_platform.c
@@ -68,42 +68,23 @@ err_free:
 	return ret;
 }
 
-static int drm_platform_set_busid(struct drm_device *dev, struct drm_master *master)
+int drm_platform_set_busid(struct drm_device *dev, struct drm_master *master)
 {
-	int len, ret, id;
-
-	master->unique_len = 13 + strlen(dev->platformdev->name);
-	master->unique_size = master->unique_len;
-	master->unique = kmalloc(master->unique_len + 1, GFP_KERNEL);
-
-	if (master->unique == NULL)
-		return -ENOMEM;
+	int id;
 
 	id = dev->platformdev->id;
-
-	/* if only a single instance of the platform device, id will be
-	 * set to -1.. use 0 instead to avoid a funny looking bus-id:
-	 */
-	if (id == -1)
+	if (id < 0)
 		id = 0;
 
-	len = snprintf(master->unique, master->unique_len,
-			"platform:%s:%02d", dev->platformdev->name, id);
-
-	if (len > master->unique_len) {
-		DRM_ERROR("Unique buffer overflowed\n");
-		ret = -EINVAL;
-		goto err;
-	}
+	master->unique = kasprintf(GFP_KERNEL, "platform:%s:%02d",
+						dev->platformdev->name, id);
+	if (!master->unique)
+		return -ENOMEM;
 
+	master->unique_len = strlen(master->unique);
 	return 0;
-err:
-	return ret;
 }
-
-static struct drm_bus drm_platform_bus = {
-	.set_busid = drm_platform_set_busid,
-};
+EXPORT_SYMBOL(drm_platform_set_busid);
 
 /**
  * drm_platform_init - Register a platform device with the DRM subsystem
@@ -120,7 +101,6 @@ int drm_platform_init(struct drm_driver *driver, struct platform_device *platfor
 {
 	DRM_DEBUG("\n");
 
-	driver->bus = &drm_platform_bus;
 	return drm_get_platform_dev(platform_device, driver);
 }
 EXPORT_SYMBOL(drm_platform_init);
diff --git a/drivers/gpu/drm/drm_prime.c b/drivers/gpu/drm/drm_prime.c
index 99d578bad17e..7482b06cd08f 100644
--- a/drivers/gpu/drm/drm_prime.c
+++ b/drivers/gpu/drm/drm_prime.c
@@ -29,6 +29,9 @@
 #include <linux/export.h>
 #include <linux/dma-buf.h>
 #include <drm/drmP.h>
+#include <drm/drm_gem.h>
+
+#include "drm_internal.h"
 
 /*
  * DMA-BUF/GEM Object references and lifetime overview:
@@ -325,7 +328,7 @@ static const struct dma_buf_ops drm_gem_prime_dmabuf_ops =  {
  */
 
 /**
- * drm_gem_prime_export - helper library implemention of the export callback
+ * drm_gem_prime_export - helper library implementation of the export callback
  * @dev: drm_device to export from
  * @obj: GEM object to export
  * @flags: flags like DRM_CLOEXEC
@@ -480,7 +483,7 @@ out_unlock:
 EXPORT_SYMBOL(drm_gem_prime_handle_to_fd);
 
 /**
- * drm_gem_prime_import - helper library implemention of the import callback
+ * drm_gem_prime_import - helper library implementation of the import callback
  * @dev: drm_device to import into
  * @dma_buf: dma-buf object to import
  *
@@ -522,7 +525,7 @@ struct drm_gem_object *drm_gem_prime_import(struct drm_device *dev,
 		goto fail_detach;
 	}
 
-	obj = dev->driver->gem_prime_import_sg_table(dev, dma_buf->size, sgt);
+	obj = dev->driver->gem_prime_import_sg_table(dev, attach, sgt);
 	if (IS_ERR(obj)) {
 		ret = PTR_ERR(obj);
 		goto fail_unmap;
@@ -666,7 +669,7 @@ int drm_prime_fd_to_handle_ioctl(struct drm_device *dev, void *data,
  * the driver is responsible for mapping the pages into the
  * importers address space for use with dma_buf itself.
  */
-struct sg_table *drm_prime_pages_to_sg(struct page **pages, int nr_pages)
+struct sg_table *drm_prime_pages_to_sg(struct page **pages, unsigned int nr_pages)
 {
 	struct sg_table *sg = NULL;
 	int ret;
diff --git a/drivers/gpu/drm/drm_probe_helper.c b/drivers/gpu/drm/drm_probe_helper.c
index db7d250f7ac7..7483a47de8e4 100644
--- a/drivers/gpu/drm/drm_probe_helper.c
+++ b/drivers/gpu/drm/drm_probe_helper.c
@@ -82,6 +82,22 @@ static void drm_mode_validate_flag(struct drm_connector *connector,
 	return;
 }
 
+static int drm_helper_probe_add_cmdline_mode(struct drm_connector *connector)
+{
+	struct drm_display_mode *mode;
+
+	if (!connector->cmdline_mode.specified)
+		return 0;
+
+	mode = drm_mode_create_from_cmdline_mode(connector->dev,
+						 &connector->cmdline_mode);
+	if (mode == NULL)
+		return 0;
+
+	drm_mode_probed_add(connector, mode);
+	return 1;
+}
+
 static int drm_helper_probe_single_connector_modes_merge_bits(struct drm_connector *connector,
 							      uint32_t maxX, uint32_t maxY, bool merge_type_bits)
 {
@@ -102,7 +118,8 @@ static int drm_helper_probe_single_connector_modes_merge_bits(struct drm_connect
 		mode->status = MODE_UNVERIFIED;
 
 	if (connector->force) {
-		if (connector->force == DRM_FORCE_ON)
+		if (connector->force == DRM_FORCE_ON ||
+		    connector->force == DRM_FORCE_ON_DIGITAL)
 			connector->status = connector_status_connected;
 		else
 			connector->status = connector_status_disconnected;
@@ -141,6 +158,7 @@ static int drm_helper_probe_single_connector_modes_merge_bits(struct drm_connect
 
 	if (count == 0 && connector->status == connector_status_connected)
 		count = drm_add_modes_noedid(connector, 1024, 768);
+	count += drm_helper_probe_add_cmdline_mode(connector);
 	if (count == 0)
 		goto prune;
 
diff --git a/drivers/gpu/drm/drm_scatter.c b/drivers/gpu/drm/drm_scatter.c
index 1c78406f6e71..4f0f3b36d537 100644
--- a/drivers/gpu/drm/drm_scatter.c
+++ b/drivers/gpu/drm/drm_scatter.c
@@ -34,6 +34,7 @@
 #include <linux/vmalloc.h>
 #include <linux/slab.h>
 #include <drm/drmP.h>
+#include "drm_legacy.h"
 
 #define DEBUG_SCATTER 0
 
@@ -78,8 +79,8 @@ void drm_legacy_sg_cleanup(struct drm_device *dev)
 # define ScatterHandle(x) (unsigned int)(x)
 #endif
 
-int drm_sg_alloc(struct drm_device *dev, void *data,
-		 struct drm_file *file_priv)
+int drm_legacy_sg_alloc(struct drm_device *dev, void *data,
+			struct drm_file *file_priv)
 {
 	struct drm_scatter_gather *request = data;
 	struct drm_sg_mem *entry;
@@ -194,8 +195,8 @@ int drm_sg_alloc(struct drm_device *dev, void *data,
 	return -ENOMEM;
 }
 
-int drm_sg_free(struct drm_device *dev, void *data,
-		struct drm_file *file_priv)
+int drm_legacy_sg_free(struct drm_device *dev, void *data,
+		       struct drm_file *file_priv)
 {
 	struct drm_scatter_gather *request = data;
 	struct drm_sg_mem *entry;
diff --git a/drivers/gpu/drm/drm_sysfs.c b/drivers/gpu/drm/drm_sysfs.c
index ab1a5f6dde8a..cc3d6d6d67e0 100644
--- a/drivers/gpu/drm/drm_sysfs.c
+++ b/drivers/gpu/drm/drm_sysfs.c
@@ -21,6 +21,7 @@
 #include <drm/drm_sysfs.h>
 #include <drm/drm_core.h>
 #include <drm/drmP.h>
+#include "drm_internal.h"
 
 #define to_drm_minor(d) dev_get_drvdata(d)
 #define to_drm_connector(d) dev_get_drvdata(d)
diff --git a/drivers/gpu/drm/drm_usb.c b/drivers/gpu/drm/drm_usb.c
deleted file mode 100644
index f2fe94aab901..000000000000
--- a/drivers/gpu/drm/drm_usb.c
+++ /dev/null
@@ -1,88 +0,0 @@
-#include <drm/drmP.h>
-#include <drm/drm_usb.h>
-#include <linux/usb.h>
-#include <linux/module.h>
-
-int drm_get_usb_dev(struct usb_interface *interface,
-		    const struct usb_device_id *id,
-		    struct drm_driver *driver)
-{
-	struct drm_device *dev;
-	int ret;
-
-	DRM_DEBUG("\n");
-
-	dev = drm_dev_alloc(driver, &interface->dev);
-	if (!dev)
-		return -ENOMEM;
-
-	dev->usbdev = interface_to_usbdev(interface);
-	usb_set_intfdata(interface, dev);
-
-	ret = drm_dev_register(dev, 0);
-	if (ret)
-		goto err_free;
-
-	DRM_INFO("Initialized %s %d.%d.%d %s on minor %d\n",
-		 driver->name, driver->major, driver->minor, driver->patchlevel,
-		 driver->date, dev->primary->index);
-
-	return 0;
-
-err_free:
-	drm_dev_unref(dev);
-	return ret;
-
-}
-EXPORT_SYMBOL(drm_get_usb_dev);
-
-static int drm_usb_set_busid(struct drm_device *dev,
-			       struct drm_master *master)
-{
-	return 0;
-}
-
-static struct drm_bus drm_usb_bus = {
-	.set_busid = drm_usb_set_busid,
-};
-
-/**
- * drm_usb_init - Register matching USB devices with the DRM subsystem
- * @driver: DRM device driver
- * @udriver: USB device driver
- *
- * Registers one or more devices matched by a USB driver with the DRM
- * subsystem.
- *
- * Return: 0 on success or a negative error code on failure.
- */
-int drm_usb_init(struct drm_driver *driver, struct usb_driver *udriver)
-{
-	int res;
-	DRM_DEBUG("\n");
-
-	driver->bus = &drm_usb_bus;
-
-	res = usb_register(udriver);
-	return res;
-}
-EXPORT_SYMBOL(drm_usb_init);
-
-/**
- * drm_usb_exit - Unregister matching USB devices from the DRM subsystem
- * @driver: DRM device driver
- * @udriver: USB device driver
- *
- * Unregisters one or more devices matched by a USB driver from the DRM
- * subsystem.
- */
-void drm_usb_exit(struct drm_driver *driver,
-		  struct usb_driver *udriver)
-{
-	usb_deregister(udriver);
-}
-EXPORT_SYMBOL(drm_usb_exit);
-
-MODULE_AUTHOR("David Airlie");
-MODULE_DESCRIPTION("USB DRM support");
-MODULE_LICENSE("GPL and additional rights");
diff --git a/drivers/gpu/drm/drm_vm.c b/drivers/gpu/drm/drm_vm.c
index 24e045c4f531..4a2c328959e5 100644
--- a/drivers/gpu/drm/drm_vm.c
+++ b/drivers/gpu/drm/drm_vm.c
@@ -35,10 +35,19 @@
 
 #include <drm/drmP.h>
 #include <linux/export.h>
+#include <linux/seq_file.h>
 #if defined(__ia64__)
 #include <linux/efi.h>
 #include <linux/slab.h>
 #endif
+#include <asm/pgtable.h>
+#include "drm_legacy.h"
+
+struct drm_vma_entry {
+	struct list_head head;
+	struct vm_area_struct *vma;
+	pid_t pid;
+};
 
 static void drm_vm_open(struct vm_area_struct *vma);
 static void drm_vm_close(struct vm_area_struct *vma);
@@ -48,15 +57,11 @@ static pgprot_t drm_io_prot(struct drm_local_map *map,
 {
 	pgprot_t tmp = vm_get_page_prot(vma->vm_flags);
 
-#if defined(__i386__) || defined(__x86_64__)
+#if defined(__i386__) || defined(__x86_64__) || defined(__powerpc__)
 	if (map->type == _DRM_REGISTERS && !(map->flags & _DRM_WRITE_COMBINING))
 		tmp = pgprot_noncached(tmp);
 	else
 		tmp = pgprot_writecombine(tmp);
-#elif defined(__powerpc__)
-	pgprot_val(tmp) |= _PAGE_NO_CACHE;
-	if (map->type == _DRM_REGISTERS)
-		pgprot_val(tmp) |= _PAGE_GUARDED;
 #elif defined(__ia64__)
 	if (efi_range_is_wc(vma->vm_start, vma->vm_end -
 				    vma->vm_start))
@@ -263,7 +268,7 @@ static void drm_vm_shm_close(struct vm_area_struct *vma)
 				dmah.vaddr = map->handle;
 				dmah.busaddr = map->offset;
 				dmah.size = map->size;
-				__drm_pci_free(dev, &dmah);
+				__drm_legacy_pci_free(dev, &dmah);
 				break;
 			}
 			kfree(map);
@@ -412,7 +417,6 @@ void drm_vm_open_locked(struct drm_device *dev,
 		list_add(&vma_entry->head, &dev->vmalist);
 	}
 }
-EXPORT_SYMBOL_GPL(drm_vm_open_locked);
 
 static void drm_vm_open(struct vm_area_struct *vma)
 {
@@ -532,7 +536,7 @@ static resource_size_t drm_core_get_reg_ofs(struct drm_device *dev)
  * according to the mapping type and remaps the pages. Finally sets the file
  * pointer and calls vm_open().
  */
-int drm_mmap_locked(struct file *filp, struct vm_area_struct *vma)
+static int drm_mmap_locked(struct file *filp, struct vm_area_struct *vma)
 {
 	struct drm_file *priv = filp->private_data;
 	struct drm_device *dev = priv->minor->dev;
@@ -646,7 +650,7 @@ int drm_mmap_locked(struct file *filp, struct vm_area_struct *vma)
 	return 0;
 }
 
-int drm_mmap(struct file *filp, struct vm_area_struct *vma)
+int drm_legacy_mmap(struct file *filp, struct vm_area_struct *vma)
 {
 	struct drm_file *priv = filp->private_data;
 	struct drm_device *dev = priv->minor->dev;
@@ -661,4 +665,69 @@ int drm_mmap(struct file *filp, struct vm_area_struct *vma)
 
 	return ret;
 }
-EXPORT_SYMBOL(drm_mmap);
+EXPORT_SYMBOL(drm_legacy_mmap);
+
+void drm_legacy_vma_flush(struct drm_device *dev)
+{
+	struct drm_vma_entry *vma, *vma_temp;
+
+	/* Clear vma list (only needed for legacy drivers) */
+	list_for_each_entry_safe(vma, vma_temp, &dev->vmalist, head) {
+		list_del(&vma->head);
+		kfree(vma);
+	}
+}
+
+int drm_vma_info(struct seq_file *m, void *data)
+{
+	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct drm_vma_entry *pt;
+	struct vm_area_struct *vma;
+	unsigned long vma_count = 0;
+#if defined(__i386__)
+	unsigned int pgprot;
+#endif
+
+	mutex_lock(&dev->struct_mutex);
+	list_for_each_entry(pt, &dev->vmalist, head)
+		vma_count++;
+
+	seq_printf(m, "vma use count: %lu, high_memory = %pK, 0x%pK\n",
+		   vma_count, high_memory,
+		   (void *)(unsigned long)virt_to_phys(high_memory));
+
+	list_for_each_entry(pt, &dev->vmalist, head) {
+		vma = pt->vma;
+		if (!vma)
+			continue;
+		seq_printf(m,
+			   "\n%5d 0x%pK-0x%pK %c%c%c%c%c%c 0x%08lx000",
+			   pt->pid,
+			   (void *)vma->vm_start, (void *)vma->vm_end,
+			   vma->vm_flags & VM_READ ? 'r' : '-',
+			   vma->vm_flags & VM_WRITE ? 'w' : '-',
+			   vma->vm_flags & VM_EXEC ? 'x' : '-',
+			   vma->vm_flags & VM_MAYSHARE ? 's' : 'p',
+			   vma->vm_flags & VM_LOCKED ? 'l' : '-',
+			   vma->vm_flags & VM_IO ? 'i' : '-',
+			   vma->vm_pgoff);
+
+#if defined(__i386__)
+		pgprot = pgprot_val(vma->vm_page_prot);
+		seq_printf(m, " %c%c%c%c%c%c%c%c%c",
+			   pgprot & _PAGE_PRESENT ? 'p' : '-',
+			   pgprot & _PAGE_RW ? 'w' : 'r',
+			   pgprot & _PAGE_USER ? 'u' : 's',
+			   pgprot & _PAGE_PWT ? 't' : 'b',
+			   pgprot & _PAGE_PCD ? 'u' : 'c',
+			   pgprot & _PAGE_ACCESSED ? 'a' : '-',
+			   pgprot & _PAGE_DIRTY ? 'd' : '-',
+			   pgprot & _PAGE_PSE ? 'm' : 'k',
+			   pgprot & _PAGE_GLOBAL ? 'g' : 'l');
+#endif
+		seq_printf(m, "\n");
+	}
+	mutex_unlock(&dev->struct_mutex);
+	return 0;
+}
diff --git a/drivers/gpu/drm/exynos/exynos_dp_core.c b/drivers/gpu/drm/exynos/exynos_dp_core.c
index 4f3c7eb2d37d..34d46aa75416 100644
--- a/drivers/gpu/drm/exynos/exynos_dp_core.c
+++ b/drivers/gpu/drm/exynos/exynos_dp_core.c
@@ -30,12 +30,17 @@
 #include <drm/drm_panel.h>
 #include <drm/bridge/ptn3460.h>
 
-#include "exynos_drm_drv.h"
 #include "exynos_dp_core.h"
 
 #define ctx_from_connector(c)	container_of(c, struct exynos_dp_device, \
 					connector)
 
+static inline struct exynos_dp_device *
+display_to_dp(struct exynos_drm_display *d)
+{
+	return container_of(d, struct exynos_dp_device, display);
+}
+
 struct bridge_init {
 	struct i2c_client *client;
 	struct device_node *node;
@@ -329,8 +334,8 @@ static int exynos_dp_link_start(struct exynos_dp_device *dp)
 		return retval;
 
 	for (lane = 0; lane < lane_count; lane++)
-		buf[lane] = DP_TRAIN_PRE_EMPHASIS_0 |
-			    DP_TRAIN_VOLTAGE_SWING_400;
+		buf[lane] = DP_TRAIN_PRE_EMPH_LEVEL_0 |
+			    DP_TRAIN_VOLTAGE_SWING_LEVEL_0;
 
 	retval = exynos_dp_write_bytes_to_dpcd(dp, DP_TRAINING_LANE0_SET,
 			lane_count, buf);
@@ -882,7 +887,7 @@ static void exynos_dp_hotplug(struct work_struct *work)
 
 static void exynos_dp_commit(struct exynos_drm_display *display)
 {
-	struct exynos_dp_device *dp = display->ctx;
+	struct exynos_dp_device *dp = display_to_dp(display);
 	int ret;
 
 	/* Keep the panel disabled while we configure video */
@@ -937,6 +942,8 @@ static enum drm_connector_status exynos_dp_detect(
 
 static void exynos_dp_connector_destroy(struct drm_connector *connector)
 {
+	drm_connector_unregister(connector);
+	drm_connector_cleanup(connector);
 }
 
 static struct drm_connector_funcs exynos_dp_connector_funcs = {
@@ -1018,7 +1025,7 @@ static int exynos_drm_attach_lcd_bridge(struct drm_device *dev,
 static int exynos_dp_create_connector(struct exynos_drm_display *display,
 				struct drm_encoder *encoder)
 {
-	struct exynos_dp_device *dp = display->ctx;
+	struct exynos_dp_device *dp = display_to_dp(display);
 	struct drm_connector *connector = &dp->connector;
 	int ret;
 
@@ -1050,33 +1057,19 @@ static int exynos_dp_create_connector(struct exynos_drm_display *display,
 
 static void exynos_dp_phy_init(struct exynos_dp_device *dp)
 {
-	if (dp->phy) {
+	if (dp->phy)
 		phy_power_on(dp->phy);
-	} else if (dp->phy_addr) {
-		u32 reg;
-
-		reg = __raw_readl(dp->phy_addr);
-		reg |= dp->enable_mask;
-		__raw_writel(reg, dp->phy_addr);
-	}
 }
 
 static void exynos_dp_phy_exit(struct exynos_dp_device *dp)
 {
-	if (dp->phy) {
+	if (dp->phy)
 		phy_power_off(dp->phy);
-	} else if (dp->phy_addr) {
-		u32 reg;
-
-		reg = __raw_readl(dp->phy_addr);
-		reg &= ~(dp->enable_mask);
-		__raw_writel(reg, dp->phy_addr);
-	}
 }
 
 static void exynos_dp_poweron(struct exynos_drm_display *display)
 {
-	struct exynos_dp_device *dp = display->ctx;
+	struct exynos_dp_device *dp = display_to_dp(display);
 
 	if (dp->dpms_mode == DRM_MODE_DPMS_ON)
 		return;
@@ -1097,7 +1090,7 @@ static void exynos_dp_poweron(struct exynos_drm_display *display)
 
 static void exynos_dp_poweroff(struct exynos_drm_display *display)
 {
-	struct exynos_dp_device *dp = display->ctx;
+	struct exynos_dp_device *dp = display_to_dp(display);
 
 	if (dp->dpms_mode != DRM_MODE_DPMS_ON)
 		return;
@@ -1122,7 +1115,7 @@ static void exynos_dp_poweroff(struct exynos_drm_display *display)
 
 static void exynos_dp_dpms(struct exynos_drm_display *display, int mode)
 {
-	struct exynos_dp_device *dp = display->ctx;
+	struct exynos_dp_device *dp = display_to_dp(display);
 
 	switch (mode) {
 	case DRM_MODE_DPMS_ON:
@@ -1145,11 +1138,6 @@ static struct exynos_drm_display_ops exynos_dp_display_ops = {
 	.commit = exynos_dp_commit,
 };
 
-static struct exynos_drm_display exynos_dp_display = {
-	.type = EXYNOS_DISPLAY_TYPE_LCD,
-	.ops = &exynos_dp_display_ops,
-};
-
 static struct video_info *exynos_dp_dt_parse_pdata(struct device *dev)
 {
 	struct device_node *dp_node = dev->of_node;
@@ -1208,44 +1196,6 @@ static struct video_info *exynos_dp_dt_parse_pdata(struct device *dev)
 	return dp_video_config;
 }
 
-static int exynos_dp_dt_parse_phydata(struct exynos_dp_device *dp)
-{
-	struct device_node *dp_phy_node = of_node_get(dp->dev->of_node);
-	u32 phy_base;
-	int ret = 0;
-
-	dp_phy_node = of_find_node_by_name(dp_phy_node, "dptx-phy");
-	if (!dp_phy_node) {
-		dp->phy = devm_phy_get(dp->dev, "dp");
-		return PTR_ERR_OR_ZERO(dp->phy);
-	}
-
-	if (of_property_read_u32(dp_phy_node, "reg", &phy_base)) {
-		dev_err(dp->dev, "failed to get reg for dptx-phy\n");
-		ret = -EINVAL;
-		goto err;
-	}
-
-	if (of_property_read_u32(dp_phy_node, "samsung,enable-mask",
-				&dp->enable_mask)) {
-		dev_err(dp->dev, "failed to get enable-mask for dptx-phy\n");
-		ret = -EINVAL;
-		goto err;
-	}
-
-	dp->phy_addr = ioremap(phy_base, SZ_4);
-	if (!dp->phy_addr) {
-		dev_err(dp->dev, "failed to ioremap dp-phy\n");
-		ret = -ENOMEM;
-		goto err;
-	}
-
-err:
-	of_node_put(dp_phy_node);
-
-	return ret;
-}
-
 static int exynos_dp_dt_parse_panel(struct exynos_dp_device *dp)
 {
 	int ret;
@@ -1261,10 +1211,10 @@ static int exynos_dp_dt_parse_panel(struct exynos_dp_device *dp)
 
 static int exynos_dp_bind(struct device *dev, struct device *master, void *data)
 {
+	struct exynos_dp_device *dp = dev_get_drvdata(dev);
 	struct platform_device *pdev = to_platform_device(dev);
 	struct drm_device *drm_dev = data;
 	struct resource *res;
-	struct exynos_dp_device *dp = exynos_dp_display.ctx;
 	unsigned int irq_flags;
 	int ret = 0;
 
@@ -1275,9 +1225,21 @@ static int exynos_dp_bind(struct device *dev, struct device *master, void *data)
 	if (IS_ERR(dp->video_info))
 		return PTR_ERR(dp->video_info);
 
-	ret = exynos_dp_dt_parse_phydata(dp);
-	if (ret)
-		return ret;
+	dp->phy = devm_phy_get(dp->dev, "dp");
+	if (IS_ERR(dp->phy)) {
+		dev_err(dp->dev, "no DP phy configured\n");
+		ret = PTR_ERR(dp->phy);
+		if (ret) {
+			/*
+			 * phy itself is not enabled, so we can move forward
+			 * assigning NULL to phy pointer.
+			 */
+			if (ret == -ENOSYS || ret == -ENODEV)
+				dp->phy = NULL;
+			else
+				return ret;
+		}
+	}
 
 	if (!dp->panel) {
 		ret = exynos_dp_dt_parse_panel(dp);
@@ -1344,22 +1306,15 @@ static int exynos_dp_bind(struct device *dev, struct device *master, void *data)
 
 	dp->drm_dev = drm_dev;
 
-	platform_set_drvdata(pdev, &exynos_dp_display);
-
-	return exynos_drm_create_enc_conn(drm_dev, &exynos_dp_display);
+	return exynos_drm_create_enc_conn(drm_dev, &dp->display);
 }
 
 static void exynos_dp_unbind(struct device *dev, struct device *master,
 				void *data)
 {
-	struct exynos_drm_display *display = dev_get_drvdata(dev);
-	struct exynos_dp_device *dp = display->ctx;
-	struct drm_encoder *encoder = dp->encoder;
-
-	exynos_dp_dpms(display, DRM_MODE_DPMS_OFF);
+	struct exynos_dp_device *dp = dev_get_drvdata(dev);
 
-	encoder->funcs->destroy(encoder);
-	drm_connector_cleanup(&dp->connector);
+	exynos_dp_dpms(&dp->display, DRM_MODE_DPMS_OFF);
 }
 
 static const struct component_ops exynos_dp_ops = {
@@ -1374,16 +1329,20 @@ static int exynos_dp_probe(struct platform_device *pdev)
 	struct exynos_dp_device *dp;
 	int ret;
 
-	ret = exynos_drm_component_add(&pdev->dev, EXYNOS_DEVICE_TYPE_CONNECTOR,
-					exynos_dp_display.type);
-	if (ret)
-		return ret;
-
 	dp = devm_kzalloc(&pdev->dev, sizeof(struct exynos_dp_device),
 				GFP_KERNEL);
 	if (!dp)
 		return -ENOMEM;
 
+	dp->display.type = EXYNOS_DISPLAY_TYPE_LCD;
+	dp->display.ops = &exynos_dp_display_ops;
+	platform_set_drvdata(pdev, dp);
+
+	ret = exynos_drm_component_add(&pdev->dev, EXYNOS_DEVICE_TYPE_CONNECTOR,
+					dp->display.type);
+	if (ret)
+		return ret;
+
 	panel_node = of_parse_phandle(dev->of_node, "panel", 0);
 	if (panel_node) {
 		dp->panel = of_drm_find_panel(panel_node);
@@ -1392,8 +1351,6 @@ static int exynos_dp_probe(struct platform_device *pdev)
 			return -EPROBE_DEFER;
 	}
 
-	exynos_dp_display.ctx = dp;
-
 	ret = component_add(&pdev->dev, &exynos_dp_ops);
 	if (ret)
 		exynos_drm_component_del(&pdev->dev,
@@ -1413,19 +1370,17 @@ static int exynos_dp_remove(struct platform_device *pdev)
 #ifdef CONFIG_PM_SLEEP
 static int exynos_dp_suspend(struct device *dev)
 {
-	struct platform_device *pdev = to_platform_device(dev);
-	struct exynos_drm_display *display = platform_get_drvdata(pdev);
+	struct exynos_dp_device *dp = dev_get_drvdata(dev);
 
-	exynos_dp_dpms(display, DRM_MODE_DPMS_OFF);
+	exynos_dp_dpms(&dp->display, DRM_MODE_DPMS_OFF);
 	return 0;
 }
 
 static int exynos_dp_resume(struct device *dev)
 {
-	struct platform_device *pdev = to_platform_device(dev);
-	struct exynos_drm_display *display = platform_get_drvdata(pdev);
+	struct exynos_dp_device *dp = dev_get_drvdata(dev);
 
-	exynos_dp_dpms(display, DRM_MODE_DPMS_ON);
+	exynos_dp_dpms(&dp->display, DRM_MODE_DPMS_ON);
 	return 0;
 }
 #endif
diff --git a/drivers/gpu/drm/exynos/exynos_dp_core.h b/drivers/gpu/drm/exynos/exynos_dp_core.h
index a1aee6931bd7..164f171168e7 100644
--- a/drivers/gpu/drm/exynos/exynos_dp_core.h
+++ b/drivers/gpu/drm/exynos/exynos_dp_core.h
@@ -17,6 +17,8 @@
 #include <drm/drm_dp_helper.h>
 #include <drm/exynos_drm.h>
 
+#include "exynos_drm_drv.h"
+
 #define DP_TIMEOUT_LOOP_COUNT 100
 #define MAX_CR_LOOP 5
 #define MAX_EQ_LOOP 5
@@ -145,6 +147,7 @@ struct link_train {
 };
 
 struct exynos_dp_device {
+	struct exynos_drm_display display;
 	struct device		*dev;
 	struct drm_device	*drm_dev;
 	struct drm_connector	connector;
@@ -153,8 +156,6 @@ struct exynos_dp_device {
 	struct clk		*clock;
 	unsigned int		irq;
 	void __iomem		*reg_base;
-	void __iomem		*phy_addr;
-	unsigned int		enable_mask;
 
 	struct video_info	*video_info;
 	struct link_train	link_train;
diff --git a/drivers/gpu/drm/exynos/exynos_drm_crtc.c b/drivers/gpu/drm/exynos/exynos_drm_crtc.c
index b68e58f78cd1..45026e693225 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_crtc.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_crtc.c
@@ -32,7 +32,6 @@ enum exynos_crtc_mode {
  * Exynos specific crtc structure.
  *
  * @drm_crtc: crtc object.
- * @drm_plane: pointer of private plane object for this crtc
  * @manager: the manager associated with this crtc
  * @pipe: a crtc index created at load() with a new crtc object creation
  *	and the crtc object would be set to private->crtc array
@@ -46,7 +45,6 @@ enum exynos_crtc_mode {
  */
 struct exynos_drm_crtc {
 	struct drm_crtc			drm_crtc;
-	struct drm_plane		*plane;
 	struct exynos_drm_manager	*manager;
 	unsigned int			pipe;
 	unsigned int			dpms;
@@ -73,13 +71,16 @@ static void exynos_drm_crtc_dpms(struct drm_crtc *crtc, int mode)
 				!atomic_read(&exynos_crtc->pending_flip),
 				HZ/20))
 			atomic_set(&exynos_crtc->pending_flip, 0);
-		drm_vblank_off(crtc->dev, exynos_crtc->pipe);
+		drm_crtc_vblank_off(crtc);
 	}
 
 	if (manager->ops->dpms)
 		manager->ops->dpms(manager, mode);
 
 	exynos_crtc->dpms = mode;
+
+	if (mode == DRM_MODE_DPMS_ON)
+		drm_crtc_vblank_on(crtc);
 }
 
 static void exynos_drm_crtc_prepare(struct drm_crtc *crtc)
@@ -94,12 +95,12 @@ static void exynos_drm_crtc_commit(struct drm_crtc *crtc)
 
 	exynos_drm_crtc_dpms(crtc, DRM_MODE_DPMS_ON);
 
-	exynos_plane_commit(exynos_crtc->plane);
+	exynos_plane_commit(crtc->primary);
 
 	if (manager->ops->commit)
 		manager->ops->commit(manager);
 
-	exynos_plane_dpms(exynos_crtc->plane, DRM_MODE_DPMS_ON);
+	exynos_plane_dpms(crtc->primary, DRM_MODE_DPMS_ON);
 }
 
 static bool
@@ -123,10 +124,9 @@ exynos_drm_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *mode,
 {
 	struct exynos_drm_crtc *exynos_crtc = to_exynos_crtc(crtc);
 	struct exynos_drm_manager *manager = exynos_crtc->manager;
-	struct drm_plane *plane = exynos_crtc->plane;
+	struct drm_framebuffer *fb = crtc->primary->fb;
 	unsigned int crtc_w;
 	unsigned int crtc_h;
-	int ret;
 
 	/*
 	 * copy the mode data adjusted by mode_fixup() into crtc->mode
@@ -134,29 +134,21 @@ exynos_drm_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *mode,
 	 */
 	memcpy(&crtc->mode, adjusted_mode, sizeof(*adjusted_mode));
 
-	crtc_w = crtc->primary->fb->width - x;
-	crtc_h = crtc->primary->fb->height - y;
+	crtc_w = fb->width - x;
+	crtc_h = fb->height - y;
 
 	if (manager->ops->mode_set)
 		manager->ops->mode_set(manager, &crtc->mode);
 
-	ret = exynos_plane_mode_set(plane, crtc, crtc->primary->fb, 0, 0, crtc_w, crtc_h,
-				    x, y, crtc_w, crtc_h);
-	if (ret)
-		return ret;
-
-	plane->crtc = crtc;
-	plane->fb = crtc->primary->fb;
-	drm_framebuffer_reference(plane->fb);
-
-	return 0;
+	return exynos_plane_mode_set(crtc->primary, crtc, fb, 0, 0,
+				     crtc_w, crtc_h, x, y, crtc_w, crtc_h);
 }
 
 static int exynos_drm_crtc_mode_set_commit(struct drm_crtc *crtc, int x, int y,
 					  struct drm_framebuffer *old_fb)
 {
 	struct exynos_drm_crtc *exynos_crtc = to_exynos_crtc(crtc);
-	struct drm_plane *plane = exynos_crtc->plane;
+	struct drm_framebuffer *fb = crtc->primary->fb;
 	unsigned int crtc_w;
 	unsigned int crtc_h;
 	int ret;
@@ -167,11 +159,11 @@ static int exynos_drm_crtc_mode_set_commit(struct drm_crtc *crtc, int x, int y,
 		return -EPERM;
 	}
 
-	crtc_w = crtc->primary->fb->width - x;
-	crtc_h = crtc->primary->fb->height - y;
+	crtc_w = fb->width - x;
+	crtc_h = fb->height - y;
 
-	ret = exynos_plane_mode_set(plane, crtc, crtc->primary->fb, 0, 0, crtc_w, crtc_h,
-				    x, y, crtc_w, crtc_h);
+	ret = exynos_plane_mode_set(crtc->primary, crtc, fb, 0, 0,
+				    crtc_w, crtc_h, x, y, crtc_w, crtc_h);
 	if (ret)
 		return ret;
 
@@ -304,8 +296,7 @@ static int exynos_drm_crtc_set_property(struct drm_crtc *crtc,
 			exynos_drm_crtc_commit(crtc);
 			break;
 		case CRTC_MODE_BLANK:
-			exynos_plane_dpms(exynos_crtc->plane,
-					  DRM_MODE_DPMS_OFF);
+			exynos_plane_dpms(crtc->primary, DRM_MODE_DPMS_OFF);
 			break;
 		default:
 			break;
@@ -351,8 +342,10 @@ static void exynos_drm_crtc_attach_mode_property(struct drm_crtc *crtc)
 int exynos_drm_crtc_create(struct exynos_drm_manager *manager)
 {
 	struct exynos_drm_crtc *exynos_crtc;
+	struct drm_plane *plane;
 	struct exynos_drm_private *private = manager->drm_dev->dev_private;
 	struct drm_crtc *crtc;
+	int ret;
 
 	exynos_crtc = kzalloc(sizeof(*exynos_crtc), GFP_KERNEL);
 	if (!exynos_crtc)
@@ -364,11 +357,11 @@ int exynos_drm_crtc_create(struct exynos_drm_manager *manager)
 	exynos_crtc->dpms = DRM_MODE_DPMS_OFF;
 	exynos_crtc->manager = manager;
 	exynos_crtc->pipe = manager->pipe;
-	exynos_crtc->plane = exynos_plane_init(manager->drm_dev,
-				1 << manager->pipe, true);
-	if (!exynos_crtc->plane) {
-		kfree(exynos_crtc);
-		return -ENOMEM;
+	plane = exynos_plane_init(manager->drm_dev, 1 << manager->pipe,
+				  DRM_PLANE_TYPE_PRIMARY);
+	if (IS_ERR(plane)) {
+		ret = PTR_ERR(plane);
+		goto err_plane;
 	}
 
 	manager->crtc = &exynos_crtc->drm_crtc;
@@ -376,12 +369,22 @@ int exynos_drm_crtc_create(struct exynos_drm_manager *manager)
 
 	private->crtc[manager->pipe] = crtc;
 
-	drm_crtc_init(manager->drm_dev, crtc, &exynos_crtc_funcs);
+	ret = drm_crtc_init_with_planes(manager->drm_dev, crtc, plane, NULL,
+					&exynos_crtc_funcs);
+	if (ret < 0)
+		goto err_crtc;
+
 	drm_crtc_helper_add(crtc, &exynos_crtc_helper_funcs);
 
 	exynos_drm_crtc_attach_mode_property(crtc);
 
 	return 0;
+
+err_crtc:
+	plane->funcs->destroy(plane);
+err_plane:
+	kfree(exynos_crtc);
+	return ret;
 }
 
 int exynos_drm_crtc_enable_vblank(struct drm_device *dev, int pipe)
diff --git a/drivers/gpu/drm/exynos/exynos_drm_crtc.h b/drivers/gpu/drm/exynos/exynos_drm_crtc.h
index 690dcddab725..e353d353836f 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_crtc.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_crtc.h
@@ -15,10 +15,7 @@
 #ifndef _EXYNOS_DRM_CRTC_H_
 #define _EXYNOS_DRM_CRTC_H_
 
-struct drm_device;
-struct drm_crtc;
-struct exynos_drm_manager;
-struct exynos_drm_overlay;
+#include "exynos_drm_drv.h"
 
 int exynos_drm_crtc_create(struct exynos_drm_manager *manager);
 int exynos_drm_crtc_enable_vblank(struct drm_device *dev, int pipe);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_dpi.c b/drivers/gpu/drm/exynos/exynos_drm_dpi.c
index fa08f05e3e34..37678cf4425a 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_dpi.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_dpi.c
@@ -22,6 +22,7 @@
 #include "exynos_drm_drv.h"
 
 struct exynos_dpi {
+	struct exynos_drm_display display;
 	struct device *dev;
 	struct device_node *panel_node;
 
@@ -35,6 +36,11 @@ struct exynos_dpi {
 
 #define connector_to_dpi(c) container_of(c, struct exynos_dpi, connector)
 
+static inline struct exynos_dpi *display_to_dpi(struct exynos_drm_display *d)
+{
+	return container_of(d, struct exynos_dpi, display);
+}
+
 static enum drm_connector_status
 exynos_dpi_detect(struct drm_connector *connector, bool force)
 {
@@ -100,7 +106,7 @@ static struct drm_connector_helper_funcs exynos_dpi_connector_helper_funcs = {
 static int exynos_dpi_create_connector(struct exynos_drm_display *display,
 				       struct drm_encoder *encoder)
 {
-	struct exynos_dpi *ctx = display->ctx;
+	struct exynos_dpi *ctx = display_to_dpi(display);
 	struct drm_connector *connector = &ctx->connector;
 	int ret;
 
@@ -141,7 +147,7 @@ static void exynos_dpi_poweroff(struct exynos_dpi *ctx)
 
 static void exynos_dpi_dpms(struct exynos_drm_display *display, int mode)
 {
-	struct exynos_dpi *ctx = display->ctx;
+	struct exynos_dpi *ctx = display_to_dpi(display);
 
 	switch (mode) {
 	case DRM_MODE_DPMS_ON:
@@ -165,11 +171,6 @@ static struct exynos_drm_display_ops exynos_dpi_display_ops = {
 	.dpms = exynos_dpi_dpms
 };
 
-static struct exynos_drm_display exynos_dpi_display = {
-	.type = EXYNOS_DISPLAY_TYPE_LCD,
-	.ops = &exynos_dpi_display_ops,
-};
-
 /* of_* functions will be removed after merge of of_graph patches */
 static struct device_node *
 of_get_child_by_name_reg(struct device_node *parent, const char *name, u32 reg)
@@ -299,20 +300,21 @@ struct exynos_drm_display *exynos_dpi_probe(struct device *dev)
 	struct exynos_dpi *ctx;
 	int ret;
 
-	ret = exynos_drm_component_add(dev,
-					EXYNOS_DEVICE_TYPE_CONNECTOR,
-					exynos_dpi_display.type);
-	if (ret)
-		return ERR_PTR(ret);
-
 	ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
 	if (!ctx)
-		goto err_del_component;
+		return ERR_PTR(-ENOMEM);
 
+	ctx->display.type = EXYNOS_DISPLAY_TYPE_LCD;
+	ctx->display.ops = &exynos_dpi_display_ops;
 	ctx->dev = dev;
-	exynos_dpi_display.ctx = ctx;
 	ctx->dpms_mode = DRM_MODE_DPMS_OFF;
 
+	ret = exynos_drm_component_add(dev,
+					EXYNOS_DEVICE_TYPE_CONNECTOR,
+					ctx->display.type);
+	if (ret)
+		return ERR_PTR(ret);
+
 	ret = exynos_dpi_parse_dt(ctx);
 	if (ret < 0) {
 		devm_kfree(dev, ctx);
@@ -328,7 +330,7 @@ struct exynos_drm_display *exynos_dpi_probe(struct device *dev)
 		}
 	}
 
-	return &exynos_dpi_display;
+	return &ctx->display;
 
 err_del_component:
 	exynos_drm_component_del(dev, EXYNOS_DEVICE_TYPE_CONNECTOR);
@@ -336,16 +338,16 @@ err_del_component:
 	return NULL;
 }
 
-int exynos_dpi_remove(struct device *dev)
+int exynos_dpi_remove(struct exynos_drm_display *display)
 {
-	struct drm_encoder *encoder = exynos_dpi_display.encoder;
-	struct exynos_dpi *ctx = exynos_dpi_display.ctx;
+	struct exynos_dpi *ctx = display_to_dpi(display);
 
-	exynos_dpi_dpms(&exynos_dpi_display, DRM_MODE_DPMS_OFF);
-	encoder->funcs->destroy(encoder);
-	drm_connector_cleanup(&ctx->connector);
+	exynos_dpi_dpms(&ctx->display, DRM_MODE_DPMS_OFF);
 
-	exynos_drm_component_del(dev, EXYNOS_DEVICE_TYPE_CONNECTOR);
+	if (ctx->panel)
+		drm_panel_detach(ctx->panel);
+
+	exynos_drm_component_del(ctx->dev, EXYNOS_DEVICE_TYPE_CONNECTOR);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.c b/drivers/gpu/drm/exynos/exynos_drm_drv.c
index 0d74e9b99c4e..1bcbe07cecfc 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_drv.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_drv.c
@@ -15,7 +15,6 @@
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
 
-#include <linux/anon_inodes.h>
 #include <linux/component.h>
 
 #include <drm/exynos_drm.h>
@@ -86,17 +85,14 @@ static int exynos_drm_load(struct drm_device *dev, unsigned long flags)
 		struct drm_plane *plane;
 		unsigned long possible_crtcs = (1 << MAX_CRTC) - 1;
 
-		plane = exynos_plane_init(dev, possible_crtcs, false);
-		if (!plane)
-			goto err_mode_config_cleanup;
-	}
-
-	/* init kms poll for handling hpd */
-	drm_kms_helper_poll_init(dev);
+		plane = exynos_plane_init(dev, possible_crtcs,
+					  DRM_PLANE_TYPE_OVERLAY);
+		if (!IS_ERR(plane))
+			continue;
 
-	ret = drm_vblank_init(dev, MAX_CRTC);
-	if (ret)
+		ret = PTR_ERR(plane);
 		goto err_mode_config_cleanup;
+	}
 
 	/* setup possible_clones. */
 	exynos_drm_encoder_setup(dev);
@@ -106,22 +102,46 @@ static int exynos_drm_load(struct drm_device *dev, unsigned long flags)
 	/* Try to bind all sub drivers. */
 	ret = component_bind_all(dev->dev, dev);
 	if (ret)
-		goto err_cleanup_vblank;
+		goto err_mode_config_cleanup;
+
+	ret = drm_vblank_init(dev, dev->mode_config.num_crtc);
+	if (ret)
+		goto err_unbind_all;
 
 	/* Probe non kms sub drivers and virtual display driver. */
 	ret = exynos_drm_device_subdrv_probe(dev);
 	if (ret)
-		goto err_unbind_all;
+		goto err_cleanup_vblank;
+
+	/*
+	 * enable drm irq mode.
+	 * - with irq_enabled = true, we can use the vblank feature.
+	 *
+	 * P.S. note that we wouldn't use drm irq handler but
+	 *	just specific driver own one instead because
+	 *	drm framework supports only one irq handler.
+	 */
+	dev->irq_enabled = true;
+
+	/*
+	 * with vblank_disable_allowed = true, vblank interrupt will be disabled
+	 * by drm timer once a current process gives up ownership of
+	 * vblank event.(after drm_vblank_put function is called)
+	 */
+	dev->vblank_disable_allowed = true;
+
+	/* init kms poll for handling hpd */
+	drm_kms_helper_poll_init(dev);
 
 	/* force connectors detection */
 	drm_helper_hpd_irq_event(dev);
 
 	return 0;
 
-err_unbind_all:
-	component_unbind_all(dev->dev, dev);
 err_cleanup_vblank:
 	drm_vblank_cleanup(dev);
+err_unbind_all:
+	component_unbind_all(dev->dev, dev);
 err_mode_config_cleanup:
 	drm_mode_config_cleanup(dev);
 	drm_release_iommu_mapping(dev);
@@ -136,23 +156,19 @@ static int exynos_drm_unload(struct drm_device *dev)
 	exynos_drm_device_subdrv_remove(dev);
 
 	exynos_drm_fbdev_fini(dev);
-	drm_vblank_cleanup(dev);
 	drm_kms_helper_poll_fini(dev);
-	drm_mode_config_cleanup(dev);
 
+	drm_vblank_cleanup(dev);
+	component_unbind_all(dev->dev, dev);
+	drm_mode_config_cleanup(dev);
 	drm_release_iommu_mapping(dev);
-	kfree(dev->dev_private);
 
-	component_unbind_all(dev->dev, dev);
+	kfree(dev->dev_private);
 	dev->dev_private = NULL;
 
 	return 0;
 }
 
-static const struct file_operations exynos_drm_gem_fops = {
-	.mmap = exynos_drm_gem_mmap_buffer,
-};
-
 static int exynos_drm_suspend(struct drm_device *dev, pm_message_t state)
 {
 	struct drm_connector *connector;
@@ -178,20 +194,21 @@ static int exynos_drm_resume(struct drm_device *dev)
 
 	drm_modeset_lock_all(dev);
 	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
-		if (connector->funcs->dpms)
-			connector->funcs->dpms(connector, connector->dpms);
+		if (connector->funcs->dpms) {
+			int dpms = connector->dpms;
+
+			connector->dpms = DRM_MODE_DPMS_OFF;
+			connector->funcs->dpms(connector, dpms);
+		}
 	}
 	drm_modeset_unlock_all(dev);
 
-	drm_helper_resume_force_mode(dev);
-
 	return 0;
 }
 
 static int exynos_drm_open(struct drm_device *dev, struct drm_file *file)
 {
 	struct drm_exynos_file_private *file_priv;
-	struct file *anon_filp;
 	int ret;
 
 	file_priv = kzalloc(sizeof(*file_priv), GFP_KERNEL);
@@ -204,21 +221,8 @@ static int exynos_drm_open(struct drm_device *dev, struct drm_file *file)
 	if (ret)
 		goto err_file_priv_free;
 
-	anon_filp = anon_inode_getfile("exynos_gem", &exynos_drm_gem_fops,
-					NULL, 0);
-	if (IS_ERR(anon_filp)) {
-		ret = PTR_ERR(anon_filp);
-		goto err_subdrv_close;
-	}
-
-	anon_filp->f_mode = FMODE_READ | FMODE_WRITE;
-	file_priv->anon_filp = anon_filp;
-
 	return ret;
 
-err_subdrv_close:
-	exynos_drm_subdrv_close(dev, file);
-
 err_file_priv_free:
 	kfree(file_priv);
 	file->driver_priv = NULL;
@@ -234,7 +238,6 @@ static void exynos_drm_preclose(struct drm_device *dev,
 static void exynos_drm_postclose(struct drm_device *dev, struct drm_file *file)
 {
 	struct exynos_drm_private *private = dev->dev_private;
-	struct drm_exynos_file_private *file_priv;
 	struct drm_pending_vblank_event *v, *vt;
 	struct drm_pending_event *e, *et;
 	unsigned long flags;
@@ -260,10 +263,6 @@ static void exynos_drm_postclose(struct drm_device *dev, struct drm_file *file)
 	}
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 
-	file_priv = file->driver_priv;
-	if (file_priv->anon_filp)
-		fput(file_priv->anon_filp);
-
 	kfree(file->driver_priv);
 	file->driver_priv = NULL;
 }
@@ -282,11 +281,6 @@ static const struct vm_operations_struct exynos_drm_gem_vm_ops = {
 static const struct drm_ioctl_desc exynos_ioctls[] = {
 	DRM_IOCTL_DEF_DRV(EXYNOS_GEM_CREATE, exynos_drm_gem_create_ioctl,
 			DRM_UNLOCKED | DRM_AUTH),
-	DRM_IOCTL_DEF_DRV(EXYNOS_GEM_MAP_OFFSET,
-			exynos_drm_gem_map_offset_ioctl, DRM_UNLOCKED |
-			DRM_AUTH),
-	DRM_IOCTL_DEF_DRV(EXYNOS_GEM_MMAP,
-			exynos_drm_gem_mmap_ioctl, DRM_UNLOCKED | DRM_AUTH),
 	DRM_IOCTL_DEF_DRV(EXYNOS_GEM_GET,
 			exynos_drm_gem_get_ioctl, DRM_UNLOCKED),
 	DRM_IOCTL_DEF_DRV(EXYNOS_VIDI_CONNECTION,
@@ -330,6 +324,7 @@ static struct drm_driver exynos_drm_driver = {
 	.preclose		= exynos_drm_preclose,
 	.lastclose		= exynos_drm_lastclose,
 	.postclose		= exynos_drm_postclose,
+	.set_busid		= drm_platform_set_busid,
 	.get_vblank_counter	= drm_vblank_count,
 	.enable_vblank		= exynos_drm_crtc_enable_vblank,
 	.disable_vblank		= exynos_drm_crtc_disable_vblank,
@@ -478,28 +473,31 @@ void exynos_drm_component_del(struct device *dev,
 			list_del(&cdev->list);
 			kfree(cdev);
 		}
-
-		break;
 	}
 
 	mutex_unlock(&drm_component_lock);
 }
 
-static int compare_of(struct device *dev, void *data)
+static int compare_dev(struct device *dev, void *data)
 {
 	return dev == (struct device *)data;
 }
 
-static int exynos_drm_add_components(struct device *dev, struct master *m)
+static struct component_match *exynos_drm_match_add(struct device *dev)
 {
+	struct component_match *match = NULL;
 	struct component_dev *cdev;
 	unsigned int attach_cnt = 0;
 
 	mutex_lock(&drm_component_lock);
 
-	list_for_each_entry(cdev, &drm_component_list, list) {
-		int ret;
+	/* Do not retry to probe if there is no any kms driver regitered. */
+	if (list_empty(&drm_component_list)) {
+		mutex_unlock(&drm_component_lock);
+		return ERR_PTR(-ENODEV);
+	}
 
+	list_for_each_entry(cdev, &drm_component_list, list) {
 		/*
 		 * Add components to master only in case that crtc and
 		 * encoder/connector device objects exist.
@@ -514,16 +512,10 @@ static int exynos_drm_add_components(struct device *dev, struct master *m)
 		/*
 		 * fimd and dpi modules have same device object so add
 		 * only crtc device object in this case.
-		 *
-		 * TODO. if dpi module follows driver-model driver then
-		 * below codes can be removed.
 		 */
 		if (cdev->crtc_dev == cdev->conn_dev) {
-			ret = component_master_add_child(m, compare_of,
-					cdev->crtc_dev);
-			if (ret < 0)
-				return ret;
-
+			component_match_add(dev, &match, compare_dev,
+						cdev->crtc_dev);
 			goto out_lock;
 		}
 
@@ -533,11 +525,8 @@ static int exynos_drm_add_components(struct device *dev, struct master *m)
 		 * connector/encoder need pipe number of crtc when they
 		 * are created.
 		 */
-		ret = component_master_add_child(m, compare_of, cdev->crtc_dev);
-		ret |= component_master_add_child(m, compare_of,
-							cdev->conn_dev);
-		if (ret < 0)
-			return ret;
+		component_match_add(dev, &match, compare_dev, cdev->crtc_dev);
+		component_match_add(dev, &match, compare_dev, cdev->conn_dev);
 
 out_lock:
 		mutex_lock(&drm_component_lock);
@@ -545,7 +534,7 @@ out_lock:
 
 	mutex_unlock(&drm_component_lock);
 
-	return attach_cnt ? 0 : -ENODEV;
+	return attach_cnt ? match : ERR_PTR(-EPROBE_DEFER);
 }
 
 static int exynos_drm_bind(struct device *dev)
@@ -559,182 +548,76 @@ static void exynos_drm_unbind(struct device *dev)
 }
 
 static const struct component_master_ops exynos_drm_ops = {
-	.add_components = exynos_drm_add_components,
 	.bind		= exynos_drm_bind,
 	.unbind		= exynos_drm_unbind,
 };
 
-static int exynos_drm_platform_probe(struct platform_device *pdev)
-{
-	int ret;
-
-	pdev->dev.coherent_dma_mask = DMA_BIT_MASK(32);
-	exynos_drm_driver.num_ioctls = ARRAY_SIZE(exynos_ioctls);
-
+static struct platform_driver *const exynos_drm_kms_drivers[] = {
 #ifdef CONFIG_DRM_EXYNOS_FIMD
-	ret = platform_driver_register(&fimd_driver);
-	if (ret < 0)
-		return ret;
+	&fimd_driver,
 #endif
-
 #ifdef CONFIG_DRM_EXYNOS_DP
-	ret = platform_driver_register(&dp_driver);
-	if (ret < 0)
-		goto err_unregister_fimd_drv;
+	&dp_driver,
 #endif
-
 #ifdef CONFIG_DRM_EXYNOS_DSI
-	ret = platform_driver_register(&dsi_driver);
-	if (ret < 0)
-		goto err_unregister_dp_drv;
+	&dsi_driver,
 #endif
-
 #ifdef CONFIG_DRM_EXYNOS_HDMI
-	ret = platform_driver_register(&mixer_driver);
-	if (ret < 0)
-		goto err_unregister_dsi_drv;
-	ret = platform_driver_register(&hdmi_driver);
-	if (ret < 0)
-		goto err_unregister_mixer_drv;
+	&mixer_driver,
+	&hdmi_driver,
 #endif
+};
 
+static struct platform_driver *const exynos_drm_non_kms_drivers[] = {
 #ifdef CONFIG_DRM_EXYNOS_G2D
-	ret = platform_driver_register(&g2d_driver);
-	if (ret < 0)
-		goto err_unregister_hdmi_drv;
+	&g2d_driver,
 #endif
-
 #ifdef CONFIG_DRM_EXYNOS_FIMC
-	ret = platform_driver_register(&fimc_driver);
-	if (ret < 0)
-		goto err_unregister_g2d_drv;
+	&fimc_driver,
 #endif
-
 #ifdef CONFIG_DRM_EXYNOS_ROTATOR
-	ret = platform_driver_register(&rotator_driver);
-	if (ret < 0)
-		goto err_unregister_fimc_drv;
+	&rotator_driver,
 #endif
-
 #ifdef CONFIG_DRM_EXYNOS_GSC
-	ret = platform_driver_register(&gsc_driver);
-	if (ret < 0)
-		goto err_unregister_rotator_drv;
-#endif
-
-#ifdef CONFIG_DRM_EXYNOS_IPP
-	ret = platform_driver_register(&ipp_driver);
-	if (ret < 0)
-		goto err_unregister_gsc_drv;
-
-	ret = exynos_platform_device_ipp_register();
-	if (ret < 0)
-		goto err_unregister_ipp_drv;
+	&gsc_driver,
 #endif
-
-	ret = component_master_add(&pdev->dev, &exynos_drm_ops);
-	if (ret < 0)
-		DRM_DEBUG_KMS("re-tried by last sub driver probed later.\n");
-
-	return 0;
-
 #ifdef CONFIG_DRM_EXYNOS_IPP
-err_unregister_ipp_drv:
-	platform_driver_unregister(&ipp_driver);
-err_unregister_gsc_drv:
-#endif
-
-#ifdef CONFIG_DRM_EXYNOS_GSC
-	platform_driver_unregister(&gsc_driver);
-err_unregister_rotator_drv:
-#endif
-
-#ifdef CONFIG_DRM_EXYNOS_ROTATOR
-	platform_driver_unregister(&rotator_driver);
-err_unregister_fimc_drv:
-#endif
-
-#ifdef CONFIG_DRM_EXYNOS_FIMC
-	platform_driver_unregister(&fimc_driver);
-err_unregister_g2d_drv:
-#endif
-
-#ifdef CONFIG_DRM_EXYNOS_G2D
-	platform_driver_unregister(&g2d_driver);
-err_unregister_hdmi_drv:
+	&ipp_driver,
 #endif
+};
 
-#ifdef CONFIG_DRM_EXYNOS_HDMI
-	platform_driver_unregister(&hdmi_driver);
-err_unregister_mixer_drv:
-	platform_driver_unregister(&mixer_driver);
-err_unregister_dsi_drv:
-#endif
+static int exynos_drm_platform_probe(struct platform_device *pdev)
+{
+	struct component_match *match;
 
-#ifdef CONFIG_DRM_EXYNOS_DSI
-	platform_driver_unregister(&dsi_driver);
-err_unregister_dp_drv:
-#endif
+	pdev->dev.coherent_dma_mask = DMA_BIT_MASK(32);
+	exynos_drm_driver.num_ioctls = ARRAY_SIZE(exynos_ioctls);
 
-#ifdef CONFIG_DRM_EXYNOS_DP
-	platform_driver_unregister(&dp_driver);
-err_unregister_fimd_drv:
-#endif
+	match = exynos_drm_match_add(&pdev->dev);
+	if (IS_ERR(match)) {
+		return PTR_ERR(match);
+	}
 
-#ifdef CONFIG_DRM_EXYNOS_FIMD
-	platform_driver_unregister(&fimd_driver);
-#endif
-	return ret;
+	return component_master_add_with_match(&pdev->dev, &exynos_drm_ops,
+					       match);
 }
 
 static int exynos_drm_platform_remove(struct platform_device *pdev)
 {
-#ifdef CONFIG_DRM_EXYNOS_IPP
-	exynos_platform_device_ipp_unregister();
-	platform_driver_unregister(&ipp_driver);
-#endif
-
-#ifdef CONFIG_DRM_EXYNOS_GSC
-	platform_driver_unregister(&gsc_driver);
-#endif
-
-#ifdef CONFIG_DRM_EXYNOS_ROTATOR
-	platform_driver_unregister(&rotator_driver);
-#endif
-
-#ifdef CONFIG_DRM_EXYNOS_FIMC
-	platform_driver_unregister(&fimc_driver);
-#endif
-
-#ifdef CONFIG_DRM_EXYNOS_G2D
-	platform_driver_unregister(&g2d_driver);
-#endif
-
-#ifdef CONFIG_DRM_EXYNOS_HDMI
-	platform_driver_unregister(&mixer_driver);
-	platform_driver_unregister(&hdmi_driver);
-#endif
-
-#ifdef CONFIG_DRM_EXYNOS_FIMD
-	platform_driver_unregister(&fimd_driver);
-#endif
-
-#ifdef CONFIG_DRM_EXYNOS_DSI
-	platform_driver_unregister(&dsi_driver);
-#endif
-
-#ifdef CONFIG_DRM_EXYNOS_DP
-	platform_driver_unregister(&dp_driver);
-#endif
 	component_master_del(&pdev->dev, &exynos_drm_ops);
 	return 0;
 }
 
+static const char * const strings[] = {
+	"samsung,exynos3",
+	"samsung,exynos4",
+	"samsung,exynos5",
+};
+
 static struct platform_driver exynos_drm_platform_driver = {
 	.probe	= exynos_drm_platform_probe,
 	.remove	= exynos_drm_platform_remove,
 	.driver	= {
-		.owner	= THIS_MODULE,
 		.name	= "exynos-drm",
 		.pm	= &exynos_drm_pm_ops,
 	},
@@ -742,31 +625,75 @@ static struct platform_driver exynos_drm_platform_driver = {
 
 static int exynos_drm_init(void)
 {
-	int ret;
+	bool is_exynos = false;
+	int ret, i, j;
+
+	/*
+	 * Register device object only in case of Exynos SoC.
+	 *
+	 * Below codes resolves temporarily infinite loop issue incurred
+	 * by Exynos drm driver when using multi-platform kernel.
+	 * So these codes will be replaced with more generic way later.
+	 */
+	for (i = 0; i < ARRAY_SIZE(strings); i++) {
+		if (of_machine_is_compatible(strings[i])) {
+			is_exynos = true;
+			break;
+		}
+	}
+
+	if (!is_exynos)
+		return -ENODEV;
 
 	exynos_drm_pdev = platform_device_register_simple("exynos-drm", -1,
 								NULL, 0);
 	if (IS_ERR(exynos_drm_pdev))
 		return PTR_ERR(exynos_drm_pdev);
 
-#ifdef CONFIG_DRM_EXYNOS_VIDI
 	ret = exynos_drm_probe_vidi();
 	if (ret < 0)
 		goto err_unregister_pd;
+
+	for (i = 0; i < ARRAY_SIZE(exynos_drm_kms_drivers); ++i) {
+		ret = platform_driver_register(exynos_drm_kms_drivers[i]);
+		if (ret < 0)
+			goto err_unregister_kms_drivers;
+	}
+
+	for (j = 0; j < ARRAY_SIZE(exynos_drm_non_kms_drivers); ++j) {
+		ret = platform_driver_register(exynos_drm_non_kms_drivers[j]);
+		if (ret < 0)
+			goto err_unregister_non_kms_drivers;
+	}
+
+#ifdef CONFIG_DRM_EXYNOS_IPP
+	ret = exynos_platform_device_ipp_register();
+	if (ret < 0)
+		goto err_unregister_non_kms_drivers;
 #endif
 
 	ret = platform_driver_register(&exynos_drm_platform_driver);
 	if (ret)
-		goto err_remove_vidi;
+		goto err_unregister_resources;
 
 	return 0;
 
-err_remove_vidi:
-#ifdef CONFIG_DRM_EXYNOS_VIDI
+err_unregister_resources:
+#ifdef CONFIG_DRM_EXYNOS_IPP
+	exynos_platform_device_ipp_unregister();
+#endif
+
+err_unregister_non_kms_drivers:
+	while (--j >= 0)
+		platform_driver_unregister(exynos_drm_non_kms_drivers[j]);
+
+err_unregister_kms_drivers:
+	while (--i >= 0)
+		platform_driver_unregister(exynos_drm_kms_drivers[i]);
+
 	exynos_drm_remove_vidi();
 
 err_unregister_pd:
-#endif
 	platform_device_unregister(exynos_drm_pdev);
 
 	return ret;
@@ -774,10 +701,22 @@ err_unregister_pd:
 
 static void exynos_drm_exit(void)
 {
+	int i;
+
+#ifdef CONFIG_DRM_EXYNOS_IPP
+	exynos_platform_device_ipp_unregister();
+#endif
+
+	for (i = ARRAY_SIZE(exynos_drm_non_kms_drivers) - 1; i >= 0; --i)
+		platform_driver_unregister(exynos_drm_non_kms_drivers[i]);
+
+	for (i = ARRAY_SIZE(exynos_drm_kms_drivers) - 1; i >= 0; --i)
+		platform_driver_unregister(exynos_drm_kms_drivers[i]);
+
 	platform_driver_unregister(&exynos_drm_platform_driver);
-#ifdef CONFIG_DRM_EXYNOS_VIDI
+
 	exynos_drm_remove_vidi();
-#endif
+
 	platform_device_unregister(exynos_drm_pdev);
 }
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.h b/drivers/gpu/drm/exynos/exynos_drm_drv.h
index 69a6fa397d75..2e5063488c50 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_drv.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_drv.h
@@ -15,6 +15,7 @@
 #ifndef _EXYNOS_DRM_DRV_H_
 #define _EXYNOS_DRM_DRV_H_
 
+#include <drm/drmP.h>
 #include <linux/module.h>
 
 #define MAX_CRTC	3
@@ -22,24 +23,6 @@
 #define MAX_FB_BUFFER	4
 #define DEFAULT_ZPOS	-1
 
-#define _wait_for(COND, MS) ({ \
-	unsigned long timeout__ = jiffies + msecs_to_jiffies(MS);	\
-	int ret__ = 0;							\
-	while (!(COND)) {						\
-		if (time_after(jiffies, timeout__)) {			\
-			ret__ = -ETIMEDOUT;				\
-			break;						\
-		}							\
-	}								\
-	ret__;								\
-})
-
-#define wait_for(COND, MS) _wait_for(COND, MS)
-
-struct drm_device;
-struct exynos_drm_overlay;
-struct drm_connector;
-
 /* This enumerates device type. */
 enum exynos_drm_device_type {
 	EXYNOS_DEVICE_TYPE_NONE,
@@ -83,10 +66,10 @@ enum exynos_drm_output_type {
  * @dma_addr: array of bus(accessed by dma) address to the memory region
  *	      allocated for a overlay.
  * @zpos: order of overlay layer(z position).
- * @default_win: a window to be enabled.
- * @color_key: color key on or off.
  * @index_color: if using color key feature then this value would be used
  *			as index color.
+ * @default_win: a window to be enabled.
+ * @color_key: color key on or off.
  * @local_path: in case of lcd type, local path mode on or off.
  * @transparency: transparency on or off.
  * @activated: activated or not.
@@ -114,19 +97,20 @@ struct exynos_drm_overlay {
 	uint32_t pixel_format;
 	dma_addr_t dma_addr[MAX_FB_BUFFER];
 	int zpos;
-
-	bool default_win;
-	bool color_key;
 	unsigned int index_color;
-	bool local_path;
-	bool transparency;
-	bool activated;
+
+	bool default_win:1;
+	bool color_key:1;
+	bool local_path:1;
+	bool transparency:1;
+	bool activated:1;
 };
 
 /*
  * Exynos DRM Display Structure.
  *	- this structure is common to analog tv, digital tv and lcd panel.
  *
+ * @create_connector: initialize and register a new connector
  * @remove: cleans up the display for removal
  * @mode_fixup: fix mode data comparing to hw specific display mode.
  * @mode_set: convert drm_display_mode to hw specific display mode and
@@ -168,7 +152,6 @@ struct exynos_drm_display {
 	struct drm_encoder *encoder;
 	struct drm_connector *connector;
 	struct exynos_drm_display_ops *ops;
-	void *ctx;
 };
 
 /*
@@ -227,7 +210,6 @@ struct exynos_drm_manager {
 	struct drm_crtc *crtc;
 	int pipe;
 	struct exynos_drm_manager_ops *ops;
-	void *ctx;
 };
 
 struct exynos_drm_g2d_private {
@@ -240,7 +222,6 @@ struct exynos_drm_g2d_private {
 struct drm_exynos_file_private {
 	struct exynos_drm_g2d_private	*g2d_priv;
 	struct device			*ipp_dev;
-	struct file			*anon_filp;
 };
 
 /*
@@ -280,8 +261,6 @@ struct exynos_drm_private {
  * @dev: pointer to device object for subdrv device driver.
  * @drm_dev: pointer to drm_device and this pointer would be set
  *	when sub driver calls exynos_drm_subdrv_register().
- * @manager: subdrv has its own manager to control a hardware appropriately
- *     and we can access a hardware drawing on this manager.
  * @probe: this callback would be called by exynos drm driver after
  *     subdrv is registered to it.
  * @remove: this callback is used to release resources created
@@ -313,45 +292,34 @@ int exynos_drm_device_subdrv_remove(struct drm_device *dev);
 int exynos_drm_subdrv_open(struct drm_device *dev, struct drm_file *file);
 void exynos_drm_subdrv_close(struct drm_device *dev, struct drm_file *file);
 
-/*
- * this function registers exynos drm hdmi platform device. It ensures only one
- * instance of the device is created.
- */
-int exynos_platform_device_hdmi_register(void);
-
-/*
- * this function unregisters exynos drm hdmi platform device if it exists.
- */
-void exynos_platform_device_hdmi_unregister(void);
-
-/*
- * this function registers exynos drm ipp platform device.
- */
+#ifdef CONFIG_DRM_EXYNOS_IPP
 int exynos_platform_device_ipp_register(void);
-
-/*
- * this function unregisters exynos drm ipp platform device if it exists.
- */
 void exynos_platform_device_ipp_unregister(void);
+#else
+static inline int exynos_platform_device_ipp_register(void) { return 0; }
+static inline void exynos_platform_device_ipp_unregister(void) {}
+#endif
+
 
 #ifdef CONFIG_DRM_EXYNOS_DPI
 struct exynos_drm_display * exynos_dpi_probe(struct device *dev);
-int exynos_dpi_remove(struct device *dev);
+int exynos_dpi_remove(struct exynos_drm_display *display);
 #else
 static inline struct exynos_drm_display *
 exynos_dpi_probe(struct device *dev) { return NULL; }
-static inline int exynos_dpi_remove(struct device *dev) { return 0; }
+static inline int exynos_dpi_remove(struct exynos_drm_display *display)
+{
+	return 0;
+}
 #endif
 
-/*
- * this function registers exynos drm vidi platform device/driver.
- */
+#ifdef CONFIG_DRM_EXYNOS_VIDI
 int exynos_drm_probe_vidi(void);
-
-/*
- * this function unregister exynos drm vidi platform device/driver.
- */
 void exynos_drm_remove_vidi(void);
+#else
+static inline int exynos_drm_probe_vidi(void) { return 0; }
+static inline void exynos_drm_remove_vidi(void) {}
+#endif
 
 /* This function creates a encoder and a connector, and initializes them. */
 int exynos_drm_create_enc_conn(struct drm_device *dev,
diff --git a/drivers/gpu/drm/exynos/exynos_drm_dsi.c b/drivers/gpu/drm/exynos/exynos_drm_dsi.c
index 442aa2d00132..05fe93dc57a8 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_dsi.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_dsi.c
@@ -114,6 +114,8 @@
 #define DSIM_SYNC_INFORM		(1 << 27)
 #define DSIM_EOT_DISABLE		(1 << 28)
 #define DSIM_MFLUSH_VS			(1 << 29)
+/* This flag is valid only for exynos3250/3472/4415/5260/5430 */
+#define DSIM_CLKLANE_STOP		(1 << 30)
 
 /* DSIM_ESCMODE */
 #define DSIM_TX_TRIGGER_RST		(1 << 4)
@@ -262,12 +264,13 @@ struct exynos_dsi_driver_data {
 	unsigned int plltmr_reg;
 
 	unsigned int has_freqband:1;
+	unsigned int has_clklane_stop:1;
 };
 
 struct exynos_dsi {
+	struct exynos_drm_display display;
 	struct mipi_dsi_host dsi_host;
 	struct drm_connector connector;
-	struct drm_encoder *encoder;
 	struct device_node *panel_node;
 	struct drm_panel *panel;
 	struct device *dev;
@@ -301,9 +304,26 @@ struct exynos_dsi {
 #define host_to_dsi(host) container_of(host, struct exynos_dsi, dsi_host)
 #define connector_to_dsi(c) container_of(c, struct exynos_dsi, connector)
 
+static inline struct exynos_dsi *display_to_dsi(struct exynos_drm_display *d)
+{
+	return container_of(d, struct exynos_dsi, display);
+}
+
+static struct exynos_dsi_driver_data exynos3_dsi_driver_data = {
+	.plltmr_reg = 0x50,
+	.has_freqband = 1,
+	.has_clklane_stop = 1,
+};
+
 static struct exynos_dsi_driver_data exynos4_dsi_driver_data = {
 	.plltmr_reg = 0x50,
 	.has_freqband = 1,
+	.has_clklane_stop = 1,
+};
+
+static struct exynos_dsi_driver_data exynos4415_dsi_driver_data = {
+	.plltmr_reg = 0x58,
+	.has_clklane_stop = 1,
 };
 
 static struct exynos_dsi_driver_data exynos5_dsi_driver_data = {
@@ -311,8 +331,12 @@ static struct exynos_dsi_driver_data exynos5_dsi_driver_data = {
 };
 
 static struct of_device_id exynos_dsi_of_match[] = {
+	{ .compatible = "samsung,exynos3250-mipi-dsi",
+	  .data = &exynos3_dsi_driver_data },
 	{ .compatible = "samsung,exynos4210-mipi-dsi",
 	  .data = &exynos4_dsi_driver_data },
+	{ .compatible = "samsung,exynos4415-mipi-dsi",
+	  .data = &exynos4415_dsi_driver_data },
 	{ .compatible = "samsung,exynos5410-mipi-dsi",
 	  .data = &exynos5_dsi_driver_data },
 	{ }
@@ -421,7 +445,7 @@ static unsigned long exynos_dsi_set_pll(struct exynos_dsi *dsi,
 	if (!fout) {
 		dev_err(dsi->dev,
 			"failed to find PLL PMS for requested frequency\n");
-		return -EFAULT;
+		return 0;
 	}
 	dev_dbg(dsi->dev, "PLL freq %lu, (p %d, m %d, s %d)\n", fout, p, m, s);
 
@@ -453,7 +477,7 @@ static unsigned long exynos_dsi_set_pll(struct exynos_dsi *dsi,
 	do {
 		if (timeout-- == 0) {
 			dev_err(dsi->dev, "PLL failed to stabilize\n");
-			return -EFAULT;
+			return 0;
 		}
 		reg = readl(dsi->reg_base + DSIM_STATUS_REG);
 	} while ((reg & DSIM_PLL_STABLE) == 0);
@@ -569,6 +593,7 @@ static void exynos_dsi_disable_clock(struct exynos_dsi *dsi)
 
 static int exynos_dsi_init_link(struct exynos_dsi *dsi)
 {
+	struct exynos_dsi_driver_data *driver_data = dsi->driver_data;
 	int timeout;
 	u32 reg;
 	u32 lanes_mask;
@@ -650,6 +675,20 @@ static int exynos_dsi_init_link(struct exynos_dsi *dsi)
 	reg |= DSIM_LANE_EN(lanes_mask);
 	writel(reg, dsi->reg_base + DSIM_CONFIG_REG);
 
+	/*
+	 * Use non-continuous clock mode if the periparal wants and
+	 * host controller supports
+	 *
+	 * In non-continous clock mode, host controller will turn off
+	 * the HS clock between high-speed transmissions to reduce
+	 * power consumption.
+	 */
+	if (driver_data->has_clklane_stop &&
+			dsi->mode_flags & MIPI_DSI_CLOCK_NON_CONTINUOUS) {
+		reg |= DSIM_CLKLANE_STOP;
+		writel(reg, dsi->reg_base + DSIM_CONFIG_REG);
+	}
+
 	/* Check clock and data lane state are stop state */
 	timeout = 100;
 	do {
@@ -1077,7 +1116,7 @@ static irqreturn_t exynos_dsi_irq(int irq, void *dev_id)
 static irqreturn_t exynos_dsi_te_irq_handler(int irq, void *dev_id)
 {
 	struct exynos_dsi *dsi = (struct exynos_dsi *)dev_id;
-	struct drm_encoder *encoder = dsi->encoder;
+	struct drm_encoder *encoder = dsi->display.encoder;
 
 	if (dsi->state & DSIM_STATE_ENABLED)
 		exynos_drm_crtc_te_handler(encoder->crtc);
@@ -1116,6 +1155,7 @@ static int exynos_dsi_init(struct exynos_dsi *dsi)
 static int exynos_dsi_register_te_irq(struct exynos_dsi *dsi)
 {
 	int ret;
+	int te_gpio_irq;
 
 	dsi->te_gpio = of_get_named_gpio(dsi->panel_node, "te-gpios", 0);
 	if (!gpio_is_valid(dsi->te_gpio)) {
@@ -1130,14 +1170,10 @@ static int exynos_dsi_register_te_irq(struct exynos_dsi *dsi)
 		goto out;
 	}
 
-	/*
-	 * This TE GPIO IRQ should not be set to IRQ_NOAUTOEN, because panel
-	 * calls drm_panel_init() first then calls mipi_dsi_attach() in probe().
-	 * It means that te_gpio is invalid when exynos_dsi_enable_irq() is
-	 * called by drm_panel_init() before panel is attached.
-	 */
-	ret = request_threaded_irq(gpio_to_irq(dsi->te_gpio),
-					exynos_dsi_te_irq_handler, NULL,
+	te_gpio_irq = gpio_to_irq(dsi->te_gpio);
+
+	irq_set_status_flags(te_gpio_irq, IRQ_NOAUTOEN);
+	ret = request_threaded_irq(te_gpio_irq, exynos_dsi_te_irq_handler, NULL,
 					IRQF_TRIGGER_RISING, "TE", dsi);
 	if (ret) {
 		dev_err(dsi->dev, "request interrupt failed with %d\n", ret);
@@ -1168,9 +1204,6 @@ static int exynos_dsi_host_attach(struct mipi_dsi_host *host,
 	dsi->mode_flags = device->mode_flags;
 	dsi->panel_node = device->dev.of_node;
 
-	if (dsi->connector.dev)
-		drm_helper_hpd_irq_event(dsi->connector.dev);
-
 	/*
 	 * This is a temporary solution and should be made by more generic way.
 	 *
@@ -1184,6 +1217,9 @@ static int exynos_dsi_host_attach(struct mipi_dsi_host *host,
 			return ret;
 	}
 
+	if (dsi->connector.dev)
+		drm_helper_hpd_irq_event(dsi->connector.dev);
+
 	return 0;
 }
 
@@ -1209,7 +1245,7 @@ static bool exynos_dsi_is_short_dsi_type(u8 type)
 }
 
 static ssize_t exynos_dsi_host_transfer(struct mipi_dsi_host *host,
-				       struct mipi_dsi_msg *msg)
+				        const struct mipi_dsi_msg *msg)
 {
 	struct exynos_dsi *dsi = host_to_dsi(host);
 	struct exynos_dsi_transfer xfer;
@@ -1342,16 +1378,17 @@ static int exynos_dsi_enable(struct exynos_dsi *dsi)
 	exynos_dsi_set_display_mode(dsi);
 	exynos_dsi_set_display_enable(dsi, true);
 
+	dsi->state |= DSIM_STATE_ENABLED;
+
 	ret = drm_panel_enable(dsi->panel);
 	if (ret < 0) {
+		dsi->state &= ~DSIM_STATE_ENABLED;
 		exynos_dsi_set_display_enable(dsi, false);
 		drm_panel_unprepare(dsi->panel);
 		exynos_dsi_poweroff(dsi);
 		return ret;
 	}
 
-	dsi->state |= DSIM_STATE_ENABLED;
-
 	return 0;
 }
 
@@ -1370,7 +1407,7 @@ static void exynos_dsi_disable(struct exynos_dsi *dsi)
 
 static void exynos_dsi_dpms(struct exynos_drm_display *display, int mode)
 {
-	struct exynos_dsi *dsi = display->ctx;
+	struct exynos_dsi *dsi = display_to_dsi(display);
 
 	if (dsi->panel) {
 		switch (mode) {
@@ -1414,6 +1451,9 @@ exynos_dsi_detect(struct drm_connector *connector, bool force)
 
 static void exynos_dsi_connector_destroy(struct drm_connector *connector)
 {
+	drm_connector_unregister(connector);
+	drm_connector_cleanup(connector);
+	connector->dev = NULL;
 }
 
 static struct drm_connector_funcs exynos_dsi_connector_funcs = {
@@ -1444,7 +1484,7 @@ exynos_dsi_best_encoder(struct drm_connector *connector)
 {
 	struct exynos_dsi *dsi = connector_to_dsi(connector);
 
-	return dsi->encoder;
+	return dsi->display.encoder;
 }
 
 static struct drm_connector_helper_funcs exynos_dsi_connector_helper_funcs = {
@@ -1456,12 +1496,10 @@ static struct drm_connector_helper_funcs exynos_dsi_connector_helper_funcs = {
 static int exynos_dsi_create_connector(struct exynos_drm_display *display,
 				       struct drm_encoder *encoder)
 {
-	struct exynos_dsi *dsi = display->ctx;
+	struct exynos_dsi *dsi = display_to_dsi(display);
 	struct drm_connector *connector = &dsi->connector;
 	int ret;
 
-	dsi->encoder = encoder;
-
 	connector->polled = DRM_CONNECTOR_POLL_HPD;
 
 	ret = drm_connector_init(encoder->dev, connector,
@@ -1482,7 +1520,7 @@ static int exynos_dsi_create_connector(struct exynos_drm_display *display,
 static void exynos_dsi_mode_set(struct exynos_drm_display *display,
 			 struct drm_display_mode *mode)
 {
-	struct exynos_dsi *dsi = display->ctx;
+	struct exynos_dsi *dsi = display_to_dsi(display);
 	struct videomode *vm = &dsi->vm;
 
 	vm->hactive = mode->hdisplay;
@@ -1501,10 +1539,6 @@ static struct exynos_drm_display_ops exynos_dsi_display_ops = {
 	.dpms = exynos_dsi_dpms
 };
 
-static struct exynos_drm_display exynos_dsi_display = {
-	.type = EXYNOS_DISPLAY_TYPE_LCD,
-	.ops = &exynos_dsi_display_ops,
-};
 MODULE_DEVICE_TABLE(of, exynos_dsi_of_match);
 
 /* of_* functions will be removed after merge of of_graph patches */
@@ -1610,34 +1644,30 @@ end:
 static int exynos_dsi_bind(struct device *dev, struct device *master,
 				void *data)
 {
+	struct exynos_drm_display *display = dev_get_drvdata(dev);
+	struct exynos_dsi *dsi = display_to_dsi(display);
 	struct drm_device *drm_dev = data;
-	struct exynos_dsi *dsi;
 	int ret;
 
-	ret = exynos_drm_create_enc_conn(drm_dev, &exynos_dsi_display);
+	ret = exynos_drm_create_enc_conn(drm_dev, display);
 	if (ret) {
 		DRM_ERROR("Encoder create [%d] failed with %d\n",
-				exynos_dsi_display.type, ret);
+			  display->type, ret);
 		return ret;
 	}
 
-	dsi = exynos_dsi_display.ctx;
-
 	return mipi_dsi_host_register(&dsi->dsi_host);
 }
 
 static void exynos_dsi_unbind(struct device *dev, struct device *master,
 				void *data)
 {
-	struct exynos_dsi *dsi = exynos_dsi_display.ctx;
-	struct drm_encoder *encoder = dsi->encoder;
+	struct exynos_drm_display *display = dev_get_drvdata(dev);
+	struct exynos_dsi *dsi = display_to_dsi(display);
 
-	exynos_dsi_dpms(&exynos_dsi_display, DRM_MODE_DPMS_OFF);
+	exynos_dsi_dpms(display, DRM_MODE_DPMS_OFF);
 
 	mipi_dsi_host_unregister(&dsi->dsi_host);
-
-	encoder->funcs->destroy(encoder);
-	drm_connector_cleanup(&dsi->connector);
 }
 
 static const struct component_ops exynos_dsi_component_ops = {
@@ -1647,22 +1677,23 @@ static const struct component_ops exynos_dsi_component_ops = {
 
 static int exynos_dsi_probe(struct platform_device *pdev)
 {
+	struct device *dev = &pdev->dev;
 	struct resource *res;
 	struct exynos_dsi *dsi;
 	int ret;
 
-	ret = exynos_drm_component_add(&pdev->dev, EXYNOS_DEVICE_TYPE_CONNECTOR,
-					exynos_dsi_display.type);
+	dsi = devm_kzalloc(dev, sizeof(*dsi), GFP_KERNEL);
+	if (!dsi)
+		return -ENOMEM;
+
+	dsi->display.type = EXYNOS_DISPLAY_TYPE_LCD;
+	dsi->display.ops = &exynos_dsi_display_ops;
+
+	ret = exynos_drm_component_add(dev, EXYNOS_DEVICE_TYPE_CONNECTOR,
+				       dsi->display.type);
 	if (ret)
 		return ret;
 
-	dsi = devm_kzalloc(&pdev->dev, sizeof(*dsi), GFP_KERNEL);
-	if (!dsi) {
-		dev_err(&pdev->dev, "failed to allocate dsi object.\n");
-		ret = -ENOMEM;
-		goto err_del_component;
-	}
-
 	/* To be checked as invalid one */
 	dsi->te_gpio = -ENOENT;
 
@@ -1671,9 +1702,9 @@ static int exynos_dsi_probe(struct platform_device *pdev)
 	INIT_LIST_HEAD(&dsi->transfer_list);
 
 	dsi->dsi_host.ops = &exynos_dsi_ops;
-	dsi->dsi_host.dev = &pdev->dev;
+	dsi->dsi_host.dev = dev;
 
-	dsi->dev = &pdev->dev;
+	dsi->dev = dev;
 	dsi->driver_data = exynos_dsi_get_driver_data(pdev);
 
 	ret = exynos_dsi_parse_dt(dsi);
@@ -1682,70 +1713,68 @@ static int exynos_dsi_probe(struct platform_device *pdev)
 
 	dsi->supplies[0].supply = "vddcore";
 	dsi->supplies[1].supply = "vddio";
-	ret = devm_regulator_bulk_get(&pdev->dev, ARRAY_SIZE(dsi->supplies),
+	ret = devm_regulator_bulk_get(dev, ARRAY_SIZE(dsi->supplies),
 				      dsi->supplies);
 	if (ret) {
-		dev_info(&pdev->dev, "failed to get regulators: %d\n", ret);
+		dev_info(dev, "failed to get regulators: %d\n", ret);
 		return -EPROBE_DEFER;
 	}
 
-	dsi->pll_clk = devm_clk_get(&pdev->dev, "pll_clk");
+	dsi->pll_clk = devm_clk_get(dev, "pll_clk");
 	if (IS_ERR(dsi->pll_clk)) {
-		dev_info(&pdev->dev, "failed to get dsi pll input clock\n");
+		dev_info(dev, "failed to get dsi pll input clock\n");
 		ret = PTR_ERR(dsi->pll_clk);
 		goto err_del_component;
 	}
 
-	dsi->bus_clk = devm_clk_get(&pdev->dev, "bus_clk");
+	dsi->bus_clk = devm_clk_get(dev, "bus_clk");
 	if (IS_ERR(dsi->bus_clk)) {
-		dev_info(&pdev->dev, "failed to get dsi bus clock\n");
+		dev_info(dev, "failed to get dsi bus clock\n");
 		ret = PTR_ERR(dsi->bus_clk);
 		goto err_del_component;
 	}
 
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	dsi->reg_base = devm_ioremap_resource(&pdev->dev, res);
+	dsi->reg_base = devm_ioremap_resource(dev, res);
 	if (IS_ERR(dsi->reg_base)) {
-		dev_err(&pdev->dev, "failed to remap io region\n");
+		dev_err(dev, "failed to remap io region\n");
 		ret = PTR_ERR(dsi->reg_base);
 		goto err_del_component;
 	}
 
-	dsi->phy = devm_phy_get(&pdev->dev, "dsim");
+	dsi->phy = devm_phy_get(dev, "dsim");
 	if (IS_ERR(dsi->phy)) {
-		dev_info(&pdev->dev, "failed to get dsim phy\n");
+		dev_info(dev, "failed to get dsim phy\n");
 		ret = PTR_ERR(dsi->phy);
 		goto err_del_component;
 	}
 
 	dsi->irq = platform_get_irq(pdev, 0);
 	if (dsi->irq < 0) {
-		dev_err(&pdev->dev, "failed to request dsi irq resource\n");
+		dev_err(dev, "failed to request dsi irq resource\n");
 		ret = dsi->irq;
 		goto err_del_component;
 	}
 
 	irq_set_status_flags(dsi->irq, IRQ_NOAUTOEN);
-	ret = devm_request_threaded_irq(&pdev->dev, dsi->irq, NULL,
+	ret = devm_request_threaded_irq(dev, dsi->irq, NULL,
 					exynos_dsi_irq, IRQF_ONESHOT,
-					dev_name(&pdev->dev), dsi);
+					dev_name(dev), dsi);
 	if (ret) {
-		dev_err(&pdev->dev, "failed to request dsi irq\n");
+		dev_err(dev, "failed to request dsi irq\n");
 		goto err_del_component;
 	}
 
-	exynos_dsi_display.ctx = dsi;
+	platform_set_drvdata(pdev, &dsi->display);
 
-	platform_set_drvdata(pdev, &exynos_dsi_display);
-
-	ret = component_add(&pdev->dev, &exynos_dsi_component_ops);
+	ret = component_add(dev, &exynos_dsi_component_ops);
 	if (ret)
 		goto err_del_component;
 
 	return ret;
 
 err_del_component:
-	exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CONNECTOR);
+	exynos_drm_component_del(dev, EXYNOS_DEVICE_TYPE_CONNECTOR);
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_encoder.h b/drivers/gpu/drm/exynos/exynos_drm_encoder.h
index b7a1620a7e79..26305d8dd93a 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_encoder.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_encoder.h
@@ -14,8 +14,6 @@
 #ifndef _EXYNOS_DRM_ENCODER_H_
 #define _EXYNOS_DRM_ENCODER_H_
 
-struct exynos_drm_manager;
-
 void exynos_drm_encoder_setup(struct drm_device *dev);
 struct drm_encoder *exynos_drm_encoder_create(struct drm_device *dev,
 			struct exynos_drm_display *mgr,
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fb.c b/drivers/gpu/drm/exynos/exynos_drm_fb.c
index 65a22cad7b36..d346d1e6eda0 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fb.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fb.c
@@ -165,6 +165,7 @@ exynos_drm_framebuffer_init(struct drm_device *dev,
 
 	ret = drm_framebuffer_init(dev, &exynos_fb->fb, &exynos_drm_fb_funcs);
 	if (ret) {
+		kfree(exynos_fb);
 		DRM_ERROR("failed to initialize framebuffer\n");
 		return ERR_PTR(ret);
 	}
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fbdev.c b/drivers/gpu/drm/exynos/exynos_drm_fbdev.c
index 32e63f60e1d1..e12ea90c6237 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fbdev.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fbdev.c
@@ -123,6 +123,7 @@ static int exynos_drm_fbdev_update(struct drm_fb_helper *helper,
 
 	fbi->screen_base = buffer->kvaddr + offset;
 	fbi->screen_size = size;
+	fbi->fix.smem_len = size;
 
 	return 0;
 }
@@ -353,9 +354,6 @@ void exynos_drm_fbdev_fini(struct drm_device *dev)
 
 	fbdev = to_exynos_fbdev(private->fb_helper);
 
-	if (fbdev->exynos_gem_obj)
-		exynos_drm_gem_destroy(fbdev->exynos_gem_obj);
-
 	exynos_drm_fbdev_destroy(dev, private->fb_helper);
 	kfree(fbdev);
 	private->fb_helper = NULL;
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimc.c b/drivers/gpu/drm/exynos/exynos_drm_fimc.c
index ec7cc9ea50df..835b6af00970 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fimc.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fimc.c
@@ -336,9 +336,6 @@ static bool fimc_check_ovf(struct fimc_context *ctx)
 		fimc_set_bits(ctx, EXYNOS_CIWDOFST,
 			EXYNOS_CIWDOFST_CLROVFIY | EXYNOS_CIWDOFST_CLROVFICB |
 			EXYNOS_CIWDOFST_CLROVFICR);
-		fimc_clear_bits(ctx, EXYNOS_CIWDOFST,
-			EXYNOS_CIWDOFST_CLROVFIY | EXYNOS_CIWDOFST_CLROVFICB |
-			EXYNOS_CIWDOFST_CLROVFICR);
 
 		dev_err(ippdrv->dev, "occurred overflow at %d, status 0x%x.\n",
 			ctx->id, status);
@@ -718,24 +715,24 @@ static int fimc_src_set_addr(struct device *dev,
 	case IPP_BUF_ENQUEUE:
 		config = &property->config[EXYNOS_DRM_OPS_SRC];
 		fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_Y],
-			EXYNOS_CIIYSA(buf_id));
+			EXYNOS_CIIYSA0);
 
 		if (config->fmt == DRM_FORMAT_YVU420) {
 			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CR],
-				EXYNOS_CIICBSA(buf_id));
+				EXYNOS_CIICBSA0);
 			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CB],
-				EXYNOS_CIICRSA(buf_id));
+				EXYNOS_CIICRSA0);
 		} else {
 			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CB],
-				EXYNOS_CIICBSA(buf_id));
+				EXYNOS_CIICBSA0);
 			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CR],
-				EXYNOS_CIICRSA(buf_id));
+				EXYNOS_CIICRSA0);
 		}
 		break;
 	case IPP_BUF_DEQUEUE:
-		fimc_write(ctx, 0x0, EXYNOS_CIIYSA(buf_id));
-		fimc_write(ctx, 0x0, EXYNOS_CIICBSA(buf_id));
-		fimc_write(ctx, 0x0, EXYNOS_CIICRSA(buf_id));
+		fimc_write(ctx, 0x0, EXYNOS_CIIYSA0);
+		fimc_write(ctx, 0x0, EXYNOS_CIICBSA0);
+		fimc_write(ctx, 0x0, EXYNOS_CIICRSA0);
 		break;
 	default:
 		/* bypass */
@@ -1122,67 +1119,34 @@ static int fimc_dst_set_size(struct device *dev, int swap,
 	return 0;
 }
 
-static int fimc_dst_get_buf_count(struct fimc_context *ctx)
-{
-	u32 cfg, buf_num;
-
-	cfg = fimc_read(ctx, EXYNOS_CIFCNTSEQ);
-
-	buf_num = hweight32(cfg);
-
-	DRM_DEBUG_KMS("buf_num[%d]\n", buf_num);
-
-	return buf_num;
-}
-
-static int fimc_dst_set_buf_seq(struct fimc_context *ctx, u32 buf_id,
+static void fimc_dst_set_buf_seq(struct fimc_context *ctx, u32 buf_id,
 		enum drm_exynos_ipp_buf_type buf_type)
 {
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	bool enable;
-	u32 cfg;
-	u32 mask = 0x00000001 << buf_id;
-	int ret = 0;
 	unsigned long flags;
+	u32 buf_num;
+	u32 cfg;
 
 	DRM_DEBUG_KMS("buf_id[%d]buf_type[%d]\n", buf_id, buf_type);
 
 	spin_lock_irqsave(&ctx->lock, flags);
 
-	/* mask register set */
 	cfg = fimc_read(ctx, EXYNOS_CIFCNTSEQ);
 
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		enable = true;
-		break;
-	case IPP_BUF_DEQUEUE:
-		enable = false;
-		break;
-	default:
-		dev_err(ippdrv->dev, "invalid buf ctrl parameter.\n");
-		ret =  -EINVAL;
-		goto err_unlock;
-	}
+	if (buf_type == IPP_BUF_ENQUEUE)
+		cfg |= (1 << buf_id);
+	else
+		cfg &= ~(1 << buf_id);
 
-	/* sequence id */
-	cfg &= ~mask;
-	cfg |= (enable << buf_id);
 	fimc_write(ctx, cfg, EXYNOS_CIFCNTSEQ);
 
-	/* interrupt enable */
-	if (buf_type == IPP_BUF_ENQUEUE &&
-	    fimc_dst_get_buf_count(ctx) >= FIMC_BUF_START)
-		fimc_mask_irq(ctx, true);
+	buf_num = hweight32(cfg);
 
-	/* interrupt disable */
-	if (buf_type == IPP_BUF_DEQUEUE &&
-	    fimc_dst_get_buf_count(ctx) <= FIMC_BUF_STOP)
+	if (buf_type == IPP_BUF_ENQUEUE && buf_num >= FIMC_BUF_START)
+		fimc_mask_irq(ctx, true);
+	else if (buf_type == IPP_BUF_DEQUEUE && buf_num <= FIMC_BUF_STOP)
 		fimc_mask_irq(ctx, false);
 
-err_unlock:
 	spin_unlock_irqrestore(&ctx->lock, flags);
-	return ret;
 }
 
 static int fimc_dst_set_addr(struct device *dev,
@@ -1240,7 +1204,9 @@ static int fimc_dst_set_addr(struct device *dev,
 		break;
 	}
 
-	return fimc_dst_set_buf_seq(ctx, buf_id, buf_type);
+	fimc_dst_set_buf_seq(ctx, buf_id, buf_type);
+
+	return 0;
 }
 
 static struct exynos_drm_ipp_ops fimc_dst_ops = {
@@ -1291,14 +1257,11 @@ static irqreturn_t fimc_irq_handler(int irq, void *dev_id)
 
 	DRM_DEBUG_KMS("buf_id[%d]\n", buf_id);
 
-	if (fimc_dst_set_buf_seq(ctx, buf_id, IPP_BUF_DEQUEUE) < 0) {
-		DRM_ERROR("failed to dequeue.\n");
-		return IRQ_HANDLED;
-	}
+	fimc_dst_set_buf_seq(ctx, buf_id, IPP_BUF_DEQUEUE);
 
 	event_work->ippdrv = ippdrv;
 	event_work->buf_id[EXYNOS_DRM_OPS_DST] = buf_id;
-	queue_work(ippdrv->event_workq, (struct work_struct *)event_work);
+	queue_work(ippdrv->event_workq, &event_work->work);
 
 	return IRQ_HANDLED;
 }
@@ -1590,11 +1553,8 @@ static int fimc_ippdrv_start(struct device *dev, enum drm_exynos_ipp_cmd cmd)
 
 	fimc_clear_bits(ctx, EXYNOS_CIOCTRL, EXYNOS_CIOCTRL_WEAVE_MASK);
 
-	if (cmd == IPP_CMD_M2M) {
-		fimc_set_bits(ctx, EXYNOS_MSCTRL, EXYNOS_MSCTRL_ENVID);
-
+	if (cmd == IPP_CMD_M2M)
 		fimc_set_bits(ctx, EXYNOS_MSCTRL, EXYNOS_MSCTRL_ENVID);
-	}
 
 	return 0;
 }
@@ -1857,7 +1817,7 @@ static int fimc_resume(struct device *dev)
 }
 #endif
 
-#ifdef CONFIG_PM_RUNTIME
+#ifdef CONFIG_PM
 static int fimc_runtime_suspend(struct device *dev)
 {
 	struct fimc_context *ctx = get_fimc_context(dev);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimd.c b/drivers/gpu/drm/exynos/exynos_drm_fimd.c
index 5d09e33fef87..e5810d13bf9c 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fimd.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fimd.c
@@ -84,8 +84,6 @@
 /* FIMD has totally five hardware windows. */
 #define WINDOWS_NR	5
 
-#define get_fimd_manager(mgr)	platform_get_drvdata(to_platform_device(dev))
-
 struct fimd_driver_data {
 	unsigned int timing_base;
 	unsigned int lcdblk_offset;
@@ -96,6 +94,7 @@ struct fimd_driver_data {
 	unsigned int has_clksel:1;
 	unsigned int has_limited_fmt:1;
 	unsigned int has_vidoutcon:1;
+	unsigned int has_vtsel:1;
 };
 
 static struct fimd_driver_data s3c64xx_fimd_driver_data = {
@@ -104,12 +103,31 @@ static struct fimd_driver_data s3c64xx_fimd_driver_data = {
 	.has_limited_fmt = 1,
 };
 
+static struct fimd_driver_data exynos3_fimd_driver_data = {
+	.timing_base = 0x20000,
+	.lcdblk_offset = 0x210,
+	.lcdblk_bypass_shift = 1,
+	.has_shadowcon = 1,
+	.has_vidoutcon = 1,
+};
+
 static struct fimd_driver_data exynos4_fimd_driver_data = {
 	.timing_base = 0x0,
 	.lcdblk_offset = 0x210,
 	.lcdblk_vt_shift = 10,
 	.lcdblk_bypass_shift = 1,
 	.has_shadowcon = 1,
+	.has_vtsel = 1,
+};
+
+static struct fimd_driver_data exynos4415_fimd_driver_data = {
+	.timing_base = 0x20000,
+	.lcdblk_offset = 0x210,
+	.lcdblk_vt_shift = 10,
+	.lcdblk_bypass_shift = 1,
+	.has_shadowcon = 1,
+	.has_vidoutcon = 1,
+	.has_vtsel = 1,
 };
 
 static struct fimd_driver_data exynos5_fimd_driver_data = {
@@ -119,6 +137,7 @@ static struct fimd_driver_data exynos5_fimd_driver_data = {
 	.lcdblk_bypass_shift = 15,
 	.has_shadowcon = 1,
 	.has_vidoutcon = 1,
+	.has_vtsel = 1,
 };
 
 struct fimd_win_data {
@@ -138,6 +157,7 @@ struct fimd_win_data {
 };
 
 struct fimd_context {
+	struct exynos_drm_manager	manager;
 	struct device			*dev;
 	struct drm_device		*drm_dev;
 	struct clk			*bus_clk;
@@ -165,11 +185,20 @@ struct fimd_context {
 	struct exynos_drm_display *display;
 };
 
+static inline struct fimd_context *mgr_to_fimd(struct exynos_drm_manager *mgr)
+{
+	return container_of(mgr, struct fimd_context, manager);
+}
+
 static const struct of_device_id fimd_driver_dt_match[] = {
 	{ .compatible = "samsung,s3c6400-fimd",
 	  .data = &s3c64xx_fimd_driver_data },
+	{ .compatible = "samsung,exynos3250-fimd",
+	  .data = &exynos3_fimd_driver_data },
 	{ .compatible = "samsung,exynos4210-fimd",
 	  .data = &exynos4_fimd_driver_data },
+	{ .compatible = "samsung,exynos4415-fimd",
+	  .data = &exynos4415_fimd_driver_data },
 	{ .compatible = "samsung,exynos5250-fimd",
 	  .data = &exynos5_fimd_driver_data },
 	{},
@@ -187,7 +216,7 @@ static inline struct fimd_driver_data *drm_fimd_get_driver_data(
 
 static void fimd_wait_for_vblank(struct exynos_drm_manager *mgr)
 {
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = mgr_to_fimd(mgr);
 
 	if (ctx->suspended)
 		return;
@@ -204,56 +233,74 @@ static void fimd_wait_for_vblank(struct exynos_drm_manager *mgr)
 		DRM_DEBUG_KMS("vblank wait timed out.\n");
 }
 
+static void fimd_enable_video_output(struct fimd_context *ctx, int win,
+					bool enable)
+{
+	u32 val = readl(ctx->regs + WINCON(win));
+
+	if (enable)
+		val |= WINCONx_ENWIN;
+	else
+		val &= ~WINCONx_ENWIN;
+
+	writel(val, ctx->regs + WINCON(win));
+}
+
+static void fimd_enable_shadow_channel_path(struct fimd_context *ctx, int win,
+						bool enable)
+{
+	u32 val = readl(ctx->regs + SHADOWCON);
+
+	if (enable)
+		val |= SHADOWCON_CHx_ENABLE(win);
+	else
+		val &= ~SHADOWCON_CHx_ENABLE(win);
+
+	writel(val, ctx->regs + SHADOWCON);
+}
 
 static void fimd_clear_channel(struct exynos_drm_manager *mgr)
 {
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = mgr_to_fimd(mgr);
 	int win, ch_enabled = 0;
 
 	DRM_DEBUG_KMS("%s\n", __FILE__);
 
 	/* Check if any channel is enabled. */
 	for (win = 0; win < WINDOWS_NR; win++) {
-		u32 val = readl(ctx->regs + SHADOWCON);
-		if (val & SHADOWCON_CHx_ENABLE(win)) {
-			val &= ~SHADOWCON_CHx_ENABLE(win);
-			writel(val, ctx->regs + SHADOWCON);
+		u32 val = readl(ctx->regs + WINCON(win));
+
+		if (val & WINCONx_ENWIN) {
+			fimd_enable_video_output(ctx, win, false);
+
+			if (ctx->driver_data->has_shadowcon)
+				fimd_enable_shadow_channel_path(ctx, win,
+								false);
+
 			ch_enabled = 1;
 		}
 	}
 
 	/* Wait for vsync, as disable channel takes effect at next vsync */
-	if (ch_enabled)
+	if (ch_enabled) {
+		unsigned int state = ctx->suspended;
+
+		ctx->suspended = 0;
 		fimd_wait_for_vblank(mgr);
+		ctx->suspended = state;
+	}
 }
 
 static int fimd_mgr_initialize(struct exynos_drm_manager *mgr,
 			struct drm_device *drm_dev)
 {
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = mgr_to_fimd(mgr);
 	struct exynos_drm_private *priv;
 	priv = drm_dev->dev_private;
 
 	mgr->drm_dev = ctx->drm_dev = drm_dev;
 	mgr->pipe = ctx->pipe = priv->pipe++;
 
-	/*
-	 * enable drm irq mode.
-	 * - with irq_enabled = true, we can use the vblank feature.
-	 *
-	 * P.S. note that we wouldn't use drm irq handler but
-	 *	just specific driver own one instead because
-	 *	drm framework supports only one irq handler.
-	 */
-	drm_dev->irq_enabled = true;
-
-	/*
-	 * with vblank_disable_allowed = true, vblank interrupt will be disabled
-	 * by drm timer once a current process gives up ownership of
-	 * vblank event.(after drm_vblank_put function is called)
-	 */
-	drm_dev->vblank_disable_allowed = true;
-
 	/* attach this sub driver to iommu mapping if supported. */
 	if (is_drm_iommu_supported(ctx->drm_dev)) {
 		/*
@@ -269,7 +316,7 @@ static int fimd_mgr_initialize(struct exynos_drm_manager *mgr,
 
 static void fimd_mgr_remove(struct exynos_drm_manager *mgr)
 {
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = mgr_to_fimd(mgr);
 
 	/* detach this sub driver from iommu mapping if supported. */
 	if (is_drm_iommu_supported(ctx->drm_dev))
@@ -309,14 +356,14 @@ static bool fimd_mode_fixup(struct exynos_drm_manager *mgr,
 static void fimd_mode_set(struct exynos_drm_manager *mgr,
 		const struct drm_display_mode *in_mode)
 {
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = mgr_to_fimd(mgr);
 
 	drm_mode_copy(&ctx->mode, in_mode);
 }
 
 static void fimd_commit(struct exynos_drm_manager *mgr)
 {
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = mgr_to_fimd(mgr);
 	struct drm_display_mode *mode = &ctx->mode;
 	struct fimd_driver_data *driver_data = ctx->driver_data;
 	void *timing_base = ctx->regs + driver_data->timing_base;
@@ -337,7 +384,8 @@ static void fimd_commit(struct exynos_drm_manager *mgr)
 		writel(0, timing_base + I80IFCONFBx(0));
 
 		/* set video type selection to I80 interface */
-		if (ctx->sysreg && regmap_update_bits(ctx->sysreg,
+		if (driver_data->has_vtsel && ctx->sysreg &&
+				regmap_update_bits(ctx->sysreg,
 					driver_data->lcdblk_offset,
 					0x3 << driver_data->lcdblk_vt_shift,
 					0x1 << driver_data->lcdblk_vt_shift)) {
@@ -415,7 +463,7 @@ static void fimd_commit(struct exynos_drm_manager *mgr)
 
 static int fimd_enable_vblank(struct exynos_drm_manager *mgr)
 {
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = mgr_to_fimd(mgr);
 	u32 val;
 
 	if (ctx->suspended)
@@ -425,12 +473,19 @@ static int fimd_enable_vblank(struct exynos_drm_manager *mgr)
 		val = readl(ctx->regs + VIDINTCON0);
 
 		val |= VIDINTCON0_INT_ENABLE;
-		val |= VIDINTCON0_INT_FRAME;
 
-		val &= ~VIDINTCON0_FRAMESEL0_MASK;
-		val |= VIDINTCON0_FRAMESEL0_VSYNC;
-		val &= ~VIDINTCON0_FRAMESEL1_MASK;
-		val |= VIDINTCON0_FRAMESEL1_NONE;
+		if (ctx->i80_if) {
+			val |= VIDINTCON0_INT_I80IFDONE;
+			val |= VIDINTCON0_INT_SYSMAINCON;
+			val &= ~VIDINTCON0_INT_SYSSUBCON;
+		} else {
+			val |= VIDINTCON0_INT_FRAME;
+
+			val &= ~VIDINTCON0_FRAMESEL0_MASK;
+			val |= VIDINTCON0_FRAMESEL0_VSYNC;
+			val &= ~VIDINTCON0_FRAMESEL1_MASK;
+			val |= VIDINTCON0_FRAMESEL1_NONE;
+		}
 
 		writel(val, ctx->regs + VIDINTCON0);
 	}
@@ -440,7 +495,7 @@ static int fimd_enable_vblank(struct exynos_drm_manager *mgr)
 
 static void fimd_disable_vblank(struct exynos_drm_manager *mgr)
 {
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = mgr_to_fimd(mgr);
 	u32 val;
 
 	if (ctx->suspended)
@@ -449,9 +504,15 @@ static void fimd_disable_vblank(struct exynos_drm_manager *mgr)
 	if (test_and_clear_bit(0, &ctx->irq_flags)) {
 		val = readl(ctx->regs + VIDINTCON0);
 
-		val &= ~VIDINTCON0_INT_FRAME;
 		val &= ~VIDINTCON0_INT_ENABLE;
 
+		if (ctx->i80_if) {
+			val &= ~VIDINTCON0_INT_I80IFDONE;
+			val &= ~VIDINTCON0_INT_SYSMAINCON;
+			val &= ~VIDINTCON0_INT_SYSSUBCON;
+		} else
+			val &= ~VIDINTCON0_INT_FRAME;
+
 		writel(val, ctx->regs + VIDINTCON0);
 	}
 }
@@ -459,7 +520,7 @@ static void fimd_disable_vblank(struct exynos_drm_manager *mgr)
 static void fimd_win_mode_set(struct exynos_drm_manager *mgr,
 			struct exynos_drm_overlay *overlay)
 {
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = mgr_to_fimd(mgr);
 	struct fimd_win_data *win_data;
 	int win;
 	unsigned long offset;
@@ -617,7 +678,7 @@ static void fimd_shadow_protect_win(struct fimd_context *ctx,
 
 static void fimd_win_commit(struct exynos_drm_manager *mgr, int zpos)
 {
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = mgr_to_fimd(mgr);
 	struct fimd_win_data *win_data;
 	int win = zpos;
 	unsigned long val, alpha, size;
@@ -724,20 +785,14 @@ static void fimd_win_commit(struct exynos_drm_manager *mgr, int zpos)
 	if (win != 0)
 		fimd_win_set_colkey(ctx, win);
 
-	/* wincon */
-	val = readl(ctx->regs + WINCON(win));
-	val |= WINCONx_ENWIN;
-	writel(val, ctx->regs + WINCON(win));
+	fimd_enable_video_output(ctx, win, true);
+
+	if (ctx->driver_data->has_shadowcon)
+		fimd_enable_shadow_channel_path(ctx, win, true);
 
 	/* Enable DMA channel and unprotect windows */
 	fimd_shadow_protect_win(ctx, win, false);
 
-	if (ctx->driver_data->has_shadowcon) {
-		val = readl(ctx->regs + SHADOWCON);
-		val |= SHADOWCON_CHx_ENABLE(win);
-		writel(val, ctx->regs + SHADOWCON);
-	}
-
 	win_data->enabled = true;
 
 	if (ctx->i80_if)
@@ -746,10 +801,9 @@ static void fimd_win_commit(struct exynos_drm_manager *mgr, int zpos)
 
 static void fimd_win_disable(struct exynos_drm_manager *mgr, int zpos)
 {
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = mgr_to_fimd(mgr);
 	struct fimd_win_data *win_data;
 	int win = zpos;
-	u32 val;
 
 	if (win == DEFAULT_ZPOS)
 		win = ctx->default_win;
@@ -768,18 +822,12 @@ static void fimd_win_disable(struct exynos_drm_manager *mgr, int zpos)
 	/* protect windows */
 	fimd_shadow_protect_win(ctx, win, true);
 
-	/* wincon */
-	val = readl(ctx->regs + WINCON(win));
-	val &= ~WINCONx_ENWIN;
-	writel(val, ctx->regs + WINCON(win));
+	fimd_enable_video_output(ctx, win, false);
 
-	/* unprotect windows */
-	if (ctx->driver_data->has_shadowcon) {
-		val = readl(ctx->regs + SHADOWCON);
-		val &= ~SHADOWCON_CHx_ENABLE(win);
-		writel(val, ctx->regs + SHADOWCON);
-	}
+	if (ctx->driver_data->has_shadowcon)
+		fimd_enable_shadow_channel_path(ctx, win, false);
 
+	/* unprotect windows */
 	fimd_shadow_protect_win(ctx, win, false);
 
 	win_data->enabled = false;
@@ -787,7 +835,7 @@ static void fimd_win_disable(struct exynos_drm_manager *mgr, int zpos)
 
 static void fimd_window_suspend(struct exynos_drm_manager *mgr)
 {
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = mgr_to_fimd(mgr);
 	struct fimd_win_data *win_data;
 	int i;
 
@@ -797,12 +845,11 @@ static void fimd_window_suspend(struct exynos_drm_manager *mgr)
 		if (win_data->enabled)
 			fimd_win_disable(mgr, i);
 	}
-	fimd_wait_for_vblank(mgr);
 }
 
 static void fimd_window_resume(struct exynos_drm_manager *mgr)
 {
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = mgr_to_fimd(mgr);
 	struct fimd_win_data *win_data;
 	int i;
 
@@ -815,7 +862,7 @@ static void fimd_window_resume(struct exynos_drm_manager *mgr)
 
 static void fimd_apply(struct exynos_drm_manager *mgr)
 {
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = mgr_to_fimd(mgr);
 	struct fimd_win_data *win_data;
 	int i;
 
@@ -832,7 +879,7 @@ static void fimd_apply(struct exynos_drm_manager *mgr)
 
 static int fimd_poweron(struct exynos_drm_manager *mgr)
 {
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = mgr_to_fimd(mgr);
 	int ret;
 
 	if (!ctx->suspended)
@@ -880,7 +927,7 @@ bus_clk_err:
 
 static int fimd_poweroff(struct exynos_drm_manager *mgr)
 {
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = mgr_to_fimd(mgr);
 
 	if (ctx->suspended)
 		return 0;
@@ -922,39 +969,41 @@ static void fimd_dpms(struct exynos_drm_manager *mgr, int mode)
 
 static void fimd_trigger(struct device *dev)
 {
-	struct exynos_drm_manager *mgr = get_fimd_manager(dev);
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = dev_get_drvdata(dev);
 	struct fimd_driver_data *driver_data = ctx->driver_data;
 	void *timing_base = ctx->regs + driver_data->timing_base;
 	u32 reg;
 
-	atomic_set(&ctx->triggering, 1);
+	 /*
+	  * Skips triggering if in triggering state, because multiple triggering
+	  * requests can cause panel reset.
+	  */
+	if (atomic_read(&ctx->triggering))
+		return;
 
-	reg = readl(ctx->regs + VIDINTCON0);
-	reg |= (VIDINTCON0_INT_ENABLE | VIDINTCON0_INT_I80IFDONE |
-						VIDINTCON0_INT_SYSMAINCON);
-	writel(reg, ctx->regs + VIDINTCON0);
+	/* Enters triggering mode */
+	atomic_set(&ctx->triggering, 1);
 
 	reg = readl(timing_base + TRIGCON);
 	reg |= (TRGMODE_I80_RGB_ENABLE_I80 | SWTRGCMD_I80_RGB_ENABLE);
 	writel(reg, timing_base + TRIGCON);
+
+	/*
+	 * Exits triggering mode if vblank is not enabled yet, because when the
+	 * VIDINTCON0 register is not set, it can not exit from triggering mode.
+	 */
+	if (!test_bit(0, &ctx->irq_flags))
+		atomic_set(&ctx->triggering, 0);
 }
 
 static void fimd_te_handler(struct exynos_drm_manager *mgr)
 {
-	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_context *ctx = mgr_to_fimd(mgr);
 
 	/* Checks the crtc is detached already from encoder */
 	if (ctx->pipe < 0 || !ctx->drm_dev)
 		return;
 
-	 /*
-	 * Skips to trigger if in triggering state, because multiple triggering
-	 * requests can cause panel reset.
-	 */
-	if (atomic_read(&ctx->triggering))
-		return;
-
 	/*
 	 * If there is a page flip request, triggers and handles the page flip
 	 * event so that current fb can be updated into panel GRAM.
@@ -966,10 +1015,10 @@ static void fimd_te_handler(struct exynos_drm_manager *mgr)
 	if (atomic_read(&ctx->wait_vsync_event)) {
 		atomic_set(&ctx->wait_vsync_event, 0);
 		wake_up(&ctx->wait_vsync_queue);
-
-		if (!atomic_read(&ctx->triggering))
-			drm_handle_vblank(ctx->drm_dev, ctx->pipe);
 	}
+
+	if (test_bit(0, &ctx->irq_flags))
+		drm_handle_vblank(ctx->drm_dev, ctx->pipe);
 }
 
 static struct exynos_drm_manager_ops fimd_manager_ops = {
@@ -986,11 +1035,6 @@ static struct exynos_drm_manager_ops fimd_manager_ops = {
 	.te_handler = fimd_te_handler,
 };
 
-static struct exynos_drm_manager fimd_manager = {
-	.type = EXYNOS_DISPLAY_TYPE_LCD,
-	.ops = &fimd_manager_ops,
-};
-
 static irqreturn_t fimd_irq_handler(int irq, void *dev_id)
 {
 	struct fimd_context *ctx = (struct fimd_context *)dev_id;
@@ -1007,16 +1051,10 @@ static irqreturn_t fimd_irq_handler(int irq, void *dev_id)
 		goto out;
 
 	if (ctx->i80_if) {
-		/* unset I80 frame done interrupt */
-		val = readl(ctx->regs + VIDINTCON0);
-		val &= ~(VIDINTCON0_INT_I80IFDONE | VIDINTCON0_INT_SYSMAINCON);
-		writel(val, ctx->regs + VIDINTCON0);
+		exynos_drm_crtc_finish_pageflip(ctx->drm_dev, ctx->pipe);
 
-		/* exit triggering mode */
+		/* Exits triggering mode */
 		atomic_set(&ctx->triggering, 0);
-
-		drm_handle_vblank(ctx->drm_dev, ctx->pipe);
-		exynos_drm_crtc_finish_pageflip(ctx->drm_dev, ctx->pipe);
 	} else {
 		drm_handle_vblank(ctx->drm_dev, ctx->pipe);
 		exynos_drm_crtc_finish_pageflip(ctx->drm_dev, ctx->pipe);
@@ -1034,11 +1072,11 @@ out:
 
 static int fimd_bind(struct device *dev, struct device *master, void *data)
 {
-	struct fimd_context *ctx = fimd_manager.ctx;
+	struct fimd_context *ctx = dev_get_drvdata(dev);
 	struct drm_device *drm_dev = data;
 
-	fimd_mgr_initialize(&fimd_manager, drm_dev);
-	exynos_drm_crtc_create(&fimd_manager);
+	fimd_mgr_initialize(&ctx->manager, drm_dev);
+	exynos_drm_crtc_create(&ctx->manager);
 	if (ctx->display)
 		exynos_drm_create_enc_conn(drm_dev, ctx->display);
 
@@ -1049,18 +1087,14 @@ static int fimd_bind(struct device *dev, struct device *master, void *data)
 static void fimd_unbind(struct device *dev, struct device *master,
 			void *data)
 {
-	struct exynos_drm_manager *mgr = dev_get_drvdata(dev);
-	struct fimd_context *ctx = fimd_manager.ctx;
-	struct drm_crtc *crtc = mgr->crtc;
+	struct fimd_context *ctx = dev_get_drvdata(dev);
 
-	fimd_dpms(mgr, DRM_MODE_DPMS_OFF);
+	fimd_dpms(&ctx->manager, DRM_MODE_DPMS_OFF);
 
 	if (ctx->display)
-		exynos_dpi_remove(dev);
-
-	fimd_mgr_remove(mgr);
+		exynos_dpi_remove(ctx->display);
 
-	crtc->funcs->destroy(crtc);
+	fimd_mgr_remove(&ctx->manager);
 }
 
 static const struct component_ops fimd_component_ops = {
@@ -1076,21 +1110,20 @@ static int fimd_probe(struct platform_device *pdev)
 	struct resource *res;
 	int ret = -EINVAL;
 
-	ret = exynos_drm_component_add(&pdev->dev, EXYNOS_DEVICE_TYPE_CRTC,
-					fimd_manager.type);
-	if (ret)
-		return ret;
-
-	if (!dev->of_node) {
-		ret = -ENODEV;
-		goto err_del_component;
-	}
+	if (!dev->of_node)
+		return -ENODEV;
 
 	ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
-	if (!ctx) {
-		ret = -ENOMEM;
-		goto err_del_component;
-	}
+	if (!ctx)
+		return -ENOMEM;
+
+	ctx->manager.type = EXYNOS_DISPLAY_TYPE_LCD;
+	ctx->manager.ops = &fimd_manager_ops;
+
+	ret = exynos_drm_component_add(dev, EXYNOS_DEVICE_TYPE_CRTC,
+				       ctx->manager.type);
+	if (ret)
+		return ret;
 
 	ctx->dev = dev;
 	ctx->suspended = true;
@@ -1179,27 +1212,27 @@ static int fimd_probe(struct platform_device *pdev)
 	init_waitqueue_head(&ctx->wait_vsync_queue);
 	atomic_set(&ctx->wait_vsync_event, 0);
 
-	platform_set_drvdata(pdev, &fimd_manager);
-
-	fimd_manager.ctx = ctx;
+	platform_set_drvdata(pdev, ctx);
 
 	ctx->display = exynos_dpi_probe(dev);
-	if (IS_ERR(ctx->display))
-		return PTR_ERR(ctx->display);
+	if (IS_ERR(ctx->display)) {
+		ret = PTR_ERR(ctx->display);
+		goto err_del_component;
+	}
 
-	pm_runtime_enable(&pdev->dev);
+	pm_runtime_enable(dev);
 
-	ret = component_add(&pdev->dev, &fimd_component_ops);
+	ret = component_add(dev, &fimd_component_ops);
 	if (ret)
 		goto err_disable_pm_runtime;
 
 	return ret;
 
 err_disable_pm_runtime:
-	pm_runtime_disable(&pdev->dev);
+	pm_runtime_disable(dev);
 
 err_del_component:
-	exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CRTC);
+	exynos_drm_component_del(dev, EXYNOS_DEVICE_TYPE_CRTC);
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_g2d.c b/drivers/gpu/drm/exynos/exynos_drm_g2d.c
index df7a77d3eff8..81a250830808 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_g2d.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_g2d.c
@@ -302,9 +302,12 @@ static void g2d_fini_cmdlist(struct g2d_data *g2d)
 	struct exynos_drm_subdrv *subdrv = &g2d->subdrv;
 
 	kfree(g2d->cmdlist_node);
-	dma_free_attrs(subdrv->drm_dev->dev, G2D_CMDLIST_POOL_SIZE,
-			g2d->cmdlist_pool_virt,
-			g2d->cmdlist_pool, &g2d->cmdlist_dma_attrs);
+
+	if (g2d->cmdlist_pool_virt && g2d->cmdlist_pool) {
+		dma_free_attrs(subdrv->drm_dev->dev, G2D_CMDLIST_POOL_SIZE,
+				g2d->cmdlist_pool_virt,
+				g2d->cmdlist_pool, &g2d->cmdlist_dma_attrs);
+	}
 }
 
 static struct g2d_cmdlist_node *g2d_get_cmdlist(struct g2d_data *g2d)
@@ -1537,7 +1540,7 @@ static int g2d_resume(struct device *dev)
 }
 #endif
 
-#ifdef CONFIG_PM_RUNTIME
+#ifdef CONFIG_PM
 static int g2d_runtime_suspend(struct device *dev)
 {
 	struct g2d_data *g2d = dev_get_drvdata(dev);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.c b/drivers/gpu/drm/exynos/exynos_drm_gem.c
index 15db80138382..0d5b9698d384 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gem.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_gem.c
@@ -318,40 +318,16 @@ void exynos_drm_gem_put_dma_addr(struct drm_device *dev,
 	drm_gem_object_unreference_unlocked(obj);
 }
 
-int exynos_drm_gem_map_offset_ioctl(struct drm_device *dev, void *data,
-				    struct drm_file *file_priv)
-{
-	struct drm_exynos_gem_map_off *args = data;
-
-	DRM_DEBUG_KMS("handle = 0x%x, offset = 0x%lx\n",
-			args->handle, (unsigned long)args->offset);
-
-	if (!(dev->driver->driver_features & DRIVER_GEM)) {
-		DRM_ERROR("does not support GEM.\n");
-		return -ENODEV;
-	}
-
-	return exynos_drm_gem_dumb_map_offset(file_priv, dev, args->handle,
-			&args->offset);
-}
-
-int exynos_drm_gem_mmap_buffer(struct file *filp,
+int exynos_drm_gem_mmap_buffer(struct exynos_drm_gem_obj *exynos_gem_obj,
 				      struct vm_area_struct *vma)
 {
-	struct drm_gem_object *obj = filp->private_data;
-	struct exynos_drm_gem_obj *exynos_gem_obj = to_exynos_gem_obj(obj);
-	struct drm_device *drm_dev = obj->dev;
+	struct drm_device *drm_dev = exynos_gem_obj->base.dev;
 	struct exynos_drm_gem_buf *buffer;
 	unsigned long vm_size;
 	int ret;
 
-	WARN_ON(!mutex_is_locked(&obj->dev->struct_mutex));
-
-	vma->vm_flags |= VM_IO | VM_DONTEXPAND | VM_DONTDUMP;
-	vma->vm_private_data = obj;
-	vma->vm_ops = drm_dev->driver->gem_vm_ops;
-
-	update_vm_cache_attr(exynos_gem_obj, vma);
+	vma->vm_flags &= ~VM_PFNMAP;
+	vma->vm_pgoff = 0;
 
 	vm_size = vma->vm_end - vma->vm_start;
 
@@ -373,60 +349,6 @@ int exynos_drm_gem_mmap_buffer(struct file *filp,
 		return ret;
 	}
 
-	/*
-	 * take a reference to this mapping of the object. And this reference
-	 * is unreferenced by the corresponding vm_close call.
-	 */
-	drm_gem_object_reference(obj);
-
-	drm_vm_open_locked(drm_dev, vma);
-
-	return 0;
-}
-
-int exynos_drm_gem_mmap_ioctl(struct drm_device *dev, void *data,
-			      struct drm_file *file_priv)
-{
-	struct drm_exynos_file_private *exynos_file_priv;
-	struct drm_exynos_gem_mmap *args = data;
-	struct drm_gem_object *obj;
-	struct file *anon_filp;
-	unsigned long addr;
-
-	if (!(dev->driver->driver_features & DRIVER_GEM)) {
-		DRM_ERROR("does not support GEM.\n");
-		return -ENODEV;
-	}
-
-	mutex_lock(&dev->struct_mutex);
-
-	obj = drm_gem_object_lookup(dev, file_priv, args->handle);
-	if (!obj) {
-		DRM_ERROR("failed to lookup gem object.\n");
-		mutex_unlock(&dev->struct_mutex);
-		return -EINVAL;
-	}
-
-	exynos_file_priv = file_priv->driver_priv;
-	anon_filp = exynos_file_priv->anon_filp;
-	anon_filp->private_data = obj;
-
-	addr = vm_mmap(anon_filp, 0, args->size, PROT_READ | PROT_WRITE,
-			MAP_SHARED, 0);
-
-	drm_gem_object_unreference(obj);
-
-	if (IS_ERR_VALUE(addr)) {
-		mutex_unlock(&dev->struct_mutex);
-		return (int)addr;
-	}
-
-	mutex_unlock(&dev->struct_mutex);
-
-	args->mapped = addr;
-
-	DRM_DEBUG_KMS("mapped = 0x%lx\n", (unsigned long)args->mapped);
-
 	return 0;
 }
 
@@ -710,16 +632,20 @@ int exynos_drm_gem_mmap(struct file *filp, struct vm_area_struct *vma)
 	exynos_gem_obj = to_exynos_gem_obj(obj);
 
 	ret = check_gem_flags(exynos_gem_obj->flags);
-	if (ret) {
-		drm_gem_vm_close(vma);
-		drm_gem_free_mmap_offset(obj);
-		return ret;
-	}
-
-	vma->vm_flags &= ~VM_PFNMAP;
-	vma->vm_flags |= VM_MIXEDMAP;
+	if (ret)
+		goto err_close_vm;
 
 	update_vm_cache_attr(exynos_gem_obj, vma);
 
+	ret = exynos_drm_gem_mmap_buffer(exynos_gem_obj, vma);
+	if (ret)
+		goto err_close_vm;
+
+	return ret;
+
+err_close_vm:
+	drm_gem_vm_close(vma);
+	drm_gem_free_mmap_offset(obj);
+
 	return ret;
 }
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.h b/drivers/gpu/drm/exynos/exynos_drm_gem.h
index 1592c0ba7de8..ec58fe9c40df 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gem.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_gem.h
@@ -12,6 +12,8 @@
 #ifndef _EXYNOS_DRM_GEM_H_
 #define _EXYNOS_DRM_GEM_H_
 
+#include <drm/drm_gem.h>
+
 #define to_exynos_gem_obj(x)	container_of(x,\
 			struct exynos_drm_gem_obj, base)
 
@@ -111,20 +113,6 @@ void exynos_drm_gem_put_dma_addr(struct drm_device *dev,
 					unsigned int gem_handle,
 					struct drm_file *filp);
 
-/* get buffer offset to map to user space. */
-int exynos_drm_gem_map_offset_ioctl(struct drm_device *dev, void *data,
-				    struct drm_file *file_priv);
-
-/*
- * mmap the physically continuous memory that a gem object contains
- * to user space.
- */
-int exynos_drm_gem_mmap_ioctl(struct drm_device *dev, void *data,
-			      struct drm_file *file_priv);
-
-int exynos_drm_gem_mmap_buffer(struct file *filp,
-				      struct vm_area_struct *vma);
-
 /* map user space allocated by malloc to pages. */
 int exynos_drm_gem_userptr_ioctl(struct drm_device *dev, void *data,
 				      struct drm_file *file_priv);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gsc.c b/drivers/gpu/drm/exynos/exynos_drm_gsc.c
index 9e3ff1672965..0261468c8019 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gsc.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_gsc.c
@@ -1326,8 +1326,7 @@ static irqreturn_t gsc_irq_handler(int irq, void *dev_id)
 			buf_id[EXYNOS_DRM_OPS_SRC];
 		event_work->buf_id[EXYNOS_DRM_OPS_DST] =
 			buf_id[EXYNOS_DRM_OPS_DST];
-		queue_work(ippdrv->event_workq,
-			(struct work_struct *)event_work);
+		queue_work(ippdrv->event_workq, &event_work->work);
 	}
 
 	return IRQ_HANDLED;
@@ -1765,7 +1764,7 @@ static int gsc_resume(struct device *dev)
 }
 #endif
 
-#ifdef CONFIG_PM_RUNTIME
+#ifdef CONFIG_PM
 static int gsc_runtime_suspend(struct device *dev)
 {
 	struct gsc_context *ctx = get_gsc_context(dev);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_iommu.h b/drivers/gpu/drm/exynos/exynos_drm_iommu.h
index 72376d41c512..35d25889b476 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_iommu.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_iommu.h
@@ -40,7 +40,6 @@ static inline bool is_drm_iommu_supported(struct drm_device *drm_dev)
 
 #else
 
-struct dma_iommu_mapping;
 static inline int drm_create_iommu_mapping(struct drm_device *drm_dev)
 {
 	return 0;
diff --git a/drivers/gpu/drm/exynos/exynos_drm_ipp.c b/drivers/gpu/drm/exynos/exynos_drm_ipp.c
index c411399070d6..d5ad17dfc24d 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_ipp.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_ipp.c
@@ -75,7 +75,6 @@ struct drm_exynos_ipp_mem_node {
 	u32	prop_id;
 	u32	buf_id;
 	struct drm_exynos_ipp_buf_info	buf_info;
-	struct drm_file		*filp;
 };
 
 /*
@@ -319,44 +318,6 @@ static void ipp_print_property(struct drm_exynos_ipp_property *property,
 		sz->hsize, sz->vsize, config->flip, config->degree);
 }
 
-static int ipp_find_and_set_property(struct drm_exynos_ipp_property *property)
-{
-	struct exynos_drm_ippdrv *ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node;
-	u32 prop_id = property->prop_id;
-
-	DRM_DEBUG_KMS("prop_id[%d]\n", prop_id);
-
-	ippdrv = ipp_find_drv_by_handle(prop_id);
-	if (IS_ERR(ippdrv)) {
-		DRM_ERROR("failed to get ipp driver.\n");
-		return -EINVAL;
-	}
-
-	/*
-	 * Find command node using command list in ippdrv.
-	 * when we find this command no using prop_id.
-	 * return property information set in this command node.
-	 */
-	mutex_lock(&ippdrv->cmd_lock);
-	list_for_each_entry(c_node, &ippdrv->cmd_list, list) {
-		if ((c_node->property.prop_id == prop_id) &&
-		    (c_node->state == IPP_STATE_STOP)) {
-			mutex_unlock(&ippdrv->cmd_lock);
-			DRM_DEBUG_KMS("found cmd[%d]ippdrv[0x%x]\n",
-				property->cmd, (int)ippdrv);
-
-			c_node->property = *property;
-			return 0;
-		}
-	}
-	mutex_unlock(&ippdrv->cmd_lock);
-
-	DRM_ERROR("failed to search property.\n");
-
-	return -EINVAL;
-}
-
 static struct drm_exynos_ipp_cmd_work *ipp_create_cmd_work(void)
 {
 	struct drm_exynos_ipp_cmd_work *cmd_work;
@@ -392,6 +353,7 @@ int exynos_drm_ipp_set_property(struct drm_device *drm_dev, void *data,
 	struct drm_exynos_ipp_property *property = data;
 	struct exynos_drm_ippdrv *ippdrv;
 	struct drm_exynos_ipp_cmd_node *c_node;
+	u32 prop_id;
 	int ret, i;
 
 	if (!ctx) {
@@ -404,6 +366,8 @@ int exynos_drm_ipp_set_property(struct drm_device *drm_dev, void *data,
 		return -EINVAL;
 	}
 
+	prop_id = property->prop_id;
+
 	/*
 	 * This is log print for user application property.
 	 * user application set various property.
@@ -412,14 +376,24 @@ int exynos_drm_ipp_set_property(struct drm_device *drm_dev, void *data,
 		ipp_print_property(property, i);
 
 	/*
-	 * set property ioctl generated new prop_id.
-	 * but in this case already asigned prop_id using old set property.
-	 * e.g PAUSE state. this case supports find current prop_id and use it
-	 * instead of allocation.
+	 * In case prop_id is not zero try to set existing property.
 	 */
-	if (property->prop_id) {
-		DRM_DEBUG_KMS("prop_id[%d]\n", property->prop_id);
-		return ipp_find_and_set_property(property);
+	if (prop_id) {
+		c_node = ipp_find_obj(&ctx->prop_idr, &ctx->prop_lock, prop_id);
+
+		if (!c_node || c_node->filp != file) {
+			DRM_DEBUG_KMS("prop_id[%d] not found\n", prop_id);
+			return -EINVAL;
+		}
+
+		if (c_node->state != IPP_STATE_STOP) {
+			DRM_DEBUG_KMS("prop_id[%d] not stopped\n", prop_id);
+			return -EINVAL;
+		}
+
+		c_node->property = *property;
+
+		return 0;
 	}
 
 	/* find ipp driver using ipp id */
@@ -445,25 +419,28 @@ int exynos_drm_ipp_set_property(struct drm_device *drm_dev, void *data,
 		property->prop_id, property->cmd, (int)ippdrv);
 
 	/* stored property information and ippdrv in private data */
-	c_node->dev = dev;
 	c_node->property = *property;
 	c_node->state = IPP_STATE_IDLE;
+	c_node->filp = file;
 
 	c_node->start_work = ipp_create_cmd_work();
 	if (IS_ERR(c_node->start_work)) {
 		DRM_ERROR("failed to create start work.\n");
+		ret = PTR_ERR(c_node->start_work);
 		goto err_remove_id;
 	}
 
 	c_node->stop_work = ipp_create_cmd_work();
 	if (IS_ERR(c_node->stop_work)) {
 		DRM_ERROR("failed to create stop work.\n");
+		ret = PTR_ERR(c_node->stop_work);
 		goto err_free_start;
 	}
 
 	c_node->event_work = ipp_create_event_work();
 	if (IS_ERR(c_node->event_work)) {
 		DRM_ERROR("failed to create event work.\n");
+		ret = PTR_ERR(c_node->event_work);
 		goto err_free_stop;
 	}
 
@@ -499,105 +476,37 @@ err_clear:
 	return ret;
 }
 
-static void ipp_clean_cmd_node(struct ipp_context *ctx,
-				struct drm_exynos_ipp_cmd_node *c_node)
-{
-	/* delete list */
-	list_del(&c_node->list);
-
-	ipp_remove_id(&ctx->prop_idr, &ctx->prop_lock,
-			c_node->property.prop_id);
-
-	/* destroy mutex */
-	mutex_destroy(&c_node->lock);
-	mutex_destroy(&c_node->mem_lock);
-	mutex_destroy(&c_node->event_lock);
-
-	/* free command node */
-	kfree(c_node->start_work);
-	kfree(c_node->stop_work);
-	kfree(c_node->event_work);
-	kfree(c_node);
-}
-
-static bool ipp_check_mem_list(struct drm_exynos_ipp_cmd_node *c_node)
-{
-	switch (c_node->property.cmd) {
-	case IPP_CMD_WB:
-		return !list_empty(&c_node->mem_list[EXYNOS_DRM_OPS_DST]);
-	case IPP_CMD_OUTPUT:
-		return !list_empty(&c_node->mem_list[EXYNOS_DRM_OPS_SRC]);
-	case IPP_CMD_M2M:
-	default:
-		return !list_empty(&c_node->mem_list[EXYNOS_DRM_OPS_SRC]) &&
-		       !list_empty(&c_node->mem_list[EXYNOS_DRM_OPS_DST]);
-	}
-}
-
-static struct drm_exynos_ipp_mem_node
-		*ipp_find_mem_node(struct drm_exynos_ipp_cmd_node *c_node,
-		struct drm_exynos_ipp_queue_buf *qbuf)
-{
-	struct drm_exynos_ipp_mem_node *m_node;
-	struct list_head *head;
-	int count = 0;
-
-	DRM_DEBUG_KMS("buf_id[%d]\n", qbuf->buf_id);
-
-	/* source/destination memory list */
-	head = &c_node->mem_list[qbuf->ops_id];
-
-	/* find memory node from memory list */
-	list_for_each_entry(m_node, head, list) {
-		DRM_DEBUG_KMS("count[%d]m_node[0x%x]\n", count++, (int)m_node);
-
-		/* compare buffer id */
-		if (m_node->buf_id == qbuf->buf_id)
-			return m_node;
-	}
-
-	return NULL;
-}
-
-static int ipp_set_mem_node(struct exynos_drm_ippdrv *ippdrv,
+static int ipp_put_mem_node(struct drm_device *drm_dev,
 		struct drm_exynos_ipp_cmd_node *c_node,
 		struct drm_exynos_ipp_mem_node *m_node)
 {
-	struct exynos_drm_ipp_ops *ops = NULL;
-	int ret = 0;
+	int i;
 
 	DRM_DEBUG_KMS("node[0x%x]\n", (int)m_node);
 
 	if (!m_node) {
-		DRM_ERROR("invalid queue node.\n");
+		DRM_ERROR("invalid dequeue node.\n");
 		return -EFAULT;
 	}
 
 	DRM_DEBUG_KMS("ops_id[%d]\n", m_node->ops_id);
 
-	/* get operations callback */
-	ops = ippdrv->ops[m_node->ops_id];
-	if (!ops) {
-		DRM_ERROR("not support ops.\n");
-		return -EFAULT;
+	/* put gem buffer */
+	for_each_ipp_planar(i) {
+		unsigned long handle = m_node->buf_info.handles[i];
+		if (handle)
+			exynos_drm_gem_put_dma_addr(drm_dev, handle,
+							c_node->filp);
 	}
 
-	/* set address and enable irq */
-	if (ops->set_addr) {
-		ret = ops->set_addr(ippdrv->dev, &m_node->buf_info,
-			m_node->buf_id, IPP_BUF_ENQUEUE);
-		if (ret) {
-			DRM_ERROR("failed to set addr.\n");
-			return ret;
-		}
-	}
+	list_del(&m_node->list);
+	kfree(m_node);
 
-	return ret;
+	return 0;
 }
 
 static struct drm_exynos_ipp_mem_node
 		*ipp_get_mem_node(struct drm_device *drm_dev,
-		struct drm_file *file,
 		struct drm_exynos_ipp_cmd_node *c_node,
 		struct drm_exynos_ipp_queue_buf *qbuf)
 {
@@ -615,6 +524,7 @@ static struct drm_exynos_ipp_mem_node
 	m_node->ops_id = qbuf->ops_id;
 	m_node->prop_id = qbuf->prop_id;
 	m_node->buf_id = qbuf->buf_id;
+	INIT_LIST_HEAD(&m_node->list);
 
 	DRM_DEBUG_KMS("m_node[0x%x]ops_id[%d]\n", (int)m_node, qbuf->ops_id);
 	DRM_DEBUG_KMS("prop_id[%d]buf_id[%d]\n", qbuf->prop_id, m_node->buf_id);
@@ -627,10 +537,11 @@ static struct drm_exynos_ipp_mem_node
 			dma_addr_t *addr;
 
 			addr = exynos_drm_gem_get_dma_addr(drm_dev,
-					qbuf->handle[i], file);
+					qbuf->handle[i], c_node->filp);
 			if (IS_ERR(addr)) {
 				DRM_ERROR("failed to get addr.\n");
-				goto err_clear;
+				ipp_put_mem_node(drm_dev, c_node, m_node);
+				return ERR_PTR(-EFAULT);
 			}
 
 			buf_info->handles[i] = qbuf->handle[i];
@@ -640,46 +551,30 @@ static struct drm_exynos_ipp_mem_node
 		}
 	}
 
-	m_node->filp = file;
 	mutex_lock(&c_node->mem_lock);
 	list_add_tail(&m_node->list, &c_node->mem_list[qbuf->ops_id]);
 	mutex_unlock(&c_node->mem_lock);
 
 	return m_node;
-
-err_clear:
-	kfree(m_node);
-	return ERR_PTR(-EFAULT);
 }
 
-static int ipp_put_mem_node(struct drm_device *drm_dev,
-		struct drm_exynos_ipp_cmd_node *c_node,
-		struct drm_exynos_ipp_mem_node *m_node)
+static void ipp_clean_mem_nodes(struct drm_device *drm_dev,
+			       struct drm_exynos_ipp_cmd_node *c_node, int ops)
 {
-	int i;
-
-	DRM_DEBUG_KMS("node[0x%x]\n", (int)m_node);
+	struct drm_exynos_ipp_mem_node *m_node, *tm_node;
+	struct list_head *head = &c_node->mem_list[ops];
 
-	if (!m_node) {
-		DRM_ERROR("invalid dequeue node.\n");
-		return -EFAULT;
-	}
+	mutex_lock(&c_node->mem_lock);
 
-	DRM_DEBUG_KMS("ops_id[%d]\n", m_node->ops_id);
+	list_for_each_entry_safe(m_node, tm_node, head, list) {
+		int ret;
 
-	/* put gem buffer */
-	for_each_ipp_planar(i) {
-		unsigned long handle = m_node->buf_info.handles[i];
-		if (handle)
-			exynos_drm_gem_put_dma_addr(drm_dev, handle,
-							m_node->filp);
+		ret = ipp_put_mem_node(drm_dev, c_node, m_node);
+		if (ret)
+			DRM_ERROR("failed to put m_node.\n");
 	}
 
-	/* delete list in queue */
-	list_del(&m_node->list);
-	kfree(m_node);
-
-	return 0;
+	mutex_unlock(&c_node->mem_lock);
 }
 
 static void ipp_free_event(struct drm_pending_event *event)
@@ -688,7 +583,6 @@ static void ipp_free_event(struct drm_pending_event *event)
 }
 
 static int ipp_get_event(struct drm_device *drm_dev,
-		struct drm_file *file,
 		struct drm_exynos_ipp_cmd_node *c_node,
 		struct drm_exynos_ipp_queue_buf *qbuf)
 {
@@ -700,7 +594,7 @@ static int ipp_get_event(struct drm_device *drm_dev,
 	e = kzalloc(sizeof(*e), GFP_KERNEL);
 	if (!e) {
 		spin_lock_irqsave(&drm_dev->event_lock, flags);
-		file->event_space += sizeof(e->event);
+		c_node->filp->event_space += sizeof(e->event);
 		spin_unlock_irqrestore(&drm_dev->event_lock, flags);
 		return -ENOMEM;
 	}
@@ -712,7 +606,7 @@ static int ipp_get_event(struct drm_device *drm_dev,
 	e->event.prop_id = qbuf->prop_id;
 	e->event.buf_id[EXYNOS_DRM_OPS_DST] = qbuf->buf_id;
 	e->base.event = &e->event.base;
-	e->base.file_priv = file;
+	e->base.file_priv = c_node->filp;
 	e->base.destroy = ipp_free_event;
 	mutex_lock(&c_node->event_lock);
 	list_add_tail(&e->base.link, &c_node->event_list);
@@ -757,6 +651,115 @@ out_unlock:
 	return;
 }
 
+static void ipp_clean_cmd_node(struct ipp_context *ctx,
+				struct drm_exynos_ipp_cmd_node *c_node)
+{
+	int i;
+
+	/* cancel works */
+	cancel_work_sync(&c_node->start_work->work);
+	cancel_work_sync(&c_node->stop_work->work);
+	cancel_work_sync(&c_node->event_work->work);
+
+	/* put event */
+	ipp_put_event(c_node, NULL);
+
+	for_each_ipp_ops(i)
+		ipp_clean_mem_nodes(ctx->subdrv.drm_dev, c_node, i);
+
+	/* delete list */
+	list_del(&c_node->list);
+
+	ipp_remove_id(&ctx->prop_idr, &ctx->prop_lock,
+			c_node->property.prop_id);
+
+	/* destroy mutex */
+	mutex_destroy(&c_node->lock);
+	mutex_destroy(&c_node->mem_lock);
+	mutex_destroy(&c_node->event_lock);
+
+	/* free command node */
+	kfree(c_node->start_work);
+	kfree(c_node->stop_work);
+	kfree(c_node->event_work);
+	kfree(c_node);
+}
+
+static bool ipp_check_mem_list(struct drm_exynos_ipp_cmd_node *c_node)
+{
+	switch (c_node->property.cmd) {
+	case IPP_CMD_WB:
+		return !list_empty(&c_node->mem_list[EXYNOS_DRM_OPS_DST]);
+	case IPP_CMD_OUTPUT:
+		return !list_empty(&c_node->mem_list[EXYNOS_DRM_OPS_SRC]);
+	case IPP_CMD_M2M:
+	default:
+		return !list_empty(&c_node->mem_list[EXYNOS_DRM_OPS_SRC]) &&
+		       !list_empty(&c_node->mem_list[EXYNOS_DRM_OPS_DST]);
+	}
+}
+
+static struct drm_exynos_ipp_mem_node
+		*ipp_find_mem_node(struct drm_exynos_ipp_cmd_node *c_node,
+		struct drm_exynos_ipp_queue_buf *qbuf)
+{
+	struct drm_exynos_ipp_mem_node *m_node;
+	struct list_head *head;
+	int count = 0;
+
+	DRM_DEBUG_KMS("buf_id[%d]\n", qbuf->buf_id);
+
+	/* source/destination memory list */
+	head = &c_node->mem_list[qbuf->ops_id];
+
+	/* find memory node from memory list */
+	list_for_each_entry(m_node, head, list) {
+		DRM_DEBUG_KMS("count[%d]m_node[0x%x]\n", count++, (int)m_node);
+
+		/* compare buffer id */
+		if (m_node->buf_id == qbuf->buf_id)
+			return m_node;
+	}
+
+	return NULL;
+}
+
+static int ipp_set_mem_node(struct exynos_drm_ippdrv *ippdrv,
+		struct drm_exynos_ipp_cmd_node *c_node,
+		struct drm_exynos_ipp_mem_node *m_node)
+{
+	struct exynos_drm_ipp_ops *ops = NULL;
+	int ret = 0;
+
+	DRM_DEBUG_KMS("node[0x%x]\n", (int)m_node);
+
+	if (!m_node) {
+		DRM_ERROR("invalid queue node.\n");
+		return -EFAULT;
+	}
+
+	DRM_DEBUG_KMS("ops_id[%d]\n", m_node->ops_id);
+
+	/* get operations callback */
+	ops = ippdrv->ops[m_node->ops_id];
+	if (!ops) {
+		DRM_ERROR("not support ops.\n");
+		return -EFAULT;
+	}
+
+	/* set address and enable irq */
+	if (ops->set_addr) {
+		ret = ops->set_addr(ippdrv->dev, &m_node->buf_info,
+			m_node->buf_id, IPP_BUF_ENQUEUE);
+		if (ret) {
+			DRM_ERROR("failed to set addr.\n");
+			return ret;
+		}
+	}
+
+	return ret;
+}
+
 static void ipp_handle_cmd_work(struct device *dev,
 		struct exynos_drm_ippdrv *ippdrv,
 		struct drm_exynos_ipp_cmd_work *cmd_work,
@@ -766,7 +769,7 @@ static void ipp_handle_cmd_work(struct device *dev,
 
 	cmd_work->ippdrv = ippdrv;
 	cmd_work->c_node = c_node;
-	queue_work(ctx->cmd_workq, (struct work_struct *)cmd_work);
+	queue_work(ctx->cmd_workq, &cmd_work->work);
 }
 
 static int ipp_queue_buf_with_run(struct device *dev,
@@ -872,7 +875,7 @@ int exynos_drm_ipp_queue_buf(struct drm_device *drm_dev, void *data,
 	/* find command node */
 	c_node = ipp_find_obj(&ctx->prop_idr, &ctx->prop_lock,
 		qbuf->prop_id);
-	if (!c_node) {
+	if (!c_node || c_node->filp != file) {
 		DRM_ERROR("failed to get command node.\n");
 		return -ENODEV;
 	}
@@ -881,7 +884,7 @@ int exynos_drm_ipp_queue_buf(struct drm_device *drm_dev, void *data,
 	switch (qbuf->buf_type) {
 	case IPP_BUF_ENQUEUE:
 		/* get memory node */
-		m_node = ipp_get_mem_node(drm_dev, file, c_node, qbuf);
+		m_node = ipp_get_mem_node(drm_dev, c_node, qbuf);
 		if (IS_ERR(m_node)) {
 			DRM_ERROR("failed to get m_node.\n");
 			return PTR_ERR(m_node);
@@ -894,7 +897,7 @@ int exynos_drm_ipp_queue_buf(struct drm_device *drm_dev, void *data,
 		 */
 		if (qbuf->ops_id == EXYNOS_DRM_OPS_DST) {
 			/* get event for destination buffer */
-			ret = ipp_get_event(drm_dev, file, c_node, qbuf);
+			ret = ipp_get_event(drm_dev, c_node, qbuf);
 			if (ret) {
 				DRM_ERROR("failed to get event.\n");
 				goto err_clean_node;
@@ -1007,7 +1010,7 @@ int exynos_drm_ipp_cmd_ctrl(struct drm_device *drm_dev, void *data,
 
 	c_node = ipp_find_obj(&ctx->prop_idr, &ctx->prop_lock,
 		cmd_ctrl->prop_id);
-	if (!c_node) {
+	if (!c_node || c_node->filp != file) {
 		DRM_ERROR("invalid command node list.\n");
 		return -ENODEV;
 	}
@@ -1257,80 +1260,39 @@ static int ipp_stop_property(struct drm_device *drm_dev,
 		struct exynos_drm_ippdrv *ippdrv,
 		struct drm_exynos_ipp_cmd_node *c_node)
 {
-	struct drm_exynos_ipp_mem_node *m_node, *tm_node;
 	struct drm_exynos_ipp_property *property = &c_node->property;
-	struct list_head *head;
-	int ret = 0, i;
+	int i;
 
 	DRM_DEBUG_KMS("prop_id[%d]\n", property->prop_id);
 
-	/* put event */
-	ipp_put_event(c_node, NULL);
-
-	mutex_lock(&c_node->mem_lock);
+	/* stop operations */
+	if (ippdrv->stop)
+		ippdrv->stop(ippdrv->dev, property->cmd);
 
 	/* check command */
 	switch (property->cmd) {
 	case IPP_CMD_M2M:
-		for_each_ipp_ops(i) {
-			/* source/destination memory list */
-			head = &c_node->mem_list[i];
-
-			list_for_each_entry_safe(m_node, tm_node,
-				head, list) {
-				ret = ipp_put_mem_node(drm_dev, c_node,
-					m_node);
-				if (ret) {
-					DRM_ERROR("failed to put m_node.\n");
-					goto err_clear;
-				}
-			}
-		}
+		for_each_ipp_ops(i)
+			ipp_clean_mem_nodes(drm_dev, c_node, i);
 		break;
 	case IPP_CMD_WB:
-		/* destination memory list */
-		head = &c_node->mem_list[EXYNOS_DRM_OPS_DST];
-
-		list_for_each_entry_safe(m_node, tm_node, head, list) {
-			ret = ipp_put_mem_node(drm_dev, c_node, m_node);
-			if (ret) {
-				DRM_ERROR("failed to put m_node.\n");
-				goto err_clear;
-			}
-		}
+		ipp_clean_mem_nodes(drm_dev, c_node, EXYNOS_DRM_OPS_DST);
 		break;
 	case IPP_CMD_OUTPUT:
-		/* source memory list */
-		head = &c_node->mem_list[EXYNOS_DRM_OPS_SRC];
-
-		list_for_each_entry_safe(m_node, tm_node, head, list) {
-			ret = ipp_put_mem_node(drm_dev, c_node, m_node);
-			if (ret) {
-				DRM_ERROR("failed to put m_node.\n");
-				goto err_clear;
-			}
-		}
+		ipp_clean_mem_nodes(drm_dev, c_node, EXYNOS_DRM_OPS_SRC);
 		break;
 	default:
 		DRM_ERROR("invalid operations.\n");
-		ret = -EINVAL;
-		goto err_clear;
+		return -EINVAL;
 	}
 
-err_clear:
-	mutex_unlock(&c_node->mem_lock);
-
-	/* stop operations */
-	if (ippdrv->stop)
-		ippdrv->stop(ippdrv->dev, property->cmd);
-
-	return ret;
+	return 0;
 }
 
 void ipp_sched_cmd(struct work_struct *work)
 {
 	struct drm_exynos_ipp_cmd_work *cmd_work =
-		(struct drm_exynos_ipp_cmd_work *)work;
+		container_of(work, struct drm_exynos_ipp_cmd_work, work);
 	struct exynos_drm_ippdrv *ippdrv;
 	struct drm_exynos_ipp_cmd_node *c_node;
 	struct drm_exynos_ipp_property *property;
@@ -1543,7 +1505,7 @@ err_event_unlock:
 void ipp_sched_event(struct work_struct *work)
 {
 	struct drm_exynos_ipp_event_work *event_work =
-		(struct drm_exynos_ipp_event_work *)work;
+		container_of(work, struct drm_exynos_ipp_event_work, work);
 	struct exynos_drm_ippdrv *ippdrv;
 	struct drm_exynos_ipp_cmd_node *c_node;
 	int ret;
@@ -1646,11 +1608,11 @@ err:
 
 static void ipp_subdrv_remove(struct drm_device *drm_dev, struct device *dev)
 {
-	struct exynos_drm_ippdrv *ippdrv;
+	struct exynos_drm_ippdrv *ippdrv, *t;
 	struct ipp_context *ctx = get_ipp_context(dev);
 
 	/* get ipp driver entry */
-	list_for_each_entry(ippdrv, &exynos_drm_ippdrv_list, drv_list) {
+	list_for_each_entry_safe(ippdrv, t, &exynos_drm_ippdrv_list, drv_list) {
 		if (is_drm_iommu_supported(drm_dev))
 			drm_iommu_detach_device(drm_dev, ippdrv->dev);
 
@@ -1677,14 +1639,11 @@ static int ipp_subdrv_open(struct drm_device *drm_dev, struct device *dev,
 static void ipp_subdrv_close(struct drm_device *drm_dev, struct device *dev,
 		struct drm_file *file)
 {
-	struct drm_exynos_file_private *file_priv = file->driver_priv;
 	struct exynos_drm_ippdrv *ippdrv = NULL;
 	struct ipp_context *ctx = get_ipp_context(dev);
 	struct drm_exynos_ipp_cmd_node *c_node, *tc_node;
 	int count = 0;
 
-	DRM_DEBUG_KMS("for priv[0x%x]\n", (int)file_priv->ipp_dev);
-
 	list_for_each_entry(ippdrv, &exynos_drm_ippdrv_list, drv_list) {
 		mutex_lock(&ippdrv->cmd_lock);
 		list_for_each_entry_safe(c_node, tc_node,
@@ -1692,7 +1651,7 @@ static void ipp_subdrv_close(struct drm_device *drm_dev, struct device *dev,
 			DRM_DEBUG_KMS("count[%d]ippdrv[0x%x]\n",
 				count++, (int)ippdrv);
 
-			if (c_node->dev == file_priv->ipp_dev) {
+			if (c_node->filp == file) {
 				/*
 				 * userland goto unnormal state. process killed.
 				 * and close the file.
@@ -1808,63 +1767,12 @@ static int ipp_remove(struct platform_device *pdev)
 	return 0;
 }
 
-static int ipp_power_ctrl(struct ipp_context *ctx, bool enable)
-{
-	DRM_DEBUG_KMS("enable[%d]\n", enable);
-
-	return 0;
-}
-
-#ifdef CONFIG_PM_SLEEP
-static int ipp_suspend(struct device *dev)
-{
-	struct ipp_context *ctx = get_ipp_context(dev);
-
-	if (pm_runtime_suspended(dev))
-		return 0;
-
-	return ipp_power_ctrl(ctx, false);
-}
-
-static int ipp_resume(struct device *dev)
-{
-	struct ipp_context *ctx = get_ipp_context(dev);
-
-	if (!pm_runtime_suspended(dev))
-		return ipp_power_ctrl(ctx, true);
-
-	return 0;
-}
-#endif
-
-#ifdef CONFIG_PM_RUNTIME
-static int ipp_runtime_suspend(struct device *dev)
-{
-	struct ipp_context *ctx = get_ipp_context(dev);
-
-	return ipp_power_ctrl(ctx, false);
-}
-
-static int ipp_runtime_resume(struct device *dev)
-{
-	struct ipp_context *ctx = get_ipp_context(dev);
-
-	return ipp_power_ctrl(ctx, true);
-}
-#endif
-
-static const struct dev_pm_ops ipp_pm_ops = {
-	SET_SYSTEM_SLEEP_PM_OPS(ipp_suspend, ipp_resume)
-	SET_RUNTIME_PM_OPS(ipp_runtime_suspend, ipp_runtime_resume, NULL)
-};
-
 struct platform_driver ipp_driver = {
 	.probe		= ipp_probe,
 	.remove		= ipp_remove,
 	.driver		= {
 		.name	= "exynos-drm-ipp",
 		.owner	= THIS_MODULE,
-		.pm	= &ipp_pm_ops,
 	},
 };
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_ipp.h b/drivers/gpu/drm/exynos/exynos_drm_ipp.h
index 6f48d62aeb30..2a61547a39d0 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_ipp.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_ipp.h
@@ -48,7 +48,6 @@ struct drm_exynos_ipp_cmd_work {
 /*
  * A structure of command node.
  *
- * @dev: IPP device.
  * @list: list head to command queue information.
  * @event_list: list head of event.
  * @mem_list: list head to source,destination memory queue information.
@@ -62,9 +61,9 @@ struct drm_exynos_ipp_cmd_work {
  * @stop_work: stop command work structure.
  * @event_work: event work structure.
  * @state: state of command node.
+ * @filp: associated file pointer.
  */
 struct drm_exynos_ipp_cmd_node {
-	struct device		*dev;
 	struct list_head	list;
 	struct list_head	event_list;
 	struct list_head	mem_list[EXYNOS_DRM_OPS_MAX];
@@ -78,6 +77,7 @@ struct drm_exynos_ipp_cmd_node {
 	struct drm_exynos_ipp_cmd_work *stop_work;
 	struct drm_exynos_ipp_event_work *event_work;
 	enum drm_exynos_ipp_state	state;
+	struct drm_file	*filp;
 };
 
 /*
diff --git a/drivers/gpu/drm/exynos/exynos_drm_plane.c b/drivers/gpu/drm/exynos/exynos_drm_plane.c
index 8371cbd7631d..c7045a663763 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_plane.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_plane.c
@@ -139,6 +139,8 @@ int exynos_plane_mode_set(struct drm_plane *plane, struct drm_crtc *crtc,
 			overlay->crtc_x, overlay->crtc_y,
 			overlay->crtc_width, overlay->crtc_height);
 
+	plane->crtc = crtc;
+
 	exynos_drm_crtc_plane_mode_set(crtc, overlay);
 
 	return 0;
@@ -187,8 +189,6 @@ exynos_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	if (ret < 0)
 		return ret;
 
-	plane->crtc = crtc;
-
 	exynos_plane_commit(plane);
 	exynos_plane_dpms(plane, DRM_MODE_DPMS_ON);
 
@@ -254,25 +254,26 @@ static void exynos_plane_attach_zpos_property(struct drm_plane *plane)
 }
 
 struct drm_plane *exynos_plane_init(struct drm_device *dev,
-				    unsigned long possible_crtcs, bool priv)
+				    unsigned long possible_crtcs,
+				    enum drm_plane_type type)
 {
 	struct exynos_plane *exynos_plane;
 	int err;
 
 	exynos_plane = kzalloc(sizeof(struct exynos_plane), GFP_KERNEL);
 	if (!exynos_plane)
-		return NULL;
+		return ERR_PTR(-ENOMEM);
 
-	err = drm_plane_init(dev, &exynos_plane->base, possible_crtcs,
-			      &exynos_plane_funcs, formats, ARRAY_SIZE(formats),
-			      priv);
+	err = drm_universal_plane_init(dev, &exynos_plane->base, possible_crtcs,
+				       &exynos_plane_funcs, formats,
+				       ARRAY_SIZE(formats), type);
 	if (err) {
 		DRM_ERROR("failed to initialize plane\n");
 		kfree(exynos_plane);
-		return NULL;
+		return ERR_PTR(err);
 	}
 
-	if (priv)
+	if (type == DRM_PLANE_TYPE_PRIMARY)
 		exynos_plane->overlay.zpos = DEFAULT_ZPOS;
 	else
 		exynos_plane_attach_zpos_property(&exynos_plane->base);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_plane.h b/drivers/gpu/drm/exynos/exynos_drm_plane.h
index 84d464c90d3d..0d1986b115f8 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_plane.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_plane.h
@@ -17,4 +17,5 @@ int exynos_plane_mode_set(struct drm_plane *plane, struct drm_crtc *crtc,
 void exynos_plane_commit(struct drm_plane *plane);
 void exynos_plane_dpms(struct drm_plane *plane, int mode);
 struct drm_plane *exynos_plane_init(struct drm_device *dev,
-				    unsigned long possible_crtcs, bool priv);
+				    unsigned long possible_crtcs,
+				    enum drm_plane_type type);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_rotator.c b/drivers/gpu/drm/exynos/exynos_drm_rotator.c
index 55af6b41c1df..425e70625388 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_rotator.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_rotator.c
@@ -156,8 +156,7 @@ static irqreturn_t rotator_irq_handler(int irq, void *arg)
 		event_work->ippdrv = ippdrv;
 		event_work->buf_id[EXYNOS_DRM_OPS_DST] =
 			rot->cur_buf_id[EXYNOS_DRM_OPS_DST];
-		queue_work(ippdrv->event_workq,
-			(struct work_struct *)event_work);
+		queue_work(ippdrv->event_workq, &event_work->work);
 	} else {
 		DRM_ERROR("the SFR is set illegally\n");
 	}
@@ -823,7 +822,7 @@ static int rotator_resume(struct device *dev)
 }
 #endif
 
-#ifdef CONFIG_PM_RUNTIME
+#ifdef CONFIG_PM
 static int rotator_runtime_suspend(struct device *dev)
 {
 	struct rot_context *rot = dev_get_drvdata(dev);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_vidi.c b/drivers/gpu/drm/exynos/exynos_drm_vidi.c
index 9528d81d8004..45899fb63272 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_vidi.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_vidi.c
@@ -14,6 +14,7 @@
 
 #include <linux/kernel.h>
 #include <linux/platform_device.h>
+#include <linux/component.h>
 
 #include <drm/exynos_drm.h>
 
@@ -28,7 +29,6 @@
 /* vidi has totally three virtual windows. */
 #define WINDOWS_NR		3
 
-#define get_vidi_mgr(dev)	platform_get_drvdata(to_platform_device(dev))
 #define ctx_from_connector(c)	container_of(c, struct vidi_context, \
 					connector)
 
@@ -47,11 +47,13 @@ struct vidi_win_data {
 };
 
 struct vidi_context {
+	struct exynos_drm_manager	manager;
+	struct exynos_drm_display	display;
+	struct platform_device		*pdev;
 	struct drm_device		*drm_dev;
 	struct drm_crtc			*crtc;
 	struct drm_encoder		*encoder;
 	struct drm_connector		connector;
-	struct exynos_drm_subdrv	subdrv;
 	struct vidi_win_data		win_data[WINDOWS_NR];
 	struct edid			*raw_edid;
 	unsigned int			clkdiv;
@@ -66,6 +68,16 @@ struct vidi_context {
 	int				pipe;
 };
 
+static inline struct vidi_context *manager_to_vidi(struct exynos_drm_manager *m)
+{
+	return container_of(m, struct vidi_context, manager);
+}
+
+static inline struct vidi_context *display_to_vidi(struct exynos_drm_display *d)
+{
+	return container_of(d, struct vidi_context, display);
+}
+
 static const char fake_edid_info[] = {
 	0x00, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x00, 0x4c, 0x2d, 0x05, 0x05,
 	0x00, 0x00, 0x00, 0x00, 0x30, 0x12, 0x01, 0x03, 0x80, 0x10, 0x09, 0x78,
@@ -93,7 +105,7 @@ static const char fake_edid_info[] = {
 
 static void vidi_apply(struct exynos_drm_manager *mgr)
 {
-	struct vidi_context *ctx = mgr->ctx;
+	struct vidi_context *ctx = manager_to_vidi(mgr);
 	struct exynos_drm_manager_ops *mgr_ops = mgr->ops;
 	struct vidi_win_data *win_data;
 	int i;
@@ -110,7 +122,7 @@ static void vidi_apply(struct exynos_drm_manager *mgr)
 
 static void vidi_commit(struct exynos_drm_manager *mgr)
 {
-	struct vidi_context *ctx = mgr->ctx;
+	struct vidi_context *ctx = manager_to_vidi(mgr);
 
 	if (ctx->suspended)
 		return;
@@ -118,7 +130,7 @@ static void vidi_commit(struct exynos_drm_manager *mgr)
 
 static int vidi_enable_vblank(struct exynos_drm_manager *mgr)
 {
-	struct vidi_context *ctx = mgr->ctx;
+	struct vidi_context *ctx = manager_to_vidi(mgr);
 
 	if (ctx->suspended)
 		return -EPERM;
@@ -140,7 +152,7 @@ static int vidi_enable_vblank(struct exynos_drm_manager *mgr)
 
 static void vidi_disable_vblank(struct exynos_drm_manager *mgr)
 {
-	struct vidi_context *ctx = mgr->ctx;
+	struct vidi_context *ctx = manager_to_vidi(mgr);
 
 	if (ctx->suspended)
 		return;
@@ -152,7 +164,7 @@ static void vidi_disable_vblank(struct exynos_drm_manager *mgr)
 static void vidi_win_mode_set(struct exynos_drm_manager *mgr,
 			struct exynos_drm_overlay *overlay)
 {
-	struct vidi_context *ctx = mgr->ctx;
+	struct vidi_context *ctx = manager_to_vidi(mgr);
 	struct vidi_win_data *win_data;
 	int win;
 	unsigned long offset;
@@ -204,7 +216,7 @@ static void vidi_win_mode_set(struct exynos_drm_manager *mgr,
 
 static void vidi_win_commit(struct exynos_drm_manager *mgr, int zpos)
 {
-	struct vidi_context *ctx = mgr->ctx;
+	struct vidi_context *ctx = manager_to_vidi(mgr);
 	struct vidi_win_data *win_data;
 	int win = zpos;
 
@@ -229,7 +241,7 @@ static void vidi_win_commit(struct exynos_drm_manager *mgr, int zpos)
 
 static void vidi_win_disable(struct exynos_drm_manager *mgr, int zpos)
 {
-	struct vidi_context *ctx = mgr->ctx;
+	struct vidi_context *ctx = manager_to_vidi(mgr);
 	struct vidi_win_data *win_data;
 	int win = zpos;
 
@@ -247,7 +259,7 @@ static void vidi_win_disable(struct exynos_drm_manager *mgr, int zpos)
 
 static int vidi_power_on(struct exynos_drm_manager *mgr, bool enable)
 {
-	struct vidi_context *ctx = mgr->ctx;
+	struct vidi_context *ctx = manager_to_vidi(mgr);
 
 	DRM_DEBUG_KMS("%s\n", __FILE__);
 
@@ -271,7 +283,7 @@ static int vidi_power_on(struct exynos_drm_manager *mgr, bool enable)
 
 static void vidi_dpms(struct exynos_drm_manager *mgr, int mode)
 {
-	struct vidi_context *ctx = mgr->ctx;
+	struct vidi_context *ctx = manager_to_vidi(mgr);
 
 	DRM_DEBUG_KMS("%d\n", mode);
 
@@ -297,29 +309,12 @@ static void vidi_dpms(struct exynos_drm_manager *mgr, int mode)
 static int vidi_mgr_initialize(struct exynos_drm_manager *mgr,
 			struct drm_device *drm_dev)
 {
-	struct vidi_context *ctx = mgr->ctx;
+	struct vidi_context *ctx = manager_to_vidi(mgr);
 	struct exynos_drm_private *priv = drm_dev->dev_private;
 
 	mgr->drm_dev = ctx->drm_dev = drm_dev;
 	mgr->pipe = ctx->pipe = priv->pipe++;
 
-	/*
-	 * enable drm irq mode.
-	 * - with irq_enabled = 1, we can use the vblank feature.
-	 *
-	 * P.S. note that we wouldn't use drm irq handler but
-	 *	just specific driver own one instead because
-	 *	drm framework supports only one irq handler.
-	 */
-	drm_dev->irq_enabled = 1;
-
-	/*
-	 * with vblank_disable_allowed = 1, vblank interrupt will be disabled
-	 * by drm timer once a current process gives up ownership of
-	 * vblank event.(after drm_vblank_put function is called)
-	 */
-	drm_dev->vblank_disable_allowed = 1;
-
 	return 0;
 }
 
@@ -333,11 +328,6 @@ static struct exynos_drm_manager_ops vidi_manager_ops = {
 	.win_disable = vidi_win_disable,
 };
 
-static struct exynos_drm_manager vidi_manager = {
-	.type = EXYNOS_DISPLAY_TYPE_VIDI,
-	.ops = &vidi_manager_ops,
-};
-
 static void vidi_fake_vblank_handler(struct work_struct *work)
 {
 	struct vidi_context *ctx = container_of(work, struct vidi_context,
@@ -366,9 +356,8 @@ static void vidi_fake_vblank_handler(struct work_struct *work)
 static int vidi_show_connection(struct device *dev,
 				struct device_attribute *attr, char *buf)
 {
+	struct vidi_context *ctx = dev_get_drvdata(dev);
 	int rc;
-	struct exynos_drm_manager *mgr = get_vidi_mgr(dev);
-	struct vidi_context *ctx = mgr->ctx;
 
 	mutex_lock(&ctx->lock);
 
@@ -383,8 +372,7 @@ static int vidi_store_connection(struct device *dev,
 				struct device_attribute *attr,
 				const char *buf, size_t len)
 {
-	struct exynos_drm_manager *mgr = get_vidi_mgr(dev);
-	struct vidi_context *ctx = mgr->ctx;
+	struct vidi_context *ctx = dev_get_drvdata(dev);
 	int ret;
 
 	ret = kstrtoint(buf, 0, &ctx->connected);
@@ -437,7 +425,7 @@ int vidi_connection_ioctl(struct drm_device *drm_dev, void *data,
 		display = exynos_drm_get_display(encoder);
 
 		if (display->type == EXYNOS_DISPLAY_TYPE_VIDI) {
-			ctx = display->ctx;
+			ctx = display_to_vidi(display);
 			break;
 		}
 	}
@@ -547,7 +535,7 @@ static struct drm_connector_helper_funcs vidi_connector_helper_funcs = {
 static int vidi_create_connector(struct exynos_drm_display *display,
 				struct drm_encoder *encoder)
 {
-	struct vidi_context *ctx = display->ctx;
+	struct vidi_context *ctx = display_to_vidi(display);
 	struct drm_connector *connector = &ctx->connector;
 	int ret;
 
@@ -573,27 +561,22 @@ static struct exynos_drm_display_ops vidi_display_ops = {
 	.create_connector = vidi_create_connector,
 };
 
-static struct exynos_drm_display vidi_display = {
-	.type = EXYNOS_DISPLAY_TYPE_VIDI,
-	.ops = &vidi_display_ops,
-};
-
-static int vidi_subdrv_probe(struct drm_device *drm_dev, struct device *dev)
+static int vidi_bind(struct device *dev, struct device *master, void *data)
 {
-	struct exynos_drm_manager *mgr = get_vidi_mgr(dev);
-	struct vidi_context *ctx = mgr->ctx;
+	struct vidi_context *ctx = dev_get_drvdata(dev);
+	struct drm_device *drm_dev = data;
 	struct drm_crtc *crtc = ctx->crtc;
 	int ret;
 
-	vidi_mgr_initialize(mgr, drm_dev);
+	vidi_mgr_initialize(&ctx->manager, drm_dev);
 
-	ret = exynos_drm_crtc_create(&vidi_manager);
+	ret = exynos_drm_crtc_create(&ctx->manager);
 	if (ret) {
 		DRM_ERROR("failed to create crtc.\n");
 		return ret;
 	}
 
-	ret = exynos_drm_create_enc_conn(drm_dev, &vidi_display);
+	ret = exynos_drm_create_enc_conn(drm_dev, &ctx->display);
 	if (ret) {
 		crtc->funcs->destroy(crtc);
 		DRM_ERROR("failed to create encoder and connector.\n");
@@ -603,9 +586,18 @@ static int vidi_subdrv_probe(struct drm_device *drm_dev, struct device *dev)
 	return 0;
 }
 
+
+static void vidi_unbind(struct device *dev, struct device *master, void *data)
+{
+}
+
+static const struct component_ops vidi_component_ops = {
+	.bind	= vidi_bind,
+	.unbind = vidi_unbind,
+};
+
 static int vidi_probe(struct platform_device *pdev)
 {
-	struct exynos_drm_subdrv *subdrv;
 	struct vidi_context *ctx;
 	int ret;
 
@@ -613,42 +605,54 @@ static int vidi_probe(struct platform_device *pdev)
 	if (!ctx)
 		return -ENOMEM;
 
+	ctx->manager.type = EXYNOS_DISPLAY_TYPE_VIDI;
+	ctx->manager.ops = &vidi_manager_ops;
+	ctx->display.type = EXYNOS_DISPLAY_TYPE_VIDI;
+	ctx->display.ops = &vidi_display_ops;
 	ctx->default_win = 0;
+	ctx->pdev = pdev;
 
-	INIT_WORK(&ctx->work, vidi_fake_vblank_handler);
-
-	vidi_manager.ctx = ctx;
-	vidi_display.ctx = ctx;
+	ret = exynos_drm_component_add(&pdev->dev, EXYNOS_DEVICE_TYPE_CRTC,
+					ctx->manager.type);
+	if (ret)
+		return ret;
 
-	mutex_init(&ctx->lock);
+	ret = exynos_drm_component_add(&pdev->dev, EXYNOS_DEVICE_TYPE_CONNECTOR,
+					ctx->display.type);
+	if (ret)
+		goto err_del_crtc_component;
 
-	platform_set_drvdata(pdev, &vidi_manager);
+	INIT_WORK(&ctx->work, vidi_fake_vblank_handler);
 
-	subdrv = &ctx->subdrv;
-	subdrv->dev = &pdev->dev;
-	subdrv->probe = vidi_subdrv_probe;
+	mutex_init(&ctx->lock);
 
-	ret = exynos_drm_subdrv_register(subdrv);
-	if (ret < 0) {
-		dev_err(&pdev->dev, "failed to register drm vidi device\n");
-		return ret;
-	}
+	platform_set_drvdata(pdev, ctx);
 
 	ret = device_create_file(&pdev->dev, &dev_attr_connection);
 	if (ret < 0) {
-		exynos_drm_subdrv_unregister(subdrv);
-		DRM_INFO("failed to create connection sysfs.\n");
+		DRM_ERROR("failed to create connection sysfs.\n");
+		goto err_del_conn_component;
 	}
 
-	return 0;
+	ret = component_add(&pdev->dev, &vidi_component_ops);
+	if (ret)
+		goto err_remove_file;
+
+	return ret;
+
+err_remove_file:
+	device_remove_file(&pdev->dev, &dev_attr_connection);
+err_del_conn_component:
+	exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CONNECTOR);
+err_del_crtc_component:
+	exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CRTC);
+
+	return ret;
 }
 
 static int vidi_remove(struct platform_device *pdev)
 {
-	struct exynos_drm_manager *mgr = platform_get_drvdata(pdev);
-	struct vidi_context *ctx = mgr->ctx;
-	struct drm_encoder *encoder = ctx->encoder;
-	struct drm_crtc *crtc = mgr->crtc;
+	struct vidi_context *ctx = platform_get_drvdata(pdev);
 
 	if (ctx->raw_edid != (struct edid *)fake_edid_info) {
 		kfree(ctx->raw_edid);
@@ -657,9 +661,9 @@ static int vidi_remove(struct platform_device *pdev)
 		return -EINVAL;
 	}
 
-	crtc->funcs->destroy(crtc);
-	encoder->funcs->destroy(encoder);
-	drm_connector_cleanup(&ctx->connector);
+	component_del(&pdev->dev, &vidi_component_ops);
+	exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CONNECTOR);
+	exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CRTC);
 
 	return 0;
 }
@@ -691,12 +695,19 @@ int exynos_drm_probe_vidi(void)
 	return ret;
 }
 
+static int exynos_drm_remove_vidi_device(struct device *dev, void *data)
+{
+	platform_device_unregister(to_platform_device(dev));
+
+	return 0;
+}
+
 void exynos_drm_remove_vidi(void)
 {
-	struct vidi_context *ctx = vidi_manager.ctx;
-	struct exynos_drm_subdrv *subdrv = &ctx->subdrv;
-	struct platform_device *pdev = to_platform_device(subdrv->dev);
+	int ret = driver_for_each_device(&vidi_driver.driver, NULL, NULL,
+					 exynos_drm_remove_vidi_device);
+	/* silence compiler warning */
+	(void)ret;
 
 	platform_driver_unregister(&vidi_driver);
-	platform_device_unregister(pdev);
 }
diff --git a/drivers/gpu/drm/exynos/exynos_hdmi.c b/drivers/gpu/drm/exynos/exynos_hdmi.c
index 562966db2aa1..98051e8e855a 100644
--- a/drivers/gpu/drm/exynos/exynos_hdmi.c
+++ b/drivers/gpu/drm/exynos/exynos_hdmi.c
@@ -49,7 +49,6 @@
 #include <linux/gpio.h>
 #include <media/s5p_hdmi.h>
 
-#define get_hdmi_display(dev)	platform_get_drvdata(to_platform_device(dev))
 #define ctx_from_connector(c)	container_of(c, struct hdmi_context, connector)
 
 #define HOTPLUG_DEBOUNCE_MS		1100
@@ -182,6 +181,7 @@ struct hdmi_conf_regs {
 };
 
 struct hdmi_context {
+	struct exynos_drm_display	display;
 	struct device			*dev;
 	struct drm_device		*drm_dev;
 	struct drm_connector		connector;
@@ -213,6 +213,11 @@ struct hdmi_context {
 	enum hdmi_type			type;
 };
 
+static inline struct hdmi_context *display_to_hdmi(struct exynos_drm_display *d)
+{
+	return container_of(d, struct hdmi_context, display);
+}
+
 struct hdmiphy_config {
 	int pixel_clock;
 	u8 conf[32];
@@ -1040,6 +1045,8 @@ static enum drm_connector_status hdmi_detect(struct drm_connector *connector,
 
 static void hdmi_connector_destroy(struct drm_connector *connector)
 {
+	drm_connector_unregister(connector);
+	drm_connector_cleanup(connector);
 }
 
 static struct drm_connector_funcs hdmi_connector_funcs = {
@@ -1121,7 +1128,7 @@ static struct drm_connector_helper_funcs hdmi_connector_helper_funcs = {
 static int hdmi_create_connector(struct exynos_drm_display *display,
 			struct drm_encoder *encoder)
 {
-	struct hdmi_context *hdata = display->ctx;
+	struct hdmi_context *hdata = display_to_hdmi(display);
 	struct drm_connector *connector = &hdata->connector;
 	int ret;
 
@@ -1662,7 +1669,6 @@ static void hdmi_mode_apply(struct hdmi_context *hdata)
 
 static void hdmiphy_conf_reset(struct hdmi_context *hdata)
 {
-	u8 buffer[2];
 	u32 reg;
 
 	clk_disable_unprepare(hdata->res.sclk_hdmi);
@@ -1670,11 +1676,8 @@ static void hdmiphy_conf_reset(struct hdmi_context *hdata)
 	clk_prepare_enable(hdata->res.sclk_hdmi);
 
 	/* operation mode */
-	buffer[0] = 0x1f;
-	buffer[1] = 0x00;
-
-	if (hdata->hdmiphy_port)
-		i2c_master_send(hdata->hdmiphy_port, buffer, 2);
+	hdmiphy_reg_writeb(hdata, HDMIPHY_MODE_SET_DONE,
+				HDMI_PHY_ENABLE_MODE_SET);
 
 	if (hdata->type == HDMI_TYPE13)
 		reg = HDMI_V13_PHY_RSTOUT;
@@ -1998,7 +2001,7 @@ static void hdmi_v14_mode_set(struct hdmi_context *hdata,
 static void hdmi_mode_set(struct exynos_drm_display *display,
 			struct drm_display_mode *mode)
 {
-	struct hdmi_context *hdata = display->ctx;
+	struct hdmi_context *hdata = display_to_hdmi(display);
 	struct drm_display_mode *m = mode;
 
 	DRM_DEBUG_KMS("xres=%d, yres=%d, refresh=%d, intl=%s\n",
@@ -2017,7 +2020,7 @@ static void hdmi_mode_set(struct exynos_drm_display *display,
 
 static void hdmi_commit(struct exynos_drm_display *display)
 {
-	struct hdmi_context *hdata = display->ctx;
+	struct hdmi_context *hdata = display_to_hdmi(display);
 
 	mutex_lock(&hdata->hdmi_mutex);
 	if (!hdata->powered) {
@@ -2031,7 +2034,7 @@ static void hdmi_commit(struct exynos_drm_display *display)
 
 static void hdmi_poweron(struct exynos_drm_display *display)
 {
-	struct hdmi_context *hdata = display->ctx;
+	struct hdmi_context *hdata = display_to_hdmi(display);
 	struct hdmi_resources *res = &hdata->res;
 
 	mutex_lock(&hdata->hdmi_mutex);
@@ -2062,7 +2065,7 @@ static void hdmi_poweron(struct exynos_drm_display *display)
 
 static void hdmi_poweroff(struct exynos_drm_display *display)
 {
-	struct hdmi_context *hdata = display->ctx;
+	struct hdmi_context *hdata = display_to_hdmi(display);
 	struct hdmi_resources *res = &hdata->res;
 
 	mutex_lock(&hdata->hdmi_mutex);
@@ -2097,7 +2100,7 @@ out:
 
 static void hdmi_dpms(struct exynos_drm_display *display, int mode)
 {
-	struct hdmi_context *hdata = display->ctx;
+	struct hdmi_context *hdata = display_to_hdmi(display);
 	struct drm_encoder *encoder = hdata->encoder;
 	struct drm_crtc *crtc = encoder->crtc;
 	struct drm_crtc_helper_funcs *funcs = NULL;
@@ -2141,11 +2144,6 @@ static struct exynos_drm_display_ops hdmi_display_ops = {
 	.commit		= hdmi_commit,
 };
 
-static struct exynos_drm_display hdmi_display = {
-	.type = EXYNOS_DISPLAY_TYPE_HDMI,
-	.ops = &hdmi_display_ops,
-};
-
 static void hdmi_hotplug_work_func(struct work_struct *work)
 {
 	struct hdmi_context *hdata;
@@ -2300,22 +2298,15 @@ MODULE_DEVICE_TABLE (of, hdmi_match_types);
 static int hdmi_bind(struct device *dev, struct device *master, void *data)
 {
 	struct drm_device *drm_dev = data;
-	struct hdmi_context *hdata;
+	struct hdmi_context *hdata = dev_get_drvdata(dev);
 
-	hdata = hdmi_display.ctx;
 	hdata->drm_dev = drm_dev;
 
-	return exynos_drm_create_enc_conn(drm_dev, &hdmi_display);
+	return exynos_drm_create_enc_conn(drm_dev, &hdata->display);
 }
 
 static void hdmi_unbind(struct device *dev, struct device *master, void *data)
 {
-	struct exynos_drm_display *display = get_hdmi_display(dev);
-	struct drm_encoder *encoder = display->encoder;
-	struct hdmi_context *hdata = display->ctx;
-
-	encoder->funcs->destroy(encoder);
-	drm_connector_cleanup(&hdata->connector);
 }
 
 static const struct component_ops hdmi_component_ops = {
@@ -2353,31 +2344,28 @@ static int hdmi_probe(struct platform_device *pdev)
 	struct resource *res;
 	int ret;
 
-	ret = exynos_drm_component_add(&pdev->dev, EXYNOS_DEVICE_TYPE_CONNECTOR,
-					hdmi_display.type);
-	if (ret)
-		return ret;
-
-	if (!dev->of_node) {
-		ret = -ENODEV;
-		goto err_del_component;
-	}
+	if (!dev->of_node)
+		return -ENODEV;
 
 	pdata = drm_hdmi_dt_parse_pdata(dev);
-	if (!pdata) {
-		ret = -EINVAL;
-		goto err_del_component;
-	}
+	if (!pdata)
+		return -EINVAL;
 
 	hdata = devm_kzalloc(dev, sizeof(struct hdmi_context), GFP_KERNEL);
-	if (!hdata) {
-		ret = -ENOMEM;
-		goto err_del_component;
-	}
+	if (!hdata)
+		return -ENOMEM;
+
+	hdata->display.type = EXYNOS_DISPLAY_TYPE_HDMI;
+	hdata->display.ops = &hdmi_display_ops;
+
+	ret = exynos_drm_component_add(&pdev->dev, EXYNOS_DEVICE_TYPE_CONNECTOR,
+					hdata->display.type);
+	if (ret)
+		return ret;
 
 	mutex_init(&hdata->hdmi_mutex);
 
-	platform_set_drvdata(pdev, &hdmi_display);
+	platform_set_drvdata(pdev, hdata);
 
 	match = of_match_node(hdmi_match_types, dev->of_node);
 	if (!match) {
@@ -2489,7 +2477,6 @@ out_get_phy_port:
 	}
 
 	pm_runtime_enable(dev);
-	hdmi_display.ctx = hdata;
 
 	ret = component_add(&pdev->dev, &hdmi_component_ops);
 	if (ret)
@@ -2514,7 +2501,7 @@ err_del_component:
 
 static int hdmi_remove(struct platform_device *pdev)
 {
-	struct hdmi_context *hdata = hdmi_display.ctx;
+	struct hdmi_context *hdata = platform_get_drvdata(pdev);
 
 	cancel_delayed_work_sync(&hdata->hotplug_work);
 
diff --git a/drivers/gpu/drm/exynos/exynos_mixer.c b/drivers/gpu/drm/exynos/exynos_mixer.c
index e8b4ec84b312..064ed6597def 100644
--- a/drivers/gpu/drm/exynos/exynos_mixer.c
+++ b/drivers/gpu/drm/exynos/exynos_mixer.c
@@ -40,8 +40,6 @@
 #include "exynos_drm_iommu.h"
 #include "exynos_mixer.h"
 
-#define get_mixer_manager(dev)	platform_get_drvdata(to_platform_device(dev))
-
 #define MIXER_WIN_NR		3
 #define MIXER_DEFAULT_WIN	0
 
@@ -86,6 +84,7 @@ enum mixer_version_id {
 };
 
 struct mixer_context {
+	struct exynos_drm_manager manager;
 	struct platform_device *pdev;
 	struct device		*dev;
 	struct drm_device	*drm_dev;
@@ -104,6 +103,11 @@ struct mixer_context {
 	atomic_t		wait_vsync_event;
 };
 
+static inline struct mixer_context *mgr_to_mixer(struct exynos_drm_manager *mgr)
+{
+	return container_of(mgr, struct mixer_context, manager);
+}
+
 struct mixer_drv_data {
 	enum mixer_version_id	version;
 	bool					is_vp_enabled;
@@ -854,7 +858,7 @@ static int mixer_initialize(struct exynos_drm_manager *mgr,
 			struct drm_device *drm_dev)
 {
 	int ret;
-	struct mixer_context *mixer_ctx = mgr->ctx;
+	struct mixer_context *mixer_ctx = mgr_to_mixer(mgr);
 	struct exynos_drm_private *priv;
 	priv = drm_dev->dev_private;
 
@@ -885,7 +889,7 @@ static int mixer_initialize(struct exynos_drm_manager *mgr,
 
 static void mixer_mgr_remove(struct exynos_drm_manager *mgr)
 {
-	struct mixer_context *mixer_ctx = mgr->ctx;
+	struct mixer_context *mixer_ctx = mgr_to_mixer(mgr);
 
 	if (is_drm_iommu_supported(mixer_ctx->drm_dev))
 		drm_iommu_detach_device(mixer_ctx->drm_dev, mixer_ctx->dev);
@@ -893,7 +897,7 @@ static void mixer_mgr_remove(struct exynos_drm_manager *mgr)
 
 static int mixer_enable_vblank(struct exynos_drm_manager *mgr)
 {
-	struct mixer_context *mixer_ctx = mgr->ctx;
+	struct mixer_context *mixer_ctx = mgr_to_mixer(mgr);
 	struct mixer_resources *res = &mixer_ctx->mixer_res;
 
 	if (!mixer_ctx->powered) {
@@ -910,7 +914,7 @@ static int mixer_enable_vblank(struct exynos_drm_manager *mgr)
 
 static void mixer_disable_vblank(struct exynos_drm_manager *mgr)
 {
-	struct mixer_context *mixer_ctx = mgr->ctx;
+	struct mixer_context *mixer_ctx = mgr_to_mixer(mgr);
 	struct mixer_resources *res = &mixer_ctx->mixer_res;
 
 	/* disable vsync interrupt */
@@ -920,7 +924,7 @@ static void mixer_disable_vblank(struct exynos_drm_manager *mgr)
 static void mixer_win_mode_set(struct exynos_drm_manager *mgr,
 			struct exynos_drm_overlay *overlay)
 {
-	struct mixer_context *mixer_ctx = mgr->ctx;
+	struct mixer_context *mixer_ctx = mgr_to_mixer(mgr);
 	struct hdmi_win_data *win_data;
 	int win;
 
@@ -971,7 +975,7 @@ static void mixer_win_mode_set(struct exynos_drm_manager *mgr,
 
 static void mixer_win_commit(struct exynos_drm_manager *mgr, int zpos)
 {
-	struct mixer_context *mixer_ctx = mgr->ctx;
+	struct mixer_context *mixer_ctx = mgr_to_mixer(mgr);
 	int win = zpos == DEFAULT_ZPOS ? MIXER_DEFAULT_WIN : zpos;
 
 	DRM_DEBUG_KMS("win: %d\n", win);
@@ -993,7 +997,7 @@ static void mixer_win_commit(struct exynos_drm_manager *mgr, int zpos)
 
 static void mixer_win_disable(struct exynos_drm_manager *mgr, int zpos)
 {
-	struct mixer_context *mixer_ctx = mgr->ctx;
+	struct mixer_context *mixer_ctx = mgr_to_mixer(mgr);
 	struct mixer_resources *res = &mixer_ctx->mixer_res;
 	int win = zpos == DEFAULT_ZPOS ? MIXER_DEFAULT_WIN : zpos;
 	unsigned long flags;
@@ -1021,7 +1025,8 @@ static void mixer_win_disable(struct exynos_drm_manager *mgr, int zpos)
 
 static void mixer_wait_for_vblank(struct exynos_drm_manager *mgr)
 {
-	struct mixer_context *mixer_ctx = mgr->ctx;
+	struct mixer_context *mixer_ctx = mgr_to_mixer(mgr);
+	int err;
 
 	mutex_lock(&mixer_ctx->mixer_mutex);
 	if (!mixer_ctx->powered) {
@@ -1030,7 +1035,11 @@ static void mixer_wait_for_vblank(struct exynos_drm_manager *mgr)
 	}
 	mutex_unlock(&mixer_ctx->mixer_mutex);
 
-	drm_vblank_get(mgr->crtc->dev, mixer_ctx->pipe);
+	err = drm_vblank_get(mgr->crtc->dev, mixer_ctx->pipe);
+	if (err < 0) {
+		DRM_DEBUG_KMS("failed to acquire vblank counter\n");
+		return;
+	}
 
 	atomic_set(&mixer_ctx->wait_vsync_event, 1);
 
@@ -1048,7 +1057,7 @@ static void mixer_wait_for_vblank(struct exynos_drm_manager *mgr)
 
 static void mixer_window_suspend(struct exynos_drm_manager *mgr)
 {
-	struct mixer_context *ctx = mgr->ctx;
+	struct mixer_context *ctx = mgr_to_mixer(mgr);
 	struct hdmi_win_data *win_data;
 	int i;
 
@@ -1062,7 +1071,7 @@ static void mixer_window_suspend(struct exynos_drm_manager *mgr)
 
 static void mixer_window_resume(struct exynos_drm_manager *mgr)
 {
-	struct mixer_context *ctx = mgr->ctx;
+	struct mixer_context *ctx = mgr_to_mixer(mgr);
 	struct hdmi_win_data *win_data;
 	int i;
 
@@ -1077,7 +1086,7 @@ static void mixer_window_resume(struct exynos_drm_manager *mgr)
 
 static void mixer_poweron(struct exynos_drm_manager *mgr)
 {
-	struct mixer_context *ctx = mgr->ctx;
+	struct mixer_context *ctx = mgr_to_mixer(mgr);
 	struct mixer_resources *res = &ctx->mixer_res;
 
 	mutex_lock(&ctx->mixer_mutex);
@@ -1111,7 +1120,7 @@ static void mixer_poweron(struct exynos_drm_manager *mgr)
 
 static void mixer_poweroff(struct exynos_drm_manager *mgr)
 {
-	struct mixer_context *ctx = mgr->ctx;
+	struct mixer_context *ctx = mgr_to_mixer(mgr);
 	struct mixer_resources *res = &ctx->mixer_res;
 
 	mutex_lock(&ctx->mixer_mutex);
@@ -1187,11 +1196,6 @@ static struct exynos_drm_manager_ops mixer_manager_ops = {
 	.win_disable		= mixer_win_disable,
 };
 
-static struct exynos_drm_manager mixer_manager = {
-	.type			= EXYNOS_DISPLAY_TYPE_HDMI,
-	.ops			= &mixer_manager_ops,
-};
-
 static struct mixer_drv_data exynos5420_mxr_drv_data = {
 	.version = MXR_VER_128_0_0_184,
 	.is_vp_enabled = 0,
@@ -1249,13 +1253,41 @@ MODULE_DEVICE_TABLE(of, mixer_match_types);
 
 static int mixer_bind(struct device *dev, struct device *manager, void *data)
 {
-	struct platform_device *pdev = to_platform_device(dev);
+	struct mixer_context *ctx = dev_get_drvdata(dev);
 	struct drm_device *drm_dev = data;
-	struct mixer_context *ctx;
-	struct mixer_drv_data *drv;
 	int ret;
 
-	dev_info(dev, "probe start\n");
+	ret = mixer_initialize(&ctx->manager, drm_dev);
+	if (ret)
+		return ret;
+
+	ret = exynos_drm_crtc_create(&ctx->manager);
+	if (ret) {
+		mixer_mgr_remove(&ctx->manager);
+		return ret;
+	}
+
+	return 0;
+}
+
+static void mixer_unbind(struct device *dev, struct device *master, void *data)
+{
+	struct mixer_context *ctx = dev_get_drvdata(dev);
+
+	mixer_mgr_remove(&ctx->manager);
+}
+
+static const struct component_ops mixer_component_ops = {
+	.bind	= mixer_bind,
+	.unbind	= mixer_unbind,
+};
+
+static int mixer_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct mixer_drv_data *drv;
+	struct mixer_context *ctx;
+	int ret;
 
 	ctx = devm_kzalloc(&pdev->dev, sizeof(*ctx), GFP_KERNEL);
 	if (!ctx) {
@@ -1265,8 +1297,12 @@ static int mixer_bind(struct device *dev, struct device *manager, void *data)
 
 	mutex_init(&ctx->mixer_mutex);
 
+	ctx->manager.type = EXYNOS_DISPLAY_TYPE_HDMI;
+	ctx->manager.ops = &mixer_manager_ops;
+
 	if (dev->of_node) {
 		const struct of_device_id *match;
+
 		match = of_match_node(mixer_match_types, dev->of_node);
 		drv = (struct mixer_drv_data *)match->data;
 	} else {
@@ -1282,60 +1318,28 @@ static int mixer_bind(struct device *dev, struct device *manager, void *data)
 	init_waitqueue_head(&ctx->wait_vsync_queue);
 	atomic_set(&ctx->wait_vsync_event, 0);
 
-	mixer_manager.ctx = ctx;
-	ret = mixer_initialize(&mixer_manager, drm_dev);
+	platform_set_drvdata(pdev, ctx);
+
+	ret = exynos_drm_component_add(&pdev->dev, EXYNOS_DEVICE_TYPE_CRTC,
+					ctx->manager.type);
 	if (ret)
 		return ret;
 
-	platform_set_drvdata(pdev, &mixer_manager);
-	ret = exynos_drm_crtc_create(&mixer_manager);
+	ret = component_add(&pdev->dev, &mixer_component_ops);
 	if (ret) {
-		mixer_mgr_remove(&mixer_manager);
+		exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CRTC);
 		return ret;
 	}
 
 	pm_runtime_enable(dev);
 
-	return 0;
-}
-
-static void mixer_unbind(struct device *dev, struct device *master, void *data)
-{
-	struct exynos_drm_manager *mgr = dev_get_drvdata(dev);
-	struct drm_crtc *crtc = mgr->crtc;
-
-	dev_info(dev, "remove successful\n");
-
-	mixer_mgr_remove(mgr);
-
-	pm_runtime_disable(dev);
-
-	crtc->funcs->destroy(crtc);
-}
-
-static const struct component_ops mixer_component_ops = {
-	.bind	= mixer_bind,
-	.unbind	= mixer_unbind,
-};
-
-static int mixer_probe(struct platform_device *pdev)
-{
-	int ret;
-
-	ret = exynos_drm_component_add(&pdev->dev, EXYNOS_DEVICE_TYPE_CRTC,
-					mixer_manager.type);
-	if (ret)
-		return ret;
-
-	ret = component_add(&pdev->dev, &mixer_component_ops);
-	if (ret)
-		exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CRTC);
-
 	return ret;
 }
 
 static int mixer_remove(struct platform_device *pdev)
 {
+	pm_runtime_disable(&pdev->dev);
+
 	component_del(&pdev->dev, &mixer_component_ops);
 	exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CRTC);
 
diff --git a/drivers/gpu/drm/gma500/Makefile b/drivers/gpu/drm/gma500/Makefile
index b15315576376..190e55f2f891 100644
--- a/drivers/gpu/drm/gma500/Makefile
+++ b/drivers/gpu/drm/gma500/Makefile
@@ -39,6 +39,7 @@ gma500_gfx-$(CONFIG_DRM_GMA3600) +=  cdv_device.o \
 gma500_gfx-$(CONFIG_DRM_GMA600) += oaktrail_device.o \
 	  oaktrail_crtc.o \
 	  oaktrail_lvds.o \
+	  oaktrail_lvds_i2c.o \
 	  oaktrail_hdmi.o \
 	  oaktrail_hdmi_i2c.o
 
diff --git a/drivers/gpu/drm/gma500/cdv_intel_dp.c b/drivers/gpu/drm/gma500/cdv_intel_dp.c
index a4cc0e60a1be..0fafb8e2483a 100644
--- a/drivers/gpu/drm/gma500/cdv_intel_dp.c
+++ b/drivers/gpu/drm/gma500/cdv_intel_dp.c
@@ -37,6 +37,201 @@
 #include "gma_display.h"
 #include <drm/drm_dp_helper.h>
 
+/**
+ * struct i2c_algo_dp_aux_data - driver interface structure for i2c over dp
+ * 				 aux algorithm
+ * @running: set by the algo indicating whether an i2c is ongoing or whether
+ * 	     the i2c bus is quiescent
+ * @address: i2c target address for the currently ongoing transfer
+ * @aux_ch: driver callback to transfer a single byte of the i2c payload
+ */
+struct i2c_algo_dp_aux_data {
+	bool running;
+	u16 address;
+	int (*aux_ch) (struct i2c_adapter *adapter,
+		       int mode, uint8_t write_byte,
+		       uint8_t *read_byte);
+};
+
+/* Run a single AUX_CH I2C transaction, writing/reading data as necessary */
+static int
+i2c_algo_dp_aux_transaction(struct i2c_adapter *adapter, int mode,
+			    uint8_t write_byte, uint8_t *read_byte)
+{
+	struct i2c_algo_dp_aux_data *algo_data = adapter->algo_data;
+	int ret;
+
+	ret = (*algo_data->aux_ch)(adapter, mode,
+				   write_byte, read_byte);
+	return ret;
+}
+
+/*
+ * I2C over AUX CH
+ */
+
+/*
+ * Send the address. If the I2C link is running, this 'restarts'
+ * the connection with the new address, this is used for doing
+ * a write followed by a read (as needed for DDC)
+ */
+static int
+i2c_algo_dp_aux_address(struct i2c_adapter *adapter, u16 address, bool reading)
+{
+	struct i2c_algo_dp_aux_data *algo_data = adapter->algo_data;
+	int mode = MODE_I2C_START;
+	int ret;
+
+	if (reading)
+		mode |= MODE_I2C_READ;
+	else
+		mode |= MODE_I2C_WRITE;
+	algo_data->address = address;
+	algo_data->running = true;
+	ret = i2c_algo_dp_aux_transaction(adapter, mode, 0, NULL);
+	return ret;
+}
+
+/*
+ * Stop the I2C transaction. This closes out the link, sending
+ * a bare address packet with the MOT bit turned off
+ */
+static void
+i2c_algo_dp_aux_stop(struct i2c_adapter *adapter, bool reading)
+{
+	struct i2c_algo_dp_aux_data *algo_data = adapter->algo_data;
+	int mode = MODE_I2C_STOP;
+
+	if (reading)
+		mode |= MODE_I2C_READ;
+	else
+		mode |= MODE_I2C_WRITE;
+	if (algo_data->running) {
+		(void) i2c_algo_dp_aux_transaction(adapter, mode, 0, NULL);
+		algo_data->running = false;
+	}
+}
+
+/*
+ * Write a single byte to the current I2C address, the
+ * the I2C link must be running or this returns -EIO
+ */
+static int
+i2c_algo_dp_aux_put_byte(struct i2c_adapter *adapter, u8 byte)
+{
+	struct i2c_algo_dp_aux_data *algo_data = adapter->algo_data;
+	int ret;
+
+	if (!algo_data->running)
+		return -EIO;
+
+	ret = i2c_algo_dp_aux_transaction(adapter, MODE_I2C_WRITE, byte, NULL);
+	return ret;
+}
+
+/*
+ * Read a single byte from the current I2C address, the
+ * I2C link must be running or this returns -EIO
+ */
+static int
+i2c_algo_dp_aux_get_byte(struct i2c_adapter *adapter, u8 *byte_ret)
+{
+	struct i2c_algo_dp_aux_data *algo_data = adapter->algo_data;
+	int ret;
+
+	if (!algo_data->running)
+		return -EIO;
+
+	ret = i2c_algo_dp_aux_transaction(adapter, MODE_I2C_READ, 0, byte_ret);
+	return ret;
+}
+
+static int
+i2c_algo_dp_aux_xfer(struct i2c_adapter *adapter,
+		     struct i2c_msg *msgs,
+		     int num)
+{
+	int ret = 0;
+	bool reading = false;
+	int m;
+	int b;
+
+	for (m = 0; m < num; m++) {
+		u16 len = msgs[m].len;
+		u8 *buf = msgs[m].buf;
+		reading = (msgs[m].flags & I2C_M_RD) != 0;
+		ret = i2c_algo_dp_aux_address(adapter, msgs[m].addr, reading);
+		if (ret < 0)
+			break;
+		if (reading) {
+			for (b = 0; b < len; b++) {
+				ret = i2c_algo_dp_aux_get_byte(adapter, &buf[b]);
+				if (ret < 0)
+					break;
+			}
+		} else {
+			for (b = 0; b < len; b++) {
+				ret = i2c_algo_dp_aux_put_byte(adapter, buf[b]);
+				if (ret < 0)
+					break;
+			}
+		}
+		if (ret < 0)
+			break;
+	}
+	if (ret >= 0)
+		ret = num;
+	i2c_algo_dp_aux_stop(adapter, reading);
+	DRM_DEBUG_KMS("dp_aux_xfer return %d\n", ret);
+	return ret;
+}
+
+static u32
+i2c_algo_dp_aux_functionality(struct i2c_adapter *adapter)
+{
+	return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL |
+	       I2C_FUNC_SMBUS_READ_BLOCK_DATA |
+	       I2C_FUNC_SMBUS_BLOCK_PROC_CALL |
+	       I2C_FUNC_10BIT_ADDR;
+}
+
+static const struct i2c_algorithm i2c_dp_aux_algo = {
+	.master_xfer	= i2c_algo_dp_aux_xfer,
+	.functionality	= i2c_algo_dp_aux_functionality,
+};
+
+static void
+i2c_dp_aux_reset_bus(struct i2c_adapter *adapter)
+{
+	(void) i2c_algo_dp_aux_address(adapter, 0, false);
+	(void) i2c_algo_dp_aux_stop(adapter, false);
+}
+
+static int
+i2c_dp_aux_prepare_bus(struct i2c_adapter *adapter)
+{
+	adapter->algo = &i2c_dp_aux_algo;
+	adapter->retries = 3;
+	i2c_dp_aux_reset_bus(adapter);
+	return 0;
+}
+
+/*
+ * FIXME: This is the old dp aux helper, gma500 is the last driver that needs to
+ * be ported over to the new helper code in drm_dp_helper.c like i915 or radeon.
+ */
+static int __deprecated
+i2c_dp_aux_add_bus(struct i2c_adapter *adapter)
+{
+	int error;
+
+	error = i2c_dp_aux_prepare_bus(adapter);
+	if (error)
+		return error;
+	error = i2c_add_adapter(adapter);
+	return error;
+}
+
 #define _wait_for(COND, MS, W) ({ \
         unsigned long timeout__ = jiffies + msecs_to_jiffies(MS);       \
         int ret__ = 0;                                                  \
@@ -1089,7 +1284,7 @@ static char	*link_train_names[] = {
 };
 #endif
 
-#define CDV_DP_VOLTAGE_MAX	    DP_TRAIN_VOLTAGE_SWING_1200
+#define CDV_DP_VOLTAGE_MAX	    DP_TRAIN_VOLTAGE_SWING_LEVEL_3
 /*
 static uint8_t
 cdv_intel_dp_pre_emphasis_max(uint8_t voltage_swing)
@@ -1276,7 +1471,7 @@ cdv_intel_dp_set_vswing_premph(struct gma_encoder *encoder, uint8_t signal_level
 		cdv_sb_write(dev, ddi_reg->VSwing2, dp_vswing_premph_table[index]);
 
 	/* ;gfx_dpio_set_reg(0x814c, 0x40802040) */
-	if ((vswing + premph) == DP_TRAIN_VOLTAGE_SWING_1200)
+	if ((vswing + premph) == DP_TRAIN_VOLTAGE_SWING_LEVEL_3)
 		cdv_sb_write(dev, ddi_reg->VSwing3, 0x70802040);
 	else
 		cdv_sb_write(dev, ddi_reg->VSwing3, 0x40802040);
diff --git a/drivers/gpu/drm/gma500/framebuffer.c b/drivers/gpu/drm/gma500/framebuffer.c
index d0dd3bea8aa5..ddd90ddbc200 100644
--- a/drivers/gpu/drm/gma500/framebuffer.c
+++ b/drivers/gpu/drm/gma500/framebuffer.c
@@ -540,7 +540,8 @@ static void psbfb_gamma_get(struct drm_crtc *crtc, u16 *red,
 static int psbfb_probe(struct drm_fb_helper *helper,
 				struct drm_fb_helper_surface_size *sizes)
 {
-	struct psb_fbdev *psb_fbdev = (struct psb_fbdev *)helper;
+	struct psb_fbdev *psb_fbdev =
+		container_of(helper, struct psb_fbdev, psb_fb_helper);
 	struct drm_device *dev = psb_fbdev->psb_fb_helper.dev;
 	struct drm_psb_private *dev_priv = dev->dev_private;
 	int bytespp;
diff --git a/drivers/gpu/drm/gma500/gtt.h b/drivers/gpu/drm/gma500/gtt.h
index f5860a739bd8..cdbb350c9d5d 100644
--- a/drivers/gpu/drm/gma500/gtt.h
+++ b/drivers/gpu/drm/gma500/gtt.h
@@ -21,6 +21,7 @@
 #define _PSB_GTT_H_
 
 #include <drm/drmP.h>
+#include <drm/drm_gem.h>
 
 /* This wants cleaning up with respect to the psb_dev and un-needed stuff */
 struct psb_gtt {
diff --git a/drivers/gpu/drm/gma500/intel_bios.c b/drivers/gpu/drm/gma500/intel_bios.c
index d3497348c4d5..63bde4e86c6a 100644
--- a/drivers/gpu/drm/gma500/intel_bios.c
+++ b/drivers/gpu/drm/gma500/intel_bios.c
@@ -116,30 +116,30 @@ parse_edp(struct drm_psb_private *dev_priv, struct bdb_header *bdb)
 
 	switch (edp_link_params->preemphasis) {
 	case 0:
-		dev_priv->edp.preemphasis = DP_TRAIN_PRE_EMPHASIS_0;
+		dev_priv->edp.preemphasis = DP_TRAIN_PRE_EMPH_LEVEL_0;
 		break;
 	case 1:
-		dev_priv->edp.preemphasis = DP_TRAIN_PRE_EMPHASIS_3_5;
+		dev_priv->edp.preemphasis = DP_TRAIN_PRE_EMPH_LEVEL_1;
 		break;
 	case 2:
-		dev_priv->edp.preemphasis = DP_TRAIN_PRE_EMPHASIS_6;
+		dev_priv->edp.preemphasis = DP_TRAIN_PRE_EMPH_LEVEL_2;
 		break;
 	case 3:
-		dev_priv->edp.preemphasis = DP_TRAIN_PRE_EMPHASIS_9_5;
+		dev_priv->edp.preemphasis = DP_TRAIN_PRE_EMPH_LEVEL_3;
 		break;
 	}
 	switch (edp_link_params->vswing) {
 	case 0:
-		dev_priv->edp.vswing = DP_TRAIN_VOLTAGE_SWING_400;
+		dev_priv->edp.vswing = DP_TRAIN_VOLTAGE_SWING_LEVEL_0;
 		break;
 	case 1:
-		dev_priv->edp.vswing = DP_TRAIN_VOLTAGE_SWING_600;
+		dev_priv->edp.vswing = DP_TRAIN_VOLTAGE_SWING_LEVEL_1;
 		break;
 	case 2:
-		dev_priv->edp.vswing = DP_TRAIN_VOLTAGE_SWING_800;
+		dev_priv->edp.vswing = DP_TRAIN_VOLTAGE_SWING_LEVEL_2;
 		break;
 	case 3:
-		dev_priv->edp.vswing = DP_TRAIN_VOLTAGE_SWING_1200;
+		dev_priv->edp.vswing = DP_TRAIN_VOLTAGE_SWING_LEVEL_3;
 		break;
 	}
 	DRM_DEBUG_KMS("VBT reports EDP: VSwing  %d, Preemph %d\n",
diff --git a/drivers/gpu/drm/gma500/mdfld_dsi_pkg_sender.c b/drivers/gpu/drm/gma500/mdfld_dsi_pkg_sender.c
index 87885d8c06e8..6b43ae3ffd73 100644
--- a/drivers/gpu/drm/gma500/mdfld_dsi_pkg_sender.c
+++ b/drivers/gpu/drm/gma500/mdfld_dsi_pkg_sender.c
@@ -25,6 +25,7 @@
  */
 
 #include <linux/freezer.h>
+#include <video/mipi_display.h>
 
 #include "mdfld_dsi_output.h"
 #include "mdfld_dsi_pkg_sender.h"
@@ -32,20 +33,6 @@
 
 #define MDFLD_DSI_READ_MAX_COUNT		5000
 
-enum data_type {
-	DSI_DT_GENERIC_SHORT_WRITE_0	= 0x03,
-	DSI_DT_GENERIC_SHORT_WRITE_1	= 0x13,
-	DSI_DT_GENERIC_SHORT_WRITE_2	= 0x23,
-	DSI_DT_GENERIC_READ_0		= 0x04,
-	DSI_DT_GENERIC_READ_1		= 0x14,
-	DSI_DT_GENERIC_READ_2		= 0x24,
-	DSI_DT_GENERIC_LONG_WRITE	= 0x29,
-	DSI_DT_DCS_SHORT_WRITE_0	= 0x05,
-	DSI_DT_DCS_SHORT_WRITE_1	= 0x15,
-	DSI_DT_DCS_READ			= 0x06,
-	DSI_DT_DCS_LONG_WRITE		= 0x39,
-};
-
 enum {
 	MDFLD_DSI_PANEL_MODE_SLEEP = 0x1,
 };
@@ -321,9 +308,9 @@ static int send_pkg_prepare(struct mdfld_dsi_pkg_sender *sender, u8 data_type,
 	u8 cmd;
 
 	switch (data_type) {
-	case DSI_DT_DCS_SHORT_WRITE_0:
-	case DSI_DT_DCS_SHORT_WRITE_1:
-	case DSI_DT_DCS_LONG_WRITE:
+	case MIPI_DSI_DCS_SHORT_WRITE:
+	case MIPI_DSI_DCS_SHORT_WRITE_PARAM:
+	case MIPI_DSI_DCS_LONG_WRITE:
 		cmd = *data;
 		break;
 	default:
@@ -334,12 +321,12 @@ static int send_pkg_prepare(struct mdfld_dsi_pkg_sender *sender, u8 data_type,
 	sender->status = MDFLD_DSI_PKG_SENDER_BUSY;
 
 	/*wait for 120 milliseconds in case exit_sleep_mode just be sent*/
-	if (unlikely(cmd == DCS_ENTER_SLEEP_MODE)) {
+	if (unlikely(cmd == MIPI_DCS_ENTER_SLEEP_MODE)) {
 		/*TODO: replace it with msleep later*/
 		mdelay(120);
 	}
 
-	if (unlikely(cmd == DCS_EXIT_SLEEP_MODE)) {
+	if (unlikely(cmd == MIPI_DCS_EXIT_SLEEP_MODE)) {
 		/*TODO: replace it with msleep later*/
 		mdelay(120);
 	}
@@ -352,9 +339,9 @@ static int send_pkg_done(struct mdfld_dsi_pkg_sender *sender, u8 data_type,
 	u8 cmd;
 
 	switch (data_type) {
-	case DSI_DT_DCS_SHORT_WRITE_0:
-	case DSI_DT_DCS_SHORT_WRITE_1:
-	case DSI_DT_DCS_LONG_WRITE:
+	case MIPI_DSI_DCS_SHORT_WRITE:
+	case MIPI_DSI_DCS_SHORT_WRITE_PARAM:
+	case MIPI_DSI_DCS_LONG_WRITE:
 		cmd = *data;
 		break;
 	default:
@@ -362,15 +349,15 @@ static int send_pkg_done(struct mdfld_dsi_pkg_sender *sender, u8 data_type,
 	}
 
 	/*update panel status*/
-	if (unlikely(cmd == DCS_ENTER_SLEEP_MODE)) {
+	if (unlikely(cmd == MIPI_DCS_ENTER_SLEEP_MODE)) {
 		sender->panel_mode |= MDFLD_DSI_PANEL_MODE_SLEEP;
 		/*TODO: replace it with msleep later*/
 		mdelay(120);
-	} else if (unlikely(cmd == DCS_EXIT_SLEEP_MODE)) {
+	} else if (unlikely(cmd == MIPI_DCS_EXIT_SLEEP_MODE)) {
 		sender->panel_mode &= ~MDFLD_DSI_PANEL_MODE_SLEEP;
 		/*TODO: replace it with msleep later*/
 		mdelay(120);
-	} else if (unlikely(cmd == DCS_SOFT_RESET)) {
+	} else if (unlikely(cmd == MIPI_DCS_SOFT_RESET)) {
 		/*TODO: replace it with msleep later*/
 		mdelay(5);
 	}
@@ -405,19 +392,19 @@ static int send_pkg(struct mdfld_dsi_pkg_sender *sender, u8 data_type,
 	}
 
 	switch (data_type) {
-	case DSI_DT_GENERIC_SHORT_WRITE_0:
-	case DSI_DT_GENERIC_SHORT_WRITE_1:
-	case DSI_DT_GENERIC_SHORT_WRITE_2:
-	case DSI_DT_GENERIC_READ_0:
-	case DSI_DT_GENERIC_READ_1:
-	case DSI_DT_GENERIC_READ_2:
-	case DSI_DT_DCS_SHORT_WRITE_0:
-	case DSI_DT_DCS_SHORT_WRITE_1:
-	case DSI_DT_DCS_READ:
+	case MIPI_DSI_GENERIC_SHORT_WRITE_0_PARAM:
+	case MIPI_DSI_GENERIC_SHORT_WRITE_1_PARAM:
+	case MIPI_DSI_GENERIC_SHORT_WRITE_2_PARAM:
+	case MIPI_DSI_GENERIC_READ_REQUEST_0_PARAM:
+	case MIPI_DSI_GENERIC_READ_REQUEST_1_PARAM:
+	case MIPI_DSI_GENERIC_READ_REQUEST_2_PARAM:
+	case MIPI_DSI_DCS_SHORT_WRITE:
+	case MIPI_DSI_DCS_SHORT_WRITE_PARAM:
+	case MIPI_DSI_DCS_READ:
 		ret = send_short_pkg(sender, data_type, data[0], data[1], hs);
 		break;
-	case DSI_DT_GENERIC_LONG_WRITE:
-	case DSI_DT_DCS_LONG_WRITE:
+	case MIPI_DSI_GENERIC_LONG_WRITE:
+	case MIPI_DSI_DCS_LONG_WRITE:
 		ret = send_long_pkg(sender, data_type, data, len, hs);
 		break;
 	}
@@ -440,7 +427,7 @@ int mdfld_dsi_send_mcs_long(struct mdfld_dsi_pkg_sender *sender, u8 *data,
 	}
 
 	spin_lock_irqsave(&sender->lock, flags);
-	send_pkg(sender, DSI_DT_DCS_LONG_WRITE, data, len, hs);
+	send_pkg(sender, MIPI_DSI_DCS_LONG_WRITE, data, len, hs);
 	spin_unlock_irqrestore(&sender->lock, flags);
 
 	return 0;
@@ -461,10 +448,10 @@ int mdfld_dsi_send_mcs_short(struct mdfld_dsi_pkg_sender *sender, u8 cmd,
 	data[0] = cmd;
 
 	if (param_num) {
-		data_type = DSI_DT_DCS_SHORT_WRITE_1;
+		data_type = MIPI_DSI_DCS_SHORT_WRITE_PARAM;
 		data[1] = param;
 	} else {
-		data_type = DSI_DT_DCS_SHORT_WRITE_0;
+		data_type = MIPI_DSI_DCS_SHORT_WRITE;
 		data[1] = 0;
 	}
 
@@ -489,17 +476,17 @@ int mdfld_dsi_send_gen_short(struct mdfld_dsi_pkg_sender *sender, u8 param0,
 
 	switch (param_num) {
 	case 0:
-		data_type = DSI_DT_GENERIC_SHORT_WRITE_0;
+		data_type = MIPI_DSI_GENERIC_SHORT_WRITE_0_PARAM;
 		data[0] = 0;
 		data[1] = 0;
 		break;
 	case 1:
-		data_type = DSI_DT_GENERIC_SHORT_WRITE_1;
+		data_type = MIPI_DSI_GENERIC_SHORT_WRITE_1_PARAM;
 		data[0] = param0;
 		data[1] = 0;
 		break;
 	case 2:
-		data_type = DSI_DT_GENERIC_SHORT_WRITE_2;
+		data_type = MIPI_DSI_GENERIC_SHORT_WRITE_2_PARAM;
 		data[0] = param0;
 		data[1] = param1;
 		break;
@@ -523,7 +510,7 @@ int mdfld_dsi_send_gen_long(struct mdfld_dsi_pkg_sender *sender, u8 *data,
 	}
 
 	spin_lock_irqsave(&sender->lock, flags);
-	send_pkg(sender, DSI_DT_GENERIC_LONG_WRITE, data, len, hs);
+	send_pkg(sender, MIPI_DSI_GENERIC_LONG_WRITE, data, len, hs);
 	spin_unlock_irqrestore(&sender->lock, flags);
 
 	return 0;
@@ -594,7 +581,7 @@ int mdfld_dsi_read_mcs(struct mdfld_dsi_pkg_sender *sender, u8 cmd,
 		return -EINVAL;
 	}
 
-	return __read_panel_data(sender, DSI_DT_DCS_READ, &cmd, 1,
+	return __read_panel_data(sender, MIPI_DSI_DCS_READ, &cmd, 1,
 				data, len, hs);
 }
 
diff --git a/drivers/gpu/drm/gma500/mdfld_dsi_pkg_sender.h b/drivers/gpu/drm/gma500/mdfld_dsi_pkg_sender.h
index 459cd7ea8b81..0478a21c15d5 100644
--- a/drivers/gpu/drm/gma500/mdfld_dsi_pkg_sender.h
+++ b/drivers/gpu/drm/gma500/mdfld_dsi_pkg_sender.h
@@ -62,18 +62,6 @@ struct mdfld_dsi_pkg_sender {
 	u32 mipi_cmd_len_reg;
 };
 
-/* DCS definitions */
-#define DCS_SOFT_RESET			0x01
-#define DCS_ENTER_SLEEP_MODE		0x10
-#define DCS_EXIT_SLEEP_MODE		0x11
-#define DCS_SET_DISPLAY_OFF		0x28
-#define DCS_SET_DISPLAY_ON		0x29
-#define DCS_SET_COLUMN_ADDRESS		0x2a
-#define DCS_SET_PAGE_ADDRESS		0x2b
-#define DCS_WRITE_MEM_START		0x2c
-#define DCS_SET_TEAR_OFF		0x34
-#define DCS_SET_TEAR_ON			0x35
-
 extern int mdfld_dsi_pkg_sender_init(struct mdfld_dsi_connector *dsi_connector,
 					int pipe);
 extern void mdfld_dsi_pkg_sender_destroy(struct mdfld_dsi_pkg_sender *sender);
diff --git a/drivers/gpu/drm/gma500/oaktrail_lvds.c b/drivers/gpu/drm/gma500/oaktrail_lvds.c
index 0d39da6e8b7a..83bbc271bcfb 100644
--- a/drivers/gpu/drm/gma500/oaktrail_lvds.c
+++ b/drivers/gpu/drm/gma500/oaktrail_lvds.c
@@ -359,22 +359,26 @@ void oaktrail_lvds_init(struct drm_device *dev,
 	 *    if closed, act like it's not there for now
 	 */
 
+	edid = NULL;
 	mutex_lock(&dev->mode_config.mutex);
 	i2c_adap = i2c_get_adapter(dev_priv->ops->i2c_bus);
-	if (i2c_adap == NULL)
-		dev_err(dev->dev, "No ddc adapter available!\n");
+	if (i2c_adap)
+		edid = drm_get_edid(connector, i2c_adap);
+	if (edid == NULL && dev_priv->lpc_gpio_base) {
+		oaktrail_lvds_i2c_init(encoder);
+		if (gma_encoder->ddc_bus != NULL) {
+			i2c_adap = &gma_encoder->ddc_bus->adapter;
+			edid = drm_get_edid(connector, i2c_adap);
+		}
+	}
 	/*
 	 * Attempt to get the fixed panel mode from DDC.  Assume that the
 	 * preferred mode is the right one.
 	 */
-	if (i2c_adap) {
-		edid = drm_get_edid(connector, i2c_adap);
-		if (edid) {
-			drm_mode_connector_update_edid_property(connector,
-									edid);
-			drm_add_edid_modes(connector, edid);
-			kfree(edid);
-		}
+	if (edid) {
+		drm_mode_connector_update_edid_property(connector, edid);
+		drm_add_edid_modes(connector, edid);
+		kfree(edid);
 
 		list_for_each_entry(scan, &connector->probed_modes, head) {
 			if (scan->type & DRM_MODE_TYPE_PREFERRED) {
@@ -383,7 +387,8 @@ void oaktrail_lvds_init(struct drm_device *dev,
 				goto out;	/* FIXME: check for quirks */
 			}
 		}
-	}
+	} else
+		dev_err(dev->dev, "No ddc adapter available!\n");
 	/*
 	 * If we didn't get EDID, try geting panel timing
 	 * from configuration data
@@ -411,8 +416,10 @@ failed_find:
 	mutex_unlock(&dev->mode_config.mutex);
 
 	dev_dbg(dev->dev, "No LVDS modes found, disabling.\n");
-	if (gma_encoder->ddc_bus)
+	if (gma_encoder->ddc_bus) {
 		psb_intel_i2c_destroy(gma_encoder->ddc_bus);
+		gma_encoder->ddc_bus = NULL;
+	}
 
 /* failed_ddc: */
 
diff --git a/drivers/gpu/drm/gma500/oaktrail_lvds_i2c.c b/drivers/gpu/drm/gma500/oaktrail_lvds_i2c.c
new file mode 100644
index 000000000000..f913a62eee5f
--- /dev/null
+++ b/drivers/gpu/drm/gma500/oaktrail_lvds_i2c.c
@@ -0,0 +1,170 @@
+/*
+ * Copyright (c) 2002-2010, Intel Corporation.
+ * Copyright (c) 2014 ATRON electronic GmbH
+ *   Author: Jan Safrata <jan.nikitenko@gmail.com>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/types.h>
+#include <linux/i2c.h>
+#include <linux/i2c-algo-bit.h>
+#include <linux/init.h>
+#include <linux/io.h>
+#include <linux/delay.h>
+
+#include <drm/drmP.h>
+#include "psb_drv.h"
+#include "psb_intel_reg.h"
+
+
+/*
+ * LPC GPIO based I2C bus for LVDS of Atom E6xx
+ */
+
+/*-----------------------------------------------------------------------------
+ * LPC Register Offsets. Used for LVDS GPIO Bit Bashing. Registers are part
+ * Atom E6xx [D31:F0]
+ ----------------------------------------------------------------------------*/
+#define RGEN    0x20
+#define RGIO    0x24
+#define RGLVL   0x28
+#define RGTPE   0x2C
+#define RGTNE   0x30
+#define RGGPE   0x34
+#define RGSMI   0x38
+#define RGTS    0x3C
+
+/* The LVDS GPIO clock lines are GPIOSUS[3]
+ * The LVDS GPIO data lines are GPIOSUS[4]
+ */
+#define GPIO_CLOCK	0x08
+#define GPIO_DATA	0x10
+
+#define LPC_READ_REG(chan, r) inl((chan)->reg + (r))
+#define LPC_WRITE_REG(chan, r, val) outl((val), (chan)->reg + (r))
+
+static int get_clock(void *data)
+{
+	struct psb_intel_i2c_chan *chan = data;
+	u32 val, tmp;
+
+	val = LPC_READ_REG(chan, RGIO);
+	val |= GPIO_CLOCK;
+	LPC_WRITE_REG(chan, RGIO, val);
+	tmp = LPC_READ_REG(chan, RGLVL);
+	val = (LPC_READ_REG(chan, RGLVL) & GPIO_CLOCK) ? 1 : 0;
+
+	return val;
+}
+
+static int get_data(void *data)
+{
+	struct psb_intel_i2c_chan *chan = data;
+	u32 val, tmp;
+
+	val = LPC_READ_REG(chan, RGIO);
+	val |= GPIO_DATA;
+	LPC_WRITE_REG(chan, RGIO, val);
+	tmp = LPC_READ_REG(chan, RGLVL);
+	val = (LPC_READ_REG(chan, RGLVL) & GPIO_DATA) ? 1 : 0;
+
+	return val;
+}
+
+static void set_clock(void *data, int state_high)
+{
+	struct psb_intel_i2c_chan *chan = data;
+	u32 val;
+
+	if (state_high) {
+		val = LPC_READ_REG(chan, RGIO);
+		val |= GPIO_CLOCK;
+		LPC_WRITE_REG(chan, RGIO, val);
+	} else {
+		val = LPC_READ_REG(chan, RGIO);
+		val &= ~GPIO_CLOCK;
+		LPC_WRITE_REG(chan, RGIO, val);
+		val = LPC_READ_REG(chan, RGLVL);
+		val &= ~GPIO_CLOCK;
+		LPC_WRITE_REG(chan, RGLVL, val);
+	}
+}
+
+static void set_data(void *data, int state_high)
+{
+	struct psb_intel_i2c_chan *chan = data;
+	u32 val;
+
+	if (state_high) {
+		val = LPC_READ_REG(chan, RGIO);
+		val |= GPIO_DATA;
+		LPC_WRITE_REG(chan, RGIO, val);
+	} else {
+		val = LPC_READ_REG(chan, RGIO);
+		val &= ~GPIO_DATA;
+		LPC_WRITE_REG(chan, RGIO, val);
+		val = LPC_READ_REG(chan, RGLVL);
+		val &= ~GPIO_DATA;
+		LPC_WRITE_REG(chan, RGLVL, val);
+	}
+}
+
+void oaktrail_lvds_i2c_init(struct drm_encoder *encoder)
+{
+	struct drm_device *dev = encoder->dev;
+	struct gma_encoder *gma_encoder = to_gma_encoder(encoder);
+	struct drm_psb_private *dev_priv = dev->dev_private;
+	struct psb_intel_i2c_chan *chan;
+
+	chan = kzalloc(sizeof(struct psb_intel_i2c_chan), GFP_KERNEL);
+	if (!chan)
+		return;
+
+	chan->drm_dev = dev;
+	chan->reg = dev_priv->lpc_gpio_base;
+	strncpy(chan->adapter.name, "gma500 LPC",  I2C_NAME_SIZE - 1);
+	chan->adapter.owner = THIS_MODULE;
+	chan->adapter.algo_data = &chan->algo;
+	chan->adapter.dev.parent = &dev->pdev->dev;
+	chan->algo.setsda = set_data;
+	chan->algo.setscl = set_clock;
+	chan->algo.getsda = get_data;
+	chan->algo.getscl = get_clock;
+	chan->algo.udelay = 100;
+	chan->algo.timeout = usecs_to_jiffies(2200);
+	chan->algo.data = chan;
+
+	i2c_set_adapdata(&chan->adapter, chan);
+
+	set_data(chan, 1);
+	set_clock(chan, 1);
+	udelay(50);
+
+	if (i2c_bit_add_bus(&chan->adapter)) {
+		kfree(chan);
+		return;
+	}
+
+	gma_encoder->ddc_bus = chan;
+}
diff --git a/drivers/gpu/drm/gma500/psb_drv.c b/drivers/gpu/drm/gma500/psb_drv.c
index eec993f93b1a..92e7e5795398 100644
--- a/drivers/gpu/drm/gma500/psb_drv.c
+++ b/drivers/gpu/drm/gma500/psb_drv.c
@@ -212,6 +212,8 @@ static int psb_driver_unload(struct drm_device *dev)
 		}
 		if (dev_priv->aux_pdev)
 			pci_dev_put(dev_priv->aux_pdev);
+		if (dev_priv->lpc_pdev)
+			pci_dev_put(dev_priv->lpc_pdev);
 
 		/* Destroy VBT data */
 		psb_intel_destroy_bios(dev);
@@ -280,6 +282,24 @@ static int psb_driver_load(struct drm_device *dev, unsigned long flags)
 			DRM_DEBUG_KMS("Couldn't find aux pci device");
 		}
 		dev_priv->gmbus_reg = dev_priv->aux_reg;
+
+		dev_priv->lpc_pdev = pci_get_bus_and_slot(0, PCI_DEVFN(31, 0));
+		if (dev_priv->lpc_pdev) {
+			pci_read_config_word(dev_priv->lpc_pdev, PSB_LPC_GBA,
+				&dev_priv->lpc_gpio_base);
+			pci_write_config_dword(dev_priv->lpc_pdev, PSB_LPC_GBA,
+				(u32)dev_priv->lpc_gpio_base | (1L<<31));
+			pci_read_config_word(dev_priv->lpc_pdev, PSB_LPC_GBA,
+				&dev_priv->lpc_gpio_base);
+			dev_priv->lpc_gpio_base &= 0xffc0;
+			if (dev_priv->lpc_gpio_base)
+				DRM_DEBUG_KMS("Found LPC GPIO at 0x%04x\n",
+						dev_priv->lpc_gpio_base);
+			else {
+				pci_dev_put(dev_priv->lpc_pdev);
+				dev_priv->lpc_pdev = NULL;
+			}
+		}
 	} else {
 		dev_priv->gmbus_reg = dev_priv->vdc_reg;
 	}
@@ -476,6 +496,7 @@ static struct drm_driver driver = {
 	.unload = psb_driver_unload,
 	.lastclose = psb_driver_lastclose,
 	.preclose = psb_driver_preclose,
+	.set_busid = drm_pci_set_busid,
 
 	.num_ioctls = ARRAY_SIZE(psb_ioctls),
 	.device_is_agp = psb_driver_device_is_agp,
diff --git a/drivers/gpu/drm/gma500/psb_drv.h b/drivers/gpu/drm/gma500/psb_drv.h
index 55ebe2bd88dd..e38057b91865 100644
--- a/drivers/gpu/drm/gma500/psb_drv.h
+++ b/drivers/gpu/drm/gma500/psb_drv.h
@@ -83,6 +83,7 @@ enum {
 #define PSB_PGETBL_CTL		 0x2020
 #define _PSB_PGETBL_ENABLED	 0x00000001
 #define PSB_SGX_2D_SLAVE_PORT	 0x4000
+#define PSB_LPC_GBA		 0x44
 
 /* TODO: To get rid of */
 #define PSB_TT_PRIV0_LIMIT	 (256*1024*1024)
@@ -441,6 +442,7 @@ struct psb_ops;
 struct drm_psb_private {
 	struct drm_device *dev;
 	struct pci_dev *aux_pdev; /* Currently only used by mrst */
+	struct pci_dev *lpc_pdev; /* Currently only used by mrst */
 	const struct psb_ops *ops;
 	const struct psb_offset *regmap;
 	
@@ -470,6 +472,7 @@ struct drm_psb_private {
 	uint8_t __iomem *sgx_reg;
 	uint8_t __iomem *vdc_reg;
 	uint8_t __iomem *aux_reg; /* Auxillary vdc pipe regs */
+	uint16_t lpc_gpio_base;
 	uint32_t gatt_free_offset;
 
 	/* Fencing / irq */
diff --git a/drivers/gpu/drm/gma500/psb_intel_display.c b/drivers/gpu/drm/gma500/psb_intel_display.c
index 87b50ba64ed4..b21a09451d1d 100644
--- a/drivers/gpu/drm/gma500/psb_intel_display.c
+++ b/drivers/gpu/drm/gma500/psb_intel_display.c
@@ -21,6 +21,7 @@
 #include <linux/i2c.h>
 
 #include <drm/drmP.h>
+#include <drm/drm_plane_helper.h>
 #include "framebuffer.h"
 #include "psb_drv.h"
 #include "psb_intel_drv.h"
diff --git a/drivers/gpu/drm/gma500/psb_intel_drv.h b/drivers/gpu/drm/gma500/psb_intel_drv.h
index 336bd3aa1a06..860dd2177ca1 100644
--- a/drivers/gpu/drm/gma500/psb_intel_drv.h
+++ b/drivers/gpu/drm/gma500/psb_intel_drv.h
@@ -223,6 +223,7 @@ extern void oaktrail_lvds_init(struct drm_device *dev,
 extern void oaktrail_wait_for_INTR_PKT_SENT(struct drm_device *dev);
 extern void oaktrail_dsi_init(struct drm_device *dev,
 			   struct psb_intel_mode_device *mode_dev);
+extern void oaktrail_lvds_i2c_init(struct drm_encoder *encoder);
 extern void mid_dsi_init(struct drm_device *dev,
 		    struct psb_intel_mode_device *mode_dev, int dsi_num);
 
diff --git a/drivers/gpu/drm/gma500/psb_intel_sdvo.c b/drivers/gpu/drm/gma500/psb_intel_sdvo.c
index 0be96fdb5e28..58529cea575d 100644
--- a/drivers/gpu/drm/gma500/psb_intel_sdvo.c
+++ b/drivers/gpu/drm/gma500/psb_intel_sdvo.c
@@ -1631,57 +1631,8 @@ static int psb_intel_sdvo_get_modes(struct drm_connector *connector)
 	return !list_empty(&connector->probed_modes);
 }
 
-static void
-psb_intel_sdvo_destroy_enhance_property(struct drm_connector *connector)
-{
-	struct psb_intel_sdvo_connector *psb_intel_sdvo_connector = to_psb_intel_sdvo_connector(connector);
-	struct drm_device *dev = connector->dev;
-
-	if (psb_intel_sdvo_connector->left)
-		drm_property_destroy(dev, psb_intel_sdvo_connector->left);
-	if (psb_intel_sdvo_connector->right)
-		drm_property_destroy(dev, psb_intel_sdvo_connector->right);
-	if (psb_intel_sdvo_connector->top)
-		drm_property_destroy(dev, psb_intel_sdvo_connector->top);
-	if (psb_intel_sdvo_connector->bottom)
-		drm_property_destroy(dev, psb_intel_sdvo_connector->bottom);
-	if (psb_intel_sdvo_connector->hpos)
-		drm_property_destroy(dev, psb_intel_sdvo_connector->hpos);
-	if (psb_intel_sdvo_connector->vpos)
-		drm_property_destroy(dev, psb_intel_sdvo_connector->vpos);
-	if (psb_intel_sdvo_connector->saturation)
-		drm_property_destroy(dev, psb_intel_sdvo_connector->saturation);
-	if (psb_intel_sdvo_connector->contrast)
-		drm_property_destroy(dev, psb_intel_sdvo_connector->contrast);
-	if (psb_intel_sdvo_connector->hue)
-		drm_property_destroy(dev, psb_intel_sdvo_connector->hue);
-	if (psb_intel_sdvo_connector->sharpness)
-		drm_property_destroy(dev, psb_intel_sdvo_connector->sharpness);
-	if (psb_intel_sdvo_connector->flicker_filter)
-		drm_property_destroy(dev, psb_intel_sdvo_connector->flicker_filter);
-	if (psb_intel_sdvo_connector->flicker_filter_2d)
-		drm_property_destroy(dev, psb_intel_sdvo_connector->flicker_filter_2d);
-	if (psb_intel_sdvo_connector->flicker_filter_adaptive)
-		drm_property_destroy(dev, psb_intel_sdvo_connector->flicker_filter_adaptive);
-	if (psb_intel_sdvo_connector->tv_luma_filter)
-		drm_property_destroy(dev, psb_intel_sdvo_connector->tv_luma_filter);
-	if (psb_intel_sdvo_connector->tv_chroma_filter)
-		drm_property_destroy(dev, psb_intel_sdvo_connector->tv_chroma_filter);
-	if (psb_intel_sdvo_connector->dot_crawl)
-		drm_property_destroy(dev, psb_intel_sdvo_connector->dot_crawl);
-	if (psb_intel_sdvo_connector->brightness)
-		drm_property_destroy(dev, psb_intel_sdvo_connector->brightness);
-}
-
 static void psb_intel_sdvo_destroy(struct drm_connector *connector)
 {
-	struct psb_intel_sdvo_connector *psb_intel_sdvo_connector = to_psb_intel_sdvo_connector(connector);
-
-	if (psb_intel_sdvo_connector->tv_format)
-		drm_property_destroy(connector->dev,
-				     psb_intel_sdvo_connector->tv_format);
-
-	psb_intel_sdvo_destroy_enhance_property(connector);
 	drm_connector_unregister(connector);
 	drm_connector_cleanup(connector);
 	kfree(connector);
diff --git a/drivers/gpu/drm/i2c/Kconfig b/drivers/gpu/drm/i2c/Kconfig
index 4d341db462a2..22c7ed63a001 100644
--- a/drivers/gpu/drm/i2c/Kconfig
+++ b/drivers/gpu/drm/i2c/Kconfig
@@ -1,6 +1,12 @@
 menu "I2C encoder or helper chips"
      depends on DRM && DRM_KMS_HELPER && I2C
 
+config DRM_I2C_ADV7511
+	tristate "AV7511 encoder"
+	select REGMAP_I2C
+	help
+	  Support for the Analog Device ADV7511(W) and ADV7513 HDMI encoders.
+
 config DRM_I2C_CH7006
 	tristate "Chrontel ch7006 TV encoder"
 	default m if DRM_NOUVEAU
diff --git a/drivers/gpu/drm/i2c/Makefile b/drivers/gpu/drm/i2c/Makefile
index 43aa33baebed..2c72eb584ab7 100644
--- a/drivers/gpu/drm/i2c/Makefile
+++ b/drivers/gpu/drm/i2c/Makefile
@@ -1,5 +1,7 @@
 ccflags-y := -Iinclude/drm
 
+obj-$(CONFIG_DRM_I2C_ADV7511) += adv7511.o
+
 ch7006-y := ch7006_drv.o ch7006_mode.o
 obj-$(CONFIG_DRM_I2C_CH7006) += ch7006.o
 
diff --git a/drivers/gpu/drm/i2c/adv7511.c b/drivers/gpu/drm/i2c/adv7511.c
new file mode 100644
index 000000000000..faf1c0c5ab2e
--- /dev/null
+++ b/drivers/gpu/drm/i2c/adv7511.c
@@ -0,0 +1,1010 @@
+/*
+ * Analog Devices ADV7511 HDMI transmitter driver
+ *
+ * Copyright 2012 Analog Devices Inc.
+ *
+ * Licensed under the GPL-2.
+ */
+
+#include <linux/device.h>
+#include <linux/gpio/consumer.h>
+#include <linux/i2c.h>
+#include <linux/module.h>
+#include <linux/regmap.h>
+#include <linux/slab.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_edid.h>
+#include <drm/drm_encoder_slave.h>
+
+#include "adv7511.h"
+
+struct adv7511 {
+	struct i2c_client *i2c_main;
+	struct i2c_client *i2c_edid;
+
+	struct regmap *regmap;
+	struct regmap *packet_memory_regmap;
+	enum drm_connector_status status;
+	int dpms_mode;
+
+	unsigned int f_tmds;
+
+	unsigned int current_edid_segment;
+	uint8_t edid_buf[256];
+
+	wait_queue_head_t wq;
+	struct drm_encoder *encoder;
+
+	bool embedded_sync;
+	enum adv7511_sync_polarity vsync_polarity;
+	enum adv7511_sync_polarity hsync_polarity;
+	bool rgb;
+
+	struct edid *edid;
+
+	struct gpio_desc *gpio_pd;
+};
+
+static struct adv7511 *encoder_to_adv7511(struct drm_encoder *encoder)
+{
+	return to_encoder_slave(encoder)->slave_priv;
+}
+
+/* ADI recommended values for proper operation. */
+static const struct reg_default adv7511_fixed_registers[] = {
+	{ 0x98, 0x03 },
+	{ 0x9a, 0xe0 },
+	{ 0x9c, 0x30 },
+	{ 0x9d, 0x61 },
+	{ 0xa2, 0xa4 },
+	{ 0xa3, 0xa4 },
+	{ 0xe0, 0xd0 },
+	{ 0xf9, 0x00 },
+	{ 0x55, 0x02 },
+};
+
+/* -----------------------------------------------------------------------------
+ * Register access
+ */
+
+static const uint8_t adv7511_register_defaults[] = {
+	0x12, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 00 */
+	0x00, 0x00, 0x01, 0x0e, 0xbc, 0x18, 0x01, 0x13,
+	0x25, 0x37, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 10 */
+	0x46, 0x62, 0x04, 0xa8, 0x00, 0x00, 0x1c, 0x84,
+	0x1c, 0xbf, 0x04, 0xa8, 0x1e, 0x70, 0x02, 0x1e, /* 20 */
+	0x00, 0x00, 0x04, 0xa8, 0x08, 0x12, 0x1b, 0xac,
+	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 30 */
+	0x00, 0x00, 0x00, 0x80, 0x00, 0x00, 0x00, 0xb0,
+	0x00, 0x50, 0x90, 0x7e, 0x79, 0x70, 0x00, 0x00, /* 40 */
+	0x00, 0xa8, 0x80, 0x00, 0x00, 0x00, 0x00, 0x00,
+	0x00, 0x00, 0x02, 0x0d, 0x00, 0x00, 0x00, 0x00, /* 50 */
+	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 60 */
+	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+	0x01, 0x0a, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 70 */
+	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 80 */
+	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+	0x00, 0x00, 0x00, 0x00, 0xc0, 0x00, 0x00, 0x00, /* 90 */
+	0x0b, 0x02, 0x00, 0x18, 0x5a, 0x60, 0x00, 0x00,
+	0x00, 0x00, 0x80, 0x80, 0x08, 0x04, 0x00, 0x00, /* a0 */
+	0x00, 0x00, 0x00, 0x40, 0x00, 0x00, 0x40, 0x14,
+	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* b0 */
+	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* c0 */
+	0x00, 0x03, 0x00, 0x00, 0x02, 0x00, 0x01, 0x04,
+	0x30, 0xff, 0x80, 0x80, 0x80, 0x00, 0x00, 0x00, /* d0 */
+	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x10, 0x01,
+	0x80, 0x75, 0x00, 0x00, 0x60, 0x00, 0x00, 0x00, /* e0 */
+	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+	0x00, 0x00, 0x00, 0x00, 0x00, 0x75, 0x11, 0x00, /* f0 */
+	0x00, 0x7c, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+};
+
+static bool adv7511_register_volatile(struct device *dev, unsigned int reg)
+{
+	switch (reg) {
+	case ADV7511_REG_CHIP_REVISION:
+	case ADV7511_REG_SPDIF_FREQ:
+	case ADV7511_REG_CTS_AUTOMATIC1:
+	case ADV7511_REG_CTS_AUTOMATIC2:
+	case ADV7511_REG_VIC_DETECTED:
+	case ADV7511_REG_VIC_SEND:
+	case ADV7511_REG_AUX_VIC_DETECTED:
+	case ADV7511_REG_STATUS:
+	case ADV7511_REG_GC(1):
+	case ADV7511_REG_INT(0):
+	case ADV7511_REG_INT(1):
+	case ADV7511_REG_PLL_STATUS:
+	case ADV7511_REG_AN(0):
+	case ADV7511_REG_AN(1):
+	case ADV7511_REG_AN(2):
+	case ADV7511_REG_AN(3):
+	case ADV7511_REG_AN(4):
+	case ADV7511_REG_AN(5):
+	case ADV7511_REG_AN(6):
+	case ADV7511_REG_AN(7):
+	case ADV7511_REG_HDCP_STATUS:
+	case ADV7511_REG_BCAPS:
+	case ADV7511_REG_BKSV(0):
+	case ADV7511_REG_BKSV(1):
+	case ADV7511_REG_BKSV(2):
+	case ADV7511_REG_BKSV(3):
+	case ADV7511_REG_BKSV(4):
+	case ADV7511_REG_DDC_STATUS:
+	case ADV7511_REG_BSTATUS(0):
+	case ADV7511_REG_BSTATUS(1):
+	case ADV7511_REG_CHIP_ID_HIGH:
+	case ADV7511_REG_CHIP_ID_LOW:
+		return true;
+	}
+
+	return false;
+}
+
+static const struct regmap_config adv7511_regmap_config = {
+	.reg_bits = 8,
+	.val_bits = 8,
+
+	.max_register = 0xff,
+	.cache_type = REGCACHE_RBTREE,
+	.reg_defaults_raw = adv7511_register_defaults,
+	.num_reg_defaults_raw = ARRAY_SIZE(adv7511_register_defaults),
+
+	.volatile_reg = adv7511_register_volatile,
+};
+
+/* -----------------------------------------------------------------------------
+ * Hardware configuration
+ */
+
+static void adv7511_set_colormap(struct adv7511 *adv7511, bool enable,
+				 const uint16_t *coeff,
+				 unsigned int scaling_factor)
+{
+	unsigned int i;
+
+	regmap_update_bits(adv7511->regmap, ADV7511_REG_CSC_UPPER(1),
+			   ADV7511_CSC_UPDATE_MODE, ADV7511_CSC_UPDATE_MODE);
+
+	if (enable) {
+		for (i = 0; i < 12; ++i) {
+			regmap_update_bits(adv7511->regmap,
+					   ADV7511_REG_CSC_UPPER(i),
+					   0x1f, coeff[i] >> 8);
+			regmap_write(adv7511->regmap,
+				     ADV7511_REG_CSC_LOWER(i),
+				     coeff[i] & 0xff);
+		}
+	}
+
+	if (enable)
+		regmap_update_bits(adv7511->regmap, ADV7511_REG_CSC_UPPER(0),
+				   0xe0, 0x80 | (scaling_factor << 5));
+	else
+		regmap_update_bits(adv7511->regmap, ADV7511_REG_CSC_UPPER(0),
+				   0x80, 0x00);
+
+	regmap_update_bits(adv7511->regmap, ADV7511_REG_CSC_UPPER(1),
+			   ADV7511_CSC_UPDATE_MODE, 0);
+}
+
+static int adv7511_packet_enable(struct adv7511 *adv7511, unsigned int packet)
+{
+	if (packet & 0xff)
+		regmap_update_bits(adv7511->regmap, ADV7511_REG_PACKET_ENABLE0,
+				   packet, 0xff);
+
+	if (packet & 0xff00) {
+		packet >>= 8;
+		regmap_update_bits(adv7511->regmap, ADV7511_REG_PACKET_ENABLE1,
+				   packet, 0xff);
+	}
+
+	return 0;
+}
+
+static int adv7511_packet_disable(struct adv7511 *adv7511, unsigned int packet)
+{
+	if (packet & 0xff)
+		regmap_update_bits(adv7511->regmap, ADV7511_REG_PACKET_ENABLE0,
+				   packet, 0x00);
+
+	if (packet & 0xff00) {
+		packet >>= 8;
+		regmap_update_bits(adv7511->regmap, ADV7511_REG_PACKET_ENABLE1,
+				   packet, 0x00);
+	}
+
+	return 0;
+}
+
+/* Coefficients for adv7511 color space conversion */
+static const uint16_t adv7511_csc_ycbcr_to_rgb[] = {
+	0x0734, 0x04ad, 0x0000, 0x1c1b,
+	0x1ddc, 0x04ad, 0x1f24, 0x0135,
+	0x0000, 0x04ad, 0x087c, 0x1b77,
+};
+
+static void adv7511_set_config_csc(struct adv7511 *adv7511,
+				   struct drm_connector *connector,
+				   bool rgb)
+{
+	struct adv7511_video_config config;
+	bool output_format_422, output_format_ycbcr;
+	unsigned int mode;
+	uint8_t infoframe[17];
+
+	if (adv7511->edid)
+		config.hdmi_mode = drm_detect_hdmi_monitor(adv7511->edid);
+	else
+		config.hdmi_mode = false;
+
+	hdmi_avi_infoframe_init(&config.avi_infoframe);
+
+	config.avi_infoframe.scan_mode = HDMI_SCAN_MODE_UNDERSCAN;
+
+	if (rgb) {
+		config.csc_enable = false;
+		config.avi_infoframe.colorspace = HDMI_COLORSPACE_RGB;
+	} else {
+		config.csc_scaling_factor = ADV7511_CSC_SCALING_4;
+		config.csc_coefficents = adv7511_csc_ycbcr_to_rgb;
+
+		if ((connector->display_info.color_formats &
+		     DRM_COLOR_FORMAT_YCRCB422) &&
+		    config.hdmi_mode) {
+			config.csc_enable = false;
+			config.avi_infoframe.colorspace =
+				HDMI_COLORSPACE_YUV422;
+		} else {
+			config.csc_enable = true;
+			config.avi_infoframe.colorspace = HDMI_COLORSPACE_RGB;
+		}
+	}
+
+	if (config.hdmi_mode) {
+		mode = ADV7511_HDMI_CFG_MODE_HDMI;
+
+		switch (config.avi_infoframe.colorspace) {
+		case HDMI_COLORSPACE_YUV444:
+			output_format_422 = false;
+			output_format_ycbcr = true;
+			break;
+		case HDMI_COLORSPACE_YUV422:
+			output_format_422 = true;
+			output_format_ycbcr = true;
+			break;
+		default:
+			output_format_422 = false;
+			output_format_ycbcr = false;
+			break;
+		}
+	} else {
+		mode = ADV7511_HDMI_CFG_MODE_DVI;
+		output_format_422 = false;
+		output_format_ycbcr = false;
+	}
+
+	adv7511_packet_disable(adv7511, ADV7511_PACKET_ENABLE_AVI_INFOFRAME);
+
+	adv7511_set_colormap(adv7511, config.csc_enable,
+			     config.csc_coefficents,
+			     config.csc_scaling_factor);
+
+	regmap_update_bits(adv7511->regmap, ADV7511_REG_VIDEO_INPUT_CFG1, 0x81,
+			   (output_format_422 << 7) | output_format_ycbcr);
+
+	regmap_update_bits(adv7511->regmap, ADV7511_REG_HDCP_HDMI_CFG,
+			   ADV7511_HDMI_CFG_MODE_MASK, mode);
+
+	hdmi_avi_infoframe_pack(&config.avi_infoframe, infoframe,
+				sizeof(infoframe));
+
+	/* The AVI infoframe id is not configurable */
+	regmap_bulk_write(adv7511->regmap, ADV7511_REG_AVI_INFOFRAME_VERSION,
+			  infoframe + 1, sizeof(infoframe) - 1);
+
+	adv7511_packet_enable(adv7511, ADV7511_PACKET_ENABLE_AVI_INFOFRAME);
+}
+
+static void adv7511_set_link_config(struct adv7511 *adv7511,
+				    const struct adv7511_link_config *config)
+{
+	/*
+	 * The input style values documented in the datasheet don't match the
+	 * hardware register field values :-(
+	 */
+	static const unsigned int input_styles[4] = { 0, 2, 1, 3 };
+
+	unsigned int clock_delay;
+	unsigned int color_depth;
+	unsigned int input_id;
+
+	clock_delay = (config->clock_delay + 1200) / 400;
+	color_depth = config->input_color_depth == 8 ? 3
+		    : (config->input_color_depth == 10 ? 1 : 2);
+
+	/* TODO Support input ID 6 */
+	if (config->input_colorspace != HDMI_COLORSPACE_YUV422)
+		input_id = config->input_clock == ADV7511_INPUT_CLOCK_DDR
+			 ? 5 : 0;
+	else if (config->input_clock == ADV7511_INPUT_CLOCK_DDR)
+		input_id = config->embedded_sync ? 8 : 7;
+	else if (config->input_clock == ADV7511_INPUT_CLOCK_2X)
+		input_id = config->embedded_sync ? 4 : 3;
+	else
+		input_id = config->embedded_sync ? 2 : 1;
+
+	regmap_update_bits(adv7511->regmap, ADV7511_REG_I2C_FREQ_ID_CFG, 0xf,
+			   input_id);
+	regmap_update_bits(adv7511->regmap, ADV7511_REG_VIDEO_INPUT_CFG1, 0x7e,
+			   (color_depth << 4) |
+			   (input_styles[config->input_style] << 2));
+	regmap_write(adv7511->regmap, ADV7511_REG_VIDEO_INPUT_CFG2,
+		     config->input_justification << 3);
+	regmap_write(adv7511->regmap, ADV7511_REG_TIMING_GEN_SEQ,
+		     config->sync_pulse << 2);
+
+	regmap_write(adv7511->regmap, 0xba, clock_delay << 5);
+
+	adv7511->embedded_sync = config->embedded_sync;
+	adv7511->hsync_polarity = config->hsync_polarity;
+	adv7511->vsync_polarity = config->vsync_polarity;
+	adv7511->rgb = config->input_colorspace == HDMI_COLORSPACE_RGB;
+}
+
+/* -----------------------------------------------------------------------------
+ * Interrupt and hotplug detection
+ */
+
+static bool adv7511_hpd(struct adv7511 *adv7511)
+{
+	unsigned int irq0;
+	int ret;
+
+	ret = regmap_read(adv7511->regmap, ADV7511_REG_INT(0), &irq0);
+	if (ret < 0)
+		return false;
+
+	if (irq0 & ADV7511_INT0_HDP) {
+		regmap_write(adv7511->regmap, ADV7511_REG_INT(0),
+			     ADV7511_INT0_HDP);
+		return true;
+	}
+
+	return false;
+}
+
+static irqreturn_t adv7511_irq_handler(int irq, void *devid)
+{
+	struct adv7511 *adv7511 = devid;
+
+	if (adv7511_hpd(adv7511))
+		drm_helper_hpd_irq_event(adv7511->encoder->dev);
+
+	wake_up_all(&adv7511->wq);
+
+	return IRQ_HANDLED;
+}
+
+static unsigned int adv7511_is_interrupt_pending(struct adv7511 *adv7511,
+						 unsigned int irq)
+{
+	unsigned int irq0, irq1;
+	unsigned int pending;
+	int ret;
+
+	ret = regmap_read(adv7511->regmap, ADV7511_REG_INT(0), &irq0);
+	if (ret < 0)
+		return 0;
+	ret = regmap_read(adv7511->regmap, ADV7511_REG_INT(1), &irq1);
+	if (ret < 0)
+		return 0;
+
+	pending = (irq1 << 8) | irq0;
+
+	return pending & irq;
+}
+
+static int adv7511_wait_for_interrupt(struct adv7511 *adv7511, int irq,
+				      int timeout)
+{
+	unsigned int pending;
+	int ret;
+
+	if (adv7511->i2c_main->irq) {
+		ret = wait_event_interruptible_timeout(adv7511->wq,
+				adv7511_is_interrupt_pending(adv7511, irq),
+				msecs_to_jiffies(timeout));
+		if (ret <= 0)
+			return 0;
+		pending = adv7511_is_interrupt_pending(adv7511, irq);
+	} else {
+		if (timeout < 25)
+			timeout = 25;
+		do {
+			pending = adv7511_is_interrupt_pending(adv7511, irq);
+			if (pending)
+				break;
+			msleep(25);
+			timeout -= 25;
+		} while (timeout >= 25);
+	}
+
+	return pending;
+}
+
+/* -----------------------------------------------------------------------------
+ * EDID retrieval
+ */
+
+static int adv7511_get_edid_block(void *data, u8 *buf, unsigned int block,
+				  size_t len)
+{
+	struct adv7511 *adv7511 = data;
+	struct i2c_msg xfer[2];
+	uint8_t offset;
+	unsigned int i;
+	int ret;
+
+	if (len > 128)
+		return -EINVAL;
+
+	if (adv7511->current_edid_segment != block / 2) {
+		unsigned int status;
+
+		ret = regmap_read(adv7511->regmap, ADV7511_REG_DDC_STATUS,
+				  &status);
+		if (ret < 0)
+			return ret;
+
+		if (status != 2) {
+			regmap_write(adv7511->regmap, ADV7511_REG_EDID_SEGMENT,
+				     block);
+			ret = adv7511_wait_for_interrupt(adv7511,
+					ADV7511_INT0_EDID_READY |
+					ADV7511_INT1_DDC_ERROR, 200);
+
+			if (!(ret & ADV7511_INT0_EDID_READY))
+				return -EIO;
+		}
+
+		regmap_write(adv7511->regmap, ADV7511_REG_INT(0),
+			     ADV7511_INT0_EDID_READY | ADV7511_INT1_DDC_ERROR);
+
+		/* Break this apart, hopefully more I2C controllers will
+		 * support 64 byte transfers than 256 byte transfers
+		 */
+
+		xfer[0].addr = adv7511->i2c_edid->addr;
+		xfer[0].flags = 0;
+		xfer[0].len = 1;
+		xfer[0].buf = &offset;
+		xfer[1].addr = adv7511->i2c_edid->addr;
+		xfer[1].flags = I2C_M_RD;
+		xfer[1].len = 64;
+		xfer[1].buf = adv7511->edid_buf;
+
+		offset = 0;
+
+		for (i = 0; i < 4; ++i) {
+			ret = i2c_transfer(adv7511->i2c_edid->adapter, xfer,
+					   ARRAY_SIZE(xfer));
+			if (ret < 0)
+				return ret;
+			else if (ret != 2)
+				return -EIO;
+
+			xfer[1].buf += 64;
+			offset += 64;
+		}
+
+		adv7511->current_edid_segment = block / 2;
+	}
+
+	if (block % 2 == 0)
+		memcpy(buf, adv7511->edid_buf, len);
+	else
+		memcpy(buf, adv7511->edid_buf + 128, len);
+
+	return 0;
+}
+
+/* -----------------------------------------------------------------------------
+ * Encoder operations
+ */
+
+static int adv7511_get_modes(struct drm_encoder *encoder,
+			     struct drm_connector *connector)
+{
+	struct adv7511 *adv7511 = encoder_to_adv7511(encoder);
+	struct edid *edid;
+	unsigned int count;
+
+	/* Reading the EDID only works if the device is powered */
+	if (adv7511->dpms_mode != DRM_MODE_DPMS_ON) {
+		regmap_write(adv7511->regmap, ADV7511_REG_INT(0),
+			     ADV7511_INT0_EDID_READY | ADV7511_INT1_DDC_ERROR);
+		regmap_update_bits(adv7511->regmap, ADV7511_REG_POWER,
+				   ADV7511_POWER_POWER_DOWN, 0);
+		adv7511->current_edid_segment = -1;
+	}
+
+	edid = drm_do_get_edid(connector, adv7511_get_edid_block, adv7511);
+
+	if (adv7511->dpms_mode != DRM_MODE_DPMS_ON)
+		regmap_update_bits(adv7511->regmap, ADV7511_REG_POWER,
+				   ADV7511_POWER_POWER_DOWN,
+				   ADV7511_POWER_POWER_DOWN);
+
+	kfree(adv7511->edid);
+	adv7511->edid = edid;
+	if (!edid)
+		return 0;
+
+	drm_mode_connector_update_edid_property(connector, edid);
+	count = drm_add_edid_modes(connector, edid);
+
+	adv7511_set_config_csc(adv7511, connector, adv7511->rgb);
+
+	return count;
+}
+
+static void adv7511_encoder_dpms(struct drm_encoder *encoder, int mode)
+{
+	struct adv7511 *adv7511 = encoder_to_adv7511(encoder);
+
+	switch (mode) {
+	case DRM_MODE_DPMS_ON:
+		adv7511->current_edid_segment = -1;
+
+		regmap_write(adv7511->regmap, ADV7511_REG_INT(0),
+			     ADV7511_INT0_EDID_READY | ADV7511_INT1_DDC_ERROR);
+		regmap_update_bits(adv7511->regmap, ADV7511_REG_POWER,
+				   ADV7511_POWER_POWER_DOWN, 0);
+		/*
+		 * Per spec it is allowed to pulse the HDP signal to indicate
+		 * that the EDID information has changed. Some monitors do this
+		 * when they wakeup from standby or are enabled. When the HDP
+		 * goes low the adv7511 is reset and the outputs are disabled
+		 * which might cause the monitor to go to standby again. To
+		 * avoid this we ignore the HDP pin for the first few seconds
+		 * after enabeling the output.
+		 */
+		regmap_update_bits(adv7511->regmap, ADV7511_REG_POWER2,
+				   ADV7511_REG_POWER2_HDP_SRC_MASK,
+				   ADV7511_REG_POWER2_HDP_SRC_NONE);
+		/* Most of the registers are reset during power down or
+		 * when HPD is low
+		 */
+		regcache_sync(adv7511->regmap);
+		break;
+	default:
+		/* TODO: setup additional power down modes */
+		regmap_update_bits(adv7511->regmap, ADV7511_REG_POWER,
+				   ADV7511_POWER_POWER_DOWN,
+				   ADV7511_POWER_POWER_DOWN);
+		regcache_mark_dirty(adv7511->regmap);
+		break;
+	}
+
+	adv7511->dpms_mode = mode;
+}
+
+static enum drm_connector_status
+adv7511_encoder_detect(struct drm_encoder *encoder,
+		       struct drm_connector *connector)
+{
+	struct adv7511 *adv7511 = encoder_to_adv7511(encoder);
+	enum drm_connector_status status;
+	unsigned int val;
+	bool hpd;
+	int ret;
+
+	ret = regmap_read(adv7511->regmap, ADV7511_REG_STATUS, &val);
+	if (ret < 0)
+		return connector_status_disconnected;
+
+	if (val & ADV7511_STATUS_HPD)
+		status = connector_status_connected;
+	else
+		status = connector_status_disconnected;
+
+	hpd = adv7511_hpd(adv7511);
+
+	/* The chip resets itself when the cable is disconnected, so in case
+	 * there is a pending HPD interrupt and the cable is connected there was
+	 * at least one transition from disconnected to connected and the chip
+	 * has to be reinitialized. */
+	if (status == connector_status_connected && hpd &&
+	    adv7511->dpms_mode == DRM_MODE_DPMS_ON) {
+		regcache_mark_dirty(adv7511->regmap);
+		adv7511_encoder_dpms(encoder, adv7511->dpms_mode);
+		adv7511_get_modes(encoder, connector);
+		if (adv7511->status == connector_status_connected)
+			status = connector_status_disconnected;
+	} else {
+		/* Renable HDP sensing */
+		regmap_update_bits(adv7511->regmap, ADV7511_REG_POWER2,
+				   ADV7511_REG_POWER2_HDP_SRC_MASK,
+				   ADV7511_REG_POWER2_HDP_SRC_BOTH);
+	}
+
+	adv7511->status = status;
+	return status;
+}
+
+static int adv7511_encoder_mode_valid(struct drm_encoder *encoder,
+				      struct drm_display_mode *mode)
+{
+	if (mode->clock > 165000)
+		return MODE_CLOCK_HIGH;
+
+	if (mode->flags & DRM_MODE_FLAG_INTERLACE)
+		return MODE_NO_INTERLACE;
+
+	return MODE_OK;
+}
+
+static void adv7511_encoder_mode_set(struct drm_encoder *encoder,
+				     struct drm_display_mode *mode,
+				     struct drm_display_mode *adj_mode)
+{
+	struct adv7511 *adv7511 = encoder_to_adv7511(encoder);
+	unsigned int low_refresh_rate;
+	unsigned int hsync_polarity = 0;
+	unsigned int vsync_polarity = 0;
+
+	if (adv7511->embedded_sync) {
+		unsigned int hsync_offset, hsync_len;
+		unsigned int vsync_offset, vsync_len;
+
+		hsync_offset = adj_mode->crtc_hsync_start -
+			       adj_mode->crtc_hdisplay;
+		vsync_offset = adj_mode->crtc_vsync_start -
+			       adj_mode->crtc_vdisplay;
+		hsync_len = adj_mode->crtc_hsync_end -
+			    adj_mode->crtc_hsync_start;
+		vsync_len = adj_mode->crtc_vsync_end -
+			    adj_mode->crtc_vsync_start;
+
+		/* The hardware vsync generator has a off-by-one bug */
+		vsync_offset += 1;
+
+		regmap_write(adv7511->regmap, ADV7511_REG_HSYNC_PLACEMENT_MSB,
+			     ((hsync_offset >> 10) & 0x7) << 5);
+		regmap_write(adv7511->regmap, ADV7511_REG_SYNC_DECODER(0),
+			     (hsync_offset >> 2) & 0xff);
+		regmap_write(adv7511->regmap, ADV7511_REG_SYNC_DECODER(1),
+			     ((hsync_offset & 0x3) << 6) |
+			     ((hsync_len >> 4) & 0x3f));
+		regmap_write(adv7511->regmap, ADV7511_REG_SYNC_DECODER(2),
+			     ((hsync_len & 0xf) << 4) |
+			     ((vsync_offset >> 6) & 0xf));
+		regmap_write(adv7511->regmap, ADV7511_REG_SYNC_DECODER(3),
+			     ((vsync_offset & 0x3f) << 2) |
+			     ((vsync_len >> 8) & 0x3));
+		regmap_write(adv7511->regmap, ADV7511_REG_SYNC_DECODER(4),
+			     vsync_len & 0xff);
+
+		hsync_polarity = !(adj_mode->flags & DRM_MODE_FLAG_PHSYNC);
+		vsync_polarity = !(adj_mode->flags & DRM_MODE_FLAG_PVSYNC);
+	} else {
+		enum adv7511_sync_polarity mode_hsync_polarity;
+		enum adv7511_sync_polarity mode_vsync_polarity;
+
+		/**
+		 * If the input signal is always low or always high we want to
+		 * invert or let it passthrough depending on the polarity of the
+		 * current mode.
+		 **/
+		if (adj_mode->flags & DRM_MODE_FLAG_NHSYNC)
+			mode_hsync_polarity = ADV7511_SYNC_POLARITY_LOW;
+		else
+			mode_hsync_polarity = ADV7511_SYNC_POLARITY_HIGH;
+
+		if (adj_mode->flags & DRM_MODE_FLAG_NVSYNC)
+			mode_vsync_polarity = ADV7511_SYNC_POLARITY_LOW;
+		else
+			mode_vsync_polarity = ADV7511_SYNC_POLARITY_HIGH;
+
+		if (adv7511->hsync_polarity != mode_hsync_polarity &&
+		    adv7511->hsync_polarity !=
+		    ADV7511_SYNC_POLARITY_PASSTHROUGH)
+			hsync_polarity = 1;
+
+		if (adv7511->vsync_polarity != mode_vsync_polarity &&
+		    adv7511->vsync_polarity !=
+		    ADV7511_SYNC_POLARITY_PASSTHROUGH)
+			vsync_polarity = 1;
+	}
+
+	if (mode->vrefresh <= 24000)
+		low_refresh_rate = ADV7511_LOW_REFRESH_RATE_24HZ;
+	else if (mode->vrefresh <= 25000)
+		low_refresh_rate = ADV7511_LOW_REFRESH_RATE_25HZ;
+	else if (mode->vrefresh <= 30000)
+		low_refresh_rate = ADV7511_LOW_REFRESH_RATE_30HZ;
+	else
+		low_refresh_rate = ADV7511_LOW_REFRESH_RATE_NONE;
+
+	regmap_update_bits(adv7511->regmap, 0xfb,
+		0x6, low_refresh_rate << 1);
+	regmap_update_bits(adv7511->regmap, 0x17,
+		0x60, (vsync_polarity << 6) | (hsync_polarity << 5));
+
+	/*
+	 * TODO Test first order 4:2:2 to 4:4:4 up conversion method, which is
+	 * supposed to give better results.
+	 */
+
+	adv7511->f_tmds = mode->clock;
+}
+
+static struct drm_encoder_slave_funcs adv7511_encoder_funcs = {
+	.dpms = adv7511_encoder_dpms,
+	.mode_valid = adv7511_encoder_mode_valid,
+	.mode_set = adv7511_encoder_mode_set,
+	.detect = adv7511_encoder_detect,
+	.get_modes = adv7511_get_modes,
+};
+
+/* -----------------------------------------------------------------------------
+ * Probe & remove
+ */
+
+static int adv7511_parse_dt(struct device_node *np,
+			    struct adv7511_link_config *config)
+{
+	const char *str;
+	int ret;
+
+	memset(config, 0, sizeof(*config));
+
+	of_property_read_u32(np, "adi,input-depth", &config->input_color_depth);
+	if (config->input_color_depth != 8 && config->input_color_depth != 10 &&
+	    config->input_color_depth != 12)
+		return -EINVAL;
+
+	ret = of_property_read_string(np, "adi,input-colorspace", &str);
+	if (ret < 0)
+		return ret;
+
+	if (!strcmp(str, "rgb"))
+		config->input_colorspace = HDMI_COLORSPACE_RGB;
+	else if (!strcmp(str, "yuv422"))
+		config->input_colorspace = HDMI_COLORSPACE_YUV422;
+	else if (!strcmp(str, "yuv444"))
+		config->input_colorspace = HDMI_COLORSPACE_YUV444;
+	else
+		return -EINVAL;
+
+	ret = of_property_read_string(np, "adi,input-clock", &str);
+	if (ret < 0)
+		return ret;
+
+	if (!strcmp(str, "1x"))
+		config->input_clock = ADV7511_INPUT_CLOCK_1X;
+	else if (!strcmp(str, "2x"))
+		config->input_clock = ADV7511_INPUT_CLOCK_2X;
+	else if (!strcmp(str, "ddr"))
+		config->input_clock = ADV7511_INPUT_CLOCK_DDR;
+	else
+		return -EINVAL;
+
+	if (config->input_colorspace == HDMI_COLORSPACE_YUV422 ||
+	    config->input_clock != ADV7511_INPUT_CLOCK_1X) {
+		ret = of_property_read_u32(np, "adi,input-style",
+					   &config->input_style);
+		if (ret)
+			return ret;
+
+		if (config->input_style < 1 || config->input_style > 3)
+			return -EINVAL;
+
+		ret = of_property_read_string(np, "adi,input-justification",
+					      &str);
+		if (ret < 0)
+			return ret;
+
+		if (!strcmp(str, "left"))
+			config->input_justification =
+				ADV7511_INPUT_JUSTIFICATION_LEFT;
+		else if (!strcmp(str, "evenly"))
+			config->input_justification =
+				ADV7511_INPUT_JUSTIFICATION_EVENLY;
+		else if (!strcmp(str, "right"))
+			config->input_justification =
+				ADV7511_INPUT_JUSTIFICATION_RIGHT;
+		else
+			return -EINVAL;
+
+	} else {
+		config->input_style = 1;
+		config->input_justification = ADV7511_INPUT_JUSTIFICATION_LEFT;
+	}
+
+	of_property_read_u32(np, "adi,clock-delay", &config->clock_delay);
+	if (config->clock_delay < -1200 || config->clock_delay > 1600)
+		return -EINVAL;
+
+	config->embedded_sync = of_property_read_bool(np, "adi,embedded-sync");
+
+	/* Hardcode the sync pulse configurations for now. */
+	config->sync_pulse = ADV7511_INPUT_SYNC_PULSE_NONE;
+	config->vsync_polarity = ADV7511_SYNC_POLARITY_PASSTHROUGH;
+	config->hsync_polarity = ADV7511_SYNC_POLARITY_PASSTHROUGH;
+
+	return 0;
+}
+
+static const int edid_i2c_addr = 0x7e;
+static const int packet_i2c_addr = 0x70;
+static const int cec_i2c_addr = 0x78;
+
+static int adv7511_probe(struct i2c_client *i2c, const struct i2c_device_id *id)
+{
+	struct adv7511_link_config link_config;
+	struct adv7511 *adv7511;
+	struct device *dev = &i2c->dev;
+	unsigned int val;
+	int ret;
+
+	if (!dev->of_node)
+		return -EINVAL;
+
+	adv7511 = devm_kzalloc(dev, sizeof(*adv7511), GFP_KERNEL);
+	if (!adv7511)
+		return -ENOMEM;
+
+	adv7511->dpms_mode = DRM_MODE_DPMS_OFF;
+	adv7511->status = connector_status_disconnected;
+
+	ret = adv7511_parse_dt(dev->of_node, &link_config);
+	if (ret)
+		return ret;
+
+	/*
+	 * The power down GPIO is optional. If present, toggle it from active to
+	 * inactive to wake up the encoder.
+	 */
+	adv7511->gpio_pd = devm_gpiod_get_optional(dev, "pd", GPIOD_OUT_HIGH);
+	if (IS_ERR(adv7511->gpio_pd))
+		return PTR_ERR(adv7511->gpio_pd);
+
+	if (adv7511->gpio_pd) {
+		mdelay(5);
+		gpiod_set_value_cansleep(adv7511->gpio_pd, 0);
+	}
+
+	adv7511->regmap = devm_regmap_init_i2c(i2c, &adv7511_regmap_config);
+	if (IS_ERR(adv7511->regmap))
+		return PTR_ERR(adv7511->regmap);
+
+	ret = regmap_read(adv7511->regmap, ADV7511_REG_CHIP_REVISION, &val);
+	if (ret)
+		return ret;
+	dev_dbg(dev, "Rev. %d\n", val);
+
+	ret = regmap_register_patch(adv7511->regmap, adv7511_fixed_registers,
+				    ARRAY_SIZE(adv7511_fixed_registers));
+	if (ret)
+		return ret;
+
+	regmap_write(adv7511->regmap, ADV7511_REG_EDID_I2C_ADDR, edid_i2c_addr);
+	regmap_write(adv7511->regmap, ADV7511_REG_PACKET_I2C_ADDR,
+		     packet_i2c_addr);
+	regmap_write(adv7511->regmap, ADV7511_REG_CEC_I2C_ADDR, cec_i2c_addr);
+	adv7511_packet_disable(adv7511, 0xffff);
+
+	adv7511->i2c_main = i2c;
+	adv7511->i2c_edid = i2c_new_dummy(i2c->adapter, edid_i2c_addr >> 1);
+	if (!adv7511->i2c_edid)
+		return -ENOMEM;
+
+	if (i2c->irq) {
+		init_waitqueue_head(&adv7511->wq);
+
+		ret = devm_request_threaded_irq(dev, i2c->irq, NULL,
+						adv7511_irq_handler,
+						IRQF_ONESHOT, dev_name(dev),
+						adv7511);
+		if (ret)
+			goto err_i2c_unregister_device;
+	}
+
+	/* CEC is unused for now */
+	regmap_write(adv7511->regmap, ADV7511_REG_CEC_CTRL,
+		     ADV7511_CEC_CTRL_POWER_DOWN);
+
+	regmap_update_bits(adv7511->regmap, ADV7511_REG_POWER,
+			   ADV7511_POWER_POWER_DOWN, ADV7511_POWER_POWER_DOWN);
+
+	adv7511->current_edid_segment = -1;
+
+	i2c_set_clientdata(i2c, adv7511);
+
+	adv7511_set_link_config(adv7511, &link_config);
+
+	return 0;
+
+err_i2c_unregister_device:
+	i2c_unregister_device(adv7511->i2c_edid);
+
+	return ret;
+}
+
+static int adv7511_remove(struct i2c_client *i2c)
+{
+	struct adv7511 *adv7511 = i2c_get_clientdata(i2c);
+
+	i2c_unregister_device(adv7511->i2c_edid);
+
+	kfree(adv7511->edid);
+
+	return 0;
+}
+
+static int adv7511_encoder_init(struct i2c_client *i2c, struct drm_device *dev,
+				struct drm_encoder_slave *encoder)
+{
+
+	struct adv7511 *adv7511 = i2c_get_clientdata(i2c);
+
+	encoder->slave_priv = adv7511;
+	encoder->slave_funcs = &adv7511_encoder_funcs;
+
+	adv7511->encoder = &encoder->base;
+
+	return 0;
+}
+
+static const struct i2c_device_id adv7511_i2c_ids[] = {
+	{ "adv7511", 0 },
+	{ "adv7511w", 0 },
+	{ "adv7513", 0 },
+	{ }
+};
+MODULE_DEVICE_TABLE(i2c, adv7511_i2c_ids);
+
+static const struct of_device_id adv7511_of_ids[] = {
+	{ .compatible = "adi,adv7511", },
+	{ .compatible = "adi,adv7511w", },
+	{ .compatible = "adi,adv7513", },
+	{ }
+};
+MODULE_DEVICE_TABLE(of, adv7511_of_ids);
+
+static struct drm_i2c_encoder_driver adv7511_driver = {
+	.i2c_driver = {
+		.driver = {
+			.name = "adv7511",
+			.of_match_table = adv7511_of_ids,
+		},
+		.id_table = adv7511_i2c_ids,
+		.probe = adv7511_probe,
+		.remove = adv7511_remove,
+	},
+
+	.encoder_init = adv7511_encoder_init,
+};
+
+static int __init adv7511_init(void)
+{
+	return drm_i2c_encoder_register(THIS_MODULE, &adv7511_driver);
+}
+module_init(adv7511_init);
+
+static void __exit adv7511_exit(void)
+{
+	drm_i2c_encoder_unregister(&adv7511_driver);
+}
+module_exit(adv7511_exit);
+
+MODULE_AUTHOR("Lars-Peter Clausen <lars@metafoo.de>");
+MODULE_DESCRIPTION("ADV7511 HDMI transmitter driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/i2c/adv7511.h b/drivers/gpu/drm/i2c/adv7511.h
new file mode 100644
index 000000000000..6599ed538426
--- /dev/null
+++ b/drivers/gpu/drm/i2c/adv7511.h
@@ -0,0 +1,289 @@
+/*
+ * Analog Devices ADV7511 HDMI transmitter driver
+ *
+ * Copyright 2012 Analog Devices Inc.
+ *
+ * Licensed under the GPL-2.
+ */
+
+#ifndef __DRM_I2C_ADV7511_H__
+#define __DRM_I2C_ADV7511_H__
+
+#include <linux/hdmi.h>
+
+#define ADV7511_REG_CHIP_REVISION		0x00
+#define ADV7511_REG_N0				0x01
+#define ADV7511_REG_N1				0x02
+#define ADV7511_REG_N2				0x03
+#define ADV7511_REG_SPDIF_FREQ			0x04
+#define ADV7511_REG_CTS_AUTOMATIC1		0x05
+#define ADV7511_REG_CTS_AUTOMATIC2		0x06
+#define ADV7511_REG_CTS_MANUAL0			0x07
+#define ADV7511_REG_CTS_MANUAL1			0x08
+#define ADV7511_REG_CTS_MANUAL2			0x09
+#define ADV7511_REG_AUDIO_SOURCE		0x0a
+#define ADV7511_REG_AUDIO_CONFIG		0x0b
+#define ADV7511_REG_I2S_CONFIG			0x0c
+#define ADV7511_REG_I2S_WIDTH			0x0d
+#define ADV7511_REG_AUDIO_SUB_SRC0		0x0e
+#define ADV7511_REG_AUDIO_SUB_SRC1		0x0f
+#define ADV7511_REG_AUDIO_SUB_SRC2		0x10
+#define ADV7511_REG_AUDIO_SUB_SRC3		0x11
+#define ADV7511_REG_AUDIO_CFG1			0x12
+#define ADV7511_REG_AUDIO_CFG2			0x13
+#define ADV7511_REG_AUDIO_CFG3			0x14
+#define ADV7511_REG_I2C_FREQ_ID_CFG		0x15
+#define ADV7511_REG_VIDEO_INPUT_CFG1		0x16
+#define ADV7511_REG_CSC_UPPER(x)		(0x18 + (x) * 2)
+#define ADV7511_REG_CSC_LOWER(x)		(0x19 + (x) * 2)
+#define ADV7511_REG_SYNC_DECODER(x)		(0x30 + (x))
+#define ADV7511_REG_DE_GENERATOR		(0x35 + (x))
+#define ADV7511_REG_PIXEL_REPETITION		0x3b
+#define ADV7511_REG_VIC_MANUAL			0x3c
+#define ADV7511_REG_VIC_SEND			0x3d
+#define ADV7511_REG_VIC_DETECTED		0x3e
+#define ADV7511_REG_AUX_VIC_DETECTED		0x3f
+#define ADV7511_REG_PACKET_ENABLE0		0x40
+#define ADV7511_REG_POWER			0x41
+#define ADV7511_REG_STATUS			0x42
+#define ADV7511_REG_EDID_I2C_ADDR		0x43
+#define ADV7511_REG_PACKET_ENABLE1		0x44
+#define ADV7511_REG_PACKET_I2C_ADDR		0x45
+#define ADV7511_REG_DSD_ENABLE			0x46
+#define ADV7511_REG_VIDEO_INPUT_CFG2		0x48
+#define ADV7511_REG_INFOFRAME_UPDATE		0x4a
+#define ADV7511_REG_GC(x)			(0x4b + (x)) /* 0x4b - 0x51 */
+#define ADV7511_REG_AVI_INFOFRAME_VERSION	0x52
+#define ADV7511_REG_AVI_INFOFRAME_LENGTH	0x53
+#define ADV7511_REG_AVI_INFOFRAME_CHECKSUM	0x54
+#define ADV7511_REG_AVI_INFOFRAME(x)		(0x55 + (x)) /* 0x55 - 0x6f */
+#define ADV7511_REG_AUDIO_INFOFRAME_VERSION	0x70
+#define ADV7511_REG_AUDIO_INFOFRAME_LENGTH	0x71
+#define ADV7511_REG_AUDIO_INFOFRAME_CHECKSUM	0x72
+#define ADV7511_REG_AUDIO_INFOFRAME(x)		(0x73 + (x)) /* 0x73 - 0x7c */
+#define ADV7511_REG_INT_ENABLE(x)		(0x94 + (x))
+#define ADV7511_REG_INT(x)			(0x96 + (x))
+#define ADV7511_REG_INPUT_CLK_DIV		0x9d
+#define ADV7511_REG_PLL_STATUS			0x9e
+#define ADV7511_REG_HDMI_POWER			0xa1
+#define ADV7511_REG_HDCP_HDMI_CFG		0xaf
+#define ADV7511_REG_AN(x)			(0xb0 + (x)) /* 0xb0 - 0xb7 */
+#define ADV7511_REG_HDCP_STATUS			0xb8
+#define ADV7511_REG_BCAPS			0xbe
+#define ADV7511_REG_BKSV(x)			(0xc0 + (x)) /* 0xc0 - 0xc3 */
+#define ADV7511_REG_EDID_SEGMENT		0xc4
+#define ADV7511_REG_DDC_STATUS			0xc8
+#define ADV7511_REG_EDID_READ_CTRL		0xc9
+#define ADV7511_REG_BSTATUS(x)			(0xca + (x)) /* 0xca - 0xcb */
+#define ADV7511_REG_TIMING_GEN_SEQ		0xd0
+#define ADV7511_REG_POWER2			0xd6
+#define ADV7511_REG_HSYNC_PLACEMENT_MSB		0xfa
+
+#define ADV7511_REG_SYNC_ADJUSTMENT(x)		(0xd7 + (x)) /* 0xd7 - 0xdc */
+#define ADV7511_REG_TMDS_CLOCK_INV		0xde
+#define ADV7511_REG_ARC_CTRL			0xdf
+#define ADV7511_REG_CEC_I2C_ADDR		0xe1
+#define ADV7511_REG_CEC_CTRL			0xe2
+#define ADV7511_REG_CHIP_ID_HIGH		0xf5
+#define ADV7511_REG_CHIP_ID_LOW			0xf6
+
+#define ADV7511_CSC_ENABLE			BIT(7)
+#define ADV7511_CSC_UPDATE_MODE			BIT(5)
+
+#define ADV7511_INT0_HDP			BIT(7)
+#define ADV7511_INT0_VSYNC			BIT(5)
+#define ADV7511_INT0_AUDIO_FIFO_FULL		BIT(4)
+#define ADV7511_INT0_EDID_READY			BIT(2)
+#define ADV7511_INT0_HDCP_AUTHENTICATED		BIT(1)
+
+#define ADV7511_INT1_DDC_ERROR			BIT(7)
+#define ADV7511_INT1_BKSV			BIT(6)
+#define ADV7511_INT1_CEC_TX_READY		BIT(5)
+#define ADV7511_INT1_CEC_TX_ARBIT_LOST		BIT(4)
+#define ADV7511_INT1_CEC_TX_RETRY_TIMEOUT	BIT(3)
+#define ADV7511_INT1_CEC_RX_READY3		BIT(2)
+#define ADV7511_INT1_CEC_RX_READY2		BIT(1)
+#define ADV7511_INT1_CEC_RX_READY1		BIT(0)
+
+#define ADV7511_ARC_CTRL_POWER_DOWN		BIT(0)
+
+#define ADV7511_CEC_CTRL_POWER_DOWN		BIT(0)
+
+#define ADV7511_POWER_POWER_DOWN		BIT(6)
+
+#define ADV7511_HDMI_CFG_MODE_MASK		0x2
+#define ADV7511_HDMI_CFG_MODE_DVI		0x0
+#define ADV7511_HDMI_CFG_MODE_HDMI		0x2
+
+#define ADV7511_AUDIO_SELECT_I2C		0x0
+#define ADV7511_AUDIO_SELECT_SPDIF		0x1
+#define ADV7511_AUDIO_SELECT_DSD		0x2
+#define ADV7511_AUDIO_SELECT_HBR		0x3
+#define ADV7511_AUDIO_SELECT_DST		0x4
+
+#define ADV7511_I2S_SAMPLE_LEN_16		0x2
+#define ADV7511_I2S_SAMPLE_LEN_20		0x3
+#define ADV7511_I2S_SAMPLE_LEN_18		0x4
+#define ADV7511_I2S_SAMPLE_LEN_22		0x5
+#define ADV7511_I2S_SAMPLE_LEN_19		0x8
+#define ADV7511_I2S_SAMPLE_LEN_23		0x9
+#define ADV7511_I2S_SAMPLE_LEN_24		0xb
+#define ADV7511_I2S_SAMPLE_LEN_17		0xc
+#define ADV7511_I2S_SAMPLE_LEN_21		0xd
+
+#define ADV7511_SAMPLE_FREQ_44100		0x0
+#define ADV7511_SAMPLE_FREQ_48000		0x2
+#define ADV7511_SAMPLE_FREQ_32000		0x3
+#define ADV7511_SAMPLE_FREQ_88200		0x8
+#define ADV7511_SAMPLE_FREQ_96000		0xa
+#define ADV7511_SAMPLE_FREQ_176400		0xc
+#define ADV7511_SAMPLE_FREQ_192000		0xe
+
+#define ADV7511_STATUS_POWER_DOWN_POLARITY	BIT(7)
+#define ADV7511_STATUS_HPD			BIT(6)
+#define ADV7511_STATUS_MONITOR_SENSE		BIT(5)
+#define ADV7511_STATUS_I2S_32BIT_MODE		BIT(3)
+
+#define ADV7511_PACKET_ENABLE_N_CTS		BIT(8+6)
+#define ADV7511_PACKET_ENABLE_AUDIO_SAMPLE	BIT(8+5)
+#define ADV7511_PACKET_ENABLE_AVI_INFOFRAME	BIT(8+4)
+#define ADV7511_PACKET_ENABLE_AUDIO_INFOFRAME	BIT(8+3)
+#define ADV7511_PACKET_ENABLE_GC		BIT(7)
+#define ADV7511_PACKET_ENABLE_SPD		BIT(6)
+#define ADV7511_PACKET_ENABLE_MPEG		BIT(5)
+#define ADV7511_PACKET_ENABLE_ACP		BIT(4)
+#define ADV7511_PACKET_ENABLE_ISRC		BIT(3)
+#define ADV7511_PACKET_ENABLE_GM		BIT(2)
+#define ADV7511_PACKET_ENABLE_SPARE2		BIT(1)
+#define ADV7511_PACKET_ENABLE_SPARE1		BIT(0)
+
+#define ADV7511_REG_POWER2_HDP_SRC_MASK		0xc0
+#define ADV7511_REG_POWER2_HDP_SRC_BOTH		0x00
+#define ADV7511_REG_POWER2_HDP_SRC_HDP		0x40
+#define ADV7511_REG_POWER2_HDP_SRC_CEC		0x80
+#define ADV7511_REG_POWER2_HDP_SRC_NONE		0xc0
+#define ADV7511_REG_POWER2_TDMS_ENABLE		BIT(4)
+#define ADV7511_REG_POWER2_GATE_INPUT_CLK	BIT(0)
+
+#define ADV7511_LOW_REFRESH_RATE_NONE		0x0
+#define ADV7511_LOW_REFRESH_RATE_24HZ		0x1
+#define ADV7511_LOW_REFRESH_RATE_25HZ		0x2
+#define ADV7511_LOW_REFRESH_RATE_30HZ		0x3
+
+#define ADV7511_AUDIO_CFG3_LEN_MASK		0x0f
+#define ADV7511_I2C_FREQ_ID_CFG_RATE_MASK	0xf0
+
+#define ADV7511_AUDIO_SOURCE_I2S		0
+#define ADV7511_AUDIO_SOURCE_SPDIF		1
+
+#define ADV7511_I2S_FORMAT_I2S			0
+#define ADV7511_I2S_FORMAT_RIGHT_J		1
+#define ADV7511_I2S_FORMAT_LEFT_J		2
+
+#define ADV7511_PACKET(p, x)	    ((p) * 0x20 + (x))
+#define ADV7511_PACKET_SDP(x)	    ADV7511_PACKET(0, x)
+#define ADV7511_PACKET_MPEG(x)	    ADV7511_PACKET(1, x)
+#define ADV7511_PACKET_ACP(x)	    ADV7511_PACKET(2, x)
+#define ADV7511_PACKET_ISRC1(x)	    ADV7511_PACKET(3, x)
+#define ADV7511_PACKET_ISRC2(x)	    ADV7511_PACKET(4, x)
+#define ADV7511_PACKET_GM(x)	    ADV7511_PACKET(5, x)
+#define ADV7511_PACKET_SPARE(x)	    ADV7511_PACKET(6, x)
+
+enum adv7511_input_clock {
+	ADV7511_INPUT_CLOCK_1X,
+	ADV7511_INPUT_CLOCK_2X,
+	ADV7511_INPUT_CLOCK_DDR,
+};
+
+enum adv7511_input_justification {
+	ADV7511_INPUT_JUSTIFICATION_EVENLY = 0,
+	ADV7511_INPUT_JUSTIFICATION_RIGHT = 1,
+	ADV7511_INPUT_JUSTIFICATION_LEFT = 2,
+};
+
+enum adv7511_input_sync_pulse {
+	ADV7511_INPUT_SYNC_PULSE_DE = 0,
+	ADV7511_INPUT_SYNC_PULSE_HSYNC = 1,
+	ADV7511_INPUT_SYNC_PULSE_VSYNC = 2,
+	ADV7511_INPUT_SYNC_PULSE_NONE = 3,
+};
+
+/**
+ * enum adv7511_sync_polarity - Polarity for the input sync signals
+ * @ADV7511_SYNC_POLARITY_PASSTHROUGH:  Sync polarity matches that of
+ *				       the currently configured mode.
+ * @ADV7511_SYNC_POLARITY_LOW:	    Sync polarity is low
+ * @ADV7511_SYNC_POLARITY_HIGH:	    Sync polarity is high
+ *
+ * If the polarity is set to either LOW or HIGH the driver will configure the
+ * ADV7511 to internally invert the sync signal if required to match the sync
+ * polarity setting for the currently selected output mode.
+ *
+ * If the polarity is set to PASSTHROUGH, the ADV7511 will route the signal
+ * unchanged. This is used when the upstream graphics core already generates
+ * the sync signals with the correct polarity.
+ */
+enum adv7511_sync_polarity {
+	ADV7511_SYNC_POLARITY_PASSTHROUGH,
+	ADV7511_SYNC_POLARITY_LOW,
+	ADV7511_SYNC_POLARITY_HIGH,
+};
+
+/**
+ * struct adv7511_link_config - Describes adv7511 hardware configuration
+ * @input_color_depth:		Number of bits per color component (8, 10 or 12)
+ * @input_colorspace:		The input colorspace (RGB, YUV444, YUV422)
+ * @input_clock:		The input video clock style (1x, 2x, DDR)
+ * @input_style:		The input component arrangement variant
+ * @input_justification:	Video input format bit justification
+ * @clock_delay:		Clock delay for the input clock (in ps)
+ * @embedded_sync:		Video input uses BT.656-style embedded sync
+ * @sync_pulse:			Select the sync pulse
+ * @vsync_polarity:		vsync input signal configuration
+ * @hsync_polarity:		hsync input signal configuration
+ */
+struct adv7511_link_config {
+	unsigned int input_color_depth;
+	enum hdmi_colorspace input_colorspace;
+	enum adv7511_input_clock input_clock;
+	unsigned int input_style;
+	enum adv7511_input_justification input_justification;
+
+	int clock_delay;
+
+	bool embedded_sync;
+	enum adv7511_input_sync_pulse sync_pulse;
+	enum adv7511_sync_polarity vsync_polarity;
+	enum adv7511_sync_polarity hsync_polarity;
+};
+
+/**
+ * enum adv7511_csc_scaling - Scaling factor for the ADV7511 CSC
+ * @ADV7511_CSC_SCALING_1: CSC results are not scaled
+ * @ADV7511_CSC_SCALING_2: CSC results are scaled by a factor of two
+ * @ADV7511_CSC_SCALING_4: CSC results are scalled by a factor of four
+ */
+enum adv7511_csc_scaling {
+	ADV7511_CSC_SCALING_1 = 0,
+	ADV7511_CSC_SCALING_2 = 1,
+	ADV7511_CSC_SCALING_4 = 2,
+};
+
+/**
+ * struct adv7511_video_config - Describes adv7511 hardware configuration
+ * @csc_enable:			Whether to enable color space conversion
+ * @csc_scaling_factor:		Color space conversion scaling factor
+ * @csc_coefficents:		Color space conversion coefficents
+ * @hdmi_mode:			Whether to use HDMI or DVI output mode
+ * @avi_infoframe:		HDMI infoframe
+ */
+struct adv7511_video_config {
+	bool csc_enable;
+	enum adv7511_csc_scaling csc_scaling_factor;
+	const uint16_t *csc_coefficents;
+
+	bool hdmi_mode;
+	struct hdmi_avi_infoframe avi_infoframe;
+};
+
+#endif /* __DRM_I2C_ADV7511_H__ */
diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c
index d4762799351d..a9041d1a8ff0 100644
--- a/drivers/gpu/drm/i2c/tda998x_drv.c
+++ b/drivers/gpu/drm/i2c/tda998x_drv.c
@@ -32,6 +32,8 @@
 struct tda998x_priv {
 	struct i2c_client *cec;
 	struct i2c_client *hdmi;
+	struct mutex mutex;
+	struct delayed_work dwork;
 	uint16_t rev;
 	uint8_t current_page;
 	int dpms;
@@ -402,9 +404,10 @@ reg_read_range(struct tda998x_priv *priv, uint16_t reg, char *buf, int cnt)
 	uint8_t addr = REG2ADDR(reg);
 	int ret;
 
+	mutex_lock(&priv->mutex);
 	ret = set_page(priv, reg);
 	if (ret < 0)
-		return ret;
+		goto out;
 
 	ret = i2c_master_send(client, &addr, sizeof(addr));
 	if (ret < 0)
@@ -414,10 +417,12 @@ reg_read_range(struct tda998x_priv *priv, uint16_t reg, char *buf, int cnt)
 	if (ret < 0)
 		goto fail;
 
-	return ret;
+	goto out;
 
 fail:
 	dev_err(&client->dev, "Error %d reading from 0x%x\n", ret, reg);
+out:
+	mutex_unlock(&priv->mutex);
 	return ret;
 }
 
@@ -431,13 +436,16 @@ reg_write_range(struct tda998x_priv *priv, uint16_t reg, uint8_t *p, int cnt)
 	buf[0] = REG2ADDR(reg);
 	memcpy(&buf[1], p, cnt);
 
+	mutex_lock(&priv->mutex);
 	ret = set_page(priv, reg);
 	if (ret < 0)
-		return;
+		goto out;
 
 	ret = i2c_master_send(client, buf, cnt + 1);
 	if (ret < 0)
 		dev_err(&client->dev, "Error %d writing to 0x%x\n", ret, reg);
+out:
+	mutex_unlock(&priv->mutex);
 }
 
 static int
@@ -459,13 +467,16 @@ reg_write(struct tda998x_priv *priv, uint16_t reg, uint8_t val)
 	uint8_t buf[] = {REG2ADDR(reg), val};
 	int ret;
 
+	mutex_lock(&priv->mutex);
 	ret = set_page(priv, reg);
 	if (ret < 0)
-		return;
+		goto out;
 
 	ret = i2c_master_send(client, buf, sizeof(buf));
 	if (ret < 0)
 		dev_err(&client->dev, "Error %d writing to 0x%x\n", ret, reg);
+out:
+	mutex_unlock(&priv->mutex);
 }
 
 static void
@@ -475,13 +486,16 @@ reg_write16(struct tda998x_priv *priv, uint16_t reg, uint16_t val)
 	uint8_t buf[] = {REG2ADDR(reg), val >> 8, val};
 	int ret;
 
+	mutex_lock(&priv->mutex);
 	ret = set_page(priv, reg);
 	if (ret < 0)
-		return;
+		goto out;
 
 	ret = i2c_master_send(client, buf, sizeof(buf));
 	if (ret < 0)
 		dev_err(&client->dev, "Error %d writing to 0x%x\n", ret, reg);
+out:
+	mutex_unlock(&priv->mutex);
 }
 
 static void
@@ -536,6 +550,17 @@ tda998x_reset(struct tda998x_priv *priv)
 	reg_write(priv, REG_MUX_VP_VIP_OUT, 0x24);
 }
 
+/* handle HDMI connect/disconnect */
+static void tda998x_hpd(struct work_struct *work)
+{
+	struct delayed_work *dwork = to_delayed_work(work);
+	struct tda998x_priv *priv =
+			container_of(dwork, struct tda998x_priv, dwork);
+
+	if (priv->encoder && priv->encoder->dev)
+		drm_kms_helper_hotplug_event(priv->encoder->dev);
+}
+
 /*
  * only 2 interrupts may occur: screen plug/unplug and EDID read
  */
@@ -559,8 +584,7 @@ static irqreturn_t tda998x_irq_thread(int irq, void *data)
 		priv->wq_edid_wait = 0;
 		wake_up(&priv->wq_edid);
 	} else if (cec != 0) {			/* HPD change */
-		if (priv->encoder && priv->encoder->dev)
-			drm_helper_hpd_irq_event(priv->encoder->dev);
+		schedule_delayed_work(&priv->dwork, HZ/10);
 	}
 	return IRQ_HANDLED;
 }
@@ -1170,8 +1194,10 @@ static void tda998x_destroy(struct tda998x_priv *priv)
 	/* disable all IRQs and free the IRQ handler */
 	cec_write(priv, REG_CEC_RXSHPDINTENA, 0);
 	reg_clear(priv, REG_INT_FLAGS_2, INT_FLAGS_2_EDID_BLK_RD);
-	if (priv->hdmi->irq)
+	if (priv->hdmi->irq) {
 		free_irq(priv->hdmi->irq, priv);
+		cancel_delayed_work_sync(&priv->dwork);
+	}
 
 	i2c_unregister_device(priv->cec);
 }
@@ -1255,6 +1281,7 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 	struct device_node *np = client->dev.of_node;
 	u32 video;
 	int rev_lo, rev_hi, ret;
+	unsigned short cec_addr;
 
 	priv->vip_cntrl_0 = VIP_CNTRL_0_SWAP_A(2) | VIP_CNTRL_0_SWAP_B(3);
 	priv->vip_cntrl_1 = VIP_CNTRL_1_SWAP_C(0) | VIP_CNTRL_1_SWAP_D(1);
@@ -1262,12 +1289,16 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 
 	priv->current_page = 0xff;
 	priv->hdmi = client;
-	priv->cec = i2c_new_dummy(client->adapter, 0x34);
+	/* CEC I2C address bound to TDA998x I2C addr by configuration pins */
+	cec_addr = 0x34 + (client->addr & 0x03);
+	priv->cec = i2c_new_dummy(client->adapter, cec_addr);
 	if (!priv->cec)
 		return -ENODEV;
 
 	priv->dpms = DRM_MODE_DPMS_OFF;
 
+	mutex_init(&priv->mutex);	/* protect the page access */
+
 	/* wake up the device: */
 	cec_write(priv, REG_CEC_ENAMODS,
 			CEC_ENAMODS_EN_RXSENS | CEC_ENAMODS_EN_HDMI);
@@ -1323,8 +1354,9 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 	if (client->irq) {
 		int irqf_trigger;
 
-		/* init read EDID waitqueue */
+		/* init read EDID waitqueue and HDP work */
 		init_waitqueue_head(&priv->wq_edid);
+		INIT_DELAYED_WORK(&priv->dwork, tda998x_hpd);
 
 		/* clear pending interrupts */
 		reg_read(priv, REG_INT_FLAGS_0);
diff --git a/drivers/gpu/drm/i810/i810_dma.c b/drivers/gpu/drm/i810/i810_dma.c
index bae897de9468..d91856779beb 100644
--- a/drivers/gpu/drm/i810/i810_dma.c
+++ b/drivers/gpu/drm/i810/i810_dma.c
@@ -213,7 +213,7 @@ static int i810_dma_cleanup(struct drm_device *dev)
 		    (drm_i810_private_t *) dev->dev_private;
 
 		if (dev_priv->ring.virtual_start)
-			drm_core_ioremapfree(&dev_priv->ring.map, dev);
+			drm_legacy_ioremapfree(&dev_priv->ring.map, dev);
 		if (dev_priv->hw_status_page) {
 			pci_free_consistent(dev->pdev, PAGE_SIZE,
 					    dev_priv->hw_status_page,
@@ -227,7 +227,7 @@ static int i810_dma_cleanup(struct drm_device *dev)
 			drm_i810_buf_priv_t *buf_priv = buf->dev_private;
 
 			if (buf_priv->kernel_virtual && buf->total)
-				drm_core_ioremapfree(&buf_priv->map, dev);
+				drm_legacy_ioremapfree(&buf_priv->map, dev);
 		}
 	}
 	return 0;
@@ -306,7 +306,7 @@ static int i810_freelist_init(struct drm_device *dev, drm_i810_private_t *dev_pr
 		buf_priv->map.flags = 0;
 		buf_priv->map.mtrr = 0;
 
-		drm_core_ioremap(&buf_priv->map, dev);
+		drm_legacy_ioremap(&buf_priv->map, dev);
 		buf_priv->kernel_virtual = buf_priv->map.handle;
 
 	}
@@ -334,7 +334,7 @@ static int i810_dma_initialize(struct drm_device *dev,
 		DRM_ERROR("can not find sarea!\n");
 		return -EINVAL;
 	}
-	dev_priv->mmio_map = drm_core_findmap(dev, init->mmio_offset);
+	dev_priv->mmio_map = drm_legacy_findmap(dev, init->mmio_offset);
 	if (!dev_priv->mmio_map) {
 		dev->dev_private = (void *)dev_priv;
 		i810_dma_cleanup(dev);
@@ -342,7 +342,7 @@ static int i810_dma_initialize(struct drm_device *dev,
 		return -EINVAL;
 	}
 	dev->agp_buffer_token = init->buffers_offset;
-	dev->agp_buffer_map = drm_core_findmap(dev, init->buffers_offset);
+	dev->agp_buffer_map = drm_legacy_findmap(dev, init->buffers_offset);
 	if (!dev->agp_buffer_map) {
 		dev->dev_private = (void *)dev_priv;
 		i810_dma_cleanup(dev);
@@ -363,7 +363,7 @@ static int i810_dma_initialize(struct drm_device *dev,
 	dev_priv->ring.map.flags = 0;
 	dev_priv->ring.map.mtrr = 0;
 
-	drm_core_ioremap(&dev_priv->ring.map, dev);
+	drm_legacy_ioremap(&dev_priv->ring.map, dev);
 
 	if (dev_priv->ring.map.handle == NULL) {
 		dev->dev_private = (void *)dev_priv;
@@ -1215,9 +1215,9 @@ void i810_driver_preclose(struct drm_device *dev, struct drm_file *file_priv)
 	}
 
 	if (file_priv->master && file_priv->master->lock.hw_lock) {
-		drm_idlelock_take(&file_priv->master->lock);
+		drm_legacy_idlelock_take(&file_priv->master->lock);
 		i810_driver_reclaim_buffers(dev, file_priv);
-		drm_idlelock_release(&file_priv->master->lock);
+		drm_legacy_idlelock_release(&file_priv->master->lock);
 	} else {
 		/* master disappeared, clean up stuff anyway and hope nothing
 		 * goes wrong */
diff --git a/drivers/gpu/drm/i810/i810_drv.c b/drivers/gpu/drm/i810/i810_drv.c
index 441ccf8f5bdc..44f4a131c8dd 100644
--- a/drivers/gpu/drm/i810/i810_drv.c
+++ b/drivers/gpu/drm/i810/i810_drv.c
@@ -47,7 +47,7 @@ static const struct file_operations i810_driver_fops = {
 	.open = drm_open,
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_mmap,
+	.mmap = drm_legacy_mmap,
 	.poll = drm_poll,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl = drm_compat_ioctl,
@@ -63,6 +63,7 @@ static struct drm_driver driver = {
 	.load = i810_driver_load,
 	.lastclose = i810_driver_lastclose,
 	.preclose = i810_driver_preclose,
+	.set_busid = drm_pci_set_busid,
 	.device_is_agp = i810_driver_device_is_agp,
 	.dma_quiescent = i810_driver_dma_quiescent,
 	.ioctls = i810_ioctls,
diff --git a/drivers/gpu/drm/i810/i810_drv.h b/drivers/gpu/drm/i810/i810_drv.h
index d4d16eddd651..93ec5dc4e7d3 100644
--- a/drivers/gpu/drm/i810/i810_drv.h
+++ b/drivers/gpu/drm/i810/i810_drv.h
@@ -32,6 +32,8 @@
 #ifndef _I810_DRV_H_
 #define _I810_DRV_H_
 
+#include <drm/drm_legacy.h>
+
 /* General customization:
  */
 
diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile
index 91bd167e1cb7..e4083e41a600 100644
--- a/drivers/gpu/drm/i915/Makefile
+++ b/drivers/gpu/drm/i915/Makefile
@@ -11,7 +11,9 @@ i915-y := i915_drv.o \
 	  i915_params.o \
           i915_suspend.o \
 	  i915_sysfs.o \
-	  intel_pm.o
+	  intel_pm.o \
+	  intel_runtime_pm.o
+
 i915-$(CONFIG_COMPAT)   += i915_ioc32.o
 i915-$(CONFIG_DEBUG_FS) += i915_debugfs.o
 
@@ -31,19 +33,25 @@ i915-y += i915_cmd_parser.o \
 	  i915_gpu_error.o \
 	  i915_irq.o \
 	  i915_trace_points.o \
+	  intel_lrc.o \
 	  intel_ringbuffer.o \
 	  intel_uncore.o
 
 # autogenerated null render state
 i915-y += intel_renderstate_gen6.o \
 	  intel_renderstate_gen7.o \
-	  intel_renderstate_gen8.o
+	  intel_renderstate_gen8.o \
+	  intel_renderstate_gen9.o
 
 # modesetting core code
-i915-y += intel_bios.o \
+i915-y += intel_audio.o \
+	  intel_bios.o \
 	  intel_display.o \
+	  intel_fifo_underrun.o \
+	  intel_frontbuffer.o \
 	  intel_modes.o \
 	  intel_overlay.o \
+	  intel_psr.o \
 	  intel_sideband.o \
 	  intel_sprite.o
 i915-$(CONFIG_ACPI)		+= intel_acpi.o intel_opregion.o
diff --git a/drivers/gpu/drm/i915/dvo_ns2501.c b/drivers/gpu/drm/i915/dvo_ns2501.c
index 74f2af7c2d3e..441630434d34 100644
--- a/drivers/gpu/drm/i915/dvo_ns2501.c
+++ b/drivers/gpu/drm/i915/dvo_ns2501.c
@@ -60,16 +60,297 @@
 
 #define NS2501_REGC 0x0c
 
+enum {
+	MODE_640x480,
+	MODE_800x600,
+	MODE_1024x768,
+};
+
+struct ns2501_reg {
+	 uint8_t offset;
+	 uint8_t value;
+};
+
+/*
+ * Magic values based on what the BIOS on
+ * Fujitsu-Siemens Lifebook S6010 programs (1024x768 panel).
+ */
+static const struct ns2501_reg regs_1024x768[][86] = {
+	[MODE_640x480] = {
+		[0] = { .offset = 0x0a, .value = 0x81, },
+		[1] = { .offset = 0x18, .value = 0x07, },
+		[2] = { .offset = 0x19, .value = 0x00, },
+		[3] = { .offset = 0x1a, .value = 0x00, },
+		[4] = { .offset = 0x1b, .value = 0x11, },
+		[5] = { .offset = 0x1c, .value = 0x54, },
+		[6] = { .offset = 0x1d, .value = 0x03, },
+		[7] = { .offset = 0x1e, .value = 0x02, },
+		[8] = { .offset = 0xf3, .value = 0x90, },
+		[9] = { .offset = 0xf9, .value = 0x00, },
+		[10] = { .offset = 0xc1, .value = 0x90, },
+		[11] = { .offset = 0xc2, .value = 0x00, },
+		[12] = { .offset = 0xc3, .value = 0x0f, },
+		[13] = { .offset = 0xc4, .value = 0x03, },
+		[14] = { .offset = 0xc5, .value = 0x16, },
+		[15] = { .offset = 0xc6, .value = 0x00, },
+		[16] = { .offset = 0xc7, .value = 0x02, },
+		[17] = { .offset = 0xc8, .value = 0x02, },
+		[18] = { .offset = 0xf4, .value = 0x00, },
+		[19] = { .offset = 0x80, .value = 0xff, },
+		[20] = { .offset = 0x81, .value = 0x07, },
+		[21] = { .offset = 0x82, .value = 0x3d, },
+		[22] = { .offset = 0x83, .value = 0x05, },
+		[23] = { .offset = 0x94, .value = 0x00, },
+		[24] = { .offset = 0x95, .value = 0x00, },
+		[25] = { .offset = 0x96, .value = 0x05, },
+		[26] = { .offset = 0x97, .value = 0x00, },
+		[27] = { .offset = 0x9a, .value = 0x88, },
+		[28] = { .offset = 0x9b, .value = 0x00, },
+		[29] = { .offset = 0x98, .value = 0x00, },
+		[30] = { .offset = 0x99, .value = 0x00, },
+		[31] = { .offset = 0xf7, .value = 0x88, },
+		[32] = { .offset = 0xf8, .value = 0x0a, },
+		[33] = { .offset = 0x9c, .value = 0x24, },
+		[34] = { .offset = 0x9d, .value = 0x00, },
+		[35] = { .offset = 0x9e, .value = 0x25, },
+		[36] = { .offset = 0x9f, .value = 0x03, },
+		[37] = { .offset = 0xa0, .value = 0x28, },
+		[38] = { .offset = 0xa1, .value = 0x01, },
+		[39] = { .offset = 0xa2, .value = 0x28, },
+		[40] = { .offset = 0xa3, .value = 0x05, },
+		[41] = { .offset = 0xb6, .value = 0x09, },
+		[42] = { .offset = 0xb8, .value = 0x00, },
+		[43] = { .offset = 0xb9, .value = 0xa0, },
+		[44] = { .offset = 0xba, .value = 0x00, },
+		[45] = { .offset = 0xbb, .value = 0x20, },
+		[46] = { .offset = 0x10, .value = 0x00, },
+		[47] = { .offset = 0x11, .value = 0xa0, },
+		[48] = { .offset = 0x12, .value = 0x02, },
+		[49] = { .offset = 0x20, .value = 0x00, },
+		[50] = { .offset = 0x22, .value = 0x00, },
+		[51] = { .offset = 0x23, .value = 0x00, },
+		[52] = { .offset = 0x24, .value = 0x00, },
+		[53] = { .offset = 0x25, .value = 0x00, },
+		[54] = { .offset = 0x8c, .value = 0x10, },
+		[55] = { .offset = 0x8d, .value = 0x02, },
+		[56] = { .offset = 0x8e, .value = 0x10, },
+		[57] = { .offset = 0x8f, .value = 0x00, },
+		[58] = { .offset = 0x90, .value = 0xff, },
+		[59] = { .offset = 0x91, .value = 0x07, },
+		[60] = { .offset = 0x92, .value = 0xa0, },
+		[61] = { .offset = 0x93, .value = 0x02, },
+		[62] = { .offset = 0xa5, .value = 0x00, },
+		[63] = { .offset = 0xa6, .value = 0x00, },
+		[64] = { .offset = 0xa7, .value = 0x00, },
+		[65] = { .offset = 0xa8, .value = 0x00, },
+		[66] = { .offset = 0xa9, .value = 0x04, },
+		[67] = { .offset = 0xaa, .value = 0x70, },
+		[68] = { .offset = 0xab, .value = 0x4f, },
+		[69] = { .offset = 0xac, .value = 0x00, },
+		[70] = { .offset = 0xa4, .value = 0x84, },
+		[71] = { .offset = 0x7e, .value = 0x18, },
+		[72] = { .offset = 0x84, .value = 0x00, },
+		[73] = { .offset = 0x85, .value = 0x00, },
+		[74] = { .offset = 0x86, .value = 0x00, },
+		[75] = { .offset = 0x87, .value = 0x00, },
+		[76] = { .offset = 0x88, .value = 0x00, },
+		[77] = { .offset = 0x89, .value = 0x00, },
+		[78] = { .offset = 0x8a, .value = 0x00, },
+		[79] = { .offset = 0x8b, .value = 0x00, },
+		[80] = { .offset = 0x26, .value = 0x00, },
+		[81] = { .offset = 0x27, .value = 0x00, },
+		[82] = { .offset = 0xad, .value = 0x00, },
+		[83] = { .offset = 0x08, .value = 0x30, }, /* 0x31 */
+		[84] = { .offset = 0x41, .value = 0x00, },
+		[85] = { .offset = 0xc0, .value = 0x05, },
+	},
+	[MODE_800x600] = {
+		[0] = { .offset = 0x0a, .value = 0x81, },
+		[1] = { .offset = 0x18, .value = 0x07, },
+		[2] = { .offset = 0x19, .value = 0x00, },
+		[3] = { .offset = 0x1a, .value = 0x00, },
+		[4] = { .offset = 0x1b, .value = 0x19, },
+		[5] = { .offset = 0x1c, .value = 0x64, },
+		[6] = { .offset = 0x1d, .value = 0x02, },
+		[7] = { .offset = 0x1e, .value = 0x02, },
+		[8] = { .offset = 0xf3, .value = 0x90, },
+		[9] = { .offset = 0xf9, .value = 0x00, },
+		[10] = { .offset = 0xc1, .value = 0xd7, },
+		[11] = { .offset = 0xc2, .value = 0x00, },
+		[12] = { .offset = 0xc3, .value = 0xf8, },
+		[13] = { .offset = 0xc4, .value = 0x03, },
+		[14] = { .offset = 0xc5, .value = 0x1a, },
+		[15] = { .offset = 0xc6, .value = 0x00, },
+		[16] = { .offset = 0xc7, .value = 0x73, },
+		[17] = { .offset = 0xc8, .value = 0x02, },
+		[18] = { .offset = 0xf4, .value = 0x00, },
+		[19] = { .offset = 0x80, .value = 0x27, },
+		[20] = { .offset = 0x81, .value = 0x03, },
+		[21] = { .offset = 0x82, .value = 0x41, },
+		[22] = { .offset = 0x83, .value = 0x05, },
+		[23] = { .offset = 0x94, .value = 0x00, },
+		[24] = { .offset = 0x95, .value = 0x00, },
+		[25] = { .offset = 0x96, .value = 0x05, },
+		[26] = { .offset = 0x97, .value = 0x00, },
+		[27] = { .offset = 0x9a, .value = 0x88, },
+		[28] = { .offset = 0x9b, .value = 0x00, },
+		[29] = { .offset = 0x98, .value = 0x00, },
+		[30] = { .offset = 0x99, .value = 0x00, },
+		[31] = { .offset = 0xf7, .value = 0x88, },
+		[32] = { .offset = 0xf8, .value = 0x06, },
+		[33] = { .offset = 0x9c, .value = 0x23, },
+		[34] = { .offset = 0x9d, .value = 0x00, },
+		[35] = { .offset = 0x9e, .value = 0x25, },
+		[36] = { .offset = 0x9f, .value = 0x03, },
+		[37] = { .offset = 0xa0, .value = 0x28, },
+		[38] = { .offset = 0xa1, .value = 0x01, },
+		[39] = { .offset = 0xa2, .value = 0x28, },
+		[40] = { .offset = 0xa3, .value = 0x05, },
+		[41] = { .offset = 0xb6, .value = 0x09, },
+		[42] = { .offset = 0xb8, .value = 0x30, },
+		[43] = { .offset = 0xb9, .value = 0xc8, },
+		[44] = { .offset = 0xba, .value = 0x00, },
+		[45] = { .offset = 0xbb, .value = 0x20, },
+		[46] = { .offset = 0x10, .value = 0x20, },
+		[47] = { .offset = 0x11, .value = 0xc8, },
+		[48] = { .offset = 0x12, .value = 0x02, },
+		[49] = { .offset = 0x20, .value = 0x00, },
+		[50] = { .offset = 0x22, .value = 0x00, },
+		[51] = { .offset = 0x23, .value = 0x00, },
+		[52] = { .offset = 0x24, .value = 0x00, },
+		[53] = { .offset = 0x25, .value = 0x00, },
+		[54] = { .offset = 0x8c, .value = 0x10, },
+		[55] = { .offset = 0x8d, .value = 0x02, },
+		[56] = { .offset = 0x8e, .value = 0x04, },
+		[57] = { .offset = 0x8f, .value = 0x00, },
+		[58] = { .offset = 0x90, .value = 0xff, },
+		[59] = { .offset = 0x91, .value = 0x07, },
+		[60] = { .offset = 0x92, .value = 0xa0, },
+		[61] = { .offset = 0x93, .value = 0x02, },
+		[62] = { .offset = 0xa5, .value = 0x00, },
+		[63] = { .offset = 0xa6, .value = 0x00, },
+		[64] = { .offset = 0xa7, .value = 0x00, },
+		[65] = { .offset = 0xa8, .value = 0x00, },
+		[66] = { .offset = 0xa9, .value = 0x83, },
+		[67] = { .offset = 0xaa, .value = 0x40, },
+		[68] = { .offset = 0xab, .value = 0x32, },
+		[69] = { .offset = 0xac, .value = 0x00, },
+		[70] = { .offset = 0xa4, .value = 0x80, },
+		[71] = { .offset = 0x7e, .value = 0x18, },
+		[72] = { .offset = 0x84, .value = 0x00, },
+		[73] = { .offset = 0x85, .value = 0x00, },
+		[74] = { .offset = 0x86, .value = 0x00, },
+		[75] = { .offset = 0x87, .value = 0x00, },
+		[76] = { .offset = 0x88, .value = 0x00, },
+		[77] = { .offset = 0x89, .value = 0x00, },
+		[78] = { .offset = 0x8a, .value = 0x00, },
+		[79] = { .offset = 0x8b, .value = 0x00, },
+		[80] = { .offset = 0x26, .value = 0x00, },
+		[81] = { .offset = 0x27, .value = 0x00, },
+		[82] = { .offset = 0xad, .value = 0x00, },
+		[83] = { .offset = 0x08, .value = 0x30, }, /* 0x31 */
+		[84] = { .offset = 0x41, .value = 0x00, },
+		[85] = { .offset = 0xc0, .value = 0x07, },
+	},
+	[MODE_1024x768] = {
+		[0] = { .offset = 0x0a, .value = 0x81, },
+		[1] = { .offset = 0x18, .value = 0x07, },
+		[2] = { .offset = 0x19, .value = 0x00, },
+		[3] = { .offset = 0x1a, .value = 0x00, },
+		[4] = { .offset = 0x1b, .value = 0x11, },
+		[5] = { .offset = 0x1c, .value = 0x54, },
+		[6] = { .offset = 0x1d, .value = 0x03, },
+		[7] = { .offset = 0x1e, .value = 0x02, },
+		[8] = { .offset = 0xf3, .value = 0x90, },
+		[9] = { .offset = 0xf9, .value = 0x00, },
+		[10] = { .offset = 0xc1, .value = 0x90, },
+		[11] = { .offset = 0xc2, .value = 0x00, },
+		[12] = { .offset = 0xc3, .value = 0x0f, },
+		[13] = { .offset = 0xc4, .value = 0x03, },
+		[14] = { .offset = 0xc5, .value = 0x16, },
+		[15] = { .offset = 0xc6, .value = 0x00, },
+		[16] = { .offset = 0xc7, .value = 0x02, },
+		[17] = { .offset = 0xc8, .value = 0x02, },
+		[18] = { .offset = 0xf4, .value = 0x00, },
+		[19] = { .offset = 0x80, .value = 0xff, },
+		[20] = { .offset = 0x81, .value = 0x07, },
+		[21] = { .offset = 0x82, .value = 0x3d, },
+		[22] = { .offset = 0x83, .value = 0x05, },
+		[23] = { .offset = 0x94, .value = 0x00, },
+		[24] = { .offset = 0x95, .value = 0x00, },
+		[25] = { .offset = 0x96, .value = 0x05, },
+		[26] = { .offset = 0x97, .value = 0x00, },
+		[27] = { .offset = 0x9a, .value = 0x88, },
+		[28] = { .offset = 0x9b, .value = 0x00, },
+		[29] = { .offset = 0x98, .value = 0x00, },
+		[30] = { .offset = 0x99, .value = 0x00, },
+		[31] = { .offset = 0xf7, .value = 0x88, },
+		[32] = { .offset = 0xf8, .value = 0x0a, },
+		[33] = { .offset = 0x9c, .value = 0x24, },
+		[34] = { .offset = 0x9d, .value = 0x00, },
+		[35] = { .offset = 0x9e, .value = 0x25, },
+		[36] = { .offset = 0x9f, .value = 0x03, },
+		[37] = { .offset = 0xa0, .value = 0x28, },
+		[38] = { .offset = 0xa1, .value = 0x01, },
+		[39] = { .offset = 0xa2, .value = 0x28, },
+		[40] = { .offset = 0xa3, .value = 0x05, },
+		[41] = { .offset = 0xb6, .value = 0x09, },
+		[42] = { .offset = 0xb8, .value = 0x00, },
+		[43] = { .offset = 0xb9, .value = 0xa0, },
+		[44] = { .offset = 0xba, .value = 0x00, },
+		[45] = { .offset = 0xbb, .value = 0x20, },
+		[46] = { .offset = 0x10, .value = 0x00, },
+		[47] = { .offset = 0x11, .value = 0xa0, },
+		[48] = { .offset = 0x12, .value = 0x02, },
+		[49] = { .offset = 0x20, .value = 0x00, },
+		[50] = { .offset = 0x22, .value = 0x00, },
+		[51] = { .offset = 0x23, .value = 0x00, },
+		[52] = { .offset = 0x24, .value = 0x00, },
+		[53] = { .offset = 0x25, .value = 0x00, },
+		[54] = { .offset = 0x8c, .value = 0x10, },
+		[55] = { .offset = 0x8d, .value = 0x02, },
+		[56] = { .offset = 0x8e, .value = 0x10, },
+		[57] = { .offset = 0x8f, .value = 0x00, },
+		[58] = { .offset = 0x90, .value = 0xff, },
+		[59] = { .offset = 0x91, .value = 0x07, },
+		[60] = { .offset = 0x92, .value = 0xa0, },
+		[61] = { .offset = 0x93, .value = 0x02, },
+		[62] = { .offset = 0xa5, .value = 0x00, },
+		[63] = { .offset = 0xa6, .value = 0x00, },
+		[64] = { .offset = 0xa7, .value = 0x00, },
+		[65] = { .offset = 0xa8, .value = 0x00, },
+		[66] = { .offset = 0xa9, .value = 0x04, },
+		[67] = { .offset = 0xaa, .value = 0x70, },
+		[68] = { .offset = 0xab, .value = 0x4f, },
+		[69] = { .offset = 0xac, .value = 0x00, },
+		[70] = { .offset = 0xa4, .value = 0x84, },
+		[71] = { .offset = 0x7e, .value = 0x18, },
+		[72] = { .offset = 0x84, .value = 0x00, },
+		[73] = { .offset = 0x85, .value = 0x00, },
+		[74] = { .offset = 0x86, .value = 0x00, },
+		[75] = { .offset = 0x87, .value = 0x00, },
+		[76] = { .offset = 0x88, .value = 0x00, },
+		[77] = { .offset = 0x89, .value = 0x00, },
+		[78] = { .offset = 0x8a, .value = 0x00, },
+		[79] = { .offset = 0x8b, .value = 0x00, },
+		[80] = { .offset = 0x26, .value = 0x00, },
+		[81] = { .offset = 0x27, .value = 0x00, },
+		[82] = { .offset = 0xad, .value = 0x00, },
+		[83] = { .offset = 0x08, .value = 0x34, }, /* 0x35 */
+		[84] = { .offset = 0x41, .value = 0x00, },
+		[85] = { .offset = 0xc0, .value = 0x01, },
+	},
+};
+
+static const struct ns2501_reg regs_init[] = {
+	[0] = { .offset = 0x35, .value = 0xff, },
+	[1] = { .offset = 0x34, .value = 0x00, },
+	[2] = { .offset = 0x08, .value = 0x30, },
+};
+
 struct ns2501_priv {
-	//I2CDevRec d;
 	bool quiet;
-	int reg_8_shadow;
-	int reg_8_set;
-	// Shadow registers for i915
-	int dvoc;
-	int pll_a;
-	int srcdim;
-	int fw_blc;
+	const struct ns2501_reg *regs;
 };
 
 #define NSPTR(d) ((NS2501Ptr)(d->DriverPrivate.ptr))
@@ -205,11 +486,9 @@ static bool ns2501_init(struct intel_dvo_device *dvo,
 		goto out;
 	}
 	ns->quiet = false;
-	ns->reg_8_set = 0;
-	ns->reg_8_shadow =
-	    NS2501_8_PD | NS2501_8_BPAS | NS2501_8_VEN | NS2501_8_HEN;
 
 	DRM_DEBUG_KMS("init ns2501 dvo controller successfully!\n");
+
 	return true;
 
 out:
@@ -242,9 +521,9 @@ static enum drm_mode_status ns2501_mode_valid(struct intel_dvo_device *dvo,
 	 * of the panel in here so we could always accept it
 	 * by disabling the scaler.
 	 */
-	if ((mode->hdisplay == 800 && mode->vdisplay == 600) ||
-	    (mode->hdisplay == 640 && mode->vdisplay == 480) ||
-	    (mode->hdisplay == 1024 && mode->vdisplay == 768)) {
+	if ((mode->hdisplay == 640 && mode->vdisplay == 480 && mode->clock == 25175) ||
+	    (mode->hdisplay == 800 && mode->vdisplay == 600 && mode->clock == 40000) ||
+	    (mode->hdisplay == 1024 && mode->vdisplay == 768 && mode->clock == 65000)) {
 		return MODE_OK;
 	} else {
 		return MODE_ONE_SIZE;	/* Is this a reasonable error? */
@@ -255,180 +534,30 @@ static void ns2501_mode_set(struct intel_dvo_device *dvo,
 			    struct drm_display_mode *mode,
 			    struct drm_display_mode *adjusted_mode)
 {
-	bool ok;
-	int retries = 10;
 	struct ns2501_priv *ns = (struct ns2501_priv *)(dvo->dev_priv);
+	int mode_idx, i;
 
 	DRM_DEBUG_KMS
 	    ("set mode (hdisplay=%d,htotal=%d,vdisplay=%d,vtotal=%d).\n",
 	     mode->hdisplay, mode->htotal, mode->vdisplay, mode->vtotal);
 
-	/*
-	 * Where do I find the native resolution for which scaling is not required???
-	 *
-	 * First trigger the DVO on as otherwise the chip does not appear on the i2c
-	 * bus.
-	 */
-	do {
-		ok = true;
-
-		if (mode->hdisplay == 800 && mode->vdisplay == 600) {
-			/* mode 277 */
-			ns->reg_8_shadow &= ~NS2501_8_BPAS;
-			DRM_DEBUG_KMS("switching to 800x600\n");
-
-			/*
-			 * No, I do not know where this data comes from.
-			 * It is just what the video bios left in the DVO, so
-			 * I'm just copying it here over.
-			 * This also means that I cannot support any other modes
-			 * except the ones supported by the bios.
-			 */
-			ok &= ns2501_writeb(dvo, 0x11, 0xc8);	// 0xc7 also works.
-			ok &= ns2501_writeb(dvo, 0x1b, 0x19);
-			ok &= ns2501_writeb(dvo, 0x1c, 0x62);	// VBIOS left 0x64 here, but 0x62 works nicer
-			ok &= ns2501_writeb(dvo, 0x1d, 0x02);
-
-			ok &= ns2501_writeb(dvo, 0x34, 0x03);
-			ok &= ns2501_writeb(dvo, 0x35, 0xff);
+	if (mode->hdisplay == 640 && mode->vdisplay == 480)
+		mode_idx = MODE_640x480;
+	else if (mode->hdisplay == 800 && mode->vdisplay == 600)
+		mode_idx = MODE_800x600;
+	else if (mode->hdisplay == 1024 && mode->vdisplay == 768)
+		mode_idx = MODE_1024x768;
+	else
+		return;
 
-			ok &= ns2501_writeb(dvo, 0x80, 0x27);
-			ok &= ns2501_writeb(dvo, 0x81, 0x03);
-			ok &= ns2501_writeb(dvo, 0x82, 0x41);
-			ok &= ns2501_writeb(dvo, 0x83, 0x05);
+	/* Hopefully doing it every time won't hurt... */
+	for (i = 0; i < ARRAY_SIZE(regs_init); i++)
+		ns2501_writeb(dvo, regs_init[i].offset, regs_init[i].value);
 
-			ok &= ns2501_writeb(dvo, 0x8d, 0x02);
-			ok &= ns2501_writeb(dvo, 0x8e, 0x04);
-			ok &= ns2501_writeb(dvo, 0x8f, 0x00);
+	ns->regs = regs_1024x768[mode_idx];
 
-			ok &= ns2501_writeb(dvo, 0x90, 0xfe);	/* vertical. VBIOS left 0xff here, but 0xfe works better */
-			ok &= ns2501_writeb(dvo, 0x91, 0x07);
-			ok &= ns2501_writeb(dvo, 0x94, 0x00);
-			ok &= ns2501_writeb(dvo, 0x95, 0x00);
-
-			ok &= ns2501_writeb(dvo, 0x96, 0x00);
-
-			ok &= ns2501_writeb(dvo, 0x99, 0x00);
-			ok &= ns2501_writeb(dvo, 0x9a, 0x88);
-
-			ok &= ns2501_writeb(dvo, 0x9c, 0x23);	/* Looks like first and last line of the image. */
-			ok &= ns2501_writeb(dvo, 0x9d, 0x00);
-			ok &= ns2501_writeb(dvo, 0x9e, 0x25);
-			ok &= ns2501_writeb(dvo, 0x9f, 0x03);
-
-			ok &= ns2501_writeb(dvo, 0xa4, 0x80);
-
-			ok &= ns2501_writeb(dvo, 0xb6, 0x00);
-
-			ok &= ns2501_writeb(dvo, 0xb9, 0xc8);	/* horizontal? */
-			ok &= ns2501_writeb(dvo, 0xba, 0x00);	/* horizontal? */
-
-			ok &= ns2501_writeb(dvo, 0xc0, 0x05);	/* horizontal? */
-			ok &= ns2501_writeb(dvo, 0xc1, 0xd7);
-
-			ok &= ns2501_writeb(dvo, 0xc2, 0x00);
-			ok &= ns2501_writeb(dvo, 0xc3, 0xf8);
-
-			ok &= ns2501_writeb(dvo, 0xc4, 0x03);
-			ok &= ns2501_writeb(dvo, 0xc5, 0x1a);
-
-			ok &= ns2501_writeb(dvo, 0xc6, 0x00);
-			ok &= ns2501_writeb(dvo, 0xc7, 0x73);
-			ok &= ns2501_writeb(dvo, 0xc8, 0x02);
-
-		} else if (mode->hdisplay == 640 && mode->vdisplay == 480) {
-			/* mode 274 */
-			DRM_DEBUG_KMS("switching to 640x480\n");
-			/*
-			 * No, I do not know where this data comes from.
-			 * It is just what the video bios left in the DVO, so
-			 * I'm just copying it here over.
-			 * This also means that I cannot support any other modes
-			 * except the ones supported by the bios.
-			 */
-			ns->reg_8_shadow &= ~NS2501_8_BPAS;
-
-			ok &= ns2501_writeb(dvo, 0x11, 0xa0);
-			ok &= ns2501_writeb(dvo, 0x1b, 0x11);
-			ok &= ns2501_writeb(dvo, 0x1c, 0x54);
-			ok &= ns2501_writeb(dvo, 0x1d, 0x03);
-
-			ok &= ns2501_writeb(dvo, 0x34, 0x03);
-			ok &= ns2501_writeb(dvo, 0x35, 0xff);
-
-			ok &= ns2501_writeb(dvo, 0x80, 0xff);
-			ok &= ns2501_writeb(dvo, 0x81, 0x07);
-			ok &= ns2501_writeb(dvo, 0x82, 0x3d);
-			ok &= ns2501_writeb(dvo, 0x83, 0x05);
-
-			ok &= ns2501_writeb(dvo, 0x8d, 0x02);
-			ok &= ns2501_writeb(dvo, 0x8e, 0x10);
-			ok &= ns2501_writeb(dvo, 0x8f, 0x00);
-
-			ok &= ns2501_writeb(dvo, 0x90, 0xff);	/* vertical */
-			ok &= ns2501_writeb(dvo, 0x91, 0x07);
-			ok &= ns2501_writeb(dvo, 0x94, 0x00);
-			ok &= ns2501_writeb(dvo, 0x95, 0x00);
-
-			ok &= ns2501_writeb(dvo, 0x96, 0x05);
-
-			ok &= ns2501_writeb(dvo, 0x99, 0x00);
-			ok &= ns2501_writeb(dvo, 0x9a, 0x88);
-
-			ok &= ns2501_writeb(dvo, 0x9c, 0x24);
-			ok &= ns2501_writeb(dvo, 0x9d, 0x00);
-			ok &= ns2501_writeb(dvo, 0x9e, 0x25);
-			ok &= ns2501_writeb(dvo, 0x9f, 0x03);
-
-			ok &= ns2501_writeb(dvo, 0xa4, 0x84);
-
-			ok &= ns2501_writeb(dvo, 0xb6, 0x09);
-
-			ok &= ns2501_writeb(dvo, 0xb9, 0xa0);	/* horizontal? */
-			ok &= ns2501_writeb(dvo, 0xba, 0x00);	/* horizontal? */
-
-			ok &= ns2501_writeb(dvo, 0xc0, 0x05);	/* horizontal? */
-			ok &= ns2501_writeb(dvo, 0xc1, 0x90);
-
-			ok &= ns2501_writeb(dvo, 0xc2, 0x00);
-			ok &= ns2501_writeb(dvo, 0xc3, 0x0f);
-
-			ok &= ns2501_writeb(dvo, 0xc4, 0x03);
-			ok &= ns2501_writeb(dvo, 0xc5, 0x16);
-
-			ok &= ns2501_writeb(dvo, 0xc6, 0x00);
-			ok &= ns2501_writeb(dvo, 0xc7, 0x02);
-			ok &= ns2501_writeb(dvo, 0xc8, 0x02);
-
-		} else if (mode->hdisplay == 1024 && mode->vdisplay == 768) {
-			/* mode 280 */
-			DRM_DEBUG_KMS("switching to 1024x768\n");
-			/*
-			 * This might or might not work, actually. I'm silently
-			 * assuming here that the native panel resolution is
-			 * 1024x768. If not, then this leaves the scaler disabled
-			 * generating a picture that is likely not the expected.
-			 *
-			 * Problem is that I do not know where to take the panel
-			 * dimensions from.
-			 *
-			 * Enable the bypass, scaling not required.
-			 *
-			 * The scaler registers are irrelevant here....
-			 *
-			 */
-			ns->reg_8_shadow |= NS2501_8_BPAS;
-			ok &= ns2501_writeb(dvo, 0x37, 0x44);
-		} else {
-			/*
-			 * Data not known. Bummer!
-			 * Hopefully, the code should not go here
-			 * as mode_OK delivered no other modes.
-			 */
-			ns->reg_8_shadow |= NS2501_8_BPAS;
-		}
-		ok &= ns2501_writeb(dvo, NS2501_REG8, ns->reg_8_shadow);
-	} while (!ok && retries--);
+	for (i = 0; i < 84; i++)
+		ns2501_writeb(dvo, ns->regs[i].offset, ns->regs[i].value);
 }
 
 /* set the NS2501 power state */
@@ -439,60 +568,46 @@ static bool ns2501_get_hw_state(struct intel_dvo_device *dvo)
 	if (!ns2501_readb(dvo, NS2501_REG8, &ch))
 		return false;
 
-	if (ch & NS2501_8_PD)
-		return true;
-	else
-		return false;
+	return ch & NS2501_8_PD;
 }
 
 /* set the NS2501 power state */
 static void ns2501_dpms(struct intel_dvo_device *dvo, bool enable)
 {
-	bool ok;
-	int retries = 10;
 	struct ns2501_priv *ns = (struct ns2501_priv *)(dvo->dev_priv);
-	unsigned char ch;
 
 	DRM_DEBUG_KMS("Trying set the dpms of the DVO to %i\n", enable);
 
-	ch = ns->reg_8_shadow;
+	if (enable) {
+		if (WARN_ON(ns->regs[83].offset != 0x08 ||
+			    ns->regs[84].offset != 0x41 ||
+			    ns->regs[85].offset != 0xc0))
+			return;
 
-	if (enable)
-		ch |= NS2501_8_PD;
-	else
-		ch &= ~NS2501_8_PD;
-
-	if (ns->reg_8_set == 0 || ns->reg_8_shadow != ch) {
-		ns->reg_8_set = 1;
-		ns->reg_8_shadow = ch;
-
-		do {
-			ok = true;
-			ok &= ns2501_writeb(dvo, NS2501_REG8, ch);
-			ok &=
-			    ns2501_writeb(dvo, 0x34,
-					  enable ? 0x03 : 0x00);
-			ok &=
-			    ns2501_writeb(dvo, 0x35,
-					  enable ? 0xff : 0x00);
-		} while (!ok && retries--);
-	}
-}
+		ns2501_writeb(dvo, 0xc0, ns->regs[85].value | 0x08);
 
-static void ns2501_dump_regs(struct intel_dvo_device *dvo)
-{
-	uint8_t val;
-
-	ns2501_readb(dvo, NS2501_FREQ_LO, &val);
-	DRM_DEBUG_KMS("NS2501_FREQ_LO: 0x%02x\n", val);
-	ns2501_readb(dvo, NS2501_FREQ_HI, &val);
-	DRM_DEBUG_KMS("NS2501_FREQ_HI: 0x%02x\n", val);
-	ns2501_readb(dvo, NS2501_REG8, &val);
-	DRM_DEBUG_KMS("NS2501_REG8: 0x%02x\n", val);
-	ns2501_readb(dvo, NS2501_REG9, &val);
-	DRM_DEBUG_KMS("NS2501_REG9: 0x%02x\n", val);
-	ns2501_readb(dvo, NS2501_REGC, &val);
-	DRM_DEBUG_KMS("NS2501_REGC: 0x%02x\n", val);
+		ns2501_writeb(dvo, 0x41, ns->regs[84].value);
+
+		ns2501_writeb(dvo, 0x34, 0x01);
+		msleep(15);
+
+		ns2501_writeb(dvo, 0x08, 0x35);
+		if (!(ns->regs[83].value & NS2501_8_BPAS))
+			ns2501_writeb(dvo, 0x08, 0x31);
+		msleep(200);
+
+		ns2501_writeb(dvo, 0x34, 0x03);
+
+		ns2501_writeb(dvo, 0xc0, ns->regs[85].value);
+	} else {
+		ns2501_writeb(dvo, 0x34, 0x01);
+		msleep(200);
+
+		ns2501_writeb(dvo, 0x08, 0x34);
+		msleep(15);
+
+		ns2501_writeb(dvo, 0x34, 0x00);
+	}
 }
 
 static void ns2501_destroy(struct intel_dvo_device *dvo)
@@ -512,6 +627,5 @@ struct intel_dvo_dev_ops ns2501_ops = {
 	.mode_set = ns2501_mode_set,
 	.dpms = ns2501_dpms,
 	.get_hw_state = ns2501_get_hw_state,
-	.dump_regs = ns2501_dump_regs,
 	.destroy = ns2501_destroy,
 };
diff --git a/drivers/gpu/drm/i915/i915_cmd_parser.c b/drivers/gpu/drm/i915/i915_cmd_parser.c
index 4b7ed5289217..22c992a78ac6 100644
--- a/drivers/gpu/drm/i915/i915_cmd_parser.c
+++ b/drivers/gpu/drm/i915/i915_cmd_parser.c
@@ -73,7 +73,7 @@
  * those commands required by the parser. This generally works because command
  * opcode ranges have standard command length encodings. So for commands that
  * the parser does not need to check, it can easily skip them. This is
- * implementated via a per-ring length decoding vfunc.
+ * implemented via a per-ring length decoding vfunc.
  *
  * Unfortunately, there are a number of commands that do not follow the standard
  * length encoding for their opcode range, primarily amongst the MI_* commands.
@@ -138,6 +138,11 @@ static const struct drm_i915_cmd_descriptor common_cmds[] = {
 			.mask = MI_GLOBAL_GTT,
 			.expected = 0,
 	      }},						       ),
+	/*
+	 * MI_BATCH_BUFFER_START requires some special handling. It's not
+	 * really a 'skip' action but it doesn't seem like it's worth adding
+	 * a new action. See i915_parse_cmds().
+	 */
 	CMD(  MI_BATCH_BUFFER_START,            SMI,   !F,  0xFF,   S  ),
 };
 
@@ -408,6 +413,8 @@ static const u32 gen7_render_regs[] = {
 	REG64(PS_INVOCATION_COUNT),
 	REG64(PS_DEPTH_COUNT),
 	OACONTROL, /* Only allowed for LRI and SRM. See below. */
+	REG64(MI_PREDICATE_SRC0),
+	REG64(MI_PREDICATE_SRC1),
 	GEN7_3DPRIM_END_OFFSET,
 	GEN7_3DPRIM_START_VERTEX,
 	GEN7_3DPRIM_VERTEX_COUNT,
@@ -838,23 +845,16 @@ finish:
  * @ring: the ring in question
  *
  * Only certain platforms require software batch buffer command parsing, and
- * only when enabled via module paramter.
+ * only when enabled via module parameter.
  *
  * Return: true if the ring requires software command parsing
  */
 bool i915_needs_cmd_parser(struct intel_engine_cs *ring)
 {
-	struct drm_i915_private *dev_priv = ring->dev->dev_private;
-
 	if (!ring->needs_cmd_parser)
 		return false;
 
-	/*
-	 * XXX: VLV is Gen7 and therefore has cmd_tables, but has PPGTT
-	 * disabled. That will cause all of the parser's PPGTT checks to
-	 * fail. For now, disable parsing when PPGTT is off.
-	 */
-	if (!dev_priv->mm.aliasing_ppgtt)
+	if (!USES_PPGTT(ring->dev))
 		return false;
 
 	return (i915.enable_cmd_parser == 1);
@@ -890,8 +890,10 @@ static bool check_cmd(const struct intel_engine_cs *ring,
 		 * OACONTROL writes to only MI_LOAD_REGISTER_IMM commands.
 		 */
 		if (reg_addr == OACONTROL) {
-			if (desc->cmd.value == MI_LOAD_REGISTER_MEM)
+			if (desc->cmd.value == MI_LOAD_REGISTER_MEM) {
+				DRM_DEBUG_DRIVER("CMD: Rejected LRM to OACONTROL\n");
 				return false;
+			}
 
 			if (desc->cmd.value == MI_LOAD_REGISTER_IMM(1))
 				*oacontrol_set = (cmd[2] != 0);
@@ -960,7 +962,8 @@ static bool check_cmd(const struct intel_engine_cs *ring,
  * Parses the specified batch buffer looking for privilege violations as
  * described in the overview.
  *
- * Return: non-zero if the parser finds violations or otherwise fails
+ * Return: non-zero if the parser finds violations or otherwise fails; -EACCES
+ * if the batch appears legal but should use hardware parsing
  */
 int i915_parse_cmds(struct intel_engine_cs *ring,
 		    struct drm_i915_gem_object *batch_obj,
@@ -1007,6 +1010,16 @@ int i915_parse_cmds(struct intel_engine_cs *ring,
 			break;
 		}
 
+		/*
+		 * If the batch buffer contains a chained batch, return an
+		 * error that tells the caller to abort and dispatch the
+		 * workload as a non-secure batch.
+		 */
+		if (desc->cmd.value == MI_BATCH_BUFFER_START) {
+			ret = -EACCES;
+			break;
+		}
+
 		if (desc->flags & CMD_DESC_FIXED)
 			length = desc->length.fixed;
 		else
@@ -1061,6 +1074,8 @@ int i915_cmd_parser_get_version(void)
 	 *
 	 * 1. Initial version. Checks batches and reports violations, but leaves
 	 *    hardware parsing enabled (so does not allow new use cases).
+	 * 2. Allow access to the MI_PREDICATE_SRC0 and
+	 *    MI_PREDICATE_SRC1 registers.
 	 */
-	return 1;
+	return 2;
 }
diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 9e737b771c40..779a275eb1fd 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -116,7 +116,7 @@ static const char *get_tiling_flag(struct drm_i915_gem_object *obj)
 
 static inline const char *get_global_flag(struct drm_i915_gem_object *obj)
 {
-	return obj->has_global_gtt_mapping ? "g" : " ";
+	return i915_gem_obj_to_ggtt(obj) ? "g" : " ";
 }
 
 static void
@@ -136,7 +136,7 @@ describe_obj(struct seq_file *m, struct drm_i915_gem_object *obj)
 		   obj->last_read_seqno,
 		   obj->last_write_seqno,
 		   obj->last_fenced_seqno,
-		   i915_cache_level_str(obj->cache_level),
+		   i915_cache_level_str(to_i915(obj->base.dev), obj->cache_level),
 		   obj->dirty ? " dirty" : "",
 		   obj->madv == I915_MADV_DONTNEED ? " purgeable" : "");
 	if (obj->base.name)
@@ -333,7 +333,7 @@ static int per_file_stats(int id, void *ptr, void *data)
 			}
 
 			ppgtt = container_of(vma->vm, struct i915_hw_ppgtt, base);
-			if (ppgtt->ctx && ppgtt->ctx->file_priv != stats->file_priv)
+			if (ppgtt->file_priv != stats->file_priv)
 				continue;
 
 			if (obj->ring) /* XXX per-vma statistic */
@@ -515,7 +515,7 @@ static int i915_gem_pageflip_info(struct seq_file *m, void *data)
 {
 	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	unsigned long flags;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *crtc;
 	int ret;
 
@@ -528,12 +528,14 @@ static int i915_gem_pageflip_info(struct seq_file *m, void *data)
 		const char plane = plane_name(crtc->plane);
 		struct intel_unpin_work *work;
 
-		spin_lock_irqsave(&dev->event_lock, flags);
+		spin_lock_irq(&dev->event_lock);
 		work = crtc->unpin_work;
 		if (work == NULL) {
 			seq_printf(m, "No flip due on pipe %c (plane %c)\n",
 				   pipe, plane);
 		} else {
+			u32 addr;
+
 			if (atomic_read(&work->pending) < INTEL_FLIP_COMPLETE) {
 				seq_printf(m, "Flip queued on pipe %c (plane %c)\n",
 					   pipe, plane);
@@ -541,26 +543,38 @@ static int i915_gem_pageflip_info(struct seq_file *m, void *data)
 				seq_printf(m, "Flip pending (waiting for vsync) on pipe %c (plane %c)\n",
 					   pipe, plane);
 			}
+			if (work->flip_queued_ring) {
+				seq_printf(m, "Flip queued on %s at seqno %u, next seqno %u [current breadcrumb %u], completed? %d\n",
+					   work->flip_queued_ring->name,
+					   work->flip_queued_seqno,
+					   dev_priv->next_seqno,
+					   work->flip_queued_ring->get_seqno(work->flip_queued_ring, true),
+					   i915_seqno_passed(work->flip_queued_ring->get_seqno(work->flip_queued_ring, true),
+							     work->flip_queued_seqno));
+			} else
+				seq_printf(m, "Flip not associated with any ring\n");
+			seq_printf(m, "Flip queued on frame %d, (was ready on frame %d), now %d\n",
+				   work->flip_queued_vblank,
+				   work->flip_ready_vblank,
+				   drm_vblank_count(dev, crtc->pipe));
 			if (work->enable_stall_check)
 				seq_puts(m, "Stall check enabled, ");
 			else
 				seq_puts(m, "Stall check waiting for page flip ioctl, ");
 			seq_printf(m, "%d prepares\n", atomic_read(&work->pending));
 
-			if (work->old_fb_obj) {
-				struct drm_i915_gem_object *obj = work->old_fb_obj;
-				if (obj)
-					seq_printf(m, "Old framebuffer gtt_offset 0x%08lx\n",
-						   i915_gem_obj_ggtt_offset(obj));
-			}
+			if (INTEL_INFO(dev)->gen >= 4)
+				addr = I915_HI_DISPBASE(I915_READ(DSPSURF(crtc->plane)));
+			else
+				addr = I915_READ(DSPADDR(crtc->plane));
+			seq_printf(m, "Current scanout address 0x%08x\n", addr);
+
 			if (work->pending_flip_obj) {
-				struct drm_i915_gem_object *obj = work->pending_flip_obj;
-				if (obj)
-					seq_printf(m, "New framebuffer gtt_offset 0x%08lx\n",
-						   i915_gem_obj_ggtt_offset(obj));
+				seq_printf(m, "New framebuffer address 0x%08lx\n", (long)work->gtt_offset);
+				seq_printf(m, "MMIO update completed? %d\n",  addr == work->gtt_offset);
 			}
 		}
-		spin_unlock_irqrestore(&dev->event_lock, flags);
+		spin_unlock_irq(&dev->event_lock);
 	}
 
 	mutex_unlock(&dev->struct_mutex);
@@ -650,7 +664,6 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 	intel_runtime_pm_get(dev_priv);
 
 	if (IS_CHERRYVIEW(dev)) {
-		int i;
 		seq_printf(m, "Master Interrupt Control:\t%08x\n",
 			   I915_READ(GEN8_MASTER_IRQ));
 
@@ -662,7 +675,7 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 			   I915_READ(VLV_IIR_RW));
 		seq_printf(m, "Display IMR:\t%08x\n",
 			   I915_READ(VLV_IMR));
-		for_each_pipe(pipe)
+		for_each_pipe(dev_priv, pipe)
 			seq_printf(m, "Pipe %c stat:\t%08x\n",
 				   pipe_name(pipe),
 				   I915_READ(PIPESTAT(pipe)));
@@ -702,7 +715,13 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 				   i, I915_READ(GEN8_GT_IER(i)));
 		}
 
-		for_each_pipe(pipe) {
+		for_each_pipe(dev_priv, pipe) {
+			if (!intel_display_power_is_enabled(dev_priv,
+						POWER_DOMAIN_PIPE(pipe))) {
+				seq_printf(m, "Pipe %c power disabled\n",
+					   pipe_name(pipe));
+				continue;
+			}
 			seq_printf(m, "Pipe %c IMR:\t%08x\n",
 				   pipe_name(pipe),
 				   I915_READ(GEN8_DE_PIPE_IMR(pipe)));
@@ -743,7 +762,7 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 			   I915_READ(VLV_IIR_RW));
 		seq_printf(m, "Display IMR:\t%08x\n",
 			   I915_READ(VLV_IMR));
-		for_each_pipe(pipe)
+		for_each_pipe(dev_priv, pipe)
 			seq_printf(m, "Pipe %c stat:\t%08x\n",
 				   pipe_name(pipe),
 				   I915_READ(PIPESTAT(pipe)));
@@ -779,7 +798,7 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 			   I915_READ(IIR));
 		seq_printf(m, "Interrupt mask:      %08x\n",
 			   I915_READ(IMR));
-		for_each_pipe(pipe)
+		for_each_pipe(dev_priv, pipe)
 			seq_printf(m, "Pipe %c stat:         %08x\n",
 				   pipe_name(pipe),
 				   I915_READ(PIPESTAT(pipe)));
@@ -927,7 +946,7 @@ static ssize_t i915_error_state_read(struct file *file, char __user *userbuf,
 	ssize_t ret_count = 0;
 	int ret;
 
-	ret = i915_error_state_buf_init(&error_str, count, *pos);
+	ret = i915_error_state_buf_init(&error_str, to_i915(error_priv->dev), count, *pos);
 	if (ret)
 		return ret;
 
@@ -1024,6 +1043,7 @@ static int i915_frequency_info(struct seq_file *m, void *unused)
 		u32 rpstat, cagf, reqf;
 		u32 rpupei, rpcurup, rpprevup;
 		u32 rpdownei, rpcurdown, rpprevdown;
+		u32 pm_ier, pm_imr, pm_isr, pm_iir, pm_mask;
 		int max_freq;
 
 		/* RPSTAT1 is in the GT power well */
@@ -1061,12 +1081,21 @@ static int i915_frequency_info(struct seq_file *m, void *unused)
 		gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 		mutex_unlock(&dev->struct_mutex);
 
+		if (IS_GEN6(dev) || IS_GEN7(dev)) {
+			pm_ier = I915_READ(GEN6_PMIER);
+			pm_imr = I915_READ(GEN6_PMIMR);
+			pm_isr = I915_READ(GEN6_PMISR);
+			pm_iir = I915_READ(GEN6_PMIIR);
+			pm_mask = I915_READ(GEN6_PMINTRMSK);
+		} else {
+			pm_ier = I915_READ(GEN8_GT_IER(2));
+			pm_imr = I915_READ(GEN8_GT_IMR(2));
+			pm_isr = I915_READ(GEN8_GT_ISR(2));
+			pm_iir = I915_READ(GEN8_GT_IIR(2));
+			pm_mask = I915_READ(GEN6_PMINTRMSK);
+		}
 		seq_printf(m, "PM IER=0x%08x IMR=0x%08x ISR=0x%08x IIR=0x%08x, MASK=0x%08x\n",
-			   I915_READ(GEN6_PMIER),
-			   I915_READ(GEN6_PMIMR),
-			   I915_READ(GEN6_PMISR),
-			   I915_READ(GEN6_PMIIR),
-			   I915_READ(GEN6_PMINTRMSK));
+			   pm_ier, pm_imr, pm_isr, pm_iir, pm_mask);
 		seq_printf(m, "GT_PERF_STATUS: 0x%08x\n", gt_perf_status);
 		seq_printf(m, "Render p-state ratio: %d\n",
 			   (gt_perf_status & 0xff00) >> 8);
@@ -1211,11 +1240,12 @@ static int vlv_drpc_info(struct seq_file *m)
 	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 rpmodectl1, rcctl1;
+	u32 rpmodectl1, rcctl1, pw_status;
 	unsigned fw_rendercount = 0, fw_mediacount = 0;
 
 	intel_runtime_pm_get(dev_priv);
 
+	pw_status = I915_READ(VLV_GTLC_PW_STATUS);
 	rpmodectl1 = I915_READ(GEN6_RP_CONTROL);
 	rcctl1 = I915_READ(GEN6_RC_CONTROL);
 
@@ -1234,11 +1264,9 @@ static int vlv_drpc_info(struct seq_file *m)
 		   yesno(rcctl1 & (GEN7_RC_CTL_TO_MODE |
 					GEN6_RC_CTL_EI_MODE(1))));
 	seq_printf(m, "Render Power Well: %s\n",
-			(I915_READ(VLV_GTLC_PW_STATUS) &
-				VLV_GTLC_PW_RENDER_STATUS_MASK) ? "Up" : "Down");
+		   (pw_status & VLV_GTLC_PW_RENDER_STATUS_MASK) ? "Up" : "Down");
 	seq_printf(m, "Media Power Well: %s\n",
-			(I915_READ(VLV_GTLC_PW_STATUS) &
-				VLV_GTLC_PW_MEDIA_STATUS_MASK) ? "Up" : "Down");
+		   (pw_status & VLV_GTLC_PW_MEDIA_STATUS_MASK) ? "Up" : "Down");
 
 	seq_printf(m, "Render RC6 residency since boot: %u\n",
 		   I915_READ(VLV_GT_RENDER_RC6));
@@ -1365,7 +1393,7 @@ static int i915_drpc_info(struct seq_file *m, void *unused)
 
 	if (IS_VALLEYVIEW(dev))
 		return vlv_drpc_info(m);
-	else if (IS_GEN6(dev) || IS_GEN7(dev))
+	else if (INTEL_INFO(dev)->gen >= 6)
 		return gen6_drpc_info(m);
 	else
 		return ironlake_drpc_info(m);
@@ -1433,6 +1461,47 @@ static int i915_fbc_status(struct seq_file *m, void *unused)
 	return 0;
 }
 
+static int i915_fbc_fc_get(void *data, u64 *val)
+{
+	struct drm_device *dev = data;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (INTEL_INFO(dev)->gen < 7 || !HAS_FBC(dev))
+		return -ENODEV;
+
+	drm_modeset_lock_all(dev);
+	*val = dev_priv->fbc.false_color;
+	drm_modeset_unlock_all(dev);
+
+	return 0;
+}
+
+static int i915_fbc_fc_set(void *data, u64 val)
+{
+	struct drm_device *dev = data;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 reg;
+
+	if (INTEL_INFO(dev)->gen < 7 || !HAS_FBC(dev))
+		return -ENODEV;
+
+	drm_modeset_lock_all(dev);
+
+	reg = I915_READ(ILK_DPFC_CONTROL);
+	dev_priv->fbc.false_color = val;
+
+	I915_WRITE(ILK_DPFC_CONTROL, val ?
+		   (reg | FBC_CTL_FALSE_COLOR) :
+		   (reg & ~FBC_CTL_FALSE_COLOR));
+
+	drm_modeset_unlock_all(dev);
+	return 0;
+}
+
+DEFINE_SIMPLE_ATTRIBUTE(i915_fbc_fc_fops,
+			i915_fbc_fc_get, i915_fbc_fc_set,
+			"%llu\n");
+
 static int i915_ips_status(struct seq_file *m, void *unused)
 {
 	struct drm_info_node *node = m->private;
@@ -1630,6 +1699,14 @@ static int i915_gem_framebuffer_info(struct seq_file *m, void *data)
 	return 0;
 }
 
+static void describe_ctx_ringbuf(struct seq_file *m,
+				 struct intel_ringbuffer *ringbuf)
+{
+	seq_printf(m, " (ringbuffer, space: %d, head: %u, tail: %u, last head: %d)",
+		   ringbuf->space, ringbuf->head, ringbuf->tail,
+		   ringbuf->last_retired_head);
+}
+
 static int i915_context_status(struct seq_file *m, void *unused)
 {
 	struct drm_info_node *node = m->private;
@@ -1656,19 +1733,198 @@ static int i915_context_status(struct seq_file *m, void *unused)
 	}
 
 	list_for_each_entry(ctx, &dev_priv->context_list, link) {
-		if (ctx->legacy_hw_ctx.rcs_state == NULL)
+		if (!i915.enable_execlists &&
+		    ctx->legacy_hw_ctx.rcs_state == NULL)
 			continue;
 
 		seq_puts(m, "HW context ");
 		describe_ctx(m, ctx);
-		for_each_ring(ring, dev_priv, i)
+		for_each_ring(ring, dev_priv, i) {
 			if (ring->default_context == ctx)
-				seq_printf(m, "(default context %s) ", ring->name);
+				seq_printf(m, "(default context %s) ",
+					   ring->name);
+		}
+
+		if (i915.enable_execlists) {
+			seq_putc(m, '\n');
+			for_each_ring(ring, dev_priv, i) {
+				struct drm_i915_gem_object *ctx_obj =
+					ctx->engine[i].state;
+				struct intel_ringbuffer *ringbuf =
+					ctx->engine[i].ringbuf;
+
+				seq_printf(m, "%s: ", ring->name);
+				if (ctx_obj)
+					describe_obj(m, ctx_obj);
+				if (ringbuf)
+					describe_ctx_ringbuf(m, ringbuf);
+				seq_putc(m, '\n');
+			}
+		} else {
+			describe_obj(m, ctx->legacy_hw_ctx.rcs_state);
+		}
+
+		seq_putc(m, '\n');
+	}
+
+	mutex_unlock(&dev->struct_mutex);
+
+	return 0;
+}
+
+static void i915_dump_lrc_obj(struct seq_file *m,
+			      struct intel_engine_cs *ring,
+			      struct drm_i915_gem_object *ctx_obj)
+{
+	struct page *page;
+	uint32_t *reg_state;
+	int j;
+	unsigned long ggtt_offset = 0;
+
+	if (ctx_obj == NULL) {
+		seq_printf(m, "Context on %s with no gem object\n",
+			   ring->name);
+		return;
+	}
+
+	seq_printf(m, "CONTEXT: %s %u\n", ring->name,
+		   intel_execlists_ctx_id(ctx_obj));
+
+	if (!i915_gem_obj_ggtt_bound(ctx_obj))
+		seq_puts(m, "\tNot bound in GGTT\n");
+	else
+		ggtt_offset = i915_gem_obj_ggtt_offset(ctx_obj);
+
+	if (i915_gem_object_get_pages(ctx_obj)) {
+		seq_puts(m, "\tFailed to get pages for context object\n");
+		return;
+	}
+
+	page = i915_gem_object_get_page(ctx_obj, 1);
+	if (!WARN_ON(page == NULL)) {
+		reg_state = kmap_atomic(page);
+
+		for (j = 0; j < 0x600 / sizeof(u32) / 4; j += 4) {
+			seq_printf(m, "\t[0x%08lx] 0x%08x 0x%08x 0x%08x 0x%08x\n",
+				   ggtt_offset + 4096 + (j * 4),
+				   reg_state[j], reg_state[j + 1],
+				   reg_state[j + 2], reg_state[j + 3]);
+		}
+		kunmap_atomic(reg_state);
+	}
+
+	seq_putc(m, '\n');
+}
+
+static int i915_dump_lrc(struct seq_file *m, void *unused)
+{
+	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
+	struct intel_context *ctx;
+	int ret, i;
+
+	if (!i915.enable_execlists) {
+		seq_printf(m, "Logical Ring Contexts are disabled\n");
+		return 0;
+	}
+
+	ret = mutex_lock_interruptible(&dev->struct_mutex);
+	if (ret)
+		return ret;
+
+	list_for_each_entry(ctx, &dev_priv->context_list, link) {
+		for_each_ring(ring, dev_priv, i) {
+			if (ring->default_context != ctx)
+				i915_dump_lrc_obj(m, ring,
+						  ctx->engine[i].state);
+		}
+	}
+
+	mutex_unlock(&dev->struct_mutex);
+
+	return 0;
+}
+
+static int i915_execlists(struct seq_file *m, void *data)
+{
+	struct drm_info_node *node = (struct drm_info_node *)m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
+	u32 status_pointer;
+	u8 read_pointer;
+	u8 write_pointer;
+	u32 status;
+	u32 ctx_id;
+	struct list_head *cursor;
+	int ring_id, i;
+	int ret;
+
+	if (!i915.enable_execlists) {
+		seq_puts(m, "Logical Ring Contexts are disabled\n");
+		return 0;
+	}
+
+	ret = mutex_lock_interruptible(&dev->struct_mutex);
+	if (ret)
+		return ret;
+
+	intel_runtime_pm_get(dev_priv);
+
+	for_each_ring(ring, dev_priv, ring_id) {
+		struct intel_ctx_submit_request *head_req = NULL;
+		int count = 0;
+		unsigned long flags;
+
+		seq_printf(m, "%s\n", ring->name);
+
+		status = I915_READ(RING_EXECLIST_STATUS(ring));
+		ctx_id = I915_READ(RING_EXECLIST_STATUS(ring) + 4);
+		seq_printf(m, "\tExeclist status: 0x%08X, context: %u\n",
+			   status, ctx_id);
+
+		status_pointer = I915_READ(RING_CONTEXT_STATUS_PTR(ring));
+		seq_printf(m, "\tStatus pointer: 0x%08X\n", status_pointer);
+
+		read_pointer = ring->next_context_status_buffer;
+		write_pointer = status_pointer & 0x07;
+		if (read_pointer > write_pointer)
+			write_pointer += 6;
+		seq_printf(m, "\tRead pointer: 0x%08X, write pointer 0x%08X\n",
+			   read_pointer, write_pointer);
+
+		for (i = 0; i < 6; i++) {
+			status = I915_READ(RING_CONTEXT_STATUS_BUF(ring) + 8*i);
+			ctx_id = I915_READ(RING_CONTEXT_STATUS_BUF(ring) + 8*i + 4);
+
+			seq_printf(m, "\tStatus buffer %d: 0x%08X, context: %u\n",
+				   i, status, ctx_id);
+		}
+
+		spin_lock_irqsave(&ring->execlist_lock, flags);
+		list_for_each(cursor, &ring->execlist_queue)
+			count++;
+		head_req = list_first_entry_or_null(&ring->execlist_queue,
+				struct intel_ctx_submit_request, execlist_link);
+		spin_unlock_irqrestore(&ring->execlist_lock, flags);
+
+		seq_printf(m, "\t%d requests in queue\n", count);
+		if (head_req) {
+			struct drm_i915_gem_object *ctx_obj;
+
+			ctx_obj = head_req->ctx->engine[ring_id].state;
+			seq_printf(m, "\tHead request id: %u\n",
+				   intel_execlists_ctx_id(ctx_obj));
+			seq_printf(m, "\tHead request tail: %u\n",
+				   head_req->tail);
+		}
 
-		describe_obj(m, ctx->legacy_hw_ctx.rcs_state);
 		seq_putc(m, '\n');
 	}
 
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
@@ -1742,6 +1998,8 @@ static int i915_swizzle_info(struct seq_file *m, void *data)
 	if (IS_GEN3(dev) || IS_GEN4(dev)) {
 		seq_printf(m, "DDC = 0x%08x\n",
 			   I915_READ(DCC));
+		seq_printf(m, "DDC2 = 0x%08x\n",
+			   I915_READ(DCC2));
 		seq_printf(m, "C0DRB3 = 0x%04x\n",
 			   I915_READ16(C0DRB3));
 		seq_printf(m, "C1DRB3 = 0x%04x\n",
@@ -1755,7 +2013,7 @@ static int i915_swizzle_info(struct seq_file *m, void *data)
 			   I915_READ(MAD_DIMM_C2));
 		seq_printf(m, "TILECTL = 0x%08x\n",
 			   I915_READ(TILECTL));
-		if (IS_GEN8(dev))
+		if (INTEL_INFO(dev)->gen >= 8)
 			seq_printf(m, "GAMTARBMODE = 0x%08x\n",
 				   I915_READ(GAMTARBMODE));
 		else
@@ -1764,6 +2022,10 @@ static int i915_swizzle_info(struct seq_file *m, void *data)
 		seq_printf(m, "DISP_ARB_CTL = 0x%08x\n",
 			   I915_READ(DISP_ARB_CTL));
 	}
+
+	if (dev_priv->quirks & QUIRK_PIN_SWIZZLED_PAGES)
+		seq_puts(m, "L-shaped memory detected\n");
+
 	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
@@ -1774,7 +2036,13 @@ static int per_file_ctx(int id, void *ptr, void *data)
 {
 	struct intel_context *ctx = ptr;
 	struct seq_file *m = data;
-	struct i915_hw_ppgtt *ppgtt = ctx_to_ppgtt(ctx);
+	struct i915_hw_ppgtt *ppgtt = ctx->ppgtt;
+
+	if (!ppgtt) {
+		seq_printf(m, "  no ppgtt for context %d\n",
+			   ctx->user_handle);
+		return 0;
+	}
 
 	if (i915_gem_context_is_default(ctx))
 		seq_puts(m, "  default context:\n");
@@ -1834,8 +2102,7 @@ static void gen6_ppgtt_info(struct seq_file *m, struct drm_device *dev)
 		seq_printf(m, "pd gtt offset: 0x%08x\n", ppgtt->pd_offset);
 
 		ppgtt->debug_dump(ppgtt, m);
-	} else
-		return;
+	}
 
 	list_for_each_entry_reverse(file, &dev->filelist, lhead) {
 		struct drm_i915_file_private *file_priv = file->driver_priv;
@@ -2392,20 +2659,91 @@ static int i915_shared_dplls_info(struct seq_file *m, void *unused)
 		struct intel_shared_dpll *pll = &dev_priv->shared_dplls[i];
 
 		seq_printf(m, "DPLL%i: %s, id: %i\n", i, pll->name, pll->id);
-		seq_printf(m, " refcount: %i, active: %i, on: %s\n", pll->refcount,
-			   pll->active, yesno(pll->on));
+		seq_printf(m, " crtc_mask: 0x%08x, active: %d, on: %s\n",
+			   pll->config.crtc_mask, pll->active, yesno(pll->on));
 		seq_printf(m, " tracked hardware state:\n");
-		seq_printf(m, " dpll:    0x%08x\n", pll->hw_state.dpll);
-		seq_printf(m, " dpll_md: 0x%08x\n", pll->hw_state.dpll_md);
-		seq_printf(m, " fp0:     0x%08x\n", pll->hw_state.fp0);
-		seq_printf(m, " fp1:     0x%08x\n", pll->hw_state.fp1);
-		seq_printf(m, " wrpll:   0x%08x\n", pll->hw_state.wrpll);
+		seq_printf(m, " dpll:    0x%08x\n", pll->config.hw_state.dpll);
+		seq_printf(m, " dpll_md: 0x%08x\n",
+			   pll->config.hw_state.dpll_md);
+		seq_printf(m, " fp0:     0x%08x\n", pll->config.hw_state.fp0);
+		seq_printf(m, " fp1:     0x%08x\n", pll->config.hw_state.fp1);
+		seq_printf(m, " wrpll:   0x%08x\n", pll->config.hw_state.wrpll);
 	}
 	drm_modeset_unlock_all(dev);
 
 	return 0;
 }
 
+static int i915_wa_registers(struct seq_file *m, void *unused)
+{
+	int i;
+	int ret;
+	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	ret = mutex_lock_interruptible(&dev->struct_mutex);
+	if (ret)
+		return ret;
+
+	intel_runtime_pm_get(dev_priv);
+
+	seq_printf(m, "Workarounds applied: %d\n", dev_priv->workarounds.count);
+	for (i = 0; i < dev_priv->workarounds.count; ++i) {
+		u32 addr, mask, value, read;
+		bool ok;
+
+		addr = dev_priv->workarounds.reg[i].addr;
+		mask = dev_priv->workarounds.reg[i].mask;
+		value = dev_priv->workarounds.reg[i].value;
+		read = I915_READ(addr);
+		ok = (value & mask) == (read & mask);
+		seq_printf(m, "0x%X: 0x%08X, mask: 0x%08X, read: 0x%08x, status: %s\n",
+			   addr, value, mask, read, ok ? "OK" : "FAIL");
+	}
+
+	intel_runtime_pm_put(dev_priv);
+	mutex_unlock(&dev->struct_mutex);
+
+	return 0;
+}
+
+static int i915_ddb_info(struct seq_file *m, void *unused)
+{
+	struct drm_info_node *node = m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct skl_ddb_allocation *ddb;
+	struct skl_ddb_entry *entry;
+	enum pipe pipe;
+	int plane;
+
+	drm_modeset_lock_all(dev);
+
+	ddb = &dev_priv->wm.skl_hw.ddb;
+
+	seq_printf(m, "%-15s%8s%8s%8s\n", "", "Start", "End", "Size");
+
+	for_each_pipe(dev_priv, pipe) {
+		seq_printf(m, "Pipe %c\n", pipe_name(pipe));
+
+		for_each_plane(pipe, plane) {
+			entry = &ddb->plane[pipe][plane];
+			seq_printf(m, "  Plane%-8d%8u%8u%8u\n", plane + 1,
+				   entry->start, entry->end,
+				   skl_ddb_entry_size(entry));
+		}
+
+		entry = &ddb->cursor[pipe];
+		seq_printf(m, "  %-13s%8u%8u%8u\n", "Cursor", entry->start,
+			   entry->end, skl_ddb_entry_size(entry));
+	}
+
+	drm_modeset_unlock_all(dev);
+
+	return 0;
+}
+
 struct pipe_crc_info {
 	const char *name;
 	struct drm_device *dev;
@@ -2667,8 +3005,7 @@ static int i9xx_pipe_crc_auto_source(struct drm_device *dev, enum pipe pipe,
 	*source = INTEL_PIPE_CRC_SOURCE_PIPE;
 
 	drm_modeset_lock_all(dev);
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		if (!encoder->base.crtc)
 			continue;
 
@@ -2700,6 +3037,8 @@ static int i9xx_pipe_crc_auto_source(struct drm_device *dev, enum pipe pipe,
 				break;
 			}
 			break;
+		default:
+			break;
 		}
 	}
 	drm_modeset_unlock_all(dev);
@@ -2987,6 +3326,8 @@ static int pipe_crc_set_source(struct drm_device *dev, enum pipe pipe,
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[pipe];
+	struct intel_crtc *crtc = to_intel_crtc(intel_get_crtc_for_pipe(dev,
+									pipe));
 	u32 val = 0; /* shut up gcc */
 	int ret;
 
@@ -2997,6 +3338,11 @@ static int pipe_crc_set_source(struct drm_device *dev, enum pipe pipe,
 	if (pipe_crc->source && source)
 		return -EINVAL;
 
+	if (!intel_display_power_is_enabled(dev_priv, POWER_DOMAIN_PIPE(pipe))) {
+		DRM_DEBUG_KMS("Trying to capture CRC while pipe is off\n");
+		return -EIO;
+	}
+
 	if (IS_GEN2(dev))
 		ret = i8xx_pipe_crc_ctl_reg(&source, &val);
 	else if (INTEL_INFO(dev)->gen < 5)
@@ -3022,6 +3368,14 @@ static int pipe_crc_set_source(struct drm_device *dev, enum pipe pipe,
 		if (!pipe_crc->entries)
 			return -ENOMEM;
 
+		/*
+		 * When IPS gets enabled, the pipe CRC changes. Since IPS gets
+		 * enabled and disabled dynamically based on package C states,
+		 * user space can't make reliable use of the CRCs, so let's just
+		 * completely disable it.
+		 */
+		hsw_disable_ips(crtc);
+
 		spin_lock_irq(&pipe_crc->lock);
 		pipe_crc->head = 0;
 		pipe_crc->tail = 0;
@@ -3060,6 +3414,8 @@ static int pipe_crc_set_source(struct drm_device *dev, enum pipe pipe,
 			vlv_undo_pipe_scramble_reset(dev, pipe);
 		else if (IS_HASWELL(dev) && pipe == PIPE_A)
 			hsw_undo_trans_edp_pipe_A_crc_wa(dev);
+
+		hsw_enable_ips(crtc);
 	}
 
 	return 0;
@@ -3237,7 +3593,7 @@ static const struct file_operations i915_display_crc_ctl_fops = {
 	.write = display_crc_ctl_write
 };
 
-static void wm_latency_show(struct seq_file *m, const uint16_t wm[5])
+static void wm_latency_show(struct seq_file *m, const uint16_t wm[8])
 {
 	struct drm_device *dev = m->private;
 	int num_levels = ilk_wm_max_level(dev) + 1;
@@ -3248,13 +3604,17 @@ static void wm_latency_show(struct seq_file *m, const uint16_t wm[5])
 	for (level = 0; level < num_levels; level++) {
 		unsigned int latency = wm[level];
 
-		/* WM1+ latency values in 0.5us units */
-		if (level > 0)
+		/*
+		 * - WM1+ latency values in 0.5us units
+		 * - latencies are in us on gen9
+		 */
+		if (INTEL_INFO(dev)->gen >= 9)
+			latency *= 10;
+		else if (level > 0)
 			latency *= 5;
 
 		seq_printf(m, "WM%d %u (%u.%u usec)\n",
-			   level, wm[level],
-			   latency / 10, latency % 10);
+			   level, wm[level], latency / 10, latency % 10);
 	}
 
 	drm_modeset_unlock_all(dev);
@@ -3263,8 +3623,15 @@ static void wm_latency_show(struct seq_file *m, const uint16_t wm[5])
 static int pri_wm_latency_show(struct seq_file *m, void *data)
 {
 	struct drm_device *dev = m->private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	const uint16_t *latencies;
 
-	wm_latency_show(m, to_i915(dev)->wm.pri_latency);
+	if (INTEL_INFO(dev)->gen >= 9)
+		latencies = dev_priv->wm.skl_latency;
+	else
+		latencies = to_i915(dev)->wm.pri_latency;
+
+	wm_latency_show(m, latencies);
 
 	return 0;
 }
@@ -3272,8 +3639,15 @@ static int pri_wm_latency_show(struct seq_file *m, void *data)
 static int spr_wm_latency_show(struct seq_file *m, void *data)
 {
 	struct drm_device *dev = m->private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	const uint16_t *latencies;
 
-	wm_latency_show(m, to_i915(dev)->wm.spr_latency);
+	if (INTEL_INFO(dev)->gen >= 9)
+		latencies = dev_priv->wm.skl_latency;
+	else
+		latencies = to_i915(dev)->wm.spr_latency;
+
+	wm_latency_show(m, latencies);
 
 	return 0;
 }
@@ -3281,8 +3655,15 @@ static int spr_wm_latency_show(struct seq_file *m, void *data)
 static int cur_wm_latency_show(struct seq_file *m, void *data)
 {
 	struct drm_device *dev = m->private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	const uint16_t *latencies;
+
+	if (INTEL_INFO(dev)->gen >= 9)
+		latencies = dev_priv->wm.skl_latency;
+	else
+		latencies = to_i915(dev)->wm.cur_latency;
 
-	wm_latency_show(m, to_i915(dev)->wm.cur_latency);
+	wm_latency_show(m, latencies);
 
 	return 0;
 }
@@ -3318,11 +3699,11 @@ static int cur_wm_latency_open(struct inode *inode, struct file *file)
 }
 
 static ssize_t wm_latency_write(struct file *file, const char __user *ubuf,
-				size_t len, loff_t *offp, uint16_t wm[5])
+				size_t len, loff_t *offp, uint16_t wm[8])
 {
 	struct seq_file *m = file->private_data;
 	struct drm_device *dev = m->private;
-	uint16_t new[5] = { 0 };
+	uint16_t new[8] = { 0 };
 	int num_levels = ilk_wm_max_level(dev) + 1;
 	int level;
 	int ret;
@@ -3336,7 +3717,9 @@ static ssize_t wm_latency_write(struct file *file, const char __user *ubuf,
 
 	tmp[len] = '\0';
 
-	ret = sscanf(tmp, "%hu %hu %hu %hu %hu", &new[0], &new[1], &new[2], &new[3], &new[4]);
+	ret = sscanf(tmp, "%hu %hu %hu %hu %hu %hu %hu %hu",
+		     &new[0], &new[1], &new[2], &new[3],
+		     &new[4], &new[5], &new[6], &new[7]);
 	if (ret != num_levels)
 		return -EINVAL;
 
@@ -3356,8 +3739,15 @@ static ssize_t pri_wm_latency_write(struct file *file, const char __user *ubuf,
 {
 	struct seq_file *m = file->private_data;
 	struct drm_device *dev = m->private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	uint16_t *latencies;
+
+	if (INTEL_INFO(dev)->gen >= 9)
+		latencies = dev_priv->wm.skl_latency;
+	else
+		latencies = to_i915(dev)->wm.pri_latency;
 
-	return wm_latency_write(file, ubuf, len, offp, to_i915(dev)->wm.pri_latency);
+	return wm_latency_write(file, ubuf, len, offp, latencies);
 }
 
 static ssize_t spr_wm_latency_write(struct file *file, const char __user *ubuf,
@@ -3365,8 +3755,15 @@ static ssize_t spr_wm_latency_write(struct file *file, const char __user *ubuf,
 {
 	struct seq_file *m = file->private_data;
 	struct drm_device *dev = m->private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	uint16_t *latencies;
+
+	if (INTEL_INFO(dev)->gen >= 9)
+		latencies = dev_priv->wm.skl_latency;
+	else
+		latencies = to_i915(dev)->wm.spr_latency;
 
-	return wm_latency_write(file, ubuf, len, offp, to_i915(dev)->wm.spr_latency);
+	return wm_latency_write(file, ubuf, len, offp, latencies);
 }
 
 static ssize_t cur_wm_latency_write(struct file *file, const char __user *ubuf,
@@ -3374,8 +3771,15 @@ static ssize_t cur_wm_latency_write(struct file *file, const char __user *ubuf,
 {
 	struct seq_file *m = file->private_data;
 	struct drm_device *dev = m->private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	uint16_t *latencies;
+
+	if (INTEL_INFO(dev)->gen >= 9)
+		latencies = dev_priv->wm.skl_latency;
+	else
+		latencies = to_i915(dev)->wm.cur_latency;
 
-	return wm_latency_write(file, ubuf, len, offp, to_i915(dev)->wm.cur_latency);
+	return wm_latency_write(file, ubuf, len, offp, latencies);
 }
 
 static const struct file_operations i915_pri_wm_latency_fops = {
@@ -3557,9 +3961,6 @@ i915_drop_caches_set(void *data, u64 val)
 {
 	struct drm_device *dev = data;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_i915_gem_object *obj, *next;
-	struct i915_address_space *vm;
-	struct i915_vma *vma, *x;
 	int ret;
 
 	DRM_DEBUG("Dropping caches: 0x%08llx\n", val);
@@ -3579,29 +3980,11 @@ i915_drop_caches_set(void *data, u64 val)
 	if (val & (DROP_RETIRE | DROP_ACTIVE))
 		i915_gem_retire_requests(dev);
 
-	if (val & DROP_BOUND) {
-		list_for_each_entry(vm, &dev_priv->vm_list, global_link) {
-			list_for_each_entry_safe(vma, x, &vm->inactive_list,
-						 mm_list) {
-				if (vma->pin_count)
-					continue;
+	if (val & DROP_BOUND)
+		i915_gem_shrink(dev_priv, LONG_MAX, I915_SHRINK_BOUND);
 
-				ret = i915_vma_unbind(vma);
-				if (ret)
-					goto unlock;
-			}
-		}
-	}
-
-	if (val & DROP_UNBOUND) {
-		list_for_each_entry_safe(obj, next, &dev_priv->mm.unbound_list,
-					 global_list)
-			if (obj->pages_pin_count == 0) {
-				ret = i915_gem_object_put_pages(obj);
-				if (ret)
-					goto unlock;
-			}
-	}
+	if (val & DROP_UNBOUND)
+		i915_gem_shrink(dev_priv, LONG_MAX, I915_SHRINK_UNBOUND);
 
 unlock:
 	mutex_unlock(&dev->struct_mutex);
@@ -3923,6 +4306,8 @@ static const struct drm_info_list i915_debugfs_list[] = {
 	{"i915_opregion", i915_opregion, 0},
 	{"i915_gem_framebuffer", i915_gem_framebuffer_info, 0},
 	{"i915_context_status", i915_context_status, 0},
+	{"i915_dump_lrc", i915_dump_lrc, 0},
+	{"i915_execlists", i915_execlists, 0},
 	{"i915_gen6_forcewake_count", i915_gen6_forcewake_count_info, 0},
 	{"i915_swizzle_info", i915_swizzle_info, 0},
 	{"i915_ppgtt_info", i915_ppgtt_info, 0},
@@ -3936,6 +4321,8 @@ static const struct drm_info_list i915_debugfs_list[] = {
 	{"i915_semaphore_status", i915_semaphore_status, 0},
 	{"i915_shared_dplls_info", i915_shared_dplls_info, 0},
 	{"i915_dp_mst_info", i915_dp_mst_info, 0},
+	{"i915_wa_registers", i915_wa_registers, 0},
+	{"i915_ddb_info", i915_ddb_info, 0},
 };
 #define I915_DEBUGFS_ENTRIES ARRAY_SIZE(i915_debugfs_list)
 
@@ -3957,6 +4344,7 @@ static const struct i915_debugfs_files {
 	{"i915_pri_wm_latency", &i915_pri_wm_latency_fops},
 	{"i915_spr_wm_latency", &i915_spr_wm_latency_fops},
 	{"i915_cur_wm_latency", &i915_cur_wm_latency_fops},
+	{"i915_fbc_false_color", &i915_fbc_fc_fops},
 };
 
 void intel_display_crc_init(struct drm_device *dev)
@@ -3964,7 +4352,7 @@ void intel_display_crc_init(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	enum pipe pipe;
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[pipe];
 
 		pipe_crc->opened = false;
diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c
index 9933c26017ed..ecee3bcc8772 100644
--- a/drivers/gpu/drm/i915/i915_dma.c
+++ b/drivers/gpu/drm/i915/i915_dma.c
@@ -28,9 +28,11 @@
 
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 
+#include <linux/async.h>
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_fb_helper.h>
+#include <drm/drm_legacy.h>
 #include "intel_drv.h"
 #include <drm/i915_drm.h>
 #include "i915_drv.h"
@@ -48,884 +50,6 @@
 #include <linux/pm_runtime.h>
 #include <linux/oom.h>
 
-#define LP_RING(d) (&((struct drm_i915_private *)(d))->ring[RCS])
-
-#define BEGIN_LP_RING(n) \
-	intel_ring_begin(LP_RING(dev_priv), (n))
-
-#define OUT_RING(x) \
-	intel_ring_emit(LP_RING(dev_priv), x)
-
-#define ADVANCE_LP_RING() \
-	__intel_ring_advance(LP_RING(dev_priv))
-
-/**
- * Lock test for when it's just for synchronization of ring access.
- *
- * In that case, we don't need to do it when GEM is initialized as nobody else
- * has access to the ring.
- */
-#define RING_LOCK_TEST_WITH_RETURN(dev, file) do {			\
-	if (LP_RING(dev->dev_private)->buffer->obj == NULL)			\
-		LOCK_TEST_WITH_RETURN(dev, file);			\
-} while (0)
-
-static inline u32
-intel_read_legacy_status_page(struct drm_i915_private *dev_priv, int reg)
-{
-	if (I915_NEED_GFX_HWS(dev_priv->dev))
-		return ioread32(dev_priv->dri1.gfx_hws_cpu_addr + reg);
-	else
-		return intel_read_status_page(LP_RING(dev_priv), reg);
-}
-
-#define READ_HWSP(dev_priv, reg) intel_read_legacy_status_page(dev_priv, reg)
-#define READ_BREADCRUMB(dev_priv) READ_HWSP(dev_priv, I915_BREADCRUMB_INDEX)
-#define I915_BREADCRUMB_INDEX		0x21
-
-void i915_update_dri1_breadcrumb(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_i915_master_private *master_priv;
-
-	/*
-	 * The dri breadcrumb update races against the drm master disappearing.
-	 * Instead of trying to fix this (this is by far not the only ums issue)
-	 * just don't do the update in kms mode.
-	 */
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		return;
-
-	if (dev->primary->master) {
-		master_priv = dev->primary->master->driver_priv;
-		if (master_priv->sarea_priv)
-			master_priv->sarea_priv->last_dispatch =
-				READ_BREADCRUMB(dev_priv);
-	}
-}
-
-static void i915_write_hws_pga(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 addr;
-
-	addr = dev_priv->status_page_dmah->busaddr;
-	if (INTEL_INFO(dev)->gen >= 4)
-		addr |= (dev_priv->status_page_dmah->busaddr >> 28) & 0xf0;
-	I915_WRITE(HWS_PGA, addr);
-}
-
-/**
- * Frees the hardware status page, whether it's a physical address or a virtual
- * address set up by the X Server.
- */
-static void i915_free_hws(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_engine_cs *ring = LP_RING(dev_priv);
-
-	if (dev_priv->status_page_dmah) {
-		drm_pci_free(dev, dev_priv->status_page_dmah);
-		dev_priv->status_page_dmah = NULL;
-	}
-
-	if (ring->status_page.gfx_addr) {
-		ring->status_page.gfx_addr = 0;
-		iounmap(dev_priv->dri1.gfx_hws_cpu_addr);
-	}
-
-	/* Need to rewrite hardware status page */
-	I915_WRITE(HWS_PGA, 0x1ffff000);
-}
-
-void i915_kernel_lost_context(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_i915_master_private *master_priv;
-	struct intel_engine_cs *ring = LP_RING(dev_priv);
-	struct intel_ringbuffer *ringbuf = ring->buffer;
-
-	/*
-	 * We should never lose context on the ring with modesetting
-	 * as we don't expose it to userspace
-	 */
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		return;
-
-	ringbuf->head = I915_READ_HEAD(ring) & HEAD_ADDR;
-	ringbuf->tail = I915_READ_TAIL(ring) & TAIL_ADDR;
-	ringbuf->space = ringbuf->head - (ringbuf->tail + I915_RING_FREE_SPACE);
-	if (ringbuf->space < 0)
-		ringbuf->space += ringbuf->size;
-
-	if (!dev->primary->master)
-		return;
-
-	master_priv = dev->primary->master->driver_priv;
-	if (ringbuf->head == ringbuf->tail && master_priv->sarea_priv)
-		master_priv->sarea_priv->perf_boxes |= I915_BOX_RING_EMPTY;
-}
-
-static int i915_dma_cleanup(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int i;
-
-	/* Make sure interrupts are disabled here because the uninstall ioctl
-	 * may not have been called from userspace and after dev_private
-	 * is freed, it's too late.
-	 */
-	if (dev->irq_enabled)
-		drm_irq_uninstall(dev);
-
-	mutex_lock(&dev->struct_mutex);
-	for (i = 0; i < I915_NUM_RINGS; i++)
-		intel_cleanup_ring_buffer(&dev_priv->ring[i]);
-	mutex_unlock(&dev->struct_mutex);
-
-	/* Clear the HWS virtual address at teardown */
-	if (I915_NEED_GFX_HWS(dev))
-		i915_free_hws(dev);
-
-	return 0;
-}
-
-static int i915_initialize(struct drm_device *dev, drm_i915_init_t *init)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_i915_master_private *master_priv = dev->primary->master->driver_priv;
-	int ret;
-
-	master_priv->sarea = drm_getsarea(dev);
-	if (master_priv->sarea) {
-		master_priv->sarea_priv = (drm_i915_sarea_t *)
-			((u8 *)master_priv->sarea->handle + init->sarea_priv_offset);
-	} else {
-		DRM_DEBUG_DRIVER("sarea not found assuming DRI2 userspace\n");
-	}
-
-	if (init->ring_size != 0) {
-		if (LP_RING(dev_priv)->buffer->obj != NULL) {
-			i915_dma_cleanup(dev);
-			DRM_ERROR("Client tried to initialize ringbuffer in "
-				  "GEM mode\n");
-			return -EINVAL;
-		}
-
-		ret = intel_render_ring_init_dri(dev,
-						 init->ring_start,
-						 init->ring_size);
-		if (ret) {
-			i915_dma_cleanup(dev);
-			return ret;
-		}
-	}
-
-	dev_priv->dri1.cpp = init->cpp;
-	dev_priv->dri1.back_offset = init->back_offset;
-	dev_priv->dri1.front_offset = init->front_offset;
-	dev_priv->dri1.current_page = 0;
-	if (master_priv->sarea_priv)
-		master_priv->sarea_priv->pf_current_page = 0;
-
-	/* Allow hardware batchbuffers unless told otherwise.
-	 */
-	dev_priv->dri1.allow_batchbuffer = 1;
-
-	return 0;
-}
-
-static int i915_dma_resume(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_engine_cs *ring = LP_RING(dev_priv);
-
-	DRM_DEBUG_DRIVER("%s\n", __func__);
-
-	if (ring->buffer->virtual_start == NULL) {
-		DRM_ERROR("can not ioremap virtual address for"
-			  " ring buffer\n");
-		return -ENOMEM;
-	}
-
-	/* Program Hardware Status Page */
-	if (!ring->status_page.page_addr) {
-		DRM_ERROR("Can not find hardware status page\n");
-		return -EINVAL;
-	}
-	DRM_DEBUG_DRIVER("hw status page @ %p\n",
-				ring->status_page.page_addr);
-	if (ring->status_page.gfx_addr != 0)
-		intel_ring_setup_status_page(ring);
-	else
-		i915_write_hws_pga(dev);
-
-	DRM_DEBUG_DRIVER("Enabled hardware status page\n");
-
-	return 0;
-}
-
-static int i915_dma_init(struct drm_device *dev, void *data,
-			 struct drm_file *file_priv)
-{
-	drm_i915_init_t *init = data;
-	int retcode = 0;
-
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		return -ENODEV;
-
-	switch (init->func) {
-	case I915_INIT_DMA:
-		retcode = i915_initialize(dev, init);
-		break;
-	case I915_CLEANUP_DMA:
-		retcode = i915_dma_cleanup(dev);
-		break;
-	case I915_RESUME_DMA:
-		retcode = i915_dma_resume(dev);
-		break;
-	default:
-		retcode = -EINVAL;
-		break;
-	}
-
-	return retcode;
-}
-
-/* Implement basically the same security restrictions as hardware does
- * for MI_BATCH_NON_SECURE.  These can be made stricter at any time.
- *
- * Most of the calculations below involve calculating the size of a
- * particular instruction.  It's important to get the size right as
- * that tells us where the next instruction to check is.  Any illegal
- * instruction detected will be given a size of zero, which is a
- * signal to abort the rest of the buffer.
- */
-static int validate_cmd(int cmd)
-{
-	switch (((cmd >> 29) & 0x7)) {
-	case 0x0:
-		switch ((cmd >> 23) & 0x3f) {
-		case 0x0:
-			return 1;	/* MI_NOOP */
-		case 0x4:
-			return 1;	/* MI_FLUSH */
-		default:
-			return 0;	/* disallow everything else */
-		}
-		break;
-	case 0x1:
-		return 0;	/* reserved */
-	case 0x2:
-		return (cmd & 0xff) + 2;	/* 2d commands */
-	case 0x3:
-		if (((cmd >> 24) & 0x1f) <= 0x18)
-			return 1;
-
-		switch ((cmd >> 24) & 0x1f) {
-		case 0x1c:
-			return 1;
-		case 0x1d:
-			switch ((cmd >> 16) & 0xff) {
-			case 0x3:
-				return (cmd & 0x1f) + 2;
-			case 0x4:
-				return (cmd & 0xf) + 2;
-			default:
-				return (cmd & 0xffff) + 2;
-			}
-		case 0x1e:
-			if (cmd & (1 << 23))
-				return (cmd & 0xffff) + 1;
-			else
-				return 1;
-		case 0x1f:
-			if ((cmd & (1 << 23)) == 0)	/* inline vertices */
-				return (cmd & 0x1ffff) + 2;
-			else if (cmd & (1 << 17))	/* indirect random */
-				if ((cmd & 0xffff) == 0)
-					return 0;	/* unknown length, too hard */
-				else
-					return (((cmd & 0xffff) + 1) / 2) + 1;
-			else
-				return 2;	/* indirect sequential */
-		default:
-			return 0;
-		}
-	default:
-		return 0;
-	}
-
-	return 0;
-}
-
-static int i915_emit_cmds(struct drm_device *dev, int *buffer, int dwords)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int i, ret;
-
-	if ((dwords+1) * sizeof(int) >= LP_RING(dev_priv)->buffer->size - 8)
-		return -EINVAL;
-
-	for (i = 0; i < dwords;) {
-		int sz = validate_cmd(buffer[i]);
-
-		if (sz == 0 || i + sz > dwords)
-			return -EINVAL;
-		i += sz;
-	}
-
-	ret = BEGIN_LP_RING((dwords+1)&~1);
-	if (ret)
-		return ret;
-
-	for (i = 0; i < dwords; i++)
-		OUT_RING(buffer[i]);
-	if (dwords & 1)
-		OUT_RING(0);
-
-	ADVANCE_LP_RING();
-
-	return 0;
-}
-
-int
-i915_emit_box(struct drm_device *dev,
-	      struct drm_clip_rect *box,
-	      int DR1, int DR4)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int ret;
-
-	if (box->y2 <= box->y1 || box->x2 <= box->x1 ||
-	    box->y2 <= 0 || box->x2 <= 0) {
-		DRM_ERROR("Bad box %d,%d..%d,%d\n",
-			  box->x1, box->y1, box->x2, box->y2);
-		return -EINVAL;
-	}
-
-	if (INTEL_INFO(dev)->gen >= 4) {
-		ret = BEGIN_LP_RING(4);
-		if (ret)
-			return ret;
-
-		OUT_RING(GFX_OP_DRAWRECT_INFO_I965);
-		OUT_RING((box->x1 & 0xffff) | (box->y1 << 16));
-		OUT_RING(((box->x2 - 1) & 0xffff) | ((box->y2 - 1) << 16));
-		OUT_RING(DR4);
-	} else {
-		ret = BEGIN_LP_RING(6);
-		if (ret)
-			return ret;
-
-		OUT_RING(GFX_OP_DRAWRECT_INFO);
-		OUT_RING(DR1);
-		OUT_RING((box->x1 & 0xffff) | (box->y1 << 16));
-		OUT_RING(((box->x2 - 1) & 0xffff) | ((box->y2 - 1) << 16));
-		OUT_RING(DR4);
-		OUT_RING(0);
-	}
-	ADVANCE_LP_RING();
-
-	return 0;
-}
-
-/* XXX: Emitting the counter should really be moved to part of the IRQ
- * emit. For now, do it in both places:
- */
-
-static void i915_emit_breadcrumb(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_i915_master_private *master_priv = dev->primary->master->driver_priv;
-
-	dev_priv->dri1.counter++;
-	if (dev_priv->dri1.counter > 0x7FFFFFFFUL)
-		dev_priv->dri1.counter = 0;
-	if (master_priv->sarea_priv)
-		master_priv->sarea_priv->last_enqueue = dev_priv->dri1.counter;
-
-	if (BEGIN_LP_RING(4) == 0) {
-		OUT_RING(MI_STORE_DWORD_INDEX);
-		OUT_RING(I915_BREADCRUMB_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
-		OUT_RING(dev_priv->dri1.counter);
-		OUT_RING(0);
-		ADVANCE_LP_RING();
-	}
-}
-
-static int i915_dispatch_cmdbuffer(struct drm_device *dev,
-				   drm_i915_cmdbuffer_t *cmd,
-				   struct drm_clip_rect *cliprects,
-				   void *cmdbuf)
-{
-	int nbox = cmd->num_cliprects;
-	int i = 0, count, ret;
-
-	if (cmd->sz & 0x3) {
-		DRM_ERROR("alignment");
-		return -EINVAL;
-	}
-
-	i915_kernel_lost_context(dev);
-
-	count = nbox ? nbox : 1;
-
-	for (i = 0; i < count; i++) {
-		if (i < nbox) {
-			ret = i915_emit_box(dev, &cliprects[i],
-					    cmd->DR1, cmd->DR4);
-			if (ret)
-				return ret;
-		}
-
-		ret = i915_emit_cmds(dev, cmdbuf, cmd->sz / 4);
-		if (ret)
-			return ret;
-	}
-
-	i915_emit_breadcrumb(dev);
-	return 0;
-}
-
-static int i915_dispatch_batchbuffer(struct drm_device *dev,
-				     drm_i915_batchbuffer_t *batch,
-				     struct drm_clip_rect *cliprects)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int nbox = batch->num_cliprects;
-	int i, count, ret;
-
-	if ((batch->start | batch->used) & 0x7) {
-		DRM_ERROR("alignment");
-		return -EINVAL;
-	}
-
-	i915_kernel_lost_context(dev);
-
-	count = nbox ? nbox : 1;
-	for (i = 0; i < count; i++) {
-		if (i < nbox) {
-			ret = i915_emit_box(dev, &cliprects[i],
-					    batch->DR1, batch->DR4);
-			if (ret)
-				return ret;
-		}
-
-		if (!IS_I830(dev) && !IS_845G(dev)) {
-			ret = BEGIN_LP_RING(2);
-			if (ret)
-				return ret;
-
-			if (INTEL_INFO(dev)->gen >= 4) {
-				OUT_RING(MI_BATCH_BUFFER_START | (2 << 6) | MI_BATCH_NON_SECURE_I965);
-				OUT_RING(batch->start);
-			} else {
-				OUT_RING(MI_BATCH_BUFFER_START | (2 << 6));
-				OUT_RING(batch->start | MI_BATCH_NON_SECURE);
-			}
-		} else {
-			ret = BEGIN_LP_RING(4);
-			if (ret)
-				return ret;
-
-			OUT_RING(MI_BATCH_BUFFER);
-			OUT_RING(batch->start | MI_BATCH_NON_SECURE);
-			OUT_RING(batch->start + batch->used - 4);
-			OUT_RING(0);
-		}
-		ADVANCE_LP_RING();
-	}
-
-
-	if (IS_G4X(dev) || IS_GEN5(dev)) {
-		if (BEGIN_LP_RING(2) == 0) {
-			OUT_RING(MI_FLUSH | MI_NO_WRITE_FLUSH | MI_INVALIDATE_ISP);
-			OUT_RING(MI_NOOP);
-			ADVANCE_LP_RING();
-		}
-	}
-
-	i915_emit_breadcrumb(dev);
-	return 0;
-}
-
-static int i915_dispatch_flip(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_i915_master_private *master_priv =
-		dev->primary->master->driver_priv;
-	int ret;
-
-	if (!master_priv->sarea_priv)
-		return -EINVAL;
-
-	DRM_DEBUG_DRIVER("%s: page=%d pfCurrentPage=%d\n",
-			  __func__,
-			 dev_priv->dri1.current_page,
-			 master_priv->sarea_priv->pf_current_page);
-
-	i915_kernel_lost_context(dev);
-
-	ret = BEGIN_LP_RING(10);
-	if (ret)
-		return ret;
-
-	OUT_RING(MI_FLUSH | MI_READ_FLUSH);
-	OUT_RING(0);
-
-	OUT_RING(CMD_OP_DISPLAYBUFFER_INFO | ASYNC_FLIP);
-	OUT_RING(0);
-	if (dev_priv->dri1.current_page == 0) {
-		OUT_RING(dev_priv->dri1.back_offset);
-		dev_priv->dri1.current_page = 1;
-	} else {
-		OUT_RING(dev_priv->dri1.front_offset);
-		dev_priv->dri1.current_page = 0;
-	}
-	OUT_RING(0);
-
-	OUT_RING(MI_WAIT_FOR_EVENT | MI_WAIT_FOR_PLANE_A_FLIP);
-	OUT_RING(0);
-
-	ADVANCE_LP_RING();
-
-	master_priv->sarea_priv->last_enqueue = dev_priv->dri1.counter++;
-
-	if (BEGIN_LP_RING(4) == 0) {
-		OUT_RING(MI_STORE_DWORD_INDEX);
-		OUT_RING(I915_BREADCRUMB_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
-		OUT_RING(dev_priv->dri1.counter);
-		OUT_RING(0);
-		ADVANCE_LP_RING();
-	}
-
-	master_priv->sarea_priv->pf_current_page = dev_priv->dri1.current_page;
-	return 0;
-}
-
-static int i915_quiescent(struct drm_device *dev)
-{
-	i915_kernel_lost_context(dev);
-	return intel_ring_idle(LP_RING(dev->dev_private));
-}
-
-static int i915_flush_ioctl(struct drm_device *dev, void *data,
-			    struct drm_file *file_priv)
-{
-	int ret;
-
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		return -ENODEV;
-
-	RING_LOCK_TEST_WITH_RETURN(dev, file_priv);
-
-	mutex_lock(&dev->struct_mutex);
-	ret = i915_quiescent(dev);
-	mutex_unlock(&dev->struct_mutex);
-
-	return ret;
-}
-
-static int i915_batchbuffer(struct drm_device *dev, void *data,
-			    struct drm_file *file_priv)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_i915_master_private *master_priv;
-	drm_i915_sarea_t *sarea_priv;
-	drm_i915_batchbuffer_t *batch = data;
-	int ret;
-	struct drm_clip_rect *cliprects = NULL;
-
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		return -ENODEV;
-
-	master_priv = dev->primary->master->driver_priv;
-	sarea_priv = (drm_i915_sarea_t *) master_priv->sarea_priv;
-
-	if (!dev_priv->dri1.allow_batchbuffer) {
-		DRM_ERROR("Batchbuffer ioctl disabled\n");
-		return -EINVAL;
-	}
-
-	DRM_DEBUG_DRIVER("i915 batchbuffer, start %x used %d cliprects %d\n",
-			batch->start, batch->used, batch->num_cliprects);
-
-	RING_LOCK_TEST_WITH_RETURN(dev, file_priv);
-
-	if (batch->num_cliprects < 0)
-		return -EINVAL;
-
-	if (batch->num_cliprects) {
-		cliprects = kcalloc(batch->num_cliprects,
-				    sizeof(*cliprects),
-				    GFP_KERNEL);
-		if (cliprects == NULL)
-			return -ENOMEM;
-
-		ret = copy_from_user(cliprects, batch->cliprects,
-				     batch->num_cliprects *
-				     sizeof(struct drm_clip_rect));
-		if (ret != 0) {
-			ret = -EFAULT;
-			goto fail_free;
-		}
-	}
-
-	mutex_lock(&dev->struct_mutex);
-	ret = i915_dispatch_batchbuffer(dev, batch, cliprects);
-	mutex_unlock(&dev->struct_mutex);
-
-	if (sarea_priv)
-		sarea_priv->last_dispatch = READ_BREADCRUMB(dev_priv);
-
-fail_free:
-	kfree(cliprects);
-
-	return ret;
-}
-
-static int i915_cmdbuffer(struct drm_device *dev, void *data,
-			  struct drm_file *file_priv)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_i915_master_private *master_priv;
-	drm_i915_sarea_t *sarea_priv;
-	drm_i915_cmdbuffer_t *cmdbuf = data;
-	struct drm_clip_rect *cliprects = NULL;
-	void *batch_data;
-	int ret;
-
-	DRM_DEBUG_DRIVER("i915 cmdbuffer, buf %p sz %d cliprects %d\n",
-			cmdbuf->buf, cmdbuf->sz, cmdbuf->num_cliprects);
-
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		return -ENODEV;
-
-	master_priv = dev->primary->master->driver_priv;
-	sarea_priv = (drm_i915_sarea_t *) master_priv->sarea_priv;
-
-	RING_LOCK_TEST_WITH_RETURN(dev, file_priv);
-
-	if (cmdbuf->num_cliprects < 0)
-		return -EINVAL;
-
-	batch_data = kmalloc(cmdbuf->sz, GFP_KERNEL);
-	if (batch_data == NULL)
-		return -ENOMEM;
-
-	ret = copy_from_user(batch_data, cmdbuf->buf, cmdbuf->sz);
-	if (ret != 0) {
-		ret = -EFAULT;
-		goto fail_batch_free;
-	}
-
-	if (cmdbuf->num_cliprects) {
-		cliprects = kcalloc(cmdbuf->num_cliprects,
-				    sizeof(*cliprects), GFP_KERNEL);
-		if (cliprects == NULL) {
-			ret = -ENOMEM;
-			goto fail_batch_free;
-		}
-
-		ret = copy_from_user(cliprects, cmdbuf->cliprects,
-				     cmdbuf->num_cliprects *
-				     sizeof(struct drm_clip_rect));
-		if (ret != 0) {
-			ret = -EFAULT;
-			goto fail_clip_free;
-		}
-	}
-
-	mutex_lock(&dev->struct_mutex);
-	ret = i915_dispatch_cmdbuffer(dev, cmdbuf, cliprects, batch_data);
-	mutex_unlock(&dev->struct_mutex);
-	if (ret) {
-		DRM_ERROR("i915_dispatch_cmdbuffer failed\n");
-		goto fail_clip_free;
-	}
-
-	if (sarea_priv)
-		sarea_priv->last_dispatch = READ_BREADCRUMB(dev_priv);
-
-fail_clip_free:
-	kfree(cliprects);
-fail_batch_free:
-	kfree(batch_data);
-
-	return ret;
-}
-
-static int i915_emit_irq(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_i915_master_private *master_priv = dev->primary->master->driver_priv;
-
-	i915_kernel_lost_context(dev);
-
-	DRM_DEBUG_DRIVER("\n");
-
-	dev_priv->dri1.counter++;
-	if (dev_priv->dri1.counter > 0x7FFFFFFFUL)
-		dev_priv->dri1.counter = 1;
-	if (master_priv->sarea_priv)
-		master_priv->sarea_priv->last_enqueue = dev_priv->dri1.counter;
-
-	if (BEGIN_LP_RING(4) == 0) {
-		OUT_RING(MI_STORE_DWORD_INDEX);
-		OUT_RING(I915_BREADCRUMB_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
-		OUT_RING(dev_priv->dri1.counter);
-		OUT_RING(MI_USER_INTERRUPT);
-		ADVANCE_LP_RING();
-	}
-
-	return dev_priv->dri1.counter;
-}
-
-static int i915_wait_irq(struct drm_device *dev, int irq_nr)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_i915_master_private *master_priv = dev->primary->master->driver_priv;
-	int ret = 0;
-	struct intel_engine_cs *ring = LP_RING(dev_priv);
-
-	DRM_DEBUG_DRIVER("irq_nr=%d breadcrumb=%d\n", irq_nr,
-		  READ_BREADCRUMB(dev_priv));
-
-	if (READ_BREADCRUMB(dev_priv) >= irq_nr) {
-		if (master_priv->sarea_priv)
-			master_priv->sarea_priv->last_dispatch = READ_BREADCRUMB(dev_priv);
-		return 0;
-	}
-
-	if (master_priv->sarea_priv)
-		master_priv->sarea_priv->perf_boxes |= I915_BOX_WAIT;
-
-	if (ring->irq_get(ring)) {
-		DRM_WAIT_ON(ret, ring->irq_queue, 3 * HZ,
-			    READ_BREADCRUMB(dev_priv) >= irq_nr);
-		ring->irq_put(ring);
-	} else if (wait_for(READ_BREADCRUMB(dev_priv) >= irq_nr, 3000))
-		ret = -EBUSY;
-
-	if (ret == -EBUSY) {
-		DRM_ERROR("EBUSY -- rec: %d emitted: %d\n",
-			  READ_BREADCRUMB(dev_priv), (int)dev_priv->dri1.counter);
-	}
-
-	return ret;
-}
-
-/* Needs the lock as it touches the ring.
- */
-static int i915_irq_emit(struct drm_device *dev, void *data,
-			 struct drm_file *file_priv)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	drm_i915_irq_emit_t *emit = data;
-	int result;
-
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		return -ENODEV;
-
-	if (!dev_priv || !LP_RING(dev_priv)->buffer->virtual_start) {
-		DRM_ERROR("called with no initialization\n");
-		return -EINVAL;
-	}
-
-	RING_LOCK_TEST_WITH_RETURN(dev, file_priv);
-
-	mutex_lock(&dev->struct_mutex);
-	result = i915_emit_irq(dev);
-	mutex_unlock(&dev->struct_mutex);
-
-	if (copy_to_user(emit->irq_seq, &result, sizeof(int))) {
-		DRM_ERROR("copy_to_user\n");
-		return -EFAULT;
-	}
-
-	return 0;
-}
-
-/* Doesn't need the hardware lock.
- */
-static int i915_irq_wait(struct drm_device *dev, void *data,
-			 struct drm_file *file_priv)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	drm_i915_irq_wait_t *irqwait = data;
-
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		return -ENODEV;
-
-	if (!dev_priv) {
-		DRM_ERROR("called with no initialization\n");
-		return -EINVAL;
-	}
-
-	return i915_wait_irq(dev, irqwait->irq_seq);
-}
-
-static int i915_vblank_pipe_get(struct drm_device *dev, void *data,
-			 struct drm_file *file_priv)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	drm_i915_vblank_pipe_t *pipe = data;
-
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		return -ENODEV;
-
-	if (!dev_priv) {
-		DRM_ERROR("called with no initialization\n");
-		return -EINVAL;
-	}
-
-	pipe->pipe = DRM_I915_VBLANK_PIPE_A | DRM_I915_VBLANK_PIPE_B;
-
-	return 0;
-}
-
-/**
- * Schedule buffer swap at given vertical blank.
- */
-static int i915_vblank_swap(struct drm_device *dev, void *data,
-		     struct drm_file *file_priv)
-{
-	/* The delayed swap mechanism was fundamentally racy, and has been
-	 * removed.  The model was that the client requested a delayed flip/swap
-	 * from the kernel, then waited for vblank before continuing to perform
-	 * rendering.  The problem was that the kernel might wake the client
-	 * up before it dispatched the vblank swap (since the lock has to be
-	 * held while touching the ringbuffer), in which case the client would
-	 * clear and start the next frame before the swap occurred, and
-	 * flicker would occur in addition to likely missing the vblank.
-	 *
-	 * In the absence of this ioctl, userland falls back to a correct path
-	 * of waiting for a vblank, then dispatching the swap on its own.
-	 * Context switching to userland and back is plenty fast enough for
-	 * meeting the requirements of vblank swapping.
-	 */
-	return -EINVAL;
-}
-
-static int i915_flip_bufs(struct drm_device *dev, void *data,
-			  struct drm_file *file_priv)
-{
-	int ret;
-
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		return -ENODEV;
-
-	DRM_DEBUG_DRIVER("%s\n", __func__);
-
-	RING_LOCK_TEST_WITH_RETURN(dev, file_priv);
-
-	mutex_lock(&dev->struct_mutex);
-	ret = i915_dispatch_flip(dev);
-	mutex_unlock(&dev->struct_mutex);
-
-	return ret;
-}
 
 static int i915_getparam(struct drm_device *dev, void *data,
 			 struct drm_file *file_priv)
@@ -934,21 +58,12 @@ static int i915_getparam(struct drm_device *dev, void *data,
 	drm_i915_getparam_t *param = data;
 	int value;
 
-	if (!dev_priv) {
-		DRM_ERROR("called with no initialization\n");
-		return -EINVAL;
-	}
-
 	switch (param->param) {
 	case I915_PARAM_IRQ_ACTIVE:
-		value = dev->pdev->irq ? 1 : 0;
-		break;
 	case I915_PARAM_ALLOW_BATCHBUFFER:
-		value = dev_priv->dri1.allow_batchbuffer ? 1 : 0;
-		break;
 	case I915_PARAM_LAST_DISPATCH:
-		value = READ_BREADCRUMB(dev_priv);
-		break;
+		/* Reject all old ums/dri params. */
+		return -ENODEV;
 	case I915_PARAM_CHIPSET_ID:
 		value = dev->pdev->device;
 		break;
@@ -999,7 +114,7 @@ static int i915_getparam(struct drm_device *dev, void *data,
 		value = HAS_WT(dev);
 		break;
 	case I915_PARAM_HAS_ALIASING_PPGTT:
-		value = dev_priv->mm.aliasing_ppgtt || USES_FULL_PPGTT(dev);
+		value = USES_PPGTT(dev);
 		break;
 	case I915_PARAM_HAS_WAIT_TIMEOUT:
 		value = 1;
@@ -1025,6 +140,9 @@ static int i915_getparam(struct drm_device *dev, void *data,
 	case I915_PARAM_CMD_PARSER_VERSION:
 		value = i915_cmd_parser_get_version();
 		break;
+	case I915_PARAM_HAS_COHERENT_PHYS_GTT:
+		value = 1;
+		break;
 	default:
 		DRM_DEBUG("Unknown parameter %d\n", param->param);
 		return -EINVAL;
@@ -1044,19 +162,13 @@ static int i915_setparam(struct drm_device *dev, void *data,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	drm_i915_setparam_t *param = data;
 
-	if (!dev_priv) {
-		DRM_ERROR("called with no initialization\n");
-		return -EINVAL;
-	}
-
 	switch (param->param) {
 	case I915_SETPARAM_USE_MI_BATCHBUFFER_START:
-		break;
 	case I915_SETPARAM_TEX_LRU_LOG_GRANULARITY:
-		break;
 	case I915_SETPARAM_ALLOW_BATCHBUFFER:
-		dev_priv->dri1.allow_batchbuffer = param->value ? 1 : 0;
-		break;
+		/* Reject all old ums/dri params. */
+		return -ENODEV;
+
 	case I915_SETPARAM_NUM_USED_FENCES:
 		if (param->value > dev_priv->num_fence_regs ||
 		    param->value < 0)
@@ -1073,54 +185,6 @@ static int i915_setparam(struct drm_device *dev, void *data,
 	return 0;
 }
 
-static int i915_set_status_page(struct drm_device *dev, void *data,
-				struct drm_file *file_priv)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	drm_i915_hws_addr_t *hws = data;
-	struct intel_engine_cs *ring;
-
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		return -ENODEV;
-
-	if (!I915_NEED_GFX_HWS(dev))
-		return -EINVAL;
-
-	if (!dev_priv) {
-		DRM_ERROR("called with no initialization\n");
-		return -EINVAL;
-	}
-
-	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
-		WARN(1, "tried to set status page when mode setting active\n");
-		return 0;
-	}
-
-	DRM_DEBUG_DRIVER("set status page addr 0x%08x\n", (u32)hws->addr);
-
-	ring = LP_RING(dev_priv);
-	ring->status_page.gfx_addr = hws->addr & (0x1ffff<<12);
-
-	dev_priv->dri1.gfx_hws_cpu_addr =
-		ioremap_wc(dev_priv->gtt.mappable_base + hws->addr, 4096);
-	if (dev_priv->dri1.gfx_hws_cpu_addr == NULL) {
-		i915_dma_cleanup(dev);
-		ring->status_page.gfx_addr = 0;
-		DRM_ERROR("can not ioremap virtual address for"
-				" G33 hw status page\n");
-		return -ENOMEM;
-	}
-
-	memset_io(dev_priv->dri1.gfx_hws_cpu_addr, 0, PAGE_SIZE);
-	I915_WRITE(HWS_PGA, ring->status_page.gfx_addr);
-
-	DRM_DEBUG_DRIVER("load hws HWS_PGA with gfx mem 0x%x\n",
-			 ring->status_page.gfx_addr);
-	DRM_DEBUG_DRIVER("load hws at %p\n",
-			 ring->status_page.page_addr);
-	return 0;
-}
-
 static int i915_get_bridge_dev(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1273,12 +337,12 @@ static void i915_switcheroo_set_state(struct pci_dev *pdev, enum vga_switcheroo_
 		dev->switch_power_state = DRM_SWITCH_POWER_CHANGING;
 		/* i915 resume handler doesn't set to D0 */
 		pci_set_power_state(dev->pdev, PCI_D0);
-		i915_resume(dev);
+		i915_resume_legacy(dev);
 		dev->switch_power_state = DRM_SWITCH_POWER_ON;
 	} else {
 		pr_err("switched off\n");
 		dev->switch_power_state = DRM_SWITCH_POWER_CHANGING;
-		i915_suspend(dev, pmm);
+		i915_suspend_legacy(dev, pmm);
 		dev->switch_power_state = DRM_SWITCH_POWER_OFF;
 	}
 }
@@ -1336,14 +400,7 @@ static int i915_load_modeset_init(struct drm_device *dev)
 
 	intel_power_domains_init_hw(dev_priv);
 
-	/*
-	 * We enable some interrupt sources in our postinstall hooks, so mark
-	 * interrupts as enabled _before_ actually enabling them to avoid
-	 * special cases in our ordering checks.
-	 */
-	dev_priv->pm._irqs_disabled = false;
-
-	ret = drm_irq_install(dev, dev->pdev->irq);
+	ret = intel_irq_install(dev_priv);
 	if (ret)
 		goto cleanup_gem_stolen;
 
@@ -1355,8 +412,6 @@ static int i915_load_modeset_init(struct drm_device *dev)
 	if (ret)
 		goto cleanup_irq;
 
-	INIT_WORK(&dev_priv->console_resume_work, intel_console_resume);
-
 	intel_modeset_gem_init(dev);
 
 	/* Always safe in the mode setting case. */
@@ -1370,7 +425,7 @@ static int i915_load_modeset_init(struct drm_device *dev)
 		goto cleanup_gem;
 
 	/* Only enable hotplug handling once the fbdev is fully set up. */
-	intel_hpd_init(dev);
+	intel_hpd_init(dev_priv);
 
 	/*
 	 * Some ports require correctly set-up hpd registers for detection to
@@ -1382,7 +437,7 @@ static int i915_load_modeset_init(struct drm_device *dev)
 	 * scanning against hotplug events. Hence do this first and ignore the
 	 * tiny window where we will loose hotplug notifactions.
 	 */
-	intel_fbdev_initial_config(dev);
+	async_schedule(intel_fbdev_initial_config, dev_priv);
 
 	drm_kms_helper_poll_init(dev);
 
@@ -1393,7 +448,6 @@ cleanup_gem:
 	i915_gem_cleanup_ringbuffer(dev);
 	i915_gem_context_fini(dev);
 	mutex_unlock(&dev->struct_mutex);
-	WARN_ON(dev_priv->mm.aliasing_ppgtt);
 cleanup_irq:
 	drm_irq_uninstall(dev);
 cleanup_gem_stolen:
@@ -1406,30 +460,6 @@ out:
 	return ret;
 }
 
-int i915_master_create(struct drm_device *dev, struct drm_master *master)
-{
-	struct drm_i915_master_private *master_priv;
-
-	master_priv = kzalloc(sizeof(*master_priv), GFP_KERNEL);
-	if (!master_priv)
-		return -ENOMEM;
-
-	master->driver_priv = master_priv;
-	return 0;
-}
-
-void i915_master_destroy(struct drm_device *dev, struct drm_master *master)
-{
-	struct drm_i915_master_private *master_priv = master->driver_priv;
-
-	if (!master_priv)
-		return;
-
-	kfree(master_priv);
-
-	master->driver_priv = NULL;
-}
-
 #if IS_ENABLED(CONFIG_FB)
 static int i915_kick_out_firmware_fb(struct drm_i915_private *dev_priv)
 {
@@ -1535,11 +565,11 @@ static void intel_device_info_runtime_init(struct drm_device *dev)
 
 	info = (struct intel_device_info *)&dev_priv->info;
 
-	if (IS_VALLEYVIEW(dev))
-		for_each_pipe(pipe)
+	if (IS_VALLEYVIEW(dev) || INTEL_INFO(dev)->gen == 9)
+		for_each_pipe(dev_priv, pipe)
 			info->num_sprites[pipe] = 2;
 	else
-		for_each_pipe(pipe)
+		for_each_pipe(dev_priv, pipe)
 			info->num_sprites[pipe] = 1;
 
 	if (i915.disable_display) {
@@ -1608,13 +638,14 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 	dev->dev_private = dev_priv;
 	dev_priv->dev = dev;
 
-	/* copy initial configuration to dev_priv->info */
+	/* Setup the write-once "constant" device info */
 	device_info = (struct intel_device_info *)&dev_priv->info;
-	*device_info = *info;
+	memcpy(device_info, info, sizeof(dev_priv->info));
+	device_info->device_id = dev->pdev->device;
 
 	spin_lock_init(&dev_priv->irq_lock);
 	spin_lock_init(&dev_priv->gpu_error.lock);
-	spin_lock_init(&dev_priv->backlight_lock);
+	mutex_init(&dev_priv->backlight_lock);
 	spin_lock_init(&dev_priv->uncore.lock);
 	spin_lock_init(&dev_priv->mm.object_stat_lock);
 	spin_lock_init(&dev_priv->mmio_flip_lock);
@@ -1670,15 +701,17 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 		goto out_regs;
 
 	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
-		ret = i915_kick_out_vgacon(dev_priv);
+		/* WARNING: Apparently we must kick fbdev drivers before vgacon,
+		 * otherwise the vga fbdev driver falls over. */
+		ret = i915_kick_out_firmware_fb(dev_priv);
 		if (ret) {
-			DRM_ERROR("failed to remove conflicting VGA console\n");
+			DRM_ERROR("failed to remove conflicting framebuffer drivers\n");
 			goto out_gtt;
 		}
 
-		ret = i915_kick_out_firmware_fb(dev_priv);
+		ret = i915_kick_out_vgacon(dev_priv);
 		if (ret) {
-			DRM_ERROR("failed to remove conflicting framebuffer drivers\n");
+			DRM_ERROR("failed to remove conflicting VGA console\n");
 			goto out_gtt;
 		}
 	}
@@ -1740,7 +773,7 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 		goto out_freewq;
 	}
 
-	intel_irq_init(dev);
+	intel_irq_init(dev_priv);
 	intel_uncore_sanitize(dev);
 
 	/* Try to make sure MCHBAR is enabled before poking at it */
@@ -1782,9 +815,6 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 			DRM_ERROR("failed to init modeset\n");
 			goto out_power_well;
 		}
-	} else {
-		/* Start out suspended in ums mode. */
-		dev_priv->ums.mm_suspended = 1;
 	}
 
 	i915_setup_sysfs(dev);
@@ -1798,12 +828,12 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 	if (IS_GEN5(dev))
 		intel_gpu_ips_init(dev_priv);
 
-	intel_init_runtime_pm(dev_priv);
+	intel_runtime_pm_enable(dev_priv);
 
 	return 0;
 
 out_power_well:
-	intel_power_domains_remove(dev_priv);
+	intel_power_domains_fini(dev_priv);
 	drm_vblank_cleanup(dev);
 out_gem_unload:
 	WARN_ON(unregister_oom_notifier(&dev_priv->mm.oom_notifier));
@@ -1822,7 +852,7 @@ out_mtrrfree:
 	arch_phys_wc_del(dev_priv->gtt.mtrr);
 	io_mapping_free(dev_priv->gtt.mappable);
 out_gtt:
-	dev_priv->gtt.base.cleanup(&dev_priv->gtt.base);
+	i915_global_gtt_cleanup(dev);
 out_regs:
 	intel_uncore_fini(dev);
 	pci_iounmap(dev->pdev, dev_priv->regs);
@@ -1846,16 +876,10 @@ int i915_driver_unload(struct drm_device *dev)
 		return ret;
 	}
 
-	intel_fini_runtime_pm(dev_priv);
+	intel_power_domains_fini(dev_priv);
 
 	intel_gpu_ips_teardown();
 
-	/* The i915.ko module is still not prepared to be loaded when
-	 * the power well is not enabled, so just enable it in case
-	 * we're going to unload/reload. */
-	intel_display_set_init_power(dev_priv, true);
-	intel_power_domains_remove(dev_priv);
-
 	i915_teardown_sysfs(dev);
 
 	WARN_ON(unregister_oom_notifier(&dev_priv->mm.oom_notifier));
@@ -1866,10 +890,13 @@ int i915_driver_unload(struct drm_device *dev)
 
 	acpi_video_unregister();
 
-	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
+	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		intel_fbdev_fini(dev);
+
+	drm_vblank_cleanup(dev);
+
+	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
 		intel_modeset_cleanup(dev);
-		cancel_work_sync(&dev_priv->console_resume_work);
 
 		/*
 		 * free the memory space allocated for the child device
@@ -1902,18 +929,10 @@ int i915_driver_unload(struct drm_device *dev)
 		mutex_lock(&dev->struct_mutex);
 		i915_gem_cleanup_ringbuffer(dev);
 		i915_gem_context_fini(dev);
-		WARN_ON(dev_priv->mm.aliasing_ppgtt);
 		mutex_unlock(&dev->struct_mutex);
 		i915_gem_cleanup_stolen(dev);
-
-		if (!I915_NEED_GFX_HWS(dev))
-			i915_free_hws(dev);
 	}
 
-	WARN_ON(!list_empty(&dev_priv->vm_list));
-
-	drm_vblank_cleanup(dev);
-
 	intel_teardown_gmbus(dev);
 	intel_teardown_mchbar(dev);
 
@@ -1921,7 +940,7 @@ int i915_driver_unload(struct drm_device *dev)
 	destroy_workqueue(dev_priv->wq);
 	pm_qos_remove_request(&dev_priv->pm_qos);
 
-	dev_priv->gtt.base.cleanup(&dev_priv->gtt.base);
+	i915_global_gtt_cleanup(dev);
 
 	intel_uncore_fini(dev);
 	if (dev_priv->regs != NULL)
@@ -1961,23 +980,8 @@ int i915_driver_open(struct drm_device *dev, struct drm_file *file)
  */
 void i915_driver_lastclose(struct drm_device *dev)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	/* On gen6+ we refuse to init without kms enabled, but then the drm core
-	 * goes right around and calls lastclose. Check for this and don't clean
-	 * up anything. */
-	if (!dev_priv)
-		return;
-
-	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
-		intel_fbdev_restore_mode(dev);
-		vga_switcheroo_process_delayed_switch();
-		return;
-	}
-
-	i915_gem_lastclose(dev);
-
-	i915_dma_cleanup(dev);
+	intel_fbdev_restore_mode(dev);
+	vga_switcheroo_process_delayed_switch();
 }
 
 void i915_driver_preclose(struct drm_device *dev, struct drm_file *file)
@@ -1986,6 +990,9 @@ void i915_driver_preclose(struct drm_device *dev, struct drm_file *file)
 	i915_gem_context_close(dev, file);
 	i915_gem_release(dev, file);
 	mutex_unlock(&dev->struct_mutex);
+
+	if (drm_core_check_feature(dev, DRIVER_MODESET))
+		intel_modeset_preclose(dev, file);
 }
 
 void i915_driver_postclose(struct drm_device *dev, struct drm_file *file)
@@ -1998,24 +1005,24 @@ void i915_driver_postclose(struct drm_device *dev, struct drm_file *file)
 }
 
 const struct drm_ioctl_desc i915_ioctls[] = {
-	DRM_IOCTL_DEF_DRV(I915_INIT, i915_dma_init, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF_DRV(I915_FLUSH, i915_flush_ioctl, DRM_AUTH),
-	DRM_IOCTL_DEF_DRV(I915_FLIP, i915_flip_bufs, DRM_AUTH),
-	DRM_IOCTL_DEF_DRV(I915_BATCHBUFFER, i915_batchbuffer, DRM_AUTH),
-	DRM_IOCTL_DEF_DRV(I915_IRQ_EMIT, i915_irq_emit, DRM_AUTH),
-	DRM_IOCTL_DEF_DRV(I915_IRQ_WAIT, i915_irq_wait, DRM_AUTH),
+	DRM_IOCTL_DEF_DRV(I915_INIT, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF_DRV(I915_FLUSH, drm_noop, DRM_AUTH),
+	DRM_IOCTL_DEF_DRV(I915_FLIP, drm_noop, DRM_AUTH),
+	DRM_IOCTL_DEF_DRV(I915_BATCHBUFFER, drm_noop, DRM_AUTH),
+	DRM_IOCTL_DEF_DRV(I915_IRQ_EMIT, drm_noop, DRM_AUTH),
+	DRM_IOCTL_DEF_DRV(I915_IRQ_WAIT, drm_noop, DRM_AUTH),
 	DRM_IOCTL_DEF_DRV(I915_GETPARAM, i915_getparam, DRM_AUTH|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_SETPARAM, i915_setparam, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
 	DRM_IOCTL_DEF_DRV(I915_ALLOC, drm_noop, DRM_AUTH),
 	DRM_IOCTL_DEF_DRV(I915_FREE, drm_noop, DRM_AUTH),
 	DRM_IOCTL_DEF_DRV(I915_INIT_HEAP, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF_DRV(I915_CMDBUFFER, i915_cmdbuffer, DRM_AUTH),
+	DRM_IOCTL_DEF_DRV(I915_CMDBUFFER, drm_noop, DRM_AUTH),
 	DRM_IOCTL_DEF_DRV(I915_DESTROY_HEAP,  drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
 	DRM_IOCTL_DEF_DRV(I915_SET_VBLANK_PIPE,  drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF_DRV(I915_GET_VBLANK_PIPE,  i915_vblank_pipe_get, DRM_AUTH),
-	DRM_IOCTL_DEF_DRV(I915_VBLANK_SWAP, i915_vblank_swap, DRM_AUTH),
-	DRM_IOCTL_DEF_DRV(I915_HWS_ADDR, i915_set_status_page, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF_DRV(I915_GEM_INIT, i915_gem_init_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY|DRM_UNLOCKED),
+	DRM_IOCTL_DEF_DRV(I915_GET_VBLANK_PIPE,  drm_noop, DRM_AUTH),
+	DRM_IOCTL_DEF_DRV(I915_VBLANK_SWAP, drm_noop, DRM_AUTH),
+	DRM_IOCTL_DEF_DRV(I915_HWS_ADDR, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF_DRV(I915_GEM_INIT, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY|DRM_UNLOCKED),
 	DRM_IOCTL_DEF_DRV(I915_GEM_EXECBUFFER, i915_gem_execbuffer, DRM_AUTH|DRM_UNLOCKED),
 	DRM_IOCTL_DEF_DRV(I915_GEM_EXECBUFFER2, i915_gem_execbuffer2, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_GEM_PIN, i915_gem_pin_ioctl, DRM_AUTH|DRM_ROOT_ONLY|DRM_UNLOCKED),
@@ -2024,8 +1031,8 @@ const struct drm_ioctl_desc i915_ioctls[] = {
 	DRM_IOCTL_DEF_DRV(I915_GEM_SET_CACHING, i915_gem_set_caching_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_GEM_GET_CACHING, i915_gem_get_caching_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_GEM_THROTTLE, i915_gem_throttle_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
-	DRM_IOCTL_DEF_DRV(I915_GEM_ENTERVT, i915_gem_entervt_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY|DRM_UNLOCKED),
-	DRM_IOCTL_DEF_DRV(I915_GEM_LEAVEVT, i915_gem_leavevt_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY|DRM_UNLOCKED),
+	DRM_IOCTL_DEF_DRV(I915_GEM_ENTERVT, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY|DRM_UNLOCKED),
+	DRM_IOCTL_DEF_DRV(I915_GEM_LEAVEVT, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY|DRM_UNLOCKED),
 	DRM_IOCTL_DEF_DRV(I915_GEM_CREATE, i915_gem_create_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_GEM_PREAD, i915_gem_pread_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_GEM_PWRITE, i915_gem_pwrite_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index e27cdbe9d524..7643300828c3 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -356,6 +356,19 @@ static const struct intel_device_info intel_cherryview_info = {
 	CURSOR_OFFSETS,
 };
 
+static const struct intel_device_info intel_skylake_info = {
+	.is_preliminary = 1,
+	.is_skylake = 1,
+	.gen = 9, .num_pipes = 3,
+	.need_gfx_hws = 1, .has_hotplug = 1,
+	.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING,
+	.has_llc = 1,
+	.has_ddi = 1,
+	.has_fbc = 1,
+	GEN_DEFAULT_PIPEOFFSETS,
+	IVB_CURSOR_OFFSETS,
+};
+
 /*
  * Make sure any device matches here are from most specific to most
  * general.  For example, since the Quanta match is based on the subsystem
@@ -392,7 +405,8 @@ static const struct intel_device_info intel_cherryview_info = {
 	INTEL_BDW_GT12D_IDS(&intel_broadwell_d_info),	\
 	INTEL_BDW_GT3M_IDS(&intel_broadwell_gt3m_info),	\
 	INTEL_BDW_GT3D_IDS(&intel_broadwell_gt3d_info), \
-	INTEL_CHV_IDS(&intel_cherryview_info)
+	INTEL_CHV_IDS(&intel_cherryview_info),	\
+	INTEL_SKL_IDS(&intel_skylake_info)
 
 static const struct pci_device_id pciidlist[] = {		/* aka */
 	INTEL_PCI_IDS,
@@ -448,19 +462,21 @@ void intel_detect_pch(struct drm_device *dev)
 			} else if (id == INTEL_PCH_LPT_DEVICE_ID_TYPE) {
 				dev_priv->pch_type = PCH_LPT;
 				DRM_DEBUG_KMS("Found LynxPoint PCH\n");
-				WARN_ON(!IS_HASWELL(dev));
-				WARN_ON(IS_ULT(dev));
-			} else if (IS_BROADWELL(dev)) {
-				dev_priv->pch_type = PCH_LPT;
-				dev_priv->pch_id =
-					INTEL_PCH_LPT_LP_DEVICE_ID_TYPE;
-				DRM_DEBUG_KMS("This is Broadwell, assuming "
-					      "LynxPoint LP PCH\n");
+				WARN_ON(!IS_HASWELL(dev) && !IS_BROADWELL(dev));
+				WARN_ON(IS_HSW_ULT(dev) || IS_BDW_ULT(dev));
 			} else if (id == INTEL_PCH_LPT_LP_DEVICE_ID_TYPE) {
 				dev_priv->pch_type = PCH_LPT;
 				DRM_DEBUG_KMS("Found LynxPoint LP PCH\n");
-				WARN_ON(!IS_HASWELL(dev));
-				WARN_ON(!IS_ULT(dev));
+				WARN_ON(!IS_HASWELL(dev) && !IS_BROADWELL(dev));
+				WARN_ON(!IS_HSW_ULT(dev) && !IS_BDW_ULT(dev));
+			} else if (id == INTEL_PCH_SPT_DEVICE_ID_TYPE) {
+				dev_priv->pch_type = PCH_SPT;
+				DRM_DEBUG_KMS("Found SunrisePoint PCH\n");
+				WARN_ON(!IS_SKYLAKE(dev));
+			} else if (id == INTEL_PCH_SPT_LP_DEVICE_ID_TYPE) {
+				dev_priv->pch_type = PCH_SPT;
+				DRM_DEBUG_KMS("Found SunrisePoint LP PCH\n");
+				WARN_ON(!IS_SKYLAKE(dev));
 			} else
 				continue;
 
@@ -481,6 +497,10 @@ bool i915_semaphore_is_enabled(struct drm_device *dev)
 	if (i915.semaphores >= 0)
 		return i915.semaphores;
 
+	/* TODO: make semaphores and Execlists play nicely together */
+	if (i915.enable_execlists)
+		return false;
+
 	/* Until we get further testing... */
 	if (IS_GEN8(dev))
 		return false;
@@ -524,7 +544,11 @@ static void intel_suspend_encoders(struct drm_i915_private *dev_priv)
 	drm_modeset_unlock_all(dev);
 }
 
-static int i915_drm_freeze(struct drm_device *dev)
+static int intel_suspend_complete(struct drm_i915_private *dev_priv);
+static int vlv_resume_prepare(struct drm_i915_private *dev_priv,
+			      bool rpm_resume);
+
+static int i915_drm_suspend(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_crtc *crtc;
@@ -554,6 +578,8 @@ static int i915_drm_freeze(struct drm_device *dev)
 			return error;
 		}
 
+		intel_suspend_gt_powersave(dev);
+
 		/*
 		 * Disable CRTCs directly since we want to preserve sw state
 		 * for _thaw. Also, power gate the CRTC power wells.
@@ -565,16 +591,12 @@ static int i915_drm_freeze(struct drm_device *dev)
 
 		intel_dp_mst_suspend(dev);
 
-		flush_delayed_work(&dev_priv->rps.delayed_resume_work);
-
-		intel_runtime_pm_disable_interrupts(dev);
+		intel_runtime_pm_disable_interrupts(dev_priv);
 		intel_hpd_cancel_work(dev_priv);
 
 		intel_suspend_encoders(dev_priv);
 
-		intel_suspend_gt_powersave(dev);
-
-		intel_modeset_suspend_hw(dev);
+		intel_suspend_hw(dev);
 	}
 
 	i915_gem_suspend_gtt_mappings(dev);
@@ -591,9 +613,7 @@ static int i915_drm_freeze(struct drm_device *dev)
 	intel_uncore_forcewake_reset(dev, false);
 	intel_opregion_fini(dev);
 
-	console_lock();
-	intel_fbdev_set_suspend(dev, FBINFO_STATE_SUSPENDED);
-	console_unlock();
+	intel_fbdev_set_suspend(dev, FBINFO_STATE_SUSPENDED, true);
 
 	dev_priv->suspend_count++;
 
@@ -602,7 +622,26 @@ static int i915_drm_freeze(struct drm_device *dev)
 	return 0;
 }
 
-int i915_suspend(struct drm_device *dev, pm_message_t state)
+static int i915_drm_suspend_late(struct drm_device *drm_dev)
+{
+	struct drm_i915_private *dev_priv = drm_dev->dev_private;
+	int ret;
+
+	ret = intel_suspend_complete(dev_priv);
+
+	if (ret) {
+		DRM_ERROR("Suspend complete failed: %d\n", ret);
+
+		return ret;
+	}
+
+	pci_disable_device(drm_dev->pdev);
+	pci_set_power_state(drm_dev->pdev, PCI_D3hot);
+
+	return 0;
+}
+
+int i915_suspend_legacy(struct drm_device *dev, pm_message_t state)
 {
 	int error;
 
@@ -612,58 +651,25 @@ int i915_suspend(struct drm_device *dev, pm_message_t state)
 		return -ENODEV;
 	}
 
-	if (state.event == PM_EVENT_PRETHAW)
-		return 0;
-
+	if (WARN_ON_ONCE(state.event != PM_EVENT_SUSPEND &&
+			 state.event != PM_EVENT_FREEZE))
+		return -EINVAL;
 
 	if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
 		return 0;
 
-	error = i915_drm_freeze(dev);
+	error = i915_drm_suspend(dev);
 	if (error)
 		return error;
 
-	if (state.event == PM_EVENT_SUSPEND) {
-		/* Shut down the device */
-		pci_disable_device(dev->pdev);
-		pci_set_power_state(dev->pdev, PCI_D3hot);
-	}
-
-	return 0;
-}
-
-void intel_console_resume(struct work_struct *work)
-{
-	struct drm_i915_private *dev_priv =
-		container_of(work, struct drm_i915_private,
-			     console_resume_work);
-	struct drm_device *dev = dev_priv->dev;
-
-	console_lock();
-	intel_fbdev_set_suspend(dev, FBINFO_STATE_RUNNING);
-	console_unlock();
-}
-
-static int i915_drm_thaw_early(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
-		hsw_disable_pc8(dev_priv);
-
-	intel_uncore_early_sanitize(dev, true);
-	intel_uncore_sanitize(dev);
-	intel_power_domains_init_hw(dev_priv);
-
-	return 0;
+	return i915_drm_suspend_late(dev);
 }
 
-static int __i915_drm_thaw(struct drm_device *dev, bool restore_gtt_mappings)
+static int i915_drm_resume(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (drm_core_check_feature(dev, DRIVER_MODESET) &&
-	    restore_gtt_mappings) {
+	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
 		mutex_lock(&dev->struct_mutex);
 		i915_gem_restore_gtt_mappings(dev);
 		mutex_unlock(&dev->struct_mutex);
@@ -684,47 +690,36 @@ static int __i915_drm_thaw(struct drm_device *dev, bool restore_gtt_mappings)
 		}
 		mutex_unlock(&dev->struct_mutex);
 
-		intel_runtime_pm_restore_interrupts(dev);
+		/* We need working interrupts for modeset enabling ... */
+		intel_runtime_pm_enable_interrupts(dev_priv);
 
 		intel_modeset_init_hw(dev);
 
-		{
-			unsigned long irqflags;
-			spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
-			if (dev_priv->display.hpd_irq_setup)
-				dev_priv->display.hpd_irq_setup(dev);
-			spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
-		}
+		spin_lock_irq(&dev_priv->irq_lock);
+		if (dev_priv->display.hpd_irq_setup)
+			dev_priv->display.hpd_irq_setup(dev);
+		spin_unlock_irq(&dev_priv->irq_lock);
 
-		intel_dp_mst_resume(dev);
 		drm_modeset_lock_all(dev);
 		intel_modeset_setup_hw_state(dev, true);
 		drm_modeset_unlock_all(dev);
 
+		intel_dp_mst_resume(dev);
+
 		/*
 		 * ... but also need to make sure that hotplug processing
 		 * doesn't cause havoc. Like in the driver load code we don't
 		 * bother with the tiny race here where we might loose hotplug
 		 * notifications.
 		 * */
-		intel_hpd_init(dev);
+		intel_hpd_init(dev_priv);
 		/* Config may have changed between suspend and resume */
 		drm_helper_hpd_irq_event(dev);
 	}
 
 	intel_opregion_init(dev);
 
-	/*
-	 * The console lock can be pretty contented on resume due
-	 * to all the printk activity.  Try to keep it out of the hot
-	 * path of resume if possible.
-	 */
-	if (console_trylock()) {
-		intel_fbdev_set_suspend(dev, FBINFO_STATE_RUNNING);
-		console_unlock();
-	} else {
-		schedule_work(&dev_priv->console_resume_work);
-	}
+	intel_fbdev_set_suspend(dev, FBINFO_STATE_RUNNING, false);
 
 	mutex_lock(&dev_priv->modeset_restore_lock);
 	dev_priv->modeset_restore = MODESET_DONE;
@@ -732,21 +727,15 @@ static int __i915_drm_thaw(struct drm_device *dev, bool restore_gtt_mappings)
 
 	intel_opregion_notify_adapter(dev, PCI_D0);
 
-	return 0;
-}
-
-static int i915_drm_thaw(struct drm_device *dev)
-{
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		i915_check_and_clear_faults(dev);
+	drm_kms_helper_poll_enable(dev);
 
-	return __i915_drm_thaw(dev, true);
+	return 0;
 }
 
-static int i915_resume_early(struct drm_device *dev)
+static int i915_drm_resume_early(struct drm_device *dev)
 {
-	if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
-		return 0;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret = 0;
 
 	/*
 	 * We have a resume ordering issue with the snd-hda driver also
@@ -762,33 +751,34 @@ static int i915_resume_early(struct drm_device *dev)
 
 	pci_set_master(dev->pdev);
 
-	return i915_drm_thaw_early(dev);
+	if (IS_VALLEYVIEW(dev_priv))
+		ret = vlv_resume_prepare(dev_priv, false);
+	if (ret)
+		DRM_ERROR("Resume prepare failed: %d,Continuing resume\n", ret);
+
+	intel_uncore_early_sanitize(dev, true);
+
+	if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
+		hsw_disable_pc8(dev_priv);
+
+	intel_uncore_sanitize(dev);
+	intel_power_domains_init_hw(dev_priv);
+
+	return ret;
 }
 
-int i915_resume(struct drm_device *dev)
+int i915_resume_legacy(struct drm_device *dev)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
-	/*
-	 * Platforms with opregion should have sane BIOS, older ones (gen3 and
-	 * earlier) need to restore the GTT mappings since the BIOS might clear
-	 * all our scratch PTEs.
-	 */
-	ret = __i915_drm_thaw(dev, !dev_priv->opregion.header);
+	if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
+		return 0;
+
+	ret = i915_drm_resume_early(dev);
 	if (ret)
 		return ret;
 
-	drm_kms_helper_poll_enable(dev);
-	return 0;
-}
-
-static int i915_resume_legacy(struct drm_device *dev)
-{
-	i915_resume_early(dev);
-	i915_resume(dev);
-
-	return 0;
+	return i915_drm_resume(dev);
 }
 
 /**
@@ -815,6 +805,8 @@ int i915_reset(struct drm_device *dev)
 	if (!i915.reset)
 		return 0;
 
+	intel_reset_gt_powersave(dev);
+
 	mutex_lock(&dev->struct_mutex);
 
 	i915_gem_reset(dev);
@@ -834,6 +826,9 @@ int i915_reset(struct drm_device *dev)
 		}
 	}
 
+	if (i915_stop_ring_allow_warn(dev_priv))
+		pr_notice("drm/i915: Resetting chip after gpu hang\n");
+
 	if (ret) {
 		DRM_ERROR("Failed to reset chip: %i\n", ret);
 		mutex_unlock(&dev->struct_mutex);
@@ -854,11 +849,14 @@ int i915_reset(struct drm_device *dev)
 	 * was running at the time of the reset (i.e. we weren't VT
 	 * switched away).
 	 */
-	if (drm_core_check_feature(dev, DRIVER_MODESET) ||
-			!dev_priv->ums.mm_suspended) {
-		dev_priv->ums.mm_suspended = 0;
+	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
+		/* Used to prevent gem_check_wedged returning -EAGAIN during gpu reset */
+		dev_priv->gpu_error.reload_in_reset = true;
 
 		ret = i915_gem_init_hw(dev);
+
+		dev_priv->gpu_error.reload_in_reset = false;
+
 		mutex_unlock(&dev->struct_mutex);
 		if (ret) {
 			DRM_ERROR("Failed hw init on reset %d\n", ret);
@@ -878,9 +876,7 @@ int i915_reset(struct drm_device *dev)
 		 * of re-init after reset.
 		 */
 		if (INTEL_INFO(dev)->gen > 5)
-			intel_reset_gt_powersave(dev);
-
-		intel_hpd_init(dev);
+			intel_enable_gt_powersave(dev);
 	} else {
 		mutex_unlock(&dev->struct_mutex);
 	}
@@ -933,14 +929,13 @@ static int i915_pm_suspend(struct device *dev)
 	if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF)
 		return 0;
 
-	return i915_drm_freeze(drm_dev);
+	return i915_drm_suspend(drm_dev);
 }
 
 static int i915_pm_suspend_late(struct device *dev)
 {
 	struct pci_dev *pdev = to_pci_dev(dev);
 	struct drm_device *drm_dev = pci_get_drvdata(pdev);
-	struct drm_i915_private *dev_priv = drm_dev->dev_private;
 
 	/*
 	 * We have a suspedn ordering issue with the snd-hda driver also
@@ -954,13 +949,7 @@ static int i915_pm_suspend_late(struct device *dev)
 	if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF)
 		return 0;
 
-	if (IS_HASWELL(drm_dev) || IS_BROADWELL(drm_dev))
-		hsw_enable_pc8(dev_priv);
-
-	pci_disable_device(pdev);
-	pci_set_power_state(pdev, PCI_D3hot);
-
-	return 0;
+	return i915_drm_suspend_late(drm_dev);
 }
 
 static int i915_pm_resume_early(struct device *dev)
@@ -968,77 +957,30 @@ static int i915_pm_resume_early(struct device *dev)
 	struct pci_dev *pdev = to_pci_dev(dev);
 	struct drm_device *drm_dev = pci_get_drvdata(pdev);
 
-	return i915_resume_early(drm_dev);
-}
-
-static int i915_pm_resume(struct device *dev)
-{
-	struct pci_dev *pdev = to_pci_dev(dev);
-	struct drm_device *drm_dev = pci_get_drvdata(pdev);
-
-	return i915_resume(drm_dev);
-}
-
-static int i915_pm_freeze(struct device *dev)
-{
-	struct pci_dev *pdev = to_pci_dev(dev);
-	struct drm_device *drm_dev = pci_get_drvdata(pdev);
-
-	if (!drm_dev || !drm_dev->dev_private) {
-		dev_err(dev, "DRM not initialized, aborting suspend.\n");
-		return -ENODEV;
-	}
-
-	return i915_drm_freeze(drm_dev);
-}
-
-static int i915_pm_thaw_early(struct device *dev)
-{
-	struct pci_dev *pdev = to_pci_dev(dev);
-	struct drm_device *drm_dev = pci_get_drvdata(pdev);
+	if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF)
+		return 0;
 
-	return i915_drm_thaw_early(drm_dev);
+	return i915_drm_resume_early(drm_dev);
 }
 
-static int i915_pm_thaw(struct device *dev)
+static int i915_pm_resume(struct device *dev)
 {
 	struct pci_dev *pdev = to_pci_dev(dev);
 	struct drm_device *drm_dev = pci_get_drvdata(pdev);
 
-	return i915_drm_thaw(drm_dev);
-}
-
-static int i915_pm_poweroff(struct device *dev)
-{
-	struct pci_dev *pdev = to_pci_dev(dev);
-	struct drm_device *drm_dev = pci_get_drvdata(pdev);
+	if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF)
+		return 0;
 
-	return i915_drm_freeze(drm_dev);
+	return i915_drm_resume(drm_dev);
 }
 
-static int hsw_runtime_suspend(struct drm_i915_private *dev_priv)
+static int hsw_suspend_complete(struct drm_i915_private *dev_priv)
 {
 	hsw_enable_pc8(dev_priv);
 
 	return 0;
 }
 
-static int snb_runtime_resume(struct drm_i915_private *dev_priv)
-{
-	struct drm_device *dev = dev_priv->dev;
-
-	intel_init_pch_refclk(dev);
-
-	return 0;
-}
-
-static int hsw_runtime_resume(struct drm_i915_private *dev_priv)
-{
-	hsw_disable_pc8(dev_priv);
-
-	return 0;
-}
-
 /*
  * Save all Gunit registers that may be lost after a D3 and a subsequent
  * S0i[R123] transition. The list of registers needing a save/restore is
@@ -1328,7 +1270,7 @@ static void vlv_check_no_gt_access(struct drm_i915_private *dev_priv)
 	I915_WRITE(VLV_GTLC_PW_STATUS, VLV_GTLC_ALLOWWAKEERR);
 }
 
-static int vlv_runtime_suspend(struct drm_i915_private *dev_priv)
+static int vlv_suspend_complete(struct drm_i915_private *dev_priv)
 {
 	u32 mask;
 	int err;
@@ -1368,7 +1310,8 @@ err1:
 	return err;
 }
 
-static int vlv_runtime_resume(struct drm_i915_private *dev_priv)
+static int vlv_resume_prepare(struct drm_i915_private *dev_priv,
+				bool rpm_resume)
 {
 	struct drm_device *dev = dev_priv->dev;
 	int err;
@@ -1393,8 +1336,10 @@ static int vlv_runtime_resume(struct drm_i915_private *dev_priv)
 
 	vlv_check_no_gt_access(dev_priv);
 
-	intel_init_clock_gating(dev);
-	i915_gem_restore_fences(dev);
+	if (rpm_resume) {
+		intel_init_clock_gating(dev);
+		i915_gem_restore_fences(dev);
+	}
 
 	return ret;
 }
@@ -1409,7 +1354,9 @@ static int intel_runtime_suspend(struct device *device)
 	if (WARN_ON_ONCE(!(dev_priv->rps.enabled && intel_enable_rc6(dev))))
 		return -ENODEV;
 
-	WARN_ON(!HAS_RUNTIME_PM(dev));
+	if (WARN_ON_ONCE(!HAS_RUNTIME_PM(dev)))
+		return -ENODEV;
+
 	assert_force_wake_inactive(dev_priv);
 
 	DRM_DEBUG_KMS("Suspending device\n");
@@ -1438,28 +1385,13 @@ static int intel_runtime_suspend(struct device *device)
 	i915_gem_release_all_mmaps(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
-	/*
-	 * rps.work can't be rearmed here, since we get here only after making
-	 * sure the GPU is idle and the RPS freq is set to the minimum. See
-	 * intel_mark_idle().
-	 */
-	cancel_work_sync(&dev_priv->rps.work);
-	intel_runtime_pm_disable_interrupts(dev);
-
-	if (IS_GEN6(dev)) {
-		ret = 0;
-	} else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
-		ret = hsw_runtime_suspend(dev_priv);
-	} else if (IS_VALLEYVIEW(dev)) {
-		ret = vlv_runtime_suspend(dev_priv);
-	} else {
-		ret = -ENODEV;
-		WARN_ON(1);
-	}
+	intel_suspend_gt_powersave(dev);
+	intel_runtime_pm_disable_interrupts(dev_priv);
 
+	ret = intel_suspend_complete(dev_priv);
 	if (ret) {
 		DRM_ERROR("Runtime suspend failed, disabling it (%d)\n", ret);
-		intel_runtime_pm_restore_interrupts(dev);
+		intel_runtime_pm_enable_interrupts(dev_priv);
 
 		return ret;
 	}
@@ -1468,13 +1400,29 @@ static int intel_runtime_suspend(struct device *device)
 	dev_priv->pm.suspended = true;
 
 	/*
-	 * current versions of firmware which depend on this opregion
-	 * notification have repurposed the D1 definition to mean
-	 * "runtime suspended" vs. what you would normally expect (D3)
-	 * to distinguish it from notifications that might be sent
-	 * via the suspend path.
+	 * FIXME: We really should find a document that references the arguments
+	 * used below!
 	 */
-	intel_opregion_notify_adapter(dev, PCI_D1);
+	if (IS_HASWELL(dev)) {
+		/*
+		 * current versions of firmware which depend on this opregion
+		 * notification have repurposed the D1 definition to mean
+		 * "runtime suspended" vs. what you would normally expect (D3)
+		 * to distinguish it from notifications that might be sent via
+		 * the suspend path.
+		 */
+		intel_opregion_notify_adapter(dev, PCI_D1);
+	} else {
+		/*
+		 * On Broadwell, if we use PCI_D1 the PCH DDI ports will stop
+		 * being detected, and the call we do at intel_runtime_resume()
+		 * won't be able to restore them. Since PCI_D3hot matches the
+		 * actual specification and appears to be working, use it. Let's
+		 * assume the other non-Haswell platforms will stay the same as
+		 * Broadwell.
+		 */
+		intel_opregion_notify_adapter(dev, PCI_D3hot);
+	}
 
 	DRM_DEBUG_KMS("Device suspended\n");
 	return 0;
@@ -1485,25 +1433,22 @@ static int intel_runtime_resume(struct device *device)
 	struct pci_dev *pdev = to_pci_dev(device);
 	struct drm_device *dev = pci_get_drvdata(pdev);
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	int ret;
+	int ret = 0;
 
-	WARN_ON(!HAS_RUNTIME_PM(dev));
+	if (WARN_ON_ONCE(!HAS_RUNTIME_PM(dev)))
+		return -ENODEV;
 
 	DRM_DEBUG_KMS("Resuming device\n");
 
 	intel_opregion_notify_adapter(dev, PCI_D0);
 	dev_priv->pm.suspended = false;
 
-	if (IS_GEN6(dev)) {
-		ret = snb_runtime_resume(dev_priv);
-	} else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
-		ret = hsw_runtime_resume(dev_priv);
-	} else if (IS_VALLEYVIEW(dev)) {
-		ret = vlv_runtime_resume(dev_priv);
-	} else {
-		WARN_ON(1);
-		ret = -ENODEV;
-	}
+	if (IS_GEN6(dev_priv))
+		intel_init_pch_refclk(dev);
+	else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
+		hsw_disable_pc8(dev_priv);
+	else if (IS_VALLEYVIEW(dev_priv))
+		ret = vlv_resume_prepare(dev_priv, true);
 
 	/*
 	 * No point of rolling back things in case of an error, as the best
@@ -1512,8 +1457,8 @@ static int intel_runtime_resume(struct device *device)
 	i915_gem_init_swizzling(dev);
 	gen6_update_ring_freq(dev);
 
-	intel_runtime_pm_restore_interrupts(dev);
-	intel_reset_gt_powersave(dev);
+	intel_runtime_pm_enable_interrupts(dev_priv);
+	intel_enable_gt_powersave(dev);
 
 	if (ret)
 		DRM_ERROR("Runtime resume failed, disabling it (%d)\n", ret);
@@ -1523,17 +1468,60 @@ static int intel_runtime_resume(struct device *device)
 	return ret;
 }
 
+/*
+ * This function implements common functionality of runtime and system
+ * suspend sequence.
+ */
+static int intel_suspend_complete(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	int ret;
+
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+		ret = hsw_suspend_complete(dev_priv);
+	else if (IS_VALLEYVIEW(dev))
+		ret = vlv_suspend_complete(dev_priv);
+	else
+		ret = 0;
+
+	return ret;
+}
+
 static const struct dev_pm_ops i915_pm_ops = {
+	/*
+	 * S0ix (via system suspend) and S3 event handlers [PMSG_SUSPEND,
+	 * PMSG_RESUME]
+	 */
 	.suspend = i915_pm_suspend,
 	.suspend_late = i915_pm_suspend_late,
 	.resume_early = i915_pm_resume_early,
 	.resume = i915_pm_resume,
-	.freeze = i915_pm_freeze,
-	.thaw_early = i915_pm_thaw_early,
-	.thaw = i915_pm_thaw,
-	.poweroff = i915_pm_poweroff,
+
+	/*
+	 * S4 event handlers
+	 * @freeze, @freeze_late    : called (1) before creating the
+	 *                            hibernation image [PMSG_FREEZE] and
+	 *                            (2) after rebooting, before restoring
+	 *                            the image [PMSG_QUIESCE]
+	 * @thaw, @thaw_early       : called (1) after creating the hibernation
+	 *                            image, before writing it [PMSG_THAW]
+	 *                            and (2) after failing to create or
+	 *                            restore the image [PMSG_RECOVER]
+	 * @poweroff, @poweroff_late: called after writing the hibernation
+	 *                            image, before rebooting [PMSG_HIBERNATE]
+	 * @restore, @restore_early : called after rebooting and restoring the
+	 *                            hibernation image [PMSG_RESTORE]
+	 */
+	.freeze = i915_pm_suspend,
+	.freeze_late = i915_pm_suspend_late,
+	.thaw_early = i915_pm_resume_early,
+	.thaw = i915_pm_resume,
+	.poweroff = i915_pm_suspend,
+	.poweroff_late = i915_pm_suspend_late,
 	.restore_early = i915_pm_resume_early,
 	.restore = i915_pm_resume,
+
+	/* S0ix (via runtime suspend) event handlers */
 	.runtime_suspend = intel_runtime_suspend,
 	.runtime_resume = intel_runtime_resume,
 };
@@ -1572,14 +1560,13 @@ static struct drm_driver driver = {
 	.lastclose = i915_driver_lastclose,
 	.preclose = i915_driver_preclose,
 	.postclose = i915_driver_postclose,
+	.set_busid = drm_pci_set_busid,
 
 	/* Used in place of i915_pm_ops for non-DRIVER_MODESET */
-	.suspend = i915_suspend,
+	.suspend = i915_suspend_legacy,
 	.resume = i915_resume_legacy,
 
 	.device_is_agp = i915_driver_device_is_agp,
-	.master_create = i915_master_create,
-	.master_destroy = i915_master_destroy,
 #if defined(CONFIG_DEBUG_FS)
 	.debugfs_init = i915_debugfs_init,
 	.debugfs_cleanup = i915_debugfs_cleanup,
@@ -1663,6 +1650,8 @@ static void __exit i915_exit(void)
 module_init(i915_init);
 module_exit(i915_exit);
 
-MODULE_AUTHOR(DRIVER_AUTHOR);
+MODULE_AUTHOR("Tungsten Graphics, Inc.");
+MODULE_AUTHOR("Intel Corporation");
+
 MODULE_DESCRIPTION(DRIVER_DESC);
 MODULE_LICENSE("GPL and additional rights");
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 3524306d8cfb..9d7a7155bf02 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -35,11 +35,15 @@
 #include "i915_reg.h"
 #include "intel_bios.h"
 #include "intel_ringbuffer.h"
+#include "intel_lrc.h"
 #include "i915_gem_gtt.h"
+#include "i915_gem_render_state.h"
 #include <linux/io-mapping.h>
 #include <linux/i2c.h>
 #include <linux/i2c-algo-bit.h>
 #include <drm/intel-gtt.h>
+#include <drm/drm_legacy.h> /* for struct drm_dma_handle */
+#include <drm/drm_gem.h>
 #include <linux/backlight.h>
 #include <linux/hashtable.h>
 #include <linux/intel-iommu.h>
@@ -49,11 +53,12 @@
 /* General customization:
  */
 
-#define DRIVER_AUTHOR		"Tungsten Graphics, Inc."
-
 #define DRIVER_NAME		"i915"
 #define DRIVER_DESC		"Intel Graphics"
-#define DRIVER_DATE		"20140725"
+#define DRIVER_DATE		"20141121"
+
+#undef WARN_ON
+#define WARN_ON(x)		WARN(x, "WARN_ON(" #x ")")
 
 enum pipe {
 	INVALID_PIPE = -1,
@@ -74,6 +79,14 @@ enum transcoder {
 };
 #define transcoder_name(t) ((t) + 'A')
 
+/*
+ * This is the maximum (across all platforms) number of planes (primary +
+ * sprites) that can be active at the same time on one pipe.
+ *
+ * This value doesn't count the cursor plane.
+ */
+#define I915_MAX_PLANES	3
+
 enum plane {
 	PLANE_A = 0,
 	PLANE_B,
@@ -162,7 +175,10 @@ enum hpd_pin {
 	 I915_GEM_DOMAIN_INSTRUCTION | \
 	 I915_GEM_DOMAIN_VERTEX)
 
-#define for_each_pipe(p) for ((p) = 0; (p) < INTEL_INFO(dev)->num_pipes; (p)++)
+#define for_each_pipe(__dev_priv, __p) \
+	for ((__p) = 0; (__p) < INTEL_INFO(__dev_priv)->num_pipes; (__p)++)
+#define for_each_plane(pipe, p) \
+	for ((p) = 0; (p) < INTEL_INFO(dev)->num_sprites[(pipe)] + 1; (p)++)
 #define for_each_sprite(p, s) for ((s) = 0; (s) < INTEL_INFO(dev)->num_sprites[(p)]; (s)++)
 
 #define for_each_crtc(dev, crtc) \
@@ -171,6 +187,11 @@ enum hpd_pin {
 #define for_each_intel_crtc(dev, intel_crtc) \
 	list_for_each_entry(intel_crtc, &dev->mode_config.crtc_list, base.head)
 
+#define for_each_intel_encoder(dev, intel_encoder)		\
+	list_for_each_entry(intel_encoder,			\
+			    &(dev)->mode_config.encoder_list,	\
+			    base.head)
+
 #define for_each_encoder_on_crtc(dev, __crtc, intel_encoder) \
 	list_for_each_entry((intel_encoder), &(dev)->mode_config.encoder_list, base.head) \
 		if ((intel_encoder)->base.crtc == (__crtc))
@@ -192,27 +213,52 @@ enum intel_dpll_id {
 	/* real shared dpll ids must be >= 0 */
 	DPLL_ID_PCH_PLL_A = 0,
 	DPLL_ID_PCH_PLL_B = 1,
+	/* hsw/bdw */
 	DPLL_ID_WRPLL1 = 0,
 	DPLL_ID_WRPLL2 = 1,
+	/* skl */
+	DPLL_ID_SKL_DPLL1 = 0,
+	DPLL_ID_SKL_DPLL2 = 1,
+	DPLL_ID_SKL_DPLL3 = 2,
 };
-#define I915_NUM_PLLS 2
+#define I915_NUM_PLLS 3
 
 struct intel_dpll_hw_state {
+	/* i9xx, pch plls */
 	uint32_t dpll;
 	uint32_t dpll_md;
 	uint32_t fp0;
 	uint32_t fp1;
+
+	/* hsw, bdw */
 	uint32_t wrpll;
+
+	/* skl */
+	/*
+	 * DPLL_CTRL1 has 6 bits for each each this DPLL. We store those in
+	 * lower part of crtl1 and they get shifted into position when writing
+	 * the register.  This allows us to easily compare the state to share
+	 * the DPLL.
+	 */
+	uint32_t ctrl1;
+	/* HDMI only, 0 when used for DP */
+	uint32_t cfgcr1, cfgcr2;
+};
+
+struct intel_shared_dpll_config {
+	unsigned crtc_mask; /* mask of CRTCs sharing this PLL */
+	struct intel_dpll_hw_state hw_state;
 };
 
 struct intel_shared_dpll {
-	int refcount; /* count of number of CRTCs sharing this PLL */
+	struct intel_shared_dpll_config config;
+	struct intel_shared_dpll_config *new_config;
+
 	int active; /* count of number of active CRTCs (i.e. DPMS on) */
 	bool on; /* is the PLL actually active? Disabled during modeset */
 	const char *name;
 	/* should match the index in the dev_priv->shared_dplls array */
 	enum intel_dpll_id id;
-	struct intel_dpll_hw_state hw_state;
 	/* The mode_set hook is optional and should be used together with the
 	 * intel_prepare_shared_dpll function. */
 	void (*mode_set)(struct drm_i915_private *dev_priv,
@@ -226,6 +272,11 @@ struct intel_shared_dpll {
 			     struct intel_dpll_hw_state *hw_state);
 };
 
+#define SKL_DPLL0 0
+#define SKL_DPLL1 1
+#define SKL_DPLL2 2
+#define SKL_DPLL3 3
+
 /* Used by dp and fdi links */
 struct intel_link_m_n {
 	uint32_t	tu;
@@ -254,7 +305,6 @@ void intel_link_compute_m_n(int bpp, int nlanes,
 #define DRIVER_PATCHLEVEL	0
 
 #define WATCH_LISTS	0
-#define WATCH_GTT	0
 
 struct opregion_header;
 struct opregion_acpi;
@@ -277,10 +327,6 @@ struct intel_opregion {
 struct intel_overlay;
 struct intel_overlay_error_state;
 
-struct drm_i915_master_private {
-	drm_local_map_t *sarea;
-	struct _drm_i915_sarea *sarea_priv;
-};
 #define I915_FENCE_REG_NONE -1
 #define I915_MAX_NUM_FENCES 32
 /* 32 fences + sign bit for FENCE_REG_NONE */
@@ -388,6 +434,7 @@ struct drm_i915_error_state {
 		pid_t pid;
 		char comm[TASK_COMM_LEN];
 	} ring[I915_NUM_RINGS];
+
 	struct drm_i915_error_buffer {
 		u32 size;
 		u32 name;
@@ -406,9 +453,11 @@ struct drm_i915_error_state {
 	} **active_bo, **pinned_bo;
 
 	u32 *active_bo_count, *pinned_bo_count;
+	u32 vm_count;
 };
 
 struct intel_connector;
+struct intel_encoder;
 struct intel_crtc_config;
 struct intel_plane_config;
 struct intel_crtc;
@@ -435,7 +484,7 @@ struct drm_i915_display_funcs {
 	 * Returns true on success, false on failure.
 	 */
 	bool (*find_dpll)(const struct intel_limit *limit,
-			  struct drm_crtc *crtc,
+			  struct intel_crtc *crtc,
 			  int target, int refclk,
 			  struct dpll *match_clock,
 			  struct dpll *best_clock);
@@ -451,15 +500,14 @@ struct drm_i915_display_funcs {
 				struct intel_crtc_config *);
 	void (*get_plane_config)(struct intel_crtc *,
 				 struct intel_plane_config *);
-	int (*crtc_mode_set)(struct drm_crtc *crtc,
-			     int x, int y,
-			     struct drm_framebuffer *old_fb);
+	int (*crtc_compute_clock)(struct intel_crtc *crtc);
 	void (*crtc_enable)(struct drm_crtc *crtc);
 	void (*crtc_disable)(struct drm_crtc *crtc);
 	void (*off)(struct drm_crtc *crtc);
-	void (*write_eld)(struct drm_connector *connector,
-			  struct drm_crtc *crtc,
-			  struct drm_display_mode *mode);
+	void (*audio_codec_enable)(struct drm_connector *connector,
+				   struct intel_encoder *encoder,
+				   struct drm_display_mode *mode);
+	void (*audio_codec_disable)(struct intel_encoder *encoder);
 	void (*fdi_link_train)(struct drm_crtc *crtc);
 	void (*init_clock_gating)(struct drm_device *dev);
 	int (*queue_flip)(struct drm_device *dev, struct drm_crtc *crtc,
@@ -477,7 +525,7 @@ struct drm_i915_display_funcs {
 	/* display clock increase/decrease */
 	/* pll clock increase/decrease */
 
-	int (*setup_backlight)(struct intel_connector *connector);
+	int (*setup_backlight)(struct intel_connector *connector, enum pipe pipe);
 	uint32_t (*get_backlight)(struct intel_connector *connector);
 	void (*set_backlight)(struct intel_connector *connector,
 			      uint32_t level);
@@ -516,6 +564,7 @@ struct intel_uncore {
 
 	unsigned fw_rendercount;
 	unsigned fw_mediacount;
+	unsigned fw_blittercount;
 
 	struct timer_list force_wake_timer;
 };
@@ -534,6 +583,7 @@ struct intel_uncore {
 	func(is_ivybridge) sep \
 	func(is_valleyview) sep \
 	func(is_haswell) sep \
+	func(is_skylake) sep \
 	func(is_preliminary) sep \
 	func(has_fbc) sep \
 	func(has_pipe_cxsr) sep \
@@ -551,6 +601,7 @@ struct intel_uncore {
 
 struct intel_device_info {
 	u32 display_mmio_offset;
+	u16 device_id;
 	u8 num_pipes:3;
 	u8 num_sprites[I915_MAX_PIPES];
 	u8 gen;
@@ -615,13 +666,22 @@ struct intel_context {
 	uint8_t remap_slice;
 	struct drm_i915_file_private *file_priv;
 	struct i915_ctx_hang_stats hang_stats;
-	struct i915_address_space *vm;
+	struct i915_hw_ppgtt *ppgtt;
 
+	/* Legacy ring buffer submission */
 	struct {
 		struct drm_i915_gem_object *rcs_state;
 		bool initialized;
 	} legacy_hw_ctx;
 
+	/* Execlists */
+	bool rcs_initialized;
+	struct {
+		struct drm_i915_gem_object *state;
+		struct intel_ringbuffer *ringbuf;
+		int unpin_count;
+	} engine[I915_NUM_RINGS];
+
 	struct list_head link;
 };
 
@@ -635,6 +695,20 @@ struct i915_fbc {
 	struct drm_mm_node compressed_fb;
 	struct drm_mm_node *compressed_llb;
 
+	bool false_color;
+
+	/* Tracks whether the HW is actually enabled, not whether the feature is
+	 * possible. */
+	bool enabled;
+
+	/* On gen8 some rings cannont perform fbc clean operation so for now
+	 * we are doing this on SW with mmio.
+	 * This variable works in the opposite information direction
+	 * of ring->fbc_dirty telling software on frontbuffer tracking
+	 * to perform the cache clean on sw side.
+	 */
+	bool need_sw_cache_clean;
+
 	struct intel_fbc_work {
 		struct delayed_work work;
 		struct drm_crtc *crtc;
@@ -676,6 +750,7 @@ enum intel_pch {
 	PCH_IBX,	/* Ibexpeak PCH */
 	PCH_CPT,	/* Cougarpoint PCH */
 	PCH_LPT,	/* Lynxpoint PCH */
+	PCH_SPT,        /* Sunrisepoint PCH */
 	PCH_NOP,
 };
 
@@ -688,6 +763,8 @@ enum intel_sbi_destination {
 #define QUIRK_LVDS_SSC_DISABLE (1<<1)
 #define QUIRK_INVERT_BRIGHTNESS (1<<2)
 #define QUIRK_BACKLIGHT_PRESENT (1<<3)
+#define QUIRK_PIPEB_FORCE (1<<4)
+#define QUIRK_PIN_SWIZZLED_PAGES (1<<5)
 
 struct intel_fbdev;
 struct intel_fbc_work;
@@ -739,7 +816,6 @@ struct i915_suspend_saved_registers {
 	u32 saveBLC_HIST_CTL;
 	u32 saveBLC_PWM_CTL;
 	u32 saveBLC_PWM_CTL2;
-	u32 saveBLC_HIST_CTL_B;
 	u32 saveBLC_CPU_PWM_CTL;
 	u32 saveBLC_CPU_PWM_CTL2;
 	u32 saveFPB0;
@@ -848,6 +924,7 @@ struct i915_suspend_saved_registers {
 	u32 savePIPEB_LINK_N1;
 	u32 saveMCHBAR_RENDER_STANDBY;
 	u32 savePCH_PORT_HOTPLUG;
+	u16 saveGCDGMBUS;
 };
 
 struct vlv_s0ix_state {
@@ -918,8 +995,12 @@ struct intel_rps_ei {
 };
 
 struct intel_gen6_power_mgmt {
-	/* work and pm_iir are protected by dev_priv->irq_lock */
+	/*
+	 * work, interrupts_enabled and pm_iir are protected by
+	 * dev_priv->irq_lock
+	 */
 	struct work_struct work;
+	bool interrupts_enabled;
 	u32 pm_iir;
 
 	/* Frequencies are stored in potentially platform dependent multiples.
@@ -1042,31 +1123,6 @@ struct i915_power_domains {
 	struct i915_power_well *power_wells;
 };
 
-struct i915_dri1_state {
-	unsigned allow_batchbuffer : 1;
-	u32 __iomem *gfx_hws_cpu_addr;
-
-	unsigned int cpp;
-	int back_offset;
-	int front_offset;
-	int current_page;
-	int page_flipping;
-
-	uint32_t counter;
-};
-
-struct i915_ums_state {
-	/**
-	 * Flag if the X Server, and thus DRM, is not currently in
-	 * control of the device.
-	 *
-	 * This is set between LeaveVT and EnterVT.  It needs to be
-	 * replaced with a semaphore.  It also needs to be
-	 * transitioned away from for kernel modesetting.
-	 */
-	int mm_suspended;
-};
-
 #define MAX_L3_SLICES 2
 struct intel_l3_parity {
 	u32 *remap_info[MAX_L3_SLICES];
@@ -1147,6 +1203,7 @@ struct i915_gem_mm {
 };
 
 struct drm_i915_error_state_buf {
+	struct drm_i915_private *i915;
 	unsigned bytes;
 	unsigned size;
 	int err;
@@ -1219,6 +1276,9 @@ struct i915_gpu_error {
 
 	/* For missed irq/seqno simulation. */
 	unsigned int test_irq_rings;
+
+	/* Used to prevent gem_check_wedged returning -EAGAIN during gpu reset   */
+	bool reload_in_reset;
 };
 
 enum modeset_restore {
@@ -1228,6 +1288,12 @@ enum modeset_restore {
 };
 
 struct ddi_vbt_port_info {
+	/*
+	 * This is an index in the HDMI/DVI DDI buffer translation table.
+	 * The special value HDMI_LEVEL_SHIFT_UNKNOWN means the VBT didn't
+	 * populate this field.
+	 */
+#define HDMI_LEVEL_SHIFT_UNKNOWN	0xff
 	uint8_t hdmi_level_shift;
 
 	uint8_t supports_dvi:1;
@@ -1318,6 +1384,49 @@ struct ilk_wm_values {
 	enum intel_ddb_partitioning partitioning;
 };
 
+struct skl_ddb_entry {
+	uint16_t start, end;	/* in number of blocks, 'end' is exclusive */
+};
+
+static inline uint16_t skl_ddb_entry_size(const struct skl_ddb_entry *entry)
+{
+	return entry->end - entry->start;
+}
+
+static inline bool skl_ddb_entry_equal(const struct skl_ddb_entry *e1,
+				       const struct skl_ddb_entry *e2)
+{
+	if (e1->start == e2->start && e1->end == e2->end)
+		return true;
+
+	return false;
+}
+
+struct skl_ddb_allocation {
+	struct skl_ddb_entry pipe[I915_MAX_PIPES];
+	struct skl_ddb_entry plane[I915_MAX_PIPES][I915_MAX_PLANES];
+	struct skl_ddb_entry cursor[I915_MAX_PIPES];
+};
+
+struct skl_wm_values {
+	bool dirty[I915_MAX_PIPES];
+	struct skl_ddb_allocation ddb;
+	uint32_t wm_linetime[I915_MAX_PIPES];
+	uint32_t plane[I915_MAX_PIPES][I915_MAX_PLANES][8];
+	uint32_t cursor[I915_MAX_PIPES][8];
+	uint32_t plane_trans[I915_MAX_PIPES][I915_MAX_PLANES];
+	uint32_t cursor_trans[I915_MAX_PIPES];
+};
+
+struct skl_wm_level {
+	bool plane_en[I915_MAX_PLANES];
+	bool cursor_en;
+	uint16_t plane_res_b[I915_MAX_PLANES];
+	uint8_t plane_res_l[I915_MAX_PLANES];
+	uint16_t cursor_res_b;
+	uint8_t cursor_res_l;
+};
+
 /*
  * This struct helps tracking the state needed for runtime PM, which puts the
  * device in PCI D3 state. Notice that when this happens, nothing on the
@@ -1330,7 +1439,7 @@ struct ilk_wm_values {
  *
  * Our driver uses the autosuspend delay feature, which means we'll only really
  * suspend if we stay with zero refcount for a certain amount of time. The
- * default value is currently very conservative (see intel_init_runtime_pm), but
+ * default value is currently very conservative (see intel_runtime_pm_enable), but
  * it can be changed with the standard runtime PM files from sysfs.
  *
  * The irqs_disabled variable becomes true exactly after we disable the IRQs and
@@ -1343,7 +1452,7 @@ struct ilk_wm_values {
  */
 struct i915_runtime_pm {
 	bool suspended;
-	bool _irqs_disabled;
+	bool irqs_enabled;
 };
 
 enum intel_pipe_crc_source {
@@ -1387,6 +1496,20 @@ struct i915_frontbuffer_tracking {
 	unsigned flip_bits;
 };
 
+struct i915_wa_reg {
+	u32 addr;
+	u32 value;
+	/* bitmask representing WA bits */
+	u32 mask;
+};
+
+#define I915_MAX_WA_REGS 16
+
+struct i915_workarounds {
+	struct i915_wa_reg reg[I915_MAX_WA_REGS];
+	u32 count;
+};
+
 struct drm_i915_private {
 	struct drm_device *dev;
 	struct kmem_cache *slab;
@@ -1421,7 +1544,7 @@ struct drm_i915_private {
 	struct drm_i915_gem_object *semaphore_obj;
 	uint32_t last_seqno, next_seqno;
 
-	drm_dma_handle_t *status_page_dmah;
+	struct drm_dma_handle *status_page_dmah;
 	struct resource mch_res;
 
 	/* protects the irq masks */
@@ -1466,21 +1589,27 @@ struct drm_i915_private {
 	struct intel_opregion opregion;
 	struct intel_vbt_data vbt;
 
+	bool preserve_bios_swizzle;
+
 	/* overlay */
 	struct intel_overlay *overlay;
 
 	/* backlight registers and fields in struct intel_panel */
-	spinlock_t backlight_lock;
+	struct mutex backlight_lock;
 
 	/* LVDS info */
 	bool no_aux_handshake;
 
+	/* protects panel power sequencer state */
+	struct mutex pps_mutex;
+
 	struct drm_i915_fence_reg fence_regs[I915_MAX_NUM_FENCES]; /* assume 965 */
 	int fence_reg_start; /* 4 if userland hasn't ioctl'd us yet */
 	int num_fence_regs; /* 8 on pre-965, 16 otherwise */
 
 	unsigned int fsb_freq, mem_freq, is_ddr3;
 	unsigned int vlv_cdclk_freq;
+	unsigned int hpll_freq;
 
 	/**
 	 * wq - Driver workqueue for GEM.
@@ -1526,6 +1655,8 @@ struct drm_i915_private {
 	struct intel_shared_dpll shared_dplls[I915_NUM_PLLS];
 	int dpio_phy_iosf_port[I915_NUM_PHYS_VLV];
 
+	struct i915_workarounds workarounds;
+
 	/* Reclocking support */
 	bool render_reclock_avail;
 	bool lvds_downclock_avail;
@@ -1561,14 +1692,9 @@ struct drm_i915_private {
 #ifdef CONFIG_DRM_I915_FBDEV
 	/* list of fbdev register on this device */
 	struct intel_fbdev *fbdev;
+	struct work_struct fbdev_suspend_work;
 #endif
 
-	/*
-	 * The console may be contended at resume, but we don't
-	 * want it to block on it.
-	 */
-	struct work_struct console_resume_work;
-
 	struct drm_property *broadcast_rgb_property;
 	struct drm_property *force_audio_property;
 
@@ -1593,9 +1719,25 @@ struct drm_i915_private {
 		uint16_t spr_latency[5];
 		/* cursor */
 		uint16_t cur_latency[5];
+		/*
+		 * Raw watermark memory latency values
+		 * for SKL for all 8 levels
+		 * in 1us units.
+		 */
+		uint16_t skl_latency[8];
+
+		/*
+		 * The skl_wm_values structure is a bit too big for stack
+		 * allocation, so we keep the staging struct where we store
+		 * intermediate results here instead.
+		 */
+		struct skl_wm_values skl_results;
 
 		/* current hardware state */
-		struct ilk_wm_values hw;
+		union {
+			struct ilk_wm_values hw;
+			struct skl_wm_values skl_hw;
+		};
 	} wm;
 
 	struct i915_runtime_pm pm;
@@ -1614,11 +1756,19 @@ struct drm_i915_private {
 	 */
 	struct workqueue_struct *dp_wq;
 
-	/* Old dri1 support infrastructure, beware the dragons ya fools entering
-	 * here! */
-	struct i915_dri1_state dri1;
-	/* Old ums support infrastructure, same warning applies. */
-	struct i915_ums_state ums;
+	/* Abstract the submission mechanism (legacy ringbuffer or execlists) away */
+	struct {
+		int (*do_execbuf)(struct drm_device *dev, struct drm_file *file,
+				  struct intel_engine_cs *ring,
+				  struct intel_context *ctx,
+				  struct drm_i915_gem_execbuffer2 *args,
+				  struct list_head *vmas,
+				  struct drm_i915_gem_object *batch_obj,
+				  u64 exec_start, u32 flags);
+		int (*init_rings)(struct drm_device *dev);
+		void (*cleanup_ring)(struct intel_engine_cs *ring);
+		void (*stop_ring)(struct intel_engine_cs *ring);
+	} gt;
 
 	/*
 	 * NOTE: This is the dri1/ums dungeon, don't add stuff here. Your patch
@@ -1761,17 +1911,8 @@ struct drm_i915_gem_object {
 	 * Only honoured if hardware has relevant pte bit
 	 */
 	unsigned long gt_ro:1;
-
-	/*
-	 * Is the GPU currently using a fence to access this buffer,
-	 */
-	unsigned int pending_fenced_gpu_access:1;
-	unsigned int fenced_gpu_access:1;
-
 	unsigned int cache_level:3;
 
-	unsigned int has_aliasing_ppgtt_mapping:1;
-	unsigned int has_global_gtt_mapping:1;
 	unsigned int has_dma_mapping:1;
 
 	unsigned int frontbuffer_bits:INTEL_FRONTBUFFER_BITS;
@@ -1804,10 +1945,10 @@ struct drm_i915_gem_object {
 	unsigned long user_pin_count;
 	struct drm_file *pin_filp;
 
-	/** for phy allocated objects */
-	drm_dma_handle_t *phys_handle;
-
 	union {
+		/** for phy allocated objects */
+		struct drm_dma_handle *phys_handle;
+
 		struct i915_gem_userptr {
 			uintptr_t ptr;
 			unsigned read_only :1;
@@ -1971,51 +2112,64 @@ struct drm_i915_cmd_table {
 	int count;
 };
 
-#define INTEL_INFO(dev)	(&to_i915(dev)->info)
-
-#define IS_I830(dev)		((dev)->pdev->device == 0x3577)
-#define IS_845G(dev)		((dev)->pdev->device == 0x2562)
+/* Note that the (struct drm_i915_private *) cast is just to shut up gcc. */
+#define __I915__(p) ({ \
+	struct drm_i915_private *__p; \
+	if (__builtin_types_compatible_p(typeof(*p), struct drm_i915_private)) \
+		__p = (struct drm_i915_private *)p; \
+	else if (__builtin_types_compatible_p(typeof(*p), struct drm_device)) \
+		__p = to_i915((struct drm_device *)p); \
+	else \
+		BUILD_BUG(); \
+	__p; \
+})
+#define INTEL_INFO(p) 	(&__I915__(p)->info)
+#define INTEL_DEVID(p)	(INTEL_INFO(p)->device_id)
+
+#define IS_I830(dev)		(INTEL_DEVID(dev) == 0x3577)
+#define IS_845G(dev)		(INTEL_DEVID(dev) == 0x2562)
 #define IS_I85X(dev)		(INTEL_INFO(dev)->is_i85x)
-#define IS_I865G(dev)		((dev)->pdev->device == 0x2572)
+#define IS_I865G(dev)		(INTEL_DEVID(dev) == 0x2572)
 #define IS_I915G(dev)		(INTEL_INFO(dev)->is_i915g)
-#define IS_I915GM(dev)		((dev)->pdev->device == 0x2592)
-#define IS_I945G(dev)		((dev)->pdev->device == 0x2772)
+#define IS_I915GM(dev)		(INTEL_DEVID(dev) == 0x2592)
+#define IS_I945G(dev)		(INTEL_DEVID(dev) == 0x2772)
 #define IS_I945GM(dev)		(INTEL_INFO(dev)->is_i945gm)
 #define IS_BROADWATER(dev)	(INTEL_INFO(dev)->is_broadwater)
 #define IS_CRESTLINE(dev)	(INTEL_INFO(dev)->is_crestline)
-#define IS_GM45(dev)		((dev)->pdev->device == 0x2A42)
+#define IS_GM45(dev)		(INTEL_DEVID(dev) == 0x2A42)
 #define IS_G4X(dev)		(INTEL_INFO(dev)->is_g4x)
-#define IS_PINEVIEW_G(dev)	((dev)->pdev->device == 0xa001)
-#define IS_PINEVIEW_M(dev)	((dev)->pdev->device == 0xa011)
+#define IS_PINEVIEW_G(dev)	(INTEL_DEVID(dev) == 0xa001)
+#define IS_PINEVIEW_M(dev)	(INTEL_DEVID(dev) == 0xa011)
 #define IS_PINEVIEW(dev)	(INTEL_INFO(dev)->is_pineview)
 #define IS_G33(dev)		(INTEL_INFO(dev)->is_g33)
-#define IS_IRONLAKE_M(dev)	((dev)->pdev->device == 0x0046)
+#define IS_IRONLAKE_M(dev)	(INTEL_DEVID(dev) == 0x0046)
 #define IS_IVYBRIDGE(dev)	(INTEL_INFO(dev)->is_ivybridge)
-#define IS_IVB_GT1(dev)		((dev)->pdev->device == 0x0156 || \
-				 (dev)->pdev->device == 0x0152 || \
-				 (dev)->pdev->device == 0x015a)
-#define IS_SNB_GT1(dev)		((dev)->pdev->device == 0x0102 || \
-				 (dev)->pdev->device == 0x0106 || \
-				 (dev)->pdev->device == 0x010A)
+#define IS_IVB_GT1(dev)		(INTEL_DEVID(dev) == 0x0156 || \
+				 INTEL_DEVID(dev) == 0x0152 || \
+				 INTEL_DEVID(dev) == 0x015a)
+#define IS_SNB_GT1(dev)		(INTEL_DEVID(dev) == 0x0102 || \
+				 INTEL_DEVID(dev) == 0x0106 || \
+				 INTEL_DEVID(dev) == 0x010A)
 #define IS_VALLEYVIEW(dev)	(INTEL_INFO(dev)->is_valleyview)
 #define IS_CHERRYVIEW(dev)	(INTEL_INFO(dev)->is_valleyview && IS_GEN8(dev))
 #define IS_HASWELL(dev)	(INTEL_INFO(dev)->is_haswell)
 #define IS_BROADWELL(dev)	(!INTEL_INFO(dev)->is_valleyview && IS_GEN8(dev))
+#define IS_SKYLAKE(dev)	(INTEL_INFO(dev)->is_skylake)
 #define IS_MOBILE(dev)		(INTEL_INFO(dev)->is_mobile)
 #define IS_HSW_EARLY_SDV(dev)	(IS_HASWELL(dev) && \
-				 ((dev)->pdev->device & 0xFF00) == 0x0C00)
+				 (INTEL_DEVID(dev) & 0xFF00) == 0x0C00)
 #define IS_BDW_ULT(dev)		(IS_BROADWELL(dev) && \
-				 (((dev)->pdev->device & 0xf) == 0x2  || \
-				 ((dev)->pdev->device & 0xf) == 0x6 || \
-				 ((dev)->pdev->device & 0xf) == 0xe))
+				 ((INTEL_DEVID(dev) & 0xf) == 0x6 ||	\
+				 (INTEL_DEVID(dev) & 0xf) == 0xe))
+#define IS_BDW_GT3(dev)		(IS_BROADWELL(dev) && \
+				 (INTEL_DEVID(dev) & 0x00F0) == 0x0020)
 #define IS_HSW_ULT(dev)		(IS_HASWELL(dev) && \
-				 ((dev)->pdev->device & 0xFF00) == 0x0A00)
-#define IS_ULT(dev)		(IS_HSW_ULT(dev) || IS_BDW_ULT(dev))
+				 (INTEL_DEVID(dev) & 0xFF00) == 0x0A00)
 #define IS_HSW_GT3(dev)		(IS_HASWELL(dev) && \
-				 ((dev)->pdev->device & 0x00F0) == 0x0020)
+				 (INTEL_DEVID(dev) & 0x00F0) == 0x0020)
 /* ULX machines are also considered ULT. */
-#define IS_HSW_ULX(dev)		((dev)->pdev->device == 0x0A0E || \
-				 (dev)->pdev->device == 0x0A1E)
+#define IS_HSW_ULX(dev)		(INTEL_DEVID(dev) == 0x0A0E || \
+				 INTEL_DEVID(dev) == 0x0A1E)
 #define IS_PRELIMINARY_HW(intel_info) ((intel_info)->is_preliminary)
 
 /*
@@ -2031,6 +2185,7 @@ struct drm_i915_cmd_table {
 #define IS_GEN6(dev)	(INTEL_INFO(dev)->gen == 6)
 #define IS_GEN7(dev)	(INTEL_INFO(dev)->gen == 7)
 #define IS_GEN8(dev)	(INTEL_INFO(dev)->gen == 8)
+#define IS_GEN9(dev)	(INTEL_INFO(dev)->gen == 9)
 
 #define RENDER_RING		(1<<RCS)
 #define BSD_RING		(1<<VCS)
@@ -2043,14 +2198,13 @@ struct drm_i915_cmd_table {
 #define HAS_VEBOX(dev)		(INTEL_INFO(dev)->ring_mask & VEBOX_RING)
 #define HAS_LLC(dev)		(INTEL_INFO(dev)->has_llc)
 #define HAS_WT(dev)		((IS_HASWELL(dev) || IS_BROADWELL(dev)) && \
-				 to_i915(dev)->ellc_size)
+				 __I915__(dev)->ellc_size)
 #define I915_NEED_GFX_HWS(dev)	(INTEL_INFO(dev)->need_gfx_hws)
 
 #define HAS_HW_CONTEXTS(dev)	(INTEL_INFO(dev)->gen >= 6)
-#define HAS_ALIASING_PPGTT(dev)	(INTEL_INFO(dev)->gen >= 6)
-#define HAS_PPGTT(dev)		(INTEL_INFO(dev)->gen >= 7 && !IS_GEN8(dev))
-#define USES_PPGTT(dev)		intel_enable_ppgtt(dev, false)
-#define USES_FULL_PPGTT(dev)	intel_enable_ppgtt(dev, true)
+#define HAS_LOGICAL_RING_CONTEXTS(dev)	(INTEL_INFO(dev)->gen >= 8)
+#define USES_PPGTT(dev)		(i915.enable_ppgtt)
+#define USES_FULL_PPGTT(dev)	(i915.enable_ppgtt == 2)
 
 #define HAS_OVERLAY(dev)		(INTEL_INFO(dev)->has_overlay)
 #define OVERLAY_NEEDS_PHYSICAL(dev)	(INTEL_INFO(dev)->overlay_needs_physical)
@@ -2081,13 +2235,15 @@ struct drm_i915_cmd_table {
 #define HAS_PIPE_CXSR(dev) (INTEL_INFO(dev)->has_pipe_cxsr)
 #define HAS_FBC(dev) (INTEL_INFO(dev)->has_fbc)
 
-#define HAS_IPS(dev)		(IS_ULT(dev) || IS_BROADWELL(dev))
+#define HAS_IPS(dev)		(IS_HSW_ULT(dev) || IS_BROADWELL(dev))
 
 #define HAS_DDI(dev)		(INTEL_INFO(dev)->has_ddi)
 #define HAS_FPGA_DBG_UNCLAIMED(dev)	(INTEL_INFO(dev)->has_fpga_dbg)
 #define HAS_PSR(dev)		(IS_HASWELL(dev) || IS_BROADWELL(dev))
 #define HAS_RUNTIME_PM(dev)	(IS_GEN6(dev) || IS_HASWELL(dev) || \
 				 IS_BROADWELL(dev) || IS_VALLEYVIEW(dev))
+#define HAS_RC6(dev)		(INTEL_INFO(dev)->gen >= 6)
+#define HAS_RC6p(dev)		(INTEL_INFO(dev)->gen == 6 || IS_IVYBRIDGE(dev))
 
 #define INTEL_PCH_DEVICE_ID_MASK		0xff00
 #define INTEL_PCH_IBX_DEVICE_ID_TYPE		0x3b00
@@ -2095,8 +2251,11 @@ struct drm_i915_cmd_table {
 #define INTEL_PCH_PPT_DEVICE_ID_TYPE		0x1e00
 #define INTEL_PCH_LPT_DEVICE_ID_TYPE		0x8c00
 #define INTEL_PCH_LPT_LP_DEVICE_ID_TYPE		0x9c00
+#define INTEL_PCH_SPT_DEVICE_ID_TYPE		0xA100
+#define INTEL_PCH_SPT_LP_DEVICE_ID_TYPE		0x9D00
 
-#define INTEL_PCH_TYPE(dev) (to_i915(dev)->pch_type)
+#define INTEL_PCH_TYPE(dev) (__I915__(dev)->pch_type)
+#define HAS_PCH_SPT(dev) (INTEL_PCH_TYPE(dev) == PCH_SPT)
 #define HAS_PCH_LPT(dev) (INTEL_PCH_TYPE(dev) == PCH_LPT)
 #define HAS_PCH_CPT(dev) (INTEL_PCH_TYPE(dev) == PCH_CPT)
 #define HAS_PCH_IBX(dev) (INTEL_PCH_TYPE(dev) == PCH_IBX)
@@ -2116,8 +2275,8 @@ struct drm_i915_cmd_table {
 extern const struct drm_ioctl_desc i915_ioctls[];
 extern int i915_max_ioctl;
 
-extern int i915_suspend(struct drm_device *dev, pm_message_t state);
-extern int i915_resume(struct drm_device *dev);
+extern int i915_suspend_legacy(struct drm_device *dev, pm_message_t state);
+extern int i915_resume_legacy(struct drm_device *dev);
 extern int i915_master_create(struct drm_device *dev, struct drm_master *master);
 extern void i915_master_destroy(struct drm_device *dev, struct drm_master *master);
 
@@ -2134,6 +2293,7 @@ struct i915_params {
 	int enable_rc6;
 	int enable_fbc;
 	int enable_ppgtt;
+	int enable_execlists;
 	int enable_psr;
 	unsigned int preliminary_hw_support;
 	int disable_power_well;
@@ -2153,8 +2313,6 @@ struct i915_params {
 extern struct i915_params i915 __read_mostly;
 
 				/* i915_dma.c */
-void i915_update_dri1_breadcrumb(struct drm_device *dev);
-extern void i915_kernel_lost_context(struct drm_device * dev);
 extern int i915_driver_load(struct drm_device *, unsigned long flags);
 extern int i915_driver_unload(struct drm_device *);
 extern int i915_driver_open(struct drm_device *dev, struct drm_file *file);
@@ -2168,9 +2326,6 @@ extern int i915_driver_device_is_agp(struct drm_device * dev);
 extern long i915_compat_ioctl(struct file *filp, unsigned int cmd,
 			      unsigned long arg);
 #endif
-extern int i915_emit_box(struct drm_device *dev,
-			 struct drm_clip_rect *box,
-			 int DR1, int DR4);
 extern int intel_gpu_reset(struct drm_device *dev);
 extern int i915_reset(struct drm_device *dev);
 extern unsigned long i915_chipset_val(struct drm_i915_private *dev_priv);
@@ -2180,18 +2335,16 @@ extern void i915_update_gfx_val(struct drm_i915_private *dev_priv);
 int vlv_force_gfx_clock(struct drm_i915_private *dev_priv, bool on);
 void intel_hpd_cancel_work(struct drm_i915_private *dev_priv);
 
-extern void intel_console_resume(struct work_struct *work);
-
 /* i915_irq.c */
 void i915_queue_hangcheck(struct drm_device *dev);
 __printf(3, 4)
 void i915_handle_error(struct drm_device *dev, bool wedged,
 		       const char *fmt, ...);
 
-void gen6_set_pm_mask(struct drm_i915_private *dev_priv, u32 pm_iir,
-							int new_delay);
-extern void intel_irq_init(struct drm_device *dev);
-extern void intel_hpd_init(struct drm_device *dev);
+extern void intel_irq_init(struct drm_i915_private *dev_priv);
+extern void intel_hpd_init(struct drm_i915_private *dev_priv);
+int intel_irq_install(struct drm_i915_private *dev_priv);
+void intel_irq_uninstall(struct drm_i915_private *dev_priv);
 
 extern void intel_uncore_sanitize(struct drm_device *dev);
 extern void intel_uncore_early_sanitize(struct drm_device *dev,
@@ -2211,10 +2364,19 @@ i915_disable_pipestat(struct drm_i915_private *dev_priv, enum pipe pipe,
 
 void valleyview_enable_display_irqs(struct drm_i915_private *dev_priv);
 void valleyview_disable_display_irqs(struct drm_i915_private *dev_priv);
+void
+ironlake_enable_display_irq(struct drm_i915_private *dev_priv, u32 mask);
+void
+ironlake_disable_display_irq(struct drm_i915_private *dev_priv, u32 mask);
+void ibx_display_interrupt_update(struct drm_i915_private *dev_priv,
+				  uint32_t interrupt_mask,
+				  uint32_t enabled_irq_mask);
+#define ibx_enable_display_interrupt(dev_priv, bits) \
+	ibx_display_interrupt_update((dev_priv), (bits), (bits))
+#define ibx_disable_display_interrupt(dev_priv, bits) \
+	ibx_display_interrupt_update((dev_priv), (bits), 0)
 
 /* i915_gem.c */
-int i915_gem_init_ioctl(struct drm_device *dev, void *data,
-			struct drm_file *file_priv);
 int i915_gem_create_ioctl(struct drm_device *dev, void *data,
 			  struct drm_file *file_priv);
 int i915_gem_pread_ioctl(struct drm_device *dev, void *data,
@@ -2229,6 +2391,20 @@ int i915_gem_set_domain_ioctl(struct drm_device *dev, void *data,
 			      struct drm_file *file_priv);
 int i915_gem_sw_finish_ioctl(struct drm_device *dev, void *data,
 			     struct drm_file *file_priv);
+void i915_gem_execbuffer_move_to_active(struct list_head *vmas,
+					struct intel_engine_cs *ring);
+void i915_gem_execbuffer_retire_commands(struct drm_device *dev,
+					 struct drm_file *file,
+					 struct intel_engine_cs *ring,
+					 struct drm_i915_gem_object *obj);
+int i915_gem_ringbuffer_submission(struct drm_device *dev,
+				   struct drm_file *file,
+				   struct intel_engine_cs *ring,
+				   struct intel_context *ctx,
+				   struct drm_i915_gem_execbuffer2 *args,
+				   struct list_head *vmas,
+				   struct drm_i915_gem_object *batch_obj,
+				   u64 exec_start, u32 flags);
 int i915_gem_execbuffer(struct drm_device *dev, void *data,
 			struct drm_file *file_priv);
 int i915_gem_execbuffer2(struct drm_device *dev, void *data,
@@ -2247,10 +2423,6 @@ int i915_gem_throttle_ioctl(struct drm_device *dev, void *data,
 			    struct drm_file *file_priv);
 int i915_gem_madvise_ioctl(struct drm_device *dev, void *data,
 			   struct drm_file *file_priv);
-int i915_gem_entervt_ioctl(struct drm_device *dev, void *data,
-			   struct drm_file *file_priv);
-int i915_gem_leavevt_ioctl(struct drm_device *dev, void *data,
-			   struct drm_file *file_priv);
 int i915_gem_set_tiling(struct drm_device *dev, void *data,
 			struct drm_file *file_priv);
 int i915_gem_get_tiling(struct drm_device *dev, void *data,
@@ -2263,6 +2435,12 @@ int i915_gem_get_aperture_ioctl(struct drm_device *dev, void *data,
 int i915_gem_wait_ioctl(struct drm_device *dev, void *data,
 			struct drm_file *file_priv);
 void i915_gem_load(struct drm_device *dev);
+unsigned long i915_gem_shrink(struct drm_i915_private *dev_priv,
+			      long target,
+			      unsigned flags);
+#define I915_SHRINK_PURGEABLE 0x1
+#define I915_SHRINK_UNBOUND 0x2
+#define I915_SHRINK_BOUND 0x4
 void *i915_gem_object_alloc(struct drm_device *dev);
 void i915_gem_object_free(struct drm_i915_gem_object *obj);
 void i915_gem_object_init(struct drm_i915_gem_object *obj,
@@ -2287,7 +2465,6 @@ int __must_check i915_vma_unbind(struct i915_vma *vma);
 int i915_gem_object_put_pages(struct drm_i915_gem_object *obj);
 void i915_gem_release_all_mmaps(struct drm_i915_private *dev_priv);
 void i915_gem_release_mmap(struct drm_i915_gem_object *obj);
-void i915_gem_lastclose(struct drm_device *dev);
 
 int i915_gem_obj_prepare_shmem_read(struct drm_i915_gem_object *obj,
 				    int *needs_clflush);
@@ -2381,6 +2558,7 @@ void i915_gem_reset(struct drm_device *dev);
 bool i915_gem_clflush_object(struct drm_i915_gem_object *obj, bool force);
 int __must_check i915_gem_object_finish_gpu(struct drm_i915_gem_object *obj);
 int __must_check i915_gem_init(struct drm_device *dev);
+int i915_gem_init_rings(struct drm_device *dev);
 int __must_check i915_gem_init_hw(struct drm_device *dev);
 int i915_gem_l3_remap(struct intel_engine_cs *ring, int slice);
 void i915_gem_init_swizzling(struct drm_device *dev);
@@ -2393,6 +2571,11 @@ int __i915_add_request(struct intel_engine_cs *ring,
 		       u32 *seqno);
 #define i915_add_request(ring, seqno) \
 	__i915_add_request(ring, NULL, NULL, seqno)
+int __i915_wait_seqno(struct intel_engine_cs *ring, u32 seqno,
+			unsigned reset_counter,
+			bool interruptible,
+			s64 *timeout,
+			struct drm_i915_file_private *file_priv);
 int __must_check i915_wait_seqno(struct intel_engine_cs *ring,
 				 uint32_t seqno);
 int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
@@ -2451,7 +2634,7 @@ static inline bool i915_gem_obj_is_pinned(struct drm_i915_gem_object *obj) {
 }
 
 /* Some GGTT VM helpers */
-#define obj_to_ggtt(obj) \
+#define i915_obj_to_ggtt(obj) \
 	(&((struct drm_i915_private *)(obj)->base.dev->dev_private)->gtt.base)
 static inline bool i915_is_ggtt(struct i915_address_space *vm)
 {
@@ -2460,21 +2643,30 @@ static inline bool i915_is_ggtt(struct i915_address_space *vm)
 	return vm == ggtt;
 }
 
+static inline struct i915_hw_ppgtt *
+i915_vm_to_ppgtt(struct i915_address_space *vm)
+{
+	WARN_ON(i915_is_ggtt(vm));
+
+	return container_of(vm, struct i915_hw_ppgtt, base);
+}
+
+
 static inline bool i915_gem_obj_ggtt_bound(struct drm_i915_gem_object *obj)
 {
-	return i915_gem_obj_bound(obj, obj_to_ggtt(obj));
+	return i915_gem_obj_bound(obj, i915_obj_to_ggtt(obj));
 }
 
 static inline unsigned long
 i915_gem_obj_ggtt_offset(struct drm_i915_gem_object *obj)
 {
-	return i915_gem_obj_offset(obj, obj_to_ggtt(obj));
+	return i915_gem_obj_offset(obj, i915_obj_to_ggtt(obj));
 }
 
 static inline unsigned long
 i915_gem_obj_ggtt_size(struct drm_i915_gem_object *obj)
 {
-	return i915_gem_obj_size(obj, obj_to_ggtt(obj));
+	return i915_gem_obj_size(obj, i915_obj_to_ggtt(obj));
 }
 
 static inline int __must_check
@@ -2482,7 +2674,8 @@ i915_gem_obj_ggtt_pin(struct drm_i915_gem_object *obj,
 		      uint32_t alignment,
 		      unsigned flags)
 {
-	return i915_gem_object_pin(obj, obj_to_ggtt(obj), alignment, flags | PIN_GLOBAL);
+	return i915_gem_object_pin(obj, i915_obj_to_ggtt(obj),
+				   alignment, flags | PIN_GLOBAL);
 }
 
 static inline int
@@ -2494,7 +2687,6 @@ i915_gem_object_ggtt_unbind(struct drm_i915_gem_object *obj)
 void i915_gem_object_ggtt_unpin(struct drm_i915_gem_object *obj);
 
 /* i915_gem_context.c */
-#define ctx_to_ppgtt(ctx) container_of((ctx)->vm, struct i915_hw_ppgtt, base)
 int __must_check i915_gem_context_init(struct drm_device *dev);
 void i915_gem_context_fini(struct drm_device *dev);
 void i915_gem_context_reset(struct drm_device *dev);
@@ -2506,6 +2698,8 @@ int i915_switch_context(struct intel_engine_cs *ring,
 struct intel_context *
 i915_gem_context_get(struct drm_i915_file_private *file_priv, u32 id);
 void i915_gem_context_free(struct kref *ctx_ref);
+struct drm_i915_gem_object *
+i915_gem_alloc_context_obj(struct drm_device *dev, size_t size);
 static inline void i915_gem_context_reference(struct intel_context *ctx)
 {
 	kref_get(&ctx->ref);
@@ -2526,8 +2720,6 @@ int i915_gem_context_create_ioctl(struct drm_device *dev, void *data,
 int i915_gem_context_destroy_ioctl(struct drm_device *dev, void *data,
 				   struct drm_file *file);
 
-/* i915_gem_render_state.c */
-int i915_gem_render_state_init(struct intel_engine_cs *ring);
 /* i915_gem_evict.c */
 int __must_check i915_gem_evict_something(struct drm_device *dev,
 					  struct i915_address_space *vm,
@@ -2595,6 +2787,7 @@ void i915_error_printf(struct drm_i915_error_state_buf *e, const char *f, ...);
 int i915_error_state_to_str(struct drm_i915_error_state_buf *estr,
 			    const struct i915_error_state_file_priv *error);
 int i915_error_state_buf_init(struct drm_i915_error_state_buf *eb,
+			      struct drm_i915_private *i915,
 			      size_t count, loff_t pos);
 static inline void i915_error_state_buf_release(
 	struct drm_i915_error_state_buf *eb)
@@ -2609,7 +2802,7 @@ void i915_error_state_put(struct i915_error_state_file_priv *error_priv);
 void i915_destroy_error_state(struct drm_device *dev);
 
 void i915_get_extra_instdone(struct drm_device *dev, uint32_t *instdone);
-const char *i915_cache_level_str(int type);
+const char *i915_cache_level_str(struct drm_i915_private *i915, int type);
 
 /* i915_cmd_parser.c */
 int i915_cmd_parser_get_version(void);
@@ -2652,7 +2845,6 @@ static inline bool intel_gmbus_is_forced_bit(struct i2c_adapter *adapter)
 extern void intel_i2c_reset(struct drm_device *dev);
 
 /* intel_opregion.c */
-struct intel_encoder;
 #ifdef CONFIG_ACPI
 extern int intel_opregion_setup(struct drm_device *dev);
 extern void intel_opregion_init(struct drm_device *dev);
@@ -2690,7 +2882,6 @@ static inline void intel_unregister_dsm_handler(void) { return; }
 
 /* modesetting */
 extern void intel_modeset_init_hw(struct drm_device *dev);
-extern void intel_modeset_suspend_hw(struct drm_device *dev);
 extern void intel_modeset_init(struct drm_device *dev);
 extern void intel_modeset_gem_init(struct drm_device *dev);
 extern void intel_modeset_cleanup(struct drm_device *dev);
@@ -2701,6 +2892,7 @@ extern void intel_modeset_setup_hw_state(struct drm_device *dev,
 extern void i915_redisable_vga(struct drm_device *dev);
 extern void i915_redisable_vga_power_on(struct drm_device *dev);
 extern bool intel_fbc_enabled(struct drm_device *dev);
+extern void bdw_fbc_sw_flush(struct drm_device *dev, u32 value);
 extern void intel_disable_fbc(struct drm_device *dev);
 extern bool ironlake_set_drps(struct drm_device *dev, u8 val);
 extern void intel_init_pch_refclk(struct drm_device *dev);
@@ -2738,8 +2930,8 @@ void gen6_gt_force_wake_get(struct drm_i915_private *dev_priv, int fw_engine);
 void gen6_gt_force_wake_put(struct drm_i915_private *dev_priv, int fw_engine);
 void assert_force_wake_inactive(struct drm_i915_private *dev_priv);
 
-int sandybridge_pcode_read(struct drm_i915_private *dev_priv, u8 mbox, u32 *val);
-int sandybridge_pcode_write(struct drm_i915_private *dev_priv, u8 mbox, u32 val);
+int sandybridge_pcode_read(struct drm_i915_private *dev_priv, u32 mbox, u32 *val);
+int sandybridge_pcode_write(struct drm_i915_private *dev_priv, u32 mbox, u32 val);
 
 /* intel_sideband.c */
 u32 vlv_punit_read(struct drm_i915_private *dev_priv, u8 addr);
@@ -2769,7 +2961,9 @@ int vlv_freq_opcode(struct drm_i915_private *dev_priv, int val);
 
 #define FORCEWAKE_RENDER	(1 << 0)
 #define FORCEWAKE_MEDIA		(1 << 1)
-#define FORCEWAKE_ALL		(FORCEWAKE_RENDER | FORCEWAKE_MEDIA)
+#define FORCEWAKE_BLITTER	(1 << 2)
+#define FORCEWAKE_ALL		(FORCEWAKE_RENDER | FORCEWAKE_MEDIA | \
+					FORCEWAKE_BLITTER)
 
 
 #define I915_READ8(reg)		dev_priv->uncore.funcs.mmio_readb(dev_priv, (reg), true)
@@ -2835,6 +3029,11 @@ static inline unsigned long msecs_to_jiffies_timeout(const unsigned int m)
 	return min_t(unsigned long, MAX_JIFFY_OFFSET, j + 1);
 }
 
+static inline unsigned long nsecs_to_jiffies_timeout(const u64 n)
+{
+        return min_t(u64, MAX_JIFFY_OFFSET, nsecs_to_jiffies64(n) + 1);
+}
+
 static inline unsigned long
 timespec_to_jiffies_timeout(const struct timespec *value)
 {
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index ad55b06a3cb1..5f614828d365 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -60,7 +60,6 @@ static unsigned long i915_gem_shrinker_scan(struct shrinker *shrinker,
 static int i915_gem_shrinker_oom(struct notifier_block *nb,
 				 unsigned long event,
 				 void *ptr);
-static unsigned long i915_gem_purge(struct drm_i915_private *dev_priv, long target);
 static unsigned long i915_gem_shrink_all(struct drm_i915_private *dev_priv);
 
 static bool cpu_cache_is_coherent(struct drm_device *dev,
@@ -161,33 +160,6 @@ i915_gem_object_is_inactive(struct drm_i915_gem_object *obj)
 }
 
 int
-i915_gem_init_ioctl(struct drm_device *dev, void *data,
-		    struct drm_file *file)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_i915_gem_init *args = data;
-
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		return -ENODEV;
-
-	if (args->gtt_start >= args->gtt_end ||
-	    (args->gtt_end | args->gtt_start) & (PAGE_SIZE - 1))
-		return -EINVAL;
-
-	/* GEM with user mode setting was never supported on ilk and later. */
-	if (INTEL_INFO(dev)->gen >= 5)
-		return -ENODEV;
-
-	mutex_lock(&dev->struct_mutex);
-	i915_gem_setup_global_gtt(dev, args->gtt_start, args->gtt_end,
-				  args->gtt_end);
-	dev_priv->gtt.mappable_end = args->gtt_end;
-	mutex_unlock(&dev->struct_mutex);
-
-	return 0;
-}
-
-int
 i915_gem_get_aperture_ioctl(struct drm_device *dev, void *data,
 			    struct drm_file *file)
 {
@@ -209,40 +181,137 @@ i915_gem_get_aperture_ioctl(struct drm_device *dev, void *data,
 	return 0;
 }
 
-static void i915_gem_object_detach_phys(struct drm_i915_gem_object *obj)
+static int
+i915_gem_object_get_pages_phys(struct drm_i915_gem_object *obj)
 {
-	drm_dma_handle_t *phys = obj->phys_handle;
+	struct address_space *mapping = file_inode(obj->base.filp)->i_mapping;
+	char *vaddr = obj->phys_handle->vaddr;
+	struct sg_table *st;
+	struct scatterlist *sg;
+	int i;
 
-	if (!phys)
-		return;
+	if (WARN_ON(i915_gem_object_needs_bit17_swizzle(obj)))
+		return -EINVAL;
+
+	for (i = 0; i < obj->base.size / PAGE_SIZE; i++) {
+		struct page *page;
+		char *src;
+
+		page = shmem_read_mapping_page(mapping, i);
+		if (IS_ERR(page))
+			return PTR_ERR(page);
+
+		src = kmap_atomic(page);
+		memcpy(vaddr, src, PAGE_SIZE);
+		drm_clflush_virt_range(vaddr, PAGE_SIZE);
+		kunmap_atomic(src);
+
+		page_cache_release(page);
+		vaddr += PAGE_SIZE;
+	}
+
+	i915_gem_chipset_flush(obj->base.dev);
+
+	st = kmalloc(sizeof(*st), GFP_KERNEL);
+	if (st == NULL)
+		return -ENOMEM;
+
+	if (sg_alloc_table(st, 1, GFP_KERNEL)) {
+		kfree(st);
+		return -ENOMEM;
+	}
+
+	sg = st->sgl;
+	sg->offset = 0;
+	sg->length = obj->base.size;
 
-	if (obj->madv == I915_MADV_WILLNEED) {
+	sg_dma_address(sg) = obj->phys_handle->busaddr;
+	sg_dma_len(sg) = obj->base.size;
+
+	obj->pages = st;
+	obj->has_dma_mapping = true;
+	return 0;
+}
+
+static void
+i915_gem_object_put_pages_phys(struct drm_i915_gem_object *obj)
+{
+	int ret;
+
+	BUG_ON(obj->madv == __I915_MADV_PURGED);
+
+	ret = i915_gem_object_set_to_cpu_domain(obj, true);
+	if (ret) {
+		/* In the event of a disaster, abandon all caches and
+		 * hope for the best.
+		 */
+		WARN_ON(ret != -EIO);
+		obj->base.read_domains = obj->base.write_domain = I915_GEM_DOMAIN_CPU;
+	}
+
+	if (obj->madv == I915_MADV_DONTNEED)
+		obj->dirty = 0;
+
+	if (obj->dirty) {
 		struct address_space *mapping = file_inode(obj->base.filp)->i_mapping;
-		char *vaddr = phys->vaddr;
+		char *vaddr = obj->phys_handle->vaddr;
 		int i;
 
 		for (i = 0; i < obj->base.size / PAGE_SIZE; i++) {
-			struct page *page = shmem_read_mapping_page(mapping, i);
-			if (!IS_ERR(page)) {
-				char *dst = kmap_atomic(page);
-				memcpy(dst, vaddr, PAGE_SIZE);
-				drm_clflush_virt_range(dst, PAGE_SIZE);
-				kunmap_atomic(dst);
-
-				set_page_dirty(page);
+			struct page *page;
+			char *dst;
+
+			page = shmem_read_mapping_page(mapping, i);
+			if (IS_ERR(page))
+				continue;
+
+			dst = kmap_atomic(page);
+			drm_clflush_virt_range(vaddr, PAGE_SIZE);
+			memcpy(dst, vaddr, PAGE_SIZE);
+			kunmap_atomic(dst);
+
+			set_page_dirty(page);
+			if (obj->madv == I915_MADV_WILLNEED)
 				mark_page_accessed(page);
-				page_cache_release(page);
-			}
+			page_cache_release(page);
 			vaddr += PAGE_SIZE;
 		}
-		i915_gem_chipset_flush(obj->base.dev);
+		obj->dirty = 0;
 	}
 
-#ifdef CONFIG_X86
-	set_memory_wb((unsigned long)phys->vaddr, phys->size / PAGE_SIZE);
-#endif
-	drm_pci_free(obj->base.dev, phys);
-	obj->phys_handle = NULL;
+	sg_free_table(obj->pages);
+	kfree(obj->pages);
+
+	obj->has_dma_mapping = false;
+}
+
+static void
+i915_gem_object_release_phys(struct drm_i915_gem_object *obj)
+{
+	drm_pci_free(obj->base.dev, obj->phys_handle);
+}
+
+static const struct drm_i915_gem_object_ops i915_gem_phys_ops = {
+	.get_pages = i915_gem_object_get_pages_phys,
+	.put_pages = i915_gem_object_put_pages_phys,
+	.release = i915_gem_object_release_phys,
+};
+
+static int
+drop_pages(struct drm_i915_gem_object *obj)
+{
+	struct i915_vma *vma, *next;
+	int ret;
+
+	drm_gem_object_reference(&obj->base);
+	list_for_each_entry_safe(vma, next, &obj->vma_list, vma_link)
+		if (i915_vma_unbind(vma))
+			break;
+
+	ret = i915_gem_object_put_pages(obj);
+	drm_gem_object_unreference(&obj->base);
+
+	return ret;
 }
 
 int
@@ -250,9 +319,7 @@ i915_gem_object_attach_phys(struct drm_i915_gem_object *obj,
 			    int align)
 {
 	drm_dma_handle_t *phys;
-	struct address_space *mapping;
-	char *vaddr;
-	int i;
+	int ret;
 
 	if (obj->phys_handle) {
 		if ((unsigned long)obj->phys_handle->vaddr & (align -1))
@@ -267,41 +334,19 @@ i915_gem_object_attach_phys(struct drm_i915_gem_object *obj,
 	if (obj->base.filp == NULL)
 		return -EINVAL;
 
+	ret = drop_pages(obj);
+	if (ret)
+		return ret;
+
 	/* create a new object */
 	phys = drm_pci_alloc(obj->base.dev, obj->base.size, align);
 	if (!phys)
 		return -ENOMEM;
 
-	vaddr = phys->vaddr;
-#ifdef CONFIG_X86
-	set_memory_wc((unsigned long)vaddr, phys->size / PAGE_SIZE);
-#endif
-	mapping = file_inode(obj->base.filp)->i_mapping;
-	for (i = 0; i < obj->base.size / PAGE_SIZE; i++) {
-		struct page *page;
-		char *src;
-
-		page = shmem_read_mapping_page(mapping, i);
-		if (IS_ERR(page)) {
-#ifdef CONFIG_X86
-			set_memory_wb((unsigned long)phys->vaddr, phys->size / PAGE_SIZE);
-#endif
-			drm_pci_free(obj->base.dev, phys);
-			return PTR_ERR(page);
-		}
-
-		src = kmap_atomic(page);
-		memcpy(vaddr, src, PAGE_SIZE);
-		kunmap_atomic(src);
-
-		mark_page_accessed(page);
-		page_cache_release(page);
-
-		vaddr += PAGE_SIZE;
-	}
-
 	obj->phys_handle = phys;
-	return 0;
+	obj->ops = &i915_gem_phys_ops;
+
+	return i915_gem_object_get_pages(obj);
 }
 
 static int
@@ -312,6 +357,14 @@ i915_gem_phys_pwrite(struct drm_i915_gem_object *obj,
 	struct drm_device *dev = obj->base.dev;
 	void *vaddr = obj->phys_handle->vaddr + args->offset;
 	char __user *user_data = to_user_ptr(args->data_ptr);
+	int ret;
+
+	/* We manually control the domain here and pretend that it
+	 * remains coherent i.e. in the GTT domain, like shmem_pwrite.
+	 */
+	ret = i915_gem_object_wait_rendering(obj, false);
+	if (ret)
+		return ret;
 
 	if (__copy_from_user_inatomic_nocache(vaddr, user_data, args->size)) {
 		unsigned long unwritten;
@@ -327,6 +380,7 @@ i915_gem_phys_pwrite(struct drm_i915_gem_object *obj,
 			return -EFAULT;
 	}
 
+	drm_clflush_virt_range(vaddr, args->size);
 	i915_gem_chipset_flush(dev);
 	return 0;
 }
@@ -994,6 +1048,7 @@ int
 i915_gem_pwrite_ioctl(struct drm_device *dev, void *data,
 		      struct drm_file *file)
 {
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_gem_pwrite *args = data;
 	struct drm_i915_gem_object *obj;
 	int ret;
@@ -1013,9 +1068,11 @@ i915_gem_pwrite_ioctl(struct drm_device *dev, void *data,
 			return -EFAULT;
 	}
 
+	intel_runtime_pm_get(dev_priv);
+
 	ret = i915_mutex_lock_interruptible(dev);
 	if (ret)
-		return ret;
+		goto put_rpm;
 
 	obj = to_intel_bo(drm_gem_object_lookup(dev, file, args->handle));
 	if (&obj->base == NULL) {
@@ -1047,11 +1104,6 @@ i915_gem_pwrite_ioctl(struct drm_device *dev, void *data,
 	 * pread/pwrite currently are reading and writing from the CPU
 	 * perspective, requiring manual detiling by the client.
 	 */
-	if (obj->phys_handle) {
-		ret = i915_gem_phys_pwrite(obj, args, file);
-		goto out;
-	}
-
 	if (obj->tiling_mode == I915_TILING_NONE &&
 	    obj->base.write_domain != I915_GEM_DOMAIN_CPU &&
 	    cpu_write_needs_clflush(obj)) {
@@ -1061,13 +1113,20 @@ i915_gem_pwrite_ioctl(struct drm_device *dev, void *data,
 		 * textures). Fallback to the shmem path in that case. */
 	}
 
-	if (ret == -EFAULT || ret == -ENOSPC)
-		ret = i915_gem_shmem_pwrite(dev, obj, args, file);
+	if (ret == -EFAULT || ret == -ENOSPC) {
+		if (obj->phys_handle)
+			ret = i915_gem_phys_pwrite(obj, args, file);
+		else
+			ret = i915_gem_shmem_pwrite(dev, obj, args, file);
+	}
 
 out:
 	drm_gem_object_unreference(&obj->base);
 unlock:
 	mutex_unlock(&dev->struct_mutex);
+put_rpm:
+	intel_runtime_pm_put(dev_priv);
+
 	return ret;
 }
 
@@ -1085,7 +1144,13 @@ i915_gem_check_wedge(struct i915_gpu_error *error,
 		if (i915_terminally_wedged(error))
 			return -EIO;
 
-		return -EAGAIN;
+		/*
+		 * Check if GPU Reset is in progress - we need intel_ring_begin
+		 * to work properly to reinit the hw state while the gpu is
+		 * still marked as reset-in-progress. Handle this with a flag.
+		 */
+		if (!error->reload_in_reset)
+			return -EAGAIN;
 	}
 
 	return 0;
@@ -1129,7 +1194,7 @@ static bool can_wait_boost(struct drm_i915_file_private *file_priv)
 }
 
 /**
- * __wait_seqno - wait until execution of seqno has finished
+ * __i915_wait_seqno - wait until execution of seqno has finished
  * @ring: the ring expected to report seqno
  * @seqno: duh!
  * @reset_counter: reset sequence associated with the given seqno
@@ -1146,7 +1211,7 @@ static bool can_wait_boost(struct drm_i915_file_private *file_priv)
  * Returns 0 if the seqno was found within the alloted time. Else returns the
  * errno with remaining time filled in timeout argument.
  */
-static int __wait_seqno(struct intel_engine_cs *ring, u32 seqno,
+int __i915_wait_seqno(struct intel_engine_cs *ring, u32 seqno,
 			unsigned reset_counter,
 			bool interruptible,
 			s64 *timeout,
@@ -1166,7 +1231,8 @@ static int __wait_seqno(struct intel_engine_cs *ring, u32 seqno,
 	if (i915_seqno_passed(ring->get_seqno(ring, true), seqno))
 		return 0;
 
-	timeout_expire = timeout ? jiffies + nsecs_to_jiffies((u64)*timeout) : 0;
+	timeout_expire = timeout ?
+		jiffies + nsecs_to_jiffies_timeout((u64)*timeout) : 0;
 
 	if (INTEL_INFO(dev)->gen >= 6 && ring->id == RCS && can_wait_boost(file_priv)) {
 		gen6_rps_boost(dev_priv);
@@ -1242,6 +1308,16 @@ static int __wait_seqno(struct intel_engine_cs *ring, u32 seqno,
 		s64 tres = *timeout - (now - before);
 
 		*timeout = tres < 0 ? 0 : tres;
+
+		/*
+		 * Apparently ktime isn't accurate enough and occasionally has a
+		 * bit of mismatch in the jiffies<->nsecs<->ktime loop. So patch
+		 * things up to make the test happy. We allow up to 1 jiffy.
+		 *
+		 * This is a regrssion from the timespec->ktime conversion.
+		 */
+		if (ret == -ETIME && *timeout < jiffies_to_usecs(1)*1000)
+			*timeout = 0;
 	}
 
 	return ret;
@@ -1257,6 +1333,7 @@ i915_wait_seqno(struct intel_engine_cs *ring, uint32_t seqno)
 	struct drm_device *dev = ring->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	bool interruptible = dev_priv->mm.interruptible;
+	unsigned reset_counter;
 	int ret;
 
 	BUG_ON(!mutex_is_locked(&dev->struct_mutex));
@@ -1270,14 +1347,13 @@ i915_wait_seqno(struct intel_engine_cs *ring, uint32_t seqno)
 	if (ret)
 		return ret;
 
-	return __wait_seqno(ring, seqno,
-			    atomic_read(&dev_priv->gpu_error.reset_counter),
-			    interruptible, NULL, NULL);
+	reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter);
+	return __i915_wait_seqno(ring, seqno, reset_counter, interruptible,
+				 NULL, NULL);
 }
 
 static int
-i915_gem_object_wait_rendering__tail(struct drm_i915_gem_object *obj,
-				     struct intel_engine_cs *ring)
+i915_gem_object_wait_rendering__tail(struct drm_i915_gem_object *obj)
 {
 	if (!obj->active)
 		return 0;
@@ -1314,7 +1390,7 @@ i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj,
 	if (ret)
 		return ret;
 
-	return i915_gem_object_wait_rendering__tail(obj, ring);
+	return i915_gem_object_wait_rendering__tail(obj);
 }
 
 /* A nonblocking variant of the above wait. This is a highly dangerous routine
@@ -1349,12 +1425,13 @@ i915_gem_object_wait_rendering__nonblocking(struct drm_i915_gem_object *obj,
 
 	reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter);
 	mutex_unlock(&dev->struct_mutex);
-	ret = __wait_seqno(ring, seqno, reset_counter, true, NULL, file_priv);
+	ret = __i915_wait_seqno(ring, seqno, reset_counter, true, NULL,
+				file_priv);
 	mutex_lock(&dev->struct_mutex);
 	if (ret)
 		return ret;
 
-	return i915_gem_object_wait_rendering__tail(obj, ring);
+	return i915_gem_object_wait_rendering__tail(obj);
 }
 
 /**
@@ -1461,6 +1538,16 @@ unlock:
  *
  * While the mapping holds a reference on the contents of the object, it doesn't
  * imply a ref on the object itself.
+ *
+ * IMPORTANT:
+ *
+ * DRM driver writers who look a this function as an example for how to do GEM
+ * mmap support, please don't implement mmap support like here. The modern way
+ * to implement DRM mmap support is with an mmap offset ioctl (like
+ * i915_gem_mmap_gtt) and then using the mmap syscall on the DRM fd directly.
+ * That way debug tooling like valgrind will understand what's going on, hiding
+ * the mmap call in a driver private ioctl will break that. The i915 driver only
+ * does cpu mmaps this way because we didn't know better.
  */
 int
 i915_gem_mmap_ioctl(struct drm_device *dev, void *data,
@@ -1735,7 +1822,11 @@ static int i915_gem_object_create_mmap_offset(struct drm_i915_gem_object *obj)
 	 * offsets on purgeable objects by truncating it and marking it purged,
 	 * which prevents userspace from ever using that object again.
 	 */
-	i915_gem_purge(dev_priv, obj->base.size >> PAGE_SHIFT);
+	i915_gem_shrink(dev_priv,
+			obj->base.size >> PAGE_SHIFT,
+			I915_SHRINK_BOUND |
+			I915_SHRINK_UNBOUND |
+			I915_SHRINK_PURGEABLE);
 	ret = drm_gem_create_mmap_offset(&obj->base);
 	if (ret != -ENOSPC)
 		goto out;
@@ -1932,12 +2023,18 @@ i915_gem_object_put_pages(struct drm_i915_gem_object *obj)
 	return 0;
 }
 
-static unsigned long
-__i915_gem_shrink(struct drm_i915_private *dev_priv, long target,
-		  bool purgeable_only)
+unsigned long
+i915_gem_shrink(struct drm_i915_private *dev_priv,
+		long target, unsigned flags)
 {
-	struct list_head still_in_list;
-	struct drm_i915_gem_object *obj;
+	const struct {
+		struct list_head *list;
+		unsigned int bit;
+	} phases[] = {
+		{ &dev_priv->mm.unbound_list, I915_SHRINK_UNBOUND },
+		{ &dev_priv->mm.bound_list, I915_SHRINK_BOUND },
+		{ NULL, 0 },
+	}, *phase;
 	unsigned long count = 0;
 
 	/*
@@ -1959,62 +2056,50 @@ __i915_gem_shrink(struct drm_i915_private *dev_priv, long target,
 	 * dev->struct_mutex and so we won't ever be able to observe an
 	 * object on the bound_list with a reference count equals 0.
 	 */
-	INIT_LIST_HEAD(&still_in_list);
-	while (count < target && !list_empty(&dev_priv->mm.unbound_list)) {
-		obj = list_first_entry(&dev_priv->mm.unbound_list,
-				       typeof(*obj), global_list);
-		list_move_tail(&obj->global_list, &still_in_list);
+	for (phase = phases; phase->list; phase++) {
+		struct list_head still_in_list;
 
-		if (!i915_gem_object_is_purgeable(obj) && purgeable_only)
+		if ((flags & phase->bit) == 0)
 			continue;
 
-		drm_gem_object_reference(&obj->base);
-
-		if (i915_gem_object_put_pages(obj) == 0)
-			count += obj->base.size >> PAGE_SHIFT;
-
-		drm_gem_object_unreference(&obj->base);
-	}
-	list_splice(&still_in_list, &dev_priv->mm.unbound_list);
-
-	INIT_LIST_HEAD(&still_in_list);
-	while (count < target && !list_empty(&dev_priv->mm.bound_list)) {
-		struct i915_vma *vma, *v;
+		INIT_LIST_HEAD(&still_in_list);
+		while (count < target && !list_empty(phase->list)) {
+			struct drm_i915_gem_object *obj;
+			struct i915_vma *vma, *v;
 
-		obj = list_first_entry(&dev_priv->mm.bound_list,
-				       typeof(*obj), global_list);
-		list_move_tail(&obj->global_list, &still_in_list);
+			obj = list_first_entry(phase->list,
+					       typeof(*obj), global_list);
+			list_move_tail(&obj->global_list, &still_in_list);
 
-		if (!i915_gem_object_is_purgeable(obj) && purgeable_only)
-			continue;
+			if (flags & I915_SHRINK_PURGEABLE &&
+			    !i915_gem_object_is_purgeable(obj))
+				continue;
 
-		drm_gem_object_reference(&obj->base);
+			drm_gem_object_reference(&obj->base);
 
-		list_for_each_entry_safe(vma, v, &obj->vma_list, vma_link)
-			if (i915_vma_unbind(vma))
-				break;
+			/* For the unbound phase, this should be a no-op! */
+			list_for_each_entry_safe(vma, v,
+						 &obj->vma_list, vma_link)
+				if (i915_vma_unbind(vma))
+					break;
 
-		if (i915_gem_object_put_pages(obj) == 0)
-			count += obj->base.size >> PAGE_SHIFT;
+			if (i915_gem_object_put_pages(obj) == 0)
+				count += obj->base.size >> PAGE_SHIFT;
 
-		drm_gem_object_unreference(&obj->base);
+			drm_gem_object_unreference(&obj->base);
+		}
+		list_splice(&still_in_list, phase->list);
 	}
-	list_splice(&still_in_list, &dev_priv->mm.bound_list);
 
 	return count;
 }
 
 static unsigned long
-i915_gem_purge(struct drm_i915_private *dev_priv, long target)
-{
-	return __i915_gem_shrink(dev_priv, target, true);
-}
-
-static unsigned long
 i915_gem_shrink_all(struct drm_i915_private *dev_priv)
 {
 	i915_gem_evict_everything(dev_priv->dev);
-	return __i915_gem_shrink(dev_priv, LONG_MAX, false);
+	return i915_gem_shrink(dev_priv, LONG_MAX,
+			       I915_SHRINK_BOUND | I915_SHRINK_UNBOUND);
 }
 
 static int
@@ -2061,7 +2146,11 @@ i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj)
 	for (i = 0; i < page_count; i++) {
 		page = shmem_read_mapping_page_gfp(mapping, i, gfp);
 		if (IS_ERR(page)) {
-			i915_gem_purge(dev_priv, page_count);
+			i915_gem_shrink(dev_priv,
+					page_count,
+					I915_SHRINK_BOUND |
+					I915_SHRINK_UNBOUND |
+					I915_SHRINK_PURGEABLE);
 			page = shmem_read_mapping_page_gfp(mapping, i, gfp);
 		}
 		if (IS_ERR(page)) {
@@ -2104,6 +2193,10 @@ i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj)
 	if (i915_gem_object_needs_bit17_swizzle(obj))
 		i915_gem_object_do_bit_17_swizzle(obj);
 
+	if (obj->tiling_mode != I915_TILING_NONE &&
+	    dev_priv->quirks & QUIRK_PIN_SWIZZLED_PAGES)
+		i915_gem_object_pin_pages(obj);
+
 	return 0;
 
 err_pages:
@@ -2163,8 +2256,6 @@ static void
 i915_gem_object_move_to_active(struct drm_i915_gem_object *obj,
 			       struct intel_engine_cs *ring)
 {
-	struct drm_device *dev = obj->base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 seqno = intel_ring_get_seqno(ring);
 
 	BUG_ON(ring == NULL);
@@ -2183,19 +2274,6 @@ i915_gem_object_move_to_active(struct drm_i915_gem_object *obj,
 	list_move_tail(&obj->ring_list, &ring->active_list);
 
 	obj->last_read_seqno = seqno;
-
-	if (obj->fenced_gpu_access) {
-		obj->last_fenced_seqno = seqno;
-
-		/* Bump MRU to take account of the delayed flush */
-		if (obj->fence_reg != I915_FENCE_REG_NONE) {
-			struct drm_i915_fence_reg *reg;
-
-			reg = &dev_priv->fence_regs[obj->fence_reg];
-			list_move_tail(&reg->lru_list,
-				       &dev_priv->mm.fence_list);
-		}
-	}
 }
 
 void i915_vma_move_to_active(struct i915_vma *vma,
@@ -2231,7 +2309,6 @@ i915_gem_object_move_to_inactive(struct drm_i915_gem_object *obj)
 	obj->base.write_domain = 0;
 
 	obj->last_fenced_seqno = 0;
-	obj->fenced_gpu_access = false;
 
 	obj->active = 0;
 	drm_gem_object_unreference(&obj->base);
@@ -2329,10 +2406,21 @@ int __i915_add_request(struct intel_engine_cs *ring,
 {
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 	struct drm_i915_gem_request *request;
+	struct intel_ringbuffer *ringbuf;
 	u32 request_ring_position, request_start;
 	int ret;
 
-	request_start = intel_ring_get_tail(ring->buffer);
+	request = ring->preallocated_lazy_request;
+	if (WARN_ON(request == NULL))
+		return -ENOMEM;
+
+	if (i915.enable_execlists) {
+		struct intel_context *ctx = request->ctx;
+		ringbuf = ctx->engine[ring->id].ringbuf;
+	} else
+		ringbuf = ring->buffer;
+
+	request_start = intel_ring_get_tail(ringbuf);
 	/*
 	 * Emit any outstanding flushes - execbuf can fail to emit the flush
 	 * after having emitted the batchbuffer command. Hence we need to fix
@@ -2340,24 +2428,32 @@ int __i915_add_request(struct intel_engine_cs *ring,
 	 * is that the flush _must_ happen before the next request, no matter
 	 * what.
 	 */
-	ret = intel_ring_flush_all_caches(ring);
-	if (ret)
-		return ret;
-
-	request = ring->preallocated_lazy_request;
-	if (WARN_ON(request == NULL))
-		return -ENOMEM;
+	if (i915.enable_execlists) {
+		ret = logical_ring_flush_all_caches(ringbuf);
+		if (ret)
+			return ret;
+	} else {
+		ret = intel_ring_flush_all_caches(ring);
+		if (ret)
+			return ret;
+	}
 
 	/* Record the position of the start of the request so that
 	 * should we detect the updated seqno part-way through the
 	 * GPU processing the request, we never over-estimate the
 	 * position of the head.
 	 */
-	request_ring_position = intel_ring_get_tail(ring->buffer);
+	request_ring_position = intel_ring_get_tail(ringbuf);
 
-	ret = ring->add_request(ring);
-	if (ret)
-		return ret;
+	if (i915.enable_execlists) {
+		ret = ring->emit_request(ringbuf);
+		if (ret)
+			return ret;
+	} else {
+		ret = ring->add_request(ring);
+		if (ret)
+			return ret;
+	}
 
 	request->seqno = intel_ring_get_seqno(ring);
 	request->ring = ring;
@@ -2372,12 +2468,14 @@ int __i915_add_request(struct intel_engine_cs *ring,
 	 */
 	request->batch_obj = obj;
 
-	/* Hold a reference to the current context so that we can inspect
-	 * it later in case a hangcheck error event fires.
-	 */
-	request->ctx = ring->last_context;
-	if (request->ctx)
-		i915_gem_context_reference(request->ctx);
+	if (!i915.enable_execlists) {
+		/* Hold a reference to the current context so that we can inspect
+		 * it later in case a hangcheck error event fires.
+		 */
+		request->ctx = ring->last_context;
+		if (request->ctx)
+			i915_gem_context_reference(request->ctx);
+	}
 
 	request->emitted_jiffies = jiffies;
 	list_add_tail(&request->list, &ring->request_list);
@@ -2397,15 +2495,13 @@ int __i915_add_request(struct intel_engine_cs *ring,
 	ring->outstanding_lazy_seqno = 0;
 	ring->preallocated_lazy_request = NULL;
 
-	if (!dev_priv->ums.mm_suspended) {
-		i915_queue_hangcheck(ring->dev);
+	i915_queue_hangcheck(ring->dev);
 
-		cancel_delayed_work_sync(&dev_priv->mm.idle_work);
-		queue_delayed_work(dev_priv->wq,
-				   &dev_priv->mm.retire_work,
-				   round_jiffies_up_relative(HZ));
-		intel_mark_busy(dev_priv->dev);
-	}
+	cancel_delayed_work_sync(&dev_priv->mm.idle_work);
+	queue_delayed_work(dev_priv->wq,
+			   &dev_priv->mm.retire_work,
+			   round_jiffies_up_relative(HZ));
+	intel_mark_busy(dev_priv->dev);
 
 	if (out_seqno)
 		*out_seqno = request->seqno;
@@ -2472,12 +2568,20 @@ static void i915_set_reset_status(struct drm_i915_private *dev_priv,
 
 static void i915_gem_free_request(struct drm_i915_gem_request *request)
 {
+	struct intel_context *ctx = request->ctx;
+
 	list_del(&request->list);
 	i915_gem_request_remove_from_client(request);
 
-	if (request->ctx)
-		i915_gem_context_unreference(request->ctx);
+	if (ctx) {
+		if (i915.enable_execlists) {
+			struct intel_engine_cs *ring = request->ring;
 
+			if (ctx != ring->default_context)
+				intel_lr_context_unpin(ring, ctx);
+		}
+		i915_gem_context_unreference(ctx);
+	}
 	kfree(request);
 }
 
@@ -2532,6 +2636,23 @@ static void i915_gem_reset_ring_cleanup(struct drm_i915_private *dev_priv,
 	}
 
 	/*
+	 * Clear the execlists queue up before freeing the requests, as those
+	 * are the ones that keep the context and ringbuffer backing objects
+	 * pinned in place.
+	 */
+	while (!list_empty(&ring->execlist_queue)) {
+		struct intel_ctx_submit_request *submit_req;
+
+		submit_req = list_first_entry(&ring->execlist_queue,
+				struct intel_ctx_submit_request,
+				execlist_link);
+		list_del(&submit_req->execlist_link);
+		intel_runtime_pm_put(dev_priv);
+		i915_gem_context_unreference(submit_req->ctx);
+		kfree(submit_req);
+	}
+
+	/*
 	 * We must free the requests after all the corresponding objects have
 	 * been moved off active lists. Which is the same order as the normal
 	 * retire_requests function does. This is important if object hold
@@ -2632,6 +2753,7 @@ i915_gem_retire_requests_ring(struct intel_engine_cs *ring)
 
 	while (!list_empty(&ring->request_list)) {
 		struct drm_i915_gem_request *request;
+		struct intel_ringbuffer *ringbuf;
 
 		request = list_first_entry(&ring->request_list,
 					   struct drm_i915_gem_request,
@@ -2641,12 +2763,24 @@ i915_gem_retire_requests_ring(struct intel_engine_cs *ring)
 			break;
 
 		trace_i915_gem_request_retire(ring, request->seqno);
+
+		/* This is one of the few common intersection points
+		 * between legacy ringbuffer submission and execlists:
+		 * we need to tell them apart in order to find the correct
+		 * ringbuffer to which the request belongs to.
+		 */
+		if (i915.enable_execlists) {
+			struct intel_context *ctx = request->ctx;
+			ringbuf = ctx->engine[ring->id].ringbuf;
+		} else
+			ringbuf = ring->buffer;
+
 		/* We know the GPU must have read the request to have
 		 * sent us the seqno + interrupt, so use the position
 		 * of tail of the request to update the last known position
 		 * of the GPU head.
 		 */
-		ring->buffer->last_retired_head = request->tail;
+		ringbuf->last_retired_head = request->tail;
 
 		i915_gem_free_request(request);
 	}
@@ -2671,6 +2805,15 @@ i915_gem_retire_requests(struct drm_device *dev)
 	for_each_ring(ring, dev_priv, i) {
 		i915_gem_retire_requests_ring(ring);
 		idle &= list_empty(&ring->request_list);
+		if (i915.enable_execlists) {
+			unsigned long flags;
+
+			spin_lock_irqsave(&ring->execlist_lock, flags);
+			idle &= list_empty(&ring->execlist_queue);
+			spin_unlock_irqrestore(&ring->execlist_lock, flags);
+
+			intel_execlists_retire_requests(ring);
+		}
 	}
 
 	if (idle)
@@ -2763,6 +2906,9 @@ i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
 	u32 seqno = 0;
 	int ret = 0;
 
+	if (args->flags != 0)
+		return -EINVAL;
+
 	ret = i915_mutex_lock_interruptible(dev);
 	if (ret)
 		return ret;
@@ -2798,8 +2944,8 @@ i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
 	reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter);
 	mutex_unlock(&dev->struct_mutex);
 
-	return __wait_seqno(ring, seqno, reset_counter, true, &args->timeout_ns,
-			    file->driver_priv);
+	return __i915_wait_seqno(ring, seqno, reset_counter, true,
+				 &args->timeout_ns, file->driver_priv);
 
 out:
 	drm_gem_object_unreference(&obj->base);
@@ -2908,6 +3054,9 @@ int i915_vma_unbind(struct i915_vma *vma)
 	 * cause memory corruption through use-after-free.
 	 */
 
+	/* Throw away the active reference before moving to the unbound list */
+	i915_gem_object_retire(obj);
+
 	if (i915_is_ggtt(vma->vm)) {
 		i915_gem_object_finish_gtt(obj);
 
@@ -2922,9 +3071,8 @@ int i915_vma_unbind(struct i915_vma *vma)
 	vma->unbind_vma(vma);
 
 	list_del_init(&vma->mm_list);
-	/* Avoid an unnecessary call to unbind on rebind. */
 	if (i915_is_ggtt(vma->vm))
-		obj->map_and_fenceable = true;
+		obj->map_and_fenceable = false;
 
 	drm_mm_remove_node(&vma->node);
 	i915_gem_vma_destroy(vma);
@@ -2953,9 +3101,11 @@ int i915_gpu_idle(struct drm_device *dev)
 
 	/* Flush everything onto the inactive list. */
 	for_each_ring(ring, dev_priv, i) {
-		ret = i915_switch_context(ring, ring->default_context);
-		if (ret)
-			return ret;
+		if (!i915.enable_execlists) {
+			ret = i915_switch_context(ring, ring->default_context);
+			if (ret)
+				return ret;
+		}
 
 		ret = intel_ring_idle(ring);
 		if (ret)
@@ -2998,6 +3148,13 @@ static void i965_write_fence_reg(struct drm_device *dev, int reg,
 		u32 size = i915_gem_obj_ggtt_size(obj);
 		uint64_t val;
 
+		/* Adjust fence size to match tiled area */
+		if (obj->tiling_mode != I915_TILING_NONE) {
+			uint32_t row_size = obj->stride *
+				(obj->tiling_mode == I915_TILING_Y ? 32 : 8);
+			size = (size / row_size) * row_size;
+		}
+
 		val = (uint64_t)((i915_gem_obj_ggtt_offset(obj) + size - 4096) &
 				 0xfffff000) << 32;
 		val |= i915_gem_obj_ggtt_offset(obj) & 0xfffff000;
@@ -3114,6 +3271,7 @@ static void i915_gem_write_fence(struct drm_device *dev, int reg,
 	     obj->stride, obj->tiling_mode);
 
 	switch (INTEL_INFO(dev)->gen) {
+	case 9:
 	case 8:
 	case 7:
 	case 6:
@@ -3169,7 +3327,6 @@ i915_gem_object_wait_fence(struct drm_i915_gem_object *obj)
 		obj->last_fenced_seqno = 0;
 	}
 
-	obj->fenced_gpu_access = false;
 	return 0;
 }
 
@@ -3276,6 +3433,9 @@ i915_gem_object_get_fence(struct drm_i915_gem_object *obj)
 			return 0;
 		}
 	} else if (enable) {
+		if (WARN_ON(!obj->map_and_fenceable))
+			return -EINVAL;
+
 		reg = i915_find_fence_reg(dev);
 		if (IS_ERR(reg))
 			return PTR_ERR(reg);
@@ -3297,17 +3457,20 @@ i915_gem_object_get_fence(struct drm_i915_gem_object *obj)
 	return 0;
 }
 
-static bool i915_gem_valid_gtt_space(struct drm_device *dev,
-				     struct drm_mm_node *gtt_space,
+static bool i915_gem_valid_gtt_space(struct i915_vma *vma,
 				     unsigned long cache_level)
 {
+	struct drm_mm_node *gtt_space = &vma->node;
 	struct drm_mm_node *other;
 
-	/* On non-LLC machines we have to be careful when putting differing
-	 * types of snoopable memory together to avoid the prefetcher
-	 * crossing memory domains and dying.
+	/*
+	 * On some machines we have to be careful when putting differing types
+	 * of snoopable memory together to avoid the prefetcher crossing memory
+	 * domains and dying. During vm initialisation, we decide whether or not
+	 * these constraints apply and set the drm_mm.color_adjust
+	 * appropriately.
 	 */
-	if (HAS_LLC(dev))
+	if (vma->vm->mm.color_adjust == NULL)
 		return true;
 
 	if (!drm_mm_node_allocated(gtt_space))
@@ -3327,46 +3490,6 @@ static bool i915_gem_valid_gtt_space(struct drm_device *dev,
 	return true;
 }
 
-static void i915_gem_verify_gtt(struct drm_device *dev)
-{
-#if WATCH_GTT
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_i915_gem_object *obj;
-	int err = 0;
-
-	list_for_each_entry(obj, &dev_priv->mm.gtt_list, global_list) {
-		if (obj->gtt_space == NULL) {
-			printk(KERN_ERR "object found on GTT list with no space reserved\n");
-			err++;
-			continue;
-		}
-
-		if (obj->cache_level != obj->gtt_space->color) {
-			printk(KERN_ERR "object reserved space [%08lx, %08lx] with wrong color, cache_level=%x, color=%lx\n",
-			       i915_gem_obj_ggtt_offset(obj),
-			       i915_gem_obj_ggtt_offset(obj) + i915_gem_obj_ggtt_size(obj),
-			       obj->cache_level,
-			       obj->gtt_space->color);
-			err++;
-			continue;
-		}
-
-		if (!i915_gem_valid_gtt_space(dev,
-					      obj->gtt_space,
-					      obj->cache_level)) {
-			printk(KERN_ERR "invalid GTT space found at [%08lx, %08lx] - color=%x\n",
-			       i915_gem_obj_ggtt_offset(obj),
-			       i915_gem_obj_ggtt_offset(obj) + i915_gem_obj_ggtt_size(obj),
-			       obj->cache_level);
-			err++;
-			continue;
-		}
-	}
-
-	WARN_ON(err);
-#endif
-}
-
 /**
  * Finds free space in the GTT aperture and binds the object there.
  */
@@ -3445,8 +3568,7 @@ search_free:
 
 		goto err_free_vma;
 	}
-	if (WARN_ON(!i915_gem_valid_gtt_space(dev, &vma->node,
-					      obj->cache_level))) {
+	if (WARN_ON(!i915_gem_valid_gtt_space(vma, obj->cache_level))) {
 		ret = -EINVAL;
 		goto err_remove_node;
 	}
@@ -3458,25 +3580,10 @@ search_free:
 	list_move_tail(&obj->global_list, &dev_priv->mm.bound_list);
 	list_add_tail(&vma->mm_list, &vm->inactive_list);
 
-	if (i915_is_ggtt(vm)) {
-		bool mappable, fenceable;
-
-		fenceable = (vma->node.size == fence_size &&
-			     (vma->node.start & (fence_alignment - 1)) == 0);
-
-		mappable = (vma->node.start + obj->base.size <=
-			    dev_priv->gtt.mappable_end);
-
-		obj->map_and_fenceable = mappable && fenceable;
-	}
-
-	WARN_ON(flags & PIN_MAPPABLE && !obj->map_and_fenceable);
-
 	trace_i915_vma_bind(vma, flags);
 	vma->bind_vma(vma, obj->cache_level,
-		      flags & (PIN_MAPPABLE | PIN_GLOBAL) ? GLOBAL_BIND : 0);
+		      flags & PIN_GLOBAL ? GLOBAL_BIND : 0);
 
-	i915_gem_verify_gtt(dev);
 	return vma;
 
 err_remove_node:
@@ -3504,7 +3611,7 @@ i915_gem_clflush_object(struct drm_i915_gem_object *obj,
 	 * Stolen memory is always coherent with the GPU as it is explicitly
 	 * marked as wc by the system, or the system is cache-coherent.
 	 */
-	if (obj->stolen)
+	if (obj->stolen || obj->phys_handle)
 		return false;
 
 	/* If the GPU is snooping the contents of the CPU cache,
@@ -3586,11 +3693,12 @@ int
 i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write)
 {
 	struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
+	struct i915_vma *vma = i915_gem_obj_to_ggtt(obj);
 	uint32_t old_write_domain, old_read_domains;
 	int ret;
 
 	/* Not valid to be called on unbound objects. */
-	if (!i915_gem_obj_bound_any(obj))
+	if (vma == NULL)
 		return -EINVAL;
 
 	if (obj->base.write_domain == I915_GEM_DOMAIN_GTT)
@@ -3632,13 +3740,9 @@ i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write)
 					    old_write_domain);
 
 	/* And bump the LRU for this access */
-	if (i915_gem_object_is_inactive(obj)) {
-		struct i915_vma *vma = i915_gem_obj_to_ggtt(obj);
-		if (vma)
-			list_move_tail(&vma->mm_list,
-				       &dev_priv->gtt.base.inactive_list);
-
-	}
+	if (i915_gem_object_is_inactive(obj))
+		list_move_tail(&vma->mm_list,
+			       &dev_priv->gtt.base.inactive_list);
 
 	return 0;
 }
@@ -3659,7 +3763,7 @@ int i915_gem_object_set_cache_level(struct drm_i915_gem_object *obj,
 	}
 
 	list_for_each_entry_safe(vma, next, &obj->vma_list, vma_link) {
-		if (!i915_gem_valid_gtt_space(dev, &vma->node, cache_level)) {
+		if (!i915_gem_valid_gtt_space(vma, cache_level)) {
 			ret = i915_vma_unbind(vma);
 			if (ret)
 				return ret;
@@ -3686,7 +3790,7 @@ int i915_gem_object_set_cache_level(struct drm_i915_gem_object *obj,
 		list_for_each_entry(vma, &obj->vma_list, vma_link)
 			if (drm_mm_node_allocated(&vma->node))
 				vma->bind_vma(vma, cache_level,
-					      obj->has_global_gtt_mapping ? GLOBAL_BIND : 0);
+						vma->bound & GLOBAL_BIND);
 	}
 
 	list_for_each_entry(vma, &obj->vma_list, vma_link)
@@ -3716,7 +3820,6 @@ int i915_gem_object_set_cache_level(struct drm_i915_gem_object *obj,
 						    old_write_domain);
 	}
 
-	i915_gem_verify_gtt(dev);
 	return 0;
 }
 
@@ -3802,9 +3905,6 @@ static bool is_pin_display(struct drm_i915_gem_object *obj)
 {
 	struct i915_vma *vma;
 
-	if (list_empty(&obj->vma_list))
-		return false;
-
 	vma = i915_gem_obj_to_ggtt(obj);
 	if (!vma)
 		return false;
@@ -4017,7 +4117,7 @@ i915_gem_ring_throttle(struct drm_device *dev, struct drm_file *file)
 	if (seqno == 0)
 		return 0;
 
-	ret = __wait_seqno(ring, seqno, reset_counter, true, NULL, NULL);
+	ret = __i915_wait_seqno(ring, seqno, reset_counter, true, NULL, NULL);
 	if (ret == 0)
 		queue_delayed_work(dev_priv->wq, &dev_priv->mm.retire_work, 0);
 
@@ -4051,6 +4151,7 @@ i915_gem_object_pin(struct drm_i915_gem_object *obj,
 {
 	struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
 	struct i915_vma *vma;
+	unsigned bound;
 	int ret;
 
 	if (WARN_ON(vm == &dev_priv->mm.aliasing_ppgtt->base))
@@ -4059,6 +4160,9 @@ i915_gem_object_pin(struct drm_i915_gem_object *obj,
 	if (WARN_ON(flags & (PIN_GLOBAL | PIN_MAPPABLE) && !i915_is_ggtt(vm)))
 		return -EINVAL;
 
+	if (WARN_ON((flags & (PIN_MAPPABLE | PIN_GLOBAL)) == PIN_MAPPABLE))
+		return -EINVAL;
+
 	vma = i915_gem_obj_to_vma(obj, vm);
 	if (vma) {
 		if (WARN_ON(vma->pin_count == DRM_I915_GEM_OBJECT_MAX_PIN_COUNT))
@@ -4080,15 +4184,39 @@ i915_gem_object_pin(struct drm_i915_gem_object *obj,
 		}
 	}
 
+	bound = vma ? vma->bound : 0;
 	if (vma == NULL || !drm_mm_node_allocated(&vma->node)) {
 		vma = i915_gem_object_bind_to_vm(obj, vm, alignment, flags);
 		if (IS_ERR(vma))
 			return PTR_ERR(vma);
 	}
 
-	if (flags & PIN_GLOBAL && !obj->has_global_gtt_mapping)
+	if (flags & PIN_GLOBAL && !(vma->bound & GLOBAL_BIND))
 		vma->bind_vma(vma, obj->cache_level, GLOBAL_BIND);
 
+	if ((bound ^ vma->bound) & GLOBAL_BIND) {
+		bool mappable, fenceable;
+		u32 fence_size, fence_alignment;
+
+		fence_size = i915_gem_get_gtt_size(obj->base.dev,
+						   obj->base.size,
+						   obj->tiling_mode);
+		fence_alignment = i915_gem_get_gtt_alignment(obj->base.dev,
+							     obj->base.size,
+							     obj->tiling_mode,
+							     true);
+
+		fenceable = (vma->node.size == fence_size &&
+			     (vma->node.start & (fence_alignment - 1)) == 0);
+
+		mappable = (vma->node.start + obj->base.size <=
+			    dev_priv->gtt.mappable_end);
+
+		obj->map_and_fenceable = mappable && fenceable;
+	}
+
+	WARN_ON(flags & PIN_MAPPABLE && !obj->map_and_fenceable);
+
 	vma->pin_count++;
 	if (flags & PIN_MAPPABLE)
 		obj->pin_mappable |= true;
@@ -4143,7 +4271,7 @@ i915_gem_pin_ioctl(struct drm_device *dev, void *data,
 	struct drm_i915_gem_object *obj;
 	int ret;
 
-	if (INTEL_INFO(dev)->gen >= 6)
+	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		return -ENODEV;
 
 	ret = i915_mutex_lock_interruptible(dev);
@@ -4199,6 +4327,9 @@ i915_gem_unpin_ioctl(struct drm_device *dev, void *data,
 	struct drm_i915_gem_object *obj;
 	int ret;
 
+	if (drm_core_check_feature(dev, DRIVER_MODESET))
+		return -ENODEV;
+
 	ret = i915_mutex_lock_interruptible(dev);
 	if (ret)
 		return ret;
@@ -4276,6 +4407,7 @@ int
 i915_gem_madvise_ioctl(struct drm_device *dev, void *data,
 		       struct drm_file *file_priv)
 {
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_gem_madvise *args = data;
 	struct drm_i915_gem_object *obj;
 	int ret;
@@ -4303,6 +4435,15 @@ i915_gem_madvise_ioctl(struct drm_device *dev, void *data,
 		goto out;
 	}
 
+	if (obj->pages &&
+	    obj->tiling_mode != I915_TILING_NONE &&
+	    dev_priv->quirks & QUIRK_PIN_SWIZZLED_PAGES) {
+		if (obj->madv == I915_MADV_WILLNEED)
+			i915_gem_object_unpin_pages(obj);
+		if (args->madv == I915_MADV_WILLNEED)
+			i915_gem_object_pin_pages(obj);
+	}
+
 	if (obj->madv != __I915_MADV_PURGED)
 		obj->madv = args->madv;
 
@@ -4331,8 +4472,6 @@ void i915_gem_object_init(struct drm_i915_gem_object *obj,
 
 	obj->fence_reg = I915_FENCE_REG_NONE;
 	obj->madv = I915_MADV_WILLNEED;
-	/* Avoid an unnecessary call to unbind on the first bind. */
-	obj->map_and_fenceable = true;
 
 	i915_gem_info_add_obj(obj->base.dev->dev_private, obj->base.size);
 }
@@ -4447,8 +4586,6 @@ void i915_gem_free_object(struct drm_gem_object *gem_obj)
 		}
 	}
 
-	i915_gem_object_detach_phys(obj);
-
 	/* Stolen objects don't hold a ref, but do hold pin count. Fix that up
 	 * before progressing. */
 	if (obj->stolen)
@@ -4456,6 +4593,11 @@ void i915_gem_free_object(struct drm_gem_object *gem_obj)
 
 	WARN_ON(obj->frontbuffer_bits);
 
+	if (obj->pages && obj->madv == I915_MADV_WILLNEED &&
+	    dev_priv->quirks & QUIRK_PIN_SWIZZLED_PAGES &&
+	    obj->tiling_mode != I915_TILING_NONE)
+		i915_gem_object_unpin_pages(obj);
+
 	if (WARN_ON(obj->pages_pin_count))
 		obj->pages_pin_count = 0;
 	if (discard_backing_storage(obj))
@@ -4493,12 +4635,18 @@ struct i915_vma *i915_gem_obj_to_vma(struct drm_i915_gem_object *obj,
 
 void i915_gem_vma_destroy(struct i915_vma *vma)
 {
+	struct i915_address_space *vm = NULL;
 	WARN_ON(vma->node.allocated);
 
 	/* Keep the vma as a placeholder in the execbuffer reservation lists */
 	if (!list_empty(&vma->exec_list))
 		return;
 
+	vm = vma->vm;
+
+	if (!i915_is_ggtt(vm))
+		i915_ppgtt_put(i915_vm_to_ppgtt(vm));
+
 	list_del(&vma->vma_link);
 
 	kfree(vma);
@@ -4512,7 +4660,7 @@ i915_gem_stop_ringbuffers(struct drm_device *dev)
 	int i;
 
 	for_each_ring(ring, dev_priv, i)
-		intel_stop_ring_buffer(ring);
+		dev_priv->gt.stop_ring(ring);
 }
 
 int
@@ -4522,9 +4670,6 @@ i915_gem_suspend(struct drm_device *dev)
 	int ret = 0;
 
 	mutex_lock(&dev->struct_mutex);
-	if (dev_priv->ums.mm_suspended)
-		goto err;
-
 	ret = i915_gpu_idle(dev);
 	if (ret)
 		goto err;
@@ -4535,15 +4680,7 @@ i915_gem_suspend(struct drm_device *dev)
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		i915_gem_evict_everything(dev);
 
-	i915_kernel_lost_context(dev);
 	i915_gem_stop_ringbuffers(dev);
-
-	/* Hack!  Don't let anybody do execbuf while we don't control the chip.
-	 * We need to replace this with a semaphore, or something.
-	 * And not confound ums.mm_suspended!
-	 */
-	dev_priv->ums.mm_suspended = !drm_core_check_feature(dev,
-							     DRIVER_MODESET);
 	mutex_unlock(&dev->struct_mutex);
 
 	del_timer_sync(&dev_priv->gpu_error.hangcheck_timer);
@@ -4629,11 +4766,46 @@ intel_enable_blt(struct drm_device *dev)
 	return true;
 }
 
-static int i915_gem_init_rings(struct drm_device *dev)
+static void init_unused_ring(struct drm_device *dev, u32 base)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	I915_WRITE(RING_CTL(base), 0);
+	I915_WRITE(RING_HEAD(base), 0);
+	I915_WRITE(RING_TAIL(base), 0);
+	I915_WRITE(RING_START(base), 0);
+}
+
+static void init_unused_rings(struct drm_device *dev)
+{
+	if (IS_I830(dev)) {
+		init_unused_ring(dev, PRB1_BASE);
+		init_unused_ring(dev, SRB0_BASE);
+		init_unused_ring(dev, SRB1_BASE);
+		init_unused_ring(dev, SRB2_BASE);
+		init_unused_ring(dev, SRB3_BASE);
+	} else if (IS_GEN2(dev)) {
+		init_unused_ring(dev, SRB0_BASE);
+		init_unused_ring(dev, SRB1_BASE);
+	} else if (IS_GEN3(dev)) {
+		init_unused_ring(dev, PRB1_BASE);
+		init_unused_ring(dev, PRB2_BASE);
+	}
+}
+
+int i915_gem_init_rings(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
+	/*
+	 * At least 830 can leave some of the unused rings
+	 * "active" (ie. head != tail) after resume which
+	 * will prevent c3 entry. Makes sure all unused rings
+	 * are totally idle.
+	 */
+	init_unused_rings(dev);
+
 	ret = intel_init_render_ring_buffer(dev);
 	if (ret)
 		return ret;
@@ -4712,24 +4884,25 @@ i915_gem_init_hw(struct drm_device *dev)
 
 	i915_gem_init_swizzling(dev);
 
-	ret = i915_gem_init_rings(dev);
+	ret = dev_priv->gt.init_rings(dev);
 	if (ret)
 		return ret;
 
 	for (i = 0; i < NUM_L3_SLICES(dev); i++)
 		i915_gem_l3_remap(&dev_priv->ring[RCS], i);
 
-	/*
-	 * XXX: Contexts should only be initialized once. Doing a switch to the
-	 * default context switch however is something we'd like to do after
-	 * reset or thaw (the latter may not actually be necessary for HW, but
-	 * goes with our code better). Context switching requires rings (for
-	 * the do_switch), but before enabling PPGTT. So don't move this.
-	 */
+	ret = i915_ppgtt_init_hw(dev);
+	if (ret && ret != -EIO) {
+		DRM_ERROR("PPGTT enable failed %d\n", ret);
+		i915_gem_cleanup_ringbuffer(dev);
+	}
+
 	ret = i915_gem_context_enable(dev_priv);
 	if (ret && ret != -EIO) {
 		DRM_ERROR("Context enable failed %d\n", ret);
 		i915_gem_cleanup_ringbuffer(dev);
+
+		return ret;
 	}
 
 	return ret;
@@ -4740,6 +4913,9 @@ int i915_gem_init(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
+	i915.enable_execlists = intel_sanitize_enable_execlists(dev,
+			i915.enable_execlists);
+
 	mutex_lock(&dev->struct_mutex);
 
 	if (IS_VALLEYVIEW(dev)) {
@@ -4750,7 +4926,24 @@ int i915_gem_init(struct drm_device *dev)
 			DRM_DEBUG_DRIVER("allow wake ack timed out\n");
 	}
 
-	i915_gem_init_userptr(dev);
+	if (!i915.enable_execlists) {
+		dev_priv->gt.do_execbuf = i915_gem_ringbuffer_submission;
+		dev_priv->gt.init_rings = i915_gem_init_rings;
+		dev_priv->gt.cleanup_ring = intel_cleanup_ring_buffer;
+		dev_priv->gt.stop_ring = intel_stop_ring_buffer;
+	} else {
+		dev_priv->gt.do_execbuf = intel_execlists_submission;
+		dev_priv->gt.init_rings = intel_logical_rings_init;
+		dev_priv->gt.cleanup_ring = intel_logical_ring_cleanup;
+		dev_priv->gt.stop_ring = intel_logical_ring_stop;
+	}
+
+	ret = i915_gem_init_userptr(dev);
+	if (ret) {
+		mutex_unlock(&dev->struct_mutex);
+		return ret;
+	}
+
 	i915_gem_init_global_gtt(dev);
 
 	ret = i915_gem_context_init(dev);
@@ -4771,9 +4964,6 @@ int i915_gem_init(struct drm_device *dev)
 	}
 	mutex_unlock(&dev->struct_mutex);
 
-	/* Allow hardware batchbuffers unless told otherwise, but not for KMS. */
-	if (!drm_core_check_feature(dev, DRIVER_MODESET))
-		dev_priv->dri1.allow_batchbuffer = 1;
 	return ret;
 }
 
@@ -4785,75 +4975,7 @@ i915_gem_cleanup_ringbuffer(struct drm_device *dev)
 	int i;
 
 	for_each_ring(ring, dev_priv, i)
-		intel_cleanup_ring_buffer(ring);
-}
-
-int
-i915_gem_entervt_ioctl(struct drm_device *dev, void *data,
-		       struct drm_file *file_priv)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int ret;
-
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		return 0;
-
-	if (i915_reset_in_progress(&dev_priv->gpu_error)) {
-		DRM_ERROR("Reenabling wedged hardware, good luck\n");
-		atomic_set(&dev_priv->gpu_error.reset_counter, 0);
-	}
-
-	mutex_lock(&dev->struct_mutex);
-	dev_priv->ums.mm_suspended = 0;
-
-	ret = i915_gem_init_hw(dev);
-	if (ret != 0) {
-		mutex_unlock(&dev->struct_mutex);
-		return ret;
-	}
-
-	BUG_ON(!list_empty(&dev_priv->gtt.base.active_list));
-
-	ret = drm_irq_install(dev, dev->pdev->irq);
-	if (ret)
-		goto cleanup_ringbuffer;
-	mutex_unlock(&dev->struct_mutex);
-
-	return 0;
-
-cleanup_ringbuffer:
-	i915_gem_cleanup_ringbuffer(dev);
-	dev_priv->ums.mm_suspended = 1;
-	mutex_unlock(&dev->struct_mutex);
-
-	return ret;
-}
-
-int
-i915_gem_leavevt_ioctl(struct drm_device *dev, void *data,
-		       struct drm_file *file_priv)
-{
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		return 0;
-
-	mutex_lock(&dev->struct_mutex);
-	drm_irq_uninstall(dev);
-	mutex_unlock(&dev->struct_mutex);
-
-	return i915_gem_suspend(dev);
-}
-
-void
-i915_gem_lastclose(struct drm_device *dev)
-{
-	int ret;
-
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		return;
-
-	ret = i915_gem_suspend(dev);
-	if (ret)
-		DRM_ERROR("failed to idle hardware: %d\n", ret);
+		dev_priv->gt.cleanup_ring(ring);
 }
 
 static void
@@ -5002,6 +5124,15 @@ int i915_gem_open(struct drm_device *dev, struct drm_file *file)
 	return ret;
 }
 
+/**
+ * i915_gem_track_fb - update frontbuffer tracking
+ * old: current GEM buffer for the frontbuffer slots
+ * new: new GEM buffer for the frontbuffer slots
+ * frontbuffer_bits: bitmask of frontbuffer slots
+ *
+ * This updates the frontbuffer tracking bits @frontbuffer_bits by clearing them
+ * from @old and setting them in @new. Both @old and @new can be NULL.
+ */
 void i915_gem_track_fb(struct drm_i915_gem_object *old,
 		       struct drm_i915_gem_object *new,
 		       unsigned frontbuffer_bits)
@@ -5024,7 +5155,7 @@ static bool mutex_is_locked_by(struct mutex *mutex, struct task_struct *task)
 	if (!mutex_is_locked(mutex))
 		return false;
 
-#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_MUTEXES)
+#if defined(CONFIG_SMP) && !defined(CONFIG_DEBUG_MUTEXES)
 	return mutex->owner == task;
 #else
 	/* Since UP may be pre-empted, we cannot assume that we own the lock */
@@ -5097,9 +5228,7 @@ unsigned long i915_gem_obj_offset(struct drm_i915_gem_object *o,
 	struct drm_i915_private *dev_priv = o->base.dev->dev_private;
 	struct i915_vma *vma;
 
-	if (!dev_priv->mm.aliasing_ppgtt ||
-	    vm == &dev_priv->mm.aliasing_ppgtt->base)
-		vm = &dev_priv->gtt.base;
+	WARN_ON(vm == &dev_priv->mm.aliasing_ppgtt->base);
 
 	list_for_each_entry(vma, &o->vma_list, vma_link) {
 		if (vma->vm == vm)
@@ -5140,9 +5269,7 @@ unsigned long i915_gem_obj_size(struct drm_i915_gem_object *o,
 	struct drm_i915_private *dev_priv = o->base.dev->dev_private;
 	struct i915_vma *vma;
 
-	if (!dev_priv->mm.aliasing_ppgtt ||
-	    vm == &dev_priv->mm.aliasing_ppgtt->base)
-		vm = &dev_priv->gtt.base;
+	WARN_ON(vm == &dev_priv->mm.aliasing_ppgtt->base);
 
 	BUG_ON(list_empty(&o->vma_list));
 
@@ -5165,11 +5292,16 @@ i915_gem_shrinker_scan(struct shrinker *shrinker, struct shrink_control *sc)
 	if (!i915_gem_shrinker_lock(dev, &unlock))
 		return SHRINK_STOP;
 
-	freed = i915_gem_purge(dev_priv, sc->nr_to_scan);
+	freed = i915_gem_shrink(dev_priv,
+				sc->nr_to_scan,
+				I915_SHRINK_BOUND |
+				I915_SHRINK_UNBOUND |
+				I915_SHRINK_PURGEABLE);
 	if (freed < sc->nr_to_scan)
-		freed += __i915_gem_shrink(dev_priv,
-					   sc->nr_to_scan - freed,
-					   false);
+		freed += i915_gem_shrink(dev_priv,
+					 sc->nr_to_scan - freed,
+					 I915_SHRINK_BOUND |
+					 I915_SHRINK_UNBOUND);
 	if (unlock)
 		mutex_unlock(&dev->struct_mutex);
 
@@ -5184,7 +5316,7 @@ i915_gem_shrinker_oom(struct notifier_block *nb, unsigned long event, void *ptr)
 	struct drm_device *dev = dev_priv->dev;
 	struct drm_i915_gem_object *obj;
 	unsigned long timeout = msecs_to_jiffies(5000) + 1;
-	unsigned long pinned, bound, unbound, freed;
+	unsigned long pinned, bound, unbound, freed_pages;
 	bool was_interruptible;
 	bool unlock;
 
@@ -5201,7 +5333,7 @@ i915_gem_shrinker_oom(struct notifier_block *nb, unsigned long event, void *ptr)
 	was_interruptible = dev_priv->mm.interruptible;
 	dev_priv->mm.interruptible = false;
 
-	freed = i915_gem_shrink_all(dev_priv);
+	freed_pages = i915_gem_shrink_all(dev_priv);
 
 	dev_priv->mm.interruptible = was_interruptible;
 
@@ -5232,14 +5364,15 @@ i915_gem_shrinker_oom(struct notifier_block *nb, unsigned long event, void *ptr)
 	if (unlock)
 		mutex_unlock(&dev->struct_mutex);
 
-	pr_info("Purging GPU memory, %lu bytes freed, %lu bytes still pinned.\n",
-		freed, pinned);
+	if (freed_pages || unbound || bound)
+		pr_info("Purging GPU memory, %lu bytes freed, %lu bytes still pinned.\n",
+			freed_pages << PAGE_SHIFT, pinned);
 	if (unbound || bound)
 		pr_err("%lu and %lu bytes still available in the "
 		       "bound and unbound GPU page lists.\n",
 		       bound, unbound);
 
-	*(unsigned long *)ptr += freed;
+	*(unsigned long *)ptr += freed_pages;
 	return NOTIFY_DONE;
 }
 
@@ -5247,14 +5380,8 @@ struct i915_vma *i915_gem_obj_to_ggtt(struct drm_i915_gem_object *obj)
 {
 	struct i915_vma *vma;
 
-	/* This WARN has probably outlived its usefulness (callers already
-	 * WARN if they don't find the GGTT vma they expect). When removing,
-	 * remember to remove the pre-check in is_pin_display() as well */
-	if (WARN_ON(list_empty(&obj->vma_list)))
-		return NULL;
-
 	vma = list_first_entry(&obj->vma_list, typeof(*vma), vma_link);
-	if (vma->vm != obj_to_ggtt(obj))
+	if (vma->vm != i915_obj_to_ggtt(obj))
 		return NULL;
 
 	return vma;
diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index 3b99390e467a..d011ec82ef1e 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -88,6 +88,7 @@
 #include <drm/drmP.h>
 #include <drm/i915_drm.h>
 #include "i915_drv.h"
+#include "i915_trace.h"
 
 /* This is a HW constraint. The value below is the largest known requirement
  * I've seen in a spec to date, and that was a workaround for a non-shipping
@@ -96,50 +97,6 @@
 #define GEN6_CONTEXT_ALIGN (64<<10)
 #define GEN7_CONTEXT_ALIGN 4096
 
-static void do_ppgtt_cleanup(struct i915_hw_ppgtt *ppgtt)
-{
-	struct drm_device *dev = ppgtt->base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct i915_address_space *vm = &ppgtt->base;
-
-	if (ppgtt == dev_priv->mm.aliasing_ppgtt ||
-	    (list_empty(&vm->active_list) && list_empty(&vm->inactive_list))) {
-		ppgtt->base.cleanup(&ppgtt->base);
-		return;
-	}
-
-	/*
-	 * Make sure vmas are unbound before we take down the drm_mm
-	 *
-	 * FIXME: Proper refcounting should take care of this, this shouldn't be
-	 * needed at all.
-	 */
-	if (!list_empty(&vm->active_list)) {
-		struct i915_vma *vma;
-
-		list_for_each_entry(vma, &vm->active_list, mm_list)
-			if (WARN_ON(list_empty(&vma->vma_link) ||
-				    list_is_singular(&vma->vma_link)))
-				break;
-
-		i915_gem_evict_vm(&ppgtt->base, true);
-	} else {
-		i915_gem_retire_requests(dev);
-		i915_gem_evict_vm(&ppgtt->base, false);
-	}
-
-	ppgtt->base.cleanup(&ppgtt->base);
-}
-
-static void ppgtt_release(struct kref *kref)
-{
-	struct i915_hw_ppgtt *ppgtt =
-		container_of(kref, struct i915_hw_ppgtt, ref);
-
-	do_ppgtt_cleanup(ppgtt);
-	kfree(ppgtt);
-}
-
 static size_t get_context_alignment(struct drm_device *dev)
 {
 	if (IS_GEN6(dev))
@@ -179,24 +136,22 @@ static int get_context_size(struct drm_device *dev)
 void i915_gem_context_free(struct kref *ctx_ref)
 {
 	struct intel_context *ctx = container_of(ctx_ref,
-						   typeof(*ctx), ref);
-	struct i915_hw_ppgtt *ppgtt = NULL;
+						 typeof(*ctx), ref);
 
-	if (ctx->legacy_hw_ctx.rcs_state) {
-		/* We refcount even the aliasing PPGTT to keep the code symmetric */
-		if (USES_PPGTT(ctx->legacy_hw_ctx.rcs_state->base.dev))
-			ppgtt = ctx_to_ppgtt(ctx);
-	}
+	trace_i915_context_free(ctx);
+
+	if (i915.enable_execlists)
+		intel_lr_context_free(ctx);
+
+	i915_ppgtt_put(ctx->ppgtt);
 
-	if (ppgtt)
-		kref_put(&ppgtt->ref, ppgtt_release);
 	if (ctx->legacy_hw_ctx.rcs_state)
 		drm_gem_object_unreference(&ctx->legacy_hw_ctx.rcs_state->base);
 	list_del(&ctx->link);
 	kfree(ctx);
 }
 
-static struct drm_i915_gem_object *
+struct drm_i915_gem_object *
 i915_gem_alloc_context_obj(struct drm_device *dev, size_t size)
 {
 	struct drm_i915_gem_object *obj;
@@ -226,29 +181,9 @@ i915_gem_alloc_context_obj(struct drm_device *dev, size_t size)
 	return obj;
 }
 
-static struct i915_hw_ppgtt *
-create_vm_for_ctx(struct drm_device *dev, struct intel_context *ctx)
-{
-	struct i915_hw_ppgtt *ppgtt;
-	int ret;
-
-	ppgtt = kzalloc(sizeof(*ppgtt), GFP_KERNEL);
-	if (!ppgtt)
-		return ERR_PTR(-ENOMEM);
-
-	ret = i915_gem_init_ppgtt(dev, ppgtt);
-	if (ret) {
-		kfree(ppgtt);
-		return ERR_PTR(ret);
-	}
-
-	ppgtt->ctx = ctx;
-	return ppgtt;
-}
-
 static struct intel_context *
 __create_hw_context(struct drm_device *dev,
-		  struct drm_i915_file_private *file_priv)
+		    struct drm_i915_file_private *file_priv)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_context *ctx;
@@ -301,11 +236,9 @@ err_out:
  */
 static struct intel_context *
 i915_gem_create_context(struct drm_device *dev,
-			struct drm_i915_file_private *file_priv,
-			bool create_vm)
+			struct drm_i915_file_private *file_priv)
 {
 	const bool is_global_default_ctx = file_priv == NULL;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_context *ctx;
 	int ret = 0;
 
@@ -331,34 +264,20 @@ i915_gem_create_context(struct drm_device *dev,
 		}
 	}
 
-	if (create_vm) {
-		struct i915_hw_ppgtt *ppgtt = create_vm_for_ctx(dev, ctx);
+	if (USES_FULL_PPGTT(dev)) {
+		struct i915_hw_ppgtt *ppgtt = i915_ppgtt_create(dev, file_priv);
 
 		if (IS_ERR_OR_NULL(ppgtt)) {
 			DRM_DEBUG_DRIVER("PPGTT setup failed (%ld)\n",
 					 PTR_ERR(ppgtt));
 			ret = PTR_ERR(ppgtt);
 			goto err_unpin;
-		} else
-			ctx->vm = &ppgtt->base;
-
-		/* This case is reserved for the global default context and
-		 * should only happen once. */
-		if (is_global_default_ctx) {
-			if (WARN_ON(dev_priv->mm.aliasing_ppgtt)) {
-				ret = -EEXIST;
-				goto err_unpin;
-			}
-
-			dev_priv->mm.aliasing_ppgtt = ppgtt;
 		}
-	} else if (USES_PPGTT(dev)) {
-		/* For platforms which only have aliasing PPGTT, we fake the
-		 * address space and refcounting. */
-		ctx->vm = &dev_priv->mm.aliasing_ppgtt->base;
-		kref_get(&dev_priv->mm.aliasing_ppgtt->ref);
-	} else
-		ctx->vm = &dev_priv->gtt.base;
+
+		ctx->ppgtt = ppgtt;
+	}
+
+	trace_i915_context_create(ctx);
 
 	return ctx;
 
@@ -375,34 +294,23 @@ void i915_gem_context_reset(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int i;
 
-	/* Prevent the hardware from restoring the last context (which hung) on
-	 * the next switch */
+	/* In execlists mode we will unreference the context when the execlist
+	 * queue is cleared and the requests destroyed.
+	 */
+	if (i915.enable_execlists)
+		return;
+
 	for (i = 0; i < I915_NUM_RINGS; i++) {
 		struct intel_engine_cs *ring = &dev_priv->ring[i];
-		struct intel_context *dctx = ring->default_context;
 		struct intel_context *lctx = ring->last_context;
 
-		/* Do a fake switch to the default context */
-		if (lctx == dctx)
-			continue;
-
-		if (!lctx)
-			continue;
+		if (lctx) {
+			if (lctx->legacy_hw_ctx.rcs_state && i == RCS)
+				i915_gem_object_ggtt_unpin(lctx->legacy_hw_ctx.rcs_state);
 
-		if (dctx->legacy_hw_ctx.rcs_state && i == RCS) {
-			WARN_ON(i915_gem_obj_ggtt_pin(dctx->legacy_hw_ctx.rcs_state,
-						      get_context_alignment(dev), 0));
-			/* Fake a finish/inactive */
-			dctx->legacy_hw_ctx.rcs_state->base.write_domain = 0;
-			dctx->legacy_hw_ctx.rcs_state->active = 0;
+			i915_gem_context_unreference(lctx);
+			ring->last_context = NULL;
 		}
-
-		if (lctx->legacy_hw_ctx.rcs_state && i == RCS)
-			i915_gem_object_ggtt_unpin(lctx->legacy_hw_ctx.rcs_state);
-
-		i915_gem_context_unreference(lctx);
-		i915_gem_context_reference(dctx);
-		ring->last_context = dctx;
 	}
 }
 
@@ -417,7 +325,11 @@ int i915_gem_context_init(struct drm_device *dev)
 	if (WARN_ON(dev_priv->ring[RCS].default_context))
 		return 0;
 
-	if (HAS_HW_CONTEXTS(dev)) {
+	if (i915.enable_execlists) {
+		/* NB: intentionally left blank. We will allocate our own
+		 * backing objects as we need them, thank you very much */
+		dev_priv->hw_context_size = 0;
+	} else if (HAS_HW_CONTEXTS(dev)) {
 		dev_priv->hw_context_size = round_up(get_context_size(dev), 4096);
 		if (dev_priv->hw_context_size > (1<<20)) {
 			DRM_DEBUG_DRIVER("Disabling HW Contexts; invalid size %d\n",
@@ -426,18 +338,23 @@ int i915_gem_context_init(struct drm_device *dev)
 		}
 	}
 
-	ctx = i915_gem_create_context(dev, NULL, USES_PPGTT(dev));
+	ctx = i915_gem_create_context(dev, NULL);
 	if (IS_ERR(ctx)) {
 		DRM_ERROR("Failed to create default global context (error %ld)\n",
 			  PTR_ERR(ctx));
 		return PTR_ERR(ctx);
 	}
 
-	/* NB: RCS will hold a ref for all rings */
-	for (i = 0; i < I915_NUM_RINGS; i++)
-		dev_priv->ring[i].default_context = ctx;
+	for (i = 0; i < I915_NUM_RINGS; i++) {
+		struct intel_engine_cs *ring = &dev_priv->ring[i];
+
+		/* NB: RCS will hold a ref for all rings */
+		ring->default_context = ctx;
+	}
 
-	DRM_DEBUG_DRIVER("%s context support initialized\n", dev_priv->hw_context_size ? "HW" : "fake");
+	DRM_DEBUG_DRIVER("%s context support initialized\n",
+			i915.enable_execlists ? "LR" :
+			dev_priv->hw_context_size ? "HW" : "fake");
 	return 0;
 }
 
@@ -489,19 +406,11 @@ int i915_gem_context_enable(struct drm_i915_private *dev_priv)
 	struct intel_engine_cs *ring;
 	int ret, i;
 
-	/* This is the only place the aliasing PPGTT gets enabled, which means
-	 * it has to happen before we bail on reset */
-	if (dev_priv->mm.aliasing_ppgtt) {
-		struct i915_hw_ppgtt *ppgtt = dev_priv->mm.aliasing_ppgtt;
-		ppgtt->enable(ppgtt);
-	}
+	BUG_ON(!dev_priv->ring[RCS].default_context);
 
-	/* FIXME: We should make this work, even in reset */
-	if (i915_reset_in_progress(&dev_priv->gpu_error))
+	if (i915.enable_execlists)
 		return 0;
 
-	BUG_ON(!dev_priv->ring[RCS].default_context);
-
 	for_each_ring(ring, dev_priv, i) {
 		ret = i915_switch_context(ring, ring->default_context);
 		if (ret)
@@ -527,7 +436,7 @@ int i915_gem_context_open(struct drm_device *dev, struct drm_file *file)
 	idr_init(&file_priv->context_idr);
 
 	mutex_lock(&dev->struct_mutex);
-	ctx = i915_gem_create_context(dev, file_priv, USES_FULL_PPGTT(dev));
+	ctx = i915_gem_create_context(dev, file_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	if (IS_ERR(ctx)) {
@@ -563,7 +472,13 @@ mi_set_context(struct intel_engine_cs *ring,
 	       struct intel_context *new_context,
 	       u32 hw_flags)
 {
-	int ret;
+	u32 flags = hw_flags | MI_MM_SPACE_GTT;
+	const int num_rings =
+		/* Use an extended w/a on ivb+ if signalling from other rings */
+		i915_semaphore_is_enabled(ring->dev) ?
+		hweight32(INTEL_INFO(ring->dev)->ring_mask) - 1 :
+		0;
+	int len, i, ret;
 
 	/* w/a: If Flush TLB Invalidation Mode is enabled, driver must do a TLB
 	 * invalidation prior to MI_SET_CONTEXT. On GEN6 we don't set the value
@@ -576,33 +491,61 @@ mi_set_context(struct intel_engine_cs *ring,
 			return ret;
 	}
 
-	ret = intel_ring_begin(ring, 6);
+	/* These flags are for resource streamer on HSW+ */
+	if (!IS_HASWELL(ring->dev) && INTEL_INFO(ring->dev)->gen < 8)
+		flags |= (MI_SAVE_EXT_STATE_EN | MI_RESTORE_EXT_STATE_EN);
+
+
+	len = 4;
+	if (INTEL_INFO(ring->dev)->gen >= 7)
+		len += 2 + (num_rings ? 4*num_rings + 2 : 0);
+
+	ret = intel_ring_begin(ring, len);
 	if (ret)
 		return ret;
 
 	/* WaProgramMiArbOnOffAroundMiSetContext:ivb,vlv,hsw,bdw,chv */
-	if (INTEL_INFO(ring->dev)->gen >= 7)
+	if (INTEL_INFO(ring->dev)->gen >= 7) {
 		intel_ring_emit(ring, MI_ARB_ON_OFF | MI_ARB_DISABLE);
-	else
-		intel_ring_emit(ring, MI_NOOP);
+		if (num_rings) {
+			struct intel_engine_cs *signaller;
+
+			intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(num_rings));
+			for_each_ring(signaller, to_i915(ring->dev), i) {
+				if (signaller == ring)
+					continue;
+
+				intel_ring_emit(ring, RING_PSMI_CTL(signaller->mmio_base));
+				intel_ring_emit(ring, _MASKED_BIT_ENABLE(GEN6_PSMI_SLEEP_MSG_DISABLE));
+			}
+		}
+	}
 
 	intel_ring_emit(ring, MI_NOOP);
 	intel_ring_emit(ring, MI_SET_CONTEXT);
 	intel_ring_emit(ring, i915_gem_obj_ggtt_offset(new_context->legacy_hw_ctx.rcs_state) |
-			MI_MM_SPACE_GTT |
-			MI_SAVE_EXT_STATE_EN |
-			MI_RESTORE_EXT_STATE_EN |
-			hw_flags);
+			flags);
 	/*
 	 * w/a: MI_SET_CONTEXT must always be followed by MI_NOOP
 	 * WaMiSetContext_Hang:snb,ivb,vlv
 	 */
 	intel_ring_emit(ring, MI_NOOP);
 
-	if (INTEL_INFO(ring->dev)->gen >= 7)
+	if (INTEL_INFO(ring->dev)->gen >= 7) {
+		if (num_rings) {
+			struct intel_engine_cs *signaller;
+
+			intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(num_rings));
+			for_each_ring(signaller, to_i915(ring->dev), i) {
+				if (signaller == ring)
+					continue;
+
+				intel_ring_emit(ring, RING_PSMI_CTL(signaller->mmio_base));
+				intel_ring_emit(ring, _MASKED_BIT_DISABLE(GEN6_PSMI_SLEEP_MSG_DISABLE));
+			}
+		}
 		intel_ring_emit(ring, MI_ARB_ON_OFF | MI_ARB_ENABLE);
-	else
-		intel_ring_emit(ring, MI_NOOP);
+	}
 
 	intel_ring_advance(ring);
 
@@ -614,9 +557,9 @@ static int do_switch(struct intel_engine_cs *ring,
 {
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 	struct intel_context *from = ring->last_context;
-	struct i915_hw_ppgtt *ppgtt = ctx_to_ppgtt(to);
 	u32 hw_flags = 0;
 	bool uninitialized = false;
+	struct i915_vma *vma;
 	int ret, i;
 
 	if (from != NULL && ring == &dev_priv->ring[RCS]) {
@@ -642,8 +585,9 @@ static int do_switch(struct intel_engine_cs *ring,
 	 */
 	from = ring->last_context;
 
-	if (USES_FULL_PPGTT(ring->dev)) {
-		ret = ppgtt->switch_mm(ppgtt, ring, false);
+	if (to->ppgtt) {
+		trace_switch_mm(ring, to);
+		ret = to->ppgtt->switch_mm(to->ppgtt, ring);
 		if (ret)
 			goto unpin_out;
 	}
@@ -666,11 +610,10 @@ static int do_switch(struct intel_engine_cs *ring,
 	if (ret)
 		goto unpin_out;
 
-	if (!to->legacy_hw_ctx.rcs_state->has_global_gtt_mapping) {
-		struct i915_vma *vma = i915_gem_obj_to_vma(to->legacy_hw_ctx.rcs_state,
-							   &dev_priv->gtt.base);
-		vma->bind_vma(vma, to->legacy_hw_ctx.rcs_state->cache_level, GLOBAL_BIND);
-	}
+	vma = i915_gem_obj_to_ggtt(to->legacy_hw_ctx.rcs_state);
+	if (!(vma->bound & GLOBAL_BIND))
+		vma->bind_vma(vma, to->legacy_hw_ctx.rcs_state->cache_level,
+				GLOBAL_BIND);
 
 	if (!to->legacy_hw_ctx.initialized || i915_gem_context_is_default(to))
 		hw_flags |= MI_RESTORE_INHIBIT;
@@ -723,6 +666,12 @@ done:
 	ring->last_context = to;
 
 	if (uninitialized) {
+		if (ring->init_context) {
+			ret = ring->init_context(ring, to);
+			if (ret)
+				DRM_ERROR("ring init context: %d\n", ret);
+		}
+
 		ret = i915_gem_render_state_init(ring);
 		if (ret)
 			DRM_ERROR("init render state: %d\n", ret);
@@ -743,14 +692,19 @@ unpin_out:
  *
  * The context life cycle is simple. The context refcount is incremented and
  * decremented by 1 and create and destroy. If the context is in use by the GPU,
- * it will have a refoucnt > 1. This allows us to destroy the context abstract
+ * it will have a refcount > 1. This allows us to destroy the context abstract
  * object while letting the normal object tracking destroy the backing BO.
+ *
+ * This function should not be used in execlists mode.  Instead the context is
+ * switched by writing to the ELSP and requests keep a reference to their
+ * context.
  */
 int i915_switch_context(struct intel_engine_cs *ring,
 			struct intel_context *to)
 {
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 
+	WARN_ON(i915.enable_execlists);
 	WARN_ON(!mutex_is_locked(&dev_priv->dev->struct_mutex));
 
 	if (to->legacy_hw_ctx.rcs_state == NULL) { /* We have the fake context */
@@ -766,9 +720,9 @@ int i915_switch_context(struct intel_engine_cs *ring,
 	return do_switch(ring, to);
 }
 
-static bool hw_context_enabled(struct drm_device *dev)
+static bool contexts_enabled(struct drm_device *dev)
 {
-	return to_i915(dev)->hw_context_size;
+	return i915.enable_execlists || to_i915(dev)->hw_context_size;
 }
 
 int i915_gem_context_create_ioctl(struct drm_device *dev, void *data,
@@ -779,14 +733,14 @@ int i915_gem_context_create_ioctl(struct drm_device *dev, void *data,
 	struct intel_context *ctx;
 	int ret;
 
-	if (!hw_context_enabled(dev))
+	if (!contexts_enabled(dev))
 		return -ENODEV;
 
 	ret = i915_mutex_lock_interruptible(dev);
 	if (ret)
 		return ret;
 
-	ctx = i915_gem_create_context(dev, file_priv, USES_FULL_PPGTT(dev));
+	ctx = i915_gem_create_context(dev, file_priv);
 	mutex_unlock(&dev->struct_mutex);
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c
index bbf4b12d842e..886ff2ee7a28 100644
--- a/drivers/gpu/drm/i915/i915_gem_evict.c
+++ b/drivers/gpu/drm/i915/i915_gem_evict.c
@@ -243,7 +243,7 @@ int
 i915_gem_evict_everything(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct i915_address_space *vm;
+	struct i915_address_space *vm, *v;
 	bool lists_empty = true;
 	int ret;
 
@@ -270,7 +270,7 @@ i915_gem_evict_everything(struct drm_device *dev)
 	i915_gem_retire_requests(dev);
 
 	/* Having flushed everything, unbind() should never raise an error */
-	list_for_each_entry(vm, &dev_priv->vm_list, global_link)
+	list_for_each_entry_safe(vm, v, &dev_priv->vm_list, global_link)
 		WARN_ON(i915_gem_evict_vm(vm, false));
 
 	return 0;
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index 60998fc4e5b2..11738316394a 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -35,6 +35,7 @@
 
 #define  __EXEC_OBJECT_HAS_PIN (1<<31)
 #define  __EXEC_OBJECT_HAS_FENCE (1<<30)
+#define  __EXEC_OBJECT_NEEDS_MAP (1<<29)
 #define  __EXEC_OBJECT_NEEDS_BIAS (1<<28)
 
 #define BATCH_OFFSET_BIAS (256*1024)
@@ -94,7 +95,6 @@ eb_lookup_vmas(struct eb_vmas *eb,
 	       struct i915_address_space *vm,
 	       struct drm_file *file)
 {
-	struct drm_i915_private *dev_priv = vm->dev->dev_private;
 	struct drm_i915_gem_object *obj;
 	struct list_head objects;
 	int i, ret;
@@ -129,20 +129,6 @@ eb_lookup_vmas(struct eb_vmas *eb,
 	i = 0;
 	while (!list_empty(&objects)) {
 		struct i915_vma *vma;
-		struct i915_address_space *bind_vm = vm;
-
-		if (exec[i].flags & EXEC_OBJECT_NEEDS_GTT &&
-		    USES_FULL_PPGTT(vm->dev)) {
-			ret = -EINVAL;
-			goto err;
-		}
-
-		/* If we have secure dispatch, or the userspace assures us that
-		 * they know what they're doing, use the GGTT VM.
-		 */
-		if (((args->flags & I915_EXEC_SECURE) &&
-		    (i == (args->buffer_count - 1))))
-			bind_vm = &dev_priv->gtt.base;
 
 		obj = list_first_entry(&objects,
 				       struct drm_i915_gem_object,
@@ -156,7 +142,7 @@ eb_lookup_vmas(struct eb_vmas *eb,
 		 * from the (obj, vm) we don't run the risk of creating
 		 * duplicated vmas for the same vm.
 		 */
-		vma = i915_gem_obj_lookup_or_create_vma(obj, bind_vm);
+		vma = i915_gem_obj_lookup_or_create_vma(obj, vm);
 		if (IS_ERR(vma)) {
 			DRM_DEBUG("Failed to lookup VMA\n");
 			ret = PTR_ERR(vma);
@@ -307,7 +293,7 @@ relocate_entry_gtt(struct drm_i915_gem_object *obj,
 	struct drm_device *dev = obj->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint64_t delta = reloc->delta + target_offset;
-	uint32_t __iomem *reloc_entry;
+	uint64_t offset;
 	void __iomem *reloc_page;
 	int ret;
 
@@ -320,25 +306,24 @@ relocate_entry_gtt(struct drm_i915_gem_object *obj,
 		return ret;
 
 	/* Map the page containing the relocation we're going to perform.  */
-	reloc->offset += i915_gem_obj_ggtt_offset(obj);
+	offset = i915_gem_obj_ggtt_offset(obj);
+	offset += reloc->offset;
 	reloc_page = io_mapping_map_atomic_wc(dev_priv->gtt.mappable,
-			reloc->offset & PAGE_MASK);
-	reloc_entry = (uint32_t __iomem *)
-		(reloc_page + offset_in_page(reloc->offset));
-	iowrite32(lower_32_bits(delta), reloc_entry);
+					      offset & PAGE_MASK);
+	iowrite32(lower_32_bits(delta), reloc_page + offset_in_page(offset));
 
 	if (INTEL_INFO(dev)->gen >= 8) {
-		reloc_entry += 1;
+		offset += sizeof(uint32_t);
 
-		if (offset_in_page(reloc->offset + sizeof(uint32_t)) == 0) {
+		if (offset_in_page(offset) == 0) {
 			io_mapping_unmap_atomic(reloc_page);
-			reloc_page = io_mapping_map_atomic_wc(
-					dev_priv->gtt.mappable,
-					reloc->offset + sizeof(uint32_t));
-			reloc_entry = reloc_page;
+			reloc_page =
+				io_mapping_map_atomic_wc(dev_priv->gtt.mappable,
+							 offset);
 		}
 
-		iowrite32(upper_32_bits(delta), reloc_entry);
+		iowrite32(upper_32_bits(delta),
+			  reloc_page + offset_in_page(offset));
 	}
 
 	io_mapping_unmap_atomic(reloc_page);
@@ -372,12 +357,9 @@ i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj,
 	 * through the ppgtt for non_secure batchbuffers. */
 	if (unlikely(IS_GEN6(dev) &&
 	    reloc->write_domain == I915_GEM_DOMAIN_INSTRUCTION &&
-	    !target_i915_obj->has_global_gtt_mapping)) {
-		struct i915_vma *vma =
-			list_first_entry(&target_i915_obj->vma_list,
-					 typeof(*vma), vma_link);
-		vma->bind_vma(vma, target_i915_obj->cache_level, GLOBAL_BIND);
-	}
+	    !(target_vma->bound & GLOBAL_BIND)))
+		target_vma->bind_vma(target_vma, target_i915_obj->cache_level,
+				GLOBAL_BIND);
 
 	/* Validate that the target is in a valid r/w GPU domain */
 	if (unlikely(reloc->write_domain & (reloc->write_domain - 1))) {
@@ -535,34 +517,18 @@ i915_gem_execbuffer_relocate(struct eb_vmas *eb)
 }
 
 static int
-need_reloc_mappable(struct i915_vma *vma)
-{
-	struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
-	return entry->relocation_count && !use_cpu_reloc(vma->obj) &&
-		i915_is_ggtt(vma->vm);
-}
-
-static int
 i915_gem_execbuffer_reserve_vma(struct i915_vma *vma,
 				struct intel_engine_cs *ring,
 				bool *need_reloc)
 {
 	struct drm_i915_gem_object *obj = vma->obj;
 	struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
-	bool has_fenced_gpu_access = INTEL_INFO(ring->dev)->gen < 4;
-	bool need_fence;
 	uint64_t flags;
 	int ret;
 
 	flags = 0;
-
-	need_fence =
-		has_fenced_gpu_access &&
-		entry->flags & EXEC_OBJECT_NEEDS_FENCE &&
-		obj->tiling_mode != I915_TILING_NONE;
-	if (need_fence || need_reloc_mappable(vma))
-		flags |= PIN_MAPPABLE;
-
+	if (entry->flags & __EXEC_OBJECT_NEEDS_MAP)
+		flags |= PIN_GLOBAL | PIN_MAPPABLE;
 	if (entry->flags & EXEC_OBJECT_NEEDS_GTT)
 		flags |= PIN_GLOBAL;
 	if (entry->flags & __EXEC_OBJECT_NEEDS_BIAS)
@@ -574,17 +540,13 @@ i915_gem_execbuffer_reserve_vma(struct i915_vma *vma,
 
 	entry->flags |= __EXEC_OBJECT_HAS_PIN;
 
-	if (has_fenced_gpu_access) {
-		if (entry->flags & EXEC_OBJECT_NEEDS_FENCE) {
-			ret = i915_gem_object_get_fence(obj);
-			if (ret)
-				return ret;
-
-			if (i915_gem_object_pin_fence(obj))
-				entry->flags |= __EXEC_OBJECT_HAS_FENCE;
+	if (entry->flags & EXEC_OBJECT_NEEDS_FENCE) {
+		ret = i915_gem_object_get_fence(obj);
+		if (ret)
+			return ret;
 
-			obj->pending_fenced_gpu_access = true;
-		}
+		if (i915_gem_object_pin_fence(obj))
+			entry->flags |= __EXEC_OBJECT_HAS_FENCE;
 	}
 
 	if (entry->offset != vma->node.start) {
@@ -601,26 +563,40 @@ i915_gem_execbuffer_reserve_vma(struct i915_vma *vma,
 }
 
 static bool
-eb_vma_misplaced(struct i915_vma *vma, bool has_fenced_gpu_access)
+need_reloc_mappable(struct i915_vma *vma)
 {
 	struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
-	struct drm_i915_gem_object *obj = vma->obj;
-	bool need_fence, need_mappable;
 
-	need_fence =
-		has_fenced_gpu_access &&
-		entry->flags & EXEC_OBJECT_NEEDS_FENCE &&
-		obj->tiling_mode != I915_TILING_NONE;
-	need_mappable = need_fence || need_reloc_mappable(vma);
+	if (entry->relocation_count == 0)
+		return false;
 
-	WARN_ON((need_mappable || need_fence) &&
+	if (!i915_is_ggtt(vma->vm))
+		return false;
+
+	/* See also use_cpu_reloc() */
+	if (HAS_LLC(vma->obj->base.dev))
+		return false;
+
+	if (vma->obj->base.write_domain == I915_GEM_DOMAIN_CPU)
+		return false;
+
+	return true;
+}
+
+static bool
+eb_vma_misplaced(struct i915_vma *vma)
+{
+	struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
+	struct drm_i915_gem_object *obj = vma->obj;
+
+	WARN_ON(entry->flags & __EXEC_OBJECT_NEEDS_MAP &&
 	       !i915_is_ggtt(vma->vm));
 
 	if (entry->alignment &&
 	    vma->node.start & (entry->alignment - 1))
 		return true;
 
-	if (need_mappable && !obj->map_and_fenceable)
+	if (entry->flags & __EXEC_OBJECT_NEEDS_MAP && !obj->map_and_fenceable)
 		return true;
 
 	if (entry->flags & __EXEC_OBJECT_NEEDS_BIAS &&
@@ -642,9 +618,6 @@ i915_gem_execbuffer_reserve(struct intel_engine_cs *ring,
 	bool has_fenced_gpu_access = INTEL_INFO(ring->dev)->gen < 4;
 	int retry;
 
-	if (list_empty(vmas))
-		return 0;
-
 	i915_gem_retire_requests_ring(ring);
 
 	vm = list_first_entry(vmas, struct i915_vma, exec_list)->vm;
@@ -658,20 +631,21 @@ i915_gem_execbuffer_reserve(struct intel_engine_cs *ring,
 		obj = vma->obj;
 		entry = vma->exec_entry;
 
+		if (!has_fenced_gpu_access)
+			entry->flags &= ~EXEC_OBJECT_NEEDS_FENCE;
 		need_fence =
-			has_fenced_gpu_access &&
 			entry->flags & EXEC_OBJECT_NEEDS_FENCE &&
 			obj->tiling_mode != I915_TILING_NONE;
 		need_mappable = need_fence || need_reloc_mappable(vma);
 
-		if (need_mappable)
+		if (need_mappable) {
+			entry->flags |= __EXEC_OBJECT_NEEDS_MAP;
 			list_move(&vma->exec_list, &ordered_vmas);
-		else
+		} else
 			list_move_tail(&vma->exec_list, &ordered_vmas);
 
 		obj->base.pending_read_domains = I915_GEM_GPU_DOMAINS & ~I915_GEM_DOMAIN_COMMAND;
 		obj->base.pending_write_domain = 0;
-		obj->pending_fenced_gpu_access = false;
 	}
 	list_splice(&ordered_vmas, vmas);
 
@@ -696,7 +670,7 @@ i915_gem_execbuffer_reserve(struct intel_engine_cs *ring,
 			if (!drm_mm_node_allocated(&vma->node))
 				continue;
 
-			if (eb_vma_misplaced(vma, has_fenced_gpu_access))
+			if (eb_vma_misplaced(vma))
 				ret = i915_vma_unbind(vma);
 			else
 				ret = i915_gem_execbuffer_reserve_vma(vma, ring, need_relocs);
@@ -744,9 +718,6 @@ i915_gem_execbuffer_relocate_slow(struct drm_device *dev,
 	int i, total, ret;
 	unsigned count = args->buffer_count;
 
-	if (WARN_ON(list_empty(&eb->vmas)))
-		return 0;
-
 	vm = list_first_entry(&eb->vmas, struct i915_vma, exec_list)->vm;
 
 	/* We may process another execbuffer during the unlock... */
@@ -890,18 +861,24 @@ i915_gem_check_execbuffer(struct drm_i915_gem_execbuffer2 *exec)
 }
 
 static int
-validate_exec_list(struct drm_i915_gem_exec_object2 *exec,
+validate_exec_list(struct drm_device *dev,
+		   struct drm_i915_gem_exec_object2 *exec,
 		   int count)
 {
-	int i;
 	unsigned relocs_total = 0;
 	unsigned relocs_max = UINT_MAX / sizeof(struct drm_i915_gem_relocation_entry);
+	unsigned invalid_flags;
+	int i;
+
+	invalid_flags = __EXEC_OBJECT_UNKNOWN_FLAGS;
+	if (USES_FULL_PPGTT(dev))
+		invalid_flags |= EXEC_OBJECT_NEEDS_GTT;
 
 	for (i = 0; i < count; i++) {
 		char __user *ptr = to_user_ptr(exec[i].relocs_ptr);
 		int length; /* limited by fault_in_pages_readable() */
 
-		if (exec[i].flags & __EXEC_OBJECT_UNKNOWN_FLAGS)
+		if (exec[i].flags & invalid_flags)
 			return -EINVAL;
 
 		/* First check for malicious input causing overflow in
@@ -951,16 +928,26 @@ i915_gem_validate_context(struct drm_device *dev, struct drm_file *file,
 		return ERR_PTR(-EIO);
 	}
 
+	if (i915.enable_execlists && !ctx->engine[ring->id].state) {
+		int ret = intel_lr_context_deferred_create(ctx, ring);
+		if (ret) {
+			DRM_DEBUG("Could not create LRC %u: %d\n", ctx_id, ret);
+			return ERR_PTR(ret);
+		}
+	}
+
 	return ctx;
 }
 
-static void
+void
 i915_gem_execbuffer_move_to_active(struct list_head *vmas,
 				   struct intel_engine_cs *ring)
 {
+	u32 seqno = intel_ring_get_seqno(ring);
 	struct i915_vma *vma;
 
 	list_for_each_entry(vma, vmas, exec_list) {
+		struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
 		struct drm_i915_gem_object *obj = vma->obj;
 		u32 old_read = obj->base.read_domains;
 		u32 old_write = obj->base.write_domain;
@@ -969,24 +956,31 @@ i915_gem_execbuffer_move_to_active(struct list_head *vmas,
 		if (obj->base.write_domain == 0)
 			obj->base.pending_read_domains |= obj->base.read_domains;
 		obj->base.read_domains = obj->base.pending_read_domains;
-		obj->fenced_gpu_access = obj->pending_fenced_gpu_access;
 
 		i915_vma_move_to_active(vma, ring);
 		if (obj->base.write_domain) {
 			obj->dirty = 1;
-			obj->last_write_seqno = intel_ring_get_seqno(ring);
+			obj->last_write_seqno = seqno;
 
 			intel_fb_obj_invalidate(obj, ring);
 
 			/* update for the implicit flush after a batch */
 			obj->base.write_domain &= ~I915_GEM_GPU_DOMAINS;
 		}
+		if (entry->flags & EXEC_OBJECT_NEEDS_FENCE) {
+			obj->last_fenced_seqno = seqno;
+			if (entry->flags & __EXEC_OBJECT_HAS_FENCE) {
+				struct drm_i915_private *dev_priv = to_i915(ring->dev);
+				list_move_tail(&dev_priv->fence_regs[obj->fence_reg].lru_list,
+					       &dev_priv->mm.fence_list);
+			}
+		}
 
 		trace_i915_gem_object_change_domain(obj, old_read, old_write);
 	}
 }
 
-static void
+void
 i915_gem_execbuffer_retire_commands(struct drm_device *dev,
 				    struct drm_file *file,
 				    struct intel_engine_cs *ring,
@@ -1027,13 +1021,54 @@ i915_reset_gen7_sol_offsets(struct drm_device *dev,
 }
 
 static int
-legacy_ringbuffer_submission(struct drm_device *dev, struct drm_file *file,
-			     struct intel_engine_cs *ring,
-			     struct intel_context *ctx,
-			     struct drm_i915_gem_execbuffer2 *args,
-			     struct list_head *vmas,
-			     struct drm_i915_gem_object *batch_obj,
-			     u64 exec_start, u32 flags)
+i915_emit_box(struct intel_engine_cs *ring,
+	      struct drm_clip_rect *box,
+	      int DR1, int DR4)
+{
+	int ret;
+
+	if (box->y2 <= box->y1 || box->x2 <= box->x1 ||
+	    box->y2 <= 0 || box->x2 <= 0) {
+		DRM_ERROR("Bad box %d,%d..%d,%d\n",
+			  box->x1, box->y1, box->x2, box->y2);
+		return -EINVAL;
+	}
+
+	if (INTEL_INFO(ring->dev)->gen >= 4) {
+		ret = intel_ring_begin(ring, 4);
+		if (ret)
+			return ret;
+
+		intel_ring_emit(ring, GFX_OP_DRAWRECT_INFO_I965);
+		intel_ring_emit(ring, (box->x1 & 0xffff) | box->y1 << 16);
+		intel_ring_emit(ring, ((box->x2 - 1) & 0xffff) | (box->y2 - 1) << 16);
+		intel_ring_emit(ring, DR4);
+	} else {
+		ret = intel_ring_begin(ring, 6);
+		if (ret)
+			return ret;
+
+		intel_ring_emit(ring, GFX_OP_DRAWRECT_INFO);
+		intel_ring_emit(ring, DR1);
+		intel_ring_emit(ring, (box->x1 & 0xffff) | box->y1 << 16);
+		intel_ring_emit(ring, ((box->x2 - 1) & 0xffff) | (box->y2 - 1) << 16);
+		intel_ring_emit(ring, DR4);
+		intel_ring_emit(ring, 0);
+	}
+	intel_ring_advance(ring);
+
+	return 0;
+}
+
+
+int
+i915_gem_ringbuffer_submission(struct drm_device *dev, struct drm_file *file,
+			       struct intel_engine_cs *ring,
+			       struct intel_context *ctx,
+			       struct drm_i915_gem_execbuffer2 *args,
+			       struct list_head *vmas,
+			       struct drm_i915_gem_object *batch_obj,
+			       u64 exec_start, u32 flags)
 {
 	struct drm_clip_rect *cliprects = NULL;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1154,7 +1189,7 @@ legacy_ringbuffer_submission(struct drm_device *dev, struct drm_file *file,
 	exec_len = args->batch_len;
 	if (cliprects) {
 		for (i = 0; i < args->num_cliprects; i++) {
-			ret = i915_emit_box(dev, &cliprects[i],
+			ret = i915_emit_box(ring, &cliprects[i],
 					    args->DR1, args->DR4);
 			if (ret)
 				goto error;
@@ -1254,7 +1289,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 	if (!i915_gem_check_execbuffer(args))
 		return -EINVAL;
 
-	ret = validate_exec_list(exec, args->buffer_count);
+	ret = validate_exec_list(dev, exec, args->buffer_count);
 	if (ret)
 		return ret;
 
@@ -1303,12 +1338,6 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 	if (ret)
 		goto pre_mutex_err;
 
-	if (dev_priv->ums.mm_suspended) {
-		mutex_unlock(&dev->struct_mutex);
-		ret = -EBUSY;
-		goto pre_mutex_err;
-	}
-
 	ctx = i915_gem_validate_context(dev, file, ring, ctx_id);
 	if (IS_ERR(ctx)) {
 		mutex_unlock(&dev->struct_mutex);
@@ -1318,8 +1347,9 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 
 	i915_gem_context_reference(ctx);
 
-	vm = ctx->vm;
-	if (!USES_FULL_PPGTT(dev))
+	if (ctx->ppgtt)
+		vm = &ctx->ppgtt->base;
+	else
 		vm = &dev_priv->gtt.base;
 
 	eb = eb_create(args);
@@ -1370,41 +1400,54 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 				      batch_obj,
 				      args->batch_start_offset,
 				      file->is_master);
-		if (ret)
-			goto err;
-
-		/*
-		 * XXX: Actually do this when enabling batch copy...
-		 *
-		 * Set the DISPATCH_SECURE bit to remove the NON_SECURE bit
-		 * from MI_BATCH_BUFFER_START commands issued in the
-		 * dispatch_execbuffer implementations. We specifically don't
-		 * want that set when the command parser is enabled.
-		 */
+		if (ret) {
+			if (ret != -EACCES)
+				goto err;
+		} else {
+			/*
+			 * XXX: Actually do this when enabling batch copy...
+			 *
+			 * Set the DISPATCH_SECURE bit to remove the NON_SECURE bit
+			 * from MI_BATCH_BUFFER_START commands issued in the
+			 * dispatch_execbuffer implementations. We specifically don't
+			 * want that set when the command parser is enabled.
+			 */
+		}
 	}
 
 	/* snb/ivb/vlv conflate the "batch in ppgtt" bit with the "non-secure
 	 * batch" bit. Hence we need to pin secure batches into the global gtt.
 	 * hsw should have this fixed, but bdw mucks it up again. */
-	if (flags & I915_DISPATCH_SECURE &&
-	    !batch_obj->has_global_gtt_mapping) {
-		/* When we have multiple VMs, we'll need to make sure that we
-		 * allocate space first */
-		struct i915_vma *vma = i915_gem_obj_to_ggtt(batch_obj);
-		BUG_ON(!vma);
-		vma->bind_vma(vma, batch_obj->cache_level, GLOBAL_BIND);
-	}
+	if (flags & I915_DISPATCH_SECURE) {
+		/*
+		 * So on first glance it looks freaky that we pin the batch here
+		 * outside of the reservation loop. But:
+		 * - The batch is already pinned into the relevant ppgtt, so we
+		 *   already have the backing storage fully allocated.
+		 * - No other BO uses the global gtt (well contexts, but meh),
+		 *   so we don't really have issues with mutliple objects not
+		 *   fitting due to fragmentation.
+		 * So this is actually safe.
+		 */
+		ret = i915_gem_obj_ggtt_pin(batch_obj, 0, 0);
+		if (ret)
+			goto err;
 
-	if (flags & I915_DISPATCH_SECURE)
 		exec_start += i915_gem_obj_ggtt_offset(batch_obj);
-	else
+	} else
 		exec_start += i915_gem_obj_offset(batch_obj, vm);
 
-	ret = legacy_ringbuffer_submission(dev, file, ring, ctx,
-			args, &eb->vmas, batch_obj, exec_start, flags);
-	if (ret)
-		goto err;
+	ret = dev_priv->gt.do_execbuf(dev, file, ring, ctx, args,
+				      &eb->vmas, batch_obj, exec_start, flags);
 
+	/*
+	 * FIXME: We crucially rely upon the active tracking for the (ppgtt)
+	 * batch vma for correctness. For less ugly and less fragility this
+	 * needs to be adjusted to also track the ggtt batch vma properly as
+	 * active.
+	 */
+	if (flags & I915_DISPATCH_SECURE)
+		i915_gem_object_ggtt_unpin(batch_obj);
 err:
 	/* the request owns the ref now */
 	i915_gem_context_unreference(ctx);
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index e42925f76b4b..171f6eafdeee 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -33,26 +33,28 @@
 static void bdw_setup_private_ppat(struct drm_i915_private *dev_priv);
 static void chv_setup_private_ppat(struct drm_i915_private *dev_priv);
 
-bool intel_enable_ppgtt(struct drm_device *dev, bool full)
+static int sanitize_enable_ppgtt(struct drm_device *dev, int enable_ppgtt)
 {
-	if (i915.enable_ppgtt == 0)
-		return false;
-
-	if (i915.enable_ppgtt == 1 && full)
-		return false;
+	bool has_aliasing_ppgtt;
+	bool has_full_ppgtt;
 
-	return true;
-}
+	has_aliasing_ppgtt = INTEL_INFO(dev)->gen >= 6;
+	has_full_ppgtt = INTEL_INFO(dev)->gen >= 7;
+	if (IS_GEN8(dev))
+		has_full_ppgtt = false; /* XXX why? */
 
-static int sanitize_enable_ppgtt(struct drm_device *dev, int enable_ppgtt)
-{
-	if (enable_ppgtt == 0 || !HAS_ALIASING_PPGTT(dev))
+	/*
+	 * We don't allow disabling PPGTT for gen9+ as it's a requirement for
+	 * execlists, the sole mechanism available to submit work.
+	 */
+	if (INTEL_INFO(dev)->gen < 9 &&
+	    (enable_ppgtt == 0 || !has_aliasing_ppgtt))
 		return 0;
 
 	if (enable_ppgtt == 1)
 		return 1;
 
-	if (enable_ppgtt == 2 && HAS_PPGTT(dev))
+	if (enable_ppgtt == 2 && has_full_ppgtt)
 		return 2;
 
 #ifdef CONFIG_INTEL_IOMMU
@@ -70,7 +72,7 @@ static int sanitize_enable_ppgtt(struct drm_device *dev, int enable_ppgtt)
 		return 0;
 	}
 
-	return HAS_ALIASING_PPGTT(dev) ? 1 : 0;
+	return has_aliasing_ppgtt ? 1 : 0;
 }
 
 
@@ -78,7 +80,6 @@ static void ppgtt_bind_vma(struct i915_vma *vma,
 			   enum i915_cache_level cache_level,
 			   u32 flags);
 static void ppgtt_unbind_vma(struct i915_vma *vma);
-static int gen8_ppgtt_enable(struct i915_hw_ppgtt *ppgtt);
 
 static inline gen8_gtt_pte_t gen8_pte_encode(dma_addr_t addr,
 					     enum i915_cache_level level,
@@ -168,9 +169,6 @@ static gen6_gtt_pte_t byt_pte_encode(dma_addr_t addr,
 	gen6_gtt_pte_t pte = valid ? GEN6_PTE_VALID : 0;
 	pte |= GEN6_PTE_ADDR_ENCODE(addr);
 
-	/* Mark the page as writeable.  Other platforms don't have a
-	 * setting for read-only/writable, so this matches that behavior.
-	 */
 	if (!(flags & PTE_READ_ONLY))
 		pte |= BYT_PTE_WRITEABLE;
 
@@ -216,19 +214,12 @@ static gen6_gtt_pte_t iris_pte_encode(dma_addr_t addr,
 
 /* Broadwell Page Directory Pointer Descriptors */
 static int gen8_write_pdp(struct intel_engine_cs *ring, unsigned entry,
-			   uint64_t val, bool synchronous)
+			   uint64_t val)
 {
-	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 	int ret;
 
 	BUG_ON(entry >= 4);
 
-	if (synchronous) {
-		I915_WRITE(GEN8_RING_PDP_UDW(ring, entry), val >> 32);
-		I915_WRITE(GEN8_RING_PDP_LDW(ring, entry), (u32)val);
-		return 0;
-	}
-
 	ret = intel_ring_begin(ring, 6);
 	if (ret)
 		return ret;
@@ -245,8 +236,7 @@ static int gen8_write_pdp(struct intel_engine_cs *ring, unsigned entry,
 }
 
 static int gen8_mm_switch(struct i915_hw_ppgtt *ppgtt,
-			  struct intel_engine_cs *ring,
-			  bool synchronous)
+			  struct intel_engine_cs *ring)
 {
 	int i, ret;
 
@@ -255,7 +245,7 @@ static int gen8_mm_switch(struct i915_hw_ppgtt *ppgtt,
 
 	for (i = used_pd - 1; i >= 0; i--) {
 		dma_addr_t addr = ppgtt->pd_dma_addr[i];
-		ret = gen8_write_pdp(ring, i, addr, synchronous);
+		ret = gen8_write_pdp(ring, i, addr);
 		if (ret)
 			return ret;
 	}
@@ -403,9 +393,6 @@ static void gen8_ppgtt_cleanup(struct i915_address_space *vm)
 	struct i915_hw_ppgtt *ppgtt =
 		container_of(vm, struct i915_hw_ppgtt, base);
 
-	list_del(&vm->global_link);
-	drm_mm_takedown(&vm->mm);
-
 	gen8_ppgtt_unmap_pages(ppgtt);
 	gen8_ppgtt_free(ppgtt);
 }
@@ -615,7 +602,6 @@ static int gen8_ppgtt_init(struct i915_hw_ppgtt *ppgtt, uint64_t size)
 		kunmap_atomic(pd_vaddr);
 	}
 
-	ppgtt->enable = gen8_ppgtt_enable;
 	ppgtt->switch_mm = gen8_mm_switch;
 	ppgtt->base.clear_range = gen8_ppgtt_clear_range;
 	ppgtt->base.insert_entries = gen8_ppgtt_insert_entries;
@@ -724,29 +710,10 @@ static uint32_t get_pd_offset(struct i915_hw_ppgtt *ppgtt)
 }
 
 static int hsw_mm_switch(struct i915_hw_ppgtt *ppgtt,
-			 struct intel_engine_cs *ring,
-			 bool synchronous)
+			 struct intel_engine_cs *ring)
 {
-	struct drm_device *dev = ppgtt->base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
-	/* If we're in reset, we can assume the GPU is sufficiently idle to
-	 * manually frob these bits. Ideally we could use the ring functions,
-	 * except our error handling makes it quite difficult (can't use
-	 * intel_ring_begin, ring->flush, or intel_ring_advance)
-	 *
-	 * FIXME: We should try not to special case reset
-	 */
-	if (synchronous ||
-	    i915_reset_in_progress(&dev_priv->gpu_error)) {
-		WARN_ON(ppgtt != dev_priv->mm.aliasing_ppgtt);
-		I915_WRITE(RING_PP_DIR_DCLV(ring), PP_DIR_DCLV_2G);
-		I915_WRITE(RING_PP_DIR_BASE(ring), get_pd_offset(ppgtt));
-		POSTING_READ(RING_PP_DIR_BASE(ring));
-		return 0;
-	}
-
 	/* NB: TLBs must be flushed and invalidated before a switch */
 	ret = ring->flush(ring, I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
 	if (ret)
@@ -768,29 +735,10 @@ static int hsw_mm_switch(struct i915_hw_ppgtt *ppgtt,
 }
 
 static int gen7_mm_switch(struct i915_hw_ppgtt *ppgtt,
-			  struct intel_engine_cs *ring,
-			  bool synchronous)
+			  struct intel_engine_cs *ring)
 {
-	struct drm_device *dev = ppgtt->base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
-	/* If we're in reset, we can assume the GPU is sufficiently idle to
-	 * manually frob these bits. Ideally we could use the ring functions,
-	 * except our error handling makes it quite difficult (can't use
-	 * intel_ring_begin, ring->flush, or intel_ring_advance)
-	 *
-	 * FIXME: We should try not to special case reset
-	 */
-	if (synchronous ||
-	    i915_reset_in_progress(&dev_priv->gpu_error)) {
-		WARN_ON(ppgtt != dev_priv->mm.aliasing_ppgtt);
-		I915_WRITE(RING_PP_DIR_DCLV(ring), PP_DIR_DCLV_2G);
-		I915_WRITE(RING_PP_DIR_BASE(ring), get_pd_offset(ppgtt));
-		POSTING_READ(RING_PP_DIR_BASE(ring));
-		return 0;
-	}
-
 	/* NB: TLBs must be flushed and invalidated before a switch */
 	ret = ring->flush(ring, I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
 	if (ret)
@@ -819,14 +767,11 @@ static int gen7_mm_switch(struct i915_hw_ppgtt *ppgtt,
 }
 
 static int gen6_mm_switch(struct i915_hw_ppgtt *ppgtt,
-			  struct intel_engine_cs *ring,
-			  bool synchronous)
+			  struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ppgtt->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (!synchronous)
-		return 0;
 
 	I915_WRITE(RING_PP_DIR_DCLV(ring), PP_DIR_DCLV_2G);
 	I915_WRITE(RING_PP_DIR_BASE(ring), get_pd_offset(ppgtt));
@@ -836,39 +781,20 @@ static int gen6_mm_switch(struct i915_hw_ppgtt *ppgtt,
 	return 0;
 }
 
-static int gen8_ppgtt_enable(struct i915_hw_ppgtt *ppgtt)
+static void gen8_ppgtt_enable(struct drm_device *dev)
 {
-	struct drm_device *dev = ppgtt->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_engine_cs *ring;
-	int j, ret;
+	int j;
 
 	for_each_ring(ring, dev_priv, j) {
 		I915_WRITE(RING_MODE_GEN7(ring),
 			   _MASKED_BIT_ENABLE(GFX_PPGTT_ENABLE));
-
-		/* We promise to do a switch later with FULL PPGTT. If this is
-		 * aliasing, this is the one and only switch we'll do */
-		if (USES_FULL_PPGTT(dev))
-			continue;
-
-		ret = ppgtt->switch_mm(ppgtt, ring, true);
-		if (ret)
-			goto err_out;
 	}
-
-	return 0;
-
-err_out:
-	for_each_ring(ring, dev_priv, j)
-		I915_WRITE(RING_MODE_GEN7(ring),
-			   _MASKED_BIT_DISABLE(GFX_PPGTT_ENABLE));
-	return ret;
 }
 
-static int gen7_ppgtt_enable(struct i915_hw_ppgtt *ppgtt)
+static void gen7_ppgtt_enable(struct drm_device *dev)
 {
-	struct drm_device *dev = ppgtt->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_engine_cs *ring;
 	uint32_t ecochk, ecobits;
@@ -887,31 +813,16 @@ static int gen7_ppgtt_enable(struct i915_hw_ppgtt *ppgtt)
 	I915_WRITE(GAM_ECOCHK, ecochk);
 
 	for_each_ring(ring, dev_priv, i) {
-		int ret;
 		/* GFX_MODE is per-ring on gen7+ */
 		I915_WRITE(RING_MODE_GEN7(ring),
 			   _MASKED_BIT_ENABLE(GFX_PPGTT_ENABLE));
-
-		/* We promise to do a switch later with FULL PPGTT. If this is
-		 * aliasing, this is the one and only switch we'll do */
-		if (USES_FULL_PPGTT(dev))
-			continue;
-
-		ret = ppgtt->switch_mm(ppgtt, ring, true);
-		if (ret)
-			return ret;
 	}
-
-	return 0;
 }
 
-static int gen6_ppgtt_enable(struct i915_hw_ppgtt *ppgtt)
+static void gen6_ppgtt_enable(struct drm_device *dev)
 {
-	struct drm_device *dev = ppgtt->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_engine_cs *ring;
 	uint32_t ecochk, gab_ctl, ecobits;
-	int i;
 
 	ecobits = I915_READ(GAC_ECO_BITS);
 	I915_WRITE(GAC_ECO_BITS, ecobits | ECOBITS_SNB_BIT |
@@ -924,14 +835,6 @@ static int gen6_ppgtt_enable(struct i915_hw_ppgtt *ppgtt)
 	I915_WRITE(GAM_ECOCHK, ecochk | ECOCHK_SNB_BIT | ECOCHK_PPGTT_CACHE64B);
 
 	I915_WRITE(GFX_MODE, _MASKED_BIT_ENABLE(GFX_PPGTT_ENABLE));
-
-	for_each_ring(ring, dev_priv, i) {
-		int ret = ppgtt->switch_mm(ppgtt, ring, true);
-		if (ret)
-			return ret;
-	}
-
-	return 0;
 }
 
 /* PPGTT support for Sandybdrige/Gen6 and later */
@@ -1029,8 +932,6 @@ static void gen6_ppgtt_cleanup(struct i915_address_space *vm)
 	struct i915_hw_ppgtt *ppgtt =
 		container_of(vm, struct i915_hw_ppgtt, base);
 
-	list_del(&vm->global_link);
-	drm_mm_takedown(&ppgtt->base.mm);
 	drm_mm_remove_node(&ppgtt->node);
 
 	gen6_ppgtt_unmap_pages(ppgtt);
@@ -1151,13 +1052,10 @@ static int gen6_ppgtt_init(struct i915_hw_ppgtt *ppgtt)
 
 	ppgtt->base.pte_encode = dev_priv->gtt.base.pte_encode;
 	if (IS_GEN6(dev)) {
-		ppgtt->enable = gen6_ppgtt_enable;
 		ppgtt->switch_mm = gen6_mm_switch;
 	} else if (IS_HASWELL(dev)) {
-		ppgtt->enable = gen7_ppgtt_enable;
 		ppgtt->switch_mm = hsw_mm_switch;
 	} else if (IS_GEN7(dev)) {
-		ppgtt->enable = gen7_ppgtt_enable;
 		ppgtt->switch_mm = gen7_mm_switch;
 	} else
 		BUG();
@@ -1188,39 +1086,118 @@ static int gen6_ppgtt_init(struct i915_hw_ppgtt *ppgtt)
 			 ppgtt->node.size >> 20,
 			 ppgtt->node.start / PAGE_SIZE);
 
+	gen6_write_pdes(ppgtt);
+	DRM_DEBUG("Adding PPGTT at offset %x\n",
+		  ppgtt->pd_offset << 10);
+
 	return 0;
 }
 
-int i915_gem_init_ppgtt(struct drm_device *dev, struct i915_hw_ppgtt *ppgtt)
+static int __hw_ppgtt_init(struct drm_device *dev, struct i915_hw_ppgtt *ppgtt)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	int ret = 0;
 
 	ppgtt->base.dev = dev;
 	ppgtt->base.scratch = dev_priv->gtt.base.scratch;
 
 	if (INTEL_INFO(dev)->gen < 8)
-		ret = gen6_ppgtt_init(ppgtt);
-	else if (IS_GEN8(dev))
-		ret = gen8_ppgtt_init(ppgtt, dev_priv->gtt.base.total);
+		return gen6_ppgtt_init(ppgtt);
+	else if (IS_GEN8(dev) || IS_GEN9(dev))
+		return gen8_ppgtt_init(ppgtt, dev_priv->gtt.base.total);
 	else
 		BUG();
+}
+int i915_ppgtt_init(struct drm_device *dev, struct i915_hw_ppgtt *ppgtt)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret = 0;
 
-	if (!ret) {
-		struct drm_i915_private *dev_priv = dev->dev_private;
+	ret = __hw_ppgtt_init(dev, ppgtt);
+	if (ret == 0) {
 		kref_init(&ppgtt->ref);
 		drm_mm_init(&ppgtt->base.mm, ppgtt->base.start,
 			    ppgtt->base.total);
 		i915_init_vm(dev_priv, &ppgtt->base);
-		if (INTEL_INFO(dev)->gen < 8) {
-			gen6_write_pdes(ppgtt);
-			DRM_DEBUG("Adding PPGTT at offset %x\n",
-				  ppgtt->pd_offset << 10);
+	}
+
+	return ret;
+}
+
+int i915_ppgtt_init_hw(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
+	struct i915_hw_ppgtt *ppgtt = dev_priv->mm.aliasing_ppgtt;
+	int i, ret = 0;
+
+	/* In the case of execlists, PPGTT is enabled by the context descriptor
+	 * and the PDPs are contained within the context itself.  We don't
+	 * need to do anything here. */
+	if (i915.enable_execlists)
+		return 0;
+
+	if (!USES_PPGTT(dev))
+		return 0;
+
+	if (IS_GEN6(dev))
+		gen6_ppgtt_enable(dev);
+	else if (IS_GEN7(dev))
+		gen7_ppgtt_enable(dev);
+	else if (INTEL_INFO(dev)->gen >= 8)
+		gen8_ppgtt_enable(dev);
+	else
+		WARN_ON(1);
+
+	if (ppgtt) {
+		for_each_ring(ring, dev_priv, i) {
+			ret = ppgtt->switch_mm(ppgtt, ring);
+			if (ret != 0)
+				return ret;
 		}
 	}
 
 	return ret;
 }
+struct i915_hw_ppgtt *
+i915_ppgtt_create(struct drm_device *dev, struct drm_i915_file_private *fpriv)
+{
+	struct i915_hw_ppgtt *ppgtt;
+	int ret;
+
+	ppgtt = kzalloc(sizeof(*ppgtt), GFP_KERNEL);
+	if (!ppgtt)
+		return ERR_PTR(-ENOMEM);
+
+	ret = i915_ppgtt_init(dev, ppgtt);
+	if (ret) {
+		kfree(ppgtt);
+		return ERR_PTR(ret);
+	}
+
+	ppgtt->file_priv = fpriv;
+
+	trace_i915_ppgtt_create(&ppgtt->base);
+
+	return ppgtt;
+}
+
+void  i915_ppgtt_release(struct kref *kref)
+{
+	struct i915_hw_ppgtt *ppgtt =
+		container_of(kref, struct i915_hw_ppgtt, ref);
+
+	trace_i915_ppgtt_release(&ppgtt->base);
+
+	/* vmas should already be unbound */
+	WARN_ON(!list_empty(&ppgtt->base.active_list));
+	WARN_ON(!list_empty(&ppgtt->base.inactive_list));
+
+	list_del(&ppgtt->base.global_link);
+	drm_mm_takedown(&ppgtt->base.mm);
+
+	ppgtt->base.cleanup(&ppgtt->base);
+	kfree(ppgtt);
+}
 
 static void
 ppgtt_bind_vma(struct i915_vma *vma,
@@ -1295,7 +1272,7 @@ void i915_check_and_clear_faults(struct drm_device *dev)
 		fault_reg = I915_READ(RING_FAULT_REG(ring));
 		if (fault_reg & RING_FAULT_VALID) {
 			DRM_DEBUG_DRIVER("Unexpected fault\n"
-					 "\tAddr: 0x%08lx\\n"
+					 "\tAddr: 0x%08lx\n"
 					 "\tAddress space: %s\n"
 					 "\tSource ID: %d\n"
 					 "\tType: %d\n",
@@ -1365,7 +1342,7 @@ void i915_gem_restore_gtt_mappings(struct drm_device *dev)
 		 * Unfortunately above, we've just wiped out the mappings
 		 * without telling our object about it. So we need to fake it.
 		 */
-		obj->has_global_gtt_mapping = 0;
+		vma->bound &= ~GLOBAL_BIND;
 		vma->bind_vma(vma, obj->cache_level, GLOBAL_BIND);
 	}
 
@@ -1562,7 +1539,7 @@ static void i915_ggtt_bind_vma(struct i915_vma *vma,
 
 	BUG_ON(!i915_is_ggtt(vma->vm));
 	intel_gtt_insert_sg_entries(vma->obj->pages, entry, flags);
-	vma->obj->has_global_gtt_mapping = 1;
+	vma->bound = GLOBAL_BIND;
 }
 
 static void i915_ggtt_clear_range(struct i915_address_space *vm,
@@ -1581,7 +1558,7 @@ static void i915_ggtt_unbind_vma(struct i915_vma *vma)
 	const unsigned int size = vma->obj->base.size >> PAGE_SHIFT;
 
 	BUG_ON(!i915_is_ggtt(vma->vm));
-	vma->obj->has_global_gtt_mapping = 0;
+	vma->bound = 0;
 	intel_gtt_clear_range(first, size);
 }
 
@@ -1609,24 +1586,24 @@ static void ggtt_bind_vma(struct i915_vma *vma,
 	 * flags. At all other times, the GPU will use the aliasing PPGTT.
 	 */
 	if (!dev_priv->mm.aliasing_ppgtt || flags & GLOBAL_BIND) {
-		if (!obj->has_global_gtt_mapping ||
+		if (!(vma->bound & GLOBAL_BIND) ||
 		    (cache_level != obj->cache_level)) {
 			vma->vm->insert_entries(vma->vm, obj->pages,
 						vma->node.start,
 						cache_level, flags);
-			obj->has_global_gtt_mapping = 1;
+			vma->bound |= GLOBAL_BIND;
 		}
 	}
 
 	if (dev_priv->mm.aliasing_ppgtt &&
-	    (!obj->has_aliasing_ppgtt_mapping ||
+	    (!(vma->bound & LOCAL_BIND) ||
 	     (cache_level != obj->cache_level))) {
 		struct i915_hw_ppgtt *appgtt = dev_priv->mm.aliasing_ppgtt;
 		appgtt->base.insert_entries(&appgtt->base,
 					    vma->obj->pages,
 					    vma->node.start,
 					    cache_level, flags);
-		vma->obj->has_aliasing_ppgtt_mapping = 1;
+		vma->bound |= LOCAL_BIND;
 	}
 }
 
@@ -1636,21 +1613,21 @@ static void ggtt_unbind_vma(struct i915_vma *vma)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_gem_object *obj = vma->obj;
 
-	if (obj->has_global_gtt_mapping) {
+	if (vma->bound & GLOBAL_BIND) {
 		vma->vm->clear_range(vma->vm,
 				     vma->node.start,
 				     obj->base.size,
 				     true);
-		obj->has_global_gtt_mapping = 0;
+		vma->bound &= ~GLOBAL_BIND;
 	}
 
-	if (obj->has_aliasing_ppgtt_mapping) {
+	if (vma->bound & LOCAL_BIND) {
 		struct i915_hw_ppgtt *appgtt = dev_priv->mm.aliasing_ppgtt;
 		appgtt->base.clear_range(&appgtt->base,
 					 vma->node.start,
 					 obj->base.size,
 					 true);
-		obj->has_aliasing_ppgtt_mapping = 0;
+		vma->bound &= ~LOCAL_BIND;
 	}
 }
 
@@ -1687,10 +1664,10 @@ static void i915_gtt_color_adjust(struct drm_mm_node *node,
 	}
 }
 
-void i915_gem_setup_global_gtt(struct drm_device *dev,
-			       unsigned long start,
-			       unsigned long mappable_end,
-			       unsigned long end)
+static int i915_gem_setup_global_gtt(struct drm_device *dev,
+				     unsigned long start,
+				     unsigned long mappable_end,
+				     unsigned long end)
 {
 	/* Let GEM Manage all of the aperture.
 	 *
@@ -1706,6 +1683,7 @@ void i915_gem_setup_global_gtt(struct drm_device *dev,
 	struct drm_mm_node *entry;
 	struct drm_i915_gem_object *obj;
 	unsigned long hole_start, hole_end;
+	int ret;
 
 	BUG_ON(mappable_end > end);
 
@@ -1717,15 +1695,17 @@ void i915_gem_setup_global_gtt(struct drm_device *dev,
 	/* Mark any preallocated objects as occupied */
 	list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) {
 		struct i915_vma *vma = i915_gem_obj_to_vma(obj, ggtt_vm);
-		int ret;
+
 		DRM_DEBUG_KMS("reserving preallocated space: %lx + %zx\n",
 			      i915_gem_obj_ggtt_offset(obj), obj->base.size);
 
 		WARN_ON(i915_gem_obj_ggtt_bound(obj));
 		ret = drm_mm_reserve_node(&ggtt_vm->mm, &vma->node);
-		if (ret)
-			DRM_DEBUG_KMS("Reservation failed\n");
-		obj->has_global_gtt_mapping = 1;
+		if (ret) {
+			DRM_DEBUG_KMS("Reservation failed: %i\n", ret);
+			return ret;
+		}
+		vma->bound |= GLOBAL_BIND;
 	}
 
 	dev_priv->gtt.base.start = start;
@@ -1741,6 +1721,22 @@ void i915_gem_setup_global_gtt(struct drm_device *dev,
 
 	/* And finally clear the reserved guard page */
 	ggtt_vm->clear_range(ggtt_vm, end - PAGE_SIZE, PAGE_SIZE, true);
+
+	if (USES_PPGTT(dev) && !USES_FULL_PPGTT(dev)) {
+		struct i915_hw_ppgtt *ppgtt;
+
+		ppgtt = kzalloc(sizeof(*ppgtt), GFP_KERNEL);
+		if (!ppgtt)
+			return -ENOMEM;
+
+		ret = __hw_ppgtt_init(dev, ppgtt);
+		if (ret != 0)
+			return ret;
+
+		dev_priv->mm.aliasing_ppgtt = ppgtt;
+	}
+
+	return 0;
 }
 
 void i915_gem_init_global_gtt(struct drm_device *dev)
@@ -1754,6 +1750,25 @@ void i915_gem_init_global_gtt(struct drm_device *dev)
 	i915_gem_setup_global_gtt(dev, 0, mappable_size, gtt_size);
 }
 
+void i915_global_gtt_cleanup(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct i915_address_space *vm = &dev_priv->gtt.base;
+
+	if (dev_priv->mm.aliasing_ppgtt) {
+		struct i915_hw_ppgtt *ppgtt = dev_priv->mm.aliasing_ppgtt;
+
+		ppgtt->base.cleanup(&ppgtt->base);
+	}
+
+	if (drm_mm_initialized(&vm->mm)) {
+		drm_mm_takedown(&vm->mm);
+		list_del(&vm->global_link);
+	}
+
+	vm->cleanup(vm);
+}
+
 static int setup_scratch_page(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1763,7 +1778,6 @@ static int setup_scratch_page(struct drm_device *dev)
 	page = alloc_page(GFP_KERNEL | GFP_DMA32 | __GFP_ZERO);
 	if (page == NULL)
 		return -ENOMEM;
-	get_page(page);
 	set_pages_uc(page, 1);
 
 #ifdef CONFIG_INTEL_IOMMU
@@ -1788,7 +1802,6 @@ static void teardown_scratch_page(struct drm_device *dev)
 	set_pages_wb(page, 1);
 	pci_unmap_page(dev->pdev, dev_priv->gtt.base.scratch.addr,
 		       PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
-	put_page(page);
 	__free_page(page);
 }
 
@@ -1858,6 +1871,18 @@ static size_t chv_get_stolen_size(u16 gmch_ctrl)
 		return (gmch_ctrl - 0x17 + 9) << 22;
 }
 
+static size_t gen9_get_stolen_size(u16 gen9_gmch_ctl)
+{
+	gen9_gmch_ctl >>= BDW_GMCH_GMS_SHIFT;
+	gen9_gmch_ctl &= BDW_GMCH_GMS_MASK;
+
+	if (gen9_gmch_ctl < 0xf0)
+		return gen9_gmch_ctl << 25; /* 32 MB units */
+	else
+		/* 4MB increments starting at 0xf0 for 4MB */
+		return (gen9_gmch_ctl - 0xf0 + 1) << 22;
+}
+
 static int ggtt_probe_common(struct drm_device *dev,
 			     size_t gtt_size)
 {
@@ -1901,6 +1926,22 @@ static void bdw_setup_private_ppat(struct drm_i915_private *dev_priv)
 	      GEN8_PPAT(6, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(2)) |
 	      GEN8_PPAT(7, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3));
 
+	if (!USES_PPGTT(dev_priv->dev))
+		/* Spec: "For GGTT, there is NO pat_sel[2:0] from the entry,
+		 * so RTL will always use the value corresponding to
+		 * pat_sel = 000".
+		 * So let's disable cache for GGTT to avoid screen corruptions.
+		 * MOCS still can be used though.
+		 * - System agent ggtt writes (i.e. cpu gtt mmaps) already work
+		 * before this patch, i.e. the same uncached + snooping access
+		 * like on gen6/7 seems to be in effect.
+		 * - So this just fixes blitter/render access. Again it looks
+		 * like it's not just uncached access, but uncached + snooping.
+		 * So we can still hold onto all our assumptions wrt cpu
+		 * clflushing on LLC machines.
+		 */
+		pat = GEN8_PPAT(0, GEN8_PPAT_UC);
+
 	/* XXX: spec defines this as 2 distinct registers. It's unclear if a 64b
 	 * write would work. */
 	I915_WRITE(GEN8_PRIVATE_PAT, pat);
@@ -1917,9 +1958,17 @@ static void chv_setup_private_ppat(struct drm_i915_private *dev_priv)
 	 * Only the snoop bit has meaning for CHV, the rest is
 	 * ignored.
 	 *
-	 * Note that the harware enforces snooping for all page
-	 * table accesses. The snoop bit is actually ignored for
-	 * PDEs.
+	 * The hardware will never snoop for certain types of accesses:
+	 * - CPU GTT (GMADR->GGTT->no snoop->memory)
+	 * - PPGTT page tables
+	 * - some other special cycles
+	 *
+	 * As with BDW, we also need to consider the following for GT accesses:
+	 * "For GGTT, there is NO pat_sel[2:0] from the entry,
+	 * so RTL will always use the value corresponding to
+	 * pat_sel = 000".
+	 * Which means we must set the snoop bit in PAT entry 0
+	 * in order to keep the global status page working.
 	 */
 	pat = GEN8_PPAT(0, CHV_PPAT_SNOOP) |
 	      GEN8_PPAT(1, 0) |
@@ -1954,7 +2003,10 @@ static int gen8_gmch_probe(struct drm_device *dev,
 
 	pci_read_config_word(dev->pdev, SNB_GMCH_CTRL, &snb_gmch_ctl);
 
-	if (IS_CHERRYVIEW(dev)) {
+	if (INTEL_INFO(dev)->gen >= 9) {
+		*stolen = gen9_get_stolen_size(snb_gmch_ctl);
+		gtt_size = gen8_get_total_gtt_size(snb_gmch_ctl);
+	} else if (IS_CHERRYVIEW(dev)) {
 		*stolen = chv_get_stolen_size(snb_gmch_ctl);
 		gtt_size = chv_get_total_gtt_size(snb_gmch_ctl);
 	} else {
@@ -2022,10 +2074,6 @@ static void gen6_gmch_remove(struct i915_address_space *vm)
 
 	struct i915_gtt *gtt = container_of(vm, struct i915_gtt, base);
 
-	if (drm_mm_initialized(&vm->mm)) {
-		drm_mm_takedown(&vm->mm);
-		list_del(&vm->global_link);
-	}
 	iounmap(gtt->gsm);
 	teardown_scratch_page(vm->dev);
 }
@@ -2058,10 +2106,6 @@ static int i915_gmch_probe(struct drm_device *dev,
 
 static void i915_gmch_remove(struct i915_address_space *vm)
 {
-	if (drm_mm_initialized(&vm->mm)) {
-		drm_mm_takedown(&vm->mm);
-		list_del(&vm->global_link);
-	}
 	intel_gmch_remove();
 }
 
@@ -2134,6 +2178,7 @@ static struct i915_vma *__i915_gem_vma_create(struct drm_i915_gem_object *obj,
 	vma->obj = obj;
 
 	switch (INTEL_INFO(vm->dev)->gen) {
+	case 9:
 	case 8:
 	case 7:
 	case 6:
@@ -2160,8 +2205,10 @@ static struct i915_vma *__i915_gem_vma_create(struct drm_i915_gem_object *obj,
 	/* Keep GGTT vmas first to make debug easier */
 	if (i915_is_ggtt(vm))
 		list_add(&vma->vma_link, &obj->vma_list);
-	else
+	else {
 		list_add_tail(&vma->vma_link, &obj->vma_list);
+		i915_ppgtt_get(i915_vm_to_ppgtt(vm));
+	}
 
 	return vma;
 }
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h
index 8d6f7c18c404..beaf4bcfdac8 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.h
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.h
@@ -34,6 +34,8 @@
 #ifndef __I915_GEM_GTT_H__
 #define __I915_GEM_GTT_H__
 
+struct drm_i915_file_private;
+
 typedef uint32_t gen6_gtt_pte_t;
 typedef uint64_t gen8_gtt_pte_t;
 typedef gen8_gtt_pte_t gen8_ppgtt_pde_t;
@@ -121,6 +123,12 @@ struct i915_vma {
 	struct drm_i915_gem_object *obj;
 	struct i915_address_space *vm;
 
+	/** Flags and address space this VMA is bound to */
+#define GLOBAL_BIND	(1<<0)
+#define LOCAL_BIND	(1<<1)
+#define PTE_READ_ONLY	(1<<2)
+	unsigned int bound : 4;
+
 	/** This object's place on the active/inactive lists */
 	struct list_head mm_list;
 
@@ -153,8 +161,6 @@ struct i915_vma {
 	 * setting the valid PTE entries to a reserved scratch page. */
 	void (*unbind_vma)(struct i915_vma *vma);
 	/* Map an object into an address space with the given cache flags. */
-#define GLOBAL_BIND (1<<0)
-#define PTE_READ_ONLY (1<<1)
 	void (*bind_vma)(struct i915_vma *vma,
 			 enum i915_cache_level cache_level,
 			 u32 flags);
@@ -258,22 +264,34 @@ struct i915_hw_ppgtt {
 		dma_addr_t *gen8_pt_dma_addr[4];
 	};
 
-	struct intel_context *ctx;
+	struct drm_i915_file_private *file_priv;
 
 	int (*enable)(struct i915_hw_ppgtt *ppgtt);
 	int (*switch_mm)(struct i915_hw_ppgtt *ppgtt,
-			 struct intel_engine_cs *ring,
-			 bool synchronous);
+			 struct intel_engine_cs *ring);
 	void (*debug_dump)(struct i915_hw_ppgtt *ppgtt, struct seq_file *m);
 };
 
 int i915_gem_gtt_init(struct drm_device *dev);
 void i915_gem_init_global_gtt(struct drm_device *dev);
-void i915_gem_setup_global_gtt(struct drm_device *dev, unsigned long start,
-			       unsigned long mappable_end, unsigned long end);
-
-bool intel_enable_ppgtt(struct drm_device *dev, bool full);
-int i915_gem_init_ppgtt(struct drm_device *dev, struct i915_hw_ppgtt *ppgtt);
+void i915_global_gtt_cleanup(struct drm_device *dev);
+
+
+int i915_ppgtt_init(struct drm_device *dev, struct i915_hw_ppgtt *ppgtt);
+int i915_ppgtt_init_hw(struct drm_device *dev);
+void i915_ppgtt_release(struct kref *kref);
+struct i915_hw_ppgtt *i915_ppgtt_create(struct drm_device *dev,
+					struct drm_i915_file_private *fpriv);
+static inline void i915_ppgtt_get(struct i915_hw_ppgtt *ppgtt)
+{
+	if (ppgtt)
+		kref_get(&ppgtt->ref);
+}
+static inline void i915_ppgtt_put(struct i915_hw_ppgtt *ppgtt)
+{
+	if (ppgtt)
+		kref_put(&ppgtt->ref, i915_ppgtt_release);
+}
 
 void i915_check_and_clear_faults(struct drm_device *dev);
 void i915_gem_suspend_gtt_mappings(struct drm_device *dev);
diff --git a/drivers/gpu/drm/i915/i915_gem_render_state.c b/drivers/gpu/drm/i915/i915_gem_render_state.c
index e60be3f552a6..98dcd94acba8 100644
--- a/drivers/gpu/drm/i915/i915_gem_render_state.c
+++ b/drivers/gpu/drm/i915/i915_gem_render_state.c
@@ -28,13 +28,6 @@
 #include "i915_drv.h"
 #include "intel_renderstate.h"
 
-struct render_state {
-	const struct intel_renderstate_rodata *rodata;
-	struct drm_i915_gem_object *obj;
-	u64 ggtt_offset;
-	int gen;
-};
-
 static const struct intel_renderstate_rodata *
 render_state_get_rodata(struct drm_device *dev, const int gen)
 {
@@ -45,6 +38,8 @@ render_state_get_rodata(struct drm_device *dev, const int gen)
 		return &gen7_null_state;
 	case 8:
 		return &gen8_null_state;
+	case 9:
+		return &gen9_null_state;
 	}
 
 	return NULL;
@@ -127,30 +122,47 @@ static int render_state_setup(struct render_state *so)
 	return 0;
 }
 
-static void render_state_fini(struct render_state *so)
+void i915_gem_render_state_fini(struct render_state *so)
 {
 	i915_gem_object_ggtt_unpin(so->obj);
 	drm_gem_object_unreference(&so->obj->base);
 }
 
-int i915_gem_render_state_init(struct intel_engine_cs *ring)
+int i915_gem_render_state_prepare(struct intel_engine_cs *ring,
+				  struct render_state *so)
 {
-	struct render_state so;
 	int ret;
 
 	if (WARN_ON(ring->id != RCS))
 		return -ENOENT;
 
-	ret = render_state_init(&so, ring->dev);
+	ret = render_state_init(so, ring->dev);
 	if (ret)
 		return ret;
 
-	if (so.rodata == NULL)
+	if (so->rodata == NULL)
 		return 0;
 
-	ret = render_state_setup(&so);
+	ret = render_state_setup(so);
+	if (ret) {
+		i915_gem_render_state_fini(so);
+		return ret;
+	}
+
+	return 0;
+}
+
+int i915_gem_render_state_init(struct intel_engine_cs *ring)
+{
+	struct render_state so;
+	int ret;
+
+	ret = i915_gem_render_state_prepare(ring, &so);
 	if (ret)
-		goto out;
+		return ret;
+
+	if (so.rodata == NULL)
+		return 0;
 
 	ret = ring->dispatch_execbuffer(ring,
 					so.ggtt_offset,
@@ -164,6 +176,6 @@ int i915_gem_render_state_init(struct intel_engine_cs *ring)
 	ret = __i915_add_request(ring, NULL, so.obj, NULL);
 	/* __i915_add_request moves object to inactive if it fails */
 out:
-	render_state_fini(&so);
+	i915_gem_render_state_fini(&so);
 	return ret;
 }
diff --git a/drivers/gpu/drm/i915/i915_gem_render_state.h b/drivers/gpu/drm/i915/i915_gem_render_state.h
new file mode 100644
index 000000000000..c44961ed3fad
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_gem_render_state.h
@@ -0,0 +1,47 @@
+/*
+ * Copyright © 2014 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#ifndef _I915_GEM_RENDER_STATE_H_
+#define _I915_GEM_RENDER_STATE_H_
+
+#include <linux/types.h>
+
+struct intel_renderstate_rodata {
+	const u32 *reloc;
+	const u32 *batch;
+	const u32 batch_items;
+};
+
+struct render_state {
+	const struct intel_renderstate_rodata *rodata;
+	struct drm_i915_gem_object *obj;
+	u64 ggtt_offset;
+	int gen;
+};
+
+int i915_gem_render_state_init(struct intel_engine_cs *ring);
+void i915_gem_render_state_fini(struct render_state *so);
+int i915_gem_render_state_prepare(struct intel_engine_cs *ring,
+				  struct render_state *so);
+
+#endif /* _I915_GEM_RENDER_STATE_H_ */
diff --git a/drivers/gpu/drm/i915/i915_gem_stolen.c b/drivers/gpu/drm/i915/i915_gem_stolen.c
index 21c025a209c0..a2045848bd1a 100644
--- a/drivers/gpu/drm/i915/i915_gem_stolen.c
+++ b/drivers/gpu/drm/i915/i915_gem_stolen.c
@@ -137,7 +137,11 @@ static unsigned long i915_stolen_to_physical(struct drm_device *dev)
 		r = devm_request_mem_region(dev->dev, base + 1,
 					    dev_priv->gtt.stolen_size - 1,
 					    "Graphics Stolen Memory");
-		if (r == NULL) {
+		/*
+		 * GEN3 firmware likes to smash pci bridges into the stolen
+		 * range. Apparently this works.
+		 */
+		if (r == NULL && !IS_GEN3(dev)) {
 			DRM_ERROR("conflict detected with stolen region: [0x%08x - 0x%08x]\n",
 				  base, base + (uint32_t)dev_priv->gtt.stolen_size);
 			base = 0;
@@ -289,6 +293,7 @@ void i915_gem_cleanup_stolen(struct drm_device *dev)
 int i915_gem_init_stolen(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 tmp;
 	int bios_reserved = 0;
 
 #ifdef CONFIG_INTEL_IOMMU
@@ -308,8 +313,16 @@ int i915_gem_init_stolen(struct drm_device *dev)
 	DRM_DEBUG_KMS("found %zd bytes of stolen memory at %08lx\n",
 		      dev_priv->gtt.stolen_size, dev_priv->mm.stolen_base);
 
-	if (IS_VALLEYVIEW(dev))
-		bios_reserved = 1024*1024; /* top 1M on VLV/BYT */
+	if (INTEL_INFO(dev)->gen >= 8) {
+		tmp = I915_READ(GEN7_BIOS_RESERVED);
+		tmp >>= GEN8_BIOS_RESERVED_SHIFT;
+		tmp &= GEN8_BIOS_RESERVED_MASK;
+		bios_reserved = (1024*1024) << tmp;
+	} else if (IS_GEN7(dev)) {
+		tmp = I915_READ(GEN7_BIOS_RESERVED);
+		bios_reserved = tmp & GEN7_BIOS_RESERVED_256K ?
+			256*1024 : 1024*1024;
+	}
 
 	if (WARN_ON(bios_reserved > dev_priv->gtt.stolen_size))
 		return 0;
@@ -524,7 +537,7 @@ i915_gem_object_create_stolen_for_preallocated(struct drm_device *dev,
 		}
 	}
 
-	obj->has_global_gtt_mapping = 1;
+	vma->bound |= GLOBAL_BIND;
 
 	list_add_tail(&obj->global_list, &dev_priv->mm.bound_list);
 	list_add_tail(&vma->mm_list, &ggtt->inactive_list);
diff --git a/drivers/gpu/drm/i915/i915_gem_tiling.c b/drivers/gpu/drm/i915/i915_gem_tiling.c
index cb150e8b4336..4727a4e2c87c 100644
--- a/drivers/gpu/drm/i915/i915_gem_tiling.c
+++ b/drivers/gpu/drm/i915/i915_gem_tiling.c
@@ -91,26 +91,44 @@ i915_gem_detect_bit_6_swizzle(struct drm_device *dev)
 	uint32_t swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
 	uint32_t swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
 
-	if (IS_VALLEYVIEW(dev)) {
+	if (INTEL_INFO(dev)->gen >= 8 || IS_VALLEYVIEW(dev)) {
+		/*
+		 * On BDW+, swizzling is not used. We leave the CPU memory
+		 * controller in charge of optimizing memory accesses without
+		 * the extra address manipulation GPU side.
+		 *
+		 * VLV and CHV don't have GPU swizzling.
+		 */
 		swizzle_x = I915_BIT_6_SWIZZLE_NONE;
 		swizzle_y = I915_BIT_6_SWIZZLE_NONE;
 	} else if (INTEL_INFO(dev)->gen >= 6) {
-		uint32_t dimm_c0, dimm_c1;
-		dimm_c0 = I915_READ(MAD_DIMM_C0);
-		dimm_c1 = I915_READ(MAD_DIMM_C1);
-		dimm_c0 &= MAD_DIMM_A_SIZE_MASK | MAD_DIMM_B_SIZE_MASK;
-		dimm_c1 &= MAD_DIMM_A_SIZE_MASK | MAD_DIMM_B_SIZE_MASK;
-		/* Enable swizzling when the channels are populated with
-		 * identically sized dimms. We don't need to check the 3rd
-		 * channel because no cpu with gpu attached ships in that
-		 * configuration. Also, swizzling only makes sense for 2
-		 * channels anyway. */
-		if (dimm_c0 == dimm_c1) {
-			swizzle_x = I915_BIT_6_SWIZZLE_9_10;
-			swizzle_y = I915_BIT_6_SWIZZLE_9;
+		if (dev_priv->preserve_bios_swizzle) {
+			if (I915_READ(DISP_ARB_CTL) &
+			    DISP_TILE_SURFACE_SWIZZLING) {
+				swizzle_x = I915_BIT_6_SWIZZLE_9_10;
+				swizzle_y = I915_BIT_6_SWIZZLE_9;
+			} else {
+				swizzle_x = I915_BIT_6_SWIZZLE_NONE;
+				swizzle_y = I915_BIT_6_SWIZZLE_NONE;
+			}
 		} else {
-			swizzle_x = I915_BIT_6_SWIZZLE_NONE;
-			swizzle_y = I915_BIT_6_SWIZZLE_NONE;
+			uint32_t dimm_c0, dimm_c1;
+			dimm_c0 = I915_READ(MAD_DIMM_C0);
+			dimm_c1 = I915_READ(MAD_DIMM_C1);
+			dimm_c0 &= MAD_DIMM_A_SIZE_MASK | MAD_DIMM_B_SIZE_MASK;
+			dimm_c1 &= MAD_DIMM_A_SIZE_MASK | MAD_DIMM_B_SIZE_MASK;
+			/* Enable swizzling when the channels are populated
+			 * with identically sized dimms. We don't need to check
+			 * the 3rd channel because no cpu with gpu attached
+			 * ships in that configuration. Also, swizzling only
+			 * makes sense for 2 channels anyway. */
+			if (dimm_c0 == dimm_c1) {
+				swizzle_x = I915_BIT_6_SWIZZLE_9_10;
+				swizzle_y = I915_BIT_6_SWIZZLE_9;
+			} else {
+				swizzle_x = I915_BIT_6_SWIZZLE_NONE;
+				swizzle_y = I915_BIT_6_SWIZZLE_NONE;
+			}
 		}
 	} else if (IS_GEN5(dev)) {
 		/* On Ironlake whatever DRAM config, GPU always do
@@ -160,6 +178,15 @@ i915_gem_detect_bit_6_swizzle(struct drm_device *dev)
 			}
 			break;
 		}
+
+		/* check for L-shaped memory aka modified enhanced addressing */
+		if (IS_GEN4(dev)) {
+			uint32_t ddc2 = I915_READ(DCC2);
+
+			if (!(ddc2 & DCC2_MODIFIED_ENHANCED_DISABLE))
+				dev_priv->quirks |= QUIRK_PIN_SWIZZLED_PAGES;
+		}
+
 		if (dcc == 0xffffffff) {
 			DRM_ERROR("Couldn't read from MCHBAR.  "
 				  "Disabling tiling.\n");
@@ -357,26 +384,22 @@ i915_gem_set_tiling(struct drm_device *dev, void *data,
 		 * has to also include the unfenced register the GPU uses
 		 * whilst executing a fenced command for an untiled object.
 		 */
-
-		obj->map_and_fenceable =
-			!i915_gem_obj_ggtt_bound(obj) ||
-			(i915_gem_obj_ggtt_offset(obj) +
-			 obj->base.size <= dev_priv->gtt.mappable_end &&
-			 i915_gem_object_fence_ok(obj, args->tiling_mode));
-
-		/* Rebind if we need a change of alignment */
-		if (!obj->map_and_fenceable) {
-			u32 unfenced_align =
-				i915_gem_get_gtt_alignment(dev, obj->base.size,
-							    args->tiling_mode,
-							    false);
-			if (i915_gem_obj_ggtt_offset(obj) & (unfenced_align - 1))
-				ret = i915_gem_object_ggtt_unbind(obj);
-		}
+		if (obj->map_and_fenceable &&
+		    !i915_gem_object_fence_ok(obj, args->tiling_mode))
+			ret = i915_gem_object_ggtt_unbind(obj);
 
 		if (ret == 0) {
+			if (obj->pages &&
+			    obj->madv == I915_MADV_WILLNEED &&
+			    dev_priv->quirks & QUIRK_PIN_SWIZZLED_PAGES) {
+				if (args->tiling_mode == I915_TILING_NONE)
+					i915_gem_object_unpin_pages(obj);
+				if (obj->tiling_mode == I915_TILING_NONE)
+					i915_gem_object_pin_pages(obj);
+			}
+
 			obj->fence_dirty =
-				obj->fenced_gpu_access ||
+				obj->last_fenced_seqno ||
 				obj->fence_reg != I915_FENCE_REG_NONE;
 
 			obj->tiling_mode = args->tiling_mode;
@@ -440,6 +463,7 @@ i915_gem_get_tiling(struct drm_device *dev, void *data,
 	}
 
 	/* Hide bit 17 from the user -- see comment in i915_gem_set_tiling */
+	args->phys_swizzle_mode = args->swizzle_mode;
 	if (args->swizzle_mode == I915_BIT_6_SWIZZLE_9_17)
 		args->swizzle_mode = I915_BIT_6_SWIZZLE_9;
 	if (args->swizzle_mode == I915_BIT_6_SWIZZLE_9_10_17)
diff --git a/drivers/gpu/drm/i915/i915_gem_userptr.c b/drivers/gpu/drm/i915/i915_gem_userptr.c
index d38413997379..d182058383a9 100644
--- a/drivers/gpu/drm/i915/i915_gem_userptr.c
+++ b/drivers/gpu/drm/i915/i915_gem_userptr.c
@@ -293,15 +293,23 @@ i915_gem_userptr_release__mmu_notifier(struct drm_i915_gem_object *obj)
 static struct i915_mmu_notifier *
 i915_mmu_notifier_find(struct i915_mm_struct *mm)
 {
-	if (mm->mn == NULL) {
-		down_write(&mm->mm->mmap_sem);
-		mutex_lock(&to_i915(mm->dev)->mm_lock);
-		if (mm->mn == NULL)
-			mm->mn = i915_mmu_notifier_create(mm->mm);
-		mutex_unlock(&to_i915(mm->dev)->mm_lock);
-		up_write(&mm->mm->mmap_sem);
+	struct i915_mmu_notifier *mn = mm->mn;
+
+	mn = mm->mn;
+	if (mn)
+		return mn;
+
+	down_write(&mm->mm->mmap_sem);
+	mutex_lock(&to_i915(mm->dev)->mm_lock);
+	if ((mn = mm->mn) == NULL) {
+		mn = i915_mmu_notifier_create(mm->mm);
+		if (!IS_ERR(mn))
+			mm->mn = mn;
 	}
-	return mm->mn;
+	mutex_unlock(&to_i915(mm->dev)->mm_lock);
+	up_write(&mm->mm->mmap_sem);
+
+	return mn;
 }
 
 static int
@@ -681,16 +689,15 @@ i915_gem_userptr_get_pages(struct drm_i915_gem_object *obj)
 static void
 i915_gem_userptr_put_pages(struct drm_i915_gem_object *obj)
 {
-	struct scatterlist *sg;
-	int i;
+	struct sg_page_iter sg_iter;
 
 	BUG_ON(obj->userptr.work != NULL);
 
 	if (obj->madv != I915_MADV_WILLNEED)
 		obj->dirty = 0;
 
-	for_each_sg(obj->pages->sgl, sg, obj->pages->nents, i) {
-		struct page *page = sg_page(sg);
+	for_each_sg_page(obj->pages->sgl, &sg_iter, obj->pages->nents, 0) {
+		struct page *page = sg_page_iter_page(&sg_iter);
 
 		if (obj->dirty)
 			set_page_dirty(page);
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index eab41f9390f8..cdaee6ce05f8 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -192,10 +192,10 @@ static void print_error_buffers(struct drm_i915_error_state_buf *m,
 				struct drm_i915_error_buffer *err,
 				int count)
 {
-	err_printf(m, "%s [%d]:\n", name, count);
+	err_printf(m, "  %s [%d]:\n", name, count);
 
 	while (count--) {
-		err_printf(m, "  %08x %8u %02x %02x %x %x",
+		err_printf(m, "    %08x %8u %02x %02x %x %x",
 			   err->gtt_offset,
 			   err->size,
 			   err->read_domains,
@@ -208,7 +208,7 @@ static void print_error_buffers(struct drm_i915_error_state_buf *m,
 		err_puts(m, err->userptr ? " userptr" : "");
 		err_puts(m, err->ring != -1 ? " " : "");
 		err_puts(m, ring_str(err->ring));
-		err_puts(m, i915_cache_level_str(err->cache_level));
+		err_puts(m, i915_cache_level_str(m->i915, err->cache_level));
 
 		if (err->name)
 			err_printf(m, " (name: %d)", err->name);
@@ -242,11 +242,15 @@ static const char *hangcheck_action_to_str(enum intel_ring_hangcheck_action a)
 
 static void i915_ring_error_state(struct drm_i915_error_state_buf *m,
 				  struct drm_device *dev,
-				  struct drm_i915_error_ring *ring)
+				  struct drm_i915_error_state *error,
+				  int ring_idx)
 {
+	struct drm_i915_error_ring *ring = &error->ring[ring_idx];
+
 	if (!ring->valid)
 		return;
 
+	err_printf(m, "%s command stream:\n", ring_str(ring_idx));
 	err_printf(m, "  HEAD: 0x%08x\n", ring->head);
 	err_printf(m, "  TAIL: 0x%08x\n", ring->tail);
 	err_printf(m, "  CTL: 0x%08x\n", ring->ctl);
@@ -388,20 +392,20 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 	if (INTEL_INFO(dev)->gen == 7)
 		err_printf(m, "ERR_INT: 0x%08x\n", error->err_int);
 
-	for (i = 0; i < ARRAY_SIZE(error->ring); i++) {
-		err_printf(m, "%s command stream:\n", ring_str(i));
-		i915_ring_error_state(m, dev, &error->ring[i]);
-	}
+	for (i = 0; i < ARRAY_SIZE(error->ring); i++)
+		i915_ring_error_state(m, dev, error, i);
+
+	for (i = 0; i < error->vm_count; i++) {
+		err_printf(m, "vm[%d]\n", i);
 
-	if (error->active_bo)
 		print_error_buffers(m, "Active",
-				    error->active_bo[0],
-				    error->active_bo_count[0]);
+				    error->active_bo[i],
+				    error->active_bo_count[i]);
 
-	if (error->pinned_bo)
 		print_error_buffers(m, "Pinned",
-				    error->pinned_bo[0],
-				    error->pinned_bo_count[0]);
+				    error->pinned_bo[i],
+				    error->pinned_bo_count[i]);
+	}
 
 	for (i = 0; i < ARRAY_SIZE(error->ring); i++) {
 		obj = error->ring[i].batchbuffer;
@@ -492,9 +496,11 @@ out:
 }
 
 int i915_error_state_buf_init(struct drm_i915_error_state_buf *ebuf,
+			      struct drm_i915_private *i915,
 			      size_t count, loff_t pos)
 {
 	memset(ebuf, 0, sizeof(*ebuf));
+	ebuf->i915 = i915;
 
 	/* We need to have enough room to store any i915_error_state printf
 	 * so that we can move it to start position.
@@ -556,24 +562,56 @@ static void i915_error_state_free(struct kref *error_ref)
 }
 
 static struct drm_i915_error_object *
-i915_error_object_create_sized(struct drm_i915_private *dev_priv,
-			       struct drm_i915_gem_object *src,
-			       struct i915_address_space *vm,
-			       const int num_pages)
+i915_error_object_create(struct drm_i915_private *dev_priv,
+			 struct drm_i915_gem_object *src,
+			 struct i915_address_space *vm)
 {
 	struct drm_i915_error_object *dst;
-	int i;
+	struct i915_vma *vma = NULL;
+	int num_pages;
+	bool use_ggtt;
+	int i = 0;
 	u32 reloc_offset;
 
 	if (src == NULL || src->pages == NULL)
 		return NULL;
 
+	num_pages = src->base.size >> PAGE_SHIFT;
+
 	dst = kmalloc(sizeof(*dst) + num_pages * sizeof(u32 *), GFP_ATOMIC);
 	if (dst == NULL)
 		return NULL;
 
-	reloc_offset = dst->gtt_offset = i915_gem_obj_offset(src, vm);
-	for (i = 0; i < num_pages; i++) {
+	if (i915_gem_obj_bound(src, vm))
+		dst->gtt_offset = i915_gem_obj_offset(src, vm);
+	else
+		dst->gtt_offset = -1;
+
+	reloc_offset = dst->gtt_offset;
+	if (i915_is_ggtt(vm))
+		vma = i915_gem_obj_to_ggtt(src);
+	use_ggtt = (src->cache_level == I915_CACHE_NONE &&
+		   vma && (vma->bound & GLOBAL_BIND) &&
+		   reloc_offset + num_pages * PAGE_SIZE <= dev_priv->gtt.mappable_end);
+
+	/* Cannot access stolen address directly, try to use the aperture */
+	if (src->stolen) {
+		use_ggtt = true;
+
+		if (!(vma && vma->bound & GLOBAL_BIND))
+			goto unwind;
+
+		reloc_offset = i915_gem_obj_ggtt_offset(src);
+		if (reloc_offset + num_pages * PAGE_SIZE > dev_priv->gtt.mappable_end)
+			goto unwind;
+	}
+
+	/* Cannot access snooped pages through the aperture */
+	if (use_ggtt && src->cache_level != I915_CACHE_NONE && !HAS_LLC(dev_priv->dev))
+		goto unwind;
+
+	dst->page_count = num_pages;
+	while (num_pages--) {
 		unsigned long flags;
 		void *d;
 
@@ -582,10 +620,7 @@ i915_error_object_create_sized(struct drm_i915_private *dev_priv,
 			goto unwind;
 
 		local_irq_save(flags);
-		if (src->cache_level == I915_CACHE_NONE &&
-		    reloc_offset < dev_priv->gtt.mappable_end &&
-		    src->has_global_gtt_mapping &&
-		    i915_is_ggtt(vm)) {
+		if (use_ggtt) {
 			void __iomem *s;
 
 			/* Simply ignore tiling or any overlapping fence.
@@ -597,14 +632,6 @@ i915_error_object_create_sized(struct drm_i915_private *dev_priv,
 						     reloc_offset);
 			memcpy_fromio(d, s, PAGE_SIZE);
 			io_mapping_unmap_atomic(s);
-		} else if (src->stolen) {
-			unsigned long offset;
-
-			offset = dev_priv->mm.stolen_base;
-			offset += src->stolen->start;
-			offset += i << PAGE_SHIFT;
-
-			memcpy_fromio(d, (void __iomem *) offset, PAGE_SIZE);
 		} else {
 			struct page *page;
 			void *s;
@@ -621,11 +648,9 @@ i915_error_object_create_sized(struct drm_i915_private *dev_priv,
 		}
 		local_irq_restore(flags);
 
-		dst->pages[i] = d;
-
+		dst->pages[i++] = d;
 		reloc_offset += PAGE_SIZE;
 	}
-	dst->page_count = num_pages;
 
 	return dst;
 
@@ -635,22 +660,19 @@ unwind:
 	kfree(dst);
 	return NULL;
 }
-#define i915_error_object_create(dev_priv, src, vm) \
-	i915_error_object_create_sized((dev_priv), (src), (vm), \
-				       (src)->base.size>>PAGE_SHIFT)
-
 #define i915_error_ggtt_object_create(dev_priv, src) \
-	i915_error_object_create_sized((dev_priv), (src), &(dev_priv)->gtt.base, \
-				       (src)->base.size>>PAGE_SHIFT)
+	i915_error_object_create((dev_priv), (src), &(dev_priv)->gtt.base)
 
 static void capture_bo(struct drm_i915_error_buffer *err,
-		       struct drm_i915_gem_object *obj)
+		       struct i915_vma *vma)
 {
+	struct drm_i915_gem_object *obj = vma->obj;
+
 	err->size = obj->base.size;
 	err->name = obj->base.name;
 	err->rseqno = obj->last_read_seqno;
 	err->wseqno = obj->last_write_seqno;
-	err->gtt_offset = i915_gem_obj_ggtt_offset(obj);
+	err->gtt_offset = vma->node.start;
 	err->read_domains = obj->base.read_domains;
 	err->write_domain = obj->base.write_domain;
 	err->fence_reg = obj->fence_reg;
@@ -674,7 +696,7 @@ static u32 capture_active_bo(struct drm_i915_error_buffer *err,
 	int i = 0;
 
 	list_for_each_entry(vma, head, mm_list) {
-		capture_bo(err++, vma->obj);
+		capture_bo(err++, vma);
 		if (++i == count)
 			break;
 	}
@@ -683,21 +705,27 @@ static u32 capture_active_bo(struct drm_i915_error_buffer *err,
 }
 
 static u32 capture_pinned_bo(struct drm_i915_error_buffer *err,
-			     int count, struct list_head *head)
+			     int count, struct list_head *head,
+			     struct i915_address_space *vm)
 {
 	struct drm_i915_gem_object *obj;
-	int i = 0;
+	struct drm_i915_error_buffer * const first = err;
+	struct drm_i915_error_buffer * const last = err + count;
 
 	list_for_each_entry(obj, head, global_list) {
-		if (!i915_gem_obj_is_pinned(obj))
-			continue;
+		struct i915_vma *vma;
 
-		capture_bo(err++, obj);
-		if (++i == count)
+		if (err == last)
 			break;
+
+		list_for_each_entry(vma, &obj->vma_list, vma_link)
+			if (vma->vm == vm && vma->pin_count > 0) {
+				capture_bo(err++, vma);
+				break;
+			}
 	}
 
-	return i;
+	return err - first;
 }
 
 /* Generate a semi-unique error code. The code is not meant to have meaning, The
@@ -741,6 +769,7 @@ static void i915_gem_record_fences(struct drm_device *dev,
 
 	/* Fences */
 	switch (INTEL_INFO(dev)->gen) {
+	case 9:
 	case 8:
 	case 7:
 	case 6:
@@ -780,9 +809,8 @@ static void gen8_record_semaphore_state(struct drm_i915_private *dev_priv,
 
 	if (!error->semaphore_obj)
 		error->semaphore_obj =
-			i915_error_object_create(dev_priv,
-						 dev_priv->semaphore_obj,
-						 &dev_priv->gtt.base);
+			i915_error_ggtt_object_create(dev_priv,
+						      dev_priv->semaphore_obj);
 
 	for_each_ring(to, dev_priv, i) {
 		int idx;
@@ -890,9 +918,6 @@ static void i915_record_ring_state(struct drm_device *dev,
 		ering->hws = I915_READ(mmio);
 	}
 
-	ering->cpu_ring_head = ring->buffer->head;
-	ering->cpu_ring_tail = ring->buffer->tail;
-
 	ering->hangcheck_score = ring->hangcheck.score;
 	ering->hangcheck_action = ring->hangcheck.action;
 
@@ -902,6 +927,7 @@ static void i915_record_ring_state(struct drm_device *dev,
 		ering->vm_info.gfx_mode = I915_READ(RING_MODE_GEN7(ring));
 
 		switch (INTEL_INFO(dev)->gen) {
+		case 9:
 		case 8:
 			for (i = 0; i < 4; i++) {
 				ering->vm_info.pdp[i] =
@@ -955,6 +981,7 @@ static void i915_gem_record_rings(struct drm_device *dev,
 
 	for (i = 0; i < I915_NUM_RINGS; i++) {
 		struct intel_engine_cs *ring = &dev_priv->ring[i];
+		struct intel_ringbuffer *rbuf;
 
 		error->ring[i].pid = -1;
 
@@ -967,6 +994,12 @@ static void i915_gem_record_rings(struct drm_device *dev,
 
 		request = i915_gem_find_active_request(ring);
 		if (request) {
+			struct i915_address_space *vm;
+
+			vm = request->ctx && request->ctx->ppgtt ?
+				&request->ctx->ppgtt->base :
+				&dev_priv->gtt.base;
+
 			/* We need to copy these to an anonymous buffer
 			 * as the simplest method to avoid being overwritten
 			 * by userspace.
@@ -974,12 +1007,9 @@ static void i915_gem_record_rings(struct drm_device *dev,
 			error->ring[i].batchbuffer =
 				i915_error_object_create(dev_priv,
 							 request->batch_obj,
-							 request->ctx ?
-							 request->ctx->vm :
-							 &dev_priv->gtt.base);
+							 vm);
 
-			if (HAS_BROKEN_CS_TLB(dev_priv->dev) &&
-			    ring->scratch.obj)
+			if (HAS_BROKEN_CS_TLB(dev_priv->dev))
 				error->ring[i].wa_batchbuffer =
 					i915_error_ggtt_object_create(dev_priv,
 							     ring->scratch.obj);
@@ -998,12 +1028,27 @@ static void i915_gem_record_rings(struct drm_device *dev,
 			}
 		}
 
+		if (i915.enable_execlists) {
+			/* TODO: This is only a small fix to keep basic error
+			 * capture working, but we need to add more information
+			 * for it to be useful (e.g. dump the context being
+			 * executed).
+			 */
+			if (request)
+				rbuf = request->ctx->engine[ring->id].ringbuf;
+			else
+				rbuf = ring->default_context->engine[ring->id].ringbuf;
+		} else
+			rbuf = ring->buffer;
+
+		error->ring[i].cpu_ring_head = rbuf->head;
+		error->ring[i].cpu_ring_tail = rbuf->tail;
+
 		error->ring[i].ringbuffer =
-			i915_error_ggtt_object_create(dev_priv, ring->buffer->obj);
+			i915_error_ggtt_object_create(dev_priv, rbuf->obj);
 
-		if (ring->status_page.obj)
-			error->ring[i].hws_page =
-				i915_error_ggtt_object_create(dev_priv, ring->status_page.obj);
+		error->ring[i].hws_page =
+			i915_error_ggtt_object_create(dev_priv, ring->status_page.obj);
 
 		i915_gem_record_active_context(ring, error, &error->ring[i]);
 
@@ -1049,9 +1094,14 @@ static void i915_gem_capture_vm(struct drm_i915_private *dev_priv,
 	list_for_each_entry(vma, &vm->active_list, mm_list)
 		i++;
 	error->active_bo_count[ndx] = i;
-	list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list)
-		if (i915_gem_obj_is_pinned(obj))
-			i++;
+
+	list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) {
+		list_for_each_entry(vma, &obj->vma_list, vma_link)
+			if (vma->vm == vm && vma->pin_count > 0) {
+				i++;
+				break;
+			}
+	}
 	error->pinned_bo_count[ndx] = i - error->active_bo_count[ndx];
 
 	if (i) {
@@ -1070,7 +1120,7 @@ static void i915_gem_capture_vm(struct drm_i915_private *dev_priv,
 		error->pinned_bo_count[ndx] =
 			capture_pinned_bo(pinned_bo,
 					  error->pinned_bo_count[ndx],
-					  &dev_priv->mm.bound_list);
+					  &dev_priv->mm.bound_list, vm);
 	error->active_bo[ndx] = active_bo;
 	error->pinned_bo[ndx] = pinned_bo;
 }
@@ -1091,8 +1141,25 @@ static void i915_gem_capture_buffers(struct drm_i915_private *dev_priv,
 	error->pinned_bo_count = kcalloc(cnt, sizeof(*error->pinned_bo_count),
 					 GFP_ATOMIC);
 
-	list_for_each_entry(vm, &dev_priv->vm_list, global_link)
-		i915_gem_capture_vm(dev_priv, error, vm, i++);
+	if (error->active_bo == NULL ||
+	    error->pinned_bo == NULL ||
+	    error->active_bo_count == NULL ||
+	    error->pinned_bo_count == NULL) {
+		kfree(error->active_bo);
+		kfree(error->active_bo_count);
+		kfree(error->pinned_bo);
+		kfree(error->pinned_bo_count);
+
+		error->active_bo = NULL;
+		error->active_bo_count = NULL;
+		error->pinned_bo = NULL;
+		error->pinned_bo_count = NULL;
+	} else {
+		list_for_each_entry(vm, &dev_priv->vm_list, global_link)
+			i915_gem_capture_vm(dev_priv, error, vm, i++);
+
+		error->vm_count = cnt;
+	}
 }
 
 /* Capture all registers which don't fit into another category. */
@@ -1176,7 +1243,8 @@ static void i915_error_capture_msg(struct drm_device *dev,
 	ecode = i915_error_generate_code(dev_priv, error, &ring_id);
 
 	len = scnprintf(error->error_msg, sizeof(error->error_msg),
-			"GPU HANG: ecode %d:0x%08x", ring_id, ecode);
+			"GPU HANG: ecode %d:%d:0x%08x",
+			INTEL_INFO(dev)->gen, ring_id, ecode);
 
 	if (ring_id != -1 && error->ring[ring_id].pid != -1)
 		len += scnprintf(error->error_msg + len,
@@ -1264,13 +1332,12 @@ void i915_error_state_get(struct drm_device *dev,
 			  struct i915_error_state_file_priv *error_priv)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	unsigned long flags;
 
-	spin_lock_irqsave(&dev_priv->gpu_error.lock, flags);
+	spin_lock_irq(&dev_priv->gpu_error.lock);
 	error_priv->error = dev_priv->gpu_error.first_error;
 	if (error_priv->error)
 		kref_get(&error_priv->error->ref);
-	spin_unlock_irqrestore(&dev_priv->gpu_error.lock, flags);
+	spin_unlock_irq(&dev_priv->gpu_error.lock);
 
 }
 
@@ -1284,22 +1351,21 @@ void i915_destroy_error_state(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_error_state *error;
-	unsigned long flags;
 
-	spin_lock_irqsave(&dev_priv->gpu_error.lock, flags);
+	spin_lock_irq(&dev_priv->gpu_error.lock);
 	error = dev_priv->gpu_error.first_error;
 	dev_priv->gpu_error.first_error = NULL;
-	spin_unlock_irqrestore(&dev_priv->gpu_error.lock, flags);
+	spin_unlock_irq(&dev_priv->gpu_error.lock);
 
 	if (error)
 		kref_put(&error->ref, i915_error_state_free);
 }
 
-const char *i915_cache_level_str(int type)
+const char *i915_cache_level_str(struct drm_i915_private *i915, int type)
 {
 	switch (type) {
 	case I915_CACHE_NONE: return " uncached";
-	case I915_CACHE_LLC: return " snooped or LLC";
+	case I915_CACHE_LLC: return HAS_LLC(i915) ? " LLC" : " snooped";
 	case I915_CACHE_L3_LLC: return " L3+LLC";
 	case I915_CACHE_WT: return " WT";
 	default: return "";
@@ -1327,6 +1393,7 @@ void i915_get_extra_instdone(struct drm_device *dev, uint32_t *instdone)
 		WARN_ONCE(1, "Unsupported platform\n");
 	case 7:
 	case 8:
+	case 9:
 		instdone[0] = I915_READ(GEN7_INSTDONE_1);
 		instdone[1] = I915_READ(GEN7_SC_INSTDONE);
 		instdone[2] = I915_READ(GEN7_SAMPLER_INSTDONE);
diff --git a/drivers/gpu/drm/i915/i915_ioc32.c b/drivers/gpu/drm/i915/i915_ioc32.c
index 2e0613e26251..176de6322e4d 100644
--- a/drivers/gpu/drm/i915/i915_ioc32.c
+++ b/drivers/gpu/drm/i915/i915_ioc32.c
@@ -189,7 +189,6 @@ static drm_ioctl_compat_t *i915_compat_ioctls[] = {
 	[DRM_I915_ALLOC] = compat_i915_alloc
 };
 
-#ifdef CONFIG_COMPAT
 /**
  * Called whenever a 32-bit process running under a 64-bit kernel
  * performs an ioctl on /dev/dri/card<n>.
@@ -218,4 +217,3 @@ long i915_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 
 	return ret;
 }
-#endif
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 0050ee9470f1..b051a238baf9 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -37,6 +37,14 @@
 #include "i915_trace.h"
 #include "intel_drv.h"
 
+/**
+ * DOC: interrupt handling
+ *
+ * These functions provide the basic support for enabling and disabling the
+ * interrupt handling support. There's a lot more functionality in i915_irq.c
+ * and related files, but that will be described in separate chapters.
+ */
+
 static const u32 hpd_ibx[] = {
 	[HPD_CRT] = SDE_CRT_HOTPLUG,
 	[HPD_SDVO_B] = SDE_SDVOB_HOTPLUG,
@@ -118,20 +126,22 @@ static const u32 hpd_status_i915[] = { /* i915 and valleyview are the same */
 
 #define GEN8_IRQ_INIT_NDX(type, which, imr_val, ier_val) do { \
 	GEN5_ASSERT_IIR_IS_ZERO(GEN8_##type##_IIR(which)); \
-	I915_WRITE(GEN8_##type##_IMR(which), (imr_val)); \
 	I915_WRITE(GEN8_##type##_IER(which), (ier_val)); \
-	POSTING_READ(GEN8_##type##_IER(which)); \
+	I915_WRITE(GEN8_##type##_IMR(which), (imr_val)); \
+	POSTING_READ(GEN8_##type##_IMR(which)); \
 } while (0)
 
 #define GEN5_IRQ_INIT(type, imr_val, ier_val) do { \
 	GEN5_ASSERT_IIR_IS_ZERO(type##IIR); \
-	I915_WRITE(type##IMR, (imr_val)); \
 	I915_WRITE(type##IER, (ier_val)); \
-	POSTING_READ(type##IER); \
+	I915_WRITE(type##IMR, (imr_val)); \
+	POSTING_READ(type##IMR); \
 } while (0)
 
+static void gen6_rps_irq_handler(struct drm_i915_private *dev_priv, u32 pm_iir);
+
 /* For display hotplug interrupt */
-static void
+void
 ironlake_enable_display_irq(struct drm_i915_private *dev_priv, u32 mask)
 {
 	assert_spin_locked(&dev_priv->irq_lock);
@@ -146,12 +156,12 @@ ironlake_enable_display_irq(struct drm_i915_private *dev_priv, u32 mask)
 	}
 }
 
-static void
+void
 ironlake_disable_display_irq(struct drm_i915_private *dev_priv, u32 mask)
 {
 	assert_spin_locked(&dev_priv->irq_lock);
 
-	if (!intel_irqs_enabled(dev_priv))
+	if (WARN_ON(!intel_irqs_enabled(dev_priv)))
 		return;
 
 	if ((dev_priv->irq_mask & mask) != mask) {
@@ -192,6 +202,21 @@ void gen5_disable_gt_irq(struct drm_i915_private *dev_priv, uint32_t mask)
 	ilk_update_gt_irq(dev_priv, mask, 0);
 }
 
+static u32 gen6_pm_iir(struct drm_i915_private *dev_priv)
+{
+	return INTEL_INFO(dev_priv)->gen >= 8 ? GEN8_GT_IIR(2) : GEN6_PMIIR;
+}
+
+static u32 gen6_pm_imr(struct drm_i915_private *dev_priv)
+{
+	return INTEL_INFO(dev_priv)->gen >= 8 ? GEN8_GT_IMR(2) : GEN6_PMIMR;
+}
+
+static u32 gen6_pm_ier(struct drm_i915_private *dev_priv)
+{
+	return INTEL_INFO(dev_priv)->gen >= 8 ? GEN8_GT_IER(2) : GEN6_PMIER;
+}
+
 /**
   * snb_update_pm_irq - update GEN6_PMIMR
   * @dev_priv: driver private
@@ -206,202 +231,107 @@ static void snb_update_pm_irq(struct drm_i915_private *dev_priv,
 
 	assert_spin_locked(&dev_priv->irq_lock);
 
-	if (WARN_ON(!intel_irqs_enabled(dev_priv)))
-		return;
-
 	new_val = dev_priv->pm_irq_mask;
 	new_val &= ~interrupt_mask;
 	new_val |= (~enabled_irq_mask & interrupt_mask);
 
 	if (new_val != dev_priv->pm_irq_mask) {
 		dev_priv->pm_irq_mask = new_val;
-		I915_WRITE(GEN6_PMIMR, dev_priv->pm_irq_mask);
-		POSTING_READ(GEN6_PMIMR);
+		I915_WRITE(gen6_pm_imr(dev_priv), dev_priv->pm_irq_mask);
+		POSTING_READ(gen6_pm_imr(dev_priv));
 	}
 }
 
 void gen6_enable_pm_irq(struct drm_i915_private *dev_priv, uint32_t mask)
 {
+	if (WARN_ON(!intel_irqs_enabled(dev_priv)))
+		return;
+
 	snb_update_pm_irq(dev_priv, mask, mask);
 }
 
-void gen6_disable_pm_irq(struct drm_i915_private *dev_priv, uint32_t mask)
+static void __gen6_disable_pm_irq(struct drm_i915_private *dev_priv,
+				  uint32_t mask)
 {
 	snb_update_pm_irq(dev_priv, mask, 0);
 }
 
-static bool ivb_can_enable_err_int(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_crtc *crtc;
-	enum pipe pipe;
-
-	assert_spin_locked(&dev_priv->irq_lock);
-
-	for_each_pipe(pipe) {
-		crtc = to_intel_crtc(dev_priv->pipe_to_crtc_mapping[pipe]);
-
-		if (crtc->cpu_fifo_underrun_disabled)
-			return false;
-	}
-
-	return true;
-}
-
-/**
-  * bdw_update_pm_irq - update GT interrupt 2
-  * @dev_priv: driver private
-  * @interrupt_mask: mask of interrupt bits to update
-  * @enabled_irq_mask: mask of interrupt bits to enable
-  *
-  * Copied from the snb function, updated with relevant register offsets
-  */
-static void bdw_update_pm_irq(struct drm_i915_private *dev_priv,
-			      uint32_t interrupt_mask,
-			      uint32_t enabled_irq_mask)
+void gen6_disable_pm_irq(struct drm_i915_private *dev_priv, uint32_t mask)
 {
-	uint32_t new_val;
-
-	assert_spin_locked(&dev_priv->irq_lock);
-
 	if (WARN_ON(!intel_irqs_enabled(dev_priv)))
 		return;
 
-	new_val = dev_priv->pm_irq_mask;
-	new_val &= ~interrupt_mask;
-	new_val |= (~enabled_irq_mask & interrupt_mask);
-
-	if (new_val != dev_priv->pm_irq_mask) {
-		dev_priv->pm_irq_mask = new_val;
-		I915_WRITE(GEN8_GT_IMR(2), dev_priv->pm_irq_mask);
-		POSTING_READ(GEN8_GT_IMR(2));
-	}
-}
-
-void gen8_enable_pm_irq(struct drm_i915_private *dev_priv, uint32_t mask)
-{
-	bdw_update_pm_irq(dev_priv, mask, mask);
+	__gen6_disable_pm_irq(dev_priv, mask);
 }
 
-void gen8_disable_pm_irq(struct drm_i915_private *dev_priv, uint32_t mask)
-{
-	bdw_update_pm_irq(dev_priv, mask, 0);
-}
-
-static bool cpt_can_enable_serr_int(struct drm_device *dev)
+void gen6_reset_rps_interrupts(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	enum pipe pipe;
-	struct intel_crtc *crtc;
+	uint32_t reg = gen6_pm_iir(dev_priv);
 
-	assert_spin_locked(&dev_priv->irq_lock);
-
-	for_each_pipe(pipe) {
-		crtc = to_intel_crtc(dev_priv->pipe_to_crtc_mapping[pipe]);
-
-		if (crtc->pch_fifo_underrun_disabled)
-			return false;
-	}
-
-	return true;
+	spin_lock_irq(&dev_priv->irq_lock);
+	I915_WRITE(reg, dev_priv->pm_rps_events);
+	I915_WRITE(reg, dev_priv->pm_rps_events);
+	POSTING_READ(reg);
+	spin_unlock_irq(&dev_priv->irq_lock);
 }
 
-void i9xx_check_fifo_underruns(struct drm_device *dev)
+void gen6_enable_rps_interrupts(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_crtc *crtc;
-	unsigned long flags;
-
-	spin_lock_irqsave(&dev_priv->irq_lock, flags);
-
-	for_each_intel_crtc(dev, crtc) {
-		u32 reg = PIPESTAT(crtc->pipe);
-		u32 pipestat;
-
-		if (crtc->cpu_fifo_underrun_disabled)
-			continue;
-
-		pipestat = I915_READ(reg) & 0xffff0000;
-		if ((pipestat & PIPE_FIFO_UNDERRUN_STATUS) == 0)
-			continue;
 
-		I915_WRITE(reg, pipestat | PIPE_FIFO_UNDERRUN_STATUS);
-		POSTING_READ(reg);
+	spin_lock_irq(&dev_priv->irq_lock);
 
-		DRM_ERROR("pipe %c underrun\n", pipe_name(crtc->pipe));
-	}
+	WARN_ON(dev_priv->rps.pm_iir);
+	WARN_ON(I915_READ(gen6_pm_iir(dev_priv)) & dev_priv->pm_rps_events);
+	dev_priv->rps.interrupts_enabled = true;
+	I915_WRITE(gen6_pm_ier(dev_priv), I915_READ(gen6_pm_ier(dev_priv)) |
+				dev_priv->pm_rps_events);
+	gen6_enable_pm_irq(dev_priv, dev_priv->pm_rps_events);
 
-	spin_unlock_irqrestore(&dev_priv->irq_lock, flags);
+	spin_unlock_irq(&dev_priv->irq_lock);
 }
 
-static void i9xx_set_fifo_underrun_reporting(struct drm_device *dev,
-					     enum pipe pipe,
-					     bool enable, bool old)
+u32 gen6_sanitize_rps_pm_mask(struct drm_i915_private *dev_priv, u32 mask)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 reg = PIPESTAT(pipe);
-	u32 pipestat = I915_READ(reg) & 0xffff0000;
+	/*
+	 * SNB,IVB can while VLV,CHV may hard hang on looping batchbuffer
+	 * if GEN6_PM_UP_EI_EXPIRED is masked.
+	 *
+	 * TODO: verify if this can be reproduced on VLV,CHV.
+	 */
+	if (INTEL_INFO(dev_priv)->gen <= 7 && !IS_HASWELL(dev_priv))
+		mask &= ~GEN6_PM_RP_UP_EI_EXPIRED;
 
-	assert_spin_locked(&dev_priv->irq_lock);
+	if (INTEL_INFO(dev_priv)->gen >= 8)
+		mask &= ~GEN8_PMINTR_REDIRECT_TO_NON_DISP;
 
-	if (enable) {
-		I915_WRITE(reg, pipestat | PIPE_FIFO_UNDERRUN_STATUS);
-		POSTING_READ(reg);
-	} else {
-		if (old && pipestat & PIPE_FIFO_UNDERRUN_STATUS)
-			DRM_ERROR("pipe %c underrun\n", pipe_name(pipe));
-	}
+	return mask;
 }
 
-static void ironlake_set_fifo_underrun_reporting(struct drm_device *dev,
-						 enum pipe pipe, bool enable)
+void gen6_disable_rps_interrupts(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	uint32_t bit = (pipe == PIPE_A) ? DE_PIPEA_FIFO_UNDERRUN :
-					  DE_PIPEB_FIFO_UNDERRUN;
-
-	if (enable)
-		ironlake_enable_display_irq(dev_priv, bit);
-	else
-		ironlake_disable_display_irq(dev_priv, bit);
-}
 
-static void ivybridge_set_fifo_underrun_reporting(struct drm_device *dev,
-						  enum pipe pipe,
-						  bool enable, bool old)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	if (enable) {
-		I915_WRITE(GEN7_ERR_INT, ERR_INT_FIFO_UNDERRUN(pipe));
+	spin_lock_irq(&dev_priv->irq_lock);
+	dev_priv->rps.interrupts_enabled = false;
+	spin_unlock_irq(&dev_priv->irq_lock);
 
-		if (!ivb_can_enable_err_int(dev))
-			return;
+	cancel_work_sync(&dev_priv->rps.work);
 
-		ironlake_enable_display_irq(dev_priv, DE_ERR_INT_IVB);
-	} else {
-		ironlake_disable_display_irq(dev_priv, DE_ERR_INT_IVB);
+	spin_lock_irq(&dev_priv->irq_lock);
 
-		if (old &&
-		    I915_READ(GEN7_ERR_INT) & ERR_INT_FIFO_UNDERRUN(pipe)) {
-			DRM_ERROR("uncleared fifo underrun on pipe %c\n",
-				  pipe_name(pipe));
-		}
-	}
-}
+	I915_WRITE(GEN6_PMINTRMSK, gen6_sanitize_rps_pm_mask(dev_priv, ~0));
 
-static void broadwell_set_fifo_underrun_reporting(struct drm_device *dev,
-						  enum pipe pipe, bool enable)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	__gen6_disable_pm_irq(dev_priv, dev_priv->pm_rps_events);
+	I915_WRITE(gen6_pm_ier(dev_priv), I915_READ(gen6_pm_ier(dev_priv)) &
+				~dev_priv->pm_rps_events);
+	I915_WRITE(gen6_pm_iir(dev_priv), dev_priv->pm_rps_events);
+	I915_WRITE(gen6_pm_iir(dev_priv), dev_priv->pm_rps_events);
 
-	assert_spin_locked(&dev_priv->irq_lock);
+	dev_priv->rps.pm_iir = 0;
 
-	if (enable)
-		dev_priv->de_irq_mask[pipe] &= ~GEN8_PIPE_FIFO_UNDERRUN;
-	else
-		dev_priv->de_irq_mask[pipe] |= GEN8_PIPE_FIFO_UNDERRUN;
-	I915_WRITE(GEN8_DE_PIPE_IMR(pipe), dev_priv->de_irq_mask[pipe]);
-	POSTING_READ(GEN8_DE_PIPE_IMR(pipe));
+	spin_unlock_irq(&dev_priv->irq_lock);
 }
 
 /**
@@ -410,9 +340,9 @@ static void broadwell_set_fifo_underrun_reporting(struct drm_device *dev,
  * @interrupt_mask: mask of interrupt bits to update
  * @enabled_irq_mask: mask of interrupt bits to enable
  */
-static void ibx_display_interrupt_update(struct drm_i915_private *dev_priv,
-					 uint32_t interrupt_mask,
-					 uint32_t enabled_irq_mask)
+void ibx_display_interrupt_update(struct drm_i915_private *dev_priv,
+				  uint32_t interrupt_mask,
+				  uint32_t enabled_irq_mask)
 {
 	uint32_t sdeimr = I915_READ(SDEIMR);
 	sdeimr &= ~interrupt_mask;
@@ -426,160 +356,6 @@ static void ibx_display_interrupt_update(struct drm_i915_private *dev_priv,
 	I915_WRITE(SDEIMR, sdeimr);
 	POSTING_READ(SDEIMR);
 }
-#define ibx_enable_display_interrupt(dev_priv, bits) \
-	ibx_display_interrupt_update((dev_priv), (bits), (bits))
-#define ibx_disable_display_interrupt(dev_priv, bits) \
-	ibx_display_interrupt_update((dev_priv), (bits), 0)
-
-static void ibx_set_fifo_underrun_reporting(struct drm_device *dev,
-					    enum transcoder pch_transcoder,
-					    bool enable)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	uint32_t bit = (pch_transcoder == TRANSCODER_A) ?
-		       SDE_TRANSA_FIFO_UNDER : SDE_TRANSB_FIFO_UNDER;
-
-	if (enable)
-		ibx_enable_display_interrupt(dev_priv, bit);
-	else
-		ibx_disable_display_interrupt(dev_priv, bit);
-}
-
-static void cpt_set_fifo_underrun_reporting(struct drm_device *dev,
-					    enum transcoder pch_transcoder,
-					    bool enable, bool old)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	if (enable) {
-		I915_WRITE(SERR_INT,
-			   SERR_INT_TRANS_FIFO_UNDERRUN(pch_transcoder));
-
-		if (!cpt_can_enable_serr_int(dev))
-			return;
-
-		ibx_enable_display_interrupt(dev_priv, SDE_ERROR_CPT);
-	} else {
-		ibx_disable_display_interrupt(dev_priv, SDE_ERROR_CPT);
-
-		if (old && I915_READ(SERR_INT) &
-		    SERR_INT_TRANS_FIFO_UNDERRUN(pch_transcoder)) {
-			DRM_ERROR("uncleared pch fifo underrun on pch transcoder %c\n",
-				  transcoder_name(pch_transcoder));
-		}
-	}
-}
-
-/**
- * intel_set_cpu_fifo_underrun_reporting - enable/disable FIFO underrun messages
- * @dev: drm device
- * @pipe: pipe
- * @enable: true if we want to report FIFO underrun errors, false otherwise
- *
- * This function makes us disable or enable CPU fifo underruns for a specific
- * pipe. Notice that on some Gens (e.g. IVB, HSW), disabling FIFO underrun
- * reporting for one pipe may also disable all the other CPU error interruts for
- * the other pipes, due to the fact that there's just one interrupt mask/enable
- * bit for all the pipes.
- *
- * Returns the previous state of underrun reporting.
- */
-static bool __intel_set_cpu_fifo_underrun_reporting(struct drm_device *dev,
-						    enum pipe pipe, bool enable)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pipe];
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	bool old;
-
-	assert_spin_locked(&dev_priv->irq_lock);
-
-	old = !intel_crtc->cpu_fifo_underrun_disabled;
-	intel_crtc->cpu_fifo_underrun_disabled = !enable;
-
-	if (INTEL_INFO(dev)->gen < 5 || IS_VALLEYVIEW(dev))
-		i9xx_set_fifo_underrun_reporting(dev, pipe, enable, old);
-	else if (IS_GEN5(dev) || IS_GEN6(dev))
-		ironlake_set_fifo_underrun_reporting(dev, pipe, enable);
-	else if (IS_GEN7(dev))
-		ivybridge_set_fifo_underrun_reporting(dev, pipe, enable, old);
-	else if (IS_GEN8(dev))
-		broadwell_set_fifo_underrun_reporting(dev, pipe, enable);
-
-	return old;
-}
-
-bool intel_set_cpu_fifo_underrun_reporting(struct drm_device *dev,
-					   enum pipe pipe, bool enable)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	unsigned long flags;
-	bool ret;
-
-	spin_lock_irqsave(&dev_priv->irq_lock, flags);
-	ret = __intel_set_cpu_fifo_underrun_reporting(dev, pipe, enable);
-	spin_unlock_irqrestore(&dev_priv->irq_lock, flags);
-
-	return ret;
-}
-
-static bool __cpu_fifo_underrun_reporting_enabled(struct drm_device *dev,
-						  enum pipe pipe)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pipe];
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-
-	return !intel_crtc->cpu_fifo_underrun_disabled;
-}
-
-/**
- * intel_set_pch_fifo_underrun_reporting - enable/disable FIFO underrun messages
- * @dev: drm device
- * @pch_transcoder: the PCH transcoder (same as pipe on IVB and older)
- * @enable: true if we want to report FIFO underrun errors, false otherwise
- *
- * This function makes us disable or enable PCH fifo underruns for a specific
- * PCH transcoder. Notice that on some PCHs (e.g. CPT/PPT), disabling FIFO
- * underrun reporting for one transcoder may also disable all the other PCH
- * error interruts for the other transcoders, due to the fact that there's just
- * one interrupt mask/enable bit for all the transcoders.
- *
- * Returns the previous state of underrun reporting.
- */
-bool intel_set_pch_fifo_underrun_reporting(struct drm_device *dev,
-					   enum transcoder pch_transcoder,
-					   bool enable)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pch_transcoder];
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	unsigned long flags;
-	bool old;
-
-	/*
-	 * NOTE: Pre-LPT has a fixed cpu pipe -> pch transcoder mapping, but LPT
-	 * has only one pch transcoder A that all pipes can use. To avoid racy
-	 * pch transcoder -> pipe lookups from interrupt code simply store the
-	 * underrun statistics in crtc A. Since we never expose this anywhere
-	 * nor use it outside of the fifo underrun code here using the "wrong"
-	 * crtc on LPT won't cause issues.
-	 */
-
-	spin_lock_irqsave(&dev_priv->irq_lock, flags);
-
-	old = !intel_crtc->pch_fifo_underrun_disabled;
-	intel_crtc->pch_fifo_underrun_disabled = !enable;
-
-	if (HAS_PCH_IBX(dev))
-		ibx_set_fifo_underrun_reporting(dev, pch_transcoder, enable);
-	else
-		cpt_set_fifo_underrun_reporting(dev, pch_transcoder, enable, old);
-
-	spin_unlock_irqrestore(&dev_priv->irq_lock, flags);
-	return old;
-}
-
 
 static void
 __i915_enable_pipestat(struct drm_i915_private *dev_priv, enum pipe pipe,
@@ -589,6 +365,7 @@ __i915_enable_pipestat(struct drm_i915_private *dev_priv, enum pipe pipe,
 	u32 pipestat = I915_READ(reg) & PIPESTAT_INT_ENABLE_MASK;
 
 	assert_spin_locked(&dev_priv->irq_lock);
+	WARN_ON(!intel_irqs_enabled(dev_priv));
 
 	if (WARN_ONCE(enable_mask & ~PIPESTAT_INT_ENABLE_MASK ||
 		      status_mask & ~PIPESTAT_INT_STATUS_MASK,
@@ -615,6 +392,7 @@ __i915_disable_pipestat(struct drm_i915_private *dev_priv, enum pipe pipe,
 	u32 pipestat = I915_READ(reg) & PIPESTAT_INT_ENABLE_MASK;
 
 	assert_spin_locked(&dev_priv->irq_lock);
+	WARN_ON(!intel_irqs_enabled(dev_priv));
 
 	if (WARN_ONCE(enable_mask & ~PIPESTAT_INT_ENABLE_MASK ||
 		      status_mask & ~PIPESTAT_INT_STATUS_MASK,
@@ -694,19 +472,18 @@ i915_disable_pipestat(struct drm_i915_private *dev_priv, enum pipe pipe,
 static void i915_enable_asle_pipestat(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	unsigned long irqflags;
 
 	if (!dev_priv->opregion.asle || !IS_MOBILE(dev))
 		return;
 
-	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+	spin_lock_irq(&dev_priv->irq_lock);
 
 	i915_enable_pipestat(dev_priv, PIPE_B, PIPE_LEGACY_BLC_EVENT_STATUS);
 	if (INTEL_INFO(dev)->gen >= 4)
 		i915_enable_pipestat(dev_priv, PIPE_A,
 				     PIPE_LEGACY_BLC_EVENT_STATUS);
 
-	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+	spin_unlock_irq(&dev_priv->irq_lock);
 }
 
 /**
@@ -1020,7 +797,7 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, int pipe,
 
 	/* In vblank? */
 	if (in_vbl)
-		ret |= DRM_SCANOUTPOS_INVBL;
+		ret |= DRM_SCANOUTPOS_IN_VBLANK;
 
 	return ret;
 }
@@ -1094,18 +871,17 @@ static void i915_digport_work_func(struct work_struct *work)
 {
 	struct drm_i915_private *dev_priv =
 		container_of(work, struct drm_i915_private, dig_port_work);
-	unsigned long irqflags;
 	u32 long_port_mask, short_port_mask;
 	struct intel_digital_port *intel_dig_port;
 	int i, ret;
 	u32 old_bits = 0;
 
-	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+	spin_lock_irq(&dev_priv->irq_lock);
 	long_port_mask = dev_priv->long_hpd_port_mask;
 	dev_priv->long_hpd_port_mask = 0;
 	short_port_mask = dev_priv->short_hpd_port_mask;
 	dev_priv->short_hpd_port_mask = 0;
-	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+	spin_unlock_irq(&dev_priv->irq_lock);
 
 	for (i = 0; i < I915_MAX_PORTS; i++) {
 		bool valid = false;
@@ -1130,9 +906,9 @@ static void i915_digport_work_func(struct work_struct *work)
 	}
 
 	if (old_bits) {
-		spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+		spin_lock_irq(&dev_priv->irq_lock);
 		dev_priv->hpd_event_bits |= old_bits;
-		spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+		spin_unlock_irq(&dev_priv->irq_lock);
 		schedule_work(&dev_priv->hotplug_work);
 	}
 }
@@ -1151,7 +927,6 @@ static void i915_hotplug_work_func(struct work_struct *work)
 	struct intel_connector *intel_connector;
 	struct intel_encoder *intel_encoder;
 	struct drm_connector *connector;
-	unsigned long irqflags;
 	bool hpd_disabled = false;
 	bool changed = false;
 	u32 hpd_event_bits;
@@ -1159,7 +934,7 @@ static void i915_hotplug_work_func(struct work_struct *work)
 	mutex_lock(&mode_config->mutex);
 	DRM_DEBUG_KMS("running encoder hotplug functions\n");
 
-	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+	spin_lock_irq(&dev_priv->irq_lock);
 
 	hpd_event_bits = dev_priv->hpd_event_bits;
 	dev_priv->hpd_event_bits = 0;
@@ -1193,7 +968,7 @@ static void i915_hotplug_work_func(struct work_struct *work)
 				 msecs_to_jiffies(I915_REENABLE_HOTPLUG_DELAY));
 	}
 
-	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+	spin_unlock_irq(&dev_priv->irq_lock);
 
 	list_for_each_entry(connector, &mode_config->connector_list, head) {
 		intel_connector = to_intel_connector(connector);
@@ -1260,11 +1035,7 @@ static void notify_ring(struct drm_device *dev,
 
 	trace_i915_gem_request_complete(ring);
 
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		intel_notify_mmio_flip(ring);
-
 	wake_up_all(&ring->irq_queue);
-	i915_queue_hangcheck(dev);
 }
 
 static u32 vlv_c0_residency(struct drm_i915_private *dev_priv,
@@ -1322,10 +1093,10 @@ static u32 vlv_c0_residency(struct drm_i915_private *dev_priv,
  * @dev_priv: DRM device private
  *
  */
-static u32 vlv_calc_delay_from_C0_counters(struct drm_i915_private *dev_priv)
+static int vlv_calc_delay_from_C0_counters(struct drm_i915_private *dev_priv)
 {
 	u32 residency_C0_up = 0, residency_C0_down = 0;
-	u8 new_delay, adj;
+	int new_delay, adj;
 
 	dev_priv->rps.ei_interrupt_count++;
 
@@ -1400,14 +1171,15 @@ static void gen6_pm_rps_work(struct work_struct *work)
 	int new_delay, adj;
 
 	spin_lock_irq(&dev_priv->irq_lock);
+	/* Speed up work cancelation during disabling rps interrupts. */
+	if (!dev_priv->rps.interrupts_enabled) {
+		spin_unlock_irq(&dev_priv->irq_lock);
+		return;
+	}
 	pm_iir = dev_priv->rps.pm_iir;
 	dev_priv->rps.pm_iir = 0;
-	if (INTEL_INFO(dev_priv->dev)->gen >= 8)
-		gen8_enable_pm_irq(dev_priv, dev_priv->pm_rps_events);
-	else {
-		/* Make sure not to corrupt PMIMR state used by ringbuffer */
-		gen6_enable_pm_irq(dev_priv, dev_priv->pm_rps_events);
-	}
+	/* Make sure not to corrupt PMIMR state used by ringbuffer on GEN6 */
+	gen6_enable_pm_irq(dev_priv, dev_priv->pm_rps_events);
 	spin_unlock_irq(&dev_priv->irq_lock);
 
 	/* Make sure we didn't queue anything we're not going to process. */
@@ -1488,7 +1260,6 @@ static void ivybridge_parity_work(struct work_struct *work)
 	u32 error_status, row, bank, subbank;
 	char *parity_event[6];
 	uint32_t misccpctl;
-	unsigned long flags;
 	uint8_t slice = 0;
 
 	/* We must turn off DOP level clock gating to access the L3 registers.
@@ -1547,9 +1318,9 @@ static void ivybridge_parity_work(struct work_struct *work)
 
 out:
 	WARN_ON(dev_priv->l3_parity.which_slice);
-	spin_lock_irqsave(&dev_priv->irq_lock, flags);
+	spin_lock_irq(&dev_priv->irq_lock);
 	gen5_enable_gt_irq(dev_priv, GT_PARITY_ERROR(dev_priv->dev));
-	spin_unlock_irqrestore(&dev_priv->irq_lock, flags);
+	spin_unlock_irq(&dev_priv->irq_lock);
 
 	mutex_unlock(&dev_priv->dev->struct_mutex);
 }
@@ -1601,32 +1372,18 @@ static void snb_gt_irq_handler(struct drm_device *dev,
 
 	if (gt_iir & (GT_BLT_CS_ERROR_INTERRUPT |
 		      GT_BSD_CS_ERROR_INTERRUPT |
-		      GT_RENDER_CS_MASTER_ERROR_INTERRUPT)) {
-		i915_handle_error(dev, false, "GT error interrupt 0x%08x",
-				  gt_iir);
-	}
+		      GT_RENDER_CS_MASTER_ERROR_INTERRUPT))
+		DRM_DEBUG("Command parser error, gt_iir 0x%08x\n", gt_iir);
 
 	if (gt_iir & GT_PARITY_ERROR(dev))
 		ivybridge_parity_error_irq_handler(dev, gt_iir);
 }
 
-static void gen8_rps_irq_handler(struct drm_i915_private *dev_priv, u32 pm_iir)
-{
-	if ((pm_iir & dev_priv->pm_rps_events) == 0)
-		return;
-
-	spin_lock(&dev_priv->irq_lock);
-	dev_priv->rps.pm_iir |= pm_iir & dev_priv->pm_rps_events;
-	gen8_disable_pm_irq(dev_priv, pm_iir & dev_priv->pm_rps_events);
-	spin_unlock(&dev_priv->irq_lock);
-
-	queue_work(dev_priv->wq, &dev_priv->rps.work);
-}
-
 static irqreturn_t gen8_gt_irq_handler(struct drm_device *dev,
 				       struct drm_i915_private *dev_priv,
 				       u32 master_ctl)
 {
+	struct intel_engine_cs *ring;
 	u32 rcs, bcs, vcs;
 	uint32_t tmp = 0;
 	irqreturn_t ret = IRQ_NONE;
@@ -1636,12 +1393,20 @@ static irqreturn_t gen8_gt_irq_handler(struct drm_device *dev,
 		if (tmp) {
 			I915_WRITE(GEN8_GT_IIR(0), tmp);
 			ret = IRQ_HANDLED;
+
 			rcs = tmp >> GEN8_RCS_IRQ_SHIFT;
-			bcs = tmp >> GEN8_BCS_IRQ_SHIFT;
+			ring = &dev_priv->ring[RCS];
 			if (rcs & GT_RENDER_USER_INTERRUPT)
-				notify_ring(dev, &dev_priv->ring[RCS]);
+				notify_ring(dev, ring);
+			if (rcs & GT_CONTEXT_SWITCH_INTERRUPT)
+				intel_execlists_handle_ctx_events(ring);
+
+			bcs = tmp >> GEN8_BCS_IRQ_SHIFT;
+			ring = &dev_priv->ring[BCS];
 			if (bcs & GT_RENDER_USER_INTERRUPT)
-				notify_ring(dev, &dev_priv->ring[BCS]);
+				notify_ring(dev, ring);
+			if (bcs & GT_CONTEXT_SWITCH_INTERRUPT)
+				intel_execlists_handle_ctx_events(ring);
 		} else
 			DRM_ERROR("The master control interrupt lied (GT0)!\n");
 	}
@@ -1651,12 +1416,20 @@ static irqreturn_t gen8_gt_irq_handler(struct drm_device *dev,
 		if (tmp) {
 			I915_WRITE(GEN8_GT_IIR(1), tmp);
 			ret = IRQ_HANDLED;
+
 			vcs = tmp >> GEN8_VCS1_IRQ_SHIFT;
+			ring = &dev_priv->ring[VCS];
 			if (vcs & GT_RENDER_USER_INTERRUPT)
-				notify_ring(dev, &dev_priv->ring[VCS]);
+				notify_ring(dev, ring);
+			if (vcs & GT_CONTEXT_SWITCH_INTERRUPT)
+				intel_execlists_handle_ctx_events(ring);
+
 			vcs = tmp >> GEN8_VCS2_IRQ_SHIFT;
+			ring = &dev_priv->ring[VCS2];
 			if (vcs & GT_RENDER_USER_INTERRUPT)
-				notify_ring(dev, &dev_priv->ring[VCS2]);
+				notify_ring(dev, ring);
+			if (vcs & GT_CONTEXT_SWITCH_INTERRUPT)
+				intel_execlists_handle_ctx_events(ring);
 		} else
 			DRM_ERROR("The master control interrupt lied (GT1)!\n");
 	}
@@ -1667,7 +1440,7 @@ static irqreturn_t gen8_gt_irq_handler(struct drm_device *dev,
 			I915_WRITE(GEN8_GT_IIR(2),
 				   tmp & dev_priv->pm_rps_events);
 			ret = IRQ_HANDLED;
-			gen8_rps_irq_handler(dev_priv, tmp);
+			gen6_rps_irq_handler(dev_priv, tmp);
 		} else
 			DRM_ERROR("The master control interrupt lied (PM)!\n");
 	}
@@ -1677,9 +1450,13 @@ static irqreturn_t gen8_gt_irq_handler(struct drm_device *dev,
 		if (tmp) {
 			I915_WRITE(GEN8_GT_IIR(3), tmp);
 			ret = IRQ_HANDLED;
+
 			vcs = tmp >> GEN8_VECS_IRQ_SHIFT;
+			ring = &dev_priv->ring[VECS];
 			if (vcs & GT_RENDER_USER_INTERRUPT)
-				notify_ring(dev, &dev_priv->ring[VECS]);
+				notify_ring(dev, ring);
+			if (vcs & GT_CONTEXT_SWITCH_INTERRUPT)
+				intel_execlists_handle_ctx_events(ring);
 		} else
 			DRM_ERROR("The master control interrupt lied (GT3)!\n");
 	}
@@ -1690,7 +1467,7 @@ static irqreturn_t gen8_gt_irq_handler(struct drm_device *dev,
 #define HPD_STORM_DETECT_PERIOD 1000
 #define HPD_STORM_THRESHOLD 5
 
-static int ilk_port_to_hotplug_shift(enum port port)
+static int pch_port_to_hotplug_shift(enum port port)
 {
 	switch (port) {
 	case PORT_A:
@@ -1706,7 +1483,7 @@ static int ilk_port_to_hotplug_shift(enum port port)
 	}
 }
 
-static int g4x_port_to_hotplug_shift(enum port port)
+static int i915_port_to_hotplug_shift(enum port port)
 {
 	switch (port) {
 	case PORT_A:
@@ -1764,15 +1541,17 @@ static inline void intel_hpd_irq_handler(struct drm_device *dev,
 		if (port && dev_priv->hpd_irq_port[port]) {
 			bool long_hpd;
 
-			if (IS_G4X(dev)) {
-				dig_shift = g4x_port_to_hotplug_shift(port);
-				long_hpd = (hotplug_trigger >> dig_shift) & PORTB_HOTPLUG_LONG_DETECT;
-			} else {
-				dig_shift = ilk_port_to_hotplug_shift(port);
+			if (HAS_PCH_SPLIT(dev)) {
+				dig_shift = pch_port_to_hotplug_shift(port);
 				long_hpd = (dig_hotplug_reg >> dig_shift) & PORTB_HOTPLUG_LONG_DETECT;
+			} else {
+				dig_shift = i915_port_to_hotplug_shift(port);
+				long_hpd = (hotplug_trigger >> dig_shift) & PORTB_HOTPLUG_LONG_DETECT;
 			}
 
-			DRM_DEBUG_DRIVER("digital hpd port %d %d\n", port, long_hpd);
+			DRM_DEBUG_DRIVER("digital hpd port %c - %s\n",
+					 port_name(port),
+					 long_hpd ? "long" : "short");
 			/* for long HPD pulses we want to have the digital queue happen,
 			   but we still want HPD storm detection to function. */
 			if (long_hpd) {
@@ -1875,7 +1654,7 @@ static void display_pipe_crc_irq_handler(struct drm_device *dev, enum pipe pipe,
 
 	if (!pipe_crc->entries) {
 		spin_unlock(&pipe_crc->lock);
-		DRM_ERROR("spurious interrupt\n");
+		DRM_DEBUG_KMS("spurious interrupt\n");
 		return;
 	}
 
@@ -1961,37 +1740,38 @@ static void i9xx_pipe_crc_irq_handler(struct drm_device *dev, enum pipe pipe)
  * the work queue. */
 static void gen6_rps_irq_handler(struct drm_i915_private *dev_priv, u32 pm_iir)
 {
+	/* TODO: RPS on GEN9+ is not supported yet. */
+	if (WARN_ONCE(INTEL_INFO(dev_priv)->gen >= 9,
+		      "GEN9+: unexpected RPS IRQ\n"))
+		return;
+
 	if (pm_iir & dev_priv->pm_rps_events) {
 		spin_lock(&dev_priv->irq_lock);
-		dev_priv->rps.pm_iir |= pm_iir & dev_priv->pm_rps_events;
 		gen6_disable_pm_irq(dev_priv, pm_iir & dev_priv->pm_rps_events);
+		if (dev_priv->rps.interrupts_enabled) {
+			dev_priv->rps.pm_iir |= pm_iir & dev_priv->pm_rps_events;
+			queue_work(dev_priv->wq, &dev_priv->rps.work);
+		}
 		spin_unlock(&dev_priv->irq_lock);
-
-		queue_work(dev_priv->wq, &dev_priv->rps.work);
 	}
 
+	if (INTEL_INFO(dev_priv)->gen >= 8)
+		return;
+
 	if (HAS_VEBOX(dev_priv->dev)) {
 		if (pm_iir & PM_VEBOX_USER_INTERRUPT)
 			notify_ring(dev_priv->dev, &dev_priv->ring[VECS]);
 
-		if (pm_iir & PM_VEBOX_CS_ERROR_INTERRUPT) {
-			i915_handle_error(dev_priv->dev, false,
-					  "VEBOX CS error interrupt 0x%08x",
-					  pm_iir);
-		}
+		if (pm_iir & PM_VEBOX_CS_ERROR_INTERRUPT)
+			DRM_DEBUG("Command parser error, pm_iir 0x%08x\n", pm_iir);
 	}
 }
 
 static bool intel_pipe_handle_vblank(struct drm_device *dev, enum pipe pipe)
 {
-	struct intel_crtc *crtc;
-
 	if (!drm_handle_vblank(dev, pipe))
 		return false;
 
-	crtc = to_intel_crtc(intel_get_crtc_for_pipe(dev, pipe));
-	wake_up(&crtc->vbl_wait);
-
 	return true;
 }
 
@@ -2002,7 +1782,7 @@ static void valleyview_pipestat_irq_handler(struct drm_device *dev, u32 iir)
 	int pipe;
 
 	spin_lock(&dev_priv->irq_lock);
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		int reg;
 		u32 mask, iir_bit = 0;
 
@@ -2013,9 +1793,9 @@ static void valleyview_pipestat_irq_handler(struct drm_device *dev, u32 iir)
 		 * we need to be careful that we only handle what we want to
 		 * handle.
 		 */
-		mask = 0;
-		if (__cpu_fifo_underrun_reporting_enabled(dev, pipe))
-			mask |= PIPE_FIFO_UNDERRUN_STATUS;
+
+		/* fifo underruns are filterered in the underrun handler. */
+		mask = PIPE_FIFO_UNDERRUN_STATUS;
 
 		switch (pipe) {
 		case PIPE_A:
@@ -2047,9 +1827,10 @@ static void valleyview_pipestat_irq_handler(struct drm_device *dev, u32 iir)
 	}
 	spin_unlock(&dev_priv->irq_lock);
 
-	for_each_pipe(pipe) {
-		if (pipe_stats[pipe] & PIPE_START_VBLANK_INTERRUPT_STATUS)
-			intel_pipe_handle_vblank(dev, pipe);
+	for_each_pipe(dev_priv, pipe) {
+		if (pipe_stats[pipe] & PIPE_START_VBLANK_INTERRUPT_STATUS &&
+		    intel_pipe_handle_vblank(dev, pipe))
+			intel_check_page_flip(dev, pipe);
 
 		if (pipe_stats[pipe] & PLANE_FLIP_DONE_INT_STATUS_VLV) {
 			intel_prepare_page_flip(dev, pipe);
@@ -2059,9 +1840,8 @@ static void valleyview_pipestat_irq_handler(struct drm_device *dev, u32 iir)
 		if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
 			i9xx_pipe_crc_irq_handler(dev, pipe);
 
-		if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS &&
-		    intel_set_cpu_fifo_underrun_reporting(dev, pipe, false))
-			DRM_ERROR("pipe %c underrun\n", pipe_name(pipe));
+		if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS)
+			intel_cpu_fifo_underrun_irq_handler(dev_priv, pipe);
 	}
 
 	if (pipe_stats[0] & PIPE_GMBUS_INTERRUPT_STATUS)
@@ -2216,7 +1996,7 @@ static void ibx_irq_handler(struct drm_device *dev, u32 pch_iir)
 		DRM_ERROR("PCH poison interrupt\n");
 
 	if (pch_iir & SDE_FDI_MASK)
-		for_each_pipe(pipe)
+		for_each_pipe(dev_priv, pipe)
 			DRM_DEBUG_DRIVER("  pipe %c FDI IIR: 0x%08x\n",
 					 pipe_name(pipe),
 					 I915_READ(FDI_RX_IIR(pipe)));
@@ -2228,14 +2008,10 @@ static void ibx_irq_handler(struct drm_device *dev, u32 pch_iir)
 		DRM_DEBUG_DRIVER("PCH transcoder CRC error interrupt\n");
 
 	if (pch_iir & SDE_TRANSA_FIFO_UNDER)
-		if (intel_set_pch_fifo_underrun_reporting(dev, TRANSCODER_A,
-							  false))
-			DRM_ERROR("PCH transcoder A FIFO underrun\n");
+		intel_pch_fifo_underrun_irq_handler(dev_priv, TRANSCODER_A);
 
 	if (pch_iir & SDE_TRANSB_FIFO_UNDER)
-		if (intel_set_pch_fifo_underrun_reporting(dev, TRANSCODER_B,
-							  false))
-			DRM_ERROR("PCH transcoder B FIFO underrun\n");
+		intel_pch_fifo_underrun_irq_handler(dev_priv, TRANSCODER_B);
 }
 
 static void ivb_err_int_handler(struct drm_device *dev)
@@ -2247,13 +2023,9 @@ static void ivb_err_int_handler(struct drm_device *dev)
 	if (err_int & ERR_INT_POISON)
 		DRM_ERROR("Poison interrupt\n");
 
-	for_each_pipe(pipe) {
-		if (err_int & ERR_INT_FIFO_UNDERRUN(pipe)) {
-			if (intel_set_cpu_fifo_underrun_reporting(dev, pipe,
-								  false))
-				DRM_ERROR("Pipe %c FIFO underrun\n",
-					  pipe_name(pipe));
-		}
+	for_each_pipe(dev_priv, pipe) {
+		if (err_int & ERR_INT_FIFO_UNDERRUN(pipe))
+			intel_cpu_fifo_underrun_irq_handler(dev_priv, pipe);
 
 		if (err_int & ERR_INT_PIPE_CRC_DONE(pipe)) {
 			if (IS_IVYBRIDGE(dev))
@@ -2275,19 +2047,13 @@ static void cpt_serr_int_handler(struct drm_device *dev)
 		DRM_ERROR("PCH poison interrupt\n");
 
 	if (serr_int & SERR_INT_TRANS_A_FIFO_UNDERRUN)
-		if (intel_set_pch_fifo_underrun_reporting(dev, TRANSCODER_A,
-							  false))
-			DRM_ERROR("PCH transcoder A FIFO underrun\n");
+		intel_pch_fifo_underrun_irq_handler(dev_priv, TRANSCODER_A);
 
 	if (serr_int & SERR_INT_TRANS_B_FIFO_UNDERRUN)
-		if (intel_set_pch_fifo_underrun_reporting(dev, TRANSCODER_B,
-							  false))
-			DRM_ERROR("PCH transcoder B FIFO underrun\n");
+		intel_pch_fifo_underrun_irq_handler(dev_priv, TRANSCODER_B);
 
 	if (serr_int & SERR_INT_TRANS_C_FIFO_UNDERRUN)
-		if (intel_set_pch_fifo_underrun_reporting(dev, TRANSCODER_C,
-							  false))
-			DRM_ERROR("PCH transcoder C FIFO underrun\n");
+		intel_pch_fifo_underrun_irq_handler(dev_priv, TRANSCODER_C);
 
 	I915_WRITE(SERR_INT, serr_int);
 }
@@ -2324,7 +2090,7 @@ static void cpt_irq_handler(struct drm_device *dev, u32 pch_iir)
 		DRM_DEBUG_DRIVER("Audio CP change interrupt\n");
 
 	if (pch_iir & SDE_FDI_MASK_CPT)
-		for_each_pipe(pipe)
+		for_each_pipe(dev_priv, pipe)
 			DRM_DEBUG_DRIVER("  pipe %c FDI IIR: 0x%08x\n",
 					 pipe_name(pipe),
 					 I915_READ(FDI_RX_IIR(pipe)));
@@ -2347,14 +2113,13 @@ static void ilk_display_irq_handler(struct drm_device *dev, u32 de_iir)
 	if (de_iir & DE_POISON)
 		DRM_ERROR("Poison interrupt\n");
 
-	for_each_pipe(pipe) {
-		if (de_iir & DE_PIPE_VBLANK(pipe))
-			intel_pipe_handle_vblank(dev, pipe);
+	for_each_pipe(dev_priv, pipe) {
+		if (de_iir & DE_PIPE_VBLANK(pipe) &&
+		    intel_pipe_handle_vblank(dev, pipe))
+			intel_check_page_flip(dev, pipe);
 
 		if (de_iir & DE_PIPE_FIFO_UNDERRUN(pipe))
-			if (intel_set_cpu_fifo_underrun_reporting(dev, pipe, false))
-				DRM_ERROR("Pipe %c FIFO underrun\n",
-					  pipe_name(pipe));
+			intel_cpu_fifo_underrun_irq_handler(dev_priv, pipe);
 
 		if (de_iir & DE_PIPE_CRC_DONE(pipe))
 			i9xx_pipe_crc_irq_handler(dev, pipe);
@@ -2397,9 +2162,10 @@ static void ivb_display_irq_handler(struct drm_device *dev, u32 de_iir)
 	if (de_iir & DE_GSE_IVB)
 		intel_opregion_asle_intr(dev);
 
-	for_each_pipe(pipe) {
-		if (de_iir & (DE_PIPE_VBLANK_IVB(pipe)))
-			intel_pipe_handle_vblank(dev, pipe);
+	for_each_pipe(dev_priv, pipe) {
+		if (de_iir & (DE_PIPE_VBLANK_IVB(pipe)) &&
+		    intel_pipe_handle_vblank(dev, pipe))
+			intel_check_page_flip(dev, pipe);
 
 		/* plane/pipes map 1:1 on ilk+ */
 		if (de_iir & DE_PLANE_FLIP_DONE_IVB(pipe)) {
@@ -2503,6 +2269,11 @@ static irqreturn_t gen8_irq_handler(int irq, void *arg)
 	irqreturn_t ret = IRQ_NONE;
 	uint32_t tmp = 0;
 	enum pipe pipe;
+	u32 aux_mask = GEN8_AUX_CHANNEL_A;
+
+	if (IS_GEN9(dev))
+		aux_mask |=  GEN9_AUX_CHANNEL_B | GEN9_AUX_CHANNEL_C |
+			GEN9_AUX_CHANNEL_D;
 
 	master_ctl = I915_READ(GEN8_MASTER_IRQ);
 	master_ctl &= ~GEN8_MASTER_IRQ_CONTROL;
@@ -2535,7 +2306,8 @@ static irqreturn_t gen8_irq_handler(int irq, void *arg)
 		if (tmp) {
 			I915_WRITE(GEN8_DE_PORT_IIR, tmp);
 			ret = IRQ_HANDLED;
-			if (tmp & GEN8_AUX_CHANNEL_A)
+
+			if (tmp & aux_mask)
 				dp_aux_irq_handler(dev);
 			else
 				DRM_ERROR("Unexpected DE Port interrupt\n");
@@ -2544,8 +2316,8 @@ static irqreturn_t gen8_irq_handler(int irq, void *arg)
 			DRM_ERROR("The master control interrupt lied (DE PORT)!\n");
 	}
 
-	for_each_pipe(pipe) {
-		uint32_t pipe_iir;
+	for_each_pipe(dev_priv, pipe) {
+		uint32_t pipe_iir, flip_done = 0, fault_errors = 0;
 
 		if (!(master_ctl & GEN8_DE_PIPE_IRQ(pipe)))
 			continue;
@@ -2554,10 +2326,17 @@ static irqreturn_t gen8_irq_handler(int irq, void *arg)
 		if (pipe_iir) {
 			ret = IRQ_HANDLED;
 			I915_WRITE(GEN8_DE_PIPE_IIR(pipe), pipe_iir);
-			if (pipe_iir & GEN8_PIPE_VBLANK)
-				intel_pipe_handle_vblank(dev, pipe);
 
-			if (pipe_iir & GEN8_PIPE_PRIMARY_FLIP_DONE) {
+			if (pipe_iir & GEN8_PIPE_VBLANK &&
+			    intel_pipe_handle_vblank(dev, pipe))
+				intel_check_page_flip(dev, pipe);
+
+			if (IS_GEN9(dev))
+				flip_done = pipe_iir & GEN9_PIPE_PLANE1_FLIP_DONE;
+			else
+				flip_done = pipe_iir & GEN8_PIPE_PRIMARY_FLIP_DONE;
+
+			if (flip_done) {
 				intel_prepare_page_flip(dev, pipe);
 				intel_finish_page_flip_plane(dev, pipe);
 			}
@@ -2565,18 +2344,20 @@ static irqreturn_t gen8_irq_handler(int irq, void *arg)
 			if (pipe_iir & GEN8_PIPE_CDCLK_CRC_DONE)
 				hsw_pipe_crc_irq_handler(dev, pipe);
 
-			if (pipe_iir & GEN8_PIPE_FIFO_UNDERRUN) {
-				if (intel_set_cpu_fifo_underrun_reporting(dev, pipe,
-									  false))
-					DRM_ERROR("Pipe %c FIFO underrun\n",
-						  pipe_name(pipe));
-			}
+			if (pipe_iir & GEN8_PIPE_FIFO_UNDERRUN)
+				intel_cpu_fifo_underrun_irq_handler(dev_priv,
+								    pipe);
 
-			if (pipe_iir & GEN8_DE_PIPE_IRQ_FAULT_ERRORS) {
+
+			if (IS_GEN9(dev))
+				fault_errors = pipe_iir & GEN9_DE_PIPE_IRQ_FAULT_ERRORS;
+			else
+				fault_errors = pipe_iir & GEN8_DE_PIPE_IRQ_FAULT_ERRORS;
+
+			if (fault_errors)
 				DRM_ERROR("Fault errors on pipe %c\n: 0x%08x",
 					  pipe_name(pipe),
 					  pipe_iir & GEN8_DE_PIPE_IRQ_FAULT_ERRORS);
-			}
 		} else
 			DRM_ERROR("The master control interrupt lied (DE PIPE)!\n");
 	}
@@ -2675,6 +2456,9 @@ static void i915_error_work_func(struct work_struct *work)
 		 * simulated reset via debugs, so get an RPM reference.
 		 */
 		intel_runtime_pm_get(dev_priv);
+
+		intel_prepare_reset(dev);
+
 		/*
 		 * All state reset _must_ be completed before we update the
 		 * reset counter, for otherwise waiters might miss the reset
@@ -2683,7 +2467,7 @@ static void i915_error_work_func(struct work_struct *work)
 		 */
 		ret = i915_reset(dev);
 
-		intel_display_handle_reset(dev);
+		intel_finish_reset(dev);
 
 		intel_runtime_pm_put(dev_priv);
 
@@ -2763,7 +2547,7 @@ static void i915_report_and_clear_eir(struct drm_device *dev)
 
 	if (eir & I915_ERROR_MEMORY_REFRESH) {
 		pr_err("memory refresh error:\n");
-		for_each_pipe(pipe)
+		for_each_pipe(dev_priv, pipe)
 			pr_err("pipe %c stat: 0x%08x\n",
 			       pipe_name(pipe), I915_READ(PIPESTAT(pipe)));
 		/* pipestat has already been acked */
@@ -2860,52 +2644,6 @@ void i915_handle_error(struct drm_device *dev, bool wedged,
 	schedule_work(&dev_priv->gpu_error.work);
 }
 
-static void __always_unused i915_pageflip_stall_check(struct drm_device *dev, int pipe)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pipe];
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct drm_i915_gem_object *obj;
-	struct intel_unpin_work *work;
-	unsigned long flags;
-	bool stall_detected;
-
-	/* Ignore early vblank irqs */
-	if (intel_crtc == NULL)
-		return;
-
-	spin_lock_irqsave(&dev->event_lock, flags);
-	work = intel_crtc->unpin_work;
-
-	if (work == NULL ||
-	    atomic_read(&work->pending) >= INTEL_FLIP_COMPLETE ||
-	    !work->enable_stall_check) {
-		/* Either the pending flip IRQ arrived, or we're too early. Don't check */
-		spin_unlock_irqrestore(&dev->event_lock, flags);
-		return;
-	}
-
-	/* Potential stall - if we see that the flip has happened, assume a missed interrupt */
-	obj = work->pending_flip_obj;
-	if (INTEL_INFO(dev)->gen >= 4) {
-		int dspsurf = DSPSURF(intel_crtc->plane);
-		stall_detected = I915_HI_DISPBASE(I915_READ(dspsurf)) ==
-					i915_gem_obj_ggtt_offset(obj);
-	} else {
-		int dspaddr = DSPADDR(intel_crtc->plane);
-		stall_detected = I915_READ(dspaddr) == (i915_gem_obj_ggtt_offset(obj) +
-							crtc->y * crtc->primary->fb->pitches[0] +
-							crtc->x * crtc->primary->fb->bits_per_pixel/8);
-	}
-
-	spin_unlock_irqrestore(&dev->event_lock, flags);
-
-	if (stall_detected) {
-		DRM_DEBUG_DRIVER("Pageflip stall detected\n");
-		intel_prepare_page_flip(dev, intel_crtc->plane);
-	}
-}
-
 /* Called from drm generic code, passed 'crtc' which
  * we use as a pipe index
  */
@@ -3354,10 +3092,15 @@ static void i915_hangcheck_elapsed(unsigned long data)
 void i915_queue_hangcheck(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct timer_list *timer = &dev_priv->gpu_error.hangcheck_timer;
+
 	if (!i915.enable_hangcheck)
 		return;
 
-	mod_timer(&dev_priv->gpu_error.hangcheck_timer,
+	/* Don't continually defer the hangcheck, but make sure it is active */
+	if (timer_pending(timer))
+		return;
+	mod_timer(timer,
 		  round_jiffies_up(jiffies + DRM_I915_HANGCHECK_JIFFIES));
 }
 
@@ -3420,10 +3163,22 @@ static void ironlake_irq_reset(struct drm_device *dev)
 	ibx_irq_reset(dev);
 }
 
+static void vlv_display_irq_reset(struct drm_i915_private *dev_priv)
+{
+	enum pipe pipe;
+
+	I915_WRITE(PORT_HOTPLUG_EN, 0);
+	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
+
+	for_each_pipe(dev_priv, pipe)
+		I915_WRITE(PIPESTAT(pipe), 0xffff);
+
+	GEN5_IRQ_RESET(VLV_);
+}
+
 static void valleyview_irq_preinstall(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	int pipe;
 
 	/* VLV magic */
 	I915_WRITE(VLV_IMR, 0);
@@ -3431,22 +3186,11 @@ static void valleyview_irq_preinstall(struct drm_device *dev)
 	I915_WRITE(RING_IMR(GEN6_BSD_RING_BASE), 0);
 	I915_WRITE(RING_IMR(BLT_RING_BASE), 0);
 
-	/* and GT */
-	I915_WRITE(GTIIR, I915_READ(GTIIR));
-	I915_WRITE(GTIIR, I915_READ(GTIIR));
-
 	gen5_gt_irq_reset(dev);
 
-	I915_WRITE(DPINVGTT, 0xff);
+	I915_WRITE(DPINVGTT, DPINVGTT_STATUS_MASK);
 
-	I915_WRITE(PORT_HOTPLUG_EN, 0);
-	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
-	for_each_pipe(pipe)
-		I915_WRITE(PIPESTAT(pipe), 0xffff);
-	I915_WRITE(VLV_IIR, 0xffffffff);
-	I915_WRITE(VLV_IMR, 0xffffffff);
-	I915_WRITE(VLV_IER, 0x0);
-	POSTING_READ(VLV_IER);
+	vlv_display_irq_reset(dev_priv);
 }
 
 static void gen8_gt_irq_reset(struct drm_i915_private *dev_priv)
@@ -3467,9 +3211,9 @@ static void gen8_irq_reset(struct drm_device *dev)
 
 	gen8_gt_irq_reset(dev_priv);
 
-	for_each_pipe(pipe)
-		if (intel_display_power_enabled(dev_priv,
-						POWER_DOMAIN_PIPE(pipe)))
+	for_each_pipe(dev_priv, pipe)
+		if (intel_display_power_is_enabled(dev_priv,
+						   POWER_DOMAIN_PIPE(pipe)))
 			GEN8_IRQ_RESET_NDX(DE_PIPE, pipe);
 
 	GEN5_IRQ_RESET(GEN8_DE_PORT_);
@@ -3481,20 +3225,19 @@ static void gen8_irq_reset(struct drm_device *dev)
 
 void gen8_irq_power_well_post_enable(struct drm_i915_private *dev_priv)
 {
-	unsigned long irqflags;
+	uint32_t extra_ier = GEN8_PIPE_VBLANK | GEN8_PIPE_FIFO_UNDERRUN;
 
-	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+	spin_lock_irq(&dev_priv->irq_lock);
 	GEN8_IRQ_INIT_NDX(DE_PIPE, PIPE_B, dev_priv->de_irq_mask[PIPE_B],
-			  ~dev_priv->de_irq_mask[PIPE_B]);
+			  ~dev_priv->de_irq_mask[PIPE_B] | extra_ier);
 	GEN8_IRQ_INIT_NDX(DE_PIPE, PIPE_C, dev_priv->de_irq_mask[PIPE_C],
-			  ~dev_priv->de_irq_mask[PIPE_C]);
-	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+			  ~dev_priv->de_irq_mask[PIPE_C] | extra_ier);
+	spin_unlock_irq(&dev_priv->irq_lock);
 }
 
 static void cherryview_irq_preinstall(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	int pipe;
 
 	I915_WRITE(GEN8_MASTER_IRQ, 0);
 	POSTING_READ(GEN8_MASTER_IRQ);
@@ -3503,37 +3246,25 @@ static void cherryview_irq_preinstall(struct drm_device *dev)
 
 	GEN5_IRQ_RESET(GEN8_PCU_);
 
-	POSTING_READ(GEN8_PCU_IIR);
-
 	I915_WRITE(DPINVGTT, DPINVGTT_STATUS_MASK_CHV);
 
-	I915_WRITE(PORT_HOTPLUG_EN, 0);
-	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
-
-	for_each_pipe(pipe)
-		I915_WRITE(PIPESTAT(pipe), 0xffff);
-
-	I915_WRITE(VLV_IMR, 0xffffffff);
-	I915_WRITE(VLV_IER, 0x0);
-	I915_WRITE(VLV_IIR, 0xffffffff);
-	POSTING_READ(VLV_IIR);
+	vlv_display_irq_reset(dev_priv);
 }
 
 static void ibx_hpd_irq_setup(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_mode_config *mode_config = &dev->mode_config;
 	struct intel_encoder *intel_encoder;
 	u32 hotplug_irqs, hotplug, enabled_irqs = 0;
 
 	if (HAS_PCH_IBX(dev)) {
 		hotplug_irqs = SDE_HOTPLUG_MASK;
-		list_for_each_entry(intel_encoder, &mode_config->encoder_list, base.head)
+		for_each_intel_encoder(dev, intel_encoder)
 			if (dev_priv->hpd_stats[intel_encoder->hpd_pin].hpd_mark == HPD_ENABLED)
 				enabled_irqs |= hpd_ibx[intel_encoder->hpd_pin];
 	} else {
 		hotplug_irqs = SDE_HOTPLUG_MASK_CPT;
-		list_for_each_entry(intel_encoder, &mode_config->encoder_list, base.head)
+		for_each_intel_encoder(dev, intel_encoder)
 			if (dev_priv->hpd_stats[intel_encoder->hpd_pin].hpd_mark == HPD_ENABLED)
 				enabled_irqs |= hpd_cpt[intel_encoder->hpd_pin];
 	}
@@ -3596,8 +3327,10 @@ static void gen5_gt_irq_postinstall(struct drm_device *dev)
 	GEN5_IRQ_INIT(GT, dev_priv->gt_irq_mask, gt_irqs);
 
 	if (INTEL_INFO(dev)->gen >= 6) {
-		pm_irqs |= dev_priv->pm_rps_events;
-
+		/*
+		 * RPS interrupts will get enabled/disabled on demand when RPS
+		 * itself is enabled/disabled.
+		 */
 		if (HAS_VEBOX(dev))
 			pm_irqs |= PM_VEBOX_USER_INTERRUPT;
 
@@ -3608,7 +3341,6 @@ static void gen5_gt_irq_postinstall(struct drm_device *dev)
 
 static int ironlake_irq_postinstall(struct drm_device *dev)
 {
-	unsigned long irqflags;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 display_mask, extra_mask;
 
@@ -3647,9 +3379,9 @@ static int ironlake_irq_postinstall(struct drm_device *dev)
 		 * spinlocking not required here for correctness since interrupt
 		 * setup is guaranteed to run in single-threaded context. But we
 		 * need it to make the assert_spin_locked happy. */
-		spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+		spin_lock_irq(&dev_priv->irq_lock);
 		ironlake_enable_display_irq(dev_priv, DE_PCU_EVENT);
-		spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+		spin_unlock_irq(&dev_priv->irq_lock);
 	}
 
 	return 0;
@@ -3659,45 +3391,51 @@ static void valleyview_display_irqs_install(struct drm_i915_private *dev_priv)
 {
 	u32 pipestat_mask;
 	u32 iir_mask;
+	enum pipe pipe;
 
 	pipestat_mask = PIPESTAT_INT_STATUS_MASK |
 			PIPE_FIFO_UNDERRUN_STATUS;
 
-	I915_WRITE(PIPESTAT(PIPE_A), pipestat_mask);
-	I915_WRITE(PIPESTAT(PIPE_B), pipestat_mask);
+	for_each_pipe(dev_priv, pipe)
+		I915_WRITE(PIPESTAT(pipe), pipestat_mask);
 	POSTING_READ(PIPESTAT(PIPE_A));
 
 	pipestat_mask = PLANE_FLIP_DONE_INT_STATUS_VLV |
 			PIPE_CRC_DONE_INTERRUPT_STATUS;
 
-	i915_enable_pipestat(dev_priv, PIPE_A, pipestat_mask |
-					       PIPE_GMBUS_INTERRUPT_STATUS);
-	i915_enable_pipestat(dev_priv, PIPE_B, pipestat_mask);
+	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_GMBUS_INTERRUPT_STATUS);
+	for_each_pipe(dev_priv, pipe)
+		      i915_enable_pipestat(dev_priv, pipe, pipestat_mask);
 
 	iir_mask = I915_DISPLAY_PORT_INTERRUPT |
 		   I915_DISPLAY_PIPE_A_EVENT_INTERRUPT |
 		   I915_DISPLAY_PIPE_B_EVENT_INTERRUPT;
+	if (IS_CHERRYVIEW(dev_priv))
+		iir_mask |= I915_DISPLAY_PIPE_C_EVENT_INTERRUPT;
 	dev_priv->irq_mask &= ~iir_mask;
 
 	I915_WRITE(VLV_IIR, iir_mask);
 	I915_WRITE(VLV_IIR, iir_mask);
-	I915_WRITE(VLV_IMR, dev_priv->irq_mask);
 	I915_WRITE(VLV_IER, ~dev_priv->irq_mask);
-	POSTING_READ(VLV_IER);
+	I915_WRITE(VLV_IMR, dev_priv->irq_mask);
+	POSTING_READ(VLV_IMR);
 }
 
 static void valleyview_display_irqs_uninstall(struct drm_i915_private *dev_priv)
 {
 	u32 pipestat_mask;
 	u32 iir_mask;
+	enum pipe pipe;
 
 	iir_mask = I915_DISPLAY_PORT_INTERRUPT |
 		   I915_DISPLAY_PIPE_A_EVENT_INTERRUPT |
 		   I915_DISPLAY_PIPE_B_EVENT_INTERRUPT;
+	if (IS_CHERRYVIEW(dev_priv))
+		iir_mask |= I915_DISPLAY_PIPE_C_EVENT_INTERRUPT;
 
 	dev_priv->irq_mask |= iir_mask;
-	I915_WRITE(VLV_IER, ~dev_priv->irq_mask);
 	I915_WRITE(VLV_IMR, dev_priv->irq_mask);
+	I915_WRITE(VLV_IER, ~dev_priv->irq_mask);
 	I915_WRITE(VLV_IIR, iir_mask);
 	I915_WRITE(VLV_IIR, iir_mask);
 	POSTING_READ(VLV_IIR);
@@ -3705,14 +3443,15 @@ static void valleyview_display_irqs_uninstall(struct drm_i915_private *dev_priv)
 	pipestat_mask = PLANE_FLIP_DONE_INT_STATUS_VLV |
 			PIPE_CRC_DONE_INTERRUPT_STATUS;
 
-	i915_disable_pipestat(dev_priv, PIPE_A, pipestat_mask |
-					        PIPE_GMBUS_INTERRUPT_STATUS);
-	i915_disable_pipestat(dev_priv, PIPE_B, pipestat_mask);
+	i915_disable_pipestat(dev_priv, PIPE_A, PIPE_GMBUS_INTERRUPT_STATUS);
+	for_each_pipe(dev_priv, pipe)
+		i915_disable_pipestat(dev_priv, pipe, pipestat_mask);
 
 	pipestat_mask = PIPESTAT_INT_STATUS_MASK |
 			PIPE_FIFO_UNDERRUN_STATUS;
-	I915_WRITE(PIPESTAT(PIPE_A), pipestat_mask);
-	I915_WRITE(PIPESTAT(PIPE_B), pipestat_mask);
+
+	for_each_pipe(dev_priv, pipe)
+		I915_WRITE(PIPESTAT(pipe), pipestat_mask);
 	POSTING_READ(PIPESTAT(PIPE_A));
 }
 
@@ -3725,7 +3464,7 @@ void valleyview_enable_display_irqs(struct drm_i915_private *dev_priv)
 
 	dev_priv->display_irqs_enabled = true;
 
-	if (dev_priv->dev->irq_enabled)
+	if (intel_irqs_enabled(dev_priv))
 		valleyview_display_irqs_install(dev_priv);
 }
 
@@ -3738,34 +3477,36 @@ void valleyview_disable_display_irqs(struct drm_i915_private *dev_priv)
 
 	dev_priv->display_irqs_enabled = false;
 
-	if (dev_priv->dev->irq_enabled)
+	if (intel_irqs_enabled(dev_priv))
 		valleyview_display_irqs_uninstall(dev_priv);
 }
 
-static int valleyview_irq_postinstall(struct drm_device *dev)
+static void vlv_display_irq_postinstall(struct drm_i915_private *dev_priv)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	unsigned long irqflags;
-
 	dev_priv->irq_mask = ~0;
 
 	I915_WRITE(PORT_HOTPLUG_EN, 0);
 	POSTING_READ(PORT_HOTPLUG_EN);
 
-	I915_WRITE(VLV_IMR, dev_priv->irq_mask);
-	I915_WRITE(VLV_IER, ~dev_priv->irq_mask);
 	I915_WRITE(VLV_IIR, 0xffffffff);
-	POSTING_READ(VLV_IER);
+	I915_WRITE(VLV_IIR, 0xffffffff);
+	I915_WRITE(VLV_IER, ~dev_priv->irq_mask);
+	I915_WRITE(VLV_IMR, dev_priv->irq_mask);
+	POSTING_READ(VLV_IMR);
 
 	/* Interrupt setup is already guaranteed to be single-threaded, this is
 	 * just to make the assert_spin_locked check happy. */
-	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+	spin_lock_irq(&dev_priv->irq_lock);
 	if (dev_priv->display_irqs_enabled)
 		valleyview_display_irqs_install(dev_priv);
-	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+	spin_unlock_irq(&dev_priv->irq_lock);
+}
 
-	I915_WRITE(VLV_IIR, 0xffffffff);
-	I915_WRITE(VLV_IIR, 0xffffffff);
+static int valleyview_irq_postinstall(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	vlv_display_irq_postinstall(dev_priv);
 
 	gen5_gt_irq_postinstall(dev);
 
@@ -3782,46 +3523,64 @@ static int valleyview_irq_postinstall(struct drm_device *dev)
 
 static void gen8_gt_irq_postinstall(struct drm_i915_private *dev_priv)
 {
-	int i;
-
 	/* These are interrupts we'll toggle with the ring mask register */
 	uint32_t gt_interrupts[] = {
 		GT_RENDER_USER_INTERRUPT << GEN8_RCS_IRQ_SHIFT |
+			GT_CONTEXT_SWITCH_INTERRUPT << GEN8_RCS_IRQ_SHIFT |
 			GT_RENDER_L3_PARITY_ERROR_INTERRUPT |
-			GT_RENDER_USER_INTERRUPT << GEN8_BCS_IRQ_SHIFT,
+			GT_RENDER_USER_INTERRUPT << GEN8_BCS_IRQ_SHIFT |
+			GT_CONTEXT_SWITCH_INTERRUPT << GEN8_BCS_IRQ_SHIFT,
 		GT_RENDER_USER_INTERRUPT << GEN8_VCS1_IRQ_SHIFT |
-			GT_RENDER_USER_INTERRUPT << GEN8_VCS2_IRQ_SHIFT,
+			GT_CONTEXT_SWITCH_INTERRUPT << GEN8_VCS1_IRQ_SHIFT |
+			GT_RENDER_USER_INTERRUPT << GEN8_VCS2_IRQ_SHIFT |
+			GT_CONTEXT_SWITCH_INTERRUPT << GEN8_VCS2_IRQ_SHIFT,
 		0,
-		GT_RENDER_USER_INTERRUPT << GEN8_VECS_IRQ_SHIFT
+		GT_RENDER_USER_INTERRUPT << GEN8_VECS_IRQ_SHIFT |
+			GT_CONTEXT_SWITCH_INTERRUPT << GEN8_VECS_IRQ_SHIFT
 		};
 
-	for (i = 0; i < ARRAY_SIZE(gt_interrupts); i++)
-		GEN8_IRQ_INIT_NDX(GT, i, ~gt_interrupts[i], gt_interrupts[i]);
-
 	dev_priv->pm_irq_mask = 0xffffffff;
+	GEN8_IRQ_INIT_NDX(GT, 0, ~gt_interrupts[0], gt_interrupts[0]);
+	GEN8_IRQ_INIT_NDX(GT, 1, ~gt_interrupts[1], gt_interrupts[1]);
+	/*
+	 * RPS interrupts will get enabled/disabled on demand when RPS itself
+	 * is enabled/disabled.
+	 */
+	GEN8_IRQ_INIT_NDX(GT, 2, dev_priv->pm_irq_mask, 0);
+	GEN8_IRQ_INIT_NDX(GT, 3, ~gt_interrupts[3], gt_interrupts[3]);
 }
 
 static void gen8_de_irq_postinstall(struct drm_i915_private *dev_priv)
 {
-	struct drm_device *dev = dev_priv->dev;
-	uint32_t de_pipe_masked = GEN8_PIPE_PRIMARY_FLIP_DONE |
-		GEN8_PIPE_CDCLK_CRC_DONE |
-		GEN8_DE_PIPE_IRQ_FAULT_ERRORS;
-	uint32_t de_pipe_enables = de_pipe_masked | GEN8_PIPE_VBLANK |
-		GEN8_PIPE_FIFO_UNDERRUN;
+	uint32_t de_pipe_masked = GEN8_PIPE_CDCLK_CRC_DONE;
+	uint32_t de_pipe_enables;
 	int pipe;
+	u32 aux_en = GEN8_AUX_CHANNEL_A;
+
+	if (IS_GEN9(dev_priv)) {
+		de_pipe_masked |= GEN9_PIPE_PLANE1_FLIP_DONE |
+				  GEN9_DE_PIPE_IRQ_FAULT_ERRORS;
+		aux_en |= GEN9_AUX_CHANNEL_B | GEN9_AUX_CHANNEL_C |
+			GEN9_AUX_CHANNEL_D;
+	} else
+		de_pipe_masked |= GEN8_PIPE_PRIMARY_FLIP_DONE |
+				  GEN8_DE_PIPE_IRQ_FAULT_ERRORS;
+
+	de_pipe_enables = de_pipe_masked | GEN8_PIPE_VBLANK |
+					   GEN8_PIPE_FIFO_UNDERRUN;
+
 	dev_priv->de_irq_mask[PIPE_A] = ~de_pipe_masked;
 	dev_priv->de_irq_mask[PIPE_B] = ~de_pipe_masked;
 	dev_priv->de_irq_mask[PIPE_C] = ~de_pipe_masked;
 
-	for_each_pipe(pipe)
-		if (intel_display_power_enabled(dev_priv,
+	for_each_pipe(dev_priv, pipe)
+		if (intel_display_power_is_enabled(dev_priv,
 				POWER_DOMAIN_PIPE(pipe)))
 			GEN8_IRQ_INIT_NDX(DE_PIPE, pipe,
 					  dev_priv->de_irq_mask[pipe],
 					  de_pipe_enables);
 
-	GEN5_IRQ_INIT(GEN8_DE_PORT_, ~GEN8_AUX_CHANNEL_A, GEN8_AUX_CHANNEL_A);
+	GEN5_IRQ_INIT(GEN8_DE_PORT_, ~aux_en, aux_en);
 }
 
 static int gen8_irq_postinstall(struct drm_device *dev)
@@ -3844,33 +3603,8 @@ static int gen8_irq_postinstall(struct drm_device *dev)
 static int cherryview_irq_postinstall(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 enable_mask = I915_DISPLAY_PORT_INTERRUPT |
-		I915_DISPLAY_PIPE_A_EVENT_INTERRUPT |
-		I915_DISPLAY_PIPE_B_EVENT_INTERRUPT |
-		I915_DISPLAY_PIPE_C_EVENT_INTERRUPT;
-	u32 pipestat_enable = PLANE_FLIP_DONE_INT_STATUS_VLV |
-		PIPE_CRC_DONE_INTERRUPT_STATUS;
-	unsigned long irqflags;
-	int pipe;
-
-	/*
-	 * Leave vblank interrupts masked initially.  enable/disable will
-	 * toggle them based on usage.
-	 */
-	dev_priv->irq_mask = ~enable_mask;
-
-	for_each_pipe(pipe)
-		I915_WRITE(PIPESTAT(pipe), 0xffff);
 
-	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
-	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_GMBUS_INTERRUPT_STATUS);
-	for_each_pipe(pipe)
-		i915_enable_pipestat(dev_priv, pipe, pipestat_enable);
-	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
-
-	I915_WRITE(VLV_IIR, 0xffffffff);
-	I915_WRITE(VLV_IMR, dev_priv->irq_mask);
-	I915_WRITE(VLV_IER, enable_mask);
+	vlv_display_irq_postinstall(dev_priv);
 
 	gen8_gt_irq_postinstall(dev_priv);
 
@@ -3890,41 +3624,39 @@ static void gen8_irq_uninstall(struct drm_device *dev)
 	gen8_irq_reset(dev);
 }
 
+static void vlv_display_irq_uninstall(struct drm_i915_private *dev_priv)
+{
+	/* Interrupt setup is already guaranteed to be single-threaded, this is
+	 * just to make the assert_spin_locked check happy. */
+	spin_lock_irq(&dev_priv->irq_lock);
+	if (dev_priv->display_irqs_enabled)
+		valleyview_display_irqs_uninstall(dev_priv);
+	spin_unlock_irq(&dev_priv->irq_lock);
+
+	vlv_display_irq_reset(dev_priv);
+
+	dev_priv->irq_mask = ~0;
+}
+
 static void valleyview_irq_uninstall(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	unsigned long irqflags;
-	int pipe;
 
 	if (!dev_priv)
 		return;
 
 	I915_WRITE(VLV_MASTER_IER, 0);
 
-	for_each_pipe(pipe)
-		I915_WRITE(PIPESTAT(pipe), 0xffff);
+	gen5_gt_irq_reset(dev);
 
 	I915_WRITE(HWSTAM, 0xffffffff);
-	I915_WRITE(PORT_HOTPLUG_EN, 0);
-	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
-
-	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
-	if (dev_priv->display_irqs_enabled)
-		valleyview_display_irqs_uninstall(dev_priv);
-	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 
-	dev_priv->irq_mask = 0;
-
-	I915_WRITE(VLV_IIR, 0xffffffff);
-	I915_WRITE(VLV_IMR, 0xffffffff);
-	I915_WRITE(VLV_IER, 0x0);
-	POSTING_READ(VLV_IER);
+	vlv_display_irq_uninstall(dev_priv);
 }
 
 static void cherryview_irq_uninstall(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	int pipe;
 
 	if (!dev_priv)
 		return;
@@ -3932,44 +3664,11 @@ static void cherryview_irq_uninstall(struct drm_device *dev)
 	I915_WRITE(GEN8_MASTER_IRQ, 0);
 	POSTING_READ(GEN8_MASTER_IRQ);
 
-#define GEN8_IRQ_FINI_NDX(type, which)				\
-do {								\
-	I915_WRITE(GEN8_##type##_IMR(which), 0xffffffff);	\
-	I915_WRITE(GEN8_##type##_IER(which), 0);		\
-	I915_WRITE(GEN8_##type##_IIR(which), 0xffffffff);	\
-	POSTING_READ(GEN8_##type##_IIR(which));			\
-	I915_WRITE(GEN8_##type##_IIR(which), 0xffffffff);	\
-} while (0)
-
-#define GEN8_IRQ_FINI(type)				\
-do {							\
-	I915_WRITE(GEN8_##type##_IMR, 0xffffffff);	\
-	I915_WRITE(GEN8_##type##_IER, 0);		\
-	I915_WRITE(GEN8_##type##_IIR, 0xffffffff);	\
-	POSTING_READ(GEN8_##type##_IIR);		\
-	I915_WRITE(GEN8_##type##_IIR, 0xffffffff);	\
-} while (0)
-
-	GEN8_IRQ_FINI_NDX(GT, 0);
-	GEN8_IRQ_FINI_NDX(GT, 1);
-	GEN8_IRQ_FINI_NDX(GT, 2);
-	GEN8_IRQ_FINI_NDX(GT, 3);
-
-	GEN8_IRQ_FINI(PCU);
-
-#undef GEN8_IRQ_FINI
-#undef GEN8_IRQ_FINI_NDX
-
-	I915_WRITE(PORT_HOTPLUG_EN, 0);
-	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
+	gen8_gt_irq_reset(dev_priv);
 
-	for_each_pipe(pipe)
-		I915_WRITE(PIPESTAT(pipe), 0xffff);
+	GEN5_IRQ_RESET(GEN8_PCU_);
 
-	I915_WRITE(VLV_IMR, 0xffffffff);
-	I915_WRITE(VLV_IER, 0x0);
-	I915_WRITE(VLV_IIR, 0xffffffff);
-	POSTING_READ(VLV_IIR);
+	vlv_display_irq_uninstall(dev_priv);
 }
 
 static void ironlake_irq_uninstall(struct drm_device *dev)
@@ -3987,7 +3686,7 @@ static void i8xx_irq_preinstall(struct drm_device * dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe;
 
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		I915_WRITE(PIPESTAT(pipe), 0);
 	I915_WRITE16(IMR, 0xffff);
 	I915_WRITE16(IER, 0x0);
@@ -3997,7 +3696,6 @@ static void i8xx_irq_preinstall(struct drm_device * dev)
 static int i8xx_irq_postinstall(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	unsigned long irqflags;
 
 	I915_WRITE16(EMR,
 		     ~(I915_ERROR_PAGE_TABLE | I915_ERROR_MEMORY_REFRESH));
@@ -4020,10 +3718,10 @@ static int i8xx_irq_postinstall(struct drm_device *dev)
 
 	/* Interrupt setup is already guaranteed to be single-threaded, this is
 	 * just to make the assert_spin_locked check happy. */
-	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+	spin_lock_irq(&dev_priv->irq_lock);
 	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_CRC_DONE_INTERRUPT_STATUS);
 	i915_enable_pipestat(dev_priv, PIPE_B, PIPE_CRC_DONE_INTERRUPT_STATUS);
-	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+	spin_unlock_irq(&dev_priv->irq_lock);
 
 	return 0;
 }
@@ -4041,9 +3739,7 @@ static bool i8xx_handle_vblank(struct drm_device *dev,
 		return false;
 
 	if ((iir & flip_pending) == 0)
-		return false;
-
-	intel_prepare_page_flip(dev, plane);
+		goto check_page_flip;
 
 	/* We detect FlipDone by looking for the change in PendingFlip from '1'
 	 * to '0' on the following vblank, i.e. IIR has the Pendingflip
@@ -4052,11 +3748,15 @@ static bool i8xx_handle_vblank(struct drm_device *dev,
 	 * an interrupt per se, we watch for the change at vblank.
 	 */
 	if (I915_READ16(ISR) & flip_pending)
-		return false;
+		goto check_page_flip;
 
+	intel_prepare_page_flip(dev, plane);
 	intel_finish_page_flip(dev, pipe);
-
 	return true;
+
+check_page_flip:
+	intel_check_page_flip(dev, pipe);
+	return false;
 }
 
 static irqreturn_t i8xx_irq_handler(int irq, void *arg)
@@ -4065,7 +3765,6 @@ static irqreturn_t i8xx_irq_handler(int irq, void *arg)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u16 iir, new_iir;
 	u32 pipe_stats[2];
-	unsigned long irqflags;
 	int pipe;
 	u16 flip_mask =
 		I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT |
@@ -4081,13 +3780,11 @@ static irqreturn_t i8xx_irq_handler(int irq, void *arg)
 		 * It doesn't set the bit in iir again, but it still produces
 		 * interrupts (for non-MSI).
 		 */
-		spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+		spin_lock(&dev_priv->irq_lock);
 		if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT)
-			i915_handle_error(dev, false,
-					  "Command parser error, iir 0x%08x",
-					  iir);
+			DRM_DEBUG("Command parser error, iir 0x%08x\n", iir);
 
-		for_each_pipe(pipe) {
+		for_each_pipe(dev_priv, pipe) {
 			int reg = PIPESTAT(pipe);
 			pipe_stats[pipe] = I915_READ(reg);
 
@@ -4097,17 +3794,15 @@ static irqreturn_t i8xx_irq_handler(int irq, void *arg)
 			if (pipe_stats[pipe] & 0x8000ffff)
 				I915_WRITE(reg, pipe_stats[pipe]);
 		}
-		spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+		spin_unlock(&dev_priv->irq_lock);
 
 		I915_WRITE16(IIR, iir & ~flip_mask);
 		new_iir = I915_READ16(IIR); /* Flush posted writes */
 
-		i915_update_dri1_breadcrumb(dev);
-
 		if (iir & I915_USER_INTERRUPT)
 			notify_ring(dev, &dev_priv->ring[RCS]);
 
-		for_each_pipe(pipe) {
+		for_each_pipe(dev_priv, pipe) {
 			int plane = pipe;
 			if (HAS_FBC(dev))
 				plane = !plane;
@@ -4119,9 +3814,9 @@ static irqreturn_t i8xx_irq_handler(int irq, void *arg)
 			if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
 				i9xx_pipe_crc_irq_handler(dev, pipe);
 
-			if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS &&
-			    intel_set_cpu_fifo_underrun_reporting(dev, pipe, false))
-				DRM_ERROR("pipe %c underrun\n", pipe_name(pipe));
+			if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS)
+				intel_cpu_fifo_underrun_irq_handler(dev_priv,
+								    pipe);
 		}
 
 		iir = new_iir;
@@ -4135,7 +3830,7 @@ static void i8xx_irq_uninstall(struct drm_device * dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe;
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		/* Clear enable bits; then clear status bits */
 		I915_WRITE(PIPESTAT(pipe), 0);
 		I915_WRITE(PIPESTAT(pipe), I915_READ(PIPESTAT(pipe)));
@@ -4156,7 +3851,7 @@ static void i915_irq_preinstall(struct drm_device * dev)
 	}
 
 	I915_WRITE16(HWSTAM, 0xeffe);
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		I915_WRITE(PIPESTAT(pipe), 0);
 	I915_WRITE(IMR, 0xffffffff);
 	I915_WRITE(IER, 0x0);
@@ -4167,7 +3862,6 @@ static int i915_irq_postinstall(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 enable_mask;
-	unsigned long irqflags;
 
 	I915_WRITE(EMR, ~(I915_ERROR_PAGE_TABLE | I915_ERROR_MEMORY_REFRESH));
 
@@ -4205,10 +3899,10 @@ static int i915_irq_postinstall(struct drm_device *dev)
 
 	/* Interrupt setup is already guaranteed to be single-threaded, this is
 	 * just to make the assert_spin_locked check happy. */
-	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+	spin_lock_irq(&dev_priv->irq_lock);
 	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_CRC_DONE_INTERRUPT_STATUS);
 	i915_enable_pipestat(dev_priv, PIPE_B, PIPE_CRC_DONE_INTERRUPT_STATUS);
-	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+	spin_unlock_irq(&dev_priv->irq_lock);
 
 	return 0;
 }
@@ -4226,9 +3920,7 @@ static bool i915_handle_vblank(struct drm_device *dev,
 		return false;
 
 	if ((iir & flip_pending) == 0)
-		return false;
-
-	intel_prepare_page_flip(dev, plane);
+		goto check_page_flip;
 
 	/* We detect FlipDone by looking for the change in PendingFlip from '1'
 	 * to '0' on the following vblank, i.e. IIR has the Pendingflip
@@ -4237,11 +3929,15 @@ static bool i915_handle_vblank(struct drm_device *dev,
 	 * an interrupt per se, we watch for the change at vblank.
 	 */
 	if (I915_READ(ISR) & flip_pending)
-		return false;
+		goto check_page_flip;
 
+	intel_prepare_page_flip(dev, plane);
 	intel_finish_page_flip(dev, pipe);
-
 	return true;
+
+check_page_flip:
+	intel_check_page_flip(dev, pipe);
+	return false;
 }
 
 static irqreturn_t i915_irq_handler(int irq, void *arg)
@@ -4249,7 +3945,6 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 	struct drm_device *dev = arg;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 iir, new_iir, pipe_stats[I915_MAX_PIPES];
-	unsigned long irqflags;
 	u32 flip_mask =
 		I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT |
 		I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT;
@@ -4265,13 +3960,11 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 		 * It doesn't set the bit in iir again, but it still produces
 		 * interrupts (for non-MSI).
 		 */
-		spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+		spin_lock(&dev_priv->irq_lock);
 		if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT)
-			i915_handle_error(dev, false,
-					  "Command parser error, iir 0x%08x",
-					  iir);
+			DRM_DEBUG("Command parser error, iir 0x%08x\n", iir);
 
-		for_each_pipe(pipe) {
+		for_each_pipe(dev_priv, pipe) {
 			int reg = PIPESTAT(pipe);
 			pipe_stats[pipe] = I915_READ(reg);
 
@@ -4281,7 +3974,7 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 				irq_received = true;
 			}
 		}
-		spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+		spin_unlock(&dev_priv->irq_lock);
 
 		if (!irq_received)
 			break;
@@ -4297,7 +3990,7 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 		if (iir & I915_USER_INTERRUPT)
 			notify_ring(dev, &dev_priv->ring[RCS]);
 
-		for_each_pipe(pipe) {
+		for_each_pipe(dev_priv, pipe) {
 			int plane = pipe;
 			if (HAS_FBC(dev))
 				plane = !plane;
@@ -4312,9 +4005,9 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 			if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
 				i9xx_pipe_crc_irq_handler(dev, pipe);
 
-			if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS &&
-			    intel_set_cpu_fifo_underrun_reporting(dev, pipe, false))
-				DRM_ERROR("pipe %c underrun\n", pipe_name(pipe));
+			if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS)
+				intel_cpu_fifo_underrun_irq_handler(dev_priv,
+								    pipe);
 		}
 
 		if (blc_event || (iir & I915_ASLE_INTERRUPT))
@@ -4339,8 +4032,6 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 		iir = new_iir;
 	} while (iir & ~flip_mask);
 
-	i915_update_dri1_breadcrumb(dev);
-
 	return ret;
 }
 
@@ -4355,7 +4046,7 @@ static void i915_irq_uninstall(struct drm_device * dev)
 	}
 
 	I915_WRITE16(HWSTAM, 0xffff);
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		/* Clear enable bits; then clear status bits */
 		I915_WRITE(PIPESTAT(pipe), 0);
 		I915_WRITE(PIPESTAT(pipe), I915_READ(PIPESTAT(pipe)));
@@ -4375,7 +4066,7 @@ static void i965_irq_preinstall(struct drm_device * dev)
 	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
 
 	I915_WRITE(HWSTAM, 0xeffe);
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		I915_WRITE(PIPESTAT(pipe), 0);
 	I915_WRITE(IMR, 0xffffffff);
 	I915_WRITE(IER, 0x0);
@@ -4387,7 +4078,6 @@ static int i965_irq_postinstall(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 enable_mask;
 	u32 error_mask;
-	unsigned long irqflags;
 
 	/* Unmask the interrupts that we always want on. */
 	dev_priv->irq_mask = ~(I915_ASLE_INTERRUPT |
@@ -4408,11 +4098,11 @@ static int i965_irq_postinstall(struct drm_device *dev)
 
 	/* Interrupt setup is already guaranteed to be single-threaded, this is
 	 * just to make the assert_spin_locked check happy. */
-	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+	spin_lock_irq(&dev_priv->irq_lock);
 	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_GMBUS_INTERRUPT_STATUS);
 	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_CRC_DONE_INTERRUPT_STATUS);
 	i915_enable_pipestat(dev_priv, PIPE_B, PIPE_CRC_DONE_INTERRUPT_STATUS);
-	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+	spin_unlock_irq(&dev_priv->irq_lock);
 
 	/*
 	 * Enable some error detection, note the instruction error mask
@@ -4444,7 +4134,6 @@ static int i965_irq_postinstall(struct drm_device *dev)
 static void i915_hpd_irq_setup(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_mode_config *mode_config = &dev->mode_config;
 	struct intel_encoder *intel_encoder;
 	u32 hotplug_en;
 
@@ -4455,7 +4144,7 @@ static void i915_hpd_irq_setup(struct drm_device *dev)
 		hotplug_en &= ~HOTPLUG_INT_EN_MASK;
 		/* Note HDMI and DP share hotplug bits */
 		/* enable bits are the same for all generations */
-		list_for_each_entry(intel_encoder, &mode_config->encoder_list, base.head)
+		for_each_intel_encoder(dev, intel_encoder)
 			if (dev_priv->hpd_stats[intel_encoder->hpd_pin].hpd_mark == HPD_ENABLED)
 				hotplug_en |= hpd_mask_i915[intel_encoder->hpd_pin];
 		/* Programming the CRT detection parameters tends
@@ -4478,7 +4167,6 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 iir, new_iir;
 	u32 pipe_stats[I915_MAX_PIPES];
-	unsigned long irqflags;
 	int ret = IRQ_NONE, pipe;
 	u32 flip_mask =
 		I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT |
@@ -4495,13 +4183,11 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 		 * It doesn't set the bit in iir again, but it still produces
 		 * interrupts (for non-MSI).
 		 */
-		spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+		spin_lock(&dev_priv->irq_lock);
 		if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT)
-			i915_handle_error(dev, false,
-					  "Command parser error, iir 0x%08x",
-					  iir);
+			DRM_DEBUG("Command parser error, iir 0x%08x\n", iir);
 
-		for_each_pipe(pipe) {
+		for_each_pipe(dev_priv, pipe) {
 			int reg = PIPESTAT(pipe);
 			pipe_stats[pipe] = I915_READ(reg);
 
@@ -4513,7 +4199,7 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 				irq_received = true;
 			}
 		}
-		spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+		spin_unlock(&dev_priv->irq_lock);
 
 		if (!irq_received)
 			break;
@@ -4532,7 +4218,7 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 		if (iir & I915_BSD_USER_INTERRUPT)
 			notify_ring(dev, &dev_priv->ring[VCS]);
 
-		for_each_pipe(pipe) {
+		for_each_pipe(dev_priv, pipe) {
 			if (pipe_stats[pipe] & PIPE_START_VBLANK_INTERRUPT_STATUS &&
 			    i915_handle_vblank(dev, pipe, pipe, iir))
 				flip_mask &= ~DISPLAY_PLANE_FLIP_PENDING(pipe);
@@ -4543,9 +4229,8 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 			if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
 				i9xx_pipe_crc_irq_handler(dev, pipe);
 
-			if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS &&
-			    intel_set_cpu_fifo_underrun_reporting(dev, pipe, false))
-				DRM_ERROR("pipe %c underrun\n", pipe_name(pipe));
+			if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS)
+				intel_cpu_fifo_underrun_irq_handler(dev_priv, pipe);
 		}
 
 		if (blc_event || (iir & I915_ASLE_INTERRUPT))
@@ -4572,8 +4257,6 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 		iir = new_iir;
 	}
 
-	i915_update_dri1_breadcrumb(dev);
-
 	return ret;
 }
 
@@ -4589,30 +4272,29 @@ static void i965_irq_uninstall(struct drm_device * dev)
 	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
 
 	I915_WRITE(HWSTAM, 0xffffffff);
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		I915_WRITE(PIPESTAT(pipe), 0);
 	I915_WRITE(IMR, 0xffffffff);
 	I915_WRITE(IER, 0x0);
 
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		I915_WRITE(PIPESTAT(pipe),
 			   I915_READ(PIPESTAT(pipe)) & 0x8000ffff);
 	I915_WRITE(IIR, I915_READ(IIR));
 }
 
-static void intel_hpd_irq_reenable(struct work_struct *work)
+static void intel_hpd_irq_reenable_work(struct work_struct *work)
 {
 	struct drm_i915_private *dev_priv =
 		container_of(work, typeof(*dev_priv),
 			     hotplug_reenable_work.work);
 	struct drm_device *dev = dev_priv->dev;
 	struct drm_mode_config *mode_config = &dev->mode_config;
-	unsigned long irqflags;
 	int i;
 
 	intel_runtime_pm_get(dev_priv);
 
-	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+	spin_lock_irq(&dev_priv->irq_lock);
 	for (i = (HPD_NONE + 1); i < HPD_NUM_PINS; i++) {
 		struct drm_connector *connector;
 
@@ -4636,14 +4318,21 @@ static void intel_hpd_irq_reenable(struct work_struct *work)
 	}
 	if (dev_priv->display.hpd_irq_setup)
 		dev_priv->display.hpd_irq_setup(dev);
-	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+	spin_unlock_irq(&dev_priv->irq_lock);
 
 	intel_runtime_pm_put(dev_priv);
 }
 
-void intel_irq_init(struct drm_device *dev)
+/**
+ * intel_irq_init - initializes irq support
+ * @dev_priv: i915 device instance
+ *
+ * This function initializes all the irq support including work items, timers
+ * and all the vtables. It does not setup the interrupt itself though.
+ */
+void intel_irq_init(struct drm_i915_private *dev_priv)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_device *dev = dev_priv->dev;
 
 	INIT_WORK(&dev_priv->hotplug_work, i915_hotplug_work_func);
 	INIT_WORK(&dev_priv->dig_port_work, i915_digport_work_func);
@@ -4652,8 +4341,8 @@ void intel_irq_init(struct drm_device *dev)
 	INIT_WORK(&dev_priv->l3_parity.error_work, ivybridge_parity_work);
 
 	/* Let's track the enabled rps events */
-	if (IS_VALLEYVIEW(dev))
-		/* WaGsvRC0ResidenncyMethod:VLV */
+	if (IS_VALLEYVIEW(dev_priv) && !IS_CHERRYVIEW(dev_priv))
+		/* WaGsvRC0ResidencyMethod:vlv */
 		dev_priv->pm_rps_events = GEN6_PM_RP_UP_EI_EXPIRED;
 	else
 		dev_priv->pm_rps_events = GEN6_PM_RPS_EVENTS;
@@ -4662,17 +4351,14 @@ void intel_irq_init(struct drm_device *dev)
 		    i915_hangcheck_elapsed,
 		    (unsigned long) dev);
 	INIT_DELAYED_WORK(&dev_priv->hotplug_reenable_work,
-			  intel_hpd_irq_reenable);
+			  intel_hpd_irq_reenable_work);
 
 	pm_qos_add_request(&dev_priv->pm_qos, PM_QOS_CPU_DMA_LATENCY, PM_QOS_DEFAULT_VALUE);
 
-	/* Haven't installed the IRQ handler yet */
-	dev_priv->pm._irqs_disabled = true;
-
-	if (IS_GEN2(dev)) {
+	if (IS_GEN2(dev_priv)) {
 		dev->max_vblank_count = 0;
 		dev->driver->get_vblank_counter = i8xx_get_vblank_counter;
-	} else if (IS_G4X(dev) || INTEL_INFO(dev)->gen >= 5) {
+	} else if (IS_G4X(dev_priv) || INTEL_INFO(dev_priv)->gen >= 5) {
 		dev->max_vblank_count = 0xffffffff; /* full 32 bit counter */
 		dev->driver->get_vblank_counter = gm45_get_vblank_counter;
 	} else {
@@ -4680,12 +4366,20 @@ void intel_irq_init(struct drm_device *dev)
 		dev->max_vblank_count = 0xffffff; /* only 24 bits of frame count */
 	}
 
+	/*
+	 * Opt out of the vblank disable timer on everything except gen2.
+	 * Gen2 doesn't have a hardware frame counter and so depends on
+	 * vblank interrupts to produce sane vblank seuquence numbers.
+	 */
+	if (!IS_GEN2(dev_priv))
+		dev->vblank_disable_immediate = true;
+
 	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
 		dev->driver->get_vblank_timestamp = i915_get_vblank_timestamp;
 		dev->driver->get_scanout_position = i915_get_crtc_scanoutpos;
 	}
 
-	if (IS_CHERRYVIEW(dev)) {
+	if (IS_CHERRYVIEW(dev_priv)) {
 		dev->driver->irq_handler = cherryview_irq_handler;
 		dev->driver->irq_preinstall = cherryview_irq_preinstall;
 		dev->driver->irq_postinstall = cherryview_irq_postinstall;
@@ -4693,7 +4387,7 @@ void intel_irq_init(struct drm_device *dev)
 		dev->driver->enable_vblank = valleyview_enable_vblank;
 		dev->driver->disable_vblank = valleyview_disable_vblank;
 		dev_priv->display.hpd_irq_setup = i915_hpd_irq_setup;
-	} else if (IS_VALLEYVIEW(dev)) {
+	} else if (IS_VALLEYVIEW(dev_priv)) {
 		dev->driver->irq_handler = valleyview_irq_handler;
 		dev->driver->irq_preinstall = valleyview_irq_preinstall;
 		dev->driver->irq_postinstall = valleyview_irq_postinstall;
@@ -4701,7 +4395,7 @@ void intel_irq_init(struct drm_device *dev)
 		dev->driver->enable_vblank = valleyview_enable_vblank;
 		dev->driver->disable_vblank = valleyview_disable_vblank;
 		dev_priv->display.hpd_irq_setup = i915_hpd_irq_setup;
-	} else if (IS_GEN8(dev)) {
+	} else if (INTEL_INFO(dev_priv)->gen >= 8) {
 		dev->driver->irq_handler = gen8_irq_handler;
 		dev->driver->irq_preinstall = gen8_irq_reset;
 		dev->driver->irq_postinstall = gen8_irq_postinstall;
@@ -4718,12 +4412,12 @@ void intel_irq_init(struct drm_device *dev)
 		dev->driver->disable_vblank = ironlake_disable_vblank;
 		dev_priv->display.hpd_irq_setup = ibx_hpd_irq_setup;
 	} else {
-		if (INTEL_INFO(dev)->gen == 2) {
+		if (INTEL_INFO(dev_priv)->gen == 2) {
 			dev->driver->irq_preinstall = i8xx_irq_preinstall;
 			dev->driver->irq_postinstall = i8xx_irq_postinstall;
 			dev->driver->irq_handler = i8xx_irq_handler;
 			dev->driver->irq_uninstall = i8xx_irq_uninstall;
-		} else if (INTEL_INFO(dev)->gen == 3) {
+		} else if (INTEL_INFO(dev_priv)->gen == 3) {
 			dev->driver->irq_preinstall = i915_irq_preinstall;
 			dev->driver->irq_postinstall = i915_irq_postinstall;
 			dev->driver->irq_uninstall = i915_irq_uninstall;
@@ -4741,12 +4435,23 @@ void intel_irq_init(struct drm_device *dev)
 	}
 }
 
-void intel_hpd_init(struct drm_device *dev)
+/**
+ * intel_hpd_init - initializes and enables hpd support
+ * @dev_priv: i915 device instance
+ *
+ * This function enables the hotplug support. It requires that interrupts have
+ * already been enabled with intel_irq_init_hw(). From this point on hotplug and
+ * poll request can run concurrently to other code, so locking rules must be
+ * obeyed.
+ *
+ * This is a separate step from interrupt enabling to simplify the locking rules
+ * in the driver load and resume code.
+ */
+void intel_hpd_init(struct drm_i915_private *dev_priv)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_device *dev = dev_priv->dev;
 	struct drm_mode_config *mode_config = &dev->mode_config;
 	struct drm_connector *connector;
-	unsigned long irqflags;
 	int i;
 
 	for (i = 1; i < HPD_NUM_PINS; i++) {
@@ -4764,27 +4469,72 @@ void intel_hpd_init(struct drm_device *dev)
 
 	/* Interrupt setup is already guaranteed to be single-threaded, this is
 	 * just to make the assert_spin_locked checks happy. */
-	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+	spin_lock_irq(&dev_priv->irq_lock);
 	if (dev_priv->display.hpd_irq_setup)
 		dev_priv->display.hpd_irq_setup(dev);
-	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+	spin_unlock_irq(&dev_priv->irq_lock);
 }
 
-/* Disable interrupts so we can allow runtime PM. */
-void intel_runtime_pm_disable_interrupts(struct drm_device *dev)
+/**
+ * intel_irq_install - enables the hardware interrupt
+ * @dev_priv: i915 device instance
+ *
+ * This function enables the hardware interrupt handling, but leaves the hotplug
+ * handling still disabled. It is called after intel_irq_init().
+ *
+ * In the driver load and resume code we need working interrupts in a few places
+ * but don't want to deal with the hassle of concurrent probe and hotplug
+ * workers. Hence the split into this two-stage approach.
+ */
+int intel_irq_install(struct drm_i915_private *dev_priv)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	/*
+	 * We enable some interrupt sources in our postinstall hooks, so mark
+	 * interrupts as enabled _before_ actually enabling them to avoid
+	 * special cases in our ordering checks.
+	 */
+	dev_priv->pm.irqs_enabled = true;
 
-	dev->driver->irq_uninstall(dev);
-	dev_priv->pm._irqs_disabled = true;
+	return drm_irq_install(dev_priv->dev, dev_priv->dev->pdev->irq);
 }
 
-/* Restore interrupts so we can recover from runtime PM. */
-void intel_runtime_pm_restore_interrupts(struct drm_device *dev)
+/**
+ * intel_irq_uninstall - finilizes all irq handling
+ * @dev_priv: i915 device instance
+ *
+ * This stops interrupt and hotplug handling and unregisters and frees all
+ * resources acquired in the init functions.
+ */
+void intel_irq_uninstall(struct drm_i915_private *dev_priv)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	drm_irq_uninstall(dev_priv->dev);
+	intel_hpd_cancel_work(dev_priv);
+	dev_priv->pm.irqs_enabled = false;
+}
 
-	dev_priv->pm._irqs_disabled = false;
-	dev->driver->irq_preinstall(dev);
-	dev->driver->irq_postinstall(dev);
+/**
+ * intel_runtime_pm_disable_interrupts - runtime interrupt disabling
+ * @dev_priv: i915 device instance
+ *
+ * This function is used to disable interrupts at runtime, both in the runtime
+ * pm and the system suspend/resume code.
+ */
+void intel_runtime_pm_disable_interrupts(struct drm_i915_private *dev_priv)
+{
+	dev_priv->dev->driver->irq_uninstall(dev_priv->dev);
+	dev_priv->pm.irqs_enabled = false;
+}
+
+/**
+ * intel_runtime_pm_enable_interrupts - runtime interrupt enabling
+ * @dev_priv: i915 device instance
+ *
+ * This function is used to enable interrupts at runtime, both in the runtime
+ * pm and the system suspend/resume code.
+ */
+void intel_runtime_pm_enable_interrupts(struct drm_i915_private *dev_priv)
+{
+	dev_priv->pm.irqs_enabled = true;
+	dev_priv->dev->driver->irq_preinstall(dev_priv->dev);
+	dev_priv->dev->driver->irq_postinstall(dev_priv->dev);
 }
diff --git a/drivers/gpu/drm/i915/i915_params.c b/drivers/gpu/drm/i915/i915_params.c
index 7f84dd263ee8..c91cb2033cc5 100644
--- a/drivers/gpu/drm/i915/i915_params.c
+++ b/drivers/gpu/drm/i915/i915_params.c
@@ -35,6 +35,7 @@ struct i915_params i915 __read_mostly = {
 	.vbt_sdvo_panel_type = -1,
 	.enable_rc6 = -1,
 	.enable_fbc = -1,
+	.enable_execlists = 0,
 	.enable_hangcheck = true,
 	.enable_ppgtt = -1,
 	.enable_psr = 0,
@@ -66,12 +67,12 @@ module_param_named(powersave, i915.powersave, int, 0600);
 MODULE_PARM_DESC(powersave,
 	"Enable powersavings, fbc, downclocking, etc. (default: true)");
 
-module_param_named(semaphores, i915.semaphores, int, 0400);
+module_param_named_unsafe(semaphores, i915.semaphores, int, 0400);
 MODULE_PARM_DESC(semaphores,
 	"Use semaphores for inter-ring sync "
 	"(default: -1 (use per-chip defaults))");
 
-module_param_named(enable_rc6, i915.enable_rc6, int, 0400);
+module_param_named_unsafe(enable_rc6, i915.enable_rc6, int, 0400);
 MODULE_PARM_DESC(enable_rc6,
 	"Enable power-saving render C-state 6. "
 	"Different stages can be selected via bitmask values "
@@ -79,7 +80,7 @@ MODULE_PARM_DESC(enable_rc6,
 	"For example, 3 would enable rc6 and deep rc6, and 7 would enable everything. "
 	"default: -1 (use per-chip default)");
 
-module_param_named(enable_fbc, i915.enable_fbc, int, 0600);
+module_param_named_unsafe(enable_fbc, i915.enable_fbc, int, 0600);
 MODULE_PARM_DESC(enable_fbc,
 	"Enable frame buffer compression for power savings "
 	"(default: -1 (use per-chip default))");
@@ -113,11 +114,16 @@ MODULE_PARM_DESC(enable_hangcheck,
 	"WARNING: Disabling this can cause system wide hangs. "
 	"(default: true)");
 
-module_param_named(enable_ppgtt, i915.enable_ppgtt, int, 0400);
+module_param_named_unsafe(enable_ppgtt, i915.enable_ppgtt, int, 0400);
 MODULE_PARM_DESC(enable_ppgtt,
 	"Override PPGTT usage. "
 	"(-1=auto [default], 0=disabled, 1=aliasing, 2=full)");
 
+module_param_named(enable_execlists, i915.enable_execlists, int, 0400);
+MODULE_PARM_DESC(enable_execlists,
+	"Override execlists usage. "
+	"(-1=auto, 0=disabled [default], 1=enabled)");
+
 module_param_named(enable_psr, i915.enable_psr, int, 0600);
 MODULE_PARM_DESC(enable_psr, "Enable PSR (default: false)");
 
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index f29b44c86a2f..172de3b3433b 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -26,14 +26,25 @@
 #define _I915_REG_H_
 
 #define _PIPE(pipe, a, b) ((a) + (pipe)*((b)-(a)))
+#define _PLANE(plane, a, b) _PIPE(plane, a, b)
 #define _TRANSCODER(tran, a, b) ((a) + (tran)*((b)-(a)))
-
 #define _PORT(port, a, b) ((a) + (port)*((b)-(a)))
 #define _PIPE3(pipe, a, b, c) ((pipe) == PIPE_A ? (a) : \
 			       (pipe) == PIPE_B ? (b) : (c))
 
-#define _MASKED_BIT_ENABLE(a) (((a) << 16) | (a))
-#define _MASKED_BIT_DISABLE(a) ((a) << 16)
+#define _MASKED_FIELD(mask, value) ({					   \
+	if (__builtin_constant_p(mask))					   \
+		BUILD_BUG_ON_MSG(((mask) & 0xffff0000), "Incorrect mask"); \
+	if (__builtin_constant_p(value))				   \
+		BUILD_BUG_ON_MSG((value) & 0xffff0000, "Incorrect value"); \
+	if (__builtin_constant_p(mask) && __builtin_constant_p(value))	   \
+		BUILD_BUG_ON_MSG((value) & ~(mask),			   \
+				 "Incorrect value for mask");		   \
+	(mask) << 16 | (value); })
+#define _MASKED_BIT_ENABLE(a)	({ typeof(a) _a = (a); _MASKED_FIELD(_a, _a); })
+#define _MASKED_BIT_DISABLE(a)	(_MASKED_FIELD((a), 0))
+
+
 
 /* PCI config space */
 
@@ -74,15 +85,17 @@
 #define   I915_GC_RENDER_CLOCK_166_MHZ	(0 << 0)
 #define   I915_GC_RENDER_CLOCK_200_MHZ	(1 << 0)
 #define   I915_GC_RENDER_CLOCK_333_MHZ	(4 << 0)
+#define GCDGMBUS 0xcc
 #define PCI_LBPC 0xf4 /* legacy/combination backlight modes, also called LBB */
 
 
 /* Graphics reset regs */
-#define I965_GDRST 0xc0 /* PCI config register */
+#define I915_GDRST 0xc0 /* PCI config register */
 #define  GRDOM_FULL	(0<<2)
 #define  GRDOM_RENDER	(1<<2)
 #define  GRDOM_MEDIA	(3<<2)
 #define  GRDOM_MASK	(3<<2)
+#define  GRDOM_RESET_STATUS (1<<1)
 #define  GRDOM_RESET_ENABLE (1<<0)
 
 #define ILK_GDSR 0x2ca4 /* MCHBAR offset */
@@ -143,6 +156,14 @@
 #define GAB_CTL				0x24000
 #define   GAB_CTL_CONT_AFTER_PAGEFAULT	(1<<8)
 
+#define GEN7_BIOS_RESERVED		0x1082C0
+#define GEN7_BIOS_RESERVED_1M		(0 << 5)
+#define GEN7_BIOS_RESERVED_256K		(1 << 5)
+#define GEN8_BIOS_RESERVED_SHIFT       7
+#define GEN7_BIOS_RESERVED_MASK        0x1
+#define GEN8_BIOS_RESERVED_MASK        0x3
+
+
 /* VGA stuff */
 
 #define VGA_ST01_MDA 0x3ba
@@ -240,6 +261,16 @@
 #define   MI_DISPLAY_FLIP_IVB_SPRITE_B (3 << 19)
 #define   MI_DISPLAY_FLIP_IVB_PLANE_C  (4 << 19)
 #define   MI_DISPLAY_FLIP_IVB_SPRITE_C (5 << 19)
+/* SKL ones */
+#define   MI_DISPLAY_FLIP_SKL_PLANE_1_A	(0 << 8)
+#define   MI_DISPLAY_FLIP_SKL_PLANE_1_B	(1 << 8)
+#define   MI_DISPLAY_FLIP_SKL_PLANE_1_C	(2 << 8)
+#define   MI_DISPLAY_FLIP_SKL_PLANE_2_A	(4 << 8)
+#define   MI_DISPLAY_FLIP_SKL_PLANE_2_B	(5 << 8)
+#define   MI_DISPLAY_FLIP_SKL_PLANE_2_C	(6 << 8)
+#define   MI_DISPLAY_FLIP_SKL_PLANE_3_A	(7 << 8)
+#define   MI_DISPLAY_FLIP_SKL_PLANE_3_B	(8 << 8)
+#define   MI_DISPLAY_FLIP_SKL_PLANE_3_C	(9 << 8)
 #define MI_SEMAPHORE_MBOX	MI_INSTR(0x16, 1) /* gen6, gen7 */
 #define   MI_SEMAPHORE_GLOBAL_GTT    (1<<22)
 #define   MI_SEMAPHORE_UPDATE	    (1<<21)
@@ -272,6 +303,7 @@
 #define   MI_SEMAPHORE_POLL		(1<<15)
 #define   MI_SEMAPHORE_SAD_GTE_SDD	(1<<12)
 #define MI_STORE_DWORD_IMM	MI_INSTR(0x20, 1)
+#define MI_STORE_DWORD_IMM_GEN8	MI_INSTR(0x20, 2)
 #define   MI_MEM_VIRTUAL	(1 << 22) /* 965+ only */
 #define MI_STORE_DWORD_INDEX	MI_INSTR(0x21, 1)
 #define   MI_STORE_DWORD_INDEX_SHIFT 2
@@ -282,6 +314,7 @@
  *   address/value pairs. Don't overdue it, though, x <= 2^4 must hold!
  */
 #define MI_LOAD_REGISTER_IMM(x)	MI_INSTR(0x22, 2*(x)-1)
+#define   MI_LRI_FORCE_POSTED		(1<<12)
 #define MI_STORE_REGISTER_MEM(x) MI_INSTR(0x24, 2*(x)-1)
 #define MI_STORE_REGISTER_MEM_GEN8(x) MI_INSTR(0x24, 3*(x)-1)
 #define   MI_SRM_LRM_GLOBAL_GTT		(1<<22)
@@ -304,6 +337,8 @@
 #define   MI_BATCH_GTT		    (2<<6) /* aliased with (1<<7) on gen4 */
 #define MI_BATCH_BUFFER_START_GEN8	MI_INSTR(0x31, 1)
 
+#define MI_PREDICATE_SRC0	(0x2400)
+#define MI_PREDICATE_SRC1	(0x2408)
 
 #define MI_PREDICATE_RESULT_2	(0x2214)
 #define  LOWER_SLICE_ENABLED	(1<<0)
@@ -360,6 +395,7 @@
 #define   PIPE_CONTROL_STORE_DATA_INDEX			(1<<21)
 #define   PIPE_CONTROL_CS_STALL				(1<<20)
 #define   PIPE_CONTROL_TLB_INVALIDATE			(1<<18)
+#define   PIPE_CONTROL_MEDIA_STATE_CLEAR		(1<<16)
 #define   PIPE_CONTROL_QW_WRITE				(1<<14)
 #define   PIPE_CONTROL_POST_SYNC_OP_MASK                (3<<14)
 #define   PIPE_CONTROL_DEPTH_STALL			(1<<13)
@@ -501,10 +537,26 @@
 #define BUNIT_REG_BISOC				0x11
 
 #define PUNIT_REG_DSPFREQ			0x36
+#define   DSPFREQSTAT_SHIFT_CHV			24
+#define   DSPFREQSTAT_MASK_CHV			(0x1f << DSPFREQSTAT_SHIFT_CHV)
+#define   DSPFREQGUAR_SHIFT_CHV			8
+#define   DSPFREQGUAR_MASK_CHV			(0x1f << DSPFREQGUAR_SHIFT_CHV)
 #define   DSPFREQSTAT_SHIFT			30
 #define   DSPFREQSTAT_MASK			(0x3 << DSPFREQSTAT_SHIFT)
 #define   DSPFREQGUAR_SHIFT			14
 #define   DSPFREQGUAR_MASK			(0x3 << DSPFREQGUAR_SHIFT)
+#define   _DP_SSC(val, pipe)			((val) << (2 * (pipe)))
+#define   DP_SSC_MASK(pipe)			_DP_SSC(0x3, (pipe))
+#define   DP_SSC_PWR_ON(pipe)			_DP_SSC(0x0, (pipe))
+#define   DP_SSC_CLK_GATE(pipe)			_DP_SSC(0x1, (pipe))
+#define   DP_SSC_RESET(pipe)			_DP_SSC(0x2, (pipe))
+#define   DP_SSC_PWR_GATE(pipe)			_DP_SSC(0x3, (pipe))
+#define   _DP_SSS(val, pipe)			((val) << (2 * (pipe) + 16))
+#define   DP_SSS_MASK(pipe)			_DP_SSS(0x3, (pipe))
+#define   DP_SSS_PWR_ON(pipe)			_DP_SSS(0x0, (pipe))
+#define   DP_SSS_CLK_GATE(pipe)			_DP_SSS(0x1, (pipe))
+#define   DP_SSS_RESET(pipe)			_DP_SSS(0x2, (pipe))
+#define   DP_SSS_PWR_GATE(pipe)			_DP_SSS(0x3, (pipe))
 
 /* See the PUNIT HAS v0.8 for the below bits */
 enum punit_power_well {
@@ -518,6 +570,11 @@ enum punit_power_well {
 	PUNIT_POWER_WELL_DPIO_TX_C_LANES_23	= 9,
 	PUNIT_POWER_WELL_DPIO_RX0		= 10,
 	PUNIT_POWER_WELL_DPIO_RX1		= 11,
+	PUNIT_POWER_WELL_DPIO_CMN_D		= 12,
+	/* FIXME: guesswork below */
+	PUNIT_POWER_WELL_DPIO_TX_D_LANES_01	= 13,
+	PUNIT_POWER_WELL_DPIO_TX_D_LANES_23	= 14,
+	PUNIT_POWER_WELL_DPIO_RX2		= 15,
 
 	PUNIT_POWER_WELL_NUM,
 };
@@ -533,6 +590,7 @@ enum punit_power_well {
 #define PUNIT_REG_GPU_LFM			0xd3
 #define PUNIT_REG_GPU_FREQ_REQ			0xd4
 #define PUNIT_REG_GPU_FREQ_STS			0xd8
+#define   GPLLENABLE				(1<<4)
 #define   GENFREQSTATUS				(1<<0)
 #define PUNIT_REG_MEDIA_TURBO_FREQ_REQ		0xdc
 #define PUNIT_REG_CZ_TIMESTAMP			0xce
@@ -641,7 +699,7 @@ enum punit_power_well {
  * need to be accessed during AUX communication,
  *
  * Generally the common lane corresponds to the pipe and
- * the spline (PCS/TX) correponds to the port.
+ * the spline (PCS/TX) corresponds to the port.
  *
  * For dual channel PHY (VLV/CHV):
  *
@@ -765,6 +823,8 @@ enum punit_power_well {
 #define _VLV_PCS_DW0_CH1		0x8400
 #define   DPIO_PCS_TX_LANE2_RESET	(1<<16)
 #define   DPIO_PCS_TX_LANE1_RESET	(1<<7)
+#define   DPIO_LEFT_TXFIFO_RST_MASTER2	(1<<4)
+#define   DPIO_RIGHT_TXFIFO_RST_MASTER2	(1<<3)
 #define VLV_PCS_DW0(ch) _PORT(ch, _VLV_PCS_DW0_CH0, _VLV_PCS_DW0_CH1)
 
 #define _VLV_PCS01_DW0_CH0		0x200
@@ -805,12 +865,31 @@ enum punit_power_well {
 
 #define _VLV_PCS_DW9_CH0		0x8224
 #define _VLV_PCS_DW9_CH1		0x8424
+#define   DPIO_PCS_TX2MARGIN_MASK	(0x7<<13)
+#define   DPIO_PCS_TX2MARGIN_000	(0<<13)
+#define   DPIO_PCS_TX2MARGIN_101	(1<<13)
+#define   DPIO_PCS_TX1MARGIN_MASK	(0x7<<10)
+#define   DPIO_PCS_TX1MARGIN_000	(0<<10)
+#define   DPIO_PCS_TX1MARGIN_101	(1<<10)
 #define	VLV_PCS_DW9(ch) _PORT(ch, _VLV_PCS_DW9_CH0, _VLV_PCS_DW9_CH1)
 
+#define _VLV_PCS01_DW9_CH0		0x224
+#define _VLV_PCS23_DW9_CH0		0x424
+#define _VLV_PCS01_DW9_CH1		0x2624
+#define _VLV_PCS23_DW9_CH1		0x2824
+#define VLV_PCS01_DW9(ch) _PORT(ch, _VLV_PCS01_DW9_CH0, _VLV_PCS01_DW9_CH1)
+#define VLV_PCS23_DW9(ch) _PORT(ch, _VLV_PCS23_DW9_CH0, _VLV_PCS23_DW9_CH1)
+
 #define _CHV_PCS_DW10_CH0		0x8228
 #define _CHV_PCS_DW10_CH1		0x8428
 #define   DPIO_PCS_SWING_CALC_TX0_TX2	(1<<30)
 #define   DPIO_PCS_SWING_CALC_TX1_TX3	(1<<31)
+#define   DPIO_PCS_TX2DEEMP_MASK	(0xf<<24)
+#define   DPIO_PCS_TX2DEEMP_9P5		(0<<24)
+#define   DPIO_PCS_TX2DEEMP_6P0		(2<<24)
+#define   DPIO_PCS_TX1DEEMP_MASK	(0xf<<16)
+#define   DPIO_PCS_TX1DEEMP_9P5		(0<<16)
+#define   DPIO_PCS_TX1DEEMP_6P0		(2<<16)
 #define CHV_PCS_DW10(ch) _PORT(ch, _CHV_PCS_DW10_CH0, _CHV_PCS_DW10_CH1)
 
 #define _VLV_PCS01_DW10_CH0		0x0228
@@ -822,8 +901,18 @@ enum punit_power_well {
 
 #define _VLV_PCS_DW11_CH0		0x822c
 #define _VLV_PCS_DW11_CH1		0x842c
+#define   DPIO_LANEDESKEW_STRAP_OVRD	(1<<3)
+#define   DPIO_LEFT_TXFIFO_RST_MASTER	(1<<1)
+#define   DPIO_RIGHT_TXFIFO_RST_MASTER	(1<<0)
 #define VLV_PCS_DW11(ch) _PORT(ch, _VLV_PCS_DW11_CH0, _VLV_PCS_DW11_CH1)
 
+#define _VLV_PCS01_DW11_CH0		0x022c
+#define _VLV_PCS23_DW11_CH0		0x042c
+#define _VLV_PCS01_DW11_CH1		0x262c
+#define _VLV_PCS23_DW11_CH1		0x282c
+#define VLV_PCS01_DW11(ch) _PORT(ch, _VLV_PCS01_DW11_CH0, _VLV_PCS01_DW11_CH1)
+#define VLV_PCS23_DW11(ch) _PORT(ch, _VLV_PCS23_DW11_CH0, _VLV_PCS23_DW11_CH1)
+
 #define _VLV_PCS_DW12_CH0		0x8230
 #define _VLV_PCS_DW12_CH1		0x8430
 #define VLV_PCS_DW12(ch) _PORT(ch, _VLV_PCS_DW12_CH0, _VLV_PCS_DW12_CH1)
@@ -838,8 +927,8 @@ enum punit_power_well {
 
 #define _VLV_TX_DW2_CH0			0x8288
 #define _VLV_TX_DW2_CH1			0x8488
-#define   DPIO_SWING_MARGIN_SHIFT	16
-#define   DPIO_SWING_MARGIN_MASK	(0xff << DPIO_SWING_MARGIN_SHIFT)
+#define   DPIO_SWING_MARGIN000_SHIFT	16
+#define   DPIO_SWING_MARGIN000_MASK	(0xff << DPIO_SWING_MARGIN000_SHIFT)
 #define   DPIO_UNIQ_TRANS_SCALE_SHIFT	8
 #define VLV_TX_DW2(ch) _PORT(ch, _VLV_TX_DW2_CH0, _VLV_TX_DW2_CH1)
 
@@ -847,12 +936,16 @@ enum punit_power_well {
 #define _VLV_TX_DW3_CH1			0x848c
 /* The following bit for CHV phy */
 #define   DPIO_TX_UNIQ_TRANS_SCALE_EN	(1<<27)
+#define   DPIO_SWING_MARGIN101_SHIFT	16
+#define   DPIO_SWING_MARGIN101_MASK	(0xff << DPIO_SWING_MARGIN101_SHIFT)
 #define VLV_TX_DW3(ch) _PORT(ch, _VLV_TX_DW3_CH0, _VLV_TX_DW3_CH1)
 
 #define _VLV_TX_DW4_CH0			0x8290
 #define _VLV_TX_DW4_CH1			0x8490
 #define   DPIO_SWING_DEEMPH9P5_SHIFT	24
 #define   DPIO_SWING_DEEMPH9P5_MASK	(0xff << DPIO_SWING_DEEMPH9P5_SHIFT)
+#define   DPIO_SWING_DEEMPH6P0_SHIFT	16
+#define   DPIO_SWING_DEEMPH6P0_MASK	(0xff << DPIO_SWING_DEEMPH6P0_SHIFT)
 #define VLV_TX_DW4(ch) _PORT(ch, _VLV_TX_DW4_CH0, _VLV_TX_DW4_CH1)
 
 #define _VLV_TX3_DW4_CH0		0x690
@@ -1003,6 +1096,13 @@ enum punit_power_well {
 #define   PGTBL_ADDRESS_LO_MASK	0xfffff000 /* bits [31:12] */
 #define   PGTBL_ADDRESS_HI_MASK	0x000000f0 /* bits [35:32] (gen4) */
 #define PGTBL_ER	0x02024
+#define PRB0_BASE (0x2030-0x30)
+#define PRB1_BASE (0x2040-0x30) /* 830,gen3 */
+#define PRB2_BASE (0x2050-0x30) /* gen3 */
+#define SRB0_BASE (0x2100-0x30) /* gen2 */
+#define SRB1_BASE (0x2110-0x30) /* gen2 */
+#define SRB2_BASE (0x2120-0x30) /* 830 */
+#define SRB3_BASE (0x2130-0x30) /* 830 */
 #define RENDER_RING_BASE	0x02000
 #define BSD_RING_BASE		0x04000
 #define GEN6_BSD_RING_BASE	0x12000
@@ -1029,6 +1129,7 @@ enum punit_power_well {
 #define GEN6_VERSYNC	(RING_SYNC_1(VEBOX_RING_BASE))
 #define GEN6_VEVSYNC	(RING_SYNC_2(VEBOX_RING_BASE))
 #define GEN6_NOSYNC 0
+#define RING_PSMI_CTL(base)	((base)+0x50)
 #define RING_MAX_IDLE(base)	((base)+0x54)
 #define RING_HWS_PGA(base)	((base)+0x80)
 #define RING_HWS_PGA_GEN6(base)	((base)+0x2080)
@@ -1064,6 +1165,7 @@ enum punit_power_well {
 #define RING_ACTHD_UDW(base)	((base)+0x5c)
 #define RING_NOPID(base)	((base)+0x94)
 #define RING_IMR(base)		((base)+0xa8)
+#define RING_HWSTAM(base)	((base)+0x98)
 #define RING_TIMESTAMP(base)	((base)+0x358)
 #define   TAIL_ADDR		0x001FFFF8
 #define   HEAD_WRAP_COUNT	0xFFE00000
@@ -1194,7 +1296,7 @@ enum punit_power_well {
 #define   GEN6_WIZ_HASHING_8x8				GEN6_WIZ_HASHING(0, 0)
 #define   GEN6_WIZ_HASHING_8x4				GEN6_WIZ_HASHING(0, 1)
 #define   GEN6_WIZ_HASHING_16x4				GEN6_WIZ_HASHING(1, 0)
-#define   GEN6_WIZ_HASHING_MASK				(GEN6_WIZ_HASHING(1, 1) << 16)
+#define   GEN6_WIZ_HASHING_MASK				GEN6_WIZ_HASHING(1, 1)
 #define   GEN6_TD_FOUR_ROW_DISPATCH_DISABLE		(1 << 5)
 
 #define GFX_MODE	0x02520
@@ -1248,6 +1350,10 @@ enum punit_power_well {
 #define   INSTPM_TLB_INVALIDATE	(1<<9)
 #define   INSTPM_SYNC_FLUSH	(1<<5)
 #define ACTHD	        0x020c8
+#define MEM_MODE	0x020cc
+#define   MEM_DISPLAY_B_TRICKLE_FEED_DISABLE (1<<3) /* 830 only */
+#define   MEM_DISPLAY_A_TRICKLE_FEED_DISABLE (1<<2) /* 830/845 only */
+#define   MEM_DISPLAY_TRICKLE_FEED_DISABLE (1<<2) /* 85x only */
 #define FW_BLC		0x020d8
 #define FW_BLC2		0x020dc
 #define FW_BLC_SELF	0x020e0 /* 915+ only */
@@ -1354,6 +1460,7 @@ enum punit_power_well {
 #define   GEN6_BLITTER_FBC_NOTIFY			(1<<3)
 
 #define GEN6_RC_SLEEP_PSMI_CONTROL	0x2050
+#define   GEN6_PSMI_SLEEP_MSG_DISABLE	(1 << 0)
 #define   GEN8_RC_SEMA_IDLE_MSG_DISABLE	(1 << 12)
 #define   GEN8_FF_DOP_CLOCK_GATE_DISABLE	(1<<10)
 
@@ -1380,6 +1487,7 @@ enum punit_power_well {
 #define GT_BSD_CS_ERROR_INTERRUPT		(1 << 15)
 #define GT_BSD_USER_INTERRUPT			(1 << 12)
 #define GT_RENDER_L3_PARITY_ERROR_INTERRUPT_S1	(1 << 11) /* hsw+; rsvd on snb, ivb, vlv */
+#define GT_CONTEXT_SWITCH_INTERRUPT		(1 <<  8)
 #define GT_RENDER_L3_PARITY_ERROR_INTERRUPT	(1 <<  5) /* !snb */
 #define GT_RENDER_PIPECTL_NOTIFY_INTERRUPT	(1 <<  4)
 #define GT_RENDER_CS_MASTER_ERROR_INTERRUPT	(1 <<  3)
@@ -1519,6 +1627,7 @@ enum punit_power_well {
 /* Framebuffer compression for Ironlake */
 #define ILK_DPFC_CB_BASE	0x43200
 #define ILK_DPFC_CONTROL	0x43208
+#define   FBC_CTL_FALSE_COLOR	(1<<10)
 /* The bit 28-8 is reserved */
 #define   DPFC_RESERVED		(0x1FFFFF00)
 #define ILK_DPFC_RECOMP_CTL	0x4320c
@@ -1675,12 +1784,9 @@ enum punit_power_well {
 #define DPIO_PHY_STATUS			(VLV_DISPLAY_BASE + 0x6240)
 #define   DPLL_PORTD_READY_MASK		(0xf)
 #define DISPLAY_PHY_CONTROL (VLV_DISPLAY_BASE + 0x60100)
-#define   PHY_COM_LANE_RESET_DEASSERT(phy, val) \
-				((phy == DPIO_PHY0) ? (val | 1) : (val | 2))
-#define   PHY_COM_LANE_RESET_ASSERT(phy, val) \
-				((phy == DPIO_PHY0) ? (val & ~1) : (val & ~2))
+#define   PHY_COM_LANE_RESET_DEASSERT(phy) (1 << (phy))
 #define DISPLAY_PHY_STATUS (VLV_DISPLAY_BASE + 0x60104)
-#define   PHY_POWERGOOD(phy)	((phy == DPIO_PHY0) ? (1<<31) : (1<<30))
+#define   PHY_POWERGOOD(phy)	(((phy) == DPIO_PHY0) ? (1<<31) : (1<<30))
 
 /*
  * The i830 generation, in LVDS mode, defines P1 as the bit number set within
@@ -1953,6 +2059,8 @@ enum punit_power_well {
 #define DCC_ADDRESSING_MODE_MASK			(3 << 0)
 #define DCC_CHANNEL_XOR_DISABLE				(1 << 10)
 #define DCC_CHANNEL_XOR_BIT_17				(1 << 9)
+#define DCC2			0x10204
+#define DCC2_MODIFIED_ENHANCED_DISABLE			(1 << 20)
 
 /* Pineview MCH register contains DDR3 setting */
 #define CSHRDDR3CTL            0x101a8
@@ -2236,7 +2344,6 @@ enum punit_power_well {
 
 #define GEN6_GT_THREAD_STATUS_REG 0x13805c
 #define GEN6_GT_THREAD_STATUS_CORE_MASK 0x7
-#define GEN6_GT_THREAD_STATUS_CORE_MASK_HSW (0x7 | (0x07 << 16))
 
 #define GEN6_GT_PERF_STATUS	(MCHBAR_MIRROR_BASE_SNB + 0x5948)
 #define GEN6_RP_STATE_LIMITS	(MCHBAR_MIRROR_BASE_SNB + 0x5994)
@@ -2397,6 +2504,7 @@ enum punit_power_well {
 #define _PIPEASRC	0x6001c
 #define _BCLRPAT_A	0x60020
 #define _VSYNCSHIFT_A	0x60028
+#define _PIPE_MULT_A	0x6002c
 
 /* Pipe B timing regs */
 #define _HTOTAL_B	0x61000
@@ -2408,6 +2516,7 @@ enum punit_power_well {
 #define _PIPEBSRC	0x6101c
 #define _BCLRPAT_B	0x61020
 #define _VSYNCSHIFT_B	0x61028
+#define _PIPE_MULT_B	0x6102c
 
 #define TRANSCODER_A_OFFSET 0x60000
 #define TRANSCODER_B_OFFSET 0x61000
@@ -2428,6 +2537,7 @@ enum punit_power_well {
 #define BCLRPAT(trans) _TRANSCODER2(trans, _BCLRPAT_A)
 #define VSYNCSHIFT(trans) _TRANSCODER2(trans, _VSYNCSHIFT_A)
 #define PIPESRC(trans) _TRANSCODER2(trans, _PIPEASRC)
+#define PIPE_MULT(trans) _TRANSCODER2(trans, _PIPE_MULT_A)
 
 /* HSW+ eDP PSR registers */
 #define EDP_PSR_BASE(dev)                       (IS_HASWELL(dev) ? 0x64800 : 0x6f800)
@@ -2457,9 +2567,7 @@ enum punit_power_well {
 
 #define EDP_PSR_AUX_CTL(dev)			(EDP_PSR_BASE(dev) + 0x10)
 #define EDP_PSR_AUX_DATA1(dev)			(EDP_PSR_BASE(dev) + 0x14)
-#define   EDP_PSR_DPCD_COMMAND		0x80060000
 #define EDP_PSR_AUX_DATA2(dev)			(EDP_PSR_BASE(dev) + 0x18)
-#define   EDP_PSR_DPCD_NORMAL_OPERATION	(1<<24)
 #define EDP_PSR_AUX_DATA3(dev)			(EDP_PSR_BASE(dev) + 0x1c)
 #define EDP_PSR_AUX_DATA4(dev)			(EDP_PSR_BASE(dev) + 0x20)
 #define EDP_PSR_AUX_DATA5(dev)			(EDP_PSR_BASE(dev) + 0x24)
@@ -3476,6 +3584,8 @@ enum punit_power_well {
 #define   DP_LINK_TRAIN_OFF		(3 << 28)
 #define   DP_LINK_TRAIN_MASK		(3 << 28)
 #define   DP_LINK_TRAIN_SHIFT		28
+#define   DP_LINK_TRAIN_PAT_3_CHV	(1 << 14)
+#define   DP_LINK_TRAIN_MASK_CHV	((3 << 28)|(1<<14))
 
 /* CPT Link training mode */
 #define   DP_LINK_TRAIN_PAT_1_CPT	(0 << 8)
@@ -3594,6 +3704,7 @@ enum punit_power_well {
 #define   DP_AUX_CH_CTL_PRECHARGE_TEST	    (1 << 11)
 #define   DP_AUX_CH_CTL_BIT_CLOCK_2X_MASK    (0x7ff)
 #define   DP_AUX_CH_CTL_BIT_CLOCK_2X_SHIFT   0
+#define   DP_AUX_CH_CTL_SYNC_PULSE_SKL(c)   ((c) - 1)
 
 /*
  * Computing GMCH M and N values for the Display Port link
@@ -3732,7 +3843,6 @@ enum punit_power_well {
 #define   PIPE_VSYNC_INTERRUPT_STATUS		(1UL<<9)
 #define   PIPE_DISPLAY_LINE_COMPARE_STATUS	(1UL<<8)
 #define   PIPE_DPST_EVENT_STATUS		(1UL<<7)
-#define   PIPE_LEGACY_BLC_EVENT_STATUS		(1UL<<6)
 #define   PIPE_A_PSR_STATUS_VLV			(1UL<<6)
 #define   PIPE_LEGACY_BLC_EVENT_STATUS		(1UL<<6)
 #define   PIPE_ODD_FIELD_INTERRUPT_STATUS	(1UL<<5)
@@ -3842,73 +3952,152 @@ enum punit_power_well {
 #define   DSPARB_BEND_SHIFT	9 /* on 855 */
 #define   DSPARB_AEND_SHIFT	0
 
+/* pnv/gen4/g4x/vlv/chv */
 #define DSPFW1			(dev_priv->info.display_mmio_offset + 0x70034)
-#define   DSPFW_SR_SHIFT	23
-#define   DSPFW_SR_MASK		(0x1ff<<23)
-#define   DSPFW_CURSORB_SHIFT	16
-#define   DSPFW_CURSORB_MASK	(0x3f<<16)
-#define   DSPFW_PLANEB_SHIFT	8
-#define   DSPFW_PLANEB_MASK	(0x7f<<8)
-#define   DSPFW_PLANEA_MASK	(0x7f)
+#define   DSPFW_SR_SHIFT		23
+#define   DSPFW_SR_MASK			(0x1ff<<23)
+#define   DSPFW_CURSORB_SHIFT		16
+#define   DSPFW_CURSORB_MASK		(0x3f<<16)
+#define   DSPFW_PLANEB_SHIFT		8
+#define   DSPFW_PLANEB_MASK		(0x7f<<8)
+#define   DSPFW_PLANEB_MASK_VLV		(0xff<<8) /* vlv/chv */
+#define   DSPFW_PLANEA_SHIFT		0
+#define   DSPFW_PLANEA_MASK		(0x7f<<0)
+#define   DSPFW_PLANEA_MASK_VLV		(0xff<<0) /* vlv/chv */
 #define DSPFW2			(dev_priv->info.display_mmio_offset + 0x70038)
-#define   DSPFW_CURSORA_MASK	0x00003f00
-#define   DSPFW_CURSORA_SHIFT	8
-#define   DSPFW_PLANEC_MASK	(0x7f)
+#define   DSPFW_FBC_SR_EN		(1<<31)	  /* g4x */
+#define   DSPFW_FBC_SR_SHIFT		28
+#define   DSPFW_FBC_SR_MASK		(0x7<<28) /* g4x */
+#define   DSPFW_FBC_HPLL_SR_SHIFT	24
+#define   DSPFW_FBC_HPLL_SR_MASK	(0xf<<24) /* g4x */
+#define   DSPFW_SPRITEB_SHIFT		(16)
+#define   DSPFW_SPRITEB_MASK		(0x7f<<16) /* g4x */
+#define   DSPFW_SPRITEB_MASK_VLV	(0xff<<16) /* vlv/chv */
+#define   DSPFW_CURSORA_SHIFT		8
+#define   DSPFW_CURSORA_MASK		(0x3f<<8)
+#define   DSPFW_PLANEC_SHIFT_OLD	0
+#define   DSPFW_PLANEC_MASK_OLD		(0x7f<<0) /* pre-gen4 sprite C */
+#define   DSPFW_SPRITEA_SHIFT		0
+#define   DSPFW_SPRITEA_MASK		(0x7f<<0) /* g4x */
+#define   DSPFW_SPRITEA_MASK_VLV	(0xff<<0) /* vlv/chv */
 #define DSPFW3			(dev_priv->info.display_mmio_offset + 0x7003c)
-#define   DSPFW_HPLL_SR_EN	(1<<31)
-#define   DSPFW_CURSOR_SR_SHIFT	24
+#define   DSPFW_HPLL_SR_EN		(1<<31)
 #define   PINEVIEW_SELF_REFRESH_EN	(1<<30)
+#define   DSPFW_CURSOR_SR_SHIFT		24
 #define   DSPFW_CURSOR_SR_MASK		(0x3f<<24)
 #define   DSPFW_HPLL_CURSOR_SHIFT	16
 #define   DSPFW_HPLL_CURSOR_MASK	(0x3f<<16)
-#define   DSPFW_HPLL_SR_MASK		(0x1ff)
-#define DSPFW4			(dev_priv->info.display_mmio_offset + 0x70070)
-#define DSPFW7			(dev_priv->info.display_mmio_offset + 0x7007c)
+#define   DSPFW_HPLL_SR_SHIFT		0
+#define   DSPFW_HPLL_SR_MASK		(0x1ff<<0)
+
+/* vlv/chv */
+#define DSPFW4			(VLV_DISPLAY_BASE + 0x70070)
+#define   DSPFW_SPRITEB_WM1_SHIFT	16
+#define   DSPFW_SPRITEB_WM1_MASK	(0xff<<16)
+#define   DSPFW_CURSORA_WM1_SHIFT	8
+#define   DSPFW_CURSORA_WM1_MASK	(0x3f<<8)
+#define   DSPFW_SPRITEA_WM1_SHIFT	0
+#define   DSPFW_SPRITEA_WM1_MASK	(0xff<<0)
+#define DSPFW5			(VLV_DISPLAY_BASE + 0x70074)
+#define   DSPFW_PLANEB_WM1_SHIFT	24
+#define   DSPFW_PLANEB_WM1_MASK		(0xff<<24)
+#define   DSPFW_PLANEA_WM1_SHIFT	16
+#define   DSPFW_PLANEA_WM1_MASK		(0xff<<16)
+#define   DSPFW_CURSORB_WM1_SHIFT	8
+#define   DSPFW_CURSORB_WM1_MASK	(0x3f<<8)
+#define   DSPFW_CURSOR_SR_WM1_SHIFT	0
+#define   DSPFW_CURSOR_SR_WM1_MASK	(0x3f<<0)
+#define DSPFW6			(VLV_DISPLAY_BASE + 0x70078)
+#define   DSPFW_SR_WM1_SHIFT		0
+#define   DSPFW_SR_WM1_MASK		(0x1ff<<0)
+#define DSPFW7			(VLV_DISPLAY_BASE + 0x7007c)
+#define DSPFW7_CHV		(VLV_DISPLAY_BASE + 0x700b4) /* wtf #1? */
+#define   DSPFW_SPRITED_WM1_SHIFT	24
+#define   DSPFW_SPRITED_WM1_MASK	(0xff<<24)
+#define   DSPFW_SPRITED_SHIFT		16
+#define   DSPFW_SPRITED_MASK		(0xff<<16)
+#define   DSPFW_SPRITEC_WM1_SHIFT	8
+#define   DSPFW_SPRITEC_WM1_MASK	(0xff<<8)
+#define   DSPFW_SPRITEC_SHIFT		0
+#define   DSPFW_SPRITEC_MASK		(0xff<<0)
+#define DSPFW8_CHV		(VLV_DISPLAY_BASE + 0x700b8)
+#define   DSPFW_SPRITEF_WM1_SHIFT	24
+#define   DSPFW_SPRITEF_WM1_MASK	(0xff<<24)
+#define   DSPFW_SPRITEF_SHIFT		16
+#define   DSPFW_SPRITEF_MASK		(0xff<<16)
+#define   DSPFW_SPRITEE_WM1_SHIFT	8
+#define   DSPFW_SPRITEE_WM1_MASK	(0xff<<8)
+#define   DSPFW_SPRITEE_SHIFT		0
+#define   DSPFW_SPRITEE_MASK		(0xff<<0)
+#define DSPFW9_CHV		(VLV_DISPLAY_BASE + 0x7007c) /* wtf #2? */
+#define   DSPFW_PLANEC_WM1_SHIFT	24
+#define   DSPFW_PLANEC_WM1_MASK		(0xff<<24)
+#define   DSPFW_PLANEC_SHIFT		16
+#define   DSPFW_PLANEC_MASK		(0xff<<16)
+#define   DSPFW_CURSORC_WM1_SHIFT	8
+#define   DSPFW_CURSORC_WM1_MASK	(0x3f<<16)
+#define   DSPFW_CURSORC_SHIFT		0
+#define   DSPFW_CURSORC_MASK		(0x3f<<0)
+
+/* vlv/chv high order bits */
+#define DSPHOWM			(VLV_DISPLAY_BASE + 0x70064)
+#define   DSPFW_SR_HI_SHIFT		24
+#define   DSPFW_SR_HI_MASK		(1<<24)
+#define   DSPFW_SPRITEF_HI_SHIFT	23
+#define   DSPFW_SPRITEF_HI_MASK		(1<<23)
+#define   DSPFW_SPRITEE_HI_SHIFT	22
+#define   DSPFW_SPRITEE_HI_MASK		(1<<22)
+#define   DSPFW_PLANEC_HI_SHIFT		21
+#define   DSPFW_PLANEC_HI_MASK		(1<<21)
+#define   DSPFW_SPRITED_HI_SHIFT	20
+#define   DSPFW_SPRITED_HI_MASK		(1<<20)
+#define   DSPFW_SPRITEC_HI_SHIFT	16
+#define   DSPFW_SPRITEC_HI_MASK		(1<<16)
+#define   DSPFW_PLANEB_HI_SHIFT		12
+#define   DSPFW_PLANEB_HI_MASK		(1<<12)
+#define   DSPFW_SPRITEB_HI_SHIFT	8
+#define   DSPFW_SPRITEB_HI_MASK		(1<<8)
+#define   DSPFW_SPRITEA_HI_SHIFT	4
+#define   DSPFW_SPRITEA_HI_MASK		(1<<4)
+#define   DSPFW_PLANEA_HI_SHIFT		0
+#define   DSPFW_PLANEA_HI_MASK		(1<<0)
+#define DSPHOWM1		(VLV_DISPLAY_BASE + 0x70068)
+#define   DSPFW_SR_WM1_HI_SHIFT		24
+#define   DSPFW_SR_WM1_HI_MASK		(1<<24)
+#define   DSPFW_SPRITEF_WM1_HI_SHIFT	23
+#define   DSPFW_SPRITEF_WM1_HI_MASK	(1<<23)
+#define   DSPFW_SPRITEE_WM1_HI_SHIFT	22
+#define   DSPFW_SPRITEE_WM1_HI_MASK	(1<<22)
+#define   DSPFW_PLANEC_WM1_HI_SHIFT	21
+#define   DSPFW_PLANEC_WM1_HI_MASK	(1<<21)
+#define   DSPFW_SPRITED_WM1_HI_SHIFT	20
+#define   DSPFW_SPRITED_WM1_HI_MASK	(1<<20)
+#define   DSPFW_SPRITEC_WM1_HI_SHIFT	16
+#define   DSPFW_SPRITEC_WM1_HI_MASK	(1<<16)
+#define   DSPFW_PLANEB_WM1_HI_SHIFT	12
+#define   DSPFW_PLANEB_WM1_HI_MASK	(1<<12)
+#define   DSPFW_SPRITEB_WM1_HI_SHIFT	8
+#define   DSPFW_SPRITEB_WM1_HI_MASK	(1<<8)
+#define   DSPFW_SPRITEA_WM1_HI_SHIFT	4
+#define   DSPFW_SPRITEA_WM1_HI_MASK	(1<<4)
+#define   DSPFW_PLANEA_WM1_HI_SHIFT	0
+#define   DSPFW_PLANEA_WM1_HI_MASK	(1<<0)
 
 /* drain latency register values*/
+#define DRAIN_LATENCY_PRECISION_16	16
 #define DRAIN_LATENCY_PRECISION_32	32
 #define DRAIN_LATENCY_PRECISION_64	64
-#define VLV_DDL1			(VLV_DISPLAY_BASE + 0x70050)
-#define DDL_CURSORA_PRECISION_64	(1<<31)
-#define DDL_CURSORA_PRECISION_32	(0<<31)
-#define DDL_CURSORA_SHIFT		24
-#define DDL_SPRITEB_PRECISION_64	(1<<23)
-#define DDL_SPRITEB_PRECISION_32	(0<<23)
-#define DDL_SPRITEB_SHIFT		16
-#define DDL_SPRITEA_PRECISION_64	(1<<15)
-#define DDL_SPRITEA_PRECISION_32	(0<<15)
-#define DDL_SPRITEA_SHIFT		8
-#define DDL_PLANEA_PRECISION_64		(1<<7)
-#define DDL_PLANEA_PRECISION_32		(0<<7)
-#define DDL_PLANEA_SHIFT		0
-
-#define VLV_DDL2			(VLV_DISPLAY_BASE + 0x70054)
-#define DDL_CURSORB_PRECISION_64	(1<<31)
-#define DDL_CURSORB_PRECISION_32	(0<<31)
-#define DDL_CURSORB_SHIFT		24
-#define DDL_SPRITED_PRECISION_64	(1<<23)
-#define DDL_SPRITED_PRECISION_32	(0<<23)
-#define DDL_SPRITED_SHIFT		16
-#define DDL_SPRITEC_PRECISION_64	(1<<15)
-#define DDL_SPRITEC_PRECISION_32	(0<<15)
-#define DDL_SPRITEC_SHIFT		8
-#define DDL_PLANEB_PRECISION_64		(1<<7)
-#define DDL_PLANEB_PRECISION_32		(0<<7)
-#define DDL_PLANEB_SHIFT		0
-
-#define VLV_DDL3			(VLV_DISPLAY_BASE + 0x70058)
-#define DDL_CURSORC_PRECISION_64	(1<<31)
-#define DDL_CURSORC_PRECISION_32	(0<<31)
-#define DDL_CURSORC_SHIFT		24
-#define DDL_SPRITEF_PRECISION_64	(1<<23)
-#define DDL_SPRITEF_PRECISION_32	(0<<23)
-#define DDL_SPRITEF_SHIFT		16
-#define DDL_SPRITEE_PRECISION_64	(1<<15)
-#define DDL_SPRITEE_PRECISION_32	(0<<15)
-#define DDL_SPRITEE_SHIFT		8
-#define DDL_PLANEC_PRECISION_64		(1<<7)
-#define DDL_PLANEC_PRECISION_32		(0<<7)
-#define DDL_PLANEC_SHIFT		0
+#define VLV_DDL(pipe)			(VLV_DISPLAY_BASE + 0x70050 + 4 * (pipe))
+#define DDL_CURSOR_PRECISION_HIGH	(1<<31)
+#define DDL_CURSOR_PRECISION_LOW	(0<<31)
+#define DDL_CURSOR_SHIFT		24
+#define DDL_SPRITE_PRECISION_HIGH(sprite)	(1<<(15+8*(sprite)))
+#define DDL_SPRITE_PRECISION_LOW(sprite)	(0<<(15+8*(sprite)))
+#define DDL_SPRITE_SHIFT(sprite)	(8+8*(sprite))
+#define DDL_PLANE_PRECISION_HIGH	(1<<7)
+#define DDL_PLANE_PRECISION_LOW		(0<<7)
+#define DDL_PLANE_SHIFT			0
+#define DRAIN_LATENCY_MASK		0x7f
 
 /* FIFO watermark sizes etc */
 #define G4X_FIFO_LINE_SIZE	64
@@ -3943,6 +4132,41 @@ enum punit_power_well {
 #define I965_CURSOR_MAX_WM	32
 #define I965_CURSOR_DFT_WM	8
 
+/* Watermark register definitions for SKL */
+#define CUR_WM_A_0		0x70140
+#define CUR_WM_B_0		0x71140
+#define PLANE_WM_1_A_0		0x70240
+#define PLANE_WM_1_B_0		0x71240
+#define PLANE_WM_2_A_0		0x70340
+#define PLANE_WM_2_B_0		0x71340
+#define PLANE_WM_TRANS_1_A_0	0x70268
+#define PLANE_WM_TRANS_1_B_0	0x71268
+#define PLANE_WM_TRANS_2_A_0	0x70368
+#define PLANE_WM_TRANS_2_B_0	0x71368
+#define CUR_WM_TRANS_A_0	0x70168
+#define CUR_WM_TRANS_B_0	0x71168
+#define   PLANE_WM_EN		(1 << 31)
+#define   PLANE_WM_LINES_SHIFT	14
+#define   PLANE_WM_LINES_MASK	0x1f
+#define   PLANE_WM_BLOCKS_MASK	0x3ff
+
+#define CUR_WM_0(pipe) _PIPE(pipe, CUR_WM_A_0, CUR_WM_B_0)
+#define CUR_WM(pipe, level) (CUR_WM_0(pipe) + ((4) * (level)))
+#define CUR_WM_TRANS(pipe) _PIPE(pipe, CUR_WM_TRANS_A_0, CUR_WM_TRANS_B_0)
+
+#define _PLANE_WM_1(pipe) _PIPE(pipe, PLANE_WM_1_A_0, PLANE_WM_1_B_0)
+#define _PLANE_WM_2(pipe) _PIPE(pipe, PLANE_WM_2_A_0, PLANE_WM_2_B_0)
+#define _PLANE_WM_BASE(pipe, plane)	\
+			_PLANE(plane, _PLANE_WM_1(pipe), _PLANE_WM_2(pipe))
+#define PLANE_WM(pipe, plane, level)	\
+			(_PLANE_WM_BASE(pipe, plane) + ((4) * (level)))
+#define _PLANE_WM_TRANS_1(pipe)	\
+			_PIPE(pipe, PLANE_WM_TRANS_1_A_0, PLANE_WM_TRANS_1_B_0)
+#define _PLANE_WM_TRANS_2(pipe)	\
+			_PIPE(pipe, PLANE_WM_TRANS_2_A_0, PLANE_WM_TRANS_2_B_0)
+#define PLANE_WM_TRANS(pipe, plane)	\
+		_PLANE(plane, _PLANE_WM_TRANS_1(pipe), _PLANE_WM_TRANS_2(pipe))
+
 /* define the Watermark register on Ironlake */
 #define WM0_PIPEA_ILK		0x45100
 #define  WM0_PIPE_PLANE_MASK	(0xffff<<16)
@@ -4026,7 +4250,8 @@ enum punit_power_well {
 /* Old style CUR*CNTR flags (desktop 8xx) */
 #define   CURSOR_ENABLE		0x80000000
 #define   CURSOR_GAMMA_ENABLE	0x40000000
-#define   CURSOR_STRIDE_MASK	0x30000000
+#define   CURSOR_STRIDE_SHIFT	28
+#define   CURSOR_STRIDE(x)	((ffs(x)-9) << CURSOR_STRIDE_SHIFT) /* 256,512,1k,2k */
 #define   CURSOR_PIPE_CSC_ENABLE (1<<24)
 #define   CURSOR_FORMAT_SHIFT	24
 #define   CURSOR_FORMAT_MASK	(0x07 << CURSOR_FORMAT_SHIFT)
@@ -4048,6 +4273,7 @@ enum punit_power_well {
 #define   MCURSOR_PIPE_A	0x00
 #define   MCURSOR_PIPE_B	(1 << 28)
 #define   MCURSOR_GAMMA_ENABLE  (1 << 26)
+#define   CURSOR_ROTATE_180	(1<<15)
 #define   CURSOR_TRICKLE_FEED_DISABLE	(1 << 14)
 #define _CURABASE		0x70084
 #define _CURAPOS		0x70088
@@ -4111,8 +4337,11 @@ enum punit_power_well {
 #define   DISPPLANE_NO_LINE_DOUBLE		0
 #define   DISPPLANE_STEREO_POLARITY_FIRST	0
 #define   DISPPLANE_STEREO_POLARITY_SECOND	(1<<18)
+#define   DISPPLANE_ALPHA_PREMULTIPLY		(1<<16) /* CHV pipe B */
+#define   DISPPLANE_ROTATE_180			(1<<15)
 #define   DISPPLANE_TRICKLE_FEED_DISABLE	(1<<14) /* Ironlake */
 #define   DISPPLANE_TILED			(1<<10)
+#define   DISPPLANE_MIRROR			(1<<8) /* CHV pipe B */
 #define _DSPAADDR				0x70184
 #define _DSPASTRIDE				0x70188
 #define _DSPAPOS				0x7018C /* reserved */
@@ -4133,6 +4362,24 @@ enum punit_power_well {
 #define DSPOFFSET(plane) _PIPE2(plane, _DSPAOFFSET)
 #define DSPSURFLIVE(plane) _PIPE2(plane, _DSPASURFLIVE)
 
+/* CHV pipe B blender and primary plane */
+#define _CHV_BLEND_A		0x60a00
+#define   CHV_BLEND_LEGACY		(0<<30)
+#define   CHV_BLEND_ANDROID		(1<<30)
+#define   CHV_BLEND_MPO			(2<<30)
+#define   CHV_BLEND_MASK		(3<<30)
+#define _CHV_CANVAS_A		0x60a04
+#define _PRIMPOS_A		0x60a08
+#define _PRIMSIZE_A		0x60a0c
+#define _PRIMCNSTALPHA_A	0x60a10
+#define   PRIM_CONST_ALPHA_ENABLE	(1<<31)
+
+#define CHV_BLEND(pipe) _TRANSCODER2(pipe, _CHV_BLEND_A)
+#define CHV_CANVAS(pipe) _TRANSCODER2(pipe, _CHV_CANVAS_A)
+#define PRIMPOS(plane) _TRANSCODER2(plane, _PRIMPOS_A)
+#define PRIMSIZE(plane) _TRANSCODER2(plane, _PRIMSIZE_A)
+#define PRIMCNSTALPHA(plane) _TRANSCODER2(plane, _PRIMCNSTALPHA_A)
+
 /* Display/Sprite base address macros */
 #define DISP_BASEADDR_MASK	(0xfffff000)
 #define I915_LO_DISPBASE(val)	(val & ~DISP_BASEADDR_MASK)
@@ -4195,6 +4442,7 @@ enum punit_power_well {
 #define   DVS_YUV_ORDER_UYVY	(1<<16)
 #define   DVS_YUV_ORDER_YVYU	(2<<16)
 #define   DVS_YUV_ORDER_VYUY	(3<<16)
+#define   DVS_ROTATE_180	(1<<15)
 #define   DVS_DEST_KEY		(1<<2)
 #define   DVS_TRICKLE_FEED_DISABLE (1<<14)
 #define   DVS_TILED		(1<<10)
@@ -4265,6 +4513,7 @@ enum punit_power_well {
 #define   SPRITE_YUV_ORDER_UYVY		(1<<16)
 #define   SPRITE_YUV_ORDER_YVYU		(2<<16)
 #define   SPRITE_YUV_ORDER_VYUY		(3<<16)
+#define   SPRITE_ROTATE_180		(1<<15)
 #define   SPRITE_TRICKLE_FEED_DISABLE	(1<<14)
 #define   SPRITE_INT_GAMMA_ENABLE	(1<<13)
 #define   SPRITE_TILED			(1<<10)
@@ -4332,13 +4581,16 @@ enum punit_power_well {
 #define   SP_FORMAT_RGBA1010102		(9<<26)
 #define   SP_FORMAT_RGBX8888		(0xe<<26)
 #define   SP_FORMAT_RGBA8888		(0xf<<26)
+#define   SP_ALPHA_PREMULTIPLY		(1<<23) /* CHV pipe B */
 #define   SP_SOURCE_KEY			(1<<22)
 #define   SP_YUV_BYTE_ORDER_MASK	(3<<16)
 #define   SP_YUV_ORDER_YUYV		(0<<16)
 #define   SP_YUV_ORDER_UYVY		(1<<16)
 #define   SP_YUV_ORDER_YVYU		(2<<16)
 #define   SP_YUV_ORDER_VYUY		(3<<16)
+#define   SP_ROTATE_180			(1<<15)
 #define   SP_TILED			(1<<10)
+#define   SP_MIRROR			(1<<8) /* CHV pipe B */
 #define _SPALINOFF		(VLV_DISPLAY_BASE + 0x72184)
 #define _SPASTRIDE		(VLV_DISPLAY_BASE + 0x72188)
 #define _SPAPOS			(VLV_DISPLAY_BASE + 0x7218c)
@@ -4349,6 +4601,7 @@ enum punit_power_well {
 #define _SPAKEYMAXVAL		(VLV_DISPLAY_BASE + 0x721a0)
 #define _SPATILEOFF		(VLV_DISPLAY_BASE + 0x721a4)
 #define _SPACONSTALPHA		(VLV_DISPLAY_BASE + 0x721a8)
+#define   SP_CONST_ALPHA_ENABLE		(1<<31)
 #define _SPAGAMC		(VLV_DISPLAY_BASE + 0x721f4)
 
 #define _SPBCNTR		(VLV_DISPLAY_BASE + 0x72280)
@@ -4377,6 +4630,195 @@ enum punit_power_well {
 #define SPCONSTALPHA(pipe, plane) _PIPE(pipe * 2 + plane, _SPACONSTALPHA, _SPBCONSTALPHA)
 #define SPGAMC(pipe, plane) _PIPE(pipe * 2 + plane, _SPAGAMC, _SPBGAMC)
 
+/*
+ * CHV pipe B sprite CSC
+ *
+ * |cr|   |c0 c1 c2|   |cr + cr_ioff|   |cr_ooff|
+ * |yg| = |c3 c4 c5| x |yg + yg_ioff| + |yg_ooff|
+ * |cb|   |c6 c7 c8|   |cb + cr_ioff|   |cb_ooff|
+ */
+#define SPCSCYGOFF(sprite)	(VLV_DISPLAY_BASE + 0x6d900 + (sprite) * 0x1000)
+#define SPCSCCBOFF(sprite)	(VLV_DISPLAY_BASE + 0x6d904 + (sprite) * 0x1000)
+#define SPCSCCROFF(sprite)	(VLV_DISPLAY_BASE + 0x6d908 + (sprite) * 0x1000)
+#define  SPCSC_OOFF(x)		(((x) & 0x7ff) << 16) /* s11 */
+#define  SPCSC_IOFF(x)		(((x) & 0x7ff) << 0) /* s11 */
+
+#define SPCSCC01(sprite)	(VLV_DISPLAY_BASE + 0x6d90c + (sprite) * 0x1000)
+#define SPCSCC23(sprite)	(VLV_DISPLAY_BASE + 0x6d910 + (sprite) * 0x1000)
+#define SPCSCC45(sprite)	(VLV_DISPLAY_BASE + 0x6d914 + (sprite) * 0x1000)
+#define SPCSCC67(sprite)	(VLV_DISPLAY_BASE + 0x6d918 + (sprite) * 0x1000)
+#define SPCSCC8(sprite)		(VLV_DISPLAY_BASE + 0x6d91c + (sprite) * 0x1000)
+#define  SPCSC_C1(x)		(((x) & 0x7fff) << 16) /* s3.12 */
+#define  SPCSC_C0(x)		(((x) & 0x7fff) << 0) /* s3.12 */
+
+#define SPCSCYGICLAMP(sprite)	(VLV_DISPLAY_BASE + 0x6d920 + (sprite) * 0x1000)
+#define SPCSCCBICLAMP(sprite)	(VLV_DISPLAY_BASE + 0x6d924 + (sprite) * 0x1000)
+#define SPCSCCRICLAMP(sprite)	(VLV_DISPLAY_BASE + 0x6d928 + (sprite) * 0x1000)
+#define  SPCSC_IMAX(x)		(((x) & 0x7ff) << 16) /* s11 */
+#define  SPCSC_IMIN(x)		(((x) & 0x7ff) << 0) /* s11 */
+
+#define SPCSCYGOCLAMP(sprite)	(VLV_DISPLAY_BASE + 0x6d92c + (sprite) * 0x1000)
+#define SPCSCCBOCLAMP(sprite)	(VLV_DISPLAY_BASE + 0x6d930 + (sprite) * 0x1000)
+#define SPCSCCROCLAMP(sprite)	(VLV_DISPLAY_BASE + 0x6d934 + (sprite) * 0x1000)
+#define  SPCSC_OMAX(x)		((x) << 16) /* u10 */
+#define  SPCSC_OMIN(x)		((x) << 0) /* u10 */
+
+/* Skylake plane registers */
+
+#define _PLANE_CTL_1_A				0x70180
+#define _PLANE_CTL_2_A				0x70280
+#define _PLANE_CTL_3_A				0x70380
+#define   PLANE_CTL_ENABLE			(1 << 31)
+#define   PLANE_CTL_PIPE_GAMMA_ENABLE		(1 << 30)
+#define   PLANE_CTL_FORMAT_MASK			(0xf << 24)
+#define   PLANE_CTL_FORMAT_YUV422		(  0 << 24)
+#define   PLANE_CTL_FORMAT_NV12			(  1 << 24)
+#define   PLANE_CTL_FORMAT_XRGB_2101010		(  2 << 24)
+#define   PLANE_CTL_FORMAT_XRGB_8888		(  4 << 24)
+#define   PLANE_CTL_FORMAT_XRGB_16161616F	(  6 << 24)
+#define   PLANE_CTL_FORMAT_AYUV			(  8 << 24)
+#define   PLANE_CTL_FORMAT_INDEXED		( 12 << 24)
+#define   PLANE_CTL_FORMAT_RGB_565		( 14 << 24)
+#define   PLANE_CTL_PIPE_CSC_ENABLE		(1 << 23)
+#define   PLANE_CTL_KEY_ENABLE_MASK		(0x3 << 21)
+#define   PLANE_CTL_KEY_ENABLE_SOURCE		(  1 << 21)
+#define   PLANE_CTL_KEY_ENABLE_DESTINATION	(  2 << 21)
+#define   PLANE_CTL_ORDER_BGRX			(0 << 20)
+#define   PLANE_CTL_ORDER_RGBX			(1 << 20)
+#define   PLANE_CTL_YUV422_ORDER_MASK		(0x3 << 16)
+#define   PLANE_CTL_YUV422_YUYV			(  0 << 16)
+#define   PLANE_CTL_YUV422_UYVY			(  1 << 16)
+#define   PLANE_CTL_YUV422_YVYU			(  2 << 16)
+#define   PLANE_CTL_YUV422_VYUY			(  3 << 16)
+#define   PLANE_CTL_DECOMPRESSION_ENABLE	(1 << 15)
+#define   PLANE_CTL_TRICKLE_FEED_DISABLE	(1 << 14)
+#define   PLANE_CTL_PLANE_GAMMA_DISABLE		(1 << 13)
+#define   PLANE_CTL_TILED_MASK			(0x7 << 10)
+#define   PLANE_CTL_TILED_LINEAR		(  0 << 10)
+#define   PLANE_CTL_TILED_X			(  1 << 10)
+#define   PLANE_CTL_TILED_Y			(  4 << 10)
+#define   PLANE_CTL_TILED_YF			(  5 << 10)
+#define   PLANE_CTL_ALPHA_MASK			(0x3 << 4)
+#define   PLANE_CTL_ALPHA_DISABLE		(  0 << 4)
+#define   PLANE_CTL_ALPHA_SW_PREMULTIPLY	(  2 << 4)
+#define   PLANE_CTL_ALPHA_HW_PREMULTIPLY	(  3 << 4)
+#define   PLANE_CTL_ROTATE_MASK			0x3
+#define   PLANE_CTL_ROTATE_0			0x0
+#define   PLANE_CTL_ROTATE_180			0x2
+#define _PLANE_STRIDE_1_A			0x70188
+#define _PLANE_STRIDE_2_A			0x70288
+#define _PLANE_STRIDE_3_A			0x70388
+#define _PLANE_POS_1_A				0x7018c
+#define _PLANE_POS_2_A				0x7028c
+#define _PLANE_POS_3_A				0x7038c
+#define _PLANE_SIZE_1_A				0x70190
+#define _PLANE_SIZE_2_A				0x70290
+#define _PLANE_SIZE_3_A				0x70390
+#define _PLANE_SURF_1_A				0x7019c
+#define _PLANE_SURF_2_A				0x7029c
+#define _PLANE_SURF_3_A				0x7039c
+#define _PLANE_OFFSET_1_A			0x701a4
+#define _PLANE_OFFSET_2_A			0x702a4
+#define _PLANE_OFFSET_3_A			0x703a4
+#define _PLANE_KEYVAL_1_A			0x70194
+#define _PLANE_KEYVAL_2_A			0x70294
+#define _PLANE_KEYMSK_1_A			0x70198
+#define _PLANE_KEYMSK_2_A			0x70298
+#define _PLANE_KEYMAX_1_A			0x701a0
+#define _PLANE_KEYMAX_2_A			0x702a0
+#define _PLANE_BUF_CFG_1_A			0x7027c
+#define _PLANE_BUF_CFG_2_A			0x7037c
+
+#define _PLANE_CTL_1_B				0x71180
+#define _PLANE_CTL_2_B				0x71280
+#define _PLANE_CTL_3_B				0x71380
+#define _PLANE_CTL_1(pipe)	_PIPE(pipe, _PLANE_CTL_1_A, _PLANE_CTL_1_B)
+#define _PLANE_CTL_2(pipe)	_PIPE(pipe, _PLANE_CTL_2_A, _PLANE_CTL_2_B)
+#define _PLANE_CTL_3(pipe)	_PIPE(pipe, _PLANE_CTL_3_A, _PLANE_CTL_3_B)
+#define PLANE_CTL(pipe, plane)	\
+	_PLANE(plane, _PLANE_CTL_1(pipe), _PLANE_CTL_2(pipe))
+
+#define _PLANE_STRIDE_1_B			0x71188
+#define _PLANE_STRIDE_2_B			0x71288
+#define _PLANE_STRIDE_3_B			0x71388
+#define _PLANE_STRIDE_1(pipe)	\
+	_PIPE(pipe, _PLANE_STRIDE_1_A, _PLANE_STRIDE_1_B)
+#define _PLANE_STRIDE_2(pipe)	\
+	_PIPE(pipe, _PLANE_STRIDE_2_A, _PLANE_STRIDE_2_B)
+#define _PLANE_STRIDE_3(pipe)	\
+	_PIPE(pipe, _PLANE_STRIDE_3_A, _PLANE_STRIDE_3_B)
+#define PLANE_STRIDE(pipe, plane)	\
+	_PLANE(plane, _PLANE_STRIDE_1(pipe), _PLANE_STRIDE_2(pipe))
+
+#define _PLANE_POS_1_B				0x7118c
+#define _PLANE_POS_2_B				0x7128c
+#define _PLANE_POS_3_B				0x7138c
+#define _PLANE_POS_1(pipe)	_PIPE(pipe, _PLANE_POS_1_A, _PLANE_POS_1_B)
+#define _PLANE_POS_2(pipe)	_PIPE(pipe, _PLANE_POS_2_A, _PLANE_POS_2_B)
+#define _PLANE_POS_3(pipe)	_PIPE(pipe, _PLANE_POS_3_A, _PLANE_POS_3_B)
+#define PLANE_POS(pipe, plane)	\
+	_PLANE(plane, _PLANE_POS_1(pipe), _PLANE_POS_2(pipe))
+
+#define _PLANE_SIZE_1_B				0x71190
+#define _PLANE_SIZE_2_B				0x71290
+#define _PLANE_SIZE_3_B				0x71390
+#define _PLANE_SIZE_1(pipe)	_PIPE(pipe, _PLANE_SIZE_1_A, _PLANE_SIZE_1_B)
+#define _PLANE_SIZE_2(pipe)	_PIPE(pipe, _PLANE_SIZE_2_A, _PLANE_SIZE_2_B)
+#define _PLANE_SIZE_3(pipe)	_PIPE(pipe, _PLANE_SIZE_3_A, _PLANE_SIZE_3_B)
+#define PLANE_SIZE(pipe, plane)	\
+	_PLANE(plane, _PLANE_SIZE_1(pipe), _PLANE_SIZE_2(pipe))
+
+#define _PLANE_SURF_1_B				0x7119c
+#define _PLANE_SURF_2_B				0x7129c
+#define _PLANE_SURF_3_B				0x7139c
+#define _PLANE_SURF_1(pipe)	_PIPE(pipe, _PLANE_SURF_1_A, _PLANE_SURF_1_B)
+#define _PLANE_SURF_2(pipe)	_PIPE(pipe, _PLANE_SURF_2_A, _PLANE_SURF_2_B)
+#define _PLANE_SURF_3(pipe)	_PIPE(pipe, _PLANE_SURF_3_A, _PLANE_SURF_3_B)
+#define PLANE_SURF(pipe, plane)	\
+	_PLANE(plane, _PLANE_SURF_1(pipe), _PLANE_SURF_2(pipe))
+
+#define _PLANE_OFFSET_1_B			0x711a4
+#define _PLANE_OFFSET_2_B			0x712a4
+#define _PLANE_OFFSET_1(pipe) _PIPE(pipe, _PLANE_OFFSET_1_A, _PLANE_OFFSET_1_B)
+#define _PLANE_OFFSET_2(pipe) _PIPE(pipe, _PLANE_OFFSET_2_A, _PLANE_OFFSET_2_B)
+#define PLANE_OFFSET(pipe, plane)	\
+	_PLANE(plane, _PLANE_OFFSET_1(pipe), _PLANE_OFFSET_2(pipe))
+
+#define _PLANE_KEYVAL_1_B			0x71194
+#define _PLANE_KEYVAL_2_B			0x71294
+#define _PLANE_KEYVAL_1(pipe) _PIPE(pipe, _PLANE_KEYVAL_1_A, _PLANE_KEYVAL_1_B)
+#define _PLANE_KEYVAL_2(pipe) _PIPE(pipe, _PLANE_KEYVAL_2_A, _PLANE_KEYVAL_2_B)
+#define PLANE_KEYVAL(pipe, plane)	\
+	_PLANE(plane, _PLANE_KEYVAL_1(pipe), _PLANE_KEYVAL_2(pipe))
+
+#define _PLANE_KEYMSK_1_B			0x71198
+#define _PLANE_KEYMSK_2_B			0x71298
+#define _PLANE_KEYMSK_1(pipe) _PIPE(pipe, _PLANE_KEYMSK_1_A, _PLANE_KEYMSK_1_B)
+#define _PLANE_KEYMSK_2(pipe) _PIPE(pipe, _PLANE_KEYMSK_2_A, _PLANE_KEYMSK_2_B)
+#define PLANE_KEYMSK(pipe, plane)	\
+	_PLANE(plane, _PLANE_KEYMSK_1(pipe), _PLANE_KEYMSK_2(pipe))
+
+#define _PLANE_KEYMAX_1_B			0x711a0
+#define _PLANE_KEYMAX_2_B			0x712a0
+#define _PLANE_KEYMAX_1(pipe) _PIPE(pipe, _PLANE_KEYMAX_1_A, _PLANE_KEYMAX_1_B)
+#define _PLANE_KEYMAX_2(pipe) _PIPE(pipe, _PLANE_KEYMAX_2_A, _PLANE_KEYMAX_2_B)
+#define PLANE_KEYMAX(pipe, plane)	\
+	_PLANE(plane, _PLANE_KEYMAX_1(pipe), _PLANE_KEYMAX_2(pipe))
+
+#define _PLANE_BUF_CFG_1_B			0x7127c
+#define _PLANE_BUF_CFG_2_B			0x7137c
+#define _PLANE_BUF_CFG_1(pipe)	\
+	_PIPE(pipe, _PLANE_BUF_CFG_1_A, _PLANE_BUF_CFG_1_B)
+#define _PLANE_BUF_CFG_2(pipe)	\
+	_PIPE(pipe, _PLANE_BUF_CFG_2_A, _PLANE_BUF_CFG_2_B)
+#define PLANE_BUF_CFG(pipe, plane)	\
+	_PLANE(plane, _PLANE_BUF_CFG_1(pipe), _PLANE_BUF_CFG_2(pipe))
+
+/* SKL new cursor registers */
+#define _CUR_BUF_CFG_A				0x7017c
+#define _CUR_BUF_CFG_B				0x7117c
+#define CUR_BUF_CFG(pipe)	_PIPE(pipe, _CUR_BUF_CFG_A, _CUR_BUF_CFG_B)
+
 /* VBIOS regs */
 #define VGACNTRL		0x71400
 # define VGA_DISP_DISABLE			(1 << 31)
@@ -4492,6 +4934,18 @@ enum punit_power_well {
 #define PF_VSCALE(pipe)		_PIPE(pipe, _PFA_VSCALE, _PFB_VSCALE)
 #define PF_HSCALE(pipe)		_PIPE(pipe, _PFA_HSCALE, _PFB_HSCALE)
 
+#define _PSA_CTL		0x68180
+#define _PSB_CTL		0x68980
+#define PS_ENABLE		(1<<31)
+#define _PSA_WIN_SZ		0x68174
+#define _PSB_WIN_SZ		0x68974
+#define _PSA_WIN_POS		0x68170
+#define _PSB_WIN_POS		0x68970
+
+#define PS_CTL(pipe)		_PIPE(pipe, _PSA_CTL, _PSB_CTL)
+#define PS_WIN_SZ(pipe)		_PIPE(pipe, _PSA_WIN_SZ, _PSB_WIN_SZ)
+#define PS_WIN_POS(pipe)	_PIPE(pipe, _PSA_WIN_POS, _PSB_WIN_POS)
+
 /* legacy palette */
 #define _LGC_PALETTE_A           0x4a000
 #define _LGC_PALETTE_B           0x4a800
@@ -4613,16 +5067,32 @@ enum punit_power_well {
 #define  GEN8_PIPE_SCAN_LINE_EVENT	(1 << 2)
 #define  GEN8_PIPE_VSYNC		(1 << 1)
 #define  GEN8_PIPE_VBLANK		(1 << 0)
+#define  GEN9_PIPE_CURSOR_FAULT		(1 << 11)
+#define  GEN9_PIPE_PLANE3_FAULT		(1 << 9)
+#define  GEN9_PIPE_PLANE2_FAULT		(1 << 8)
+#define  GEN9_PIPE_PLANE1_FAULT		(1 << 7)
+#define  GEN9_PIPE_PLANE3_FLIP_DONE	(1 << 5)
+#define  GEN9_PIPE_PLANE2_FLIP_DONE	(1 << 4)
+#define  GEN9_PIPE_PLANE1_FLIP_DONE	(1 << 3)
+#define  GEN9_PIPE_PLANE_FLIP_DONE(p)	(1 << (3 + p))
 #define GEN8_DE_PIPE_IRQ_FAULT_ERRORS \
 	(GEN8_PIPE_CURSOR_FAULT | \
 	 GEN8_PIPE_SPRITE_FAULT | \
 	 GEN8_PIPE_PRIMARY_FAULT)
+#define GEN9_DE_PIPE_IRQ_FAULT_ERRORS \
+	(GEN9_PIPE_CURSOR_FAULT | \
+	 GEN9_PIPE_PLANE3_FAULT | \
+	 GEN9_PIPE_PLANE2_FAULT | \
+	 GEN9_PIPE_PLANE1_FAULT)
 
 #define GEN8_DE_PORT_ISR 0x44440
 #define GEN8_DE_PORT_IMR 0x44444
 #define GEN8_DE_PORT_IIR 0x44448
 #define GEN8_DE_PORT_IER 0x4444c
 #define  GEN8_PORT_DP_A_HOTPLUG		(1 << 3)
+#define  GEN9_AUX_CHANNEL_D		(1 << 27)
+#define  GEN9_AUX_CHANNEL_C		(1 << 26)
+#define  GEN9_AUX_CHANNEL_B		(1 << 25)
 #define  GEN8_AUX_CHANNEL_A		(1 << 0)
 
 #define GEN8_DE_MISC_ISR 0x44460
@@ -4706,6 +5176,8 @@ enum punit_power_well {
 /* GEN8 chicken */
 #define HDC_CHICKEN0				0x7300
 #define  HDC_FORCE_NON_COHERENT			(1<<4)
+#define  HDC_DONOT_FETCH_MEM_WHEN_MASKED	(1<<11)
+#define  HDC_FENCE_DEST_SLM_DISABLE		(1<<14)
 
 /* WaCatErrorRejectionIssue */
 #define GEN7_SQ_CHICKEN_MBCUNIT_CONFIG		0x9030
@@ -5246,8 +5718,7 @@ enum punit_power_well {
 #define PIPEA_PP_STATUS         (VLV_DISPLAY_BASE + 0x61200)
 #define PIPEA_PP_CONTROL        (VLV_DISPLAY_BASE + 0x61204)
 #define PIPEA_PP_ON_DELAYS      (VLV_DISPLAY_BASE + 0x61208)
-#define  PANEL_PORT_SELECT_DPB_VLV	(1 << 30)
-#define  PANEL_PORT_SELECT_DPC_VLV	(2 << 30)
+#define  PANEL_PORT_SELECT_VLV(port)	((port) << 30)
 #define PIPEA_PP_OFF_DELAYS     (VLV_DISPLAY_BASE + 0x6120c)
 #define PIPEA_PP_DIVISOR        (VLV_DISPLAY_BASE + 0x61210)
 
@@ -5407,8 +5878,13 @@ enum punit_power_well {
 #define   VLV_GTLC_ALLOWWAKEERR			(1 << 1)
 #define   VLV_GTLC_PW_MEDIA_STATUS_MASK		(1 << 5)
 #define   VLV_GTLC_PW_RENDER_STATUS_MASK	(1 << 7)
-#define VLV_GTLC_SURVIVABILITY_REG              0x130098
 #define  FORCEWAKE_MT				0xa188 /* multi-threaded */
+#define  FORCEWAKE_MEDIA_GEN9			0xa270
+#define  FORCEWAKE_RENDER_GEN9			0xa278
+#define  FORCEWAKE_BLITTER_GEN9			0xa188
+#define  FORCEWAKE_ACK_MEDIA_GEN9		0x0D88
+#define  FORCEWAKE_ACK_RENDER_GEN9		0x0D84
+#define  FORCEWAKE_ACK_BLITTER_GEN9		0x130044
 #define   FORCEWAKE_KERNEL			0x1
 #define   FORCEWAKE_USER			0x2
 #define  FORCEWAKE_MT_ACK			0x130040
@@ -5545,12 +6021,6 @@ enum punit_power_well {
 						 GEN6_PM_RP_DOWN_THRESHOLD | \
 						 GEN6_PM_RP_DOWN_TIMEOUT)
 
-#define CHV_CZ_CLOCK_FREQ_MODE_200			200
-#define CHV_CZ_CLOCK_FREQ_MODE_267			267
-#define CHV_CZ_CLOCK_FREQ_MODE_320			320
-#define CHV_CZ_CLOCK_FREQ_MODE_333			333
-#define CHV_CZ_CLOCK_FREQ_MODE_400			400
-
 #define GEN7_GT_SCRATCH_BASE			0x4F100
 #define GEN7_GT_SCRATCH_REG_NUM			8
 
@@ -5586,9 +6056,17 @@ enum punit_power_well {
 #define   GEN6_ENCODE_RC6_VID(mv)		(((mv) - 245) / 5)
 #define   GEN6_DECODE_RC6_VID(vids)		(((vids) * 5) + 245)
 #define   DISPLAY_IPS_CONTROL			0x19
+#define	  HSW_PCODE_DYNAMIC_DUTY_CYCLE_CONTROL	0x1A
 #define GEN6_PCODE_DATA				0x138128
 #define   GEN6_PCODE_FREQ_IA_RATIO_SHIFT	8
 #define   GEN6_PCODE_FREQ_RING_RATIO_SHIFT	16
+#define GEN6_PCODE_DATA1			0x13812C
+
+#define   GEN9_PCODE_READ_MEM_LATENCY		0x6
+#define   GEN9_MEM_LATENCY_LEVEL_MASK		0xFF
+#define   GEN9_MEM_LATENCY_LEVEL_1_5_SHIFT	8
+#define   GEN9_MEM_LATENCY_LEVEL_2_6_SHIFT	16
+#define   GEN9_MEM_LATENCY_LEVEL_3_7_SHIFT	24
 
 #define GEN6_GT_CORE_STATUS		0x138060
 #define   GEN6_CORE_CPD_STATE_MASK	(7<<4)
@@ -5626,6 +6104,9 @@ enum punit_power_well {
 #define   GEN7_SINGLE_SUBSCAN_DISPATCH_ENABLE	(1<<10)
 #define   GEN7_PSD_SINGLE_PORT_DISPATCH_ENABLE	(1<<3)
 
+#define GEN9_HALF_SLICE_CHICKEN5	0xe188
+#define   GEN9_DG_MIRROR_FIX_ENABLE	(1<<5)
+
 #define GEN8_ROW_CHICKEN		0xe4f0
 #define   PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE	(1<<8)
 #define   STALL_DOP_GATING_DISABLE		(1<<5)
@@ -5641,57 +6122,58 @@ enum punit_power_well {
 #define   GEN8_CENTROID_PIXEL_OPT_DIS	(1<<8)
 #define   GEN8_SAMPLER_POWER_BYPASS_DIS	(1<<1)
 
+/* Audio */
 #define G4X_AUD_VID_DID			(dev_priv->info.display_mmio_offset + 0x62020)
-#define INTEL_AUDIO_DEVCL		0x808629FB
-#define INTEL_AUDIO_DEVBLC		0x80862801
-#define INTEL_AUDIO_DEVCTG		0x80862802
+#define   INTEL_AUDIO_DEVCL		0x808629FB
+#define   INTEL_AUDIO_DEVBLC		0x80862801
+#define   INTEL_AUDIO_DEVCTG		0x80862802
 
 #define G4X_AUD_CNTL_ST			0x620B4
-#define G4X_ELDV_DEVCL_DEVBLC		(1 << 13)
-#define G4X_ELDV_DEVCTG			(1 << 14)
-#define G4X_ELD_ADDR			(0xf << 5)
-#define G4X_ELD_ACK			(1 << 4)
+#define   G4X_ELDV_DEVCL_DEVBLC		(1 << 13)
+#define   G4X_ELDV_DEVCTG		(1 << 14)
+#define   G4X_ELD_ADDR_MASK		(0xf << 5)
+#define   G4X_ELD_ACK			(1 << 4)
 #define G4X_HDMIW_HDMIEDID		0x6210C
 
-#define IBX_HDMIW_HDMIEDID_A		0xE2050
-#define IBX_HDMIW_HDMIEDID_B		0xE2150
+#define _IBX_HDMIW_HDMIEDID_A		0xE2050
+#define _IBX_HDMIW_HDMIEDID_B		0xE2150
 #define IBX_HDMIW_HDMIEDID(pipe) _PIPE(pipe, \
-					IBX_HDMIW_HDMIEDID_A, \
-					IBX_HDMIW_HDMIEDID_B)
-#define IBX_AUD_CNTL_ST_A		0xE20B4
-#define IBX_AUD_CNTL_ST_B		0xE21B4
+					_IBX_HDMIW_HDMIEDID_A, \
+					_IBX_HDMIW_HDMIEDID_B)
+#define _IBX_AUD_CNTL_ST_A		0xE20B4
+#define _IBX_AUD_CNTL_ST_B		0xE21B4
 #define IBX_AUD_CNTL_ST(pipe) _PIPE(pipe, \
-					IBX_AUD_CNTL_ST_A, \
-					IBX_AUD_CNTL_ST_B)
-#define IBX_ELD_BUFFER_SIZE		(0x1f << 10)
-#define IBX_ELD_ADDRESS			(0x1f << 5)
-#define IBX_ELD_ACK			(1 << 4)
+					_IBX_AUD_CNTL_ST_A, \
+					_IBX_AUD_CNTL_ST_B)
+#define   IBX_ELD_BUFFER_SIZE_MASK	(0x1f << 10)
+#define   IBX_ELD_ADDRESS_MASK		(0x1f << 5)
+#define   IBX_ELD_ACK			(1 << 4)
 #define IBX_AUD_CNTL_ST2		0xE20C0
-#define IBX_ELD_VALIDB			(1 << 0)
-#define IBX_CP_READYB			(1 << 1)
+#define   IBX_CP_READY(port)		((1 << 1) << (((port) - 1) * 4))
+#define   IBX_ELD_VALID(port)		((1 << 0) << (((port) - 1) * 4))
 
-#define CPT_HDMIW_HDMIEDID_A		0xE5050
-#define CPT_HDMIW_HDMIEDID_B		0xE5150
+#define _CPT_HDMIW_HDMIEDID_A		0xE5050
+#define _CPT_HDMIW_HDMIEDID_B		0xE5150
 #define CPT_HDMIW_HDMIEDID(pipe) _PIPE(pipe, \
-					CPT_HDMIW_HDMIEDID_A, \
-					CPT_HDMIW_HDMIEDID_B)
-#define CPT_AUD_CNTL_ST_A		0xE50B4
-#define CPT_AUD_CNTL_ST_B		0xE51B4
+					_CPT_HDMIW_HDMIEDID_A, \
+					_CPT_HDMIW_HDMIEDID_B)
+#define _CPT_AUD_CNTL_ST_A		0xE50B4
+#define _CPT_AUD_CNTL_ST_B		0xE51B4
 #define CPT_AUD_CNTL_ST(pipe) _PIPE(pipe, \
-					CPT_AUD_CNTL_ST_A, \
-					CPT_AUD_CNTL_ST_B)
+					_CPT_AUD_CNTL_ST_A, \
+					_CPT_AUD_CNTL_ST_B)
 #define CPT_AUD_CNTRL_ST2		0xE50C0
 
-#define VLV_HDMIW_HDMIEDID_A		(VLV_DISPLAY_BASE + 0x62050)
-#define VLV_HDMIW_HDMIEDID_B		(VLV_DISPLAY_BASE + 0x62150)
+#define _VLV_HDMIW_HDMIEDID_A		(VLV_DISPLAY_BASE + 0x62050)
+#define _VLV_HDMIW_HDMIEDID_B		(VLV_DISPLAY_BASE + 0x62150)
 #define VLV_HDMIW_HDMIEDID(pipe) _PIPE(pipe, \
-					VLV_HDMIW_HDMIEDID_A, \
-					VLV_HDMIW_HDMIEDID_B)
-#define VLV_AUD_CNTL_ST_A		(VLV_DISPLAY_BASE + 0x620B4)
-#define VLV_AUD_CNTL_ST_B		(VLV_DISPLAY_BASE + 0x621B4)
+					_VLV_HDMIW_HDMIEDID_A, \
+					_VLV_HDMIW_HDMIEDID_B)
+#define _VLV_AUD_CNTL_ST_A		(VLV_DISPLAY_BASE + 0x620B4)
+#define _VLV_AUD_CNTL_ST_B		(VLV_DISPLAY_BASE + 0x621B4)
 #define VLV_AUD_CNTL_ST(pipe) _PIPE(pipe, \
-					VLV_AUD_CNTL_ST_A, \
-					VLV_AUD_CNTL_ST_B)
+					_VLV_AUD_CNTL_ST_A, \
+					_VLV_AUD_CNTL_ST_B)
 #define VLV_AUD_CNTL_ST2		(VLV_DISPLAY_BASE + 0x620C0)
 
 /* These are the 4 32-bit write offset registers for each stream
@@ -5700,28 +6182,28 @@ enum punit_power_well {
  */
 #define GEN7_SO_WRITE_OFFSET(n)		(0x5280 + (n) * 4)
 
-#define IBX_AUD_CONFIG_A			0xe2000
-#define IBX_AUD_CONFIG_B			0xe2100
+#define _IBX_AUD_CONFIG_A		0xe2000
+#define _IBX_AUD_CONFIG_B		0xe2100
 #define IBX_AUD_CFG(pipe) _PIPE(pipe, \
-					IBX_AUD_CONFIG_A, \
-					IBX_AUD_CONFIG_B)
-#define CPT_AUD_CONFIG_A			0xe5000
-#define CPT_AUD_CONFIG_B			0xe5100
+					_IBX_AUD_CONFIG_A, \
+					_IBX_AUD_CONFIG_B)
+#define _CPT_AUD_CONFIG_A		0xe5000
+#define _CPT_AUD_CONFIG_B		0xe5100
 #define CPT_AUD_CFG(pipe) _PIPE(pipe, \
-					CPT_AUD_CONFIG_A, \
-					CPT_AUD_CONFIG_B)
-#define VLV_AUD_CONFIG_A		(VLV_DISPLAY_BASE + 0x62000)
-#define VLV_AUD_CONFIG_B		(VLV_DISPLAY_BASE + 0x62100)
+					_CPT_AUD_CONFIG_A, \
+					_CPT_AUD_CONFIG_B)
+#define _VLV_AUD_CONFIG_A		(VLV_DISPLAY_BASE + 0x62000)
+#define _VLV_AUD_CONFIG_B		(VLV_DISPLAY_BASE + 0x62100)
 #define VLV_AUD_CFG(pipe) _PIPE(pipe, \
-					VLV_AUD_CONFIG_A, \
-					VLV_AUD_CONFIG_B)
+					_VLV_AUD_CONFIG_A, \
+					_VLV_AUD_CONFIG_B)
 
 #define   AUD_CONFIG_N_VALUE_INDEX		(1 << 29)
 #define   AUD_CONFIG_N_PROG_ENABLE		(1 << 28)
 #define   AUD_CONFIG_UPPER_N_SHIFT		20
-#define   AUD_CONFIG_UPPER_N_VALUE		(0xff << 20)
+#define   AUD_CONFIG_UPPER_N_MASK		(0xff << 20)
 #define   AUD_CONFIG_LOWER_N_SHIFT		4
-#define   AUD_CONFIG_LOWER_N_VALUE		(0xfff << 4)
+#define   AUD_CONFIG_LOWER_N_MASK		(0xfff << 4)
 #define   AUD_CONFIG_PIXEL_CLOCK_HDMI_SHIFT	16
 #define   AUD_CONFIG_PIXEL_CLOCK_HDMI_MASK	(0xf << 16)
 #define   AUD_CONFIG_PIXEL_CLOCK_HDMI_25175	(0 << 16)
@@ -5737,52 +6219,44 @@ enum punit_power_well {
 #define   AUD_CONFIG_DISABLE_NCTS		(1 << 3)
 
 /* HSW Audio */
-#define   HSW_AUD_CONFIG_A		0x65000 /* Audio Configuration Transcoder A */
-#define   HSW_AUD_CONFIG_B		0x65100 /* Audio Configuration Transcoder B */
-#define   HSW_AUD_CFG(pipe) _PIPE(pipe, \
-					HSW_AUD_CONFIG_A, \
-					HSW_AUD_CONFIG_B)
-
-#define   HSW_AUD_MISC_CTRL_A		0x65010 /* Audio Misc Control Convert 1 */
-#define   HSW_AUD_MISC_CTRL_B		0x65110 /* Audio Misc Control Convert 2 */
-#define   HSW_AUD_MISC_CTRL(pipe) _PIPE(pipe, \
-					HSW_AUD_MISC_CTRL_A, \
-					HSW_AUD_MISC_CTRL_B)
-
-#define   HSW_AUD_DIP_ELD_CTRL_ST_A	0x650b4 /* Audio DIP and ELD Control State Transcoder A */
-#define   HSW_AUD_DIP_ELD_CTRL_ST_B	0x651b4 /* Audio DIP and ELD Control State Transcoder B */
-#define   HSW_AUD_DIP_ELD_CTRL(pipe) _PIPE(pipe, \
-					HSW_AUD_DIP_ELD_CTRL_ST_A, \
-					HSW_AUD_DIP_ELD_CTRL_ST_B)
+#define _HSW_AUD_CONFIG_A		0x65000
+#define _HSW_AUD_CONFIG_B		0x65100
+#define HSW_AUD_CFG(pipe) _PIPE(pipe, \
+					_HSW_AUD_CONFIG_A, \
+					_HSW_AUD_CONFIG_B)
+
+#define _HSW_AUD_MISC_CTRL_A		0x65010
+#define _HSW_AUD_MISC_CTRL_B		0x65110
+#define HSW_AUD_MISC_CTRL(pipe) _PIPE(pipe, \
+					_HSW_AUD_MISC_CTRL_A, \
+					_HSW_AUD_MISC_CTRL_B)
+
+#define _HSW_AUD_DIP_ELD_CTRL_ST_A	0x650b4
+#define _HSW_AUD_DIP_ELD_CTRL_ST_B	0x651b4
+#define HSW_AUD_DIP_ELD_CTRL(pipe) _PIPE(pipe, \
+					_HSW_AUD_DIP_ELD_CTRL_ST_A, \
+					_HSW_AUD_DIP_ELD_CTRL_ST_B)
 
 /* Audio Digital Converter */
-#define   HSW_AUD_DIG_CNVT_1		0x65080 /* Audio Converter 1 */
-#define   HSW_AUD_DIG_CNVT_2		0x65180 /* Audio Converter 1 */
-#define   AUD_DIG_CNVT(pipe) _PIPE(pipe, \
-					HSW_AUD_DIG_CNVT_1, \
-					HSW_AUD_DIG_CNVT_2)
-#define   DIP_PORT_SEL_MASK		0x3
-
-#define   HSW_AUD_EDID_DATA_A		0x65050
-#define   HSW_AUD_EDID_DATA_B		0x65150
-#define   HSW_AUD_EDID_DATA(pipe) _PIPE(pipe, \
-					HSW_AUD_EDID_DATA_A, \
-					HSW_AUD_EDID_DATA_B)
-
-#define   HSW_AUD_PIPE_CONV_CFG		0x6507c /* Audio pipe and converter configs */
-#define   HSW_AUD_PIN_ELD_CP_VLD	0x650c0 /* Audio ELD and CP Ready Status */
-#define   AUDIO_INACTIVE_C		(1<<11)
-#define   AUDIO_INACTIVE_B		(1<<7)
-#define   AUDIO_INACTIVE_A		(1<<3)
-#define   AUDIO_OUTPUT_ENABLE_A		(1<<2)
-#define   AUDIO_OUTPUT_ENABLE_B		(1<<6)
-#define   AUDIO_OUTPUT_ENABLE_C		(1<<10)
-#define   AUDIO_ELD_VALID_A		(1<<0)
-#define   AUDIO_ELD_VALID_B		(1<<4)
-#define   AUDIO_ELD_VALID_C		(1<<8)
-#define   AUDIO_CP_READY_A		(1<<1)
-#define   AUDIO_CP_READY_B		(1<<5)
-#define   AUDIO_CP_READY_C		(1<<9)
+#define _HSW_AUD_DIG_CNVT_1		0x65080
+#define _HSW_AUD_DIG_CNVT_2		0x65180
+#define AUD_DIG_CNVT(pipe) _PIPE(pipe, \
+					_HSW_AUD_DIG_CNVT_1, \
+					_HSW_AUD_DIG_CNVT_2)
+#define DIP_PORT_SEL_MASK		0x3
+
+#define _HSW_AUD_EDID_DATA_A		0x65050
+#define _HSW_AUD_EDID_DATA_B		0x65150
+#define HSW_AUD_EDID_DATA(pipe) _PIPE(pipe, \
+					_HSW_AUD_EDID_DATA_A, \
+					_HSW_AUD_EDID_DATA_B)
+
+#define HSW_AUD_PIPE_CONV_CFG		0x6507c
+#define HSW_AUD_PIN_ELD_CP_VLD		0x650c0
+#define   AUDIO_INACTIVE(trans)		((1 << 3) << ((trans) * 4))
+#define   AUDIO_OUTPUT_ENABLE(trans)	((1 << 2) << ((trans) * 4))
+#define   AUDIO_CP_READY(trans)		((1 << 1) << ((trans) * 4))
+#define   AUDIO_ELD_VALID(trans)	((1 << 0) << ((trans) * 4))
 
 /* HSW Power Wells */
 #define HSW_PWR_WELL_BIOS			0x45400 /* CTL1 */
@@ -5866,15 +6340,7 @@ enum punit_power_well {
 #define DDI_BUF_CTL_B				0x64100
 #define DDI_BUF_CTL(port) _PORT(port, DDI_BUF_CTL_A, DDI_BUF_CTL_B)
 #define  DDI_BUF_CTL_ENABLE			(1<<31)
-#define  DDI_BUF_EMP_400MV_0DB_HSW		(0<<24)   /* Sel0 */
-#define  DDI_BUF_EMP_400MV_3_5DB_HSW		(1<<24)   /* Sel1 */
-#define  DDI_BUF_EMP_400MV_6DB_HSW		(2<<24)   /* Sel2 */
-#define  DDI_BUF_EMP_400MV_9_5DB_HSW		(3<<24)   /* Sel3 */
-#define  DDI_BUF_EMP_600MV_0DB_HSW		(4<<24)   /* Sel4 */
-#define  DDI_BUF_EMP_600MV_3_5DB_HSW		(5<<24)   /* Sel5 */
-#define  DDI_BUF_EMP_600MV_6DB_HSW		(6<<24)   /* Sel6 */
-#define  DDI_BUF_EMP_800MV_0DB_HSW		(7<<24)   /* Sel7 */
-#define  DDI_BUF_EMP_800MV_3_5DB_HSW		(8<<24)   /* Sel8 */
+#define  DDI_BUF_TRANS_SELECT(n)	((n) << 24)
 #define  DDI_BUF_EMP_MASK			(0xf<<24)
 #define  DDI_BUF_PORT_REVERSAL			(1<<16)
 #define  DDI_BUF_IS_IDLE			(1<<7)
@@ -6008,6 +6474,83 @@ enum punit_power_well {
 #define  LCPLL_CD_SOURCE_FCLK		(1<<21)
 #define  LCPLL_CD_SOURCE_FCLK_DONE	(1<<19)
 
+/*
+ * SKL Clocks
+ */
+
+/* CDCLK_CTL */
+#define CDCLK_CTL			0x46000
+#define  CDCLK_FREQ_SEL_MASK		(3<<26)
+#define  CDCLK_FREQ_450_432		(0<<26)
+#define  CDCLK_FREQ_540			(1<<26)
+#define  CDCLK_FREQ_337_308		(2<<26)
+#define  CDCLK_FREQ_675_617		(3<<26)
+#define  CDCLK_FREQ_DECIMAL_MASK	(0x7ff)
+
+/* LCPLL_CTL */
+#define LCPLL1_CTL		0x46010
+#define LCPLL2_CTL		0x46014
+#define  LCPLL_PLL_ENABLE	(1<<31)
+
+/* DPLL control1 */
+#define DPLL_CTRL1		0x6C058
+#define  DPLL_CTRL1_HDMI_MODE(id)		(1<<((id)*6+5))
+#define  DPLL_CTRL1_SSC(id)			(1<<((id)*6+4))
+#define  DPLL_CRTL1_LINK_RATE_MASK(id)		(7<<((id)*6+1))
+#define  DPLL_CRTL1_LINK_RATE_SHIFT(id)		((id)*6+1)
+#define  DPLL_CRTL1_LINK_RATE(linkrate, id)	((linkrate)<<((id)*6+1))
+#define  DPLL_CTRL1_OVERRIDE(id)		(1<<((id)*6))
+#define  DPLL_CRTL1_LINK_RATE_2700		0
+#define  DPLL_CRTL1_LINK_RATE_1350		1
+#define  DPLL_CRTL1_LINK_RATE_810		2
+#define  DPLL_CRTL1_LINK_RATE_1620		3
+#define  DPLL_CRTL1_LINK_RATE_1080		4
+#define  DPLL_CRTL1_LINK_RATE_2160		5
+
+/* DPLL control2 */
+#define DPLL_CTRL2				0x6C05C
+#define  DPLL_CTRL2_DDI_CLK_OFF(port)		(1<<(port+15))
+#define  DPLL_CTRL2_DDI_CLK_SEL_MASK(port)	(3<<((port)*3+1))
+#define  DPLL_CTRL2_DDI_CLK_SEL_SHIFT(port)    ((port)*3+1)
+#define  DPLL_CTRL2_DDI_CLK_SEL(clk, port)	(clk<<((port)*3+1))
+#define  DPLL_CTRL2_DDI_SEL_OVERRIDE(port)     (1<<((port)*3))
+
+/* DPLL Status */
+#define DPLL_STATUS	0x6C060
+#define  DPLL_LOCK(id) (1<<((id)*8))
+
+/* DPLL cfg */
+#define DPLL1_CFGCR1	0x6C040
+#define DPLL2_CFGCR1	0x6C048
+#define DPLL3_CFGCR1	0x6C050
+#define  DPLL_CFGCR1_FREQ_ENABLE	(1<<31)
+#define  DPLL_CFGCR1_DCO_FRACTION_MASK	(0x7fff<<9)
+#define  DPLL_CFGCR1_DCO_FRACTION(x)	(x<<9)
+#define  DPLL_CFGCR1_DCO_INTEGER_MASK	(0x1ff)
+
+#define DPLL1_CFGCR2	0x6C044
+#define DPLL2_CFGCR2	0x6C04C
+#define DPLL3_CFGCR2	0x6C054
+#define  DPLL_CFGCR2_QDIV_RATIO_MASK	(0xff<<8)
+#define  DPLL_CFGCR2_QDIV_RATIO(x)	(x<<8)
+#define  DPLL_CFGCR2_QDIV_MODE(x)	(x<<7)
+#define  DPLL_CFGCR2_KDIV_MASK		(3<<5)
+#define  DPLL_CFGCR2_KDIV(x)		(x<<5)
+#define  DPLL_CFGCR2_KDIV_5 (0<<5)
+#define  DPLL_CFGCR2_KDIV_2 (1<<5)
+#define  DPLL_CFGCR2_KDIV_3 (2<<5)
+#define  DPLL_CFGCR2_KDIV_1 (3<<5)
+#define  DPLL_CFGCR2_PDIV_MASK		(7<<2)
+#define  DPLL_CFGCR2_PDIV(x)		(x<<2)
+#define  DPLL_CFGCR2_PDIV_1 (0<<2)
+#define  DPLL_CFGCR2_PDIV_2 (1<<2)
+#define  DPLL_CFGCR2_PDIV_3 (2<<2)
+#define  DPLL_CFGCR2_PDIV_7 (4<<2)
+#define  DPLL_CFGCR2_CENTRAL_FREQ_MASK	(3)
+
+#define GET_CFG_CR1_REG(id) (DPLL1_CFGCR1 + (id - SKL_DPLL1) * 8)
+#define GET_CFG_CR2_REG(id) (DPLL1_CFGCR2 + (id - SKL_DPLL1) * 8)
+
 /* Please see hsw_read_dcomp() and hsw_write_dcomp() before using this register,
  * since on HSW we can't write to it using I915_WRITE. */
 #define D_COMP_HSW			(MCHBAR_MIRROR_BASE_SNB + 0x5F0C)
diff --git a/drivers/gpu/drm/i915/i915_suspend.c b/drivers/gpu/drm/i915/i915_suspend.c
index 043123c77a1f..26368822a33f 100644
--- a/drivers/gpu/drm/i915/i915_suspend.c
+++ b/drivers/gpu/drm/i915/i915_suspend.c
@@ -203,34 +203,19 @@ static void i915_save_display(struct drm_device *dev)
 		i915_save_display_reg(dev);
 
 	/* LVDS state */
-	if (HAS_PCH_SPLIT(dev)) {
-		dev_priv->regfile.savePP_CONTROL = I915_READ(PCH_PP_CONTROL);
-		if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev))
-			dev_priv->regfile.saveLVDS = I915_READ(PCH_LVDS);
-	} else if (IS_VALLEYVIEW(dev)) {
-		dev_priv->regfile.savePP_CONTROL = I915_READ(PP_CONTROL);
-		dev_priv->regfile.savePFIT_PGM_RATIOS = I915_READ(PFIT_PGM_RATIOS);
-
-		dev_priv->regfile.saveBLC_HIST_CTL =
-			I915_READ(VLV_BLC_HIST_CTL(PIPE_A));
-		dev_priv->regfile.saveBLC_HIST_CTL_B =
-			I915_READ(VLV_BLC_HIST_CTL(PIPE_B));
-	} else {
-		dev_priv->regfile.savePP_CONTROL = I915_READ(PP_CONTROL);
-		dev_priv->regfile.savePFIT_PGM_RATIOS = I915_READ(PFIT_PGM_RATIOS);
-		dev_priv->regfile.saveBLC_HIST_CTL = I915_READ(BLC_HIST_CTL);
-		if (IS_MOBILE(dev) && !IS_I830(dev))
-			dev_priv->regfile.saveLVDS = I915_READ(LVDS);
-	}
-
-	if (!IS_I830(dev) && !IS_845G(dev) && !HAS_PCH_SPLIT(dev))
-		dev_priv->regfile.savePFIT_CONTROL = I915_READ(PFIT_CONTROL);
+	if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev))
+		dev_priv->regfile.saveLVDS = I915_READ(PCH_LVDS);
+	else if (INTEL_INFO(dev)->gen <= 4 && IS_MOBILE(dev) && !IS_I830(dev))
+		dev_priv->regfile.saveLVDS = I915_READ(LVDS);
 
+	/* Panel power sequencer */
 	if (HAS_PCH_SPLIT(dev)) {
+		dev_priv->regfile.savePP_CONTROL = I915_READ(PCH_PP_CONTROL);
 		dev_priv->regfile.savePP_ON_DELAYS = I915_READ(PCH_PP_ON_DELAYS);
 		dev_priv->regfile.savePP_OFF_DELAYS = I915_READ(PCH_PP_OFF_DELAYS);
 		dev_priv->regfile.savePP_DIVISOR = I915_READ(PCH_PP_DIVISOR);
-	} else {
+	} else if (!IS_VALLEYVIEW(dev)) {
+		dev_priv->regfile.savePP_CONTROL = I915_READ(PP_CONTROL);
 		dev_priv->regfile.savePP_ON_DELAYS = I915_READ(PP_ON_DELAYS);
 		dev_priv->regfile.savePP_OFF_DELAYS = I915_READ(PP_OFF_DELAYS);
 		dev_priv->regfile.savePP_DIVISOR = I915_READ(PP_DIVISOR);
@@ -259,29 +244,19 @@ static void i915_restore_display(struct drm_device *dev)
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		mask = ~LVDS_PORT_EN;
 
+	/* LVDS state */
 	if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev))
 		I915_WRITE(PCH_LVDS, dev_priv->regfile.saveLVDS & mask);
 	else if (INTEL_INFO(dev)->gen <= 4 && IS_MOBILE(dev) && !IS_I830(dev))
 		I915_WRITE(LVDS, dev_priv->regfile.saveLVDS & mask);
 
-	if (!IS_I830(dev) && !IS_845G(dev) && !HAS_PCH_SPLIT(dev))
-		I915_WRITE(PFIT_CONTROL, dev_priv->regfile.savePFIT_CONTROL);
-
+	/* Panel power sequencer */
 	if (HAS_PCH_SPLIT(dev)) {
 		I915_WRITE(PCH_PP_ON_DELAYS, dev_priv->regfile.savePP_ON_DELAYS);
 		I915_WRITE(PCH_PP_OFF_DELAYS, dev_priv->regfile.savePP_OFF_DELAYS);
 		I915_WRITE(PCH_PP_DIVISOR, dev_priv->regfile.savePP_DIVISOR);
 		I915_WRITE(PCH_PP_CONTROL, dev_priv->regfile.savePP_CONTROL);
-		I915_WRITE(RSTDBYCTL,
-			   dev_priv->regfile.saveMCHBAR_RENDER_STANDBY);
-	} else if (IS_VALLEYVIEW(dev)) {
-		I915_WRITE(VLV_BLC_HIST_CTL(PIPE_A),
-			   dev_priv->regfile.saveBLC_HIST_CTL);
-		I915_WRITE(VLV_BLC_HIST_CTL(PIPE_B),
-			   dev_priv->regfile.saveBLC_HIST_CTL);
-	} else {
-		I915_WRITE(PFIT_PGM_RATIOS, dev_priv->regfile.savePFIT_PGM_RATIOS);
-		I915_WRITE(BLC_HIST_CTL, dev_priv->regfile.saveBLC_HIST_CTL);
+	} else if (!IS_VALLEYVIEW(dev)) {
 		I915_WRITE(PP_ON_DELAYS, dev_priv->regfile.savePP_ON_DELAYS);
 		I915_WRITE(PP_OFF_DELAYS, dev_priv->regfile.savePP_OFF_DELAYS);
 		I915_WRITE(PP_DIVISOR, dev_priv->regfile.savePP_DIVISOR);
@@ -328,6 +303,10 @@ int i915_save_state(struct drm_device *dev)
 		}
 	}
 
+	if (IS_GEN4(dev))
+		pci_read_config_word(dev->pdev, GCDGMBUS,
+				     &dev_priv->regfile.saveGCDGMBUS);
+
 	/* Cache mode state */
 	if (INTEL_INFO(dev)->gen < 7)
 		dev_priv->regfile.saveCACHE_MODE_0 = I915_READ(CACHE_MODE_0);
@@ -356,6 +335,10 @@ int i915_restore_state(struct drm_device *dev)
 	mutex_lock(&dev->struct_mutex);
 
 	i915_gem_restore_fences(dev);
+
+	if (IS_GEN4(dev))
+		pci_write_config_word(dev->pdev, GCDGMBUS,
+				      dev_priv->regfile.saveGCDGMBUS);
 	i915_restore_display(dev);
 
 	if (!drm_core_check_feature(dev, DRIVER_MODESET)) {
@@ -368,6 +351,8 @@ int i915_restore_state(struct drm_device *dev)
 			I915_WRITE(_FDI_RXA_IMR, dev_priv->regfile.saveFDI_RXA_IMR);
 			I915_WRITE(_FDI_RXB_IMR, dev_priv->regfile.saveFDI_RXB_IMR);
 			I915_WRITE(PCH_PORT_HOTPLUG, dev_priv->regfile.savePCH_PORT_HOTPLUG);
+			I915_WRITE(RSTDBYCTL,
+				   dev_priv->regfile.saveMCHBAR_RENDER_STANDBY);
 		} else {
 			I915_WRITE(IER, dev_priv->regfile.saveIER);
 			I915_WRITE(IMR, dev_priv->regfile.saveIMR);
diff --git a/drivers/gpu/drm/i915/i915_sysfs.c b/drivers/gpu/drm/i915/i915_sysfs.c
index ae7fd8fc27f0..4a5af695307e 100644
--- a/drivers/gpu/drm/i915/i915_sysfs.c
+++ b/drivers/gpu/drm/i915/i915_sysfs.c
@@ -139,8 +139,6 @@ static DEVICE_ATTR(rc6pp_residency_ms, S_IRUGO, show_rc6pp_ms, NULL);
 static struct attribute *rc6_attrs[] = {
 	&dev_attr_rc6_enable.attr,
 	&dev_attr_rc6_residency_ms.attr,
-	&dev_attr_rc6p_residency_ms.attr,
-	&dev_attr_rc6pp_residency_ms.attr,
 	NULL
 };
 
@@ -148,6 +146,17 @@ static struct attribute_group rc6_attr_group = {
 	.name = power_group_name,
 	.attrs =  rc6_attrs
 };
+
+static struct attribute *rc6p_attrs[] = {
+	&dev_attr_rc6p_residency_ms.attr,
+	&dev_attr_rc6pp_residency_ms.attr,
+	NULL
+};
+
+static struct attribute_group rc6p_attr_group = {
+	.name = power_group_name,
+	.attrs =  rc6p_attrs
+};
 #endif
 
 static int l3_access_valid(struct drm_device *dev, loff_t offset)
@@ -540,7 +549,7 @@ static ssize_t error_state_read(struct file *filp, struct kobject *kobj,
 
 	memset(&error_priv, 0, sizeof(error_priv));
 
-	ret = i915_error_state_buf_init(&error_str, count, off);
+	ret = i915_error_state_buf_init(&error_str, to_i915(dev), count, off);
 	if (ret)
 		return ret;
 
@@ -595,12 +604,18 @@ void i915_setup_sysfs(struct drm_device *dev)
 	int ret;
 
 #ifdef CONFIG_PM
-	if (INTEL_INFO(dev)->gen >= 6) {
+	if (HAS_RC6(dev)) {
 		ret = sysfs_merge_group(&dev->primary->kdev->kobj,
 					&rc6_attr_group);
 		if (ret)
 			DRM_ERROR("RC6 residency sysfs setup failed\n");
 	}
+	if (HAS_RC6p(dev)) {
+		ret = sysfs_merge_group(&dev->primary->kdev->kobj,
+					&rc6p_attr_group);
+		if (ret)
+			DRM_ERROR("RC6p residency sysfs setup failed\n");
+	}
 #endif
 	if (HAS_L3_DPF(dev)) {
 		ret = device_create_bin_file(dev->primary->kdev, &dpf_attrs);
@@ -640,5 +655,6 @@ void i915_teardown_sysfs(struct drm_device *dev)
 	device_remove_bin_file(dev->primary->kdev,  &dpf_attrs);
 #ifdef CONFIG_PM
 	sysfs_unmerge_group(&dev->primary->kdev->kobj, &rc6_attr_group);
+	sysfs_unmerge_group(&dev->primary->kdev->kobj, &rc6p_attr_group);
 #endif
 }
diff --git a/drivers/gpu/drm/i915/i915_trace.h b/drivers/gpu/drm/i915/i915_trace.h
index f5aa0067755a..751d4ad14d62 100644
--- a/drivers/gpu/drm/i915/i915_trace.h
+++ b/drivers/gpu/drm/i915/i915_trace.h
@@ -587,6 +587,110 @@ TRACE_EVENT(intel_gpu_freq_change,
 	    TP_printk("new_freq=%u", __entry->freq)
 );
 
+/**
+ * DOC: i915_ppgtt_create and i915_ppgtt_release tracepoints
+ *
+ * With full ppgtt enabled each process using drm will allocate at least one
+ * translation table. With these traces it is possible to keep track of the
+ * allocation and of the lifetime of the tables; this can be used during
+ * testing/debug to verify that we are not leaking ppgtts.
+ * These traces identify the ppgtt through the vm pointer, which is also printed
+ * by the i915_vma_bind and i915_vma_unbind tracepoints.
+ */
+DECLARE_EVENT_CLASS(i915_ppgtt,
+	TP_PROTO(struct i915_address_space *vm),
+	TP_ARGS(vm),
+
+	TP_STRUCT__entry(
+			__field(struct i915_address_space *, vm)
+			__field(u32, dev)
+	),
+
+	TP_fast_assign(
+			__entry->vm = vm;
+			__entry->dev = vm->dev->primary->index;
+	),
+
+	TP_printk("dev=%u, vm=%p", __entry->dev, __entry->vm)
+)
+
+DEFINE_EVENT(i915_ppgtt, i915_ppgtt_create,
+	TP_PROTO(struct i915_address_space *vm),
+	TP_ARGS(vm)
+);
+
+DEFINE_EVENT(i915_ppgtt, i915_ppgtt_release,
+	TP_PROTO(struct i915_address_space *vm),
+	TP_ARGS(vm)
+);
+
+/**
+ * DOC: i915_context_create and i915_context_free tracepoints
+ *
+ * These tracepoints are used to track creation and deletion of contexts.
+ * If full ppgtt is enabled, they also print the address of the vm assigned to
+ * the context.
+ */
+DECLARE_EVENT_CLASS(i915_context,
+	TP_PROTO(struct intel_context *ctx),
+	TP_ARGS(ctx),
+
+	TP_STRUCT__entry(
+			__field(u32, dev)
+			__field(struct intel_context *, ctx)
+			__field(struct i915_address_space *, vm)
+	),
+
+	TP_fast_assign(
+			__entry->ctx = ctx;
+			__entry->vm = ctx->ppgtt ? &ctx->ppgtt->base : NULL;
+			__entry->dev = ctx->file_priv->dev_priv->dev->primary->index;
+	),
+
+	TP_printk("dev=%u, ctx=%p, ctx_vm=%p",
+		  __entry->dev, __entry->ctx, __entry->vm)
+)
+
+DEFINE_EVENT(i915_context, i915_context_create,
+	TP_PROTO(struct intel_context *ctx),
+	TP_ARGS(ctx)
+);
+
+DEFINE_EVENT(i915_context, i915_context_free,
+	TP_PROTO(struct intel_context *ctx),
+	TP_ARGS(ctx)
+);
+
+/**
+ * DOC: switch_mm tracepoint
+ *
+ * This tracepoint allows tracking of the mm switch, which is an important point
+ * in the lifetime of the vm in the legacy submission path. This tracepoint is
+ * called only if full ppgtt is enabled.
+ */
+TRACE_EVENT(switch_mm,
+	TP_PROTO(struct intel_engine_cs *ring, struct intel_context *to),
+
+	TP_ARGS(ring, to),
+
+	TP_STRUCT__entry(
+			__field(u32, ring)
+			__field(struct intel_context *, to)
+			__field(struct i915_address_space *, vm)
+			__field(u32, dev)
+	),
+
+	TP_fast_assign(
+			__entry->ring = ring->id;
+			__entry->to = to;
+			__entry->vm = to->ppgtt? &to->ppgtt->base : NULL;
+			__entry->dev = ring->dev->primary->index;
+	),
+
+	TP_printk("dev=%u, ring=%u, ctx=%p, ctx_vm=%p",
+		  __entry->dev, __entry->ring, __entry->to, __entry->vm)
+);
+
 #endif /* _I915_TRACE_H_ */
 
 /* This part must be outside protection */
diff --git a/drivers/gpu/drm/i915/i915_ums.c b/drivers/gpu/drm/i915/i915_ums.c
index 480da593e6c0..d10fe3e9c49f 100644
--- a/drivers/gpu/drm/i915/i915_ums.c
+++ b/drivers/gpu/drm/i915/i915_ums.c
@@ -270,6 +270,12 @@ void i915_save_display_reg(struct drm_device *dev)
 	}
 	/* FIXME: regfile.save TV & SDVO state */
 
+	/* Panel fitter */
+	if (!IS_I830(dev) && !IS_845G(dev) && !HAS_PCH_SPLIT(dev)) {
+		dev_priv->regfile.savePFIT_CONTROL = I915_READ(PFIT_CONTROL);
+		dev_priv->regfile.savePFIT_PGM_RATIOS = I915_READ(PFIT_PGM_RATIOS);
+	}
+
 	/* Backlight */
 	if (INTEL_INFO(dev)->gen <= 4)
 		pci_read_config_byte(dev->pdev, PCI_LBPC,
@@ -284,6 +290,7 @@ void i915_save_display_reg(struct drm_device *dev)
 		dev_priv->regfile.saveBLC_PWM_CTL = I915_READ(BLC_PWM_CTL);
 		if (INTEL_INFO(dev)->gen >= 4)
 			dev_priv->regfile.saveBLC_PWM_CTL2 = I915_READ(BLC_PWM_CTL2);
+		dev_priv->regfile.saveBLC_HIST_CTL = I915_READ(BLC_HIST_CTL);
 	}
 
 	return;
@@ -313,6 +320,13 @@ void i915_restore_display_reg(struct drm_device *dev)
 		if (INTEL_INFO(dev)->gen >= 4)
 			I915_WRITE(BLC_PWM_CTL2, dev_priv->regfile.saveBLC_PWM_CTL2);
 		I915_WRITE(BLC_PWM_CTL, dev_priv->regfile.saveBLC_PWM_CTL);
+		I915_WRITE(BLC_HIST_CTL, dev_priv->regfile.saveBLC_HIST_CTL);
+	}
+
+	/* Panel fitter */
+	if (!IS_I830(dev) && !IS_845G(dev) && !HAS_PCH_SPLIT(dev)) {
+		I915_WRITE(PFIT_PGM_RATIOS, dev_priv->regfile.savePFIT_PGM_RATIOS);
+		I915_WRITE(PFIT_CONTROL, dev_priv->regfile.savePFIT_CONTROL);
 	}
 
 	/* Display port ratios (must be done before clock is set) */
diff --git a/drivers/gpu/drm/i915/intel_audio.c b/drivers/gpu/drm/i915/intel_audio.c
new file mode 100644
index 000000000000..2c7ed5cb29c0
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_audio.c
@@ -0,0 +1,463 @@
+/*
+ * Copyright © 2014 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#include <linux/kernel.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_edid.h>
+#include "intel_drv.h"
+#include "i915_drv.h"
+
+/**
+ * DOC: High Definition Audio over HDMI and Display Port
+ *
+ * The graphics and audio drivers together support High Definition Audio over
+ * HDMI and Display Port. The audio programming sequences are divided into audio
+ * codec and controller enable and disable sequences. The graphics driver
+ * handles the audio codec sequences, while the audio driver handles the audio
+ * controller sequences.
+ *
+ * The disable sequences must be performed before disabling the transcoder or
+ * port. The enable sequences may only be performed after enabling the
+ * transcoder and port, and after completed link training.
+ *
+ * The codec and controller sequences could be done either parallel or serial,
+ * but generally the ELDV/PD change in the codec sequence indicates to the audio
+ * driver that the controller sequence should start. Indeed, most of the
+ * co-operation between the graphics and audio drivers is handled via audio
+ * related registers. (The notable exception is the power management, not
+ * covered here.)
+ */
+
+static const struct {
+	int clock;
+	u32 config;
+} hdmi_audio_clock[] = {
+	{ DIV_ROUND_UP(25200 * 1000, 1001), AUD_CONFIG_PIXEL_CLOCK_HDMI_25175 },
+	{ 25200, AUD_CONFIG_PIXEL_CLOCK_HDMI_25200 }, /* default per bspec */
+	{ 27000, AUD_CONFIG_PIXEL_CLOCK_HDMI_27000 },
+	{ 27000 * 1001 / 1000, AUD_CONFIG_PIXEL_CLOCK_HDMI_27027 },
+	{ 54000, AUD_CONFIG_PIXEL_CLOCK_HDMI_54000 },
+	{ 54000 * 1001 / 1000, AUD_CONFIG_PIXEL_CLOCK_HDMI_54054 },
+	{ DIV_ROUND_UP(74250 * 1000, 1001), AUD_CONFIG_PIXEL_CLOCK_HDMI_74176 },
+	{ 74250, AUD_CONFIG_PIXEL_CLOCK_HDMI_74250 },
+	{ DIV_ROUND_UP(148500 * 1000, 1001), AUD_CONFIG_PIXEL_CLOCK_HDMI_148352 },
+	{ 148500, AUD_CONFIG_PIXEL_CLOCK_HDMI_148500 },
+};
+
+/* get AUD_CONFIG_PIXEL_CLOCK_HDMI_* value for mode */
+static u32 audio_config_hdmi_pixel_clock(struct drm_display_mode *mode)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(hdmi_audio_clock); i++) {
+		if (mode->clock == hdmi_audio_clock[i].clock)
+			break;
+	}
+
+	if (i == ARRAY_SIZE(hdmi_audio_clock)) {
+		DRM_DEBUG_KMS("HDMI audio pixel clock setting for %d not found, falling back to defaults\n", mode->clock);
+		i = 1;
+	}
+
+	DRM_DEBUG_KMS("Configuring HDMI audio for pixel clock %d (0x%08x)\n",
+		      hdmi_audio_clock[i].clock,
+		      hdmi_audio_clock[i].config);
+
+	return hdmi_audio_clock[i].config;
+}
+
+static bool intel_eld_uptodate(struct drm_connector *connector,
+			       int reg_eldv, uint32_t bits_eldv,
+			       int reg_elda, uint32_t bits_elda,
+			       int reg_edid)
+{
+	struct drm_i915_private *dev_priv = connector->dev->dev_private;
+	uint8_t *eld = connector->eld;
+	uint32_t tmp;
+	int i;
+
+	tmp = I915_READ(reg_eldv);
+	tmp &= bits_eldv;
+
+	if (!tmp)
+		return false;
+
+	tmp = I915_READ(reg_elda);
+	tmp &= ~bits_elda;
+	I915_WRITE(reg_elda, tmp);
+
+	for (i = 0; i < drm_eld_size(eld) / 4; i++)
+		if (I915_READ(reg_edid) != *((uint32_t *)eld + i))
+			return false;
+
+	return true;
+}
+
+static void g4x_audio_codec_disable(struct intel_encoder *encoder)
+{
+	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
+	uint32_t eldv, tmp;
+
+	DRM_DEBUG_KMS("Disable audio codec\n");
+
+	tmp = I915_READ(G4X_AUD_VID_DID);
+	if (tmp == INTEL_AUDIO_DEVBLC || tmp == INTEL_AUDIO_DEVCL)
+		eldv = G4X_ELDV_DEVCL_DEVBLC;
+	else
+		eldv = G4X_ELDV_DEVCTG;
+
+	/* Invalidate ELD */
+	tmp = I915_READ(G4X_AUD_CNTL_ST);
+	tmp &= ~eldv;
+	I915_WRITE(G4X_AUD_CNTL_ST, tmp);
+}
+
+static void g4x_audio_codec_enable(struct drm_connector *connector,
+				   struct intel_encoder *encoder,
+				   struct drm_display_mode *mode)
+{
+	struct drm_i915_private *dev_priv = connector->dev->dev_private;
+	uint8_t *eld = connector->eld;
+	uint32_t eldv;
+	uint32_t tmp;
+	int len, i;
+
+	DRM_DEBUG_KMS("Enable audio codec, %u bytes ELD\n", eld[2]);
+
+	tmp = I915_READ(G4X_AUD_VID_DID);
+	if (tmp == INTEL_AUDIO_DEVBLC || tmp == INTEL_AUDIO_DEVCL)
+		eldv = G4X_ELDV_DEVCL_DEVBLC;
+	else
+		eldv = G4X_ELDV_DEVCTG;
+
+	if (intel_eld_uptodate(connector,
+			       G4X_AUD_CNTL_ST, eldv,
+			       G4X_AUD_CNTL_ST, G4X_ELD_ADDR_MASK,
+			       G4X_HDMIW_HDMIEDID))
+		return;
+
+	tmp = I915_READ(G4X_AUD_CNTL_ST);
+	tmp &= ~(eldv | G4X_ELD_ADDR_MASK);
+	len = (tmp >> 9) & 0x1f;		/* ELD buffer size */
+	I915_WRITE(G4X_AUD_CNTL_ST, tmp);
+
+	len = min(drm_eld_size(eld) / 4, len);
+	DRM_DEBUG_DRIVER("ELD size %d\n", len);
+	for (i = 0; i < len; i++)
+		I915_WRITE(G4X_HDMIW_HDMIEDID, *((uint32_t *)eld + i));
+
+	tmp = I915_READ(G4X_AUD_CNTL_ST);
+	tmp |= eldv;
+	I915_WRITE(G4X_AUD_CNTL_ST, tmp);
+}
+
+static void hsw_audio_codec_disable(struct intel_encoder *encoder)
+{
+	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
+	enum pipe pipe = intel_crtc->pipe;
+	uint32_t tmp;
+
+	DRM_DEBUG_KMS("Disable audio codec on pipe %c\n", pipe_name(pipe));
+
+	/* Disable timestamps */
+	tmp = I915_READ(HSW_AUD_CFG(pipe));
+	tmp &= ~AUD_CONFIG_N_VALUE_INDEX;
+	tmp |= AUD_CONFIG_N_PROG_ENABLE;
+	tmp &= ~AUD_CONFIG_UPPER_N_MASK;
+	tmp &= ~AUD_CONFIG_LOWER_N_MASK;
+	if (intel_pipe_has_type(intel_crtc, INTEL_OUTPUT_DISPLAYPORT))
+		tmp |= AUD_CONFIG_N_VALUE_INDEX;
+	I915_WRITE(HSW_AUD_CFG(pipe), tmp);
+
+	/* Invalidate ELD */
+	tmp = I915_READ(HSW_AUD_PIN_ELD_CP_VLD);
+	tmp &= ~AUDIO_ELD_VALID(pipe);
+	tmp &= ~AUDIO_OUTPUT_ENABLE(pipe);
+	I915_WRITE(HSW_AUD_PIN_ELD_CP_VLD, tmp);
+}
+
+static void hsw_audio_codec_enable(struct drm_connector *connector,
+				   struct intel_encoder *encoder,
+				   struct drm_display_mode *mode)
+{
+	struct drm_i915_private *dev_priv = connector->dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
+	enum pipe pipe = intel_crtc->pipe;
+	const uint8_t *eld = connector->eld;
+	uint32_t tmp;
+	int len, i;
+
+	DRM_DEBUG_KMS("Enable audio codec on pipe %c, %u bytes ELD\n",
+		      pipe_name(pipe), drm_eld_size(eld));
+
+	/* Enable audio presence detect, invalidate ELD */
+	tmp = I915_READ(HSW_AUD_PIN_ELD_CP_VLD);
+	tmp |= AUDIO_OUTPUT_ENABLE(pipe);
+	tmp &= ~AUDIO_ELD_VALID(pipe);
+	I915_WRITE(HSW_AUD_PIN_ELD_CP_VLD, tmp);
+
+	/*
+	 * FIXME: We're supposed to wait for vblank here, but we have vblanks
+	 * disabled during the mode set. The proper fix would be to push the
+	 * rest of the setup into a vblank work item, queued here, but the
+	 * infrastructure is not there yet.
+	 */
+
+	/* Reset ELD write address */
+	tmp = I915_READ(HSW_AUD_DIP_ELD_CTRL(pipe));
+	tmp &= ~IBX_ELD_ADDRESS_MASK;
+	I915_WRITE(HSW_AUD_DIP_ELD_CTRL(pipe), tmp);
+
+	/* Up to 84 bytes of hw ELD buffer */
+	len = min(drm_eld_size(eld), 84);
+	for (i = 0; i < len / 4; i++)
+		I915_WRITE(HSW_AUD_EDID_DATA(pipe), *((uint32_t *)eld + i));
+
+	/* ELD valid */
+	tmp = I915_READ(HSW_AUD_PIN_ELD_CP_VLD);
+	tmp |= AUDIO_ELD_VALID(pipe);
+	I915_WRITE(HSW_AUD_PIN_ELD_CP_VLD, tmp);
+
+	/* Enable timestamps */
+	tmp = I915_READ(HSW_AUD_CFG(pipe));
+	tmp &= ~AUD_CONFIG_N_VALUE_INDEX;
+	tmp &= ~AUD_CONFIG_N_PROG_ENABLE;
+	tmp &= ~AUD_CONFIG_PIXEL_CLOCK_HDMI_MASK;
+	if (intel_pipe_has_type(intel_crtc, INTEL_OUTPUT_DISPLAYPORT))
+		tmp |= AUD_CONFIG_N_VALUE_INDEX;
+	else
+		tmp |= audio_config_hdmi_pixel_clock(mode);
+	I915_WRITE(HSW_AUD_CFG(pipe), tmp);
+}
+
+static void ilk_audio_codec_disable(struct intel_encoder *encoder)
+{
+	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
+	struct intel_digital_port *intel_dig_port =
+		enc_to_dig_port(&encoder->base);
+	enum port port = intel_dig_port->port;
+	enum pipe pipe = intel_crtc->pipe;
+	uint32_t tmp, eldv;
+	int aud_config;
+	int aud_cntrl_st2;
+
+	DRM_DEBUG_KMS("Disable audio codec on port %c, pipe %c\n",
+		      port_name(port), pipe_name(pipe));
+
+	if (HAS_PCH_IBX(dev_priv->dev)) {
+		aud_config = IBX_AUD_CFG(pipe);
+		aud_cntrl_st2 = IBX_AUD_CNTL_ST2;
+	} else if (IS_VALLEYVIEW(dev_priv)) {
+		aud_config = VLV_AUD_CFG(pipe);
+		aud_cntrl_st2 = VLV_AUD_CNTL_ST2;
+	} else {
+		aud_config = CPT_AUD_CFG(pipe);
+		aud_cntrl_st2 = CPT_AUD_CNTRL_ST2;
+	}
+
+	/* Disable timestamps */
+	tmp = I915_READ(aud_config);
+	tmp &= ~AUD_CONFIG_N_VALUE_INDEX;
+	tmp |= AUD_CONFIG_N_PROG_ENABLE;
+	tmp &= ~AUD_CONFIG_UPPER_N_MASK;
+	tmp &= ~AUD_CONFIG_LOWER_N_MASK;
+	if (intel_pipe_has_type(intel_crtc, INTEL_OUTPUT_DISPLAYPORT))
+		tmp |= AUD_CONFIG_N_VALUE_INDEX;
+	I915_WRITE(aud_config, tmp);
+
+	if (WARN_ON(!port)) {
+		eldv = IBX_ELD_VALID(PORT_B) | IBX_ELD_VALID(PORT_C) |
+			IBX_ELD_VALID(PORT_D);
+	} else {
+		eldv = IBX_ELD_VALID(port);
+	}
+
+	/* Invalidate ELD */
+	tmp = I915_READ(aud_cntrl_st2);
+	tmp &= ~eldv;
+	I915_WRITE(aud_cntrl_st2, tmp);
+}
+
+static void ilk_audio_codec_enable(struct drm_connector *connector,
+				   struct intel_encoder *encoder,
+				   struct drm_display_mode *mode)
+{
+	struct drm_i915_private *dev_priv = connector->dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
+	struct intel_digital_port *intel_dig_port =
+		enc_to_dig_port(&encoder->base);
+	enum port port = intel_dig_port->port;
+	enum pipe pipe = intel_crtc->pipe;
+	uint8_t *eld = connector->eld;
+	uint32_t eldv;
+	uint32_t tmp;
+	int len, i;
+	int hdmiw_hdmiedid;
+	int aud_config;
+	int aud_cntl_st;
+	int aud_cntrl_st2;
+
+	DRM_DEBUG_KMS("Enable audio codec on port %c, pipe %c, %u bytes ELD\n",
+		      port_name(port), pipe_name(pipe), drm_eld_size(eld));
+
+	/*
+	 * FIXME: We're supposed to wait for vblank here, but we have vblanks
+	 * disabled during the mode set. The proper fix would be to push the
+	 * rest of the setup into a vblank work item, queued here, but the
+	 * infrastructure is not there yet.
+	 */
+
+	if (HAS_PCH_IBX(connector->dev)) {
+		hdmiw_hdmiedid = IBX_HDMIW_HDMIEDID(pipe);
+		aud_config = IBX_AUD_CFG(pipe);
+		aud_cntl_st = IBX_AUD_CNTL_ST(pipe);
+		aud_cntrl_st2 = IBX_AUD_CNTL_ST2;
+	} else if (IS_VALLEYVIEW(connector->dev)) {
+		hdmiw_hdmiedid = VLV_HDMIW_HDMIEDID(pipe);
+		aud_config = VLV_AUD_CFG(pipe);
+		aud_cntl_st = VLV_AUD_CNTL_ST(pipe);
+		aud_cntrl_st2 = VLV_AUD_CNTL_ST2;
+	} else {
+		hdmiw_hdmiedid = CPT_HDMIW_HDMIEDID(pipe);
+		aud_config = CPT_AUD_CFG(pipe);
+		aud_cntl_st = CPT_AUD_CNTL_ST(pipe);
+		aud_cntrl_st2 = CPT_AUD_CNTRL_ST2;
+	}
+
+	if (WARN_ON(!port)) {
+		eldv = IBX_ELD_VALID(PORT_B) | IBX_ELD_VALID(PORT_C) |
+			IBX_ELD_VALID(PORT_D);
+	} else {
+		eldv = IBX_ELD_VALID(port);
+	}
+
+	/* Invalidate ELD */
+	tmp = I915_READ(aud_cntrl_st2);
+	tmp &= ~eldv;
+	I915_WRITE(aud_cntrl_st2, tmp);
+
+	/* Reset ELD write address */
+	tmp = I915_READ(aud_cntl_st);
+	tmp &= ~IBX_ELD_ADDRESS_MASK;
+	I915_WRITE(aud_cntl_st, tmp);
+
+	/* Up to 84 bytes of hw ELD buffer */
+	len = min(drm_eld_size(eld), 84);
+	for (i = 0; i < len / 4; i++)
+		I915_WRITE(hdmiw_hdmiedid, *((uint32_t *)eld + i));
+
+	/* ELD valid */
+	tmp = I915_READ(aud_cntrl_st2);
+	tmp |= eldv;
+	I915_WRITE(aud_cntrl_st2, tmp);
+
+	/* Enable timestamps */
+	tmp = I915_READ(aud_config);
+	tmp &= ~AUD_CONFIG_N_VALUE_INDEX;
+	tmp &= ~AUD_CONFIG_N_PROG_ENABLE;
+	tmp &= ~AUD_CONFIG_PIXEL_CLOCK_HDMI_MASK;
+	if (intel_pipe_has_type(intel_crtc, INTEL_OUTPUT_DISPLAYPORT))
+		tmp |= AUD_CONFIG_N_VALUE_INDEX;
+	else
+		tmp |= audio_config_hdmi_pixel_clock(mode);
+	I915_WRITE(aud_config, tmp);
+}
+
+/**
+ * intel_audio_codec_enable - Enable the audio codec for HD audio
+ * @intel_encoder: encoder on which to enable audio
+ *
+ * The enable sequences may only be performed after enabling the transcoder and
+ * port, and after completed link training.
+ */
+void intel_audio_codec_enable(struct intel_encoder *intel_encoder)
+{
+	struct drm_encoder *encoder = &intel_encoder->base;
+	struct intel_crtc *crtc = to_intel_crtc(encoder->crtc);
+	struct drm_display_mode *mode = &crtc->config.adjusted_mode;
+	struct drm_connector *connector;
+	struct drm_device *dev = encoder->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	connector = drm_select_eld(encoder, mode);
+	if (!connector)
+		return;
+
+	DRM_DEBUG_DRIVER("ELD on [CONNECTOR:%d:%s], [ENCODER:%d:%s]\n",
+			 connector->base.id,
+			 connector->name,
+			 connector->encoder->base.id,
+			 connector->encoder->name);
+
+	/* ELD Conn_Type */
+	connector->eld[5] &= ~(3 << 2);
+	if (intel_pipe_has_type(crtc, INTEL_OUTPUT_DISPLAYPORT))
+		connector->eld[5] |= (1 << 2);
+
+	connector->eld[6] = drm_av_sync_delay(connector, mode) / 2;
+
+	if (dev_priv->display.audio_codec_enable)
+		dev_priv->display.audio_codec_enable(connector, intel_encoder, mode);
+}
+
+/**
+ * intel_audio_codec_disable - Disable the audio codec for HD audio
+ * @encoder: encoder on which to disable audio
+ *
+ * The disable sequences must be performed before disabling the transcoder or
+ * port.
+ */
+void intel_audio_codec_disable(struct intel_encoder *encoder)
+{
+	struct drm_device *dev = encoder->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (dev_priv->display.audio_codec_disable)
+		dev_priv->display.audio_codec_disable(encoder);
+}
+
+/**
+ * intel_init_audio - Set up chip specific audio functions
+ * @dev: drm device
+ */
+void intel_init_audio(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (IS_G4X(dev)) {
+		dev_priv->display.audio_codec_enable = g4x_audio_codec_enable;
+		dev_priv->display.audio_codec_disable = g4x_audio_codec_disable;
+	} else if (IS_VALLEYVIEW(dev)) {
+		dev_priv->display.audio_codec_enable = ilk_audio_codec_enable;
+		dev_priv->display.audio_codec_disable = ilk_audio_codec_disable;
+	} else if (IS_HASWELL(dev) || INTEL_INFO(dev)->gen >= 8) {
+		dev_priv->display.audio_codec_enable = hsw_audio_codec_enable;
+		dev_priv->display.audio_codec_disable = hsw_audio_codec_disable;
+	} else if (HAS_PCH_SPLIT(dev)) {
+		dev_priv->display.audio_codec_enable = ilk_audio_codec_enable;
+		dev_priv->display.audio_codec_disable = ilk_audio_codec_disable;
+	}
+}
diff --git a/drivers/gpu/drm/i915/intel_bios.c b/drivers/gpu/drm/i915/intel_bios.c
index eee79e1c3222..a4bd90f36a03 100644
--- a/drivers/gpu/drm/i915/intel_bios.c
+++ b/drivers/gpu/drm/i915/intel_bios.c
@@ -627,16 +627,16 @@ parse_edp(struct drm_i915_private *dev_priv, struct bdb_header *bdb)
 
 	switch (edp_link_params->preemphasis) {
 	case EDP_PREEMPHASIS_NONE:
-		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPHASIS_0;
+		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPH_LEVEL_0;
 		break;
 	case EDP_PREEMPHASIS_3_5dB:
-		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPHASIS_3_5;
+		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPH_LEVEL_1;
 		break;
 	case EDP_PREEMPHASIS_6dB:
-		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPHASIS_6;
+		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPH_LEVEL_2;
 		break;
 	case EDP_PREEMPHASIS_9_5dB:
-		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPHASIS_9_5;
+		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPH_LEVEL_3;
 		break;
 	default:
 		DRM_DEBUG_KMS("VBT has unknown eDP pre-emphasis value %u\n",
@@ -646,16 +646,16 @@ parse_edp(struct drm_i915_private *dev_priv, struct bdb_header *bdb)
 
 	switch (edp_link_params->vswing) {
 	case EDP_VSWING_0_4V:
-		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_400;
+		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_LEVEL_0;
 		break;
 	case EDP_VSWING_0_6V:
-		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_600;
+		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_LEVEL_1;
 		break;
 	case EDP_VSWING_0_8V:
-		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_800;
+		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_LEVEL_2;
 		break;
 	case EDP_VSWING_1_2V:
-		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_1200;
+		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_LEVEL_3;
 		break;
 	default:
 		DRM_DEBUG_KMS("VBT has unknown eDP voltage swing value %u\n",
@@ -946,7 +946,7 @@ static void parse_ddi_port(struct drm_i915_private *dev_priv, enum port port,
 		DRM_DEBUG_KMS("Analog port %c is also DP or TMDS compatible\n",
 			      port_name(port));
 	if (is_dvi && (port == PORT_A || port == PORT_E))
-		DRM_DEBUG_KMS("Port %c is TMDS compabile\n", port_name(port));
+		DRM_DEBUG_KMS("Port %c is TMDS compatible\n", port_name(port));
 	if (!is_dvi && !is_dp && !is_crt)
 		DRM_DEBUG_KMS("Port %c is not DP/TMDS/CRT compatible\n",
 			      port_name(port));
@@ -976,12 +976,10 @@ static void parse_ddi_port(struct drm_i915_private *dev_priv, enum port port,
 	if (bdb->version >= 158) {
 		/* The VBT HDMI level shift values match the table we have. */
 		hdmi_level_shift = child->raw[7] & 0xF;
-		if (hdmi_level_shift < 0xC) {
-			DRM_DEBUG_KMS("VBT HDMI level shift for port %c: %d\n",
-				      port_name(port),
-				      hdmi_level_shift);
-			info->hdmi_level_shift = hdmi_level_shift;
-		}
+		DRM_DEBUG_KMS("VBT HDMI level shift for port %c: %d\n",
+			      port_name(port),
+			      hdmi_level_shift);
+		info->hdmi_level_shift = hdmi_level_shift;
 	}
 }
 
@@ -1114,8 +1112,7 @@ init_vbt_defaults(struct drm_i915_private *dev_priv)
 		struct ddi_vbt_port_info *info =
 			&dev_priv->vbt.ddi_port_info[port];
 
-		/* Recommended BSpec default: 800mV 0dB. */
-		info->hdmi_level_shift = 6;
+		info->hdmi_level_shift = HDMI_LEVEL_SHIFT_UNKNOWN;
 
 		info->supports_dvi = (port != PORT_A && port != PORT_E);
 		info->supports_hdmi = info->supports_dvi;
diff --git a/drivers/gpu/drm/i915/intel_bios.h b/drivers/gpu/drm/i915/intel_bios.h
index b98667796337..7603765c91fc 100644
--- a/drivers/gpu/drm/i915/intel_bios.h
+++ b/drivers/gpu/drm/i915/intel_bios.h
@@ -46,7 +46,7 @@ struct bdb_header {
 	u16 version;			/**< decimal */
 	u16 header_size;		/**< in bytes */
 	u16 bdb_size;			/**< in bytes */
-};
+} __packed;
 
 /* strictly speaking, this is a "skip" block, but it has interesting info */
 struct vbios_data {
@@ -252,7 +252,7 @@ union child_device_config {
 	/* This one should also be safe to use anywhere, even without version
 	 * checks. */
 	struct common_child_dev_config common;
-};
+} __packed;
 
 struct bdb_general_definitions {
 	/* DDC GPIO */
@@ -802,7 +802,8 @@ struct mipi_config {
 
 	u16 rsvd4;
 
-	u8 rsvd5[5];
+	u8 rsvd5;
+	u32 target_burst_mode_freq;
 	u32 dsi_ddr_clk;
 	u32 bridge_ref_clk;
 
@@ -887,12 +888,12 @@ struct mipi_pps_data {
 	u16 bl_disable_delay;
 	u16 panel_off_delay;
 	u16 panel_power_cycle_delay;
-};
+} __packed;
 
 struct bdb_mipi_config {
 	struct mipi_config config[MAX_MIPI_CONFIGURATIONS];
 	struct mipi_pps_data pps[MAX_MIPI_CONFIGURATIONS];
-};
+} __packed;
 
 /* Block 53 contains MIPI sequences as needed by the panel
  * for enabling it. This block can be variable in size and
@@ -901,7 +902,7 @@ struct bdb_mipi_config {
 struct bdb_mipi_sequence {
 	u8 version;
 	u8 data[0];
-};
+} __packed;
 
 /* MIPI Sequnece Block definitions */
 enum mipi_seq {
diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c
index 9212e6504e0f..a9af9a4866db 100644
--- a/drivers/gpu/drm/i915/intel_crt.c
+++ b/drivers/gpu/drm/i915/intel_crt.c
@@ -72,7 +72,7 @@ static bool intel_crt_get_hw_state(struct intel_encoder *encoder,
 	u32 tmp;
 
 	power_domain = intel_display_port_power_domain(encoder);
-	if (!intel_display_power_enabled(dev_priv, power_domain))
+	if (!intel_display_power_is_enabled(dev_priv, power_domain))
 		return false;
 
 	tmp = I915_READ(crt->adpa_reg);
@@ -775,7 +775,7 @@ static void intel_crt_reset(struct drm_connector *connector)
 		I915_WRITE(crt->adpa_reg, adpa);
 		POSTING_READ(crt->adpa_reg);
 
-		DRM_DEBUG_KMS("pch crt adpa set to 0x%x\n", adpa);
+		DRM_DEBUG_KMS("crt adpa set to 0x%x\n", adpa);
 		crt->force_hotplug_required = 1;
 	}
 
diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c
index 5db0b5552e39..e6b45cd150d3 100644
--- a/drivers/gpu/drm/i915/intel_ddi.c
+++ b/drivers/gpu/drm/i915/intel_ddi.c
@@ -28,87 +28,129 @@
 #include "i915_drv.h"
 #include "intel_drv.h"
 
+struct ddi_buf_trans {
+	u32 trans1;	/* balance leg enable, de-emph level */
+	u32 trans2;	/* vref sel, vswing */
+};
+
 /* HDMI/DVI modes ignore everything but the last 2 items. So we share
  * them for both DP and FDI transports, allowing those ports to
  * automatically adapt to HDMI connections as well
  */
-static const u32 hsw_ddi_translations_dp[] = {
-	0x00FFFFFF, 0x0006000E,		/* DP parameters */
-	0x00D75FFF, 0x0005000A,
-	0x00C30FFF, 0x00040006,
-	0x80AAAFFF, 0x000B0000,
-	0x00FFFFFF, 0x0005000A,
-	0x00D75FFF, 0x000C0004,
-	0x80C30FFF, 0x000B0000,
-	0x00FFFFFF, 0x00040006,
-	0x80D75FFF, 0x000B0000,
+static const struct ddi_buf_trans hsw_ddi_translations_dp[] = {
+	{ 0x00FFFFFF, 0x0006000E },
+	{ 0x00D75FFF, 0x0005000A },
+	{ 0x00C30FFF, 0x00040006 },
+	{ 0x80AAAFFF, 0x000B0000 },
+	{ 0x00FFFFFF, 0x0005000A },
+	{ 0x00D75FFF, 0x000C0004 },
+	{ 0x80C30FFF, 0x000B0000 },
+	{ 0x00FFFFFF, 0x00040006 },
+	{ 0x80D75FFF, 0x000B0000 },
+};
+
+static const struct ddi_buf_trans hsw_ddi_translations_fdi[] = {
+	{ 0x00FFFFFF, 0x0007000E },
+	{ 0x00D75FFF, 0x000F000A },
+	{ 0x00C30FFF, 0x00060006 },
+	{ 0x00AAAFFF, 0x001E0000 },
+	{ 0x00FFFFFF, 0x000F000A },
+	{ 0x00D75FFF, 0x00160004 },
+	{ 0x00C30FFF, 0x001E0000 },
+	{ 0x00FFFFFF, 0x00060006 },
+	{ 0x00D75FFF, 0x001E0000 },
+};
+
+static const struct ddi_buf_trans hsw_ddi_translations_hdmi[] = {
+					/* Idx	NT mV d	T mV d	db	*/
+	{ 0x00FFFFFF, 0x0006000E },	/* 0:	400	400	0	*/
+	{ 0x00E79FFF, 0x000E000C },	/* 1:	400	500	2	*/
+	{ 0x00D75FFF, 0x0005000A },	/* 2:	400	600	3.5	*/
+	{ 0x00FFFFFF, 0x0005000A },	/* 3:	600	600	0	*/
+	{ 0x00E79FFF, 0x001D0007 },	/* 4:	600	750	2	*/
+	{ 0x00D75FFF, 0x000C0004 },	/* 5:	600	900	3.5	*/
+	{ 0x00FFFFFF, 0x00040006 },	/* 6:	800	800	0	*/
+	{ 0x80E79FFF, 0x00030002 },	/* 7:	800	1000	2	*/
+	{ 0x00FFFFFF, 0x00140005 },	/* 8:	850	850	0	*/
+	{ 0x00FFFFFF, 0x000C0004 },	/* 9:	900	900	0	*/
+	{ 0x00FFFFFF, 0x001C0003 },	/* 10:	950	950	0	*/
+	{ 0x80FFFFFF, 0x00030002 },	/* 11:	1000	1000	0	*/
+};
+
+static const struct ddi_buf_trans bdw_ddi_translations_edp[] = {
+	{ 0x00FFFFFF, 0x00000012 },
+	{ 0x00EBAFFF, 0x00020011 },
+	{ 0x00C71FFF, 0x0006000F },
+	{ 0x00AAAFFF, 0x000E000A },
+	{ 0x00FFFFFF, 0x00020011 },
+	{ 0x00DB6FFF, 0x0005000F },
+	{ 0x00BEEFFF, 0x000A000C },
+	{ 0x00FFFFFF, 0x0005000F },
+	{ 0x00DB6FFF, 0x000A000C },
 };
 
-static const u32 hsw_ddi_translations_fdi[] = {
-	0x00FFFFFF, 0x0007000E,		/* FDI parameters */
-	0x00D75FFF, 0x000F000A,
-	0x00C30FFF, 0x00060006,
-	0x00AAAFFF, 0x001E0000,
-	0x00FFFFFF, 0x000F000A,
-	0x00D75FFF, 0x00160004,
-	0x00C30FFF, 0x001E0000,
-	0x00FFFFFF, 0x00060006,
-	0x00D75FFF, 0x001E0000,
+static const struct ddi_buf_trans bdw_ddi_translations_dp[] = {
+	{ 0x00FFFFFF, 0x0007000E },
+	{ 0x00D75FFF, 0x000E000A },
+	{ 0x00BEFFFF, 0x00140006 },
+	{ 0x80B2CFFF, 0x001B0002 },
+	{ 0x00FFFFFF, 0x000E000A },
+	{ 0x00DB6FFF, 0x00160005 },
+	{ 0x80C71FFF, 0x001A0002 },
+	{ 0x00F7DFFF, 0x00180004 },
+	{ 0x80D75FFF, 0x001B0002 },
 };
 
-static const u32 hsw_ddi_translations_hdmi[] = {
-				/* Idx	NT mV diff	T mV diff	db  */
-	0x00FFFFFF, 0x0006000E, /* 0:	400		400		0   */
-	0x00E79FFF, 0x000E000C, /* 1:	400		500		2   */
-	0x00D75FFF, 0x0005000A, /* 2:	400		600		3.5 */
-	0x00FFFFFF, 0x0005000A, /* 3:	600		600		0   */
-	0x00E79FFF, 0x001D0007, /* 4:	600		750		2   */
-	0x00D75FFF, 0x000C0004, /* 5:	600		900		3.5 */
-	0x00FFFFFF, 0x00040006, /* 6:	800		800		0   */
-	0x80E79FFF, 0x00030002, /* 7:	800		1000		2   */
-	0x00FFFFFF, 0x00140005, /* 8:	850		850		0   */
-	0x00FFFFFF, 0x000C0004, /* 9:	900		900		0   */
-	0x00FFFFFF, 0x001C0003, /* 10:	950		950		0   */
-	0x80FFFFFF, 0x00030002, /* 11:	1000		1000		0   */
+static const struct ddi_buf_trans bdw_ddi_translations_fdi[] = {
+	{ 0x00FFFFFF, 0x0001000E },
+	{ 0x00D75FFF, 0x0004000A },
+	{ 0x00C30FFF, 0x00070006 },
+	{ 0x00AAAFFF, 0x000C0000 },
+	{ 0x00FFFFFF, 0x0004000A },
+	{ 0x00D75FFF, 0x00090004 },
+	{ 0x00C30FFF, 0x000C0000 },
+	{ 0x00FFFFFF, 0x00070006 },
+	{ 0x00D75FFF, 0x000C0000 },
 };
 
-static const u32 bdw_ddi_translations_edp[] = {
-	0x00FFFFFF, 0x00000012,		/* eDP parameters */
-	0x00EBAFFF, 0x00020011,
-	0x00C71FFF, 0x0006000F,
-	0x00AAAFFF, 0x000E000A,
-	0x00FFFFFF, 0x00020011,
-	0x00DB6FFF, 0x0005000F,
-	0x00BEEFFF, 0x000A000C,
-	0x00FFFFFF, 0x0005000F,
-	0x00DB6FFF, 0x000A000C,
-	0x00FFFFFF, 0x00140006		/* HDMI parameters 800mV 0dB*/
+static const struct ddi_buf_trans bdw_ddi_translations_hdmi[] = {
+					/* Idx	NT mV d	T mV df	db	*/
+	{ 0x00FFFFFF, 0x0007000E },	/* 0:	400	400	0	*/
+	{ 0x00D75FFF, 0x000E000A },	/* 1:	400	600	3.5	*/
+	{ 0x00BEFFFF, 0x00140006 },	/* 2:	400	800	6	*/
+	{ 0x00FFFFFF, 0x0009000D },	/* 3:	450	450	0	*/
+	{ 0x00FFFFFF, 0x000E000A },	/* 4:	600	600	0	*/
+	{ 0x00D7FFFF, 0x00140006 },	/* 5:	600	800	2.5	*/
+	{ 0x80CB2FFF, 0x001B0002 },	/* 6:	600	1000	4.5	*/
+	{ 0x00FFFFFF, 0x00140006 },	/* 7:	800	800	0	*/
+	{ 0x80E79FFF, 0x001B0002 },	/* 8:	800	1000	2	*/
+	{ 0x80FFFFFF, 0x001B0002 },	/* 9:	1000	1000	0	*/
 };
 
-static const u32 bdw_ddi_translations_dp[] = {
-	0x00FFFFFF, 0x0007000E,		/* DP parameters */
-	0x00D75FFF, 0x000E000A,
-	0x00BEFFFF, 0x00140006,
-	0x80B2CFFF, 0x001B0002,
-	0x00FFFFFF, 0x000E000A,
-	0x00D75FFF, 0x00180004,
-	0x80CB2FFF, 0x001B0002,
-	0x00F7DFFF, 0x00180004,
-	0x80D75FFF, 0x001B0002,
-	0x00FFFFFF, 0x00140006		/* HDMI parameters 800mV 0dB*/
+static const struct ddi_buf_trans skl_ddi_translations_dp[] = {
+	{ 0x00000018, 0x000000a0 },
+	{ 0x00004014, 0x00000098 },
+	{ 0x00006012, 0x00000088 },
+	{ 0x00008010, 0x00000080 },
+	{ 0x00000018, 0x00000098 },
+	{ 0x00004014, 0x00000088 },
+	{ 0x00006012, 0x00000080 },
+	{ 0x00000018, 0x00000088 },
+	{ 0x00004014, 0x00000080 },
 };
 
-static const u32 bdw_ddi_translations_fdi[] = {
-	0x00FFFFFF, 0x0001000E,		/* FDI parameters */
-	0x00D75FFF, 0x0004000A,
-	0x00C30FFF, 0x00070006,
-	0x00AAAFFF, 0x000C0000,
-	0x00FFFFFF, 0x0004000A,
-	0x00D75FFF, 0x00090004,
-	0x00C30FFF, 0x000C0000,
-	0x00FFFFFF, 0x00070006,
-	0x00D75FFF, 0x000C0000,
-	0x00FFFFFF, 0x00140006		/* HDMI parameters 800mV 0dB*/
+static const struct ddi_buf_trans skl_ddi_translations_hdmi[] = {
+					/* Idx	NT mV   T mV    db  */
+	{ 0x00000018, 0x000000a0 },	/* 0:	400	400	0   */
+	{ 0x00004014, 0x00000098 },	/* 1:	400	600	3.5 */
+	{ 0x00006012, 0x00000088 },	/* 2:	400	800	6   */
+	{ 0x00000018, 0x0000003c },	/* 3:	450	450	0   */
+	{ 0x00000018, 0x00000098 },	/* 4:	600	600	0   */
+	{ 0x00003015, 0x00000088 },	/* 5:	600	800	2.5 */
+	{ 0x00005013, 0x00000080 },	/* 6:	600	1000	4.5 */
+	{ 0x00000018, 0x00000088 },	/* 7:	800	800	0   */
+	{ 0x00000096, 0x00000080 },	/* 8:	800	1000	2   */
+	{ 0x00000018, 0x00000080 },	/* 9:	1200	1200	0   */
 };
 
 enum port intel_ddi_get_encoder_port(struct intel_encoder *intel_encoder)
@@ -145,26 +187,43 @@ static void intel_prepare_ddi_buffers(struct drm_device *dev, enum port port)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 reg;
-	int i;
+	int i, n_hdmi_entries, hdmi_800mV_0dB;
 	int hdmi_level = dev_priv->vbt.ddi_port_info[port].hdmi_level_shift;
-	const u32 *ddi_translations_fdi;
-	const u32 *ddi_translations_dp;
-	const u32 *ddi_translations_edp;
-	const u32 *ddi_translations;
-
-	if (IS_BROADWELL(dev)) {
+	const struct ddi_buf_trans *ddi_translations_fdi;
+	const struct ddi_buf_trans *ddi_translations_dp;
+	const struct ddi_buf_trans *ddi_translations_edp;
+	const struct ddi_buf_trans *ddi_translations_hdmi;
+	const struct ddi_buf_trans *ddi_translations;
+
+	if (IS_SKYLAKE(dev)) {
+		ddi_translations_fdi = NULL;
+		ddi_translations_dp = skl_ddi_translations_dp;
+		ddi_translations_edp = skl_ddi_translations_dp;
+		ddi_translations_hdmi = skl_ddi_translations_hdmi;
+		n_hdmi_entries = ARRAY_SIZE(skl_ddi_translations_hdmi);
+		hdmi_800mV_0dB = 7;
+	} else if (IS_BROADWELL(dev)) {
 		ddi_translations_fdi = bdw_ddi_translations_fdi;
 		ddi_translations_dp = bdw_ddi_translations_dp;
 		ddi_translations_edp = bdw_ddi_translations_edp;
+		ddi_translations_hdmi = bdw_ddi_translations_hdmi;
+		n_hdmi_entries = ARRAY_SIZE(bdw_ddi_translations_hdmi);
+		hdmi_800mV_0dB = 7;
 	} else if (IS_HASWELL(dev)) {
 		ddi_translations_fdi = hsw_ddi_translations_fdi;
 		ddi_translations_dp = hsw_ddi_translations_dp;
 		ddi_translations_edp = hsw_ddi_translations_dp;
+		ddi_translations_hdmi = hsw_ddi_translations_hdmi;
+		n_hdmi_entries = ARRAY_SIZE(hsw_ddi_translations_hdmi);
+		hdmi_800mV_0dB = 6;
 	} else {
 		WARN(1, "ddi translation table missing\n");
 		ddi_translations_edp = bdw_ddi_translations_dp;
 		ddi_translations_fdi = bdw_ddi_translations_fdi;
 		ddi_translations_dp = bdw_ddi_translations_dp;
+		ddi_translations_hdmi = bdw_ddi_translations_hdmi;
+		n_hdmi_entries = ARRAY_SIZE(bdw_ddi_translations_hdmi);
+		hdmi_800mV_0dB = 7;
 	}
 
 	switch (port) {
@@ -182,7 +241,10 @@ static void intel_prepare_ddi_buffers(struct drm_device *dev, enum port port)
 			ddi_translations = ddi_translations_dp;
 		break;
 	case PORT_E:
-		ddi_translations = ddi_translations_fdi;
+		if (ddi_translations_fdi)
+			ddi_translations = ddi_translations_fdi;
+		else
+			ddi_translations = ddi_translations_dp;
 		break;
 	default:
 		BUG();
@@ -190,14 +252,22 @@ static void intel_prepare_ddi_buffers(struct drm_device *dev, enum port port)
 
 	for (i = 0, reg = DDI_BUF_TRANS(port);
 	     i < ARRAY_SIZE(hsw_ddi_translations_fdi); i++) {
-		I915_WRITE(reg, ddi_translations[i]);
+		I915_WRITE(reg, ddi_translations[i].trans1);
 		reg += 4;
-	}
-	/* Entry 9 is for HDMI: */
-	for (i = 0; i < 2; i++) {
-		I915_WRITE(reg, hsw_ddi_translations_hdmi[hdmi_level * 2 + i]);
+		I915_WRITE(reg, ddi_translations[i].trans2);
 		reg += 4;
 	}
+
+	/* Choose a good default if VBT is badly populated */
+	if (hdmi_level == HDMI_LEVEL_SHIFT_UNKNOWN ||
+	    hdmi_level >= n_hdmi_entries)
+		hdmi_level = hdmi_800mV_0dB;
+
+	/* Entry 9 is for HDMI: */
+	I915_WRITE(reg, ddi_translations_hdmi[hdmi_level].trans1);
+	reg += 4;
+	I915_WRITE(reg, ddi_translations_hdmi[hdmi_level].trans2);
+	reg += 4;
 }
 
 /* Program DDI buffers translations for DP. By default, program ports A-D in DP
@@ -214,18 +284,6 @@ void intel_prepare_ddi(struct drm_device *dev)
 		intel_prepare_ddi_buffers(dev, port);
 }
 
-static const long hsw_ddi_buf_ctl_values[] = {
-	DDI_BUF_EMP_400MV_0DB_HSW,
-	DDI_BUF_EMP_400MV_3_5DB_HSW,
-	DDI_BUF_EMP_400MV_6DB_HSW,
-	DDI_BUF_EMP_400MV_9_5DB_HSW,
-	DDI_BUF_EMP_600MV_0DB_HSW,
-	DDI_BUF_EMP_600MV_3_5DB_HSW,
-	DDI_BUF_EMP_600MV_6DB_HSW,
-	DDI_BUF_EMP_800MV_0DB_HSW,
-	DDI_BUF_EMP_800MV_3_5DB_HSW
-};
-
 static void intel_wait_ddi_buf_idle(struct drm_i915_private *dev_priv,
 				    enum port port)
 {
@@ -285,7 +343,7 @@ void hsw_fdi_link_train(struct drm_crtc *crtc)
 
 	/* Start the training iterating through available voltages and emphasis,
 	 * testing each value twice. */
-	for (i = 0; i < ARRAY_SIZE(hsw_ddi_buf_ctl_values) * 2; i++) {
+	for (i = 0; i < ARRAY_SIZE(hsw_ddi_translations_fdi) * 2; i++) {
 		/* Configure DP_TP_CTL with auto-training */
 		I915_WRITE(DP_TP_CTL(PORT_E),
 					DP_TP_CTL_FDI_AUTOTRAIN |
@@ -300,7 +358,7 @@ void hsw_fdi_link_train(struct drm_crtc *crtc)
 		I915_WRITE(DDI_BUF_CTL(PORT_E),
 			   DDI_BUF_CTL_ENABLE |
 			   ((intel_crtc->config.fdi_lanes - 1) << 1) |
-			   hsw_ddi_buf_ctl_values[i / 2]);
+			   DDI_BUF_TRANS_SELECT(i / 2));
 		POSTING_READ(DDI_BUF_CTL(PORT_E));
 
 		udelay(600);
@@ -375,7 +433,7 @@ void intel_ddi_init_dp_buf_reg(struct intel_encoder *encoder)
 		enc_to_dig_port(&encoder->base);
 
 	intel_dp->DP = intel_dig_port->saved_port_bits |
-		DDI_BUF_CTL_ENABLE | DDI_BUF_EMP_400MV_0DB_HSW;
+		DDI_BUF_CTL_ENABLE | DDI_BUF_TRANS_SELECT(0);
 	intel_dp->DP |= DDI_PORT_WIDTH(intel_dp->lane_count);
 
 }
@@ -401,8 +459,29 @@ intel_ddi_get_crtc_encoder(struct drm_crtc *crtc)
 	return ret;
 }
 
+static struct intel_encoder *
+intel_ddi_get_crtc_new_encoder(struct intel_crtc *crtc)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct intel_encoder *intel_encoder, *ret = NULL;
+	int num_encoders = 0;
+
+	for_each_intel_encoder(dev, intel_encoder) {
+		if (intel_encoder->new_crtc == crtc) {
+			ret = intel_encoder;
+			num_encoders++;
+		}
+	}
+
+	WARN(num_encoders != 1, "%d encoders on crtc for pipe %c\n", num_encoders,
+	     pipe_name(crtc->pipe));
+
+	BUG_ON(ret == NULL);
+	return ret;
+}
+
 #define LC_FREQ 2700
-#define LC_FREQ_2K (LC_FREQ * 2000)
+#define LC_FREQ_2K U64_C(LC_FREQ * 2000)
 
 #define P_MIN 2
 #define P_MAX 64
@@ -414,7 +493,11 @@ intel_ddi_get_crtc_encoder(struct drm_crtc *crtc)
 #define VCO_MIN 2400
 #define VCO_MAX 4800
 
-#define ABS_DIFF(a, b) ((a > b) ? (a - b) : (b - a))
+#define abs_diff(a, b) ({			\
+	typeof(a) __a = (a);			\
+	typeof(b) __b = (b);			\
+	(void) (&__a == &__b);			\
+	__a > __b ? (__a - __b) : (__b - __a); })
 
 struct wrpll_rnp {
 	unsigned p, n2, r2;
@@ -524,9 +607,9 @@ static void wrpll_update_rnp(uint64_t freq2k, unsigned budget,
 	 */
 	a = freq2k * budget * p * r2;
 	b = freq2k * budget * best->p * best->r2;
-	diff = ABS_DIFF((freq2k * p * r2), (LC_FREQ_2K * n2));
-	diff_best = ABS_DIFF((freq2k * best->p * best->r2),
-			     (LC_FREQ_2K * best->n2));
+	diff = abs_diff(freq2k * p * r2, LC_FREQ_2K * n2);
+	diff_best = abs_diff(freq2k * best->p * best->r2,
+			     LC_FREQ_2K * best->n2);
 	c = 1000000 * diff;
 	d = 1000000 * diff_best;
 
@@ -587,8 +670,113 @@ static int intel_ddi_calc_wrpll_link(struct drm_i915_private *dev_priv,
 	return (refclk * n * 100) / (p * r);
 }
 
-void intel_ddi_clock_get(struct intel_encoder *encoder,
-			 struct intel_crtc_config *pipe_config)
+static int skl_calc_wrpll_link(struct drm_i915_private *dev_priv,
+			       uint32_t dpll)
+{
+	uint32_t cfgcr1_reg, cfgcr2_reg;
+	uint32_t cfgcr1_val, cfgcr2_val;
+	uint32_t p0, p1, p2, dco_freq;
+
+	cfgcr1_reg = GET_CFG_CR1_REG(dpll);
+	cfgcr2_reg = GET_CFG_CR2_REG(dpll);
+
+	cfgcr1_val = I915_READ(cfgcr1_reg);
+	cfgcr2_val = I915_READ(cfgcr2_reg);
+
+	p0 = cfgcr2_val & DPLL_CFGCR2_PDIV_MASK;
+	p2 = cfgcr2_val & DPLL_CFGCR2_KDIV_MASK;
+
+	if (cfgcr2_val &  DPLL_CFGCR2_QDIV_MODE(1))
+		p1 = (cfgcr2_val & DPLL_CFGCR2_QDIV_RATIO_MASK) >> 8;
+	else
+		p1 = 1;
+
+
+	switch (p0) {
+	case DPLL_CFGCR2_PDIV_1:
+		p0 = 1;
+		break;
+	case DPLL_CFGCR2_PDIV_2:
+		p0 = 2;
+		break;
+	case DPLL_CFGCR2_PDIV_3:
+		p0 = 3;
+		break;
+	case DPLL_CFGCR2_PDIV_7:
+		p0 = 7;
+		break;
+	}
+
+	switch (p2) {
+	case DPLL_CFGCR2_KDIV_5:
+		p2 = 5;
+		break;
+	case DPLL_CFGCR2_KDIV_2:
+		p2 = 2;
+		break;
+	case DPLL_CFGCR2_KDIV_3:
+		p2 = 3;
+		break;
+	case DPLL_CFGCR2_KDIV_1:
+		p2 = 1;
+		break;
+	}
+
+	dco_freq = (cfgcr1_val & DPLL_CFGCR1_DCO_INTEGER_MASK) * 24 * 1000;
+
+	dco_freq += (((cfgcr1_val & DPLL_CFGCR1_DCO_FRACTION_MASK) >> 9) * 24 *
+		1000) / 0x8000;
+
+	return dco_freq / (p0 * p1 * p2 * 5);
+}
+
+
+static void skl_ddi_clock_get(struct intel_encoder *encoder,
+				struct intel_crtc_config *pipe_config)
+{
+	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
+	int link_clock = 0;
+	uint32_t dpll_ctl1, dpll;
+
+	dpll = pipe_config->ddi_pll_sel;
+
+	dpll_ctl1 = I915_READ(DPLL_CTRL1);
+
+	if (dpll_ctl1 & DPLL_CTRL1_HDMI_MODE(dpll)) {
+		link_clock = skl_calc_wrpll_link(dev_priv, dpll);
+	} else {
+		link_clock = dpll_ctl1 & DPLL_CRTL1_LINK_RATE_MASK(dpll);
+		link_clock >>= DPLL_CRTL1_LINK_RATE_SHIFT(dpll);
+
+		switch (link_clock) {
+		case DPLL_CRTL1_LINK_RATE_810:
+			link_clock = 81000;
+			break;
+		case DPLL_CRTL1_LINK_RATE_1350:
+			link_clock = 135000;
+			break;
+		case DPLL_CRTL1_LINK_RATE_2700:
+			link_clock = 270000;
+			break;
+		default:
+			WARN(1, "Unsupported link rate\n");
+			break;
+		}
+		link_clock *= 2;
+	}
+
+	pipe_config->port_clock = link_clock;
+
+	if (pipe_config->has_dp_encoder)
+		pipe_config->adjusted_mode.crtc_clock =
+			intel_dotclock_calculate(pipe_config->port_clock,
+						 &pipe_config->dp_m_n);
+	else
+		pipe_config->adjusted_mode.crtc_clock = pipe_config->port_clock;
+}
+
+static void hsw_ddi_clock_get(struct intel_encoder *encoder,
+			      struct intel_crtc_config *pipe_config)
 {
 	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
 	int link_clock = 0;
@@ -643,9 +831,15 @@ void intel_ddi_clock_get(struct intel_encoder *encoder,
 		pipe_config->adjusted_mode.crtc_clock = pipe_config->port_clock;
 }
 
+void intel_ddi_clock_get(struct intel_encoder *encoder,
+			 struct intel_crtc_config *pipe_config)
+{
+	hsw_ddi_clock_get(encoder, pipe_config);
+}
+
 static void
-intel_ddi_calculate_wrpll(int clock /* in Hz */,
-			  unsigned *r2_out, unsigned *n2_out, unsigned *p_out)
+hsw_ddi_calculate_wrpll(int clock /* in Hz */,
+			unsigned *r2_out, unsigned *n2_out, unsigned *p_out)
 {
 	uint64_t freq2k;
 	unsigned p, n2, r2;
@@ -708,33 +902,23 @@ intel_ddi_calculate_wrpll(int clock /* in Hz */,
 	*r2_out = best.r2;
 }
 
-/*
- * Tries to find a PLL for the CRTC. If it finds, it increases the refcount and
- * stores it in intel_crtc->ddi_pll_sel, so other mode sets won't be able to
- * steal the selected PLL. You need to call intel_ddi_pll_enable to actually
- * enable the PLL.
- */
-bool intel_ddi_pll_select(struct intel_crtc *intel_crtc)
+static bool
+hsw_ddi_pll_select(struct intel_crtc *intel_crtc,
+		   struct intel_encoder *intel_encoder,
+		   int clock)
 {
-	struct drm_crtc *crtc = &intel_crtc->base;
-	struct intel_encoder *intel_encoder = intel_ddi_get_crtc_encoder(crtc);
-	int type = intel_encoder->type;
-	int clock = intel_crtc->config.port_clock;
-
-	intel_put_shared_dpll(intel_crtc);
-
-	if (type == INTEL_OUTPUT_HDMI) {
+	if (intel_encoder->type == INTEL_OUTPUT_HDMI) {
 		struct intel_shared_dpll *pll;
 		uint32_t val;
 		unsigned p, n2, r2;
 
-		intel_ddi_calculate_wrpll(clock * 1000, &r2, &n2, &p);
+		hsw_ddi_calculate_wrpll(clock * 1000, &r2, &n2, &p);
 
 		val = WRPLL_PLL_ENABLE | WRPLL_PLL_LCPLL |
 		      WRPLL_DIVIDER_REFERENCE(r2) | WRPLL_DIVIDER_FEEDBACK(n2) |
 		      WRPLL_DIVIDER_POST(p);
 
-		intel_crtc->config.dpll_hw_state.wrpll = val;
+		intel_crtc->new_config->dpll_hw_state.wrpll = val;
 
 		pll = intel_get_shared_dpll(intel_crtc);
 		if (pll == NULL) {
@@ -743,12 +927,255 @@ bool intel_ddi_pll_select(struct intel_crtc *intel_crtc)
 			return false;
 		}
 
-		intel_crtc->config.ddi_pll_sel = PORT_CLK_SEL_WRPLL(pll->id);
+		intel_crtc->new_config->ddi_pll_sel = PORT_CLK_SEL_WRPLL(pll->id);
+	}
+
+	return true;
+}
+
+struct skl_wrpll_params {
+	uint32_t        dco_fraction;
+	uint32_t        dco_integer;
+	uint32_t        qdiv_ratio;
+	uint32_t        qdiv_mode;
+	uint32_t        kdiv;
+	uint32_t        pdiv;
+	uint32_t        central_freq;
+};
+
+static void
+skl_ddi_calculate_wrpll(int clock /* in Hz */,
+			struct skl_wrpll_params *wrpll_params)
+{
+	uint64_t afe_clock = clock * 5; /* AFE Clock is 5x Pixel clock */
+	uint64_t dco_central_freq[3] = {8400000000ULL,
+					9000000000ULL,
+					9600000000ULL};
+	uint32_t min_dco_deviation = 400;
+	uint32_t min_dco_index = 3;
+	uint32_t P0[4] = {1, 2, 3, 7};
+	uint32_t P2[4] = {1, 2, 3, 5};
+	bool found = false;
+	uint32_t candidate_p = 0;
+	uint32_t candidate_p0[3] = {0}, candidate_p1[3] = {0};
+	uint32_t candidate_p2[3] = {0};
+	uint32_t dco_central_freq_deviation[3];
+	uint32_t i, P1, k, dco_count;
+	bool retry_with_odd = false;
+	uint64_t dco_freq;
+
+	/* Determine P0, P1 or P2 */
+	for (dco_count = 0; dco_count < 3; dco_count++) {
+		found = false;
+		candidate_p =
+			div64_u64(dco_central_freq[dco_count], afe_clock);
+		if (retry_with_odd == false)
+			candidate_p = (candidate_p % 2 == 0 ?
+				candidate_p : candidate_p + 1);
+
+		for (P1 = 1; P1 < candidate_p; P1++) {
+			for (i = 0; i < 4; i++) {
+				if (!(P0[i] != 1 || P1 == 1))
+					continue;
+
+				for (k = 0; k < 4; k++) {
+					if (P1 != 1 && P2[k] != 2)
+						continue;
+
+					if (candidate_p == P0[i] * P1 * P2[k]) {
+						/* Found possible P0, P1, P2 */
+						found = true;
+						candidate_p0[dco_count] = P0[i];
+						candidate_p1[dco_count] = P1;
+						candidate_p2[dco_count] = P2[k];
+						goto found;
+					}
+
+				}
+			}
+		}
+
+found:
+		if (found) {
+			dco_central_freq_deviation[dco_count] =
+				div64_u64(10000 *
+					  abs_diff((candidate_p * afe_clock),
+						   dco_central_freq[dco_count]),
+					  dco_central_freq[dco_count]);
+
+			if (dco_central_freq_deviation[dco_count] <
+				min_dco_deviation) {
+				min_dco_deviation =
+					dco_central_freq_deviation[dco_count];
+				min_dco_index = dco_count;
+			}
+		}
+
+		if (min_dco_index > 2 && dco_count == 2) {
+			retry_with_odd = true;
+			dco_count = 0;
+		}
+	}
+
+	if (min_dco_index > 2) {
+		WARN(1, "No valid values found for the given pixel clock\n");
+	} else {
+		 wrpll_params->central_freq = dco_central_freq[min_dco_index];
+
+		 switch (dco_central_freq[min_dco_index]) {
+		 case 9600000000ULL:
+			wrpll_params->central_freq = 0;
+			break;
+		 case 9000000000ULL:
+			wrpll_params->central_freq = 1;
+			break;
+		 case 8400000000ULL:
+			wrpll_params->central_freq = 3;
+		 }
+
+		 switch (candidate_p0[min_dco_index]) {
+		 case 1:
+			wrpll_params->pdiv = 0;
+			break;
+		 case 2:
+			wrpll_params->pdiv = 1;
+			break;
+		 case 3:
+			wrpll_params->pdiv = 2;
+			break;
+		 case 7:
+			wrpll_params->pdiv = 4;
+			break;
+		 default:
+			WARN(1, "Incorrect PDiv\n");
+		 }
+
+		 switch (candidate_p2[min_dco_index]) {
+		 case 5:
+			wrpll_params->kdiv = 0;
+			break;
+		 case 2:
+			wrpll_params->kdiv = 1;
+			break;
+		 case 3:
+			wrpll_params->kdiv = 2;
+			break;
+		 case 1:
+			wrpll_params->kdiv = 3;
+			break;
+		 default:
+			WARN(1, "Incorrect KDiv\n");
+		 }
+
+		 wrpll_params->qdiv_ratio = candidate_p1[min_dco_index];
+		 wrpll_params->qdiv_mode =
+			(wrpll_params->qdiv_ratio == 1) ? 0 : 1;
+
+		 dco_freq = candidate_p0[min_dco_index] *
+			 candidate_p1[min_dco_index] *
+			 candidate_p2[min_dco_index] * afe_clock;
+
+		/*
+		* Intermediate values are in Hz.
+		* Divide by MHz to match bsepc
+		*/
+		 wrpll_params->dco_integer = div_u64(dco_freq, (24 * MHz(1)));
+		 wrpll_params->dco_fraction =
+			 div_u64(((div_u64(dco_freq, 24) -
+				   wrpll_params->dco_integer * MHz(1)) * 0x8000), MHz(1));
+
+	}
+}
+
+
+static bool
+skl_ddi_pll_select(struct intel_crtc *intel_crtc,
+		   struct intel_encoder *intel_encoder,
+		   int clock)
+{
+	struct intel_shared_dpll *pll;
+	uint32_t ctrl1, cfgcr1, cfgcr2;
+
+	/*
+	 * See comment in intel_dpll_hw_state to understand why we always use 0
+	 * as the DPLL id in this function.
+	 */
+
+	ctrl1 = DPLL_CTRL1_OVERRIDE(0);
+
+	if (intel_encoder->type == INTEL_OUTPUT_HDMI) {
+		struct skl_wrpll_params wrpll_params = { 0, };
+
+		ctrl1 |= DPLL_CTRL1_HDMI_MODE(0);
+
+		skl_ddi_calculate_wrpll(clock * 1000, &wrpll_params);
+
+		cfgcr1 = DPLL_CFGCR1_FREQ_ENABLE |
+			 DPLL_CFGCR1_DCO_FRACTION(wrpll_params.dco_fraction) |
+			 wrpll_params.dco_integer;
+
+		cfgcr2 = DPLL_CFGCR2_QDIV_RATIO(wrpll_params.qdiv_ratio) |
+			 DPLL_CFGCR2_QDIV_MODE(wrpll_params.qdiv_mode) |
+			 DPLL_CFGCR2_KDIV(wrpll_params.kdiv) |
+			 DPLL_CFGCR2_PDIV(wrpll_params.pdiv) |
+			 wrpll_params.central_freq;
+	} else if (intel_encoder->type == INTEL_OUTPUT_DISPLAYPORT) {
+		struct drm_encoder *encoder = &intel_encoder->base;
+		struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
+
+		switch (intel_dp->link_bw) {
+		case DP_LINK_BW_1_62:
+			ctrl1 |= DPLL_CRTL1_LINK_RATE(DPLL_CRTL1_LINK_RATE_810, 0);
+			break;
+		case DP_LINK_BW_2_7:
+			ctrl1 |= DPLL_CRTL1_LINK_RATE(DPLL_CRTL1_LINK_RATE_1350, 0);
+			break;
+		case DP_LINK_BW_5_4:
+			ctrl1 |= DPLL_CRTL1_LINK_RATE(DPLL_CRTL1_LINK_RATE_2700, 0);
+			break;
+		}
+
+		cfgcr1 = cfgcr2 = 0;
+	} else /* eDP */
+		return true;
+
+	intel_crtc->new_config->dpll_hw_state.ctrl1 = ctrl1;
+	intel_crtc->new_config->dpll_hw_state.cfgcr1 = cfgcr1;
+	intel_crtc->new_config->dpll_hw_state.cfgcr2 = cfgcr2;
+
+	pll = intel_get_shared_dpll(intel_crtc);
+	if (pll == NULL) {
+		DRM_DEBUG_DRIVER("failed to find PLL for pipe %c\n",
+				 pipe_name(intel_crtc->pipe));
+		return false;
 	}
 
+	/* shared DPLL id 0 is DPLL 1 */
+	intel_crtc->new_config->ddi_pll_sel = pll->id + 1;
+
 	return true;
 }
 
+/*
+ * Tries to find a *shared* PLL for the CRTC and store it in
+ * intel_crtc->ddi_pll_sel.
+ *
+ * For private DPLLs, compute_config() should do the selection for us. This
+ * function should be folded into compute_config() eventually.
+ */
+bool intel_ddi_pll_select(struct intel_crtc *intel_crtc)
+{
+	struct drm_device *dev = intel_crtc->base.dev;
+	struct intel_encoder *intel_encoder =
+		intel_ddi_get_crtc_new_encoder(intel_crtc);
+	int clock = intel_crtc->new_config->port_clock;
+
+	if (IS_SKYLAKE(dev))
+		return skl_ddi_pll_select(intel_crtc, intel_encoder, clock);
+	else
+		return hsw_ddi_pll_select(intel_crtc, intel_encoder, clock);
+}
+
 void intel_ddi_set_pipe_settings(struct drm_crtc *crtc)
 {
 	struct drm_i915_private *dev_priv = crtc->dev->dev_private;
@@ -921,7 +1348,7 @@ bool intel_ddi_connector_get_hw_state(struct intel_connector *intel_connector)
 	uint32_t tmp;
 
 	power_domain = intel_display_port_power_domain(intel_encoder);
-	if (!intel_display_power_enabled(dev_priv, power_domain))
+	if (!intel_display_power_is_enabled(dev_priv, power_domain))
 		return false;
 
 	if (!intel_encoder->get_hw_state(intel_encoder, &pipe))
@@ -967,7 +1394,7 @@ bool intel_ddi_get_hw_state(struct intel_encoder *encoder,
 	int i;
 
 	power_domain = intel_display_port_power_domain(encoder);
-	if (!intel_display_power_enabled(dev_priv, power_domain))
+	if (!intel_display_power_is_enabled(dev_priv, power_domain))
 		return false;
 
 	tmp = I915_READ(DDI_BUF_CTL(port));
@@ -1038,27 +1465,53 @@ void intel_ddi_disable_pipe_clock(struct intel_crtc *intel_crtc)
 static void intel_ddi_pre_enable(struct intel_encoder *intel_encoder)
 {
 	struct drm_encoder *encoder = &intel_encoder->base;
-	struct drm_i915_private *dev_priv = encoder->dev->dev_private;
+	struct drm_device *dev = encoder->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *crtc = to_intel_crtc(encoder->crtc);
 	enum port port = intel_ddi_get_encoder_port(intel_encoder);
 	int type = intel_encoder->type;
 
-	if (crtc->config.has_audio) {
-		DRM_DEBUG_DRIVER("Audio on pipe %c on DDI\n",
-				 pipe_name(crtc->pipe));
-
-		/* write eld */
-		DRM_DEBUG_DRIVER("DDI audio: write eld information\n");
-		intel_write_eld(encoder, &crtc->config.adjusted_mode);
-	}
-
 	if (type == INTEL_OUTPUT_EDP) {
 		struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
 		intel_edp_panel_on(intel_dp);
 	}
 
-	WARN_ON(crtc->config.ddi_pll_sel == PORT_CLK_SEL_NONE);
-	I915_WRITE(PORT_CLK_SEL(port), crtc->config.ddi_pll_sel);
+	if (IS_SKYLAKE(dev)) {
+		uint32_t dpll = crtc->config.ddi_pll_sel;
+		uint32_t val;
+
+		/*
+		 * DPLL0 is used for eDP and is the only "private" DPLL (as
+		 * opposed to shared) on SKL
+		 */
+		if (type == INTEL_OUTPUT_EDP) {
+			WARN_ON(dpll != SKL_DPLL0);
+
+			val = I915_READ(DPLL_CTRL1);
+
+			val &= ~(DPLL_CTRL1_HDMI_MODE(dpll) |
+				 DPLL_CTRL1_SSC(dpll) |
+				 DPLL_CRTL1_LINK_RATE_MASK(dpll));
+			val |= crtc->config.dpll_hw_state.ctrl1 << (dpll * 6);
+
+			I915_WRITE(DPLL_CTRL1, val);
+			POSTING_READ(DPLL_CTRL1);
+		}
+
+		/* DDI -> PLL mapping  */
+		val = I915_READ(DPLL_CTRL2);
+
+		val &= ~(DPLL_CTRL2_DDI_CLK_OFF(port) |
+			DPLL_CTRL2_DDI_CLK_SEL_MASK(port));
+		val |= (DPLL_CTRL2_DDI_CLK_SEL(dpll, port) |
+			DPLL_CTRL2_DDI_SEL_OVERRIDE(port));
+
+		I915_WRITE(DPLL_CTRL2, val);
+
+	} else {
+		WARN_ON(crtc->config.ddi_pll_sel == PORT_CLK_SEL_NONE);
+		I915_WRITE(PORT_CLK_SEL(port), crtc->config.ddi_pll_sel);
+	}
 
 	if (type == INTEL_OUTPUT_DISPLAYPORT || type == INTEL_OUTPUT_EDP) {
 		struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
@@ -1068,7 +1521,7 @@ static void intel_ddi_pre_enable(struct intel_encoder *intel_encoder)
 		intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_ON);
 		intel_dp_start_link_train(intel_dp);
 		intel_dp_complete_link_train(intel_dp);
-		if (port != PORT_A)
+		if (port != PORT_A || INTEL_INFO(dev)->gen >= 9)
 			intel_dp_stop_link_train(intel_dp);
 	} else if (type == INTEL_OUTPUT_HDMI) {
 		struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(encoder);
@@ -1082,7 +1535,8 @@ static void intel_ddi_pre_enable(struct intel_encoder *intel_encoder)
 static void intel_ddi_post_disable(struct intel_encoder *intel_encoder)
 {
 	struct drm_encoder *encoder = &intel_encoder->base;
-	struct drm_i915_private *dev_priv = encoder->dev->dev_private;
+	struct drm_device *dev = encoder->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	enum port port = intel_ddi_get_encoder_port(intel_encoder);
 	int type = intel_encoder->type;
 	uint32_t val;
@@ -1110,7 +1564,11 @@ static void intel_ddi_post_disable(struct intel_encoder *intel_encoder)
 		intel_edp_panel_off(intel_dp);
 	}
 
-	I915_WRITE(PORT_CLK_SEL(port), PORT_CLK_SEL_NONE);
+	if (IS_SKYLAKE(dev))
+		I915_WRITE(DPLL_CTRL2, (I915_READ(DPLL_CTRL2) |
+					DPLL_CTRL2_DDI_CLK_OFF(port)));
+	else
+		I915_WRITE(PORT_CLK_SEL(port), PORT_CLK_SEL_NONE);
 }
 
 static void intel_enable_ddi(struct intel_encoder *intel_encoder)
@@ -1118,12 +1576,10 @@ static void intel_enable_ddi(struct intel_encoder *intel_encoder)
 	struct drm_encoder *encoder = &intel_encoder->base;
 	struct drm_crtc *crtc = encoder->crtc;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
 	struct drm_device *dev = encoder->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	enum port port = intel_ddi_get_encoder_port(intel_encoder);
 	int type = intel_encoder->type;
-	uint32_t tmp;
 
 	if (type == INTEL_OUTPUT_HDMI) {
 		struct intel_digital_port *intel_dig_port =
@@ -1139,18 +1595,16 @@ static void intel_enable_ddi(struct intel_encoder *intel_encoder)
 	} else if (type == INTEL_OUTPUT_EDP) {
 		struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
 
-		if (port == PORT_A)
+		if (port == PORT_A && INTEL_INFO(dev)->gen < 9)
 			intel_dp_stop_link_train(intel_dp);
 
 		intel_edp_backlight_on(intel_dp);
-		intel_edp_psr_enable(intel_dp);
+		intel_psr_enable(intel_dp);
 	}
 
 	if (intel_crtc->config.has_audio) {
 		intel_display_power_get(dev_priv, POWER_DOMAIN_AUDIO);
-		tmp = I915_READ(HSW_AUD_PIN_ELD_CP_VLD);
-		tmp |= ((AUDIO_OUTPUT_ENABLE_A | AUDIO_ELD_VALID_A) << (pipe * 4));
-		I915_WRITE(HSW_AUD_PIN_ELD_CP_VLD, tmp);
+		intel_audio_codec_enable(intel_encoder);
 	}
 }
 
@@ -1159,61 +1613,126 @@ static void intel_disable_ddi(struct intel_encoder *intel_encoder)
 	struct drm_encoder *encoder = &intel_encoder->base;
 	struct drm_crtc *crtc = encoder->crtc;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
 	int type = intel_encoder->type;
 	struct drm_device *dev = encoder->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	uint32_t tmp;
 
-	/* We can't touch HSW_AUD_PIN_ELD_CP_VLD uncionditionally because this
-	 * register is part of the power well on Haswell. */
 	if (intel_crtc->config.has_audio) {
-		tmp = I915_READ(HSW_AUD_PIN_ELD_CP_VLD);
-		tmp &= ~((AUDIO_OUTPUT_ENABLE_A | AUDIO_ELD_VALID_A) <<
-			 (pipe * 4));
-		I915_WRITE(HSW_AUD_PIN_ELD_CP_VLD, tmp);
+		intel_audio_codec_disable(intel_encoder);
 		intel_display_power_put(dev_priv, POWER_DOMAIN_AUDIO);
 	}
 
 	if (type == INTEL_OUTPUT_EDP) {
 		struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
 
-		intel_edp_psr_disable(intel_dp);
+		intel_psr_disable(intel_dp);
 		intel_edp_backlight_off(intel_dp);
 	}
 }
 
-int intel_ddi_get_cdclk_freq(struct drm_i915_private *dev_priv)
+static int skl_get_cdclk_freq(struct drm_i915_private *dev_priv)
+{
+	uint32_t lcpll1 = I915_READ(LCPLL1_CTL);
+	uint32_t cdctl = I915_READ(CDCLK_CTL);
+	uint32_t linkrate;
+
+	if (!(lcpll1 & LCPLL_PLL_ENABLE)) {
+		WARN(1, "LCPLL1 not enabled\n");
+		return 24000; /* 24MHz is the cd freq with NSSC ref */
+	}
+
+	if ((cdctl & CDCLK_FREQ_SEL_MASK) == CDCLK_FREQ_540)
+		return 540000;
+
+	linkrate = (I915_READ(DPLL_CTRL1) &
+		    DPLL_CRTL1_LINK_RATE_MASK(SKL_DPLL0)) >> 1;
+
+	if (linkrate == DPLL_CRTL1_LINK_RATE_2160 ||
+	    linkrate == DPLL_CRTL1_LINK_RATE_1080) {
+		/* vco 8640 */
+		switch (cdctl & CDCLK_FREQ_SEL_MASK) {
+		case CDCLK_FREQ_450_432:
+			return 432000;
+		case CDCLK_FREQ_337_308:
+			return 308570;
+		case CDCLK_FREQ_675_617:
+			return 617140;
+		default:
+			WARN(1, "Unknown cd freq selection\n");
+		}
+	} else {
+		/* vco 8100 */
+		switch (cdctl & CDCLK_FREQ_SEL_MASK) {
+		case CDCLK_FREQ_450_432:
+			return 450000;
+		case CDCLK_FREQ_337_308:
+			return 337500;
+		case CDCLK_FREQ_675_617:
+			return 675000;
+		default:
+			WARN(1, "Unknown cd freq selection\n");
+		}
+	}
+
+	/* error case, do as if DPLL0 isn't enabled */
+	return 24000;
+}
+
+static int bdw_get_cdclk_freq(struct drm_i915_private *dev_priv)
+{
+	uint32_t lcpll = I915_READ(LCPLL_CTL);
+	uint32_t freq = lcpll & LCPLL_CLK_FREQ_MASK;
+
+	if (lcpll & LCPLL_CD_SOURCE_FCLK)
+		return 800000;
+	else if (I915_READ(FUSE_STRAP) & HSW_CDCLK_LIMIT)
+		return 450000;
+	else if (freq == LCPLL_CLK_FREQ_450)
+		return 450000;
+	else if (freq == LCPLL_CLK_FREQ_54O_BDW)
+		return 540000;
+	else if (freq == LCPLL_CLK_FREQ_337_5_BDW)
+		return 337500;
+	else
+		return 675000;
+}
+
+static int hsw_get_cdclk_freq(struct drm_i915_private *dev_priv)
 {
 	struct drm_device *dev = dev_priv->dev;
 	uint32_t lcpll = I915_READ(LCPLL_CTL);
 	uint32_t freq = lcpll & LCPLL_CLK_FREQ_MASK;
 
-	if (lcpll & LCPLL_CD_SOURCE_FCLK) {
+	if (lcpll & LCPLL_CD_SOURCE_FCLK)
 		return 800000;
-	} else if (I915_READ(FUSE_STRAP) & HSW_CDCLK_LIMIT) {
+	else if (I915_READ(FUSE_STRAP) & HSW_CDCLK_LIMIT)
 		return 450000;
-	} else if (freq == LCPLL_CLK_FREQ_450) {
+	else if (freq == LCPLL_CLK_FREQ_450)
 		return 450000;
-	} else if (IS_HASWELL(dev)) {
-		if (IS_ULT(dev))
-			return 337500;
-		else
-			return 540000;
-	} else {
-		if (freq == LCPLL_CLK_FREQ_54O_BDW)
-			return 540000;
-		else if (freq == LCPLL_CLK_FREQ_337_5_BDW)
-			return 337500;
-		else
-			return 675000;
-	}
+	else if (IS_HSW_ULT(dev))
+		return 337500;
+	else
+		return 540000;
+}
+
+int intel_ddi_get_cdclk_freq(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+
+	if (IS_SKYLAKE(dev))
+		return skl_get_cdclk_freq(dev_priv);
+
+	if (IS_BROADWELL(dev))
+		return bdw_get_cdclk_freq(dev_priv);
+
+	/* Haswell */
+	return hsw_get_cdclk_freq(dev_priv);
 }
 
 static void hsw_ddi_pll_enable(struct drm_i915_private *dev_priv,
 			       struct intel_shared_dpll *pll)
 {
-	I915_WRITE(WRPLL_CTL(pll->id), pll->hw_state.wrpll);
+	I915_WRITE(WRPLL_CTL(pll->id), pll->config.hw_state.wrpll);
 	POSTING_READ(WRPLL_CTL(pll->id));
 	udelay(20);
 }
@@ -1234,7 +1753,7 @@ static bool hsw_ddi_pll_get_hw_state(struct drm_i915_private *dev_priv,
 {
 	uint32_t val;
 
-	if (!intel_display_power_enabled(dev_priv, POWER_DOMAIN_PLLS))
+	if (!intel_display_power_is_enabled(dev_priv, POWER_DOMAIN_PLLS))
 		return false;
 
 	val = I915_READ(WRPLL_CTL(pll->id));
@@ -1248,10 +1767,8 @@ static const char * const hsw_ddi_pll_names[] = {
 	"WRPLL 2",
 };
 
-void intel_ddi_pll_init(struct drm_device *dev)
+static void hsw_shared_dplls_init(struct drm_i915_private *dev_priv)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	uint32_t val = I915_READ(LCPLL_CTL);
 	int i;
 
 	dev_priv->num_shared_dpll = 2;
@@ -1264,20 +1781,158 @@ void intel_ddi_pll_init(struct drm_device *dev)
 		dev_priv->shared_dplls[i].get_hw_state =
 			hsw_ddi_pll_get_hw_state;
 	}
+}
 
-	/* The LCPLL register should be turned on by the BIOS. For now let's
-	 * just check its state and print errors in case something is wrong.
-	 * Don't even try to turn it on.
-	 */
+static const char * const skl_ddi_pll_names[] = {
+	"DPLL 1",
+	"DPLL 2",
+	"DPLL 3",
+};
+
+struct skl_dpll_regs {
+	u32 ctl, cfgcr1, cfgcr2;
+};
+
+/* this array is indexed by the *shared* pll id */
+static const struct skl_dpll_regs skl_dpll_regs[3] = {
+	{
+		/* DPLL 1 */
+		.ctl = LCPLL2_CTL,
+		.cfgcr1 = DPLL1_CFGCR1,
+		.cfgcr2 = DPLL1_CFGCR2,
+	},
+	{
+		/* DPLL 2 */
+		.ctl = WRPLL_CTL1,
+		.cfgcr1 = DPLL2_CFGCR1,
+		.cfgcr2 = DPLL2_CFGCR2,
+	},
+	{
+		/* DPLL 3 */
+		.ctl = WRPLL_CTL2,
+		.cfgcr1 = DPLL3_CFGCR1,
+		.cfgcr2 = DPLL3_CFGCR2,
+	},
+};
+
+static void skl_ddi_pll_enable(struct drm_i915_private *dev_priv,
+			       struct intel_shared_dpll *pll)
+{
+	uint32_t val;
+	unsigned int dpll;
+	const struct skl_dpll_regs *regs = skl_dpll_regs;
+
+	/* DPLL0 is not part of the shared DPLLs, so pll->id is 0 for DPLL1 */
+	dpll = pll->id + 1;
+
+	val = I915_READ(DPLL_CTRL1);
+
+	val &= ~(DPLL_CTRL1_HDMI_MODE(dpll) | DPLL_CTRL1_SSC(dpll) |
+		 DPLL_CRTL1_LINK_RATE_MASK(dpll));
+	val |= pll->config.hw_state.ctrl1 << (dpll * 6);
+
+	I915_WRITE(DPLL_CTRL1, val);
+	POSTING_READ(DPLL_CTRL1);
+
+	I915_WRITE(regs[pll->id].cfgcr1, pll->config.hw_state.cfgcr1);
+	I915_WRITE(regs[pll->id].cfgcr2, pll->config.hw_state.cfgcr2);
+	POSTING_READ(regs[pll->id].cfgcr1);
+	POSTING_READ(regs[pll->id].cfgcr2);
+
+	/* the enable bit is always bit 31 */
+	I915_WRITE(regs[pll->id].ctl,
+		   I915_READ(regs[pll->id].ctl) | LCPLL_PLL_ENABLE);
+
+	if (wait_for(I915_READ(DPLL_STATUS) & DPLL_LOCK(dpll), 5))
+		DRM_ERROR("DPLL %d not locked\n", dpll);
+}
+
+static void skl_ddi_pll_disable(struct drm_i915_private *dev_priv,
+				struct intel_shared_dpll *pll)
+{
+	const struct skl_dpll_regs *regs = skl_dpll_regs;
+
+	/* the enable bit is always bit 31 */
+	I915_WRITE(regs[pll->id].ctl,
+		   I915_READ(regs[pll->id].ctl) & ~LCPLL_PLL_ENABLE);
+	POSTING_READ(regs[pll->id].ctl);
+}
+
+static bool skl_ddi_pll_get_hw_state(struct drm_i915_private *dev_priv,
+				     struct intel_shared_dpll *pll,
+				     struct intel_dpll_hw_state *hw_state)
+{
+	uint32_t val;
+	unsigned int dpll;
+	const struct skl_dpll_regs *regs = skl_dpll_regs;
+
+	if (!intel_display_power_is_enabled(dev_priv, POWER_DOMAIN_PLLS))
+		return false;
+
+	/* DPLL0 is not part of the shared DPLLs, so pll->id is 0 for DPLL1 */
+	dpll = pll->id + 1;
+
+	val = I915_READ(regs[pll->id].ctl);
+	if (!(val & LCPLL_PLL_ENABLE))
+		return false;
+
+	val = I915_READ(DPLL_CTRL1);
+	hw_state->ctrl1 = (val >> (dpll * 6)) & 0x3f;
+
+	/* avoid reading back stale values if HDMI mode is not enabled */
+	if (val & DPLL_CTRL1_HDMI_MODE(dpll)) {
+		hw_state->cfgcr1 = I915_READ(regs[pll->id].cfgcr1);
+		hw_state->cfgcr2 = I915_READ(regs[pll->id].cfgcr2);
+	}
+
+	return true;
+}
+
+static void skl_shared_dplls_init(struct drm_i915_private *dev_priv)
+{
+	int i;
+
+	dev_priv->num_shared_dpll = 3;
+
+	for (i = 0; i < dev_priv->num_shared_dpll; i++) {
+		dev_priv->shared_dplls[i].id = i;
+		dev_priv->shared_dplls[i].name = skl_ddi_pll_names[i];
+		dev_priv->shared_dplls[i].disable = skl_ddi_pll_disable;
+		dev_priv->shared_dplls[i].enable = skl_ddi_pll_enable;
+		dev_priv->shared_dplls[i].get_hw_state =
+			skl_ddi_pll_get_hw_state;
+	}
+}
+
+void intel_ddi_pll_init(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	uint32_t val = I915_READ(LCPLL_CTL);
+
+	if (IS_SKYLAKE(dev))
+		skl_shared_dplls_init(dev_priv);
+	else
+		hsw_shared_dplls_init(dev_priv);
 
 	DRM_DEBUG_KMS("CDCLK running at %dKHz\n",
 		      intel_ddi_get_cdclk_freq(dev_priv));
 
-	if (val & LCPLL_CD_SOURCE_FCLK)
-		DRM_ERROR("CDCLK source is not LCPLL\n");
+	if (IS_SKYLAKE(dev)) {
+		if (!(I915_READ(LCPLL1_CTL) & LCPLL_PLL_ENABLE))
+			DRM_ERROR("LCPLL1 is disabled\n");
+	} else {
+		/*
+		 * The LCPLL register should be turned on by the BIOS. For now
+		 * let's just check its state and print errors in case
+		 * something is wrong.  Don't even try to turn it on.
+		 */
 
-	if (val & LCPLL_PLL_DISABLE)
-		DRM_ERROR("LCPLL is disabled\n");
+		if (val & LCPLL_CD_SOURCE_FCLK)
+			DRM_ERROR("CDCLK source is not LCPLL\n");
+
+		if (val & LCPLL_PLL_DISABLE)
+			DRM_ERROR("LCPLL is disabled\n");
+	}
 }
 
 void intel_ddi_prepare_link_retrain(struct drm_encoder *encoder)
@@ -1372,7 +2027,9 @@ void intel_ddi_get_config(struct intel_encoder *encoder,
 	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
 	enum transcoder cpu_transcoder = intel_crtc->config.cpu_transcoder;
+	struct intel_hdmi *intel_hdmi;
 	u32 temp, flags = 0;
+	struct drm_device *dev = dev_priv->dev;
 
 	temp = I915_READ(TRANS_DDI_FUNC_CTL(cpu_transcoder));
 	if (temp & TRANS_DDI_PHSYNC)
@@ -1406,6 +2063,11 @@ void intel_ddi_get_config(struct intel_encoder *encoder,
 	switch (temp & TRANS_DDI_MODE_SELECT_MASK) {
 	case TRANS_DDI_MODE_SELECT_HDMI:
 		pipe_config->has_hdmi_sink = true;
+		intel_hdmi = enc_to_intel_hdmi(&encoder->base);
+
+		if (intel_hdmi->infoframe_enabled(&encoder->base))
+			pipe_config->has_infoframe = true;
+		break;
 	case TRANS_DDI_MODE_SELECT_DVI:
 	case TRANS_DDI_MODE_SELECT_FDI:
 		break;
@@ -1418,9 +2080,9 @@ void intel_ddi_get_config(struct intel_encoder *encoder,
 		break;
 	}
 
-	if (intel_display_power_enabled(dev_priv, POWER_DOMAIN_AUDIO)) {
+	if (intel_display_power_is_enabled(dev_priv, POWER_DOMAIN_AUDIO)) {
 		temp = I915_READ(HSW_AUD_PIN_ELD_CP_VLD);
-		if (temp & (AUDIO_OUTPUT_ENABLE_A << (intel_crtc->pipe * 4)))
+		if (temp & AUDIO_OUTPUT_ENABLE(intel_crtc->pipe))
 			pipe_config->has_audio = true;
 	}
 
@@ -1444,7 +2106,10 @@ void intel_ddi_get_config(struct intel_encoder *encoder,
 		dev_priv->vbt.edp_bpp = pipe_config->pipe_bpp;
 	}
 
-	intel_ddi_clock_get(encoder, pipe_config);
+	if (INTEL_INFO(dev)->gen <= 8)
+		hsw_ddi_clock_get(encoder, pipe_config);
+	else
+		skl_ddi_clock_get(encoder, pipe_config);
 }
 
 static void intel_ddi_destroy(struct drm_encoder *encoder)
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index d8324c69fa86..e7a16f119a29 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -73,11 +73,6 @@ static const uint32_t intel_cursor_formats[] = {
 	DRM_FORMAT_ARGB8888,
 };
 
-#define DIV_ROUND_CLOSEST_ULL(ll, d)	\
-({ unsigned long long _tmp = (ll)+(d)/2; do_div(_tmp, d); _tmp; })
-
-static void intel_increase_pllclock(struct drm_device *dev,
-				    enum pipe pipe);
 static void intel_crtc_update_cursor(struct drm_crtc *crtc, bool on);
 
 static void i9xx_crtc_clock_get(struct intel_crtc *crtc,
@@ -91,15 +86,18 @@ static int intel_framebuffer_init(struct drm_device *dev,
 				  struct intel_framebuffer *ifb,
 				  struct drm_mode_fb_cmd2 *mode_cmd,
 				  struct drm_i915_gem_object *obj);
-static void intel_dp_set_m_n(struct intel_crtc *crtc);
 static void i9xx_set_pipeconf(struct intel_crtc *intel_crtc);
 static void intel_set_pipe_timings(struct intel_crtc *intel_crtc);
 static void intel_cpu_transcoder_set_m_n(struct intel_crtc *crtc,
-					 struct intel_link_m_n *m_n);
+					 struct intel_link_m_n *m_n,
+					 struct intel_link_m_n *m2_n2);
 static void ironlake_set_pipeconf(struct drm_crtc *crtc);
 static void haswell_set_pipeconf(struct drm_crtc *crtc);
 static void intel_set_pipe_csc(struct drm_crtc *crtc);
-static void vlv_prepare_pll(struct intel_crtc *crtc);
+static void vlv_prepare_pll(struct intel_crtc *crtc,
+			    const struct intel_crtc_config *pipe_config);
+static void chv_prepare_pll(struct intel_crtc *crtc,
+			    const struct intel_crtc_config *pipe_config);
 
 static struct intel_encoder *intel_find_encoder(struct intel_connector *connector, int pipe)
 {
@@ -410,25 +408,43 @@ static void vlv_clock(int refclk, intel_clock_t *clock)
 /**
  * Returns whether any output on the specified pipe is of the specified type
  */
-static bool intel_pipe_has_type(struct drm_crtc *crtc, int type)
+bool intel_pipe_has_type(struct intel_crtc *crtc, enum intel_output_type type)
 {
-	struct drm_device *dev = crtc->dev;
+	struct drm_device *dev = crtc->base.dev;
 	struct intel_encoder *encoder;
 
-	for_each_encoder_on_crtc(dev, crtc, encoder)
+	for_each_encoder_on_crtc(dev, &crtc->base, encoder)
 		if (encoder->type == type)
 			return true;
 
 	return false;
 }
 
-static const intel_limit_t *intel_ironlake_limit(struct drm_crtc *crtc,
+/**
+ * Returns whether any output on the specified pipe will have the specified
+ * type after a staged modeset is complete, i.e., the same as
+ * intel_pipe_has_type() but looking at encoder->new_crtc instead of
+ * encoder->crtc.
+ */
+static bool intel_pipe_will_have_type(struct intel_crtc *crtc, int type)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct intel_encoder *encoder;
+
+	for_each_intel_encoder(dev, encoder)
+		if (encoder->new_crtc == crtc && encoder->type == type)
+			return true;
+
+	return false;
+}
+
+static const intel_limit_t *intel_ironlake_limit(struct intel_crtc *crtc,
 						int refclk)
 {
-	struct drm_device *dev = crtc->dev;
+	struct drm_device *dev = crtc->base.dev;
 	const intel_limit_t *limit;
 
-	if (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS)) {
+	if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS)) {
 		if (intel_is_dual_link_lvds(dev)) {
 			if (refclk == 100000)
 				limit = &intel_limits_ironlake_dual_lvds_100m;
@@ -446,20 +462,20 @@ static const intel_limit_t *intel_ironlake_limit(struct drm_crtc *crtc,
 	return limit;
 }
 
-static const intel_limit_t *intel_g4x_limit(struct drm_crtc *crtc)
+static const intel_limit_t *intel_g4x_limit(struct intel_crtc *crtc)
 {
-	struct drm_device *dev = crtc->dev;
+	struct drm_device *dev = crtc->base.dev;
 	const intel_limit_t *limit;
 
-	if (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS)) {
+	if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS)) {
 		if (intel_is_dual_link_lvds(dev))
 			limit = &intel_limits_g4x_dual_channel_lvds;
 		else
 			limit = &intel_limits_g4x_single_channel_lvds;
-	} else if (intel_pipe_has_type(crtc, INTEL_OUTPUT_HDMI) ||
-		   intel_pipe_has_type(crtc, INTEL_OUTPUT_ANALOG)) {
+	} else if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_HDMI) ||
+		   intel_pipe_will_have_type(crtc, INTEL_OUTPUT_ANALOG)) {
 		limit = &intel_limits_g4x_hdmi;
-	} else if (intel_pipe_has_type(crtc, INTEL_OUTPUT_SDVO)) {
+	} else if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_SDVO)) {
 		limit = &intel_limits_g4x_sdvo;
 	} else /* The option is for other outputs */
 		limit = &intel_limits_i9xx_sdvo;
@@ -467,9 +483,9 @@ static const intel_limit_t *intel_g4x_limit(struct drm_crtc *crtc)
 	return limit;
 }
 
-static const intel_limit_t *intel_limit(struct drm_crtc *crtc, int refclk)
+static const intel_limit_t *intel_limit(struct intel_crtc *crtc, int refclk)
 {
-	struct drm_device *dev = crtc->dev;
+	struct drm_device *dev = crtc->base.dev;
 	const intel_limit_t *limit;
 
 	if (HAS_PCH_SPLIT(dev))
@@ -477,7 +493,7 @@ static const intel_limit_t *intel_limit(struct drm_crtc *crtc, int refclk)
 	else if (IS_G4X(dev)) {
 		limit = intel_g4x_limit(crtc);
 	} else if (IS_PINEVIEW(dev)) {
-		if (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS))
+		if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS))
 			limit = &intel_limits_pineview_lvds;
 		else
 			limit = &intel_limits_pineview_sdvo;
@@ -486,14 +502,14 @@ static const intel_limit_t *intel_limit(struct drm_crtc *crtc, int refclk)
 	} else if (IS_VALLEYVIEW(dev)) {
 		limit = &intel_limits_vlv;
 	} else if (!IS_GEN2(dev)) {
-		if (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS))
+		if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS))
 			limit = &intel_limits_i9xx_lvds;
 		else
 			limit = &intel_limits_i9xx_sdvo;
 	} else {
-		if (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS))
+		if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS))
 			limit = &intel_limits_i8xx_lvds;
-		else if (intel_pipe_has_type(crtc, INTEL_OUTPUT_DVO))
+		else if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_DVO))
 			limit = &intel_limits_i8xx_dvo;
 		else
 			limit = &intel_limits_i8xx_dac;
@@ -580,15 +596,15 @@ static bool intel_PLL_is_valid(struct drm_device *dev,
 }
 
 static bool
-i9xx_find_best_dpll(const intel_limit_t *limit, struct drm_crtc *crtc,
+i9xx_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc,
 		    int target, int refclk, intel_clock_t *match_clock,
 		    intel_clock_t *best_clock)
 {
-	struct drm_device *dev = crtc->dev;
+	struct drm_device *dev = crtc->base.dev;
 	intel_clock_t clock;
 	int err = target;
 
-	if (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS)) {
+	if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS)) {
 		/*
 		 * For LVDS just rely on its current settings for dual-channel.
 		 * We haven't figured out how to reliably set up different
@@ -641,15 +657,15 @@ i9xx_find_best_dpll(const intel_limit_t *limit, struct drm_crtc *crtc,
 }
 
 static bool
-pnv_find_best_dpll(const intel_limit_t *limit, struct drm_crtc *crtc,
+pnv_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc,
 		   int target, int refclk, intel_clock_t *match_clock,
 		   intel_clock_t *best_clock)
 {
-	struct drm_device *dev = crtc->dev;
+	struct drm_device *dev = crtc->base.dev;
 	intel_clock_t clock;
 	int err = target;
 
-	if (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS)) {
+	if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS)) {
 		/*
 		 * For LVDS just rely on its current settings for dual-channel.
 		 * We haven't figured out how to reliably set up different
@@ -700,11 +716,11 @@ pnv_find_best_dpll(const intel_limit_t *limit, struct drm_crtc *crtc,
 }
 
 static bool
-g4x_find_best_dpll(const intel_limit_t *limit, struct drm_crtc *crtc,
+g4x_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc,
 		   int target, int refclk, intel_clock_t *match_clock,
 		   intel_clock_t *best_clock)
 {
-	struct drm_device *dev = crtc->dev;
+	struct drm_device *dev = crtc->base.dev;
 	intel_clock_t clock;
 	int max_n;
 	bool found;
@@ -712,7 +728,7 @@ g4x_find_best_dpll(const intel_limit_t *limit, struct drm_crtc *crtc,
 	int err_most = (target >> 8) + (target >> 9);
 	found = false;
 
-	if (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS)) {
+	if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS)) {
 		if (intel_is_dual_link_lvds(dev))
 			clock.p2 = limit->p2.p2_fast;
 		else
@@ -757,11 +773,11 @@ g4x_find_best_dpll(const intel_limit_t *limit, struct drm_crtc *crtc,
 }
 
 static bool
-vlv_find_best_dpll(const intel_limit_t *limit, struct drm_crtc *crtc,
+vlv_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc,
 		   int target, int refclk, intel_clock_t *match_clock,
 		   intel_clock_t *best_clock)
 {
-	struct drm_device *dev = crtc->dev;
+	struct drm_device *dev = crtc->base.dev;
 	intel_clock_t clock;
 	unsigned int bestppm = 1000000;
 	/* min update 19.2 MHz */
@@ -814,11 +830,11 @@ vlv_find_best_dpll(const intel_limit_t *limit, struct drm_crtc *crtc,
 }
 
 static bool
-chv_find_best_dpll(const intel_limit_t *limit, struct drm_crtc *crtc,
+chv_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc,
 		   int target, int refclk, intel_clock_t *match_clock,
 		   intel_clock_t *best_clock)
 {
-	struct drm_device *dev = crtc->dev;
+	struct drm_device *dev = crtc->base.dev;
 	intel_clock_t clock;
 	uint64_t m2;
 	int found = false;
@@ -891,58 +907,6 @@ enum transcoder intel_pipe_to_cpu_transcoder(struct drm_i915_private *dev_priv,
 	return intel_crtc->config.cpu_transcoder;
 }
 
-static void g4x_wait_for_vblank(struct drm_device *dev, int pipe)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 frame, frame_reg = PIPE_FRMCOUNT_GM45(pipe);
-
-	frame = I915_READ(frame_reg);
-
-	if (wait_for(I915_READ_NOTRACE(frame_reg) != frame, 50))
-		WARN(1, "vblank wait timed out\n");
-}
-
-/**
- * intel_wait_for_vblank - wait for vblank on a given pipe
- * @dev: drm device
- * @pipe: pipe to wait for
- *
- * Wait for vblank to occur on a given pipe.  Needed for various bits of
- * mode setting code.
- */
-void intel_wait_for_vblank(struct drm_device *dev, int pipe)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int pipestat_reg = PIPESTAT(pipe);
-
-	if (IS_G4X(dev) || INTEL_INFO(dev)->gen >= 5) {
-		g4x_wait_for_vblank(dev, pipe);
-		return;
-	}
-
-	/* Clear existing vblank status. Note this will clear any other
-	 * sticky status fields as well.
-	 *
-	 * This races with i915_driver_irq_handler() with the result
-	 * that either function could miss a vblank event.  Here it is not
-	 * fatal, as we will either wait upon the next vblank interrupt or
-	 * timeout.  Generally speaking intel_wait_for_vblank() is only
-	 * called during modeset at which time the GPU should be idle and
-	 * should *not* be performing page flips and thus not waiting on
-	 * vblanks...
-	 * Currently, the result of us stealing a vblank from the irq
-	 * handler is that a single frame will be skipped during swapbuffers.
-	 */
-	I915_WRITE(pipestat_reg,
-		   I915_READ(pipestat_reg) | PIPE_VBLANK_INTERRUPT_STATUS);
-
-	/* Wait for vblank interrupt bit to set */
-	if (wait_for(I915_READ(pipestat_reg) &
-		     PIPE_VBLANK_INTERRUPT_STATUS,
-		     50))
-		DRM_DEBUG_KMS("vblank wait timed out\n");
-}
-
 static bool pipe_dsl_stopped(struct drm_device *dev, enum pipe pipe)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -964,8 +928,7 @@ static bool pipe_dsl_stopped(struct drm_device *dev, enum pipe pipe)
 
 /*
  * intel_wait_for_pipe_off - wait for pipe to turn off
- * @dev: drm device
- * @pipe: pipe to wait for
+ * @crtc: crtc whose pipe to wait for
  *
  * After disabling a pipe, we can't wait for vblank in the usual way,
  * spinning on the vblank interrupt status bit, since we won't actually
@@ -979,11 +942,12 @@ static bool pipe_dsl_stopped(struct drm_device *dev, enum pipe pipe)
  *   ends up stopping at the start of the next frame).
  *
  */
-void intel_wait_for_pipe_off(struct drm_device *dev, int pipe)
+static void intel_wait_for_pipe_off(struct intel_crtc *crtc)
 {
+	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	enum transcoder cpu_transcoder = intel_pipe_to_cpu_transcoder(dev_priv,
-								      pipe);
+	enum transcoder cpu_transcoder = crtc->config.cpu_transcoder;
+	enum pipe pipe = crtc->pipe;
 
 	if (INTEL_INFO(dev)->gen >= 4) {
 		int reg = PIPECONF(cpu_transcoder);
@@ -1189,30 +1153,43 @@ void assert_fdi_rx_pll(struct drm_i915_private *dev_priv,
 	     state_string(state), state_string(cur_state));
 }
 
-static void assert_panel_unlocked(struct drm_i915_private *dev_priv,
-				  enum pipe pipe)
+void assert_panel_unlocked(struct drm_i915_private *dev_priv,
+			   enum pipe pipe)
 {
-	int pp_reg, lvds_reg;
+	struct drm_device *dev = dev_priv->dev;
+	int pp_reg;
 	u32 val;
 	enum pipe panel_pipe = PIPE_A;
 	bool locked = true;
 
-	if (HAS_PCH_SPLIT(dev_priv->dev)) {
+	if (WARN_ON(HAS_DDI(dev)))
+		return;
+
+	if (HAS_PCH_SPLIT(dev)) {
+		u32 port_sel;
+
 		pp_reg = PCH_PP_CONTROL;
-		lvds_reg = PCH_LVDS;
+		port_sel = I915_READ(PCH_PP_ON_DELAYS) & PANEL_PORT_SELECT_MASK;
+
+		if (port_sel == PANEL_PORT_SELECT_LVDS &&
+		    I915_READ(PCH_LVDS) & LVDS_PIPEB_SELECT)
+			panel_pipe = PIPE_B;
+		/* XXX: else fix for eDP */
+	} else if (IS_VALLEYVIEW(dev)) {
+		/* presumably write lock depends on pipe, not port select */
+		pp_reg = VLV_PIPE_PP_CONTROL(pipe);
+		panel_pipe = pipe;
 	} else {
 		pp_reg = PP_CONTROL;
-		lvds_reg = LVDS;
+		if (I915_READ(LVDS) & LVDS_PIPEB_SELECT)
+			panel_pipe = PIPE_B;
 	}
 
 	val = I915_READ(pp_reg);
 	if (!(val & PANEL_POWER_ON) ||
-	    ((val & PANEL_UNLOCK_REGS) == PANEL_UNLOCK_REGS))
+	    ((val & PANEL_UNLOCK_MASK) == PANEL_UNLOCK_REGS))
 		locked = false;
 
-	if (I915_READ(lvds_reg) & LVDS_PIPEB_SELECT)
-		panel_pipe = PIPE_B;
-
 	WARN(panel_pipe == pipe && locked,
 	     "panel assertion failure, pipe %c regs locked\n",
 	     pipe_name(pipe));
@@ -1245,11 +1222,12 @@ void assert_pipe(struct drm_i915_private *dev_priv,
 	enum transcoder cpu_transcoder = intel_pipe_to_cpu_transcoder(dev_priv,
 								      pipe);
 
-	/* if we need the pipe A quirk it must be always on */
-	if (pipe == PIPE_A && dev_priv->quirks & QUIRK_PIPEA_FORCE)
+	/* if we need the pipe quirk it must be always on */
+	if ((pipe == PIPE_A && dev_priv->quirks & QUIRK_PIPEA_FORCE) ||
+	    (pipe == PIPE_B && dev_priv->quirks & QUIRK_PIPEB_FORCE))
 		state = true;
 
-	if (!intel_display_power_enabled(dev_priv,
+	if (!intel_display_power_is_enabled(dev_priv,
 				POWER_DOMAIN_TRANSCODER(cpu_transcoder))) {
 		cur_state = false;
 	} else {
@@ -1300,7 +1278,7 @@ static void assert_planes_disabled(struct drm_i915_private *dev_priv,
 	}
 
 	/* Need to check both planes against the pipe */
-	for_each_pipe(i) {
+	for_each_pipe(dev_priv, i) {
 		reg = DSPCNTR(i);
 		val = I915_READ(reg);
 		cur_pipe = (val & DISPPLANE_SEL_PIPE_MASK) >>
@@ -1318,7 +1296,14 @@ static void assert_sprites_disabled(struct drm_i915_private *dev_priv,
 	int reg, sprite;
 	u32 val;
 
-	if (IS_VALLEYVIEW(dev)) {
+	if (INTEL_INFO(dev)->gen >= 9) {
+		for_each_sprite(pipe, sprite) {
+			val = I915_READ(PLANE_CTL(pipe, sprite));
+			WARN(val & PLANE_CTL_ENABLE,
+			     "plane %d assertion failure, should be off on pipe %c but is still active\n",
+			     sprite, pipe_name(pipe));
+		}
+	} else if (IS_VALLEYVIEW(dev)) {
 		for_each_sprite(pipe, sprite) {
 			reg = SPCNTR(pipe, sprite);
 			val = I915_READ(reg);
@@ -1341,6 +1326,12 @@ static void assert_sprites_disabled(struct drm_i915_private *dev_priv,
 	}
 }
 
+static void assert_vblank_disabled(struct drm_crtc *crtc)
+{
+	if (WARN_ON(drm_crtc_vblank_get(crtc) == 0))
+		drm_crtc_vblank_put(crtc);
+}
+
 static void ibx_assert_pch_refclk_enabled(struct drm_i915_private *dev_priv)
 {
 	u32 val;
@@ -1513,40 +1504,13 @@ static void intel_init_dpio(struct drm_device *dev)
 	}
 }
 
-static void intel_reset_dpio(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	if (IS_CHERRYVIEW(dev)) {
-		enum dpio_phy phy;
-		u32 val;
-
-		for (phy = DPIO_PHY0; phy < I915_NUM_PHYS_VLV; phy++) {
-			/* Poll for phypwrgood signal */
-			if (wait_for(I915_READ(DISPLAY_PHY_STATUS) &
-						PHY_POWERGOOD(phy), 1))
-				DRM_ERROR("Display PHY %d is not power up\n", phy);
-
-			/*
-			 * Deassert common lane reset for PHY.
-			 *
-			 * This should only be done on init and resume from S3
-			 * with both PLLs disabled, or we risk losing DPIO and
-			 * PLL synchronization.
-			 */
-			val = I915_READ(DISPLAY_PHY_CONTROL);
-			I915_WRITE(DISPLAY_PHY_CONTROL,
-				PHY_COM_LANE_RESET_DEASSERT(phy, val));
-		}
-	}
-}
-
-static void vlv_enable_pll(struct intel_crtc *crtc)
+static void vlv_enable_pll(struct intel_crtc *crtc,
+			   const struct intel_crtc_config *pipe_config)
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int reg = DPLL(crtc->pipe);
-	u32 dpll = crtc->config.dpll_hw_state.dpll;
+	u32 dpll = pipe_config->dpll_hw_state.dpll;
 
 	assert_pipe_disabled(dev_priv, crtc->pipe);
 
@@ -1554,7 +1518,7 @@ static void vlv_enable_pll(struct intel_crtc *crtc)
 	BUG_ON(!IS_VALLEYVIEW(dev_priv->dev));
 
 	/* PLL is protected by panel, make sure we can write it */
-	if (IS_MOBILE(dev_priv->dev) && !IS_I830(dev_priv->dev))
+	if (IS_MOBILE(dev_priv->dev))
 		assert_panel_unlocked(dev_priv, crtc->pipe);
 
 	I915_WRITE(reg, dpll);
@@ -1564,7 +1528,7 @@ static void vlv_enable_pll(struct intel_crtc *crtc)
 	if (wait_for(((I915_READ(reg) & DPLL_LOCK_VLV) == DPLL_LOCK_VLV), 1))
 		DRM_ERROR("DPLL %d failed to lock\n", crtc->pipe);
 
-	I915_WRITE(DPLL_MD(crtc->pipe), crtc->config.dpll_hw_state.dpll_md);
+	I915_WRITE(DPLL_MD(crtc->pipe), pipe_config->dpll_hw_state.dpll_md);
 	POSTING_READ(DPLL_MD(crtc->pipe));
 
 	/* We do this three times for luck */
@@ -1579,7 +1543,8 @@ static void vlv_enable_pll(struct intel_crtc *crtc)
 	udelay(150); /* wait for warmup */
 }
 
-static void chv_enable_pll(struct intel_crtc *crtc)
+static void chv_enable_pll(struct intel_crtc *crtc,
+			   const struct intel_crtc_config *pipe_config)
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1604,19 +1569,31 @@ static void chv_enable_pll(struct intel_crtc *crtc)
 	udelay(1);
 
 	/* Enable PLL */
-	I915_WRITE(DPLL(pipe), crtc->config.dpll_hw_state.dpll);
+	I915_WRITE(DPLL(pipe), pipe_config->dpll_hw_state.dpll);
 
 	/* Check PLL is locked */
 	if (wait_for(((I915_READ(DPLL(pipe)) & DPLL_LOCK_VLV) == DPLL_LOCK_VLV), 1))
 		DRM_ERROR("PLL %d failed to lock\n", pipe);
 
 	/* not sure when this should be written */
-	I915_WRITE(DPLL_MD(pipe), crtc->config.dpll_hw_state.dpll_md);
+	I915_WRITE(DPLL_MD(pipe), pipe_config->dpll_hw_state.dpll_md);
 	POSTING_READ(DPLL_MD(pipe));
 
 	mutex_unlock(&dev_priv->dpio_lock);
 }
 
+static int intel_num_dvo_pipes(struct drm_device *dev)
+{
+	struct intel_crtc *crtc;
+	int count = 0;
+
+	for_each_intel_crtc(dev, crtc)
+		count += crtc->active &&
+			intel_pipe_has_type(crtc, INTEL_OUTPUT_DVO);
+
+	return count;
+}
+
 static void i9xx_enable_pll(struct intel_crtc *crtc)
 {
 	struct drm_device *dev = crtc->base.dev;
@@ -1633,7 +1610,18 @@ static void i9xx_enable_pll(struct intel_crtc *crtc)
 	if (IS_MOBILE(dev) && !IS_I830(dev))
 		assert_panel_unlocked(dev_priv, crtc->pipe);
 
-	I915_WRITE(reg, dpll);
+	/* Enable DVO 2x clock on both PLLs if necessary */
+	if (IS_I830(dev) && intel_num_dvo_pipes(dev) > 0) {
+		/*
+		 * It appears to be important that we don't enable this
+		 * for the current pipe before otherwise configuring the
+		 * PLL. No idea how this should be handled if multiple
+		 * DVO outputs are enabled simultaneosly.
+		 */
+		dpll |= DPLL_DVO_2X_MODE;
+		I915_WRITE(DPLL(!crtc->pipe),
+			   I915_READ(DPLL(!crtc->pipe)) | DPLL_DVO_2X_MODE);
+	}
 
 	/* Wait for the clocks to stabilize. */
 	POSTING_READ(reg);
@@ -1672,10 +1660,25 @@ static void i9xx_enable_pll(struct intel_crtc *crtc)
  *
  * Note!  This is for pre-ILK only.
  */
-static void i9xx_disable_pll(struct drm_i915_private *dev_priv, enum pipe pipe)
+static void i9xx_disable_pll(struct intel_crtc *crtc)
 {
-	/* Don't disable pipe A or pipe A PLLs if needed */
-	if (pipe == PIPE_A && (dev_priv->quirks & QUIRK_PIPEA_FORCE))
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum pipe pipe = crtc->pipe;
+
+	/* Disable DVO 2x clock on both PLLs if necessary */
+	if (IS_I830(dev) &&
+	    intel_pipe_has_type(crtc, INTEL_OUTPUT_DVO) &&
+	    intel_num_dvo_pipes(dev) == 1) {
+		I915_WRITE(DPLL(PIPE_B),
+			   I915_READ(DPLL(PIPE_B)) & ~DPLL_DVO_2X_MODE);
+		I915_WRITE(DPLL(PIPE_A),
+			   I915_READ(DPLL(PIPE_A)) & ~DPLL_DVO_2X_MODE);
+	}
+
+	/* Don't disable pipe or pipe PLLs if needed */
+	if ((pipe == PIPE_A && dev_priv->quirks & QUIRK_PIPEA_FORCE) ||
+	    (pipe == PIPE_B && dev_priv->quirks & QUIRK_PIPEB_FORCE))
 		return;
 
 	/* Make sure the pipe isn't still relying on us */
@@ -1712,7 +1715,7 @@ static void chv_disable_pll(struct drm_i915_private *dev_priv, enum pipe pipe)
 	assert_pipe_disabled(dev_priv, pipe);
 
 	/* Set PLL en = 0 */
-	val = DPLL_SSC_REF_CLOCK_CHV;
+	val = DPLL_SSC_REF_CLOCK_CHV | DPLL_REFA_CLK_ENABLE_VLV;
 	if (pipe != PIPE_A)
 		val |= DPLL_INTEGRATED_CRI_CLK_VLV;
 	I915_WRITE(DPLL(pipe), val);
@@ -1776,7 +1779,7 @@ static void intel_prepare_shared_dpll(struct intel_crtc *crtc)
 	if (WARN_ON(pll == NULL))
 		return;
 
-	WARN_ON(!pll->refcount);
+	WARN_ON(!pll->config.crtc_mask);
 	if (pll->active == 0) {
 		DRM_DEBUG_DRIVER("setting up %s\n", pll->name);
 		WARN_ON(pll->on);
@@ -1803,10 +1806,10 @@ static void intel_enable_shared_dpll(struct intel_crtc *crtc)
 	if (WARN_ON(pll == NULL))
 		return;
 
-	if (WARN_ON(pll->refcount == 0))
+	if (WARN_ON(pll->config.crtc_mask == 0))
 		return;
 
-	DRM_DEBUG_KMS("enable %s (active %d, on? %d)for crtc %d\n",
+	DRM_DEBUG_KMS("enable %s (active %d, on? %d) for crtc %d\n",
 		      pll->name, pll->active, pll->on,
 		      crtc->base.base.id);
 
@@ -1824,7 +1827,7 @@ static void intel_enable_shared_dpll(struct intel_crtc *crtc)
 	pll->on = true;
 }
 
-void intel_disable_shared_dpll(struct intel_crtc *crtc)
+static void intel_disable_shared_dpll(struct intel_crtc *crtc)
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1835,7 +1838,7 @@ void intel_disable_shared_dpll(struct intel_crtc *crtc)
 	if (WARN_ON(pll == NULL))
 	       return;
 
-	if (WARN_ON(pll->refcount == 0))
+	if (WARN_ON(pll->config.crtc_mask == 0))
 		return;
 
 	DRM_DEBUG_KMS("disable %s (active %d, on? %d) for crtc %d\n",
@@ -1868,7 +1871,7 @@ static void ironlake_enable_pch_transcoder(struct drm_i915_private *dev_priv,
 	uint32_t reg, val, pipeconf_val;
 
 	/* PCH only available on ILK+ */
-	BUG_ON(INTEL_INFO(dev)->gen < 5);
+	BUG_ON(!HAS_PCH_SPLIT(dev));
 
 	/* Make sure PCH DPLL is enabled */
 	assert_shared_dpll_enabled(dev_priv,
@@ -1903,7 +1906,7 @@ static void ironlake_enable_pch_transcoder(struct drm_i915_private *dev_priv,
 	val &= ~TRANS_INTERLACE_MASK;
 	if ((pipeconf_val & PIPECONF_INTERLACE_MASK) == PIPECONF_INTERLACED_ILK)
 		if (HAS_PCH_IBX(dev_priv->dev) &&
-		    intel_pipe_has_type(crtc, INTEL_OUTPUT_SDVO))
+		    intel_pipe_has_type(intel_crtc, INTEL_OUTPUT_SDVO))
 			val |= TRANS_LEGACY_INTERLACED_ILK;
 		else
 			val |= TRANS_INTERLACED;
@@ -1921,7 +1924,7 @@ static void lpt_enable_pch_transcoder(struct drm_i915_private *dev_priv,
 	u32 val, pipeconf_val;
 
 	/* PCH only available on ILK+ */
-	BUG_ON(INTEL_INFO(dev_priv->dev)->gen < 5);
+	BUG_ON(!HAS_PCH_SPLIT(dev_priv->dev));
 
 	/* FDI must be feeding us bits for PCH ports */
 	assert_fdi_tx_enabled(dev_priv, (enum pipe) cpu_transcoder);
@@ -2026,7 +2029,7 @@ static void intel_enable_pipe(struct intel_crtc *crtc)
 	 * need the check.
 	 */
 	if (!HAS_PCH_SPLIT(dev_priv->dev))
-		if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_DSI))
+		if (intel_pipe_has_type(crtc, INTEL_OUTPUT_DSI))
 			assert_dsi_pll_enabled(dev_priv);
 		else
 			assert_pll_enabled(dev_priv, pipe);
@@ -2043,8 +2046,8 @@ static void intel_enable_pipe(struct intel_crtc *crtc)
 	reg = PIPECONF(cpu_transcoder);
 	val = I915_READ(reg);
 	if (val & PIPECONF_ENABLE) {
-		WARN_ON(!(pipe == PIPE_A &&
-			  dev_priv->quirks & QUIRK_PIPEA_FORCE));
+		WARN_ON(!((pipe == PIPE_A && dev_priv->quirks & QUIRK_PIPEA_FORCE) ||
+			  (pipe == PIPE_B && dev_priv->quirks & QUIRK_PIPEB_FORCE)));
 		return;
 	}
 
@@ -2054,21 +2057,19 @@ static void intel_enable_pipe(struct intel_crtc *crtc)
 
 /**
  * intel_disable_pipe - disable a pipe, asserting requirements
- * @dev_priv: i915 private structure
- * @pipe: pipe to disable
- *
- * Disable @pipe, making sure that various hardware specific requirements
- * are met, if applicable, e.g. plane disabled, panel fitter off, etc.
+ * @crtc: crtc whose pipes is to be disabled
  *
- * @pipe should be %PIPE_A or %PIPE_B.
+ * Disable the pipe of @crtc, making sure that various hardware
+ * specific requirements are met, if applicable, e.g. plane
+ * disabled, panel fitter off, etc.
  *
  * Will wait until the pipe has shut down before returning.
  */
-static void intel_disable_pipe(struct drm_i915_private *dev_priv,
-			       enum pipe pipe)
+static void intel_disable_pipe(struct intel_crtc *crtc)
 {
-	enum transcoder cpu_transcoder = intel_pipe_to_cpu_transcoder(dev_priv,
-								      pipe);
+	struct drm_i915_private *dev_priv = crtc->base.dev->dev_private;
+	enum transcoder cpu_transcoder = crtc->config.cpu_transcoder;
+	enum pipe pipe = crtc->pipe;
 	int reg;
 	u32 val;
 
@@ -2080,17 +2081,26 @@ static void intel_disable_pipe(struct drm_i915_private *dev_priv,
 	assert_cursor_disabled(dev_priv, pipe);
 	assert_sprites_disabled(dev_priv, pipe);
 
-	/* Don't disable pipe A or pipe A PLLs if needed */
-	if (pipe == PIPE_A && (dev_priv->quirks & QUIRK_PIPEA_FORCE))
-		return;
-
 	reg = PIPECONF(cpu_transcoder);
 	val = I915_READ(reg);
 	if ((val & PIPECONF_ENABLE) == 0)
 		return;
 
-	I915_WRITE(reg, val & ~PIPECONF_ENABLE);
-	intel_wait_for_pipe_off(dev_priv->dev, pipe);
+	/*
+	 * Double wide has implications for planes
+	 * so best keep it disabled when not needed.
+	 */
+	if (crtc->config.double_wide)
+		val &= ~PIPECONF_DOUBLE_WIDE;
+
+	/* Don't disable pipe or pipe PLLs if needed */
+	if (!(pipe == PIPE_A && dev_priv->quirks & QUIRK_PIPEA_FORCE) &&
+	    !(pipe == PIPE_B && dev_priv->quirks & QUIRK_PIPEB_FORCE))
+		val &= ~PIPECONF_ENABLE;
+
+	I915_WRITE(reg, val);
+	if ((val & PIPECONF_ENABLE) == 0)
+		intel_wait_for_pipe_off(crtc);
 }
 
 /*
@@ -2109,35 +2119,28 @@ void intel_flush_primary_plane(struct drm_i915_private *dev_priv,
 
 /**
  * intel_enable_primary_hw_plane - enable the primary plane on a given pipe
- * @dev_priv: i915 private structure
- * @plane: plane to enable
- * @pipe: pipe being fed
+ * @plane:  plane to be enabled
+ * @crtc: crtc for the plane
  *
- * Enable @plane on @pipe, making sure that @pipe is running first.
+ * Enable @plane on @crtc, making sure that the pipe is running first.
  */
-static void intel_enable_primary_hw_plane(struct drm_i915_private *dev_priv,
-					  enum plane plane, enum pipe pipe)
+static void intel_enable_primary_hw_plane(struct drm_plane *plane,
+					  struct drm_crtc *crtc)
 {
-	struct drm_device *dev = dev_priv->dev;
-	struct intel_crtc *intel_crtc =
-		to_intel_crtc(dev_priv->pipe_to_crtc_mapping[pipe]);
-	int reg;
-	u32 val;
+	struct drm_device *dev = plane->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 
 	/* If the pipe isn't enabled, we can't pump pixels and may hang */
-	assert_pipe_enabled(dev_priv, pipe);
+	assert_pipe_enabled(dev_priv, intel_crtc->pipe);
 
 	if (intel_crtc->primary_enabled)
 		return;
 
 	intel_crtc->primary_enabled = true;
 
-	reg = DSPCNTR(plane);
-	val = I915_READ(reg);
-	WARN_ON(val & DISPLAY_PLANE_ENABLE);
-
-	I915_WRITE(reg, val | DISPLAY_PLANE_ENABLE);
-	intel_flush_primary_plane(dev_priv, plane);
+	dev_priv->display.update_primary_plane(crtc, plane->fb,
+					       crtc->x, crtc->y);
 
 	/*
 	 * BDW signals flip done immediately if the plane
@@ -2150,31 +2153,27 @@ static void intel_enable_primary_hw_plane(struct drm_i915_private *dev_priv,
 
 /**
  * intel_disable_primary_hw_plane - disable the primary hardware plane
- * @dev_priv: i915 private structure
- * @plane: plane to disable
- * @pipe: pipe consuming the data
+ * @plane: plane to be disabled
+ * @crtc: crtc for the plane
  *
- * Disable @plane; should be an independent operation.
+ * Disable @plane on @crtc, making sure that the pipe is running first.
  */
-static void intel_disable_primary_hw_plane(struct drm_i915_private *dev_priv,
-					   enum plane plane, enum pipe pipe)
+static void intel_disable_primary_hw_plane(struct drm_plane *plane,
+					   struct drm_crtc *crtc)
 {
-	struct intel_crtc *intel_crtc =
-		to_intel_crtc(dev_priv->pipe_to_crtc_mapping[pipe]);
-	int reg;
-	u32 val;
+	struct drm_device *dev = plane->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+
+	assert_pipe_enabled(dev_priv, intel_crtc->pipe);
 
 	if (!intel_crtc->primary_enabled)
 		return;
 
 	intel_crtc->primary_enabled = false;
 
-	reg = DSPCNTR(plane);
-	val = I915_READ(reg);
-	WARN_ON((val & DISPLAY_PLANE_ENABLE) == 0);
-
-	I915_WRITE(reg, val & ~DISPLAY_PLANE_ENABLE);
-	intel_flush_primary_plane(dev_priv, plane);
+	dev_priv->display.update_primary_plane(crtc, plane->fb,
+					       crtc->x, crtc->y);
 }
 
 static bool need_vtd_wa(struct drm_device *dev)
@@ -2195,11 +2194,13 @@ static int intel_align_height(struct drm_device *dev, int height, bool tiled)
 }
 
 int
-intel_pin_and_fence_fb_obj(struct drm_device *dev,
-			   struct drm_i915_gem_object *obj,
+intel_pin_and_fence_fb_obj(struct drm_plane *plane,
+			   struct drm_framebuffer *fb,
 			   struct intel_engine_cs *pipelined)
 {
+	struct drm_device *dev = fb->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_i915_gem_object *obj = intel_fb_obj(fb);
 	u32 alignment;
 	int ret;
 
@@ -2207,7 +2208,9 @@ intel_pin_and_fence_fb_obj(struct drm_device *dev,
 
 	switch (obj->tiling_mode) {
 	case I915_TILING_NONE:
-		if (IS_BROADWATER(dev) || IS_CRESTLINE(dev))
+		if (INTEL_INFO(dev)->gen >= 9)
+			alignment = 256 * 1024;
+		else if (IS_BROADWATER(dev) || IS_CRESTLINE(dev))
 			alignment = 128 * 1024;
 		else if (INTEL_INFO(dev)->gen >= 4)
 			alignment = 4 * 1024;
@@ -2215,8 +2218,12 @@ intel_pin_and_fence_fb_obj(struct drm_device *dev,
 			alignment = 64 * 1024;
 		break;
 	case I915_TILING_X:
-		/* pin() will align the object as required by fence */
-		alignment = 0;
+		if (INTEL_INFO(dev)->gen >= 9)
+			alignment = 256 * 1024;
+		else {
+			/* pin() will align the object as required by fence */
+			alignment = 0;
+		}
 		break;
 	case I915_TILING_Y:
 		WARN(1, "Y tiled bo slipped through, driver bug!\n");
@@ -2376,6 +2383,7 @@ static void intel_find_plane_obj(struct intel_crtc *intel_crtc,
 				 struct intel_plane_config *plane_config)
 {
 	struct drm_device *dev = intel_crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_crtc *c;
 	struct intel_crtc *i;
 	struct drm_i915_gem_object *obj;
@@ -2407,6 +2415,9 @@ static void intel_find_plane_obj(struct intel_crtc *intel_crtc,
 			continue;
 
 		if (i915_gem_obj_ggtt_offset(obj) == plane_config->base) {
+			if (obj->tiling_mode != I915_TILING_NONE)
+				dev_priv->preserve_bios_swizzle = true;
+
 			drm_framebuffer_reference(c->primary->fb);
 			intel_crtc->base.primary->fb = c->primary->fb;
 			obj->frontbuffer_bits |= INTEL_FRONTBUFFER_PRIMARY(intel_crtc->pipe);
@@ -2422,16 +2433,52 @@ static void i9xx_update_primary_plane(struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct drm_i915_gem_object *obj = intel_fb_obj(fb);
+	struct drm_i915_gem_object *obj;
 	int plane = intel_crtc->plane;
 	unsigned long linear_offset;
 	u32 dspcntr;
-	u32 reg;
+	u32 reg = DSPCNTR(plane);
+	int pixel_size;
+
+	if (!intel_crtc->primary_enabled) {
+		I915_WRITE(reg, 0);
+		if (INTEL_INFO(dev)->gen >= 4)
+			I915_WRITE(DSPSURF(plane), 0);
+		else
+			I915_WRITE(DSPADDR(plane), 0);
+		POSTING_READ(reg);
+		return;
+	}
+
+	obj = intel_fb_obj(fb);
+	if (WARN_ON(obj == NULL))
+		return;
+
+	pixel_size = drm_format_plane_cpp(fb->pixel_format, 0);
+
+	dspcntr = DISPPLANE_GAMMA_ENABLE;
+
+	dspcntr |= DISPLAY_PLANE_ENABLE;
+
+	if (INTEL_INFO(dev)->gen < 4) {
+		if (intel_crtc->pipe == PIPE_B)
+			dspcntr |= DISPPLANE_SEL_PIPE_B;
+
+		/* pipesrc and dspsize control the size that is scaled from,
+		 * which should always be the user's requested size.
+		 */
+		I915_WRITE(DSPSIZE(plane),
+			   ((intel_crtc->config.pipe_src_h - 1) << 16) |
+			   (intel_crtc->config.pipe_src_w - 1));
+		I915_WRITE(DSPPOS(plane), 0);
+	} else if (IS_CHERRYVIEW(dev) && plane == PLANE_B) {
+		I915_WRITE(PRIMSIZE(plane),
+			   ((intel_crtc->config.pipe_src_h - 1) << 16) |
+			   (intel_crtc->config.pipe_src_w - 1));
+		I915_WRITE(PRIMPOS(plane), 0);
+		I915_WRITE(PRIMCNSTALPHA(plane), 0);
+	}
 
-	reg = DSPCNTR(plane);
-	dspcntr = I915_READ(reg);
-	/* Mask out pixel format bits in case we change it */
-	dspcntr &= ~DISPPLANE_PIXFORMAT_MASK;
 	switch (fb->pixel_format) {
 	case DRM_FORMAT_C8:
 		dspcntr |= DISPPLANE_8BPP;
@@ -2463,30 +2510,40 @@ static void i9xx_update_primary_plane(struct drm_crtc *crtc,
 		BUG();
 	}
 
-	if (INTEL_INFO(dev)->gen >= 4) {
-		if (obj->tiling_mode != I915_TILING_NONE)
-			dspcntr |= DISPPLANE_TILED;
-		else
-			dspcntr &= ~DISPPLANE_TILED;
-	}
+	if (INTEL_INFO(dev)->gen >= 4 &&
+	    obj->tiling_mode != I915_TILING_NONE)
+		dspcntr |= DISPPLANE_TILED;
 
 	if (IS_G4X(dev))
 		dspcntr |= DISPPLANE_TRICKLE_FEED_DISABLE;
 
-	I915_WRITE(reg, dspcntr);
-
-	linear_offset = y * fb->pitches[0] + x * (fb->bits_per_pixel / 8);
+	linear_offset = y * fb->pitches[0] + x * pixel_size;
 
 	if (INTEL_INFO(dev)->gen >= 4) {
 		intel_crtc->dspaddr_offset =
 			intel_gen4_compute_page_offset(&x, &y, obj->tiling_mode,
-						       fb->bits_per_pixel / 8,
+						       pixel_size,
 						       fb->pitches[0]);
 		linear_offset -= intel_crtc->dspaddr_offset;
 	} else {
 		intel_crtc->dspaddr_offset = linear_offset;
 	}
 
+	if (to_intel_plane(crtc->primary)->rotation == BIT(DRM_ROTATE_180)) {
+		dspcntr |= DISPPLANE_ROTATE_180;
+
+		x += (intel_crtc->config.pipe_src_w - 1);
+		y += (intel_crtc->config.pipe_src_h - 1);
+
+		/* Finding the last pixel of the last line of the display
+		data and adding to linear_offset*/
+		linear_offset +=
+			(intel_crtc->config.pipe_src_h - 1) * fb->pitches[0] +
+			(intel_crtc->config.pipe_src_w - 1) * pixel_size;
+	}
+
+	I915_WRITE(reg, dspcntr);
+
 	DRM_DEBUG_KMS("Writing base %08lX %08lX %d %d %d\n",
 		      i915_gem_obj_ggtt_offset(obj), linear_offset, x, y,
 		      fb->pitches[0]);
@@ -2508,16 +2565,33 @@ static void ironlake_update_primary_plane(struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct drm_i915_gem_object *obj = intel_fb_obj(fb);
+	struct drm_i915_gem_object *obj;
 	int plane = intel_crtc->plane;
 	unsigned long linear_offset;
 	u32 dspcntr;
-	u32 reg;
+	u32 reg = DSPCNTR(plane);
+	int pixel_size;
+
+	if (!intel_crtc->primary_enabled) {
+		I915_WRITE(reg, 0);
+		I915_WRITE(DSPSURF(plane), 0);
+		POSTING_READ(reg);
+		return;
+	}
+
+	obj = intel_fb_obj(fb);
+	if (WARN_ON(obj == NULL))
+		return;
+
+	pixel_size = drm_format_plane_cpp(fb->pixel_format, 0);
+
+	dspcntr = DISPPLANE_GAMMA_ENABLE;
+
+	dspcntr |= DISPLAY_PLANE_ENABLE;
+
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+		dspcntr |= DISPPLANE_PIPE_CSC_ENABLE;
 
-	reg = DSPCNTR(plane);
-	dspcntr = I915_READ(reg);
-	/* Mask out pixel format bits in case we change it */
-	dspcntr &= ~DISPPLANE_PIXFORMAT_MASK;
 	switch (fb->pixel_format) {
 	case DRM_FORMAT_C8:
 		dspcntr |= DISPPLANE_8BPP;
@@ -2547,22 +2621,32 @@ static void ironlake_update_primary_plane(struct drm_crtc *crtc,
 
 	if (obj->tiling_mode != I915_TILING_NONE)
 		dspcntr |= DISPPLANE_TILED;
-	else
-		dspcntr &= ~DISPPLANE_TILED;
 
-	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
-		dspcntr &= ~DISPPLANE_TRICKLE_FEED_DISABLE;
-	else
+	if (!IS_HASWELL(dev) && !IS_BROADWELL(dev))
 		dspcntr |= DISPPLANE_TRICKLE_FEED_DISABLE;
 
-	I915_WRITE(reg, dspcntr);
-
-	linear_offset = y * fb->pitches[0] + x * (fb->bits_per_pixel / 8);
+	linear_offset = y * fb->pitches[0] + x * pixel_size;
 	intel_crtc->dspaddr_offset =
 		intel_gen4_compute_page_offset(&x, &y, obj->tiling_mode,
-					       fb->bits_per_pixel / 8,
+					       pixel_size,
 					       fb->pitches[0]);
 	linear_offset -= intel_crtc->dspaddr_offset;
+	if (to_intel_plane(crtc->primary)->rotation == BIT(DRM_ROTATE_180)) {
+		dspcntr |= DISPPLANE_ROTATE_180;
+
+		if (!IS_HASWELL(dev) && !IS_BROADWELL(dev)) {
+			x += (intel_crtc->config.pipe_src_w - 1);
+			y += (intel_crtc->config.pipe_src_h - 1);
+
+			/* Finding the last pixel of the last line of the display
+			data and adding to linear_offset*/
+			linear_offset +=
+				(intel_crtc->config.pipe_src_h - 1) * fb->pitches[0] +
+				(intel_crtc->config.pipe_src_w - 1) * pixel_size;
+		}
+	}
+
+	I915_WRITE(reg, dspcntr);
 
 	DRM_DEBUG_KMS("Writing base %08lX %08lX %d %d %d\n",
 		      i915_gem_obj_ggtt_offset(obj), linear_offset, x, y,
@@ -2579,6 +2663,92 @@ static void ironlake_update_primary_plane(struct drm_crtc *crtc,
 	POSTING_READ(reg);
 }
 
+static void skylake_update_primary_plane(struct drm_crtc *crtc,
+					 struct drm_framebuffer *fb,
+					 int x, int y)
+{
+	struct drm_device *dev = crtc->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	struct intel_framebuffer *intel_fb;
+	struct drm_i915_gem_object *obj;
+	int pipe = intel_crtc->pipe;
+	u32 plane_ctl, stride;
+
+	if (!intel_crtc->primary_enabled) {
+		I915_WRITE(PLANE_CTL(pipe, 0), 0);
+		I915_WRITE(PLANE_SURF(pipe, 0), 0);
+		POSTING_READ(PLANE_CTL(pipe, 0));
+		return;
+	}
+
+	plane_ctl = PLANE_CTL_ENABLE |
+		    PLANE_CTL_PIPE_GAMMA_ENABLE |
+		    PLANE_CTL_PIPE_CSC_ENABLE;
+
+	switch (fb->pixel_format) {
+	case DRM_FORMAT_RGB565:
+		plane_ctl |= PLANE_CTL_FORMAT_RGB_565;
+		break;
+	case DRM_FORMAT_XRGB8888:
+		plane_ctl |= PLANE_CTL_FORMAT_XRGB_8888;
+		break;
+	case DRM_FORMAT_XBGR8888:
+		plane_ctl |= PLANE_CTL_ORDER_RGBX;
+		plane_ctl |= PLANE_CTL_FORMAT_XRGB_8888;
+		break;
+	case DRM_FORMAT_XRGB2101010:
+		plane_ctl |= PLANE_CTL_FORMAT_XRGB_2101010;
+		break;
+	case DRM_FORMAT_XBGR2101010:
+		plane_ctl |= PLANE_CTL_ORDER_RGBX;
+		plane_ctl |= PLANE_CTL_FORMAT_XRGB_2101010;
+		break;
+	default:
+		BUG();
+	}
+
+	intel_fb = to_intel_framebuffer(fb);
+	obj = intel_fb->obj;
+
+	/*
+	 * The stride is either expressed as a multiple of 64 bytes chunks for
+	 * linear buffers or in number of tiles for tiled buffers.
+	 */
+	switch (obj->tiling_mode) {
+	case I915_TILING_NONE:
+		stride = fb->pitches[0] >> 6;
+		break;
+	case I915_TILING_X:
+		plane_ctl |= PLANE_CTL_TILED_X;
+		stride = fb->pitches[0] >> 9;
+		break;
+	default:
+		BUG();
+	}
+
+	plane_ctl |= PLANE_CTL_PLANE_GAMMA_DISABLE;
+	if (to_intel_plane(crtc->primary)->rotation == BIT(DRM_ROTATE_180))
+		plane_ctl |= PLANE_CTL_ROTATE_180;
+
+	I915_WRITE(PLANE_CTL(pipe, 0), plane_ctl);
+
+	DRM_DEBUG_KMS("Writing base %08lX %d,%d,%d,%d pitch=%d\n",
+		      i915_gem_obj_ggtt_offset(obj),
+		      x, y, fb->width, fb->height,
+		      fb->pitches[0]);
+
+	I915_WRITE(PLANE_POS(pipe, 0), 0);
+	I915_WRITE(PLANE_OFFSET(pipe, 0), (y << 16) | x);
+	I915_WRITE(PLANE_SIZE(pipe, 0),
+		   (intel_crtc->config.pipe_src_h - 1) << 16 |
+		   (intel_crtc->config.pipe_src_w - 1));
+	I915_WRITE(PLANE_STRIDE(pipe, 0), stride);
+	I915_WRITE(PLANE_SURF(pipe, 0), i915_gem_obj_ggtt_offset(obj));
+
+	POSTING_READ(PLANE_SURF(pipe, 0));
+}
+
 /* Assume fb object is pinned & idle & fenced and just update base pointers */
 static int
 intel_pipe_set_base_atomic(struct drm_crtc *crtc, struct drm_framebuffer *fb,
@@ -2589,32 +2759,16 @@ intel_pipe_set_base_atomic(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 
 	if (dev_priv->display.disable_fbc)
 		dev_priv->display.disable_fbc(dev);
-	intel_increase_pllclock(dev, to_intel_crtc(crtc)->pipe);
 
 	dev_priv->display.update_primary_plane(crtc, fb, x, y);
 
 	return 0;
 }
 
-void intel_display_handle_reset(struct drm_device *dev)
+static void intel_complete_page_flips(struct drm_device *dev)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_crtc *crtc;
 
-	/*
-	 * Flips in the rings have been nuked by the reset,
-	 * so complete all pending flips so that user space
-	 * will get its events and not get stuck.
-	 *
-	 * Also update the base address of all primary
-	 * planes to the the last fb to make sure we're
-	 * showing the correct fb after a reset.
-	 *
-	 * Need to make two loops over the crtcs so that we
-	 * don't try to grab a crtc mutex before the
-	 * pending_flip_queue really got woken up.
-	 */
-
 	for_each_crtc(dev, crtc) {
 		struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 		enum plane plane = intel_crtc->plane;
@@ -2622,6 +2776,12 @@ void intel_display_handle_reset(struct drm_device *dev)
 		intel_prepare_page_flip(dev, plane);
 		intel_finish_page_flip_plane(dev, plane);
 	}
+}
+
+static void intel_update_primary_planes(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_crtc *crtc;
 
 	for_each_crtc(dev, crtc) {
 		struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
@@ -2641,6 +2801,79 @@ void intel_display_handle_reset(struct drm_device *dev)
 	}
 }
 
+void intel_prepare_reset(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct intel_crtc *crtc;
+
+	/* no reset support for gen2 */
+	if (IS_GEN2(dev))
+		return;
+
+	/* reset doesn't touch the display */
+	if (INTEL_INFO(dev)->gen >= 5 || IS_G4X(dev))
+		return;
+
+	drm_modeset_lock_all(dev);
+
+	/*
+	 * Disabling the crtcs gracefully seems nicer. Also the
+	 * g33 docs say we should at least disable all the planes.
+	 */
+	for_each_intel_crtc(dev, crtc) {
+		if (crtc->active)
+			dev_priv->display.crtc_disable(&crtc->base);
+	}
+}
+
+void intel_finish_reset(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = to_i915(dev);
+
+	/*
+	 * Flips in the rings will be nuked by the reset,
+	 * so complete all pending flips so that user space
+	 * will get its events and not get stuck.
+	 */
+	intel_complete_page_flips(dev);
+
+	/* no reset support for gen2 */
+	if (IS_GEN2(dev))
+		return;
+
+	/* reset doesn't touch the display */
+	if (INTEL_INFO(dev)->gen >= 5 || IS_G4X(dev)) {
+		/*
+		 * Flips in the rings have been nuked by the reset,
+		 * so update the base address of all primary
+		 * planes to the the last fb to make sure we're
+		 * showing the correct fb after a reset.
+		 */
+		intel_update_primary_planes(dev);
+		return;
+	}
+
+	/*
+	 * The display has been reset as well,
+	 * so need a full re-initialization.
+	 */
+	intel_runtime_pm_disable_interrupts(dev_priv);
+	intel_runtime_pm_enable_interrupts(dev_priv);
+
+	intel_modeset_init_hw(dev);
+
+	spin_lock_irq(&dev_priv->irq_lock);
+	if (dev_priv->display.hpd_irq_setup)
+		dev_priv->display.hpd_irq_setup(dev);
+	spin_unlock_irq(&dev_priv->irq_lock);
+
+	intel_modeset_setup_hw_state(dev, true);
+
+	intel_hpd_init(dev_priv);
+
+	drm_modeset_unlock_all(dev);
+}
+
 static int
 intel_finish_fb(struct drm_framebuffer *old_fb)
 {
@@ -2669,20 +2902,58 @@ static bool intel_crtc_has_pending_flip(struct drm_crtc *crtc)
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	unsigned long flags;
 	bool pending;
 
 	if (i915_reset_in_progress(&dev_priv->gpu_error) ||
 	    intel_crtc->reset_counter != atomic_read(&dev_priv->gpu_error.reset_counter))
 		return false;
 
-	spin_lock_irqsave(&dev->event_lock, flags);
+	spin_lock_irq(&dev->event_lock);
 	pending = to_intel_crtc(crtc)->unpin_work != NULL;
-	spin_unlock_irqrestore(&dev->event_lock, flags);
+	spin_unlock_irq(&dev->event_lock);
 
 	return pending;
 }
 
+static void intel_update_pipe_size(struct intel_crtc *crtc)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	const struct drm_display_mode *adjusted_mode;
+
+	if (!i915.fastboot)
+		return;
+
+	/*
+	 * Update pipe size and adjust fitter if needed: the reason for this is
+	 * that in compute_mode_changes we check the native mode (not the pfit
+	 * mode) to see if we can flip rather than do a full mode set. In the
+	 * fastboot case, we'll flip, but if we don't update the pipesrc and
+	 * pfit state, we'll end up with a big fb scanned out into the wrong
+	 * sized surface.
+	 *
+	 * To fix this properly, we need to hoist the checks up into
+	 * compute_mode_changes (or above), check the actual pfit state and
+	 * whether the platform allows pfit disable with pipe active, and only
+	 * then update the pipesrc and pfit state, even on the flip path.
+	 */
+
+	adjusted_mode = &crtc->config.adjusted_mode;
+
+	I915_WRITE(PIPESRC(crtc->pipe),
+		   ((adjusted_mode->crtc_hdisplay - 1) << 16) |
+		   (adjusted_mode->crtc_vdisplay - 1));
+	if (!crtc->config.pch_pfit.enabled &&
+	    (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS) ||
+	     intel_pipe_has_type(crtc, INTEL_OUTPUT_EDP))) {
+		I915_WRITE(PF_CTL(crtc->pipe), 0);
+		I915_WRITE(PF_WIN_POS(crtc->pipe), 0);
+		I915_WRITE(PF_WIN_SZ(crtc->pipe), 0);
+	}
+	crtc->config.pipe_src_w = adjusted_mode->crtc_hdisplay;
+	crtc->config.pipe_src_h = adjusted_mode->crtc_vdisplay;
+}
+
 static int
 intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 		    struct drm_framebuffer *fb)
@@ -2692,7 +2963,6 @@ intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	enum pipe pipe = intel_crtc->pipe;
 	struct drm_framebuffer *old_fb = crtc->primary->fb;
-	struct drm_i915_gem_object *obj = intel_fb_obj(fb);
 	struct drm_i915_gem_object *old_obj = intel_fb_obj(old_fb);
 	int ret;
 
@@ -2715,9 +2985,9 @@ intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 	}
 
 	mutex_lock(&dev->struct_mutex);
-	ret = intel_pin_and_fence_fb_obj(dev, obj, NULL);
+	ret = intel_pin_and_fence_fb_obj(crtc->primary, fb, NULL);
 	if (ret == 0)
-		i915_gem_track_fb(old_obj, obj,
+		i915_gem_track_fb(old_obj, intel_fb_obj(fb),
 				  INTEL_FRONTBUFFER_PRIMARY(pipe));
 	mutex_unlock(&dev->struct_mutex);
 	if (ret != 0) {
@@ -2725,37 +2995,6 @@ intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 		return ret;
 	}
 
-	/*
-	 * Update pipe size and adjust fitter if needed: the reason for this is
-	 * that in compute_mode_changes we check the native mode (not the pfit
-	 * mode) to see if we can flip rather than do a full mode set. In the
-	 * fastboot case, we'll flip, but if we don't update the pipesrc and
-	 * pfit state, we'll end up with a big fb scanned out into the wrong
-	 * sized surface.
-	 *
-	 * To fix this properly, we need to hoist the checks up into
-	 * compute_mode_changes (or above), check the actual pfit state and
-	 * whether the platform allows pfit disable with pipe active, and only
-	 * then update the pipesrc and pfit state, even on the flip path.
-	 */
-	if (i915.fastboot) {
-		const struct drm_display_mode *adjusted_mode =
-			&intel_crtc->config.adjusted_mode;
-
-		I915_WRITE(PIPESRC(intel_crtc->pipe),
-			   ((adjusted_mode->crtc_hdisplay - 1) << 16) |
-			   (adjusted_mode->crtc_vdisplay - 1));
-		if (!intel_crtc->config.pch_pfit.enabled &&
-		    (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS) ||
-		     intel_pipe_has_type(crtc, INTEL_OUTPUT_EDP))) {
-			I915_WRITE(PF_CTL(intel_crtc->pipe), 0);
-			I915_WRITE(PF_WIN_POS(intel_crtc->pipe), 0);
-			I915_WRITE(PF_WIN_SZ(intel_crtc->pipe), 0);
-		}
-		intel_crtc->config.pipe_src_w = adjusted_mode->crtc_hdisplay;
-		intel_crtc->config.pipe_src_h = adjusted_mode->crtc_vdisplay;
-	}
-
 	dev_priv->display.update_primary_plane(crtc, fb, x, y);
 
 	if (intel_crtc->active)
@@ -3346,23 +3585,53 @@ bool intel_has_pending_fb_unpin(struct drm_device *dev)
 	return false;
 }
 
+static void page_flip_completed(struct intel_crtc *intel_crtc)
+{
+	struct drm_i915_private *dev_priv = to_i915(intel_crtc->base.dev);
+	struct intel_unpin_work *work = intel_crtc->unpin_work;
+
+	/* ensure that the unpin work is consistent wrt ->pending. */
+	smp_rmb();
+	intel_crtc->unpin_work = NULL;
+
+	if (work->event)
+		drm_send_vblank_event(intel_crtc->base.dev,
+				      intel_crtc->pipe,
+				      work->event);
+
+	drm_crtc_vblank_put(&intel_crtc->base);
+
+	wake_up_all(&dev_priv->pending_flip_queue);
+	queue_work(dev_priv->wq, &work->work);
+
+	trace_i915_flip_complete(intel_crtc->plane,
+				 work->pending_flip_obj);
+}
+
 void intel_crtc_wait_for_pending_flips(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (crtc->primary->fb == NULL)
-		return;
-
 	WARN_ON(waitqueue_active(&dev_priv->pending_flip_queue));
+	if (WARN_ON(wait_event_timeout(dev_priv->pending_flip_queue,
+				       !intel_crtc_has_pending_flip(crtc),
+				       60*HZ) == 0)) {
+		struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 
-	WARN_ON(wait_event_timeout(dev_priv->pending_flip_queue,
-				   !intel_crtc_has_pending_flip(crtc),
-				   60*HZ) == 0);
+		spin_lock_irq(&dev->event_lock);
+		if (intel_crtc->unpin_work) {
+			WARN_ONCE(1, "Removing stuck page flip\n");
+			page_flip_completed(intel_crtc);
+		}
+		spin_unlock_irq(&dev->event_lock);
+	}
 
-	mutex_lock(&dev->struct_mutex);
-	intel_finish_fb(crtc->primary->fb);
-	mutex_unlock(&dev->struct_mutex);
+	if (crtc->primary->fb) {
+		mutex_lock(&dev->struct_mutex);
+		intel_finish_fb(crtc->primary->fb);
+		mutex_unlock(&dev->struct_mutex);
+	}
 }
 
 /* Program iCLKIP clock to the desired frequency */
@@ -3580,9 +3849,7 @@ static void ironlake_pch_enable(struct drm_crtc *crtc)
 	intel_fdi_normal_train(crtc);
 
 	/* For PCH DP, enable TRANS_DP_CTL */
-	if (HAS_PCH_CPT(dev) &&
-	    (intel_pipe_has_type(crtc, INTEL_OUTPUT_DISPLAYPORT) ||
-	     intel_pipe_has_type(crtc, INTEL_OUTPUT_EDP))) {
+	if (HAS_PCH_CPT(dev) && intel_crtc->config.has_dp_encoder) {
 		u32 bpc = (I915_READ(PIPECONF(pipe)) & PIPECONF_BPC_MASK) >> 5;
 		reg = TRANS_DP_CTL(pipe);
 		temp = I915_READ(reg);
@@ -3642,12 +3909,13 @@ void intel_put_shared_dpll(struct intel_crtc *crtc)
 	if (pll == NULL)
 		return;
 
-	if (pll->refcount == 0) {
-		WARN(1, "bad %s refcount\n", pll->name);
+	if (!(pll->config.crtc_mask & (1 << crtc->pipe))) {
+		WARN(1, "bad %s crtc mask\n", pll->name);
 		return;
 	}
 
-	if (--pll->refcount == 0) {
+	pll->config.crtc_mask &= ~(1 << crtc->pipe);
+	if (pll->config.crtc_mask == 0) {
 		WARN_ON(pll->on);
 		WARN_ON(pll->active);
 	}
@@ -3658,15 +3926,9 @@ void intel_put_shared_dpll(struct intel_crtc *crtc)
 struct intel_shared_dpll *intel_get_shared_dpll(struct intel_crtc *crtc)
 {
 	struct drm_i915_private *dev_priv = crtc->base.dev->dev_private;
-	struct intel_shared_dpll *pll = intel_crtc_to_shared_dpll(crtc);
+	struct intel_shared_dpll *pll;
 	enum intel_dpll_id i;
 
-	if (pll) {
-		DRM_DEBUG_KMS("CRTC:%d dropping existing %s\n",
-			      crtc->base.base.id, pll->name);
-		intel_put_shared_dpll(crtc);
-	}
-
 	if (HAS_PCH_IBX(dev_priv->dev)) {
 		/* Ironlake PCH has a fixed PLL->PCH pipe mapping. */
 		i = (enum intel_dpll_id) crtc->pipe;
@@ -3675,7 +3937,7 @@ struct intel_shared_dpll *intel_get_shared_dpll(struct intel_crtc *crtc)
 		DRM_DEBUG_KMS("CRTC:%d using pre-allocated %s\n",
 			      crtc->base.base.id, pll->name);
 
-		WARN_ON(pll->refcount);
+		WARN_ON(pll->new_config->crtc_mask);
 
 		goto found;
 	}
@@ -3684,15 +3946,16 @@ struct intel_shared_dpll *intel_get_shared_dpll(struct intel_crtc *crtc)
 		pll = &dev_priv->shared_dplls[i];
 
 		/* Only want to check enabled timings first */
-		if (pll->refcount == 0)
+		if (pll->new_config->crtc_mask == 0)
 			continue;
 
-		if (memcmp(&crtc->config.dpll_hw_state, &pll->hw_state,
-			   sizeof(pll->hw_state)) == 0) {
-			DRM_DEBUG_KMS("CRTC:%d sharing existing %s (refcount %d, ative %d)\n",
-				      crtc->base.base.id,
-				      pll->name, pll->refcount, pll->active);
-
+		if (memcmp(&crtc->new_config->dpll_hw_state,
+			   &pll->new_config->hw_state,
+			   sizeof(pll->new_config->hw_state)) == 0) {
+			DRM_DEBUG_KMS("CRTC:%d sharing existing %s (crtc mask 0x%08x, ative %d)\n",
+				      crtc->base.base.id, pll->name,
+				      pll->new_config->crtc_mask,
+				      pll->active);
 			goto found;
 		}
 	}
@@ -3700,7 +3963,7 @@ struct intel_shared_dpll *intel_get_shared_dpll(struct intel_crtc *crtc)
 	/* Ok no matching timings, maybe there's a free one? */
 	for (i = 0; i < dev_priv->num_shared_dpll; i++) {
 		pll = &dev_priv->shared_dplls[i];
-		if (pll->refcount == 0) {
+		if (pll->new_config->crtc_mask == 0) {
 			DRM_DEBUG_KMS("CRTC:%d allocated %s\n",
 				      crtc->base.base.id, pll->name);
 			goto found;
@@ -3710,18 +3973,86 @@ struct intel_shared_dpll *intel_get_shared_dpll(struct intel_crtc *crtc)
 	return NULL;
 
 found:
-	if (pll->refcount == 0)
-		pll->hw_state = crtc->config.dpll_hw_state;
+	if (pll->new_config->crtc_mask == 0)
+		pll->new_config->hw_state = crtc->new_config->dpll_hw_state;
 
-	crtc->config.shared_dpll = i;
+	crtc->new_config->shared_dpll = i;
 	DRM_DEBUG_DRIVER("using %s for pipe %c\n", pll->name,
 			 pipe_name(crtc->pipe));
 
-	pll->refcount++;
+	pll->new_config->crtc_mask |= 1 << crtc->pipe;
 
 	return pll;
 }
 
+/**
+ * intel_shared_dpll_start_config - start a new PLL staged config
+ * @dev_priv: DRM device
+ * @clear_pipes: mask of pipes that will have their PLLs freed
+ *
+ * Starts a new PLL staged config, copying the current config but
+ * releasing the references of pipes specified in clear_pipes.
+ */
+static int intel_shared_dpll_start_config(struct drm_i915_private *dev_priv,
+					  unsigned clear_pipes)
+{
+	struct intel_shared_dpll *pll;
+	enum intel_dpll_id i;
+
+	for (i = 0; i < dev_priv->num_shared_dpll; i++) {
+		pll = &dev_priv->shared_dplls[i];
+
+		pll->new_config = kmemdup(&pll->config, sizeof pll->config,
+					  GFP_KERNEL);
+		if (!pll->new_config)
+			goto cleanup;
+
+		pll->new_config->crtc_mask &= ~clear_pipes;
+	}
+
+	return 0;
+
+cleanup:
+	while (--i >= 0) {
+		pll = &dev_priv->shared_dplls[i];
+		kfree(pll->new_config);
+		pll->new_config = NULL;
+	}
+
+	return -ENOMEM;
+}
+
+static void intel_shared_dpll_commit(struct drm_i915_private *dev_priv)
+{
+	struct intel_shared_dpll *pll;
+	enum intel_dpll_id i;
+
+	for (i = 0; i < dev_priv->num_shared_dpll; i++) {
+		pll = &dev_priv->shared_dplls[i];
+
+		WARN_ON(pll->new_config == &pll->config);
+
+		pll->config = *pll->new_config;
+		kfree(pll->new_config);
+		pll->new_config = NULL;
+	}
+}
+
+static void intel_shared_dpll_abort_config(struct drm_i915_private *dev_priv)
+{
+	struct intel_shared_dpll *pll;
+	enum intel_dpll_id i;
+
+	for (i = 0; i < dev_priv->num_shared_dpll; i++) {
+		pll = &dev_priv->shared_dplls[i];
+
+		WARN_ON(pll->new_config == &pll->config);
+
+		kfree(pll->new_config);
+		pll->new_config = NULL;
+	}
+}
+
 static void cpt_verify_modeset(struct drm_device *dev, int pipe)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -3736,6 +4067,19 @@ static void cpt_verify_modeset(struct drm_device *dev, int pipe)
 	}
 }
 
+static void skylake_pfit_enable(struct intel_crtc *crtc)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int pipe = crtc->pipe;
+
+	if (crtc->config.pch_pfit.enabled) {
+		I915_WRITE(PS_CTL(pipe), PS_ENABLE);
+		I915_WRITE(PS_WIN_POS(pipe), crtc->config.pch_pfit.pos);
+		I915_WRITE(PS_WIN_SZ(pipe), crtc->config.pch_pfit.size);
+	}
+}
+
 static void ironlake_pfit_enable(struct intel_crtc *crtc)
 {
 	struct drm_device *dev = crtc->base.dev;
@@ -3859,7 +4203,7 @@ static void intel_crtc_load_lut(struct drm_crtc *crtc)
 		return;
 
 	if (!HAS_PCH_SPLIT(dev_priv->dev)) {
-		if (intel_pipe_has_type(crtc, INTEL_OUTPUT_DSI))
+		if (intel_pipe_has_type(intel_crtc, INTEL_OUTPUT_DSI))
 			assert_dsi_pll_enabled(dev_priv);
 		else
 			assert_pll_enabled(dev_priv, pipe);
@@ -3911,14 +4255,10 @@ static void intel_crtc_dpms_overlay(struct intel_crtc *intel_crtc, bool enable)
 static void intel_crtc_enable_planes(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int pipe = intel_crtc->pipe;
-	int plane = intel_crtc->plane;
-
-	drm_vblank_on(dev, pipe);
 
-	intel_enable_primary_hw_plane(dev_priv, plane, pipe);
+	intel_enable_primary_hw_plane(crtc->primary, crtc);
 	intel_enable_planes(crtc);
 	intel_crtc_update_cursor(crtc, true);
 	intel_crtc_dpms_overlay(intel_crtc, true);
@@ -3955,7 +4295,7 @@ static void intel_crtc_disable_planes(struct drm_crtc *crtc)
 	intel_crtc_dpms_overlay(intel_crtc, false);
 	intel_crtc_update_cursor(crtc, false);
 	intel_disable_planes(crtc);
-	intel_disable_primary_hw_plane(dev_priv, plane, pipe);
+	intel_disable_primary_hw_plane(crtc->primary, crtc);
 
 	/*
 	 * FIXME: Once we grow proper nuclear flip support out of this we need
@@ -3963,8 +4303,6 @@ static void intel_crtc_disable_planes(struct drm_crtc *crtc)
 	 * consider this a flip to a NULL plane.
 	 */
 	intel_frontbuffer_flip(dev, INTEL_FRONTBUFFER_ALL_MASK(pipe));
-
-	drm_vblank_off(dev, pipe);
 }
 
 static void ironlake_crtc_enable(struct drm_crtc *crtc)
@@ -3974,7 +4312,6 @@ static void ironlake_crtc_enable(struct drm_crtc *crtc)
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_encoder *encoder;
 	int pipe = intel_crtc->pipe;
-	enum plane plane = intel_crtc->plane;
 
 	WARN_ON(!crtc->enabled);
 
@@ -3991,22 +4328,15 @@ static void ironlake_crtc_enable(struct drm_crtc *crtc)
 
 	if (intel_crtc->config.has_pch_encoder) {
 		intel_cpu_transcoder_set_m_n(intel_crtc,
-					     &intel_crtc->config.fdi_m_n);
+				     &intel_crtc->config.fdi_m_n, NULL);
 	}
 
 	ironlake_set_pipeconf(crtc);
 
-	/* Set up the display plane register */
-	I915_WRITE(DSPCNTR(plane), DISPPLANE_GAMMA_ENABLE);
-	POSTING_READ(DSPCNTR(plane));
-
-	dev_priv->display.update_primary_plane(crtc, crtc->primary->fb,
-					       crtc->x, crtc->y);
-
 	intel_crtc->active = true;
 
-	intel_set_cpu_fifo_underrun_reporting(dev, pipe, true);
-	intel_set_pch_fifo_underrun_reporting(dev, pipe, true);
+	intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true);
+	intel_set_pch_fifo_underrun_reporting(dev_priv, pipe, true);
 
 	for_each_encoder_on_crtc(dev, crtc, encoder)
 		if (encoder->pre_enable)
@@ -4042,6 +4372,9 @@ static void ironlake_crtc_enable(struct drm_crtc *crtc)
 	if (HAS_PCH_CPT(dev))
 		cpt_verify_modeset(dev, intel_crtc->pipe);
 
+	assert_vblank_disabled(crtc);
+	drm_crtc_vblank_on(crtc);
+
 	intel_crtc_enable_planes(crtc);
 }
 
@@ -4087,7 +4420,6 @@ static void haswell_crtc_enable(struct drm_crtc *crtc)
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_encoder *encoder;
 	int pipe = intel_crtc->pipe;
-	enum plane plane = intel_crtc->plane;
 
 	WARN_ON(!crtc->enabled);
 
@@ -4102,37 +4434,39 @@ static void haswell_crtc_enable(struct drm_crtc *crtc)
 
 	intel_set_pipe_timings(intel_crtc);
 
+	if (intel_crtc->config.cpu_transcoder != TRANSCODER_EDP) {
+		I915_WRITE(PIPE_MULT(intel_crtc->config.cpu_transcoder),
+			   intel_crtc->config.pixel_multiplier - 1);
+	}
+
 	if (intel_crtc->config.has_pch_encoder) {
 		intel_cpu_transcoder_set_m_n(intel_crtc,
-					     &intel_crtc->config.fdi_m_n);
+				     &intel_crtc->config.fdi_m_n, NULL);
 	}
 
 	haswell_set_pipeconf(crtc);
 
 	intel_set_pipe_csc(crtc);
 
-	/* Set up the display plane register */
-	I915_WRITE(DSPCNTR(plane), DISPPLANE_GAMMA_ENABLE | DISPPLANE_PIPE_CSC_ENABLE);
-	POSTING_READ(DSPCNTR(plane));
-
-	dev_priv->display.update_primary_plane(crtc, crtc->primary->fb,
-					       crtc->x, crtc->y);
-
 	intel_crtc->active = true;
 
-	intel_set_cpu_fifo_underrun_reporting(dev, pipe, true);
+	intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true);
 	for_each_encoder_on_crtc(dev, crtc, encoder)
 		if (encoder->pre_enable)
 			encoder->pre_enable(encoder);
 
 	if (intel_crtc->config.has_pch_encoder) {
-		intel_set_pch_fifo_underrun_reporting(dev, TRANSCODER_A, true);
+		intel_set_pch_fifo_underrun_reporting(dev_priv, TRANSCODER_A,
+						      true);
 		dev_priv->display.fdi_link_train(crtc);
 	}
 
 	intel_ddi_enable_pipe_clock(intel_crtc);
 
-	ironlake_pfit_enable(intel_crtc);
+	if (IS_SKYLAKE(dev))
+		skylake_pfit_enable(intel_crtc);
+	else
+		ironlake_pfit_enable(intel_crtc);
 
 	/*
 	 * On ILK+ LUT must be loaded before the pipe is running but with
@@ -4157,12 +4491,30 @@ static void haswell_crtc_enable(struct drm_crtc *crtc)
 		intel_opregion_notify_encoder(encoder, true);
 	}
 
+	assert_vblank_disabled(crtc);
+	drm_crtc_vblank_on(crtc);
+
 	/* If we change the relative order between pipe/planes enabling, we need
 	 * to change the workaround. */
 	haswell_mode_set_planes_workaround(intel_crtc);
 	intel_crtc_enable_planes(crtc);
 }
 
+static void skylake_pfit_disable(struct intel_crtc *crtc)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int pipe = crtc->pipe;
+
+	/* To avoid upsetting the power well on haswell only disable the pfit if
+	 * it's in use. The hw state code will make sure we get this right. */
+	if (crtc->config.pch_pfit.enabled) {
+		I915_WRITE(PS_CTL(pipe), 0);
+		I915_WRITE(PS_WIN_POS(pipe), 0);
+		I915_WRITE(PS_WIN_SZ(pipe), 0);
+	}
+}
+
 static void ironlake_pfit_disable(struct intel_crtc *crtc)
 {
 	struct drm_device *dev = crtc->base.dev;
@@ -4192,13 +4544,17 @@ static void ironlake_crtc_disable(struct drm_crtc *crtc)
 
 	intel_crtc_disable_planes(crtc);
 
+	drm_crtc_vblank_off(crtc);
+	assert_vblank_disabled(crtc);
+
 	for_each_encoder_on_crtc(dev, crtc, encoder)
 		encoder->disable(encoder);
 
 	if (intel_crtc->config.has_pch_encoder)
-		intel_set_pch_fifo_underrun_reporting(dev, pipe, false);
+		intel_set_pch_fifo_underrun_reporting(dev_priv, pipe, false);
+
+	intel_disable_pipe(intel_crtc);
 
-	intel_disable_pipe(dev_priv, pipe);
 	ironlake_pfit_disable(intel_crtc);
 
 	for_each_encoder_on_crtc(dev, crtc, encoder)
@@ -4209,7 +4565,6 @@ static void ironlake_crtc_disable(struct drm_crtc *crtc)
 		ironlake_fdi_disable(crtc);
 
 		ironlake_disable_pch_transcoder(dev_priv, pipe);
-		intel_set_pch_fifo_underrun_reporting(dev, pipe, true);
 
 		if (HAS_PCH_CPT(dev)) {
 			/* disable TRANS_DP_CTL */
@@ -4246,7 +4601,6 @@ static void haswell_crtc_disable(struct drm_crtc *crtc)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_encoder *encoder;
-	int pipe = intel_crtc->pipe;
 	enum transcoder cpu_transcoder = intel_crtc->config.cpu_transcoder;
 
 	if (!intel_crtc->active)
@@ -4254,27 +4608,33 @@ static void haswell_crtc_disable(struct drm_crtc *crtc)
 
 	intel_crtc_disable_planes(crtc);
 
+	drm_crtc_vblank_off(crtc);
+	assert_vblank_disabled(crtc);
+
 	for_each_encoder_on_crtc(dev, crtc, encoder) {
 		intel_opregion_notify_encoder(encoder, false);
 		encoder->disable(encoder);
 	}
 
 	if (intel_crtc->config.has_pch_encoder)
-		intel_set_pch_fifo_underrun_reporting(dev, TRANSCODER_A, false);
-	intel_disable_pipe(dev_priv, pipe);
+		intel_set_pch_fifo_underrun_reporting(dev_priv, TRANSCODER_A,
+						      false);
+	intel_disable_pipe(intel_crtc);
 
 	if (intel_crtc->config.dp_encoder_is_mst)
 		intel_ddi_set_vc_payload_alloc(crtc, false);
 
 	intel_ddi_disable_transcoder_func(dev_priv, cpu_transcoder);
 
-	ironlake_pfit_disable(intel_crtc);
+	if (IS_SKYLAKE(dev))
+		skylake_pfit_disable(intel_crtc);
+	else
+		ironlake_pfit_disable(intel_crtc);
 
 	intel_ddi_disable_pipe_clock(intel_crtc);
 
 	if (intel_crtc->config.has_pch_encoder) {
 		lpt_disable_pch_transcoder(dev_priv);
-		intel_set_pch_fifo_underrun_reporting(dev, TRANSCODER_A, true);
 		intel_ddi_fdi_disable(crtc);
 	}
 
@@ -4395,20 +4755,6 @@ static unsigned long get_crtc_power_domains(struct drm_crtc *crtc)
 	return mask;
 }
 
-void intel_display_set_init_power(struct drm_i915_private *dev_priv,
-				  bool enable)
-{
-	if (dev_priv->power_domains.init_power_on == enable)
-		return;
-
-	if (enable)
-		intel_display_power_get(dev_priv, POWER_DOMAIN_INIT);
-	else
-		intel_display_power_put(dev_priv, POWER_DOMAIN_INIT);
-
-	dev_priv->power_domains.init_power_on = enable;
-}
-
 static void modeset_update_crtc_power_domains(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -4431,6 +4777,9 @@ static void modeset_update_crtc_power_domains(struct drm_device *dev)
 			intel_display_power_get(dev_priv, domain);
 	}
 
+	if (dev_priv->display.modeset_global_resources)
+		dev_priv->display.modeset_global_resources(dev);
+
 	for_each_intel_crtc(dev, crtc) {
 		enum intel_display_power_domain domain;
 
@@ -4462,7 +4811,7 @@ static void vlv_update_cdclk(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	dev_priv->vlv_cdclk_freq = dev_priv->display.get_display_clock_speed(dev);
-	DRM_DEBUG_DRIVER("Current CD clock rate: %d kHz",
+	DRM_DEBUG_DRIVER("Current CD clock rate: %d kHz\n",
 			 dev_priv->vlv_cdclk_freq);
 
 	/*
@@ -4470,7 +4819,7 @@ static void vlv_update_cdclk(struct drm_device *dev)
 	 * BSpec erroneously claims we should aim for 4MHz, but
 	 * in fact 1MHz is the correct frequency.
 	 */
-	I915_WRITE(GMBUSFREQ_VLV, dev_priv->vlv_cdclk_freq);
+	I915_WRITE(GMBUSFREQ_VLV, DIV_ROUND_UP(dev_priv->vlv_cdclk_freq, 1000));
 }
 
 /* Adjust CDclk dividers to allow high res or save power if possible */
@@ -4501,10 +4850,9 @@ static void valleyview_set_cdclk(struct drm_device *dev, int cdclk)
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	if (cdclk == 400000) {
-		u32 divider, vco;
+		u32 divider;
 
-		vco = valleyview_get_vco(dev_priv);
-		divider = DIV_ROUND_CLOSEST(vco << 1, cdclk) - 1;
+		divider = DIV_ROUND_CLOSEST(dev_priv->hpll_freq << 1, cdclk) - 1;
 
 		mutex_lock(&dev_priv->dpio_lock);
 		/* adjust cdclk divider */
@@ -4539,11 +4887,55 @@ static void valleyview_set_cdclk(struct drm_device *dev, int cdclk)
 	vlv_update_cdclk(dev);
 }
 
+static void cherryview_set_cdclk(struct drm_device *dev, int cdclk)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 val, cmd;
+
+	WARN_ON(dev_priv->display.get_display_clock_speed(dev) != dev_priv->vlv_cdclk_freq);
+
+	switch (cdclk) {
+	case 400000:
+		cmd = 3;
+		break;
+	case 333333:
+	case 320000:
+		cmd = 2;
+		break;
+	case 266667:
+		cmd = 1;
+		break;
+	case 200000:
+		cmd = 0;
+		break;
+	default:
+		WARN_ON(1);
+		return;
+	}
+
+	mutex_lock(&dev_priv->rps.hw_lock);
+	val = vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ);
+	val &= ~DSPFREQGUAR_MASK_CHV;
+	val |= (cmd << DSPFREQGUAR_SHIFT_CHV);
+	vlv_punit_write(dev_priv, PUNIT_REG_DSPFREQ, val);
+	if (wait_for((vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ) &
+		      DSPFREQSTAT_MASK_CHV) == (cmd << DSPFREQSTAT_SHIFT_CHV),
+		     50)) {
+		DRM_ERROR("timed out waiting for CDclk change\n");
+	}
+	mutex_unlock(&dev_priv->rps.hw_lock);
+
+	vlv_update_cdclk(dev);
+}
+
 static int valleyview_calc_cdclk(struct drm_i915_private *dev_priv,
 				 int max_pixclk)
 {
-	int vco = valleyview_get_vco(dev_priv);
-	int freq_320 = (vco <<  1) % 320000 != 0 ? 333333 : 320000;
+	int freq_320 = (dev_priv->hpll_freq <<  1) % 320000 != 0 ? 333333 : 320000;
+
+	/* FIXME: Punit isn't quite ready yet */
+	if (IS_CHERRYVIEW(dev_priv->dev))
+		return 400000;
 
 	/*
 	 * Really only a few cases to deal with, as only 4 CDclks are supported:
@@ -4607,59 +4999,67 @@ static void valleyview_modeset_global_resources(struct drm_device *dev)
 	int max_pixclk = intel_mode_max_pixclk(dev_priv);
 	int req_cdclk = valleyview_calc_cdclk(dev_priv, max_pixclk);
 
-	if (req_cdclk != dev_priv->vlv_cdclk_freq)
-		valleyview_set_cdclk(dev, req_cdclk);
-	modeset_update_crtc_power_domains(dev);
+	if (req_cdclk != dev_priv->vlv_cdclk_freq) {
+		/*
+		 * FIXME: We can end up here with all power domains off, yet
+		 * with a CDCLK frequency other than the minimum. To account
+		 * for this take the PIPE-A power domain, which covers the HW
+		 * blocks needed for the following programming. This can be
+		 * removed once it's guaranteed that we get here either with
+		 * the minimum CDCLK set, or the required power domains
+		 * enabled.
+		 */
+		intel_display_power_get(dev_priv, POWER_DOMAIN_PIPE_A);
+
+		if (IS_CHERRYVIEW(dev))
+			cherryview_set_cdclk(dev, req_cdclk);
+		else
+			valleyview_set_cdclk(dev, req_cdclk);
+
+		intel_display_power_put(dev_priv, POWER_DOMAIN_PIPE_A);
+	}
 }
 
 static void valleyview_crtc_enable(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_encoder *encoder;
 	int pipe = intel_crtc->pipe;
-	int plane = intel_crtc->plane;
 	bool is_dsi;
-	u32 dspcntr;
 
 	WARN_ON(!crtc->enabled);
 
 	if (intel_crtc->active)
 		return;
 
-	is_dsi = intel_pipe_has_type(crtc, INTEL_OUTPUT_DSI);
-
-	if (!is_dsi && !IS_CHERRYVIEW(dev))
-		vlv_prepare_pll(intel_crtc);
+	is_dsi = intel_pipe_has_type(intel_crtc, INTEL_OUTPUT_DSI);
 
-	/* Set up the display plane register */
-	dspcntr = DISPPLANE_GAMMA_ENABLE;
+	if (!is_dsi) {
+		if (IS_CHERRYVIEW(dev))
+			chv_prepare_pll(intel_crtc, &intel_crtc->config);
+		else
+			vlv_prepare_pll(intel_crtc, &intel_crtc->config);
+	}
 
 	if (intel_crtc->config.has_dp_encoder)
 		intel_dp_set_m_n(intel_crtc);
 
 	intel_set_pipe_timings(intel_crtc);
 
-	/* pipesrc and dspsize control the size that is scaled from,
-	 * which should always be the user's requested size.
-	 */
-	I915_WRITE(DSPSIZE(plane),
-		   ((intel_crtc->config.pipe_src_h - 1) << 16) |
-		   (intel_crtc->config.pipe_src_w - 1));
-	I915_WRITE(DSPPOS(plane), 0);
-
-	i9xx_set_pipeconf(intel_crtc);
+	if (IS_CHERRYVIEW(dev) && pipe == PIPE_B) {
+		struct drm_i915_private *dev_priv = dev->dev_private;
 
-	I915_WRITE(DSPCNTR(plane), dspcntr);
-	POSTING_READ(DSPCNTR(plane));
+		I915_WRITE(CHV_BLEND(pipe), CHV_BLEND_LEGACY);
+		I915_WRITE(CHV_CANVAS(pipe), 0);
+	}
 
-	dev_priv->display.update_primary_plane(crtc, crtc->primary->fb,
-					       crtc->x, crtc->y);
+	i9xx_set_pipeconf(intel_crtc);
 
 	intel_crtc->active = true;
 
-	intel_set_cpu_fifo_underrun_reporting(dev, pipe, true);
+	intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true);
 
 	for_each_encoder_on_crtc(dev, crtc, encoder)
 		if (encoder->pre_pll_enable)
@@ -4667,9 +5067,9 @@ static void valleyview_crtc_enable(struct drm_crtc *crtc)
 
 	if (!is_dsi) {
 		if (IS_CHERRYVIEW(dev))
-			chv_enable_pll(intel_crtc);
+			chv_enable_pll(intel_crtc, &intel_crtc->config);
 		else
-			vlv_enable_pll(intel_crtc);
+			vlv_enable_pll(intel_crtc, &intel_crtc->config);
 	}
 
 	for_each_encoder_on_crtc(dev, crtc, encoder)
@@ -4686,10 +5086,13 @@ static void valleyview_crtc_enable(struct drm_crtc *crtc)
 	for_each_encoder_on_crtc(dev, crtc, encoder)
 		encoder->enable(encoder);
 
+	assert_vblank_disabled(crtc);
+	drm_crtc_vblank_on(crtc);
+
 	intel_crtc_enable_planes(crtc);
 
 	/* Underruns don't raise interrupts, so check manually. */
-	i9xx_check_fifo_underruns(dev);
+	i9xx_check_fifo_underruns(dev_priv);
 }
 
 static void i9xx_set_pll_dividers(struct intel_crtc *crtc)
@@ -4704,12 +5107,10 @@ static void i9xx_set_pll_dividers(struct intel_crtc *crtc)
 static void i9xx_crtc_enable(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_encoder *encoder;
 	int pipe = intel_crtc->pipe;
-	int plane = intel_crtc->plane;
-	u32 dspcntr;
 
 	WARN_ON(!crtc->enabled);
 
@@ -4718,39 +5119,17 @@ static void i9xx_crtc_enable(struct drm_crtc *crtc)
 
 	i9xx_set_pll_dividers(intel_crtc);
 
-	/* Set up the display plane register */
-	dspcntr = DISPPLANE_GAMMA_ENABLE;
-
-	if (pipe == 0)
-		dspcntr &= ~DISPPLANE_SEL_PIPE_MASK;
-	else
-		dspcntr |= DISPPLANE_SEL_PIPE_B;
-
 	if (intel_crtc->config.has_dp_encoder)
 		intel_dp_set_m_n(intel_crtc);
 
 	intel_set_pipe_timings(intel_crtc);
 
-	/* pipesrc and dspsize control the size that is scaled from,
-	 * which should always be the user's requested size.
-	 */
-	I915_WRITE(DSPSIZE(plane),
-		   ((intel_crtc->config.pipe_src_h - 1) << 16) |
-		   (intel_crtc->config.pipe_src_w - 1));
-	I915_WRITE(DSPPOS(plane), 0);
-
 	i9xx_set_pipeconf(intel_crtc);
 
-	I915_WRITE(DSPCNTR(plane), dspcntr);
-	POSTING_READ(DSPCNTR(plane));
-
-	dev_priv->display.update_primary_plane(crtc, crtc->primary->fb,
-					       crtc->x, crtc->y);
-
 	intel_crtc->active = true;
 
 	if (!IS_GEN2(dev))
-		intel_set_cpu_fifo_underrun_reporting(dev, pipe, true);
+		intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true);
 
 	for_each_encoder_on_crtc(dev, crtc, encoder)
 		if (encoder->pre_enable)
@@ -4768,6 +5147,9 @@ static void i9xx_crtc_enable(struct drm_crtc *crtc)
 	for_each_encoder_on_crtc(dev, crtc, encoder)
 		encoder->enable(encoder);
 
+	assert_vblank_disabled(crtc);
+	drm_crtc_vblank_on(crtc);
+
 	intel_crtc_enable_planes(crtc);
 
 	/*
@@ -4778,10 +5160,10 @@ static void i9xx_crtc_enable(struct drm_crtc *crtc)
 	 * but leave the pipe running.
 	 */
 	if (IS_GEN2(dev))
-		intel_set_cpu_fifo_underrun_reporting(dev, pipe, true);
+		intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true);
 
 	/* Underruns don't raise interrupts, so check manually. */
-	i9xx_check_fifo_underruns(dev);
+	i9xx_check_fifo_underruns(dev_priv);
 }
 
 static void i9xx_pfit_disable(struct intel_crtc *crtc)
@@ -4817,7 +5199,7 @@ static void i9xx_crtc_disable(struct drm_crtc *crtc)
 	 * but leave the pipe running.
 	 */
 	if (IS_GEN2(dev))
-		intel_set_cpu_fifo_underrun_reporting(dev, pipe, false);
+		intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, false);
 
 	/*
 	 * Vblank time updates from the shadow to live plane control register
@@ -4831,9 +5213,6 @@ static void i9xx_crtc_disable(struct drm_crtc *crtc)
 	intel_set_memory_cxsr(dev_priv, false);
 	intel_crtc_disable_planes(crtc);
 
-	for_each_encoder_on_crtc(dev, crtc, encoder)
-		encoder->disable(encoder);
-
 	/*
 	 * On gen2 planes are double buffered but the pipe isn't, so we must
 	 * wait for planes to fully turn off before disabling the pipe.
@@ -4842,7 +5221,13 @@ static void i9xx_crtc_disable(struct drm_crtc *crtc)
 	 */
 	intel_wait_for_vblank(dev, pipe);
 
-	intel_disable_pipe(dev_priv, pipe);
+	drm_crtc_vblank_off(crtc);
+	assert_vblank_disabled(crtc);
+
+	for_each_encoder_on_crtc(dev, crtc, encoder)
+		encoder->disable(encoder);
+
+	intel_disable_pipe(intel_crtc);
 
 	i9xx_pfit_disable(intel_crtc);
 
@@ -4850,17 +5235,17 @@ static void i9xx_crtc_disable(struct drm_crtc *crtc)
 		if (encoder->post_disable)
 			encoder->post_disable(encoder);
 
-	if (!intel_pipe_has_type(crtc, INTEL_OUTPUT_DSI)) {
+	if (!intel_pipe_has_type(intel_crtc, INTEL_OUTPUT_DSI)) {
 		if (IS_CHERRYVIEW(dev))
 			chv_disable_pll(dev_priv, pipe);
 		else if (IS_VALLEYVIEW(dev))
 			vlv_disable_pll(dev_priv, pipe);
 		else
-			i9xx_disable_pll(dev_priv, pipe);
+			i9xx_disable_pll(intel_crtc);
 	}
 
 	if (!IS_GEN2(dev))
-		intel_set_cpu_fifo_underrun_reporting(dev, pipe, false);
+		intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, false);
 
 	intel_crtc->active = false;
 	intel_update_watermarks(crtc);
@@ -4874,36 +5259,6 @@ static void i9xx_crtc_off(struct drm_crtc *crtc)
 {
 }
 
-static void intel_crtc_update_sarea(struct drm_crtc *crtc,
-				    bool enabled)
-{
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_master_private *master_priv;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
-
-	if (!dev->primary->master)
-		return;
-
-	master_priv = dev->primary->master->driver_priv;
-	if (!master_priv->sarea_priv)
-		return;
-
-	switch (pipe) {
-	case 0:
-		master_priv->sarea_priv->pipeA_w = enabled ? crtc->mode.hdisplay : 0;
-		master_priv->sarea_priv->pipeA_h = enabled ? crtc->mode.vdisplay : 0;
-		break;
-	case 1:
-		master_priv->sarea_priv->pipeB_w = enabled ? crtc->mode.hdisplay : 0;
-		master_priv->sarea_priv->pipeB_h = enabled ? crtc->mode.vdisplay : 0;
-		break;
-	default:
-		DRM_ERROR("Can't update pipe %c in SAREA\n", pipe_name(pipe));
-		break;
-	}
-}
-
 /* Master function to enable/disable CRTC and corresponding power wells */
 void intel_crtc_control(struct drm_crtc *crtc, bool enable)
 {
@@ -4947,8 +5302,6 @@ void intel_crtc_update_dpms(struct drm_crtc *crtc)
 		enable |= intel_encoder->connectors_active;
 
 	intel_crtc_control(crtc, enable);
-
-	intel_crtc_update_sarea(crtc, enable);
 }
 
 static void intel_crtc_disable(struct drm_crtc *crtc)
@@ -4963,7 +5316,6 @@ static void intel_crtc_disable(struct drm_crtc *crtc)
 	WARN_ON(!crtc->enabled);
 
 	dev_priv->display.crtc_disable(crtc);
-	intel_crtc_update_sarea(crtc, false);
 	dev_priv->display.off(crtc);
 
 	if (crtc->primary->fb) {
@@ -5202,11 +5554,11 @@ static int intel_crtc_compute_config(struct intel_crtc *crtc,
 				     struct intel_crtc_config *pipe_config)
 {
 	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_display_mode *adjusted_mode = &pipe_config->adjusted_mode;
 
 	/* FIXME should check pixel clock limits on all platforms */
 	if (INTEL_INFO(dev)->gen < 4) {
-		struct drm_i915_private *dev_priv = dev->dev_private;
 		int clock_limit =
 			dev_priv->display.get_display_clock_speed(dev);
 
@@ -5233,7 +5585,7 @@ static int intel_crtc_compute_config(struct intel_crtc *crtc,
 	 * - LVDS dual channel mode
 	 * - Double wide pipe
 	 */
-	if ((intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_LVDS) &&
+	if ((intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS) &&
 	     intel_is_dual_link_lvds(dev)) || pipe_config->double_wide)
 		pipe_config->pipe_src_w &= ~1;
 
@@ -5255,13 +5607,6 @@ static int intel_crtc_compute_config(struct intel_crtc *crtc,
 	if (HAS_IPS(dev))
 		hsw_compute_ips_config(crtc, pipe_config);
 
-	/*
-	 * XXX: PCH/WRPLL clock sharing is done in ->mode_set, so make sure the
-	 * old clock survives for now.
-	 */
-	if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev) || HAS_DDI(dev))
-		pipe_config->shared_dpll = crtc->config.shared_dpll;
-
 	if (pipe_config->has_pch_encoder)
 		return ironlake_fdi_compute_config(crtc, pipe_config);
 
@@ -5271,10 +5616,16 @@ static int intel_crtc_compute_config(struct intel_crtc *crtc,
 static int valleyview_get_display_clock_speed(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	int vco = valleyview_get_vco(dev_priv);
 	u32 val;
 	int divider;
 
+	/* FIXME: Punit isn't quite ready yet */
+	if (IS_CHERRYVIEW(dev))
+		return 400000;
+
+	if (dev_priv->hpll_freq == 0)
+		dev_priv->hpll_freq = valleyview_get_vco(dev_priv);
+
 	mutex_lock(&dev_priv->dpio_lock);
 	val = vlv_cck_read(dev_priv, CCK_DISPLAY_CLOCK_CONTROL);
 	mutex_unlock(&dev_priv->dpio_lock);
@@ -5285,7 +5636,7 @@ static int valleyview_get_display_clock_speed(struct drm_device *dev)
 	     (divider << DISPLAY_FREQUENCY_STATUS_SHIFT),
 	     "cdclk change in progress\n");
 
-	return DIV_ROUND_CLOSEST(vco << 1, divider + 1);
+	return DIV_ROUND_CLOSEST(dev_priv->hpll_freq << 1, divider + 1);
 }
 
 static int i945_get_display_clock_speed(struct drm_device *dev)
@@ -5417,15 +5768,15 @@ static inline bool intel_panel_use_ssc(struct drm_i915_private *dev_priv)
 		&& !(dev_priv->quirks & QUIRK_LVDS_SSC_DISABLE);
 }
 
-static int i9xx_get_refclk(struct drm_crtc *crtc, int num_connectors)
+static int i9xx_get_refclk(struct intel_crtc *crtc, int num_connectors)
 {
-	struct drm_device *dev = crtc->dev;
+	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int refclk;
 
 	if (IS_VALLEYVIEW(dev)) {
 		refclk = 100000;
-	} else if (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS) &&
+	} else if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS) &&
 	    intel_panel_use_ssc(dev_priv) && num_connectors < 2) {
 		refclk = dev_priv->vbt.lvds_ssc_freq;
 		DRM_DEBUG_KMS("using SSC reference clock of %d kHz\n", refclk);
@@ -5455,24 +5806,24 @@ static void i9xx_update_pll_dividers(struct intel_crtc *crtc,
 	u32 fp, fp2 = 0;
 
 	if (IS_PINEVIEW(dev)) {
-		fp = pnv_dpll_compute_fp(&crtc->config.dpll);
+		fp = pnv_dpll_compute_fp(&crtc->new_config->dpll);
 		if (reduced_clock)
 			fp2 = pnv_dpll_compute_fp(reduced_clock);
 	} else {
-		fp = i9xx_dpll_compute_fp(&crtc->config.dpll);
+		fp = i9xx_dpll_compute_fp(&crtc->new_config->dpll);
 		if (reduced_clock)
 			fp2 = i9xx_dpll_compute_fp(reduced_clock);
 	}
 
-	crtc->config.dpll_hw_state.fp0 = fp;
+	crtc->new_config->dpll_hw_state.fp0 = fp;
 
 	crtc->lowfreq_avail = false;
-	if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_LVDS) &&
+	if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS) &&
 	    reduced_clock && i915.powersave) {
-		crtc->config.dpll_hw_state.fp1 = fp2;
+		crtc->new_config->dpll_hw_state.fp1 = fp2;
 		crtc->lowfreq_avail = true;
 	} else {
-		crtc->config.dpll_hw_state.fp1 = fp;
+		crtc->new_config->dpll_hw_state.fp1 = fp;
 	}
 }
 
@@ -5519,7 +5870,8 @@ static void intel_pch_transcoder_set_m_n(struct intel_crtc *crtc,
 }
 
 static void intel_cpu_transcoder_set_m_n(struct intel_crtc *crtc,
-					 struct intel_link_m_n *m_n)
+					 struct intel_link_m_n *m_n,
+					 struct intel_link_m_n *m2_n2)
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -5531,6 +5883,18 @@ static void intel_cpu_transcoder_set_m_n(struct intel_crtc *crtc,
 		I915_WRITE(PIPE_DATA_N1(transcoder), m_n->gmch_n);
 		I915_WRITE(PIPE_LINK_M1(transcoder), m_n->link_m);
 		I915_WRITE(PIPE_LINK_N1(transcoder), m_n->link_n);
+		/* M2_N2 registers to be set only for gen < 8 (M2_N2 available
+		 * for gen < 8) and if DRRS is supported (to make sure the
+		 * registers are not unnecessarily accessed).
+		 */
+		if (m2_n2 && INTEL_INFO(dev)->gen < 8 &&
+			crtc->config.has_drrs) {
+			I915_WRITE(PIPE_DATA_M2(transcoder),
+					TU_SIZE(m2_n2->tu) | m2_n2->gmch_m);
+			I915_WRITE(PIPE_DATA_N2(transcoder), m2_n2->gmch_n);
+			I915_WRITE(PIPE_LINK_M2(transcoder), m2_n2->link_m);
+			I915_WRITE(PIPE_LINK_N2(transcoder), m2_n2->link_n);
+		}
 	} else {
 		I915_WRITE(PIPE_DATA_M_G4X(pipe), TU_SIZE(m_n->tu) | m_n->gmch_m);
 		I915_WRITE(PIPE_DATA_N_G4X(pipe), m_n->gmch_n);
@@ -5539,15 +5903,17 @@ static void intel_cpu_transcoder_set_m_n(struct intel_crtc *crtc,
 	}
 }
 
-static void intel_dp_set_m_n(struct intel_crtc *crtc)
+void intel_dp_set_m_n(struct intel_crtc *crtc)
 {
 	if (crtc->config.has_pch_encoder)
 		intel_pch_transcoder_set_m_n(crtc, &crtc->config.dp_m_n);
 	else
-		intel_cpu_transcoder_set_m_n(crtc, &crtc->config.dp_m_n);
+		intel_cpu_transcoder_set_m_n(crtc, &crtc->config.dp_m_n,
+						   &crtc->config.dp_m2_n2);
 }
 
-static void vlv_update_pll(struct intel_crtc *crtc)
+static void vlv_update_pll(struct intel_crtc *crtc,
+			   struct intel_crtc_config *pipe_config)
 {
 	u32 dpll, dpll_md;
 
@@ -5562,14 +5928,15 @@ static void vlv_update_pll(struct intel_crtc *crtc)
 	if (crtc->pipe == PIPE_B)
 		dpll |= DPLL_INTEGRATED_CRI_CLK_VLV;
 	dpll |= DPLL_VCO_ENABLE;
-	crtc->config.dpll_hw_state.dpll = dpll;
+	pipe_config->dpll_hw_state.dpll = dpll;
 
-	dpll_md = (crtc->config.pixel_multiplier - 1)
+	dpll_md = (pipe_config->pixel_multiplier - 1)
 		<< DPLL_MD_UDI_MULTIPLIER_SHIFT;
-	crtc->config.dpll_hw_state.dpll_md = dpll_md;
+	pipe_config->dpll_hw_state.dpll_md = dpll_md;
 }
 
-static void vlv_prepare_pll(struct intel_crtc *crtc)
+static void vlv_prepare_pll(struct intel_crtc *crtc,
+			    const struct intel_crtc_config *pipe_config)
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -5580,11 +5947,11 @@ static void vlv_prepare_pll(struct intel_crtc *crtc)
 
 	mutex_lock(&dev_priv->dpio_lock);
 
-	bestn = crtc->config.dpll.n;
-	bestm1 = crtc->config.dpll.m1;
-	bestm2 = crtc->config.dpll.m2;
-	bestp1 = crtc->config.dpll.p1;
-	bestp2 = crtc->config.dpll.p2;
+	bestn = pipe_config->dpll.n;
+	bestm1 = pipe_config->dpll.m1;
+	bestm2 = pipe_config->dpll.m2;
+	bestp1 = pipe_config->dpll.p1;
+	bestp2 = pipe_config->dpll.p2;
 
 	/* See eDP HDMI DPIO driver vbios notes doc */
 
@@ -5621,17 +5988,16 @@ static void vlv_prepare_pll(struct intel_crtc *crtc)
 	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW3(pipe), mdiv);
 
 	/* Set HBR and RBR LPF coefficients */
-	if (crtc->config.port_clock == 162000 ||
-	    intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_ANALOG) ||
-	    intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_HDMI))
+	if (pipe_config->port_clock == 162000 ||
+	    intel_pipe_has_type(crtc, INTEL_OUTPUT_ANALOG) ||
+	    intel_pipe_has_type(crtc, INTEL_OUTPUT_HDMI))
 		vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW10(pipe),
 				 0x009f0003);
 	else
 		vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW10(pipe),
 				 0x00d0000f);
 
-	if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_EDP) ||
-	    intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_DISPLAYPORT)) {
+	if (crtc->config.has_dp_encoder) {
 		/* Use SSC source */
 		if (pipe == PIPE_A)
 			vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW5(pipe),
@@ -5651,8 +6017,8 @@ static void vlv_prepare_pll(struct intel_crtc *crtc)
 
 	coreclk = vlv_dpio_read(dev_priv, pipe, VLV_PLL_DW7(pipe));
 	coreclk = (coreclk & 0x0000ff00) | 0x01c00000;
-	if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_DISPLAYPORT) ||
-	    intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_EDP))
+	if (intel_pipe_has_type(crtc, INTEL_OUTPUT_DISPLAYPORT) ||
+	    intel_pipe_has_type(crtc, INTEL_OUTPUT_EDP))
 		coreclk |= 0x01000000;
 	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW7(pipe), coreclk);
 
@@ -5660,7 +6026,21 @@ static void vlv_prepare_pll(struct intel_crtc *crtc)
 	mutex_unlock(&dev_priv->dpio_lock);
 }
 
-static void chv_update_pll(struct intel_crtc *crtc)
+static void chv_update_pll(struct intel_crtc *crtc,
+			   struct intel_crtc_config *pipe_config)
+{
+	pipe_config->dpll_hw_state.dpll = DPLL_SSC_REF_CLOCK_CHV |
+		DPLL_REFA_CLK_ENABLE_VLV | DPLL_VGA_MODE_DIS |
+		DPLL_VCO_ENABLE;
+	if (crtc->pipe != PIPE_A)
+		pipe_config->dpll_hw_state.dpll |= DPLL_INTEGRATED_CRI_CLK_VLV;
+
+	pipe_config->dpll_hw_state.dpll_md =
+		(pipe_config->pixel_multiplier - 1) << DPLL_MD_UDI_MULTIPLIER_SHIFT;
+}
+
+static void chv_prepare_pll(struct intel_crtc *crtc,
+			    const struct intel_crtc_config *pipe_config)
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -5671,27 +6051,18 @@ static void chv_update_pll(struct intel_crtc *crtc)
 	u32 bestn, bestm1, bestm2, bestp1, bestp2, bestm2_frac;
 	int refclk;
 
-	crtc->config.dpll_hw_state.dpll = DPLL_SSC_REF_CLOCK_CHV |
-		DPLL_REFA_CLK_ENABLE_VLV | DPLL_VGA_MODE_DIS |
-		DPLL_VCO_ENABLE;
-	if (pipe != PIPE_A)
-		crtc->config.dpll_hw_state.dpll |= DPLL_INTEGRATED_CRI_CLK_VLV;
-
-	crtc->config.dpll_hw_state.dpll_md =
-		(crtc->config.pixel_multiplier - 1) << DPLL_MD_UDI_MULTIPLIER_SHIFT;
-
-	bestn = crtc->config.dpll.n;
-	bestm2_frac = crtc->config.dpll.m2 & 0x3fffff;
-	bestm1 = crtc->config.dpll.m1;
-	bestm2 = crtc->config.dpll.m2 >> 22;
-	bestp1 = crtc->config.dpll.p1;
-	bestp2 = crtc->config.dpll.p2;
+	bestn = pipe_config->dpll.n;
+	bestm2_frac = pipe_config->dpll.m2 & 0x3fffff;
+	bestm1 = pipe_config->dpll.m1;
+	bestm2 = pipe_config->dpll.m2 >> 22;
+	bestp1 = pipe_config->dpll.p1;
+	bestp2 = pipe_config->dpll.p2;
 
 	/*
 	 * Enable Refclk and SSC
 	 */
 	I915_WRITE(dpll_reg,
-		   crtc->config.dpll_hw_state.dpll & ~DPLL_VCO_ENABLE);
+		   pipe_config->dpll_hw_state.dpll & ~DPLL_VCO_ENABLE);
 
 	mutex_lock(&dev_priv->dpio_lock);
 
@@ -5719,7 +6090,7 @@ static void chv_update_pll(struct intel_crtc *crtc)
 		       (2 << DPIO_CHV_FEEDFWD_GAIN_SHIFT));
 
 	/* Loop filter */
-	refclk = i9xx_get_refclk(&crtc->base, 0);
+	refclk = i9xx_get_refclk(crtc, 0);
 	loopfilter = 5 << DPIO_CHV_PROP_COEFF_SHIFT |
 		2 << DPIO_CHV_GAIN_CTRL_SHIFT;
 	if (refclk == 100000)
@@ -5739,6 +6110,53 @@ static void chv_update_pll(struct intel_crtc *crtc)
 	mutex_unlock(&dev_priv->dpio_lock);
 }
 
+/**
+ * vlv_force_pll_on - forcibly enable just the PLL
+ * @dev_priv: i915 private structure
+ * @pipe: pipe PLL to enable
+ * @dpll: PLL configuration
+ *
+ * Enable the PLL for @pipe using the supplied @dpll config. To be used
+ * in cases where we need the PLL enabled even when @pipe is not going to
+ * be enabled.
+ */
+void vlv_force_pll_on(struct drm_device *dev, enum pipe pipe,
+		      const struct dpll *dpll)
+{
+	struct intel_crtc *crtc =
+		to_intel_crtc(intel_get_crtc_for_pipe(dev, pipe));
+	struct intel_crtc_config pipe_config = {
+		.pixel_multiplier = 1,
+		.dpll = *dpll,
+	};
+
+	if (IS_CHERRYVIEW(dev)) {
+		chv_update_pll(crtc, &pipe_config);
+		chv_prepare_pll(crtc, &pipe_config);
+		chv_enable_pll(crtc, &pipe_config);
+	} else {
+		vlv_update_pll(crtc, &pipe_config);
+		vlv_prepare_pll(crtc, &pipe_config);
+		vlv_enable_pll(crtc, &pipe_config);
+	}
+}
+
+/**
+ * vlv_force_pll_off - forcibly disable just the PLL
+ * @dev_priv: i915 private structure
+ * @pipe: pipe PLL to disable
+ *
+ * Disable the PLL for @pipe. To be used in cases where we need
+ * the PLL enabled even when @pipe is not going to be enabled.
+ */
+void vlv_force_pll_off(struct drm_device *dev, enum pipe pipe)
+{
+	if (IS_CHERRYVIEW(dev))
+		chv_disable_pll(to_i915(dev), pipe);
+	else
+		vlv_disable_pll(to_i915(dev), pipe);
+}
+
 static void i9xx_update_pll(struct intel_crtc *crtc,
 			    intel_clock_t *reduced_clock,
 			    int num_connectors)
@@ -5747,29 +6165,29 @@ static void i9xx_update_pll(struct intel_crtc *crtc,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 dpll;
 	bool is_sdvo;
-	struct dpll *clock = &crtc->config.dpll;
+	struct dpll *clock = &crtc->new_config->dpll;
 
 	i9xx_update_pll_dividers(crtc, reduced_clock);
 
-	is_sdvo = intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_SDVO) ||
-		intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_HDMI);
+	is_sdvo = intel_pipe_will_have_type(crtc, INTEL_OUTPUT_SDVO) ||
+		intel_pipe_will_have_type(crtc, INTEL_OUTPUT_HDMI);
 
 	dpll = DPLL_VGA_MODE_DIS;
 
-	if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_LVDS))
+	if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS))
 		dpll |= DPLLB_MODE_LVDS;
 	else
 		dpll |= DPLLB_MODE_DAC_SERIAL;
 
 	if (IS_I945G(dev) || IS_I945GM(dev) || IS_G33(dev)) {
-		dpll |= (crtc->config.pixel_multiplier - 1)
+		dpll |= (crtc->new_config->pixel_multiplier - 1)
 			<< SDVO_MULTIPLIER_SHIFT_HIRES;
 	}
 
 	if (is_sdvo)
 		dpll |= DPLL_SDVO_HIGH_SPEED;
 
-	if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_DISPLAYPORT))
+	if (crtc->new_config->has_dp_encoder)
 		dpll |= DPLL_SDVO_HIGH_SPEED;
 
 	/* compute bitmask from p1 value */
@@ -5797,21 +6215,21 @@ static void i9xx_update_pll(struct intel_crtc *crtc,
 	if (INTEL_INFO(dev)->gen >= 4)
 		dpll |= (6 << PLL_LOAD_PULSE_PHASE_SHIFT);
 
-	if (crtc->config.sdvo_tv_clock)
+	if (crtc->new_config->sdvo_tv_clock)
 		dpll |= PLL_REF_INPUT_TVCLKINBC;
-	else if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_LVDS) &&
+	else if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS) &&
 		 intel_panel_use_ssc(dev_priv) && num_connectors < 2)
 		dpll |= PLLB_REF_INPUT_SPREADSPECTRUMIN;
 	else
 		dpll |= PLL_REF_INPUT_DREFCLK;
 
 	dpll |= DPLL_VCO_ENABLE;
-	crtc->config.dpll_hw_state.dpll = dpll;
+	crtc->new_config->dpll_hw_state.dpll = dpll;
 
 	if (INTEL_INFO(dev)->gen >= 4) {
-		u32 dpll_md = (crtc->config.pixel_multiplier - 1)
+		u32 dpll_md = (crtc->new_config->pixel_multiplier - 1)
 			<< DPLL_MD_UDI_MULTIPLIER_SHIFT;
-		crtc->config.dpll_hw_state.dpll_md = dpll_md;
+		crtc->new_config->dpll_hw_state.dpll_md = dpll_md;
 	}
 }
 
@@ -5822,13 +6240,13 @@ static void i8xx_update_pll(struct intel_crtc *crtc,
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 dpll;
-	struct dpll *clock = &crtc->config.dpll;
+	struct dpll *clock = &crtc->new_config->dpll;
 
 	i9xx_update_pll_dividers(crtc, reduced_clock);
 
 	dpll = DPLL_VGA_MODE_DIS;
 
-	if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_LVDS)) {
+	if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS)) {
 		dpll |= (1 << (clock->p1 - 1)) << DPLL_FPA01_P1_POST_DIV_SHIFT;
 	} else {
 		if (clock->p1 == 2)
@@ -5839,17 +6257,17 @@ static void i8xx_update_pll(struct intel_crtc *crtc,
 			dpll |= PLL_P2_DIVIDE_BY_4;
 	}
 
-	if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_DVO))
+	if (!IS_I830(dev) && intel_pipe_will_have_type(crtc, INTEL_OUTPUT_DVO))
 		dpll |= DPLL_DVO_2X_MODE;
 
-	if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_LVDS) &&
+	if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS) &&
 		 intel_panel_use_ssc(dev_priv) && num_connectors < 2)
 		dpll |= PLLB_REF_INPUT_SPREADSPECTRUMIN;
 	else
 		dpll |= PLL_REF_INPUT_DREFCLK;
 
 	dpll |= DPLL_VCO_ENABLE;
-	crtc->config.dpll_hw_state.dpll = dpll;
+	crtc->new_config->dpll_hw_state.dpll = dpll;
 }
 
 static void intel_set_pipe_timings(struct intel_crtc *intel_crtc)
@@ -5873,7 +6291,7 @@ static void intel_set_pipe_timings(struct intel_crtc *intel_crtc)
 		crtc_vtotal -= 1;
 		crtc_vblank_end -= 1;
 
-		if (intel_pipe_has_type(&intel_crtc->base, INTEL_OUTPUT_SDVO))
+		if (intel_pipe_has_type(intel_crtc, INTEL_OUTPUT_SDVO))
 			vsyncshift = (adjusted_mode->crtc_htotal - 1) / 2;
 		else
 			vsyncshift = adjusted_mode->crtc_hsync_start -
@@ -5990,9 +6408,9 @@ static void i9xx_set_pipeconf(struct intel_crtc *intel_crtc)
 
 	pipeconf = 0;
 
-	if (dev_priv->quirks & QUIRK_PIPEA_FORCE &&
-	    I915_READ(PIPECONF(intel_crtc->pipe)) & PIPECONF_ENABLE)
-		pipeconf |= PIPECONF_ENABLE;
+	if ((intel_crtc->pipe == PIPE_A && dev_priv->quirks & QUIRK_PIPEA_FORCE) ||
+	    (intel_crtc->pipe == PIPE_B && dev_priv->quirks & QUIRK_PIPEB_FORCE))
+		pipeconf |= I915_READ(PIPECONF(intel_crtc->pipe)) & PIPECONF_ENABLE;
 
 	if (intel_crtc->config.double_wide)
 		pipeconf |= PIPECONF_DOUBLE_WIDE;
@@ -6031,7 +6449,7 @@ static void i9xx_set_pipeconf(struct intel_crtc *intel_crtc)
 
 	if (intel_crtc->config.adjusted_mode.flags & DRM_MODE_FLAG_INTERLACE) {
 		if (INTEL_INFO(dev)->gen < 4 ||
-		    intel_pipe_has_type(&intel_crtc->base, INTEL_OUTPUT_SDVO))
+		    intel_pipe_has_type(intel_crtc, INTEL_OUTPUT_SDVO))
 			pipeconf |= PIPECONF_INTERLACE_W_FIELD_INDICATION;
 		else
 			pipeconf |= PIPECONF_INTERLACE_W_SYNC_SHIFT;
@@ -6045,13 +6463,10 @@ static void i9xx_set_pipeconf(struct intel_crtc *intel_crtc)
 	POSTING_READ(PIPECONF(intel_crtc->pipe));
 }
 
-static int i9xx_crtc_mode_set(struct drm_crtc *crtc,
-			      int x, int y,
-			      struct drm_framebuffer *fb)
+static int i9xx_crtc_compute_clock(struct intel_crtc *crtc)
 {
-	struct drm_device *dev = crtc->dev;
+	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int refclk, num_connectors = 0;
 	intel_clock_t clock, reduced_clock;
 	bool ok, has_reduced_clock = false;
@@ -6059,7 +6474,10 @@ static int i9xx_crtc_mode_set(struct drm_crtc *crtc,
 	struct intel_encoder *encoder;
 	const intel_limit_t *limit;
 
-	for_each_encoder_on_crtc(dev, crtc, encoder) {
+	for_each_intel_encoder(dev, encoder) {
+		if (encoder->new_crtc != crtc)
+			continue;
+
 		switch (encoder->type) {
 		case INTEL_OUTPUT_LVDS:
 			is_lvds = true;
@@ -6067,6 +6485,8 @@ static int i9xx_crtc_mode_set(struct drm_crtc *crtc,
 		case INTEL_OUTPUT_DSI:
 			is_dsi = true;
 			break;
+		default:
+			break;
 		}
 
 		num_connectors++;
@@ -6075,7 +6495,7 @@ static int i9xx_crtc_mode_set(struct drm_crtc *crtc,
 	if (is_dsi)
 		return 0;
 
-	if (!intel_crtc->config.clock_set) {
+	if (!crtc->new_config->clock_set) {
 		refclk = i9xx_get_refclk(crtc, num_connectors);
 
 		/*
@@ -6086,7 +6506,7 @@ static int i9xx_crtc_mode_set(struct drm_crtc *crtc,
 		 */
 		limit = intel_limit(crtc, refclk);
 		ok = dev_priv->display.find_dpll(limit, crtc,
-						 intel_crtc->config.port_clock,
+						 crtc->new_config->port_clock,
 						 refclk, NULL, &clock);
 		if (!ok) {
 			DRM_ERROR("Couldn't find PLL settings for mode!\n");
@@ -6107,23 +6527,23 @@ static int i9xx_crtc_mode_set(struct drm_crtc *crtc,
 							    &reduced_clock);
 		}
 		/* Compat-code for transition, will disappear. */
-		intel_crtc->config.dpll.n = clock.n;
-		intel_crtc->config.dpll.m1 = clock.m1;
-		intel_crtc->config.dpll.m2 = clock.m2;
-		intel_crtc->config.dpll.p1 = clock.p1;
-		intel_crtc->config.dpll.p2 = clock.p2;
+		crtc->new_config->dpll.n = clock.n;
+		crtc->new_config->dpll.m1 = clock.m1;
+		crtc->new_config->dpll.m2 = clock.m2;
+		crtc->new_config->dpll.p1 = clock.p1;
+		crtc->new_config->dpll.p2 = clock.p2;
 	}
 
 	if (IS_GEN2(dev)) {
-		i8xx_update_pll(intel_crtc,
+		i8xx_update_pll(crtc,
 				has_reduced_clock ? &reduced_clock : NULL,
 				num_connectors);
 	} else if (IS_CHERRYVIEW(dev)) {
-		chv_update_pll(intel_crtc);
+		chv_update_pll(crtc, crtc->new_config);
 	} else if (IS_VALLEYVIEW(dev)) {
-		vlv_update_pll(intel_crtc);
+		vlv_update_pll(crtc, crtc->new_config);
 	} else {
-		i9xx_update_pll(intel_crtc,
+		i9xx_update_pll(crtc,
 				has_reduced_clock ? &reduced_clock : NULL,
 				num_connectors);
 	}
@@ -6235,7 +6655,7 @@ static void i9xx_get_plane_config(struct intel_crtc *crtc,
 	crtc->base.primary->fb->height = ((val >> 0) & 0xfff) + 1;
 
 	val = I915_READ(DSPSTRIDE(pipe));
-	crtc->base.primary->fb->pitches[0] = val & 0xffffff80;
+	crtc->base.primary->fb->pitches[0] = val & 0xffffffc0;
 
 	aligned_height = intel_align_height(dev, crtc->base.primary->fb->height,
 					    plane_config->tiled);
@@ -6289,8 +6709,8 @@ static bool i9xx_get_pipe_config(struct intel_crtc *crtc,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint32_t tmp;
 
-	if (!intel_display_power_enabled(dev_priv,
-					 POWER_DOMAIN_PIPE(crtc->pipe)))
+	if (!intel_display_power_is_enabled(dev_priv,
+					    POWER_DOMAIN_PIPE(crtc->pipe)))
 		return false;
 
 	pipe_config->cpu_transcoder = (enum transcoder) crtc->pipe;
@@ -6345,6 +6765,14 @@ static bool i9xx_get_pipe_config(struct intel_crtc *crtc,
 	}
 	pipe_config->dpll_hw_state.dpll = I915_READ(DPLL(crtc->pipe));
 	if (!IS_VALLEYVIEW(dev)) {
+		/*
+		 * DPLL_DVO_2X_MODE must be enabled for both DPLLs
+		 * on 830. Filter it out here so that we don't
+		 * report errors due to that.
+		 */
+		if (IS_I830(dev))
+			pipe_config->dpll_hw_state.dpll &= ~DPLL_DVO_2X_MODE;
+
 		pipe_config->dpll_hw_state.fp0 = I915_READ(FP0(crtc->pipe));
 		pipe_config->dpll_hw_state.fp1 = I915_READ(FP1(crtc->pipe));
 	} else {
@@ -6367,7 +6795,6 @@ static bool i9xx_get_pipe_config(struct intel_crtc *crtc,
 static void ironlake_init_pch_refclk(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_mode_config *mode_config = &dev->mode_config;
 	struct intel_encoder *encoder;
 	u32 val, final;
 	bool has_lvds = false;
@@ -6377,8 +6804,7 @@ static void ironlake_init_pch_refclk(struct drm_device *dev)
 	bool can_ssc = false;
 
 	/* We need to take the global config into account */
-	list_for_each_entry(encoder, &mode_config->encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		switch (encoder->type) {
 		case INTEL_OUTPUT_LVDS:
 			has_panel = true;
@@ -6389,6 +6815,8 @@ static void ironlake_init_pch_refclk(struct drm_device *dev)
 			if (enc_to_dig_port(&encoder->base)->port == PORT_A)
 				has_cpu_edp = true;
 			break;
+		default:
+			break;
 		}
 	}
 
@@ -6685,15 +7113,16 @@ static void lpt_disable_clkout_dp(struct drm_device *dev)
 
 static void lpt_init_pch_refclk(struct drm_device *dev)
 {
-	struct drm_mode_config *mode_config = &dev->mode_config;
 	struct intel_encoder *encoder;
 	bool has_vga = false;
 
-	list_for_each_entry(encoder, &mode_config->encoder_list, base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		switch (encoder->type) {
 		case INTEL_OUTPUT_ANALOG:
 			has_vga = true;
 			break;
+		default:
+			break;
 		}
 	}
 
@@ -6722,11 +7151,16 @@ static int ironlake_get_refclk(struct drm_crtc *crtc)
 	int num_connectors = 0;
 	bool is_lvds = false;
 
-	for_each_encoder_on_crtc(dev, crtc, encoder) {
+	for_each_intel_encoder(dev, encoder) {
+		if (encoder->new_crtc != to_intel_crtc(crtc))
+			continue;
+
 		switch (encoder->type) {
 		case INTEL_OUTPUT_LVDS:
 			is_lvds = true;
 			break;
+		default:
+			break;
 		}
 		num_connectors++;
 	}
@@ -6871,7 +7305,7 @@ static void haswell_set_pipeconf(struct drm_crtc *crtc)
 	I915_WRITE(GAMMA_MODE(intel_crtc->pipe), GAMMA_MODE_MODE_8BIT);
 	POSTING_READ(GAMMA_MODE(intel_crtc->pipe));
 
-	if (IS_BROADWELL(dev)) {
+	if (IS_BROADWELL(dev) || INTEL_INFO(dev)->gen >= 9) {
 		val = 0;
 
 		switch (intel_crtc->config.pipe_bpp) {
@@ -6906,18 +7340,12 @@ static bool ironlake_compute_clocks(struct drm_crtc *crtc,
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_encoder *intel_encoder;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int refclk;
 	const intel_limit_t *limit;
 	bool ret, is_lvds = false;
 
-	for_each_encoder_on_crtc(dev, crtc, intel_encoder) {
-		switch (intel_encoder->type) {
-		case INTEL_OUTPUT_LVDS:
-			is_lvds = true;
-			break;
-		}
-	}
+	is_lvds = intel_pipe_will_have_type(intel_crtc, INTEL_OUTPUT_LVDS);
 
 	refclk = ironlake_get_refclk(crtc);
 
@@ -6926,9 +7354,9 @@ static bool ironlake_compute_clocks(struct drm_crtc *crtc,
 	 * refclk, or FALSE.  The returned values represent the clock equation:
 	 * reflck * (5 * (m1 + 2) + (m2 + 2)) / (n + 2) / p1 / p2.
 	 */
-	limit = intel_limit(crtc, refclk);
-	ret = dev_priv->display.find_dpll(limit, crtc,
-					  to_intel_crtc(crtc)->config.port_clock,
+	limit = intel_limit(intel_crtc, refclk);
+	ret = dev_priv->display.find_dpll(limit, intel_crtc,
+					  intel_crtc->new_config->port_clock,
 					  refclk, NULL, clock);
 	if (!ret)
 		return false;
@@ -6941,7 +7369,7 @@ static bool ironlake_compute_clocks(struct drm_crtc *crtc,
 		 * downclock feature.
 		*/
 		*has_reduced_clock =
-			dev_priv->display.find_dpll(limit, crtc,
+			dev_priv->display.find_dpll(limit, intel_crtc,
 						    dev_priv->lvds_downclock,
 						    refclk, clock,
 						    reduced_clock);
@@ -6978,7 +7406,10 @@ static uint32_t ironlake_compute_dpll(struct intel_crtc *intel_crtc,
 	int factor, num_connectors = 0;
 	bool is_lvds = false, is_sdvo = false;
 
-	for_each_encoder_on_crtc(dev, crtc, intel_encoder) {
+	for_each_intel_encoder(dev, intel_encoder) {
+		if (intel_encoder->new_crtc != to_intel_crtc(crtc))
+			continue;
+
 		switch (intel_encoder->type) {
 		case INTEL_OUTPUT_LVDS:
 			is_lvds = true;
@@ -6987,6 +7418,8 @@ static uint32_t ironlake_compute_dpll(struct intel_crtc *intel_crtc,
 		case INTEL_OUTPUT_HDMI:
 			is_sdvo = true;
 			break;
+		default:
+			break;
 		}
 
 		num_connectors++;
@@ -6999,10 +7432,10 @@ static uint32_t ironlake_compute_dpll(struct intel_crtc *intel_crtc,
 		     dev_priv->vbt.lvds_ssc_freq == 100000) ||
 		    (HAS_PCH_IBX(dev) && intel_is_dual_link_lvds(dev)))
 			factor = 25;
-	} else if (intel_crtc->config.sdvo_tv_clock)
+	} else if (intel_crtc->new_config->sdvo_tv_clock)
 		factor = 20;
 
-	if (ironlake_needs_fb_cb_tune(&intel_crtc->config.dpll, factor))
+	if (ironlake_needs_fb_cb_tune(&intel_crtc->new_config->dpll, factor))
 		*fp |= FP_CB_TUNE;
 
 	if (fp2 && (reduced_clock->m < factor * reduced_clock->n))
@@ -7015,20 +7448,20 @@ static uint32_t ironlake_compute_dpll(struct intel_crtc *intel_crtc,
 	else
 		dpll |= DPLLB_MODE_DAC_SERIAL;
 
-	dpll |= (intel_crtc->config.pixel_multiplier - 1)
+	dpll |= (intel_crtc->new_config->pixel_multiplier - 1)
 		<< PLL_REF_SDVO_HDMI_MULTIPLIER_SHIFT;
 
 	if (is_sdvo)
 		dpll |= DPLL_SDVO_HIGH_SPEED;
-	if (intel_crtc->config.has_dp_encoder)
+	if (intel_crtc->new_config->has_dp_encoder)
 		dpll |= DPLL_SDVO_HIGH_SPEED;
 
 	/* compute bitmask from p1 value */
-	dpll |= (1 << (intel_crtc->config.dpll.p1 - 1)) << DPLL_FPA01_P1_POST_DIV_SHIFT;
+	dpll |= (1 << (intel_crtc->new_config->dpll.p1 - 1)) << DPLL_FPA01_P1_POST_DIV_SHIFT;
 	/* also FPA1 */
-	dpll |= (1 << (intel_crtc->config.dpll.p1 - 1)) << DPLL_FPA1_P1_POST_DIV_SHIFT;
+	dpll |= (1 << (intel_crtc->new_config->dpll.p1 - 1)) << DPLL_FPA1_P1_POST_DIV_SHIFT;
 
-	switch (intel_crtc->config.dpll.p2) {
+	switch (intel_crtc->new_config->dpll.p2) {
 	case 5:
 		dpll |= DPLL_DAC_SERIAL_P2_CLOCK_DIV_5;
 		break;
@@ -7051,78 +7484,64 @@ static uint32_t ironlake_compute_dpll(struct intel_crtc *intel_crtc,
 	return dpll | DPLL_VCO_ENABLE;
 }
 
-static int ironlake_crtc_mode_set(struct drm_crtc *crtc,
-				  int x, int y,
-				  struct drm_framebuffer *fb)
+static int ironlake_crtc_compute_clock(struct intel_crtc *crtc)
 {
-	struct drm_device *dev = crtc->dev;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int num_connectors = 0;
+	struct drm_device *dev = crtc->base.dev;
 	intel_clock_t clock, reduced_clock;
 	u32 dpll = 0, fp = 0, fp2 = 0;
 	bool ok, has_reduced_clock = false;
 	bool is_lvds = false;
-	struct intel_encoder *encoder;
 	struct intel_shared_dpll *pll;
 
-	for_each_encoder_on_crtc(dev, crtc, encoder) {
-		switch (encoder->type) {
-		case INTEL_OUTPUT_LVDS:
-			is_lvds = true;
-			break;
-		}
-
-		num_connectors++;
-	}
+	is_lvds = intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS);
 
 	WARN(!(HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev)),
 	     "Unexpected PCH type %d\n", INTEL_PCH_TYPE(dev));
 
-	ok = ironlake_compute_clocks(crtc, &clock,
+	ok = ironlake_compute_clocks(&crtc->base, &clock,
 				     &has_reduced_clock, &reduced_clock);
-	if (!ok && !intel_crtc->config.clock_set) {
+	if (!ok && !crtc->new_config->clock_set) {
 		DRM_ERROR("Couldn't find PLL settings for mode!\n");
 		return -EINVAL;
 	}
 	/* Compat-code for transition, will disappear. */
-	if (!intel_crtc->config.clock_set) {
-		intel_crtc->config.dpll.n = clock.n;
-		intel_crtc->config.dpll.m1 = clock.m1;
-		intel_crtc->config.dpll.m2 = clock.m2;
-		intel_crtc->config.dpll.p1 = clock.p1;
-		intel_crtc->config.dpll.p2 = clock.p2;
+	if (!crtc->new_config->clock_set) {
+		crtc->new_config->dpll.n = clock.n;
+		crtc->new_config->dpll.m1 = clock.m1;
+		crtc->new_config->dpll.m2 = clock.m2;
+		crtc->new_config->dpll.p1 = clock.p1;
+		crtc->new_config->dpll.p2 = clock.p2;
 	}
 
 	/* CPU eDP is the only output that doesn't need a PCH PLL of its own. */
-	if (intel_crtc->config.has_pch_encoder) {
-		fp = i9xx_dpll_compute_fp(&intel_crtc->config.dpll);
+	if (crtc->new_config->has_pch_encoder) {
+		fp = i9xx_dpll_compute_fp(&crtc->new_config->dpll);
 		if (has_reduced_clock)
 			fp2 = i9xx_dpll_compute_fp(&reduced_clock);
 
-		dpll = ironlake_compute_dpll(intel_crtc,
+		dpll = ironlake_compute_dpll(crtc,
 					     &fp, &reduced_clock,
 					     has_reduced_clock ? &fp2 : NULL);
 
-		intel_crtc->config.dpll_hw_state.dpll = dpll;
-		intel_crtc->config.dpll_hw_state.fp0 = fp;
+		crtc->new_config->dpll_hw_state.dpll = dpll;
+		crtc->new_config->dpll_hw_state.fp0 = fp;
 		if (has_reduced_clock)
-			intel_crtc->config.dpll_hw_state.fp1 = fp2;
+			crtc->new_config->dpll_hw_state.fp1 = fp2;
 		else
-			intel_crtc->config.dpll_hw_state.fp1 = fp;
+			crtc->new_config->dpll_hw_state.fp1 = fp;
 
-		pll = intel_get_shared_dpll(intel_crtc);
+		pll = intel_get_shared_dpll(crtc);
 		if (pll == NULL) {
 			DRM_DEBUG_DRIVER("failed to find PLL for pipe %c\n",
-					 pipe_name(intel_crtc->pipe));
+					 pipe_name(crtc->pipe));
 			return -EINVAL;
 		}
-	} else
-		intel_put_shared_dpll(intel_crtc);
+	}
 
 	if (is_lvds && has_reduced_clock && i915.powersave)
-		intel_crtc->lowfreq_avail = true;
+		crtc->lowfreq_avail = true;
 	else
-		intel_crtc->lowfreq_avail = false;
+		crtc->lowfreq_avail = false;
 
 	return 0;
 }
@@ -7145,7 +7564,8 @@ static void intel_pch_transcoder_get_m_n(struct intel_crtc *crtc,
 
 static void intel_cpu_transcoder_get_m_n(struct intel_crtc *crtc,
 					 enum transcoder transcoder,
-					 struct intel_link_m_n *m_n)
+					 struct intel_link_m_n *m_n,
+					 struct intel_link_m_n *m2_n2)
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -7159,6 +7579,20 @@ static void intel_cpu_transcoder_get_m_n(struct intel_crtc *crtc,
 		m_n->gmch_n = I915_READ(PIPE_DATA_N1(transcoder));
 		m_n->tu = ((I915_READ(PIPE_DATA_M1(transcoder))
 			    & TU_SIZE_MASK) >> TU_SIZE_SHIFT) + 1;
+		/* Read M2_N2 registers only for gen < 8 (M2_N2 available for
+		 * gen < 8) and if DRRS is supported (to make sure the
+		 * registers are not unnecessarily read).
+		 */
+		if (m2_n2 && INTEL_INFO(dev)->gen < 8 &&
+			crtc->config.has_drrs) {
+			m2_n2->link_m = I915_READ(PIPE_LINK_M2(transcoder));
+			m2_n2->link_n =	I915_READ(PIPE_LINK_N2(transcoder));
+			m2_n2->gmch_m =	I915_READ(PIPE_DATA_M2(transcoder))
+					& ~TU_SIZE_MASK;
+			m2_n2->gmch_n =	I915_READ(PIPE_DATA_N2(transcoder));
+			m2_n2->tu = ((I915_READ(PIPE_DATA_M2(transcoder))
+					& TU_SIZE_MASK) >> TU_SIZE_SHIFT) + 1;
+		}
 	} else {
 		m_n->link_m = I915_READ(PIPE_LINK_M_G4X(pipe));
 		m_n->link_n = I915_READ(PIPE_LINK_N_G4X(pipe));
@@ -7177,14 +7611,31 @@ void intel_dp_get_m_n(struct intel_crtc *crtc,
 		intel_pch_transcoder_get_m_n(crtc, &pipe_config->dp_m_n);
 	else
 		intel_cpu_transcoder_get_m_n(crtc, pipe_config->cpu_transcoder,
-					     &pipe_config->dp_m_n);
+					     &pipe_config->dp_m_n,
+					     &pipe_config->dp_m2_n2);
 }
 
 static void ironlake_get_fdi_m_n_config(struct intel_crtc *crtc,
 					struct intel_crtc_config *pipe_config)
 {
 	intel_cpu_transcoder_get_m_n(crtc, pipe_config->cpu_transcoder,
-				     &pipe_config->fdi_m_n);
+				     &pipe_config->fdi_m_n, NULL);
+}
+
+static void skylake_get_pfit_config(struct intel_crtc *crtc,
+				    struct intel_crtc_config *pipe_config)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	uint32_t tmp;
+
+	tmp = I915_READ(PS_CTL(crtc->pipe));
+
+	if (tmp & PS_ENABLE) {
+		pipe_config->pch_pfit.enabled = true;
+		pipe_config->pch_pfit.pos = I915_READ(PS_WIN_POS(crtc->pipe));
+		pipe_config->pch_pfit.size = I915_READ(PS_WIN_SZ(crtc->pipe));
+	}
 }
 
 static void ironlake_get_pfit_config(struct intel_crtc *crtc,
@@ -7255,7 +7706,7 @@ static void ironlake_get_plane_config(struct intel_crtc *crtc,
 	crtc->base.primary->fb->height = ((val >> 0) & 0xfff) + 1;
 
 	val = I915_READ(DSPSTRIDE(pipe));
-	crtc->base.primary->fb->pitches[0] = val & 0xffffff80;
+	crtc->base.primary->fb->pitches[0] = val & 0xffffffc0;
 
 	aligned_height = intel_align_height(dev, crtc->base.primary->fb->height,
 					    plane_config->tiled);
@@ -7278,8 +7729,8 @@ static bool ironlake_get_pipe_config(struct intel_crtc *crtc,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint32_t tmp;
 
-	if (!intel_display_power_enabled(dev_priv,
-					 POWER_DOMAIN_PIPE(crtc->pipe)))
+	if (!intel_display_power_is_enabled(dev_priv,
+					    POWER_DOMAIN_PIPE(crtc->pipe)))
 		return false;
 
 	pipe_config->cpu_transcoder = (enum transcoder) crtc->pipe;
@@ -7472,7 +7923,6 @@ static void hsw_disable_lcpll(struct drm_i915_private *dev_priv,
 static void hsw_restore_lcpll(struct drm_i915_private *dev_priv)
 {
 	uint32_t val;
-	unsigned long irqflags;
 
 	val = I915_READ(LCPLL_CTL);
 
@@ -7492,10 +7942,10 @@ static void hsw_restore_lcpll(struct drm_i915_private *dev_priv)
 	 * to call special forcewake code that doesn't touch runtime PM and
 	 * doesn't enable the forcewake delayed work.
 	 */
-	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
+	spin_lock_irq(&dev_priv->uncore.lock);
 	if (dev_priv->uncore.forcewake_count++ == 0)
 		dev_priv->uncore.funcs.force_wake_get(dev_priv, FORCEWAKE_ALL);
-	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
+	spin_unlock_irq(&dev_priv->uncore.lock);
 
 	if (val & LCPLL_POWER_DOWN_ALLOW) {
 		val &= ~LCPLL_POWER_DOWN_ALLOW;
@@ -7526,10 +7976,10 @@ static void hsw_restore_lcpll(struct drm_i915_private *dev_priv)
 	}
 
 	/* See the big comment above. */
-	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
+	spin_lock_irq(&dev_priv->uncore.lock);
 	if (--dev_priv->uncore.forcewake_count == 0)
 		dev_priv->uncore.funcs.force_wake_put(dev_priv, FORCEWAKE_ALL);
-	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
+	spin_unlock_irq(&dev_priv->uncore.lock);
 }
 
 /*
@@ -7591,28 +8041,52 @@ void hsw_disable_pc8(struct drm_i915_private *dev_priv)
 	intel_prepare_ddi(dev);
 }
 
-static void snb_modeset_global_resources(struct drm_device *dev)
+static int haswell_crtc_compute_clock(struct intel_crtc *crtc)
 {
-	modeset_update_crtc_power_domains(dev);
-}
+	if (!intel_ddi_pll_select(crtc))
+		return -EINVAL;
 
-static void haswell_modeset_global_resources(struct drm_device *dev)
-{
-	modeset_update_crtc_power_domains(dev);
+	crtc->lowfreq_avail = false;
+
+	return 0;
 }
 
-static int haswell_crtc_mode_set(struct drm_crtc *crtc,
-				 int x, int y,
-				 struct drm_framebuffer *fb)
+static void skylake_get_ddi_pll(struct drm_i915_private *dev_priv,
+				enum port port,
+				struct intel_crtc_config *pipe_config)
 {
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	u32 temp;
 
-	if (!intel_ddi_pll_select(intel_crtc))
-		return -EINVAL;
+	temp = I915_READ(DPLL_CTRL2) & DPLL_CTRL2_DDI_CLK_SEL_MASK(port);
+	pipe_config->ddi_pll_sel = temp >> (port * 3 + 1);
+
+	switch (pipe_config->ddi_pll_sel) {
+	case SKL_DPLL1:
+		pipe_config->shared_dpll = DPLL_ID_SKL_DPLL1;
+		break;
+	case SKL_DPLL2:
+		pipe_config->shared_dpll = DPLL_ID_SKL_DPLL2;
+		break;
+	case SKL_DPLL3:
+		pipe_config->shared_dpll = DPLL_ID_SKL_DPLL3;
+		break;
+	}
+}
 
-	intel_crtc->lowfreq_avail = false;
+static void haswell_get_ddi_pll(struct drm_i915_private *dev_priv,
+				enum port port,
+				struct intel_crtc_config *pipe_config)
+{
+	pipe_config->ddi_pll_sel = I915_READ(PORT_CLK_SEL(port));
 
-	return 0;
+	switch (pipe_config->ddi_pll_sel) {
+	case PORT_CLK_SEL_WRPLL1:
+		pipe_config->shared_dpll = DPLL_ID_WRPLL1;
+		break;
+	case PORT_CLK_SEL_WRPLL2:
+		pipe_config->shared_dpll = DPLL_ID_WRPLL2;
+		break;
+	}
 }
 
 static void haswell_get_ddi_port_state(struct intel_crtc *crtc,
@@ -7628,16 +8102,10 @@ static void haswell_get_ddi_port_state(struct intel_crtc *crtc,
 
 	port = (tmp & TRANS_DDI_PORT_MASK) >> TRANS_DDI_PORT_SHIFT;
 
-	pipe_config->ddi_pll_sel = I915_READ(PORT_CLK_SEL(port));
-
-	switch (pipe_config->ddi_pll_sel) {
-	case PORT_CLK_SEL_WRPLL1:
-		pipe_config->shared_dpll = DPLL_ID_WRPLL1;
-		break;
-	case PORT_CLK_SEL_WRPLL2:
-		pipe_config->shared_dpll = DPLL_ID_WRPLL2;
-		break;
-	}
+	if (IS_SKYLAKE(dev))
+		skylake_get_ddi_pll(dev_priv, port, pipe_config);
+	else
+		haswell_get_ddi_pll(dev_priv, port, pipe_config);
 
 	if (pipe_config->shared_dpll >= 0) {
 		pll = &dev_priv->shared_dplls[pipe_config->shared_dpll];
@@ -7651,7 +8119,8 @@ static void haswell_get_ddi_port_state(struct intel_crtc *crtc,
 	 * DDI E. So just check whether this pipe is wired to DDI E and whether
 	 * the PCH transcoder is on.
 	 */
-	if ((port == PORT_E) && I915_READ(LPT_TRANSCONF) & TRANS_ENABLE) {
+	if (INTEL_INFO(dev)->gen < 9 &&
+	    (port == PORT_E) && I915_READ(LPT_TRANSCONF) & TRANS_ENABLE) {
 		pipe_config->has_pch_encoder = true;
 
 		tmp = I915_READ(FDI_RX_CTL(PIPE_A));
@@ -7670,7 +8139,7 @@ static bool haswell_get_pipe_config(struct intel_crtc *crtc,
 	enum intel_display_power_domain pfit_domain;
 	uint32_t tmp;
 
-	if (!intel_display_power_enabled(dev_priv,
+	if (!intel_display_power_is_enabled(dev_priv,
 					 POWER_DOMAIN_PIPE(crtc->pipe)))
 		return false;
 
@@ -7699,7 +8168,7 @@ static bool haswell_get_pipe_config(struct intel_crtc *crtc,
 			pipe_config->cpu_transcoder = TRANSCODER_EDP;
 	}
 
-	if (!intel_display_power_enabled(dev_priv,
+	if (!intel_display_power_is_enabled(dev_priv,
 			POWER_DOMAIN_TRANSCODER(pipe_config->cpu_transcoder)))
 		return false;
 
@@ -7712,353 +8181,82 @@ static bool haswell_get_pipe_config(struct intel_crtc *crtc,
 	intel_get_pipe_timings(crtc, pipe_config);
 
 	pfit_domain = POWER_DOMAIN_PIPE_PANEL_FITTER(crtc->pipe);
-	if (intel_display_power_enabled(dev_priv, pfit_domain))
-		ironlake_get_pfit_config(crtc, pipe_config);
+	if (intel_display_power_is_enabled(dev_priv, pfit_domain)) {
+		if (IS_SKYLAKE(dev))
+			skylake_get_pfit_config(crtc, pipe_config);
+		else
+			ironlake_get_pfit_config(crtc, pipe_config);
+	}
 
 	if (IS_HASWELL(dev))
 		pipe_config->ips_enabled = hsw_crtc_supports_ips(crtc) &&
 			(I915_READ(IPS_CTL) & IPS_ENABLE);
 
-	pipe_config->pixel_multiplier = 1;
-
-	return true;
-}
-
-static struct {
-	int clock;
-	u32 config;
-} hdmi_audio_clock[] = {
-	{ DIV_ROUND_UP(25200 * 1000, 1001), AUD_CONFIG_PIXEL_CLOCK_HDMI_25175 },
-	{ 25200, AUD_CONFIG_PIXEL_CLOCK_HDMI_25200 }, /* default per bspec */
-	{ 27000, AUD_CONFIG_PIXEL_CLOCK_HDMI_27000 },
-	{ 27000 * 1001 / 1000, AUD_CONFIG_PIXEL_CLOCK_HDMI_27027 },
-	{ 54000, AUD_CONFIG_PIXEL_CLOCK_HDMI_54000 },
-	{ 54000 * 1001 / 1000, AUD_CONFIG_PIXEL_CLOCK_HDMI_54054 },
-	{ DIV_ROUND_UP(74250 * 1000, 1001), AUD_CONFIG_PIXEL_CLOCK_HDMI_74176 },
-	{ 74250, AUD_CONFIG_PIXEL_CLOCK_HDMI_74250 },
-	{ DIV_ROUND_UP(148500 * 1000, 1001), AUD_CONFIG_PIXEL_CLOCK_HDMI_148352 },
-	{ 148500, AUD_CONFIG_PIXEL_CLOCK_HDMI_148500 },
-};
-
-/* get AUD_CONFIG_PIXEL_CLOCK_HDMI_* value for mode */
-static u32 audio_config_hdmi_pixel_clock(struct drm_display_mode *mode)
-{
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(hdmi_audio_clock); i++) {
-		if (mode->clock == hdmi_audio_clock[i].clock)
-			break;
-	}
-
-	if (i == ARRAY_SIZE(hdmi_audio_clock)) {
-		DRM_DEBUG_KMS("HDMI audio pixel clock setting for %d not found, falling back to defaults\n", mode->clock);
-		i = 1;
+	if (pipe_config->cpu_transcoder != TRANSCODER_EDP) {
+		pipe_config->pixel_multiplier =
+			I915_READ(PIPE_MULT(pipe_config->cpu_transcoder)) + 1;
+	} else {
+		pipe_config->pixel_multiplier = 1;
 	}
 
-	DRM_DEBUG_KMS("Configuring HDMI audio for pixel clock %d (0x%08x)\n",
-		      hdmi_audio_clock[i].clock,
-		      hdmi_audio_clock[i].config);
-
-	return hdmi_audio_clock[i].config;
-}
-
-static bool intel_eld_uptodate(struct drm_connector *connector,
-			       int reg_eldv, uint32_t bits_eldv,
-			       int reg_elda, uint32_t bits_elda,
-			       int reg_edid)
-{
-	struct drm_i915_private *dev_priv = connector->dev->dev_private;
-	uint8_t *eld = connector->eld;
-	uint32_t i;
-
-	i = I915_READ(reg_eldv);
-	i &= bits_eldv;
-
-	if (!eld[0])
-		return !i;
-
-	if (!i)
-		return false;
-
-	i = I915_READ(reg_elda);
-	i &= ~bits_elda;
-	I915_WRITE(reg_elda, i);
-
-	for (i = 0; i < eld[2]; i++)
-		if (I915_READ(reg_edid) != *((uint32_t *)eld + i))
-			return false;
-
 	return true;
 }
 
-static void g4x_write_eld(struct drm_connector *connector,
-			  struct drm_crtc *crtc,
-			  struct drm_display_mode *mode)
+static void i845_update_cursor(struct drm_crtc *crtc, u32 base)
 {
-	struct drm_i915_private *dev_priv = connector->dev->dev_private;
-	uint8_t *eld = connector->eld;
-	uint32_t eldv;
-	uint32_t len;
-	uint32_t i;
-
-	i = I915_READ(G4X_AUD_VID_DID);
-
-	if (i == INTEL_AUDIO_DEVBLC || i == INTEL_AUDIO_DEVCL)
-		eldv = G4X_ELDV_DEVCL_DEVBLC;
-	else
-		eldv = G4X_ELDV_DEVCTG;
-
-	if (intel_eld_uptodate(connector,
-			       G4X_AUD_CNTL_ST, eldv,
-			       G4X_AUD_CNTL_ST, G4X_ELD_ADDR,
-			       G4X_HDMIW_HDMIEDID))
-		return;
-
-	i = I915_READ(G4X_AUD_CNTL_ST);
-	i &= ~(eldv | G4X_ELD_ADDR);
-	len = (i >> 9) & 0x1f;		/* ELD buffer size */
-	I915_WRITE(G4X_AUD_CNTL_ST, i);
-
-	if (!eld[0])
-		return;
-
-	len = min_t(uint8_t, eld[2], len);
-	DRM_DEBUG_DRIVER("ELD size %d\n", len);
-	for (i = 0; i < len; i++)
-		I915_WRITE(G4X_HDMIW_HDMIEDID, *((uint32_t *)eld + i));
-
-	i = I915_READ(G4X_AUD_CNTL_ST);
-	i |= eldv;
-	I915_WRITE(G4X_AUD_CNTL_ST, i);
-}
-
-static void haswell_write_eld(struct drm_connector *connector,
-			      struct drm_crtc *crtc,
-			      struct drm_display_mode *mode)
-{
-	struct drm_i915_private *dev_priv = connector->dev->dev_private;
-	uint8_t *eld = connector->eld;
-	uint32_t eldv;
-	uint32_t i;
-	int len;
-	int pipe = to_intel_crtc(crtc)->pipe;
-	int tmp;
-
-	int hdmiw_hdmiedid = HSW_AUD_EDID_DATA(pipe);
-	int aud_cntl_st = HSW_AUD_DIP_ELD_CTRL(pipe);
-	int aud_config = HSW_AUD_CFG(pipe);
-	int aud_cntrl_st2 = HSW_AUD_PIN_ELD_CP_VLD;
-
-	/* Audio output enable */
-	DRM_DEBUG_DRIVER("HDMI audio: enable codec\n");
-	tmp = I915_READ(aud_cntrl_st2);
-	tmp |= (AUDIO_OUTPUT_ENABLE_A << (pipe * 4));
-	I915_WRITE(aud_cntrl_st2, tmp);
-	POSTING_READ(aud_cntrl_st2);
-
-	assert_pipe_disabled(dev_priv, to_intel_crtc(crtc)->pipe);
-
-	/* Set ELD valid state */
-	tmp = I915_READ(aud_cntrl_st2);
-	DRM_DEBUG_DRIVER("HDMI audio: pin eld vld status=0x%08x\n", tmp);
-	tmp |= (AUDIO_ELD_VALID_A << (pipe * 4));
-	I915_WRITE(aud_cntrl_st2, tmp);
-	tmp = I915_READ(aud_cntrl_st2);
-	DRM_DEBUG_DRIVER("HDMI audio: eld vld status=0x%08x\n", tmp);
-
-	/* Enable HDMI mode */
-	tmp = I915_READ(aud_config);
-	DRM_DEBUG_DRIVER("HDMI audio: audio conf: 0x%08x\n", tmp);
-	/* clear N_programing_enable and N_value_index */
-	tmp &= ~(AUD_CONFIG_N_VALUE_INDEX | AUD_CONFIG_N_PROG_ENABLE);
-	I915_WRITE(aud_config, tmp);
-
-	DRM_DEBUG_DRIVER("ELD on pipe %c\n", pipe_name(pipe));
-
-	eldv = AUDIO_ELD_VALID_A << (pipe * 4);
-
-	if (intel_pipe_has_type(crtc, INTEL_OUTPUT_DISPLAYPORT)) {
-		DRM_DEBUG_DRIVER("ELD: DisplayPort detected\n");
-		eld[5] |= (1 << 2);	/* Conn_Type, 0x1 = DisplayPort */
-		I915_WRITE(aud_config, AUD_CONFIG_N_VALUE_INDEX); /* 0x1 = DP */
-	} else {
-		I915_WRITE(aud_config, audio_config_hdmi_pixel_clock(mode));
-	}
-
-	if (intel_eld_uptodate(connector,
-			       aud_cntrl_st2, eldv,
-			       aud_cntl_st, IBX_ELD_ADDRESS,
-			       hdmiw_hdmiedid))
-		return;
-
-	i = I915_READ(aud_cntrl_st2);
-	i &= ~eldv;
-	I915_WRITE(aud_cntrl_st2, i);
-
-	if (!eld[0])
-		return;
+	struct drm_device *dev = crtc->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	uint32_t cntl = 0, size = 0;
 
-	i = I915_READ(aud_cntl_st);
-	i &= ~IBX_ELD_ADDRESS;
-	I915_WRITE(aud_cntl_st, i);
-	i = (i >> 29) & DIP_PORT_SEL_MASK;		/* DIP_Port_Select, 0x1 = PortB */
-	DRM_DEBUG_DRIVER("port num:%d\n", i);
-
-	len = min_t(uint8_t, eld[2], 21);	/* 84 bytes of hw ELD buffer */
-	DRM_DEBUG_DRIVER("ELD size %d\n", len);
-	for (i = 0; i < len; i++)
-		I915_WRITE(hdmiw_hdmiedid, *((uint32_t *)eld + i));
-
-	i = I915_READ(aud_cntrl_st2);
-	i |= eldv;
-	I915_WRITE(aud_cntrl_st2, i);
-
-}
-
-static void ironlake_write_eld(struct drm_connector *connector,
-			       struct drm_crtc *crtc,
-			       struct drm_display_mode *mode)
-{
-	struct drm_i915_private *dev_priv = connector->dev->dev_private;
-	uint8_t *eld = connector->eld;
-	uint32_t eldv;
-	uint32_t i;
-	int len;
-	int hdmiw_hdmiedid;
-	int aud_config;
-	int aud_cntl_st;
-	int aud_cntrl_st2;
-	int pipe = to_intel_crtc(crtc)->pipe;
-
-	if (HAS_PCH_IBX(connector->dev)) {
-		hdmiw_hdmiedid = IBX_HDMIW_HDMIEDID(pipe);
-		aud_config = IBX_AUD_CFG(pipe);
-		aud_cntl_st = IBX_AUD_CNTL_ST(pipe);
-		aud_cntrl_st2 = IBX_AUD_CNTL_ST2;
-	} else if (IS_VALLEYVIEW(connector->dev)) {
-		hdmiw_hdmiedid = VLV_HDMIW_HDMIEDID(pipe);
-		aud_config = VLV_AUD_CFG(pipe);
-		aud_cntl_st = VLV_AUD_CNTL_ST(pipe);
-		aud_cntrl_st2 = VLV_AUD_CNTL_ST2;
-	} else {
-		hdmiw_hdmiedid = CPT_HDMIW_HDMIEDID(pipe);
-		aud_config = CPT_AUD_CFG(pipe);
-		aud_cntl_st = CPT_AUD_CNTL_ST(pipe);
-		aud_cntrl_st2 = CPT_AUD_CNTRL_ST2;
-	}
+	if (base) {
+		unsigned int width = intel_crtc->cursor_width;
+		unsigned int height = intel_crtc->cursor_height;
+		unsigned int stride = roundup_pow_of_two(width) * 4;
 
-	DRM_DEBUG_DRIVER("ELD on pipe %c\n", pipe_name(pipe));
+		switch (stride) {
+		default:
+			WARN_ONCE(1, "Invalid cursor width/stride, width=%u, stride=%u\n",
+				  width, stride);
+			stride = 256;
+			/* fallthrough */
+		case 256:
+		case 512:
+		case 1024:
+		case 2048:
+			break;
+		}
 
-	if (IS_VALLEYVIEW(connector->dev))  {
-		struct intel_encoder *intel_encoder;
-		struct intel_digital_port *intel_dig_port;
+		cntl |= CURSOR_ENABLE |
+			CURSOR_GAMMA_ENABLE |
+			CURSOR_FORMAT_ARGB |
+			CURSOR_STRIDE(stride);
 
-		intel_encoder = intel_attached_encoder(connector);
-		intel_dig_port = enc_to_dig_port(&intel_encoder->base);
-		i = intel_dig_port->port;
-	} else {
-		i = I915_READ(aud_cntl_st);
-		i = (i >> 29) & DIP_PORT_SEL_MASK;
-		/* DIP_Port_Select, 0x1 = PortB */
+		size = (height << 12) | width;
 	}
 
-	if (!i) {
-		DRM_DEBUG_DRIVER("Audio directed to unknown port\n");
-		/* operate blindly on all ports */
-		eldv = IBX_ELD_VALIDB;
-		eldv |= IBX_ELD_VALIDB << 4;
-		eldv |= IBX_ELD_VALIDB << 8;
-	} else {
-		DRM_DEBUG_DRIVER("ELD on port %c\n", port_name(i));
-		eldv = IBX_ELD_VALIDB << ((i - 1) * 4);
+	if (intel_crtc->cursor_cntl != 0 &&
+	    (intel_crtc->cursor_base != base ||
+	     intel_crtc->cursor_size != size ||
+	     intel_crtc->cursor_cntl != cntl)) {
+		/* On these chipsets we can only modify the base/size/stride
+		 * whilst the cursor is disabled.
+		 */
+		I915_WRITE(_CURACNTR, 0);
+		POSTING_READ(_CURACNTR);
+		intel_crtc->cursor_cntl = 0;
 	}
 
-	if (intel_pipe_has_type(crtc, INTEL_OUTPUT_DISPLAYPORT)) {
-		DRM_DEBUG_DRIVER("ELD: DisplayPort detected\n");
-		eld[5] |= (1 << 2);	/* Conn_Type, 0x1 = DisplayPort */
-		I915_WRITE(aud_config, AUD_CONFIG_N_VALUE_INDEX); /* 0x1 = DP */
-	} else {
-		I915_WRITE(aud_config, audio_config_hdmi_pixel_clock(mode));
+	if (intel_crtc->cursor_base != base) {
+		I915_WRITE(_CURABASE, base);
+		intel_crtc->cursor_base = base;
 	}
 
-	if (intel_eld_uptodate(connector,
-			       aud_cntrl_st2, eldv,
-			       aud_cntl_st, IBX_ELD_ADDRESS,
-			       hdmiw_hdmiedid))
-		return;
-
-	i = I915_READ(aud_cntrl_st2);
-	i &= ~eldv;
-	I915_WRITE(aud_cntrl_st2, i);
-
-	if (!eld[0])
-		return;
-
-	i = I915_READ(aud_cntl_st);
-	i &= ~IBX_ELD_ADDRESS;
-	I915_WRITE(aud_cntl_st, i);
-
-	len = min_t(uint8_t, eld[2], 21);	/* 84 bytes of hw ELD buffer */
-	DRM_DEBUG_DRIVER("ELD size %d\n", len);
-	for (i = 0; i < len; i++)
-		I915_WRITE(hdmiw_hdmiedid, *((uint32_t *)eld + i));
-
-	i = I915_READ(aud_cntrl_st2);
-	i |= eldv;
-	I915_WRITE(aud_cntrl_st2, i);
-}
-
-void intel_write_eld(struct drm_encoder *encoder,
-		     struct drm_display_mode *mode)
-{
-	struct drm_crtc *crtc = encoder->crtc;
-	struct drm_connector *connector;
-	struct drm_device *dev = encoder->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	connector = drm_select_eld(encoder, mode);
-	if (!connector)
-		return;
-
-	DRM_DEBUG_DRIVER("ELD on [CONNECTOR:%d:%s], [ENCODER:%d:%s]\n",
-			 connector->base.id,
-			 connector->name,
-			 connector->encoder->base.id,
-			 connector->encoder->name);
-
-	connector->eld[6] = drm_av_sync_delay(connector, mode) / 2;
-
-	if (dev_priv->display.write_eld)
-		dev_priv->display.write_eld(connector, crtc, mode);
-}
-
-static void i845_update_cursor(struct drm_crtc *crtc, u32 base)
-{
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	uint32_t cntl;
-
-	if (base != intel_crtc->cursor_base) {
-		/* On these chipsets we can only modify the base whilst
-		 * the cursor is disabled.
-		 */
-		if (intel_crtc->cursor_cntl) {
-			I915_WRITE(_CURACNTR, 0);
-			POSTING_READ(_CURACNTR);
-			intel_crtc->cursor_cntl = 0;
-		}
-
-		I915_WRITE(_CURABASE, base);
-		POSTING_READ(_CURABASE);
+	if (intel_crtc->cursor_size != size) {
+		I915_WRITE(CURSIZE, size);
+		intel_crtc->cursor_size = size;
 	}
 
-	/* XXX width must be 64, stride 256 => 0x00 << 28 */
-	cntl = 0;
-	if (base)
-		cntl = (CURSOR_ENABLE |
-			CURSOR_GAMMA_ENABLE |
-			CURSOR_FORMAT_ARGB);
 	if (intel_crtc->cursor_cntl != cntl) {
 		I915_WRITE(_CURACNTR, cntl);
 		POSTING_READ(_CURACNTR);
@@ -8092,46 +8290,13 @@ static void i9xx_update_cursor(struct drm_crtc *crtc, u32 base)
 				return;
 		}
 		cntl |= pipe << 28; /* Connect to correct pipe */
-	}
-	if (intel_crtc->cursor_cntl != cntl) {
-		I915_WRITE(CURCNTR(pipe), cntl);
-		POSTING_READ(CURCNTR(pipe));
-		intel_crtc->cursor_cntl = cntl;
-	}
-
-	/* and commit changes on next vblank */
-	I915_WRITE(CURBASE(pipe), base);
-	POSTING_READ(CURBASE(pipe));
-}
 
-static void ivb_update_cursor(struct drm_crtc *crtc, u32 base)
-{
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
-	uint32_t cntl;
-
-	cntl = 0;
-	if (base) {
-		cntl = MCURSOR_GAMMA_ENABLE;
-		switch (intel_crtc->cursor_width) {
-			case 64:
-				cntl |= CURSOR_MODE_64_ARGB_AX;
-				break;
-			case 128:
-				cntl |= CURSOR_MODE_128_ARGB_AX;
-				break;
-			case 256:
-				cntl |= CURSOR_MODE_256_ARGB_AX;
-				break;
-			default:
-				WARN_ON(1);
-				return;
-		}
+		if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+			cntl |= CURSOR_PIPE_CSC_ENABLE;
 	}
-	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
-		cntl |= CURSOR_PIPE_CSC_ENABLE;
+
+	if (to_intel_plane(crtc->cursor)->rotation == BIT(DRM_ROTATE_180))
+		cntl |= CURSOR_ROTATE_180;
 
 	if (intel_crtc->cursor_cntl != cntl) {
 		I915_WRITE(CURCNTR(pipe), cntl);
@@ -8142,6 +8307,8 @@ static void ivb_update_cursor(struct drm_crtc *crtc, u32 base)
 	/* and commit changes on next vblank */
 	I915_WRITE(CURBASE(pipe), base);
 	POSTING_READ(CURBASE(pipe));
+
+	intel_crtc->cursor_base = base;
 }
 
 /* If no-part of the cursor is visible on the framebuffer, then the GPU may hang... */
@@ -8188,28 +8355,62 @@ static void intel_crtc_update_cursor(struct drm_crtc *crtc,
 
 	I915_WRITE(CURPOS(pipe), pos);
 
-	if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev) || IS_BROADWELL(dev))
-		ivb_update_cursor(crtc, base);
-	else if (IS_845G(dev) || IS_I865G(dev))
+	/* ILK+ do this automagically */
+	if (HAS_GMCH_DISPLAY(dev) &&
+		to_intel_plane(crtc->cursor)->rotation == BIT(DRM_ROTATE_180)) {
+		base += (intel_crtc->cursor_height *
+			intel_crtc->cursor_width - 1) * 4;
+	}
+
+	if (IS_845G(dev) || IS_I865G(dev))
 		i845_update_cursor(crtc, base);
 	else
 		i9xx_update_cursor(crtc, base);
-	intel_crtc->cursor_base = base;
 }
 
-/*
- * intel_crtc_cursor_set_obj - Set cursor to specified GEM object
- *
- * Note that the object's reference will be consumed if the update fails.  If
- * the update succeeds, the reference of the old object (if any) will be
- * consumed.
- */
+static bool cursor_size_ok(struct drm_device *dev,
+			   uint32_t width, uint32_t height)
+{
+	if (width == 0 || height == 0)
+		return false;
+
+	/*
+	 * 845g/865g are special in that they are only limited by
+	 * the width of their cursors, the height is arbitrary up to
+	 * the precision of the register. Everything else requires
+	 * square cursors, limited to a few power-of-two sizes.
+	 */
+	if (IS_845G(dev) || IS_I865G(dev)) {
+		if ((width & 63) != 0)
+			return false;
+
+		if (width > (IS_845G(dev) ? 64 : 512))
+			return false;
+
+		if (height > 1023)
+			return false;
+	} else {
+		switch (width | height) {
+		case 256:
+		case 128:
+			if (IS_GEN2(dev))
+				return false;
+		case 64:
+			break;
+		default:
+			return false;
+		}
+	}
+
+	return true;
+}
+
 static int intel_crtc_cursor_set_obj(struct drm_crtc *crtc,
 				     struct drm_i915_gem_object *obj,
 				     uint32_t width, uint32_t height)
 {
 	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	enum pipe pipe = intel_crtc->pipe;
 	unsigned old_width;
@@ -8220,36 +8421,15 @@ static int intel_crtc_cursor_set_obj(struct drm_crtc *crtc,
 	if (!obj) {
 		DRM_DEBUG_KMS("cursor off\n");
 		addr = 0;
-		obj = NULL;
 		mutex_lock(&dev->struct_mutex);
 		goto finish;
 	}
 
-	/* Check for which cursor types we support */
-	if (!((width == 64 && height == 64) ||
-			(width == 128 && height == 128 && !IS_GEN2(dev)) ||
-			(width == 256 && height == 256 && !IS_GEN2(dev)))) {
-		DRM_DEBUG("Cursor dimension not supported\n");
-		return -EINVAL;
-	}
-
-	if (obj->base.size < width * height * 4) {
-		DRM_DEBUG_KMS("buffer is too small\n");
-		ret = -ENOMEM;
-		goto fail;
-	}
-
 	/* we only need to pin inside GTT if cursor is non-phy */
 	mutex_lock(&dev->struct_mutex);
 	if (!INTEL_INFO(dev)->cursor_needs_physical) {
 		unsigned alignment;
 
-		if (obj->tiling_mode) {
-			DRM_DEBUG_KMS("cursor cannot be tiled\n");
-			ret = -EINVAL;
-			goto fail_locked;
-		}
-
 		/*
 		 * Global gtt pte registers are special registers which actually
 		 * forward writes to a chunk of system memory. Which means that
@@ -8295,9 +8475,6 @@ static int intel_crtc_cursor_set_obj(struct drm_crtc *crtc,
 		addr = obj->phys_handle->busaddr;
 	}
 
-	if (IS_GEN2(dev))
-		I915_WRITE(CURSIZE, (height << 12) | width);
-
  finish:
 	if (intel_crtc->cursor_bo) {
 		if (!INTEL_INFO(dev)->cursor_needs_physical)
@@ -8319,17 +8496,15 @@ static int intel_crtc_cursor_set_obj(struct drm_crtc *crtc,
 		if (old_width != width)
 			intel_update_watermarks(crtc);
 		intel_crtc_update_cursor(crtc, intel_crtc->cursor_bo != NULL);
-	}
 
-	intel_frontbuffer_flip(dev, INTEL_FRONTBUFFER_CURSOR(pipe));
+		intel_frontbuffer_flip(dev, INTEL_FRONTBUFFER_CURSOR(pipe));
+	}
 
 	return 0;
 fail_unpin:
 	i915_gem_object_unpin_from_display_plane(obj);
 fail_locked:
 	mutex_unlock(&dev->struct_mutex);
-fail:
-	drm_gem_object_unreference_unlocked(&obj->base);
 	return ret;
 }
 
@@ -8364,7 +8539,7 @@ __intel_framebuffer_create(struct drm_device *dev,
 
 	intel_fb = kzalloc(sizeof(*intel_fb), GFP_KERNEL);
 	if (!intel_fb) {
-		drm_gem_object_unreference_unlocked(&obj->base);
+		drm_gem_object_unreference(&obj->base);
 		return ERR_PTR(-ENOMEM);
 	}
 
@@ -8374,7 +8549,7 @@ __intel_framebuffer_create(struct drm_device *dev,
 
 	return &intel_fb->base;
 err:
-	drm_gem_object_unreference_unlocked(&obj->base);
+	drm_gem_object_unreference(&obj->base);
 	kfree(intel_fb);
 
 	return ERR_PTR(ret);
@@ -8507,6 +8682,9 @@ retry:
 		ret = drm_modeset_lock(&crtc->mutex, ctx);
 		if (ret)
 			goto fail_unlock;
+		ret = drm_modeset_lock(&crtc->primary->mutex, ctx);
+		if (ret)
+			goto fail_unlock;
 
 		old->dpms_mode = connector->dpms;
 		old->load_detect_temp = false;
@@ -8544,6 +8722,9 @@ retry:
 	ret = drm_modeset_lock(&crtc->mutex, ctx);
 	if (ret)
 		goto fail_unlock;
+	ret = drm_modeset_lock(&crtc->primary->mutex, ctx);
+	if (ret)
+		goto fail_unlock;
 	intel_encoder->new_crtc = to_intel_crtc(crtc);
 	to_intel_connector(connector)->new_encoder = intel_encoder;
 
@@ -8826,35 +9007,6 @@ struct drm_display_mode *intel_crtc_mode_get(struct drm_device *dev,
 	return mode;
 }
 
-static void intel_increase_pllclock(struct drm_device *dev,
-				    enum pipe pipe)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int dpll_reg = DPLL(pipe);
-	int dpll;
-
-	if (!HAS_GMCH_DISPLAY(dev))
-		return;
-
-	if (!dev_priv->lvds_downclock_avail)
-		return;
-
-	dpll = I915_READ(dpll_reg);
-	if (!HAS_PIPE_CXSR(dev) && (dpll & DISPLAY_RATE_SELECT_FPA1)) {
-		DRM_DEBUG_DRIVER("upclocking LVDS\n");
-
-		assert_panel_unlocked(dev_priv, pipe);
-
-		dpll &= ~DISPLAY_RATE_SELECT_FPA1;
-		I915_WRITE(dpll_reg, dpll);
-		intel_wait_for_vblank(dev, pipe);
-
-		dpll = I915_READ(dpll_reg);
-		if (dpll & DISPLAY_RATE_SELECT_FPA1)
-			DRM_DEBUG_DRIVER("failed to upclock LVDS!\n");
-	}
-}
-
 static void intel_decrease_pllclock(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
@@ -8930,190 +9082,16 @@ out:
 	intel_runtime_pm_put(dev_priv);
 }
 
-
-/**
- * intel_mark_fb_busy - mark given planes as busy
- * @dev: DRM device
- * @frontbuffer_bits: bits for the affected planes
- * @ring: optional ring for asynchronous commands
- *
- * This function gets called every time the screen contents change. It can be
- * used to keep e.g. the update rate at the nominal refresh rate with DRRS.
- */
-static void intel_mark_fb_busy(struct drm_device *dev,
-			       unsigned frontbuffer_bits,
-			       struct intel_engine_cs *ring)
-{
-	enum pipe pipe;
-
-	if (!i915.powersave)
-		return;
-
-	for_each_pipe(pipe) {
-		if (!(frontbuffer_bits & INTEL_FRONTBUFFER_ALL_MASK(pipe)))
-			continue;
-
-		intel_increase_pllclock(dev, pipe);
-		if (ring && intel_fbc_enabled(dev))
-			ring->fbc_dirty = true;
-	}
-}
-
-/**
- * intel_fb_obj_invalidate - invalidate frontbuffer object
- * @obj: GEM object to invalidate
- * @ring: set for asynchronous rendering
- *
- * This function gets called every time rendering on the given object starts and
- * frontbuffer caching (fbc, low refresh rate for DRRS, panel self refresh) must
- * be invalidated. If @ring is non-NULL any subsequent invalidation will be delayed
- * until the rendering completes or a flip on this frontbuffer plane is
- * scheduled.
- */
-void intel_fb_obj_invalidate(struct drm_i915_gem_object *obj,
-			     struct intel_engine_cs *ring)
-{
-	struct drm_device *dev = obj->base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	WARN_ON(!mutex_is_locked(&dev->struct_mutex));
-
-	if (!obj->frontbuffer_bits)
-		return;
-
-	if (ring) {
-		mutex_lock(&dev_priv->fb_tracking.lock);
-		dev_priv->fb_tracking.busy_bits
-			|= obj->frontbuffer_bits;
-		dev_priv->fb_tracking.flip_bits
-			&= ~obj->frontbuffer_bits;
-		mutex_unlock(&dev_priv->fb_tracking.lock);
-	}
-
-	intel_mark_fb_busy(dev, obj->frontbuffer_bits, ring);
-
-	intel_edp_psr_invalidate(dev, obj->frontbuffer_bits);
-}
-
-/**
- * intel_frontbuffer_flush - flush frontbuffer
- * @dev: DRM device
- * @frontbuffer_bits: frontbuffer plane tracking bits
- *
- * This function gets called every time rendering on the given planes has
- * completed and frontbuffer caching can be started again. Flushes will get
- * delayed if they're blocked by some oustanding asynchronous rendering.
- *
- * Can be called without any locks held.
- */
-void intel_frontbuffer_flush(struct drm_device *dev,
-			     unsigned frontbuffer_bits)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	/* Delay flushing when rings are still busy.*/
-	mutex_lock(&dev_priv->fb_tracking.lock);
-	frontbuffer_bits &= ~dev_priv->fb_tracking.busy_bits;
-	mutex_unlock(&dev_priv->fb_tracking.lock);
-
-	intel_mark_fb_busy(dev, frontbuffer_bits, NULL);
-
-	intel_edp_psr_flush(dev, frontbuffer_bits);
-}
-
-/**
- * intel_fb_obj_flush - flush frontbuffer object
- * @obj: GEM object to flush
- * @retire: set when retiring asynchronous rendering
- *
- * This function gets called every time rendering on the given object has
- * completed and frontbuffer caching can be started again. If @retire is true
- * then any delayed flushes will be unblocked.
- */
-void intel_fb_obj_flush(struct drm_i915_gem_object *obj,
-			bool retire)
-{
-	struct drm_device *dev = obj->base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	unsigned frontbuffer_bits;
-
-	WARN_ON(!mutex_is_locked(&dev->struct_mutex));
-
-	if (!obj->frontbuffer_bits)
-		return;
-
-	frontbuffer_bits = obj->frontbuffer_bits;
-
-	if (retire) {
-		mutex_lock(&dev_priv->fb_tracking.lock);
-		/* Filter out new bits since rendering started. */
-		frontbuffer_bits &= dev_priv->fb_tracking.busy_bits;
-
-		dev_priv->fb_tracking.busy_bits &= ~frontbuffer_bits;
-		mutex_unlock(&dev_priv->fb_tracking.lock);
-	}
-
-	intel_frontbuffer_flush(dev, frontbuffer_bits);
-}
-
-/**
- * intel_frontbuffer_flip_prepare - prepare asnychronous frontbuffer flip
- * @dev: DRM device
- * @frontbuffer_bits: frontbuffer plane tracking bits
- *
- * This function gets called after scheduling a flip on @obj. The actual
- * frontbuffer flushing will be delayed until completion is signalled with
- * intel_frontbuffer_flip_complete. If an invalidate happens in between this
- * flush will be cancelled.
- *
- * Can be called without any locks held.
- */
-void intel_frontbuffer_flip_prepare(struct drm_device *dev,
-				    unsigned frontbuffer_bits)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	mutex_lock(&dev_priv->fb_tracking.lock);
-	dev_priv->fb_tracking.flip_bits
-		|= frontbuffer_bits;
-	mutex_unlock(&dev_priv->fb_tracking.lock);
-}
-
-/**
- * intel_frontbuffer_flip_complete - complete asynchronous frontbuffer flush
- * @dev: DRM device
- * @frontbuffer_bits: frontbuffer plane tracking bits
- *
- * This function gets called after the flip has been latched and will complete
- * on the next vblank. It will execute the fush if it hasn't been cancalled yet.
- *
- * Can be called without any locks held.
- */
-void intel_frontbuffer_flip_complete(struct drm_device *dev,
-				     unsigned frontbuffer_bits)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	mutex_lock(&dev_priv->fb_tracking.lock);
-	/* Mask any cancelled flips. */
-	frontbuffer_bits &= dev_priv->fb_tracking.flip_bits;
-	dev_priv->fb_tracking.flip_bits &= ~frontbuffer_bits;
-	mutex_unlock(&dev_priv->fb_tracking.lock);
-
-	intel_frontbuffer_flush(dev, frontbuffer_bits);
-}
-
 static void intel_crtc_destroy(struct drm_crtc *crtc)
 {
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct intel_unpin_work *work;
-	unsigned long flags;
 
-	spin_lock_irqsave(&dev->event_lock, flags);
+	spin_lock_irq(&dev->event_lock);
 	work = intel_crtc->unpin_work;
 	intel_crtc->unpin_work = NULL;
-	spin_unlock_irqrestore(&dev->event_lock, flags);
+	spin_unlock_irq(&dev->event_lock);
 
 	if (work) {
 		cancel_work_sync(&work->work);
@@ -9151,7 +9129,6 @@ static void intel_unpin_work_fn(struct work_struct *__work)
 static void do_intel_finish_page_flip(struct drm_device *dev,
 				      struct drm_crtc *crtc)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_unpin_work *work;
 	unsigned long flags;
@@ -9160,6 +9137,10 @@ static void do_intel_finish_page_flip(struct drm_device *dev,
 	if (intel_crtc == NULL)
 		return;
 
+	/*
+	 * This is called both by irq handlers and the reset code (to complete
+	 * lost pageflips) so needs the full irqsave spinlocks.
+	 */
 	spin_lock_irqsave(&dev->event_lock, flags);
 	work = intel_crtc->unpin_work;
 
@@ -9171,23 +9152,9 @@ static void do_intel_finish_page_flip(struct drm_device *dev,
 		return;
 	}
 
-	/* and that the unpin work is consistent wrt ->pending. */
-	smp_rmb();
-
-	intel_crtc->unpin_work = NULL;
-
-	if (work->event)
-		drm_send_vblank_event(dev, intel_crtc->pipe, work->event);
-
-	drm_crtc_vblank_put(crtc);
+	page_flip_completed(intel_crtc);
 
 	spin_unlock_irqrestore(&dev->event_lock, flags);
-
-	wake_up_all(&dev_priv->pending_flip_queue);
-
-	queue_work(dev_priv->wq, &work->work);
-
-	trace_i915_flip_complete(intel_crtc->plane, work->pending_flip_obj);
 }
 
 void intel_finish_page_flip(struct drm_device *dev, int pipe)
@@ -9217,6 +9184,10 @@ static bool page_flip_finished(struct intel_crtc *crtc)
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
+	if (i915_reset_in_progress(&dev_priv->gpu_error) ||
+	    crtc->reset_counter != atomic_read(&dev_priv->gpu_error.reset_counter))
+		return true;
+
 	/*
 	 * The relevant registers doen't exist on pre-ctg.
 	 * As the flip done interrupt doesn't trigger for mmio
@@ -9255,7 +9226,12 @@ void intel_prepare_page_flip(struct drm_device *dev, int plane)
 		to_intel_crtc(dev_priv->plane_to_crtc_mapping[plane]);
 	unsigned long flags;
 
-	/* NB: An MMIO update of the plane base pointer will also
+
+	/*
+	 * This is called both by irq handlers and the reset code (to complete
+	 * lost pageflips) so needs the full irqsave spinlocks.
+	 *
+	 * NB: An MMIO update of the plane base pointer will also
 	 * generate a page-flip completion irq, i.e. every modeset
 	 * is also accompanied by a spurious intel_prepare_page_flip().
 	 */
@@ -9532,6 +9508,8 @@ static bool use_mmio_flip(struct intel_engine_cs *ring,
 		return false;
 	else if (i915.use_mmio_flip > 0)
 		return true;
+	else if (i915.enable_execlists)
+		return true;
 	else
 		return ring != obj->ring;
 }
@@ -9543,115 +9521,128 @@ static void intel_do_mmio_flip(struct intel_crtc *intel_crtc)
 	struct intel_framebuffer *intel_fb =
 		to_intel_framebuffer(intel_crtc->base.primary->fb);
 	struct drm_i915_gem_object *obj = intel_fb->obj;
+	bool atomic_update;
+	u32 start_vbl_count;
 	u32 dspcntr;
 	u32 reg;
 
 	intel_mark_page_flip_active(intel_crtc);
 
+	atomic_update = intel_pipe_update_start(intel_crtc, &start_vbl_count);
+
 	reg = DSPCNTR(intel_crtc->plane);
 	dspcntr = I915_READ(reg);
 
-	if (INTEL_INFO(dev)->gen >= 4) {
-		if (obj->tiling_mode != I915_TILING_NONE)
-			dspcntr |= DISPPLANE_TILED;
-		else
-			dspcntr &= ~DISPPLANE_TILED;
-	}
+	if (obj->tiling_mode != I915_TILING_NONE)
+		dspcntr |= DISPPLANE_TILED;
+	else
+		dspcntr &= ~DISPPLANE_TILED;
+
 	I915_WRITE(reg, dspcntr);
 
 	I915_WRITE(DSPSURF(intel_crtc->plane),
 		   intel_crtc->unpin_work->gtt_offset);
 	POSTING_READ(DSPSURF(intel_crtc->plane));
+
+	if (atomic_update)
+		intel_pipe_update_end(intel_crtc, start_vbl_count);
 }
 
-static int intel_postpone_flip(struct drm_i915_gem_object *obj)
+static void intel_mmio_flip_work_func(struct work_struct *work)
 {
+	struct intel_crtc *intel_crtc =
+		container_of(work, struct intel_crtc, mmio_flip.work);
 	struct intel_engine_cs *ring;
-	int ret;
-
-	lockdep_assert_held(&obj->base.dev->struct_mutex);
+	uint32_t seqno;
 
-	if (!obj->last_write_seqno)
-		return 0;
-
-	ring = obj->ring;
-
-	if (i915_seqno_passed(ring->get_seqno(ring, true),
-			      obj->last_write_seqno))
-		return 0;
-
-	ret = i915_gem_check_olr(ring, obj->last_write_seqno);
-	if (ret)
-		return ret;
+	seqno = intel_crtc->mmio_flip.seqno;
+	ring = intel_crtc->mmio_flip.ring;
 
-	if (WARN_ON(!ring->irq_get(ring)))
-		return 0;
+	if (seqno)
+		WARN_ON(__i915_wait_seqno(ring, seqno,
+					  intel_crtc->reset_counter,
+					  false, NULL, NULL) != 0);
 
-	return 1;
+	intel_do_mmio_flip(intel_crtc);
 }
 
-void intel_notify_mmio_flip(struct intel_engine_cs *ring)
+static int intel_queue_mmio_flip(struct drm_device *dev,
+				 struct drm_crtc *crtc,
+				 struct drm_framebuffer *fb,
+				 struct drm_i915_gem_object *obj,
+				 struct intel_engine_cs *ring,
+				 uint32_t flags)
 {
-	struct drm_i915_private *dev_priv = to_i915(ring->dev);
-	struct intel_crtc *intel_crtc;
-	unsigned long irq_flags;
-	u32 seqno;
-
-	seqno = ring->get_seqno(ring, false);
-
-	spin_lock_irqsave(&dev_priv->mmio_flip_lock, irq_flags);
-	for_each_intel_crtc(ring->dev, intel_crtc) {
-		struct intel_mmio_flip *mmio_flip;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 
-		mmio_flip = &intel_crtc->mmio_flip;
-		if (mmio_flip->seqno == 0)
-			continue;
+	intel_crtc->mmio_flip.seqno = obj->last_write_seqno;
+	intel_crtc->mmio_flip.ring = obj->ring;
 
-		if (ring->id != mmio_flip->ring_id)
-			continue;
+	schedule_work(&intel_crtc->mmio_flip.work);
 
-		if (i915_seqno_passed(seqno, mmio_flip->seqno)) {
-			intel_do_mmio_flip(intel_crtc);
-			mmio_flip->seqno = 0;
-			ring->irq_put(ring);
-		}
-	}
-	spin_unlock_irqrestore(&dev_priv->mmio_flip_lock, irq_flags);
+	return 0;
 }
 
-static int intel_queue_mmio_flip(struct drm_device *dev,
+static int intel_gen9_queue_flip(struct drm_device *dev,
 				 struct drm_crtc *crtc,
 				 struct drm_framebuffer *fb,
 				 struct drm_i915_gem_object *obj,
 				 struct intel_engine_cs *ring,
 				 uint32_t flags)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	unsigned long irq_flags;
+	uint32_t plane = 0, stride;
 	int ret;
 
-	if (WARN_ON(intel_crtc->mmio_flip.seqno))
-		return -EBUSY;
+	switch(intel_crtc->pipe) {
+	case PIPE_A:
+		plane = MI_DISPLAY_FLIP_SKL_PLANE_1_A;
+		break;
+	case PIPE_B:
+		plane = MI_DISPLAY_FLIP_SKL_PLANE_1_B;
+		break;
+	case PIPE_C:
+		plane = MI_DISPLAY_FLIP_SKL_PLANE_1_C;
+		break;
+	default:
+		WARN_ONCE(1, "unknown plane in flip command\n");
+		return -ENODEV;
+	}
 
-	ret = intel_postpone_flip(obj);
-	if (ret < 0)
-		return ret;
-	if (ret == 0) {
-		intel_do_mmio_flip(intel_crtc);
-		return 0;
+	switch (obj->tiling_mode) {
+	case I915_TILING_NONE:
+		stride = fb->pitches[0] >> 6;
+		break;
+	case I915_TILING_X:
+		stride = fb->pitches[0] >> 9;
+		break;
+	default:
+		WARN_ONCE(1, "unknown tiling in flip command\n");
+		return -ENODEV;
 	}
 
-	spin_lock_irqsave(&dev_priv->mmio_flip_lock, irq_flags);
-	intel_crtc->mmio_flip.seqno = obj->last_write_seqno;
-	intel_crtc->mmio_flip.ring_id = obj->ring->id;
-	spin_unlock_irqrestore(&dev_priv->mmio_flip_lock, irq_flags);
+	ret = intel_ring_begin(ring, 10);
+	if (ret)
+		return ret;
+
+	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(1));
+	intel_ring_emit(ring, DERRMR);
+	intel_ring_emit(ring, ~(DERRMR_PIPEA_PRI_FLIP_DONE |
+				DERRMR_PIPEB_PRI_FLIP_DONE |
+				DERRMR_PIPEC_PRI_FLIP_DONE));
+	intel_ring_emit(ring, MI_STORE_REGISTER_MEM_GEN8(1) |
+			      MI_SRM_LRM_GLOBAL_GTT);
+	intel_ring_emit(ring, DERRMR);
+	intel_ring_emit(ring, ring->scratch.gtt_offset + 256);
+	intel_ring_emit(ring, 0);
+
+	intel_ring_emit(ring, MI_DISPLAY_FLIP_I915 | plane);
+	intel_ring_emit(ring, stride << 6 | obj->tiling_mode);
+	intel_ring_emit(ring, intel_crtc->unpin_work->gtt_offset);
+
+	intel_mark_page_flip_active(intel_crtc);
+	__intel_ring_advance(ring);
 
-	/*
-	 * Double check to catch cases where irq fired before
-	 * mmio flip data was ready
-	 */
-	intel_notify_mmio_flip(obj->ring);
 	return 0;
 }
 
@@ -9665,6 +9656,66 @@ static int intel_default_queue_flip(struct drm_device *dev,
 	return -ENODEV;
 }
 
+static bool __intel_pageflip_stall_check(struct drm_device *dev,
+					 struct drm_crtc *crtc)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	struct intel_unpin_work *work = intel_crtc->unpin_work;
+	u32 addr;
+
+	if (atomic_read(&work->pending) >= INTEL_FLIP_COMPLETE)
+		return true;
+
+	if (!work->enable_stall_check)
+		return false;
+
+	if (work->flip_ready_vblank == 0) {
+		if (work->flip_queued_ring &&
+		    !i915_seqno_passed(work->flip_queued_ring->get_seqno(work->flip_queued_ring, true),
+				       work->flip_queued_seqno))
+			return false;
+
+		work->flip_ready_vblank = drm_vblank_count(dev, intel_crtc->pipe);
+	}
+
+	if (drm_vblank_count(dev, intel_crtc->pipe) - work->flip_ready_vblank < 3)
+		return false;
+
+	/* Potential stall - if we see that the flip has happened,
+	 * assume a missed interrupt. */
+	if (INTEL_INFO(dev)->gen >= 4)
+		addr = I915_HI_DISPBASE(I915_READ(DSPSURF(intel_crtc->plane)));
+	else
+		addr = I915_READ(DSPADDR(intel_crtc->plane));
+
+	/* There is a potential issue here with a false positive after a flip
+	 * to the same address. We could address this by checking for a
+	 * non-incrementing frame counter.
+	 */
+	return addr == work->gtt_offset;
+}
+
+void intel_check_page_flip(struct drm_device *dev, int pipe)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pipe];
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+
+	WARN_ON(!in_irq());
+
+	if (crtc == NULL)
+		return;
+
+	spin_lock(&dev->event_lock);
+	if (intel_crtc->unpin_work && __intel_pageflip_stall_check(dev, crtc)) {
+		WARN_ONCE(1, "Kicking stuck page flip: queued at %d, now %d\n",
+			 intel_crtc->unpin_work->flip_queued_vblank, drm_vblank_count(dev, pipe));
+		page_flip_completed(intel_crtc);
+	}
+	spin_unlock(&dev->event_lock);
+}
+
 static int intel_crtc_page_flip(struct drm_crtc *crtc,
 				struct drm_framebuffer *fb,
 				struct drm_pending_vblank_event *event,
@@ -9678,7 +9729,6 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 	enum pipe pipe = intel_crtc->pipe;
 	struct intel_unpin_work *work;
 	struct intel_engine_cs *ring;
-	unsigned long flags;
 	int ret;
 
 	/*
@@ -9719,17 +9769,25 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 		goto free_work;
 
 	/* We borrow the event spin lock for protecting unpin_work */
-	spin_lock_irqsave(&dev->event_lock, flags);
+	spin_lock_irq(&dev->event_lock);
 	if (intel_crtc->unpin_work) {
-		spin_unlock_irqrestore(&dev->event_lock, flags);
-		kfree(work);
-		drm_crtc_vblank_put(crtc);
+		/* Before declaring the flip queue wedged, check if
+		 * the hardware completed the operation behind our backs.
+		 */
+		if (__intel_pageflip_stall_check(dev, crtc)) {
+			DRM_DEBUG_DRIVER("flip queue: previous flip completed, continuing\n");
+			page_flip_completed(intel_crtc);
+		} else {
+			DRM_DEBUG_DRIVER("flip queue: crtc already busy\n");
+			spin_unlock_irq(&dev->event_lock);
 
-		DRM_DEBUG_DRIVER("flip queue: crtc already busy\n");
-		return -EBUSY;
+			drm_crtc_vblank_put(crtc);
+			kfree(work);
+			return -EBUSY;
+		}
 	}
 	intel_crtc->unpin_work = work;
-	spin_unlock_irqrestore(&dev->event_lock, flags);
+	spin_unlock_irq(&dev->event_lock);
 
 	if (atomic_read(&intel_crtc->unpin_work_count) >= 2)
 		flush_workqueue(dev_priv->wq);
@@ -9746,8 +9804,6 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 
 	work->pending_flip_obj = obj;
 
-	work->enable_stall_check = true;
-
 	atomic_inc(&intel_crtc->unpin_work_count);
 	intel_crtc->reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter);
 
@@ -9759,7 +9815,7 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 		if (obj->tiling_mode != work->old_fb_obj->tiling_mode)
 			/* vlv: DISPLAY_FLIP fails to change tiling */
 			ring = NULL;
-	} else if (IS_IVYBRIDGE(dev)) {
+	} else if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev)) {
 		ring = &dev_priv->ring[BCS];
 	} else if (INTEL_INFO(dev)->gen >= 7) {
 		ring = obj->ring;
@@ -9769,21 +9825,33 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 		ring = &dev_priv->ring[RCS];
 	}
 
-	ret = intel_pin_and_fence_fb_obj(dev, obj, ring);
+	ret = intel_pin_and_fence_fb_obj(crtc->primary, fb, ring);
 	if (ret)
 		goto cleanup_pending;
 
 	work->gtt_offset =
 		i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset;
 
-	if (use_mmio_flip(ring, obj))
+	if (use_mmio_flip(ring, obj)) {
 		ret = intel_queue_mmio_flip(dev, crtc, fb, obj, ring,
 					    page_flip_flags);
-	else
+		if (ret)
+			goto cleanup_unpin;
+
+		work->flip_queued_seqno = obj->last_write_seqno;
+		work->flip_queued_ring = obj->ring;
+	} else {
 		ret = dev_priv->display.queue_flip(dev, crtc, fb, obj, ring,
-				page_flip_flags);
-	if (ret)
-		goto cleanup_unpin;
+						   page_flip_flags);
+		if (ret)
+			goto cleanup_unpin;
+
+		work->flip_queued_seqno = intel_ring_get_seqno(ring);
+		work->flip_queued_ring = ring;
+	}
+
+	work->flip_queued_vblank = drm_vblank_count(dev, intel_crtc->pipe);
+	work->enable_stall_check = true;
 
 	i915_gem_track_fb(work->old_fb_obj, obj,
 			  INTEL_FRONTBUFFER_PRIMARY(pipe));
@@ -9806,9 +9874,9 @@ cleanup_pending:
 	mutex_unlock(&dev->struct_mutex);
 
 cleanup:
-	spin_lock_irqsave(&dev->event_lock, flags);
+	spin_lock_irq(&dev->event_lock);
 	intel_crtc->unpin_work = NULL;
-	spin_unlock_irqrestore(&dev->event_lock, flags);
+	spin_unlock_irq(&dev->event_lock);
 
 	drm_crtc_vblank_put(crtc);
 free_work:
@@ -9818,8 +9886,11 @@ free_work:
 out_hang:
 		intel_crtc_wait_for_pending_flips(crtc);
 		ret = intel_pipe_set_base(crtc, crtc->x, crtc->y, fb);
-		if (ret == 0 && event)
+		if (ret == 0 && event) {
+			spin_lock_irq(&dev->event_lock);
 			drm_send_vblank_event(dev, pipe, event);
+			spin_unlock_irq(&dev->event_lock);
+		}
 	}
 	return ret;
 }
@@ -9847,8 +9918,7 @@ static void intel_modeset_update_staged_output_state(struct drm_device *dev)
 			to_intel_encoder(connector->base.encoder);
 	}
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		encoder->new_crtc =
 			to_intel_crtc(encoder->base.crtc);
 	}
@@ -9879,8 +9949,7 @@ static void intel_modeset_commit_output_state(struct drm_device *dev)
 		connector->base.encoder = &connector->new_encoder->base;
 	}
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		encoder->base.crtc = &encoder->new_crtc->base;
 	}
 
@@ -10007,6 +10076,19 @@ static void intel_dump_pipe_config(struct intel_crtc *crtc,
 		      pipe_config->dp_m_n.gmch_m, pipe_config->dp_m_n.gmch_n,
 		      pipe_config->dp_m_n.link_m, pipe_config->dp_m_n.link_n,
 		      pipe_config->dp_m_n.tu);
+
+	DRM_DEBUG_KMS("dp: %i, gmch_m2: %u, gmch_n2: %u, link_m2: %u, link_n2: %u, tu2: %u\n",
+		      pipe_config->has_dp_encoder,
+		      pipe_config->dp_m2_n2.gmch_m,
+		      pipe_config->dp_m2_n2.gmch_n,
+		      pipe_config->dp_m2_n2.link_m,
+		      pipe_config->dp_m2_n2.link_n,
+		      pipe_config->dp_m2_n2.tu);
+
+	DRM_DEBUG_KMS("audio: %i, infoframes: %i\n",
+		      pipe_config->has_audio,
+		      pipe_config->has_infoframe);
+
 	DRM_DEBUG_KMS("requested mode:\n");
 	drm_mode_debug_printmodeline(&pipe_config->requested_mode);
 	DRM_DEBUG_KMS("adjusted mode:\n");
@@ -10041,8 +10123,7 @@ static bool check_single_encoder_cloning(struct intel_crtc *crtc,
 	struct drm_device *dev = crtc->base.dev;
 	struct intel_encoder *source_encoder;
 
-	list_for_each_entry(source_encoder,
-			    &dev->mode_config.encoder_list, base.head) {
+	for_each_intel_encoder(dev, source_encoder) {
 		if (source_encoder->new_crtc != crtc)
 			continue;
 
@@ -10058,8 +10139,7 @@ static bool check_encoder_cloning(struct intel_crtc *crtc)
 	struct drm_device *dev = crtc->base.dev;
 	struct intel_encoder *encoder;
 
-	list_for_each_entry(encoder,
-			    &dev->mode_config.encoder_list, base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		if (encoder->new_crtc != crtc)
 			continue;
 
@@ -10070,6 +10150,48 @@ static bool check_encoder_cloning(struct intel_crtc *crtc)
 	return true;
 }
 
+static bool check_digital_port_conflicts(struct drm_device *dev)
+{
+	struct intel_connector *connector;
+	unsigned int used_ports = 0;
+
+	/*
+	 * Walk the connector list instead of the encoder
+	 * list to detect the problem on ddi platforms
+	 * where there's just one encoder per digital port.
+	 */
+	list_for_each_entry(connector,
+			    &dev->mode_config.connector_list, base.head) {
+		struct intel_encoder *encoder = connector->new_encoder;
+
+		if (!encoder)
+			continue;
+
+		WARN_ON(!encoder->new_crtc);
+
+		switch (encoder->type) {
+			unsigned int port_mask;
+		case INTEL_OUTPUT_UNKNOWN:
+			if (WARN_ON(!HAS_DDI(dev)))
+				break;
+		case INTEL_OUTPUT_DISPLAYPORT:
+		case INTEL_OUTPUT_HDMI:
+		case INTEL_OUTPUT_EDP:
+			port_mask = 1 << enc_to_dig_port(&encoder->base)->port;
+
+			/* the same port mustn't appear more than once */
+			if (used_ports & port_mask)
+				return false;
+
+			used_ports |= port_mask;
+		default:
+			break;
+		}
+	}
+
+	return true;
+}
+
 static struct intel_crtc_config *
 intel_modeset_pipe_config(struct drm_crtc *crtc,
 			  struct drm_framebuffer *fb,
@@ -10086,6 +10208,11 @@ intel_modeset_pipe_config(struct drm_crtc *crtc,
 		return ERR_PTR(-EINVAL);
 	}
 
+	if (!check_digital_port_conflicts(dev)) {
+		DRM_DEBUG_KMS("rejecting conflicting digital port configuration\n");
+		return ERR_PTR(-EINVAL);
+	}
+
 	pipe_config = kzalloc(sizeof(*pipe_config), GFP_KERNEL);
 	if (!pipe_config)
 		return ERR_PTR(-ENOMEM);
@@ -10143,8 +10270,7 @@ encoder_retry:
 	 * adjust it according to limitations or connector properties, and also
 	 * a chance to reject the mode entirely.
 	 */
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 
 		if (&encoder->new_crtc->base != crtc)
 			continue;
@@ -10222,8 +10348,7 @@ intel_modeset_affected_pipes(struct drm_crtc *crtc, unsigned *modeset_pipes,
 				1 << connector->new_encoder->new_crtc->pipe;
 	}
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		if (encoder->base.crtc == &encoder->new_crtc->base)
 			continue;
 
@@ -10293,12 +10418,14 @@ static bool intel_crtc_in_use(struct drm_crtc *crtc)
 static void
 intel_modeset_update_state(struct drm_device *dev, unsigned prepare_pipes)
 {
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_encoder *intel_encoder;
 	struct intel_crtc *intel_crtc;
 	struct drm_connector *connector;
 
-	list_for_each_entry(intel_encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	intel_shared_dpll_commit(dev_priv);
+
+	for_each_intel_encoder(dev, intel_encoder) {
 		if (!intel_encoder->base.crtc)
 			continue;
 
@@ -10387,6 +10514,22 @@ intel_pipe_config_compare(struct drm_device *dev,
 		return false; \
 	}
 
+/* This is required for BDW+ where there is only one set of registers for
+ * switching between high and low RR.
+ * This macro can be used whenever a comparison has to be made between one
+ * hw state and multiple sw state variables.
+ */
+#define PIPE_CONF_CHECK_I_ALT(name, alt_name) \
+	if ((current_config->name != pipe_config->name) && \
+		(current_config->alt_name != pipe_config->name)) { \
+			DRM_ERROR("mismatch in " #name " " \
+				  "(expected %i or %i, found %i)\n", \
+				  current_config->name, \
+				  current_config->alt_name, \
+				  pipe_config->name); \
+			return false; \
+	}
+
 #define PIPE_CONF_CHECK_FLAGS(name, mask)	\
 	if ((current_config->name ^ pipe_config->name) & (mask)) { \
 		DRM_ERROR("mismatch in " #name "(" #mask ") "	   \
@@ -10419,11 +10562,28 @@ intel_pipe_config_compare(struct drm_device *dev,
 	PIPE_CONF_CHECK_I(fdi_m_n.tu);
 
 	PIPE_CONF_CHECK_I(has_dp_encoder);
-	PIPE_CONF_CHECK_I(dp_m_n.gmch_m);
-	PIPE_CONF_CHECK_I(dp_m_n.gmch_n);
-	PIPE_CONF_CHECK_I(dp_m_n.link_m);
-	PIPE_CONF_CHECK_I(dp_m_n.link_n);
-	PIPE_CONF_CHECK_I(dp_m_n.tu);
+
+	if (INTEL_INFO(dev)->gen < 8) {
+		PIPE_CONF_CHECK_I(dp_m_n.gmch_m);
+		PIPE_CONF_CHECK_I(dp_m_n.gmch_n);
+		PIPE_CONF_CHECK_I(dp_m_n.link_m);
+		PIPE_CONF_CHECK_I(dp_m_n.link_n);
+		PIPE_CONF_CHECK_I(dp_m_n.tu);
+
+		if (current_config->has_drrs) {
+			PIPE_CONF_CHECK_I(dp_m2_n2.gmch_m);
+			PIPE_CONF_CHECK_I(dp_m2_n2.gmch_n);
+			PIPE_CONF_CHECK_I(dp_m2_n2.link_m);
+			PIPE_CONF_CHECK_I(dp_m2_n2.link_n);
+			PIPE_CONF_CHECK_I(dp_m2_n2.tu);
+		}
+	} else {
+		PIPE_CONF_CHECK_I_ALT(dp_m_n.gmch_m, dp_m2_n2.gmch_m);
+		PIPE_CONF_CHECK_I_ALT(dp_m_n.gmch_n, dp_m2_n2.gmch_n);
+		PIPE_CONF_CHECK_I_ALT(dp_m_n.link_m, dp_m2_n2.link_m);
+		PIPE_CONF_CHECK_I_ALT(dp_m_n.link_n, dp_m2_n2.link_n);
+		PIPE_CONF_CHECK_I_ALT(dp_m_n.tu, dp_m2_n2.tu);
+	}
 
 	PIPE_CONF_CHECK_I(adjusted_mode.crtc_hdisplay);
 	PIPE_CONF_CHECK_I(adjusted_mode.crtc_htotal);
@@ -10444,6 +10604,7 @@ intel_pipe_config_compare(struct drm_device *dev,
 	if ((INTEL_INFO(dev)->gen < 8 && !IS_HASWELL(dev)) ||
 	    IS_VALLEYVIEW(dev))
 		PIPE_CONF_CHECK_I(limited_color_range);
+	PIPE_CONF_CHECK_I(has_infoframe);
 
 	PIPE_CONF_CHECK_I(has_audio);
 
@@ -10500,6 +10661,9 @@ intel_pipe_config_compare(struct drm_device *dev,
 	PIPE_CONF_CHECK_X(dpll_hw_state.fp0);
 	PIPE_CONF_CHECK_X(dpll_hw_state.fp1);
 	PIPE_CONF_CHECK_X(dpll_hw_state.wrpll);
+	PIPE_CONF_CHECK_X(dpll_hw_state.ctrl1);
+	PIPE_CONF_CHECK_X(dpll_hw_state.cfgcr1);
+	PIPE_CONF_CHECK_X(dpll_hw_state.cfgcr2);
 
 	if (IS_G4X(dev) || INTEL_INFO(dev)->gen >= 5)
 		PIPE_CONF_CHECK_I(pipe_bpp);
@@ -10509,6 +10673,7 @@ intel_pipe_config_compare(struct drm_device *dev,
 
 #undef PIPE_CONF_CHECK_X
 #undef PIPE_CONF_CHECK_I
+#undef PIPE_CONF_CHECK_I_ALT
 #undef PIPE_CONF_CHECK_FLAGS
 #undef PIPE_CONF_CHECK_CLOCK_FUZZY
 #undef PIPE_CONF_QUIRK
@@ -10516,6 +10681,56 @@ intel_pipe_config_compare(struct drm_device *dev,
 	return true;
 }
 
+static void check_wm_state(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct skl_ddb_allocation hw_ddb, *sw_ddb;
+	struct intel_crtc *intel_crtc;
+	int plane;
+
+	if (INTEL_INFO(dev)->gen < 9)
+		return;
+
+	skl_ddb_get_hw_state(dev_priv, &hw_ddb);
+	sw_ddb = &dev_priv->wm.skl_hw.ddb;
+
+	for_each_intel_crtc(dev, intel_crtc) {
+		struct skl_ddb_entry *hw_entry, *sw_entry;
+		const enum pipe pipe = intel_crtc->pipe;
+
+		if (!intel_crtc->active)
+			continue;
+
+		/* planes */
+		for_each_plane(pipe, plane) {
+			hw_entry = &hw_ddb.plane[pipe][plane];
+			sw_entry = &sw_ddb->plane[pipe][plane];
+
+			if (skl_ddb_entry_equal(hw_entry, sw_entry))
+				continue;
+
+			DRM_ERROR("mismatch in DDB state pipe %c plane %d "
+				  "(expected (%u,%u), found (%u,%u))\n",
+				  pipe_name(pipe), plane + 1,
+				  sw_entry->start, sw_entry->end,
+				  hw_entry->start, hw_entry->end);
+		}
+
+		/* cursor */
+		hw_entry = &hw_ddb.cursor[pipe];
+		sw_entry = &sw_ddb->cursor[pipe];
+
+		if (skl_ddb_entry_equal(hw_entry, sw_entry))
+			continue;
+
+		DRM_ERROR("mismatch in DDB state pipe %c cursor "
+			  "(expected (%u,%u), found (%u,%u))\n",
+			  pipe_name(pipe),
+			  sw_entry->start, sw_entry->end,
+			  hw_entry->start, hw_entry->end);
+	}
+}
+
 static void
 check_connector_state(struct drm_device *dev)
 {
@@ -10538,8 +10753,7 @@ check_encoder_state(struct drm_device *dev)
 	struct intel_encoder *encoder;
 	struct intel_connector *connector;
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		bool enabled = false;
 		bool active = false;
 		enum pipe pipe, tracked_pipe;
@@ -10618,8 +10832,7 @@ check_crtc_state(struct drm_device *dev)
 		WARN(crtc->active && !crtc->base.enabled,
 		     "active crtc, but not enabled in sw tracking\n");
 
-		list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-				    base.head) {
+		for_each_intel_encoder(dev, encoder) {
 			if (encoder->base.crtc != &crtc->base)
 				continue;
 			enabled = true;
@@ -10637,12 +10850,12 @@ check_crtc_state(struct drm_device *dev)
 		active = dev_priv->display.get_pipe_config(crtc,
 							   &pipe_config);
 
-		/* hw state is inconsistent with the pipe A quirk */
-		if (crtc->pipe == PIPE_A && dev_priv->quirks & QUIRK_PIPEA_FORCE)
+		/* hw state is inconsistent with the pipe quirk */
+		if ((crtc->pipe == PIPE_A && dev_priv->quirks & QUIRK_PIPEA_FORCE) ||
+		    (crtc->pipe == PIPE_B && dev_priv->quirks & QUIRK_PIPEB_FORCE))
 			active = crtc->active;
 
-		list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-				    base.head) {
+		for_each_intel_encoder(dev, encoder) {
 			enum pipe pipe;
 			if (encoder->base.crtc != &crtc->base)
 				continue;
@@ -10684,9 +10897,9 @@ check_shared_dpll_state(struct drm_device *dev)
 
 		active = pll->get_hw_state(dev_priv, pll, &dpll_hw_state);
 
-		WARN(pll->active > pll->refcount,
+		WARN(pll->active > hweight32(pll->config.crtc_mask),
 		     "more active pll users than references: %i vs %i\n",
-		     pll->active, pll->refcount);
+		     pll->active, hweight32(pll->config.crtc_mask));
 		WARN(pll->active && !pll->on,
 		     "pll in active use but not on in sw tracking\n");
 		WARN(pll->on && !pll->active,
@@ -10704,11 +10917,11 @@ check_shared_dpll_state(struct drm_device *dev)
 		WARN(pll->active != active_crtcs,
 		     "pll active crtcs mismatch (expected %i, found %i)\n",
 		     pll->active, active_crtcs);
-		WARN(pll->refcount != enabled_crtcs,
+		WARN(hweight32(pll->config.crtc_mask) != enabled_crtcs,
 		     "pll enabled crtcs mismatch (expected %i, found %i)\n",
-		     pll->refcount, enabled_crtcs);
+		     hweight32(pll->config.crtc_mask), enabled_crtcs);
 
-		WARN(pll->on && memcmp(&pll->hw_state, &dpll_hw_state,
+		WARN(pll->on && memcmp(&pll->config.hw_state, &dpll_hw_state,
 				       sizeof(dpll_hw_state)),
 		     "pll hw state mismatch\n");
 	}
@@ -10717,6 +10930,7 @@ check_shared_dpll_state(struct drm_device *dev)
 void
 intel_modeset_check_state(struct drm_device *dev)
 {
+	check_wm_state(dev);
 	check_connector_state(dev);
 	check_encoder_state(dev);
 	check_crtc_state(dev);
@@ -10767,50 +10981,67 @@ static void update_scanline_offset(struct intel_crtc *crtc)
 
 		crtc->scanline_offset = vtotal - 1;
 	} else if (HAS_DDI(dev) &&
-		   intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_HDMI)) {
+		   intel_pipe_has_type(crtc, INTEL_OUTPUT_HDMI)) {
 		crtc->scanline_offset = 2;
 	} else
 		crtc->scanline_offset = 1;
 }
 
+static struct intel_crtc_config *
+intel_modeset_compute_config(struct drm_crtc *crtc,
+			     struct drm_display_mode *mode,
+			     struct drm_framebuffer *fb,
+			     unsigned *modeset_pipes,
+			     unsigned *prepare_pipes,
+			     unsigned *disable_pipes)
+{
+	struct intel_crtc_config *pipe_config = NULL;
+
+	intel_modeset_affected_pipes(crtc, modeset_pipes,
+				     prepare_pipes, disable_pipes);
+
+	if ((*modeset_pipes) == 0)
+		goto out;
+
+	/*
+	 * Note this needs changes when we start tracking multiple modes
+	 * and crtcs.  At that point we'll need to compute the whole config
+	 * (i.e. one pipe_config for each crtc) rather than just the one
+	 * for this crtc.
+	 */
+	pipe_config = intel_modeset_pipe_config(crtc, fb, mode);
+	if (IS_ERR(pipe_config)) {
+		goto out;
+	}
+	intel_dump_pipe_config(to_intel_crtc(crtc), pipe_config,
+			       "[modeset]");
+
+out:
+	return pipe_config;
+}
+
 static int __intel_set_mode(struct drm_crtc *crtc,
 			    struct drm_display_mode *mode,
-			    int x, int y, struct drm_framebuffer *fb)
+			    int x, int y, struct drm_framebuffer *fb,
+			    struct intel_crtc_config *pipe_config,
+			    unsigned modeset_pipes,
+			    unsigned prepare_pipes,
+			    unsigned disable_pipes)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_display_mode *saved_mode;
-	struct intel_crtc_config *pipe_config = NULL;
 	struct intel_crtc *intel_crtc;
-	unsigned disable_pipes, prepare_pipes, modeset_pipes;
 	int ret = 0;
 
 	saved_mode = kmalloc(sizeof(*saved_mode), GFP_KERNEL);
 	if (!saved_mode)
 		return -ENOMEM;
 
-	intel_modeset_affected_pipes(crtc, &modeset_pipes,
-				     &prepare_pipes, &disable_pipes);
-
 	*saved_mode = crtc->mode;
 
-	/* Hack: Because we don't (yet) support global modeset on multiple
-	 * crtcs, we don't keep track of the new mode for more than one crtc.
-	 * Hence simply check whether any bit is set in modeset_pipes in all the
-	 * pieces of code that are not yet converted to deal with mutliple crtcs
-	 * changing their mode at the same time. */
-	if (modeset_pipes) {
-		pipe_config = intel_modeset_pipe_config(crtc, fb, mode);
-		if (IS_ERR(pipe_config)) {
-			ret = PTR_ERR(pipe_config);
-			pipe_config = NULL;
-
-			goto out;
-		}
-		intel_dump_pipe_config(to_intel_crtc(crtc), pipe_config,
-				       "[modeset]");
+	if (modeset_pipes)
 		to_intel_crtc(crtc)->new_config = pipe_config;
-	}
 
 	/*
 	 * See if the config requires any additional preparation, e.g.
@@ -10826,6 +11057,22 @@ static int __intel_set_mode(struct drm_crtc *crtc,
 		prepare_pipes &= ~disable_pipes;
 	}
 
+	if (dev_priv->display.crtc_compute_clock) {
+		unsigned clear_pipes = modeset_pipes | disable_pipes;
+
+		ret = intel_shared_dpll_start_config(dev_priv, clear_pipes);
+		if (ret)
+			goto done;
+
+		for_each_intel_crtc_masked(dev, modeset_pipes, intel_crtc) {
+			ret = dev_priv->display.crtc_compute_clock(intel_crtc);
+			if (ret) {
+				intel_shared_dpll_abort_config(dev_priv);
+				goto done;
+			}
+		}
+	}
+
 	for_each_intel_crtc_masked(dev, disable_pipes, intel_crtc)
 		intel_crtc_disable(&intel_crtc->base);
 
@@ -10836,6 +11083,10 @@ static int __intel_set_mode(struct drm_crtc *crtc,
 
 	/* crtc->mode is already used by the ->mode_set callbacks, hence we need
 	 * to set it here already despite that we pass it down the callchain.
+	 *
+	 * Note we'll need to fix this up when we start tracking multiple
+	 * pipes; here we assume a single modeset_pipe and only track the
+	 * single crtc and mode.
 	 */
 	if (modeset_pipes) {
 		crtc->mode = *mode;
@@ -10857,8 +11108,7 @@ static int __intel_set_mode(struct drm_crtc *crtc,
 	 * update the the output configuration. */
 	intel_modeset_update_state(dev, prepare_pipes);
 
-	if (dev_priv->display.modeset_global_resources)
-		dev_priv->display.modeset_global_resources(dev);
+	modeset_update_crtc_power_domains(dev);
 
 	/* Set up the DPLL and any encoders state that needs to adjust or depend
 	 * on the DPLL.
@@ -10869,9 +11119,7 @@ static int __intel_set_mode(struct drm_crtc *crtc,
 		struct drm_i915_gem_object *obj = intel_fb_obj(fb);
 
 		mutex_lock(&dev->struct_mutex);
-		ret = intel_pin_and_fence_fb_obj(dev,
-						 obj,
-						 NULL);
+		ret = intel_pin_and_fence_fb_obj(crtc->primary, fb, NULL);
 		if (ret != 0) {
 			DRM_ERROR("pin & fence failed\n");
 			mutex_unlock(&dev->struct_mutex);
@@ -10886,11 +11134,6 @@ static int __intel_set_mode(struct drm_crtc *crtc,
 		crtc->primary->fb = fb;
 		crtc->x = x;
 		crtc->y = y;
-
-		ret = dev_priv->display.crtc_mode_set(&intel_crtc->base,
-						      x, y, fb);
-		if (ret)
-			goto done;
 	}
 
 	/* Now enable the clocks, plane, pipe, and connectors that we set up. */
@@ -10905,19 +11148,23 @@ done:
 	if (ret && crtc->enabled)
 		crtc->mode = *saved_mode;
 
-out:
 	kfree(pipe_config);
 	kfree(saved_mode);
 	return ret;
 }
 
-static int intel_set_mode(struct drm_crtc *crtc,
-			  struct drm_display_mode *mode,
-			  int x, int y, struct drm_framebuffer *fb)
+static int intel_set_mode_pipes(struct drm_crtc *crtc,
+				struct drm_display_mode *mode,
+				int x, int y, struct drm_framebuffer *fb,
+				struct intel_crtc_config *pipe_config,
+				unsigned modeset_pipes,
+				unsigned prepare_pipes,
+				unsigned disable_pipes)
 {
 	int ret;
 
-	ret = __intel_set_mode(crtc, mode, x, y, fb);
+	ret = __intel_set_mode(crtc, mode, x, y, fb, pipe_config, modeset_pipes,
+			       prepare_pipes, disable_pipes);
 
 	if (ret == 0)
 		intel_modeset_check_state(crtc->dev);
@@ -10925,6 +11172,26 @@ static int intel_set_mode(struct drm_crtc *crtc,
 	return ret;
 }
 
+static int intel_set_mode(struct drm_crtc *crtc,
+			  struct drm_display_mode *mode,
+			  int x, int y, struct drm_framebuffer *fb)
+{
+	struct intel_crtc_config *pipe_config;
+	unsigned modeset_pipes, prepare_pipes, disable_pipes;
+
+	pipe_config = intel_modeset_compute_config(crtc, mode, fb,
+						   &modeset_pipes,
+						   &prepare_pipes,
+						   &disable_pipes);
+
+	if (IS_ERR(pipe_config))
+		return PTR_ERR(pipe_config);
+
+	return intel_set_mode_pipes(crtc, mode, x, y, fb, pipe_config,
+				    modeset_pipes, prepare_pipes,
+				    disable_pipes);
+}
+
 void intel_crtc_restore_mode(struct drm_crtc *crtc)
 {
 	intel_set_mode(crtc, &crtc->mode, crtc->x, crtc->y, crtc->primary->fb);
@@ -11010,7 +11277,7 @@ static void intel_set_config_restore_state(struct drm_device *dev,
 	}
 
 	count = 0;
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		encoder->new_crtc =
 			to_intel_crtc(config->save_encoder_crtcs[count++]);
 	}
@@ -11169,8 +11436,7 @@ intel_modeset_stage_output_state(struct drm_device *dev,
 	}
 
 	/* Check for any encoders that needs to be disabled. */
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		int num_connectors = 0;
 		list_for_each_entry(connector,
 				    &dev->mode_config.connector_list,
@@ -11203,9 +11469,7 @@ intel_modeset_stage_output_state(struct drm_device *dev,
 	for_each_intel_crtc(dev, crtc) {
 		crtc->new_enabled = false;
 
-		list_for_each_entry(encoder,
-				    &dev->mode_config.encoder_list,
-				    base.head) {
+		for_each_intel_encoder(dev, encoder) {
 			if (encoder->new_crtc == crtc) {
 				crtc->new_enabled = true;
 				break;
@@ -11242,7 +11506,7 @@ static void disable_crtc_nofb(struct intel_crtc *crtc)
 			connector->new_encoder = NULL;
 	}
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		if (encoder->new_crtc == crtc)
 			encoder->new_crtc = NULL;
 	}
@@ -11256,6 +11520,8 @@ static int intel_crtc_set_config(struct drm_mode_set *set)
 	struct drm_device *dev;
 	struct drm_mode_set save_set;
 	struct intel_set_config *config;
+	struct intel_crtc_config *pipe_config;
+	unsigned modeset_pipes, prepare_pipes, disable_pipes;
 	int ret;
 
 	BUG_ON(!set);
@@ -11301,11 +11567,39 @@ static int intel_crtc_set_config(struct drm_mode_set *set)
 	if (ret)
 		goto fail;
 
+	pipe_config = intel_modeset_compute_config(set->crtc, set->mode,
+						   set->fb,
+						   &modeset_pipes,
+						   &prepare_pipes,
+						   &disable_pipes);
+	if (IS_ERR(pipe_config)) {
+		ret = PTR_ERR(pipe_config);
+		goto fail;
+	} else if (pipe_config) {
+		if (pipe_config->has_audio !=
+		    to_intel_crtc(set->crtc)->config.has_audio)
+			config->mode_changed = true;
+
+		/*
+		 * Note we have an issue here with infoframes: current code
+		 * only updates them on the full mode set path per hw
+		 * requirements.  So here we should be checking for any
+		 * required changes and forcing a mode set.
+		 */
+	}
+
+	/* set_mode will free it in the mode_changed case */
+	if (!config->mode_changed)
+		kfree(pipe_config);
+
+	intel_update_pipe_size(to_intel_crtc(set->crtc));
+
 	if (config->mode_changed) {
-		ret = intel_set_mode(set->crtc, set->mode,
-				     set->x, set->y, set->fb);
+		ret = intel_set_mode_pipes(set->crtc, set->mode,
+					   set->x, set->y, set->fb, pipe_config,
+					   modeset_pipes, prepare_pipes,
+					   disable_pipes);
 	} else if (config->fb_changed) {
-		struct drm_i915_private *dev_priv = dev->dev_private;
 		struct intel_crtc *intel_crtc = to_intel_crtc(set->crtc);
 
 		intel_crtc_wait_for_pending_flips(set->crtc);
@@ -11319,8 +11613,7 @@ static int intel_crtc_set_config(struct drm_mode_set *set)
 		 */
 		if (!intel_crtc->primary_enabled && ret == 0) {
 			WARN_ON(!intel_crtc->active);
-			intel_enable_primary_hw_plane(dev_priv, intel_crtc->plane,
-						      intel_crtc->pipe);
+			intel_enable_primary_hw_plane(set->crtc->primary, set->crtc);
 		}
 
 		/*
@@ -11375,7 +11668,7 @@ static bool ibx_pch_dpll_get_hw_state(struct drm_i915_private *dev_priv,
 {
 	uint32_t val;
 
-	if (!intel_display_power_enabled(dev_priv, POWER_DOMAIN_PLLS))
+	if (!intel_display_power_is_enabled(dev_priv, POWER_DOMAIN_PLLS))
 		return false;
 
 	val = I915_READ(PCH_DPLL(pll->id));
@@ -11389,8 +11682,8 @@ static bool ibx_pch_dpll_get_hw_state(struct drm_i915_private *dev_priv,
 static void ibx_pch_dpll_mode_set(struct drm_i915_private *dev_priv,
 				  struct intel_shared_dpll *pll)
 {
-	I915_WRITE(PCH_FP0(pll->id), pll->hw_state.fp0);
-	I915_WRITE(PCH_FP1(pll->id), pll->hw_state.fp1);
+	I915_WRITE(PCH_FP0(pll->id), pll->config.hw_state.fp0);
+	I915_WRITE(PCH_FP1(pll->id), pll->config.hw_state.fp1);
 }
 
 static void ibx_pch_dpll_enable(struct drm_i915_private *dev_priv,
@@ -11399,7 +11692,7 @@ static void ibx_pch_dpll_enable(struct drm_i915_private *dev_priv,
 	/* PCH refclock must be enabled first */
 	ibx_assert_pch_refclk_enabled(dev_priv);
 
-	I915_WRITE(PCH_DPLL(pll->id), pll->hw_state.dpll);
+	I915_WRITE(PCH_DPLL(pll->id), pll->config.hw_state.dpll);
 
 	/* Wait for the clocks to stabilize. */
 	POSTING_READ(PCH_DPLL(pll->id));
@@ -11410,7 +11703,7 @@ static void ibx_pch_dpll_enable(struct drm_i915_private *dev_priv,
 	 *
 	 * So write it again.
 	 */
-	I915_WRITE(PCH_DPLL(pll->id), pll->hw_state.dpll);
+	I915_WRITE(PCH_DPLL(pll->id), pll->config.hw_state.dpll);
 	POSTING_READ(PCH_DPLL(pll->id));
 	udelay(200);
 }
@@ -11473,8 +11766,6 @@ static int
 intel_primary_plane_disable(struct drm_plane *plane)
 {
 	struct drm_device *dev = plane->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_plane *intel_plane = to_intel_plane(plane);
 	struct intel_crtc *intel_crtc;
 
 	if (!plane->fb)
@@ -11497,8 +11788,8 @@ intel_primary_plane_disable(struct drm_plane *plane)
 		goto disable_unpin;
 
 	intel_crtc_wait_for_pending_flips(plane->crtc);
-	intel_disable_primary_hw_plane(dev_priv, intel_plane->plane,
-				       intel_plane->pipe);
+	intel_disable_primary_hw_plane(plane, plane->crtc);
+
 disable_unpin:
 	mutex_lock(&dev->struct_mutex);
 	i915_gem_track_fb(intel_fb_obj(plane->fb), NULL,
@@ -11511,123 +11802,195 @@ disable_unpin:
 }
 
 static int
-intel_primary_plane_setplane(struct drm_plane *plane, struct drm_crtc *crtc,
-			     struct drm_framebuffer *fb, int crtc_x, int crtc_y,
-			     unsigned int crtc_w, unsigned int crtc_h,
-			     uint32_t src_x, uint32_t src_y,
-			     uint32_t src_w, uint32_t src_h)
+intel_check_primary_plane(struct drm_plane *plane,
+			  struct intel_plane_state *state)
+{
+	struct drm_crtc *crtc = state->crtc;
+	struct drm_framebuffer *fb = state->fb;
+	struct drm_rect *dest = &state->dst;
+	struct drm_rect *src = &state->src;
+	const struct drm_rect *clip = &state->clip;
+
+	return drm_plane_helper_check_update(plane, crtc, fb,
+					     src, dest, clip,
+					     DRM_PLANE_HELPER_NO_SCALING,
+					     DRM_PLANE_HELPER_NO_SCALING,
+					     false, true, &state->visible);
+}
+
+static int
+intel_prepare_primary_plane(struct drm_plane *plane,
+			    struct intel_plane_state *state)
 {
+	struct drm_crtc *crtc = state->crtc;
+	struct drm_framebuffer *fb = state->fb;
 	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct intel_plane *intel_plane = to_intel_plane(plane);
+	enum pipe pipe = intel_crtc->pipe;
 	struct drm_i915_gem_object *obj = intel_fb_obj(fb);
 	struct drm_i915_gem_object *old_obj = intel_fb_obj(plane->fb);
-	struct drm_rect dest = {
-		/* integer pixels */
-		.x1 = crtc_x,
-		.y1 = crtc_y,
-		.x2 = crtc_x + crtc_w,
-		.y2 = crtc_y + crtc_h,
-	};
-	struct drm_rect src = {
-		/* 16.16 fixed point */
-		.x1 = src_x,
-		.y1 = src_y,
-		.x2 = src_x + src_w,
-		.y2 = src_y + src_h,
-	};
-	const struct drm_rect clip = {
-		/* integer pixels */
-		.x2 = intel_crtc->active ? intel_crtc->config.pipe_src_w : 0,
-		.y2 = intel_crtc->active ? intel_crtc->config.pipe_src_h : 0,
-	};
-	bool visible;
 	int ret;
 
-	ret = drm_plane_helper_check_update(plane, crtc, fb,
-					    &src, &dest, &clip,
-					    DRM_PLANE_HELPER_NO_SCALING,
-					    DRM_PLANE_HELPER_NO_SCALING,
-					    false, true, &visible);
+	intel_crtc_wait_for_pending_flips(crtc);
 
-	if (ret)
-		return ret;
+	if (intel_crtc_has_pending_flip(crtc)) {
+		DRM_ERROR("pipe is still busy with an old pageflip\n");
+		return -EBUSY;
+	}
 
-	/*
-	 * If the CRTC isn't enabled, we're just pinning the framebuffer,
-	 * updating the fb pointer, and returning without touching the
-	 * hardware.  This allows us to later do a drmModeSetCrtc with fb=-1 to
-	 * turn on the display with all planes setup as desired.
-	 */
-	if (!crtc->enabled) {
+	if (old_obj != obj) {
 		mutex_lock(&dev->struct_mutex);
+		ret = intel_pin_and_fence_fb_obj(plane, fb, NULL);
+		if (ret == 0)
+			i915_gem_track_fb(old_obj, obj,
+					  INTEL_FRONTBUFFER_PRIMARY(pipe));
+		mutex_unlock(&dev->struct_mutex);
+		if (ret != 0) {
+			DRM_DEBUG_KMS("pin & fence failed\n");
+			return ret;
+		}
+	}
+
+	return 0;
+}
+
+static void
+intel_commit_primary_plane(struct drm_plane *plane,
+			   struct intel_plane_state *state)
+{
+	struct drm_crtc *crtc = state->crtc;
+	struct drm_framebuffer *fb = state->fb;
+	struct drm_device *dev = crtc->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	enum pipe pipe = intel_crtc->pipe;
+	struct drm_framebuffer *old_fb = plane->fb;
+	struct drm_i915_gem_object *obj = intel_fb_obj(fb);
+	struct drm_i915_gem_object *old_obj = intel_fb_obj(plane->fb);
+	struct intel_plane *intel_plane = to_intel_plane(plane);
+	struct drm_rect *src = &state->src;
+
+	crtc->primary->fb = fb;
+	crtc->x = src->x1 >> 16;
+	crtc->y = src->y1 >> 16;
+
+	intel_plane->crtc_x = state->orig_dst.x1;
+	intel_plane->crtc_y = state->orig_dst.y1;
+	intel_plane->crtc_w = drm_rect_width(&state->orig_dst);
+	intel_plane->crtc_h = drm_rect_height(&state->orig_dst);
+	intel_plane->src_x = state->orig_src.x1;
+	intel_plane->src_y = state->orig_src.y1;
+	intel_plane->src_w = drm_rect_width(&state->orig_src);
+	intel_plane->src_h = drm_rect_height(&state->orig_src);
+	intel_plane->obj = obj;
 
+	if (intel_crtc->active) {
 		/*
-		 * If we already called setplane while the crtc was disabled,
-		 * we may have an fb pinned; unpin it.
+		 * FBC does not work on some platforms for rotated
+		 * planes, so disable it when rotation is not 0 and
+		 * update it when rotation is set back to 0.
+		 *
+		 * FIXME: This is redundant with the fbc update done in
+		 * the primary plane enable function except that that
+		 * one is done too late. We eventually need to unify
+		 * this.
 		 */
-		if (plane->fb)
-			intel_unpin_fb_obj(old_obj);
+		if (intel_crtc->primary_enabled &&
+		    INTEL_INFO(dev)->gen <= 4 && !IS_G4X(dev) &&
+		    dev_priv->fbc.plane == intel_crtc->plane &&
+		    intel_plane->rotation != BIT(DRM_ROTATE_0)) {
+			intel_disable_fbc(dev);
+		}
 
-		i915_gem_track_fb(old_obj, obj,
-				  INTEL_FRONTBUFFER_PRIMARY(intel_crtc->pipe));
+		if (state->visible) {
+			bool was_enabled = intel_crtc->primary_enabled;
 
-		/* Pin and return without programming hardware */
-		ret = intel_pin_and_fence_fb_obj(dev, obj, NULL);
-		mutex_unlock(&dev->struct_mutex);
+			/* FIXME: kill this fastboot hack */
+			intel_update_pipe_size(intel_crtc);
 
-		return ret;
+			intel_crtc->primary_enabled = true;
+
+			dev_priv->display.update_primary_plane(crtc, plane->fb,
+					crtc->x, crtc->y);
+
+			/*
+			 * BDW signals flip done immediately if the plane
+			 * is disabled, even if the plane enable is already
+			 * armed to occur at the next vblank :(
+			 */
+			if (IS_BROADWELL(dev) && !was_enabled)
+				intel_wait_for_vblank(dev, intel_crtc->pipe);
+		} else {
+			/*
+			 * If clipping results in a non-visible primary plane,
+			 * we'll disable the primary plane.  Note that this is
+			 * a bit different than what happens if userspace
+			 * explicitly disables the plane by passing fb=0
+			 * because plane->fb still gets set and pinned.
+			 */
+			intel_disable_primary_hw_plane(plane, crtc);
+		}
+
+		intel_frontbuffer_flip(dev, INTEL_FRONTBUFFER_PRIMARY(pipe));
+
+		mutex_lock(&dev->struct_mutex);
+		intel_update_fbc(dev);
+		mutex_unlock(&dev->struct_mutex);
 	}
 
-	intel_crtc_wait_for_pending_flips(crtc);
+	if (old_fb && old_fb != fb) {
+		if (intel_crtc->active)
+			intel_wait_for_vblank(dev, intel_crtc->pipe);
 
-	/*
-	 * If clipping results in a non-visible primary plane, we'll disable
-	 * the primary plane.  Note that this is a bit different than what
-	 * happens if userspace explicitly disables the plane by passing fb=0
-	 * because plane->fb still gets set and pinned.
-	 */
-	if (!visible) {
 		mutex_lock(&dev->struct_mutex);
+		intel_unpin_fb_obj(old_obj);
+		mutex_unlock(&dev->struct_mutex);
+	}
+}
 
-		/*
-		 * Try to pin the new fb first so that we can bail out if we
-		 * fail.
-		 */
-		if (plane->fb != fb) {
-			ret = intel_pin_and_fence_fb_obj(dev, obj, NULL);
-			if (ret) {
-				mutex_unlock(&dev->struct_mutex);
-				return ret;
-			}
-		}
+static int
+intel_primary_plane_setplane(struct drm_plane *plane, struct drm_crtc *crtc,
+			     struct drm_framebuffer *fb, int crtc_x, int crtc_y,
+			     unsigned int crtc_w, unsigned int crtc_h,
+			     uint32_t src_x, uint32_t src_y,
+			     uint32_t src_w, uint32_t src_h)
+{
+	struct intel_plane_state state;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	int ret;
 
-		i915_gem_track_fb(old_obj, obj,
-				  INTEL_FRONTBUFFER_PRIMARY(intel_crtc->pipe));
+	state.crtc = crtc;
+	state.fb = fb;
 
-		if (intel_crtc->primary_enabled)
-			intel_disable_primary_hw_plane(dev_priv,
-						       intel_plane->plane,
-						       intel_plane->pipe);
+	/* sample coordinates in 16.16 fixed point */
+	state.src.x1 = src_x;
+	state.src.x2 = src_x + src_w;
+	state.src.y1 = src_y;
+	state.src.y2 = src_y + src_h;
 
+	/* integer pixels */
+	state.dst.x1 = crtc_x;
+	state.dst.x2 = crtc_x + crtc_w;
+	state.dst.y1 = crtc_y;
+	state.dst.y2 = crtc_y + crtc_h;
 
-		if (plane->fb != fb)
-			if (plane->fb)
-				intel_unpin_fb_obj(old_obj);
+	state.clip.x1 = 0;
+	state.clip.y1 = 0;
+	state.clip.x2 = intel_crtc->active ? intel_crtc->config.pipe_src_w : 0;
+	state.clip.y2 = intel_crtc->active ? intel_crtc->config.pipe_src_h : 0;
 
-		mutex_unlock(&dev->struct_mutex);
+	state.orig_src = state.src;
+	state.orig_dst = state.dst;
 
-		return 0;
-	}
+	ret = intel_check_primary_plane(plane, &state);
+	if (ret)
+		return ret;
 
-	ret = intel_pipe_set_base(crtc, src.x1, src.y1, fb);
+	ret = intel_prepare_primary_plane(plane, &state);
 	if (ret)
 		return ret;
 
-	if (!intel_crtc->primary_enabled)
-		intel_enable_primary_hw_plane(dev_priv, intel_crtc->plane,
-					      intel_crtc->pipe);
+	intel_commit_primary_plane(plane, &state);
 
 	return 0;
 }
@@ -11644,6 +12007,7 @@ static const struct drm_plane_funcs intel_primary_plane_funcs = {
 	.update_plane = intel_primary_plane_setplane,
 	.disable_plane = intel_primary_plane_disable,
 	.destroy = intel_plane_destroy,
+	.set_property = intel_plane_set_property
 };
 
 static struct drm_plane *intel_primary_plane_create(struct drm_device *dev,
@@ -11661,6 +12025,7 @@ static struct drm_plane *intel_primary_plane_create(struct drm_device *dev,
 	primary->max_downscale = 1;
 	primary->pipe = pipe;
 	primary->plane = pipe;
+	primary->rotation = BIT(DRM_ROTATE_0);
 	if (HAS_FBC(dev) && INTEL_INFO(dev)->gen < 4)
 		primary->plane = !pipe;
 
@@ -11676,6 +12041,19 @@ static struct drm_plane *intel_primary_plane_create(struct drm_device *dev,
 				 &intel_primary_plane_funcs,
 				 intel_primary_formats, num_formats,
 				 DRM_PLANE_TYPE_PRIMARY);
+
+	if (INTEL_INFO(dev)->gen >= 4) {
+		if (!dev->mode_config.rotation_property)
+			dev->mode_config.rotation_property =
+				drm_mode_create_rotation_property(dev,
+							BIT(DRM_ROTATE_0) |
+							BIT(DRM_ROTATE_180));
+		if (dev->mode_config.rotation_property)
+			drm_object_attach_property(&primary->base.base,
+				dev->mode_config.rotation_property,
+				primary->rotation);
+	}
+
 	return &primary->base;
 }
 
@@ -11691,58 +12069,146 @@ intel_cursor_plane_disable(struct drm_plane *plane)
 }
 
 static int
-intel_cursor_plane_update(struct drm_plane *plane, struct drm_crtc *crtc,
-			  struct drm_framebuffer *fb, int crtc_x, int crtc_y,
-			  unsigned int crtc_w, unsigned int crtc_h,
-			  uint32_t src_x, uint32_t src_y,
-			  uint32_t src_w, uint32_t src_h)
+intel_check_cursor_plane(struct drm_plane *plane,
+			 struct intel_plane_state *state)
 {
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb);
-	struct drm_i915_gem_object *obj = intel_fb->obj;
-	struct drm_rect dest = {
-		/* integer pixels */
-		.x1 = crtc_x,
-		.y1 = crtc_y,
-		.x2 = crtc_x + crtc_w,
-		.y2 = crtc_y + crtc_h,
-	};
-	struct drm_rect src = {
-		/* 16.16 fixed point */
-		.x1 = src_x,
-		.y1 = src_y,
-		.x2 = src_x + src_w,
-		.y2 = src_y + src_h,
-	};
-	const struct drm_rect clip = {
-		/* integer pixels */
-		.x2 = intel_crtc->active ? intel_crtc->config.pipe_src_w : 0,
-		.y2 = intel_crtc->active ? intel_crtc->config.pipe_src_h : 0,
-	};
-	bool visible;
+	struct drm_crtc *crtc = state->crtc;
+	struct drm_device *dev = crtc->dev;
+	struct drm_framebuffer *fb = state->fb;
+	struct drm_rect *dest = &state->dst;
+	struct drm_rect *src = &state->src;
+	const struct drm_rect *clip = &state->clip;
+	struct drm_i915_gem_object *obj = intel_fb_obj(fb);
+	int crtc_w, crtc_h;
+	unsigned stride;
 	int ret;
 
 	ret = drm_plane_helper_check_update(plane, crtc, fb,
-					    &src, &dest, &clip,
+					    src, dest, clip,
 					    DRM_PLANE_HELPER_NO_SCALING,
 					    DRM_PLANE_HELPER_NO_SCALING,
-					    true, true, &visible);
+					    true, true, &state->visible);
 	if (ret)
 		return ret;
 
-	crtc->cursor_x = crtc_x;
-	crtc->cursor_y = crtc_y;
+
+	/* if we want to turn off the cursor ignore width and height */
+	if (!obj)
+		return 0;
+
+	/* Check for which cursor types we support */
+	crtc_w = drm_rect_width(&state->orig_dst);
+	crtc_h = drm_rect_height(&state->orig_dst);
+	if (!cursor_size_ok(dev, crtc_w, crtc_h)) {
+		DRM_DEBUG("Cursor dimension not supported\n");
+		return -EINVAL;
+	}
+
+	stride = roundup_pow_of_two(crtc_w) * 4;
+	if (obj->base.size < stride * crtc_h) {
+		DRM_DEBUG_KMS("buffer is too small\n");
+		return -ENOMEM;
+	}
+
+	if (fb == crtc->cursor->fb)
+		return 0;
+
+	/* we only need to pin inside GTT if cursor is non-phy */
+	mutex_lock(&dev->struct_mutex);
+	if (!INTEL_INFO(dev)->cursor_needs_physical && obj->tiling_mode) {
+		DRM_DEBUG_KMS("cursor cannot be tiled\n");
+		ret = -EINVAL;
+	}
+	mutex_unlock(&dev->struct_mutex);
+
+	return ret;
+}
+
+static int
+intel_commit_cursor_plane(struct drm_plane *plane,
+			  struct intel_plane_state *state)
+{
+	struct drm_crtc *crtc = state->crtc;
+	struct drm_framebuffer *fb = state->fb;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	struct intel_plane *intel_plane = to_intel_plane(plane);
+	struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb);
+	struct drm_i915_gem_object *obj = intel_fb->obj;
+	int crtc_w, crtc_h;
+
+	crtc->cursor_x = state->orig_dst.x1;
+	crtc->cursor_y = state->orig_dst.y1;
+
+	intel_plane->crtc_x = state->orig_dst.x1;
+	intel_plane->crtc_y = state->orig_dst.y1;
+	intel_plane->crtc_w = drm_rect_width(&state->orig_dst);
+	intel_plane->crtc_h = drm_rect_height(&state->orig_dst);
+	intel_plane->src_x = state->orig_src.x1;
+	intel_plane->src_y = state->orig_src.y1;
+	intel_plane->src_w = drm_rect_width(&state->orig_src);
+	intel_plane->src_h = drm_rect_height(&state->orig_src);
+	intel_plane->obj = obj;
+
 	if (fb != crtc->cursor->fb) {
+		crtc_w = drm_rect_width(&state->orig_dst);
+		crtc_h = drm_rect_height(&state->orig_dst);
 		return intel_crtc_cursor_set_obj(crtc, obj, crtc_w, crtc_h);
 	} else {
-		intel_crtc_update_cursor(crtc, visible);
+		intel_crtc_update_cursor(crtc, state->visible);
+
+		intel_frontbuffer_flip(crtc->dev,
+				       INTEL_FRONTBUFFER_CURSOR(intel_crtc->pipe));
+
 		return 0;
 	}
 }
+
+static int
+intel_cursor_plane_update(struct drm_plane *plane, struct drm_crtc *crtc,
+			  struct drm_framebuffer *fb, int crtc_x, int crtc_y,
+			  unsigned int crtc_w, unsigned int crtc_h,
+			  uint32_t src_x, uint32_t src_y,
+			  uint32_t src_w, uint32_t src_h)
+{
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	struct intel_plane_state state;
+	int ret;
+
+	state.crtc = crtc;
+	state.fb = fb;
+
+	/* sample coordinates in 16.16 fixed point */
+	state.src.x1 = src_x;
+	state.src.x2 = src_x + src_w;
+	state.src.y1 = src_y;
+	state.src.y2 = src_y + src_h;
+
+	/* integer pixels */
+	state.dst.x1 = crtc_x;
+	state.dst.x2 = crtc_x + crtc_w;
+	state.dst.y1 = crtc_y;
+	state.dst.y2 = crtc_y + crtc_h;
+
+	state.clip.x1 = 0;
+	state.clip.y1 = 0;
+	state.clip.x2 = intel_crtc->active ? intel_crtc->config.pipe_src_w : 0;
+	state.clip.y2 = intel_crtc->active ? intel_crtc->config.pipe_src_h : 0;
+
+	state.orig_src = state.src;
+	state.orig_dst = state.dst;
+
+	ret = intel_check_cursor_plane(plane, &state);
+	if (ret)
+		return ret;
+
+	return intel_commit_cursor_plane(plane, &state);
+}
+
 static const struct drm_plane_funcs intel_cursor_plane_funcs = {
 	.update_plane = intel_cursor_plane_update,
 	.disable_plane = intel_cursor_plane_disable,
 	.destroy = intel_plane_destroy,
+	.set_property = intel_plane_set_property,
 };
 
 static struct drm_plane *intel_cursor_plane_create(struct drm_device *dev,
@@ -11758,12 +12224,26 @@ static struct drm_plane *intel_cursor_plane_create(struct drm_device *dev,
 	cursor->max_downscale = 1;
 	cursor->pipe = pipe;
 	cursor->plane = pipe;
+	cursor->rotation = BIT(DRM_ROTATE_0);
 
 	drm_universal_plane_init(dev, &cursor->base, 0,
 				 &intel_cursor_plane_funcs,
 				 intel_cursor_formats,
 				 ARRAY_SIZE(intel_cursor_formats),
 				 DRM_PLANE_TYPE_CURSOR);
+
+	if (INTEL_INFO(dev)->gen >= 4) {
+		if (!dev->mode_config.rotation_property)
+			dev->mode_config.rotation_property =
+				drm_mode_create_rotation_property(dev,
+							BIT(DRM_ROTATE_0) |
+							BIT(DRM_ROTATE_180));
+		if (dev->mode_config.rotation_property)
+			drm_object_attach_property(&cursor->base.base,
+				dev->mode_config.rotation_property,
+				cursor->rotation);
+	}
+
 	return &cursor->base;
 }
 
@@ -11812,14 +12292,15 @@ static void intel_crtc_init(struct drm_device *dev, int pipe)
 
 	intel_crtc->cursor_base = ~0;
 	intel_crtc->cursor_cntl = ~0;
-
-	init_waitqueue_head(&intel_crtc->vbl_wait);
+	intel_crtc->cursor_size = ~0;
 
 	BUG_ON(pipe >= ARRAY_SIZE(dev_priv->plane_to_crtc_mapping) ||
 	       dev_priv->plane_to_crtc_mapping[intel_crtc->plane] != NULL);
 	dev_priv->plane_to_crtc_mapping[intel_crtc->plane] = &intel_crtc->base;
 	dev_priv->pipe_to_crtc_mapping[intel_crtc->pipe] = &intel_crtc->base;
 
+	INIT_WORK(&intel_crtc->mmio_flip.work, intel_mmio_flip_work_func);
+
 	drm_crtc_helper_add(&intel_crtc->base, &intel_helper_funcs);
 
 	WARN_ON(drm_crtc_index(&intel_crtc->base) != intel_crtc->pipe);
@@ -11840,7 +12321,7 @@ enum pipe intel_get_pipe_from_connector(struct intel_connector *connector)
 
 	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
 
-	if (!encoder)
+	if (!encoder || WARN_ON(!encoder->crtc))
 		return INVALID_PIPE;
 
 	return to_intel_crtc(encoder->crtc)->pipe;
@@ -11876,8 +12357,7 @@ static int intel_encoder_clones(struct intel_encoder *encoder)
 	int index_mask = 0;
 	int entry = 0;
 
-	list_for_each_entry(source_encoder,
-			    &dev->mode_config.encoder_list, base.head) {
+	for_each_intel_encoder(dev, source_encoder) {
 		if (encoders_cloneable(encoder, source_encoder))
 			index_mask |= (1 << entry);
 
@@ -11929,7 +12409,10 @@ static bool intel_crt_present(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (IS_ULT(dev))
+	if (INTEL_INFO(dev)->gen >= 9)
+		return false;
+
+	if (IS_HSW_ULT(dev) || IS_BDW_ULT(dev))
 		return false;
 
 	if (IS_CHERRYVIEW(dev))
@@ -11999,27 +12482,36 @@ static void intel_setup_outputs(struct drm_device *dev)
 		if (I915_READ(PCH_DP_D) & DP_DETECTED)
 			intel_dp_init(dev, PCH_DP_D, PORT_D);
 	} else if (IS_VALLEYVIEW(dev)) {
-		if (I915_READ(VLV_DISPLAY_BASE + GEN4_HDMIB) & SDVO_DETECTED) {
+		/*
+		 * The DP_DETECTED bit is the latched state of the DDC
+		 * SDA pin at boot. However since eDP doesn't require DDC
+		 * (no way to plug in a DP->HDMI dongle) the DDC pins for
+		 * eDP ports may have been muxed to an alternate function.
+		 * Thus we can't rely on the DP_DETECTED bit alone to detect
+		 * eDP ports. Consult the VBT as well as DP_DETECTED to
+		 * detect eDP ports.
+		 */
+		if (I915_READ(VLV_DISPLAY_BASE + GEN4_HDMIB) & SDVO_DETECTED)
 			intel_hdmi_init(dev, VLV_DISPLAY_BASE + GEN4_HDMIB,
 					PORT_B);
-			if (I915_READ(VLV_DISPLAY_BASE + DP_B) & DP_DETECTED)
-				intel_dp_init(dev, VLV_DISPLAY_BASE + DP_B, PORT_B);
-		}
+		if (I915_READ(VLV_DISPLAY_BASE + DP_B) & DP_DETECTED ||
+		    intel_dp_is_edp(dev, PORT_B))
+			intel_dp_init(dev, VLV_DISPLAY_BASE + DP_B, PORT_B);
 
-		if (I915_READ(VLV_DISPLAY_BASE + GEN4_HDMIC) & SDVO_DETECTED) {
+		if (I915_READ(VLV_DISPLAY_BASE + GEN4_HDMIC) & SDVO_DETECTED)
 			intel_hdmi_init(dev, VLV_DISPLAY_BASE + GEN4_HDMIC,
 					PORT_C);
-			if (I915_READ(VLV_DISPLAY_BASE + DP_C) & DP_DETECTED)
-				intel_dp_init(dev, VLV_DISPLAY_BASE + DP_C, PORT_C);
-		}
+		if (I915_READ(VLV_DISPLAY_BASE + DP_C) & DP_DETECTED ||
+		    intel_dp_is_edp(dev, PORT_C))
+			intel_dp_init(dev, VLV_DISPLAY_BASE + DP_C, PORT_C);
 
 		if (IS_CHERRYVIEW(dev)) {
-			if (I915_READ(VLV_DISPLAY_BASE + CHV_HDMID) & SDVO_DETECTED) {
+			if (I915_READ(VLV_DISPLAY_BASE + CHV_HDMID) & SDVO_DETECTED)
 				intel_hdmi_init(dev, VLV_DISPLAY_BASE + CHV_HDMID,
 						PORT_D);
-				if (I915_READ(VLV_DISPLAY_BASE + DP_D) & DP_DETECTED)
-					intel_dp_init(dev, VLV_DISPLAY_BASE + DP_D, PORT_D);
-			}
+			/* eDP not supported on port D, so don't check VBT */
+			if (I915_READ(VLV_DISPLAY_BASE + DP_D) & DP_DETECTED)
+				intel_dp_init(dev, VLV_DISPLAY_BASE + DP_D, PORT_D);
 		}
 
 		intel_dsi_init(dev);
@@ -12064,9 +12556,9 @@ static void intel_setup_outputs(struct drm_device *dev)
 	if (SUPPORTS_TV(dev))
 		intel_tv_init(dev);
 
-	intel_edp_psr_init(dev);
+	intel_psr_init(dev);
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		encoder->base.possible_crtcs = encoder->crtc_mask;
 		encoder->base.possible_clones =
 			intel_encoder_clones(encoder);
@@ -12268,16 +12760,22 @@ static void intel_init_display(struct drm_device *dev)
 	if (HAS_DDI(dev)) {
 		dev_priv->display.get_pipe_config = haswell_get_pipe_config;
 		dev_priv->display.get_plane_config = ironlake_get_plane_config;
-		dev_priv->display.crtc_mode_set = haswell_crtc_mode_set;
+		dev_priv->display.crtc_compute_clock =
+			haswell_crtc_compute_clock;
 		dev_priv->display.crtc_enable = haswell_crtc_enable;
 		dev_priv->display.crtc_disable = haswell_crtc_disable;
 		dev_priv->display.off = ironlake_crtc_off;
-		dev_priv->display.update_primary_plane =
-			ironlake_update_primary_plane;
+		if (INTEL_INFO(dev)->gen >= 9)
+			dev_priv->display.update_primary_plane =
+				skylake_update_primary_plane;
+		else
+			dev_priv->display.update_primary_plane =
+				ironlake_update_primary_plane;
 	} else if (HAS_PCH_SPLIT(dev)) {
 		dev_priv->display.get_pipe_config = ironlake_get_pipe_config;
 		dev_priv->display.get_plane_config = ironlake_get_plane_config;
-		dev_priv->display.crtc_mode_set = ironlake_crtc_mode_set;
+		dev_priv->display.crtc_compute_clock =
+			ironlake_crtc_compute_clock;
 		dev_priv->display.crtc_enable = ironlake_crtc_enable;
 		dev_priv->display.crtc_disable = ironlake_crtc_disable;
 		dev_priv->display.off = ironlake_crtc_off;
@@ -12286,7 +12784,7 @@ static void intel_init_display(struct drm_device *dev)
 	} else if (IS_VALLEYVIEW(dev)) {
 		dev_priv->display.get_pipe_config = i9xx_get_pipe_config;
 		dev_priv->display.get_plane_config = i9xx_get_plane_config;
-		dev_priv->display.crtc_mode_set = i9xx_crtc_mode_set;
+		dev_priv->display.crtc_compute_clock = i9xx_crtc_compute_clock;
 		dev_priv->display.crtc_enable = valleyview_crtc_enable;
 		dev_priv->display.crtc_disable = i9xx_crtc_disable;
 		dev_priv->display.off = i9xx_crtc_off;
@@ -12295,7 +12793,7 @@ static void intel_init_display(struct drm_device *dev)
 	} else {
 		dev_priv->display.get_pipe_config = i9xx_get_pipe_config;
 		dev_priv->display.get_plane_config = i9xx_get_plane_config;
-		dev_priv->display.crtc_mode_set = i9xx_crtc_mode_set;
+		dev_priv->display.crtc_compute_clock = i9xx_crtc_compute_clock;
 		dev_priv->display.crtc_enable = i9xx_crtc_enable;
 		dev_priv->display.crtc_disable = i9xx_crtc_disable;
 		dev_priv->display.off = i9xx_crtc_off;
@@ -12332,33 +12830,20 @@ static void intel_init_display(struct drm_device *dev)
 		dev_priv->display.get_display_clock_speed =
 			i830_get_display_clock_speed;
 
-	if (HAS_PCH_SPLIT(dev)) {
-		if (IS_GEN5(dev)) {
-			dev_priv->display.fdi_link_train = ironlake_fdi_link_train;
-			dev_priv->display.write_eld = ironlake_write_eld;
-		} else if (IS_GEN6(dev)) {
-			dev_priv->display.fdi_link_train = gen6_fdi_link_train;
-			dev_priv->display.write_eld = ironlake_write_eld;
-			dev_priv->display.modeset_global_resources =
-				snb_modeset_global_resources;
-		} else if (IS_IVYBRIDGE(dev)) {
-			/* FIXME: detect B0+ stepping and use auto training */
-			dev_priv->display.fdi_link_train = ivb_manual_fdi_link_train;
-			dev_priv->display.write_eld = ironlake_write_eld;
-			dev_priv->display.modeset_global_resources =
-				ivb_modeset_global_resources;
-		} else if (IS_HASWELL(dev) || IS_GEN8(dev)) {
-			dev_priv->display.fdi_link_train = hsw_fdi_link_train;
-			dev_priv->display.write_eld = haswell_write_eld;
-			dev_priv->display.modeset_global_resources =
-				haswell_modeset_global_resources;
-		}
-	} else if (IS_G4X(dev)) {
-		dev_priv->display.write_eld = g4x_write_eld;
+	if (IS_GEN5(dev)) {
+		dev_priv->display.fdi_link_train = ironlake_fdi_link_train;
+	} else if (IS_GEN6(dev)) {
+		dev_priv->display.fdi_link_train = gen6_fdi_link_train;
+	} else if (IS_IVYBRIDGE(dev)) {
+		/* FIXME: detect B0+ stepping and use auto training */
+		dev_priv->display.fdi_link_train = ivb_manual_fdi_link_train;
+		dev_priv->display.modeset_global_resources =
+			ivb_modeset_global_resources;
+	} else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
+		dev_priv->display.fdi_link_train = hsw_fdi_link_train;
 	} else if (IS_VALLEYVIEW(dev)) {
 		dev_priv->display.modeset_global_resources =
 			valleyview_modeset_global_resources;
-		dev_priv->display.write_eld = ironlake_write_eld;
 	}
 
 	/* Default just returns -ENODEV to indicate unsupported */
@@ -12385,9 +12870,14 @@ static void intel_init_display(struct drm_device *dev)
 	case 8: /* FIXME(BDW): Check that the gen8 RCS flip works. */
 		dev_priv->display.queue_flip = intel_gen7_queue_flip;
 		break;
+	case 9:
+		dev_priv->display.queue_flip = intel_gen9_queue_flip;
+		break;
 	}
 
 	intel_panel_init_backlight_funcs(dev);
+
+	mutex_init(&dev_priv->pps_mutex);
 }
 
 /*
@@ -12403,6 +12893,14 @@ static void quirk_pipea_force(struct drm_device *dev)
 	DRM_INFO("applying pipe a force quirk\n");
 }
 
+static void quirk_pipeb_force(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	dev_priv->quirks |= QUIRK_PIPEB_FORCE;
+	DRM_INFO("applying pipe b force quirk\n");
+}
+
 /*
  * Some machines (Lenovo U160) do not work with SSC on LVDS for some reason
  */
@@ -12477,6 +12975,12 @@ static struct intel_quirk intel_quirks[] = {
 	/* ThinkPad T60 needs pipe A force quirk (bug #16494) */
 	{ 0x2782, 0x17aa, 0x201a, quirk_pipea_force },
 
+	/* 830 needs to leave pipe A & dpll A up */
+	{ 0x3577, PCI_ANY_ID, PCI_ANY_ID, quirk_pipea_force },
+
+	/* 830 needs to leave pipe B & dpll B up */
+	{ 0x3577, PCI_ANY_ID, PCI_ANY_ID, quirk_pipeb_force },
+
 	/* Lenovo U160 cannot use SSC on LVDS */
 	{ 0x0046, 0x17aa, 0x3920, quirk_ssc_force_disable },
 
@@ -12507,6 +13011,9 @@ static struct intel_quirk intel_quirks[] = {
 	/* Acer C720 Chromebook (Core i3 4005U) */
 	{ 0x0a16, 0x1025, 0x0a11, quirk_backlight_present },
 
+	/* Apple Macbook 2,1 (Core 2 T7400) */
+	{ 0x27a2, 0x8086, 0x7270, quirk_backlight_present },
+
 	/* Toshiba CB35 Chromebook (Celeron 2955U) */
 	{ 0x0a06, 0x1179, 0x0a88, quirk_backlight_present },
 
@@ -12563,16 +13070,9 @@ void intel_modeset_init_hw(struct drm_device *dev)
 
 	intel_init_clock_gating(dev);
 
-	intel_reset_dpio(dev);
-
 	intel_enable_gt_powersave(dev);
 }
 
-void intel_modeset_suspend_hw(struct drm_device *dev)
-{
-	intel_suspend_hw(dev);
-}
-
 void intel_modeset_init(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -12598,6 +13098,7 @@ void intel_modeset_init(struct drm_device *dev)
 		return;
 
 	intel_init_display(dev);
+	intel_init_audio(dev);
 
 	if (IS_GEN2(dev)) {
 		dev->mode_config.max_width = 2048;
@@ -12610,7 +13111,10 @@ void intel_modeset_init(struct drm_device *dev)
 		dev->mode_config.max_height = 8192;
 	}
 
-	if (IS_GEN2(dev)) {
+	if (IS_845G(dev) || IS_I865G(dev)) {
+		dev->mode_config.cursor_width = IS_845G(dev) ? 64 : 512;
+		dev->mode_config.cursor_height = 1023;
+	} else if (IS_GEN2(dev)) {
 		dev->mode_config.cursor_width = GEN2_CURSOR_WIDTH;
 		dev->mode_config.cursor_height = GEN2_CURSOR_HEIGHT;
 	} else {
@@ -12624,7 +13128,7 @@ void intel_modeset_init(struct drm_device *dev)
 		      INTEL_INFO(dev)->num_pipes,
 		      INTEL_INFO(dev)->num_pipes > 1 ? "s" : "");
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		intel_crtc_init(dev, pipe);
 		for_each_sprite(pipe, sprite) {
 			ret = intel_plane_init(dev, pipe, sprite);
@@ -12635,7 +13139,6 @@ void intel_modeset_init(struct drm_device *dev)
 	}
 
 	intel_init_dpio(dev);
-	intel_reset_dpio(dev);
 
 	intel_shared_dpll_init(dev);
 
@@ -12730,9 +13233,10 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc)
 	I915_WRITE(reg, I915_READ(reg) & ~PIPECONF_FRAME_START_DELAY_MASK);
 
 	/* restore vblank interrupts to correct state */
-	if (crtc->active)
+	if (crtc->active) {
+		update_scanline_offset(crtc);
 		drm_vblank_on(dev, crtc->pipe);
-	else
+	} else
 		drm_vblank_off(dev, crtc->pipe);
 
 	/* We need to sanitize the plane -> pipe mapping first because this will
@@ -12815,7 +13319,7 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc)
 		}
 	}
 
-	if (crtc->active || IS_VALLEYVIEW(dev) || INTEL_INFO(dev)->gen < 5) {
+	if (crtc->active || HAS_GMCH_DISPLAY(dev)) {
 		/*
 		 * We start out with underrun reporting disabled to avoid races.
 		 * For correct bookkeeping mark this on active crtcs.
@@ -12831,8 +13335,6 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc)
 		 */
 		crtc->cpu_fifo_underrun_disabled = true;
 		crtc->pch_fifo_underrun_disabled = true;
-
-		update_scanline_offset(crtc);
 	}
 }
 
@@ -12905,7 +13407,7 @@ void i915_redisable_vga(struct drm_device *dev)
 	 * level, just check if the power well is enabled instead of trying to
 	 * follow the "don't touch the power well if we don't need it" policy
 	 * the rest of the driver uses. */
-	if (!intel_display_power_enabled(dev_priv, POWER_DOMAIN_VGA))
+	if (!intel_display_power_is_enabled(dev_priv, POWER_DOMAIN_VGA))
 		return;
 
 	i915_redisable_vga_power_on(dev);
@@ -12949,23 +13451,25 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 	for (i = 0; i < dev_priv->num_shared_dpll; i++) {
 		struct intel_shared_dpll *pll = &dev_priv->shared_dplls[i];
 
-		pll->on = pll->get_hw_state(dev_priv, pll, &pll->hw_state);
+		pll->on = pll->get_hw_state(dev_priv, pll,
+					    &pll->config.hw_state);
 		pll->active = 0;
+		pll->config.crtc_mask = 0;
 		for_each_intel_crtc(dev, crtc) {
-			if (crtc->active && intel_crtc_to_shared_dpll(crtc) == pll)
+			if (crtc->active && intel_crtc_to_shared_dpll(crtc) == pll) {
 				pll->active++;
+				pll->config.crtc_mask |= 1 << crtc->pipe;
+			}
 		}
-		pll->refcount = pll->active;
 
-		DRM_DEBUG_KMS("%s hw state readout: refcount %i, on %i\n",
-			      pll->name, pll->refcount, pll->on);
+		DRM_DEBUG_KMS("%s hw state readout: crtc_mask 0x%08x, on %i\n",
+			      pll->name, pll->config.crtc_mask, pll->on);
 
-		if (pll->refcount)
+		if (pll->config.crtc_mask)
 			intel_display_power_get(dev_priv, POWER_DOMAIN_PLLS);
 	}
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		pipe = 0;
 
 		if (encoder->get_hw_state(encoder, &pipe)) {
@@ -13029,12 +13533,11 @@ void intel_modeset_setup_hw_state(struct drm_device *dev,
 	}
 
 	/* HW state is read out, now we need to sanitize this mess. */
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		intel_sanitize_encoder(encoder);
 	}
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		crtc = to_intel_crtc(dev_priv->pipe_to_crtc_mapping[pipe]);
 		intel_sanitize_crtc(crtc);
 		intel_dump_pipe_config(crtc, &crtc->config, "[setup_hw_state]");
@@ -13052,7 +13555,9 @@ void intel_modeset_setup_hw_state(struct drm_device *dev,
 		pll->on = false;
 	}
 
-	if (HAS_PCH_SPLIT(dev))
+	if (IS_GEN9(dev))
+		skl_wm_get_hw_state(dev);
+	else if (HAS_PCH_SPLIT(dev))
 		ilk_wm_get_hw_state(dev);
 
 	if (force_restore) {
@@ -13062,12 +13567,12 @@ void intel_modeset_setup_hw_state(struct drm_device *dev,
 		 * We need to use raw interfaces for restoring state to avoid
 		 * checking (bogus) intermediate states.
 		 */
-		for_each_pipe(pipe) {
+		for_each_pipe(dev_priv, pipe) {
 			struct drm_crtc *crtc =
 				dev_priv->pipe_to_crtc_mapping[pipe];
 
-			__intel_set_mode(crtc, &crtc->mode, crtc->x, crtc->y,
-					 crtc->primary->fb);
+			intel_set_mode(crtc, &crtc->mode, crtc->x, crtc->y,
+				       crtc->primary->fb);
 		}
 	} else {
 		intel_modeset_update_staged_output_state(dev);
@@ -13078,6 +13583,7 @@ void intel_modeset_setup_hw_state(struct drm_device *dev,
 
 void intel_modeset_gem_init(struct drm_device *dev)
 {
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_crtc *c;
 	struct drm_i915_gem_object *obj;
 
@@ -13085,6 +13591,16 @@ void intel_modeset_gem_init(struct drm_device *dev)
 	intel_init_gt_powersave(dev);
 	mutex_unlock(&dev->struct_mutex);
 
+	/*
+	 * There may be no VBT; and if the BIOS enabled SSC we can
+	 * just keep using it to avoid unnecessary flicker.  Whereas if the
+	 * BIOS isn't using it, don't assume it will work even if the VBT
+	 * indicates as much.
+	 */
+	if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev))
+		dev_priv->vbt.lvds_use_ssc = !!(I915_READ(PCH_DREF_CONTROL) &
+						DREF_SSC1_ENABLE);
+
 	intel_modeset_init_hw(dev);
 
 	intel_setup_overlay(dev);
@@ -13100,7 +13616,9 @@ void intel_modeset_gem_init(struct drm_device *dev)
 		if (obj == NULL)
 			continue;
 
-		if (intel_pin_and_fence_fb_obj(dev, obj, NULL)) {
+		if (intel_pin_and_fence_fb_obj(c->primary,
+					       c->primary->fb,
+					       NULL)) {
 			DRM_ERROR("failed to pin boot fb on pipe %d\n",
 				  to_intel_crtc(c)->pipe);
 			drm_framebuffer_unreference(c->primary->fb);
@@ -13108,6 +13626,8 @@ void intel_modeset_gem_init(struct drm_device *dev)
 		}
 	}
 	mutex_unlock(&dev->struct_mutex);
+
+	intel_backlight_register(dev);
 }
 
 void intel_connector_unregister(struct intel_connector *intel_connector)
@@ -13123,14 +13643,16 @@ void intel_modeset_cleanup(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_connector *connector;
 
+	intel_disable_gt_powersave(dev);
+
+	intel_backlight_unregister(dev);
+
 	/*
 	 * Interrupts and polling as the first thing to avoid creating havoc.
-	 * Too much stuff here (turning of rps, connectors, ...) would
+	 * Too much stuff here (turning of connectors, ...) would
 	 * experience fancy races otherwise.
 	 */
-	drm_irq_uninstall(dev);
-	intel_hpd_cancel_work(dev_priv);
-	dev_priv->pm._irqs_disabled = true;
+	intel_irq_uninstall(dev_priv);
 
 	/*
 	 * Due to the hpd irq storm handling the hotplug work can re-arm the
@@ -13144,8 +13666,6 @@ void intel_modeset_cleanup(struct drm_device *dev)
 
 	intel_disable_fbc(dev);
 
-	intel_disable_gt_powersave(dev);
-
 	ironlake_teardown_rc6(dev);
 
 	mutex_unlock(&dev->struct_mutex);
@@ -13283,10 +13803,10 @@ intel_display_capture_error_state(struct drm_device *dev)
 	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
 		error->power_well_driver = I915_READ(HSW_PWR_WELL_DRIVER);
 
-	for_each_pipe(i) {
+	for_each_pipe(dev_priv, i) {
 		error->pipe[i].power_domain_on =
-			intel_display_power_enabled_unlocked(dev_priv,
-							   POWER_DOMAIN_PIPE(i));
+			__intel_display_power_is_enabled(dev_priv,
+							 POWER_DOMAIN_PIPE(i));
 		if (!error->pipe[i].power_domain_on)
 			continue;
 
@@ -13321,7 +13841,7 @@ intel_display_capture_error_state(struct drm_device *dev)
 		enum transcoder cpu_transcoder = transcoders[i];
 
 		error->transcoder[i].power_domain_on =
-			intel_display_power_enabled_unlocked(dev_priv,
+			__intel_display_power_is_enabled(dev_priv,
 				POWER_DOMAIN_TRANSCODER(cpu_transcoder));
 		if (!error->transcoder[i].power_domain_on)
 			continue;
@@ -13347,6 +13867,7 @@ intel_display_print_error_state(struct drm_i915_error_state_buf *m,
 				struct drm_device *dev,
 				struct intel_display_error_state *error)
 {
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int i;
 
 	if (!error)
@@ -13356,7 +13877,7 @@ intel_display_print_error_state(struct drm_i915_error_state_buf *m,
 	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
 		err_printf(m, "PWR_WELL_CTL2: %08x\n",
 			   error->power_well_driver);
-	for_each_pipe(i) {
+	for_each_pipe(dev_priv, i) {
 		err_printf(m, "Pipe [%d]:\n", i);
 		err_printf(m, "  Power: %s\n",
 			   error->pipe[i].power_domain_on ? "on" : "off");
@@ -13397,3 +13918,24 @@ intel_display_print_error_state(struct drm_i915_error_state_buf *m,
 		err_printf(m, "  VSYNC: %08x\n", error->transcoder[i].vsync);
 	}
 }
+
+void intel_modeset_preclose(struct drm_device *dev, struct drm_file *file)
+{
+	struct intel_crtc *crtc;
+
+	for_each_intel_crtc(dev, crtc) {
+		struct intel_unpin_work *work;
+
+		spin_lock_irq(&dev->event_lock);
+
+		work = crtc->unpin_work;
+
+		if (work && work->event &&
+		    work->event->base.file_priv == file) {
+			kfree(work->event);
+			work->event = NULL;
+		}
+
+		spin_unlock_irq(&dev->event_lock);
+	}
+}
diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index fdff1d420c14..5cecc20efa71 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -111,8 +111,11 @@ static struct intel_dp *intel_attached_dp(struct drm_connector *connector)
 }
 
 static void intel_dp_link_down(struct intel_dp *intel_dp);
-static bool _edp_panel_vdd_on(struct intel_dp *intel_dp);
+static bool edp_panel_vdd_on(struct intel_dp *intel_dp);
 static void edp_panel_vdd_off(struct intel_dp *intel_dp, bool sync);
+static void vlv_init_panel_power_sequencer(struct intel_dp *intel_dp);
+static void vlv_steal_power_sequencer(struct drm_device *dev,
+				      enum pipe pipe);
 
 int
 intel_dp_max_link_bw(struct intel_dp *intel_dp)
@@ -224,8 +227,7 @@ intel_dp_mode_valid(struct drm_connector *connector,
 	return MODE_OK;
 }
 
-static uint32_t
-pack_aux(uint8_t *src, int src_bytes)
+uint32_t intel_dp_pack_aux(const uint8_t *src, int src_bytes)
 {
 	int	i;
 	uint32_t v = 0;
@@ -237,8 +239,7 @@ pack_aux(uint8_t *src, int src_bytes)
 	return v;
 }
 
-static void
-unpack_aux(uint32_t src, uint8_t *dst, int dst_bytes)
+void intel_dp_unpack_aux(uint32_t src, uint8_t *dst, int dst_bytes)
 {
 	int i;
 	if (dst_bytes > 4)
@@ -283,39 +284,275 @@ intel_hrawclk(struct drm_device *dev)
 
 static void
 intel_dp_init_panel_power_sequencer(struct drm_device *dev,
-				    struct intel_dp *intel_dp,
-				    struct edp_power_seq *out);
+				    struct intel_dp *intel_dp);
 static void
 intel_dp_init_panel_power_sequencer_registers(struct drm_device *dev,
-					      struct intel_dp *intel_dp,
-					      struct edp_power_seq *out);
+					      struct intel_dp *intel_dp);
+
+static void pps_lock(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct intel_encoder *encoder = &intel_dig_port->base;
+	struct drm_device *dev = encoder->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum intel_display_power_domain power_domain;
+
+	/*
+	 * See vlv_power_sequencer_reset() why we need
+	 * a power domain reference here.
+	 */
+	power_domain = intel_display_port_power_domain(encoder);
+	intel_display_power_get(dev_priv, power_domain);
+
+	mutex_lock(&dev_priv->pps_mutex);
+}
+
+static void pps_unlock(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct intel_encoder *encoder = &intel_dig_port->base;
+	struct drm_device *dev = encoder->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum intel_display_power_domain power_domain;
+
+	mutex_unlock(&dev_priv->pps_mutex);
+
+	power_domain = intel_display_port_power_domain(encoder);
+	intel_display_power_put(dev_priv, power_domain);
+}
+
+static void
+vlv_power_sequencer_kick(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = intel_dig_port->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum pipe pipe = intel_dp->pps_pipe;
+	bool pll_enabled;
+	uint32_t DP;
+
+	if (WARN(I915_READ(intel_dp->output_reg) & DP_PORT_EN,
+		 "skipping pipe %c power seqeuncer kick due to port %c being active\n",
+		 pipe_name(pipe), port_name(intel_dig_port->port)))
+		return;
+
+	DRM_DEBUG_KMS("kicking pipe %c power sequencer for port %c\n",
+		      pipe_name(pipe), port_name(intel_dig_port->port));
+
+	/* Preserve the BIOS-computed detected bit. This is
+	 * supposed to be read-only.
+	 */
+	DP = I915_READ(intel_dp->output_reg) & DP_DETECTED;
+	DP |= DP_VOLTAGE_0_4 | DP_PRE_EMPHASIS_0;
+	DP |= DP_PORT_WIDTH(1);
+	DP |= DP_LINK_TRAIN_PAT_1;
+
+	if (IS_CHERRYVIEW(dev))
+		DP |= DP_PIPE_SELECT_CHV(pipe);
+	else if (pipe == PIPE_B)
+		DP |= DP_PIPEB_SELECT;
+
+	pll_enabled = I915_READ(DPLL(pipe)) & DPLL_VCO_ENABLE;
+
+	/*
+	 * The DPLL for the pipe must be enabled for this to work.
+	 * So enable temporarily it if it's not already enabled.
+	 */
+	if (!pll_enabled)
+		vlv_force_pll_on(dev, pipe, IS_CHERRYVIEW(dev) ?
+				 &chv_dpll[0].dpll : &vlv_dpll[0].dpll);
+
+	/*
+	 * Similar magic as in intel_dp_enable_port().
+	 * We _must_ do this port enable + disable trick
+	 * to make this power seqeuencer lock onto the port.
+	 * Otherwise even VDD force bit won't work.
+	 */
+	I915_WRITE(intel_dp->output_reg, DP);
+	POSTING_READ(intel_dp->output_reg);
+
+	I915_WRITE(intel_dp->output_reg, DP | DP_PORT_EN);
+	POSTING_READ(intel_dp->output_reg);
+
+	I915_WRITE(intel_dp->output_reg, DP & ~DP_PORT_EN);
+	POSTING_READ(intel_dp->output_reg);
+
+	if (!pll_enabled)
+		vlv_force_pll_off(dev, pipe);
+}
 
 static enum pipe
 vlv_power_sequencer_pipe(struct intel_dp *intel_dp)
 {
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
-	struct drm_crtc *crtc = intel_dig_port->base.base.crtc;
 	struct drm_device *dev = intel_dig_port->base.base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	enum port port = intel_dig_port->port;
+	struct intel_encoder *encoder;
+	unsigned int pipes = (1 << PIPE_A) | (1 << PIPE_B);
 	enum pipe pipe;
 
-	/* modeset should have pipe */
-	if (crtc)
-		return to_intel_crtc(crtc)->pipe;
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
+	/* We should never land here with regular DP ports */
+	WARN_ON(!is_edp(intel_dp));
+
+	if (intel_dp->pps_pipe != INVALID_PIPE)
+		return intel_dp->pps_pipe;
+
+	/*
+	 * We don't have power sequencer currently.
+	 * Pick one that's not used by other ports.
+	 */
+	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
+			    base.head) {
+		struct intel_dp *tmp;
+
+		if (encoder->type != INTEL_OUTPUT_EDP)
+			continue;
+
+		tmp = enc_to_intel_dp(&encoder->base);
+
+		if (tmp->pps_pipe != INVALID_PIPE)
+			pipes &= ~(1 << tmp->pps_pipe);
+	}
+
+	/*
+	 * Didn't find one. This should not happen since there
+	 * are two power sequencers and up to two eDP ports.
+	 */
+	if (WARN_ON(pipes == 0))
+		pipe = PIPE_A;
+	else
+		pipe = ffs(pipes) - 1;
+
+	vlv_steal_power_sequencer(dev, pipe);
+	intel_dp->pps_pipe = pipe;
+
+	DRM_DEBUG_KMS("picked pipe %c power sequencer for port %c\n",
+		      pipe_name(intel_dp->pps_pipe),
+		      port_name(intel_dig_port->port));
+
+	/* init power sequencer on this pipe and port */
+	intel_dp_init_panel_power_sequencer(dev, intel_dp);
+	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp);
+
+	/*
+	 * Even vdd force doesn't work until we've made
+	 * the power sequencer lock in on the port.
+	 */
+	vlv_power_sequencer_kick(intel_dp);
+
+	return intel_dp->pps_pipe;
+}
+
+typedef bool (*vlv_pipe_check)(struct drm_i915_private *dev_priv,
+			       enum pipe pipe);
+
+static bool vlv_pipe_has_pp_on(struct drm_i915_private *dev_priv,
+			       enum pipe pipe)
+{
+	return I915_READ(VLV_PIPE_PP_STATUS(pipe)) & PP_ON;
+}
+
+static bool vlv_pipe_has_vdd_on(struct drm_i915_private *dev_priv,
+				enum pipe pipe)
+{
+	return I915_READ(VLV_PIPE_PP_CONTROL(pipe)) & EDP_FORCE_VDD;
+}
+
+static bool vlv_pipe_any(struct drm_i915_private *dev_priv,
+			 enum pipe pipe)
+{
+	return true;
+}
+
+static enum pipe
+vlv_initial_pps_pipe(struct drm_i915_private *dev_priv,
+		     enum port port,
+		     vlv_pipe_check pipe_check)
+{
+	enum pipe pipe;
 
-	/* init time, try to find a pipe with this port selected */
 	for (pipe = PIPE_A; pipe <= PIPE_B; pipe++) {
 		u32 port_sel = I915_READ(VLV_PIPE_PP_ON_DELAYS(pipe)) &
 			PANEL_PORT_SELECT_MASK;
-		if (port_sel == PANEL_PORT_SELECT_DPB_VLV && port == PORT_B)
-			return pipe;
-		if (port_sel == PANEL_PORT_SELECT_DPC_VLV && port == PORT_C)
-			return pipe;
+
+		if (port_sel != PANEL_PORT_SELECT_VLV(port))
+			continue;
+
+		if (!pipe_check(dev_priv, pipe))
+			continue;
+
+		return pipe;
 	}
 
-	/* shrug */
-	return PIPE_A;
+	return INVALID_PIPE;
+}
+
+static void
+vlv_initial_power_sequencer_setup(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = intel_dig_port->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum port port = intel_dig_port->port;
+
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
+	/* try to find a pipe with this port selected */
+	/* first pick one where the panel is on */
+	intel_dp->pps_pipe = vlv_initial_pps_pipe(dev_priv, port,
+						  vlv_pipe_has_pp_on);
+	/* didn't find one? pick one where vdd is on */
+	if (intel_dp->pps_pipe == INVALID_PIPE)
+		intel_dp->pps_pipe = vlv_initial_pps_pipe(dev_priv, port,
+							  vlv_pipe_has_vdd_on);
+	/* didn't find one? pick one with just the correct port */
+	if (intel_dp->pps_pipe == INVALID_PIPE)
+		intel_dp->pps_pipe = vlv_initial_pps_pipe(dev_priv, port,
+							  vlv_pipe_any);
+
+	/* didn't find one? just let vlv_power_sequencer_pipe() pick one when needed */
+	if (intel_dp->pps_pipe == INVALID_PIPE) {
+		DRM_DEBUG_KMS("no initial power sequencer for port %c\n",
+			      port_name(port));
+		return;
+	}
+
+	DRM_DEBUG_KMS("initial power sequencer for port %c: pipe %c\n",
+		      port_name(port), pipe_name(intel_dp->pps_pipe));
+
+	intel_dp_init_panel_power_sequencer(dev, intel_dp);
+	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp);
+}
+
+void vlv_power_sequencer_reset(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct intel_encoder *encoder;
+
+	if (WARN_ON(!IS_VALLEYVIEW(dev)))
+		return;
+
+	/*
+	 * We can't grab pps_mutex here due to deadlock with power_domain
+	 * mutex when power_domain functions are called while holding pps_mutex.
+	 * That also means that in order to use pps_pipe the code needs to
+	 * hold both a power domain reference and pps_mutex, and the power domain
+	 * reference get/put must be done while _not_ holding pps_mutex.
+	 * pps_{lock,unlock}() do these steps in the correct order, so one
+	 * should use them always.
+	 */
+
+	list_for_each_entry(encoder, &dev->mode_config.encoder_list, base.head) {
+		struct intel_dp *intel_dp;
+
+		if (encoder->type != INTEL_OUTPUT_EDP)
+			continue;
+
+		intel_dp = enc_to_intel_dp(&encoder->base);
+		intel_dp->pps_pipe = INVALID_PIPE;
+	}
 }
 
 static u32 _pp_ctrl_reg(struct intel_dp *intel_dp)
@@ -349,12 +586,15 @@ static int edp_notify_handler(struct notifier_block *this, unsigned long code,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 pp_div;
 	u32 pp_ctrl_reg, pp_div_reg;
-	enum pipe pipe = vlv_power_sequencer_pipe(intel_dp);
 
 	if (!is_edp(intel_dp) || code != SYS_RESTART)
 		return 0;
 
+	pps_lock(intel_dp);
+
 	if (IS_VALLEYVIEW(dev)) {
+		enum pipe pipe = vlv_power_sequencer_pipe(intel_dp);
+
 		pp_ctrl_reg = VLV_PIPE_PP_CONTROL(pipe);
 		pp_div_reg  = VLV_PIPE_PP_DIVISOR(pipe);
 		pp_div = I915_READ(pp_div_reg);
@@ -366,6 +606,8 @@ static int edp_notify_handler(struct notifier_block *this, unsigned long code,
 		msleep(intel_dp->panel_power_cycle_delay);
 	}
 
+	pps_unlock(intel_dp);
+
 	return 0;
 }
 
@@ -374,6 +616,12 @@ static bool edp_have_panel_power(struct intel_dp *intel_dp)
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
+	if (IS_VALLEYVIEW(dev) &&
+	    intel_dp->pps_pipe == INVALID_PIPE)
+		return false;
+
 	return (I915_READ(_pp_stat_reg(intel_dp)) & PP_ON) != 0;
 }
 
@@ -381,13 +629,14 @@ static bool edp_have_panel_vdd(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
-	struct intel_encoder *intel_encoder = &intel_dig_port->base;
-	enum intel_display_power_domain power_domain;
 
-	power_domain = intel_display_port_power_domain(intel_encoder);
-	return intel_display_power_enabled(dev_priv, power_domain) &&
-	       (I915_READ(_pp_ctrl_reg(intel_dp)) & EDP_FORCE_VDD) != 0;
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
+	if (IS_VALLEYVIEW(dev) &&
+	    intel_dp->pps_pipe == INVALID_PIPE)
+		return false;
+
+	return I915_READ(_pp_ctrl_reg(intel_dp)) & EDP_FORCE_VDD;
 }
 
 static void
@@ -488,6 +737,16 @@ static uint32_t vlv_get_aux_clock_divider(struct intel_dp *intel_dp, int index)
 	return index ? 0 : 100;
 }
 
+static uint32_t skl_get_aux_clock_divider(struct intel_dp *intel_dp, int index)
+{
+	/*
+	 * SKL doesn't need us to program the AUX clock divider (Hardware will
+	 * derive the clock from CDCLK automatically). We still implement the
+	 * get_aux_clock_divider vfunc to plug-in into the existing code.
+	 */
+	return index ? 0 : 1;
+}
+
 static uint32_t i9xx_get_aux_send_ctl(struct intel_dp *intel_dp,
 				      bool has_aux_irq,
 				      int send_bytes,
@@ -518,9 +777,24 @@ static uint32_t i9xx_get_aux_send_ctl(struct intel_dp *intel_dp,
 	       (aux_clock_divider << DP_AUX_CH_CTL_BIT_CLOCK_2X_SHIFT);
 }
 
+static uint32_t skl_get_aux_send_ctl(struct intel_dp *intel_dp,
+				      bool has_aux_irq,
+				      int send_bytes,
+				      uint32_t unused)
+{
+	return DP_AUX_CH_CTL_SEND_BUSY |
+	       DP_AUX_CH_CTL_DONE |
+	       (has_aux_irq ? DP_AUX_CH_CTL_INTERRUPT : 0) |
+	       DP_AUX_CH_CTL_TIME_OUT_ERROR |
+	       DP_AUX_CH_CTL_TIME_OUT_1600us |
+	       DP_AUX_CH_CTL_RECEIVE_ERROR |
+	       (send_bytes << DP_AUX_CH_CTL_MESSAGE_SIZE_SHIFT) |
+	       DP_AUX_CH_CTL_SYNC_PULSE_SKL(32);
+}
+
 static int
 intel_dp_aux_ch(struct intel_dp *intel_dp,
-		uint8_t *send, int send_bytes,
+		const uint8_t *send, int send_bytes,
 		uint8_t *recv, int recv_size)
 {
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
@@ -535,7 +809,15 @@ intel_dp_aux_ch(struct intel_dp *intel_dp,
 	bool has_aux_irq = HAS_AUX_IRQ(dev);
 	bool vdd;
 
-	vdd = _edp_panel_vdd_on(intel_dp);
+	pps_lock(intel_dp);
+
+	/*
+	 * We will be called with VDD already enabled for dpcd/edid/oui reads.
+	 * In such cases we want to leave VDD enabled and it's up to upper layers
+	 * to turn it off. But for eg. i2c-dev access we need to turn it on/off
+	 * ourselves.
+	 */
+	vdd = edp_panel_vdd_on(intel_dp);
 
 	/* dp aux is extremely sensitive to irq latency, hence request the
 	 * lowest possible wakeup latency and so prevent the cpu from going into
@@ -579,7 +861,8 @@ intel_dp_aux_ch(struct intel_dp *intel_dp,
 			/* Load the send data into the aux channel data registers */
 			for (i = 0; i < send_bytes; i += 4)
 				I915_WRITE(ch_data + i,
-					   pack_aux(send + i, send_bytes - i));
+					   intel_dp_pack_aux(send + i,
+							     send_bytes - i));
 
 			/* Send the command and wait for it to complete */
 			I915_WRITE(ch_ctl, send_ctl);
@@ -633,8 +916,8 @@ intel_dp_aux_ch(struct intel_dp *intel_dp,
 		recv_bytes = recv_size;
 
 	for (i = 0; i < recv_bytes; i += 4)
-		unpack_aux(I915_READ(ch_data + i),
-			   recv + i, recv_bytes - i);
+		intel_dp_unpack_aux(I915_READ(ch_data + i),
+				    recv + i, recv_bytes - i);
 
 	ret = recv_bytes;
 out:
@@ -644,6 +927,8 @@ out:
 	if (vdd)
 		edp_panel_vdd_off(intel_dp, false);
 
+	pps_unlock(intel_dp);
+
 	return ret;
 }
 
@@ -742,7 +1027,16 @@ intel_dp_aux_init(struct intel_dp *intel_dp, struct intel_connector *connector)
 		BUG();
 	}
 
-	if (!HAS_DDI(dev))
+	/*
+	 * The AUX_CTL register is usually DP_CTL + 0x10.
+	 *
+	 * On Haswell and Broadwell though:
+	 *   - Both port A DDI_BUF_CTL and DDI_AUX_CTL are on the CPU
+	 *   - Port B/C/D AUX channels are on the PCH, DDI_BUF_CTL on the CPU
+	 *
+	 * Skylake moves AUX_CTL back next to DDI_BUF_CTL, on the CPU.
+	 */
+	if (!IS_HASWELL(dev) && !IS_BROADWELL(dev))
 		intel_dp->aux_ch_ctl_reg = intel_dp->output_reg + 0x10;
 
 	intel_dp->aux.name = name;
@@ -780,6 +1074,33 @@ intel_dp_connector_unregister(struct intel_connector *intel_connector)
 }
 
 static void
+skl_edp_set_pll_config(struct intel_crtc_config *pipe_config, int link_bw)
+{
+	u32 ctrl1;
+
+	pipe_config->ddi_pll_sel = SKL_DPLL0;
+	pipe_config->dpll_hw_state.cfgcr1 = 0;
+	pipe_config->dpll_hw_state.cfgcr2 = 0;
+
+	ctrl1 = DPLL_CTRL1_OVERRIDE(SKL_DPLL0);
+	switch (link_bw) {
+	case DP_LINK_BW_1_62:
+		ctrl1 |= DPLL_CRTL1_LINK_RATE(DPLL_CRTL1_LINK_RATE_810,
+					      SKL_DPLL0);
+		break;
+	case DP_LINK_BW_2_7:
+		ctrl1 |= DPLL_CRTL1_LINK_RATE(DPLL_CRTL1_LINK_RATE_1350,
+					      SKL_DPLL0);
+		break;
+	case DP_LINK_BW_5_4:
+		ctrl1 |= DPLL_CRTL1_LINK_RATE(DPLL_CRTL1_LINK_RATE_2700,
+					      SKL_DPLL0);
+		break;
+	}
+	pipe_config->dpll_hw_state.ctrl1 = ctrl1;
+}
+
+static void
 hsw_dp_set_ddi_pll_sel(struct intel_crtc_config *pipe_config, int link_bw)
 {
 	switch (link_bw) {
@@ -828,20 +1149,6 @@ intel_dp_set_clock(struct intel_encoder *encoder,
 	}
 }
 
-static void
-intel_dp_set_m2_n2(struct intel_crtc *crtc, struct intel_link_m_n *m_n)
-{
-	struct drm_device *dev = crtc->base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	enum transcoder transcoder = crtc->config.cpu_transcoder;
-
-	I915_WRITE(PIPE_DATA_M2(transcoder),
-		TU_SIZE(m_n->tu) | m_n->gmch_m);
-	I915_WRITE(PIPE_DATA_N2(transcoder), m_n->gmch_n);
-	I915_WRITE(PIPE_LINK_M2(transcoder), m_n->link_m);
-	I915_WRITE(PIPE_LINK_N2(transcoder), m_n->link_n);
-}
-
 bool
 intel_dp_compute_config(struct intel_encoder *encoder,
 			struct intel_crtc_config *pipe_config)
@@ -867,6 +1174,7 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 		pipe_config->has_pch_encoder = true;
 
 	pipe_config->has_dp_encoder = true;
+	pipe_config->has_drrs = false;
 	pipe_config->has_audio = intel_dp->has_audio;
 
 	if (is_edp(intel_dp) && intel_connector->panel.fixed_mode) {
@@ -898,23 +1206,15 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 			bpp = dev_priv->vbt.edp_bpp;
 		}
 
-		if (IS_BROADWELL(dev)) {
-			/* Yes, it's an ugly hack. */
-			min_lane_count = max_lane_count;
-			DRM_DEBUG_KMS("forcing lane count to max (%u) on BDW\n",
-				      min_lane_count);
-		} else if (dev_priv->vbt.edp_lanes) {
-			min_lane_count = min(dev_priv->vbt.edp_lanes,
-					     max_lane_count);
-			DRM_DEBUG_KMS("using min %u lanes per VBT\n",
-				      min_lane_count);
-		}
-
-		if (dev_priv->vbt.edp_rate) {
-			min_clock = min(dev_priv->vbt.edp_rate >> 3, max_clock);
-			DRM_DEBUG_KMS("using min %02x link bw per VBT\n",
-				      bws[min_clock]);
-		}
+		/*
+		 * Use the maximum clock and number of lanes the eDP panel
+		 * advertizes being capable of. The panels are generally
+		 * designed to support only a single clock and lane
+		 * configuration, and typically these values correspond to the
+		 * native resolution of the panel.
+		 */
+		min_lane_count = max_lane_count;
+		min_clock = max_clock;
 	}
 
 	for (; bpp >= 6*3; bpp -= 2*3) {
@@ -970,13 +1270,16 @@ found:
 
 	if (intel_connector->panel.downclock_mode != NULL &&
 		intel_dp->drrs_state.type == SEAMLESS_DRRS_SUPPORT) {
+			pipe_config->has_drrs = true;
 			intel_link_compute_m_n(bpp, lane_count,
 				intel_connector->panel.downclock_mode->clock,
 				pipe_config->port_clock,
 				&pipe_config->dp_m2_n2);
 	}
 
-	if (HAS_DDI(dev))
+	if (IS_SKYLAKE(dev) && is_edp(intel_dp))
+		skl_edp_set_pll_config(pipe_config, intel_dp->link_bw);
+	else if (IS_HASWELL(dev) || IS_BROADWELL(dev))
 		hsw_dp_set_ddi_pll_sel(pipe_config, intel_dp->link_bw);
 	else
 		intel_dp_set_clock(encoder, pipe_config, intel_dp->link_bw);
@@ -1049,12 +1352,8 @@ static void intel_dp_prepare(struct intel_encoder *encoder)
 	intel_dp->DP |= DP_VOLTAGE_0_4 | DP_PRE_EMPHASIS_0;
 	intel_dp->DP |= DP_PORT_WIDTH(intel_dp->lane_count);
 
-	if (crtc->config.has_audio) {
-		DRM_DEBUG_DRIVER("Enabling DP audio on pipe %c\n",
-				 pipe_name(crtc->pipe));
+	if (crtc->config.has_audio)
 		intel_dp->DP |= DP_AUDIO_OUTPUT_ENABLE;
-		intel_write_eld(&encoder->base, adjusted_mode);
-	}
 
 	/* Split out the IBX/CPU vs CPT settings */
 
@@ -1110,6 +1409,8 @@ static void wait_panel_status(struct intel_dp *intel_dp,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 pp_stat_reg, pp_ctrl_reg;
 
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
 	pp_stat_reg = _pp_stat_reg(intel_dp);
 	pp_ctrl_reg = _pp_ctrl_reg(intel_dp);
 
@@ -1173,13 +1474,20 @@ static  u32 ironlake_get_pp_control(struct intel_dp *intel_dp)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 control;
 
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
 	control = I915_READ(_pp_ctrl_reg(intel_dp));
 	control &= ~PANEL_UNLOCK_MASK;
 	control |= PANEL_UNLOCK_REGS;
 	return control;
 }
 
-static bool _edp_panel_vdd_on(struct intel_dp *intel_dp)
+/*
+ * Must be paired with edp_panel_vdd_off().
+ * Must hold pps_mutex around the whole on/off sequence.
+ * Can be nested with intel_edp_panel_vdd_{on,off}() calls.
+ */
+static bool edp_panel_vdd_on(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
@@ -1190,9 +1498,12 @@ static bool _edp_panel_vdd_on(struct intel_dp *intel_dp)
 	u32 pp_stat_reg, pp_ctrl_reg;
 	bool need_to_disable = !intel_dp->want_panel_vdd;
 
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
 	if (!is_edp(intel_dp))
 		return false;
 
+	cancel_delayed_work(&intel_dp->panel_vdd_work);
 	intel_dp->want_panel_vdd = true;
 
 	if (edp_have_panel_vdd(intel_dp))
@@ -1201,7 +1512,8 @@ static bool _edp_panel_vdd_on(struct intel_dp *intel_dp)
 	power_domain = intel_display_port_power_domain(intel_encoder);
 	intel_display_power_get(dev_priv, power_domain);
 
-	DRM_DEBUG_KMS("Turning eDP VDD on\n");
+	DRM_DEBUG_KMS("Turning eDP port %c VDD on\n",
+		      port_name(intel_dig_port->port));
 
 	if (!edp_have_panel_power(intel_dp))
 		wait_panel_power_cycle(intel_dp);
@@ -1220,69 +1532,86 @@ static bool _edp_panel_vdd_on(struct intel_dp *intel_dp)
 	 * If the panel wasn't on, delay before accessing aux channel
 	 */
 	if (!edp_have_panel_power(intel_dp)) {
-		DRM_DEBUG_KMS("eDP was not running\n");
+		DRM_DEBUG_KMS("eDP port %c panel power wasn't enabled\n",
+			      port_name(intel_dig_port->port));
 		msleep(intel_dp->panel_power_up_delay);
 	}
 
 	return need_to_disable;
 }
 
+/*
+ * Must be paired with intel_edp_panel_vdd_off() or
+ * intel_edp_panel_off().
+ * Nested calls to these functions are not allowed since
+ * we drop the lock. Caller must use some higher level
+ * locking to prevent nested calls from other threads.
+ */
 void intel_edp_panel_vdd_on(struct intel_dp *intel_dp)
 {
-	if (is_edp(intel_dp)) {
-		bool vdd = _edp_panel_vdd_on(intel_dp);
+	bool vdd;
 
-		WARN(!vdd, "eDP VDD already requested on\n");
-	}
+	if (!is_edp(intel_dp))
+		return;
+
+	pps_lock(intel_dp);
+	vdd = edp_panel_vdd_on(intel_dp);
+	pps_unlock(intel_dp);
+
+	WARN(!vdd, "eDP port %c VDD already requested on\n",
+	     port_name(dp_to_dig_port(intel_dp)->port));
 }
 
 static void edp_panel_vdd_off_sync(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_digital_port *intel_dig_port =
+		dp_to_dig_port(intel_dp);
+	struct intel_encoder *intel_encoder = &intel_dig_port->base;
+	enum intel_display_power_domain power_domain;
 	u32 pp;
 	u32 pp_stat_reg, pp_ctrl_reg;
 
-	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
+	lockdep_assert_held(&dev_priv->pps_mutex);
 
-	if (!intel_dp->want_panel_vdd && edp_have_panel_vdd(intel_dp)) {
-		struct intel_digital_port *intel_dig_port =
-						dp_to_dig_port(intel_dp);
-		struct intel_encoder *intel_encoder = &intel_dig_port->base;
-		enum intel_display_power_domain power_domain;
+	WARN_ON(intel_dp->want_panel_vdd);
 
-		DRM_DEBUG_KMS("Turning eDP VDD off\n");
+	if (!edp_have_panel_vdd(intel_dp))
+		return;
 
-		pp = ironlake_get_pp_control(intel_dp);
-		pp &= ~EDP_FORCE_VDD;
+	DRM_DEBUG_KMS("Turning eDP port %c VDD off\n",
+		      port_name(intel_dig_port->port));
 
-		pp_ctrl_reg = _pp_ctrl_reg(intel_dp);
-		pp_stat_reg = _pp_stat_reg(intel_dp);
+	pp = ironlake_get_pp_control(intel_dp);
+	pp &= ~EDP_FORCE_VDD;
 
-		I915_WRITE(pp_ctrl_reg, pp);
-		POSTING_READ(pp_ctrl_reg);
+	pp_ctrl_reg = _pp_ctrl_reg(intel_dp);
+	pp_stat_reg = _pp_stat_reg(intel_dp);
 
-		/* Make sure sequencer is idle before allowing subsequent activity */
-		DRM_DEBUG_KMS("PP_STATUS: 0x%08x PP_CONTROL: 0x%08x\n",
-		I915_READ(pp_stat_reg), I915_READ(pp_ctrl_reg));
+	I915_WRITE(pp_ctrl_reg, pp);
+	POSTING_READ(pp_ctrl_reg);
 
-		if ((pp & POWER_TARGET_ON) == 0)
-			intel_dp->last_power_cycle = jiffies;
+	/* Make sure sequencer is idle before allowing subsequent activity */
+	DRM_DEBUG_KMS("PP_STATUS: 0x%08x PP_CONTROL: 0x%08x\n",
+	I915_READ(pp_stat_reg), I915_READ(pp_ctrl_reg));
 
-		power_domain = intel_display_port_power_domain(intel_encoder);
-		intel_display_power_put(dev_priv, power_domain);
-	}
+	if ((pp & POWER_TARGET_ON) == 0)
+		intel_dp->last_power_cycle = jiffies;
+
+	power_domain = intel_display_port_power_domain(intel_encoder);
+	intel_display_power_put(dev_priv, power_domain);
 }
 
 static void edp_panel_vdd_work(struct work_struct *__work)
 {
 	struct intel_dp *intel_dp = container_of(to_delayed_work(__work),
 						 struct intel_dp, panel_vdd_work);
-	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 
-	drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
-	edp_panel_vdd_off_sync(intel_dp);
-	drm_modeset_unlock(&dev->mode_config.connection_mutex);
+	pps_lock(intel_dp);
+	if (!intel_dp->want_panel_vdd)
+		edp_panel_vdd_off_sync(intel_dp);
+	pps_unlock(intel_dp);
 }
 
 static void edp_panel_vdd_schedule_off(struct intel_dp *intel_dp)
@@ -1298,12 +1627,23 @@ static void edp_panel_vdd_schedule_off(struct intel_dp *intel_dp)
 	schedule_delayed_work(&intel_dp->panel_vdd_work, delay);
 }
 
+/*
+ * Must be paired with edp_panel_vdd_on().
+ * Must hold pps_mutex around the whole on/off sequence.
+ * Can be nested with intel_edp_panel_vdd_{on,off}() calls.
+ */
 static void edp_panel_vdd_off(struct intel_dp *intel_dp, bool sync)
 {
+	struct drm_i915_private *dev_priv =
+		intel_dp_to_dev(intel_dp)->dev_private;
+
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
 	if (!is_edp(intel_dp))
 		return;
 
-	WARN(!intel_dp->want_panel_vdd, "eDP VDD not forced on");
+	WARN(!intel_dp->want_panel_vdd, "eDP port %c VDD not forced on",
+	     port_name(dp_to_dig_port(intel_dp)->port));
 
 	intel_dp->want_panel_vdd = false;
 
@@ -1313,22 +1653,25 @@ static void edp_panel_vdd_off(struct intel_dp *intel_dp, bool sync)
 		edp_panel_vdd_schedule_off(intel_dp);
 }
 
-void intel_edp_panel_on(struct intel_dp *intel_dp)
+static void edp_panel_on(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 pp;
 	u32 pp_ctrl_reg;
 
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
 	if (!is_edp(intel_dp))
 		return;
 
-	DRM_DEBUG_KMS("Turn eDP power on\n");
+	DRM_DEBUG_KMS("Turn eDP port %c panel power on\n",
+		      port_name(dp_to_dig_port(intel_dp)->port));
 
-	if (edp_have_panel_power(intel_dp)) {
-		DRM_DEBUG_KMS("eDP power already on\n");
+	if (WARN(edp_have_panel_power(intel_dp),
+		 "eDP port %c panel power already on\n",
+		 port_name(dp_to_dig_port(intel_dp)->port)))
 		return;
-	}
 
 	wait_panel_power_cycle(intel_dp);
 
@@ -1358,7 +1701,18 @@ void intel_edp_panel_on(struct intel_dp *intel_dp)
 	}
 }
 
-void intel_edp_panel_off(struct intel_dp *intel_dp)
+void intel_edp_panel_on(struct intel_dp *intel_dp)
+{
+	if (!is_edp(intel_dp))
+		return;
+
+	pps_lock(intel_dp);
+	edp_panel_on(intel_dp);
+	pps_unlock(intel_dp);
+}
+
+
+static void edp_panel_off(struct intel_dp *intel_dp)
 {
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
 	struct intel_encoder *intel_encoder = &intel_dig_port->base;
@@ -1368,12 +1722,16 @@ void intel_edp_panel_off(struct intel_dp *intel_dp)
 	u32 pp;
 	u32 pp_ctrl_reg;
 
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
 	if (!is_edp(intel_dp))
 		return;
 
-	DRM_DEBUG_KMS("Turn eDP power off\n");
+	DRM_DEBUG_KMS("Turn eDP port %c panel power off\n",
+		      port_name(dp_to_dig_port(intel_dp)->port));
 
-	WARN(!intel_dp->want_panel_vdd, "Need VDD to turn off panel\n");
+	WARN(!intel_dp->want_panel_vdd, "Need eDP port %c VDD to turn off panel\n",
+	     port_name(dp_to_dig_port(intel_dp)->port));
 
 	pp = ironlake_get_pp_control(intel_dp);
 	/* We need to switch off panel power _and_ force vdd, for otherwise some
@@ -1396,7 +1754,18 @@ void intel_edp_panel_off(struct intel_dp *intel_dp)
 	intel_display_power_put(dev_priv, power_domain);
 }
 
-void intel_edp_backlight_on(struct intel_dp *intel_dp)
+void intel_edp_panel_off(struct intel_dp *intel_dp)
+{
+	if (!is_edp(intel_dp))
+		return;
+
+	pps_lock(intel_dp);
+	edp_panel_off(intel_dp);
+	pps_unlock(intel_dp);
+}
+
+/* Enable backlight in the panel power control. */
+static void _intel_edp_backlight_on(struct intel_dp *intel_dp)
 {
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
 	struct drm_device *dev = intel_dig_port->base.base.dev;
@@ -1404,13 +1773,6 @@ void intel_edp_backlight_on(struct intel_dp *intel_dp)
 	u32 pp;
 	u32 pp_ctrl_reg;
 
-	if (!is_edp(intel_dp))
-		return;
-
-	DRM_DEBUG_KMS("\n");
-
-	intel_panel_enable_backlight(intel_dp->attached_connector);
-
 	/*
 	 * If we enable the backlight right away following a panel power
 	 * on, we may see slight flicker as the panel syncs with the eDP
@@ -1418,6 +1780,9 @@ void intel_edp_backlight_on(struct intel_dp *intel_dp)
 	 * allowing it to appear.
 	 */
 	wait_backlight_on(intel_dp);
+
+	pps_lock(intel_dp);
+
 	pp = ironlake_get_pp_control(intel_dp);
 	pp |= EDP_BLC_ENABLE;
 
@@ -1425,9 +1790,24 @@ void intel_edp_backlight_on(struct intel_dp *intel_dp)
 
 	I915_WRITE(pp_ctrl_reg, pp);
 	POSTING_READ(pp_ctrl_reg);
+
+	pps_unlock(intel_dp);
 }
 
-void intel_edp_backlight_off(struct intel_dp *intel_dp)
+/* Enable backlight PWM and backlight PP control. */
+void intel_edp_backlight_on(struct intel_dp *intel_dp)
+{
+	if (!is_edp(intel_dp))
+		return;
+
+	DRM_DEBUG_KMS("\n");
+
+	intel_panel_enable_backlight(intel_dp->attached_connector);
+	_intel_edp_backlight_on(intel_dp);
+}
+
+/* Disable backlight in the panel power control. */
+static void _intel_edp_backlight_off(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1437,7 +1817,8 @@ void intel_edp_backlight_off(struct intel_dp *intel_dp)
 	if (!is_edp(intel_dp))
 		return;
 
-	DRM_DEBUG_KMS("\n");
+	pps_lock(intel_dp);
+
 	pp = ironlake_get_pp_control(intel_dp);
 	pp &= ~EDP_BLC_ENABLE;
 
@@ -1445,13 +1826,51 @@ void intel_edp_backlight_off(struct intel_dp *intel_dp)
 
 	I915_WRITE(pp_ctrl_reg, pp);
 	POSTING_READ(pp_ctrl_reg);
-	intel_dp->last_backlight_off = jiffies;
 
+	pps_unlock(intel_dp);
+
+	intel_dp->last_backlight_off = jiffies;
 	edp_wait_backlight_off(intel_dp);
+}
+
+/* Disable backlight PP control and backlight PWM. */
+void intel_edp_backlight_off(struct intel_dp *intel_dp)
+{
+	if (!is_edp(intel_dp))
+		return;
 
+	DRM_DEBUG_KMS("\n");
+
+	_intel_edp_backlight_off(intel_dp);
 	intel_panel_disable_backlight(intel_dp->attached_connector);
 }
 
+/*
+ * Hook for controlling the panel power control backlight through the bl_power
+ * sysfs attribute. Take care to handle multiple calls.
+ */
+static void intel_edp_backlight_power(struct intel_connector *connector,
+				      bool enable)
+{
+	struct intel_dp *intel_dp = intel_attached_dp(&connector->base);
+	bool is_enabled;
+
+	pps_lock(intel_dp);
+	is_enabled = ironlake_get_pp_control(intel_dp) & EDP_BLC_ENABLE;
+	pps_unlock(intel_dp);
+
+	if (is_enabled == enable)
+		return;
+
+	DRM_DEBUG_KMS("panel power control backlight %s\n",
+		      enable ? "enable" : "disable");
+
+	if (enable)
+		_intel_edp_backlight_on(intel_dp);
+	else
+		_intel_edp_backlight_off(intel_dp);
+}
+
 static void ironlake_edp_pll_on(struct intel_dp *intel_dp)
 {
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
@@ -1515,8 +1934,6 @@ void intel_dp_sink_dpms(struct intel_dp *intel_dp, int mode)
 	if (mode != DRM_MODE_DPMS_ON) {
 		ret = drm_dp_dpcd_writeb(&intel_dp->aux, DP_SET_POWER,
 					 DP_SET_POWER_D3);
-		if (ret != 1)
-			DRM_DEBUG_DRIVER("failed to write sink power state\n");
 	} else {
 		/*
 		 * When turning on, we need to retry for 1ms to give the sink
@@ -1530,6 +1947,10 @@ void intel_dp_sink_dpms(struct intel_dp *intel_dp, int mode)
 			msleep(1);
 		}
 	}
+
+	if (ret != 1)
+		DRM_DEBUG_KMS("failed to %s sink power state\n",
+			      mode == DRM_MODE_DPMS_ON ? "enable" : "disable");
 }
 
 static bool intel_dp_get_hw_state(struct intel_encoder *encoder,
@@ -1543,7 +1964,7 @@ static bool intel_dp_get_hw_state(struct intel_encoder *encoder,
 	u32 tmp;
 
 	power_domain = intel_display_port_power_domain(encoder);
-	if (!intel_display_power_enabled(dev_priv, power_domain))
+	if (!intel_display_power_is_enabled(dev_priv, power_domain))
 		return false;
 
 	tmp = I915_READ(intel_dp->output_reg);
@@ -1576,7 +1997,7 @@ static bool intel_dp_get_hw_state(struct intel_encoder *encoder,
 			return true;
 		}
 
-		for_each_pipe(i) {
+		for_each_pipe(dev_priv, i) {
 			trans_dp = I915_READ(TRANS_DP_CTL(i));
 			if ((trans_dp & TRANS_DP_PORT_SEL_MASK) == trans_sel) {
 				*pipe = i;
@@ -1675,369 +2096,14 @@ static void intel_dp_get_config(struct intel_encoder *encoder,
 	}
 }
 
-static bool is_edp_psr(struct intel_dp *intel_dp)
-{
-	return intel_dp->psr_dpcd[0] & DP_PSR_IS_SUPPORTED;
-}
-
-static bool intel_edp_is_psr_enabled(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	if (!HAS_PSR(dev))
-		return false;
-
-	return I915_READ(EDP_PSR_CTL(dev)) & EDP_PSR_ENABLE;
-}
-
-static void intel_edp_psr_write_vsc(struct intel_dp *intel_dp,
-				    struct edp_vsc_psr *vsc_psr)
-{
-	struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp);
-	struct drm_device *dev = dig_port->base.base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_crtc *crtc = to_intel_crtc(dig_port->base.base.crtc);
-	u32 ctl_reg = HSW_TVIDEO_DIP_CTL(crtc->config.cpu_transcoder);
-	u32 data_reg = HSW_TVIDEO_DIP_VSC_DATA(crtc->config.cpu_transcoder);
-	uint32_t *data = (uint32_t *) vsc_psr;
-	unsigned int i;
-
-	/* As per BSPec (Pipe Video Data Island Packet), we need to disable
-	   the video DIP being updated before program video DIP data buffer
-	   registers for DIP being updated. */
-	I915_WRITE(ctl_reg, 0);
-	POSTING_READ(ctl_reg);
-
-	for (i = 0; i < VIDEO_DIP_VSC_DATA_SIZE; i += 4) {
-		if (i < sizeof(struct edp_vsc_psr))
-			I915_WRITE(data_reg + i, *data++);
-		else
-			I915_WRITE(data_reg + i, 0);
-	}
-
-	I915_WRITE(ctl_reg, VIDEO_DIP_ENABLE_VSC_HSW);
-	POSTING_READ(ctl_reg);
-}
-
-static void intel_edp_psr_setup(struct intel_dp *intel_dp)
-{
-	struct drm_device *dev = intel_dp_to_dev(intel_dp);
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct edp_vsc_psr psr_vsc;
-
-	/* Prepare VSC packet as per EDP 1.3 spec, Table 3.10 */
-	memset(&psr_vsc, 0, sizeof(psr_vsc));
-	psr_vsc.sdp_header.HB0 = 0;
-	psr_vsc.sdp_header.HB1 = 0x7;
-	psr_vsc.sdp_header.HB2 = 0x2;
-	psr_vsc.sdp_header.HB3 = 0x8;
-	intel_edp_psr_write_vsc(intel_dp, &psr_vsc);
-
-	/* Avoid continuous PSR exit by masking memup and hpd */
-	I915_WRITE(EDP_PSR_DEBUG_CTL(dev), EDP_PSR_DEBUG_MASK_MEMUP |
-		   EDP_PSR_DEBUG_MASK_HPD | EDP_PSR_DEBUG_MASK_LPSP);
-}
-
-static void intel_edp_psr_enable_sink(struct intel_dp *intel_dp)
-{
-	struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp);
-	struct drm_device *dev = dig_port->base.base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	uint32_t aux_clock_divider;
-	int precharge = 0x3;
-	int msg_size = 5;       /* Header(4) + Message(1) */
-	bool only_standby = false;
-
-	aux_clock_divider = intel_dp->get_aux_clock_divider(intel_dp, 0);
-
-	if (IS_BROADWELL(dev) && dig_port->port != PORT_A)
-		only_standby = true;
-
-	/* Enable PSR in sink */
-	if (intel_dp->psr_dpcd[1] & DP_PSR_NO_TRAIN_ON_EXIT || only_standby)
-		drm_dp_dpcd_writeb(&intel_dp->aux, DP_PSR_EN_CFG,
-				   DP_PSR_ENABLE & ~DP_PSR_MAIN_LINK_ACTIVE);
-	else
-		drm_dp_dpcd_writeb(&intel_dp->aux, DP_PSR_EN_CFG,
-				   DP_PSR_ENABLE | DP_PSR_MAIN_LINK_ACTIVE);
-
-	/* Setup AUX registers */
-	I915_WRITE(EDP_PSR_AUX_DATA1(dev), EDP_PSR_DPCD_COMMAND);
-	I915_WRITE(EDP_PSR_AUX_DATA2(dev), EDP_PSR_DPCD_NORMAL_OPERATION);
-	I915_WRITE(EDP_PSR_AUX_CTL(dev),
-		   DP_AUX_CH_CTL_TIME_OUT_400us |
-		   (msg_size << DP_AUX_CH_CTL_MESSAGE_SIZE_SHIFT) |
-		   (precharge << DP_AUX_CH_CTL_PRECHARGE_2US_SHIFT) |
-		   (aux_clock_divider << DP_AUX_CH_CTL_BIT_CLOCK_2X_SHIFT));
-}
-
-static void intel_edp_psr_enable_source(struct intel_dp *intel_dp)
-{
-	struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp);
-	struct drm_device *dev = dig_port->base.base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	uint32_t max_sleep_time = 0x1f;
-	uint32_t idle_frames = 1;
-	uint32_t val = 0x0;
-	const uint32_t link_entry_time = EDP_PSR_MIN_LINK_ENTRY_TIME_8_LINES;
-	bool only_standby = false;
-
-	if (IS_BROADWELL(dev) && dig_port->port != PORT_A)
-		only_standby = true;
-
-	if (intel_dp->psr_dpcd[1] & DP_PSR_NO_TRAIN_ON_EXIT || only_standby) {
-		val |= EDP_PSR_LINK_STANDBY;
-		val |= EDP_PSR_TP2_TP3_TIME_0us;
-		val |= EDP_PSR_TP1_TIME_0us;
-		val |= EDP_PSR_SKIP_AUX_EXIT;
-		val |= IS_BROADWELL(dev) ? BDW_PSR_SINGLE_FRAME : 0;
-	} else
-		val |= EDP_PSR_LINK_DISABLE;
-
-	I915_WRITE(EDP_PSR_CTL(dev), val |
-		   (IS_BROADWELL(dev) ? 0 : link_entry_time) |
-		   max_sleep_time << EDP_PSR_MAX_SLEEP_TIME_SHIFT |
-		   idle_frames << EDP_PSR_IDLE_FRAME_SHIFT |
-		   EDP_PSR_ENABLE);
-}
-
-static bool intel_edp_psr_match_conditions(struct intel_dp *intel_dp)
-{
-	struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp);
-	struct drm_device *dev = dig_port->base.base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_crtc *crtc = dig_port->base.base.crtc;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-
-	lockdep_assert_held(&dev_priv->psr.lock);
-	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
-	WARN_ON(!drm_modeset_is_locked(&crtc->mutex));
-
-	dev_priv->psr.source_ok = false;
-
-	if (IS_HASWELL(dev) && dig_port->port != PORT_A) {
-		DRM_DEBUG_KMS("HSW ties PSR to DDI A (eDP)\n");
-		return false;
-	}
-
-	if (!i915.enable_psr) {
-		DRM_DEBUG_KMS("PSR disable by flag\n");
-		return false;
-	}
-
-	/* Below limitations aren't valid for Broadwell */
-	if (IS_BROADWELL(dev))
-		goto out;
-
-	if (I915_READ(HSW_STEREO_3D_CTL(intel_crtc->config.cpu_transcoder)) &
-	    S3D_ENABLE) {
-		DRM_DEBUG_KMS("PSR condition failed: Stereo 3D is Enabled\n");
-		return false;
-	}
-
-	if (intel_crtc->config.adjusted_mode.flags & DRM_MODE_FLAG_INTERLACE) {
-		DRM_DEBUG_KMS("PSR condition failed: Interlaced is Enabled\n");
-		return false;
-	}
-
- out:
-	dev_priv->psr.source_ok = true;
-	return true;
-}
-
-static void intel_edp_psr_do_enable(struct intel_dp *intel_dp)
-{
-	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
-	struct drm_device *dev = intel_dig_port->base.base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	WARN_ON(I915_READ(EDP_PSR_CTL(dev)) & EDP_PSR_ENABLE);
-	WARN_ON(dev_priv->psr.active);
-	lockdep_assert_held(&dev_priv->psr.lock);
-
-	/* Enable PSR on the panel */
-	intel_edp_psr_enable_sink(intel_dp);
-
-	/* Enable PSR on the host */
-	intel_edp_psr_enable_source(intel_dp);
-
-	dev_priv->psr.active = true;
-}
-
-void intel_edp_psr_enable(struct intel_dp *intel_dp)
-{
-	struct drm_device *dev = intel_dp_to_dev(intel_dp);
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	if (!HAS_PSR(dev)) {
-		DRM_DEBUG_KMS("PSR not supported on this platform\n");
-		return;
-	}
-
-	if (!is_edp_psr(intel_dp)) {
-		DRM_DEBUG_KMS("PSR not supported by this panel\n");
-		return;
-	}
-
-	mutex_lock(&dev_priv->psr.lock);
-	if (dev_priv->psr.enabled) {
-		DRM_DEBUG_KMS("PSR already in use\n");
-		mutex_unlock(&dev_priv->psr.lock);
-		return;
-	}
-
-	dev_priv->psr.busy_frontbuffer_bits = 0;
-
-	/* Setup PSR once */
-	intel_edp_psr_setup(intel_dp);
-
-	if (intel_edp_psr_match_conditions(intel_dp))
-		dev_priv->psr.enabled = intel_dp;
-	mutex_unlock(&dev_priv->psr.lock);
-}
-
-void intel_edp_psr_disable(struct intel_dp *intel_dp)
-{
-	struct drm_device *dev = intel_dp_to_dev(intel_dp);
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	mutex_lock(&dev_priv->psr.lock);
-	if (!dev_priv->psr.enabled) {
-		mutex_unlock(&dev_priv->psr.lock);
-		return;
-	}
-
-	if (dev_priv->psr.active) {
-		I915_WRITE(EDP_PSR_CTL(dev),
-			   I915_READ(EDP_PSR_CTL(dev)) & ~EDP_PSR_ENABLE);
-
-		/* Wait till PSR is idle */
-		if (_wait_for((I915_READ(EDP_PSR_STATUS_CTL(dev)) &
-			       EDP_PSR_STATUS_STATE_MASK) == 0, 2000, 10))
-			DRM_ERROR("Timed out waiting for PSR Idle State\n");
-
-		dev_priv->psr.active = false;
-	} else {
-		WARN_ON(I915_READ(EDP_PSR_CTL(dev)) & EDP_PSR_ENABLE);
-	}
-
-	dev_priv->psr.enabled = NULL;
-	mutex_unlock(&dev_priv->psr.lock);
-
-	cancel_delayed_work_sync(&dev_priv->psr.work);
-}
-
-static void intel_edp_psr_work(struct work_struct *work)
-{
-	struct drm_i915_private *dev_priv =
-		container_of(work, typeof(*dev_priv), psr.work.work);
-	struct intel_dp *intel_dp = dev_priv->psr.enabled;
-
-	mutex_lock(&dev_priv->psr.lock);
-	intel_dp = dev_priv->psr.enabled;
-
-	if (!intel_dp)
-		goto unlock;
-
-	/*
-	 * The delayed work can race with an invalidate hence we need to
-	 * recheck. Since psr_flush first clears this and then reschedules we
-	 * won't ever miss a flush when bailing out here.
-	 */
-	if (dev_priv->psr.busy_frontbuffer_bits)
-		goto unlock;
-
-	intel_edp_psr_do_enable(intel_dp);
-unlock:
-	mutex_unlock(&dev_priv->psr.lock);
-}
-
-static void intel_edp_psr_do_exit(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	if (dev_priv->psr.active) {
-		u32 val = I915_READ(EDP_PSR_CTL(dev));
-
-		WARN_ON(!(val & EDP_PSR_ENABLE));
-
-		I915_WRITE(EDP_PSR_CTL(dev), val & ~EDP_PSR_ENABLE);
-
-		dev_priv->psr.active = false;
-	}
-
-}
-
-void intel_edp_psr_invalidate(struct drm_device *dev,
-			      unsigned frontbuffer_bits)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_crtc *crtc;
-	enum pipe pipe;
-
-	mutex_lock(&dev_priv->psr.lock);
-	if (!dev_priv->psr.enabled) {
-		mutex_unlock(&dev_priv->psr.lock);
-		return;
-	}
-
-	crtc = dp_to_dig_port(dev_priv->psr.enabled)->base.base.crtc;
-	pipe = to_intel_crtc(crtc)->pipe;
-
-	intel_edp_psr_do_exit(dev);
-
-	frontbuffer_bits &= INTEL_FRONTBUFFER_ALL_MASK(pipe);
-
-	dev_priv->psr.busy_frontbuffer_bits |= frontbuffer_bits;
-	mutex_unlock(&dev_priv->psr.lock);
-}
-
-void intel_edp_psr_flush(struct drm_device *dev,
-			 unsigned frontbuffer_bits)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_crtc *crtc;
-	enum pipe pipe;
-
-	mutex_lock(&dev_priv->psr.lock);
-	if (!dev_priv->psr.enabled) {
-		mutex_unlock(&dev_priv->psr.lock);
-		return;
-	}
-
-	crtc = dp_to_dig_port(dev_priv->psr.enabled)->base.base.crtc;
-	pipe = to_intel_crtc(crtc)->pipe;
-	dev_priv->psr.busy_frontbuffer_bits &= ~frontbuffer_bits;
-
-	/*
-	 * On Haswell sprite plane updates don't result in a psr invalidating
-	 * signal in the hardware. Which means we need to manually fake this in
-	 * software for all flushes, not just when we've seen a preceding
-	 * invalidation through frontbuffer rendering.
-	 */
-	if (IS_HASWELL(dev) &&
-	    (frontbuffer_bits & INTEL_FRONTBUFFER_SPRITE(pipe)))
-		intel_edp_psr_do_exit(dev);
-
-	if (!dev_priv->psr.active && !dev_priv->psr.busy_frontbuffer_bits)
-		schedule_delayed_work(&dev_priv->psr.work,
-				      msecs_to_jiffies(100));
-	mutex_unlock(&dev_priv->psr.lock);
-}
-
-void intel_edp_psr_init(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	INIT_DELAYED_WORK(&dev_priv->psr.work, intel_edp_psr_work);
-	mutex_init(&dev_priv->psr.lock);
-}
-
 static void intel_disable_dp(struct intel_encoder *encoder)
 {
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
-	enum port port = dp_to_dig_port(intel_dp)->port;
 	struct drm_device *dev = encoder->base.dev;
+	struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
+
+	if (crtc->config.has_audio)
+		intel_audio_codec_disable(encoder);
 
 	/* Make sure the panel is off before trying to change the mode. But also
 	 * ensure that we have vdd while we switch off the panel. */
@@ -2046,21 +2112,19 @@ static void intel_disable_dp(struct intel_encoder *encoder)
 	intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_OFF);
 	intel_edp_panel_off(intel_dp);
 
-	/* cpu edp my only be disable _after_ the cpu pipe/plane is disabled. */
-	if (!(port == PORT_A || IS_VALLEYVIEW(dev)))
+	/* disable the port before the pipe on g4x */
+	if (INTEL_INFO(dev)->gen < 5)
 		intel_dp_link_down(intel_dp);
 }
 
-static void g4x_post_disable_dp(struct intel_encoder *encoder)
+static void ilk_post_disable_dp(struct intel_encoder *encoder)
 {
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 	enum port port = dp_to_dig_port(intel_dp)->port;
 
-	if (port != PORT_A)
-		return;
-
 	intel_dp_link_down(intel_dp);
-	ironlake_edp_pll_off(intel_dp);
+	if (port == PORT_A)
+		ironlake_edp_pll_off(intel_dp);
 }
 
 static void vlv_post_disable_dp(struct intel_encoder *encoder)
@@ -2106,23 +2170,150 @@ static void chv_post_disable_dp(struct intel_encoder *encoder)
 	mutex_unlock(&dev_priv->dpio_lock);
 }
 
+static void
+_intel_dp_set_link_train(struct intel_dp *intel_dp,
+			 uint32_t *DP,
+			 uint8_t dp_train_pat)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = intel_dig_port->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum port port = intel_dig_port->port;
+
+	if (HAS_DDI(dev)) {
+		uint32_t temp = I915_READ(DP_TP_CTL(port));
+
+		if (dp_train_pat & DP_LINK_SCRAMBLING_DISABLE)
+			temp |= DP_TP_CTL_SCRAMBLE_DISABLE;
+		else
+			temp &= ~DP_TP_CTL_SCRAMBLE_DISABLE;
+
+		temp &= ~DP_TP_CTL_LINK_TRAIN_MASK;
+		switch (dp_train_pat & DP_TRAINING_PATTERN_MASK) {
+		case DP_TRAINING_PATTERN_DISABLE:
+			temp |= DP_TP_CTL_LINK_TRAIN_NORMAL;
+
+			break;
+		case DP_TRAINING_PATTERN_1:
+			temp |= DP_TP_CTL_LINK_TRAIN_PAT1;
+			break;
+		case DP_TRAINING_PATTERN_2:
+			temp |= DP_TP_CTL_LINK_TRAIN_PAT2;
+			break;
+		case DP_TRAINING_PATTERN_3:
+			temp |= DP_TP_CTL_LINK_TRAIN_PAT3;
+			break;
+		}
+		I915_WRITE(DP_TP_CTL(port), temp);
+
+	} else if (HAS_PCH_CPT(dev) && (IS_GEN7(dev) || port != PORT_A)) {
+		*DP &= ~DP_LINK_TRAIN_MASK_CPT;
+
+		switch (dp_train_pat & DP_TRAINING_PATTERN_MASK) {
+		case DP_TRAINING_PATTERN_DISABLE:
+			*DP |= DP_LINK_TRAIN_OFF_CPT;
+			break;
+		case DP_TRAINING_PATTERN_1:
+			*DP |= DP_LINK_TRAIN_PAT_1_CPT;
+			break;
+		case DP_TRAINING_PATTERN_2:
+			*DP |= DP_LINK_TRAIN_PAT_2_CPT;
+			break;
+		case DP_TRAINING_PATTERN_3:
+			DRM_ERROR("DP training pattern 3 not supported\n");
+			*DP |= DP_LINK_TRAIN_PAT_2_CPT;
+			break;
+		}
+
+	} else {
+		if (IS_CHERRYVIEW(dev))
+			*DP &= ~DP_LINK_TRAIN_MASK_CHV;
+		else
+			*DP &= ~DP_LINK_TRAIN_MASK;
+
+		switch (dp_train_pat & DP_TRAINING_PATTERN_MASK) {
+		case DP_TRAINING_PATTERN_DISABLE:
+			*DP |= DP_LINK_TRAIN_OFF;
+			break;
+		case DP_TRAINING_PATTERN_1:
+			*DP |= DP_LINK_TRAIN_PAT_1;
+			break;
+		case DP_TRAINING_PATTERN_2:
+			*DP |= DP_LINK_TRAIN_PAT_2;
+			break;
+		case DP_TRAINING_PATTERN_3:
+			if (IS_CHERRYVIEW(dev)) {
+				*DP |= DP_LINK_TRAIN_PAT_3_CHV;
+			} else {
+				DRM_ERROR("DP training pattern 3 not supported\n");
+				*DP |= DP_LINK_TRAIN_PAT_2;
+			}
+			break;
+		}
+	}
+}
+
+static void intel_dp_enable_port(struct intel_dp *intel_dp)
+{
+	struct drm_device *dev = intel_dp_to_dev(intel_dp);
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	/* enable with pattern 1 (as per spec) */
+	_intel_dp_set_link_train(intel_dp, &intel_dp->DP,
+				 DP_TRAINING_PATTERN_1);
+
+	I915_WRITE(intel_dp->output_reg, intel_dp->DP);
+	POSTING_READ(intel_dp->output_reg);
+
+	/*
+	 * Magic for VLV/CHV. We _must_ first set up the register
+	 * without actually enabling the port, and then do another
+	 * write to enable the port. Otherwise link training will
+	 * fail when the power sequencer is freshly used for this port.
+	 */
+	intel_dp->DP |= DP_PORT_EN;
+
+	I915_WRITE(intel_dp->output_reg, intel_dp->DP);
+	POSTING_READ(intel_dp->output_reg);
+}
+
 static void intel_enable_dp(struct intel_encoder *encoder)
 {
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
 	uint32_t dp_reg = I915_READ(intel_dp->output_reg);
 
 	if (WARN_ON(dp_reg & DP_PORT_EN))
 		return;
 
-	intel_edp_panel_vdd_on(intel_dp);
+	pps_lock(intel_dp);
+
+	if (IS_VALLEYVIEW(dev))
+		vlv_init_panel_power_sequencer(intel_dp);
+
+	intel_dp_enable_port(intel_dp);
+
+	edp_panel_vdd_on(intel_dp);
+	edp_panel_on(intel_dp);
+	edp_panel_vdd_off(intel_dp, true);
+
+	pps_unlock(intel_dp);
+
+	if (IS_VALLEYVIEW(dev))
+		vlv_wait_port_ready(dev_priv, dp_to_dig_port(intel_dp));
+
 	intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_ON);
 	intel_dp_start_link_train(intel_dp);
-	intel_edp_panel_on(intel_dp);
-	edp_panel_vdd_off(intel_dp, true);
 	intel_dp_complete_link_train(intel_dp);
 	intel_dp_stop_link_train(intel_dp);
+
+	if (crtc->config.has_audio) {
+		DRM_DEBUG_DRIVER("Enabling DP audio on pipe %c\n",
+				 pipe_name(crtc->pipe));
+		intel_audio_codec_enable(encoder);
+	}
 }
 
 static void g4x_enable_dp(struct intel_encoder *encoder)
@@ -2154,6 +2345,110 @@ static void g4x_pre_enable_dp(struct intel_encoder *encoder)
 	}
 }
 
+static void vlv_detach_power_sequencer(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct drm_i915_private *dev_priv = intel_dig_port->base.base.dev->dev_private;
+	enum pipe pipe = intel_dp->pps_pipe;
+	int pp_on_reg = VLV_PIPE_PP_ON_DELAYS(pipe);
+
+	edp_panel_vdd_off_sync(intel_dp);
+
+	/*
+	 * VLV seems to get confused when multiple power seqeuencers
+	 * have the same port selected (even if only one has power/vdd
+	 * enabled). The failure manifests as vlv_wait_port_ready() failing
+	 * CHV on the other hand doesn't seem to mind having the same port
+	 * selected in multiple power seqeuencers, but let's clear the
+	 * port select always when logically disconnecting a power sequencer
+	 * from a port.
+	 */
+	DRM_DEBUG_KMS("detaching pipe %c power sequencer from port %c\n",
+		      pipe_name(pipe), port_name(intel_dig_port->port));
+	I915_WRITE(pp_on_reg, 0);
+	POSTING_READ(pp_on_reg);
+
+	intel_dp->pps_pipe = INVALID_PIPE;
+}
+
+static void vlv_steal_power_sequencer(struct drm_device *dev,
+				      enum pipe pipe)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_encoder *encoder;
+
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
+	if (WARN_ON(pipe != PIPE_A && pipe != PIPE_B))
+		return;
+
+	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
+			    base.head) {
+		struct intel_dp *intel_dp;
+		enum port port;
+
+		if (encoder->type != INTEL_OUTPUT_EDP)
+			continue;
+
+		intel_dp = enc_to_intel_dp(&encoder->base);
+		port = dp_to_dig_port(intel_dp)->port;
+
+		if (intel_dp->pps_pipe != pipe)
+			continue;
+
+		DRM_DEBUG_KMS("stealing pipe %c power sequencer from port %c\n",
+			      pipe_name(pipe), port_name(port));
+
+		WARN(encoder->connectors_active,
+		     "stealing pipe %c power sequencer from active eDP port %c\n",
+		     pipe_name(pipe), port_name(port));
+
+		/* make sure vdd is off before we steal it */
+		vlv_detach_power_sequencer(intel_dp);
+	}
+}
+
+static void vlv_init_panel_power_sequencer(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct intel_encoder *encoder = &intel_dig_port->base;
+	struct drm_device *dev = encoder->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
+
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
+	if (!is_edp(intel_dp))
+		return;
+
+	if (intel_dp->pps_pipe == crtc->pipe)
+		return;
+
+	/*
+	 * If another power sequencer was being used on this
+	 * port previously make sure to turn off vdd there while
+	 * we still have control of it.
+	 */
+	if (intel_dp->pps_pipe != INVALID_PIPE)
+		vlv_detach_power_sequencer(intel_dp);
+
+	/*
+	 * We may be stealing the power
+	 * sequencer from another port.
+	 */
+	vlv_steal_power_sequencer(dev, crtc->pipe);
+
+	/* now it's all ours */
+	intel_dp->pps_pipe = crtc->pipe;
+
+	DRM_DEBUG_KMS("initializing pipe %c power sequencer for port %c\n",
+		      pipe_name(intel_dp->pps_pipe), port_name(intel_dig_port->port));
+
+	/* init power sequencer on this pipe and port */
+	intel_dp_init_panel_power_sequencer(dev, intel_dp);
+	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp);
+}
+
 static void vlv_pre_enable_dp(struct intel_encoder *encoder)
 {
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
@@ -2163,7 +2458,6 @@ static void vlv_pre_enable_dp(struct intel_encoder *encoder)
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
 	enum dpio_channel port = vlv_dport_to_channel(dport);
 	int pipe = intel_crtc->pipe;
-	struct edp_power_seq power_seq;
 	u32 val;
 
 	mutex_lock(&dev_priv->dpio_lock);
@@ -2181,16 +2475,7 @@ static void vlv_pre_enable_dp(struct intel_encoder *encoder)
 
 	mutex_unlock(&dev_priv->dpio_lock);
 
-	if (is_edp(intel_dp)) {
-		/* init power sequencer on this pipe and port */
-		intel_dp_init_panel_power_sequencer(dev, intel_dp, &power_seq);
-		intel_dp_init_panel_power_sequencer_registers(dev, intel_dp,
-							      &power_seq);
-	}
-
 	intel_enable_dp(encoder);
-
-	vlv_wait_port_ready(dev_priv, dport);
 }
 
 static void vlv_dp_pre_pll_enable(struct intel_encoder *encoder)
@@ -2229,7 +2514,6 @@ static void chv_pre_enable_dp(struct intel_encoder *encoder)
 	struct intel_digital_port *dport = dp_to_dig_port(intel_dp);
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct edp_power_seq power_seq;
 	struct intel_crtc *intel_crtc =
 		to_intel_crtc(encoder->base.crtc);
 	enum dpio_channel ch = vlv_dport_to_channel(dport);
@@ -2239,6 +2523,15 @@ static void chv_pre_enable_dp(struct intel_encoder *encoder)
 
 	mutex_lock(&dev_priv->dpio_lock);
 
+	/* allow hardware to manage TX FIFO reset source */
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW11(ch));
+	val &= ~DPIO_LANEDESKEW_STRAP_OVRD;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW11(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW11(ch));
+	val &= ~DPIO_LANEDESKEW_STRAP_OVRD;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW11(ch), val);
+
 	/* Deassert soft data lane reset*/
 	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW1(ch));
 	val |= CHV_PCS_REQ_SOFTRESET_EN;
@@ -2274,16 +2567,7 @@ static void chv_pre_enable_dp(struct intel_encoder *encoder)
 
 	mutex_unlock(&dev_priv->dpio_lock);
 
-	if (is_edp(intel_dp)) {
-		/* init power sequencer on this pipe and port */
-		intel_dp_init_panel_power_sequencer(dev, intel_dp, &power_seq);
-		intel_dp_init_panel_power_sequencer_registers(dev, intel_dp,
-							      &power_seq);
-	}
-
 	intel_enable_dp(encoder);
-
-	vlv_wait_port_ready(dev_priv, dport);
 }
 
 static void chv_dp_pre_pll_enable(struct intel_encoder *encoder)
@@ -2297,6 +2581,8 @@ static void chv_dp_pre_pll_enable(struct intel_encoder *encoder)
 	enum pipe pipe = intel_crtc->pipe;
 	u32 val;
 
+	intel_dp_prepare(encoder);
+
 	mutex_lock(&dev_priv->dpio_lock);
 
 	/* program left/right clock distribution */
@@ -2364,6 +2650,13 @@ intel_dp_dpcd_read_wake(struct drm_dp_aux *aux, unsigned int offset,
 	ssize_t ret;
 	int i;
 
+	/*
+	 * Sometime we just get the same incorrect byte repeated
+	 * over the entire buffer. Doing just one throw away read
+	 * initially seems to "solve" it.
+	 */
+	drm_dp_dpcd_read(aux, DP_DPCD_REV, buffer, 1);
+
 	for (i = 0; i < 3; i++) {
 		ret = drm_dp_dpcd_read(aux, offset, buffer, size);
 		if (ret == size)
@@ -2394,14 +2687,16 @@ intel_dp_voltage_max(struct intel_dp *intel_dp)
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	enum port port = dp_to_dig_port(intel_dp)->port;
 
-	if (IS_VALLEYVIEW(dev))
-		return DP_TRAIN_VOLTAGE_SWING_1200;
+	if (INTEL_INFO(dev)->gen >= 9)
+		return DP_TRAIN_VOLTAGE_SWING_LEVEL_2;
+	else if (IS_VALLEYVIEW(dev))
+		return DP_TRAIN_VOLTAGE_SWING_LEVEL_3;
 	else if (IS_GEN7(dev) && port == PORT_A)
-		return DP_TRAIN_VOLTAGE_SWING_800;
+		return DP_TRAIN_VOLTAGE_SWING_LEVEL_2;
 	else if (HAS_PCH_CPT(dev) && port != PORT_A)
-		return DP_TRAIN_VOLTAGE_SWING_1200;
+		return DP_TRAIN_VOLTAGE_SWING_LEVEL_3;
 	else
-		return DP_TRAIN_VOLTAGE_SWING_800;
+		return DP_TRAIN_VOLTAGE_SWING_LEVEL_2;
 }
 
 static uint8_t
@@ -2410,51 +2705,62 @@ intel_dp_pre_emphasis_max(struct intel_dp *intel_dp, uint8_t voltage_swing)
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	enum port port = dp_to_dig_port(intel_dp)->port;
 
-	if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
+	if (INTEL_INFO(dev)->gen >= 9) {
 		switch (voltage_swing & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
-			return DP_TRAIN_PRE_EMPHASIS_9_5;
-		case DP_TRAIN_VOLTAGE_SWING_600:
-			return DP_TRAIN_PRE_EMPHASIS_6;
-		case DP_TRAIN_VOLTAGE_SWING_800:
-			return DP_TRAIN_PRE_EMPHASIS_3_5;
-		case DP_TRAIN_VOLTAGE_SWING_1200:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
+			return DP_TRAIN_PRE_EMPH_LEVEL_3;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
+			return DP_TRAIN_PRE_EMPH_LEVEL_2;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
+			return DP_TRAIN_PRE_EMPH_LEVEL_1;
 		default:
-			return DP_TRAIN_PRE_EMPHASIS_0;
+			return DP_TRAIN_PRE_EMPH_LEVEL_0;
+		}
+	} else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
+		switch (voltage_swing & DP_TRAIN_VOLTAGE_SWING_MASK) {
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
+			return DP_TRAIN_PRE_EMPH_LEVEL_3;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
+			return DP_TRAIN_PRE_EMPH_LEVEL_2;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
+			return DP_TRAIN_PRE_EMPH_LEVEL_1;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_3:
+		default:
+			return DP_TRAIN_PRE_EMPH_LEVEL_0;
 		}
 	} else if (IS_VALLEYVIEW(dev)) {
 		switch (voltage_swing & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
-			return DP_TRAIN_PRE_EMPHASIS_9_5;
-		case DP_TRAIN_VOLTAGE_SWING_600:
-			return DP_TRAIN_PRE_EMPHASIS_6;
-		case DP_TRAIN_VOLTAGE_SWING_800:
-			return DP_TRAIN_PRE_EMPHASIS_3_5;
-		case DP_TRAIN_VOLTAGE_SWING_1200:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
+			return DP_TRAIN_PRE_EMPH_LEVEL_3;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
+			return DP_TRAIN_PRE_EMPH_LEVEL_2;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
+			return DP_TRAIN_PRE_EMPH_LEVEL_1;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_3:
 		default:
-			return DP_TRAIN_PRE_EMPHASIS_0;
+			return DP_TRAIN_PRE_EMPH_LEVEL_0;
 		}
 	} else if (IS_GEN7(dev) && port == PORT_A) {
 		switch (voltage_swing & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
-			return DP_TRAIN_PRE_EMPHASIS_6;
-		case DP_TRAIN_VOLTAGE_SWING_600:
-		case DP_TRAIN_VOLTAGE_SWING_800:
-			return DP_TRAIN_PRE_EMPHASIS_3_5;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
+			return DP_TRAIN_PRE_EMPH_LEVEL_2;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
+			return DP_TRAIN_PRE_EMPH_LEVEL_1;
 		default:
-			return DP_TRAIN_PRE_EMPHASIS_0;
+			return DP_TRAIN_PRE_EMPH_LEVEL_0;
 		}
 	} else {
 		switch (voltage_swing & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
-			return DP_TRAIN_PRE_EMPHASIS_6;
-		case DP_TRAIN_VOLTAGE_SWING_600:
-			return DP_TRAIN_PRE_EMPHASIS_6;
-		case DP_TRAIN_VOLTAGE_SWING_800:
-			return DP_TRAIN_PRE_EMPHASIS_3_5;
-		case DP_TRAIN_VOLTAGE_SWING_1200:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
+			return DP_TRAIN_PRE_EMPH_LEVEL_2;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
+			return DP_TRAIN_PRE_EMPH_LEVEL_2;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
+			return DP_TRAIN_PRE_EMPH_LEVEL_1;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_3:
 		default:
-			return DP_TRAIN_PRE_EMPHASIS_0;
+			return DP_TRAIN_PRE_EMPH_LEVEL_0;
 		}
 	}
 }
@@ -2473,22 +2779,22 @@ static uint32_t intel_vlv_signal_levels(struct intel_dp *intel_dp)
 	int pipe = intel_crtc->pipe;
 
 	switch (train_set & DP_TRAIN_PRE_EMPHASIS_MASK) {
-	case DP_TRAIN_PRE_EMPHASIS_0:
+	case DP_TRAIN_PRE_EMPH_LEVEL_0:
 		preemph_reg_value = 0x0004000;
 		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 			demph_reg_value = 0x2B405555;
 			uniqtranscale_reg_value = 0x552AB83A;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_600:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
 			demph_reg_value = 0x2B404040;
 			uniqtranscale_reg_value = 0x5548B83A;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_800:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
 			demph_reg_value = 0x2B245555;
 			uniqtranscale_reg_value = 0x5560B83A;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_1200:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_3:
 			demph_reg_value = 0x2B405555;
 			uniqtranscale_reg_value = 0x5598DA3A;
 			break;
@@ -2496,18 +2802,18 @@ static uint32_t intel_vlv_signal_levels(struct intel_dp *intel_dp)
 			return 0;
 		}
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_3_5:
+	case DP_TRAIN_PRE_EMPH_LEVEL_1:
 		preemph_reg_value = 0x0002000;
 		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 			demph_reg_value = 0x2B404040;
 			uniqtranscale_reg_value = 0x5552B83A;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_600:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
 			demph_reg_value = 0x2B404848;
 			uniqtranscale_reg_value = 0x5580B83A;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_800:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
 			demph_reg_value = 0x2B404040;
 			uniqtranscale_reg_value = 0x55ADDA3A;
 			break;
@@ -2515,14 +2821,14 @@ static uint32_t intel_vlv_signal_levels(struct intel_dp *intel_dp)
 			return 0;
 		}
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_6:
+	case DP_TRAIN_PRE_EMPH_LEVEL_2:
 		preemph_reg_value = 0x0000000;
 		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 			demph_reg_value = 0x2B305555;
 			uniqtranscale_reg_value = 0x5570B83A;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_600:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
 			demph_reg_value = 0x2B2B4040;
 			uniqtranscale_reg_value = 0x55ADDA3A;
 			break;
@@ -2530,10 +2836,10 @@ static uint32_t intel_vlv_signal_levels(struct intel_dp *intel_dp)
 			return 0;
 		}
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_9_5:
+	case DP_TRAIN_PRE_EMPH_LEVEL_3:
 		preemph_reg_value = 0x0006000;
 		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 			demph_reg_value = 0x1B405555;
 			uniqtranscale_reg_value = 0x55ADDA3A;
 			break;
@@ -2572,21 +2878,21 @@ static uint32_t intel_chv_signal_levels(struct intel_dp *intel_dp)
 	int i;
 
 	switch (train_set & DP_TRAIN_PRE_EMPHASIS_MASK) {
-	case DP_TRAIN_PRE_EMPHASIS_0:
+	case DP_TRAIN_PRE_EMPH_LEVEL_0:
 		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 			deemph_reg_value = 128;
 			margin_reg_value = 52;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_600:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
 			deemph_reg_value = 128;
 			margin_reg_value = 77;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_800:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
 			deemph_reg_value = 128;
 			margin_reg_value = 102;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_1200:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_3:
 			deemph_reg_value = 128;
 			margin_reg_value = 154;
 			/* FIXME extra to set for 1200 */
@@ -2595,17 +2901,17 @@ static uint32_t intel_chv_signal_levels(struct intel_dp *intel_dp)
 			return 0;
 		}
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_3_5:
+	case DP_TRAIN_PRE_EMPH_LEVEL_1:
 		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 			deemph_reg_value = 85;
 			margin_reg_value = 78;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_600:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
 			deemph_reg_value = 85;
 			margin_reg_value = 116;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_800:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
 			deemph_reg_value = 85;
 			margin_reg_value = 154;
 			break;
@@ -2613,13 +2919,13 @@ static uint32_t intel_chv_signal_levels(struct intel_dp *intel_dp)
 			return 0;
 		}
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_6:
+	case DP_TRAIN_PRE_EMPH_LEVEL_2:
 		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 			deemph_reg_value = 64;
 			margin_reg_value = 104;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_600:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
 			deemph_reg_value = 64;
 			margin_reg_value = 154;
 			break;
@@ -2627,9 +2933,9 @@ static uint32_t intel_chv_signal_levels(struct intel_dp *intel_dp)
 			return 0;
 		}
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_9_5:
+	case DP_TRAIN_PRE_EMPH_LEVEL_3:
 		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 			deemph_reg_value = 43;
 			margin_reg_value = 154;
 			break;
@@ -2646,12 +2952,26 @@ static uint32_t intel_chv_signal_levels(struct intel_dp *intel_dp)
 	/* Clear calc init */
 	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW10(ch));
 	val &= ~(DPIO_PCS_SWING_CALC_TX0_TX2 | DPIO_PCS_SWING_CALC_TX1_TX3);
+	val &= ~(DPIO_PCS_TX1DEEMP_MASK | DPIO_PCS_TX2DEEMP_MASK);
+	val |= DPIO_PCS_TX1DEEMP_9P5 | DPIO_PCS_TX2DEEMP_9P5;
 	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW10(ch), val);
 
 	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW10(ch));
 	val &= ~(DPIO_PCS_SWING_CALC_TX0_TX2 | DPIO_PCS_SWING_CALC_TX1_TX3);
+	val &= ~(DPIO_PCS_TX1DEEMP_MASK | DPIO_PCS_TX2DEEMP_MASK);
+	val |= DPIO_PCS_TX1DEEMP_9P5 | DPIO_PCS_TX2DEEMP_9P5;
 	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW10(ch), val);
 
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW9(ch));
+	val &= ~(DPIO_PCS_TX1MARGIN_MASK | DPIO_PCS_TX2MARGIN_MASK);
+	val |= DPIO_PCS_TX1MARGIN_000 | DPIO_PCS_TX2MARGIN_000;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW9(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW9(ch));
+	val &= ~(DPIO_PCS_TX1MARGIN_MASK | DPIO_PCS_TX2MARGIN_MASK);
+	val |= DPIO_PCS_TX1MARGIN_000 | DPIO_PCS_TX2MARGIN_000;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW9(ch), val);
+
 	/* Program swing deemph */
 	for (i = 0; i < 4; i++) {
 		val = vlv_dpio_read(dev_priv, pipe, CHV_TX_DW4(ch, i));
@@ -2663,8 +2983,8 @@ static uint32_t intel_chv_signal_levels(struct intel_dp *intel_dp)
 	/* Program swing margin */
 	for (i = 0; i < 4; i++) {
 		val = vlv_dpio_read(dev_priv, pipe, CHV_TX_DW2(ch, i));
-		val &= ~DPIO_SWING_MARGIN_MASK;
-		val |= margin_reg_value << DPIO_SWING_MARGIN_SHIFT;
+		val &= ~DPIO_SWING_MARGIN000_MASK;
+		val |= margin_reg_value << DPIO_SWING_MARGIN000_SHIFT;
 		vlv_dpio_write(dev_priv, pipe, CHV_TX_DW2(ch, i), val);
 	}
 
@@ -2676,9 +2996,9 @@ static uint32_t intel_chv_signal_levels(struct intel_dp *intel_dp)
 	}
 
 	if (((train_set & DP_TRAIN_PRE_EMPHASIS_MASK)
-			== DP_TRAIN_PRE_EMPHASIS_0) &&
+			== DP_TRAIN_PRE_EMPH_LEVEL_0) &&
 		((train_set & DP_TRAIN_VOLTAGE_SWING_MASK)
-			== DP_TRAIN_VOLTAGE_SWING_1200)) {
+			== DP_TRAIN_VOLTAGE_SWING_LEVEL_3)) {
 
 		/*
 		 * The document said it needs to set bit 27 for ch0 and bit 26
@@ -2757,32 +3077,32 @@ intel_gen4_signal_levels(uint8_t train_set)
 	uint32_t	signal_levels = 0;
 
 	switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-	case DP_TRAIN_VOLTAGE_SWING_400:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 	default:
 		signal_levels |= DP_VOLTAGE_0_4;
 		break;
-	case DP_TRAIN_VOLTAGE_SWING_600:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
 		signal_levels |= DP_VOLTAGE_0_6;
 		break;
-	case DP_TRAIN_VOLTAGE_SWING_800:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
 		signal_levels |= DP_VOLTAGE_0_8;
 		break;
-	case DP_TRAIN_VOLTAGE_SWING_1200:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_3:
 		signal_levels |= DP_VOLTAGE_1_2;
 		break;
 	}
 	switch (train_set & DP_TRAIN_PRE_EMPHASIS_MASK) {
-	case DP_TRAIN_PRE_EMPHASIS_0:
+	case DP_TRAIN_PRE_EMPH_LEVEL_0:
 	default:
 		signal_levels |= DP_PRE_EMPHASIS_0;
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_3_5:
+	case DP_TRAIN_PRE_EMPH_LEVEL_1:
 		signal_levels |= DP_PRE_EMPHASIS_3_5;
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_6:
+	case DP_TRAIN_PRE_EMPH_LEVEL_2:
 		signal_levels |= DP_PRE_EMPHASIS_6;
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_9_5:
+	case DP_TRAIN_PRE_EMPH_LEVEL_3:
 		signal_levels |= DP_PRE_EMPHASIS_9_5;
 		break;
 	}
@@ -2796,19 +3116,19 @@ intel_gen6_edp_signal_levels(uint8_t train_set)
 	int signal_levels = train_set & (DP_TRAIN_VOLTAGE_SWING_MASK |
 					 DP_TRAIN_PRE_EMPHASIS_MASK);
 	switch (signal_levels) {
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_0:
-	case DP_TRAIN_VOLTAGE_SWING_600 | DP_TRAIN_PRE_EMPHASIS_0:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_0:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1 | DP_TRAIN_PRE_EMPH_LEVEL_0:
 		return EDP_LINK_TRAIN_400_600MV_0DB_SNB_B;
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_3_5:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_1:
 		return EDP_LINK_TRAIN_400MV_3_5DB_SNB_B;
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_6:
-	case DP_TRAIN_VOLTAGE_SWING_600 | DP_TRAIN_PRE_EMPHASIS_6:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_2:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1 | DP_TRAIN_PRE_EMPH_LEVEL_2:
 		return EDP_LINK_TRAIN_400_600MV_6DB_SNB_B;
-	case DP_TRAIN_VOLTAGE_SWING_600 | DP_TRAIN_PRE_EMPHASIS_3_5:
-	case DP_TRAIN_VOLTAGE_SWING_800 | DP_TRAIN_PRE_EMPHASIS_3_5:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1 | DP_TRAIN_PRE_EMPH_LEVEL_1:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_2 | DP_TRAIN_PRE_EMPH_LEVEL_1:
 		return EDP_LINK_TRAIN_600_800MV_3_5DB_SNB_B;
-	case DP_TRAIN_VOLTAGE_SWING_800 | DP_TRAIN_PRE_EMPHASIS_0:
-	case DP_TRAIN_VOLTAGE_SWING_1200 | DP_TRAIN_PRE_EMPHASIS_0:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_2 | DP_TRAIN_PRE_EMPH_LEVEL_0:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_3 | DP_TRAIN_PRE_EMPH_LEVEL_0:
 		return EDP_LINK_TRAIN_800_1200MV_0DB_SNB_B;
 	default:
 		DRM_DEBUG_KMS("Unsupported voltage swing/pre-emphasis level:"
@@ -2824,21 +3144,21 @@ intel_gen7_edp_signal_levels(uint8_t train_set)
 	int signal_levels = train_set & (DP_TRAIN_VOLTAGE_SWING_MASK |
 					 DP_TRAIN_PRE_EMPHASIS_MASK);
 	switch (signal_levels) {
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_0:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_0:
 		return EDP_LINK_TRAIN_400MV_0DB_IVB;
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_3_5:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_1:
 		return EDP_LINK_TRAIN_400MV_3_5DB_IVB;
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_6:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_2:
 		return EDP_LINK_TRAIN_400MV_6DB_IVB;
 
-	case DP_TRAIN_VOLTAGE_SWING_600 | DP_TRAIN_PRE_EMPHASIS_0:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1 | DP_TRAIN_PRE_EMPH_LEVEL_0:
 		return EDP_LINK_TRAIN_600MV_0DB_IVB;
-	case DP_TRAIN_VOLTAGE_SWING_600 | DP_TRAIN_PRE_EMPHASIS_3_5:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1 | DP_TRAIN_PRE_EMPH_LEVEL_1:
 		return EDP_LINK_TRAIN_600MV_3_5DB_IVB;
 
-	case DP_TRAIN_VOLTAGE_SWING_800 | DP_TRAIN_PRE_EMPHASIS_0:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_2 | DP_TRAIN_PRE_EMPH_LEVEL_0:
 		return EDP_LINK_TRAIN_800MV_0DB_IVB;
-	case DP_TRAIN_VOLTAGE_SWING_800 | DP_TRAIN_PRE_EMPHASIS_3_5:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_2 | DP_TRAIN_PRE_EMPH_LEVEL_1:
 		return EDP_LINK_TRAIN_800MV_3_5DB_IVB;
 
 	default:
@@ -2855,30 +3175,30 @@ intel_hsw_signal_levels(uint8_t train_set)
 	int signal_levels = train_set & (DP_TRAIN_VOLTAGE_SWING_MASK |
 					 DP_TRAIN_PRE_EMPHASIS_MASK);
 	switch (signal_levels) {
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_0:
-		return DDI_BUF_EMP_400MV_0DB_HSW;
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_3_5:
-		return DDI_BUF_EMP_400MV_3_5DB_HSW;
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_6:
-		return DDI_BUF_EMP_400MV_6DB_HSW;
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_9_5:
-		return DDI_BUF_EMP_400MV_9_5DB_HSW;
-
-	case DP_TRAIN_VOLTAGE_SWING_600 | DP_TRAIN_PRE_EMPHASIS_0:
-		return DDI_BUF_EMP_600MV_0DB_HSW;
-	case DP_TRAIN_VOLTAGE_SWING_600 | DP_TRAIN_PRE_EMPHASIS_3_5:
-		return DDI_BUF_EMP_600MV_3_5DB_HSW;
-	case DP_TRAIN_VOLTAGE_SWING_600 | DP_TRAIN_PRE_EMPHASIS_6:
-		return DDI_BUF_EMP_600MV_6DB_HSW;
-
-	case DP_TRAIN_VOLTAGE_SWING_800 | DP_TRAIN_PRE_EMPHASIS_0:
-		return DDI_BUF_EMP_800MV_0DB_HSW;
-	case DP_TRAIN_VOLTAGE_SWING_800 | DP_TRAIN_PRE_EMPHASIS_3_5:
-		return DDI_BUF_EMP_800MV_3_5DB_HSW;
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_0:
+		return DDI_BUF_TRANS_SELECT(0);
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_1:
+		return DDI_BUF_TRANS_SELECT(1);
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_2:
+		return DDI_BUF_TRANS_SELECT(2);
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_3:
+		return DDI_BUF_TRANS_SELECT(3);
+
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1 | DP_TRAIN_PRE_EMPH_LEVEL_0:
+		return DDI_BUF_TRANS_SELECT(4);
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1 | DP_TRAIN_PRE_EMPH_LEVEL_1:
+		return DDI_BUF_TRANS_SELECT(5);
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1 | DP_TRAIN_PRE_EMPH_LEVEL_2:
+		return DDI_BUF_TRANS_SELECT(6);
+
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_2 | DP_TRAIN_PRE_EMPH_LEVEL_0:
+		return DDI_BUF_TRANS_SELECT(7);
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_2 | DP_TRAIN_PRE_EMPH_LEVEL_1:
+		return DDI_BUF_TRANS_SELECT(8);
 	default:
 		DRM_DEBUG_KMS("Unsupported voltage swing/pre-emphasis level:"
 			      "0x%x\n", signal_levels);
-		return DDI_BUF_EMP_400MV_0DB_HSW;
+		return DDI_BUF_TRANS_SELECT(0);
 	}
 }
 
@@ -2892,7 +3212,7 @@ intel_dp_set_signal_levels(struct intel_dp *intel_dp, uint32_t *DP)
 	uint32_t signal_levels, mask;
 	uint8_t train_set = intel_dp->train_set[0];
 
-	if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev) || INTEL_INFO(dev)->gen >= 9) {
 		signal_levels = intel_hsw_signal_levels(train_set);
 		mask = DDI_BUF_EMP_MASK;
 	} else if (IS_CHERRYVIEW(dev)) {
@@ -2925,74 +3245,10 @@ intel_dp_set_link_train(struct intel_dp *intel_dp,
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
 	struct drm_device *dev = intel_dig_port->base.base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	enum port port = intel_dig_port->port;
 	uint8_t buf[sizeof(intel_dp->train_set) + 1];
 	int ret, len;
 
-	if (HAS_DDI(dev)) {
-		uint32_t temp = I915_READ(DP_TP_CTL(port));
-
-		if (dp_train_pat & DP_LINK_SCRAMBLING_DISABLE)
-			temp |= DP_TP_CTL_SCRAMBLE_DISABLE;
-		else
-			temp &= ~DP_TP_CTL_SCRAMBLE_DISABLE;
-
-		temp &= ~DP_TP_CTL_LINK_TRAIN_MASK;
-		switch (dp_train_pat & DP_TRAINING_PATTERN_MASK) {
-		case DP_TRAINING_PATTERN_DISABLE:
-			temp |= DP_TP_CTL_LINK_TRAIN_NORMAL;
-
-			break;
-		case DP_TRAINING_PATTERN_1:
-			temp |= DP_TP_CTL_LINK_TRAIN_PAT1;
-			break;
-		case DP_TRAINING_PATTERN_2:
-			temp |= DP_TP_CTL_LINK_TRAIN_PAT2;
-			break;
-		case DP_TRAINING_PATTERN_3:
-			temp |= DP_TP_CTL_LINK_TRAIN_PAT3;
-			break;
-		}
-		I915_WRITE(DP_TP_CTL(port), temp);
-
-	} else if (HAS_PCH_CPT(dev) && (IS_GEN7(dev) || port != PORT_A)) {
-		*DP &= ~DP_LINK_TRAIN_MASK_CPT;
-
-		switch (dp_train_pat & DP_TRAINING_PATTERN_MASK) {
-		case DP_TRAINING_PATTERN_DISABLE:
-			*DP |= DP_LINK_TRAIN_OFF_CPT;
-			break;
-		case DP_TRAINING_PATTERN_1:
-			*DP |= DP_LINK_TRAIN_PAT_1_CPT;
-			break;
-		case DP_TRAINING_PATTERN_2:
-			*DP |= DP_LINK_TRAIN_PAT_2_CPT;
-			break;
-		case DP_TRAINING_PATTERN_3:
-			DRM_ERROR("DP training pattern 3 not supported\n");
-			*DP |= DP_LINK_TRAIN_PAT_2_CPT;
-			break;
-		}
-
-	} else {
-		*DP &= ~DP_LINK_TRAIN_MASK;
-
-		switch (dp_train_pat & DP_TRAINING_PATTERN_MASK) {
-		case DP_TRAINING_PATTERN_DISABLE:
-			*DP |= DP_LINK_TRAIN_OFF;
-			break;
-		case DP_TRAINING_PATTERN_1:
-			*DP |= DP_LINK_TRAIN_PAT_1;
-			break;
-		case DP_TRAINING_PATTERN_2:
-			*DP |= DP_LINK_TRAIN_PAT_2;
-			break;
-		case DP_TRAINING_PATTERN_3:
-			DRM_ERROR("DP training pattern 3 not supported\n");
-			*DP |= DP_LINK_TRAIN_PAT_2;
-			break;
-		}
-	}
+	_intel_dp_set_link_train(intel_dp, DP, dp_train_pat);
 
 	I915_WRITE(intel_dp->output_reg, *DP);
 	POSTING_READ(intel_dp->output_reg);
@@ -3220,7 +3476,6 @@ intel_dp_complete_link_train(struct intel_dp *intel_dp)
 
 		/* Try 5 times, then try clock recovery if that fails */
 		if (tries > 5) {
-			intel_dp_link_down(intel_dp);
 			intel_dp_start_link_train(intel_dp);
 			intel_dp_set_link_train(intel_dp, &DP,
 						training_pattern |
@@ -3276,7 +3531,10 @@ intel_dp_link_down(struct intel_dp *intel_dp)
 		DP &= ~DP_LINK_TRAIN_MASK_CPT;
 		I915_WRITE(intel_dp->output_reg, DP | DP_LINK_TRAIN_PAT_IDLE_CPT);
 	} else {
-		DP &= ~DP_LINK_TRAIN_MASK;
+		if (IS_CHERRYVIEW(dev))
+			DP &= ~DP_LINK_TRAIN_MASK_CHV;
+		else
+			DP &= ~DP_LINK_TRAIN_MASK;
 		I915_WRITE(intel_dp->output_reg, DP | DP_LINK_TRAIN_PAT_IDLE);
 	}
 	POSTING_READ(intel_dp->output_reg);
@@ -3322,15 +3580,11 @@ intel_dp_get_dpcd(struct intel_dp *intel_dp)
 	struct drm_device *dev = dig_port->base.base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	char dpcd_hex_dump[sizeof(intel_dp->dpcd) * 3];
-
 	if (intel_dp_dpcd_read_wake(&intel_dp->aux, 0x000, intel_dp->dpcd,
 				    sizeof(intel_dp->dpcd)) < 0)
 		return false; /* aux transfer failed */
 
-	hex_dump_to_buffer(intel_dp->dpcd, sizeof(intel_dp->dpcd),
-			   32, 1, dpcd_hex_dump, sizeof(dpcd_hex_dump), false);
-	DRM_DEBUG_KMS("DPCD: %s\n", dpcd_hex_dump);
+	DRM_DEBUG_KMS("DPCD: %*ph\n", (int) sizeof(intel_dp->dpcd), intel_dp->dpcd);
 
 	if (intel_dp->dpcd[DP_DPCD_REV] == 0)
 		return false; /* DPCD not present */
@@ -3347,11 +3601,12 @@ intel_dp_get_dpcd(struct intel_dp *intel_dp)
 		}
 	}
 
-	/* Training Pattern 3 support */
+	/* Training Pattern 3 support, both source and sink */
 	if (intel_dp->dpcd[DP_DPCD_REV] >= 0x12 &&
-	    intel_dp->dpcd[DP_MAX_LANE_COUNT] & DP_TPS3_SUPPORTED) {
+	    intel_dp->dpcd[DP_MAX_LANE_COUNT] & DP_TPS3_SUPPORTED &&
+	    (IS_HASWELL(dev_priv) || INTEL_INFO(dev_priv)->gen >= 8)) {
 		intel_dp->use_tps3 = true;
-		DRM_DEBUG_KMS("Displayport TPS3 supported");
+		DRM_DEBUG_KMS("Displayport TPS3 supported\n");
 	} else
 		intel_dp->use_tps3 = false;
 
@@ -3378,8 +3633,6 @@ intel_dp_probe_oui(struct intel_dp *intel_dp)
 	if (!(intel_dp->dpcd[DP_DOWN_STREAM_PORT_COUNT] & DP_OUI_SUPPORT))
 		return;
 
-	intel_edp_panel_vdd_on(intel_dp);
-
 	if (intel_dp_dpcd_read_wake(&intel_dp->aux, DP_SINK_OUI, buf, 3) == 3)
 		DRM_DEBUG_KMS("Sink OUI: %02hx%02hx%02hx\n",
 			      buf[0], buf[1], buf[2]);
@@ -3387,8 +3640,6 @@ intel_dp_probe_oui(struct intel_dp *intel_dp)
 	if (intel_dp_dpcd_read_wake(&intel_dp->aux, DP_BRANCH_OUI, buf, 3) == 3)
 		DRM_DEBUG_KMS("Branch OUI: %02hx%02hx%02hx\n",
 			      buf[0], buf[1], buf[2]);
-
-	edp_panel_vdd_off(intel_dp, false);
 }
 
 static bool
@@ -3402,7 +3653,6 @@ intel_dp_probe_mst(struct intel_dp *intel_dp)
 	if (intel_dp->dpcd[DP_DPCD_REV] < 0x12)
 		return false;
 
-	_edp_panel_vdd_on(intel_dp);
 	if (intel_dp_dpcd_read_wake(&intel_dp->aux, DP_MSTM_CAP, buf, 1)) {
 		if (buf[0] & DP_MST_CAP) {
 			DRM_DEBUG_KMS("Sink is MST capable\n");
@@ -3412,7 +3662,6 @@ intel_dp_probe_mst(struct intel_dp *intel_dp)
 			intel_dp->is_mst = false;
 		}
 	}
-	edp_panel_vdd_off(intel_dp, false);
 
 	drm_dp_mst_topology_mgr_set_mst(&intel_dp->mst_mgr, intel_dp->is_mst);
 	return intel_dp->is_mst;
@@ -3424,26 +3673,48 @@ int intel_dp_sink_crc(struct intel_dp *intel_dp, u8 *crc)
 	struct drm_device *dev = intel_dig_port->base.base.dev;
 	struct intel_crtc *intel_crtc =
 		to_intel_crtc(intel_dig_port->base.base.crtc);
-	u8 buf[1];
+	u8 buf;
+	int test_crc_count;
+	int attempts = 6;
 
-	if (drm_dp_dpcd_readb(&intel_dp->aux, DP_TEST_SINK_MISC, buf) < 0)
-		return -EAGAIN;
+	if (drm_dp_dpcd_readb(&intel_dp->aux, DP_TEST_SINK_MISC, &buf) < 0)
+		return -EIO;
 
-	if (!(buf[0] & DP_TEST_CRC_SUPPORTED))
+	if (!(buf & DP_TEST_CRC_SUPPORTED))
 		return -ENOTTY;
 
+	if (drm_dp_dpcd_readb(&intel_dp->aux, DP_TEST_SINK, &buf) < 0)
+		return -EIO;
+
 	if (drm_dp_dpcd_writeb(&intel_dp->aux, DP_TEST_SINK,
-			       DP_TEST_SINK_START) < 0)
-		return -EAGAIN;
+				buf | DP_TEST_SINK_START) < 0)
+		return -EIO;
+
+	if (drm_dp_dpcd_readb(&intel_dp->aux, DP_TEST_SINK_MISC, &buf) < 0)
+		return -EIO;
+	test_crc_count = buf & DP_TEST_COUNT_MASK;
+
+	do {
+		if (drm_dp_dpcd_readb(&intel_dp->aux,
+				      DP_TEST_SINK_MISC, &buf) < 0)
+			return -EIO;
+		intel_wait_for_vblank(dev, intel_crtc->pipe);
+	} while (--attempts && (buf & DP_TEST_COUNT_MASK) == test_crc_count);
 
-	/* Wait 2 vblanks to be sure we will have the correct CRC value */
-	intel_wait_for_vblank(dev, intel_crtc->pipe);
-	intel_wait_for_vblank(dev, intel_crtc->pipe);
+	if (attempts == 0) {
+		DRM_DEBUG_KMS("Panel is unable to calculate CRC after 6 vblanks\n");
+		return -ETIMEDOUT;
+	}
 
 	if (drm_dp_dpcd_read(&intel_dp->aux, DP_TEST_CRC_R_CR, crc, 6) < 0)
-		return -EAGAIN;
+		return -EIO;
+
+	if (drm_dp_dpcd_readb(&intel_dp->aux, DP_TEST_SINK, &buf) < 0)
+		return -EIO;
+	if (drm_dp_dpcd_writeb(&intel_dp->aux, DP_TEST_SINK,
+			       buf & ~DP_TEST_SINK_START) < 0)
+		return -EIO;
 
-	drm_dp_dpcd_writeb(&intel_dp->aux, DP_TEST_SINK, 0);
 	return 0;
 }
 
@@ -3644,20 +3915,24 @@ intel_dp_detect_dpcd(struct intel_dp *intel_dp)
 }
 
 static enum drm_connector_status
+edp_detect(struct intel_dp *intel_dp)
+{
+	struct drm_device *dev = intel_dp_to_dev(intel_dp);
+	enum drm_connector_status status;
+
+	status = intel_panel_detect(dev);
+	if (status == connector_status_unknown)
+		status = connector_status_connected;
+
+	return status;
+}
+
+static enum drm_connector_status
 ironlake_dp_detect(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
-	enum drm_connector_status status;
-
-	/* Can't disconnect eDP, but you can close the lid... */
-	if (is_edp(intel_dp)) {
-		status = intel_panel_detect(dev);
-		if (status == connector_status_unknown)
-			status = connector_status_connected;
-		return status;
-	}
 
 	if (!ibx_digital_port_connected(dev_priv, intel_dig_port))
 		return connector_status_disconnected;
@@ -3733,9 +4008,9 @@ g4x_dp_detect(struct intel_dp *intel_dp)
 }
 
 static struct edid *
-intel_dp_get_edid(struct drm_connector *connector, struct i2c_adapter *adapter)
+intel_dp_get_edid(struct intel_dp *intel_dp)
 {
-	struct intel_connector *intel_connector = to_intel_connector(connector);
+	struct intel_connector *intel_connector = intel_dp->attached_connector;
 
 	/* use cached edid if we have one */
 	if (intel_connector->edid) {
@@ -3744,27 +4019,55 @@ intel_dp_get_edid(struct drm_connector *connector, struct i2c_adapter *adapter)
 			return NULL;
 
 		return drm_edid_duplicate(intel_connector->edid);
-	}
+	} else
+		return drm_get_edid(&intel_connector->base,
+				    &intel_dp->aux.ddc);
+}
 
-	return drm_get_edid(connector, adapter);
+static void
+intel_dp_set_edid(struct intel_dp *intel_dp)
+{
+	struct intel_connector *intel_connector = intel_dp->attached_connector;
+	struct edid *edid;
+
+	edid = intel_dp_get_edid(intel_dp);
+	intel_connector->detect_edid = edid;
+
+	if (intel_dp->force_audio != HDMI_AUDIO_AUTO)
+		intel_dp->has_audio = intel_dp->force_audio == HDMI_AUDIO_ON;
+	else
+		intel_dp->has_audio = drm_detect_monitor_audio(edid);
 }
 
-static int
-intel_dp_get_edid_modes(struct drm_connector *connector, struct i2c_adapter *adapter)
+static void
+intel_dp_unset_edid(struct intel_dp *intel_dp)
 {
-	struct intel_connector *intel_connector = to_intel_connector(connector);
+	struct intel_connector *intel_connector = intel_dp->attached_connector;
 
-	/* use cached edid if we have one */
-	if (intel_connector->edid) {
-		/* invalid edid */
-		if (IS_ERR(intel_connector->edid))
-			return 0;
+	kfree(intel_connector->detect_edid);
+	intel_connector->detect_edid = NULL;
 
-		return intel_connector_update_modes(connector,
-						    intel_connector->edid);
-	}
+	intel_dp->has_audio = false;
+}
+
+static enum intel_display_power_domain
+intel_dp_power_get(struct intel_dp *dp)
+{
+	struct intel_encoder *encoder = &dp_to_dig_port(dp)->base;
+	enum intel_display_power_domain power_domain;
 
-	return intel_ddc_get_modes(connector, adapter);
+	power_domain = intel_display_port_power_domain(encoder);
+	intel_display_power_get(to_i915(encoder->base.dev), power_domain);
+
+	return power_domain;
+}
+
+static void
+intel_dp_power_put(struct intel_dp *dp,
+		   enum intel_display_power_domain power_domain)
+{
+	struct intel_encoder *encoder = &dp_to_dig_port(dp)->base;
+	intel_display_power_put(to_i915(encoder->base.dev), power_domain);
 }
 
 static enum drm_connector_status
@@ -3774,33 +4077,30 @@ intel_dp_detect(struct drm_connector *connector, bool force)
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
 	struct intel_encoder *intel_encoder = &intel_dig_port->base;
 	struct drm_device *dev = connector->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	enum drm_connector_status status;
 	enum intel_display_power_domain power_domain;
-	struct edid *edid = NULL;
 	bool ret;
 
-	power_domain = intel_display_port_power_domain(intel_encoder);
-	intel_display_power_get(dev_priv, power_domain);
-
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
 		      connector->base.id, connector->name);
+	intel_dp_unset_edid(intel_dp);
 
 	if (intel_dp->is_mst) {
 		/* MST devices are disconnected from a monitor POV */
 		if (intel_encoder->type != INTEL_OUTPUT_EDP)
 			intel_encoder->type = INTEL_OUTPUT_DISPLAYPORT;
-		status = connector_status_disconnected;
-		goto out;
+		return connector_status_disconnected;
 	}
 
-	intel_dp->has_audio = false;
+	power_domain = intel_dp_power_get(intel_dp);
 
-	if (HAS_PCH_SPLIT(dev))
+	/* Can't disconnect eDP, but you can close the lid... */
+	if (is_edp(intel_dp))
+		status = edp_detect(intel_dp);
+	else if (HAS_PCH_SPLIT(dev))
 		status = ironlake_dp_detect(intel_dp);
 	else
 		status = g4x_dp_detect(intel_dp);
-
 	if (status != connector_status_connected)
 		goto out;
 
@@ -3816,82 +4116,78 @@ intel_dp_detect(struct drm_connector *connector, bool force)
 		goto out;
 	}
 
-	if (intel_dp->force_audio != HDMI_AUDIO_AUTO) {
-		intel_dp->has_audio = (intel_dp->force_audio == HDMI_AUDIO_ON);
-	} else {
-		edid = intel_dp_get_edid(connector, &intel_dp->aux.ddc);
-		if (edid) {
-			intel_dp->has_audio = drm_detect_monitor_audio(edid);
-			kfree(edid);
-		}
-	}
+	intel_dp_set_edid(intel_dp);
 
 	if (intel_encoder->type != INTEL_OUTPUT_EDP)
 		intel_encoder->type = INTEL_OUTPUT_DISPLAYPORT;
 	status = connector_status_connected;
 
 out:
-	intel_display_power_put(dev_priv, power_domain);
+	intel_dp_power_put(intel_dp, power_domain);
 	return status;
 }
 
-static int intel_dp_get_modes(struct drm_connector *connector)
+static void
+intel_dp_force(struct drm_connector *connector)
 {
 	struct intel_dp *intel_dp = intel_attached_dp(connector);
-	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
-	struct intel_encoder *intel_encoder = &intel_dig_port->base;
-	struct intel_connector *intel_connector = to_intel_connector(connector);
-	struct drm_device *dev = connector->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_encoder *intel_encoder = &dp_to_dig_port(intel_dp)->base;
 	enum intel_display_power_domain power_domain;
-	int ret;
 
-	/* We should parse the EDID data and find out if it has an audio sink
-	 */
+	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
+		      connector->base.id, connector->name);
+	intel_dp_unset_edid(intel_dp);
 
-	power_domain = intel_display_port_power_domain(intel_encoder);
-	intel_display_power_get(dev_priv, power_domain);
+	if (connector->status != connector_status_connected)
+		return;
 
-	ret = intel_dp_get_edid_modes(connector, &intel_dp->aux.ddc);
-	intel_display_power_put(dev_priv, power_domain);
-	if (ret)
-		return ret;
+	power_domain = intel_dp_power_get(intel_dp);
+
+	intel_dp_set_edid(intel_dp);
+
+	intel_dp_power_put(intel_dp, power_domain);
+
+	if (intel_encoder->type != INTEL_OUTPUT_EDP)
+		intel_encoder->type = INTEL_OUTPUT_DISPLAYPORT;
+}
+
+static int intel_dp_get_modes(struct drm_connector *connector)
+{
+	struct intel_connector *intel_connector = to_intel_connector(connector);
+	struct edid *edid;
+
+	edid = intel_connector->detect_edid;
+	if (edid) {
+		int ret = intel_connector_update_modes(connector, edid);
+		if (ret)
+			return ret;
+	}
 
 	/* if eDP has no EDID, fall back to fixed mode */
-	if (is_edp(intel_dp) && intel_connector->panel.fixed_mode) {
+	if (is_edp(intel_attached_dp(connector)) &&
+	    intel_connector->panel.fixed_mode) {
 		struct drm_display_mode *mode;
-		mode = drm_mode_duplicate(dev,
+
+		mode = drm_mode_duplicate(connector->dev,
 					  intel_connector->panel.fixed_mode);
 		if (mode) {
 			drm_mode_probed_add(connector, mode);
 			return 1;
 		}
 	}
+
 	return 0;
 }
 
 static bool
 intel_dp_detect_audio(struct drm_connector *connector)
 {
-	struct intel_dp *intel_dp = intel_attached_dp(connector);
-	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
-	struct intel_encoder *intel_encoder = &intel_dig_port->base;
-	struct drm_device *dev = connector->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	enum intel_display_power_domain power_domain;
-	struct edid *edid;
 	bool has_audio = false;
+	struct edid *edid;
 
-	power_domain = intel_display_port_power_domain(intel_encoder);
-	intel_display_power_get(dev_priv, power_domain);
-
-	edid = intel_dp_get_edid(connector, &intel_dp->aux.ddc);
-	if (edid) {
+	edid = to_intel_connector(connector)->detect_edid;
+	if (edid)
 		has_audio = drm_detect_monitor_audio(edid);
-		kfree(edid);
-	}
-
-	intel_display_power_put(dev_priv, power_domain);
 
 	return has_audio;
 }
@@ -3989,6 +4285,8 @@ intel_dp_connector_destroy(struct drm_connector *connector)
 {
 	struct intel_connector *intel_connector = to_intel_connector(connector);
 
+	kfree(intel_connector->detect_edid);
+
 	if (!IS_ERR_OR_NULL(intel_connector->edid))
 		kfree(intel_connector->edid);
 
@@ -4005,16 +4303,20 @@ void intel_dp_encoder_destroy(struct drm_encoder *encoder)
 {
 	struct intel_digital_port *intel_dig_port = enc_to_dig_port(encoder);
 	struct intel_dp *intel_dp = &intel_dig_port->dp;
-	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 
 	drm_dp_aux_unregister(&intel_dp->aux);
 	intel_dp_mst_encoder_cleanup(intel_dig_port);
 	drm_encoder_cleanup(encoder);
 	if (is_edp(intel_dp)) {
 		cancel_delayed_work_sync(&intel_dp->panel_vdd_work);
-		drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
+		/*
+		 * vdd might still be enabled do to the delayed vdd off.
+		 * Make sure vdd is actually turned off here.
+		 */
+		pps_lock(intel_dp);
 		edp_panel_vdd_off_sync(intel_dp);
-		drm_modeset_unlock(&dev->mode_config.connection_mutex);
+		pps_unlock(intel_dp);
+
 		if (intel_dp->edp_notifier.notifier_call) {
 			unregister_reboot_notifier(&intel_dp->edp_notifier);
 			intel_dp->edp_notifier.notifier_call = NULL;
@@ -4030,17 +4332,68 @@ static void intel_dp_encoder_suspend(struct intel_encoder *intel_encoder)
 	if (!is_edp(intel_dp))
 		return;
 
+	/*
+	 * vdd might still be enabled do to the delayed vdd off.
+	 * Make sure vdd is actually turned off here.
+	 */
+	cancel_delayed_work_sync(&intel_dp->panel_vdd_work);
+	pps_lock(intel_dp);
 	edp_panel_vdd_off_sync(intel_dp);
+	pps_unlock(intel_dp);
+}
+
+static void intel_edp_panel_vdd_sanitize(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = intel_dig_port->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum intel_display_power_domain power_domain;
+
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
+	if (!edp_have_panel_vdd(intel_dp))
+		return;
+
+	/*
+	 * The VDD bit needs a power domain reference, so if the bit is
+	 * already enabled when we boot or resume, grab this reference and
+	 * schedule a vdd off, so we don't hold on to the reference
+	 * indefinitely.
+	 */
+	DRM_DEBUG_KMS("VDD left on by BIOS, adjusting state tracking\n");
+	power_domain = intel_display_port_power_domain(&intel_dig_port->base);
+	intel_display_power_get(dev_priv, power_domain);
+
+	edp_panel_vdd_schedule_off(intel_dp);
 }
 
 static void intel_dp_encoder_reset(struct drm_encoder *encoder)
 {
-	intel_edp_panel_vdd_sanitize(to_intel_encoder(encoder));
+	struct intel_dp *intel_dp;
+
+	if (to_intel_encoder(encoder)->type != INTEL_OUTPUT_EDP)
+		return;
+
+	intel_dp = enc_to_intel_dp(encoder);
+
+	pps_lock(intel_dp);
+
+	/*
+	 * Read out the current power sequencer assignment,
+	 * in case the BIOS did something with it.
+	 */
+	if (IS_VALLEYVIEW(encoder->dev))
+		vlv_initial_power_sequencer_setup(intel_dp);
+
+	intel_edp_panel_vdd_sanitize(intel_dp);
+
+	pps_unlock(intel_dp);
 }
 
 static const struct drm_connector_funcs intel_dp_connector_funcs = {
 	.dpms = intel_connector_dpms,
 	.detect = intel_dp_detect,
+	.force = intel_dp_force,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.set_property = intel_dp_set_property,
 	.destroy = intel_dp_connector_destroy,
@@ -4076,7 +4429,20 @@ intel_dp_hpd_pulse(struct intel_digital_port *intel_dig_port, bool long_hpd)
 	if (intel_dig_port->base.type != INTEL_OUTPUT_EDP)
 		intel_dig_port->base.type = INTEL_OUTPUT_DISPLAYPORT;
 
-	DRM_DEBUG_KMS("got hpd irq on port %d - %s\n", intel_dig_port->port,
+	if (long_hpd && intel_dig_port->base.type == INTEL_OUTPUT_EDP) {
+		/*
+		 * vdd off can generate a long pulse on eDP which
+		 * would require vdd on to handle it, and thus we
+		 * would end up in an endless cycle of
+		 * "vdd off -> long hpd -> vdd on -> detect -> vdd off -> ..."
+		 */
+		DRM_DEBUG_KMS("ignoring long hpd on eDP port %c\n",
+			      port_name(intel_dig_port->port));
+		return false;
+	}
+
+	DRM_DEBUG_KMS("got hpd irq on port %c - %s\n",
+		      port_name(intel_dig_port->port),
 		      long_hpd ? "long" : "short");
 
 	power_domain = intel_display_port_power_domain(intel_encoder);
@@ -4208,14 +4574,20 @@ static void intel_dp_init_panel_power_timestamps(struct intel_dp *intel_dp)
 
 static void
 intel_dp_init_panel_power_sequencer(struct drm_device *dev,
-				    struct intel_dp *intel_dp,
-				    struct edp_power_seq *out)
+				    struct intel_dp *intel_dp)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct edp_power_seq cur, vbt, spec, final;
+	struct edp_power_seq cur, vbt, spec,
+		*final = &intel_dp->pps_delays;
 	u32 pp_on, pp_off, pp_div, pp;
 	int pp_ctrl_reg, pp_on_reg, pp_off_reg, pp_div_reg;
 
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
+	/* already initialized? */
+	if (final->t11_t12 != 0)
+		return;
+
 	if (HAS_PCH_SPLIT(dev)) {
 		pp_ctrl_reg = PCH_PP_CONTROL;
 		pp_on_reg = PCH_PP_ON_DELAYS;
@@ -4277,7 +4649,7 @@ intel_dp_init_panel_power_sequencer(struct drm_device *dev,
 
 	/* Use the max of the register settings and vbt. If both are
 	 * unset, fall back to the spec limits. */
-#define assign_final(field)	final.field = (max(cur.field, vbt.field) == 0 ? \
+#define assign_final(field)	final->field = (max(cur.field, vbt.field) == 0 ? \
 				       spec.field : \
 				       max(cur.field, vbt.field))
 	assign_final(t1_t3);
@@ -4287,7 +4659,7 @@ intel_dp_init_panel_power_sequencer(struct drm_device *dev,
 	assign_final(t11_t12);
 #undef assign_final
 
-#define get_delay(field)	(DIV_ROUND_UP(final.field, 10))
+#define get_delay(field)	(DIV_ROUND_UP(final->field, 10))
 	intel_dp->panel_power_up_delay = get_delay(t1_t3);
 	intel_dp->backlight_on_delay = get_delay(t8);
 	intel_dp->backlight_off_delay = get_delay(t9);
@@ -4301,20 +4673,20 @@ intel_dp_init_panel_power_sequencer(struct drm_device *dev,
 
 	DRM_DEBUG_KMS("backlight on delay %d, off delay %d\n",
 		      intel_dp->backlight_on_delay, intel_dp->backlight_off_delay);
-
-	if (out)
-		*out = final;
 }
 
 static void
 intel_dp_init_panel_power_sequencer_registers(struct drm_device *dev,
-					      struct intel_dp *intel_dp,
-					      struct edp_power_seq *seq)
+					      struct intel_dp *intel_dp)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 pp_on, pp_off, pp_div, port_sel = 0;
 	int div = HAS_PCH_SPLIT(dev) ? intel_pch_rawclk(dev) : intel_hrawclk(dev);
 	int pp_on_reg, pp_off_reg, pp_div_reg;
+	enum port port = dp_to_dig_port(intel_dp)->port;
+	const struct edp_power_seq *seq = &intel_dp->pps_delays;
+
+	lockdep_assert_held(&dev_priv->pps_mutex);
 
 	if (HAS_PCH_SPLIT(dev)) {
 		pp_on_reg = PCH_PP_ON_DELAYS;
@@ -4349,12 +4721,9 @@ intel_dp_init_panel_power_sequencer_registers(struct drm_device *dev,
 	/* Haswell doesn't have any port selection bits for the panel
 	 * power sequencer any more. */
 	if (IS_VALLEYVIEW(dev)) {
-		if (dp_to_dig_port(intel_dp)->port == PORT_B)
-			port_sel = PANEL_PORT_SELECT_DPB_VLV;
-		else
-			port_sel = PANEL_PORT_SELECT_DPC_VLV;
+		port_sel = PANEL_PORT_SELECT_VLV(port);
 	} else if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev)) {
-		if (dp_to_dig_port(intel_dp)->port == PORT_A)
+		if (port == PORT_A)
 			port_sel = PANEL_PORT_SELECT_DPA;
 		else
 			port_sel = PANEL_PORT_SELECT_DPD;
@@ -4398,7 +4767,7 @@ void intel_dp_set_drrs_state(struct drm_device *dev, int refresh_rate)
 	 * hard to tell without seeing the user of this function of this code.
 	 * Check locking and ordering once that lands.
 	 */
-	if (INTEL_INFO(dev)->gen < 8 && intel_edp_is_psr_enabled(dev)) {
+	if (INTEL_INFO(dev)->gen < 8 && intel_psr_is_enabled(dev)) {
 		DRM_DEBUG_KMS("DRRS is disabled as PSR is enabled\n");
 		return;
 	}
@@ -4438,7 +4807,7 @@ void intel_dp_set_drrs_state(struct drm_device *dev, int refresh_rate)
 		val = I915_READ(reg);
 		if (index > DRRS_HIGH_RR) {
 			val |= PIPECONF_EDP_RR_MODE_SWITCH;
-			intel_dp_set_m2_n2(intel_crtc, &config->dp_m2_n2);
+			intel_dp_set_m_n(intel_crtc);
 		} else {
 			val &= ~PIPECONF_EDP_RR_MODE_SWITCH;
 		}
@@ -4478,7 +4847,7 @@ intel_dp_drrs_init(struct intel_digital_port *intel_dig_port,
 	}
 
 	if (dev_priv->vbt.drrs_type != SEAMLESS_DRRS_SUPPORT) {
-		DRM_INFO("VBT doesn't support DRRS\n");
+		DRM_DEBUG_KMS("VBT doesn't support DRRS\n");
 		return NULL;
 	}
 
@@ -4486,7 +4855,7 @@ intel_dp_drrs_init(struct intel_digital_port *intel_dig_port,
 					(dev, fixed_mode, connector);
 
 	if (!downclock_mode) {
-		DRM_INFO("DRRS not supported\n");
+		DRM_DEBUG_KMS("DRRS not supported\n");
 		return NULL;
 	}
 
@@ -4497,39 +4866,12 @@ intel_dp_drrs_init(struct intel_digital_port *intel_dig_port,
 	intel_dp->drrs_state.type = dev_priv->vbt.drrs_type;
 
 	intel_dp->drrs_state.refresh_rate_type = DRRS_HIGH_RR;
-	DRM_INFO("seamless DRRS supported for eDP panel.\n");
+	DRM_DEBUG_KMS("seamless DRRS supported for eDP panel.\n");
 	return downclock_mode;
 }
 
-void intel_edp_panel_vdd_sanitize(struct intel_encoder *intel_encoder)
-{
-	struct drm_device *dev = intel_encoder->base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_dp *intel_dp;
-	enum intel_display_power_domain power_domain;
-
-	if (intel_encoder->type != INTEL_OUTPUT_EDP)
-		return;
-
-	intel_dp = enc_to_intel_dp(&intel_encoder->base);
-	if (!edp_have_panel_vdd(intel_dp))
-		return;
-	/*
-	 * The VDD bit needs a power domain reference, so if the bit is
-	 * already enabled when we boot or resume, grab this reference and
-	 * schedule a vdd off, so we don't hold on to the reference
-	 * indefinitely.
-	 */
-	DRM_DEBUG_KMS("VDD left on by BIOS, adjusting state tracking\n");
-	power_domain = intel_display_port_power_domain(intel_encoder);
-	intel_display_power_get(dev_priv, power_domain);
-
-	edp_panel_vdd_schedule_off(intel_dp);
-}
-
 static bool intel_edp_init_connector(struct intel_dp *intel_dp,
-				     struct intel_connector *intel_connector,
-				     struct edp_power_seq *power_seq)
+				     struct intel_connector *intel_connector)
 {
 	struct drm_connector *connector = &intel_connector->base;
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
@@ -4541,18 +4883,19 @@ static bool intel_edp_init_connector(struct intel_dp *intel_dp,
 	bool has_dpcd;
 	struct drm_display_mode *scan;
 	struct edid *edid;
+	enum pipe pipe = INVALID_PIPE;
 
 	intel_dp->drrs_state.type = DRRS_NOT_SUPPORTED;
 
 	if (!is_edp(intel_dp))
 		return true;
 
-	intel_edp_panel_vdd_sanitize(intel_encoder);
+	pps_lock(intel_dp);
+	intel_edp_panel_vdd_sanitize(intel_dp);
+	pps_unlock(intel_dp);
 
 	/* Cache DPCD and EDID for edp. */
-	intel_edp_panel_vdd_on(intel_dp);
 	has_dpcd = intel_dp_get_dpcd(intel_dp);
-	edp_panel_vdd_off(intel_dp, false);
 
 	if (has_dpcd) {
 		if (intel_dp->dpcd[DP_DPCD_REV] >= 0x11)
@@ -4566,7 +4909,9 @@ static bool intel_edp_init_connector(struct intel_dp *intel_dp,
 	}
 
 	/* We now know it's not a ghost, init power sequence regs. */
-	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp, power_seq);
+	pps_lock(intel_dp);
+	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp);
+	pps_unlock(intel_dp);
 
 	mutex_lock(&dev->mode_config.mutex);
 	edid = drm_get_edid(connector, &intel_dp->aux.ddc);
@@ -4607,10 +4952,30 @@ static bool intel_edp_init_connector(struct intel_dp *intel_dp,
 	if (IS_VALLEYVIEW(dev)) {
 		intel_dp->edp_notifier.notifier_call = edp_notify_handler;
 		register_reboot_notifier(&intel_dp->edp_notifier);
+
+		/*
+		 * Figure out the current pipe for the initial backlight setup.
+		 * If the current pipe isn't valid, try the PPS pipe, and if that
+		 * fails just assume pipe A.
+		 */
+		if (IS_CHERRYVIEW(dev))
+			pipe = DP_PORT_TO_PIPE_CHV(intel_dp->DP);
+		else
+			pipe = PORT_TO_PIPE(intel_dp->DP);
+
+		if (pipe != PIPE_A && pipe != PIPE_B)
+			pipe = intel_dp->pps_pipe;
+
+		if (pipe != PIPE_A && pipe != PIPE_B)
+			pipe = PIPE_A;
+
+		DRM_DEBUG_KMS("using pipe %c for initial backlight setup\n",
+			      pipe_name(pipe));
 	}
 
 	intel_panel_init(&intel_connector->panel, fixed_mode, downclock_mode);
-	intel_panel_setup_backlight(connector);
+	intel_connector->panel.backlight_power = intel_edp_backlight_power;
+	intel_panel_setup_backlight(connector, pipe);
 
 	return true;
 }
@@ -4625,11 +4990,14 @@ intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
 	struct drm_device *dev = intel_encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	enum port port = intel_dig_port->port;
-	struct edp_power_seq power_seq = { 0 };
 	int type;
 
+	intel_dp->pps_pipe = INVALID_PIPE;
+
 	/* intel_dp vfuncs */
-	if (IS_VALLEYVIEW(dev))
+	if (INTEL_INFO(dev)->gen >= 9)
+		intel_dp->get_aux_clock_divider = skl_get_aux_clock_divider;
+	else if (IS_VALLEYVIEW(dev))
 		intel_dp->get_aux_clock_divider = vlv_get_aux_clock_divider;
 	else if (IS_HASWELL(dev) || IS_BROADWELL(dev))
 		intel_dp->get_aux_clock_divider = hsw_get_aux_clock_divider;
@@ -4638,7 +5006,10 @@ intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
 	else
 		intel_dp->get_aux_clock_divider = i9xx_get_aux_clock_divider;
 
-	intel_dp->get_aux_send_ctl = i9xx_get_aux_send_ctl;
+	if (INTEL_INFO(dev)->gen >= 9)
+		intel_dp->get_aux_send_ctl = skl_get_aux_send_ctl;
+	else
+		intel_dp->get_aux_send_ctl = i9xx_get_aux_send_ctl;
 
 	/* Preserve the current hw state. */
 	intel_dp->DP = I915_READ(intel_dp->output_reg);
@@ -4657,6 +5028,11 @@ intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
 	if (type == DRM_MODE_CONNECTOR_eDP)
 		intel_encoder->type = INTEL_OUTPUT_EDP;
 
+	/* eDP only on port B and/or C on vlv/chv */
+	if (WARN_ON(IS_VALLEYVIEW(dev) && is_edp(intel_dp) &&
+		    port != PORT_B && port != PORT_C))
+		return false;
+
 	DRM_DEBUG_KMS("Adding %s connector on port %c\n",
 			type == DRM_MODE_CONNECTOR_eDP ? "eDP" : "DP",
 			port_name(port));
@@ -4698,8 +5074,13 @@ intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
 	}
 
 	if (is_edp(intel_dp)) {
+		pps_lock(intel_dp);
 		intel_dp_init_panel_power_timestamps(intel_dp);
-		intel_dp_init_panel_power_sequencer(dev, intel_dp, &power_seq);
+		if (IS_VALLEYVIEW(dev))
+			vlv_initial_power_sequencer_setup(intel_dp);
+		else
+			intel_dp_init_panel_power_sequencer(dev, intel_dp);
+		pps_unlock(intel_dp);
 	}
 
 	intel_dp_aux_init(intel_dp, intel_connector);
@@ -4707,17 +5088,22 @@ intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
 	/* init MST on ports that can support it */
 	if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
 		if (port == PORT_B || port == PORT_C || port == PORT_D) {
-			intel_dp_mst_encoder_init(intel_dig_port, intel_connector->base.base.id);
+			intel_dp_mst_encoder_init(intel_dig_port,
+						  intel_connector->base.base.id);
 		}
 	}
 
-	if (!intel_edp_init_connector(intel_dp, intel_connector, &power_seq)) {
+	if (!intel_edp_init_connector(intel_dp, intel_connector)) {
 		drm_dp_aux_unregister(&intel_dp->aux);
 		if (is_edp(intel_dp)) {
 			cancel_delayed_work_sync(&intel_dp->panel_vdd_work);
-			drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
+			/*
+			 * vdd might still be enabled do to the delayed vdd off.
+			 * Make sure vdd is actually turned off here.
+			 */
+			pps_lock(intel_dp);
 			edp_panel_vdd_off_sync(intel_dp);
-			drm_modeset_unlock(&dev->mode_config.connection_mutex);
+			pps_unlock(intel_dp);
 		}
 		drm_connector_unregister(connector);
 		drm_connector_cleanup(connector);
@@ -4781,7 +5167,8 @@ intel_dp_init(struct drm_device *dev, int output_reg, enum port port)
 	} else {
 		intel_encoder->pre_enable = g4x_pre_enable_dp;
 		intel_encoder->enable = g4x_enable_dp;
-		intel_encoder->post_disable = g4x_post_disable_dp;
+		if (INTEL_INFO(dev)->gen >= 5)
+			intel_encoder->post_disable = ilk_post_disable_dp;
 	}
 
 	intel_dig_port->port = port;
diff --git a/drivers/gpu/drm/i915/intel_dp_mst.c b/drivers/gpu/drm/i915/intel_dp_mst.c
index d9a7a7865f66..7f8c6a66680a 100644
--- a/drivers/gpu/drm/i915/intel_dp_mst.c
+++ b/drivers/gpu/drm/i915/intel_dp_mst.c
@@ -278,20 +278,12 @@ static int intel_dp_mst_get_ddc_modes(struct drm_connector *connector)
 }
 
 static enum drm_connector_status
-intel_mst_port_dp_detect(struct drm_connector *connector)
+intel_dp_mst_detect(struct drm_connector *connector, bool force)
 {
 	struct intel_connector *intel_connector = to_intel_connector(connector);
 	struct intel_dp *intel_dp = intel_connector->mst_port;
 
-	return drm_dp_mst_detect_port(&intel_dp->mst_mgr, intel_connector->port);
-}
-
-static enum drm_connector_status
-intel_dp_mst_detect(struct drm_connector *connector, bool force)
-{
-	enum drm_connector_status status;
-	status = intel_mst_port_dp_detect(connector);
-	return status;
+	return drm_dp_mst_detect_port(connector, &intel_dp->mst_mgr, intel_connector->port);
 }
 
 static int
@@ -393,7 +385,7 @@ static void intel_connector_remove_from_fbdev(struct intel_connector *connector)
 #endif
 }
 
-static struct drm_connector *intel_dp_add_mst_connector(struct drm_dp_mst_topology_mgr *mgr, struct drm_dp_mst_port *port, char *pathprop)
+static struct drm_connector *intel_dp_add_mst_connector(struct drm_dp_mst_topology_mgr *mgr, struct drm_dp_mst_port *port, const char *pathprop)
 {
 	struct intel_dp *intel_dp = container_of(mgr, struct intel_dp, mst_mgr);
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
@@ -422,6 +414,8 @@ static struct drm_connector *intel_dp_add_mst_connector(struct drm_dp_mst_topolo
 	intel_dp_add_properties(intel_dp, connector);
 
 	drm_object_attach_property(&connector->base, dev->mode_config.path_property, 0);
+	drm_object_attach_property(&connector->base, dev->mode_config.tile_property, 0);
+
 	drm_mode_connector_set_path_property(connector, pathprop);
 	drm_reinit_primary_mode_group(dev);
 	mutex_lock(&dev->mode_config.mutex);
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index b8c8bbd8e5f9..3b40a17b8852 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -25,6 +25,7 @@
 #ifndef __INTEL_DRV_H__
 #define __INTEL_DRV_H__
 
+#include <linux/async.h>
 #include <linux/i2c.h>
 #include <linux/hdmi.h>
 #include <drm/i915_drm.h>
@@ -33,6 +34,10 @@
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_dp_mst_helper.h>
+#include <drm/drm_rect.h>
+
+#define DIV_ROUND_CLOSEST_ULL(ll, d)	\
+({ unsigned long long _tmp = (ll)+(d)/2; do_div(_tmp, d); _tmp; })
 
 /**
  * _wait_for - magic (register) wait macro
@@ -89,18 +94,20 @@
 
 /* these are outputs from the chip - integrated only
    external chips are via DVO or SDVO output */
-#define INTEL_OUTPUT_UNUSED 0
-#define INTEL_OUTPUT_ANALOG 1
-#define INTEL_OUTPUT_DVO 2
-#define INTEL_OUTPUT_SDVO 3
-#define INTEL_OUTPUT_LVDS 4
-#define INTEL_OUTPUT_TVOUT 5
-#define INTEL_OUTPUT_HDMI 6
-#define INTEL_OUTPUT_DISPLAYPORT 7
-#define INTEL_OUTPUT_EDP 8
-#define INTEL_OUTPUT_DSI 9
-#define INTEL_OUTPUT_UNKNOWN 10
-#define INTEL_OUTPUT_DP_MST 11
+enum intel_output_type {
+	INTEL_OUTPUT_UNUSED = 0,
+	INTEL_OUTPUT_ANALOG = 1,
+	INTEL_OUTPUT_DVO = 2,
+	INTEL_OUTPUT_SDVO = 3,
+	INTEL_OUTPUT_LVDS = 4,
+	INTEL_OUTPUT_TVOUT = 5,
+	INTEL_OUTPUT_HDMI = 6,
+	INTEL_OUTPUT_DISPLAYPORT = 7,
+	INTEL_OUTPUT_EDP = 8,
+	INTEL_OUTPUT_DSI = 9,
+	INTEL_OUTPUT_UNKNOWN = 10,
+	INTEL_OUTPUT_DP_MST = 11,
+};
 
 #define INTEL_DVO_CHIP_NONE 0
 #define INTEL_DVO_CHIP_LVDS 1
@@ -131,7 +138,7 @@ struct intel_encoder {
 	 */
 	struct intel_crtc *new_crtc;
 
-	int type;
+	enum intel_output_type type;
 	unsigned int cloneable;
 	bool connectors_active;
 	void (*hot_plug)(struct intel_encoder *);
@@ -179,6 +186,8 @@ struct intel_panel {
 		bool active_low_pwm;
 		struct backlight_device *device;
 	} backlight;
+
+	void (*backlight_power)(struct intel_connector *, bool enable);
 };
 
 struct intel_connector {
@@ -211,6 +220,7 @@ struct intel_connector {
 
 	/* Cached EDID for eDP and LVDS. May hold ERR_PTR for invalid EDID. */
 	struct edid *edid;
+	struct edid *detect_edid;
 
 	/* since POLL and HPD connectors may use the same HPD line keep the native
 	   state of connector->polled in case hotplug storm detection changes it */
@@ -233,6 +243,17 @@ typedef struct dpll {
 	int	p;
 } intel_clock_t;
 
+struct intel_plane_state {
+	struct drm_crtc *crtc;
+	struct drm_framebuffer *fb;
+	struct drm_rect src;
+	struct drm_rect dst;
+	struct drm_rect clip;
+	struct drm_rect orig_src;
+	struct drm_rect orig_dst;
+	bool visible;
+};
+
 struct intel_plane_config {
 	bool tiled;
 	int size;
@@ -271,6 +292,9 @@ struct intel_crtc_config {
 	 * between pch encoders and cpu encoders. */
 	bool has_pch_encoder;
 
+	/* Are we sending infoframes on the attached port */
+	bool has_infoframe;
+
 	/* CPU Transcoder for the pipe. Currently this can only differ from the
 	 * pipe on Haswell (where we have a special eDP transcoder). */
 	enum transcoder cpu_transcoder;
@@ -319,7 +343,10 @@ struct intel_crtc_config {
 	/* Selected dpll when shared or DPLL_ID_PRIVATE. */
 	enum intel_dpll_id shared_dpll;
 
-	/* PORT_CLK_SEL for DDI ports. */
+	/*
+	 * - PORT_CLK_SEL for DDI ports on HSW/BDW.
+	 * - enum skl_dpll on SKL
+	 */
 	uint32_t ddi_pll_sel;
 
 	/* Actual register state of the dpll, for shared dpll cross-checking. */
@@ -330,6 +357,7 @@ struct intel_crtc_config {
 
 	/* m2_n2 for eDP downclock */
 	struct intel_link_m_n dp_m2_n2;
+	bool has_drrs;
 
 	/*
 	 * Frequence the dpll for the port should run at. Differs from the
@@ -379,7 +407,14 @@ struct intel_pipe_wm {
 
 struct intel_mmio_flip {
 	u32 seqno;
-	u32 ring_id;
+	struct intel_engine_cs *ring;
+	struct work_struct work;
+};
+
+struct skl_pipe_wm {
+	struct skl_wm_level wm[8];
+	struct skl_wm_level trans_wm;
+	uint32_t linetime;
 };
 
 struct intel_crtc {
@@ -410,6 +445,7 @@ struct intel_crtc {
 	uint32_t cursor_addr;
 	int16_t cursor_width, cursor_height;
 	uint32_t cursor_cntl;
+	uint32_t cursor_size;
 	uint32_t cursor_base;
 
 	struct intel_plane_config plane_config;
@@ -428,10 +464,10 @@ struct intel_crtc {
 	struct {
 		/* watermarks currently being used  */
 		struct intel_pipe_wm active;
+		/* SKL wm values currently in use */
+		struct skl_pipe_wm skl_active;
 	} wm;
 
-	wait_queue_head_t vbl_wait;
-
 	int scanline_offset;
 	struct intel_mmio_flip mmio_flip;
 };
@@ -455,6 +491,7 @@ struct intel_plane {
 	unsigned int crtc_w, crtc_h;
 	uint32_t src_x, src_y;
 	uint32_t src_w, src_h;
+	unsigned int rotation;
 
 	/* Since we need to change the watermarks before/after
 	 * enabling/disabling the planes, we need to store the parameters here
@@ -521,6 +558,7 @@ struct intel_hdmi {
 	void (*set_infoframes)(struct drm_encoder *encoder,
 			       bool enable,
 			       struct drm_display_mode *adjusted_mode);
+	bool (*infoframe_enabled)(struct drm_encoder *encoder);
 };
 
 struct intel_dp_mst_encoder;
@@ -565,6 +603,13 @@ struct intel_dp {
 
 	struct notifier_block edp_notifier;
 
+	/*
+	 * Pipe whose power sequencer is currently locked into
+	 * this port. Only relevant on VLV/CHV.
+	 */
+	enum pipe pps_pipe;
+	struct edp_power_seq pps_delays;
+
 	bool use_tps3;
 	bool can_mst; /* this port supports mst */
 	bool is_mst;
@@ -663,6 +708,10 @@ struct intel_unpin_work {
 #define INTEL_FLIP_COMPLETE	2
 	u32 flip_count;
 	u32 gtt_offset;
+	struct intel_engine_cs *flip_queued_ring;
+	u32 flip_queued_seqno;
+	int flip_queued_vblank;
+	int flip_ready_vblank;
 	bool enable_stall_check;
 };
 
@@ -716,32 +765,48 @@ hdmi_to_dig_port(struct intel_hdmi *intel_hdmi)
 	return container_of(intel_hdmi, struct intel_digital_port, hdmi);
 }
 
+/*
+ * Returns the number of planes for this pipe, ie the number of sprites + 1
+ * (primary plane). This doesn't count the cursor plane then.
+ */
+static inline unsigned int intel_num_planes(struct intel_crtc *crtc)
+{
+	return INTEL_INFO(crtc->base.dev)->num_sprites[crtc->pipe] + 1;
+}
 
-/* i915_irq.c */
-bool intel_set_cpu_fifo_underrun_reporting(struct drm_device *dev,
+/* intel_fifo_underrun.c */
+bool intel_set_cpu_fifo_underrun_reporting(struct drm_i915_private *dev_priv,
 					   enum pipe pipe, bool enable);
-bool intel_set_pch_fifo_underrun_reporting(struct drm_device *dev,
+bool intel_set_pch_fifo_underrun_reporting(struct drm_i915_private *dev_priv,
 					   enum transcoder pch_transcoder,
 					   bool enable);
+void intel_cpu_fifo_underrun_irq_handler(struct drm_i915_private *dev_priv,
+					 enum pipe pipe);
+void intel_pch_fifo_underrun_irq_handler(struct drm_i915_private *dev_priv,
+					 enum transcoder pch_transcoder);
+void i9xx_check_fifo_underruns(struct drm_i915_private *dev_priv);
+
+/* i915_irq.c */
 void gen5_enable_gt_irq(struct drm_i915_private *dev_priv, uint32_t mask);
 void gen5_disable_gt_irq(struct drm_i915_private *dev_priv, uint32_t mask);
 void gen6_enable_pm_irq(struct drm_i915_private *dev_priv, uint32_t mask);
 void gen6_disable_pm_irq(struct drm_i915_private *dev_priv, uint32_t mask);
-void gen8_enable_pm_irq(struct drm_i915_private *dev_priv, uint32_t mask);
-void gen8_disable_pm_irq(struct drm_i915_private *dev_priv, uint32_t mask);
-void intel_runtime_pm_disable_interrupts(struct drm_device *dev);
-void intel_runtime_pm_restore_interrupts(struct drm_device *dev);
+void gen6_reset_rps_interrupts(struct drm_device *dev);
+void gen6_enable_rps_interrupts(struct drm_device *dev);
+void gen6_disable_rps_interrupts(struct drm_device *dev);
+u32 gen6_sanitize_rps_pm_mask(struct drm_i915_private *dev_priv, u32 mask);
+void intel_runtime_pm_disable_interrupts(struct drm_i915_private *dev_priv);
+void intel_runtime_pm_enable_interrupts(struct drm_i915_private *dev_priv);
 static inline bool intel_irqs_enabled(struct drm_i915_private *dev_priv)
 {
 	/*
 	 * We only use drm_irq_uninstall() at unload and VT switch, so
 	 * this is the only thing we need to check.
 	 */
-	return !dev_priv->pm._irqs_disabled;
+	return dev_priv->pm.irqs_enabled;
 }
 
 int intel_get_crtc_scanline(struct intel_crtc *crtc);
-void i9xx_check_fifo_underruns(struct drm_device *dev);
 void gen8_irq_power_well_post_enable(struct drm_i915_private *dev_priv);
 
 /* intel_crt.c */
@@ -774,11 +839,7 @@ void intel_ddi_clock_get(struct intel_encoder *encoder,
 			 struct intel_crtc_config *pipe_config);
 void intel_ddi_set_vc_payload_alloc(struct drm_crtc *crtc, bool state);
 
-/* intel_display.c */
-const char *intel_output_name(int output);
-bool intel_has_pending_fb_unpin(struct drm_device *dev);
-int intel_pch_rawclk(struct drm_device *dev);
-void intel_mark_busy(struct drm_device *dev);
+/* intel_frontbuffer.c */
 void intel_fb_obj_invalidate(struct drm_i915_gem_object *obj,
 			     struct intel_engine_cs *ring);
 void intel_frontbuffer_flip_prepare(struct drm_device *dev,
@@ -788,7 +849,7 @@ void intel_frontbuffer_flip_complete(struct drm_device *dev,
 void intel_frontbuffer_flush(struct drm_device *dev,
 			     unsigned frontbuffer_bits);
 /**
- * intel_frontbuffer_flip - prepare frontbuffer flip
+ * intel_frontbuffer_flip - synchronous frontbuffer flip
  * @dev: DRM device
  * @frontbuffer_bits: frontbuffer plane tracking bits
  *
@@ -806,6 +867,18 @@ void intel_frontbuffer_flip(struct drm_device *dev,
 }
 
 void intel_fb_obj_flush(struct drm_i915_gem_object *obj, bool retire);
+
+
+/* intel_audio.c */
+void intel_init_audio(struct drm_device *dev);
+void intel_audio_codec_enable(struct intel_encoder *encoder);
+void intel_audio_codec_disable(struct intel_encoder *encoder);
+
+/* intel_display.c */
+const char *intel_output_name(int output);
+bool intel_has_pending_fb_unpin(struct drm_device *dev);
+int intel_pch_rawclk(struct drm_device *dev);
+void intel_mark_busy(struct drm_device *dev);
 void intel_mark_idle(struct drm_device *dev);
 void intel_crtc_restore_mode(struct drm_crtc *crtc);
 void intel_crtc_control(struct drm_crtc *crtc, bool enable);
@@ -826,8 +899,12 @@ int intel_get_pipe_from_crtc_id(struct drm_device *dev, void *data,
 				struct drm_file *file_priv);
 enum transcoder intel_pipe_to_cpu_transcoder(struct drm_i915_private *dev_priv,
 					     enum pipe pipe);
-void intel_wait_for_vblank(struct drm_device *dev, int pipe);
-void intel_wait_for_pipe_off(struct drm_device *dev, int pipe);
+bool intel_pipe_has_type(struct intel_crtc *crtc, enum intel_output_type type);
+static inline void
+intel_wait_for_vblank(struct drm_device *dev, int pipe)
+{
+	drm_wait_one_vblank(dev, pipe);
+}
 int ironlake_get_lanes_required(int target_clock, int link_bw, int bpp);
 void vlv_wait_port_ready(struct drm_i915_private *dev_priv,
 			 struct intel_digital_port *dport);
@@ -837,8 +914,8 @@ bool intel_get_load_detect_pipe(struct drm_connector *connector,
 				struct drm_modeset_acquire_ctx *ctx);
 void intel_release_load_detect_pipe(struct drm_connector *connector,
 				    struct intel_load_detect_pipe *old);
-int intel_pin_and_fence_fb_obj(struct drm_device *dev,
-			       struct drm_i915_gem_object *obj,
+int intel_pin_and_fence_fb_obj(struct drm_plane *plane,
+			       struct drm_framebuffer *fb,
 			       struct intel_engine_cs *pipelined);
 void intel_unpin_fb_obj(struct drm_i915_gem_object *obj);
 struct drm_framebuffer *
@@ -848,6 +925,7 @@ __intel_framebuffer_create(struct drm_device *dev,
 void intel_prepare_page_flip(struct drm_device *dev, int plane);
 void intel_finish_page_flip(struct drm_device *dev, int pipe);
 void intel_finish_page_flip_plane(struct drm_device *dev, int plane);
+void intel_check_page_flip(struct drm_device *dev, int pipe);
 
 /* shared dpll functions */
 struct intel_shared_dpll *intel_crtc_to_shared_dpll(struct intel_crtc *crtc);
@@ -859,7 +937,13 @@ void assert_shared_dpll(struct drm_i915_private *dev_priv,
 struct intel_shared_dpll *intel_get_shared_dpll(struct intel_crtc *crtc);
 void intel_put_shared_dpll(struct intel_crtc *crtc);
 
+void vlv_force_pll_on(struct drm_device *dev, enum pipe pipe,
+		      const struct dpll *dpll);
+void vlv_force_pll_off(struct drm_device *dev, enum pipe pipe);
+
 /* modesetting asserts */
+void assert_panel_unlocked(struct drm_i915_private *dev_priv,
+			   enum pipe pipe);
 void assert_pll(struct drm_i915_private *dev_priv,
 		enum pipe pipe, bool state);
 #define assert_pll_enabled(d, p) assert_pll(d, p, true)
@@ -871,17 +955,17 @@ void assert_fdi_rx_pll(struct drm_i915_private *dev_priv,
 void assert_pipe(struct drm_i915_private *dev_priv, enum pipe pipe, bool state);
 #define assert_pipe_enabled(d, p) assert_pipe(d, p, true)
 #define assert_pipe_disabled(d, p) assert_pipe(d, p, false)
-void intel_write_eld(struct drm_encoder *encoder,
-		     struct drm_display_mode *mode);
 unsigned long intel_gen4_compute_page_offset(int *x, int *y,
 					     unsigned int tiling_mode,
 					     unsigned int bpp,
 					     unsigned int pitch);
-void intel_display_handle_reset(struct drm_device *dev);
+void intel_prepare_reset(struct drm_device *dev);
+void intel_finish_reset(struct drm_device *dev);
 void hsw_enable_pc8(struct drm_i915_private *dev_priv);
 void hsw_disable_pc8(struct drm_i915_private *dev_priv);
 void intel_dp_get_m_n(struct intel_crtc *crtc,
 		      struct intel_crtc_config *pipe_config);
+void intel_dp_set_m_n(struct intel_crtc *crtc);
 int intel_dotclock_calculate(int link_freq, const struct intel_link_m_n *m_n);
 void
 ironlake_check_encoder_dotclock(const struct intel_crtc_config *pipe_config,
@@ -889,14 +973,13 @@ ironlake_check_encoder_dotclock(const struct intel_crtc_config *pipe_config,
 bool intel_crtc_active(struct drm_crtc *crtc);
 void hsw_enable_ips(struct intel_crtc *crtc);
 void hsw_disable_ips(struct intel_crtc *crtc);
-void intel_display_set_init_power(struct drm_i915_private *dev, bool enable);
 enum intel_display_power_domain
 intel_display_port_power_domain(struct intel_encoder *intel_encoder);
 void intel_mode_from_pipe_config(struct drm_display_mode *mode,
 				 struct intel_crtc_config *pipe_config);
 int intel_format_to_fourcc(int format);
 void intel_crtc_wait_for_pending_flips(struct drm_crtc *crtc);
-
+void intel_modeset_preclose(struct drm_device *dev, struct drm_file *file);
 
 /* intel_dp.c */
 void intel_dp_init(struct drm_device *dev, int output_reg, enum port port);
@@ -917,24 +1000,18 @@ bool intel_dp_hpd_pulse(struct intel_digital_port *intel_dig_port,
 void intel_edp_backlight_on(struct intel_dp *intel_dp);
 void intel_edp_backlight_off(struct intel_dp *intel_dp);
 void intel_edp_panel_vdd_on(struct intel_dp *intel_dp);
-void intel_edp_panel_vdd_sanitize(struct intel_encoder *intel_encoder);
 void intel_edp_panel_on(struct intel_dp *intel_dp);
 void intel_edp_panel_off(struct intel_dp *intel_dp);
-void intel_edp_psr_enable(struct intel_dp *intel_dp);
-void intel_edp_psr_disable(struct intel_dp *intel_dp);
 void intel_dp_set_drrs_state(struct drm_device *dev, int refresh_rate);
-void intel_edp_psr_invalidate(struct drm_device *dev,
-			      unsigned frontbuffer_bits);
-void intel_edp_psr_flush(struct drm_device *dev,
-			 unsigned frontbuffer_bits);
-void intel_edp_psr_init(struct drm_device *dev);
-
-int intel_dp_handle_hpd_irq(struct intel_digital_port *digport, bool long_hpd);
 void intel_dp_add_properties(struct intel_dp *intel_dp, struct drm_connector *connector);
 void intel_dp_mst_suspend(struct drm_device *dev);
 void intel_dp_mst_resume(struct drm_device *dev);
 int intel_dp_max_link_bw(struct intel_dp *intel_dp);
 void intel_dp_hot_plug(struct intel_encoder *intel_encoder);
+void vlv_power_sequencer_reset(struct drm_i915_private *dev_priv);
+uint32_t intel_dp_pack_aux(const uint8_t *src, int src_bytes);
+void intel_dp_unpack_aux(uint32_t src, uint8_t *dst, int dst_bytes);
+
 /* intel_dp_mst.c */
 int intel_dp_mst_encoder_init(struct intel_digital_port *intel_dig_port, int conn_id);
 void intel_dp_mst_encoder_cleanup(struct intel_digital_port *intel_dig_port);
@@ -949,9 +1026,9 @@ void intel_dvo_init(struct drm_device *dev);
 /* legacy fbdev emulation in intel_fbdev.c */
 #ifdef CONFIG_DRM_I915_FBDEV
 extern int intel_fbdev_init(struct drm_device *dev);
-extern void intel_fbdev_initial_config(struct drm_device *dev);
+extern void intel_fbdev_initial_config(void *data, async_cookie_t cookie);
 extern void intel_fbdev_fini(struct drm_device *dev);
-extern void intel_fbdev_set_suspend(struct drm_device *dev, int state);
+extern void intel_fbdev_set_suspend(struct drm_device *dev, int state, bool synchronous);
 extern void intel_fbdev_output_poll_changed(struct drm_device *dev);
 extern void intel_fbdev_restore_mode(struct drm_device *dev);
 #else
@@ -960,7 +1037,7 @@ static inline int intel_fbdev_init(struct drm_device *dev)
 	return 0;
 }
 
-static inline void intel_fbdev_initial_config(struct drm_device *dev)
+static inline void intel_fbdev_initial_config(void *data, async_cookie_t cookie)
 {
 }
 
@@ -968,7 +1045,7 @@ static inline void intel_fbdev_fini(struct drm_device *dev)
 {
 }
 
-static inline void intel_fbdev_set_suspend(struct drm_device *dev, int state)
+static inline void intel_fbdev_set_suspend(struct drm_device *dev, int state, bool synchronous)
 {
 }
 
@@ -1024,7 +1101,7 @@ void intel_gmch_panel_fitting(struct intel_crtc *crtc,
 			      int fitting_mode);
 void intel_panel_set_backlight_acpi(struct intel_connector *connector,
 				    u32 level, u32 max);
-int intel_panel_setup_backlight(struct drm_connector *connector);
+int intel_panel_setup_backlight(struct drm_connector *connector, enum pipe pipe);
 void intel_panel_enable_backlight(struct intel_connector *connector);
 void intel_panel_disable_backlight(struct intel_connector *connector);
 void intel_panel_destroy_backlight(struct drm_connector *connector);
@@ -1034,6 +1111,41 @@ extern struct drm_display_mode *intel_find_panel_downclock(
 				struct drm_device *dev,
 				struct drm_display_mode *fixed_mode,
 				struct drm_connector *connector);
+void intel_backlight_register(struct drm_device *dev);
+void intel_backlight_unregister(struct drm_device *dev);
+
+
+/* intel_psr.c */
+bool intel_psr_is_enabled(struct drm_device *dev);
+void intel_psr_enable(struct intel_dp *intel_dp);
+void intel_psr_disable(struct intel_dp *intel_dp);
+void intel_psr_invalidate(struct drm_device *dev,
+			      unsigned frontbuffer_bits);
+void intel_psr_flush(struct drm_device *dev,
+			 unsigned frontbuffer_bits);
+void intel_psr_init(struct drm_device *dev);
+
+/* intel_runtime_pm.c */
+int intel_power_domains_init(struct drm_i915_private *);
+void intel_power_domains_fini(struct drm_i915_private *);
+void intel_power_domains_init_hw(struct drm_i915_private *dev_priv);
+void intel_runtime_pm_enable(struct drm_i915_private *dev_priv);
+
+bool intel_display_power_is_enabled(struct drm_i915_private *dev_priv,
+				    enum intel_display_power_domain domain);
+bool __intel_display_power_is_enabled(struct drm_i915_private *dev_priv,
+				      enum intel_display_power_domain domain);
+void intel_display_power_get(struct drm_i915_private *dev_priv,
+			     enum intel_display_power_domain domain);
+void intel_display_power_put(struct drm_i915_private *dev_priv,
+			     enum intel_display_power_domain domain);
+void intel_aux_display_runtime_get(struct drm_i915_private *dev_priv);
+void intel_aux_display_runtime_put(struct drm_i915_private *dev_priv);
+void intel_runtime_pm_get(struct drm_i915_private *dev_priv);
+void intel_runtime_pm_get_noresume(struct drm_i915_private *dev_priv);
+void intel_runtime_pm_put(struct drm_i915_private *dev_priv);
+
+void intel_display_set_init_power(struct drm_i915_private *dev, bool enable);
 
 /* intel_pm.c */
 void intel_init_clock_gating(struct drm_device *dev);
@@ -1052,17 +1164,6 @@ bool intel_fbc_enabled(struct drm_device *dev);
 void intel_update_fbc(struct drm_device *dev);
 void intel_gpu_ips_init(struct drm_i915_private *dev_priv);
 void intel_gpu_ips_teardown(void);
-int intel_power_domains_init(struct drm_i915_private *);
-void intel_power_domains_remove(struct drm_i915_private *);
-bool intel_display_power_enabled(struct drm_i915_private *dev_priv,
-				 enum intel_display_power_domain domain);
-bool intel_display_power_enabled_unlocked(struct drm_i915_private *dev_priv,
-					  enum intel_display_power_domain domain);
-void intel_display_power_get(struct drm_i915_private *dev_priv,
-			     enum intel_display_power_domain domain);
-void intel_display_power_put(struct drm_i915_private *dev_priv,
-			     enum intel_display_power_domain domain);
-void intel_power_domains_init_hw(struct drm_i915_private *dev_priv);
 void intel_init_gt_powersave(struct drm_device *dev);
 void intel_cleanup_gt_powersave(struct drm_device *dev);
 void intel_enable_gt_powersave(struct drm_device *dev);
@@ -1073,14 +1174,10 @@ void ironlake_teardown_rc6(struct drm_device *dev);
 void gen6_update_ring_freq(struct drm_device *dev);
 void gen6_rps_idle(struct drm_i915_private *dev_priv);
 void gen6_rps_boost(struct drm_i915_private *dev_priv);
-void intel_aux_display_runtime_get(struct drm_i915_private *dev_priv);
-void intel_aux_display_runtime_put(struct drm_i915_private *dev_priv);
-void intel_runtime_pm_get(struct drm_i915_private *dev_priv);
-void intel_runtime_pm_get_noresume(struct drm_i915_private *dev_priv);
-void intel_runtime_pm_put(struct drm_i915_private *dev_priv);
-void intel_init_runtime_pm(struct drm_i915_private *dev_priv);
-void intel_fini_runtime_pm(struct drm_i915_private *dev_priv);
 void ilk_wm_get_hw_state(struct drm_device *dev);
+void skl_wm_get_hw_state(struct drm_device *dev);
+void skl_ddb_get_hw_state(struct drm_i915_private *dev_priv,
+			  struct skl_ddb_allocation *ddb /* out */);
 
 
 /* intel_sdvo.c */
@@ -1091,13 +1188,18 @@ bool intel_sdvo_init(struct drm_device *dev, uint32_t sdvo_reg, bool is_sdvob);
 int intel_plane_init(struct drm_device *dev, enum pipe pipe, int plane);
 void intel_flush_primary_plane(struct drm_i915_private *dev_priv,
 			       enum plane plane);
-void intel_plane_restore(struct drm_plane *plane);
+int intel_plane_set_property(struct drm_plane *plane,
+			     struct drm_property *prop,
+			     uint64_t val);
+int intel_plane_restore(struct drm_plane *plane);
 void intel_plane_disable(struct drm_plane *plane);
 int intel_sprite_set_colorkey(struct drm_device *dev, void *data,
 			      struct drm_file *file_priv);
 int intel_sprite_get_colorkey(struct drm_device *dev, void *data,
 			      struct drm_file *file_priv);
-
+bool intel_pipe_update_start(struct intel_crtc *crtc,
+			     uint32_t *start_vbl_count);
+void intel_pipe_update_end(struct intel_crtc *crtc, u32 start_vbl_count);
 
 /* intel_tv.c */
 void intel_tv_init(struct drm_device *dev);
diff --git a/drivers/gpu/drm/i915/intel_dsi.c b/drivers/gpu/drm/i915/intel_dsi.c
index 670c29a7b5dd..0b184079de14 100644
--- a/drivers/gpu/drm/i915/intel_dsi.c
+++ b/drivers/gpu/drm/i915/intel_dsi.c
@@ -184,7 +184,7 @@ static void intel_dsi_pre_enable(struct intel_encoder *encoder)
 
 	/* update the hw state for DPLL */
 	intel_crtc->config.dpll_hw_state.dpll = DPLL_INTEGRATED_CLOCK_VLV |
-						DPLL_REFA_CLK_ENABLE_VLV;
+		DPLL_REFA_CLK_ENABLE_VLV;
 
 	tmp = I915_READ(DSPCLK_GATE_D);
 	tmp |= DPOUNIT_CLOCK_GATE_DISABLE;
@@ -259,8 +259,8 @@ static void intel_dsi_disable(struct intel_encoder *encoder)
 	temp = I915_READ(MIPI_CTRL(pipe));
 	temp &= ~ESCAPE_CLOCK_DIVIDER_MASK;
 	I915_WRITE(MIPI_CTRL(pipe), temp |
-			intel_dsi->escape_clk_div <<
-			ESCAPE_CLOCK_DIVIDER_SHIFT);
+		   intel_dsi->escape_clk_div <<
+		   ESCAPE_CLOCK_DIVIDER_SHIFT);
 
 	I915_WRITE(MIPI_EOT_DISABLE(pipe), CLOCKSTOP);
 
@@ -297,7 +297,7 @@ static void intel_dsi_clear_device_ready(struct intel_encoder *encoder)
 	usleep_range(2000, 2500);
 
 	if (wait_for(((I915_READ(MIPI_PORT_CTRL(pipe)) & AFE_LATCHOUT)
-					== 0x00000), 30))
+		      == 0x00000), 30))
 		DRM_ERROR("DSI LP not going Low\n");
 
 	val = I915_READ(MIPI_PORT_CTRL(pipe));
@@ -344,7 +344,7 @@ static bool intel_dsi_get_hw_state(struct intel_encoder *encoder,
 	DRM_DEBUG_KMS("\n");
 
 	power_domain = intel_display_port_power_domain(encoder);
-	if (!intel_display_power_enabled(dev_priv, power_domain))
+	if (!intel_display_power_is_enabled(dev_priv, power_domain))
 		return false;
 
 	/* XXX: this only works for one DSI output */
@@ -423,9 +423,11 @@ static u16 txclkesc(u32 divider, unsigned int us)
 }
 
 /* return pixels in terms of txbyteclkhs */
-static u16 txbyteclkhs(u16 pixels, int bpp, int lane_count)
+static u16 txbyteclkhs(u16 pixels, int bpp, int lane_count,
+		       u16 burst_mode_ratio)
 {
-	return DIV_ROUND_UP(DIV_ROUND_UP(pixels * bpp, 8), lane_count);
+	return DIV_ROUND_UP(DIV_ROUND_UP(pixels * bpp * burst_mode_ratio,
+					 8 * 100), lane_count);
 }
 
 static void set_dsi_timings(struct drm_encoder *encoder,
@@ -451,10 +453,12 @@ static void set_dsi_timings(struct drm_encoder *encoder,
 	vbp = mode->vtotal - mode->vsync_end;
 
 	/* horizontal values are in terms of high speed byte clock */
-	hactive = txbyteclkhs(hactive, bpp, lane_count);
-	hfp = txbyteclkhs(hfp, bpp, lane_count);
-	hsync = txbyteclkhs(hsync, bpp, lane_count);
-	hbp = txbyteclkhs(hbp, bpp, lane_count);
+	hactive = txbyteclkhs(hactive, bpp, lane_count,
+			      intel_dsi->burst_mode_ratio);
+	hfp = txbyteclkhs(hfp, bpp, lane_count, intel_dsi->burst_mode_ratio);
+	hsync = txbyteclkhs(hsync, bpp, lane_count,
+			    intel_dsi->burst_mode_ratio);
+	hbp = txbyteclkhs(hbp, bpp, lane_count, intel_dsi->burst_mode_ratio);
 
 	I915_WRITE(MIPI_HACTIVE_AREA_COUNT(pipe), hactive);
 	I915_WRITE(MIPI_HFP_COUNT(pipe), hfp);
@@ -541,12 +545,14 @@ static void intel_dsi_prepare(struct intel_encoder *intel_encoder)
 	    intel_dsi->video_mode_format == VIDEO_MODE_BURST) {
 		I915_WRITE(MIPI_HS_TX_TIMEOUT(pipe),
 			   txbyteclkhs(adjusted_mode->htotal, bpp,
-				       intel_dsi->lane_count) + 1);
+				       intel_dsi->lane_count,
+				       intel_dsi->burst_mode_ratio) + 1);
 	} else {
 		I915_WRITE(MIPI_HS_TX_TIMEOUT(pipe),
 			   txbyteclkhs(adjusted_mode->vtotal *
 				       adjusted_mode->htotal,
-				       bpp, intel_dsi->lane_count) + 1);
+				       bpp, intel_dsi->lane_count,
+				       intel_dsi->burst_mode_ratio) + 1);
 	}
 	I915_WRITE(MIPI_LP_RX_TIMEOUT(pipe), intel_dsi->lp_rx_timeout);
 	I915_WRITE(MIPI_TURN_AROUND_TIMEOUT(pipe), intel_dsi->turn_arnd_val);
@@ -576,7 +582,7 @@ static void intel_dsi_prepare(struct intel_encoder *intel_encoder)
 	 * XXX: write MIPI_STOP_STATE_STALL?
 	 */
 	I915_WRITE(MIPI_HIGH_LOW_SWITCH_COUNT(pipe),
-						intel_dsi->hs_to_lp_count);
+		   intel_dsi->hs_to_lp_count);
 
 	/* XXX: low power clock equivalence in terms of byte clock. the number
 	 * of byte clocks occupied in one low power clock. based on txbyteclkhs
@@ -601,10 +607,10 @@ static void intel_dsi_prepare(struct intel_encoder *intel_encoder)
 		 * 64 like 1366 x 768. Enable RANDOM resolution support for such
 		 * panels by default */
 		I915_WRITE(MIPI_VIDEO_MODE_FORMAT(pipe),
-				intel_dsi->video_frmt_cfg_bits |
-				intel_dsi->video_mode_format |
-				IP_TG_CONFIG |
-				RANDOM_DPI_DISPLAY_RESOLUTION);
+			   intel_dsi->video_frmt_cfg_bits |
+			   intel_dsi->video_mode_format |
+			   IP_TG_CONFIG |
+			   RANDOM_DPI_DISPLAY_RESOLUTION);
 }
 
 static void intel_dsi_pre_pll_enable(struct intel_encoder *encoder)
diff --git a/drivers/gpu/drm/i915/intel_dsi.h b/drivers/gpu/drm/i915/intel_dsi.h
index fd51867fd0d3..657eb5c1b9d8 100644
--- a/drivers/gpu/drm/i915/intel_dsi.h
+++ b/drivers/gpu/drm/i915/intel_dsi.h
@@ -116,6 +116,8 @@ struct intel_dsi {
 	u16 clk_hs_to_lp_count;
 
 	u16 init_count;
+	u32 pclk;
+	u16 burst_mode_ratio;
 
 	/* all delays in ms */
 	u16 backlight_off_delay;
diff --git a/drivers/gpu/drm/i915/intel_dsi_cmd.c b/drivers/gpu/drm/i915/intel_dsi_cmd.c
index 7f1430ac8543..f4767fd2ebeb 100644
--- a/drivers/gpu/drm/i915/intel_dsi_cmd.c
+++ b/drivers/gpu/drm/i915/intel_dsi_cmd.c
@@ -430,7 +430,7 @@ void wait_for_dsi_fifo_empty(struct intel_dsi *intel_dsi)
 	u32 mask;
 
 	mask = LP_CTRL_FIFO_EMPTY | HS_CTRL_FIFO_EMPTY |
-					LP_DATA_FIFO_EMPTY | HS_DATA_FIFO_EMPTY;
+		LP_DATA_FIFO_EMPTY | HS_DATA_FIFO_EMPTY;
 
 	if (wait_for((I915_READ(MIPI_GEN_FIFO_STAT(pipe)) & mask) == mask, 100))
 		DRM_ERROR("DPI FIFOs are not empty\n");
diff --git a/drivers/gpu/drm/i915/intel_dsi_panel_vbt.c b/drivers/gpu/drm/i915/intel_dsi_panel_vbt.c
index 47c7584a4aa0..f6bdd44069ce 100644
--- a/drivers/gpu/drm/i915/intel_dsi_panel_vbt.c
+++ b/drivers/gpu/drm/i915/intel_dsi_panel_vbt.c
@@ -271,6 +271,8 @@ static bool generic_init(struct intel_dsi_device *dsi)
 	u32 ths_prepare_ns, tclk_trail_ns;
 	u32 tclk_prepare_clkzero, ths_prepare_hszero;
 	u32 lp_to_hs_switch, hs_to_lp_switch;
+	u32 pclk, computed_ddr;
+	u16 burst_mode_ratio;
 
 	DRM_DEBUG_KMS("\n");
 
@@ -284,8 +286,6 @@ static bool generic_init(struct intel_dsi_device *dsi)
 	else if (intel_dsi->pixel_format == VID_MODE_FORMAT_RGB565)
 		bits_per_pixel = 16;
 
-	bitrate = (mode->clock * bits_per_pixel) / intel_dsi->lane_count;
-
 	intel_dsi->operation_mode = mipi_config->is_cmd_mode;
 	intel_dsi->video_mode_format = mipi_config->video_transfer_mode;
 	intel_dsi->escape_clk_div = mipi_config->byte_clk_sel;
@@ -297,6 +297,40 @@ static bool generic_init(struct intel_dsi_device *dsi)
 	intel_dsi->video_frmt_cfg_bits =
 		mipi_config->bta_enabled ? DISABLE_VIDEO_BTA : 0;
 
+	pclk = mode->clock;
+
+	/* Burst Mode Ratio
+	 * Target ddr frequency from VBT / non burst ddr freq
+	 * multiply by 100 to preserve remainder
+	 */
+	if (intel_dsi->video_mode_format == VIDEO_MODE_BURST) {
+		if (mipi_config->target_burst_mode_freq) {
+			computed_ddr =
+				(pclk * bits_per_pixel) / intel_dsi->lane_count;
+
+			if (mipi_config->target_burst_mode_freq <
+								computed_ddr) {
+				DRM_ERROR("Burst mode freq is less than computed\n");
+				return false;
+			}
+
+			burst_mode_ratio = DIV_ROUND_UP(
+				mipi_config->target_burst_mode_freq * 100,
+				computed_ddr);
+
+			pclk = DIV_ROUND_UP(pclk * burst_mode_ratio, 100);
+		} else {
+			DRM_ERROR("Burst mode target is not set\n");
+			return false;
+		}
+	} else
+		burst_mode_ratio = 100;
+
+	intel_dsi->burst_mode_ratio = burst_mode_ratio;
+	intel_dsi->pclk = pclk;
+
+	bitrate = (pclk * bits_per_pixel) / intel_dsi->lane_count;
+
 	switch (intel_dsi->escape_clk_div) {
 	case 0:
 		tlpx_ns = 50;
diff --git a/drivers/gpu/drm/i915/intel_dsi_pll.c b/drivers/gpu/drm/i915/intel_dsi_pll.c
index d8bb1ea2f0da..fa7a6ca34cd6 100644
--- a/drivers/gpu/drm/i915/intel_dsi_pll.c
+++ b/drivers/gpu/drm/i915/intel_dsi_pll.c
@@ -134,8 +134,7 @@ static u32 dsi_rr_formula(const struct drm_display_mode *mode,
 #else
 
 /* Get DSI clock from pixel clock */
-static u32 dsi_clk_from_pclk(const struct drm_display_mode *mode,
-			  int pixel_format, int lane_count)
+static u32 dsi_clk_from_pclk(u32 pclk, int pixel_format, int lane_count)
 {
 	u32 dsi_clk_khz;
 	u32 bpp;
@@ -156,7 +155,7 @@ static u32 dsi_clk_from_pclk(const struct drm_display_mode *mode,
 
 	/* DSI data rate = pixel clock * bits per pixel / lane count
 	   pixel clock is converted from KHz to Hz */
-	dsi_clk_khz = DIV_ROUND_CLOSEST(mode->clock * bpp, lane_count);
+	dsi_clk_khz = DIV_ROUND_CLOSEST(pclk * bpp, lane_count);
 
 	return dsi_clk_khz;
 }
@@ -191,7 +190,7 @@ static int dsi_calc_mnp(u32 dsi_clk, struct dsi_mnp *dsi_mnp)
 	for (m = 62; m <= 92; m++) {
 		for (p = 2; p <= 6; p++) {
 			/* Find the optimal m and p divisors
-			with minimal error +/- the required clock */
+			   with minimal error +/- the required clock */
 			calc_dsi_clk = (m * ref_clk) / p;
 			if (calc_dsi_clk == target_dsi_clk) {
 				calc_m = m;
@@ -228,15 +227,13 @@ static int dsi_calc_mnp(u32 dsi_clk, struct dsi_mnp *dsi_mnp)
 static void vlv_configure_dsi_pll(struct intel_encoder *encoder)
 {
 	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
-	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
-	const struct drm_display_mode *mode = &intel_crtc->config.adjusted_mode;
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
 	int ret;
 	struct dsi_mnp dsi_mnp;
 	u32 dsi_clk;
 
-	dsi_clk = dsi_clk_from_pclk(mode, intel_dsi->pixel_format,
-						intel_dsi->lane_count);
+	dsi_clk = dsi_clk_from_pclk(intel_dsi->pclk, intel_dsi->pixel_format,
+				    intel_dsi->lane_count);
 
 	ret = dsi_calc_mnp(dsi_clk, &dsi_mnp);
 	if (ret) {
@@ -318,8 +315,8 @@ static void assert_bpp_mismatch(int pixel_format, int pipe_bpp)
 	}
 
 	WARN(bpp != pipe_bpp,
-		"bpp match assertion failure (expected %d, current %d)\n",
-		bpp, pipe_bpp);
+	     "bpp match assertion failure (expected %d, current %d)\n",
+	     bpp, pipe_bpp);
 }
 
 u32 vlv_get_dsi_pclk(struct intel_encoder *encoder, int pipe_bpp)
diff --git a/drivers/gpu/drm/i915/intel_dvo.c b/drivers/gpu/drm/i915/intel_dvo.c
index 56b47d2ffaf7..e40e3df33517 100644
--- a/drivers/gpu/drm/i915/intel_dvo.c
+++ b/drivers/gpu/drm/i915/intel_dvo.c
@@ -85,7 +85,7 @@ static const struct intel_dvo_device intel_dvo_devices[] = {
 	{
 	        .type = INTEL_DVO_CHIP_TMDS,
 		.name = "ns2501",
-		.dvo_reg = DVOC,
+		.dvo_reg = DVOB,
 		.slave_addr = NS2501_ADDR,
 		.dev_ops = &ns2501_ops,
        }
@@ -185,12 +185,13 @@ static void intel_enable_dvo(struct intel_encoder *encoder)
 	u32 dvo_reg = intel_dvo->dev.dvo_reg;
 	u32 temp = I915_READ(dvo_reg);
 
-	I915_WRITE(dvo_reg, temp | DVO_ENABLE);
-	I915_READ(dvo_reg);
 	intel_dvo->dev.dev_ops->mode_set(&intel_dvo->dev,
 					 &crtc->config.requested_mode,
 					 &crtc->config.adjusted_mode);
 
+	I915_WRITE(dvo_reg, temp | DVO_ENABLE);
+	I915_READ(dvo_reg);
+
 	intel_dvo->dev.dev_ops->dpms(&intel_dvo->dev, true);
 }
 
@@ -226,10 +227,6 @@ static void intel_dvo_dpms(struct drm_connector *connector, int mode)
 
 		intel_crtc_update_dpms(crtc);
 
-		intel_dvo->dev.dev_ops->mode_set(&intel_dvo->dev,
-						 &config->requested_mode,
-						 &config->adjusted_mode);
-
 		intel_dvo->dev.dev_ops->dpms(&intel_dvo->dev, true);
 	} else {
 		intel_dvo->dev.dev_ops->dpms(&intel_dvo->dev, false);
diff --git a/drivers/gpu/drm/i915/intel_fbdev.c b/drivers/gpu/drm/i915/intel_fbdev.c
index f475414671d8..850cf7d6578c 100644
--- a/drivers/gpu/drm/i915/intel_fbdev.c
+++ b/drivers/gpu/drm/i915/intel_fbdev.c
@@ -24,8 +24,10 @@
  *     David Airlie
  */
 
+#include <linux/async.h>
 #include <linux/module.h>
 #include <linux/kernel.h>
+#include <linux/console.h>
 #include <linux/errno.h>
 #include <linux/string.h>
 #include <linux/mm.h>
@@ -117,25 +119,25 @@ static int intelfb_alloc(struct drm_fb_helper *helper,
 		goto out;
 	}
 
-	/* Flush everything out, we'll be doing GTT only from now on */
-	ret = intel_pin_and_fence_fb_obj(dev, obj, NULL);
-	if (ret) {
-		DRM_ERROR("failed to pin obj: %d\n", ret);
-		goto out_unref;
-	}
-
 	fb = __intel_framebuffer_create(dev, &mode_cmd, obj);
 	if (IS_ERR(fb)) {
 		ret = PTR_ERR(fb);
-		goto out_unpin;
+		goto out_unref;
+	}
+
+	/* Flush everything out, we'll be doing GTT only from now on */
+	ret = intel_pin_and_fence_fb_obj(NULL, fb, NULL);
+	if (ret) {
+		DRM_ERROR("failed to pin obj: %d\n", ret);
+		goto out_fb;
 	}
 
 	ifbdev->fb = to_intel_framebuffer(fb);
 
 	return 0;
 
-out_unpin:
-	i915_gem_object_ggtt_unpin(obj);
+out_fb:
+	drm_framebuffer_remove(fb);
 out_unref:
 	drm_gem_object_unreference(&obj->base);
 out:
@@ -322,6 +324,7 @@ intel_fb_helper_crtc(struct drm_fb_helper *fb_helper, struct drm_crtc *crtc)
 static bool intel_fb_initial_config(struct drm_fb_helper *fb_helper,
 				    struct drm_fb_helper_crtc **crtcs,
 				    struct drm_display_mode **modes,
+				    struct drm_fb_offset *offsets,
 				    bool *enabled, int width, int height)
 {
 	struct drm_device *dev = fb_helper->dev;
@@ -330,24 +333,8 @@ static bool intel_fb_initial_config(struct drm_fb_helper *fb_helper,
 	bool fallback = true;
 	int num_connectors_enabled = 0;
 	int num_connectors_detected = 0;
-
-	/*
-	 * If the user specified any force options, just bail here
-	 * and use that config.
-	 */
-	for (i = 0; i < fb_helper->connector_count; i++) {
-		struct drm_fb_helper_connector *fb_conn;
-		struct drm_connector *connector;
-
-		fb_conn = fb_helper->connector_info[i];
-		connector = fb_conn->connector;
-
-		if (!enabled[i])
-			continue;
-
-		if (connector->force != DRM_FORCE_UNSPECIFIED)
-			return false;
-	}
+	uint64_t conn_configured = 0, mask;
+	int pass = 0;
 
 	save_enabled = kcalloc(dev->mode_config.num_connector, sizeof(bool),
 			       GFP_KERNEL);
@@ -355,7 +342,8 @@ static bool intel_fb_initial_config(struct drm_fb_helper *fb_helper,
 		return false;
 
 	memcpy(save_enabled, enabled, dev->mode_config.num_connector);
-
+	mask = (1 << fb_helper->connector_count) - 1;
+retry:
 	for (i = 0; i < fb_helper->connector_count; i++) {
 		struct drm_fb_helper_connector *fb_conn;
 		struct drm_connector *connector;
@@ -365,20 +353,38 @@ static bool intel_fb_initial_config(struct drm_fb_helper *fb_helper,
 		fb_conn = fb_helper->connector_info[i];
 		connector = fb_conn->connector;
 
+		if (conn_configured & (1 << i))
+			continue;
+
+		if (pass == 0 && !connector->has_tile)
+			continue;
+
 		if (connector->status == connector_status_connected)
 			num_connectors_detected++;
 
 		if (!enabled[i]) {
 			DRM_DEBUG_KMS("connector %s not enabled, skipping\n",
 				      connector->name);
+			conn_configured |= (1 << i);
+			continue;
+		}
+
+		if (connector->force == DRM_FORCE_OFF) {
+			DRM_DEBUG_KMS("connector %s is disabled by user, skipping\n",
+				      connector->name);
+			enabled[i] = false;
 			continue;
 		}
 
 		encoder = connector->encoder;
 		if (!encoder || WARN_ON(!encoder->crtc)) {
+			if (connector->force > DRM_FORCE_OFF)
+				goto bail;
+
 			DRM_DEBUG_KMS("connector %s has no encoder or crtc, skipping\n",
 				      connector->name);
 			enabled[i] = false;
+			conn_configured |= (1 << i);
 			continue;
 		}
 
@@ -394,8 +400,7 @@ static bool intel_fb_initial_config(struct drm_fb_helper *fb_helper,
 		for (j = 0; j < fb_helper->connector_count; j++) {
 			if (crtcs[j] == new_crtc) {
 				DRM_DEBUG_KMS("fallback: cloned configuration\n");
-				fallback = true;
-				goto out;
+				goto bail;
 			}
 		}
 
@@ -407,8 +412,8 @@ static bool intel_fb_initial_config(struct drm_fb_helper *fb_helper,
 
 		/* try for preferred next */
 		if (!modes[i]) {
-			DRM_DEBUG_KMS("looking for preferred mode on connector %s\n",
-				      connector->name);
+			DRM_DEBUG_KMS("looking for preferred mode on connector %s %d\n",
+				      connector->name, connector->has_tile);
 			modes[i] = drm_has_preferred_mode(fb_conn, width,
 							  height);
 		}
@@ -451,6 +456,12 @@ static bool intel_fb_initial_config(struct drm_fb_helper *fb_helper,
 			      modes[i]->flags & DRM_MODE_FLAG_INTERLACE ? "i" :"");
 
 		fallback = false;
+		conn_configured |= (1 << i);
+	}
+
+	if ((conn_configured & mask) != mask) {
+		pass++;
+		goto retry;
 	}
 
 	/*
@@ -466,8 +477,8 @@ static bool intel_fb_initial_config(struct drm_fb_helper *fb_helper,
 		fallback = true;
 	}
 
-out:
 	if (fallback) {
+bail:
 		DRM_DEBUG_KMS("Not using firmware configuration\n");
 		memcpy(enabled, save_enabled, dev->mode_config.num_connector);
 		kfree(save_enabled);
@@ -636,6 +647,15 @@ out:
 	return false;
 }
 
+static void intel_fbdev_suspend_worker(struct work_struct *work)
+{
+	intel_fbdev_set_suspend(container_of(work,
+					     struct drm_i915_private,
+					     fbdev_suspend_work)->dev,
+				FBINFO_STATE_RUNNING,
+				true);
+}
+
 int intel_fbdev_init(struct drm_device *dev)
 {
 	struct intel_fbdev *ifbdev;
@@ -662,14 +682,16 @@ int intel_fbdev_init(struct drm_device *dev)
 	}
 
 	dev_priv->fbdev = ifbdev;
+	INIT_WORK(&dev_priv->fbdev_suspend_work, intel_fbdev_suspend_worker);
+
 	drm_fb_helper_single_add_all_connectors(&ifbdev->helper);
 
 	return 0;
 }
 
-void intel_fbdev_initial_config(struct drm_device *dev)
+void intel_fbdev_initial_config(void *data, async_cookie_t cookie)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = data;
 	struct intel_fbdev *ifbdev = dev_priv->fbdev;
 
 	/* Due to peculiar init order wrt to hpd handling this is separate. */
@@ -682,12 +704,15 @@ void intel_fbdev_fini(struct drm_device *dev)
 	if (!dev_priv->fbdev)
 		return;
 
+	flush_work(&dev_priv->fbdev_suspend_work);
+
+	async_synchronize_full();
 	intel_fbdev_destroy(dev, dev_priv->fbdev);
 	kfree(dev_priv->fbdev);
 	dev_priv->fbdev = NULL;
 }
 
-void intel_fbdev_set_suspend(struct drm_device *dev, int state)
+void intel_fbdev_set_suspend(struct drm_device *dev, int state, bool synchronous)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_fbdev *ifbdev = dev_priv->fbdev;
@@ -698,6 +723,33 @@ void intel_fbdev_set_suspend(struct drm_device *dev, int state)
 
 	info = ifbdev->helper.fbdev;
 
+	if (synchronous) {
+		/* Flush any pending work to turn the console on, and then
+		 * wait to turn it off. It must be synchronous as we are
+		 * about to suspend or unload the driver.
+		 *
+		 * Note that from within the work-handler, we cannot flush
+		 * ourselves, so only flush outstanding work upon suspend!
+		 */
+		if (state != FBINFO_STATE_RUNNING)
+			flush_work(&dev_priv->fbdev_suspend_work);
+		console_lock();
+	} else {
+		/*
+		 * The console lock can be pretty contented on resume due
+		 * to all the printk activity.  Try to keep it out of the hot
+		 * path of resume if possible.
+		 */
+		WARN_ON(state != FBINFO_STATE_RUNNING);
+		if (!console_trylock()) {
+			/* Don't block our own workqueue as this can
+			 * be run in parallel with other i915.ko tasks.
+			 */
+			schedule_work(&dev_priv->fbdev_suspend_work);
+			return;
+		}
+	}
+
 	/* On resume from hibernation: If the object is shmemfs backed, it has
 	 * been restored from swap. If the object is stolen however, it will be
 	 * full of whatever garbage was left in there.
@@ -706,6 +758,7 @@ void intel_fbdev_set_suspend(struct drm_device *dev, int state)
 		memset_io(info->screen_base, 0, info->screen_size);
 
 	fb_set_suspend(info, state);
+	console_unlock();
 }
 
 void intel_fbdev_output_poll_changed(struct drm_device *dev)
diff --git a/drivers/gpu/drm/i915/intel_fifo_underrun.c b/drivers/gpu/drm/i915/intel_fifo_underrun.c
new file mode 100644
index 000000000000..77af512d2d35
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_fifo_underrun.c
@@ -0,0 +1,381 @@
+/*
+ * Copyright © 2014 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ * Authors:
+ *    Daniel Vetter <daniel.vetter@ffwll.ch>
+ *
+ */
+
+#include "i915_drv.h"
+#include "intel_drv.h"
+
+/**
+ * DOC: fifo underrun handling
+ *
+ * The i915 driver checks for display fifo underruns using the interrupt signals
+ * provided by the hardware. This is enabled by default and fairly useful to
+ * debug display issues, especially watermark settings.
+ *
+ * If an underrun is detected this is logged into dmesg. To avoid flooding logs
+ * and occupying the cpu underrun interrupts are disabled after the first
+ * occurrence until the next modeset on a given pipe.
+ *
+ * Note that underrun detection on gmch platforms is a bit more ugly since there
+ * is no interrupt (despite that the signalling bit is in the PIPESTAT pipe
+ * interrupt register). Also on some other platforms underrun interrupts are
+ * shared, which means that if we detect an underrun we need to disable underrun
+ * reporting on all pipes.
+ *
+ * The code also supports underrun detection on the PCH transcoder.
+ */
+
+static bool ivb_can_enable_err_int(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *crtc;
+	enum pipe pipe;
+
+	assert_spin_locked(&dev_priv->irq_lock);
+
+	for_each_pipe(dev_priv, pipe) {
+		crtc = to_intel_crtc(dev_priv->pipe_to_crtc_mapping[pipe]);
+
+		if (crtc->cpu_fifo_underrun_disabled)
+			return false;
+	}
+
+	return true;
+}
+
+static bool cpt_can_enable_serr_int(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum pipe pipe;
+	struct intel_crtc *crtc;
+
+	assert_spin_locked(&dev_priv->irq_lock);
+
+	for_each_pipe(dev_priv, pipe) {
+		crtc = to_intel_crtc(dev_priv->pipe_to_crtc_mapping[pipe]);
+
+		if (crtc->pch_fifo_underrun_disabled)
+			return false;
+	}
+
+	return true;
+}
+
+/**
+ * i9xx_check_fifo_underruns - check for fifo underruns
+ * @dev_priv: i915 device instance
+ *
+ * This function checks for fifo underruns on GMCH platforms. This needs to be
+ * done manually on modeset to make sure that we catch all underruns since they
+ * do not generate an interrupt by themselves on these platforms.
+ */
+void i9xx_check_fifo_underruns(struct drm_i915_private *dev_priv)
+{
+	struct intel_crtc *crtc;
+
+	spin_lock_irq(&dev_priv->irq_lock);
+
+	for_each_intel_crtc(dev_priv->dev, crtc) {
+		u32 reg = PIPESTAT(crtc->pipe);
+		u32 pipestat;
+
+		if (crtc->cpu_fifo_underrun_disabled)
+			continue;
+
+		pipestat = I915_READ(reg) & 0xffff0000;
+		if ((pipestat & PIPE_FIFO_UNDERRUN_STATUS) == 0)
+			continue;
+
+		I915_WRITE(reg, pipestat | PIPE_FIFO_UNDERRUN_STATUS);
+		POSTING_READ(reg);
+
+		DRM_ERROR("pipe %c underrun\n", pipe_name(crtc->pipe));
+	}
+
+	spin_unlock_irq(&dev_priv->irq_lock);
+}
+
+static void i9xx_set_fifo_underrun_reporting(struct drm_device *dev,
+					     enum pipe pipe,
+					     bool enable, bool old)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 reg = PIPESTAT(pipe);
+	u32 pipestat = I915_READ(reg) & 0xffff0000;
+
+	assert_spin_locked(&dev_priv->irq_lock);
+
+	if (enable) {
+		I915_WRITE(reg, pipestat | PIPE_FIFO_UNDERRUN_STATUS);
+		POSTING_READ(reg);
+	} else {
+		if (old && pipestat & PIPE_FIFO_UNDERRUN_STATUS)
+			DRM_ERROR("pipe %c underrun\n", pipe_name(pipe));
+	}
+}
+
+static void ironlake_set_fifo_underrun_reporting(struct drm_device *dev,
+						 enum pipe pipe, bool enable)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	uint32_t bit = (pipe == PIPE_A) ? DE_PIPEA_FIFO_UNDERRUN :
+					  DE_PIPEB_FIFO_UNDERRUN;
+
+	if (enable)
+		ironlake_enable_display_irq(dev_priv, bit);
+	else
+		ironlake_disable_display_irq(dev_priv, bit);
+}
+
+static void ivybridge_set_fifo_underrun_reporting(struct drm_device *dev,
+						  enum pipe pipe,
+						  bool enable, bool old)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	if (enable) {
+		I915_WRITE(GEN7_ERR_INT, ERR_INT_FIFO_UNDERRUN(pipe));
+
+		if (!ivb_can_enable_err_int(dev))
+			return;
+
+		ironlake_enable_display_irq(dev_priv, DE_ERR_INT_IVB);
+	} else {
+		ironlake_disable_display_irq(dev_priv, DE_ERR_INT_IVB);
+
+		if (old &&
+		    I915_READ(GEN7_ERR_INT) & ERR_INT_FIFO_UNDERRUN(pipe)) {
+			DRM_ERROR("uncleared fifo underrun on pipe %c\n",
+				  pipe_name(pipe));
+		}
+	}
+}
+
+static void broadwell_set_fifo_underrun_reporting(struct drm_device *dev,
+						  enum pipe pipe, bool enable)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	assert_spin_locked(&dev_priv->irq_lock);
+
+	if (enable)
+		dev_priv->de_irq_mask[pipe] &= ~GEN8_PIPE_FIFO_UNDERRUN;
+	else
+		dev_priv->de_irq_mask[pipe] |= GEN8_PIPE_FIFO_UNDERRUN;
+	I915_WRITE(GEN8_DE_PIPE_IMR(pipe), dev_priv->de_irq_mask[pipe]);
+	POSTING_READ(GEN8_DE_PIPE_IMR(pipe));
+}
+
+static void ibx_set_fifo_underrun_reporting(struct drm_device *dev,
+					    enum transcoder pch_transcoder,
+					    bool enable)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	uint32_t bit = (pch_transcoder == TRANSCODER_A) ?
+		       SDE_TRANSA_FIFO_UNDER : SDE_TRANSB_FIFO_UNDER;
+
+	if (enable)
+		ibx_enable_display_interrupt(dev_priv, bit);
+	else
+		ibx_disable_display_interrupt(dev_priv, bit);
+}
+
+static void cpt_set_fifo_underrun_reporting(struct drm_device *dev,
+					    enum transcoder pch_transcoder,
+					    bool enable, bool old)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (enable) {
+		I915_WRITE(SERR_INT,
+			   SERR_INT_TRANS_FIFO_UNDERRUN(pch_transcoder));
+
+		if (!cpt_can_enable_serr_int(dev))
+			return;
+
+		ibx_enable_display_interrupt(dev_priv, SDE_ERROR_CPT);
+	} else {
+		ibx_disable_display_interrupt(dev_priv, SDE_ERROR_CPT);
+
+		if (old && I915_READ(SERR_INT) &
+		    SERR_INT_TRANS_FIFO_UNDERRUN(pch_transcoder)) {
+			DRM_ERROR("uncleared pch fifo underrun on pch transcoder %c\n",
+				  transcoder_name(pch_transcoder));
+		}
+	}
+}
+
+static bool __intel_set_cpu_fifo_underrun_reporting(struct drm_device *dev,
+						    enum pipe pipe, bool enable)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pipe];
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	bool old;
+
+	assert_spin_locked(&dev_priv->irq_lock);
+
+	old = !intel_crtc->cpu_fifo_underrun_disabled;
+	intel_crtc->cpu_fifo_underrun_disabled = !enable;
+
+	if (HAS_GMCH_DISPLAY(dev))
+		i9xx_set_fifo_underrun_reporting(dev, pipe, enable, old);
+	else if (IS_GEN5(dev) || IS_GEN6(dev))
+		ironlake_set_fifo_underrun_reporting(dev, pipe, enable);
+	else if (IS_GEN7(dev))
+		ivybridge_set_fifo_underrun_reporting(dev, pipe, enable, old);
+	else if (IS_GEN8(dev) || IS_GEN9(dev))
+		broadwell_set_fifo_underrun_reporting(dev, pipe, enable);
+
+	return old;
+}
+
+/**
+ * intel_set_cpu_fifo_underrun_reporting - set cpu fifo underrrun reporting state
+ * @dev_priv: i915 device instance
+ * @pipe: (CPU) pipe to set state for
+ * @enable: whether underruns should be reported or not
+ *
+ * This function sets the fifo underrun state for @pipe. It is used in the
+ * modeset code to avoid false positives since on many platforms underruns are
+ * expected when disabling or enabling the pipe.
+ *
+ * Notice that on some platforms disabling underrun reports for one pipe
+ * disables for all due to shared interrupts. Actual reporting is still per-pipe
+ * though.
+ *
+ * Returns the previous state of underrun reporting.
+ */
+bool intel_set_cpu_fifo_underrun_reporting(struct drm_i915_private *dev_priv,
+					   enum pipe pipe, bool enable)
+{
+	unsigned long flags;
+	bool ret;
+
+	spin_lock_irqsave(&dev_priv->irq_lock, flags);
+	ret = __intel_set_cpu_fifo_underrun_reporting(dev_priv->dev, pipe,
+						      enable);
+	spin_unlock_irqrestore(&dev_priv->irq_lock, flags);
+
+	return ret;
+}
+
+static bool
+__cpu_fifo_underrun_reporting_enabled(struct drm_i915_private *dev_priv,
+				      enum pipe pipe)
+{
+	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pipe];
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+
+	return !intel_crtc->cpu_fifo_underrun_disabled;
+}
+
+/**
+ * intel_set_pch_fifo_underrun_reporting - set PCH fifo underrun reporting state
+ * @dev_priv: i915 device instance
+ * @pch_transcoder: the PCH transcoder (same as pipe on IVB and older)
+ * @enable: whether underruns should be reported or not
+ *
+ * This function makes us disable or enable PCH fifo underruns for a specific
+ * PCH transcoder. Notice that on some PCHs (e.g. CPT/PPT), disabling FIFO
+ * underrun reporting for one transcoder may also disable all the other PCH
+ * error interruts for the other transcoders, due to the fact that there's just
+ * one interrupt mask/enable bit for all the transcoders.
+ *
+ * Returns the previous state of underrun reporting.
+ */
+bool intel_set_pch_fifo_underrun_reporting(struct drm_i915_private *dev_priv,
+					   enum transcoder pch_transcoder,
+					   bool enable)
+{
+	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pch_transcoder];
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	unsigned long flags;
+	bool old;
+
+	/*
+	 * NOTE: Pre-LPT has a fixed cpu pipe -> pch transcoder mapping, but LPT
+	 * has only one pch transcoder A that all pipes can use. To avoid racy
+	 * pch transcoder -> pipe lookups from interrupt code simply store the
+	 * underrun statistics in crtc A. Since we never expose this anywhere
+	 * nor use it outside of the fifo underrun code here using the "wrong"
+	 * crtc on LPT won't cause issues.
+	 */
+
+	spin_lock_irqsave(&dev_priv->irq_lock, flags);
+
+	old = !intel_crtc->pch_fifo_underrun_disabled;
+	intel_crtc->pch_fifo_underrun_disabled = !enable;
+
+	if (HAS_PCH_IBX(dev_priv->dev))
+		ibx_set_fifo_underrun_reporting(dev_priv->dev, pch_transcoder,
+						enable);
+	else
+		cpt_set_fifo_underrun_reporting(dev_priv->dev, pch_transcoder,
+						enable, old);
+
+	spin_unlock_irqrestore(&dev_priv->irq_lock, flags);
+	return old;
+}
+
+/**
+ * intel_pch_fifo_underrun_irq_handler - handle PCH fifo underrun interrupt
+ * @dev_priv: i915 device instance
+ * @pipe: (CPU) pipe to set state for
+ *
+ * This handles a CPU fifo underrun interrupt, generating an underrun warning
+ * into dmesg if underrun reporting is enabled and then disables the underrun
+ * interrupt to avoid an irq storm.
+ */
+void intel_cpu_fifo_underrun_irq_handler(struct drm_i915_private *dev_priv,
+					 enum pipe pipe)
+{
+	/* GMCH can't disable fifo underruns, filter them. */
+	if (HAS_GMCH_DISPLAY(dev_priv->dev) &&
+	    !__cpu_fifo_underrun_reporting_enabled(dev_priv, pipe))
+		return;
+
+	if (intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, false))
+		DRM_ERROR("CPU pipe %c FIFO underrun\n",
+			  pipe_name(pipe));
+}
+
+/**
+ * intel_pch_fifo_underrun_irq_handler - handle PCH fifo underrun interrupt
+ * @dev_priv: i915 device instance
+ * @pch_transcoder: the PCH transcoder (same as pipe on IVB and older)
+ *
+ * This handles a PCH fifo underrun interrupt, generating an underrun warning
+ * into dmesg if underrun reporting is enabled and then disables the underrun
+ * interrupt to avoid an irq storm.
+ */
+void intel_pch_fifo_underrun_irq_handler(struct drm_i915_private *dev_priv,
+					 enum transcoder pch_transcoder)
+{
+	if (intel_set_pch_fifo_underrun_reporting(dev_priv, pch_transcoder,
+						  false))
+		DRM_ERROR("PCH transcoder %c FIFO underrun\n",
+			  transcoder_name(pch_transcoder));
+}
diff --git a/drivers/gpu/drm/i915/intel_frontbuffer.c b/drivers/gpu/drm/i915/intel_frontbuffer.c
new file mode 100644
index 000000000000..79f6d72179c5
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_frontbuffer.c
@@ -0,0 +1,279 @@
+/*
+ * Copyright © 2014 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
+ * Authors:
+ *	Daniel Vetter <daniel.vetter@ffwll.ch>
+ */
+
+/**
+ * DOC: frontbuffer tracking
+ *
+ * Many features require us to track changes to the currently active
+ * frontbuffer, especially rendering targeted at the frontbuffer.
+ *
+ * To be able to do so GEM tracks frontbuffers using a bitmask for all possible
+ * frontbuffer slots through i915_gem_track_fb(). The function in this file are
+ * then called when the contents of the frontbuffer are invalidated, when
+ * frontbuffer rendering has stopped again to flush out all the changes and when
+ * the frontbuffer is exchanged with a flip. Subsystems interested in
+ * frontbuffer changes (e.g. PSR, FBC, DRRS) should directly put their callbacks
+ * into the relevant places and filter for the frontbuffer slots that they are
+ * interested int.
+ *
+ * On a high level there are two types of powersaving features. The first one
+ * work like a special cache (FBC and PSR) and are interested when they should
+ * stop caching and when to restart caching. This is done by placing callbacks
+ * into the invalidate and the flush functions: At invalidate the caching must
+ * be stopped and at flush time it can be restarted. And maybe they need to know
+ * when the frontbuffer changes (e.g. when the hw doesn't initiate an invalidate
+ * and flush on its own) which can be achieved with placing callbacks into the
+ * flip functions.
+ *
+ * The other type of display power saving feature only cares about busyness
+ * (e.g. DRRS). In that case all three (invalidate, flush and flip) indicate
+ * busyness. There is no direct way to detect idleness. Instead an idle timer
+ * work delayed work should be started from the flush and flip functions and
+ * cancelled as soon as busyness is detected.
+ *
+ * Note that there's also an older frontbuffer activity tracking scheme which
+ * just tracks general activity. This is done by the various mark_busy and
+ * mark_idle functions. For display power management features using these
+ * functions is deprecated and should be avoided.
+ */
+
+#include <drm/drmP.h>
+
+#include "intel_drv.h"
+#include "i915_drv.h"
+
+static void intel_increase_pllclock(struct drm_device *dev,
+				    enum pipe pipe)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int dpll_reg = DPLL(pipe);
+	int dpll;
+
+	if (!HAS_GMCH_DISPLAY(dev))
+		return;
+
+	if (!dev_priv->lvds_downclock_avail)
+		return;
+
+	dpll = I915_READ(dpll_reg);
+	if (!HAS_PIPE_CXSR(dev) && (dpll & DISPLAY_RATE_SELECT_FPA1)) {
+		DRM_DEBUG_DRIVER("upclocking LVDS\n");
+
+		assert_panel_unlocked(dev_priv, pipe);
+
+		dpll &= ~DISPLAY_RATE_SELECT_FPA1;
+		I915_WRITE(dpll_reg, dpll);
+		intel_wait_for_vblank(dev, pipe);
+
+		dpll = I915_READ(dpll_reg);
+		if (dpll & DISPLAY_RATE_SELECT_FPA1)
+			DRM_DEBUG_DRIVER("failed to upclock LVDS!\n");
+	}
+}
+
+/**
+ * intel_mark_fb_busy - mark given planes as busy
+ * @dev: DRM device
+ * @frontbuffer_bits: bits for the affected planes
+ * @ring: optional ring for asynchronous commands
+ *
+ * This function gets called every time the screen contents change. It can be
+ * used to keep e.g. the update rate at the nominal refresh rate with DRRS.
+ */
+static void intel_mark_fb_busy(struct drm_device *dev,
+			       unsigned frontbuffer_bits,
+			       struct intel_engine_cs *ring)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum pipe pipe;
+
+	if (!i915.powersave)
+		return;
+
+	for_each_pipe(dev_priv, pipe) {
+		if (!(frontbuffer_bits & INTEL_FRONTBUFFER_ALL_MASK(pipe)))
+			continue;
+
+		intel_increase_pllclock(dev, pipe);
+		if (ring && intel_fbc_enabled(dev))
+			ring->fbc_dirty = true;
+	}
+}
+
+/**
+ * intel_fb_obj_invalidate - invalidate frontbuffer object
+ * @obj: GEM object to invalidate
+ * @ring: set for asynchronous rendering
+ *
+ * This function gets called every time rendering on the given object starts and
+ * frontbuffer caching (fbc, low refresh rate for DRRS, panel self refresh) must
+ * be invalidated. If @ring is non-NULL any subsequent invalidation will be delayed
+ * until the rendering completes or a flip on this frontbuffer plane is
+ * scheduled.
+ */
+void intel_fb_obj_invalidate(struct drm_i915_gem_object *obj,
+			     struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = obj->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	WARN_ON(!mutex_is_locked(&dev->struct_mutex));
+
+	if (!obj->frontbuffer_bits)
+		return;
+
+	if (ring) {
+		mutex_lock(&dev_priv->fb_tracking.lock);
+		dev_priv->fb_tracking.busy_bits
+			|= obj->frontbuffer_bits;
+		dev_priv->fb_tracking.flip_bits
+			&= ~obj->frontbuffer_bits;
+		mutex_unlock(&dev_priv->fb_tracking.lock);
+	}
+
+	intel_mark_fb_busy(dev, obj->frontbuffer_bits, ring);
+
+	intel_psr_invalidate(dev, obj->frontbuffer_bits);
+}
+
+/**
+ * intel_frontbuffer_flush - flush frontbuffer
+ * @dev: DRM device
+ * @frontbuffer_bits: frontbuffer plane tracking bits
+ *
+ * This function gets called every time rendering on the given planes has
+ * completed and frontbuffer caching can be started again. Flushes will get
+ * delayed if they're blocked by some outstanding asynchronous rendering.
+ *
+ * Can be called without any locks held.
+ */
+void intel_frontbuffer_flush(struct drm_device *dev,
+			     unsigned frontbuffer_bits)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	/* Delay flushing when rings are still busy.*/
+	mutex_lock(&dev_priv->fb_tracking.lock);
+	frontbuffer_bits &= ~dev_priv->fb_tracking.busy_bits;
+	mutex_unlock(&dev_priv->fb_tracking.lock);
+
+	intel_mark_fb_busy(dev, frontbuffer_bits, NULL);
+
+	intel_psr_flush(dev, frontbuffer_bits);
+
+	/*
+	 * FIXME: Unconditional fbc flushing here is a rather gross hack and
+	 * needs to be reworked into a proper frontbuffer tracking scheme like
+	 * psr employs.
+	 */
+	if (dev_priv->fbc.need_sw_cache_clean) {
+		dev_priv->fbc.need_sw_cache_clean = false;
+		bdw_fbc_sw_flush(dev, FBC_REND_CACHE_CLEAN);
+	}
+}
+
+/**
+ * intel_fb_obj_flush - flush frontbuffer object
+ * @obj: GEM object to flush
+ * @retire: set when retiring asynchronous rendering
+ *
+ * This function gets called every time rendering on the given object has
+ * completed and frontbuffer caching can be started again. If @retire is true
+ * then any delayed flushes will be unblocked.
+ */
+void intel_fb_obj_flush(struct drm_i915_gem_object *obj,
+			bool retire)
+{
+	struct drm_device *dev = obj->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	unsigned frontbuffer_bits;
+
+	WARN_ON(!mutex_is_locked(&dev->struct_mutex));
+
+	if (!obj->frontbuffer_bits)
+		return;
+
+	frontbuffer_bits = obj->frontbuffer_bits;
+
+	if (retire) {
+		mutex_lock(&dev_priv->fb_tracking.lock);
+		/* Filter out new bits since rendering started. */
+		frontbuffer_bits &= dev_priv->fb_tracking.busy_bits;
+
+		dev_priv->fb_tracking.busy_bits &= ~frontbuffer_bits;
+		mutex_unlock(&dev_priv->fb_tracking.lock);
+	}
+
+	intel_frontbuffer_flush(dev, frontbuffer_bits);
+}
+
+/**
+ * intel_frontbuffer_flip_prepare - prepare asynchronous frontbuffer flip
+ * @dev: DRM device
+ * @frontbuffer_bits: frontbuffer plane tracking bits
+ *
+ * This function gets called after scheduling a flip on @obj. The actual
+ * frontbuffer flushing will be delayed until completion is signalled with
+ * intel_frontbuffer_flip_complete. If an invalidate happens in between this
+ * flush will be cancelled.
+ *
+ * Can be called without any locks held.
+ */
+void intel_frontbuffer_flip_prepare(struct drm_device *dev,
+				    unsigned frontbuffer_bits)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	mutex_lock(&dev_priv->fb_tracking.lock);
+	dev_priv->fb_tracking.flip_bits |= frontbuffer_bits;
+	/* Remove stale busy bits due to the old buffer. */
+	dev_priv->fb_tracking.busy_bits &= ~frontbuffer_bits;
+	mutex_unlock(&dev_priv->fb_tracking.lock);
+}
+
+/**
+ * intel_frontbuffer_flip_complete - complete asynchronous frontbuffer flip
+ * @dev: DRM device
+ * @frontbuffer_bits: frontbuffer plane tracking bits
+ *
+ * This function gets called after the flip has been latched and will complete
+ * on the next vblank. It will execute the flush if it hasn't been cancelled yet.
+ *
+ * Can be called without any locks held.
+ */
+void intel_frontbuffer_flip_complete(struct drm_device *dev,
+				     unsigned frontbuffer_bits)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	mutex_lock(&dev_priv->fb_tracking.lock);
+	/* Mask any cancelled flips. */
+	frontbuffer_bits &= dev_priv->fb_tracking.flip_bits;
+	dev_priv->fb_tracking.flip_bits &= ~frontbuffer_bits;
+	mutex_unlock(&dev_priv->fb_tracking.lock);
+
+	intel_frontbuffer_flush(dev, frontbuffer_bits);
+}
diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c
index 5a9de21637b7..3abc2000fce9 100644
--- a/drivers/gpu/drm/i915/intel_hdmi.c
+++ b/drivers/gpu/drm/i915/intel_hdmi.c
@@ -166,6 +166,19 @@ static void g4x_write_infoframe(struct drm_encoder *encoder,
 	POSTING_READ(VIDEO_DIP_CTL);
 }
 
+static bool g4x_infoframe_enabled(struct drm_encoder *encoder)
+{
+	struct drm_device *dev = encoder->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_digital_port *intel_dig_port = enc_to_dig_port(encoder);
+	u32 val = I915_READ(VIDEO_DIP_CTL);
+
+	if (VIDEO_DIP_PORT(intel_dig_port->port) == (val & VIDEO_DIP_PORT_MASK))
+		return val & VIDEO_DIP_ENABLE;
+
+	return false;
+}
+
 static void ibx_write_infoframe(struct drm_encoder *encoder,
 				enum hdmi_infoframe_type type,
 				const void *frame, ssize_t len)
@@ -204,6 +217,17 @@ static void ibx_write_infoframe(struct drm_encoder *encoder,
 	POSTING_READ(reg);
 }
 
+static bool ibx_infoframe_enabled(struct drm_encoder *encoder)
+{
+	struct drm_device *dev = encoder->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->crtc);
+	int reg = TVIDEO_DIP_CTL(intel_crtc->pipe);
+	u32 val = I915_READ(reg);
+
+	return val & VIDEO_DIP_ENABLE;
+}
+
 static void cpt_write_infoframe(struct drm_encoder *encoder,
 				enum hdmi_infoframe_type type,
 				const void *frame, ssize_t len)
@@ -245,6 +269,17 @@ static void cpt_write_infoframe(struct drm_encoder *encoder,
 	POSTING_READ(reg);
 }
 
+static bool cpt_infoframe_enabled(struct drm_encoder *encoder)
+{
+	struct drm_device *dev = encoder->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->crtc);
+	int reg = TVIDEO_DIP_CTL(intel_crtc->pipe);
+	u32 val = I915_READ(reg);
+
+	return val & VIDEO_DIP_ENABLE;
+}
+
 static void vlv_write_infoframe(struct drm_encoder *encoder,
 				enum hdmi_infoframe_type type,
 				const void *frame, ssize_t len)
@@ -283,6 +318,17 @@ static void vlv_write_infoframe(struct drm_encoder *encoder,
 	POSTING_READ(reg);
 }
 
+static bool vlv_infoframe_enabled(struct drm_encoder *encoder)
+{
+	struct drm_device *dev = encoder->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->crtc);
+	int reg = VLV_TVIDEO_DIP_CTL(intel_crtc->pipe);
+	u32 val = I915_READ(reg);
+
+	return val & VIDEO_DIP_ENABLE;
+}
+
 static void hsw_write_infoframe(struct drm_encoder *encoder,
 				enum hdmi_infoframe_type type,
 				const void *frame, ssize_t len)
@@ -320,6 +366,18 @@ static void hsw_write_infoframe(struct drm_encoder *encoder,
 	POSTING_READ(ctl_reg);
 }
 
+static bool hsw_infoframe_enabled(struct drm_encoder *encoder)
+{
+	struct drm_device *dev = encoder->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->crtc);
+	u32 ctl_reg = HSW_TVIDEO_DIP_CTL(intel_crtc->config.cpu_transcoder);
+	u32 val = I915_READ(ctl_reg);
+
+	return val & (VIDEO_DIP_ENABLE_AVI_HSW | VIDEO_DIP_ENABLE_SPD_HSW |
+		      VIDEO_DIP_ENABLE_VS_HSW);
+}
+
 /*
  * The data we write to the DIP data buffer registers is 1 byte bigger than the
  * HDMI infoframe size because of an ECC/reserved byte at position 3 (starting
@@ -661,14 +719,6 @@ static void intel_hdmi_prepare(struct intel_encoder *encoder)
 	if (crtc->config.has_hdmi_sink)
 		hdmi_val |= HDMI_MODE_SELECT_HDMI;
 
-	if (crtc->config.has_audio) {
-		WARN_ON(!crtc->config.has_hdmi_sink);
-		DRM_DEBUG_DRIVER("Enabling HDMI audio on pipe %c\n",
-				 pipe_name(crtc->pipe));
-		hdmi_val |= SDVO_AUDIO_ENABLE;
-		intel_write_eld(&encoder->base, adjusted_mode);
-	}
-
 	if (HAS_PCH_CPT(dev))
 		hdmi_val |= SDVO_PIPE_SEL_CPT(crtc->pipe);
 	else if (IS_CHERRYVIEW(dev))
@@ -690,7 +740,7 @@ static bool intel_hdmi_get_hw_state(struct intel_encoder *encoder,
 	u32 tmp;
 
 	power_domain = intel_display_port_power_domain(encoder);
-	if (!intel_display_power_enabled(dev_priv, power_domain))
+	if (!intel_display_power_is_enabled(dev_priv, power_domain))
 		return false;
 
 	tmp = I915_READ(intel_hdmi->hdmi_reg);
@@ -732,6 +782,9 @@ static void intel_hdmi_get_config(struct intel_encoder *encoder,
 	if (tmp & HDMI_MODE_SELECT_HDMI)
 		pipe_config->has_hdmi_sink = true;
 
+	if (intel_hdmi->infoframe_enabled(&encoder->base))
+		pipe_config->has_infoframe = true;
+
 	if (tmp & SDVO_AUDIO_ENABLE)
 		pipe_config->has_audio = true;
 
@@ -791,6 +844,13 @@ static void intel_enable_hdmi(struct intel_encoder *encoder)
 		I915_WRITE(intel_hdmi->hdmi_reg, temp);
 		POSTING_READ(intel_hdmi->hdmi_reg);
 	}
+
+	if (intel_crtc->config.has_audio) {
+		WARN_ON(!intel_crtc->config.has_hdmi_sink);
+		DRM_DEBUG_DRIVER("Enabling HDMI audio on pipe %c\n",
+				 pipe_name(intel_crtc->pipe));
+		intel_audio_codec_enable(encoder);
+	}
 }
 
 static void vlv_enable_hdmi(struct intel_encoder *encoder)
@@ -802,9 +862,13 @@ static void intel_disable_hdmi(struct intel_encoder *encoder)
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base);
+	struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
 	u32 temp;
 	u32 enable_bits = SDVO_ENABLE | SDVO_AUDIO_ENABLE;
 
+	if (crtc->config.has_audio)
+		intel_audio_codec_disable(encoder);
+
 	temp = I915_READ(intel_hdmi->hdmi_reg);
 
 	/* HW workaround for IBX, we need to move the port to transcoder A
@@ -869,10 +933,15 @@ static enum drm_mode_status
 intel_hdmi_mode_valid(struct drm_connector *connector,
 		      struct drm_display_mode *mode)
 {
-	if (mode->clock > hdmi_portclock_limit(intel_attached_hdmi(connector),
-					       true))
+	int clock = mode->clock;
+
+	if (mode->flags & DRM_MODE_FLAG_DBLCLK)
+		clock *= 2;
+
+	if (clock > hdmi_portclock_limit(intel_attached_hdmi(connector),
+					 true))
 		return MODE_CLOCK_HIGH;
-	if (mode->clock < 20000)
+	if (clock < 20000)
 		return MODE_CLOCK_LOW;
 
 	if (mode->flags & DRM_MODE_FLAG_DBLSCAN)
@@ -890,7 +959,7 @@ static bool hdmi_12bpc_possible(struct intel_crtc *crtc)
 	if (HAS_GMCH_DISPLAY(dev))
 		return false;
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		if (encoder->new_crtc != crtc)
 			continue;
 
@@ -917,6 +986,9 @@ bool intel_hdmi_compute_config(struct intel_encoder *encoder,
 
 	pipe_config->has_hdmi_sink = intel_hdmi->has_hdmi_sink;
 
+	if (pipe_config->has_hdmi_sink)
+		pipe_config->has_infoframe = true;
+
 	if (intel_hdmi->color_range_auto) {
 		/* See CEA-861-E - 5.1 Default Encoding Parameters */
 		if (pipe_config->has_hdmi_sink &&
@@ -926,6 +998,10 @@ bool intel_hdmi_compute_config(struct intel_encoder *encoder,
 			intel_hdmi->color_range = 0;
 	}
 
+	if (adjusted_mode->flags & DRM_MODE_FLAG_DBLCLK) {
+		pipe_config->pixel_multiplier = 2;
+	}
+
 	if (intel_hdmi->color_range)
 		pipe_config->limited_color_range = true;
 
@@ -967,104 +1043,117 @@ bool intel_hdmi_compute_config(struct intel_encoder *encoder,
 	return true;
 }
 
-static enum drm_connector_status
-intel_hdmi_detect(struct drm_connector *connector, bool force)
+static void
+intel_hdmi_unset_edid(struct drm_connector *connector)
 {
-	struct drm_device *dev = connector->dev;
 	struct intel_hdmi *intel_hdmi = intel_attached_hdmi(connector);
-	struct intel_digital_port *intel_dig_port =
-		hdmi_to_dig_port(intel_hdmi);
-	struct intel_encoder *intel_encoder = &intel_dig_port->base;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct edid *edid;
-	enum intel_display_power_domain power_domain;
-	enum drm_connector_status status = connector_status_disconnected;
 
-	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
-		      connector->base.id, connector->name);
+	intel_hdmi->has_hdmi_sink = false;
+	intel_hdmi->has_audio = false;
+	intel_hdmi->rgb_quant_range_selectable = false;
+
+	kfree(to_intel_connector(connector)->detect_edid);
+	to_intel_connector(connector)->detect_edid = NULL;
+}
+
+static bool
+intel_hdmi_set_edid(struct drm_connector *connector)
+{
+	struct drm_i915_private *dev_priv = to_i915(connector->dev);
+	struct intel_hdmi *intel_hdmi = intel_attached_hdmi(connector);
+	struct intel_encoder *intel_encoder =
+		&hdmi_to_dig_port(intel_hdmi)->base;
+	enum intel_display_power_domain power_domain;
+	struct edid *edid;
+	bool connected = false;
 
 	power_domain = intel_display_port_power_domain(intel_encoder);
 	intel_display_power_get(dev_priv, power_domain);
 
-	intel_hdmi->has_hdmi_sink = false;
-	intel_hdmi->has_audio = false;
-	intel_hdmi->rgb_quant_range_selectable = false;
 	edid = drm_get_edid(connector,
 			    intel_gmbus_get_adapter(dev_priv,
 						    intel_hdmi->ddc_bus));
 
-	if (edid) {
-		if (edid->input & DRM_EDID_INPUT_DIGITAL) {
-			status = connector_status_connected;
-			if (intel_hdmi->force_audio != HDMI_AUDIO_OFF_DVI)
-				intel_hdmi->has_hdmi_sink =
-						drm_detect_hdmi_monitor(edid);
-			intel_hdmi->has_audio = drm_detect_monitor_audio(edid);
-			intel_hdmi->rgb_quant_range_selectable =
-				drm_rgb_quant_range_selectable(edid);
-		}
-		kfree(edid);
-	}
+	intel_display_power_put(dev_priv, power_domain);
+
+	to_intel_connector(connector)->detect_edid = edid;
+	if (edid && edid->input & DRM_EDID_INPUT_DIGITAL) {
+		intel_hdmi->rgb_quant_range_selectable =
+			drm_rgb_quant_range_selectable(edid);
 
-	if (status == connector_status_connected) {
+		intel_hdmi->has_audio = drm_detect_monitor_audio(edid);
 		if (intel_hdmi->force_audio != HDMI_AUDIO_AUTO)
 			intel_hdmi->has_audio =
-				(intel_hdmi->force_audio == HDMI_AUDIO_ON);
-		intel_encoder->type = INTEL_OUTPUT_HDMI;
+				intel_hdmi->force_audio == HDMI_AUDIO_ON;
+
+		if (intel_hdmi->force_audio != HDMI_AUDIO_OFF_DVI)
+			intel_hdmi->has_hdmi_sink =
+				drm_detect_hdmi_monitor(edid);
+
+		connected = true;
 	}
 
-	intel_display_power_put(dev_priv, power_domain);
+	return connected;
+}
+
+static enum drm_connector_status
+intel_hdmi_detect(struct drm_connector *connector, bool force)
+{
+	enum drm_connector_status status;
+
+	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
+		      connector->base.id, connector->name);
+
+	intel_hdmi_unset_edid(connector);
+
+	if (intel_hdmi_set_edid(connector)) {
+		struct intel_hdmi *intel_hdmi = intel_attached_hdmi(connector);
+
+		hdmi_to_dig_port(intel_hdmi)->base.type = INTEL_OUTPUT_HDMI;
+		status = connector_status_connected;
+	} else
+		status = connector_status_disconnected;
 
 	return status;
 }
 
-static int intel_hdmi_get_modes(struct drm_connector *connector)
+static void
+intel_hdmi_force(struct drm_connector *connector)
 {
-	struct intel_encoder *intel_encoder = intel_attached_encoder(connector);
-	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&intel_encoder->base);
-	struct drm_i915_private *dev_priv = connector->dev->dev_private;
-	enum intel_display_power_domain power_domain;
-	int ret;
+	struct intel_hdmi *intel_hdmi = intel_attached_hdmi(connector);
 
-	/* We should parse the EDID data and find out if it's an HDMI sink so
-	 * we can send audio to it.
-	 */
+	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
+		      connector->base.id, connector->name);
 
-	power_domain = intel_display_port_power_domain(intel_encoder);
-	intel_display_power_get(dev_priv, power_domain);
+	intel_hdmi_unset_edid(connector);
 
-	ret = intel_ddc_get_modes(connector,
-				   intel_gmbus_get_adapter(dev_priv,
-							   intel_hdmi->ddc_bus));
+	if (connector->status != connector_status_connected)
+		return;
 
-	intel_display_power_put(dev_priv, power_domain);
+	intel_hdmi_set_edid(connector);
+	hdmi_to_dig_port(intel_hdmi)->base.type = INTEL_OUTPUT_HDMI;
+}
+
+static int intel_hdmi_get_modes(struct drm_connector *connector)
+{
+	struct edid *edid;
+
+	edid = to_intel_connector(connector)->detect_edid;
+	if (edid == NULL)
+		return 0;
 
-	return ret;
+	return intel_connector_update_modes(connector, edid);
 }
 
 static bool
 intel_hdmi_detect_audio(struct drm_connector *connector)
 {
-	struct intel_encoder *intel_encoder = intel_attached_encoder(connector);
-	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&intel_encoder->base);
-	struct drm_i915_private *dev_priv = connector->dev->dev_private;
-	enum intel_display_power_domain power_domain;
-	struct edid *edid;
 	bool has_audio = false;
+	struct edid *edid;
 
-	power_domain = intel_display_port_power_domain(intel_encoder);
-	intel_display_power_get(dev_priv, power_domain);
-
-	edid = drm_get_edid(connector,
-			    intel_gmbus_get_adapter(dev_priv,
-						    intel_hdmi->ddc_bus));
-	if (edid) {
-		if (edid->input & DRM_EDID_INPUT_DIGITAL)
-			has_audio = drm_detect_monitor_audio(edid);
-		kfree(edid);
-	}
-
-	intel_display_power_put(dev_priv, power_domain);
+	edid = to_intel_connector(connector)->detect_edid;
+	if (edid && edid->input & DRM_EDID_INPUT_DIGITAL)
+		has_audio = drm_detect_monitor_audio(edid);
 
 	return has_audio;
 }
@@ -1265,6 +1354,8 @@ static void chv_hdmi_pre_pll_enable(struct intel_encoder *encoder)
 	enum pipe pipe = intel_crtc->pipe;
 	u32 val;
 
+	intel_hdmi_prepare(encoder);
+
 	mutex_lock(&dev_priv->dpio_lock);
 
 	/* program left/right clock distribution */
@@ -1370,10 +1461,13 @@ static void chv_hdmi_post_disable(struct intel_encoder *encoder)
 static void chv_hdmi_pre_enable(struct intel_encoder *encoder)
 {
 	struct intel_digital_port *dport = enc_to_dig_port(&encoder->base);
+	struct intel_hdmi *intel_hdmi = &dport->hdmi;
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc =
 		to_intel_crtc(encoder->base.crtc);
+	struct drm_display_mode *adjusted_mode =
+		&intel_crtc->config.adjusted_mode;
 	enum dpio_channel ch = vlv_dport_to_channel(dport);
 	int pipe = intel_crtc->pipe;
 	int data, i;
@@ -1381,6 +1475,15 @@ static void chv_hdmi_pre_enable(struct intel_encoder *encoder)
 
 	mutex_lock(&dev_priv->dpio_lock);
 
+	/* allow hardware to manage TX FIFO reset source */
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW11(ch));
+	val &= ~DPIO_LANEDESKEW_STRAP_OVRD;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW11(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW11(ch));
+	val &= ~DPIO_LANEDESKEW_STRAP_OVRD;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW11(ch), val);
+
 	/* Deassert soft data lane reset*/
 	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW1(ch));
 	val |= CHV_PCS_REQ_SOFTRESET_EN;
@@ -1417,12 +1520,26 @@ static void chv_hdmi_pre_enable(struct intel_encoder *encoder)
 	/* Clear calc init */
 	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW10(ch));
 	val &= ~(DPIO_PCS_SWING_CALC_TX0_TX2 | DPIO_PCS_SWING_CALC_TX1_TX3);
+	val &= ~(DPIO_PCS_TX1DEEMP_MASK | DPIO_PCS_TX2DEEMP_MASK);
+	val |= DPIO_PCS_TX1DEEMP_9P5 | DPIO_PCS_TX2DEEMP_9P5;
 	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW10(ch), val);
 
 	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW10(ch));
 	val &= ~(DPIO_PCS_SWING_CALC_TX0_TX2 | DPIO_PCS_SWING_CALC_TX1_TX3);
+	val &= ~(DPIO_PCS_TX1DEEMP_MASK | DPIO_PCS_TX2DEEMP_MASK);
+	val |= DPIO_PCS_TX1DEEMP_9P5 | DPIO_PCS_TX2DEEMP_9P5;
 	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW10(ch), val);
 
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW9(ch));
+	val &= ~(DPIO_PCS_TX1MARGIN_MASK | DPIO_PCS_TX2MARGIN_MASK);
+	val |= DPIO_PCS_TX1MARGIN_000 | DPIO_PCS_TX2MARGIN_000;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW9(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW9(ch));
+	val &= ~(DPIO_PCS_TX1MARGIN_MASK | DPIO_PCS_TX2MARGIN_MASK);
+	val |= DPIO_PCS_TX1MARGIN_000 | DPIO_PCS_TX2MARGIN_000;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW9(ch), val);
+
 	/* FIXME: Program the support xxx V-dB */
 	/* Use 800mV-0dB */
 	for (i = 0; i < 4; i++) {
@@ -1434,8 +1551,8 @@ static void chv_hdmi_pre_enable(struct intel_encoder *encoder)
 
 	for (i = 0; i < 4; i++) {
 		val = vlv_dpio_read(dev_priv, pipe, CHV_TX_DW2(ch, i));
-		val &= ~DPIO_SWING_MARGIN_MASK;
-		val |= 102 << DPIO_SWING_MARGIN_SHIFT;
+		val &= ~DPIO_SWING_MARGIN000_MASK;
+		val |= 102 << DPIO_SWING_MARGIN000_SHIFT;
 		vlv_dpio_write(dev_priv, pipe, CHV_TX_DW2(ch, i), val);
 	}
 
@@ -1475,6 +1592,10 @@ static void chv_hdmi_pre_enable(struct intel_encoder *encoder)
 
 	mutex_unlock(&dev_priv->dpio_lock);
 
+	intel_hdmi->set_infoframes(&encoder->base,
+				   intel_crtc->config.has_hdmi_sink,
+				   adjusted_mode);
+
 	intel_enable_hdmi(encoder);
 
 	vlv_wait_port_ready(dev_priv, dport);
@@ -1482,6 +1603,7 @@ static void chv_hdmi_pre_enable(struct intel_encoder *encoder)
 
 static void intel_hdmi_destroy(struct drm_connector *connector)
 {
+	kfree(to_intel_connector(connector)->detect_edid);
 	drm_connector_cleanup(connector);
 	kfree(connector);
 }
@@ -1489,6 +1611,7 @@ static void intel_hdmi_destroy(struct drm_connector *connector)
 static const struct drm_connector_funcs intel_hdmi_connector_funcs = {
 	.dpms = intel_connector_dpms,
 	.detect = intel_hdmi_detect,
+	.force = intel_hdmi_force,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.set_property = intel_hdmi_set_property,
 	.destroy = intel_hdmi_destroy,
@@ -1567,18 +1690,23 @@ void intel_hdmi_init_connector(struct intel_digital_port *intel_dig_port,
 	if (IS_VALLEYVIEW(dev)) {
 		intel_hdmi->write_infoframe = vlv_write_infoframe;
 		intel_hdmi->set_infoframes = vlv_set_infoframes;
+		intel_hdmi->infoframe_enabled = vlv_infoframe_enabled;
 	} else if (IS_G4X(dev)) {
 		intel_hdmi->write_infoframe = g4x_write_infoframe;
 		intel_hdmi->set_infoframes = g4x_set_infoframes;
+		intel_hdmi->infoframe_enabled = g4x_infoframe_enabled;
 	} else if (HAS_DDI(dev)) {
 		intel_hdmi->write_infoframe = hsw_write_infoframe;
 		intel_hdmi->set_infoframes = hsw_set_infoframes;
+		intel_hdmi->infoframe_enabled = hsw_infoframe_enabled;
 	} else if (HAS_PCH_IBX(dev)) {
 		intel_hdmi->write_infoframe = ibx_write_infoframe;
 		intel_hdmi->set_infoframes = ibx_set_infoframes;
+		intel_hdmi->infoframe_enabled = ibx_infoframe_enabled;
 	} else {
 		intel_hdmi->write_infoframe = cpt_write_infoframe;
 		intel_hdmi->set_infoframes = cpt_set_infoframes;
+		intel_hdmi->infoframe_enabled = cpt_infoframe_enabled;
 	}
 
 	if (HAS_DDI(dev))
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
new file mode 100644
index 000000000000..e588376227ea
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -0,0 +1,1942 @@
+/*
+ * Copyright © 2014 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ * Authors:
+ *    Ben Widawsky <ben@bwidawsk.net>
+ *    Michel Thierry <michel.thierry@intel.com>
+ *    Thomas Daniel <thomas.daniel@intel.com>
+ *    Oscar Mateo <oscar.mateo@intel.com>
+ *
+ */
+
+/**
+ * DOC: Logical Rings, Logical Ring Contexts and Execlists
+ *
+ * Motivation:
+ * GEN8 brings an expansion of the HW contexts: "Logical Ring Contexts".
+ * These expanded contexts enable a number of new abilities, especially
+ * "Execlists" (also implemented in this file).
+ *
+ * One of the main differences with the legacy HW contexts is that logical
+ * ring contexts incorporate many more things to the context's state, like
+ * PDPs or ringbuffer control registers:
+ *
+ * The reason why PDPs are included in the context is straightforward: as
+ * PPGTTs (per-process GTTs) are actually per-context, having the PDPs
+ * contained there mean you don't need to do a ppgtt->switch_mm yourself,
+ * instead, the GPU will do it for you on the context switch.
+ *
+ * But, what about the ringbuffer control registers (head, tail, etc..)?
+ * shouldn't we just need a set of those per engine command streamer? This is
+ * where the name "Logical Rings" starts to make sense: by virtualizing the
+ * rings, the engine cs shifts to a new "ring buffer" with every context
+ * switch. When you want to submit a workload to the GPU you: A) choose your
+ * context, B) find its appropriate virtualized ring, C) write commands to it
+ * and then, finally, D) tell the GPU to switch to that context.
+ *
+ * Instead of the legacy MI_SET_CONTEXT, the way you tell the GPU to switch
+ * to a contexts is via a context execution list, ergo "Execlists".
+ *
+ * LRC implementation:
+ * Regarding the creation of contexts, we have:
+ *
+ * - One global default context.
+ * - One local default context for each opened fd.
+ * - One local extra context for each context create ioctl call.
+ *
+ * Now that ringbuffers belong per-context (and not per-engine, like before)
+ * and that contexts are uniquely tied to a given engine (and not reusable,
+ * like before) we need:
+ *
+ * - One ringbuffer per-engine inside each context.
+ * - One backing object per-engine inside each context.
+ *
+ * The global default context starts its life with these new objects fully
+ * allocated and populated. The local default context for each opened fd is
+ * more complex, because we don't know at creation time which engine is going
+ * to use them. To handle this, we have implemented a deferred creation of LR
+ * contexts:
+ *
+ * The local context starts its life as a hollow or blank holder, that only
+ * gets populated for a given engine once we receive an execbuffer. If later
+ * on we receive another execbuffer ioctl for the same context but a different
+ * engine, we allocate/populate a new ringbuffer and context backing object and
+ * so on.
+ *
+ * Finally, regarding local contexts created using the ioctl call: as they are
+ * only allowed with the render ring, we can allocate & populate them right
+ * away (no need to defer anything, at least for now).
+ *
+ * Execlists implementation:
+ * Execlists are the new method by which, on gen8+ hardware, workloads are
+ * submitted for execution (as opposed to the legacy, ringbuffer-based, method).
+ * This method works as follows:
+ *
+ * When a request is committed, its commands (the BB start and any leading or
+ * trailing commands, like the seqno breadcrumbs) are placed in the ringbuffer
+ * for the appropriate context. The tail pointer in the hardware context is not
+ * updated at this time, but instead, kept by the driver in the ringbuffer
+ * structure. A structure representing this request is added to a request queue
+ * for the appropriate engine: this structure contains a copy of the context's
+ * tail after the request was written to the ring buffer and a pointer to the
+ * context itself.
+ *
+ * If the engine's request queue was empty before the request was added, the
+ * queue is processed immediately. Otherwise the queue will be processed during
+ * a context switch interrupt. In any case, elements on the queue will get sent
+ * (in pairs) to the GPU's ExecLists Submit Port (ELSP, for short) with a
+ * globally unique 20-bits submission ID.
+ *
+ * When execution of a request completes, the GPU updates the context status
+ * buffer with a context complete event and generates a context switch interrupt.
+ * During the interrupt handling, the driver examines the events in the buffer:
+ * for each context complete event, if the announced ID matches that on the head
+ * of the request queue, then that request is retired and removed from the queue.
+ *
+ * After processing, if any requests were retired and the queue is not empty
+ * then a new execution list can be submitted. The two requests at the front of
+ * the queue are next to be submitted but since a context may not occur twice in
+ * an execution list, if subsequent requests have the same ID as the first then
+ * the two requests must be combined. This is done simply by discarding requests
+ * at the head of the queue until either only one requests is left (in which case
+ * we use a NULL second context) or the first two requests have unique IDs.
+ *
+ * By always executing the first two requests in the queue the driver ensures
+ * that the GPU is kept as busy as possible. In the case where a single context
+ * completes but a second context is still executing, the request for this second
+ * context will be at the head of the queue when we remove the first one. This
+ * request will then be resubmitted along with a new request for a different context,
+ * which will cause the hardware to continue executing the second request and queue
+ * the new request (the GPU detects the condition of a context getting preempted
+ * with the same context and optimizes the context switch flow by not doing
+ * preemption, but just sampling the new tail pointer).
+ *
+ */
+
+#include <drm/drmP.h>
+#include <drm/i915_drm.h>
+#include "i915_drv.h"
+
+#define GEN9_LR_CONTEXT_RENDER_SIZE (22 * PAGE_SIZE)
+#define GEN8_LR_CONTEXT_RENDER_SIZE (20 * PAGE_SIZE)
+#define GEN8_LR_CONTEXT_OTHER_SIZE (2 * PAGE_SIZE)
+
+#define RING_EXECLIST_QFULL		(1 << 0x2)
+#define RING_EXECLIST1_VALID		(1 << 0x3)
+#define RING_EXECLIST0_VALID		(1 << 0x4)
+#define RING_EXECLIST_ACTIVE_STATUS	(3 << 0xE)
+#define RING_EXECLIST1_ACTIVE		(1 << 0x11)
+#define RING_EXECLIST0_ACTIVE		(1 << 0x12)
+
+#define GEN8_CTX_STATUS_IDLE_ACTIVE	(1 << 0)
+#define GEN8_CTX_STATUS_PREEMPTED	(1 << 1)
+#define GEN8_CTX_STATUS_ELEMENT_SWITCH	(1 << 2)
+#define GEN8_CTX_STATUS_ACTIVE_IDLE	(1 << 3)
+#define GEN8_CTX_STATUS_COMPLETE	(1 << 4)
+#define GEN8_CTX_STATUS_LITE_RESTORE	(1 << 15)
+
+#define CTX_LRI_HEADER_0		0x01
+#define CTX_CONTEXT_CONTROL		0x02
+#define CTX_RING_HEAD			0x04
+#define CTX_RING_TAIL			0x06
+#define CTX_RING_BUFFER_START		0x08
+#define CTX_RING_BUFFER_CONTROL		0x0a
+#define CTX_BB_HEAD_U			0x0c
+#define CTX_BB_HEAD_L			0x0e
+#define CTX_BB_STATE			0x10
+#define CTX_SECOND_BB_HEAD_U		0x12
+#define CTX_SECOND_BB_HEAD_L		0x14
+#define CTX_SECOND_BB_STATE		0x16
+#define CTX_BB_PER_CTX_PTR		0x18
+#define CTX_RCS_INDIRECT_CTX		0x1a
+#define CTX_RCS_INDIRECT_CTX_OFFSET	0x1c
+#define CTX_LRI_HEADER_1		0x21
+#define CTX_CTX_TIMESTAMP		0x22
+#define CTX_PDP3_UDW			0x24
+#define CTX_PDP3_LDW			0x26
+#define CTX_PDP2_UDW			0x28
+#define CTX_PDP2_LDW			0x2a
+#define CTX_PDP1_UDW			0x2c
+#define CTX_PDP1_LDW			0x2e
+#define CTX_PDP0_UDW			0x30
+#define CTX_PDP0_LDW			0x32
+#define CTX_LRI_HEADER_2		0x41
+#define CTX_R_PWR_CLK_STATE		0x42
+#define CTX_GPGPU_CSR_BASE_ADDRESS	0x44
+
+#define GEN8_CTX_VALID (1<<0)
+#define GEN8_CTX_FORCE_PD_RESTORE (1<<1)
+#define GEN8_CTX_FORCE_RESTORE (1<<2)
+#define GEN8_CTX_L3LLC_COHERENT (1<<5)
+#define GEN8_CTX_PRIVILEGE (1<<8)
+enum {
+	ADVANCED_CONTEXT = 0,
+	LEGACY_CONTEXT,
+	ADVANCED_AD_CONTEXT,
+	LEGACY_64B_CONTEXT
+};
+#define GEN8_CTX_MODE_SHIFT 3
+enum {
+	FAULT_AND_HANG = 0,
+	FAULT_AND_HALT, /* Debug only */
+	FAULT_AND_STREAM,
+	FAULT_AND_CONTINUE /* Unsupported */
+};
+#define GEN8_CTX_ID_SHIFT 32
+
+static int intel_lr_context_pin(struct intel_engine_cs *ring,
+		struct intel_context *ctx);
+
+/**
+ * intel_sanitize_enable_execlists() - sanitize i915.enable_execlists
+ * @dev: DRM device.
+ * @enable_execlists: value of i915.enable_execlists module parameter.
+ *
+ * Only certain platforms support Execlists (the prerequisites being
+ * support for Logical Ring Contexts and Aliasing PPGTT or better),
+ * and only when enabled via module parameter.
+ *
+ * Return: 1 if Execlists is supported and has to be enabled.
+ */
+int intel_sanitize_enable_execlists(struct drm_device *dev, int enable_execlists)
+{
+	WARN_ON(i915.enable_ppgtt == -1);
+
+	if (INTEL_INFO(dev)->gen >= 9)
+		return 1;
+
+	if (enable_execlists == 0)
+		return 0;
+
+	if (HAS_LOGICAL_RING_CONTEXTS(dev) && USES_PPGTT(dev) &&
+	    i915.use_mmio_flip >= 0)
+		return 1;
+
+	return 0;
+}
+
+/**
+ * intel_execlists_ctx_id() - get the Execlists Context ID
+ * @ctx_obj: Logical Ring Context backing object.
+ *
+ * Do not confuse with ctx->id! Unfortunately we have a name overload
+ * here: the old context ID we pass to userspace as a handler so that
+ * they can refer to a context, and the new context ID we pass to the
+ * ELSP so that the GPU can inform us of the context status via
+ * interrupts.
+ *
+ * Return: 20-bits globally unique context ID.
+ */
+u32 intel_execlists_ctx_id(struct drm_i915_gem_object *ctx_obj)
+{
+	u32 lrca = i915_gem_obj_ggtt_offset(ctx_obj);
+
+	/* LRCA is required to be 4K aligned so the more significant 20 bits
+	 * are globally unique */
+	return lrca >> 12;
+}
+
+static uint64_t execlists_ctx_descriptor(struct drm_i915_gem_object *ctx_obj)
+{
+	uint64_t desc;
+	uint64_t lrca = i915_gem_obj_ggtt_offset(ctx_obj);
+
+	WARN_ON(lrca & 0xFFFFFFFF00000FFFULL);
+
+	desc = GEN8_CTX_VALID;
+	desc |= LEGACY_CONTEXT << GEN8_CTX_MODE_SHIFT;
+	desc |= GEN8_CTX_L3LLC_COHERENT;
+	desc |= GEN8_CTX_PRIVILEGE;
+	desc |= lrca;
+	desc |= (u64)intel_execlists_ctx_id(ctx_obj) << GEN8_CTX_ID_SHIFT;
+
+	/* TODO: WaDisableLiteRestore when we start using semaphore
+	 * signalling between Command Streamers */
+	/* desc |= GEN8_CTX_FORCE_RESTORE; */
+
+	return desc;
+}
+
+static void execlists_elsp_write(struct intel_engine_cs *ring,
+				 struct drm_i915_gem_object *ctx_obj0,
+				 struct drm_i915_gem_object *ctx_obj1)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	uint64_t temp = 0;
+	uint32_t desc[4];
+	unsigned long flags;
+
+	/* XXX: You must always write both descriptors in the order below. */
+	if (ctx_obj1)
+		temp = execlists_ctx_descriptor(ctx_obj1);
+	else
+		temp = 0;
+	desc[1] = (u32)(temp >> 32);
+	desc[0] = (u32)temp;
+
+	temp = execlists_ctx_descriptor(ctx_obj0);
+	desc[3] = (u32)(temp >> 32);
+	desc[2] = (u32)temp;
+
+	/* Set Force Wakeup bit to prevent GT from entering C6 while ELSP writes
+	 * are in progress.
+	 *
+	 * The other problem is that we can't just call gen6_gt_force_wake_get()
+	 * because that function calls intel_runtime_pm_get(), which might sleep.
+	 * Instead, we do the runtime_pm_get/put when creating/destroying requests.
+	 */
+	spin_lock_irqsave(&dev_priv->uncore.lock, flags);
+	if (IS_CHERRYVIEW(dev) || INTEL_INFO(dev)->gen >= 9) {
+		if (dev_priv->uncore.fw_rendercount++ == 0)
+			dev_priv->uncore.funcs.force_wake_get(dev_priv,
+							      FORCEWAKE_RENDER);
+		if (dev_priv->uncore.fw_mediacount++ == 0)
+			dev_priv->uncore.funcs.force_wake_get(dev_priv,
+							      FORCEWAKE_MEDIA);
+		if (INTEL_INFO(dev)->gen >= 9) {
+			if (dev_priv->uncore.fw_blittercount++ == 0)
+				dev_priv->uncore.funcs.force_wake_get(dev_priv,
+							FORCEWAKE_BLITTER);
+		}
+	} else {
+		if (dev_priv->uncore.forcewake_count++ == 0)
+			dev_priv->uncore.funcs.force_wake_get(dev_priv,
+							      FORCEWAKE_ALL);
+	}
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, flags);
+
+	I915_WRITE(RING_ELSP(ring), desc[1]);
+	I915_WRITE(RING_ELSP(ring), desc[0]);
+	I915_WRITE(RING_ELSP(ring), desc[3]);
+	/* The context is automatically loaded after the following */
+	I915_WRITE(RING_ELSP(ring), desc[2]);
+
+	/* ELSP is a wo register, so use another nearby reg for posting instead */
+	POSTING_READ(RING_EXECLIST_STATUS(ring));
+
+	/* Release Force Wakeup (see the big comment above). */
+	spin_lock_irqsave(&dev_priv->uncore.lock, flags);
+	if (IS_CHERRYVIEW(dev) || INTEL_INFO(dev)->gen >= 9) {
+		if (--dev_priv->uncore.fw_rendercount == 0)
+			dev_priv->uncore.funcs.force_wake_put(dev_priv,
+							      FORCEWAKE_RENDER);
+		if (--dev_priv->uncore.fw_mediacount == 0)
+			dev_priv->uncore.funcs.force_wake_put(dev_priv,
+							      FORCEWAKE_MEDIA);
+		if (INTEL_INFO(dev)->gen >= 9) {
+			if (--dev_priv->uncore.fw_blittercount == 0)
+				dev_priv->uncore.funcs.force_wake_put(dev_priv,
+							FORCEWAKE_BLITTER);
+		}
+	} else {
+		if (--dev_priv->uncore.forcewake_count == 0)
+			dev_priv->uncore.funcs.force_wake_put(dev_priv,
+							      FORCEWAKE_ALL);
+	}
+
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, flags);
+}
+
+static int execlists_update_context(struct drm_i915_gem_object *ctx_obj,
+				    struct drm_i915_gem_object *ring_obj,
+				    u32 tail)
+{
+	struct page *page;
+	uint32_t *reg_state;
+
+	page = i915_gem_object_get_page(ctx_obj, 1);
+	reg_state = kmap_atomic(page);
+
+	reg_state[CTX_RING_TAIL+1] = tail;
+	reg_state[CTX_RING_BUFFER_START+1] = i915_gem_obj_ggtt_offset(ring_obj);
+
+	kunmap_atomic(reg_state);
+
+	return 0;
+}
+
+static void execlists_submit_contexts(struct intel_engine_cs *ring,
+				      struct intel_context *to0, u32 tail0,
+				      struct intel_context *to1, u32 tail1)
+{
+	struct drm_i915_gem_object *ctx_obj0 = to0->engine[ring->id].state;
+	struct intel_ringbuffer *ringbuf0 = to0->engine[ring->id].ringbuf;
+	struct drm_i915_gem_object *ctx_obj1 = NULL;
+	struct intel_ringbuffer *ringbuf1 = NULL;
+
+	BUG_ON(!ctx_obj0);
+	WARN_ON(!i915_gem_obj_is_pinned(ctx_obj0));
+	WARN_ON(!i915_gem_obj_is_pinned(ringbuf0->obj));
+
+	execlists_update_context(ctx_obj0, ringbuf0->obj, tail0);
+
+	if (to1) {
+		ringbuf1 = to1->engine[ring->id].ringbuf;
+		ctx_obj1 = to1->engine[ring->id].state;
+		BUG_ON(!ctx_obj1);
+		WARN_ON(!i915_gem_obj_is_pinned(ctx_obj1));
+		WARN_ON(!i915_gem_obj_is_pinned(ringbuf1->obj));
+
+		execlists_update_context(ctx_obj1, ringbuf1->obj, tail1);
+	}
+
+	execlists_elsp_write(ring, ctx_obj0, ctx_obj1);
+}
+
+static void execlists_context_unqueue(struct intel_engine_cs *ring)
+{
+	struct intel_ctx_submit_request *req0 = NULL, *req1 = NULL;
+	struct intel_ctx_submit_request *cursor = NULL, *tmp = NULL;
+
+	assert_spin_locked(&ring->execlist_lock);
+
+	if (list_empty(&ring->execlist_queue))
+		return;
+
+	/* Try to read in pairs */
+	list_for_each_entry_safe(cursor, tmp, &ring->execlist_queue,
+				 execlist_link) {
+		if (!req0) {
+			req0 = cursor;
+		} else if (req0->ctx == cursor->ctx) {
+			/* Same ctx: ignore first request, as second request
+			 * will update tail past first request's workload */
+			cursor->elsp_submitted = req0->elsp_submitted;
+			list_del(&req0->execlist_link);
+			list_add_tail(&req0->execlist_link,
+				&ring->execlist_retired_req_list);
+			req0 = cursor;
+		} else {
+			req1 = cursor;
+			break;
+		}
+	}
+
+	WARN_ON(req1 && req1->elsp_submitted);
+
+	execlists_submit_contexts(ring, req0->ctx, req0->tail,
+				  req1 ? req1->ctx : NULL,
+				  req1 ? req1->tail : 0);
+
+	req0->elsp_submitted++;
+	if (req1)
+		req1->elsp_submitted++;
+}
+
+static bool execlists_check_remove_request(struct intel_engine_cs *ring,
+					   u32 request_id)
+{
+	struct intel_ctx_submit_request *head_req;
+
+	assert_spin_locked(&ring->execlist_lock);
+
+	head_req = list_first_entry_or_null(&ring->execlist_queue,
+					    struct intel_ctx_submit_request,
+					    execlist_link);
+
+	if (head_req != NULL) {
+		struct drm_i915_gem_object *ctx_obj =
+				head_req->ctx->engine[ring->id].state;
+		if (intel_execlists_ctx_id(ctx_obj) == request_id) {
+			WARN(head_req->elsp_submitted == 0,
+			     "Never submitted head request\n");
+
+			if (--head_req->elsp_submitted <= 0) {
+				list_del(&head_req->execlist_link);
+				list_add_tail(&head_req->execlist_link,
+					&ring->execlist_retired_req_list);
+				return true;
+			}
+		}
+	}
+
+	return false;
+}
+
+/**
+ * intel_execlists_handle_ctx_events() - handle Context Switch interrupts
+ * @ring: Engine Command Streamer to handle.
+ *
+ * Check the unread Context Status Buffers and manage the submission of new
+ * contexts to the ELSP accordingly.
+ */
+void intel_execlists_handle_ctx_events(struct intel_engine_cs *ring)
+{
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+	u32 status_pointer;
+	u8 read_pointer;
+	u8 write_pointer;
+	u32 status;
+	u32 status_id;
+	u32 submit_contexts = 0;
+
+	status_pointer = I915_READ(RING_CONTEXT_STATUS_PTR(ring));
+
+	read_pointer = ring->next_context_status_buffer;
+	write_pointer = status_pointer & 0x07;
+	if (read_pointer > write_pointer)
+		write_pointer += 6;
+
+	spin_lock(&ring->execlist_lock);
+
+	while (read_pointer < write_pointer) {
+		read_pointer++;
+		status = I915_READ(RING_CONTEXT_STATUS_BUF(ring) +
+				(read_pointer % 6) * 8);
+		status_id = I915_READ(RING_CONTEXT_STATUS_BUF(ring) +
+				(read_pointer % 6) * 8 + 4);
+
+		if (status & GEN8_CTX_STATUS_PREEMPTED) {
+			if (status & GEN8_CTX_STATUS_LITE_RESTORE) {
+				if (execlists_check_remove_request(ring, status_id))
+					WARN(1, "Lite Restored request removed from queue\n");
+			} else
+				WARN(1, "Preemption without Lite Restore\n");
+		}
+
+		 if ((status & GEN8_CTX_STATUS_ACTIVE_IDLE) ||
+		     (status & GEN8_CTX_STATUS_ELEMENT_SWITCH)) {
+			if (execlists_check_remove_request(ring, status_id))
+				submit_contexts++;
+		}
+	}
+
+	if (submit_contexts != 0)
+		execlists_context_unqueue(ring);
+
+	spin_unlock(&ring->execlist_lock);
+
+	WARN(submit_contexts > 2, "More than two context complete events?\n");
+	ring->next_context_status_buffer = write_pointer % 6;
+
+	I915_WRITE(RING_CONTEXT_STATUS_PTR(ring),
+		   ((u32)ring->next_context_status_buffer & 0x07) << 8);
+}
+
+static int execlists_context_queue(struct intel_engine_cs *ring,
+				   struct intel_context *to,
+				   u32 tail)
+{
+	struct intel_ctx_submit_request *req = NULL, *cursor;
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+	unsigned long flags;
+	int num_elements = 0;
+
+	req = kzalloc(sizeof(*req), GFP_KERNEL);
+	if (req == NULL)
+		return -ENOMEM;
+	req->ctx = to;
+	i915_gem_context_reference(req->ctx);
+
+	if (to != ring->default_context)
+		intel_lr_context_pin(ring, to);
+
+	req->ring = ring;
+	req->tail = tail;
+
+	intel_runtime_pm_get(dev_priv);
+
+	spin_lock_irqsave(&ring->execlist_lock, flags);
+
+	list_for_each_entry(cursor, &ring->execlist_queue, execlist_link)
+		if (++num_elements > 2)
+			break;
+
+	if (num_elements > 2) {
+		struct intel_ctx_submit_request *tail_req;
+
+		tail_req = list_last_entry(&ring->execlist_queue,
+					   struct intel_ctx_submit_request,
+					   execlist_link);
+
+		if (to == tail_req->ctx) {
+			WARN(tail_req->elsp_submitted != 0,
+				"More than 2 already-submitted reqs queued\n");
+			list_del(&tail_req->execlist_link);
+			list_add_tail(&tail_req->execlist_link,
+				&ring->execlist_retired_req_list);
+		}
+	}
+
+	list_add_tail(&req->execlist_link, &ring->execlist_queue);
+	if (num_elements == 0)
+		execlists_context_unqueue(ring);
+
+	spin_unlock_irqrestore(&ring->execlist_lock, flags);
+
+	return 0;
+}
+
+static int logical_ring_invalidate_all_caches(struct intel_ringbuffer *ringbuf)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	uint32_t flush_domains;
+	int ret;
+
+	flush_domains = 0;
+	if (ring->gpu_caches_dirty)
+		flush_domains = I915_GEM_GPU_DOMAINS;
+
+	ret = ring->emit_flush(ringbuf, I915_GEM_GPU_DOMAINS, flush_domains);
+	if (ret)
+		return ret;
+
+	ring->gpu_caches_dirty = false;
+	return 0;
+}
+
+static int execlists_move_to_gpu(struct intel_ringbuffer *ringbuf,
+				 struct list_head *vmas)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	struct i915_vma *vma;
+	uint32_t flush_domains = 0;
+	bool flush_chipset = false;
+	int ret;
+
+	list_for_each_entry(vma, vmas, exec_list) {
+		struct drm_i915_gem_object *obj = vma->obj;
+
+		ret = i915_gem_object_sync(obj, ring);
+		if (ret)
+			return ret;
+
+		if (obj->base.write_domain & I915_GEM_DOMAIN_CPU)
+			flush_chipset |= i915_gem_clflush_object(obj, false);
+
+		flush_domains |= obj->base.write_domain;
+	}
+
+	if (flush_domains & I915_GEM_DOMAIN_GTT)
+		wmb();
+
+	/* Unconditionally invalidate gpu caches and ensure that we do flush
+	 * any residual writes from the previous batch.
+	 */
+	return logical_ring_invalidate_all_caches(ringbuf);
+}
+
+/**
+ * execlists_submission() - submit a batchbuffer for execution, Execlists style
+ * @dev: DRM device.
+ * @file: DRM file.
+ * @ring: Engine Command Streamer to submit to.
+ * @ctx: Context to employ for this submission.
+ * @args: execbuffer call arguments.
+ * @vmas: list of vmas.
+ * @batch_obj: the batchbuffer to submit.
+ * @exec_start: batchbuffer start virtual address pointer.
+ * @flags: translated execbuffer call flags.
+ *
+ * This is the evil twin version of i915_gem_ringbuffer_submission. It abstracts
+ * away the submission details of the execbuffer ioctl call.
+ *
+ * Return: non-zero if the submission fails.
+ */
+int intel_execlists_submission(struct drm_device *dev, struct drm_file *file,
+			       struct intel_engine_cs *ring,
+			       struct intel_context *ctx,
+			       struct drm_i915_gem_execbuffer2 *args,
+			       struct list_head *vmas,
+			       struct drm_i915_gem_object *batch_obj,
+			       u64 exec_start, u32 flags)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_ringbuffer *ringbuf = ctx->engine[ring->id].ringbuf;
+	int instp_mode;
+	u32 instp_mask;
+	int ret;
+
+	instp_mode = args->flags & I915_EXEC_CONSTANTS_MASK;
+	instp_mask = I915_EXEC_CONSTANTS_MASK;
+	switch (instp_mode) {
+	case I915_EXEC_CONSTANTS_REL_GENERAL:
+	case I915_EXEC_CONSTANTS_ABSOLUTE:
+	case I915_EXEC_CONSTANTS_REL_SURFACE:
+		if (instp_mode != 0 && ring != &dev_priv->ring[RCS]) {
+			DRM_DEBUG("non-0 rel constants mode on non-RCS\n");
+			return -EINVAL;
+		}
+
+		if (instp_mode != dev_priv->relative_constants_mode) {
+			if (instp_mode == I915_EXEC_CONSTANTS_REL_SURFACE) {
+				DRM_DEBUG("rel surface constants mode invalid on gen5+\n");
+				return -EINVAL;
+			}
+
+			/* The HW changed the meaning on this bit on gen6 */
+			instp_mask &= ~I915_EXEC_CONSTANTS_REL_SURFACE;
+		}
+		break;
+	default:
+		DRM_DEBUG("execbuf with unknown constants: %d\n", instp_mode);
+		return -EINVAL;
+	}
+
+	if (args->num_cliprects != 0) {
+		DRM_DEBUG("clip rectangles are only valid on pre-gen5\n");
+		return -EINVAL;
+	} else {
+		if (args->DR4 == 0xffffffff) {
+			DRM_DEBUG("UXA submitting garbage DR4, fixing up\n");
+			args->DR4 = 0;
+		}
+
+		if (args->DR1 || args->DR4 || args->cliprects_ptr) {
+			DRM_DEBUG("0 cliprects but dirt in cliprects fields\n");
+			return -EINVAL;
+		}
+	}
+
+	if (args->flags & I915_EXEC_GEN7_SOL_RESET) {
+		DRM_DEBUG("sol reset is gen7 only\n");
+		return -EINVAL;
+	}
+
+	ret = execlists_move_to_gpu(ringbuf, vmas);
+	if (ret)
+		return ret;
+
+	if (ring == &dev_priv->ring[RCS] &&
+	    instp_mode != dev_priv->relative_constants_mode) {
+		ret = intel_logical_ring_begin(ringbuf, 4);
+		if (ret)
+			return ret;
+
+		intel_logical_ring_emit(ringbuf, MI_NOOP);
+		intel_logical_ring_emit(ringbuf, MI_LOAD_REGISTER_IMM(1));
+		intel_logical_ring_emit(ringbuf, INSTPM);
+		intel_logical_ring_emit(ringbuf, instp_mask << 16 | instp_mode);
+		intel_logical_ring_advance(ringbuf);
+
+		dev_priv->relative_constants_mode = instp_mode;
+	}
+
+	ret = ring->emit_bb_start(ringbuf, exec_start, flags);
+	if (ret)
+		return ret;
+
+	i915_gem_execbuffer_move_to_active(vmas, ring);
+	i915_gem_execbuffer_retire_commands(dev, file, ring, batch_obj);
+
+	return 0;
+}
+
+void intel_execlists_retire_requests(struct intel_engine_cs *ring)
+{
+	struct intel_ctx_submit_request *req, *tmp;
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+	unsigned long flags;
+	struct list_head retired_list;
+
+	WARN_ON(!mutex_is_locked(&ring->dev->struct_mutex));
+	if (list_empty(&ring->execlist_retired_req_list))
+		return;
+
+	INIT_LIST_HEAD(&retired_list);
+	spin_lock_irqsave(&ring->execlist_lock, flags);
+	list_replace_init(&ring->execlist_retired_req_list, &retired_list);
+	spin_unlock_irqrestore(&ring->execlist_lock, flags);
+
+	list_for_each_entry_safe(req, tmp, &retired_list, execlist_link) {
+		struct intel_context *ctx = req->ctx;
+		struct drm_i915_gem_object *ctx_obj =
+				ctx->engine[ring->id].state;
+
+		if (ctx_obj && (ctx != ring->default_context))
+			intel_lr_context_unpin(ring, ctx);
+		intel_runtime_pm_put(dev_priv);
+		i915_gem_context_unreference(req->ctx);
+		list_del(&req->execlist_link);
+		kfree(req);
+	}
+}
+
+void intel_logical_ring_stop(struct intel_engine_cs *ring)
+{
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+	int ret;
+
+	if (!intel_ring_initialized(ring))
+		return;
+
+	ret = intel_ring_idle(ring);
+	if (ret && !i915_reset_in_progress(&to_i915(ring->dev)->gpu_error))
+		DRM_ERROR("failed to quiesce %s whilst cleaning up: %d\n",
+			  ring->name, ret);
+
+	/* TODO: Is this correct with Execlists enabled? */
+	I915_WRITE_MODE(ring, _MASKED_BIT_ENABLE(STOP_RING));
+	if (wait_for_atomic((I915_READ_MODE(ring) & MODE_IDLE) != 0, 1000)) {
+		DRM_ERROR("%s :timed out trying to stop ring\n", ring->name);
+		return;
+	}
+	I915_WRITE_MODE(ring, _MASKED_BIT_DISABLE(STOP_RING));
+}
+
+int logical_ring_flush_all_caches(struct intel_ringbuffer *ringbuf)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	int ret;
+
+	if (!ring->gpu_caches_dirty)
+		return 0;
+
+	ret = ring->emit_flush(ringbuf, 0, I915_GEM_GPU_DOMAINS);
+	if (ret)
+		return ret;
+
+	ring->gpu_caches_dirty = false;
+	return 0;
+}
+
+/**
+ * intel_logical_ring_advance_and_submit() - advance the tail and submit the workload
+ * @ringbuf: Logical Ringbuffer to advance.
+ *
+ * The tail is updated in our logical ringbuffer struct, not in the actual context. What
+ * really happens during submission is that the context and current tail will be placed
+ * on a queue waiting for the ELSP to be ready to accept a new context submission. At that
+ * point, the tail *inside* the context is updated and the ELSP written to.
+ */
+void intel_logical_ring_advance_and_submit(struct intel_ringbuffer *ringbuf)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	struct intel_context *ctx = ringbuf->FIXME_lrc_ctx;
+
+	intel_logical_ring_advance(ringbuf);
+
+	if (intel_ring_stopped(ring))
+		return;
+
+	execlists_context_queue(ring, ctx, ringbuf->tail);
+}
+
+static int intel_lr_context_pin(struct intel_engine_cs *ring,
+		struct intel_context *ctx)
+{
+	struct drm_i915_gem_object *ctx_obj = ctx->engine[ring->id].state;
+	struct intel_ringbuffer *ringbuf = ctx->engine[ring->id].ringbuf;
+	int ret = 0;
+
+	WARN_ON(!mutex_is_locked(&ring->dev->struct_mutex));
+	if (ctx->engine[ring->id].unpin_count++ == 0) {
+		ret = i915_gem_obj_ggtt_pin(ctx_obj,
+				GEN8_LR_CONTEXT_ALIGN, 0);
+		if (ret)
+			goto reset_unpin_count;
+
+		ret = intel_pin_and_map_ringbuffer_obj(ring->dev, ringbuf);
+		if (ret)
+			goto unpin_ctx_obj;
+	}
+
+	return ret;
+
+unpin_ctx_obj:
+	i915_gem_object_ggtt_unpin(ctx_obj);
+reset_unpin_count:
+	ctx->engine[ring->id].unpin_count = 0;
+
+	return ret;
+}
+
+void intel_lr_context_unpin(struct intel_engine_cs *ring,
+		struct intel_context *ctx)
+{
+	struct drm_i915_gem_object *ctx_obj = ctx->engine[ring->id].state;
+	struct intel_ringbuffer *ringbuf = ctx->engine[ring->id].ringbuf;
+
+	if (ctx_obj) {
+		WARN_ON(!mutex_is_locked(&ring->dev->struct_mutex));
+		if (--ctx->engine[ring->id].unpin_count == 0) {
+			intel_unpin_ringbuffer_obj(ringbuf);
+			i915_gem_object_ggtt_unpin(ctx_obj);
+		}
+	}
+}
+
+static int logical_ring_alloc_seqno(struct intel_engine_cs *ring,
+				    struct intel_context *ctx)
+{
+	int ret;
+
+	if (ring->outstanding_lazy_seqno)
+		return 0;
+
+	if (ring->preallocated_lazy_request == NULL) {
+		struct drm_i915_gem_request *request;
+
+		request = kmalloc(sizeof(*request), GFP_KERNEL);
+		if (request == NULL)
+			return -ENOMEM;
+
+		if (ctx != ring->default_context) {
+			ret = intel_lr_context_pin(ring, ctx);
+			if (ret) {
+				kfree(request);
+				return ret;
+			}
+		}
+
+		/* Hold a reference to the context this request belongs to
+		 * (we will need it when the time comes to emit/retire the
+		 * request).
+		 */
+		request->ctx = ctx;
+		i915_gem_context_reference(request->ctx);
+
+		ring->preallocated_lazy_request = request;
+	}
+
+	return i915_gem_get_seqno(ring->dev, &ring->outstanding_lazy_seqno);
+}
+
+static int logical_ring_wait_request(struct intel_ringbuffer *ringbuf,
+				     int bytes)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	struct drm_i915_gem_request *request;
+	u32 seqno = 0;
+	int ret;
+
+	if (ringbuf->last_retired_head != -1) {
+		ringbuf->head = ringbuf->last_retired_head;
+		ringbuf->last_retired_head = -1;
+
+		ringbuf->space = intel_ring_space(ringbuf);
+		if (ringbuf->space >= bytes)
+			return 0;
+	}
+
+	list_for_each_entry(request, &ring->request_list, list) {
+		if (__intel_ring_space(request->tail, ringbuf->tail,
+				       ringbuf->size) >= bytes) {
+			seqno = request->seqno;
+			break;
+		}
+	}
+
+	if (seqno == 0)
+		return -ENOSPC;
+
+	ret = i915_wait_seqno(ring, seqno);
+	if (ret)
+		return ret;
+
+	i915_gem_retire_requests_ring(ring);
+	ringbuf->head = ringbuf->last_retired_head;
+	ringbuf->last_retired_head = -1;
+
+	ringbuf->space = intel_ring_space(ringbuf);
+	return 0;
+}
+
+static int logical_ring_wait_for_space(struct intel_ringbuffer *ringbuf,
+				       int bytes)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	unsigned long end;
+	int ret;
+
+	ret = logical_ring_wait_request(ringbuf, bytes);
+	if (ret != -ENOSPC)
+		return ret;
+
+	/* Force the context submission in case we have been skipping it */
+	intel_logical_ring_advance_and_submit(ringbuf);
+
+	/* With GEM the hangcheck timer should kick us out of the loop,
+	 * leaving it early runs the risk of corrupting GEM state (due
+	 * to running on almost untested codepaths). But on resume
+	 * timers don't work yet, so prevent a complete hang in that
+	 * case by choosing an insanely large timeout. */
+	end = jiffies + 60 * HZ;
+
+	do {
+		ringbuf->head = I915_READ_HEAD(ring);
+		ringbuf->space = intel_ring_space(ringbuf);
+		if (ringbuf->space >= bytes) {
+			ret = 0;
+			break;
+		}
+
+		msleep(1);
+
+		if (dev_priv->mm.interruptible && signal_pending(current)) {
+			ret = -ERESTARTSYS;
+			break;
+		}
+
+		ret = i915_gem_check_wedge(&dev_priv->gpu_error,
+					   dev_priv->mm.interruptible);
+		if (ret)
+			break;
+
+		if (time_after(jiffies, end)) {
+			ret = -EBUSY;
+			break;
+		}
+	} while (1);
+
+	return ret;
+}
+
+static int logical_ring_wrap_buffer(struct intel_ringbuffer *ringbuf)
+{
+	uint32_t __iomem *virt;
+	int rem = ringbuf->size - ringbuf->tail;
+
+	if (ringbuf->space < rem) {
+		int ret = logical_ring_wait_for_space(ringbuf, rem);
+
+		if (ret)
+			return ret;
+	}
+
+	virt = ringbuf->virtual_start + ringbuf->tail;
+	rem /= 4;
+	while (rem--)
+		iowrite32(MI_NOOP, virt++);
+
+	ringbuf->tail = 0;
+	ringbuf->space = intel_ring_space(ringbuf);
+
+	return 0;
+}
+
+static int logical_ring_prepare(struct intel_ringbuffer *ringbuf, int bytes)
+{
+	int ret;
+
+	if (unlikely(ringbuf->tail + bytes > ringbuf->effective_size)) {
+		ret = logical_ring_wrap_buffer(ringbuf);
+		if (unlikely(ret))
+			return ret;
+	}
+
+	if (unlikely(ringbuf->space < bytes)) {
+		ret = logical_ring_wait_for_space(ringbuf, bytes);
+		if (unlikely(ret))
+			return ret;
+	}
+
+	return 0;
+}
+
+/**
+ * intel_logical_ring_begin() - prepare the logical ringbuffer to accept some commands
+ *
+ * @ringbuf: Logical ringbuffer.
+ * @num_dwords: number of DWORDs that we plan to write to the ringbuffer.
+ *
+ * The ringbuffer might not be ready to accept the commands right away (maybe it needs to
+ * be wrapped, or wait a bit for the tail to be updated). This function takes care of that
+ * and also preallocates a request (every workload submission is still mediated through
+ * requests, same as it did with legacy ringbuffer submission).
+ *
+ * Return: non-zero if the ringbuffer is not ready to be written to.
+ */
+int intel_logical_ring_begin(struct intel_ringbuffer *ringbuf, int num_dwords)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
+
+	ret = i915_gem_check_wedge(&dev_priv->gpu_error,
+				   dev_priv->mm.interruptible);
+	if (ret)
+		return ret;
+
+	ret = logical_ring_prepare(ringbuf, num_dwords * sizeof(uint32_t));
+	if (ret)
+		return ret;
+
+	/* Preallocate the olr before touching the ring */
+	ret = logical_ring_alloc_seqno(ring, ringbuf->FIXME_lrc_ctx);
+	if (ret)
+		return ret;
+
+	ringbuf->space -= num_dwords * sizeof(uint32_t);
+	return 0;
+}
+
+static int intel_logical_ring_workarounds_emit(struct intel_engine_cs *ring,
+					       struct intel_context *ctx)
+{
+	int ret, i;
+	struct intel_ringbuffer *ringbuf = ctx->engine[ring->id].ringbuf;
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct i915_workarounds *w = &dev_priv->workarounds;
+
+	if (WARN_ON(w->count == 0))
+		return 0;
+
+	ring->gpu_caches_dirty = true;
+	ret = logical_ring_flush_all_caches(ringbuf);
+	if (ret)
+		return ret;
+
+	ret = intel_logical_ring_begin(ringbuf, w->count * 2 + 2);
+	if (ret)
+		return ret;
+
+	intel_logical_ring_emit(ringbuf, MI_LOAD_REGISTER_IMM(w->count));
+	for (i = 0; i < w->count; i++) {
+		intel_logical_ring_emit(ringbuf, w->reg[i].addr);
+		intel_logical_ring_emit(ringbuf, w->reg[i].value);
+	}
+	intel_logical_ring_emit(ringbuf, MI_NOOP);
+
+	intel_logical_ring_advance(ringbuf);
+
+	ring->gpu_caches_dirty = true;
+	ret = logical_ring_flush_all_caches(ringbuf);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+static int gen8_init_common_ring(struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	I915_WRITE_IMR(ring, ~(ring->irq_enable_mask | ring->irq_keep_mask));
+	I915_WRITE(RING_HWSTAM(ring->mmio_base), 0xffffffff);
+
+	I915_WRITE(RING_MODE_GEN7(ring),
+		   _MASKED_BIT_DISABLE(GFX_REPLAY_MODE) |
+		   _MASKED_BIT_ENABLE(GFX_RUN_LIST_ENABLE));
+	POSTING_READ(RING_MODE_GEN7(ring));
+	DRM_DEBUG_DRIVER("Execlists enabled for %s\n", ring->name);
+
+	memset(&ring->hangcheck, 0, sizeof(ring->hangcheck));
+
+	return 0;
+}
+
+static int gen8_init_render_ring(struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
+
+	ret = gen8_init_common_ring(ring);
+	if (ret)
+		return ret;
+
+	/* We need to disable the AsyncFlip performance optimisations in order
+	 * to use MI_WAIT_FOR_EVENT within the CS. It should already be
+	 * programmed to '1' on all products.
+	 *
+	 * WaDisableAsyncFlipPerfMode:snb,ivb,hsw,vlv,bdw,chv
+	 */
+	I915_WRITE(MI_MODE, _MASKED_BIT_ENABLE(ASYNC_FLIP_PERF_DISABLE));
+
+	ret = intel_init_pipe_control(ring);
+	if (ret)
+		return ret;
+
+	I915_WRITE(INSTPM, _MASKED_BIT_ENABLE(INSTPM_FORCE_ORDERING));
+
+	return init_workarounds_ring(ring);
+}
+
+static int gen8_emit_bb_start(struct intel_ringbuffer *ringbuf,
+			      u64 offset, unsigned flags)
+{
+	bool ppgtt = !(flags & I915_DISPATCH_SECURE);
+	int ret;
+
+	ret = intel_logical_ring_begin(ringbuf, 4);
+	if (ret)
+		return ret;
+
+	/* FIXME(BDW): Address space and security selectors. */
+	intel_logical_ring_emit(ringbuf, MI_BATCH_BUFFER_START_GEN8 | (ppgtt<<8));
+	intel_logical_ring_emit(ringbuf, lower_32_bits(offset));
+	intel_logical_ring_emit(ringbuf, upper_32_bits(offset));
+	intel_logical_ring_emit(ringbuf, MI_NOOP);
+	intel_logical_ring_advance(ringbuf);
+
+	return 0;
+}
+
+static bool gen8_logical_ring_get_irq(struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	unsigned long flags;
+
+	if (WARN_ON(!intel_irqs_enabled(dev_priv)))
+		return false;
+
+	spin_lock_irqsave(&dev_priv->irq_lock, flags);
+	if (ring->irq_refcount++ == 0) {
+		I915_WRITE_IMR(ring, ~(ring->irq_enable_mask | ring->irq_keep_mask));
+		POSTING_READ(RING_IMR(ring->mmio_base));
+	}
+	spin_unlock_irqrestore(&dev_priv->irq_lock, flags);
+
+	return true;
+}
+
+static void gen8_logical_ring_put_irq(struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	unsigned long flags;
+
+	spin_lock_irqsave(&dev_priv->irq_lock, flags);
+	if (--ring->irq_refcount == 0) {
+		I915_WRITE_IMR(ring, ~ring->irq_keep_mask);
+		POSTING_READ(RING_IMR(ring->mmio_base));
+	}
+	spin_unlock_irqrestore(&dev_priv->irq_lock, flags);
+}
+
+static int gen8_emit_flush(struct intel_ringbuffer *ringbuf,
+			   u32 invalidate_domains,
+			   u32 unused)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	uint32_t cmd;
+	int ret;
+
+	ret = intel_logical_ring_begin(ringbuf, 4);
+	if (ret)
+		return ret;
+
+	cmd = MI_FLUSH_DW + 1;
+
+	if (ring == &dev_priv->ring[VCS]) {
+		if (invalidate_domains & I915_GEM_GPU_DOMAINS)
+			cmd |= MI_INVALIDATE_TLB | MI_INVALIDATE_BSD |
+				MI_FLUSH_DW_STORE_INDEX |
+				MI_FLUSH_DW_OP_STOREDW;
+	} else {
+		if (invalidate_domains & I915_GEM_DOMAIN_RENDER)
+			cmd |= MI_INVALIDATE_TLB | MI_FLUSH_DW_STORE_INDEX |
+				MI_FLUSH_DW_OP_STOREDW;
+	}
+
+	intel_logical_ring_emit(ringbuf, cmd);
+	intel_logical_ring_emit(ringbuf,
+				I915_GEM_HWS_SCRATCH_ADDR |
+				MI_FLUSH_DW_USE_GTT);
+	intel_logical_ring_emit(ringbuf, 0); /* upper addr */
+	intel_logical_ring_emit(ringbuf, 0); /* value */
+	intel_logical_ring_advance(ringbuf);
+
+	return 0;
+}
+
+static int gen8_emit_flush_render(struct intel_ringbuffer *ringbuf,
+				  u32 invalidate_domains,
+				  u32 flush_domains)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	u32 scratch_addr = ring->scratch.gtt_offset + 2 * CACHELINE_BYTES;
+	u32 flags = 0;
+	int ret;
+
+	flags |= PIPE_CONTROL_CS_STALL;
+
+	if (flush_domains) {
+		flags |= PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH;
+		flags |= PIPE_CONTROL_DEPTH_CACHE_FLUSH;
+	}
+
+	if (invalidate_domains) {
+		flags |= PIPE_CONTROL_TLB_INVALIDATE;
+		flags |= PIPE_CONTROL_INSTRUCTION_CACHE_INVALIDATE;
+		flags |= PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE;
+		flags |= PIPE_CONTROL_VF_CACHE_INVALIDATE;
+		flags |= PIPE_CONTROL_CONST_CACHE_INVALIDATE;
+		flags |= PIPE_CONTROL_STATE_CACHE_INVALIDATE;
+		flags |= PIPE_CONTROL_QW_WRITE;
+		flags |= PIPE_CONTROL_GLOBAL_GTT_IVB;
+	}
+
+	ret = intel_logical_ring_begin(ringbuf, 6);
+	if (ret)
+		return ret;
+
+	intel_logical_ring_emit(ringbuf, GFX_OP_PIPE_CONTROL(6));
+	intel_logical_ring_emit(ringbuf, flags);
+	intel_logical_ring_emit(ringbuf, scratch_addr);
+	intel_logical_ring_emit(ringbuf, 0);
+	intel_logical_ring_emit(ringbuf, 0);
+	intel_logical_ring_emit(ringbuf, 0);
+	intel_logical_ring_advance(ringbuf);
+
+	return 0;
+}
+
+static u32 gen8_get_seqno(struct intel_engine_cs *ring, bool lazy_coherency)
+{
+	return intel_read_status_page(ring, I915_GEM_HWS_INDEX);
+}
+
+static void gen8_set_seqno(struct intel_engine_cs *ring, u32 seqno)
+{
+	intel_write_status_page(ring, I915_GEM_HWS_INDEX, seqno);
+}
+
+static int gen8_emit_request(struct intel_ringbuffer *ringbuf)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	u32 cmd;
+	int ret;
+
+	ret = intel_logical_ring_begin(ringbuf, 6);
+	if (ret)
+		return ret;
+
+	cmd = MI_STORE_DWORD_IMM_GEN8;
+	cmd |= MI_GLOBAL_GTT;
+
+	intel_logical_ring_emit(ringbuf, cmd);
+	intel_logical_ring_emit(ringbuf,
+				(ring->status_page.gfx_addr +
+				(I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT)));
+	intel_logical_ring_emit(ringbuf, 0);
+	intel_logical_ring_emit(ringbuf, ring->outstanding_lazy_seqno);
+	intel_logical_ring_emit(ringbuf, MI_USER_INTERRUPT);
+	intel_logical_ring_emit(ringbuf, MI_NOOP);
+	intel_logical_ring_advance_and_submit(ringbuf);
+
+	return 0;
+}
+
+/**
+ * intel_logical_ring_cleanup() - deallocate the Engine Command Streamer
+ *
+ * @ring: Engine Command Streamer.
+ *
+ */
+void intel_logical_ring_cleanup(struct intel_engine_cs *ring)
+{
+	struct drm_i915_private *dev_priv;
+
+	if (!intel_ring_initialized(ring))
+		return;
+
+	dev_priv = ring->dev->dev_private;
+
+	intel_logical_ring_stop(ring);
+	WARN_ON((I915_READ_MODE(ring) & MODE_IDLE) == 0);
+	ring->preallocated_lazy_request = NULL;
+	ring->outstanding_lazy_seqno = 0;
+
+	if (ring->cleanup)
+		ring->cleanup(ring);
+
+	i915_cmd_parser_fini_ring(ring);
+
+	if (ring->status_page.obj) {
+		kunmap(sg_page(ring->status_page.obj->pages->sgl));
+		ring->status_page.obj = NULL;
+	}
+}
+
+static int logical_ring_init(struct drm_device *dev, struct intel_engine_cs *ring)
+{
+	int ret;
+
+	/* Intentionally left blank. */
+	ring->buffer = NULL;
+
+	ring->dev = dev;
+	INIT_LIST_HEAD(&ring->active_list);
+	INIT_LIST_HEAD(&ring->request_list);
+	init_waitqueue_head(&ring->irq_queue);
+
+	INIT_LIST_HEAD(&ring->execlist_queue);
+	INIT_LIST_HEAD(&ring->execlist_retired_req_list);
+	spin_lock_init(&ring->execlist_lock);
+	ring->next_context_status_buffer = 0;
+
+	ret = i915_cmd_parser_init_ring(ring);
+	if (ret)
+		return ret;
+
+	if (ring->init) {
+		ret = ring->init(ring);
+		if (ret)
+			return ret;
+	}
+
+	ret = intel_lr_context_deferred_create(ring->default_context, ring);
+
+	return ret;
+}
+
+static int logical_render_ring_init(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[RCS];
+
+	ring->name = "render ring";
+	ring->id = RCS;
+	ring->mmio_base = RENDER_RING_BASE;
+	ring->irq_enable_mask =
+		GT_RENDER_USER_INTERRUPT << GEN8_RCS_IRQ_SHIFT;
+	ring->irq_keep_mask =
+		GT_CONTEXT_SWITCH_INTERRUPT << GEN8_RCS_IRQ_SHIFT;
+	if (HAS_L3_DPF(dev))
+		ring->irq_keep_mask |= GT_RENDER_L3_PARITY_ERROR_INTERRUPT;
+
+	ring->init = gen8_init_render_ring;
+	ring->init_context = intel_logical_ring_workarounds_emit;
+	ring->cleanup = intel_fini_pipe_control;
+	ring->get_seqno = gen8_get_seqno;
+	ring->set_seqno = gen8_set_seqno;
+	ring->emit_request = gen8_emit_request;
+	ring->emit_flush = gen8_emit_flush_render;
+	ring->irq_get = gen8_logical_ring_get_irq;
+	ring->irq_put = gen8_logical_ring_put_irq;
+	ring->emit_bb_start = gen8_emit_bb_start;
+
+	return logical_ring_init(dev, ring);
+}
+
+static int logical_bsd_ring_init(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[VCS];
+
+	ring->name = "bsd ring";
+	ring->id = VCS;
+	ring->mmio_base = GEN6_BSD_RING_BASE;
+	ring->irq_enable_mask =
+		GT_RENDER_USER_INTERRUPT << GEN8_VCS1_IRQ_SHIFT;
+	ring->irq_keep_mask =
+		GT_CONTEXT_SWITCH_INTERRUPT << GEN8_VCS1_IRQ_SHIFT;
+
+	ring->init = gen8_init_common_ring;
+	ring->get_seqno = gen8_get_seqno;
+	ring->set_seqno = gen8_set_seqno;
+	ring->emit_request = gen8_emit_request;
+	ring->emit_flush = gen8_emit_flush;
+	ring->irq_get = gen8_logical_ring_get_irq;
+	ring->irq_put = gen8_logical_ring_put_irq;
+	ring->emit_bb_start = gen8_emit_bb_start;
+
+	return logical_ring_init(dev, ring);
+}
+
+static int logical_bsd2_ring_init(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[VCS2];
+
+	ring->name = "bds2 ring";
+	ring->id = VCS2;
+	ring->mmio_base = GEN8_BSD2_RING_BASE;
+	ring->irq_enable_mask =
+		GT_RENDER_USER_INTERRUPT << GEN8_VCS2_IRQ_SHIFT;
+	ring->irq_keep_mask =
+		GT_CONTEXT_SWITCH_INTERRUPT << GEN8_VCS2_IRQ_SHIFT;
+
+	ring->init = gen8_init_common_ring;
+	ring->get_seqno = gen8_get_seqno;
+	ring->set_seqno = gen8_set_seqno;
+	ring->emit_request = gen8_emit_request;
+	ring->emit_flush = gen8_emit_flush;
+	ring->irq_get = gen8_logical_ring_get_irq;
+	ring->irq_put = gen8_logical_ring_put_irq;
+	ring->emit_bb_start = gen8_emit_bb_start;
+
+	return logical_ring_init(dev, ring);
+}
+
+static int logical_blt_ring_init(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[BCS];
+
+	ring->name = "blitter ring";
+	ring->id = BCS;
+	ring->mmio_base = BLT_RING_BASE;
+	ring->irq_enable_mask =
+		GT_RENDER_USER_INTERRUPT << GEN8_BCS_IRQ_SHIFT;
+	ring->irq_keep_mask =
+		GT_CONTEXT_SWITCH_INTERRUPT << GEN8_BCS_IRQ_SHIFT;
+
+	ring->init = gen8_init_common_ring;
+	ring->get_seqno = gen8_get_seqno;
+	ring->set_seqno = gen8_set_seqno;
+	ring->emit_request = gen8_emit_request;
+	ring->emit_flush = gen8_emit_flush;
+	ring->irq_get = gen8_logical_ring_get_irq;
+	ring->irq_put = gen8_logical_ring_put_irq;
+	ring->emit_bb_start = gen8_emit_bb_start;
+
+	return logical_ring_init(dev, ring);
+}
+
+static int logical_vebox_ring_init(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[VECS];
+
+	ring->name = "video enhancement ring";
+	ring->id = VECS;
+	ring->mmio_base = VEBOX_RING_BASE;
+	ring->irq_enable_mask =
+		GT_RENDER_USER_INTERRUPT << GEN8_VECS_IRQ_SHIFT;
+	ring->irq_keep_mask =
+		GT_CONTEXT_SWITCH_INTERRUPT << GEN8_VECS_IRQ_SHIFT;
+
+	ring->init = gen8_init_common_ring;
+	ring->get_seqno = gen8_get_seqno;
+	ring->set_seqno = gen8_set_seqno;
+	ring->emit_request = gen8_emit_request;
+	ring->emit_flush = gen8_emit_flush;
+	ring->irq_get = gen8_logical_ring_get_irq;
+	ring->irq_put = gen8_logical_ring_put_irq;
+	ring->emit_bb_start = gen8_emit_bb_start;
+
+	return logical_ring_init(dev, ring);
+}
+
+/**
+ * intel_logical_rings_init() - allocate, populate and init the Engine Command Streamers
+ * @dev: DRM device.
+ *
+ * This function inits the engines for an Execlists submission style (the equivalent in the
+ * legacy ringbuffer submission world would be i915_gem_init_rings). It does it only for
+ * those engines that are present in the hardware.
+ *
+ * Return: non-zero if the initialization failed.
+ */
+int intel_logical_rings_init(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
+
+	ret = logical_render_ring_init(dev);
+	if (ret)
+		return ret;
+
+	if (HAS_BSD(dev)) {
+		ret = logical_bsd_ring_init(dev);
+		if (ret)
+			goto cleanup_render_ring;
+	}
+
+	if (HAS_BLT(dev)) {
+		ret = logical_blt_ring_init(dev);
+		if (ret)
+			goto cleanup_bsd_ring;
+	}
+
+	if (HAS_VEBOX(dev)) {
+		ret = logical_vebox_ring_init(dev);
+		if (ret)
+			goto cleanup_blt_ring;
+	}
+
+	if (HAS_BSD2(dev)) {
+		ret = logical_bsd2_ring_init(dev);
+		if (ret)
+			goto cleanup_vebox_ring;
+	}
+
+	ret = i915_gem_set_seqno(dev, ((u32)~0 - 0x1000));
+	if (ret)
+		goto cleanup_bsd2_ring;
+
+	return 0;
+
+cleanup_bsd2_ring:
+	intel_logical_ring_cleanup(&dev_priv->ring[VCS2]);
+cleanup_vebox_ring:
+	intel_logical_ring_cleanup(&dev_priv->ring[VECS]);
+cleanup_blt_ring:
+	intel_logical_ring_cleanup(&dev_priv->ring[BCS]);
+cleanup_bsd_ring:
+	intel_logical_ring_cleanup(&dev_priv->ring[VCS]);
+cleanup_render_ring:
+	intel_logical_ring_cleanup(&dev_priv->ring[RCS]);
+
+	return ret;
+}
+
+int intel_lr_context_render_state_init(struct intel_engine_cs *ring,
+				       struct intel_context *ctx)
+{
+	struct intel_ringbuffer *ringbuf = ctx->engine[ring->id].ringbuf;
+	struct render_state so;
+	struct drm_i915_file_private *file_priv = ctx->file_priv;
+	struct drm_file *file = file_priv ? file_priv->file : NULL;
+	int ret;
+
+	ret = i915_gem_render_state_prepare(ring, &so);
+	if (ret)
+		return ret;
+
+	if (so.rodata == NULL)
+		return 0;
+
+	ret = ring->emit_bb_start(ringbuf,
+			so.ggtt_offset,
+			I915_DISPATCH_SECURE);
+	if (ret)
+		goto out;
+
+	i915_vma_move_to_active(i915_gem_obj_to_ggtt(so.obj), ring);
+
+	ret = __i915_add_request(ring, file, so.obj, NULL);
+	/* intel_logical_ring_add_request moves object to inactive if it
+	 * fails */
+out:
+	i915_gem_render_state_fini(&so);
+	return ret;
+}
+
+static int
+populate_lr_context(struct intel_context *ctx, struct drm_i915_gem_object *ctx_obj,
+		    struct intel_engine_cs *ring, struct intel_ringbuffer *ringbuf)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct i915_hw_ppgtt *ppgtt = ctx->ppgtt;
+	struct page *page;
+	uint32_t *reg_state;
+	int ret;
+
+	if (!ppgtt)
+		ppgtt = dev_priv->mm.aliasing_ppgtt;
+
+	ret = i915_gem_object_set_to_cpu_domain(ctx_obj, true);
+	if (ret) {
+		DRM_DEBUG_DRIVER("Could not set to CPU domain\n");
+		return ret;
+	}
+
+	ret = i915_gem_object_get_pages(ctx_obj);
+	if (ret) {
+		DRM_DEBUG_DRIVER("Could not get object pages\n");
+		return ret;
+	}
+
+	i915_gem_object_pin_pages(ctx_obj);
+
+	/* The second page of the context object contains some fields which must
+	 * be set up prior to the first execution. */
+	page = i915_gem_object_get_page(ctx_obj, 1);
+	reg_state = kmap_atomic(page);
+
+	/* A context is actually a big batch buffer with several MI_LOAD_REGISTER_IMM
+	 * commands followed by (reg, value) pairs. The values we are setting here are
+	 * only for the first context restore: on a subsequent save, the GPU will
+	 * recreate this batchbuffer with new values (including all the missing
+	 * MI_LOAD_REGISTER_IMM commands that we are not initializing here). */
+	if (ring->id == RCS)
+		reg_state[CTX_LRI_HEADER_0] = MI_LOAD_REGISTER_IMM(14);
+	else
+		reg_state[CTX_LRI_HEADER_0] = MI_LOAD_REGISTER_IMM(11);
+	reg_state[CTX_LRI_HEADER_0] |= MI_LRI_FORCE_POSTED;
+	reg_state[CTX_CONTEXT_CONTROL] = RING_CONTEXT_CONTROL(ring);
+	reg_state[CTX_CONTEXT_CONTROL+1] =
+			_MASKED_BIT_ENABLE((1<<3) | MI_RESTORE_INHIBIT);
+	reg_state[CTX_RING_HEAD] = RING_HEAD(ring->mmio_base);
+	reg_state[CTX_RING_HEAD+1] = 0;
+	reg_state[CTX_RING_TAIL] = RING_TAIL(ring->mmio_base);
+	reg_state[CTX_RING_TAIL+1] = 0;
+	reg_state[CTX_RING_BUFFER_START] = RING_START(ring->mmio_base);
+	/* Ring buffer start address is not known until the buffer is pinned.
+	 * It is written to the context image in execlists_update_context()
+	 */
+	reg_state[CTX_RING_BUFFER_CONTROL] = RING_CTL(ring->mmio_base);
+	reg_state[CTX_RING_BUFFER_CONTROL+1] =
+			((ringbuf->size - PAGE_SIZE) & RING_NR_PAGES) | RING_VALID;
+	reg_state[CTX_BB_HEAD_U] = ring->mmio_base + 0x168;
+	reg_state[CTX_BB_HEAD_U+1] = 0;
+	reg_state[CTX_BB_HEAD_L] = ring->mmio_base + 0x140;
+	reg_state[CTX_BB_HEAD_L+1] = 0;
+	reg_state[CTX_BB_STATE] = ring->mmio_base + 0x110;
+	reg_state[CTX_BB_STATE+1] = (1<<5);
+	reg_state[CTX_SECOND_BB_HEAD_U] = ring->mmio_base + 0x11c;
+	reg_state[CTX_SECOND_BB_HEAD_U+1] = 0;
+	reg_state[CTX_SECOND_BB_HEAD_L] = ring->mmio_base + 0x114;
+	reg_state[CTX_SECOND_BB_HEAD_L+1] = 0;
+	reg_state[CTX_SECOND_BB_STATE] = ring->mmio_base + 0x118;
+	reg_state[CTX_SECOND_BB_STATE+1] = 0;
+	if (ring->id == RCS) {
+		/* TODO: according to BSpec, the register state context
+		 * for CHV does not have these. OTOH, these registers do
+		 * exist in CHV. I'm waiting for a clarification */
+		reg_state[CTX_BB_PER_CTX_PTR] = ring->mmio_base + 0x1c0;
+		reg_state[CTX_BB_PER_CTX_PTR+1] = 0;
+		reg_state[CTX_RCS_INDIRECT_CTX] = ring->mmio_base + 0x1c4;
+		reg_state[CTX_RCS_INDIRECT_CTX+1] = 0;
+		reg_state[CTX_RCS_INDIRECT_CTX_OFFSET] = ring->mmio_base + 0x1c8;
+		reg_state[CTX_RCS_INDIRECT_CTX_OFFSET+1] = 0;
+	}
+	reg_state[CTX_LRI_HEADER_1] = MI_LOAD_REGISTER_IMM(9);
+	reg_state[CTX_LRI_HEADER_1] |= MI_LRI_FORCE_POSTED;
+	reg_state[CTX_CTX_TIMESTAMP] = ring->mmio_base + 0x3a8;
+	reg_state[CTX_CTX_TIMESTAMP+1] = 0;
+	reg_state[CTX_PDP3_UDW] = GEN8_RING_PDP_UDW(ring, 3);
+	reg_state[CTX_PDP3_LDW] = GEN8_RING_PDP_LDW(ring, 3);
+	reg_state[CTX_PDP2_UDW] = GEN8_RING_PDP_UDW(ring, 2);
+	reg_state[CTX_PDP2_LDW] = GEN8_RING_PDP_LDW(ring, 2);
+	reg_state[CTX_PDP1_UDW] = GEN8_RING_PDP_UDW(ring, 1);
+	reg_state[CTX_PDP1_LDW] = GEN8_RING_PDP_LDW(ring, 1);
+	reg_state[CTX_PDP0_UDW] = GEN8_RING_PDP_UDW(ring, 0);
+	reg_state[CTX_PDP0_LDW] = GEN8_RING_PDP_LDW(ring, 0);
+	reg_state[CTX_PDP3_UDW+1] = upper_32_bits(ppgtt->pd_dma_addr[3]);
+	reg_state[CTX_PDP3_LDW+1] = lower_32_bits(ppgtt->pd_dma_addr[3]);
+	reg_state[CTX_PDP2_UDW+1] = upper_32_bits(ppgtt->pd_dma_addr[2]);
+	reg_state[CTX_PDP2_LDW+1] = lower_32_bits(ppgtt->pd_dma_addr[2]);
+	reg_state[CTX_PDP1_UDW+1] = upper_32_bits(ppgtt->pd_dma_addr[1]);
+	reg_state[CTX_PDP1_LDW+1] = lower_32_bits(ppgtt->pd_dma_addr[1]);
+	reg_state[CTX_PDP0_UDW+1] = upper_32_bits(ppgtt->pd_dma_addr[0]);
+	reg_state[CTX_PDP0_LDW+1] = lower_32_bits(ppgtt->pd_dma_addr[0]);
+	if (ring->id == RCS) {
+		reg_state[CTX_LRI_HEADER_2] = MI_LOAD_REGISTER_IMM(1);
+		reg_state[CTX_R_PWR_CLK_STATE] = 0x20c8;
+		reg_state[CTX_R_PWR_CLK_STATE+1] = 0;
+	}
+
+	kunmap_atomic(reg_state);
+
+	ctx_obj->dirty = 1;
+	set_page_dirty(page);
+	i915_gem_object_unpin_pages(ctx_obj);
+
+	return 0;
+}
+
+/**
+ * intel_lr_context_free() - free the LRC specific bits of a context
+ * @ctx: the LR context to free.
+ *
+ * The real context freeing is done in i915_gem_context_free: this only
+ * takes care of the bits that are LRC related: the per-engine backing
+ * objects and the logical ringbuffer.
+ */
+void intel_lr_context_free(struct intel_context *ctx)
+{
+	int i;
+
+	for (i = 0; i < I915_NUM_RINGS; i++) {
+		struct drm_i915_gem_object *ctx_obj = ctx->engine[i].state;
+
+		if (ctx_obj) {
+			struct intel_ringbuffer *ringbuf =
+					ctx->engine[i].ringbuf;
+			struct intel_engine_cs *ring = ringbuf->ring;
+
+			if (ctx == ring->default_context) {
+				intel_unpin_ringbuffer_obj(ringbuf);
+				i915_gem_object_ggtt_unpin(ctx_obj);
+			}
+			intel_destroy_ringbuffer_obj(ringbuf);
+			kfree(ringbuf);
+			drm_gem_object_unreference(&ctx_obj->base);
+		}
+	}
+}
+
+static uint32_t get_lr_context_size(struct intel_engine_cs *ring)
+{
+	int ret = 0;
+
+	WARN_ON(INTEL_INFO(ring->dev)->gen < 8);
+
+	switch (ring->id) {
+	case RCS:
+		if (INTEL_INFO(ring->dev)->gen >= 9)
+			ret = GEN9_LR_CONTEXT_RENDER_SIZE;
+		else
+			ret = GEN8_LR_CONTEXT_RENDER_SIZE;
+		break;
+	case VCS:
+	case BCS:
+	case VECS:
+	case VCS2:
+		ret = GEN8_LR_CONTEXT_OTHER_SIZE;
+		break;
+	}
+
+	return ret;
+}
+
+static void lrc_setup_hardware_status_page(struct intel_engine_cs *ring,
+		struct drm_i915_gem_object *default_ctx_obj)
+{
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+
+	/* The status page is offset 0 from the default context object
+	 * in LRC mode. */
+	ring->status_page.gfx_addr = i915_gem_obj_ggtt_offset(default_ctx_obj);
+	ring->status_page.page_addr =
+			kmap(sg_page(default_ctx_obj->pages->sgl));
+	ring->status_page.obj = default_ctx_obj;
+
+	I915_WRITE(RING_HWS_PGA(ring->mmio_base),
+			(u32)ring->status_page.gfx_addr);
+	POSTING_READ(RING_HWS_PGA(ring->mmio_base));
+}
+
+/**
+ * intel_lr_context_deferred_create() - create the LRC specific bits of a context
+ * @ctx: LR context to create.
+ * @ring: engine to be used with the context.
+ *
+ * This function can be called more than once, with different engines, if we plan
+ * to use the context with them. The context backing objects and the ringbuffers
+ * (specially the ringbuffer backing objects) suck a lot of memory up, and that's why
+ * the creation is a deferred call: it's better to make sure first that we need to use
+ * a given ring with the context.
+ *
+ * Return: non-zero on error.
+ */
+int intel_lr_context_deferred_create(struct intel_context *ctx,
+				     struct intel_engine_cs *ring)
+{
+	const bool is_global_default_ctx = (ctx == ring->default_context);
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_gem_object *ctx_obj;
+	uint32_t context_size;
+	struct intel_ringbuffer *ringbuf;
+	int ret;
+
+	WARN_ON(ctx->legacy_hw_ctx.rcs_state != NULL);
+	if (ctx->engine[ring->id].state)
+		return 0;
+
+	context_size = round_up(get_lr_context_size(ring), 4096);
+
+	ctx_obj = i915_gem_alloc_context_obj(dev, context_size);
+	if (IS_ERR(ctx_obj)) {
+		ret = PTR_ERR(ctx_obj);
+		DRM_DEBUG_DRIVER("Alloc LRC backing obj failed: %d\n", ret);
+		return ret;
+	}
+
+	if (is_global_default_ctx) {
+		ret = i915_gem_obj_ggtt_pin(ctx_obj, GEN8_LR_CONTEXT_ALIGN, 0);
+		if (ret) {
+			DRM_DEBUG_DRIVER("Pin LRC backing obj failed: %d\n",
+					ret);
+			drm_gem_object_unreference(&ctx_obj->base);
+			return ret;
+		}
+	}
+
+	ringbuf = kzalloc(sizeof(*ringbuf), GFP_KERNEL);
+	if (!ringbuf) {
+		DRM_DEBUG_DRIVER("Failed to allocate ringbuffer %s\n",
+				ring->name);
+		ret = -ENOMEM;
+		goto error_unpin_ctx;
+	}
+
+	ringbuf->ring = ring;
+	ringbuf->FIXME_lrc_ctx = ctx;
+
+	ringbuf->size = 32 * PAGE_SIZE;
+	ringbuf->effective_size = ringbuf->size;
+	ringbuf->head = 0;
+	ringbuf->tail = 0;
+	ringbuf->space = ringbuf->size;
+	ringbuf->last_retired_head = -1;
+
+	if (ringbuf->obj == NULL) {
+		ret = intel_alloc_ringbuffer_obj(dev, ringbuf);
+		if (ret) {
+			DRM_DEBUG_DRIVER(
+				"Failed to allocate ringbuffer obj %s: %d\n",
+				ring->name, ret);
+			goto error_free_rbuf;
+		}
+
+		if (is_global_default_ctx) {
+			ret = intel_pin_and_map_ringbuffer_obj(dev, ringbuf);
+			if (ret) {
+				DRM_ERROR(
+					"Failed to pin and map ringbuffer %s: %d\n",
+					ring->name, ret);
+				goto error_destroy_rbuf;
+			}
+		}
+
+	}
+
+	ret = populate_lr_context(ctx, ctx_obj, ring, ringbuf);
+	if (ret) {
+		DRM_DEBUG_DRIVER("Failed to populate LRC: %d\n", ret);
+		goto error;
+	}
+
+	ctx->engine[ring->id].ringbuf = ringbuf;
+	ctx->engine[ring->id].state = ctx_obj;
+
+	if (ctx == ring->default_context)
+		lrc_setup_hardware_status_page(ring, ctx_obj);
+
+	if (ring->id == RCS && !ctx->rcs_initialized) {
+		if (ring->init_context) {
+			ret = ring->init_context(ring, ctx);
+			if (ret)
+				DRM_ERROR("ring init context: %d\n", ret);
+		}
+
+		ret = intel_lr_context_render_state_init(ring, ctx);
+		if (ret) {
+			DRM_ERROR("Init render state failed: %d\n", ret);
+			ctx->engine[ring->id].ringbuf = NULL;
+			ctx->engine[ring->id].state = NULL;
+			goto error;
+		}
+		ctx->rcs_initialized = true;
+	}
+
+	return 0;
+
+error:
+	if (is_global_default_ctx)
+		intel_unpin_ringbuffer_obj(ringbuf);
+error_destroy_rbuf:
+	intel_destroy_ringbuffer_obj(ringbuf);
+error_free_rbuf:
+	kfree(ringbuf);
+error_unpin_ctx:
+	if (is_global_default_ctx)
+		i915_gem_object_ggtt_unpin(ctx_obj);
+	drm_gem_object_unreference(&ctx_obj->base);
+	return ret;
+}
diff --git a/drivers/gpu/drm/i915/intel_lrc.h b/drivers/gpu/drm/i915/intel_lrc.h
new file mode 100644
index 000000000000..14b216b9be7f
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_lrc.h
@@ -0,0 +1,118 @@
+/*
+ * Copyright © 2014 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#ifndef _INTEL_LRC_H_
+#define _INTEL_LRC_H_
+
+#define GEN8_LR_CONTEXT_ALIGN 4096
+
+/* Execlists regs */
+#define RING_ELSP(ring)			((ring)->mmio_base+0x230)
+#define RING_EXECLIST_STATUS(ring)	((ring)->mmio_base+0x234)
+#define RING_CONTEXT_CONTROL(ring)	((ring)->mmio_base+0x244)
+#define RING_CONTEXT_STATUS_BUF(ring)	((ring)->mmio_base+0x370)
+#define RING_CONTEXT_STATUS_PTR(ring)	((ring)->mmio_base+0x3a0)
+
+/* Logical Rings */
+void intel_logical_ring_stop(struct intel_engine_cs *ring);
+void intel_logical_ring_cleanup(struct intel_engine_cs *ring);
+int intel_logical_rings_init(struct drm_device *dev);
+
+int logical_ring_flush_all_caches(struct intel_ringbuffer *ringbuf);
+void intel_logical_ring_advance_and_submit(struct intel_ringbuffer *ringbuf);
+/**
+ * intel_logical_ring_advance() - advance the ringbuffer tail
+ * @ringbuf: Ringbuffer to advance.
+ *
+ * The tail is only updated in our logical ringbuffer struct.
+ */
+static inline void intel_logical_ring_advance(struct intel_ringbuffer *ringbuf)
+{
+	ringbuf->tail &= ringbuf->size - 1;
+}
+/**
+ * intel_logical_ring_emit() - write a DWORD to the ringbuffer.
+ * @ringbuf: Ringbuffer to write to.
+ * @data: DWORD to write.
+ */
+static inline void intel_logical_ring_emit(struct intel_ringbuffer *ringbuf,
+					   u32 data)
+{
+	iowrite32(data, ringbuf->virtual_start + ringbuf->tail);
+	ringbuf->tail += 4;
+}
+int intel_logical_ring_begin(struct intel_ringbuffer *ringbuf, int num_dwords);
+
+/* Logical Ring Contexts */
+int intel_lr_context_render_state_init(struct intel_engine_cs *ring,
+				       struct intel_context *ctx);
+void intel_lr_context_free(struct intel_context *ctx);
+int intel_lr_context_deferred_create(struct intel_context *ctx,
+				     struct intel_engine_cs *ring);
+void intel_lr_context_unpin(struct intel_engine_cs *ring,
+		struct intel_context *ctx);
+
+/* Execlists */
+int intel_sanitize_enable_execlists(struct drm_device *dev, int enable_execlists);
+int intel_execlists_submission(struct drm_device *dev, struct drm_file *file,
+			       struct intel_engine_cs *ring,
+			       struct intel_context *ctx,
+			       struct drm_i915_gem_execbuffer2 *args,
+			       struct list_head *vmas,
+			       struct drm_i915_gem_object *batch_obj,
+			       u64 exec_start, u32 flags);
+u32 intel_execlists_ctx_id(struct drm_i915_gem_object *ctx_obj);
+
+/**
+ * struct intel_ctx_submit_request - queued context submission request
+ * @ctx: Context to submit to the ELSP.
+ * @ring: Engine to submit it to.
+ * @tail: how far in the context's ringbuffer this request goes to.
+ * @execlist_link: link in the submission queue.
+ * @work: workqueue for processing this request in a bottom half.
+ * @elsp_submitted: no. of times this request has been sent to the ELSP.
+ *
+ * The ELSP only accepts two elements at a time, so we queue context/tail
+ * pairs on a given queue (ring->execlist_queue) until the hardware is
+ * available. The queue serves a double purpose: we also use it to keep track
+ * of the up to 2 contexts currently in the hardware (usually one in execution
+ * and the other queued up by the GPU): We only remove elements from the head
+ * of the queue when the hardware informs us that an element has been
+ * completed.
+ *
+ * All accesses to the queue are mediated by a spinlock (ring->execlist_lock).
+ */
+struct intel_ctx_submit_request {
+	struct intel_context *ctx;
+	struct intel_engine_cs *ring;
+	u32 tail;
+
+	struct list_head execlist_link;
+
+	int elsp_submitted;
+};
+
+void intel_execlists_handle_ctx_events(struct intel_engine_cs *ring);
+void intel_execlists_retire_requests(struct intel_engine_cs *ring);
+
+#endif /* _INTEL_LRC_H_ */
diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c
index fdf40267249c..14654d628ca4 100644
--- a/drivers/gpu/drm/i915/intel_lvds.c
+++ b/drivers/gpu/drm/i915/intel_lvds.c
@@ -76,7 +76,7 @@ static bool intel_lvds_get_hw_state(struct intel_encoder *encoder,
 	u32 tmp;
 
 	power_domain = intel_display_port_power_domain(encoder);
-	if (!intel_display_power_enabled(dev_priv, power_domain))
+	if (!intel_display_power_is_enabled(dev_priv, power_domain))
 		return false;
 
 	tmp = I915_READ(lvds_encoder->reg);
@@ -823,8 +823,7 @@ bool intel_is_dual_link_lvds(struct drm_device *dev)
 	struct intel_encoder *encoder;
 	struct intel_lvds_encoder *lvds_encoder;
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		if (encoder->type == INTEL_OUTPUT_LVDS) {
 			lvds_encoder = to_lvds_encoder(&encoder->base);
 
@@ -900,6 +899,17 @@ void intel_lvds_init(struct drm_device *dev)
 	int pipe;
 	u8 pin;
 
+	/*
+	 * Unlock registers and just leave them unlocked. Do this before
+	 * checking quirk lists to avoid bogus WARNINGs.
+	 */
+	if (HAS_PCH_SPLIT(dev)) {
+		I915_WRITE(PCH_PP_CONTROL,
+			   I915_READ(PCH_PP_CONTROL) | PANEL_UNLOCK_REGS);
+	} else {
+		I915_WRITE(PP_CONTROL,
+			   I915_READ(PP_CONTROL) | PANEL_UNLOCK_REGS);
+	}
 	if (!intel_lvds_supported(dev))
 		return;
 
@@ -1098,17 +1108,6 @@ out:
 	lvds_encoder->a3_power = I915_READ(lvds_encoder->reg) &
 				 LVDS_A3_POWER_MASK;
 
-	/*
-	 * Unlock registers and just
-	 * leave them unlocked
-	 */
-	if (HAS_PCH_SPLIT(dev)) {
-		I915_WRITE(PCH_PP_CONTROL,
-			   I915_READ(PCH_PP_CONTROL) | PANEL_UNLOCK_REGS);
-	} else {
-		I915_WRITE(PP_CONTROL,
-			   I915_READ(PP_CONTROL) | PANEL_UNLOCK_REGS);
-	}
 	lvds_connector->lid_notifier.notifier_call = intel_lid_notify;
 	if (acpi_lid_notifier_register(&lvds_connector->lid_notifier)) {
 		DRM_DEBUG_KMS("lid notifier registration failed\n");
@@ -1117,7 +1116,7 @@ out:
 	drm_connector_register(connector);
 
 	intel_panel_init(&intel_connector->panel, fixed_mode, downclock_mode);
-	intel_panel_setup_backlight(connector);
+	intel_panel_setup_backlight(connector, INVALID_PIPE);
 
 	return;
 
diff --git a/drivers/gpu/drm/i915/intel_panel.c b/drivers/gpu/drm/i915/intel_panel.c
index 8e374449c6b5..dfb783a8f2c3 100644
--- a/drivers/gpu/drm/i915/intel_panel.c
+++ b/drivers/gpu/drm/i915/intel_panel.c
@@ -419,9 +419,8 @@ static uint32_t scale(uint32_t source_val,
 	source_val = clamp(source_val, source_min, source_max);
 
 	/* avoid overflows */
-	target_val = (uint64_t)(source_val - source_min) *
-		(target_max - target_min);
-	do_div(target_val, source_max - source_min);
+	target_val = DIV_ROUND_CLOSEST_ULL((uint64_t)(source_val - source_min) *
+			(target_max - target_min), source_max - source_min);
 	target_val += target_min;
 
 	return target_val;
@@ -522,6 +521,9 @@ static u32 _vlv_get_backlight(struct drm_device *dev, enum pipe pipe)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
+	if (WARN_ON(pipe != PIPE_A && pipe != PIPE_B))
+		return 0;
+
 	return I915_READ(VLV_BLC_PWM_CTL(pipe)) & BACKLIGHT_DUTY_CYCLE_MASK;
 }
 
@@ -537,15 +539,17 @@ static u32 intel_panel_get_backlight(struct intel_connector *connector)
 {
 	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 val;
-	unsigned long flags;
+	struct intel_panel *panel = &connector->panel;
+	u32 val = 0;
 
-	spin_lock_irqsave(&dev_priv->backlight_lock, flags);
+	mutex_lock(&dev_priv->backlight_lock);
 
-	val = dev_priv->display.get_backlight(connector);
-	val = intel_panel_compute_brightness(connector, val);
+	if (panel->backlight.enabled) {
+		val = dev_priv->display.get_backlight(connector);
+		val = intel_panel_compute_brightness(connector, val);
+	}
 
-	spin_unlock_irqrestore(&dev_priv->backlight_lock, flags);
+	mutex_unlock(&dev_priv->backlight_lock);
 
 	DRM_DEBUG_DRIVER("get backlight PWM = %d\n", val);
 	return val;
@@ -604,6 +608,9 @@ static void vlv_set_backlight(struct intel_connector *connector, u32 level)
 	enum pipe pipe = intel_get_pipe_from_connector(connector);
 	u32 tmp;
 
+	if (WARN_ON(pipe != PIPE_A && pipe != PIPE_B))
+		return;
+
 	tmp = I915_READ(VLV_BLC_PWM_CTL(pipe)) & ~BACKLIGHT_DUTY_CYCLE_MASK;
 	I915_WRITE(VLV_BLC_PWM_CTL(pipe), tmp | level);
 }
@@ -627,14 +634,12 @@ static void intel_panel_set_backlight(struct intel_connector *connector,
 	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_panel *panel = &connector->panel;
-	enum pipe pipe = intel_get_pipe_from_connector(connector);
 	u32 hw_level;
-	unsigned long flags;
 
-	if (!panel->backlight.present || pipe == INVALID_PIPE)
+	if (!panel->backlight.present)
 		return;
 
-	spin_lock_irqsave(&dev_priv->backlight_lock, flags);
+	mutex_lock(&dev_priv->backlight_lock);
 
 	WARN_ON(panel->backlight.max == 0);
 
@@ -644,7 +649,7 @@ static void intel_panel_set_backlight(struct intel_connector *connector,
 	if (panel->backlight.enabled)
 		intel_panel_actually_set_backlight(connector, hw_level);
 
-	spin_unlock_irqrestore(&dev_priv->backlight_lock, flags);
+	mutex_unlock(&dev_priv->backlight_lock);
 }
 
 /* set backlight brightness to level in range [0..max], assuming hw min is
@@ -658,12 +663,17 @@ void intel_panel_set_backlight_acpi(struct intel_connector *connector,
 	struct intel_panel *panel = &connector->panel;
 	enum pipe pipe = intel_get_pipe_from_connector(connector);
 	u32 hw_level;
-	unsigned long flags;
 
+	/*
+	 * INVALID_PIPE may occur during driver init because
+	 * connection_mutex isn't held across the entire backlight
+	 * setup + modeset readout, and the BIOS can issue the
+	 * requests at any time.
+	 */
 	if (!panel->backlight.present || pipe == INVALID_PIPE)
 		return;
 
-	spin_lock_irqsave(&dev_priv->backlight_lock, flags);
+	mutex_lock(&dev_priv->backlight_lock);
 
 	WARN_ON(panel->backlight.max == 0);
 
@@ -679,7 +689,7 @@ void intel_panel_set_backlight_acpi(struct intel_connector *connector,
 	if (panel->backlight.enabled)
 		intel_panel_actually_set_backlight(connector, hw_level);
 
-	spin_unlock_irqrestore(&dev_priv->backlight_lock, flags);
+	mutex_unlock(&dev_priv->backlight_lock);
 }
 
 static void pch_disable_backlight(struct intel_connector *connector)
@@ -721,6 +731,9 @@ static void vlv_disable_backlight(struct intel_connector *connector)
 	enum pipe pipe = intel_get_pipe_from_connector(connector);
 	u32 tmp;
 
+	if (WARN_ON(pipe != PIPE_A && pipe != PIPE_B))
+		return;
+
 	intel_panel_actually_set_backlight(connector, 0);
 
 	tmp = I915_READ(VLV_BLC_PWM_CTL2(pipe));
@@ -732,10 +745,8 @@ void intel_panel_disable_backlight(struct intel_connector *connector)
 	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_panel *panel = &connector->panel;
-	enum pipe pipe = intel_get_pipe_from_connector(connector);
-	unsigned long flags;
 
-	if (!panel->backlight.present || pipe == INVALID_PIPE)
+	if (!panel->backlight.present)
 		return;
 
 	/*
@@ -749,12 +760,14 @@ void intel_panel_disable_backlight(struct intel_connector *connector)
 		return;
 	}
 
-	spin_lock_irqsave(&dev_priv->backlight_lock, flags);
+	mutex_lock(&dev_priv->backlight_lock);
 
+	if (panel->backlight.device)
+		panel->backlight.device->props.power = FB_BLANK_POWERDOWN;
 	panel->backlight.enabled = false;
 	dev_priv->display.disable_backlight(connector);
 
-	spin_unlock_irqrestore(&dev_priv->backlight_lock, flags);
+	mutex_unlock(&dev_priv->backlight_lock);
 }
 
 static void bdw_enable_backlight(struct intel_connector *connector)
@@ -778,8 +791,9 @@ static void bdw_enable_backlight(struct intel_connector *connector)
 	if (panel->backlight.active_low_pwm)
 		pch_ctl1 |= BLM_PCH_POLARITY;
 
-	/* BDW always uses the pch pwm controls. */
-	pch_ctl1 |= BLM_PCH_OVERRIDE_ENABLE;
+	/* After LPT, override is the default. */
+	if (HAS_PCH_LPT(dev_priv))
+		pch_ctl1 |= BLM_PCH_OVERRIDE_ENABLE;
 
 	I915_WRITE(BLC_PWM_PCH_CTL1, pch_ctl1);
 	POSTING_READ(BLC_PWM_PCH_CTL1);
@@ -908,6 +922,9 @@ static void vlv_enable_backlight(struct intel_connector *connector)
 	enum pipe pipe = intel_get_pipe_from_connector(connector);
 	u32 ctl, ctl2;
 
+	if (WARN_ON(pipe != PIPE_A && pipe != PIPE_B))
+		return;
+
 	ctl2 = I915_READ(VLV_BLC_PWM_CTL2(pipe));
 	if (ctl2 & BLM_PWM_ENABLE) {
 		DRM_DEBUG_KMS("backlight already enabled\n");
@@ -935,18 +952,17 @@ void intel_panel_enable_backlight(struct intel_connector *connector)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_panel *panel = &connector->panel;
 	enum pipe pipe = intel_get_pipe_from_connector(connector);
-	unsigned long flags;
 
-	if (!panel->backlight.present || pipe == INVALID_PIPE)
+	if (!panel->backlight.present)
 		return;
 
 	DRM_DEBUG_KMS("pipe %c\n", pipe_name(pipe));
 
-	spin_lock_irqsave(&dev_priv->backlight_lock, flags);
+	mutex_lock(&dev_priv->backlight_lock);
 
 	WARN_ON(panel->backlight.max == 0);
 
-	if (panel->backlight.level == 0) {
+	if (panel->backlight.level <= panel->backlight.min) {
 		panel->backlight.level = panel->backlight.max;
 		if (panel->backlight.device)
 			panel->backlight.device->props.brightness =
@@ -957,14 +973,17 @@ void intel_panel_enable_backlight(struct intel_connector *connector)
 
 	dev_priv->display.enable_backlight(connector);
 	panel->backlight.enabled = true;
+	if (panel->backlight.device)
+		panel->backlight.device->props.power = FB_BLANK_UNBLANK;
 
-	spin_unlock_irqrestore(&dev_priv->backlight_lock, flags);
+	mutex_unlock(&dev_priv->backlight_lock);
 }
 
 #if IS_ENABLED(CONFIG_BACKLIGHT_CLASS_DEVICE)
 static int intel_backlight_device_update_status(struct backlight_device *bd)
 {
 	struct intel_connector *connector = bl_get_data(bd);
+	struct intel_panel *panel = &connector->panel;
 	struct drm_device *dev = connector->base.dev;
 
 	drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
@@ -972,6 +991,23 @@ static int intel_backlight_device_update_status(struct backlight_device *bd)
 		      bd->props.brightness, bd->props.max_brightness);
 	intel_panel_set_backlight(connector, bd->props.brightness,
 				  bd->props.max_brightness);
+
+	/*
+	 * Allow flipping bl_power as a sub-state of enabled. Sadly the
+	 * backlight class device does not make it easy to to differentiate
+	 * between callbacks for brightness and bl_power, so our backlight_power
+	 * callback needs to take this into account.
+	 */
+	if (panel->backlight.enabled) {
+		if (panel->backlight_power) {
+			bool enable = bd->props.power == FB_BLANK_UNBLANK &&
+				bd->props.brightness != 0;
+			panel->backlight_power(connector, enable);
+		}
+	} else {
+		bd->props.power = FB_BLANK_POWERDOWN;
+	}
+
 	drm_modeset_unlock(&dev->mode_config.connection_mutex);
 	return 0;
 }
@@ -1009,6 +1045,9 @@ static int intel_backlight_device_register(struct intel_connector *connector)
 	if (WARN_ON(panel->backlight.device))
 		return -ENODEV;
 
+	if (!panel->backlight.present)
+		return 0;
+
 	WARN_ON(panel->backlight.max == 0);
 
 	memset(&props, 0, sizeof(props));
@@ -1023,6 +1062,11 @@ static int intel_backlight_device_register(struct intel_connector *connector)
 					    panel->backlight.level,
 					    props.max_brightness);
 
+	if (panel->backlight.enabled)
+		props.power = FB_BLANK_UNBLANK;
+	else
+		props.power = FB_BLANK_POWERDOWN;
+
 	/*
 	 * Note: using the same name independent of the connector prevents
 	 * registration of multiple backlight devices in the driver.
@@ -1039,6 +1083,10 @@ static int intel_backlight_device_register(struct intel_connector *connector)
 		panel->backlight.device = NULL;
 		return -ENODEV;
 	}
+
+	DRM_DEBUG_KMS("Connector %s backlight sysfs interface registered\n",
+		      connector->base.name);
+
 	return 0;
 }
 
@@ -1072,15 +1120,28 @@ static u32 get_backlight_min_vbt(struct intel_connector *connector)
 	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_panel *panel = &connector->panel;
+	int min;
 
 	WARN_ON(panel->backlight.max == 0);
 
+	/*
+	 * XXX: If the vbt value is 255, it makes min equal to max, which leads
+	 * to problems. There are such machines out there. Either our
+	 * interpretation is wrong or the vbt has bogus data. Or both. Safeguard
+	 * against this by letting the minimum be at most (arbitrarily chosen)
+	 * 25% of the max.
+	 */
+	min = clamp_t(int, dev_priv->vbt.backlight.min_brightness, 0, 64);
+	if (min != dev_priv->vbt.backlight.min_brightness) {
+		DRM_DEBUG_KMS("clamping VBT min backlight %d/255 to %d/255\n",
+			      dev_priv->vbt.backlight.min_brightness, min);
+	}
+
 	/* vbt value is a coefficient in range [0..255] */
-	return scale(dev_priv->vbt.backlight.min_brightness, 0, 255,
-		     0, panel->backlight.max);
+	return scale(min, 0, 255, 0, panel->backlight.max);
 }
 
-static int bdw_setup_backlight(struct intel_connector *connector)
+static int bdw_setup_backlight(struct intel_connector *connector, enum pipe unused)
 {
 	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1106,7 +1167,7 @@ static int bdw_setup_backlight(struct intel_connector *connector)
 	return 0;
 }
 
-static int pch_setup_backlight(struct intel_connector *connector)
+static int pch_setup_backlight(struct intel_connector *connector, enum pipe unused)
 {
 	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1133,7 +1194,7 @@ static int pch_setup_backlight(struct intel_connector *connector)
 	return 0;
 }
 
-static int i9xx_setup_backlight(struct intel_connector *connector)
+static int i9xx_setup_backlight(struct intel_connector *connector, enum pipe unused)
 {
 	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1165,7 +1226,7 @@ static int i9xx_setup_backlight(struct intel_connector *connector)
 	return 0;
 }
 
-static int i965_setup_backlight(struct intel_connector *connector)
+static int i965_setup_backlight(struct intel_connector *connector, enum pipe unused)
 {
 	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1195,37 +1256,40 @@ static int i965_setup_backlight(struct intel_connector *connector)
 	return 0;
 }
 
-static int vlv_setup_backlight(struct intel_connector *connector)
+static int vlv_setup_backlight(struct intel_connector *connector, enum pipe pipe)
 {
 	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_panel *panel = &connector->panel;
-	enum pipe pipe;
+	enum pipe p;
 	u32 ctl, ctl2, val;
 
-	for_each_pipe(pipe) {
-		u32 cur_val = I915_READ(VLV_BLC_PWM_CTL(pipe));
+	for_each_pipe(dev_priv, p) {
+		u32 cur_val = I915_READ(VLV_BLC_PWM_CTL(p));
 
 		/* Skip if the modulation freq is already set */
 		if (cur_val & ~BACKLIGHT_DUTY_CYCLE_MASK)
 			continue;
 
 		cur_val &= BACKLIGHT_DUTY_CYCLE_MASK;
-		I915_WRITE(VLV_BLC_PWM_CTL(pipe), (0xf42 << 16) |
+		I915_WRITE(VLV_BLC_PWM_CTL(p), (0xf42 << 16) |
 			   cur_val);
 	}
 
-	ctl2 = I915_READ(VLV_BLC_PWM_CTL2(PIPE_A));
+	if (WARN_ON(pipe != PIPE_A && pipe != PIPE_B))
+		return -ENODEV;
+
+	ctl2 = I915_READ(VLV_BLC_PWM_CTL2(pipe));
 	panel->backlight.active_low_pwm = ctl2 & BLM_POLARITY_I965;
 
-	ctl = I915_READ(VLV_BLC_PWM_CTL(PIPE_A));
+	ctl = I915_READ(VLV_BLC_PWM_CTL(pipe));
 	panel->backlight.max = ctl >> 16;
 	if (!panel->backlight.max)
 		return -ENODEV;
 
 	panel->backlight.min = get_backlight_min_vbt(connector);
 
-	val = _vlv_get_backlight(dev, PIPE_A);
+	val = _vlv_get_backlight(dev, pipe);
 	panel->backlight.level = intel_panel_compute_brightness(connector, val);
 
 	panel->backlight.enabled = (ctl2 & BLM_PWM_ENABLE) &&
@@ -1234,13 +1298,12 @@ static int vlv_setup_backlight(struct intel_connector *connector)
 	return 0;
 }
 
-int intel_panel_setup_backlight(struct drm_connector *connector)
+int intel_panel_setup_backlight(struct drm_connector *connector, enum pipe pipe)
 {
 	struct drm_device *dev = connector->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_connector *intel_connector = to_intel_connector(connector);
 	struct intel_panel *panel = &intel_connector->panel;
-	unsigned long flags;
 	int ret;
 
 	if (!dev_priv->vbt.backlight.present) {
@@ -1253,9 +1316,9 @@ int intel_panel_setup_backlight(struct drm_connector *connector)
 	}
 
 	/* set level and max in panel struct */
-	spin_lock_irqsave(&dev_priv->backlight_lock, flags);
-	ret = dev_priv->display.setup_backlight(intel_connector);
-	spin_unlock_irqrestore(&dev_priv->backlight_lock, flags);
+	mutex_lock(&dev_priv->backlight_lock);
+	ret = dev_priv->display.setup_backlight(intel_connector, pipe);
+	mutex_unlock(&dev_priv->backlight_lock);
 
 	if (ret) {
 		DRM_DEBUG_KMS("failed to setup backlight for connector %s\n",
@@ -1263,15 +1326,12 @@ int intel_panel_setup_backlight(struct drm_connector *connector)
 		return ret;
 	}
 
-	intel_backlight_device_register(intel_connector);
-
 	panel->backlight.present = true;
 
-	DRM_DEBUG_KMS("backlight initialized, %s, brightness %u/%u, "
-		      "sysfs interface %sregistered\n",
+	DRM_DEBUG_KMS("Connector %s backlight initialized, %s, brightness %u/%u\n",
+		      connector->name,
 		      panel->backlight.enabled ? "enabled" : "disabled",
-		      panel->backlight.level, panel->backlight.max,
-		      panel->backlight.device ? "" : "not ");
+		      panel->backlight.level, panel->backlight.max);
 
 	return 0;
 }
@@ -1282,7 +1342,6 @@ void intel_panel_destroy_backlight(struct drm_connector *connector)
 	struct intel_panel *panel = &intel_connector->panel;
 
 	panel->backlight.present = false;
-	intel_backlight_device_unregister(intel_connector);
 }
 
 /* Set up chip specific backlight functions */
@@ -1290,7 +1349,7 @@ void intel_panel_init_backlight_funcs(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (IS_BROADWELL(dev)) {
+	if (IS_BROADWELL(dev) || (INTEL_INFO(dev)->gen >= 9)) {
 		dev_priv->display.setup_backlight = bdw_setup_backlight;
 		dev_priv->display.enable_backlight = bdw_enable_backlight;
 		dev_priv->display.disable_backlight = pch_disable_backlight;
@@ -1345,3 +1404,19 @@ void intel_panel_fini(struct intel_panel *panel)
 		drm_mode_destroy(intel_connector->base.dev,
 				panel->downclock_mode);
 }
+
+void intel_backlight_register(struct drm_device *dev)
+{
+	struct intel_connector *connector;
+
+	list_for_each_entry(connector, &dev->mode_config.connector_list, base.head)
+		intel_backlight_device_register(connector);
+}
+
+void intel_backlight_unregister(struct drm_device *dev)
+{
+	struct intel_connector *connector;
+
+	list_for_each_entry(connector, &dev->mode_config.connector_list, base.head)
+		intel_backlight_device_unregister(connector);
+}
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index 40c12295c0bd..bf814a64582a 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -30,9 +30,6 @@
 #include "intel_drv.h"
 #include "../../../platform/x86/intel_ips.h"
 #include <linux/module.h>
-#include <linux/vgaarb.h>
-#include <drm/i915_powerwell.h>
-#include <linux/pm_runtime.h>
 
 /**
  * RC6 is a special power stage which allows the GPU to enter an very
@@ -66,11 +63,37 @@
  * i915.i915_enable_fbc parameter
  */
 
+static void gen9_init_clock_gating(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	/*
+	 * WaDisableSDEUnitClockGating:skl
+	 * This seems to be a pre-production w/a.
+	 */
+	I915_WRITE(GEN8_UCGCTL6, I915_READ(GEN8_UCGCTL6) |
+		   GEN8_SDEUNIT_CLOCK_GATE_DISABLE);
+
+	/*
+	 * WaDisableDgMirrorFixInHalfSliceChicken5:skl
+	 * This is a pre-production w/a.
+	 */
+	I915_WRITE(GEN9_HALF_SLICE_CHICKEN5,
+		   I915_READ(GEN9_HALF_SLICE_CHICKEN5) &
+		   ~GEN9_DG_MIRROR_FIX_ENABLE);
+
+	/* Wa4x4STCOptimizationDisable:skl */
+	I915_WRITE(CACHE_MODE_1,
+		   _MASKED_BIT_ENABLE(GEN8_4x4_STC_OPTIMIZATION_DISABLE));
+}
+
 static void i8xx_disable_fbc(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 fbc_ctl;
 
+	dev_priv->fbc.enabled = false;
+
 	/* Disable compression */
 	fbc_ctl = I915_READ(FBC_CONTROL);
 	if ((fbc_ctl & FBC_CTL_EN) == 0)
@@ -99,6 +122,8 @@ static void i8xx_enable_fbc(struct drm_crtc *crtc)
 	int i;
 	u32 fbc_ctl;
 
+	dev_priv->fbc.enabled = true;
+
 	cfb_pitch = dev_priv->fbc.size / FBC_LL_SIZE;
 	if (fb->pitches[0] < cfb_pitch)
 		cfb_pitch = fb->pitches[0];
@@ -153,6 +178,8 @@ static void g4x_enable_fbc(struct drm_crtc *crtc)
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	u32 dpfc_ctl;
 
+	dev_priv->fbc.enabled = true;
+
 	dpfc_ctl = DPFC_CTL_PLANE(intel_crtc->plane) | DPFC_SR_EN;
 	if (drm_format_plane_cpp(fb->pixel_format, 0) == 2)
 		dpfc_ctl |= DPFC_CTL_LIMIT_2X;
@@ -173,6 +200,8 @@ static void g4x_disable_fbc(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 dpfc_ctl;
 
+	dev_priv->fbc.enabled = false;
+
 	/* Disable compression */
 	dpfc_ctl = I915_READ(DPFC_CONTROL);
 	if (dpfc_ctl & DPFC_CTL_EN) {
@@ -224,6 +253,8 @@ static void ironlake_enable_fbc(struct drm_crtc *crtc)
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	u32 dpfc_ctl;
 
+	dev_priv->fbc.enabled = true;
+
 	dpfc_ctl = DPFC_CTL_PLANE(intel_crtc->plane);
 	if (drm_format_plane_cpp(fb->pixel_format, 0) == 2)
 		dev_priv->fbc.threshold++;
@@ -264,6 +295,8 @@ static void ironlake_disable_fbc(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 dpfc_ctl;
 
+	dev_priv->fbc.enabled = false;
+
 	/* Disable compression */
 	dpfc_ctl = I915_READ(ILK_DPFC_CONTROL);
 	if (dpfc_ctl & DPFC_CTL_EN) {
@@ -290,6 +323,8 @@ static void gen7_enable_fbc(struct drm_crtc *crtc)
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	u32 dpfc_ctl;
 
+	dev_priv->fbc.enabled = true;
+
 	dpfc_ctl = IVB_DPFC_CTL_PLANE(intel_crtc->plane);
 	if (drm_format_plane_cpp(fb->pixel_format, 0) == 2)
 		dev_priv->fbc.threshold++;
@@ -309,6 +344,9 @@ static void gen7_enable_fbc(struct drm_crtc *crtc)
 
 	dpfc_ctl |= IVB_DPFC_CTL_FENCE_EN;
 
+	if (dev_priv->fbc.false_color)
+		dpfc_ctl |= FBC_CTL_FALSE_COLOR;
+
 	I915_WRITE(ILK_DPFC_CONTROL, dpfc_ctl | DPFC_CTL_EN);
 
 	if (IS_IVYBRIDGE(dev)) {
@@ -336,10 +374,20 @@ bool intel_fbc_enabled(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (!dev_priv->display.fbc_enabled)
-		return false;
+	return dev_priv->fbc.enabled;
+}
+
+void bdw_fbc_sw_flush(struct drm_device *dev, u32 value)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (!IS_GEN8(dev))
+		return;
+
+	if (!intel_fbc_enabled(dev))
+		return;
 
-	return dev_priv->display.fbc_enabled(dev);
+	I915_WRITE(MSG_FBC_REND_STATE, value);
 }
 
 static void intel_fbc_work_fn(struct work_struct *__work)
@@ -578,6 +626,12 @@ void intel_update_fbc(struct drm_device *dev)
 			DRM_DEBUG_KMS("framebuffer not tiled or fenced, disabling compression\n");
 		goto out_disable;
 	}
+	if (INTEL_INFO(dev)->gen <= 4 && !IS_G4X(dev) &&
+	    to_intel_plane(crtc->primary)->rotation != BIT(DRM_ROTATE_0)) {
+		if (set_no_fbc_reason(dev_priv, FBC_UNSUPPORTED_MODE))
+			DRM_DEBUG_KMS("Rotation unsupported, disabling\n");
+		goto out_disable;
+	}
 
 	/* If the kernel debugger is active, always disable compression */
 	if (in_dbg_master())
@@ -853,7 +907,7 @@ void intel_set_memory_cxsr(struct drm_i915_private *dev_priv, bool enable)
  * A value of 5us seems to be a good balance; safe for very low end
  * platforms but not overly aggressive on lower latency configs.
  */
-static const int latency_ns = 5000;
+static const int pessimal_latency_ns = 5000;
 
 static int i9xx_get_fifo_size(struct drm_device *dev, int plane)
 {
@@ -982,13 +1036,20 @@ static const struct intel_watermark_params i915_wm_info = {
 	.guard_size = 2,
 	.cacheline_size = I915_FIFO_LINE_SIZE,
 };
-static const struct intel_watermark_params i830_wm_info = {
+static const struct intel_watermark_params i830_a_wm_info = {
 	.fifo_size = I855GM_FIFO_SIZE,
 	.max_wm = I915_MAX_WM,
 	.default_wm = 1,
 	.guard_size = 2,
 	.cacheline_size = I830_FIFO_LINE_SIZE,
 };
+static const struct intel_watermark_params i830_bc_wm_info = {
+	.fifo_size = I855GM_FIFO_SIZE,
+	.max_wm = I915_MAX_WM/2,
+	.default_wm = 1,
+	.guard_size = 2,
+	.cacheline_size = I830_FIFO_LINE_SIZE,
+};
 static const struct intel_watermark_params i845_wm_info = {
 	.fifo_size = I830_FIFO_SIZE,
 	.max_wm = I915_MAX_WM,
@@ -1044,6 +1105,17 @@ static unsigned long intel_calculate_wm(unsigned long clock_in_khz,
 		wm_size = wm->max_wm;
 	if (wm_size <= 0)
 		wm_size = wm->default_wm;
+
+	/*
+	 * Bspec seems to indicate that the value shouldn't be lower than
+	 * 'burst size + 1'. Certainly 830 is quite unhappy with low values.
+	 * Lets go for 8 which is the burst size since certain platforms
+	 * already use a hardcoded 8 (which is what the spec says should be
+	 * done).
+	 */
+	if (wm_size <= 8)
+		wm_size = 8;
+
 	return wm_size;
 }
 
@@ -1268,33 +1340,32 @@ static bool g4x_compute_srwm(struct drm_device *dev,
 			      display, cursor);
 }
 
-static bool vlv_compute_drain_latency(struct drm_device *dev,
-				     int plane,
-				     int *plane_prec_mult,
-				     int *plane_dl,
-				     int *cursor_prec_mult,
-				     int *cursor_dl)
+static bool vlv_compute_drain_latency(struct drm_crtc *crtc,
+				      int pixel_size,
+				      int *prec_mult,
+				      int *drain_latency)
 {
-	struct drm_crtc *crtc;
-	int clock, pixel_size;
+	struct drm_device *dev = crtc->dev;
 	int entries;
+	int clock = to_intel_crtc(crtc)->config.adjusted_mode.crtc_clock;
 
-	crtc = intel_get_crtc_for_plane(dev, plane);
-	if (!intel_crtc_active(crtc))
+	if (WARN(clock == 0, "Pixel clock is zero!\n"))
 		return false;
 
-	clock = to_intel_crtc(crtc)->config.adjusted_mode.crtc_clock;
-	pixel_size = crtc->primary->fb->bits_per_pixel / 8;	/* BPP */
+	if (WARN(pixel_size == 0, "Pixel size is zero!\n"))
+		return false;
 
-	entries = (clock / 1000) * pixel_size;
-	*plane_prec_mult = (entries > 128) ?
-		DRAIN_LATENCY_PRECISION_64 : DRAIN_LATENCY_PRECISION_32;
-	*plane_dl = (64 * (*plane_prec_mult) * 4) / entries;
+	entries = DIV_ROUND_UP(clock, 1000) * pixel_size;
+	if (IS_CHERRYVIEW(dev))
+		*prec_mult = (entries > 128) ? DRAIN_LATENCY_PRECISION_32 :
+					       DRAIN_LATENCY_PRECISION_16;
+	else
+		*prec_mult = (entries > 128) ? DRAIN_LATENCY_PRECISION_64 :
+					       DRAIN_LATENCY_PRECISION_32;
+	*drain_latency = (64 * (*prec_mult) * 4) / entries;
 
-	entries = (clock / 1000) * 4;	/* BPP is always 4 for cursor */
-	*cursor_prec_mult = (entries > 128) ?
-		DRAIN_LATENCY_PRECISION_64 : DRAIN_LATENCY_PRECISION_32;
-	*cursor_dl = (64 * (*cursor_prec_mult) * 4) / entries;
+	if (*drain_latency > DRAIN_LATENCY_MASK)
+		*drain_latency = DRAIN_LATENCY_MASK;
 
 	return true;
 }
@@ -1307,39 +1378,51 @@ static bool vlv_compute_drain_latency(struct drm_device *dev,
  * latency value.
  */
 
-static void vlv_update_drain_latency(struct drm_device *dev)
+static void vlv_update_drain_latency(struct drm_crtc *crtc)
 {
+	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	int planea_prec, planea_dl, planeb_prec, planeb_dl;
-	int cursora_prec, cursora_dl, cursorb_prec, cursorb_dl;
-	int plane_prec_mult, cursor_prec_mult; /* Precision multiplier is
-							either 16 or 32 */
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	int pixel_size;
+	int drain_latency;
+	enum pipe pipe = intel_crtc->pipe;
+	int plane_prec, prec_mult, plane_dl;
+	const int high_precision = IS_CHERRYVIEW(dev) ?
+		DRAIN_LATENCY_PRECISION_32 : DRAIN_LATENCY_PRECISION_64;
 
-	/* For plane A, Cursor A */
-	if (vlv_compute_drain_latency(dev, 0, &plane_prec_mult, &planea_dl,
-				      &cursor_prec_mult, &cursora_dl)) {
-		cursora_prec = (cursor_prec_mult == DRAIN_LATENCY_PRECISION_32) ?
-			DDL_CURSORA_PRECISION_32 : DDL_CURSORA_PRECISION_64;
-		planea_prec = (plane_prec_mult == DRAIN_LATENCY_PRECISION_32) ?
-			DDL_PLANEA_PRECISION_32 : DDL_PLANEA_PRECISION_64;
+	plane_dl = I915_READ(VLV_DDL(pipe)) & ~(DDL_PLANE_PRECISION_HIGH |
+		   DRAIN_LATENCY_MASK | DDL_CURSOR_PRECISION_HIGH |
+		   (DRAIN_LATENCY_MASK << DDL_CURSOR_SHIFT));
 
-		I915_WRITE(VLV_DDL1, cursora_prec |
-				(cursora_dl << DDL_CURSORA_SHIFT) |
-				planea_prec | planea_dl);
+	if (!intel_crtc_active(crtc)) {
+		I915_WRITE(VLV_DDL(pipe), plane_dl);
+		return;
 	}
 
-	/* For plane B, Cursor B */
-	if (vlv_compute_drain_latency(dev, 1, &plane_prec_mult, &planeb_dl,
-				      &cursor_prec_mult, &cursorb_dl)) {
-		cursorb_prec = (cursor_prec_mult == DRAIN_LATENCY_PRECISION_32) ?
-			DDL_CURSORB_PRECISION_32 : DDL_CURSORB_PRECISION_64;
-		planeb_prec = (plane_prec_mult == DRAIN_LATENCY_PRECISION_32) ?
-			DDL_PLANEB_PRECISION_32 : DDL_PLANEB_PRECISION_64;
+	/* Primary plane Drain Latency */
+	pixel_size = crtc->primary->fb->bits_per_pixel / 8;	/* BPP */
+	if (vlv_compute_drain_latency(crtc, pixel_size, &prec_mult, &drain_latency)) {
+		plane_prec = (prec_mult == high_precision) ?
+					   DDL_PLANE_PRECISION_HIGH :
+					   DDL_PLANE_PRECISION_LOW;
+		plane_dl |= plane_prec | drain_latency;
+	}
 
-		I915_WRITE(VLV_DDL2, cursorb_prec |
-				(cursorb_dl << DDL_CURSORB_SHIFT) |
-				planeb_prec | planeb_dl);
+	/* Cursor Drain Latency
+	 * BPP is always 4 for cursor
+	 */
+	pixel_size = 4;
+
+	/* Program cursor DL only if it is enabled */
+	if (intel_crtc->cursor_base &&
+	    vlv_compute_drain_latency(crtc, pixel_size, &prec_mult, &drain_latency)) {
+		plane_prec = (prec_mult == high_precision) ?
+					   DDL_CURSOR_PRECISION_HIGH :
+					   DDL_CURSOR_PRECISION_LOW;
+		plane_dl |= plane_prec | (drain_latency << DDL_CURSOR_SHIFT);
 	}
+
+	I915_WRITE(VLV_DDL(pipe), plane_dl);
 }
 
 #define single_plane_enabled(mask) is_power_of_2(mask)
@@ -1355,20 +1438,92 @@ static void valleyview_update_wm(struct drm_crtc *crtc)
 	unsigned int enabled = 0;
 	bool cxsr_enabled;
 
-	vlv_update_drain_latency(dev);
+	vlv_update_drain_latency(crtc);
+
+	if (g4x_compute_wm0(dev, PIPE_A,
+			    &valleyview_wm_info, pessimal_latency_ns,
+			    &valleyview_cursor_wm_info, pessimal_latency_ns,
+			    &planea_wm, &cursora_wm))
+		enabled |= 1 << PIPE_A;
+
+	if (g4x_compute_wm0(dev, PIPE_B,
+			    &valleyview_wm_info, pessimal_latency_ns,
+			    &valleyview_cursor_wm_info, pessimal_latency_ns,
+			    &planeb_wm, &cursorb_wm))
+		enabled |= 1 << PIPE_B;
+
+	if (single_plane_enabled(enabled) &&
+	    g4x_compute_srwm(dev, ffs(enabled) - 1,
+			     sr_latency_ns,
+			     &valleyview_wm_info,
+			     &valleyview_cursor_wm_info,
+			     &plane_sr, &ignore_cursor_sr) &&
+	    g4x_compute_srwm(dev, ffs(enabled) - 1,
+			     2*sr_latency_ns,
+			     &valleyview_wm_info,
+			     &valleyview_cursor_wm_info,
+			     &ignore_plane_sr, &cursor_sr)) {
+		cxsr_enabled = true;
+	} else {
+		cxsr_enabled = false;
+		intel_set_memory_cxsr(dev_priv, false);
+		plane_sr = cursor_sr = 0;
+	}
+
+	DRM_DEBUG_KMS("Setting FIFO watermarks - A: plane=%d, cursor=%d, "
+		      "B: plane=%d, cursor=%d, SR: plane=%d, cursor=%d\n",
+		      planea_wm, cursora_wm,
+		      planeb_wm, cursorb_wm,
+		      plane_sr, cursor_sr);
+
+	I915_WRITE(DSPFW1,
+		   (plane_sr << DSPFW_SR_SHIFT) |
+		   (cursorb_wm << DSPFW_CURSORB_SHIFT) |
+		   (planeb_wm << DSPFW_PLANEB_SHIFT) |
+		   (planea_wm << DSPFW_PLANEA_SHIFT));
+	I915_WRITE(DSPFW2,
+		   (I915_READ(DSPFW2) & ~DSPFW_CURSORA_MASK) |
+		   (cursora_wm << DSPFW_CURSORA_SHIFT));
+	I915_WRITE(DSPFW3,
+		   (I915_READ(DSPFW3) & ~DSPFW_CURSOR_SR_MASK) |
+		   (cursor_sr << DSPFW_CURSOR_SR_SHIFT));
+
+	if (cxsr_enabled)
+		intel_set_memory_cxsr(dev_priv, true);
+}
+
+static void cherryview_update_wm(struct drm_crtc *crtc)
+{
+	struct drm_device *dev = crtc->dev;
+	static const int sr_latency_ns = 12000;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int planea_wm, planeb_wm, planec_wm;
+	int cursora_wm, cursorb_wm, cursorc_wm;
+	int plane_sr, cursor_sr;
+	int ignore_plane_sr, ignore_cursor_sr;
+	unsigned int enabled = 0;
+	bool cxsr_enabled;
+
+	vlv_update_drain_latency(crtc);
 
 	if (g4x_compute_wm0(dev, PIPE_A,
-			    &valleyview_wm_info, latency_ns,
-			    &valleyview_cursor_wm_info, latency_ns,
+			    &valleyview_wm_info, pessimal_latency_ns,
+			    &valleyview_cursor_wm_info, pessimal_latency_ns,
 			    &planea_wm, &cursora_wm))
 		enabled |= 1 << PIPE_A;
 
 	if (g4x_compute_wm0(dev, PIPE_B,
-			    &valleyview_wm_info, latency_ns,
-			    &valleyview_cursor_wm_info, latency_ns,
+			    &valleyview_wm_info, pessimal_latency_ns,
+			    &valleyview_cursor_wm_info, pessimal_latency_ns,
 			    &planeb_wm, &cursorb_wm))
 		enabled |= 1 << PIPE_B;
 
+	if (g4x_compute_wm0(dev, PIPE_C,
+			    &valleyview_wm_info, pessimal_latency_ns,
+			    &valleyview_cursor_wm_info, pessimal_latency_ns,
+			    &planec_wm, &cursorc_wm))
+		enabled |= 1 << PIPE_C;
+
 	if (single_plane_enabled(enabled) &&
 	    g4x_compute_srwm(dev, ffs(enabled) - 1,
 			     sr_latency_ns,
@@ -1387,27 +1542,68 @@ static void valleyview_update_wm(struct drm_crtc *crtc)
 		plane_sr = cursor_sr = 0;
 	}
 
-	DRM_DEBUG_KMS("Setting FIFO watermarks - A: plane=%d, cursor=%d, B: plane=%d, cursor=%d, SR: plane=%d, cursor=%d\n",
+	DRM_DEBUG_KMS("Setting FIFO watermarks - A: plane=%d, cursor=%d, "
+		      "B: plane=%d, cursor=%d, C: plane=%d, cursor=%d, "
+		      "SR: plane=%d, cursor=%d\n",
 		      planea_wm, cursora_wm,
 		      planeb_wm, cursorb_wm,
+		      planec_wm, cursorc_wm,
 		      plane_sr, cursor_sr);
 
 	I915_WRITE(DSPFW1,
 		   (plane_sr << DSPFW_SR_SHIFT) |
 		   (cursorb_wm << DSPFW_CURSORB_SHIFT) |
 		   (planeb_wm << DSPFW_PLANEB_SHIFT) |
-		   planea_wm);
+		   (planea_wm << DSPFW_PLANEA_SHIFT));
 	I915_WRITE(DSPFW2,
 		   (I915_READ(DSPFW2) & ~DSPFW_CURSORA_MASK) |
 		   (cursora_wm << DSPFW_CURSORA_SHIFT));
 	I915_WRITE(DSPFW3,
 		   (I915_READ(DSPFW3) & ~DSPFW_CURSOR_SR_MASK) |
 		   (cursor_sr << DSPFW_CURSOR_SR_SHIFT));
+	I915_WRITE(DSPFW9_CHV,
+		   (I915_READ(DSPFW9_CHV) & ~(DSPFW_PLANEC_MASK |
+					      DSPFW_CURSORC_MASK)) |
+		   (planec_wm << DSPFW_PLANEC_SHIFT) |
+		   (cursorc_wm << DSPFW_CURSORC_SHIFT));
 
 	if (cxsr_enabled)
 		intel_set_memory_cxsr(dev_priv, true);
 }
 
+static void valleyview_update_sprite_wm(struct drm_plane *plane,
+					struct drm_crtc *crtc,
+					uint32_t sprite_width,
+					uint32_t sprite_height,
+					int pixel_size,
+					bool enabled, bool scaled)
+{
+	struct drm_device *dev = crtc->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int pipe = to_intel_plane(plane)->pipe;
+	int sprite = to_intel_plane(plane)->plane;
+	int drain_latency;
+	int plane_prec;
+	int sprite_dl;
+	int prec_mult;
+	const int high_precision = IS_CHERRYVIEW(dev) ?
+		DRAIN_LATENCY_PRECISION_32 : DRAIN_LATENCY_PRECISION_64;
+
+	sprite_dl = I915_READ(VLV_DDL(pipe)) & ~(DDL_SPRITE_PRECISION_HIGH(sprite) |
+		    (DRAIN_LATENCY_MASK << DDL_SPRITE_SHIFT(sprite)));
+
+	if (enabled && vlv_compute_drain_latency(crtc, pixel_size, &prec_mult,
+						 &drain_latency)) {
+		plane_prec = (prec_mult == high_precision) ?
+					   DDL_SPRITE_PRECISION_HIGH(sprite) :
+					   DDL_SPRITE_PRECISION_LOW(sprite);
+		sprite_dl |= plane_prec |
+			     (drain_latency << DDL_SPRITE_SHIFT(sprite));
+	}
+
+	I915_WRITE(VLV_DDL(pipe), sprite_dl);
+}
+
 static void g4x_update_wm(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
@@ -1419,14 +1615,14 @@ static void g4x_update_wm(struct drm_crtc *crtc)
 	bool cxsr_enabled;
 
 	if (g4x_compute_wm0(dev, PIPE_A,
-			    &g4x_wm_info, latency_ns,
-			    &g4x_cursor_wm_info, latency_ns,
+			    &g4x_wm_info, pessimal_latency_ns,
+			    &g4x_cursor_wm_info, pessimal_latency_ns,
 			    &planea_wm, &cursora_wm))
 		enabled |= 1 << PIPE_A;
 
 	if (g4x_compute_wm0(dev, PIPE_B,
-			    &g4x_wm_info, latency_ns,
-			    &g4x_cursor_wm_info, latency_ns,
+			    &g4x_wm_info, pessimal_latency_ns,
+			    &g4x_cursor_wm_info, pessimal_latency_ns,
 			    &planeb_wm, &cursorb_wm))
 		enabled |= 1 << PIPE_B;
 
@@ -1443,7 +1639,8 @@ static void g4x_update_wm(struct drm_crtc *crtc)
 		plane_sr = cursor_sr = 0;
 	}
 
-	DRM_DEBUG_KMS("Setting FIFO watermarks - A: plane=%d, cursor=%d, B: plane=%d, cursor=%d, SR: plane=%d, cursor=%d\n",
+	DRM_DEBUG_KMS("Setting FIFO watermarks - A: plane=%d, cursor=%d, "
+		      "B: plane=%d, cursor=%d, SR: plane=%d, cursor=%d\n",
 		      planea_wm, cursora_wm,
 		      planeb_wm, cursorb_wm,
 		      plane_sr, cursor_sr);
@@ -1452,7 +1649,7 @@ static void g4x_update_wm(struct drm_crtc *crtc)
 		   (plane_sr << DSPFW_SR_SHIFT) |
 		   (cursorb_wm << DSPFW_CURSORB_SHIFT) |
 		   (planeb_wm << DSPFW_PLANEB_SHIFT) |
-		   planea_wm);
+		   (planea_wm << DSPFW_PLANEA_SHIFT));
 	I915_WRITE(DSPFW2,
 		   (I915_READ(DSPFW2) & ~DSPFW_CURSORA_MASK) |
 		   (cursora_wm << DSPFW_CURSORA_SHIFT));
@@ -1526,8 +1723,11 @@ static void i965_update_wm(struct drm_crtc *unused_crtc)
 
 	/* 965 has limitations... */
 	I915_WRITE(DSPFW1, (srwm << DSPFW_SR_SHIFT) |
-		   (8 << 16) | (8 << 8) | (8 << 0));
-	I915_WRITE(DSPFW2, (8 << 8) | (8 << 0));
+		   (8 << DSPFW_CURSORB_SHIFT) |
+		   (8 << DSPFW_PLANEB_SHIFT) |
+		   (8 << DSPFW_PLANEA_SHIFT));
+	I915_WRITE(DSPFW2, (8 << DSPFW_CURSORA_SHIFT) |
+		   (8 << DSPFW_PLANEC_SHIFT_OLD));
 	/* update cursor SR watermark */
 	I915_WRITE(DSPFW3, (cursor_sr << DSPFW_CURSOR_SR_SHIFT));
 
@@ -1552,7 +1752,7 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc)
 	else if (!IS_GEN2(dev))
 		wm_info = &i915_wm_info;
 	else
-		wm_info = &i830_wm_info;
+		wm_info = &i830_a_wm_info;
 
 	fifo_size = dev_priv->display.get_fifo_size(dev, 0);
 	crtc = intel_get_crtc_for_plane(dev, 0);
@@ -1565,10 +1765,16 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc)
 		adjusted_mode = &to_intel_crtc(crtc)->config.adjusted_mode;
 		planea_wm = intel_calculate_wm(adjusted_mode->crtc_clock,
 					       wm_info, fifo_size, cpp,
-					       latency_ns);
+					       pessimal_latency_ns);
 		enabled = crtc;
-	} else
+	} else {
 		planea_wm = fifo_size - wm_info->guard_size;
+		if (planea_wm > (long)wm_info->max_wm)
+			planea_wm = wm_info->max_wm;
+	}
+
+	if (IS_GEN2(dev))
+		wm_info = &i830_bc_wm_info;
 
 	fifo_size = dev_priv->display.get_fifo_size(dev, 1);
 	crtc = intel_get_crtc_for_plane(dev, 1);
@@ -1581,13 +1787,16 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc)
 		adjusted_mode = &to_intel_crtc(crtc)->config.adjusted_mode;
 		planeb_wm = intel_calculate_wm(adjusted_mode->crtc_clock,
 					       wm_info, fifo_size, cpp,
-					       latency_ns);
+					       pessimal_latency_ns);
 		if (enabled == NULL)
 			enabled = crtc;
 		else
 			enabled = NULL;
-	} else
+	} else {
 		planeb_wm = fifo_size - wm_info->guard_size;
+		if (planeb_wm > (long)wm_info->max_wm)
+			planeb_wm = wm_info->max_wm;
+	}
 
 	DRM_DEBUG_KMS("FIFO watermarks - A: %d, B: %d\n", planea_wm, planeb_wm);
 
@@ -1674,7 +1883,7 @@ static void i845_update_wm(struct drm_crtc *unused_crtc)
 	planea_wm = intel_calculate_wm(adjusted_mode->crtc_clock,
 				       &i845_wm_info,
 				       dev_priv->display.get_fifo_size(dev, 0),
-				       4, latency_ns);
+				       4, pessimal_latency_ns);
 	fwater_lo = I915_READ(FW_BLC) & ~0xfff;
 	fwater_lo |= (3<<8) | planea_wm;
 
@@ -1751,6 +1960,14 @@ static uint32_t ilk_wm_fbc(uint32_t pri_val, uint32_t horiz_pixels,
 	return DIV_ROUND_UP(pri_val * 64, horiz_pixels * bytes_per_pixel) + 2;
 }
 
+struct skl_pipe_wm_parameters {
+	bool active;
+	uint32_t pipe_htotal;
+	uint32_t pixel_rate; /* in KHz */
+	struct intel_plane_wm_parameters plane[I915_MAX_PLANES];
+	struct intel_plane_wm_parameters cursor;
+};
+
 struct ilk_pipe_wm_parameters {
 	bool active;
 	uint32_t pipe_htotal;
@@ -2062,11 +2279,82 @@ hsw_compute_linetime_wm(struct drm_device *dev, struct drm_crtc *crtc)
 	       PIPE_WM_LINETIME_TIME(linetime);
 }
 
-static void intel_read_wm_latency(struct drm_device *dev, uint16_t wm[5])
+static void intel_read_wm_latency(struct drm_device *dev, uint16_t wm[8])
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
+	if (IS_GEN9(dev)) {
+		uint32_t val;
+		int ret, i;
+		int level, max_level = ilk_wm_max_level(dev);
+
+		/* read the first set of memory latencies[0:3] */
+		val = 0; /* data0 to be programmed to 0 for first set */
+		mutex_lock(&dev_priv->rps.hw_lock);
+		ret = sandybridge_pcode_read(dev_priv,
+					     GEN9_PCODE_READ_MEM_LATENCY,
+					     &val);
+		mutex_unlock(&dev_priv->rps.hw_lock);
+
+		if (ret) {
+			DRM_ERROR("SKL Mailbox read error = %d\n", ret);
+			return;
+		}
+
+		wm[0] = val & GEN9_MEM_LATENCY_LEVEL_MASK;
+		wm[1] = (val >> GEN9_MEM_LATENCY_LEVEL_1_5_SHIFT) &
+				GEN9_MEM_LATENCY_LEVEL_MASK;
+		wm[2] = (val >> GEN9_MEM_LATENCY_LEVEL_2_6_SHIFT) &
+				GEN9_MEM_LATENCY_LEVEL_MASK;
+		wm[3] = (val >> GEN9_MEM_LATENCY_LEVEL_3_7_SHIFT) &
+				GEN9_MEM_LATENCY_LEVEL_MASK;
+
+		/* read the second set of memory latencies[4:7] */
+		val = 1; /* data0 to be programmed to 1 for second set */
+		mutex_lock(&dev_priv->rps.hw_lock);
+		ret = sandybridge_pcode_read(dev_priv,
+					     GEN9_PCODE_READ_MEM_LATENCY,
+					     &val);
+		mutex_unlock(&dev_priv->rps.hw_lock);
+		if (ret) {
+			DRM_ERROR("SKL Mailbox read error = %d\n", ret);
+			return;
+		}
+
+		wm[4] = val & GEN9_MEM_LATENCY_LEVEL_MASK;
+		wm[5] = (val >> GEN9_MEM_LATENCY_LEVEL_1_5_SHIFT) &
+				GEN9_MEM_LATENCY_LEVEL_MASK;
+		wm[6] = (val >> GEN9_MEM_LATENCY_LEVEL_2_6_SHIFT) &
+				GEN9_MEM_LATENCY_LEVEL_MASK;
+		wm[7] = (val >> GEN9_MEM_LATENCY_LEVEL_3_7_SHIFT) &
+				GEN9_MEM_LATENCY_LEVEL_MASK;
+
+		/*
+		 * punit doesn't take into account the read latency so we need
+		 * to add 2us to the various latency levels we retrieve from
+		 * the punit.
+		 *   - W0 is a bit special in that it's the only level that
+		 *   can't be disabled if we want to have display working, so
+		 *   we always add 2us there.
+		 *   - For levels >=1, punit returns 0us latency when they are
+		 *   disabled, so we respect that and don't add 2us then
+		 *
+		 * Additionally, if a level n (n > 1) has a 0us latency, all
+		 * levels m (m >= n) need to be disabled. We make sure to
+		 * sanitize the values out of the punit to satisfy this
+		 * requirement.
+		 */
+		wm[0] += 2;
+		for (level = 1; level <= max_level; level++)
+			if (wm[level] != 0)
+				wm[level] += 2;
+			else {
+				for (i = level + 1; i <= max_level; i++)
+					wm[i] = 0;
+
+				break;
+			}
+	} else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
 		uint64_t sskpd = I915_READ64(MCH_SSKPD);
 
 		wm[0] = (sskpd >> 56) & 0xFF;
@@ -2114,7 +2402,9 @@ static void intel_fixup_cur_wm_latency(struct drm_device *dev, uint16_t wm[5])
 int ilk_wm_max_level(const struct drm_device *dev)
 {
 	/* how many WM levels are we expecting */
-	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+	if (IS_GEN9(dev))
+		return 7;
+	else if (IS_HASWELL(dev) || IS_BROADWELL(dev))
 		return 4;
 	else if (INTEL_INFO(dev)->gen >= 6)
 		return 3;
@@ -2124,7 +2414,7 @@ int ilk_wm_max_level(const struct drm_device *dev)
 
 static void intel_print_wm_latency(struct drm_device *dev,
 				   const char *name,
-				   const uint16_t wm[5])
+				   const uint16_t wm[8])
 {
 	int level, max_level = ilk_wm_max_level(dev);
 
@@ -2137,8 +2427,13 @@ static void intel_print_wm_latency(struct drm_device *dev,
 			continue;
 		}
 
-		/* WM1+ latency values in 0.5us units */
-		if (level > 0)
+		/*
+		 * - latencies are in us on gen9.
+		 * - before then, WM1+ latency values are in 0.5us units
+		 */
+		if (IS_GEN9(dev))
+			latency *= 10;
+		else if (level > 0)
 			latency *= 5;
 
 		DRM_DEBUG_KMS("%s WM%d latency %u (%u.%u usec)\n",
@@ -2206,6 +2501,14 @@ static void ilk_setup_wm_latency(struct drm_device *dev)
 		snb_wm_latency_quirk(dev);
 }
 
+static void skl_setup_wm_latency(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	intel_read_wm_latency(dev, dev_priv->wm.skl_latency);
+	intel_print_wm_latency(dev, "Gen9 Plane", dev_priv->wm.skl_latency);
+}
+
 static void ilk_compute_wm_parameters(struct drm_crtc *crtc,
 				      struct ilk_pipe_wm_parameters *p)
 {
@@ -2527,7 +2830,7 @@ static struct intel_pipe_wm *ilk_find_best_result(struct drm_device *dev,
 #define WM_DIRTY_FBC (1 << 24)
 #define WM_DIRTY_DDB (1 << 25)
 
-static unsigned int ilk_compute_wm_dirty(struct drm_device *dev,
+static unsigned int ilk_compute_wm_dirty(struct drm_i915_private *dev_priv,
 					 const struct ilk_wm_values *old,
 					 const struct ilk_wm_values *new)
 {
@@ -2535,7 +2838,7 @@ static unsigned int ilk_compute_wm_dirty(struct drm_device *dev,
 	enum pipe pipe;
 	int wm_lp;
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		if (old->wm_linetime[pipe] != new->wm_linetime[pipe]) {
 			dirty |= WM_DIRTY_LINETIME(pipe);
 			/* Must disable LP1+ watermarks too */
@@ -2621,7 +2924,7 @@ static void ilk_write_wm_values(struct drm_i915_private *dev_priv,
 	unsigned int dirty;
 	uint32_t val;
 
-	dirty = ilk_compute_wm_dirty(dev, previous, results);
+	dirty = ilk_compute_wm_dirty(dev_priv, previous, results);
 	if (!dirty)
 		return;
 
@@ -2696,6 +2999,769 @@ static bool ilk_disable_lp_wm(struct drm_device *dev)
 	return _ilk_disable_lp_wm(dev_priv, WM_DIRTY_LP_ALL);
 }
 
+/*
+ * On gen9, we need to allocate Display Data Buffer (DDB) portions to the
+ * different active planes.
+ */
+
+#define SKL_DDB_SIZE		896	/* in blocks */
+
+static void
+skl_ddb_get_pipe_allocation_limits(struct drm_device *dev,
+				   struct drm_crtc *for_crtc,
+				   const struct intel_wm_config *config,
+				   const struct skl_pipe_wm_parameters *params,
+				   struct skl_ddb_entry *alloc /* out */)
+{
+	struct drm_crtc *crtc;
+	unsigned int pipe_size, ddb_size;
+	int nth_active_pipe;
+
+	if (!params->active) {
+		alloc->start = 0;
+		alloc->end = 0;
+		return;
+	}
+
+	ddb_size = SKL_DDB_SIZE;
+
+	ddb_size -= 4; /* 4 blocks for bypass path allocation */
+
+	nth_active_pipe = 0;
+	for_each_crtc(dev, crtc) {
+		if (!intel_crtc_active(crtc))
+			continue;
+
+		if (crtc == for_crtc)
+			break;
+
+		nth_active_pipe++;
+	}
+
+	pipe_size = ddb_size / config->num_pipes_active;
+	alloc->start = nth_active_pipe * ddb_size / config->num_pipes_active;
+	alloc->end = alloc->start + pipe_size;
+}
+
+static unsigned int skl_cursor_allocation(const struct intel_wm_config *config)
+{
+	if (config->num_pipes_active == 1)
+		return 32;
+
+	return 8;
+}
+
+static void skl_ddb_entry_init_from_hw(struct skl_ddb_entry *entry, u32 reg)
+{
+	entry->start = reg & 0x3ff;
+	entry->end = (reg >> 16) & 0x3ff;
+	if (entry->end)
+		entry->end += 1;
+}
+
+void skl_ddb_get_hw_state(struct drm_i915_private *dev_priv,
+			  struct skl_ddb_allocation *ddb /* out */)
+{
+	struct drm_device *dev = dev_priv->dev;
+	enum pipe pipe;
+	int plane;
+	u32 val;
+
+	for_each_pipe(dev_priv, pipe) {
+		for_each_plane(pipe, plane) {
+			val = I915_READ(PLANE_BUF_CFG(pipe, plane));
+			skl_ddb_entry_init_from_hw(&ddb->plane[pipe][plane],
+						   val);
+		}
+
+		val = I915_READ(CUR_BUF_CFG(pipe));
+		skl_ddb_entry_init_from_hw(&ddb->cursor[pipe], val);
+	}
+}
+
+static unsigned int
+skl_plane_relative_data_rate(const struct intel_plane_wm_parameters *p)
+{
+	return p->horiz_pixels * p->vert_pixels * p->bytes_per_pixel;
+}
+
+/*
+ * We don't overflow 32 bits. Worst case is 3 planes enabled, each fetching
+ * a 8192x4096@32bpp framebuffer:
+ *   3 * 4096 * 8192  * 4 < 2^32
+ */
+static unsigned int
+skl_get_total_relative_data_rate(struct intel_crtc *intel_crtc,
+				 const struct skl_pipe_wm_parameters *params)
+{
+	unsigned int total_data_rate = 0;
+	int plane;
+
+	for (plane = 0; plane < intel_num_planes(intel_crtc); plane++) {
+		const struct intel_plane_wm_parameters *p;
+
+		p = &params->plane[plane];
+		if (!p->enabled)
+			continue;
+
+		total_data_rate += skl_plane_relative_data_rate(p);
+	}
+
+	return total_data_rate;
+}
+
+static void
+skl_allocate_pipe_ddb(struct drm_crtc *crtc,
+		      const struct intel_wm_config *config,
+		      const struct skl_pipe_wm_parameters *params,
+		      struct skl_ddb_allocation *ddb /* out */)
+{
+	struct drm_device *dev = crtc->dev;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	enum pipe pipe = intel_crtc->pipe;
+	struct skl_ddb_entry *alloc = &ddb->pipe[pipe];
+	uint16_t alloc_size, start, cursor_blocks;
+	unsigned int total_data_rate;
+	int plane;
+
+	skl_ddb_get_pipe_allocation_limits(dev, crtc, config, params, alloc);
+	alloc_size = skl_ddb_entry_size(alloc);
+	if (alloc_size == 0) {
+		memset(ddb->plane[pipe], 0, sizeof(ddb->plane[pipe]));
+		memset(&ddb->cursor[pipe], 0, sizeof(ddb->cursor[pipe]));
+		return;
+	}
+
+	cursor_blocks = skl_cursor_allocation(config);
+	ddb->cursor[pipe].start = alloc->end - cursor_blocks;
+	ddb->cursor[pipe].end = alloc->end;
+
+	alloc_size -= cursor_blocks;
+	alloc->end -= cursor_blocks;
+
+	/*
+	 * Each active plane get a portion of the remaining space, in
+	 * proportion to the amount of data they need to fetch from memory.
+	 *
+	 * FIXME: we may not allocate every single block here.
+	 */
+	total_data_rate = skl_get_total_relative_data_rate(intel_crtc, params);
+
+	start = alloc->start;
+	for (plane = 0; plane < intel_num_planes(intel_crtc); plane++) {
+		const struct intel_plane_wm_parameters *p;
+		unsigned int data_rate;
+		uint16_t plane_blocks;
+
+		p = &params->plane[plane];
+		if (!p->enabled)
+			continue;
+
+		data_rate = skl_plane_relative_data_rate(p);
+
+		/*
+		 * promote the expression to 64 bits to avoid overflowing, the
+		 * result is < available as data_rate / total_data_rate < 1
+		 */
+		plane_blocks = div_u64((uint64_t)alloc_size * data_rate,
+				       total_data_rate);
+
+		ddb->plane[pipe][plane].start = start;
+		ddb->plane[pipe][plane].end = start + plane_blocks;
+
+		start += plane_blocks;
+	}
+
+}
+
+static uint32_t skl_pipe_pixel_rate(const struct intel_crtc_config *config)
+{
+	/* TODO: Take into account the scalers once we support them */
+	return config->adjusted_mode.crtc_clock;
+}
+
+/*
+ * The max latency should be 257 (max the punit can code is 255 and we add 2us
+ * for the read latency) and bytes_per_pixel should always be <= 8, so that
+ * should allow pixel_rate up to ~2 GHz which seems sufficient since max
+ * 2xcdclk is 1350 MHz and the pixel rate should never exceed that.
+*/
+static uint32_t skl_wm_method1(uint32_t pixel_rate, uint8_t bytes_per_pixel,
+			       uint32_t latency)
+{
+	uint32_t wm_intermediate_val, ret;
+
+	if (latency == 0)
+		return UINT_MAX;
+
+	wm_intermediate_val = latency * pixel_rate * bytes_per_pixel;
+	ret = DIV_ROUND_UP(wm_intermediate_val, 1000);
+
+	return ret;
+}
+
+static uint32_t skl_wm_method2(uint32_t pixel_rate, uint32_t pipe_htotal,
+			       uint32_t horiz_pixels, uint8_t bytes_per_pixel,
+			       uint32_t latency)
+{
+	uint32_t ret, plane_bytes_per_line, wm_intermediate_val;
+
+	if (latency == 0)
+		return UINT_MAX;
+
+	plane_bytes_per_line = horiz_pixels * bytes_per_pixel;
+	wm_intermediate_val = latency * pixel_rate;
+	ret = DIV_ROUND_UP(wm_intermediate_val, pipe_htotal * 1000) *
+				plane_bytes_per_line;
+
+	return ret;
+}
+
+static bool skl_ddb_allocation_changed(const struct skl_ddb_allocation *new_ddb,
+				       const struct intel_crtc *intel_crtc)
+{
+	struct drm_device *dev = intel_crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	const struct skl_ddb_allocation *cur_ddb = &dev_priv->wm.skl_hw.ddb;
+	enum pipe pipe = intel_crtc->pipe;
+
+	if (memcmp(new_ddb->plane[pipe], cur_ddb->plane[pipe],
+		   sizeof(new_ddb->plane[pipe])))
+		return true;
+
+	if (memcmp(&new_ddb->cursor[pipe], &cur_ddb->cursor[pipe],
+		    sizeof(new_ddb->cursor[pipe])))
+		return true;
+
+	return false;
+}
+
+static void skl_compute_wm_global_parameters(struct drm_device *dev,
+					     struct intel_wm_config *config)
+{
+	struct drm_crtc *crtc;
+	struct drm_plane *plane;
+
+	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
+		config->num_pipes_active += intel_crtc_active(crtc);
+
+	/* FIXME: I don't think we need those two global parameters on SKL */
+	list_for_each_entry(plane, &dev->mode_config.plane_list, head) {
+		struct intel_plane *intel_plane = to_intel_plane(plane);
+
+		config->sprites_enabled |= intel_plane->wm.enabled;
+		config->sprites_scaled |= intel_plane->wm.scaled;
+	}
+}
+
+static void skl_compute_wm_pipe_parameters(struct drm_crtc *crtc,
+					   struct skl_pipe_wm_parameters *p)
+{
+	struct drm_device *dev = crtc->dev;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	enum pipe pipe = intel_crtc->pipe;
+	struct drm_plane *plane;
+	int i = 1; /* Index for sprite planes start */
+
+	p->active = intel_crtc_active(crtc);
+	if (p->active) {
+		p->pipe_htotal = intel_crtc->config.adjusted_mode.crtc_htotal;
+		p->pixel_rate = skl_pipe_pixel_rate(&intel_crtc->config);
+
+		/*
+		 * For now, assume primary and cursor planes are always enabled.
+		 */
+		p->plane[0].enabled = true;
+		p->plane[0].bytes_per_pixel =
+			crtc->primary->fb->bits_per_pixel / 8;
+		p->plane[0].horiz_pixels = intel_crtc->config.pipe_src_w;
+		p->plane[0].vert_pixels = intel_crtc->config.pipe_src_h;
+
+		p->cursor.enabled = true;
+		p->cursor.bytes_per_pixel = 4;
+		p->cursor.horiz_pixels = intel_crtc->cursor_width ?
+					 intel_crtc->cursor_width : 64;
+	}
+
+	list_for_each_entry(plane, &dev->mode_config.plane_list, head) {
+		struct intel_plane *intel_plane = to_intel_plane(plane);
+
+		if (intel_plane->pipe == pipe)
+			p->plane[i++] = intel_plane->wm;
+	}
+}
+
+static bool skl_compute_plane_wm(struct skl_pipe_wm_parameters *p,
+				 struct intel_plane_wm_parameters *p_params,
+				 uint16_t ddb_allocation,
+				 uint32_t mem_value,
+				 uint16_t *out_blocks, /* out */
+				 uint8_t *out_lines /* out */)
+{
+	uint32_t method1, method2, plane_bytes_per_line, res_blocks, res_lines;
+	uint32_t result_bytes;
+
+	if (mem_value == 0 || !p->active || !p_params->enabled)
+		return false;
+
+	method1 = skl_wm_method1(p->pixel_rate,
+				 p_params->bytes_per_pixel,
+				 mem_value);
+	method2 = skl_wm_method2(p->pixel_rate,
+				 p->pipe_htotal,
+				 p_params->horiz_pixels,
+				 p_params->bytes_per_pixel,
+				 mem_value);
+
+	plane_bytes_per_line = p_params->horiz_pixels *
+					p_params->bytes_per_pixel;
+
+	/* For now xtile and linear */
+	if (((ddb_allocation * 512) / plane_bytes_per_line) >= 1)
+		result_bytes = min(method1, method2);
+	else
+		result_bytes = method1;
+
+	res_blocks = DIV_ROUND_UP(result_bytes, 512) + 1;
+	res_lines = DIV_ROUND_UP(result_bytes, plane_bytes_per_line);
+
+	if (res_blocks > ddb_allocation || res_lines > 31)
+		return false;
+
+	*out_blocks = res_blocks;
+	*out_lines = res_lines;
+
+	return true;
+}
+
+static void skl_compute_wm_level(const struct drm_i915_private *dev_priv,
+				 struct skl_ddb_allocation *ddb,
+				 struct skl_pipe_wm_parameters *p,
+				 enum pipe pipe,
+				 int level,
+				 int num_planes,
+				 struct skl_wm_level *result)
+{
+	uint16_t latency = dev_priv->wm.skl_latency[level];
+	uint16_t ddb_blocks;
+	int i;
+
+	for (i = 0; i < num_planes; i++) {
+		ddb_blocks = skl_ddb_entry_size(&ddb->plane[pipe][i]);
+
+		result->plane_en[i] = skl_compute_plane_wm(p, &p->plane[i],
+						ddb_blocks,
+						latency,
+						&result->plane_res_b[i],
+						&result->plane_res_l[i]);
+	}
+
+	ddb_blocks = skl_ddb_entry_size(&ddb->cursor[pipe]);
+	result->cursor_en = skl_compute_plane_wm(p, &p->cursor, ddb_blocks,
+						 latency, &result->cursor_res_b,
+						 &result->cursor_res_l);
+}
+
+static uint32_t
+skl_compute_linetime_wm(struct drm_crtc *crtc, struct skl_pipe_wm_parameters *p)
+{
+	if (!intel_crtc_active(crtc))
+		return 0;
+
+	return DIV_ROUND_UP(8 * p->pipe_htotal * 1000, p->pixel_rate);
+
+}
+
+static void skl_compute_transition_wm(struct drm_crtc *crtc,
+				      struct skl_pipe_wm_parameters *params,
+				      struct skl_wm_level *trans_wm /* out */)
+{
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	int i;
+
+	if (!params->active)
+		return;
+
+	/* Until we know more, just disable transition WMs */
+	for (i = 0; i < intel_num_planes(intel_crtc); i++)
+		trans_wm->plane_en[i] = false;
+	trans_wm->cursor_en = false;
+}
+
+static void skl_compute_pipe_wm(struct drm_crtc *crtc,
+				struct skl_ddb_allocation *ddb,
+				struct skl_pipe_wm_parameters *params,
+				struct skl_pipe_wm *pipe_wm)
+{
+	struct drm_device *dev = crtc->dev;
+	const struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	int level, max_level = ilk_wm_max_level(dev);
+
+	for (level = 0; level <= max_level; level++) {
+		skl_compute_wm_level(dev_priv, ddb, params, intel_crtc->pipe,
+				     level, intel_num_planes(intel_crtc),
+				     &pipe_wm->wm[level]);
+	}
+	pipe_wm->linetime = skl_compute_linetime_wm(crtc, params);
+
+	skl_compute_transition_wm(crtc, params, &pipe_wm->trans_wm);
+}
+
+static void skl_compute_wm_results(struct drm_device *dev,
+				   struct skl_pipe_wm_parameters *p,
+				   struct skl_pipe_wm *p_wm,
+				   struct skl_wm_values *r,
+				   struct intel_crtc *intel_crtc)
+{
+	int level, max_level = ilk_wm_max_level(dev);
+	enum pipe pipe = intel_crtc->pipe;
+	uint32_t temp;
+	int i;
+
+	for (level = 0; level <= max_level; level++) {
+		for (i = 0; i < intel_num_planes(intel_crtc); i++) {
+			temp = 0;
+
+			temp |= p_wm->wm[level].plane_res_l[i] <<
+					PLANE_WM_LINES_SHIFT;
+			temp |= p_wm->wm[level].plane_res_b[i];
+			if (p_wm->wm[level].plane_en[i])
+				temp |= PLANE_WM_EN;
+
+			r->plane[pipe][i][level] = temp;
+		}
+
+		temp = 0;
+
+		temp |= p_wm->wm[level].cursor_res_l << PLANE_WM_LINES_SHIFT;
+		temp |= p_wm->wm[level].cursor_res_b;
+
+		if (p_wm->wm[level].cursor_en)
+			temp |= PLANE_WM_EN;
+
+		r->cursor[pipe][level] = temp;
+
+	}
+
+	/* transition WMs */
+	for (i = 0; i < intel_num_planes(intel_crtc); i++) {
+		temp = 0;
+		temp |= p_wm->trans_wm.plane_res_l[i] << PLANE_WM_LINES_SHIFT;
+		temp |= p_wm->trans_wm.plane_res_b[i];
+		if (p_wm->trans_wm.plane_en[i])
+			temp |= PLANE_WM_EN;
+
+		r->plane_trans[pipe][i] = temp;
+	}
+
+	temp = 0;
+	temp |= p_wm->trans_wm.cursor_res_l << PLANE_WM_LINES_SHIFT;
+	temp |= p_wm->trans_wm.cursor_res_b;
+	if (p_wm->trans_wm.cursor_en)
+		temp |= PLANE_WM_EN;
+
+	r->cursor_trans[pipe] = temp;
+
+	r->wm_linetime[pipe] = p_wm->linetime;
+}
+
+static void skl_ddb_entry_write(struct drm_i915_private *dev_priv, uint32_t reg,
+				const struct skl_ddb_entry *entry)
+{
+	if (entry->end)
+		I915_WRITE(reg, (entry->end - 1) << 16 | entry->start);
+	else
+		I915_WRITE(reg, 0);
+}
+
+static void skl_write_wm_values(struct drm_i915_private *dev_priv,
+				const struct skl_wm_values *new)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct intel_crtc *crtc;
+
+	list_for_each_entry(crtc, &dev->mode_config.crtc_list, base.head) {
+		int i, level, max_level = ilk_wm_max_level(dev);
+		enum pipe pipe = crtc->pipe;
+
+		if (!new->dirty[pipe])
+			continue;
+
+		I915_WRITE(PIPE_WM_LINETIME(pipe), new->wm_linetime[pipe]);
+
+		for (level = 0; level <= max_level; level++) {
+			for (i = 0; i < intel_num_planes(crtc); i++)
+				I915_WRITE(PLANE_WM(pipe, i, level),
+					   new->plane[pipe][i][level]);
+			I915_WRITE(CUR_WM(pipe, level),
+				   new->cursor[pipe][level]);
+		}
+		for (i = 0; i < intel_num_planes(crtc); i++)
+			I915_WRITE(PLANE_WM_TRANS(pipe, i),
+				   new->plane_trans[pipe][i]);
+		I915_WRITE(CUR_WM_TRANS(pipe), new->cursor_trans[pipe]);
+
+		for (i = 0; i < intel_num_planes(crtc); i++)
+			skl_ddb_entry_write(dev_priv,
+					    PLANE_BUF_CFG(pipe, i),
+					    &new->ddb.plane[pipe][i]);
+
+		skl_ddb_entry_write(dev_priv, CUR_BUF_CFG(pipe),
+				    &new->ddb.cursor[pipe]);
+	}
+}
+
+/*
+ * When setting up a new DDB allocation arrangement, we need to correctly
+ * sequence the times at which the new allocations for the pipes are taken into
+ * account or we'll have pipes fetching from space previously allocated to
+ * another pipe.
+ *
+ * Roughly the sequence looks like:
+ *  1. re-allocate the pipe(s) with the allocation being reduced and not
+ *     overlapping with a previous light-up pipe (another way to put it is:
+ *     pipes with their new allocation strickly included into their old ones).
+ *  2. re-allocate the other pipes that get their allocation reduced
+ *  3. allocate the pipes having their allocation increased
+ *
+ * Steps 1. and 2. are here to take care of the following case:
+ * - Initially DDB looks like this:
+ *     |   B    |   C    |
+ * - enable pipe A.
+ * - pipe B has a reduced DDB allocation that overlaps with the old pipe C
+ *   allocation
+ *     |  A  |  B  |  C  |
+ *
+ * We need to sequence the re-allocation: C, B, A (and not B, C, A).
+ */
+
+static void
+skl_wm_flush_pipe(struct drm_i915_private *dev_priv, enum pipe pipe, int pass)
+{
+	struct drm_device *dev = dev_priv->dev;
+	int plane;
+
+	DRM_DEBUG_KMS("flush pipe %c (pass %d)\n", pipe_name(pipe), pass);
+
+	for_each_plane(pipe, plane) {
+		I915_WRITE(PLANE_SURF(pipe, plane),
+			   I915_READ(PLANE_SURF(pipe, plane)));
+	}
+	I915_WRITE(CURBASE(pipe), I915_READ(CURBASE(pipe)));
+}
+
+static bool
+skl_ddb_allocation_included(const struct skl_ddb_allocation *old,
+			    const struct skl_ddb_allocation *new,
+			    enum pipe pipe)
+{
+	uint16_t old_size, new_size;
+
+	old_size = skl_ddb_entry_size(&old->pipe[pipe]);
+	new_size = skl_ddb_entry_size(&new->pipe[pipe]);
+
+	return old_size != new_size &&
+	       new->pipe[pipe].start >= old->pipe[pipe].start &&
+	       new->pipe[pipe].end <= old->pipe[pipe].end;
+}
+
+static void skl_flush_wm_values(struct drm_i915_private *dev_priv,
+				struct skl_wm_values *new_values)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct skl_ddb_allocation *cur_ddb, *new_ddb;
+	bool reallocated[I915_MAX_PIPES] = {false, false, false};
+	struct intel_crtc *crtc;
+	enum pipe pipe;
+
+	new_ddb = &new_values->ddb;
+	cur_ddb = &dev_priv->wm.skl_hw.ddb;
+
+	/*
+	 * First pass: flush the pipes with the new allocation contained into
+	 * the old space.
+	 *
+	 * We'll wait for the vblank on those pipes to ensure we can safely
+	 * re-allocate the freed space without this pipe fetching from it.
+	 */
+	for_each_intel_crtc(dev, crtc) {
+		if (!crtc->active)
+			continue;
+
+		pipe = crtc->pipe;
+
+		if (!skl_ddb_allocation_included(cur_ddb, new_ddb, pipe))
+			continue;
+
+		skl_wm_flush_pipe(dev_priv, pipe, 1);
+		intel_wait_for_vblank(dev, pipe);
+
+		reallocated[pipe] = true;
+	}
+
+
+	/*
+	 * Second pass: flush the pipes that are having their allocation
+	 * reduced, but overlapping with a previous allocation.
+	 *
+	 * Here as well we need to wait for the vblank to make sure the freed
+	 * space is not used anymore.
+	 */
+	for_each_intel_crtc(dev, crtc) {
+		if (!crtc->active)
+			continue;
+
+		pipe = crtc->pipe;
+
+		if (reallocated[pipe])
+			continue;
+
+		if (skl_ddb_entry_size(&new_ddb->pipe[pipe]) <
+		    skl_ddb_entry_size(&cur_ddb->pipe[pipe])) {
+			skl_wm_flush_pipe(dev_priv, pipe, 2);
+			intel_wait_for_vblank(dev, pipe);
+		}
+
+		reallocated[pipe] = true;
+	}
+
+	/*
+	 * Third pass: flush the pipes that got more space allocated.
+	 *
+	 * We don't need to actively wait for the update here, next vblank
+	 * will just get more DDB space with the correct WM values.
+	 */
+	for_each_intel_crtc(dev, crtc) {
+		if (!crtc->active)
+			continue;
+
+		pipe = crtc->pipe;
+
+		/*
+		 * At this point, only the pipes more space than before are
+		 * left to re-allocate.
+		 */
+		if (reallocated[pipe])
+			continue;
+
+		skl_wm_flush_pipe(dev_priv, pipe, 3);
+	}
+}
+
+static bool skl_update_pipe_wm(struct drm_crtc *crtc,
+			       struct skl_pipe_wm_parameters *params,
+			       struct intel_wm_config *config,
+			       struct skl_ddb_allocation *ddb, /* out */
+			       struct skl_pipe_wm *pipe_wm /* out */)
+{
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+
+	skl_compute_wm_pipe_parameters(crtc, params);
+	skl_allocate_pipe_ddb(crtc, config, params, ddb);
+	skl_compute_pipe_wm(crtc, ddb, params, pipe_wm);
+
+	if (!memcmp(&intel_crtc->wm.skl_active, pipe_wm, sizeof(*pipe_wm)))
+		return false;
+
+	intel_crtc->wm.skl_active = *pipe_wm;
+	return true;
+}
+
+static void skl_update_other_pipe_wm(struct drm_device *dev,
+				     struct drm_crtc *crtc,
+				     struct intel_wm_config *config,
+				     struct skl_wm_values *r)
+{
+	struct intel_crtc *intel_crtc;
+	struct intel_crtc *this_crtc = to_intel_crtc(crtc);
+
+	/*
+	 * If the WM update hasn't changed the allocation for this_crtc (the
+	 * crtc we are currently computing the new WM values for), other
+	 * enabled crtcs will keep the same allocation and we don't need to
+	 * recompute anything for them.
+	 */
+	if (!skl_ddb_allocation_changed(&r->ddb, this_crtc))
+		return;
+
+	/*
+	 * Otherwise, because of this_crtc being freshly enabled/disabled, the
+	 * other active pipes need new DDB allocation and WM values.
+	 */
+	list_for_each_entry(intel_crtc, &dev->mode_config.crtc_list,
+				base.head) {
+		struct skl_pipe_wm_parameters params = {};
+		struct skl_pipe_wm pipe_wm = {};
+		bool wm_changed;
+
+		if (this_crtc->pipe == intel_crtc->pipe)
+			continue;
+
+		if (!intel_crtc->active)
+			continue;
+
+		wm_changed = skl_update_pipe_wm(&intel_crtc->base,
+						&params, config,
+						&r->ddb, &pipe_wm);
+
+		/*
+		 * If we end up re-computing the other pipe WM values, it's
+		 * because it was really needed, so we expect the WM values to
+		 * be different.
+		 */
+		WARN_ON(!wm_changed);
+
+		skl_compute_wm_results(dev, &params, &pipe_wm, r, intel_crtc);
+		r->dirty[intel_crtc->pipe] = true;
+	}
+}
+
+static void skl_update_wm(struct drm_crtc *crtc)
+{
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	struct drm_device *dev = crtc->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct skl_pipe_wm_parameters params = {};
+	struct skl_wm_values *results = &dev_priv->wm.skl_results;
+	struct skl_pipe_wm pipe_wm = {};
+	struct intel_wm_config config = {};
+
+	memset(results, 0, sizeof(*results));
+
+	skl_compute_wm_global_parameters(dev, &config);
+
+	if (!skl_update_pipe_wm(crtc, &params, &config,
+				&results->ddb, &pipe_wm))
+		return;
+
+	skl_compute_wm_results(dev, &params, &pipe_wm, results, intel_crtc);
+	results->dirty[intel_crtc->pipe] = true;
+
+	skl_update_other_pipe_wm(dev, crtc, &config, results);
+	skl_write_wm_values(dev_priv, results);
+	skl_flush_wm_values(dev_priv, results);
+
+	/* store the new configuration */
+	dev_priv->wm.skl_hw = *results;
+}
+
+static void
+skl_update_sprite_wm(struct drm_plane *plane, struct drm_crtc *crtc,
+		     uint32_t sprite_width, uint32_t sprite_height,
+		     int pixel_size, bool enabled, bool scaled)
+{
+	struct intel_plane *intel_plane = to_intel_plane(plane);
+
+	intel_plane->wm.enabled = enabled;
+	intel_plane->wm.scaled = scaled;
+	intel_plane->wm.horiz_pixels = sprite_width;
+	intel_plane->wm.vert_pixels = sprite_height;
+	intel_plane->wm.bytes_per_pixel = pixel_size;
+
+	skl_update_wm(crtc);
+}
+
 static void ilk_update_wm(struct drm_crtc *crtc)
 {
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
@@ -2770,6 +3836,113 @@ ilk_update_sprite_wm(struct drm_plane *plane,
 	ilk_update_wm(crtc);
 }
 
+static void skl_pipe_wm_active_state(uint32_t val,
+				     struct skl_pipe_wm *active,
+				     bool is_transwm,
+				     bool is_cursor,
+				     int i,
+				     int level)
+{
+	bool is_enabled = (val & PLANE_WM_EN) != 0;
+
+	if (!is_transwm) {
+		if (!is_cursor) {
+			active->wm[level].plane_en[i] = is_enabled;
+			active->wm[level].plane_res_b[i] =
+					val & PLANE_WM_BLOCKS_MASK;
+			active->wm[level].plane_res_l[i] =
+					(val >> PLANE_WM_LINES_SHIFT) &
+						PLANE_WM_LINES_MASK;
+		} else {
+			active->wm[level].cursor_en = is_enabled;
+			active->wm[level].cursor_res_b =
+					val & PLANE_WM_BLOCKS_MASK;
+			active->wm[level].cursor_res_l =
+					(val >> PLANE_WM_LINES_SHIFT) &
+						PLANE_WM_LINES_MASK;
+		}
+	} else {
+		if (!is_cursor) {
+			active->trans_wm.plane_en[i] = is_enabled;
+			active->trans_wm.plane_res_b[i] =
+					val & PLANE_WM_BLOCKS_MASK;
+			active->trans_wm.plane_res_l[i] =
+					(val >> PLANE_WM_LINES_SHIFT) &
+						PLANE_WM_LINES_MASK;
+		} else {
+			active->trans_wm.cursor_en = is_enabled;
+			active->trans_wm.cursor_res_b =
+					val & PLANE_WM_BLOCKS_MASK;
+			active->trans_wm.cursor_res_l =
+					(val >> PLANE_WM_LINES_SHIFT) &
+						PLANE_WM_LINES_MASK;
+		}
+	}
+}
+
+static void skl_pipe_wm_get_hw_state(struct drm_crtc *crtc)
+{
+	struct drm_device *dev = crtc->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct skl_wm_values *hw = &dev_priv->wm.skl_hw;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	struct skl_pipe_wm *active = &intel_crtc->wm.skl_active;
+	enum pipe pipe = intel_crtc->pipe;
+	int level, i, max_level;
+	uint32_t temp;
+
+	max_level = ilk_wm_max_level(dev);
+
+	hw->wm_linetime[pipe] = I915_READ(PIPE_WM_LINETIME(pipe));
+
+	for (level = 0; level <= max_level; level++) {
+		for (i = 0; i < intel_num_planes(intel_crtc); i++)
+			hw->plane[pipe][i][level] =
+					I915_READ(PLANE_WM(pipe, i, level));
+		hw->cursor[pipe][level] = I915_READ(CUR_WM(pipe, level));
+	}
+
+	for (i = 0; i < intel_num_planes(intel_crtc); i++)
+		hw->plane_trans[pipe][i] = I915_READ(PLANE_WM_TRANS(pipe, i));
+	hw->cursor_trans[pipe] = I915_READ(CUR_WM_TRANS(pipe));
+
+	if (!intel_crtc_active(crtc))
+		return;
+
+	hw->dirty[pipe] = true;
+
+	active->linetime = hw->wm_linetime[pipe];
+
+	for (level = 0; level <= max_level; level++) {
+		for (i = 0; i < intel_num_planes(intel_crtc); i++) {
+			temp = hw->plane[pipe][i][level];
+			skl_pipe_wm_active_state(temp, active, false,
+						false, i, level);
+		}
+		temp = hw->cursor[pipe][level];
+		skl_pipe_wm_active_state(temp, active, false, true, i, level);
+	}
+
+	for (i = 0; i < intel_num_planes(intel_crtc); i++) {
+		temp = hw->plane_trans[pipe][i];
+		skl_pipe_wm_active_state(temp, active, true, false, i, 0);
+	}
+
+	temp = hw->cursor_trans[pipe];
+	skl_pipe_wm_active_state(temp, active, true, true, i, 0);
+}
+
+void skl_wm_get_hw_state(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct skl_ddb_allocation *ddb = &dev_priv->wm.skl_hw.ddb;
+	struct drm_crtc *crtc;
+
+	skl_ddb_get_hw_state(dev_priv, ddb);
+	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
+		skl_pipe_wm_get_hw_state(crtc);
+}
+
 static void ilk_pipe_wm_get_hw_state(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
@@ -3190,16 +4363,7 @@ static u32 gen6_rps_pm_mask(struct drm_i915_private *dev_priv, u8 val)
 	mask |= dev_priv->pm_rps_events & (GEN6_PM_RP_DOWN_EI_EXPIRED | GEN6_PM_RP_UP_EI_EXPIRED);
 	mask &= dev_priv->pm_rps_events;
 
-	/* IVB and SNB hard hangs on looping batchbuffer
-	 * if GEN6_PM_UP_EI_EXPIRED is masked.
-	 */
-	if (INTEL_INFO(dev_priv->dev)->gen <= 7 && !IS_HASWELL(dev_priv->dev))
-		mask |= GEN6_PM_RP_UP_EI_EXPIRED;
-
-	if (IS_GEN8(dev_priv->dev))
-		mask |= GEN8_PMINTR_REDIRECT_TO_NON_DISP;
-
-	return ~mask;
+	return gen6_sanitize_rps_pm_mask(dev_priv, ~mask);
 }
 
 /* gen6_set_rps is called to update the frequency request, but should also be
@@ -3268,7 +4432,8 @@ static void vlv_set_rps_idle(struct drm_i915_private *dev_priv)
 		return;
 
 	/* Mask turbo interrupt so that they will not come in between */
-	I915_WRITE(GEN6_PMINTRMSK, 0xffffffff);
+	I915_WRITE(GEN6_PMINTRMSK,
+		   gen6_sanitize_rps_pm_mask(dev_priv, ~0));
 
 	vlv_force_gfx_clock(dev_priv, true);
 
@@ -3278,7 +4443,7 @@ static void vlv_set_rps_idle(struct drm_i915_private *dev_priv)
 					dev_priv->rps.min_freq_softlimit);
 
 	if (wait_for(((vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS))
-				& GENFREQSTATUS) == 0, 5))
+				& GENFREQSTATUS) == 0, 100))
 		DRM_ERROR("timed out waiting for Punit\n");
 
 	vlv_force_gfx_clock(dev_priv, false);
@@ -3327,10 +4492,9 @@ void valleyview_set_rps(struct drm_device *dev, u8 val)
 	WARN_ON(val > dev_priv->rps.max_freq_softlimit);
 	WARN_ON(val < dev_priv->rps.min_freq_softlimit);
 
-	DRM_DEBUG_DRIVER("GPU freq request from %d MHz (%u) to %d MHz (%u)\n",
-			 vlv_gpu_freq(dev_priv, dev_priv->rps.cur_freq),
-			 dev_priv->rps.cur_freq,
-			 vlv_gpu_freq(dev_priv, val), val);
+	if (WARN_ONCE(IS_CHERRYVIEW(dev) && (val & 1),
+		      "Odd GPU freq value\n"))
+		val &= ~1;
 
 	if (val != dev_priv->rps.cur_freq)
 		vlv_punit_write(dev_priv, PUNIT_REG_GPU_FREQ_REQ, val);
@@ -3341,43 +4505,11 @@ void valleyview_set_rps(struct drm_device *dev, u8 val)
 	trace_intel_gpu_freq_change(vlv_gpu_freq(dev_priv, val));
 }
 
-static void gen8_disable_rps_interrupts(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	I915_WRITE(GEN6_PMINTRMSK, ~GEN8_PMINTR_REDIRECT_TO_NON_DISP);
-	I915_WRITE(GEN8_GT_IER(2), I915_READ(GEN8_GT_IER(2)) &
-				   ~dev_priv->pm_rps_events);
-	/* Complete PM interrupt masking here doesn't race with the rps work
-	 * item again unmasking PM interrupts because that is using a different
-	 * register (GEN8_GT_IMR(2)) to mask PM interrupts. The only risk is in
-	 * leaving stale bits in GEN8_GT_IIR(2) and GEN8_GT_IMR(2) which
-	 * gen8_enable_rps will clean up. */
-
-	spin_lock_irq(&dev_priv->irq_lock);
-	dev_priv->rps.pm_iir = 0;
-	spin_unlock_irq(&dev_priv->irq_lock);
-
-	I915_WRITE(GEN8_GT_IIR(2), dev_priv->pm_rps_events);
-}
-
-static void gen6_disable_rps_interrupts(struct drm_device *dev)
+static void gen9_disable_rps(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	I915_WRITE(GEN6_PMINTRMSK, 0xffffffff);
-	I915_WRITE(GEN6_PMIER, I915_READ(GEN6_PMIER) &
-				~dev_priv->pm_rps_events);
-	/* Complete PM interrupt masking here doesn't race with the rps work
-	 * item again unmasking PM interrupts because that is using a different
-	 * register (PMIMR) to mask PM interrupts. The only risk is in leaving
-	 * stale bits in PMIIR and PMIMR which gen6_enable_rps will clean up. */
-
-	spin_lock_irq(&dev_priv->irq_lock);
-	dev_priv->rps.pm_iir = 0;
-	spin_unlock_irq(&dev_priv->irq_lock);
-
-	I915_WRITE(GEN6_PMIIR, dev_priv->pm_rps_events);
+	I915_WRITE(GEN6_RC_CONTROL, 0);
 }
 
 static void gen6_disable_rps(struct drm_device *dev)
@@ -3386,11 +4518,6 @@ static void gen6_disable_rps(struct drm_device *dev)
 
 	I915_WRITE(GEN6_RC_CONTROL, 0);
 	I915_WRITE(GEN6_RPNSWREQ, 1 << 31);
-
-	if (IS_BROADWELL(dev))
-		gen8_disable_rps_interrupts(dev);
-	else
-		gen6_disable_rps_interrupts(dev);
 }
 
 static void cherryview_disable_rps(struct drm_device *dev)
@@ -3398,17 +4525,19 @@ static void cherryview_disable_rps(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	I915_WRITE(GEN6_RC_CONTROL, 0);
-
-	gen8_disable_rps_interrupts(dev);
 }
 
 static void valleyview_disable_rps(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
+	/* we're doing forcewake before Disabling RC6,
+	 * This what the BIOS expects when going into suspend */
+	gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL);
+
 	I915_WRITE(GEN6_RC_CONTROL, 0);
 
-	gen6_disable_rps_interrupts(dev);
+	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 }
 
 static void intel_print_rc6_info(struct drm_device *dev, u32 mode)
@@ -3419,10 +4548,15 @@ static void intel_print_rc6_info(struct drm_device *dev, u32 mode)
 		else
 			mode = 0;
 	}
-	DRM_DEBUG_KMS("Enabling RC6 states: RC6 %s, RC6p %s, RC6pp %s\n",
-		      (mode & GEN6_RC_CTL_RC6_ENABLE) ? "on" : "off",
-		      (mode & GEN6_RC_CTL_RC6p_ENABLE) ? "on" : "off",
-		      (mode & GEN6_RC_CTL_RC6pp_ENABLE) ? "on" : "off");
+	if (HAS_RC6p(dev))
+		DRM_DEBUG_KMS("Enabling RC6 states: RC6 %s RC6p %s RC6pp %s\n",
+			      (mode & GEN6_RC_CTL_RC6_ENABLE) ? "on" : "off",
+			      (mode & GEN6_RC_CTL_RC6p_ENABLE) ? "on" : "off",
+			      (mode & GEN6_RC_CTL_RC6pp_ENABLE) ? "on" : "off");
+
+	else
+		DRM_DEBUG_KMS("Enabling RC6 states: RC6 %s\n",
+			      (mode & GEN6_RC_CTL_RC6_ENABLE) ? "on" : "off");
 }
 
 static int sanitize_rc6_option(const struct drm_device *dev, int enable_rc6)
@@ -3439,7 +4573,7 @@ static int sanitize_rc6_option(const struct drm_device *dev, int enable_rc6)
 	if (enable_rc6 >= 0) {
 		int mask;
 
-		if (INTEL_INFO(dev)->gen == 6 || IS_IVYBRIDGE(dev))
+		if (HAS_RC6p(dev))
 			mask = INTEL_RC6_ENABLE | INTEL_RC6p_ENABLE |
 			       INTEL_RC6pp_ENABLE;
 		else
@@ -3467,54 +4601,92 @@ int intel_enable_rc6(const struct drm_device *dev)
 	return i915.enable_rc6;
 }
 
-static void gen8_enable_rps_interrupts(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	spin_lock_irq(&dev_priv->irq_lock);
-	WARN_ON(dev_priv->rps.pm_iir);
-	gen8_enable_pm_irq(dev_priv, dev_priv->pm_rps_events);
-	I915_WRITE(GEN8_GT_IIR(2), dev_priv->pm_rps_events);
-	spin_unlock_irq(&dev_priv->irq_lock);
-}
-
-static void gen6_enable_rps_interrupts(struct drm_device *dev)
+static void gen6_init_rps_frequencies(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	uint32_t rp_state_cap;
+	u32 ddcc_status = 0;
+	int ret;
 
-	spin_lock_irq(&dev_priv->irq_lock);
-	WARN_ON(dev_priv->rps.pm_iir);
-	gen6_enable_pm_irq(dev_priv, dev_priv->pm_rps_events);
-	I915_WRITE(GEN6_PMIIR, dev_priv->pm_rps_events);
-	spin_unlock_irq(&dev_priv->irq_lock);
-}
-
-static void parse_rp_state_cap(struct drm_i915_private *dev_priv, u32 rp_state_cap)
-{
+	rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
 	/* All of these values are in units of 50MHz */
 	dev_priv->rps.cur_freq		= 0;
-	/* static values from HW: RP0 < RPe < RP1 < RPn (min_freq) */
-	dev_priv->rps.rp1_freq		= (rp_state_cap >>  8) & 0xff;
+	/* static values from HW: RP0 > RP1 > RPn (min_freq) */
 	dev_priv->rps.rp0_freq		= (rp_state_cap >>  0) & 0xff;
+	dev_priv->rps.rp1_freq		= (rp_state_cap >>  8) & 0xff;
 	dev_priv->rps.min_freq		= (rp_state_cap >> 16) & 0xff;
-	/* XXX: only BYT has a special efficient freq */
-	dev_priv->rps.efficient_freq	= dev_priv->rps.rp1_freq;
 	/* hw_max = RP0 until we check for overclocking */
 	dev_priv->rps.max_freq		= dev_priv->rps.rp0_freq;
 
+	dev_priv->rps.efficient_freq = dev_priv->rps.rp1_freq;
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
+		ret = sandybridge_pcode_read(dev_priv,
+					HSW_PCODE_DYNAMIC_DUTY_CYCLE_CONTROL,
+					&ddcc_status);
+		if (0 == ret)
+			dev_priv->rps.efficient_freq =
+				(ddcc_status >> 8) & 0xff;
+	}
+
 	/* Preserve min/max settings in case of re-init */
 	if (dev_priv->rps.max_freq_softlimit == 0)
 		dev_priv->rps.max_freq_softlimit = dev_priv->rps.max_freq;
 
-	if (dev_priv->rps.min_freq_softlimit == 0)
-		dev_priv->rps.min_freq_softlimit = dev_priv->rps.min_freq;
+	if (dev_priv->rps.min_freq_softlimit == 0) {
+		if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+			dev_priv->rps.min_freq_softlimit =
+				/* max(RPe, 450 MHz) */
+				max(dev_priv->rps.efficient_freq, (u8) 9);
+		else
+			dev_priv->rps.min_freq_softlimit =
+				dev_priv->rps.min_freq;
+	}
+}
+
+static void gen9_enable_rps(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
+	uint32_t rc6_mask = 0;
+	int unused;
+
+	/* 1a: Software RC state - RC0 */
+	I915_WRITE(GEN6_RC_STATE, 0);
+
+	/* 1b: Get forcewake during program sequence. Although the driver
+	 * hasn't enabled a state yet where we need forcewake, BIOS may have.*/
+	gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL);
+
+	/* 2a: Disable RC states. */
+	I915_WRITE(GEN6_RC_CONTROL, 0);
+
+	/* 2b: Program RC6 thresholds.*/
+	I915_WRITE(GEN6_RC6_WAKE_RATE_LIMIT, 54 << 16);
+	I915_WRITE(GEN6_RC_EVALUATION_INTERVAL, 125000); /* 12500 * 1280ns */
+	I915_WRITE(GEN6_RC_IDLE_HYSTERSIS, 25); /* 25 * 1280ns */
+	for_each_ring(ring, dev_priv, unused)
+		I915_WRITE(RING_MAX_IDLE(ring->mmio_base), 10);
+	I915_WRITE(GEN6_RC_SLEEP, 0);
+	I915_WRITE(GEN6_RC6_THRESHOLD, 37500); /* 37.5/125ms per EI */
+
+	/* 3a: Enable RC6 */
+	if (intel_enable_rc6(dev) & INTEL_RC6_ENABLE)
+		rc6_mask = GEN6_RC_CTL_RC6_ENABLE;
+	DRM_INFO("RC6 %s\n", (rc6_mask & GEN6_RC_CTL_RC6_ENABLE) ?
+			"on" : "off");
+	I915_WRITE(GEN6_RC_CONTROL, GEN6_RC_CTL_HW_ENABLE |
+				   GEN6_RC_CTL_EI_MODE(1) |
+				   rc6_mask);
+
+	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
+
 }
 
 static void gen8_enable_rps(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_engine_cs *ring;
-	uint32_t rc6_mask = 0, rp_state_cap;
+	uint32_t rc6_mask = 0;
 	int unused;
 
 	/* 1a: Software RC state - RC0 */
@@ -3527,8 +4699,8 @@ static void gen8_enable_rps(struct drm_device *dev)
 	/* 2a: Disable RC states. */
 	I915_WRITE(GEN6_RC_CONTROL, 0);
 
-	rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
-	parse_rp_state_cap(dev_priv, rp_state_cap);
+	/* Initialize rps frequencies */
+	gen6_init_rps_frequencies(dev);
 
 	/* 2b: Program RC6 thresholds.*/
 	I915_WRITE(GEN6_RC6_WAKE_RATE_LIMIT, 40 << 16);
@@ -3586,9 +4758,8 @@ static void gen8_enable_rps(struct drm_device *dev)
 
 	/* 6: Ring frequency + overclocking (our driver does this later */
 
-	gen6_set_rps(dev, (I915_READ(GEN6_GT_PERF_STATUS) & 0xff00) >> 8);
-
-	gen8_enable_rps_interrupts(dev);
+	dev_priv->rps.power = HIGH_POWER; /* force a reset */
+	gen6_set_rps(dev_priv->dev, dev_priv->rps.min_freq_softlimit);
 
 	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 }
@@ -3597,8 +4768,6 @@ static void gen6_enable_rps(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_engine_cs *ring;
-	u32 rp_state_cap;
-	u32 gt_perf_status;
 	u32 rc6vids, pcu_mbox = 0, rc6_mask = 0;
 	u32 gtfifodbg;
 	int rc6_mode;
@@ -3622,10 +4791,8 @@ static void gen6_enable_rps(struct drm_device *dev)
 
 	gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL);
 
-	rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
-	gt_perf_status = I915_READ(GEN6_GT_PERF_STATUS);
-
-	parse_rp_state_cap(dev_priv, rp_state_cap);
+	/* Initialize rps frequencies */
+	gen6_init_rps_frequencies(dev);
 
 	/* disable the counters and set deterministic thresholds */
 	I915_WRITE(GEN6_RC_CONTROL, 0);
@@ -3688,8 +4855,6 @@ static void gen6_enable_rps(struct drm_device *dev)
 	dev_priv->rps.power = HIGH_POWER; /* force a reset */
 	gen6_set_rps(dev_priv->dev, dev_priv->rps.min_freq_softlimit);
 
-	gen6_enable_rps_interrupts(dev);
-
 	rc6vids = 0;
 	ret = sandybridge_pcode_read(dev_priv, GEN6_PCODE_READ_RC6VIDS, &rc6vids);
 	if (IS_GEN6(dev) && ret) {
@@ -3742,9 +4907,9 @@ static void __gen6_update_ring_freq(struct drm_device *dev)
 	 * to use for memory access.  We do this by specifying the IA frequency
 	 * the PCU should use as a reference to determine the ring frequency.
 	 */
-	for (gpu_freq = dev_priv->rps.max_freq_softlimit; gpu_freq >= dev_priv->rps.min_freq_softlimit;
+	for (gpu_freq = dev_priv->rps.max_freq; gpu_freq >= dev_priv->rps.min_freq;
 	     gpu_freq--) {
-		int diff = dev_priv->rps.max_freq_softlimit - gpu_freq;
+		int diff = dev_priv->rps.max_freq - gpu_freq;
 		unsigned int ia_freq = 0, ring_freq = 0;
 
 		if (INTEL_INFO(dev)->gen >= 8) {
@@ -3899,12 +5064,15 @@ static void cherryview_setup_pctx(struct drm_device *dev)
 
 	pcbr = I915_READ(VLV_PCBR);
 	if ((pcbr >> VLV_PCBR_ADDR_SHIFT) == 0) {
+		DRM_DEBUG_DRIVER("BIOS didn't set up PCBR, fixing up\n");
 		paddr = (dev_priv->mm.stolen_base +
 			 (gtt->stolen_size - pctx_size));
 
 		pctx_paddr = (paddr & (~4095));
 		I915_WRITE(VLV_PCBR, pctx_paddr);
 	}
+
+	DRM_DEBUG_DRIVER("PCBR: 0x%08x\n", I915_READ(VLV_PCBR));
 }
 
 static void valleyview_setup_pctx(struct drm_device *dev)
@@ -3930,6 +5098,8 @@ static void valleyview_setup_pctx(struct drm_device *dev)
 		goto out;
 	}
 
+	DRM_DEBUG_DRIVER("BIOS didn't set up PCBR, fixing up\n");
+
 	/*
 	 * From the Gunit register HAS:
 	 * The Gfx driver is expected to program this register and ensure
@@ -3948,6 +5118,7 @@ static void valleyview_setup_pctx(struct drm_device *dev)
 	I915_WRITE(VLV_PCBR, pctx_paddr);
 
 out:
+	DRM_DEBUG_DRIVER("PCBR: 0x%08x\n", I915_READ(VLV_PCBR));
 	dev_priv->vlv_pctx = pctx;
 }
 
@@ -3965,11 +5136,27 @@ static void valleyview_cleanup_pctx(struct drm_device *dev)
 static void valleyview_init_gt_powersave(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 val;
 
 	valleyview_setup_pctx(dev);
 
 	mutex_lock(&dev_priv->rps.hw_lock);
 
+	val = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS);
+	switch ((val >> 6) & 3) {
+	case 0:
+	case 1:
+		dev_priv->mem_freq = 800;
+		break;
+	case 2:
+		dev_priv->mem_freq = 1066;
+		break;
+	case 3:
+		dev_priv->mem_freq = 1333;
+		break;
+	}
+	DRM_DEBUG_DRIVER("DDR speed: %d MHz\n", dev_priv->mem_freq);
+
 	dev_priv->rps.max_freq = valleyview_rps_max_freq(dev_priv);
 	dev_priv->rps.rp0_freq = dev_priv->rps.max_freq;
 	DRM_DEBUG_DRIVER("max GPU freq: %d MHz (%u)\n",
@@ -4004,11 +5191,41 @@ static void valleyview_init_gt_powersave(struct drm_device *dev)
 static void cherryview_init_gt_powersave(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 val;
 
 	cherryview_setup_pctx(dev);
 
 	mutex_lock(&dev_priv->rps.hw_lock);
 
+	mutex_lock(&dev_priv->dpio_lock);
+	val = vlv_cck_read(dev_priv, CCK_FUSE_REG);
+	mutex_unlock(&dev_priv->dpio_lock);
+
+	switch ((val >> 2) & 0x7) {
+	case 0:
+	case 1:
+		dev_priv->rps.cz_freq = 200;
+		dev_priv->mem_freq = 1600;
+		break;
+	case 2:
+		dev_priv->rps.cz_freq = 267;
+		dev_priv->mem_freq = 1600;
+		break;
+	case 3:
+		dev_priv->rps.cz_freq = 333;
+		dev_priv->mem_freq = 2000;
+		break;
+	case 4:
+		dev_priv->rps.cz_freq = 320;
+		dev_priv->mem_freq = 1600;
+		break;
+	case 5:
+		dev_priv->rps.cz_freq = 400;
+		dev_priv->mem_freq = 1600;
+		break;
+	}
+	DRM_DEBUG_DRIVER("DDR speed: %d MHz\n", dev_priv->mem_freq);
+
 	dev_priv->rps.max_freq = cherryview_rps_max_freq(dev_priv);
 	dev_priv->rps.rp0_freq = dev_priv->rps.max_freq;
 	DRM_DEBUG_DRIVER("max GPU freq: %d MHz (%u)\n",
@@ -4030,6 +5247,12 @@ static void cherryview_init_gt_powersave(struct drm_device *dev)
 			 vlv_gpu_freq(dev_priv, dev_priv->rps.min_freq),
 			 dev_priv->rps.min_freq);
 
+	WARN_ONCE((dev_priv->rps.max_freq |
+		   dev_priv->rps.efficient_freq |
+		   dev_priv->rps.rp1_freq |
+		   dev_priv->rps.min_freq) & 1,
+		  "Odd GPU freq values\n");
+
 	/* Preserve min/max settings in case of re-init */
 	if (dev_priv->rps.max_freq_softlimit == 0)
 		dev_priv->rps.max_freq_softlimit = dev_priv->rps.max_freq;
@@ -4087,8 +5310,6 @@ static void cherryview_enable_rps(struct drm_device *dev)
 	/* For now we assume BIOS is allocating and populating the PCBR  */
 	pcbr = I915_READ(VLV_PCBR);
 
-	DRM_DEBUG_DRIVER("PCBR offset : 0x%x\n", pcbr);
-
 	/* 3: Enable RC6 */
 	if ((intel_enable_rc6(dev) & INTEL_RC6_ENABLE) &&
 						(pcbr >> VLV_PCBR_ADDR_SHIFT))
@@ -4118,7 +5339,10 @@ static void cherryview_enable_rps(struct drm_device *dev)
 
 	val = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS);
 
-	DRM_DEBUG_DRIVER("GPLL enabled? %s\n", val & 0x10 ? "yes" : "no");
+	/* RPS code assumes GPLL is used */
+	WARN_ONCE((val & GPLLENABLE) == 0, "GPLL not enabled\n");
+
+	DRM_DEBUG_DRIVER("GPLL enabled? %s\n", val & GPLLENABLE ? "yes" : "no");
 	DRM_DEBUG_DRIVER("GPU status: 0x%08x\n", val);
 
 	dev_priv->rps.cur_freq = (val >> 8) & 0xff;
@@ -4132,8 +5356,6 @@ static void cherryview_enable_rps(struct drm_device *dev)
 
 	valleyview_set_rps(dev_priv->dev, dev_priv->rps.efficient_freq);
 
-	gen8_enable_rps_interrupts(dev);
-
 	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 }
 
@@ -4198,7 +5420,10 @@ static void valleyview_enable_rps(struct drm_device *dev)
 
 	val = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS);
 
-	DRM_DEBUG_DRIVER("GPLL enabled? %s\n", val & 0x10 ? "yes" : "no");
+	/* RPS code assumes GPLL is used */
+	WARN_ONCE((val & GPLLENABLE) == 0, "GPLL not enabled\n");
+
+	DRM_DEBUG_DRIVER("GPLL enabled? %s\n", val & GPLLENABLE ? "yes" : "no");
 	DRM_DEBUG_DRIVER("GPU status: 0x%08x\n", val);
 
 	dev_priv->rps.cur_freq = (val >> 8) & 0xff;
@@ -4212,8 +5437,6 @@ static void valleyview_enable_rps(struct drm_device *dev)
 
 	valleyview_set_rps(dev_priv->dev, dev_priv->rps.efficient_freq);
 
-	gen6_enable_rps_interrupts(dev);
-
 	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 }
 
@@ -4960,6 +6183,20 @@ void intel_cleanup_gt_powersave(struct drm_device *dev)
 		valleyview_cleanup_gt_powersave(dev);
 }
 
+static void gen6_suspend_rps(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	flush_delayed_work(&dev_priv->rps.delayed_resume_work);
+
+	/*
+	 * TODO: disable RPS interrupts on GEN9+ too once RPS support
+	 * is added for it.
+	 */
+	if (INTEL_INFO(dev)->gen < 9)
+		gen6_disable_rps_interrupts(dev);
+}
+
 /**
  * intel_suspend_gt_powersave - suspend PM work and helper threads
  * @dev: drm device
@@ -4972,12 +6209,10 @@ void intel_suspend_gt_powersave(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	/* Interrupts should be disabled already to avoid re-arming. */
-	WARN_ON(intel_irqs_enabled(dev_priv));
-
-	flush_delayed_work(&dev_priv->rps.delayed_resume_work);
+	if (INTEL_INFO(dev)->gen < 6)
+		return;
 
-	cancel_work_sync(&dev_priv->rps.work);
+	gen6_suspend_rps(dev);
 
 	/* Force GPU to min freq during suspend */
 	gen6_rps_idle(dev_priv);
@@ -4987,9 +6222,6 @@ void intel_disable_gt_powersave(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	/* Interrupts should be disabled already to avoid re-arming. */
-	WARN_ON(intel_irqs_enabled(dev_priv));
-
 	if (IS_IRONLAKE_M(dev)) {
 		ironlake_disable_drps(dev);
 		ironlake_disable_rc6(dev);
@@ -4997,12 +6229,15 @@ void intel_disable_gt_powersave(struct drm_device *dev)
 		intel_suspend_gt_powersave(dev);
 
 		mutex_lock(&dev_priv->rps.hw_lock);
-		if (IS_CHERRYVIEW(dev))
+		if (INTEL_INFO(dev)->gen >= 9)
+			gen9_disable_rps(dev);
+		else if (IS_CHERRYVIEW(dev))
 			cherryview_disable_rps(dev);
 		else if (IS_VALLEYVIEW(dev))
 			valleyview_disable_rps(dev);
 		else
 			gen6_disable_rps(dev);
+
 		dev_priv->rps.enabled = false;
 		mutex_unlock(&dev_priv->rps.hw_lock);
 	}
@@ -5017,10 +6252,19 @@ static void intel_gen6_powersave_work(struct work_struct *work)
 
 	mutex_lock(&dev_priv->rps.hw_lock);
 
+	/*
+	 * TODO: reset/enable RPS interrupts on GEN9+ too, once RPS support is
+	 * added for it.
+	 */
+	if (INTEL_INFO(dev)->gen < 9)
+		gen6_reset_rps_interrupts(dev);
+
 	if (IS_CHERRYVIEW(dev)) {
 		cherryview_enable_rps(dev);
 	} else if (IS_VALLEYVIEW(dev)) {
 		valleyview_enable_rps(dev);
+	} else if (INTEL_INFO(dev)->gen >= 9) {
+		gen9_enable_rps(dev);
 	} else if (IS_BROADWELL(dev)) {
 		gen8_enable_rps(dev);
 		__gen6_update_ring_freq(dev);
@@ -5029,6 +6273,10 @@ static void intel_gen6_powersave_work(struct work_struct *work)
 		__gen6_update_ring_freq(dev);
 	}
 	dev_priv->rps.enabled = true;
+
+	if (INTEL_INFO(dev)->gen < 9)
+		gen6_enable_rps_interrupts(dev);
+
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	intel_runtime_pm_put(dev_priv);
@@ -5067,8 +6315,11 @@ void intel_reset_gt_powersave(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
+	if (INTEL_INFO(dev)->gen < 6)
+		return;
+
+	gen6_suspend_rps(dev);
 	dev_priv->rps.enabled = false;
-	intel_enable_gt_powersave(dev);
 }
 
 static void ibx_init_clock_gating(struct drm_device *dev)
@@ -5088,7 +6339,7 @@ static void g4x_disable_trickle_feed(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe;
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		I915_WRITE(DSPCNTR(pipe),
 			   I915_READ(DSPCNTR(pipe)) |
 			   DISPPLANE_TRICKLE_FEED_DISABLE);
@@ -5203,7 +6454,7 @@ static void cpt_init_clock_gating(struct drm_device *dev)
 	/* The below fixes the weird display corruption, a few pixels shifted
 	 * downward, on (only) LVDS of some HP laptops with IVY.
 	 */
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		val = I915_READ(TRANS_CHICKEN2(pipe));
 		val |= TRANS_CHICKEN2_TIMING_OVERRIDE;
 		val &= ~TRANS_CHICKEN2_FDI_POLARITY_REVERSED;
@@ -5215,7 +6466,7 @@ static void cpt_init_clock_gating(struct drm_device *dev)
 		I915_WRITE(TRANS_CHICKEN2(pipe), val);
 	}
 	/* WADP0ClockGatingDisable */
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		I915_WRITE(TRANS_CHICKEN1(pipe),
 			   TRANS_CHICKEN1_DP0UNIT_GC_DISABLE);
 	}
@@ -5247,11 +6498,6 @@ static void gen6_init_clock_gating(struct drm_device *dev)
 	I915_WRITE(_3D_CHICKEN,
 		   _MASKED_BIT_ENABLE(_3D_CHICKEN_HIZ_PLANE_DISABLE_MSAA_4X_SNB));
 
-	/* WaSetupGtModeTdRowDispatch:snb */
-	if (IS_SNB_GT1(dev))
-		I915_WRITE(GEN6_GT_MODE,
-			   _MASKED_BIT_ENABLE(GEN6_TD_FOUR_ROW_DISPATCH_DISABLE));
-
 	/* WaDisable_RenderCache_OperationalFlush:snb */
 	I915_WRITE(CACHE_MODE_0, _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE));
 
@@ -5264,7 +6510,7 @@ static void gen6_init_clock_gating(struct drm_device *dev)
 	 * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM).
 	 */
 	I915_WRITE(GEN6_GT_MODE,
-		   GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4);
+		   _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4));
 
 	ilk_init_lp_watermarks(dev);
 
@@ -5383,7 +6629,7 @@ static void lpt_suspend_hw(struct drm_device *dev)
 	}
 }
 
-static void gen8_init_clock_gating(struct drm_device *dev)
+static void broadwell_init_clock_gating(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	enum pipe pipe;
@@ -5392,41 +6638,6 @@ static void gen8_init_clock_gating(struct drm_device *dev)
 	I915_WRITE(WM2_LP_ILK, 0);
 	I915_WRITE(WM1_LP_ILK, 0);
 
-	/* FIXME(BDW): Check all the w/a, some might only apply to
-	 * pre-production hw. */
-
-	/* WaDisablePartialInstShootdown:bdw */
-	I915_WRITE(GEN8_ROW_CHICKEN,
-		   _MASKED_BIT_ENABLE(PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE));
-
-	/* WaDisableThreadStallDopClockGating:bdw */
-	/* FIXME: Unclear whether we really need this on production bdw. */
-	I915_WRITE(GEN8_ROW_CHICKEN,
-		   _MASKED_BIT_ENABLE(STALL_DOP_GATING_DISABLE));
-
-	/*
-	 * This GEN8_CENTROID_PIXEL_OPT_DIS W/A is only needed for
-	 * pre-production hardware
-	 */
-	I915_WRITE(HALF_SLICE_CHICKEN3,
-		   _MASKED_BIT_ENABLE(GEN8_CENTROID_PIXEL_OPT_DIS));
-	I915_WRITE(HALF_SLICE_CHICKEN3,
-		   _MASKED_BIT_ENABLE(GEN8_SAMPLER_POWER_BYPASS_DIS));
-	I915_WRITE(GAMTARBMODE, _MASKED_BIT_ENABLE(ARB_MODE_BWGTLB_DISABLE));
-
-	I915_WRITE(_3D_CHICKEN3,
-		   _MASKED_BIT_ENABLE(_3D_CHICKEN_SDE_LIMIT_FIFO_POLY_DEPTH(2)));
-
-	I915_WRITE(COMMON_SLICE_CHICKEN2,
-		   _MASKED_BIT_ENABLE(GEN8_CSC2_SBE_VUE_CACHE_CONSERVATIVE));
-
-	I915_WRITE(GEN7_HALF_SLICE_CHICKEN1,
-		   _MASKED_BIT_ENABLE(GEN7_SINGLE_SUBSCAN_DISPATCH_ENABLE));
-
-	/* WaDisableDopClockGating:bdw May not be needed for production */
-	I915_WRITE(GEN7_ROW_CHICKEN2,
-		   _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE));
-
 	/* WaSwitchSolVfFArbitrationPriority:bdw */
 	I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | HSW_ECOCHK_ARB_PRIO_SOL);
 
@@ -5435,37 +6646,18 @@ static void gen8_init_clock_gating(struct drm_device *dev)
 		   I915_READ(CHICKEN_PAR1_1) | DPA_MASK_VBLANK_SRD);
 
 	/* WaPsrDPRSUnmaskVBlankInSRD:bdw */
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		I915_WRITE(CHICKEN_PIPESL_1(pipe),
 			   I915_READ(CHICKEN_PIPESL_1(pipe)) |
 			   BDW_DPRS_MASK_VBLANK_SRD);
 	}
 
-	/* Use Force Non-Coherent whenever executing a 3D context. This is a
-	 * workaround for for a possible hang in the unlikely event a TLB
-	 * invalidation occurs during a PSD flush.
-	 */
-	I915_WRITE(HDC_CHICKEN0,
-		   I915_READ(HDC_CHICKEN0) |
-		   _MASKED_BIT_ENABLE(HDC_FORCE_NON_COHERENT));
-
 	/* WaVSRefCountFullforceMissDisable:bdw */
 	/* WaDSRefCountFullforceMissDisable:bdw */
 	I915_WRITE(GEN7_FF_THREAD_MODE,
 		   I915_READ(GEN7_FF_THREAD_MODE) &
 		   ~(GEN8_FF_DS_REF_CNT_FFME | GEN7_FF_VS_REF_CNT_FFME));
 
-	/*
-	 * BSpec recommends 8x4 when MSAA is used,
-	 * however in practice 16x4 seems fastest.
-	 *
-	 * Note that PS/WM thread counts depend on the WIZ hashing
-	 * disable bit, which we don't touch here, but it's good
-	 * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM).
-	 */
-	I915_WRITE(GEN7_GT_MODE,
-		   GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4);
-
 	I915_WRITE(GEN6_RC_SLEEP_PSMI_CONTROL,
 		   _MASKED_BIT_ENABLE(GEN8_RC_SEMA_IDLE_MSG_DISABLE));
 
@@ -5473,9 +6665,7 @@ static void gen8_init_clock_gating(struct drm_device *dev)
 	I915_WRITE(GEN8_UCGCTL6, I915_READ(GEN8_UCGCTL6) |
 		   GEN8_SDEUNIT_CLOCK_GATE_DISABLE);
 
-	/* Wa4x4STCOptimizationDisable:bdw */
-	I915_WRITE(CACHE_MODE_1,
-		   _MASKED_BIT_ENABLE(GEN8_4x4_STC_OPTIMIZATION_DISABLE));
+	lpt_init_clock_gating(dev);
 }
 
 static void haswell_init_clock_gating(struct drm_device *dev)
@@ -5518,7 +6708,7 @@ static void haswell_init_clock_gating(struct drm_device *dev)
 	 * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM).
 	 */
 	I915_WRITE(GEN7_GT_MODE,
-		   GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4);
+		   _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4));
 
 	/* WaSwitchSolVfFArbitrationPriority:hsw */
 	I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | HSW_ECOCHK_ARB_PRIO_SOL);
@@ -5615,7 +6805,7 @@ static void ivybridge_init_clock_gating(struct drm_device *dev)
 	 * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM).
 	 */
 	I915_WRITE(GEN7_GT_MODE,
-		   GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4);
+		   _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4));
 
 	snpcr = I915_READ(GEN6_MBCUNIT_SNPCR);
 	snpcr &= ~GEN6_MBC_SNPCR_MASK;
@@ -5631,24 +6821,6 @@ static void ivybridge_init_clock_gating(struct drm_device *dev)
 static void valleyview_init_clock_gating(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 val;
-
-	mutex_lock(&dev_priv->rps.hw_lock);
-	val = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS);
-	mutex_unlock(&dev_priv->rps.hw_lock);
-	switch ((val >> 6) & 3) {
-	case 0:
-	case 1:
-		dev_priv->mem_freq = 800;
-		break;
-	case 2:
-		dev_priv->mem_freq = 1066;
-		break;
-	case 3:
-		dev_priv->mem_freq = 1333;
-		break;
-	}
-	DRM_DEBUG_DRIVER("DDR speed: %d MHz", dev_priv->mem_freq);
 
 	I915_WRITE(DSPCLK_GATE_D, VRHUNIT_CLOCK_GATE_DISABLE);
 
@@ -5724,48 +6896,11 @@ static void valleyview_init_clock_gating(struct drm_device *dev)
 static void cherryview_init_clock_gating(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 val;
-
-	mutex_lock(&dev_priv->rps.hw_lock);
-	val = vlv_punit_read(dev_priv, CCK_FUSE_REG);
-	mutex_unlock(&dev_priv->rps.hw_lock);
-	switch ((val >> 2) & 0x7) {
-	case 0:
-	case 1:
-			dev_priv->rps.cz_freq = CHV_CZ_CLOCK_FREQ_MODE_200;
-			dev_priv->mem_freq = 1600;
-			break;
-	case 2:
-			dev_priv->rps.cz_freq = CHV_CZ_CLOCK_FREQ_MODE_267;
-			dev_priv->mem_freq = 1600;
-			break;
-	case 3:
-			dev_priv->rps.cz_freq = CHV_CZ_CLOCK_FREQ_MODE_333;
-			dev_priv->mem_freq = 2000;
-			break;
-	case 4:
-			dev_priv->rps.cz_freq = CHV_CZ_CLOCK_FREQ_MODE_320;
-			dev_priv->mem_freq = 1600;
-			break;
-	case 5:
-			dev_priv->rps.cz_freq = CHV_CZ_CLOCK_FREQ_MODE_400;
-			dev_priv->mem_freq = 1600;
-			break;
-	}
-	DRM_DEBUG_DRIVER("DDR speed: %d MHz", dev_priv->mem_freq);
 
 	I915_WRITE(DSPCLK_GATE_D, VRHUNIT_CLOCK_GATE_DISABLE);
 
 	I915_WRITE(MI_ARB_VLV, MI_ARB_DISPLAY_TRICKLE_FEED_DISABLE);
 
-	/* WaDisablePartialInstShootdown:chv */
-	I915_WRITE(GEN8_ROW_CHICKEN,
-		   _MASKED_BIT_ENABLE(PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE));
-
-	/* WaDisableThreadStallDopClockGating:chv */
-	I915_WRITE(GEN8_ROW_CHICKEN,
-		   _MASKED_BIT_ENABLE(STALL_DOP_GATING_DISABLE));
-
 	/* WaVSRefCountFullforceMissDisable:chv */
 	/* WaDSRefCountFullforceMissDisable:chv */
 	I915_WRITE(GEN7_FF_THREAD_MODE,
@@ -5783,24 +6918,6 @@ static void cherryview_init_clock_gating(struct drm_device *dev)
 	/* WaDisableSDEUnitClockGating:chv */
 	I915_WRITE(GEN8_UCGCTL6, I915_READ(GEN8_UCGCTL6) |
 		   GEN8_SDEUNIT_CLOCK_GATE_DISABLE);
-
-	/* WaDisableSamplerPowerBypass:chv (pre-production hw) */
-	I915_WRITE(HALF_SLICE_CHICKEN3,
-		   _MASKED_BIT_ENABLE(GEN8_SAMPLER_POWER_BYPASS_DIS));
-
-	/* WaDisableGunitClockGating:chv (pre-production hw) */
-	I915_WRITE(VLV_GUNIT_CLOCK_GATE, I915_READ(VLV_GUNIT_CLOCK_GATE) |
-		   GINT_DIS);
-
-	/* WaDisableFfDopClockGating:chv (pre-production hw) */
-	I915_WRITE(GEN6_RC_SLEEP_PSMI_CONTROL,
-		   _MASKED_BIT_ENABLE(GEN8_FF_DOP_CLOCK_GATE_DISABLE));
-
-	/* WaDisableDopClockGating:chv (pre-production hw) */
-	I915_WRITE(GEN7_ROW_CHICKEN2,
-		   _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE));
-	I915_WRITE(GEN6_UCGCTL1, I915_READ(GEN6_UCGCTL1) |
-		   GEN6_EU_TCUNIT_CLOCK_GATE_DISABLE);
 }
 
 static void g4x_init_clock_gating(struct drm_device *dev)
@@ -5883,6 +7000,9 @@ static void gen3_init_clock_gating(struct drm_device *dev)
 
 	/* On GEN3 we really need to make sure the ARB C3 LP bit is set */
 	I915_WRITE(MI_ARB_STATE, _MASKED_BIT_ENABLE(MI_ARB_C3_LP_WRITE_ENABLE));
+
+	I915_WRITE(MI_ARB_STATE,
+		   _MASKED_BIT_ENABLE(MI_ARB_DISPLAY_TRICKLE_FEED_DISABLE));
 }
 
 static void i85x_init_clock_gating(struct drm_device *dev)
@@ -5894,6 +7014,9 @@ static void i85x_init_clock_gating(struct drm_device *dev)
 	/* interrupts should cause a wake up from C3 */
 	I915_WRITE(MI_STATE, _MASKED_BIT_ENABLE(MI_AGPBUSY_INT_EN) |
 		   _MASKED_BIT_DISABLE(MI_AGPBUSY_830_MODE));
+
+	I915_WRITE(MEM_MODE,
+		   _MASKED_BIT_ENABLE(MEM_DISPLAY_TRICKLE_FEED_DISABLE));
 }
 
 static void i830_init_clock_gating(struct drm_device *dev)
@@ -5901,6 +7024,10 @@ static void i830_init_clock_gating(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	I915_WRITE(DSPCLK_GATE_D, OVRUNIT_CLOCK_GATE_DISABLE);
+
+	I915_WRITE(MEM_MODE,
+		   _MASKED_BIT_ENABLE(MEM_DISPLAY_A_TRICKLE_FEED_DISABLE) |
+		   _MASKED_BIT_ENABLE(MEM_DISPLAY_B_TRICKLE_FEED_DISABLE));
 }
 
 void intel_init_clock_gating(struct drm_device *dev)
@@ -5916,863 +7043,35 @@ void intel_suspend_hw(struct drm_device *dev)
 		lpt_suspend_hw(dev);
 }
 
-#define for_each_power_well(i, power_well, domain_mask, power_domains)	\
-	for (i = 0;							\
-	     i < (power_domains)->power_well_count &&			\
-		 ((power_well) = &(power_domains)->power_wells[i]);	\
-	     i++)							\
-		if ((power_well)->domains & (domain_mask))
-
-#define for_each_power_well_rev(i, power_well, domain_mask, power_domains) \
-	for (i = (power_domains)->power_well_count - 1;			 \
-	     i >= 0 && ((power_well) = &(power_domains)->power_wells[i]);\
-	     i--)							 \
-		if ((power_well)->domains & (domain_mask))
-
-/**
- * We should only use the power well if we explicitly asked the hardware to
- * enable it, so check if it's enabled and also check if we've requested it to
- * be enabled.
- */
-static bool hsw_power_well_enabled(struct drm_i915_private *dev_priv,
-				   struct i915_power_well *power_well)
-{
-	return I915_READ(HSW_PWR_WELL_DRIVER) ==
-		     (HSW_PWR_WELL_ENABLE_REQUEST | HSW_PWR_WELL_STATE_ENABLED);
-}
-
-bool intel_display_power_enabled_unlocked(struct drm_i915_private *dev_priv,
-					  enum intel_display_power_domain domain)
-{
-	struct i915_power_domains *power_domains;
-	struct i915_power_well *power_well;
-	bool is_enabled;
-	int i;
-
-	if (dev_priv->pm.suspended)
-		return false;
-
-	power_domains = &dev_priv->power_domains;
-
-	is_enabled = true;
-
-	for_each_power_well_rev(i, power_well, BIT(domain), power_domains) {
-		if (power_well->always_on)
-			continue;
-
-		if (!power_well->hw_enabled) {
-			is_enabled = false;
-			break;
-		}
-	}
-
-	return is_enabled;
-}
-
-bool intel_display_power_enabled(struct drm_i915_private *dev_priv,
-				 enum intel_display_power_domain domain)
-{
-	struct i915_power_domains *power_domains;
-	bool ret;
-
-	power_domains = &dev_priv->power_domains;
-
-	mutex_lock(&power_domains->lock);
-	ret = intel_display_power_enabled_unlocked(dev_priv, domain);
-	mutex_unlock(&power_domains->lock);
-
-	return ret;
-}
-
-/*
- * Starting with Haswell, we have a "Power Down Well" that can be turned off
- * when not needed anymore. We have 4 registers that can request the power well
- * to be enabled, and it will only be disabled if none of the registers is
- * requesting it to be enabled.
- */
-static void hsw_power_well_post_enable(struct drm_i915_private *dev_priv)
-{
-	struct drm_device *dev = dev_priv->dev;
-
-	/*
-	 * After we re-enable the power well, if we touch VGA register 0x3d5
-	 * we'll get unclaimed register interrupts. This stops after we write
-	 * anything to the VGA MSR register. The vgacon module uses this
-	 * register all the time, so if we unbind our driver and, as a
-	 * consequence, bind vgacon, we'll get stuck in an infinite loop at
-	 * console_unlock(). So make here we touch the VGA MSR register, making
-	 * sure vgacon can keep working normally without triggering interrupts
-	 * and error messages.
-	 */
-	vga_get_uninterruptible(dev->pdev, VGA_RSRC_LEGACY_IO);
-	outb(inb(VGA_MSR_READ), VGA_MSR_WRITE);
-	vga_put(dev->pdev, VGA_RSRC_LEGACY_IO);
-
-	if (IS_BROADWELL(dev))
-		gen8_irq_power_well_post_enable(dev_priv);
-}
-
-static void hsw_set_power_well(struct drm_i915_private *dev_priv,
-			       struct i915_power_well *power_well, bool enable)
-{
-	bool is_enabled, enable_requested;
-	uint32_t tmp;
-
-	tmp = I915_READ(HSW_PWR_WELL_DRIVER);
-	is_enabled = tmp & HSW_PWR_WELL_STATE_ENABLED;
-	enable_requested = tmp & HSW_PWR_WELL_ENABLE_REQUEST;
-
-	if (enable) {
-		if (!enable_requested)
-			I915_WRITE(HSW_PWR_WELL_DRIVER,
-				   HSW_PWR_WELL_ENABLE_REQUEST);
-
-		if (!is_enabled) {
-			DRM_DEBUG_KMS("Enabling power well\n");
-			if (wait_for((I915_READ(HSW_PWR_WELL_DRIVER) &
-				      HSW_PWR_WELL_STATE_ENABLED), 20))
-				DRM_ERROR("Timeout enabling power well\n");
-		}
-
-		hsw_power_well_post_enable(dev_priv);
-	} else {
-		if (enable_requested) {
-			I915_WRITE(HSW_PWR_WELL_DRIVER, 0);
-			POSTING_READ(HSW_PWR_WELL_DRIVER);
-			DRM_DEBUG_KMS("Requesting to disable the power well\n");
-		}
-	}
-}
-
-static void hsw_power_well_sync_hw(struct drm_i915_private *dev_priv,
-				   struct i915_power_well *power_well)
-{
-	hsw_set_power_well(dev_priv, power_well, power_well->count > 0);
-
-	/*
-	 * We're taking over the BIOS, so clear any requests made by it since
-	 * the driver is in charge now.
-	 */
-	if (I915_READ(HSW_PWR_WELL_BIOS) & HSW_PWR_WELL_ENABLE_REQUEST)
-		I915_WRITE(HSW_PWR_WELL_BIOS, 0);
-}
-
-static void hsw_power_well_enable(struct drm_i915_private *dev_priv,
-				  struct i915_power_well *power_well)
-{
-	hsw_set_power_well(dev_priv, power_well, true);
-}
-
-static void hsw_power_well_disable(struct drm_i915_private *dev_priv,
-				   struct i915_power_well *power_well)
-{
-	hsw_set_power_well(dev_priv, power_well, false);
-}
-
-static void i9xx_always_on_power_well_noop(struct drm_i915_private *dev_priv,
-					   struct i915_power_well *power_well)
-{
-}
-
-static bool i9xx_always_on_power_well_enabled(struct drm_i915_private *dev_priv,
-					     struct i915_power_well *power_well)
-{
-	return true;
-}
-
-static void vlv_set_power_well(struct drm_i915_private *dev_priv,
-			       struct i915_power_well *power_well, bool enable)
-{
-	enum punit_power_well power_well_id = power_well->data;
-	u32 mask;
-	u32 state;
-	u32 ctrl;
-
-	mask = PUNIT_PWRGT_MASK(power_well_id);
-	state = enable ? PUNIT_PWRGT_PWR_ON(power_well_id) :
-			 PUNIT_PWRGT_PWR_GATE(power_well_id);
-
-	mutex_lock(&dev_priv->rps.hw_lock);
-
-#define COND \
-	((vlv_punit_read(dev_priv, PUNIT_REG_PWRGT_STATUS) & mask) == state)
-
-	if (COND)
-		goto out;
-
-	ctrl = vlv_punit_read(dev_priv, PUNIT_REG_PWRGT_CTRL);
-	ctrl &= ~mask;
-	ctrl |= state;
-	vlv_punit_write(dev_priv, PUNIT_REG_PWRGT_CTRL, ctrl);
-
-	if (wait_for(COND, 100))
-		DRM_ERROR("timout setting power well state %08x (%08x)\n",
-			  state,
-			  vlv_punit_read(dev_priv, PUNIT_REG_PWRGT_CTRL));
-
-#undef COND
-
-out:
-	mutex_unlock(&dev_priv->rps.hw_lock);
-}
-
-static void vlv_power_well_sync_hw(struct drm_i915_private *dev_priv,
-				   struct i915_power_well *power_well)
-{
-	vlv_set_power_well(dev_priv, power_well, power_well->count > 0);
-}
-
-static void vlv_power_well_enable(struct drm_i915_private *dev_priv,
-				  struct i915_power_well *power_well)
-{
-	vlv_set_power_well(dev_priv, power_well, true);
-}
-
-static void vlv_power_well_disable(struct drm_i915_private *dev_priv,
-				   struct i915_power_well *power_well)
-{
-	vlv_set_power_well(dev_priv, power_well, false);
-}
-
-static bool vlv_power_well_enabled(struct drm_i915_private *dev_priv,
-				   struct i915_power_well *power_well)
-{
-	int power_well_id = power_well->data;
-	bool enabled = false;
-	u32 mask;
-	u32 state;
-	u32 ctrl;
-
-	mask = PUNIT_PWRGT_MASK(power_well_id);
-	ctrl = PUNIT_PWRGT_PWR_ON(power_well_id);
-
-	mutex_lock(&dev_priv->rps.hw_lock);
-
-	state = vlv_punit_read(dev_priv, PUNIT_REG_PWRGT_STATUS) & mask;
-	/*
-	 * We only ever set the power-on and power-gate states, anything
-	 * else is unexpected.
-	 */
-	WARN_ON(state != PUNIT_PWRGT_PWR_ON(power_well_id) &&
-		state != PUNIT_PWRGT_PWR_GATE(power_well_id));
-	if (state == ctrl)
-		enabled = true;
-
-	/*
-	 * A transient state at this point would mean some unexpected party
-	 * is poking at the power controls too.
-	 */
-	ctrl = vlv_punit_read(dev_priv, PUNIT_REG_PWRGT_CTRL) & mask;
-	WARN_ON(ctrl != state);
-
-	mutex_unlock(&dev_priv->rps.hw_lock);
-
-	return enabled;
-}
-
-static void vlv_display_power_well_enable(struct drm_i915_private *dev_priv,
-					  struct i915_power_well *power_well)
-{
-	WARN_ON_ONCE(power_well->data != PUNIT_POWER_WELL_DISP2D);
-
-	vlv_set_power_well(dev_priv, power_well, true);
-
-	spin_lock_irq(&dev_priv->irq_lock);
-	valleyview_enable_display_irqs(dev_priv);
-	spin_unlock_irq(&dev_priv->irq_lock);
-
-	/*
-	 * During driver initialization/resume we can avoid restoring the
-	 * part of the HW/SW state that will be inited anyway explicitly.
-	 */
-	if (dev_priv->power_domains.initializing)
-		return;
-
-	intel_hpd_init(dev_priv->dev);
-
-	i915_redisable_vga_power_on(dev_priv->dev);
-}
-
-static void vlv_display_power_well_disable(struct drm_i915_private *dev_priv,
-					   struct i915_power_well *power_well)
-{
-	WARN_ON_ONCE(power_well->data != PUNIT_POWER_WELL_DISP2D);
-
-	spin_lock_irq(&dev_priv->irq_lock);
-	valleyview_disable_display_irqs(dev_priv);
-	spin_unlock_irq(&dev_priv->irq_lock);
-
-	vlv_set_power_well(dev_priv, power_well, false);
-}
-
-static void vlv_dpio_cmn_power_well_enable(struct drm_i915_private *dev_priv,
-					   struct i915_power_well *power_well)
-{
-	WARN_ON_ONCE(power_well->data != PUNIT_POWER_WELL_DPIO_CMN_BC);
-
-	/*
-	 * Enable the CRI clock source so we can get at the
-	 * display and the reference clock for VGA
-	 * hotplug / manual detection.
-	 */
-	I915_WRITE(DPLL(PIPE_B), I915_READ(DPLL(PIPE_B)) |
-		   DPLL_REFA_CLK_ENABLE_VLV | DPLL_INTEGRATED_CRI_CLK_VLV);
-	udelay(1); /* >10ns for cmnreset, >0ns for sidereset */
-
-	vlv_set_power_well(dev_priv, power_well, true);
-
-	/*
-	 * From VLV2A0_DP_eDP_DPIO_driver_vbios_notes_10.docx -
-	 *  6.	De-assert cmn_reset/side_reset. Same as VLV X0.
-	 *   a.	GUnit 0x2110 bit[0] set to 1 (def 0)
-	 *   b.	The other bits such as sfr settings / modesel may all
-	 *	be set to 0.
-	 *
-	 * This should only be done on init and resume from S3 with
-	 * both PLLs disabled, or we risk losing DPIO and PLL
-	 * synchronization.
-	 */
-	I915_WRITE(DPIO_CTL, I915_READ(DPIO_CTL) | DPIO_CMNRST);
-}
-
-static void vlv_dpio_cmn_power_well_disable(struct drm_i915_private *dev_priv,
-					    struct i915_power_well *power_well)
-{
-	struct drm_device *dev = dev_priv->dev;
-	enum pipe pipe;
-
-	WARN_ON_ONCE(power_well->data != PUNIT_POWER_WELL_DPIO_CMN_BC);
-
-	for_each_pipe(pipe)
-		assert_pll_disabled(dev_priv, pipe);
-
-	/* Assert common reset */
-	I915_WRITE(DPIO_CTL, I915_READ(DPIO_CTL) & ~DPIO_CMNRST);
-
-	vlv_set_power_well(dev_priv, power_well, false);
-}
-
-static void check_power_well_state(struct drm_i915_private *dev_priv,
-				   struct i915_power_well *power_well)
+static void intel_init_fbc(struct drm_i915_private *dev_priv)
 {
-	bool enabled = power_well->ops->is_enabled(dev_priv, power_well);
-
-	if (power_well->always_on || !i915.disable_power_well) {
-		if (!enabled)
-			goto mismatch;
-
+	if (!HAS_FBC(dev_priv)) {
+		dev_priv->fbc.enabled = false;
 		return;
 	}
 
-	if (enabled != (power_well->count > 0))
-		goto mismatch;
-
-	return;
-
-mismatch:
-	WARN(1, "state mismatch for '%s' (always_on %d hw state %d use-count %d disable_power_well %d\n",
-		  power_well->name, power_well->always_on, enabled,
-		  power_well->count, i915.disable_power_well);
-}
-
-void intel_display_power_get(struct drm_i915_private *dev_priv,
-			     enum intel_display_power_domain domain)
-{
-	struct i915_power_domains *power_domains;
-	struct i915_power_well *power_well;
-	int i;
-
-	intel_runtime_pm_get(dev_priv);
-
-	power_domains = &dev_priv->power_domains;
-
-	mutex_lock(&power_domains->lock);
-
-	for_each_power_well(i, power_well, BIT(domain), power_domains) {
-		if (!power_well->count++) {
-			DRM_DEBUG_KMS("enabling %s\n", power_well->name);
-			power_well->ops->enable(dev_priv, power_well);
-			power_well->hw_enabled = true;
-		}
-
-		check_power_well_state(dev_priv, power_well);
-	}
-
-	power_domains->domain_use_count[domain]++;
-
-	mutex_unlock(&power_domains->lock);
-}
-
-void intel_display_power_put(struct drm_i915_private *dev_priv,
-			     enum intel_display_power_domain domain)
-{
-	struct i915_power_domains *power_domains;
-	struct i915_power_well *power_well;
-	int i;
-
-	power_domains = &dev_priv->power_domains;
-
-	mutex_lock(&power_domains->lock);
-
-	WARN_ON(!power_domains->domain_use_count[domain]);
-	power_domains->domain_use_count[domain]--;
-
-	for_each_power_well_rev(i, power_well, BIT(domain), power_domains) {
-		WARN_ON(!power_well->count);
-
-		if (!--power_well->count && i915.disable_power_well) {
-			DRM_DEBUG_KMS("disabling %s\n", power_well->name);
-			power_well->hw_enabled = false;
-			power_well->ops->disable(dev_priv, power_well);
-		}
-
-		check_power_well_state(dev_priv, power_well);
-	}
-
-	mutex_unlock(&power_domains->lock);
-
-	intel_runtime_pm_put(dev_priv);
-}
-
-static struct i915_power_domains *hsw_pwr;
-
-/* Display audio driver power well request */
-int i915_request_power_well(void)
-{
-	struct drm_i915_private *dev_priv;
-
-	if (!hsw_pwr)
-		return -ENODEV;
-
-	dev_priv = container_of(hsw_pwr, struct drm_i915_private,
-				power_domains);
-	intel_display_power_get(dev_priv, POWER_DOMAIN_AUDIO);
-	return 0;
-}
-EXPORT_SYMBOL_GPL(i915_request_power_well);
-
-/* Display audio driver power well release */
-int i915_release_power_well(void)
-{
-	struct drm_i915_private *dev_priv;
-
-	if (!hsw_pwr)
-		return -ENODEV;
-
-	dev_priv = container_of(hsw_pwr, struct drm_i915_private,
-				power_domains);
-	intel_display_power_put(dev_priv, POWER_DOMAIN_AUDIO);
-	return 0;
-}
-EXPORT_SYMBOL_GPL(i915_release_power_well);
-
-/*
- * Private interface for the audio driver to get CDCLK in kHz.
- *
- * Caller must request power well using i915_request_power_well() prior to
- * making the call.
- */
-int i915_get_cdclk_freq(void)
-{
-	struct drm_i915_private *dev_priv;
-
-	if (!hsw_pwr)
-		return -ENODEV;
-
-	dev_priv = container_of(hsw_pwr, struct drm_i915_private,
-				power_domains);
-
-	return intel_ddi_get_cdclk_freq(dev_priv);
-}
-EXPORT_SYMBOL_GPL(i915_get_cdclk_freq);
-
-
-#define POWER_DOMAIN_MASK (BIT(POWER_DOMAIN_NUM) - 1)
-
-#define HSW_ALWAYS_ON_POWER_DOMAINS (			\
-	BIT(POWER_DOMAIN_PIPE_A) |			\
-	BIT(POWER_DOMAIN_TRANSCODER_EDP) |		\
-	BIT(POWER_DOMAIN_PORT_DDI_A_2_LANES) |		\
-	BIT(POWER_DOMAIN_PORT_DDI_A_4_LANES) |		\
-	BIT(POWER_DOMAIN_PORT_DDI_B_2_LANES) |		\
-	BIT(POWER_DOMAIN_PORT_DDI_B_4_LANES) |		\
-	BIT(POWER_DOMAIN_PORT_DDI_C_2_LANES) |		\
-	BIT(POWER_DOMAIN_PORT_DDI_C_4_LANES) |		\
-	BIT(POWER_DOMAIN_PORT_DDI_D_2_LANES) |		\
-	BIT(POWER_DOMAIN_PORT_DDI_D_4_LANES) |		\
-	BIT(POWER_DOMAIN_PORT_CRT) |			\
-	BIT(POWER_DOMAIN_PLLS) |			\
-	BIT(POWER_DOMAIN_INIT))
-#define HSW_DISPLAY_POWER_DOMAINS (				\
-	(POWER_DOMAIN_MASK & ~HSW_ALWAYS_ON_POWER_DOMAINS) |	\
-	BIT(POWER_DOMAIN_INIT))
-
-#define BDW_ALWAYS_ON_POWER_DOMAINS (			\
-	HSW_ALWAYS_ON_POWER_DOMAINS |			\
-	BIT(POWER_DOMAIN_PIPE_A_PANEL_FITTER))
-#define BDW_DISPLAY_POWER_DOMAINS (				\
-	(POWER_DOMAIN_MASK & ~BDW_ALWAYS_ON_POWER_DOMAINS) |	\
-	BIT(POWER_DOMAIN_INIT))
-
-#define VLV_ALWAYS_ON_POWER_DOMAINS	BIT(POWER_DOMAIN_INIT)
-#define VLV_DISPLAY_POWER_DOMAINS	POWER_DOMAIN_MASK
-
-#define VLV_DPIO_CMN_BC_POWER_DOMAINS (		\
-	BIT(POWER_DOMAIN_PORT_DDI_B_2_LANES) |	\
-	BIT(POWER_DOMAIN_PORT_DDI_B_4_LANES) |	\
-	BIT(POWER_DOMAIN_PORT_DDI_C_2_LANES) |	\
-	BIT(POWER_DOMAIN_PORT_DDI_C_4_LANES) |	\
-	BIT(POWER_DOMAIN_PORT_CRT) |		\
-	BIT(POWER_DOMAIN_INIT))
-
-#define VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS (	\
-	BIT(POWER_DOMAIN_PORT_DDI_B_2_LANES) |	\
-	BIT(POWER_DOMAIN_PORT_DDI_B_4_LANES) |	\
-	BIT(POWER_DOMAIN_INIT))
-
-#define VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS (	\
-	BIT(POWER_DOMAIN_PORT_DDI_B_4_LANES) |	\
-	BIT(POWER_DOMAIN_INIT))
-
-#define VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS (	\
-	BIT(POWER_DOMAIN_PORT_DDI_C_2_LANES) |	\
-	BIT(POWER_DOMAIN_PORT_DDI_C_4_LANES) |	\
-	BIT(POWER_DOMAIN_INIT))
-
-#define VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS (	\
-	BIT(POWER_DOMAIN_PORT_DDI_C_4_LANES) |	\
-	BIT(POWER_DOMAIN_INIT))
-
-static const struct i915_power_well_ops i9xx_always_on_power_well_ops = {
-	.sync_hw = i9xx_always_on_power_well_noop,
-	.enable = i9xx_always_on_power_well_noop,
-	.disable = i9xx_always_on_power_well_noop,
-	.is_enabled = i9xx_always_on_power_well_enabled,
-};
-
-static struct i915_power_well i9xx_always_on_power_well[] = {
-	{
-		.name = "always-on",
-		.always_on = 1,
-		.domains = POWER_DOMAIN_MASK,
-		.ops = &i9xx_always_on_power_well_ops,
-	},
-};
-
-static const struct i915_power_well_ops hsw_power_well_ops = {
-	.sync_hw = hsw_power_well_sync_hw,
-	.enable = hsw_power_well_enable,
-	.disable = hsw_power_well_disable,
-	.is_enabled = hsw_power_well_enabled,
-};
-
-static struct i915_power_well hsw_power_wells[] = {
-	{
-		.name = "always-on",
-		.always_on = 1,
-		.domains = HSW_ALWAYS_ON_POWER_DOMAINS,
-		.ops = &i9xx_always_on_power_well_ops,
-	},
-	{
-		.name = "display",
-		.domains = HSW_DISPLAY_POWER_DOMAINS,
-		.ops = &hsw_power_well_ops,
-	},
-};
-
-static struct i915_power_well bdw_power_wells[] = {
-	{
-		.name = "always-on",
-		.always_on = 1,
-		.domains = BDW_ALWAYS_ON_POWER_DOMAINS,
-		.ops = &i9xx_always_on_power_well_ops,
-	},
-	{
-		.name = "display",
-		.domains = BDW_DISPLAY_POWER_DOMAINS,
-		.ops = &hsw_power_well_ops,
-	},
-};
-
-static const struct i915_power_well_ops vlv_display_power_well_ops = {
-	.sync_hw = vlv_power_well_sync_hw,
-	.enable = vlv_display_power_well_enable,
-	.disable = vlv_display_power_well_disable,
-	.is_enabled = vlv_power_well_enabled,
-};
-
-static const struct i915_power_well_ops vlv_dpio_cmn_power_well_ops = {
-	.sync_hw = vlv_power_well_sync_hw,
-	.enable = vlv_dpio_cmn_power_well_enable,
-	.disable = vlv_dpio_cmn_power_well_disable,
-	.is_enabled = vlv_power_well_enabled,
-};
-
-static const struct i915_power_well_ops vlv_dpio_power_well_ops = {
-	.sync_hw = vlv_power_well_sync_hw,
-	.enable = vlv_power_well_enable,
-	.disable = vlv_power_well_disable,
-	.is_enabled = vlv_power_well_enabled,
-};
-
-static struct i915_power_well vlv_power_wells[] = {
-	{
-		.name = "always-on",
-		.always_on = 1,
-		.domains = VLV_ALWAYS_ON_POWER_DOMAINS,
-		.ops = &i9xx_always_on_power_well_ops,
-	},
-	{
-		.name = "display",
-		.domains = VLV_DISPLAY_POWER_DOMAINS,
-		.data = PUNIT_POWER_WELL_DISP2D,
-		.ops = &vlv_display_power_well_ops,
-	},
-	{
-		.name = "dpio-tx-b-01",
-		.domains = VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS |
-			   VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS |
-			   VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS |
-			   VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS,
-		.ops = &vlv_dpio_power_well_ops,
-		.data = PUNIT_POWER_WELL_DPIO_TX_B_LANES_01,
-	},
-	{
-		.name = "dpio-tx-b-23",
-		.domains = VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS |
-			   VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS |
-			   VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS |
-			   VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS,
-		.ops = &vlv_dpio_power_well_ops,
-		.data = PUNIT_POWER_WELL_DPIO_TX_B_LANES_23,
-	},
-	{
-		.name = "dpio-tx-c-01",
-		.domains = VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS |
-			   VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS |
-			   VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS |
-			   VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS,
-		.ops = &vlv_dpio_power_well_ops,
-		.data = PUNIT_POWER_WELL_DPIO_TX_C_LANES_01,
-	},
-	{
-		.name = "dpio-tx-c-23",
-		.domains = VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS |
-			   VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS |
-			   VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS |
-			   VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS,
-		.ops = &vlv_dpio_power_well_ops,
-		.data = PUNIT_POWER_WELL_DPIO_TX_C_LANES_23,
-	},
-	{
-		.name = "dpio-common",
-		.domains = VLV_DPIO_CMN_BC_POWER_DOMAINS,
-		.data = PUNIT_POWER_WELL_DPIO_CMN_BC,
-		.ops = &vlv_dpio_cmn_power_well_ops,
-	},
-};
-
-static struct i915_power_well *lookup_power_well(struct drm_i915_private *dev_priv,
-						 enum punit_power_well power_well_id)
-{
-	struct i915_power_domains *power_domains = &dev_priv->power_domains;
-	struct i915_power_well *power_well;
-	int i;
-
-	for_each_power_well(i, power_well, POWER_DOMAIN_MASK, power_domains) {
-		if (power_well->data == power_well_id)
-			return power_well;
-	}
-
-	return NULL;
-}
-
-#define set_power_wells(power_domains, __power_wells) ({		\
-	(power_domains)->power_wells = (__power_wells);			\
-	(power_domains)->power_well_count = ARRAY_SIZE(__power_wells);	\
-})
-
-int intel_power_domains_init(struct drm_i915_private *dev_priv)
-{
-	struct i915_power_domains *power_domains = &dev_priv->power_domains;
-
-	mutex_init(&power_domains->lock);
-
-	/*
-	 * The enabling order will be from lower to higher indexed wells,
-	 * the disabling order is reversed.
-	 */
-	if (IS_HASWELL(dev_priv->dev)) {
-		set_power_wells(power_domains, hsw_power_wells);
-		hsw_pwr = power_domains;
-	} else if (IS_BROADWELL(dev_priv->dev)) {
-		set_power_wells(power_domains, bdw_power_wells);
-		hsw_pwr = power_domains;
-	} else if (IS_VALLEYVIEW(dev_priv->dev)) {
-		set_power_wells(power_domains, vlv_power_wells);
+	if (INTEL_INFO(dev_priv)->gen >= 7) {
+		dev_priv->display.fbc_enabled = ironlake_fbc_enabled;
+		dev_priv->display.enable_fbc = gen7_enable_fbc;
+		dev_priv->display.disable_fbc = ironlake_disable_fbc;
+	} else if (INTEL_INFO(dev_priv)->gen >= 5) {
+		dev_priv->display.fbc_enabled = ironlake_fbc_enabled;
+		dev_priv->display.enable_fbc = ironlake_enable_fbc;
+		dev_priv->display.disable_fbc = ironlake_disable_fbc;
+	} else if (IS_GM45(dev_priv)) {
+		dev_priv->display.fbc_enabled = g4x_fbc_enabled;
+		dev_priv->display.enable_fbc = g4x_enable_fbc;
+		dev_priv->display.disable_fbc = g4x_disable_fbc;
 	} else {
-		set_power_wells(power_domains, i9xx_always_on_power_well);
-	}
-
-	return 0;
-}
-
-void intel_power_domains_remove(struct drm_i915_private *dev_priv)
-{
-	hsw_pwr = NULL;
-}
-
-static void intel_power_domains_resume(struct drm_i915_private *dev_priv)
-{
-	struct i915_power_domains *power_domains = &dev_priv->power_domains;
-	struct i915_power_well *power_well;
-	int i;
-
-	mutex_lock(&power_domains->lock);
-	for_each_power_well(i, power_well, POWER_DOMAIN_MASK, power_domains) {
-		power_well->ops->sync_hw(dev_priv, power_well);
-		power_well->hw_enabled = power_well->ops->is_enabled(dev_priv,
-								     power_well);
-	}
-	mutex_unlock(&power_domains->lock);
-}
-
-static void vlv_cmnlane_wa(struct drm_i915_private *dev_priv)
-{
-	struct i915_power_well *cmn =
-		lookup_power_well(dev_priv, PUNIT_POWER_WELL_DPIO_CMN_BC);
-	struct i915_power_well *disp2d =
-		lookup_power_well(dev_priv, PUNIT_POWER_WELL_DISP2D);
-
-	/* nothing to do if common lane is already off */
-	if (!cmn->ops->is_enabled(dev_priv, cmn))
-		return;
-
-	/* If the display might be already active skip this */
-	if (disp2d->ops->is_enabled(dev_priv, disp2d) &&
-	    I915_READ(DPIO_CTL) & DPIO_CMNRST)
-		return;
-
-	DRM_DEBUG_KMS("toggling display PHY side reset\n");
-
-	/* cmnlane needs DPLL registers */
-	disp2d->ops->enable(dev_priv, disp2d);
-
-	/*
-	 * From VLV2A0_DP_eDP_HDMI_DPIO_driver_vbios_notes_11.docx:
-	 * Need to assert and de-assert PHY SB reset by gating the
-	 * common lane power, then un-gating it.
-	 * Simply ungating isn't enough to reset the PHY enough to get
-	 * ports and lanes running.
-	 */
-	cmn->ops->disable(dev_priv, cmn);
-}
-
-void intel_power_domains_init_hw(struct drm_i915_private *dev_priv)
-{
-	struct drm_device *dev = dev_priv->dev;
-	struct i915_power_domains *power_domains = &dev_priv->power_domains;
-
-	power_domains->initializing = true;
+		dev_priv->display.fbc_enabled = i8xx_fbc_enabled;
+		dev_priv->display.enable_fbc = i8xx_enable_fbc;
+		dev_priv->display.disable_fbc = i8xx_disable_fbc;
 
-	if (IS_VALLEYVIEW(dev) && !IS_CHERRYVIEW(dev)) {
-		mutex_lock(&power_domains->lock);
-		vlv_cmnlane_wa(dev_priv);
-		mutex_unlock(&power_domains->lock);
+		/* This value was pulled out of someone's hat */
+		I915_WRITE(FBC_CONTROL, 500 << FBC_CTL_INTERVAL_SHIFT);
 	}
 
-	/* For now, we need the power well to be always enabled. */
-	intel_display_set_init_power(dev_priv, true);
-	intel_power_domains_resume(dev_priv);
-	power_domains->initializing = false;
-}
-
-void intel_aux_display_runtime_get(struct drm_i915_private *dev_priv)
-{
-	intel_runtime_pm_get(dev_priv);
-}
-
-void intel_aux_display_runtime_put(struct drm_i915_private *dev_priv)
-{
-	intel_runtime_pm_put(dev_priv);
-}
-
-void intel_runtime_pm_get(struct drm_i915_private *dev_priv)
-{
-	struct drm_device *dev = dev_priv->dev;
-	struct device *device = &dev->pdev->dev;
-
-	if (!HAS_RUNTIME_PM(dev))
-		return;
-
-	pm_runtime_get_sync(device);
-	WARN(dev_priv->pm.suspended, "Device still suspended.\n");
-}
-
-void intel_runtime_pm_get_noresume(struct drm_i915_private *dev_priv)
-{
-	struct drm_device *dev = dev_priv->dev;
-	struct device *device = &dev->pdev->dev;
-
-	if (!HAS_RUNTIME_PM(dev))
-		return;
-
-	WARN(dev_priv->pm.suspended, "Getting nosync-ref while suspended.\n");
-	pm_runtime_get_noresume(device);
-}
-
-void intel_runtime_pm_put(struct drm_i915_private *dev_priv)
-{
-	struct drm_device *dev = dev_priv->dev;
-	struct device *device = &dev->pdev->dev;
-
-	if (!HAS_RUNTIME_PM(dev))
-		return;
-
-	pm_runtime_mark_last_busy(device);
-	pm_runtime_put_autosuspend(device);
-}
-
-void intel_init_runtime_pm(struct drm_i915_private *dev_priv)
-{
-	struct drm_device *dev = dev_priv->dev;
-	struct device *device = &dev->pdev->dev;
-
-	if (!HAS_RUNTIME_PM(dev))
-		return;
-
-	pm_runtime_set_active(device);
-
-	/*
-	 * RPM depends on RC6 to save restore the GT HW context, so make RC6 a
-	 * requirement.
-	 */
-	if (!intel_enable_rc6(dev)) {
-		DRM_INFO("RC6 disabled, disabling runtime PM support\n");
-		return;
-	}
-
-	pm_runtime_set_autosuspend_delay(device, 10000); /* 10s */
-	pm_runtime_mark_last_busy(device);
-	pm_runtime_use_autosuspend(device);
-
-	pm_runtime_put_autosuspend(device);
-}
-
-void intel_fini_runtime_pm(struct drm_i915_private *dev_priv)
-{
-	struct drm_device *dev = dev_priv->dev;
-	struct device *device = &dev->pdev->dev;
-
-	if (!HAS_RUNTIME_PM(dev))
-		return;
-
-	if (!intel_enable_rc6(dev))
-		return;
-
-	/* Make sure we're not suspended first. */
-	pm_runtime_get_sync(device);
-	pm_runtime_disable(device);
+	dev_priv->fbc.enabled = dev_priv->display.fbc_enabled(dev_priv->dev);
 }
 
 /* Set up chip specific power management-related functions */
@@ -6780,28 +7079,7 @@ void intel_init_pm(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (HAS_FBC(dev)) {
-		if (INTEL_INFO(dev)->gen >= 7) {
-			dev_priv->display.fbc_enabled = ironlake_fbc_enabled;
-			dev_priv->display.enable_fbc = gen7_enable_fbc;
-			dev_priv->display.disable_fbc = ironlake_disable_fbc;
-		} else if (INTEL_INFO(dev)->gen >= 5) {
-			dev_priv->display.fbc_enabled = ironlake_fbc_enabled;
-			dev_priv->display.enable_fbc = ironlake_enable_fbc;
-			dev_priv->display.disable_fbc = ironlake_disable_fbc;
-		} else if (IS_GM45(dev)) {
-			dev_priv->display.fbc_enabled = g4x_fbc_enabled;
-			dev_priv->display.enable_fbc = g4x_enable_fbc;
-			dev_priv->display.disable_fbc = g4x_disable_fbc;
-		} else {
-			dev_priv->display.fbc_enabled = i8xx_fbc_enabled;
-			dev_priv->display.enable_fbc = i8xx_enable_fbc;
-			dev_priv->display.disable_fbc = i8xx_disable_fbc;
-
-			/* This value was pulled out of someone's hat */
-			I915_WRITE(FBC_CONTROL, 500 << FBC_CTL_INTERVAL_SHIFT);
-		}
-	}
+	intel_init_fbc(dev_priv);
 
 	/* For cxsr */
 	if (IS_PINEVIEW(dev))
@@ -6810,7 +7088,13 @@ void intel_init_pm(struct drm_device *dev)
 		i915_ironlake_get_mem_freq(dev);
 
 	/* For FIFO watermark updates */
-	if (HAS_PCH_SPLIT(dev)) {
+	if (INTEL_INFO(dev)->gen >= 9) {
+		skl_setup_wm_latency(dev);
+
+		dev_priv->display.init_clock_gating = gen9_init_clock_gating;
+		dev_priv->display.update_wm = skl_update_wm;
+		dev_priv->display.update_sprite_wm = skl_update_sprite_wm;
+	} else if (HAS_PCH_SPLIT(dev)) {
 		ilk_setup_wm_latency(dev);
 
 		if ((IS_GEN5(dev) && dev_priv->wm.pri_latency[1] &&
@@ -6833,13 +7117,15 @@ void intel_init_pm(struct drm_device *dev)
 		else if (IS_HASWELL(dev))
 			dev_priv->display.init_clock_gating = haswell_init_clock_gating;
 		else if (INTEL_INFO(dev)->gen == 8)
-			dev_priv->display.init_clock_gating = gen8_init_clock_gating;
+			dev_priv->display.init_clock_gating = broadwell_init_clock_gating;
 	} else if (IS_CHERRYVIEW(dev)) {
-		dev_priv->display.update_wm = valleyview_update_wm;
+		dev_priv->display.update_wm = cherryview_update_wm;
+		dev_priv->display.update_sprite_wm = valleyview_update_sprite_wm;
 		dev_priv->display.init_clock_gating =
 			cherryview_init_clock_gating;
 	} else if (IS_VALLEYVIEW(dev)) {
 		dev_priv->display.update_wm = valleyview_update_wm;
+		dev_priv->display.update_sprite_wm = valleyview_update_sprite_wm;
 		dev_priv->display.init_clock_gating =
 			valleyview_init_clock_gating;
 	} else if (IS_PINEVIEW(dev)) {
@@ -6889,7 +7175,7 @@ void intel_init_pm(struct drm_device *dev)
 	}
 }
 
-int sandybridge_pcode_read(struct drm_i915_private *dev_priv, u8 mbox, u32 *val)
+int sandybridge_pcode_read(struct drm_i915_private *dev_priv, u32 mbox, u32 *val)
 {
 	WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock));
 
@@ -6899,6 +7185,7 @@ int sandybridge_pcode_read(struct drm_i915_private *dev_priv, u8 mbox, u32 *val)
 	}
 
 	I915_WRITE(GEN6_PCODE_DATA, *val);
+	I915_WRITE(GEN6_PCODE_DATA1, 0);
 	I915_WRITE(GEN6_PCODE_MAILBOX, GEN6_PCODE_READY | mbox);
 
 	if (wait_for((I915_READ(GEN6_PCODE_MAILBOX) & GEN6_PCODE_READY) == 0,
@@ -6913,7 +7200,7 @@ int sandybridge_pcode_read(struct drm_i915_private *dev_priv, u8 mbox, u32 *val)
 	return 0;
 }
 
-int sandybridge_pcode_write(struct drm_i915_private *dev_priv, u8 mbox, u32 val)
+int sandybridge_pcode_write(struct drm_i915_private *dev_priv, u32 mbox, u32 val)
 {
 	WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock));
 
@@ -6936,98 +7223,66 @@ int sandybridge_pcode_write(struct drm_i915_private *dev_priv, u8 mbox, u32 val)
 	return 0;
 }
 
-static int byt_gpu_freq(struct drm_i915_private *dev_priv, int val)
+static int vlv_gpu_freq_div(unsigned int czclk_freq)
 {
-	int div;
-
-	/* 4 x czclk */
-	switch (dev_priv->mem_freq) {
-	case 800:
-		div = 10;
-		break;
-	case 1066:
-		div = 12;
-		break;
-	case 1333:
-		div = 16;
-		break;
+	switch (czclk_freq) {
+	case 200:
+		return 10;
+	case 267:
+		return 12;
+	case 320:
+	case 333:
+		return 16;
+	case 400:
+		return 20;
 	default:
 		return -1;
 	}
+}
 
-	return DIV_ROUND_CLOSEST(dev_priv->mem_freq * (val + 6 - 0xbd), 4 * div);
+static int byt_gpu_freq(struct drm_i915_private *dev_priv, int val)
+{
+	int div, czclk_freq = DIV_ROUND_CLOSEST(dev_priv->mem_freq, 4);
+
+	div = vlv_gpu_freq_div(czclk_freq);
+	if (div < 0)
+		return div;
+
+	return DIV_ROUND_CLOSEST(czclk_freq * (val + 6 - 0xbd), div);
 }
 
 static int byt_freq_opcode(struct drm_i915_private *dev_priv, int val)
 {
-	int mul;
+	int mul, czclk_freq = DIV_ROUND_CLOSEST(dev_priv->mem_freq, 4);
 
-	/* 4 x czclk */
-	switch (dev_priv->mem_freq) {
-	case 800:
-		mul = 10;
-		break;
-	case 1066:
-		mul = 12;
-		break;
-	case 1333:
-		mul = 16;
-		break;
-	default:
-		return -1;
-	}
+	mul = vlv_gpu_freq_div(czclk_freq);
+	if (mul < 0)
+		return mul;
 
-	return DIV_ROUND_CLOSEST(4 * mul * val, dev_priv->mem_freq) + 0xbd - 6;
+	return DIV_ROUND_CLOSEST(mul * val, czclk_freq) + 0xbd - 6;
 }
 
 static int chv_gpu_freq(struct drm_i915_private *dev_priv, int val)
 {
-	int div, freq;
+	int div, czclk_freq = dev_priv->rps.cz_freq;
 
-	switch (dev_priv->rps.cz_freq) {
-	case 200:
-		div = 5;
-		break;
-	case 267:
-		div = 6;
-		break;
-	case 320:
-	case 333:
-	case 400:
-		div = 8;
-		break;
-	default:
-		return -1;
-	}
+	div = vlv_gpu_freq_div(czclk_freq) / 2;
+	if (div < 0)
+		return div;
 
-	freq = (DIV_ROUND_CLOSEST((dev_priv->rps.cz_freq * val), 2 * div) / 2);
-
-	return freq;
+	return DIV_ROUND_CLOSEST(czclk_freq * val, 2 * div) / 2;
 }
 
 static int chv_freq_opcode(struct drm_i915_private *dev_priv, int val)
 {
-	int mul, opcode;
-
-	switch (dev_priv->rps.cz_freq) {
-	case 200:
-		mul = 5;
-		break;
-	case 267:
-		mul = 6;
-		break;
-	case 320:
-	case 333:
-	case 400:
-		mul = 8;
-		break;
-	default:
-		return -1;
-	}
+	int mul, czclk_freq = dev_priv->rps.cz_freq;
 
-	opcode = (DIV_ROUND_CLOSEST((val * 2 * mul), dev_priv->rps.cz_freq) * 2);
+	mul = vlv_gpu_freq_div(czclk_freq) / 2;
+	if (mul < 0)
+		return mul;
 
-	return opcode;
+	/* CHV needs even values */
+	return DIV_ROUND_CLOSEST(val * 2 * mul, czclk_freq) * 2;
 }
 
 int vlv_gpu_freq(struct drm_i915_private *dev_priv, int val)
@@ -7064,5 +7319,4 @@ void intel_pm_setup(struct drm_device *dev)
 			  intel_gen6_powersave_work);
 
 	dev_priv->pm.suspended = false;
-	dev_priv->pm._irqs_disabled = false;
 }
diff --git a/drivers/gpu/drm/i915/intel_psr.c b/drivers/gpu/drm/i915/intel_psr.c
new file mode 100644
index 000000000000..716b8a961eea
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_psr.c
@@ -0,0 +1,481 @@
+/*
+ * Copyright © 2014 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+/**
+ * DOC: Panel Self Refresh (PSR/SRD)
+ *
+ * Since Haswell Display controller supports Panel Self-Refresh on display
+ * panels witch have a remote frame buffer (RFB) implemented according to PSR
+ * spec in eDP1.3. PSR feature allows the display to go to lower standby states
+ * when system is idle but display is on as it eliminates display refresh
+ * request to DDR memory completely as long as the frame buffer for that
+ * display is unchanged.
+ *
+ * Panel Self Refresh must be supported by both Hardware (source) and
+ * Panel (sink).
+ *
+ * PSR saves power by caching the framebuffer in the panel RFB, which allows us
+ * to power down the link and memory controller. For DSI panels the same idea
+ * is called "manual mode".
+ *
+ * The implementation uses the hardware-based PSR support which automatically
+ * enters/exits self-refresh mode. The hardware takes care of sending the
+ * required DP aux message and could even retrain the link (that part isn't
+ * enabled yet though). The hardware also keeps track of any frontbuffer
+ * changes to know when to exit self-refresh mode again. Unfortunately that
+ * part doesn't work too well, hence why the i915 PSR support uses the
+ * software frontbuffer tracking to make sure it doesn't miss a screen
+ * update. For this integration intel_psr_invalidate() and intel_psr_flush()
+ * get called by the frontbuffer tracking code. Note that because of locking
+ * issues the self-refresh re-enable code is done from a work queue, which
+ * must be correctly synchronized/cancelled when shutting down the pipe."
+ */
+
+#include <drm/drmP.h>
+
+#include "intel_drv.h"
+#include "i915_drv.h"
+
+static bool is_edp_psr(struct intel_dp *intel_dp)
+{
+	return intel_dp->psr_dpcd[0] & DP_PSR_IS_SUPPORTED;
+}
+
+bool intel_psr_is_enabled(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (!HAS_PSR(dev))
+		return false;
+
+	return I915_READ(EDP_PSR_CTL(dev)) & EDP_PSR_ENABLE;
+}
+
+static void intel_psr_write_vsc(struct intel_dp *intel_dp,
+				    struct edp_vsc_psr *vsc_psr)
+{
+	struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = dig_port->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *crtc = to_intel_crtc(dig_port->base.base.crtc);
+	u32 ctl_reg = HSW_TVIDEO_DIP_CTL(crtc->config.cpu_transcoder);
+	u32 data_reg = HSW_TVIDEO_DIP_VSC_DATA(crtc->config.cpu_transcoder);
+	uint32_t *data = (uint32_t *) vsc_psr;
+	unsigned int i;
+
+	/* As per BSPec (Pipe Video Data Island Packet), we need to disable
+	   the video DIP being updated before program video DIP data buffer
+	   registers for DIP being updated. */
+	I915_WRITE(ctl_reg, 0);
+	POSTING_READ(ctl_reg);
+
+	for (i = 0; i < VIDEO_DIP_VSC_DATA_SIZE; i += 4) {
+		if (i < sizeof(struct edp_vsc_psr))
+			I915_WRITE(data_reg + i, *data++);
+		else
+			I915_WRITE(data_reg + i, 0);
+	}
+
+	I915_WRITE(ctl_reg, VIDEO_DIP_ENABLE_VSC_HSW);
+	POSTING_READ(ctl_reg);
+}
+
+static void intel_psr_setup_vsc(struct intel_dp *intel_dp)
+{
+	struct edp_vsc_psr psr_vsc;
+
+	/* Prepare VSC packet as per EDP 1.3 spec, Table 3.10 */
+	memset(&psr_vsc, 0, sizeof(psr_vsc));
+	psr_vsc.sdp_header.HB0 = 0;
+	psr_vsc.sdp_header.HB1 = 0x7;
+	psr_vsc.sdp_header.HB2 = 0x2;
+	psr_vsc.sdp_header.HB3 = 0x8;
+	intel_psr_write_vsc(intel_dp, &psr_vsc);
+}
+
+static void intel_psr_enable_sink(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = dig_port->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	uint32_t aux_clock_divider;
+	int precharge = 0x3;
+	bool only_standby = false;
+	static const uint8_t aux_msg[] = {
+		[0] = DP_AUX_NATIVE_WRITE << 4,
+		[1] = DP_SET_POWER >> 8,
+		[2] = DP_SET_POWER & 0xff,
+		[3] = 1 - 1,
+		[4] = DP_SET_POWER_D0,
+	};
+	int i;
+
+	BUILD_BUG_ON(sizeof(aux_msg) > 20);
+
+	aux_clock_divider = intel_dp->get_aux_clock_divider(intel_dp, 0);
+
+	if (IS_BROADWELL(dev) && dig_port->port != PORT_A)
+		only_standby = true;
+
+	/* Enable PSR in sink */
+	if (intel_dp->psr_dpcd[1] & DP_PSR_NO_TRAIN_ON_EXIT || only_standby)
+		drm_dp_dpcd_writeb(&intel_dp->aux, DP_PSR_EN_CFG,
+				   DP_PSR_ENABLE & ~DP_PSR_MAIN_LINK_ACTIVE);
+	else
+		drm_dp_dpcd_writeb(&intel_dp->aux, DP_PSR_EN_CFG,
+				   DP_PSR_ENABLE | DP_PSR_MAIN_LINK_ACTIVE);
+
+	/* Setup AUX registers */
+	for (i = 0; i < sizeof(aux_msg); i += 4)
+		I915_WRITE(EDP_PSR_AUX_DATA1(dev) + i,
+			   intel_dp_pack_aux(&aux_msg[i], sizeof(aux_msg) - i));
+
+	I915_WRITE(EDP_PSR_AUX_CTL(dev),
+		   DP_AUX_CH_CTL_TIME_OUT_400us |
+		   (sizeof(aux_msg) << DP_AUX_CH_CTL_MESSAGE_SIZE_SHIFT) |
+		   (precharge << DP_AUX_CH_CTL_PRECHARGE_2US_SHIFT) |
+		   (aux_clock_divider << DP_AUX_CH_CTL_BIT_CLOCK_2X_SHIFT));
+}
+
+static void intel_psr_enable_source(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = dig_port->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	uint32_t max_sleep_time = 0x1f;
+	uint32_t idle_frames = 1;
+	uint32_t val = 0x0;
+	const uint32_t link_entry_time = EDP_PSR_MIN_LINK_ENTRY_TIME_8_LINES;
+	bool only_standby = false;
+
+	if (IS_BROADWELL(dev) && dig_port->port != PORT_A)
+		only_standby = true;
+
+	if (intel_dp->psr_dpcd[1] & DP_PSR_NO_TRAIN_ON_EXIT || only_standby) {
+		val |= EDP_PSR_LINK_STANDBY;
+		val |= EDP_PSR_TP2_TP3_TIME_0us;
+		val |= EDP_PSR_TP1_TIME_0us;
+		val |= EDP_PSR_SKIP_AUX_EXIT;
+		val |= IS_BROADWELL(dev) ? BDW_PSR_SINGLE_FRAME : 0;
+	} else
+		val |= EDP_PSR_LINK_DISABLE;
+
+	I915_WRITE(EDP_PSR_CTL(dev), val |
+		   (IS_BROADWELL(dev) ? 0 : link_entry_time) |
+		   max_sleep_time << EDP_PSR_MAX_SLEEP_TIME_SHIFT |
+		   idle_frames << EDP_PSR_IDLE_FRAME_SHIFT |
+		   EDP_PSR_ENABLE);
+}
+
+static bool intel_psr_match_conditions(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = dig_port->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_crtc *crtc = dig_port->base.base.crtc;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+
+	lockdep_assert_held(&dev_priv->psr.lock);
+	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
+	WARN_ON(!drm_modeset_is_locked(&crtc->mutex));
+
+	dev_priv->psr.source_ok = false;
+
+	if (IS_HASWELL(dev) && dig_port->port != PORT_A) {
+		DRM_DEBUG_KMS("HSW ties PSR to DDI A (eDP)\n");
+		return false;
+	}
+
+	if (!i915.enable_psr) {
+		DRM_DEBUG_KMS("PSR disable by flag\n");
+		return false;
+	}
+
+	/* Below limitations aren't valid for Broadwell */
+	if (IS_BROADWELL(dev))
+		goto out;
+
+	if (I915_READ(HSW_STEREO_3D_CTL(intel_crtc->config.cpu_transcoder)) &
+	    S3D_ENABLE) {
+		DRM_DEBUG_KMS("PSR condition failed: Stereo 3D is Enabled\n");
+		return false;
+	}
+
+	if (intel_crtc->config.adjusted_mode.flags & DRM_MODE_FLAG_INTERLACE) {
+		DRM_DEBUG_KMS("PSR condition failed: Interlaced is Enabled\n");
+		return false;
+	}
+
+ out:
+	dev_priv->psr.source_ok = true;
+	return true;
+}
+
+static void intel_psr_do_enable(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = intel_dig_port->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	WARN_ON(I915_READ(EDP_PSR_CTL(dev)) & EDP_PSR_ENABLE);
+	WARN_ON(dev_priv->psr.active);
+	lockdep_assert_held(&dev_priv->psr.lock);
+
+	/* Enable/Re-enable PSR on the host */
+	intel_psr_enable_source(intel_dp);
+
+	dev_priv->psr.active = true;
+}
+
+/**
+ * intel_psr_enable - Enable PSR
+ * @intel_dp: Intel DP
+ *
+ * This function can only be called after the pipe is fully trained and enabled.
+ */
+void intel_psr_enable(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = intel_dig_port->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (!HAS_PSR(dev)) {
+		DRM_DEBUG_KMS("PSR not supported on this platform\n");
+		return;
+	}
+
+	if (!is_edp_psr(intel_dp)) {
+		DRM_DEBUG_KMS("PSR not supported by this panel\n");
+		return;
+	}
+
+	mutex_lock(&dev_priv->psr.lock);
+	if (dev_priv->psr.enabled) {
+		DRM_DEBUG_KMS("PSR already in use\n");
+		goto unlock;
+	}
+
+	if (!intel_psr_match_conditions(intel_dp))
+		goto unlock;
+
+	dev_priv->psr.busy_frontbuffer_bits = 0;
+
+	intel_psr_setup_vsc(intel_dp);
+
+	/* Avoid continuous PSR exit by masking memup and hpd */
+	I915_WRITE(EDP_PSR_DEBUG_CTL(dev), EDP_PSR_DEBUG_MASK_MEMUP |
+		   EDP_PSR_DEBUG_MASK_HPD | EDP_PSR_DEBUG_MASK_LPSP);
+
+	/* Enable PSR on the panel */
+	intel_psr_enable_sink(intel_dp);
+
+	dev_priv->psr.enabled = intel_dp;
+unlock:
+	mutex_unlock(&dev_priv->psr.lock);
+}
+
+/**
+ * intel_psr_disable - Disable PSR
+ * @intel_dp: Intel DP
+ *
+ * This function needs to be called before disabling pipe.
+ */
+void intel_psr_disable(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = intel_dig_port->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	mutex_lock(&dev_priv->psr.lock);
+	if (!dev_priv->psr.enabled) {
+		mutex_unlock(&dev_priv->psr.lock);
+		return;
+	}
+
+	if (dev_priv->psr.active) {
+		I915_WRITE(EDP_PSR_CTL(dev),
+			   I915_READ(EDP_PSR_CTL(dev)) & ~EDP_PSR_ENABLE);
+
+		/* Wait till PSR is idle */
+		if (_wait_for((I915_READ(EDP_PSR_STATUS_CTL(dev)) &
+			       EDP_PSR_STATUS_STATE_MASK) == 0, 2000, 10))
+			DRM_ERROR("Timed out waiting for PSR Idle State\n");
+
+		dev_priv->psr.active = false;
+	} else {
+		WARN_ON(I915_READ(EDP_PSR_CTL(dev)) & EDP_PSR_ENABLE);
+	}
+
+	dev_priv->psr.enabled = NULL;
+	mutex_unlock(&dev_priv->psr.lock);
+
+	cancel_delayed_work_sync(&dev_priv->psr.work);
+}
+
+static void intel_psr_work(struct work_struct *work)
+{
+	struct drm_i915_private *dev_priv =
+		container_of(work, typeof(*dev_priv), psr.work.work);
+	struct intel_dp *intel_dp = dev_priv->psr.enabled;
+
+	/* We have to make sure PSR is ready for re-enable
+	 * otherwise it keeps disabled until next full enable/disable cycle.
+	 * PSR might take some time to get fully disabled
+	 * and be ready for re-enable.
+	 */
+	if (wait_for((I915_READ(EDP_PSR_STATUS_CTL(dev_priv->dev)) &
+		      EDP_PSR_STATUS_STATE_MASK) == 0, 50)) {
+		DRM_ERROR("Timed out waiting for PSR Idle for re-enable\n");
+		return;
+	}
+
+	mutex_lock(&dev_priv->psr.lock);
+	intel_dp = dev_priv->psr.enabled;
+
+	if (!intel_dp)
+		goto unlock;
+
+	/*
+	 * The delayed work can race with an invalidate hence we need to
+	 * recheck. Since psr_flush first clears this and then reschedules we
+	 * won't ever miss a flush when bailing out here.
+	 */
+	if (dev_priv->psr.busy_frontbuffer_bits)
+		goto unlock;
+
+	intel_psr_do_enable(intel_dp);
+unlock:
+	mutex_unlock(&dev_priv->psr.lock);
+}
+
+static void intel_psr_exit(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (dev_priv->psr.active) {
+		u32 val = I915_READ(EDP_PSR_CTL(dev));
+
+		WARN_ON(!(val & EDP_PSR_ENABLE));
+
+		I915_WRITE(EDP_PSR_CTL(dev), val & ~EDP_PSR_ENABLE);
+
+		dev_priv->psr.active = false;
+	}
+
+}
+
+/**
+ * intel_psr_invalidate - Invalidade PSR
+ * @dev: DRM device
+ * @frontbuffer_bits: frontbuffer plane tracking bits
+ *
+ * Since the hardware frontbuffer tracking has gaps we need to integrate
+ * with the software frontbuffer tracking. This function gets called every
+ * time frontbuffer rendering starts and a buffer gets dirtied. PSR must be
+ * disabled if the frontbuffer mask contains a buffer relevant to PSR.
+ *
+ * Dirty frontbuffers relevant to PSR are tracked in busy_frontbuffer_bits."
+ */
+void intel_psr_invalidate(struct drm_device *dev,
+			      unsigned frontbuffer_bits)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_crtc *crtc;
+	enum pipe pipe;
+
+	mutex_lock(&dev_priv->psr.lock);
+	if (!dev_priv->psr.enabled) {
+		mutex_unlock(&dev_priv->psr.lock);
+		return;
+	}
+
+	crtc = dp_to_dig_port(dev_priv->psr.enabled)->base.base.crtc;
+	pipe = to_intel_crtc(crtc)->pipe;
+
+	intel_psr_exit(dev);
+
+	frontbuffer_bits &= INTEL_FRONTBUFFER_ALL_MASK(pipe);
+
+	dev_priv->psr.busy_frontbuffer_bits |= frontbuffer_bits;
+	mutex_unlock(&dev_priv->psr.lock);
+}
+
+/**
+ * intel_psr_flush - Flush PSR
+ * @dev: DRM device
+ * @frontbuffer_bits: frontbuffer plane tracking bits
+ *
+ * Since the hardware frontbuffer tracking has gaps we need to integrate
+ * with the software frontbuffer tracking. This function gets called every
+ * time frontbuffer rendering has completed and flushed out to memory. PSR
+ * can be enabled again if no other frontbuffer relevant to PSR is dirty.
+ *
+ * Dirty frontbuffers relevant to PSR are tracked in busy_frontbuffer_bits.
+ */
+void intel_psr_flush(struct drm_device *dev,
+			 unsigned frontbuffer_bits)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_crtc *crtc;
+	enum pipe pipe;
+
+	mutex_lock(&dev_priv->psr.lock);
+	if (!dev_priv->psr.enabled) {
+		mutex_unlock(&dev_priv->psr.lock);
+		return;
+	}
+
+	crtc = dp_to_dig_port(dev_priv->psr.enabled)->base.base.crtc;
+	pipe = to_intel_crtc(crtc)->pipe;
+	dev_priv->psr.busy_frontbuffer_bits &= ~frontbuffer_bits;
+
+	/*
+	 * On Haswell sprite plane updates don't result in a psr invalidating
+	 * signal in the hardware. Which means we need to manually fake this in
+	 * software for all flushes, not just when we've seen a preceding
+	 * invalidation through frontbuffer rendering.
+	 */
+	if (IS_HASWELL(dev) &&
+	    (frontbuffer_bits & INTEL_FRONTBUFFER_SPRITE(pipe)))
+		intel_psr_exit(dev);
+
+	if (!dev_priv->psr.active && !dev_priv->psr.busy_frontbuffer_bits)
+		schedule_delayed_work(&dev_priv->psr.work,
+				      msecs_to_jiffies(100));
+	mutex_unlock(&dev_priv->psr.lock);
+}
+
+/**
+ * intel_psr_init - Init basic PSR work and mutex.
+ * @dev: DRM device
+ *
+ * This function is  called only once at driver load to initialize basic
+ * PSR stuff.
+ */
+void intel_psr_init(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	INIT_DELAYED_WORK(&dev_priv->psr.work, intel_psr_work);
+	mutex_init(&dev_priv->psr.lock);
+}
diff --git a/drivers/gpu/drm/i915/intel_renderstate.h b/drivers/gpu/drm/i915/intel_renderstate.h
index fd4f66231d30..5bd69852752c 100644
--- a/drivers/gpu/drm/i915/intel_renderstate.h
+++ b/drivers/gpu/drm/i915/intel_renderstate.h
@@ -24,17 +24,12 @@
 #ifndef _INTEL_RENDERSTATE_H
 #define _INTEL_RENDERSTATE_H
 
-#include <linux/types.h>
-
-struct intel_renderstate_rodata {
-	const u32 *reloc;
-	const u32 *batch;
-	const u32 batch_items;
-};
+#include "i915_drv.h"
 
 extern const struct intel_renderstate_rodata gen6_null_state;
 extern const struct intel_renderstate_rodata gen7_null_state;
 extern const struct intel_renderstate_rodata gen8_null_state;
+extern const struct intel_renderstate_rodata gen9_null_state;
 
 #define RO_RENDERSTATE(_g)						\
 	const struct intel_renderstate_rodata gen ## _g ## _null_state = { \
diff --git a/drivers/gpu/drm/i915/intel_renderstate_gen8.c b/drivers/gpu/drm/i915/intel_renderstate_gen8.c
index 75ef1b5de45c..78011d73fa9f 100644
--- a/drivers/gpu/drm/i915/intel_renderstate_gen8.c
+++ b/drivers/gpu/drm/i915/intel_renderstate_gen8.c
@@ -1,16 +1,134 @@
 #include "intel_renderstate.h"
 
 static const u32 gen8_null_state_relocs[] = {
-	0x00000048,
-	0x00000050,
-	0x00000060,
-	0x000003ec,
+	0x00000798,
+	0x000007a4,
+	0x000007ac,
+	0x000007bc,
 	-1,
 };
 
 static const u32 gen8_null_state_batch[] = {
+	0x7a000004,
+	0x01000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x69040000,
-	0x61020001,
+	0x78140000,
+	0x04000000,
+	0x7820000a,
+	0x00000000,
+	0x00000000,
+	0x80000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78130002,
+	0x00000000,
+	0x00000000,
+	0x02001808,
+	0x781f0002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78510009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78100007,
+	0x00000000,
+	0x00000000,
+	0x00010000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x781b0007,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000800,
+	0x00000000,
+	0x78110008,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x781e0003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x781d0007,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78120002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78500003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x781c0002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x780c0000,
+	0x00000000,
+	0x78520003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78300000,
+	0x08010040,
+	0x78310000,
+	0x1e000000,
+	0x78320000,
+	0x1e000000,
+	0x78330000,
+	0x1e000000,
+	0x79190002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x791a0002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x791b0002,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x79120000,
@@ -23,48 +141,435 @@ static const u32 gen8_null_state_batch[] = {
 	0x00000000,
 	0x79160000,
 	0x00000000,
-	0x6101000e,
-	0x00000001,
+	0x78150009,
 	0x00000000,
-	0x00000001,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78190009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x781a0009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78160009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78170009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78490001,
+	0x00000000,
+	0x00000000,
+	0x784a0000,
+	0x00000000,
+	0x784b0000,
+	0x00000004,
+	0x79170101,
+	0x00000000,
+	0x00000080,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x79180006,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x79180006,
+	0x20000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x79180006,
+	0x40000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x79180006,
+	0x60000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x6101000e,
 	0x00000001,	 /* reloc */
 	0x00000000,
+	0x00000000,
 	0x00000001,	 /* reloc */
 	0x00000000,
+	0x00000001,	 /* reloc */
 	0x00000000,
+	0x00000001,
 	0x00000000,
 	0x00000001,	 /* reloc */
 	0x00000000,
-	0xfffff001,
 	0x00001001,
-	0xfffff001,
 	0x00001001,
-	0x78230000,
-	0x000006e0,
-	0x78210000,
-	0x00000700,
-	0x78300000,
-	0x08010040,
-	0x78330000,
-	0x08000000,
-	0x78310000,
-	0x08000000,
-	0x78320000,
-	0x08000000,
-	0x78240000,
-	0x00000641,
-	0x780e0000,
-	0x00000601,
+	0x00000001,
+	0x00001001,
+	0x61020001,
+	0x00000000,
+	0x00000000,
+	0x79000002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78050006,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x79040002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x79040002,
+	0x40000000,
+	0x00000000,
+	0x00000000,
+	0x79040002,
+	0x80000000,
+	0x00000000,
+	0x00000000,
+	0x79040002,
+	0xc0000000,
+	0x00000000,
+	0x00000000,
+	0x79080001,
+	0x00000000,
+	0x00000000,
+	0x790a0001,
+	0x00000000,
+	0x00000000,
+	0x78060003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78070003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78040001,
+	0x00000000,
+	0x00000000,
+	0x79110000,
+	0x00000000,
 	0x780d0000,
 	0x00000000,
-	0x78180000,
-	0x00000001,
-	0x78520003,
+	0x79060000,
 	0x00000000,
+	0x7907001f,
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x78190009,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -75,7 +580,6 @@ static const u32 gen8_null_state_batch[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x781b0007,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -84,26 +588,22 @@ static const u32 gen8_null_state_batch[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x78270000,
 	0x00000000,
-	0x782c0000,
 	0x00000000,
-	0x781c0002,
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x78160009,
 	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x7902000f,
 	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x78110008,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -113,12 +613,10 @@ static const u32 gen8_null_state_batch[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x78290000,
 	0x00000000,
-	0x782e0000,
 	0x00000000,
-	0x781a0009,
 	0x00000000,
+	0x790c000f,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -128,7 +626,6 @@ static const u32 gen8_null_state_batch[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x781d0007,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -136,153 +633,153 @@ static const u32 gen8_null_state_batch[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x780a0003,
 	0x00000000,
-	0x78280000,
 	0x00000000,
-	0x782d0000,
 	0x00000000,
-	0x78260000,
 	0x00000000,
-	0x782b0000,
+	0x78080083,
+	0x00004000,
 	0x00000000,
-	0x78150009,
 	0x00000000,
 	0x00000000,
+	0x04004000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x08004000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x0c004000,
 	0x00000000,
 	0x00000000,
-	0x78100007,
 	0x00000000,
+	0x10004000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x14004000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x18004000,
 	0x00000000,
-	0x781e0003,
 	0x00000000,
 	0x00000000,
+	0x1c004000,
 	0x00000000,
 	0x00000000,
-	0x78120002,
 	0x00000000,
+	0x20004000,
 	0x00000000,
 	0x00000000,
-	0x781f0002,
-	0x30400820,
 	0x00000000,
+	0x24004000,
 	0x00000000,
-	0x78510009,
 	0x00000000,
 	0x00000000,
+	0x28004000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x2c004000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x30004000,
 	0x00000000,
 	0x00000000,
-	0x78500003,
-	0x00210000,
 	0x00000000,
+	0x34004000,
 	0x00000000,
 	0x00000000,
-	0x78130002,
 	0x00000000,
+	0x38004000,
 	0x00000000,
 	0x00000000,
-	0x782a0000,
-	0x00000480,
-	0x782f0000,
-	0x00000540,
-	0x78140000,
-	0x00000800,
-	0x78170009,
 	0x00000000,
+	0x3c004000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x40004000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x44004000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x7820000a,
-	0x00000580,
+	0x48004000,
 	0x00000000,
-	0x08080000,
 	0x00000000,
 	0x00000000,
-	0x1f000002,
-	0x00060000,
+	0x4c004000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x50004000,
 	0x00000000,
-	0x784d0000,
-	0x40000000,
-	0x784f0000,
-	0x80000100,
-	0x780f0000,
-	0x00000740,
-	0x78050006,
 	0x00000000,
 	0x00000000,
+	0x54004000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x58004000,
 	0x00000000,
 	0x00000000,
-	0x78070003,
 	0x00000000,
+	0x5c004000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x78060003,
+	0x60004000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x64004000,
 	0x00000000,
-	0x78040001,
 	0x00000000,
-	0x00000001,
-	0x79000002,
-	0xffffffff,
+	0x00000000,
+	0x68004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x6c004000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x78080003,
-	0x00006000,
-	0x000005e0,	 /* reloc */
+	0x70004000,
 	0x00000000,
 	0x00000000,
-	0x78090005,
+	0x00000000,
+	0x74004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x7c004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x80004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78090043,
 	0x02000000,
 	0x22220000,
-	0x02f60000,
-	0x11230000,
-	0x02850004,
-	0x11230000,
-	0x784b0000,
-	0x0000000f,
-	0x78490001,
 	0x00000000,
 	0x00000000,
-	0x7b000005,
 	0x00000000,
-	0x00000003,
 	0x00000000,
-	0x00000001,
 	0x00000000,
 	0x00000000,
-	0x05000000,	 /* cmds end */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -297,8 +794,6 @@ static const u32 gen8_null_state_batch[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x000004c0,	 /* state start */
-	0x00000500,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -345,46 +840,65 @@ static const u32 gen8_null_state_batch[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x680b0001,
+	0x78260000,
+	0x00000000,
+	0x78270000,
+	0x00000000,
+	0x78280000,
+	0x00000000,
+	0x78290000,
+	0x00000000,
+	0x782a0000,
+	0x00000000,
+	0x780e0000,
+	0x00000dc1,
+	0x78240000,
+	0x00000e01,
+	0x784f0000,
+	0x80000100,
+	0x784d0000,
+	0x40000000,
+	0x782b0000,
+	0x00000000,
+	0x782c0000,
+	0x00000000,
+	0x782d0000,
 	0x00000000,
+	0x782e0000,
 	0x00000000,
+	0x782f0000,
 	0x00000000,
-	0x00000092,
+	0x780f0000,
 	0x00000000,
+	0x78230000,
+	0x00000e60,
+	0x78210000,
+	0x00000e80,
+	0x7b000005,
+	0x00000004,
+	0x00000001,
 	0x00000000,
+	0x00000001,
 	0x00000000,
 	0x00000000,
+	0x05000000,	 /* cmds end */
 	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x00000000,	 /* state start */
+	0x00000000,
+	0x3f800000,
+	0x3f800000,
+	0x3f800000,
+	0x3f800000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x0060005a,
-	0x21403ae8,
-	0x3a0000c0,
-	0x008d0040,
-	0x0060005a,
-	0x21603ae8,
-	0x3a0000c0,
-	0x008d0080,
-	0x0060005a,
-	0x21803ae8,
-	0x3a0000d0,
-	0x008d0040,
-	0x0060005a,
-	0x21a03ae8,
-	0x3a0000d0,
-	0x008d0080,
-	0x02800031,
-	0x2e0022e8,
-	0x0e000140,
-	0x08840001,
-	0x05800031,
-	0x200022e0,
-	0x0e000e00,
-	0x90031000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -410,38 +924,6 @@ static const u32 gen8_null_state_batch[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x06200000,
-	0x00000002,
-	0x06200000,
-	0x00000002,
-	0x06200000,
-	0x00000002,
-	0x06200000,
-	0x00000002,
-	0x06200000,
-	0x00000002,
-	0x06200000,
-	0x00000002,
-	0x06200000,
-	0x00000002,
-	0x06200000,
-	0x00000002,
-	0x06200000,
-	0x00000002,
-	0x06200000,
-	0x00000002,
-	0x06200000,
-	0x00000002,
-	0x06200000,
-	0x00000002,
-	0x06200000,
-	0x00000002,
-	0x06200000,
-	0x00000002,
-	0x06200000,
-	0x00000002,
-	0x06200000,
-	0x00000002,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -449,8 +931,6 @@ static const u32 gen8_null_state_batch[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0xf99a130c,
-	0x799a130c,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -466,9 +946,7 @@ static const u32 gen8_null_state_batch[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x3f800000,
 	0x00000000,
-	0x3f800000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/i915/intel_renderstate_gen9.c b/drivers/gpu/drm/i915/intel_renderstate_gen9.c
new file mode 100644
index 000000000000..875075373807
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_renderstate_gen9.c
@@ -0,0 +1,974 @@
+#include "intel_renderstate.h"
+
+static const u32 gen9_null_state_relocs[] = {
+	0x000007a8,
+	0x000007b4,
+	0x000007bc,
+	0x000007cc,
+	-1,
+};
+
+static const u32 gen9_null_state_batch[] = {
+	0x7a000004,
+	0x01000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x69040300,
+	0x78140000,
+	0x04000000,
+	0x7820000a,
+	0x00000000,
+	0x00000000,
+	0x80000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78130002,
+	0x00000000,
+	0x00000000,
+	0x02001808,
+	0x781f0004,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78510009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78100007,
+	0x00000000,
+	0x00000000,
+	0x00010000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x781b0007,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000800,
+	0x00000000,
+	0x78110008,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x781e0003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x781d0009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78120002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78500003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x781c0002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x780c0000,
+	0x00000000,
+	0x78520003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78300000,
+	0x08010040,
+	0x78310000,
+	0x1e000000,
+	0x78320000,
+	0x1e000000,
+	0x78330000,
+	0x1e000000,
+	0x79190002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x791a0002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x791b0002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x79120000,
+	0x00000000,
+	0x79130000,
+	0x00000000,
+	0x79140000,
+	0x00000000,
+	0x79150000,
+	0x00000000,
+	0x79160000,
+	0x00000000,
+	0x78150009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78190009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x781a0009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78160009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78170009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78490001,
+	0x00000000,
+	0x00000000,
+	0x784a0000,
+	0x00000000,
+	0x784b0000,
+	0x00000004,
+	0x79170101,
+	0x00000000,
+	0x00000080,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x79180006,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x79180006,
+	0x20000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x79180006,
+	0x40000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x79180006,
+	0x60000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x61010011,
+	0x00000001,	 /* reloc */
+	0x00000000,
+	0x00000000,
+	0x00000001,	 /* reloc */
+	0x00000000,
+	0x00000001,	 /* reloc */
+	0x00000000,
+	0x00000001,
+	0x00000000,
+	0x00000001,	 /* reloc */
+	0x00000000,
+	0x00001001,
+	0x00001001,
+	0x00000001,
+	0x00001001,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x61020001,
+	0x00000000,
+	0x00000000,
+	0x79000002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78050006,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x79040002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x79040002,
+	0x40000000,
+	0x00000000,
+	0x00000000,
+	0x79040002,
+	0x80000000,
+	0x00000000,
+	0x00000000,
+	0x79040002,
+	0xc0000000,
+	0x00000000,
+	0x00000000,
+	0x79080001,
+	0x00000000,
+	0x00000000,
+	0x790a0001,
+	0x00000000,
+	0x00000000,
+	0x78060003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78070003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78040001,
+	0x00000000,
+	0x00000000,
+	0x79110000,
+	0x00000000,
+	0x780d0000,
+	0x00000000,
+	0x79060000,
+	0x00000000,
+	0x7907001f,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x7902000f,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x790c000f,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x780a0003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78080083,
+	0x00004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x04004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x08004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x0c004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x10004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x14004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x18004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x1c004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x20004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x24004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x28004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x2c004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x30004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x34004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x38004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x3c004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x40004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x44004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x48004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x4c004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x50004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x54004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x58004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x5c004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x60004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x64004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x68004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x6c004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x70004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x74004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x7c004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x80004000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78090043,
+	0x02000000,
+	0x22220000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78550003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x680b0001,
+	0x780e0000,
+	0x00000e01,
+	0x78240000,
+	0x00000e41,
+	0x784f0000,
+	0x80000100,
+	0x784d0000,
+	0x40000000,
+	0x782b0000,
+	0x00000000,
+	0x782c0000,
+	0x00000000,
+	0x782d0000,
+	0x00000000,
+	0x782e0000,
+	0x00000000,
+	0x782f0000,
+	0x00000000,
+	0x780f0000,
+	0x00000000,
+	0x78230000,
+	0x00000ea0,
+	0x78210000,
+	0x00000ec0,
+	0x78260000,
+	0x00000000,
+	0x78270000,
+	0x00000000,
+	0x78280000,
+	0x00000000,
+	0x78290000,
+	0x00000000,
+	0x782a0000,
+	0x00000000,
+	0x7b000005,
+	0x00000004,
+	0x00000001,
+	0x00000000,
+	0x00000001,
+	0x00000000,
+	0x00000000,
+	0x05000000,	 /* cmds end */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,	 /* state start */
+	0x00000000,
+	0x3f800000,
+	0x3f800000,
+	0x3f800000,
+	0x3f800000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,	 /* state end */
+};
+
+RO_RENDERSTATE(9);
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 47a126a0493f..c7bc93d28d84 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -33,14 +33,24 @@
 #include "i915_trace.h"
 #include "intel_drv.h"
 
-/* Early gen2 devices have a cacheline of just 32 bytes, using 64 is overkill,
- * but keeps the logic simple. Indeed, the whole purpose of this macro is just
- * to give some inclination as to some of the magic values used in the various
- * workarounds!
- */
-#define CACHELINE_BYTES 64
+bool
+intel_ring_initialized(struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+
+	if (!dev)
+		return false;
 
-static inline int __ring_space(int head, int tail, int size)
+	if (i915.enable_execlists) {
+		struct intel_context *dctx = ring->default_context;
+		struct intel_ringbuffer *ringbuf = dctx->engine[ring->id].ringbuf;
+
+		return ringbuf->obj;
+	} else
+		return ring->buffer && ring->buffer->obj;
+}
+
+int __intel_ring_space(int head, int tail, int size)
 {
 	int space = head - (tail + I915_RING_FREE_SPACE);
 	if (space < 0)
@@ -48,12 +58,13 @@ static inline int __ring_space(int head, int tail, int size)
 	return space;
 }
 
-static inline int ring_space(struct intel_ringbuffer *ringbuf)
+int intel_ring_space(struct intel_ringbuffer *ringbuf)
 {
-	return __ring_space(ringbuf->head & HEAD_ADDR, ringbuf->tail, ringbuf->size);
+	return __intel_ring_space(ringbuf->head & HEAD_ADDR,
+				  ringbuf->tail, ringbuf->size);
 }
 
-static bool intel_ring_stopped(struct intel_engine_cs *ring)
+bool intel_ring_stopped(struct intel_engine_cs *ring)
 {
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 	return dev_priv->gpu_error.stop_rings & intel_ring_flag(ring);
@@ -351,12 +362,15 @@ gen7_render_ring_flush(struct intel_engine_cs *ring,
 		flags |= PIPE_CONTROL_VF_CACHE_INVALIDATE;
 		flags |= PIPE_CONTROL_CONST_CACHE_INVALIDATE;
 		flags |= PIPE_CONTROL_STATE_CACHE_INVALIDATE;
+		flags |= PIPE_CONTROL_MEDIA_STATE_CLEAR;
 		/*
 		 * TLB invalidate requires a post-sync write.
 		 */
 		flags |= PIPE_CONTROL_QW_WRITE;
 		flags |= PIPE_CONTROL_GLOBAL_GTT_IVB;
 
+		flags |= PIPE_CONTROL_STALL_AT_SCOREBOARD;
+
 		/* Workaround: we must issue a pipe_control with CS-stall bit
 		 * set before a pipe_control command that has the state cache
 		 * invalidate bit set. */
@@ -433,7 +447,14 @@ gen8_render_ring_flush(struct intel_engine_cs *ring,
 			return ret;
 	}
 
-	return gen8_emit_pipe_control(ring, flags, scratch_addr);
+	ret = gen8_emit_pipe_control(ring, flags, scratch_addr);
+	if (ret)
+		return ret;
+
+	if (!invalidate_domains && flush_domains)
+		return gen7_ring_fbc_flush(ring, FBC_REND_NUKE);
+
+	return 0;
 }
 
 static void ring_write_tail(struct intel_engine_cs *ring,
@@ -476,9 +497,14 @@ static bool stop_ring(struct intel_engine_cs *ring)
 
 	if (!IS_GEN2(ring->dev)) {
 		I915_WRITE_MODE(ring, _MASKED_BIT_ENABLE(STOP_RING));
-		if (wait_for_atomic((I915_READ_MODE(ring) & MODE_IDLE) != 0, 1000)) {
-			DRM_ERROR("%s :timed out trying to stop ring\n", ring->name);
-			return false;
+		if (wait_for((I915_READ_MODE(ring) & MODE_IDLE) != 0, 1000)) {
+			DRM_ERROR("%s : timed out trying to stop ring\n", ring->name);
+			/* Sometimes we observe that the idle flag is not
+			 * set even though the ring is empty. So double
+			 * check before giving up.
+			 */
+			if (I915_READ_HEAD(ring) != I915_READ_TAIL(ring))
+				return false;
 		}
 	}
 
@@ -540,6 +566,14 @@ static int init_ring_common(struct intel_engine_cs *ring)
 	 * also enforces ordering), otherwise the hw might lose the new ring
 	 * register values. */
 	I915_WRITE_START(ring, i915_gem_obj_ggtt_offset(obj));
+
+	/* WaClearRingBufHeadRegAtInit:ctg,elk */
+	if (I915_READ_HEAD(ring))
+		DRM_DEBUG("%s initialization failed [head=%08x], fudging\n",
+			  ring->name, I915_READ_HEAD(ring));
+	I915_WRITE_HEAD(ring, 0);
+	(void)I915_READ_HEAD(ring);
+
 	I915_WRITE_CTL(ring,
 			((ringbuf->size - PAGE_SIZE) & RING_NR_PAGES)
 			| RING_VALID);
@@ -558,14 +592,10 @@ static int init_ring_common(struct intel_engine_cs *ring)
 		goto out;
 	}
 
-	if (!drm_core_check_feature(ring->dev, DRIVER_MODESET))
-		i915_kernel_lost_context(ring->dev);
-	else {
-		ringbuf->head = I915_READ_HEAD(ring);
-		ringbuf->tail = I915_READ_TAIL(ring) & TAIL_ADDR;
-		ringbuf->space = ring_space(ringbuf);
-		ringbuf->last_retired_head = -1;
-	}
+	ringbuf->head = I915_READ_HEAD(ring);
+	ringbuf->tail = I915_READ_TAIL(ring) & TAIL_ADDR;
+	ringbuf->space = intel_ring_space(ringbuf);
+	ringbuf->last_retired_head = -1;
 
 	memset(&ring->hangcheck, 0, sizeof(ring->hangcheck));
 
@@ -575,8 +605,25 @@ out:
 	return ret;
 }
 
-static int
-init_pipe_control(struct intel_engine_cs *ring)
+void
+intel_fini_pipe_control(struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+
+	if (ring->scratch.obj == NULL)
+		return;
+
+	if (INTEL_INFO(dev)->gen >= 5) {
+		kunmap(sg_page(ring->scratch.obj->pages->sgl));
+		i915_gem_object_ggtt_unpin(ring->scratch.obj);
+	}
+
+	drm_gem_object_unreference(&ring->scratch.obj->base);
+	ring->scratch.obj = NULL;
+}
+
+int
+intel_init_pipe_control(struct intel_engine_cs *ring)
 {
 	int ret;
 
@@ -617,6 +664,170 @@ err:
 	return ret;
 }
 
+static int intel_ring_workarounds_emit(struct intel_engine_cs *ring,
+				       struct intel_context *ctx)
+{
+	int ret, i;
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct i915_workarounds *w = &dev_priv->workarounds;
+
+	if (WARN_ON(w->count == 0))
+		return 0;
+
+	ring->gpu_caches_dirty = true;
+	ret = intel_ring_flush_all_caches(ring);
+	if (ret)
+		return ret;
+
+	ret = intel_ring_begin(ring, (w->count * 2 + 2));
+	if (ret)
+		return ret;
+
+	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(w->count));
+	for (i = 0; i < w->count; i++) {
+		intel_ring_emit(ring, w->reg[i].addr);
+		intel_ring_emit(ring, w->reg[i].value);
+	}
+	intel_ring_emit(ring, MI_NOOP);
+
+	intel_ring_advance(ring);
+
+	ring->gpu_caches_dirty = true;
+	ret = intel_ring_flush_all_caches(ring);
+	if (ret)
+		return ret;
+
+	DRM_DEBUG_DRIVER("Number of Workarounds emitted: %d\n", w->count);
+
+	return 0;
+}
+
+static int wa_add(struct drm_i915_private *dev_priv,
+		  const u32 addr, const u32 mask, const u32 val)
+{
+	const u32 idx = dev_priv->workarounds.count;
+
+	if (WARN_ON(idx >= I915_MAX_WA_REGS))
+		return -ENOSPC;
+
+	dev_priv->workarounds.reg[idx].addr = addr;
+	dev_priv->workarounds.reg[idx].value = val;
+	dev_priv->workarounds.reg[idx].mask = mask;
+
+	dev_priv->workarounds.count++;
+
+	return 0;
+}
+
+#define WA_REG(addr, mask, val) { \
+		const int r = wa_add(dev_priv, (addr), (mask), (val)); \
+		if (r) \
+			return r; \
+	}
+
+#define WA_SET_BIT_MASKED(addr, mask) \
+	WA_REG(addr, (mask), _MASKED_BIT_ENABLE(mask))
+
+#define WA_CLR_BIT_MASKED(addr, mask) \
+	WA_REG(addr, (mask), _MASKED_BIT_DISABLE(mask))
+
+#define WA_SET_FIELD_MASKED(addr, mask, value) \
+	WA_REG(addr, mask, _MASKED_FIELD(mask, value))
+
+#define WA_SET_BIT(addr, mask) WA_REG(addr, mask, I915_READ(addr) | (mask))
+#define WA_CLR_BIT(addr, mask) WA_REG(addr, mask, I915_READ(addr) & ~(mask))
+
+#define WA_WRITE(addr, val) WA_REG(addr, 0xffffffff, val)
+
+static int bdw_init_workarounds(struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	/* WaDisablePartialInstShootdown:bdw */
+	/* WaDisableThreadStallDopClockGating:bdw (pre-production) */
+	WA_SET_BIT_MASKED(GEN8_ROW_CHICKEN,
+			  PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE |
+			  STALL_DOP_GATING_DISABLE);
+
+	/* WaDisableDopClockGating:bdw */
+	WA_SET_BIT_MASKED(GEN7_ROW_CHICKEN2,
+			  DOP_CLOCK_GATING_DISABLE);
+
+	WA_SET_BIT_MASKED(HALF_SLICE_CHICKEN3,
+			  GEN8_SAMPLER_POWER_BYPASS_DIS);
+
+	/* Use Force Non-Coherent whenever executing a 3D context. This is a
+	 * workaround for for a possible hang in the unlikely event a TLB
+	 * invalidation occurs during a PSD flush.
+	 */
+	/* WaDisableFenceDestinationToSLM:bdw (GT3 pre-production) */
+	WA_SET_BIT_MASKED(HDC_CHICKEN0,
+			  HDC_FORCE_NON_COHERENT |
+			  (IS_BDW_GT3(dev) ? HDC_FENCE_DEST_SLM_DISABLE : 0));
+
+	/* Wa4x4STCOptimizationDisable:bdw */
+	WA_SET_BIT_MASKED(CACHE_MODE_1,
+			  GEN8_4x4_STC_OPTIMIZATION_DISABLE);
+
+	/*
+	 * BSpec recommends 8x4 when MSAA is used,
+	 * however in practice 16x4 seems fastest.
+	 *
+	 * Note that PS/WM thread counts depend on the WIZ hashing
+	 * disable bit, which we don't touch here, but it's good
+	 * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM).
+	 */
+	WA_SET_FIELD_MASKED(GEN7_GT_MODE,
+			    GEN6_WIZ_HASHING_MASK,
+			    GEN6_WIZ_HASHING_16x4);
+
+	return 0;
+}
+
+static int chv_init_workarounds(struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	/* WaDisablePartialInstShootdown:chv */
+	/* WaDisableThreadStallDopClockGating:chv */
+	WA_SET_BIT_MASKED(GEN8_ROW_CHICKEN,
+			  PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE |
+			  STALL_DOP_GATING_DISABLE);
+
+	/* Use Force Non-Coherent whenever executing a 3D context. This is a
+	 * workaround for a possible hang in the unlikely event a TLB
+	 * invalidation occurs during a PSD flush.
+	 */
+	/* WaForceEnableNonCoherent:chv */
+	/* WaHdcDisableFetchWhenMasked:chv */
+	WA_SET_BIT_MASKED(HDC_CHICKEN0,
+			  HDC_FORCE_NON_COHERENT |
+			  HDC_DONOT_FETCH_MEM_WHEN_MASKED);
+
+	return 0;
+}
+
+int init_workarounds_ring(struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	WARN_ON(ring->id != RCS);
+
+	dev_priv->workarounds.count = 0;
+
+	if (IS_BROADWELL(dev))
+		return bdw_init_workarounds(ring);
+
+	if (IS_CHERRYVIEW(dev))
+		return chv_init_workarounds(ring);
+
+	return 0;
+}
+
 static int init_render_ring(struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ring->dev;
@@ -635,7 +846,7 @@ static int init_render_ring(struct intel_engine_cs *ring)
 	 *
 	 * WaDisableAsyncFlipPerfMode:snb,ivb,hsw,vlv,bdw,chv
 	 */
-	if (INTEL_INFO(dev)->gen >= 6)
+	if (INTEL_INFO(dev)->gen >= 6 && INTEL_INFO(dev)->gen < 9)
 		I915_WRITE(MI_MODE, _MASKED_BIT_ENABLE(ASYNC_FLIP_PERF_DISABLE));
 
 	/* Required for the hardware to program scanline values for waiting */
@@ -651,7 +862,7 @@ static int init_render_ring(struct intel_engine_cs *ring)
 			   _MASKED_BIT_ENABLE(GFX_REPLAY_MODE));
 
 	if (INTEL_INFO(dev)->gen >= 5) {
-		ret = init_pipe_control(ring);
+		ret = intel_init_pipe_control(ring);
 		if (ret)
 			return ret;
 	}
@@ -672,7 +883,7 @@ static int init_render_ring(struct intel_engine_cs *ring)
 	if (HAS_L3_DPF(dev))
 		I915_WRITE_IMR(ring, ~GT_PARITY_ERROR(dev));
 
-	return ret;
+	return init_workarounds_ring(ring);
 }
 
 static void render_ring_cleanup(struct intel_engine_cs *ring)
@@ -686,16 +897,7 @@ static void render_ring_cleanup(struct intel_engine_cs *ring)
 		dev_priv->semaphore_obj = NULL;
 	}
 
-	if (ring->scratch.obj == NULL)
-		return;
-
-	if (INTEL_INFO(dev)->gen >= 5) {
-		kunmap(sg_page(ring->scratch.obj->pages->sgl));
-		i915_gem_object_ggtt_unpin(ring->scratch.obj);
-	}
-
-	drm_gem_object_unreference(&ring->scratch.obj->base);
-	ring->scratch.obj = NULL;
+	intel_fini_pipe_control(ring);
 }
 
 static int gen8_rcs_signal(struct intel_engine_cs *signaller,
@@ -1018,7 +1220,7 @@ gen5_ring_get_irq(struct intel_engine_cs *ring)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long flags;
 
-	if (!dev->irq_enabled)
+	if (WARN_ON(!intel_irqs_enabled(dev_priv)))
 		return false;
 
 	spin_lock_irqsave(&dev_priv->irq_lock, flags);
@@ -1049,7 +1251,7 @@ i9xx_ring_get_irq(struct intel_engine_cs *ring)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long flags;
 
-	if (!dev->irq_enabled)
+	if (!intel_irqs_enabled(dev_priv))
 		return false;
 
 	spin_lock_irqsave(&dev_priv->irq_lock, flags);
@@ -1086,7 +1288,7 @@ i8xx_ring_get_irq(struct intel_engine_cs *ring)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long flags;
 
-	if (!dev->irq_enabled)
+	if (!intel_irqs_enabled(dev_priv))
 		return false;
 
 	spin_lock_irqsave(&dev_priv->irq_lock, flags);
@@ -1220,8 +1422,8 @@ gen6_ring_get_irq(struct intel_engine_cs *ring)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long flags;
 
-	if (!dev->irq_enabled)
-	       return false;
+	if (WARN_ON(!intel_irqs_enabled(dev_priv)))
+		return false;
 
 	spin_lock_irqsave(&dev_priv->irq_lock, flags);
 	if (ring->irq_refcount++ == 0) {
@@ -1263,7 +1465,7 @@ hsw_vebox_get_irq(struct intel_engine_cs *ring)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long flags;
 
-	if (!dev->irq_enabled)
+	if (WARN_ON(!intel_irqs_enabled(dev_priv)))
 		return false;
 
 	spin_lock_irqsave(&dev_priv->irq_lock, flags);
@@ -1283,9 +1485,6 @@ hsw_vebox_put_irq(struct intel_engine_cs *ring)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long flags;
 
-	if (!dev->irq_enabled)
-		return;
-
 	spin_lock_irqsave(&dev_priv->irq_lock, flags);
 	if (--ring->irq_refcount == 0) {
 		I915_WRITE_IMR(ring, ~0);
@@ -1301,7 +1500,7 @@ gen8_ring_get_irq(struct intel_engine_cs *ring)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long flags;
 
-	if (!dev->irq_enabled)
+	if (WARN_ON(!intel_irqs_enabled(dev_priv)))
 		return false;
 
 	spin_lock_irqsave(&dev_priv->irq_lock, flags);
@@ -1526,26 +1725,50 @@ static int init_phys_status_page(struct intel_engine_cs *ring)
 	return 0;
 }
 
-static void intel_destroy_ringbuffer_obj(struct intel_ringbuffer *ringbuf)
+void intel_unpin_ringbuffer_obj(struct intel_ringbuffer *ringbuf)
 {
-	if (!ringbuf->obj)
-		return;
-
 	iounmap(ringbuf->virtual_start);
+	ringbuf->virtual_start = NULL;
 	i915_gem_object_ggtt_unpin(ringbuf->obj);
-	drm_gem_object_unreference(&ringbuf->obj->base);
-	ringbuf->obj = NULL;
 }
 
-static int intel_alloc_ringbuffer_obj(struct drm_device *dev,
-				      struct intel_ringbuffer *ringbuf)
+int intel_pin_and_map_ringbuffer_obj(struct drm_device *dev,
+				     struct intel_ringbuffer *ringbuf)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct drm_i915_gem_object *obj;
+	struct drm_i915_gem_object *obj = ringbuf->obj;
 	int ret;
 
-	if (ringbuf->obj)
-		return 0;
+	ret = i915_gem_obj_ggtt_pin(obj, PAGE_SIZE, PIN_MAPPABLE);
+	if (ret)
+		return ret;
+
+	ret = i915_gem_object_set_to_gtt_domain(obj, true);
+	if (ret) {
+		i915_gem_object_ggtt_unpin(obj);
+		return ret;
+	}
+
+	ringbuf->virtual_start = ioremap_wc(dev_priv->gtt.mappable_base +
+			i915_gem_obj_ggtt_offset(obj), ringbuf->size);
+	if (ringbuf->virtual_start == NULL) {
+		i915_gem_object_ggtt_unpin(obj);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+void intel_destroy_ringbuffer_obj(struct intel_ringbuffer *ringbuf)
+{
+	drm_gem_object_unreference(&ringbuf->obj->base);
+	ringbuf->obj = NULL;
+}
+
+int intel_alloc_ringbuffer_obj(struct drm_device *dev,
+			       struct intel_ringbuffer *ringbuf)
+{
+	struct drm_i915_gem_object *obj;
 
 	obj = NULL;
 	if (!HAS_LLC(dev))
@@ -1558,30 +1781,9 @@ static int intel_alloc_ringbuffer_obj(struct drm_device *dev,
 	/* mark ring buffers as read-only from GPU side by default */
 	obj->gt_ro = 1;
 
-	ret = i915_gem_obj_ggtt_pin(obj, PAGE_SIZE, PIN_MAPPABLE);
-	if (ret)
-		goto err_unref;
-
-	ret = i915_gem_object_set_to_gtt_domain(obj, true);
-	if (ret)
-		goto err_unpin;
-
-	ringbuf->virtual_start =
-		ioremap_wc(dev_priv->gtt.mappable_base + i915_gem_obj_ggtt_offset(obj),
-				ringbuf->size);
-	if (ringbuf->virtual_start == NULL) {
-		ret = -EINVAL;
-		goto err_unpin;
-	}
-
 	ringbuf->obj = obj;
-	return 0;
 
-err_unpin:
-	i915_gem_object_ggtt_unpin(obj);
-err_unref:
-	drm_gem_object_unreference(&obj->base);
-	return ret;
+	return 0;
 }
 
 static int intel_init_ring_buffer(struct drm_device *dev,
@@ -1600,7 +1802,9 @@ static int intel_init_ring_buffer(struct drm_device *dev,
 	ring->dev = dev;
 	INIT_LIST_HEAD(&ring->active_list);
 	INIT_LIST_HEAD(&ring->request_list);
+	INIT_LIST_HEAD(&ring->execlist_queue);
 	ringbuf->size = 32 * PAGE_SIZE;
+	ringbuf->ring = ring;
 	memset(ring->semaphore.sync_seqno, 0, sizeof(ring->semaphore.sync_seqno));
 
 	init_waitqueue_head(&ring->irq_queue);
@@ -1616,10 +1820,21 @@ static int intel_init_ring_buffer(struct drm_device *dev,
 			goto error;
 	}
 
-	ret = intel_alloc_ringbuffer_obj(dev, ringbuf);
-	if (ret) {
-		DRM_ERROR("Failed to allocate ringbuffer %s: %d\n", ring->name, ret);
-		goto error;
+	if (ringbuf->obj == NULL) {
+		ret = intel_alloc_ringbuffer_obj(dev, ringbuf);
+		if (ret) {
+			DRM_ERROR("Failed to allocate ringbuffer %s: %d\n",
+					ring->name, ret);
+			goto error;
+		}
+
+		ret = intel_pin_and_map_ringbuffer_obj(dev, ringbuf);
+		if (ret) {
+			DRM_ERROR("Failed to pin and map ringbuffer %s: %d\n",
+					ring->name, ret);
+			intel_destroy_ringbuffer_obj(ringbuf);
+			goto error;
+		}
 	}
 
 	/* Workaround an erratum on the i830 which causes a hang if
@@ -1648,15 +1863,19 @@ error:
 
 void intel_cleanup_ring_buffer(struct intel_engine_cs *ring)
 {
-	struct drm_i915_private *dev_priv = to_i915(ring->dev);
-	struct intel_ringbuffer *ringbuf = ring->buffer;
+	struct drm_i915_private *dev_priv;
+	struct intel_ringbuffer *ringbuf;
 
 	if (!intel_ring_initialized(ring))
 		return;
 
+	dev_priv = to_i915(ring->dev);
+	ringbuf = ring->buffer;
+
 	intel_stop_ring_buffer(ring);
 	WARN_ON(!IS_GEN2(ring->dev) && (I915_READ_MODE(ring) & MODE_IDLE) == 0);
 
+	intel_unpin_ringbuffer_obj(ringbuf);
 	intel_destroy_ringbuffer_obj(ringbuf);
 	ring->preallocated_lazy_request = NULL;
 	ring->outstanding_lazy_seqno = 0;
@@ -1683,13 +1902,14 @@ static int intel_ring_wait_request(struct intel_engine_cs *ring, int n)
 		ringbuf->head = ringbuf->last_retired_head;
 		ringbuf->last_retired_head = -1;
 
-		ringbuf->space = ring_space(ringbuf);
+		ringbuf->space = intel_ring_space(ringbuf);
 		if (ringbuf->space >= n)
 			return 0;
 	}
 
 	list_for_each_entry(request, &ring->request_list, list) {
-		if (__ring_space(request->tail, ringbuf->tail, ringbuf->size) >= n) {
+		if (__intel_ring_space(request->tail, ringbuf->tail,
+				       ringbuf->size) >= n) {
 			seqno = request->seqno;
 			break;
 		}
@@ -1706,7 +1926,7 @@ static int intel_ring_wait_request(struct intel_engine_cs *ring, int n)
 	ringbuf->head = ringbuf->last_retired_head;
 	ringbuf->last_retired_head = -1;
 
-	ringbuf->space = ring_space(ringbuf);
+	ringbuf->space = intel_ring_space(ringbuf);
 	return 0;
 }
 
@@ -1735,19 +1955,12 @@ static int ring_wait_for_space(struct intel_engine_cs *ring, int n)
 	trace_i915_ring_wait_begin(ring);
 	do {
 		ringbuf->head = I915_READ_HEAD(ring);
-		ringbuf->space = ring_space(ringbuf);
+		ringbuf->space = intel_ring_space(ringbuf);
 		if (ringbuf->space >= n) {
 			ret = 0;
 			break;
 		}
 
-		if (!drm_core_check_feature(dev, DRIVER_MODESET) &&
-		    dev->primary->master) {
-			struct drm_i915_master_private *master_priv = dev->primary->master->driver_priv;
-			if (master_priv->sarea_priv)
-				master_priv->sarea_priv->perf_boxes |= I915_BOX_WAIT;
-		}
-
 		msleep(1);
 
 		if (dev_priv->mm.interruptible && signal_pending(current)) {
@@ -1787,7 +2000,7 @@ static int intel_wrap_ring_buffer(struct intel_engine_cs *ring)
 		iowrite32(MI_NOOP, virt++);
 
 	ringbuf->tail = 0;
-	ringbuf->space = ring_space(ringbuf);
+	ringbuf->space = intel_ring_space(ringbuf);
 
 	return 0;
 }
@@ -1992,9 +2205,7 @@ gen8_ring_dispatch_execbuffer(struct intel_engine_cs *ring,
 			      u64 offset, u32 len,
 			      unsigned flags)
 {
-	struct drm_i915_private *dev_priv = ring->dev->dev_private;
-	bool ppgtt = dev_priv->mm.aliasing_ppgtt != NULL &&
-		!(flags & I915_DISPATCH_SECURE);
+	bool ppgtt = USES_PPGTT(ring->dev) && !(flags & I915_DISPATCH_SECURE);
 	int ret;
 
 	ret = intel_ring_begin(ring, 4);
@@ -2023,8 +2234,9 @@ hsw_ring_dispatch_execbuffer(struct intel_engine_cs *ring,
 		return ret;
 
 	intel_ring_emit(ring,
-			MI_BATCH_BUFFER_START | MI_BATCH_PPGTT_HSW |
-			(flags & I915_DISPATCH_SECURE ? 0 : MI_BATCH_NON_SECURE_HSW));
+			MI_BATCH_BUFFER_START |
+			(flags & I915_DISPATCH_SECURE ?
+			 0 : MI_BATCH_PPGTT_HSW | MI_BATCH_NON_SECURE_HSW));
 	/* bit0-7 is the length on GEN6+ */
 	intel_ring_emit(ring, offset);
 	intel_ring_advance(ring);
@@ -2059,6 +2271,7 @@ static int gen6_ring_flush(struct intel_engine_cs *ring,
 			   u32 invalidate, u32 flush)
 {
 	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint32_t cmd;
 	int ret;
 
@@ -2089,8 +2302,12 @@ static int gen6_ring_flush(struct intel_engine_cs *ring,
 	}
 	intel_ring_advance(ring);
 
-	if (IS_GEN7(dev) && !invalidate && flush)
-		return gen7_ring_fbc_flush(ring, FBC_REND_CACHE_CLEAN);
+	if (!invalidate && flush) {
+		if (IS_GEN7(dev))
+			return gen7_ring_fbc_flush(ring, FBC_REND_CACHE_CLEAN);
+		else if (IS_BROADWELL(dev))
+			dev_priv->fbc.need_sw_cache_clean = true;
+	}
 
 	return 0;
 }
@@ -2123,6 +2340,8 @@ int intel_init_render_ring_buffer(struct drm_device *dev)
 					dev_priv->semaphore_obj = obj;
 			}
 		}
+
+		ring->init_context = intel_ring_workarounds_emit;
 		ring->add_request = gen6_add_request;
 		ring->flush = gen8_render_ring_flush;
 		ring->irq_get = gen8_ring_get_irq;
@@ -2232,91 +2451,6 @@ int intel_init_render_ring_buffer(struct drm_device *dev)
 	return intel_init_ring_buffer(dev, ring);
 }
 
-int intel_render_ring_init_dri(struct drm_device *dev, u64 start, u32 size)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_engine_cs *ring = &dev_priv->ring[RCS];
-	struct intel_ringbuffer *ringbuf = ring->buffer;
-	int ret;
-
-	if (ringbuf == NULL) {
-		ringbuf = kzalloc(sizeof(*ringbuf), GFP_KERNEL);
-		if (!ringbuf)
-			return -ENOMEM;
-		ring->buffer = ringbuf;
-	}
-
-	ring->name = "render ring";
-	ring->id = RCS;
-	ring->mmio_base = RENDER_RING_BASE;
-
-	if (INTEL_INFO(dev)->gen >= 6) {
-		/* non-kms not supported on gen6+ */
-		ret = -ENODEV;
-		goto err_ringbuf;
-	}
-
-	/* Note: gem is not supported on gen5/ilk without kms (the corresponding
-	 * gem_init ioctl returns with -ENODEV). Hence we do not need to set up
-	 * the special gen5 functions. */
-	ring->add_request = i9xx_add_request;
-	if (INTEL_INFO(dev)->gen < 4)
-		ring->flush = gen2_render_ring_flush;
-	else
-		ring->flush = gen4_render_ring_flush;
-	ring->get_seqno = ring_get_seqno;
-	ring->set_seqno = ring_set_seqno;
-	if (IS_GEN2(dev)) {
-		ring->irq_get = i8xx_ring_get_irq;
-		ring->irq_put = i8xx_ring_put_irq;
-	} else {
-		ring->irq_get = i9xx_ring_get_irq;
-		ring->irq_put = i9xx_ring_put_irq;
-	}
-	ring->irq_enable_mask = I915_USER_INTERRUPT;
-	ring->write_tail = ring_write_tail;
-	if (INTEL_INFO(dev)->gen >= 4)
-		ring->dispatch_execbuffer = i965_dispatch_execbuffer;
-	else if (IS_I830(dev) || IS_845G(dev))
-		ring->dispatch_execbuffer = i830_dispatch_execbuffer;
-	else
-		ring->dispatch_execbuffer = i915_dispatch_execbuffer;
-	ring->init = init_render_ring;
-	ring->cleanup = render_ring_cleanup;
-
-	ring->dev = dev;
-	INIT_LIST_HEAD(&ring->active_list);
-	INIT_LIST_HEAD(&ring->request_list);
-
-	ringbuf->size = size;
-	ringbuf->effective_size = ringbuf->size;
-	if (IS_I830(ring->dev) || IS_845G(ring->dev))
-		ringbuf->effective_size -= 2 * CACHELINE_BYTES;
-
-	ringbuf->virtual_start = ioremap_wc(start, size);
-	if (ringbuf->virtual_start == NULL) {
-		DRM_ERROR("can not ioremap virtual address for"
-			  " ring buffer\n");
-		ret = -ENOMEM;
-		goto err_ringbuf;
-	}
-
-	if (!I915_NEED_GFX_HWS(dev)) {
-		ret = init_phys_status_page(ring);
-		if (ret)
-			goto err_vstart;
-	}
-
-	return 0;
-
-err_vstart:
-	iounmap(ringbuf->virtual_start);
-err_ringbuf:
-	kfree(ringbuf);
-	ring->buffer = NULL;
-	return ret;
-}
-
 int intel_init_bsd_ring_buffer(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 70525d0c2c74..fe426cff598b 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -5,6 +5,13 @@
 
 #define I915_CMD_HASH_ORDER 9
 
+/* Early gen2 devices have a cacheline of just 32 bytes, using 64 is overkill,
+ * but keeps the logic simple. Indeed, the whole purpose of this macro is just
+ * to give some inclination as to some of the magic values used in the various
+ * workarounds!
+ */
+#define CACHELINE_BYTES 64
+
 /*
  * Gen2 BSpec "1. Programming Environment" / 1.4.4.6 "Ring Buffer Use"
  * Gen3 BSpec "vol1c Memory Interface Functions" / 2.3.4.5 "Ring Buffer Use"
@@ -90,6 +97,15 @@ struct intel_ringbuffer {
 	struct drm_i915_gem_object *obj;
 	void __iomem *virtual_start;
 
+	struct intel_engine_cs *ring;
+
+	/*
+	 * FIXME: This backpointer is an artifact of the history of how the
+	 * execlist patches came into being. It will get removed once the basic
+	 * code has landed.
+	 */
+	struct intel_context *FIXME_lrc_ctx;
+
 	u32 head;
 	u32 tail;
 	int space;
@@ -132,6 +148,9 @@ struct  intel_engine_cs {
 
 	int		(*init)(struct intel_engine_cs *ring);
 
+	int		(*init_context)(struct intel_engine_cs *ring,
+					struct intel_context *ctx);
+
 	void		(*write_tail)(struct intel_engine_cs *ring,
 				      u32 value);
 	int __must_check (*flush)(struct intel_engine_cs *ring,
@@ -214,6 +233,19 @@ struct  intel_engine_cs {
 				  unsigned int num_dwords);
 	} semaphore;
 
+	/* Execlists */
+	spinlock_t execlist_lock;
+	struct list_head execlist_queue;
+	struct list_head execlist_retired_req_list;
+	u8 next_context_status_buffer;
+	u32             irq_keep_mask; /* bitmask for interrupts that should not be masked */
+	int		(*emit_request)(struct intel_ringbuffer *ringbuf);
+	int		(*emit_flush)(struct intel_ringbuffer *ringbuf,
+				      u32 invalidate_domains,
+				      u32 flush_domains);
+	int		(*emit_bb_start)(struct intel_ringbuffer *ringbuf,
+					 u64 offset, unsigned flags);
+
 	/**
 	 * List of objects currently involved in rendering from the
 	 * ringbuffer.
@@ -287,11 +319,7 @@ struct  intel_engine_cs {
 	u32 (*get_cmd_length_mask)(u32 cmd_header);
 };
 
-static inline bool
-intel_ring_initialized(struct intel_engine_cs *ring)
-{
-	return ring->buffer && ring->buffer->obj;
-}
+bool intel_ring_initialized(struct intel_engine_cs *ring);
 
 static inline unsigned
 intel_ring_flag(struct intel_engine_cs *ring)
@@ -355,6 +383,13 @@ intel_write_status_page(struct intel_engine_cs *ring,
 #define I915_GEM_HWS_SCRATCH_INDEX	0x30
 #define I915_GEM_HWS_SCRATCH_ADDR (I915_GEM_HWS_SCRATCH_INDEX << MI_STORE_DWORD_INDEX_SHIFT)
 
+void intel_unpin_ringbuffer_obj(struct intel_ringbuffer *ringbuf);
+int intel_pin_and_map_ringbuffer_obj(struct drm_device *dev,
+				     struct intel_ringbuffer *ringbuf);
+void intel_destroy_ringbuffer_obj(struct intel_ringbuffer *ringbuf);
+int intel_alloc_ringbuffer_obj(struct drm_device *dev,
+			       struct intel_ringbuffer *ringbuf);
+
 void intel_stop_ring_buffer(struct intel_engine_cs *ring);
 void intel_cleanup_ring_buffer(struct intel_engine_cs *ring);
 
@@ -372,6 +407,9 @@ static inline void intel_ring_advance(struct intel_engine_cs *ring)
 	struct intel_ringbuffer *ringbuf = ring->buffer;
 	ringbuf->tail &= ringbuf->size - 1;
 }
+int __intel_ring_space(int head, int tail, int size);
+int intel_ring_space(struct intel_ringbuffer *ringbuf);
+bool intel_ring_stopped(struct intel_engine_cs *ring);
 void __intel_ring_advance(struct intel_engine_cs *ring);
 
 int __must_check intel_ring_idle(struct intel_engine_cs *ring);
@@ -379,6 +417,9 @@ void intel_ring_init_seqno(struct intel_engine_cs *ring, u32 seqno);
 int intel_ring_flush_all_caches(struct intel_engine_cs *ring);
 int intel_ring_invalidate_all_caches(struct intel_engine_cs *ring);
 
+void intel_fini_pipe_control(struct intel_engine_cs *ring);
+int intel_init_pipe_control(struct intel_engine_cs *ring);
+
 int intel_init_render_ring_buffer(struct drm_device *dev);
 int intel_init_bsd_ring_buffer(struct drm_device *dev);
 int intel_init_bsd2_ring_buffer(struct drm_device *dev);
@@ -388,6 +429,8 @@ int intel_init_vebox_ring_buffer(struct drm_device *dev);
 u64 intel_ring_get_active_head(struct intel_engine_cs *ring);
 void intel_ring_setup_status_page(struct intel_engine_cs *ring);
 
+int init_workarounds_ring(struct intel_engine_cs *ring);
+
 static inline u32 intel_ring_get_tail(struct intel_ringbuffer *ringbuf)
 {
 	return ringbuf->tail;
@@ -405,7 +448,4 @@ static inline void i915_trace_irq_get(struct intel_engine_cs *ring, u32 seqno)
 		ring->trace_irq_seqno = seqno;
 }
 
-/* DRI warts */
-int intel_render_ring_init_dri(struct drm_device *dev, u64 start, u32 size);
-
 #endif /* _INTEL_RINGBUFFER_H_ */
diff --git a/drivers/gpu/drm/i915/intel_runtime_pm.c b/drivers/gpu/drm/i915/intel_runtime_pm.c
new file mode 100644
index 000000000000..ac6da7102fbb
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_runtime_pm.c
@@ -0,0 +1,1379 @@
+/*
+ * Copyright © 2012-2014 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ * Authors:
+ *    Eugeni Dodonov <eugeni.dodonov@intel.com>
+ *    Daniel Vetter <daniel.vetter@ffwll.ch>
+ *
+ */
+
+#include <linux/pm_runtime.h>
+#include <linux/vgaarb.h>
+
+#include "i915_drv.h"
+#include "intel_drv.h"
+#include <drm/i915_powerwell.h>
+
+/**
+ * DOC: runtime pm
+ *
+ * The i915 driver supports dynamic enabling and disabling of entire hardware
+ * blocks at runtime. This is especially important on the display side where
+ * software is supposed to control many power gates manually on recent hardware,
+ * since on the GT side a lot of the power management is done by the hardware.
+ * But even there some manual control at the device level is required.
+ *
+ * Since i915 supports a diverse set of platforms with a unified codebase and
+ * hardware engineers just love to shuffle functionality around between power
+ * domains there's a sizeable amount of indirection required. This file provides
+ * generic functions to the driver for grabbing and releasing references for
+ * abstract power domains. It then maps those to the actual power wells
+ * present for a given platform.
+ */
+
+static struct i915_power_domains *hsw_pwr;
+
+#define for_each_power_well(i, power_well, domain_mask, power_domains)	\
+	for (i = 0;							\
+	     i < (power_domains)->power_well_count &&			\
+		 ((power_well) = &(power_domains)->power_wells[i]);	\
+	     i++)							\
+		if ((power_well)->domains & (domain_mask))
+
+#define for_each_power_well_rev(i, power_well, domain_mask, power_domains) \
+	for (i = (power_domains)->power_well_count - 1;			 \
+	     i >= 0 && ((power_well) = &(power_domains)->power_wells[i]);\
+	     i--)							 \
+		if ((power_well)->domains & (domain_mask))
+
+/*
+ * We should only use the power well if we explicitly asked the hardware to
+ * enable it, so check if it's enabled and also check if we've requested it to
+ * be enabled.
+ */
+static bool hsw_power_well_enabled(struct drm_i915_private *dev_priv,
+				   struct i915_power_well *power_well)
+{
+	return I915_READ(HSW_PWR_WELL_DRIVER) ==
+		     (HSW_PWR_WELL_ENABLE_REQUEST | HSW_PWR_WELL_STATE_ENABLED);
+}
+
+/**
+ * __intel_display_power_is_enabled - unlocked check for a power domain
+ * @dev_priv: i915 device instance
+ * @domain: power domain to check
+ *
+ * This is the unlocked version of intel_display_power_is_enabled() and should
+ * only be used from error capture and recovery code where deadlocks are
+ * possible.
+ *
+ * Returns:
+ * True when the power domain is enabled, false otherwise.
+ */
+bool __intel_display_power_is_enabled(struct drm_i915_private *dev_priv,
+				      enum intel_display_power_domain domain)
+{
+	struct i915_power_domains *power_domains;
+	struct i915_power_well *power_well;
+	bool is_enabled;
+	int i;
+
+	if (dev_priv->pm.suspended)
+		return false;
+
+	power_domains = &dev_priv->power_domains;
+
+	is_enabled = true;
+
+	for_each_power_well_rev(i, power_well, BIT(domain), power_domains) {
+		if (power_well->always_on)
+			continue;
+
+		if (!power_well->hw_enabled) {
+			is_enabled = false;
+			break;
+		}
+	}
+
+	return is_enabled;
+}
+
+/**
+ * intel_display_power_is_enabled - unlocked check for a power domain
+ * @dev_priv: i915 device instance
+ * @domain: power domain to check
+ *
+ * This function can be used to check the hw power domain state. It is mostly
+ * used in hardware state readout functions. Everywhere else code should rely
+ * upon explicit power domain reference counting to ensure that the hardware
+ * block is powered up before accessing it.
+ *
+ * Callers must hold the relevant modesetting locks to ensure that concurrent
+ * threads can't disable the power well while the caller tries to read a few
+ * registers.
+ *
+ * Returns:
+ * True when the power domain is enabled, false otherwise.
+ */
+bool intel_display_power_is_enabled(struct drm_i915_private *dev_priv,
+				    enum intel_display_power_domain domain)
+{
+	struct i915_power_domains *power_domains;
+	bool ret;
+
+	power_domains = &dev_priv->power_domains;
+
+	mutex_lock(&power_domains->lock);
+	ret = __intel_display_power_is_enabled(dev_priv, domain);
+	mutex_unlock(&power_domains->lock);
+
+	return ret;
+}
+
+/**
+ * intel_display_set_init_power - set the initial power domain state
+ * @dev_priv: i915 device instance
+ * @enable: whether to enable or disable the initial power domain state
+ *
+ * For simplicity our driver load/unload and system suspend/resume code assumes
+ * that all power domains are always enabled. This functions controls the state
+ * of this little hack. While the initial power domain state is enabled runtime
+ * pm is effectively disabled.
+ */
+void intel_display_set_init_power(struct drm_i915_private *dev_priv,
+				  bool enable)
+{
+	if (dev_priv->power_domains.init_power_on == enable)
+		return;
+
+	if (enable)
+		intel_display_power_get(dev_priv, POWER_DOMAIN_INIT);
+	else
+		intel_display_power_put(dev_priv, POWER_DOMAIN_INIT);
+
+	dev_priv->power_domains.init_power_on = enable;
+}
+
+/*
+ * Starting with Haswell, we have a "Power Down Well" that can be turned off
+ * when not needed anymore. We have 4 registers that can request the power well
+ * to be enabled, and it will only be disabled if none of the registers is
+ * requesting it to be enabled.
+ */
+static void hsw_power_well_post_enable(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+
+	/*
+	 * After we re-enable the power well, if we touch VGA register 0x3d5
+	 * we'll get unclaimed register interrupts. This stops after we write
+	 * anything to the VGA MSR register. The vgacon module uses this
+	 * register all the time, so if we unbind our driver and, as a
+	 * consequence, bind vgacon, we'll get stuck in an infinite loop at
+	 * console_unlock(). So make here we touch the VGA MSR register, making
+	 * sure vgacon can keep working normally without triggering interrupts
+	 * and error messages.
+	 */
+	vga_get_uninterruptible(dev->pdev, VGA_RSRC_LEGACY_IO);
+	outb(inb(VGA_MSR_READ), VGA_MSR_WRITE);
+	vga_put(dev->pdev, VGA_RSRC_LEGACY_IO);
+
+	if (IS_BROADWELL(dev) || (INTEL_INFO(dev)->gen >= 9))
+		gen8_irq_power_well_post_enable(dev_priv);
+}
+
+static void hsw_set_power_well(struct drm_i915_private *dev_priv,
+			       struct i915_power_well *power_well, bool enable)
+{
+	bool is_enabled, enable_requested;
+	uint32_t tmp;
+
+	tmp = I915_READ(HSW_PWR_WELL_DRIVER);
+	is_enabled = tmp & HSW_PWR_WELL_STATE_ENABLED;
+	enable_requested = tmp & HSW_PWR_WELL_ENABLE_REQUEST;
+
+	if (enable) {
+		if (!enable_requested)
+			I915_WRITE(HSW_PWR_WELL_DRIVER,
+				   HSW_PWR_WELL_ENABLE_REQUEST);
+
+		if (!is_enabled) {
+			DRM_DEBUG_KMS("Enabling power well\n");
+			if (wait_for((I915_READ(HSW_PWR_WELL_DRIVER) &
+				      HSW_PWR_WELL_STATE_ENABLED), 20))
+				DRM_ERROR("Timeout enabling power well\n");
+			hsw_power_well_post_enable(dev_priv);
+		}
+
+	} else {
+		if (enable_requested) {
+			I915_WRITE(HSW_PWR_WELL_DRIVER, 0);
+			POSTING_READ(HSW_PWR_WELL_DRIVER);
+			DRM_DEBUG_KMS("Requesting to disable the power well\n");
+		}
+	}
+}
+
+static void hsw_power_well_sync_hw(struct drm_i915_private *dev_priv,
+				   struct i915_power_well *power_well)
+{
+	hsw_set_power_well(dev_priv, power_well, power_well->count > 0);
+
+	/*
+	 * We're taking over the BIOS, so clear any requests made by it since
+	 * the driver is in charge now.
+	 */
+	if (I915_READ(HSW_PWR_WELL_BIOS) & HSW_PWR_WELL_ENABLE_REQUEST)
+		I915_WRITE(HSW_PWR_WELL_BIOS, 0);
+}
+
+static void hsw_power_well_enable(struct drm_i915_private *dev_priv,
+				  struct i915_power_well *power_well)
+{
+	hsw_set_power_well(dev_priv, power_well, true);
+}
+
+static void hsw_power_well_disable(struct drm_i915_private *dev_priv,
+				   struct i915_power_well *power_well)
+{
+	hsw_set_power_well(dev_priv, power_well, false);
+}
+
+static void i9xx_always_on_power_well_noop(struct drm_i915_private *dev_priv,
+					   struct i915_power_well *power_well)
+{
+}
+
+static bool i9xx_always_on_power_well_enabled(struct drm_i915_private *dev_priv,
+					     struct i915_power_well *power_well)
+{
+	return true;
+}
+
+static void vlv_set_power_well(struct drm_i915_private *dev_priv,
+			       struct i915_power_well *power_well, bool enable)
+{
+	enum punit_power_well power_well_id = power_well->data;
+	u32 mask;
+	u32 state;
+	u32 ctrl;
+
+	mask = PUNIT_PWRGT_MASK(power_well_id);
+	state = enable ? PUNIT_PWRGT_PWR_ON(power_well_id) :
+			 PUNIT_PWRGT_PWR_GATE(power_well_id);
+
+	mutex_lock(&dev_priv->rps.hw_lock);
+
+#define COND \
+	((vlv_punit_read(dev_priv, PUNIT_REG_PWRGT_STATUS) & mask) == state)
+
+	if (COND)
+		goto out;
+
+	ctrl = vlv_punit_read(dev_priv, PUNIT_REG_PWRGT_CTRL);
+	ctrl &= ~mask;
+	ctrl |= state;
+	vlv_punit_write(dev_priv, PUNIT_REG_PWRGT_CTRL, ctrl);
+
+	if (wait_for(COND, 100))
+		DRM_ERROR("timout setting power well state %08x (%08x)\n",
+			  state,
+			  vlv_punit_read(dev_priv, PUNIT_REG_PWRGT_CTRL));
+
+#undef COND
+
+out:
+	mutex_unlock(&dev_priv->rps.hw_lock);
+}
+
+static void vlv_power_well_sync_hw(struct drm_i915_private *dev_priv,
+				   struct i915_power_well *power_well)
+{
+	vlv_set_power_well(dev_priv, power_well, power_well->count > 0);
+}
+
+static void vlv_power_well_enable(struct drm_i915_private *dev_priv,
+				  struct i915_power_well *power_well)
+{
+	vlv_set_power_well(dev_priv, power_well, true);
+}
+
+static void vlv_power_well_disable(struct drm_i915_private *dev_priv,
+				   struct i915_power_well *power_well)
+{
+	vlv_set_power_well(dev_priv, power_well, false);
+}
+
+static bool vlv_power_well_enabled(struct drm_i915_private *dev_priv,
+				   struct i915_power_well *power_well)
+{
+	int power_well_id = power_well->data;
+	bool enabled = false;
+	u32 mask;
+	u32 state;
+	u32 ctrl;
+
+	mask = PUNIT_PWRGT_MASK(power_well_id);
+	ctrl = PUNIT_PWRGT_PWR_ON(power_well_id);
+
+	mutex_lock(&dev_priv->rps.hw_lock);
+
+	state = vlv_punit_read(dev_priv, PUNIT_REG_PWRGT_STATUS) & mask;
+	/*
+	 * We only ever set the power-on and power-gate states, anything
+	 * else is unexpected.
+	 */
+	WARN_ON(state != PUNIT_PWRGT_PWR_ON(power_well_id) &&
+		state != PUNIT_PWRGT_PWR_GATE(power_well_id));
+	if (state == ctrl)
+		enabled = true;
+
+	/*
+	 * A transient state at this point would mean some unexpected party
+	 * is poking at the power controls too.
+	 */
+	ctrl = vlv_punit_read(dev_priv, PUNIT_REG_PWRGT_CTRL) & mask;
+	WARN_ON(ctrl != state);
+
+	mutex_unlock(&dev_priv->rps.hw_lock);
+
+	return enabled;
+}
+
+static void vlv_display_power_well_enable(struct drm_i915_private *dev_priv,
+					  struct i915_power_well *power_well)
+{
+	WARN_ON_ONCE(power_well->data != PUNIT_POWER_WELL_DISP2D);
+
+	vlv_set_power_well(dev_priv, power_well, true);
+
+	spin_lock_irq(&dev_priv->irq_lock);
+	valleyview_enable_display_irqs(dev_priv);
+	spin_unlock_irq(&dev_priv->irq_lock);
+
+	/*
+	 * During driver initialization/resume we can avoid restoring the
+	 * part of the HW/SW state that will be inited anyway explicitly.
+	 */
+	if (dev_priv->power_domains.initializing)
+		return;
+
+	intel_hpd_init(dev_priv);
+
+	i915_redisable_vga_power_on(dev_priv->dev);
+}
+
+static void vlv_display_power_well_disable(struct drm_i915_private *dev_priv,
+					   struct i915_power_well *power_well)
+{
+	WARN_ON_ONCE(power_well->data != PUNIT_POWER_WELL_DISP2D);
+
+	spin_lock_irq(&dev_priv->irq_lock);
+	valleyview_disable_display_irqs(dev_priv);
+	spin_unlock_irq(&dev_priv->irq_lock);
+
+	vlv_set_power_well(dev_priv, power_well, false);
+
+	vlv_power_sequencer_reset(dev_priv);
+}
+
+static void vlv_dpio_cmn_power_well_enable(struct drm_i915_private *dev_priv,
+					   struct i915_power_well *power_well)
+{
+	WARN_ON_ONCE(power_well->data != PUNIT_POWER_WELL_DPIO_CMN_BC);
+
+	/*
+	 * Enable the CRI clock source so we can get at the
+	 * display and the reference clock for VGA
+	 * hotplug / manual detection.
+	 */
+	I915_WRITE(DPLL(PIPE_B), I915_READ(DPLL(PIPE_B)) |
+		   DPLL_REFA_CLK_ENABLE_VLV | DPLL_INTEGRATED_CRI_CLK_VLV);
+	udelay(1); /* >10ns for cmnreset, >0ns for sidereset */
+
+	vlv_set_power_well(dev_priv, power_well, true);
+
+	/*
+	 * From VLV2A0_DP_eDP_DPIO_driver_vbios_notes_10.docx -
+	 *  6.	De-assert cmn_reset/side_reset. Same as VLV X0.
+	 *   a.	GUnit 0x2110 bit[0] set to 1 (def 0)
+	 *   b.	The other bits such as sfr settings / modesel may all
+	 *	be set to 0.
+	 *
+	 * This should only be done on init and resume from S3 with
+	 * both PLLs disabled, or we risk losing DPIO and PLL
+	 * synchronization.
+	 */
+	I915_WRITE(DPIO_CTL, I915_READ(DPIO_CTL) | DPIO_CMNRST);
+}
+
+static void vlv_dpio_cmn_power_well_disable(struct drm_i915_private *dev_priv,
+					    struct i915_power_well *power_well)
+{
+	enum pipe pipe;
+
+	WARN_ON_ONCE(power_well->data != PUNIT_POWER_WELL_DPIO_CMN_BC);
+
+	for_each_pipe(dev_priv, pipe)
+		assert_pll_disabled(dev_priv, pipe);
+
+	/* Assert common reset */
+	I915_WRITE(DPIO_CTL, I915_READ(DPIO_CTL) & ~DPIO_CMNRST);
+
+	vlv_set_power_well(dev_priv, power_well, false);
+}
+
+static void chv_dpio_cmn_power_well_enable(struct drm_i915_private *dev_priv,
+					   struct i915_power_well *power_well)
+{
+	enum dpio_phy phy;
+
+	WARN_ON_ONCE(power_well->data != PUNIT_POWER_WELL_DPIO_CMN_BC &&
+		     power_well->data != PUNIT_POWER_WELL_DPIO_CMN_D);
+
+	/*
+	 * Enable the CRI clock source so we can get at the
+	 * display and the reference clock for VGA
+	 * hotplug / manual detection.
+	 */
+	if (power_well->data == PUNIT_POWER_WELL_DPIO_CMN_BC) {
+		phy = DPIO_PHY0;
+		I915_WRITE(DPLL(PIPE_B), I915_READ(DPLL(PIPE_B)) |
+			   DPLL_REFA_CLK_ENABLE_VLV);
+		I915_WRITE(DPLL(PIPE_B), I915_READ(DPLL(PIPE_B)) |
+			   DPLL_REFA_CLK_ENABLE_VLV | DPLL_INTEGRATED_CRI_CLK_VLV);
+	} else {
+		phy = DPIO_PHY1;
+		I915_WRITE(DPLL(PIPE_C), I915_READ(DPLL(PIPE_C)) |
+			   DPLL_REFA_CLK_ENABLE_VLV | DPLL_INTEGRATED_CRI_CLK_VLV);
+	}
+	udelay(1); /* >10ns for cmnreset, >0ns for sidereset */
+	vlv_set_power_well(dev_priv, power_well, true);
+
+	/* Poll for phypwrgood signal */
+	if (wait_for(I915_READ(DISPLAY_PHY_STATUS) & PHY_POWERGOOD(phy), 1))
+		DRM_ERROR("Display PHY %d is not power up\n", phy);
+
+	I915_WRITE(DISPLAY_PHY_CONTROL, I915_READ(DISPLAY_PHY_CONTROL) |
+		   PHY_COM_LANE_RESET_DEASSERT(phy));
+}
+
+static void chv_dpio_cmn_power_well_disable(struct drm_i915_private *dev_priv,
+					    struct i915_power_well *power_well)
+{
+	enum dpio_phy phy;
+
+	WARN_ON_ONCE(power_well->data != PUNIT_POWER_WELL_DPIO_CMN_BC &&
+		     power_well->data != PUNIT_POWER_WELL_DPIO_CMN_D);
+
+	if (power_well->data == PUNIT_POWER_WELL_DPIO_CMN_BC) {
+		phy = DPIO_PHY0;
+		assert_pll_disabled(dev_priv, PIPE_A);
+		assert_pll_disabled(dev_priv, PIPE_B);
+	} else {
+		phy = DPIO_PHY1;
+		assert_pll_disabled(dev_priv, PIPE_C);
+	}
+
+	I915_WRITE(DISPLAY_PHY_CONTROL, I915_READ(DISPLAY_PHY_CONTROL) &
+		   ~PHY_COM_LANE_RESET_DEASSERT(phy));
+
+	vlv_set_power_well(dev_priv, power_well, false);
+}
+
+static bool chv_pipe_power_well_enabled(struct drm_i915_private *dev_priv,
+					struct i915_power_well *power_well)
+{
+	enum pipe pipe = power_well->data;
+	bool enabled;
+	u32 state, ctrl;
+
+	mutex_lock(&dev_priv->rps.hw_lock);
+
+	state = vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ) & DP_SSS_MASK(pipe);
+	/*
+	 * We only ever set the power-on and power-gate states, anything
+	 * else is unexpected.
+	 */
+	WARN_ON(state != DP_SSS_PWR_ON(pipe) && state != DP_SSS_PWR_GATE(pipe));
+	enabled = state == DP_SSS_PWR_ON(pipe);
+
+	/*
+	 * A transient state at this point would mean some unexpected party
+	 * is poking at the power controls too.
+	 */
+	ctrl = vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ) & DP_SSC_MASK(pipe);
+	WARN_ON(ctrl << 16 != state);
+
+	mutex_unlock(&dev_priv->rps.hw_lock);
+
+	return enabled;
+}
+
+static void chv_set_pipe_power_well(struct drm_i915_private *dev_priv,
+				    struct i915_power_well *power_well,
+				    bool enable)
+{
+	enum pipe pipe = power_well->data;
+	u32 state;
+	u32 ctrl;
+
+	state = enable ? DP_SSS_PWR_ON(pipe) : DP_SSS_PWR_GATE(pipe);
+
+	mutex_lock(&dev_priv->rps.hw_lock);
+
+#define COND \
+	((vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ) & DP_SSS_MASK(pipe)) == state)
+
+	if (COND)
+		goto out;
+
+	ctrl = vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ);
+	ctrl &= ~DP_SSC_MASK(pipe);
+	ctrl |= enable ? DP_SSC_PWR_ON(pipe) : DP_SSC_PWR_GATE(pipe);
+	vlv_punit_write(dev_priv, PUNIT_REG_DSPFREQ, ctrl);
+
+	if (wait_for(COND, 100))
+		DRM_ERROR("timout setting power well state %08x (%08x)\n",
+			  state,
+			  vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ));
+
+#undef COND
+
+out:
+	mutex_unlock(&dev_priv->rps.hw_lock);
+}
+
+static void chv_pipe_power_well_sync_hw(struct drm_i915_private *dev_priv,
+					struct i915_power_well *power_well)
+{
+	chv_set_pipe_power_well(dev_priv, power_well, power_well->count > 0);
+}
+
+static void chv_pipe_power_well_enable(struct drm_i915_private *dev_priv,
+				       struct i915_power_well *power_well)
+{
+	WARN_ON_ONCE(power_well->data != PIPE_A &&
+		     power_well->data != PIPE_B &&
+		     power_well->data != PIPE_C);
+
+	chv_set_pipe_power_well(dev_priv, power_well, true);
+
+	if (power_well->data == PIPE_A) {
+		spin_lock_irq(&dev_priv->irq_lock);
+		valleyview_enable_display_irqs(dev_priv);
+		spin_unlock_irq(&dev_priv->irq_lock);
+
+		/*
+		 * During driver initialization/resume we can avoid restoring the
+		 * part of the HW/SW state that will be inited anyway explicitly.
+		 */
+		if (dev_priv->power_domains.initializing)
+			return;
+
+		intel_hpd_init(dev_priv);
+
+		i915_redisable_vga_power_on(dev_priv->dev);
+	}
+}
+
+static void chv_pipe_power_well_disable(struct drm_i915_private *dev_priv,
+					struct i915_power_well *power_well)
+{
+	WARN_ON_ONCE(power_well->data != PIPE_A &&
+		     power_well->data != PIPE_B &&
+		     power_well->data != PIPE_C);
+
+	if (power_well->data == PIPE_A) {
+		spin_lock_irq(&dev_priv->irq_lock);
+		valleyview_disable_display_irqs(dev_priv);
+		spin_unlock_irq(&dev_priv->irq_lock);
+	}
+
+	chv_set_pipe_power_well(dev_priv, power_well, false);
+
+	if (power_well->data == PIPE_A)
+		vlv_power_sequencer_reset(dev_priv);
+}
+
+/**
+ * intel_display_power_get - grab a power domain reference
+ * @dev_priv: i915 device instance
+ * @domain: power domain to reference
+ *
+ * This function grabs a power domain reference for @domain and ensures that the
+ * power domain and all its parents are powered up. Therefore users should only
+ * grab a reference to the innermost power domain they need.
+ *
+ * Any power domain reference obtained by this function must have a symmetric
+ * call to intel_display_power_put() to release the reference again.
+ */
+void intel_display_power_get(struct drm_i915_private *dev_priv,
+			     enum intel_display_power_domain domain)
+{
+	struct i915_power_domains *power_domains;
+	struct i915_power_well *power_well;
+	int i;
+
+	intel_runtime_pm_get(dev_priv);
+
+	power_domains = &dev_priv->power_domains;
+
+	mutex_lock(&power_domains->lock);
+
+	for_each_power_well(i, power_well, BIT(domain), power_domains) {
+		if (!power_well->count++) {
+			DRM_DEBUG_KMS("enabling %s\n", power_well->name);
+			power_well->ops->enable(dev_priv, power_well);
+			power_well->hw_enabled = true;
+		}
+	}
+
+	power_domains->domain_use_count[domain]++;
+
+	mutex_unlock(&power_domains->lock);
+}
+
+/**
+ * intel_display_power_put - release a power domain reference
+ * @dev_priv: i915 device instance
+ * @domain: power domain to reference
+ *
+ * This function drops the power domain reference obtained by
+ * intel_display_power_get() and might power down the corresponding hardware
+ * block right away if this is the last reference.
+ */
+void intel_display_power_put(struct drm_i915_private *dev_priv,
+			     enum intel_display_power_domain domain)
+{
+	struct i915_power_domains *power_domains;
+	struct i915_power_well *power_well;
+	int i;
+
+	power_domains = &dev_priv->power_domains;
+
+	mutex_lock(&power_domains->lock);
+
+	WARN_ON(!power_domains->domain_use_count[domain]);
+	power_domains->domain_use_count[domain]--;
+
+	for_each_power_well_rev(i, power_well, BIT(domain), power_domains) {
+		WARN_ON(!power_well->count);
+
+		if (!--power_well->count && i915.disable_power_well) {
+			DRM_DEBUG_KMS("disabling %s\n", power_well->name);
+			power_well->hw_enabled = false;
+			power_well->ops->disable(dev_priv, power_well);
+		}
+	}
+
+	mutex_unlock(&power_domains->lock);
+
+	intel_runtime_pm_put(dev_priv);
+}
+
+#define POWER_DOMAIN_MASK (BIT(POWER_DOMAIN_NUM) - 1)
+
+#define HSW_ALWAYS_ON_POWER_DOMAINS (			\
+	BIT(POWER_DOMAIN_PIPE_A) |			\
+	BIT(POWER_DOMAIN_TRANSCODER_EDP) |		\
+	BIT(POWER_DOMAIN_PORT_DDI_A_2_LANES) |		\
+	BIT(POWER_DOMAIN_PORT_DDI_A_4_LANES) |		\
+	BIT(POWER_DOMAIN_PORT_DDI_B_2_LANES) |		\
+	BIT(POWER_DOMAIN_PORT_DDI_B_4_LANES) |		\
+	BIT(POWER_DOMAIN_PORT_DDI_C_2_LANES) |		\
+	BIT(POWER_DOMAIN_PORT_DDI_C_4_LANES) |		\
+	BIT(POWER_DOMAIN_PORT_DDI_D_2_LANES) |		\
+	BIT(POWER_DOMAIN_PORT_DDI_D_4_LANES) |		\
+	BIT(POWER_DOMAIN_PORT_CRT) |			\
+	BIT(POWER_DOMAIN_PLLS) |			\
+	BIT(POWER_DOMAIN_INIT))
+#define HSW_DISPLAY_POWER_DOMAINS (				\
+	(POWER_DOMAIN_MASK & ~HSW_ALWAYS_ON_POWER_DOMAINS) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define BDW_ALWAYS_ON_POWER_DOMAINS (			\
+	HSW_ALWAYS_ON_POWER_DOMAINS |			\
+	BIT(POWER_DOMAIN_PIPE_A_PANEL_FITTER))
+#define BDW_DISPLAY_POWER_DOMAINS (				\
+	(POWER_DOMAIN_MASK & ~BDW_ALWAYS_ON_POWER_DOMAINS) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define VLV_ALWAYS_ON_POWER_DOMAINS	BIT(POWER_DOMAIN_INIT)
+#define VLV_DISPLAY_POWER_DOMAINS	POWER_DOMAIN_MASK
+
+#define VLV_DPIO_CMN_BC_POWER_DOMAINS (		\
+	BIT(POWER_DOMAIN_PORT_DDI_B_2_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_B_4_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_C_2_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_C_4_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_CRT) |		\
+	BIT(POWER_DOMAIN_INIT))
+
+#define VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PORT_DDI_B_2_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_B_4_LANES) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PORT_DDI_B_4_LANES) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PORT_DDI_C_2_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_C_4_LANES) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PORT_DDI_C_4_LANES) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define CHV_PIPE_A_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PIPE_A) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define CHV_PIPE_B_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PIPE_B) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define CHV_PIPE_C_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PIPE_C) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define CHV_DPIO_CMN_BC_POWER_DOMAINS (		\
+	BIT(POWER_DOMAIN_PORT_DDI_B_2_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_B_4_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_C_2_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_C_4_LANES) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define CHV_DPIO_CMN_D_POWER_DOMAINS (		\
+	BIT(POWER_DOMAIN_PORT_DDI_D_2_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_D_4_LANES) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define CHV_DPIO_TX_D_LANES_01_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PORT_DDI_D_2_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_D_4_LANES) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define CHV_DPIO_TX_D_LANES_23_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PORT_DDI_D_4_LANES) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+static const struct i915_power_well_ops i9xx_always_on_power_well_ops = {
+	.sync_hw = i9xx_always_on_power_well_noop,
+	.enable = i9xx_always_on_power_well_noop,
+	.disable = i9xx_always_on_power_well_noop,
+	.is_enabled = i9xx_always_on_power_well_enabled,
+};
+
+static const struct i915_power_well_ops chv_pipe_power_well_ops = {
+	.sync_hw = chv_pipe_power_well_sync_hw,
+	.enable = chv_pipe_power_well_enable,
+	.disable = chv_pipe_power_well_disable,
+	.is_enabled = chv_pipe_power_well_enabled,
+};
+
+static const struct i915_power_well_ops chv_dpio_cmn_power_well_ops = {
+	.sync_hw = vlv_power_well_sync_hw,
+	.enable = chv_dpio_cmn_power_well_enable,
+	.disable = chv_dpio_cmn_power_well_disable,
+	.is_enabled = vlv_power_well_enabled,
+};
+
+static struct i915_power_well i9xx_always_on_power_well[] = {
+	{
+		.name = "always-on",
+		.always_on = 1,
+		.domains = POWER_DOMAIN_MASK,
+		.ops = &i9xx_always_on_power_well_ops,
+	},
+};
+
+static const struct i915_power_well_ops hsw_power_well_ops = {
+	.sync_hw = hsw_power_well_sync_hw,
+	.enable = hsw_power_well_enable,
+	.disable = hsw_power_well_disable,
+	.is_enabled = hsw_power_well_enabled,
+};
+
+static struct i915_power_well hsw_power_wells[] = {
+	{
+		.name = "always-on",
+		.always_on = 1,
+		.domains = HSW_ALWAYS_ON_POWER_DOMAINS,
+		.ops = &i9xx_always_on_power_well_ops,
+	},
+	{
+		.name = "display",
+		.domains = HSW_DISPLAY_POWER_DOMAINS,
+		.ops = &hsw_power_well_ops,
+	},
+};
+
+static struct i915_power_well bdw_power_wells[] = {
+	{
+		.name = "always-on",
+		.always_on = 1,
+		.domains = BDW_ALWAYS_ON_POWER_DOMAINS,
+		.ops = &i9xx_always_on_power_well_ops,
+	},
+	{
+		.name = "display",
+		.domains = BDW_DISPLAY_POWER_DOMAINS,
+		.ops = &hsw_power_well_ops,
+	},
+};
+
+static const struct i915_power_well_ops vlv_display_power_well_ops = {
+	.sync_hw = vlv_power_well_sync_hw,
+	.enable = vlv_display_power_well_enable,
+	.disable = vlv_display_power_well_disable,
+	.is_enabled = vlv_power_well_enabled,
+};
+
+static const struct i915_power_well_ops vlv_dpio_cmn_power_well_ops = {
+	.sync_hw = vlv_power_well_sync_hw,
+	.enable = vlv_dpio_cmn_power_well_enable,
+	.disable = vlv_dpio_cmn_power_well_disable,
+	.is_enabled = vlv_power_well_enabled,
+};
+
+static const struct i915_power_well_ops vlv_dpio_power_well_ops = {
+	.sync_hw = vlv_power_well_sync_hw,
+	.enable = vlv_power_well_enable,
+	.disable = vlv_power_well_disable,
+	.is_enabled = vlv_power_well_enabled,
+};
+
+static struct i915_power_well vlv_power_wells[] = {
+	{
+		.name = "always-on",
+		.always_on = 1,
+		.domains = VLV_ALWAYS_ON_POWER_DOMAINS,
+		.ops = &i9xx_always_on_power_well_ops,
+	},
+	{
+		.name = "display",
+		.domains = VLV_DISPLAY_POWER_DOMAINS,
+		.data = PUNIT_POWER_WELL_DISP2D,
+		.ops = &vlv_display_power_well_ops,
+	},
+	{
+		.name = "dpio-tx-b-01",
+		.domains = VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_B_LANES_01,
+	},
+	{
+		.name = "dpio-tx-b-23",
+		.domains = VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_B_LANES_23,
+	},
+	{
+		.name = "dpio-tx-c-01",
+		.domains = VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_C_LANES_01,
+	},
+	{
+		.name = "dpio-tx-c-23",
+		.domains = VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_C_LANES_23,
+	},
+	{
+		.name = "dpio-common",
+		.domains = VLV_DPIO_CMN_BC_POWER_DOMAINS,
+		.data = PUNIT_POWER_WELL_DPIO_CMN_BC,
+		.ops = &vlv_dpio_cmn_power_well_ops,
+	},
+};
+
+static struct i915_power_well chv_power_wells[] = {
+	{
+		.name = "always-on",
+		.always_on = 1,
+		.domains = VLV_ALWAYS_ON_POWER_DOMAINS,
+		.ops = &i9xx_always_on_power_well_ops,
+	},
+#if 0
+	{
+		.name = "display",
+		.domains = VLV_DISPLAY_POWER_DOMAINS,
+		.data = PUNIT_POWER_WELL_DISP2D,
+		.ops = &vlv_display_power_well_ops,
+	},
+#endif
+	{
+		.name = "pipe-a",
+		/*
+		 * FIXME: pipe A power well seems to be the new disp2d well.
+		 * At least all registers seem to be housed there. Figure
+		 * out if this a a temporary situation in pre-production
+		 * hardware or a permanent state of affairs.
+		 */
+		.domains = CHV_PIPE_A_POWER_DOMAINS | VLV_DISPLAY_POWER_DOMAINS,
+		.data = PIPE_A,
+		.ops = &chv_pipe_power_well_ops,
+	},
+#if 0
+	{
+		.name = "pipe-b",
+		.domains = CHV_PIPE_B_POWER_DOMAINS,
+		.data = PIPE_B,
+		.ops = &chv_pipe_power_well_ops,
+	},
+	{
+		.name = "pipe-c",
+		.domains = CHV_PIPE_C_POWER_DOMAINS,
+		.data = PIPE_C,
+		.ops = &chv_pipe_power_well_ops,
+	},
+#endif
+	{
+		.name = "dpio-common-bc",
+		/*
+		 * XXX: cmnreset for one PHY seems to disturb the other.
+		 * As a workaround keep both powered on at the same
+		 * time for now.
+		 */
+		.domains = CHV_DPIO_CMN_BC_POWER_DOMAINS | CHV_DPIO_CMN_D_POWER_DOMAINS,
+		.data = PUNIT_POWER_WELL_DPIO_CMN_BC,
+		.ops = &chv_dpio_cmn_power_well_ops,
+	},
+	{
+		.name = "dpio-common-d",
+		/*
+		 * XXX: cmnreset for one PHY seems to disturb the other.
+		 * As a workaround keep both powered on at the same
+		 * time for now.
+		 */
+		.domains = CHV_DPIO_CMN_BC_POWER_DOMAINS | CHV_DPIO_CMN_D_POWER_DOMAINS,
+		.data = PUNIT_POWER_WELL_DPIO_CMN_D,
+		.ops = &chv_dpio_cmn_power_well_ops,
+	},
+#if 0
+	{
+		.name = "dpio-tx-b-01",
+		.domains = VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_B_LANES_01,
+	},
+	{
+		.name = "dpio-tx-b-23",
+		.domains = VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_B_LANES_23,
+	},
+	{
+		.name = "dpio-tx-c-01",
+		.domains = VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_C_LANES_01,
+	},
+	{
+		.name = "dpio-tx-c-23",
+		.domains = VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_C_LANES_23,
+	},
+	{
+		.name = "dpio-tx-d-01",
+		.domains = CHV_DPIO_TX_D_LANES_01_POWER_DOMAINS |
+			   CHV_DPIO_TX_D_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_D_LANES_01,
+	},
+	{
+		.name = "dpio-tx-d-23",
+		.domains = CHV_DPIO_TX_D_LANES_01_POWER_DOMAINS |
+			   CHV_DPIO_TX_D_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_D_LANES_23,
+	},
+#endif
+};
+
+static struct i915_power_well *lookup_power_well(struct drm_i915_private *dev_priv,
+						 enum punit_power_well power_well_id)
+{
+	struct i915_power_domains *power_domains = &dev_priv->power_domains;
+	struct i915_power_well *power_well;
+	int i;
+
+	for_each_power_well(i, power_well, POWER_DOMAIN_MASK, power_domains) {
+		if (power_well->data == power_well_id)
+			return power_well;
+	}
+
+	return NULL;
+}
+
+#define set_power_wells(power_domains, __power_wells) ({		\
+	(power_domains)->power_wells = (__power_wells);			\
+	(power_domains)->power_well_count = ARRAY_SIZE(__power_wells);	\
+})
+
+/**
+ * intel_power_domains_init - initializes the power domain structures
+ * @dev_priv: i915 device instance
+ *
+ * Initializes the power domain structures for @dev_priv depending upon the
+ * supported platform.
+ */
+int intel_power_domains_init(struct drm_i915_private *dev_priv)
+{
+	struct i915_power_domains *power_domains = &dev_priv->power_domains;
+
+	mutex_init(&power_domains->lock);
+
+	/*
+	 * The enabling order will be from lower to higher indexed wells,
+	 * the disabling order is reversed.
+	 */
+	if (IS_HASWELL(dev_priv->dev)) {
+		set_power_wells(power_domains, hsw_power_wells);
+		hsw_pwr = power_domains;
+	} else if (IS_BROADWELL(dev_priv->dev)) {
+		set_power_wells(power_domains, bdw_power_wells);
+		hsw_pwr = power_domains;
+	} else if (IS_CHERRYVIEW(dev_priv->dev)) {
+		set_power_wells(power_domains, chv_power_wells);
+	} else if (IS_VALLEYVIEW(dev_priv->dev)) {
+		set_power_wells(power_domains, vlv_power_wells);
+	} else {
+		set_power_wells(power_domains, i9xx_always_on_power_well);
+	}
+
+	return 0;
+}
+
+static void intel_runtime_pm_disable(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct device *device = &dev->pdev->dev;
+
+	if (!HAS_RUNTIME_PM(dev))
+		return;
+
+	if (!intel_enable_rc6(dev))
+		return;
+
+	/* Make sure we're not suspended first. */
+	pm_runtime_get_sync(device);
+	pm_runtime_disable(device);
+}
+
+/**
+ * intel_power_domains_fini - finalizes the power domain structures
+ * @dev_priv: i915 device instance
+ *
+ * Finalizes the power domain structures for @dev_priv depending upon the
+ * supported platform. This function also disables runtime pm and ensures that
+ * the device stays powered up so that the driver can be reloaded.
+ */
+void intel_power_domains_fini(struct drm_i915_private *dev_priv)
+{
+	intel_runtime_pm_disable(dev_priv);
+
+	/* The i915.ko module is still not prepared to be loaded when
+	 * the power well is not enabled, so just enable it in case
+	 * we're going to unload/reload. */
+	intel_display_set_init_power(dev_priv, true);
+
+	hsw_pwr = NULL;
+}
+
+static void intel_power_domains_resume(struct drm_i915_private *dev_priv)
+{
+	struct i915_power_domains *power_domains = &dev_priv->power_domains;
+	struct i915_power_well *power_well;
+	int i;
+
+	mutex_lock(&power_domains->lock);
+	for_each_power_well(i, power_well, POWER_DOMAIN_MASK, power_domains) {
+		power_well->ops->sync_hw(dev_priv, power_well);
+		power_well->hw_enabled = power_well->ops->is_enabled(dev_priv,
+								     power_well);
+	}
+	mutex_unlock(&power_domains->lock);
+}
+
+static void vlv_cmnlane_wa(struct drm_i915_private *dev_priv)
+{
+	struct i915_power_well *cmn =
+		lookup_power_well(dev_priv, PUNIT_POWER_WELL_DPIO_CMN_BC);
+	struct i915_power_well *disp2d =
+		lookup_power_well(dev_priv, PUNIT_POWER_WELL_DISP2D);
+
+	/* If the display might be already active skip this */
+	if (cmn->ops->is_enabled(dev_priv, cmn) &&
+	    disp2d->ops->is_enabled(dev_priv, disp2d) &&
+	    I915_READ(DPIO_CTL) & DPIO_CMNRST)
+		return;
+
+	DRM_DEBUG_KMS("toggling display PHY side reset\n");
+
+	/* cmnlane needs DPLL registers */
+	disp2d->ops->enable(dev_priv, disp2d);
+
+	/*
+	 * From VLV2A0_DP_eDP_HDMI_DPIO_driver_vbios_notes_11.docx:
+	 * Need to assert and de-assert PHY SB reset by gating the
+	 * common lane power, then un-gating it.
+	 * Simply ungating isn't enough to reset the PHY enough to get
+	 * ports and lanes running.
+	 */
+	cmn->ops->disable(dev_priv, cmn);
+}
+
+/**
+ * intel_power_domains_init_hw - initialize hardware power domain state
+ * @dev_priv: i915 device instance
+ *
+ * This function initializes the hardware power domain state and enables all
+ * power domains using intel_display_set_init_power().
+ */
+void intel_power_domains_init_hw(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct i915_power_domains *power_domains = &dev_priv->power_domains;
+
+	power_domains->initializing = true;
+
+	if (IS_VALLEYVIEW(dev) && !IS_CHERRYVIEW(dev)) {
+		mutex_lock(&power_domains->lock);
+		vlv_cmnlane_wa(dev_priv);
+		mutex_unlock(&power_domains->lock);
+	}
+
+	/* For now, we need the power well to be always enabled. */
+	intel_display_set_init_power(dev_priv, true);
+	intel_power_domains_resume(dev_priv);
+	power_domains->initializing = false;
+}
+
+/**
+ * intel_aux_display_runtime_get - grab an auxilliary power domain reference
+ * @dev_priv: i915 device instance
+ *
+ * This function grabs a power domain reference for the auxiliary power domain
+ * (for access to the GMBUS and DP AUX blocks) and ensures that it and all its
+ * parents are powered up. Therefore users should only grab a reference to the
+ * innermost power domain they need.
+ *
+ * Any power domain reference obtained by this function must have a symmetric
+ * call to intel_aux_display_runtime_put() to release the reference again.
+ */
+void intel_aux_display_runtime_get(struct drm_i915_private *dev_priv)
+{
+	intel_runtime_pm_get(dev_priv);
+}
+
+/**
+ * intel_aux_display_runtime_put - release an auxilliary power domain reference
+ * @dev_priv: i915 device instance
+ *
+ * This function drops the auxilliary power domain reference obtained by
+ * intel_aux_display_runtime_get() and might power down the corresponding
+ * hardware block right away if this is the last reference.
+ */
+void intel_aux_display_runtime_put(struct drm_i915_private *dev_priv)
+{
+	intel_runtime_pm_put(dev_priv);
+}
+
+/**
+ * intel_runtime_pm_get - grab a runtime pm reference
+ * @dev_priv: i915 device instance
+ *
+ * This function grabs a device-level runtime pm reference (mostly used for GEM
+ * code to ensure the GTT or GT is on) and ensures that it is powered up.
+ *
+ * Any runtime pm reference obtained by this function must have a symmetric
+ * call to intel_runtime_pm_put() to release the reference again.
+ */
+void intel_runtime_pm_get(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct device *device = &dev->pdev->dev;
+
+	if (!HAS_RUNTIME_PM(dev))
+		return;
+
+	pm_runtime_get_sync(device);
+	WARN(dev_priv->pm.suspended, "Device still suspended.\n");
+}
+
+/**
+ * intel_runtime_pm_get_noresume - grab a runtime pm reference
+ * @dev_priv: i915 device instance
+ *
+ * This function grabs a device-level runtime pm reference (mostly used for GEM
+ * code to ensure the GTT or GT is on).
+ *
+ * It will _not_ power up the device but instead only check that it's powered
+ * on.  Therefore it is only valid to call this functions from contexts where
+ * the device is known to be powered up and where trying to power it up would
+ * result in hilarity and deadlocks. That pretty much means only the system
+ * suspend/resume code where this is used to grab runtime pm references for
+ * delayed setup down in work items.
+ *
+ * Any runtime pm reference obtained by this function must have a symmetric
+ * call to intel_runtime_pm_put() to release the reference again.
+ */
+void intel_runtime_pm_get_noresume(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct device *device = &dev->pdev->dev;
+
+	if (!HAS_RUNTIME_PM(dev))
+		return;
+
+	WARN(dev_priv->pm.suspended, "Getting nosync-ref while suspended.\n");
+	pm_runtime_get_noresume(device);
+}
+
+/**
+ * intel_runtime_pm_put - release a runtime pm reference
+ * @dev_priv: i915 device instance
+ *
+ * This function drops the device-level runtime pm reference obtained by
+ * intel_runtime_pm_get() and might power down the corresponding
+ * hardware block right away if this is the last reference.
+ */
+void intel_runtime_pm_put(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct device *device = &dev->pdev->dev;
+
+	if (!HAS_RUNTIME_PM(dev))
+		return;
+
+	pm_runtime_mark_last_busy(device);
+	pm_runtime_put_autosuspend(device);
+}
+
+/**
+ * intel_runtime_pm_enable - enable runtime pm
+ * @dev_priv: i915 device instance
+ *
+ * This function enables runtime pm at the end of the driver load sequence.
+ *
+ * Note that this function does currently not enable runtime pm for the
+ * subordinate display power domains. That is only done on the first modeset
+ * using intel_display_set_init_power().
+ */
+void intel_runtime_pm_enable(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct device *device = &dev->pdev->dev;
+
+	if (!HAS_RUNTIME_PM(dev))
+		return;
+
+	pm_runtime_set_active(device);
+
+	/*
+	 * RPM depends on RC6 to save restore the GT HW context, so make RC6 a
+	 * requirement.
+	 */
+	if (!intel_enable_rc6(dev)) {
+		DRM_INFO("RC6 disabled, disabling runtime PM support\n");
+		return;
+	}
+
+	pm_runtime_set_autosuspend_delay(device, 10000); /* 10s */
+	pm_runtime_mark_last_busy(device);
+	pm_runtime_use_autosuspend(device);
+
+	pm_runtime_put_autosuspend(device);
+}
+
+/* Display audio driver power well request */
+int i915_request_power_well(void)
+{
+	struct drm_i915_private *dev_priv;
+
+	if (!hsw_pwr)
+		return -ENODEV;
+
+	dev_priv = container_of(hsw_pwr, struct drm_i915_private,
+				power_domains);
+	intel_display_power_get(dev_priv, POWER_DOMAIN_AUDIO);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(i915_request_power_well);
+
+/* Display audio driver power well release */
+int i915_release_power_well(void)
+{
+	struct drm_i915_private *dev_priv;
+
+	if (!hsw_pwr)
+		return -ENODEV;
+
+	dev_priv = container_of(hsw_pwr, struct drm_i915_private,
+				power_domains);
+	intel_display_power_put(dev_priv, POWER_DOMAIN_AUDIO);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(i915_release_power_well);
+
+/*
+ * Private interface for the audio driver to get CDCLK in kHz.
+ *
+ * Caller must request power well using i915_request_power_well() prior to
+ * making the call.
+ */
+int i915_get_cdclk_freq(void)
+{
+	struct drm_i915_private *dev_priv;
+
+	if (!hsw_pwr)
+		return -ENODEV;
+
+	dev_priv = container_of(hsw_pwr, struct drm_i915_private,
+				power_domains);
+
+	return intel_ddi_get_cdclk_freq(dev_priv);
+}
+EXPORT_SYMBOL_GPL(i915_get_cdclk_freq);
diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c
index 9350edd6728d..6d7a277458b5 100644
--- a/drivers/gpu/drm/i915/intel_sdvo.c
+++ b/drivers/gpu/drm/i915/intel_sdvo.c
@@ -1991,57 +1991,10 @@ static int intel_sdvo_get_modes(struct drm_connector *connector)
 	return !list_empty(&connector->probed_modes);
 }
 
-static void
-intel_sdvo_destroy_enhance_property(struct drm_connector *connector)
-{
-	struct intel_sdvo_connector *intel_sdvo_connector = to_intel_sdvo_connector(connector);
-	struct drm_device *dev = connector->dev;
-
-	if (intel_sdvo_connector->left)
-		drm_property_destroy(dev, intel_sdvo_connector->left);
-	if (intel_sdvo_connector->right)
-		drm_property_destroy(dev, intel_sdvo_connector->right);
-	if (intel_sdvo_connector->top)
-		drm_property_destroy(dev, intel_sdvo_connector->top);
-	if (intel_sdvo_connector->bottom)
-		drm_property_destroy(dev, intel_sdvo_connector->bottom);
-	if (intel_sdvo_connector->hpos)
-		drm_property_destroy(dev, intel_sdvo_connector->hpos);
-	if (intel_sdvo_connector->vpos)
-		drm_property_destroy(dev, intel_sdvo_connector->vpos);
-	if (intel_sdvo_connector->saturation)
-		drm_property_destroy(dev, intel_sdvo_connector->saturation);
-	if (intel_sdvo_connector->contrast)
-		drm_property_destroy(dev, intel_sdvo_connector->contrast);
-	if (intel_sdvo_connector->hue)
-		drm_property_destroy(dev, intel_sdvo_connector->hue);
-	if (intel_sdvo_connector->sharpness)
-		drm_property_destroy(dev, intel_sdvo_connector->sharpness);
-	if (intel_sdvo_connector->flicker_filter)
-		drm_property_destroy(dev, intel_sdvo_connector->flicker_filter);
-	if (intel_sdvo_connector->flicker_filter_2d)
-		drm_property_destroy(dev, intel_sdvo_connector->flicker_filter_2d);
-	if (intel_sdvo_connector->flicker_filter_adaptive)
-		drm_property_destroy(dev, intel_sdvo_connector->flicker_filter_adaptive);
-	if (intel_sdvo_connector->tv_luma_filter)
-		drm_property_destroy(dev, intel_sdvo_connector->tv_luma_filter);
-	if (intel_sdvo_connector->tv_chroma_filter)
-		drm_property_destroy(dev, intel_sdvo_connector->tv_chroma_filter);
-	if (intel_sdvo_connector->dot_crawl)
-		drm_property_destroy(dev, intel_sdvo_connector->dot_crawl);
-	if (intel_sdvo_connector->brightness)
-		drm_property_destroy(dev, intel_sdvo_connector->brightness);
-}
-
 static void intel_sdvo_destroy(struct drm_connector *connector)
 {
 	struct intel_sdvo_connector *intel_sdvo_connector = to_intel_sdvo_connector(connector);
 
-	if (intel_sdvo_connector->tv_format)
-		drm_property_destroy(connector->dev,
-				     intel_sdvo_connector->tv_format);
-
-	intel_sdvo_destroy_enhance_property(connector);
 	drm_connector_cleanup(connector);
 	kfree(intel_sdvo_connector);
 }
diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c
index 168c6652cda1..7d9c340f7693 100644
--- a/drivers/gpu/drm/i915/intel_sprite.c
+++ b/drivers/gpu/drm/i915/intel_sprite.c
@@ -37,6 +37,20 @@
 #include <drm/i915_drm.h>
 #include "i915_drv.h"
 
+static bool
+format_is_yuv(uint32_t format)
+{
+	switch (format) {
+	case DRM_FORMAT_YUYV:
+	case DRM_FORMAT_UYVY:
+	case DRM_FORMAT_VYUY:
+	case DRM_FORMAT_YVYU:
+		return true;
+	default:
+		return false;
+	}
+}
+
 static int usecs_to_scanlines(const struct drm_display_mode *mode, int usecs)
 {
 	/* paranoia */
@@ -46,17 +60,32 @@ static int usecs_to_scanlines(const struct drm_display_mode *mode, int usecs)
 	return DIV_ROUND_UP(usecs * mode->crtc_clock, 1000 * mode->crtc_htotal);
 }
 
-static bool intel_pipe_update_start(struct intel_crtc *crtc, uint32_t *start_vbl_count)
+/**
+ * intel_pipe_update_start() - start update of a set of display registers
+ * @crtc: the crtc of which the registers are going to be updated
+ * @start_vbl_count: vblank counter return pointer used for error checking
+ *
+ * Mark the start of an update to pipe registers that should be updated
+ * atomically regarding vblank. If the next vblank will happens within
+ * the next 100 us, this function waits until the vblank passes.
+ *
+ * After a successful call to this function, interrupts will be disabled
+ * until a subsequent call to intel_pipe_update_end(). That is done to
+ * avoid random delays. The value written to @start_vbl_count should be
+ * supplied to intel_pipe_update_end() for error checking.
+ *
+ * Return: true if the call was successful
+ */
+bool intel_pipe_update_start(struct intel_crtc *crtc, uint32_t *start_vbl_count)
 {
 	struct drm_device *dev = crtc->base.dev;
 	const struct drm_display_mode *mode = &crtc->config.adjusted_mode;
 	enum pipe pipe = crtc->pipe;
 	long timeout = msecs_to_jiffies_timeout(1);
 	int scanline, min, max, vblank_start;
+	wait_queue_head_t *wq = drm_crtc_vblank_waitqueue(&crtc->base);
 	DEFINE_WAIT(wait);
 
-	WARN_ON(!drm_modeset_is_locked(&crtc->base.mutex));
-
 	vblank_start = mode->crtc_vblank_start;
 	if (mode->flags & DRM_MODE_FLAG_INTERLACE)
 		vblank_start = DIV_ROUND_UP(vblank_start, 2);
@@ -81,7 +110,7 @@ static bool intel_pipe_update_start(struct intel_crtc *crtc, uint32_t *start_vbl
 		 * other CPUs can see the task state update by the time we
 		 * read the scanline.
 		 */
-		prepare_to_wait(&crtc->vbl_wait, &wait, TASK_UNINTERRUPTIBLE);
+		prepare_to_wait(wq, &wait, TASK_UNINTERRUPTIBLE);
 
 		scanline = intel_get_crtc_scanline(crtc);
 		if (scanline < min || scanline > max)
@@ -100,7 +129,7 @@ static bool intel_pipe_update_start(struct intel_crtc *crtc, uint32_t *start_vbl
 		local_irq_disable();
 	}
 
-	finish_wait(&crtc->vbl_wait, &wait);
+	finish_wait(wq, &wait);
 
 	drm_vblank_put(dev, pipe);
 
@@ -111,7 +140,16 @@ static bool intel_pipe_update_start(struct intel_crtc *crtc, uint32_t *start_vbl
 	return true;
 }
 
-static void intel_pipe_update_end(struct intel_crtc *crtc, u32 start_vbl_count)
+/**
+ * intel_pipe_update_end() - end update of a set of display registers
+ * @crtc: the crtc of which the registers were updated
+ * @start_vbl_count: start vblank counter (used for error checking)
+ *
+ * Mark the end of an update started with intel_pipe_update_start(). This
+ * re-enables interrupts and verifies the update was actually completed
+ * before a vblank using the value of @start_vbl_count.
+ */
+void intel_pipe_update_end(struct intel_crtc *crtc, u32 start_vbl_count)
 {
 	struct drm_device *dev = crtc->base.dev;
 	enum pipe pipe = crtc->pipe;
@@ -138,6 +176,226 @@ static void intel_update_primary_plane(struct intel_crtc *crtc)
 }
 
 static void
+skl_update_plane(struct drm_plane *drm_plane, struct drm_crtc *crtc,
+		 struct drm_framebuffer *fb,
+		 struct drm_i915_gem_object *obj, int crtc_x, int crtc_y,
+		 unsigned int crtc_w, unsigned int crtc_h,
+		 uint32_t x, uint32_t y,
+		 uint32_t src_w, uint32_t src_h)
+{
+	struct drm_device *dev = drm_plane->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_plane *intel_plane = to_intel_plane(drm_plane);
+	const int pipe = intel_plane->pipe;
+	const int plane = intel_plane->plane + 1;
+	u32 plane_ctl, stride;
+	int pixel_size = drm_format_plane_cpp(fb->pixel_format, 0);
+
+	plane_ctl = I915_READ(PLANE_CTL(pipe, plane));
+
+	/* Mask out pixel format bits in case we change it */
+	plane_ctl &= ~PLANE_CTL_FORMAT_MASK;
+	plane_ctl &= ~PLANE_CTL_ORDER_RGBX;
+	plane_ctl &= ~PLANE_CTL_YUV422_ORDER_MASK;
+	plane_ctl &= ~PLANE_CTL_TILED_MASK;
+	plane_ctl &= ~PLANE_CTL_ALPHA_MASK;
+	plane_ctl &= ~PLANE_CTL_ROTATE_MASK;
+
+	/* Trickle feed has to be enabled */
+	plane_ctl &= ~PLANE_CTL_TRICKLE_FEED_DISABLE;
+
+	switch (fb->pixel_format) {
+	case DRM_FORMAT_RGB565:
+		plane_ctl |= PLANE_CTL_FORMAT_RGB_565;
+		break;
+	case DRM_FORMAT_XBGR8888:
+		plane_ctl |= PLANE_CTL_FORMAT_XRGB_8888 | PLANE_CTL_ORDER_RGBX;
+		break;
+	case DRM_FORMAT_XRGB8888:
+		plane_ctl |= PLANE_CTL_FORMAT_XRGB_8888;
+		break;
+	/*
+	 * XXX: For ARBG/ABGR formats we default to expecting scanout buffers
+	 * to be already pre-multiplied. We need to add a knob (or a different
+	 * DRM_FORMAT) for user-space to configure that.
+	 */
+	case DRM_FORMAT_ABGR8888:
+		plane_ctl |= PLANE_CTL_FORMAT_XRGB_8888 |
+			     PLANE_CTL_ORDER_RGBX |
+			     PLANE_CTL_ALPHA_SW_PREMULTIPLY;
+		break;
+	case DRM_FORMAT_ARGB8888:
+		plane_ctl |= PLANE_CTL_FORMAT_XRGB_8888 |
+			     PLANE_CTL_ALPHA_SW_PREMULTIPLY;
+		break;
+	case DRM_FORMAT_YUYV:
+		plane_ctl |= PLANE_CTL_FORMAT_YUV422 | PLANE_CTL_YUV422_YUYV;
+		break;
+	case DRM_FORMAT_YVYU:
+		plane_ctl |= PLANE_CTL_FORMAT_YUV422 | PLANE_CTL_YUV422_YVYU;
+		break;
+	case DRM_FORMAT_UYVY:
+		plane_ctl |= PLANE_CTL_FORMAT_YUV422 | PLANE_CTL_YUV422_UYVY;
+		break;
+	case DRM_FORMAT_VYUY:
+		plane_ctl |= PLANE_CTL_FORMAT_YUV422 | PLANE_CTL_YUV422_VYUY;
+		break;
+	default:
+		BUG();
+	}
+
+	switch (obj->tiling_mode) {
+	case I915_TILING_NONE:
+		stride = fb->pitches[0] >> 6;
+		break;
+	case I915_TILING_X:
+		plane_ctl |= PLANE_CTL_TILED_X;
+		stride = fb->pitches[0] >> 9;
+		break;
+	default:
+		BUG();
+	}
+	if (intel_plane->rotation == BIT(DRM_ROTATE_180))
+		plane_ctl |= PLANE_CTL_ROTATE_180;
+
+	plane_ctl |= PLANE_CTL_ENABLE;
+	plane_ctl |= PLANE_CTL_PIPE_CSC_ENABLE;
+
+	intel_update_sprite_watermarks(drm_plane, crtc, src_w, src_h,
+				       pixel_size, true,
+				       src_w != crtc_w || src_h != crtc_h);
+
+	/* Sizes are 0 based */
+	src_w--;
+	src_h--;
+	crtc_w--;
+	crtc_h--;
+
+	I915_WRITE(PLANE_OFFSET(pipe, plane), (y << 16) | x);
+	I915_WRITE(PLANE_STRIDE(pipe, plane), stride);
+	I915_WRITE(PLANE_POS(pipe, plane), (crtc_y << 16) | crtc_x);
+	I915_WRITE(PLANE_SIZE(pipe, plane), (crtc_h << 16) | crtc_w);
+	I915_WRITE(PLANE_CTL(pipe, plane), plane_ctl);
+	I915_WRITE(PLANE_SURF(pipe, plane), i915_gem_obj_ggtt_offset(obj));
+	POSTING_READ(PLANE_SURF(pipe, plane));
+}
+
+static void
+skl_disable_plane(struct drm_plane *drm_plane, struct drm_crtc *crtc)
+{
+	struct drm_device *dev = drm_plane->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_plane *intel_plane = to_intel_plane(drm_plane);
+	const int pipe = intel_plane->pipe;
+	const int plane = intel_plane->plane + 1;
+
+	I915_WRITE(PLANE_CTL(pipe, plane),
+		   I915_READ(PLANE_CTL(pipe, plane)) & ~PLANE_CTL_ENABLE);
+
+	/* Activate double buffered register update */
+	I915_WRITE(PLANE_CTL(pipe, plane), 0);
+	POSTING_READ(PLANE_CTL(pipe, plane));
+
+	intel_update_sprite_watermarks(drm_plane, crtc, 0, 0, 0, false, false);
+}
+
+static int
+skl_update_colorkey(struct drm_plane *drm_plane,
+		    struct drm_intel_sprite_colorkey *key)
+{
+	struct drm_device *dev = drm_plane->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_plane *intel_plane = to_intel_plane(drm_plane);
+	const int pipe = intel_plane->pipe;
+	const int plane = intel_plane->plane;
+	u32 plane_ctl;
+
+	I915_WRITE(PLANE_KEYVAL(pipe, plane), key->min_value);
+	I915_WRITE(PLANE_KEYMAX(pipe, plane), key->max_value);
+	I915_WRITE(PLANE_KEYMSK(pipe, plane), key->channel_mask);
+
+	plane_ctl = I915_READ(PLANE_CTL(pipe, plane));
+	plane_ctl &= ~PLANE_CTL_KEY_ENABLE_MASK;
+	if (key->flags & I915_SET_COLORKEY_DESTINATION)
+		plane_ctl |= PLANE_CTL_KEY_ENABLE_DESTINATION;
+	else if (key->flags & I915_SET_COLORKEY_SOURCE)
+		plane_ctl |= PLANE_CTL_KEY_ENABLE_SOURCE;
+	I915_WRITE(PLANE_CTL(pipe, plane), plane_ctl);
+
+	POSTING_READ(PLANE_CTL(pipe, plane));
+
+	return 0;
+}
+
+static void
+skl_get_colorkey(struct drm_plane *drm_plane,
+		 struct drm_intel_sprite_colorkey *key)
+{
+	struct drm_device *dev = drm_plane->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_plane *intel_plane = to_intel_plane(drm_plane);
+	const int pipe = intel_plane->pipe;
+	const int plane = intel_plane->plane;
+	u32 plane_ctl;
+
+	key->min_value = I915_READ(PLANE_KEYVAL(pipe, plane));
+	key->max_value = I915_READ(PLANE_KEYMAX(pipe, plane));
+	key->channel_mask = I915_READ(PLANE_KEYMSK(pipe, plane));
+
+	plane_ctl = I915_READ(PLANE_CTL(pipe, plane));
+
+	switch (plane_ctl & PLANE_CTL_KEY_ENABLE_MASK) {
+	case PLANE_CTL_KEY_ENABLE_DESTINATION:
+		key->flags = I915_SET_COLORKEY_DESTINATION;
+		break;
+	case PLANE_CTL_KEY_ENABLE_SOURCE:
+		key->flags = I915_SET_COLORKEY_SOURCE;
+		break;
+	default:
+		key->flags = I915_SET_COLORKEY_NONE;
+	}
+}
+
+static void
+chv_update_csc(struct intel_plane *intel_plane, uint32_t format)
+{
+	struct drm_i915_private *dev_priv = intel_plane->base.dev->dev_private;
+	int plane = intel_plane->plane;
+
+	/* Seems RGB data bypasses the CSC always */
+	if (!format_is_yuv(format))
+		return;
+
+	/*
+	 * BT.601 limited range YCbCr -> full range RGB
+	 *
+	 * |r|   | 6537 4769     0|   |cr  |
+	 * |g| = |-3330 4769 -1605| x |y-64|
+	 * |b|   |    0 4769  8263|   |cb  |
+	 *
+	 * Cb and Cr apparently come in as signed already, so no
+	 * need for any offset. For Y we need to remove the offset.
+	 */
+	I915_WRITE(SPCSCYGOFF(plane), SPCSC_OOFF(0) | SPCSC_IOFF(-64));
+	I915_WRITE(SPCSCCBOFF(plane), SPCSC_OOFF(0) | SPCSC_IOFF(0));
+	I915_WRITE(SPCSCCROFF(plane), SPCSC_OOFF(0) | SPCSC_IOFF(0));
+
+	I915_WRITE(SPCSCC01(plane), SPCSC_C1(4769) | SPCSC_C0(6537));
+	I915_WRITE(SPCSCC23(plane), SPCSC_C1(-3330) | SPCSC_C0(0));
+	I915_WRITE(SPCSCC45(plane), SPCSC_C1(-1605) | SPCSC_C0(4769));
+	I915_WRITE(SPCSCC67(plane), SPCSC_C1(4769) | SPCSC_C0(0));
+	I915_WRITE(SPCSCC8(plane), SPCSC_C0(8263));
+
+	I915_WRITE(SPCSCYGICLAMP(plane), SPCSC_IMAX(940) | SPCSC_IMIN(64));
+	I915_WRITE(SPCSCCBICLAMP(plane), SPCSC_IMAX(448) | SPCSC_IMIN(-448));
+	I915_WRITE(SPCSCCRICLAMP(plane), SPCSC_IMAX(448) | SPCSC_IMIN(-448));
+
+	I915_WRITE(SPCSCYGOCLAMP(plane), SPCSC_OMAX(1023) | SPCSC_OMIN(0));
+	I915_WRITE(SPCSCCBOCLAMP(plane), SPCSC_OMAX(1023) | SPCSC_OMIN(0));
+	I915_WRITE(SPCSCCROCLAMP(plane), SPCSC_OMAX(1023) | SPCSC_OMIN(0));
+}
+
+static void
 vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc,
 		 struct drm_framebuffer *fb,
 		 struct drm_i915_gem_object *obj, int crtc_x, int crtc_y,
@@ -163,6 +421,7 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc,
 	sprctl &= ~SP_PIXFORMAT_MASK;
 	sprctl &= ~SP_YUV_BYTE_ORDER_MASK;
 	sprctl &= ~SP_TILED;
+	sprctl &= ~SP_ROTATE_180;
 
 	switch (fb->pixel_format) {
 	case DRM_FORMAT_YUYV:
@@ -235,10 +494,21 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc,
 							fb->pitches[0]);
 	linear_offset -= sprsurf_offset;
 
+	if (intel_plane->rotation == BIT(DRM_ROTATE_180)) {
+		sprctl |= SP_ROTATE_180;
+
+		x += src_w;
+		y += src_h;
+		linear_offset += src_h * fb->pitches[0] + src_w * pixel_size;
+	}
+
 	atomic_update = intel_pipe_update_start(intel_crtc, &start_vbl_count);
 
 	intel_update_primary_plane(intel_crtc);
 
+	if (IS_CHERRYVIEW(dev) && pipe == PIPE_B)
+		chv_update_csc(intel_plane, fb->pixel_format);
+
 	I915_WRITE(SPSTRIDE(pipe, plane), fb->pitches[0]);
 	I915_WRITE(SPPOS(pipe, plane), (crtc_y << 16) | crtc_x);
 
@@ -247,6 +517,8 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc,
 	else
 		I915_WRITE(SPLINOFF(pipe, plane), linear_offset);
 
+	I915_WRITE(SPCONSTALPHA(pipe, plane), 0);
+
 	I915_WRITE(SPSIZE(pipe, plane), (crtc_h << 16) | crtc_w);
 	I915_WRITE(SPCNTR(pipe, plane), sprctl);
 	I915_WRITE(SPSURF(pipe, plane), i915_gem_obj_ggtt_offset(obj) +
@@ -364,6 +636,7 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	sprctl &= ~SPRITE_RGB_ORDER_RGBX;
 	sprctl &= ~SPRITE_YUV_BYTE_ORDER_MASK;
 	sprctl &= ~SPRITE_TILED;
+	sprctl &= ~SPRITE_ROTATE_180;
 
 	switch (fb->pixel_format) {
 	case DRM_FORMAT_XBGR8888:
@@ -426,6 +699,18 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 					       pixel_size, fb->pitches[0]);
 	linear_offset -= sprsurf_offset;
 
+	if (intel_plane->rotation == BIT(DRM_ROTATE_180)) {
+		sprctl |= SPRITE_ROTATE_180;
+
+		/* HSW and BDW does this automagically in hardware */
+		if (!IS_HASWELL(dev) && !IS_BROADWELL(dev)) {
+			x += src_w;
+			y += src_h;
+			linear_offset += src_h * fb->pitches[0] +
+				src_w * pixel_size;
+		}
+	}
+
 	atomic_update = intel_pipe_update_start(intel_crtc, &start_vbl_count);
 
 	intel_update_primary_plane(intel_crtc);
@@ -571,6 +856,7 @@ ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	dvscntr &= ~DVS_RGB_ORDER_XBGR;
 	dvscntr &= ~DVS_YUV_BYTE_ORDER_MASK;
 	dvscntr &= ~DVS_TILED;
+	dvscntr &= ~DVS_ROTATE_180;
 
 	switch (fb->pixel_format) {
 	case DRM_FORMAT_XBGR8888:
@@ -628,6 +914,14 @@ ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 					       pixel_size, fb->pitches[0]);
 	linear_offset -= dvssurf_offset;
 
+	if (intel_plane->rotation == BIT(DRM_ROTATE_180)) {
+		dvscntr |= DVS_ROTATE_180;
+
+		x += src_w;
+		y += src_h;
+		linear_offset += src_h * fb->pitches[0] + src_w * pixel_size;
+	}
+
 	atomic_update = intel_pipe_update_start(intel_crtc, &start_vbl_count);
 
 	intel_update_primary_plane(intel_crtc);
@@ -789,20 +1083,6 @@ ilk_get_colorkey(struct drm_plane *plane, struct drm_intel_sprite_colorkey *key)
 		key->flags = I915_SET_COLORKEY_NONE;
 }
 
-static bool
-format_is_yuv(uint32_t format)
-{
-	switch (format) {
-	case DRM_FORMAT_YUYV:
-	case DRM_FORMAT_UYVY:
-	case DRM_FORMAT_VYUY:
-	case DRM_FORMAT_YVYU:
-		return true;
-	default:
-		return false;
-	}
-}
-
 static bool colorkey_enabled(struct intel_plane *intel_plane)
 {
 	struct drm_intel_sprite_colorkey key;
@@ -813,57 +1093,23 @@ static bool colorkey_enabled(struct intel_plane *intel_plane)
 }
 
 static int
-intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
-		   struct drm_framebuffer *fb, int crtc_x, int crtc_y,
-		   unsigned int crtc_w, unsigned int crtc_h,
-		   uint32_t src_x, uint32_t src_y,
-		   uint32_t src_w, uint32_t src_h)
+intel_check_sprite_plane(struct drm_plane *plane,
+			 struct intel_plane_state *state)
 {
-	struct drm_device *dev = plane->dev;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(state->crtc);
 	struct intel_plane *intel_plane = to_intel_plane(plane);
-	enum pipe pipe = intel_crtc->pipe;
-	struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb);
-	struct drm_i915_gem_object *obj = intel_fb->obj;
-	struct drm_i915_gem_object *old_obj = intel_plane->obj;
-	int ret;
-	bool primary_enabled;
-	bool visible;
+	struct drm_framebuffer *fb = state->fb;
+	struct drm_i915_gem_object *obj = intel_fb_obj(fb);
+	int crtc_x, crtc_y;
+	unsigned int crtc_w, crtc_h;
+	uint32_t src_x, src_y, src_w, src_h;
+	struct drm_rect *src = &state->src;
+	struct drm_rect *dst = &state->dst;
+	struct drm_rect *orig_src = &state->orig_src;
+	const struct drm_rect *clip = &state->clip;
 	int hscale, vscale;
 	int max_scale, min_scale;
 	int pixel_size = drm_format_plane_cpp(fb->pixel_format, 0);
-	struct drm_rect src = {
-		/* sample coordinates in 16.16 fixed point */
-		.x1 = src_x,
-		.x2 = src_x + src_w,
-		.y1 = src_y,
-		.y2 = src_y + src_h,
-	};
-	struct drm_rect dst = {
-		/* integer pixels */
-		.x1 = crtc_x,
-		.x2 = crtc_x + crtc_w,
-		.y1 = crtc_y,
-		.y2 = crtc_y + crtc_h,
-	};
-	const struct drm_rect clip = {
-		.x2 = intel_crtc->active ? intel_crtc->config.pipe_src_w : 0,
-		.y2 = intel_crtc->active ? intel_crtc->config.pipe_src_h : 0,
-	};
-	const struct {
-		int crtc_x, crtc_y;
-		unsigned int crtc_w, crtc_h;
-		uint32_t src_x, src_y, src_w, src_h;
-	} orig = {
-		.crtc_x = crtc_x,
-		.crtc_y = crtc_y,
-		.crtc_w = crtc_w,
-		.crtc_h = crtc_h,
-		.src_x = src_x,
-		.src_y = src_y,
-		.src_w = src_w,
-		.src_h = src_h,
-	};
 
 	/* Don't modify another pipe's plane */
 	if (intel_plane->pipe != intel_crtc->pipe) {
@@ -895,49 +1141,55 @@ intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	max_scale = intel_plane->max_downscale << 16;
 	min_scale = intel_plane->can_scale ? 1 : (1 << 16);
 
-	hscale = drm_rect_calc_hscale_relaxed(&src, &dst, min_scale, max_scale);
+	drm_rect_rotate(src, fb->width << 16, fb->height << 16,
+			intel_plane->rotation);
+
+	hscale = drm_rect_calc_hscale_relaxed(src, dst, min_scale, max_scale);
 	BUG_ON(hscale < 0);
 
-	vscale = drm_rect_calc_vscale_relaxed(&src, &dst, min_scale, max_scale);
+	vscale = drm_rect_calc_vscale_relaxed(src, dst, min_scale, max_scale);
 	BUG_ON(vscale < 0);
 
-	visible = drm_rect_clip_scaled(&src, &dst, &clip, hscale, vscale);
+	state->visible =  drm_rect_clip_scaled(src, dst, clip, hscale, vscale);
 
-	crtc_x = dst.x1;
-	crtc_y = dst.y1;
-	crtc_w = drm_rect_width(&dst);
-	crtc_h = drm_rect_height(&dst);
+	crtc_x = dst->x1;
+	crtc_y = dst->y1;
+	crtc_w = drm_rect_width(dst);
+	crtc_h = drm_rect_height(dst);
 
-	if (visible) {
+	if (state->visible) {
 		/* check again in case clipping clamped the results */
-		hscale = drm_rect_calc_hscale(&src, &dst, min_scale, max_scale);
+		hscale = drm_rect_calc_hscale(src, dst, min_scale, max_scale);
 		if (hscale < 0) {
 			DRM_DEBUG_KMS("Horizontal scaling factor out of limits\n");
-			drm_rect_debug_print(&src, true);
-			drm_rect_debug_print(&dst, false);
+			drm_rect_debug_print(src, true);
+			drm_rect_debug_print(dst, false);
 
 			return hscale;
 		}
 
-		vscale = drm_rect_calc_vscale(&src, &dst, min_scale, max_scale);
+		vscale = drm_rect_calc_vscale(src, dst, min_scale, max_scale);
 		if (vscale < 0) {
 			DRM_DEBUG_KMS("Vertical scaling factor out of limits\n");
-			drm_rect_debug_print(&src, true);
-			drm_rect_debug_print(&dst, false);
+			drm_rect_debug_print(src, true);
+			drm_rect_debug_print(dst, false);
 
 			return vscale;
 		}
 
 		/* Make the source viewport size an exact multiple of the scaling factors. */
-		drm_rect_adjust_size(&src,
-				     drm_rect_width(&dst) * hscale - drm_rect_width(&src),
-				     drm_rect_height(&dst) * vscale - drm_rect_height(&src));
+		drm_rect_adjust_size(src,
+				     drm_rect_width(dst) * hscale - drm_rect_width(src),
+				     drm_rect_height(dst) * vscale - drm_rect_height(src));
+
+		drm_rect_rotate_inv(src, fb->width << 16, fb->height << 16,
+				    intel_plane->rotation);
 
 		/* sanity check to make sure the src viewport wasn't enlarged */
-		WARN_ON(src.x1 < (int) src_x ||
-			src.y1 < (int) src_y ||
-			src.x2 > (int) (src_x + src_w) ||
-			src.y2 > (int) (src_y + src_h));
+		WARN_ON(src->x1 < (int) orig_src->x1 ||
+			src->y1 < (int) orig_src->y1 ||
+			src->x2 > (int) orig_src->x2 ||
+			src->y2 > (int) orig_src->y2);
 
 		/*
 		 * Hardware doesn't handle subpixel coordinates.
@@ -945,10 +1197,10 @@ intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 		 * increase the source viewport size, because that could
 		 * push the downscaling factor out of bounds.
 		 */
-		src_x = src.x1 >> 16;
-		src_w = drm_rect_width(&src) >> 16;
-		src_y = src.y1 >> 16;
-		src_h = drm_rect_height(&src) >> 16;
+		src_x = src->x1 >> 16;
+		src_w = drm_rect_width(src) >> 16;
+		src_y = src->y1 >> 16;
+		src_h = drm_rect_height(src) >> 16;
 
 		if (format_is_yuv(fb->pixel_format)) {
 			src_x &= ~1;
@@ -962,12 +1214,12 @@ intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 				crtc_w &= ~1;
 
 			if (crtc_w == 0)
-				visible = false;
+				state->visible = false;
 		}
 	}
 
 	/* Check size restrictions when scaling */
-	if (visible && (src_w != crtc_w || src_h != crtc_h)) {
+	if (state->visible && (src_w != crtc_w || src_h != crtc_h)) {
 		unsigned int width_bytes;
 
 		WARN_ON(!intel_plane->can_scale);
@@ -975,12 +1227,13 @@ intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 		/* FIXME interlacing min height is 6 */
 
 		if (crtc_w < 3 || crtc_h < 3)
-			visible = false;
+			state->visible = false;
 
 		if (src_w < 3 || src_h < 3)
-			visible = false;
+			state->visible = false;
 
-		width_bytes = ((src_x * pixel_size) & 63) + src_w * pixel_size;
+		width_bytes = ((src_x * pixel_size) & 63) +
+					src_w * pixel_size;
 
 		if (src_w > 2048 || src_h > 2048 ||
 		    width_bytes > 4096 || fb->pitches[0] > 4096) {
@@ -989,42 +1242,90 @@ intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 		}
 	}
 
-	dst.x1 = crtc_x;
-	dst.x2 = crtc_x + crtc_w;
-	dst.y1 = crtc_y;
-	dst.y2 = crtc_y + crtc_h;
+	if (state->visible) {
+		src->x1 = src_x;
+		src->x2 = src_x + src_w;
+		src->y1 = src_y;
+		src->y2 = src_y + src_h;
+	}
 
-	/*
-	 * If the sprite is completely covering the primary plane,
-	 * we can disable the primary and save power.
-	 */
-	primary_enabled = !drm_rect_equals(&dst, &clip) || colorkey_enabled(intel_plane);
-	WARN_ON(!primary_enabled && !visible && intel_crtc->active);
+	dst->x1 = crtc_x;
+	dst->x2 = crtc_x + crtc_w;
+	dst->y1 = crtc_y;
+	dst->y2 = crtc_y + crtc_h;
 
-	mutex_lock(&dev->struct_mutex);
+	return 0;
+}
 
-	/* Note that this will apply the VT-d workaround for scanouts,
-	 * which is more restrictive than required for sprites. (The
-	 * primary plane requires 256KiB alignment with 64 PTE padding,
-	 * the sprite planes only require 128KiB alignment and 32 PTE padding.
-	 */
-	ret = intel_pin_and_fence_fb_obj(dev, obj, NULL);
+static int
+intel_prepare_sprite_plane(struct drm_plane *plane,
+			   struct intel_plane_state *state)
+{
+	struct drm_device *dev = plane->dev;
+	struct drm_crtc *crtc = state->crtc;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	struct intel_plane *intel_plane = to_intel_plane(plane);
+	enum pipe pipe = intel_crtc->pipe;
+	struct drm_framebuffer *fb = state->fb;
+	struct drm_i915_gem_object *obj = intel_fb_obj(fb);
+	struct drm_i915_gem_object *old_obj = intel_plane->obj;
+	int ret;
 
-	i915_gem_track_fb(old_obj, obj,
-			  INTEL_FRONTBUFFER_SPRITE(pipe));
-	mutex_unlock(&dev->struct_mutex);
+	if (old_obj != obj) {
+		mutex_lock(&dev->struct_mutex);
 
-	if (ret)
-		return ret;
+		/* Note that this will apply the VT-d workaround for scanouts,
+		 * which is more restrictive than required for sprites. (The
+		 * primary plane requires 256KiB alignment with 64 PTE padding,
+		 * the sprite planes only require 128KiB alignment and 32 PTE
+		 * padding.
+		 */
+		ret = intel_pin_and_fence_fb_obj(plane, fb, NULL);
+		if (ret == 0)
+			i915_gem_track_fb(old_obj, obj,
+					  INTEL_FRONTBUFFER_SPRITE(pipe));
+		mutex_unlock(&dev->struct_mutex);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+static void
+intel_commit_sprite_plane(struct drm_plane *plane,
+			  struct intel_plane_state *state)
+{
+	struct drm_device *dev = plane->dev;
+	struct drm_crtc *crtc = state->crtc;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	struct intel_plane *intel_plane = to_intel_plane(plane);
+	enum pipe pipe = intel_crtc->pipe;
+	struct drm_framebuffer *fb = state->fb;
+	struct drm_i915_gem_object *obj = intel_fb_obj(fb);
+	struct drm_i915_gem_object *old_obj = intel_plane->obj;
+	int crtc_x, crtc_y;
+	unsigned int crtc_w, crtc_h;
+	uint32_t src_x, src_y, src_w, src_h;
+	struct drm_rect *dst = &state->dst;
+	const struct drm_rect *clip = &state->clip;
+	bool primary_enabled;
 
-	intel_plane->crtc_x = orig.crtc_x;
-	intel_plane->crtc_y = orig.crtc_y;
-	intel_plane->crtc_w = orig.crtc_w;
-	intel_plane->crtc_h = orig.crtc_h;
-	intel_plane->src_x = orig.src_x;
-	intel_plane->src_y = orig.src_y;
-	intel_plane->src_w = orig.src_w;
-	intel_plane->src_h = orig.src_h;
+	/*
+	 * If the sprite is completely covering the primary plane,
+	 * we can disable the primary and save power.
+	 */
+	primary_enabled = !drm_rect_equals(dst, clip) || colorkey_enabled(intel_plane);
+	WARN_ON(!primary_enabled && !state->visible && intel_crtc->active);
+
+	intel_plane->crtc_x = state->orig_dst.x1;
+	intel_plane->crtc_y = state->orig_dst.y1;
+	intel_plane->crtc_w = drm_rect_width(&state->orig_dst);
+	intel_plane->crtc_h = drm_rect_height(&state->orig_dst);
+	intel_plane->src_x = state->orig_src.x1;
+	intel_plane->src_y = state->orig_src.y1;
+	intel_plane->src_w = drm_rect_width(&state->orig_src);
+	intel_plane->src_h = drm_rect_height(&state->orig_src);
 	intel_plane->obj = obj;
 
 	if (intel_crtc->active) {
@@ -1038,12 +1339,22 @@ intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 		if (primary_was_enabled && !primary_enabled)
 			intel_pre_disable_primary(crtc);
 
-		if (visible)
+		if (state->visible) {
+			crtc_x = state->dst.x1;
+			crtc_y = state->dst.y1;
+			crtc_w = drm_rect_width(&state->dst);
+			crtc_h = drm_rect_height(&state->dst);
+			src_x = state->src.x1;
+			src_y = state->src.y1;
+			src_w = drm_rect_width(&state->src);
+			src_h = drm_rect_height(&state->src);
 			intel_plane->update_plane(plane, crtc, fb, obj,
 						  crtc_x, crtc_y, crtc_w, crtc_h,
 						  src_x, src_y, src_w, src_h);
-		else
+		} else {
 			intel_plane->disable_plane(plane, crtc);
+		}
+
 
 		intel_frontbuffer_flip(dev, INTEL_FRONTBUFFER_SPRITE(pipe));
 
@@ -1052,21 +1363,65 @@ intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	}
 
 	/* Unpin old obj after new one is active to avoid ugliness */
-	if (old_obj) {
+	if (old_obj && old_obj != obj) {
+
 		/*
 		 * It's fairly common to simply update the position of
 		 * an existing object.  In that case, we don't need to
 		 * wait for vblank to avoid ugliness, we only need to
 		 * do the pin & ref bookkeeping.
 		 */
-		if (old_obj != obj && intel_crtc->active)
+		if (intel_crtc->active)
 			intel_wait_for_vblank(dev, intel_crtc->pipe);
 
 		mutex_lock(&dev->struct_mutex);
 		intel_unpin_fb_obj(old_obj);
 		mutex_unlock(&dev->struct_mutex);
 	}
+}
 
+static int
+intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
+		   struct drm_framebuffer *fb, int crtc_x, int crtc_y,
+		   unsigned int crtc_w, unsigned int crtc_h,
+		   uint32_t src_x, uint32_t src_y,
+		   uint32_t src_w, uint32_t src_h)
+{
+	struct intel_plane_state state;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	int ret;
+
+	state.crtc = crtc;
+	state.fb = fb;
+
+	/* sample coordinates in 16.16 fixed point */
+	state.src.x1 = src_x;
+	state.src.x2 = src_x + src_w;
+	state.src.y1 = src_y;
+	state.src.y2 = src_y + src_h;
+
+	/* integer pixels */
+	state.dst.x1 = crtc_x;
+	state.dst.x2 = crtc_x + crtc_w;
+	state.dst.y1 = crtc_y;
+	state.dst.y2 = crtc_y + crtc_h;
+
+	state.clip.x1 = 0;
+	state.clip.y1 = 0;
+	state.clip.x2 = intel_crtc->active ? intel_crtc->config.pipe_src_w : 0;
+	state.clip.y2 = intel_crtc->active ? intel_crtc->config.pipe_src_h : 0;
+	state.orig_src = state.src;
+	state.orig_dst = state.dst;
+
+	ret = intel_check_sprite_plane(plane, &state);
+	if (ret)
+		return ret;
+
+	ret = intel_prepare_sprite_plane(plane, &state);
+	if (ret)
+		return ret;
+
+	intel_commit_sprite_plane(plane, &state);
 	return 0;
 }
 
@@ -1180,18 +1535,45 @@ out_unlock:
 	return ret;
 }
 
-void intel_plane_restore(struct drm_plane *plane)
+int intel_plane_set_property(struct drm_plane *plane,
+			     struct drm_property *prop,
+			     uint64_t val)
+{
+	struct drm_device *dev = plane->dev;
+	struct intel_plane *intel_plane = to_intel_plane(plane);
+	uint64_t old_val;
+	int ret = -ENOENT;
+
+	if (prop == dev->mode_config.rotation_property) {
+		/* exactly one rotation angle please */
+		if (hweight32(val & 0xf) != 1)
+			return -EINVAL;
+
+		if (intel_plane->rotation == val)
+			return 0;
+
+		old_val = intel_plane->rotation;
+		intel_plane->rotation = val;
+		ret = intel_plane_restore(plane);
+		if (ret)
+			intel_plane->rotation = old_val;
+	}
+
+	return ret;
+}
+
+int intel_plane_restore(struct drm_plane *plane)
 {
 	struct intel_plane *intel_plane = to_intel_plane(plane);
 
 	if (!plane->crtc || !plane->fb)
-		return;
+		return 0;
 
-	intel_update_plane(plane, plane->crtc, plane->fb,
-			   intel_plane->crtc_x, intel_plane->crtc_y,
-			   intel_plane->crtc_w, intel_plane->crtc_h,
-			   intel_plane->src_x, intel_plane->src_y,
-			   intel_plane->src_w, intel_plane->src_h);
+	return plane->funcs->update_plane(plane, plane->crtc, plane->fb,
+				  intel_plane->crtc_x, intel_plane->crtc_y,
+				  intel_plane->crtc_w, intel_plane->crtc_h,
+				  intel_plane->src_x, intel_plane->src_y,
+				  intel_plane->src_w, intel_plane->src_h);
 }
 
 void intel_plane_disable(struct drm_plane *plane)
@@ -1206,6 +1588,7 @@ static const struct drm_plane_funcs intel_plane_funcs = {
 	.update_plane = intel_update_plane,
 	.disable_plane = intel_disable_plane,
 	.destroy = intel_destroy_plane,
+	.set_property = intel_plane_set_property,
 };
 
 static uint32_t ilk_plane_formats[] = {
@@ -1239,6 +1622,18 @@ static uint32_t vlv_plane_formats[] = {
 	DRM_FORMAT_VYUY,
 };
 
+static uint32_t skl_plane_formats[] = {
+	DRM_FORMAT_RGB565,
+	DRM_FORMAT_ABGR8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_YUYV,
+	DRM_FORMAT_YVYU,
+	DRM_FORMAT_UYVY,
+	DRM_FORMAT_VYUY,
+};
+
 int
 intel_plane_init(struct drm_device *dev, enum pipe pipe, int plane)
 {
@@ -1302,7 +1697,21 @@ intel_plane_init(struct drm_device *dev, enum pipe pipe, int plane)
 			num_plane_formats = ARRAY_SIZE(snb_plane_formats);
 		}
 		break;
-
+	case 9:
+		/*
+		 * FIXME: Skylake planes can be scaled (with some restrictions),
+		 * but this is for another time.
+		 */
+		intel_plane->can_scale = false;
+		intel_plane->max_downscale = 1;
+		intel_plane->update_plane = skl_update_plane;
+		intel_plane->disable_plane = skl_disable_plane;
+		intel_plane->update_colorkey = skl_update_colorkey;
+		intel_plane->get_colorkey = skl_get_colorkey;
+
+		plane_formats = skl_plane_formats;
+		num_plane_formats = ARRAY_SIZE(skl_plane_formats);
+		break;
 	default:
 		kfree(intel_plane);
 		return -ENODEV;
@@ -1310,13 +1719,28 @@ intel_plane_init(struct drm_device *dev, enum pipe pipe, int plane)
 
 	intel_plane->pipe = pipe;
 	intel_plane->plane = plane;
+	intel_plane->rotation = BIT(DRM_ROTATE_0);
 	possible_crtcs = (1 << pipe);
-	ret = drm_plane_init(dev, &intel_plane->base, possible_crtcs,
-			     &intel_plane_funcs,
-			     plane_formats, num_plane_formats,
-			     false);
-	if (ret)
+	ret = drm_universal_plane_init(dev, &intel_plane->base, possible_crtcs,
+				       &intel_plane_funcs,
+				       plane_formats, num_plane_formats,
+				       DRM_PLANE_TYPE_OVERLAY);
+	if (ret) {
 		kfree(intel_plane);
+		goto out;
+	}
+
+	if (!dev->mode_config.rotation_property)
+		dev->mode_config.rotation_property =
+			drm_mode_create_rotation_property(dev,
+							  BIT(DRM_ROTATE_0) |
+							  BIT(DRM_ROTATE_180));
+
+	if (dev->mode_config.rotation_property)
+		drm_object_attach_property(&intel_plane->base.base,
+					   dev->mode_config.rotation_property,
+					   intel_plane->rotation);
 
+ out:
 	return ret;
 }
diff --git a/drivers/gpu/drm/i915/intel_tv.c b/drivers/gpu/drm/i915/intel_tv.c
index c14341ca3ef9..6f5f59b880f5 100644
--- a/drivers/gpu/drm/i915/intel_tv.c
+++ b/drivers/gpu/drm/i915/intel_tv.c
@@ -1182,18 +1182,17 @@ intel_tv_detect_type(struct intel_tv *intel_tv,
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct drm_device *dev = encoder->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	unsigned long irqflags;
 	u32 tv_ctl, save_tv_ctl;
 	u32 tv_dac, save_tv_dac;
 	int type;
 
 	/* Disable TV interrupts around load detect or we'll recurse */
 	if (connector->polled & DRM_CONNECTOR_POLL_HPD) {
-		spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+		spin_lock_irq(&dev_priv->irq_lock);
 		i915_disable_pipestat(dev_priv, 0,
 				      PIPE_HOTPLUG_INTERRUPT_STATUS |
 				      PIPE_HOTPLUG_TV_INTERRUPT_STATUS);
-		spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+		spin_unlock_irq(&dev_priv->irq_lock);
 	}
 
 	save_tv_dac = tv_dac = I915_READ(TV_DAC);
@@ -1266,11 +1265,11 @@ intel_tv_detect_type(struct intel_tv *intel_tv,
 
 	/* Restore interrupt config */
 	if (connector->polled & DRM_CONNECTOR_POLL_HPD) {
-		spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+		spin_lock_irq(&dev_priv->irq_lock);
 		i915_enable_pipestat(dev_priv, 0,
 				     PIPE_HOTPLUG_INTERRUPT_STATUS |
 				     PIPE_HOTPLUG_TV_INTERRUPT_STATUS);
-		spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+		spin_unlock_irq(&dev_priv->irq_lock);
 	}
 
 	return type;
diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c
index e81bc3bdc533..46de8d75b4bf 100644
--- a/drivers/gpu/drm/i915/intel_uncore.c
+++ b/drivers/gpu/drm/i915/intel_uncore.c
@@ -43,23 +43,17 @@
 static void
 assert_device_not_suspended(struct drm_i915_private *dev_priv)
 {
-	WARN(HAS_RUNTIME_PM(dev_priv->dev) && dev_priv->pm.suspended,
-	     "Device suspended\n");
+	WARN_ONCE(HAS_RUNTIME_PM(dev_priv->dev) && dev_priv->pm.suspended,
+		  "Device suspended\n");
 }
 
 static void __gen6_gt_wait_for_thread_c0(struct drm_i915_private *dev_priv)
 {
-	u32 gt_thread_status_mask;
-
-	if (IS_HASWELL(dev_priv->dev))
-		gt_thread_status_mask = GEN6_GT_THREAD_STATUS_CORE_MASK_HSW;
-	else
-		gt_thread_status_mask = GEN6_GT_THREAD_STATUS_CORE_MASK;
-
 	/* w/a for a sporadic read returning 0 by waiting for the GT
 	 * thread to wake up.
 	 */
-	if (wait_for_atomic_us((__raw_i915_read32(dev_priv, GEN6_GT_THREAD_STATUS_REG) & gt_thread_status_mask) == 0, 500))
+	if (wait_for_atomic_us((__raw_i915_read32(dev_priv, GEN6_GT_THREAD_STATUS_REG) &
+				GEN6_GT_THREAD_STATUS_CORE_MASK) == 0, 500))
 		DRM_ERROR("GT thread status wait timed out\n");
 }
 
@@ -101,7 +95,7 @@ static void __gen7_gt_force_wake_mt_get(struct drm_i915_private *dev_priv,
 {
 	u32 forcewake_ack;
 
-	if (IS_HASWELL(dev_priv->dev) || IS_GEN8(dev_priv->dev))
+	if (IS_HASWELL(dev_priv->dev) || IS_BROADWELL(dev_priv->dev))
 		forcewake_ack = FORCEWAKE_ACK_HSW;
 	else
 		forcewake_ack = FORCEWAKE_MT_ACK;
@@ -120,8 +114,7 @@ static void __gen7_gt_force_wake_mt_get(struct drm_i915_private *dev_priv,
 		DRM_ERROR("Timed out waiting for forcewake to ack request.\n");
 
 	/* WaRsForcewakeWaitTC0:ivb,hsw */
-	if (INTEL_INFO(dev_priv->dev)->gen < 8)
-		__gen6_gt_wait_for_thread_c0(dev_priv);
+	__gen6_gt_wait_for_thread_c0(dev_priv);
 }
 
 static void gen6_gt_check_fifodbg(struct drm_i915_private *dev_priv)
@@ -229,10 +222,6 @@ static void __vlv_force_wake_get(struct drm_i915_private *dev_priv,
 					FORCEWAKE_ACK_TIMEOUT_MS))
 			DRM_ERROR("Timed out: waiting for media to ack.\n");
 	}
-
-	/* WaRsForcewakeWaitTC0:vlv */
-	if (!IS_CHERRYVIEW(dev_priv->dev))
-		__gen6_gt_wait_for_thread_c0(dev_priv);
 }
 
 static void __vlv_force_wake_put(struct drm_i915_private *dev_priv,
@@ -299,6 +288,154 @@ static void vlv_force_wake_put(struct drm_i915_private *dev_priv, int fw_engine)
 	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 }
 
+static void __gen9_gt_force_wake_mt_reset(struct drm_i915_private *dev_priv)
+{
+	__raw_i915_write32(dev_priv, FORCEWAKE_RENDER_GEN9,
+			_MASKED_BIT_DISABLE(0xffff));
+
+	__raw_i915_write32(dev_priv, FORCEWAKE_MEDIA_GEN9,
+			_MASKED_BIT_DISABLE(0xffff));
+
+	__raw_i915_write32(dev_priv, FORCEWAKE_BLITTER_GEN9,
+			_MASKED_BIT_DISABLE(0xffff));
+}
+
+static void
+__gen9_force_wake_get(struct drm_i915_private *dev_priv, int fw_engine)
+{
+	/* Check for Render Engine */
+	if (FORCEWAKE_RENDER & fw_engine) {
+		if (wait_for_atomic((__raw_i915_read32(dev_priv,
+						FORCEWAKE_ACK_RENDER_GEN9) &
+						FORCEWAKE_KERNEL) == 0,
+					FORCEWAKE_ACK_TIMEOUT_MS))
+			DRM_ERROR("Timed out: Render forcewake old ack to clear.\n");
+
+		__raw_i915_write32(dev_priv, FORCEWAKE_RENDER_GEN9,
+				   _MASKED_BIT_ENABLE(FORCEWAKE_KERNEL));
+
+		if (wait_for_atomic((__raw_i915_read32(dev_priv,
+						FORCEWAKE_ACK_RENDER_GEN9) &
+						FORCEWAKE_KERNEL),
+					FORCEWAKE_ACK_TIMEOUT_MS))
+			DRM_ERROR("Timed out: waiting for Render to ack.\n");
+	}
+
+	/* Check for Media Engine */
+	if (FORCEWAKE_MEDIA & fw_engine) {
+		if (wait_for_atomic((__raw_i915_read32(dev_priv,
+						FORCEWAKE_ACK_MEDIA_GEN9) &
+						FORCEWAKE_KERNEL) == 0,
+					FORCEWAKE_ACK_TIMEOUT_MS))
+			DRM_ERROR("Timed out: Media forcewake old ack to clear.\n");
+
+		__raw_i915_write32(dev_priv, FORCEWAKE_MEDIA_GEN9,
+				   _MASKED_BIT_ENABLE(FORCEWAKE_KERNEL));
+
+		if (wait_for_atomic((__raw_i915_read32(dev_priv,
+						FORCEWAKE_ACK_MEDIA_GEN9) &
+						FORCEWAKE_KERNEL),
+					FORCEWAKE_ACK_TIMEOUT_MS))
+			DRM_ERROR("Timed out: waiting for Media to ack.\n");
+	}
+
+	/* Check for Blitter Engine */
+	if (FORCEWAKE_BLITTER & fw_engine) {
+		if (wait_for_atomic((__raw_i915_read32(dev_priv,
+						FORCEWAKE_ACK_BLITTER_GEN9) &
+						FORCEWAKE_KERNEL) == 0,
+					FORCEWAKE_ACK_TIMEOUT_MS))
+			DRM_ERROR("Timed out: Blitter forcewake old ack to clear.\n");
+
+		__raw_i915_write32(dev_priv, FORCEWAKE_BLITTER_GEN9,
+				   _MASKED_BIT_ENABLE(FORCEWAKE_KERNEL));
+
+		if (wait_for_atomic((__raw_i915_read32(dev_priv,
+						FORCEWAKE_ACK_BLITTER_GEN9) &
+						FORCEWAKE_KERNEL),
+					FORCEWAKE_ACK_TIMEOUT_MS))
+			DRM_ERROR("Timed out: waiting for Blitter to ack.\n");
+	}
+}
+
+static void
+__gen9_force_wake_put(struct drm_i915_private *dev_priv, int fw_engine)
+{
+	/* Check for Render Engine */
+	if (FORCEWAKE_RENDER & fw_engine)
+		__raw_i915_write32(dev_priv, FORCEWAKE_RENDER_GEN9,
+				_MASKED_BIT_DISABLE(FORCEWAKE_KERNEL));
+
+	/* Check for Media Engine */
+	if (FORCEWAKE_MEDIA & fw_engine)
+		__raw_i915_write32(dev_priv, FORCEWAKE_MEDIA_GEN9,
+				_MASKED_BIT_DISABLE(FORCEWAKE_KERNEL));
+
+	/* Check for Blitter Engine */
+	if (FORCEWAKE_BLITTER & fw_engine)
+		__raw_i915_write32(dev_priv, FORCEWAKE_BLITTER_GEN9,
+				_MASKED_BIT_DISABLE(FORCEWAKE_KERNEL));
+}
+
+static void
+gen9_force_wake_get(struct drm_i915_private *dev_priv, int fw_engine)
+{
+	unsigned long irqflags;
+
+	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
+
+	if (FORCEWAKE_RENDER & fw_engine) {
+		if (dev_priv->uncore.fw_rendercount++ == 0)
+			dev_priv->uncore.funcs.force_wake_get(dev_priv,
+							FORCEWAKE_RENDER);
+	}
+
+	if (FORCEWAKE_MEDIA & fw_engine) {
+		if (dev_priv->uncore.fw_mediacount++ == 0)
+			dev_priv->uncore.funcs.force_wake_get(dev_priv,
+							FORCEWAKE_MEDIA);
+	}
+
+	if (FORCEWAKE_BLITTER & fw_engine) {
+		if (dev_priv->uncore.fw_blittercount++ == 0)
+			dev_priv->uncore.funcs.force_wake_get(dev_priv,
+							FORCEWAKE_BLITTER);
+	}
+
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
+}
+
+static void
+gen9_force_wake_put(struct drm_i915_private *dev_priv, int fw_engine)
+{
+	unsigned long irqflags;
+
+	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
+
+	if (FORCEWAKE_RENDER & fw_engine) {
+		WARN_ON(dev_priv->uncore.fw_rendercount == 0);
+		if (--dev_priv->uncore.fw_rendercount == 0)
+			dev_priv->uncore.funcs.force_wake_put(dev_priv,
+							FORCEWAKE_RENDER);
+	}
+
+	if (FORCEWAKE_MEDIA & fw_engine) {
+		WARN_ON(dev_priv->uncore.fw_mediacount == 0);
+		if (--dev_priv->uncore.fw_mediacount == 0)
+			dev_priv->uncore.funcs.force_wake_put(dev_priv,
+							FORCEWAKE_MEDIA);
+	}
+
+	if (FORCEWAKE_BLITTER & fw_engine) {
+		WARN_ON(dev_priv->uncore.fw_blittercount == 0);
+		if (--dev_priv->uncore.fw_blittercount == 0)
+			dev_priv->uncore.funcs.force_wake_put(dev_priv,
+							FORCEWAKE_BLITTER);
+	}
+
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
+}
+
 static void gen6_force_wake_timer(unsigned long arg)
 {
 	struct drm_i915_private *dev_priv = (void *)arg;
@@ -334,9 +471,12 @@ void intel_uncore_forcewake_reset(struct drm_device *dev, bool restore)
 	else if (IS_GEN6(dev) || IS_GEN7(dev))
 		__gen6_gt_force_wake_reset(dev_priv);
 
-	if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev) || IS_GEN8(dev))
+	if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev) || IS_BROADWELL(dev))
 		__gen7_gt_force_wake_mt_reset(dev_priv);
 
+	if (IS_GEN9(dev))
+		__gen9_gt_force_wake_mt_reset(dev_priv);
+
 	if (restore) { /* If reset with a user forcewake, try to restore */
 		unsigned fw = 0;
 
@@ -346,6 +486,15 @@ void intel_uncore_forcewake_reset(struct drm_device *dev, bool restore)
 
 			if (dev_priv->uncore.fw_mediacount)
 				fw |= FORCEWAKE_MEDIA;
+		} else if (IS_GEN9(dev)) {
+			if (dev_priv->uncore.fw_rendercount)
+				fw |= FORCEWAKE_RENDER;
+
+			if (dev_priv->uncore.fw_mediacount)
+				fw |= FORCEWAKE_MEDIA;
+
+			if (dev_priv->uncore.fw_blittercount)
+				fw |= FORCEWAKE_BLITTER;
 		} else {
 			if (dev_priv->uncore.forcewake_count)
 				fw = FORCEWAKE_ALL;
@@ -363,7 +512,8 @@ void intel_uncore_forcewake_reset(struct drm_device *dev, bool restore)
 	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 }
 
-void intel_uncore_early_sanitize(struct drm_device *dev, bool restore_forcewake)
+static void __intel_uncore_early_sanitize(struct drm_device *dev,
+					  bool restore_forcewake)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
@@ -389,6 +539,12 @@ void intel_uncore_early_sanitize(struct drm_device *dev, bool restore_forcewake)
 	intel_uncore_forcewake_reset(dev, restore_forcewake);
 }
 
+void intel_uncore_early_sanitize(struct drm_device *dev, bool restore_forcewake)
+{
+	__intel_uncore_early_sanitize(dev, restore_forcewake);
+	i915_check_and_clear_faults(dev);
+}
+
 void intel_uncore_sanitize(struct drm_device *dev)
 {
 	/* BIOS often leaves RC6 enabled, but disable it for hw init */
@@ -410,6 +566,10 @@ void gen6_gt_force_wake_get(struct drm_i915_private *dev_priv, int fw_engine)
 
 	intel_runtime_pm_get(dev_priv);
 
+	/* Redirect to Gen9 specific routine */
+	if (IS_GEN9(dev_priv->dev))
+		return gen9_force_wake_get(dev_priv, fw_engine);
+
 	/* Redirect to VLV specific routine */
 	if (IS_VALLEYVIEW(dev_priv->dev))
 		return vlv_force_wake_get(dev_priv, fw_engine);
@@ -431,6 +591,12 @@ void gen6_gt_force_wake_put(struct drm_i915_private *dev_priv, int fw_engine)
 	if (!dev_priv->uncore.funcs.force_wake_put)
 		return;
 
+	/* Redirect to Gen9 specific routine */
+	if (IS_GEN9(dev_priv->dev)) {
+		gen9_force_wake_put(dev_priv, fw_engine);
+		goto out;
+	}
+
 	/* Redirect to VLV specific routine */
 	if (IS_VALLEYVIEW(dev_priv->dev)) {
 		vlv_force_wake_put(dev_priv, fw_engine);
@@ -504,6 +670,38 @@ void assert_force_wake_inactive(struct drm_i915_private *dev_priv)
 	 REG_RANGE((reg), 0x14000, 0x14400) || \
 	 REG_RANGE((reg), 0x22000, 0x24000))
 
+#define FORCEWAKE_GEN9_UNCORE_RANGE_OFFSET(reg) \
+	REG_RANGE((reg), 0xB00,  0x2000)
+
+#define FORCEWAKE_GEN9_RENDER_RANGE_OFFSET(reg) \
+	(REG_RANGE((reg), 0x2000, 0x2700) || \
+	 REG_RANGE((reg), 0x3000, 0x4000) || \
+	 REG_RANGE((reg), 0x5200, 0x8000) || \
+	 REG_RANGE((reg), 0x8140, 0x8160) || \
+	 REG_RANGE((reg), 0x8300, 0x8500) || \
+	 REG_RANGE((reg), 0x8C00, 0x8D00) || \
+	 REG_RANGE((reg), 0xB000, 0xB480) || \
+	 REG_RANGE((reg), 0xE000, 0xE900) || \
+	 REG_RANGE((reg), 0x24400, 0x24800))
+
+#define FORCEWAKE_GEN9_MEDIA_RANGE_OFFSET(reg) \
+	(REG_RANGE((reg), 0x8130, 0x8140) || \
+	 REG_RANGE((reg), 0x8800, 0x8A00) || \
+	 REG_RANGE((reg), 0xD000, 0xD800) || \
+	 REG_RANGE((reg), 0x12000, 0x14000) || \
+	 REG_RANGE((reg), 0x1A000, 0x1EA00) || \
+	 REG_RANGE((reg), 0x30000, 0x40000))
+
+#define FORCEWAKE_GEN9_COMMON_RANGE_OFFSET(reg) \
+	REG_RANGE((reg), 0x9400, 0x9800)
+
+#define FORCEWAKE_GEN9_BLITTER_RANGE_OFFSET(reg) \
+	((reg) < 0x40000 &&\
+	 !FORCEWAKE_GEN9_UNCORE_RANGE_OFFSET(reg) && \
+	 !FORCEWAKE_GEN9_RENDER_RANGE_OFFSET(reg) && \
+	 !FORCEWAKE_GEN9_MEDIA_RANGE_OFFSET(reg) && \
+	 !FORCEWAKE_GEN9_COMMON_RANGE_OFFSET(reg))
+
 static void
 ilk_dummy_write(struct drm_i915_private *dev_priv)
 {
@@ -634,6 +832,45 @@ chv_read##x(struct drm_i915_private *dev_priv, off_t reg, bool trace) { \
 	REG_READ_FOOTER; \
 }
 
+#define SKL_NEEDS_FORCE_WAKE(dev_priv, reg)	\
+	 ((reg) < 0x40000 && !FORCEWAKE_GEN9_UNCORE_RANGE_OFFSET(reg))
+
+#define __gen9_read(x) \
+static u##x \
+gen9_read##x(struct drm_i915_private *dev_priv, off_t reg, bool trace) { \
+	REG_READ_HEADER(x); \
+	if (!SKL_NEEDS_FORCE_WAKE((dev_priv), (reg))) { \
+		val = __raw_i915_read##x(dev_priv, reg); \
+	} else { \
+		unsigned fwengine = 0; \
+		if (FORCEWAKE_GEN9_RENDER_RANGE_OFFSET(reg)) { \
+			if (dev_priv->uncore.fw_rendercount == 0) \
+				fwengine = FORCEWAKE_RENDER; \
+		} else if (FORCEWAKE_GEN9_MEDIA_RANGE_OFFSET(reg)) { \
+			if (dev_priv->uncore.fw_mediacount == 0) \
+				fwengine = FORCEWAKE_MEDIA; \
+		} else if (FORCEWAKE_GEN9_COMMON_RANGE_OFFSET(reg)) { \
+			if (dev_priv->uncore.fw_rendercount == 0) \
+				fwengine |= FORCEWAKE_RENDER; \
+			if (dev_priv->uncore.fw_mediacount == 0) \
+				fwengine |= FORCEWAKE_MEDIA; \
+		} else { \
+			if (dev_priv->uncore.fw_blittercount == 0) \
+				fwengine = FORCEWAKE_BLITTER; \
+		} \
+		if (fwengine) \
+			dev_priv->uncore.funcs.force_wake_get(dev_priv, fwengine); \
+		val = __raw_i915_read##x(dev_priv, reg); \
+		if (fwengine) \
+			dev_priv->uncore.funcs.force_wake_put(dev_priv, fwengine); \
+	} \
+	REG_READ_FOOTER; \
+}
+
+__gen9_read(8)
+__gen9_read(16)
+__gen9_read(32)
+__gen9_read(64)
 __chv_read(8)
 __chv_read(16)
 __chv_read(32)
@@ -655,6 +892,7 @@ __gen4_read(16)
 __gen4_read(32)
 __gen4_read(64)
 
+#undef __gen9_read
 #undef __chv_read
 #undef __vlv_read
 #undef __gen6_read
@@ -792,6 +1030,69 @@ chv_write##x(struct drm_i915_private *dev_priv, off_t reg, u##x val, bool trace)
 	REG_WRITE_FOOTER; \
 }
 
+static const u32 gen9_shadowed_regs[] = {
+	RING_TAIL(RENDER_RING_BASE),
+	RING_TAIL(GEN6_BSD_RING_BASE),
+	RING_TAIL(VEBOX_RING_BASE),
+	RING_TAIL(BLT_RING_BASE),
+	FORCEWAKE_BLITTER_GEN9,
+	FORCEWAKE_RENDER_GEN9,
+	FORCEWAKE_MEDIA_GEN9,
+	GEN6_RPNSWREQ,
+	GEN6_RC_VIDEO_FREQ,
+	/* TODO: Other registers are not yet used */
+};
+
+static bool is_gen9_shadowed(struct drm_i915_private *dev_priv, u32 reg)
+{
+	int i;
+	for (i = 0; i < ARRAY_SIZE(gen9_shadowed_regs); i++)
+		if (reg == gen9_shadowed_regs[i])
+			return true;
+
+	return false;
+}
+
+#define __gen9_write(x) \
+static void \
+gen9_write##x(struct drm_i915_private *dev_priv, off_t reg, u##x val, \
+		bool trace) { \
+	REG_WRITE_HEADER; \
+	if (!SKL_NEEDS_FORCE_WAKE((dev_priv), (reg)) || \
+			is_gen9_shadowed(dev_priv, reg)) { \
+		__raw_i915_write##x(dev_priv, reg, val); \
+	} else { \
+		unsigned fwengine = 0; \
+		if (FORCEWAKE_GEN9_RENDER_RANGE_OFFSET(reg)) { \
+			if (dev_priv->uncore.fw_rendercount == 0) \
+				fwengine = FORCEWAKE_RENDER; \
+		} else if (FORCEWAKE_GEN9_MEDIA_RANGE_OFFSET(reg)) { \
+			if (dev_priv->uncore.fw_mediacount == 0) \
+				fwengine = FORCEWAKE_MEDIA; \
+		} else if (FORCEWAKE_GEN9_COMMON_RANGE_OFFSET(reg)) { \
+			if (dev_priv->uncore.fw_rendercount == 0) \
+				fwengine |= FORCEWAKE_RENDER; \
+			if (dev_priv->uncore.fw_mediacount == 0) \
+				fwengine |= FORCEWAKE_MEDIA; \
+		} else { \
+			if (dev_priv->uncore.fw_blittercount == 0) \
+				fwengine = FORCEWAKE_BLITTER; \
+		} \
+		if (fwengine) \
+			dev_priv->uncore.funcs.force_wake_get(dev_priv, \
+					fwengine); \
+		__raw_i915_write##x(dev_priv, reg, val); \
+		if (fwengine) \
+			dev_priv->uncore.funcs.force_wake_put(dev_priv, \
+					fwengine); \
+	} \
+	REG_WRITE_FOOTER; \
+}
+
+__gen9_write(8)
+__gen9_write(16)
+__gen9_write(32)
+__gen9_write(64)
 __chv_write(8)
 __chv_write(16)
 __chv_write(32)
@@ -817,6 +1118,7 @@ __gen4_write(16)
 __gen4_write(32)
 __gen4_write(64)
 
+#undef __gen9_write
 #undef __chv_write
 #undef __gen8_write
 #undef __hsw_write
@@ -826,6 +1128,22 @@ __gen4_write(64)
 #undef REG_WRITE_FOOTER
 #undef REG_WRITE_HEADER
 
+#define ASSIGN_WRITE_MMIO_VFUNCS(x) \
+do { \
+	dev_priv->uncore.funcs.mmio_writeb = x##_write8; \
+	dev_priv->uncore.funcs.mmio_writew = x##_write16; \
+	dev_priv->uncore.funcs.mmio_writel = x##_write32; \
+	dev_priv->uncore.funcs.mmio_writeq = x##_write64; \
+} while (0)
+
+#define ASSIGN_READ_MMIO_VFUNCS(x) \
+do { \
+	dev_priv->uncore.funcs.mmio_readb = x##_read8; \
+	dev_priv->uncore.funcs.mmio_readw = x##_read16; \
+	dev_priv->uncore.funcs.mmio_readl = x##_read32; \
+	dev_priv->uncore.funcs.mmio_readq = x##_read64; \
+} while (0)
+
 void intel_uncore_init(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -833,12 +1151,15 @@ void intel_uncore_init(struct drm_device *dev)
 	setup_timer(&dev_priv->uncore.force_wake_timer,
 		    gen6_force_wake_timer, (unsigned long)dev_priv);
 
-	intel_uncore_early_sanitize(dev, false);
+	__intel_uncore_early_sanitize(dev, false);
 
-	if (IS_VALLEYVIEW(dev)) {
+	if (IS_GEN9(dev)) {
+		dev_priv->uncore.funcs.force_wake_get = __gen9_force_wake_get;
+		dev_priv->uncore.funcs.force_wake_put = __gen9_force_wake_put;
+	} else if (IS_VALLEYVIEW(dev)) {
 		dev_priv->uncore.funcs.force_wake_get = __vlv_force_wake_get;
 		dev_priv->uncore.funcs.force_wake_put = __vlv_force_wake_put;
-	} else if (IS_HASWELL(dev) || IS_GEN8(dev)) {
+	} else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
 		dev_priv->uncore.funcs.force_wake_get = __gen7_gt_force_wake_mt_get;
 		dev_priv->uncore.funcs.force_wake_put = __gen7_gt_force_wake_mt_put;
 	} else if (IS_IVYBRIDGE(dev)) {
@@ -881,77 +1202,52 @@ void intel_uncore_init(struct drm_device *dev)
 
 	switch (INTEL_INFO(dev)->gen) {
 	default:
+		WARN_ON(1);
+		return;
+	case 9:
+		ASSIGN_WRITE_MMIO_VFUNCS(gen9);
+		ASSIGN_READ_MMIO_VFUNCS(gen9);
+		break;
+	case 8:
 		if (IS_CHERRYVIEW(dev)) {
-			dev_priv->uncore.funcs.mmio_writeb  = chv_write8;
-			dev_priv->uncore.funcs.mmio_writew  = chv_write16;
-			dev_priv->uncore.funcs.mmio_writel  = chv_write32;
-			dev_priv->uncore.funcs.mmio_writeq  = chv_write64;
-			dev_priv->uncore.funcs.mmio_readb  = chv_read8;
-			dev_priv->uncore.funcs.mmio_readw  = chv_read16;
-			dev_priv->uncore.funcs.mmio_readl  = chv_read32;
-			dev_priv->uncore.funcs.mmio_readq  = chv_read64;
+			ASSIGN_WRITE_MMIO_VFUNCS(chv);
+			ASSIGN_READ_MMIO_VFUNCS(chv);
 
 		} else {
-			dev_priv->uncore.funcs.mmio_writeb  = gen8_write8;
-			dev_priv->uncore.funcs.mmio_writew  = gen8_write16;
-			dev_priv->uncore.funcs.mmio_writel  = gen8_write32;
-			dev_priv->uncore.funcs.mmio_writeq  = gen8_write64;
-			dev_priv->uncore.funcs.mmio_readb  = gen6_read8;
-			dev_priv->uncore.funcs.mmio_readw  = gen6_read16;
-			dev_priv->uncore.funcs.mmio_readl  = gen6_read32;
-			dev_priv->uncore.funcs.mmio_readq  = gen6_read64;
+			ASSIGN_WRITE_MMIO_VFUNCS(gen8);
+			ASSIGN_READ_MMIO_VFUNCS(gen6);
 		}
 		break;
 	case 7:
 	case 6:
 		if (IS_HASWELL(dev)) {
-			dev_priv->uncore.funcs.mmio_writeb  = hsw_write8;
-			dev_priv->uncore.funcs.mmio_writew  = hsw_write16;
-			dev_priv->uncore.funcs.mmio_writel  = hsw_write32;
-			dev_priv->uncore.funcs.mmio_writeq  = hsw_write64;
+			ASSIGN_WRITE_MMIO_VFUNCS(hsw);
 		} else {
-			dev_priv->uncore.funcs.mmio_writeb  = gen6_write8;
-			dev_priv->uncore.funcs.mmio_writew  = gen6_write16;
-			dev_priv->uncore.funcs.mmio_writel  = gen6_write32;
-			dev_priv->uncore.funcs.mmio_writeq  = gen6_write64;
+			ASSIGN_WRITE_MMIO_VFUNCS(gen6);
 		}
 
 		if (IS_VALLEYVIEW(dev)) {
-			dev_priv->uncore.funcs.mmio_readb  = vlv_read8;
-			dev_priv->uncore.funcs.mmio_readw  = vlv_read16;
-			dev_priv->uncore.funcs.mmio_readl  = vlv_read32;
-			dev_priv->uncore.funcs.mmio_readq  = vlv_read64;
+			ASSIGN_READ_MMIO_VFUNCS(vlv);
 		} else {
-			dev_priv->uncore.funcs.mmio_readb  = gen6_read8;
-			dev_priv->uncore.funcs.mmio_readw  = gen6_read16;
-			dev_priv->uncore.funcs.mmio_readl  = gen6_read32;
-			dev_priv->uncore.funcs.mmio_readq  = gen6_read64;
+			ASSIGN_READ_MMIO_VFUNCS(gen6);
 		}
 		break;
 	case 5:
-		dev_priv->uncore.funcs.mmio_writeb  = gen5_write8;
-		dev_priv->uncore.funcs.mmio_writew  = gen5_write16;
-		dev_priv->uncore.funcs.mmio_writel  = gen5_write32;
-		dev_priv->uncore.funcs.mmio_writeq  = gen5_write64;
-		dev_priv->uncore.funcs.mmio_readb  = gen5_read8;
-		dev_priv->uncore.funcs.mmio_readw  = gen5_read16;
-		dev_priv->uncore.funcs.mmio_readl  = gen5_read32;
-		dev_priv->uncore.funcs.mmio_readq  = gen5_read64;
+		ASSIGN_WRITE_MMIO_VFUNCS(gen5);
+		ASSIGN_READ_MMIO_VFUNCS(gen5);
 		break;
 	case 4:
 	case 3:
 	case 2:
-		dev_priv->uncore.funcs.mmio_writeb  = gen4_write8;
-		dev_priv->uncore.funcs.mmio_writew  = gen4_write16;
-		dev_priv->uncore.funcs.mmio_writel  = gen4_write32;
-		dev_priv->uncore.funcs.mmio_writeq  = gen4_write64;
-		dev_priv->uncore.funcs.mmio_readb  = gen4_read8;
-		dev_priv->uncore.funcs.mmio_readw  = gen4_read16;
-		dev_priv->uncore.funcs.mmio_readl  = gen4_read32;
-		dev_priv->uncore.funcs.mmio_readq  = gen4_read64;
+		ASSIGN_WRITE_MMIO_VFUNCS(gen4);
+		ASSIGN_READ_MMIO_VFUNCS(gen4);
 		break;
 	}
+
+	i915_check_and_clear_faults(dev);
 }
+#undef ASSIGN_WRITE_MMIO_VFUNCS
+#undef ASSIGN_READ_MMIO_VFUNCS
 
 void intel_uncore_fini(struct drm_device *dev)
 {
@@ -968,7 +1264,7 @@ static const struct register_whitelist {
 	/* supported gens, 0x10 for 4, 0x30 for 4 and 5, etc. */
 	uint32_t gen_bitmask;
 } whitelist[] = {
-	{ RING_TIMESTAMP(RENDER_RING_BASE), 8, GEN_RANGE(4, 8) },
+	{ RING_TIMESTAMP(RENDER_RING_BASE), 8, GEN_RANGE(4, 9) },
 };
 
 int i915_reg_read_ioctl(struct drm_device *dev,
@@ -1053,41 +1349,34 @@ int i915_get_reset_stats_ioctl(struct drm_device *dev,
 	return 0;
 }
 
-static int i965_reset_complete(struct drm_device *dev)
+static int i915_reset_complete(struct drm_device *dev)
 {
 	u8 gdrst;
-	pci_read_config_byte(dev->pdev, I965_GDRST, &gdrst);
-	return (gdrst & GRDOM_RESET_ENABLE) == 0;
+	pci_read_config_byte(dev->pdev, I915_GDRST, &gdrst);
+	return (gdrst & GRDOM_RESET_STATUS) == 0;
 }
 
-static int i965_do_reset(struct drm_device *dev)
+static int i915_do_reset(struct drm_device *dev)
 {
-	int ret;
-
-	/* FIXME: i965g/gm need a display save/restore for gpu reset. */
-	return -ENODEV;
+	/* assert reset for at least 20 usec */
+	pci_write_config_byte(dev->pdev, I915_GDRST, GRDOM_RESET_ENABLE);
+	udelay(20);
+	pci_write_config_byte(dev->pdev, I915_GDRST, 0);
 
-	/*
-	 * Set the domains we want to reset (GRDOM/bits 2 and 3) as
-	 * well as the reset bit (GR/bit 0).  Setting the GR bit
-	 * triggers the reset; when done, the hardware will clear it.
-	 */
-	pci_write_config_byte(dev->pdev, I965_GDRST,
-			      GRDOM_RENDER | GRDOM_RESET_ENABLE);
-	ret =  wait_for(i965_reset_complete(dev), 500);
-	if (ret)
-		return ret;
-
-	pci_write_config_byte(dev->pdev, I965_GDRST,
-			      GRDOM_MEDIA | GRDOM_RESET_ENABLE);
-
-	ret =  wait_for(i965_reset_complete(dev), 500);
-	if (ret)
-		return ret;
+	return wait_for(i915_reset_complete(dev), 500);
+}
 
-	pci_write_config_byte(dev->pdev, I965_GDRST, 0);
+static int g4x_reset_complete(struct drm_device *dev)
+{
+	u8 gdrst;
+	pci_read_config_byte(dev->pdev, I915_GDRST, &gdrst);
+	return (gdrst & GRDOM_RESET_ENABLE) == 0;
+}
 
-	return 0;
+static int g33_do_reset(struct drm_device *dev)
+{
+	pci_write_config_byte(dev->pdev, I915_GDRST, GRDOM_RESET_ENABLE);
+	return wait_for(g4x_reset_complete(dev), 500);
 }
 
 static int g4x_do_reset(struct drm_device *dev)
@@ -1095,9 +1384,9 @@ static int g4x_do_reset(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
-	pci_write_config_byte(dev->pdev, I965_GDRST,
+	pci_write_config_byte(dev->pdev, I915_GDRST,
 			      GRDOM_RENDER | GRDOM_RESET_ENABLE);
-	ret =  wait_for(i965_reset_complete(dev), 500);
+	ret =  wait_for(g4x_reset_complete(dev), 500);
 	if (ret)
 		return ret;
 
@@ -1105,9 +1394,9 @@ static int g4x_do_reset(struct drm_device *dev)
 	I915_WRITE(VDECCLK_GATE_D, I915_READ(VDECCLK_GATE_D) | VCP_UNIT_CLOCK_GATE_DISABLE);
 	POSTING_READ(VDECCLK_GATE_D);
 
-	pci_write_config_byte(dev->pdev, I965_GDRST,
+	pci_write_config_byte(dev->pdev, I915_GDRST,
 			      GRDOM_MEDIA | GRDOM_RESET_ENABLE);
-	ret =  wait_for(i965_reset_complete(dev), 500);
+	ret =  wait_for(g4x_reset_complete(dev), 500);
 	if (ret)
 		return ret;
 
@@ -1115,7 +1404,7 @@ static int g4x_do_reset(struct drm_device *dev)
 	I915_WRITE(VDECCLK_GATE_D, I915_READ(VDECCLK_GATE_D) & ~VCP_UNIT_CLOCK_GATE_DISABLE);
 	POSTING_READ(VDECCLK_GATE_D);
 
-	pci_write_config_byte(dev->pdev, I965_GDRST, 0);
+	pci_write_config_byte(dev->pdev, I915_GDRST, 0);
 
 	return 0;
 }
@@ -1173,8 +1462,10 @@ int intel_gpu_reset(struct drm_device *dev)
 		return ironlake_do_reset(dev);
 	else if (IS_G4X(dev))
 		return g4x_do_reset(dev);
-	else if (IS_GEN4(dev))
-		return i965_do_reset(dev);
+	else if (IS_G33(dev))
+		return g33_do_reset(dev);
+	else if (INTEL_INFO(dev)->gen >= 3)
+		return i915_do_reset(dev);
 	else
 		return -ENODEV;
 }
diff --git a/drivers/staging/imx-drm/Kconfig b/drivers/gpu/drm/imx/Kconfig
index 82fb758a29bc..ab31848e92cf 100644
--- a/drivers/staging/imx-drm/Kconfig
+++ b/drivers/gpu/drm/imx/Kconfig
@@ -6,6 +6,7 @@ config DRM_IMX
 	select DRM_GEM_CMA_HELPER
 	select DRM_KMS_CMA_HELPER
 	depends on DRM && (ARCH_MXC || ARCH_MULTIPLATFORM)
+	depends on IMX_IPUV3_CORE
 	help
 	  enable i.MX graphics support
 
@@ -40,11 +41,11 @@ config DRM_IMX_LDB
 	  found on i.MX53 and i.MX6 processors.
 
 config DRM_IMX_IPUV3
-	tristate "DRM Support for i.MX IPUv3"
+	tristate
 	depends on DRM_IMX
 	depends on IMX_IPUV3_CORE
-	help
-	  Choose this if you have a i.MX5 or i.MX6 processor.
+	default y if DRM_IMX=y
+	default m if DRM_IMX=m
 
 config DRM_IMX_HDMI
 	tristate "Freescale i.MX DRM HDMI"
diff --git a/drivers/staging/imx-drm/Makefile b/drivers/gpu/drm/imx/Makefile
index 582c438d8cbd..582c438d8cbd 100644
--- a/drivers/staging/imx-drm/Makefile
+++ b/drivers/gpu/drm/imx/Makefile
diff --git a/drivers/staging/imx-drm/imx-drm-core.c b/drivers/gpu/drm/imx/imx-drm-core.c
index 6b22106534d8..b250130debc8 100644
--- a/drivers/staging/imx-drm/imx-drm-core.c
+++ b/drivers/gpu/drm/imx/imx-drm-core.c
@@ -24,13 +24,12 @@
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_gem_cma_helper.h>
 #include <drm/drm_fb_cma_helper.h>
+#include <drm/drm_plane_helper.h>
 
 #include "imx-drm.h"
 
 #define MAX_CRTC	4
 
-struct imx_drm_crtc;
-
 struct imx_drm_component {
 	struct device_node *of_node;
 	struct list_head list;
@@ -87,6 +86,8 @@ static int imx_drm_driver_unload(struct drm_device *drm)
 	drm_vblank_cleanup(drm);
 	drm_mode_config_cleanup(drm);
 
+	platform_set_drvdata(drm->platformdev, NULL);
+
 	return 0;
 }
 
@@ -427,6 +428,7 @@ static uint32_t imx_drm_find_crtc_mask(struct imx_drm_device *imxdrm,
 
 	for (i = 0; i < MAX_CRTC; i++) {
 		struct imx_drm_crtc *imx_drm_crtc = imxdrm->crtc[i];
+
 		if (imx_drm_crtc && imx_drm_crtc->port == port)
 			return drm_crtc_mask(imx_drm_crtc->crtc);
 	}
@@ -438,6 +440,7 @@ static struct device_node *imx_drm_of_get_next_endpoint(
 		const struct device_node *parent, struct device_node *prev)
 {
 	struct device_node *node = of_graph_get_next_endpoint(parent, prev);
+
 	of_node_put(prev);
 	return node;
 }
@@ -471,8 +474,7 @@ int imx_drm_encoder_parse_of(struct drm_device *drm,
 		crtc_mask |= mask;
 	}
 
-	if (ep)
-		of_node_put(ep);
+	of_node_put(ep);
 	if (i == 0)
 		return -ENOENT;
 
@@ -528,6 +530,7 @@ static struct drm_driver imx_drm_driver = {
 	.unload			= imx_drm_driver_unload,
 	.lastclose		= imx_drm_driver_lastclose,
 	.preclose		= imx_drm_driver_preclose,
+	.set_busid		= drm_platform_set_busid,
 	.gem_free_object	= drm_gem_cma_free_object,
 	.gem_vm_ops		= &drm_gem_cma_vm_ops,
 	.dumb_create		= drm_gem_cma_dumb_create,
@@ -628,7 +631,8 @@ static int imx_drm_platform_probe(struct platform_device *pdev)
 				continue;
 			}
 
-			component_match_add(&pdev->dev, &match, compare_of, remote);
+			component_match_add(&pdev->dev, &match, compare_of,
+					    remote);
 			of_node_put(remote);
 		}
 		of_node_put(port);
@@ -647,6 +651,36 @@ static int imx_drm_platform_remove(struct platform_device *pdev)
 	return 0;
 }
 
+#ifdef CONFIG_PM_SLEEP
+static int imx_drm_suspend(struct device *dev)
+{
+	struct drm_device *drm_dev = dev_get_drvdata(dev);
+
+	/* The drm_dev is NULL before .load hook is called */
+	if (drm_dev == NULL)
+		return 0;
+
+	drm_kms_helper_poll_disable(drm_dev);
+
+	return 0;
+}
+
+static int imx_drm_resume(struct device *dev)
+{
+	struct drm_device *drm_dev = dev_get_drvdata(dev);
+
+	if (drm_dev == NULL)
+		return 0;
+
+	drm_helper_resume_force_mode(drm_dev);
+	drm_kms_helper_poll_enable(drm_dev);
+
+	return 0;
+}
+#endif
+
+static SIMPLE_DEV_PM_OPS(imx_drm_pm_ops, imx_drm_suspend, imx_drm_resume);
+
 static const struct of_device_id imx_drm_dt_ids[] = {
 	{ .compatible = "fsl,imx-display-subsystem", },
 	{ /* sentinel */ },
@@ -657,8 +691,8 @@ static struct platform_driver imx_drm_pdrv = {
 	.probe		= imx_drm_platform_probe,
 	.remove		= imx_drm_platform_remove,
 	.driver		= {
-		.owner	= THIS_MODULE,
 		.name	= "imx-drm",
+		.pm	= &imx_drm_pm_ops,
 		.of_match_table = imx_drm_dt_ids,
 	},
 };
diff --git a/drivers/staging/imx-drm/imx-drm.h b/drivers/gpu/drm/imx/imx-drm.h
index 7453ae00c412..7453ae00c412 100644
--- a/drivers/staging/imx-drm/imx-drm.h
+++ b/drivers/gpu/drm/imx/imx-drm.h
diff --git a/drivers/staging/imx-drm/imx-hdmi.c b/drivers/gpu/drm/imx/imx-hdmi.c
index 18c9ccd460b7..ddc53e039530 100644
--- a/drivers/staging/imx-drm/imx-hdmi.c
+++ b/drivers/gpu/drm/imx/imx-hdmi.c
@@ -323,8 +323,7 @@ static unsigned int hdmi_compute_cts(unsigned int freq, unsigned long pixel_clk,
 	}
 	if (ratio == 100)
 		return cts;
-	else
-		return (cts * ratio) / 100;
+	return (cts * ratio) / 100;
 }
 
 static void hdmi_set_clk_regenerator(struct imx_hdmi *hdmi,
@@ -1755,7 +1754,6 @@ static struct platform_driver imx_hdmi_driver = {
 	.remove = imx_hdmi_platform_remove,
 	.driver = {
 		.name = "imx-hdmi",
-		.owner = THIS_MODULE,
 		.of_match_table = imx_hdmi_dt_ids,
 	},
 };
diff --git a/drivers/staging/imx-drm/imx-hdmi.h b/drivers/gpu/drm/imx/imx-hdmi.h
index 39b677689db6..39b677689db6 100644
--- a/drivers/staging/imx-drm/imx-hdmi.h
+++ b/drivers/gpu/drm/imx/imx-hdmi.h
diff --git a/drivers/staging/imx-drm/imx-ldb.c b/drivers/gpu/drm/imx/imx-ldb.c
index 4662e00b456a..c60460043e24 100644
--- a/drivers/staging/imx-drm/imx-ldb.c
+++ b/drivers/gpu/drm/imx/imx-ldb.c
@@ -11,11 +11,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- * MA 02110-1301, USA.
  */
 
 #include <linux/module.h>
@@ -604,7 +599,6 @@ static struct platform_driver imx_ldb_driver = {
 	.driver		= {
 		.of_match_table = imx_ldb_dt_ids,
 		.name	= DRIVER_NAME,
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/staging/imx-drm/imx-tve.c b/drivers/gpu/drm/imx/imx-tve.c
index c628fcdc22ae..a729f4f7074c 100644
--- a/drivers/staging/imx-drm/imx-tve.c
+++ b/drivers/gpu/drm/imx/imx-tve.c
@@ -11,11 +11,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- * MA 02110-1301, USA.
  */
 
 #include <linux/clk.h>
@@ -133,6 +128,7 @@ static void tve_lock(void *__tve)
 __acquires(&tve->lock)
 {
 	struct imx_tve *tve = __tve;
+
 	spin_lock(&tve->lock);
 }
 
@@ -140,6 +136,7 @@ static void tve_unlock(void *__tve)
 __releases(&tve->lock)
 {
 	struct imx_tve *tve = __tve;
+
 	spin_unlock(&tve->lock);
 }
 
@@ -430,8 +427,7 @@ static long clk_tve_di_round_rate(struct clk_hw *hw, unsigned long rate,
 		return *prate / 4;
 	else if (div >= 2)
 		return *prate / 2;
-	else
-		return *prate;
+	return *prate;
 }
 
 static int clk_tve_di_set_rate(struct clk_hw *hw, unsigned long rate,
@@ -536,7 +532,7 @@ static struct regmap_config tve_regmap_config = {
 	.max_register = 0xdc,
 };
 
-static const char *imx_tve_modes[] = {
+static const char * const imx_tve_modes[] = {
 	[TVE_MODE_TVOUT]  = "tvout",
 	[TVE_MODE_VGA] = "vga",
 };
@@ -664,7 +660,8 @@ static int imx_tve_bind(struct device *dev, struct device *master, void *data)
 
 	ret = regmap_read(tve->regmap, TVE_COM_CONF_REG, &val);
 	if (ret < 0) {
-		dev_err(dev, "failed to read configuration register: %d\n", ret);
+		dev_err(dev, "failed to read configuration register: %d\n",
+			ret);
 		return ret;
 	}
 	if (val != 0x00100000) {
@@ -723,7 +720,6 @@ static struct platform_driver imx_tve_driver = {
 	.driver		= {
 		.of_match_table = imx_tve_dt_ids,
 		.name	= "imx-tve",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/staging/imx-drm/ipuv3-crtc.c b/drivers/gpu/drm/imx/ipuv3-crtc.c
index 720868bff35b..ebee59cb96d8 100644
--- a/drivers/staging/imx-drm/ipuv3-crtc.c
+++ b/drivers/gpu/drm/imx/ipuv3-crtc.c
@@ -11,11 +11,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- * MA 02110-1301, USA.
  */
 #include <linux/component.h>
 #include <linux/module.h>
@@ -201,7 +196,8 @@ static int ipu_crtc_mode_set(struct drm_crtc *crtc,
 		return ret;
 	}
 
-	return ipu_plane_mode_set(ipu_crtc->plane[0], crtc, mode, crtc->primary->fb,
+	return ipu_plane_mode_set(ipu_crtc->plane[0], crtc, mode,
+				  crtc->primary->fb,
 				  0, 0, mode->hdisplay, mode->vdisplay,
 				  x, y, mode->hdisplay, mode->vdisplay);
 }
@@ -226,9 +222,11 @@ static irqreturn_t ipu_irq_handler(int irq, void *dev_id)
 	imx_drm_handle_vblank(ipu_crtc->imx_crtc);
 
 	if (ipu_crtc->newfb) {
+		struct ipu_plane *plane = ipu_crtc->plane[0];
+
 		ipu_crtc->newfb = NULL;
-		ipu_plane_set_base(ipu_crtc->plane[0], ipu_crtc->base.primary->fb,
-				ipu_crtc->plane[0]->x, ipu_crtc->plane[0]->y);
+		ipu_plane_set_base(plane, ipu_crtc->base.primary->fb,
+				   plane->x, plane->y);
 		ipu_crtc_handle_pageflip(ipu_crtc);
 	}
 
diff --git a/drivers/staging/imx-drm/ipuv3-plane.c b/drivers/gpu/drm/imx/ipuv3-plane.c
index 50de10a550e9..6987e16fe99b 100644
--- a/drivers/staging/imx-drm/ipuv3-plane.c
+++ b/drivers/gpu/drm/imx/ipuv3-plane.c
@@ -62,9 +62,9 @@ static inline int calc_bandwidth(int width, int height, unsigned int vref)
 int ipu_plane_set_base(struct ipu_plane *ipu_plane, struct drm_framebuffer *fb,
 		       int x, int y)
 {
-	struct ipu_ch_param __iomem *cpmem;
 	struct drm_gem_cma_object *cma_obj;
 	unsigned long eba;
+	int active;
 
 	cma_obj = drm_fb_cma_get_gem_obj(fb, 0);
 	if (!cma_obj) {
@@ -75,13 +75,17 @@ int ipu_plane_set_base(struct ipu_plane *ipu_plane, struct drm_framebuffer *fb,
 	dev_dbg(ipu_plane->base.dev->dev, "phys = %pad, x = %d, y = %d",
 		&cma_obj->paddr, x, y);
 
-	cpmem = ipu_get_cpmem(ipu_plane->ipu_ch);
-	ipu_cpmem_set_stride(cpmem, fb->pitches[0]);
-
 	eba = cma_obj->paddr + fb->offsets[0] +
 	      fb->pitches[0] * y + (fb->bits_per_pixel >> 3) * x;
-	ipu_cpmem_set_buffer(cpmem, 0, eba);
-	ipu_cpmem_set_buffer(cpmem, 1, eba);
+
+	if (ipu_plane->enabled) {
+		active = ipu_idmac_get_current_buffer(ipu_plane->ipu_ch);
+		ipu_cpmem_set_buffer(ipu_plane->ipu_ch, !active, eba);
+		ipu_idmac_select_buffer(ipu_plane->ipu_ch, !active);
+	} else {
+		ipu_cpmem_set_buffer(ipu_plane->ipu_ch, 0, eba);
+		ipu_cpmem_set_buffer(ipu_plane->ipu_ch, 1, eba);
+	}
 
 	/* cache offsets for subsequent pageflips */
 	ipu_plane->x = x;
@@ -97,7 +101,6 @@ int ipu_plane_mode_set(struct ipu_plane *ipu_plane, struct drm_crtc *crtc,
 		       uint32_t src_x, uint32_t src_y,
 		       uint32_t src_w, uint32_t src_h)
 {
-	struct ipu_ch_param __iomem *cpmem;
 	struct device *dev = ipu_plane->base.dev->dev;
 	int ret;
 
@@ -140,6 +143,18 @@ int ipu_plane_mode_set(struct ipu_plane *ipu_plane, struct drm_crtc *crtc,
 	if (crtc_h < 2)
 		return -EINVAL;
 
+	/*
+	 * since we cannot touch active IDMAC channels, we do not support
+	 * resizing the enabled plane or changing its format
+	 */
+	if (ipu_plane->enabled) {
+		if (src_w != ipu_plane->w || src_h != ipu_plane->h ||
+		    fb->pixel_format != ipu_plane->base.fb->pixel_format)
+			return -EINVAL;
+
+		return ipu_plane_set_base(ipu_plane, fb, src_x, src_y);
+	}
+
 	switch (ipu_plane->dp_flow) {
 	case IPU_DP_FLOW_SYNC_BG:
 		ret = ipu_dp_setup_channel(ipu_plane->dp,
@@ -151,14 +166,22 @@ int ipu_plane_mode_set(struct ipu_plane *ipu_plane, struct drm_crtc *crtc,
 				ret);
 			return ret;
 		}
-		ipu_dp_set_global_alpha(ipu_plane->dp, 1, 0, 1);
+		ipu_dp_set_global_alpha(ipu_plane->dp, true, 0, true);
 		break;
 	case IPU_DP_FLOW_SYNC_FG:
 		ipu_dp_setup_channel(ipu_plane->dp,
 				ipu_drm_fourcc_to_colorspace(fb->pixel_format),
 				IPUV3_COLORSPACE_UNKNOWN);
 		ipu_dp_set_window_pos(ipu_plane->dp, crtc_x, crtc_y);
-		break;
+		/* Enable local alpha on partial plane */
+		switch (fb->pixel_format) {
+		case DRM_FORMAT_ARGB8888:
+		case DRM_FORMAT_ABGR8888:
+			ipu_dp_set_global_alpha(ipu_plane->dp, false, 0, false);
+			break;
+		default:
+			break;
+		}
 	}
 
 	ret = ipu_dmfc_init_channel(ipu_plane->dmfc, crtc_w);
@@ -175,21 +198,25 @@ int ipu_plane_mode_set(struct ipu_plane *ipu_plane, struct drm_crtc *crtc,
 		return ret;
 	}
 
-	cpmem = ipu_get_cpmem(ipu_plane->ipu_ch);
-	ipu_ch_param_zero(cpmem);
-	ipu_cpmem_set_resolution(cpmem, src_w, src_h);
-	ret = ipu_cpmem_set_fmt(cpmem, fb->pixel_format);
+	ipu_cpmem_zero(ipu_plane->ipu_ch);
+	ipu_cpmem_set_resolution(ipu_plane->ipu_ch, src_w, src_h);
+	ret = ipu_cpmem_set_fmt(ipu_plane->ipu_ch, fb->pixel_format);
 	if (ret < 0) {
 		dev_err(dev, "unsupported pixel format 0x%08x\n",
 			fb->pixel_format);
 		return ret;
 	}
 	ipu_cpmem_set_high_priority(ipu_plane->ipu_ch);
+	ipu_idmac_set_double_buffer(ipu_plane->ipu_ch, 1);
+	ipu_cpmem_set_stride(ipu_plane->ipu_ch, fb->pitches[0]);
 
 	ret = ipu_plane_set_base(ipu_plane, fb, src_x, src_y);
 	if (ret < 0)
 		return ret;
 
+	ipu_plane->w = src_w;
+	ipu_plane->h = src_h;
+
 	return 0;
 }
 
@@ -263,29 +290,6 @@ void ipu_plane_disable(struct ipu_plane *ipu_plane)
 		ipu_dp_disable(ipu_plane->ipu);
 }
 
-static void ipu_plane_dpms(struct ipu_plane *ipu_plane, int mode)
-{
-	bool enable;
-
-	DRM_DEBUG_KMS("mode = %d", mode);
-
-	enable = (mode == DRM_MODE_DPMS_ON);
-
-	if (enable == ipu_plane->enabled)
-		return;
-
-	if (enable) {
-		ipu_plane_enable(ipu_plane);
-	} else {
-		ipu_plane_disable(ipu_plane);
-
-		ipu_idmac_put(ipu_plane->ipu_ch);
-		ipu_dmfc_put(ipu_plane->dmfc);
-		if (ipu_plane->dp)
-			ipu_dp_put(ipu_plane->dp);
-	}
-}
-
 /*
  * drm_plane API
  */
@@ -319,7 +323,8 @@ static int ipu_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 				plane->crtc, crtc);
 	plane->crtc = crtc;
 
-	ipu_plane_dpms(ipu_plane, DRM_MODE_DPMS_ON);
+	if (!ipu_plane->enabled)
+		ipu_plane_enable(ipu_plane);
 
 	return 0;
 }
@@ -330,7 +335,8 @@ static int ipu_disable_plane(struct drm_plane *plane)
 
 	DRM_DEBUG_KMS("[%d] %s\n", __LINE__, __func__);
 
-	ipu_plane_dpms(ipu_plane, DRM_MODE_DPMS_OFF);
+	if (ipu_plane->enabled)
+		ipu_plane_disable(ipu_plane);
 
 	ipu_plane_put_resources(ipu_plane);
 
diff --git a/drivers/staging/imx-drm/ipuv3-plane.h b/drivers/gpu/drm/imx/ipuv3-plane.h
index c0aae5bcb5d4..af125fb40ef5 100644
--- a/drivers/staging/imx-drm/ipuv3-plane.h
+++ b/drivers/gpu/drm/imx/ipuv3-plane.h
@@ -26,6 +26,8 @@ struct ipu_plane {
 
 	int			x;
 	int			y;
+	int			w;
+	int			h;
 
 	bool			enabled;
 };
diff --git a/drivers/staging/imx-drm/parallel-display.c b/drivers/gpu/drm/imx/parallel-display.c
index 4ca61afdf622..796c3c1c170a 100644
--- a/drivers/staging/imx-drm/parallel-display.c
+++ b/drivers/gpu/drm/imx/parallel-display.c
@@ -11,11 +11,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- * MA 02110-1301, USA.
  */
 
 #include <linux/component.h>
@@ -70,6 +65,7 @@ static int imx_pd_connector_get_modes(struct drm_connector *connector)
 
 	if (imxpd->mode_valid) {
 		struct drm_display_mode *mode = drm_mode_create(connector->dev);
+
 		if (!mode)
 			return -EINVAL;
 		drm_mode_copy(mode, &imxpd->mode);
@@ -80,6 +76,7 @@ static int imx_pd_connector_get_modes(struct drm_connector *connector)
 
 	if (np) {
 		struct drm_display_mode *mode = drm_mode_create(connector->dev);
+
 		if (!mode)
 			return -EINVAL;
 		of_get_drm_display_mode(np, &imxpd->mode, OF_USE_NATIVE_MODE);
@@ -126,6 +123,10 @@ static void imx_pd_encoder_prepare(struct drm_encoder *encoder)
 
 static void imx_pd_encoder_commit(struct drm_encoder *encoder)
 {
+	struct imx_parallel_display *imxpd = enc_to_imxpd(encoder);
+
+	drm_panel_prepare(imxpd->panel);
+	drm_panel_enable(imxpd->panel);
 }
 
 static void imx_pd_encoder_mode_set(struct drm_encoder *encoder,
@@ -136,6 +137,10 @@ static void imx_pd_encoder_mode_set(struct drm_encoder *encoder,
 
 static void imx_pd_encoder_disable(struct drm_encoder *encoder)
 {
+	struct imx_parallel_display *imxpd = enc_to_imxpd(encoder);
+
+	drm_panel_disable(imxpd->panel);
+	drm_panel_unprepare(imxpd->panel);
 }
 
 static struct drm_connector_funcs imx_pd_connector_funcs = {
@@ -226,7 +231,8 @@ static int imx_pd_bind(struct device *dev, struct device *master, void *data)
 		else if (!strcmp(fmt, "bgr666"))
 			imxpd->interface_pix_fmt = V4L2_PIX_FMT_BGR666;
 		else if (!strcmp(fmt, "lvds666"))
-			imxpd->interface_pix_fmt = v4l2_fourcc('L', 'V', 'D', '6');
+			imxpd->interface_pix_fmt =
+					v4l2_fourcc('L', 'V', 'D', '6');
 	}
 
 	panel_node = of_parse_phandle(np, "fsl,panel", 0);
@@ -281,7 +287,6 @@ static struct platform_driver imx_pd_driver = {
 	.driver		= {
 		.of_match_table = imx_pd_dt_ids,
 		.name	= "imx-parallel-display",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/gpu/drm/mga/mga_dma.c b/drivers/gpu/drm/mga/mga_dma.c
index c3bf059ba720..8cfa9cb74c86 100644
--- a/drivers/gpu/drm/mga/mga_dma.c
+++ b/drivers/gpu/drm/mga/mga_dma.c
@@ -502,31 +502,31 @@ static int mga_do_agp_dma_bootstrap(struct drm_device *dev,
 		return err;
 	}
 
-	/* Make drm_addbufs happy by not trying to create a mapping for less
-	 * than a page.
+	/* Make drm_legacy_addbufs happy by not trying to create a mapping for
+	 * less than a page.
 	 */
 	if (warp_size < PAGE_SIZE)
 		warp_size = PAGE_SIZE;
 
 	offset = 0;
-	err = drm_addmap(dev, offset, warp_size,
-			 _DRM_AGP, _DRM_READ_ONLY, &dev_priv->warp);
+	err = drm_legacy_addmap(dev, offset, warp_size,
+				_DRM_AGP, _DRM_READ_ONLY, &dev_priv->warp);
 	if (err) {
 		DRM_ERROR("Unable to map WARP microcode: %d\n", err);
 		return err;
 	}
 
 	offset += warp_size;
-	err = drm_addmap(dev, offset, dma_bs->primary_size,
-			 _DRM_AGP, _DRM_READ_ONLY, &dev_priv->primary);
+	err = drm_legacy_addmap(dev, offset, dma_bs->primary_size,
+				_DRM_AGP, _DRM_READ_ONLY, &dev_priv->primary);
 	if (err) {
 		DRM_ERROR("Unable to map primary DMA region: %d\n", err);
 		return err;
 	}
 
 	offset += dma_bs->primary_size;
-	err = drm_addmap(dev, offset, secondary_size,
-			 _DRM_AGP, 0, &dev->agp_buffer_map);
+	err = drm_legacy_addmap(dev, offset, secondary_size,
+				_DRM_AGP, 0, &dev->agp_buffer_map);
 	if (err) {
 		DRM_ERROR("Unable to map secondary DMA region: %d\n", err);
 		return err;
@@ -538,7 +538,7 @@ static int mga_do_agp_dma_bootstrap(struct drm_device *dev,
 	req.flags = _DRM_AGP_BUFFER;
 	req.agp_start = offset;
 
-	err = drm_addbufs_agp(dev, &req);
+	err = drm_legacy_addbufs_agp(dev, &req);
 	if (err) {
 		DRM_ERROR("Unable to add secondary DMA buffers: %d\n", err);
 		return err;
@@ -559,16 +559,16 @@ static int mga_do_agp_dma_bootstrap(struct drm_device *dev,
 	}
 
 	offset += secondary_size;
-	err = drm_addmap(dev, offset, agp_size - offset,
-			 _DRM_AGP, 0, &dev_priv->agp_textures);
+	err = drm_legacy_addmap(dev, offset, agp_size - offset,
+				_DRM_AGP, 0, &dev_priv->agp_textures);
 	if (err) {
 		DRM_ERROR("Unable to map AGP texture region %d\n", err);
 		return err;
 	}
 
-	drm_core_ioremap(dev_priv->warp, dev);
-	drm_core_ioremap(dev_priv->primary, dev);
-	drm_core_ioremap(dev->agp_buffer_map, dev);
+	drm_legacy_ioremap(dev_priv->warp, dev);
+	drm_legacy_ioremap(dev_priv->primary, dev);
+	drm_legacy_ioremap(dev->agp_buffer_map, dev);
 
 	if (!dev_priv->warp->handle ||
 	    !dev_priv->primary->handle || !dev->agp_buffer_map->handle) {
@@ -602,7 +602,7 @@ static int mga_do_agp_dma_bootstrap(struct drm_device *dev,
  *
  * \todo
  * Determine whether the maximum address passed to drm_pci_alloc is correct.
- * The same goes for drm_addbufs_pci.
+ * The same goes for drm_legacy_addbufs_pci.
  *
  * \sa mga_do_dma_bootstrap, mga_do_agp_dma_bootstrap
  */
@@ -622,15 +622,15 @@ static int mga_do_pci_dma_bootstrap(struct drm_device *dev,
 		return -EFAULT;
 	}
 
-	/* Make drm_addbufs happy by not trying to create a mapping for less
-	 * than a page.
+	/* Make drm_legacy_addbufs happy by not trying to create a mapping for
+	 * less than a page.
 	 */
 	if (warp_size < PAGE_SIZE)
 		warp_size = PAGE_SIZE;
 
 	/* The proper alignment is 0x100 for this mapping */
-	err = drm_addmap(dev, 0, warp_size, _DRM_CONSISTENT,
-			 _DRM_READ_ONLY, &dev_priv->warp);
+	err = drm_legacy_addmap(dev, 0, warp_size, _DRM_CONSISTENT,
+				_DRM_READ_ONLY, &dev_priv->warp);
 	if (err != 0) {
 		DRM_ERROR("Unable to create mapping for WARP microcode: %d\n",
 			  err);
@@ -645,8 +645,8 @@ static int mga_do_pci_dma_bootstrap(struct drm_device *dev,
 	for (primary_size = dma_bs->primary_size; primary_size != 0;
 	     primary_size >>= 1) {
 		/* The proper alignment for this mapping is 0x04 */
-		err = drm_addmap(dev, 0, primary_size, _DRM_CONSISTENT,
-				 _DRM_READ_ONLY, &dev_priv->primary);
+		err = drm_legacy_addmap(dev, 0, primary_size, _DRM_CONSISTENT,
+					_DRM_READ_ONLY, &dev_priv->primary);
 		if (!err)
 			break;
 	}
@@ -669,7 +669,7 @@ static int mga_do_pci_dma_bootstrap(struct drm_device *dev,
 		req.count = bin_count;
 		req.size = dma_bs->secondary_bin_size;
 
-		err = drm_addbufs_pci(dev, &req);
+		err = drm_legacy_addbufs_pci(dev, &req);
 		if (!err)
 			break;
 	}
@@ -708,15 +708,16 @@ static int mga_do_dma_bootstrap(struct drm_device *dev,
 	/* The first steps are the same for both PCI and AGP based DMA.  Map
 	 * the cards MMIO registers and map a status page.
 	 */
-	err = drm_addmap(dev, dev_priv->mmio_base, dev_priv->mmio_size,
-			 _DRM_REGISTERS, _DRM_READ_ONLY, &dev_priv->mmio);
+	err = drm_legacy_addmap(dev, dev_priv->mmio_base, dev_priv->mmio_size,
+				_DRM_REGISTERS, _DRM_READ_ONLY,
+				&dev_priv->mmio);
 	if (err) {
 		DRM_ERROR("Unable to map MMIO region: %d\n", err);
 		return err;
 	}
 
-	err = drm_addmap(dev, 0, SAREA_MAX, _DRM_SHM,
-			 _DRM_READ_ONLY | _DRM_LOCKED | _DRM_KERNEL,
+	err = drm_legacy_addmap(dev, 0, SAREA_MAX, _DRM_SHM,
+				_DRM_READ_ONLY | _DRM_LOCKED | _DRM_KERNEL,
 			 &dev_priv->status);
 	if (err) {
 		DRM_ERROR("Unable to map status region: %d\n", err);
@@ -809,7 +810,7 @@ static int mga_do_init_dma(struct drm_device *dev, drm_mga_init_t *init)
 	dev_priv->texture_offset = init->texture_offset[0];
 	dev_priv->texture_size = init->texture_size[0];
 
-	dev_priv->sarea = drm_getsarea(dev);
+	dev_priv->sarea = drm_legacy_getsarea(dev);
 	if (!dev_priv->sarea) {
 		DRM_ERROR("failed to find sarea!\n");
 		return -EINVAL;
@@ -820,37 +821,37 @@ static int mga_do_init_dma(struct drm_device *dev, drm_mga_init_t *init)
 		dev_priv->dma_access = MGA_PAGPXFER;
 		dev_priv->wagp_enable = MGA_WAGP_ENABLE;
 
-		dev_priv->status = drm_core_findmap(dev, init->status_offset);
+		dev_priv->status = drm_legacy_findmap(dev, init->status_offset);
 		if (!dev_priv->status) {
 			DRM_ERROR("failed to find status page!\n");
 			return -EINVAL;
 		}
-		dev_priv->mmio = drm_core_findmap(dev, init->mmio_offset);
+		dev_priv->mmio = drm_legacy_findmap(dev, init->mmio_offset);
 		if (!dev_priv->mmio) {
 			DRM_ERROR("failed to find mmio region!\n");
 			return -EINVAL;
 		}
-		dev_priv->warp = drm_core_findmap(dev, init->warp_offset);
+		dev_priv->warp = drm_legacy_findmap(dev, init->warp_offset);
 		if (!dev_priv->warp) {
 			DRM_ERROR("failed to find warp microcode region!\n");
 			return -EINVAL;
 		}
-		dev_priv->primary = drm_core_findmap(dev, init->primary_offset);
+		dev_priv->primary = drm_legacy_findmap(dev, init->primary_offset);
 		if (!dev_priv->primary) {
 			DRM_ERROR("failed to find primary dma region!\n");
 			return -EINVAL;
 		}
 		dev->agp_buffer_token = init->buffers_offset;
 		dev->agp_buffer_map =
-		    drm_core_findmap(dev, init->buffers_offset);
+		    drm_legacy_findmap(dev, init->buffers_offset);
 		if (!dev->agp_buffer_map) {
 			DRM_ERROR("failed to find dma buffer region!\n");
 			return -EINVAL;
 		}
 
-		drm_core_ioremap(dev_priv->warp, dev);
-		drm_core_ioremap(dev_priv->primary, dev);
-		drm_core_ioremap(dev->agp_buffer_map, dev);
+		drm_legacy_ioremap(dev_priv->warp, dev);
+		drm_legacy_ioremap(dev_priv->primary, dev);
+		drm_legacy_ioremap(dev->agp_buffer_map, dev);
 	}
 
 	dev_priv->sarea_priv =
@@ -936,14 +937,14 @@ static int mga_do_cleanup_dma(struct drm_device *dev, int full_cleanup)
 
 		if ((dev_priv->warp != NULL)
 		    && (dev_priv->warp->type != _DRM_CONSISTENT))
-			drm_core_ioremapfree(dev_priv->warp, dev);
+			drm_legacy_ioremapfree(dev_priv->warp, dev);
 
 		if ((dev_priv->primary != NULL)
 		    && (dev_priv->primary->type != _DRM_CONSISTENT))
-			drm_core_ioremapfree(dev_priv->primary, dev);
+			drm_legacy_ioremapfree(dev_priv->primary, dev);
 
 		if (dev->agp_buffer_map != NULL)
-			drm_core_ioremapfree(dev->agp_buffer_map, dev);
+			drm_legacy_ioremapfree(dev->agp_buffer_map, dev);
 
 		if (dev_priv->used_new_dma_init) {
 #if __OS_HAS_AGP
diff --git a/drivers/gpu/drm/mga/mga_drv.c b/drivers/gpu/drm/mga/mga_drv.c
index 6b1a87c8aac5..5e2f131a6a72 100644
--- a/drivers/gpu/drm/mga/mga_drv.c
+++ b/drivers/gpu/drm/mga/mga_drv.c
@@ -48,7 +48,7 @@ static const struct file_operations mga_driver_fops = {
 	.open = drm_open,
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_mmap,
+	.mmap = drm_legacy_mmap,
 	.poll = drm_poll,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl = mga_compat_ioctl,
@@ -64,6 +64,7 @@ static struct drm_driver driver = {
 	.load = mga_driver_load,
 	.unload = mga_driver_unload,
 	.lastclose = mga_driver_lastclose,
+	.set_busid = drm_pci_set_busid,
 	.dma_quiescent = mga_driver_dma_quiescent,
 	.device_is_agp = mga_driver_device_is_agp,
 	.get_vblank_counter = mga_get_vblank_counter,
diff --git a/drivers/gpu/drm/mga/mga_drv.h b/drivers/gpu/drm/mga/mga_drv.h
index fe453213600a..b4a2014917e5 100644
--- a/drivers/gpu/drm/mga/mga_drv.h
+++ b/drivers/gpu/drm/mga/mga_drv.h
@@ -31,6 +31,8 @@
 #ifndef __MGA_DRV_H__
 #define __MGA_DRV_H__
 
+#include <drm/drm_legacy.h>
+
 /* General customization:
  */
 
diff --git a/drivers/gpu/drm/mgag200/mgag200_drv.c b/drivers/gpu/drm/mgag200/mgag200_drv.c
index 2d75d6df0789..97745991544d 100644
--- a/drivers/gpu/drm/mgag200/mgag200_drv.c
+++ b/drivers/gpu/drm/mgag200/mgag200_drv.c
@@ -91,6 +91,7 @@ static struct drm_driver driver = {
 	.driver_features = DRIVER_GEM | DRIVER_MODESET,
 	.load = mgag200_driver_load,
 	.unload = mgag200_driver_unload,
+	.set_busid = drm_pci_set_busid,
 	.fops = &mgag200_driver_fops,
 	.name = DRIVER_NAME,
 	.desc = DRIVER_DESC,
diff --git a/drivers/gpu/drm/mgag200/mgag200_drv.h b/drivers/gpu/drm/mgag200/mgag200_drv.h
index 80de23d9b9c9..e9eea1d4e7c3 100644
--- a/drivers/gpu/drm/mgag200/mgag200_drv.h
+++ b/drivers/gpu/drm/mgag200/mgag200_drv.h
@@ -22,6 +22,8 @@
 #include <drm/ttm/ttm_memory.h>
 #include <drm/ttm/ttm_module.h>
 
+#include <drm/drm_gem.h>
+
 #include <linux/i2c.h>
 #include <linux/i2c-algo-bit.h>
 
@@ -190,8 +192,6 @@ struct mga_device {
 	resource_size_t			rmmio_size;
 	void __iomem			*rmmio;
 
-	drm_local_map_t			*framebuffer;
-
 	struct mga_mc			mc;
 	struct mga_mode_info		mode_info;
 
@@ -224,7 +224,7 @@ struct mgag200_bo {
 	struct ttm_placement placement;
 	struct ttm_bo_kmap_obj kmap;
 	struct drm_gem_object gem;
-	u32 placements[3];
+	struct ttm_place placements[3];
 	int pin_count;
 };
 #define gem_to_mga_bo(gobj) container_of((gobj), struct mgag200_bo, gem)
diff --git a/drivers/gpu/drm/mgag200/mgag200_fb.c b/drivers/gpu/drm/mgag200/mgag200_fb.c
index 5451dc58eff1..4415af3666ab 100644
--- a/drivers/gpu/drm/mgag200/mgag200_fb.c
+++ b/drivers/gpu/drm/mgag200/mgag200_fb.c
@@ -158,7 +158,8 @@ static int mgag200fb_create_object(struct mga_fbdev *afbdev,
 static int mgag200fb_create(struct drm_fb_helper *helper,
 			   struct drm_fb_helper_surface_size *sizes)
 {
-	struct mga_fbdev *mfbdev = (struct mga_fbdev *)helper;
+	struct mga_fbdev *mfbdev =
+		container_of(helper, struct mga_fbdev, helper);
 	struct drm_device *dev = mfbdev->helper.dev;
 	struct drm_mode_fb_cmd2 mode_cmd;
 	struct mga_device *mdev = dev->dev_private;
diff --git a/drivers/gpu/drm/mgag200/mgag200_mode.c b/drivers/gpu/drm/mgag200/mgag200_mode.c
index 45f04dea0ac2..9872ba9abf1a 100644
--- a/drivers/gpu/drm/mgag200/mgag200_mode.c
+++ b/drivers/gpu/drm/mgag200/mgag200_mode.c
@@ -15,6 +15,7 @@
 
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_plane_helper.h>
 
 #include "mgag200_drv.h"
 
@@ -1483,11 +1484,7 @@ static int mga_vga_mode_valid(struct drm_connector *connector,
 {
 	struct drm_device *dev = connector->dev;
 	struct mga_device *mdev = (struct mga_device*)dev->dev_private;
-	struct mga_fbdev *mfbdev = mdev->mfbdev;
-	struct drm_fb_helper *fb_helper = &mfbdev->helper;
-	struct drm_fb_helper_connector *fb_helper_conn = NULL;
 	int bpp = 32;
-	int i = 0;
 
 	if (IS_G200_SE(mdev)) {
 		if (mdev->unique_rev_id == 0x01) {
@@ -1537,21 +1534,14 @@ static int mga_vga_mode_valid(struct drm_connector *connector,
 	}
 
 	/* Validate the mode input by the user */
-	for (i = 0; i < fb_helper->connector_count; i++) {
-		if (fb_helper->connector_info[i]->connector == connector) {
-			/* Found the helper for this connector */
-			fb_helper_conn = fb_helper->connector_info[i];
-			if (fb_helper_conn->cmdline_mode.specified) {
-				if (fb_helper_conn->cmdline_mode.bpp_specified) {
-					bpp = fb_helper_conn->cmdline_mode.bpp;
-				}
-			}
-		}
+	if (connector->cmdline_mode.specified) {
+		if (connector->cmdline_mode.bpp_specified)
+			bpp = connector->cmdline_mode.bpp;
 	}
 
 	if ((mode->hdisplay * mode->vdisplay * (bpp/8)) > mdev->mc.vram_size) {
-		if (fb_helper_conn)
-			fb_helper_conn->cmdline_mode.specified = false;
+		if (connector->cmdline_mode.specified)
+			connector->cmdline_mode.specified = false;
 		return MODE_BAD;
 	}
 
diff --git a/drivers/gpu/drm/mgag200/mgag200_ttm.c b/drivers/gpu/drm/mgag200/mgag200_ttm.c
index 5a00e90696de..d16964ea0ed4 100644
--- a/drivers/gpu/drm/mgag200/mgag200_ttm.c
+++ b/drivers/gpu/drm/mgag200/mgag200_ttm.c
@@ -293,18 +293,22 @@ void mgag200_mm_fini(struct mga_device *mdev)
 void mgag200_ttm_placement(struct mgag200_bo *bo, int domain)
 {
 	u32 c = 0;
-	bo->placement.fpfn = 0;
-	bo->placement.lpfn = 0;
+	unsigned i;
+
 	bo->placement.placement = bo->placements;
 	bo->placement.busy_placement = bo->placements;
 	if (domain & TTM_PL_FLAG_VRAM)
-		bo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_VRAM;
+		bo->placements[c++].flags = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_VRAM;
 	if (domain & TTM_PL_FLAG_SYSTEM)
-		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		bo->placements[c++].flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
 	if (!c)
-		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		bo->placements[c++].flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
 	bo->placement.num_placement = c;
 	bo->placement.num_busy_placement = c;
+	for (i = 0; i < c; ++i) {
+		bo->placements[i].fpfn = 0;
+		bo->placements[i].lpfn = 0;
+	}
 }
 
 int mgag200_bo_create(struct drm_device *dev, int size, int align,
@@ -335,7 +339,7 @@ int mgag200_bo_create(struct drm_device *dev, int size, int align,
 	ret = ttm_bo_init(&mdev->ttm.bdev, &mgabo->bo, size,
 			  ttm_bo_type_device, &mgabo->placement,
 			  align >> PAGE_SHIFT, false, NULL, acc_size,
-			  NULL, mgag200_bo_ttm_destroy);
+			  NULL, NULL, mgag200_bo_ttm_destroy);
 	if (ret)
 		return ret;
 
@@ -361,7 +365,7 @@ int mgag200_bo_pin(struct mgag200_bo *bo, u32 pl_flag, u64 *gpu_addr)
 
 	mgag200_ttm_placement(bo, pl_flag);
 	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret)
 		return ret;
@@ -384,7 +388,7 @@ int mgag200_bo_unpin(struct mgag200_bo *bo)
 		return 0;
 
 	for (i = 0; i < bo->placement.num_placement ; i++)
-		bo->placements[i] &= ~TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags &= ~TTM_PL_FLAG_NO_EVICT;
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret)
 		return ret;
@@ -408,7 +412,7 @@ int mgag200_bo_push_sysram(struct mgag200_bo *bo)
 
 	mgag200_ttm_placement(bo, TTM_PL_FLAG_SYSTEM);
 	for (i = 0; i < bo->placement.num_placement ; i++)
-		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret) {
@@ -424,7 +428,7 @@ int mgag200_mmap(struct file *filp, struct vm_area_struct *vma)
 	struct mga_device *mdev;
 
 	if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET))
-		return drm_mmap(filp, vma);
+		return -EINVAL;
 
 	file_priv = filp->private_data;
 	mdev = file_priv->minor->dev->dev_private;
diff --git a/drivers/gpu/drm/msm/Kconfig b/drivers/gpu/drm/msm/Kconfig
index c99c50de3226..5b2a1ff95d3d 100644
--- a/drivers/gpu/drm/msm/Kconfig
+++ b/drivers/gpu/drm/msm/Kconfig
@@ -3,7 +3,9 @@ config DRM_MSM
 	tristate "MSM DRM"
 	depends on DRM
 	depends on ARCH_QCOM || (ARM && COMPILE_TEST)
+	select REGULATOR
 	select DRM_KMS_HELPER
+	select DRM_PANEL
 	select SHMEM
 	select TMPFS
 	default y
diff --git a/drivers/gpu/drm/msm/Makefile b/drivers/gpu/drm/msm/Makefile
index 93ca49c8df44..143d988f8add 100644
--- a/drivers/gpu/drm/msm/Makefile
+++ b/drivers/gpu/drm/msm/Makefile
@@ -4,8 +4,10 @@ ifeq (, $(findstring -W,$(EXTRA_CFLAGS)))
 endif
 
 msm-y := \
+	adreno/adreno_device.o \
 	adreno/adreno_gpu.o \
 	adreno/a3xx_gpu.o \
+	adreno/a4xx_gpu.o \
 	hdmi/hdmi.o \
 	hdmi/hdmi_audio.o \
 	hdmi/hdmi_bridge.o \
@@ -18,15 +20,20 @@ msm-y := \
 	mdp/mdp_kms.o \
 	mdp/mdp4/mdp4_crtc.o \
 	mdp/mdp4/mdp4_dtv_encoder.o \
+	mdp/mdp4/mdp4_lcdc_encoder.o \
+	mdp/mdp4/mdp4_lvds_connector.o \
 	mdp/mdp4/mdp4_irq.o \
 	mdp/mdp4/mdp4_kms.o \
 	mdp/mdp4/mdp4_plane.o \
+	mdp/mdp5/mdp5_cfg.o \
+	mdp/mdp5/mdp5_ctl.o \
 	mdp/mdp5/mdp5_crtc.o \
 	mdp/mdp5/mdp5_encoder.o \
 	mdp/mdp5/mdp5_irq.o \
 	mdp/mdp5/mdp5_kms.o \
 	mdp/mdp5/mdp5_plane.o \
 	mdp/mdp5/mdp5_smp.o \
+	msm_atomic.o \
 	msm_drv.o \
 	msm_fb.o \
 	msm_gem.o \
@@ -39,5 +46,6 @@ msm-y := \
 	msm_ringbuffer.o
 
 msm-$(CONFIG_DRM_MSM_FBDEV) += msm_fbdev.o
+msm-$(CONFIG_COMMON_CLK) += mdp/mdp4/mdp4_lvds_pll.o
 
 obj-$(CONFIG_DRM_MSM)	+= msm.o
diff --git a/drivers/gpu/drm/msm/adreno/a2xx.xml.h b/drivers/gpu/drm/msm/adreno/a2xx.xml.h
index a8a144b38eaa..22882cc0a573 100644
--- a/drivers/gpu/drm/msm/adreno/a2xx.xml.h
+++ b/drivers/gpu/drm/msm/adreno/a2xx.xml.h
@@ -11,10 +11,10 @@ The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/adreno.xml               (    364 bytes, from 2013-11-30 14:47:15)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml  (   1453 bytes, from 2013-03-31 16:51:27)
 - /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml          (  32901 bytes, from 2014-06-02 15:21:30)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (   9859 bytes, from 2014-06-02 15:21:30)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  14477 bytes, from 2014-05-16 11:51:57)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  58020 bytes, from 2014-06-25 12:57:16)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (  26602 bytes, from 2014-06-25 12:57:16)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (  10551 bytes, from 2014-11-13 22:44:30)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  15053 bytes, from 2014-11-09 15:45:47)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  63169 bytes, from 2014-11-13 22:44:18)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (  49097 bytes, from 2014-11-14 15:38:00)
 
 Copyright (C) 2013-2014 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -926,11 +926,11 @@ static inline uint32_t A2XX_VGT_DRAW_INITIATOR_INDEX_SIZE(enum pc_di_index_size
 #define A2XX_VGT_DRAW_INITIATOR_NOT_EOP				0x00001000
 #define A2XX_VGT_DRAW_INITIATOR_SMALL_INDEX			0x00002000
 #define A2XX_VGT_DRAW_INITIATOR_PRE_DRAW_INITIATOR_ENABLE	0x00004000
-#define A2XX_VGT_DRAW_INITIATOR_NUM_INDICES__MASK		0xffff0000
-#define A2XX_VGT_DRAW_INITIATOR_NUM_INDICES__SHIFT		16
-static inline uint32_t A2XX_VGT_DRAW_INITIATOR_NUM_INDICES(uint32_t val)
+#define A2XX_VGT_DRAW_INITIATOR_NUM_INSTANCES__MASK		0xff000000
+#define A2XX_VGT_DRAW_INITIATOR_NUM_INSTANCES__SHIFT		24
+static inline uint32_t A2XX_VGT_DRAW_INITIATOR_NUM_INSTANCES(uint32_t val)
 {
-	return ((val) << A2XX_VGT_DRAW_INITIATOR_NUM_INDICES__SHIFT) & A2XX_VGT_DRAW_INITIATOR_NUM_INDICES__MASK;
+	return ((val) << A2XX_VGT_DRAW_INITIATOR_NUM_INSTANCES__SHIFT) & A2XX_VGT_DRAW_INITIATOR_NUM_INSTANCES__MASK;
 }
 
 #define REG_A2XX_VGT_IMMED_DATA					0x000021fd
@@ -1243,13 +1243,13 @@ static inline uint32_t A2XX_CLEAR_COLOR_ALPHA(uint32_t val)
 #define A2XX_PA_SU_POINT_SIZE_HEIGHT__SHIFT			0
 static inline uint32_t A2XX_PA_SU_POINT_SIZE_HEIGHT(float val)
 {
-	return ((((uint32_t)(val * 8.0))) << A2XX_PA_SU_POINT_SIZE_HEIGHT__SHIFT) & A2XX_PA_SU_POINT_SIZE_HEIGHT__MASK;
+	return ((((uint32_t)(val * 16.0))) << A2XX_PA_SU_POINT_SIZE_HEIGHT__SHIFT) & A2XX_PA_SU_POINT_SIZE_HEIGHT__MASK;
 }
 #define A2XX_PA_SU_POINT_SIZE_WIDTH__MASK			0xffff0000
 #define A2XX_PA_SU_POINT_SIZE_WIDTH__SHIFT			16
 static inline uint32_t A2XX_PA_SU_POINT_SIZE_WIDTH(float val)
 {
-	return ((((uint32_t)(val * 8.0))) << A2XX_PA_SU_POINT_SIZE_WIDTH__SHIFT) & A2XX_PA_SU_POINT_SIZE_WIDTH__MASK;
+	return ((((uint32_t)(val * 16.0))) << A2XX_PA_SU_POINT_SIZE_WIDTH__SHIFT) & A2XX_PA_SU_POINT_SIZE_WIDTH__MASK;
 }
 
 #define REG_A2XX_PA_SU_POINT_MINMAX				0x00002281
@@ -1257,13 +1257,13 @@ static inline uint32_t A2XX_PA_SU_POINT_SIZE_WIDTH(float val)
 #define A2XX_PA_SU_POINT_MINMAX_MIN__SHIFT			0
 static inline uint32_t A2XX_PA_SU_POINT_MINMAX_MIN(float val)
 {
-	return ((((uint32_t)(val * 8.0))) << A2XX_PA_SU_POINT_MINMAX_MIN__SHIFT) & A2XX_PA_SU_POINT_MINMAX_MIN__MASK;
+	return ((((uint32_t)(val * 16.0))) << A2XX_PA_SU_POINT_MINMAX_MIN__SHIFT) & A2XX_PA_SU_POINT_MINMAX_MIN__MASK;
 }
 #define A2XX_PA_SU_POINT_MINMAX_MAX__MASK			0xffff0000
 #define A2XX_PA_SU_POINT_MINMAX_MAX__SHIFT			16
 static inline uint32_t A2XX_PA_SU_POINT_MINMAX_MAX(float val)
 {
-	return ((((uint32_t)(val * 8.0))) << A2XX_PA_SU_POINT_MINMAX_MAX__SHIFT) & A2XX_PA_SU_POINT_MINMAX_MAX__MASK;
+	return ((((uint32_t)(val * 16.0))) << A2XX_PA_SU_POINT_MINMAX_MAX__SHIFT) & A2XX_PA_SU_POINT_MINMAX_MAX__MASK;
 }
 
 #define REG_A2XX_PA_SU_LINE_CNTL				0x00002282
@@ -1271,7 +1271,7 @@ static inline uint32_t A2XX_PA_SU_POINT_MINMAX_MAX(float val)
 #define A2XX_PA_SU_LINE_CNTL_WIDTH__SHIFT			0
 static inline uint32_t A2XX_PA_SU_LINE_CNTL_WIDTH(float val)
 {
-	return ((((uint32_t)(val * 8.0))) << A2XX_PA_SU_LINE_CNTL_WIDTH__SHIFT) & A2XX_PA_SU_LINE_CNTL_WIDTH__MASK;
+	return ((((uint32_t)(val * 16.0))) << A2XX_PA_SU_LINE_CNTL_WIDTH__SHIFT) & A2XX_PA_SU_LINE_CNTL_WIDTH__MASK;
 }
 
 #define REG_A2XX_PA_SC_LINE_STIPPLE				0x00002283
diff --git a/drivers/gpu/drm/msm/adreno/a3xx.xml.h b/drivers/gpu/drm/msm/adreno/a3xx.xml.h
index 303e8a9e91a5..109e9a263daf 100644
--- a/drivers/gpu/drm/msm/adreno/a3xx.xml.h
+++ b/drivers/gpu/drm/msm/adreno/a3xx.xml.h
@@ -11,10 +11,10 @@ The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/adreno.xml               (    364 bytes, from 2013-11-30 14:47:15)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml  (   1453 bytes, from 2013-03-31 16:51:27)
 - /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml          (  32901 bytes, from 2014-06-02 15:21:30)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (   9859 bytes, from 2014-06-02 15:21:30)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  14477 bytes, from 2014-05-16 11:51:57)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  58020 bytes, from 2014-06-25 12:57:16)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (  26602 bytes, from 2014-06-25 12:57:16)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (  10551 bytes, from 2014-11-13 22:44:30)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  15053 bytes, from 2014-11-09 15:45:47)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  63169 bytes, from 2014-11-13 22:44:18)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (  49097 bytes, from 2014-11-14 15:38:00)
 
 Copyright (C) 2013-2014 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -86,6 +86,14 @@ enum a3xx_vtx_fmt {
 	VFMT_NORM_USHORT_16_16 = 29,
 	VFMT_NORM_USHORT_16_16_16 = 30,
 	VFMT_NORM_USHORT_16_16_16_16 = 31,
+	VFMT_UINT_32 = 32,
+	VFMT_UINT_32_32 = 33,
+	VFMT_UINT_32_32_32 = 34,
+	VFMT_UINT_32_32_32_32 = 35,
+	VFMT_INT_32 = 36,
+	VFMT_INT_32_32 = 37,
+	VFMT_INT_32_32_32 = 38,
+	VFMT_INT_32_32_32_32 = 39,
 	VFMT_UBYTE_8 = 40,
 	VFMT_UBYTE_8_8 = 41,
 	VFMT_UBYTE_8_8_8 = 42,
@@ -112,7 +120,9 @@ enum a3xx_tex_fmt {
 	TFMT_NORM_USHORT_565 = 4,
 	TFMT_NORM_USHORT_5551 = 6,
 	TFMT_NORM_USHORT_4444 = 7,
+	TFMT_NORM_USHORT_Z16 = 9,
 	TFMT_NORM_UINT_X8Z24 = 10,
+	TFMT_FLOAT_Z32 = 11,
 	TFMT_NORM_UINT_NV12_UV_TILED = 17,
 	TFMT_NORM_UINT_NV12_Y_TILED = 19,
 	TFMT_NORM_UINT_NV12_UV = 21,
@@ -121,18 +131,38 @@ enum a3xx_tex_fmt {
 	TFMT_NORM_UINT_I420_U = 26,
 	TFMT_NORM_UINT_I420_V = 27,
 	TFMT_NORM_UINT_2_10_10_10 = 41,
+	TFMT_FLOAT_9_9_9_E5 = 42,
+	TFMT_FLOAT_10_11_11 = 43,
 	TFMT_NORM_UINT_A8 = 44,
 	TFMT_NORM_UINT_L8_A8 = 47,
 	TFMT_NORM_UINT_8 = 48,
 	TFMT_NORM_UINT_8_8 = 49,
 	TFMT_NORM_UINT_8_8_8 = 50,
 	TFMT_NORM_UINT_8_8_8_8 = 51,
+	TFMT_NORM_SINT_8_8 = 53,
+	TFMT_NORM_SINT_8_8_8_8 = 55,
+	TFMT_UINT_8_8 = 57,
+	TFMT_UINT_8_8_8_8 = 59,
+	TFMT_SINT_8_8 = 61,
+	TFMT_SINT_8_8_8_8 = 63,
 	TFMT_FLOAT_16 = 64,
 	TFMT_FLOAT_16_16 = 65,
 	TFMT_FLOAT_16_16_16_16 = 67,
+	TFMT_UINT_16 = 68,
+	TFMT_UINT_16_16 = 69,
+	TFMT_UINT_16_16_16_16 = 71,
+	TFMT_SINT_16 = 72,
+	TFMT_SINT_16_16 = 73,
+	TFMT_SINT_16_16_16_16 = 75,
 	TFMT_FLOAT_32 = 84,
 	TFMT_FLOAT_32_32 = 85,
 	TFMT_FLOAT_32_32_32_32 = 87,
+	TFMT_UINT_32 = 88,
+	TFMT_UINT_32_32 = 89,
+	TFMT_UINT_32_32_32_32 = 91,
+	TFMT_SINT_32 = 92,
+	TFMT_SINT_32_32 = 93,
+	TFMT_SINT_32_32_32_32 = 95,
 };
 
 enum a3xx_tex_fetchsize {
@@ -145,19 +175,34 @@ enum a3xx_tex_fetchsize {
 };
 
 enum a3xx_color_fmt {
+	RB_R5G6B5_UNORM = 0,
+	RB_R5G5B5A1_UNORM = 1,
+	RB_R4G4B4A4_UNORM = 3,
 	RB_R8G8B8_UNORM = 4,
 	RB_R8G8B8A8_UNORM = 8,
-	RB_Z16_UNORM = 12,
+	RB_R8G8B8A8_UINT = 10,
+	RB_R8G8B8A8_SINT = 11,
+	RB_R8G8_UNORM = 12,
+	RB_R8_UINT = 14,
+	RB_R8_SINT = 15,
+	RB_R10G10B10A2_UNORM = 16,
 	RB_A8_UNORM = 20,
+	RB_R8_UNORM = 21,
 	RB_R16G16B16A16_FLOAT = 27,
+	RB_R11G11B10_FLOAT = 28,
+	RB_R16_SINT = 40,
+	RB_R16G16_SINT = 41,
+	RB_R16G16B16A16_SINT = 43,
+	RB_R16_UINT = 44,
+	RB_R16G16_UINT = 45,
+	RB_R16G16B16A16_UINT = 47,
 	RB_R32G32B32A32_FLOAT = 51,
-};
-
-enum a3xx_color_swap {
-	WZYX = 0,
-	WXYZ = 1,
-	ZYXW = 2,
-	XYZW = 3,
+	RB_R32_SINT = 52,
+	RB_R32G32_SINT = 53,
+	RB_R32G32B32A32_SINT = 55,
+	RB_R32_UINT = 56,
+	RB_R32G32_UINT = 57,
+	RB_R32G32B32A32_UINT = 59,
 };
 
 enum a3xx_sp_perfcounter_select {
@@ -194,6 +239,11 @@ enum a3xx_rb_blend_opcode {
 	BLEND_MAX_DST_SRC = 4,
 };
 
+enum a3xx_intp_mode {
+	SMOOTH = 0,
+	FLAT = 1,
+};
+
 enum a3xx_tex_filter {
 	A3XX_TEX_NEAREST = 0,
 	A3XX_TEX_LINEAR = 1,
@@ -536,6 +586,10 @@ enum a3xx_tex_type {
 
 #define REG_A3XX_CP_MEQ_DATA					0x000001db
 
+#define REG_A3XX_CP_WFI_PEND_CTR				0x000001f5
+
+#define REG_A3XX_RBBM_PM_OVERRIDE2				0x0000039d
+
 #define REG_A3XX_CP_PERFCOUNTER_SELECT				0x00000445
 
 #define REG_A3XX_CP_HW_FAULT					0x0000045c
@@ -550,6 +604,12 @@ static inline uint32_t REG_A3XX_CP_PROTECT_REG(uint32_t i0) { return 0x00000460
 
 #define REG_A3XX_CP_AHB_FAULT					0x0000054d
 
+#define REG_A3XX_SQ_GPR_MANAGEMENT				0x00000d00
+
+#define REG_A3XX_SQ_INST_STORE_MANAGMENT			0x00000d02
+
+#define REG_A3XX_TP0_CHICKEN					0x00000e1e
+
 #define REG_A3XX_SP_GLOBAL_MEM_SIZE				0x00000e22
 
 #define REG_A3XX_SP_GLOBAL_MEM_ADDR				0x00000e23
@@ -632,13 +692,13 @@ static inline uint32_t A3XX_GRAS_CL_VPORT_ZSCALE(float val)
 #define A3XX_GRAS_SU_POINT_MINMAX_MIN__SHIFT			0
 static inline uint32_t A3XX_GRAS_SU_POINT_MINMAX_MIN(float val)
 {
-	return ((((uint32_t)(val * 8.0))) << A3XX_GRAS_SU_POINT_MINMAX_MIN__SHIFT) & A3XX_GRAS_SU_POINT_MINMAX_MIN__MASK;
+	return ((((uint32_t)(val * 16.0))) << A3XX_GRAS_SU_POINT_MINMAX_MIN__SHIFT) & A3XX_GRAS_SU_POINT_MINMAX_MIN__MASK;
 }
 #define A3XX_GRAS_SU_POINT_MINMAX_MAX__MASK			0xffff0000
 #define A3XX_GRAS_SU_POINT_MINMAX_MAX__SHIFT			16
 static inline uint32_t A3XX_GRAS_SU_POINT_MINMAX_MAX(float val)
 {
-	return ((((uint32_t)(val * 8.0))) << A3XX_GRAS_SU_POINT_MINMAX_MAX__SHIFT) & A3XX_GRAS_SU_POINT_MINMAX_MAX__MASK;
+	return ((((uint32_t)(val * 16.0))) << A3XX_GRAS_SU_POINT_MINMAX_MAX__SHIFT) & A3XX_GRAS_SU_POINT_MINMAX_MAX__MASK;
 }
 
 #define REG_A3XX_GRAS_SU_POINT_SIZE				0x00002069
@@ -646,7 +706,7 @@ static inline uint32_t A3XX_GRAS_SU_POINT_MINMAX_MAX(float val)
 #define A3XX_GRAS_SU_POINT_SIZE__SHIFT				0
 static inline uint32_t A3XX_GRAS_SU_POINT_SIZE(float val)
 {
-	return ((((uint32_t)(val * 8.0))) << A3XX_GRAS_SU_POINT_SIZE__SHIFT) & A3XX_GRAS_SU_POINT_SIZE__MASK;
+	return ((((int32_t)(val * 16.0))) << A3XX_GRAS_SU_POINT_SIZE__SHIFT) & A3XX_GRAS_SU_POINT_SIZE__MASK;
 }
 
 #define REG_A3XX_GRAS_SU_POLY_OFFSET_SCALE			0x0000206c
@@ -654,7 +714,7 @@ static inline uint32_t A3XX_GRAS_SU_POINT_SIZE(float val)
 #define A3XX_GRAS_SU_POLY_OFFSET_SCALE_VAL__SHIFT		0
 static inline uint32_t A3XX_GRAS_SU_POLY_OFFSET_SCALE_VAL(float val)
 {
-	return ((((uint32_t)(val * 40.0))) << A3XX_GRAS_SU_POLY_OFFSET_SCALE_VAL__SHIFT) & A3XX_GRAS_SU_POLY_OFFSET_SCALE_VAL__MASK;
+	return ((((int32_t)(val * 16384.0))) << A3XX_GRAS_SU_POLY_OFFSET_SCALE_VAL__SHIFT) & A3XX_GRAS_SU_POLY_OFFSET_SCALE_VAL__MASK;
 }
 
 #define REG_A3XX_GRAS_SU_POLY_OFFSET_OFFSET			0x0000206d
@@ -662,7 +722,7 @@ static inline uint32_t A3XX_GRAS_SU_POLY_OFFSET_SCALE_VAL(float val)
 #define A3XX_GRAS_SU_POLY_OFFSET_OFFSET__SHIFT			0
 static inline uint32_t A3XX_GRAS_SU_POLY_OFFSET_OFFSET(float val)
 {
-	return ((((uint32_t)(val * 44.0))) << A3XX_GRAS_SU_POLY_OFFSET_OFFSET__SHIFT) & A3XX_GRAS_SU_POLY_OFFSET_OFFSET__MASK;
+	return ((((int32_t)(val * 16384.0))) << A3XX_GRAS_SU_POLY_OFFSET_OFFSET__SHIFT) & A3XX_GRAS_SU_POLY_OFFSET_OFFSET__MASK;
 }
 
 #define REG_A3XX_GRAS_SU_MODE_CONTROL				0x00002070
@@ -673,7 +733,7 @@ static inline uint32_t A3XX_GRAS_SU_POLY_OFFSET_OFFSET(float val)
 #define A3XX_GRAS_SU_MODE_CONTROL_LINEHALFWIDTH__SHIFT		3
 static inline uint32_t A3XX_GRAS_SU_MODE_CONTROL_LINEHALFWIDTH(float val)
 {
-	return ((((uint32_t)(val * 4.0))) << A3XX_GRAS_SU_MODE_CONTROL_LINEHALFWIDTH__SHIFT) & A3XX_GRAS_SU_MODE_CONTROL_LINEHALFWIDTH__MASK;
+	return ((((int32_t)(val * 4.0))) << A3XX_GRAS_SU_MODE_CONTROL_LINEHALFWIDTH__SHIFT) & A3XX_GRAS_SU_MODE_CONTROL_LINEHALFWIDTH__MASK;
 }
 #define A3XX_GRAS_SU_MODE_CONTROL_POLY_OFFSET			0x00000800
 
@@ -863,6 +923,7 @@ static inline uint32_t A3XX_RB_MRT_BUF_INFO_COLOR_SWAP(enum a3xx_color_swap val)
 {
 	return ((val) << A3XX_RB_MRT_BUF_INFO_COLOR_SWAP__SHIFT) & A3XX_RB_MRT_BUF_INFO_COLOR_SWAP__MASK;
 }
+#define A3XX_RB_MRT_BUF_INFO_COLOR_SRGB				0x00004000
 #define A3XX_RB_MRT_BUF_INFO_COLOR_BUF_PITCH__MASK		0xfffe0000
 #define A3XX_RB_MRT_BUF_INFO_COLOR_BUF_PITCH__SHIFT		17
 static inline uint32_t A3XX_RB_MRT_BUF_INFO_COLOR_BUF_PITCH(uint32_t val)
@@ -1001,6 +1062,7 @@ static inline uint32_t A3XX_RB_COPY_CONTROL_FASTCLEAR(uint32_t val)
 {
 	return ((val) << A3XX_RB_COPY_CONTROL_FASTCLEAR__SHIFT) & A3XX_RB_COPY_CONTROL_FASTCLEAR__MASK;
 }
+#define A3XX_RB_COPY_CONTROL_UNK12				0x00001000
 #define A3XX_RB_COPY_CONTROL_GMEM_BASE__MASK			0xffffc000
 #define A3XX_RB_COPY_CONTROL_GMEM_BASE__SHIFT			14
 static inline uint32_t A3XX_RB_COPY_CONTROL_GMEM_BASE(uint32_t val)
@@ -1079,7 +1141,7 @@ static inline uint32_t A3XX_RB_DEPTH_CONTROL_ZFUNC(enum adreno_compare_func val)
 #define REG_A3XX_RB_DEPTH_CLEAR					0x00002101
 
 #define REG_A3XX_RB_DEPTH_INFO					0x00002102
-#define A3XX_RB_DEPTH_INFO_DEPTH_FORMAT__MASK			0x00000001
+#define A3XX_RB_DEPTH_INFO_DEPTH_FORMAT__MASK			0x00000003
 #define A3XX_RB_DEPTH_INFO_DEPTH_FORMAT__SHIFT			0
 static inline uint32_t A3XX_RB_DEPTH_INFO_DEPTH_FORMAT(enum adreno_rb_depth_format val)
 {
@@ -1265,6 +1327,7 @@ static inline uint32_t A3XX_PC_PRIM_VTX_CNTL_POLYMODE_BACK_PTYPE(enum adreno_pa_
 {
 	return ((val) << A3XX_PC_PRIM_VTX_CNTL_POLYMODE_BACK_PTYPE__SHIFT) & A3XX_PC_PRIM_VTX_CNTL_POLYMODE_BACK_PTYPE__MASK;
 }
+#define A3XX_PC_PRIM_VTX_CNTL_PRIMITIVE_RESTART			0x00100000
 #define A3XX_PC_PRIM_VTX_CNTL_PROVOKING_VTX_LAST		0x02000000
 #define A3XX_PC_PRIM_VTX_CNTL_PSIZE				0x04000000
 
@@ -1281,7 +1344,12 @@ static inline uint32_t A3XX_HLSQ_CONTROL_0_REG_FSTHREADSIZE(enum a3xx_threadsize
 #define A3XX_HLSQ_CONTROL_0_REG_SPSHADERRESTART			0x00000200
 #define A3XX_HLSQ_CONTROL_0_REG_RESERVED2			0x00000400
 #define A3XX_HLSQ_CONTROL_0_REG_CHUNKDISABLE			0x04000000
-#define A3XX_HLSQ_CONTROL_0_REG_CONSTSWITCHMODE			0x08000000
+#define A3XX_HLSQ_CONTROL_0_REG_CONSTMODE__MASK			0x08000000
+#define A3XX_HLSQ_CONTROL_0_REG_CONSTMODE__SHIFT		27
+static inline uint32_t A3XX_HLSQ_CONTROL_0_REG_CONSTMODE(uint32_t val)
+{
+	return ((val) << A3XX_HLSQ_CONTROL_0_REG_CONSTMODE__SHIFT) & A3XX_HLSQ_CONTROL_0_REG_CONSTMODE__MASK;
+}
 #define A3XX_HLSQ_CONTROL_0_REG_LAZYUPDATEDISABLE		0x10000000
 #define A3XX_HLSQ_CONTROL_0_REG_SPCONSTFULLUPDATE		0x20000000
 #define A3XX_HLSQ_CONTROL_0_REG_TPFULLUPDATE			0x40000000
@@ -1484,6 +1552,8 @@ static inline uint32_t A3XX_VFD_CONTROL_1_REGID4INST(uint32_t val)
 
 #define REG_A3XX_VFD_INDEX_OFFSET				0x00002245
 
+#define REG_A3XX_VFD_INDEX_OFFSET				0x00002245
+
 static inline uint32_t REG_A3XX_VFD_FETCH(uint32_t i0) { return 0x00002246 + 0x2*i0; }
 
 static inline uint32_t REG_A3XX_VFD_FETCH_INSTR_0(uint32_t i0) { return 0x00002246 + 0x2*i0; }
@@ -1537,6 +1607,7 @@ static inline uint32_t A3XX_VFD_DECODE_INSTR_REGID(uint32_t val)
 {
 	return ((val) << A3XX_VFD_DECODE_INSTR_REGID__SHIFT) & A3XX_VFD_DECODE_INSTR_REGID__MASK;
 }
+#define A3XX_VFD_DECODE_INSTR_INT				0x00100000
 #define A3XX_VFD_DECODE_INSTR_SWAP__MASK			0x00c00000
 #define A3XX_VFD_DECODE_INSTR_SWAP__SHIFT			22
 static inline uint32_t A3XX_VFD_DECODE_INSTR_SWAP(enum a3xx_color_swap val)
@@ -1604,6 +1675,102 @@ static inline uint32_t A3XX_VPC_PACK_NUMNONPOSVSVAR(uint32_t val)
 static inline uint32_t REG_A3XX_VPC_VARYING_INTERP(uint32_t i0) { return 0x00002282 + 0x1*i0; }
 
 static inline uint32_t REG_A3XX_VPC_VARYING_INTERP_MODE(uint32_t i0) { return 0x00002282 + 0x1*i0; }
+#define A3XX_VPC_VARYING_INTERP_MODE_C0__MASK			0x00000003
+#define A3XX_VPC_VARYING_INTERP_MODE_C0__SHIFT			0
+static inline uint32_t A3XX_VPC_VARYING_INTERP_MODE_C0(enum a3xx_intp_mode val)
+{
+	return ((val) << A3XX_VPC_VARYING_INTERP_MODE_C0__SHIFT) & A3XX_VPC_VARYING_INTERP_MODE_C0__MASK;
+}
+#define A3XX_VPC_VARYING_INTERP_MODE_C1__MASK			0x0000000c
+#define A3XX_VPC_VARYING_INTERP_MODE_C1__SHIFT			2
+static inline uint32_t A3XX_VPC_VARYING_INTERP_MODE_C1(enum a3xx_intp_mode val)
+{
+	return ((val) << A3XX_VPC_VARYING_INTERP_MODE_C1__SHIFT) & A3XX_VPC_VARYING_INTERP_MODE_C1__MASK;
+}
+#define A3XX_VPC_VARYING_INTERP_MODE_C2__MASK			0x00000030
+#define A3XX_VPC_VARYING_INTERP_MODE_C2__SHIFT			4
+static inline uint32_t A3XX_VPC_VARYING_INTERP_MODE_C2(enum a3xx_intp_mode val)
+{
+	return ((val) << A3XX_VPC_VARYING_INTERP_MODE_C2__SHIFT) & A3XX_VPC_VARYING_INTERP_MODE_C2__MASK;
+}
+#define A3XX_VPC_VARYING_INTERP_MODE_C3__MASK			0x000000c0
+#define A3XX_VPC_VARYING_INTERP_MODE_C3__SHIFT			6
+static inline uint32_t A3XX_VPC_VARYING_INTERP_MODE_C3(enum a3xx_intp_mode val)
+{
+	return ((val) << A3XX_VPC_VARYING_INTERP_MODE_C3__SHIFT) & A3XX_VPC_VARYING_INTERP_MODE_C3__MASK;
+}
+#define A3XX_VPC_VARYING_INTERP_MODE_C4__MASK			0x00000300
+#define A3XX_VPC_VARYING_INTERP_MODE_C4__SHIFT			8
+static inline uint32_t A3XX_VPC_VARYING_INTERP_MODE_C4(enum a3xx_intp_mode val)
+{
+	return ((val) << A3XX_VPC_VARYING_INTERP_MODE_C4__SHIFT) & A3XX_VPC_VARYING_INTERP_MODE_C4__MASK;
+}
+#define A3XX_VPC_VARYING_INTERP_MODE_C5__MASK			0x00000c00
+#define A3XX_VPC_VARYING_INTERP_MODE_C5__SHIFT			10
+static inline uint32_t A3XX_VPC_VARYING_INTERP_MODE_C5(enum a3xx_intp_mode val)
+{
+	return ((val) << A3XX_VPC_VARYING_INTERP_MODE_C5__SHIFT) & A3XX_VPC_VARYING_INTERP_MODE_C5__MASK;
+}
+#define A3XX_VPC_VARYING_INTERP_MODE_C6__MASK			0x00003000
+#define A3XX_VPC_VARYING_INTERP_MODE_C6__SHIFT			12
+static inline uint32_t A3XX_VPC_VARYING_INTERP_MODE_C6(enum a3xx_intp_mode val)
+{
+	return ((val) << A3XX_VPC_VARYING_INTERP_MODE_C6__SHIFT) & A3XX_VPC_VARYING_INTERP_MODE_C6__MASK;
+}
+#define A3XX_VPC_VARYING_INTERP_MODE_C7__MASK			0x0000c000
+#define A3XX_VPC_VARYING_INTERP_MODE_C7__SHIFT			14
+static inline uint32_t A3XX_VPC_VARYING_INTERP_MODE_C7(enum a3xx_intp_mode val)
+{
+	return ((val) << A3XX_VPC_VARYING_INTERP_MODE_C7__SHIFT) & A3XX_VPC_VARYING_INTERP_MODE_C7__MASK;
+}
+#define A3XX_VPC_VARYING_INTERP_MODE_C8__MASK			0x00030000
+#define A3XX_VPC_VARYING_INTERP_MODE_C8__SHIFT			16
+static inline uint32_t A3XX_VPC_VARYING_INTERP_MODE_C8(enum a3xx_intp_mode val)
+{
+	return ((val) << A3XX_VPC_VARYING_INTERP_MODE_C8__SHIFT) & A3XX_VPC_VARYING_INTERP_MODE_C8__MASK;
+}
+#define A3XX_VPC_VARYING_INTERP_MODE_C9__MASK			0x000c0000
+#define A3XX_VPC_VARYING_INTERP_MODE_C9__SHIFT			18
+static inline uint32_t A3XX_VPC_VARYING_INTERP_MODE_C9(enum a3xx_intp_mode val)
+{
+	return ((val) << A3XX_VPC_VARYING_INTERP_MODE_C9__SHIFT) & A3XX_VPC_VARYING_INTERP_MODE_C9__MASK;
+}
+#define A3XX_VPC_VARYING_INTERP_MODE_CA__MASK			0x00300000
+#define A3XX_VPC_VARYING_INTERP_MODE_CA__SHIFT			20
+static inline uint32_t A3XX_VPC_VARYING_INTERP_MODE_CA(enum a3xx_intp_mode val)
+{
+	return ((val) << A3XX_VPC_VARYING_INTERP_MODE_CA__SHIFT) & A3XX_VPC_VARYING_INTERP_MODE_CA__MASK;
+}
+#define A3XX_VPC_VARYING_INTERP_MODE_CB__MASK			0x00c00000
+#define A3XX_VPC_VARYING_INTERP_MODE_CB__SHIFT			22
+static inline uint32_t A3XX_VPC_VARYING_INTERP_MODE_CB(enum a3xx_intp_mode val)
+{
+	return ((val) << A3XX_VPC_VARYING_INTERP_MODE_CB__SHIFT) & A3XX_VPC_VARYING_INTERP_MODE_CB__MASK;
+}
+#define A3XX_VPC_VARYING_INTERP_MODE_CC__MASK			0x03000000
+#define A3XX_VPC_VARYING_INTERP_MODE_CC__SHIFT			24
+static inline uint32_t A3XX_VPC_VARYING_INTERP_MODE_CC(enum a3xx_intp_mode val)
+{
+	return ((val) << A3XX_VPC_VARYING_INTERP_MODE_CC__SHIFT) & A3XX_VPC_VARYING_INTERP_MODE_CC__MASK;
+}
+#define A3XX_VPC_VARYING_INTERP_MODE_CD__MASK			0x0c000000
+#define A3XX_VPC_VARYING_INTERP_MODE_CD__SHIFT			26
+static inline uint32_t A3XX_VPC_VARYING_INTERP_MODE_CD(enum a3xx_intp_mode val)
+{
+	return ((val) << A3XX_VPC_VARYING_INTERP_MODE_CD__SHIFT) & A3XX_VPC_VARYING_INTERP_MODE_CD__MASK;
+}
+#define A3XX_VPC_VARYING_INTERP_MODE_CE__MASK			0x30000000
+#define A3XX_VPC_VARYING_INTERP_MODE_CE__SHIFT			28
+static inline uint32_t A3XX_VPC_VARYING_INTERP_MODE_CE(enum a3xx_intp_mode val)
+{
+	return ((val) << A3XX_VPC_VARYING_INTERP_MODE_CE__SHIFT) & A3XX_VPC_VARYING_INTERP_MODE_CE__MASK;
+}
+#define A3XX_VPC_VARYING_INTERP_MODE_CF__MASK			0xc0000000
+#define A3XX_VPC_VARYING_INTERP_MODE_CF__SHIFT			30
+static inline uint32_t A3XX_VPC_VARYING_INTERP_MODE_CF(enum a3xx_intp_mode val)
+{
+	return ((val) << A3XX_VPC_VARYING_INTERP_MODE_CF__SHIFT) & A3XX_VPC_VARYING_INTERP_MODE_CF__MASK;
+}
 
 static inline uint32_t REG_A3XX_VPC_VARYING_PS_REPL(uint32_t i0) { return 0x00002286 + 0x1*i0; }
 
@@ -1696,7 +1863,7 @@ static inline uint32_t A3XX_SP_VS_CTRL_REG1_CONSTFOOTPRINT(uint32_t val)
 {
 	return ((val) << A3XX_SP_VS_CTRL_REG1_CONSTFOOTPRINT__SHIFT) & A3XX_SP_VS_CTRL_REG1_CONSTFOOTPRINT__MASK;
 }
-#define A3XX_SP_VS_CTRL_REG1_INITIALOUTSTANDING__MASK		0x3f000000
+#define A3XX_SP_VS_CTRL_REG1_INITIALOUTSTANDING__MASK		0x7f000000
 #define A3XX_SP_VS_CTRL_REG1_INITIALOUTSTANDING__SHIFT		24
 static inline uint32_t A3XX_SP_VS_CTRL_REG1_INITIALOUTSTANDING(uint32_t val)
 {
@@ -1928,6 +2095,8 @@ static inline uint32_t A3XX_SP_FS_MRT_REG_REGID(uint32_t val)
 	return ((val) << A3XX_SP_FS_MRT_REG_REGID__SHIFT) & A3XX_SP_FS_MRT_REG_REGID__MASK;
 }
 #define A3XX_SP_FS_MRT_REG_HALF_PRECISION			0x00000100
+#define A3XX_SP_FS_MRT_REG_SINT					0x00000400
+#define A3XX_SP_FS_MRT_REG_UINT					0x00000800
 
 static inline uint32_t REG_A3XX_SP_FS_IMAGE_OUTPUT(uint32_t i0) { return 0x000022f4 + 0x1*i0; }
 
@@ -1947,6 +2116,8 @@ static inline uint32_t A3XX_SP_FS_LENGTH_REG_SHADERLENGTH(uint32_t val)
 	return ((val) << A3XX_SP_FS_LENGTH_REG_SHADERLENGTH__SHIFT) & A3XX_SP_FS_LENGTH_REG_SHADERLENGTH__MASK;
 }
 
+#define REG_A3XX_PA_SC_AA_CONFIG				0x00002301
+
 #define REG_A3XX_TPL1_TP_VS_TEX_OFFSET				0x00002340
 #define A3XX_TPL1_TP_VS_TEX_OFFSET_SAMPLEROFFSET__MASK		0x000000ff
 #define A3XX_TPL1_TP_VS_TEX_OFFSET_SAMPLEROFFSET__SHIFT		0
@@ -2297,11 +2468,11 @@ static inline uint32_t A3XX_VGT_DRAW_INITIATOR_INDEX_SIZE(enum pc_di_index_size
 #define A3XX_VGT_DRAW_INITIATOR_NOT_EOP				0x00001000
 #define A3XX_VGT_DRAW_INITIATOR_SMALL_INDEX			0x00002000
 #define A3XX_VGT_DRAW_INITIATOR_PRE_DRAW_INITIATOR_ENABLE	0x00004000
-#define A3XX_VGT_DRAW_INITIATOR_NUM_INDICES__MASK		0xffff0000
-#define A3XX_VGT_DRAW_INITIATOR_NUM_INDICES__SHIFT		16
-static inline uint32_t A3XX_VGT_DRAW_INITIATOR_NUM_INDICES(uint32_t val)
+#define A3XX_VGT_DRAW_INITIATOR_NUM_INSTANCES__MASK		0xff000000
+#define A3XX_VGT_DRAW_INITIATOR_NUM_INSTANCES__SHIFT		24
+static inline uint32_t A3XX_VGT_DRAW_INITIATOR_NUM_INSTANCES(uint32_t val)
 {
-	return ((val) << A3XX_VGT_DRAW_INITIATOR_NUM_INDICES__SHIFT) & A3XX_VGT_DRAW_INITIATOR_NUM_INDICES__MASK;
+	return ((val) << A3XX_VGT_DRAW_INITIATOR_NUM_INSTANCES__SHIFT) & A3XX_VGT_DRAW_INITIATOR_NUM_INSTANCES__MASK;
 }
 
 #define REG_A3XX_VGT_IMMED_DATA					0x000021fd
@@ -2347,17 +2518,23 @@ static inline uint32_t A3XX_TEX_SAMP_0_COMPARE_FUNC(enum adreno_compare_func val
 #define A3XX_TEX_SAMP_0_UNNORM_COORDS				0x80000000
 
 #define REG_A3XX_TEX_SAMP_1					0x00000001
+#define A3XX_TEX_SAMP_1_LOD_BIAS__MASK				0x000007ff
+#define A3XX_TEX_SAMP_1_LOD_BIAS__SHIFT				0
+static inline uint32_t A3XX_TEX_SAMP_1_LOD_BIAS(float val)
+{
+	return ((((int32_t)(val * 64.0))) << A3XX_TEX_SAMP_1_LOD_BIAS__SHIFT) & A3XX_TEX_SAMP_1_LOD_BIAS__MASK;
+}
 #define A3XX_TEX_SAMP_1_MAX_LOD__MASK				0x003ff000
 #define A3XX_TEX_SAMP_1_MAX_LOD__SHIFT				12
 static inline uint32_t A3XX_TEX_SAMP_1_MAX_LOD(float val)
 {
-	return ((((uint32_t)(val * 12.0))) << A3XX_TEX_SAMP_1_MAX_LOD__SHIFT) & A3XX_TEX_SAMP_1_MAX_LOD__MASK;
+	return ((((uint32_t)(val * 64.0))) << A3XX_TEX_SAMP_1_MAX_LOD__SHIFT) & A3XX_TEX_SAMP_1_MAX_LOD__MASK;
 }
 #define A3XX_TEX_SAMP_1_MIN_LOD__MASK				0xffc00000
 #define A3XX_TEX_SAMP_1_MIN_LOD__SHIFT				22
 static inline uint32_t A3XX_TEX_SAMP_1_MIN_LOD(float val)
 {
-	return ((((uint32_t)(val * 12.0))) << A3XX_TEX_SAMP_1_MIN_LOD__SHIFT) & A3XX_TEX_SAMP_1_MIN_LOD__MASK;
+	return ((((uint32_t)(val * 64.0))) << A3XX_TEX_SAMP_1_MIN_LOD__SHIFT) & A3XX_TEX_SAMP_1_MIN_LOD__MASK;
 }
 
 #define REG_A3XX_TEX_CONST_0					0x00000000
@@ -2448,6 +2625,24 @@ static inline uint32_t A3XX_TEX_CONST_2_SWAP(enum a3xx_color_swap val)
 }
 
 #define REG_A3XX_TEX_CONST_3					0x00000003
+#define A3XX_TEX_CONST_3_LAYERSZ1__MASK				0x0000000f
+#define A3XX_TEX_CONST_3_LAYERSZ1__SHIFT			0
+static inline uint32_t A3XX_TEX_CONST_3_LAYERSZ1(uint32_t val)
+{
+	return ((val >> 12) << A3XX_TEX_CONST_3_LAYERSZ1__SHIFT) & A3XX_TEX_CONST_3_LAYERSZ1__MASK;
+}
+#define A3XX_TEX_CONST_3_DEPTH__MASK				0x0ffe0000
+#define A3XX_TEX_CONST_3_DEPTH__SHIFT				17
+static inline uint32_t A3XX_TEX_CONST_3_DEPTH(uint32_t val)
+{
+	return ((val) << A3XX_TEX_CONST_3_DEPTH__SHIFT) & A3XX_TEX_CONST_3_DEPTH__MASK;
+}
+#define A3XX_TEX_CONST_3_LAYERSZ2__MASK				0xf0000000
+#define A3XX_TEX_CONST_3_LAYERSZ2__SHIFT			28
+static inline uint32_t A3XX_TEX_CONST_3_LAYERSZ2(uint32_t val)
+{
+	return ((val >> 12) << A3XX_TEX_CONST_3_LAYERSZ2__SHIFT) & A3XX_TEX_CONST_3_LAYERSZ2__MASK;
+}
 
 
 #endif /* A3XX_XML */
diff --git a/drivers/gpu/drm/msm/adreno/a3xx_gpu.c b/drivers/gpu/drm/msm/adreno/a3xx_gpu.c
index 2773600c9488..b66c53bdc039 100644
--- a/drivers/gpu/drm/msm/adreno/a3xx_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/a3xx_gpu.c
@@ -2,6 +2,8 @@
  * Copyright (C) 2013 Red Hat
  * Author: Rob Clark <robdclark@gmail.com>
  *
+ * Copyright (c) 2014 The Linux Foundation. All rights reserved.
+ *
  * This program is free software; you can redistribute it and/or modify it
  * under the terms of the GNU General Public License version 2 as published by
  * the Free Software Foundation.
@@ -35,10 +37,8 @@
 	 A3XX_INT0_CP_AHB_ERROR_HALT |     \
 	 A3XX_INT0_UCHE_OOB_ACCESS)
 
+extern bool hang_debug;
 
-static bool hang_debug = false;
-MODULE_PARM_DESC(hang_debug, "Dump registers when hang is detected (can be slow!)");
-module_param_named(hang_debug, hang_debug, bool, 0600);
 static void a3xx_dump(struct msm_gpu *gpu);
 
 static void a3xx_me_init(struct msm_gpu *gpu)
@@ -387,59 +387,115 @@ static const unsigned int a3xx_registers[] = {
 	0x2750, 0x2756, 0x2760, 0x2760, 0x300c, 0x300e, 0x301c, 0x301d,
 	0x302a, 0x302a, 0x302c, 0x302d, 0x3030, 0x3031, 0x3034, 0x3036,
 	0x303c, 0x303c, 0x305e, 0x305f,
+	~0   /* sentinel */
 };
 
 #ifdef CONFIG_DEBUG_FS
 static void a3xx_show(struct msm_gpu *gpu, struct seq_file *m)
 {
-	int i;
-
-	adreno_show(gpu, m);
-
 	gpu->funcs->pm_resume(gpu);
-
 	seq_printf(m, "status:   %08x\n",
 			gpu_read(gpu, REG_A3XX_RBBM_STATUS));
-
-	/* dump these out in a form that can be parsed by demsm: */
-	seq_printf(m, "IO:region %s 00000000 00020000\n", gpu->name);
-	for (i = 0; i < ARRAY_SIZE(a3xx_registers); i += 2) {
-		uint32_t start = a3xx_registers[i];
-		uint32_t end   = a3xx_registers[i+1];
-		uint32_t addr;
-
-		for (addr = start; addr <= end; addr++) {
-			uint32_t val = gpu_read(gpu, addr);
-			seq_printf(m, "IO:R %08x %08x\n", addr<<2, val);
-		}
-	}
-
 	gpu->funcs->pm_suspend(gpu);
+	adreno_show(gpu, m);
 }
 #endif
 
 /* would be nice to not have to duplicate the _show() stuff with printk(): */
 static void a3xx_dump(struct msm_gpu *gpu)
 {
-	int i;
-
-	adreno_dump(gpu);
 	printk("status:   %08x\n",
 			gpu_read(gpu, REG_A3XX_RBBM_STATUS));
-
-	/* dump these out in a form that can be parsed by demsm: */
-	printk("IO:region %s 00000000 00020000\n", gpu->name);
-	for (i = 0; i < ARRAY_SIZE(a3xx_registers); i += 2) {
-		uint32_t start = a3xx_registers[i];
-		uint32_t end   = a3xx_registers[i+1];
-		uint32_t addr;
-
-		for (addr = start; addr <= end; addr++) {
-			uint32_t val = gpu_read(gpu, addr);
-			printk("IO:R %08x %08x\n", addr<<2, val);
-		}
-	}
+	adreno_dump(gpu);
 }
+/* Register offset defines for A3XX */
+static const unsigned int a3xx_register_offsets[REG_ADRENO_REGISTER_MAX] = {
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_DEBUG, REG_AXXX_CP_DEBUG),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_ME_RAM_WADDR, REG_AXXX_CP_ME_RAM_WADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_ME_RAM_DATA, REG_AXXX_CP_ME_RAM_DATA),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_PFP_UCODE_DATA,
+			REG_A3XX_CP_PFP_UCODE_DATA),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_PFP_UCODE_ADDR,
+			REG_A3XX_CP_PFP_UCODE_ADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_WFI_PEND_CTR, REG_A3XX_CP_WFI_PEND_CTR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_BASE, REG_AXXX_CP_RB_BASE),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_RPTR_ADDR, REG_AXXX_CP_RB_RPTR_ADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_RPTR, REG_AXXX_CP_RB_RPTR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_WPTR, REG_AXXX_CP_RB_WPTR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_PROTECT_CTRL, REG_A3XX_CP_PROTECT_CTRL),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_ME_CNTL, REG_AXXX_CP_ME_CNTL),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_CNTL, REG_AXXX_CP_RB_CNTL),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_IB1_BASE, REG_AXXX_CP_IB1_BASE),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_IB1_BUFSZ, REG_AXXX_CP_IB1_BUFSZ),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_IB2_BASE, REG_AXXX_CP_IB2_BASE),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_IB2_BUFSZ, REG_AXXX_CP_IB2_BUFSZ),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_TIMESTAMP, REG_AXXX_CP_SCRATCH_REG0),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_ME_RAM_RADDR, REG_AXXX_CP_ME_RAM_RADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_SCRATCH_ADDR, REG_AXXX_SCRATCH_ADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_SCRATCH_UMSK, REG_AXXX_SCRATCH_UMSK),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_ROQ_ADDR, REG_A3XX_CP_ROQ_ADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_ROQ_DATA, REG_A3XX_CP_ROQ_DATA),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_MERCIU_ADDR, REG_A3XX_CP_MERCIU_ADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_MERCIU_DATA, REG_A3XX_CP_MERCIU_DATA),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_MERCIU_DATA2, REG_A3XX_CP_MERCIU_DATA2),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_MEQ_ADDR, REG_A3XX_CP_MEQ_ADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_MEQ_DATA, REG_A3XX_CP_MEQ_DATA),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_HW_FAULT, REG_A3XX_CP_HW_FAULT),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_PROTECT_STATUS,
+			REG_A3XX_CP_PROTECT_STATUS),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_STATUS, REG_A3XX_RBBM_STATUS),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_CTL,
+			REG_A3XX_RBBM_PERFCTR_CTL),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_CMD0,
+			REG_A3XX_RBBM_PERFCTR_LOAD_CMD0),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_CMD1,
+			REG_A3XX_RBBM_PERFCTR_LOAD_CMD1),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_PWR_1_LO,
+			REG_A3XX_RBBM_PERFCTR_PWR_1_LO),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_INT_0_MASK, REG_A3XX_RBBM_INT_0_MASK),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_INT_0_STATUS,
+			REG_A3XX_RBBM_INT_0_STATUS),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_AHB_ERROR_STATUS,
+			REG_A3XX_RBBM_AHB_ERROR_STATUS),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_AHB_CMD, REG_A3XX_RBBM_AHB_CMD),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_INT_CLEAR_CMD,
+			REG_A3XX_RBBM_INT_CLEAR_CMD),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_CLOCK_CTL, REG_A3XX_RBBM_CLOCK_CTL),
+	REG_ADRENO_DEFINE(REG_ADRENO_VPC_DEBUG_RAM_SEL,
+			REG_A3XX_VPC_VPC_DEBUG_RAM_SEL),
+	REG_ADRENO_DEFINE(REG_ADRENO_VPC_DEBUG_RAM_READ,
+			REG_A3XX_VPC_VPC_DEBUG_RAM_READ),
+	REG_ADRENO_DEFINE(REG_ADRENO_VSC_SIZE_ADDRESS,
+			REG_A3XX_VSC_SIZE_ADDRESS),
+	REG_ADRENO_DEFINE(REG_ADRENO_VFD_CONTROL_0, REG_A3XX_VFD_CONTROL_0),
+	REG_ADRENO_DEFINE(REG_ADRENO_VFD_INDEX_MAX, REG_A3XX_VFD_INDEX_MAX),
+	REG_ADRENO_DEFINE(REG_ADRENO_SP_VS_PVT_MEM_ADDR_REG,
+			REG_A3XX_SP_VS_PVT_MEM_ADDR_REG),
+	REG_ADRENO_DEFINE(REG_ADRENO_SP_FS_PVT_MEM_ADDR_REG,
+			REG_A3XX_SP_FS_PVT_MEM_ADDR_REG),
+	REG_ADRENO_DEFINE(REG_ADRENO_SP_VS_OBJ_START_REG,
+			REG_A3XX_SP_VS_OBJ_START_REG),
+	REG_ADRENO_DEFINE(REG_ADRENO_SP_FS_OBJ_START_REG,
+			REG_A3XX_SP_FS_OBJ_START_REG),
+	REG_ADRENO_DEFINE(REG_ADRENO_PA_SC_AA_CONFIG, REG_A3XX_PA_SC_AA_CONFIG),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PM_OVERRIDE2,
+			REG_A3XX_RBBM_PM_OVERRIDE2),
+	REG_ADRENO_DEFINE(REG_ADRENO_SCRATCH_REG2, REG_AXXX_CP_SCRATCH_REG2),
+	REG_ADRENO_DEFINE(REG_ADRENO_SQ_GPR_MANAGEMENT,
+			REG_A3XX_SQ_GPR_MANAGEMENT),
+	REG_ADRENO_DEFINE(REG_ADRENO_SQ_INST_STORE_MANAGMENT,
+			REG_A3XX_SQ_INST_STORE_MANAGMENT),
+	REG_ADRENO_DEFINE(REG_ADRENO_TP0_CHICKEN, REG_A3XX_TP0_CHICKEN),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_RBBM_CTL, REG_A3XX_RBBM_RBBM_CTL),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_SW_RESET_CMD,
+			REG_A3XX_RBBM_SW_RESET_CMD),
+	REG_ADRENO_DEFINE(REG_ADRENO_UCHE_INVALIDATE0,
+			REG_A3XX_UCHE_CACHE_INVALIDATE0_REG),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_VALUE_LO,
+			REG_A3XX_RBBM_PERFCTR_LOAD_VALUE_LO),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_VALUE_HI,
+			REG_A3XX_RBBM_PERFCTR_LOAD_VALUE_HI),
+};
 
 static const struct adreno_gpu_funcs funcs = {
 	.base = {
@@ -474,7 +530,6 @@ struct msm_gpu *a3xx_gpu_init(struct drm_device *dev)
 	struct msm_gpu *gpu;
 	struct msm_drm_private *priv = dev->dev_private;
 	struct platform_device *pdev = priv->gpu_pdev;
-	struct adreno_platform_config *config;
 	int ret;
 
 	if (!pdev) {
@@ -483,8 +538,6 @@ struct msm_gpu *a3xx_gpu_init(struct drm_device *dev)
 		goto fail;
 	}
 
-	config = pdev->dev.platform_data;
-
 	a3xx_gpu = kzalloc(sizeof(*a3xx_gpu), GFP_KERNEL);
 	if (!a3xx_gpu) {
 		ret = -ENOMEM;
@@ -496,20 +549,13 @@ struct msm_gpu *a3xx_gpu_init(struct drm_device *dev)
 
 	a3xx_gpu->pdev = pdev;
 
-	gpu->fast_rate = config->fast_rate;
-	gpu->slow_rate = config->slow_rate;
-	gpu->bus_freq  = config->bus_freq;
-#ifdef CONFIG_MSM_BUS_SCALING
-	gpu->bus_scale_table = config->bus_scale_table;
-#endif
-
-	DBG("fast_rate=%u, slow_rate=%u, bus_freq=%u",
-			gpu->fast_rate, gpu->slow_rate, gpu->bus_freq);
-
 	gpu->perfcntrs = perfcntrs;
 	gpu->num_perfcntrs = ARRAY_SIZE(perfcntrs);
 
-	ret = adreno_gpu_init(dev, pdev, adreno_gpu, &funcs, config->rev);
+	adreno_gpu->registers = a3xx_registers;
+	adreno_gpu->reg_offsets = a3xx_register_offsets;
+
+	ret = adreno_gpu_init(dev, pdev, adreno_gpu, &funcs);
 	if (ret)
 		goto fail;
 
@@ -549,158 +595,3 @@ fail:
 
 	return ERR_PTR(ret);
 }
-
-/*
- * The a3xx device:
- */
-
-#if defined(CONFIG_MSM_BUS_SCALING) && !defined(CONFIG_OF)
-#  include <mach/kgsl.h>
-#endif
-
-static void set_gpu_pdev(struct drm_device *dev,
-		struct platform_device *pdev)
-{
-	struct msm_drm_private *priv = dev->dev_private;
-	priv->gpu_pdev = pdev;
-}
-
-static int a3xx_bind(struct device *dev, struct device *master, void *data)
-{
-	static struct adreno_platform_config config = {};
-#ifdef CONFIG_OF
-	struct device_node *child, *node = dev->of_node;
-	u32 val;
-	int ret;
-
-	ret = of_property_read_u32(node, "qcom,chipid", &val);
-	if (ret) {
-		dev_err(dev, "could not find chipid: %d\n", ret);
-		return ret;
-	}
-
-	config.rev = ADRENO_REV((val >> 24) & 0xff,
-			(val >> 16) & 0xff, (val >> 8) & 0xff, val & 0xff);
-
-	/* find clock rates: */
-	config.fast_rate = 0;
-	config.slow_rate = ~0;
-	for_each_child_of_node(node, child) {
-		if (of_device_is_compatible(child, "qcom,gpu-pwrlevels")) {
-			struct device_node *pwrlvl;
-			for_each_child_of_node(child, pwrlvl) {
-				ret = of_property_read_u32(pwrlvl, "qcom,gpu-freq", &val);
-				if (ret) {
-					dev_err(dev, "could not find gpu-freq: %d\n", ret);
-					return ret;
-				}
-				config.fast_rate = max(config.fast_rate, val);
-				config.slow_rate = min(config.slow_rate, val);
-			}
-		}
-	}
-
-	if (!config.fast_rate) {
-		dev_err(dev, "could not find clk rates\n");
-		return -ENXIO;
-	}
-
-#else
-	struct kgsl_device_platform_data *pdata = dev->platform_data;
-	uint32_t version = socinfo_get_version();
-	if (cpu_is_apq8064ab()) {
-		config.fast_rate = 450000000;
-		config.slow_rate = 27000000;
-		config.bus_freq  = 4;
-		config.rev = ADRENO_REV(3, 2, 1, 0);
-	} else if (cpu_is_apq8064()) {
-		config.fast_rate = 400000000;
-		config.slow_rate = 27000000;
-		config.bus_freq  = 4;
-
-		if (SOCINFO_VERSION_MAJOR(version) == 2)
-			config.rev = ADRENO_REV(3, 2, 0, 2);
-		else if ((SOCINFO_VERSION_MAJOR(version) == 1) &&
-				(SOCINFO_VERSION_MINOR(version) == 1))
-			config.rev = ADRENO_REV(3, 2, 0, 1);
-		else
-			config.rev = ADRENO_REV(3, 2, 0, 0);
-
-	} else if (cpu_is_msm8960ab()) {
-		config.fast_rate = 400000000;
-		config.slow_rate = 320000000;
-		config.bus_freq  = 4;
-
-		if (SOCINFO_VERSION_MINOR(version) == 0)
-			config.rev = ADRENO_REV(3, 2, 1, 0);
-		else
-			config.rev = ADRENO_REV(3, 2, 1, 1);
-
-	} else if (cpu_is_msm8930()) {
-		config.fast_rate = 400000000;
-		config.slow_rate = 27000000;
-		config.bus_freq  = 3;
-
-		if ((SOCINFO_VERSION_MAJOR(version) == 1) &&
-			(SOCINFO_VERSION_MINOR(version) == 2))
-			config.rev = ADRENO_REV(3, 0, 5, 2);
-		else
-			config.rev = ADRENO_REV(3, 0, 5, 0);
-
-	}
-#  ifdef CONFIG_MSM_BUS_SCALING
-	config.bus_scale_table = pdata->bus_scale_table;
-#  endif
-#endif
-	dev->platform_data = &config;
-	set_gpu_pdev(dev_get_drvdata(master), to_platform_device(dev));
-	return 0;
-}
-
-static void a3xx_unbind(struct device *dev, struct device *master,
-		void *data)
-{
-	set_gpu_pdev(dev_get_drvdata(master), NULL);
-}
-
-static const struct component_ops a3xx_ops = {
-		.bind   = a3xx_bind,
-		.unbind = a3xx_unbind,
-};
-
-static int a3xx_probe(struct platform_device *pdev)
-{
-	return component_add(&pdev->dev, &a3xx_ops);
-}
-
-static int a3xx_remove(struct platform_device *pdev)
-{
-	component_del(&pdev->dev, &a3xx_ops);
-	return 0;
-}
-
-static const struct of_device_id dt_match[] = {
-	{ .compatible = "qcom,adreno-3xx" },
-	/* for backwards compat w/ downstream kgsl DT files: */
-	{ .compatible = "qcom,kgsl-3d0" },
-	{}
-};
-
-static struct platform_driver a3xx_driver = {
-	.probe = a3xx_probe,
-	.remove = a3xx_remove,
-	.driver = {
-		.name = "kgsl-3d0",
-		.of_match_table = dt_match,
-	},
-};
-
-void __init a3xx_register(void)
-{
-	platform_driver_register(&a3xx_driver);
-}
-
-void __exit a3xx_unregister(void)
-{
-	platform_driver_unregister(&a3xx_driver);
-}
diff --git a/drivers/gpu/drm/msm/adreno/a4xx.xml.h b/drivers/gpu/drm/msm/adreno/a4xx.xml.h
new file mode 100644
index 000000000000..5a24c416d2dd
--- /dev/null
+++ b/drivers/gpu/drm/msm/adreno/a4xx.xml.h
@@ -0,0 +1,2144 @@
+#ifndef A4XX_XML
+#define A4XX_XML
+
+/* Autogenerated file, DO NOT EDIT manually!
+
+This file was generated by the rules-ng-ng headergen tool in this git repository:
+http://github.com/freedreno/envytools/
+git clone https://github.com/freedreno/envytools.git
+
+The rules-ng-ng source files this header was generated from are:
+- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml               (    364 bytes, from 2013-11-30 14:47:15)
+- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml  (   1453 bytes, from 2013-03-31 16:51:27)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml          (  32901 bytes, from 2014-06-02 15:21:30)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (  10551 bytes, from 2014-11-13 22:44:30)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  15053 bytes, from 2014-11-09 15:45:47)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  63169 bytes, from 2014-11-13 22:44:18)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (  49097 bytes, from 2014-11-14 15:38:00)
+
+Copyright (C) 2013-2014 by the following authors:
+- Rob Clark <robdclark@gmail.com> (robclark)
+
+Permission is hereby granted, free of charge, to any person obtaining
+a copy of this software and associated documentation files (the
+"Software"), to deal in the Software without restriction, including
+without limitation the rights to use, copy, modify, merge, publish,
+distribute, sublicense, and/or sell copies of the Software, and to
+permit persons to whom the Software is furnished to do so, subject to
+the following conditions:
+
+The above copyright notice and this permission notice (including the
+next paragraph) shall be included in all copies or substantial
+portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+IN NO EVENT SHALL THE COPYRIGHT OWNER(S) AND/OR ITS SUPPLIERS BE
+LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
+OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
+WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+*/
+
+
+enum a4xx_color_fmt {
+	RB4_A8_UNORM = 1,
+	RB4_R5G6R5_UNORM = 14,
+	RB4_Z16_UNORM = 15,
+	RB4_R8G8B8_UNORM = 25,
+	RB4_R8G8B8A8_UNORM = 26,
+};
+
+enum a4xx_tile_mode {
+	TILE4_LINEAR = 0,
+	TILE4_3 = 3,
+};
+
+enum a4xx_rb_blend_opcode {
+	BLEND_DST_PLUS_SRC = 0,
+	BLEND_SRC_MINUS_DST = 1,
+	BLEND_DST_MINUS_SRC = 2,
+	BLEND_MIN_DST_SRC = 3,
+	BLEND_MAX_DST_SRC = 4,
+};
+
+enum a4xx_vtx_fmt {
+	VFMT4_FLOAT_32 = 1,
+	VFMT4_FLOAT_32_32 = 2,
+	VFMT4_FLOAT_32_32_32 = 3,
+	VFMT4_FLOAT_32_32_32_32 = 4,
+	VFMT4_FLOAT_16 = 5,
+	VFMT4_FLOAT_16_16 = 6,
+	VFMT4_FLOAT_16_16_16 = 7,
+	VFMT4_FLOAT_16_16_16_16 = 8,
+	VFMT4_FIXED_32 = 9,
+	VFMT4_FIXED_32_32 = 10,
+	VFMT4_FIXED_32_32_32 = 11,
+	VFMT4_FIXED_32_32_32_32 = 12,
+	VFMT4_SHORT_16 = 16,
+	VFMT4_SHORT_16_16 = 17,
+	VFMT4_SHORT_16_16_16 = 18,
+	VFMT4_SHORT_16_16_16_16 = 19,
+	VFMT4_USHORT_16 = 20,
+	VFMT4_USHORT_16_16 = 21,
+	VFMT4_USHORT_16_16_16 = 22,
+	VFMT4_USHORT_16_16_16_16 = 23,
+	VFMT4_NORM_SHORT_16 = 24,
+	VFMT4_NORM_SHORT_16_16 = 25,
+	VFMT4_NORM_SHORT_16_16_16 = 26,
+	VFMT4_NORM_SHORT_16_16_16_16 = 27,
+	VFMT4_NORM_USHORT_16 = 28,
+	VFMT4_NORM_USHORT_16_16 = 29,
+	VFMT4_NORM_USHORT_16_16_16 = 30,
+	VFMT4_NORM_USHORT_16_16_16_16 = 31,
+	VFMT4_UBYTE_8 = 40,
+	VFMT4_UBYTE_8_8 = 41,
+	VFMT4_UBYTE_8_8_8 = 42,
+	VFMT4_UBYTE_8_8_8_8 = 43,
+	VFMT4_NORM_UBYTE_8 = 44,
+	VFMT4_NORM_UBYTE_8_8 = 45,
+	VFMT4_NORM_UBYTE_8_8_8 = 46,
+	VFMT4_NORM_UBYTE_8_8_8_8 = 47,
+	VFMT4_BYTE_8 = 48,
+	VFMT4_BYTE_8_8 = 49,
+	VFMT4_BYTE_8_8_8 = 50,
+	VFMT4_BYTE_8_8_8_8 = 51,
+	VFMT4_NORM_BYTE_8 = 52,
+	VFMT4_NORM_BYTE_8_8 = 53,
+	VFMT4_NORM_BYTE_8_8_8 = 54,
+	VFMT4_NORM_BYTE_8_8_8_8 = 55,
+	VFMT4_UINT_10_10_10_2 = 60,
+	VFMT4_NORM_UINT_10_10_10_2 = 61,
+	VFMT4_INT_10_10_10_2 = 62,
+	VFMT4_NORM_INT_10_10_10_2 = 63,
+};
+
+enum a4xx_tex_fmt {
+	TFMT4_NORM_USHORT_565 = 11,
+	TFMT4_NORM_USHORT_5551 = 10,
+	TFMT4_NORM_USHORT_4444 = 8,
+	TFMT4_NORM_UINT_X8Z24 = 71,
+	TFMT4_NORM_UINT_2_10_10_10 = 33,
+	TFMT4_NORM_UINT_A8 = 3,
+	TFMT4_NORM_UINT_L8_A8 = 13,
+	TFMT4_NORM_UINT_8 = 4,
+	TFMT4_NORM_UINT_8_8_8_8 = 28,
+	TFMT4_FLOAT_16 = 20,
+	TFMT4_FLOAT_16_16 = 40,
+	TFMT4_FLOAT_16_16_16_16 = 53,
+	TFMT4_FLOAT_32 = 43,
+	TFMT4_FLOAT_32_32 = 56,
+	TFMT4_FLOAT_32_32_32_32 = 63,
+};
+
+enum a4xx_depth_format {
+	DEPTH4_NONE = 0,
+	DEPTH4_16 = 1,
+	DEPTH4_24_8 = 2,
+};
+
+enum a4xx_tex_filter {
+	A4XX_TEX_NEAREST = 0,
+	A4XX_TEX_LINEAR = 1,
+};
+
+enum a4xx_tex_clamp {
+	A4XX_TEX_REPEAT = 0,
+	A4XX_TEX_CLAMP_TO_EDGE = 1,
+	A4XX_TEX_MIRROR_REPEAT = 2,
+	A4XX_TEX_CLAMP_NONE = 3,
+};
+
+enum a4xx_tex_swiz {
+	A4XX_TEX_X = 0,
+	A4XX_TEX_Y = 1,
+	A4XX_TEX_Z = 2,
+	A4XX_TEX_W = 3,
+	A4XX_TEX_ZERO = 4,
+	A4XX_TEX_ONE = 5,
+};
+
+enum a4xx_tex_type {
+	A4XX_TEX_1D = 0,
+	A4XX_TEX_2D = 1,
+	A4XX_TEX_CUBE = 2,
+	A4XX_TEX_3D = 3,
+};
+
+#define A4XX_CGC_HLSQ_EARLY_CYC__MASK				0x00700000
+#define A4XX_CGC_HLSQ_EARLY_CYC__SHIFT				20
+static inline uint32_t A4XX_CGC_HLSQ_EARLY_CYC(uint32_t val)
+{
+	return ((val) << A4XX_CGC_HLSQ_EARLY_CYC__SHIFT) & A4XX_CGC_HLSQ_EARLY_CYC__MASK;
+}
+#define A4XX_INT0_RBBM_GPU_IDLE					0x00000001
+#define A4XX_INT0_RBBM_AHB_ERROR				0x00000002
+#define A4XX_INT0_RBBM_REG_TIMEOUT				0x00000004
+#define A4XX_INT0_RBBM_ME_MS_TIMEOUT				0x00000008
+#define A4XX_INT0_RBBM_PFP_MS_TIMEOUT				0x00000010
+#define A4XX_INT0_RBBM_ATB_BUS_OVERFLOW				0x00000020
+#define A4XX_INT0_VFD_ERROR					0x00000040
+#define A4XX_INT0_CP_SW_INT					0x00000080
+#define A4XX_INT0_CP_T0_PACKET_IN_IB				0x00000100
+#define A4XX_INT0_CP_OPCODE_ERROR				0x00000200
+#define A4XX_INT0_CP_RESERVED_BIT_ERROR				0x00000400
+#define A4XX_INT0_CP_HW_FAULT					0x00000800
+#define A4XX_INT0_CP_DMA					0x00001000
+#define A4XX_INT0_CP_IB2_INT					0x00002000
+#define A4XX_INT0_CP_IB1_INT					0x00004000
+#define A4XX_INT0_CP_RB_INT					0x00008000
+#define A4XX_INT0_CP_REG_PROTECT_FAULT				0x00010000
+#define A4XX_INT0_CP_RB_DONE_TS					0x00020000
+#define A4XX_INT0_CP_VS_DONE_TS					0x00040000
+#define A4XX_INT0_CP_PS_DONE_TS					0x00080000
+#define A4XX_INT0_CACHE_FLUSH_TS				0x00100000
+#define A4XX_INT0_CP_AHB_ERROR_HALT				0x00200000
+#define A4XX_INT0_MISC_HANG_DETECT				0x01000000
+#define A4XX_INT0_UCHE_OOB_ACCESS				0x02000000
+#define REG_A4XX_RB_GMEM_BASE_ADDR				0x00000cc0
+
+#define REG_A4XX_RB_PERFCTR_RB_SEL_0				0x00000cc7
+
+#define REG_A4XX_RB_PERFCTR_RB_SEL_1				0x00000cc8
+
+#define REG_A4XX_RB_PERFCTR_RB_SEL_2				0x00000cc9
+
+#define REG_A4XX_RB_PERFCTR_RB_SEL_3				0x00000cca
+
+#define REG_A4XX_RB_PERFCTR_RB_SEL_4				0x00000ccb
+
+#define REG_A4XX_RB_PERFCTR_RB_SEL_5				0x00000ccc
+
+#define REG_A4XX_RB_PERFCTR_RB_SEL_6				0x00000ccd
+
+#define REG_A4XX_RB_PERFCTR_RB_SEL_7				0x00000cce
+
+#define REG_A4XX_RB_PERFCTR_CCU_SEL_3				0x00000cd2
+
+#define REG_A4XX_RB_FRAME_BUFFER_DIMENSION			0x00000ce0
+#define A4XX_RB_FRAME_BUFFER_DIMENSION_WIDTH__MASK		0x00003fff
+#define A4XX_RB_FRAME_BUFFER_DIMENSION_WIDTH__SHIFT		0
+static inline uint32_t A4XX_RB_FRAME_BUFFER_DIMENSION_WIDTH(uint32_t val)
+{
+	return ((val) << A4XX_RB_FRAME_BUFFER_DIMENSION_WIDTH__SHIFT) & A4XX_RB_FRAME_BUFFER_DIMENSION_WIDTH__MASK;
+}
+#define A4XX_RB_FRAME_BUFFER_DIMENSION_HEIGHT__MASK		0x3fff0000
+#define A4XX_RB_FRAME_BUFFER_DIMENSION_HEIGHT__SHIFT		16
+static inline uint32_t A4XX_RB_FRAME_BUFFER_DIMENSION_HEIGHT(uint32_t val)
+{
+	return ((val) << A4XX_RB_FRAME_BUFFER_DIMENSION_HEIGHT__SHIFT) & A4XX_RB_FRAME_BUFFER_DIMENSION_HEIGHT__MASK;
+}
+
+#define REG_A4XX_RB_CLEAR_COLOR_DW0				0x000020cc
+
+#define REG_A4XX_RB_CLEAR_COLOR_DW1				0x000020cd
+
+#define REG_A4XX_RB_CLEAR_COLOR_DW2				0x000020ce
+
+#define REG_A4XX_RB_CLEAR_COLOR_DW3				0x000020cf
+
+#define REG_A4XX_RB_MODE_CONTROL				0x000020a0
+#define A4XX_RB_MODE_CONTROL_WIDTH__MASK			0x0000003f
+#define A4XX_RB_MODE_CONTROL_WIDTH__SHIFT			0
+static inline uint32_t A4XX_RB_MODE_CONTROL_WIDTH(uint32_t val)
+{
+	return ((val >> 5) << A4XX_RB_MODE_CONTROL_WIDTH__SHIFT) & A4XX_RB_MODE_CONTROL_WIDTH__MASK;
+}
+#define A4XX_RB_MODE_CONTROL_HEIGHT__MASK			0x00003f00
+#define A4XX_RB_MODE_CONTROL_HEIGHT__SHIFT			8
+static inline uint32_t A4XX_RB_MODE_CONTROL_HEIGHT(uint32_t val)
+{
+	return ((val >> 5) << A4XX_RB_MODE_CONTROL_HEIGHT__SHIFT) & A4XX_RB_MODE_CONTROL_HEIGHT__MASK;
+}
+
+#define REG_A4XX_RB_RENDER_CONTROL				0x000020a1
+#define A4XX_RB_RENDER_CONTROL_BINNING_PASS			0x00000001
+#define A4XX_RB_RENDER_CONTROL_DISABLE_COLOR_PIPE		0x00000020
+
+#define REG_A4XX_RB_MSAA_CONTROL				0x000020a2
+#define A4XX_RB_MSAA_CONTROL_DISABLE				0x00001000
+#define A4XX_RB_MSAA_CONTROL_SAMPLES__MASK			0x0000e000
+#define A4XX_RB_MSAA_CONTROL_SAMPLES__SHIFT			13
+static inline uint32_t A4XX_RB_MSAA_CONTROL_SAMPLES(uint32_t val)
+{
+	return ((val) << A4XX_RB_MSAA_CONTROL_SAMPLES__SHIFT) & A4XX_RB_MSAA_CONTROL_SAMPLES__MASK;
+}
+
+#define REG_A4XX_RB_MSAA_CONTROL2				0x000020a3
+#define A4XX_RB_MSAA_CONTROL2_MSAA_SAMPLES__MASK		0x00000380
+#define A4XX_RB_MSAA_CONTROL2_MSAA_SAMPLES__SHIFT		7
+static inline uint32_t A4XX_RB_MSAA_CONTROL2_MSAA_SAMPLES(uint32_t val)
+{
+	return ((val) << A4XX_RB_MSAA_CONTROL2_MSAA_SAMPLES__SHIFT) & A4XX_RB_MSAA_CONTROL2_MSAA_SAMPLES__MASK;
+}
+#define A4XX_RB_MSAA_CONTROL2_VARYING				0x00001000
+
+static inline uint32_t REG_A4XX_RB_MRT(uint32_t i0) { return 0x000020a4 + 0x5*i0; }
+
+static inline uint32_t REG_A4XX_RB_MRT_CONTROL(uint32_t i0) { return 0x000020a4 + 0x5*i0; }
+#define A4XX_RB_MRT_CONTROL_READ_DEST_ENABLE			0x00000008
+#define A4XX_RB_MRT_CONTROL_BLEND				0x00000010
+#define A4XX_RB_MRT_CONTROL_BLEND2				0x00000020
+#define A4XX_RB_MRT_CONTROL_FASTCLEAR				0x00000400
+#define A4XX_RB_MRT_CONTROL_B11					0x00000800
+#define A4XX_RB_MRT_CONTROL_COMPONENT_ENABLE__MASK		0x0f000000
+#define A4XX_RB_MRT_CONTROL_COMPONENT_ENABLE__SHIFT		24
+static inline uint32_t A4XX_RB_MRT_CONTROL_COMPONENT_ENABLE(uint32_t val)
+{
+	return ((val) << A4XX_RB_MRT_CONTROL_COMPONENT_ENABLE__SHIFT) & A4XX_RB_MRT_CONTROL_COMPONENT_ENABLE__MASK;
+}
+
+static inline uint32_t REG_A4XX_RB_MRT_BUF_INFO(uint32_t i0) { return 0x000020a5 + 0x5*i0; }
+#define A4XX_RB_MRT_BUF_INFO_COLOR_FORMAT__MASK			0x0000003f
+#define A4XX_RB_MRT_BUF_INFO_COLOR_FORMAT__SHIFT		0
+static inline uint32_t A4XX_RB_MRT_BUF_INFO_COLOR_FORMAT(enum a4xx_color_fmt val)
+{
+	return ((val) << A4XX_RB_MRT_BUF_INFO_COLOR_FORMAT__SHIFT) & A4XX_RB_MRT_BUF_INFO_COLOR_FORMAT__MASK;
+}
+#define A4XX_RB_MRT_BUF_INFO_DITHER_MODE__MASK			0x00000600
+#define A4XX_RB_MRT_BUF_INFO_DITHER_MODE__SHIFT			9
+static inline uint32_t A4XX_RB_MRT_BUF_INFO_DITHER_MODE(enum adreno_rb_dither_mode val)
+{
+	return ((val) << A4XX_RB_MRT_BUF_INFO_DITHER_MODE__SHIFT) & A4XX_RB_MRT_BUF_INFO_DITHER_MODE__MASK;
+}
+#define A4XX_RB_MRT_BUF_INFO_COLOR_SWAP__MASK			0x00001800
+#define A4XX_RB_MRT_BUF_INFO_COLOR_SWAP__SHIFT			11
+static inline uint32_t A4XX_RB_MRT_BUF_INFO_COLOR_SWAP(enum a3xx_color_swap val)
+{
+	return ((val) << A4XX_RB_MRT_BUF_INFO_COLOR_SWAP__SHIFT) & A4XX_RB_MRT_BUF_INFO_COLOR_SWAP__MASK;
+}
+#define A4XX_RB_MRT_BUF_INFO_COLOR_BUF_PITCH__MASK		0x007fc000
+#define A4XX_RB_MRT_BUF_INFO_COLOR_BUF_PITCH__SHIFT		14
+static inline uint32_t A4XX_RB_MRT_BUF_INFO_COLOR_BUF_PITCH(uint32_t val)
+{
+	return ((val >> 4) << A4XX_RB_MRT_BUF_INFO_COLOR_BUF_PITCH__SHIFT) & A4XX_RB_MRT_BUF_INFO_COLOR_BUF_PITCH__MASK;
+}
+
+static inline uint32_t REG_A4XX_RB_MRT_BASE(uint32_t i0) { return 0x000020a6 + 0x5*i0; }
+
+static inline uint32_t REG_A4XX_RB_MRT_CONTROL3(uint32_t i0) { return 0x000020a7 + 0x5*i0; }
+#define A4XX_RB_MRT_CONTROL3_STRIDE__MASK			0x0001fff8
+#define A4XX_RB_MRT_CONTROL3_STRIDE__SHIFT			3
+static inline uint32_t A4XX_RB_MRT_CONTROL3_STRIDE(uint32_t val)
+{
+	return ((val) << A4XX_RB_MRT_CONTROL3_STRIDE__SHIFT) & A4XX_RB_MRT_CONTROL3_STRIDE__MASK;
+}
+
+static inline uint32_t REG_A4XX_RB_MRT_BLEND_CONTROL(uint32_t i0) { return 0x000020a8 + 0x5*i0; }
+#define A4XX_RB_MRT_BLEND_CONTROL_RGB_SRC_FACTOR__MASK		0x0000001f
+#define A4XX_RB_MRT_BLEND_CONTROL_RGB_SRC_FACTOR__SHIFT		0
+static inline uint32_t A4XX_RB_MRT_BLEND_CONTROL_RGB_SRC_FACTOR(enum adreno_rb_blend_factor val)
+{
+	return ((val) << A4XX_RB_MRT_BLEND_CONTROL_RGB_SRC_FACTOR__SHIFT) & A4XX_RB_MRT_BLEND_CONTROL_RGB_SRC_FACTOR__MASK;
+}
+#define A4XX_RB_MRT_BLEND_CONTROL_RGB_BLEND_OPCODE__MASK	0x000000e0
+#define A4XX_RB_MRT_BLEND_CONTROL_RGB_BLEND_OPCODE__SHIFT	5
+static inline uint32_t A4XX_RB_MRT_BLEND_CONTROL_RGB_BLEND_OPCODE(enum a4xx_rb_blend_opcode val)
+{
+	return ((val) << A4XX_RB_MRT_BLEND_CONTROL_RGB_BLEND_OPCODE__SHIFT) & A4XX_RB_MRT_BLEND_CONTROL_RGB_BLEND_OPCODE__MASK;
+}
+#define A4XX_RB_MRT_BLEND_CONTROL_RGB_DEST_FACTOR__MASK		0x00001f00
+#define A4XX_RB_MRT_BLEND_CONTROL_RGB_DEST_FACTOR__SHIFT	8
+static inline uint32_t A4XX_RB_MRT_BLEND_CONTROL_RGB_DEST_FACTOR(enum adreno_rb_blend_factor val)
+{
+	return ((val) << A4XX_RB_MRT_BLEND_CONTROL_RGB_DEST_FACTOR__SHIFT) & A4XX_RB_MRT_BLEND_CONTROL_RGB_DEST_FACTOR__MASK;
+}
+#define A4XX_RB_MRT_BLEND_CONTROL_ALPHA_SRC_FACTOR__MASK	0x001f0000
+#define A4XX_RB_MRT_BLEND_CONTROL_ALPHA_SRC_FACTOR__SHIFT	16
+static inline uint32_t A4XX_RB_MRT_BLEND_CONTROL_ALPHA_SRC_FACTOR(enum adreno_rb_blend_factor val)
+{
+	return ((val) << A4XX_RB_MRT_BLEND_CONTROL_ALPHA_SRC_FACTOR__SHIFT) & A4XX_RB_MRT_BLEND_CONTROL_ALPHA_SRC_FACTOR__MASK;
+}
+#define A4XX_RB_MRT_BLEND_CONTROL_ALPHA_BLEND_OPCODE__MASK	0x00e00000
+#define A4XX_RB_MRT_BLEND_CONTROL_ALPHA_BLEND_OPCODE__SHIFT	21
+static inline uint32_t A4XX_RB_MRT_BLEND_CONTROL_ALPHA_BLEND_OPCODE(enum a4xx_rb_blend_opcode val)
+{
+	return ((val) << A4XX_RB_MRT_BLEND_CONTROL_ALPHA_BLEND_OPCODE__SHIFT) & A4XX_RB_MRT_BLEND_CONTROL_ALPHA_BLEND_OPCODE__MASK;
+}
+#define A4XX_RB_MRT_BLEND_CONTROL_ALPHA_DEST_FACTOR__MASK	0x1f000000
+#define A4XX_RB_MRT_BLEND_CONTROL_ALPHA_DEST_FACTOR__SHIFT	24
+static inline uint32_t A4XX_RB_MRT_BLEND_CONTROL_ALPHA_DEST_FACTOR(enum adreno_rb_blend_factor val)
+{
+	return ((val) << A4XX_RB_MRT_BLEND_CONTROL_ALPHA_DEST_FACTOR__SHIFT) & A4XX_RB_MRT_BLEND_CONTROL_ALPHA_DEST_FACTOR__MASK;
+}
+
+#define REG_A4XX_RB_ALPHA_CONTROL				0x000020f8
+#define A4XX_RB_ALPHA_CONTROL_ALPHA_TEST			0x00000100
+#define A4XX_RB_ALPHA_CONTROL_ALPHA_TEST_FUNC__MASK		0x00000e00
+#define A4XX_RB_ALPHA_CONTROL_ALPHA_TEST_FUNC__SHIFT		9
+static inline uint32_t A4XX_RB_ALPHA_CONTROL_ALPHA_TEST_FUNC(enum adreno_compare_func val)
+{
+	return ((val) << A4XX_RB_ALPHA_CONTROL_ALPHA_TEST_FUNC__SHIFT) & A4XX_RB_ALPHA_CONTROL_ALPHA_TEST_FUNC__MASK;
+}
+
+#define REG_A4XX_RB_FS_OUTPUT					0x000020f9
+#define A4XX_RB_FS_OUTPUT_ENABLE_COLOR_PIPE			0x00000001
+#define A4XX_RB_FS_OUTPUT_FAST_CLEAR				0x00000100
+#define A4XX_RB_FS_OUTPUT_SAMPLE_MASK__MASK			0xffff0000
+#define A4XX_RB_FS_OUTPUT_SAMPLE_MASK__SHIFT			16
+static inline uint32_t A4XX_RB_FS_OUTPUT_SAMPLE_MASK(uint32_t val)
+{
+	return ((val) << A4XX_RB_FS_OUTPUT_SAMPLE_MASK__SHIFT) & A4XX_RB_FS_OUTPUT_SAMPLE_MASK__MASK;
+}
+
+#define REG_A4XX_RB_RENDER_CONTROL3				0x000020fb
+#define A4XX_RB_RENDER_CONTROL3_COMPONENT_ENABLE__MASK		0x0000001f
+#define A4XX_RB_RENDER_CONTROL3_COMPONENT_ENABLE__SHIFT		0
+static inline uint32_t A4XX_RB_RENDER_CONTROL3_COMPONENT_ENABLE(uint32_t val)
+{
+	return ((val) << A4XX_RB_RENDER_CONTROL3_COMPONENT_ENABLE__SHIFT) & A4XX_RB_RENDER_CONTROL3_COMPONENT_ENABLE__MASK;
+}
+
+#define REG_A4XX_RB_COPY_CONTROL				0x000020fc
+#define A4XX_RB_COPY_CONTROL_MSAA_RESOLVE__MASK			0x00000003
+#define A4XX_RB_COPY_CONTROL_MSAA_RESOLVE__SHIFT		0
+static inline uint32_t A4XX_RB_COPY_CONTROL_MSAA_RESOLVE(enum a3xx_msaa_samples val)
+{
+	return ((val) << A4XX_RB_COPY_CONTROL_MSAA_RESOLVE__SHIFT) & A4XX_RB_COPY_CONTROL_MSAA_RESOLVE__MASK;
+}
+#define A4XX_RB_COPY_CONTROL_MODE__MASK				0x00000070
+#define A4XX_RB_COPY_CONTROL_MODE__SHIFT			4
+static inline uint32_t A4XX_RB_COPY_CONTROL_MODE(enum adreno_rb_copy_control_mode val)
+{
+	return ((val) << A4XX_RB_COPY_CONTROL_MODE__SHIFT) & A4XX_RB_COPY_CONTROL_MODE__MASK;
+}
+#define A4XX_RB_COPY_CONTROL_FASTCLEAR__MASK			0x00000f00
+#define A4XX_RB_COPY_CONTROL_FASTCLEAR__SHIFT			8
+static inline uint32_t A4XX_RB_COPY_CONTROL_FASTCLEAR(uint32_t val)
+{
+	return ((val) << A4XX_RB_COPY_CONTROL_FASTCLEAR__SHIFT) & A4XX_RB_COPY_CONTROL_FASTCLEAR__MASK;
+}
+#define A4XX_RB_COPY_CONTROL_GMEM_BASE__MASK			0xffffc000
+#define A4XX_RB_COPY_CONTROL_GMEM_BASE__SHIFT			14
+static inline uint32_t A4XX_RB_COPY_CONTROL_GMEM_BASE(uint32_t val)
+{
+	return ((val >> 14) << A4XX_RB_COPY_CONTROL_GMEM_BASE__SHIFT) & A4XX_RB_COPY_CONTROL_GMEM_BASE__MASK;
+}
+
+#define REG_A4XX_RB_COPY_DEST_BASE				0x000020fd
+#define A4XX_RB_COPY_DEST_BASE_BASE__MASK			0xfffffff0
+#define A4XX_RB_COPY_DEST_BASE_BASE__SHIFT			4
+static inline uint32_t A4XX_RB_COPY_DEST_BASE_BASE(uint32_t val)
+{
+	return ((val >> 4) << A4XX_RB_COPY_DEST_BASE_BASE__SHIFT) & A4XX_RB_COPY_DEST_BASE_BASE__MASK;
+}
+
+#define REG_A4XX_RB_COPY_DEST_PITCH				0x000020fe
+#define A4XX_RB_COPY_DEST_PITCH_PITCH__MASK			0xffffffff
+#define A4XX_RB_COPY_DEST_PITCH_PITCH__SHIFT			0
+static inline uint32_t A4XX_RB_COPY_DEST_PITCH_PITCH(uint32_t val)
+{
+	return ((val >> 5) << A4XX_RB_COPY_DEST_PITCH_PITCH__SHIFT) & A4XX_RB_COPY_DEST_PITCH_PITCH__MASK;
+}
+
+#define REG_A4XX_RB_COPY_DEST_INFO				0x000020ff
+#define A4XX_RB_COPY_DEST_INFO_FORMAT__MASK			0x000000fc
+#define A4XX_RB_COPY_DEST_INFO_FORMAT__SHIFT			2
+static inline uint32_t A4XX_RB_COPY_DEST_INFO_FORMAT(enum a4xx_color_fmt val)
+{
+	return ((val) << A4XX_RB_COPY_DEST_INFO_FORMAT__SHIFT) & A4XX_RB_COPY_DEST_INFO_FORMAT__MASK;
+}
+#define A4XX_RB_COPY_DEST_INFO_SWAP__MASK			0x00000300
+#define A4XX_RB_COPY_DEST_INFO_SWAP__SHIFT			8
+static inline uint32_t A4XX_RB_COPY_DEST_INFO_SWAP(enum a3xx_color_swap val)
+{
+	return ((val) << A4XX_RB_COPY_DEST_INFO_SWAP__SHIFT) & A4XX_RB_COPY_DEST_INFO_SWAP__MASK;
+}
+#define A4XX_RB_COPY_DEST_INFO_DITHER_MODE__MASK		0x00000c00
+#define A4XX_RB_COPY_DEST_INFO_DITHER_MODE__SHIFT		10
+static inline uint32_t A4XX_RB_COPY_DEST_INFO_DITHER_MODE(enum adreno_rb_dither_mode val)
+{
+	return ((val) << A4XX_RB_COPY_DEST_INFO_DITHER_MODE__SHIFT) & A4XX_RB_COPY_DEST_INFO_DITHER_MODE__MASK;
+}
+#define A4XX_RB_COPY_DEST_INFO_COMPONENT_ENABLE__MASK		0x0003c000
+#define A4XX_RB_COPY_DEST_INFO_COMPONENT_ENABLE__SHIFT		14
+static inline uint32_t A4XX_RB_COPY_DEST_INFO_COMPONENT_ENABLE(uint32_t val)
+{
+	return ((val) << A4XX_RB_COPY_DEST_INFO_COMPONENT_ENABLE__SHIFT) & A4XX_RB_COPY_DEST_INFO_COMPONENT_ENABLE__MASK;
+}
+#define A4XX_RB_COPY_DEST_INFO_ENDIAN__MASK			0x001c0000
+#define A4XX_RB_COPY_DEST_INFO_ENDIAN__SHIFT			18
+static inline uint32_t A4XX_RB_COPY_DEST_INFO_ENDIAN(enum adreno_rb_surface_endian val)
+{
+	return ((val) << A4XX_RB_COPY_DEST_INFO_ENDIAN__SHIFT) & A4XX_RB_COPY_DEST_INFO_ENDIAN__MASK;
+}
+#define A4XX_RB_COPY_DEST_INFO_TILE__MASK			0x03000000
+#define A4XX_RB_COPY_DEST_INFO_TILE__SHIFT			24
+static inline uint32_t A4XX_RB_COPY_DEST_INFO_TILE(enum a4xx_tile_mode val)
+{
+	return ((val) << A4XX_RB_COPY_DEST_INFO_TILE__SHIFT) & A4XX_RB_COPY_DEST_INFO_TILE__MASK;
+}
+
+#define REG_A4XX_RB_FS_OUTPUT_REG				0x00002100
+#define A4XX_RB_FS_OUTPUT_REG_COLOR_PIPE_ENABLE			0x00000001
+#define A4XX_RB_FS_OUTPUT_REG_FRAG_WRITES_Z			0x00000020
+
+#define REG_A4XX_RB_DEPTH_CONTROL				0x00002101
+#define A4XX_RB_DEPTH_CONTROL_FRAG_WRITES_Z			0x00000001
+#define A4XX_RB_DEPTH_CONTROL_Z_ENABLE				0x00000002
+#define A4XX_RB_DEPTH_CONTROL_Z_WRITE_ENABLE			0x00000004
+#define A4XX_RB_DEPTH_CONTROL_ZFUNC__MASK			0x00000070
+#define A4XX_RB_DEPTH_CONTROL_ZFUNC__SHIFT			4
+static inline uint32_t A4XX_RB_DEPTH_CONTROL_ZFUNC(enum adreno_compare_func val)
+{
+	return ((val) << A4XX_RB_DEPTH_CONTROL_ZFUNC__SHIFT) & A4XX_RB_DEPTH_CONTROL_ZFUNC__MASK;
+}
+#define A4XX_RB_DEPTH_CONTROL_BF_ENABLE				0x00000080
+#define A4XX_RB_DEPTH_CONTROL_EARLY_Z_DISABLE			0x00010000
+#define A4XX_RB_DEPTH_CONTROL_Z_TEST_ENABLE			0x80000000
+
+#define REG_A4XX_RB_DEPTH_CLEAR					0x00002102
+
+#define REG_A4XX_RB_DEPTH_INFO					0x00002103
+#define A4XX_RB_DEPTH_INFO_DEPTH_FORMAT__MASK			0x00000003
+#define A4XX_RB_DEPTH_INFO_DEPTH_FORMAT__SHIFT			0
+static inline uint32_t A4XX_RB_DEPTH_INFO_DEPTH_FORMAT(enum a4xx_depth_format val)
+{
+	return ((val) << A4XX_RB_DEPTH_INFO_DEPTH_FORMAT__SHIFT) & A4XX_RB_DEPTH_INFO_DEPTH_FORMAT__MASK;
+}
+#define A4XX_RB_DEPTH_INFO_DEPTH_BASE__MASK			0xfffff000
+#define A4XX_RB_DEPTH_INFO_DEPTH_BASE__SHIFT			12
+static inline uint32_t A4XX_RB_DEPTH_INFO_DEPTH_BASE(uint32_t val)
+{
+	return ((val >> 12) << A4XX_RB_DEPTH_INFO_DEPTH_BASE__SHIFT) & A4XX_RB_DEPTH_INFO_DEPTH_BASE__MASK;
+}
+
+#define REG_A4XX_RB_DEPTH_PITCH					0x00002104
+#define A4XX_RB_DEPTH_PITCH__MASK				0xffffffff
+#define A4XX_RB_DEPTH_PITCH__SHIFT				0
+static inline uint32_t A4XX_RB_DEPTH_PITCH(uint32_t val)
+{
+	return ((val >> 4) << A4XX_RB_DEPTH_PITCH__SHIFT) & A4XX_RB_DEPTH_PITCH__MASK;
+}
+
+#define REG_A4XX_RB_DEPTH_PITCH2				0x00002105
+#define A4XX_RB_DEPTH_PITCH2__MASK				0xffffffff
+#define A4XX_RB_DEPTH_PITCH2__SHIFT				0
+static inline uint32_t A4XX_RB_DEPTH_PITCH2(uint32_t val)
+{
+	return ((val >> 4) << A4XX_RB_DEPTH_PITCH2__SHIFT) & A4XX_RB_DEPTH_PITCH2__MASK;
+}
+
+#define REG_A4XX_RB_STENCIL_CONTROL				0x00002106
+#define A4XX_RB_STENCIL_CONTROL_STENCIL_ENABLE			0x00000001
+#define A4XX_RB_STENCIL_CONTROL_STENCIL_ENABLE_BF		0x00000002
+#define A4XX_RB_STENCIL_CONTROL_STENCIL_READ			0x00000004
+#define A4XX_RB_STENCIL_CONTROL_FUNC__MASK			0x00000700
+#define A4XX_RB_STENCIL_CONTROL_FUNC__SHIFT			8
+static inline uint32_t A4XX_RB_STENCIL_CONTROL_FUNC(enum adreno_compare_func val)
+{
+	return ((val) << A4XX_RB_STENCIL_CONTROL_FUNC__SHIFT) & A4XX_RB_STENCIL_CONTROL_FUNC__MASK;
+}
+#define A4XX_RB_STENCIL_CONTROL_FAIL__MASK			0x00003800
+#define A4XX_RB_STENCIL_CONTROL_FAIL__SHIFT			11
+static inline uint32_t A4XX_RB_STENCIL_CONTROL_FAIL(enum adreno_stencil_op val)
+{
+	return ((val) << A4XX_RB_STENCIL_CONTROL_FAIL__SHIFT) & A4XX_RB_STENCIL_CONTROL_FAIL__MASK;
+}
+#define A4XX_RB_STENCIL_CONTROL_ZPASS__MASK			0x0001c000
+#define A4XX_RB_STENCIL_CONTROL_ZPASS__SHIFT			14
+static inline uint32_t A4XX_RB_STENCIL_CONTROL_ZPASS(enum adreno_stencil_op val)
+{
+	return ((val) << A4XX_RB_STENCIL_CONTROL_ZPASS__SHIFT) & A4XX_RB_STENCIL_CONTROL_ZPASS__MASK;
+}
+#define A4XX_RB_STENCIL_CONTROL_ZFAIL__MASK			0x000e0000
+#define A4XX_RB_STENCIL_CONTROL_ZFAIL__SHIFT			17
+static inline uint32_t A4XX_RB_STENCIL_CONTROL_ZFAIL(enum adreno_stencil_op val)
+{
+	return ((val) << A4XX_RB_STENCIL_CONTROL_ZFAIL__SHIFT) & A4XX_RB_STENCIL_CONTROL_ZFAIL__MASK;
+}
+#define A4XX_RB_STENCIL_CONTROL_FUNC_BF__MASK			0x00700000
+#define A4XX_RB_STENCIL_CONTROL_FUNC_BF__SHIFT			20
+static inline uint32_t A4XX_RB_STENCIL_CONTROL_FUNC_BF(enum adreno_compare_func val)
+{
+	return ((val) << A4XX_RB_STENCIL_CONTROL_FUNC_BF__SHIFT) & A4XX_RB_STENCIL_CONTROL_FUNC_BF__MASK;
+}
+#define A4XX_RB_STENCIL_CONTROL_FAIL_BF__MASK			0x03800000
+#define A4XX_RB_STENCIL_CONTROL_FAIL_BF__SHIFT			23
+static inline uint32_t A4XX_RB_STENCIL_CONTROL_FAIL_BF(enum adreno_stencil_op val)
+{
+	return ((val) << A4XX_RB_STENCIL_CONTROL_FAIL_BF__SHIFT) & A4XX_RB_STENCIL_CONTROL_FAIL_BF__MASK;
+}
+#define A4XX_RB_STENCIL_CONTROL_ZPASS_BF__MASK			0x1c000000
+#define A4XX_RB_STENCIL_CONTROL_ZPASS_BF__SHIFT			26
+static inline uint32_t A4XX_RB_STENCIL_CONTROL_ZPASS_BF(enum adreno_stencil_op val)
+{
+	return ((val) << A4XX_RB_STENCIL_CONTROL_ZPASS_BF__SHIFT) & A4XX_RB_STENCIL_CONTROL_ZPASS_BF__MASK;
+}
+#define A4XX_RB_STENCIL_CONTROL_ZFAIL_BF__MASK			0xe0000000
+#define A4XX_RB_STENCIL_CONTROL_ZFAIL_BF__SHIFT			29
+static inline uint32_t A4XX_RB_STENCIL_CONTROL_ZFAIL_BF(enum adreno_stencil_op val)
+{
+	return ((val) << A4XX_RB_STENCIL_CONTROL_ZFAIL_BF__SHIFT) & A4XX_RB_STENCIL_CONTROL_ZFAIL_BF__MASK;
+}
+
+#define REG_A4XX_RB_STENCIL_CONTROL2				0x00002107
+#define A4XX_RB_STENCIL_CONTROL2_STENCIL_BUFFER			0x00000001
+
+#define REG_A4XX_RB_STENCILREFMASK				0x0000210b
+#define A4XX_RB_STENCILREFMASK_STENCILREF__MASK			0x000000ff
+#define A4XX_RB_STENCILREFMASK_STENCILREF__SHIFT		0
+static inline uint32_t A4XX_RB_STENCILREFMASK_STENCILREF(uint32_t val)
+{
+	return ((val) << A4XX_RB_STENCILREFMASK_STENCILREF__SHIFT) & A4XX_RB_STENCILREFMASK_STENCILREF__MASK;
+}
+#define A4XX_RB_STENCILREFMASK_STENCILMASK__MASK		0x0000ff00
+#define A4XX_RB_STENCILREFMASK_STENCILMASK__SHIFT		8
+static inline uint32_t A4XX_RB_STENCILREFMASK_STENCILMASK(uint32_t val)
+{
+	return ((val) << A4XX_RB_STENCILREFMASK_STENCILMASK__SHIFT) & A4XX_RB_STENCILREFMASK_STENCILMASK__MASK;
+}
+#define A4XX_RB_STENCILREFMASK_STENCILWRITEMASK__MASK		0x00ff0000
+#define A4XX_RB_STENCILREFMASK_STENCILWRITEMASK__SHIFT		16
+static inline uint32_t A4XX_RB_STENCILREFMASK_STENCILWRITEMASK(uint32_t val)
+{
+	return ((val) << A4XX_RB_STENCILREFMASK_STENCILWRITEMASK__SHIFT) & A4XX_RB_STENCILREFMASK_STENCILWRITEMASK__MASK;
+}
+
+#define REG_A4XX_RB_STENCILREFMASK_BF				0x0000210c
+#define A4XX_RB_STENCILREFMASK_BF_STENCILREF__MASK		0x000000ff
+#define A4XX_RB_STENCILREFMASK_BF_STENCILREF__SHIFT		0
+static inline uint32_t A4XX_RB_STENCILREFMASK_BF_STENCILREF(uint32_t val)
+{
+	return ((val) << A4XX_RB_STENCILREFMASK_BF_STENCILREF__SHIFT) & A4XX_RB_STENCILREFMASK_BF_STENCILREF__MASK;
+}
+#define A4XX_RB_STENCILREFMASK_BF_STENCILMASK__MASK		0x0000ff00
+#define A4XX_RB_STENCILREFMASK_BF_STENCILMASK__SHIFT		8
+static inline uint32_t A4XX_RB_STENCILREFMASK_BF_STENCILMASK(uint32_t val)
+{
+	return ((val) << A4XX_RB_STENCILREFMASK_BF_STENCILMASK__SHIFT) & A4XX_RB_STENCILREFMASK_BF_STENCILMASK__MASK;
+}
+#define A4XX_RB_STENCILREFMASK_BF_STENCILWRITEMASK__MASK	0x00ff0000
+#define A4XX_RB_STENCILREFMASK_BF_STENCILWRITEMASK__SHIFT	16
+static inline uint32_t A4XX_RB_STENCILREFMASK_BF_STENCILWRITEMASK(uint32_t val)
+{
+	return ((val) << A4XX_RB_STENCILREFMASK_BF_STENCILWRITEMASK__SHIFT) & A4XX_RB_STENCILREFMASK_BF_STENCILWRITEMASK__MASK;
+}
+
+#define REG_A4XX_RB_BIN_OFFSET					0x0000210d
+#define A4XX_RB_BIN_OFFSET_WINDOW_OFFSET_DISABLE		0x80000000
+#define A4XX_RB_BIN_OFFSET_X__MASK				0x00007fff
+#define A4XX_RB_BIN_OFFSET_X__SHIFT				0
+static inline uint32_t A4XX_RB_BIN_OFFSET_X(uint32_t val)
+{
+	return ((val) << A4XX_RB_BIN_OFFSET_X__SHIFT) & A4XX_RB_BIN_OFFSET_X__MASK;
+}
+#define A4XX_RB_BIN_OFFSET_Y__MASK				0x7fff0000
+#define A4XX_RB_BIN_OFFSET_Y__SHIFT				16
+static inline uint32_t A4XX_RB_BIN_OFFSET_Y(uint32_t val)
+{
+	return ((val) << A4XX_RB_BIN_OFFSET_Y__SHIFT) & A4XX_RB_BIN_OFFSET_Y__MASK;
+}
+
+#define REG_A4XX_RB_VPORT_Z_CLAMP_MAX_15			0x0000213f
+
+#define REG_A4XX_RBBM_HW_VERSION				0x00000000
+
+#define REG_A4XX_RBBM_HW_CONFIGURATION				0x00000002
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_CTL_TP(uint32_t i0) { return 0x00000004 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_CTL_TP_REG(uint32_t i0) { return 0x00000004 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_CTL2_TP(uint32_t i0) { return 0x00000008 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_CTL2_TP_REG(uint32_t i0) { return 0x00000008 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_HYST_TP(uint32_t i0) { return 0x0000000c + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_HYST_TP_REG(uint32_t i0) { return 0x0000000c + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_DELAY_TP(uint32_t i0) { return 0x00000010 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_DELAY_TP_REG(uint32_t i0) { return 0x00000010 + 0x1*i0; }
+
+#define REG_A4XX_RBBM_CLOCK_CTL_UCHE 				0x00000014
+
+#define REG_A4XX_RBBM_CLOCK_CTL2_UCHE				0x00000015
+
+#define REG_A4XX_RBBM_CLOCK_CTL3_UCHE				0x00000016
+
+#define REG_A4XX_RBBM_CLOCK_CTL4_UCHE				0x00000017
+
+#define REG_A4XX_RBBM_CLOCK_HYST_UCHE				0x00000018
+
+#define REG_A4XX_RBBM_CLOCK_DELAY_UCHE				0x00000019
+
+#define REG_A4XX_RBBM_CLOCK_MODE_GPC				0x0000001a
+
+#define REG_A4XX_RBBM_CLOCK_DELAY_GPC				0x0000001b
+
+#define REG_A4XX_RBBM_CLOCK_HYST_GPC				0x0000001c
+
+#define REG_A4XX_RBBM_CLOCK_CTL_TSE_RAS_RBBM			0x0000001d
+
+#define REG_A4XX_RBBM_CLOCK_HYST_TSE_RAS_RBBM			0x0000001e
+
+#define REG_A4XX_RBBM_CLOCK_DELAY_TSE_RAS_RBBM			0x0000001f
+
+#define REG_A4XX_RBBM_CLOCK_CTL					0x00000020
+
+#define REG_A4XX_RBBM_SP_HYST_CNT				0x00000021
+
+#define REG_A4XX_RBBM_SW_RESET_CMD				0x00000022
+
+#define REG_A4XX_RBBM_AHB_CTL0					0x00000023
+
+#define REG_A4XX_RBBM_AHB_CTL1					0x00000024
+
+#define REG_A4XX_RBBM_AHB_CMD					0x00000025
+
+#define REG_A4XX_RBBM_RB_SUB_BLOCK_SEL_CTL			0x00000026
+
+#define REG_A4XX_RBBM_RAM_ACC_63_32				0x00000028
+
+#define REG_A4XX_RBBM_WAIT_IDLE_CLOCKS_CTL			0x0000002b
+
+#define REG_A4XX_RBBM_INTERFACE_HANG_INT_CTL			0x0000002f
+
+#define REG_A4XX_RBBM_INTERFACE_HANG_MASK_CTL4			0x00000034
+
+#define REG_A4XX_RBBM_INT_CLEAR_CMD				0x00000036
+
+#define REG_A4XX_RBBM_INT_0_MASK				0x00000037
+
+#define REG_A4XX_RBBM_RBBM_CTL					0x0000003e
+
+#define REG_A4XX_RBBM_AHB_DEBUG_CTL				0x0000003f
+
+#define REG_A4XX_RBBM_VBIF_DEBUG_CTL				0x00000041
+
+#define REG_A4XX_RBBM_CLOCK_CTL2				0x00000042
+
+#define REG_A4XX_RBBM_BLOCK_SW_RESET_CMD			0x00000045
+
+#define REG_A4XX_RBBM_RESET_CYCLES				0x00000047
+
+#define REG_A4XX_RBBM_EXT_TRACE_BUS_CTL				0x00000049
+
+#define REG_A4XX_RBBM_CFG_DEBBUS_SEL_A				0x0000004a
+
+#define REG_A4XX_RBBM_CFG_DEBBUS_SEL_B				0x0000004b
+
+#define REG_A4XX_RBBM_CFG_DEBBUS_SEL_C				0x0000004c
+
+#define REG_A4XX_RBBM_CFG_DEBBUS_SEL_D				0x0000004d
+
+#define REG_A4XX_RBBM_PERFCTR_CP_0_LO				0x0000009c
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_CTL_SP(uint32_t i0) { return 0x00000068 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_CTL_SP_REG(uint32_t i0) { return 0x00000068 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_CTL2_SP(uint32_t i0) { return 0x0000006c + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_CTL2_SP_REG(uint32_t i0) { return 0x0000006c + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_HYST_SP(uint32_t i0) { return 0x00000070 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_HYST_SP_REG(uint32_t i0) { return 0x00000070 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_DELAY_SP(uint32_t i0) { return 0x00000074 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_DELAY_SP_REG(uint32_t i0) { return 0x00000074 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_CTL_RB(uint32_t i0) { return 0x00000078 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_CTL_RB_REG(uint32_t i0) { return 0x00000078 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_CTL2_RB(uint32_t i0) { return 0x0000007c + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_CTL2_RB_REG(uint32_t i0) { return 0x0000007c + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_CTL_MARB_CCU(uint32_t i0) { return 0x00000082 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_CTL_MARB_CCU_REG(uint32_t i0) { return 0x00000082 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_HYST_RB_MARB_CCU(uint32_t i0) { return 0x00000086 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_HYST_RB_MARB_CCU_REG(uint32_t i0) { return 0x00000086 + 0x1*i0; }
+
+#define REG_A4XX_RBBM_CLOCK_HYST_COM_DCOM			0x00000080
+
+#define REG_A4XX_RBBM_CLOCK_CTL_COM_DCOM			0x00000081
+
+#define REG_A4XX_RBBM_CLOCK_CTL_HLSQ				0x0000008a
+
+#define REG_A4XX_RBBM_CLOCK_HYST_HLSQ				0x0000008b
+
+#define REG_A4XX_RBBM_CLOCK_DELAY_HLSQ				0x0000008c
+
+#define REG_A4XX_RBBM_CLOCK_DELAY_COM_DCOM			0x0000008d
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_DELAY_RB_MARB_CCU_L1(uint32_t i0) { return 0x0000008e + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_RBBM_CLOCK_DELAY_RB_MARB_CCU_L1_REG(uint32_t i0) { return 0x0000008e + 0x1*i0; }
+
+#define REG_A4XX_RBBM_PERFCTR_PWR_1_LO				0x00000168
+
+#define REG_A4XX_RBBM_PERFCTR_CTL				0x00000170
+
+#define REG_A4XX_RBBM_PERFCTR_LOAD_CMD0				0x00000171
+
+#define REG_A4XX_RBBM_PERFCTR_LOAD_CMD1				0x00000172
+
+#define REG_A4XX_RBBM_PERFCTR_LOAD_CMD2				0x00000173
+
+#define REG_A4XX_RBBM_PERFCTR_LOAD_VALUE_LO			0x00000174
+
+#define REG_A4XX_RBBM_PERFCTR_LOAD_VALUE_HI			0x00000175
+
+#define REG_A4XX_RBBM_GPU_BUSY_MASKED				0x0000017a
+
+#define REG_A4XX_RBBM_INT_0_STATUS				0x0000017d
+
+#define REG_A4XX_RBBM_CLOCK_STATUS				0x00000182
+
+#define REG_A4XX_RBBM_AHB_STATUS				0x00000189
+
+#define REG_A4XX_RBBM_AHB_ME_SPLIT_STATUS			0x0000018c
+
+#define REG_A4XX_RBBM_AHB_PFP_SPLIT_STATUS			0x0000018d
+
+#define REG_A4XX_RBBM_AHB_ERROR_STATUS				0x0000018f
+
+#define REG_A4XX_RBBM_STATUS					0x00000191
+#define A4XX_RBBM_STATUS_HI_BUSY				0x00000001
+#define A4XX_RBBM_STATUS_CP_ME_BUSY				0x00000002
+#define A4XX_RBBM_STATUS_CP_PFP_BUSY				0x00000004
+#define A4XX_RBBM_STATUS_CP_NRT_BUSY				0x00004000
+#define A4XX_RBBM_STATUS_VBIF_BUSY				0x00008000
+#define A4XX_RBBM_STATUS_TSE_BUSY				0x00010000
+#define A4XX_RBBM_STATUS_RAS_BUSY				0x00020000
+#define A4XX_RBBM_STATUS_RB_BUSY				0x00040000
+#define A4XX_RBBM_STATUS_PC_DCALL_BUSY				0x00080000
+#define A4XX_RBBM_STATUS_PC_VSD_BUSY				0x00100000
+#define A4XX_RBBM_STATUS_VFD_BUSY				0x00200000
+#define A4XX_RBBM_STATUS_VPC_BUSY				0x00400000
+#define A4XX_RBBM_STATUS_UCHE_BUSY				0x00800000
+#define A4XX_RBBM_STATUS_SP_BUSY				0x01000000
+#define A4XX_RBBM_STATUS_TPL1_BUSY				0x02000000
+#define A4XX_RBBM_STATUS_MARB_BUSY				0x04000000
+#define A4XX_RBBM_STATUS_VSC_BUSY				0x08000000
+#define A4XX_RBBM_STATUS_ARB_BUSY				0x10000000
+#define A4XX_RBBM_STATUS_HLSQ_BUSY				0x20000000
+#define A4XX_RBBM_STATUS_GPU_BUSY_NOHC				0x40000000
+#define A4XX_RBBM_STATUS_GPU_BUSY				0x80000000
+
+#define REG_A4XX_RBBM_INTERFACE_RRDY_STATUS5			0x0000019f
+
+#define REG_A4XX_CP_SCRATCH_UMASK				0x00000228
+
+#define REG_A4XX_CP_SCRATCH_ADDR				0x00000229
+
+#define REG_A4XX_CP_RB_BASE					0x00000200
+
+#define REG_A4XX_CP_RB_CNTL					0x00000201
+
+#define REG_A4XX_CP_RB_WPTR					0x00000205
+
+#define REG_A4XX_CP_RB_RPTR_ADDR				0x00000203
+
+#define REG_A4XX_CP_RB_RPTR					0x00000204
+
+#define REG_A4XX_CP_IB1_BASE					0x00000206
+
+#define REG_A4XX_CP_IB1_BUFSZ					0x00000207
+
+#define REG_A4XX_CP_IB2_BASE					0x00000208
+
+#define REG_A4XX_CP_IB2_BUFSZ					0x00000209
+
+#define REG_A4XX_CP_ME_RB_DONE_DATA				0x00000217
+
+#define REG_A4XX_CP_QUEUE_THRESH2				0x00000219
+
+#define REG_A4XX_CP_MERCIU_SIZE					0x0000021b
+
+#define REG_A4XX_CP_ROQ_ADDR					0x0000021c
+
+#define REG_A4XX_CP_ROQ_DATA					0x0000021d
+
+#define REG_A4XX_CP_MEQ_ADDR 					0x0000021e
+
+#define REG_A4XX_CP_MEQ_DATA 					0x0000021f
+
+#define REG_A4XX_CP_MERCIU_ADDR					0x00000220
+
+#define REG_A4XX_CP_MERCIU_DATA					0x00000221
+
+#define REG_A4XX_CP_MERCIU_DATA2				0x00000222
+
+#define REG_A4XX_CP_PFP_UCODE_ADDR				0x00000223
+
+#define REG_A4XX_CP_PFP_UCODE_DATA				0x00000224
+
+#define REG_A4XX_CP_ME_RAM_WADDR				0x00000225
+
+#define REG_A4XX_CP_ME_RAM_RADDR				0x00000226
+
+#define REG_A4XX_CP_ME_RAM_DATA					0x00000227
+
+#define REG_A4XX_CP_PREEMPT					0x0000022a
+
+#define REG_A4XX_CP_CNTL					0x0000022c
+
+#define REG_A4XX_CP_ME_CNTL					0x0000022d
+
+#define REG_A4XX_CP_DEBUG					0x0000022e
+
+#define REG_A4XX_CP_DEBUG_ECO_CONTROL				0x00000231
+
+#define REG_A4XX_CP_DRAW_STATE_ADDR				0x00000232
+
+#define REG_A4XX_CP_PROTECT_REG_0				0x00000240
+
+static inline uint32_t REG_A4XX_CP_PROTECT(uint32_t i0) { return 0x00000240 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_CP_PROTECT_REG(uint32_t i0) { return 0x00000240 + 0x1*i0; }
+
+#define REG_A4XX_CP_PROTECT_CTRL				0x00000250
+
+#define REG_A4XX_CP_ST_BASE					0x000004c0
+
+#define REG_A4XX_CP_STQ_AVAIL					0x000004ce
+
+#define REG_A4XX_CP_MERCIU_STAT					0x000004d0
+
+#define REG_A4XX_CP_WFI_PEND_CTR				0x000004d2
+
+#define REG_A4XX_CP_HW_FAULT					0x000004d8
+
+#define REG_A4XX_CP_PROTECT_STATUS				0x000004da
+
+#define REG_A4XX_CP_EVENTS_IN_FLIGHT				0x000004dd
+
+#define REG_A4XX_CP_PERFCTR_CP_SEL_0				0x00000500
+
+#define REG_A4XX_CP_PERFCOMBINER_SELECT				0x0000050b
+
+static inline uint32_t REG_A4XX_CP_SCRATCH(uint32_t i0) { return 0x00000578 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_CP_SCRATCH_REG(uint32_t i0) { return 0x00000578 + 0x1*i0; }
+
+#define REG_A4XX_SP_VS_STATUS					0x00000ec0
+
+#define REG_A4XX_SP_PERFCTR_SP_SEL_11				0x00000ecf
+
+#define REG_A4XX_SP_SP_CTRL_REG					0x000022c0
+#define A4XX_SP_SP_CTRL_REG_BINNING_PASS			0x00080000
+
+#define REG_A4XX_SP_INSTR_CACHE_CTRL				0x000022c1
+
+#define REG_A4XX_SP_VS_CTRL_REG0				0x000022c4
+#define A4XX_SP_VS_CTRL_REG0_THREADMODE__MASK			0x00000001
+#define A4XX_SP_VS_CTRL_REG0_THREADMODE__SHIFT			0
+static inline uint32_t A4XX_SP_VS_CTRL_REG0_THREADMODE(enum a3xx_threadmode val)
+{
+	return ((val) << A4XX_SP_VS_CTRL_REG0_THREADMODE__SHIFT) & A4XX_SP_VS_CTRL_REG0_THREADMODE__MASK;
+}
+#define A4XX_SP_VS_CTRL_REG0_VARYING				0x00000002
+#define A4XX_SP_VS_CTRL_REG0_CACHEINVALID			0x00000004
+#define A4XX_SP_VS_CTRL_REG0_HALFREGFOOTPRINT__MASK		0x000003f0
+#define A4XX_SP_VS_CTRL_REG0_HALFREGFOOTPRINT__SHIFT		4
+static inline uint32_t A4XX_SP_VS_CTRL_REG0_HALFREGFOOTPRINT(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_CTRL_REG0_HALFREGFOOTPRINT__SHIFT) & A4XX_SP_VS_CTRL_REG0_HALFREGFOOTPRINT__MASK;
+}
+#define A4XX_SP_VS_CTRL_REG0_FULLREGFOOTPRINT__MASK		0x0003fc00
+#define A4XX_SP_VS_CTRL_REG0_FULLREGFOOTPRINT__SHIFT		10
+static inline uint32_t A4XX_SP_VS_CTRL_REG0_FULLREGFOOTPRINT(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_CTRL_REG0_FULLREGFOOTPRINT__SHIFT) & A4XX_SP_VS_CTRL_REG0_FULLREGFOOTPRINT__MASK;
+}
+#define A4XX_SP_VS_CTRL_REG0_INOUTREGOVERLAP__MASK		0x000c0000
+#define A4XX_SP_VS_CTRL_REG0_INOUTREGOVERLAP__SHIFT		18
+static inline uint32_t A4XX_SP_VS_CTRL_REG0_INOUTREGOVERLAP(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_CTRL_REG0_INOUTREGOVERLAP__SHIFT) & A4XX_SP_VS_CTRL_REG0_INOUTREGOVERLAP__MASK;
+}
+#define A4XX_SP_VS_CTRL_REG0_THREADSIZE__MASK			0x00100000
+#define A4XX_SP_VS_CTRL_REG0_THREADSIZE__SHIFT			20
+static inline uint32_t A4XX_SP_VS_CTRL_REG0_THREADSIZE(enum a3xx_threadsize val)
+{
+	return ((val) << A4XX_SP_VS_CTRL_REG0_THREADSIZE__SHIFT) & A4XX_SP_VS_CTRL_REG0_THREADSIZE__MASK;
+}
+#define A4XX_SP_VS_CTRL_REG0_SUPERTHREADMODE			0x00200000
+#define A4XX_SP_VS_CTRL_REG0_PIXLODENABLE			0x00400000
+
+#define REG_A4XX_SP_VS_CTRL_REG1				0x000022c5
+#define A4XX_SP_VS_CTRL_REG1_CONSTLENGTH__MASK			0x000000ff
+#define A4XX_SP_VS_CTRL_REG1_CONSTLENGTH__SHIFT			0
+static inline uint32_t A4XX_SP_VS_CTRL_REG1_CONSTLENGTH(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_CTRL_REG1_CONSTLENGTH__SHIFT) & A4XX_SP_VS_CTRL_REG1_CONSTLENGTH__MASK;
+}
+#define A4XX_SP_VS_CTRL_REG1_INITIALOUTSTANDING__MASK		0x7f000000
+#define A4XX_SP_VS_CTRL_REG1_INITIALOUTSTANDING__SHIFT		24
+static inline uint32_t A4XX_SP_VS_CTRL_REG1_INITIALOUTSTANDING(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_CTRL_REG1_INITIALOUTSTANDING__SHIFT) & A4XX_SP_VS_CTRL_REG1_INITIALOUTSTANDING__MASK;
+}
+
+#define REG_A4XX_SP_VS_PARAM_REG				0x000022c6
+#define A4XX_SP_VS_PARAM_REG_POSREGID__MASK			0x000000ff
+#define A4XX_SP_VS_PARAM_REG_POSREGID__SHIFT			0
+static inline uint32_t A4XX_SP_VS_PARAM_REG_POSREGID(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_PARAM_REG_POSREGID__SHIFT) & A4XX_SP_VS_PARAM_REG_POSREGID__MASK;
+}
+#define A4XX_SP_VS_PARAM_REG_PSIZEREGID__MASK			0x0000ff00
+#define A4XX_SP_VS_PARAM_REG_PSIZEREGID__SHIFT			8
+static inline uint32_t A4XX_SP_VS_PARAM_REG_PSIZEREGID(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_PARAM_REG_PSIZEREGID__SHIFT) & A4XX_SP_VS_PARAM_REG_PSIZEREGID__MASK;
+}
+#define A4XX_SP_VS_PARAM_REG_TOTALVSOUTVAR__MASK		0xfff00000
+#define A4XX_SP_VS_PARAM_REG_TOTALVSOUTVAR__SHIFT		20
+static inline uint32_t A4XX_SP_VS_PARAM_REG_TOTALVSOUTVAR(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_PARAM_REG_TOTALVSOUTVAR__SHIFT) & A4XX_SP_VS_PARAM_REG_TOTALVSOUTVAR__MASK;
+}
+
+static inline uint32_t REG_A4XX_SP_VS_OUT(uint32_t i0) { return 0x000022c7 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_SP_VS_OUT_REG(uint32_t i0) { return 0x000022c7 + 0x1*i0; }
+#define A4XX_SP_VS_OUT_REG_A_REGID__MASK			0x000001ff
+#define A4XX_SP_VS_OUT_REG_A_REGID__SHIFT			0
+static inline uint32_t A4XX_SP_VS_OUT_REG_A_REGID(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_OUT_REG_A_REGID__SHIFT) & A4XX_SP_VS_OUT_REG_A_REGID__MASK;
+}
+#define A4XX_SP_VS_OUT_REG_A_COMPMASK__MASK			0x00001e00
+#define A4XX_SP_VS_OUT_REG_A_COMPMASK__SHIFT			9
+static inline uint32_t A4XX_SP_VS_OUT_REG_A_COMPMASK(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_OUT_REG_A_COMPMASK__SHIFT) & A4XX_SP_VS_OUT_REG_A_COMPMASK__MASK;
+}
+#define A4XX_SP_VS_OUT_REG_B_REGID__MASK			0x01ff0000
+#define A4XX_SP_VS_OUT_REG_B_REGID__SHIFT			16
+static inline uint32_t A4XX_SP_VS_OUT_REG_B_REGID(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_OUT_REG_B_REGID__SHIFT) & A4XX_SP_VS_OUT_REG_B_REGID__MASK;
+}
+#define A4XX_SP_VS_OUT_REG_B_COMPMASK__MASK			0x1e000000
+#define A4XX_SP_VS_OUT_REG_B_COMPMASK__SHIFT			25
+static inline uint32_t A4XX_SP_VS_OUT_REG_B_COMPMASK(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_OUT_REG_B_COMPMASK__SHIFT) & A4XX_SP_VS_OUT_REG_B_COMPMASK__MASK;
+}
+
+static inline uint32_t REG_A4XX_SP_VS_VPC_DST(uint32_t i0) { return 0x000022d8 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_SP_VS_VPC_DST_REG(uint32_t i0) { return 0x000022d8 + 0x1*i0; }
+#define A4XX_SP_VS_VPC_DST_REG_OUTLOC0__MASK			0x000000ff
+#define A4XX_SP_VS_VPC_DST_REG_OUTLOC0__SHIFT			0
+static inline uint32_t A4XX_SP_VS_VPC_DST_REG_OUTLOC0(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_VPC_DST_REG_OUTLOC0__SHIFT) & A4XX_SP_VS_VPC_DST_REG_OUTLOC0__MASK;
+}
+#define A4XX_SP_VS_VPC_DST_REG_OUTLOC1__MASK			0x0000ff00
+#define A4XX_SP_VS_VPC_DST_REG_OUTLOC1__SHIFT			8
+static inline uint32_t A4XX_SP_VS_VPC_DST_REG_OUTLOC1(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_VPC_DST_REG_OUTLOC1__SHIFT) & A4XX_SP_VS_VPC_DST_REG_OUTLOC1__MASK;
+}
+#define A4XX_SP_VS_VPC_DST_REG_OUTLOC2__MASK			0x00ff0000
+#define A4XX_SP_VS_VPC_DST_REG_OUTLOC2__SHIFT			16
+static inline uint32_t A4XX_SP_VS_VPC_DST_REG_OUTLOC2(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_VPC_DST_REG_OUTLOC2__SHIFT) & A4XX_SP_VS_VPC_DST_REG_OUTLOC2__MASK;
+}
+#define A4XX_SP_VS_VPC_DST_REG_OUTLOC3__MASK			0xff000000
+#define A4XX_SP_VS_VPC_DST_REG_OUTLOC3__SHIFT			24
+static inline uint32_t A4XX_SP_VS_VPC_DST_REG_OUTLOC3(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_VPC_DST_REG_OUTLOC3__SHIFT) & A4XX_SP_VS_VPC_DST_REG_OUTLOC3__MASK;
+}
+
+#define REG_A4XX_SP_VS_OBJ_OFFSET_REG				0x000022e0
+#define A4XX_SP_VS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__MASK	0x01ff0000
+#define A4XX_SP_VS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__SHIFT	16
+static inline uint32_t A4XX_SP_VS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__SHIFT) & A4XX_SP_VS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__MASK;
+}
+#define A4XX_SP_VS_OBJ_OFFSET_REG_SHADEROBJOFFSET__MASK		0xfe000000
+#define A4XX_SP_VS_OBJ_OFFSET_REG_SHADEROBJOFFSET__SHIFT	25
+static inline uint32_t A4XX_SP_VS_OBJ_OFFSET_REG_SHADEROBJOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_SP_VS_OBJ_OFFSET_REG_SHADEROBJOFFSET__SHIFT) & A4XX_SP_VS_OBJ_OFFSET_REG_SHADEROBJOFFSET__MASK;
+}
+
+#define REG_A4XX_SP_VS_OBJ_START				0x000022e1
+
+#define REG_A4XX_SP_VS_PVT_MEM_PARAM				0x000022e2
+
+#define REG_A4XX_SP_VS_PVT_MEM_ADDR				0x000022e3
+
+#define REG_A4XX_SP_VS_LENGTH_REG				0x000022e5
+
+#define REG_A4XX_SP_FS_CTRL_REG0				0x000022e8
+#define A4XX_SP_FS_CTRL_REG0_THREADMODE__MASK			0x00000001
+#define A4XX_SP_FS_CTRL_REG0_THREADMODE__SHIFT			0
+static inline uint32_t A4XX_SP_FS_CTRL_REG0_THREADMODE(enum a3xx_threadmode val)
+{
+	return ((val) << A4XX_SP_FS_CTRL_REG0_THREADMODE__SHIFT) & A4XX_SP_FS_CTRL_REG0_THREADMODE__MASK;
+}
+#define A4XX_SP_FS_CTRL_REG0_VARYING				0x00000002
+#define A4XX_SP_FS_CTRL_REG0_CACHEINVALID			0x00000004
+#define A4XX_SP_FS_CTRL_REG0_HALFREGFOOTPRINT__MASK		0x000003f0
+#define A4XX_SP_FS_CTRL_REG0_HALFREGFOOTPRINT__SHIFT		4
+static inline uint32_t A4XX_SP_FS_CTRL_REG0_HALFREGFOOTPRINT(uint32_t val)
+{
+	return ((val) << A4XX_SP_FS_CTRL_REG0_HALFREGFOOTPRINT__SHIFT) & A4XX_SP_FS_CTRL_REG0_HALFREGFOOTPRINT__MASK;
+}
+#define A4XX_SP_FS_CTRL_REG0_FULLREGFOOTPRINT__MASK		0x0003fc00
+#define A4XX_SP_FS_CTRL_REG0_FULLREGFOOTPRINT__SHIFT		10
+static inline uint32_t A4XX_SP_FS_CTRL_REG0_FULLREGFOOTPRINT(uint32_t val)
+{
+	return ((val) << A4XX_SP_FS_CTRL_REG0_FULLREGFOOTPRINT__SHIFT) & A4XX_SP_FS_CTRL_REG0_FULLREGFOOTPRINT__MASK;
+}
+#define A4XX_SP_FS_CTRL_REG0_INOUTREGOVERLAP__MASK		0x000c0000
+#define A4XX_SP_FS_CTRL_REG0_INOUTREGOVERLAP__SHIFT		18
+static inline uint32_t A4XX_SP_FS_CTRL_REG0_INOUTREGOVERLAP(uint32_t val)
+{
+	return ((val) << A4XX_SP_FS_CTRL_REG0_INOUTREGOVERLAP__SHIFT) & A4XX_SP_FS_CTRL_REG0_INOUTREGOVERLAP__MASK;
+}
+#define A4XX_SP_FS_CTRL_REG0_THREADSIZE__MASK			0x00100000
+#define A4XX_SP_FS_CTRL_REG0_THREADSIZE__SHIFT			20
+static inline uint32_t A4XX_SP_FS_CTRL_REG0_THREADSIZE(enum a3xx_threadsize val)
+{
+	return ((val) << A4XX_SP_FS_CTRL_REG0_THREADSIZE__SHIFT) & A4XX_SP_FS_CTRL_REG0_THREADSIZE__MASK;
+}
+#define A4XX_SP_FS_CTRL_REG0_SUPERTHREADMODE			0x00200000
+#define A4XX_SP_FS_CTRL_REG0_PIXLODENABLE			0x00400000
+
+#define REG_A4XX_SP_FS_CTRL_REG1				0x000022e9
+#define A4XX_SP_FS_CTRL_REG1_CONSTLENGTH__MASK			0x000000ff
+#define A4XX_SP_FS_CTRL_REG1_CONSTLENGTH__SHIFT			0
+static inline uint32_t A4XX_SP_FS_CTRL_REG1_CONSTLENGTH(uint32_t val)
+{
+	return ((val) << A4XX_SP_FS_CTRL_REG1_CONSTLENGTH__SHIFT) & A4XX_SP_FS_CTRL_REG1_CONSTLENGTH__MASK;
+}
+#define A4XX_SP_FS_CTRL_REG1_VARYING				0x00100000
+
+#define REG_A4XX_SP_FS_OBJ_OFFSET_REG				0x000022ea
+#define A4XX_SP_FS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__MASK	0x01ff0000
+#define A4XX_SP_FS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__SHIFT	16
+static inline uint32_t A4XX_SP_FS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_SP_FS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__SHIFT) & A4XX_SP_FS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__MASK;
+}
+#define A4XX_SP_FS_OBJ_OFFSET_REG_SHADEROBJOFFSET__MASK		0xfe000000
+#define A4XX_SP_FS_OBJ_OFFSET_REG_SHADEROBJOFFSET__SHIFT	25
+static inline uint32_t A4XX_SP_FS_OBJ_OFFSET_REG_SHADEROBJOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_SP_FS_OBJ_OFFSET_REG_SHADEROBJOFFSET__SHIFT) & A4XX_SP_FS_OBJ_OFFSET_REG_SHADEROBJOFFSET__MASK;
+}
+
+#define REG_A4XX_SP_FS_OBJ_START				0x000022eb
+
+#define REG_A4XX_SP_FS_PVT_MEM_PARAM				0x000022ec
+
+#define REG_A4XX_SP_FS_PVT_MEM_ADDR				0x000022ed
+
+#define REG_A4XX_SP_FS_LENGTH_REG				0x000022ef
+
+#define REG_A4XX_SP_FS_OUTPUT_REG				0x000022f0
+#define A4XX_SP_FS_OUTPUT_REG_DEPTH_ENABLE			0x00000080
+#define A4XX_SP_FS_OUTPUT_REG_DEPTH_REGID__MASK			0x0000ff00
+#define A4XX_SP_FS_OUTPUT_REG_DEPTH_REGID__SHIFT		8
+static inline uint32_t A4XX_SP_FS_OUTPUT_REG_DEPTH_REGID(uint32_t val)
+{
+	return ((val) << A4XX_SP_FS_OUTPUT_REG_DEPTH_REGID__SHIFT) & A4XX_SP_FS_OUTPUT_REG_DEPTH_REGID__MASK;
+}
+
+static inline uint32_t REG_A4XX_SP_FS_MRT(uint32_t i0) { return 0x000022f1 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_SP_FS_MRT_REG(uint32_t i0) { return 0x000022f1 + 0x1*i0; }
+#define A4XX_SP_FS_MRT_REG_REGID__MASK				0x000000ff
+#define A4XX_SP_FS_MRT_REG_REGID__SHIFT				0
+static inline uint32_t A4XX_SP_FS_MRT_REG_REGID(uint32_t val)
+{
+	return ((val) << A4XX_SP_FS_MRT_REG_REGID__SHIFT) & A4XX_SP_FS_MRT_REG_REGID__MASK;
+}
+#define A4XX_SP_FS_MRT_REG_HALF_PRECISION			0x00000100
+#define A4XX_SP_FS_MRT_REG_MRTFORMAT__MASK			0x0003f000
+#define A4XX_SP_FS_MRT_REG_MRTFORMAT__SHIFT			12
+static inline uint32_t A4XX_SP_FS_MRT_REG_MRTFORMAT(enum a4xx_color_fmt val)
+{
+	return ((val) << A4XX_SP_FS_MRT_REG_MRTFORMAT__SHIFT) & A4XX_SP_FS_MRT_REG_MRTFORMAT__MASK;
+}
+
+#define REG_A4XX_SP_HS_OBJ_OFFSET_REG				0x0000230d
+#define A4XX_SP_HS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__MASK	0x01ff0000
+#define A4XX_SP_HS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__SHIFT	16
+static inline uint32_t A4XX_SP_HS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_SP_HS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__SHIFT) & A4XX_SP_HS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__MASK;
+}
+#define A4XX_SP_HS_OBJ_OFFSET_REG_SHADEROBJOFFSET__MASK		0xfe000000
+#define A4XX_SP_HS_OBJ_OFFSET_REG_SHADEROBJOFFSET__SHIFT	25
+static inline uint32_t A4XX_SP_HS_OBJ_OFFSET_REG_SHADEROBJOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_SP_HS_OBJ_OFFSET_REG_SHADEROBJOFFSET__SHIFT) & A4XX_SP_HS_OBJ_OFFSET_REG_SHADEROBJOFFSET__MASK;
+}
+
+#define REG_A4XX_SP_DS_OBJ_OFFSET_REG				0x00002334
+#define A4XX_SP_DS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__MASK	0x01ff0000
+#define A4XX_SP_DS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__SHIFT	16
+static inline uint32_t A4XX_SP_DS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_SP_DS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__SHIFT) & A4XX_SP_DS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__MASK;
+}
+#define A4XX_SP_DS_OBJ_OFFSET_REG_SHADEROBJOFFSET__MASK		0xfe000000
+#define A4XX_SP_DS_OBJ_OFFSET_REG_SHADEROBJOFFSET__SHIFT	25
+static inline uint32_t A4XX_SP_DS_OBJ_OFFSET_REG_SHADEROBJOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_SP_DS_OBJ_OFFSET_REG_SHADEROBJOFFSET__SHIFT) & A4XX_SP_DS_OBJ_OFFSET_REG_SHADEROBJOFFSET__MASK;
+}
+
+#define REG_A4XX_SP_GS_OBJ_OFFSET_REG				0x0000235b
+#define A4XX_SP_GS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__MASK	0x01ff0000
+#define A4XX_SP_GS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__SHIFT	16
+static inline uint32_t A4XX_SP_GS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_SP_GS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__SHIFT) & A4XX_SP_GS_OBJ_OFFSET_REG_CONSTOBJECTOFFSET__MASK;
+}
+#define A4XX_SP_GS_OBJ_OFFSET_REG_SHADEROBJOFFSET__MASK		0xfe000000
+#define A4XX_SP_GS_OBJ_OFFSET_REG_SHADEROBJOFFSET__SHIFT	25
+static inline uint32_t A4XX_SP_GS_OBJ_OFFSET_REG_SHADEROBJOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_SP_GS_OBJ_OFFSET_REG_SHADEROBJOFFSET__SHIFT) & A4XX_SP_GS_OBJ_OFFSET_REG_SHADEROBJOFFSET__MASK;
+}
+
+#define REG_A4XX_SP_GS_LENGTH_REG				0x00002360
+
+#define REG_A4XX_VPC_DEBUG_RAM_SEL				0x00000e60
+
+#define REG_A4XX_VPC_DEBUG_RAM_READ				0x00000e61
+
+#define REG_A4XX_VPC_DEBUG_ECO_CONTROL				0x00000e64
+
+#define REG_A4XX_VPC_PERFCTR_VPC_SEL_3				0x00000e68
+
+#define REG_A4XX_VPC_ATTR					0x00002140
+#define A4XX_VPC_ATTR_TOTALATTR__MASK				0x000001ff
+#define A4XX_VPC_ATTR_TOTALATTR__SHIFT				0
+static inline uint32_t A4XX_VPC_ATTR_TOTALATTR(uint32_t val)
+{
+	return ((val) << A4XX_VPC_ATTR_TOTALATTR__SHIFT) & A4XX_VPC_ATTR_TOTALATTR__MASK;
+}
+#define A4XX_VPC_ATTR_PSIZE					0x00000200
+#define A4XX_VPC_ATTR_THRDASSIGN__MASK				0x00003000
+#define A4XX_VPC_ATTR_THRDASSIGN__SHIFT				12
+static inline uint32_t A4XX_VPC_ATTR_THRDASSIGN(uint32_t val)
+{
+	return ((val) << A4XX_VPC_ATTR_THRDASSIGN__SHIFT) & A4XX_VPC_ATTR_THRDASSIGN__MASK;
+}
+#define A4XX_VPC_ATTR_ENABLE					0x02000000
+
+#define REG_A4XX_VPC_PACK					0x00002141
+#define A4XX_VPC_PACK_NUMBYPASSVAR__MASK			0x000000ff
+#define A4XX_VPC_PACK_NUMBYPASSVAR__SHIFT			0
+static inline uint32_t A4XX_VPC_PACK_NUMBYPASSVAR(uint32_t val)
+{
+	return ((val) << A4XX_VPC_PACK_NUMBYPASSVAR__SHIFT) & A4XX_VPC_PACK_NUMBYPASSVAR__MASK;
+}
+#define A4XX_VPC_PACK_NUMFPNONPOSVAR__MASK			0x0000ff00
+#define A4XX_VPC_PACK_NUMFPNONPOSVAR__SHIFT			8
+static inline uint32_t A4XX_VPC_PACK_NUMFPNONPOSVAR(uint32_t val)
+{
+	return ((val) << A4XX_VPC_PACK_NUMFPNONPOSVAR__SHIFT) & A4XX_VPC_PACK_NUMFPNONPOSVAR__MASK;
+}
+#define A4XX_VPC_PACK_NUMNONPOSVSVAR__MASK			0x00ff0000
+#define A4XX_VPC_PACK_NUMNONPOSVSVAR__SHIFT			16
+static inline uint32_t A4XX_VPC_PACK_NUMNONPOSVSVAR(uint32_t val)
+{
+	return ((val) << A4XX_VPC_PACK_NUMNONPOSVSVAR__SHIFT) & A4XX_VPC_PACK_NUMNONPOSVSVAR__MASK;
+}
+
+static inline uint32_t REG_A4XX_VPC_VARYING_INTERP(uint32_t i0) { return 0x00002142 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_VPC_VARYING_INTERP_MODE(uint32_t i0) { return 0x00002142 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_VPC_VARYING_PS_REPL(uint32_t i0) { return 0x0000214a + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_VPC_VARYING_PS_REPL_MODE(uint32_t i0) { return 0x0000214a + 0x1*i0; }
+
+#define REG_A4XX_VPC_SO_FLUSH_WADDR_3				0x0000216e
+
+#define REG_A4XX_VSC_BIN_SIZE					0x00000c00
+#define A4XX_VSC_BIN_SIZE_WIDTH__MASK				0x0000001f
+#define A4XX_VSC_BIN_SIZE_WIDTH__SHIFT				0
+static inline uint32_t A4XX_VSC_BIN_SIZE_WIDTH(uint32_t val)
+{
+	return ((val >> 5) << A4XX_VSC_BIN_SIZE_WIDTH__SHIFT) & A4XX_VSC_BIN_SIZE_WIDTH__MASK;
+}
+#define A4XX_VSC_BIN_SIZE_HEIGHT__MASK				0x000003e0
+#define A4XX_VSC_BIN_SIZE_HEIGHT__SHIFT				5
+static inline uint32_t A4XX_VSC_BIN_SIZE_HEIGHT(uint32_t val)
+{
+	return ((val >> 5) << A4XX_VSC_BIN_SIZE_HEIGHT__SHIFT) & A4XX_VSC_BIN_SIZE_HEIGHT__MASK;
+}
+
+#define REG_A4XX_VSC_SIZE_ADDRESS				0x00000c01
+
+#define REG_A4XX_VSC_SIZE_ADDRESS2				0x00000c02
+
+#define REG_A4XX_VSC_DEBUG_ECO_CONTROL				0x00000c03
+
+static inline uint32_t REG_A4XX_VSC_PIPE_CONFIG(uint32_t i0) { return 0x00000c08 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_VSC_PIPE_CONFIG_REG(uint32_t i0) { return 0x00000c08 + 0x1*i0; }
+#define A4XX_VSC_PIPE_CONFIG_REG_X__MASK			0x000003ff
+#define A4XX_VSC_PIPE_CONFIG_REG_X__SHIFT			0
+static inline uint32_t A4XX_VSC_PIPE_CONFIG_REG_X(uint32_t val)
+{
+	return ((val) << A4XX_VSC_PIPE_CONFIG_REG_X__SHIFT) & A4XX_VSC_PIPE_CONFIG_REG_X__MASK;
+}
+#define A4XX_VSC_PIPE_CONFIG_REG_Y__MASK			0x000ffc00
+#define A4XX_VSC_PIPE_CONFIG_REG_Y__SHIFT			10
+static inline uint32_t A4XX_VSC_PIPE_CONFIG_REG_Y(uint32_t val)
+{
+	return ((val) << A4XX_VSC_PIPE_CONFIG_REG_Y__SHIFT) & A4XX_VSC_PIPE_CONFIG_REG_Y__MASK;
+}
+#define A4XX_VSC_PIPE_CONFIG_REG_W__MASK			0x00f00000
+#define A4XX_VSC_PIPE_CONFIG_REG_W__SHIFT			20
+static inline uint32_t A4XX_VSC_PIPE_CONFIG_REG_W(uint32_t val)
+{
+	return ((val) << A4XX_VSC_PIPE_CONFIG_REG_W__SHIFT) & A4XX_VSC_PIPE_CONFIG_REG_W__MASK;
+}
+#define A4XX_VSC_PIPE_CONFIG_REG_H__MASK			0x0f000000
+#define A4XX_VSC_PIPE_CONFIG_REG_H__SHIFT			24
+static inline uint32_t A4XX_VSC_PIPE_CONFIG_REG_H(uint32_t val)
+{
+	return ((val) << A4XX_VSC_PIPE_CONFIG_REG_H__SHIFT) & A4XX_VSC_PIPE_CONFIG_REG_H__MASK;
+}
+
+static inline uint32_t REG_A4XX_VSC_PIPE_DATA_ADDRESS(uint32_t i0) { return 0x00000c10 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_VSC_PIPE_DATA_ADDRESS_REG(uint32_t i0) { return 0x00000c10 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_VSC_PIPE_DATA_LENGTH(uint32_t i0) { return 0x00000c18 + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_VSC_PIPE_DATA_LENGTH_REG(uint32_t i0) { return 0x00000c18 + 0x1*i0; }
+
+#define REG_A4XX_VSC_PIPE_PARTIAL_POSN_1			0x00000c41
+
+#define REG_A4XX_VSC_PERFCTR_VSC_SEL_0				0x00000c50
+
+#define REG_A4XX_VSC_PERFCTR_VSC_SEL_1				0x00000c51
+
+#define REG_A4XX_VFD_DEBUG_CONTROL				0x00000e40
+
+#define REG_A4XX_VFD_PERFCTR_VFD_SEL_7				0x00000e4a
+
+#define REG_A4XX_VFD_CONTROL_0					0x00002200
+#define A4XX_VFD_CONTROL_0_TOTALATTRTOVS__MASK			0x000000ff
+#define A4XX_VFD_CONTROL_0_TOTALATTRTOVS__SHIFT			0
+static inline uint32_t A4XX_VFD_CONTROL_0_TOTALATTRTOVS(uint32_t val)
+{
+	return ((val) << A4XX_VFD_CONTROL_0_TOTALATTRTOVS__SHIFT) & A4XX_VFD_CONTROL_0_TOTALATTRTOVS__MASK;
+}
+#define A4XX_VFD_CONTROL_0_BYPASSATTROVS__MASK			0x0001fe00
+#define A4XX_VFD_CONTROL_0_BYPASSATTROVS__SHIFT			9
+static inline uint32_t A4XX_VFD_CONTROL_0_BYPASSATTROVS(uint32_t val)
+{
+	return ((val) << A4XX_VFD_CONTROL_0_BYPASSATTROVS__SHIFT) & A4XX_VFD_CONTROL_0_BYPASSATTROVS__MASK;
+}
+#define A4XX_VFD_CONTROL_0_STRMDECINSTRCNT__MASK		0x03f00000
+#define A4XX_VFD_CONTROL_0_STRMDECINSTRCNT__SHIFT		20
+static inline uint32_t A4XX_VFD_CONTROL_0_STRMDECINSTRCNT(uint32_t val)
+{
+	return ((val) << A4XX_VFD_CONTROL_0_STRMDECINSTRCNT__SHIFT) & A4XX_VFD_CONTROL_0_STRMDECINSTRCNT__MASK;
+}
+#define A4XX_VFD_CONTROL_0_STRMFETCHINSTRCNT__MASK		0xfc000000
+#define A4XX_VFD_CONTROL_0_STRMFETCHINSTRCNT__SHIFT		26
+static inline uint32_t A4XX_VFD_CONTROL_0_STRMFETCHINSTRCNT(uint32_t val)
+{
+	return ((val) << A4XX_VFD_CONTROL_0_STRMFETCHINSTRCNT__SHIFT) & A4XX_VFD_CONTROL_0_STRMFETCHINSTRCNT__MASK;
+}
+
+#define REG_A4XX_VFD_CONTROL_1					0x00002201
+#define A4XX_VFD_CONTROL_1_MAXSTORAGE__MASK			0x0000ffff
+#define A4XX_VFD_CONTROL_1_MAXSTORAGE__SHIFT			0
+static inline uint32_t A4XX_VFD_CONTROL_1_MAXSTORAGE(uint32_t val)
+{
+	return ((val) << A4XX_VFD_CONTROL_1_MAXSTORAGE__SHIFT) & A4XX_VFD_CONTROL_1_MAXSTORAGE__MASK;
+}
+#define A4XX_VFD_CONTROL_1_REGID4VTX__MASK			0x00ff0000
+#define A4XX_VFD_CONTROL_1_REGID4VTX__SHIFT			16
+static inline uint32_t A4XX_VFD_CONTROL_1_REGID4VTX(uint32_t val)
+{
+	return ((val) << A4XX_VFD_CONTROL_1_REGID4VTX__SHIFT) & A4XX_VFD_CONTROL_1_REGID4VTX__MASK;
+}
+#define A4XX_VFD_CONTROL_1_REGID4INST__MASK			0xff000000
+#define A4XX_VFD_CONTROL_1_REGID4INST__SHIFT			24
+static inline uint32_t A4XX_VFD_CONTROL_1_REGID4INST(uint32_t val)
+{
+	return ((val) << A4XX_VFD_CONTROL_1_REGID4INST__SHIFT) & A4XX_VFD_CONTROL_1_REGID4INST__MASK;
+}
+
+#define REG_A4XX_VFD_CONTROL_2					0x00002202
+
+#define REG_A4XX_VFD_CONTROL_3					0x00002203
+
+#define REG_A4XX_VFD_CONTROL_4					0x00002204
+
+#define REG_A4XX_VFD_INDEX_OFFSET				0x00002208
+
+static inline uint32_t REG_A4XX_VFD_FETCH(uint32_t i0) { return 0x0000220a + 0x4*i0; }
+
+static inline uint32_t REG_A4XX_VFD_FETCH_INSTR_0(uint32_t i0) { return 0x0000220a + 0x4*i0; }
+#define A4XX_VFD_FETCH_INSTR_0_FETCHSIZE__MASK			0x0000007f
+#define A4XX_VFD_FETCH_INSTR_0_FETCHSIZE__SHIFT			0
+static inline uint32_t A4XX_VFD_FETCH_INSTR_0_FETCHSIZE(uint32_t val)
+{
+	return ((val) << A4XX_VFD_FETCH_INSTR_0_FETCHSIZE__SHIFT) & A4XX_VFD_FETCH_INSTR_0_FETCHSIZE__MASK;
+}
+#define A4XX_VFD_FETCH_INSTR_0_BUFSTRIDE__MASK			0x0001ff80
+#define A4XX_VFD_FETCH_INSTR_0_BUFSTRIDE__SHIFT			7
+static inline uint32_t A4XX_VFD_FETCH_INSTR_0_BUFSTRIDE(uint32_t val)
+{
+	return ((val) << A4XX_VFD_FETCH_INSTR_0_BUFSTRIDE__SHIFT) & A4XX_VFD_FETCH_INSTR_0_BUFSTRIDE__MASK;
+}
+#define A4XX_VFD_FETCH_INSTR_0_SWITCHNEXT			0x00080000
+#define A4XX_VFD_FETCH_INSTR_0_STEPRATE__MASK			0xff000000
+#define A4XX_VFD_FETCH_INSTR_0_STEPRATE__SHIFT			24
+static inline uint32_t A4XX_VFD_FETCH_INSTR_0_STEPRATE(uint32_t val)
+{
+	return ((val) << A4XX_VFD_FETCH_INSTR_0_STEPRATE__SHIFT) & A4XX_VFD_FETCH_INSTR_0_STEPRATE__MASK;
+}
+
+static inline uint32_t REG_A4XX_VFD_FETCH_INSTR_1(uint32_t i0) { return 0x0000220b + 0x4*i0; }
+
+static inline uint32_t REG_A4XX_VFD_FETCH_INSTR_2(uint32_t i0) { return 0x0000220c + 0x4*i0; }
+#define A4XX_VFD_FETCH_INSTR_2_SIZE__MASK			0xfffffff0
+#define A4XX_VFD_FETCH_INSTR_2_SIZE__SHIFT			4
+static inline uint32_t A4XX_VFD_FETCH_INSTR_2_SIZE(uint32_t val)
+{
+	return ((val >> 4) << A4XX_VFD_FETCH_INSTR_2_SIZE__SHIFT) & A4XX_VFD_FETCH_INSTR_2_SIZE__MASK;
+}
+
+static inline uint32_t REG_A4XX_VFD_FETCH_INSTR_3(uint32_t i0) { return 0x0000220d + 0x4*i0; }
+
+static inline uint32_t REG_A4XX_VFD_DECODE(uint32_t i0) { return 0x0000228a + 0x1*i0; }
+
+static inline uint32_t REG_A4XX_VFD_DECODE_INSTR(uint32_t i0) { return 0x0000228a + 0x1*i0; }
+#define A4XX_VFD_DECODE_INSTR_WRITEMASK__MASK			0x0000000f
+#define A4XX_VFD_DECODE_INSTR_WRITEMASK__SHIFT			0
+static inline uint32_t A4XX_VFD_DECODE_INSTR_WRITEMASK(uint32_t val)
+{
+	return ((val) << A4XX_VFD_DECODE_INSTR_WRITEMASK__SHIFT) & A4XX_VFD_DECODE_INSTR_WRITEMASK__MASK;
+}
+#define A4XX_VFD_DECODE_INSTR_CONSTFILL				0x00000010
+#define A4XX_VFD_DECODE_INSTR_FORMAT__MASK			0x00000fc0
+#define A4XX_VFD_DECODE_INSTR_FORMAT__SHIFT			6
+static inline uint32_t A4XX_VFD_DECODE_INSTR_FORMAT(enum a4xx_vtx_fmt val)
+{
+	return ((val) << A4XX_VFD_DECODE_INSTR_FORMAT__SHIFT) & A4XX_VFD_DECODE_INSTR_FORMAT__MASK;
+}
+#define A4XX_VFD_DECODE_INSTR_REGID__MASK			0x000ff000
+#define A4XX_VFD_DECODE_INSTR_REGID__SHIFT			12
+static inline uint32_t A4XX_VFD_DECODE_INSTR_REGID(uint32_t val)
+{
+	return ((val) << A4XX_VFD_DECODE_INSTR_REGID__SHIFT) & A4XX_VFD_DECODE_INSTR_REGID__MASK;
+}
+#define A4XX_VFD_DECODE_INSTR_SWAP__MASK			0x00c00000
+#define A4XX_VFD_DECODE_INSTR_SWAP__SHIFT			22
+static inline uint32_t A4XX_VFD_DECODE_INSTR_SWAP(enum a3xx_color_swap val)
+{
+	return ((val) << A4XX_VFD_DECODE_INSTR_SWAP__SHIFT) & A4XX_VFD_DECODE_INSTR_SWAP__MASK;
+}
+#define A4XX_VFD_DECODE_INSTR_SHIFTCNT__MASK			0x1f000000
+#define A4XX_VFD_DECODE_INSTR_SHIFTCNT__SHIFT			24
+static inline uint32_t A4XX_VFD_DECODE_INSTR_SHIFTCNT(uint32_t val)
+{
+	return ((val) << A4XX_VFD_DECODE_INSTR_SHIFTCNT__SHIFT) & A4XX_VFD_DECODE_INSTR_SHIFTCNT__MASK;
+}
+#define A4XX_VFD_DECODE_INSTR_LASTCOMPVALID			0x20000000
+#define A4XX_VFD_DECODE_INSTR_SWITCHNEXT			0x40000000
+
+#define REG_A4XX_TPL1_DEBUG_ECO_CONTROL				0x00000f00
+
+#define REG_A4XX_TPL1_PERFCTR_TP_SEL_7				0x00000f0b
+
+#define REG_A4XX_TPL1_TP_TEX_OFFSET				0x00002380
+
+#define REG_A4XX_TPL1_TP_CS_TEXMEMOBJ_BASE_ADDR			0x000023a6
+
+#define REG_A4XX_GRAS_TSE_STATUS				0x00000c80
+
+#define REG_A4XX_GRAS_DEBUG_ECO_CONTROL				0x00000c81
+
+#define REG_A4XX_GRAS_PERFCTR_TSE_SEL_0				0x00000c88
+
+#define REG_A4XX_GRAS_PERFCTR_TSE_SEL_3				0x00000c8b
+
+#define REG_A4XX_GRAS_CL_CLIP_CNTL				0x00002000
+
+#define REG_A4XX_GRAS_CLEAR_CNTL				0x00002003
+#define A4XX_GRAS_CLEAR_CNTL_NOT_FASTCLEAR			0x00000001
+
+#define REG_A4XX_GRAS_CL_GB_CLIP_ADJ				0x00002004
+#define A4XX_GRAS_CL_GB_CLIP_ADJ_HORZ__MASK			0x000003ff
+#define A4XX_GRAS_CL_GB_CLIP_ADJ_HORZ__SHIFT			0
+static inline uint32_t A4XX_GRAS_CL_GB_CLIP_ADJ_HORZ(uint32_t val)
+{
+	return ((val) << A4XX_GRAS_CL_GB_CLIP_ADJ_HORZ__SHIFT) & A4XX_GRAS_CL_GB_CLIP_ADJ_HORZ__MASK;
+}
+#define A4XX_GRAS_CL_GB_CLIP_ADJ_VERT__MASK			0x000ffc00
+#define A4XX_GRAS_CL_GB_CLIP_ADJ_VERT__SHIFT			10
+static inline uint32_t A4XX_GRAS_CL_GB_CLIP_ADJ_VERT(uint32_t val)
+{
+	return ((val) << A4XX_GRAS_CL_GB_CLIP_ADJ_VERT__SHIFT) & A4XX_GRAS_CL_GB_CLIP_ADJ_VERT__MASK;
+}
+
+#define REG_A4XX_GRAS_CL_VPORT_XOFFSET_0			0x00002008
+#define A4XX_GRAS_CL_VPORT_XOFFSET_0__MASK			0xffffffff
+#define A4XX_GRAS_CL_VPORT_XOFFSET_0__SHIFT			0
+static inline uint32_t A4XX_GRAS_CL_VPORT_XOFFSET_0(float val)
+{
+	return ((fui(val)) << A4XX_GRAS_CL_VPORT_XOFFSET_0__SHIFT) & A4XX_GRAS_CL_VPORT_XOFFSET_0__MASK;
+}
+
+#define REG_A4XX_GRAS_CL_VPORT_XSCALE_0				0x00002009
+#define A4XX_GRAS_CL_VPORT_XSCALE_0__MASK			0xffffffff
+#define A4XX_GRAS_CL_VPORT_XSCALE_0__SHIFT			0
+static inline uint32_t A4XX_GRAS_CL_VPORT_XSCALE_0(float val)
+{
+	return ((fui(val)) << A4XX_GRAS_CL_VPORT_XSCALE_0__SHIFT) & A4XX_GRAS_CL_VPORT_XSCALE_0__MASK;
+}
+
+#define REG_A4XX_GRAS_CL_VPORT_YOFFSET_0			0x0000200a
+#define A4XX_GRAS_CL_VPORT_YOFFSET_0__MASK			0xffffffff
+#define A4XX_GRAS_CL_VPORT_YOFFSET_0__SHIFT			0
+static inline uint32_t A4XX_GRAS_CL_VPORT_YOFFSET_0(float val)
+{
+	return ((fui(val)) << A4XX_GRAS_CL_VPORT_YOFFSET_0__SHIFT) & A4XX_GRAS_CL_VPORT_YOFFSET_0__MASK;
+}
+
+#define REG_A4XX_GRAS_CL_VPORT_YSCALE_0				0x0000200b
+#define A4XX_GRAS_CL_VPORT_YSCALE_0__MASK			0xffffffff
+#define A4XX_GRAS_CL_VPORT_YSCALE_0__SHIFT			0
+static inline uint32_t A4XX_GRAS_CL_VPORT_YSCALE_0(float val)
+{
+	return ((fui(val)) << A4XX_GRAS_CL_VPORT_YSCALE_0__SHIFT) & A4XX_GRAS_CL_VPORT_YSCALE_0__MASK;
+}
+
+#define REG_A4XX_GRAS_CL_VPORT_ZOFFSET_0			0x0000200c
+#define A4XX_GRAS_CL_VPORT_ZOFFSET_0__MASK			0xffffffff
+#define A4XX_GRAS_CL_VPORT_ZOFFSET_0__SHIFT			0
+static inline uint32_t A4XX_GRAS_CL_VPORT_ZOFFSET_0(float val)
+{
+	return ((fui(val)) << A4XX_GRAS_CL_VPORT_ZOFFSET_0__SHIFT) & A4XX_GRAS_CL_VPORT_ZOFFSET_0__MASK;
+}
+
+#define REG_A4XX_GRAS_CL_VPORT_ZSCALE_0				0x0000200d
+#define A4XX_GRAS_CL_VPORT_ZSCALE_0__MASK			0xffffffff
+#define A4XX_GRAS_CL_VPORT_ZSCALE_0__SHIFT			0
+static inline uint32_t A4XX_GRAS_CL_VPORT_ZSCALE_0(float val)
+{
+	return ((fui(val)) << A4XX_GRAS_CL_VPORT_ZSCALE_0__SHIFT) & A4XX_GRAS_CL_VPORT_ZSCALE_0__MASK;
+}
+
+#define REG_A4XX_GRAS_SU_POINT_MINMAX				0x00002070
+#define A4XX_GRAS_SU_POINT_MINMAX_MIN__MASK			0x0000ffff
+#define A4XX_GRAS_SU_POINT_MINMAX_MIN__SHIFT			0
+static inline uint32_t A4XX_GRAS_SU_POINT_MINMAX_MIN(float val)
+{
+	return ((((uint32_t)(val * 16.0))) << A4XX_GRAS_SU_POINT_MINMAX_MIN__SHIFT) & A4XX_GRAS_SU_POINT_MINMAX_MIN__MASK;
+}
+#define A4XX_GRAS_SU_POINT_MINMAX_MAX__MASK			0xffff0000
+#define A4XX_GRAS_SU_POINT_MINMAX_MAX__SHIFT			16
+static inline uint32_t A4XX_GRAS_SU_POINT_MINMAX_MAX(float val)
+{
+	return ((((uint32_t)(val * 16.0))) << A4XX_GRAS_SU_POINT_MINMAX_MAX__SHIFT) & A4XX_GRAS_SU_POINT_MINMAX_MAX__MASK;
+}
+
+#define REG_A4XX_GRAS_SU_POINT_SIZE				0x00002071
+#define A4XX_GRAS_SU_POINT_SIZE__MASK				0xffffffff
+#define A4XX_GRAS_SU_POINT_SIZE__SHIFT				0
+static inline uint32_t A4XX_GRAS_SU_POINT_SIZE(float val)
+{
+	return ((((int32_t)(val * 16.0))) << A4XX_GRAS_SU_POINT_SIZE__SHIFT) & A4XX_GRAS_SU_POINT_SIZE__MASK;
+}
+
+#define REG_A4XX_GRAS_ALPHA_CONTROL				0x00002073
+#define A4XX_GRAS_ALPHA_CONTROL_ALPHA_TEST_ENABLE		0x00000004
+
+#define REG_A4XX_GRAS_SU_POLY_OFFSET_SCALE			0x00002074
+#define A4XX_GRAS_SU_POLY_OFFSET_SCALE__MASK			0xffffffff
+#define A4XX_GRAS_SU_POLY_OFFSET_SCALE__SHIFT			0
+static inline uint32_t A4XX_GRAS_SU_POLY_OFFSET_SCALE(float val)
+{
+	return ((fui(val)) << A4XX_GRAS_SU_POLY_OFFSET_SCALE__SHIFT) & A4XX_GRAS_SU_POLY_OFFSET_SCALE__MASK;
+}
+
+#define REG_A4XX_GRAS_SU_POLY_OFFSET_OFFSET			0x00002075
+#define A4XX_GRAS_SU_POLY_OFFSET_OFFSET__MASK			0xffffffff
+#define A4XX_GRAS_SU_POLY_OFFSET_OFFSET__SHIFT			0
+static inline uint32_t A4XX_GRAS_SU_POLY_OFFSET_OFFSET(float val)
+{
+	return ((fui(val)) << A4XX_GRAS_SU_POLY_OFFSET_OFFSET__SHIFT) & A4XX_GRAS_SU_POLY_OFFSET_OFFSET__MASK;
+}
+
+#define REG_A4XX_GRAS_SC_EXTENT_WINDOW_TL			0x0000209f
+
+#define REG_A4XX_GRAS_SC_SCREEN_SCISSOR_TL			0x0000207c
+#define A4XX_GRAS_SC_SCREEN_SCISSOR_TL_WINDOW_OFFSET_DISABLE	0x80000000
+#define A4XX_GRAS_SC_SCREEN_SCISSOR_TL_X__MASK			0x00007fff
+#define A4XX_GRAS_SC_SCREEN_SCISSOR_TL_X__SHIFT			0
+static inline uint32_t A4XX_GRAS_SC_SCREEN_SCISSOR_TL_X(uint32_t val)
+{
+	return ((val) << A4XX_GRAS_SC_SCREEN_SCISSOR_TL_X__SHIFT) & A4XX_GRAS_SC_SCREEN_SCISSOR_TL_X__MASK;
+}
+#define A4XX_GRAS_SC_SCREEN_SCISSOR_TL_Y__MASK			0x7fff0000
+#define A4XX_GRAS_SC_SCREEN_SCISSOR_TL_Y__SHIFT			16
+static inline uint32_t A4XX_GRAS_SC_SCREEN_SCISSOR_TL_Y(uint32_t val)
+{
+	return ((val) << A4XX_GRAS_SC_SCREEN_SCISSOR_TL_Y__SHIFT) & A4XX_GRAS_SC_SCREEN_SCISSOR_TL_Y__MASK;
+}
+
+#define REG_A4XX_GRAS_SC_SCREEN_SCISSOR_BR			0x0000207d
+#define A4XX_GRAS_SC_SCREEN_SCISSOR_BR_WINDOW_OFFSET_DISABLE	0x80000000
+#define A4XX_GRAS_SC_SCREEN_SCISSOR_BR_X__MASK			0x00007fff
+#define A4XX_GRAS_SC_SCREEN_SCISSOR_BR_X__SHIFT			0
+static inline uint32_t A4XX_GRAS_SC_SCREEN_SCISSOR_BR_X(uint32_t val)
+{
+	return ((val) << A4XX_GRAS_SC_SCREEN_SCISSOR_BR_X__SHIFT) & A4XX_GRAS_SC_SCREEN_SCISSOR_BR_X__MASK;
+}
+#define A4XX_GRAS_SC_SCREEN_SCISSOR_BR_Y__MASK			0x7fff0000
+#define A4XX_GRAS_SC_SCREEN_SCISSOR_BR_Y__SHIFT			16
+static inline uint32_t A4XX_GRAS_SC_SCREEN_SCISSOR_BR_Y(uint32_t val)
+{
+	return ((val) << A4XX_GRAS_SC_SCREEN_SCISSOR_BR_Y__SHIFT) & A4XX_GRAS_SC_SCREEN_SCISSOR_BR_Y__MASK;
+}
+
+#define REG_A4XX_GRAS_SC_WINDOW_SCISSOR_BR			0x0000209c
+#define A4XX_GRAS_SC_WINDOW_SCISSOR_BR_WINDOW_OFFSET_DISABLE	0x80000000
+#define A4XX_GRAS_SC_WINDOW_SCISSOR_BR_X__MASK			0x00007fff
+#define A4XX_GRAS_SC_WINDOW_SCISSOR_BR_X__SHIFT			0
+static inline uint32_t A4XX_GRAS_SC_WINDOW_SCISSOR_BR_X(uint32_t val)
+{
+	return ((val) << A4XX_GRAS_SC_WINDOW_SCISSOR_BR_X__SHIFT) & A4XX_GRAS_SC_WINDOW_SCISSOR_BR_X__MASK;
+}
+#define A4XX_GRAS_SC_WINDOW_SCISSOR_BR_Y__MASK			0x7fff0000
+#define A4XX_GRAS_SC_WINDOW_SCISSOR_BR_Y__SHIFT			16
+static inline uint32_t A4XX_GRAS_SC_WINDOW_SCISSOR_BR_Y(uint32_t val)
+{
+	return ((val) << A4XX_GRAS_SC_WINDOW_SCISSOR_BR_Y__SHIFT) & A4XX_GRAS_SC_WINDOW_SCISSOR_BR_Y__MASK;
+}
+
+#define REG_A4XX_GRAS_SC_WINDOW_SCISSOR_TL			0x0000209d
+#define A4XX_GRAS_SC_WINDOW_SCISSOR_TL_WINDOW_OFFSET_DISABLE	0x80000000
+#define A4XX_GRAS_SC_WINDOW_SCISSOR_TL_X__MASK			0x00007fff
+#define A4XX_GRAS_SC_WINDOW_SCISSOR_TL_X__SHIFT			0
+static inline uint32_t A4XX_GRAS_SC_WINDOW_SCISSOR_TL_X(uint32_t val)
+{
+	return ((val) << A4XX_GRAS_SC_WINDOW_SCISSOR_TL_X__SHIFT) & A4XX_GRAS_SC_WINDOW_SCISSOR_TL_X__MASK;
+}
+#define A4XX_GRAS_SC_WINDOW_SCISSOR_TL_Y__MASK			0x7fff0000
+#define A4XX_GRAS_SC_WINDOW_SCISSOR_TL_Y__SHIFT			16
+static inline uint32_t A4XX_GRAS_SC_WINDOW_SCISSOR_TL_Y(uint32_t val)
+{
+	return ((val) << A4XX_GRAS_SC_WINDOW_SCISSOR_TL_Y__SHIFT) & A4XX_GRAS_SC_WINDOW_SCISSOR_TL_Y__MASK;
+}
+
+#define REG_A4XX_GRAS_DEPTH_CONTROL				0x00002077
+#define A4XX_GRAS_DEPTH_CONTROL_FORMAT__MASK			0x00000003
+#define A4XX_GRAS_DEPTH_CONTROL_FORMAT__SHIFT			0
+static inline uint32_t A4XX_GRAS_DEPTH_CONTROL_FORMAT(enum a4xx_depth_format val)
+{
+	return ((val) << A4XX_GRAS_DEPTH_CONTROL_FORMAT__SHIFT) & A4XX_GRAS_DEPTH_CONTROL_FORMAT__MASK;
+}
+
+#define REG_A4XX_GRAS_SU_MODE_CONTROL				0x00002078
+#define A4XX_GRAS_SU_MODE_CONTROL_CULL_FRONT			0x00000001
+#define A4XX_GRAS_SU_MODE_CONTROL_CULL_BACK			0x00000002
+#define A4XX_GRAS_SU_MODE_CONTROL_FRONT_CW			0x00000004
+#define A4XX_GRAS_SU_MODE_CONTROL_LINEHALFWIDTH__MASK		0x000007f8
+#define A4XX_GRAS_SU_MODE_CONTROL_LINEHALFWIDTH__SHIFT		3
+static inline uint32_t A4XX_GRAS_SU_MODE_CONTROL_LINEHALFWIDTH(float val)
+{
+	return ((((int32_t)(val * 4.0))) << A4XX_GRAS_SU_MODE_CONTROL_LINEHALFWIDTH__SHIFT) & A4XX_GRAS_SU_MODE_CONTROL_LINEHALFWIDTH__MASK;
+}
+#define A4XX_GRAS_SU_MODE_CONTROL_POLY_OFFSET			0x00000800
+#define A4XX_GRAS_SU_MODE_CONTROL_RENDERING_PASS		0x00100000
+
+#define REG_A4XX_GRAS_SC_CONTROL				0x0000207b
+#define A4XX_GRAS_SC_CONTROL_RENDER_MODE__MASK			0x0000000c
+#define A4XX_GRAS_SC_CONTROL_RENDER_MODE__SHIFT			2
+static inline uint32_t A4XX_GRAS_SC_CONTROL_RENDER_MODE(enum a3xx_render_mode val)
+{
+	return ((val) << A4XX_GRAS_SC_CONTROL_RENDER_MODE__SHIFT) & A4XX_GRAS_SC_CONTROL_RENDER_MODE__MASK;
+}
+#define A4XX_GRAS_SC_CONTROL_MSAA_SAMPLES__MASK			0x00000380
+#define A4XX_GRAS_SC_CONTROL_MSAA_SAMPLES__SHIFT		7
+static inline uint32_t A4XX_GRAS_SC_CONTROL_MSAA_SAMPLES(uint32_t val)
+{
+	return ((val) << A4XX_GRAS_SC_CONTROL_MSAA_SAMPLES__SHIFT) & A4XX_GRAS_SC_CONTROL_MSAA_SAMPLES__MASK;
+}
+#define A4XX_GRAS_SC_CONTROL_MSAA_DISABLE			0x00000800
+#define A4XX_GRAS_SC_CONTROL_RASTER_MODE__MASK			0x0000f000
+#define A4XX_GRAS_SC_CONTROL_RASTER_MODE__SHIFT			12
+static inline uint32_t A4XX_GRAS_SC_CONTROL_RASTER_MODE(uint32_t val)
+{
+	return ((val) << A4XX_GRAS_SC_CONTROL_RASTER_MODE__SHIFT) & A4XX_GRAS_SC_CONTROL_RASTER_MODE__MASK;
+}
+
+#define REG_A4XX_UCHE_CACHE_MODE_CONTROL			0x00000e80
+
+#define REG_A4XX_UCHE_TRAP_BASE_LO				0x00000e83
+
+#define REG_A4XX_UCHE_TRAP_BASE_HI				0x00000e84
+
+#define REG_A4XX_UCHE_CACHE_STATUS				0x00000e88
+
+#define REG_A4XX_UCHE_INVALIDATE0				0x00000e8a
+
+#define REG_A4XX_UCHE_INVALIDATE1				0x00000e8b
+
+#define REG_A4XX_UCHE_CACHE_WAYS_VFD				0x00000e8c
+
+#define REG_A4XX_UCHE_PERFCTR_UCHE_SEL_7			0x00000e95
+
+#define REG_A4XX_HLSQ_TIMEOUT_THRESHOLD				0x00000e00
+
+#define REG_A4XX_HLSQ_DEBUG_ECO_CONTROL				0x00000e04
+
+#define REG_A4XX_HLSQ_PERF_PIPE_MASK				0x00000e0e
+
+#define REG_A4XX_HLSQ_CONTROL_0_REG				0x000023c0
+#define A4XX_HLSQ_CONTROL_0_REG_FSTHREADSIZE__MASK		0x00000010
+#define A4XX_HLSQ_CONTROL_0_REG_FSTHREADSIZE__SHIFT		4
+static inline uint32_t A4XX_HLSQ_CONTROL_0_REG_FSTHREADSIZE(enum a3xx_threadsize val)
+{
+	return ((val) << A4XX_HLSQ_CONTROL_0_REG_FSTHREADSIZE__SHIFT) & A4XX_HLSQ_CONTROL_0_REG_FSTHREADSIZE__MASK;
+}
+#define A4XX_HLSQ_CONTROL_0_REG_FSSUPERTHREADENABLE		0x00000040
+#define A4XX_HLSQ_CONTROL_0_REG_SPSHADERRESTART			0x00000200
+#define A4XX_HLSQ_CONTROL_0_REG_RESERVED2			0x00000400
+#define A4XX_HLSQ_CONTROL_0_REG_CHUNKDISABLE			0x04000000
+#define A4XX_HLSQ_CONTROL_0_REG_CONSTMODE__MASK			0x08000000
+#define A4XX_HLSQ_CONTROL_0_REG_CONSTMODE__SHIFT		27
+static inline uint32_t A4XX_HLSQ_CONTROL_0_REG_CONSTMODE(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_CONTROL_0_REG_CONSTMODE__SHIFT) & A4XX_HLSQ_CONTROL_0_REG_CONSTMODE__MASK;
+}
+#define A4XX_HLSQ_CONTROL_0_REG_LAZYUPDATEDISABLE		0x10000000
+#define A4XX_HLSQ_CONTROL_0_REG_SPCONSTFULLUPDATE		0x20000000
+#define A4XX_HLSQ_CONTROL_0_REG_TPFULLUPDATE			0x40000000
+#define A4XX_HLSQ_CONTROL_0_REG_SINGLECONTEXT			0x80000000
+
+#define REG_A4XX_HLSQ_CONTROL_1_REG				0x000023c1
+#define A4XX_HLSQ_CONTROL_1_REG_VSTHREADSIZE__MASK		0x00000040
+#define A4XX_HLSQ_CONTROL_1_REG_VSTHREADSIZE__SHIFT		6
+static inline uint32_t A4XX_HLSQ_CONTROL_1_REG_VSTHREADSIZE(enum a3xx_threadsize val)
+{
+	return ((val) << A4XX_HLSQ_CONTROL_1_REG_VSTHREADSIZE__SHIFT) & A4XX_HLSQ_CONTROL_1_REG_VSTHREADSIZE__MASK;
+}
+#define A4XX_HLSQ_CONTROL_1_REG_VSSUPERTHREADENABLE		0x00000100
+#define A4XX_HLSQ_CONTROL_1_REG_RESERVED1			0x00000200
+#define A4XX_HLSQ_CONTROL_1_REG_ZWCOORD				0x02000000
+
+#define REG_A4XX_HLSQ_CONTROL_2_REG				0x000023c2
+#define A4XX_HLSQ_CONTROL_2_REG_PRIMALLOCTHRESHOLD__MASK	0xfc000000
+#define A4XX_HLSQ_CONTROL_2_REG_PRIMALLOCTHRESHOLD__SHIFT	26
+static inline uint32_t A4XX_HLSQ_CONTROL_2_REG_PRIMALLOCTHRESHOLD(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_CONTROL_2_REG_PRIMALLOCTHRESHOLD__SHIFT) & A4XX_HLSQ_CONTROL_2_REG_PRIMALLOCTHRESHOLD__MASK;
+}
+
+#define REG_A4XX_HLSQ_CONTROL_3_REG				0x000023c3
+#define A4XX_HLSQ_CONTROL_3_REG_REGID__MASK			0x000000ff
+#define A4XX_HLSQ_CONTROL_3_REG_REGID__SHIFT			0
+static inline uint32_t A4XX_HLSQ_CONTROL_3_REG_REGID(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_CONTROL_3_REG_REGID__SHIFT) & A4XX_HLSQ_CONTROL_3_REG_REGID__MASK;
+}
+
+#define REG_A4XX_HLSQ_VS_CONTROL_REG				0x000023c5
+#define A4XX_HLSQ_VS_CONTROL_REG_CONSTLENGTH__MASK		0x000000ff
+#define A4XX_HLSQ_VS_CONTROL_REG_CONSTLENGTH__SHIFT		0
+static inline uint32_t A4XX_HLSQ_VS_CONTROL_REG_CONSTLENGTH(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_VS_CONTROL_REG_CONSTLENGTH__SHIFT) & A4XX_HLSQ_VS_CONTROL_REG_CONSTLENGTH__MASK;
+}
+#define A4XX_HLSQ_VS_CONTROL_REG_CONSTOBJECTOFFSET__MASK	0x0000ff00
+#define A4XX_HLSQ_VS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT	8
+static inline uint32_t A4XX_HLSQ_VS_CONTROL_REG_CONSTOBJECTOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_VS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT) & A4XX_HLSQ_VS_CONTROL_REG_CONSTOBJECTOFFSET__MASK;
+}
+#define A4XX_HLSQ_VS_CONTROL_REG_SHADEROBJOFFSET__MASK		0x00fe0000
+#define A4XX_HLSQ_VS_CONTROL_REG_SHADEROBJOFFSET__SHIFT		17
+static inline uint32_t A4XX_HLSQ_VS_CONTROL_REG_SHADEROBJOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_VS_CONTROL_REG_SHADEROBJOFFSET__SHIFT) & A4XX_HLSQ_VS_CONTROL_REG_SHADEROBJOFFSET__MASK;
+}
+#define A4XX_HLSQ_VS_CONTROL_REG_INSTRLENGTH__MASK		0xff000000
+#define A4XX_HLSQ_VS_CONTROL_REG_INSTRLENGTH__SHIFT		24
+static inline uint32_t A4XX_HLSQ_VS_CONTROL_REG_INSTRLENGTH(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_VS_CONTROL_REG_INSTRLENGTH__SHIFT) & A4XX_HLSQ_VS_CONTROL_REG_INSTRLENGTH__MASK;
+}
+
+#define REG_A4XX_HLSQ_FS_CONTROL_REG				0x000023c6
+#define A4XX_HLSQ_FS_CONTROL_REG_CONSTLENGTH__MASK		0x000000ff
+#define A4XX_HLSQ_FS_CONTROL_REG_CONSTLENGTH__SHIFT		0
+static inline uint32_t A4XX_HLSQ_FS_CONTROL_REG_CONSTLENGTH(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_FS_CONTROL_REG_CONSTLENGTH__SHIFT) & A4XX_HLSQ_FS_CONTROL_REG_CONSTLENGTH__MASK;
+}
+#define A4XX_HLSQ_FS_CONTROL_REG_CONSTOBJECTOFFSET__MASK	0x0000ff00
+#define A4XX_HLSQ_FS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT	8
+static inline uint32_t A4XX_HLSQ_FS_CONTROL_REG_CONSTOBJECTOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_FS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT) & A4XX_HLSQ_FS_CONTROL_REG_CONSTOBJECTOFFSET__MASK;
+}
+#define A4XX_HLSQ_FS_CONTROL_REG_SHADEROBJOFFSET__MASK		0x00fe0000
+#define A4XX_HLSQ_FS_CONTROL_REG_SHADEROBJOFFSET__SHIFT		17
+static inline uint32_t A4XX_HLSQ_FS_CONTROL_REG_SHADEROBJOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_FS_CONTROL_REG_SHADEROBJOFFSET__SHIFT) & A4XX_HLSQ_FS_CONTROL_REG_SHADEROBJOFFSET__MASK;
+}
+#define A4XX_HLSQ_FS_CONTROL_REG_INSTRLENGTH__MASK		0xff000000
+#define A4XX_HLSQ_FS_CONTROL_REG_INSTRLENGTH__SHIFT		24
+static inline uint32_t A4XX_HLSQ_FS_CONTROL_REG_INSTRLENGTH(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_FS_CONTROL_REG_INSTRLENGTH__SHIFT) & A4XX_HLSQ_FS_CONTROL_REG_INSTRLENGTH__MASK;
+}
+
+#define REG_A4XX_HLSQ_HS_CONTROL_REG				0x000023c7
+#define A4XX_HLSQ_HS_CONTROL_REG_CONSTLENGTH__MASK		0x000000ff
+#define A4XX_HLSQ_HS_CONTROL_REG_CONSTLENGTH__SHIFT		0
+static inline uint32_t A4XX_HLSQ_HS_CONTROL_REG_CONSTLENGTH(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_HS_CONTROL_REG_CONSTLENGTH__SHIFT) & A4XX_HLSQ_HS_CONTROL_REG_CONSTLENGTH__MASK;
+}
+#define A4XX_HLSQ_HS_CONTROL_REG_CONSTOBJECTOFFSET__MASK	0x0000ff00
+#define A4XX_HLSQ_HS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT	8
+static inline uint32_t A4XX_HLSQ_HS_CONTROL_REG_CONSTOBJECTOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_HS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT) & A4XX_HLSQ_HS_CONTROL_REG_CONSTOBJECTOFFSET__MASK;
+}
+#define A4XX_HLSQ_HS_CONTROL_REG_SHADEROBJOFFSET__MASK		0x00fe0000
+#define A4XX_HLSQ_HS_CONTROL_REG_SHADEROBJOFFSET__SHIFT		17
+static inline uint32_t A4XX_HLSQ_HS_CONTROL_REG_SHADEROBJOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_HS_CONTROL_REG_SHADEROBJOFFSET__SHIFT) & A4XX_HLSQ_HS_CONTROL_REG_SHADEROBJOFFSET__MASK;
+}
+#define A4XX_HLSQ_HS_CONTROL_REG_INSTRLENGTH__MASK		0xff000000
+#define A4XX_HLSQ_HS_CONTROL_REG_INSTRLENGTH__SHIFT		24
+static inline uint32_t A4XX_HLSQ_HS_CONTROL_REG_INSTRLENGTH(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_HS_CONTROL_REG_INSTRLENGTH__SHIFT) & A4XX_HLSQ_HS_CONTROL_REG_INSTRLENGTH__MASK;
+}
+
+#define REG_A4XX_HLSQ_DS_CONTROL_REG				0x000023c8
+#define A4XX_HLSQ_DS_CONTROL_REG_CONSTLENGTH__MASK		0x000000ff
+#define A4XX_HLSQ_DS_CONTROL_REG_CONSTLENGTH__SHIFT		0
+static inline uint32_t A4XX_HLSQ_DS_CONTROL_REG_CONSTLENGTH(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_DS_CONTROL_REG_CONSTLENGTH__SHIFT) & A4XX_HLSQ_DS_CONTROL_REG_CONSTLENGTH__MASK;
+}
+#define A4XX_HLSQ_DS_CONTROL_REG_CONSTOBJECTOFFSET__MASK	0x0000ff00
+#define A4XX_HLSQ_DS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT	8
+static inline uint32_t A4XX_HLSQ_DS_CONTROL_REG_CONSTOBJECTOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_DS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT) & A4XX_HLSQ_DS_CONTROL_REG_CONSTOBJECTOFFSET__MASK;
+}
+#define A4XX_HLSQ_DS_CONTROL_REG_SHADEROBJOFFSET__MASK		0x00fe0000
+#define A4XX_HLSQ_DS_CONTROL_REG_SHADEROBJOFFSET__SHIFT		17
+static inline uint32_t A4XX_HLSQ_DS_CONTROL_REG_SHADEROBJOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_DS_CONTROL_REG_SHADEROBJOFFSET__SHIFT) & A4XX_HLSQ_DS_CONTROL_REG_SHADEROBJOFFSET__MASK;
+}
+#define A4XX_HLSQ_DS_CONTROL_REG_INSTRLENGTH__MASK		0xff000000
+#define A4XX_HLSQ_DS_CONTROL_REG_INSTRLENGTH__SHIFT		24
+static inline uint32_t A4XX_HLSQ_DS_CONTROL_REG_INSTRLENGTH(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_DS_CONTROL_REG_INSTRLENGTH__SHIFT) & A4XX_HLSQ_DS_CONTROL_REG_INSTRLENGTH__MASK;
+}
+
+#define REG_A4XX_HLSQ_GS_CONTROL_REG				0x000023c9
+#define A4XX_HLSQ_GS_CONTROL_REG_CONSTLENGTH__MASK		0x000000ff
+#define A4XX_HLSQ_GS_CONTROL_REG_CONSTLENGTH__SHIFT		0
+static inline uint32_t A4XX_HLSQ_GS_CONTROL_REG_CONSTLENGTH(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_GS_CONTROL_REG_CONSTLENGTH__SHIFT) & A4XX_HLSQ_GS_CONTROL_REG_CONSTLENGTH__MASK;
+}
+#define A4XX_HLSQ_GS_CONTROL_REG_CONSTOBJECTOFFSET__MASK	0x0000ff00
+#define A4XX_HLSQ_GS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT	8
+static inline uint32_t A4XX_HLSQ_GS_CONTROL_REG_CONSTOBJECTOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_GS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT) & A4XX_HLSQ_GS_CONTROL_REG_CONSTOBJECTOFFSET__MASK;
+}
+#define A4XX_HLSQ_GS_CONTROL_REG_SHADEROBJOFFSET__MASK		0x00fe0000
+#define A4XX_HLSQ_GS_CONTROL_REG_SHADEROBJOFFSET__SHIFT		17
+static inline uint32_t A4XX_HLSQ_GS_CONTROL_REG_SHADEROBJOFFSET(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_GS_CONTROL_REG_SHADEROBJOFFSET__SHIFT) & A4XX_HLSQ_GS_CONTROL_REG_SHADEROBJOFFSET__MASK;
+}
+#define A4XX_HLSQ_GS_CONTROL_REG_INSTRLENGTH__MASK		0xff000000
+#define A4XX_HLSQ_GS_CONTROL_REG_INSTRLENGTH__SHIFT		24
+static inline uint32_t A4XX_HLSQ_GS_CONTROL_REG_INSTRLENGTH(uint32_t val)
+{
+	return ((val) << A4XX_HLSQ_GS_CONTROL_REG_INSTRLENGTH__SHIFT) & A4XX_HLSQ_GS_CONTROL_REG_INSTRLENGTH__MASK;
+}
+
+#define REG_A4XX_HLSQ_UPDATE_CONTROL				0x000023db
+
+#define REG_A4XX_PC_BINNING_COMMAND				0x00000d00
+#define A4XX_PC_BINNING_COMMAND_BINNING_ENABLE			0x00000001
+
+#define REG_A4XX_PC_DRAWCALL_SETUP_OVERRIDE			0x00000d0c
+
+#define REG_A4XX_PC_PERFCTR_PC_SEL_0				0x00000d10
+
+#define REG_A4XX_PC_PERFCTR_PC_SEL_7				0x00000d17
+
+#define REG_A4XX_PC_BIN_BASE					0x000021c0
+
+#define REG_A4XX_PC_PRIM_VTX_CNTL				0x000021c4
+#define A4XX_PC_PRIM_VTX_CNTL_VAROUT				0x00000001
+#define A4XX_PC_PRIM_VTX_CNTL_PROVOKING_VTX_LAST		0x02000000
+#define A4XX_PC_PRIM_VTX_CNTL_PSIZE				0x04000000
+
+#define REG_A4XX_UNKNOWN_21C5					0x000021c5
+
+#define REG_A4XX_PC_RESTART_INDEX				0x000021c6
+
+#define REG_A4XX_PC_GS_PARAM					0x000021e5
+
+#define REG_A4XX_PC_HS_PARAM					0x000021e7
+
+#define REG_A4XX_VBIF_VERSION					0x00003000
+
+#define REG_A4XX_VBIF_CLKON					0x00003001
+#define A4XX_VBIF_CLKON_FORCE_ON_TESTBUS			0x00000001
+
+#define REG_A4XX_VBIF_ABIT_SORT					0x0000301c
+
+#define REG_A4XX_VBIF_ABIT_SORT_CONF				0x0000301d
+
+#define REG_A4XX_VBIF_GATE_OFF_WRREQ_EN				0x0000302a
+
+#define REG_A4XX_VBIF_IN_RD_LIM_CONF0				0x0000302c
+
+#define REG_A4XX_VBIF_IN_RD_LIM_CONF1				0x0000302d
+
+#define REG_A4XX_VBIF_IN_WR_LIM_CONF0				0x00003030
+
+#define REG_A4XX_VBIF_IN_WR_LIM_CONF1				0x00003031
+
+#define REG_A4XX_VBIF_ROUND_ROBIN_QOS_ARB			0x00003049
+
+#define REG_A4XX_UNKNOWN_0CC5					0x00000cc5
+
+#define REG_A4XX_UNKNOWN_0CC6					0x00000cc6
+
+#define REG_A4XX_UNKNOWN_0D01					0x00000d01
+
+#define REG_A4XX_UNKNOWN_0E05					0x00000e05
+
+#define REG_A4XX_UNKNOWN_0E42					0x00000e42
+
+#define REG_A4XX_UNKNOWN_0EC2					0x00000ec2
+
+#define REG_A4XX_UNKNOWN_0EC3					0x00000ec3
+
+#define REG_A4XX_UNKNOWN_0F03					0x00000f03
+
+#define REG_A4XX_UNKNOWN_2001					0x00002001
+
+#define REG_A4XX_UNKNOWN_209B					0x0000209b
+
+#define REG_A4XX_UNKNOWN_20EF					0x000020ef
+
+#define REG_A4XX_UNKNOWN_20F0					0x000020f0
+
+#define REG_A4XX_UNKNOWN_20F1					0x000020f1
+
+#define REG_A4XX_UNKNOWN_20F2					0x000020f2
+
+#define REG_A4XX_UNKNOWN_20F3					0x000020f3
+
+#define REG_A4XX_UNKNOWN_20F4					0x000020f4
+
+#define REG_A4XX_UNKNOWN_20F5					0x000020f5
+
+#define REG_A4XX_UNKNOWN_20F6					0x000020f6
+
+#define REG_A4XX_UNKNOWN_20F7					0x000020f7
+
+#define REG_A4XX_UNKNOWN_2152					0x00002152
+
+#define REG_A4XX_UNKNOWN_2153					0x00002153
+
+#define REG_A4XX_UNKNOWN_2154					0x00002154
+
+#define REG_A4XX_UNKNOWN_2155					0x00002155
+
+#define REG_A4XX_UNKNOWN_2156					0x00002156
+
+#define REG_A4XX_UNKNOWN_2157					0x00002157
+
+#define REG_A4XX_UNKNOWN_21C3					0x000021c3
+
+#define REG_A4XX_UNKNOWN_21E6					0x000021e6
+
+#define REG_A4XX_UNKNOWN_2209					0x00002209
+
+#define REG_A4XX_UNKNOWN_22D7					0x000022d7
+
+#define REG_A4XX_UNKNOWN_2381					0x00002381
+
+#define REG_A4XX_UNKNOWN_23A0					0x000023a0
+
+#define REG_A4XX_TEX_SAMP_0					0x00000000
+#define A4XX_TEX_SAMP_0_XY_MAG__MASK				0x00000006
+#define A4XX_TEX_SAMP_0_XY_MAG__SHIFT				1
+static inline uint32_t A4XX_TEX_SAMP_0_XY_MAG(enum a4xx_tex_filter val)
+{
+	return ((val) << A4XX_TEX_SAMP_0_XY_MAG__SHIFT) & A4XX_TEX_SAMP_0_XY_MAG__MASK;
+}
+#define A4XX_TEX_SAMP_0_XY_MIN__MASK				0x00000018
+#define A4XX_TEX_SAMP_0_XY_MIN__SHIFT				3
+static inline uint32_t A4XX_TEX_SAMP_0_XY_MIN(enum a4xx_tex_filter val)
+{
+	return ((val) << A4XX_TEX_SAMP_0_XY_MIN__SHIFT) & A4XX_TEX_SAMP_0_XY_MIN__MASK;
+}
+#define A4XX_TEX_SAMP_0_WRAP_S__MASK				0x000000e0
+#define A4XX_TEX_SAMP_0_WRAP_S__SHIFT				5
+static inline uint32_t A4XX_TEX_SAMP_0_WRAP_S(enum a4xx_tex_clamp val)
+{
+	return ((val) << A4XX_TEX_SAMP_0_WRAP_S__SHIFT) & A4XX_TEX_SAMP_0_WRAP_S__MASK;
+}
+#define A4XX_TEX_SAMP_0_WRAP_T__MASK				0x00000700
+#define A4XX_TEX_SAMP_0_WRAP_T__SHIFT				8
+static inline uint32_t A4XX_TEX_SAMP_0_WRAP_T(enum a4xx_tex_clamp val)
+{
+	return ((val) << A4XX_TEX_SAMP_0_WRAP_T__SHIFT) & A4XX_TEX_SAMP_0_WRAP_T__MASK;
+}
+#define A4XX_TEX_SAMP_0_WRAP_R__MASK				0x00003800
+#define A4XX_TEX_SAMP_0_WRAP_R__SHIFT				11
+static inline uint32_t A4XX_TEX_SAMP_0_WRAP_R(enum a4xx_tex_clamp val)
+{
+	return ((val) << A4XX_TEX_SAMP_0_WRAP_R__SHIFT) & A4XX_TEX_SAMP_0_WRAP_R__MASK;
+}
+
+#define REG_A4XX_TEX_SAMP_1					0x00000001
+#define A4XX_TEX_SAMP_1_COMPARE_FUNC__MASK			0x0000000e
+#define A4XX_TEX_SAMP_1_COMPARE_FUNC__SHIFT			1
+static inline uint32_t A4XX_TEX_SAMP_1_COMPARE_FUNC(enum adreno_compare_func val)
+{
+	return ((val) << A4XX_TEX_SAMP_1_COMPARE_FUNC__SHIFT) & A4XX_TEX_SAMP_1_COMPARE_FUNC__MASK;
+}
+#define A4XX_TEX_SAMP_1_MAX_LOD__MASK				0x000fff00
+#define A4XX_TEX_SAMP_1_MAX_LOD__SHIFT				8
+static inline uint32_t A4XX_TEX_SAMP_1_MAX_LOD(float val)
+{
+	return ((((uint32_t)(val * 64.0))) << A4XX_TEX_SAMP_1_MAX_LOD__SHIFT) & A4XX_TEX_SAMP_1_MAX_LOD__MASK;
+}
+#define A4XX_TEX_SAMP_1_MIN_LOD__MASK				0xfff00000
+#define A4XX_TEX_SAMP_1_MIN_LOD__SHIFT				20
+static inline uint32_t A4XX_TEX_SAMP_1_MIN_LOD(float val)
+{
+	return ((((uint32_t)(val * 64.0))) << A4XX_TEX_SAMP_1_MIN_LOD__SHIFT) & A4XX_TEX_SAMP_1_MIN_LOD__MASK;
+}
+
+#define REG_A4XX_TEX_CONST_0					0x00000000
+#define A4XX_TEX_CONST_0_TILED					0x00000001
+#define A4XX_TEX_CONST_0_SWIZ_X__MASK				0x00000070
+#define A4XX_TEX_CONST_0_SWIZ_X__SHIFT				4
+static inline uint32_t A4XX_TEX_CONST_0_SWIZ_X(enum a4xx_tex_swiz val)
+{
+	return ((val) << A4XX_TEX_CONST_0_SWIZ_X__SHIFT) & A4XX_TEX_CONST_0_SWIZ_X__MASK;
+}
+#define A4XX_TEX_CONST_0_SWIZ_Y__MASK				0x00000380
+#define A4XX_TEX_CONST_0_SWIZ_Y__SHIFT				7
+static inline uint32_t A4XX_TEX_CONST_0_SWIZ_Y(enum a4xx_tex_swiz val)
+{
+	return ((val) << A4XX_TEX_CONST_0_SWIZ_Y__SHIFT) & A4XX_TEX_CONST_0_SWIZ_Y__MASK;
+}
+#define A4XX_TEX_CONST_0_SWIZ_Z__MASK				0x00001c00
+#define A4XX_TEX_CONST_0_SWIZ_Z__SHIFT				10
+static inline uint32_t A4XX_TEX_CONST_0_SWIZ_Z(enum a4xx_tex_swiz val)
+{
+	return ((val) << A4XX_TEX_CONST_0_SWIZ_Z__SHIFT) & A4XX_TEX_CONST_0_SWIZ_Z__MASK;
+}
+#define A4XX_TEX_CONST_0_SWIZ_W__MASK				0x0000e000
+#define A4XX_TEX_CONST_0_SWIZ_W__SHIFT				13
+static inline uint32_t A4XX_TEX_CONST_0_SWIZ_W(enum a4xx_tex_swiz val)
+{
+	return ((val) << A4XX_TEX_CONST_0_SWIZ_W__SHIFT) & A4XX_TEX_CONST_0_SWIZ_W__MASK;
+}
+#define A4XX_TEX_CONST_0_FMT__MASK				0x1fc00000
+#define A4XX_TEX_CONST_0_FMT__SHIFT				22
+static inline uint32_t A4XX_TEX_CONST_0_FMT(enum a4xx_tex_fmt val)
+{
+	return ((val) << A4XX_TEX_CONST_0_FMT__SHIFT) & A4XX_TEX_CONST_0_FMT__MASK;
+}
+#define A4XX_TEX_CONST_0_TYPE__MASK				0x60000000
+#define A4XX_TEX_CONST_0_TYPE__SHIFT				29
+static inline uint32_t A4XX_TEX_CONST_0_TYPE(enum a4xx_tex_type val)
+{
+	return ((val) << A4XX_TEX_CONST_0_TYPE__SHIFT) & A4XX_TEX_CONST_0_TYPE__MASK;
+}
+
+#define REG_A4XX_TEX_CONST_1					0x00000001
+#define A4XX_TEX_CONST_1_HEIGHT__MASK				0x00007fff
+#define A4XX_TEX_CONST_1_HEIGHT__SHIFT				0
+static inline uint32_t A4XX_TEX_CONST_1_HEIGHT(uint32_t val)
+{
+	return ((val) << A4XX_TEX_CONST_1_HEIGHT__SHIFT) & A4XX_TEX_CONST_1_HEIGHT__MASK;
+}
+#define A4XX_TEX_CONST_1_WIDTH__MASK				0x1fff8000
+#define A4XX_TEX_CONST_1_WIDTH__SHIFT				15
+static inline uint32_t A4XX_TEX_CONST_1_WIDTH(uint32_t val)
+{
+	return ((val) << A4XX_TEX_CONST_1_WIDTH__SHIFT) & A4XX_TEX_CONST_1_WIDTH__MASK;
+}
+
+#define REG_A4XX_TEX_CONST_2					0x00000002
+#define A4XX_TEX_CONST_2_PITCH__MASK				0x3ffffe00
+#define A4XX_TEX_CONST_2_PITCH__SHIFT				9
+static inline uint32_t A4XX_TEX_CONST_2_PITCH(uint32_t val)
+{
+	return ((val) << A4XX_TEX_CONST_2_PITCH__SHIFT) & A4XX_TEX_CONST_2_PITCH__MASK;
+}
+#define A4XX_TEX_CONST_2_SWAP__MASK				0xc0000000
+#define A4XX_TEX_CONST_2_SWAP__SHIFT				30
+static inline uint32_t A4XX_TEX_CONST_2_SWAP(enum a3xx_color_swap val)
+{
+	return ((val) << A4XX_TEX_CONST_2_SWAP__SHIFT) & A4XX_TEX_CONST_2_SWAP__MASK;
+}
+
+#define REG_A4XX_TEX_CONST_3					0x00000003
+#define A4XX_TEX_CONST_3_LAYERSZ__MASK				0x0000000f
+#define A4XX_TEX_CONST_3_LAYERSZ__SHIFT				0
+static inline uint32_t A4XX_TEX_CONST_3_LAYERSZ(uint32_t val)
+{
+	return ((val >> 12) << A4XX_TEX_CONST_3_LAYERSZ__SHIFT) & A4XX_TEX_CONST_3_LAYERSZ__MASK;
+}
+
+#define REG_A4XX_TEX_CONST_4					0x00000004
+#define A4XX_TEX_CONST_4_BASE__MASK				0xffffffff
+#define A4XX_TEX_CONST_4_BASE__SHIFT				0
+static inline uint32_t A4XX_TEX_CONST_4_BASE(uint32_t val)
+{
+	return ((val) << A4XX_TEX_CONST_4_BASE__SHIFT) & A4XX_TEX_CONST_4_BASE__MASK;
+}
+
+#define REG_A4XX_TEX_CONST_5					0x00000005
+
+#define REG_A4XX_TEX_CONST_6					0x00000006
+
+#define REG_A4XX_TEX_CONST_7					0x00000007
+
+
+#endif /* A4XX_XML */
diff --git a/drivers/gpu/drm/msm/adreno/a4xx_gpu.c b/drivers/gpu/drm/msm/adreno/a4xx_gpu.c
new file mode 100644
index 000000000000..91221836c5ad
--- /dev/null
+++ b/drivers/gpu/drm/msm/adreno/a4xx_gpu.c
@@ -0,0 +1,604 @@
+/* Copyright (c) 2014 The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ */
+#include "a4xx_gpu.h"
+#ifdef CONFIG_MSM_OCMEM
+#  include <soc/qcom/ocmem.h>
+#endif
+
+#define A4XX_INT0_MASK \
+	(A4XX_INT0_RBBM_AHB_ERROR |        \
+	 A4XX_INT0_RBBM_ATB_BUS_OVERFLOW | \
+	 A4XX_INT0_CP_T0_PACKET_IN_IB |    \
+	 A4XX_INT0_CP_OPCODE_ERROR |       \
+	 A4XX_INT0_CP_RESERVED_BIT_ERROR | \
+	 A4XX_INT0_CP_HW_FAULT |           \
+	 A4XX_INT0_CP_IB1_INT |            \
+	 A4XX_INT0_CP_IB2_INT |            \
+	 A4XX_INT0_CP_RB_INT |             \
+	 A4XX_INT0_CP_REG_PROTECT_FAULT |  \
+	 A4XX_INT0_CP_AHB_ERROR_HALT |     \
+	 A4XX_INT0_UCHE_OOB_ACCESS)
+
+extern bool hang_debug;
+static void a4xx_dump(struct msm_gpu *gpu);
+
+/*
+ * a4xx_enable_hwcg() - Program the clock control registers
+ * @device: The adreno device pointer
+ */
+static void a4xx_enable_hwcg(struct msm_gpu *gpu)
+{
+	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
+	unsigned int i;
+	for (i = 0; i < 4; i++)
+		gpu_write(gpu, REG_A4XX_RBBM_CLOCK_CTL_TP(i), 0x02222202);
+	for (i = 0; i < 4; i++)
+		gpu_write(gpu, REG_A4XX_RBBM_CLOCK_CTL2_TP(i), 0x00002222);
+	for (i = 0; i < 4; i++)
+		gpu_write(gpu, REG_A4XX_RBBM_CLOCK_HYST_TP(i), 0x0E739CE7);
+	for (i = 0; i < 4; i++)
+		gpu_write(gpu, REG_A4XX_RBBM_CLOCK_DELAY_TP(i), 0x00111111);
+	for (i = 0; i < 4; i++)
+		gpu_write(gpu, REG_A4XX_RBBM_CLOCK_CTL_SP(i), 0x22222222);
+	for (i = 0; i < 4; i++)
+		gpu_write(gpu, REG_A4XX_RBBM_CLOCK_CTL2_SP(i), 0x00222222);
+	for (i = 0; i < 4; i++)
+		gpu_write(gpu, REG_A4XX_RBBM_CLOCK_HYST_SP(i), 0x00000104);
+	for (i = 0; i < 4; i++)
+		gpu_write(gpu, REG_A4XX_RBBM_CLOCK_DELAY_SP(i), 0x00000081);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_CTL_UCHE, 0x22222222);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_CTL2_UCHE, 0x02222222);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_CTL3_UCHE, 0x00000000);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_CTL4_UCHE, 0x00000000);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_HYST_UCHE, 0x00004444);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_DELAY_UCHE, 0x00001112);
+	for (i = 0; i < 4; i++)
+		gpu_write(gpu, REG_A4XX_RBBM_CLOCK_CTL_RB(i), 0x22222222);
+
+	/* Disable L1 clocking in A420 due to CCU issues with it */
+	for (i = 0; i < 4; i++) {
+		if (adreno_is_a420(adreno_gpu)) {
+			gpu_write(gpu, REG_A4XX_RBBM_CLOCK_CTL2_RB(i),
+					0x00002020);
+		} else {
+			gpu_write(gpu, REG_A4XX_RBBM_CLOCK_CTL2_RB(i),
+					0x00022020);
+		}
+	}
+
+	for (i = 0; i < 4; i++) {
+		gpu_write(gpu, REG_A4XX_RBBM_CLOCK_CTL_MARB_CCU(i),
+				0x00000922);
+	}
+
+	for (i = 0; i < 4; i++) {
+		gpu_write(gpu, REG_A4XX_RBBM_CLOCK_HYST_RB_MARB_CCU(i),
+				0x00000000);
+	}
+
+	for (i = 0; i < 4; i++) {
+		gpu_write(gpu, REG_A4XX_RBBM_CLOCK_DELAY_RB_MARB_CCU_L1(i),
+				0x00000001);
+	}
+
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_MODE_GPC, 0x02222222);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_HYST_GPC, 0x04100104);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_DELAY_GPC, 0x00022222);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_CTL_COM_DCOM, 0x00000022);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_HYST_COM_DCOM, 0x0000010F);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_DELAY_COM_DCOM, 0x00000022);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_CTL_TSE_RAS_RBBM, 0x00222222);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_HYST_TSE_RAS_RBBM, 0x00004104);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_DELAY_TSE_RAS_RBBM, 0x00000222);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_CTL_HLSQ , 0x00000000);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_HYST_HLSQ, 0x00000000);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_DELAY_HLSQ, 0x00020000);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_CTL, 0xAAAAAAAA);
+	gpu_write(gpu, REG_A4XX_RBBM_CLOCK_CTL2, 0);
+}
+
+static void a4xx_me_init(struct msm_gpu *gpu)
+{
+	struct msm_ringbuffer *ring = gpu->rb;
+
+	OUT_PKT3(ring, CP_ME_INIT, 17);
+	OUT_RING(ring, 0x000003f7);
+	OUT_RING(ring, 0x00000000);
+	OUT_RING(ring, 0x00000000);
+	OUT_RING(ring, 0x00000000);
+	OUT_RING(ring, 0x00000080);
+	OUT_RING(ring, 0x00000100);
+	OUT_RING(ring, 0x00000180);
+	OUT_RING(ring, 0x00006600);
+	OUT_RING(ring, 0x00000150);
+	OUT_RING(ring, 0x0000014e);
+	OUT_RING(ring, 0x00000154);
+	OUT_RING(ring, 0x00000001);
+	OUT_RING(ring, 0x00000000);
+	OUT_RING(ring, 0x00000000);
+	OUT_RING(ring, 0x00000000);
+	OUT_RING(ring, 0x00000000);
+	OUT_RING(ring, 0x00000000);
+
+	gpu->funcs->flush(gpu);
+	gpu->funcs->idle(gpu);
+}
+
+static int a4xx_hw_init(struct msm_gpu *gpu)
+{
+	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
+	struct a4xx_gpu *a4xx_gpu = to_a4xx_gpu(adreno_gpu);
+	uint32_t *ptr, len;
+	int i, ret;
+
+	if (adreno_is_a4xx(adreno_gpu)) {
+		gpu_write(gpu, REG_A4XX_VBIF_ABIT_SORT, 0x0001001F);
+		gpu_write(gpu, REG_A4XX_VBIF_ABIT_SORT_CONF, 0x000000A4);
+		gpu_write(gpu, REG_A4XX_VBIF_GATE_OFF_WRREQ_EN, 0x00000001);
+		gpu_write(gpu, REG_A4XX_VBIF_IN_RD_LIM_CONF0, 0x18181818);
+		gpu_write(gpu, REG_A4XX_VBIF_IN_RD_LIM_CONF1, 0x00000018);
+		gpu_write(gpu, REG_A4XX_VBIF_IN_WR_LIM_CONF0, 0x18181818);
+		gpu_write(gpu, REG_A4XX_VBIF_IN_WR_LIM_CONF1, 0x00000018);
+		gpu_write(gpu, REG_A4XX_VBIF_ROUND_ROBIN_QOS_ARB, 0x00000003);
+	} else {
+		BUG();
+	}
+
+	/* Make all blocks contribute to the GPU BUSY perf counter */
+	gpu_write(gpu, REG_A4XX_RBBM_GPU_BUSY_MASKED, 0xffffffff);
+
+	/* Tune the hystersis counters for SP and CP idle detection */
+	gpu_write(gpu, REG_A4XX_RBBM_SP_HYST_CNT, 0x10);
+	gpu_write(gpu, REG_A4XX_RBBM_WAIT_IDLE_CLOCKS_CTL, 0x10);
+
+	 /* Enable the RBBM error reporting bits */
+	gpu_write(gpu, REG_A4XX_RBBM_AHB_CTL0, 0x00000001);
+
+	/* Enable AHB error reporting*/
+	gpu_write(gpu, REG_A4XX_RBBM_AHB_CTL1, 0xa6ffffff);
+
+	/* Enable power counters*/
+	gpu_write(gpu, REG_A4XX_RBBM_RBBM_CTL, 0x00000030);
+
+	/*
+	 * Turn on hang detection - this spews a lot of useful information
+	 * into the RBBM registers on a hang:
+	 */
+	gpu_write(gpu, REG_A4XX_RBBM_INTERFACE_HANG_INT_CTL,
+			(1 << 30) | 0xFFFF);
+
+	gpu_write(gpu, REG_A4XX_RB_GMEM_BASE_ADDR,
+			(unsigned int)(a4xx_gpu->ocmem_base >> 14));
+
+	/* Turn on performance counters: */
+	gpu_write(gpu, REG_A4XX_RBBM_PERFCTR_CTL, 0x01);
+
+	/* Disable L2 bypass to avoid UCHE out of bounds errors */
+	gpu_write(gpu, REG_A4XX_UCHE_TRAP_BASE_LO, 0xffff0000);
+	gpu_write(gpu, REG_A4XX_UCHE_TRAP_BASE_HI, 0xffff0000);
+
+	gpu_write(gpu, REG_A4XX_CP_DEBUG, (1 << 25) |
+			(adreno_is_a420(adreno_gpu) ? (1 << 29) : 0));
+
+	a4xx_enable_hwcg(gpu);
+
+	/*
+	 * For A420 set RBBM_CLOCK_DELAY_HLSQ.CGC_HLSQ_TP_EARLY_CYC >= 2
+	 * due to timing issue with HLSQ_TP_CLK_EN
+	 */
+	if (adreno_is_a420(adreno_gpu)) {
+		unsigned int val;
+		val = gpu_read(gpu, REG_A4XX_RBBM_CLOCK_DELAY_HLSQ);
+		val &= ~A4XX_CGC_HLSQ_EARLY_CYC__MASK;
+		val |= 2 << A4XX_CGC_HLSQ_EARLY_CYC__SHIFT;
+		gpu_write(gpu, REG_A4XX_RBBM_CLOCK_DELAY_HLSQ, val);
+	}
+
+	ret = adreno_hw_init(gpu);
+	if (ret)
+		return ret;
+
+	/* setup access protection: */
+	gpu_write(gpu, REG_A4XX_CP_PROTECT_CTRL, 0x00000007);
+
+	/* RBBM registers */
+	gpu_write(gpu, REG_A4XX_CP_PROTECT(0), 0x62000010);
+	gpu_write(gpu, REG_A4XX_CP_PROTECT(1), 0x63000020);
+	gpu_write(gpu, REG_A4XX_CP_PROTECT(2), 0x64000040);
+	gpu_write(gpu, REG_A4XX_CP_PROTECT(3), 0x65000080);
+	gpu_write(gpu, REG_A4XX_CP_PROTECT(4), 0x66000100);
+	gpu_write(gpu, REG_A4XX_CP_PROTECT(5), 0x64000200);
+
+	/* CP registers */
+	gpu_write(gpu, REG_A4XX_CP_PROTECT(6), 0x67000800);
+	gpu_write(gpu, REG_A4XX_CP_PROTECT(7), 0x64001600);
+
+
+	/* RB registers */
+	gpu_write(gpu, REG_A4XX_CP_PROTECT(8), 0x60003300);
+
+	/* HLSQ registers */
+	gpu_write(gpu, REG_A4XX_CP_PROTECT(9), 0x60003800);
+
+	/* VPC registers */
+	gpu_write(gpu, REG_A4XX_CP_PROTECT(10), 0x61003980);
+
+	/* SMMU registers */
+	gpu_write(gpu, REG_A4XX_CP_PROTECT(11), 0x6e010000);
+
+	gpu_write(gpu, REG_A4XX_RBBM_INT_0_MASK, A4XX_INT0_MASK);
+
+	ret = adreno_hw_init(gpu);
+	if (ret)
+		return ret;
+
+	/* Load PM4: */
+	ptr = (uint32_t *)(adreno_gpu->pm4->data);
+	len = adreno_gpu->pm4->size / 4;
+	DBG("loading PM4 ucode version: %u", ptr[0]);
+	gpu_write(gpu, REG_A4XX_CP_ME_RAM_WADDR, 0);
+	for (i = 1; i < len; i++)
+		gpu_write(gpu, REG_A4XX_CP_ME_RAM_DATA, ptr[i]);
+
+	/* Load PFP: */
+	ptr = (uint32_t *)(adreno_gpu->pfp->data);
+	len = adreno_gpu->pfp->size / 4;
+	DBG("loading PFP ucode version: %u", ptr[0]);
+
+	gpu_write(gpu, REG_A4XX_CP_PFP_UCODE_ADDR, 0);
+	for (i = 1; i < len; i++)
+		gpu_write(gpu, REG_A4XX_CP_PFP_UCODE_DATA, ptr[i]);
+
+	/* clear ME_HALT to start micro engine */
+	gpu_write(gpu, REG_A4XX_CP_ME_CNTL, 0);
+
+	a4xx_me_init(gpu);
+	return 0;
+}
+
+static void a4xx_recover(struct msm_gpu *gpu)
+{
+	/* dump registers before resetting gpu, if enabled: */
+	if (hang_debug)
+		a4xx_dump(gpu);
+
+	gpu_write(gpu, REG_A4XX_RBBM_SW_RESET_CMD, 1);
+	gpu_read(gpu, REG_A4XX_RBBM_SW_RESET_CMD);
+	gpu_write(gpu, REG_A4XX_RBBM_SW_RESET_CMD, 0);
+	adreno_recover(gpu);
+}
+
+static void a4xx_destroy(struct msm_gpu *gpu)
+{
+	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
+	struct a4xx_gpu *a4xx_gpu = to_a4xx_gpu(adreno_gpu);
+
+	DBG("%s", gpu->name);
+
+	adreno_gpu_cleanup(adreno_gpu);
+
+#ifdef CONFIG_MSM_OCMEM
+	if (a4xx_gpu->ocmem_base)
+		ocmem_free(OCMEM_GRAPHICS, a4xx_gpu->ocmem_hdl);
+#endif
+
+	kfree(a4xx_gpu);
+}
+
+static void a4xx_idle(struct msm_gpu *gpu)
+{
+	/* wait for ringbuffer to drain: */
+	adreno_idle(gpu);
+
+	/* then wait for GPU to finish: */
+	if (spin_until(!(gpu_read(gpu, REG_A4XX_RBBM_STATUS) &
+					A4XX_RBBM_STATUS_GPU_BUSY)))
+		DRM_ERROR("%s: timeout waiting for GPU to idle!\n", gpu->name);
+
+	/* TODO maybe we need to reset GPU here to recover from hang? */
+}
+
+static irqreturn_t a4xx_irq(struct msm_gpu *gpu)
+{
+	uint32_t status;
+
+	status = gpu_read(gpu, REG_A4XX_RBBM_INT_0_STATUS);
+	DBG("%s: Int status %08x", gpu->name, status);
+
+	gpu_write(gpu, REG_A4XX_RBBM_INT_CLEAR_CMD, status);
+
+	msm_gpu_retire(gpu);
+
+	return IRQ_HANDLED;
+}
+
+static const unsigned int a4xx_registers[] = {
+	/* RBBM */
+	0x0000, 0x0002, 0x0004, 0x0021, 0x0023, 0x0024, 0x0026, 0x0026,
+	0x0028, 0x002B, 0x002E, 0x0034, 0x0037, 0x0044, 0x0047, 0x0066,
+	0x0068, 0x0095, 0x009C, 0x0170, 0x0174, 0x01AF,
+	/* CP */
+	0x0200, 0x0233, 0x0240, 0x0250, 0x04C0, 0x04DD, 0x0500, 0x050B,
+	0x0578, 0x058F,
+	/* VSC */
+	0x0C00, 0x0C03, 0x0C08, 0x0C41, 0x0C50, 0x0C51,
+	/* GRAS */
+	0x0C80, 0x0C81, 0x0C88, 0x0C8F,
+	/* RB */
+	0x0CC0, 0x0CC0, 0x0CC4, 0x0CD2,
+	/* PC */
+	0x0D00, 0x0D0C, 0x0D10, 0x0D17, 0x0D20, 0x0D23,
+	/* VFD */
+	0x0E40, 0x0E4A,
+	/* VPC */
+	0x0E60, 0x0E61, 0x0E63, 0x0E68,
+	/* UCHE */
+	0x0E80, 0x0E84, 0x0E88, 0x0E95,
+	/* VMIDMT */
+	0x1000, 0x1000, 0x1002, 0x1002, 0x1004, 0x1004, 0x1008, 0x100A,
+	0x100C, 0x100D, 0x100F, 0x1010, 0x1012, 0x1016, 0x1024, 0x1024,
+	0x1027, 0x1027, 0x1100, 0x1100, 0x1102, 0x1102, 0x1104, 0x1104,
+	0x1110, 0x1110, 0x1112, 0x1116, 0x1124, 0x1124, 0x1300, 0x1300,
+	0x1380, 0x1380,
+	/* GRAS CTX 0 */
+	0x2000, 0x2004, 0x2008, 0x2067, 0x2070, 0x2078, 0x207B, 0x216E,
+	/* PC CTX 0 */
+	0x21C0, 0x21C6, 0x21D0, 0x21D0, 0x21D9, 0x21D9, 0x21E5, 0x21E7,
+	/* VFD CTX 0 */
+	0x2200, 0x2204, 0x2208, 0x22A9,
+	/* GRAS CTX 1 */
+	0x2400, 0x2404, 0x2408, 0x2467, 0x2470, 0x2478, 0x247B, 0x256E,
+	/* PC CTX 1 */
+	0x25C0, 0x25C6, 0x25D0, 0x25D0, 0x25D9, 0x25D9, 0x25E5, 0x25E7,
+	/* VFD CTX 1 */
+	0x2600, 0x2604, 0x2608, 0x26A9,
+	/* XPU */
+	0x2C00, 0x2C01, 0x2C10, 0x2C10, 0x2C12, 0x2C16, 0x2C1D, 0x2C20,
+	0x2C28, 0x2C28, 0x2C30, 0x2C30, 0x2C32, 0x2C36, 0x2C40, 0x2C40,
+	0x2C50, 0x2C50, 0x2C52, 0x2C56, 0x2C80, 0x2C80, 0x2C94, 0x2C95,
+	/* VBIF */
+	0x3000, 0x3007, 0x300C, 0x3014, 0x3018, 0x301D, 0x3020, 0x3022,
+	0x3024, 0x3026, 0x3028, 0x302A, 0x302C, 0x302D, 0x3030, 0x3031,
+	0x3034, 0x3036, 0x3038, 0x3038, 0x303C, 0x303D, 0x3040, 0x3040,
+	0x3049, 0x3049, 0x3058, 0x3058, 0x305B, 0x3061, 0x3064, 0x3068,
+	0x306C, 0x306D, 0x3080, 0x3088, 0x308B, 0x308C, 0x3090, 0x3094,
+	0x3098, 0x3098, 0x309C, 0x309C, 0x30C0, 0x30C0, 0x30C8, 0x30C8,
+	0x30D0, 0x30D0, 0x30D8, 0x30D8, 0x30E0, 0x30E0, 0x3100, 0x3100,
+	0x3108, 0x3108, 0x3110, 0x3110, 0x3118, 0x3118, 0x3120, 0x3120,
+	0x3124, 0x3125, 0x3129, 0x3129, 0x3131, 0x3131, 0x330C, 0x330C,
+	0x3310, 0x3310, 0x3400, 0x3401, 0x3410, 0x3410, 0x3412, 0x3416,
+	0x341D, 0x3420, 0x3428, 0x3428, 0x3430, 0x3430, 0x3432, 0x3436,
+	0x3440, 0x3440, 0x3450, 0x3450, 0x3452, 0x3456, 0x3480, 0x3480,
+	0x3494, 0x3495, 0x4000, 0x4000, 0x4002, 0x4002, 0x4004, 0x4004,
+	0x4008, 0x400A, 0x400C, 0x400D, 0x400F, 0x4012, 0x4014, 0x4016,
+	0x401D, 0x401D, 0x4020, 0x4027, 0x4060, 0x4062, 0x4200, 0x4200,
+	0x4300, 0x4300, 0x4400, 0x4400, 0x4500, 0x4500, 0x4800, 0x4802,
+	0x480F, 0x480F, 0x4811, 0x4811, 0x4813, 0x4813, 0x4815, 0x4816,
+	0x482B, 0x482B, 0x4857, 0x4857, 0x4883, 0x4883, 0x48AF, 0x48AF,
+	0x48C5, 0x48C5, 0x48E5, 0x48E5, 0x4905, 0x4905, 0x4925, 0x4925,
+	0x4945, 0x4945, 0x4950, 0x4950, 0x495B, 0x495B, 0x4980, 0x498E,
+	0x4B00, 0x4B00, 0x4C00, 0x4C00, 0x4D00, 0x4D00, 0x4E00, 0x4E00,
+	0x4E80, 0x4E80, 0x4F00, 0x4F00, 0x4F08, 0x4F08, 0x4F10, 0x4F10,
+	0x4F18, 0x4F18, 0x4F20, 0x4F20, 0x4F30, 0x4F30, 0x4F60, 0x4F60,
+	0x4F80, 0x4F81, 0x4F88, 0x4F89, 0x4FEE, 0x4FEE, 0x4FF3, 0x4FF3,
+	0x6000, 0x6001, 0x6008, 0x600F, 0x6014, 0x6016, 0x6018, 0x601B,
+	0x61FD, 0x61FD, 0x623C, 0x623C, 0x6380, 0x6380, 0x63A0, 0x63A0,
+	0x63C0, 0x63C1, 0x63C8, 0x63C9, 0x63D0, 0x63D4, 0x63D6, 0x63D6,
+	0x63EE, 0x63EE, 0x6400, 0x6401, 0x6408, 0x640F, 0x6414, 0x6416,
+	0x6418, 0x641B, 0x65FD, 0x65FD, 0x663C, 0x663C, 0x6780, 0x6780,
+	0x67A0, 0x67A0, 0x67C0, 0x67C1, 0x67C8, 0x67C9, 0x67D0, 0x67D4,
+	0x67D6, 0x67D6, 0x67EE, 0x67EE, 0x6800, 0x6801, 0x6808, 0x680F,
+	0x6814, 0x6816, 0x6818, 0x681B, 0x69FD, 0x69FD, 0x6A3C, 0x6A3C,
+	0x6B80, 0x6B80, 0x6BA0, 0x6BA0, 0x6BC0, 0x6BC1, 0x6BC8, 0x6BC9,
+	0x6BD0, 0x6BD4, 0x6BD6, 0x6BD6, 0x6BEE, 0x6BEE,
+	~0 /* sentinel */
+};
+
+#ifdef CONFIG_DEBUG_FS
+static void a4xx_show(struct msm_gpu *gpu, struct seq_file *m)
+{
+	gpu->funcs->pm_resume(gpu);
+
+	seq_printf(m, "status:   %08x\n",
+			gpu_read(gpu, REG_A4XX_RBBM_STATUS));
+	gpu->funcs->pm_suspend(gpu);
+
+	adreno_show(gpu, m);
+
+}
+#endif
+
+/* Register offset defines for A4XX, in order of enum adreno_regs */
+static const unsigned int a4xx_register_offsets[REG_ADRENO_REGISTER_MAX] = {
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_DEBUG, REG_A4XX_CP_DEBUG),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_ME_RAM_WADDR, REG_A4XX_CP_ME_RAM_WADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_ME_RAM_DATA, REG_A4XX_CP_ME_RAM_DATA),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_PFP_UCODE_DATA,
+			REG_A4XX_CP_PFP_UCODE_DATA),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_PFP_UCODE_ADDR,
+			REG_A4XX_CP_PFP_UCODE_ADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_WFI_PEND_CTR, REG_A4XX_CP_WFI_PEND_CTR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_BASE, REG_A4XX_CP_RB_BASE),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_RPTR_ADDR, REG_A4XX_CP_RB_RPTR_ADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_RPTR, REG_A4XX_CP_RB_RPTR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_WPTR, REG_A4XX_CP_RB_WPTR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_PROTECT_CTRL, REG_A4XX_CP_PROTECT_CTRL),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_ME_CNTL, REG_A4XX_CP_ME_CNTL),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_CNTL, REG_A4XX_CP_RB_CNTL),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_IB1_BASE, REG_A4XX_CP_IB1_BASE),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_IB1_BUFSZ, REG_A4XX_CP_IB1_BUFSZ),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_IB2_BASE, REG_A4XX_CP_IB2_BASE),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_IB2_BUFSZ, REG_A4XX_CP_IB2_BUFSZ),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_TIMESTAMP, REG_AXXX_CP_SCRATCH_REG0),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_ME_RAM_RADDR, REG_A4XX_CP_ME_RAM_RADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_ROQ_ADDR, REG_A4XX_CP_ROQ_ADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_ROQ_DATA, REG_A4XX_CP_ROQ_DATA),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_MERCIU_ADDR, REG_A4XX_CP_MERCIU_ADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_MERCIU_DATA, REG_A4XX_CP_MERCIU_DATA),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_MERCIU_DATA2, REG_A4XX_CP_MERCIU_DATA2),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_MEQ_ADDR, REG_A4XX_CP_MEQ_ADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_MEQ_DATA, REG_A4XX_CP_MEQ_DATA),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_HW_FAULT, REG_A4XX_CP_HW_FAULT),
+	REG_ADRENO_DEFINE(REG_ADRENO_CP_PROTECT_STATUS,
+			REG_A4XX_CP_PROTECT_STATUS),
+	REG_ADRENO_DEFINE(REG_ADRENO_SCRATCH_ADDR, REG_A4XX_CP_SCRATCH_ADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_SCRATCH_UMSK, REG_A4XX_CP_SCRATCH_UMASK),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_STATUS, REG_A4XX_RBBM_STATUS),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_CTL,
+			REG_A4XX_RBBM_PERFCTR_CTL),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_CMD0,
+			REG_A4XX_RBBM_PERFCTR_LOAD_CMD0),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_CMD1,
+			REG_A4XX_RBBM_PERFCTR_LOAD_CMD1),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_CMD2,
+			REG_A4XX_RBBM_PERFCTR_LOAD_CMD2),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_PWR_1_LO,
+			REG_A4XX_RBBM_PERFCTR_PWR_1_LO),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_INT_0_MASK, REG_A4XX_RBBM_INT_0_MASK),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_INT_0_STATUS,
+			REG_A4XX_RBBM_INT_0_STATUS),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_AHB_ERROR_STATUS,
+			REG_A4XX_RBBM_AHB_ERROR_STATUS),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_AHB_CMD, REG_A4XX_RBBM_AHB_CMD),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_CLOCK_CTL, REG_A4XX_RBBM_CLOCK_CTL),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_AHB_ME_SPLIT_STATUS,
+			REG_A4XX_RBBM_AHB_ME_SPLIT_STATUS),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_AHB_PFP_SPLIT_STATUS,
+			REG_A4XX_RBBM_AHB_PFP_SPLIT_STATUS),
+	REG_ADRENO_DEFINE(REG_ADRENO_VPC_DEBUG_RAM_SEL,
+			REG_A4XX_VPC_DEBUG_RAM_SEL),
+	REG_ADRENO_DEFINE(REG_ADRENO_VPC_DEBUG_RAM_READ,
+			REG_A4XX_VPC_DEBUG_RAM_READ),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_INT_CLEAR_CMD,
+			REG_A4XX_RBBM_INT_CLEAR_CMD),
+	REG_ADRENO_DEFINE(REG_ADRENO_VSC_SIZE_ADDRESS,
+			REG_A4XX_VSC_SIZE_ADDRESS),
+	REG_ADRENO_DEFINE(REG_ADRENO_VFD_CONTROL_0, REG_A4XX_VFD_CONTROL_0),
+	REG_ADRENO_DEFINE(REG_ADRENO_SP_VS_PVT_MEM_ADDR_REG,
+			REG_A4XX_SP_VS_PVT_MEM_ADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_SP_FS_PVT_MEM_ADDR_REG,
+			REG_A4XX_SP_FS_PVT_MEM_ADDR),
+	REG_ADRENO_DEFINE(REG_ADRENO_SP_VS_OBJ_START_REG,
+			REG_A4XX_SP_VS_OBJ_START),
+	REG_ADRENO_DEFINE(REG_ADRENO_SP_FS_OBJ_START_REG,
+			REG_A4XX_SP_FS_OBJ_START),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_RBBM_CTL, REG_A4XX_RBBM_RBBM_CTL),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_SW_RESET_CMD,
+			REG_A4XX_RBBM_SW_RESET_CMD),
+	REG_ADRENO_DEFINE(REG_ADRENO_UCHE_INVALIDATE0,
+			REG_A4XX_UCHE_INVALIDATE0),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_VALUE_LO,
+			REG_A4XX_RBBM_PERFCTR_LOAD_VALUE_LO),
+	REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_VALUE_HI,
+			REG_A4XX_RBBM_PERFCTR_LOAD_VALUE_HI),
+};
+
+static void a4xx_dump(struct msm_gpu *gpu)
+{
+	adreno_dump(gpu);
+	printk("status:   %08x\n",
+			gpu_read(gpu, REG_A4XX_RBBM_STATUS));
+	adreno_dump(gpu);
+}
+
+static const struct adreno_gpu_funcs funcs = {
+	.base = {
+		.get_param = adreno_get_param,
+		.hw_init = a4xx_hw_init,
+		.pm_suspend = msm_gpu_pm_suspend,
+		.pm_resume = msm_gpu_pm_resume,
+		.recover = a4xx_recover,
+		.last_fence = adreno_last_fence,
+		.submit = adreno_submit,
+		.flush = adreno_flush,
+		.idle = a4xx_idle,
+		.irq = a4xx_irq,
+		.destroy = a4xx_destroy,
+#ifdef CONFIG_DEBUG_FS
+		.show = a4xx_show,
+#endif
+	},
+};
+
+struct msm_gpu *a4xx_gpu_init(struct drm_device *dev)
+{
+	struct a4xx_gpu *a4xx_gpu = NULL;
+	struct adreno_gpu *adreno_gpu;
+	struct msm_gpu *gpu;
+	struct msm_drm_private *priv = dev->dev_private;
+	struct platform_device *pdev = priv->gpu_pdev;
+	int ret;
+
+	if (!pdev) {
+		dev_err(dev->dev, "no a4xx device\n");
+		ret = -ENXIO;
+		goto fail;
+	}
+
+	a4xx_gpu = kzalloc(sizeof(*a4xx_gpu), GFP_KERNEL);
+	if (!a4xx_gpu) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	adreno_gpu = &a4xx_gpu->base;
+	gpu = &adreno_gpu->base;
+
+	a4xx_gpu->pdev = pdev;
+
+	gpu->perfcntrs = NULL;
+	gpu->num_perfcntrs = 0;
+
+	adreno_gpu->registers = a4xx_registers;
+	adreno_gpu->reg_offsets = a4xx_register_offsets;
+
+	ret = adreno_gpu_init(dev, pdev, adreno_gpu, &funcs);
+	if (ret)
+		goto fail;
+
+	/* if needed, allocate gmem: */
+	if (adreno_is_a4xx(adreno_gpu)) {
+#ifdef CONFIG_MSM_OCMEM
+		/* TODO this is different/missing upstream: */
+		struct ocmem_buf *ocmem_hdl =
+				ocmem_allocate(OCMEM_GRAPHICS, adreno_gpu->gmem);
+
+		a4xx_gpu->ocmem_hdl = ocmem_hdl;
+		a4xx_gpu->ocmem_base = ocmem_hdl->addr;
+		adreno_gpu->gmem = ocmem_hdl->len;
+		DBG("using %dK of OCMEM at 0x%08x", adreno_gpu->gmem / 1024,
+				a4xx_gpu->ocmem_base);
+#endif
+	}
+
+	if (!gpu->mmu) {
+		/* TODO we think it is possible to configure the GPU to
+		 * restrict access to VRAM carveout.  But the required
+		 * registers are unknown.  For now just bail out and
+		 * limp along with just modesetting.  If it turns out
+		 * to not be possible to restrict access, then we must
+		 * implement a cmdstream validator.
+		 */
+		dev_err(dev->dev, "No memory protection without IOMMU\n");
+		ret = -ENXIO;
+		goto fail;
+	}
+
+	return gpu;
+
+fail:
+	if (a4xx_gpu)
+		a4xx_destroy(&a4xx_gpu->base.base);
+
+	return ERR_PTR(ret);
+}
diff --git a/drivers/gpu/drm/msm/adreno/a4xx_gpu.h b/drivers/gpu/drm/msm/adreno/a4xx_gpu.h
new file mode 100644
index 000000000000..01247204ac92
--- /dev/null
+++ b/drivers/gpu/drm/msm/adreno/a4xx_gpu.h
@@ -0,0 +1,34 @@
+/* Copyright (c) 2014 The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ */
+#ifndef __A4XX_GPU_H__
+#define __A4XX_GPU_H__
+
+#include "adreno_gpu.h"
+
+/* arrg, somehow fb.h is getting pulled in: */
+#undef ROP_COPY
+#undef ROP_XOR
+
+#include "a4xx.xml.h"
+
+struct a4xx_gpu {
+	struct adreno_gpu base;
+	struct platform_device *pdev;
+
+	/* if OCMEM is used for GMEM: */
+	uint32_t ocmem_base;
+	void *ocmem_hdl;
+};
+#define to_a4xx_gpu(x) container_of(x, struct a4xx_gpu, base)
+
+#endif /* __A4XX_GPU_H__ */
diff --git a/drivers/gpu/drm/msm/adreno/adreno_common.xml.h b/drivers/gpu/drm/msm/adreno/adreno_common.xml.h
index 9de19ac2e86c..a4b33af9338d 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_common.xml.h
+++ b/drivers/gpu/drm/msm/adreno/adreno_common.xml.h
@@ -11,10 +11,10 @@ The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/adreno.xml               (    364 bytes, from 2013-11-30 14:47:15)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml  (   1453 bytes, from 2013-03-31 16:51:27)
 - /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml          (  32901 bytes, from 2014-06-02 15:21:30)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (   9859 bytes, from 2014-06-02 15:21:30)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  14477 bytes, from 2014-05-16 11:51:57)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  58020 bytes, from 2014-06-25 12:57:16)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (  26602 bytes, from 2014-06-25 12:57:16)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (  10551 bytes, from 2014-11-13 22:44:30)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  15053 bytes, from 2014-11-09 15:45:47)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  63169 bytes, from 2014-11-13 22:44:18)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (  49097 bytes, from 2014-11-14 15:38:00)
 
 Copyright (C) 2013-2014 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -105,6 +105,7 @@ enum adreno_rb_dither_mode {
 enum adreno_rb_depth_format {
 	DEPTHX_16 = 0,
 	DEPTHX_24_8 = 1,
+	DEPTHX_32 = 2,
 };
 
 enum adreno_rb_copy_control_mode {
@@ -132,6 +133,7 @@ enum a3xx_threadmode {
 };
 
 enum a3xx_instrbuffermode {
+	CACHE = 0,
 	BUFFER = 1,
 };
 
@@ -140,6 +142,13 @@ enum a3xx_threadsize {
 	FOUR_QUADS = 1,
 };
 
+enum a3xx_color_swap {
+	WZYX = 0,
+	WXYZ = 1,
+	ZYXW = 2,
+	XYZW = 3,
+};
+
 #define REG_AXXX_CP_RB_BASE					0x000001c0
 
 #define REG_AXXX_CP_RB_CNTL					0x000001c1
diff --git a/drivers/gpu/drm/msm/adreno/adreno_device.c b/drivers/gpu/drm/msm/adreno/adreno_device.c
new file mode 100644
index 000000000000..be83dee83d08
--- /dev/null
+++ b/drivers/gpu/drm/msm/adreno/adreno_device.c
@@ -0,0 +1,298 @@
+/*
+ * Copyright (C) 2013-2014 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * Copyright (c) 2014 The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "adreno_gpu.h"
+
+#if defined(CONFIG_MSM_BUS_SCALING) && !defined(CONFIG_OF)
+#  include <mach/kgsl.h>
+#endif
+
+#define ANY_ID 0xff
+
+bool hang_debug = false;
+MODULE_PARM_DESC(hang_debug, "Dump registers when hang is detected (can be slow!)");
+module_param_named(hang_debug, hang_debug, bool, 0600);
+
+struct msm_gpu *a3xx_gpu_init(struct drm_device *dev);
+struct msm_gpu *a4xx_gpu_init(struct drm_device *dev);
+
+static const struct adreno_info gpulist[] = {
+	{
+		.rev   = ADRENO_REV(3, 0, 5, ANY_ID),
+		.revn  = 305,
+		.name  = "A305",
+		.pm4fw = "a300_pm4.fw",
+		.pfpfw = "a300_pfp.fw",
+		.gmem  = SZ_256K,
+		.init  = a3xx_gpu_init,
+	}, {
+		.rev   = ADRENO_REV(3, 2, ANY_ID, ANY_ID),
+		.revn  = 320,
+		.name  = "A320",
+		.pm4fw = "a300_pm4.fw",
+		.pfpfw = "a300_pfp.fw",
+		.gmem  = SZ_512K,
+		.init  = a3xx_gpu_init,
+	}, {
+		.rev   = ADRENO_REV(3, 3, 0, ANY_ID),
+		.revn  = 330,
+		.name  = "A330",
+		.pm4fw = "a330_pm4.fw",
+		.pfpfw = "a330_pfp.fw",
+		.gmem  = SZ_1M,
+		.init  = a3xx_gpu_init,
+	}, {
+		.rev   = ADRENO_REV(4, 2, 0, ANY_ID),
+		.revn  = 420,
+		.name  = "A420",
+		.pm4fw = "a420_pm4.fw",
+		.pfpfw = "a420_pfp.fw",
+		.gmem  = (SZ_1M + SZ_512K),
+		.init  = a4xx_gpu_init,
+	},
+};
+
+MODULE_FIRMWARE("a300_pm4.fw");
+MODULE_FIRMWARE("a300_pfp.fw");
+MODULE_FIRMWARE("a330_pm4.fw");
+MODULE_FIRMWARE("a330_pfp.fw");
+MODULE_FIRMWARE("a420_pm4.fw");
+MODULE_FIRMWARE("a420_pfp.fw");
+
+static inline bool _rev_match(uint8_t entry, uint8_t id)
+{
+	return (entry == ANY_ID) || (entry == id);
+}
+
+const struct adreno_info *adreno_info(struct adreno_rev rev)
+{
+	int i;
+
+	/* identify gpu: */
+	for (i = 0; i < ARRAY_SIZE(gpulist); i++) {
+		const struct adreno_info *info = &gpulist[i];
+		if (_rev_match(info->rev.core, rev.core) &&
+				_rev_match(info->rev.major, rev.major) &&
+				_rev_match(info->rev.minor, rev.minor) &&
+				_rev_match(info->rev.patchid, rev.patchid))
+			return info;
+	}
+
+	return NULL;
+}
+
+struct msm_gpu *adreno_load_gpu(struct drm_device *dev)
+{
+	struct msm_drm_private *priv = dev->dev_private;
+	struct platform_device *pdev = priv->gpu_pdev;
+	struct adreno_platform_config *config;
+	struct adreno_rev rev;
+	const struct adreno_info *info;
+	struct msm_gpu *gpu = NULL;
+
+	if (!pdev) {
+		dev_err(dev->dev, "no adreno device\n");
+		return NULL;
+	}
+
+	config = pdev->dev.platform_data;
+	rev = config->rev;
+	info = adreno_info(config->rev);
+
+	if (!info) {
+		dev_warn(dev->dev, "Unknown GPU revision: %u.%u.%u.%u\n",
+				rev.core, rev.major, rev.minor, rev.patchid);
+		return NULL;
+	}
+
+	DBG("Found GPU: %u.%u.%u.%u",  rev.core, rev.major,
+			rev.minor, rev.patchid);
+
+	gpu = info->init(dev);
+	if (IS_ERR(gpu)) {
+		dev_warn(dev->dev, "failed to load adreno gpu\n");
+		gpu = NULL;
+		/* not fatal */
+	}
+
+	if (gpu) {
+		int ret;
+		mutex_lock(&dev->struct_mutex);
+		gpu->funcs->pm_resume(gpu);
+		mutex_unlock(&dev->struct_mutex);
+		ret = gpu->funcs->hw_init(gpu);
+		if (ret) {
+			dev_err(dev->dev, "gpu hw init failed: %d\n", ret);
+			gpu->funcs->destroy(gpu);
+			gpu = NULL;
+		} else {
+			/* give inactive pm a chance to kick in: */
+			msm_gpu_retire(gpu);
+		}
+	}
+
+	return gpu;
+}
+
+static void set_gpu_pdev(struct drm_device *dev,
+		struct platform_device *pdev)
+{
+	struct msm_drm_private *priv = dev->dev_private;
+	priv->gpu_pdev = pdev;
+}
+
+static int adreno_bind(struct device *dev, struct device *master, void *data)
+{
+	static struct adreno_platform_config config = {};
+#ifdef CONFIG_OF
+	struct device_node *child, *node = dev->of_node;
+	u32 val;
+	int ret;
+
+	ret = of_property_read_u32(node, "qcom,chipid", &val);
+	if (ret) {
+		dev_err(dev, "could not find chipid: %d\n", ret);
+		return ret;
+	}
+
+	config.rev = ADRENO_REV((val >> 24) & 0xff,
+			(val >> 16) & 0xff, (val >> 8) & 0xff, val & 0xff);
+
+	/* find clock rates: */
+	config.fast_rate = 0;
+	config.slow_rate = ~0;
+	for_each_child_of_node(node, child) {
+		if (of_device_is_compatible(child, "qcom,gpu-pwrlevels")) {
+			struct device_node *pwrlvl;
+			for_each_child_of_node(child, pwrlvl) {
+				ret = of_property_read_u32(pwrlvl, "qcom,gpu-freq", &val);
+				if (ret) {
+					dev_err(dev, "could not find gpu-freq: %d\n", ret);
+					return ret;
+				}
+				config.fast_rate = max(config.fast_rate, val);
+				config.slow_rate = min(config.slow_rate, val);
+			}
+		}
+	}
+
+	if (!config.fast_rate) {
+		dev_err(dev, "could not find clk rates\n");
+		return -ENXIO;
+	}
+
+#else
+	struct kgsl_device_platform_data *pdata = dev->platform_data;
+	uint32_t version = socinfo_get_version();
+	if (cpu_is_apq8064ab()) {
+		config.fast_rate = 450000000;
+		config.slow_rate = 27000000;
+		config.bus_freq  = 4;
+		config.rev = ADRENO_REV(3, 2, 1, 0);
+	} else if (cpu_is_apq8064()) {
+		config.fast_rate = 400000000;
+		config.slow_rate = 27000000;
+		config.bus_freq  = 4;
+
+		if (SOCINFO_VERSION_MAJOR(version) == 2)
+			config.rev = ADRENO_REV(3, 2, 0, 2);
+		else if ((SOCINFO_VERSION_MAJOR(version) == 1) &&
+				(SOCINFO_VERSION_MINOR(version) == 1))
+			config.rev = ADRENO_REV(3, 2, 0, 1);
+		else
+			config.rev = ADRENO_REV(3, 2, 0, 0);
+
+	} else if (cpu_is_msm8960ab()) {
+		config.fast_rate = 400000000;
+		config.slow_rate = 320000000;
+		config.bus_freq  = 4;
+
+		if (SOCINFO_VERSION_MINOR(version) == 0)
+			config.rev = ADRENO_REV(3, 2, 1, 0);
+		else
+			config.rev = ADRENO_REV(3, 2, 1, 1);
+
+	} else if (cpu_is_msm8930()) {
+		config.fast_rate = 400000000;
+		config.slow_rate = 27000000;
+		config.bus_freq  = 3;
+
+		if ((SOCINFO_VERSION_MAJOR(version) == 1) &&
+			(SOCINFO_VERSION_MINOR(version) == 2))
+			config.rev = ADRENO_REV(3, 0, 5, 2);
+		else
+			config.rev = ADRENO_REV(3, 0, 5, 0);
+
+	}
+#  ifdef CONFIG_MSM_BUS_SCALING
+	config.bus_scale_table = pdata->bus_scale_table;
+#  endif
+#endif
+	dev->platform_data = &config;
+	set_gpu_pdev(dev_get_drvdata(master), to_platform_device(dev));
+	return 0;
+}
+
+static void adreno_unbind(struct device *dev, struct device *master,
+		void *data)
+{
+	set_gpu_pdev(dev_get_drvdata(master), NULL);
+}
+
+static const struct component_ops a3xx_ops = {
+		.bind   = adreno_bind,
+		.unbind = adreno_unbind,
+};
+
+static int adreno_probe(struct platform_device *pdev)
+{
+	return component_add(&pdev->dev, &a3xx_ops);
+}
+
+static int adreno_remove(struct platform_device *pdev)
+{
+	component_del(&pdev->dev, &a3xx_ops);
+	return 0;
+}
+
+static const struct of_device_id dt_match[] = {
+	{ .compatible = "qcom,adreno-3xx" },
+	/* for backwards compat w/ downstream kgsl DT files: */
+	{ .compatible = "qcom,kgsl-3d0" },
+	{}
+};
+
+static struct platform_driver adreno_driver = {
+	.probe = adreno_probe,
+	.remove = adreno_remove,
+	.driver = {
+		.name = "adreno",
+		.of_match_table = dt_match,
+	},
+};
+
+void __init adreno_register(void)
+{
+	platform_driver_register(&adreno_driver);
+}
+
+void __exit adreno_unregister(void)
+{
+	platform_driver_unregister(&adreno_driver);
+}
diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
index 655ce5b14ad0..94a5bee69fe7 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
@@ -2,6 +2,8 @@
  * Copyright (C) 2013 Red Hat
  * Author: Rob Clark <robdclark@gmail.com>
  *
+ * Copyright (c) 2014 The Linux Foundation. All rights reserved.
+ *
  * This program is free software; you can redistribute it and/or modify it
  * under the terms of the GNU General Public License version 2 as published by
  * the Free Software Foundation.
@@ -19,46 +21,6 @@
 #include "msm_gem.h"
 #include "msm_mmu.h"
 
-struct adreno_info {
-	struct adreno_rev rev;
-	uint32_t revn;
-	const char *name;
-	const char *pm4fw, *pfpfw;
-	uint32_t gmem;
-};
-
-#define ANY_ID 0xff
-
-static const struct adreno_info gpulist[] = {
-	{
-		.rev   = ADRENO_REV(3, 0, 5, ANY_ID),
-		.revn  = 305,
-		.name  = "A305",
-		.pm4fw = "a300_pm4.fw",
-		.pfpfw = "a300_pfp.fw",
-		.gmem  = SZ_256K,
-	}, {
-		.rev   = ADRENO_REV(3, 2, ANY_ID, ANY_ID),
-		.revn  = 320,
-		.name  = "A320",
-		.pm4fw = "a300_pm4.fw",
-		.pfpfw = "a300_pfp.fw",
-		.gmem  = SZ_512K,
-	}, {
-		.rev   = ADRENO_REV(3, 3, 0, ANY_ID),
-		.revn  = 330,
-		.name  = "A330",
-		.pm4fw = "a330_pm4.fw",
-		.pfpfw = "a330_pfp.fw",
-		.gmem  = SZ_1M,
-	},
-};
-
-MODULE_FIRMWARE("a300_pm4.fw");
-MODULE_FIRMWARE("a300_pfp.fw");
-MODULE_FIRMWARE("a330_pm4.fw");
-MODULE_FIRMWARE("a330_pfp.fw");
-
 #define RB_SIZE    SZ_32K
 #define RB_BLKSIZE 16
 
@@ -103,19 +65,21 @@ int adreno_hw_init(struct msm_gpu *gpu)
 	}
 
 	/* Setup REG_CP_RB_CNTL: */
-	gpu_write(gpu, REG_AXXX_CP_RB_CNTL,
+	adreno_gpu_write(adreno_gpu, REG_ADRENO_CP_RB_CNTL,
 			/* size is log2(quad-words): */
 			AXXX_CP_RB_CNTL_BUFSZ(ilog2(gpu->rb->size / 8)) |
 			AXXX_CP_RB_CNTL_BLKSZ(ilog2(RB_BLKSIZE / 8)));
 
 	/* Setup ringbuffer address: */
-	gpu_write(gpu, REG_AXXX_CP_RB_BASE, gpu->rb_iova);
-	gpu_write(gpu, REG_AXXX_CP_RB_RPTR_ADDR, rbmemptr(adreno_gpu, rptr));
+	adreno_gpu_write(adreno_gpu, REG_ADRENO_CP_RB_BASE, gpu->rb_iova);
+	adreno_gpu_write(adreno_gpu, REG_ADRENO_CP_RB_RPTR_ADDR,
+			rbmemptr(adreno_gpu, rptr));
 
 	/* Setup scratch/timestamp: */
-	gpu_write(gpu, REG_AXXX_SCRATCH_ADDR, rbmemptr(adreno_gpu, fence));
+	adreno_gpu_write(adreno_gpu, REG_ADRENO_SCRATCH_ADDR,
+			rbmemptr(adreno_gpu, fence));
 
-	gpu_write(gpu, REG_AXXX_SCRATCH_UMSK, 0x1);
+	adreno_gpu_write(adreno_gpu, REG_ADRENO_SCRATCH_UMSK, 0x1);
 
 	return 0;
 }
@@ -191,7 +155,7 @@ int adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 	OUT_PKT0(ring, REG_AXXX_CP_SCRATCH_REG2, 1);
 	OUT_RING(ring, submit->fence);
 
-	if (adreno_is_a3xx(adreno_gpu)) {
+	if (adreno_is_a3xx(adreno_gpu) || adreno_is_a4xx(adreno_gpu)) {
 		/* Flush HLSQ lazy updates to make sure there is nothing
 		 * pending for indirect loads after the timestamp has
 		 * passed:
@@ -228,12 +192,13 @@ int adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 
 void adreno_flush(struct msm_gpu *gpu)
 {
+	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
 	uint32_t wptr = get_wptr(gpu->rb);
 
 	/* ensure writes to ringbuffer have hit system memory: */
 	mb();
 
-	gpu_write(gpu, REG_AXXX_CP_RB_WPTR, wptr);
+	adreno_gpu_write(adreno_gpu, REG_ADRENO_CP_RB_WPTR, wptr);
 }
 
 void adreno_idle(struct msm_gpu *gpu)
@@ -252,6 +217,7 @@ void adreno_idle(struct msm_gpu *gpu)
 void adreno_show(struct msm_gpu *gpu, struct seq_file *m)
 {
 	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
+	int i;
 
 	seq_printf(m, "revision: %d (%d.%d.%d.%d)\n",
 			adreno_gpu->info->revn, adreno_gpu->rev.core,
@@ -263,6 +229,23 @@ void adreno_show(struct msm_gpu *gpu, struct seq_file *m)
 	seq_printf(m, "rptr:     %d\n", adreno_gpu->memptrs->rptr);
 	seq_printf(m, "wptr:     %d\n", adreno_gpu->memptrs->wptr);
 	seq_printf(m, "rb wptr:  %d\n", get_wptr(gpu->rb));
+
+	gpu->funcs->pm_resume(gpu);
+
+	/* dump these out in a form that can be parsed by demsm: */
+	seq_printf(m, "IO:region %s 00000000 00020000\n", gpu->name);
+	for (i = 0; adreno_gpu->registers[i] != ~0; i += 2) {
+		uint32_t start = adreno_gpu->registers[i];
+		uint32_t end   = adreno_gpu->registers[i+1];
+		uint32_t addr;
+
+		for (addr = start; addr <= end; addr++) {
+			uint32_t val = gpu_read(gpu, addr);
+			seq_printf(m, "IO:R %08x %08x\n", addr<<2, val);
+		}
+	}
+
+	gpu->funcs->pm_suspend(gpu);
 }
 #endif
 
@@ -270,6 +253,7 @@ void adreno_show(struct msm_gpu *gpu, struct seq_file *m)
 void adreno_dump(struct msm_gpu *gpu)
 {
 	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
+	int i;
 
 	printk("revision: %d (%d.%d.%d.%d)\n",
 			adreno_gpu->info->revn, adreno_gpu->rev.core,
@@ -282,6 +266,18 @@ void adreno_dump(struct msm_gpu *gpu)
 	printk("wptr:     %d\n", adreno_gpu->memptrs->wptr);
 	printk("rb wptr:  %d\n", get_wptr(gpu->rb));
 
+	/* dump these out in a form that can be parsed by demsm: */
+	printk("IO:region %s 00000000 00020000\n", gpu->name);
+	for (i = 0; adreno_gpu->registers[i] != ~0; i += 2) {
+		uint32_t start = adreno_gpu->registers[i];
+		uint32_t end   = adreno_gpu->registers[i+1];
+		uint32_t addr;
+
+		for (addr = start; addr <= end; addr++) {
+			uint32_t val = gpu_read(gpu, addr);
+			printk("IO:R %08x %08x\n", addr<<2, val);
+		}
+	}
 }
 
 static uint32_t ring_freewords(struct msm_gpu *gpu)
@@ -304,65 +300,51 @@ static const char *iommu_ports[] = {
 		"gfx3d1_user", "gfx3d1_priv",
 };
 
-static inline bool _rev_match(uint8_t entry, uint8_t id)
-{
-	return (entry == ANY_ID) || (entry == id);
-}
-
 int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev,
-		struct adreno_gpu *gpu, const struct adreno_gpu_funcs *funcs,
-		struct adreno_rev rev)
+		struct adreno_gpu *adreno_gpu, const struct adreno_gpu_funcs *funcs)
 {
+	struct adreno_platform_config *config = pdev->dev.platform_data;
+	struct msm_gpu *gpu = &adreno_gpu->base;
 	struct msm_mmu *mmu;
-	int i, ret;
-
-	/* identify gpu: */
-	for (i = 0; i < ARRAY_SIZE(gpulist); i++) {
-		const struct adreno_info *info = &gpulist[i];
-		if (_rev_match(info->rev.core, rev.core) &&
-				_rev_match(info->rev.major, rev.major) &&
-				_rev_match(info->rev.minor, rev.minor) &&
-				_rev_match(info->rev.patchid, rev.patchid)) {
-			gpu->info = info;
-			gpu->revn = info->revn;
-			break;
-		}
-	}
+	int ret;
 
-	if (i == ARRAY_SIZE(gpulist)) {
-		dev_err(drm->dev, "Unknown GPU revision: %u.%u.%u.%u\n",
-				rev.core, rev.major, rev.minor, rev.patchid);
-		return -ENXIO;
-	}
+	adreno_gpu->funcs = funcs;
+	adreno_gpu->info = adreno_info(config->rev);
+	adreno_gpu->gmem = adreno_gpu->info->gmem;
+	adreno_gpu->revn = adreno_gpu->info->revn;
+	adreno_gpu->rev = config->rev;
+
+	gpu->fast_rate = config->fast_rate;
+	gpu->slow_rate = config->slow_rate;
+	gpu->bus_freq  = config->bus_freq;
+#ifdef CONFIG_MSM_BUS_SCALING
+	gpu->bus_scale_table = config->bus_scale_table;
+#endif
 
-	DBG("Found GPU: %s (%u.%u.%u.%u)", gpu->info->name,
-			rev.core, rev.major, rev.minor, rev.patchid);
+	DBG("fast_rate=%u, slow_rate=%u, bus_freq=%u",
+			gpu->fast_rate, gpu->slow_rate, gpu->bus_freq);
 
-	gpu->funcs = funcs;
-	gpu->gmem = gpu->info->gmem;
-	gpu->rev = rev;
+	ret = msm_gpu_init(drm, pdev, &adreno_gpu->base, &funcs->base,
+			adreno_gpu->info->name, "kgsl_3d0_reg_memory", "kgsl_3d0_irq",
+			RB_SIZE);
+	if (ret)
+		return ret;
 
-	ret = request_firmware(&gpu->pm4, gpu->info->pm4fw, drm->dev);
+	ret = request_firmware(&adreno_gpu->pm4, adreno_gpu->info->pm4fw, drm->dev);
 	if (ret) {
 		dev_err(drm->dev, "failed to load %s PM4 firmware: %d\n",
-				gpu->info->pm4fw, ret);
+				adreno_gpu->info->pm4fw, ret);
 		return ret;
 	}
 
-	ret = request_firmware(&gpu->pfp, gpu->info->pfpfw, drm->dev);
+	ret = request_firmware(&adreno_gpu->pfp, adreno_gpu->info->pfpfw, drm->dev);
 	if (ret) {
 		dev_err(drm->dev, "failed to load %s PFP firmware: %d\n",
-				gpu->info->pfpfw, ret);
+				adreno_gpu->info->pfpfw, ret);
 		return ret;
 	}
 
-	ret = msm_gpu_init(drm, pdev, &gpu->base, &funcs->base,
-			gpu->info->name, "kgsl_3d0_reg_memory", "kgsl_3d0_irq",
-			RB_SIZE);
-	if (ret)
-		return ret;
-
-	mmu = gpu->base.mmu;
+	mmu = gpu->mmu;
 	if (mmu) {
 		ret = mmu->funcs->attach(mmu, iommu_ports,
 				ARRAY_SIZE(iommu_ports));
@@ -371,24 +353,24 @@ int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 	}
 
 	mutex_lock(&drm->struct_mutex);
-	gpu->memptrs_bo = msm_gem_new(drm, sizeof(*gpu->memptrs),
+	adreno_gpu->memptrs_bo = msm_gem_new(drm, sizeof(*adreno_gpu->memptrs),
 			MSM_BO_UNCACHED);
 	mutex_unlock(&drm->struct_mutex);
-	if (IS_ERR(gpu->memptrs_bo)) {
-		ret = PTR_ERR(gpu->memptrs_bo);
-		gpu->memptrs_bo = NULL;
+	if (IS_ERR(adreno_gpu->memptrs_bo)) {
+		ret = PTR_ERR(adreno_gpu->memptrs_bo);
+		adreno_gpu->memptrs_bo = NULL;
 		dev_err(drm->dev, "could not allocate memptrs: %d\n", ret);
 		return ret;
 	}
 
-	gpu->memptrs = msm_gem_vaddr(gpu->memptrs_bo);
-	if (!gpu->memptrs) {
+	adreno_gpu->memptrs = msm_gem_vaddr(adreno_gpu->memptrs_bo);
+	if (!adreno_gpu->memptrs) {
 		dev_err(drm->dev, "could not vmap memptrs\n");
 		return -ENOMEM;
 	}
 
-	ret = msm_gem_get_iova(gpu->memptrs_bo, gpu->base.id,
-			&gpu->memptrs_iova);
+	ret = msm_gem_get_iova(adreno_gpu->memptrs_bo, gpu->id,
+			&adreno_gpu->memptrs_iova);
 	if (ret) {
 		dev_err(drm->dev, "could not map memptrs: %d\n", ret);
 		return ret;
@@ -404,9 +386,7 @@ void adreno_gpu_cleanup(struct adreno_gpu *gpu)
 			msm_gem_put_iova(gpu->memptrs_bo, gpu->base.id);
 		drm_gem_object_unreference(gpu->memptrs_bo);
 	}
-	if (gpu->pm4)
-		release_firmware(gpu->pm4);
-	if (gpu->pfp)
-		release_firmware(gpu->pfp);
+	release_firmware(gpu->pm4);
+	release_firmware(gpu->pfp);
 	msm_gpu_cleanup(&gpu->base);
 }
diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.h b/drivers/gpu/drm/msm/adreno/adreno_gpu.h
index 63c36ce33020..a0cc30977e67 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_gpu.h
+++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.h
@@ -2,6 +2,8 @@
  * Copyright (C) 2013 Red Hat
  * Author: Rob Clark <robdclark@gmail.com>
  *
+ * Copyright (c) 2014 The Linux Foundation. All rights reserved.
+ *
  * This program is free software; you can redistribute it and/or modify it
  * under the terms of the GNU General Public License version 2 as published by
  * the Free Software Foundation.
@@ -25,6 +27,81 @@
 #include "adreno_common.xml.h"
 #include "adreno_pm4.xml.h"
 
+#define REG_ADRENO_DEFINE(_offset, _reg) [_offset] = (_reg) + 1
+/**
+ * adreno_regs: List of registers that are used in across all
+ * 3D devices. Each device type has different offset value for the same
+ * register, so an array of register offsets are declared for every device
+ * and are indexed by the enumeration values defined in this enum
+ */
+enum adreno_regs {
+	REG_ADRENO_CP_DEBUG,
+	REG_ADRENO_CP_ME_RAM_WADDR,
+	REG_ADRENO_CP_ME_RAM_DATA,
+	REG_ADRENO_CP_PFP_UCODE_DATA,
+	REG_ADRENO_CP_PFP_UCODE_ADDR,
+	REG_ADRENO_CP_WFI_PEND_CTR,
+	REG_ADRENO_CP_RB_BASE,
+	REG_ADRENO_CP_RB_RPTR_ADDR,
+	REG_ADRENO_CP_RB_RPTR,
+	REG_ADRENO_CP_RB_WPTR,
+	REG_ADRENO_CP_PROTECT_CTRL,
+	REG_ADRENO_CP_ME_CNTL,
+	REG_ADRENO_CP_RB_CNTL,
+	REG_ADRENO_CP_IB1_BASE,
+	REG_ADRENO_CP_IB1_BUFSZ,
+	REG_ADRENO_CP_IB2_BASE,
+	REG_ADRENO_CP_IB2_BUFSZ,
+	REG_ADRENO_CP_TIMESTAMP,
+	REG_ADRENO_CP_ME_RAM_RADDR,
+	REG_ADRENO_CP_ROQ_ADDR,
+	REG_ADRENO_CP_ROQ_DATA,
+	REG_ADRENO_CP_MERCIU_ADDR,
+	REG_ADRENO_CP_MERCIU_DATA,
+	REG_ADRENO_CP_MERCIU_DATA2,
+	REG_ADRENO_CP_MEQ_ADDR,
+	REG_ADRENO_CP_MEQ_DATA,
+	REG_ADRENO_CP_HW_FAULT,
+	REG_ADRENO_CP_PROTECT_STATUS,
+	REG_ADRENO_SCRATCH_ADDR,
+	REG_ADRENO_SCRATCH_UMSK,
+	REG_ADRENO_SCRATCH_REG2,
+	REG_ADRENO_RBBM_STATUS,
+	REG_ADRENO_RBBM_PERFCTR_CTL,
+	REG_ADRENO_RBBM_PERFCTR_LOAD_CMD0,
+	REG_ADRENO_RBBM_PERFCTR_LOAD_CMD1,
+	REG_ADRENO_RBBM_PERFCTR_LOAD_CMD2,
+	REG_ADRENO_RBBM_PERFCTR_PWR_1_LO,
+	REG_ADRENO_RBBM_INT_0_MASK,
+	REG_ADRENO_RBBM_INT_0_STATUS,
+	REG_ADRENO_RBBM_AHB_ERROR_STATUS,
+	REG_ADRENO_RBBM_PM_OVERRIDE2,
+	REG_ADRENO_RBBM_AHB_CMD,
+	REG_ADRENO_RBBM_INT_CLEAR_CMD,
+	REG_ADRENO_RBBM_SW_RESET_CMD,
+	REG_ADRENO_RBBM_CLOCK_CTL,
+	REG_ADRENO_RBBM_AHB_ME_SPLIT_STATUS,
+	REG_ADRENO_RBBM_AHB_PFP_SPLIT_STATUS,
+	REG_ADRENO_VPC_DEBUG_RAM_SEL,
+	REG_ADRENO_VPC_DEBUG_RAM_READ,
+	REG_ADRENO_VSC_SIZE_ADDRESS,
+	REG_ADRENO_VFD_CONTROL_0,
+	REG_ADRENO_VFD_INDEX_MAX,
+	REG_ADRENO_SP_VS_PVT_MEM_ADDR_REG,
+	REG_ADRENO_SP_FS_PVT_MEM_ADDR_REG,
+	REG_ADRENO_SP_VS_OBJ_START_REG,
+	REG_ADRENO_SP_FS_OBJ_START_REG,
+	REG_ADRENO_PA_SC_AA_CONFIG,
+	REG_ADRENO_SQ_GPR_MANAGEMENT,
+	REG_ADRENO_SQ_INST_STORE_MANAGMENT,
+	REG_ADRENO_TP0_CHICKEN,
+	REG_ADRENO_RBBM_RBBM_CTL,
+	REG_ADRENO_UCHE_INVALIDATE0,
+	REG_ADRENO_RBBM_PERFCTR_LOAD_VALUE_LO,
+	REG_ADRENO_RBBM_PERFCTR_LOAD_VALUE_HI,
+	REG_ADRENO_REGISTER_MAX,
+};
+
 struct adreno_rev {
 	uint8_t  core;
 	uint8_t  major;
@@ -39,7 +116,16 @@ struct adreno_gpu_funcs {
 	struct msm_gpu_funcs base;
 };
 
-struct adreno_info;
+struct adreno_info {
+	struct adreno_rev rev;
+	uint32_t revn;
+	const char *name;
+	const char *pm4fw, *pfpfw;
+	uint32_t gmem;
+	struct msm_gpu *(*init)(struct drm_device *dev);
+};
+
+const struct adreno_info *adreno_info(struct adreno_rev rev);
 
 struct adreno_rbmemptrs {
 	volatile uint32_t rptr;
@@ -55,6 +141,9 @@ struct adreno_gpu {
 	uint32_t revn;  /* numeric revision name */
 	const struct adreno_gpu_funcs *funcs;
 
+	/* interesting register offsets to dump: */
+	const unsigned int *registers;
+
 	/* firmware: */
 	const struct firmware *pm4, *pfp;
 
@@ -64,6 +153,13 @@ struct adreno_gpu {
 	struct adreno_rbmemptrs *memptrs;
 	struct drm_gem_object *memptrs_bo;
 	uint32_t memptrs_iova;
+
+	/*
+	 * Register offsets are different between some GPUs.
+	 * GPU specific offsets will be exported by GPU specific
+	 * code (a3xx_gpu.c) and stored in this common location.
+	 */
+	const unsigned int *reg_offsets;
 };
 #define to_adreno_gpu(x) container_of(x, struct adreno_gpu, base)
 
@@ -116,6 +212,16 @@ static inline bool adreno_is_a330v2(struct adreno_gpu *gpu)
 	return adreno_is_a330(gpu) && (gpu->rev.patchid > 0);
 }
 
+static inline bool adreno_is_a4xx(struct adreno_gpu *gpu)
+{
+	return (gpu->revn >= 400) && (gpu->revn < 500);
+}
+
+static inline int adreno_is_a420(struct adreno_gpu *gpu)
+{
+	return gpu->revn == 420;
+}
+
 int adreno_get_param(struct msm_gpu *gpu, uint32_t param, uint64_t *value);
 int adreno_hw_init(struct msm_gpu *gpu);
 uint32_t adreno_last_fence(struct msm_gpu *gpu);
@@ -131,8 +237,7 @@ void adreno_dump(struct msm_gpu *gpu);
 void adreno_wait_ring(struct msm_gpu *gpu, uint32_t ndwords);
 
 int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev,
-		struct adreno_gpu *gpu, const struct adreno_gpu_funcs *funcs,
-		struct adreno_rev rev);
+		struct adreno_gpu *gpu, const struct adreno_gpu_funcs *funcs);
 void adreno_gpu_cleanup(struct adreno_gpu *gpu);
 
 
@@ -160,5 +265,37 @@ OUT_PKT3(struct msm_ringbuffer *ring, uint8_t opcode, uint16_t cnt)
 	OUT_RING(ring, CP_TYPE3_PKT | ((cnt-1) << 16) | ((opcode & 0xFF) << 8));
 }
 
+/*
+ * adreno_checkreg_off() - Checks the validity of a register enum
+ * @gpu:		Pointer to struct adreno_gpu
+ * @offset_name:	The register enum that is checked
+ */
+static inline bool adreno_reg_check(struct adreno_gpu *gpu,
+		enum adreno_regs offset_name)
+{
+	if (offset_name >= REG_ADRENO_REGISTER_MAX ||
+			!gpu->reg_offsets[offset_name]) {
+		BUG();
+	}
+	return true;
+}
+
+static inline u32 adreno_gpu_read(struct adreno_gpu *gpu,
+		enum adreno_regs offset_name)
+{
+	u32 reg = gpu->reg_offsets[offset_name];
+	u32 val = 0;
+	if(adreno_reg_check(gpu,offset_name))
+		val = gpu_read(&gpu->base, reg - 1);
+	return val;
+}
+
+static inline void adreno_gpu_write(struct adreno_gpu *gpu,
+		enum adreno_regs offset_name, u32 data)
+{
+	u32 reg = gpu->reg_offsets[offset_name];
+	if(adreno_reg_check(gpu, offset_name))
+		gpu_write(&gpu->base, reg - 1, data);
+}
 
 #endif /* __ADRENO_GPU_H__ */
diff --git a/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h b/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h
index 4eee0ec8f069..6a75cee94d81 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h
+++ b/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h
@@ -11,10 +11,10 @@ The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/adreno.xml               (    364 bytes, from 2013-11-30 14:47:15)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml  (   1453 bytes, from 2013-03-31 16:51:27)
 - /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml          (  32901 bytes, from 2014-06-02 15:21:30)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (   9859 bytes, from 2014-06-02 15:21:30)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  14477 bytes, from 2014-05-16 11:51:57)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  58020 bytes, from 2014-06-25 12:57:16)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (  26602 bytes, from 2014-06-25 12:57:16)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (  10551 bytes, from 2014-11-13 22:44:30)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  15053 bytes, from 2014-11-09 15:45:47)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  63169 bytes, from 2014-11-13 22:44:18)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (  49097 bytes, from 2014-11-14 15:38:00)
 
 Copyright (C) 2013-2014 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -157,18 +157,23 @@ enum adreno_pm4_type3_packets {
 	CP_IM_STORE = 44,
 	CP_SET_DRAW_INIT_FLAGS = 75,
 	CP_SET_PROTECTED_MODE = 95,
+	CP_BOOTSTRAP_UCODE = 111,
 	CP_LOAD_STATE = 48,
 	CP_COND_INDIRECT_BUFFER_PFE = 58,
 	CP_COND_INDIRECT_BUFFER_PFD = 50,
 	CP_INDIRECT_BUFFER_PFE = 63,
 	CP_SET_BIN = 76,
 	CP_TEST_TWO_MEMS = 113,
+	CP_REG_WR_NO_CTXT = 120,
+	CP_RECORD_PFP_TIMESTAMP = 17,
 	CP_WAIT_FOR_ME = 19,
 	CP_SET_DRAW_STATE = 67,
 	CP_DRAW_INDX_OFFSET = 56,
 	CP_DRAW_INDIRECT = 40,
 	CP_DRAW_INDX_INDIRECT = 41,
 	CP_DRAW_AUTO = 36,
+	CP_UNKNOWN_1A = 26,
+	CP_WIDE_REG_WRITE = 116,
 	IN_IB_PREFETCH_END = 23,
 	IN_SUBBLK_PREFETCH = 31,
 	IN_INSTR_PREFETCH = 32,
@@ -274,11 +279,11 @@ static inline uint32_t CP_DRAW_INDX_1_INDEX_SIZE(enum pc_di_index_size val)
 #define CP_DRAW_INDX_1_NOT_EOP					0x00001000
 #define CP_DRAW_INDX_1_SMALL_INDEX				0x00002000
 #define CP_DRAW_INDX_1_PRE_DRAW_INITIATOR_ENABLE		0x00004000
-#define CP_DRAW_INDX_1_NUM_INDICES__MASK			0xffff0000
-#define CP_DRAW_INDX_1_NUM_INDICES__SHIFT			16
-static inline uint32_t CP_DRAW_INDX_1_NUM_INDICES(uint32_t val)
+#define CP_DRAW_INDX_1_NUM_INSTANCES__MASK			0xff000000
+#define CP_DRAW_INDX_1_NUM_INSTANCES__SHIFT			24
+static inline uint32_t CP_DRAW_INDX_1_NUM_INSTANCES(uint32_t val)
 {
-	return ((val) << CP_DRAW_INDX_1_NUM_INDICES__SHIFT) & CP_DRAW_INDX_1_NUM_INDICES__MASK;
+	return ((val) << CP_DRAW_INDX_1_NUM_INSTANCES__SHIFT) & CP_DRAW_INDX_1_NUM_INSTANCES__MASK;
 }
 
 #define REG_CP_DRAW_INDX_2					0x00000002
@@ -289,20 +294,20 @@ static inline uint32_t CP_DRAW_INDX_2_NUM_INDICES(uint32_t val)
 	return ((val) << CP_DRAW_INDX_2_NUM_INDICES__SHIFT) & CP_DRAW_INDX_2_NUM_INDICES__MASK;
 }
 
-#define REG_CP_DRAW_INDX_2					0x00000002
-#define CP_DRAW_INDX_2_INDX_BASE__MASK				0xffffffff
-#define CP_DRAW_INDX_2_INDX_BASE__SHIFT				0
-static inline uint32_t CP_DRAW_INDX_2_INDX_BASE(uint32_t val)
+#define REG_CP_DRAW_INDX_3					0x00000003
+#define CP_DRAW_INDX_3_INDX_BASE__MASK				0xffffffff
+#define CP_DRAW_INDX_3_INDX_BASE__SHIFT				0
+static inline uint32_t CP_DRAW_INDX_3_INDX_BASE(uint32_t val)
 {
-	return ((val) << CP_DRAW_INDX_2_INDX_BASE__SHIFT) & CP_DRAW_INDX_2_INDX_BASE__MASK;
+	return ((val) << CP_DRAW_INDX_3_INDX_BASE__SHIFT) & CP_DRAW_INDX_3_INDX_BASE__MASK;
 }
 
-#define REG_CP_DRAW_INDX_2					0x00000002
-#define CP_DRAW_INDX_2_INDX_SIZE__MASK				0xffffffff
-#define CP_DRAW_INDX_2_INDX_SIZE__SHIFT				0
-static inline uint32_t CP_DRAW_INDX_2_INDX_SIZE(uint32_t val)
+#define REG_CP_DRAW_INDX_4					0x00000004
+#define CP_DRAW_INDX_4_INDX_SIZE__MASK				0xffffffff
+#define CP_DRAW_INDX_4_INDX_SIZE__SHIFT				0
+static inline uint32_t CP_DRAW_INDX_4_INDX_SIZE(uint32_t val)
 {
-	return ((val) << CP_DRAW_INDX_2_INDX_SIZE__SHIFT) & CP_DRAW_INDX_2_INDX_SIZE__MASK;
+	return ((val) << CP_DRAW_INDX_4_INDX_SIZE__SHIFT) & CP_DRAW_INDX_4_INDX_SIZE__MASK;
 }
 
 #define REG_CP_DRAW_INDX_2_0					0x00000000
@@ -341,11 +346,11 @@ static inline uint32_t CP_DRAW_INDX_2_1_INDEX_SIZE(enum pc_di_index_size val)
 #define CP_DRAW_INDX_2_1_NOT_EOP				0x00001000
 #define CP_DRAW_INDX_2_1_SMALL_INDEX				0x00002000
 #define CP_DRAW_INDX_2_1_PRE_DRAW_INITIATOR_ENABLE		0x00004000
-#define CP_DRAW_INDX_2_1_NUM_INDICES__MASK			0xffff0000
-#define CP_DRAW_INDX_2_1_NUM_INDICES__SHIFT			16
-static inline uint32_t CP_DRAW_INDX_2_1_NUM_INDICES(uint32_t val)
+#define CP_DRAW_INDX_2_1_NUM_INSTANCES__MASK			0xff000000
+#define CP_DRAW_INDX_2_1_NUM_INSTANCES__SHIFT			24
+static inline uint32_t CP_DRAW_INDX_2_1_NUM_INSTANCES(uint32_t val)
 {
-	return ((val) << CP_DRAW_INDX_2_1_NUM_INDICES__SHIFT) & CP_DRAW_INDX_2_1_NUM_INDICES__MASK;
+	return ((val) << CP_DRAW_INDX_2_1_NUM_INSTANCES__SHIFT) & CP_DRAW_INDX_2_1_NUM_INSTANCES__MASK;
 }
 
 #define REG_CP_DRAW_INDX_2_2					0x00000002
@@ -384,11 +389,11 @@ static inline uint32_t CP_DRAW_INDX_OFFSET_0_INDEX_SIZE(enum pc_di_index_size va
 #define CP_DRAW_INDX_OFFSET_0_NOT_EOP				0x00001000
 #define CP_DRAW_INDX_OFFSET_0_SMALL_INDEX			0x00002000
 #define CP_DRAW_INDX_OFFSET_0_PRE_DRAW_INITIATOR_ENABLE		0x00004000
-#define CP_DRAW_INDX_OFFSET_0_NUM_INDICES__MASK			0xffff0000
-#define CP_DRAW_INDX_OFFSET_0_NUM_INDICES__SHIFT		16
-static inline uint32_t CP_DRAW_INDX_OFFSET_0_NUM_INDICES(uint32_t val)
+#define CP_DRAW_INDX_OFFSET_0_NUM_INSTANCES__MASK		0xffff0000
+#define CP_DRAW_INDX_OFFSET_0_NUM_INSTANCES__SHIFT		16
+static inline uint32_t CP_DRAW_INDX_OFFSET_0_NUM_INSTANCES(uint32_t val)
 {
-	return ((val) << CP_DRAW_INDX_OFFSET_0_NUM_INDICES__SHIFT) & CP_DRAW_INDX_OFFSET_0_NUM_INDICES__MASK;
+	return ((val) << CP_DRAW_INDX_OFFSET_0_NUM_INSTANCES__SHIFT) & CP_DRAW_INDX_OFFSET_0_NUM_INSTANCES__MASK;
 }
 
 #define REG_CP_DRAW_INDX_OFFSET_1				0x00000001
@@ -401,20 +406,22 @@ static inline uint32_t CP_DRAW_INDX_OFFSET_2_NUM_INDICES(uint32_t val)
 	return ((val) << CP_DRAW_INDX_OFFSET_2_NUM_INDICES__SHIFT) & CP_DRAW_INDX_OFFSET_2_NUM_INDICES__MASK;
 }
 
-#define REG_CP_DRAW_INDX_OFFSET_2				0x00000002
-#define CP_DRAW_INDX_OFFSET_2_INDX_BASE__MASK			0xffffffff
-#define CP_DRAW_INDX_OFFSET_2_INDX_BASE__SHIFT			0
-static inline uint32_t CP_DRAW_INDX_OFFSET_2_INDX_BASE(uint32_t val)
+#define REG_CP_DRAW_INDX_OFFSET_3				0x00000003
+
+#define REG_CP_DRAW_INDX_OFFSET_4				0x00000004
+#define CP_DRAW_INDX_OFFSET_4_INDX_BASE__MASK			0xffffffff
+#define CP_DRAW_INDX_OFFSET_4_INDX_BASE__SHIFT			0
+static inline uint32_t CP_DRAW_INDX_OFFSET_4_INDX_BASE(uint32_t val)
 {
-	return ((val) << CP_DRAW_INDX_OFFSET_2_INDX_BASE__SHIFT) & CP_DRAW_INDX_OFFSET_2_INDX_BASE__MASK;
+	return ((val) << CP_DRAW_INDX_OFFSET_4_INDX_BASE__SHIFT) & CP_DRAW_INDX_OFFSET_4_INDX_BASE__MASK;
 }
 
-#define REG_CP_DRAW_INDX_OFFSET_2				0x00000002
-#define CP_DRAW_INDX_OFFSET_2_INDX_SIZE__MASK			0xffffffff
-#define CP_DRAW_INDX_OFFSET_2_INDX_SIZE__SHIFT			0
-static inline uint32_t CP_DRAW_INDX_OFFSET_2_INDX_SIZE(uint32_t val)
+#define REG_CP_DRAW_INDX_OFFSET_5				0x00000005
+#define CP_DRAW_INDX_OFFSET_5_INDX_SIZE__MASK			0xffffffff
+#define CP_DRAW_INDX_OFFSET_5_INDX_SIZE__SHIFT			0
+static inline uint32_t CP_DRAW_INDX_OFFSET_5_INDX_SIZE(uint32_t val)
 {
-	return ((val) << CP_DRAW_INDX_OFFSET_2_INDX_SIZE__SHIFT) & CP_DRAW_INDX_OFFSET_2_INDX_SIZE__MASK;
+	return ((val) << CP_DRAW_INDX_OFFSET_5_INDX_SIZE__SHIFT) & CP_DRAW_INDX_OFFSET_5_INDX_SIZE__MASK;
 }
 
 #define REG_CP_SET_DRAW_STATE_0					0x00000000
diff --git a/drivers/gpu/drm/msm/dsi/dsi.xml.h b/drivers/gpu/drm/msm/dsi/dsi.xml.h
index 0f1f5b9459a5..448438b759b4 100644
--- a/drivers/gpu/drm/msm/dsi/dsi.xml.h
+++ b/drivers/gpu/drm/msm/dsi/dsi.xml.h
@@ -10,14 +10,14 @@ git clone https://github.com/freedreno/envytools.git
 The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-06-25 12:55:02)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  20136 bytes, from 2014-10-31 16:51:39)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1940 bytes, from 2014-10-31 16:51:39)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  23963 bytes, from 2014-10-31 16:51:46)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
-- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1686 bytes, from 2014-10-31 16:48:57)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-06-25 12:53:44)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-07-17 15:33:30)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
diff --git a/drivers/gpu/drm/msm/dsi/mmss_cc.xml.h b/drivers/gpu/drm/msm/dsi/mmss_cc.xml.h
index d468f86f637c..c102a7f074ac 100644
--- a/drivers/gpu/drm/msm/dsi/mmss_cc.xml.h
+++ b/drivers/gpu/drm/msm/dsi/mmss_cc.xml.h
@@ -10,16 +10,16 @@ git clone https://github.com/freedreno/envytools.git
 The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-06-25 12:55:02)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  20136 bytes, from 2014-10-31 16:51:39)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1940 bytes, from 2014-10-31 16:51:39)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  23963 bytes, from 2014-10-31 16:51:46)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
-- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1686 bytes, from 2014-10-31 16:48:57)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-06-25 12:53:44)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-07-17 15:33:30)
 
-Copyright (C) 2013 by the following authors:
+Copyright (C) 2013-2014 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
 
 Permission is hereby granted, free of charge, to any person obtaining
@@ -112,5 +112,11 @@ static inline uint32_t MMSS_CC_CLK_NS_VAL(uint32_t val)
 	return ((val) << MMSS_CC_CLK_NS_VAL__SHIFT) & MMSS_CC_CLK_NS_VAL__MASK;
 }
 
+#define REG_MMSS_CC_DSI2_PIXEL_CC				0x00000094
+
+#define REG_MMSS_CC_DSI2_PIXEL_NS				0x000000e4
+
+#define REG_MMSS_CC_DSI2_PIXEL_CC2				0x00000264
+
 
 #endif /* MMSS_CC_XML */
diff --git a/drivers/gpu/drm/msm/dsi/sfpb.xml.h b/drivers/gpu/drm/msm/dsi/sfpb.xml.h
index da8740054cdf..a900134bdf33 100644
--- a/drivers/gpu/drm/msm/dsi/sfpb.xml.h
+++ b/drivers/gpu/drm/msm/dsi/sfpb.xml.h
@@ -10,14 +10,14 @@ git clone https://github.com/freedreno/envytools.git
 The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-06-25 12:55:02)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  20136 bytes, from 2014-10-31 16:51:39)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1940 bytes, from 2014-10-31 16:51:39)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  23963 bytes, from 2014-10-31 16:51:46)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
-- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1686 bytes, from 2014-10-31 16:48:57)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-06-25 12:53:44)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-07-17 15:33:30)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi.c b/drivers/gpu/drm/msm/hdmi/hdmi.c
index c6c9b02e0ada..062c68725376 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi.c
+++ b/drivers/gpu/drm/msm/hdmi/hdmi.c
@@ -15,6 +15,7 @@
  * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
+#include <linux/of_irq.h>
 #include "hdmi.h"
 
 void hdmi_set_mode(struct hdmi *hdmi, bool power_on)
@@ -39,7 +40,7 @@ void hdmi_set_mode(struct hdmi *hdmi, bool power_on)
 			power_on ? "Enable" : "Disable", ctrl);
 }
 
-irqreturn_t hdmi_irq(int irq, void *dev_id)
+static irqreturn_t hdmi_irq(int irq, void *dev_id)
 {
 	struct hdmi *hdmi = dev_id;
 
@@ -54,9 +55,8 @@ irqreturn_t hdmi_irq(int irq, void *dev_id)
 	return IRQ_HANDLED;
 }
 
-void hdmi_destroy(struct kref *kref)
+static void hdmi_destroy(struct hdmi *hdmi)
 {
-	struct hdmi *hdmi = container_of(kref, struct hdmi, refcount);
 	struct hdmi_phy *phy = hdmi->phy;
 
 	if (phy)
@@ -68,37 +68,24 @@ void hdmi_destroy(struct kref *kref)
 	platform_set_drvdata(hdmi->pdev, NULL);
 }
 
-/* initialize connector */
-struct hdmi *hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
+/* construct hdmi at bind/probe time, grab all the resources.  If
+ * we are to EPROBE_DEFER we want to do it here, rather than later
+ * at modeset_init() time
+ */
+static struct hdmi *hdmi_init(struct platform_device *pdev)
 {
+	struct hdmi_platform_config *config = pdev->dev.platform_data;
 	struct hdmi *hdmi = NULL;
-	struct msm_drm_private *priv = dev->dev_private;
-	struct platform_device *pdev = priv->hdmi_pdev;
-	struct hdmi_platform_config *config;
 	int i, ret;
 
-	if (!pdev) {
-		dev_err(dev->dev, "no hdmi device\n");
-		ret = -ENXIO;
-		goto fail;
-	}
-
-	config = pdev->dev.platform_data;
-
-	hdmi = kzalloc(sizeof(*hdmi), GFP_KERNEL);
+	hdmi = devm_kzalloc(&pdev->dev, sizeof(*hdmi), GFP_KERNEL);
 	if (!hdmi) {
 		ret = -ENOMEM;
 		goto fail;
 	}
 
-	kref_init(&hdmi->refcount);
-
-	hdmi->dev = dev;
 	hdmi->pdev = pdev;
 	hdmi->config = config;
-	hdmi->encoder = encoder;
-
-	hdmi_audio_infoframe_init(&hdmi->audio.infoframe);
 
 	/* not sure about which phy maps to which msm.. probably I miss some */
 	if (config->phy_init)
@@ -108,7 +95,7 @@ struct hdmi *hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 
 	if (IS_ERR(hdmi->phy)) {
 		ret = PTR_ERR(hdmi->phy);
-		dev_err(dev->dev, "failed to load phy: %d\n", ret);
+		dev_err(&pdev->dev, "failed to load phy: %d\n", ret);
 		hdmi->phy = NULL;
 		goto fail;
 	}
@@ -123,11 +110,11 @@ struct hdmi *hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 	for (i = 0; i < config->hpd_reg_cnt; i++) {
 		struct regulator *reg;
 
-		reg = devm_regulator_get_exclusive(&pdev->dev,
+		reg = devm_regulator_get(&pdev->dev,
 				config->hpd_reg_names[i]);
 		if (IS_ERR(reg)) {
 			ret = PTR_ERR(reg);
-			dev_err(dev->dev, "failed to get hpd regulator: %s (%d)\n",
+			dev_err(&pdev->dev, "failed to get hpd regulator: %s (%d)\n",
 					config->hpd_reg_names[i], ret);
 			goto fail;
 		}
@@ -139,11 +126,11 @@ struct hdmi *hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 	for (i = 0; i < config->pwr_reg_cnt; i++) {
 		struct regulator *reg;
 
-		reg = devm_regulator_get_exclusive(&pdev->dev,
+		reg = devm_regulator_get(&pdev->dev,
 				config->pwr_reg_names[i]);
 		if (IS_ERR(reg)) {
 			ret = PTR_ERR(reg);
-			dev_err(dev->dev, "failed to get pwr regulator: %s (%d)\n",
+			dev_err(&pdev->dev, "failed to get pwr regulator: %s (%d)\n",
 					config->pwr_reg_names[i], ret);
 			goto fail;
 		}
@@ -158,7 +145,7 @@ struct hdmi *hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 		clk = devm_clk_get(&pdev->dev, config->hpd_clk_names[i]);
 		if (IS_ERR(clk)) {
 			ret = PTR_ERR(clk);
-			dev_err(dev->dev, "failed to get hpd clk: %s (%d)\n",
+			dev_err(&pdev->dev, "failed to get hpd clk: %s (%d)\n",
 					config->hpd_clk_names[i], ret);
 			goto fail;
 		}
@@ -173,7 +160,7 @@ struct hdmi *hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 		clk = devm_clk_get(&pdev->dev, config->pwr_clk_names[i]);
 		if (IS_ERR(clk)) {
 			ret = PTR_ERR(clk);
-			dev_err(dev->dev, "failed to get pwr clk: %s (%d)\n",
+			dev_err(&pdev->dev, "failed to get pwr clk: %s (%d)\n",
 					config->pwr_clk_names[i], ret);
 			goto fail;
 		}
@@ -184,11 +171,40 @@ struct hdmi *hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 	hdmi->i2c = hdmi_i2c_init(hdmi);
 	if (IS_ERR(hdmi->i2c)) {
 		ret = PTR_ERR(hdmi->i2c);
-		dev_err(dev->dev, "failed to get i2c: %d\n", ret);
+		dev_err(&pdev->dev, "failed to get i2c: %d\n", ret);
 		hdmi->i2c = NULL;
 		goto fail;
 	}
 
+	return hdmi;
+
+fail:
+	if (hdmi)
+		hdmi_destroy(hdmi);
+
+	return ERR_PTR(ret);
+}
+
+/* Second part of initialization, the drm/kms level modeset_init,
+ * constructs/initializes mode objects, etc, is called from master
+ * driver (not hdmi sub-device's probe/bind!)
+ *
+ * Any resource (regulator/clk/etc) which could be missing at boot
+ * should be handled in hdmi_init() so that failure happens from
+ * hdmi sub-device's probe.
+ */
+int hdmi_modeset_init(struct hdmi *hdmi,
+		struct drm_device *dev, struct drm_encoder *encoder)
+{
+	struct msm_drm_private *priv = dev->dev_private;
+	struct platform_device *pdev = hdmi->pdev;
+	int ret;
+
+	hdmi->dev = dev;
+	hdmi->encoder = encoder;
+
+	hdmi_audio_infoframe_init(&hdmi->audio.infoframe);
+
 	hdmi->bridge = hdmi_bridge_init(hdmi);
 	if (IS_ERR(hdmi->bridge)) {
 		ret = PTR_ERR(hdmi->bridge);
@@ -205,22 +221,20 @@ struct hdmi *hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 		goto fail;
 	}
 
-	if (!config->shared_irq) {
-		hdmi->irq = platform_get_irq(pdev, 0);
-		if (hdmi->irq < 0) {
-			ret = hdmi->irq;
-			dev_err(dev->dev, "failed to get irq: %d\n", ret);
-			goto fail;
-		}
+	hdmi->irq = irq_of_parse_and_map(pdev->dev.of_node, 0);
+	if (hdmi->irq < 0) {
+		ret = hdmi->irq;
+		dev_err(dev->dev, "failed to get irq: %d\n", ret);
+		goto fail;
+	}
 
-		ret = devm_request_threaded_irq(&pdev->dev, hdmi->irq,
-				NULL, hdmi_irq, IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
-				"hdmi_isr", hdmi);
-		if (ret < 0) {
-			dev_err(dev->dev, "failed to request IRQ%u: %d\n",
-					hdmi->irq, ret);
-			goto fail;
-		}
+	ret = devm_request_irq(&pdev->dev, hdmi->irq,
+			hdmi_irq, IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
+			"hdmi_isr", hdmi);
+	if (ret < 0) {
+		dev_err(dev->dev, "failed to request IRQ%u: %d\n",
+				hdmi->irq, ret);
+		goto fail;
 	}
 
 	encoder->bridge = hdmi->bridge;
@@ -230,19 +244,20 @@ struct hdmi *hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 
 	platform_set_drvdata(pdev, hdmi);
 
-	return hdmi;
+	return 0;
 
 fail:
-	if (hdmi) {
-		/* bridge/connector are normally destroyed by drm: */
-		if (hdmi->bridge)
-			hdmi->bridge->funcs->destroy(hdmi->bridge);
-		if (hdmi->connector)
-			hdmi->connector->funcs->destroy(hdmi->connector);
-		hdmi_destroy(&hdmi->refcount);
+	/* bridge/connector are normally destroyed by drm: */
+	if (hdmi->bridge) {
+		hdmi->bridge->funcs->destroy(hdmi->bridge);
+		hdmi->bridge = NULL;
+	}
+	if (hdmi->connector) {
+		hdmi->connector->funcs->destroy(hdmi->connector);
+		hdmi->connector = NULL;
 	}
 
-	return ERR_PTR(ret);
+	return ret;
 }
 
 /*
@@ -251,13 +266,6 @@ fail:
 
 #include <linux/of_gpio.h>
 
-static void set_hdmi_pdev(struct drm_device *dev,
-		struct platform_device *pdev)
-{
-	struct msm_drm_private *priv = dev->dev_private;
-	priv->hdmi_pdev = pdev;
-}
-
 #ifdef CONFIG_OF
 static int get_gpio(struct device *dev, struct device_node *of_node, const char *name)
 {
@@ -278,7 +286,10 @@ static int get_gpio(struct device *dev, struct device_node *of_node, const char
 
 static int hdmi_bind(struct device *dev, struct device *master, void *data)
 {
+	struct drm_device *drm = dev_get_drvdata(master);
+	struct msm_drm_private *priv = drm->dev_private;
 	static struct hdmi_platform_config config = {};
+	struct hdmi *hdmi;
 #ifdef CONFIG_OF
 	struct device_node *of_node = dev->of_node;
 
@@ -298,7 +309,6 @@ static int hdmi_bind(struct device *dev, struct device *master, void *data)
 		config.hpd_clk_cnt   = ARRAY_SIZE(hpd_clk_names);
 		config.pwr_clk_names = pwr_clk_names;
 		config.pwr_clk_cnt   = ARRAY_SIZE(pwr_clk_names);
-		config.shared_irq    = true;
 	} else if (of_device_is_compatible(of_node, "qcom,hdmi-tx-8960")) {
 		static const char *hpd_clk_names[] = {"core_clk", "master_iface_clk", "slave_iface_clk"};
 		static const char *hpd_reg_names[] = {"core-vdda", "hdmi-mux"};
@@ -369,14 +379,22 @@ static int hdmi_bind(struct device *dev, struct device *master, void *data)
 	}
 #endif
 	dev->platform_data = &config;
-	set_hdmi_pdev(dev_get_drvdata(master), to_platform_device(dev));
+	hdmi = hdmi_init(to_platform_device(dev));
+	if (IS_ERR(hdmi))
+		return PTR_ERR(hdmi);
+	priv->hdmi = hdmi;
 	return 0;
 }
 
 static void hdmi_unbind(struct device *dev, struct device *master,
 		void *data)
 {
-	set_hdmi_pdev(dev_get_drvdata(master), NULL);
+	struct drm_device *drm = dev_get_drvdata(master);
+	struct msm_drm_private *priv = drm->dev_private;
+	if (priv->hdmi) {
+		hdmi_destroy(priv->hdmi);
+		priv->hdmi = NULL;
+	}
 }
 
 static const struct component_ops hdmi_ops = {
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi.h b/drivers/gpu/drm/msm/hdmi/hdmi.h
index b981995410b5..43e654f751b7 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi.h
+++ b/drivers/gpu/drm/msm/hdmi/hdmi.h
@@ -38,8 +38,6 @@ struct hdmi_audio {
 };
 
 struct hdmi {
-	struct kref refcount;
-
 	struct drm_device *dev;
 	struct platform_device *pdev;
 
@@ -97,13 +95,9 @@ struct hdmi_platform_config {
 	/* gpio's: */
 	int ddc_clk_gpio, ddc_data_gpio, hpd_gpio, mux_en_gpio, mux_sel_gpio;
 	int mux_lpm_gpio;
-
-	/* older devices had their own irq, mdp5+ it is shared w/ mdp: */
-	bool shared_irq;
 };
 
 void hdmi_set_mode(struct hdmi *hdmi, bool power_on);
-void hdmi_destroy(struct kref *kref);
 
 static inline void hdmi_write(struct hdmi *hdmi, u32 reg, u32 data)
 {
@@ -115,17 +109,6 @@ static inline u32 hdmi_read(struct hdmi *hdmi, u32 reg)
 	return msm_readl(hdmi->mmio + reg);
 }
 
-static inline struct hdmi * hdmi_reference(struct hdmi *hdmi)
-{
-	kref_get(&hdmi->refcount);
-	return hdmi;
-}
-
-static inline void hdmi_unreference(struct hdmi *hdmi)
-{
-	kref_put(&hdmi->refcount, hdmi_destroy);
-}
-
 /*
  * The phy appears to be different, for example between 8960 and 8x60,
  * so split the phy related functions out and load the correct one at
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi.xml.h b/drivers/gpu/drm/msm/hdmi/hdmi.xml.h
index e89fe053d375..5b0844befbab 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi.xml.h
+++ b/drivers/gpu/drm/msm/hdmi/hdmi.xml.h
@@ -10,14 +10,14 @@ git clone https://github.com/freedreno/envytools.git
 The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-06-25 12:55:02)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  20136 bytes, from 2014-10-31 16:51:39)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1940 bytes, from 2014-10-31 16:51:39)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  23963 bytes, from 2014-10-31 16:51:46)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
-- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1686 bytes, from 2014-10-31 16:48:57)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-06-25 12:53:44)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-07-17 15:33:30)
 
 Copyright (C) 2013-2014 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi_bridge.c b/drivers/gpu/drm/msm/hdmi/hdmi_bridge.c
index f6cf745c249e..6902ad6da710 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi_bridge.c
+++ b/drivers/gpu/drm/msm/hdmi/hdmi_bridge.c
@@ -26,7 +26,6 @@ struct hdmi_bridge {
 static void hdmi_bridge_destroy(struct drm_bridge *bridge)
 {
 	struct hdmi_bridge *hdmi_bridge = to_hdmi_bridge(bridge);
-	hdmi_unreference(hdmi_bridge->hdmi);
 	drm_bridge_cleanup(bridge);
 	kfree(hdmi_bridge);
 }
@@ -218,7 +217,7 @@ struct drm_bridge *hdmi_bridge_init(struct hdmi *hdmi)
 		goto fail;
 	}
 
-	hdmi_bridge->hdmi = hdmi_reference(hdmi);
+	hdmi_bridge->hdmi = hdmi;
 
 	bridge = &hdmi_bridge->base;
 
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi_connector.c b/drivers/gpu/drm/msm/hdmi/hdmi_connector.c
index 4aca2a3c667c..b4e70e0e3cfa 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi_connector.c
+++ b/drivers/gpu/drm/msm/hdmi/hdmi_connector.c
@@ -141,6 +141,15 @@ static int hpd_enable(struct hdmi_connector *hdmi_connector)
 	uint32_t hpd_ctrl;
 	int i, ret;
 
+	for (i = 0; i < config->hpd_reg_cnt; i++) {
+		ret = regulator_enable(hdmi->hpd_regs[i]);
+		if (ret) {
+			dev_err(dev->dev, "failed to enable hpd regulator: %s (%d)\n",
+					config->hpd_reg_names[i], ret);
+			goto fail;
+		}
+	}
+
 	ret = gpio_config(hdmi, true);
 	if (ret) {
 		dev_err(dev->dev, "failed to configure GPIOs: %d\n", ret);
@@ -164,15 +173,6 @@ static int hpd_enable(struct hdmi_connector *hdmi_connector)
 		}
 	}
 
-	for (i = 0; i < config->hpd_reg_cnt; i++) {
-		ret = regulator_enable(hdmi->hpd_regs[i]);
-		if (ret) {
-			dev_err(dev->dev, "failed to enable hpd regulator: %s (%d)\n",
-					config->hpd_reg_names[i], ret);
-			goto fail;
-		}
-	}
-
 	hdmi_set_mode(hdmi, false);
 	phy->funcs->reset(phy);
 	hdmi_set_mode(hdmi, true);
@@ -200,7 +200,7 @@ fail:
 	return ret;
 }
 
-static int hdp_disable(struct hdmi_connector *hdmi_connector)
+static void hdp_disable(struct hdmi_connector *hdmi_connector)
 {
 	struct hdmi *hdmi = hdmi_connector->hdmi;
 	const struct hdmi_platform_config *config = hdmi->config;
@@ -212,28 +212,19 @@ static int hdp_disable(struct hdmi_connector *hdmi_connector)
 
 	hdmi_set_mode(hdmi, false);
 
-	for (i = 0; i < config->hpd_reg_cnt; i++) {
-		ret = regulator_disable(hdmi->hpd_regs[i]);
-		if (ret) {
-			dev_err(dev->dev, "failed to disable hpd regulator: %s (%d)\n",
-					config->hpd_reg_names[i], ret);
-			goto fail;
-		}
-	}
-
 	for (i = 0; i < config->hpd_clk_cnt; i++)
 		clk_disable_unprepare(hdmi->hpd_clks[i]);
 
 	ret = gpio_config(hdmi, false);
-	if (ret) {
-		dev_err(dev->dev, "failed to unconfigure GPIOs: %d\n", ret);
-		goto fail;
-	}
+	if (ret)
+		dev_warn(dev->dev, "failed to unconfigure GPIOs: %d\n", ret);
 
-	return 0;
-
-fail:
-	return ret;
+	for (i = 0; i < config->hpd_reg_cnt; i++) {
+		ret = regulator_disable(hdmi->hpd_regs[i]);
+		if (ret)
+			dev_warn(dev->dev, "failed to disable hpd regulator: %s (%d)\n",
+					config->hpd_reg_names[i], ret);
+	}
 }
 
 static void
@@ -260,11 +251,11 @@ void hdmi_connector_irq(struct drm_connector *connector)
 			(hpd_int_status & HDMI_HPD_INT_STATUS_INT)) {
 		bool detected = !!(hpd_int_status & HDMI_HPD_INT_STATUS_CABLE_DETECTED);
 
-		DBG("status=%04x, ctrl=%04x", hpd_int_status, hpd_int_ctrl);
-
-		/* ack the irq: */
+		/* ack & disable (temporarily) HPD events: */
 		hdmi_write(hdmi, REG_HDMI_HPD_INT_CTRL,
-				hpd_int_ctrl | HDMI_HPD_INT_CTRL_INT_ACK);
+			HDMI_HPD_INT_CTRL_INT_ACK);
+
+		DBG("status=%04x, ctrl=%04x", hpd_int_status, hpd_int_ctrl);
 
 		/* detect disconnect if we are connected or visa versa: */
 		hpd_int_ctrl = HDMI_HPD_INT_CTRL_INT_EN;
@@ -330,8 +321,6 @@ static void hdmi_connector_destroy(struct drm_connector *connector)
 	drm_connector_unregister(connector);
 	drm_connector_cleanup(connector);
 
-	hdmi_unreference(hdmi_connector->hdmi);
-
 	kfree(hdmi_connector);
 }
 
@@ -401,6 +390,9 @@ static const struct drm_connector_funcs hdmi_connector_funcs = {
 	.detect = hdmi_connector_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.destroy = hdmi_connector_destroy,
+	.reset = drm_atomic_helper_connector_reset,
+	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
 };
 
 static const struct drm_connector_helper_funcs hdmi_connector_helper_funcs = {
@@ -422,7 +414,7 @@ struct drm_connector *hdmi_connector_init(struct hdmi *hdmi)
 		goto fail;
 	}
 
-	hdmi_connector->hdmi = hdmi_reference(hdmi);
+	hdmi_connector->hdmi = hdmi;
 	INIT_WORK(&hdmi_connector->hpd_work, hotplug_work);
 
 	connector = &hdmi_connector->base;
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi_phy_8960.c b/drivers/gpu/drm/msm/hdmi/hdmi_phy_8960.c
index f408b69486a8..eeed006eed13 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi_phy_8960.c
+++ b/drivers/gpu/drm/msm/hdmi/hdmi_phy_8960.c
@@ -510,7 +510,7 @@ struct hdmi_phy *hdmi_phy_8960_init(struct hdmi *hdmi)
 
 #ifdef CONFIG_COMMON_CLK
 	phy_8960->pll_hw.init = &pll_init;
-	phy_8960->pll = devm_clk_register(hdmi->dev->dev, &phy_8960->pll_hw);
+	phy_8960->pll = devm_clk_register(&hdmi->pdev->dev, &phy_8960->pll_hw);
 	if (IS_ERR(phy_8960->pll)) {
 		ret = PTR_ERR(phy_8960->pll);
 		phy_8960->pll = NULL;
diff --git a/drivers/gpu/drm/msm/hdmi/qfprom.xml.h b/drivers/gpu/drm/msm/hdmi/qfprom.xml.h
index bd81db6a7829..29bd796797de 100644
--- a/drivers/gpu/drm/msm/hdmi/qfprom.xml.h
+++ b/drivers/gpu/drm/msm/hdmi/qfprom.xml.h
@@ -10,14 +10,14 @@ git clone https://github.com/freedreno/envytools.git
 The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-06-25 12:55:02)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  20136 bytes, from 2014-10-31 16:51:39)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1940 bytes, from 2014-10-31 16:51:39)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  23963 bytes, from 2014-10-31 16:51:46)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
-- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1686 bytes, from 2014-10-31 16:48:57)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-06-25 12:53:44)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-07-17 15:33:30)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4.xml.h b/drivers/gpu/drm/msm/mdp/mdp4/mdp4.xml.h
index 122208e8a2ee..a4a7f8c7122a 100644
--- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4.xml.h
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4.xml.h
@@ -10,16 +10,16 @@ git clone https://github.com/freedreno/envytools.git
 The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-06-25 12:55:02)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  20136 bytes, from 2014-10-31 16:51:39)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1940 bytes, from 2014-10-31 16:51:39)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  23963 bytes, from 2014-10-31 16:51:46)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
-- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1686 bytes, from 2014-10-31 16:48:57)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-06-25 12:53:44)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-07-17 15:33:30)
 
-Copyright (C) 2013 by the following authors:
+Copyright (C) 2013-2014 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
 
 Permission is hereby granted, free of charge, to any person obtaining
@@ -871,6 +871,101 @@ static inline uint32_t MDP4_LCDC_UNDERFLOW_CLR_COLOR(uint32_t val)
 #define MDP4_LCDC_CTRL_POLARITY_VSYNC_LOW			0x00000002
 #define MDP4_LCDC_CTRL_POLARITY_DATA_EN_LOW			0x00000004
 
+#define REG_MDP4_LCDC_LVDS_INTF_CTL				0x000c2000
+#define MDP4_LCDC_LVDS_INTF_CTL_MODE_SEL			0x00000004
+#define MDP4_LCDC_LVDS_INTF_CTL_RGB_OUT				0x00000008
+#define MDP4_LCDC_LVDS_INTF_CTL_CH_SWAP				0x00000010
+#define MDP4_LCDC_LVDS_INTF_CTL_CH1_RES_BIT			0x00000020
+#define MDP4_LCDC_LVDS_INTF_CTL_CH2_RES_BIT			0x00000040
+#define MDP4_LCDC_LVDS_INTF_CTL_ENABLE				0x00000080
+#define MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE0_EN		0x00000100
+#define MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE1_EN		0x00000200
+#define MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE2_EN		0x00000400
+#define MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE3_EN		0x00000800
+#define MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE0_EN		0x00001000
+#define MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE1_EN		0x00002000
+#define MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE2_EN		0x00004000
+#define MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE3_EN		0x00008000
+#define MDP4_LCDC_LVDS_INTF_CTL_CH1_CLK_LANE_EN			0x00010000
+#define MDP4_LCDC_LVDS_INTF_CTL_CH2_CLK_LANE_EN			0x00020000
+
+static inline uint32_t REG_MDP4_LCDC_LVDS_MUX_CTL(uint32_t i0) { return 0x000c2014 + 0x8*i0; }
+
+static inline uint32_t REG_MDP4_LCDC_LVDS_MUX_CTL_3_TO_0(uint32_t i0) { return 0x000c2014 + 0x8*i0; }
+#define MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0__MASK		0x000000ff
+#define MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0__SHIFT		0
+static inline uint32_t MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0(uint32_t val)
+{
+	return ((val) << MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0__SHIFT) & MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0__MASK;
+}
+#define MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1__MASK		0x0000ff00
+#define MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1__SHIFT		8
+static inline uint32_t MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1(uint32_t val)
+{
+	return ((val) << MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1__SHIFT) & MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1__MASK;
+}
+#define MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2__MASK		0x00ff0000
+#define MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2__SHIFT		16
+static inline uint32_t MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2(uint32_t val)
+{
+	return ((val) << MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2__SHIFT) & MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2__MASK;
+}
+#define MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3__MASK		0xff000000
+#define MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3__SHIFT		24
+static inline uint32_t MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3(uint32_t val)
+{
+	return ((val) << MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3__SHIFT) & MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3__MASK;
+}
+
+static inline uint32_t REG_MDP4_LCDC_LVDS_MUX_CTL_6_TO_4(uint32_t i0) { return 0x000c2018 + 0x8*i0; }
+#define MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4__MASK		0x000000ff
+#define MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4__SHIFT		0
+static inline uint32_t MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4(uint32_t val)
+{
+	return ((val) << MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4__SHIFT) & MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4__MASK;
+}
+#define MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5__MASK		0x0000ff00
+#define MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5__SHIFT		8
+static inline uint32_t MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5(uint32_t val)
+{
+	return ((val) << MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5__SHIFT) & MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5__MASK;
+}
+#define MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6__MASK		0x00ff0000
+#define MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6__SHIFT		16
+static inline uint32_t MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6(uint32_t val)
+{
+	return ((val) << MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6__SHIFT) & MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6__MASK;
+}
+
+#define REG_MDP4_LCDC_LVDS_PHY_RESET				0x000c2034
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_0				0x000c3000
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_1				0x000c3004
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_2				0x000c3008
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_3				0x000c300c
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_5				0x000c3014
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_6				0x000c3018
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_7				0x000c301c
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_8				0x000c3020
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_9				0x000c3024
+
+#define REG_MDP4_LVDS_PHY_PLL_LOCKED				0x000c3080
+
+#define REG_MDP4_LVDS_PHY_CFG2					0x000c3108
+
+#define REG_MDP4_LVDS_PHY_CFG0					0x000c3100
+#define MDP4_LVDS_PHY_CFG0_SERIALIZATION_ENBLE			0x00000010
+#define MDP4_LVDS_PHY_CFG0_CHANNEL0				0x00000040
+#define MDP4_LVDS_PHY_CFG0_CHANNEL1				0x00000080
+
 #define REG_MDP4_DTV						0x000d0000
 
 #define REG_MDP4_DTV_ENABLE					0x000d0000
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c
index c6c80ea28c35..3449213f1e76 100644
--- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c
@@ -25,8 +25,6 @@
 struct mdp4_crtc {
 	struct drm_crtc base;
 	char name[8];
-	struct drm_plane *plane;
-	struct drm_plane *planes[8];
 	int id;
 	int ovlp;
 	enum mdp4_dma dma;
@@ -52,25 +50,11 @@ struct mdp4_crtc {
 
 	/* if there is a pending flip, these will be non-null: */
 	struct drm_pending_vblank_event *event;
-	struct msm_fence_cb pageflip_cb;
 
 #define PENDING_CURSOR 0x1
 #define PENDING_FLIP   0x2
 	atomic_t pending;
 
-	/* the fb that we logically (from PoV of KMS API) hold a ref
-	 * to.  Which we may not yet be scanning out (we may still
-	 * be scanning out previous in case of page_flip while waiting
-	 * for gpu rendering to complete:
-	 */
-	struct drm_framebuffer *fb;
-
-	/* the fb that we currently hold a scanout ref to: */
-	struct drm_framebuffer *scanout_fb;
-
-	/* for unref'ing framebuffers after scanout completes: */
-	struct drm_flip_work unref_fb_work;
-
 	/* for unref'ing cursor bo's after scanout completes: */
 	struct drm_flip_work unref_cursor_work;
 
@@ -97,15 +81,14 @@ static void crtc_flush(struct drm_crtc *crtc)
 {
 	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
 	struct mdp4_kms *mdp4_kms = get_kms(crtc);
-	uint32_t i, flush = 0;
+	struct drm_plane *plane;
+	uint32_t flush = 0;
 
-	for (i = 0; i < ARRAY_SIZE(mdp4_crtc->planes); i++) {
-		struct drm_plane *plane = mdp4_crtc->planes[i];
-		if (plane) {
-			enum mdp4_pipe pipe_id = mdp4_plane_pipe(plane);
-			flush |= pipe2flush(pipe_id);
-		}
+	drm_atomic_crtc_for_each_plane(plane, crtc) {
+		enum mdp4_pipe pipe_id = mdp4_plane_pipe(plane);
+		flush |= pipe2flush(pipe_id);
 	}
+
 	flush |= ovlp2flush(mdp4_crtc->ovlp);
 
 	DBG("%s: flush=%08x", mdp4_crtc->name, flush);
@@ -113,47 +96,6 @@ static void crtc_flush(struct drm_crtc *crtc)
 	mdp4_write(mdp4_kms, REG_MDP4_OVERLAY_FLUSH, flush);
 }
 
-static void update_fb(struct drm_crtc *crtc, struct drm_framebuffer *new_fb)
-{
-	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
-	struct drm_framebuffer *old_fb = mdp4_crtc->fb;
-
-	/* grab reference to incoming scanout fb: */
-	drm_framebuffer_reference(new_fb);
-	mdp4_crtc->base.primary->fb = new_fb;
-	mdp4_crtc->fb = new_fb;
-
-	if (old_fb)
-		drm_flip_work_queue(&mdp4_crtc->unref_fb_work, old_fb);
-}
-
-/* unlike update_fb(), take a ref to the new scanout fb *before* updating
- * plane, then call this.  Needed to ensure we don't unref the buffer that
- * is actually still being scanned out.
- *
- * Note that this whole thing goes away with atomic.. since we can defer
- * calling into driver until rendering is done.
- */
-static void update_scanout(struct drm_crtc *crtc, struct drm_framebuffer *fb)
-{
-	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
-
-	/* flush updates, to make sure hw is updated to new scanout fb,
-	 * so that we can safely queue unref to current fb (ie. next
-	 * vblank we know hw is done w/ previous scanout_fb).
-	 */
-	crtc_flush(crtc);
-
-	if (mdp4_crtc->scanout_fb)
-		drm_flip_work_queue(&mdp4_crtc->unref_fb_work,
-				mdp4_crtc->scanout_fb);
-
-	mdp4_crtc->scanout_fb = fb;
-
-	/* enable vblank to complete flip: */
-	request_pending(crtc, PENDING_FLIP);
-}
-
 /* if file!=NULL, this is preclose potential cancel-flip path */
 static void complete_flip(struct drm_crtc *crtc, struct drm_file *file)
 {
@@ -171,38 +113,13 @@ static void complete_flip(struct drm_crtc *crtc, struct drm_file *file)
 		 */
 		if (!file || (event->base.file_priv == file)) {
 			mdp4_crtc->event = NULL;
+			DBG("%s: send event: %p", mdp4_crtc->name, event);
 			drm_send_vblank_event(dev, mdp4_crtc->id, event);
 		}
 	}
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 }
 
-static void pageflip_cb(struct msm_fence_cb *cb)
-{
-	struct mdp4_crtc *mdp4_crtc =
-		container_of(cb, struct mdp4_crtc, pageflip_cb);
-	struct drm_crtc *crtc = &mdp4_crtc->base;
-	struct drm_framebuffer *fb = crtc->primary->fb;
-
-	if (!fb)
-		return;
-
-	drm_framebuffer_reference(fb);
-	mdp4_plane_set_scanout(mdp4_crtc->plane, fb);
-	update_scanout(crtc, fb);
-}
-
-static void unref_fb_worker(struct drm_flip_work *work, void *val)
-{
-	struct mdp4_crtc *mdp4_crtc =
-		container_of(work, struct mdp4_crtc, unref_fb_work);
-	struct drm_device *dev = mdp4_crtc->base.dev;
-
-	mutex_lock(&dev->mode_config.mutex);
-	drm_framebuffer_unreference(val);
-	mutex_unlock(&dev->mode_config.mutex);
-}
-
 static void unref_cursor_worker(struct drm_flip_work *work, void *val)
 {
 	struct mdp4_crtc *mdp4_crtc =
@@ -218,7 +135,6 @@ static void mdp4_crtc_destroy(struct drm_crtc *crtc)
 	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
 
 	drm_crtc_cleanup(crtc);
-	drm_flip_work_cleanup(&mdp4_crtc->unref_fb_work);
 	drm_flip_work_cleanup(&mdp4_crtc->unref_cursor_work);
 
 	kfree(mdp4_crtc);
@@ -251,26 +167,53 @@ static bool mdp4_crtc_mode_fixup(struct drm_crtc *crtc,
 	return true;
 }
 
-static void blend_setup(struct drm_crtc *crtc)
+/* statically (for now) map planes to mixer stage (z-order): */
+static const int idxs[] = {
+		[VG1]  = 1,
+		[VG2]  = 2,
+		[RGB1] = 0,
+		[RGB2] = 0,
+		[RGB3] = 0,
+		[VG3]  = 3,
+		[VG4]  = 4,
+
+};
+
+/* setup mixer config, for which we need to consider all crtc's and
+ * the planes attached to them
+ *
+ * TODO may possibly need some extra locking here
+ */
+static void setup_mixer(struct mdp4_kms *mdp4_kms)
 {
-	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
-	struct mdp4_kms *mdp4_kms = get_kms(crtc);
-	int i, ovlp = mdp4_crtc->ovlp;
+	struct drm_mode_config *config = &mdp4_kms->dev->mode_config;
+	struct drm_crtc *crtc;
 	uint32_t mixer_cfg = 0;
 	static const enum mdp_mixer_stage_id stages[] = {
 			STAGE_BASE, STAGE0, STAGE1, STAGE2, STAGE3,
 	};
-	/* statically (for now) map planes to mixer stage (z-order): */
-	static const int idxs[] = {
-			[VG1]  = 1,
-			[VG2]  = 2,
-			[RGB1] = 0,
-			[RGB2] = 0,
-			[RGB3] = 0,
-			[VG3]  = 3,
-			[VG4]  = 4,
 
-	};
+	list_for_each_entry(crtc, &config->crtc_list, head) {
+		struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
+		struct drm_plane *plane;
+
+		drm_atomic_crtc_for_each_plane(plane, crtc) {
+			enum mdp4_pipe pipe_id = mdp4_plane_pipe(plane);
+			int idx = idxs[pipe_id];
+			mixer_cfg = mixercfg(mixer_cfg, mdp4_crtc->mixer,
+					pipe_id, stages[idx]);
+		}
+	}
+
+	mdp4_write(mdp4_kms, REG_MDP4_LAYERMIXER_IN_CFG, mixer_cfg);
+}
+
+static void blend_setup(struct drm_crtc *crtc)
+{
+	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
+	struct mdp4_kms *mdp4_kms = get_kms(crtc);
+	struct drm_plane *plane;
+	int i, ovlp = mdp4_crtc->ovlp;
 	bool alpha[4]= { false, false, false, false };
 
 	mdp4_write(mdp4_kms, REG_MDP4_OVLP_TRANSP_LOW0(ovlp), 0);
@@ -278,26 +221,16 @@ static void blend_setup(struct drm_crtc *crtc)
 	mdp4_write(mdp4_kms, REG_MDP4_OVLP_TRANSP_HIGH0(ovlp), 0);
 	mdp4_write(mdp4_kms, REG_MDP4_OVLP_TRANSP_HIGH1(ovlp), 0);
 
-	/* TODO single register for all CRTCs, so this won't work properly
-	 * when multiple CRTCs are active..
-	 */
-	for (i = 0; i < ARRAY_SIZE(mdp4_crtc->planes); i++) {
-		struct drm_plane *plane = mdp4_crtc->planes[i];
-		if (plane) {
-			enum mdp4_pipe pipe_id = mdp4_plane_pipe(plane);
-			int idx = idxs[pipe_id];
-			if (idx > 0) {
-				const struct mdp_format *format =
+	drm_atomic_crtc_for_each_plane(plane, crtc) {
+		enum mdp4_pipe pipe_id = mdp4_plane_pipe(plane);
+		int idx = idxs[pipe_id];
+		if (idx > 0) {
+			const struct mdp_format *format =
 					to_mdp_format(msm_framebuffer_format(plane->fb));
-				alpha[idx-1] = format->alpha_enable;
-			}
-			mixer_cfg |= mixercfg(mdp4_crtc->mixer, pipe_id, stages[idx]);
+			alpha[idx-1] = format->alpha_enable;
 		}
 	}
 
-	/* this shouldn't happen.. and seems to cause underflow: */
-	WARN_ON(!mixer_cfg);
-
 	for (i = 0; i < 4; i++) {
 		uint32_t op;
 
@@ -320,21 +253,21 @@ static void blend_setup(struct drm_crtc *crtc)
 		mdp4_write(mdp4_kms, REG_MDP4_OVLP_STAGE_TRANSP_HIGH1(ovlp, i), 0);
 	}
 
-	mdp4_write(mdp4_kms, REG_MDP4_LAYERMIXER_IN_CFG, mixer_cfg);
+	setup_mixer(mdp4_kms);
 }
 
-static int mdp4_crtc_mode_set(struct drm_crtc *crtc,
-		struct drm_display_mode *mode,
-		struct drm_display_mode *adjusted_mode,
-		int x, int y,
-		struct drm_framebuffer *old_fb)
+static void mdp4_crtc_mode_set_nofb(struct drm_crtc *crtc)
 {
 	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
 	struct mdp4_kms *mdp4_kms = get_kms(crtc);
 	enum mdp4_dma dma = mdp4_crtc->dma;
-	int ret, ovlp = mdp4_crtc->ovlp;
+	int ovlp = mdp4_crtc->ovlp;
+	struct drm_display_mode *mode;
+
+	if (WARN_ON(!crtc->state))
+		return;
 
-	mode = adjusted_mode;
+	mode = &crtc->state->adjusted_mode;
 
 	DBG("%s: set mode: %d:\"%s\" %d %d %d %d %d %d %d %d %d %d 0x%x 0x%x",
 			mdp4_crtc->name, mode->base.id, mode->name,
@@ -345,28 +278,13 @@ static int mdp4_crtc_mode_set(struct drm_crtc *crtc,
 			mode->vsync_end, mode->vtotal,
 			mode->type, mode->flags);
 
-	/* grab extra ref for update_scanout() */
-	drm_framebuffer_reference(crtc->primary->fb);
-
-	ret = mdp4_plane_mode_set(mdp4_crtc->plane, crtc, crtc->primary->fb,
-			0, 0, mode->hdisplay, mode->vdisplay,
-			x << 16, y << 16,
-			mode->hdisplay << 16, mode->vdisplay << 16);
-	if (ret) {
-		drm_framebuffer_unreference(crtc->primary->fb);
-		dev_err(crtc->dev->dev, "%s: failed to set mode on plane: %d\n",
-				mdp4_crtc->name, ret);
-		return ret;
-	}
-
 	mdp4_write(mdp4_kms, REG_MDP4_DMA_SRC_SIZE(dma),
 			MDP4_DMA_SRC_SIZE_WIDTH(mode->hdisplay) |
 			MDP4_DMA_SRC_SIZE_HEIGHT(mode->vdisplay));
 
 	/* take data from pipe: */
 	mdp4_write(mdp4_kms, REG_MDP4_DMA_SRC_BASE(dma), 0);
-	mdp4_write(mdp4_kms, REG_MDP4_DMA_SRC_STRIDE(dma),
-			crtc->primary->fb->pitches[0]);
+	mdp4_write(mdp4_kms, REG_MDP4_DMA_SRC_STRIDE(dma), 0);
 	mdp4_write(mdp4_kms, REG_MDP4_DMA_DST_SIZE(dma),
 			MDP4_DMA_DST_SIZE_WIDTH(0) |
 			MDP4_DMA_DST_SIZE_HEIGHT(0));
@@ -375,8 +293,7 @@ static int mdp4_crtc_mode_set(struct drm_crtc *crtc,
 	mdp4_write(mdp4_kms, REG_MDP4_OVLP_SIZE(ovlp),
 			MDP4_OVLP_SIZE_WIDTH(mode->hdisplay) |
 			MDP4_OVLP_SIZE_HEIGHT(mode->vdisplay));
-	mdp4_write(mdp4_kms, REG_MDP4_OVLP_STRIDE(ovlp),
-			crtc->primary->fb->pitches[0]);
+	mdp4_write(mdp4_kms, REG_MDP4_OVLP_STRIDE(ovlp), 0);
 
 	mdp4_write(mdp4_kms, REG_MDP4_OVLP_CFG(ovlp), 1);
 
@@ -385,11 +302,6 @@ static int mdp4_crtc_mode_set(struct drm_crtc *crtc,
 		mdp4_write(mdp4_kms, REG_MDP4_DMA_E_QUANT(1), 0x00ff0000);
 		mdp4_write(mdp4_kms, REG_MDP4_DMA_E_QUANT(2), 0x00ff0000);
 	}
-
-	update_fb(crtc, crtc->primary->fb);
-	update_scanout(crtc, crtc->primary->fb);
-
-	return 0;
 }
 
 static void mdp4_crtc_prepare(struct drm_crtc *crtc)
@@ -411,60 +323,42 @@ static void mdp4_crtc_commit(struct drm_crtc *crtc)
 	drm_crtc_vblank_put(crtc);
 }
 
-static int mdp4_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
-		struct drm_framebuffer *old_fb)
+static void mdp4_crtc_load_lut(struct drm_crtc *crtc)
 {
-	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
-	struct drm_plane *plane = mdp4_crtc->plane;
-	struct drm_display_mode *mode = &crtc->mode;
-	int ret;
-
-	/* grab extra ref for update_scanout() */
-	drm_framebuffer_reference(crtc->primary->fb);
-
-	ret = mdp4_plane_mode_set(plane, crtc, crtc->primary->fb,
-			0, 0, mode->hdisplay, mode->vdisplay,
-			x << 16, y << 16,
-			mode->hdisplay << 16, mode->vdisplay << 16);
-	if (ret) {
-		drm_framebuffer_unreference(crtc->primary->fb);
-		return ret;
-	}
-
-	update_fb(crtc, crtc->primary->fb);
-	update_scanout(crtc, crtc->primary->fb);
+}
 
+static int mdp4_crtc_atomic_check(struct drm_crtc *crtc,
+		struct drm_crtc_state *state)
+{
+	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
+	DBG("%s: check", mdp4_crtc->name);
+	// TODO anything else to check?
 	return 0;
 }
 
-static void mdp4_crtc_load_lut(struct drm_crtc *crtc)
+static void mdp4_crtc_atomic_begin(struct drm_crtc *crtc)
 {
+	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
+	DBG("%s: begin", mdp4_crtc->name);
 }
 
-static int mdp4_crtc_page_flip(struct drm_crtc *crtc,
-		struct drm_framebuffer *new_fb,
-		struct drm_pending_vblank_event *event,
-		uint32_t page_flip_flags)
+static void mdp4_crtc_atomic_flush(struct drm_crtc *crtc)
 {
 	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
-	struct drm_gem_object *obj;
 	unsigned long flags;
 
-	if (mdp4_crtc->event) {
-		dev_err(dev->dev, "already pending flip!\n");
-		return -EBUSY;
-	}
+	DBG("%s: event: %p", mdp4_crtc->name, crtc->state->event);
 
-	obj = msm_framebuffer_bo(new_fb, 0);
+	WARN_ON(mdp4_crtc->event);
 
 	spin_lock_irqsave(&dev->event_lock, flags);
-	mdp4_crtc->event = event;
+	mdp4_crtc->event = crtc->state->event;
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 
-	update_fb(crtc, new_fb);
-
-	return msm_gem_queue_inactive_cb(obj, &mdp4_crtc->pageflip_cb);
+	blend_setup(crtc);
+	crtc_flush(crtc);
+	request_pending(crtc, PENDING_FLIP);
 }
 
 static int mdp4_crtc_set_property(struct drm_crtc *crtc,
@@ -602,22 +496,29 @@ static int mdp4_crtc_cursor_move(struct drm_crtc *crtc, int x, int y)
 }
 
 static const struct drm_crtc_funcs mdp4_crtc_funcs = {
-	.set_config = drm_crtc_helper_set_config,
+	.set_config = drm_atomic_helper_set_config,
 	.destroy = mdp4_crtc_destroy,
-	.page_flip = mdp4_crtc_page_flip,
+	.page_flip = drm_atomic_helper_page_flip,
 	.set_property = mdp4_crtc_set_property,
 	.cursor_set = mdp4_crtc_cursor_set,
 	.cursor_move = mdp4_crtc_cursor_move,
+	.reset = drm_atomic_helper_crtc_reset,
+	.atomic_duplicate_state = drm_atomic_helper_crtc_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_crtc_destroy_state,
 };
 
 static const struct drm_crtc_helper_funcs mdp4_crtc_helper_funcs = {
 	.dpms = mdp4_crtc_dpms,
 	.mode_fixup = mdp4_crtc_mode_fixup,
-	.mode_set = mdp4_crtc_mode_set,
+	.mode_set_nofb = mdp4_crtc_mode_set_nofb,
+	.mode_set = drm_helper_crtc_mode_set,
+	.mode_set_base = drm_helper_crtc_mode_set_base,
 	.prepare = mdp4_crtc_prepare,
 	.commit = mdp4_crtc_commit,
-	.mode_set_base = mdp4_crtc_mode_set_base,
 	.load_lut = mdp4_crtc_load_lut,
+	.atomic_check = mdp4_crtc_atomic_check,
+	.atomic_begin = mdp4_crtc_atomic_begin,
+	.atomic_flush = mdp4_crtc_atomic_flush,
 };
 
 static void mdp4_crtc_vblank_irq(struct mdp_irq *irq, uint32_t irqstatus)
@@ -633,7 +534,6 @@ static void mdp4_crtc_vblank_irq(struct mdp_irq *irq, uint32_t irqstatus)
 
 	if (pending & PENDING_FLIP) {
 		complete_flip(crtc, NULL);
-		drm_flip_work_commit(&mdp4_crtc->unref_fb_work, priv->wq);
 	}
 
 	if (pending & PENDING_CURSOR) {
@@ -658,7 +558,8 @@ uint32_t mdp4_crtc_vblank(struct drm_crtc *crtc)
 
 void mdp4_crtc_cancel_pending_flip(struct drm_crtc *crtc, struct drm_file *file)
 {
-	DBG("cancel: %p", file);
+	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
+	DBG("%s: cancel: %p", mdp4_crtc->name, file);
 	complete_flip(crtc, file);
 }
 
@@ -672,7 +573,7 @@ void mdp4_crtc_set_config(struct drm_crtc *crtc, uint32_t config)
 }
 
 /* set interface for routing crtc->encoder: */
-void mdp4_crtc_set_intf(struct drm_crtc *crtc, enum mdp4_intf intf)
+void mdp4_crtc_set_intf(struct drm_crtc *crtc, enum mdp4_intf intf, int mixer)
 {
 	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
 	struct mdp4_kms *mdp4_kms = get_kms(crtc);
@@ -698,15 +599,13 @@ void mdp4_crtc_set_intf(struct drm_crtc *crtc, enum mdp4_intf intf)
 	if (intf == INTF_DSI_VIDEO) {
 		intf_sel &= ~MDP4_DISP_INTF_SEL_DSI_CMD;
 		intf_sel |= MDP4_DISP_INTF_SEL_DSI_VIDEO;
-		mdp4_crtc->mixer = 0;
 	} else if (intf == INTF_DSI_CMD) {
 		intf_sel &= ~MDP4_DISP_INTF_SEL_DSI_VIDEO;
 		intf_sel |= MDP4_DISP_INTF_SEL_DSI_CMD;
-		mdp4_crtc->mixer = 0;
-	} else if (intf == INTF_LCDC_DTV){
-		mdp4_crtc->mixer = 1;
 	}
 
+	mdp4_crtc->mixer = mixer;
+
 	blend_setup(crtc);
 
 	DBG("%s: intf_sel=%08x", mdp4_crtc->name, intf_sel);
@@ -714,35 +613,6 @@ void mdp4_crtc_set_intf(struct drm_crtc *crtc, enum mdp4_intf intf)
 	mdp4_write(mdp4_kms, REG_MDP4_DISP_INTF_SEL, intf_sel);
 }
 
-static void set_attach(struct drm_crtc *crtc, enum mdp4_pipe pipe_id,
-		struct drm_plane *plane)
-{
-	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
-
-	BUG_ON(pipe_id >= ARRAY_SIZE(mdp4_crtc->planes));
-
-	if (mdp4_crtc->planes[pipe_id] == plane)
-		return;
-
-	mdp4_crtc->planes[pipe_id] = plane;
-	blend_setup(crtc);
-	if (mdp4_crtc->enabled && (plane != mdp4_crtc->plane))
-		crtc_flush(crtc);
-}
-
-void mdp4_crtc_attach(struct drm_crtc *crtc, struct drm_plane *plane)
-{
-	set_attach(crtc, mdp4_plane_pipe(plane), plane);
-}
-
-void mdp4_crtc_detach(struct drm_crtc *crtc, struct drm_plane *plane)
-{
-	/* don't actually detatch our primary plane: */
-	if (to_mdp4_crtc(crtc)->plane == plane)
-		return;
-	set_attach(crtc, mdp4_plane_pipe(plane), NULL);
-}
-
 static const char *dma_names[] = {
 		"DMA_P", "DMA_S", "DMA_E",
 };
@@ -754,17 +624,13 @@ struct drm_crtc *mdp4_crtc_init(struct drm_device *dev,
 {
 	struct drm_crtc *crtc = NULL;
 	struct mdp4_crtc *mdp4_crtc;
-	int ret;
 
 	mdp4_crtc = kzalloc(sizeof(*mdp4_crtc), GFP_KERNEL);
-	if (!mdp4_crtc) {
-		ret = -ENOMEM;
-		goto fail;
-	}
+	if (!mdp4_crtc)
+		return ERR_PTR(-ENOMEM);
 
 	crtc = &mdp4_crtc->base;
 
-	mdp4_crtc->plane = plane;
 	mdp4_crtc->id = id;
 
 	mdp4_crtc->ovlp = ovlp_id;
@@ -781,26 +647,14 @@ struct drm_crtc *mdp4_crtc_init(struct drm_device *dev,
 
 	spin_lock_init(&mdp4_crtc->cursor.lock);
 
-	ret = drm_flip_work_init(&mdp4_crtc->unref_fb_work, 16,
-			"unref fb", unref_fb_worker);
-	if (ret)
-		goto fail;
-
-	ret = drm_flip_work_init(&mdp4_crtc->unref_cursor_work, 64,
+	drm_flip_work_init(&mdp4_crtc->unref_cursor_work,
 			"unref cursor", unref_cursor_worker);
 
-	INIT_FENCE_CB(&mdp4_crtc->pageflip_cb, pageflip_cb);
-
 	drm_crtc_init_with_planes(dev, crtc, plane, NULL, &mdp4_crtc_funcs);
 	drm_crtc_helper_add(crtc, &mdp4_crtc_helper_funcs);
+	plane->crtc = crtc;
 
-	mdp4_plane_install_properties(mdp4_crtc->plane, &crtc->base);
+	mdp4_plane_install_properties(plane, &crtc->base);
 
 	return crtc;
-
-fail:
-	if (crtc)
-		mdp4_crtc_destroy(crtc);
-
-	return ERR_PTR(ret);
 }
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c
index 067ed03b35fe..c3878420180b 100644
--- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c
@@ -233,7 +233,7 @@ static void mdp4_dtv_encoder_commit(struct drm_encoder *encoder)
 			MDP4_DMA_CONFIG_G_BPC(BPC8) |
 			MDP4_DMA_CONFIG_B_BPC(BPC8) |
 			MDP4_DMA_CONFIG_PACK(0x21));
-	mdp4_crtc_set_intf(encoder->crtc, INTF_LCDC_DTV);
+	mdp4_crtc_set_intf(encoder->crtc, INTF_LCDC_DTV, 1);
 	mdp4_dtv_encoder_dpms(encoder, DRM_MODE_DPMS_ON);
 }
 
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.c
index 733646c0d3f8..a62109e4ae0d 100644
--- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.c
@@ -106,6 +106,7 @@ static int mdp4_hw_init(struct msm_kms *kms)
 
 	if (mdp4_kms->rev >= 2)
 		mdp4_write(mdp4_kms, REG_MDP4_LAYERMIXER_IN_CFG_UPDATE_METHOD, 1);
+	mdp4_write(mdp4_kms, REG_MDP4_LAYERMIXER_IN_CFG, 0);
 
 	/* disable CSC matrix / YUV by default: */
 	mdp4_write(mdp4_kms, REG_MDP4_PIPE_OP_MODE(VG1), 0);
@@ -196,6 +197,28 @@ int mdp4_enable(struct mdp4_kms *mdp4_kms)
 	return 0;
 }
 
+#ifdef CONFIG_OF
+static struct drm_panel *detect_panel(struct drm_device *dev, const char *name)
+{
+	struct device_node *n;
+	struct drm_panel *panel = NULL;
+
+	n = of_parse_phandle(dev->dev->of_node, name, 0);
+	if (n) {
+		panel = of_drm_find_panel(n);
+		if (!panel)
+			panel = ERR_PTR(-EPROBE_DEFER);
+	}
+
+	return panel;
+}
+#else
+static struct drm_panel *detect_panel(struct drm_device *dev, const char *name)
+{
+	// ??? maybe use a module param to specify which panel is attached?
+}
+#endif
+
 static int modeset_init(struct mdp4_kms *mdp4_kms)
 {
 	struct drm_device *dev = mdp4_kms->dev;
@@ -203,14 +226,10 @@ static int modeset_init(struct mdp4_kms *mdp4_kms)
 	struct drm_plane *plane;
 	struct drm_crtc *crtc;
 	struct drm_encoder *encoder;
-	struct hdmi *hdmi;
+	struct drm_connector *connector;
+	struct drm_panel *panel;
 	int ret;
 
-	/*
-	 *  NOTE: this is a bit simplistic until we add support
-	 * for more than just RGB1->DMA_E->DTV->HDMI
-	 */
-
 	/* construct non-private planes: */
 	plane = mdp4_plane_init(dev, VG1, false);
 	if (IS_ERR(plane)) {
@@ -228,7 +247,57 @@ static int modeset_init(struct mdp4_kms *mdp4_kms)
 	}
 	priv->planes[priv->num_planes++] = plane;
 
-	/* the CRTCs get constructed with a private plane: */
+	/*
+	 * Setup the LCDC/LVDS path: RGB2 -> DMA_P -> LCDC -> LVDS:
+	 */
+
+	panel = detect_panel(dev, "qcom,lvds-panel");
+	if (IS_ERR(panel)) {
+		ret = PTR_ERR(panel);
+		dev_err(dev->dev, "failed to detect LVDS panel: %d\n", ret);
+		goto fail;
+	}
+
+	plane = mdp4_plane_init(dev, RGB2, true);
+	if (IS_ERR(plane)) {
+		dev_err(dev->dev, "failed to construct plane for RGB2\n");
+		ret = PTR_ERR(plane);
+		goto fail;
+	}
+
+	crtc  = mdp4_crtc_init(dev, plane, priv->num_crtcs, 0, DMA_P);
+	if (IS_ERR(crtc)) {
+		dev_err(dev->dev, "failed to construct crtc for DMA_P\n");
+		ret = PTR_ERR(crtc);
+		goto fail;
+	}
+
+	encoder = mdp4_lcdc_encoder_init(dev, panel);
+	if (IS_ERR(encoder)) {
+		dev_err(dev->dev, "failed to construct LCDC encoder\n");
+		ret = PTR_ERR(encoder);
+		goto fail;
+	}
+
+	/* LCDC can be hooked to DMA_P: */
+	encoder->possible_crtcs = 1 << priv->num_crtcs;
+
+	priv->crtcs[priv->num_crtcs++] = crtc;
+	priv->encoders[priv->num_encoders++] = encoder;
+
+	connector = mdp4_lvds_connector_init(dev, panel, encoder);
+	if (IS_ERR(connector)) {
+		ret = PTR_ERR(connector);
+		dev_err(dev->dev, "failed to initialize LVDS connector: %d\n", ret);
+		goto fail;
+	}
+
+	priv->connectors[priv->num_connectors++] = connector;
+
+	/*
+	 * Setup DTV/HDMI path: RGB1 -> DMA_E -> DTV -> HDMI:
+	 */
+
 	plane = mdp4_plane_init(dev, RGB1, true);
 	if (IS_ERR(plane)) {
 		dev_err(dev->dev, "failed to construct plane for RGB1\n");
@@ -242,7 +311,6 @@ static int modeset_init(struct mdp4_kms *mdp4_kms)
 		ret = PTR_ERR(crtc);
 		goto fail;
 	}
-	priv->crtcs[priv->num_crtcs++] = crtc;
 
 	encoder = mdp4_dtv_encoder_init(dev);
 	if (IS_ERR(encoder)) {
@@ -250,14 +318,20 @@ static int modeset_init(struct mdp4_kms *mdp4_kms)
 		ret = PTR_ERR(encoder);
 		goto fail;
 	}
-	encoder->possible_crtcs = 0x1;     /* DTV can be hooked to DMA_E */
+
+	/* DTV can be hooked to DMA_E: */
+	encoder->possible_crtcs = 1 << priv->num_crtcs;
+
+	priv->crtcs[priv->num_crtcs++] = crtc;
 	priv->encoders[priv->num_encoders++] = encoder;
 
-	hdmi = hdmi_init(dev, encoder);
-	if (IS_ERR(hdmi)) {
-		ret = PTR_ERR(hdmi);
-		dev_err(dev->dev, "failed to initialize HDMI: %d\n", ret);
-		goto fail;
+	if (priv->hdmi) {
+		/* Construct bridge/connector for HDMI: */
+		ret = hdmi_modeset_init(priv->hdmi, dev, encoder);
+		if (ret) {
+			dev_err(dev->dev, "failed to initialize HDMI: %d\n", ret);
+			goto fail;
+		}
 	}
 
 	return 0;
@@ -308,6 +382,10 @@ struct msm_kms *mdp4_kms_init(struct drm_device *dev)
 	if (IS_ERR(mdp4_kms->dsi_pll_vddio))
 		mdp4_kms->dsi_pll_vddio = NULL;
 
+	/* NOTE: driver for this regulator still missing upstream.. use
+	 * _get_exclusive() and ignore the error if it does not exist
+	 * (and hope that the bootloader left it on for us)
+	 */
 	mdp4_kms->vdd = devm_regulator_get_exclusive(&pdev->dev, "vdd");
 	if (IS_ERR(mdp4_kms->vdd))
 		mdp4_kms->vdd = NULL;
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h
index 3225da804c61..cbd77bc626d5 100644
--- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h
@@ -23,6 +23,8 @@
 #include "mdp/mdp_kms.h"
 #include "mdp4.xml.h"
 
+#include "drm_panel.h"
+
 struct mdp4_kms {
 	struct mdp_kms base;
 
@@ -74,7 +76,7 @@ static inline uint32_t pipe2flush(enum mdp4_pipe pipe)
 	case VG1:      return MDP4_OVERLAY_FLUSH_VG1;
 	case VG2:      return MDP4_OVERLAY_FLUSH_VG2;
 	case RGB1:     return MDP4_OVERLAY_FLUSH_RGB1;
-	case RGB2:     return MDP4_OVERLAY_FLUSH_RGB1;
+	case RGB2:     return MDP4_OVERLAY_FLUSH_RGB2;
 	default:       return 0;
 	}
 }
@@ -108,38 +110,50 @@ static inline uint32_t dma2err(enum mdp4_dma dma)
 	}
 }
 
-static inline uint32_t mixercfg(int mixer, enum mdp4_pipe pipe,
-		enum mdp_mixer_stage_id stage)
+static inline uint32_t mixercfg(uint32_t mixer_cfg, int mixer,
+		enum mdp4_pipe pipe, enum mdp_mixer_stage_id stage)
 {
-	uint32_t mixer_cfg = 0;
-
 	switch (pipe) {
 	case VG1:
-		mixer_cfg = MDP4_LAYERMIXER_IN_CFG_PIPE0(stage) |
+		mixer_cfg &= ~(MDP4_LAYERMIXER_IN_CFG_PIPE0__MASK |
+				MDP4_LAYERMIXER_IN_CFG_PIPE0_MIXER1);
+		mixer_cfg |= MDP4_LAYERMIXER_IN_CFG_PIPE0(stage) |
 			COND(mixer == 1, MDP4_LAYERMIXER_IN_CFG_PIPE0_MIXER1);
 		break;
 	case VG2:
-		mixer_cfg = MDP4_LAYERMIXER_IN_CFG_PIPE1(stage) |
+		mixer_cfg &= ~(MDP4_LAYERMIXER_IN_CFG_PIPE1__MASK |
+				MDP4_LAYERMIXER_IN_CFG_PIPE1_MIXER1);
+		mixer_cfg |= MDP4_LAYERMIXER_IN_CFG_PIPE1(stage) |
 			COND(mixer == 1, MDP4_LAYERMIXER_IN_CFG_PIPE1_MIXER1);
 		break;
 	case RGB1:
-		mixer_cfg = MDP4_LAYERMIXER_IN_CFG_PIPE2(stage) |
+		mixer_cfg &= ~(MDP4_LAYERMIXER_IN_CFG_PIPE2__MASK |
+				MDP4_LAYERMIXER_IN_CFG_PIPE2_MIXER1);
+		mixer_cfg |= MDP4_LAYERMIXER_IN_CFG_PIPE2(stage) |
 			COND(mixer == 1, MDP4_LAYERMIXER_IN_CFG_PIPE2_MIXER1);
 		break;
 	case RGB2:
-		mixer_cfg = MDP4_LAYERMIXER_IN_CFG_PIPE3(stage) |
+		mixer_cfg &= ~(MDP4_LAYERMIXER_IN_CFG_PIPE3__MASK |
+				MDP4_LAYERMIXER_IN_CFG_PIPE3_MIXER1);
+		mixer_cfg |= MDP4_LAYERMIXER_IN_CFG_PIPE3(stage) |
 			COND(mixer == 1, MDP4_LAYERMIXER_IN_CFG_PIPE3_MIXER1);
 		break;
 	case RGB3:
-		mixer_cfg = MDP4_LAYERMIXER_IN_CFG_PIPE4(stage) |
+		mixer_cfg &= ~(MDP4_LAYERMIXER_IN_CFG_PIPE4__MASK |
+				MDP4_LAYERMIXER_IN_CFG_PIPE4_MIXER1);
+		mixer_cfg |= MDP4_LAYERMIXER_IN_CFG_PIPE4(stage) |
 			COND(mixer == 1, MDP4_LAYERMIXER_IN_CFG_PIPE4_MIXER1);
 		break;
 	case VG3:
-		mixer_cfg = MDP4_LAYERMIXER_IN_CFG_PIPE5(stage) |
+		mixer_cfg &= ~(MDP4_LAYERMIXER_IN_CFG_PIPE5__MASK |
+				MDP4_LAYERMIXER_IN_CFG_PIPE5_MIXER1);
+		mixer_cfg |= MDP4_LAYERMIXER_IN_CFG_PIPE5(stage) |
 			COND(mixer == 1, MDP4_LAYERMIXER_IN_CFG_PIPE5_MIXER1);
 		break;
 	case VG4:
-		mixer_cfg = MDP4_LAYERMIXER_IN_CFG_PIPE6(stage) |
+		mixer_cfg &= ~(MDP4_LAYERMIXER_IN_CFG_PIPE6__MASK |
+				MDP4_LAYERMIXER_IN_CFG_PIPE6_MIXER1);
+		mixer_cfg |= MDP4_LAYERMIXER_IN_CFG_PIPE6(stage) |
 			COND(mixer == 1, MDP4_LAYERMIXER_IN_CFG_PIPE6_MIXER1);
 		break;
 	default:
@@ -173,14 +187,6 @@ uint32_t mdp4_get_formats(enum mdp4_pipe pipe_id, uint32_t *pixel_formats,
 
 void mdp4_plane_install_properties(struct drm_plane *plane,
 		struct drm_mode_object *obj);
-void mdp4_plane_set_scanout(struct drm_plane *plane,
-		struct drm_framebuffer *fb);
-int mdp4_plane_mode_set(struct drm_plane *plane,
-		struct drm_crtc *crtc, struct drm_framebuffer *fb,
-		int crtc_x, int crtc_y,
-		unsigned int crtc_w, unsigned int crtc_h,
-		uint32_t src_x, uint32_t src_y,
-		uint32_t src_w, uint32_t src_h);
 enum mdp4_pipe mdp4_plane_pipe(struct drm_plane *plane);
 struct drm_plane *mdp4_plane_init(struct drm_device *dev,
 		enum mdp4_pipe pipe_id, bool private_plane);
@@ -188,9 +194,7 @@ struct drm_plane *mdp4_plane_init(struct drm_device *dev,
 uint32_t mdp4_crtc_vblank(struct drm_crtc *crtc);
 void mdp4_crtc_cancel_pending_flip(struct drm_crtc *crtc, struct drm_file *file);
 void mdp4_crtc_set_config(struct drm_crtc *crtc, uint32_t config);
-void mdp4_crtc_set_intf(struct drm_crtc *crtc, enum mdp4_intf intf);
-void mdp4_crtc_attach(struct drm_crtc *crtc, struct drm_plane *plane);
-void mdp4_crtc_detach(struct drm_crtc *crtc, struct drm_plane *plane);
+void mdp4_crtc_set_intf(struct drm_crtc *crtc, enum mdp4_intf intf, int mixer);
 struct drm_crtc *mdp4_crtc_init(struct drm_device *dev,
 		struct drm_plane *plane, int id, int ovlp_id,
 		enum mdp4_dma dma_id);
@@ -198,6 +202,22 @@ struct drm_crtc *mdp4_crtc_init(struct drm_device *dev,
 long mdp4_dtv_round_pixclk(struct drm_encoder *encoder, unsigned long rate);
 struct drm_encoder *mdp4_dtv_encoder_init(struct drm_device *dev);
 
+long mdp4_lcdc_round_pixclk(struct drm_encoder *encoder, unsigned long rate);
+struct drm_encoder *mdp4_lcdc_encoder_init(struct drm_device *dev,
+		struct drm_panel *panel);
+
+struct drm_connector *mdp4_lvds_connector_init(struct drm_device *dev,
+		struct drm_panel *panel, struct drm_encoder *encoder);
+
+#ifdef CONFIG_COMMON_CLK
+struct clk *mpd4_lvds_pll_init(struct drm_device *dev);
+#else
+static inline struct clk *mpd4_lvds_pll_init(struct drm_device *dev)
+{
+	return ERR_PTR(-ENODEV);
+}
+#endif
+
 #ifdef CONFIG_MSM_BUS_SCALING
 static inline int match_dev_name(struct device *dev, void *data)
 {
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lcdc_encoder.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lcdc_encoder.c
new file mode 100644
index 000000000000..41f6436754fc
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lcdc_encoder.c
@@ -0,0 +1,506 @@
+/*
+ * Copyright (C) 2014 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ * Author: Vinay Simha <vinaysimha@inforcecomputing.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "mdp4_kms.h"
+
+#include "drm_crtc.h"
+#include "drm_crtc_helper.h"
+
+struct mdp4_lcdc_encoder {
+	struct drm_encoder base;
+	struct drm_panel *panel;
+	struct clk *lcdc_clk;
+	unsigned long int pixclock;
+	struct regulator *regs[3];
+	bool enabled;
+	uint32_t bsc;
+};
+#define to_mdp4_lcdc_encoder(x) container_of(x, struct mdp4_lcdc_encoder, base)
+
+static struct mdp4_kms *get_kms(struct drm_encoder *encoder)
+{
+	struct msm_drm_private *priv = encoder->dev->dev_private;
+	return to_mdp4_kms(to_mdp_kms(priv->kms));
+}
+
+#ifdef CONFIG_MSM_BUS_SCALING
+#include <mach/board.h>
+static void bs_init(struct mdp4_lcdc_encoder *mdp4_lcdc_encoder)
+{
+	struct drm_device *dev = mdp4_lcdc_encoder->base.dev;
+	struct lcdc_platform_data *lcdc_pdata = mdp4_find_pdata("lvds.0");
+
+	if (!lcdc_pdata) {
+		dev_err(dev->dev, "could not find lvds pdata\n");
+		return;
+	}
+
+	if (lcdc_pdata->bus_scale_table) {
+		mdp4_lcdc_encoder->bsc = msm_bus_scale_register_client(
+				lcdc_pdata->bus_scale_table);
+		DBG("lvds : bus scale client: %08x", mdp4_lcdc_encoder->bsc);
+	}
+}
+
+static void bs_fini(struct mdp4_lcdc_encoder *mdp4_lcdc_encoder)
+{
+	if (mdp4_lcdc_encoder->bsc) {
+		msm_bus_scale_unregister_client(mdp4_lcdc_encoder->bsc);
+		mdp4_lcdc_encoder->bsc = 0;
+	}
+}
+
+static void bs_set(struct mdp4_lcdc_encoder *mdp4_lcdc_encoder, int idx)
+{
+	if (mdp4_lcdc_encoder->bsc) {
+		DBG("set bus scaling: %d", idx);
+		msm_bus_scale_client_update_request(mdp4_lcdc_encoder->bsc, idx);
+	}
+}
+#else
+static void bs_init(struct mdp4_lcdc_encoder *mdp4_lcdc_encoder) {}
+static void bs_fini(struct mdp4_lcdc_encoder *mdp4_lcdc_encoder) {}
+static void bs_set(struct mdp4_lcdc_encoder *mdp4_lcdc_encoder, int idx) {}
+#endif
+
+static void mdp4_lcdc_encoder_destroy(struct drm_encoder *encoder)
+{
+	struct mdp4_lcdc_encoder *mdp4_lcdc_encoder =
+			to_mdp4_lcdc_encoder(encoder);
+	bs_fini(mdp4_lcdc_encoder);
+	drm_encoder_cleanup(encoder);
+	kfree(mdp4_lcdc_encoder);
+}
+
+static const struct drm_encoder_funcs mdp4_lcdc_encoder_funcs = {
+	.destroy = mdp4_lcdc_encoder_destroy,
+};
+
+/* this should probably be a helper: */
+struct drm_connector *get_connector(struct drm_encoder *encoder)
+{
+	struct drm_device *dev = encoder->dev;
+	struct drm_connector *connector;
+
+	list_for_each_entry(connector, &dev->mode_config.connector_list, head)
+		if (connector->encoder == encoder)
+			return connector;
+
+	return NULL;
+}
+
+static void setup_phy(struct drm_encoder *encoder)
+{
+	struct drm_device *dev = encoder->dev;
+	struct drm_connector *connector = get_connector(encoder);
+	struct mdp4_kms *mdp4_kms = get_kms(encoder);
+	uint32_t lvds_intf = 0, lvds_phy_cfg0 = 0;
+	int bpp, nchan, swap;
+
+	if (!connector)
+		return;
+
+	bpp = 3 * connector->display_info.bpc;
+
+	if (!bpp)
+		bpp = 18;
+
+	/* TODO, these should come from panel somehow: */
+	nchan = 1;
+	swap = 0;
+
+	switch (bpp) {
+	case 24:
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_3_TO_0(0),
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0(0x08) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1(0x05) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2(0x04) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3(0x03));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_6_TO_4(0),
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4(0x02) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5(0x01) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6(0x00));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_3_TO_0(1),
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0(0x11) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1(0x10) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2(0x0d) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3(0x0c));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_6_TO_4(1),
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4(0x0b) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5(0x0a) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6(0x09));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_3_TO_0(2),
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0(0x1a) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1(0x19) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2(0x18) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3(0x15));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_6_TO_4(2),
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4(0x14) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5(0x13) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6(0x12));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_3_TO_0(3),
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0(0x1b) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1(0x17) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2(0x16) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3(0x0f));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_6_TO_4(3),
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4(0x0e) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5(0x07) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6(0x06));
+		if (nchan == 2) {
+			lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE3_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE2_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE1_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE0_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE3_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE2_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE1_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE0_EN;
+		} else {
+			lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE3_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE2_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE1_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE0_EN;
+		}
+		break;
+
+	case 18:
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_3_TO_0(0),
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0(0x0a) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1(0x07) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2(0x06) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3(0x05));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_6_TO_4(0),
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4(0x04) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5(0x03) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6(0x02));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_3_TO_0(1),
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0(0x13) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1(0x12) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2(0x0f) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3(0x0e));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_6_TO_4(1),
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4(0x0d) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5(0x0c) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6(0x0b));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_3_TO_0(2),
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0(0x1a) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1(0x19) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2(0x18) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3(0x17));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_6_TO_4(2),
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4(0x16) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5(0x15) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6(0x14));
+		if (nchan == 2) {
+			lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE2_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE1_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE0_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE2_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE1_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE0_EN;
+		} else {
+			lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE2_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE1_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE0_EN;
+		}
+		lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_RGB_OUT;
+		break;
+
+	default:
+		dev_err(dev->dev, "unknown bpp: %d\n", bpp);
+		return;
+	}
+
+	switch (nchan) {
+	case 1:
+		lvds_phy_cfg0 = MDP4_LVDS_PHY_CFG0_CHANNEL0;
+		lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_CH1_CLK_LANE_EN |
+				MDP4_LCDC_LVDS_INTF_CTL_MODE_SEL;
+		break;
+	case 2:
+		lvds_phy_cfg0 = MDP4_LVDS_PHY_CFG0_CHANNEL0 |
+				MDP4_LVDS_PHY_CFG0_CHANNEL1;
+		lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_CH2_CLK_LANE_EN |
+				MDP4_LCDC_LVDS_INTF_CTL_CH1_CLK_LANE_EN;
+		break;
+	default:
+		dev_err(dev->dev, "unknown # of channels: %d\n", nchan);
+		return;
+	}
+
+	if (swap)
+		lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_CH_SWAP;
+
+	lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_ENABLE;
+
+	mdp4_write(mdp4_kms, REG_MDP4_LVDS_PHY_CFG0, lvds_phy_cfg0);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_INTF_CTL, lvds_intf);
+	mdp4_write(mdp4_kms, REG_MDP4_LVDS_PHY_CFG2, 0x30);
+
+	mb();
+	udelay(1);
+	lvds_phy_cfg0 |= MDP4_LVDS_PHY_CFG0_SERIALIZATION_ENBLE;
+	mdp4_write(mdp4_kms, REG_MDP4_LVDS_PHY_CFG0, lvds_phy_cfg0);
+}
+
+static void mdp4_lcdc_encoder_dpms(struct drm_encoder *encoder, int mode)
+{
+	struct drm_device *dev = encoder->dev;
+	struct mdp4_lcdc_encoder *mdp4_lcdc_encoder =
+			to_mdp4_lcdc_encoder(encoder);
+	struct mdp4_kms *mdp4_kms = get_kms(encoder);
+	struct drm_panel *panel = mdp4_lcdc_encoder->panel;
+	bool enabled = (mode == DRM_MODE_DPMS_ON);
+	int i, ret;
+
+	DBG("mode=%d", mode);
+
+	if (enabled == mdp4_lcdc_encoder->enabled)
+		return;
+
+	if (enabled) {
+		unsigned long pc = mdp4_lcdc_encoder->pixclock;
+		int ret;
+
+		bs_set(mdp4_lcdc_encoder, 1);
+
+		for (i = 0; i < ARRAY_SIZE(mdp4_lcdc_encoder->regs); i++) {
+			ret = regulator_enable(mdp4_lcdc_encoder->regs[i]);
+			if (ret)
+				dev_err(dev->dev, "failed to enable regulator: %d\n", ret);
+		}
+
+		DBG("setting lcdc_clk=%lu", pc);
+		ret = clk_set_rate(mdp4_lcdc_encoder->lcdc_clk, pc);
+		if (ret)
+			dev_err(dev->dev, "failed to configure lcdc_clk: %d\n", ret);
+		ret = clk_prepare_enable(mdp4_lcdc_encoder->lcdc_clk);
+		if (ret)
+			dev_err(dev->dev, "failed to enable lcdc_clk: %d\n", ret);
+
+		if (panel)
+			drm_panel_enable(panel);
+
+		setup_phy(encoder);
+
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_ENABLE, 1);
+	} else {
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_ENABLE, 0);
+
+		if (panel)
+			drm_panel_disable(panel);
+
+		/*
+		 * Wait for a vsync so we know the ENABLE=0 latched before
+		 * the (connector) source of the vsync's gets disabled,
+		 * otherwise we end up in a funny state if we re-enable
+		 * before the disable latches, which results that some of
+		 * the settings changes for the new modeset (like new
+		 * scanout buffer) don't latch properly..
+		 */
+		mdp_irq_wait(&mdp4_kms->base, MDP4_IRQ_PRIMARY_VSYNC);
+
+		clk_disable_unprepare(mdp4_lcdc_encoder->lcdc_clk);
+
+		for (i = 0; i < ARRAY_SIZE(mdp4_lcdc_encoder->regs); i++) {
+			ret = regulator_disable(mdp4_lcdc_encoder->regs[i]);
+			if (ret)
+				dev_err(dev->dev, "failed to disable regulator: %d\n", ret);
+		}
+
+		bs_set(mdp4_lcdc_encoder, 0);
+	}
+
+	mdp4_lcdc_encoder->enabled = enabled;
+}
+
+static bool mdp4_lcdc_encoder_mode_fixup(struct drm_encoder *encoder,
+		const struct drm_display_mode *mode,
+		struct drm_display_mode *adjusted_mode)
+{
+	return true;
+}
+
+static void mdp4_lcdc_encoder_mode_set(struct drm_encoder *encoder,
+		struct drm_display_mode *mode,
+		struct drm_display_mode *adjusted_mode)
+{
+	struct mdp4_lcdc_encoder *mdp4_lcdc_encoder =
+			to_mdp4_lcdc_encoder(encoder);
+	struct mdp4_kms *mdp4_kms = get_kms(encoder);
+	uint32_t lcdc_hsync_skew, vsync_period, vsync_len, ctrl_pol;
+	uint32_t display_v_start, display_v_end;
+	uint32_t hsync_start_x, hsync_end_x;
+
+	mode = adjusted_mode;
+
+	DBG("set mode: %d:\"%s\" %d %d %d %d %d %d %d %d %d %d 0x%x 0x%x",
+			mode->base.id, mode->name,
+			mode->vrefresh, mode->clock,
+			mode->hdisplay, mode->hsync_start,
+			mode->hsync_end, mode->htotal,
+			mode->vdisplay, mode->vsync_start,
+			mode->vsync_end, mode->vtotal,
+			mode->type, mode->flags);
+
+	mdp4_lcdc_encoder->pixclock = mode->clock * 1000;
+
+	DBG("pixclock=%lu", mdp4_lcdc_encoder->pixclock);
+
+	ctrl_pol = 0;
+	if (mode->flags & DRM_MODE_FLAG_NHSYNC)
+		ctrl_pol |= MDP4_LCDC_CTRL_POLARITY_HSYNC_LOW;
+	if (mode->flags & DRM_MODE_FLAG_NVSYNC)
+		ctrl_pol |= MDP4_LCDC_CTRL_POLARITY_VSYNC_LOW;
+	/* probably need to get DATA_EN polarity from panel.. */
+
+	lcdc_hsync_skew = 0;  /* get this from panel? */
+
+	hsync_start_x = (mode->htotal - mode->hsync_start);
+	hsync_end_x = mode->htotal - (mode->hsync_start - mode->hdisplay) - 1;
+
+	vsync_period = mode->vtotal * mode->htotal;
+	vsync_len = (mode->vsync_end - mode->vsync_start) * mode->htotal;
+	display_v_start = (mode->vtotal - mode->vsync_start) * mode->htotal + lcdc_hsync_skew;
+	display_v_end = vsync_period - ((mode->vsync_start - mode->vdisplay) * mode->htotal) + lcdc_hsync_skew - 1;
+
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_HSYNC_CTRL,
+			MDP4_LCDC_HSYNC_CTRL_PULSEW(mode->hsync_end - mode->hsync_start) |
+			MDP4_LCDC_HSYNC_CTRL_PERIOD(mode->htotal));
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_VSYNC_PERIOD, vsync_period);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_VSYNC_LEN, vsync_len);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_DISPLAY_HCTRL,
+			MDP4_LCDC_DISPLAY_HCTRL_START(hsync_start_x) |
+			MDP4_LCDC_DISPLAY_HCTRL_END(hsync_end_x));
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_DISPLAY_VSTART, display_v_start);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_DISPLAY_VEND, display_v_end);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_BORDER_CLR, 0);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_UNDERFLOW_CLR,
+			MDP4_LCDC_UNDERFLOW_CLR_ENABLE_RECOVERY |
+			MDP4_LCDC_UNDERFLOW_CLR_COLOR(0xff));
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_HSYNC_SKEW, lcdc_hsync_skew);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_CTRL_POLARITY, ctrl_pol);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_ACTIVE_HCTL,
+			MDP4_LCDC_ACTIVE_HCTL_START(0) |
+			MDP4_LCDC_ACTIVE_HCTL_END(0));
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_ACTIVE_VSTART, 0);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_ACTIVE_VEND, 0);
+}
+
+static void mdp4_lcdc_encoder_prepare(struct drm_encoder *encoder)
+{
+	mdp4_lcdc_encoder_dpms(encoder, DRM_MODE_DPMS_OFF);
+}
+
+static void mdp4_lcdc_encoder_commit(struct drm_encoder *encoder)
+{
+	/* TODO: hard-coded for 18bpp: */
+	mdp4_crtc_set_config(encoder->crtc,
+			MDP4_DMA_CONFIG_R_BPC(BPC6) |
+			MDP4_DMA_CONFIG_G_BPC(BPC6) |
+			MDP4_DMA_CONFIG_B_BPC(BPC6) |
+			MDP4_DMA_CONFIG_PACK_ALIGN_MSB |
+			MDP4_DMA_CONFIG_PACK(0x21) |
+			MDP4_DMA_CONFIG_DEFLKR_EN |
+			MDP4_DMA_CONFIG_DITHER_EN);
+	mdp4_crtc_set_intf(encoder->crtc, INTF_LCDC_DTV, 0);
+	mdp4_lcdc_encoder_dpms(encoder, DRM_MODE_DPMS_ON);
+}
+
+static const struct drm_encoder_helper_funcs mdp4_lcdc_encoder_helper_funcs = {
+	.dpms = mdp4_lcdc_encoder_dpms,
+	.mode_fixup = mdp4_lcdc_encoder_mode_fixup,
+	.mode_set = mdp4_lcdc_encoder_mode_set,
+	.prepare = mdp4_lcdc_encoder_prepare,
+	.commit = mdp4_lcdc_encoder_commit,
+};
+
+long mdp4_lcdc_round_pixclk(struct drm_encoder *encoder, unsigned long rate)
+{
+	struct mdp4_lcdc_encoder *mdp4_lcdc_encoder =
+			to_mdp4_lcdc_encoder(encoder);
+	return clk_round_rate(mdp4_lcdc_encoder->lcdc_clk, rate);
+}
+
+/* initialize encoder */
+struct drm_encoder *mdp4_lcdc_encoder_init(struct drm_device *dev,
+		struct drm_panel *panel)
+{
+	struct drm_encoder *encoder = NULL;
+	struct mdp4_lcdc_encoder *mdp4_lcdc_encoder;
+	struct regulator *reg;
+	int ret;
+
+	mdp4_lcdc_encoder = kzalloc(sizeof(*mdp4_lcdc_encoder), GFP_KERNEL);
+	if (!mdp4_lcdc_encoder) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	mdp4_lcdc_encoder->panel = panel;
+
+	encoder = &mdp4_lcdc_encoder->base;
+
+	drm_encoder_init(dev, encoder, &mdp4_lcdc_encoder_funcs,
+			 DRM_MODE_ENCODER_LVDS);
+	drm_encoder_helper_add(encoder, &mdp4_lcdc_encoder_helper_funcs);
+
+	/* TODO: do we need different pll in other cases? */
+	mdp4_lcdc_encoder->lcdc_clk = mpd4_lvds_pll_init(dev);
+	if (IS_ERR(mdp4_lcdc_encoder->lcdc_clk)) {
+		dev_err(dev->dev, "failed to get lvds_clk\n");
+		ret = PTR_ERR(mdp4_lcdc_encoder->lcdc_clk);
+		goto fail;
+	}
+
+	/* TODO: different regulators in other cases? */
+	reg = devm_regulator_get(dev->dev, "lvds-vccs-3p3v");
+	if (IS_ERR(reg)) {
+		ret = PTR_ERR(reg);
+		dev_err(dev->dev, "failed to get lvds-vccs-3p3v: %d\n", ret);
+		goto fail;
+	}
+	mdp4_lcdc_encoder->regs[0] = reg;
+
+	reg = devm_regulator_get(dev->dev, "lvds-pll-vdda");
+	if (IS_ERR(reg)) {
+		ret = PTR_ERR(reg);
+		dev_err(dev->dev, "failed to get lvds-pll-vdda: %d\n", ret);
+		goto fail;
+	}
+	mdp4_lcdc_encoder->regs[1] = reg;
+
+	reg = devm_regulator_get(dev->dev, "lvds-vdda");
+	if (IS_ERR(reg)) {
+		ret = PTR_ERR(reg);
+		dev_err(dev->dev, "failed to get lvds-vdda: %d\n", ret);
+		goto fail;
+	}
+	mdp4_lcdc_encoder->regs[2] = reg;
+
+	bs_init(mdp4_lcdc_encoder);
+
+	return encoder;
+
+fail:
+	if (encoder)
+		mdp4_lcdc_encoder_destroy(encoder);
+
+	return ERR_PTR(ret);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lvds_connector.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lvds_connector.c
new file mode 100644
index 000000000000..4ddc28e1275b
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lvds_connector.c
@@ -0,0 +1,154 @@
+/*
+ * Copyright (C) 2014 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ * Author: Vinay Simha <vinaysimha@inforcecomputing.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/gpio.h>
+
+#include "mdp4_kms.h"
+
+struct mdp4_lvds_connector {
+	struct drm_connector base;
+	struct drm_encoder *encoder;
+	struct drm_panel *panel;
+};
+#define to_mdp4_lvds_connector(x) container_of(x, struct mdp4_lvds_connector, base)
+
+static enum drm_connector_status mdp4_lvds_connector_detect(
+		struct drm_connector *connector, bool force)
+{
+	struct mdp4_lvds_connector *mdp4_lvds_connector =
+			to_mdp4_lvds_connector(connector);
+
+	return mdp4_lvds_connector->panel ?
+			connector_status_connected :
+			connector_status_disconnected;
+}
+
+static void mdp4_lvds_connector_destroy(struct drm_connector *connector)
+{
+	struct mdp4_lvds_connector *mdp4_lvds_connector =
+			to_mdp4_lvds_connector(connector);
+	struct drm_panel *panel = mdp4_lvds_connector->panel;
+
+	if (panel)
+		drm_panel_detach(panel);
+
+	drm_connector_unregister(connector);
+	drm_connector_cleanup(connector);
+
+	kfree(mdp4_lvds_connector);
+}
+
+static int mdp4_lvds_connector_get_modes(struct drm_connector *connector)
+{
+	struct mdp4_lvds_connector *mdp4_lvds_connector =
+			to_mdp4_lvds_connector(connector);
+	struct drm_panel *panel = mdp4_lvds_connector->panel;
+	int ret = 0;
+
+	if (panel)
+		ret = panel->funcs->get_modes(panel);
+
+	return ret;
+}
+
+static int mdp4_lvds_connector_mode_valid(struct drm_connector *connector,
+				 struct drm_display_mode *mode)
+{
+	struct mdp4_lvds_connector *mdp4_lvds_connector =
+			to_mdp4_lvds_connector(connector);
+	struct drm_encoder *encoder = mdp4_lvds_connector->encoder;
+	long actual, requested;
+
+	requested = 1000 * mode->clock;
+	actual = mdp4_lcdc_round_pixclk(encoder, requested);
+
+	DBG("requested=%ld, actual=%ld", requested, actual);
+
+	if (actual != requested)
+		return MODE_CLOCK_RANGE;
+
+	return MODE_OK;
+}
+
+static struct drm_encoder *
+mdp4_lvds_connector_best_encoder(struct drm_connector *connector)
+{
+	struct mdp4_lvds_connector *mdp4_lvds_connector =
+			to_mdp4_lvds_connector(connector);
+	return mdp4_lvds_connector->encoder;
+}
+
+static const struct drm_connector_funcs mdp4_lvds_connector_funcs = {
+	.dpms = drm_helper_connector_dpms,
+	.detect = mdp4_lvds_connector_detect,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.destroy = mdp4_lvds_connector_destroy,
+	.reset = drm_atomic_helper_connector_reset,
+	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
+};
+
+static const struct drm_connector_helper_funcs mdp4_lvds_connector_helper_funcs = {
+	.get_modes = mdp4_lvds_connector_get_modes,
+	.mode_valid = mdp4_lvds_connector_mode_valid,
+	.best_encoder = mdp4_lvds_connector_best_encoder,
+};
+
+/* initialize connector */
+struct drm_connector *mdp4_lvds_connector_init(struct drm_device *dev,
+		struct drm_panel *panel, struct drm_encoder *encoder)
+{
+	struct drm_connector *connector = NULL;
+	struct mdp4_lvds_connector *mdp4_lvds_connector;
+	int ret;
+
+	mdp4_lvds_connector = kzalloc(sizeof(*mdp4_lvds_connector), GFP_KERNEL);
+	if (!mdp4_lvds_connector) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	mdp4_lvds_connector->encoder = encoder;
+	mdp4_lvds_connector->panel = panel;
+
+	connector = &mdp4_lvds_connector->base;
+
+	drm_connector_init(dev, connector, &mdp4_lvds_connector_funcs,
+			DRM_MODE_CONNECTOR_LVDS);
+	drm_connector_helper_add(connector, &mdp4_lvds_connector_helper_funcs);
+
+	connector->polled = 0;
+
+	connector->interlace_allowed = 0;
+	connector->doublescan_allowed = 0;
+
+	drm_connector_register(connector);
+
+	drm_mode_connector_attach_encoder(connector, encoder);
+
+	if (panel)
+		drm_panel_attach(panel, connector);
+
+	return connector;
+
+fail:
+	if (connector)
+		mdp4_lvds_connector_destroy(connector);
+
+	return ERR_PTR(ret);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lvds_pll.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lvds_pll.c
new file mode 100644
index 000000000000..ce4245971673
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lvds_pll.c
@@ -0,0 +1,172 @@
+/*
+ * Copyright (C) 2014 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/clk.h>
+#include <linux/clk-provider.h>
+
+#include "mdp4_kms.h"
+
+struct mdp4_lvds_pll {
+	struct clk_hw pll_hw;
+	struct drm_device *dev;
+	unsigned long pixclk;
+};
+#define to_mdp4_lvds_pll(x) container_of(x, struct mdp4_lvds_pll, pll_hw)
+
+static struct mdp4_kms *get_kms(struct mdp4_lvds_pll *lvds_pll)
+{
+	struct msm_drm_private *priv = lvds_pll->dev->dev_private;
+	return to_mdp4_kms(to_mdp_kms(priv->kms));
+}
+
+struct pll_rate {
+	unsigned long rate;
+	struct {
+		uint32_t val;
+		uint32_t reg;
+	} conf[32];
+};
+
+/* NOTE: keep sorted highest freq to lowest: */
+static const struct pll_rate freqtbl[] = {
+	{ 72000000, {
+		{ 0x8f, REG_MDP4_LVDS_PHY_PLL_CTRL_1 },
+		{ 0x30, REG_MDP4_LVDS_PHY_PLL_CTRL_2 },
+		{ 0xc6, REG_MDP4_LVDS_PHY_PLL_CTRL_3 },
+		{ 0x10, REG_MDP4_LVDS_PHY_PLL_CTRL_5 },
+		{ 0x07, REG_MDP4_LVDS_PHY_PLL_CTRL_6 },
+		{ 0x62, REG_MDP4_LVDS_PHY_PLL_CTRL_7 },
+		{ 0x41, REG_MDP4_LVDS_PHY_PLL_CTRL_8 },
+		{ 0x0d, REG_MDP4_LVDS_PHY_PLL_CTRL_9 },
+		{ 0, 0 } }
+	},
+};
+
+static const struct pll_rate *find_rate(unsigned long rate)
+{
+	int i;
+	for (i = 1; i < ARRAY_SIZE(freqtbl); i++)
+		if (rate > freqtbl[i].rate)
+			return &freqtbl[i-1];
+	return &freqtbl[i-1];
+}
+
+static int mpd4_lvds_pll_enable(struct clk_hw *hw)
+{
+	struct mdp4_lvds_pll *lvds_pll = to_mdp4_lvds_pll(hw);
+	struct mdp4_kms *mdp4_kms = get_kms(lvds_pll);
+	const struct pll_rate *pll_rate = find_rate(lvds_pll->pixclk);
+	int i;
+
+	DBG("pixclk=%lu (%lu)", lvds_pll->pixclk, pll_rate->rate);
+
+	if (WARN_ON(!pll_rate))
+		return -EINVAL;
+
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_PHY_RESET, 0x33);
+
+	for (i = 0; pll_rate->conf[i].reg; i++)
+		mdp4_write(mdp4_kms, pll_rate->conf[i].reg, pll_rate->conf[i].val);
+
+	mdp4_write(mdp4_kms, REG_MDP4_LVDS_PHY_PLL_CTRL_0, 0x01);
+
+	/* Wait until LVDS PLL is locked and ready */
+	while (!mdp4_read(mdp4_kms, REG_MDP4_LVDS_PHY_PLL_LOCKED))
+		cpu_relax();
+
+	return 0;
+}
+
+static void mpd4_lvds_pll_disable(struct clk_hw *hw)
+{
+	struct mdp4_lvds_pll *lvds_pll = to_mdp4_lvds_pll(hw);
+	struct mdp4_kms *mdp4_kms = get_kms(lvds_pll);
+
+	DBG("");
+
+	mdp4_write(mdp4_kms, REG_MDP4_LVDS_PHY_CFG0, 0x0);
+	mdp4_write(mdp4_kms, REG_MDP4_LVDS_PHY_PLL_CTRL_0, 0x0);
+}
+
+static unsigned long mpd4_lvds_pll_recalc_rate(struct clk_hw *hw,
+				unsigned long parent_rate)
+{
+	struct mdp4_lvds_pll *lvds_pll = to_mdp4_lvds_pll(hw);
+	return lvds_pll->pixclk;
+}
+
+static long mpd4_lvds_pll_round_rate(struct clk_hw *hw, unsigned long rate,
+		unsigned long *parent_rate)
+{
+	const struct pll_rate *pll_rate = find_rate(rate);
+	return pll_rate->rate;
+}
+
+static int mpd4_lvds_pll_set_rate(struct clk_hw *hw, unsigned long rate,
+		unsigned long parent_rate)
+{
+	struct mdp4_lvds_pll *lvds_pll = to_mdp4_lvds_pll(hw);
+	lvds_pll->pixclk = rate;
+	return 0;
+}
+
+
+static const struct clk_ops mpd4_lvds_pll_ops = {
+	.enable = mpd4_lvds_pll_enable,
+	.disable = mpd4_lvds_pll_disable,
+	.recalc_rate = mpd4_lvds_pll_recalc_rate,
+	.round_rate = mpd4_lvds_pll_round_rate,
+	.set_rate = mpd4_lvds_pll_set_rate,
+};
+
+static const char *mpd4_lvds_pll_parents[] = {
+	"pxo",
+};
+
+static struct clk_init_data pll_init = {
+	.name = "mpd4_lvds_pll",
+	.ops = &mpd4_lvds_pll_ops,
+	.parent_names = mpd4_lvds_pll_parents,
+	.num_parents = ARRAY_SIZE(mpd4_lvds_pll_parents),
+};
+
+struct clk *mpd4_lvds_pll_init(struct drm_device *dev)
+{
+	struct mdp4_lvds_pll *lvds_pll;
+	struct clk *clk;
+	int ret;
+
+	lvds_pll = devm_kzalloc(dev->dev, sizeof(*lvds_pll), GFP_KERNEL);
+	if (!lvds_pll) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	lvds_pll->dev = dev;
+
+	lvds_pll->pll_hw.init = &pll_init;
+	clk = devm_clk_register(dev->dev, &lvds_pll->pll_hw);
+	if (IS_ERR(clk)) {
+		ret = PTR_ERR(clk);
+		goto fail;
+	}
+
+	return clk;
+
+fail:
+	return ERR_PTR(ret);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_plane.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_plane.c
index 66f33dba1ebb..1e5ebe83647d 100644
--- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_plane.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_plane.c
@@ -31,47 +31,26 @@ struct mdp4_plane {
 };
 #define to_mdp4_plane(x) container_of(x, struct mdp4_plane, base)
 
-static struct mdp4_kms *get_kms(struct drm_plane *plane)
-{
-	struct msm_drm_private *priv = plane->dev->dev_private;
-	return to_mdp4_kms(to_mdp_kms(priv->kms));
-}
-
-static int mdp4_plane_update(struct drm_plane *plane,
+static void mdp4_plane_set_scanout(struct drm_plane *plane,
+		struct drm_framebuffer *fb);
+static int mdp4_plane_mode_set(struct drm_plane *plane,
 		struct drm_crtc *crtc, struct drm_framebuffer *fb,
 		int crtc_x, int crtc_y,
 		unsigned int crtc_w, unsigned int crtc_h,
 		uint32_t src_x, uint32_t src_y,
-		uint32_t src_w, uint32_t src_h)
-{
-	struct mdp4_plane *mdp4_plane = to_mdp4_plane(plane);
-
-	mdp4_plane->enabled = true;
-
-	if (plane->fb)
-		drm_framebuffer_unreference(plane->fb);
-
-	drm_framebuffer_reference(fb);
-
-	return mdp4_plane_mode_set(plane, crtc, fb,
-			crtc_x, crtc_y, crtc_w, crtc_h,
-			src_x, src_y, src_w, src_h);
-}
+		uint32_t src_w, uint32_t src_h);
 
-static int mdp4_plane_disable(struct drm_plane *plane)
+static struct mdp4_kms *get_kms(struct drm_plane *plane)
 {
-	struct mdp4_plane *mdp4_plane = to_mdp4_plane(plane);
-	DBG("%s: disable", mdp4_plane->name);
-	if (plane->crtc)
-		mdp4_crtc_detach(plane->crtc, plane);
-	return 0;
+	struct msm_drm_private *priv = plane->dev->dev_private;
+	return to_mdp4_kms(to_mdp_kms(priv->kms));
 }
 
 static void mdp4_plane_destroy(struct drm_plane *plane)
 {
 	struct mdp4_plane *mdp4_plane = to_mdp4_plane(plane);
 
-	mdp4_plane_disable(plane);
+	drm_plane_helper_disable(plane);
 	drm_plane_cleanup(plane);
 
 	kfree(mdp4_plane);
@@ -92,19 +71,75 @@ int mdp4_plane_set_property(struct drm_plane *plane,
 }
 
 static const struct drm_plane_funcs mdp4_plane_funcs = {
-		.update_plane = mdp4_plane_update,
-		.disable_plane = mdp4_plane_disable,
+		.update_plane = drm_atomic_helper_update_plane,
+		.disable_plane = drm_atomic_helper_disable_plane,
 		.destroy = mdp4_plane_destroy,
 		.set_property = mdp4_plane_set_property,
+		.reset = drm_atomic_helper_plane_reset,
+		.atomic_duplicate_state = drm_atomic_helper_plane_duplicate_state,
+		.atomic_destroy_state = drm_atomic_helper_plane_destroy_state,
 };
 
-void mdp4_plane_set_scanout(struct drm_plane *plane,
+static int mdp4_plane_prepare_fb(struct drm_plane *plane,
+		struct drm_framebuffer *fb)
+{
+	struct mdp4_plane *mdp4_plane = to_mdp4_plane(plane);
+	struct mdp4_kms *mdp4_kms = get_kms(plane);
+
+	DBG("%s: prepare: FB[%u]", mdp4_plane->name, fb->base.id);
+	return msm_framebuffer_prepare(fb, mdp4_kms->id);
+}
+
+static void mdp4_plane_cleanup_fb(struct drm_plane *plane,
+		struct drm_framebuffer *fb)
+{
+	struct mdp4_plane *mdp4_plane = to_mdp4_plane(plane);
+	struct mdp4_kms *mdp4_kms = get_kms(plane);
+
+	DBG("%s: cleanup: FB[%u]", mdp4_plane->name, fb->base.id);
+	msm_framebuffer_cleanup(fb, mdp4_kms->id);
+}
+
+
+static int mdp4_plane_atomic_check(struct drm_plane *plane,
+		struct drm_plane_state *state)
+{
+	return 0;
+}
+
+static void mdp4_plane_atomic_update(struct drm_plane *plane,
+				     struct drm_plane_state *old_state)
+{
+	struct drm_plane_state *state = plane->state;
+	int ret;
+
+	ret = mdp4_plane_mode_set(plane,
+			state->crtc, state->fb,
+			state->crtc_x, state->crtc_y,
+			state->crtc_w, state->crtc_h,
+			state->src_x,  state->src_y,
+			state->src_w, state->src_h);
+	/* atomic_check should have ensured that this doesn't fail */
+	WARN_ON(ret < 0);
+}
+
+static const struct drm_plane_helper_funcs mdp4_plane_helper_funcs = {
+		.prepare_fb = mdp4_plane_prepare_fb,
+		.cleanup_fb = mdp4_plane_cleanup_fb,
+		.atomic_check = mdp4_plane_atomic_check,
+		.atomic_update = mdp4_plane_atomic_update,
+};
+
+static void mdp4_plane_set_scanout(struct drm_plane *plane,
 		struct drm_framebuffer *fb)
 {
 	struct mdp4_plane *mdp4_plane = to_mdp4_plane(plane);
 	struct mdp4_kms *mdp4_kms = get_kms(plane);
 	enum mdp4_pipe pipe = mdp4_plane->pipe;
-	uint32_t iova;
+	uint32_t iova = msm_framebuffer_iova(fb, mdp4_kms->id, 0);
+
+	DBG("%s: set_scanout: %08x (%u)", mdp4_plane->name,
+			iova, fb->pitches[0]);
 
 	mdp4_write(mdp4_kms, REG_MDP4_PIPE_SRC_STRIDE_A(pipe),
 			MDP4_PIPE_SRC_STRIDE_A_P0(fb->pitches[0]) |
@@ -114,7 +149,6 @@ void mdp4_plane_set_scanout(struct drm_plane *plane,
 			MDP4_PIPE_SRC_STRIDE_B_P2(fb->pitches[2]) |
 			MDP4_PIPE_SRC_STRIDE_B_P3(fb->pitches[3]));
 
-	msm_gem_get_iova(msm_framebuffer_bo(fb, 0), mdp4_kms->id, &iova);
 	mdp4_write(mdp4_kms, REG_MDP4_PIPE_SRCP0_BASE(pipe), iova);
 
 	plane->fb = fb;
@@ -122,7 +156,7 @@ void mdp4_plane_set_scanout(struct drm_plane *plane,
 
 #define MDP4_VG_PHASE_STEP_DEFAULT	0x20000000
 
-int mdp4_plane_mode_set(struct drm_plane *plane,
+static int mdp4_plane_mode_set(struct drm_plane *plane,
 		struct drm_crtc *crtc, struct drm_framebuffer *fb,
 		int crtc_x, int crtc_y,
 		unsigned int crtc_w, unsigned int crtc_h,
@@ -137,6 +171,11 @@ int mdp4_plane_mode_set(struct drm_plane *plane,
 	uint32_t phasex_step = MDP4_VG_PHASE_STEP_DEFAULT;
 	uint32_t phasey_step = MDP4_VG_PHASE_STEP_DEFAULT;
 
+	if (!(crtc && fb)) {
+		DBG("%s: disabled!", mdp4_plane->name);
+		return 0;
+	}
+
 	/* src values are in Q16 fixed point, convert to integer: */
 	src_x = src_x >> 16;
 	src_y = src_y >> 16;
@@ -197,9 +236,6 @@ int mdp4_plane_mode_set(struct drm_plane *plane,
 	mdp4_write(mdp4_kms, REG_MDP4_PIPE_PHASEX_STEP(pipe), phasex_step);
 	mdp4_write(mdp4_kms, REG_MDP4_PIPE_PHASEY_STEP(pipe), phasey_step);
 
-	/* TODO detach from old crtc (if we had more than one) */
-	mdp4_crtc_attach(crtc, plane);
-
 	return 0;
 }
 
@@ -239,9 +275,12 @@ struct drm_plane *mdp4_plane_init(struct drm_device *dev,
 			ARRAY_SIZE(mdp4_plane->formats));
 
 	type = private_plane ? DRM_PLANE_TYPE_PRIMARY : DRM_PLANE_TYPE_OVERLAY;
-	drm_universal_plane_init(dev, plane, 0xff, &mdp4_plane_funcs,
-				 mdp4_plane->formats, mdp4_plane->nformats,
-				 type);
+	ret = drm_universal_plane_init(dev, plane, 0xff, &mdp4_plane_funcs,
+				 mdp4_plane->formats, mdp4_plane->nformats, type);
+	if (ret)
+		goto fail;
+
+	drm_plane_helper_add(plane, &mdp4_plane_helper_funcs);
 
 	mdp4_plane_install_properties(plane, &plane->base);
 
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5.xml.h b/drivers/gpu/drm/msm/mdp/mdp5/mdp5.xml.h
index 67f4f896ba8c..e87ef5512cb0 100644
--- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5.xml.h
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5.xml.h
@@ -10,14 +10,14 @@ git clone https://github.com/freedreno/envytools.git
 The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-06-25 12:55:02)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  20136 bytes, from 2014-10-31 16:51:39)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1940 bytes, from 2014-10-31 16:51:39)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  23963 bytes, from 2014-10-31 16:51:46)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
-- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1686 bytes, from 2014-10-31 16:48:57)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-06-25 12:53:44)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-07-17 15:33:30)
 
 Copyright (C) 2013-2014 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_cfg.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_cfg.c
new file mode 100644
index 000000000000..b0a44310cf2a
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_cfg.c
@@ -0,0 +1,207 @@
+/*
+ * Copyright (c) 2014 The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include "mdp5_kms.h"
+#include "mdp5_cfg.h"
+
+struct mdp5_cfg_handler {
+	int revision;
+	struct mdp5_cfg config;
+};
+
+/* mdp5_cfg must be exposed (used in mdp5.xml.h) */
+const struct mdp5_cfg_hw *mdp5_cfg = NULL;
+
+const struct mdp5_cfg_hw msm8x74_config = {
+	.name = "msm8x74",
+	.smp = {
+		.mmb_count = 22,
+		.mmb_size = 4096,
+	},
+	.ctl = {
+		.count = 5,
+		.base = { 0x00600, 0x00700, 0x00800, 0x00900, 0x00a00 },
+	},
+	.pipe_vig = {
+		.count = 3,
+		.base = { 0x01200, 0x01600, 0x01a00 },
+	},
+	.pipe_rgb = {
+		.count = 3,
+		.base = { 0x01e00, 0x02200, 0x02600 },
+	},
+	.pipe_dma = {
+		.count = 2,
+		.base = { 0x02a00, 0x02e00 },
+	},
+	.lm = {
+		.count = 5,
+		.base = { 0x03200, 0x03600, 0x03a00, 0x03e00, 0x04200 },
+		.nb_stages = 5,
+	},
+	.dspp = {
+		.count = 3,
+		.base = { 0x04600, 0x04a00, 0x04e00 },
+	},
+	.ad = {
+		.count = 2,
+		.base = { 0x13100, 0x13300 }, /* NOTE: no ad in v1.0 */
+	},
+	.intf = {
+		.count = 4,
+		.base = { 0x12500, 0x12700, 0x12900, 0x12b00 },
+	},
+	.max_clk = 200000000,
+};
+
+const struct mdp5_cfg_hw apq8084_config = {
+	.name = "apq8084",
+	.smp = {
+		.mmb_count = 44,
+		.mmb_size = 8192,
+		.reserved_state[0] = GENMASK(7, 0),	/* first 8 MMBs */
+		.reserved[CID_RGB0] = 2,
+		.reserved[CID_RGB1] = 2,
+		.reserved[CID_RGB2] = 2,
+		.reserved[CID_RGB3] = 2,
+	},
+	.ctl = {
+		.count = 5,
+		.base = { 0x00600, 0x00700, 0x00800, 0x00900, 0x00a00 },
+	},
+	.pipe_vig = {
+		.count = 4,
+		.base = { 0x01200, 0x01600, 0x01a00, 0x01e00 },
+	},
+	.pipe_rgb = {
+		.count = 4,
+		.base = { 0x02200, 0x02600, 0x02a00, 0x02e00 },
+	},
+	.pipe_dma = {
+		.count = 2,
+		.base = { 0x03200, 0x03600 },
+	},
+	.lm = {
+		.count = 6,
+		.base = { 0x03a00, 0x03e00, 0x04200, 0x04600, 0x04a00, 0x04e00 },
+		.nb_stages = 5,
+	},
+	.dspp = {
+		.count = 4,
+		.base = { 0x05200, 0x05600, 0x05a00, 0x05e00 },
+
+	},
+	.ad = {
+		.count = 3,
+		.base = { 0x13500, 0x13700, 0x13900 },
+	},
+	.intf = {
+		.count = 5,
+		.base = { 0x12500, 0x12700, 0x12900, 0x12b00, 0x12d00 },
+	},
+	.max_clk = 320000000,
+};
+
+static const struct mdp5_cfg_handler cfg_handlers[] = {
+	{ .revision = 0, .config = { .hw = &msm8x74_config } },
+	{ .revision = 2, .config = { .hw = &msm8x74_config } },
+	{ .revision = 3, .config = { .hw = &apq8084_config } },
+};
+
+
+static struct mdp5_cfg_platform *mdp5_get_config(struct platform_device *dev);
+
+const struct mdp5_cfg_hw *mdp5_cfg_get_hw_config(struct mdp5_cfg_handler *cfg_handler)
+{
+	return cfg_handler->config.hw;
+}
+
+struct mdp5_cfg *mdp5_cfg_get_config(struct mdp5_cfg_handler *cfg_handler)
+{
+	return &cfg_handler->config;
+}
+
+int mdp5_cfg_get_hw_rev(struct mdp5_cfg_handler *cfg_handler)
+{
+	return cfg_handler->revision;
+}
+
+void mdp5_cfg_destroy(struct mdp5_cfg_handler *cfg_handler)
+{
+	kfree(cfg_handler);
+}
+
+struct mdp5_cfg_handler *mdp5_cfg_init(struct mdp5_kms *mdp5_kms,
+		uint32_t major, uint32_t minor)
+{
+	struct drm_device *dev = mdp5_kms->dev;
+	struct platform_device *pdev = dev->platformdev;
+	struct mdp5_cfg_handler *cfg_handler;
+	struct mdp5_cfg_platform *pconfig;
+	int i, ret = 0;
+
+	cfg_handler = kzalloc(sizeof(*cfg_handler), GFP_KERNEL);
+	if (unlikely(!cfg_handler)) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	if (major != 1) {
+		dev_err(dev->dev, "unexpected MDP major version: v%d.%d\n",
+				major, minor);
+		ret = -ENXIO;
+		goto fail;
+	}
+
+	/* only after mdp5_cfg global pointer's init can we access the hw */
+	for (i = 0; i < ARRAY_SIZE(cfg_handlers); i++) {
+		if (cfg_handlers[i].revision != minor)
+			continue;
+		mdp5_cfg = cfg_handlers[i].config.hw;
+
+		break;
+	}
+	if (unlikely(!mdp5_cfg)) {
+		dev_err(dev->dev, "unexpected MDP minor revision: v%d.%d\n",
+				major, minor);
+		ret = -ENXIO;
+		goto fail;
+	}
+
+	cfg_handler->revision = minor;
+	cfg_handler->config.hw = mdp5_cfg;
+
+	pconfig = mdp5_get_config(pdev);
+	memcpy(&cfg_handler->config.platform, pconfig, sizeof(*pconfig));
+
+	DBG("MDP5: %s hw config selected", mdp5_cfg->name);
+
+	return cfg_handler;
+
+fail:
+	if (cfg_handler)
+		mdp5_cfg_destroy(cfg_handler);
+
+	return NULL;
+}
+
+static struct mdp5_cfg_platform *mdp5_get_config(struct platform_device *dev)
+{
+	static struct mdp5_cfg_platform config = {};
+#ifdef CONFIG_OF
+	/* TODO */
+#endif
+	config.iommu = iommu_domain_alloc(&platform_bus_type);
+
+	return &config;
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_cfg.h b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_cfg.h
new file mode 100644
index 000000000000..dba4d52cceeb
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_cfg.h
@@ -0,0 +1,91 @@
+/*
+ * Copyright (c) 2014 The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#ifndef __MDP5_CFG_H__
+#define __MDP5_CFG_H__
+
+#include "msm_drv.h"
+
+/*
+ * mdp5_cfg
+ *
+ * This module configures the dynamic offsets used by mdp5.xml.h
+ * (initialized in mdp5_cfg.c)
+ */
+extern const struct mdp5_cfg_hw *mdp5_cfg;
+
+#define MAX_CTL			8
+#define MAX_BASES		8
+#define MAX_SMP_BLOCKS		44
+#define MAX_CLIENTS		32
+
+typedef DECLARE_BITMAP(mdp5_smp_state_t, MAX_SMP_BLOCKS);
+
+#define MDP5_SUB_BLOCK_DEFINITION \
+	int count; \
+	uint32_t base[MAX_BASES]
+
+struct mdp5_sub_block {
+	MDP5_SUB_BLOCK_DEFINITION;
+};
+
+struct mdp5_lm_block {
+	MDP5_SUB_BLOCK_DEFINITION;
+	uint32_t nb_stages;		/* number of stages per blender */
+};
+
+struct mdp5_smp_block {
+	int mmb_count;			/* number of SMP MMBs */
+	int mmb_size;			/* MMB: size in bytes */
+	mdp5_smp_state_t reserved_state;/* SMP MMBs statically allocated */
+	int reserved[MAX_CLIENTS];	/* # of MMBs allocated per client */
+};
+
+struct mdp5_cfg_hw {
+	char  *name;
+
+	struct mdp5_smp_block smp;
+	struct mdp5_sub_block ctl;
+	struct mdp5_sub_block pipe_vig;
+	struct mdp5_sub_block pipe_rgb;
+	struct mdp5_sub_block pipe_dma;
+	struct mdp5_lm_block  lm;
+	struct mdp5_sub_block dspp;
+	struct mdp5_sub_block ad;
+	struct mdp5_sub_block intf;
+
+	uint32_t max_clk;
+};
+
+/* platform config data (ie. from DT, or pdata) */
+struct mdp5_cfg_platform {
+	struct iommu_domain *iommu;
+};
+
+struct mdp5_cfg {
+	const struct mdp5_cfg_hw *hw;
+	struct mdp5_cfg_platform platform;
+};
+
+struct mdp5_kms;
+struct mdp5_cfg_handler;
+
+const struct mdp5_cfg_hw *mdp5_cfg_get_hw_config(struct mdp5_cfg_handler *cfg_hnd);
+struct mdp5_cfg *mdp5_cfg_get_config(struct mdp5_cfg_handler *cfg_hnd);
+int mdp5_cfg_get_hw_rev(struct mdp5_cfg_handler *cfg_hnd);
+
+struct mdp5_cfg_handler *mdp5_cfg_init(struct mdp5_kms *mdp5_kms,
+		uint32_t major, uint32_t minor);
+void mdp5_cfg_destroy(struct mdp5_cfg_handler *cfg_hnd);
+
+#endif /* __MDP5_CFG_H__ */
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c
index ebe2e60f3ab1..f021f960a8a2 100644
--- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c
@@ -1,4 +1,5 @@
 /*
+ * Copyright (c) 2014 The Linux Foundation. All rights reserved.
  * Copyright (C) 2013 Red Hat
  * Author: Rob Clark <robdclark@gmail.com>
  *
@@ -17,43 +18,35 @@
 
 #include "mdp5_kms.h"
 
+#include <linux/sort.h>
 #include <drm/drm_mode.h>
 #include "drm_crtc.h"
 #include "drm_crtc_helper.h"
 #include "drm_flip_work.h"
 
+#define SSPP_MAX	(SSPP_RGB3 + 1) /* TODO: Add SSPP_MAX in mdp5.xml.h */
+
 struct mdp5_crtc {
 	struct drm_crtc base;
 	char name[8];
-	struct drm_plane *plane;
-	struct drm_plane *planes[8];
 	int id;
 	bool enabled;
 
-	/* which mixer/encoder we route output to: */
-	int mixer;
+	/* layer mixer used for this CRTC (+ its lock): */
+#define GET_LM_ID(crtc_id)	((crtc_id == 3) ? 5 : crtc_id)
+	int lm;
+	spinlock_t lm_lock;	/* protect REG_MDP5_LM_* registers */
+
+	/* CTL used for this CRTC: */
+	struct mdp5_ctl *ctl;
 
 	/* if there is a pending flip, these will be non-null: */
 	struct drm_pending_vblank_event *event;
-	struct msm_fence_cb pageflip_cb;
 
 #define PENDING_CURSOR 0x1
 #define PENDING_FLIP   0x2
 	atomic_t pending;
 
-	/* the fb that we logically (from PoV of KMS API) hold a ref
-	 * to.  Which we may not yet be scanning out (we may still
-	 * be scanning out previous in case of page_flip while waiting
-	 * for gpu rendering to complete:
-	 */
-	struct drm_framebuffer *fb;
-
-	/* the fb that we currently hold a scanout ref to: */
-	struct drm_framebuffer *scanout_fb;
-
-	/* for unref'ing framebuffers after scanout completes: */
-	struct drm_flip_work unref_fb_work;
-
 	struct mdp_irq vblank;
 	struct mdp_irq err;
 };
@@ -73,67 +66,38 @@ static void request_pending(struct drm_crtc *crtc, uint32_t pending)
 	mdp_irq_register(&get_kms(crtc)->base, &mdp5_crtc->vblank);
 }
 
-static void crtc_flush(struct drm_crtc *crtc)
-{
-	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
-	struct mdp5_kms *mdp5_kms = get_kms(crtc);
-	int id = mdp5_crtc->id;
-	uint32_t i, flush = 0;
-
-	for (i = 0; i < ARRAY_SIZE(mdp5_crtc->planes); i++) {
-		struct drm_plane *plane = mdp5_crtc->planes[i];
-		if (plane) {
-			enum mdp5_pipe pipe = mdp5_plane_pipe(plane);
-			flush |= pipe2flush(pipe);
-		}
-	}
-	flush |= mixer2flush(mdp5_crtc->id);
-	flush |= MDP5_CTL_FLUSH_CTL;
-
-	DBG("%s: flush=%08x", mdp5_crtc->name, flush);
-
-	mdp5_write(mdp5_kms, REG_MDP5_CTL_FLUSH(id), flush);
-}
+#define mdp5_lm_get_flush(lm)	mdp_ctl_flush_mask_lm(lm)
 
-static void update_fb(struct drm_crtc *crtc, struct drm_framebuffer *new_fb)
+static void crtc_flush(struct drm_crtc *crtc, u32 flush_mask)
 {
 	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
-	struct drm_framebuffer *old_fb = mdp5_crtc->fb;
 
-	/* grab reference to incoming scanout fb: */
-	drm_framebuffer_reference(new_fb);
-	mdp5_crtc->base.primary->fb = new_fb;
-	mdp5_crtc->fb = new_fb;
-
-	if (old_fb)
-		drm_flip_work_queue(&mdp5_crtc->unref_fb_work, old_fb);
+	DBG("%s: flush=%08x", mdp5_crtc->name, flush_mask);
+	mdp5_ctl_commit(mdp5_crtc->ctl, flush_mask);
 }
 
-/* unlike update_fb(), take a ref to the new scanout fb *before* updating
- * plane, then call this.  Needed to ensure we don't unref the buffer that
- * is actually still being scanned out.
- *
- * Note that this whole thing goes away with atomic.. since we can defer
- * calling into driver until rendering is done.
+/*
+ * flush updates, to make sure hw is updated to new scanout fb,
+ * so that we can safely queue unref to current fb (ie. next
+ * vblank we know hw is done w/ previous scanout_fb).
  */
-static void update_scanout(struct drm_crtc *crtc, struct drm_framebuffer *fb)
+static void crtc_flush_all(struct drm_crtc *crtc)
 {
 	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct drm_plane *plane;
+	uint32_t flush_mask = 0;
 
-	/* flush updates, to make sure hw is updated to new scanout fb,
-	 * so that we can safely queue unref to current fb (ie. next
-	 * vblank we know hw is done w/ previous scanout_fb).
-	 */
-	crtc_flush(crtc);
-
-	if (mdp5_crtc->scanout_fb)
-		drm_flip_work_queue(&mdp5_crtc->unref_fb_work,
-				mdp5_crtc->scanout_fb);
+	/* we could have already released CTL in the disable path: */
+	if (!mdp5_crtc->ctl)
+		return;
 
-	mdp5_crtc->scanout_fb = fb;
+	drm_atomic_crtc_for_each_plane(plane, crtc) {
+		flush_mask |= mdp5_plane_get_flush(plane);
+	}
+	flush_mask |= mdp5_ctl_get_flush(mdp5_crtc->ctl);
+	flush_mask |= mdp5_lm_get_flush(mdp5_crtc->lm);
 
-	/* enable vblank to complete flip: */
-	request_pending(crtc, PENDING_FLIP);
+	crtc_flush(crtc, flush_mask);
 }
 
 /* if file!=NULL, this is preclose potential cancel-flip path */
@@ -142,7 +106,8 @@ static void complete_flip(struct drm_crtc *crtc, struct drm_file *file)
 	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct drm_pending_vblank_event *event;
-	unsigned long flags, i;
+	struct drm_plane *plane;
+	unsigned long flags;
 
 	spin_lock_irqsave(&dev->event_lock, flags);
 	event = mdp5_crtc->event;
@@ -153,50 +118,22 @@ static void complete_flip(struct drm_crtc *crtc, struct drm_file *file)
 		 */
 		if (!file || (event->base.file_priv == file)) {
 			mdp5_crtc->event = NULL;
+			DBG("%s: send event: %p", mdp5_crtc->name, event);
 			drm_send_vblank_event(dev, mdp5_crtc->id, event);
 		}
 	}
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 
-	for (i = 0; i < ARRAY_SIZE(mdp5_crtc->planes); i++) {
-		struct drm_plane *plane = mdp5_crtc->planes[i];
-		if (plane)
-			mdp5_plane_complete_flip(plane);
+	drm_atomic_crtc_for_each_plane(plane, crtc) {
+		mdp5_plane_complete_flip(plane);
 	}
 }
 
-static void pageflip_cb(struct msm_fence_cb *cb)
-{
-	struct mdp5_crtc *mdp5_crtc =
-		container_of(cb, struct mdp5_crtc, pageflip_cb);
-	struct drm_crtc *crtc = &mdp5_crtc->base;
-	struct drm_framebuffer *fb = mdp5_crtc->fb;
-
-	if (!fb)
-		return;
-
-	drm_framebuffer_reference(fb);
-	mdp5_plane_set_scanout(mdp5_crtc->plane, fb);
-	update_scanout(crtc, fb);
-}
-
-static void unref_fb_worker(struct drm_flip_work *work, void *val)
-{
-	struct mdp5_crtc *mdp5_crtc =
-		container_of(work, struct mdp5_crtc, unref_fb_work);
-	struct drm_device *dev = mdp5_crtc->base.dev;
-
-	mutex_lock(&dev->mode_config.mutex);
-	drm_framebuffer_unreference(val);
-	mutex_unlock(&dev->mode_config.mutex);
-}
-
 static void mdp5_crtc_destroy(struct drm_crtc *crtc)
 {
 	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
 
 	drm_crtc_cleanup(crtc);
-	drm_flip_work_cleanup(&mdp5_crtc->unref_fb_work);
 
 	kfree(mdp5_crtc);
 }
@@ -214,6 +151,8 @@ static void mdp5_crtc_dpms(struct drm_crtc *crtc, int mode)
 			mdp5_enable(mdp5_kms);
 			mdp_irq_register(&mdp5_kms->base, &mdp5_crtc->err);
 		} else {
+			/* set STAGE_UNUSED for all layers */
+			mdp5_ctl_blend(mdp5_crtc->ctl, mdp5_crtc->lm, 0x00000000);
 			mdp_irq_unregister(&mdp5_kms->base, &mdp5_crtc->err);
 			mdp5_disable(mdp5_kms);
 		}
@@ -228,54 +167,78 @@ static bool mdp5_crtc_mode_fixup(struct drm_crtc *crtc,
 	return true;
 }
 
+/*
+ * blend_setup() - blend all the planes of a CRTC
+ *
+ * When border is enabled, the border color will ALWAYS be the base layer.
+ * Therefore, the first plane (private RGB pipe) will start at STAGE0.
+ * If disabled, the first plane starts at STAGE_BASE.
+ *
+ * Note:
+ * Border is not enabled here because the private plane is exactly
+ * the CRTC resolution.
+ */
 static void blend_setup(struct drm_crtc *crtc)
 {
 	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
 	struct mdp5_kms *mdp5_kms = get_kms(crtc);
-	int id = mdp5_crtc->id;
+	struct drm_plane *plane;
+	const struct mdp5_cfg_hw *hw_cfg;
+	uint32_t lm = mdp5_crtc->lm, blend_cfg = 0;
+	unsigned long flags;
+#define blender(stage)	((stage) - STAGE_BASE)
 
-	/*
-	 * Hard-coded setup for now until I figure out how the
-	 * layer-mixer works
-	 */
+	hw_cfg = mdp5_cfg_get_hw_config(mdp5_kms->cfg);
 
-	/* LM[id]: */
-	mdp5_write(mdp5_kms, REG_MDP5_LM_BLEND_COLOR_OUT(id),
-			MDP5_LM_BLEND_COLOR_OUT_STAGE0_FG_ALPHA);
-	mdp5_write(mdp5_kms, REG_MDP5_LM_BLEND_OP_MODE(id, 0),
-			MDP5_LM_BLEND_OP_MODE_FG_ALPHA(FG_CONST) |
-			MDP5_LM_BLEND_OP_MODE_BG_ALPHA(FG_PIXEL) |
-			MDP5_LM_BLEND_OP_MODE_BG_INV_ALPHA);
-	mdp5_write(mdp5_kms, REG_MDP5_LM_BLEND_FG_ALPHA(id, 0), 0xff);
-	mdp5_write(mdp5_kms, REG_MDP5_LM_BLEND_BG_ALPHA(id, 0), 0x00);
-
-	/* NOTE: seems that LM[n] and CTL[m], we do not need n==m.. but
-	 * we want to be setting CTL[m].LAYER[n].  Not sure what the
-	 * point of having CTL[m].LAYER[o] (for o!=n).. maybe that is
-	 * used when chaining up mixers for high resolution displays?
-	 */
+	spin_lock_irqsave(&mdp5_crtc->lm_lock, flags);
 
-	/* CTL[id]: */
-	mdp5_write(mdp5_kms, REG_MDP5_CTL_LAYER_REG(id, 0),
-			MDP5_CTL_LAYER_REG_RGB0(STAGE0) |
-			MDP5_CTL_LAYER_REG_BORDER_COLOR);
-	mdp5_write(mdp5_kms, REG_MDP5_CTL_LAYER_REG(id, 1), 0);
-	mdp5_write(mdp5_kms, REG_MDP5_CTL_LAYER_REG(id, 2), 0);
-	mdp5_write(mdp5_kms, REG_MDP5_CTL_LAYER_REG(id, 3), 0);
-	mdp5_write(mdp5_kms, REG_MDP5_CTL_LAYER_REG(id, 4), 0);
+	/* ctl could be released already when we are shutting down: */
+	if (!mdp5_crtc->ctl)
+		goto out;
+
+	drm_atomic_crtc_for_each_plane(plane, crtc) {
+		enum mdp_mixer_stage_id stage =
+			to_mdp5_plane_state(plane->state)->stage;
+
+		/*
+		 * Note: This cannot happen with current implementation but
+		 * we need to check this condition once z property is added
+		 */
+		BUG_ON(stage > hw_cfg->lm.nb_stages);
+
+		/* LM */
+		mdp5_write(mdp5_kms,
+				REG_MDP5_LM_BLEND_OP_MODE(lm, blender(stage)),
+				MDP5_LM_BLEND_OP_MODE_FG_ALPHA(FG_CONST) |
+				MDP5_LM_BLEND_OP_MODE_BG_ALPHA(BG_CONST));
+		mdp5_write(mdp5_kms, REG_MDP5_LM_BLEND_FG_ALPHA(lm,
+				blender(stage)), 0xff);
+		mdp5_write(mdp5_kms, REG_MDP5_LM_BLEND_BG_ALPHA(lm,
+				blender(stage)), 0x00);
+		/* CTL */
+		blend_cfg |= mdp_ctl_blend_mask(mdp5_plane_pipe(plane), stage);
+		DBG("%s: blending pipe %s on stage=%d", mdp5_crtc->name,
+				pipe2name(mdp5_plane_pipe(plane)), stage);
+	}
+
+	DBG("%s: lm%d: blend config = 0x%08x", mdp5_crtc->name, lm, blend_cfg);
+	mdp5_ctl_blend(mdp5_crtc->ctl, lm, blend_cfg);
+
+out:
+	spin_unlock_irqrestore(&mdp5_crtc->lm_lock, flags);
 }
 
-static int mdp5_crtc_mode_set(struct drm_crtc *crtc,
-		struct drm_display_mode *mode,
-		struct drm_display_mode *adjusted_mode,
-		int x, int y,
-		struct drm_framebuffer *old_fb)
+static void mdp5_crtc_mode_set_nofb(struct drm_crtc *crtc)
 {
 	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
 	struct mdp5_kms *mdp5_kms = get_kms(crtc);
-	int ret;
+	unsigned long flags;
+	struct drm_display_mode *mode;
 
-	mode = adjusted_mode;
+	if (WARN_ON(!crtc->state))
+		return;
+
+	mode = &crtc->state->adjusted_mode;
 
 	DBG("%s: set mode: %d:\"%s\" %d %d %d %d %d %d %d %d %d %d 0x%x 0x%x",
 			mdp5_crtc->name, mode->base.id, mode->name,
@@ -286,28 +249,11 @@ static int mdp5_crtc_mode_set(struct drm_crtc *crtc,
 			mode->vsync_end, mode->vtotal,
 			mode->type, mode->flags);
 
-	/* grab extra ref for update_scanout() */
-	drm_framebuffer_reference(crtc->primary->fb);
-
-	ret = mdp5_plane_mode_set(mdp5_crtc->plane, crtc, crtc->primary->fb,
-			0, 0, mode->hdisplay, mode->vdisplay,
-			x << 16, y << 16,
-			mode->hdisplay << 16, mode->vdisplay << 16);
-	if (ret) {
-		drm_framebuffer_unreference(crtc->primary->fb);
-		dev_err(crtc->dev->dev, "%s: failed to set mode on plane: %d\n",
-				mdp5_crtc->name, ret);
-		return ret;
-	}
-
-	mdp5_write(mdp5_kms, REG_MDP5_LM_OUT_SIZE(mdp5_crtc->id),
+	spin_lock_irqsave(&mdp5_crtc->lm_lock, flags);
+	mdp5_write(mdp5_kms, REG_MDP5_LM_OUT_SIZE(mdp5_crtc->lm),
 			MDP5_LM_OUT_SIZE_WIDTH(mode->hdisplay) |
 			MDP5_LM_OUT_SIZE_HEIGHT(mode->vdisplay));
-
-	update_fb(crtc, crtc->primary->fb);
-	update_scanout(crtc, crtc->primary->fb);
-
-	return 0;
+	spin_unlock_irqrestore(&mdp5_crtc->lm_lock, flags);
 }
 
 static void mdp5_crtc_prepare(struct drm_crtc *crtc)
@@ -321,66 +267,114 @@ static void mdp5_crtc_prepare(struct drm_crtc *crtc)
 
 static void mdp5_crtc_commit(struct drm_crtc *crtc)
 {
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	DBG("%s", mdp5_crtc->name);
 	mdp5_crtc_dpms(crtc, DRM_MODE_DPMS_ON);
-	crtc_flush(crtc);
+	crtc_flush_all(crtc);
 	/* drop the ref to mdp clk's that we got in prepare: */
 	mdp5_disable(get_kms(crtc));
 }
 
-static int mdp5_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
-		struct drm_framebuffer *old_fb)
+static void mdp5_crtc_load_lut(struct drm_crtc *crtc)
+{
+}
+
+struct plane_state {
+	struct drm_plane *plane;
+	struct mdp5_plane_state *state;
+};
+
+static int pstate_cmp(const void *a, const void *b)
+{
+	struct plane_state *pa = (struct plane_state *)a;
+	struct plane_state *pb = (struct plane_state *)b;
+	return pa->state->zpos - pb->state->zpos;
+}
+
+static int mdp5_crtc_atomic_check(struct drm_crtc *crtc,
+		struct drm_crtc_state *state)
 {
 	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
-	struct drm_plane *plane = mdp5_crtc->plane;
-	struct drm_display_mode *mode = &crtc->mode;
-	int ret;
-
-	/* grab extra ref for update_scanout() */
-	drm_framebuffer_reference(crtc->primary->fb);
-
-	ret = mdp5_plane_mode_set(plane, crtc, crtc->primary->fb,
-			0, 0, mode->hdisplay, mode->vdisplay,
-			x << 16, y << 16,
-			mode->hdisplay << 16, mode->vdisplay << 16);
-	if (ret) {
-		drm_framebuffer_unreference(crtc->primary->fb);
-		return ret;
+	struct mdp5_kms *mdp5_kms = get_kms(crtc);
+	struct drm_plane *plane;
+	struct drm_device *dev = crtc->dev;
+	struct plane_state pstates[STAGE3 + 1];
+	int cnt = 0, i;
+
+	DBG("%s: check", mdp5_crtc->name);
+
+	/* request a free CTL, if none is already allocated for this CRTC */
+	if (state->enable && !mdp5_crtc->ctl) {
+		mdp5_crtc->ctl = mdp5_ctlm_request(mdp5_kms->ctlm, crtc);
+		if (WARN_ON(!mdp5_crtc->ctl))
+			return -EINVAL;
 	}
 
-	update_fb(crtc, crtc->primary->fb);
-	update_scanout(crtc, crtc->primary->fb);
+	/* verify that there are not too many planes attached to crtc
+	 * and that we don't have conflicting mixer stages:
+	 */
+	drm_atomic_crtc_state_for_each_plane(plane, state) {
+		struct drm_plane_state *pstate;
+
+		if (cnt >= ARRAY_SIZE(pstates)) {
+			dev_err(dev->dev, "too many planes!\n");
+			return -EINVAL;
+		}
+
+		pstate = state->state->plane_states[drm_plane_index(plane)];
+
+		/* plane might not have changed, in which case take
+		 * current state:
+		 */
+		if (!pstate)
+			pstate = plane->state;
+
+		pstates[cnt].plane = plane;
+		pstates[cnt].state = to_mdp5_plane_state(pstate);
+
+		cnt++;
+	}
+
+	sort(pstates, cnt, sizeof(pstates[0]), pstate_cmp, NULL);
+
+	for (i = 0; i < cnt; i++) {
+		pstates[i].state->stage = STAGE_BASE + i;
+		DBG("%s: assign pipe %s on stage=%d", mdp5_crtc->name,
+				pipe2name(mdp5_plane_pipe(pstates[i].plane)),
+				pstates[i].state->stage);
+	}
 
 	return 0;
 }
 
-static void mdp5_crtc_load_lut(struct drm_crtc *crtc)
+static void mdp5_crtc_atomic_begin(struct drm_crtc *crtc)
 {
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	DBG("%s: begin", mdp5_crtc->name);
 }
 
-static int mdp5_crtc_page_flip(struct drm_crtc *crtc,
-		struct drm_framebuffer *new_fb,
-		struct drm_pending_vblank_event *event,
-		uint32_t page_flip_flags)
+static void mdp5_crtc_atomic_flush(struct drm_crtc *crtc)
 {
 	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
-	struct drm_gem_object *obj;
 	unsigned long flags;
 
-	if (mdp5_crtc->event) {
-		dev_err(dev->dev, "already pending flip!\n");
-		return -EBUSY;
-	}
+	DBG("%s: event: %p", mdp5_crtc->name, crtc->state->event);
 
-	obj = msm_framebuffer_bo(new_fb, 0);
+	WARN_ON(mdp5_crtc->event);
 
 	spin_lock_irqsave(&dev->event_lock, flags);
-	mdp5_crtc->event = event;
+	mdp5_crtc->event = crtc->state->event;
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 
-	update_fb(crtc, new_fb);
+	blend_setup(crtc);
+	crtc_flush_all(crtc);
+	request_pending(crtc, PENDING_FLIP);
 
-	return msm_gem_queue_inactive_cb(obj, &mdp5_crtc->pageflip_cb);
+	if (mdp5_crtc->ctl && !crtc->state->enable) {
+		mdp5_ctl_release(mdp5_crtc->ctl);
+		mdp5_crtc->ctl = NULL;
+	}
 }
 
 static int mdp5_crtc_set_property(struct drm_crtc *crtc,
@@ -391,27 +385,33 @@ static int mdp5_crtc_set_property(struct drm_crtc *crtc,
 }
 
 static const struct drm_crtc_funcs mdp5_crtc_funcs = {
-	.set_config = drm_crtc_helper_set_config,
+	.set_config = drm_atomic_helper_set_config,
 	.destroy = mdp5_crtc_destroy,
-	.page_flip = mdp5_crtc_page_flip,
+	.page_flip = drm_atomic_helper_page_flip,
 	.set_property = mdp5_crtc_set_property,
+	.reset = drm_atomic_helper_crtc_reset,
+	.atomic_duplicate_state = drm_atomic_helper_crtc_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_crtc_destroy_state,
 };
 
 static const struct drm_crtc_helper_funcs mdp5_crtc_helper_funcs = {
 	.dpms = mdp5_crtc_dpms,
 	.mode_fixup = mdp5_crtc_mode_fixup,
-	.mode_set = mdp5_crtc_mode_set,
+	.mode_set_nofb = mdp5_crtc_mode_set_nofb,
+	.mode_set = drm_helper_crtc_mode_set,
+	.mode_set_base = drm_helper_crtc_mode_set_base,
 	.prepare = mdp5_crtc_prepare,
 	.commit = mdp5_crtc_commit,
-	.mode_set_base = mdp5_crtc_mode_set_base,
 	.load_lut = mdp5_crtc_load_lut,
+	.atomic_check = mdp5_crtc_atomic_check,
+	.atomic_begin = mdp5_crtc_atomic_begin,
+	.atomic_flush = mdp5_crtc_atomic_flush,
 };
 
 static void mdp5_crtc_vblank_irq(struct mdp_irq *irq, uint32_t irqstatus)
 {
 	struct mdp5_crtc *mdp5_crtc = container_of(irq, struct mdp5_crtc, vblank);
 	struct drm_crtc *crtc = &mdp5_crtc->base;
-	struct msm_drm_private *priv = crtc->dev->dev_private;
 	unsigned pending;
 
 	mdp_irq_unregister(&get_kms(crtc)->base, &mdp5_crtc->vblank);
@@ -420,16 +420,14 @@ static void mdp5_crtc_vblank_irq(struct mdp_irq *irq, uint32_t irqstatus)
 
 	if (pending & PENDING_FLIP) {
 		complete_flip(crtc, NULL);
-		drm_flip_work_commit(&mdp5_crtc->unref_fb_work, priv->wq);
 	}
 }
 
 static void mdp5_crtc_err_irq(struct mdp_irq *irq, uint32_t irqstatus)
 {
 	struct mdp5_crtc *mdp5_crtc = container_of(irq, struct mdp5_crtc, err);
-	struct drm_crtc *crtc = &mdp5_crtc->base;
+
 	DBG("%s: error: %08x", mdp5_crtc->name, irqstatus);
-	crtc_flush(crtc);
 }
 
 uint32_t mdp5_crtc_vblank(struct drm_crtc *crtc)
@@ -450,19 +448,16 @@ void mdp5_crtc_set_intf(struct drm_crtc *crtc, int intf,
 {
 	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
 	struct mdp5_kms *mdp5_kms = get_kms(crtc);
-	static const enum mdp5_intfnum intfnum[] = {
-			INTF0, INTF1, INTF2, INTF3,
-	};
+	uint32_t flush_mask = 0;
 	uint32_t intf_sel;
+	unsigned long flags;
 
 	/* now that we know what irq's we want: */
 	mdp5_crtc->err.irqmask = intf2err(intf);
 	mdp5_crtc->vblank.irqmask = intf2vblank(intf);
+	mdp_irq_update(&mdp5_kms->base);
 
-	/* when called from modeset_init(), skip the rest until later: */
-	if (!mdp5_kms)
-		return;
-
+	spin_lock_irqsave(&mdp5_kms->resource_lock, flags);
 	intf_sel = mdp5_read(mdp5_kms, REG_MDP5_DISP_INTF_SEL);
 
 	switch (intf) {
@@ -487,45 +482,25 @@ void mdp5_crtc_set_intf(struct drm_crtc *crtc, int intf,
 		break;
 	}
 
-	blend_setup(crtc);
+	mdp5_write(mdp5_kms, REG_MDP5_DISP_INTF_SEL, intf_sel);
+	spin_unlock_irqrestore(&mdp5_kms->resource_lock, flags);
 
 	DBG("%s: intf_sel=%08x", mdp5_crtc->name, intf_sel);
+	mdp5_ctl_set_intf(mdp5_crtc->ctl, intf);
+	flush_mask |= mdp5_ctl_get_flush(mdp5_crtc->ctl);
+	flush_mask |= mdp5_lm_get_flush(mdp5_crtc->lm);
 
-	mdp5_write(mdp5_kms, REG_MDP5_DISP_INTF_SEL, intf_sel);
-	mdp5_write(mdp5_kms, REG_MDP5_CTL_OP(mdp5_crtc->id),
-			MDP5_CTL_OP_MODE(MODE_NONE) |
-			MDP5_CTL_OP_INTF_NUM(intfnum[intf]));
-
-	crtc_flush(crtc);
+	crtc_flush(crtc, flush_mask);
 }
 
-static void set_attach(struct drm_crtc *crtc, enum mdp5_pipe pipe_id,
-		struct drm_plane *plane)
+int mdp5_crtc_get_lm(struct drm_crtc *crtc)
 {
 	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
 
-	BUG_ON(pipe_id >= ARRAY_SIZE(mdp5_crtc->planes));
-
-	if (mdp5_crtc->planes[pipe_id] == plane)
-		return;
-
-	mdp5_crtc->planes[pipe_id] = plane;
-	blend_setup(crtc);
-	if (mdp5_crtc->enabled && (plane != mdp5_crtc->plane))
-		crtc_flush(crtc);
-}
+	if (WARN_ON(!crtc))
+		return -EINVAL;
 
-void mdp5_crtc_attach(struct drm_crtc *crtc, struct drm_plane *plane)
-{
-	set_attach(crtc, mdp5_plane_pipe(plane), plane);
-}
-
-void mdp5_crtc_detach(struct drm_crtc *crtc, struct drm_plane *plane)
-{
-	/* don't actually detatch our primary plane: */
-	if (to_mdp5_crtc(crtc)->plane == plane)
-		return;
-	set_attach(crtc, mdp5_plane_pipe(plane), NULL);
+	return mdp5_crtc->lm;
 }
 
 /* initialize crtc */
@@ -534,18 +509,17 @@ struct drm_crtc *mdp5_crtc_init(struct drm_device *dev,
 {
 	struct drm_crtc *crtc = NULL;
 	struct mdp5_crtc *mdp5_crtc;
-	int ret;
 
 	mdp5_crtc = kzalloc(sizeof(*mdp5_crtc), GFP_KERNEL);
-	if (!mdp5_crtc) {
-		ret = -ENOMEM;
-		goto fail;
-	}
+	if (!mdp5_crtc)
+		return ERR_PTR(-ENOMEM);
 
 	crtc = &mdp5_crtc->base;
 
-	mdp5_crtc->plane = plane;
 	mdp5_crtc->id = id;
+	mdp5_crtc->lm = GET_LM_ID(id);
+
+	spin_lock_init(&mdp5_crtc->lm_lock);
 
 	mdp5_crtc->vblank.irq = mdp5_crtc_vblank_irq;
 	mdp5_crtc->err.irq = mdp5_crtc_err_irq;
@@ -553,23 +527,11 @@ struct drm_crtc *mdp5_crtc_init(struct drm_device *dev,
 	snprintf(mdp5_crtc->name, sizeof(mdp5_crtc->name), "%s:%d",
 			pipe2name(mdp5_plane_pipe(plane)), id);
 
-	ret = drm_flip_work_init(&mdp5_crtc->unref_fb_work, 16,
-			"unref fb", unref_fb_worker);
-	if (ret)
-		goto fail;
-
-	INIT_FENCE_CB(&mdp5_crtc->pageflip_cb, pageflip_cb);
-
 	drm_crtc_init_with_planes(dev, crtc, plane, NULL, &mdp5_crtc_funcs);
 	drm_crtc_helper_add(crtc, &mdp5_crtc_helper_funcs);
+	plane->crtc = crtc;
 
-	mdp5_plane_install_properties(mdp5_crtc->plane, &crtc->base);
+	mdp5_plane_install_properties(plane, &crtc->base);
 
 	return crtc;
-
-fail:
-	if (crtc)
-		mdp5_crtc_destroy(crtc);
-
-	return ERR_PTR(ret);
 }
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_ctl.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_ctl.c
new file mode 100644
index 000000000000..dea4505ac963
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_ctl.c
@@ -0,0 +1,322 @@
+/*
+ * Copyright (c) 2014 The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include "mdp5_kms.h"
+#include "mdp5_ctl.h"
+
+/*
+ * CTL - MDP Control Pool Manager
+ *
+ * Controls are shared between all CRTCs.
+ *
+ * They are intended to be used for data path configuration.
+ * The top level register programming describes the complete data path for
+ * a specific data path ID - REG_MDP5_CTL_*(<id>, ...)
+ *
+ * Hardware capabilities determine the number of concurrent data paths
+ *
+ * In certain use cases (high-resolution dual pipe), one single CTL can be
+ * shared across multiple CRTCs.
+ *
+ * Because the number of CTLs can be less than the number of CRTCs,
+ * CTLs are dynamically allocated from a pool of CTLs, only once a CRTC is
+ * requested by the client (in mdp5_crtc_mode_set()).
+ */
+
+struct mdp5_ctl {
+	struct mdp5_ctl_manager *ctlm;
+
+	u32 id;
+
+	/* whether this CTL has been allocated or not: */
+	bool busy;
+
+	/* memory output connection (@see mdp5_ctl_mode): */
+	u32 mode;
+
+	/* REG_MDP5_CTL_*(<id>) registers access info + lock: */
+	spinlock_t hw_lock;
+	u32 reg_offset;
+
+	/* flush mask used to commit CTL registers */
+	u32 flush_mask;
+
+	bool cursor_on;
+
+	struct drm_crtc *crtc;
+};
+
+struct mdp5_ctl_manager {
+	struct drm_device *dev;
+
+	/* number of CTL / Layer Mixers in this hw config: */
+	u32 nlm;
+	u32 nctl;
+
+	/* pool of CTLs + lock to protect resource allocation (ctls[i].busy) */
+	spinlock_t pool_lock;
+	struct mdp5_ctl ctls[MAX_CTL];
+};
+
+static inline
+struct mdp5_kms *get_kms(struct mdp5_ctl_manager *ctl_mgr)
+{
+	struct msm_drm_private *priv = ctl_mgr->dev->dev_private;
+
+	return to_mdp5_kms(to_mdp_kms(priv->kms));
+}
+
+static inline
+void ctl_write(struct mdp5_ctl *ctl, u32 reg, u32 data)
+{
+	struct mdp5_kms *mdp5_kms = get_kms(ctl->ctlm);
+
+	(void)ctl->reg_offset; /* TODO use this instead of mdp5_write */
+	mdp5_write(mdp5_kms, reg, data);
+}
+
+static inline
+u32 ctl_read(struct mdp5_ctl *ctl, u32 reg)
+{
+	struct mdp5_kms *mdp5_kms = get_kms(ctl->ctlm);
+
+	(void)ctl->reg_offset; /* TODO use this instead of mdp5_write */
+	return mdp5_read(mdp5_kms, reg);
+}
+
+
+int mdp5_ctl_set_intf(struct mdp5_ctl *ctl, enum mdp5_intf intf)
+{
+	unsigned long flags;
+	static const enum mdp5_intfnum intfnum[] = {
+			INTF0, INTF1, INTF2, INTF3,
+	};
+
+	spin_lock_irqsave(&ctl->hw_lock, flags);
+	ctl_write(ctl, REG_MDP5_CTL_OP(ctl->id),
+			MDP5_CTL_OP_MODE(ctl->mode) |
+			MDP5_CTL_OP_INTF_NUM(intfnum[intf]));
+	spin_unlock_irqrestore(&ctl->hw_lock, flags);
+
+	return 0;
+}
+
+int mdp5_ctl_set_cursor(struct mdp5_ctl *ctl, bool enable)
+{
+	struct mdp5_ctl_manager *ctl_mgr = ctl->ctlm;
+	unsigned long flags;
+	u32 blend_cfg;
+	int lm;
+
+	lm = mdp5_crtc_get_lm(ctl->crtc);
+	if (unlikely(WARN_ON(lm < 0))) {
+		dev_err(ctl_mgr->dev->dev, "CTL %d cannot find LM: %d",
+				ctl->id, lm);
+		return -EINVAL;
+	}
+
+	spin_lock_irqsave(&ctl->hw_lock, flags);
+
+	blend_cfg = ctl_read(ctl, REG_MDP5_CTL_LAYER_REG(ctl->id, lm));
+
+	if (enable)
+		blend_cfg |=  MDP5_CTL_LAYER_REG_CURSOR_OUT;
+	else
+		blend_cfg &= ~MDP5_CTL_LAYER_REG_CURSOR_OUT;
+
+	ctl_write(ctl, REG_MDP5_CTL_LAYER_REG(ctl->id, lm), blend_cfg);
+
+	spin_unlock_irqrestore(&ctl->hw_lock, flags);
+
+	ctl->cursor_on = enable;
+
+	return 0;
+}
+
+
+int mdp5_ctl_blend(struct mdp5_ctl *ctl, u32 lm, u32 blend_cfg)
+{
+	unsigned long flags;
+
+	if (ctl->cursor_on)
+		blend_cfg |=  MDP5_CTL_LAYER_REG_CURSOR_OUT;
+	else
+		blend_cfg &= ~MDP5_CTL_LAYER_REG_CURSOR_OUT;
+
+	spin_lock_irqsave(&ctl->hw_lock, flags);
+	ctl_write(ctl, REG_MDP5_CTL_LAYER_REG(ctl->id, lm), blend_cfg);
+	spin_unlock_irqrestore(&ctl->hw_lock, flags);
+
+	return 0;
+}
+
+int mdp5_ctl_commit(struct mdp5_ctl *ctl, u32 flush_mask)
+{
+	struct mdp5_ctl_manager *ctl_mgr = ctl->ctlm;
+	unsigned long flags;
+
+	if (flush_mask & MDP5_CTL_FLUSH_CURSOR_DUMMY) {
+		int lm = mdp5_crtc_get_lm(ctl->crtc);
+
+		if (unlikely(WARN_ON(lm < 0))) {
+			dev_err(ctl_mgr->dev->dev, "CTL %d cannot find LM: %d",
+					ctl->id, lm);
+			return -EINVAL;
+		}
+
+		/* for current targets, cursor bit is the same as LM bit */
+		flush_mask |= mdp_ctl_flush_mask_lm(lm);
+	}
+
+	spin_lock_irqsave(&ctl->hw_lock, flags);
+	ctl_write(ctl, REG_MDP5_CTL_FLUSH(ctl->id), flush_mask);
+	spin_unlock_irqrestore(&ctl->hw_lock, flags);
+
+	return 0;
+}
+
+u32 mdp5_ctl_get_flush(struct mdp5_ctl *ctl)
+{
+	return ctl->flush_mask;
+}
+
+void mdp5_ctl_release(struct mdp5_ctl *ctl)
+{
+	struct mdp5_ctl_manager *ctl_mgr = ctl->ctlm;
+	unsigned long flags;
+
+	if (unlikely(WARN_ON(ctl->id >= MAX_CTL) || !ctl->busy)) {
+		dev_err(ctl_mgr->dev->dev, "CTL %d in bad state (%d)",
+				ctl->id, ctl->busy);
+		return;
+	}
+
+	spin_lock_irqsave(&ctl_mgr->pool_lock, flags);
+	ctl->busy = false;
+	spin_unlock_irqrestore(&ctl_mgr->pool_lock, flags);
+
+	DBG("CTL %d released", ctl->id);
+}
+
+/*
+ * mdp5_ctl_request() - CTL dynamic allocation
+ *
+ * Note: Current implementation considers that we can only have one CRTC per CTL
+ *
+ * @return first free CTL
+ */
+struct mdp5_ctl *mdp5_ctlm_request(struct mdp5_ctl_manager *ctl_mgr,
+		struct drm_crtc *crtc)
+{
+	struct mdp5_ctl *ctl = NULL;
+	unsigned long flags;
+	int c;
+
+	spin_lock_irqsave(&ctl_mgr->pool_lock, flags);
+
+	for (c = 0; c < ctl_mgr->nctl; c++)
+		if (!ctl_mgr->ctls[c].busy)
+			break;
+
+	if (unlikely(c >= ctl_mgr->nctl)) {
+		dev_err(ctl_mgr->dev->dev, "No more CTL available!");
+		goto unlock;
+	}
+
+	ctl = &ctl_mgr->ctls[c];
+
+	ctl->crtc = crtc;
+	ctl->busy = true;
+	DBG("CTL %d allocated", ctl->id);
+
+unlock:
+	spin_unlock_irqrestore(&ctl_mgr->pool_lock, flags);
+	return ctl;
+}
+
+void mdp5_ctlm_hw_reset(struct mdp5_ctl_manager *ctl_mgr)
+{
+	unsigned long flags;
+	int c;
+
+	for (c = 0; c < ctl_mgr->nctl; c++) {
+		struct mdp5_ctl *ctl = &ctl_mgr->ctls[c];
+
+		spin_lock_irqsave(&ctl->hw_lock, flags);
+		ctl_write(ctl, REG_MDP5_CTL_OP(ctl->id), 0);
+		spin_unlock_irqrestore(&ctl->hw_lock, flags);
+	}
+}
+
+void mdp5_ctlm_destroy(struct mdp5_ctl_manager *ctl_mgr)
+{
+	kfree(ctl_mgr);
+}
+
+struct mdp5_ctl_manager *mdp5_ctlm_init(struct drm_device *dev,
+		void __iomem *mmio_base, const struct mdp5_cfg_hw *hw_cfg)
+{
+	struct mdp5_ctl_manager *ctl_mgr;
+	const struct mdp5_sub_block *ctl_cfg = &hw_cfg->ctl;
+	unsigned long flags;
+	int c, ret;
+
+	ctl_mgr = kzalloc(sizeof(*ctl_mgr), GFP_KERNEL);
+	if (!ctl_mgr) {
+		dev_err(dev->dev, "failed to allocate CTL manager\n");
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	if (unlikely(WARN_ON(ctl_cfg->count > MAX_CTL))) {
+		dev_err(dev->dev, "Increase static pool size to at least %d\n",
+				ctl_cfg->count);
+		ret = -ENOSPC;
+		goto fail;
+	}
+
+	/* initialize the CTL manager: */
+	ctl_mgr->dev = dev;
+	ctl_mgr->nlm = hw_cfg->lm.count;
+	ctl_mgr->nctl = ctl_cfg->count;
+	spin_lock_init(&ctl_mgr->pool_lock);
+
+	/* initialize each CTL of the pool: */
+	spin_lock_irqsave(&ctl_mgr->pool_lock, flags);
+	for (c = 0; c < ctl_mgr->nctl; c++) {
+		struct mdp5_ctl *ctl = &ctl_mgr->ctls[c];
+
+		if (WARN_ON(!ctl_cfg->base[c])) {
+			dev_err(dev->dev, "CTL_%d: base is null!\n", c);
+			ret = -EINVAL;
+			goto fail;
+		}
+		ctl->ctlm = ctl_mgr;
+		ctl->id = c;
+		ctl->mode = MODE_NONE;
+		ctl->reg_offset = ctl_cfg->base[c];
+		ctl->flush_mask = MDP5_CTL_FLUSH_CTL;
+		ctl->busy = false;
+		spin_lock_init(&ctl->hw_lock);
+	}
+	spin_unlock_irqrestore(&ctl_mgr->pool_lock, flags);
+	DBG("Pool of %d CTLs created.", ctl_mgr->nctl);
+
+	return ctl_mgr;
+
+fail:
+	if (ctl_mgr)
+		mdp5_ctlm_destroy(ctl_mgr);
+
+	return ERR_PTR(ret);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_ctl.h b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_ctl.h
new file mode 100644
index 000000000000..1018519b6af2
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_ctl.h
@@ -0,0 +1,122 @@
+/*
+ * Copyright (c) 2014 The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#ifndef __MDP5_CTL_H__
+#define __MDP5_CTL_H__
+
+#include "msm_drv.h"
+
+/*
+ * CTL Manager prototypes:
+ * mdp5_ctlm_init() returns a ctlm (CTL Manager) handler,
+ * which is then used to call the other mdp5_ctlm_*(ctlm, ...) functions.
+ */
+struct mdp5_ctl_manager;
+struct mdp5_ctl_manager *mdp5_ctlm_init(struct drm_device *dev,
+		void __iomem *mmio_base, const struct mdp5_cfg_hw *hw_cfg);
+void mdp5_ctlm_hw_reset(struct mdp5_ctl_manager *ctlm);
+void mdp5_ctlm_destroy(struct mdp5_ctl_manager *ctlm);
+
+/*
+ * CTL prototypes:
+ * mdp5_ctl_request(ctlm, ...) returns a ctl (CTL resource) handler,
+ * which is then used to call the other mdp5_ctl_*(ctl, ...) functions.
+ */
+struct mdp5_ctl *mdp5_ctlm_request(struct mdp5_ctl_manager *ctlm, struct drm_crtc *crtc);
+
+int mdp5_ctl_set_intf(struct mdp5_ctl *ctl, enum mdp5_intf intf);
+
+int mdp5_ctl_set_cursor(struct mdp5_ctl *ctl, bool enable);
+
+/* @blend_cfg: see LM blender config definition below */
+int mdp5_ctl_blend(struct mdp5_ctl *ctl, u32 lm, u32 blend_cfg);
+
+/* @flush_mask: see CTL flush masks definitions below */
+int mdp5_ctl_commit(struct mdp5_ctl *ctl, u32 flush_mask);
+u32 mdp5_ctl_get_flush(struct mdp5_ctl *ctl);
+
+void mdp5_ctl_release(struct mdp5_ctl *ctl);
+
+/*
+ * blend_cfg (LM blender config):
+ *
+ * The function below allows the caller of mdp5_ctl_blend() to specify how pipes
+ * are being blended according to their stage (z-order), through @blend_cfg arg.
+ */
+static inline u32 mdp_ctl_blend_mask(enum mdp5_pipe pipe,
+		enum mdp_mixer_stage_id stage)
+{
+	switch (pipe) {
+	case SSPP_VIG0: return MDP5_CTL_LAYER_REG_VIG0(stage);
+	case SSPP_VIG1: return MDP5_CTL_LAYER_REG_VIG1(stage);
+	case SSPP_VIG2: return MDP5_CTL_LAYER_REG_VIG2(stage);
+	case SSPP_RGB0: return MDP5_CTL_LAYER_REG_RGB0(stage);
+	case SSPP_RGB1: return MDP5_CTL_LAYER_REG_RGB1(stage);
+	case SSPP_RGB2: return MDP5_CTL_LAYER_REG_RGB2(stage);
+	case SSPP_DMA0: return MDP5_CTL_LAYER_REG_DMA0(stage);
+	case SSPP_DMA1: return MDP5_CTL_LAYER_REG_DMA1(stage);
+	case SSPP_VIG3: return MDP5_CTL_LAYER_REG_VIG3(stage);
+	case SSPP_RGB3: return MDP5_CTL_LAYER_REG_RGB3(stage);
+	default:	return 0;
+	}
+}
+
+/*
+ * flush_mask (CTL flush masks):
+ *
+ * The following functions allow each DRM entity to get and store
+ * their own flush mask.
+ * Once stored, these masks will then be accessed through each DRM's
+ * interface and used by the caller of mdp5_ctl_commit() to specify
+ * which block(s) need to be flushed through @flush_mask parameter.
+ */
+
+#define MDP5_CTL_FLUSH_CURSOR_DUMMY	0x80000000
+
+static inline u32 mdp_ctl_flush_mask_cursor(int cursor_id)
+{
+	/* TODO: use id once multiple cursor support is present */
+	(void)cursor_id;
+
+	return MDP5_CTL_FLUSH_CURSOR_DUMMY;
+}
+
+static inline u32 mdp_ctl_flush_mask_lm(int lm)
+{
+	switch (lm) {
+	case 0:  return MDP5_CTL_FLUSH_LM0;
+	case 1:  return MDP5_CTL_FLUSH_LM1;
+	case 2:  return MDP5_CTL_FLUSH_LM2;
+	case 5:  return MDP5_CTL_FLUSH_LM5;
+	default: return 0;
+	}
+}
+
+static inline u32 mdp_ctl_flush_mask_pipe(enum mdp5_pipe pipe)
+{
+	switch (pipe) {
+	case SSPP_VIG0: return MDP5_CTL_FLUSH_VIG0;
+	case SSPP_VIG1: return MDP5_CTL_FLUSH_VIG1;
+	case SSPP_VIG2: return MDP5_CTL_FLUSH_VIG2;
+	case SSPP_RGB0: return MDP5_CTL_FLUSH_RGB0;
+	case SSPP_RGB1: return MDP5_CTL_FLUSH_RGB1;
+	case SSPP_RGB2: return MDP5_CTL_FLUSH_RGB2;
+	case SSPP_DMA0: return MDP5_CTL_FLUSH_DMA0;
+	case SSPP_DMA1: return MDP5_CTL_FLUSH_DMA1;
+	case SSPP_VIG3: return MDP5_CTL_FLUSH_VIG3;
+	case SSPP_RGB3: return MDP5_CTL_FLUSH_RGB3;
+	default:        return 0;
+	}
+}
+
+#endif /* __MDP5_CTL_H__ */
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c
index edec7bfaa952..0254bfdeb92f 100644
--- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c
@@ -24,6 +24,7 @@ struct mdp5_encoder {
 	struct drm_encoder base;
 	int intf;
 	enum mdp5_intf intf_id;
+	spinlock_t intf_lock;	/* protect REG_MDP5_INTF_* registers */
 	bool enabled;
 	uint32_t bsc;
 };
@@ -115,6 +116,7 @@ static void mdp5_encoder_dpms(struct drm_encoder *encoder, int mode)
 	struct mdp5_kms *mdp5_kms = get_kms(encoder);
 	int intf = mdp5_encoder->intf;
 	bool enabled = (mode == DRM_MODE_DPMS_ON);
+	unsigned long flags;
 
 	DBG("mode=%d", mode);
 
@@ -123,9 +125,24 @@ static void mdp5_encoder_dpms(struct drm_encoder *encoder, int mode)
 
 	if (enabled) {
 		bs_set(mdp5_encoder, 1);
+		spin_lock_irqsave(&mdp5_encoder->intf_lock, flags);
 		mdp5_write(mdp5_kms, REG_MDP5_INTF_TIMING_ENGINE_EN(intf), 1);
+		spin_unlock_irqrestore(&mdp5_encoder->intf_lock, flags);
 	} else {
+		spin_lock_irqsave(&mdp5_encoder->intf_lock, flags);
 		mdp5_write(mdp5_kms, REG_MDP5_INTF_TIMING_ENGINE_EN(intf), 0);
+		spin_unlock_irqrestore(&mdp5_encoder->intf_lock, flags);
+
+		/*
+		 * Wait for a vsync so we know the ENABLE=0 latched before
+		 * the (connector) source of the vsync's gets disabled,
+		 * otherwise we end up in a funny state if we re-enable
+		 * before the disable latches, which results that some of
+		 * the settings changes for the new modeset (like new
+		 * scanout buffer) don't latch properly..
+		 */
+		mdp_irq_wait(&mdp5_kms->base, intf2vblank(intf));
+
 		bs_set(mdp5_encoder, 0);
 	}
 
@@ -150,6 +167,7 @@ static void mdp5_encoder_mode_set(struct drm_encoder *encoder,
 	uint32_t display_v_start, display_v_end;
 	uint32_t hsync_start_x, hsync_end_x;
 	uint32_t format;
+	unsigned long flags;
 
 	mode = adjusted_mode;
 
@@ -180,6 +198,8 @@ static void mdp5_encoder_mode_set(struct drm_encoder *encoder,
 	display_v_start = (mode->vtotal - mode->vsync_start) * mode->htotal + dtv_hsync_skew;
 	display_v_end = vsync_period - ((mode->vsync_start - mode->vdisplay) * mode->htotal) + dtv_hsync_skew - 1;
 
+	spin_lock_irqsave(&mdp5_encoder->intf_lock, flags);
+
 	mdp5_write(mdp5_kms, REG_MDP5_INTF_HSYNC_CTL(intf),
 			MDP5_INTF_HSYNC_CTL_PULSEW(mode->hsync_end - mode->hsync_start) |
 			MDP5_INTF_HSYNC_CTL_PERIOD(mode->htotal));
@@ -201,6 +221,8 @@ static void mdp5_encoder_mode_set(struct drm_encoder *encoder,
 	mdp5_write(mdp5_kms, REG_MDP5_INTF_ACTIVE_VEND_F0(intf), 0);
 	mdp5_write(mdp5_kms, REG_MDP5_INTF_PANEL_FORMAT(intf), format);
 	mdp5_write(mdp5_kms, REG_MDP5_INTF_FRAME_LINE_COUNT_EN(intf), 0x3);  /* frame+line? */
+
+	spin_unlock_irqrestore(&mdp5_encoder->intf_lock, flags);
 }
 
 static void mdp5_encoder_prepare(struct drm_encoder *encoder)
@@ -242,6 +264,8 @@ struct drm_encoder *mdp5_encoder_init(struct drm_device *dev, int intf,
 	mdp5_encoder->intf_id = intf_id;
 	encoder = &mdp5_encoder->base;
 
+	spin_lock_init(&mdp5_encoder->intf_lock);
+
 	drm_encoder_init(dev, encoder, &mdp5_encoder_funcs,
 			 DRM_MODE_ENCODER_TMDS);
 	drm_encoder_helper_add(encoder, &mdp5_encoder_helper_funcs);
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_irq.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_irq.c
index f2b985bc2adf..70ac81edd40f 100644
--- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_irq.c
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_irq.c
@@ -15,6 +15,8 @@
  * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
+#include <linux/irqdomain.h>
+#include <linux/irq.h>
 
 #include "msm_drv.h"
 #include "mdp5_kms.h"
@@ -88,11 +90,17 @@ irqreturn_t mdp5_irq(struct msm_kms *kms)
 
 	VERB("intr=%08x", intr);
 
-	if (intr & MDP5_HW_INTR_STATUS_INTR_MDP)
+	if (intr & MDP5_HW_INTR_STATUS_INTR_MDP) {
 		mdp5_irq_mdp(mdp_kms);
+		intr &= ~MDP5_HW_INTR_STATUS_INTR_MDP;
+	}
 
-	if (intr & MDP5_HW_INTR_STATUS_INTR_HDMI)
-		hdmi_irq(0, mdp5_kms->hdmi);
+	while (intr) {
+		irq_hw_number_t hwirq = fls(intr) - 1;
+		generic_handle_irq(irq_find_mapping(
+				mdp5_kms->irqcontroller.domain, hwirq));
+		intr &= ~(1 << hwirq);
+	}
 
 	return IRQ_HANDLED;
 }
@@ -109,3 +117,82 @@ void mdp5_disable_vblank(struct msm_kms *kms, struct drm_crtc *crtc)
 	mdp_update_vblank_mask(to_mdp_kms(kms),
 			mdp5_crtc_vblank(crtc), false);
 }
+
+/*
+ * interrupt-controller implementation, so sub-blocks (hdmi/eDP/dsi/etc)
+ * can register to get their irq's delivered
+ */
+
+#define VALID_IRQS  (MDP5_HW_INTR_STATUS_INTR_DSI0 | \
+		MDP5_HW_INTR_STATUS_INTR_DSI1 | \
+		MDP5_HW_INTR_STATUS_INTR_HDMI | \
+		MDP5_HW_INTR_STATUS_INTR_EDP)
+
+static void mdp5_hw_mask_irq(struct irq_data *irqd)
+{
+	struct mdp5_kms *mdp5_kms = irq_data_get_irq_chip_data(irqd);
+	smp_mb__before_atomic();
+	clear_bit(irqd->hwirq, &mdp5_kms->irqcontroller.enabled_mask);
+	smp_mb__after_atomic();
+}
+
+static void mdp5_hw_unmask_irq(struct irq_data *irqd)
+{
+	struct mdp5_kms *mdp5_kms = irq_data_get_irq_chip_data(irqd);
+	smp_mb__before_atomic();
+	set_bit(irqd->hwirq, &mdp5_kms->irqcontroller.enabled_mask);
+	smp_mb__after_atomic();
+}
+
+static struct irq_chip mdp5_hw_irq_chip = {
+	.name		= "mdp5",
+	.irq_mask	= mdp5_hw_mask_irq,
+	.irq_unmask	= mdp5_hw_unmask_irq,
+};
+
+static int mdp5_hw_irqdomain_map(struct irq_domain *d,
+		unsigned int irq, irq_hw_number_t hwirq)
+{
+	struct mdp5_kms *mdp5_kms = d->host_data;
+
+	if (!(VALID_IRQS & (1 << hwirq)))
+		return -EPERM;
+
+	irq_set_chip_and_handler(irq, &mdp5_hw_irq_chip, handle_level_irq);
+	irq_set_chip_data(irq, mdp5_kms);
+	set_irq_flags(irq, IRQF_VALID);
+
+	return 0;
+}
+
+static struct irq_domain_ops mdp5_hw_irqdomain_ops = {
+	.map = mdp5_hw_irqdomain_map,
+	.xlate = irq_domain_xlate_onecell,
+};
+
+
+int mdp5_irq_domain_init(struct mdp5_kms *mdp5_kms)
+{
+	struct device *dev = mdp5_kms->dev->dev;
+	struct irq_domain *d;
+
+	d = irq_domain_add_linear(dev->of_node, 32,
+			&mdp5_hw_irqdomain_ops, mdp5_kms);
+	if (!d) {
+		dev_err(dev, "mdp5 irq domain add failed\n");
+		return -ENXIO;
+	}
+
+	mdp5_kms->irqcontroller.enabled_mask = 0;
+	mdp5_kms->irqcontroller.domain = d;
+
+	return 0;
+}
+
+void mdp5_irq_domain_fini(struct mdp5_kms *mdp5_kms)
+{
+	if (mdp5_kms->irqcontroller.domain) {
+		irq_domain_remove(mdp5_kms->irqcontroller.domain);
+		mdp5_kms->irqcontroller.domain = NULL;
+	}
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c
index 31a2c6331a1d..9f01a4f21af2 100644
--- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c
@@ -1,4 +1,5 @@
 /*
+ * Copyright (c) 2014, The Linux Foundation. All rights reserved.
  * Copyright (C) 2013 Red Hat
  * Author: Rob Clark <robdclark@gmail.com>
  *
@@ -24,145 +25,11 @@ static const char *iommu_ports[] = {
 		"mdp_0",
 };
 
-static struct mdp5_platform_config *mdp5_get_config(struct platform_device *dev);
-
-const struct mdp5_config *mdp5_cfg;
-
-static const struct mdp5_config msm8x74_config = {
-	.name = "msm8x74",
-	.ctl = {
-		.count = 5,
-		.base = { 0x00600, 0x00700, 0x00800, 0x00900, 0x00a00 },
-	},
-	.pipe_vig = {
-		.count = 3,
-		.base = { 0x01200, 0x01600, 0x01a00 },
-	},
-	.pipe_rgb = {
-		.count = 3,
-		.base = { 0x01e00, 0x02200, 0x02600 },
-	},
-	.pipe_dma = {
-		.count = 2,
-		.base = { 0x02a00, 0x02e00 },
-	},
-	.lm = {
-		.count = 5,
-		.base = { 0x03200, 0x03600, 0x03a00, 0x03e00, 0x04200 },
-	},
-	.dspp = {
-		.count = 3,
-		.base = { 0x04600, 0x04a00, 0x04e00 },
-	},
-	.ad = {
-		.count = 2,
-		.base = { 0x13100, 0x13300 }, /* NOTE: no ad in v1.0 */
-	},
-	.intf = {
-		.count = 4,
-		.base = { 0x12500, 0x12700, 0x12900, 0x12b00 },
-	},
-};
-
-static const struct mdp5_config apq8084_config = {
-	.name = "apq8084",
-	.ctl = {
-		.count = 5,
-		.base = { 0x00600, 0x00700, 0x00800, 0x00900, 0x00a00 },
-	},
-	.pipe_vig = {
-		.count = 4,
-		.base = { 0x01200, 0x01600, 0x01a00, 0x01e00 },
-	},
-	.pipe_rgb = {
-		.count = 4,
-		.base = { 0x02200, 0x02600, 0x02a00, 0x02e00 },
-	},
-	.pipe_dma = {
-		.count = 2,
-		.base = { 0x03200, 0x03600 },
-	},
-	.lm = {
-		.count = 6,
-		.base = { 0x03a00, 0x03e00, 0x04200, 0x04600, 0x04a00, 0x04e00 },
-	},
-	.dspp = {
-		.count = 4,
-		.base = { 0x05200, 0x05600, 0x05a00, 0x05e00 },
-
-	},
-	.ad = {
-		.count = 3,
-		.base = { 0x13500, 0x13700, 0x13900 },
-	},
-	.intf = {
-		.count = 5,
-		.base = { 0x12500, 0x12700, 0x12900, 0x12b00, 0x12d00 },
-	},
-};
-
-struct mdp5_config_entry {
-	int revision;
-	const struct mdp5_config *config;
-};
-
-static const struct mdp5_config_entry mdp5_configs[] = {
-	{ .revision = 0, .config = &msm8x74_config },
-	{ .revision = 2, .config = &msm8x74_config },
-	{ .revision = 3, .config = &apq8084_config },
-};
-
-static int mdp5_select_hw_cfg(struct msm_kms *kms)
-{
-	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
-	struct drm_device *dev = mdp5_kms->dev;
-	uint32_t version, major, minor;
-	int i, ret = 0;
-
-	mdp5_enable(mdp5_kms);
-	version = mdp5_read(mdp5_kms, REG_MDP5_MDP_VERSION);
-	mdp5_disable(mdp5_kms);
-
-	major = FIELD(version, MDP5_MDP_VERSION_MAJOR);
-	minor = FIELD(version, MDP5_MDP_VERSION_MINOR);
-
-	DBG("found MDP5 version v%d.%d", major, minor);
-
-	if (major != 1) {
-		dev_err(dev->dev, "unexpected MDP major version: v%d.%d\n",
-				major, minor);
-		ret = -ENXIO;
-		goto out;
-	}
-
-	mdp5_kms->rev = minor;
-
-	/* only after mdp5_cfg global pointer's init can we access the hw */
-	for (i = 0; i < ARRAY_SIZE(mdp5_configs); i++) {
-		if (mdp5_configs[i].revision != minor)
-			continue;
-		mdp5_kms->hw_cfg = mdp5_cfg = mdp5_configs[i].config;
-		break;
-	}
-	if (unlikely(!mdp5_kms->hw_cfg)) {
-		dev_err(dev->dev, "unexpected MDP minor revision: v%d.%d\n",
-				major, minor);
-		ret = -ENXIO;
-		goto out;
-	}
-
-	DBG("MDP5: %s config selected", mdp5_kms->hw_cfg->name);
-
-	return 0;
-out:
-	return ret;
-}
-
 static int mdp5_hw_init(struct msm_kms *kms)
 {
 	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
 	struct drm_device *dev = mdp5_kms->dev;
-	int i;
+	unsigned long flags;
 
 	pm_runtime_get_sync(dev->dev);
 
@@ -190,10 +57,11 @@ static int mdp5_hw_init(struct msm_kms *kms)
 	 * care.
 	 */
 
+	spin_lock_irqsave(&mdp5_kms->resource_lock, flags);
 	mdp5_write(mdp5_kms, REG_MDP5_DISP_INTF_SEL, 0);
+	spin_unlock_irqrestore(&mdp5_kms->resource_lock, flags);
 
-	for (i = 0; i < mdp5_kms->hw_cfg->ctl.count; i++)
-		mdp5_write(mdp5_kms, REG_MDP5_CTL_OP(i), 0);
+	mdp5_ctlm_hw_reset(mdp5_kms->ctlm);
 
 	pm_runtime_put_sync(dev->dev);
 
@@ -221,10 +89,20 @@ static void mdp5_destroy(struct msm_kms *kms)
 	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
 	struct msm_mmu *mmu = mdp5_kms->mmu;
 
+	mdp5_irq_domain_fini(mdp5_kms);
+
 	if (mmu) {
 		mmu->funcs->detach(mmu, iommu_ports, ARRAY_SIZE(iommu_ports));
 		mmu->funcs->destroy(mmu);
 	}
+
+	if (mdp5_kms->ctlm)
+		mdp5_ctlm_destroy(mdp5_kms->ctlm);
+	if (mdp5_kms->smp)
+		mdp5_smp_destroy(mdp5_kms->smp);
+	if (mdp5_kms->cfg)
+		mdp5_cfg_destroy(mdp5_kms->cfg);
+
 	kfree(mdp5_kms);
 }
 
@@ -274,17 +152,31 @@ static int modeset_init(struct mdp5_kms *mdp5_kms)
 	static const enum mdp5_pipe crtcs[] = {
 			SSPP_RGB0, SSPP_RGB1, SSPP_RGB2, SSPP_RGB3,
 	};
+	static const enum mdp5_pipe pub_planes[] = {
+			SSPP_VIG0, SSPP_VIG1, SSPP_VIG2, SSPP_VIG3,
+	};
 	struct drm_device *dev = mdp5_kms->dev;
 	struct msm_drm_private *priv = dev->dev_private;
 	struct drm_encoder *encoder;
+	const struct mdp5_cfg_hw *hw_cfg;
 	int i, ret;
 
-	/* construct CRTCs: */
-	for (i = 0; i < mdp5_kms->hw_cfg->pipe_rgb.count; i++) {
+	hw_cfg = mdp5_cfg_get_hw_config(mdp5_kms->cfg);
+
+	/* register our interrupt-controller for hdmi/eDP/dsi/etc
+	 * to use for irqs routed through mdp:
+	 */
+	ret = mdp5_irq_domain_init(mdp5_kms);
+	if (ret)
+		goto fail;
+
+	/* construct CRTCs and their private planes: */
+	for (i = 0; i < hw_cfg->pipe_rgb.count; i++) {
 		struct drm_plane *plane;
 		struct drm_crtc *crtc;
 
-		plane = mdp5_plane_init(dev, crtcs[i], true);
+		plane = mdp5_plane_init(dev, crtcs[i], true,
+				hw_cfg->pipe_rgb.base[i]);
 		if (IS_ERR(plane)) {
 			ret = PTR_ERR(plane);
 			dev_err(dev->dev, "failed to construct plane for %s (%d)\n",
@@ -302,6 +194,20 @@ static int modeset_init(struct mdp5_kms *mdp5_kms)
 		priv->crtcs[priv->num_crtcs++] = crtc;
 	}
 
+	/* Construct public planes: */
+	for (i = 0; i < hw_cfg->pipe_vig.count; i++) {
+		struct drm_plane *plane;
+
+		plane = mdp5_plane_init(dev, pub_planes[i], false,
+				hw_cfg->pipe_vig.base[i]);
+		if (IS_ERR(plane)) {
+			ret = PTR_ERR(plane);
+			dev_err(dev->dev, "failed to construct %s plane: %d\n",
+					pipe2name(pub_planes[i]), ret);
+			goto fail;
+		}
+	}
+
 	/* Construct encoder for HDMI: */
 	encoder = mdp5_encoder_init(dev, 3, INTF_HDMI);
 	if (IS_ERR(encoder)) {
@@ -310,25 +216,16 @@ static int modeset_init(struct mdp5_kms *mdp5_kms)
 		goto fail;
 	}
 
-	/* NOTE: the vsync and error irq's are actually associated with
-	 * the INTF/encoder.. the easiest way to deal with this (ie. what
-	 * we do now) is assume a fixed relationship between crtc's and
-	 * encoders.  I'm not sure if there is ever a need to more freely
-	 * assign crtcs to encoders, but if there is then we need to take
-	 * care of error and vblank irq's that the crtc has registered,
-	 * and also update user-requested vblank_mask.
-	 */
-	encoder->possible_crtcs = BIT(0);
-	mdp5_crtc_set_intf(priv->crtcs[0], 3, INTF_HDMI);
-
+	encoder->possible_crtcs = (1 << priv->num_crtcs) - 1;;
 	priv->encoders[priv->num_encoders++] = encoder;
 
 	/* Construct bridge/connector for HDMI: */
-	mdp5_kms->hdmi = hdmi_init(dev, encoder);
-	if (IS_ERR(mdp5_kms->hdmi)) {
-		ret = PTR_ERR(mdp5_kms->hdmi);
-		dev_err(dev->dev, "failed to initialize HDMI: %d\n", ret);
-		goto fail;
+	if (priv->hdmi) {
+		ret = hdmi_modeset_init(priv->hdmi, dev, encoder);
+		if (ret) {
+			dev_err(dev->dev, "failed to initialize HDMI: %d\n", ret);
+			goto fail;
+		}
 	}
 
 	return 0;
@@ -337,6 +234,21 @@ fail:
 	return ret;
 }
 
+static void read_hw_revision(struct mdp5_kms *mdp5_kms,
+		uint32_t *major, uint32_t *minor)
+{
+	uint32_t version;
+
+	mdp5_enable(mdp5_kms);
+	version = mdp5_read(mdp5_kms, REG_MDP5_MDP_VERSION);
+	mdp5_disable(mdp5_kms);
+
+	*major = FIELD(version, MDP5_MDP_VERSION_MAJOR);
+	*minor = FIELD(version, MDP5_MDP_VERSION_MINOR);
+
+	DBG("MDP5 version v%d.%d", *major, *minor);
+}
+
 static int get_clk(struct platform_device *pdev, struct clk **clkp,
 		const char *name)
 {
@@ -353,10 +265,11 @@ static int get_clk(struct platform_device *pdev, struct clk **clkp,
 struct msm_kms *mdp5_kms_init(struct drm_device *dev)
 {
 	struct platform_device *pdev = dev->platformdev;
-	struct mdp5_platform_config *config = mdp5_get_config(pdev);
+	struct mdp5_cfg *config;
 	struct mdp5_kms *mdp5_kms;
 	struct msm_kms *kms = NULL;
 	struct msm_mmu *mmu;
+	uint32_t major, minor;
 	int i, ret;
 
 	mdp5_kms = kzalloc(sizeof(*mdp5_kms), GFP_KERNEL);
@@ -366,12 +279,13 @@ struct msm_kms *mdp5_kms_init(struct drm_device *dev)
 		goto fail;
 	}
 
+	spin_lock_init(&mdp5_kms->resource_lock);
+
 	mdp_kms_init(&mdp5_kms->base, &kms_funcs);
 
 	kms = &mdp5_kms->base.base;
 
 	mdp5_kms->dev = dev;
-	mdp5_kms->smp_blk_cnt = config->smp_blk_cnt;
 
 	mdp5_kms->mmio = msm_ioremap(pdev, "mdp_phys", "MDP5");
 	if (IS_ERR(mdp5_kms->mmio)) {
@@ -416,24 +330,52 @@ struct msm_kms *mdp5_kms_init(struct drm_device *dev)
 	if (ret)
 		goto fail;
 
-	ret = clk_set_rate(mdp5_kms->src_clk, config->max_clk);
+	/* we need to set a default rate before enabling.  Set a safe
+	 * rate first, then figure out hw revision, and then set a
+	 * more optimal rate:
+	 */
+	clk_set_rate(mdp5_kms->src_clk, 200000000);
 
-	ret = mdp5_select_hw_cfg(kms);
-	if (ret)
+	read_hw_revision(mdp5_kms, &major, &minor);
+
+	mdp5_kms->cfg = mdp5_cfg_init(mdp5_kms, major, minor);
+	if (IS_ERR(mdp5_kms->cfg)) {
+		ret = PTR_ERR(mdp5_kms->cfg);
+		mdp5_kms->cfg = NULL;
 		goto fail;
+	}
+
+	config = mdp5_cfg_get_config(mdp5_kms->cfg);
+
+	/* TODO: compute core clock rate at runtime */
+	clk_set_rate(mdp5_kms->src_clk, config->hw->max_clk);
+
+	mdp5_kms->smp = mdp5_smp_init(mdp5_kms->dev, &config->hw->smp);
+	if (IS_ERR(mdp5_kms->smp)) {
+		ret = PTR_ERR(mdp5_kms->smp);
+		mdp5_kms->smp = NULL;
+		goto fail;
+	}
+
+	mdp5_kms->ctlm = mdp5_ctlm_init(dev, mdp5_kms->mmio, config->hw);
+	if (IS_ERR(mdp5_kms->ctlm)) {
+		ret = PTR_ERR(mdp5_kms->ctlm);
+		mdp5_kms->ctlm = NULL;
+		goto fail;
+	}
 
 	/* make sure things are off before attaching iommu (bootloader could
 	 * have left things on, in which case we'll start getting faults if
 	 * we don't disable):
 	 */
 	mdp5_enable(mdp5_kms);
-	for (i = 0; i < mdp5_kms->hw_cfg->intf.count; i++)
+	for (i = 0; i < config->hw->intf.count; i++)
 		mdp5_write(mdp5_kms, REG_MDP5_INTF_TIMING_ENGINE_EN(i), 0);
 	mdp5_disable(mdp5_kms);
 	mdelay(16);
 
-	if (config->iommu) {
-		mmu = msm_iommu_new(&pdev->dev, config->iommu);
+	if (config->platform.iommu) {
+		mmu = msm_iommu_new(&pdev->dev, config->platform.iommu);
 		if (IS_ERR(mmu)) {
 			ret = PTR_ERR(mmu);
 			dev_err(dev->dev, "failed to init iommu: %d\n", ret);
@@ -474,18 +416,3 @@ fail:
 		mdp5_destroy(kms);
 	return ERR_PTR(ret);
 }
-
-static struct mdp5_platform_config *mdp5_get_config(struct platform_device *dev)
-{
-	static struct mdp5_platform_config config = {};
-#ifdef CONFIG_OF
-	/* TODO */
-#endif
-	config.iommu = iommu_domain_alloc(&platform_bus_type);
-	/* TODO hard-coded in downstream mdss, but should it be? */
-	config.max_clk = 200000000;
-	/* TODO get from DT: */
-	config.smp_blk_cnt = 22;
-
-	return &config;
-}
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h
index 5bf340dd0f00..dd69c77c0d64 100644
--- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h
@@ -21,25 +21,9 @@
 #include "msm_drv.h"
 #include "msm_kms.h"
 #include "mdp/mdp_kms.h"
-/* dynamic offsets used by mdp5.xml.h (initialized in mdp5_kms.c) */
-#define MDP5_MAX_BASES		8
-struct mdp5_sub_block {
-	int	count;
-	uint32_t base[MDP5_MAX_BASES];
-};
-struct mdp5_config {
-	char  *name;
-	struct mdp5_sub_block ctl;
-	struct mdp5_sub_block pipe_vig;
-	struct mdp5_sub_block pipe_rgb;
-	struct mdp5_sub_block pipe_dma;
-	struct mdp5_sub_block lm;
-	struct mdp5_sub_block dspp;
-	struct mdp5_sub_block ad;
-	struct mdp5_sub_block intf;
-};
-extern const struct mdp5_config *mdp5_cfg;
+#include "mdp5_cfg.h"	/* must be included before mdp5.xml.h */
 #include "mdp5.xml.h"
+#include "mdp5_ctl.h"
 #include "mdp5_smp.h"
 
 struct mdp5_kms {
@@ -47,17 +31,14 @@ struct mdp5_kms {
 
 	struct drm_device *dev;
 
-	int rev;
-	const struct mdp5_config *hw_cfg;
+	struct mdp5_cfg_handler *cfg;
 
 	/* mapper-id used to request GEM buffer mapped for scanout: */
 	int id;
 	struct msm_mmu *mmu;
 
-	/* for tracking smp allocation amongst pipes: */
-	mdp5_smp_state_t smp_state;
-	struct mdp5_client_smp_state smp_client_state[CID_MAX];
-	int smp_blk_cnt;
+	struct mdp5_smp *smp;
+	struct mdp5_ctl_manager *ctlm;
 
 	/* io/register spaces: */
 	void __iomem *mmio, *vbif;
@@ -71,18 +52,47 @@ struct mdp5_kms {
 	struct clk *lut_clk;
 	struct clk *vsync_clk;
 
-	struct hdmi *hdmi;
+	/*
+	 * lock to protect access to global resources: ie., following register:
+	 *	- REG_MDP5_DISP_INTF_SEL
+	 */
+	spinlock_t resource_lock;
 
 	struct mdp_irq error_handler;
+
+	struct {
+		volatile unsigned long enabled_mask;
+		struct irq_domain *domain;
+	} irqcontroller;
 };
 #define to_mdp5_kms(x) container_of(x, struct mdp5_kms, base)
 
-/* platform config data (ie. from DT, or pdata) */
-struct mdp5_platform_config {
-	struct iommu_domain *iommu;
-	uint32_t max_clk;
-	int smp_blk_cnt;
+struct mdp5_plane_state {
+	struct drm_plane_state base;
+
+	/* "virtual" zpos.. we calculate actual mixer-stage at runtime
+	 * by sorting the attached planes by zpos and then assigning
+	 * mixer stage lowest to highest.  Private planes get default
+	 * zpos of zero, and public planes a unique value that is
+	 * greater than zero.  This way, things work out if a naive
+	 * userspace assigns planes to a crtc without setting zpos.
+	 */
+	int zpos;
+
+	/* the actual mixer stage, calculated in crtc->atomic_check()
+	 * NOTE: this should move to mdp5_crtc_state, when that exists
+	 */
+	enum mdp_mixer_stage_id stage;
+
+	/* some additional transactional status to help us know in the
+	 * apply path whether we need to update SMP allocation, and
+	 * whether current update is still pending:
+	 */
+	bool mode_changed : 1;
+	bool pending : 1;
 };
+#define to_mdp5_plane_state(x) \
+		container_of(x, struct mdp5_plane_state, base)
 
 static inline void mdp5_write(struct mdp5_kms *mdp5_kms, u32 reg, u32 data)
 {
@@ -107,23 +117,6 @@ static inline const char *pipe2name(enum mdp5_pipe pipe)
 	return names[pipe];
 }
 
-static inline uint32_t pipe2flush(enum mdp5_pipe pipe)
-{
-	switch (pipe) {
-	case SSPP_VIG0: return MDP5_CTL_FLUSH_VIG0;
-	case SSPP_VIG1: return MDP5_CTL_FLUSH_VIG1;
-	case SSPP_VIG2: return MDP5_CTL_FLUSH_VIG2;
-	case SSPP_RGB0: return MDP5_CTL_FLUSH_RGB0;
-	case SSPP_RGB1: return MDP5_CTL_FLUSH_RGB1;
-	case SSPP_RGB2: return MDP5_CTL_FLUSH_RGB2;
-	case SSPP_DMA0: return MDP5_CTL_FLUSH_DMA0;
-	case SSPP_DMA1: return MDP5_CTL_FLUSH_DMA1;
-	case SSPP_VIG3: return MDP5_CTL_FLUSH_VIG3;
-	case SSPP_RGB3: return MDP5_CTL_FLUSH_RGB3;
-	default:        return 0;
-	}
-}
-
 static inline int pipe2nclients(enum mdp5_pipe pipe)
 {
 	switch (pipe) {
@@ -137,34 +130,6 @@ static inline int pipe2nclients(enum mdp5_pipe pipe)
 	}
 }
 
-static inline enum mdp5_client_id pipe2client(enum mdp5_pipe pipe, int plane)
-{
-	WARN_ON(plane >= pipe2nclients(pipe));
-	switch (pipe) {
-	case SSPP_VIG0: return CID_VIG0_Y + plane;
-	case SSPP_VIG1: return CID_VIG1_Y + plane;
-	case SSPP_VIG2: return CID_VIG2_Y + plane;
-	case SSPP_RGB0: return CID_RGB0;
-	case SSPP_RGB1: return CID_RGB1;
-	case SSPP_RGB2: return CID_RGB2;
-	case SSPP_DMA0: return CID_DMA0_Y + plane;
-	case SSPP_DMA1: return CID_DMA1_Y + plane;
-	case SSPP_VIG3: return CID_VIG3_Y + plane;
-	case SSPP_RGB3: return CID_RGB3;
-	default:        return CID_UNUSED;
-	}
-}
-
-static inline uint32_t mixer2flush(int lm)
-{
-	switch (lm) {
-	case 0:  return MDP5_CTL_FLUSH_LM0;
-	case 1:  return MDP5_CTL_FLUSH_LM1;
-	case 2:  return MDP5_CTL_FLUSH_LM2;
-	default: return 0;
-	}
-}
-
 static inline uint32_t intf2err(int intf)
 {
 	switch (intf) {
@@ -197,6 +162,8 @@ void mdp5_irq_uninstall(struct msm_kms *kms);
 irqreturn_t mdp5_irq(struct msm_kms *kms);
 int mdp5_enable_vblank(struct msm_kms *kms, struct drm_crtc *crtc);
 void mdp5_disable_vblank(struct msm_kms *kms, struct drm_crtc *crtc);
+int mdp5_irq_domain_init(struct mdp5_kms *mdp5_kms);
+void mdp5_irq_domain_fini(struct mdp5_kms *mdp5_kms);
 
 static inline
 uint32_t mdp5_get_formats(enum mdp5_pipe pipe, uint32_t *pixel_formats,
@@ -210,26 +177,18 @@ uint32_t mdp5_get_formats(enum mdp5_pipe pipe, uint32_t *pixel_formats,
 
 void mdp5_plane_install_properties(struct drm_plane *plane,
 		struct drm_mode_object *obj);
-void mdp5_plane_set_scanout(struct drm_plane *plane,
-		struct drm_framebuffer *fb);
-int mdp5_plane_mode_set(struct drm_plane *plane,
-		struct drm_crtc *crtc, struct drm_framebuffer *fb,
-		int crtc_x, int crtc_y,
-		unsigned int crtc_w, unsigned int crtc_h,
-		uint32_t src_x, uint32_t src_y,
-		uint32_t src_w, uint32_t src_h);
+uint32_t mdp5_plane_get_flush(struct drm_plane *plane);
 void mdp5_plane_complete_flip(struct drm_plane *plane);
 enum mdp5_pipe mdp5_plane_pipe(struct drm_plane *plane);
 struct drm_plane *mdp5_plane_init(struct drm_device *dev,
-		enum mdp5_pipe pipe, bool private_plane);
+		enum mdp5_pipe pipe, bool private_plane, uint32_t reg_offset);
 
 uint32_t mdp5_crtc_vblank(struct drm_crtc *crtc);
 
+int mdp5_crtc_get_lm(struct drm_crtc *crtc);
 void mdp5_crtc_cancel_pending_flip(struct drm_crtc *crtc, struct drm_file *file);
 void mdp5_crtc_set_intf(struct drm_crtc *crtc, int intf,
 		enum mdp5_intf intf_id);
-void mdp5_crtc_attach(struct drm_crtc *crtc, struct drm_plane *plane);
-void mdp5_crtc_detach(struct drm_crtc *crtc, struct drm_plane *plane);
 struct drm_crtc *mdp5_crtc_init(struct drm_device *dev,
 		struct drm_plane *plane, int id);
 
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c
index f3daec4412ad..26e5fdea6594 100644
--- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c
@@ -1,4 +1,5 @@
 /*
+ * Copyright (c) 2014 The Linux Foundation. All rights reserved.
  * Copyright (C) 2013 Red Hat
  * Author: Rob Clark <robdclark@gmail.com>
  *
@@ -17,6 +18,7 @@
 
 #include "mdp5_kms.h"
 
+#define MAX_PLANE	4
 
 struct mdp5_plane {
 	struct drm_plane base;
@@ -24,6 +26,11 @@ struct mdp5_plane {
 
 	enum mdp5_pipe pipe;
 
+	spinlock_t pipe_lock;	/* protect REG_MDP5_PIPE_* registers */
+	uint32_t reg_offset;
+
+	uint32_t flush_mask;	/* used to commit pipe registers */
+
 	uint32_t nformats;
 	uint32_t formats[32];
 
@@ -31,31 +38,24 @@ struct mdp5_plane {
 };
 #define to_mdp5_plane(x) container_of(x, struct mdp5_plane, base)
 
+static int mdp5_plane_mode_set(struct drm_plane *plane,
+		struct drm_crtc *crtc, struct drm_framebuffer *fb,
+		int crtc_x, int crtc_y,
+		unsigned int crtc_w, unsigned int crtc_h,
+		uint32_t src_x, uint32_t src_y,
+		uint32_t src_w, uint32_t src_h);
+static void set_scanout_locked(struct drm_plane *plane,
+		struct drm_framebuffer *fb);
+
 static struct mdp5_kms *get_kms(struct drm_plane *plane)
 {
 	struct msm_drm_private *priv = plane->dev->dev_private;
 	return to_mdp5_kms(to_mdp_kms(priv->kms));
 }
 
-static int mdp5_plane_update(struct drm_plane *plane,
-		struct drm_crtc *crtc, struct drm_framebuffer *fb,
-		int crtc_x, int crtc_y,
-		unsigned int crtc_w, unsigned int crtc_h,
-		uint32_t src_x, uint32_t src_y,
-		uint32_t src_w, uint32_t src_h)
+static bool plane_enabled(struct drm_plane_state *state)
 {
-	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
-
-	mdp5_plane->enabled = true;
-
-	if (plane->fb)
-		drm_framebuffer_unreference(plane->fb);
-
-	drm_framebuffer_reference(fb);
-
-	return mdp5_plane_mode_set(plane, crtc, fb,
-			crtc_x, crtc_y, crtc_w, crtc_h,
-			src_x, src_y, src_w, src_h);
+	return state->fb && state->crtc;
 }
 
 static int mdp5_plane_disable(struct drm_plane *plane)
@@ -63,21 +63,13 @@ static int mdp5_plane_disable(struct drm_plane *plane)
 	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
 	struct mdp5_kms *mdp5_kms = get_kms(plane);
 	enum mdp5_pipe pipe = mdp5_plane->pipe;
-	int i;
 
 	DBG("%s: disable", mdp5_plane->name);
 
-	/* update our SMP request to zero (release all our blks): */
-	for (i = 0; i < pipe2nclients(pipe); i++)
-		mdp5_smp_request(mdp5_kms, pipe2client(pipe, i), 0);
-
-	/* TODO detaching now will cause us not to get the last
-	 * vblank and mdp5_smp_commit().. so other planes will
-	 * still see smp blocks previously allocated to us as
-	 * in-use..
-	 */
-	if (plane->crtc)
-		mdp5_crtc_detach(plane->crtc, plane);
+	if (mdp5_kms) {
+		/* Release the memory we requested earlier from the SMP: */
+		mdp5_smp_release(mdp5_kms->smp, pipe);
+	}
 
 	return 0;
 }
@@ -85,11 +77,8 @@ static int mdp5_plane_disable(struct drm_plane *plane)
 static void mdp5_plane_destroy(struct drm_plane *plane)
 {
 	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
-	struct msm_drm_private *priv = plane->dev->dev_private;
-
-	if (priv->kms)
-		mdp5_plane_disable(plane);
 
+	drm_plane_helper_disable(plane);
 	drm_plane_cleanup(plane);
 
 	kfree(mdp5_plane);
@@ -109,109 +98,186 @@ int mdp5_plane_set_property(struct drm_plane *plane,
 	return -EINVAL;
 }
 
+static void mdp5_plane_reset(struct drm_plane *plane)
+{
+	struct mdp5_plane_state *mdp5_state;
+
+	if (plane->state && plane->state->fb)
+		drm_framebuffer_unreference(plane->state->fb);
+
+	kfree(to_mdp5_plane_state(plane->state));
+	mdp5_state = kzalloc(sizeof(*mdp5_state), GFP_KERNEL);
+
+	if (plane->type == DRM_PLANE_TYPE_PRIMARY) {
+		mdp5_state->zpos = 0;
+	} else {
+		mdp5_state->zpos = 1 + drm_plane_index(plane);
+	}
+
+	plane->state = &mdp5_state->base;
+}
+
+static struct drm_plane_state *
+mdp5_plane_duplicate_state(struct drm_plane *plane)
+{
+	struct mdp5_plane_state *mdp5_state;
+
+	if (WARN_ON(!plane->state))
+		return NULL;
+
+	mdp5_state = kmemdup(to_mdp5_plane_state(plane->state),
+			sizeof(*mdp5_state), GFP_KERNEL);
+
+	if (mdp5_state && mdp5_state->base.fb)
+		drm_framebuffer_reference(mdp5_state->base.fb);
+
+	mdp5_state->mode_changed = false;
+	mdp5_state->pending = false;
+
+	return &mdp5_state->base;
+}
+
+static void mdp5_plane_destroy_state(struct drm_plane *plane,
+		struct drm_plane_state *state)
+{
+	if (state->fb)
+		drm_framebuffer_unreference(state->fb);
+
+	kfree(to_mdp5_plane_state(state));
+}
+
 static const struct drm_plane_funcs mdp5_plane_funcs = {
-		.update_plane = mdp5_plane_update,
-		.disable_plane = mdp5_plane_disable,
+		.update_plane = drm_atomic_helper_update_plane,
+		.disable_plane = drm_atomic_helper_disable_plane,
 		.destroy = mdp5_plane_destroy,
 		.set_property = mdp5_plane_set_property,
+		.reset = mdp5_plane_reset,
+		.atomic_duplicate_state = mdp5_plane_duplicate_state,
+		.atomic_destroy_state = mdp5_plane_destroy_state,
 };
 
-void mdp5_plane_set_scanout(struct drm_plane *plane,
+static int mdp5_plane_prepare_fb(struct drm_plane *plane,
 		struct drm_framebuffer *fb)
 {
 	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
 	struct mdp5_kms *mdp5_kms = get_kms(plane);
-	enum mdp5_pipe pipe = mdp5_plane->pipe;
-	uint32_t nplanes = drm_format_num_planes(fb->pixel_format);
-	uint32_t iova[4];
-	int i;
-
-	for (i = 0; i < nplanes; i++) {
-		struct drm_gem_object *bo = msm_framebuffer_bo(fb, i);
-		msm_gem_get_iova(bo, mdp5_kms->id, &iova[i]);
-	}
-	for (; i < 4; i++)
-		iova[i] = 0;
 
-	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_STRIDE_A(pipe),
-			MDP5_PIPE_SRC_STRIDE_A_P0(fb->pitches[0]) |
-			MDP5_PIPE_SRC_STRIDE_A_P1(fb->pitches[1]));
-
-	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_STRIDE_B(pipe),
-			MDP5_PIPE_SRC_STRIDE_B_P2(fb->pitches[2]) |
-			MDP5_PIPE_SRC_STRIDE_B_P3(fb->pitches[3]));
-
-	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC0_ADDR(pipe), iova[0]);
-	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC1_ADDR(pipe), iova[1]);
-	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC2_ADDR(pipe), iova[2]);
-	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC3_ADDR(pipe), iova[3]);
-
-	plane->fb = fb;
+	DBG("%s: prepare: FB[%u]", mdp5_plane->name, fb->base.id);
+	return msm_framebuffer_prepare(fb, mdp5_kms->id);
 }
 
-/* NOTE: looks like if horizontal decimation is used (if we supported that)
- * then the width used to calculate SMP block requirements is the post-
- * decimated width.  Ie. SMP buffering sits downstream of decimation (which
- * presumably happens during the dma from scanout buffer).
- */
-static int request_smp_blocks(struct drm_plane *plane, uint32_t format,
-		uint32_t nplanes, uint32_t width)
+static void mdp5_plane_cleanup_fb(struct drm_plane *plane,
+		struct drm_framebuffer *fb)
 {
-	struct drm_device *dev = plane->dev;
 	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
 	struct mdp5_kms *mdp5_kms = get_kms(plane);
-	enum mdp5_pipe pipe = mdp5_plane->pipe;
-	int i, hsub, nlines, nblks, ret;
 
-	hsub = drm_format_horz_chroma_subsampling(format);
+	DBG("%s: cleanup: FB[%u]", mdp5_plane->name, fb->base.id);
+	msm_framebuffer_cleanup(fb, mdp5_kms->id);
+}
 
-	/* different if BWC (compressed framebuffer?) enabled: */
-	nlines = 2;
+static int mdp5_plane_atomic_check(struct drm_plane *plane,
+		struct drm_plane_state *state)
+{
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+	struct drm_plane_state *old_state = plane->state;
 
-	for (i = 0, nblks = 0; i < nplanes; i++) {
-		int n, fetch_stride, cpp;
+	DBG("%s: check (%d -> %d)", mdp5_plane->name,
+			plane_enabled(old_state), plane_enabled(state));
 
-		cpp = drm_format_plane_cpp(format, i);
-		fetch_stride = width * cpp / (i ? hsub : 1);
+	if (plane_enabled(state) && plane_enabled(old_state)) {
+		/* we cannot change SMP block configuration during scanout: */
+		bool full_modeset = false;
+		if (state->fb->pixel_format != old_state->fb->pixel_format) {
+			DBG("%s: pixel_format change!", mdp5_plane->name);
+			full_modeset = true;
+		}
+		if (state->src_w != old_state->src_w) {
+			DBG("%s: src_w change!", mdp5_plane->name);
+			full_modeset = true;
+		}
+		if (to_mdp5_plane_state(old_state)->pending) {
+			DBG("%s: still pending!", mdp5_plane->name);
+			full_modeset = true;
+		}
+		if (full_modeset) {
+			struct drm_crtc_state *crtc_state =
+					drm_atomic_get_crtc_state(state->state, state->crtc);
+			crtc_state->mode_changed = true;
+			to_mdp5_plane_state(state)->mode_changed = true;
+		}
+	} else {
+		to_mdp5_plane_state(state)->mode_changed = true;
+	}
 
-		n = DIV_ROUND_UP(fetch_stride * nlines, SMP_BLK_SIZE);
+	return 0;
+}
 
-		/* for hw rev v1.00 */
-		if (mdp5_kms->rev == 0)
-			n = roundup_pow_of_two(n);
+static void mdp5_plane_atomic_update(struct drm_plane *plane,
+				     struct drm_plane_state *old_state)
+{
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+	struct drm_plane_state *state = plane->state;
 
-		DBG("%s[%d]: request %d SMP blocks", mdp5_plane->name, i, n);
-		ret = mdp5_smp_request(mdp5_kms, pipe2client(pipe, i), n);
-		if (ret) {
-			dev_err(dev->dev, "Could not allocate %d SMP blocks: %d\n",
-					n, ret);
-			return ret;
-		}
+	DBG("%s: update", mdp5_plane->name);
 
-		nblks += n;
+	if (!plane_enabled(state)) {
+		to_mdp5_plane_state(state)->pending = true;
+		mdp5_plane_disable(plane);
+	} else if (to_mdp5_plane_state(state)->mode_changed) {
+		int ret;
+		to_mdp5_plane_state(state)->pending = true;
+		ret = mdp5_plane_mode_set(plane,
+				state->crtc, state->fb,
+				state->crtc_x, state->crtc_y,
+				state->crtc_w, state->crtc_h,
+				state->src_x,  state->src_y,
+				state->src_w, state->src_h);
+		/* atomic_check should have ensured that this doesn't fail */
+		WARN_ON(ret < 0);
+	} else {
+		unsigned long flags;
+		spin_lock_irqsave(&mdp5_plane->pipe_lock, flags);
+		set_scanout_locked(plane, state->fb);
+		spin_unlock_irqrestore(&mdp5_plane->pipe_lock, flags);
 	}
-
-	/* in success case, return total # of blocks allocated: */
-	return nblks;
 }
 
-static void set_fifo_thresholds(struct drm_plane *plane, int nblks)
+static const struct drm_plane_helper_funcs mdp5_plane_helper_funcs = {
+		.prepare_fb = mdp5_plane_prepare_fb,
+		.cleanup_fb = mdp5_plane_cleanup_fb,
+		.atomic_check = mdp5_plane_atomic_check,
+		.atomic_update = mdp5_plane_atomic_update,
+};
+
+static void set_scanout_locked(struct drm_plane *plane,
+		struct drm_framebuffer *fb)
 {
 	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
 	struct mdp5_kms *mdp5_kms = get_kms(plane);
 	enum mdp5_pipe pipe = mdp5_plane->pipe;
-	uint32_t val;
 
-	/* 1/4 of SMP pool that is being fetched */
-	val = (nblks * SMP_ENTRIES_PER_BLK) / 4;
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_STRIDE_A(pipe),
+			MDP5_PIPE_SRC_STRIDE_A_P0(fb->pitches[0]) |
+			MDP5_PIPE_SRC_STRIDE_A_P1(fb->pitches[1]));
 
-	mdp5_write(mdp5_kms, REG_MDP5_PIPE_REQPRIO_FIFO_WM_0(pipe), val * 1);
-	mdp5_write(mdp5_kms, REG_MDP5_PIPE_REQPRIO_FIFO_WM_1(pipe), val * 2);
-	mdp5_write(mdp5_kms, REG_MDP5_PIPE_REQPRIO_FIFO_WM_2(pipe), val * 3);
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_STRIDE_B(pipe),
+			MDP5_PIPE_SRC_STRIDE_B_P2(fb->pitches[2]) |
+			MDP5_PIPE_SRC_STRIDE_B_P3(fb->pitches[3]));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC0_ADDR(pipe),
+			msm_framebuffer_iova(fb, mdp5_kms->id, 0));
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC1_ADDR(pipe),
+			msm_framebuffer_iova(fb, mdp5_kms->id, 1));
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC2_ADDR(pipe),
+			msm_framebuffer_iova(fb, mdp5_kms->id, 2));
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC3_ADDR(pipe),
+			msm_framebuffer_iova(fb, mdp5_kms->id, 4));
 
+	plane->fb = fb;
 }
 
-int mdp5_plane_mode_set(struct drm_plane *plane,
+static int mdp5_plane_mode_set(struct drm_plane *plane,
 		struct drm_crtc *crtc, struct drm_framebuffer *fb,
 		int crtc_x, int crtc_y,
 		unsigned int crtc_w, unsigned int crtc_h,
@@ -225,7 +291,8 @@ int mdp5_plane_mode_set(struct drm_plane *plane,
 	uint32_t nplanes, config = 0;
 	uint32_t phasex_step = 0, phasey_step = 0;
 	uint32_t hdecm = 0, vdecm = 0;
-	int i, nblks;
+	unsigned long flags;
+	int ret;
 
 	nplanes = drm_format_num_planes(fb->pixel_format);
 
@@ -243,12 +310,11 @@ int mdp5_plane_mode_set(struct drm_plane *plane,
 			fb->base.id, src_x, src_y, src_w, src_h,
 			crtc->base.id, crtc_x, crtc_y, crtc_w, crtc_h);
 
-	/*
-	 * Calculate and request required # of smp blocks:
-	 */
-	nblks = request_smp_blocks(plane, fb->pixel_format, nplanes, src_w);
-	if (nblks < 0)
-		return nblks;
+	/* Request some memory from the SMP: */
+	ret = mdp5_smp_request(mdp5_kms->smp,
+			mdp5_plane->pipe, fb->pixel_format, src_w);
+	if (ret)
+		return ret;
 
 	/*
 	 * Currently we update the hw for allocations/requests immediately,
@@ -256,8 +322,7 @@ int mdp5_plane_mode_set(struct drm_plane *plane,
 	 * would move into atomic->check_plane_state(), while updating the
 	 * hw would remain here:
 	 */
-	for (i = 0; i < pipe2nclients(pipe); i++)
-		mdp5_smp_configure(mdp5_kms, pipe2client(pipe, i));
+	mdp5_smp_configure(mdp5_kms->smp, pipe);
 
 	if (src_w != crtc_w) {
 		config |= MDP5_PIPE_SCALE_CONFIG_SCALEX_EN;
@@ -269,6 +334,8 @@ int mdp5_plane_mode_set(struct drm_plane *plane,
 		/* TODO calc phasey_step, vdecm */
 	}
 
+	spin_lock_irqsave(&mdp5_plane->pipe_lock, flags);
+
 	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_IMG_SIZE(pipe),
 			MDP5_PIPE_SRC_IMG_SIZE_WIDTH(src_w) |
 			MDP5_PIPE_SRC_IMG_SIZE_HEIGHT(src_h));
@@ -289,8 +356,6 @@ int mdp5_plane_mode_set(struct drm_plane *plane,
 			MDP5_PIPE_OUT_XY_X(crtc_x) |
 			MDP5_PIPE_OUT_XY_Y(crtc_y));
 
-	mdp5_plane_set_scanout(plane, fb);
-
 	format = to_mdp_format(msm_framebuffer_format(fb));
 
 	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_FORMAT(pipe),
@@ -330,22 +395,24 @@ int mdp5_plane_mode_set(struct drm_plane *plane,
 			MDP5_PIPE_SCALE_CONFIG_SCALEX_MAX_FILTER(SCALE_FILTER_NEAREST) |
 			MDP5_PIPE_SCALE_CONFIG_SCALEY_MAX_FILTER(SCALE_FILTER_NEAREST));
 
-	set_fifo_thresholds(plane, nblks);
+	set_scanout_locked(plane, fb);
 
-	/* TODO detach from old crtc (if we had more than one) */
-	mdp5_crtc_attach(crtc, plane);
+	spin_unlock_irqrestore(&mdp5_plane->pipe_lock, flags);
 
-	return 0;
+	return ret;
 }
 
 void mdp5_plane_complete_flip(struct drm_plane *plane)
 {
 	struct mdp5_kms *mdp5_kms = get_kms(plane);
-	enum mdp5_pipe pipe = to_mdp5_plane(plane)->pipe;
-	int i;
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+	enum mdp5_pipe pipe = mdp5_plane->pipe;
+
+	DBG("%s: complete flip", mdp5_plane->name);
 
-	for (i = 0; i < pipe2nclients(pipe); i++)
-		mdp5_smp_commit(mdp5_kms, pipe2client(pipe, i));
+	mdp5_smp_commit(mdp5_kms->smp, pipe);
+
+	to_mdp5_plane_state(plane->state)->pending = false;
 }
 
 enum mdp5_pipe mdp5_plane_pipe(struct drm_plane *plane)
@@ -354,9 +421,16 @@ enum mdp5_pipe mdp5_plane_pipe(struct drm_plane *plane)
 	return mdp5_plane->pipe;
 }
 
+uint32_t mdp5_plane_get_flush(struct drm_plane *plane)
+{
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+
+	return mdp5_plane->flush_mask;
+}
+
 /* initialize plane */
 struct drm_plane *mdp5_plane_init(struct drm_device *dev,
-		enum mdp5_pipe pipe, bool private_plane)
+		enum mdp5_pipe pipe, bool private_plane, uint32_t reg_offset)
 {
 	struct drm_plane *plane = NULL;
 	struct mdp5_plane *mdp5_plane;
@@ -377,10 +451,18 @@ struct drm_plane *mdp5_plane_init(struct drm_device *dev,
 	mdp5_plane->nformats = mdp5_get_formats(pipe, mdp5_plane->formats,
 			ARRAY_SIZE(mdp5_plane->formats));
 
+	mdp5_plane->flush_mask = mdp_ctl_flush_mask_pipe(pipe);
+	mdp5_plane->reg_offset = reg_offset;
+	spin_lock_init(&mdp5_plane->pipe_lock);
+
 	type = private_plane ? DRM_PLANE_TYPE_PRIMARY : DRM_PLANE_TYPE_OVERLAY;
-	drm_universal_plane_init(dev, plane, 0xff, &mdp5_plane_funcs,
+	ret = drm_universal_plane_init(dev, plane, 0xff, &mdp5_plane_funcs,
 				 mdp5_plane->formats, mdp5_plane->nformats,
 				 type);
+	if (ret)
+		goto fail;
+
+	drm_plane_helper_add(plane, &mdp5_plane_helper_funcs);
 
 	mdp5_plane_install_properties(plane, &plane->base);
 
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.c
index 2d0236b963a6..bf551885e019 100644
--- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.c
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.c
@@ -1,4 +1,5 @@
 /*
+ * Copyright (c) 2014, The Linux Foundation. All rights reserved.
  * Copyright (C) 2013 Red Hat
  * Author: Rob Clark <robdclark@gmail.com>
  *
@@ -29,8 +30,11 @@
  * Based on the size of the attached scanout buffer, a certain # of
  * blocks must be allocated to that client out of the shared pool.
  *
- * For each block, it can be either free, or pending/in-use by a
- * client.  The updates happen in three steps:
+ * In some hw, some blocks are statically allocated for certain pipes
+ * and CANNOT be re-allocated (eg: MMB0 and MMB1 both tied to RGB0).
+ *
+ * For each block that can be dynamically allocated, it can be either
+ * free, or pending/in-use by a client. The updates happen in three steps:
  *
  *  1) mdp5_smp_request():
  *     When plane scanout is setup, calculate required number of
@@ -61,21 +65,68 @@
  * inuse and pending state of all clients..
  */
 
-static DEFINE_SPINLOCK(smp_lock);
+struct mdp5_smp {
+	struct drm_device *dev;
+
+	int blk_cnt;
+	int blk_size;
+
+	spinlock_t state_lock;
+	mdp5_smp_state_t state; /* to track smp allocation amongst pipes: */
+
+	struct mdp5_client_smp_state client_state[CID_MAX];
+};
 
+static inline
+struct mdp5_kms *get_kms(struct mdp5_smp *smp)
+{
+	struct msm_drm_private *priv = smp->dev->dev_private;
+
+	return to_mdp5_kms(to_mdp_kms(priv->kms));
+}
+
+static inline enum mdp5_client_id pipe2client(enum mdp5_pipe pipe, int plane)
+{
+	WARN_ON(plane >= pipe2nclients(pipe));
+	switch (pipe) {
+	case SSPP_VIG0: return CID_VIG0_Y + plane;
+	case SSPP_VIG1: return CID_VIG1_Y + plane;
+	case SSPP_VIG2: return CID_VIG2_Y + plane;
+	case SSPP_RGB0: return CID_RGB0;
+	case SSPP_RGB1: return CID_RGB1;
+	case SSPP_RGB2: return CID_RGB2;
+	case SSPP_DMA0: return CID_DMA0_Y + plane;
+	case SSPP_DMA1: return CID_DMA1_Y + plane;
+	case SSPP_VIG3: return CID_VIG3_Y + plane;
+	case SSPP_RGB3: return CID_RGB3;
+	default:        return CID_UNUSED;
+	}
+}
 
 /* step #1: update # of blocks pending for the client: */
-int mdp5_smp_request(struct mdp5_kms *mdp5_kms,
+static int smp_request_block(struct mdp5_smp *smp,
 		enum mdp5_client_id cid, int nblks)
 {
-	struct mdp5_client_smp_state *ps = &mdp5_kms->smp_client_state[cid];
-	int i, ret, avail, cur_nblks, cnt = mdp5_kms->smp_blk_cnt;
+	struct mdp5_kms *mdp5_kms = get_kms(smp);
+	const struct mdp5_cfg_hw *hw_cfg;
+	struct mdp5_client_smp_state *ps = &smp->client_state[cid];
+	int i, ret, avail, cur_nblks, cnt = smp->blk_cnt;
+	int reserved;
 	unsigned long flags;
 
-	spin_lock_irqsave(&smp_lock, flags);
+	hw_cfg = mdp5_cfg_get_hw_config(mdp5_kms->cfg);
+	reserved = hw_cfg->smp.reserved[cid];
+
+	spin_lock_irqsave(&smp->state_lock, flags);
 
-	avail = cnt - bitmap_weight(mdp5_kms->smp_state, cnt);
+	nblks -= reserved;
+	if (reserved)
+		DBG("%d MMBs allocated (%d reserved)", nblks, reserved);
+
+	avail = cnt - bitmap_weight(smp->state, cnt);
 	if (nblks > avail) {
+		dev_err(mdp5_kms->dev->dev, "out of blks (req=%d > avail=%d)\n",
+				nblks, avail);
 		ret = -ENOSPC;
 		goto fail;
 	}
@@ -84,9 +135,9 @@ int mdp5_smp_request(struct mdp5_kms *mdp5_kms,
 	if (nblks > cur_nblks) {
 		/* grow the existing pending reservation: */
 		for (i = cur_nblks; i < nblks; i++) {
-			int blk = find_first_zero_bit(mdp5_kms->smp_state, cnt);
+			int blk = find_first_zero_bit(smp->state, cnt);
 			set_bit(blk, ps->pending);
-			set_bit(blk, mdp5_kms->smp_state);
+			set_bit(blk, smp->state);
 		}
 	} else {
 		/* shrink the existing pending reservation: */
@@ -98,15 +149,88 @@ int mdp5_smp_request(struct mdp5_kms *mdp5_kms,
 	}
 
 fail:
-	spin_unlock_irqrestore(&smp_lock, flags);
+	spin_unlock_irqrestore(&smp->state_lock, flags);
+	return 0;
+}
+
+static void set_fifo_thresholds(struct mdp5_smp *smp,
+		enum mdp5_pipe pipe, int nblks)
+{
+	struct mdp5_kms *mdp5_kms = get_kms(smp);
+	u32 smp_entries_per_blk = smp->blk_size / (128 / BITS_PER_BYTE);
+	u32 val;
+
+	/* 1/4 of SMP pool that is being fetched */
+	val = (nblks * smp_entries_per_blk) / 4;
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_REQPRIO_FIFO_WM_0(pipe), val * 1);
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_REQPRIO_FIFO_WM_1(pipe), val * 2);
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_REQPRIO_FIFO_WM_2(pipe), val * 3);
+}
+
+/*
+ * NOTE: looks like if horizontal decimation is used (if we supported that)
+ * then the width used to calculate SMP block requirements is the post-
+ * decimated width.  Ie. SMP buffering sits downstream of decimation (which
+ * presumably happens during the dma from scanout buffer).
+ */
+int mdp5_smp_request(struct mdp5_smp *smp, enum mdp5_pipe pipe, u32 fmt, u32 width)
+{
+	struct mdp5_kms *mdp5_kms = get_kms(smp);
+	struct drm_device *dev = mdp5_kms->dev;
+	int rev = mdp5_cfg_get_hw_rev(mdp5_kms->cfg);
+	int i, hsub, nplanes, nlines, nblks, ret;
+
+	nplanes = drm_format_num_planes(fmt);
+	hsub = drm_format_horz_chroma_subsampling(fmt);
+
+	/* different if BWC (compressed framebuffer?) enabled: */
+	nlines = 2;
+
+	for (i = 0, nblks = 0; i < nplanes; i++) {
+		int n, fetch_stride, cpp;
+
+		cpp = drm_format_plane_cpp(fmt, i);
+		fetch_stride = width * cpp / (i ? hsub : 1);
+
+		n = DIV_ROUND_UP(fetch_stride * nlines, smp->blk_size);
+
+		/* for hw rev v1.00 */
+		if (rev == 0)
+			n = roundup_pow_of_two(n);
+
+		DBG("%s[%d]: request %d SMP blocks", pipe2name(pipe), i, n);
+		ret = smp_request_block(smp, pipe2client(pipe, i), n);
+		if (ret) {
+			dev_err(dev->dev, "Cannot allocate %d SMP blocks: %d\n",
+					n, ret);
+			return ret;
+		}
+
+		nblks += n;
+	}
+
+	set_fifo_thresholds(smp, pipe, nblks);
+
 	return 0;
 }
 
-static void update_smp_state(struct mdp5_kms *mdp5_kms,
+/* Release SMP blocks for all clients of the pipe */
+void mdp5_smp_release(struct mdp5_smp *smp, enum mdp5_pipe pipe)
+{
+	int i, nblks;
+
+	for (i = 0, nblks = 0; i < pipe2nclients(pipe); i++)
+		smp_request_block(smp, pipe2client(pipe, i), 0);
+	set_fifo_thresholds(smp, pipe, 0);
+}
+
+static void update_smp_state(struct mdp5_smp *smp,
 		enum mdp5_client_id cid, mdp5_smp_state_t *assigned)
 {
-	int cnt = mdp5_kms->smp_blk_cnt;
-	uint32_t blk, val;
+	struct mdp5_kms *mdp5_kms = get_kms(smp);
+	int cnt = smp->blk_cnt;
+	u32 blk, val;
 
 	for_each_set_bit(blk, *assigned, cnt) {
 		int idx = blk / 3;
@@ -135,39 +259,80 @@ static void update_smp_state(struct mdp5_kms *mdp5_kms,
 }
 
 /* step #2: configure hw for union(pending, inuse): */
-void mdp5_smp_configure(struct mdp5_kms *mdp5_kms, enum mdp5_client_id cid)
+void mdp5_smp_configure(struct mdp5_smp *smp, enum mdp5_pipe pipe)
 {
-	struct mdp5_client_smp_state *ps = &mdp5_kms->smp_client_state[cid];
-	int cnt = mdp5_kms->smp_blk_cnt;
+	int cnt = smp->blk_cnt;
 	mdp5_smp_state_t assigned;
+	int i;
 
-	bitmap_or(assigned, ps->inuse, ps->pending, cnt);
-	update_smp_state(mdp5_kms, cid, &assigned);
+	for (i = 0; i < pipe2nclients(pipe); i++) {
+		enum mdp5_client_id cid = pipe2client(pipe, i);
+		struct mdp5_client_smp_state *ps = &smp->client_state[cid];
+
+		bitmap_or(assigned, ps->inuse, ps->pending, cnt);
+		update_smp_state(smp, cid, &assigned);
+	}
 }
 
 /* step #3: after vblank, copy pending -> inuse: */
-void mdp5_smp_commit(struct mdp5_kms *mdp5_kms, enum mdp5_client_id cid)
+void mdp5_smp_commit(struct mdp5_smp *smp, enum mdp5_pipe pipe)
 {
-	struct mdp5_client_smp_state *ps = &mdp5_kms->smp_client_state[cid];
-	int cnt = mdp5_kms->smp_blk_cnt;
+	int cnt = smp->blk_cnt;
 	mdp5_smp_state_t released;
+	int i;
+
+	for (i = 0; i < pipe2nclients(pipe); i++) {
+		enum mdp5_client_id cid = pipe2client(pipe, i);
+		struct mdp5_client_smp_state *ps = &smp->client_state[cid];
+
+		/*
+		 * Figure out if there are any blocks we where previously
+		 * using, which can be released and made available to other
+		 * clients:
+		 */
+		if (bitmap_andnot(released, ps->inuse, ps->pending, cnt)) {
+			unsigned long flags;
+
+			spin_lock_irqsave(&smp->state_lock, flags);
+			/* clear released blocks: */
+			bitmap_andnot(smp->state, smp->state, released, cnt);
+			spin_unlock_irqrestore(&smp->state_lock, flags);
 
-	/*
-	 * Figure out if there are any blocks we where previously
-	 * using, which can be released and made available to other
-	 * clients:
-	 */
-	if (bitmap_andnot(released, ps->inuse, ps->pending, cnt)) {
-		unsigned long flags;
-
-		spin_lock_irqsave(&smp_lock, flags);
-		/* clear released blocks: */
-		bitmap_andnot(mdp5_kms->smp_state, mdp5_kms->smp_state,
-				released, cnt);
-		spin_unlock_irqrestore(&smp_lock, flags);
-
-		update_smp_state(mdp5_kms, CID_UNUSED, &released);
+			update_smp_state(smp, CID_UNUSED, &released);
+		}
+
+		bitmap_copy(ps->inuse, ps->pending, cnt);
 	}
+}
+
+void mdp5_smp_destroy(struct mdp5_smp *smp)
+{
+	kfree(smp);
+}
+
+struct mdp5_smp *mdp5_smp_init(struct drm_device *dev, const struct mdp5_smp_block *cfg)
+{
+	struct mdp5_smp *smp = NULL;
+	int ret;
+
+	smp = kzalloc(sizeof(*smp), GFP_KERNEL);
+	if (unlikely(!smp)) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	smp->dev = dev;
+	smp->blk_cnt = cfg->mmb_count;
+	smp->blk_size = cfg->mmb_size;
+
+	/* statically tied MMBs cannot be re-allocated: */
+	bitmap_copy(smp->state, cfg->reserved_state, smp->blk_cnt);
+	spin_lock_init(&smp->state_lock);
+
+	return smp;
+fail:
+	if (smp)
+		mdp5_smp_destroy(smp);
 
-	bitmap_copy(ps->inuse, ps->pending, cnt);
+	return ERR_PTR(ret);
 }
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.h b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.h
index 0ab739e1a1dd..e47179f63585 100644
--- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.h
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.h
@@ -1,4 +1,5 @@
 /*
+ * Copyright (c) 2014, The Linux Foundation. All rights reserved.
  * Copyright (C) 2013 Red Hat
  * Author: Rob Clark <robdclark@gmail.com>
  *
@@ -20,22 +21,26 @@
 
 #include "msm_drv.h"
 
-#define MAX_SMP_BLOCKS  22
-#define SMP_BLK_SIZE    4096
-#define SMP_ENTRIES_PER_BLK (SMP_BLK_SIZE / 16)
-
-typedef DECLARE_BITMAP(mdp5_smp_state_t, MAX_SMP_BLOCKS);
-
 struct mdp5_client_smp_state {
 	mdp5_smp_state_t inuse;
 	mdp5_smp_state_t pending;
 };
 
 struct mdp5_kms;
+struct mdp5_smp;
+
+/*
+ * SMP module prototypes:
+ * mdp5_smp_init() returns a SMP @handler,
+ * which is then used to call the other mdp5_smp_*(handler, ...) functions.
+ */
 
-int mdp5_smp_request(struct mdp5_kms *mdp5_kms, enum mdp5_client_id cid, int nblks);
-void mdp5_smp_configure(struct mdp5_kms *mdp5_kms, enum mdp5_client_id cid);
-void mdp5_smp_commit(struct mdp5_kms *mdp5_kms, enum mdp5_client_id cid);
+struct mdp5_smp *mdp5_smp_init(struct drm_device *dev, const struct mdp5_smp_block *cfg);
+void  mdp5_smp_destroy(struct mdp5_smp *smp);
 
+int  mdp5_smp_request(struct mdp5_smp *smp, enum mdp5_pipe pipe, u32 fmt, u32 width);
+void mdp5_smp_configure(struct mdp5_smp *smp, enum mdp5_pipe pipe);
+void mdp5_smp_commit(struct mdp5_smp *smp, enum mdp5_pipe pipe);
+void mdp5_smp_release(struct mdp5_smp *smp, enum mdp5_pipe pipe);
 
 #endif /* __MDP5_SMP_H__ */
diff --git a/drivers/gpu/drm/msm/mdp/mdp_kms.c b/drivers/gpu/drm/msm/mdp/mdp_kms.c
index 03455b64a245..2a731722d840 100644
--- a/drivers/gpu/drm/msm/mdp/mdp_kms.c
+++ b/drivers/gpu/drm/msm/mdp/mdp_kms.c
@@ -42,7 +42,10 @@ static void update_irq(struct mdp_kms *mdp_kms)
 	mdp_kms->funcs->set_irqmask(mdp_kms, irqmask);
 }
 
-static void update_irq_unlocked(struct mdp_kms *mdp_kms)
+/* if an mdp_irq's irqmask has changed, such as when mdp5 crtc<->encoder
+ * link changes, this must be called to figure out the new global irqmask
+ */
+void mdp_irq_update(struct mdp_kms *mdp_kms)
 {
 	unsigned long flags;
 	spin_lock_irqsave(&list_lock, flags);
@@ -122,7 +125,7 @@ void mdp_irq_register(struct mdp_kms *mdp_kms, struct mdp_irq *irq)
 	spin_unlock_irqrestore(&list_lock, flags);
 
 	if (needs_update)
-		update_irq_unlocked(mdp_kms);
+		mdp_irq_update(mdp_kms);
 }
 
 void mdp_irq_unregister(struct mdp_kms *mdp_kms, struct mdp_irq *irq)
@@ -141,5 +144,5 @@ void mdp_irq_unregister(struct mdp_kms *mdp_kms, struct mdp_irq *irq)
 	spin_unlock_irqrestore(&list_lock, flags);
 
 	if (needs_update)
-		update_irq_unlocked(mdp_kms);
+		mdp_irq_update(mdp_kms);
 }
diff --git a/drivers/gpu/drm/msm/mdp/mdp_kms.h b/drivers/gpu/drm/msm/mdp/mdp_kms.h
index 99557b5ad4fd..b268ce95d394 100644
--- a/drivers/gpu/drm/msm/mdp/mdp_kms.h
+++ b/drivers/gpu/drm/msm/mdp/mdp_kms.h
@@ -75,7 +75,7 @@ void mdp_update_vblank_mask(struct mdp_kms *mdp_kms, uint32_t mask, bool enable)
 void mdp_irq_wait(struct mdp_kms *mdp_kms, uint32_t irqmask);
 void mdp_irq_register(struct mdp_kms *mdp_kms, struct mdp_irq *irq);
 void mdp_irq_unregister(struct mdp_kms *mdp_kms, struct mdp_irq *irq);
-
+void mdp_irq_update(struct mdp_kms *mdp_kms);
 
 /*
  * pixel format helpers:
diff --git a/drivers/gpu/drm/msm/msm_atomic.c b/drivers/gpu/drm/msm/msm_atomic.c
new file mode 100644
index 000000000000..191968256c58
--- /dev/null
+++ b/drivers/gpu/drm/msm/msm_atomic.c
@@ -0,0 +1,230 @@
+/*
+ * Copyright (C) 2014 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "msm_drv.h"
+#include "msm_kms.h"
+#include "msm_gem.h"
+
+struct msm_commit {
+	struct drm_atomic_state *state;
+	uint32_t fence;
+	struct msm_fence_cb fence_cb;
+	uint32_t crtc_mask;
+};
+
+static void fence_cb(struct msm_fence_cb *cb);
+
+/* block until specified crtcs are no longer pending update, and
+ * atomically mark them as pending update
+ */
+static int start_atomic(struct msm_drm_private *priv, uint32_t crtc_mask)
+{
+	int ret;
+
+	spin_lock(&priv->pending_crtcs_event.lock);
+	ret = wait_event_interruptible_locked(priv->pending_crtcs_event,
+			!(priv->pending_crtcs & crtc_mask));
+	if (ret == 0) {
+		DBG("start: %08x", crtc_mask);
+		priv->pending_crtcs |= crtc_mask;
+	}
+	spin_unlock(&priv->pending_crtcs_event.lock);
+
+	return ret;
+}
+
+/* clear specified crtcs (no longer pending update)
+ */
+static void end_atomic(struct msm_drm_private *priv, uint32_t crtc_mask)
+{
+	spin_lock(&priv->pending_crtcs_event.lock);
+	DBG("end: %08x", crtc_mask);
+	priv->pending_crtcs &= ~crtc_mask;
+	wake_up_all_locked(&priv->pending_crtcs_event);
+	spin_unlock(&priv->pending_crtcs_event.lock);
+}
+
+static struct msm_commit *new_commit(struct drm_atomic_state *state)
+{
+	struct msm_commit *c = kzalloc(sizeof(*c), GFP_KERNEL);
+
+	if (!c)
+		return NULL;
+
+	c->state = state;
+	/* TODO we might need a way to indicate to run the cb on a
+	 * different wq so wait_for_vblanks() doesn't block retiring
+	 * bo's..
+	 */
+	INIT_FENCE_CB(&c->fence_cb, fence_cb);
+
+	return c;
+}
+
+/* The (potentially) asynchronous part of the commit.  At this point
+ * nothing can fail short of armageddon.
+ */
+static void complete_commit(struct msm_commit *c)
+{
+	struct drm_atomic_state *state = c->state;
+	struct drm_device *dev = state->dev;
+
+	drm_atomic_helper_commit_pre_planes(dev, state);
+
+	drm_atomic_helper_commit_planes(dev, state);
+
+	drm_atomic_helper_commit_post_planes(dev, state);
+
+	/* NOTE: _wait_for_vblanks() only waits for vblank on
+	 * enabled CRTCs.  So we end up faulting when disabling
+	 * due to (potentially) unref'ing the outgoing fb's
+	 * before the vblank when the disable has latched.
+	 *
+	 * But if it did wait on disabled (or newly disabled)
+	 * CRTCs, that would be racy (ie. we could have missed
+	 * the irq.  We need some way to poll for pipe shut
+	 * down.  Or just live with occasionally hitting the
+	 * timeout in the CRTC disable path (which really should
+	 * not be critical path)
+	 */
+
+	drm_atomic_helper_wait_for_vblanks(dev, state);
+
+	drm_atomic_helper_cleanup_planes(dev, state);
+
+	drm_atomic_state_free(state);
+
+	end_atomic(dev->dev_private, c->crtc_mask);
+
+	kfree(c);
+}
+
+static void fence_cb(struct msm_fence_cb *cb)
+{
+	struct msm_commit *c =
+			container_of(cb, struct msm_commit, fence_cb);
+	complete_commit(c);
+}
+
+static void add_fb(struct msm_commit *c, struct drm_framebuffer *fb)
+{
+	struct drm_gem_object *obj = msm_framebuffer_bo(fb, 0);
+	c->fence = max(c->fence, msm_gem_fence(to_msm_bo(obj), MSM_PREP_READ));
+}
+
+
+/**
+ * drm_atomic_helper_commit - commit validated state object
+ * @dev: DRM device
+ * @state: the driver state object
+ * @async: asynchronous commit
+ *
+ * This function commits a with drm_atomic_helper_check() pre-validated state
+ * object. This can still fail when e.g. the framebuffer reservation fails. For
+ * now this doesn't implement asynchronous commits.
+ *
+ * RETURNS
+ * Zero for success or -errno.
+ */
+int msm_atomic_commit(struct drm_device *dev,
+		struct drm_atomic_state *state, bool async)
+{
+	int nplanes = dev->mode_config.num_total_plane;
+	int ncrtcs = dev->mode_config.num_crtc;
+	struct msm_commit *c;
+	int i, ret;
+
+	ret = drm_atomic_helper_prepare_planes(dev, state);
+	if (ret)
+		return ret;
+
+	c = new_commit(state);
+	if (!c)
+		return -ENOMEM;
+
+	/*
+	 * Figure out what crtcs we have:
+	 */
+	for (i = 0; i < ncrtcs; i++) {
+		struct drm_crtc *crtc = state->crtcs[i];
+		if (!crtc)
+			continue;
+		c->crtc_mask |= (1 << drm_crtc_index(crtc));
+	}
+
+	/*
+	 * Figure out what fence to wait for:
+	 */
+	for (i = 0; i < nplanes; i++) {
+		struct drm_plane *plane = state->planes[i];
+		struct drm_plane_state *new_state = state->plane_states[i];
+
+		if (!plane)
+			continue;
+
+		if ((plane->state->fb != new_state->fb) && new_state->fb)
+			add_fb(c, new_state->fb);
+	}
+
+	/*
+	 * Wait for pending updates on any of the same crtc's and then
+	 * mark our set of crtc's as busy:
+	 */
+	ret = start_atomic(dev->dev_private, c->crtc_mask);
+	if (ret)
+		return ret;
+
+	/*
+	 * This is the point of no return - everything below never fails except
+	 * when the hw goes bonghits. Which means we can commit the new state on
+	 * the software side now.
+	 */
+
+	drm_atomic_helper_swap_state(dev, state);
+
+	/*
+	 * Everything below can be run asynchronously without the need to grab
+	 * any modeset locks at all under one conditions: It must be guaranteed
+	 * that the asynchronous work has either been cancelled (if the driver
+	 * supports it, which at least requires that the framebuffers get
+	 * cleaned up with drm_atomic_helper_cleanup_planes()) or completed
+	 * before the new state gets committed on the software side with
+	 * drm_atomic_helper_swap_state().
+	 *
+	 * This scheme allows new atomic state updates to be prepared and
+	 * checked in parallel to the asynchronous completion of the previous
+	 * update. Which is important since compositors need to figure out the
+	 * composition of the next frame right after having submitted the
+	 * current layout.
+	 */
+
+	if (async) {
+		msm_queue_fence_cb(dev, &c->fence_cb, c->fence);
+		return 0;
+	}
+
+	ret = msm_wait_fence_interruptable(dev, c->fence, NULL);
+	if (ret) {
+		WARN_ON(ret);  // TODO unswap state back?  or??
+		kfree(c);
+		return ret;
+	}
+
+	complete_commit(c);
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c
index fcf95680413d..9a61546a0b05 100644
--- a/drivers/gpu/drm/msm/msm_drv.c
+++ b/drivers/gpu/drm/msm/msm_drv.c
@@ -29,6 +29,8 @@ static void msm_fb_output_poll_changed(struct drm_device *dev)
 static const struct drm_mode_config_funcs mode_config_funcs = {
 	.fb_create = msm_framebuffer_create,
 	.output_poll_changed = msm_fb_output_poll_changed,
+	.atomic_check = drm_atomic_helper_check,
+	.atomic_commit = msm_atomic_commit,
 };
 
 int msm_register_mmu(struct drm_device *dev, struct msm_mmu *mmu)
@@ -191,6 +193,7 @@ static int msm_load(struct drm_device *dev, unsigned long flags)
 
 	priv->wq = alloc_ordered_workqueue("msm", 0);
 	init_waitqueue_head(&priv->fence_event);
+	init_waitqueue_head(&priv->pending_crtcs_event);
 
 	INIT_LIST_HEAD(&priv->inactive_list);
 	INIT_LIST_HEAD(&priv->fence_cbs);
@@ -280,7 +283,7 @@ static int msm_load(struct drm_device *dev, unsigned long flags)
 	dev->mode_config.max_height = 2048;
 	dev->mode_config.funcs = &mode_config_funcs;
 
-	ret = drm_vblank_init(dev, 1);
+	ret = drm_vblank_init(dev, priv->num_crtcs);
 	if (ret < 0) {
 		dev_err(dev->dev, "failed to initialize vblank\n");
 		goto fail;
@@ -294,6 +297,8 @@ static int msm_load(struct drm_device *dev, unsigned long flags)
 		goto fail;
 	}
 
+	drm_mode_config_reset(dev);
+
 #ifdef CONFIG_DRM_MSM_FBDEV
 	priv->fbdev = msm_fbdev_init(dev);
 #endif
@@ -315,39 +320,12 @@ static void load_gpu(struct drm_device *dev)
 {
 	static DEFINE_MUTEX(init_lock);
 	struct msm_drm_private *priv = dev->dev_private;
-	struct msm_gpu *gpu;
 
 	mutex_lock(&init_lock);
 
-	if (priv->gpu)
-		goto out;
-
-	gpu = a3xx_gpu_init(dev);
-	if (IS_ERR(gpu)) {
-		dev_warn(dev->dev, "failed to load a3xx gpu\n");
-		gpu = NULL;
-		/* not fatal */
-	}
-
-	if (gpu) {
-		int ret;
-		mutex_lock(&dev->struct_mutex);
-		gpu->funcs->pm_resume(gpu);
-		mutex_unlock(&dev->struct_mutex);
-		ret = gpu->funcs->hw_init(gpu);
-		if (ret) {
-			dev_err(dev->dev, "gpu hw init failed: %d\n", ret);
-			gpu->funcs->destroy(gpu);
-			gpu = NULL;
-		} else {
-			/* give inactive pm a chance to kick in: */
-			msm_gpu_retire(gpu);
-		}
-	}
-
-	priv->gpu = gpu;
+	if (!priv->gpu)
+		priv->gpu = adreno_load_gpu(dev);
 
-out:
 	mutex_unlock(&init_lock);
 }
 
@@ -646,6 +624,26 @@ int msm_wait_fence_interruptable(struct drm_device *dev, uint32_t fence,
 	return ret;
 }
 
+int msm_queue_fence_cb(struct drm_device *dev,
+		struct msm_fence_cb *cb, uint32_t fence)
+{
+	struct msm_drm_private *priv = dev->dev_private;
+	int ret = 0;
+
+	mutex_lock(&dev->struct_mutex);
+	if (!list_empty(&cb->work.entry)) {
+		ret = -EINVAL;
+	} else if (fence > priv->completed_fence) {
+		cb->fence = fence;
+		list_add_tail(&cb->work.entry, &priv->fence_cbs);
+	} else {
+		queue_work(priv->wq, &cb->work);
+	}
+	mutex_unlock(&dev->struct_mutex);
+
+	return ret;
+}
+
 /* called from workqueue */
 void msm_update_fence(struct drm_device *dev, uint32_t fence)
 {
@@ -836,6 +834,7 @@ static struct drm_driver msm_driver = {
 	.open               = msm_open,
 	.preclose           = msm_preclose,
 	.lastclose          = msm_lastclose,
+	.set_busid          = drm_platform_set_busid,
 	.irq_handler        = msm_irq,
 	.irq_preinstall     = msm_irq_preinstall,
 	.irq_postinstall    = msm_irq_postinstall,
@@ -858,6 +857,7 @@ static struct drm_driver msm_driver = {
 	.gem_prime_import_sg_table = msm_gem_prime_import_sg_table,
 	.gem_prime_vmap     = msm_gem_prime_vmap,
 	.gem_prime_vunmap   = msm_gem_prime_vunmap,
+	.gem_prime_mmap     = msm_gem_prime_mmap,
 #ifdef CONFIG_DEBUG_FS
 	.debugfs_init       = msm_debugfs_init,
 	.debugfs_cleanup    = msm_debugfs_cleanup,
@@ -1013,7 +1013,6 @@ static struct platform_driver msm_platform_driver = {
 	.probe      = msm_pdev_probe,
 	.remove     = msm_pdev_remove,
 	.driver     = {
-		.owner  = THIS_MODULE,
 		.name   = "msm",
 		.of_match_table = dt_match,
 		.pm     = &msm_pm_ops,
@@ -1025,7 +1024,7 @@ static int __init msm_drm_register(void)
 {
 	DBG("init");
 	hdmi_register();
-	a3xx_register();
+	adreno_register();
 	return platform_driver_register(&msm_platform_driver);
 }
 
@@ -1034,7 +1033,7 @@ static void __exit msm_drm_unregister(void)
 	DBG("fini");
 	platform_driver_unregister(&msm_platform_driver);
 	hdmi_unregister();
-	a3xx_unregister();
+	adreno_unregister();
 }
 
 module_init(msm_drm_register);
diff --git a/drivers/gpu/drm/msm/msm_drv.h b/drivers/gpu/drm/msm/msm_drv.h
index 8a2c5fd0893e..b69ef2d5a26c 100644
--- a/drivers/gpu/drm/msm/msm_drv.h
+++ b/drivers/gpu/drm/msm/msm_drv.h
@@ -32,15 +32,6 @@
 #include <linux/types.h>
 #include <asm/sizes.h>
 
-
-#if defined(CONFIG_COMPILE_TEST) && !defined(CONFIG_ARCH_QCOM)
-/* stubs we need for compile-test: */
-static inline struct device *msm_iommu_get_ctx(const char *ctx_name)
-{
-	return NULL;
-}
-#endif
-
 #ifndef CONFIG_OF
 #include <mach/board.h>
 #include <mach/socinfo.h>
@@ -48,9 +39,13 @@ static inline struct device *msm_iommu_get_ctx(const char *ctx_name)
 #endif
 
 #include <drm/drmP.h>
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_plane_helper.h>
 #include <drm/drm_fb_helper.h>
 #include <drm/msm_drm.h>
+#include <drm/drm_gem.h>
 
 struct msm_kms;
 struct msm_gpu;
@@ -74,7 +69,12 @@ struct msm_drm_private {
 	struct msm_kms *kms;
 
 	/* subordinate devices, if present: */
-	struct platform_device *hdmi_pdev, *gpu_pdev;
+	struct platform_device *gpu_pdev;
+
+	/* possibly this should be in the kms component, but it is
+	 * shared by both mdp4 and mdp5..
+	 */
+	struct hdmi *hdmi;
 
 	/* when we have more than one 'msm_gpu' these need to be an array: */
 	struct msm_gpu *gpu;
@@ -96,6 +96,10 @@ struct msm_drm_private {
 	/* callbacks deferred until bo is inactive: */
 	struct list_head fence_cbs;
 
+	/* crtcs pending async atomic updates: */
+	uint32_t pending_crtcs;
+	wait_queue_head_t pending_crtcs_event;
+
 	/* registered MMUs: */
 	unsigned int num_mmus;
 	struct msm_mmu *mmus[NUM_DOMAINS];
@@ -144,21 +148,29 @@ void __msm_fence_worker(struct work_struct *work);
 		(_cb)->func = _func;                         \
 	} while (0)
 
+int msm_atomic_commit(struct drm_device *dev,
+		struct drm_atomic_state *state, bool async);
+
 int msm_register_mmu(struct drm_device *dev, struct msm_mmu *mmu);
 
 int msm_wait_fence_interruptable(struct drm_device *dev, uint32_t fence,
 		struct timespec *timeout);
+int msm_queue_fence_cb(struct drm_device *dev,
+		struct msm_fence_cb *cb, uint32_t fence);
 void msm_update_fence(struct drm_device *dev, uint32_t fence);
 
 int msm_ioctl_gem_submit(struct drm_device *dev, void *data,
 		struct drm_file *file);
 
+int msm_gem_mmap_obj(struct drm_gem_object *obj,
+			struct vm_area_struct *vma);
 int msm_gem_mmap(struct file *filp, struct vm_area_struct *vma);
 int msm_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
 uint64_t msm_gem_mmap_offset(struct drm_gem_object *obj);
 int msm_gem_get_iova_locked(struct drm_gem_object *obj, int id,
 		uint32_t *iova);
 int msm_gem_get_iova(struct drm_gem_object *obj, int id, uint32_t *iova);
+uint32_t msm_gem_iova(struct drm_gem_object *obj, int id);
 struct page **msm_gem_get_pages(struct drm_gem_object *obj);
 void msm_gem_put_pages(struct drm_gem_object *obj);
 void msm_gem_put_iova(struct drm_gem_object *obj, int id);
@@ -169,8 +181,9 @@ int msm_gem_dumb_map_offset(struct drm_file *file, struct drm_device *dev,
 struct sg_table *msm_gem_prime_get_sg_table(struct drm_gem_object *obj);
 void *msm_gem_prime_vmap(struct drm_gem_object *obj);
 void msm_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr);
+int msm_gem_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma);
 struct drm_gem_object *msm_gem_prime_import_sg_table(struct drm_device *dev,
-		size_t size, struct sg_table *sg);
+		struct dma_buf_attachment *attach, struct sg_table *sg);
 int msm_gem_prime_pin(struct drm_gem_object *obj);
 void msm_gem_prime_unpin(struct drm_gem_object *obj);
 void *msm_gem_vaddr_locked(struct drm_gem_object *obj);
@@ -191,6 +204,9 @@ struct drm_gem_object *msm_gem_new(struct drm_device *dev,
 struct drm_gem_object *msm_gem_import(struct drm_device *dev,
 		uint32_t size, struct sg_table *sgt);
 
+int msm_framebuffer_prepare(struct drm_framebuffer *fb, int id);
+void msm_framebuffer_cleanup(struct drm_framebuffer *fb, int id);
+uint32_t msm_framebuffer_iova(struct drm_framebuffer *fb, int id, int plane);
 struct drm_gem_object *msm_framebuffer_bo(struct drm_framebuffer *fb, int plane);
 const struct msm_format *msm_framebuffer_format(struct drm_framebuffer *fb);
 struct drm_framebuffer *msm_framebuffer_init(struct drm_device *dev,
@@ -201,8 +217,8 @@ struct drm_framebuffer *msm_framebuffer_create(struct drm_device *dev,
 struct drm_fb_helper *msm_fbdev_init(struct drm_device *dev);
 
 struct hdmi;
-struct hdmi *hdmi_init(struct drm_device *dev, struct drm_encoder *encoder);
-irqreturn_t hdmi_irq(int irq, void *dev_id);
+int hdmi_modeset_init(struct hdmi *hdmi, struct drm_device *dev,
+		struct drm_encoder *encoder);
 void __init hdmi_register(void);
 void __exit hdmi_unregister(void);
 
diff --git a/drivers/gpu/drm/msm/msm_fb.c b/drivers/gpu/drm/msm/msm_fb.c
index 81bafdf19ab3..84dec161d836 100644
--- a/drivers/gpu/drm/msm/msm_fb.c
+++ b/drivers/gpu/drm/msm/msm_fb.c
@@ -24,7 +24,7 @@
 struct msm_framebuffer {
 	struct drm_framebuffer base;
 	const struct msm_format *format;
-	struct drm_gem_object *planes[2];
+	struct drm_gem_object *planes[3];
 };
 #define to_msm_framebuffer(x) container_of(x, struct msm_framebuffer, base)
 
@@ -87,6 +87,44 @@ void msm_framebuffer_describe(struct drm_framebuffer *fb, struct seq_file *m)
 }
 #endif
 
+/* prepare/pin all the fb's bo's for scanout.  Note that it is not valid
+ * to prepare an fb more multiple different initiator 'id's.  But that
+ * should be fine, since only the scanout (mdpN) side of things needs
+ * this, the gpu doesn't care about fb's.
+ */
+int msm_framebuffer_prepare(struct drm_framebuffer *fb, int id)
+{
+	struct msm_framebuffer *msm_fb = to_msm_framebuffer(fb);
+	int ret, i, n = drm_format_num_planes(fb->pixel_format);
+	uint32_t iova;
+
+	for (i = 0; i < n; i++) {
+		ret = msm_gem_get_iova(msm_fb->planes[i], id, &iova);
+		DBG("FB[%u]: iova[%d]: %08x (%d)", fb->base.id, i, iova, ret);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+void msm_framebuffer_cleanup(struct drm_framebuffer *fb, int id)
+{
+	struct msm_framebuffer *msm_fb = to_msm_framebuffer(fb);
+	int i, n = drm_format_num_planes(fb->pixel_format);
+
+	for (i = 0; i < n; i++)
+		msm_gem_put_iova(msm_fb->planes[i], id);
+}
+
+uint32_t msm_framebuffer_iova(struct drm_framebuffer *fb, int id, int plane)
+{
+	struct msm_framebuffer *msm_fb = to_msm_framebuffer(fb);
+	if (!msm_fb->planes[plane])
+		return 0;
+	return msm_gem_iova(msm_fb->planes[plane], id);
+}
+
 struct drm_gem_object *msm_framebuffer_bo(struct drm_framebuffer *fb, int plane)
 {
 	struct msm_framebuffer *msm_fb = to_msm_framebuffer(fb);
@@ -166,6 +204,11 @@ struct drm_framebuffer *msm_framebuffer_init(struct drm_device *dev,
 
 	msm_fb->format = format;
 
+	if (n > ARRAY_SIZE(msm_fb->planes)) {
+		ret = -EINVAL;
+		goto fail;
+	}
+
 	for (i = 0; i < n; i++) {
 		unsigned int width = mode_cmd->width / (i ? hsub : 1);
 		unsigned int height = mode_cmd->height / (i ? vsub : 1);
diff --git a/drivers/gpu/drm/msm/msm_fbdev.c b/drivers/gpu/drm/msm/msm_fbdev.c
index ab5bfd2d0ebf..1f3af13ccede 100644
--- a/drivers/gpu/drm/msm/msm_fbdev.c
+++ b/drivers/gpu/drm/msm/msm_fbdev.c
@@ -93,9 +93,6 @@ static int msm_fbdev_create(struct drm_fb_helper *helper,
 	uint32_t paddr;
 	int ret, size;
 
-	sizes->surface_bpp = 32;
-	sizes->surface_depth = 24;
-
 	DBG("create fbdev: %dx%d@%d (%dx%d)", sizes->surface_width,
 			sizes->surface_height, sizes->surface_bpp,
 			sizes->fb_width, sizes->fb_height);
@@ -193,8 +190,7 @@ fail_unlock:
 fail:
 
 	if (ret) {
-		if (fbi)
-			framebuffer_release(fbi);
+		framebuffer_release(fbi);
 		if (fb) {
 			drm_framebuffer_unregister_private(fb);
 			drm_framebuffer_remove(fb);
diff --git a/drivers/gpu/drm/msm/msm_gem.c b/drivers/gpu/drm/msm/msm_gem.c
index 4b1b82adabde..49dea4fb55ac 100644
--- a/drivers/gpu/drm/msm/msm_gem.c
+++ b/drivers/gpu/drm/msm/msm_gem.c
@@ -309,6 +309,7 @@ int msm_gem_get_iova_locked(struct drm_gem_object *obj, int id,
 	return ret;
 }
 
+/* get iova, taking a reference.  Should have a matching put */
 int msm_gem_get_iova(struct drm_gem_object *obj, int id, uint32_t *iova)
 {
 	struct msm_gem_object *msm_obj = to_msm_bo(obj);
@@ -328,6 +329,16 @@ int msm_gem_get_iova(struct drm_gem_object *obj, int id, uint32_t *iova)
 	return ret;
 }
 
+/* get iova without taking a reference, used in places where you have
+ * already done a 'msm_gem_get_iova()'.
+ */
+uint32_t msm_gem_iova(struct drm_gem_object *obj, int id)
+{
+	struct msm_gem_object *msm_obj = to_msm_bo(obj);
+	WARN_ON(!msm_obj->domain[id].iova);
+	return msm_obj->domain[id].iova;
+}
+
 void msm_gem_put_iova(struct drm_gem_object *obj, int id)
 {
 	// XXX TODO ..
@@ -397,23 +408,10 @@ void *msm_gem_vaddr(struct drm_gem_object *obj)
 int msm_gem_queue_inactive_cb(struct drm_gem_object *obj,
 		struct msm_fence_cb *cb)
 {
-	struct drm_device *dev = obj->dev;
-	struct msm_drm_private *priv = dev->dev_private;
 	struct msm_gem_object *msm_obj = to_msm_bo(obj);
-	int ret = 0;
-
-	mutex_lock(&dev->struct_mutex);
-	if (!list_empty(&cb->work.entry)) {
-		ret = -EINVAL;
-	} else if (is_active(msm_obj)) {
-		cb->fence = max(msm_obj->read_fence, msm_obj->write_fence);
-		list_add_tail(&cb->work.entry, &priv->fence_cbs);
-	} else {
-		queue_work(priv->wq, &cb->work);
-	}
-	mutex_unlock(&dev->struct_mutex);
-
-	return ret;
+	uint32_t fence = msm_gem_fence(msm_obj,
+			MSM_PREP_READ | MSM_PREP_WRITE);
+	return msm_queue_fence_cb(obj->dev, cb, fence);
 }
 
 void msm_gem_move_to_active(struct drm_gem_object *obj,
@@ -452,12 +450,8 @@ int msm_gem_cpu_prep(struct drm_gem_object *obj, uint32_t op,
 	int ret = 0;
 
 	if (is_active(msm_obj)) {
-		uint32_t fence = 0;
+		uint32_t fence = msm_gem_fence(msm_obj, op);
 
-		if (op & MSM_PREP_READ)
-			fence = msm_obj->write_fence;
-		if (op & MSM_PREP_WRITE)
-			fence = max(fence, msm_obj->read_fence);
 		if (op & MSM_PREP_NOSYNC)
 			timeout = NULL;
 
@@ -525,13 +519,11 @@ void msm_gem_free_object(struct drm_gem_object *obj)
 	for (id = 0; id < ARRAY_SIZE(msm_obj->domain); id++) {
 		struct msm_mmu *mmu = priv->mmus[id];
 		if (mmu && msm_obj->domain[id].iova) {
-			uint32_t offset = (uint32_t)mmap_offset(obj);
+			uint32_t offset = msm_obj->domain[id].iova;
 			mmu->funcs->unmap(mmu, offset, msm_obj->sgt, obj->size);
 		}
 	}
 
-	drm_gem_free_mmap_offset(obj);
-
 	if (obj->import_attach) {
 		if (msm_obj->vaddr)
 			dma_buf_vunmap(obj->import_attach->dmabuf, msm_obj->vaddr);
@@ -543,8 +535,7 @@ void msm_gem_free_object(struct drm_gem_object *obj)
 			drm_free_large(msm_obj->pages);
 
 	} else {
-		if (msm_obj->vaddr)
-			vunmap(msm_obj->vaddr);
+		vunmap(msm_obj->vaddr);
 		put_pages(obj);
 	}
 
diff --git a/drivers/gpu/drm/msm/msm_gem.h b/drivers/gpu/drm/msm/msm_gem.h
index bfb052688f8e..8fbbd0594c46 100644
--- a/drivers/gpu/drm/msm/msm_gem.h
+++ b/drivers/gpu/drm/msm/msm_gem.h
@@ -70,6 +70,19 @@ static inline bool is_active(struct msm_gem_object *msm_obj)
 	return msm_obj->gpu != NULL;
 }
 
+static inline uint32_t msm_gem_fence(struct msm_gem_object *msm_obj,
+		uint32_t op)
+{
+	uint32_t fence = 0;
+
+	if (op & MSM_PREP_READ)
+		fence = msm_obj->write_fence;
+	if (op & MSM_PREP_WRITE)
+		fence = max(fence, msm_obj->read_fence);
+
+	return fence;
+}
+
 #define MAX_CMDS 4
 
 /* Created per submit-ioctl, to track bo's and cmdstream bufs, etc,
diff --git a/drivers/gpu/drm/msm/msm_gem_prime.c b/drivers/gpu/drm/msm/msm_gem_prime.c
index d48f9fc5129b..dd7a7ab603e2 100644
--- a/drivers/gpu/drm/msm/msm_gem_prime.c
+++ b/drivers/gpu/drm/msm/msm_gem_prime.c
@@ -18,6 +18,7 @@
 #include "msm_drv.h"
 #include "msm_gem.h"
 
+#include <linux/dma-buf.h>
 
 struct sg_table *msm_gem_prime_get_sg_table(struct drm_gem_object *obj)
 {
@@ -36,10 +37,23 @@ void msm_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr)
 	/* TODO msm_gem_vunmap() */
 }
 
+int msm_gem_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma)
+{
+	int ret;
+
+	mutex_lock(&obj->dev->struct_mutex);
+	ret = drm_gem_mmap_obj(obj, obj->size, vma);
+	mutex_unlock(&obj->dev->struct_mutex);
+	if (ret < 0)
+		return ret;
+
+	return msm_gem_mmap_obj(vma->vm_private_data, vma);
+}
+
 struct drm_gem_object *msm_gem_prime_import_sg_table(struct drm_device *dev,
-		size_t size, struct sg_table *sg)
+		struct dma_buf_attachment *attach, struct sg_table *sg)
 {
-	return msm_gem_import(dev, size, sg);
+	return msm_gem_import(dev, attach->dmabuf->size, sg);
 }
 
 int msm_gem_prime_pin(struct drm_gem_object *obj)
diff --git a/drivers/gpu/drm/msm/msm_gpu.h b/drivers/gpu/drm/msm/msm_gpu.h
index 9b579b792840..fd1e4b4a6d40 100644
--- a/drivers/gpu/drm/msm/msm_gpu.h
+++ b/drivers/gpu/drm/msm/msm_gpu.h
@@ -166,8 +166,8 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		const char *name, const char *ioname, const char *irqname, int ringsz);
 void msm_gpu_cleanup(struct msm_gpu *gpu);
 
-struct msm_gpu *a3xx_gpu_init(struct drm_device *dev);
-void __init a3xx_register(void);
-void __exit a3xx_unregister(void);
+struct msm_gpu *adreno_load_gpu(struct drm_device *dev);
+void __init adreno_register(void);
+void __exit adreno_unregister(void);
 
 #endif /* __MSM_GPU_H__ */
diff --git a/drivers/gpu/drm/nouveau/Makefile b/drivers/gpu/drm/nouveau/Makefile
index f5d7f7ce4bc6..6461e3565afe 100644
--- a/drivers/gpu/drm/nouveau/Makefile
+++ b/drivers/gpu/drm/nouveau/Makefile
@@ -38,19 +38,33 @@ nouveau-y += core/subdev/bios/dcb.o
 nouveau-y += core/subdev/bios/disp.o
 nouveau-y += core/subdev/bios/dp.o
 nouveau-y += core/subdev/bios/extdev.o
+nouveau-y += core/subdev/bios/fan.o
 nouveau-y += core/subdev/bios/gpio.o
 nouveau-y += core/subdev/bios/i2c.o
+nouveau-y += core/subdev/bios/image.o
 nouveau-y += core/subdev/bios/init.o
 nouveau-y += core/subdev/bios/mxm.o
+nouveau-y += core/subdev/bios/npde.o
+nouveau-y += core/subdev/bios/pcir.o
 nouveau-y += core/subdev/bios/perf.o
 nouveau-y += core/subdev/bios/pll.o
+nouveau-y += core/subdev/bios/pmu.o
 nouveau-y += core/subdev/bios/ramcfg.o
 nouveau-y += core/subdev/bios/rammap.o
+nouveau-y += core/subdev/bios/shadow.o
+nouveau-y += core/subdev/bios/shadowacpi.o
+nouveau-y += core/subdev/bios/shadowof.o
+nouveau-y += core/subdev/bios/shadowpci.o
+nouveau-y += core/subdev/bios/shadowramin.o
+nouveau-y += core/subdev/bios/shadowrom.o
 nouveau-y += core/subdev/bios/timing.o
 nouveau-y += core/subdev/bios/therm.o
 nouveau-y += core/subdev/bios/vmap.o
 nouveau-y += core/subdev/bios/volt.o
 nouveau-y += core/subdev/bios/xpio.o
+nouveau-y += core/subdev/bios/M0203.o
+nouveau-y += core/subdev/bios/M0205.o
+nouveau-y += core/subdev/bios/M0209.o
 nouveau-y += core/subdev/bios/P0260.o
 nouveau-y += core/subdev/bus/hwsq.o
 nouveau-y += core/subdev/bus/nv04.o
@@ -83,6 +97,7 @@ nouveau-y += core/subdev/devinit/nva3.o
 nouveau-y += core/subdev/devinit/nvaf.o
 nouveau-y += core/subdev/devinit/nvc0.o
 nouveau-y += core/subdev/devinit/gm107.o
+nouveau-y += core/subdev/devinit/gm204.o
 nouveau-y += core/subdev/fb/base.o
 nouveau-y += core/subdev/fb/nv04.o
 nouveau-y += core/subdev/fb/nv10.o
@@ -124,12 +139,18 @@ nouveau-y += core/subdev/fb/ramnvc0.o
 nouveau-y += core/subdev/fb/ramnve0.o
 nouveau-y += core/subdev/fb/ramgk20a.o
 nouveau-y += core/subdev/fb/ramgm107.o
+nouveau-y += core/subdev/fb/sddr2.o
 nouveau-y += core/subdev/fb/sddr3.o
+nouveau-y += core/subdev/fb/gddr3.o
 nouveau-y += core/subdev/fb/gddr5.o
+nouveau-y += core/subdev/fuse/base.o
+nouveau-y += core/subdev/fuse/g80.o
+nouveau-y += core/subdev/fuse/gf100.o
+nouveau-y += core/subdev/fuse/gm107.o
 nouveau-y += core/subdev/gpio/base.o
 nouveau-y += core/subdev/gpio/nv10.o
 nouveau-y += core/subdev/gpio/nv50.o
-nouveau-y += core/subdev/gpio/nv92.o
+nouveau-y += core/subdev/gpio/nv94.o
 nouveau-y += core/subdev/gpio/nvd0.o
 nouveau-y += core/subdev/gpio/nve0.o
 nouveau-y += core/subdev/i2c/base.o
@@ -139,6 +160,7 @@ nouveau-y += core/subdev/i2c/bit.o
 nouveau-y += core/subdev/i2c/pad.o
 nouveau-y += core/subdev/i2c/padnv04.o
 nouveau-y += core/subdev/i2c/padnv94.o
+nouveau-y += core/subdev/i2c/padgm204.o
 nouveau-y += core/subdev/i2c/nv04.o
 nouveau-y += core/subdev/i2c/nv4e.o
 nouveau-y += core/subdev/i2c/nv50.o
@@ -146,6 +168,7 @@ nouveau-y += core/subdev/i2c/nv94.o
 nouveau-y += core/subdev/i2c/nvd0.o
 nouveau-y += core/subdev/i2c/gf117.o
 nouveau-y += core/subdev/i2c/nve0.o
+nouveau-y += core/subdev/i2c/gm204.o
 nouveau-y += core/subdev/ibus/nvc0.o
 nouveau-y += core/subdev/ibus/nve0.o
 nouveau-y += core/subdev/ibus/gk20a.o
@@ -190,6 +213,7 @@ nouveau-y += core/subdev/therm/nv50.o
 nouveau-y += core/subdev/therm/nv84.o
 nouveau-y += core/subdev/therm/nva3.o
 nouveau-y += core/subdev/therm/nvd0.o
+nouveau-y += core/subdev/therm/gm107.o
 nouveau-y += core/subdev/timer/base.o
 nouveau-y += core/subdev/timer/nv04.o
 nouveau-y += core/subdev/timer/gk20a.o
@@ -202,6 +226,7 @@ nouveau-y += core/subdev/vm/nvc0.o
 nouveau-y += core/subdev/volt/base.o
 nouveau-y += core/subdev/volt/gpio.o
 nouveau-y += core/subdev/volt/nv40.o
+nouveau-y += core/subdev/volt/gk20a.o
 
 nouveau-y += core/engine/falcon.o
 nouveau-y += core/engine/xtensa.o
@@ -245,6 +270,7 @@ nouveau-y += core/engine/disp/nvd0.o
 nouveau-y += core/engine/disp/nve0.o
 nouveau-y += core/engine/disp/nvf0.o
 nouveau-y += core/engine/disp/gm107.o
+nouveau-y += core/engine/disp/gm204.o
 nouveau-y += core/engine/disp/dacnv50.o
 nouveau-y += core/engine/disp/dport.o
 nouveau-y += core/engine/disp/hdanva3.o
@@ -252,10 +278,12 @@ nouveau-y += core/engine/disp/hdanvd0.o
 nouveau-y += core/engine/disp/hdminv84.o
 nouveau-y += core/engine/disp/hdminva3.o
 nouveau-y += core/engine/disp/hdminvd0.o
+nouveau-y += core/engine/disp/hdminve0.o
 nouveau-y += core/engine/disp/piornv50.o
 nouveau-y += core/engine/disp/sornv50.o
 nouveau-y += core/engine/disp/sornv94.o
 nouveau-y += core/engine/disp/sornvd0.o
+nouveau-y += core/engine/disp/sorgm204.o
 nouveau-y += core/engine/disp/vga.o
 nouveau-y += core/engine/fifo/base.o
 nouveau-y += core/engine/fifo/nv04.o
diff --git a/drivers/gpu/drm/nouveau/core/core/client.c b/drivers/gpu/drm/nouveau/core/core/client.c
index 68bf06768123..e962433294c3 100644
--- a/drivers/gpu/drm/nouveau/core/core/client.c
+++ b/drivers/gpu/drm/nouveau/core/core/client.c
@@ -91,9 +91,10 @@ nvkm_client_notify_del(struct nouveau_client *client, int index)
 }
 
 int
-nvkm_client_notify_new(struct nouveau_client *client,
+nvkm_client_notify_new(struct nouveau_object *object,
 		       struct nvkm_event *event, void *data, u32 size)
 {
+	struct nouveau_client *client = nouveau_client(object);
 	struct nvkm_client_notify *notify;
 	union {
 		struct nvif_notify_req_v0 v0;
@@ -127,8 +128,8 @@ nvkm_client_notify_new(struct nouveau_client *client,
 	}
 
 	if (ret == 0) {
-		ret = nvkm_notify_init(event, nvkm_client_notify, false,
-				       data, size, reply, &notify->n);
+		ret = nvkm_notify_init(object, event, nvkm_client_notify,
+				       false, data, size, reply, &notify->n);
 		if (ret == 0) {
 			client->notify[index] = notify;
 			notify->client = client;
diff --git a/drivers/gpu/drm/nouveau/core/core/event.c b/drivers/gpu/drm/nouveau/core/core/event.c
index 0540a48c5678..760947e380c9 100644
--- a/drivers/gpu/drm/nouveau/core/core/event.c
+++ b/drivers/gpu/drm/nouveau/core/core/event.c
@@ -20,13 +20,13 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 
-#include <core/os.h>
+#include <core/object.h>
 #include <core/event.h>
 
 void
 nvkm_event_put(struct nvkm_event *event, u32 types, int index)
 {
-	BUG_ON(!spin_is_locked(&event->refs_lock));
+	assert_spin_locked(&event->refs_lock);
 	while (types) {
 		int type = __ffs(types); types &= ~(1 << type);
 		if (--event->refs[index * event->types_nr + type] == 0) {
@@ -39,7 +39,7 @@ nvkm_event_put(struct nvkm_event *event, u32 types, int index)
 void
 nvkm_event_get(struct nvkm_event *event, u32 types, int index)
 {
-	BUG_ON(!spin_is_locked(&event->refs_lock));
+	assert_spin_locked(&event->refs_lock);
 	while (types) {
 		int type = __ffs(types); types &= ~(1 << type);
 		if (++event->refs[index * event->types_nr + type] == 1) {
diff --git a/drivers/gpu/drm/nouveau/core/core/gpuobj.c b/drivers/gpu/drm/nouveau/core/core/gpuobj.c
index 560b2214cf1c..daee87702502 100644
--- a/drivers/gpu/drm/nouveau/core/core/gpuobj.c
+++ b/drivers/gpu/drm/nouveau/core/core/gpuobj.c
@@ -115,7 +115,7 @@ nouveau_gpuobj_create_(struct nouveau_object *parent,
 	gpuobj->size = size;
 
 	if (heap) {
-		ret = nouveau_mm_head(heap, 1, size, size,
+		ret = nouveau_mm_head(heap, 0, 1, size, size,
 				      max(align, (u32)1), &gpuobj->node);
 		if (ret)
 			return ret;
diff --git a/drivers/gpu/drm/nouveau/core/core/handle.c b/drivers/gpu/drm/nouveau/core/core/handle.c
index a490b805d7e3..13f816cb08bd 100644
--- a/drivers/gpu/drm/nouveau/core/core/handle.c
+++ b/drivers/gpu/drm/nouveau/core/core/handle.c
@@ -222,116 +222,3 @@ nouveau_handle_put(struct nouveau_handle *handle)
 	if (handle)
 		nouveau_namedb_put(handle);
 }
-
-int
-nouveau_handle_new(struct nouveau_object *client, u32 _parent, u32 _handle,
-		   u16 _oclass, void *data, u32 size,
-		   struct nouveau_object **pobject)
-{
-	struct nouveau_object *parent = NULL;
-	struct nouveau_object *engctx = NULL;
-	struct nouveau_object *object = NULL;
-	struct nouveau_object *engine;
-	struct nouveau_oclass *oclass;
-	struct nouveau_handle *handle;
-	int ret;
-
-	/* lookup parent object and ensure it *is* a parent */
-	parent = nouveau_handle_ref(client, _parent);
-	if (!parent) {
-		nv_error(client, "parent 0x%08x not found\n", _parent);
-		return -ENOENT;
-	}
-
-	if (!nv_iclass(parent, NV_PARENT_CLASS)) {
-		nv_error(parent, "cannot have children\n");
-		ret = -EINVAL;
-		goto fail_class;
-	}
-
-	/* check that parent supports the requested subclass */
-	ret = nouveau_parent_sclass(parent, _oclass, &engine, &oclass);
-	if (ret) {
-		nv_debug(parent, "illegal class 0x%04x\n", _oclass);
-		goto fail_class;
-	}
-
-	/* make sure engine init has been completed *before* any objects
-	 * it controls are created - the constructors may depend on
-	 * state calculated at init (ie. default context construction)
-	 */
-	if (engine) {
-		ret = nouveau_object_inc(engine);
-		if (ret)
-			goto fail_class;
-	}
-
-	/* if engine requires it, create a context object to insert
-	 * between the parent and its children (eg. PGRAPH context)
-	 */
-	if (engine && nv_engine(engine)->cclass) {
-		ret = nouveau_object_ctor(parent, engine,
-					  nv_engine(engine)->cclass,
-					  data, size, &engctx);
-		if (ret)
-			goto fail_engctx;
-	} else {
-		nouveau_object_ref(parent, &engctx);
-	}
-
-	/* finally, create new object and bind it to its handle */
-	ret = nouveau_object_ctor(engctx, engine, oclass, data, size, &object);
-	*pobject = object;
-	if (ret)
-		goto fail_ctor;
-
-	ret = nouveau_object_inc(object);
-	if (ret)
-		goto fail_init;
-
-	ret = nouveau_handle_create(parent, _parent, _handle, object, &handle);
-	if (ret)
-		goto fail_handle;
-
-	ret = nouveau_handle_init(handle);
-	if (ret)
-		nouveau_handle_destroy(handle);
-
-fail_handle:
-	nouveau_object_dec(object, false);
-fail_init:
-	nouveau_object_ref(NULL, &object);
-fail_ctor:
-	nouveau_object_ref(NULL, &engctx);
-fail_engctx:
-	if (engine)
-		nouveau_object_dec(engine, false);
-fail_class:
-	nouveau_object_ref(NULL, &parent);
-	return ret;
-}
-
-int
-nouveau_handle_del(struct nouveau_object *client, u32 _parent, u32 _handle)
-{
-	struct nouveau_object *parent = NULL;
-	struct nouveau_object *namedb = NULL;
-	struct nouveau_handle *handle = NULL;
-
-	parent = nouveau_handle_ref(client, _parent);
-	if (!parent)
-		return -ENOENT;
-
-	namedb = nv_pclass(parent, NV_NAMEDB_CLASS);
-	if (namedb) {
-		handle = nouveau_namedb_get(nv_namedb(namedb), _handle);
-		if (handle) {
-			nouveau_namedb_put(handle);
-			nouveau_handle_fini(handle, false);
-			nouveau_handle_destroy(handle);
-		}
-	}
-
-	nouveau_object_ref(NULL, &parent);
-	return handle ? 0 : -EINVAL;
-}
diff --git a/drivers/gpu/drm/nouveau/core/core/ioctl.c b/drivers/gpu/drm/nouveau/core/core/ioctl.c
index f7e19bfb489c..692aa92dd850 100644
--- a/drivers/gpu/drm/nouveau/core/core/ioctl.c
+++ b/drivers/gpu/drm/nouveau/core/core/ioctl.c
@@ -349,7 +349,6 @@ nvkm_ioctl_unmap(struct nouveau_handle *handle, void *data, u32 size)
 static int
 nvkm_ioctl_ntfy_new(struct nouveau_handle *handle, void *data, u32 size)
 {
-	struct nouveau_client *client = nouveau_client(handle->object);
 	struct nouveau_object *object = handle->object;
 	struct nouveau_ofuncs *ofuncs = object->oclass->ofuncs;
 	union {
@@ -365,7 +364,7 @@ nvkm_ioctl_ntfy_new(struct nouveau_handle *handle, void *data, u32 size)
 		if (ret = -ENODEV, ofuncs->ntfy)
 			ret = ofuncs->ntfy(object, args->v0.event, &event);
 		if (ret == 0) {
-			ret = nvkm_client_notify_new(client, event, data, size);
+			ret = nvkm_client_notify_new(object, event, data, size);
 			if (ret >= 0) {
 				args->v0.index = ret;
 				ret = 0;
diff --git a/drivers/gpu/drm/nouveau/core/core/mm.c b/drivers/gpu/drm/nouveau/core/core/mm.c
index 7a4e0891c5f8..b4f5db66d5b5 100644
--- a/drivers/gpu/drm/nouveau/core/core/mm.c
+++ b/drivers/gpu/drm/nouveau/core/core/mm.c
@@ -28,6 +28,24 @@
 #define node(root, dir) ((root)->nl_entry.dir == &mm->nodes) ? NULL : \
 	list_entry((root)->nl_entry.dir, struct nouveau_mm_node, nl_entry)
 
+static void
+nouveau_mm_dump(struct nouveau_mm *mm, const char *header)
+{
+	struct nouveau_mm_node *node;
+
+	printk(KERN_ERR "nouveau: %s\n", header);
+	printk(KERN_ERR "nouveau: node list:\n");
+	list_for_each_entry(node, &mm->nodes, nl_entry) {
+		printk(KERN_ERR "nouveau: \t%08x %08x %d\n",
+		       node->offset, node->length, node->type);
+	}
+	printk(KERN_ERR "nouveau: free list:\n");
+	list_for_each_entry(node, &mm->free, fl_entry) {
+		printk(KERN_ERR "nouveau: \t%08x %08x %d\n",
+		       node->offset, node->length, node->type);
+	}
+}
+
 void
 nouveau_mm_free(struct nouveau_mm *mm, struct nouveau_mm_node **pthis)
 {
@@ -37,29 +55,29 @@ nouveau_mm_free(struct nouveau_mm *mm, struct nouveau_mm_node **pthis)
 		struct nouveau_mm_node *prev = node(this, prev);
 		struct nouveau_mm_node *next = node(this, next);
 
-		if (prev && prev->type == 0) {
+		if (prev && prev->type == NVKM_MM_TYPE_NONE) {
 			prev->length += this->length;
 			list_del(&this->nl_entry);
 			kfree(this); this = prev;
 		}
 
-		if (next && next->type == 0) {
+		if (next && next->type == NVKM_MM_TYPE_NONE) {
 			next->offset  = this->offset;
 			next->length += this->length;
-			if (this->type == 0)
+			if (this->type == NVKM_MM_TYPE_NONE)
 				list_del(&this->fl_entry);
 			list_del(&this->nl_entry);
 			kfree(this); this = NULL;
 		}
 
-		if (this && this->type != 0) {
+		if (this && this->type != NVKM_MM_TYPE_NONE) {
 			list_for_each_entry(prev, &mm->free, fl_entry) {
 				if (this->offset < prev->offset)
 					break;
 			}
 
 			list_add_tail(&this->fl_entry, &prev->fl_entry);
-			this->type = 0;
+			this->type = NVKM_MM_TYPE_NONE;
 		}
 	}
 
@@ -80,27 +98,32 @@ region_head(struct nouveau_mm *mm, struct nouveau_mm_node *a, u32 size)
 
 	b->offset = a->offset;
 	b->length = size;
+	b->heap   = a->heap;
 	b->type   = a->type;
 	a->offset += size;
 	a->length -= size;
 	list_add_tail(&b->nl_entry, &a->nl_entry);
-	if (b->type == 0)
+	if (b->type == NVKM_MM_TYPE_NONE)
 		list_add_tail(&b->fl_entry, &a->fl_entry);
 	return b;
 }
 
 int
-nouveau_mm_head(struct nouveau_mm *mm, u8 type, u32 size_max, u32 size_min,
-		u32 align, struct nouveau_mm_node **pnode)
+nouveau_mm_head(struct nouveau_mm *mm, u8 heap, u8 type, u32 size_max,
+		u32 size_min, u32 align, struct nouveau_mm_node **pnode)
 {
 	struct nouveau_mm_node *prev, *this, *next;
 	u32 mask = align - 1;
 	u32 splitoff;
 	u32 s, e;
 
-	BUG_ON(!type);
+	BUG_ON(type == NVKM_MM_TYPE_NONE || type == NVKM_MM_TYPE_HOLE);
 
 	list_for_each_entry(this, &mm->free, fl_entry) {
+		if (unlikely(heap != NVKM_MM_HEAP_ANY)) {
+			if (this->heap != heap)
+				continue;
+		}
 		e = this->offset + this->length;
 		s = this->offset;
 
@@ -149,27 +172,32 @@ region_tail(struct nouveau_mm *mm, struct nouveau_mm_node *a, u32 size)
 	a->length -= size;
 	b->offset  = a->offset + a->length;
 	b->length  = size;
+	b->heap    = a->heap;
 	b->type    = a->type;
 
 	list_add(&b->nl_entry, &a->nl_entry);
-	if (b->type == 0)
+	if (b->type == NVKM_MM_TYPE_NONE)
 		list_add(&b->fl_entry, &a->fl_entry);
 	return b;
 }
 
 int
-nouveau_mm_tail(struct nouveau_mm *mm, u8 type, u32 size_max, u32 size_min,
-		u32 align, struct nouveau_mm_node **pnode)
+nouveau_mm_tail(struct nouveau_mm *mm, u8 heap, u8 type, u32 size_max,
+		u32 size_min, u32 align, struct nouveau_mm_node **pnode)
 {
 	struct nouveau_mm_node *prev, *this, *next;
 	u32 mask = align - 1;
 
-	BUG_ON(!type);
+	BUG_ON(type == NVKM_MM_TYPE_NONE || type == NVKM_MM_TYPE_HOLE);
 
 	list_for_each_entry_reverse(this, &mm->free, fl_entry) {
 		u32 e = this->offset + this->length;
 		u32 s = this->offset;
 		u32 c = 0, a;
+		if (unlikely(heap != NVKM_MM_HEAP_ANY)) {
+			if (this->heap != heap)
+				continue;
+		}
 
 		prev = node(this, prev);
 		if (prev && prev->type != type)
@@ -209,9 +237,23 @@ nouveau_mm_tail(struct nouveau_mm *mm, u8 type, u32 size_max, u32 size_min,
 int
 nouveau_mm_init(struct nouveau_mm *mm, u32 offset, u32 length, u32 block)
 {
-	struct nouveau_mm_node *node;
+	struct nouveau_mm_node *node, *prev;
+	u32 next;
 
-	if (block) {
+	if (nouveau_mm_initialised(mm)) {
+		prev = list_last_entry(&mm->nodes, typeof(*node), nl_entry);
+		next = prev->offset + prev->length;
+		if (next != offset) {
+			BUG_ON(next > offset);
+			if (!(node = kzalloc(sizeof(*node), GFP_KERNEL)))
+				return -ENOMEM;
+			node->type   = NVKM_MM_TYPE_HOLE;
+			node->offset = next;
+			node->length = offset - next;
+			list_add_tail(&node->nl_entry, &mm->nodes);
+		}
+		BUG_ON(block != mm->block_size);
+	} else {
 		INIT_LIST_HEAD(&mm->nodes);
 		INIT_LIST_HEAD(&mm->free);
 		mm->block_size = block;
@@ -230,25 +272,32 @@ nouveau_mm_init(struct nouveau_mm *mm, u32 offset, u32 length, u32 block)
 
 	list_add_tail(&node->nl_entry, &mm->nodes);
 	list_add_tail(&node->fl_entry, &mm->free);
-	mm->heap_nodes++;
+	node->heap = ++mm->heap_nodes;
 	return 0;
 }
 
 int
 nouveau_mm_fini(struct nouveau_mm *mm)
 {
-	if (nouveau_mm_initialised(mm)) {
-		struct nouveau_mm_node *node, *heap =
-			list_first_entry(&mm->nodes, typeof(*heap), nl_entry);
-		int nodes = 0;
+	struct nouveau_mm_node *node, *temp;
+	int nodes = 0;
 
-		list_for_each_entry(node, &mm->nodes, nl_entry) {
-			if (WARN_ON(nodes++ == mm->heap_nodes))
+	if (!nouveau_mm_initialised(mm))
+		return 0;
+
+	list_for_each_entry(node, &mm->nodes, nl_entry) {
+		if (node->type != NVKM_MM_TYPE_HOLE) {
+			if (++nodes > mm->heap_nodes) {
+				nouveau_mm_dump(mm, "mm not clean!");
 				return -EBUSY;
+			}
 		}
-
-		kfree(heap);
 	}
 
+	list_for_each_entry_safe(node, temp, &mm->nodes, nl_entry) {
+		list_del(&node->nl_entry);
+		kfree(node);
+	}
+	mm->heap_nodes = 0;
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/core/notify.c b/drivers/gpu/drm/nouveau/core/core/notify.c
index 76adb81bdea2..839a32577680 100644
--- a/drivers/gpu/drm/nouveau/core/core/notify.c
+++ b/drivers/gpu/drm/nouveau/core/core/notify.c
@@ -98,7 +98,7 @@ nvkm_notify_send(struct nvkm_notify *notify, void *data, u32 size)
 	struct nvkm_event *event = notify->event;
 	unsigned long flags;
 
-	BUG_ON(!spin_is_locked(&event->list_lock));
+	assert_spin_locked(&event->list_lock);
 	BUG_ON(size != notify->size);
 
 	spin_lock_irqsave(&event->refs_lock, flags);
@@ -134,14 +134,15 @@ nvkm_notify_fini(struct nvkm_notify *notify)
 }
 
 int
-nvkm_notify_init(struct nvkm_event *event, int (*func)(struct nvkm_notify *),
-		 bool work, void *data, u32 size, u32 reply,
+nvkm_notify_init(struct nouveau_object *object, struct nvkm_event *event,
+		 int (*func)(struct nvkm_notify *), bool work,
+		 void *data, u32 size, u32 reply,
 		 struct nvkm_notify *notify)
 {
 	unsigned long flags;
 	int ret = -ENODEV;
 	if ((notify->event = event), event->refs) {
-		ret = event->func->ctor(data, size, notify);
+		ret = event->func->ctor(object, data, size, notify);
 		if (ret == 0 && (ret = -EINVAL, notify->size == reply)) {
 			notify->flags = 0;
 			notify->block = 1;
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/base.c b/drivers/gpu/drm/nouveau/core/engine/device/base.c
index 8928f7981d4a..137e0b0faeae 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/base.c
@@ -29,6 +29,7 @@
 #include <nvif/unpack.h>
 #include <nvif/class.h>
 
+#include <subdev/bios.h>
 #include <subdev/fb.h>
 #include <subdev/instmem.h>
 
@@ -138,7 +139,7 @@ nouveau_devobj_info(struct nouveau_object *object, void *data, u32 size)
 	}
 
 	args->v0.chipset  = device->chipset;
-	args->v0.revision = device->chipset >= 0x10 ? nv_rd32(device, 0) : 0x00;
+	args->v0.revision = device->chiprev;
 	if (pfb)  args->v0.ram_size = args->v0.ram_user = pfb->ram->size;
 	else      args->v0.ram_size = args->v0.ram_user = 0;
 	if (imem) args->v0.ram_user = args->v0.ram_user - imem->reserved;
@@ -222,6 +223,7 @@ static const u64 disable_map[] = {
 	[NVDEV_SUBDEV_VOLT]	= NV_DEVICE_V0_DISABLE_CORE,
 	[NVDEV_SUBDEV_THERM]	= NV_DEVICE_V0_DISABLE_CORE,
 	[NVDEV_SUBDEV_PWR]	= NV_DEVICE_V0_DISABLE_CORE,
+	[NVDEV_SUBDEV_FUSE]	= NV_DEVICE_V0_DISABLE_CORE,
 	[NVDEV_ENGINE_DMAOBJ]	= NV_DEVICE_V0_DISABLE_CORE,
 	[NVDEV_ENGINE_PERFMON]  = NV_DEVICE_V0_DISABLE_CORE,
 	[NVDEV_ENGINE_FIFO]	= NV_DEVICE_V0_DISABLE_FIFO,
@@ -235,6 +237,7 @@ static const u64 disable_map[] = {
 	[NVDEV_ENGINE_PPP]	= NV_DEVICE_V0_DISABLE_PPP,
 	[NVDEV_ENGINE_COPY0]	= NV_DEVICE_V0_DISABLE_COPY0,
 	[NVDEV_ENGINE_COPY1]	= NV_DEVICE_V0_DISABLE_COPY1,
+	[NVDEV_ENGINE_COPY2]	= NV_DEVICE_V0_DISABLE_COPY1,
 	[NVDEV_ENGINE_VIC]	= NV_DEVICE_V0_DISABLE_VIC,
 	[NVDEV_ENGINE_VENC]	= NV_DEVICE_V0_DISABLE_VENC,
 	[NVDEV_ENGINE_DISP]	= NV_DEVICE_V0_DISABLE_DISP,
@@ -352,12 +355,14 @@ nouveau_devobj_ctor(struct nouveau_object *parent,
 		/* determine chipset and derive architecture from it */
 		if ((boot0 & 0x1f000000) > 0) {
 			device->chipset = (boot0 & 0x1ff00000) >> 20;
+			device->chiprev = (boot0 & 0x000000ff);
 			switch (device->chipset & 0x1f0) {
 			case 0x010: {
 				if (0x461 & (1 << (device->chipset & 0xf)))
 					device->card_type = NV_10;
 				else
 					device->card_type = NV_11;
+				device->chiprev = 0x00;
 				break;
 			}
 			case 0x020: device->card_type = NV_20; break;
@@ -373,7 +378,8 @@ nouveau_devobj_ctor(struct nouveau_object *parent,
 			case 0x0e0:
 			case 0x0f0:
 			case 0x100: device->card_type = NV_E0; break;
-			case 0x110: device->card_type = GM100; break;
+			case 0x110:
+			case 0x120: device->card_type = GM100; break;
 			default:
 				break;
 			}
@@ -427,6 +433,10 @@ nouveau_devobj_ctor(struct nouveau_object *parent,
 		}
 
 		nv_debug(device, "crystal freq: %dKHz\n", device->crystal);
+	} else
+	if ( (args->v0.disable & NV_DEVICE_V0_DISABLE_IDENTIFY)) {
+		device->cname = "NULL";
+		device->oclass[NVDEV_SUBDEV_VBIOS] = &nouveau_bios_oclass;
 	}
 
 	if (!(args->v0.disable & NV_DEVICE_V0_DISABLE_MMIO) &&
@@ -505,7 +515,8 @@ nouveau_device_sclass[] = {
 };
 
 static int
-nouveau_device_event_ctor(void *data, u32 size, struct nvkm_notify *notify)
+nouveau_device_event_ctor(struct nouveau_object *object, void *data, u32 size,
+			  struct nvkm_notify *notify)
 {
 	if (!WARN_ON(size != 0)) {
 		notify->size  = 0;
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/gm100.c b/drivers/gpu/drm/nouveau/core/engine/device/gm100.c
index 377ec0b8851e..4e74a3376de8 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/gm100.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/gm100.c
@@ -26,6 +26,7 @@
 #include <subdev/bus.h>
 #include <subdev/gpio.h>
 #include <subdev/i2c.h>
+#include <subdev/fuse.h>
 #include <subdev/clock.h>
 #include <subdev/therm.h>
 #include <subdev/mxm.h>
@@ -62,10 +63,9 @@ gm100_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nvd0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gm107_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
-#if 0
-		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
-#endif
+		device->oclass[NVDEV_SUBDEV_THERM  ] = &gm107_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
 		device->oclass[NVDEV_SUBDEV_DEVINIT] =  gm107_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  gk20a_mc_oclass;
@@ -77,8 +77,9 @@ gm100_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
-#if 0
 		device->oclass[NVDEV_SUBDEV_PWR    ] =  nv108_pwr_oclass;
+
+#if 0
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 #endif
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] =  nvd0_dmaeng_oclass;
@@ -97,6 +98,49 @@ gm100_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
 #endif
 		break;
+	case 0x124:
+		device->cname = "GM204";
+		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  gm204_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gm107_fuse_oclass;
+#if 0
+		/* looks to be some non-trivial changes */
+		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
+		/* priv ring says no to 0x10eb14 writes */
+		device->oclass[NVDEV_SUBDEV_THERM  ] = &gm107_therm_oclass;
+#endif
+		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  gm204_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_MC     ] =  gk20a_mc_oclass;
+		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
+		device->oclass[NVDEV_SUBDEV_TIMER  ] = &gk20a_timer_oclass;
+		device->oclass[NVDEV_SUBDEV_FB     ] =  gm107_fb_oclass;
+		device->oclass[NVDEV_SUBDEV_LTC    ] =  gm107_ltc_oclass;
+		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nve0_ibus_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
+		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
+		device->oclass[NVDEV_SUBDEV_PWR    ] =  nv108_pwr_oclass;
+#if 0
+		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
+#endif
+		device->oclass[NVDEV_ENGINE_DMAOBJ ] =  nvd0_dmaeng_oclass;
+#if 0
+		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv108_fifo_oclass;
+		device->oclass[NVDEV_ENGINE_SW     ] =  nvc0_software_oclass;
+		device->oclass[NVDEV_ENGINE_GR     ] =  gm107_graph_oclass;
+#endif
+		device->oclass[NVDEV_ENGINE_DISP   ] =  gm204_disp_oclass;
+#if 0
+		device->oclass[NVDEV_ENGINE_COPY0  ] = &gm204_copy0_oclass;
+		device->oclass[NVDEV_ENGINE_COPY1  ] = &gm204_copy1_oclass;
+		device->oclass[NVDEV_ENGINE_COPY2  ] = &gm204_copy2_oclass;
+		device->oclass[NVDEV_ENGINE_BSP    ] = &nve0_bsp_oclass;
+		device->oclass[NVDEV_ENGINE_VP     ] = &nve0_vp_oclass;
+		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
+#endif
+		break;
 	default:
 		nv_fatal(device, "unknown Maxwell chipset\n");
 		return -EINVAL;
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nv50.c b/drivers/gpu/drm/nouveau/core/engine/device/nv50.c
index 932f84fae459..96f568d1321b 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nv50.c
@@ -26,6 +26,7 @@
 #include <subdev/bus.h>
 #include <subdev/gpio.h>
 #include <subdev/i2c.h>
+#include <subdev/fuse.h>
 #include <subdev/clock.h>
 #include <subdev/therm.h>
 #include <subdev/mxm.h>
@@ -62,6 +63,7 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv50_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv50_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv50_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv50_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -87,6 +89,7 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv50_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv50_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -115,6 +118,7 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv50_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv50_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -141,8 +145,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0x92:
 		device->cname = "G92";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv50_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv50_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -169,8 +174,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0x94:
 		device->cname = "G94";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -197,8 +203,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0x96:
 		device->cname = "G96";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -225,8 +232,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0x98:
 		device->cname = "G98";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -253,8 +261,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0xa0:
 		device->cname = "G200";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv50_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -281,8 +290,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0xaa:
 		device->cname = "MCP77/MCP78";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nvaa_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -309,8 +319,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0xac:
 		device->cname = "MCP79/MCP7A";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nvaa_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -337,8 +348,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0xa3:
 		device->cname = "GT215";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nva3_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -367,8 +379,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0xa5:
 		device->cname = "GT216";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nva3_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -396,8 +409,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0xa8:
 		device->cname = "GT218";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nva3_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -425,8 +439,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0xaf:
 		device->cname = "MCP89";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nva3_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nvc0.c b/drivers/gpu/drm/nouveau/core/engine/device/nvc0.c
index b4a2917ce555..72a40f95d048 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nvc0.c
@@ -26,6 +26,7 @@
 #include <subdev/bus.h>
 #include <subdev/gpio.h>
 #include <subdev/i2c.h>
+#include <subdev/fuse.h>
 #include <subdev/clock.h>
 #include <subdev/therm.h>
 #include <subdev/mxm.h>
@@ -60,8 +61,9 @@ nvc0_identify(struct nouveau_device *device)
 	case 0xc0:
 		device->cname = "GF100";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -92,8 +94,9 @@ nvc0_identify(struct nouveau_device *device)
 	case 0xc4:
 		device->cname = "GF104";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -124,8 +127,9 @@ nvc0_identify(struct nouveau_device *device)
 	case 0xc3:
 		device->cname = "GF106";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -155,8 +159,9 @@ nvc0_identify(struct nouveau_device *device)
 	case 0xce:
 		device->cname = "GF114";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -187,8 +192,9 @@ nvc0_identify(struct nouveau_device *device)
 	case 0xcf:
 		device->cname = "GF116";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -212,15 +218,15 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nvc0_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nvc0_copy0_oclass;
-		device->oclass[NVDEV_ENGINE_COPY1  ] = &nvc0_copy1_oclass;
 		device->oclass[NVDEV_ENGINE_DISP   ] =  nva3_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] = &nvc0_perfmon_oclass;
 		break;
 	case 0xc1:
 		device->cname = "GF108";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -250,8 +256,9 @@ nvc0_identify(struct nouveau_device *device)
 	case 0xc8:
 		device->cname = "GF110";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -284,6 +291,7 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nvd0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nvd0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -315,6 +323,7 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nvd0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  gf117_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nve0.c b/drivers/gpu/drm/nouveau/core/engine/device/nve0.c
index cdf9147f32a1..732922690653 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nve0.c
@@ -26,6 +26,7 @@
 #include <subdev/bus.h>
 #include <subdev/gpio.h>
 #include <subdev/i2c.h>
+#include <subdev/fuse.h>
 #include <subdev/clock.h>
 #include <subdev/therm.h>
 #include <subdev/mxm.h>
@@ -62,6 +63,7 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nve0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -95,6 +97,7 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nve0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -128,6 +131,7 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nve0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -161,6 +165,7 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &gk20a_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  gk20a_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &gk20a_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  gk20a_fb_oclass;
 		device->oclass[NVDEV_SUBDEV_LTC    ] =  gk104_ltc_oclass;
@@ -174,12 +179,14 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_GR     ] =  gk20a_graph_oclass;
 		device->oclass[NVDEV_ENGINE_COPY2  ] = &nve0_copy2_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] = &nve0_perfmon_oclass;
+		device->oclass[NVDEV_SUBDEV_VOLT   ] = &gk20a_volt_oclass;
 		break;
 	case 0xf0:
 		device->cname = "GK110";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nve0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -213,6 +220,7 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nvd0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -241,11 +249,45 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] = &nvf0_perfmon_oclass;
 		break;
+	case 0x106:
+		device->cname = "GK208B";
+		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nve0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
+		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
+		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
+		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_MC     ] =  gk20a_mc_oclass;
+		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
+		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
+		device->oclass[NVDEV_SUBDEV_FB     ] =  nve0_fb_oclass;
+		device->oclass[NVDEV_SUBDEV_LTC    ] =  gk104_ltc_oclass;
+		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nve0_ibus_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
+		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
+		device->oclass[NVDEV_SUBDEV_PWR    ] =  nv108_pwr_oclass;
+		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
+		device->oclass[NVDEV_ENGINE_DMAOBJ ] =  nvd0_dmaeng_oclass;
+		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv108_fifo_oclass;
+		device->oclass[NVDEV_ENGINE_SW     ] =  nvc0_software_oclass;
+		device->oclass[NVDEV_ENGINE_GR     ] =  nv108_graph_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nvf0_disp_oclass;
+		device->oclass[NVDEV_ENGINE_COPY0  ] = &nve0_copy0_oclass;
+		device->oclass[NVDEV_ENGINE_COPY1  ] = &nve0_copy1_oclass;
+		device->oclass[NVDEV_ENGINE_COPY2  ] = &nve0_copy2_oclass;
+		device->oclass[NVDEV_ENGINE_BSP    ] = &nve0_bsp_oclass;
+		device->oclass[NVDEV_ENGINE_VP     ] = &nve0_vp_oclass;
+		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
+		break;
 	case 0x108:
 		device->cname = "GK208";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nve0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/base.c b/drivers/gpu/drm/nouveau/core/engine/disp/base.c
index 22d55f6cde50..64b84667f3a5 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/base.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/base.c
@@ -32,7 +32,8 @@
 #include "conn.h"
 
 int
-nouveau_disp_vblank_ctor(void *data, u32 size, struct nvkm_notify *notify)
+nouveau_disp_vblank_ctor(struct nouveau_object *object, void *data, u32 size,
+			 struct nvkm_notify *notify)
 {
 	struct nouveau_disp *disp =
 		container_of(notify->event, typeof(*disp), vblank);
@@ -61,7 +62,8 @@ nouveau_disp_vblank(struct nouveau_disp *disp, int head)
 }
 
 static int
-nouveau_disp_hpd_ctor(void *data, u32 size, struct nvkm_notify *notify)
+nouveau_disp_hpd_ctor(struct nouveau_object *object, void *data, u32 size,
+		      struct nvkm_notify *notify)
 {
 	struct nouveau_disp *disp =
 		container_of(notify->event, typeof(*disp), hpd);
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/conn.c b/drivers/gpu/drm/nouveau/core/engine/disp/conn.c
index 3d1070228977..1496b567dd4a 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/conn.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/conn.c
@@ -126,8 +126,8 @@ nvkm_connector_create_(struct nouveau_object *parent,
 			return 0;
 		}
 
-		ret = nvkm_notify_init(&gpio->event, nvkm_connector_hpd, true,
-				       &(struct nvkm_gpio_ntfy_req) {
+		ret = nvkm_notify_init(NULL, &gpio->event, nvkm_connector_hpd,
+				       true, &(struct nvkm_gpio_ntfy_req) {
 					.mask = NVKM_GPIO_TOGGLED,
 					.line = func.line,
 				       },
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/dport.c b/drivers/gpu/drm/nouveau/core/engine/disp/dport.c
index 39890221b91c..16db08dfba6e 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/dport.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/dport.c
@@ -28,7 +28,7 @@
 #include <subdev/bios/init.h>
 #include <subdev/i2c.h>
 
-#include <engine/disp.h>
+#include "nv50.h"
 
 #include <nvif/class.h>
 
@@ -326,7 +326,7 @@ void
 nouveau_dp_train(struct work_struct *w)
 {
 	struct nvkm_output_dp *outp = container_of(w, typeof(*outp), lt.work);
-	struct nouveau_disp *disp = nouveau_disp(outp);
+	struct nv50_disp_priv *priv = (void *)nouveau_disp(outp);
 	const struct dp_rates *cfg = nouveau_dp_rates;
 	struct dp_state _dp = {
 		.outp = outp,
@@ -334,8 +334,11 @@ nouveau_dp_train(struct work_struct *w)
 	u32 datarate = 0;
 	int ret;
 
+	if (!outp->base.info.location && priv->sor.magic)
+		priv->sor.magic(&outp->base);
+
 	/* bring capabilities within encoder limits */
-	if (nv_mclass(disp) < GF110_DISP)
+	if (nv_mclass(priv) < GF110_DISP)
 		outp->dpcd[2] &= ~DPCD_RC02_TPS3_SUPPORTED;
 	if ((outp->dpcd[2] & 0x1f) > outp->base.info.dpconf.link_nr) {
 		outp->dpcd[2] &= ~DPCD_RC02_MAX_LANE_COUNT;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/gm107.c b/drivers/gpu/drm/nouveau/core/engine/disp/gm107.c
index d54da8b5f87e..e2ad0543fb31 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/gm107.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/gm107.c
@@ -35,8 +35,8 @@
 
 static struct nouveau_oclass
 gm107_disp_sclass[] = {
-	{ GM107_DISP_CORE_CHANNEL_DMA, &nvd0_disp_mast_ofuncs.base },
-	{ GK110_DISP_BASE_CHANNEL_DMA, &nvd0_disp_sync_ofuncs.base },
+	{ GM107_DISP_CORE_CHANNEL_DMA, &nvd0_disp_core_ofuncs.base },
+	{ GK110_DISP_BASE_CHANNEL_DMA, &nvd0_disp_base_ofuncs.base },
 	{ GK104_DISP_OVERLAY_CONTROL_DMA, &nvd0_disp_ovly_ofuncs.base },
 	{ GK104_DISP_OVERLAY, &nvd0_disp_oimm_ofuncs.base },
 	{ GK104_DISP_CURSOR, &nvd0_disp_curs_ofuncs.base },
@@ -44,8 +44,8 @@ gm107_disp_sclass[] = {
 };
 
 static struct nouveau_oclass
-gm107_disp_base_oclass[] = {
-	{ GM107_DISP, &nvd0_disp_base_ofuncs },
+gm107_disp_main_oclass[] = {
+	{ GM107_DISP, &nvd0_disp_main_ofuncs },
 	{}
 };
 
@@ -68,7 +68,11 @@ gm107_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	nv_engine(priv)->sclass = gm107_disp_base_oclass;
+	ret = nvkm_event_init(&nvd0_disp_chan_uevent, 1, 17, &priv->uevent);
+	if (ret)
+		return ret;
+
+	nv_engine(priv)->sclass = gm107_disp_main_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nvd0_disp_intr;
 	INIT_WORK(&priv->supervisor, nvd0_disp_intr_supervisor);
@@ -80,7 +84,7 @@ gm107_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	priv->dac.sense = nv50_dac_sense;
 	priv->sor.power = nv50_sor_power;
 	priv->sor.hda_eld = nvd0_hda_eld;
-	priv->sor.hdmi = nvd0_hdmi_ctrl;
+	priv->sor.hdmi = nve0_hdmi_ctrl;
 	return 0;
 }
 
@@ -95,9 +99,9 @@ gm107_disp_oclass = &(struct nv50_disp_impl) {
 	},
 	.base.vblank = &nvd0_disp_vblank_func,
 	.base.outp =  nvd0_disp_outp_sclass,
-	.mthd.core = &nve0_disp_mast_mthd_chan,
-	.mthd.base = &nvd0_disp_sync_mthd_chan,
+	.mthd.core = &nve0_disp_core_mthd_chan,
+	.mthd.base = &nvd0_disp_base_mthd_chan,
 	.mthd.ovly = &nve0_disp_ovly_mthd_chan,
 	.mthd.prev = -0x020000,
-	.head.scanoutpos = nvd0_disp_base_scanoutpos,
+	.head.scanoutpos = nvd0_disp_main_scanoutpos,
 }.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/gm204.c b/drivers/gpu/drm/nouveau/core/engine/disp/gm204.c
new file mode 100644
index 000000000000..672ded79b2a9
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/gm204.c
@@ -0,0 +1,114 @@
+/*
+ * Copyright 2012 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include <engine/software.h>
+#include <engine/disp.h>
+
+#include <nvif/class.h>
+
+#include "nv50.h"
+
+/*******************************************************************************
+ * Base display object
+ ******************************************************************************/
+
+static struct nouveau_oclass
+gm204_disp_sclass[] = {
+	{ GM204_DISP_CORE_CHANNEL_DMA, &nvd0_disp_core_ofuncs.base },
+	{ GK110_DISP_BASE_CHANNEL_DMA, &nvd0_disp_base_ofuncs.base },
+	{ GK104_DISP_OVERLAY_CONTROL_DMA, &nvd0_disp_ovly_ofuncs.base },
+	{ GK104_DISP_OVERLAY, &nvd0_disp_oimm_ofuncs.base },
+	{ GK104_DISP_CURSOR, &nvd0_disp_curs_ofuncs.base },
+	{}
+};
+
+static struct nouveau_oclass
+gm204_disp_main_oclass[] = {
+	{ GM204_DISP, &nvd0_disp_main_ofuncs },
+	{}
+};
+
+/*******************************************************************************
+ * Display engine implementation
+ ******************************************************************************/
+
+static int
+gm204_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+	       struct nouveau_oclass *oclass, void *data, u32 size,
+	       struct nouveau_object **pobject)
+{
+	struct nv50_disp_priv *priv;
+	int heads = nv_rd32(parent, 0x022448);
+	int ret;
+
+	ret = nouveau_disp_create(parent, engine, oclass, heads,
+				  "PDISP", "display", &priv);
+	*pobject = nv_object(priv);
+	if (ret)
+		return ret;
+
+	ret = nvkm_event_init(&nvd0_disp_chan_uevent, 1, 17, &priv->uevent);
+	if (ret)
+		return ret;
+
+	nv_engine(priv)->sclass = gm204_disp_main_oclass;
+	nv_engine(priv)->cclass = &nv50_disp_cclass;
+	nv_subdev(priv)->intr = nvd0_disp_intr;
+	INIT_WORK(&priv->supervisor, nvd0_disp_intr_supervisor);
+	priv->sclass = gm204_disp_sclass;
+	priv->head.nr = heads;
+	priv->dac.nr = 3;
+	priv->sor.nr = 4;
+	priv->dac.power = nv50_dac_power;
+	priv->dac.sense = nv50_dac_sense;
+	priv->sor.power = nv50_sor_power;
+	priv->sor.hda_eld = nvd0_hda_eld;
+	priv->sor.hdmi = nvd0_hdmi_ctrl;
+	priv->sor.magic = gm204_sor_magic;
+	return 0;
+}
+
+struct nouveau_oclass *
+gm204_disp_outp_sclass[] = {
+	&gm204_sor_dp_impl.base.base,
+	NULL
+};
+
+struct nouveau_oclass *
+gm204_disp_oclass = &(struct nv50_disp_impl) {
+	.base.base.handle = NV_ENGINE(DISP, 0x07),
+	.base.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = gm204_disp_ctor,
+		.dtor = _nouveau_disp_dtor,
+		.init = _nouveau_disp_init,
+		.fini = _nouveau_disp_fini,
+	},
+	.base.vblank = &nvd0_disp_vblank_func,
+	.base.outp =  gm204_disp_outp_sclass,
+	.mthd.core = &nve0_disp_core_mthd_chan,
+	.mthd.base = &nvd0_disp_base_mthd_chan,
+	.mthd.ovly = &nve0_disp_ovly_mthd_chan,
+	.mthd.prev = -0x020000,
+	.head.scanoutpos = nvd0_disp_main_scanoutpos,
+}.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/hdanva3.c b/drivers/gpu/drm/nouveau/core/engine/disp/hdanva3.c
index 8b4e06abe533..fe9ef5894dd4 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/hdanva3.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/hdanva3.c
@@ -26,6 +26,8 @@
 #include <nvif/unpack.h>
 #include <nvif/class.h>
 
+#include <subdev/timer.h>
+
 #include "nv50.h"
 
 int
@@ -46,16 +48,21 @@ nva3_hda_eld(NV50_DISP_MTHD_V1)
 		return ret;
 
 	if (size && args->v0.data[0]) {
+		if (outp->info.type == DCB_OUTPUT_DP) {
+			nv_mask(priv, 0x61c1e0 + soff, 0x8000000d, 0x80000001);
+			nv_wait(priv, 0x61c1e0 + soff, 0x80000000, 0x00000000);
+		}
 		for (i = 0; i < size; i++)
 			nv_wr32(priv, 0x61c440 + soff, (i << 8) | args->v0.data[0]);
 		for (; i < 0x60; i++)
 			nv_wr32(priv, 0x61c440 + soff, (i << 8));
 		nv_mask(priv, 0x61c448 + soff, 0x80000003, 0x80000003);
-	} else
-	if (size) {
-		nv_mask(priv, 0x61c448 + soff, 0x80000003, 0x80000001);
 	} else {
-		nv_mask(priv, 0x61c448 + soff, 0x80000003, 0x80000000);
+		if (outp->info.type == DCB_OUTPUT_DP) {
+			nv_mask(priv, 0x61c1e0 + soff, 0x80000001, 0x80000000);
+			nv_wait(priv, 0x61c1e0 + soff, 0x80000000, 0x00000000);
+		}
+		nv_mask(priv, 0x61c448 + soff, 0x80000003, 0x80000000 | !!size);
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/hdanvd0.c b/drivers/gpu/drm/nouveau/core/engine/disp/hdanvd0.c
index baf558fc12fb..1d4e8432d857 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/hdanvd0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/hdanvd0.c
@@ -26,10 +26,7 @@
 #include <nvif/unpack.h>
 #include <nvif/class.h>
 
-#include <subdev/bios.h>
-#include <subdev/bios/dcb.h>
-#include <subdev/bios/dp.h>
-#include <subdev/bios/init.h>
+#include <subdev/timer.h>
 
 #include "nv50.h"
 
@@ -40,6 +37,7 @@ nvd0_hda_eld(NV50_DISP_MTHD_V1)
 		struct nv50_disp_sor_hda_eld_v0 v0;
 	} *args = data;
 	const u32 soff = outp->or * 0x030;
+	const u32 hoff = head * 0x800;
 	int ret, i;
 
 	nv_ioctl(object, "disp sor hda eld size %d\n", size);
@@ -51,16 +49,22 @@ nvd0_hda_eld(NV50_DISP_MTHD_V1)
 		return ret;
 
 	if (size && args->v0.data[0]) {
+		if (outp->info.type == DCB_OUTPUT_DP) {
+			nv_mask(priv, 0x616618 + hoff, 0x8000000c, 0x80000001);
+			nv_wait(priv, 0x616618 + hoff, 0x80000000, 0x00000000);
+		}
+		nv_mask(priv, 0x616548 + hoff, 0x00000070, 0x00000000);
 		for (i = 0; i < size; i++)
 			nv_wr32(priv, 0x10ec00 + soff, (i << 8) | args->v0.data[i]);
 		for (; i < 0x60; i++)
 			nv_wr32(priv, 0x10ec00 + soff, (i << 8));
 		nv_mask(priv, 0x10ec10 + soff, 0x80000003, 0x80000003);
-	} else
-	if (size) {
-		nv_mask(priv, 0x10ec10 + soff, 0x80000003, 0x80000001);
 	} else {
-		nv_mask(priv, 0x10ec10 + soff, 0x80000003, 0x80000000);
+		if (outp->info.type == DCB_OUTPUT_DP) {
+			nv_mask(priv, 0x616618 + hoff, 0x80000001, 0x80000000);
+			nv_wait(priv, 0x616618 + hoff, 0x80000000, 0x00000000);
+		}
+		nv_mask(priv, 0x10ec10 + soff, 0x80000003, 0x80000000 | !!size);
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/hdminvd0.c b/drivers/gpu/drm/nouveau/core/engine/disp/hdminvd0.c
index 3106d295b48d..bac4fc4570f0 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/hdminvd0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/hdminvd0.c
@@ -75,8 +75,5 @@ nvd0_hdmi_ctrl(NV50_DISP_MTHD_V1)
 
 	/* HDMI_CTRL */
 	nv_mask(priv, 0x616798 + hoff, 0x401f007f, ctrl);
-
-	/* NFI, audio doesn't work without it though.. */
-	nv_mask(priv, 0x616548 + hoff, 0x00000070, 0x00000000);
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/hdminve0.c b/drivers/gpu/drm/nouveau/core/engine/disp/hdminve0.c
new file mode 100644
index 000000000000..528d14ec2f7f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/hdminve0.c
@@ -0,0 +1,83 @@
+/*
+ * Copyright 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include <core/client.h>
+#include <nvif/unpack.h>
+#include <nvif/class.h>
+
+#include "nv50.h"
+
+int
+nve0_hdmi_ctrl(NV50_DISP_MTHD_V1)
+{
+	const u32 hoff = (head * 0x800);
+	const u32 hdmi = (head * 0x400);
+	union {
+		struct nv50_disp_sor_hdmi_pwr_v0 v0;
+	} *args = data;
+	u32 ctrl;
+	int ret;
+
+	nv_ioctl(object, "disp sor hdmi ctrl size %d\n", size);
+	if (nvif_unpack(args->v0, 0, 0, false)) {
+		nv_ioctl(object, "disp sor hdmi ctrl vers %d state %d "
+				 "max_ac_packet %d rekey %d\n",
+			 args->v0.version, args->v0.state,
+			 args->v0.max_ac_packet, args->v0.rekey);
+		if (args->v0.max_ac_packet > 0x1f || args->v0.rekey > 0x7f)
+			return -EINVAL;
+		ctrl  = 0x40000000 * !!args->v0.state;
+		ctrl |= args->v0.max_ac_packet << 16;
+		ctrl |= args->v0.rekey;
+	} else
+		return ret;
+
+	if (!(ctrl & 0x40000000)) {
+		nv_mask(priv, 0x616798 + hoff, 0x40000000, 0x00000000);
+		nv_mask(priv, 0x6900c0 + hdmi, 0x00000001, 0x00000000);
+		nv_mask(priv, 0x690000 + hdmi, 0x00000001, 0x00000000);
+		return 0;
+	}
+
+	/* AVI InfoFrame */
+	nv_mask(priv, 0x690000 + hdmi, 0x00000001, 0x00000000);
+	nv_wr32(priv, 0x690008 + hdmi, 0x000d0282);
+	nv_wr32(priv, 0x69000c + hdmi, 0x0000006f);
+	nv_wr32(priv, 0x690010 + hdmi, 0x00000000);
+	nv_wr32(priv, 0x690014 + hdmi, 0x00000000);
+	nv_wr32(priv, 0x690018 + hdmi, 0x00000000);
+	nv_mask(priv, 0x690000 + hdmi, 0x00000001, 0x00000001);
+
+	/* ??? InfoFrame? */
+	nv_mask(priv, 0x6900c0 + hdmi, 0x00000001, 0x00000000);
+	nv_wr32(priv, 0x6900cc + hdmi, 0x00000010);
+	nv_mask(priv, 0x6900c0 + hdmi, 0x00000001, 0x00000001);
+
+	/* ??? */
+	nv_wr32(priv, 0x690080 + hdmi, 0x82000000);
+
+	/* HDMI_CTRL */
+	nv_mask(priv, 0x616798 + hoff, 0x401f007f, ctrl);
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c
index f8cbb512132f..44a8290aaea5 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c
@@ -29,6 +29,7 @@
 #include <core/enum.h>
 #include <nvif/unpack.h>
 #include <nvif/class.h>
+#include <nvif/event.h>
 
 #include <subdev/bios.h>
 #include <subdev/bios/dcb.h>
@@ -82,6 +83,73 @@ nv50_disp_chan_destroy(struct nv50_disp_chan *chan)
 	nouveau_namedb_destroy(&chan->base);
 }
 
+static void
+nv50_disp_chan_uevent_fini(struct nvkm_event *event, int type, int index)
+{
+	struct nv50_disp_priv *priv = container_of(event, typeof(*priv), uevent);
+	nv_mask(priv, 0x610028, 0x00000001 << index, 0x00000000 << index);
+	nv_wr32(priv, 0x610020, 0x00000001 << index);
+}
+
+static void
+nv50_disp_chan_uevent_init(struct nvkm_event *event, int types, int index)
+{
+	struct nv50_disp_priv *priv = container_of(event, typeof(*priv), uevent);
+	nv_wr32(priv, 0x610020, 0x00000001 << index);
+	nv_mask(priv, 0x610028, 0x00000001 << index, 0x00000001 << index);
+}
+
+void
+nv50_disp_chan_uevent_send(struct nv50_disp_priv *priv, int chid)
+{
+	struct nvif_notify_uevent_rep {
+	} rep;
+
+	nvkm_event_send(&priv->uevent, 1, chid, &rep, sizeof(rep));
+}
+
+int
+nv50_disp_chan_uevent_ctor(struct nouveau_object *object, void *data, u32 size,
+			   struct nvkm_notify *notify)
+{
+	struct nv50_disp_dmac *dmac = (void *)object;
+	union {
+		struct nvif_notify_uevent_req none;
+	} *args = data;
+	int ret;
+
+	if (nvif_unvers(args->none)) {
+		notify->size  = sizeof(struct nvif_notify_uevent_rep);
+		notify->types = 1;
+		notify->index = dmac->base.chid;
+		return 0;
+	}
+
+	return ret;
+}
+
+const struct nvkm_event_func
+nv50_disp_chan_uevent = {
+	.ctor = nv50_disp_chan_uevent_ctor,
+	.init = nv50_disp_chan_uevent_init,
+	.fini = nv50_disp_chan_uevent_fini,
+};
+
+int
+nv50_disp_chan_ntfy(struct nouveau_object *object, u32 type,
+		    struct nvkm_event **pevent)
+{
+	struct nv50_disp_priv *priv = (void *)object->engine;
+	switch (type) {
+	case NV50_DISP_CORE_CHANNEL_DMA_V0_NTFY_UEVENT:
+		*pevent = &priv->uevent;
+		return 0;
+	default:
+		break;
+	}
+	return -EINVAL;
+}
+
 int
 nv50_disp_chan_map(struct nouveau_object *object, u64 *addr, u32 *size)
 {
@@ -195,7 +263,7 @@ nv50_disp_dmac_init(struct nouveau_object *object)
 		return ret;
 
 	/* enable error reporting */
-	nv_mask(priv, 0x610028, 0x00010001 << chid, 0x00010001 << chid);
+	nv_mask(priv, 0x610028, 0x00010000 << chid, 0x00010000 << chid);
 
 	/* initialise channel for dma command submission */
 	nv_wr32(priv, 0x610204 + (chid * 0x0010), dmac->push);
@@ -232,7 +300,7 @@ nv50_disp_dmac_fini(struct nouveau_object *object, bool suspend)
 			return -EBUSY;
 	}
 
-	/* disable error reporting */
+	/* disable error reporting and completion notifications */
 	nv_mask(priv, 0x610028, 0x00010001 << chid, 0x00000000 << chid);
 
 	return nv50_disp_chan_fini(&dmac->base, suspend);
@@ -308,7 +376,7 @@ nv50_disp_mthd_chan(struct nv50_disp_priv *priv, int debug, int head,
 }
 
 const struct nv50_disp_mthd_list
-nv50_disp_mast_mthd_base = {
+nv50_disp_core_mthd_base = {
 	.mthd = 0x0000,
 	.addr = 0x000000,
 	.data = {
@@ -321,7 +389,7 @@ nv50_disp_mast_mthd_base = {
 };
 
 static const struct nv50_disp_mthd_list
-nv50_disp_mast_mthd_dac = {
+nv50_disp_core_mthd_dac = {
 	.mthd = 0x0080,
 	.addr = 0x000008,
 	.data = {
@@ -333,7 +401,7 @@ nv50_disp_mast_mthd_dac = {
 };
 
 const struct nv50_disp_mthd_list
-nv50_disp_mast_mthd_sor = {
+nv50_disp_core_mthd_sor = {
 	.mthd = 0x0040,
 	.addr = 0x000008,
 	.data = {
@@ -343,7 +411,7 @@ nv50_disp_mast_mthd_sor = {
 };
 
 const struct nv50_disp_mthd_list
-nv50_disp_mast_mthd_pior = {
+nv50_disp_core_mthd_pior = {
 	.mthd = 0x0040,
 	.addr = 0x000008,
 	.data = {
@@ -353,7 +421,7 @@ nv50_disp_mast_mthd_pior = {
 };
 
 static const struct nv50_disp_mthd_list
-nv50_disp_mast_mthd_head = {
+nv50_disp_core_mthd_head = {
 	.mthd = 0x0400,
 	.addr = 0x000540,
 	.data = {
@@ -400,21 +468,21 @@ nv50_disp_mast_mthd_head = {
 };
 
 static const struct nv50_disp_mthd_chan
-nv50_disp_mast_mthd_chan = {
+nv50_disp_core_mthd_chan = {
 	.name = "Core",
 	.addr = 0x000000,
 	.data = {
-		{ "Global", 1, &nv50_disp_mast_mthd_base },
-		{    "DAC", 3, &nv50_disp_mast_mthd_dac  },
-		{    "SOR", 2, &nv50_disp_mast_mthd_sor  },
-		{   "PIOR", 3, &nv50_disp_mast_mthd_pior },
-		{   "HEAD", 2, &nv50_disp_mast_mthd_head },
+		{ "Global", 1, &nv50_disp_core_mthd_base },
+		{    "DAC", 3, &nv50_disp_core_mthd_dac  },
+		{    "SOR", 2, &nv50_disp_core_mthd_sor  },
+		{   "PIOR", 3, &nv50_disp_core_mthd_pior },
+		{   "HEAD", 2, &nv50_disp_core_mthd_head },
 		{}
 	}
 };
 
 int
-nv50_disp_mast_ctor(struct nouveau_object *parent,
+nv50_disp_core_ctor(struct nouveau_object *parent,
 		    struct nouveau_object *engine,
 		    struct nouveau_oclass *oclass, void *data, u32 size,
 		    struct nouveau_object **pobject)
@@ -443,7 +511,7 @@ nv50_disp_mast_ctor(struct nouveau_object *parent,
 }
 
 static int
-nv50_disp_mast_init(struct nouveau_object *object)
+nv50_disp_core_init(struct nouveau_object *object)
 {
 	struct nv50_disp_priv *priv = (void *)object->engine;
 	struct nv50_disp_dmac *mast = (void *)object;
@@ -454,7 +522,7 @@ nv50_disp_mast_init(struct nouveau_object *object)
 		return ret;
 
 	/* enable error reporting */
-	nv_mask(priv, 0x610028, 0x00010001, 0x00010001);
+	nv_mask(priv, 0x610028, 0x00010000, 0x00010000);
 
 	/* attempt to unstick channel from some unknown state */
 	if ((nv_rd32(priv, 0x610200) & 0x009f0000) == 0x00020000)
@@ -480,7 +548,7 @@ nv50_disp_mast_init(struct nouveau_object *object)
 }
 
 static int
-nv50_disp_mast_fini(struct nouveau_object *object, bool suspend)
+nv50_disp_core_fini(struct nouveau_object *object, bool suspend)
 {
 	struct nv50_disp_priv *priv = (void *)object->engine;
 	struct nv50_disp_dmac *mast = (void *)object;
@@ -494,19 +562,20 @@ nv50_disp_mast_fini(struct nouveau_object *object, bool suspend)
 			return -EBUSY;
 	}
 
-	/* disable error reporting */
+	/* disable error reporting and completion notifications */
 	nv_mask(priv, 0x610028, 0x00010001, 0x00000000);
 
 	return nv50_disp_chan_fini(&mast->base, suspend);
 }
 
 struct nv50_disp_chan_impl
-nv50_disp_mast_ofuncs = {
-	.base.ctor = nv50_disp_mast_ctor,
+nv50_disp_core_ofuncs = {
+	.base.ctor = nv50_disp_core_ctor,
 	.base.dtor = nv50_disp_dmac_dtor,
-	.base.init = nv50_disp_mast_init,
-	.base.fini = nv50_disp_mast_fini,
+	.base.init = nv50_disp_core_init,
+	.base.fini = nv50_disp_core_fini,
 	.base.map  = nv50_disp_chan_map,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
 	.chid = 0,
@@ -519,7 +588,7 @@ nv50_disp_mast_ofuncs = {
  ******************************************************************************/
 
 static const struct nv50_disp_mthd_list
-nv50_disp_sync_mthd_base = {
+nv50_disp_base_mthd_base = {
 	.mthd = 0x0000,
 	.addr = 0x000000,
 	.data = {
@@ -544,7 +613,7 @@ nv50_disp_sync_mthd_base = {
 };
 
 const struct nv50_disp_mthd_list
-nv50_disp_sync_mthd_image = {
+nv50_disp_base_mthd_image = {
 	.mthd = 0x0400,
 	.addr = 0x000000,
 	.data = {
@@ -558,18 +627,18 @@ nv50_disp_sync_mthd_image = {
 };
 
 static const struct nv50_disp_mthd_chan
-nv50_disp_sync_mthd_chan = {
+nv50_disp_base_mthd_chan = {
 	.name = "Base",
 	.addr = 0x000540,
 	.data = {
-		{ "Global", 1, &nv50_disp_sync_mthd_base },
-		{  "Image", 2, &nv50_disp_sync_mthd_image },
+		{ "Global", 1, &nv50_disp_base_mthd_base },
+		{  "Image", 2, &nv50_disp_base_mthd_image },
 		{}
 	}
 };
 
 int
-nv50_disp_sync_ctor(struct nouveau_object *parent,
+nv50_disp_base_ctor(struct nouveau_object *parent,
 		    struct nouveau_object *engine,
 		    struct nouveau_oclass *oclass, void *data, u32 size,
 		    struct nouveau_object **pobject)
@@ -602,11 +671,12 @@ nv50_disp_sync_ctor(struct nouveau_object *parent,
 }
 
 struct nv50_disp_chan_impl
-nv50_disp_sync_ofuncs = {
-	.base.ctor = nv50_disp_sync_ctor,
+nv50_disp_base_ofuncs = {
+	.base.ctor = nv50_disp_base_ctor,
 	.base.dtor = nv50_disp_dmac_dtor,
 	.base.init = nv50_disp_dmac_init,
 	.base.fini = nv50_disp_dmac_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -696,6 +766,7 @@ nv50_disp_ovly_ofuncs = {
 	.base.dtor = nv50_disp_dmac_dtor,
 	.base.init = nv50_disp_dmac_init,
 	.base.fini = nv50_disp_dmac_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -813,6 +884,7 @@ nv50_disp_oimm_ofuncs = {
 	.base.dtor = nv50_disp_pioc_dtor,
 	.base.init = nv50_disp_pioc_init,
 	.base.fini = nv50_disp_pioc_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -860,6 +932,7 @@ nv50_disp_curs_ofuncs = {
 	.base.dtor = nv50_disp_pioc_dtor,
 	.base.init = nv50_disp_pioc_init,
 	.base.fini = nv50_disp_pioc_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -871,7 +944,7 @@ nv50_disp_curs_ofuncs = {
  ******************************************************************************/
 
 int
-nv50_disp_base_scanoutpos(NV50_DISP_MTHD_V0)
+nv50_disp_main_scanoutpos(NV50_DISP_MTHD_V0)
 {
 	const u32 blanke = nv_rd32(priv, 0x610aec + (head * 0x540));
 	const u32 blanks = nv_rd32(priv, 0x610af4 + (head * 0x540));
@@ -903,7 +976,7 @@ nv50_disp_base_scanoutpos(NV50_DISP_MTHD_V0)
 }
 
 int
-nv50_disp_base_mthd(struct nouveau_object *object, u32 mthd,
+nv50_disp_main_mthd(struct nouveau_object *object, u32 mthd,
 		    void *data, u32 size)
 {
 	const struct nv50_disp_impl *impl = (void *)nv_oclass(object->engine);
@@ -1027,7 +1100,7 @@ nv50_disp_base_mthd(struct nouveau_object *object, u32 mthd,
 }
 
 int
-nv50_disp_base_ctor(struct nouveau_object *parent,
+nv50_disp_main_ctor(struct nouveau_object *parent,
 		    struct nouveau_object *engine,
 		    struct nouveau_oclass *oclass, void *data, u32 size,
 		    struct nouveau_object **pobject)
@@ -1047,7 +1120,7 @@ nv50_disp_base_ctor(struct nouveau_object *parent,
 }
 
 void
-nv50_disp_base_dtor(struct nouveau_object *object)
+nv50_disp_main_dtor(struct nouveau_object *object)
 {
 	struct nv50_disp_base *base = (void *)object;
 	nouveau_ramht_ref(NULL, &base->ramht);
@@ -1055,7 +1128,7 @@ nv50_disp_base_dtor(struct nouveau_object *object)
 }
 
 static int
-nv50_disp_base_init(struct nouveau_object *object)
+nv50_disp_main_init(struct nouveau_object *object)
 {
 	struct nv50_disp_priv *priv = (void *)object->engine;
 	struct nv50_disp_base *base = (void *)object;
@@ -1123,7 +1196,7 @@ nv50_disp_base_init(struct nouveau_object *object)
 }
 
 static int
-nv50_disp_base_fini(struct nouveau_object *object, bool suspend)
+nv50_disp_main_fini(struct nouveau_object *object, bool suspend)
 {
 	struct nv50_disp_priv *priv = (void *)object->engine;
 	struct nv50_disp_base *base = (void *)object;
@@ -1136,25 +1209,25 @@ nv50_disp_base_fini(struct nouveau_object *object, bool suspend)
 }
 
 struct nouveau_ofuncs
-nv50_disp_base_ofuncs = {
-	.ctor = nv50_disp_base_ctor,
-	.dtor = nv50_disp_base_dtor,
-	.init = nv50_disp_base_init,
-	.fini = nv50_disp_base_fini,
-	.mthd = nv50_disp_base_mthd,
+nv50_disp_main_ofuncs = {
+	.ctor = nv50_disp_main_ctor,
+	.dtor = nv50_disp_main_dtor,
+	.init = nv50_disp_main_init,
+	.fini = nv50_disp_main_fini,
+	.mthd = nv50_disp_main_mthd,
 	.ntfy = nouveau_disp_ntfy,
 };
 
 static struct nouveau_oclass
-nv50_disp_base_oclass[] = {
-	{ NV50_DISP, &nv50_disp_base_ofuncs },
+nv50_disp_main_oclass[] = {
+	{ NV50_DISP, &nv50_disp_main_ofuncs },
 	{}
 };
 
 static struct nouveau_oclass
 nv50_disp_sclass[] = {
-	{ NV50_DISP_CORE_CHANNEL_DMA, &nv50_disp_mast_ofuncs.base },
-	{ NV50_DISP_BASE_CHANNEL_DMA, &nv50_disp_sync_ofuncs.base },
+	{ NV50_DISP_CORE_CHANNEL_DMA, &nv50_disp_core_ofuncs.base },
+	{ NV50_DISP_BASE_CHANNEL_DMA, &nv50_disp_base_ofuncs.base },
 	{ NV50_DISP_OVERLAY_CHANNEL_DMA, &nv50_disp_ovly_ofuncs.base },
 	{ NV50_DISP_OVERLAY, &nv50_disp_oimm_ofuncs.base },
 	{ NV50_DISP_CURSOR, &nv50_disp_curs_ofuncs.base },
@@ -1559,7 +1632,7 @@ nv50_disp_intr_unk20_1(struct nv50_disp_priv *priv, int head)
 }
 
 static void
-nv50_disp_intr_unk20_2_dp(struct nv50_disp_priv *priv,
+nv50_disp_intr_unk20_2_dp(struct nv50_disp_priv *priv, int head,
 			  struct dcb_output *outp, u32 pclk)
 {
 	const int link = !(outp->sorconf.link & 1);
@@ -1568,24 +1641,36 @@ nv50_disp_intr_unk20_2_dp(struct nv50_disp_priv *priv,
 	const u32 loff = (link * 0x080) + soff;
 	const u32 ctrl = nv_rd32(priv, 0x610794 + (or * 8));
 	const u32 symbol = 100000;
-	u32 dpctrl = nv_rd32(priv, 0x61c10c + loff) & 0x0000f0000;
+	const s32 vactive = nv_rd32(priv, 0x610af8 + (head * 0x540)) & 0xffff;
+	const s32 vblanke = nv_rd32(priv, 0x610ae8 + (head * 0x540)) & 0xffff;
+	const s32 vblanks = nv_rd32(priv, 0x610af0 + (head * 0x540)) & 0xffff;
+	u32 dpctrl = nv_rd32(priv, 0x61c10c + loff);
 	u32 clksor = nv_rd32(priv, 0x614300 + soff);
 	int bestTU = 0, bestVTUi = 0, bestVTUf = 0, bestVTUa = 0;
 	int TU, VTUi, VTUf, VTUa;
 	u64 link_data_rate, link_ratio, unk;
 	u32 best_diff = 64 * symbol;
 	u32 link_nr, link_bw, bits;
-
-	/* calculate packed data rate for each lane */
-	if      (dpctrl > 0x00030000) link_nr = 4;
-	else if (dpctrl > 0x00010000) link_nr = 2;
-	else			      link_nr = 1;
-
-	if (clksor & 0x000c0000)
-		link_bw = 270000;
-	else
-		link_bw = 162000;
-
+	u64 value;
+
+	link_bw = (clksor & 0x000c0000) ? 270000 : 162000;
+	link_nr = hweight32(dpctrl & 0x000f0000);
+
+	/* symbols/hblank - algorithm taken from comments in tegra driver */
+	value = vblanke + vactive - vblanks - 7;
+	value = value * link_bw;
+	do_div(value, pclk);
+	value = value - (3 * !!(dpctrl & 0x00004000)) - (12 / link_nr);
+	nv_mask(priv, 0x61c1e8 + soff, 0x0000ffff, value);
+
+	/* symbols/vblank - algorithm taken from comments in tegra driver */
+	value = vblanks - vblanke - 25;
+	value = value * link_bw;
+	do_div(value, pclk);
+	value = value - ((36 / link_nr) + 3) - 1;
+	nv_mask(priv, 0x61c1ec + soff, 0x00ffffff, value);
+
+	/* watermark / activesym */
 	if      ((ctrl & 0xf0000) == 0x60000) bits = 30;
 	else if ((ctrl & 0xf0000) == 0x50000) bits = 24;
 	else                                  bits = 18;
@@ -1731,7 +1816,7 @@ nv50_disp_intr_unk20_2(struct nv50_disp_priv *priv, int head)
 	} else
 	if (!outp->info.location) {
 		if (outp->info.type == DCB_OUTPUT_DP)
-			nv50_disp_intr_unk20_2_dp(priv, &outp->info, pclk);
+			nv50_disp_intr_unk20_2_dp(priv, head, &outp->info, pclk);
 		oreg = 0x614300 + (ffs(outp->info.or) - 1) * 0x800;
 		oval = (conf & 0x0100) ? 0x00000101 : 0x00000000;
 		hval = 0x00000000;
@@ -1847,6 +1932,12 @@ nv50_disp_intr(struct nouveau_subdev *subdev)
 		intr0 &= ~(0x00010000 << chid);
 	}
 
+	while (intr0 & 0x0000001f) {
+		u32 chid = __ffs(intr0 & 0x0000001f);
+		nv50_disp_chan_uevent_send(priv, chid);
+		intr0 &= ~(0x00000001 << chid);
+	}
+
 	if (intr1 & 0x00000004) {
 		nouveau_disp_vblank(&priv->base, 0);
 		nv_wr32(priv, 0x610024, 0x00000004);
@@ -1881,7 +1972,11 @@ nv50_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	nv_engine(priv)->sclass = nv50_disp_base_oclass;
+	ret = nvkm_event_init(&nv50_disp_chan_uevent, 1, 9, &priv->uevent);
+	if (ret)
+		return ret;
+
+	nv_engine(priv)->sclass = nv50_disp_main_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nv50_disp_intr;
 	INIT_WORK(&priv->supervisor, nv50_disp_intr_supervisor);
@@ -1914,9 +2009,9 @@ nv50_disp_oclass = &(struct nv50_disp_impl) {
 	},
 	.base.vblank = &nv50_disp_vblank_func,
 	.base.outp =  nv50_disp_outp_sclass,
-	.mthd.core = &nv50_disp_mast_mthd_chan,
-	.mthd.base = &nv50_disp_sync_mthd_chan,
+	.mthd.core = &nv50_disp_core_mthd_chan,
+	.mthd.base = &nv50_disp_base_mthd_chan,
 	.mthd.ovly = &nv50_disp_ovly_mthd_chan,
 	.mthd.prev = 0x000004,
-	.head.scanoutpos = nv50_disp_base_scanoutpos,
+	.head.scanoutpos = nv50_disp_main_scanoutpos,
 }.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.h b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.h
index 8ab14461f70c..7f08078ee925 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.h
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.h
@@ -26,6 +26,8 @@ struct nv50_disp_priv {
 	struct work_struct supervisor;
 	u32 super;
 
+	struct nvkm_event uevent;
+
 	struct {
 		int nr;
 	} head;
@@ -40,6 +42,7 @@ struct nv50_disp_priv {
 		int (*hda_eld)(NV50_DISP_MTHD_V1);
 		int (*hdmi)(NV50_DISP_MTHD_V1);
 		u32 lvdsconf;
+		void (*magic)(struct nvkm_output *);
 	} sor;
 	struct {
 		int nr;
@@ -61,10 +64,10 @@ struct nv50_disp_impl {
 	} head;
 };
 
-int nv50_disp_base_scanoutpos(NV50_DISP_MTHD_V0);
-int nv50_disp_base_mthd(struct nouveau_object *, u32, void *, u32);
+int nv50_disp_main_scanoutpos(NV50_DISP_MTHD_V0);
+int nv50_disp_main_mthd(struct nouveau_object *, u32, void *, u32);
 
-int nvd0_disp_base_scanoutpos(NV50_DISP_MTHD_V0);
+int nvd0_disp_main_scanoutpos(NV50_DISP_MTHD_V0);
 
 int nv50_dac_power(NV50_DISP_MTHD_V1);
 int nv50_dac_sense(NV50_DISP_MTHD_V1);
@@ -75,6 +78,7 @@ int nvd0_hda_eld(NV50_DISP_MTHD_V1);
 int nv84_hdmi_ctrl(NV50_DISP_MTHD_V1);
 int nva3_hdmi_ctrl(NV50_DISP_MTHD_V1);
 int nvd0_hdmi_ctrl(NV50_DISP_MTHD_V1);
+int nve0_hdmi_ctrl(NV50_DISP_MTHD_V1);
 
 int nv50_sor_power(NV50_DISP_MTHD_V1);
 
@@ -116,9 +120,16 @@ struct nv50_disp_chan {
 	int chid;
 };
 
+int  nv50_disp_chan_ntfy(struct nouveau_object *, u32, struct nvkm_event **);
 int  nv50_disp_chan_map(struct nouveau_object *, u64 *, u32 *);
 u32  nv50_disp_chan_rd32(struct nouveau_object *, u64);
 void nv50_disp_chan_wr32(struct nouveau_object *, u64, u32);
+extern const struct nvkm_event_func nv50_disp_chan_uevent;
+int  nv50_disp_chan_uevent_ctor(struct nouveau_object *, void *, u32,
+				struct nvkm_notify *);
+void nv50_disp_chan_uevent_send(struct nv50_disp_priv *, int);
+
+extern const struct nvkm_event_func nvd0_disp_chan_uevent;
 
 #define nv50_disp_chan_init(a)                                                 \
 	nouveau_namedb_init(&(a)->base)
@@ -159,18 +170,18 @@ struct nv50_disp_mthd_chan {
 	} data[];
 };
 
-extern struct nv50_disp_chan_impl nv50_disp_mast_ofuncs;
-int nv50_disp_mast_ctor(struct nouveau_object *, struct nouveau_object *,
+extern struct nv50_disp_chan_impl nv50_disp_core_ofuncs;
+int nv50_disp_core_ctor(struct nouveau_object *, struct nouveau_object *,
 			struct nouveau_oclass *, void *, u32,
 			struct nouveau_object **);
-extern const struct nv50_disp_mthd_list nv50_disp_mast_mthd_base;
-extern const struct nv50_disp_mthd_list nv50_disp_mast_mthd_sor;
-extern const struct nv50_disp_mthd_list nv50_disp_mast_mthd_pior;
-extern struct nv50_disp_chan_impl nv50_disp_sync_ofuncs;
-int nv50_disp_sync_ctor(struct nouveau_object *, struct nouveau_object *,
+extern const struct nv50_disp_mthd_list nv50_disp_core_mthd_base;
+extern const struct nv50_disp_mthd_list nv50_disp_core_mthd_sor;
+extern const struct nv50_disp_mthd_list nv50_disp_core_mthd_pior;
+extern struct nv50_disp_chan_impl nv50_disp_base_ofuncs;
+int nv50_disp_base_ctor(struct nouveau_object *, struct nouveau_object *,
 			struct nouveau_oclass *, void *, u32,
 			struct nouveau_object **);
-extern const struct nv50_disp_mthd_list nv50_disp_sync_mthd_image;
+extern const struct nv50_disp_mthd_list nv50_disp_base_mthd_image;
 extern struct nv50_disp_chan_impl nv50_disp_ovly_ofuncs;
 int nv50_disp_ovly_ctor(struct nouveau_object *, struct nouveau_object *,
 			struct nouveau_oclass *, void *, u32,
@@ -184,12 +195,12 @@ extern struct nv50_disp_chan_impl nv50_disp_curs_ofuncs;
 int nv50_disp_curs_ctor(struct nouveau_object *, struct nouveau_object *,
 			struct nouveau_oclass *, void *, u32,
 			struct nouveau_object **);
-extern struct nouveau_ofuncs nv50_disp_base_ofuncs;
-int  nv50_disp_base_ctor(struct nouveau_object *, struct nouveau_object *,
+extern struct nouveau_ofuncs nv50_disp_main_ofuncs;
+int  nv50_disp_main_ctor(struct nouveau_object *, struct nouveau_object *,
 			 struct nouveau_oclass *, void *, u32,
 			 struct nouveau_object **);
-void nv50_disp_base_dtor(struct nouveau_object *);
-extern struct nouveau_omthds nv50_disp_base_omthds[];
+void nv50_disp_main_dtor(struct nouveau_object *);
+extern struct nouveau_omthds nv50_disp_main_omthds[];
 extern struct nouveau_oclass nv50_disp_cclass;
 void nv50_disp_mthd_chan(struct nv50_disp_priv *, int debug, int head,
 			 const struct nv50_disp_mthd_chan *);
@@ -197,31 +208,31 @@ void nv50_disp_intr_supervisor(struct work_struct *);
 void nv50_disp_intr(struct nouveau_subdev *);
 extern const struct nvkm_event_func nv50_disp_vblank_func;
 
-extern const struct nv50_disp_mthd_chan nv84_disp_mast_mthd_chan;
-extern const struct nv50_disp_mthd_list nv84_disp_mast_mthd_dac;
-extern const struct nv50_disp_mthd_list nv84_disp_mast_mthd_head;
-extern const struct nv50_disp_mthd_chan nv84_disp_sync_mthd_chan;
+extern const struct nv50_disp_mthd_chan nv84_disp_core_mthd_chan;
+extern const struct nv50_disp_mthd_list nv84_disp_core_mthd_dac;
+extern const struct nv50_disp_mthd_list nv84_disp_core_mthd_head;
+extern const struct nv50_disp_mthd_chan nv84_disp_base_mthd_chan;
 extern const struct nv50_disp_mthd_chan nv84_disp_ovly_mthd_chan;
 
-extern const struct nv50_disp_mthd_chan nv94_disp_mast_mthd_chan;
+extern const struct nv50_disp_mthd_chan nv94_disp_core_mthd_chan;
 
-extern struct nv50_disp_chan_impl nvd0_disp_mast_ofuncs;
-extern const struct nv50_disp_mthd_list nvd0_disp_mast_mthd_base;
-extern const struct nv50_disp_mthd_list nvd0_disp_mast_mthd_dac;
-extern const struct nv50_disp_mthd_list nvd0_disp_mast_mthd_sor;
-extern const struct nv50_disp_mthd_list nvd0_disp_mast_mthd_pior;
-extern struct nv50_disp_chan_impl nvd0_disp_sync_ofuncs;
+extern struct nv50_disp_chan_impl nvd0_disp_core_ofuncs;
+extern const struct nv50_disp_mthd_list nvd0_disp_core_mthd_base;
+extern const struct nv50_disp_mthd_list nvd0_disp_core_mthd_dac;
+extern const struct nv50_disp_mthd_list nvd0_disp_core_mthd_sor;
+extern const struct nv50_disp_mthd_list nvd0_disp_core_mthd_pior;
+extern struct nv50_disp_chan_impl nvd0_disp_base_ofuncs;
 extern struct nv50_disp_chan_impl nvd0_disp_ovly_ofuncs;
-extern const struct nv50_disp_mthd_chan nvd0_disp_sync_mthd_chan;
+extern const struct nv50_disp_mthd_chan nvd0_disp_base_mthd_chan;
 extern struct nv50_disp_chan_impl nvd0_disp_oimm_ofuncs;
 extern struct nv50_disp_chan_impl nvd0_disp_curs_ofuncs;
-extern struct nouveau_ofuncs nvd0_disp_base_ofuncs;
+extern struct nouveau_ofuncs nvd0_disp_main_ofuncs;
 extern struct nouveau_oclass nvd0_disp_cclass;
 void nvd0_disp_intr_supervisor(struct work_struct *);
 void nvd0_disp_intr(struct nouveau_subdev *);
 extern const struct nvkm_event_func nvd0_disp_vblank_func;
 
-extern const struct nv50_disp_mthd_chan nve0_disp_mast_mthd_chan;
+extern const struct nv50_disp_mthd_chan nve0_disp_core_mthd_chan;
 extern const struct nv50_disp_mthd_chan nve0_disp_ovly_mthd_chan;
 
 extern struct nvkm_output_dp_impl nv50_pior_dp_impl;
@@ -232,6 +243,10 @@ int nv94_sor_dp_lnk_pwr(struct nvkm_output_dp *, int);
 extern struct nouveau_oclass *nv94_disp_outp_sclass[];
 
 extern struct nvkm_output_dp_impl nvd0_sor_dp_impl;
+int nvd0_sor_dp_lnk_ctl(struct nvkm_output_dp *, int, int, bool);
 extern struct nouveau_oclass *nvd0_disp_outp_sclass[];
 
+void gm204_sor_magic(struct nvkm_output *outp);
+extern struct nvkm_output_dp_impl gm204_sor_dp_impl;
+
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv84.c b/drivers/gpu/drm/nouveau/core/engine/disp/nv84.c
index 788ced1b6182..13eff5e4ee51 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv84.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv84.c
@@ -34,7 +34,7 @@
  ******************************************************************************/
 
 const struct nv50_disp_mthd_list
-nv84_disp_mast_mthd_dac = {
+nv84_disp_core_mthd_dac = {
 	.mthd = 0x0080,
 	.addr = 0x000008,
 	.data = {
@@ -46,7 +46,7 @@ nv84_disp_mast_mthd_dac = {
 };
 
 const struct nv50_disp_mthd_list
-nv84_disp_mast_mthd_head = {
+nv84_disp_core_mthd_head = {
 	.mthd = 0x0400,
 	.addr = 0x000540,
 	.data = {
@@ -98,15 +98,15 @@ nv84_disp_mast_mthd_head = {
 };
 
 const struct nv50_disp_mthd_chan
-nv84_disp_mast_mthd_chan = {
+nv84_disp_core_mthd_chan = {
 	.name = "Core",
 	.addr = 0x000000,
 	.data = {
-		{ "Global", 1, &nv50_disp_mast_mthd_base },
-		{    "DAC", 3, &nv84_disp_mast_mthd_dac  },
-		{    "SOR", 2, &nv50_disp_mast_mthd_sor  },
-		{   "PIOR", 3, &nv50_disp_mast_mthd_pior },
-		{   "HEAD", 2, &nv84_disp_mast_mthd_head },
+		{ "Global", 1, &nv50_disp_core_mthd_base },
+		{    "DAC", 3, &nv84_disp_core_mthd_dac  },
+		{    "SOR", 2, &nv50_disp_core_mthd_sor  },
+		{   "PIOR", 3, &nv50_disp_core_mthd_pior },
+		{   "HEAD", 2, &nv84_disp_core_mthd_head },
 		{}
 	}
 };
@@ -116,7 +116,7 @@ nv84_disp_mast_mthd_chan = {
  ******************************************************************************/
 
 static const struct nv50_disp_mthd_list
-nv84_disp_sync_mthd_base = {
+nv84_disp_base_mthd_base = {
 	.mthd = 0x0000,
 	.addr = 0x000000,
 	.data = {
@@ -146,12 +146,12 @@ nv84_disp_sync_mthd_base = {
 };
 
 const struct nv50_disp_mthd_chan
-nv84_disp_sync_mthd_chan = {
+nv84_disp_base_mthd_chan = {
 	.name = "Base",
 	.addr = 0x000540,
 	.data = {
-		{ "Global", 1, &nv84_disp_sync_mthd_base },
-		{  "Image", 2, &nv50_disp_sync_mthd_image },
+		{ "Global", 1, &nv84_disp_base_mthd_base },
+		{  "Image", 2, &nv50_disp_base_mthd_image },
 		{}
 	}
 };
@@ -204,8 +204,8 @@ nv84_disp_ovly_mthd_chan = {
 
 static struct nouveau_oclass
 nv84_disp_sclass[] = {
-	{ G82_DISP_CORE_CHANNEL_DMA, &nv50_disp_mast_ofuncs.base },
-	{ G82_DISP_BASE_CHANNEL_DMA, &nv50_disp_sync_ofuncs.base },
+	{ G82_DISP_CORE_CHANNEL_DMA, &nv50_disp_core_ofuncs.base },
+	{ G82_DISP_BASE_CHANNEL_DMA, &nv50_disp_base_ofuncs.base },
 	{ G82_DISP_OVERLAY_CHANNEL_DMA, &nv50_disp_ovly_ofuncs.base },
 	{ G82_DISP_OVERLAY, &nv50_disp_oimm_ofuncs.base },
 	{ G82_DISP_CURSOR, &nv50_disp_curs_ofuncs.base },
@@ -213,8 +213,8 @@ nv84_disp_sclass[] = {
 };
 
 static struct nouveau_oclass
-nv84_disp_base_oclass[] = {
-	{ G82_DISP, &nv50_disp_base_ofuncs },
+nv84_disp_main_oclass[] = {
+	{ G82_DISP, &nv50_disp_main_ofuncs },
 	{}
 };
 
@@ -236,7 +236,11 @@ nv84_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	nv_engine(priv)->sclass = nv84_disp_base_oclass;
+	ret = nvkm_event_init(&nv50_disp_chan_uevent, 1, 9, &priv->uevent);
+	if (ret)
+		return ret;
+
+	nv_engine(priv)->sclass = nv84_disp_main_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nv50_disp_intr;
 	INIT_WORK(&priv->supervisor, nv50_disp_intr_supervisor);
@@ -264,9 +268,9 @@ nv84_disp_oclass = &(struct nv50_disp_impl) {
 	},
 	.base.vblank = &nv50_disp_vblank_func,
 	.base.outp =  nv50_disp_outp_sclass,
-	.mthd.core = &nv84_disp_mast_mthd_chan,
-	.mthd.base = &nv84_disp_sync_mthd_chan,
+	.mthd.core = &nv84_disp_core_mthd_chan,
+	.mthd.base = &nv84_disp_base_mthd_chan,
 	.mthd.ovly = &nv84_disp_ovly_mthd_chan,
 	.mthd.prev = 0x000004,
-	.head.scanoutpos = nv50_disp_base_scanoutpos,
+	.head.scanoutpos = nv50_disp_main_scanoutpos,
 }.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv94.c b/drivers/gpu/drm/nouveau/core/engine/disp/nv94.c
index fa79de906eae..2bb7ac5cd0e6 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv94.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv94.c
@@ -34,7 +34,7 @@
  ******************************************************************************/
 
 const struct nv50_disp_mthd_list
-nv94_disp_mast_mthd_sor = {
+nv94_disp_core_mthd_sor = {
 	.mthd = 0x0040,
 	.addr = 0x000008,
 	.data = {
@@ -44,15 +44,15 @@ nv94_disp_mast_mthd_sor = {
 };
 
 const struct nv50_disp_mthd_chan
-nv94_disp_mast_mthd_chan = {
+nv94_disp_core_mthd_chan = {
 	.name = "Core",
 	.addr = 0x000000,
 	.data = {
-		{ "Global", 1, &nv50_disp_mast_mthd_base },
-		{    "DAC", 3, &nv84_disp_mast_mthd_dac  },
-		{    "SOR", 4, &nv94_disp_mast_mthd_sor  },
-		{   "PIOR", 3, &nv50_disp_mast_mthd_pior },
-		{   "HEAD", 2, &nv84_disp_mast_mthd_head },
+		{ "Global", 1, &nv50_disp_core_mthd_base },
+		{    "DAC", 3, &nv84_disp_core_mthd_dac  },
+		{    "SOR", 4, &nv94_disp_core_mthd_sor  },
+		{   "PIOR", 3, &nv50_disp_core_mthd_pior },
+		{   "HEAD", 2, &nv84_disp_core_mthd_head },
 		{}
 	}
 };
@@ -63,8 +63,8 @@ nv94_disp_mast_mthd_chan = {
 
 static struct nouveau_oclass
 nv94_disp_sclass[] = {
-	{ GT206_DISP_CORE_CHANNEL_DMA, &nv50_disp_mast_ofuncs.base },
-	{ GT200_DISP_BASE_CHANNEL_DMA, &nv50_disp_sync_ofuncs.base },
+	{ GT206_DISP_CORE_CHANNEL_DMA, &nv50_disp_core_ofuncs.base },
+	{ GT200_DISP_BASE_CHANNEL_DMA, &nv50_disp_base_ofuncs.base },
 	{ GT200_DISP_OVERLAY_CHANNEL_DMA, &nv50_disp_ovly_ofuncs.base },
 	{ G82_DISP_OVERLAY, &nv50_disp_oimm_ofuncs.base },
 	{ G82_DISP_CURSOR, &nv50_disp_curs_ofuncs.base },
@@ -72,8 +72,8 @@ nv94_disp_sclass[] = {
 };
 
 static struct nouveau_oclass
-nv94_disp_base_oclass[] = {
-	{ GT206_DISP, &nv50_disp_base_ofuncs },
+nv94_disp_main_oclass[] = {
+	{ GT206_DISP, &nv50_disp_main_ofuncs },
 	{}
 };
 
@@ -95,7 +95,11 @@ nv94_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	nv_engine(priv)->sclass = nv94_disp_base_oclass;
+	ret = nvkm_event_init(&nv50_disp_chan_uevent, 1, 9, &priv->uevent);
+	if (ret)
+		return ret;
+
+	nv_engine(priv)->sclass = nv94_disp_main_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nv50_disp_intr;
 	INIT_WORK(&priv->supervisor, nv50_disp_intr_supervisor);
@@ -130,9 +134,9 @@ nv94_disp_oclass = &(struct nv50_disp_impl) {
 	},
 	.base.vblank = &nv50_disp_vblank_func,
 	.base.outp =  nv94_disp_outp_sclass,
-	.mthd.core = &nv94_disp_mast_mthd_chan,
-	.mthd.base = &nv84_disp_sync_mthd_chan,
+	.mthd.core = &nv94_disp_core_mthd_chan,
+	.mthd.base = &nv84_disp_base_mthd_chan,
 	.mthd.ovly = &nv84_disp_ovly_mthd_chan,
 	.mthd.prev = 0x000004,
-	.head.scanoutpos = nv50_disp_base_scanoutpos,
+	.head.scanoutpos = nv50_disp_main_scanoutpos,
 }.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nva0.c b/drivers/gpu/drm/nouveau/core/engine/disp/nva0.c
index 7af15f5d48dc..b32456c9494f 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nva0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nva0.c
@@ -80,8 +80,8 @@ nva0_disp_ovly_mthd_chan = {
 
 static struct nouveau_oclass
 nva0_disp_sclass[] = {
-	{ GT200_DISP_CORE_CHANNEL_DMA, &nv50_disp_mast_ofuncs.base },
-	{ GT200_DISP_BASE_CHANNEL_DMA, &nv50_disp_sync_ofuncs.base },
+	{ GT200_DISP_CORE_CHANNEL_DMA, &nv50_disp_core_ofuncs.base },
+	{ GT200_DISP_BASE_CHANNEL_DMA, &nv50_disp_base_ofuncs.base },
 	{ GT200_DISP_OVERLAY_CHANNEL_DMA, &nv50_disp_ovly_ofuncs.base },
 	{ G82_DISP_OVERLAY, &nv50_disp_oimm_ofuncs.base },
 	{ G82_DISP_CURSOR, &nv50_disp_curs_ofuncs.base },
@@ -89,8 +89,8 @@ nva0_disp_sclass[] = {
 };
 
 static struct nouveau_oclass
-nva0_disp_base_oclass[] = {
-	{ GT200_DISP, &nv50_disp_base_ofuncs },
+nva0_disp_main_oclass[] = {
+	{ GT200_DISP, &nv50_disp_main_ofuncs },
 	{}
 };
 
@@ -112,7 +112,11 @@ nva0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	nv_engine(priv)->sclass = nva0_disp_base_oclass;
+	ret = nvkm_event_init(&nv50_disp_chan_uevent, 1, 9, &priv->uevent);
+	if (ret)
+		return ret;
+
+	nv_engine(priv)->sclass = nva0_disp_main_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nv50_disp_intr;
 	INIT_WORK(&priv->supervisor, nv50_disp_intr_supervisor);
@@ -140,9 +144,9 @@ nva0_disp_oclass = &(struct nv50_disp_impl) {
 	},
 	.base.vblank = &nv50_disp_vblank_func,
 	.base.outp =  nv50_disp_outp_sclass,
-	.mthd.core = &nv84_disp_mast_mthd_chan,
-	.mthd.base = &nv84_disp_sync_mthd_chan,
+	.mthd.core = &nv84_disp_core_mthd_chan,
+	.mthd.base = &nv84_disp_base_mthd_chan,
 	.mthd.ovly = &nva0_disp_ovly_mthd_chan,
 	.mthd.prev = 0x000004,
-	.head.scanoutpos = nv50_disp_base_scanoutpos,
+	.head.scanoutpos = nv50_disp_main_scanoutpos,
 }.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nva3.c b/drivers/gpu/drm/nouveau/core/engine/disp/nva3.c
index 6bd39448f8da..951d79f9b781 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nva3.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nva3.c
@@ -35,8 +35,8 @@
 
 static struct nouveau_oclass
 nva3_disp_sclass[] = {
-	{ GT214_DISP_CORE_CHANNEL_DMA, &nv50_disp_mast_ofuncs.base },
-	{ GT214_DISP_BASE_CHANNEL_DMA, &nv50_disp_sync_ofuncs.base },
+	{ GT214_DISP_CORE_CHANNEL_DMA, &nv50_disp_core_ofuncs.base },
+	{ GT214_DISP_BASE_CHANNEL_DMA, &nv50_disp_base_ofuncs.base },
 	{ GT214_DISP_OVERLAY_CHANNEL_DMA, &nv50_disp_ovly_ofuncs.base },
 	{ GT214_DISP_OVERLAY, &nv50_disp_oimm_ofuncs.base },
 	{ GT214_DISP_CURSOR, &nv50_disp_curs_ofuncs.base },
@@ -44,8 +44,8 @@ nva3_disp_sclass[] = {
 };
 
 static struct nouveau_oclass
-nva3_disp_base_oclass[] = {
-	{ GT214_DISP, &nv50_disp_base_ofuncs },
+nva3_disp_main_oclass[] = {
+	{ GT214_DISP, &nv50_disp_main_ofuncs },
 	{}
 };
 
@@ -67,7 +67,11 @@ nva3_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	nv_engine(priv)->sclass = nva3_disp_base_oclass;
+	ret = nvkm_event_init(&nv50_disp_chan_uevent, 1, 9, &priv->uevent);
+	if (ret)
+		return ret;
+
+	nv_engine(priv)->sclass = nva3_disp_main_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nv50_disp_intr;
 	INIT_WORK(&priv->supervisor, nv50_disp_intr_supervisor);
@@ -96,9 +100,9 @@ nva3_disp_oclass = &(struct nv50_disp_impl) {
 	},
 	.base.vblank = &nv50_disp_vblank_func,
 	.base.outp =  nv94_disp_outp_sclass,
-	.mthd.core = &nv94_disp_mast_mthd_chan,
-	.mthd.base = &nv84_disp_sync_mthd_chan,
+	.mthd.core = &nv94_disp_core_mthd_chan,
+	.mthd.base = &nv84_disp_base_mthd_chan,
 	.mthd.ovly = &nv84_disp_ovly_mthd_chan,
 	.mthd.prev = 0x000004,
-	.head.scanoutpos = nv50_disp_base_scanoutpos,
+	.head.scanoutpos = nv50_disp_main_scanoutpos,
 }.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c b/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c
index a4bb3c774ee1..181a2d57e356 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c
@@ -43,6 +43,33 @@
 #include "nv50.h"
 
 /*******************************************************************************
+ * EVO channel base class
+ ******************************************************************************/
+
+static void
+nvd0_disp_chan_uevent_fini(struct nvkm_event *event, int type, int index)
+{
+	struct nv50_disp_priv *priv = container_of(event, typeof(*priv), uevent);
+	nv_mask(priv, 0x610090, 0x00000001 << index, 0x00000000 << index);
+	nv_wr32(priv, 0x61008c, 0x00000001 << index);
+}
+
+static void
+nvd0_disp_chan_uevent_init(struct nvkm_event *event, int types, int index)
+{
+	struct nv50_disp_priv *priv = container_of(event, typeof(*priv), uevent);
+	nv_wr32(priv, 0x61008c, 0x00000001 << index);
+	nv_mask(priv, 0x610090, 0x00000001 << index, 0x00000001 << index);
+}
+
+const struct nvkm_event_func
+nvd0_disp_chan_uevent = {
+	.ctor = nv50_disp_chan_uevent_ctor,
+	.init = nvd0_disp_chan_uevent_init,
+	.fini = nvd0_disp_chan_uevent_fini,
+};
+
+/*******************************************************************************
  * EVO DMA channel base class
  ******************************************************************************/
 
@@ -77,7 +104,6 @@ nvd0_disp_dmac_init(struct nouveau_object *object)
 		return ret;
 
 	/* enable error reporting */
-	nv_mask(priv, 0x610090, 0x00000001 << chid, 0x00000001 << chid);
 	nv_mask(priv, 0x6100a0, 0x00000001 << chid, 0x00000001 << chid);
 
 	/* initialise channel for dma command submission */
@@ -115,7 +141,7 @@ nvd0_disp_dmac_fini(struct nouveau_object *object, bool suspend)
 			return -EBUSY;
 	}
 
-	/* disable error reporting */
+	/* disable error reporting and completion notification */
 	nv_mask(priv, 0x610090, 0x00000001 << chid, 0x00000000);
 	nv_mask(priv, 0x6100a0, 0x00000001 << chid, 0x00000000);
 
@@ -127,7 +153,7 @@ nvd0_disp_dmac_fini(struct nouveau_object *object, bool suspend)
  ******************************************************************************/
 
 const struct nv50_disp_mthd_list
-nvd0_disp_mast_mthd_base = {
+nvd0_disp_core_mthd_base = {
 	.mthd = 0x0000,
 	.addr = 0x000000,
 	.data = {
@@ -140,7 +166,7 @@ nvd0_disp_mast_mthd_base = {
 };
 
 const struct nv50_disp_mthd_list
-nvd0_disp_mast_mthd_dac = {
+nvd0_disp_core_mthd_dac = {
 	.mthd = 0x0020,
 	.addr = 0x000020,
 	.data = {
@@ -153,7 +179,7 @@ nvd0_disp_mast_mthd_dac = {
 };
 
 const struct nv50_disp_mthd_list
-nvd0_disp_mast_mthd_sor = {
+nvd0_disp_core_mthd_sor = {
 	.mthd = 0x0020,
 	.addr = 0x000020,
 	.data = {
@@ -166,7 +192,7 @@ nvd0_disp_mast_mthd_sor = {
 };
 
 const struct nv50_disp_mthd_list
-nvd0_disp_mast_mthd_pior = {
+nvd0_disp_core_mthd_pior = {
 	.mthd = 0x0020,
 	.addr = 0x000020,
 	.data = {
@@ -179,7 +205,7 @@ nvd0_disp_mast_mthd_pior = {
 };
 
 static const struct nv50_disp_mthd_list
-nvd0_disp_mast_mthd_head = {
+nvd0_disp_core_mthd_head = {
 	.mthd = 0x0300,
 	.addr = 0x000300,
 	.data = {
@@ -253,21 +279,21 @@ nvd0_disp_mast_mthd_head = {
 };
 
 static const struct nv50_disp_mthd_chan
-nvd0_disp_mast_mthd_chan = {
+nvd0_disp_core_mthd_chan = {
 	.name = "Core",
 	.addr = 0x000000,
 	.data = {
-		{ "Global", 1, &nvd0_disp_mast_mthd_base },
-		{    "DAC", 3, &nvd0_disp_mast_mthd_dac  },
-		{    "SOR", 8, &nvd0_disp_mast_mthd_sor  },
-		{   "PIOR", 4, &nvd0_disp_mast_mthd_pior },
-		{   "HEAD", 4, &nvd0_disp_mast_mthd_head },
+		{ "Global", 1, &nvd0_disp_core_mthd_base },
+		{    "DAC", 3, &nvd0_disp_core_mthd_dac  },
+		{    "SOR", 8, &nvd0_disp_core_mthd_sor  },
+		{   "PIOR", 4, &nvd0_disp_core_mthd_pior },
+		{   "HEAD", 4, &nvd0_disp_core_mthd_head },
 		{}
 	}
 };
 
 static int
-nvd0_disp_mast_init(struct nouveau_object *object)
+nvd0_disp_core_init(struct nouveau_object *object)
 {
 	struct nv50_disp_priv *priv = (void *)object->engine;
 	struct nv50_disp_dmac *mast = (void *)object;
@@ -278,7 +304,6 @@ nvd0_disp_mast_init(struct nouveau_object *object)
 		return ret;
 
 	/* enable error reporting */
-	nv_mask(priv, 0x610090, 0x00000001, 0x00000001);
 	nv_mask(priv, 0x6100a0, 0x00000001, 0x00000001);
 
 	/* initialise channel for dma command submission */
@@ -299,7 +324,7 @@ nvd0_disp_mast_init(struct nouveau_object *object)
 }
 
 static int
-nvd0_disp_mast_fini(struct nouveau_object *object, bool suspend)
+nvd0_disp_core_fini(struct nouveau_object *object, bool suspend)
 {
 	struct nv50_disp_priv *priv = (void *)object->engine;
 	struct nv50_disp_dmac *mast = (void *)object;
@@ -313,7 +338,7 @@ nvd0_disp_mast_fini(struct nouveau_object *object, bool suspend)
 			return -EBUSY;
 	}
 
-	/* disable error reporting */
+	/* disable error reporting and completion notification */
 	nv_mask(priv, 0x610090, 0x00000001, 0x00000000);
 	nv_mask(priv, 0x6100a0, 0x00000001, 0x00000000);
 
@@ -321,11 +346,12 @@ nvd0_disp_mast_fini(struct nouveau_object *object, bool suspend)
 }
 
 struct nv50_disp_chan_impl
-nvd0_disp_mast_ofuncs = {
-	.base.ctor = nv50_disp_mast_ctor,
+nvd0_disp_core_ofuncs = {
+	.base.ctor = nv50_disp_core_ctor,
 	.base.dtor = nv50_disp_dmac_dtor,
-	.base.init = nvd0_disp_mast_init,
-	.base.fini = nvd0_disp_mast_fini,
+	.base.init = nvd0_disp_core_init,
+	.base.fini = nvd0_disp_core_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -339,7 +365,7 @@ nvd0_disp_mast_ofuncs = {
  ******************************************************************************/
 
 static const struct nv50_disp_mthd_list
-nvd0_disp_sync_mthd_base = {
+nvd0_disp_base_mthd_base = {
 	.mthd = 0x0000,
 	.addr = 0x000000,
 	.data = {
@@ -389,7 +415,7 @@ nvd0_disp_sync_mthd_base = {
 };
 
 static const struct nv50_disp_mthd_list
-nvd0_disp_sync_mthd_image = {
+nvd0_disp_base_mthd_image = {
 	.mthd = 0x0400,
 	.addr = 0x000400,
 	.data = {
@@ -403,22 +429,23 @@ nvd0_disp_sync_mthd_image = {
 };
 
 const struct nv50_disp_mthd_chan
-nvd0_disp_sync_mthd_chan = {
+nvd0_disp_base_mthd_chan = {
 	.name = "Base",
 	.addr = 0x001000,
 	.data = {
-		{ "Global", 1, &nvd0_disp_sync_mthd_base },
-		{  "Image", 2, &nvd0_disp_sync_mthd_image },
+		{ "Global", 1, &nvd0_disp_base_mthd_base },
+		{  "Image", 2, &nvd0_disp_base_mthd_image },
 		{}
 	}
 };
 
 struct nv50_disp_chan_impl
-nvd0_disp_sync_ofuncs = {
-	.base.ctor = nv50_disp_sync_ctor,
+nvd0_disp_base_ofuncs = {
+	.base.ctor = nv50_disp_base_ctor,
 	.base.dtor = nv50_disp_dmac_dtor,
 	.base.init = nvd0_disp_dmac_init,
 	.base.fini = nvd0_disp_dmac_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -499,6 +526,7 @@ nvd0_disp_ovly_ofuncs = {
 	.base.dtor = nv50_disp_dmac_dtor,
 	.base.init = nvd0_disp_dmac_init,
 	.base.fini = nvd0_disp_dmac_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -524,7 +552,6 @@ nvd0_disp_pioc_init(struct nouveau_object *object)
 		return ret;
 
 	/* enable error reporting */
-	nv_mask(priv, 0x610090, 0x00000001 << chid, 0x00000001 << chid);
 	nv_mask(priv, 0x6100a0, 0x00000001 << chid, 0x00000001 << chid);
 
 	/* activate channel */
@@ -553,7 +580,7 @@ nvd0_disp_pioc_fini(struct nouveau_object *object, bool suspend)
 			return -EBUSY;
 	}
 
-	/* disable error reporting */
+	/* disable error reporting and completion notification */
 	nv_mask(priv, 0x610090, 0x00000001 << chid, 0x00000000);
 	nv_mask(priv, 0x6100a0, 0x00000001 << chid, 0x00000000);
 
@@ -570,6 +597,7 @@ nvd0_disp_oimm_ofuncs = {
 	.base.dtor = nv50_disp_pioc_dtor,
 	.base.init = nvd0_disp_pioc_init,
 	.base.fini = nvd0_disp_pioc_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -586,6 +614,7 @@ nvd0_disp_curs_ofuncs = {
 	.base.dtor = nv50_disp_pioc_dtor,
 	.base.init = nvd0_disp_pioc_init,
 	.base.fini = nvd0_disp_pioc_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -597,7 +626,7 @@ nvd0_disp_curs_ofuncs = {
  ******************************************************************************/
 
 int
-nvd0_disp_base_scanoutpos(NV50_DISP_MTHD_V0)
+nvd0_disp_main_scanoutpos(NV50_DISP_MTHD_V0)
 {
 	const u32 total  = nv_rd32(priv, 0x640414 + (head * 0x300));
 	const u32 blanke = nv_rd32(priv, 0x64041c + (head * 0x300));
@@ -629,7 +658,7 @@ nvd0_disp_base_scanoutpos(NV50_DISP_MTHD_V0)
 }
 
 static int
-nvd0_disp_base_init(struct nouveau_object *object)
+nvd0_disp_main_init(struct nouveau_object *object)
 {
 	struct nv50_disp_priv *priv = (void *)object->engine;
 	struct nv50_disp_base *base = (void *)object;
@@ -698,7 +727,7 @@ nvd0_disp_base_init(struct nouveau_object *object)
 }
 
 static int
-nvd0_disp_base_fini(struct nouveau_object *object, bool suspend)
+nvd0_disp_main_fini(struct nouveau_object *object, bool suspend)
 {
 	struct nv50_disp_priv *priv = (void *)object->engine;
 	struct nv50_disp_base *base = (void *)object;
@@ -710,25 +739,25 @@ nvd0_disp_base_fini(struct nouveau_object *object, bool suspend)
 }
 
 struct nouveau_ofuncs
-nvd0_disp_base_ofuncs = {
-	.ctor = nv50_disp_base_ctor,
-	.dtor = nv50_disp_base_dtor,
-	.init = nvd0_disp_base_init,
-	.fini = nvd0_disp_base_fini,
-	.mthd = nv50_disp_base_mthd,
+nvd0_disp_main_ofuncs = {
+	.ctor = nv50_disp_main_ctor,
+	.dtor = nv50_disp_main_dtor,
+	.init = nvd0_disp_main_init,
+	.fini = nvd0_disp_main_fini,
+	.mthd = nv50_disp_main_mthd,
 	.ntfy = nouveau_disp_ntfy,
 };
 
 static struct nouveau_oclass
-nvd0_disp_base_oclass[] = {
-	{ GF110_DISP, &nvd0_disp_base_ofuncs },
+nvd0_disp_main_oclass[] = {
+	{ GF110_DISP, &nvd0_disp_main_ofuncs },
 	{}
 };
 
 static struct nouveau_oclass
 nvd0_disp_sclass[] = {
-	{ GF110_DISP_CORE_CHANNEL_DMA, &nvd0_disp_mast_ofuncs.base },
-	{ GF110_DISP_BASE_CHANNEL_DMA, &nvd0_disp_sync_ofuncs.base },
+	{ GF110_DISP_CORE_CHANNEL_DMA, &nvd0_disp_core_ofuncs.base },
+	{ GF110_DISP_BASE_CHANNEL_DMA, &nvd0_disp_base_ofuncs.base },
 	{ GF110_DISP_OVERLAY_CONTROL_DMA, &nvd0_disp_ovly_ofuncs.base },
 	{ GF110_DISP_OVERLAY, &nvd0_disp_oimm_ofuncs.base },
 	{ GF110_DISP_CURSOR, &nvd0_disp_curs_ofuncs.base },
@@ -949,6 +978,9 @@ nvd0_disp_intr_unk2_2_tu(struct nv50_disp_priv *priv, int head,
 	const int or = ffs(outp->or) - 1;
 	const u32 ctrl = nv_rd32(priv, 0x660200 + (or   * 0x020));
 	const u32 conf = nv_rd32(priv, 0x660404 + (head * 0x300));
+	const s32 vactive = nv_rd32(priv, 0x660414 + (head * 0x300)) & 0xffff;
+	const s32 vblanke = nv_rd32(priv, 0x66041c + (head * 0x300)) & 0xffff;
+	const s32 vblanks = nv_rd32(priv, 0x660420 + (head * 0x300)) & 0xffff;
 	const u32 pclk = nv_rd32(priv, 0x660450 + (head * 0x300)) / 1000;
 	const u32 link = ((ctrl & 0xf00) == 0x800) ? 0 : 1;
 	const u32 hoff = (head * 0x800);
@@ -956,23 +988,35 @@ nvd0_disp_intr_unk2_2_tu(struct nv50_disp_priv *priv, int head,
 	const u32 loff = (link * 0x080) + soff;
 	const u32 symbol = 100000;
 	const u32 TU = 64;
-	u32 dpctrl = nv_rd32(priv, 0x61c10c + loff) & 0x000f0000;
+	u32 dpctrl = nv_rd32(priv, 0x61c10c + loff);
 	u32 clksor = nv_rd32(priv, 0x612300 + soff);
 	u32 datarate, link_nr, link_bw, bits;
 	u64 ratio, value;
 
+	link_nr  = hweight32(dpctrl & 0x000f0000);
+	link_bw  = (clksor & 0x007c0000) >> 18;
+	link_bw *= 27000;
+
+	/* symbols/hblank - algorithm taken from comments in tegra driver */
+	value = vblanke + vactive - vblanks - 7;
+	value = value * link_bw;
+	do_div(value, pclk);
+	value = value - (3 * !!(dpctrl & 0x00004000)) - (12 / link_nr);
+	nv_mask(priv, 0x616620 + hoff, 0x0000ffff, value);
+
+	/* symbols/vblank - algorithm taken from comments in tegra driver */
+	value = vblanks - vblanke - 25;
+	value = value * link_bw;
+	do_div(value, pclk);
+	value = value - ((36 / link_nr) + 3) - 1;
+	nv_mask(priv, 0x616624 + hoff, 0x00ffffff, value);
+
+	/* watermark */
 	if      ((conf & 0x3c0) == 0x180) bits = 30;
 	else if ((conf & 0x3c0) == 0x140) bits = 24;
 	else                              bits = 18;
 	datarate = (pclk * bits) / 8;
 
-	if      (dpctrl > 0x00030000) link_nr = 4;
-	else if (dpctrl > 0x00010000) link_nr = 2;
-	else			      link_nr = 1;
-
-	link_bw  = (clksor & 0x007c0000) >> 18;
-	link_bw *= 27000;
-
 	ratio  = datarate;
 	ratio *= symbol;
 	do_div(ratio, link_nr * link_bw);
@@ -1013,6 +1057,9 @@ nvd0_disp_intr_unk2_2(struct nv50_disp_priv *priv, int head)
 
 		if (nvkm_output_dp_train(outp, pclk, true))
 			ERR("link not trained before attach\n");
+	} else {
+		if (priv->sor.magic)
+			priv->sor.magic(outp);
 	}
 
 	exec_clkcmp(priv, head, 0, pclk, &conf);
@@ -1021,10 +1068,18 @@ nvd0_disp_intr_unk2_2(struct nv50_disp_priv *priv, int head)
 		addr = 0x612280 + (ffs(outp->info.or) - 1) * 0x800;
 		data = 0x00000000;
 	} else {
-		if (outp->info.type == DCB_OUTPUT_DP)
-			nvd0_disp_intr_unk2_2_tu(priv, head, &outp->info);
 		addr = 0x612300 + (ffs(outp->info.or) - 1) * 0x800;
 		data = (conf & 0x0100) ? 0x00000101 : 0x00000000;
+		switch (outp->info.type) {
+		case DCB_OUTPUT_TMDS:
+			nv_mask(priv, addr, 0x007c0000, 0x00280000);
+			break;
+		case DCB_OUTPUT_DP:
+			nvd0_disp_intr_unk2_2_tu(priv, head, &outp->info);
+			break;
+		default:
+			break;
+		}
 	}
 
 	nv_mask(priv, addr, 0x00000707, data);
@@ -1153,7 +1208,11 @@ nvd0_disp_intr(struct nouveau_subdev *subdev)
 
 	if (intr & 0x00000001) {
 		u32 stat = nv_rd32(priv, 0x61008c);
-		nv_wr32(priv, 0x61008c, stat);
+		while (stat) {
+			int chid = __ffs(stat); stat &= ~(1 << chid);
+			nv50_disp_chan_uevent_send(priv, chid);
+			nv_wr32(priv, 0x61008c, 1 << chid);
+		}
 		intr &= ~0x00000001;
 	}
 
@@ -1209,7 +1268,11 @@ nvd0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	nv_engine(priv)->sclass = nvd0_disp_base_oclass;
+	ret = nvkm_event_init(&nvd0_disp_chan_uevent, 1, 17, &priv->uevent);
+	if (ret)
+		return ret;
+
+	nv_engine(priv)->sclass = nvd0_disp_main_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nvd0_disp_intr;
 	INIT_WORK(&priv->supervisor, nvd0_disp_intr_supervisor);
@@ -1242,9 +1305,9 @@ nvd0_disp_oclass = &(struct nv50_disp_impl) {
 	},
 	.base.vblank = &nvd0_disp_vblank_func,
 	.base.outp =  nvd0_disp_outp_sclass,
-	.mthd.core = &nvd0_disp_mast_mthd_chan,
-	.mthd.base = &nvd0_disp_sync_mthd_chan,
+	.mthd.core = &nvd0_disp_core_mthd_chan,
+	.mthd.base = &nvd0_disp_base_mthd_chan,
 	.mthd.ovly = &nvd0_disp_ovly_mthd_chan,
 	.mthd.prev = -0x020000,
-	.head.scanoutpos = nvd0_disp_base_scanoutpos,
+	.head.scanoutpos = nvd0_disp_main_scanoutpos,
 }.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nve0.c b/drivers/gpu/drm/nouveau/core/engine/disp/nve0.c
index 47fef1e398c4..55debec7e68f 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nve0.c
@@ -34,7 +34,7 @@
  ******************************************************************************/
 
 static const struct nv50_disp_mthd_list
-nve0_disp_mast_mthd_head = {
+nve0_disp_core_mthd_head = {
 	.mthd = 0x0300,
 	.addr = 0x000300,
 	.data = {
@@ -113,15 +113,15 @@ nve0_disp_mast_mthd_head = {
 };
 
 const struct nv50_disp_mthd_chan
-nve0_disp_mast_mthd_chan = {
+nve0_disp_core_mthd_chan = {
 	.name = "Core",
 	.addr = 0x000000,
 	.data = {
-		{ "Global", 1, &nvd0_disp_mast_mthd_base },
-		{    "DAC", 3, &nvd0_disp_mast_mthd_dac  },
-		{    "SOR", 8, &nvd0_disp_mast_mthd_sor  },
-		{   "PIOR", 4, &nvd0_disp_mast_mthd_pior },
-		{   "HEAD", 4, &nve0_disp_mast_mthd_head },
+		{ "Global", 1, &nvd0_disp_core_mthd_base },
+		{    "DAC", 3, &nvd0_disp_core_mthd_dac  },
+		{    "SOR", 8, &nvd0_disp_core_mthd_sor  },
+		{   "PIOR", 4, &nvd0_disp_core_mthd_pior },
+		{   "HEAD", 4, &nve0_disp_core_mthd_head },
 		{}
 	}
 };
@@ -200,8 +200,8 @@ nve0_disp_ovly_mthd_chan = {
 
 static struct nouveau_oclass
 nve0_disp_sclass[] = {
-	{ GK104_DISP_CORE_CHANNEL_DMA, &nvd0_disp_mast_ofuncs.base },
-	{ GK104_DISP_BASE_CHANNEL_DMA, &nvd0_disp_sync_ofuncs.base },
+	{ GK104_DISP_CORE_CHANNEL_DMA, &nvd0_disp_core_ofuncs.base },
+	{ GK104_DISP_BASE_CHANNEL_DMA, &nvd0_disp_base_ofuncs.base },
 	{ GK104_DISP_OVERLAY_CONTROL_DMA, &nvd0_disp_ovly_ofuncs.base },
 	{ GK104_DISP_OVERLAY, &nvd0_disp_oimm_ofuncs.base },
 	{ GK104_DISP_CURSOR, &nvd0_disp_curs_ofuncs.base },
@@ -209,8 +209,8 @@ nve0_disp_sclass[] = {
 };
 
 static struct nouveau_oclass
-nve0_disp_base_oclass[] = {
-	{ GK104_DISP, &nvd0_disp_base_ofuncs },
+nve0_disp_main_oclass[] = {
+	{ GK104_DISP, &nvd0_disp_main_ofuncs },
 	{}
 };
 
@@ -233,7 +233,11 @@ nve0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	nv_engine(priv)->sclass = nve0_disp_base_oclass;
+	ret = nvkm_event_init(&nvd0_disp_chan_uevent, 1, 17, &priv->uevent);
+	if (ret)
+		return ret;
+
+	nv_engine(priv)->sclass = nve0_disp_main_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nvd0_disp_intr;
 	INIT_WORK(&priv->supervisor, nvd0_disp_intr_supervisor);
@@ -245,7 +249,7 @@ nve0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	priv->dac.sense = nv50_dac_sense;
 	priv->sor.power = nv50_sor_power;
 	priv->sor.hda_eld = nvd0_hda_eld;
-	priv->sor.hdmi = nvd0_hdmi_ctrl;
+	priv->sor.hdmi = nve0_hdmi_ctrl;
 	return 0;
 }
 
@@ -260,9 +264,9 @@ nve0_disp_oclass = &(struct nv50_disp_impl) {
 	},
 	.base.vblank = &nvd0_disp_vblank_func,
 	.base.outp =  nvd0_disp_outp_sclass,
-	.mthd.core = &nve0_disp_mast_mthd_chan,
-	.mthd.base = &nvd0_disp_sync_mthd_chan,
+	.mthd.core = &nve0_disp_core_mthd_chan,
+	.mthd.base = &nvd0_disp_base_mthd_chan,
 	.mthd.ovly = &nve0_disp_ovly_mthd_chan,
 	.mthd.prev = -0x020000,
-	.head.scanoutpos = nvd0_disp_base_scanoutpos,
+	.head.scanoutpos = nvd0_disp_main_scanoutpos,
 }.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nvf0.c b/drivers/gpu/drm/nouveau/core/engine/disp/nvf0.c
index 04bda4ac4ed3..3e7e2d28744c 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nvf0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nvf0.c
@@ -35,8 +35,8 @@
 
 static struct nouveau_oclass
 nvf0_disp_sclass[] = {
-	{ GK110_DISP_CORE_CHANNEL_DMA, &nvd0_disp_mast_ofuncs.base },
-	{ GK110_DISP_BASE_CHANNEL_DMA, &nvd0_disp_sync_ofuncs.base },
+	{ GK110_DISP_CORE_CHANNEL_DMA, &nvd0_disp_core_ofuncs.base },
+	{ GK110_DISP_BASE_CHANNEL_DMA, &nvd0_disp_base_ofuncs.base },
 	{ GK104_DISP_OVERLAY_CONTROL_DMA, &nvd0_disp_ovly_ofuncs.base },
 	{ GK104_DISP_OVERLAY, &nvd0_disp_oimm_ofuncs.base },
 	{ GK104_DISP_CURSOR, &nvd0_disp_curs_ofuncs.base },
@@ -44,8 +44,8 @@ nvf0_disp_sclass[] = {
 };
 
 static struct nouveau_oclass
-nvf0_disp_base_oclass[] = {
-	{ GK110_DISP, &nvd0_disp_base_ofuncs },
+nvf0_disp_main_oclass[] = {
+	{ GK110_DISP, &nvd0_disp_main_ofuncs },
 	{}
 };
 
@@ -68,7 +68,11 @@ nvf0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	nv_engine(priv)->sclass = nvf0_disp_base_oclass;
+	ret = nvkm_event_init(&nvd0_disp_chan_uevent, 1, 17, &priv->uevent);
+	if (ret)
+		return ret;
+
+	nv_engine(priv)->sclass = nvf0_disp_main_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nvd0_disp_intr;
 	INIT_WORK(&priv->supervisor, nvd0_disp_intr_supervisor);
@@ -80,7 +84,7 @@ nvf0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	priv->dac.sense = nv50_dac_sense;
 	priv->sor.power = nv50_sor_power;
 	priv->sor.hda_eld = nvd0_hda_eld;
-	priv->sor.hdmi = nvd0_hdmi_ctrl;
+	priv->sor.hdmi = nve0_hdmi_ctrl;
 	return 0;
 }
 
@@ -95,9 +99,9 @@ nvf0_disp_oclass = &(struct nv50_disp_impl) {
 	},
 	.base.vblank = &nvd0_disp_vblank_func,
 	.base.outp =  nvd0_disp_outp_sclass,
-	.mthd.core = &nve0_disp_mast_mthd_chan,
-	.mthd.base = &nvd0_disp_sync_mthd_chan,
+	.mthd.core = &nve0_disp_core_mthd_chan,
+	.mthd.base = &nvd0_disp_base_mthd_chan,
 	.mthd.ovly = &nve0_disp_ovly_mthd_chan,
 	.mthd.prev = -0x020000,
-	.head.scanoutpos = nvd0_disp_base_scanoutpos,
+	.head.scanoutpos = nvd0_disp_main_scanoutpos,
 }.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/outp.c b/drivers/gpu/drm/nouveau/core/engine/disp/outp.c
index a5ff00a9cedc..bbd9b6fdc90f 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/outp.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/outp.c
@@ -85,7 +85,10 @@ nvkm_output_create_(struct nouveau_object *parent,
 	    dcbE->sorconf.link : 0, dcbE->connector, dcbE->i2c_index,
 	    dcbE->bus, dcbE->heads);
 
-	outp->port = i2c->find(i2c, outp->info.i2c_index);
+	if (outp->info.type != DCB_OUTPUT_DP)
+		outp->port = i2c->find(i2c, NV_I2C_PORT(outp->info.i2c_index));
+	else
+		outp->port = i2c->find(i2c, NV_I2C_AUX(outp->info.i2c_index));
 	outp->edid = outp->port;
 
 	data = nvbios_connEp(bios, outp->info.connector, &ver, &hdr, &connE);
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/outpdp.c b/drivers/gpu/drm/nouveau/core/engine/disp/outpdp.c
index 6f6e2a898270..667a9070e006 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/outpdp.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/outpdp.c
@@ -254,7 +254,7 @@ nvkm_output_dp_create_(struct nouveau_object *parent,
 	atomic_set(&outp->lt.done, 0);
 
 	/* link maintenance */
-	ret = nvkm_notify_init(&i2c->event, nvkm_output_dp_irq, true,
+	ret = nvkm_notify_init(NULL, &i2c->event, nvkm_output_dp_irq, true,
 			       &(struct nvkm_i2c_ntfy_req) {
 				.mask = NVKM_I2C_IRQ,
 				.port = outp->base.edid->index,
@@ -268,7 +268,7 @@ nvkm_output_dp_create_(struct nouveau_object *parent,
 	}
 
 	/* hotplug detect, replaces gpio-based mechanism with aux events */
-	ret = nvkm_notify_init(&i2c->event, nvkm_output_dp_hpd, true,
+	ret = nvkm_notify_init(NULL, &i2c->event, nvkm_output_dp_hpd, true,
 			       &(struct nvkm_i2c_ntfy_req) {
 				.mask = NVKM_I2C_PLUG | NVKM_I2C_UNPLUG,
 				.port = outp->base.edid->index,
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/priv.h b/drivers/gpu/drm/nouveau/core/engine/disp/priv.h
index dbd43ae9df81..6a0511d54ce6 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/priv.h
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/priv.h
@@ -40,7 +40,8 @@ int  _nouveau_disp_fini(struct nouveau_object *, bool);
 extern struct nouveau_oclass *nvkm_output_oclass;
 extern struct nouveau_oclass *nvkm_connector_oclass;
 
-int  nouveau_disp_vblank_ctor(void *data, u32 size, struct nvkm_notify *);
+int  nouveau_disp_vblank_ctor(struct nouveau_object *, void *data, u32 size,
+			      struct nvkm_notify *);
 void nouveau_disp_vblank(struct nouveau_disp *, int head);
 int  nouveau_disp_ntfy(struct nouveau_object *, u32, struct nvkm_event **);
 
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/sorgm204.c b/drivers/gpu/drm/nouveau/core/engine/disp/sorgm204.c
new file mode 100644
index 000000000000..0b4fad39e9a6
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/sorgm204.c
@@ -0,0 +1,144 @@
+/*
+ * Copyright 2012 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include <core/os.h>
+
+#include <subdev/bios.h>
+#include <subdev/bios/dcb.h>
+#include <subdev/bios/dp.h>
+#include <subdev/bios/init.h>
+#include <subdev/timer.h>
+
+#include "nv50.h"
+
+static inline u32
+gm204_sor_soff(struct nvkm_output_dp *outp)
+{
+	return (ffs(outp->base.info.or) - 1) * 0x800;
+}
+
+static inline u32
+gm204_sor_loff(struct nvkm_output_dp *outp)
+{
+	return gm204_sor_soff(outp) + !(outp->base.info.sorconf.link & 1) * 0x80;
+}
+
+void
+gm204_sor_magic(struct nvkm_output *outp)
+{
+	struct nv50_disp_priv *priv = (void *)nouveau_disp(outp);
+	const u32 soff = outp->or * 0x100;
+	const u32 data = outp->or + 1;
+	if (outp->info.sorconf.link & 1)
+		nv_mask(priv, 0x612308 + soff, 0x0000001f, 0x00000000 | data);
+	if (outp->info.sorconf.link & 2)
+		nv_mask(priv, 0x612388 + soff, 0x0000001f, 0x00000010 | data);
+}
+
+static inline u32
+gm204_sor_dp_lane_map(struct nv50_disp_priv *priv, u8 lane)
+{
+	return lane * 0x08;
+}
+
+static int
+gm204_sor_dp_pattern(struct nvkm_output_dp *outp, int pattern)
+{
+	struct nv50_disp_priv *priv = (void *)nouveau_disp(outp);
+	const u32 soff = gm204_sor_soff(outp);
+	const u32 data = 0x01010101 * pattern;
+	if (outp->base.info.sorconf.link & 1)
+		nv_mask(priv, 0x61c110 + soff, 0x0f0f0f0f, data);
+	else
+		nv_mask(priv, 0x61c12c + soff, 0x0f0f0f0f, data);
+	return 0;
+}
+
+static int
+gm204_sor_dp_lnk_pwr(struct nvkm_output_dp *outp, int nr)
+{
+	struct nv50_disp_priv *priv = (void *)nouveau_disp(outp);
+	const u32 soff = gm204_sor_soff(outp);
+	const u32 loff = gm204_sor_loff(outp);
+	u32 mask = 0, i;
+
+	for (i = 0; i < nr; i++)
+		mask |= 1 << (gm204_sor_dp_lane_map(priv, i) >> 3);
+
+	nv_mask(priv, 0x61c130 + loff, 0x0000000f, mask);
+	nv_mask(priv, 0x61c034 + soff, 0x80000000, 0x80000000);
+	nv_wait(priv, 0x61c034 + soff, 0x80000000, 0x00000000);
+	return 0;
+}
+
+static int
+gm204_sor_dp_drv_ctl(struct nvkm_output_dp *outp, int ln, int vs, int pe, int pc)
+{
+	struct nv50_disp_priv *priv = (void *)nouveau_disp(outp);
+	struct nouveau_bios *bios = nouveau_bios(priv);
+	const u32 shift = gm204_sor_dp_lane_map(priv, ln);
+	const u32 loff = gm204_sor_loff(outp);
+	u32 addr, data[4];
+	u8  ver, hdr, cnt, len;
+	struct nvbios_dpout info;
+	struct nvbios_dpcfg ocfg;
+
+	addr = nvbios_dpout_match(bios, outp->base.info.hasht,
+					outp->base.info.hashm,
+				 &ver, &hdr, &cnt, &len, &info);
+	if (!addr)
+		return -ENODEV;
+
+	addr = nvbios_dpcfg_match(bios, addr, pc, vs, pe,
+				 &ver, &hdr, &cnt, &len, &ocfg);
+	if (!addr)
+		return -EINVAL;
+
+	data[0] = nv_rd32(priv, 0x61c118 + loff) & ~(0x000000ff << shift);
+	data[1] = nv_rd32(priv, 0x61c120 + loff) & ~(0x000000ff << shift);
+	data[2] = nv_rd32(priv, 0x61c130 + loff);
+	if ((data[2] & 0x0000ff00) < (ocfg.tx_pu << 8) || ln == 0)
+		data[2] = (data[2] & ~0x0000ff00) | (ocfg.tx_pu << 8);
+	nv_wr32(priv, 0x61c118 + loff, data[0] | (ocfg.dc << shift));
+	nv_wr32(priv, 0x61c120 + loff, data[1] | (ocfg.pe << shift));
+	nv_wr32(priv, 0x61c130 + loff, data[2] | (ocfg.tx_pu << 8));
+	data[3] = nv_rd32(priv, 0x61c13c + loff) & ~(0x000000ff << shift);
+	nv_wr32(priv, 0x61c13c + loff, data[3] | (ocfg.pc << shift));
+	return 0;
+}
+
+struct nvkm_output_dp_impl
+gm204_sor_dp_impl = {
+	.base.base.handle = DCB_OUTPUT_DP,
+	.base.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = _nvkm_output_dp_ctor,
+		.dtor = _nvkm_output_dp_dtor,
+		.init = _nvkm_output_dp_init,
+		.fini = _nvkm_output_dp_fini,
+	},
+	.pattern = gm204_sor_dp_pattern,
+	.lnk_pwr = gm204_sor_dp_lnk_pwr,
+	.lnk_ctl = nvd0_sor_dp_lnk_ctl,
+	.drv_ctl = gm204_sor_dp_drv_ctl,
+};
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/sornvd0.c b/drivers/gpu/drm/nouveau/core/engine/disp/sornvd0.c
index 7b7bbc3e459e..fdab2939070c 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/sornvd0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/sornvd0.c
@@ -60,7 +60,7 @@ nvd0_sor_dp_pattern(struct nvkm_output_dp *outp, int pattern)
 	return 0;
 }
 
-static int
+int
 nvd0_sor_dp_lnk_ctl(struct nvkm_output_dp *outp, int nr, int bw, bool ef)
 {
 	struct nv50_disp_priv *priv = (void *)nouveau_disp(outp);
diff --git a/drivers/gpu/drm/nouveau/core/engine/dmaobj/nvd0.c b/drivers/gpu/drm/nouveau/core/engine/dmaobj/nvd0.c
index 3fc4f0b0eaca..19f5f6522962 100644
--- a/drivers/gpu/drm/nouveau/core/engine/dmaobj/nvd0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/dmaobj/nvd0.c
@@ -51,6 +51,7 @@ nvd0_dmaobj_bind(struct nouveau_dmaobj *dmaobj,
 		case GK104_DISP_CORE_CHANNEL_DMA:
 		case GK110_DISP_CORE_CHANNEL_DMA:
 		case GM107_DISP_CORE_CHANNEL_DMA:
+		case GM204_DISP_CORE_CHANNEL_DMA:
 		case GF110_DISP_BASE_CHANNEL_DMA:
 		case GK104_DISP_BASE_CHANNEL_DMA:
 		case GK110_DISP_BASE_CHANNEL_DMA:
diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/base.c b/drivers/gpu/drm/nouveau/core/engine/fifo/base.c
index 0f999fc45ab9..ac8375cf4eef 100644
--- a/drivers/gpu/drm/nouveau/core/engine/fifo/base.c
+++ b/drivers/gpu/drm/nouveau/core/engine/fifo/base.c
@@ -34,7 +34,8 @@
 #include <engine/fifo.h>
 
 static int
-nouveau_fifo_event_ctor(void *data, u32 size, struct nvkm_notify *notify)
+nouveau_fifo_event_ctor(struct nouveau_object *object, void *data, u32 size,
+			struct nvkm_notify *notify)
 {
 	if (size == 0) {
 		notify->size  = 0;
@@ -170,7 +171,8 @@ _nouveau_fifo_channel_wr32(struct nouveau_object *object, u64 addr, u32 data)
 }
 
 int
-nouveau_fifo_uevent_ctor(void *data, u32 size, struct nvkm_notify *notify)
+nouveau_fifo_uevent_ctor(struct nouveau_object *object, void *data, u32 size,
+			 struct nvkm_notify *notify)
 {
 	union {
 		struct nvif_notify_uevent_req none;
diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/nv04.c b/drivers/gpu/drm/nouveau/core/engine/fifo/nv04.c
index 5ae6a43893b5..1931057f9962 100644
--- a/drivers/gpu/drm/nouveau/core/engine/fifo/nv04.c
+++ b/drivers/gpu/drm/nouveau/core/engine/fifo/nv04.c
@@ -551,8 +551,8 @@ nv04_fifo_intr(struct nouveau_subdev *subdev)
 			}
 
 			if (status & 0x40000000) {
-				nouveau_fifo_uevent(&priv->base);
 				nv_wr32(priv, 0x002100, 0x40000000);
+				nouveau_fifo_uevent(&priv->base);
 				status &= ~0x40000000;
 			}
 		}
diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/nvc0.c b/drivers/gpu/drm/nouveau/core/engine/fifo/nvc0.c
index 1fe1f8fbda0c..074d434c3077 100644
--- a/drivers/gpu/drm/nouveau/core/engine/fifo/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/fifo/nvc0.c
@@ -740,6 +740,8 @@ nvc0_fifo_intr_engine_unit(struct nvc0_fifo_priv *priv, int engn)
 	u32 inte = nv_rd32(priv, 0x002628);
 	u32 unkn;
 
+	nv_wr32(priv, 0x0025a8 + (engn * 0x04), intr);
+
 	for (unkn = 0; unkn < 8; unkn++) {
 		u32 ints = (intr >> (unkn * 0x04)) & inte;
 		if (ints & 0x1) {
@@ -751,8 +753,6 @@ nvc0_fifo_intr_engine_unit(struct nvc0_fifo_priv *priv, int engn)
 			nv_mask(priv, 0x002628, ints, 0);
 		}
 	}
-
-	nv_wr32(priv, 0x0025a8 + (engn * 0x04), intr);
 }
 
 static void
diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.c b/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.c
index d2f0fd39c145..6a8db7c80bd1 100644
--- a/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.c
@@ -792,7 +792,7 @@ nve0_fifo_intr_fault(struct nve0_fifo_priv *priv, int unit)
 	nouveau_engctx_put(engctx);
 }
 
-static const struct nouveau_bitfield nve0_fifo_pbdma_intr[] = {
+static const struct nouveau_bitfield nve0_fifo_pbdma_intr_0[] = {
 	{ 0x00000001, "MEMREQ" },
 	{ 0x00000002, "MEMACK_TIMEOUT" },
 	{ 0x00000004, "MEMACK_EXTRA" },
@@ -827,9 +827,10 @@ static const struct nouveau_bitfield nve0_fifo_pbdma_intr[] = {
 };
 
 static void
-nve0_fifo_intr_pbdma(struct nve0_fifo_priv *priv, int unit)
+nve0_fifo_intr_pbdma_0(struct nve0_fifo_priv *priv, int unit)
 {
-	u32 stat = nv_rd32(priv, 0x040108 + (unit * 0x2000));
+	u32 mask = nv_rd32(priv, 0x04010c + (unit * 0x2000));
+	u32 stat = nv_rd32(priv, 0x040108 + (unit * 0x2000)) & mask;
 	u32 addr = nv_rd32(priv, 0x0400c0 + (unit * 0x2000));
 	u32 data = nv_rd32(priv, 0x0400c4 + (unit * 0x2000));
 	u32 chid = nv_rd32(priv, 0x040120 + (unit * 0x2000)) & 0xfff;
@@ -840,11 +841,12 @@ nve0_fifo_intr_pbdma(struct nve0_fifo_priv *priv, int unit)
 	if (stat & 0x00800000) {
 		if (!nve0_fifo_swmthd(priv, chid, mthd, data))
 			show &= ~0x00800000;
+		nv_wr32(priv, 0x0400c0 + (unit * 0x2000), 0x80600008);
 	}
 
 	if (show) {
 		nv_error(priv, "PBDMA%d:", unit);
-		nouveau_bitfield_print(nve0_fifo_pbdma_intr, show);
+		nouveau_bitfield_print(nve0_fifo_pbdma_intr_0, show);
 		pr_cont("\n");
 		nv_error(priv,
 			 "PBDMA%d: ch %d [%s] subc %d mthd 0x%04x data 0x%08x\n",
@@ -853,10 +855,37 @@ nve0_fifo_intr_pbdma(struct nve0_fifo_priv *priv, int unit)
 			 subc, mthd, data);
 	}
 
-	nv_wr32(priv, 0x0400c0 + (unit * 0x2000), 0x80600008);
 	nv_wr32(priv, 0x040108 + (unit * 0x2000), stat);
 }
 
+static const struct nouveau_bitfield nve0_fifo_pbdma_intr_1[] = {
+	{ 0x00000001, "HCE_RE_ILLEGAL_OP" },
+	{ 0x00000002, "HCE_RE_ALIGNB" },
+	{ 0x00000004, "HCE_PRIV" },
+	{ 0x00000008, "HCE_ILLEGAL_MTHD" },
+	{ 0x00000010, "HCE_ILLEGAL_CLASS" },
+	{}
+};
+
+static void
+nve0_fifo_intr_pbdma_1(struct nve0_fifo_priv *priv, int unit)
+{
+	u32 mask = nv_rd32(priv, 0x04014c + (unit * 0x2000));
+	u32 stat = nv_rd32(priv, 0x040148 + (unit * 0x2000)) & mask;
+	u32 chid = nv_rd32(priv, 0x040120 + (unit * 0x2000)) & 0xfff;
+
+	if (stat) {
+		nv_error(priv, "PBDMA%d:", unit);
+		nouveau_bitfield_print(nve0_fifo_pbdma_intr_1, stat);
+		pr_cont("\n");
+		nv_error(priv, "PBDMA%d: ch %d %08x %08x\n", unit, chid,
+			 nv_rd32(priv, 0x040150 + (unit * 0x2000)),
+			 nv_rd32(priv, 0x040154 + (unit * 0x2000)));
+	}
+
+	nv_wr32(priv, 0x040148 + (unit * 0x2000), stat);
+}
+
 static void
 nve0_fifo_intr_runlist(struct nve0_fifo_priv *priv)
 {
@@ -939,7 +968,8 @@ nve0_fifo_intr(struct nouveau_subdev *subdev)
 		u32 mask = nv_rd32(priv, 0x0025a0);
 		while (mask) {
 			u32 unit = __ffs(mask);
-			nve0_fifo_intr_pbdma(priv, unit);
+			nve0_fifo_intr_pbdma_0(priv, unit);
+			nve0_fifo_intr_pbdma_1(priv, unit);
 			nv_wr32(priv, 0x0025a0, (1 << unit));
 			mask &= ~(1 << unit);
 		}
@@ -952,8 +982,8 @@ nve0_fifo_intr(struct nouveau_subdev *subdev)
 	}
 
 	if (stat & 0x80000000) {
-		nve0_fifo_intr_engine(priv);
 		nv_wr32(priv, 0x002100, 0x80000000);
+		nve0_fifo_intr_engine(priv);
 		stat &= ~0x80000000;
 	}
 
@@ -1022,6 +1052,12 @@ nve0_fifo_init(struct nouveau_object *object)
 		nv_wr32(priv, 0x04010c + (i * 0x2000), 0xfffffeff); /* INTREN */
 	}
 
+	/* PBDMA[n].HCE */
+	for (i = 0; i < priv->spoon_nr; i++) {
+		nv_wr32(priv, 0x040148 + (i * 0x2000), 0xffffffff); /* INTR */
+		nv_wr32(priv, 0x04014c + (i * 0x2000), 0xffffffff); /* INTREN */
+	}
+
 	nv_wr32(priv, 0x002254, 0x10000000 | priv->user.bar.offset >> 12);
 
 	nv_wr32(priv, 0x002100, 0xffffffff);
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnv50.c b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnv50.c
index 552fdbd45ebe..1d0e33fb5f61 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnv50.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnv50.c
@@ -113,6 +113,8 @@
 #define IS_NVA3F(x) (((x) > 0xa0 && (x) < 0xaa) || (x) == 0xaf)
 #define IS_NVAAF(x) ((x) >= 0xaa && (x) <= 0xac)
 
+#include <subdev/fb.h>
+
 /*
  * This code deals with PGRAPH contexts on NV50 family cards. Like NV40, it's
  * the GPU itself that does context-switching, but it needs a special
@@ -569,8 +571,12 @@ nv50_graph_construct_mmio(struct nouveau_grctx *ctx)
 			gr_def(ctx, 0x407d08, 0x00010040);
 		else if (device->chipset < 0xa0)
 			gr_def(ctx, 0x407d08, 0x00390040);
-		else
-			gr_def(ctx, 0x407d08, 0x003d0040);
+		else {
+			if (nouveau_fb(device)->ram->type != NV_MEM_TYPE_GDDR5)
+				gr_def(ctx, 0x407d08, 0x003d0040);
+			else
+				gr_def(ctx, 0x407d08, 0x003c0040);
+		}
 		gr_def(ctx, 0x407d0c, 0x00000022);
 	}
 
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.c b/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.c
index 30fd1dc64f93..17251e4b9e86 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.c
@@ -1557,7 +1557,7 @@ nvc0_graph_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		    nvc0_graph_ctor_fw(priv, "fuc409d", &priv->fuc409d) ||
 		    nvc0_graph_ctor_fw(priv, "fuc41ac", &priv->fuc41ac) ||
 		    nvc0_graph_ctor_fw(priv, "fuc41ad", &priv->fuc41ad))
-			return -EINVAL;
+			return -ENODEV;
 		priv->firmware = true;
 	}
 
diff --git a/drivers/gpu/drm/nouveau/core/engine/software/nv50.c b/drivers/gpu/drm/nouveau/core/engine/software/nv50.c
index 4d2994d8cc32..a0fec205f9db 100644
--- a/drivers/gpu/drm/nouveau/core/engine/software/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/engine/software/nv50.c
@@ -175,7 +175,8 @@ nv50_software_context_ctor(struct nouveau_object *parent,
 		return ret;
 
 	for (i = 0; pdisp && i < pdisp->vblank.index_nr; i++) {
-		ret = nvkm_notify_init(&pdisp->vblank, pclass->vblank, false,
+		ret = nvkm_notify_init(NULL, &pdisp->vblank, pclass->vblank,
+				       false,
 				       &(struct nvif_notify_head_req_v0) {
 					.head = i,
 				       },
diff --git a/drivers/gpu/drm/nouveau/core/include/core/client.h b/drivers/gpu/drm/nouveau/core/include/core/client.h
index 1794a05205d8..b0ce9f6680b5 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/client.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/client.h
@@ -48,7 +48,7 @@ int  nouveau_client_init(struct nouveau_client *);
 int  nouveau_client_fini(struct nouveau_client *, bool suspend);
 const char *nouveau_client_name(void *obj);
 
-int nvkm_client_notify_new(struct nouveau_client *, struct nvkm_event *,
+int nvkm_client_notify_new(struct nouveau_object *, struct nvkm_event *,
 			   void *data, u32 size);
 int nvkm_client_notify_del(struct nouveau_client *, int index);
 int nvkm_client_notify_get(struct nouveau_client *, int index);
diff --git a/drivers/gpu/drm/nouveau/core/include/core/device.h b/drivers/gpu/drm/nouveau/core/include/core/device.h
index 8743766454a5..2ec2e50d3676 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/device.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/device.h
@@ -16,6 +16,7 @@ enum nv_subdev_type {
 	 * to during POST.
 	 */
 	NVDEV_SUBDEV_DEVINIT,
+	NVDEV_SUBDEV_IBUS,
 	NVDEV_SUBDEV_GPIO,
 	NVDEV_SUBDEV_I2C,
 	NVDEV_SUBDEV_DEVINIT_LAST = NVDEV_SUBDEV_I2C,
@@ -24,13 +25,13 @@ enum nv_subdev_type {
 	 * been created, and are allowed to assume any subdevs in the
 	 * list above them exist and have been initialised.
 	 */
+	NVDEV_SUBDEV_FUSE,
 	NVDEV_SUBDEV_MXM,
 	NVDEV_SUBDEV_MC,
 	NVDEV_SUBDEV_BUS,
 	NVDEV_SUBDEV_TIMER,
 	NVDEV_SUBDEV_FB,
 	NVDEV_SUBDEV_LTC,
-	NVDEV_SUBDEV_IBUS,
 	NVDEV_SUBDEV_INSTMEM,
 	NVDEV_SUBDEV_VM,
 	NVDEV_SUBDEV_BAR,
@@ -91,6 +92,7 @@ struct nouveau_device {
 		GM100    = 0x110,
 	} card_type;
 	u32 chipset;
+	u8  chiprev;
 	u32 crystal;
 
 	struct nouveau_oclass *oclass[NVDEV_SUBDEV_NR];
@@ -157,6 +159,12 @@ nv_device_is_pci(struct nouveau_device *device)
 	return device->pdev != NULL;
 }
 
+static inline bool
+nv_device_is_cpu_coherent(struct nouveau_device *device)
+{
+	return (!IS_ENABLED(CONFIG_ARM) && nv_device_is_pci(device));
+}
+
 static inline struct device *
 nv_device_base(struct nouveau_device *device)
 {
diff --git a/drivers/gpu/drm/nouveau/core/include/core/event.h b/drivers/gpu/drm/nouveau/core/include/core/event.h
index 51e55d03330a..92876528972f 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/event.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/event.h
@@ -4,7 +4,8 @@
 #include <core/notify.h>
 
 struct nvkm_event_func {
-	int  (*ctor)(void *data, u32 size, struct nvkm_notify *);
+	int  (*ctor)(struct nouveau_object *, void *data, u32 size,
+		     struct nvkm_notify *);
 	void (*send)(void *data, u32 size, struct nvkm_notify *);
 	void (*init)(struct nvkm_event *, int type, int index);
 	void (*fini)(struct nvkm_event *, int type, int index);
diff --git a/drivers/gpu/drm/nouveau/core/include/core/handle.h b/drivers/gpu/drm/nouveau/core/include/core/handle.h
index ceb67d770875..d22a59138a9b 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/handle.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/handle.h
@@ -23,11 +23,6 @@ void nouveau_handle_destroy(struct nouveau_handle *);
 int  nouveau_handle_init(struct nouveau_handle *);
 int  nouveau_handle_fini(struct nouveau_handle *, bool suspend);
 
-int  nouveau_handle_new(struct nouveau_object *, u32 parent, u32 handle,
-			u16 oclass, void *data, u32 size,
-			struct nouveau_object **);
-int  nouveau_handle_del(struct nouveau_object *, u32 parent, u32 handle);
-
 struct nouveau_object *
 nouveau_handle_ref(struct nouveau_object *, u32 name);
 
diff --git a/drivers/gpu/drm/nouveau/core/include/core/mm.h b/drivers/gpu/drm/nouveau/core/include/core/mm.h
index 2bf7d0e32261..bfe6931544fe 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/mm.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/mm.h
@@ -6,6 +6,10 @@ struct nouveau_mm_node {
 	struct list_head fl_entry;
 	struct list_head rl_entry;
 
+#define NVKM_MM_HEAP_ANY 0x00
+	u8  heap;
+#define NVKM_MM_TYPE_NONE 0x00
+#define NVKM_MM_TYPE_HOLE 0xff
 	u8  type;
 	u32 offset;
 	u32 length;
@@ -27,10 +31,10 @@ nouveau_mm_initialised(struct nouveau_mm *mm)
 
 int  nouveau_mm_init(struct nouveau_mm *, u32 offset, u32 length, u32 block);
 int  nouveau_mm_fini(struct nouveau_mm *);
-int  nouveau_mm_head(struct nouveau_mm *, u8 type, u32 size_max, u32 size_min,
-		     u32 align, struct nouveau_mm_node **);
-int  nouveau_mm_tail(struct nouveau_mm *, u8 type, u32 size_max, u32 size_min,
-		     u32 align, struct nouveau_mm_node **);
+int  nouveau_mm_head(struct nouveau_mm *, u8 heap, u8 type, u32 size_max,
+		     u32 size_min, u32 align, struct nouveau_mm_node **);
+int  nouveau_mm_tail(struct nouveau_mm *, u8 heap, u8 type, u32 size_max,
+		     u32 size_min, u32 align, struct nouveau_mm_node **);
 void nouveau_mm_free(struct nouveau_mm *, struct nouveau_mm_node **);
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/core/notify.h b/drivers/gpu/drm/nouveau/core/include/core/notify.h
index 1262d8f020f3..a7c3c5f578cc 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/notify.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/notify.h
@@ -25,8 +25,9 @@ struct nvkm_notify {
 	const void *data;
 };
 
-int  nvkm_notify_init(struct nvkm_event *, int (*func)(struct nvkm_notify *),
-		      bool work, void *data, u32 size, u32 reply,
+int  nvkm_notify_init(struct nouveau_object *, struct nvkm_event *,
+		      int (*func)(struct nvkm_notify *), bool work,
+		      void *data, u32 size, u32 reply,
 		      struct nvkm_notify *);
 void nvkm_notify_fini(struct nvkm_notify *);
 void nvkm_notify_get(struct nvkm_notify *);
diff --git a/drivers/gpu/drm/nouveau/core/include/core/object.h b/drivers/gpu/drm/nouveau/core/include/core/object.h
index d7039482d6fd..2e2afa502c99 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/object.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/object.h
@@ -203,21 +203,4 @@ nv_memcmp(void *obj, u32 addr, const char *str, u32 len)
 	return 0;
 }
 
-#include <core/handle.h>
-
-static inline int
-nouveau_object_new(struct nouveau_object *client, u32 parent, u32 handle,
-		   u16 oclass, void *data, u32 size,
-		   struct nouveau_object **pobject)
-{
-	return nouveau_handle_new(client, parent, handle, oclass,
-				  data, size, pobject);
-}
-
-static inline int
-nouveau_object_del(struct nouveau_object *client, u32 parent, u32 handle)
-{
-	return nouveau_handle_del(client, parent, handle);
-}
-
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/engine/disp.h b/drivers/gpu/drm/nouveau/core/include/engine/disp.h
index 7a64f347b385..fc307f1317ff 100644
--- a/drivers/gpu/drm/nouveau/core/include/engine/disp.h
+++ b/drivers/gpu/drm/nouveau/core/include/engine/disp.h
@@ -31,5 +31,6 @@ extern struct nouveau_oclass *nvd0_disp_oclass;
 extern struct nouveau_oclass *nve0_disp_oclass;
 extern struct nouveau_oclass *nvf0_disp_oclass;
 extern struct nouveau_oclass *gm107_disp_oclass;
+extern struct nouveau_oclass *gm204_disp_oclass;
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/engine/fifo.h b/drivers/gpu/drm/nouveau/core/include/engine/fifo.h
index e5e4d930b2c2..2007453f6fce 100644
--- a/drivers/gpu/drm/nouveau/core/include/engine/fifo.h
+++ b/drivers/gpu/drm/nouveau/core/include/engine/fifo.h
@@ -116,7 +116,8 @@ extern struct nouveau_oclass *nve0_fifo_oclass;
 extern struct nouveau_oclass *gk20a_fifo_oclass;
 extern struct nouveau_oclass *nv108_fifo_oclass;
 
-int  nouveau_fifo_uevent_ctor(void *, u32, struct nvkm_notify *);
+int  nouveau_fifo_uevent_ctor(struct nouveau_object *, void *, u32,
+			      struct nvkm_notify *);
 void nouveau_fifo_uevent(struct nouveau_fifo *);
 
 void nv04_fifo_intr(struct nouveau_subdev *);
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bar.h b/drivers/gpu/drm/nouveau/core/include/subdev/bar.h
index be037fac534c..257ddf6d36d4 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bar.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bar.h
@@ -12,7 +12,6 @@ struct nouveau_bar {
 
 	int (*alloc)(struct nouveau_bar *, struct nouveau_object *,
 		     struct nouveau_mem *, struct nouveau_object **);
-	void __iomem *iomem;
 
 	int (*kmap)(struct nouveau_bar *, struct nouveau_mem *,
 		    u32 flags, struct nouveau_vma *);
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/M0203.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/M0203.h
new file mode 100644
index 000000000000..1f84d3612dd8
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/M0203.h
@@ -0,0 +1,31 @@
+#ifndef __NVBIOS_M0203_H__
+#define __NVBIOS_M0203_H__
+
+struct nvbios_M0203T {
+#define M0203T_TYPE_RAMCFG 0x00
+	u8  type;
+	u16 pointer;
+};
+
+u32 nvbios_M0203Te(struct nouveau_bios *, u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+u32 nvbios_M0203Tp(struct nouveau_bios *, u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		   struct nvbios_M0203T *);
+
+struct nvbios_M0203E {
+#define M0203E_TYPE_DDR2  0x0
+#define M0203E_TYPE_DDR3  0x1
+#define M0203E_TYPE_GDDR3 0x2
+#define M0203E_TYPE_GDDR5 0x3
+#define M0203E_TYPE_SKIP  0xf
+	u8 type;
+	u8 strap;
+	u8 group;
+};
+
+u32 nvbios_M0203Ee(struct nouveau_bios *, int idx, u8 *ver, u8 *hdr);
+u32 nvbios_M0203Ep(struct nouveau_bios *, int idx, u8 *ver, u8 *hdr,
+		   struct nvbios_M0203E *);
+u32 nvbios_M0203Em(struct nouveau_bios *, u8 ramcfg, u8 *ver, u8 *hdr,
+		   struct nvbios_M0203E *);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/M0205.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/M0205.h
new file mode 100644
index 000000000000..e171120cec81
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/M0205.h
@@ -0,0 +1,32 @@
+#ifndef __NVBIOS_M0205_H__
+#define __NVBIOS_M0205_H__
+
+struct nvbios_M0205T {
+	u16 freq;
+};
+
+u32 nvbios_M0205Te(struct nouveau_bios *,
+		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *snr, u8 *ssz);
+u32 nvbios_M0205Tp(struct nouveau_bios *,
+		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *snr, u8 *ssz,
+		   struct nvbios_M0205T *);
+
+struct nvbios_M0205E {
+	u8 type;
+};
+
+u32 nvbios_M0205Ee(struct nouveau_bios *, int idx,
+		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+u32 nvbios_M0205Ep(struct nouveau_bios *, int idx,
+		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		   struct nvbios_M0205E *);
+
+struct nvbios_M0205S {
+	u8 data;
+};
+
+u32 nvbios_M0205Se(struct nouveau_bios *, int ent, int idx, u8 *ver, u8 *hdr);
+u32 nvbios_M0205Sp(struct nouveau_bios *, int ent, int idx, u8 *ver, u8 *hdr,
+		   struct nvbios_M0205S *);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/M0209.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/M0209.h
new file mode 100644
index 000000000000..67dc50d837bc
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/M0209.h
@@ -0,0 +1,30 @@
+#ifndef __NVBIOS_M0209_H__
+#define __NVBIOS_M0209_H__
+
+u32 nvbios_M0209Te(struct nouveau_bios *,
+		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *snr, u8 *ssz);
+
+struct nvbios_M0209E {
+	u8 v00_40;
+	u8 bits;
+	u8 modulo;
+	u8 v02_40;
+	u8 v02_07;
+	u8 v03;
+};
+
+u32 nvbios_M0209Ee(struct nouveau_bios *, int idx,
+		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+u32 nvbios_M0209Ep(struct nouveau_bios *, int idx,
+		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		   struct nvbios_M0209E *);
+
+struct nvbios_M0209S {
+	u32 data[0x200];
+};
+
+u32 nvbios_M0209Se(struct nouveau_bios *, int ent, int idx, u8 *ver, u8 *hdr);
+u32 nvbios_M0209Sp(struct nouveau_bios *, int ent, int idx, u8 *ver, u8 *hdr,
+		   struct nvbios_M0209S *);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/fan.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/fan.h
new file mode 100644
index 000000000000..119d0874e041
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/fan.h
@@ -0,0 +1,8 @@
+#ifndef __NVBIOS_FAN_H__
+#define __NVBIOS_FAN_H__
+
+#include <subdev/bios/therm.h>
+
+u16 nvbios_fan_parse(struct nouveau_bios *bios, struct nvbios_therm_fan *fan);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/i2c.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/i2c.h
index 10b57a19a7de..c9bb112895af 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bios/i2c.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/i2c.h
@@ -4,11 +4,14 @@
 struct nouveau_bios;
 
 enum dcb_i2c_type {
-	DCB_I2C_NV04_BIT = 0,
-	DCB_I2C_NV4E_BIT = 4,
-	DCB_I2C_NVIO_BIT = 5,
-	DCB_I2C_NVIO_AUX = 6,
-	DCB_I2C_UNUSED = 0xff
+	/* matches bios type field prior to ccb 4.1 */
+	DCB_I2C_NV04_BIT = 0x00,
+	DCB_I2C_NV4E_BIT = 0x04,
+	DCB_I2C_NVIO_BIT = 0x05,
+	DCB_I2C_NVIO_AUX = 0x06,
+	/* made up - mostly */
+	DCB_I2C_PMGR     = 0x80,
+	DCB_I2C_UNUSED   = 0xff
 };
 
 struct dcb_i2c_entry {
@@ -16,6 +19,7 @@ struct dcb_i2c_entry {
 	u8 drive;
 	u8 sense;
 	u8 share;
+	u8 auxch;
 };
 
 u16 dcb_i2c_table(struct nouveau_bios *, u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/image.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/image.h
new file mode 100644
index 000000000000..3348b4580843
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/image.h
@@ -0,0 +1,13 @@
+#ifndef __NVBIOS_IMAGE_H__
+#define __NVBIOS_IMAGE_H__
+
+struct nvbios_image {
+	u32  base;
+	u32  size;
+	u8   type;
+	bool last;
+};
+
+bool nvbios_image(struct nouveau_bios *, int, struct nvbios_image *);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/npde.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/npde.h
new file mode 100644
index 000000000000..b18413d951e5
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/npde.h
@@ -0,0 +1,12 @@
+#ifndef __NVBIOS_NPDE_H__
+#define __NVBIOS_NPDE_H__
+
+struct nvbios_npdeT {
+	u32 image_size;
+	bool last;
+};
+
+u32 nvbios_npdeTe(struct nouveau_bios *, u32);
+u32 nvbios_npdeTp(struct nouveau_bios *, u32, struct nvbios_npdeT *);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/pcir.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/pcir.h
new file mode 100644
index 000000000000..3d634a06dca1
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/pcir.h
@@ -0,0 +1,18 @@
+#ifndef __NVBIOS_PCIR_H__
+#define __NVBIOS_PCIR_H__
+
+struct nvbios_pcirT {
+	u16 vendor_id;
+	u16 device_id;
+	u8  class_code[3];
+	u32 image_size;
+	u16 image_rev;
+	u8  image_type;
+	bool last;
+};
+
+u32 nvbios_pcirTe(struct nouveau_bios *, u32, u8 *ver, u16 *hdr);
+u32 nvbios_pcirTp(struct nouveau_bios *, u32, u8 *ver, u16 *hdr,
+		  struct nvbios_pcirT *);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/pmu.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/pmu.h
new file mode 100644
index 000000000000..9de593deaea8
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/pmu.h
@@ -0,0 +1,37 @@
+#ifndef __NVBIOS_PMU_H__
+#define __NVBIOS_PMU_H__
+
+struct nvbios_pmuT {
+};
+
+u32 nvbios_pmuTe(struct nouveau_bios *, u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+u32 nvbios_pmuTp(struct nouveau_bios *, u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		 struct nvbios_pmuT *);
+
+struct nvbios_pmuE {
+	u8  type;
+	u32 data;
+};
+
+u32 nvbios_pmuEe(struct nouveau_bios *, int idx, u8 *ver, u8 *hdr);
+u32 nvbios_pmuEp(struct nouveau_bios *, int idx, u8 *ver, u8 *hdr,
+		 struct nvbios_pmuE *);
+
+struct nvbios_pmuR {
+	u32 boot_addr_pmu;
+	u32 boot_addr;
+	u32 boot_size;
+	u32 code_addr_pmu;
+	u32 code_addr;
+	u32 code_size;
+	u32 init_addr_pmu;
+
+	u32 data_addr_pmu;
+	u32 data_addr;
+	u32 data_size;
+	u32 args_addr_pmu;
+};
+
+bool nvbios_pmuRm(struct nouveau_bios *, u8 type, struct nvbios_pmuR *);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/ramcfg.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/ramcfg.h
index c086ac6d677d..4a0e0ceb41ba 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bios/ramcfg.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/ramcfg.h
@@ -4,60 +4,139 @@
 struct nouveau_bios;
 
 struct nvbios_ramcfg {
-	unsigned rammap_11_08_01:1;
-	unsigned rammap_11_08_0c:2;
-	unsigned rammap_11_08_10:1;
-	unsigned rammap_11_11_0c:2;
+	unsigned rammap_ver;
+	unsigned rammap_hdr;
+	unsigned rammap_min;
+	unsigned rammap_max;
+	union {
+		struct {
+			unsigned rammap_10_04_02:1;
+			unsigned rammap_10_04_08:1;
+		};
+		struct {
+			unsigned rammap_11_08_01:1;
+			unsigned rammap_11_08_0c:2;
+			unsigned rammap_11_08_10:1;
+			unsigned rammap_11_09_01ff:9;
+			unsigned rammap_11_0a_03fe:9;
+			unsigned rammap_11_0a_0400:1;
+			unsigned rammap_11_0a_0800:1;
+			unsigned rammap_11_0b_01f0:5;
+			unsigned rammap_11_0b_0200:1;
+			unsigned rammap_11_0b_0400:1;
+			unsigned rammap_11_0b_0800:1;
+			unsigned rammap_11_0d:8;
+			unsigned rammap_11_0e:8;
+			unsigned rammap_11_0f:8;
+			unsigned rammap_11_11_0c:2;
+		};
+	};
 
-	unsigned ramcfg_11_01_01:1;
-	unsigned ramcfg_11_01_02:1;
-	unsigned ramcfg_11_01_04:1;
-	unsigned ramcfg_11_01_08:1;
-	unsigned ramcfg_11_01_10:1;
-	unsigned ramcfg_11_01_20:1;
-	unsigned ramcfg_11_01_40:1;
-	unsigned ramcfg_11_01_80:1;
-	unsigned ramcfg_11_02_03:2;
-	unsigned ramcfg_11_02_04:1;
-	unsigned ramcfg_11_02_08:1;
-	unsigned ramcfg_11_02_10:1;
-	unsigned ramcfg_11_02_40:1;
-	unsigned ramcfg_11_02_80:1;
-	unsigned ramcfg_11_03_0f:4;
-	unsigned ramcfg_11_03_30:2;
-	unsigned ramcfg_11_03_c0:2;
-	unsigned ramcfg_11_03_f0:4;
-	unsigned ramcfg_11_04:8;
-	unsigned ramcfg_11_06:8;
-	unsigned ramcfg_11_07_02:1;
-	unsigned ramcfg_11_07_04:1;
-	unsigned ramcfg_11_07_08:1;
-	unsigned ramcfg_11_07_10:1;
-	unsigned ramcfg_11_07_40:1;
-	unsigned ramcfg_11_07_80:1;
-	unsigned ramcfg_11_08_01:1;
-	unsigned ramcfg_11_08_02:1;
-	unsigned ramcfg_11_08_04:1;
-	unsigned ramcfg_11_08_08:1;
-	unsigned ramcfg_11_08_10:1;
-	unsigned ramcfg_11_08_20:1;
-	unsigned ramcfg_11_09:8;
+	unsigned ramcfg_ver;
+	unsigned ramcfg_hdr;
+	unsigned ramcfg_timing;
+	union {
+		struct {
+			unsigned ramcfg_10_02_01:1;
+			unsigned ramcfg_10_02_02:1;
+			unsigned ramcfg_10_02_04:1;
+			unsigned ramcfg_10_02_08:1;
+			unsigned ramcfg_10_02_10:1;
+			unsigned ramcfg_10_02_20:1;
+			unsigned ramcfg_10_DLLoff:1;
+			unsigned ramcfg_10_03_0f:4;
+			unsigned ramcfg_10_04_01:1;
+			unsigned ramcfg_10_05:8;
+			unsigned ramcfg_10_06:8;
+			unsigned ramcfg_10_07:8;
+			unsigned ramcfg_10_08:8;
+			unsigned ramcfg_10_09_0f:4;
+			unsigned ramcfg_10_09_f0:4;
+		};
+		struct {
+			unsigned ramcfg_11_01_01:1;
+			unsigned ramcfg_11_01_02:1;
+			unsigned ramcfg_11_01_04:1;
+			unsigned ramcfg_11_01_08:1;
+			unsigned ramcfg_11_01_10:1;
+			unsigned ramcfg_11_01_20:1;
+			unsigned ramcfg_11_01_40:1;
+			unsigned ramcfg_11_01_80:1;
+			unsigned ramcfg_11_02_03:2;
+			unsigned ramcfg_11_02_04:1;
+			unsigned ramcfg_11_02_08:1;
+			unsigned ramcfg_11_02_10:1;
+			unsigned ramcfg_11_02_40:1;
+			unsigned ramcfg_11_02_80:1;
+			unsigned ramcfg_11_03_0f:4;
+			unsigned ramcfg_11_03_30:2;
+			unsigned ramcfg_11_03_c0:2;
+			unsigned ramcfg_11_03_f0:4;
+			unsigned ramcfg_11_04:8;
+			unsigned ramcfg_11_06:8;
+			unsigned ramcfg_11_07_02:1;
+			unsigned ramcfg_11_07_04:1;
+			unsigned ramcfg_11_07_08:1;
+			unsigned ramcfg_11_07_10:1;
+			unsigned ramcfg_11_07_40:1;
+			unsigned ramcfg_11_07_80:1;
+			unsigned ramcfg_11_08_01:1;
+			unsigned ramcfg_11_08_02:1;
+			unsigned ramcfg_11_08_04:1;
+			unsigned ramcfg_11_08_08:1;
+			unsigned ramcfg_11_08_10:1;
+			unsigned ramcfg_11_08_20:1;
+			unsigned ramcfg_11_09:8;
+		};
+	};
 
+	unsigned timing_ver;
+	unsigned timing_hdr;
 	unsigned timing[11];
-	unsigned timing_20_2e_03:2;
-	unsigned timing_20_2e_30:2;
-	unsigned timing_20_2e_c0:2;
-	unsigned timing_20_2f_03:2;
-	unsigned timing_20_2c_003f:6;
-	unsigned timing_20_2c_1fc0:7;
-	unsigned timing_20_30_f8:5;
-	unsigned timing_20_30_07:3;
-	unsigned timing_20_31_0007:3;
-	unsigned timing_20_31_0078:4;
-	unsigned timing_20_31_0780:4;
-	unsigned timing_20_31_0800:1;
-	unsigned timing_20_31_7000:3;
-	unsigned timing_20_31_8000:1;
+	union {
+		struct {
+			unsigned timing_10_WR:8;
+			unsigned timing_10_WTR:8;
+			unsigned timing_10_CL:8;
+			unsigned timing_10_RC:8;
+			/*empty: 4 */
+			unsigned timing_10_RFC:8;        /* Byte 5 */
+			/*empty: 6 */
+			unsigned timing_10_RAS:8;        /* Byte 7 */
+			/*empty: 8 */
+			unsigned timing_10_RP:8;         /* Byte 9 */
+			unsigned timing_10_RCDRD:8;
+			unsigned timing_10_RCDWR:8;
+			unsigned timing_10_RRD:8;
+			unsigned timing_10_13:8;
+			unsigned timing_10_ODT:3;
+			/* empty: 15 */
+			unsigned timing_10_16:8;
+			/* empty: 17 */
+			unsigned timing_10_18:8;
+			unsigned timing_10_CWL:8;
+			unsigned timing_10_20:8;
+			unsigned timing_10_21:8;
+			/* empty: 22, 23 */
+			unsigned timing_10_24:8;
+		};
+		struct {
+			unsigned timing_20_2e_03:2;
+			unsigned timing_20_2e_30:2;
+			unsigned timing_20_2e_c0:2;
+			unsigned timing_20_2f_03:2;
+			unsigned timing_20_2c_003f:6;
+			unsigned timing_20_2c_1fc0:7;
+			unsigned timing_20_30_f8:5;
+			unsigned timing_20_30_07:3;
+			unsigned timing_20_31_0007:3;
+			unsigned timing_20_31_0078:4;
+			unsigned timing_20_31_0780:4;
+			unsigned timing_20_31_0800:1;
+			unsigned timing_20_31_7000:3;
+			unsigned timing_20_31_8000:1;
+		};
+	};
 };
 
 u8 nvbios_ramcfg_count(struct nouveau_bios *);
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/rammap.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/rammap.h
index 5bdf8e4db40a..47e021d3e20d 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bios/rammap.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/rammap.h
@@ -8,9 +8,10 @@ u32 nvbios_rammapTe(struct nouveau_bios *, u8 *ver, u8 *hdr,
 
 u32 nvbios_rammapEe(struct nouveau_bios *, int idx,
 		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+u32 nvbios_rammapEp(struct nouveau_bios *, int idx,
+		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		    struct nvbios_ramcfg *);
 u32 nvbios_rammapEm(struct nouveau_bios *, u16 mhz,
-		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
-u32 nvbios_rammapEp(struct nouveau_bios *, u16 mhz,
 		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
 		    struct nvbios_ramcfg *);
 
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/therm.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/therm.h
index 8dc5051df55d..295d093f3b30 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bios/therm.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/therm.h
@@ -23,6 +23,12 @@ struct nvbios_therm_sensor {
 	struct nvbios_therm_threshold thrs_shutdown;
 };
 
+enum nvbios_therm_fan_type {
+	NVBIOS_THERM_FAN_UNK = 0,
+	NVBIOS_THERM_FAN_TOGGLE = 1,
+	NVBIOS_THERM_FAN_PWM = 2,
+};
+
 /* no vbios have more than 6 */
 #define NOUVEAU_TEMP_FAN_TRIP_MAX 10
 struct nouveau_therm_trip_point {
@@ -38,7 +44,9 @@ enum nvbios_therm_fan_mode {
 };
 
 struct nvbios_therm_fan {
-	u16 pwm_freq;
+	enum nvbios_therm_fan_type type;
+
+	u32 pwm_freq;
 
 	u8 min_duty;
 	u8 max_duty;
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/clock.h b/drivers/gpu/drm/nouveau/core/include/subdev/clock.h
index a5ca00dd2f61..36ed035d4d42 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/clock.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/clock.h
@@ -29,6 +29,7 @@ enum nv_clk_src {
 	nv_clk_src_mdiv,
 
 	nv_clk_src_core,
+	nv_clk_src_core_intm,
 	nv_clk_src_shader,
 
 	nv_clk_src_mem,
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/devinit.h b/drivers/gpu/drm/nouveau/core/include/subdev/devinit.h
index e292271a84e4..e007a9d44683 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/devinit.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/devinit.h
@@ -30,5 +30,6 @@ extern struct nouveau_oclass *nva3_devinit_oclass;
 extern struct nouveau_oclass *nvaf_devinit_oclass;
 extern struct nouveau_oclass *nvc0_devinit_oclass;
 extern struct nouveau_oclass *gm107_devinit_oclass;
+extern struct nouveau_oclass *gm204_devinit_oclass;
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/fb.h b/drivers/gpu/drm/nouveau/core/include/subdev/fb.h
index 871e73914b24..8d0032f15205 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/fb.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/fb.h
@@ -111,6 +111,7 @@ extern struct nouveau_oclass *gm107_fb_oclass;
 #include <subdev/bios/ramcfg.h>
 
 struct nouveau_ram_data {
+	struct list_head head;
 	struct nvbios_ramcfg bios;
 	u32 freq;
 };
@@ -136,6 +137,7 @@ struct nouveau_ram {
 
 	int ranks;
 	int parts;
+	int part_mask;
 
 	int  (*get)(struct nouveau_fb *, u64 size, u32 align,
 		    u32 size_nc, u32 type, struct nouveau_mem **);
@@ -144,11 +146,6 @@ struct nouveau_ram {
 	int  (*calc)(struct nouveau_fb *, u32 freq);
 	int  (*prog)(struct nouveau_fb *);
 	void (*tidy)(struct nouveau_fb *);
-	struct {
-		u8  version;
-		u32 data;
-		u8  size;
-	} rammap, ramcfg, timing;
 	u32 freq;
 	u32 mr[16];
 	u32 mr1_nuts;
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/fb/regsnv04.h b/drivers/gpu/drm/nouveau/core/include/subdev/fb/regsnv04.h
new file mode 100644
index 000000000000..0f7fc0c52ab2
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/fb/regsnv04.h
@@ -0,0 +1,21 @@
+#ifndef __NOUVEAU_FB_REGS_04_H__
+#define __NOUVEAU_FB_REGS_04_H__
+
+#define NV04_PFB_BOOT_0						0x00100000
+#	define NV04_PFB_BOOT_0_RAM_AMOUNT			0x00000003
+#	define NV04_PFB_BOOT_0_RAM_AMOUNT_32MB			0x00000000
+#	define NV04_PFB_BOOT_0_RAM_AMOUNT_4MB			0x00000001
+#	define NV04_PFB_BOOT_0_RAM_AMOUNT_8MB			0x00000002
+#	define NV04_PFB_BOOT_0_RAM_AMOUNT_16MB			0x00000003
+#	define NV04_PFB_BOOT_0_RAM_WIDTH_128			0x00000004
+#	define NV04_PFB_BOOT_0_RAM_TYPE				0x00000028
+#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_8MBIT		0x00000000
+#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_16MBIT		0x00000008
+#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_16MBIT_4BANK	0x00000010
+#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_16MBIT		0x00000018
+#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_64MBIT		0x00000020
+#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_64MBITX16		0x00000028
+#	define NV04_PFB_BOOT_0_UMA_ENABLE			0x00000100
+#	define NV04_PFB_BOOT_0_UMA_SIZE				0x0000f000
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/fuse.h b/drivers/gpu/drm/nouveau/core/include/subdev/fuse.h
new file mode 100644
index 000000000000..2b1ddb2a9a7d
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/fuse.h
@@ -0,0 +1,30 @@
+#ifndef __NOUVEAU_FUSE_H__
+#define __NOUVEAU_FUSE_H__
+
+#include <core/subdev.h>
+#include <core/device.h>
+
+struct nouveau_fuse {
+	struct nouveau_subdev base;
+};
+
+static inline struct nouveau_fuse *
+nouveau_fuse(void *obj)
+{
+	return (void *)nv_device(obj)->subdev[NVDEV_SUBDEV_FUSE];
+}
+
+#define nouveau_fuse_create(p, e, o, d)                                        \
+	nouveau_fuse_create_((p), (e), (o), sizeof(**d), (void **)d)
+
+int  nouveau_fuse_create_(struct nouveau_object *, struct nouveau_object *,
+			  struct nouveau_oclass *, int, void **);
+void _nouveau_fuse_dtor(struct nouveau_object *);
+int  _nouveau_fuse_init(struct nouveau_object *);
+#define _nouveau_fuse_fini _nouveau_subdev_fini
+
+extern struct nouveau_oclass g80_fuse_oclass;
+extern struct nouveau_oclass gf100_fuse_oclass;
+extern struct nouveau_oclass gm107_fuse_oclass;
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/gpio.h b/drivers/gpu/drm/nouveau/core/include/subdev/gpio.h
index b73733d21cc7..f855140dbcb7 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/gpio.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/gpio.h
@@ -40,7 +40,7 @@ nouveau_gpio(void *obj)
 
 extern struct nouveau_oclass *nv10_gpio_oclass;
 extern struct nouveau_oclass *nv50_gpio_oclass;
-extern struct nouveau_oclass *nv92_gpio_oclass;
+extern struct nouveau_oclass *nv94_gpio_oclass;
 extern struct nouveau_oclass *nvd0_gpio_oclass;
 extern struct nouveau_oclass *nve0_gpio_oclass;
 
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/i2c.h b/drivers/gpu/drm/nouveau/core/include/subdev/i2c.h
index 1b937c2c25ae..d94ccacb40bf 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/i2c.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/i2c.h
@@ -8,6 +8,8 @@
 #include <subdev/bios/i2c.h>
 
 #define NV_I2C_PORT(n)    (0x00 + (n))
+#define NV_I2C_AUX(n)     (0x10 + (n))
+#define NV_I2C_EXT(n)     (0x20 + (n))
 #define NV_I2C_DEFAULT(n) (0x80 + (n))
 
 #define NV_I2C_TYPE_DCBI2C(n) (0x0000 | (n))
@@ -89,6 +91,7 @@ extern struct nouveau_oclass *nv94_i2c_oclass;
 extern struct nouveau_oclass *nvd0_i2c_oclass;
 extern struct nouveau_oclass *gf117_i2c_oclass;
 extern struct nouveau_oclass *nve0_i2c_oclass;
+extern struct nouveau_oclass *gm204_i2c_oclass;
 
 static inline int
 nv_rdi2cr(struct nouveau_i2c_port *port, u8 addr, u8 reg)
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/pwr.h b/drivers/gpu/drm/nouveau/core/include/subdev/pwr.h
index f73feec151db..f2427bf5aeed 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/pwr.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/pwr.h
@@ -47,5 +47,10 @@ void nouveau_memx_wr32(struct nouveau_memx *, u32 addr, u32 data);
 void nouveau_memx_wait(struct nouveau_memx *,
 		       u32 addr, u32 mask, u32 data, u32 nsec);
 void nouveau_memx_nsec(struct nouveau_memx *, u32 nsec);
+void nouveau_memx_wait_vblank(struct nouveau_memx *);
+void nouveau_memx_train(struct nouveau_memx *);
+int  nouveau_memx_train_result(struct nouveau_pwr *, u32 *, int);
+void nouveau_memx_block(struct nouveau_memx *);
+void nouveau_memx_unblock(struct nouveau_memx *);
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/therm.h b/drivers/gpu/drm/nouveau/core/include/subdev/therm.h
index d4a68179e586..a437597dcafc 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/therm.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/therm.h
@@ -78,5 +78,6 @@ extern struct nouveau_oclass nv50_therm_oclass;
 extern struct nouveau_oclass nv84_therm_oclass;
 extern struct nouveau_oclass nva3_therm_oclass;
 extern struct nouveau_oclass nvd0_therm_oclass;
+extern struct nouveau_oclass gm107_therm_oclass;
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/volt.h b/drivers/gpu/drm/nouveau/core/include/subdev/volt.h
index 820b62ffd75b..67db5e58880d 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/volt.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/volt.h
@@ -52,6 +52,7 @@ int  _nouveau_volt_init(struct nouveau_object *);
 #define _nouveau_volt_fini _nouveau_subdev_fini
 
 extern struct nouveau_oclass nv40_volt_oclass;
+extern struct nouveau_oclass gk20a_volt_oclass;
 
 int nouveau_voltgpio_init(struct nouveau_volt *);
 int nouveau_voltgpio_get(struct nouveau_volt *);
diff --git a/drivers/gpu/drm/nouveau/core/os.h b/drivers/gpu/drm/nouveau/core/os.h
index ccfa21d72ddc..bdd05ee7ec72 100644
--- a/drivers/gpu/drm/nouveau/core/os.h
+++ b/drivers/gpu/drm/nouveau/core/os.h
@@ -23,6 +23,7 @@
 #include <linux/pm_runtime.h>
 #include <linux/power_supply.h>
 #include <linux/clk.h>
+#include <linux/regulator/consumer.h>
 
 #include <asm/unaligned.h>
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bar/base.c b/drivers/gpu/drm/nouveau/core/subdev/bar/base.c
index 8bcbdf39cfb2..b1adc69efd88 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bar/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bar/base.c
@@ -38,10 +38,12 @@ struct nouveau_barobj {
 static int
 nouveau_barobj_ctor(struct nouveau_object *parent,
 		    struct nouveau_object *engine,
-		    struct nouveau_oclass *oclass, void *mem, u32 size,
+		    struct nouveau_oclass *oclass, void *data, u32 size,
 		    struct nouveau_object **pobject)
 {
+	struct nouveau_device *device = nv_device(parent);
 	struct nouveau_bar *bar = (void *)engine;
+	struct nouveau_mem *mem = data;
 	struct nouveau_barobj *barobj;
 	int ret;
 
@@ -54,7 +56,13 @@ nouveau_barobj_ctor(struct nouveau_object *parent,
 	if (ret)
 		return ret;
 
-	barobj->iomem = bar->iomem + (u32)barobj->vma.offset;
+	barobj->iomem = ioremap(nv_device_resource_start(device, 3) +
+				(u32)barobj->vma.offset, mem->size << 12);
+	if (!barobj->iomem) {
+		nv_warn(bar, "PRAMIN ioremap failed\n");
+		return -ENOMEM;
+	}
+
 	return 0;
 }
 
@@ -63,8 +71,11 @@ nouveau_barobj_dtor(struct nouveau_object *object)
 {
 	struct nouveau_bar *bar = (void *)object->engine;
 	struct nouveau_barobj *barobj = (void *)object;
-	if (barobj->vma.node)
+	if (barobj->vma.node) {
+		if (barobj->iomem)
+			iounmap(barobj->iomem);
 		bar->unmap(bar, &barobj->vma);
+	}
 	nouveau_object_destroy(&barobj->base);
 }
 
@@ -99,12 +110,11 @@ nouveau_bar_alloc(struct nouveau_bar *bar, struct nouveau_object *parent,
 		  struct nouveau_mem *mem, struct nouveau_object **pobject)
 {
 	struct nouveau_object *engine = nv_object(bar);
-	int ret = -ENOMEM;
-	if (bar->iomem) {
-		ret = nouveau_object_ctor(parent, engine,
-					  &nouveau_barobj_oclass,
-					  mem, 0, pobject);
-	}
+	struct nouveau_object *gpuobj;
+	int ret = nouveau_object_ctor(parent, engine, &nouveau_barobj_oclass,
+				      mem, 0, &gpuobj);
+	if (ret == 0)
+		*pobject = gpuobj;
 	return ret;
 }
 
@@ -113,7 +123,6 @@ nouveau_bar_create_(struct nouveau_object *parent,
 		    struct nouveau_object *engine,
 		    struct nouveau_oclass *oclass, int length, void **pobject)
 {
-	struct nouveau_device *device = nv_device(parent);
 	struct nouveau_bar *bar;
 	int ret;
 
@@ -123,21 +132,12 @@ nouveau_bar_create_(struct nouveau_object *parent,
 	if (ret)
 		return ret;
 
-	if (nv_device_resource_len(device, 3) != 0) {
-		bar->iomem = ioremap(nv_device_resource_start(device, 3),
-				     nv_device_resource_len(device, 3));
-		if (!bar->iomem)
-			nv_warn(bar, "PRAMIN ioremap failed\n");
-	}
-
 	return 0;
 }
 
 void
 nouveau_bar_destroy(struct nouveau_bar *bar)
 {
-	if (bar->iomem)
-		iounmap(bar->iomem);
 	nouveau_subdev_destroy(&bar->base);
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/M0203.c b/drivers/gpu/drm/nouveau/core/subdev/bios/M0203.c
new file mode 100644
index 000000000000..28906b16d4e5
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/M0203.c
@@ -0,0 +1,129 @@
+/*
+ * Copyright 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include <subdev/bios.h>
+#include <subdev/bios/bit.h>
+#include <subdev/bios/M0203.h>
+
+u32
+nvbios_M0203Te(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
+{
+	struct bit_entry bit_M;
+	u32 data = 0x00000000;
+
+	if (!bit_entry(bios, 'M', &bit_M)) {
+		if (bit_M.version == 2 && bit_M.length > 0x04)
+			data = nv_ro16(bios, bit_M.offset + 0x03);
+		if (data) {
+			*ver = nv_ro08(bios, data + 0x00);
+			switch (*ver) {
+			case 0x10:
+				*hdr = nv_ro08(bios, data + 0x01);
+				*len = nv_ro08(bios, data + 0x02);
+				*cnt = nv_ro08(bios, data + 0x03);
+				return data;
+			default:
+				break;
+			}
+		}
+	}
+
+	return 0x00000000;
+}
+
+u32
+nvbios_M0203Tp(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+	       struct nvbios_M0203T *info)
+{
+	u32 data = nvbios_M0203Te(bios, ver, hdr, cnt, len);
+	memset(info, 0x00, sizeof(*info));
+	switch (!!data * *ver) {
+	case 0x10:
+		info->type    = nv_ro08(bios, data + 0x04);
+		info->pointer = nv_ro16(bios, data + 0x05);
+		break;
+	default:
+		break;
+	}
+	return data;
+}
+
+u32
+nvbios_M0203Ee(struct nouveau_bios *bios, int idx, u8 *ver, u8 *hdr)
+{
+	u8  cnt, len;
+	u32 data = nvbios_M0203Te(bios, ver, hdr, &cnt, &len);
+	if (data && idx < cnt) {
+		data = data + *hdr + idx * len;
+		*hdr = len;
+		return data;
+	}
+	return 0x00000000;
+}
+
+u32
+nvbios_M0203Ep(struct nouveau_bios *bios, int idx, u8 *ver, u8 *hdr,
+	       struct nvbios_M0203E *info)
+{
+	u32 data = nvbios_M0203Ee(bios, idx, ver, hdr);
+	memset(info, 0x00, sizeof(*info));
+	switch (!!data * *ver) {
+	case 0x10:
+		info->type  = (nv_ro08(bios, data + 0x00) & 0x0f) >> 0;
+		info->strap = (nv_ro08(bios, data + 0x00) & 0xf0) >> 4;
+		info->group = (nv_ro08(bios, data + 0x01) & 0x0f) >> 0;
+		return data;
+	default:
+		break;
+	}
+	return 0x00000000;
+}
+
+u32
+nvbios_M0203Em(struct nouveau_bios *bios, u8 ramcfg, u8 *ver, u8 *hdr,
+	       struct nvbios_M0203E *info)
+{
+	struct nvbios_M0203T M0203T;
+	u8  cnt, len, idx = 0xff;
+	u32 data;
+
+	if (!nvbios_M0203Tp(bios, ver, hdr, &cnt, &len, &M0203T)) {
+		nv_warn(bios, "M0203T not found\n");
+		return 0x00000000;
+	}
+
+	while ((data = nvbios_M0203Ep(bios, ++idx, ver, hdr, info))) {
+		switch (M0203T.type) {
+		case M0203T_TYPE_RAMCFG:
+			if (info->strap != ramcfg)
+				continue;
+			return data;
+		default:
+			nv_warn(bios, "M0203T type %02x\n", M0203T.type);
+			return 0x00000000;
+		}
+	}
+
+	return data;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/M0205.c b/drivers/gpu/drm/nouveau/core/subdev/bios/M0205.c
new file mode 100644
index 000000000000..ac9617c5fc2a
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/M0205.c
@@ -0,0 +1,136 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include <subdev/bios.h>
+#include <subdev/bios/bit.h>
+#include <subdev/bios/M0205.h>
+
+u32
+nvbios_M0205Te(struct nouveau_bios *bios,
+	       u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *snr, u8 *ssz)
+{
+	struct bit_entry bit_M;
+	u32 data = 0x00000000;
+
+	if (!bit_entry(bios, 'M', &bit_M)) {
+		if (bit_M.version == 2 && bit_M.length > 0x08)
+			data = nv_ro32(bios, bit_M.offset + 0x05);
+		if (data) {
+			*ver = nv_ro08(bios, data + 0x00);
+			switch (*ver) {
+			case 0x10:
+				*hdr = nv_ro08(bios, data + 0x01);
+				*len = nv_ro08(bios, data + 0x02);
+				*ssz = nv_ro08(bios, data + 0x03);
+				*snr = nv_ro08(bios, data + 0x04);
+				*cnt = nv_ro08(bios, data + 0x05);
+				return data;
+			default:
+				break;
+			}
+		}
+	}
+
+	return 0x00000000;
+}
+
+u32
+nvbios_M0205Tp(struct nouveau_bios *bios,
+	       u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *snr, u8 *ssz,
+	       struct nvbios_M0205T *info)
+{
+	u32 data = nvbios_M0205Te(bios, ver, hdr, cnt, len, snr, ssz);
+	memset(info, 0x00, sizeof(*info));
+	switch (!!data * *ver) {
+	case 0x10:
+		info->freq = nv_ro16(bios, data + 0x06);
+		break;
+	default:
+		break;
+	}
+	return data;
+}
+
+u32
+nvbios_M0205Ee(struct nouveau_bios *bios, int idx,
+	       u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
+{
+	u8  snr, ssz;
+	u32 data = nvbios_M0205Te(bios, ver, hdr, cnt, len, &snr, &ssz);
+	if (data && idx < *cnt) {
+		data = data + *hdr + idx * (*len + (snr * ssz));
+		*hdr = *len;
+		*cnt = snr;
+		*len = ssz;
+		return data;
+	}
+	return 0x00000000;
+}
+
+u32
+nvbios_M0205Ep(struct nouveau_bios *bios, int idx,
+	       u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+	       struct nvbios_M0205E *info)
+{
+	u32 data = nvbios_M0205Ee(bios, idx, ver, hdr, cnt, len);
+	memset(info, 0x00, sizeof(*info));
+	switch (!!data * *ver) {
+	case 0x10:
+		info->type = nv_ro08(bios, data + 0x00) & 0x0f;
+		return data;
+	default:
+		break;
+	}
+	return 0x00000000;
+}
+
+u32
+nvbios_M0205Se(struct nouveau_bios *bios, int ent, int idx, u8 *ver, u8 *hdr)
+{
+
+	u8  cnt, len;
+	u32 data = nvbios_M0205Ee(bios, ent, ver, hdr, &cnt, &len);
+	if (data && idx < cnt) {
+		data = data + *hdr + idx * len;
+		*hdr = len;
+		return data;
+	}
+	return 0x00000000;
+}
+
+u32
+nvbios_M0205Sp(struct nouveau_bios *bios, int ent, int idx, u8 *ver, u8 *hdr,
+	       struct nvbios_M0205S *info)
+{
+	u32 data = nvbios_M0205Se(bios, ent, idx, ver, hdr);
+	memset(info, 0x00, sizeof(*info));
+	switch (!!data * *ver) {
+	case 0x10:
+		info->data = nv_ro08(bios, data + 0x00);
+		return data;
+	default:
+		break;
+	}
+	return 0x00000000;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/M0209.c b/drivers/gpu/drm/nouveau/core/subdev/bios/M0209.c
new file mode 100644
index 000000000000..b142a510e89f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/M0209.c
@@ -0,0 +1,137 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include <subdev/bios.h>
+#include <subdev/bios/bit.h>
+#include <subdev/bios/M0209.h>
+
+u32
+nvbios_M0209Te(struct nouveau_bios *bios,
+	       u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *snr, u8 *ssz)
+{
+	struct bit_entry bit_M;
+	u32 data = 0x00000000;
+
+	if (!bit_entry(bios, 'M', &bit_M)) {
+		if (bit_M.version == 2 && bit_M.length > 0x0c)
+			data = nv_ro32(bios, bit_M.offset + 0x09);
+		if (data) {
+			*ver = nv_ro08(bios, data + 0x00);
+			switch (*ver) {
+			case 0x10:
+				*hdr = nv_ro08(bios, data + 0x01);
+				*len = nv_ro08(bios, data + 0x02);
+				*ssz = nv_ro08(bios, data + 0x03);
+				*snr = 1;
+				*cnt = nv_ro08(bios, data + 0x04);
+				return data;
+			default:
+				break;
+			}
+		}
+	}
+
+	return 0x00000000;
+}
+
+u32
+nvbios_M0209Ee(struct nouveau_bios *bios, int idx,
+	       u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
+{
+	u8  snr, ssz;
+	u32 data = nvbios_M0209Te(bios, ver, hdr, cnt, len, &snr, &ssz);
+	if (data && idx < *cnt) {
+		data = data + *hdr + idx * (*len + (snr * ssz));
+		*hdr = *len;
+		*cnt = snr;
+		*len = ssz;
+		return data;
+	}
+	return 0x00000000;
+}
+
+u32
+nvbios_M0209Ep(struct nouveau_bios *bios, int idx,
+	       u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+	       struct nvbios_M0209E *info)
+{
+	u32 data = nvbios_M0209Ee(bios, idx, ver, hdr, cnt, len);
+	memset(info, 0x00, sizeof(*info));
+	switch (!!data * *ver) {
+	case 0x10:
+		info->v00_40 = (nv_ro08(bios, data + 0x00) & 0x40) >> 6;
+		info->bits   =  nv_ro08(bios, data + 0x00) & 0x3f;
+		info->modulo =  nv_ro08(bios, data + 0x01);
+		info->v02_40 = (nv_ro08(bios, data + 0x02) & 0x40) >> 6;
+		info->v02_07 =  nv_ro08(bios, data + 0x02) & 0x07;
+		info->v03    =  nv_ro08(bios, data + 0x03);
+		return data;
+	default:
+		break;
+	}
+	return 0x00000000;
+}
+
+u32
+nvbios_M0209Se(struct nouveau_bios *bios, int ent, int idx, u8 *ver, u8 *hdr)
+{
+
+	u8  cnt, len;
+	u32 data = nvbios_M0209Ee(bios, ent, ver, hdr, &cnt, &len);
+	if (data && idx < cnt) {
+		data = data + *hdr + idx * len;
+		*hdr = len;
+		return data;
+	}
+	return 0x00000000;
+}
+
+u32
+nvbios_M0209Sp(struct nouveau_bios *bios, int ent, int idx, u8 *ver, u8 *hdr,
+	       struct nvbios_M0209S *info)
+{
+	struct nvbios_M0209E M0209E;
+	u8  cnt, len;
+	u32 data = nvbios_M0209Ep(bios, ent, ver, hdr, &cnt, &len, &M0209E);
+	if (data) {
+		u32 i, data = nvbios_M0209Se(bios, ent, idx, ver, hdr);
+		memset(info, 0x00, sizeof(*info));
+		switch (!!data * *ver) {
+		case 0x10:
+			for (i = 0; i < ARRAY_SIZE(info->data); i++) {
+				u32 bits = (i % M0209E.modulo) * M0209E.bits;
+				u32 mask = (1ULL << M0209E.bits) - 1;
+				u16  off = bits / 8;
+				u8   mod = bits % 8;
+				info->data[i] = nv_ro32(bios, data + off);
+				info->data[i] = info->data[i] >> mod;
+				info->data[i] = info->data[i] & mask;
+			}
+			return data;
+		default:
+			break;
+		}
+	}
+	return 0x00000000;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/base.c b/drivers/gpu/drm/nouveau/core/subdev/bios/base.c
index d45704a2c2df..7df3a273553d 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/base.c
@@ -31,6 +31,8 @@
 #include <subdev/bios/bmp.h>
 #include <subdev/bios/bit.h>
 
+#include "priv.h"
+
 u8
 nvbios_checksum(const u8 *data, int size)
 {
@@ -56,362 +58,21 @@ nvbios_findstr(const u8 *data, int size, const char *str, int len)
 	return 0;
 }
 
-#if defined(__powerpc__)
-static void
-nouveau_bios_shadow_of(struct nouveau_bios *bios)
+int
+nvbios_extend(struct nouveau_bios *bios, u32 length)
 {
-	struct pci_dev *pdev = nv_device(bios)->pdev;
-	struct device_node *dn;
-	const u32 *data;
-	int size;
-
-	dn = pci_device_to_OF_node(pdev);
-	if (!dn) {
-		nv_info(bios, "Unable to get the OF node\n");
-		return;
-	}
-
-	data = of_get_property(dn, "NVDA,BMP", &size);
-	if (data && size) {
-		bios->size = size;
-		bios->data = kmalloc(bios->size, GFP_KERNEL);
-		if (bios->data)
-			memcpy(bios->data, data, size);
-	}
-}
-#endif
-
-static void
-nouveau_bios_shadow_pramin(struct nouveau_bios *bios)
-{
-	struct nouveau_device *device = nv_device(bios);
-	u64 addr = 0;
-	u32 bar0 = 0;
-	int i;
-
-	if (device->card_type >= NV_50) {
-		if (device->card_type >= NV_C0 && device->card_type < GM100) {
-			if (nv_rd32(bios, 0x022500) & 0x00000001)
-				return;
-		} else
-		if (device->card_type >= GM100) {
-			if (nv_rd32(bios, 0x021c04) & 0x00000001)
-				return;
-		}
-
-		addr = nv_rd32(bios, 0x619f04);
-		if (!(addr & 0x00000008)) {
-			nv_debug(bios, "... not enabled\n");
-			return;
+	if (bios->size < length) {
+		u8 *prev = bios->data;
+		if (!(bios->data = kmalloc(length, GFP_KERNEL))) {
+			bios->data = prev;
+			return -ENOMEM;
 		}
-		if ( (addr & 0x00000003) != 1) {
-			nv_debug(bios, "... not in vram\n");
-			return;
-		}
-
-		addr = (addr & 0xffffff00) << 8;
-		if (!addr) {
-			addr  = (u64)nv_rd32(bios, 0x001700) << 16;
-			addr += 0xf0000;
-		}
-
-		bar0 = nv_mask(bios, 0x001700, 0xffffffff, addr >> 16);
-	}
-
-	/* bail if no rom signature */
-	if (nv_rd08(bios, 0x700000) != 0x55 ||
-	    nv_rd08(bios, 0x700001) != 0xaa)
-		goto out;
-
-	bios->size = nv_rd08(bios, 0x700002) * 512;
-	if (!bios->size)
-		goto out;
-
-	bios->data = kmalloc(bios->size, GFP_KERNEL);
-	if (bios->data) {
-		for (i = 0; i < bios->size; i++)
-			nv_wo08(bios, i, nv_rd08(bios, 0x700000 + i));
-	}
-
-out:
-	if (device->card_type >= NV_50)
-		nv_wr32(bios, 0x001700, bar0);
-}
-
-static void
-nouveau_bios_shadow_prom(struct nouveau_bios *bios)
-{
-	struct nouveau_device *device = nv_device(bios);
-	u32 pcireg, access;
-	u16 pcir;
-	int i;
-
-	/* there is no prom on nv4x IGP's */
-	if (device->card_type == NV_40 && device->chipset >= 0x4c)
-		return;
-
-	/* enable access to rom */
-	if (device->card_type >= NV_50)
-		pcireg = 0x088050;
-	else
-		pcireg = 0x001850;
-	access = nv_mask(bios, pcireg, 0x00000001, 0x00000000);
-
-	/* WARNING: PROM accesses should always be 32-bits aligned. Other
-	 * accesses work on most chipset but do not on Kepler chipsets
-	 */
-
-	/* bail if no rom signature, with a workaround for a PROM reading
-	 * issue on some chipsets.  the first read after a period of
-	 * inactivity returns the wrong result, so retry the first header
-	 * byte a few times before giving up as a workaround
-	 */
-	i = 16;
-	do {
-		u32 data = le32_to_cpu(nv_rd32(bios, 0x300000)) & 0xffff;
-		if (data == 0xaa55)
-			break;
-	} while (i--);
-
-	if (!i)
-		goto out;
-
-	/* read entire bios image to system memory */
-	bios->size = (le32_to_cpu(nv_rd32(bios, 0x300000)) >> 16) & 0xff;
-	bios->size = bios->size * 512;
-	if (!bios->size)
-		goto out;
-
-	bios->data = kmalloc(bios->size, GFP_KERNEL);
-	if (!bios->data)
-		goto out;
-
-	for (i = 0; i < bios->size; i += 4)
-		((u32 *)bios->data)[i/4] = nv_rd32(bios, 0x300000 + i);
-
-	/* check the PCI record header */
-	pcir = nv_ro16(bios, 0x0018);
-	if (bios->data[pcir + 0] != 'P' ||
-	    bios->data[pcir + 1] != 'C' ||
-	    bios->data[pcir + 2] != 'I' ||
-	    bios->data[pcir + 3] != 'R') {
-		bios->size = 0;
-		kfree(bios->data);
-	}
-
-out:
-	/* disable access to rom */
-	nv_wr32(bios, pcireg, access);
-}
-
-#if defined(CONFIG_ACPI) && defined(CONFIG_X86)
-int nouveau_acpi_get_bios_chunk(uint8_t *bios, int offset, int len);
-bool nouveau_acpi_rom_supported(struct pci_dev *pdev);
-#else
-static inline bool
-nouveau_acpi_rom_supported(struct pci_dev *pdev) {
-	return false;
-}
-
-static inline int
-nouveau_acpi_get_bios_chunk(uint8_t *bios, int offset, int len) {
-	return -EINVAL;
-}
-#endif
-
-static void
-nouveau_bios_shadow_acpi(struct nouveau_bios *bios)
-{
-	struct pci_dev *pdev = nv_device(bios)->pdev;
-	int ret, cnt, i;
-
-	if (!nouveau_acpi_rom_supported(pdev)) {
-		bios->data = NULL;
-		return;
-	}
-
-	bios->size = 0;
-	bios->data = kmalloc(4096, GFP_KERNEL);
-	if (bios->data) {
-		if (nouveau_acpi_get_bios_chunk(bios->data, 0, 4096) == 4096)
-			bios->size = bios->data[2] * 512;
-		kfree(bios->data);
+		memcpy(bios->data, prev, bios->size);
+		bios->size = length;
+		kfree(prev);
+		return 1;
 	}
-
-	if (!bios->size)
-		return;
-
-	bios->data = kmalloc(bios->size, GFP_KERNEL);
-	if (bios->data) {
-		/* disobey the acpi spec - much faster on at least w530 ... */
-		ret = nouveau_acpi_get_bios_chunk(bios->data, 0, bios->size);
-		if (ret != bios->size ||
-		    nvbios_checksum(bios->data, bios->size)) {
-			/* ... that didn't work, ok, i'll be good now */
-			for (i = 0; i < bios->size; i += cnt) {
-				cnt = min((bios->size - i), (u32)4096);
-				ret = nouveau_acpi_get_bios_chunk(bios->data, i, cnt);
-				if (ret != cnt)
-					break;
-			}
-		}
-	}
-}
-
-static void
-nouveau_bios_shadow_pci(struct nouveau_bios *bios)
-{
-	struct pci_dev *pdev = nv_device(bios)->pdev;
-	size_t size;
-
-	if (!pci_enable_rom(pdev)) {
-		void __iomem *rom = pci_map_rom(pdev, &size);
-		if (rom && size) {
-			bios->data = kmalloc(size, GFP_KERNEL);
-			if (bios->data) {
-				memcpy_fromio(bios->data, rom, size);
-				bios->size = size;
-			}
-		}
-		if (rom)
-			pci_unmap_rom(pdev, rom);
-
-		pci_disable_rom(pdev);
-	}
-}
-
-static void
-nouveau_bios_shadow_platform(struct nouveau_bios *bios)
-{
-	struct pci_dev *pdev = nv_device(bios)->pdev;
-	size_t size;
-
-	void __iomem *rom = pci_platform_rom(pdev, &size);
-	if (rom && size) {
-		bios->data = kmalloc(size, GFP_KERNEL);
-		if (bios->data) {
-			memcpy_fromio(bios->data, rom, size);
-			bios->size = size;
-		}
-	}
-}
-
-static int
-nouveau_bios_score(struct nouveau_bios *bios, const bool writeable)
-{
-	if (bios->size < 3 || !bios->data || bios->data[0] != 0x55 ||
-			bios->data[1] != 0xAA) {
-		nv_info(bios, "... signature not found\n");
-		return 0;
-	}
-
-	if (nvbios_checksum(bios->data,
-			min_t(u32, bios->data[2] * 512, bios->size))) {
-		nv_info(bios, "... checksum invalid\n");
-		/* if a ro image is somewhat bad, it's probably all rubbish */
-		return writeable ? 2 : 1;
-	}
-
-	nv_info(bios, "... appears to be valid\n");
-	return 3;
-}
-
-struct methods {
-	const char desc[16];
-	void (*shadow)(struct nouveau_bios *);
-	const bool rw;
-	int score;
-	u32 size;
-	u8 *data;
-};
-
-static int
-nouveau_bios_shadow(struct nouveau_bios *bios)
-{
-	struct methods shadow_methods[] = {
-#if defined(__powerpc__)
-		{ "OpenFirmware", nouveau_bios_shadow_of, true, 0, 0, NULL },
-#endif
-		{ "PRAMIN", nouveau_bios_shadow_pramin, true, 0, 0, NULL },
-		{ "PROM", nouveau_bios_shadow_prom, false, 0, 0, NULL },
-		{ "ACPI", nouveau_bios_shadow_acpi, true, 0, 0, NULL },
-		{ "PCIROM", nouveau_bios_shadow_pci, true, 0, 0, NULL },
-		{ "PLATFORM", nouveau_bios_shadow_platform, true, 0, 0, NULL },
-		{}
-	};
-	struct methods *mthd, *best;
-	const struct firmware *fw;
-	const char *optarg;
-	int optlen, ret;
-	char *source;
-
-	optarg = nouveau_stropt(nv_device(bios)->cfgopt, "NvBios", &optlen);
-	source = optarg ? kstrndup(optarg, optlen, GFP_KERNEL) : NULL;
-	if (source) {
-		/* try to match one of the built-in methods */
-		mthd = shadow_methods;
-		do {
-			if (strcasecmp(source, mthd->desc))
-				continue;
-			nv_info(bios, "source: %s\n", mthd->desc);
-
-			mthd->shadow(bios);
-			mthd->score = nouveau_bios_score(bios, mthd->rw);
-			if (mthd->score) {
-				kfree(source);
-				return 0;
-			}
-		} while ((++mthd)->shadow);
-
-		/* attempt to load firmware image */
-		ret = request_firmware(&fw, source, &nv_device(bios)->pdev->dev);
-		if (ret == 0) {
-			bios->size = fw->size;
-			bios->data = kmemdup(fw->data, fw->size, GFP_KERNEL);
-			release_firmware(fw);
-
-			nv_info(bios, "image: %s\n", source);
-			if (nouveau_bios_score(bios, 1)) {
-				kfree(source);
-				return 0;
-			}
-
-			kfree(bios->data);
-			bios->data = NULL;
-		}
-
-		nv_error(bios, "source \'%s\' invalid\n", source);
-		kfree(source);
-	}
-
-	mthd = shadow_methods;
-	do {
-		nv_info(bios, "checking %s for image...\n", mthd->desc);
-		mthd->shadow(bios);
-		mthd->score = nouveau_bios_score(bios, mthd->rw);
-		mthd->size = bios->size;
-		mthd->data = bios->data;
-		bios->data = NULL;
-	} while (mthd->score != 3 && (++mthd)->shadow);
-
-	mthd = shadow_methods;
-	best = mthd;
-	do {
-		if (mthd->score > best->score) {
-			kfree(best->data);
-			best = mthd;
-		}
-	} while ((++mthd)->shadow);
-
-	if (best->score) {
-		nv_info(bios, "using image from %s\n", best->desc);
-		bios->size = best->size;
-		bios->data = best->data;
-		return 0;
-	}
-
-	nv_error(bios, "unable to locate usable image\n");
-	return -EINVAL;
+	return 0;
 }
 
 static u8
@@ -472,7 +133,7 @@ nouveau_bios_ctor(struct nouveau_object *parent,
 	if (ret)
 		return ret;
 
-	ret = nouveau_bios_shadow(bios);
+	ret = nvbios_shadow(bios);
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/dcb.c b/drivers/gpu/drm/nouveau/core/subdev/bios/dcb.c
index 88606bfaf847..96099aff8b41 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/dcb.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/dcb.c
@@ -42,7 +42,7 @@ dcb_table(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
 
 	*ver = nv_ro08(bios, dcb);
 
-	if (*ver >= 0x41) {
+	if (*ver >= 0x42) {
 		nv_warn(bios, "DCB version 0x%02x unknown\n", *ver);
 		return 0x0000;
 	} else
@@ -124,6 +124,7 @@ dcb_outp_parse(struct nouveau_bios *bios, u8 idx, u8 *ver, u8 *len,
 	       struct dcb_output *outp)
 {
 	u16 dcb = dcb_outp(bios, idx, ver, len);
+	memset(outp, 0x00, sizeof(*outp));
 	if (dcb) {
 		if (*ver >= 0x20) {
 			u32 conn = nv_ro32(bios, dcb + 0x00);
@@ -156,17 +157,20 @@ dcb_outp_parse(struct nouveau_bios *bios, u8 idx, u8 *ver, u8 *len,
 					break;
 				}
 
-				switch (conf & 0x0f000000) {
-				case 0x0f000000:
-					outp->dpconf.link_nr = 4;
-					break;
-				case 0x03000000:
-					outp->dpconf.link_nr = 2;
-					break;
-				case 0x01000000:
-				default:
-					outp->dpconf.link_nr = 1;
-					break;
+				outp->dpconf.link_nr = (conf & 0x0f000000) >> 24;
+				if (*ver < 0x41) {
+					switch (outp->dpconf.link_nr) {
+					case 0x0f:
+						outp->dpconf.link_nr = 4;
+						break;
+					case 0x03:
+						outp->dpconf.link_nr = 2;
+						break;
+					case 0x01:
+					default:
+						outp->dpconf.link_nr = 1;
+						break;
+					}
 				}
 
 				/* fall-through... */
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/disp.c b/drivers/gpu/drm/nouveau/core/subdev/bios/disp.c
index 7f16e52d9bea..51f355599694 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/disp.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/disp.c
@@ -40,6 +40,7 @@ nvbios_disp_table(struct nouveau_bios *bios,
 				switch (*ver) {
 				case 0x20:
 				case 0x21:
+				case 0x22:
 					*hdr = nv_ro08(bios, data + 0x01);
 					*len = nv_ro08(bios, data + 0x02);
 					*cnt = nv_ro08(bios, data + 0x03);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/dp.c b/drivers/gpu/drm/nouveau/core/subdev/bios/dp.c
index f309dd657250..cef53f81f12b 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/dp.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/dp.c
@@ -41,6 +41,7 @@ nvbios_dp_table(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
 				case 0x21:
 				case 0x30:
 				case 0x40:
+				case 0x41:
 					*hdr = nv_ro08(bios, data + 0x01);
 					*len = nv_ro08(bios, data + 0x02);
 					*cnt = nv_ro08(bios, data + 0x03);
@@ -70,6 +71,7 @@ nvbios_dpout_entry(struct nouveau_bios *bios, u8 idx,
 			*cnt = nv_ro08(bios, outp + 0x04);
 			break;
 		case 0x40:
+		case 0x41:
 			*hdr = nv_ro08(bios, data + 0x04);
 			*cnt = 0;
 			*len = 0;
@@ -108,6 +110,7 @@ nvbios_dpout_parse(struct nouveau_bios *bios, u8 idx,
 				info->script[4] = nv_ro16(bios, data + 0x10);
 			break;
 		case 0x40:
+		case 0x41:
 			info->flags     = nv_ro08(bios, data + 0x04);
 			info->script[0] = nv_ro16(bios, data + 0x05);
 			info->script[1] = nv_ro16(bios, data + 0x07);
@@ -172,10 +175,11 @@ nvbios_dpcfg_parse(struct nouveau_bios *bios, u16 outp, u8 idx,
 			break;
 		case 0x30:
 		case 0x40:
+		case 0x41:
 			info->pc    = nv_ro08(bios, data + 0x00);
 			info->dc    = nv_ro08(bios, data + 0x01);
 			info->pe    = nv_ro08(bios, data + 0x02);
-			info->tx_pu = nv_ro08(bios, data + 0x03);
+			info->tx_pu = nv_ro08(bios, data + 0x03) & 0x0f;
 			break;
 		default:
 			data = 0x0000;
@@ -194,6 +198,10 @@ nvbios_dpcfg_match(struct nouveau_bios *bios, u16 outp, u8 pc, u8 vs, u8 pe,
 	u16 data;
 
 	if (*ver >= 0x30) {
+		/*XXX: there's a second set of these on at least 4.1, that
+		 *     i've witnessed nvidia using instead of the first
+		 *     on gm204.  figure out what/why
+		 */
 		const u8 vsoff[] = { 0, 4, 7, 9 };
 		idx = (pc * 10) + vsoff[vs] + pe;
 	} else {
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/extdev.c b/drivers/gpu/drm/nouveau/core/subdev/bios/extdev.c
index b2a676e53580..49285d4f7ca5 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/extdev.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/extdev.c
@@ -90,7 +90,7 @@ nvbios_extdev_find(struct nouveau_bios *bios, enum nvbios_extdev_type type,
 	u16 entry;
 
 	i = 0;
-	while (!(entry = nvbios_extdev_entry(bios, i++, &ver, &len))) {
+	while ((entry = nvbios_extdev_entry(bios, i++, &ver, &len))) {
 		extdev_parse_entry(bios, entry, func);
 		if (func->type == type)
 			return 0;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/fan.c b/drivers/gpu/drm/nouveau/core/subdev/bios/fan.c
new file mode 100644
index 000000000000..e419892240f5
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/fan.c
@@ -0,0 +1,93 @@
+/*
+ * Copyright 2014 Martin Peres
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Martin Peres
+ */
+
+#include <subdev/bios.h>
+#include <subdev/bios/bit.h>
+#include <subdev/bios/fan.h>
+
+u16
+nvbios_fan_table(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
+{
+	struct bit_entry bit_P;
+	u16 fan = 0x0000;
+
+	if (!bit_entry(bios, 'P', &bit_P)) {
+		if (bit_P.version == 2 && bit_P.length >= 0x5a)
+			fan = nv_ro16(bios, bit_P.offset + 0x58);
+
+		if (fan) {
+			*ver = nv_ro08(bios, fan + 0);
+			switch (*ver) {
+			case 0x10:
+				*hdr = nv_ro08(bios, fan + 1);
+				*len = nv_ro08(bios, fan + 2);
+				*cnt = nv_ro08(bios, fan + 3);
+				return fan;
+			default:
+				break;
+			}
+		}
+	}
+
+	return 0x0000;
+}
+
+u16
+nvbios_fan_entry(struct nouveau_bios *bios, int idx, u8 *ver, u8 *hdr,
+		 u8 *cnt, u8 *len)
+{
+	u16 data = nvbios_fan_table(bios, ver, hdr, cnt, len);
+	if (data && idx < *cnt)
+		return data + *hdr + (idx * (*len));
+	return 0x0000;
+}
+
+u16
+nvbios_fan_parse(struct nouveau_bios *bios, struct nvbios_therm_fan *fan)
+{
+	u8 ver, hdr, cnt, len;
+
+	u16 data = nvbios_fan_entry(bios, 0, &ver, &hdr, &cnt, &len);
+	if (data) {
+		u8 type = nv_ro08(bios, data + 0x00);
+		switch (type) {
+		case 0:
+			fan->type = NVBIOS_THERM_FAN_TOGGLE;
+			break;
+		case 1:
+		case 2:
+			/* TODO: Understand the difference between the two! */
+			fan->type = NVBIOS_THERM_FAN_PWM;
+			break;
+		default:
+			fan->type = NVBIOS_THERM_FAN_UNK;
+		}
+
+		fan->min_duty = nv_ro08(bios, data + 0x02);
+		fan->max_duty = nv_ro08(bios, data + 0x03);
+
+		fan->pwm_freq = nv_ro32(bios, data + 0x0b) & 0xffffff;
+	}
+	return data;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/i2c.c b/drivers/gpu/drm/nouveau/core/subdev/bios/i2c.c
index cfb9288c6d28..282320ba9264 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/i2c.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/i2c.c
@@ -39,6 +39,11 @@ dcb_i2c_table(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
 			i2c = nv_ro16(bios, dcb + 4);
 	}
 
+	if (i2c && *ver >= 0x42) {
+		nv_warn(bios, "ccb %02x not supported\n", *ver);
+		return 0x0000;
+	}
+
 	if (i2c && *ver >= 0x30) {
 		*ver = nv_ro08(bios, i2c + 0);
 		*hdr = nv_ro08(bios, i2c + 1);
@@ -70,14 +75,25 @@ dcb_i2c_parse(struct nouveau_bios *bios, u8 idx, struct dcb_i2c_entry *info)
 	u8  ver, len;
 	u16 ent = dcb_i2c_entry(bios, idx, &ver, &len);
 	if (ent) {
-		info->type  = nv_ro08(bios, ent + 3);
-		info->share = DCB_I2C_UNUSED;
-		if (ver < 0x30) {
-			info->type &= 0x07;
+		if (ver >= 0x41) {
+			if (!(nv_ro32(bios, ent) & 0x80000000))
+				info->type = DCB_I2C_UNUSED;
+			else
+				info->type = DCB_I2C_PMGR;
+		} else
+		if (ver >= 0x30) {
+			info->type = nv_ro08(bios, ent + 0x03);
+		} else {
+			info->type = nv_ro08(bios, ent + 0x03) & 0x07;
 			if (info->type == 0x07)
 				info->type = DCB_I2C_UNUSED;
 		}
 
+		info->drive = DCB_I2C_UNUSED;
+		info->sense = DCB_I2C_UNUSED;
+		info->share = DCB_I2C_UNUSED;
+		info->auxch = DCB_I2C_UNUSED;
+
 		switch (info->type) {
 		case DCB_I2C_NV04_BIT:
 			info->drive = nv_ro08(bios, ent + 0);
@@ -87,12 +103,23 @@ dcb_i2c_parse(struct nouveau_bios *bios, u8 idx, struct dcb_i2c_entry *info)
 			info->drive = nv_ro08(bios, ent + 1);
 			return 0;
 		case DCB_I2C_NVIO_BIT:
-		case DCB_I2C_NVIO_AUX:
 			info->drive = nv_ro08(bios, ent + 0) & 0x0f;
-			if (nv_ro08(bios, ent + 1) & 0x01) {
-				info->share  = nv_ro08(bios, ent + 1) >> 1;
-				info->share &= 0x0f;
-			}
+			if (nv_ro08(bios, ent + 1) & 0x01)
+				info->share = nv_ro08(bios, ent + 1) >> 1;
+			return 0;
+		case DCB_I2C_NVIO_AUX:
+			info->auxch = nv_ro08(bios, ent + 0) & 0x0f;
+			if (nv_ro08(bios, ent + 1) & 0x01)
+					info->share = info->auxch;
+			return 0;
+		case DCB_I2C_PMGR:
+			info->drive = (nv_ro16(bios, ent + 0) & 0x01f) >> 0;
+			if (info->drive == 0x1f)
+				info->drive = DCB_I2C_UNUSED;
+			info->auxch = (nv_ro16(bios, ent + 0) & 0x3e0) >> 5;
+			if (info->auxch == 0x1f)
+				info->auxch = DCB_I2C_UNUSED;
+			info->share = info->auxch;
 			return 0;
 		case DCB_I2C_UNUSED:
 			return 0;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/image.c b/drivers/gpu/drm/nouveau/core/subdev/bios/image.c
new file mode 100644
index 000000000000..373f9a564ac9
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/image.c
@@ -0,0 +1,78 @@
+/*
+ * Copyright 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs <bskeggs@redhat.com>
+ */
+
+#include <subdev/bios.h>
+#include <subdev/bios/image.h>
+#include <subdev/bios/pcir.h>
+#include <subdev/bios/npde.h>
+
+static bool
+nvbios_imagen(struct nouveau_bios *bios, struct nvbios_image *image)
+{
+	struct nvbios_pcirT pcir;
+	struct nvbios_npdeT npde;
+	u8  ver;
+	u16 hdr;
+	u32 data;
+
+	switch ((data = nv_ro16(bios, image->base + 0x00))) {
+	case 0xaa55:
+	case 0xbb77:
+	case 0x4e56: /* NV */
+		break;
+	default:
+		nv_debug(bios, "%08x: ROM signature (%04x) unknown\n",
+			 image->base, data);
+		return false;
+	}
+
+	if (!(data = nvbios_pcirTp(bios, image->base, &ver, &hdr, &pcir)))
+		return false;
+	image->size = pcir.image_size;
+	image->type = pcir.image_type;
+	image->last = pcir.last;
+
+	if (image->type != 0x70) {
+		if (!(data = nvbios_npdeTp(bios, image->base, &npde)))
+			return true;
+		image->size = npde.image_size;
+		image->last = npde.last;
+	} else {
+		image->last = true;
+	}
+
+	return true;
+}
+
+bool
+nvbios_image(struct nouveau_bios *bios, int idx, struct nvbios_image *image)
+{
+	memset(image, 0x00, sizeof(*image));
+	do {
+		image->base += image->size;
+		if (image->last || !nvbios_imagen(bios, image))
+			return false;
+	} while(idx--);
+	return true;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/init.c b/drivers/gpu/drm/nouveau/core/subdev/bios/init.c
index 626380f9e4c0..c6579ef32cd1 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/init.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/init.c
@@ -255,6 +255,8 @@ init_i2c(struct nvbios_init *init, int index)
 		}
 
 		index = init->outp->i2c_index;
+		if (init->outp->type == DCB_OUTPUT_DP)
+			index += NV_I2C_AUX(0);
 	}
 
 	return i2c->find(i2c, index);
@@ -278,7 +280,7 @@ init_wri2cr(struct nvbios_init *init, u8 index, u8 addr, u8 reg, u8 val)
 	return -ENODEV;
 }
 
-static int
+static u8
 init_rdauxr(struct nvbios_init *init, u32 addr)
 {
 	struct nouveau_i2c_port *port = init_i2c(init, -2);
@@ -286,20 +288,24 @@ init_rdauxr(struct nvbios_init *init, u32 addr)
 
 	if (port && init_exec(init)) {
 		int ret = nv_rdaux(port, addr, &data, 1);
-		if (ret)
-			return ret;
-		return data;
+		if (ret == 0)
+			return data;
+		trace("auxch read failed with %d\n", ret);
 	}
 
-	return -ENODEV;
+	return 0x00;
 }
 
 static int
 init_wrauxr(struct nvbios_init *init, u32 addr, u8 data)
 {
 	struct nouveau_i2c_port *port = init_i2c(init, -2);
-	if (port && init_exec(init))
-		return nv_wraux(port, addr, &data, 1);
+	if (port && init_exec(init)) {
+		int ret = nv_wraux(port, addr, &data, 1);
+		if (ret)
+			trace("auxch write failed with %d\n", ret);
+		return ret;
+	}
 	return -ENODEV;
 }
 
@@ -838,6 +844,40 @@ init_io_or(struct nvbios_init *init)
 }
 
 /**
+ * INIT_ANDN_REG - opcode 0x47
+ *
+ */
+static void
+init_andn_reg(struct nvbios_init *init)
+{
+	struct nouveau_bios *bios = init->bios;
+	u32  reg = nv_ro32(bios, init->offset + 1);
+	u32 mask = nv_ro32(bios, init->offset + 5);
+
+	trace("ANDN_REG\tR[0x%06x] &= ~0x%08x\n", reg, mask);
+	init->offset += 9;
+
+	init_mask(init, reg, mask, 0);
+}
+
+/**
+ * INIT_OR_REG - opcode 0x48
+ *
+ */
+static void
+init_or_reg(struct nvbios_init *init)
+{
+	struct nouveau_bios *bios = init->bios;
+	u32  reg = nv_ro32(bios, init->offset + 1);
+	u32 mask = nv_ro32(bios, init->offset + 5);
+
+	trace("OR_REG\tR[0x%06x] |= 0x%08x\n", reg, mask);
+	init->offset += 9;
+
+	init_mask(init, reg, 0, mask);
+}
+
+/**
  * INIT_INDEX_ADDRESS_LATCHED - opcode 0x49
  *
  */
@@ -2068,6 +2108,8 @@ static struct nvbios_init_opcode {
 	[0x3a] = { init_dp_condition },
 	[0x3b] = { init_io_mask_or },
 	[0x3c] = { init_io_or },
+	[0x47] = { init_andn_reg },
+	[0x48] = { init_or_reg },
 	[0x49] = { init_idx_addr_latched },
 	[0x4a] = { init_io_restrict_pll2 },
 	[0x4b] = { init_pll2 },
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/npde.c b/drivers/gpu/drm/nouveau/core/subdev/bios/npde.c
new file mode 100644
index 000000000000..d694716a166c
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/npde.c
@@ -0,0 +1,59 @@
+/*
+ * Copyright 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs <bskeggs@redhat.com>
+ */
+
+#include <subdev/bios.h>
+#include <subdev/bios/npde.h>
+#include <subdev/bios/pcir.h>
+
+u32
+nvbios_npdeTe(struct nouveau_bios *bios, u32 base)
+{
+	struct nvbios_pcirT pcir;
+	u8  ver; u16 hdr;
+	u32 data = nvbios_pcirTp(bios, base, &ver, &hdr, &pcir);
+	if (data = (data + hdr + 0x0f) & ~0x0f, data) {
+		switch (nv_ro32(bios, data + 0x00)) {
+		case 0x4544504e: /* NPDE */
+			break;
+		default:
+			nv_debug(bios, "%08x: NPDE signature (%08x) unknown\n",
+				 data, nv_ro32(bios, data + 0x00));
+			data = 0;
+			break;
+		}
+	}
+	return data;
+}
+
+u32
+nvbios_npdeTp(struct nouveau_bios *bios, u32 base, struct nvbios_npdeT *info)
+{
+	u32 data = nvbios_npdeTe(bios, base);
+	memset(info, 0x00, sizeof(*info));
+	if (data) {
+		info->image_size = nv_ro16(bios, data + 0x08) * 512;
+		info->last = nv_ro08(bios, data + 0x0a) & 0x80;
+	}
+	return data;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/pcir.c b/drivers/gpu/drm/nouveau/core/subdev/bios/pcir.c
new file mode 100644
index 000000000000..91dae26bc50f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/pcir.c
@@ -0,0 +1,69 @@
+/*
+ * Copyright 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs <bskeggs@redhat.com>
+ */
+
+#include <subdev/bios.h>
+#include <subdev/bios/pcir.h>
+
+u32
+nvbios_pcirTe(struct nouveau_bios *bios, u32 base, u8 *ver, u16 *hdr)
+{
+	u32 data = nv_ro16(bios, base + 0x18);
+	if (data) {
+		data += base;
+		switch (nv_ro32(bios, data + 0x00)) {
+		case 0x52494350: /* PCIR */
+		case 0x53494752: /* RGIS */
+		case 0x5344504e: /* NPDS */
+			*hdr = nv_ro16(bios, data + 0x0a);
+			*ver = nv_ro08(bios, data + 0x0c);
+			break;
+		default:
+			nv_debug(bios, "%08x: PCIR signature (%08x) unknown\n",
+				 data, nv_ro32(bios, data + 0x00));
+			data = 0;
+			break;
+		}
+	}
+	return data;
+}
+
+u32
+nvbios_pcirTp(struct nouveau_bios *bios, u32 base, u8 *ver, u16 *hdr,
+	      struct nvbios_pcirT *info)
+{
+	u32 data = nvbios_pcirTe(bios, base, ver, hdr);
+	memset(info, 0x00, sizeof(*info));
+	if (data) {
+		info->vendor_id = nv_ro16(bios, data + 0x04);
+		info->device_id = nv_ro16(bios, data + 0x06);
+		info->class_code[0] = nv_ro08(bios, data + 0x0d);
+		info->class_code[1] = nv_ro08(bios, data + 0x0e);
+		info->class_code[2] = nv_ro08(bios, data + 0x0f);
+		info->image_size = nv_ro16(bios, data + 0x10) * 512;
+		info->image_rev = nv_ro16(bios, data + 0x12);
+		info->image_type = nv_ro08(bios, data + 0x14);
+		info->last = nv_ro08(bios, data + 0x15) & 0x80;
+	}
+	return data;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/pmu.c b/drivers/gpu/drm/nouveau/core/subdev/bios/pmu.c
new file mode 100644
index 000000000000..66c56ba07d1b
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/pmu.c
@@ -0,0 +1,135 @@
+/*
+ * Copyright 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs <bskeggs@redhat.com>
+ */
+
+#include <subdev/bios.h>
+#include <subdev/bios/bit.h>
+#include <subdev/bios/image.h>
+#include <subdev/bios/pmu.h>
+
+static u32
+weirdo_pointer(struct nouveau_bios *bios, u32 data)
+{
+	struct nvbios_image image;
+	int idx = 0;
+	if (nvbios_image(bios, idx++, &image)) {
+		data -= image.size;
+		while (nvbios_image(bios, idx++, &image)) {
+			if (image.type == 0xe0)
+				return image.base + data;
+		}
+	}
+	return 0;
+}
+
+u32
+nvbios_pmuTe(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
+{
+	struct bit_entry bit_p;
+	u32 data = 0;
+
+	if (!bit_entry(bios, 'p', &bit_p)) {
+		if (bit_p.version == 2 && bit_p.length >= 4)
+			data = nv_ro32(bios, bit_p.offset + 0x00);
+		if ((data = weirdo_pointer(bios, data))) {
+			*ver = nv_ro08(bios, data + 0x00); /* maybe? */
+			*hdr = nv_ro08(bios, data + 0x01);
+			*len = nv_ro08(bios, data + 0x02);
+			*cnt = nv_ro08(bios, data + 0x03);
+		}
+	}
+
+	return data;
+}
+
+u32
+nvbios_pmuTp(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+	     struct nvbios_pmuT *info)
+{
+	u32 data = nvbios_pmuTe(bios, ver, hdr, cnt, len);
+	memset(info, 0x00, sizeof(*info));
+	switch (!!data * *ver) {
+	default:
+		break;
+	}
+	return data;
+}
+
+u32
+nvbios_pmuEe(struct nouveau_bios *bios, int idx, u8 *ver, u8 *hdr)
+{
+	u8  cnt, len;
+	u32 data = nvbios_pmuTe(bios, ver, hdr, &cnt, &len);
+	if (data && idx < cnt) {
+		data = data + *hdr + (idx * len);
+		*hdr = len;
+		return data;
+	}
+	return 0;
+}
+
+u32
+nvbios_pmuEp(struct nouveau_bios *bios, int idx, u8 *ver, u8 *hdr,
+	     struct nvbios_pmuE *info)
+{
+	u32 data = nvbios_pmuEe(bios, idx, ver, hdr);
+	memset(info, 0x00, sizeof(*info));
+	switch (!!data * *ver) {
+	default:
+		info->type = nv_ro08(bios, data + 0x00);
+		info->data = nv_ro32(bios, data + 0x02);
+		break;
+	}
+	return data;
+}
+
+bool
+nvbios_pmuRm(struct nouveau_bios *bios, u8 type, struct nvbios_pmuR *info)
+{
+	struct nvbios_pmuE pmuE;
+	u8  ver, hdr, idx = 0;
+	u32 data;
+	memset(info, 0x00, sizeof(*info));
+	while ((data = nvbios_pmuEp(bios, idx++, &ver, &hdr, &pmuE))) {
+		if ( pmuE.type == type &&
+		    (data = weirdo_pointer(bios, pmuE.data))) {
+			info->init_addr_pmu = nv_ro32(bios, data + 0x08);
+			info->args_addr_pmu = nv_ro32(bios, data + 0x0c);
+			info->boot_addr     = data + 0x30;
+			info->boot_addr_pmu = nv_ro32(bios, data + 0x10) +
+					      nv_ro32(bios, data + 0x18);
+			info->boot_size     = nv_ro32(bios, data + 0x1c) -
+					      nv_ro32(bios, data + 0x18);
+			info->code_addr     = info->boot_addr + info->boot_size;
+			info->code_addr_pmu = info->boot_addr_pmu +
+					      info->boot_size;
+			info->code_size     = nv_ro32(bios, data + 0x20);
+			info->data_addr     = data + 0x30 +
+					      nv_ro32(bios, data + 0x24);
+			info->data_addr_pmu = nv_ro32(bios, data + 0x28);
+			info->data_size     = nv_ro32(bios, data + 0x2c);
+			return true;
+		}
+	}
+	return false;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/priv.h b/drivers/gpu/drm/nouveau/core/subdev/bios/priv.h
new file mode 100644
index 000000000000..187d225bd1e9
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/priv.h
@@ -0,0 +1,25 @@
+#ifndef __NVKM_BIOS_PRIV_H__
+#define __NVKM_BIOS_PRIV_H__
+
+#include <subdev/bios.h>
+
+struct nvbios_source {
+	const char *name;
+	void *(*init)(struct nouveau_bios *, const char *);
+	void  (*fini)(void *);
+	u32   (*read)(void *, u32 offset, u32 length, struct nouveau_bios *);
+	bool rw;
+};
+
+int nvbios_extend(struct nouveau_bios *, u32 length);
+int nvbios_shadow(struct nouveau_bios *);
+
+extern const struct nvbios_source nvbios_rom;
+extern const struct nvbios_source nvbios_ramin;
+extern const struct nvbios_source nvbios_acpi_fast;
+extern const struct nvbios_source nvbios_acpi_slow;
+extern const struct nvbios_source nvbios_pcirom;
+extern const struct nvbios_source nvbios_platform;
+extern const struct nvbios_source nvbios_of;
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/ramcfg.c b/drivers/gpu/drm/nouveau/core/subdev/bios/ramcfg.c
index 6c401f70ab99..1623c8dfe797 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/ramcfg.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/ramcfg.c
@@ -25,6 +25,7 @@
 #include <subdev/bios.h>
 #include <subdev/bios/bit.h>
 #include <subdev/bios/ramcfg.h>
+#include <subdev/bios/M0203.h>
 
 static u8
 nvbios_ramcfg_strap(struct nouveau_subdev *subdev)
@@ -54,12 +55,22 @@ nvbios_ramcfg_index(struct nouveau_subdev *subdev)
 	u8 strap = nvbios_ramcfg_strap(subdev);
 	u32 xlat = 0x00000000;
 	struct bit_entry bit_M;
+	struct nvbios_M0203E M0203E;
+	u8 ver, hdr;
 
 	if (!bit_entry(bios, 'M', &bit_M)) {
 		if (bit_M.version == 1 && bit_M.length >= 5)
 			xlat = nv_ro16(bios, bit_M.offset + 3);
-		if (bit_M.version == 2 && bit_M.length >= 3)
+		if (bit_M.version == 2 && bit_M.length >= 3) {
+			/*XXX: is M ever shorter than this?
+			 *     if not - what is xlat used for now?
+			 *     also - sigh..
+			 */
+			if (bit_M.length >= 7 &&
+			    nvbios_M0203Em(bios, strap, &ver, &hdr, &M0203E))
+				return M0203E.group;
 			xlat = nv_ro16(bios, bit_M.offset + 1);
+		}
 	}
 
 	if (xlat)
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c b/drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c
index 1811b2cb0472..c5685228c322 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c
@@ -75,31 +75,39 @@ nvbios_rammapEe(struct nouveau_bios *bios, int idx,
 }
 
 u32
-nvbios_rammapEm(struct nouveau_bios *bios, u16 khz,
-		u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
-{
-	int idx = 0;
-	u32 data;
-	while ((data = nvbios_rammapEe(bios, idx++, ver, hdr, cnt, len))) {
-		if (khz >= nv_ro16(bios, data + 0x00) &&
-		    khz <= nv_ro16(bios, data + 0x02))
-			break;
-	}
-	return data;
-}
-
-u32
-nvbios_rammapEp(struct nouveau_bios *bios, u16 khz,
+nvbios_rammapEp(struct nouveau_bios *bios, int idx,
 		u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
 		struct nvbios_ramcfg *p)
 {
-	u32 data = nvbios_rammapEm(bios, khz, ver, hdr, cnt, len);
+	u32 data = nvbios_rammapEe(bios, idx, ver, hdr, cnt, len), temp;
 	memset(p, 0x00, sizeof(*p));
+	p->rammap_ver = *ver;
+	p->rammap_hdr = *hdr;
 	switch (!!data * *ver) {
+	case 0x10:
+		p->rammap_min      =  nv_ro16(bios, data + 0x00);
+		p->rammap_max      =  nv_ro16(bios, data + 0x02);
+		p->rammap_10_04_02 = (nv_ro08(bios, data + 0x04) & 0x02) >> 1;
+		p->rammap_10_04_08 = (nv_ro08(bios, data + 0x04) & 0x08) >> 3;
+		break;
 	case 0x11:
+		p->rammap_min      =  nv_ro16(bios, data + 0x00);
+		p->rammap_max      =  nv_ro16(bios, data + 0x02);
 		p->rammap_11_08_01 = (nv_ro08(bios, data + 0x08) & 0x01) >> 0;
 		p->rammap_11_08_0c = (nv_ro08(bios, data + 0x08) & 0x0c) >> 2;
 		p->rammap_11_08_10 = (nv_ro08(bios, data + 0x08) & 0x10) >> 4;
+		temp = nv_ro32(bios, data + 0x09);
+		p->rammap_11_09_01ff = (temp & 0x000001ff) >> 0;
+		p->rammap_11_0a_03fe = (temp & 0x0003fe00) >> 9;
+		p->rammap_11_0a_0400 = (temp & 0x00040000) >> 18;
+		p->rammap_11_0a_0800 = (temp & 0x00080000) >> 19;
+		p->rammap_11_0b_01f0 = (temp & 0x01f00000) >> 20;
+		p->rammap_11_0b_0200 = (temp & 0x02000000) >> 25;
+		p->rammap_11_0b_0400 = (temp & 0x04000000) >> 26;
+		p->rammap_11_0b_0800 = (temp & 0x08000000) >> 27;
+		p->rammap_11_0d    =  nv_ro08(bios, data + 0x0d);
+		p->rammap_11_0e    =  nv_ro08(bios, data + 0x0e);
+		p->rammap_11_0f    =  nv_ro08(bios, data + 0x0f);
 		p->rammap_11_11_0c = (nv_ro08(bios, data + 0x11) & 0x0c) >> 2;
 		break;
 	default:
@@ -110,6 +118,20 @@ nvbios_rammapEp(struct nouveau_bios *bios, u16 khz,
 }
 
 u32
+nvbios_rammapEm(struct nouveau_bios *bios, u16 mhz,
+		u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		struct nvbios_ramcfg *info)
+{
+	int idx = 0;
+	u32 data;
+	while ((data = nvbios_rammapEp(bios, idx++, ver, hdr, cnt, len, info))) {
+		if (mhz >= info->rammap_min && mhz <= info->rammap_max)
+			break;
+	}
+	return data;
+}
+
+u32
 nvbios_rammapSe(struct nouveau_bios *bios, u32 data,
 		u8 ever, u8 ehdr, u8 ecnt, u8 elen, int idx,
 		u8 *ver, u8 *hdr)
@@ -129,8 +151,29 @@ nvbios_rammapSp(struct nouveau_bios *bios, u32 data,
 		u8 *ver, u8 *hdr, struct nvbios_ramcfg *p)
 {
 	data = nvbios_rammapSe(bios, data, ever, ehdr, ecnt, elen, idx, ver, hdr);
+	p->ramcfg_ver = *ver;
+	p->ramcfg_hdr = *hdr;
 	switch (!!data * *ver) {
+	case 0x10:
+		p->ramcfg_timing   =  nv_ro08(bios, data + 0x01);
+		p->ramcfg_10_02_01 = (nv_ro08(bios, data + 0x02) & 0x01) >> 0;
+		p->ramcfg_10_02_02 = (nv_ro08(bios, data + 0x02) & 0x02) >> 1;
+		p->ramcfg_10_02_04 = (nv_ro08(bios, data + 0x02) & 0x04) >> 2;
+		p->ramcfg_10_02_08 = (nv_ro08(bios, data + 0x02) & 0x08) >> 3;
+		p->ramcfg_10_02_10 = (nv_ro08(bios, data + 0x02) & 0x10) >> 4;
+		p->ramcfg_10_02_20 = (nv_ro08(bios, data + 0x02) & 0x20) >> 5;
+		p->ramcfg_10_DLLoff = (nv_ro08(bios, data + 0x02) & 0x40) >> 6;
+		p->ramcfg_10_03_0f = (nv_ro08(bios, data + 0x03) & 0x0f) >> 0;
+		p->ramcfg_10_04_01 = (nv_ro08(bios, data + 0x04) & 0x01) >> 0;
+		p->ramcfg_10_05    = (nv_ro08(bios, data + 0x05) & 0xff) >> 0;
+		p->ramcfg_10_06    = (nv_ro08(bios, data + 0x06) & 0xff) >> 0;
+		p->ramcfg_10_07    = (nv_ro08(bios, data + 0x07) & 0xff) >> 0;
+		p->ramcfg_10_08    = (nv_ro08(bios, data + 0x08) & 0xff) >> 0;
+		p->ramcfg_10_09_0f = (nv_ro08(bios, data + 0x09) & 0x0f) >> 0;
+		p->ramcfg_10_09_f0 = (nv_ro08(bios, data + 0x09) & 0xf0) >> 4;
+		break;
 	case 0x11:
+		p->ramcfg_timing   =  nv_ro08(bios, data + 0x00);
 		p->ramcfg_11_01_01 = (nv_ro08(bios, data + 0x01) & 0x01) >> 0;
 		p->ramcfg_11_01_02 = (nv_ro08(bios, data + 0x01) & 0x02) >> 1;
 		p->ramcfg_11_01_04 = (nv_ro08(bios, data + 0x01) & 0x04) >> 2;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/shadow.c b/drivers/gpu/drm/nouveau/core/subdev/bios/shadow.c
new file mode 100644
index 000000000000..bb9e0018d936
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/shadow.c
@@ -0,0 +1,270 @@
+/*
+ * Copyright 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs <bskeggs@redhat.com>
+ */
+
+#include "priv.h"
+#include <core/option.h>
+#include <subdev/bios/image.h>
+
+struct shadow {
+	struct nouveau_oclass base;
+	u32 skip;
+	const struct nvbios_source *func;
+	void *data;
+	u32 size;
+	int score;
+};
+
+static bool
+shadow_fetch(struct nouveau_bios *bios, u32 upto)
+{
+	struct shadow *mthd = (void *)nv_object(bios)->oclass;
+	const u32 limit = (upto + 3) & ~3;
+	const u32 start = bios->size;
+	void *data = mthd->data;
+	if (nvbios_extend(bios, limit) > 0) {
+		u32 read = mthd->func->read(data, start, limit - start, bios);
+		bios->size = start + read;
+	}
+	return bios->size >= limit;
+}
+
+static u8
+shadow_rd08(struct nouveau_object *object, u64 addr)
+{
+	struct nouveau_bios *bios = (void *)object;
+	if (shadow_fetch(bios, addr + 1))
+		return bios->data[addr];
+	return 0x00;
+}
+
+static u16
+shadow_rd16(struct nouveau_object *object, u64 addr)
+{
+	struct nouveau_bios *bios = (void *)object;
+	if (shadow_fetch(bios, addr + 2))
+		return get_unaligned_le16(&bios->data[addr]);
+	return 0x0000;
+}
+
+static u32
+shadow_rd32(struct nouveau_object *object, u64 addr)
+{
+	struct nouveau_bios *bios = (void *)object;
+	if (shadow_fetch(bios, addr + 4))
+		return get_unaligned_le32(&bios->data[addr]);
+	return 0x00000000;
+}
+
+static struct nouveau_oclass
+shadow_class = {
+	.handle = NV_SUBDEV(VBIOS, 0x00),
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.rd08 = shadow_rd08,
+		.rd16 = shadow_rd16,
+		.rd32 = shadow_rd32,
+	},
+};
+
+static int
+shadow_image(struct nouveau_bios *bios, int idx, struct shadow *mthd)
+{
+	struct nvbios_image image;
+	int score = 1;
+
+	if (!nvbios_image(bios, idx, &image)) {
+		nv_debug(bios, "image %d invalid\n", idx);
+		return 0;
+	}
+	nv_debug(bios, "%08x: type %02x, %d bytes\n",
+		 image.base, image.type, image.size);
+
+	if (!shadow_fetch(bios, image.size)) {
+		nv_debug(bios, "%08x: fetch failed\n", image.base);
+		return 0;
+	}
+
+	switch (image.type) {
+	case 0x00:
+		if (nvbios_checksum(&bios->data[image.base], image.size)) {
+			nv_debug(bios, "%08x: checksum failed\n", image.base);
+			if (mthd->func->rw)
+				score += 1;
+			score += 1;
+		} else {
+			score += 3;
+		}
+		break;
+	default:
+		score += 3;
+		break;
+	}
+
+	if (!image.last)
+		score += shadow_image(bios, idx + 1, mthd);
+	return score;
+}
+
+static int
+shadow_score(struct nouveau_bios *bios, struct shadow *mthd)
+{
+	struct nouveau_oclass *oclass = nv_object(bios)->oclass;
+	int score;
+	nv_object(bios)->oclass = &mthd->base;
+	score = shadow_image(bios, 0, mthd);
+	nv_object(bios)->oclass = oclass;
+	return score;
+
+}
+
+static int
+shadow_method(struct nouveau_bios *bios, struct shadow *mthd, const char *name)
+{
+	const struct nvbios_source *func = mthd->func;
+	if (func->name) {
+		nv_debug(bios, "trying %s...\n", name ? name : func->name);
+		if (func->init) {
+			mthd->data = func->init(bios, name);
+			if (IS_ERR(mthd->data)) {
+				mthd->data = NULL;
+				return 0;
+			}
+		}
+		mthd->score = shadow_score(bios, mthd);
+		if (func->fini)
+			func->fini(mthd->data);
+		nv_debug(bios, "scored %d\n", mthd->score);
+		mthd->data = bios->data;
+		mthd->size = bios->size;
+		bios->data  = NULL;
+		bios->size  = 0;
+	}
+	return mthd->score;
+}
+
+static u32
+shadow_fw_read(void *data, u32 offset, u32 length, struct nouveau_bios *bios)
+{
+	const struct firmware *fw = data;
+	if (offset + length <= fw->size) {
+		memcpy(bios->data + offset, fw->data + offset, length);
+		return length;
+	}
+	return 0;
+}
+
+static void *
+shadow_fw_init(struct nouveau_bios *bios, const char *name)
+{
+	struct device *dev = &nv_device(bios)->pdev->dev;
+	const struct firmware *fw;
+	int ret = request_firmware(&fw, name, dev);
+	if (ret)
+		return ERR_PTR(-ENOENT);
+	return (void *)fw;
+}
+
+static const struct nvbios_source
+shadow_fw = {
+	.name = "firmware",
+	.init = shadow_fw_init,
+	.fini = (void(*)(void *))release_firmware,
+	.read = shadow_fw_read,
+	.rw = false,
+};
+
+int
+nvbios_shadow(struct nouveau_bios *bios)
+{
+	struct shadow mthds[] = {
+		{ shadow_class, 0, &nvbios_of },
+		{ shadow_class, 0, &nvbios_ramin },
+		{ shadow_class, 0, &nvbios_rom },
+		{ shadow_class, 0, &nvbios_acpi_fast },
+		{ shadow_class, 4, &nvbios_acpi_slow },
+		{ shadow_class, 1, &nvbios_pcirom },
+		{ shadow_class, 1, &nvbios_platform },
+		{ shadow_class }
+	}, *mthd = mthds, *best = NULL;
+	const char *optarg;
+	char *source;
+	int optlen;
+
+	/* handle user-specified bios source */
+	optarg = nouveau_stropt(nv_device(bios)->cfgopt, "NvBios", &optlen);
+	source = optarg ? kstrndup(optarg, optlen, GFP_KERNEL) : NULL;
+	if (source) {
+		/* try to match one of the built-in methods */
+		for (mthd = mthds; mthd->func; mthd++) {
+			if (mthd->func->name &&
+			    !strcasecmp(source, mthd->func->name)) {
+				best = mthd;
+				if (shadow_method(bios, mthd, NULL))
+					break;
+			}
+		}
+
+		/* otherwise, attempt to load as firmware */
+		if (!best && (best = mthd)) {
+			mthd->func = &shadow_fw;
+			shadow_method(bios, mthd, source);
+			mthd->func = NULL;
+		}
+
+		if (!best->score) {
+			nv_error(bios, "%s invalid\n", source);
+			kfree(source);
+			source = NULL;
+		}
+	}
+
+	/* scan all potential bios sources, looking for best image */
+	if (!best || !best->score) {
+		for (mthd = mthds, best = mthd; mthd->func; mthd++) {
+			if (!mthd->skip || best->score < mthd->skip) {
+				if (shadow_method(bios, mthd, NULL)) {
+					if (mthd->score > best->score)
+						best = mthd;
+				}
+			}
+		}
+	}
+
+	/* cleanup the ones we didn't use */
+	for (mthd = mthds; mthd->func; mthd++) {
+		if (mthd != best)
+			kfree(mthd->data);
+	}
+
+	if (!best->score) {
+		nv_fatal(bios, "unable to locate usable image\n");
+		return -EINVAL;
+	}
+
+	nv_info(bios, "using image from %s\n", best->func ?
+		best->func->name : source);
+	bios->data = best->data;
+	bios->size = best->size;
+	kfree(source);
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/shadowacpi.c b/drivers/gpu/drm/nouveau/core/subdev/bios/shadowacpi.c
new file mode 100644
index 000000000000..bc130c12ec06
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/shadowacpi.c
@@ -0,0 +1,111 @@
+/*
+ * Copyright 2012 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include "priv.h"
+
+#if defined(CONFIG_ACPI) && defined(CONFIG_X86)
+int nouveau_acpi_get_bios_chunk(uint8_t *bios, int offset, int len);
+bool nouveau_acpi_rom_supported(struct pci_dev *pdev);
+#else
+static inline bool
+nouveau_acpi_rom_supported(struct pci_dev *pdev)
+{
+	return false;
+}
+
+static inline int
+nouveau_acpi_get_bios_chunk(uint8_t *bios, int offset, int len)
+{
+	return -EINVAL;
+}
+#endif
+
+/* This version of the shadow function disobeys the ACPI spec and tries
+ * to fetch in units of more than 4KiB at a time.  This is a LOT faster
+ * on some systems, such as Lenovo W530.
+ */
+static u32
+acpi_read_fast(void *data, u32 offset, u32 length, struct nouveau_bios *bios)
+{
+	u32 limit = (offset + length + 0xfff) & ~0xfff;
+	u32 start = offset & ~0x00000fff;
+	u32 fetch = limit - start;
+
+	if (nvbios_extend(bios, limit) > 0) {
+		int ret = nouveau_acpi_get_bios_chunk(bios->data, start, fetch);
+		if (ret == fetch)
+			return fetch;
+	}
+
+	return 0;
+}
+
+/* Other systems, such as the one in fdo#55948, will report a success
+ * but only return 4KiB of data.  The common bios fetching logic will
+ * detect an invalid image, and fall back to this version of the read
+ * function.
+ */
+static u32
+acpi_read_slow(void *data, u32 offset, u32 length, struct nouveau_bios *bios)
+{
+	u32 limit = (offset + length + 0xfff) & ~0xfff;
+	u32 start = offset & ~0xfff;
+	u32 fetch = 0;
+
+	if (nvbios_extend(bios, limit) > 0) {
+		while (start + fetch < limit) {
+			int ret = nouveau_acpi_get_bios_chunk(bios->data,
+							      start + fetch,
+							      0x1000);
+			if (ret != 0x1000)
+				break;
+			fetch += 0x1000;
+		}
+	}
+
+	return fetch;
+}
+
+static void *
+acpi_init(struct nouveau_bios *bios, const char *name)
+{
+	if (!nouveau_acpi_rom_supported(nv_device(bios)->pdev))
+		return ERR_PTR(-ENODEV);
+	return NULL;
+}
+
+const struct nvbios_source
+nvbios_acpi_fast = {
+	.name = "ACPI",
+	.init = acpi_init,
+	.read = acpi_read_fast,
+	.rw = false,
+};
+
+const struct nvbios_source
+nvbios_acpi_slow = {
+	.name = "ACPI",
+	.init = acpi_init,
+	.read = acpi_read_slow,
+	.rw = false,
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/shadowof.c b/drivers/gpu/drm/nouveau/core/subdev/bios/shadowof.c
new file mode 100644
index 000000000000..3abe487a6025
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/shadowof.c
@@ -0,0 +1,71 @@
+/*
+ * Copyright 2012 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include "priv.h"
+
+#if defined(__powerpc__)
+struct priv {
+	const void __iomem *data;
+	int size;
+};
+
+static u32
+of_read(void *data, u32 offset, u32 length, struct nouveau_bios *bios)
+{
+	struct priv *priv = data;
+	if (offset + length <= priv->size) {
+		memcpy_fromio(bios->data + offset, priv->data + offset, length);
+		return length;
+	}
+	return 0;
+}
+
+static void *
+of_init(struct nouveau_bios *bios, const char *name)
+{
+	struct pci_dev *pdev = nv_device(bios)->pdev;
+	struct device_node *dn;
+	struct priv *priv;
+	if (!(dn = pci_device_to_OF_node(pdev)))
+		return ERR_PTR(-ENODEV);
+	if (!(priv = kzalloc(sizeof(*priv), GFP_KERNEL)))
+		return ERR_PTR(-ENOMEM);
+	if ((priv->data = of_get_property(dn, "NVDA,BMP", &priv->size)))
+		return priv;
+	kfree(priv);
+	return ERR_PTR(-EINVAL);
+}
+
+const struct nvbios_source
+nvbios_of = {
+	.name = "OpenFirmware",
+	.init = of_init,
+	.fini = (void(*)(void *))kfree,
+	.read = of_read,
+	.rw = false,
+};
+#else
+const struct nvbios_source
+nvbios_of = {
+};
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/shadowpci.c b/drivers/gpu/drm/nouveau/core/subdev/bios/shadowpci.c
new file mode 100644
index 000000000000..1d0389c0abef
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/shadowpci.c
@@ -0,0 +1,108 @@
+/*
+ * Copyright 2012 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include "priv.h"
+
+struct priv {
+	struct pci_dev *pdev;
+	void __iomem *rom;
+	size_t size;
+};
+
+static u32
+pcirom_read(void *data, u32 offset, u32 length, struct nouveau_bios *bios)
+{
+	struct priv *priv = data;
+	if (offset + length <= priv->size) {
+		memcpy_fromio(bios->data + offset, priv->rom + offset, length);
+		return length;
+	}
+	return 0;
+}
+
+static void
+pcirom_fini(void *data)
+{
+	struct priv *priv = data;
+	pci_unmap_rom(priv->pdev, priv->rom);
+	pci_disable_rom(priv->pdev);
+	kfree(priv);
+}
+
+static void *
+pcirom_init(struct nouveau_bios *bios, const char *name)
+{
+	struct pci_dev *pdev = nv_device(bios)->pdev;
+	struct priv *priv = NULL;
+	int ret;
+
+	if (!(ret = pci_enable_rom(pdev))) {
+		if (ret = -ENOMEM,
+		    (priv = kmalloc(sizeof(*priv), GFP_KERNEL))) {
+			if (ret = -EFAULT,
+			    (priv->rom = pci_map_rom(pdev, &priv->size))) {
+				priv->pdev = pdev;
+				return priv;
+			}
+			kfree(priv);
+		}
+		pci_disable_rom(pdev);
+	}
+
+	return ERR_PTR(ret);
+}
+
+const struct nvbios_source
+nvbios_pcirom = {
+	.name = "PCIROM",
+	.init = pcirom_init,
+	.fini = pcirom_fini,
+	.read = pcirom_read,
+	.rw = true,
+};
+
+static void *
+platform_init(struct nouveau_bios *bios, const char *name)
+{
+	struct pci_dev *pdev = nv_device(bios)->pdev;
+	struct priv *priv;
+	int ret = -ENOMEM;
+
+	if ((priv = kmalloc(sizeof(*priv), GFP_KERNEL))) {
+		if (ret = -ENODEV,
+		    (priv->rom = pci_platform_rom(pdev, &priv->size)))
+			return priv;
+		kfree(priv);
+	}
+
+	return ERR_PTR(ret);
+}
+
+const struct nvbios_source
+nvbios_platform = {
+	.name = "PLATFORM",
+	.init = platform_init,
+	.fini = (void(*)(void *))kfree,
+	.read = pcirom_read,
+	.rw = true,
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/shadowramin.c b/drivers/gpu/drm/nouveau/core/subdev/bios/shadowramin.c
new file mode 100644
index 000000000000..a7a890fad1e5
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/shadowramin.c
@@ -0,0 +1,114 @@
+/*
+ * Copyright 2012 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include "priv.h"
+
+struct priv {
+	struct nouveau_bios *bios;
+	u32 bar0;
+};
+
+static u32
+pramin_read(void *data, u32 offset, u32 length, struct nouveau_bios *bios)
+{
+	u32 i;
+	if (offset + length <= 0x00100000) {
+		for (i = offset; i < offset + length; i += 4)
+			*(u32 *)&bios->data[i] = nv_rd32(bios, 0x700000 + i);
+		return length;
+	}
+	return 0;
+}
+
+static void
+pramin_fini(void *data)
+{
+	struct priv *priv = data;
+	if (priv) {
+		nv_wr32(priv->bios, 0x001700, priv->bar0);
+		kfree(priv);
+	}
+}
+
+static void *
+pramin_init(struct nouveau_bios *bios, const char *name)
+{
+	struct priv *priv = NULL;
+	u64 addr = 0;
+
+	/* PRAMIN always potentially available prior to nv50 */
+	if (nv_device(bios)->card_type < NV_50)
+		return NULL;
+
+	/* we can't get the bios image pointer without PDISP */
+	if (nv_device(bios)->card_type >= GM100)
+		addr = nv_rd32(bios, 0x021c04);
+	else
+	if (nv_device(bios)->card_type >= NV_C0)
+		addr = nv_rd32(bios, 0x022500);
+	if (addr & 0x00000001) {
+		nv_debug(bios, "... display disabled\n");
+		return ERR_PTR(-ENODEV);
+	}
+
+	/* check that the window is enabled and in vram, particularly
+	 * important as we don't want to be touching vram on an
+	 * uninitialised board
+	 */
+	addr = nv_rd32(bios, 0x619f04);
+	if (!(addr & 0x00000008)) {
+		nv_debug(bios, "... not enabled\n");
+		return ERR_PTR(-ENODEV);
+	}
+	if ( (addr & 0x00000003) != 1) {
+		nv_debug(bios, "... not in vram\n");
+		return ERR_PTR(-ENODEV);
+	}
+
+	/* some alternate method inherited from xf86-video-nv... */
+	addr = (addr & 0xffffff00) << 8;
+	if (!addr) {
+		addr  = (u64)nv_rd32(bios, 0x001700) << 16;
+		addr += 0xf0000;
+	}
+
+	/* modify bar0 PRAMIN window to cover the bios image */
+	if (!(priv = kmalloc(sizeof(*priv), GFP_KERNEL))) {
+		nv_error(bios, "... out of memory\n");
+		return ERR_PTR(-ENOMEM);
+	}
+
+	priv->bios = bios;
+	priv->bar0 = nv_rd32(bios, 0x001700);
+	nv_wr32(bios, 0x001700, addr >> 16);
+	return priv;
+}
+
+const struct nvbios_source
+nvbios_ramin = {
+	.name = "PRAMIN",
+	.init = pramin_init,
+	.fini = pramin_fini,
+	.read = pramin_read,
+	.rw = true,
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/shadowrom.c b/drivers/gpu/drm/nouveau/core/subdev/bios/shadowrom.c
new file mode 100644
index 000000000000..b7992bc3ffa5
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/shadowrom.c
@@ -0,0 +1,69 @@
+/*
+ * Copyright 2012 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include "priv.h"
+
+static u32
+prom_read(void *data, u32 offset, u32 length, struct nouveau_bios *bios)
+{
+	u32 i;
+	if (offset + length <= 0x00100000) {
+		for (i = offset; i < offset + length; i += 4)
+			*(u32 *)&bios->data[i] = nv_rd32(bios, 0x300000 + i);
+		return length;
+	}
+	return 0;
+}
+
+static void
+prom_fini(void *data)
+{
+	struct nouveau_bios *bios = data;
+	if (nv_device(bios)->card_type < NV_50)
+		nv_mask(bios, 0x001850, 0x00000001, 0x00000001);
+	else
+		nv_mask(bios, 0x088050, 0x00000001, 0x00000001);
+}
+
+static void *
+prom_init(struct nouveau_bios *bios, const char *name)
+{
+	if (nv_device(bios)->card_type < NV_50) {
+		if (nv_device(bios)->card_type == NV_40 &&
+		    nv_device(bios)->chipset >= 0x4c)
+			return ERR_PTR(-ENODEV);
+		nv_mask(bios, 0x001850, 0x00000001, 0x00000000);
+	} else {
+		nv_mask(bios, 0x088050, 0x00000001, 0x00000000);
+	}
+	return bios;
+}
+
+const struct nvbios_source
+nvbios_rom = {
+	.name = "PROM",
+	.init = prom_init,
+	.fini = prom_fini,
+	.read = prom_read,
+	.rw = false,
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/timing.c b/drivers/gpu/drm/nouveau/core/subdev/bios/timing.c
index 350d44ab2ba2..8521eca1ed9c 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/timing.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/timing.c
@@ -89,7 +89,49 @@ nvbios_timingEp(struct nouveau_bios *bios, int idx,
 		struct nvbios_ramcfg *p)
 {
 	u16 data = nvbios_timingEe(bios, idx, ver, hdr, cnt, len), temp;
+	p->timing_ver = *ver;
+	p->timing_hdr = *hdr;
 	switch (!!data * *ver) {
+	case 0x10:
+		p->timing_10_WR    = nv_ro08(bios, data + 0x00);
+		p->timing_10_WTR   = nv_ro08(bios, data + 0x01);
+		p->timing_10_CL    = nv_ro08(bios, data + 0x02);
+		p->timing_10_RC    = nv_ro08(bios, data + 0x03);
+		p->timing_10_RFC   = nv_ro08(bios, data + 0x05);
+		p->timing_10_RAS   = nv_ro08(bios, data + 0x07);
+		p->timing_10_RP    = nv_ro08(bios, data + 0x09);
+		p->timing_10_RCDRD = nv_ro08(bios, data + 0x0a);
+		p->timing_10_RCDWR = nv_ro08(bios, data + 0x0b);
+		p->timing_10_RRD   = nv_ro08(bios, data + 0x0c);
+		p->timing_10_13    = nv_ro08(bios, data + 0x0d);
+		p->timing_10_ODT   = nv_ro08(bios, data + 0x0e) & 0x07;
+
+		p->timing_10_24  = 0xff;
+		p->timing_10_21  = 0;
+		p->timing_10_20  = 0;
+		p->timing_10_CWL = 0;
+		p->timing_10_18  = 0;
+		p->timing_10_16  = 0;
+
+		switch (min_t(u8, *hdr, 25)) {
+		case 25:
+			p->timing_10_24  = nv_ro08(bios, data + 0x18);
+		case 24:
+		case 23:
+		case 22:
+			p->timing_10_21  = nv_ro08(bios, data + 0x15);
+		case 21:
+			p->timing_10_20  = nv_ro08(bios, data + 0x14);
+		case 20:
+			p->timing_10_CWL = nv_ro08(bios, data + 0x13);
+		case 19:
+			p->timing_10_18  = nv_ro08(bios, data + 0x12);
+		case 18:
+		case 17:
+			p->timing_10_16  = nv_ro08(bios, data + 0x10);
+		}
+
+		break;
 	case 0x20:
 		p->timing[0] = nv_ro32(bios, data + 0x00);
 		p->timing[1] = nv_ro32(bios, data + 0x04);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/base.c b/drivers/gpu/drm/nouveau/core/subdev/clock/base.c
index a276a711294a..e51b72d47129 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/base.c
@@ -573,7 +573,7 @@ nouveau_clock_create_(struct nouveau_object *parent,
 
 	clk->allow_reclock = allow_reclock;
 
-	ret = nvkm_notify_init(&device->event, nouveau_clock_pwrsrc, true,
+	ret = nvkm_notify_init(NULL, &device->event, nouveau_clock_pwrsrc, true,
 			       NULL, 0, 0, &clk->pwrsrc_ntfy);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/gk20a.c b/drivers/gpu/drm/nouveau/core/subdev/clock/gk20a.c
index 425a8d5e9129..fb4fad374bdd 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/gk20a.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/gk20a.c
@@ -109,7 +109,7 @@ struct gk20a_clk_pllg_params {
 };
 
 static const struct gk20a_clk_pllg_params gk20a_pllg_params = {
-	.min_vco = 1000, .max_vco = 1700,
+	.min_vco = 1000, .max_vco = 2064,
 	.min_u = 12, .max_u = 38,
 	.min_m = 1, .max_m = 255,
 	.min_n = 8, .max_n = 255,
@@ -470,76 +470,91 @@ gk20a_pstates[] = {
 	{
 		.base = {
 			.domain[nv_clk_src_gpc] = 72000,
+			.voltage = 0,
 		},
 	},
 	{
 		.base = {
 			.domain[nv_clk_src_gpc] = 108000,
+			.voltage = 1,
 		},
 	},
 	{
 		.base = {
 			.domain[nv_clk_src_gpc] = 180000,
+			.voltage = 2,
 		},
 	},
 	{
 		.base = {
 			.domain[nv_clk_src_gpc] = 252000,
+			.voltage = 3,
 		},
 	},
 	{
 		.base = {
 			.domain[nv_clk_src_gpc] = 324000,
+			.voltage = 4,
 		},
 	},
 	{
 		.base = {
 			.domain[nv_clk_src_gpc] = 396000,
+			.voltage = 5,
 		},
 	},
 	{
 		.base = {
 			.domain[nv_clk_src_gpc] = 468000,
+			.voltage = 6,
 		},
 	},
 	{
 		.base = {
 			.domain[nv_clk_src_gpc] = 540000,
+			.voltage = 7,
 		},
 	},
 	{
 		.base = {
 			.domain[nv_clk_src_gpc] = 612000,
+			.voltage = 8,
 		},
 	},
 	{
 		.base = {
 			.domain[nv_clk_src_gpc] = 648000,
+			.voltage = 9,
 		},
 	},
 	{
 		.base = {
 			.domain[nv_clk_src_gpc] = 684000,
+			.voltage = 10,
 		},
 	},
 	{
 		.base = {
 			.domain[nv_clk_src_gpc] = 708000,
+			.voltage = 11,
 		},
 	},
 	{
 		.base = {
 			.domain[nv_clk_src_gpc] = 756000,
+			.voltage = 12,
 		},
 	},
 	{
 		.base = {
 			.domain[nv_clk_src_gpc] = 804000,
+			.voltage = 13,
 		},
 	},
 	{
 		.base = {
 			.domain[nv_clk_src_gpc] = 852000,
+			.voltage = 14,
 		},
 	},
 };
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.c b/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.c
index 087012b18956..07ad01247675 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.c
@@ -20,8 +20,10 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  * Authors: Ben Skeggs
+ *          Roy Spliet
  */
 
+#include <engine/fifo.h>
 #include <subdev/bios.h>
 #include <subdev/bios/pll.h>
 #include <subdev/timer.h>
@@ -42,9 +44,17 @@ static u32
 read_vco(struct nva3_clock_priv *priv, int clk)
 {
 	u32 sctl = nv_rd32(priv, 0x4120 + (clk * 4));
-	if ((sctl & 0x00000030) != 0x00000030)
+
+	switch (sctl & 0x00000030) {
+	case 0x00000000:
+		return nv_device(priv)->crystal;
+	case 0x00000020:
 		return read_pll(priv, 0x41, 0x00e820);
-	return read_pll(priv, 0x42, 0x00e8a0);
+	case 0x00000030:
+		return read_pll(priv, 0x42, 0x00e8a0);
+	default:
+		return 0;
+	}
 }
 
 static u32
@@ -66,14 +76,25 @@ read_clk(struct nva3_clock_priv *priv, int clk, bool ignore_en)
 	if (!ignore_en && !(sctl & 0x00000100))
 		return 0;
 
+	/* out_alt */
+	if (sctl & 0x00000400)
+		return 108000;
+
+	/* vco_out */
 	switch (sctl & 0x00003000) {
 	case 0x00000000:
-		return nv_device(priv)->crystal;
+		if (!(sctl & 0x00000200))
+			return nv_device(priv)->crystal;
+		return 0;
 	case 0x00002000:
 		if (sctl & 0x00000040)
 			return 108000;
 		return 100000;
 	case 0x00003000:
+		/* vco_enable */
+		if (!(sctl & 0x00000001))
+			return 0;
+
 		sclk = read_vco(priv, clk);
 		sdiv = ((sctl & 0x003f0000) >> 16) + 2;
 		return (sclk * 2) / sdiv;
@@ -95,7 +116,9 @@ read_pll(struct nva3_clock_priv *priv, int clk, u32 pll)
 			N = (coef & 0x0000ff00) >> 8;
 			P = (coef & 0x003f0000) >> 16;
 
-			/* no post-divider on these.. */
+			/* no post-divider on these..
+			 * XXX: it looks more like two post-"dividers" that
+			 * cross each other out in the default RPLL config */
 			if ((pll & 0x00ff00) == 0x00e800)
 				P = 1;
 
@@ -114,13 +137,13 @@ static int
 nva3_clock_read(struct nouveau_clock *clk, enum nv_clk_src src)
 {
 	struct nva3_clock_priv *priv = (void *)clk;
+	u32 hsrc;
 
 	switch (src) {
 	case nv_clk_src_crystal:
 		return nv_device(priv)->crystal;
-	case nv_clk_src_href:
-		return 100000;
 	case nv_clk_src_core:
+	case nv_clk_src_core_intm:
 		return read_pll(priv, 0x00, 0x4200);
 	case nv_clk_src_shader:
 		return read_pll(priv, 0x01, 0x4220);
@@ -132,24 +155,33 @@ nva3_clock_read(struct nouveau_clock *clk, enum nv_clk_src src)
 		return read_clk(priv, 0x21, false);
 	case nv_clk_src_daemon:
 		return read_clk(priv, 0x25, false);
+	case nv_clk_src_host:
+		hsrc = (nv_rd32(priv, 0xc040) & 0x30000000) >> 28;
+		switch (hsrc) {
+		case 0:
+			return read_clk(priv, 0x1d, false);
+		case 2:
+		case 3:
+			return 277000;
+		default:
+			nv_error(clk, "unknown HOST clock source %d\n", hsrc);
+			return -EINVAL;
+		}
 	default:
 		nv_error(clk, "invalid clock source %d\n", src);
 		return -EINVAL;
 	}
+
+	return 0;
 }
 
 int
-nva3_clock_info(struct nouveau_clock *clock, int clk, u32 pll, u32 khz,
+nva3_clk_info(struct nouveau_clock *clock, int clk, u32 khz,
 		struct nva3_clock_info *info)
 {
-	struct nouveau_bios *bios = nouveau_bios(clock);
 	struct nva3_clock_priv *priv = (void *)clock;
-	struct nvbios_pll limits;
-	u32 oclk, sclk, sdiv;
-	int P, N, M, diff;
-	int ret;
+	u32 oclk, sclk, sdiv, diff;
 
-	info->pll = 0;
 	info->clk = 0;
 
 	switch (khz) {
@@ -164,43 +196,69 @@ nva3_clock_info(struct nouveau_clock *clock, int clk, u32 pll, u32 khz,
 		return khz;
 	default:
 		sclk = read_vco(priv, clk);
-		sdiv = min((sclk * 2) / (khz - 2999), (u32)65);
-		/* if the clock has a PLL attached, and we can get a within
-		 * [-2, 3) MHz of a divider, we'll disable the PLL and use
-		 * the divider instead.
-		 *
-		 * divider can go as low as 2, limited here because NVIDIA
+		sdiv = min((sclk * 2) / khz, (u32)65);
+		oclk = (sclk * 2) / sdiv;
+		diff = ((khz + 3000) - oclk);
+
+		/* When imprecise, play it safe and aim for a clock lower than
+		 * desired rather than higher */
+		if (diff < 0) {
+			sdiv++;
+			oclk = (sclk * 2) / sdiv;
+		}
+
+		/* divider can go as low as 2, limited here because NVIDIA
 		 * and the VBIOS on my NVA8 seem to prefer using the PLL
 		 * for 810MHz - is there a good reason?
-		 */
+		 * XXX: PLLs with refclk 810MHz?  */
 		if (sdiv > 4) {
-			oclk = (sclk * 2) / sdiv;
-			diff = khz - oclk;
-			if (!pll || (diff >= -2000 && diff < 3000)) {
-				info->clk = (((sdiv - 2) << 16) | 0x00003100);
-				return oclk;
-			}
+			info->clk = (((sdiv - 2) << 16) | 0x00003100);
+			return oclk;
 		}
 
-		if (!pll)
-			return -ERANGE;
 		break;
 	}
 
+	return -ERANGE;
+}
+
+int
+nva3_pll_info(struct nouveau_clock *clock, int clk, u32 pll, u32 khz,
+		struct nva3_clock_info *info)
+{
+	struct nouveau_bios *bios = nouveau_bios(clock);
+	struct nva3_clock_priv *priv = (void *)clock;
+	struct nvbios_pll limits;
+	int P, N, M, diff;
+	int ret;
+
+	info->pll = 0;
+
+	/* If we can get a within [-2, 3) MHz of a divider, we'll disable the
+	 * PLL and use the divider instead. */
+	ret = nva3_clk_info(clock, clk, khz, info);
+	diff = khz - ret;
+	if (!pll || (diff >= -2000 && diff < 3000)) {
+		goto out;
+	}
+
+	/* Try with PLL */
 	ret = nvbios_pll_parse(bios, pll, &limits);
 	if (ret)
 		return ret;
 
-	limits.refclk = read_clk(priv, clk - 0x10, true);
-	if (!limits.refclk)
+	ret = nva3_clk_info(clock, clk - 0x10, limits.refclk, info);
+	if (ret != limits.refclk)
 		return -EINVAL;
 
 	ret = nva3_pll_calc(nv_subdev(priv), &limits, khz, &N, NULL, &M, &P);
 	if (ret >= 0) {
-		info->clk = nv_rd32(priv, 0x4120 + (clk * 4));
 		info->pll = (P << 16) | (N << 8) | M;
 	}
 
+out:
+	info->fb_delay = max(((khz + 7566) / 15133), (u32) 18);
+
 	return ret ? ret : -ERANGE;
 }
 
@@ -208,13 +266,76 @@ static int
 calc_clk(struct nva3_clock_priv *priv, struct nouveau_cstate *cstate,
 	 int clk, u32 pll, int idx)
 {
-	int ret = nva3_clock_info(&priv->base, clk, pll, cstate->domain[idx],
+	int ret = nva3_pll_info(&priv->base, clk, pll, cstate->domain[idx],
 				  &priv->eng[idx]);
 	if (ret >= 0)
 		return 0;
 	return ret;
 }
 
+static int
+calc_host(struct nva3_clock_priv *priv, struct nouveau_cstate *cstate)
+{
+	int ret = 0;
+	u32 kHz = cstate->domain[nv_clk_src_host];
+	struct nva3_clock_info *info = &priv->eng[nv_clk_src_host];
+
+	if (kHz == 277000) {
+		info->clk = 0;
+		info->host_out = NVA3_HOST_277;
+		return 0;
+	}
+
+	info->host_out = NVA3_HOST_CLK;
+
+	ret = nva3_clk_info(&priv->base, 0x1d, kHz, info);
+	if (ret >= 0)
+		return 0;
+	return ret;
+}
+
+int
+nva3_clock_pre(struct nouveau_clock *clk, unsigned long *flags)
+{
+	struct nouveau_fifo *pfifo = nouveau_fifo(clk);
+
+	/* halt and idle execution engines */
+	nv_mask(clk, 0x020060, 0x00070000, 0x00000000);
+	nv_mask(clk, 0x002504, 0x00000001, 0x00000001);
+	/* Wait until the interrupt handler is finished */
+	if (!nv_wait(clk, 0x000100, 0xffffffff, 0x00000000))
+		return -EBUSY;
+
+	if (pfifo)
+		pfifo->pause(pfifo, flags);
+
+	if (!nv_wait(clk, 0x002504, 0x00000010, 0x00000010))
+		return -EIO;
+	if (!nv_wait(clk, 0x00251c, 0x0000003f, 0x0000003f))
+		return -EIO;
+
+	return 0;
+}
+
+void
+nva3_clock_post(struct nouveau_clock *clk, unsigned long *flags)
+{
+	struct nouveau_fifo *pfifo = nouveau_fifo(clk);
+
+	if (pfifo && flags)
+		pfifo->start(pfifo, flags);
+
+	nv_mask(clk, 0x002504, 0x00000001, 0x00000000);
+	nv_mask(clk, 0x020060, 0x00070000, 0x00040000);
+}
+
+static void
+disable_clk_src(struct nva3_clock_priv *priv, u32 src)
+{
+	nv_mask(priv, src, 0x00000100, 0x00000000);
+	nv_mask(priv, src, 0x00000001, 0x00000000);
+}
+
 static void
 prog_pll(struct nva3_clock_priv *priv, int clk, u32 pll, int idx)
 {
@@ -223,24 +344,35 @@ prog_pll(struct nva3_clock_priv *priv, int clk, u32 pll, int idx)
 	const u32 src1 = 0x004160 + (clk * 4);
 	const u32 ctrl = pll + 0;
 	const u32 coef = pll + 4;
+	u32 bypass;
 
 	if (info->pll) {
-		nv_mask(priv, src0, 0x00000101, 0x00000101);
+		/* Always start from a non-PLL clock */
+		bypass = nv_rd32(priv, ctrl)  & 0x00000008;
+		if (!bypass) {
+			nv_mask(priv, src1, 0x00000101, 0x00000101);
+			nv_mask(priv, ctrl, 0x00000008, 0x00000008);
+			udelay(20);
+		}
+
+		nv_mask(priv, src0, 0x003f3141, 0x00000101 | info->clk);
 		nv_wr32(priv, coef, info->pll);
 		nv_mask(priv, ctrl, 0x00000015, 0x00000015);
 		nv_mask(priv, ctrl, 0x00000010, 0x00000000);
-		nv_wait(priv, ctrl, 0x00020000, 0x00020000);
+		if (!nv_wait(priv, ctrl, 0x00020000, 0x00020000)) {
+			nv_mask(priv, ctrl, 0x00000010, 0x00000010);
+			nv_mask(priv, src0, 0x00000101, 0x00000000);
+			return;
+		}
 		nv_mask(priv, ctrl, 0x00000010, 0x00000010);
 		nv_mask(priv, ctrl, 0x00000008, 0x00000000);
-		nv_mask(priv, src1, 0x00000100, 0x00000000);
-		nv_mask(priv, src1, 0x00000001, 0x00000000);
+		disable_clk_src(priv, src1);
 	} else {
 		nv_mask(priv, src1, 0x003f3141, 0x00000101 | info->clk);
 		nv_mask(priv, ctrl, 0x00000018, 0x00000018);
 		udelay(20);
 		nv_mask(priv, ctrl, 0x00000001, 0x00000000);
-		nv_mask(priv, src0, 0x00000100, 0x00000000);
-		nv_mask(priv, src0, 0x00000001, 0x00000000);
+		disable_clk_src(priv, src0);
 	}
 }
 
@@ -251,18 +383,72 @@ prog_clk(struct nva3_clock_priv *priv, int clk, int idx)
 	nv_mask(priv, 0x004120 + (clk * 4), 0x003f3141, 0x00000101 | info->clk);
 }
 
+static void
+prog_host(struct nva3_clock_priv *priv)
+{
+	struct nva3_clock_info *info = &priv->eng[nv_clk_src_host];
+	u32 hsrc = (nv_rd32(priv, 0xc040));
+
+	switch (info->host_out) {
+	case NVA3_HOST_277:
+		if ((hsrc & 0x30000000) == 0) {
+			nv_wr32(priv, 0xc040, hsrc | 0x20000000);
+			disable_clk_src(priv, 0x4194);
+		}
+		break;
+	case NVA3_HOST_CLK:
+		prog_clk(priv, 0x1d, nv_clk_src_host);
+		if ((hsrc & 0x30000000) >= 0x20000000) {
+			nv_wr32(priv, 0xc040, hsrc & ~0x30000000);
+		}
+		break;
+	default:
+		break;
+	}
+
+	/* This seems to be a clock gating factor on idle, always set to 64 */
+	nv_wr32(priv, 0xc044, 0x3e);
+}
+
+static void
+prog_core(struct nva3_clock_priv *priv, int idx)
+{
+	struct nva3_clock_info *info = &priv->eng[idx];
+	u32 fb_delay = nv_rd32(priv, 0x10002c);
+
+	if (fb_delay < info->fb_delay)
+		nv_wr32(priv, 0x10002c, info->fb_delay);
+
+	prog_pll(priv, 0x00, 0x004200, idx);
+
+	if (fb_delay > info->fb_delay)
+		nv_wr32(priv, 0x10002c, info->fb_delay);
+}
+
 static int
 nva3_clock_calc(struct nouveau_clock *clk, struct nouveau_cstate *cstate)
 {
 	struct nva3_clock_priv *priv = (void *)clk;
+	struct nva3_clock_info *core = &priv->eng[nv_clk_src_core];
 	int ret;
 
 	if ((ret = calc_clk(priv, cstate, 0x10, 0x4200, nv_clk_src_core)) ||
 	    (ret = calc_clk(priv, cstate, 0x11, 0x4220, nv_clk_src_shader)) ||
 	    (ret = calc_clk(priv, cstate, 0x20, 0x0000, nv_clk_src_disp)) ||
-	    (ret = calc_clk(priv, cstate, 0x21, 0x0000, nv_clk_src_vdec)))
+	    (ret = calc_clk(priv, cstate, 0x21, 0x0000, nv_clk_src_vdec)) ||
+	    (ret = calc_host(priv, cstate)))
 		return ret;
 
+	/* XXX: Should be reading the highest bit in the VBIOS clock to decide
+	 * whether to use a PLL or not... but using a PLL defeats the purpose */
+	if (core->pll) {
+		ret = nva3_clk_info(clk, 0x10,
+				cstate->domain[nv_clk_src_core_intm],
+				&priv->eng[nv_clk_src_core_intm]);
+		if (ret < 0)
+			return ret;
+	}
+
 	return 0;
 }
 
@@ -270,11 +456,31 @@ static int
 nva3_clock_prog(struct nouveau_clock *clk)
 {
 	struct nva3_clock_priv *priv = (void *)clk;
-	prog_pll(priv, 0x00, 0x004200, nv_clk_src_core);
+	struct nva3_clock_info *core = &priv->eng[nv_clk_src_core];
+	int ret = 0;
+	unsigned long flags;
+	unsigned long *f = &flags;
+
+	ret = nva3_clock_pre(clk, f);
+	if (ret)
+		goto out;
+
+	if (core->pll)
+		prog_core(priv, nv_clk_src_core_intm);
+
+	prog_core(priv,  nv_clk_src_core);
 	prog_pll(priv, 0x01, 0x004220, nv_clk_src_shader);
 	prog_clk(priv, 0x20, nv_clk_src_disp);
 	prog_clk(priv, 0x21, nv_clk_src_vdec);
-	return 0;
+	prog_host(priv);
+
+out:
+	if (ret == -EBUSY)
+		f = NULL;
+
+	nva3_clock_post(clk, f);
+
+	return ret;
 }
 
 static void
@@ -284,13 +490,14 @@ nva3_clock_tidy(struct nouveau_clock *clk)
 
 static struct nouveau_clocks
 nva3_domain[] = {
-	{ nv_clk_src_crystal, 0xff },
-	{ nv_clk_src_href   , 0xff },
-	{ nv_clk_src_core   , 0x00, 0, "core", 1000 },
-	{ nv_clk_src_shader , 0x01, 0, "shader", 1000 },
-	{ nv_clk_src_mem    , 0x02, 0, "memory", 1000 },
-	{ nv_clk_src_vdec   , 0x03 },
-	{ nv_clk_src_disp   , 0x04 },
+	{ nv_clk_src_crystal  , 0xff },
+	{ nv_clk_src_core     , 0x00, 0, "core", 1000 },
+	{ nv_clk_src_shader   , 0x01, 0, "shader", 1000 },
+	{ nv_clk_src_mem      , 0x02, 0, "memory", 1000 },
+	{ nv_clk_src_vdec     , 0x03 },
+	{ nv_clk_src_disp     , 0x04 },
+	{ nv_clk_src_host     , 0x05 },
+	{ nv_clk_src_core_intm, 0x06 },
 	{ nv_clk_src_max }
 };
 
@@ -303,7 +510,7 @@ nva3_clock_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	int ret;
 
 	ret = nouveau_clock_create(parent, engine, oclass, nva3_domain, NULL, 0,
-				   false, &priv);
+				   true, &priv);
 	*pobject = nv_object(priv);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.h b/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.h
index 6229a509b42e..a45a1038b12f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.h
@@ -6,9 +6,15 @@
 struct nva3_clock_info {
 	u32 clk;
 	u32 pll;
+	enum {
+		NVA3_HOST_277,
+		NVA3_HOST_CLK,
+	} host_out;
+	u32 fb_delay;
 };
 
-int nva3_clock_info(struct nouveau_clock *, int, u32, u32,
+int nva3_pll_info(struct nouveau_clock *, int, u32, u32,
 		    struct nva3_clock_info *);
-
+int nva3_clock_pre(struct nouveau_clock *clk, unsigned long *flags);
+void nva3_clock_post(struct nouveau_clock *clk, unsigned long *flags);
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/nvaa.c b/drivers/gpu/drm/nouveau/core/subdev/clock/nvaa.c
index 74e19731b1b7..54aeab8005a0 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/nvaa.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/nvaa.c
@@ -28,6 +28,7 @@
 #include <subdev/timer.h>
 #include <subdev/clock.h>
 
+#include "nva3.h"
 #include "pll.h"
 
 struct nvaa_clock_priv {
@@ -299,25 +300,14 @@ static int
 nvaa_clock_prog(struct nouveau_clock *clk)
 {
 	struct nvaa_clock_priv *priv = (void *)clk;
-	struct nouveau_fifo *pfifo = nouveau_fifo(clk);
+	u32 pllmask = 0, mast;
 	unsigned long flags;
-	u32 pllmask = 0, mast, ptherm_gate;
-	int ret = -EBUSY;
-
-	/* halt and idle execution engines */
-	ptherm_gate = nv_mask(clk, 0x020060, 0x00070000, 0x00000000);
-	nv_mask(clk, 0x002504, 0x00000001, 0x00000001);
-	/* Wait until the interrupt handler is finished */
-	if (!nv_wait(clk, 0x000100, 0xffffffff, 0x00000000))
-		goto resume;
-
-	if (pfifo)
-		pfifo->pause(pfifo, &flags);
+	unsigned long *f = &flags;
+	int ret = 0;
 
-	if (!nv_wait(clk, 0x002504, 0x00000010, 0x00000010))
-		goto resume;
-	if (!nv_wait(clk, 0x00251c, 0x0000003f, 0x0000003f))
-		goto resume;
+	ret = nva3_clock_pre(clk, f);
+	if (ret)
+		goto out;
 
 	/* First switch to safe clocks: href */
 	mast = nv_mask(clk, 0xc054, 0x03400e70, 0x03400640);
@@ -375,15 +365,8 @@ nvaa_clock_prog(struct nouveau_clock *clk)
 	}
 
 	nv_wr32(clk, 0xc054, mast);
-	ret = 0;
 
 resume:
-	if (pfifo)
-		pfifo->start(pfifo, &flags);
-
-	nv_mask(clk, 0x002504, 0x00000001, 0x00000000);
-	nv_wr32(clk, 0x020060, ptherm_gate);
-
 	/* Disable some PLLs and dividers when unused */
 	if (priv->csrc != nv_clk_src_core) {
 		nv_wr32(clk, 0x4040, 0x00000000);
@@ -395,6 +378,12 @@ resume:
 		nv_mask(clk, 0x4020, 0x80000000, 0x00000000);
 	}
 
+out:
+	if (ret == -EBUSY)
+		f = NULL;
+
+	nva3_clock_post(clk, f);
+
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/base.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/base.c
index 239acfe876c3..0e45cee82463 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/base.c
@@ -24,8 +24,6 @@
 
 #include <core/option.h>
 
-#include <subdev/bios.h>
-#include <subdev/bios/init.h>
 #include <subdev/vga.h>
 
 #include "priv.h"
@@ -56,7 +54,7 @@ _nouveau_devinit_init(struct nouveau_object *object)
 	if (ret)
 		return ret;
 
-	ret = nvbios_init(&devinit->base, devinit->post);
+	ret = impl->post(&devinit->base, devinit->post);
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/fbmem.h b/drivers/gpu/drm/nouveau/core/subdev/devinit/fbmem.h
index 4fe49cf4c99a..6103484fea72 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/fbmem.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/fbmem.h
@@ -26,22 +26,8 @@
 
 #include <core/device.h>
 
-#define NV04_PFB_BOOT_0						0x00100000
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT			0x00000003
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT_32MB			0x00000000
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT_4MB			0x00000001
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT_8MB			0x00000002
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT_16MB			0x00000003
-#	define NV04_PFB_BOOT_0_RAM_WIDTH_128			0x00000004
-#	define NV04_PFB_BOOT_0_RAM_TYPE				0x00000028
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_8MBIT		0x00000000
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_16MBIT		0x00000008
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_16MBIT_4BANK	0x00000010
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_16MBIT		0x00000018
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_64MBIT		0x00000020
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_64MBITX16		0x00000028
-#	define NV04_PFB_BOOT_0_UMA_ENABLE			0x00000100
-#	define NV04_PFB_BOOT_0_UMA_SIZE				0x0000f000
+#include <subdev/fb/regsnv04.h>
+
 #define NV04_PFB_DEBUG_0					0x00100080
 #	define NV04_PFB_DEBUG_0_PAGE_MODE			0x00000001
 #	define NV04_PFB_DEBUG_0_REFRESH_OFF			0x00000010
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/gm107.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/gm107.c
index c69bc7f54e37..4ba43d6a1ec8 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/gm107.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/gm107.c
@@ -24,7 +24,7 @@
 
 #include "nv50.h"
 
-static u64
+u64
 gm107_devinit_disable(struct nouveau_devinit *devinit)
 {
 	struct nv50_devinit_priv *priv = (void *)devinit;
@@ -53,4 +53,5 @@ gm107_devinit_oclass = &(struct nouveau_devinit_impl) {
 	},
 	.pll_set = nvc0_devinit_pll_set,
 	.disable = gm107_devinit_disable,
+	.post = nvbios_init,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/gm204.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/gm204.c
new file mode 100644
index 000000000000..e44a86662a2a
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/gm204.c
@@ -0,0 +1,173 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include <subdev/bios.h>
+#include <subdev/bios/bit.h>
+#include <subdev/bios/pmu.h>
+
+#include "nv50.h"
+
+static void
+pmu_code(struct nv50_devinit_priv *priv, u32 pmu, u32 img, u32 len, bool sec)
+{
+	struct nouveau_bios *bios = nouveau_bios(priv);
+	int i;
+
+	nv_wr32(priv, 0x10a180, 0x01000000 | (sec ? 0x10000000 : 0) | pmu);
+	for (i = 0; i < len; i += 4) {
+		if ((i & 0xff) == 0)
+			nv_wr32(priv, 0x10a188, (pmu + i) >> 8);
+		nv_wr32(priv, 0x10a184, nv_ro32(bios, img + i));
+	}
+
+	while (i & 0xff) {
+		nv_wr32(priv, 0x10a184, 0x00000000);
+		i += 4;
+	}
+}
+
+static void
+pmu_data(struct nv50_devinit_priv *priv, u32 pmu, u32 img, u32 len)
+{
+	struct nouveau_bios *bios = nouveau_bios(priv);
+	int i;
+
+	nv_wr32(priv, 0x10a1c0, 0x01000000 | pmu);
+	for (i = 0; i < len; i += 4)
+		nv_wr32(priv, 0x10a1c4, nv_ro32(bios, img + i));
+}
+
+static u32
+pmu_args(struct nv50_devinit_priv *priv, u32 argp, u32 argi)
+{
+	nv_wr32(priv, 0x10a1c0, argp);
+	nv_wr32(priv, 0x10a1c0, nv_rd32(priv, 0x10a1c4) + argi);
+	return nv_rd32(priv, 0x10a1c4);
+}
+
+static void
+pmu_exec(struct nv50_devinit_priv *priv, u32 init_addr)
+{
+	nv_wr32(priv, 0x10a104, init_addr);
+	nv_wr32(priv, 0x10a10c, 0x00000000);
+	nv_wr32(priv, 0x10a100, 0x00000002);
+}
+
+static int
+pmu_load(struct nv50_devinit_priv *priv, u8 type, bool post,
+	 u32 *init_addr_pmu, u32 *args_addr_pmu)
+{
+	struct nouveau_bios *bios = nouveau_bios(priv);
+	struct nvbios_pmuR pmu;
+
+	if (!nvbios_pmuRm(bios, type, &pmu)) {
+		nv_error(priv, "VBIOS PMU fuc %02x not found\n", type);
+		return -EINVAL;
+	}
+
+	if (!post)
+		return 0;
+
+	pmu_code(priv, pmu.boot_addr_pmu, pmu.boot_addr, pmu.boot_size, false);
+	pmu_code(priv, pmu.code_addr_pmu, pmu.code_addr, pmu.code_size, true);
+	pmu_data(priv, pmu.data_addr_pmu, pmu.data_addr, pmu.data_size);
+
+	if (init_addr_pmu) {
+		*init_addr_pmu = pmu.init_addr_pmu;
+		*args_addr_pmu = pmu.args_addr_pmu;
+		return 0;
+	}
+
+	return pmu_exec(priv, pmu.init_addr_pmu), 0;
+}
+
+static int
+gm204_devinit_post(struct nouveau_subdev *subdev, bool post)
+{
+	struct nv50_devinit_priv *priv = (void *)nouveau_devinit(subdev);
+	struct nouveau_bios *bios = nouveau_bios(priv);
+	struct bit_entry bit_I;
+	u32 init, args;
+	int ret;
+
+	if (bit_entry(bios, 'I', &bit_I) || bit_I.version != 1 ||
+					    bit_I.length < 0x1c) {
+		nv_error(priv, "VBIOS PMU init data not found\n");
+		return -EINVAL;
+	}
+
+	/* reset PMU and load init table parser ucode */
+	if (post) {
+		nv_mask(priv, 0x000200, 0x00002000, 0x00000000);
+		nv_mask(priv, 0x000200, 0x00002000, 0x00002000);
+		nv_rd32(priv, 0x000200);
+		while (nv_rd32(priv, 0x10a10c) & 0x00000006) {
+		}
+	}
+
+	ret = pmu_load(priv, 0x04, post, &init, &args);
+	if (ret)
+		return ret;
+
+	/* upload first chunk of init data */
+	if (post) {
+		u32 pmu = pmu_args(priv, args + 0x08, 0x08);
+		u32 img = nv_ro16(bios, bit_I.offset + 0x14);
+		u32 len = nv_ro16(bios, bit_I.offset + 0x16);
+		pmu_data(priv, pmu, img, len);
+	}
+
+	/* upload second chunk of init data */
+	if (post) {
+		u32 pmu = pmu_args(priv, args + 0x08, 0x10);
+		u32 img = nv_ro16(bios, bit_I.offset + 0x18);
+		u32 len = nv_ro16(bios, bit_I.offset + 0x1a);
+		pmu_data(priv, pmu, img, len);
+	}
+
+	/* execute init tables */
+	if (post) {
+		nv_wr32(priv, 0x10a040, 0x00005000);
+		pmu_exec(priv, init);
+		while (!(nv_rd32(priv, 0x10a040) & 0x00002000)) {
+		}
+	}
+
+	/* load and execute some other ucode image (bios therm?) */
+	return pmu_load(priv, 0x01, post, NULL, NULL);
+}
+
+struct nouveau_oclass *
+gm204_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x07),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv50_devinit_ctor,
+		.dtor = _nouveau_devinit_dtor,
+		.init = nv50_devinit_init,
+		.fini = _nouveau_devinit_fini,
+	},
+	.pll_set = nvc0_devinit_pll_set,
+	.disable = gm107_devinit_disable,
+	.post = gm204_devinit_post,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.c
index 052ad690b468..65651c50f6ea 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.c
@@ -464,4 +464,5 @@ nv04_devinit_oclass = &(struct nouveau_devinit_impl) {
 	},
 	.meminit = nv04_devinit_meminit,
 	.pll_set = nv04_devinit_pll_set,
+	.post = nvbios_init,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv05.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv05.c
index 4a19c10e5178..a2007a3efc4d 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv05.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv05.c
@@ -136,4 +136,5 @@ nv05_devinit_oclass = &(struct nouveau_devinit_impl) {
 	},
 	.meminit = nv05_devinit_meminit,
 	.pll_set = nv04_devinit_pll_set,
+	.post = nvbios_init,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv10.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv10.c
index 3b8d657da279..178b46f79b50 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv10.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv10.c
@@ -107,4 +107,5 @@ nv10_devinit_oclass = &(struct nouveau_devinit_impl) {
 	},
 	.meminit = nv10_devinit_meminit,
 	.pll_set = nv04_devinit_pll_set,
+	.post = nvbios_init,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv1a.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv1a.c
index 526d0c6faacd..995dd97af3e9 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv1a.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv1a.c
@@ -34,4 +34,5 @@ nv1a_devinit_oclass = &(struct nouveau_devinit_impl) {
 		.fini = nv04_devinit_fini,
 	},
 	.pll_set = nv04_devinit_pll_set,
+	.post = nvbios_init,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv20.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv20.c
index 04bc9732644c..915089fb46f7 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv20.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv20.c
@@ -71,4 +71,5 @@ nv20_devinit_oclass = &(struct nouveau_devinit_impl) {
 	},
 	.meminit = nv20_devinit_meminit,
 	.pll_set = nv04_devinit_pll_set,
+	.post = nvbios_init,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.c
index b46c62a1d5d8..968334d1dca4 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.c
@@ -26,6 +26,7 @@
 #include <subdev/bios/dcb.h>
 #include <subdev/bios/disp.h>
 #include <subdev/bios/init.h>
+#include <subdev/ibus.h>
 #include <subdev/vga.h>
 
 #include "nv50.h"
@@ -91,6 +92,7 @@ int
 nv50_devinit_init(struct nouveau_object *object)
 {
 	struct nouveau_bios *bios = nouveau_bios(object);
+	struct nouveau_ibus *ibus = nouveau_ibus(object);
 	struct nv50_devinit_priv *priv = (void *)object;
 	struct nvbios_outp info;
 	struct dcb_output outp;
@@ -105,6 +107,13 @@ nv50_devinit_init(struct nouveau_object *object)
 		}
 	}
 
+	/* some boards appear to require certain priv register timeouts
+	 * to be bumped before runing devinit scripts.  not a clue why
+	 * the vbios engineers didn't make the scripts just work...
+	 */
+	if (priv->base.post && ibus)
+		nv_ofuncs(ibus)->init(nv_object(ibus));
+
 	ret = nouveau_devinit_init(&priv->base);
 	if (ret)
 		return ret;
@@ -160,4 +169,5 @@ nv50_devinit_oclass = &(struct nouveau_devinit_impl) {
 	},
 	.pll_set = nv50_devinit_pll_set,
 	.disable = nv50_devinit_disable,
+	.post = nvbios_init,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.h b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.h
index 51d5076333ec..f412bb7f780e 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.h
@@ -18,4 +18,6 @@ int  nva3_devinit_pll_set(struct nouveau_devinit *, u32, u32);
 
 int  nvc0_devinit_pll_set(struct nouveau_devinit *, u32, u32);
 
+u64  gm107_devinit_disable(struct nouveau_devinit *);
+
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv84.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv84.c
index 787422505d87..a7c80ded77cd 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv84.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv84.c
@@ -60,4 +60,5 @@ nv84_devinit_oclass = &(struct nouveau_devinit_impl) {
 	},
 	.pll_set = nv50_devinit_pll_set,
 	.disable = nv84_devinit_disable,
+	.post = nvbios_init,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv98.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv98.c
index 2b0e963fc6f0..a773253a17f6 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv98.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv98.c
@@ -59,4 +59,5 @@ nv98_devinit_oclass = &(struct nouveau_devinit_impl) {
 	},
 	.pll_set = nv50_devinit_pll_set,
 	.disable = nv98_devinit_disable,
+	.post = nvbios_init,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nva3.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nva3.c
index 006cf348bda7..b9cd9e53f760 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nva3.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nva3.c
@@ -142,4 +142,5 @@ nva3_devinit_oclass = &(struct nouveau_devinit_impl) {
 	.pll_set = nva3_devinit_pll_set,
 	.disable = nva3_devinit_disable,
 	.mmio    = nva3_devinit_mmio,
+	.post = nvbios_init,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nvaf.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nvaf.c
index 4fc68d27eff3..3729846a8e5c 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nvaf.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nvaf.c
@@ -60,4 +60,5 @@ nvaf_devinit_oclass = &(struct nouveau_devinit_impl) {
 	},
 	.pll_set = nva3_devinit_pll_set,
 	.disable = nvaf_devinit_disable,
+	.post = nvbios_init,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nvc0.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nvc0.c
index 30c765747eea..80bd7f5eda3d 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nvc0.c
@@ -115,4 +115,5 @@ nvc0_devinit_oclass = &(struct nouveau_devinit_impl) {
 	},
 	.pll_set = nvc0_devinit_pll_set,
 	.disable = nvc0_devinit_disable,
+	.post = nvbios_init,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/priv.h b/drivers/gpu/drm/nouveau/core/subdev/devinit/priv.h
index f0e8683ad840..cbcd51852472 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/priv.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/priv.h
@@ -3,6 +3,7 @@
 
 #include <subdev/bios.h>
 #include <subdev/bios/pll.h>
+#include <subdev/bios/init.h>
 #include <subdev/clock/pll.h>
 #include <subdev/devinit.h>
 
@@ -12,6 +13,7 @@ struct nouveau_devinit_impl {
 	int  (*pll_set)(struct nouveau_devinit *, u32 type, u32 freq);
 	u64  (*disable)(struct nouveau_devinit *);
 	u32  (*mmio)(struct nouveau_devinit *, u32);
+	int  (*post)(struct nouveau_subdev *, bool);
 };
 
 #define nouveau_devinit_create(p,e,o,d)                                        \
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/base.c b/drivers/gpu/drm/nouveau/core/subdev/fb/base.c
index f009d8a39d9d..c866148c440f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/base.c
@@ -23,37 +23,30 @@
  */
 
 #include <subdev/bios.h>
-#include <subdev/bios/bit.h>
+#include <subdev/bios/M0203.h>
 
 #include "priv.h"
 
 int
 nouveau_fb_bios_memtype(struct nouveau_bios *bios)
 {
-	struct bit_entry M;
-	u8 ramcfg;
-
-	ramcfg = (nv_rd32(bios, 0x101000) & 0x0000003c) >> 2;
-	if (!bit_entry(bios, 'M', &M) && M.version == 2 && M.length >= 5) {
-		u16 table   = nv_ro16(bios, M.offset + 3);
-		u8  version = nv_ro08(bios, table + 0);
-		u8  header  = nv_ro08(bios, table + 1);
-		u8  record  = nv_ro08(bios, table + 2);
-		u8  entries = nv_ro08(bios, table + 3);
-		if (table && version == 0x10 && ramcfg < entries) {
-			u16 entry = table + header + (ramcfg * record);
-			switch (nv_ro08(bios, entry) & 0x0f) {
-			case 0: return NV_MEM_TYPE_DDR2;
-			case 1: return NV_MEM_TYPE_DDR3;
-			case 2: return NV_MEM_TYPE_GDDR3;
-			case 3: return NV_MEM_TYPE_GDDR5;
-			default:
-				break;
-			}
-
+	const u8 ramcfg = (nv_rd32(bios, 0x101000) & 0x0000003c) >> 2;
+	struct nvbios_M0203E M0203E;
+	u8 ver, hdr;
+
+	if (nvbios_M0203Em(bios, ramcfg, &ver, &hdr, &M0203E)) {
+		switch (M0203E.type) {
+		case M0203E_TYPE_DDR2 : return NV_MEM_TYPE_DDR2;
+		case M0203E_TYPE_DDR3 : return NV_MEM_TYPE_DDR3;
+		case M0203E_TYPE_GDDR3: return NV_MEM_TYPE_GDDR3;
+		case M0203E_TYPE_GDDR5: return NV_MEM_TYPE_GDDR5;
+		default:
+			nv_warn(bios, "M0203E type %02x\n", M0203E.type);
+			return NV_MEM_TYPE_UNKNOWN;
 		}
 	}
 
+	nv_warn(bios, "M0203E not matched!\n");
 	return NV_MEM_TYPE_UNKNOWN;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/gddr3.c b/drivers/gpu/drm/nouveau/core/subdev/fb/gddr3.c
new file mode 100644
index 000000000000..d85a25d027ee
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/gddr3.c
@@ -0,0 +1,117 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs <bskeggs@redhat.com>
+ * 	    Roy Spliet <rspliet@eclipso.eu>
+ */
+
+#include <subdev/bios.h>
+#include "priv.h"
+
+struct ramxlat {
+	int id;
+	u8 enc;
+};
+
+static inline int
+ramxlat(const struct ramxlat *xlat, int id)
+{
+	while (xlat->id >= 0) {
+		if (xlat->id == id)
+			return xlat->enc;
+		xlat++;
+	}
+	return -EINVAL;
+}
+
+static const struct ramxlat
+ramgddr3_cl_lo[] = {
+	{ 7, 7 }, { 8, 0 }, { 9, 1 }, { 10, 2 }, { 11, 3 },
+	/* the below are mentioned in some, but not all, gddr3 docs */
+	{ 12, 4 }, { 13, 5 }, { 14, 6 },
+	/* XXX: Per Samsung docs, are these used? They overlap with Qimonda */
+	/* { 4, 4 }, { 5, 5 }, { 6, 6 }, { 12, 8 }, { 13, 9 }, { 14, 10 },
+	 * { 15, 11 }, */
+	{ -1 }
+};
+
+static const struct ramxlat
+ramgddr3_cl_hi[] = {
+	{ 10, 2 }, { 11, 3 }, { 12, 4 }, { 13, 5 }, { 14, 6 }, { 15, 7 },
+	{ 16, 0 }, { 17, 1 },
+	{ -1 }
+};
+
+static const struct ramxlat
+ramgddr3_wr_lo[] = {
+	{ 5, 2 }, { 7, 4 }, { 8, 5 }, { 9, 6 }, { 10, 7 },
+	{ 11, 0 },
+	/* the below are mentioned in some, but not all, gddr3 docs */
+	{ 4, 1 }, { 6, 3 }, { 12, 1 }, { 13 , 2 },
+	{ -1 }
+};
+
+int
+nouveau_gddr3_calc(struct nouveau_ram *ram)
+{
+	int CL, WR, CWL, DLL = 0, ODT = 0, hi;
+
+	switch (ram->next->bios.timing_ver) {
+	case 0x10:
+		CWL = ram->next->bios.timing_10_CWL;
+		CL  = ram->next->bios.timing_10_CL;
+		WR  = ram->next->bios.timing_10_WR;
+		DLL = !ram->next->bios.ramcfg_10_DLLoff;
+		ODT = ram->next->bios.timing_10_ODT;
+		break;
+	case 0x20:
+		CWL = (ram->next->bios.timing[1] & 0x00000f80) >> 7;
+		CL  = (ram->next->bios.timing[1] & 0x0000001f) >> 0;
+		WR  = (ram->next->bios.timing[2] & 0x007f0000) >> 16;
+		/* XXX: Get these values from the VBIOS instead */
+		DLL = !(ram->mr[1] & 0x1);
+		ODT =  (ram->mr[1] & 0x004) >> 2 |
+		       (ram->mr[1] & 0x040) >> 5 |
+		       (ram->mr[1] & 0x200) >> 7;
+		break;
+	default:
+		return -ENOSYS;
+	}
+
+	hi = ram->mr[2] & 0x1;
+	CL  = ramxlat(hi ? ramgddr3_cl_hi : ramgddr3_cl_lo, CL);
+	WR  = ramxlat(ramgddr3_wr_lo, WR);
+	if (CL < 0 || CWL < 1 || CWL > 7 || WR < 0)
+		return -EINVAL;
+
+	ram->mr[0] &= ~0xf74;
+	ram->mr[0] |= (CWL & 0x07) << 9;
+	ram->mr[0] |= (CL & 0x07) << 4;
+	ram->mr[0] |= (CL & 0x08) >> 1;
+
+	ram->mr[1] &= ~0x3fc;
+	ram->mr[1] |= (ODT & 0x03) << 2;
+	ram->mr[1] |= (ODT & 0x03) << 8;
+	ram->mr[1] |= (WR  & 0x03) << 4;
+	ram->mr[1] |= (WR  & 0x04) << 5;
+	ram->mr[1] |= !DLL << 6;
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/gddr5.c b/drivers/gpu/drm/nouveau/core/subdev/fb/gddr5.c
index 66fe959b4f74..7fbbe05d5c60 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/gddr5.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/gddr5.c
@@ -40,7 +40,7 @@ nouveau_gddr5_calc(struct nouveau_ram *ram, bool nuts)
 	int WL, CL, WR, at[2], dt, ds;
 	int rq = ram->freq < 1000000; /* XXX */
 
-	switch (ram->ramcfg.version) {
+	switch (ram->next->bios.ramcfg_ver) {
 	case 0x11:
 		pd =  ram->next->bios.ramcfg_11_01_80;
 		lf =  ram->next->bios.ramcfg_11_01_40;
@@ -54,7 +54,7 @@ nouveau_gddr5_calc(struct nouveau_ram *ram, bool nuts)
 		return -ENOSYS;
 	}
 
-	switch (ram->timing.version) {
+	switch (ram->next->bios.timing_ver) {
 	case 0x20:
 		WL = (ram->next->bios.timing[1] & 0x00000f80) >> 7;
 		CL = (ram->next->bios.timing[1] & 0x0000001f);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/gk20a.c b/drivers/gpu/drm/nouveau/core/subdev/fb/gk20a.c
index a16024a74771..fde42e4d1b56 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/gk20a.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/gk20a.c
@@ -27,6 +27,20 @@ struct gk20a_fb_priv {
 };
 
 static int
+gk20a_fb_init(struct nouveau_object *object)
+{
+	struct gk20a_fb_priv *priv = (void *)object;
+	int ret;
+
+	ret = nouveau_fb_init(&priv->base);
+	if (ret)
+		return ret;
+
+	nv_mask(priv, 0x100c80, 0x00000001, 0x00000000); /* 128KiB lpg */
+	return 0;
+}
+
+static int
 gk20a_fb_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	     struct nouveau_oclass *oclass, void *data, u32 size,
 	     struct nouveau_object **pobject)
@@ -48,7 +62,7 @@ gk20a_fb_oclass = &(struct nouveau_fb_impl) {
 	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = gk20a_fb_ctor,
 		.dtor = _nouveau_fb_dtor,
-		.init = _nouveau_fb_init,
+		.init = gk20a_fb_init,
 		.fini = _nouveau_fb_fini,
 	},
 	.memtype = nvc0_fb_memtype_valid,
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nv20.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nv20.c
index f003c1b1893f..2209ade63339 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nv20.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nv20.c
@@ -45,7 +45,7 @@ nv20_fb_tile_comp(struct nouveau_fb *pfb, int i, u32 size, u32 flags,
 {
 	u32 tiles = DIV_ROUND_UP(size, 0x40);
 	u32 tags  = round_up(tiles / pfb->ram->parts, 0x40);
-	if (!nouveau_mm_head(&pfb->tags, 1, tags, tags, 1, &tile->tag)) {
+	if (!nouveau_mm_head(&pfb->tags, 0, 1, tags, tags, 1, &tile->tag)) {
 		if (!(flags & 2)) tile->zcomp = 0x00000000; /* Z16 */
 		else              tile->zcomp = 0x04000000; /* Z24S8 */
 		tile->zcomp |= tile->tag->offset;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nv25.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nv25.c
index f34f4223210b..e2a66c355c50 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nv25.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nv25.c
@@ -32,7 +32,7 @@ nv25_fb_tile_comp(struct nouveau_fb *pfb, int i, u32 size, u32 flags,
 {
 	u32 tiles = DIV_ROUND_UP(size, 0x40);
 	u32 tags  = round_up(tiles / pfb->ram->parts, 0x40);
-	if (!nouveau_mm_head(&pfb->tags, 1, tags, tags, 1, &tile->tag)) {
+	if (!nouveau_mm_head(&pfb->tags, 0, 1, tags, tags, 1, &tile->tag)) {
 		if (!(flags & 2)) tile->zcomp = 0x00100000; /* Z16 */
 		else              tile->zcomp = 0x00200000; /* Z24S8 */
 		tile->zcomp |= tile->tag->offset;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nv30.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nv30.c
index 69093f7151f0..cbec402ba5b9 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nv30.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nv30.c
@@ -51,7 +51,7 @@ nv30_fb_tile_comp(struct nouveau_fb *pfb, int i, u32 size, u32 flags,
 {
 	u32 tiles = DIV_ROUND_UP(size, 0x40);
 	u32 tags  = round_up(tiles / pfb->ram->parts, 0x40);
-	if (!nouveau_mm_head(&pfb->tags, 1, tags, tags, 1, &tile->tag)) {
+	if (!nouveau_mm_head(&pfb->tags, 0, 1, tags, tags, 1, &tile->tag)) {
 		if (flags & 2) tile->zcomp |= 0x01000000; /* Z16 */
 		else           tile->zcomp |= 0x02000000; /* Z24S8 */
 		tile->zcomp |= ((tile->tag->offset           ) >> 6);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nv35.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nv35.c
index 161b06e8fc3f..b2cf8c69fb2e 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nv35.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nv35.c
@@ -32,7 +32,7 @@ nv35_fb_tile_comp(struct nouveau_fb *pfb, int i, u32 size, u32 flags,
 {
 	u32 tiles = DIV_ROUND_UP(size, 0x40);
 	u32 tags  = round_up(tiles / pfb->ram->parts, 0x40);
-	if (!nouveau_mm_head(&pfb->tags, 1, tags, tags, 1, &tile->tag)) {
+	if (!nouveau_mm_head(&pfb->tags, 0, 1, tags, tags, 1, &tile->tag)) {
 		if (flags & 2) tile->zcomp |= 0x04000000; /* Z16 */
 		else           tile->zcomp |= 0x08000000; /* Z24S8 */
 		tile->zcomp |= ((tile->tag->offset           ) >> 6);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nv36.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nv36.c
index 2dd3d0aab6bb..b4cdae2a3b2f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nv36.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nv36.c
@@ -32,7 +32,7 @@ nv36_fb_tile_comp(struct nouveau_fb *pfb, int i, u32 size, u32 flags,
 {
 	u32 tiles = DIV_ROUND_UP(size, 0x40);
 	u32 tags  = round_up(tiles / pfb->ram->parts, 0x40);
-	if (!nouveau_mm_head(&pfb->tags, 1, tags, tags, 1, &tile->tag)) {
+	if (!nouveau_mm_head(&pfb->tags, 0, 1, tags, tags, 1, &tile->tag)) {
 		if (flags & 2) tile->zcomp |= 0x10000000; /* Z16 */
 		else           tile->zcomp |= 0x20000000; /* Z24S8 */
 		tile->zcomp |= ((tile->tag->offset           ) >> 6);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nv40.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nv40.c
index 95a115ab0c86..52814258c212 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nv40.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nv40.c
@@ -33,7 +33,7 @@ nv40_fb_tile_comp(struct nouveau_fb *pfb, int i, u32 size, u32 flags,
 	u32 tiles = DIV_ROUND_UP(size, 0x80);
 	u32 tags  = round_up(tiles / pfb->ram->parts, 0x100);
 	if ( (flags & 2) &&
-	    !nouveau_mm_head(&pfb->tags, 1, tags, tags, 1, &tile->tag)) {
+	    !nouveau_mm_head(&pfb->tags, 0, 1, tags, tags, 1, &tile->tag)) {
 		tile->zcomp  = 0x28000000; /* Z24S8_SPLIT_GRAD */
 		tile->zcomp |= ((tile->tag->offset           ) >> 8);
 		tile->zcomp |= ((tile->tag->offset + tags - 1) >> 8) << 13;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/priv.h b/drivers/gpu/drm/nouveau/core/subdev/fb/priv.h
index 82273f832e42..283863f7aa9b 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/priv.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/priv.h
@@ -35,7 +35,9 @@ extern struct nouveau_oclass nve0_ram_oclass;
 extern struct nouveau_oclass gk20a_ram_oclass;
 extern struct nouveau_oclass gm107_ram_oclass;
 
+int nouveau_sddr2_calc(struct nouveau_ram *ram);
 int nouveau_sddr3_calc(struct nouveau_ram *ram);
+int nouveau_gddr3_calc(struct nouveau_ram *ram);
 int nouveau_gddr5_calc(struct nouveau_ram *ram, bool nuts);
 
 #define nouveau_fb_create(p,e,c,d)                                             \
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramfuc.h b/drivers/gpu/drm/nouveau/core/subdev/fb/ramfuc.h
index 2af9cfd2c60f..0ac7256443bb 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramfuc.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramfuc.h
@@ -12,16 +12,32 @@ struct ramfuc {
 struct ramfuc_reg {
 	int sequence;
 	bool force;
-	u32 addr[2];
+	u32 addr;
+	u32 stride; /* in bytes */
+	u32 mask;
 	u32 data;
 };
 
 static inline struct ramfuc_reg
+ramfuc_stride(u32 addr, u32 stride, u32 mask)
+{
+	return (struct ramfuc_reg) {
+		.sequence = 0,
+		.addr = addr,
+		.stride = stride,
+		.mask = mask,
+		.data = 0xdeadbeef,
+	};
+}
+
+static inline struct ramfuc_reg
 ramfuc_reg2(u32 addr1, u32 addr2)
 {
 	return (struct ramfuc_reg) {
 		.sequence = 0,
-		.addr = { addr1, addr2 },
+		.addr = addr1,
+		.stride = addr2 - addr1,
+		.mask = 0x3,
 		.data = 0xdeadbeef,
 	};
 }
@@ -29,7 +45,13 @@ ramfuc_reg2(u32 addr1, u32 addr2)
 static noinline struct ramfuc_reg
 ramfuc_reg(u32 addr)
 {
-	return ramfuc_reg2(addr, addr);
+	return (struct ramfuc_reg) {
+		.sequence = 0,
+		.addr = addr,
+		.stride = 0,
+		.mask = 0x1,
+		.data = 0xdeadbeef,
+	};
 }
 
 static inline int
@@ -62,18 +84,25 @@ static inline u32
 ramfuc_rd32(struct ramfuc *ram, struct ramfuc_reg *reg)
 {
 	if (reg->sequence != ram->sequence)
-		reg->data = nv_rd32(ram->pfb, reg->addr[0]);
+		reg->data = nv_rd32(ram->pfb, reg->addr);
 	return reg->data;
 }
 
 static inline void
 ramfuc_wr32(struct ramfuc *ram, struct ramfuc_reg *reg, u32 data)
 {
+	unsigned int mask, off = 0;
+
 	reg->sequence = ram->sequence;
 	reg->data = data;
-	if (reg->addr[0] != reg->addr[1])
-		nouveau_memx_wr32(ram->memx, reg->addr[1], reg->data);
-	nouveau_memx_wr32(ram->memx, reg->addr[0], reg->data);
+
+	for (mask = reg->mask; mask > 0; mask = (mask & ~1) >> 1) {
+		if (mask & 1) {
+			nouveau_memx_wr32(ram->memx, reg->addr+off, reg->data);
+		}
+
+		off += reg->stride;
+	}
 }
 
 static inline void
@@ -105,14 +134,51 @@ ramfuc_nsec(struct ramfuc *ram, u32 nsec)
 	nouveau_memx_nsec(ram->memx, nsec);
 }
 
-#define ram_init(s,p)       ramfuc_init(&(s)->base, (p))
-#define ram_exec(s,e)       ramfuc_exec(&(s)->base, (e))
-#define ram_have(s,r)       ((s)->r_##r.addr[0] != 0x000000)
-#define ram_rd32(s,r)       ramfuc_rd32(&(s)->base, &(s)->r_##r)
-#define ram_wr32(s,r,d)     ramfuc_wr32(&(s)->base, &(s)->r_##r, (d))
-#define ram_nuke(s,r)       ramfuc_nuke(&(s)->base, &(s)->r_##r)
-#define ram_mask(s,r,m,d)   ramfuc_mask(&(s)->base, &(s)->r_##r, (m), (d))
-#define ram_wait(s,r,m,d,n) ramfuc_wait(&(s)->base, (r), (m), (d), (n))
-#define ram_nsec(s,n)       ramfuc_nsec(&(s)->base, (n))
+static inline void
+ramfuc_wait_vblank(struct ramfuc *ram)
+{
+	nouveau_memx_wait_vblank(ram->memx);
+}
+
+static inline void
+ramfuc_train(struct ramfuc *ram)
+{
+	nouveau_memx_train(ram->memx);
+}
+
+static inline int
+ramfuc_train_result(struct nouveau_fb *pfb, u32 *result, u32 rsize)
+{
+	struct nouveau_pwr *ppwr = nouveau_pwr(pfb);
+
+	return nouveau_memx_train_result(ppwr, result, rsize);
+}
+
+static inline void
+ramfuc_block(struct ramfuc *ram)
+{
+	nouveau_memx_block(ram->memx);
+}
+
+static inline void
+ramfuc_unblock(struct ramfuc *ram)
+{
+	nouveau_memx_unblock(ram->memx);
+}
+
+#define ram_init(s,p)        ramfuc_init(&(s)->base, (p))
+#define ram_exec(s,e)        ramfuc_exec(&(s)->base, (e))
+#define ram_have(s,r)        ((s)->r_##r.addr != 0x000000)
+#define ram_rd32(s,r)        ramfuc_rd32(&(s)->base, &(s)->r_##r)
+#define ram_wr32(s,r,d)      ramfuc_wr32(&(s)->base, &(s)->r_##r, (d))
+#define ram_nuke(s,r)        ramfuc_nuke(&(s)->base, &(s)->r_##r)
+#define ram_mask(s,r,m,d)    ramfuc_mask(&(s)->base, &(s)->r_##r, (m), (d))
+#define ram_wait(s,r,m,d,n)  ramfuc_wait(&(s)->base, (r), (m), (d), (n))
+#define ram_nsec(s,n)        ramfuc_nsec(&(s)->base, (n))
+#define ram_wait_vblank(s)   ramfuc_wait_vblank(&(s)->base)
+#define ram_train(s)         ramfuc_train(&(s)->base)
+#define ram_train_result(s,r,l) ramfuc_train_result((s), (r), (l))
+#define ram_block(s)         ramfuc_block(&(s)->base)
+#define ram_unblock(s)       ramfuc_unblock(&(s)->base)
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv04.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv04.c
index e781080d3327..1972268d1410 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv04.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv04.c
@@ -22,22 +22,7 @@
  * Authors: Ben Skeggs
  */
 
-#define NV04_PFB_BOOT_0						0x00100000
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT			0x00000003
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT_32MB			0x00000000
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT_4MB			0x00000001
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT_8MB			0x00000002
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT_16MB			0x00000003
-#	define NV04_PFB_BOOT_0_RAM_WIDTH_128			0x00000004
-#	define NV04_PFB_BOOT_0_RAM_TYPE				0x00000028
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_8MBIT		0x00000000
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_16MBIT		0x00000008
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_16MBIT_4BANK	0x00000010
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_16MBIT		0x00000018
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_64MBIT		0x00000020
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_64MBITX16		0x00000028
-#	define NV04_PFB_BOOT_0_UMA_ENABLE			0x00000100
-#	define NV04_PFB_BOOT_0_UMA_SIZE				0x0000f000
+#include <subdev/fb/regsnv04.h>
 
 #include "priv.h"
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv50.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv50.c
index e5d12c24cc43..64a983c96625 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv50.c
@@ -280,7 +280,7 @@ nv50_ram_get(struct nouveau_fb *pfb, u64 size, u32 align, u32 ncmin,
 		if (align == 16) {
 			int n = (max >> 4) * comp;
 
-			ret = nouveau_mm_head(tags, 1, n, n, 1, &mem->tag);
+			ret = nouveau_mm_head(tags, 0, 1, n, n, 1, &mem->tag);
 			if (ret)
 				mem->tag = NULL;
 		}
@@ -296,9 +296,9 @@ nv50_ram_get(struct nouveau_fb *pfb, u64 size, u32 align, u32 ncmin,
 	type = nv50_fb_memtype[type];
 	do {
 		if (back)
-			ret = nouveau_mm_tail(heap, type, max, min, align, &r);
+			ret = nouveau_mm_tail(heap, 0, type, max, min, align, &r);
 		else
-			ret = nouveau_mm_head(heap, type, max, min, align, &r);
+			ret = nouveau_mm_head(heap, 0, type, max, min, align, &r);
 		if (ret) {
 			mutex_unlock(&pfb->base.mutex);
 			pfb->ram->put(pfb, &mem);
@@ -319,27 +319,22 @@ nv50_ram_get(struct nouveau_fb *pfb, u64 size, u32 align, u32 ncmin,
 static u32
 nv50_fb_vram_rblock(struct nouveau_fb *pfb, struct nouveau_ram *ram)
 {
-	int i, parts, colbits, rowbitsa, rowbitsb, banks;
+	int colbits, rowbitsa, rowbitsb, banks;
 	u64 rowsize, predicted;
-	u32 r0, r4, rt, ru, rblock_size;
+	u32 r0, r4, rt, rblock_size;
 
 	r0 = nv_rd32(pfb, 0x100200);
 	r4 = nv_rd32(pfb, 0x100204);
 	rt = nv_rd32(pfb, 0x100250);
-	ru = nv_rd32(pfb, 0x001540);
-	nv_debug(pfb, "memcfg 0x%08x 0x%08x 0x%08x 0x%08x\n", r0, r4, rt, ru);
-
-	for (i = 0, parts = 0; i < 8; i++) {
-		if (ru & (0x00010000 << i))
-			parts++;
-	}
+	nv_debug(pfb, "memcfg 0x%08x 0x%08x 0x%08x 0x%08x\n", r0, r4, rt,
+			nv_rd32(pfb, 0x001540));
 
 	colbits  =  (r4 & 0x0000f000) >> 12;
 	rowbitsa = ((r4 & 0x000f0000) >> 16) + 8;
 	rowbitsb = ((r4 & 0x00f00000) >> 20) + 8;
 	banks    = 1 << (((r4 & 0x03000000) >> 24) + 2);
 
-	rowsize = parts * banks * (1 << colbits) * 8;
+	rowsize = ram->parts * banks * (1 << colbits) * 8;
 	predicted = rowsize << rowbitsa;
 	if (r0 & 0x00000004)
 		predicted += rowsize << rowbitsb;
@@ -376,6 +371,9 @@ nv50_ram_create_(struct nouveau_object *parent, struct nouveau_object *engine,
 	ram->size = nv_rd32(pfb, 0x10020c);
 	ram->size = (ram->size & 0xffffff00) | ((ram->size & 0x000000ff) << 32);
 
+	ram->part_mask = (nv_rd32(pfb, 0x001540) & 0x00ff0000) >> 16;
+	ram->parts = hweight8(ram->part_mask);
+
 	switch (nv_rd32(pfb, 0x100714) & 0x00000007) {
 	case 0: ram->type = NV_MEM_TYPE_DDR1; break;
 	case 1:
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnva3.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnva3.c
index 8076fb195dd5..3b38a538845d 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnva3.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnva3.c
@@ -20,79 +20,512 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  * Authors: Ben Skeggs
+ * 	    Roy Spliet <rspliet@eclipso.eu>
  */
 
 #include <subdev/bios.h>
 #include <subdev/bios/bit.h>
 #include <subdev/bios/pll.h>
 #include <subdev/bios/rammap.h>
+#include <subdev/bios/M0205.h>
 #include <subdev/bios/timing.h>
 
 #include <subdev/clock/nva3.h>
 #include <subdev/clock/pll.h>
 
+#include <subdev/gpio.h>
+
+#include <subdev/timer.h>
+
+#include <engine/fifo.h>
+
 #include <core/option.h>
 
 #include "ramfuc.h"
 
 #include "nv50.h"
 
+/* XXX: Remove when memx gains GPIO support */
+extern int nv50_gpio_location(int line, u32 *reg, u32 *shift);
+
 struct nva3_ramfuc {
 	struct ramfuc base;
+	struct ramfuc_reg r_0x001610;
+	struct ramfuc_reg r_0x001700;
+	struct ramfuc_reg r_0x002504;
 	struct ramfuc_reg r_0x004000;
 	struct ramfuc_reg r_0x004004;
 	struct ramfuc_reg r_0x004018;
 	struct ramfuc_reg r_0x004128;
 	struct ramfuc_reg r_0x004168;
+	struct ramfuc_reg r_0x100080;
 	struct ramfuc_reg r_0x100200;
 	struct ramfuc_reg r_0x100210;
 	struct ramfuc_reg r_0x100220[9];
+	struct ramfuc_reg r_0x100264;
 	struct ramfuc_reg r_0x1002d0;
 	struct ramfuc_reg r_0x1002d4;
 	struct ramfuc_reg r_0x1002dc;
 	struct ramfuc_reg r_0x10053c;
 	struct ramfuc_reg r_0x1005a0;
 	struct ramfuc_reg r_0x1005a4;
+	struct ramfuc_reg r_0x100700;
 	struct ramfuc_reg r_0x100714;
 	struct ramfuc_reg r_0x100718;
 	struct ramfuc_reg r_0x10071c;
+	struct ramfuc_reg r_0x100720;
 	struct ramfuc_reg r_0x100760;
 	struct ramfuc_reg r_0x1007a0;
 	struct ramfuc_reg r_0x1007e0;
+	struct ramfuc_reg r_0x100da0;
 	struct ramfuc_reg r_0x10f804;
 	struct ramfuc_reg r_0x1110e0;
 	struct ramfuc_reg r_0x111100;
 	struct ramfuc_reg r_0x111104;
+	struct ramfuc_reg r_0x1111e0;
+	struct ramfuc_reg r_0x111400;
 	struct ramfuc_reg r_0x611200;
 	struct ramfuc_reg r_mr[4];
+	struct ramfuc_reg r_gpioFBVREF;
+};
+
+struct nva3_ltrain {
+	enum {
+		NVA3_TRAIN_UNKNOWN,
+		NVA3_TRAIN_UNSUPPORTED,
+		NVA3_TRAIN_ONCE,
+		NVA3_TRAIN_EXEC,
+		NVA3_TRAIN_DONE
+	} state;
+	u32 r_100720;
+	u32 r_1111e0;
+	u32 r_111400;
+	struct nouveau_mem *mem;
 };
 
 struct nva3_ram {
 	struct nouveau_ram base;
 	struct nva3_ramfuc fuc;
+	struct nva3_ltrain ltrain;
 };
 
+void
+nva3_link_train_calc(u32 *vals, struct nva3_ltrain *train)
+{
+	int i, lo, hi;
+	u8 median[8], bins[4] = {0, 0, 0, 0}, bin = 0, qty = 0;
+
+	for (i = 0; i < 8; i++) {
+		for (lo = 0; lo < 0x40; lo++) {
+			if (!(vals[lo] & 0x80000000))
+				continue;
+			if (vals[lo] & (0x101 << i))
+				break;
+		}
+
+		if (lo == 0x40)
+			return;
+
+		for (hi = lo + 1; hi < 0x40; hi++) {
+			if (!(vals[lo] & 0x80000000))
+				continue;
+			if (!(vals[hi] & (0x101 << i))) {
+				hi--;
+				break;
+			}
+		}
+
+		median[i] = ((hi - lo) >> 1) + lo;
+		bins[(median[i] & 0xf0) >> 4]++;
+		median[i] += 0x30;
+	}
+
+	/* Find the best value for 0x1111e0 */
+	for (i = 0; i < 4; i++) {
+		if (bins[i] > qty) {
+			bin = i + 3;
+			qty = bins[i];
+		}
+	}
+
+	train->r_100720 = 0;
+	for (i = 0; i < 8; i++) {
+		median[i] = max(median[i], (u8) (bin << 4));
+		median[i] = min(median[i], (u8) ((bin << 4) | 0xf));
+
+		train->r_100720 |= ((median[i] & 0x0f) << (i << 2));
+	}
+
+	train->r_1111e0 = 0x02000000 | (bin * 0x101);
+	train->r_111400 = 0x0;
+}
+
+/*
+ * Link training for (at least) DDR3
+ */
+int
+nva3_link_train(struct nouveau_fb *pfb)
+{
+	struct nouveau_bios *bios = nouveau_bios(pfb);
+	struct nva3_ram *ram = (void *)pfb->ram;
+	struct nouveau_clock *clk = nouveau_clock(pfb);
+	struct nva3_ltrain *train = &ram->ltrain;
+	struct nouveau_device *device = nv_device(pfb);
+	struct nva3_ramfuc *fuc = &ram->fuc;
+	u32 *result, r1700;
+	int ret, i;
+	struct nvbios_M0205T M0205T = { 0 };
+	u8 ver, hdr, cnt, len, snr, ssz;
+	unsigned int clk_current;
+	unsigned long flags;
+	unsigned long *f = &flags;
+
+	if (nouveau_boolopt(device->cfgopt, "NvMemExec", true) != true)
+		return -ENOSYS;
+
+	/* XXX: Multiple partitions? */
+	result = kmalloc(64 * sizeof(u32), GFP_KERNEL);
+	if (!result)
+		return -ENOMEM;
+
+	train->state = NVA3_TRAIN_EXEC;
+
+	/* Clock speeds for training and back */
+	nvbios_M0205Tp(bios, &ver, &hdr, &cnt, &len, &snr, &ssz, &M0205T);
+	if (M0205T.freq == 0)
+		return -ENOENT;
+
+	clk_current = clk->read(clk, nv_clk_src_mem);
+
+	ret = nva3_clock_pre(clk, f);
+	if (ret)
+		goto out;
+
+	/* First: clock up/down */
+	ret = ram->base.calc(pfb, (u32) M0205T.freq * 1000);
+	if (ret)
+		goto out;
+
+	/* Do this *after* calc, eliminates write in script */
+	nv_wr32(pfb, 0x111400, 0x00000000);
+	/* XXX: Magic writes that improve train reliability? */
+	nv_mask(pfb, 0x100674, 0x0000ffff, 0x00000000);
+	nv_mask(pfb, 0x1005e4, 0x0000ffff, 0x00000000);
+	nv_mask(pfb, 0x100b0c, 0x000000ff, 0x00000000);
+	nv_wr32(pfb, 0x100c04, 0x00000400);
+
+	/* Now the training script */
+	r1700 = ram_rd32(fuc, 0x001700);
+
+	ram_mask(fuc, 0x100200, 0x00000800, 0x00000000);
+	ram_wr32(fuc, 0x611200, 0x3300);
+	ram_wait_vblank(fuc);
+	ram_wait(fuc, 0x611200, 0x00000003, 0x00000000, 500000);
+	ram_mask(fuc, 0x001610, 0x00000083, 0x00000003);
+	ram_mask(fuc, 0x100080, 0x00000020, 0x00000000);
+	ram_mask(fuc, 0x10f804, 0x80000000, 0x00000000);
+	ram_wr32(fuc, 0x001700, 0x00000000);
+
+	ram_train(fuc);
+
+	/* Reset */
+	ram_mask(fuc, 0x10f804, 0x80000000, 0x80000000);
+	ram_wr32(fuc, 0x10053c, 0x0);
+	ram_wr32(fuc, 0x100720, train->r_100720);
+	ram_wr32(fuc, 0x1111e0, train->r_1111e0);
+	ram_wr32(fuc, 0x111400, train->r_111400);
+	ram_nuke(fuc, 0x100080);
+	ram_mask(fuc, 0x100080, 0x00000020, 0x00000020);
+	ram_nsec(fuc, 1000);
+
+	ram_wr32(fuc, 0x001700, r1700);
+	ram_mask(fuc, 0x001610, 0x00000083, 0x00000080);
+	ram_wr32(fuc, 0x611200, 0x3330);
+	ram_mask(fuc, 0x100200, 0x00000800, 0x00000800);
+
+	ram_exec(fuc, true);
+
+	ram->base.calc(pfb, clk_current);
+	ram_exec(fuc, true);
+
+	/* Post-processing, avoids flicker */
+	nv_mask(pfb, 0x616308, 0x10, 0x10);
+	nv_mask(pfb, 0x616b08, 0x10, 0x10);
+
+	nva3_clock_post(clk, f);
+
+	ram_train_result(pfb, result, 64);
+	for (i = 0; i < 64; i++)
+		nv_debug(pfb, "Train: %08x", result[i]);
+	nva3_link_train_calc(result, train);
+
+	nv_debug(pfb, "Train: %08x %08x %08x", train->r_100720,
+			train->r_1111e0, train->r_111400);
+
+	kfree(result);
+
+	train->state = NVA3_TRAIN_DONE;
+
+	return ret;
+
+out:
+	if(ret == -EBUSY)
+		f = NULL;
+
+	train->state = NVA3_TRAIN_UNSUPPORTED;
+
+	nva3_clock_post(clk, f);
+	return ret;
+}
+
+int
+nva3_link_train_init(struct nouveau_fb *pfb)
+{
+	static const u32 pattern[16] = {
+		0xaaaaaaaa, 0xcccccccc, 0xdddddddd, 0xeeeeeeee,
+		0x00000000, 0x11111111, 0x44444444, 0xdddddddd,
+		0x33333333, 0x55555555, 0x77777777, 0x66666666,
+		0x99999999, 0x88888888, 0xeeeeeeee, 0xbbbbbbbb,
+	};
+	struct nouveau_bios *bios = nouveau_bios(pfb);
+	struct nva3_ram *ram = (void *)pfb->ram;
+	struct nva3_ltrain *train = &ram->ltrain;
+	struct nouveau_mem *mem;
+	struct nvbios_M0205E M0205E;
+	u8 ver, hdr, cnt, len;
+	u32 r001700;
+	int ret, i = 0;
+
+	train->state = NVA3_TRAIN_UNSUPPORTED;
+
+	/* We support type "5"
+	 * XXX: training pattern table appears to be unused for this routine */
+	if (!nvbios_M0205Ep(bios, i, &ver, &hdr, &cnt, &len, &M0205E))
+		return -ENOENT;
+
+	if (M0205E.type != 5)
+		return 0;
+
+	train->state = NVA3_TRAIN_ONCE;
+
+	ret = pfb->ram->get(pfb, 0x8000, 0x10000, 0, 0x800, &ram->ltrain.mem);
+	if (ret)
+		return ret;
+
+	mem = ram->ltrain.mem;
+
+	nv_wr32(pfb, 0x100538, 0x10000000 | (mem->offset >> 16));
+	nv_wr32(pfb, 0x1005a8, 0x0000ffff);
+	nv_mask(pfb, 0x10f800, 0x00000001, 0x00000001);
+
+	for (i = 0; i < 0x30; i++) {
+		nv_wr32(pfb, 0x10f8c0, (i << 8) | i);
+		nv_wr32(pfb, 0x10f900, pattern[i % 16]);
+	}
+
+	for (i = 0; i < 0x30; i++) {
+		nv_wr32(pfb, 0x10f8e0, (i << 8) | i);
+		nv_wr32(pfb, 0x10f920, pattern[i % 16]);
+	}
+
+	/* And upload the pattern */
+	r001700 = nv_rd32(pfb, 0x1700);
+	nv_wr32(pfb, 0x1700, mem->offset >> 16);
+	for (i = 0; i < 16; i++)
+		nv_wr32(pfb, 0x700000 + (i << 2), pattern[i]);
+	for (i = 0; i < 16; i++)
+		nv_wr32(pfb, 0x700100 + (i << 2), pattern[i]);
+	nv_wr32(pfb, 0x1700, r001700);
+
+	train->r_100720 = nv_rd32(pfb, 0x100720);
+	train->r_1111e0 = nv_rd32(pfb, 0x1111e0);
+	train->r_111400 = nv_rd32(pfb, 0x111400);
+
+	return 0;
+}
+
+void
+nva3_link_train_fini(struct nouveau_fb *pfb)
+{
+	struct nva3_ram *ram = (void *)pfb->ram;
+
+	if (ram->ltrain.mem)
+		pfb->ram->put(pfb, &ram->ltrain.mem);
+}
+
+/*
+ * RAM reclocking
+ */
+#define T(t) cfg->timing_10_##t
+static int
+nva3_ram_timing_calc(struct nouveau_fb *pfb, u32 *timing)
+{
+	struct nva3_ram *ram = (void *)pfb->ram;
+	struct nvbios_ramcfg *cfg = &ram->base.target.bios;
+	int tUNK_base, tUNK_40_0, prevCL;
+	u32 cur2, cur3, cur7, cur8;
+
+	cur2 = nv_rd32(pfb, 0x100228);
+	cur3 = nv_rd32(pfb, 0x10022c);
+	cur7 = nv_rd32(pfb, 0x10023c);
+	cur8 = nv_rd32(pfb, 0x100240);
+
+
+	switch ((!T(CWL)) * ram->base.type) {
+	case NV_MEM_TYPE_DDR2:
+		T(CWL) = T(CL) - 1;
+		break;
+	case NV_MEM_TYPE_GDDR3:
+		T(CWL) = ((cur2 & 0xff000000) >> 24) + 1;
+		break;
+	}
+
+	prevCL = (cur3 & 0x000000ff) + 1;
+	tUNK_base = ((cur7 & 0x00ff0000) >> 16) - prevCL;
+
+	timing[0] = (T(RP) << 24 | T(RAS) << 16 | T(RFC) << 8 | T(RC));
+	timing[1] = (T(WR) + 1 + T(CWL)) << 24 |
+		    max_t(u8,T(18), 1) << 16 |
+		    (T(WTR) + 1 + T(CWL)) << 8 |
+		    (5 + T(CL) - T(CWL));
+	timing[2] = (T(CWL) - 1) << 24 |
+		    (T(RRD) << 16) |
+		    (T(RCDWR) << 8) |
+		    T(RCDRD);
+	timing[3] = (cur3 & 0x00ff0000) |
+		    (0x30 + T(CL)) << 24 |
+		    (0xb + T(CL)) << 8 |
+		    (T(CL) - 1);
+	timing[4] = T(20) << 24 |
+		    T(21) << 16 |
+		    T(13) << 8 |
+		    T(13);
+	timing[5] = T(RFC) << 24 |
+		    max_t(u8,T(RCDRD), T(RCDWR)) << 16 |
+		    max_t(u8, (T(CWL) + 6), (T(CL) + 2)) << 8 |
+		    T(RP);
+	timing[6] = (0x5a + T(CL)) << 16 |
+		    max_t(u8, 1, (6 - T(CL) + T(CWL))) << 8 |
+		    (0x50 + T(CL) - T(CWL));
+	timing[7] = (cur7 & 0xff000000) |
+		    ((tUNK_base + T(CL)) << 16) |
+		    0x202;
+	timing[8] = cur8 & 0xffffff00;
+
+	switch (ram->base.type) {
+	case NV_MEM_TYPE_DDR2:
+	case NV_MEM_TYPE_GDDR3:
+		tUNK_40_0 = prevCL - (cur8 & 0xff);
+		if (tUNK_40_0 > 0)
+			timing[8] |= T(CL);
+		break;
+	default:
+		break;
+	}
+
+	nv_debug(pfb, "Entry: 220: %08x %08x %08x %08x\n",
+			timing[0], timing[1], timing[2], timing[3]);
+	nv_debug(pfb, "  230: %08x %08x %08x %08x\n",
+			timing[4], timing[5], timing[6], timing[7]);
+	nv_debug(pfb, "  240: %08x\n", timing[8]);
+	return 0;
+}
+#undef T
+
+static void
+nouveau_sddr2_dll_reset(struct nva3_ramfuc *fuc)
+{
+	ram_mask(fuc, mr[0], 0x100, 0x100);
+	ram_nsec(fuc, 1000);
+	ram_mask(fuc, mr[0], 0x100, 0x000);
+	ram_nsec(fuc, 1000);
+}
+
+static void
+nouveau_sddr3_dll_disable(struct nva3_ramfuc *fuc, u32 *mr)
+{
+	u32 mr1_old = ram_rd32(fuc, mr[1]);
+
+	if (!(mr1_old & 0x1)) {
+		ram_wr32(fuc, 0x1002d4, 0x00000001);
+		ram_wr32(fuc, mr[1], mr[1]);
+		ram_nsec(fuc, 1000);
+	}
+}
+
+static void
+nouveau_gddr3_dll_disable(struct nva3_ramfuc *fuc, u32 *mr)
+{
+	u32 mr1_old = ram_rd32(fuc, mr[1]);
+
+	if (!(mr1_old & 0x40)) {
+		ram_wr32(fuc, mr[1], mr[1]);
+		ram_nsec(fuc, 1000);
+	}
+}
+
+static void
+nva3_ram_lock_pll(struct nva3_ramfuc *fuc, struct nva3_clock_info *mclk)
+{
+	ram_wr32(fuc, 0x004004, mclk->pll);
+	ram_mask(fuc, 0x004000, 0x00000001, 0x00000001);
+	ram_mask(fuc, 0x004000, 0x00000010, 0x00000000);
+	ram_wait(fuc, 0x004000, 0x00020000, 0x00020000, 64000);
+	ram_mask(fuc, 0x004000, 0x00000010, 0x00000010);
+}
+
+static void
+nva3_ram_fbvref(struct nva3_ramfuc *fuc, u32 val)
+{
+	struct nouveau_gpio *gpio = nouveau_gpio(fuc->base.pfb);
+	struct dcb_gpio_func func;
+	u32 reg, sh, gpio_val;
+	int ret;
+
+	if (gpio->get(gpio, 0, 0x2e, DCB_GPIO_UNUSED) != val) {
+		ret = gpio->find(gpio, 0, 0x2e, DCB_GPIO_UNUSED, &func);
+		if (ret)
+			return;
+
+		nv50_gpio_location(func.line, &reg, &sh);
+		gpio_val = ram_rd32(fuc, gpioFBVREF);
+		if (gpio_val & (8 << sh))
+			val = !val;
+
+		ram_mask(fuc, gpioFBVREF, (0x3 << sh), ((val | 0x2) << sh));
+		ram_nsec(fuc, 20000);
+	}
+}
+
 static int
 nva3_ram_calc(struct nouveau_fb *pfb, u32 freq)
 {
 	struct nouveau_bios *bios = nouveau_bios(pfb);
 	struct nva3_ram *ram = (void *)pfb->ram;
 	struct nva3_ramfuc *fuc = &ram->fuc;
+	struct nva3_ltrain *train = &ram->ltrain;
 	struct nva3_clock_info mclk;
-	u8  ver, cnt, len, strap;
+	struct nouveau_ram_data *next;
+	u8  ver, hdr, cnt, len, strap;
 	u32 data;
-	struct {
-		u32 data;
-		u8  size;
-	} rammap, ramcfg, timing;
-	u32 r004018, r100760, ctrl;
+	u32 r004018, r100760, r100da0, r111100, ctrl;
 	u32 unk714, unk718, unk71c;
-	int ret;
+	int ret, i;
+	u32 timing[9];
+	bool pll2pll;
+
+	next = &ram->base.target;
+	next->freq = freq;
+	ram->base.next = next;
+
+	if (ram->ltrain.state == NVA3_TRAIN_ONCE)
+		nva3_link_train(pfb);
 
 	/* lookup memory config data relevant to the target frequency */
-	rammap.data = nvbios_rammapEm(bios, freq / 1000, &ver, &rammap.size,
-				     &cnt, &ramcfg.size);
-	if (!rammap.data || ver != 0x10 || rammap.size < 0x0e) {
+	i = 0;
+	data = nvbios_rammapEm(bios, freq / 1000, &ver, &hdr, &cnt, &len,
+				      &next->bios);
+	if (!data || ver != 0x10 || hdr < 0x05) {
 		nv_error(pfb, "invalid/missing rammap entry\n");
 		return -EINVAL;
 	}
@@ -104,78 +537,123 @@ nva3_ram_calc(struct nouveau_fb *pfb, u32 freq)
 		return -EINVAL;
 	}
 
-	ramcfg.data = rammap.data + rammap.size + (strap * ramcfg.size);
-	if (!ramcfg.data || ver != 0x10 || ramcfg.size < 0x0e) {
+	data = nvbios_rammapSp(bios, data, ver, hdr, cnt, len, strap,
+			       &ver, &hdr, &next->bios);
+	if (!data || ver != 0x10 || hdr < 0x09) {
 		nv_error(pfb, "invalid/missing ramcfg entry\n");
 		return -EINVAL;
 	}
 
 	/* lookup memory timings, if bios says they're present */
-	strap = nv_ro08(bios, ramcfg.data + 0x01);
-	if (strap != 0xff) {
-		timing.data = nvbios_timingEe(bios, strap, &ver, &timing.size,
-					     &cnt, &len);
-		if (!timing.data || ver != 0x10 || timing.size < 0x19) {
+	if (next->bios.ramcfg_timing != 0xff) {
+		data = nvbios_timingEp(bios, next->bios.ramcfg_timing,
+				       &ver, &hdr, &cnt, &len,
+				       &next->bios);
+		if (!data || ver != 0x10 || hdr < 0x17) {
 			nv_error(pfb, "invalid/missing timing entry\n");
 			return -EINVAL;
 		}
-	} else {
-		timing.data = 0;
 	}
 
-	ret = nva3_clock_info(nouveau_clock(pfb), 0x12, 0x4000, freq, &mclk);
+	ret = nva3_pll_info(nouveau_clock(pfb), 0x12, 0x4000, freq, &mclk);
 	if (ret < 0) {
 		nv_error(pfb, "failed mclk calculation\n");
 		return ret;
 	}
 
+	nva3_ram_timing_calc(pfb, timing);
+
 	ret = ram_init(fuc, pfb);
 	if (ret)
 		return ret;
 
+	/* Determine ram-specific MR values */
+	ram->base.mr[0] = ram_rd32(fuc, mr[0]);
+	ram->base.mr[1] = ram_rd32(fuc, mr[1]);
+	ram->base.mr[2] = ram_rd32(fuc, mr[2]);
+
+	switch (ram->base.type) {
+	case NV_MEM_TYPE_DDR2:
+		ret = nouveau_sddr2_calc(&ram->base);
+		break;
+	case NV_MEM_TYPE_DDR3:
+		ret = nouveau_sddr3_calc(&ram->base);
+		break;
+	case NV_MEM_TYPE_GDDR3:
+		ret = nouveau_gddr3_calc(&ram->base);
+		break;
+	default:
+		ret = -ENOSYS;
+		break;
+	}
+
+	if (ret)
+		return ret;
+
 	/* XXX: where the fuck does 750MHz come from? */
 	if (freq <= 750000) {
 		r004018 = 0x10000000;
 		r100760 = 0x22222222;
+		r100da0 = 0x00000010;
 	} else {
 		r004018 = 0x00000000;
 		r100760 = 0x00000000;
+		r100da0 = 0x00000000;
 	}
 
+	if (!next->bios.ramcfg_10_DLLoff)
+		r004018 |= 0x00004000;
+
+	/* pll2pll requires to switch to a safe clock first */
 	ctrl = ram_rd32(fuc, 0x004000);
-	if (ctrl & 0x00000008) {
-		if (mclk.pll) {
-			ram_mask(fuc, 0x004128, 0x00000101, 0x00000101);
-			ram_wr32(fuc, 0x004004, mclk.pll);
-			ram_wr32(fuc, 0x004000, (ctrl |= 0x00000001));
-			ram_wr32(fuc, 0x004000, (ctrl &= 0xffffffef));
-			ram_wait(fuc, 0x004000, 0x00020000, 0x00020000, 64000);
-			ram_wr32(fuc, 0x004000, (ctrl |= 0x00000010));
-			ram_wr32(fuc, 0x004018, 0x00005000 | r004018);
-			ram_wr32(fuc, 0x004000, (ctrl |= 0x00000004));
-		}
-	} else {
-		u32 ssel = 0x00000101;
-		if (mclk.clk)
-			ssel |= mclk.clk;
-		else
-			ssel |= 0x00080000; /* 324MHz, shouldn't matter... */
-		ram_mask(fuc, 0x004168, 0x003f3141, ctrl);
-	}
+	pll2pll = (!(ctrl & 0x00000008)) && mclk.pll;
 
-	if ( (nv_ro08(bios, ramcfg.data + 0x02) & 0x10)) {
+	/* Pre, NVIDIA does this outside the script */
+	if (next->bios.ramcfg_10_02_10) {
 		ram_mask(fuc, 0x111104, 0x00000600, 0x00000000);
 	} else {
 		ram_mask(fuc, 0x111100, 0x40000000, 0x40000000);
 		ram_mask(fuc, 0x111104, 0x00000180, 0x00000000);
 	}
+	/* Always disable this bit during reclock */
+	ram_mask(fuc, 0x100200, 0x00000800, 0x00000000);
 
-	if (!(nv_ro08(bios, rammap.data + 0x04) & 0x02))
-		ram_mask(fuc, 0x100200, 0x00000800, 0x00000000);
-	ram_wr32(fuc, 0x611200, 0x00003300);
-	if (!(nv_ro08(bios, ramcfg.data + 0x02) & 0x10))
-		ram_wr32(fuc, 0x111100, 0x4c020000); /*XXX*/
+	/* If switching from non-pll to pll, lock before disabling FB */
+	if (mclk.pll && !pll2pll) {
+		ram_mask(fuc, 0x004128, 0x003f3141, mclk.clk | 0x00000101);
+		nva3_ram_lock_pll(fuc, &mclk);
+	}
+
+	/* Start with disabling some CRTCs and PFIFO? */
+	ram_wait_vblank(fuc);
+	ram_wr32(fuc, 0x611200, 0x3300);
+	ram_mask(fuc, 0x002504, 0x1, 0x1);
+	ram_nsec(fuc, 10000);
+	ram_wait(fuc, 0x002504, 0x10, 0x10, 20000); /* XXX: or longer? */
+	ram_block(fuc);
+	ram_nsec(fuc, 2000);
 
+	if (!next->bios.ramcfg_10_02_10) {
+		if (ram->base.type == NV_MEM_TYPE_GDDR3)
+			ram_mask(fuc, 0x111100, 0x04020000, 0x00020000);
+		else
+			ram_mask(fuc, 0x111100, 0x04020000, 0x04020000);
+	}
+
+	/* If we're disabling the DLL, do it now */
+	switch (next->bios.ramcfg_10_DLLoff * ram->base.type) {
+	case NV_MEM_TYPE_DDR3:
+		nouveau_sddr3_dll_disable(fuc, ram->base.mr);
+		break;
+	case NV_MEM_TYPE_GDDR3:
+		nouveau_gddr3_dll_disable(fuc, ram->base.mr);
+		break;
+	}
+
+	if (fuc->r_gpioFBVREF.addr && next->bios.timing_10_ODT)
+		nva3_ram_fbvref(fuc, 0);
+
+	/* Brace RAM for impact */
 	ram_wr32(fuc, 0x1002d4, 0x00000001);
 	ram_wr32(fuc, 0x1002d0, 0x00000001);
 	ram_wr32(fuc, 0x1002d0, 0x00000001);
@@ -183,38 +661,57 @@ nva3_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	ram_wr32(fuc, 0x1002dc, 0x00000001);
 	ram_nsec(fuc, 2000);
 
-	ctrl = ram_rd32(fuc, 0x004000);
-	if (!(ctrl & 0x00000008) && mclk.pll) {
-		ram_wr32(fuc, 0x004000, (ctrl |=  0x00000008));
+	if (nv_device(pfb)->chipset == 0xa3 && freq <= 500000)
+		ram_mask(fuc, 0x100700, 0x00000006, 0x00000006);
+
+	/* Fiddle with clocks */
+	/* There's 4 scenario's
+	 * pll->pll: first switch to a 324MHz clock, set up new PLL, switch
+	 * clk->pll: Set up new PLL, switch
+	 * pll->clk: Set up clock, switch
+	 * clk->clk: Overwrite ctrl and other bits, switch */
+
+	/* Switch to regular clock - 324MHz */
+	if (pll2pll) {
+		ram_mask(fuc, 0x004000, 0x00000004, 0x00000004);
+		ram_mask(fuc, 0x004168, 0x003f3141, 0x00083101);
+		ram_mask(fuc, 0x004000, 0x00000008, 0x00000008);
 		ram_mask(fuc, 0x1110e0, 0x00088000, 0x00088000);
 		ram_wr32(fuc, 0x004018, 0x00001000);
-		ram_wr32(fuc, 0x004000, (ctrl &= ~0x00000001));
-		ram_wr32(fuc, 0x004004, mclk.pll);
-		ram_wr32(fuc, 0x004000, (ctrl |=  0x00000001));
-		udelay(64);
-		ram_wr32(fuc, 0x004018, 0x00005000 | r004018);
-		udelay(20);
-	} else
-	if (!mclk.pll) {
-		ram_mask(fuc, 0x004168, 0x003f3040, mclk.clk);
-		ram_wr32(fuc, 0x004000, (ctrl |= 0x00000008));
+		nva3_ram_lock_pll(fuc, &mclk);
+	}
+
+	if (mclk.pll) {
+		ram_mask(fuc, 0x004000, 0x00000105, 0x00000105);
+		ram_wr32(fuc, 0x004018, 0x00001000 | r004018);
+		ram_wr32(fuc, 0x100da0, r100da0);
+	} else {
+		ram_mask(fuc, 0x004168, 0x003f3141, mclk.clk | 0x00000101);
+		ram_mask(fuc, 0x004000, 0x00000108, 0x00000008);
 		ram_mask(fuc, 0x1110e0, 0x00088000, 0x00088000);
-		ram_wr32(fuc, 0x004018, 0x0000d000 | r004018);
-	}
-
-	if ( (nv_ro08(bios, rammap.data + 0x04) & 0x08)) {
-		u32 unk5a0 = (nv_ro16(bios, ramcfg.data + 0x05) << 8) |
-			      nv_ro08(bios, ramcfg.data + 0x05);
-		u32 unk5a4 = (nv_ro16(bios, ramcfg.data + 0x07));
-		u32 unk804 = (nv_ro08(bios, ramcfg.data + 0x09) & 0xf0) << 16 |
-			     (nv_ro08(bios, ramcfg.data + 0x03) & 0x0f) << 16 |
-			     (nv_ro08(bios, ramcfg.data + 0x09) & 0x0f) |
-			     0x80000000;
-		ram_wr32(fuc, 0x1005a0, unk5a0);
-		ram_wr32(fuc, 0x1005a4, unk5a4);
-		ram_wr32(fuc, 0x10f804, unk804);
+		ram_wr32(fuc, 0x004018, 0x00009000 | r004018);
+		ram_wr32(fuc, 0x100da0, r100da0);
+	}
+	ram_nsec(fuc, 20000);
+
+	if (next->bios.rammap_10_04_08) {
+		ram_wr32(fuc, 0x1005a0, next->bios.ramcfg_10_06 << 16 |
+					next->bios.ramcfg_10_05 << 8 |
+					next->bios.ramcfg_10_05);
+		ram_wr32(fuc, 0x1005a4, next->bios.ramcfg_10_08 << 8 |
+					next->bios.ramcfg_10_07);
+		ram_wr32(fuc, 0x10f804, next->bios.ramcfg_10_09_f0 << 20 |
+					next->bios.ramcfg_10_03_0f << 16 |
+					next->bios.ramcfg_10_09_0f |
+					0x80000000);
 		ram_mask(fuc, 0x10053c, 0x00001000, 0x00000000);
 	} else {
+		if (train->state == NVA3_TRAIN_DONE) {
+			ram_wr32(fuc, 0x100080, 0x1020);
+			ram_mask(fuc, 0x111400, 0xffffffff, train->r_111400);
+			ram_mask(fuc, 0x1111e0, 0xffffffff, train->r_1111e0);
+			ram_mask(fuc, 0x100720, 0xffffffff, train->r_100720);
+		}
 		ram_mask(fuc, 0x10053c, 0x00001000, 0x00001000);
 		ram_mask(fuc, 0x10f804, 0x80000000, 0x00000000);
 		ram_mask(fuc, 0x100760, 0x22222222, r100760);
@@ -222,69 +719,134 @@ nva3_ram_calc(struct nouveau_fb *pfb, u32 freq)
 		ram_mask(fuc, 0x1007e0, 0x22222222, r100760);
 	}
 
+	if (nv_device(pfb)->chipset == 0xa3 && freq > 500000) {
+		ram_mask(fuc, 0x100700, 0x00000006, 0x00000000);
+	}
+
+	/* Final switch */
 	if (mclk.pll) {
 		ram_mask(fuc, 0x1110e0, 0x00088000, 0x00011000);
-		ram_wr32(fuc, 0x004000, (ctrl &= ~0x00000008));
+		ram_mask(fuc, 0x004000, 0x00000008, 0x00000000);
 	}
 
-	/*XXX: LEAVE */
 	ram_wr32(fuc, 0x1002dc, 0x00000000);
 	ram_wr32(fuc, 0x1002d4, 0x00000001);
 	ram_wr32(fuc, 0x100210, 0x80000000);
-	ram_nsec(fuc, 1000);
-	ram_nsec(fuc, 1000);
+	ram_nsec(fuc, 2000);
 
-	ram_mask(fuc, mr[2], 0x00000000, 0x00000000);
-	ram_nsec(fuc, 1000);
-	ram_nuke(fuc, mr[0]);
-	ram_mask(fuc, mr[0], 0x00000000, 0x00000000);
-	ram_nsec(fuc, 1000);
+	/* Set RAM MR parameters and timings */
+	for (i = 2; i >= 0; i--) {
+		if (ram_rd32(fuc, mr[i]) != ram->base.mr[i]) {
+			ram_wr32(fuc, mr[i], ram->base.mr[i]);
+			ram_nsec(fuc, 1000);
+		}
+	}
 
-	ram_mask(fuc, 0x100220[3], 0x00000000, 0x00000000);
-	ram_mask(fuc, 0x100220[1], 0x00000000, 0x00000000);
-	ram_mask(fuc, 0x100220[6], 0x00000000, 0x00000000);
-	ram_mask(fuc, 0x100220[7], 0x00000000, 0x00000000);
-	ram_mask(fuc, 0x100220[2], 0x00000000, 0x00000000);
-	ram_mask(fuc, 0x100220[4], 0x00000000, 0x00000000);
-	ram_mask(fuc, 0x100220[5], 0x00000000, 0x00000000);
-	ram_mask(fuc, 0x100220[0], 0x00000000, 0x00000000);
-	ram_mask(fuc, 0x100220[8], 0x00000000, 0x00000000);
-
-	data = (nv_ro08(bios, ramcfg.data + 0x02) & 0x08) ? 0x00000000 : 0x00001000;
-	ram_mask(fuc, 0x100200, 0x00001000, data);
-
-	unk714 = ram_rd32(fuc, 0x100714) & ~0xf0000010;
-	unk718 = ram_rd32(fuc, 0x100718) & ~0x00000100;
-	unk71c = ram_rd32(fuc, 0x10071c) & ~0x00000100;
-	if ( (nv_ro08(bios, ramcfg.data + 0x02) & 0x20))
-		unk714 |= 0xf0000000;
-	if (!(nv_ro08(bios, ramcfg.data + 0x02) & 0x04))
-		unk714 |= 0x00000010;
-	ram_wr32(fuc, 0x100714, unk714);
+	ram_wr32(fuc, 0x100220[3], timing[3]);
+	ram_wr32(fuc, 0x100220[1], timing[1]);
+	ram_wr32(fuc, 0x100220[6], timing[6]);
+	ram_wr32(fuc, 0x100220[7], timing[7]);
+	ram_wr32(fuc, 0x100220[2], timing[2]);
+	ram_wr32(fuc, 0x100220[4], timing[4]);
+	ram_wr32(fuc, 0x100220[5], timing[5]);
+	ram_wr32(fuc, 0x100220[0], timing[0]);
+	ram_wr32(fuc, 0x100220[8], timing[8]);
 
-	if (nv_ro08(bios, ramcfg.data + 0x02) & 0x01)
-		unk71c |= 0x00000100;
-	ram_wr32(fuc, 0x10071c, unk71c);
+	/* Misc */
+	ram_mask(fuc, 0x100200, 0x00001000, !next->bios.ramcfg_10_02_08 << 12);
+
+	/* XXX: A lot of "chipset"/"ram type" specific stuff...? */
+	unk714  = ram_rd32(fuc, 0x100714) & ~0xf0000130;
+	unk718  = ram_rd32(fuc, 0x100718) & ~0x00000100;
+	unk71c  = ram_rd32(fuc, 0x10071c) & ~0x00000100;
+	r111100 = ram_rd32(fuc, 0x111100) & ~0x3a800000;
+
+	if (next->bios.ramcfg_10_02_04) {
+		switch (ram->base.type) {
+		case NV_MEM_TYPE_DDR3:
+			if (nv_device(pfb)->chipset != 0xa8)
+				r111100 |= 0x00000004;
+			/* no break */
+		case NV_MEM_TYPE_DDR2:
+			r111100 |= 0x08000000;
+			break;
+		default:
+			break;
+		}
+	} else {
+		switch (ram->base.type) {
+		case NV_MEM_TYPE_DDR2:
+			r111100 |= 0x1a800000;
+			unk714  |= 0x00000010;
+			break;
+		case NV_MEM_TYPE_DDR3:
+			if (nv_device(pfb)->chipset == 0xa8) {
+				r111100 |=  0x08000000;
+			} else {
+				r111100 &= ~0x00000004;
+				r111100 |=  0x12800000;
+			}
+			unk714  |= 0x00000010;
+			break;
+		case NV_MEM_TYPE_GDDR3:
+			r111100 |= 0x30000000;
+			unk714  |= 0x00000020;
+			break;
+		default:
+			break;
+		}
+	}
+
+	unk714 |= (next->bios.ramcfg_10_04_01) << 8;
 
-	if (nv_ro08(bios, ramcfg.data + 0x02) & 0x02)
+	if (next->bios.ramcfg_10_02_20)
+		unk714 |= 0xf0000000;
+	if (next->bios.ramcfg_10_02_02)
 		unk718 |= 0x00000100;
-	ram_wr32(fuc, 0x100718, unk718);
+	if (next->bios.ramcfg_10_02_01)
+		unk71c |= 0x00000100;
+	if (next->bios.timing_10_24 != 0xff) {
+		unk718 &= ~0xf0000000;
+		unk718 |= next->bios.timing_10_24 << 28;
+	}
+	if (next->bios.ramcfg_10_02_10)
+		r111100 &= ~0x04020000;
 
-	if (nv_ro08(bios, ramcfg.data + 0x02) & 0x10)
-		ram_wr32(fuc, 0x111100, 0x48000000); /*XXX*/
+	ram_mask(fuc, 0x100714, 0xffffffff, unk714);
+	ram_mask(fuc, 0x10071c, 0xffffffff, unk71c);
+	ram_mask(fuc, 0x100718, 0xffffffff, unk718);
+	ram_mask(fuc, 0x111100, 0xffffffff, r111100);
 
-	ram_mask(fuc, mr[0], 0x100, 0x100);
-	ram_nsec(fuc, 1000);
-	ram_mask(fuc, mr[0], 0x100, 0x000);
-	ram_nsec(fuc, 1000);
+	if (fuc->r_gpioFBVREF.addr && !next->bios.timing_10_ODT)
+		nva3_ram_fbvref(fuc, 1);
 
-	ram_nsec(fuc, 2000);
-	ram_nsec(fuc, 12000);
+	/* Reset DLL */
+	if (!next->bios.ramcfg_10_DLLoff)
+		nouveau_sddr2_dll_reset(fuc);
 
-	ram_wr32(fuc, 0x611200, 0x00003330);
-	if ( (nv_ro08(bios, rammap.data + 0x04) & 0x02))
+	if (ram->base.type == NV_MEM_TYPE_GDDR3) {
+		ram_nsec(fuc, 31000);
+	} else {
+		ram_nsec(fuc, 14000);
+	}
+
+	if (ram->base.type == NV_MEM_TYPE_DDR3) {
+		ram_wr32(fuc, 0x100264, 0x1);
+		ram_nsec(fuc, 2000);
+	}
+
+	ram_nuke(fuc, 0x100700);
+	ram_mask(fuc, 0x100700, 0x01000000, 0x01000000);
+	ram_mask(fuc, 0x100700, 0x01000000, 0x00000000);
+
+	/* Re-enable FB */
+	ram_unblock(fuc);
+	ram_wr32(fuc, 0x611200, 0x3330);
+
+	/* Post fiddlings */
+	if (next->bios.rammap_10_04_02)
 		ram_mask(fuc, 0x100200, 0x00000800, 0x00000800);
-	if ( (nv_ro08(bios, ramcfg.data + 0x02) & 0x10)) {
+	if (next->bios.ramcfg_10_02_10) {
 		ram_mask(fuc, 0x111104, 0x00000180, 0x00000180);
 		ram_mask(fuc, 0x111100, 0x40000000, 0x00000000);
 	} else {
@@ -309,7 +871,22 @@ nva3_ram_prog(struct nouveau_fb *pfb)
 	struct nouveau_device *device = nv_device(pfb);
 	struct nva3_ram *ram = (void *)pfb->ram;
 	struct nva3_ramfuc *fuc = &ram->fuc;
-	ram_exec(fuc, nouveau_boolopt(device->cfgopt, "NvMemExec", true));
+	bool exec = nouveau_boolopt(device->cfgopt, "NvMemExec", true);
+
+	if (exec) {
+		nv_mask(pfb, 0x001534, 0x2, 0x2);
+
+		ram_exec(fuc, true);
+
+		/* Post-processing, avoids flicker */
+		nv_mask(pfb, 0x002504, 0x1, 0x0);
+		nv_mask(pfb, 0x001534, 0x2, 0x0);
+
+		nv_mask(pfb, 0x616308, 0x10, 0x10);
+		nv_mask(pfb, 0x616b08, 0x10, 0x10);
+	} else {
+		ram_exec(fuc, false);
+	}
 	return 0;
 }
 
@@ -326,38 +903,24 @@ nva3_ram_init(struct nouveau_object *object)
 {
 	struct nouveau_fb *pfb = (void *)object->parent;
 	struct nva3_ram   *ram = (void *)object;
-	int ret, i;
+	int ret;
 
 	ret = nouveau_ram_init(&ram->base);
 	if (ret)
 		return ret;
 
-	/* prepare for ddr link training, and load training patterns */
-	switch (ram->base.type) {
-	case NV_MEM_TYPE_DDR3: {
-		if (nv_device(pfb)->chipset == 0xa8) {
-			static const u32 pattern[16] = {
-				0xaaaaaaaa, 0xcccccccc, 0xdddddddd, 0xeeeeeeee,
-				0x00000000, 0x11111111, 0x44444444, 0xdddddddd,
-				0x33333333, 0x55555555, 0x77777777, 0x66666666,
-				0x99999999, 0x88888888, 0xeeeeeeee, 0xbbbbbbbb,
-			};
-
-			nv_wr32(pfb, 0x100538, 0x10001ff6); /*XXX*/
-			nv_wr32(pfb, 0x1005a8, 0x0000ffff);
-			nv_mask(pfb, 0x10f800, 0x00000001, 0x00000001);
-			for (i = 0; i < 0x30; i++) {
-				nv_wr32(pfb, 0x10f8c0, (i << 8) | i);
-				nv_wr32(pfb, 0x10f8e0, (i << 8) | i);
-				nv_wr32(pfb, 0x10f900, pattern[i % 16]);
-				nv_wr32(pfb, 0x10f920, pattern[i % 16]);
-			}
-		}
-	}
-		break;
-	default:
-		break;
-	}
+	nva3_link_train_init(pfb);
+
+	return 0;
+}
+
+static int
+nva3_ram_fini(struct nouveau_object *object, bool suspend)
+{
+	struct nouveau_fb *pfb = (void *)object->parent;
+
+	if (!suspend)
+		nva3_link_train_fini(pfb);
 
 	return 0;
 }
@@ -367,8 +930,12 @@ nva3_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	      struct nouveau_oclass *oclass, void *data, u32 datasize,
 	      struct nouveau_object **pobject)
 {
+	struct nouveau_fb *pfb = nouveau_fb(parent);
+	struct nouveau_gpio *gpio = nouveau_gpio(pfb);
+	struct dcb_gpio_func func;
 	struct nva3_ram *ram;
 	int ret, i;
+	u32 reg, shift;
 
 	ret = nv50_ram_create(parent, engine, oclass, &ram);
 	*pobject = nv_object(ram);
@@ -376,7 +943,9 @@ nva3_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		return ret;
 
 	switch (ram->base.type) {
+	case NV_MEM_TYPE_DDR2:
 	case NV_MEM_TYPE_DDR3:
+	case NV_MEM_TYPE_GDDR3:
 		ram->base.calc = nva3_ram_calc;
 		ram->base.prog = nva3_ram_prog;
 		ram->base.tidy = nva3_ram_tidy;
@@ -386,31 +955,41 @@ nva3_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		return 0;
 	}
 
+	ram->fuc.r_0x001610 = ramfuc_reg(0x001610);
+	ram->fuc.r_0x001700 = ramfuc_reg(0x001700);
+	ram->fuc.r_0x002504 = ramfuc_reg(0x002504);
 	ram->fuc.r_0x004000 = ramfuc_reg(0x004000);
 	ram->fuc.r_0x004004 = ramfuc_reg(0x004004);
 	ram->fuc.r_0x004018 = ramfuc_reg(0x004018);
 	ram->fuc.r_0x004128 = ramfuc_reg(0x004128);
 	ram->fuc.r_0x004168 = ramfuc_reg(0x004168);
+	ram->fuc.r_0x100080 = ramfuc_reg(0x100080);
 	ram->fuc.r_0x100200 = ramfuc_reg(0x100200);
 	ram->fuc.r_0x100210 = ramfuc_reg(0x100210);
 	for (i = 0; i < 9; i++)
 		ram->fuc.r_0x100220[i] = ramfuc_reg(0x100220 + (i * 4));
+	ram->fuc.r_0x100264 = ramfuc_reg(0x100264);
 	ram->fuc.r_0x1002d0 = ramfuc_reg(0x1002d0);
 	ram->fuc.r_0x1002d4 = ramfuc_reg(0x1002d4);
 	ram->fuc.r_0x1002dc = ramfuc_reg(0x1002dc);
 	ram->fuc.r_0x10053c = ramfuc_reg(0x10053c);
 	ram->fuc.r_0x1005a0 = ramfuc_reg(0x1005a0);
 	ram->fuc.r_0x1005a4 = ramfuc_reg(0x1005a4);
+	ram->fuc.r_0x100700 = ramfuc_reg(0x100700);
 	ram->fuc.r_0x100714 = ramfuc_reg(0x100714);
 	ram->fuc.r_0x100718 = ramfuc_reg(0x100718);
 	ram->fuc.r_0x10071c = ramfuc_reg(0x10071c);
-	ram->fuc.r_0x100760 = ramfuc_reg(0x100760);
-	ram->fuc.r_0x1007a0 = ramfuc_reg(0x1007a0);
-	ram->fuc.r_0x1007e0 = ramfuc_reg(0x1007e0);
+	ram->fuc.r_0x100720 = ramfuc_reg(0x100720);
+	ram->fuc.r_0x100760 = ramfuc_stride(0x100760, 4, ram->base.part_mask);
+	ram->fuc.r_0x1007a0 = ramfuc_stride(0x1007a0, 4, ram->base.part_mask);
+	ram->fuc.r_0x1007e0 = ramfuc_stride(0x1007e0, 4, ram->base.part_mask);
+	ram->fuc.r_0x100da0 = ramfuc_stride(0x100da0, 4, ram->base.part_mask);
 	ram->fuc.r_0x10f804 = ramfuc_reg(0x10f804);
-	ram->fuc.r_0x1110e0 = ramfuc_reg(0x1110e0);
+	ram->fuc.r_0x1110e0 = ramfuc_stride(0x1110e0, 4, ram->base.part_mask);
 	ram->fuc.r_0x111100 = ramfuc_reg(0x111100);
 	ram->fuc.r_0x111104 = ramfuc_reg(0x111104);
+	ram->fuc.r_0x1111e0 = ramfuc_reg(0x1111e0);
+	ram->fuc.r_0x111400 = ramfuc_reg(0x111400);
 	ram->fuc.r_0x611200 = ramfuc_reg(0x611200);
 
 	if (ram->base.ranks > 1) {
@@ -425,6 +1004,12 @@ nva3_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		ram->fuc.r_mr[3] = ramfuc_reg(0x1002e4);
 	}
 
+	ret = gpio->find(gpio, 0, 0x2e, DCB_GPIO_UNUSED, &func);
+	if (ret == 0) {
+		nv50_gpio_location(func.line, &reg, &shift);
+		ram->fuc.r_gpioFBVREF = ramfuc_reg(reg);
+	}
+
 	return 0;
 }
 
@@ -434,6 +1019,6 @@ nva3_ram_oclass = {
 		.ctor = nva3_ram_ctor,
 		.dtor = _nouveau_ram_dtor,
 		.init = nva3_ram_init,
-		.fini = _nouveau_ram_fini,
+		.fini = nva3_ram_fini,
 	},
 };
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvaa.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvaa.c
index 00f2ca7e44a5..033a8e999497 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvaa.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvaa.c
@@ -24,34 +24,71 @@
 
 #include "nv50.h"
 
+struct nvaa_ram_priv {
+	struct nouveau_ram base;
+	u64 poller_base;
+};
+
 static int
 nvaa_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	      struct nouveau_oclass *oclass, void *data, u32 datasize,
 	      struct nouveau_object **pobject)
 {
-	const u32 rsvd_head = ( 256 * 1024) >> 12; /* vga memory */
-	const u32 rsvd_tail = (1024 * 1024) >> 12; /* vbios etc */
+	u32 rsvd_head = ( 256 * 1024); /* vga memory */
+	u32 rsvd_tail = (1024 * 1024); /* vbios etc */
 	struct nouveau_fb *pfb = nouveau_fb(parent);
-	struct nouveau_ram *ram;
+	struct nvaa_ram_priv *priv;
 	int ret;
 
-	ret = nouveau_ram_create(parent, engine, oclass, &ram);
-	*pobject = nv_object(ram);
+	ret = nouveau_ram_create(parent, engine, oclass, &priv);
+	*pobject = nv_object(priv);
 	if (ret)
 		return ret;
 
-	ram->size = nv_rd32(pfb, 0x10020c);
-	ram->size = (ram->size & 0xffffff00) | ((ram->size & 0x000000ff) << 32);
+	priv->base.type   = NV_MEM_TYPE_STOLEN;
+	priv->base.stolen = (u64)nv_rd32(pfb, 0x100e10) << 12;
+	priv->base.size   = (u64)nv_rd32(pfb, 0x100e14) << 12;
 
-	ret = nouveau_mm_init(&pfb->vram, rsvd_head, (ram->size >> 12) -
-			      (rsvd_head + rsvd_tail), 1);
+	rsvd_tail += 0x1000;
+	priv->poller_base = priv->base.size - rsvd_tail;
+
+	ret = nouveau_mm_init(&pfb->vram, rsvd_head >> 12,
+			      (priv->base.size  - (rsvd_head + rsvd_tail)) >> 12,
+			      1);
 	if (ret)
 		return ret;
 
-	ram->type   = NV_MEM_TYPE_STOLEN;
-	ram->stolen = (u64)nv_rd32(pfb, 0x100e10) << 12;
-	ram->get = nv50_ram_get;
-	ram->put = nv50_ram_put;
+	priv->base.get = nv50_ram_get;
+	priv->base.put = nv50_ram_put;
+	return 0;
+}
+
+static int
+nvaa_ram_init(struct nouveau_object *object)
+{
+	struct nouveau_fb *pfb = nouveau_fb(object);
+	struct nvaa_ram_priv *priv = (void *)object;
+	int ret;
+	u64 dniso, hostnb, flush;
+
+	ret = nouveau_ram_init(&priv->base);
+	if (ret)
+		return ret;
+
+	dniso  = ((priv->base.size - (priv->poller_base + 0x00)) >> 5) - 1;
+	hostnb = ((priv->base.size - (priv->poller_base + 0x20)) >> 5) - 1;
+	flush  = ((priv->base.size - (priv->poller_base + 0x40)) >> 5) - 1;
+
+	/* Enable NISO poller for various clients and set their associated
+	 * read address, only for MCP77/78 and MCP79/7A. (fd#25701)
+	 */
+	nv_wr32(pfb, 0x100c18, dniso);
+	nv_mask(pfb, 0x100c14, 0x00000000, 0x00000001);
+	nv_wr32(pfb, 0x100c1c, hostnb);
+	nv_mask(pfb, 0x100c14, 0x00000000, 0x00000002);
+	nv_wr32(pfb, 0x100c24, flush);
+	nv_mask(pfb, 0x100c14, 0x00000000, 0x00010000);
+
 	return 0;
 }
 
@@ -60,7 +97,7 @@ nvaa_ram_oclass = {
 	.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nvaa_ram_ctor,
 		.dtor = _nouveau_ram_dtor,
-		.init = _nouveau_ram_init,
+		.init = nvaa_ram_init,
 		.fini = _nouveau_ram_fini,
 	},
 };
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvc0.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvc0.c
index 2b284b192763..735cb9580abe 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvc0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvc0.c
@@ -133,6 +133,7 @@ nvc0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	struct nouveau_bios *bios = nouveau_bios(pfb);
 	struct nvc0_ram *ram = (void *)pfb->ram;
 	struct nvc0_ramfuc *fuc = &ram->fuc;
+	struct nvbios_ramcfg cfg;
 	u8  ver, cnt, len, strap;
 	struct {
 		u32 data;
@@ -145,7 +146,7 @@ nvc0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 
 	/* lookup memory config data relevant to the target frequency */
 	rammap.data = nvbios_rammapEm(bios, freq / 1000, &ver, &rammap.size,
-				     &cnt, &ramcfg.size);
+				     &cnt, &ramcfg.size, &cfg);
 	if (!rammap.data || ver != 0x10 || rammap.size < 0x0e) {
 		nv_error(pfb, "invalid/missing rammap entry\n");
 		return -EINVAL;
@@ -483,9 +484,9 @@ nvc0_ram_get(struct nouveau_fb *pfb, u64 size, u32 align, u32 ncmin,
 
 	do {
 		if (back)
-			ret = nouveau_mm_tail(mm, 1, size, ncmin, align, &r);
+			ret = nouveau_mm_tail(mm, 0, 1, size, ncmin, align, &r);
 		else
-			ret = nouveau_mm_head(mm, 1, size, ncmin, align, &r);
+			ret = nouveau_mm_head(mm, 0, 1, size, ncmin, align, &r);
 		if (ret) {
 			mutex_unlock(&pfb->base.mutex);
 			pfb->ram->put(pfb, &mem);
@@ -562,7 +563,7 @@ nvc0_ram_create_(struct nouveau_object *parent, struct nouveau_object *engine,
 		offset = (0x0200000000ULL >> 12) + (bsize << 8);
 		length = (ram->size >> 12) - ((bsize * parts) << 8) - rsvd_tail;
 
-		ret = nouveau_mm_init(&pfb->vram, offset, length, 0);
+		ret = nouveau_mm_init(&pfb->vram, offset, length, 1);
 		if (ret)
 			nouveau_mm_fini(&pfb->vram);
 	}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnve0.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnve0.c
index c5b46e302319..6bae474abb44 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnve0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnve0.c
@@ -29,6 +29,8 @@
 #include <subdev/bios/init.h>
 #include <subdev/bios/rammap.h>
 #include <subdev/bios/timing.h>
+#include <subdev/bios/M0205.h>
+#include <subdev/bios/M0209.h>
 
 #include <subdev/clock.h>
 #include <subdev/clock/pll.h>
@@ -41,14 +43,6 @@
 
 #include "ramfuc.h"
 
-/* binary driver only executes this path if the condition (a) is true
- * for any configuration (combination of rammap+ramcfg+timing) that
- * can be reached on a given card.  for now, we will execute the branch
- * unconditionally in the hope that a "false everywhere" in the bios
- * tables doesn't actually mean "don't touch this".
- */
-#define NOTE00(a) 1
-
 struct nve0_ramfuc {
 	struct ramfuc base;
 
@@ -134,10 +128,12 @@ struct nve0_ram {
 	struct nouveau_ram base;
 	struct nve0_ramfuc fuc;
 
+	struct list_head cfg;
 	u32 parts;
 	u32 pmask;
 	u32 pnuts;
 
+	struct nvbios_ramcfg diff;
 	int from;
 	int mode;
 	int N1, fN1, M1, P1;
@@ -241,7 +237,7 @@ nve0_ram_nuts(struct nve0_ram *ram, struct ramfuc_reg *reg,
 {
 	struct nve0_fb_priv *priv = (void *)nouveau_fb(ram);
 	struct ramfuc *fuc = &ram->fuc.base;
-	u32 addr = 0x110000 + (reg->addr[0] & 0xfff);
+	u32 addr = 0x110000 + (reg->addr & 0xfff);
 	u32 mask = _mask | _copy;
 	u32 data = (_data & _mask) | (reg->data & _copy);
 	u32 i;
@@ -268,6 +264,7 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	u32 mask, data;
 
 	ram_mask(fuc, 0x10f808, 0x40000000, 0x40000000);
+	ram_block(fuc);
 	ram_wr32(fuc, 0x62c000, 0x0f0f0000);
 
 	/* MR1: turn termination on early, for some reason.. */
@@ -478,7 +475,7 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	ram_mask(fuc, 0x10f2e8, 0xffffffff, next->bios.timing[9]);
 
 	data = mask = 0x00000000;
-	if (NOTE00(ramcfg_08_20)) {
+	if (ram->diff.ramcfg_11_08_20) {
 		if (next->bios.ramcfg_11_08_20)
 			data |= 0x01000000;
 		mask |= 0x01000000;
@@ -486,11 +483,11 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	ram_mask(fuc, 0x10f200, mask, data);
 
 	data = mask = 0x00000000;
-	if (NOTE00(ramcfg_02_03 != 0)) {
+	if (ram->diff.ramcfg_11_02_03) {
 		data |= next->bios.ramcfg_11_02_03 << 8;
 		mask |= 0x00000300;
 	}
-	if (NOTE00(ramcfg_01_10)) {
+	if (ram->diff.ramcfg_11_01_10) {
 		if (next->bios.ramcfg_11_01_10)
 			data |= 0x70000000;
 		mask |= 0x70000000;
@@ -498,11 +495,11 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	ram_mask(fuc, 0x10f604, mask, data);
 
 	data = mask = 0x00000000;
-	if (NOTE00(timing_30_07 != 0)) {
+	if (ram->diff.timing_20_30_07) {
 		data |= next->bios.timing_20_30_07 << 28;
 		mask |= 0x70000000;
 	}
-	if (NOTE00(ramcfg_01_01)) {
+	if (ram->diff.ramcfg_11_01_01) {
 		if (next->bios.ramcfg_11_01_01)
 			data |= 0x00000100;
 		mask |= 0x00000100;
@@ -510,11 +507,11 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	ram_mask(fuc, 0x10f614, mask, data);
 
 	data = mask = 0x00000000;
-	if (NOTE00(timing_30_07 != 0)) {
+	if (ram->diff.timing_20_30_07) {
 		data |= next->bios.timing_20_30_07 << 28;
 		mask |= 0x70000000;
 	}
-	if (NOTE00(ramcfg_01_02)) {
+	if (ram->diff.ramcfg_11_01_02) {
 		if (next->bios.ramcfg_11_01_02)
 			data |= 0x00000100;
 		mask |= 0x00000100;
@@ -548,11 +545,11 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	ram_wr32(fuc, 0x10f870, 0x11111111 * next->bios.ramcfg_11_03_0f);
 
 	data = mask = 0x00000000;
-	if (NOTE00(ramcfg_02_03 != 0)) {
+	if (ram->diff.ramcfg_11_02_03) {
 		data |= next->bios.ramcfg_11_02_03;
 		mask |= 0x00000003;
 	}
-	if (NOTE00(ramcfg_01_10)) {
+	if (ram->diff.ramcfg_11_01_10) {
 		if (next->bios.ramcfg_11_01_10)
 			data |= 0x00000004;
 		mask |= 0x00000004;
@@ -666,6 +663,7 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	if (next->bios.ramcfg_11_07_02)
 		nve0_ram_train(fuc, 0x80020000, 0x01000000);
 
+	ram_unblock(fuc);
 	ram_wr32(fuc, 0x62c000, 0x0f0f0f00);
 
 	if (next->bios.rammap_11_08_01)
@@ -695,6 +693,7 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 	u32 mask, data;
 
 	ram_mask(fuc, 0x10f808, 0x40000000, 0x40000000);
+	ram_block(fuc);
 	ram_wr32(fuc, 0x62c000, 0x0f0f0000);
 
 	if (vc == 1 && ram_have(fuc, gpio2E)) {
@@ -917,6 +916,7 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 	ram_mask(fuc, 0x10f200, 0x80000000, 0x00000000);
 	ram_nsec(fuc, 1000);
 
+	ram_unblock(fuc);
 	ram_wr32(fuc, 0x62c000, 0x0f0f0f00);
 
 	if (next->bios.rammap_11_08_01)
@@ -932,58 +932,24 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
  ******************************************************************************/
 
 static int
-nve0_ram_calc_data(struct nouveau_fb *pfb, u32 freq,
+nve0_ram_calc_data(struct nouveau_fb *pfb, u32 khz,
 		   struct nouveau_ram_data *data)
 {
-	struct nouveau_bios *bios = nouveau_bios(pfb);
 	struct nve0_ram *ram = (void *)pfb->ram;
-	u8 strap, cnt, len;
-
-	/* lookup memory config data relevant to the target frequency */
-	ram->base.rammap.data = nvbios_rammapEp(bios, freq / 1000,
-					       &ram->base.rammap.version,
-					       &ram->base.rammap.size,
-					       &cnt, &len, &data->bios);
-	if (!ram->base.rammap.data || ram->base.rammap.version != 0x11 ||
-	     ram->base.rammap.size < 0x09) {
-		nv_error(pfb, "invalid/missing rammap entry\n");
-		return -EINVAL;
-	}
-
-	/* locate specific data set for the attached memory */
-	strap = nvbios_ramcfg_index(nv_subdev(pfb));
-	ram->base.ramcfg.data = nvbios_rammapSp(bios, ram->base.rammap.data,
-						ram->base.rammap.version,
-						ram->base.rammap.size,
-						cnt, len, strap,
-						&ram->base.ramcfg.version,
-						&ram->base.ramcfg.size,
-						&data->bios);
-	if (!ram->base.ramcfg.data || ram->base.ramcfg.version != 0x11 ||
-	     ram->base.ramcfg.size < 0x08) {
-		nv_error(pfb, "invalid/missing ramcfg entry\n");
-		return -EINVAL;
-	}
-
-	/* lookup memory timings, if bios says they're present */
-	strap = nv_ro08(bios, ram->base.ramcfg.data + 0x00);
-	if (strap != 0xff) {
-		ram->base.timing.data =
-			nvbios_timingEp(bios, strap, &ram->base.timing.version,
-				       &ram->base.timing.size, &cnt, &len,
-				       &data->bios);
-		if (!ram->base.timing.data ||
-		     ram->base.timing.version != 0x20 ||
-		     ram->base.timing.size < 0x33) {
-			nv_error(pfb, "invalid/missing timing entry\n");
-			return -EINVAL;
+	struct nouveau_ram_data *cfg;
+	u32 mhz = khz / 1000;
+
+	list_for_each_entry(cfg, &ram->cfg, head) {
+		if (mhz >= cfg->bios.rammap_min &&
+		    mhz <= cfg->bios.rammap_max) {
+			*data = *cfg;
+			data->freq = khz;
+			return 0;
 		}
-	} else {
-		ram->base.timing.data = 0;
 	}
 
-	data->freq = freq;
-	return 0;
+	nv_error(ram, "ramcfg data for %dMHz not found\n", mhz);
+	return -EINVAL;
 }
 
 static int
@@ -1106,13 +1072,99 @@ nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	return nve0_ram_calc_xits(pfb, ram->base.next);
 }
 
+static void
+nve0_ram_prog_0(struct nouveau_fb *pfb, u32 freq)
+{
+	struct nve0_ram *ram = (void *)pfb->ram;
+	struct nouveau_ram_data *cfg;
+	u32 mhz = freq / 1000;
+	u32 mask, data;
+
+	list_for_each_entry(cfg, &ram->cfg, head) {
+		if (mhz >= cfg->bios.rammap_min &&
+		    mhz <= cfg->bios.rammap_max)
+			break;
+	}
+
+	if (&cfg->head == &ram->cfg)
+		return;
+
+	if (mask = 0, data = 0, ram->diff.rammap_11_0a_03fe) {
+		data |= cfg->bios.rammap_11_0a_03fe << 12;
+		mask |= 0x001ff000;
+	}
+	if (ram->diff.rammap_11_09_01ff) {
+		data |= cfg->bios.rammap_11_09_01ff;
+		mask |= 0x000001ff;
+	}
+	nv_mask(pfb, 0x10f468, mask, data);
+
+	if (mask = 0, data = 0, ram->diff.rammap_11_0a_0400) {
+		data |= cfg->bios.rammap_11_0a_0400;
+		mask |= 0x00000001;
+	}
+	nv_mask(pfb, 0x10f420, mask, data);
+
+	if (mask = 0, data = 0, ram->diff.rammap_11_0a_0800) {
+		data |= cfg->bios.rammap_11_0a_0800;
+		mask |= 0x00000001;
+	}
+	nv_mask(pfb, 0x10f430, mask, data);
+
+	if (mask = 0, data = 0, ram->diff.rammap_11_0b_01f0) {
+		data |= cfg->bios.rammap_11_0b_01f0;
+		mask |= 0x0000001f;
+	}
+	nv_mask(pfb, 0x10f400, mask, data);
+
+	if (mask = 0, data = 0, ram->diff.rammap_11_0b_0200) {
+		data |= cfg->bios.rammap_11_0b_0200 << 9;
+		mask |= 0x00000200;
+	}
+	nv_mask(pfb, 0x10f410, mask, data);
+
+	if (mask = 0, data = 0, ram->diff.rammap_11_0d) {
+		data |= cfg->bios.rammap_11_0d << 16;
+		mask |= 0x00ff0000;
+	}
+	if (ram->diff.rammap_11_0f) {
+		data |= cfg->bios.rammap_11_0f << 8;
+		mask |= 0x0000ff00;
+	}
+	nv_mask(pfb, 0x10f440, mask, data);
+
+	if (mask = 0, data = 0, ram->diff.rammap_11_0e) {
+		data |= cfg->bios.rammap_11_0e << 8;
+		mask |= 0x0000ff00;
+	}
+	if (ram->diff.rammap_11_0b_0800) {
+		data |= cfg->bios.rammap_11_0b_0800 << 7;
+		mask |= 0x00000080;
+	}
+	if (ram->diff.rammap_11_0b_0400) {
+		data |= cfg->bios.rammap_11_0b_0400 << 5;
+		mask |= 0x00000020;
+	}
+	nv_mask(pfb, 0x10f444, mask, data);
+}
+
 static int
 nve0_ram_prog(struct nouveau_fb *pfb)
 {
 	struct nouveau_device *device = nv_device(pfb);
 	struct nve0_ram *ram = (void *)pfb->ram;
 	struct nve0_ramfuc *fuc = &ram->fuc;
-	ram_exec(fuc, nouveau_boolopt(device->cfgopt, "NvMemExec", true));
+	struct nouveau_ram_data *next = ram->base.next;
+
+	if (!nouveau_boolopt(device->cfgopt, "NvMemExec", true)) {
+		ram_exec(fuc, false);
+		return (ram->base.next == &ram->base.xition);
+	}
+
+	nve0_ram_prog_0(pfb, 1000);
+	ram_exec(fuc, true);
+	nve0_ram_prog_0(pfb, next->freq);
+
 	return (ram->base.next == &ram->base.xition);
 }
 
@@ -1125,24 +1177,147 @@ nve0_ram_tidy(struct nouveau_fb *pfb)
 	ram_exec(fuc, false);
 }
 
+struct nve0_ram_train {
+	u16 mask;
+	struct nvbios_M0209S remap;
+	struct nvbios_M0209S type00;
+	struct nvbios_M0209S type01;
+	struct nvbios_M0209S type04;
+	struct nvbios_M0209S type06;
+	struct nvbios_M0209S type07;
+	struct nvbios_M0209S type08;
+	struct nvbios_M0209S type09;
+};
+
+static int
+nve0_ram_train_type(struct nouveau_fb *pfb, int i, u8 ramcfg,
+		    struct nve0_ram_train *train)
+{
+	struct nouveau_bios *bios = nouveau_bios(pfb);
+	struct nvbios_M0205E M0205E;
+	struct nvbios_M0205S M0205S;
+	struct nvbios_M0209E M0209E;
+	struct nvbios_M0209S *remap = &train->remap;
+	struct nvbios_M0209S *value;
+	u8  ver, hdr, cnt, len;
+	u32 data;
+
+	/* determine type of data for this index */
+	if (!(data = nvbios_M0205Ep(bios, i, &ver, &hdr, &cnt, &len, &M0205E)))
+		return -ENOENT;
+
+	switch (M0205E.type) {
+	case 0x00: value = &train->type00; break;
+	case 0x01: value = &train->type01; break;
+	case 0x04: value = &train->type04; break;
+	case 0x06: value = &train->type06; break;
+	case 0x07: value = &train->type07; break;
+	case 0x08: value = &train->type08; break;
+	case 0x09: value = &train->type09; break;
+	default:
+		return 0;
+	}
+
+	/* training data index determined by ramcfg strap */
+	if (!(data = nvbios_M0205Sp(bios, i, ramcfg, &ver, &hdr, &M0205S)))
+		return -EINVAL;
+	i = M0205S.data;
+
+	/* training data format information */
+	if (!(data = nvbios_M0209Ep(bios, i, &ver, &hdr, &cnt, &len, &M0209E)))
+		return -EINVAL;
+
+	/* ... and the raw data */
+	if (!(data = nvbios_M0209Sp(bios, i, 0, &ver, &hdr, value)))
+		return -EINVAL;
+
+	if (M0209E.v02_07 == 2) {
+		/* of course! why wouldn't we have a pointer to another entry
+		 * in the same table, and use the first one as an array of
+		 * remap indices...
+		 */
+		if (!(data = nvbios_M0209Sp(bios, M0209E.v03, 0, &ver, &hdr,
+					    remap)))
+			return -EINVAL;
+
+		for (i = 0; i < ARRAY_SIZE(value->data); i++)
+			value->data[i] = remap->data[value->data[i]];
+	} else
+	if (M0209E.v02_07 != 1)
+		return -EINVAL;
+
+	train->mask |= 1 << M0205E.type;
+	return 0;
+}
+
+static int
+nve0_ram_train_init_0(struct nouveau_fb *pfb, struct nve0_ram_train *train)
+{
+	int i, j;
+
+	if ((train->mask & 0x03d3) != 0x03d3) {
+		nv_warn(pfb, "missing link training data\n");
+		return -EINVAL;
+	}
+
+	for (i = 0; i < 0x30; i++) {
+		for (j = 0; j < 8; j += 4) {
+			nv_wr32(pfb, 0x10f968 + j, 0x00000000 | (i << 8));
+			nv_wr32(pfb, 0x10f920 + j, 0x00000000 |
+						   train->type08.data[i] << 4 |
+						   train->type06.data[i]);
+			nv_wr32(pfb, 0x10f918 + j, train->type00.data[i]);
+			nv_wr32(pfb, 0x10f920 + j, 0x00000100 |
+						   train->type09.data[i] << 4 |
+						   train->type07.data[i]);
+			nv_wr32(pfb, 0x10f918 + j, train->type01.data[i]);
+		}
+	}
+
+	for (j = 0; j < 8; j += 4) {
+		for (i = 0; i < 0x100; i++) {
+			nv_wr32(pfb, 0x10f968 + j, i);
+			nv_wr32(pfb, 0x10f900 + j, train->type04.data[i]);
+		}
+	}
+
+	return 0;
+}
+
+static int
+nve0_ram_train_init(struct nouveau_fb *pfb)
+{
+	u8 ramcfg = nvbios_ramcfg_index(nv_subdev(pfb));
+	struct nve0_ram_train *train;
+	int ret = -ENOMEM, i;
+
+	if ((train = kzalloc(sizeof(*train), GFP_KERNEL))) {
+		for (i = 0; i < 0x100; i++) {
+			ret = nve0_ram_train_type(pfb, i, ramcfg, train);
+			if (ret && ret != -ENOENT)
+				break;
+		}
+	}
+
+	switch (pfb->ram->type) {
+	case NV_MEM_TYPE_GDDR5:
+		ret = nve0_ram_train_init_0(pfb, train);
+		break;
+	default:
+		ret = 0;
+		break;
+	}
+
+	kfree(train);
+	return ret;
+}
+
 int
 nve0_ram_init(struct nouveau_object *object)
 {
 	struct nouveau_fb *pfb = (void *)object->parent;
 	struct nve0_ram *ram   = (void *)object;
 	struct nouveau_bios *bios = nouveau_bios(pfb);
-	static const u8  train0[] = {
-		0x00, 0xff, 0xff, 0x00, 0xff, 0x00,
-		0x00, 0xff, 0xff, 0x00, 0xff, 0x00,
-	};
-	static const u32 train1[] = {
-		0x00000000, 0xffffffff,
-		0x55555555, 0xaaaaaaaa,
-		0x33333333, 0xcccccccc,
-		0xf0f0f0f0, 0x0f0f0f0f,
-		0x00ff00ff, 0xff00ff00,
-		0x0000ffff, 0xffff0000,
-	};
 	u8  ver, hdr, cnt, len, snr, ssz;
 	u32 data, save;
 	int ret, i;
@@ -1168,51 +1343,107 @@ nve0_ram_init(struct nouveau_object *object)
 
 	cnt  = nv_ro08(bios, data + 0x14); /* guess at count */
 	data = nv_ro32(bios, data + 0x10); /* guess u32... */
-	save = nv_rd32(pfb, 0x10f65c);
-	for (i = 0; i < cnt; i++) {
-		nv_mask(pfb, 0x10f65c, 0x000000f0, i << 4);
-		nvbios_exec(&(struct nvbios_init) {
-				.subdev = nv_subdev(pfb),
-				.bios = bios,
-				.offset = nv_ro32(bios, data), /* guess u32 */
-				.execute = 1,
-			    });
-		data += 4;
-	}
-	nv_wr32(pfb, 0x10f65c, save);
+	save = nv_rd32(pfb, 0x10f65c) & 0x000000f0;
+	for (i = 0; i < cnt; i++, data += 4) {
+		if (i != save >> 4) {
+			nv_mask(pfb, 0x10f65c, 0x000000f0, i << 4);
+			nvbios_exec(&(struct nvbios_init) {
+					.subdev = nv_subdev(pfb),
+					.bios = bios,
+					.offset = nv_ro32(bios, data),
+					.execute = 1,
+				    });
+		}
+	}
+	nv_mask(pfb, 0x10f65c, 0x000000f0, save);
 	nv_mask(pfb, 0x10f584, 0x11000000, 0x00000000);
+	nv_wr32(pfb, 0x10ecc0, 0xffffffff);
+	nv_mask(pfb, 0x10f160, 0x00000010, 0x00000010);
 
-	switch (ram->base.type) {
-	case NV_MEM_TYPE_GDDR5:
-		for (i = 0; i < 0x30; i++) {
-			nv_wr32(pfb, 0x10f968, 0x00000000 | (i << 8));
-			nv_wr32(pfb, 0x10f920, 0x00000000 | train0[i % 12]);
-			nv_wr32(pfb, 0x10f918,              train1[i % 12]);
-			nv_wr32(pfb, 0x10f920, 0x00000100 | train0[i % 12]);
-			nv_wr32(pfb, 0x10f918,              train1[i % 12]);
-
-			nv_wr32(pfb, 0x10f96c, 0x00000000 | (i << 8));
-			nv_wr32(pfb, 0x10f924, 0x00000000 | train0[i % 12]);
-			nv_wr32(pfb, 0x10f91c,              train1[i % 12]);
-			nv_wr32(pfb, 0x10f924, 0x00000100 | train0[i % 12]);
-			nv_wr32(pfb, 0x10f91c,              train1[i % 12]);
-		}
+	return nve0_ram_train_init(pfb);
+}
 
-		for (i = 0; i < 0x100; i++) {
-			nv_wr32(pfb, 0x10f968, i);
-			nv_wr32(pfb, 0x10f900, train1[2 + (i & 1)]);
-		}
+static int
+nve0_ram_ctor_data(struct nve0_ram *ram, u8 ramcfg, int i)
+{
+	struct nouveau_fb *pfb = (void *)nv_object(ram)->parent;
+	struct nouveau_bios *bios = nouveau_bios(pfb);
+	struct nouveau_ram_data *cfg;
+	struct nvbios_ramcfg *d = &ram->diff;
+	struct nvbios_ramcfg *p, *n;
+	u8  ver, hdr, cnt, len;
+	u32 data;
+	int ret;
 
-		for (i = 0; i < 0x100; i++) {
-			nv_wr32(pfb, 0x10f96c, i);
-			nv_wr32(pfb, 0x10f900, train1[2 + (i & 1)]);
-		}
-		break;
-	default:
-		break;
+	if (!(cfg = kmalloc(sizeof(*cfg), GFP_KERNEL)))
+		return -ENOMEM;
+	p = &list_last_entry(&ram->cfg, typeof(*cfg), head)->bios;
+	n = &cfg->bios;
+
+	/* memory config data for a range of target frequencies */
+	data = nvbios_rammapEp(bios, i, &ver, &hdr, &cnt, &len, &cfg->bios);
+	if (ret = -ENOENT, !data)
+		goto done;
+	if (ret = -ENOSYS, ver != 0x11 || hdr < 0x12)
+		goto done;
+
+	/* ... and a portion specific to the attached memory */
+	data = nvbios_rammapSp(bios, data, ver, hdr, cnt, len, ramcfg,
+			       &ver, &hdr, &cfg->bios);
+	if (ret = -EINVAL, !data)
+		goto done;
+	if (ret = -ENOSYS, ver != 0x11 || hdr < 0x0a)
+		goto done;
+
+	/* lookup memory timings, if bios says they're present */
+	if (cfg->bios.ramcfg_timing != 0xff) {
+		data = nvbios_timingEp(bios, cfg->bios.ramcfg_timing,
+				       &ver, &hdr, &cnt, &len,
+				       &cfg->bios);
+		if (ret = -EINVAL, !data)
+			goto done;
+		if (ret = -ENOSYS, ver != 0x20 || hdr < 0x33)
+			goto done;
 	}
 
-	return 0;
+	list_add_tail(&cfg->head, &ram->cfg);
+	if (ret = 0, i == 0)
+		goto done;
+
+	d->rammap_11_0a_03fe |= p->rammap_11_0a_03fe != n->rammap_11_0a_03fe;
+	d->rammap_11_09_01ff |= p->rammap_11_09_01ff != n->rammap_11_09_01ff;
+	d->rammap_11_0a_0400 |= p->rammap_11_0a_0400 != n->rammap_11_0a_0400;
+	d->rammap_11_0a_0800 |= p->rammap_11_0a_0800 != n->rammap_11_0a_0800;
+	d->rammap_11_0b_01f0 |= p->rammap_11_0b_01f0 != n->rammap_11_0b_01f0;
+	d->rammap_11_0b_0200 |= p->rammap_11_0b_0200 != n->rammap_11_0b_0200;
+	d->rammap_11_0d |= p->rammap_11_0d != n->rammap_11_0d;
+	d->rammap_11_0f |= p->rammap_11_0f != n->rammap_11_0f;
+	d->rammap_11_0e |= p->rammap_11_0e != n->rammap_11_0e;
+	d->rammap_11_0b_0800 |= p->rammap_11_0b_0800 != n->rammap_11_0b_0800;
+	d->rammap_11_0b_0400 |= p->rammap_11_0b_0400 != n->rammap_11_0b_0400;
+	d->ramcfg_11_01_01 |= p->ramcfg_11_01_01 != n->ramcfg_11_01_01;
+	d->ramcfg_11_01_02 |= p->ramcfg_11_01_02 != n->ramcfg_11_01_02;
+	d->ramcfg_11_01_10 |= p->ramcfg_11_01_10 != n->ramcfg_11_01_10;
+	d->ramcfg_11_02_03 |= p->ramcfg_11_02_03 != n->ramcfg_11_02_03;
+	d->ramcfg_11_08_20 |= p->ramcfg_11_08_20 != n->ramcfg_11_08_20;
+	d->timing_20_30_07 |= p->timing_20_30_07 != n->timing_20_30_07;
+done:
+	if (ret)
+		kfree(cfg);
+	return ret;
+}
+
+static void
+nve0_ram_dtor(struct nouveau_object *object)
+{
+	struct nve0_ram *ram = (void *)object;
+	struct nouveau_ram_data *cfg, *tmp;
+
+	list_for_each_entry_safe(cfg, tmp, &ram->cfg, head) {
+		kfree(cfg);
+	}
+
+	nouveau_ram_destroy(&ram->base);
 }
 
 static int
@@ -1226,6 +1457,7 @@ nve0_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct dcb_gpio_func func;
 	struct nve0_ram *ram;
 	int ret, i;
+	u8  ramcfg = nvbios_ramcfg_index(nv_subdev(pfb));
 	u32 tmp;
 
 	ret = nvc0_ram_create(parent, engine, oclass, 0x022554, &ram);
@@ -1233,6 +1465,8 @@ nve0_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
+	INIT_LIST_HEAD(&ram->cfg);
+
 	switch (ram->base.type) {
 	case NV_MEM_TYPE_DDR3:
 	case NV_MEM_TYPE_GDDR5:
@@ -1264,7 +1498,26 @@ nve0_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		}
 	}
 
-	// parse bios data for both pll's
+	/* parse bios data for all rammap table entries up-front, and
+	 * build information on whether certain fields differ between
+	 * any of the entries.
+	 *
+	 * the binary driver appears to completely ignore some fields
+	 * when all entries contain the same value.  at first, it was
+	 * hoped that these were mere optimisations and the bios init
+	 * tables had configured as per the values here, but there is
+	 * evidence now to suggest that this isn't the case and we do
+	 * need to treat this condition as a "don't touch" indicator.
+	 */
+	for (i = 0; !ret; i++) {
+		ret = nve0_ram_ctor_data(ram, ramcfg, i);
+		if (ret && ret != -ENOENT) {
+			nv_error(pfb, "failed to parse ramcfg data\n");
+			return ret;
+		}
+	}
+
+	/* parse bios data for both pll's */
 	ret = nvbios_pll_parse(bios, 0x0c, &ram->fuc.refpll);
 	if (ret) {
 		nv_error(pfb, "mclk refpll data not found\n");
@@ -1277,6 +1530,7 @@ nve0_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		return ret;
 	}
 
+	/* lookup memory voltage gpios */
 	ret = gpio->find(gpio, 0, 0x18, DCB_GPIO_UNUSED, &func);
 	if (ret == 0) {
 		ram->fuc.r_gpioMV = ramfuc_reg(0x00d610 + (func.line * 0x04));
@@ -1385,7 +1639,7 @@ nve0_ram_oclass = {
 	.handle = 0,
 	.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nve0_ram_ctor,
-		.dtor = _nouveau_ram_dtor,
+		.dtor = nve0_ram_dtor,
 		.init = nve0_ram_init,
 		.fini = _nouveau_ram_fini,
 	}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/sddr2.c b/drivers/gpu/drm/nouveau/core/subdev/fb/sddr2.c
new file mode 100644
index 000000000000..252575f3aa29
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/sddr2.c
@@ -0,0 +1,94 @@
+/*
+ * Copyright 2014 Roy Spliet
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Roy Spliet <rspliet@eclipso.eu>
+ *          Ben Skeggs
+ */
+
+#include "priv.h"
+
+struct ramxlat {
+	int id;
+	u8 enc;
+};
+
+static inline int
+ramxlat(const struct ramxlat *xlat, int id)
+{
+	while (xlat->id >= 0) {
+		if (xlat->id == id)
+			return xlat->enc;
+		xlat++;
+	}
+	return -EINVAL;
+}
+
+static const struct ramxlat
+ramddr2_cl[] = {
+	{ 2, 2 }, { 3, 3 }, { 4, 4 }, { 5, 5 }, { 6, 6 },
+	/* The following are available in some, but not all DDR2 docs */
+	{ 7, 7 },
+	{ -1 }
+};
+
+static const struct ramxlat
+ramddr2_wr[] = {
+	{ 2, 1 }, { 3, 2 }, { 4, 3 }, { 5, 4 }, { 6, 5 },
+	/* The following are available in some, but not all DDR2 docs */
+	{ 7, 6 },
+	{ -1 }
+};
+
+int
+nouveau_sddr2_calc(struct nouveau_ram *ram)
+{
+	int CL, WR, DLL = 0, ODT = 0;
+
+	switch (ram->next->bios.timing_ver) {
+	case 0x10:
+		CL  = ram->next->bios.timing_10_CL;
+		WR  = ram->next->bios.timing_10_WR;
+		DLL = !ram->next->bios.ramcfg_10_DLLoff;
+		ODT = ram->next->bios.timing_10_ODT & 3;
+		break;
+	case 0x20:
+		CL  = (ram->next->bios.timing[1] & 0x0000001f);
+		WR  = (ram->next->bios.timing[2] & 0x007f0000) >> 16;
+		break;
+	default:
+		return -ENOSYS;
+	}
+
+	CL  = ramxlat(ramddr2_cl, CL);
+	WR  = ramxlat(ramddr2_wr, WR);
+	if (CL < 0 || WR < 0)
+		return -EINVAL;
+
+	ram->mr[0] &= ~0xf70;
+	ram->mr[0] |= (WR & 0x07) << 9;
+	ram->mr[0] |= (CL & 0x07) << 4;
+
+	ram->mr[1] &= ~0x045;
+	ram->mr[1] |= (ODT & 0x1) << 2;
+	ram->mr[1] |= (ODT & 0x2) << 5;
+	ram->mr[1] |= !DLL;
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/sddr3.c b/drivers/gpu/drm/nouveau/core/subdev/fb/sddr3.c
index ebd4cd9c35d9..a2dca4869e52 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/sddr3.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/sddr3.c
@@ -20,9 +20,9 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  * Authors: Ben Skeggs <bskeggs@redhat.com>
+ * 	    Roy Spliet <rspliet@eclipso.eu>
  */
 
-#include <subdev/bios.h>
 #include "priv.h"
 
 struct ramxlat {
@@ -69,31 +69,52 @@ ramddr3_cwl[] = {
 int
 nouveau_sddr3_calc(struct nouveau_ram *ram)
 {
-	struct nouveau_bios *bios = nouveau_bios(ram);
-	int WL, CL, WR;
+	int CWL, CL, WR, DLL = 0, ODT = 0;
 
-	switch (!!ram->timing.data * ram->timing.version) {
+	switch (ram->next->bios.timing_ver) {
+	case 0x10:
+		if (ram->next->bios.timing_hdr < 0x17) {
+			/* XXX: NV50: Get CWL from the timing register */
+			return -ENOSYS;
+		}
+		CWL = ram->next->bios.timing_10_CWL;
+		CL  = ram->next->bios.timing_10_CL;
+		WR  = ram->next->bios.timing_10_WR;
+		DLL = !ram->next->bios.ramcfg_10_DLLoff;
+		ODT = ram->next->bios.timing_10_ODT;
+		break;
 	case 0x20:
-		WL = (nv_ro16(bios, ram->timing.data + 0x04) & 0x0f80) >> 7;
-		CL =  nv_ro08(bios, ram->timing.data + 0x04) & 0x1f;
-		WR =  nv_ro08(bios, ram->timing.data + 0x0a) & 0x7f;
+		CWL = (ram->next->bios.timing[1] & 0x00000f80) >> 7;
+		CL  = (ram->next->bios.timing[1] & 0x0000001f) >> 0;
+		WR  = (ram->next->bios.timing[2] & 0x007f0000) >> 16;
+		/* XXX: Get these values from the VBIOS instead */
+		DLL = !(ram->mr[1] & 0x1);
+		ODT =   (ram->mr[1] & 0x004) >> 2 |
+			(ram->mr[1] & 0x040) >> 5 |
+		        (ram->mr[1] & 0x200) >> 7;
 		break;
 	default:
 		return -ENOSYS;
 	}
 
-	WL = ramxlat(ramddr3_cwl, WL);
-	CL = ramxlat(ramddr3_cl, CL);
-	WR = ramxlat(ramddr3_wr, WR);
-	if (WL < 0 || CL < 0 || WR < 0)
+	CWL = ramxlat(ramddr3_cwl, CWL);
+	CL  = ramxlat(ramddr3_cl, CL);
+	WR  = ramxlat(ramddr3_wr, WR);
+	if (CL < 0 || CWL < 0 || WR < 0)
 		return -EINVAL;
 
-	ram->mr[0] &= ~0xe74;
+	ram->mr[0] &= ~0xf74;
 	ram->mr[0] |= (WR & 0x07) << 9;
 	ram->mr[0] |= (CL & 0x0e) << 3;
 	ram->mr[0] |= (CL & 0x01) << 2;
 
+	ram->mr[1] &= ~0x245;
+	ram->mr[1] |= (ODT & 0x1) << 2;
+	ram->mr[1] |= (ODT & 0x2) << 5;
+	ram->mr[1] |= (ODT & 0x4) << 7;
+	ram->mr[1] |= !DLL;
+
 	ram->mr[2] &= ~0x038;
-	ram->mr[2] |= (WL & 0x07) << 3;
+	ram->mr[2] |= (CWL & 0x07) << 3;
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fuse/base.c b/drivers/gpu/drm/nouveau/core/subdev/fuse/base.c
new file mode 100644
index 000000000000..9e8e92127715
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/fuse/base.c
@@ -0,0 +1,54 @@
+/*
+ * Copyright 2014 Martin Peres
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Martin Peres
+ */
+
+#include <subdev/fuse.h>
+
+int
+_nouveau_fuse_init(struct nouveau_object *object)
+{
+	struct nouveau_fuse *fuse = (void *)object;
+	return nouveau_subdev_init(&fuse->base);
+}
+
+void
+_nouveau_fuse_dtor(struct nouveau_object *object)
+{
+	struct nouveau_fuse *fuse = (void *)object;
+	nouveau_subdev_destroy(&fuse->base);
+}
+
+int
+nouveau_fuse_create_(struct nouveau_object *parent,
+		     struct nouveau_object *engine,
+		     struct nouveau_oclass *oclass, int length, void **pobject)
+{
+	struct nouveau_fuse *fuse;
+	int ret;
+
+	ret = nouveau_subdev_create_(parent, engine, oclass, 0, "FUSE",
+				     "fuse", length, pobject);
+	fuse = *pobject;
+
+	return ret;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fuse/g80.c b/drivers/gpu/drm/nouveau/core/subdev/fuse/g80.c
new file mode 100644
index 000000000000..a374ade485be
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/fuse/g80.c
@@ -0,0 +1,81 @@
+/*
+ * Copyright 2014 Martin Peres
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Martin Peres
+ */
+
+#include "priv.h"
+
+struct g80_fuse_priv {
+	struct nouveau_fuse base;
+
+	spinlock_t fuse_enable_lock;
+};
+
+static u32
+g80_fuse_rd32(struct nouveau_object *object, u64 addr)
+{
+	struct g80_fuse_priv *priv = (void *)object;
+	unsigned long flags;
+	u32 fuse_enable, val;
+
+	spin_lock_irqsave(&priv->fuse_enable_lock, flags);
+
+	/* racy if another part of nouveau start writing to this reg */
+	fuse_enable = nv_mask(priv, 0x1084, 0x800, 0x800);
+	val = nv_rd32(priv, 0x21000 + addr);
+	nv_wr32(priv, 0x1084, fuse_enable);
+
+	spin_unlock_irqrestore(&priv->fuse_enable_lock, flags);
+
+	return val;
+}
+
+
+static int
+g80_fuse_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+	       struct nouveau_oclass *oclass, void *data, u32 size,
+	       struct nouveau_object **pobject)
+{
+	struct g80_fuse_priv *priv;
+	int ret;
+
+	ret = nouveau_fuse_create(parent, engine, oclass, &priv);
+	*pobject = nv_object(priv);
+	if (ret)
+		return ret;
+
+	spin_lock_init(&priv->fuse_enable_lock);
+
+	return 0;
+}
+
+struct nouveau_oclass
+g80_fuse_oclass = {
+	.handle = NV_SUBDEV(FUSE, 0x50),
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = g80_fuse_ctor,
+		.dtor = _nouveau_fuse_dtor,
+		.init = _nouveau_fuse_init,
+		.fini = _nouveau_fuse_fini,
+		.rd32 = g80_fuse_rd32,
+	},
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fuse/gf100.c b/drivers/gpu/drm/nouveau/core/subdev/fuse/gf100.c
new file mode 100644
index 000000000000..5ed03f54b3d4
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/fuse/gf100.c
@@ -0,0 +1,83 @@
+/*
+ * Copyright 2014 Martin Peres
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Martin Peres
+ */
+
+#include "priv.h"
+
+struct gf100_fuse_priv {
+	struct nouveau_fuse base;
+
+	spinlock_t fuse_enable_lock;
+};
+
+static u32
+gf100_fuse_rd32(struct nouveau_object *object, u64 addr)
+{
+	struct gf100_fuse_priv *priv = (void *)object;
+	unsigned long flags;
+	u32 fuse_enable, unk, val;
+
+	spin_lock_irqsave(&priv->fuse_enable_lock, flags);
+
+	/* racy if another part of nouveau start writing to these regs */
+	fuse_enable = nv_mask(priv, 0x22400, 0x800, 0x800);
+	unk = nv_mask(priv, 0x21000, 0x1, 0x1);
+	val = nv_rd32(priv, 0x21100 + addr);
+	nv_wr32(priv, 0x21000, unk);
+	nv_wr32(priv, 0x22400, fuse_enable);
+
+	spin_unlock_irqrestore(&priv->fuse_enable_lock, flags);
+
+	return val;
+}
+
+
+static int
+gf100_fuse_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+	       struct nouveau_oclass *oclass, void *data, u32 size,
+	       struct nouveau_object **pobject)
+{
+	struct gf100_fuse_priv *priv;
+	int ret;
+
+	ret = nouveau_fuse_create(parent, engine, oclass, &priv);
+	*pobject = nv_object(priv);
+	if (ret)
+		return ret;
+
+	spin_lock_init(&priv->fuse_enable_lock);
+
+	return 0;
+}
+
+struct nouveau_oclass
+gf100_fuse_oclass = {
+	.handle = NV_SUBDEV(FUSE, 0xC0),
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = gf100_fuse_ctor,
+		.dtor = _nouveau_fuse_dtor,
+		.init = _nouveau_fuse_init,
+		.fini = _nouveau_fuse_fini,
+		.rd32 = gf100_fuse_rd32,
+	},
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fuse/gm107.c b/drivers/gpu/drm/nouveau/core/subdev/fuse/gm107.c
new file mode 100644
index 000000000000..4f1a636c6538
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/fuse/gm107.c
@@ -0,0 +1,66 @@
+/*
+ * Copyright 2014 Martin Peres
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Martin Peres
+ */
+
+#include "priv.h"
+
+struct gm107_fuse_priv {
+	struct nouveau_fuse base;
+};
+
+static u32
+gm107_fuse_rd32(struct nouveau_object *object, u64 addr)
+{
+	struct gf100_fuse_priv *priv = (void *)object;
+
+	return nv_rd32(priv, 0x21100 + addr);
+}
+
+
+static int
+gm107_fuse_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+	       struct nouveau_oclass *oclass, void *data, u32 size,
+	       struct nouveau_object **pobject)
+{
+	struct gm107_fuse_priv *priv;
+	int ret;
+
+	ret = nouveau_fuse_create(parent, engine, oclass, &priv);
+	*pobject = nv_object(priv);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+struct nouveau_oclass
+gm107_fuse_oclass = {
+	.handle = NV_SUBDEV(FUSE, 0x117),
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = gm107_fuse_ctor,
+		.dtor = _nouveau_fuse_dtor,
+		.init = _nouveau_fuse_init,
+		.fini = _nouveau_fuse_fini,
+		.rd32 = gm107_fuse_rd32,
+	},
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fuse/priv.h b/drivers/gpu/drm/nouveau/core/subdev/fuse/priv.h
new file mode 100644
index 000000000000..d2085411a5cb
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/fuse/priv.h
@@ -0,0 +1,9 @@
+#ifndef __NVKM_FUSE_PRIV_H__
+#define __NVKM_FUSE_PRIV_H__
+
+#include <subdev/fuse.h>
+
+int _nouveau_fuse_init(struct nouveau_object *object);
+void _nouveau_fuse_dtor(struct nouveau_object *object);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/gpio/base.c b/drivers/gpu/drm/nouveau/core/subdev/gpio/base.c
index b1e3ed7c8beb..7ad99b763f4c 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/gpio/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/gpio/base.c
@@ -122,7 +122,8 @@ nouveau_gpio_intr_init(struct nvkm_event *event, int type, int index)
 }
 
 static int
-nouveau_gpio_intr_ctor(void *data, u32 size, struct nvkm_notify *notify)
+nouveau_gpio_intr_ctor(struct nouveau_object *object, void *data, u32 size,
+		       struct nvkm_notify *notify)
 {
 	struct nvkm_gpio_ntfy_req *req = data;
 	if (!WARN_ON(size != sizeof(*req))) {
diff --git a/drivers/gpu/drm/nouveau/core/subdev/gpio/nv50.c b/drivers/gpu/drm/nouveau/core/subdev/gpio/nv50.c
index 1864fa98e6b1..2e30d5a62d6e 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/gpio/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/gpio/nv50.c
@@ -54,7 +54,7 @@ nv50_gpio_reset(struct nouveau_gpio *gpio, u8 match)
 	}
 }
 
-static int
+int
 nv50_gpio_location(int line, u32 *reg, u32 *shift)
 {
 	const u32 nv50_gpio_reg[4] = { 0xe104, 0xe108, 0xe280, 0xe284 };
diff --git a/drivers/gpu/drm/nouveau/core/subdev/gpio/nv92.c b/drivers/gpu/drm/nouveau/core/subdev/gpio/nv94.c
index 252083d376f5..cae404ccadac 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/gpio/nv92.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/gpio/nv94.c
@@ -25,7 +25,7 @@
 #include "priv.h"
 
 void
-nv92_gpio_intr_stat(struct nouveau_gpio *gpio, u32 *hi, u32 *lo)
+nv94_gpio_intr_stat(struct nouveau_gpio *gpio, u32 *hi, u32 *lo)
 {
 	u32 intr0 = nv_rd32(gpio, 0x00e054);
 	u32 intr1 = nv_rd32(gpio, 0x00e074);
@@ -38,7 +38,7 @@ nv92_gpio_intr_stat(struct nouveau_gpio *gpio, u32 *hi, u32 *lo)
 }
 
 void
-nv92_gpio_intr_mask(struct nouveau_gpio *gpio, u32 type, u32 mask, u32 data)
+nv94_gpio_intr_mask(struct nouveau_gpio *gpio, u32 type, u32 mask, u32 data)
 {
 	u32 inte0 = nv_rd32(gpio, 0x00e050);
 	u32 inte1 = nv_rd32(gpio, 0x00e070);
@@ -57,8 +57,8 @@ nv92_gpio_intr_mask(struct nouveau_gpio *gpio, u32 type, u32 mask, u32 data)
 }
 
 struct nouveau_oclass *
-nv92_gpio_oclass = &(struct nouveau_gpio_impl) {
-	.base.handle = NV_SUBDEV(GPIO, 0x92),
+nv94_gpio_oclass = &(struct nouveau_gpio_impl) {
+	.base.handle = NV_SUBDEV(GPIO, 0x94),
 	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = _nouveau_gpio_ctor,
 		.dtor = _nouveau_gpio_dtor,
@@ -66,8 +66,8 @@ nv92_gpio_oclass = &(struct nouveau_gpio_impl) {
 		.fini = _nouveau_gpio_fini,
 	},
 	.lines = 32,
-	.intr_stat = nv92_gpio_intr_stat,
-	.intr_mask = nv92_gpio_intr_mask,
+	.intr_stat = nv94_gpio_intr_stat,
+	.intr_mask = nv94_gpio_intr_mask,
 	.drive = nv50_gpio_drive,
 	.sense = nv50_gpio_sense,
 	.reset = nv50_gpio_reset,
diff --git a/drivers/gpu/drm/nouveau/core/subdev/gpio/nvd0.c b/drivers/gpu/drm/nouveau/core/subdev/gpio/nvd0.c
index a4682b0956ad..480d6d2af770 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/gpio/nvd0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/gpio/nvd0.c
@@ -77,8 +77,8 @@ nvd0_gpio_oclass = &(struct nouveau_gpio_impl) {
 		.fini = _nouveau_gpio_fini,
 	},
 	.lines = 32,
-	.intr_stat = nv92_gpio_intr_stat,
-	.intr_mask = nv92_gpio_intr_mask,
+	.intr_stat = nv94_gpio_intr_stat,
+	.intr_mask = nv94_gpio_intr_mask,
 	.drive = nvd0_gpio_drive,
 	.sense = nvd0_gpio_sense,
 	.reset = nvd0_gpio_reset,
diff --git a/drivers/gpu/drm/nouveau/core/subdev/gpio/priv.h b/drivers/gpu/drm/nouveau/core/subdev/gpio/priv.h
index e1724dfc86ae..bff98b86e2b5 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/gpio/priv.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/gpio/priv.h
@@ -56,8 +56,8 @@ void nv50_gpio_reset(struct nouveau_gpio *, u8);
 int  nv50_gpio_drive(struct nouveau_gpio *, int, int, int);
 int  nv50_gpio_sense(struct nouveau_gpio *, int);
 
-void nv92_gpio_intr_stat(struct nouveau_gpio *, u32 *, u32 *);
-void nv92_gpio_intr_mask(struct nouveau_gpio *, u32, u32, u32);
+void nv94_gpio_intr_stat(struct nouveau_gpio *, u32 *, u32 *);
+void nv94_gpio_intr_mask(struct nouveau_gpio *, u32, u32, u32);
 
 void nvd0_gpio_reset(struct nouveau_gpio *, u8);
 int  nvd0_gpio_drive(struct nouveau_gpio *, int, int, int);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/base.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/base.c
index a652cafde3d6..0dc605db7ec8 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/base.c
@@ -23,6 +23,7 @@
  */
 
 #include <core/option.h>
+#include <core/object.h>
 #include <core/event.h>
 
 #include <subdev/bios.h>
@@ -346,7 +347,8 @@ nouveau_i2c_intr_init(struct nvkm_event *event, int type, int index)
 }
 
 static int
-nouveau_i2c_intr_ctor(void *data, u32 size, struct nvkm_notify *notify)
+nouveau_i2c_intr_ctor(struct nouveau_object *object, void *data, u32 size,
+		      struct nvkm_notify *notify)
 {
 	struct nvkm_i2c_ntfy_req *req = data;
 	if (!WARN_ON(size != sizeof(*req))) {
@@ -471,18 +473,56 @@ nouveau_i2c_extdev_sclass[] = {
 	nouveau_anx9805_sclass,
 };
 
+static void
+nouveau_i2c_create_port(struct nouveau_i2c *i2c, int index, u8 type,
+			struct dcb_i2c_entry *info)
+{
+	const struct nouveau_i2c_impl *impl = (void *)nv_oclass(i2c);
+	struct nouveau_oclass *oclass;
+	struct nouveau_object *parent;
+	struct nouveau_object *object;
+	int ret, pad;
+
+	if (info->share != DCB_I2C_UNUSED) {
+		pad    = info->share;
+		oclass = impl->pad_s;
+	} else {
+		if (type != DCB_I2C_NVIO_AUX)
+			pad = 0x100 + info->drive;
+		else
+			pad = 0x100 + info->auxch;
+		oclass = impl->pad_x;
+	}
+
+	ret = nouveau_object_ctor(NULL, nv_object(i2c), oclass, NULL, pad,
+				 &parent);
+	if (ret < 0)
+		return;
+
+	oclass = impl->sclass;
+	do {
+		ret = -EINVAL;
+		if (oclass->handle == type) {
+			ret = nouveau_object_ctor(parent, nv_object(i2c),
+						  oclass, info, index,
+						 &object);
+		}
+	} while (ret && (++oclass)->handle);
+
+	nouveau_object_ref(NULL, &parent);
+}
+
 int
 nouveau_i2c_create_(struct nouveau_object *parent,
 		    struct nouveau_object *engine,
 		    struct nouveau_oclass *oclass,
 		    int length, void **pobject)
 {
-	const struct nouveau_i2c_impl *impl = (void *)oclass;
 	struct nouveau_bios *bios = nouveau_bios(parent);
 	struct nouveau_i2c *i2c;
 	struct nouveau_object *object;
 	struct dcb_i2c_entry info;
-	int ret, i, j, index = -1, pad;
+	int ret, i, j, index = -1;
 	struct dcb_output outp;
 	u8  ver, hdr;
 	u32 data;
@@ -505,43 +545,40 @@ nouveau_i2c_create_(struct nouveau_object *parent,
 	INIT_LIST_HEAD(&i2c->ports);
 
 	while (!dcb_i2c_parse(bios, ++index, &info)) {
-		if (info.type == DCB_I2C_UNUSED)
+		switch (info.type) {
+		case DCB_I2C_NV04_BIT:
+		case DCB_I2C_NV4E_BIT:
+		case DCB_I2C_NVIO_BIT:
+			nouveau_i2c_create_port(i2c, NV_I2C_PORT(index),
+						info.type, &info);
+			break;
+		case DCB_I2C_NVIO_AUX:
+			nouveau_i2c_create_port(i2c, NV_I2C_AUX(index),
+						info.type, &info);
+			break;
+		case DCB_I2C_PMGR:
+			if (info.drive != DCB_I2C_UNUSED) {
+				nouveau_i2c_create_port(i2c, NV_I2C_PORT(index),
+							DCB_I2C_NVIO_BIT,
+							&info);
+			}
+			if (info.auxch != DCB_I2C_UNUSED) {
+				nouveau_i2c_create_port(i2c, NV_I2C_AUX(index),
+							DCB_I2C_NVIO_AUX,
+							&info);
+			}
+			break;
+		case DCB_I2C_UNUSED:
+		default:
 			continue;
-
-		if (info.share != DCB_I2C_UNUSED) {
-			if (info.type == DCB_I2C_NVIO_AUX)
-				pad = info.drive;
-			else
-				pad = info.share;
-			oclass = impl->pad_s;
-		} else {
-			pad = 0x100 + info.drive;
-			oclass = impl->pad_x;
 		}
-
-		ret = nouveau_object_ctor(NULL, *pobject, oclass,
-					  NULL, pad, &parent);
-		if (ret < 0)
-			continue;
-
-		oclass = impl->sclass;
-		do {
-			ret = -EINVAL;
-			if (oclass->handle == info.type) {
-				ret = nouveau_object_ctor(parent, *pobject,
-							  oclass, &info,
-							  index, &object);
-			}
-		} while (ret && (++oclass)->handle);
-
-		nouveau_object_ref(NULL, &parent);
 	}
 
 	/* in addition to the busses specified in the i2c table, there
 	 * may be ddc/aux channels hiding behind external tmds/dp/etc
 	 * transmitters.
 	 */
-	index = ((index + 0x0f) / 0x10) * 0x10;
+	index = NV_I2C_EXT(0);
 	i = -1;
 	while ((data = dcb_outp_parse(bios, ++i, &ver, &hdr, &outp))) {
 		if (!outp.location || !outp.extdev)
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/gm204.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/gm204.c
new file mode 100644
index 000000000000..06a2b87ccbf1
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/gm204.c
@@ -0,0 +1,221 @@
+/*
+ * Copyright 2012 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "nv50.h"
+
+#define AUX_DBG(fmt, args...) nv_debug(aux, "AUXCH(%d): " fmt, ch, ##args)
+#define AUX_ERR(fmt, args...) nv_error(aux, "AUXCH(%d): " fmt, ch, ##args)
+
+static void
+auxch_fini(struct nouveau_i2c *aux, int ch)
+{
+	nv_mask(aux, 0x00d954 + (ch * 0x50), 0x00310000, 0x00000000);
+}
+
+static int
+auxch_init(struct nouveau_i2c *aux, int ch)
+{
+	const u32 unksel = 1; /* nfi which to use, or if it matters.. */
+	const u32 ureq = unksel ? 0x00100000 : 0x00200000;
+	const u32 urep = unksel ? 0x01000000 : 0x02000000;
+	u32 ctrl, timeout;
+
+	/* wait up to 1ms for any previous transaction to be done... */
+	timeout = 1000;
+	do {
+		ctrl = nv_rd32(aux, 0x00d954 + (ch * 0x50));
+		udelay(1);
+		if (!timeout--) {
+			AUX_ERR("begin idle timeout 0x%08x\n", ctrl);
+			return -EBUSY;
+		}
+	} while (ctrl & 0x03010000);
+
+	/* set some magic, and wait up to 1ms for it to appear */
+	nv_mask(aux, 0x00d954 + (ch * 0x50), 0x00300000, ureq);
+	timeout = 1000;
+	do {
+		ctrl = nv_rd32(aux, 0x00d954 + (ch * 0x50));
+		udelay(1);
+		if (!timeout--) {
+			AUX_ERR("magic wait 0x%08x\n", ctrl);
+			auxch_fini(aux, ch);
+			return -EBUSY;
+		}
+	} while ((ctrl & 0x03000000) != urep);
+
+	return 0;
+}
+
+int
+gm204_aux(struct nouveau_i2c_port *base, bool retry,
+	 u8 type, u32 addr, u8 *data, u8 size)
+{
+	struct nouveau_i2c *aux = nouveau_i2c(base);
+	struct nv50_i2c_port *port = (void *)base;
+	u32 ctrl, stat, timeout, retries;
+	u32 xbuf[4] = {};
+	int ch = port->addr;
+	int ret, i;
+
+	AUX_DBG("%d: 0x%08x %d\n", type, addr, size);
+
+	ret = auxch_init(aux, ch);
+	if (ret)
+		goto out;
+
+	stat = nv_rd32(aux, 0x00d958 + (ch * 0x50));
+	if (!(stat & 0x10000000)) {
+		AUX_DBG("sink not detected\n");
+		ret = -ENXIO;
+		goto out;
+	}
+
+	if (!(type & 1)) {
+		memcpy(xbuf, data, size);
+		for (i = 0; i < 16; i += 4) {
+			AUX_DBG("wr 0x%08x\n", xbuf[i / 4]);
+			nv_wr32(aux, 0x00d930 + (ch * 0x50) + i, xbuf[i / 4]);
+		}
+	}
+
+	ctrl  = nv_rd32(aux, 0x00d954 + (ch * 0x50));
+	ctrl &= ~0x0001f0ff;
+	ctrl |= type << 12;
+	ctrl |= size - 1;
+	nv_wr32(aux, 0x00d950 + (ch * 0x50), addr);
+
+	/* (maybe) retry transaction a number of times on failure... */
+	for (retries = 0; !ret && retries < 32; retries++) {
+		/* reset, and delay a while if this is a retry */
+		nv_wr32(aux, 0x00d954 + (ch * 0x50), 0x80000000 | ctrl);
+		nv_wr32(aux, 0x00d954 + (ch * 0x50), 0x00000000 | ctrl);
+		if (retries)
+			udelay(400);
+
+		/* transaction request, wait up to 1ms for it to complete */
+		nv_wr32(aux, 0x00d954 + (ch * 0x50), 0x00010000 | ctrl);
+
+		timeout = 1000;
+		do {
+			ctrl = nv_rd32(aux, 0x00d954 + (ch * 0x50));
+			udelay(1);
+			if (!timeout--) {
+				AUX_ERR("tx req timeout 0x%08x\n", ctrl);
+				ret = -EIO;
+				goto out;
+			}
+		} while (ctrl & 0x00010000);
+		ret = 1;
+
+		/* read status, and check if transaction completed ok */
+		stat = nv_mask(aux, 0x00d958 + (ch * 0x50), 0, 0);
+		if ((stat & 0x000f0000) == 0x00080000 ||
+		    (stat & 0x000f0000) == 0x00020000)
+			ret = retry ? 0 : 1;
+		if ((stat & 0x00000100))
+			ret = -ETIMEDOUT;
+		if ((stat & 0x00000e00))
+			ret = -EIO;
+
+		AUX_DBG("%02d 0x%08x 0x%08x\n", retries, ctrl, stat);
+	}
+
+	if (type & 1) {
+		for (i = 0; i < 16; i += 4) {
+			xbuf[i / 4] = nv_rd32(aux, 0x00d940 + (ch * 0x50) + i);
+			AUX_DBG("rd 0x%08x\n", xbuf[i / 4]);
+		}
+		memcpy(data, xbuf, size);
+	}
+
+out:
+	auxch_fini(aux, ch);
+	return ret < 0 ? ret : (stat & 0x000f0000) >> 16;
+}
+
+static const struct nouveau_i2c_func
+gm204_aux_func = {
+	.aux       = gm204_aux,
+};
+
+int
+gm204_aux_port_ctor(struct nouveau_object *parent,
+		    struct nouveau_object *engine,
+		    struct nouveau_oclass *oclass, void *data, u32 index,
+		    struct nouveau_object **pobject)
+{
+	struct dcb_i2c_entry *info = data;
+	struct nv50_i2c_port *port;
+	int ret;
+
+	ret = nouveau_i2c_port_create(parent, engine, oclass, index,
+				      &nouveau_i2c_aux_algo, &gm204_aux_func,
+				      &port);
+	*pobject = nv_object(port);
+	if (ret)
+		return ret;
+
+	port->base.aux = info->auxch;
+	port->addr = info->auxch;
+	return 0;
+}
+
+struct nouveau_oclass
+gm204_i2c_sclass[] = {
+	{ .handle = NV_I2C_TYPE_DCBI2C(DCB_I2C_NVIO_BIT),
+	  .ofuncs = &(struct nouveau_ofuncs) {
+		  .ctor = nvd0_i2c_port_ctor,
+		  .dtor = _nouveau_i2c_port_dtor,
+		  .init = nv50_i2c_port_init,
+		  .fini = _nouveau_i2c_port_fini,
+	  },
+	},
+	{ .handle = NV_I2C_TYPE_DCBI2C(DCB_I2C_NVIO_AUX),
+	  .ofuncs = &(struct nouveau_ofuncs) {
+		  .ctor = gm204_aux_port_ctor,
+		  .dtor = _nouveau_i2c_port_dtor,
+		  .init = _nouveau_i2c_port_init,
+		  .fini = _nouveau_i2c_port_fini,
+	  },
+	},
+	{}
+};
+
+struct nouveau_oclass *
+gm204_i2c_oclass = &(struct nouveau_i2c_impl) {
+	.base.handle = NV_SUBDEV(I2C, 0x24),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = _nouveau_i2c_ctor,
+		.dtor = _nouveau_i2c_dtor,
+		.init = _nouveau_i2c_init,
+		.fini = _nouveau_i2c_fini,
+	},
+	.sclass = gm204_i2c_sclass,
+	.pad_x = &nv04_i2c_pad_oclass,
+	.pad_s = &gm204_i2c_pad_oclass,
+	.aux = 8,
+	.aux_stat = nve0_aux_stat,
+	.aux_mask = nve0_aux_mask,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/nv50.h b/drivers/gpu/drm/nouveau/core/subdev/i2c/nv50.h
index 5d2a77421c74..9ef965692fb1 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/nv50.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/nv50.h
@@ -10,8 +10,6 @@ struct nv50_i2c_priv {
 struct nv50_i2c_port {
 	struct nouveau_i2c_port base;
 	u32 addr;
-	u32 ctrl;
-	u32 data;
 	u32 state;
 };
 
@@ -29,4 +27,8 @@ int  nv94_aux_port_ctor(struct nouveau_object *, struct nouveau_object *,
 void nv94_i2c_acquire(struct nouveau_i2c_port *);
 void nv94_i2c_release(struct nouveau_i2c_port *);
 
+int  nvd0_i2c_port_ctor(struct nouveau_object *, struct nouveau_object *,
+			struct nouveau_oclass *, void *, u32,
+			struct nouveau_object **);
+
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/nv94.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/nv94.c
index f59c3a255462..e383ee81f4d2 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/nv94.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/nv94.c
@@ -214,10 +214,6 @@ nv94_i2c_port_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 
 	port->state = 7;
 	port->addr = nv50_i2c_addr[info->drive];
-	if (info->share != DCB_I2C_UNUSED) {
-		port->ctrl = 0x00e500 + (info->share * 0x50);
-		port->data = 0x0000e001;
-	}
 	return 0;
 }
 
@@ -242,13 +238,8 @@ nv94_aux_port_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	port->base.aux = info->drive;
-	port->addr = info->drive;
-	if (info->share != DCB_I2C_UNUSED) {
-		port->ctrl = 0x00e500 + (info->drive * 0x50);
-		port->data = 0x00002002;
-	}
-
+	port->base.aux = info->auxch;
+	port->addr = info->auxch;
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/nvd0.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/nvd0.c
index 364ddb1c5f03..fd99380502ec 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/nvd0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/nvd0.c
@@ -48,7 +48,7 @@ nvd0_i2c_func = {
 	.sense_sda = nvd0_i2c_sense_sda,
 };
 
-static int
+int
 nvd0_i2c_port_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		   struct nouveau_oclass *oclass, void *data, u32 index,
 		   struct nouveau_object **pobject)
@@ -66,10 +66,6 @@ nvd0_i2c_port_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 
 	port->state = 0x00000007;
 	port->addr = 0x00d014 + (info->drive * 0x20);
-	if (info->share != DCB_I2C_UNUSED) {
-		port->ctrl = 0x00e500 + (info->share * 0x50);
-		port->data = 0x0000e001;
-	}
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/nve0.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/nve0.c
index cae77e1ad8dc..25fe5c2d110e 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/nve0.c
@@ -24,7 +24,7 @@
 
 #include "nv50.h"
 
-static void
+void
 nve0_aux_stat(struct nouveau_i2c *i2c, u32 *hi, u32 *lo, u32 *rq, u32 *tx)
 {
 	u32 intr = nv_rd32(i2c, 0x00dc60);
@@ -38,7 +38,7 @@ nve0_aux_stat(struct nouveau_i2c *i2c, u32 *hi, u32 *lo, u32 *rq, u32 *tx)
 	nv_wr32(i2c, 0x00dc60, intr);
 }
 
-static void
+void
 nve0_aux_mask(struct nouveau_i2c *i2c, u32 type, u32 mask, u32 data)
 {
 	u32 temp = nv_rd32(i2c, 0x00dc68), i;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/padgm204.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/padgm204.c
new file mode 100644
index 000000000000..f0e6fbbaa8cd
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/padgm204.c
@@ -0,0 +1,86 @@
+/*
+ * Copyright 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "pad.h"
+
+struct gm204_i2c_pad {
+	struct nvkm_i2c_pad base;
+	int addr;
+};
+
+static int
+gm204_i2c_pad_fini(struct nouveau_object *object, bool suspend)
+{
+	struct nouveau_i2c *i2c = (void *)object->engine;
+	struct gm204_i2c_pad *pad = (void *)object;
+	nv_mask(i2c, 0x00d97c + pad->addr, 0x00000001, 0x00000001);
+	return nvkm_i2c_pad_fini(&pad->base, suspend);
+}
+
+static int
+gm204_i2c_pad_init(struct nouveau_object *object)
+{
+	struct nouveau_i2c *i2c = (void *)object->engine;
+	struct gm204_i2c_pad *pad = (void *)object;
+
+	switch (nv_oclass(pad->base.next)->handle) {
+	case NV_I2C_TYPE_DCBI2C(DCB_I2C_NVIO_AUX):
+		nv_mask(i2c, 0x00d970 + pad->addr, 0x0000c003, 0x00000002);
+		break;
+	case NV_I2C_TYPE_DCBI2C(DCB_I2C_NVIO_BIT):
+	default:
+		nv_mask(i2c, 0x00d970 + pad->addr, 0x0000c003, 0x0000c001);
+		break;
+	}
+
+	nv_mask(i2c, 0x00d97c + pad->addr, 0x00000001, 0x00000000);
+	return nvkm_i2c_pad_init(&pad->base);
+}
+
+static int
+gm204_i2c_pad_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+		  struct nouveau_oclass *oclass, void *data, u32 index,
+		  struct nouveau_object **pobject)
+{
+	struct gm204_i2c_pad *pad;
+	int ret;
+
+	ret = nvkm_i2c_pad_create(parent, engine, oclass, index, &pad);
+	*pobject = nv_object(pad);
+	if (ret)
+		return ret;
+
+	pad->addr = index * 0x50;;
+	return 0;
+}
+
+struct nouveau_oclass
+gm204_i2c_pad_oclass = {
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = gm204_i2c_pad_ctor,
+		.dtor = _nvkm_i2c_pad_dtor,
+		.init = gm204_i2c_pad_init,
+		.fini = gm204_i2c_pad_fini,
+	},
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/priv.h b/drivers/gpu/drm/nouveau/core/subdev/i2c/priv.h
index 780090b6425a..4fe7ae3fde4e 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/priv.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/priv.h
@@ -5,6 +5,7 @@
 
 extern struct nouveau_oclass nv04_i2c_pad_oclass;
 extern struct nouveau_oclass nv94_i2c_pad_oclass;
+extern struct nouveau_oclass gm204_i2c_pad_oclass;
 
 #define nouveau_i2c_port_create(p,e,o,i,a,f,d)                                 \
 	nouveau_i2c_port_create_((p), (e), (o), (i), (a), (f),                 \
@@ -82,4 +83,7 @@ struct nouveau_i2c_impl {
 void nv94_aux_stat(struct nouveau_i2c *, u32 *, u32 *, u32 *, u32 *);
 void nv94_aux_mask(struct nouveau_i2c *, u32, u32, u32);
 
+void nve0_aux_stat(struct nouveau_i2c *, u32 *, u32 *, u32 *, u32 *);
+void nve0_aux_mask(struct nouveau_i2c *, u32, u32, u32);
+
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.c b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.c
index 7b64befee48f..e8b1401c59c0 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.c
@@ -69,7 +69,7 @@ nv04_instobj_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	ret = nouveau_mm_head(&priv->heap, 1, args->size, args->size,
+	ret = nouveau_mm_head(&priv->heap, 0, 1, args->size, args->size,
 			      args->align, &node->mem);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/ltc/base.c b/drivers/gpu/drm/nouveau/core/subdev/ltc/base.c
index 32ed442c5913..7fa331516f84 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/ltc/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/ltc/base.c
@@ -31,7 +31,7 @@ nvkm_ltc_tags_alloc(struct nouveau_ltc *ltc, u32 n,
 	struct nvkm_ltc_priv *priv = (void *)ltc;
 	int ret;
 
-	ret = nouveau_mm_head(&priv->tags, 1, n, n, 1, pnode);
+	ret = nouveau_mm_head(&priv->tags, 0, 1, n, n, 1, pnode);
 	if (ret)
 		*pnode = NULL;
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/ltc/gf100.c b/drivers/gpu/drm/nouveau/core/subdev/ltc/gf100.c
index d5d65285efe5..2db0977284f8 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/ltc/gf100.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/ltc/gf100.c
@@ -62,16 +62,38 @@ gf100_ltc_zbc_clear_depth(struct nvkm_ltc_priv *priv, int i, const u32 depth)
 	nv_wr32(priv, 0x17ea58, depth);
 }
 
+static const struct nouveau_bitfield
+gf100_ltc_lts_intr_name[] = {
+	{ 0x00000001, "IDLE_ERROR_IQ" },
+	{ 0x00000002, "IDLE_ERROR_CBC" },
+	{ 0x00000004, "IDLE_ERROR_TSTG" },
+	{ 0x00000008, "IDLE_ERROR_DSTG" },
+	{ 0x00000010, "EVICTED_CB" },
+	{ 0x00000020, "ILLEGAL_COMPSTAT" },
+	{ 0x00000040, "BLOCKLINEAR_CB" },
+	{ 0x00000100, "ECC_SEC_ERROR" },
+	{ 0x00000200, "ECC_DED_ERROR" },
+	{ 0x00000400, "DEBUG" },
+	{ 0x00000800, "ATOMIC_TO_Z" },
+	{ 0x00001000, "ILLEGAL_ATOMIC" },
+	{ 0x00002000, "BLKACTIVITY_ERR" },
+	{}
+};
+
 static void
-gf100_ltc_lts_isr(struct nvkm_ltc_priv *priv, int ltc, int lts)
+gf100_ltc_lts_intr(struct nvkm_ltc_priv *priv, int ltc, int lts)
 {
 	u32 base = 0x141000 + (ltc * 0x2000) + (lts * 0x400);
-	u32 stat = nv_rd32(priv, base + 0x020);
+	u32 intr = nv_rd32(priv, base + 0x020);
+	u32 stat = intr & 0x0000ffff;
 
 	if (stat) {
-		nv_info(priv, "LTC%d_LTS%d: 0x%08x\n", ltc, lts, stat);
-		nv_wr32(priv, base + 0x020, stat);
+		nv_info(priv, "LTC%d_LTS%d:", ltc, lts);
+		nouveau_bitfield_print(gf100_ltc_lts_intr_name, stat);
+		pr_cont("\n");
 	}
+
+	nv_wr32(priv, base + 0x020, intr);
 }
 
 void
@@ -84,14 +106,9 @@ gf100_ltc_intr(struct nouveau_subdev *subdev)
 	while (mask) {
 		u32 lts, ltc = __ffs(mask);
 		for (lts = 0; lts < priv->lts_nr; lts++)
-			gf100_ltc_lts_isr(priv, ltc, lts);
+			gf100_ltc_lts_intr(priv, ltc, lts);
 		mask &= ~(1 << ltc);
 	}
-
-	/* we do something horribly wrong and upset PMFB a lot, so mask off
-	 * interrupts from it after the first one until it's fixed
-	 */
-	nv_mask(priv, 0x000640, 0x02000000, 0x00000000);
 }
 
 static int
@@ -153,7 +170,7 @@ gf100_ltc_init_tag_ram(struct nouveau_fb *pfb, struct nvkm_ltc_priv *priv)
 	tag_size += tag_align;
 	tag_size  = (tag_size + 0xfff) >> 12; /* round up */
 
-	ret = nouveau_mm_tail(&pfb->vram, 1, tag_size, tag_size, 1,
+	ret = nouveau_mm_tail(&pfb->vram, 1, 1, tag_size, tag_size, 1,
 	                      &priv->tag_ram);
 	if (ret) {
 		priv->num_tags = 0;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/ltc/gm107.c b/drivers/gpu/drm/nouveau/core/subdev/ltc/gm107.c
index a4de64289762..89fc4238f50c 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/ltc/gm107.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/ltc/gm107.c
@@ -87,11 +87,6 @@ gm107_ltc_intr(struct nouveau_subdev *subdev)
 			gm107_ltc_lts_isr(priv, ltc, lts);
 		mask &= ~(1 << ltc);
 	}
-
-	/* we do something horribly wrong and upset PMFB a lot, so mask off
-	 * interrupts from it after the first one until it's fixed
-	 */
-	nv_mask(priv, 0x000640, 0x02000000, 0x00000000);
 }
 
 static int
diff --git a/drivers/gpu/drm/nouveau/core/subdev/ltc/priv.h b/drivers/gpu/drm/nouveau/core/subdev/ltc/priv.h
index 594924f39126..41f179d93da6 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/ltc/priv.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/ltc/priv.h
@@ -4,6 +4,8 @@
 #include <subdev/ltc.h>
 #include <subdev/fb.h>
 
+#include <core/enum.h>
+
 struct nvkm_ltc_priv {
 	struct nouveau_ltc base;
 	u32 ltc_nr;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/mc/nv4c.c b/drivers/gpu/drm/nouveau/core/subdev/mc/nv4c.c
index a75c35ccf25c..165401c4045c 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/mc/nv4c.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/mc/nv4c.c
@@ -24,13 +24,6 @@
 
 #include "nv04.h"
 
-static void
-nv4c_mc_msi_rearm(struct nouveau_mc *pmc)
-{
-	struct nv04_mc_priv *priv = (void *)pmc;
-	nv_wr08(priv, 0x088050, 0xff);
-}
-
 struct nouveau_oclass *
 nv4c_mc_oclass = &(struct nouveau_mc_oclass) {
 	.base.handle = NV_SUBDEV(MC, 0x4c),
@@ -41,5 +34,4 @@ nv4c_mc_oclass = &(struct nouveau_mc_oclass) {
 		.fini = _nouveau_mc_fini,
 	},
 	.intr = nv04_mc_intr,
-	.msi_rearm = nv4c_mc_msi_rearm,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/base.c b/drivers/gpu/drm/nouveau/core/subdev/pwr/base.c
index 69f1f34f6931..0ab55f27ec45 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/base.c
@@ -203,6 +203,8 @@ _nouveau_pwr_init(struct nouveau_object *object)
 	nv_wait(ppwr, 0x10a04c, 0xffffffff, 0x00000000);
 	nv_mask(ppwr, 0x000200, 0x00002000, 0x00000000);
 	nv_mask(ppwr, 0x000200, 0x00002000, 0x00002000);
+	nv_rd32(ppwr, 0x000200);
+	nv_wait(ppwr, 0x10a10c, 0x00000006, 0x00000000);
 
 	/* upload data segment */
 	nv_wr32(ppwr, 0x10a1c0, 0x01000000);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/arith.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/arith.fuc
new file mode 100644
index 000000000000..214a6d9e088d
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/arith.fuc
@@ -0,0 +1,94 @@
+/*
+ * Copyright 2014 Martin Peres <martin.peres@free.fr>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the folloing conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Martin Peres
+ */
+
+/******************************************************************************
+ * arith data segment
+ *****************************************************************************/
+#ifdef INCLUDE_PROC
+#endif
+
+#ifdef INCLUDE_DATA
+#endif
+
+/******************************************************************************
+ * arith code segment
+ *****************************************************************************/
+#ifdef INCLUDE_CODE
+
+// does a 32x32 -> 64 multiplication
+//
+// A * B = A_lo * B_lo
+//        + ( A_hi * B_lo ) << 16
+//        + ( A_lo * B_hi ) << 16
+//        + ( A_hi * B_hi ) << 32
+//
+// $r15 - current
+// $r14 - A
+// $r13 - B
+// $r12 - mul_lo (return)
+// $r11 - mul_hi (return)
+// $r0  - zero
+mulu32_32_64:
+	push $r1 // A_hi
+	push $r2 // B_hi
+	push $r3 // tmp0
+	push $r4 // tmp1
+
+	shr b32 $r1 $r14 16
+	shr b32 $r2 $r13 16
+
+	clear b32 $r12
+	clear b32 $r11
+
+	// A_lo * B_lo
+	mulu $r12 $r14 $r13
+
+	// ( A_hi * B_lo ) << 16
+	mulu $r3 $r1 $r13 // tmp0 = A_hi * B_lo
+	mov b32 $r4 $r3
+	and $r3 0xffff // tmp0 = tmp0_lo
+	shl b32 $r3 16
+	shr b32 $r4 16 // tmp1 = tmp0_hi
+	add b32 $r12 $r3
+	adc b32 $r11 $r4
+
+	// ( A_lo * B_hi ) << 16
+	mulu $r3 $r14 $r2 // tmp0 = A_lo * B_hi
+	mov b32 $r4 $r3
+	and $r3 0xffff // tmp0 = tmp0_lo
+	shl b32 $r3 16
+	shr b32 $r4 16 // tmp1 = tmp0_hi
+	add b32 $r12 $r3
+	adc b32 $r11 $r4
+
+	// ( A_hi * B_hi ) << 32
+	mulu $r3 $r1 $r2 // tmp0 = A_hi * B_hi
+	add b32 $r11 $r3
+
+	pop $r4
+	pop $r3
+	pop $r2
+	pop $r1
+	ret
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/kernel.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/kernel.fuc
index 8f29badd785f..5cf5be63cbef 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/kernel.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/kernel.fuc
@@ -98,12 +98,16 @@ wr32:
 // $r14 - ns
 // $r0  - zero
 nsec:
+	push $r9
+	push $r8
 	nv_iord($r8, NV_PPWR_TIMER_LOW)
 	nsec_loop:
 		nv_iord($r9, NV_PPWR_TIMER_LOW)
 		sub b32 $r9 $r8
 		cmp b32 $r9 $r14
 		bra l #nsec_loop
+	pop $r8
+	pop $r9
 	ret
 
 // busy-wait for a period of time
@@ -115,6 +119,8 @@ nsec:
 // $r11 - timeout (ns)
 // $r0  - zero
 wait:
+	push $r9
+	push $r8
 	nv_iord($r8, NV_PPWR_TIMER_LOW)
 	wait_loop:
 		nv_rd32($r10, $r14)
@@ -126,6 +132,8 @@ wait:
 		cmp b32 $r9 $r11
 		bra l #wait_loop
 	wait_done:
+	pop $r8
+	pop $r9
 	ret
 
 // $r15 - current (kern)
@@ -242,12 +250,89 @@ intr:
 	bclr $flags $p0
 	iret
 
-// request the current process be sent a message after a timeout expires
+// calculate the number of ticks in the specified nanoseconds delay
+//
+// $r15 - current
+// $r14 - ns
+// $r14 - ticks (return)
+// $r0  - zero
+ticks_from_ns:
+	push $r12
+	push $r11
+
+	/* try not losing precision (multiply then divide) */
+	imm32($r13, HW_TICKS_PER_US)
+	call #mulu32_32_64
+
+	/* use an immeditate, it's ok because HW_TICKS_PER_US < 16 bits */
+	div $r12 $r12 1000
+
+	/* check if there wasn't any overflow */
+	cmpu b32 $r11 0
+	bra e #ticks_from_ns_quit
+
+	/* let's divide then multiply, too bad for the precision! */
+	div $r14 $r14 1000
+	imm32($r13, HW_TICKS_PER_US)
+	call #mulu32_32_64
+
+	/* this cannot overflow as long as HW_TICKS_PER_US < 1000 */
+
+ticks_from_ns_quit:
+	mov b32 $r14 $r12
+	pop $r11
+	pop $r12
+	ret
+
+// calculate the number of ticks in the specified microsecond delay
+//
+// $r15 - current
+// $r14 - us
+// $r14 - ticks (return)
+// $r0  - zero
+ticks_from_us:
+	push $r12
+	push $r11
+
+	/* simply multiply $us by HW_TICKS_PER_US */
+	imm32($r13, HW_TICKS_PER_US)
+	call #mulu32_32_64
+	mov b32 $r14 $r12
+
+	/* check if there wasn't any overflow */
+	cmpu b32 $r11 0
+	bra e #ticks_from_us_quit
+
+	/* Overflow! */
+	clear b32 $r14
+
+ticks_from_us_quit:
+	pop $r11
+	pop $r12
+	ret
+
+// calculate the number of ticks in the specified microsecond delay
 //
 // $r15 - current
 // $r14 - ticks
+// $r14 - us (return)
+// $r0  - zero
+ticks_to_us:
+	/* simply divide $ticks by HW_TICKS_PER_US */
+	imm32($r13, HW_TICKS_PER_US)
+	div $r14 $r14 $r13
+
+	ret
+
+// request the current process be sent a message after a timeout expires
+//
+// $r15 - current
+// $r14 - ticks (make sure it is < 2^31 to avoid any possible overflow)
 // $r0  - zero
 timer:
+	push $r9
+	push $r8
+
 	// interrupts off to prevent racing with timer isr
 	bclr $flags ie0
 
@@ -255,13 +340,22 @@ timer:
 	ld b32 $r8 D[$r15 + #proc_time]
 	cmp b32 $r8 0
 	bra g #timer_done
-	st b32 D[$r15 + #proc_time] $r14
 
-	// halt watchdog timer temporarily and check for a pending
-	// interrupt.  if there's one already pending, we can just
-	// bail since the timer isr will queue the next soonest
-	// right after it's done
+	// halt watchdog timer temporarily
+	clear b32 $r8
 	nv_iowr(NV_PPWR_WATCHDOG_ENABLE, $r8)
+
+	// find out how much time elapsed since the last update
+	// of the watchdog and add this time to the wanted ticks
+	nv_iord($r8, NV_PPWR_WATCHDOG_TIME)
+	ld b32 $r9 D[$r0 + #time_prev]
+	sub b32 $r9 $r8
+	add b32 $r14 $r9
+	st b32 D[$r15 + #proc_time] $r14
+
+	// check for a pending interrupt.  if there's one already
+	// pending, we can just bail since the timer isr will
+	// queue the next soonest right after it's done
 	nv_iord($r8, NV_PPWR_INTR)
 	and $r8 NV_PPWR_INTR_WATCHDOG
 	bra nz #timer_enable
@@ -272,10 +366,10 @@ timer:
 	cmp b32 $r14 $r0
 	bra e #timer_reset
 	cmp b32 $r14 $r8
-	bra l #timer_done
-	timer_reset:
-	nv_iowr(NV_PPWR_WATCHDOG_TIME, $r14)
-	st b32 D[$r0 + #time_prev] $r14
+	bra g #timer_enable
+		timer_reset:
+		nv_iowr(NV_PPWR_WATCHDOG_TIME, $r14)
+		st b32 D[$r0 + #time_prev] $r14
 
 	// re-enable the watchdog timer
 	timer_enable:
@@ -285,6 +379,9 @@ timer:
 	// interrupts back on
 	timer_done:
 	bset $flags ie0
+
+	pop $r8
+	pop $r9
 	ret
 
 // send message to another process
@@ -371,6 +468,9 @@ send:
 // $r14 - process
 // $r0  - zero
 recv:
+	push $r9
+	push $r8
+
 	ld b32 $r8 D[$r14 + #proc_qget]
 	ld b32 $r9 D[$r14 + #proc_qput]
 	bclr $flags $p1
@@ -403,6 +503,8 @@ recv:
 		bset $flags $p1
 		pop $r15
 	recv_done:
+	pop $r8
+	pop $r9
 	ret
 
 init:
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/macros.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/macros.fuc
index 5668e045bac1..96fc984dafdc 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/macros.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/macros.fuc
@@ -250,3 +250,23 @@
 */	st b32 D[$r0] reg /*
 */	clear b32 $r0
 #endif
+
+#define st(size, addr, reg) /*
+*/	movw $r0 addr /*
+*/	st size D[$r0] reg /*
+*/	clear b32 $r0
+
+#define ld(size, reg, addr) /*
+*/	movw $r0 addr /*
+*/	ld size reg D[$r0] /*
+*/	clear b32 $r0
+
+// does a 64+64 -> 64 unsigned addition (C = A + B)
+#define addu64(reg_a_c_hi, reg_a_c_lo, b_hi, b_lo) /*
+*/    add b32 reg_a_c_lo b_lo /*
+*/    adc b32 reg_a_c_hi b_hi
+
+// does a 64+64 -> 64 substraction (C = A - B)
+#define subu64(reg_a_c_hi, reg_a_c_lo, b_hi, b_lo) /*
+*/    sub b32 reg_a_c_lo b_lo /*
+*/    sbb b32 reg_a_c_hi b_hi
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/memx.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/memx.fuc
index d43741eccb11..ec03f9a4290b 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/memx.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/memx.fuc
@@ -43,20 +43,31 @@ process(PROC_MEMX, #memx_init, #memx_recv)
 */	.b32 func
 
 memx_func_head:
-handler(ENTER , 0x0001, 0x0000, #memx_func_enter)
+handler(ENTER , 0x0000, 0x0000, #memx_func_enter)
 memx_func_next:
 handler(LEAVE , 0x0000, 0x0000, #memx_func_leave)
 handler(WR32  , 0x0000, 0x0002, #memx_func_wr32)
 handler(WAIT  , 0x0004, 0x0000, #memx_func_wait)
 handler(DELAY , 0x0001, 0x0000, #memx_func_delay)
+handler(VBLANK, 0x0001, 0x0000, #memx_func_wait_vblank)
+handler(TRAIN , 0x0000, 0x0000, #memx_func_train)
 memx_func_tail:
 
 .equ #memx_func_size #memx_func_next - #memx_func_head
 .equ #memx_func_num (#memx_func_tail - #memx_func_head) / #memx_func_size
 
+memx_ts_start:
+.b32 0
+memx_ts_end:
+.b32 0
+
 memx_data_head:
 .skip 0x0800
 memx_data_tail:
+
+memx_train_head:
+.skip 0x0100
+memx_train_tail:
 #endif
 
 /******************************************************************************
@@ -67,19 +78,44 @@ memx_data_tail:
 //
 // $r15 - current (memx)
 // $r4  - packet length
-//	+00: bitmask of heads to wait for vblank on
 // $r3  - opcode desciption
 // $r0  - zero
 memx_func_enter:
+#if NVKM_PPWR_CHIPSET == GT215
+	movw $r8 0x1610
+	nv_rd32($r7, $r8)
+	imm32($r6, 0xfffffffc)
+	and $r7 $r6
+	movw $r6 0x2
+	or $r7 $r6
+	nv_wr32($r8, $r7)
+#else
+	movw $r6 0x001620
+	imm32($r7, ~0x00000aa2);
+	nv_rd32($r8, $r6)
+	and $r8 $r7
+	nv_wr32($r6, $r8)
+
+	imm32($r7, ~0x00000001)
+	nv_rd32($r8, $r6)
+	and $r8 $r7
+	nv_wr32($r6, $r8)
+
+	movw $r6 0x0026f0
+	nv_rd32($r8, $r6)
+	and $r8 $r7
+	nv_wr32($r6, $r8)
+#endif
+
 	mov $r6 NV_PPWR_OUTPUT_SET_FB_PAUSE
 	nv_iowr(NV_PPWR_OUTPUT_SET, $r6)
 	memx_func_enter_wait:
 		nv_iord($r6, NV_PPWR_OUTPUT)
 		and $r6 NV_PPWR_OUTPUT_FB_PAUSE
 		bra z #memx_func_enter_wait
-	//XXX: TODO
-	ld b32 $r6 D[$r1 + 0x00]
-	add b32 $r1 0x04
+
+	nv_iord($r6, NV_PPWR_TIMER_LOW)
+	st b32 D[$r0 + #memx_ts_start] $r6
 	ret
 
 // description
@@ -89,14 +125,93 @@ memx_func_enter:
 // $r3  - opcode desciption
 // $r0  - zero
 memx_func_leave:
+	nv_iord($r6, NV_PPWR_TIMER_LOW)
+	st b32 D[$r0 + #memx_ts_end] $r6
+
 	mov $r6 NV_PPWR_OUTPUT_CLR_FB_PAUSE
 	nv_iowr(NV_PPWR_OUTPUT_CLR, $r6)
 	memx_func_leave_wait:
 		nv_iord($r6, NV_PPWR_OUTPUT)
 		and $r6 NV_PPWR_OUTPUT_FB_PAUSE
 		bra nz #memx_func_leave_wait
+
+#if NVKM_PPWR_CHIPSET == GT215
+	movw $r8 0x1610
+	nv_rd32($r7, $r8)
+	imm32($r6, 0xffffffcc)
+	and $r7 $r6
+	nv_wr32($r8, $r7)
+#else
+	movw $r6 0x0026f0
+	imm32($r7, 0x00000001)
+	nv_rd32($r8, $r6)
+	or $r8 $r7
+	nv_wr32($r6, $r8)
+
+	movw $r6 0x001620
+	nv_rd32($r8, $r6)
+	or $r8 $r7
+	nv_wr32($r6, $r8)
+
+	imm32($r7, 0x00000aa2);
+	nv_rd32($r8, $r6)
+	or $r8 $r7
+	nv_wr32($r6, $r8)
+#endif
 	ret
 
+#if NVKM_PPWR_CHIPSET < GF119
+// description
+//
+// $r15 - current (memx)
+// $r4  - packet length
+//	+00: head to wait for vblank on
+// $r3  - opcode desciption
+// $r0  - zero
+memx_func_wait_vblank:
+	ld b32 $r6 D[$r1 + 0x00]
+	cmp b32 $r6 0x0
+	bra z #memx_func_wait_vblank_head0
+	cmp b32 $r6 0x1
+	bra z #memx_func_wait_vblank_head1
+	bra #memx_func_wait_vblank_fini
+
+	memx_func_wait_vblank_head1:
+	movw $r7 0x20
+	bra #memx_func_wait_vblank_0
+
+	memx_func_wait_vblank_head0:
+	movw $r7 0x8
+
+	memx_func_wait_vblank_0:
+		nv_iord($r6, NV_PPWR_INPUT)
+		and $r6 $r7
+		bra nz #memx_func_wait_vblank_0
+
+	memx_func_wait_vblank_1:
+		nv_iord($r6, NV_PPWR_INPUT)
+		and $r6 $r7
+		bra z #memx_func_wait_vblank_1
+
+	memx_func_wait_vblank_fini:
+	add b32 $r1 0x4
+	ret
+
+#else
+
+// XXX: currently no-op
+//
+// $r15 - current (memx)
+// $r4  - packet length
+//	+00: head to wait for vblank on
+// $r3  - opcode desciption
+// $r0  - zero
+memx_func_wait_vblank:
+	add b32 $r1 0x4
+	ret
+
+#endif
+
 // description
 //
 // $r15 - current (memx)
@@ -150,6 +265,101 @@ memx_func_delay:
 // description
 //
 // $r15 - current (memx)
+// $r4  - packet length
+// $r3  - opcode desciption
+// $r0  - zero
+memx_func_train:
+#if NVKM_PPWR_CHIPSET == GT215
+// $r5 - outer loop counter
+// $r6 - inner loop counter
+// $r7 - entry counter (#memx_train_head + $r7)
+	movw $r5 0x3
+	movw $r7 0x0
+
+// Read random memory to wake up... things
+	imm32($r9, 0x700000)
+	nv_rd32($r8,$r9)
+	movw $r14 0x2710
+	call(nsec)
+
+	memx_func_train_loop_outer:
+		mulu $r8 $r5 0x101
+		sethi $r8 0x02000000
+		imm32($r9, 0x1111e0)
+		nv_wr32($r9, $r8)
+		push $r5
+
+		movw $r6 0x0
+		memx_func_train_loop_inner:
+			movw $r8 0x1111
+			mulu $r9 $r6 $r8
+			shl b32 $r8 $r9 0x10
+			or $r8 $r9
+			imm32($r9, 0x100720)
+			nv_wr32($r9, $r8)
+
+			imm32($r9, 0x100080)
+			nv_rd32($r8, $r9)
+			or $r8 $r8 0x20
+			nv_wr32($r9, $r8)
+
+			imm32($r9, 0x10053c)
+			imm32($r8, 0x80003002)
+			nv_wr32($r9, $r8)
+
+			imm32($r14, 0x100560)
+			imm32($r13, 0x80000000)
+			add b32 $r12 $r13 0
+			imm32($r11, 0x001e8480)
+			call(wait)
+
+			// $r5 - inner inner loop counter
+			// $r9 - result
+			movw $r5 0
+			imm32($r9, 0x8300ffff)
+			memx_func_train_loop_4x:
+				imm32($r10, 0x100080)
+				nv_rd32($r8, $r10)
+				imm32($r11, 0xffffffdf)
+				and $r8 $r11
+				nv_wr32($r10, $r8)
+
+				imm32($r10, 0x10053c)
+				imm32($r8, 0x80003002)
+				nv_wr32($r10, $r8)
+
+				imm32($r14, 0x100560)
+				imm32($r13, 0x80000000)
+				mov b32 $r12 $r13
+				imm32($r11, 0x00002710)
+				call(wait)
+
+				nv_rd32($r13, $r14)
+				and $r9 $r9 $r13
+
+				add b32 $r5 1
+				cmp b16 $r5 0x4
+				bra l #memx_func_train_loop_4x
+
+			add b32 $r10 $r7 #memx_train_head
+			st b32 D[$r10 + 0] $r9
+			add b32 $r6 1
+			add b32 $r7 4
+
+			cmp b16 $r6 0x10
+			bra l #memx_func_train_loop_inner
+
+		pop $r5
+		add b32 $r5 1
+		cmp b16 $r5 7
+		bra l #memx_func_train_loop_outer
+
+#endif
+	ret
+
+// description
+//
+// $r15 - current (memx)
 // $r14 - sender process name
 // $r13 - message (exec)
 // $r12 - head of script
@@ -160,14 +370,17 @@ memx_exec:
 	push $r13
 	mov b32 $r1 $r12
 	mov b32 $r2 $r11
+
 	memx_exec_next:
-		// fetch the packet header, and locate opcode info
+		// fetch the packet header
 		ld b32 $r3 D[$r1]
 		add b32 $r1 4
-		shr b32 $r4 $r3 16
-		mulu $r3 #memx_func_size
+		extr $r4 $r3 16:31
+		extr $r3 $r3 0:15
 
 		// execute the opcode handler
+		sub b32 $r3 1
+		mulu $r3 #memx_func_size
 		ld b32 $r5 D[$r3 + #memx_func_head + #memx_func]
 		call $r5
 
@@ -176,6 +389,10 @@ memx_exec:
 		bra l #memx_exec_next
 
 	// send completion reply
+	ld b32 $r11 D[$r0 + #memx_ts_start]
+	ld b32 $r12 D[$r0 + #memx_ts_end]
+	sub b32 $r12 $r11
+	nv_iord($r11, NV_PPWR_INPUT)
 	pop $r13
 	pop $r14
 	call(send)
@@ -190,8 +407,19 @@ memx_exec:
 // $r11 - data1
 // $r0  - zero
 memx_info:
+	cmp b16 $r12 0x1
+	bra e #memx_info_train
+
+	memx_info_data:
 	mov $r12 #memx_data_head
 	mov $r11 #memx_data_tail - #memx_data_head
+	bra #memx_info_send
+
+	memx_info_train:
+	mov $r12 #memx_train_head
+	mov $r11 #memx_train_tail - #memx_train_head
+
+	memx_info_send:
 	call(send)
 	ret
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc
index 17a8a383d91a..b439519ec866 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc
@@ -23,6 +23,7 @@
  */
 
 #define NVKM_PPWR_CHIPSET GK208
+#define HW_TICKS_PER_US 324
 
 #define NVKM_FALCON_PC24
 #define NVKM_FALCON_UNSHIFTED_IO
@@ -34,6 +35,7 @@
 .section #nv108_pwr_data
 #define INCLUDE_PROC
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
@@ -44,6 +46,7 @@
 
 #define INCLUDE_DATA
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
@@ -56,6 +59,7 @@
 .section #nv108_pwr_code
 #define INCLUDE_CODE
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc.h b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc.h
index 986495d533dd..713e11e2953d 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc.h
@@ -24,8 +24,8 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 /* 0x0058: proc_list_head */
 	0x54534f48,
-	0x00000379,
-	0x0000032a,
+	0x00000453,
+	0x00000404,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -46,8 +46,8 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x584d454d,
-	0x00000464,
-	0x00000456,
+	0x0000062d,
+	0x0000061f,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -68,8 +68,8 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x46524550,
-	0x00000468,
-	0x00000466,
+	0x00000631,
+	0x0000062f,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -90,8 +90,8 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x5f433249,
-	0x0000086c,
-	0x00000713,
+	0x00000a35,
+	0x000008dc,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -112,8 +112,8 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x54534554,
-	0x0000088d,
-	0x0000086e,
+	0x00000a56,
+	0x00000a37,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -134,8 +134,8 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x454c4449,
-	0x00000898,
-	0x00000896,
+	0x00000a61,
+	0x00000a5f,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -227,24 +227,98 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 /* 0x0370: memx_func_head */
-	0x00010000,
-	0x00000000,
-	0x000003a9,
-/* 0x037c: memx_func_next */
 	0x00000001,
 	0x00000000,
-	0x000003c7,
+	0x00000483,
+/* 0x037c: memx_func_next */
 	0x00000002,
+	0x00000000,
+	0x00000500,
+	0x00000003,
 	0x00000002,
-	0x000003df,
-	0x00040003,
+	0x00000580,
+	0x00040004,
+	0x00000000,
+	0x0000059d,
+	0x00010005,
+	0x00000000,
+	0x000005b7,
+	0x00010006,
+	0x00000000,
+	0x0000057b,
+	0x00000007,
+	0x00000000,
+	0x000005c3,
+/* 0x03c4: memx_func_tail */
+/* 0x03c4: memx_ts_start */
+	0x00000000,
+/* 0x03c8: memx_ts_end */
+	0x00000000,
+/* 0x03cc: memx_data_head */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
-	0x000003fc,
-	0x00010004,
 	0x00000000,
-	0x00000416,
-/* 0x03ac: memx_func_tail */
-/* 0x03ac: memx_data_head */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -693,6 +767,8 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0bcc: memx_data_tail */
+/* 0x0bcc: memx_train_head */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -757,8 +833,8 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0bac: memx_data_tail */
-/* 0x0bac: i2c_scl_map */
+/* 0x0ccc: memx_train_tail */
+/* 0x0ccc: i2c_scl_map */
 	0x00000400,
 	0x00000800,
 	0x00001000,
@@ -769,7 +845,7 @@ uint32_t nv108_pwr_data[] = {
 	0x00020000,
 	0x00040000,
 	0x00080000,
-/* 0x0bd4: i2c_sda_map */
+/* 0x0cf4: i2c_sda_map */
 	0x00100000,
 	0x00200000,
 	0x00400000,
@@ -781,10 +857,66 @@ uint32_t nv108_pwr_data[] = {
 	0x10000000,
 	0x20000000,
 	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 };
 
 uint32_t nv108_pwr_code[] = {
-	0x02910ef5,
+	0x031c0ef5,
 /* 0x0004: rd32 */
 	0xf607a040,
 	0x04bd000e,
@@ -812,15 +944,18 @@ uint32_t nv108_pwr_code[] = {
 	0x7000d4f1,
 	0xf8f61bf4,
 /* 0x005d: nsec */
-	0xcf2c0800,
-/* 0x0062: nsec_loop */
+	0xf990f900,
+	0xcf2c0880,
+/* 0x0066: nsec_loop */
 	0x2c090088,
 	0xbb0099cf,
 	0x9ea60298,
-	0xf8f61ef4,
-/* 0x0071: wait */
-	0xcf2c0800,
-/* 0x0076: wait_loop */
+	0xfcf61ef4,
+	0xf890fc80,
+/* 0x0079: wait */
+	0xf990f900,
+	0xcf2c0880,
+/* 0x0082: wait_loop */
 	0xeeb20088,
 	0x0000047e,
 	0xadfddab2,
@@ -828,28 +963,29 @@ uint32_t nv108_pwr_code[] = {
 	0x2c09100b,
 	0xbb0099cf,
 	0x9ba60298,
-/* 0x0093: wait_done */
-	0xf8e61ef4,
-/* 0x0095: intr_watchdog */
+/* 0x009f: wait_done */
+	0xfce61ef4,
+	0xf890fc80,
+/* 0x00a5: intr_watchdog */
 	0x03e99800,
 	0xf40096b0,
 	0x0a98280b,
 	0x029abb9a,
 	0x0d0e1cf4,
-	0x01de7e01,
+	0x02617e01,
 	0xf494bd00,
-/* 0x00b2: intr_watchdog_next_time */
+/* 0x00c2: intr_watchdog_next_time */
 	0x0a98140e,
 	0x00a6b09b,
 	0xa6080bf4,
 	0x061cf49a,
-/* 0x00c0: intr_watchdog_next_time_set */
-/* 0x00c3: intr_watchdog_next_proc */
+/* 0x00d0: intr_watchdog_next_time_set */
+/* 0x00d3: intr_watchdog_next_proc */
 	0xb59b09b5,
 	0xe0b603e9,
 	0x68e6b158,
 	0xc81bf402,
-/* 0x00d2: intr */
+/* 0x00e2: intr */
 	0x00f900f8,
 	0x80f904bd,
 	0xa0f990f9,
@@ -865,13 +1001,13 @@ uint32_t nv108_pwr_code[] = {
 	0xc40088cf,
 	0x0bf40289,
 	0x9b00b51f,
-	0x957e580e,
+	0xa57e580e,
 	0x09980000,
 	0x0096b09b,
 	0x000d0bf4,
 	0x0009f634,
 	0x09b504bd,
-/* 0x0125: intr_skip_watchdog */
+/* 0x0135: intr_skip_watchdog */
 	0x0089e49a,
 	0x360bf408,
 	0xcf068849,
@@ -881,20 +1017,20 @@ uint32_t nv108_pwr_code[] = {
 	0xc0f900cc,
 	0xf14f484e,
 	0x0d5453e3,
-	0x023f7e00,
+	0x02c27e00,
 	0x40c0fc00,
 	0x0cf604c0,
-/* 0x0157: intr_subintr_skip_fifo */
+/* 0x0167: intr_subintr_skip_fifo */
 	0x4004bd00,
 	0x09f60688,
-/* 0x015f: intr_skip_subintr */
+/* 0x016f: intr_skip_subintr */
 	0xc404bd00,
 	0x0bf42089,
 	0xbfa4f107,
-/* 0x0169: intr_skip_pause */
+/* 0x0179: intr_skip_pause */
 	0x4089c4ff,
 	0xf1070bf4,
-/* 0x0173: intr_skip_user0 */
+/* 0x0183: intr_skip_user0 */
 	0x00ffbfa4,
 	0x0008f604,
 	0x80fc04bd,
@@ -904,549 +1040,684 @@ uint32_t nv108_pwr_code[] = {
 	0xfca0fcb0,
 	0xfc80fc90,
 	0x0032f400,
-/* 0x0196: timer */
-	0x32f401f8,
-	0x03f89810,
-	0xf40086b0,
-	0xfeb53a1c,
-	0xf6380003,
+/* 0x01a6: ticks_from_ns */
+	0xc0f901f8,
+	0xd7f1b0f9,
+	0xd3f00144,
+	0x7721f500,
+	0xe8ccec03,
+	0x00b4b003,
+	0xec120bf4,
+	0xf103e8ee,
+	0xf00144d7,
+	0x21f500d3,
+/* 0x01ce: ticks_from_ns_quit */
+	0xceb20377,
+	0xc0fcb0fc,
+/* 0x01d6: ticks_from_us */
+	0xc0f900f8,
+	0xd7f1b0f9,
+	0xd3f00144,
+	0x7721f500,
+	0xb0ceb203,
+	0x0bf400b4,
+/* 0x01ef: ticks_from_us_quit */
+	0xfce4bd05,
+	0xf8c0fcb0,
+/* 0x01f5: ticks_to_us */
+	0x44d7f100,
+	0x00d3f001,
+	0xf8ecedff,
+/* 0x0201: timer */
+	0xf990f900,
+	0x1032f480,
+	0xb003f898,
+	0x1cf40086,
+	0x0084bd4a,
+	0x0008f638,
+	0x340804bd,
+	0x980088cf,
+	0x98bb9a09,
+	0x00e9bb02,
+	0x0803feb5,
+	0x0088cf08,
+	0xf40284f0,
+	0x34081c1b,
+	0xa60088cf,
+	0x080bf4e0,
+	0x1cf4e8a6,
+/* 0x0245: timer_reset */
+	0xf634000d,
+	0x04bd000e,
+/* 0x024f: timer_enable */
+	0x089a0eb5,
+	0xf6380001,
 	0x04bd0008,
-	0x88cf0808,
-	0x0284f000,
-	0x081c1bf4,
-	0x0088cf34,
-	0x0bf4e0a6,
-	0xf4e8a608,
-/* 0x01c6: timer_reset */
-	0x3400161e,
-	0xbd000ef6,
-	0x9a0eb504,
-/* 0x01d0: timer_enable */
-	0x38000108,
-	0xbd0008f6,
-/* 0x01d9: timer_done */
-	0x1031f404,
-/* 0x01de: send_proc */
-	0x80f900f8,
-	0xe89890f9,
-	0x04e99805,
-	0xa60486f0,
-	0x2a0bf489,
-	0x940398c4,
-	0x80b60488,
-	0x008ebb18,
-	0xb500fa98,
-	0x8db5008a,
-	0x028cb501,
-	0xb6038bb5,
-	0x94f00190,
-	0x04e9b507,
-/* 0x0217: send_done */
-	0xfc0231f4,
-	0xf880fc90,
-/* 0x021d: find */
-	0x0880f900,
-	0x0131f458,
-/* 0x0224: find_loop */
-	0xa6008a98,
-	0x100bf4ae,
-	0xb15880b6,
-	0xf4026886,
-	0x32f4f11b,
-/* 0x0239: find_done */
-	0xfc8eb201,
-/* 0x023f: send */
-	0x7e00f880,
-	0xf400021d,
-	0x00f89b01,
-/* 0x0248: recv */
-	0x9805e898,
-	0x32f404e9,
-	0xf489a601,
-	0x89c43c0b,
-	0x0180b603,
-	0xb50784f0,
-	0xea9805e8,
-	0xfef0f902,
-	0xf0f9018f,
-	0x9994efb2,
-	0x00e9bb04,
-	0x9818e0b6,
-	0xec9803eb,
-	0x01ed9802,
-	0xf900ee98,
-	0xfef0fca5,
-	0x31f400f8,
-/* 0x028f: recv_done */
-	0xf8f0fc01,
-/* 0x0291: init */
-	0x01084100,
-	0xe70011cf,
-	0xb6010911,
-	0x14fe0814,
-	0x00e04100,
-	0x000013f0,
-	0x0001f61c,
-	0xff0104bd,
-	0x01f61400,
-	0x0104bd00,
-	0x0015f102,
-	0xf6100008,
-	0x04bd0001,
-	0xf000d241,
-	0x10fe0013,
-	0x1031f400,
-	0x38000101,
+/* 0x0258: timer_done */
+	0xfc1031f4,
+	0xf890fc80,
+/* 0x0261: send_proc */
+	0xf980f900,
+	0x05e89890,
+	0xf004e998,
+	0x89a60486,
+	0xc42a0bf4,
+	0x88940398,
+	0x1880b604,
+	0x98008ebb,
+	0x8ab500fa,
+	0x018db500,
+	0xb5028cb5,
+	0x90b6038b,
+	0x0794f001,
+	0xf404e9b5,
+/* 0x029a: send_done */
+	0x90fc0231,
+	0x00f880fc,
+/* 0x02a0: find */
+	0x580880f9,
+/* 0x02a7: find_loop */
+	0x980131f4,
+	0xaea6008a,
+	0xb6100bf4,
+	0x86b15880,
+	0x1bf40268,
+	0x0132f4f1,
+/* 0x02bc: find_done */
+	0x80fc8eb2,
+/* 0x02c2: send */
+	0xa07e00f8,
+	0x01f40002,
+/* 0x02cb: recv */
+	0xf900f89b,
+	0x9880f990,
+	0xe99805e8,
+	0x0132f404,
+	0x0bf489a6,
+	0x0389c43c,
+	0xf00180b6,
+	0xe8b50784,
+	0x02ea9805,
+	0x8ffef0f9,
+	0xb2f0f901,
+	0x049994ef,
+	0xb600e9bb,
+	0xeb9818e0,
+	0x02ec9803,
+	0x9801ed98,
+	0xa5f900ee,
+	0xf8fef0fc,
+	0x0131f400,
+/* 0x0316: recv_done */
+	0x80fcf0fc,
+	0x00f890fc,
+/* 0x031c: init */
+	0xcf010841,
+	0x11e70011,
+	0x14b60109,
+	0x0014fe08,
+	0xf000e041,
+	0x1c000013,
 	0xbd0001f6,
-/* 0x02db: init_proc */
-	0x98580f04,
-	0x16b001f1,
-	0xfa0bf400,
-	0xf0b615f9,
-	0xf20ef458,
-/* 0x02ec: host_send */
-	0xcf04b041,
-	0xa0420011,
-	0x0022cf04,
-	0x0bf412a6,
-	0x071ec42e,
-	0xb704ee94,
-	0x980270e0,
-	0xec9803eb,
-	0x01ed9802,
-	0x7e00ee98,
-	0xb600023f,
-	0x1ec40110,
-	0x04b0400f,
-	0xbd000ef6,
-	0xc70ef404,
-/* 0x0328: host_send_done */
-/* 0x032a: host_recv */
-	0x494100f8,
-	0x5413f14e,
-	0xf4e1a652,
-/* 0x0336: host_recv_wait */
-	0xcc41b90b,
-	0x0011cf04,
-	0xcf04c842,
-	0x16f00022,
-	0xf412a608,
-	0x23c4ef0b,
-	0x0434b607,
-	0x02f030b7,
-	0xb5033bb5,
-	0x3db5023c,
-	0x003eb501,
-	0xf00120b6,
-	0xc8400f24,
-	0x0002f604,
-	0x400204bd,
-	0x02f60000,
-	0xf804bd00,
-/* 0x0379: host_init */
-	0x00804100,
-	0xf11014b6,
-	0x40027015,
-	0x01f604d0,
+	0x00ff0104,
+	0x0001f614,
+	0x020104bd,
+	0x080015f1,
+	0x01f61000,
 	0x4104bd00,
+	0x13f000e2,
+	0x0010fe00,
+	0x011031f4,
+	0xf6380001,
+	0x04bd0001,
+/* 0x0366: init_proc */
+	0xf198580f,
+	0x0016b001,
+	0xf9fa0bf4,
+	0x58f0b615,
+/* 0x0377: mulu32_32_64 */
+	0xf9f20ef4,
+	0xf920f910,
+	0x9540f930,
+	0xd29510e1,
+	0xbdc4bd10,
+	0xc0edffb4,
+	0xb2301dff,
+	0xff34f134,
+	0x1034b6ff,
+	0xbb1045b6,
+	0xb4bb00c3,
+	0x30e2ff01,
+	0x34f134b2,
+	0x34b6ffff,
+	0x1045b610,
+	0xbb00c3bb,
+	0x12ff01b4,
+	0x00b3bb30,
+	0x30fc40fc,
+	0x10fc20fc,
+/* 0x03c6: host_send */
+	0xb04100f8,
+	0x0011cf04,
+	0xcf04a042,
+	0x12a60022,
+	0xc42e0bf4,
+	0xee94071e,
+	0x70e0b704,
+	0x03eb9802,
+	0x9802ec98,
+	0xee9801ed,
+	0x02c27e00,
+	0x0110b600,
+	0x400f1ec4,
+	0x0ef604b0,
+	0xf404bd00,
+/* 0x0402: host_send_done */
+	0x00f8c70e,
+/* 0x0404: host_recv */
+	0xf14e4941,
+	0xa6525413,
+	0xb90bf4e1,
+/* 0x0410: host_recv_wait */
+	0xcf04cc41,
+	0xc8420011,
+	0x0022cf04,
+	0xa60816f0,
+	0xef0bf412,
+	0xb60723c4,
+	0x30b70434,
+	0x3bb502f0,
+	0x023cb503,
+	0xb5013db5,
+	0x20b6003e,
+	0x0f24f001,
+	0xf604c840,
+	0x04bd0002,
+	0x00004002,
+	0xbd0002f6,
+/* 0x0453: host_init */
+	0x4100f804,
 	0x14b60080,
-	0xf015f110,
-	0x04dc4002,
+	0x7015f110,
+	0x04d04002,
 	0xbd0001f6,
-	0x40010104,
-	0x01f604c4,
-	0xf804bd00,
-/* 0x03a9: memx_func_enter */
-	0x40040600,
-	0x06f607e0,
-/* 0x03b3: memx_func_enter_wait */
-	0x4604bd00,
-	0x66cf07c0,
-	0x0464f000,
-	0x98f70bf4,
-	0x10b60016,
-/* 0x03c7: memx_func_leave */
-	0x0600f804,
-	0x07e44004,
-	0xbd0006f6,
-/* 0x03d1: memx_func_leave_wait */
-	0x07c04604,
-	0xf00066cf,
-	0x1bf40464,
-/* 0x03df: memx_func_wr32 */
-	0x9800f8f7,
-	0x15980016,
-	0x0810b601,
-	0x50f960f9,
+	0x00804104,
+	0xf11014b6,
+	0x4002f015,
+	0x01f604dc,
+	0x0104bd00,
+	0x04c44001,
+	0xbd0001f6,
+/* 0x0483: memx_func_enter */
+	0xf100f804,
+	0xf1162067,
+	0xf1f55d77,
+	0xb2ffff73,
+	0x00047e6e,
+	0xfdd8b200,
+	0x60f90487,
+	0xd0fc80f9,
+	0x2e7ee0fc,
+	0x77f10000,
+	0x73f1fffe,
+	0x6eb2ffff,
+	0x0000047e,
+	0x87fdd8b2,
+	0xf960f904,
+	0xfcd0fc80,
+	0x002e7ee0,
+	0xf067f100,
+	0x7e6eb226,
+	0xb2000004,
+	0x0487fdd8,
+	0x80f960f9,
 	0xe0fcd0fc,
 	0x00002e7e,
-	0xf40242b6,
-	0x00f8e81b,
-/* 0x03fc: memx_func_wait */
-	0x88cf2c08,
-	0x001e9800,
-	0x98011d98,
-	0x1b98021c,
-	0x1010b603,
-	0x0000717e,
-/* 0x0416: memx_func_delay */
-	0x1e9800f8,
-	0x0410b600,
-	0x00005d7e,
-/* 0x0422: memx_exec */
-	0xe0f900f8,
-	0xc1b2d0f9,
-/* 0x042a: memx_exec_next */
-	0x1398b2b2,
-	0x0410b600,
-	0xf0103495,
-	0x35980c30,
-	0xa655f9de,
-	0xed1ef412,
+	0xe0400406,
+	0x0006f607,
+/* 0x04ea: memx_func_enter_wait */
+	0xc04604bd,
+	0x0066cf07,
+	0xf40464f0,
+	0x2c06f70b,
+	0xb50066cf,
+	0x00f8f106,
+/* 0x0500: memx_func_leave */
+	0x66cf2c06,
+	0xf206b500,
+	0xe4400406,
+	0x0006f607,
+/* 0x0512: memx_func_leave_wait */
+	0xc04604bd,
+	0x0066cf07,
+	0xf40464f0,
+	0x67f1f71b,
+	0x77f126f0,
+	0x73f00001,
+	0x7e6eb200,
+	0xb2000004,
+	0x0587fdd8,
+	0x80f960f9,
 	0xe0fcd0fc,
-	0x00023f7e,
-/* 0x044a: memx_info */
-	0xac4c00f8,
+	0x00002e7e,
+	0x162067f1,
+	0x047e6eb2,
+	0xd8b20000,
+	0xf90587fd,
+	0xfc80f960,
+	0x7ee0fcd0,
+	0xf100002e,
+	0xf00aa277,
+	0x6eb20073,
+	0x0000047e,
+	0x87fdd8b2,
+	0xf960f905,
+	0xfcd0fc80,
+	0x002e7ee0,
+/* 0x057b: memx_func_wait_vblank */
+	0xb600f800,
+	0x00f80410,
+/* 0x0580: memx_func_wr32 */
+	0x98001698,
+	0x10b60115,
+	0xf960f908,
+	0xfcd0fc50,
+	0x002e7ee0,
+	0x0242b600,
+	0xf8e81bf4,
+/* 0x059d: memx_func_wait */
+	0xcf2c0800,
+	0x1e980088,
+	0x011d9800,
+	0x98021c98,
+	0x10b6031b,
+	0x00797e10,
+/* 0x05b7: memx_func_delay */
+	0x9800f800,
+	0x10b6001e,
+	0x005d7e04,
+/* 0x05c3: memx_func_train */
+	0xf800f800,
+/* 0x05c5: memx_exec */
+	0xf9e0f900,
+	0xb2c1b2d0,
+/* 0x05cd: memx_exec_next */
+	0x001398b2,
+	0xe70410b6,
+	0xe701f034,
+	0xb601e033,
+	0x30f00132,
+	0xde35980c,
+	0x12a655f9,
+	0x98e51ef4,
+	0x0c98f10b,
+	0x02cbbbf2,
+	0xcf07c44b,
+	0xd0fc00bb,
+	0xc27ee0fc,
+	0x00f80002,
+/* 0x0604: memx_info */
+	0xf401c670,
+/* 0x060a: memx_info_data */
+	0xcc4c0c0b,
 	0x08004b03,
-	0x00023f7e,
-/* 0x0456: memx_recv */
-	0xd6b000f8,
-	0xc90bf401,
-	0xf400d6b0,
-	0x00f8eb0b,
-/* 0x0464: memx_init */
-/* 0x0466: perf_recv */
-	0x00f800f8,
-/* 0x0468: perf_init */
-/* 0x046a: i2c_drive_scl */
-	0x36b000f8,
-	0x0d0bf400,
-	0xf607e040,
-	0x04bd0001,
-/* 0x047a: i2c_drive_scl_lo */
-	0xe44000f8,
-	0x0001f607,
-	0x00f804bd,
-/* 0x0484: i2c_drive_sda */
-	0xf40036b0,
-	0xe0400d0b,
-	0x0002f607,
-	0x00f804bd,
-/* 0x0494: i2c_drive_sda_lo */
-	0xf607e440,
-	0x04bd0002,
-/* 0x049e: i2c_sense_scl */
-	0x32f400f8,
-	0x07c44301,
-	0xfd0033cf,
-	0x0bf40431,
-	0x0131f406,
-/* 0x04b0: i2c_sense_scl_done */
-/* 0x04b2: i2c_sense_sda */
-	0x32f400f8,
-	0x07c44301,
-	0xfd0033cf,
-	0x0bf40432,
-	0x0131f406,
-/* 0x04c4: i2c_sense_sda_done */
-/* 0x04c6: i2c_raise_scl */
-	0x40f900f8,
-	0x03089844,
-	0x046a7e01,
-/* 0x04d1: i2c_raise_scl_wait */
-	0x03e84e00,
-	0x00005d7e,
-	0x00049e7e,
-	0xb60901f4,
-	0x1bf40142,
-/* 0x04e5: i2c_raise_scl_done */
-	0xf840fcef,
-/* 0x04e9: i2c_start */
-	0x049e7e00,
-	0x0d11f400,
-	0x0004b27e,
-	0xf40611f4,
-/* 0x04fa: i2c_start_rep */
-	0x00032e0e,
-	0x00046a7e,
-	0x847e0103,
-	0x76bb0004,
-	0x0465b600,
-	0x659450f9,
-	0x0256bb04,
-	0x75fd50bd,
-	0x7e50fc04,
-	0xb60004c6,
-	0x11f40464,
-/* 0x0525: i2c_start_send */
-	0x7e00031d,
-	0x4e000484,
-	0x5d7e1388,
-	0x00030000,
-	0x00046a7e,
-	0x7e13884e,
-/* 0x053f: i2c_start_out */
-	0xf800005d,
-/* 0x0541: i2c_stop */
-	0x7e000300,
-	0x0300046a,
-	0x04847e00,
-	0x03e84e00,
-	0x00005d7e,
-	0x6a7e0103,
-	0x884e0004,
-	0x005d7e13,
+/* 0x0613: memx_info_train */
+	0x4c090ef4,
+	0x004b0bcc,
+/* 0x0619: memx_info_send */
+	0x02c27e01,
+/* 0x061f: memx_recv */
+	0xb000f800,
+	0x0bf401d6,
+	0x00d6b0a3,
+	0xf8dc0bf4,
+/* 0x062d: memx_init */
+/* 0x062f: perf_recv */
+	0xf800f800,
+/* 0x0631: perf_init */
+/* 0x0633: i2c_drive_scl */
+	0xb000f800,
+	0x0bf40036,
+	0x07e0400d,
+	0xbd0001f6,
+/* 0x0643: i2c_drive_scl_lo */
+	0x4000f804,
+	0x01f607e4,
+	0xf804bd00,
+/* 0x064d: i2c_drive_sda */
+	0x0036b000,
+	0x400d0bf4,
+	0x02f607e0,
+	0xf804bd00,
+/* 0x065d: i2c_drive_sda_lo */
+	0x07e44000,
+	0xbd0002f6,
+/* 0x0667: i2c_sense_scl */
+	0xf400f804,
+	0xc4430132,
+	0x0033cf07,
+	0xf40431fd,
+	0x31f4060b,
+/* 0x0679: i2c_sense_scl_done */
+/* 0x067b: i2c_sense_sda */
+	0xf400f801,
+	0xc4430132,
+	0x0033cf07,
+	0xf40432fd,
+	0x31f4060b,
+/* 0x068d: i2c_sense_sda_done */
+/* 0x068f: i2c_raise_scl */
+	0xf900f801,
+	0x08984440,
+	0x337e0103,
+/* 0x069a: i2c_raise_scl_wait */
+	0xe84e0006,
+	0x005d7e03,
+	0x06677e00,
+	0x0901f400,
+	0xf40142b6,
+/* 0x06ae: i2c_raise_scl_done */
+	0x40fcef1b,
+/* 0x06b2: i2c_start */
+	0x677e00f8,
+	0x11f40006,
+	0x067b7e0d,
+	0x0611f400,
+/* 0x06c3: i2c_start_rep */
+	0x032e0ef4,
+	0x06337e00,
 	0x7e010300,
-	0x4e000484,
-	0x5d7e1388,
-	0x00f80000,
-/* 0x0570: i2c_bitw */
-	0x0004847e,
-	0x7e03e84e,
-	0xbb00005d,
+	0xbb00064d,
 	0x65b60076,
 	0x9450f904,
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x0004c67e,
+	0x00068f7e,
 	0xf40464b6,
-	0x884e1711,
-	0x005d7e13,
-	0x7e000300,
-	0x4e00046a,
-	0x5d7e1388,
-/* 0x05ae: i2c_bitw_out */
-	0x00f80000,
-/* 0x05b0: i2c_bitr */
-	0x847e0103,
-	0xe84e0004,
+/* 0x06ee: i2c_start_send */
+	0x00031d11,
+	0x00064d7e,
+	0x7e13884e,
+	0x0300005d,
+	0x06337e00,
+	0x13884e00,
+	0x00005d7e,
+/* 0x0708: i2c_start_out */
+/* 0x070a: i2c_stop */
+	0x000300f8,
+	0x0006337e,
+	0x4d7e0003,
+	0xe84e0006,
 	0x005d7e03,
-	0x0076bb00,
-	0xf90465b6,
-	0x04659450,
-	0xbd0256bb,
-	0x0475fd50,
-	0xc67e50fc,
-	0x64b60004,
-	0x1a11f404,
-	0x0004b27e,
-	0x6a7e0003,
-	0x884e0004,
-	0x005d7e13,
-	0x013cf000,
-/* 0x05f3: i2c_bitr_done */
-	0xf80131f4,
-/* 0x05f5: i2c_get_byte */
-	0x04000500,
-/* 0x05f9: i2c_get_byte_next */
-	0x0154b608,
+	0x7e010300,
+	0x4e000633,
+	0x5d7e1388,
+	0x01030000,
+	0x00064d7e,
+	0x7e13884e,
+	0xf800005d,
+/* 0x0739: i2c_bitw */
+	0x064d7e00,
+	0x03e84e00,
+	0x00005d7e,
 	0xb60076bb,
 	0x50f90465,
 	0xbb046594,
 	0x50bd0256,
 	0xfc0475fd,
-	0x05b07e50,
+	0x068f7e50,
 	0x0464b600,
-	0xfd2a11f4,
-	0x42b60553,
-	0xd81bf401,
-	0x76bb0103,
+	0x4e1711f4,
+	0x5d7e1388,
+	0x00030000,
+	0x0006337e,
+	0x7e13884e,
+/* 0x0777: i2c_bitw_out */
+	0xf800005d,
+/* 0x0779: i2c_bitr */
+	0x7e010300,
+	0x4e00064d,
+	0x5d7e03e8,
+	0x76bb0000,
 	0x0465b600,
 	0x659450f9,
 	0x0256bb04,
 	0x75fd50bd,
 	0x7e50fc04,
-	0xb6000570,
-/* 0x0642: i2c_get_byte_done */
-	0x00f80464,
-/* 0x0644: i2c_put_byte */
-/* 0x0646: i2c_put_byte_next */
-	0x42b60804,
-	0x3854ff01,
-	0xb60076bb,
-	0x50f90465,
-	0xbb046594,
-	0x50bd0256,
-	0xfc0475fd,
-	0x05707e50,
-	0x0464b600,
-	0xb03411f4,
-	0x1bf40046,
-	0x0076bbd8,
+	0xb600068f,
+	0x11f40464,
+	0x067b7e1a,
+	0x7e000300,
+	0x4e000633,
+	0x5d7e1388,
+	0x3cf00000,
+	0x0131f401,
+/* 0x07bc: i2c_bitr_done */
+/* 0x07be: i2c_get_byte */
+	0x000500f8,
+/* 0x07c2: i2c_get_byte_next */
+	0x54b60804,
+	0x0076bb01,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x797e50fc,
+	0x64b60007,
+	0x2a11f404,
+	0xb60553fd,
+	0x1bf40142,
+	0xbb0103d8,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x0007397e,
+/* 0x080b: i2c_get_byte_done */
+	0xf80464b6,
+/* 0x080d: i2c_put_byte */
+/* 0x080f: i2c_put_byte_next */
+	0xb6080400,
+	0x54ff0142,
+	0x0076bb38,
 	0xf90465b6,
 	0x04659450,
 	0xbd0256bb,
 	0x0475fd50,
-	0xb07e50fc,
-	0x64b60005,
-	0x0f11f404,
-	0xb00076bb,
-	0x1bf40136,
-	0x0132f406,
-/* 0x069c: i2c_put_byte_done */
-/* 0x069e: i2c_addr */
-	0x76bb00f8,
+	0x397e50fc,
+	0x64b60007,
+	0x3411f404,
+	0xf40046b0,
+	0x76bbd81b,
 	0x0465b600,
 	0x659450f9,
 	0x0256bb04,
 	0x75fd50bd,
 	0x7e50fc04,
-	0xb60004e9,
+	0xb6000779,
 	0x11f40464,
-	0x2ec3e729,
-	0x0134b601,
-	0xbb0553fd,
+	0x0076bb0f,
+	0xf40136b0,
+	0x32f4061b,
+/* 0x0865: i2c_put_byte_done */
+/* 0x0867: i2c_addr */
+	0xbb00f801,
 	0x65b60076,
 	0x9450f904,
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x0006447e,
-/* 0x06e3: i2c_addr_done */
-	0xf80464b6,
-/* 0x06e5: i2c_acquire_addr */
-	0xf8cec700,
-	0xb705e4b6,
-	0xf8d014e0,
-/* 0x06f1: i2c_acquire */
-	0x06e57e00,
-	0x00047e00,
-	0x03d9f000,
-	0x00002e7e,
-/* 0x0702: i2c_release */
-	0xe57e00f8,
-	0x047e0006,
-	0xdaf00000,
+	0x0006b27e,
+	0xf40464b6,
+	0xc3e72911,
+	0x34b6012e,
+	0x0553fd01,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x080d7e50,
+	0x0464b600,
+/* 0x08ac: i2c_addr_done */
+/* 0x08ae: i2c_acquire_addr */
+	0xcec700f8,
+	0x05e4b6f8,
+	0xd014e0b7,
+/* 0x08ba: i2c_acquire */
+	0xae7e00f8,
+	0x047e0008,
+	0xd9f00000,
 	0x002e7e03,
-/* 0x0713: i2c_recv */
-	0xf400f800,
-	0xc1c70132,
-	0x0214b6f8,
-	0xf52816b0,
-	0xb801371f,
-	0x000bd413,
-	0xb8003298,
-	0x000bac13,
-	0xf4003198,
-	0xd0f90231,
-	0xd0f9e0f9,
-	0x000067f1,
-	0x100063f1,
-	0xbb016792,
+/* 0x08cb: i2c_release */
+	0x7e00f800,
+	0x7e0008ae,
+	0xf0000004,
+	0x2e7e03da,
+	0x00f80000,
+/* 0x08dc: i2c_recv */
+	0xc70132f4,
+	0x14b6f8c1,
+	0x2816b002,
+	0x01371ff5,
+	0x0cf413b8,
+	0x00329800,
+	0x0ccc13b8,
+	0x00319800,
+	0xf90231f4,
+	0xf9e0f9d0,
+	0x0067f1d0,
+	0x0063f100,
+	0x01679210,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x08ba7e50,
+	0x0464b600,
+	0xd6b0d0fc,
+	0xb01bf500,
+	0xbb000500,
 	0x65b60076,
 	0x9450f904,
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x0006f17e,
-	0xfc0464b6,
-	0x00d6b0d0,
-	0x00b01bf5,
-	0x76bb0005,
+	0x0008677e,
+	0xf50464b6,
+	0xc700cc11,
+	0x76bbe0c5,
 	0x0465b600,
 	0x659450f9,
 	0x0256bb04,
 	0x75fd50bd,
 	0x7e50fc04,
-	0xb600069e,
+	0xb600080d,
 	0x11f50464,
-	0xc5c700cc,
-	0x0076bbe0,
-	0xf90465b6,
-	0x04659450,
-	0xbd0256bb,
-	0x0475fd50,
-	0x447e50fc,
-	0x64b60006,
-	0xa911f504,
-	0xbb010500,
-	0x65b60076,
-	0x9450f904,
-	0x56bb0465,
-	0xfd50bd02,
-	0x50fc0475,
-	0x00069e7e,
-	0xf50464b6,
-	0xbb008711,
-	0x65b60076,
-	0x9450f904,
-	0x56bb0465,
-	0xfd50bd02,
-	0x50fc0475,
-	0x0005f57e,
-	0xf40464b6,
-	0x5bcb6711,
-	0x0076bbe0,
-	0xf90465b6,
-	0x04659450,
-	0xbd0256bb,
-	0x0475fd50,
-	0x417e50fc,
-	0x64b60005,
-	0xbd5bb204,
-	0x410ef474,
-/* 0x0818: i2c_recv_not_rd08 */
-	0xf401d6b0,
-	0x00053b1b,
-	0x00069e7e,
-	0xc73211f4,
-	0x447ee0c5,
-	0x11f40006,
-	0x7e000528,
-	0xf400069e,
-	0xb5c71f11,
-	0x06447ee0,
-	0x1511f400,
-	0x0005417e,
-	0xc5c774bd,
-	0x091bf408,
-	0xf40232f4,
-/* 0x0856: i2c_recv_not_wr08 */
-/* 0x0856: i2c_recv_done */
-	0xcec7030e,
-	0x07027ef8,
-	0xfce0fc00,
-	0x0912f4d0,
-	0x3f7e7cb2,
-/* 0x086a: i2c_recv_exit */
-	0x00f80002,
-/* 0x086c: i2c_init */
-/* 0x086e: test_recv */
-	0x584100f8,
-	0x0011cf04,
-	0x400110b6,
-	0x01f60458,
-	0xf104bd00,
-	0xf1d900e7,
-	0x7e134fe3,
-	0xf8000196,
-/* 0x088d: test_init */
-	0x08004e00,
-	0x0001967e,
-/* 0x0896: idle_recv */
-	0x00f800f8,
-/* 0x0898: idle */
-	0x410031f4,
-	0x11cf0454,
+	0x010500a9,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x08677e50,
+	0x0464b600,
+	0x008711f5,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x07be7e50,
+	0x0464b600,
+	0xcb6711f4,
+	0x76bbe05b,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0x7e50fc04,
+	0xb600070a,
+	0x5bb20464,
+	0x0ef474bd,
+/* 0x09e1: i2c_recv_not_rd08 */
+	0x01d6b041,
+	0x053b1bf4,
+	0x08677e00,
+	0x3211f400,
+	0x7ee0c5c7,
+	0xf400080d,
+	0x00052811,
+	0x0008677e,
+	0xc71f11f4,
+	0x0d7ee0b5,
+	0x11f40008,
+	0x070a7e15,
+	0xc774bd00,
+	0x1bf408c5,
+	0x0232f409,
+/* 0x0a1f: i2c_recv_not_wr08 */
+/* 0x0a1f: i2c_recv_done */
+	0xc7030ef4,
+	0xcb7ef8ce,
+	0xe0fc0008,
+	0x12f4d0fc,
+	0x7e7cb209,
+/* 0x0a33: i2c_recv_exit */
+	0xf80002c2,
+/* 0x0a35: i2c_init */
+/* 0x0a37: test_recv */
+	0x4100f800,
+	0x11cf0458,
 	0x0110b600,
-	0xf6045440,
+	0xf6045840,
 	0x04bd0001,
-/* 0x08ac: idle_loop */
-	0x32f45801,
-/* 0x08b1: idle_proc */
-/* 0x08b1: idle_proc_exec */
-	0xb210f902,
-	0x02487e1e,
-	0xf410fc00,
-	0x31f40911,
-	0xf00ef402,
-/* 0x08c4: idle_proc_next */
-	0xa65810b6,
-	0xe81bf41f,
-	0xf4e002f4,
-	0x0ef40028,
-	0x000000c6,
+	0xd900e7f1,
+	0x134fe3f1,
+	0x0002017e,
+/* 0x0a56: test_init */
+	0x004e00f8,
+	0x02017e08,
+/* 0x0a5f: idle_recv */
+	0xf800f800,
+/* 0x0a61: idle */
+	0x0031f400,
+	0xcf045441,
+	0x10b60011,
+	0x04544001,
+	0xbd0001f6,
+/* 0x0a75: idle_loop */
+	0xf4580104,
+/* 0x0a7a: idle_proc */
+/* 0x0a7a: idle_proc_exec */
+	0x10f90232,
+	0xcb7e1eb2,
+	0x10fc0002,
+	0xf40911f4,
+	0x0ef40231,
+/* 0x0a8d: idle_proc_next */
+	0x5810b6f0,
+	0x1bf41fa6,
+	0xe002f4e8,
+	0xf40028f4,
+	0x0000c60e,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc
index 6744fcc06151..daa06c1c655e 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc
@@ -23,6 +23,7 @@
  */
 
 #define NVKM_PPWR_CHIPSET GT215
+#define HW_TICKS_PER_US 203 // should be 202.5
 
 //#define NVKM_FALCON_PC24
 //#define NVKM_FALCON_UNSHIFTED_IO
@@ -34,6 +35,7 @@
 .section #nva3_pwr_data
 #define INCLUDE_PROC
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
@@ -44,6 +46,7 @@
 
 #define INCLUDE_DATA
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
@@ -56,6 +59,7 @@
 .section #nva3_pwr_code
 #define INCLUDE_CODE
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc.h b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc.h
index e087ce3041be..d1f9b6cb66d7 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc.h
@@ -24,8 +24,8 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 /* 0x0058: proc_list_head */
 	0x54534f48,
-	0x00000430,
-	0x000003cd,
+	0x00000512,
+	0x000004af,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -46,8 +46,8 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x584d454d,
-	0x00000542,
-	0x00000534,
+	0x00000842,
+	0x00000834,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -68,8 +68,8 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x46524550,
-	0x00000546,
-	0x00000544,
+	0x00000846,
+	0x00000844,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -90,8 +90,8 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x5f433249,
-	0x00000976,
-	0x00000819,
+	0x00000c76,
+	0x00000b19,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -112,8 +112,8 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x54534554,
-	0x0000099f,
-	0x00000978,
+	0x00000c9f,
+	0x00000c78,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -134,8 +134,8 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x454c4449,
-	0x000009ab,
-	0x000009a9,
+	0x00000cab,
+	0x00000ca9,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -227,24 +227,88 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 /* 0x0370: memx_func_head */
-	0x00010000,
-	0x00000000,
-	0x0000046f,
-/* 0x037c: memx_func_next */
 	0x00000001,
 	0x00000000,
-	0x00000496,
+	0x00000551,
+/* 0x037c: memx_func_next */
 	0x00000002,
+	0x00000000,
+	0x000005a8,
+	0x00000003,
 	0x00000002,
-	0x000004b7,
-	0x00040003,
+	0x0000063a,
+	0x00040004,
+	0x00000000,
+	0x00000656,
+	0x00010005,
+	0x00000000,
+	0x00000673,
+	0x00010006,
+	0x00000000,
+	0x000005f8,
+	0x00000007,
+	0x00000000,
+	0x0000067e,
+/* 0x03c4: memx_func_tail */
+/* 0x03c4: memx_ts_start */
+	0x00000000,
+/* 0x03c8: memx_ts_end */
+	0x00000000,
+/* 0x03cc: memx_data_head */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
-	0x000004d3,
-	0x00010004,
 	0x00000000,
-	0x000004f0,
-/* 0x03ac: memx_func_tail */
-/* 0x03ac: memx_data_head */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -703,6 +767,8 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0bcc: memx_data_tail */
+/* 0x0bcc: memx_train_head */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -757,8 +823,18 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0bac: memx_data_tail */
-/* 0x0bac: i2c_scl_map */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+/* 0x0ccc: memx_train_tail */
+/* 0x0ccc: i2c_scl_map */
 	0x00001000,
 	0x00004000,
 	0x00010000,
@@ -769,7 +845,7 @@ uint32_t nva3_pwr_data[] = {
 	0x01000000,
 	0x04000000,
 	0x10000000,
-/* 0x0bd4: i2c_sda_map */
+/* 0x0cf4: i2c_sda_map */
 	0x00002000,
 	0x00008000,
 	0x00020000,
@@ -780,7 +856,7 @@ uint32_t nva3_pwr_data[] = {
 	0x02000000,
 	0x08000000,
 	0x20000000,
-/* 0x0bfc: i2c_ctrl */
+/* 0x0d1c: i2c_ctrl */
 	0x0000e138,
 	0x0000e150,
 	0x0000e168,
@@ -838,18 +914,10 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
 };
 
 uint32_t nva3_pwr_code[] = {
-	0x030d0ef5,
+	0x039e0ef5,
 /* 0x0004: rd32 */
 	0x07a007f1,
 	0xd00604b6,
@@ -885,19 +953,22 @@ uint32_t nva3_pwr_code[] = {
 	0xd4f100dd,
 	0x1bf47000,
 /* 0x007f: nsec */
-	0xf000f8f2,
+	0xf900f8f2,
+	0xf080f990,
 	0x84b62c87,
 	0x0088cf06,
-/* 0x0088: nsec_loop */
+/* 0x008c: nsec_loop */
 	0xb62c97f0,
 	0x99cf0694,
 	0x0298bb00,
 	0xf4069eb8,
-	0x00f8f11e,
-/* 0x009c: wait */
+	0x80fcf11e,
+	0x00f890fc,
+/* 0x00a4: wait */
+	0x80f990f9,
 	0xb62c87f0,
 	0x88cf0684,
-/* 0x00a5: wait_loop */
+/* 0x00b1: wait_loop */
 	0x02eeb900,
 	0xb90421f4,
 	0xadfd02da,
@@ -907,28 +978,29 @@ uint32_t nva3_pwr_code[] = {
 	0x0099cf06,
 	0xb80298bb,
 	0x1ef4069b,
-/* 0x00c9: wait_done */
-/* 0x00cb: intr_watchdog */
-	0x9800f8df,
+/* 0x00d5: wait_done */
+	0xfc80fcdf,
+/* 0x00db: intr_watchdog */
+	0x9800f890,
 	0x96b003e9,
 	0x2a0bf400,
 	0xbb9a0a98,
 	0x1cf4029a,
 	0x01d7f00f,
-	0x025421f5,
+	0x02dd21f5,
 	0x0ef494bd,
-/* 0x00e9: intr_watchdog_next_time */
+/* 0x00f9: intr_watchdog_next_time */
 	0x9b0a9815,
 	0xf400a6b0,
 	0x9ab8090b,
 	0x061cf406,
-/* 0x00f8: intr_watchdog_next_time_set */
-/* 0x00fb: intr_watchdog_next_proc */
+/* 0x0108: intr_watchdog_next_time_set */
+/* 0x010b: intr_watchdog_next_proc */
 	0x809b0980,
 	0xe0b603e9,
 	0x68e6b158,
 	0xc61bf402,
-/* 0x010a: intr */
+/* 0x011a: intr */
 	0x00f900f8,
 	0x80f904bd,
 	0xa0f990f9,
@@ -948,13 +1020,13 @@ uint32_t nva3_pwr_code[] = {
 	0xf40289c4,
 	0x0080230b,
 	0x58e7f09b,
-	0x98cb21f4,
+	0x98db21f4,
 	0x96b09b09,
 	0x110bf400,
 	0xb63407f0,
 	0x09d00604,
 	0x8004bd00,
-/* 0x016e: intr_skip_watchdog */
+/* 0x017e: intr_skip_watchdog */
 	0x89e49a09,
 	0x0bf40800,
 	0x8897f148,
@@ -967,22 +1039,22 @@ uint32_t nva3_pwr_code[] = {
 	0x48e7f1c0,
 	0x53e3f14f,
 	0x00d7f054,
-	0x02b921f5,
+	0x034221f5,
 	0x07f1c0fc,
 	0x04b604c0,
 	0x000cd006,
-/* 0x01ae: intr_subintr_skip_fifo */
+/* 0x01be: intr_subintr_skip_fifo */
 	0x07f104bd,
 	0x04b60688,
 	0x0009d006,
-/* 0x01ba: intr_skip_subintr */
+/* 0x01ca: intr_skip_subintr */
 	0x89c404bd,
 	0x070bf420,
 	0xffbfa4f1,
-/* 0x01c4: intr_skip_pause */
+/* 0x01d4: intr_skip_pause */
 	0xf44089c4,
 	0xa4f1070b,
-/* 0x01ce: intr_skip_user0 */
+/* 0x01de: intr_skip_user0 */
 	0x07f0ffbf,
 	0x0604b604,
 	0xbd0008d0,
@@ -993,344 +1065,551 @@ uint32_t nva3_pwr_code[] = {
 	0x90fca0fc,
 	0x00fc80fc,
 	0xf80032f4,
-/* 0x01f5: timer */
-	0x1032f401,
-	0xb003f898,
-	0x1cf40086,
-	0x03fe8051,
+/* 0x0205: ticks_from_ns */
+	0xf9c0f901,
+	0xcbd7f1b0,
+	0x00d3f000,
+	0x041321f5,
+	0x03e8ccec,
+	0xf400b4b0,
+	0xeeec120b,
+	0xd7f103e8,
+	0xd3f000cb,
+	0x1321f500,
+/* 0x022d: ticks_from_ns_quit */
+	0x02ceb904,
+	0xc0fcb0fc,
+/* 0x0236: ticks_from_us */
+	0xc0f900f8,
+	0xd7f1b0f9,
+	0xd3f000cb,
+	0x1321f500,
+	0x02ceb904,
+	0xf400b4b0,
+	0xe4bd050b,
+/* 0x0250: ticks_from_us_quit */
+	0xc0fcb0fc,
+/* 0x0256: ticks_to_us */
+	0xd7f100f8,
+	0xd3f000cb,
+	0xecedff00,
+/* 0x0262: timer */
+	0x90f900f8,
+	0x32f480f9,
+	0x03f89810,
+	0xf40086b0,
+	0x84bd651c,
 	0xb63807f0,
 	0x08d00604,
 	0xf004bd00,
-	0x84b60887,
+	0x84b63487,
 	0x0088cf06,
-	0xf40284f0,
-	0x87f0261b,
-	0x0684b634,
-	0xb80088cf,
-	0x0bf406e0,
-	0x06e8b809,
-/* 0x0233: timer_reset */
-	0xf01f1ef4,
-	0x04b63407,
-	0x000ed006,
-	0x0e8004bd,
-/* 0x0241: timer_enable */
-	0x0187f09a,
+	0xbb9a0998,
+	0xe9bb0298,
+	0x03fe8000,
+	0xb60887f0,
+	0x88cf0684,
+	0x0284f000,
+	0xf0261bf4,
+	0x84b63487,
+	0x0088cf06,
+	0xf406e0b8,
+	0xe8b8090b,
+	0x111cf406,
+/* 0x02b8: timer_reset */
+	0xb63407f0,
+	0x0ed00604,
+	0x8004bd00,
+/* 0x02c6: timer_enable */
+	0x87f09a0e,
+	0x3807f001,
+	0xd00604b6,
+	0x04bd0008,
+/* 0x02d4: timer_done */
+	0xfc1031f4,
+	0xf890fc80,
+/* 0x02dd: send_proc */
+	0xf980f900,
+	0x05e89890,
+	0xf004e998,
+	0x89b80486,
+	0x2a0bf406,
+	0x940398c4,
+	0x80b60488,
+	0x008ebb18,
+	0x8000fa98,
+	0x8d80008a,
+	0x028c8001,
+	0xb6038b80,
+	0x94f00190,
+	0x04e98007,
+/* 0x0317: send_done */
+	0xfc0231f4,
+	0xf880fc90,
+/* 0x031d: find */
+	0xf080f900,
+	0x31f45887,
+/* 0x0325: find_loop */
+	0x008a9801,
+	0xf406aeb8,
+	0x80b6100b,
+	0x6886b158,
+	0xf01bf402,
+/* 0x033b: find_done */
+	0xb90132f4,
+	0x80fc028e,
+/* 0x0342: send */
+	0x21f500f8,
+	0x01f4031d,
+/* 0x034b: recv */
+	0xf900f897,
+	0x9880f990,
+	0xe99805e8,
+	0x0132f404,
+	0xf40689b8,
+	0x89c43d0b,
+	0x0180b603,
+	0x800784f0,
+	0xea9805e8,
+	0xfef0f902,
+	0xf0f9018f,
+	0x9402efb9,
+	0xe9bb0499,
+	0x18e0b600,
+	0x9803eb98,
+	0xed9802ec,
+	0x00ee9801,
+	0xf0fca5f9,
+	0xf400f8fe,
+	0xf0fc0131,
+/* 0x0398: recv_done */
+	0x90fc80fc,
+/* 0x039e: init */
+	0x17f100f8,
+	0x14b60108,
+	0x0011cf06,
+	0x010911e7,
+	0xfe0814b6,
+	0x17f10014,
+	0x13f000e0,
+	0x1c07f000,
+	0xd00604b6,
+	0x04bd0001,
+	0xf0ff17f0,
+	0x04b61407,
+	0x0001d006,
+	0x17f004bd,
+	0x0015f102,
+	0x1007f008,
+	0xd00604b6,
+	0x04bd0001,
+	0x011a17f1,
+	0xfe0013f0,
+	0x31f40010,
+	0x0117f010,
 	0xb63807f0,
-	0x08d00604,
-/* 0x024f: timer_done */
-	0xf404bd00,
-	0x00f81031,
-/* 0x0254: send_proc */
-	0x90f980f9,
-	0x9805e898,
-	0x86f004e9,
-	0x0689b804,
-	0xc42a0bf4,
-	0x88940398,
-	0x1880b604,
-	0x98008ebb,
-	0x8a8000fa,
-	0x018d8000,
-	0x80028c80,
-	0x90b6038b,
-	0x0794f001,
-	0xf404e980,
-/* 0x028e: send_done */
-	0x90fc0231,
-	0x00f880fc,
-/* 0x0294: find */
-	0x87f080f9,
-	0x0131f458,
-/* 0x029c: find_loop */
-	0xb8008a98,
-	0x0bf406ae,
-	0x5880b610,
-	0x026886b1,
-	0xf4f01bf4,
-/* 0x02b2: find_done */
-	0x8eb90132,
-	0xf880fc02,
-/* 0x02b9: send */
-	0x9421f500,
-	0x9701f402,
-/* 0x02c2: recv */
-	0xe89800f8,
-	0x04e99805,
-	0xb80132f4,
-	0x0bf40689,
-	0x0389c43d,
-	0xf00180b6,
-	0xe8800784,
-	0x02ea9805,
-	0x8ffef0f9,
-	0xb9f0f901,
-	0x999402ef,
-	0x00e9bb04,
-	0x9818e0b6,
-	0xec9803eb,
-	0x01ed9802,
-	0xf900ee98,
-	0xfef0fca5,
-	0x31f400f8,
-/* 0x030b: recv_done */
-	0xf8f0fc01,
-/* 0x030d: init */
-	0x0817f100,
-	0x0614b601,
-	0xe70011cf,
-	0xb6010911,
-	0x14fe0814,
-	0xe017f100,
-	0x0013f000,
-	0xb61c07f0,
 	0x01d00604,
 	0xf004bd00,
-	0x07f0ff17,
-	0x0604b614,
-	0xbd0001d0,
-	0x0217f004,
-	0x080015f1,
-	0xb61007f0,
-	0x01d00604,
-	0xf104bd00,
-	0xf0010a17,
-	0x10fe0013,
-	0x1031f400,
-	0xf00117f0,
-	0x04b63807,
-	0x0001d006,
-	0xf7f004bd,
-/* 0x0371: init_proc */
-	0x01f19858,
-	0xf40016b0,
-	0x15f9fa0b,
-	0xf458f0b6,
-/* 0x0382: host_send */
-	0x17f1f20e,
-	0x14b604b0,
-	0x0011cf06,
-	0x04a027f1,
-	0xcf0624b6,
-	0x12b80022,
-	0x320bf406,
-	0x94071ec4,
-	0xe0b704ee,
-	0xeb980270,
-	0x02ec9803,
-	0x9801ed98,
-	0x21f500ee,
-	0x10b602b9,
-	0x0f1ec401,
-	0x04b007f1,
-	0xd00604b6,
-	0x04bd000e,
-/* 0x03cb: host_send_done */
-	0xf8ba0ef4,
-/* 0x03cd: host_recv */
-	0x4917f100,
-	0x5413f14e,
-	0x06e1b852,
-/* 0x03db: host_recv_wait */
-	0xf1aa0bf4,
-	0xb604cc17,
-	0x11cf0614,
-	0xc827f100,
-	0x0624b604,
-	0xf00022cf,
-	0x12b80816,
-	0xe60bf406,
-	0xb60723c4,
-	0x30b70434,
-	0x3b8002f0,
-	0x023c8003,
-	0x80013d80,
-	0x20b6003e,
-	0x0f24f001,
-	0x04c807f1,
+/* 0x0402: init_proc */
+	0xf19858f7,
+	0x0016b001,
+	0xf9fa0bf4,
+	0x58f0b615,
+/* 0x0413: mulu32_32_64 */
+	0xf9f20ef4,
+	0xf920f910,
+	0x9540f930,
+	0xd29510e1,
+	0xbdc4bd10,
+	0xc0edffb4,
+	0xb9301dff,
+	0x34f10234,
+	0x34b6ffff,
+	0x1045b610,
+	0xbb00c3bb,
+	0xe2ff01b4,
+	0x0234b930,
+	0xffff34f1,
+	0xb61034b6,
+	0xc3bb1045,
+	0x01b4bb00,
+	0xbb3012ff,
+	0x40fc00b3,
+	0x20fc30fc,
+	0x00f810fc,
+/* 0x0464: host_send */
+	0x04b017f1,
+	0xcf0614b6,
+	0x27f10011,
+	0x24b604a0,
+	0x0022cf06,
+	0xf40612b8,
+	0x1ec4320b,
+	0x04ee9407,
+	0x0270e0b7,
+	0x9803eb98,
+	0xed9802ec,
+	0x00ee9801,
+	0x034221f5,
+	0xc40110b6,
+	0x07f10f1e,
+	0x04b604b0,
+	0x000ed006,
+	0x0ef404bd,
+/* 0x04ad: host_send_done */
+/* 0x04af: host_recv */
+	0xf100f8ba,
+	0xf14e4917,
+	0xb8525413,
+	0x0bf406e1,
+/* 0x04bd: host_recv_wait */
+	0xcc17f1aa,
+	0x0614b604,
+	0xf10011cf,
+	0xb604c827,
+	0x22cf0624,
+	0x0816f000,
+	0xf40612b8,
+	0x23c4e60b,
+	0x0434b607,
+	0x02f030b7,
+	0x80033b80,
+	0x3d80023c,
+	0x003e8001,
+	0xf00120b6,
+	0x07f10f24,
+	0x04b604c8,
+	0x0002d006,
+	0x27f004bd,
+	0x0007f040,
 	0xd00604b6,
 	0x04bd0002,
-	0xf04027f0,
-	0x04b60007,
-	0x0002d006,
-	0x00f804bd,
-/* 0x0430: host_init */
-	0x008017f1,
-	0xf11014b6,
-	0xf1027015,
-	0xb604d007,
-	0x01d00604,
-	0xf104bd00,
-	0xb6008017,
-	0x15f11014,
-	0x07f102f0,
-	0x04b604dc,
-	0x0001d006,
-	0x17f004bd,
-	0xc407f101,
+/* 0x0512: host_init */
+	0x17f100f8,
+	0x14b60080,
+	0x7015f110,
+	0xd007f102,
 	0x0604b604,
 	0xbd0001d0,
-/* 0x046f: memx_func_enter */
-	0xf000f804,
+	0x8017f104,
+	0x1014b600,
+	0x02f015f1,
+	0x04dc07f1,
+	0xd00604b6,
+	0x04bd0001,
+	0xf10117f0,
+	0xb604c407,
+	0x01d00604,
+	0xf804bd00,
+/* 0x0551: memx_func_enter */
+	0x1087f100,
+	0x028eb916,
+	0xb90421f4,
+	0x67f102d7,
+	0x63f1fffc,
+	0x76fdffff,
+	0x0267f104,
+	0x0576fd00,
+	0x70f980f9,
+	0xe0fcd0fc,
+	0xf03f21f4,
 	0x07f10467,
 	0x04b607e0,
 	0x0006d006,
-/* 0x047e: memx_func_enter_wait */
+/* 0x058a: memx_func_enter_wait */
 	0x67f104bd,
 	0x64b607c0,
 	0x0066cf06,
 	0xf40464f0,
-	0x1698f30b,
-	0x0410b600,
-/* 0x0496: memx_func_leave */
-	0x67f000f8,
-	0xe407f104,
-	0x0604b607,
-	0xbd0006d0,
-/* 0x04a5: memx_func_leave_wait */
-	0xc067f104,
+	0x67f0f30b,
+	0x0664b62c,
+	0x800066cf,
+	0x00f8f106,
+/* 0x05a8: memx_func_leave */
+	0xb62c67f0,
+	0x66cf0664,
+	0xf2068000,
+	0xf10467f0,
+	0xb607e407,
+	0x06d00604,
+/* 0x05c3: memx_func_leave_wait */
+	0xf104bd00,
+	0xb607c067,
+	0x66cf0664,
+	0x0464f000,
+	0xf1f31bf4,
+	0xb9161087,
+	0x21f4028e,
+	0x02d7b904,
+	0xffcc67f1,
+	0xffff63f1,
+	0xf90476fd,
+	0xfc70f980,
+	0xf4e0fcd0,
+	0x00f83f21,
+/* 0x05f8: memx_func_wait_vblank */
+	0xb0001698,
+	0x0bf40066,
+	0x0166b013,
+	0xf4060bf4,
+/* 0x060a: memx_func_wait_vblank_head1 */
+	0x77f12e0e,
+	0x0ef40020,
+/* 0x0611: memx_func_wait_vblank_head0 */
+	0x0877f107,
+/* 0x0615: memx_func_wait_vblank_0 */
+	0xc467f100,
 	0x0664b607,
-	0xf00066cf,
-	0x1bf40464,
-/* 0x04b7: memx_func_wr32 */
-	0x9800f8f3,
-	0x15980016,
-	0x0810b601,
-	0x50f960f9,
+	0xfd0066cf,
+	0x1bf40467,
+/* 0x0625: memx_func_wait_vblank_1 */
+	0xc467f1f3,
+	0x0664b607,
+	0xfd0066cf,
+	0x0bf40467,
+/* 0x0635: memx_func_wait_vblank_fini */
+	0x0410b6f3,
+/* 0x063a: memx_func_wr32 */
+	0x169800f8,
+	0x01159800,
+	0xf90810b6,
+	0xfc50f960,
+	0xf4e0fcd0,
+	0x42b63f21,
+	0xe91bf402,
+/* 0x0656: memx_func_wait */
+	0x87f000f8,
+	0x0684b62c,
+	0x980088cf,
+	0x1d98001e,
+	0x021c9801,
+	0xb6031b98,
+	0x21f41010,
+/* 0x0673: memx_func_delay */
+	0x9800f8a4,
+	0x10b6001e,
+	0x7f21f404,
+/* 0x067e: memx_func_train */
+	0x57f100f8,
+	0x77f10003,
+	0x97f10000,
+	0x93f00000,
+	0x029eb970,
+	0xb90421f4,
+	0xe7f102d8,
+	0x21f42710,
+/* 0x069d: memx_func_train_loop_outer */
+	0x0158e07f,
+	0x0083f101,
+	0xe097f102,
+	0x1193f011,
+	0x80f990f9,
 	0xe0fcd0fc,
-	0xb63f21f4,
-	0x1bf40242,
-/* 0x04d3: memx_func_wait */
-	0xf000f8e9,
-	0x84b62c87,
-	0x0088cf06,
-	0x98001e98,
-	0x1c98011d,
-	0x031b9802,
-	0xf41010b6,
-	0x00f89c21,
-/* 0x04f0: memx_func_delay */
-	0xb6001e98,
-	0x21f40410,
-/* 0x04fb: memx_exec */
-	0xf900f87f,
+	0xf93f21f4,
+	0x0067f150,
+/* 0x06bd: memx_func_train_loop_inner */
+	0x1187f100,
+	0x9068ff11,
+	0xfd109894,
+	0x97f10589,
+	0x93f00720,
+	0xf990f910,
+	0xfcd0fc80,
+	0x3f21f4e0,
+	0x008097f1,
+	0xb91093f0,
+	0x21f4029e,
+	0x02d8b904,
+	0xf92088c5,
+	0xfc80f990,
+	0xf4e0fcd0,
+	0x97f13f21,
+	0x93f0053c,
+	0x0287f110,
+	0x0083f130,
+	0xf990f980,
+	0xfcd0fc80,
+	0x3f21f4e0,
+	0x0560e7f1,
+	0xf110e3f0,
+	0xf10000d7,
+	0x908000d3,
+	0xb7f100dc,
+	0xb3f08480,
+	0xa421f41e,
+	0x000057f1,
+	0xffff97f1,
+	0x830093f1,
+/* 0x073c: memx_func_train_loop_4x */
+	0x0080a7f1,
+	0xb910a3f0,
+	0x21f402ae,
+	0x02d8b904,
+	0xffdfb7f1,
+	0xffffb3f1,
+	0xf9048bfd,
+	0xfc80f9a0,
+	0xf4e0fcd0,
+	0xa7f13f21,
+	0xa3f0053c,
+	0x0287f110,
+	0x0083f130,
+	0xf9a0f980,
+	0xfcd0fc80,
+	0x3f21f4e0,
+	0x0560e7f1,
+	0xf110e3f0,
+	0xf10000d7,
+	0xb98000d3,
+	0xb7f102dc,
+	0xb3f02710,
+	0xa421f400,
+	0xf402eeb9,
+	0xddb90421,
+	0x949dff02,
+	0x700150b6,
+	0x1ef40456,
+	0xcc7aa092,
+	0x00a9800b,
+	0xb60160b6,
+	0x66700470,
+	0x001ef510,
+	0xb650fcff,
+	0x56700150,
+	0xd41ef507,
+/* 0x07cf: memx_exec */
+	0xf900f8fe,
 	0xb9d0f9e0,
 	0xb2b902c1,
-/* 0x0505: memx_exec_next */
+/* 0x07d9: memx_exec_next */
 	0x00139802,
-	0x950410b6,
-	0x30f01034,
+	0xe70410b6,
+	0xe701f034,
+	0xb601e033,
+	0x30f00132,
 	0xde35980c,
 	0x12b855f9,
-	0xec1ef406,
-	0xe0fcd0fc,
-	0x02b921f5,
-/* 0x0526: memx_info */
-	0xc7f100f8,
-	0xb7f103ac,
-	0x21f50800,
-	0x00f802b9,
-/* 0x0534: memx_recv */
+	0xe41ef406,
+	0x98f10b98,
+	0xcbbbf20c,
+	0xc4b7f102,
+	0x06b4b607,
+	0xfc00bbcf,
+	0xf5e0fcd0,
+	0xf8034221,
+/* 0x0815: memx_info */
+	0x01c67000,
+/* 0x081b: memx_info_data */
+	0xf10e0bf4,
+	0xf103ccc7,
+	0xf40800b7,
+/* 0x0826: memx_info_train */
+	0xc7f10b0e,
+	0xb7f10bcc,
+/* 0x082e: memx_info_send */
+	0x21f50100,
+	0x00f80342,
+/* 0x0834: memx_recv */
 	0xf401d6b0,
-	0xd6b0c40b,
-	0xe90bf400,
-/* 0x0542: memx_init */
+	0xd6b0980b,
+	0xd80bf400,
+/* 0x0842: memx_init */
 	0x00f800f8,
-/* 0x0544: perf_recv */
-/* 0x0546: perf_init */
+/* 0x0844: perf_recv */
+/* 0x0846: perf_init */
 	0x00f800f8,
-/* 0x0548: i2c_drive_scl */
+/* 0x0848: i2c_drive_scl */
 	0xf40036b0,
 	0x07f1110b,
 	0x04b607e0,
 	0x0001d006,
 	0x00f804bd,
-/* 0x055c: i2c_drive_scl_lo */
+/* 0x085c: i2c_drive_scl_lo */
 	0x07e407f1,
 	0xd00604b6,
 	0x04bd0001,
-/* 0x056a: i2c_drive_sda */
+/* 0x086a: i2c_drive_sda */
 	0x36b000f8,
 	0x110bf400,
 	0x07e007f1,
 	0xd00604b6,
 	0x04bd0002,
-/* 0x057e: i2c_drive_sda_lo */
+/* 0x087e: i2c_drive_sda_lo */
 	0x07f100f8,
 	0x04b607e4,
 	0x0002d006,
 	0x00f804bd,
-/* 0x058c: i2c_sense_scl */
+/* 0x088c: i2c_sense_scl */
 	0xf10132f4,
 	0xb607c437,
 	0x33cf0634,
 	0x0431fd00,
 	0xf4060bf4,
-/* 0x05a2: i2c_sense_scl_done */
+/* 0x08a2: i2c_sense_scl_done */
 	0x00f80131,
-/* 0x05a4: i2c_sense_sda */
+/* 0x08a4: i2c_sense_sda */
 	0xf10132f4,
 	0xb607c437,
 	0x33cf0634,
 	0x0432fd00,
 	0xf4060bf4,
-/* 0x05ba: i2c_sense_sda_done */
+/* 0x08ba: i2c_sense_sda_done */
 	0x00f80131,
-/* 0x05bc: i2c_raise_scl */
+/* 0x08bc: i2c_raise_scl */
 	0x47f140f9,
 	0x37f00898,
 	0x4821f501,
-/* 0x05c9: i2c_raise_scl_wait */
-	0xe8e7f105,
+/* 0x08c9: i2c_raise_scl_wait */
+	0xe8e7f108,
 	0x7f21f403,
-	0x058c21f5,
+	0x088c21f5,
 	0xb60901f4,
 	0x1bf40142,
-/* 0x05dd: i2c_raise_scl_done */
+/* 0x08dd: i2c_raise_scl_done */
 	0xf840fcef,
-/* 0x05e1: i2c_start */
+/* 0x08e1: i2c_start */
 	0x8c21f500,
-	0x0d11f405,
-	0x05a421f5,
+	0x0d11f408,
+	0x08a421f5,
 	0xf40611f4,
-/* 0x05f2: i2c_start_rep */
+/* 0x08f2: i2c_start_rep */
 	0x37f0300e,
 	0x4821f500,
-	0x0137f005,
-	0x056a21f5,
+	0x0137f008,
+	0x086a21f5,
 	0xb60076bb,
 	0x50f90465,
 	0xbb046594,
 	0x50bd0256,
 	0xfc0475fd,
 	0xbc21f550,
-	0x0464b605,
-/* 0x061f: i2c_start_send */
+	0x0464b608,
+/* 0x091f: i2c_start_send */
 	0xf01f11f4,
 	0x21f50037,
-	0xe7f1056a,
+	0xe7f1086a,
 	0x21f41388,
 	0x0037f07f,
-	0x054821f5,
+	0x084821f5,
 	0x1388e7f1,
-/* 0x063b: i2c_start_out */
+/* 0x093b: i2c_start_out */
 	0xf87f21f4,
-/* 0x063d: i2c_stop */
+/* 0x093d: i2c_stop */
 	0x0037f000,
-	0x054821f5,
+	0x084821f5,
 	0xf50037f0,
-	0xf1056a21,
+	0xf1086a21,
 	0xf403e8e7,
 	0x37f07f21,
 	0x4821f501,
-	0x88e7f105,
+	0x88e7f108,
 	0x7f21f413,
 	0xf50137f0,
-	0xf1056a21,
+	0xf1086a21,
 	0xf41388e7,
 	0x00f87f21,
-/* 0x0670: i2c_bitw */
-	0x056a21f5,
+/* 0x0970: i2c_bitw */
+	0x086a21f5,
 	0x03e8e7f1,
 	0xbb7f21f4,
 	0x65b60076,
@@ -1338,18 +1617,18 @@ uint32_t nva3_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x05bc21f5,
+	0x08bc21f5,
 	0xf40464b6,
 	0xe7f11811,
 	0x21f41388,
 	0x0037f07f,
-	0x054821f5,
+	0x084821f5,
 	0x1388e7f1,
-/* 0x06af: i2c_bitw_out */
+/* 0x09af: i2c_bitw_out */
 	0xf87f21f4,
-/* 0x06b1: i2c_bitr */
+/* 0x09b1: i2c_bitr */
 	0x0137f000,
-	0x056a21f5,
+	0x086a21f5,
 	0x03e8e7f1,
 	0xbb7f21f4,
 	0x65b60076,
@@ -1357,19 +1636,19 @@ uint32_t nva3_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x05bc21f5,
+	0x08bc21f5,
 	0xf40464b6,
 	0x21f51b11,
-	0x37f005a4,
+	0x37f008a4,
 	0x4821f500,
-	0x88e7f105,
+	0x88e7f108,
 	0x7f21f413,
 	0xf4013cf0,
-/* 0x06f6: i2c_bitr_done */
+/* 0x09f6: i2c_bitr_done */
 	0x00f80131,
-/* 0x06f8: i2c_get_byte */
+/* 0x09f8: i2c_get_byte */
 	0xf00057f0,
-/* 0x06fe: i2c_get_byte_next */
+/* 0x09fe: i2c_get_byte_next */
 	0x54b60847,
 	0x0076bb01,
 	0xf90465b6,
@@ -1377,7 +1656,7 @@ uint32_t nva3_pwr_code[] = {
 	0xbd0256bb,
 	0x0475fd50,
 	0x21f550fc,
-	0x64b606b1,
+	0x64b609b1,
 	0x2b11f404,
 	0xb60553fd,
 	0x1bf40142,
@@ -1388,11 +1667,11 @@ uint32_t nva3_pwr_code[] = {
 	0x50bd0256,
 	0xfc0475fd,
 	0x7021f550,
-	0x0464b606,
-/* 0x0748: i2c_get_byte_done */
-/* 0x074a: i2c_put_byte */
+	0x0464b609,
+/* 0x0a48: i2c_get_byte_done */
+/* 0x0a4a: i2c_put_byte */
 	0x47f000f8,
-/* 0x074d: i2c_put_byte_next */
+/* 0x0a4d: i2c_put_byte_next */
 	0x0142b608,
 	0xbb3854ff,
 	0x65b60076,
@@ -1400,7 +1679,7 @@ uint32_t nva3_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x067021f5,
+	0x097021f5,
 	0xf40464b6,
 	0x46b03411,
 	0xd81bf400,
@@ -1410,20 +1689,20 @@ uint32_t nva3_pwr_code[] = {
 	0x50bd0256,
 	0xfc0475fd,
 	0xb121f550,
-	0x0464b606,
+	0x0464b609,
 	0xbb0f11f4,
 	0x36b00076,
 	0x061bf401,
-/* 0x07a3: i2c_put_byte_done */
+/* 0x0aa3: i2c_put_byte_done */
 	0xf80132f4,
-/* 0x07a5: i2c_addr */
+/* 0x0aa5: i2c_addr */
 	0x0076bb00,
 	0xf90465b6,
 	0x04659450,
 	0xbd0256bb,
 	0x0475fd50,
 	0x21f550fc,
-	0x64b605e1,
+	0x64b608e1,
 	0x2911f404,
 	0x012ec3e7,
 	0xfd0134b6,
@@ -1433,31 +1712,31 @@ uint32_t nva3_pwr_code[] = {
 	0x0256bb04,
 	0x75fd50bd,
 	0xf550fc04,
-	0xb6074a21,
-/* 0x07ea: i2c_addr_done */
+	0xb60a4a21,
+/* 0x0aea: i2c_addr_done */
 	0x00f80464,
-/* 0x07ec: i2c_acquire_addr */
+/* 0x0aec: i2c_acquire_addr */
 	0xb6f8cec7,
 	0xe0b702e4,
-	0xee980bfc,
-/* 0x07fb: i2c_acquire */
+	0xee980d1c,
+/* 0x0afb: i2c_acquire */
 	0xf500f800,
-	0xf407ec21,
+	0xf40aec21,
 	0xd9f00421,
 	0x3f21f403,
-/* 0x080a: i2c_release */
+/* 0x0b0a: i2c_release */
 	0x21f500f8,
-	0x21f407ec,
+	0x21f40aec,
 	0x03daf004,
 	0xf83f21f4,
-/* 0x0819: i2c_recv */
+/* 0x0b19: i2c_recv */
 	0x0132f400,
 	0xb6f8c1c7,
 	0x16b00214,
 	0x3a1ff528,
-	0xd413a001,
-	0x0032980b,
-	0x0bac13a0,
+	0xf413a001,
+	0x0032980c,
+	0x0ccc13a0,
 	0xf4003198,
 	0xd0f90231,
 	0xd0f9e0f9,
@@ -1469,7 +1748,7 @@ uint32_t nva3_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x07fb21f5,
+	0x0afb21f5,
 	0xfc0464b6,
 	0x00d6b0d0,
 	0x00b31bf5,
@@ -1479,7 +1758,7 @@ uint32_t nva3_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x07a521f5,
+	0x0aa521f5,
 	0xf50464b6,
 	0xc700d011,
 	0x76bbe0c5,
@@ -1488,7 +1767,7 @@ uint32_t nva3_pwr_code[] = {
 	0x0256bb04,
 	0x75fd50bd,
 	0xf550fc04,
-	0xb6074a21,
+	0xb60a4a21,
 	0x11f50464,
 	0x57f000ad,
 	0x0076bb01,
@@ -1497,7 +1776,7 @@ uint32_t nva3_pwr_code[] = {
 	0xbd0256bb,
 	0x0475fd50,
 	0x21f550fc,
-	0x64b607a5,
+	0x64b60aa5,
 	0x8a11f504,
 	0x0076bb00,
 	0xf90465b6,
@@ -1505,7 +1784,7 @@ uint32_t nva3_pwr_code[] = {
 	0xbd0256bb,
 	0x0475fd50,
 	0x21f550fc,
-	0x64b606f8,
+	0x64b609f8,
 	0x6a11f404,
 	0xbbe05bcb,
 	0x65b60076,
@@ -1513,38 +1792,38 @@ uint32_t nva3_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x063d21f5,
+	0x093d21f5,
 	0xb90464b6,
 	0x74bd025b,
-/* 0x091f: i2c_recv_not_rd08 */
+/* 0x0c1f: i2c_recv_not_rd08 */
 	0xb0430ef4,
 	0x1bf401d6,
 	0x0057f03d,
-	0x07a521f5,
+	0x0aa521f5,
 	0xc73311f4,
 	0x21f5e0c5,
-	0x11f4074a,
+	0x11f40a4a,
 	0x0057f029,
-	0x07a521f5,
+	0x0aa521f5,
 	0xc71f11f4,
 	0x21f5e0b5,
-	0x11f4074a,
+	0x11f40a4a,
 	0x3d21f515,
-	0xc774bd06,
+	0xc774bd09,
 	0x1bf408c5,
 	0x0232f409,
-/* 0x095f: i2c_recv_not_wr08 */
-/* 0x095f: i2c_recv_done */
+/* 0x0c5f: i2c_recv_not_wr08 */
+/* 0x0c5f: i2c_recv_done */
 	0xc7030ef4,
 	0x21f5f8ce,
-	0xe0fc080a,
+	0xe0fc0b0a,
 	0x12f4d0fc,
 	0x027cb90a,
-	0x02b921f5,
-/* 0x0974: i2c_recv_exit */
-/* 0x0976: i2c_init */
+	0x034221f5,
+/* 0x0c74: i2c_recv_exit */
+/* 0x0c76: i2c_init */
 	0x00f800f8,
-/* 0x0978: test_recv */
+/* 0x0c78: test_recv */
 	0x05d817f1,
 	0xcf0614b6,
 	0x10b60011,
@@ -1553,13 +1832,13 @@ uint32_t nva3_pwr_code[] = {
 	0xbd0001d0,
 	0x00e7f104,
 	0x4fe3f1d9,
-	0xf521f513,
-/* 0x099f: test_init */
-	0xf100f801,
+	0x6221f513,
+/* 0x0c9f: test_init */
+	0xf100f802,
 	0xf50800e7,
-	0xf801f521,
-/* 0x09a9: idle_recv */
-/* 0x09ab: idle */
+	0xf8026221,
+/* 0x0ca9: idle_recv */
+/* 0x0cab: idle */
 	0xf400f800,
 	0x17f10031,
 	0x14b605d4,
@@ -1567,17 +1846,17 @@ uint32_t nva3_pwr_code[] = {
 	0xf10110b6,
 	0xb605d407,
 	0x01d00604,
-/* 0x09c7: idle_loop */
+/* 0x0cc7: idle_loop */
 	0xf004bd00,
 	0x32f45817,
-/* 0x09cd: idle_proc */
-/* 0x09cd: idle_proc_exec */
+/* 0x0ccd: idle_proc */
+/* 0x0ccd: idle_proc_exec */
 	0xb910f902,
 	0x21f5021e,
-	0x10fc02c2,
+	0x10fc034b,
 	0xf40911f4,
 	0x0ef40231,
-/* 0x09e1: idle_proc_next */
+/* 0x0ce1: idle_proc_next */
 	0x5810b6ef,
 	0xf4061fb8,
 	0x02f4e61b,
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc
index 48f79434a449..21bf8cc7618f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc
@@ -23,6 +23,7 @@
  */
 
 #define NVKM_PPWR_CHIPSET GF100
+#define HW_TICKS_PER_US 203 // should be 202.5
 
 //#define NVKM_FALCON_PC24
 //#define NVKM_FALCON_UNSHIFTED_IO
@@ -34,6 +35,7 @@
 .section #nvc0_pwr_data
 #define INCLUDE_PROC
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
@@ -44,6 +46,7 @@
 
 #define INCLUDE_DATA
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
@@ -56,6 +59,7 @@
 .section #nvc0_pwr_code
 #define INCLUDE_CODE
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc.h b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc.h
index 0773ff0e3dc3..90221d973f84 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc.h
@@ -24,8 +24,8 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 /* 0x0058: proc_list_head */
 	0x54534f48,
-	0x00000430,
-	0x000003cd,
+	0x00000512,
+	0x000004af,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -46,8 +46,8 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x584d454d,
-	0x00000542,
-	0x00000534,
+	0x0000075e,
+	0x00000750,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -68,8 +68,8 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x46524550,
-	0x00000546,
-	0x00000544,
+	0x00000762,
+	0x00000760,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -90,8 +90,8 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x5f433249,
-	0x00000976,
-	0x00000819,
+	0x00000b92,
+	0x00000a35,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -112,8 +112,8 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x54534554,
-	0x0000099f,
-	0x00000978,
+	0x00000bbb,
+	0x00000b94,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -134,8 +134,8 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x454c4449,
-	0x000009ab,
-	0x000009a9,
+	0x00000bc7,
+	0x00000bc5,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -227,24 +227,74 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 /* 0x0370: memx_func_head */
-	0x00010000,
-	0x00000000,
-	0x0000046f,
-/* 0x037c: memx_func_next */
 	0x00000001,
 	0x00000000,
-	0x00000496,
+	0x00000551,
+/* 0x037c: memx_func_next */
 	0x00000002,
+	0x00000000,
+	0x000005db,
+	0x00000003,
 	0x00000002,
-	0x000004b7,
-	0x00040003,
+	0x000006a5,
+	0x00040004,
+	0x00000000,
+	0x000006c1,
+	0x00010005,
+	0x00000000,
+	0x000006de,
+	0x00010006,
+	0x00000000,
+	0x00000663,
+	0x00000007,
+	0x00000000,
+	0x000006e9,
+/* 0x03c4: memx_func_tail */
+/* 0x03c4: memx_ts_start */
+	0x00000000,
+/* 0x03c8: memx_ts_end */
+	0x00000000,
+/* 0x03cc: memx_data_head */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
-	0x000004d3,
-	0x00010004,
 	0x00000000,
-	0x000004f0,
-/* 0x03ac: memx_func_tail */
-/* 0x03ac: memx_data_head */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -717,6 +767,8 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0bcc: memx_data_tail */
+/* 0x0bcc: memx_train_head */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -757,8 +809,32 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0bac: memx_data_tail */
-/* 0x0bac: i2c_scl_map */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+/* 0x0ccc: memx_train_tail */
+/* 0x0ccc: i2c_scl_map */
 	0x00001000,
 	0x00004000,
 	0x00010000,
@@ -769,7 +845,7 @@ uint32_t nvc0_pwr_data[] = {
 	0x01000000,
 	0x04000000,
 	0x10000000,
-/* 0x0bd4: i2c_sda_map */
+/* 0x0cf4: i2c_sda_map */
 	0x00002000,
 	0x00008000,
 	0x00020000,
@@ -780,7 +856,7 @@ uint32_t nvc0_pwr_data[] = {
 	0x02000000,
 	0x08000000,
 	0x20000000,
-/* 0x0bfc: i2c_ctrl */
+/* 0x0d1c: i2c_ctrl */
 	0x0000e138,
 	0x0000e150,
 	0x0000e168,
@@ -838,18 +914,10 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
 };
 
 uint32_t nvc0_pwr_code[] = {
-	0x030d0ef5,
+	0x039e0ef5,
 /* 0x0004: rd32 */
 	0x07a007f1,
 	0xd00604b6,
@@ -885,19 +953,22 @@ uint32_t nvc0_pwr_code[] = {
 	0xd4f100dd,
 	0x1bf47000,
 /* 0x007f: nsec */
-	0xf000f8f2,
+	0xf900f8f2,
+	0xf080f990,
 	0x84b62c87,
 	0x0088cf06,
-/* 0x0088: nsec_loop */
+/* 0x008c: nsec_loop */
 	0xb62c97f0,
 	0x99cf0694,
 	0x0298bb00,
 	0xf4069eb8,
-	0x00f8f11e,
-/* 0x009c: wait */
+	0x80fcf11e,
+	0x00f890fc,
+/* 0x00a4: wait */
+	0x80f990f9,
 	0xb62c87f0,
 	0x88cf0684,
-/* 0x00a5: wait_loop */
+/* 0x00b1: wait_loop */
 	0x02eeb900,
 	0xb90421f4,
 	0xadfd02da,
@@ -907,28 +978,29 @@ uint32_t nvc0_pwr_code[] = {
 	0x0099cf06,
 	0xb80298bb,
 	0x1ef4069b,
-/* 0x00c9: wait_done */
-/* 0x00cb: intr_watchdog */
-	0x9800f8df,
+/* 0x00d5: wait_done */
+	0xfc80fcdf,
+/* 0x00db: intr_watchdog */
+	0x9800f890,
 	0x96b003e9,
 	0x2a0bf400,
 	0xbb9a0a98,
 	0x1cf4029a,
 	0x01d7f00f,
-	0x025421f5,
+	0x02dd21f5,
 	0x0ef494bd,
-/* 0x00e9: intr_watchdog_next_time */
+/* 0x00f9: intr_watchdog_next_time */
 	0x9b0a9815,
 	0xf400a6b0,
 	0x9ab8090b,
 	0x061cf406,
-/* 0x00f8: intr_watchdog_next_time_set */
-/* 0x00fb: intr_watchdog_next_proc */
+/* 0x0108: intr_watchdog_next_time_set */
+/* 0x010b: intr_watchdog_next_proc */
 	0x809b0980,
 	0xe0b603e9,
 	0x68e6b158,
 	0xc61bf402,
-/* 0x010a: intr */
+/* 0x011a: intr */
 	0x00f900f8,
 	0x80f904bd,
 	0xa0f990f9,
@@ -948,13 +1020,13 @@ uint32_t nvc0_pwr_code[] = {
 	0xf40289c4,
 	0x0080230b,
 	0x58e7f09b,
-	0x98cb21f4,
+	0x98db21f4,
 	0x96b09b09,
 	0x110bf400,
 	0xb63407f0,
 	0x09d00604,
 	0x8004bd00,
-/* 0x016e: intr_skip_watchdog */
+/* 0x017e: intr_skip_watchdog */
 	0x89e49a09,
 	0x0bf40800,
 	0x8897f148,
@@ -967,22 +1039,22 @@ uint32_t nvc0_pwr_code[] = {
 	0x48e7f1c0,
 	0x53e3f14f,
 	0x00d7f054,
-	0x02b921f5,
+	0x034221f5,
 	0x07f1c0fc,
 	0x04b604c0,
 	0x000cd006,
-/* 0x01ae: intr_subintr_skip_fifo */
+/* 0x01be: intr_subintr_skip_fifo */
 	0x07f104bd,
 	0x04b60688,
 	0x0009d006,
-/* 0x01ba: intr_skip_subintr */
+/* 0x01ca: intr_skip_subintr */
 	0x89c404bd,
 	0x070bf420,
 	0xffbfa4f1,
-/* 0x01c4: intr_skip_pause */
+/* 0x01d4: intr_skip_pause */
 	0xf44089c4,
 	0xa4f1070b,
-/* 0x01ce: intr_skip_user0 */
+/* 0x01de: intr_skip_user0 */
 	0x07f0ffbf,
 	0x0604b604,
 	0xbd0008d0,
@@ -993,344 +1065,491 @@ uint32_t nvc0_pwr_code[] = {
 	0x90fca0fc,
 	0x00fc80fc,
 	0xf80032f4,
-/* 0x01f5: timer */
-	0x1032f401,
-	0xb003f898,
-	0x1cf40086,
-	0x03fe8051,
+/* 0x0205: ticks_from_ns */
+	0xf9c0f901,
+	0xcbd7f1b0,
+	0x00d3f000,
+	0x041321f5,
+	0x03e8ccec,
+	0xf400b4b0,
+	0xeeec120b,
+	0xd7f103e8,
+	0xd3f000cb,
+	0x1321f500,
+/* 0x022d: ticks_from_ns_quit */
+	0x02ceb904,
+	0xc0fcb0fc,
+/* 0x0236: ticks_from_us */
+	0xc0f900f8,
+	0xd7f1b0f9,
+	0xd3f000cb,
+	0x1321f500,
+	0x02ceb904,
+	0xf400b4b0,
+	0xe4bd050b,
+/* 0x0250: ticks_from_us_quit */
+	0xc0fcb0fc,
+/* 0x0256: ticks_to_us */
+	0xd7f100f8,
+	0xd3f000cb,
+	0xecedff00,
+/* 0x0262: timer */
+	0x90f900f8,
+	0x32f480f9,
+	0x03f89810,
+	0xf40086b0,
+	0x84bd651c,
 	0xb63807f0,
 	0x08d00604,
 	0xf004bd00,
-	0x84b60887,
+	0x84b63487,
 	0x0088cf06,
-	0xf40284f0,
-	0x87f0261b,
-	0x0684b634,
-	0xb80088cf,
-	0x0bf406e0,
-	0x06e8b809,
-/* 0x0233: timer_reset */
-	0xf01f1ef4,
-	0x04b63407,
-	0x000ed006,
-	0x0e8004bd,
-/* 0x0241: timer_enable */
-	0x0187f09a,
+	0xbb9a0998,
+	0xe9bb0298,
+	0x03fe8000,
+	0xb60887f0,
+	0x88cf0684,
+	0x0284f000,
+	0xf0261bf4,
+	0x84b63487,
+	0x0088cf06,
+	0xf406e0b8,
+	0xe8b8090b,
+	0x111cf406,
+/* 0x02b8: timer_reset */
+	0xb63407f0,
+	0x0ed00604,
+	0x8004bd00,
+/* 0x02c6: timer_enable */
+	0x87f09a0e,
+	0x3807f001,
+	0xd00604b6,
+	0x04bd0008,
+/* 0x02d4: timer_done */
+	0xfc1031f4,
+	0xf890fc80,
+/* 0x02dd: send_proc */
+	0xf980f900,
+	0x05e89890,
+	0xf004e998,
+	0x89b80486,
+	0x2a0bf406,
+	0x940398c4,
+	0x80b60488,
+	0x008ebb18,
+	0x8000fa98,
+	0x8d80008a,
+	0x028c8001,
+	0xb6038b80,
+	0x94f00190,
+	0x04e98007,
+/* 0x0317: send_done */
+	0xfc0231f4,
+	0xf880fc90,
+/* 0x031d: find */
+	0xf080f900,
+	0x31f45887,
+/* 0x0325: find_loop */
+	0x008a9801,
+	0xf406aeb8,
+	0x80b6100b,
+	0x6886b158,
+	0xf01bf402,
+/* 0x033b: find_done */
+	0xb90132f4,
+	0x80fc028e,
+/* 0x0342: send */
+	0x21f500f8,
+	0x01f4031d,
+/* 0x034b: recv */
+	0xf900f897,
+	0x9880f990,
+	0xe99805e8,
+	0x0132f404,
+	0xf40689b8,
+	0x89c43d0b,
+	0x0180b603,
+	0x800784f0,
+	0xea9805e8,
+	0xfef0f902,
+	0xf0f9018f,
+	0x9402efb9,
+	0xe9bb0499,
+	0x18e0b600,
+	0x9803eb98,
+	0xed9802ec,
+	0x00ee9801,
+	0xf0fca5f9,
+	0xf400f8fe,
+	0xf0fc0131,
+/* 0x0398: recv_done */
+	0x90fc80fc,
+/* 0x039e: init */
+	0x17f100f8,
+	0x14b60108,
+	0x0011cf06,
+	0x010911e7,
+	0xfe0814b6,
+	0x17f10014,
+	0x13f000e0,
+	0x1c07f000,
+	0xd00604b6,
+	0x04bd0001,
+	0xf0ff17f0,
+	0x04b61407,
+	0x0001d006,
+	0x17f004bd,
+	0x0015f102,
+	0x1007f008,
+	0xd00604b6,
+	0x04bd0001,
+	0x011a17f1,
+	0xfe0013f0,
+	0x31f40010,
+	0x0117f010,
 	0xb63807f0,
-	0x08d00604,
-/* 0x024f: timer_done */
-	0xf404bd00,
-	0x00f81031,
-/* 0x0254: send_proc */
-	0x90f980f9,
-	0x9805e898,
-	0x86f004e9,
-	0x0689b804,
-	0xc42a0bf4,
-	0x88940398,
-	0x1880b604,
-	0x98008ebb,
-	0x8a8000fa,
-	0x018d8000,
-	0x80028c80,
-	0x90b6038b,
-	0x0794f001,
-	0xf404e980,
-/* 0x028e: send_done */
-	0x90fc0231,
-	0x00f880fc,
-/* 0x0294: find */
-	0x87f080f9,
-	0x0131f458,
-/* 0x029c: find_loop */
-	0xb8008a98,
-	0x0bf406ae,
-	0x5880b610,
-	0x026886b1,
-	0xf4f01bf4,
-/* 0x02b2: find_done */
-	0x8eb90132,
-	0xf880fc02,
-/* 0x02b9: send */
-	0x9421f500,
-	0x9701f402,
-/* 0x02c2: recv */
-	0xe89800f8,
-	0x04e99805,
-	0xb80132f4,
-	0x0bf40689,
-	0x0389c43d,
-	0xf00180b6,
-	0xe8800784,
-	0x02ea9805,
-	0x8ffef0f9,
-	0xb9f0f901,
-	0x999402ef,
-	0x00e9bb04,
-	0x9818e0b6,
-	0xec9803eb,
-	0x01ed9802,
-	0xf900ee98,
-	0xfef0fca5,
-	0x31f400f8,
-/* 0x030b: recv_done */
-	0xf8f0fc01,
-/* 0x030d: init */
-	0x0817f100,
-	0x0614b601,
-	0xe70011cf,
-	0xb6010911,
-	0x14fe0814,
-	0xe017f100,
-	0x0013f000,
-	0xb61c07f0,
 	0x01d00604,
 	0xf004bd00,
-	0x07f0ff17,
-	0x0604b614,
-	0xbd0001d0,
-	0x0217f004,
-	0x080015f1,
-	0xb61007f0,
-	0x01d00604,
-	0xf104bd00,
-	0xf0010a17,
-	0x10fe0013,
-	0x1031f400,
-	0xf00117f0,
-	0x04b63807,
-	0x0001d006,
-	0xf7f004bd,
-/* 0x0371: init_proc */
-	0x01f19858,
-	0xf40016b0,
-	0x15f9fa0b,
-	0xf458f0b6,
-/* 0x0382: host_send */
-	0x17f1f20e,
-	0x14b604b0,
-	0x0011cf06,
-	0x04a027f1,
-	0xcf0624b6,
-	0x12b80022,
-	0x320bf406,
-	0x94071ec4,
-	0xe0b704ee,
-	0xeb980270,
-	0x02ec9803,
-	0x9801ed98,
-	0x21f500ee,
-	0x10b602b9,
-	0x0f1ec401,
-	0x04b007f1,
-	0xd00604b6,
-	0x04bd000e,
-/* 0x03cb: host_send_done */
-	0xf8ba0ef4,
-/* 0x03cd: host_recv */
-	0x4917f100,
-	0x5413f14e,
-	0x06e1b852,
-/* 0x03db: host_recv_wait */
-	0xf1aa0bf4,
-	0xb604cc17,
-	0x11cf0614,
-	0xc827f100,
-	0x0624b604,
-	0xf00022cf,
-	0x12b80816,
-	0xe60bf406,
-	0xb60723c4,
-	0x30b70434,
-	0x3b8002f0,
-	0x023c8003,
-	0x80013d80,
-	0x20b6003e,
-	0x0f24f001,
-	0x04c807f1,
+/* 0x0402: init_proc */
+	0xf19858f7,
+	0x0016b001,
+	0xf9fa0bf4,
+	0x58f0b615,
+/* 0x0413: mulu32_32_64 */
+	0xf9f20ef4,
+	0xf920f910,
+	0x9540f930,
+	0xd29510e1,
+	0xbdc4bd10,
+	0xc0edffb4,
+	0xb9301dff,
+	0x34f10234,
+	0x34b6ffff,
+	0x1045b610,
+	0xbb00c3bb,
+	0xe2ff01b4,
+	0x0234b930,
+	0xffff34f1,
+	0xb61034b6,
+	0xc3bb1045,
+	0x01b4bb00,
+	0xbb3012ff,
+	0x40fc00b3,
+	0x20fc30fc,
+	0x00f810fc,
+/* 0x0464: host_send */
+	0x04b017f1,
+	0xcf0614b6,
+	0x27f10011,
+	0x24b604a0,
+	0x0022cf06,
+	0xf40612b8,
+	0x1ec4320b,
+	0x04ee9407,
+	0x0270e0b7,
+	0x9803eb98,
+	0xed9802ec,
+	0x00ee9801,
+	0x034221f5,
+	0xc40110b6,
+	0x07f10f1e,
+	0x04b604b0,
+	0x000ed006,
+	0x0ef404bd,
+/* 0x04ad: host_send_done */
+/* 0x04af: host_recv */
+	0xf100f8ba,
+	0xf14e4917,
+	0xb8525413,
+	0x0bf406e1,
+/* 0x04bd: host_recv_wait */
+	0xcc17f1aa,
+	0x0614b604,
+	0xf10011cf,
+	0xb604c827,
+	0x22cf0624,
+	0x0816f000,
+	0xf40612b8,
+	0x23c4e60b,
+	0x0434b607,
+	0x02f030b7,
+	0x80033b80,
+	0x3d80023c,
+	0x003e8001,
+	0xf00120b6,
+	0x07f10f24,
+	0x04b604c8,
+	0x0002d006,
+	0x27f004bd,
+	0x0007f040,
 	0xd00604b6,
 	0x04bd0002,
-	0xf04027f0,
-	0x04b60007,
-	0x0002d006,
-	0x00f804bd,
-/* 0x0430: host_init */
-	0x008017f1,
-	0xf11014b6,
-	0xf1027015,
-	0xb604d007,
-	0x01d00604,
-	0xf104bd00,
-	0xb6008017,
-	0x15f11014,
-	0x07f102f0,
-	0x04b604dc,
-	0x0001d006,
-	0x17f004bd,
-	0xc407f101,
+/* 0x0512: host_init */
+	0x17f100f8,
+	0x14b60080,
+	0x7015f110,
+	0xd007f102,
 	0x0604b604,
 	0xbd0001d0,
-/* 0x046f: memx_func_enter */
-	0xf000f804,
+	0x8017f104,
+	0x1014b600,
+	0x02f015f1,
+	0x04dc07f1,
+	0xd00604b6,
+	0x04bd0001,
+	0xf10117f0,
+	0xb604c407,
+	0x01d00604,
+	0xf804bd00,
+/* 0x0551: memx_func_enter */
+	0x2067f100,
+	0x5d77f116,
+	0xff73f1f5,
+	0x026eb9ff,
+	0xb90421f4,
+	0x87fd02d8,
+	0xf960f904,
+	0xfcd0fc80,
+	0x3f21f4e0,
+	0xfffe77f1,
+	0xffff73f1,
+	0xf4026eb9,
+	0xd8b90421,
+	0x0487fd02,
+	0x80f960f9,
+	0xe0fcd0fc,
+	0xf13f21f4,
+	0xb926f067,
+	0x21f4026e,
+	0x02d8b904,
+	0xf90487fd,
+	0xfc80f960,
+	0xf4e0fcd0,
+	0x67f03f21,
+	0xe007f104,
+	0x0604b607,
+	0xbd0006d0,
+/* 0x05bd: memx_func_enter_wait */
+	0xc067f104,
+	0x0664b607,
+	0xf00066cf,
+	0x0bf40464,
+	0x2c67f0f3,
+	0xcf0664b6,
+	0x06800066,
+/* 0x05db: memx_func_leave */
+	0xf000f8f1,
+	0x64b62c67,
+	0x0066cf06,
+	0xf0f20680,
 	0x07f10467,
-	0x04b607e0,
+	0x04b607e4,
 	0x0006d006,
-/* 0x047e: memx_func_enter_wait */
+/* 0x05f6: memx_func_leave_wait */
 	0x67f104bd,
 	0x64b607c0,
 	0x0066cf06,
 	0xf40464f0,
-	0x1698f30b,
-	0x0410b600,
-/* 0x0496: memx_func_leave */
-	0x67f000f8,
-	0xe407f104,
-	0x0604b607,
-	0xbd0006d0,
-/* 0x04a5: memx_func_leave_wait */
-	0xc067f104,
-	0x0664b607,
-	0xf00066cf,
-	0x1bf40464,
-/* 0x04b7: memx_func_wr32 */
-	0x9800f8f3,
-	0x15980016,
-	0x0810b601,
-	0x50f960f9,
+	0x67f1f31b,
+	0x77f126f0,
+	0x73f00001,
+	0x026eb900,
+	0xb90421f4,
+	0x87fd02d8,
+	0xf960f905,
+	0xfcd0fc80,
+	0x3f21f4e0,
+	0x162067f1,
+	0xf4026eb9,
+	0xd8b90421,
+	0x0587fd02,
+	0x80f960f9,
 	0xe0fcd0fc,
-	0xb63f21f4,
-	0x1bf40242,
-/* 0x04d3: memx_func_wait */
-	0xf000f8e9,
-	0x84b62c87,
-	0x0088cf06,
-	0x98001e98,
-	0x1c98011d,
-	0x031b9802,
-	0xf41010b6,
-	0x00f89c21,
-/* 0x04f0: memx_func_delay */
-	0xb6001e98,
-	0x21f40410,
-/* 0x04fb: memx_exec */
-	0xf900f87f,
+	0xf13f21f4,
+	0xf00aa277,
+	0x6eb90073,
+	0x0421f402,
+	0xfd02d8b9,
+	0x60f90587,
+	0xd0fc80f9,
+	0x21f4e0fc,
+/* 0x0663: memx_func_wait_vblank */
+	0x9800f83f,
+	0x66b00016,
+	0x130bf400,
+	0xf40166b0,
+	0x0ef4060b,
+/* 0x0675: memx_func_wait_vblank_head1 */
+	0x2077f12e,
+	0x070ef400,
+/* 0x067c: memx_func_wait_vblank_head0 */
+	0x000877f1,
+/* 0x0680: memx_func_wait_vblank_0 */
+	0x07c467f1,
+	0xcf0664b6,
+	0x67fd0066,
+	0xf31bf404,
+/* 0x0690: memx_func_wait_vblank_1 */
+	0x07c467f1,
+	0xcf0664b6,
+	0x67fd0066,
+	0xf30bf404,
+/* 0x06a0: memx_func_wait_vblank_fini */
+	0xf80410b6,
+/* 0x06a5: memx_func_wr32 */
+	0x00169800,
+	0xb6011598,
+	0x60f90810,
+	0xd0fc50f9,
+	0x21f4e0fc,
+	0x0242b63f,
+	0xf8e91bf4,
+/* 0x06c1: memx_func_wait */
+	0x2c87f000,
+	0xcf0684b6,
+	0x1e980088,
+	0x011d9800,
+	0x98021c98,
+	0x10b6031b,
+	0xa421f410,
+/* 0x06de: memx_func_delay */
+	0x1e9800f8,
+	0x0410b600,
+	0xf87f21f4,
+/* 0x06e9: memx_func_train */
+/* 0x06eb: memx_exec */
+	0xf900f800,
 	0xb9d0f9e0,
 	0xb2b902c1,
-/* 0x0505: memx_exec_next */
+/* 0x06f5: memx_exec_next */
 	0x00139802,
-	0x950410b6,
-	0x30f01034,
+	0xe70410b6,
+	0xe701f034,
+	0xb601e033,
+	0x30f00132,
 	0xde35980c,
 	0x12b855f9,
-	0xec1ef406,
-	0xe0fcd0fc,
-	0x02b921f5,
-/* 0x0526: memx_info */
-	0xc7f100f8,
-	0xb7f103ac,
-	0x21f50800,
-	0x00f802b9,
-/* 0x0534: memx_recv */
+	0xe41ef406,
+	0x98f10b98,
+	0xcbbbf20c,
+	0xc4b7f102,
+	0x06b4b607,
+	0xfc00bbcf,
+	0xf5e0fcd0,
+	0xf8034221,
+/* 0x0731: memx_info */
+	0x01c67000,
+/* 0x0737: memx_info_data */
+	0xf10e0bf4,
+	0xf103ccc7,
+	0xf40800b7,
+/* 0x0742: memx_info_train */
+	0xc7f10b0e,
+	0xb7f10bcc,
+/* 0x074a: memx_info_send */
+	0x21f50100,
+	0x00f80342,
+/* 0x0750: memx_recv */
 	0xf401d6b0,
-	0xd6b0c40b,
-	0xe90bf400,
-/* 0x0542: memx_init */
+	0xd6b0980b,
+	0xd80bf400,
+/* 0x075e: memx_init */
 	0x00f800f8,
-/* 0x0544: perf_recv */
-/* 0x0546: perf_init */
+/* 0x0760: perf_recv */
+/* 0x0762: perf_init */
 	0x00f800f8,
-/* 0x0548: i2c_drive_scl */
+/* 0x0764: i2c_drive_scl */
 	0xf40036b0,
 	0x07f1110b,
 	0x04b607e0,
 	0x0001d006,
 	0x00f804bd,
-/* 0x055c: i2c_drive_scl_lo */
+/* 0x0778: i2c_drive_scl_lo */
 	0x07e407f1,
 	0xd00604b6,
 	0x04bd0001,
-/* 0x056a: i2c_drive_sda */
+/* 0x0786: i2c_drive_sda */
 	0x36b000f8,
 	0x110bf400,
 	0x07e007f1,
 	0xd00604b6,
 	0x04bd0002,
-/* 0x057e: i2c_drive_sda_lo */
+/* 0x079a: i2c_drive_sda_lo */
 	0x07f100f8,
 	0x04b607e4,
 	0x0002d006,
 	0x00f804bd,
-/* 0x058c: i2c_sense_scl */
+/* 0x07a8: i2c_sense_scl */
 	0xf10132f4,
 	0xb607c437,
 	0x33cf0634,
 	0x0431fd00,
 	0xf4060bf4,
-/* 0x05a2: i2c_sense_scl_done */
+/* 0x07be: i2c_sense_scl_done */
 	0x00f80131,
-/* 0x05a4: i2c_sense_sda */
+/* 0x07c0: i2c_sense_sda */
 	0xf10132f4,
 	0xb607c437,
 	0x33cf0634,
 	0x0432fd00,
 	0xf4060bf4,
-/* 0x05ba: i2c_sense_sda_done */
+/* 0x07d6: i2c_sense_sda_done */
 	0x00f80131,
-/* 0x05bc: i2c_raise_scl */
+/* 0x07d8: i2c_raise_scl */
 	0x47f140f9,
 	0x37f00898,
-	0x4821f501,
-/* 0x05c9: i2c_raise_scl_wait */
-	0xe8e7f105,
+	0x6421f501,
+/* 0x07e5: i2c_raise_scl_wait */
+	0xe8e7f107,
 	0x7f21f403,
-	0x058c21f5,
+	0x07a821f5,
 	0xb60901f4,
 	0x1bf40142,
-/* 0x05dd: i2c_raise_scl_done */
+/* 0x07f9: i2c_raise_scl_done */
 	0xf840fcef,
-/* 0x05e1: i2c_start */
-	0x8c21f500,
-	0x0d11f405,
-	0x05a421f5,
+/* 0x07fd: i2c_start */
+	0xa821f500,
+	0x0d11f407,
+	0x07c021f5,
 	0xf40611f4,
-/* 0x05f2: i2c_start_rep */
+/* 0x080e: i2c_start_rep */
 	0x37f0300e,
-	0x4821f500,
-	0x0137f005,
-	0x056a21f5,
+	0x6421f500,
+	0x0137f007,
+	0x078621f5,
 	0xb60076bb,
 	0x50f90465,
 	0xbb046594,
 	0x50bd0256,
 	0xfc0475fd,
-	0xbc21f550,
-	0x0464b605,
-/* 0x061f: i2c_start_send */
+	0xd821f550,
+	0x0464b607,
+/* 0x083b: i2c_start_send */
 	0xf01f11f4,
 	0x21f50037,
-	0xe7f1056a,
+	0xe7f10786,
 	0x21f41388,
 	0x0037f07f,
-	0x054821f5,
+	0x076421f5,
 	0x1388e7f1,
-/* 0x063b: i2c_start_out */
+/* 0x0857: i2c_start_out */
 	0xf87f21f4,
-/* 0x063d: i2c_stop */
+/* 0x0859: i2c_stop */
 	0x0037f000,
-	0x054821f5,
+	0x076421f5,
 	0xf50037f0,
-	0xf1056a21,
+	0xf1078621,
 	0xf403e8e7,
 	0x37f07f21,
-	0x4821f501,
-	0x88e7f105,
+	0x6421f501,
+	0x88e7f107,
 	0x7f21f413,
 	0xf50137f0,
-	0xf1056a21,
+	0xf1078621,
 	0xf41388e7,
 	0x00f87f21,
-/* 0x0670: i2c_bitw */
-	0x056a21f5,
+/* 0x088c: i2c_bitw */
+	0x078621f5,
 	0x03e8e7f1,
 	0xbb7f21f4,
 	0x65b60076,
@@ -1338,18 +1557,18 @@ uint32_t nvc0_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x05bc21f5,
+	0x07d821f5,
 	0xf40464b6,
 	0xe7f11811,
 	0x21f41388,
 	0x0037f07f,
-	0x054821f5,
+	0x076421f5,
 	0x1388e7f1,
-/* 0x06af: i2c_bitw_out */
+/* 0x08cb: i2c_bitw_out */
 	0xf87f21f4,
-/* 0x06b1: i2c_bitr */
+/* 0x08cd: i2c_bitr */
 	0x0137f000,
-	0x056a21f5,
+	0x078621f5,
 	0x03e8e7f1,
 	0xbb7f21f4,
 	0x65b60076,
@@ -1357,19 +1576,19 @@ uint32_t nvc0_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x05bc21f5,
+	0x07d821f5,
 	0xf40464b6,
 	0x21f51b11,
-	0x37f005a4,
-	0x4821f500,
-	0x88e7f105,
+	0x37f007c0,
+	0x6421f500,
+	0x88e7f107,
 	0x7f21f413,
 	0xf4013cf0,
-/* 0x06f6: i2c_bitr_done */
+/* 0x0912: i2c_bitr_done */
 	0x00f80131,
-/* 0x06f8: i2c_get_byte */
+/* 0x0914: i2c_get_byte */
 	0xf00057f0,
-/* 0x06fe: i2c_get_byte_next */
+/* 0x091a: i2c_get_byte_next */
 	0x54b60847,
 	0x0076bb01,
 	0xf90465b6,
@@ -1377,7 +1596,7 @@ uint32_t nvc0_pwr_code[] = {
 	0xbd0256bb,
 	0x0475fd50,
 	0x21f550fc,
-	0x64b606b1,
+	0x64b608cd,
 	0x2b11f404,
 	0xb60553fd,
 	0x1bf40142,
@@ -1387,12 +1606,12 @@ uint32_t nvc0_pwr_code[] = {
 	0xbb046594,
 	0x50bd0256,
 	0xfc0475fd,
-	0x7021f550,
-	0x0464b606,
-/* 0x0748: i2c_get_byte_done */
-/* 0x074a: i2c_put_byte */
+	0x8c21f550,
+	0x0464b608,
+/* 0x0964: i2c_get_byte_done */
+/* 0x0966: i2c_put_byte */
 	0x47f000f8,
-/* 0x074d: i2c_put_byte_next */
+/* 0x0969: i2c_put_byte_next */
 	0x0142b608,
 	0xbb3854ff,
 	0x65b60076,
@@ -1400,7 +1619,7 @@ uint32_t nvc0_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x067021f5,
+	0x088c21f5,
 	0xf40464b6,
 	0x46b03411,
 	0xd81bf400,
@@ -1409,21 +1628,21 @@ uint32_t nvc0_pwr_code[] = {
 	0xbb046594,
 	0x50bd0256,
 	0xfc0475fd,
-	0xb121f550,
-	0x0464b606,
+	0xcd21f550,
+	0x0464b608,
 	0xbb0f11f4,
 	0x36b00076,
 	0x061bf401,
-/* 0x07a3: i2c_put_byte_done */
+/* 0x09bf: i2c_put_byte_done */
 	0xf80132f4,
-/* 0x07a5: i2c_addr */
+/* 0x09c1: i2c_addr */
 	0x0076bb00,
 	0xf90465b6,
 	0x04659450,
 	0xbd0256bb,
 	0x0475fd50,
 	0x21f550fc,
-	0x64b605e1,
+	0x64b607fd,
 	0x2911f404,
 	0x012ec3e7,
 	0xfd0134b6,
@@ -1433,31 +1652,31 @@ uint32_t nvc0_pwr_code[] = {
 	0x0256bb04,
 	0x75fd50bd,
 	0xf550fc04,
-	0xb6074a21,
-/* 0x07ea: i2c_addr_done */
+	0xb6096621,
+/* 0x0a06: i2c_addr_done */
 	0x00f80464,
-/* 0x07ec: i2c_acquire_addr */
+/* 0x0a08: i2c_acquire_addr */
 	0xb6f8cec7,
 	0xe0b702e4,
-	0xee980bfc,
-/* 0x07fb: i2c_acquire */
+	0xee980d1c,
+/* 0x0a17: i2c_acquire */
 	0xf500f800,
-	0xf407ec21,
+	0xf40a0821,
 	0xd9f00421,
 	0x3f21f403,
-/* 0x080a: i2c_release */
+/* 0x0a26: i2c_release */
 	0x21f500f8,
-	0x21f407ec,
+	0x21f40a08,
 	0x03daf004,
 	0xf83f21f4,
-/* 0x0819: i2c_recv */
+/* 0x0a35: i2c_recv */
 	0x0132f400,
 	0xb6f8c1c7,
 	0x16b00214,
 	0x3a1ff528,
-	0xd413a001,
-	0x0032980b,
-	0x0bac13a0,
+	0xf413a001,
+	0x0032980c,
+	0x0ccc13a0,
 	0xf4003198,
 	0xd0f90231,
 	0xd0f9e0f9,
@@ -1469,7 +1688,7 @@ uint32_t nvc0_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x07fb21f5,
+	0x0a1721f5,
 	0xfc0464b6,
 	0x00d6b0d0,
 	0x00b31bf5,
@@ -1479,7 +1698,7 @@ uint32_t nvc0_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x07a521f5,
+	0x09c121f5,
 	0xf50464b6,
 	0xc700d011,
 	0x76bbe0c5,
@@ -1488,7 +1707,7 @@ uint32_t nvc0_pwr_code[] = {
 	0x0256bb04,
 	0x75fd50bd,
 	0xf550fc04,
-	0xb6074a21,
+	0xb6096621,
 	0x11f50464,
 	0x57f000ad,
 	0x0076bb01,
@@ -1497,7 +1716,7 @@ uint32_t nvc0_pwr_code[] = {
 	0xbd0256bb,
 	0x0475fd50,
 	0x21f550fc,
-	0x64b607a5,
+	0x64b609c1,
 	0x8a11f504,
 	0x0076bb00,
 	0xf90465b6,
@@ -1505,7 +1724,7 @@ uint32_t nvc0_pwr_code[] = {
 	0xbd0256bb,
 	0x0475fd50,
 	0x21f550fc,
-	0x64b606f8,
+	0x64b60914,
 	0x6a11f404,
 	0xbbe05bcb,
 	0x65b60076,
@@ -1513,38 +1732,38 @@ uint32_t nvc0_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x063d21f5,
+	0x085921f5,
 	0xb90464b6,
 	0x74bd025b,
-/* 0x091f: i2c_recv_not_rd08 */
+/* 0x0b3b: i2c_recv_not_rd08 */
 	0xb0430ef4,
 	0x1bf401d6,
 	0x0057f03d,
-	0x07a521f5,
+	0x09c121f5,
 	0xc73311f4,
 	0x21f5e0c5,
-	0x11f4074a,
+	0x11f40966,
 	0x0057f029,
-	0x07a521f5,
+	0x09c121f5,
 	0xc71f11f4,
 	0x21f5e0b5,
-	0x11f4074a,
-	0x3d21f515,
-	0xc774bd06,
+	0x11f40966,
+	0x5921f515,
+	0xc774bd08,
 	0x1bf408c5,
 	0x0232f409,
-/* 0x095f: i2c_recv_not_wr08 */
-/* 0x095f: i2c_recv_done */
+/* 0x0b7b: i2c_recv_not_wr08 */
+/* 0x0b7b: i2c_recv_done */
 	0xc7030ef4,
 	0x21f5f8ce,
-	0xe0fc080a,
+	0xe0fc0a26,
 	0x12f4d0fc,
 	0x027cb90a,
-	0x02b921f5,
-/* 0x0974: i2c_recv_exit */
-/* 0x0976: i2c_init */
+	0x034221f5,
+/* 0x0b90: i2c_recv_exit */
+/* 0x0b92: i2c_init */
 	0x00f800f8,
-/* 0x0978: test_recv */
+/* 0x0b94: test_recv */
 	0x05d817f1,
 	0xcf0614b6,
 	0x10b60011,
@@ -1553,13 +1772,13 @@ uint32_t nvc0_pwr_code[] = {
 	0xbd0001d0,
 	0x00e7f104,
 	0x4fe3f1d9,
-	0xf521f513,
-/* 0x099f: test_init */
-	0xf100f801,
+	0x6221f513,
+/* 0x0bbb: test_init */
+	0xf100f802,
 	0xf50800e7,
-	0xf801f521,
-/* 0x09a9: idle_recv */
-/* 0x09ab: idle */
+	0xf8026221,
+/* 0x0bc5: idle_recv */
+/* 0x0bc7: idle */
 	0xf400f800,
 	0x17f10031,
 	0x14b605d4,
@@ -1567,17 +1786,17 @@ uint32_t nvc0_pwr_code[] = {
 	0xf10110b6,
 	0xb605d407,
 	0x01d00604,
-/* 0x09c7: idle_loop */
+/* 0x0be3: idle_loop */
 	0xf004bd00,
 	0x32f45817,
-/* 0x09cd: idle_proc */
-/* 0x09cd: idle_proc_exec */
+/* 0x0be9: idle_proc */
+/* 0x0be9: idle_proc_exec */
 	0xb910f902,
 	0x21f5021e,
-	0x10fc02c2,
+	0x10fc034b,
 	0xf40911f4,
 	0x0ef40231,
-/* 0x09e1: idle_proc_next */
+/* 0x0bfd: idle_proc_next */
 	0x5810b6ef,
 	0xf4061fb8,
 	0x02f4e61b,
@@ -1586,4 +1805,61 @@ uint32_t nvc0_pwr_code[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 };
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc
index 8a89dfe41ce1..b85443261569 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc
@@ -23,6 +23,7 @@
  */
 
 #define NVKM_PPWR_CHIPSET GF119
+#define HW_TICKS_PER_US 324
 
 //#define NVKM_FALCON_PC24
 #define NVKM_FALCON_UNSHIFTED_IO
@@ -34,6 +35,7 @@
 .section #nvd0_pwr_data
 #define INCLUDE_PROC
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
@@ -44,6 +46,7 @@
 
 #define INCLUDE_DATA
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
@@ -56,6 +59,7 @@
 .section #nvd0_pwr_code
 #define INCLUDE_CODE
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc.h b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc.h
index 8d369b3faaba..7e16aab44d85 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc.h
@@ -24,8 +24,8 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 /* 0x0058: proc_list_head */
 	0x54534f48,
-	0x000003be,
-	0x00000367,
+	0x0000049d,
+	0x00000446,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -46,8 +46,8 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x584d454d,
-	0x000004b8,
-	0x000004aa,
+	0x0000068b,
+	0x0000067d,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -68,8 +68,8 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x46524550,
-	0x000004bc,
-	0x000004ba,
+	0x0000068f,
+	0x0000068d,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -90,8 +90,8 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x5f433249,
-	0x000008d7,
-	0x0000077a,
+	0x00000aaa,
+	0x0000094d,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -112,8 +112,8 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x54534554,
-	0x000008fa,
-	0x000008d9,
+	0x00000acd,
+	0x00000aac,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -134,8 +134,8 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x454c4449,
-	0x00000906,
-	0x00000904,
+	0x00000ad9,
+	0x00000ad7,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -227,24 +227,63 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 /* 0x0370: memx_func_head */
-	0x00010000,
-	0x00000000,
-	0x000003f4,
-/* 0x037c: memx_func_next */
 	0x00000001,
 	0x00000000,
-	0x00000415,
+	0x000004d3,
+/* 0x037c: memx_func_next */
 	0x00000002,
+	0x00000000,
+	0x00000554,
+	0x00000003,
 	0x00000002,
-	0x00000430,
-	0x00040003,
+	0x000005d8,
+	0x00040004,
+	0x00000000,
+	0x000005f4,
+	0x00010005,
+	0x00000000,
+	0x0000060e,
+	0x00010006,
+	0x00000000,
+	0x000005d3,
+	0x00000007,
+	0x00000000,
+	0x00000619,
+/* 0x03c4: memx_func_tail */
+/* 0x03c4: memx_ts_start */
+	0x00000000,
+/* 0x03c8: memx_ts_end */
+	0x00000000,
+/* 0x03cc: memx_data_head */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
-	0x0000044c,
-	0x00010004,
 	0x00000000,
-	0x00000466,
-/* 0x03ac: memx_func_tail */
-/* 0x03ac: memx_data_head */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -728,6 +767,8 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0bcc: memx_data_tail */
+/* 0x0bcc: memx_train_head */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -757,8 +798,43 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0bac: memx_data_tail */
-/* 0x0bac: i2c_scl_map */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+/* 0x0ccc: memx_train_tail */
+/* 0x0ccc: i2c_scl_map */
 	0x00000400,
 	0x00000800,
 	0x00001000,
@@ -769,7 +845,7 @@ uint32_t nvd0_pwr_data[] = {
 	0x00020000,
 	0x00040000,
 	0x00080000,
-/* 0x0bd4: i2c_sda_map */
+/* 0x0cf4: i2c_sda_map */
 	0x00100000,
 	0x00200000,
 	0x00400000,
@@ -781,10 +857,66 @@ uint32_t nvd0_pwr_data[] = {
 	0x10000000,
 	0x20000000,
 	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 };
 
 uint32_t nvd0_pwr_code[] = {
-	0x02bf0ef5,
+	0x034d0ef5,
 /* 0x0004: rd32 */
 	0x07a007f1,
 	0xbd000ed0,
@@ -814,17 +946,20 @@ uint32_t nvd0_pwr_code[] = {
 	0xd4f100dd,
 	0x1bf47000,
 /* 0x0067: nsec */
-	0xf000f8f5,
+	0xf900f8f5,
+	0xf080f990,
 	0x88cf2c87,
-/* 0x006d: nsec_loop */
+/* 0x0071: nsec_loop */
 	0x2c97f000,
 	0xbb0099cf,
 	0x9eb80298,
 	0xf41ef406,
-/* 0x007e: wait */
-	0x87f000f8,
+	0x90fc80fc,
+/* 0x0086: wait */
+	0x90f900f8,
+	0x87f080f9,
 	0x0088cf2c,
-/* 0x0084: wait_loop */
+/* 0x0090: wait_loop */
 	0xf402eeb9,
 	0xdab90421,
 	0x04adfd02,
@@ -833,28 +968,29 @@ uint32_t nvd0_pwr_code[] = {
 	0x0099cf2c,
 	0xb80298bb,
 	0x1ef4069b,
-/* 0x00a5: wait_done */
-/* 0x00a7: intr_watchdog */
-	0x9800f8e2,
+/* 0x00b1: wait_done */
+	0xfc80fce2,
+/* 0x00b7: intr_watchdog */
+	0x9800f890,
 	0x96b003e9,
 	0x2a0bf400,
 	0xbb9a0a98,
 	0x1cf4029a,
 	0x01d7f00f,
-	0x020621f5,
+	0x028c21f5,
 	0x0ef494bd,
-/* 0x00c5: intr_watchdog_next_time */
+/* 0x00d5: intr_watchdog_next_time */
 	0x9b0a9815,
 	0xf400a6b0,
 	0x9ab8090b,
 	0x061cf406,
-/* 0x00d4: intr_watchdog_next_time_set */
-/* 0x00d7: intr_watchdog_next_proc */
+/* 0x00e4: intr_watchdog_next_time_set */
+/* 0x00e7: intr_watchdog_next_proc */
 	0x809b0980,
 	0xe0b603e9,
 	0x68e6b158,
 	0xc61bf402,
-/* 0x00e6: intr */
+/* 0x00f6: intr */
 	0x00f900f8,
 	0x80f904bd,
 	0xa0f990f9,
@@ -872,12 +1008,12 @@ uint32_t nvd0_pwr_code[] = {
 	0x0bf40289,
 	0x9b008020,
 	0xf458e7f0,
-	0x0998a721,
+	0x0998b721,
 	0x0096b09b,
 	0xf00e0bf4,
 	0x09d03407,
 	0x8004bd00,
-/* 0x013e: intr_skip_watchdog */
+/* 0x014e: intr_skip_watchdog */
 	0x89e49a09,
 	0x0bf40800,
 	0x8897f13c,
@@ -889,20 +1025,20 @@ uint32_t nvd0_pwr_code[] = {
 	0xf14f48e7,
 	0xf05453e3,
 	0x21f500d7,
-	0xc0fc026b,
+	0xc0fc02f1,
 	0x04c007f1,
 	0xbd000cd0,
-/* 0x0175: intr_subintr_skip_fifo */
+/* 0x0185: intr_subintr_skip_fifo */
 	0x8807f104,
 	0x0009d006,
-/* 0x017e: intr_skip_subintr */
+/* 0x018e: intr_skip_subintr */
 	0x89c404bd,
 	0x070bf420,
 	0xffbfa4f1,
-/* 0x0188: intr_skip_pause */
+/* 0x0198: intr_skip_pause */
 	0xf44089c4,
 	0xa4f1070b,
-/* 0x0192: intr_skip_user0 */
+/* 0x01a2: intr_skip_user0 */
 	0x07f0ffbf,
 	0x0008d004,
 	0x80fc04bd,
@@ -912,189 +1048,298 @@ uint32_t nvd0_pwr_code[] = {
 	0xfca0fcb0,
 	0xfc80fc90,
 	0x0032f400,
-/* 0x01b6: timer */
-	0x32f401f8,
-	0x03f89810,
-	0xf40086b0,
-	0xfe80421c,
-	0x3807f003,
+/* 0x01c6: ticks_from_ns */
+	0xc0f901f8,
+	0xd7f1b0f9,
+	0xd3f00144,
+	0xb321f500,
+	0xe8ccec03,
+	0x00b4b003,
+	0xec120bf4,
+	0xf103e8ee,
+	0xf00144d7,
+	0x21f500d3,
+/* 0x01ee: ticks_from_ns_quit */
+	0xceb903b3,
+	0xfcb0fc02,
+/* 0x01f7: ticks_from_us */
+	0xf900f8c0,
+	0xf1b0f9c0,
+	0xf00144d7,
+	0x21f500d3,
+	0xceb903b3,
+	0x00b4b002,
+	0xbd050bf4,
+/* 0x0211: ticks_from_us_quit */
+	0xfcb0fce4,
+/* 0x0217: ticks_to_us */
+	0xf100f8c0,
+	0xf00144d7,
+	0xedff00d3,
+/* 0x0223: timer */
+	0xf900f8ec,
+	0xf480f990,
+	0xf8981032,
+	0x0086b003,
+	0xbd531cf4,
+	0x3807f084,
 	0xbd0008d0,
-	0x0887f004,
-	0xf00088cf,
-	0x1bf40284,
-	0x3487f020,
-	0xb80088cf,
-	0x0bf406e0,
-	0x06e8b809,
-/* 0x01eb: timer_reset */
-	0xf0191ef4,
-	0x0ed03407,
-	0x8004bd00,
-/* 0x01f6: timer_enable */
-	0x87f09a0e,
-	0x3807f001,
-	0xbd0008d0,
-/* 0x0201: timer_done */
-	0x1031f404,
-/* 0x0206: send_proc */
-	0x80f900f8,
-	0xe89890f9,
+	0x3487f004,
+	0x980088cf,
+	0x98bb9a09,
+	0x00e9bb02,
+	0xf003fe80,
+	0x88cf0887,
+	0x0284f000,
+	0xf0201bf4,
+	0x88cf3487,
+	0x06e0b800,
+	0xb8090bf4,
+	0x1cf406e8,
+/* 0x026d: timer_reset */
+	0x3407f00e,
+	0xbd000ed0,
+	0x9a0e8004,
+/* 0x0278: timer_enable */
+	0xf00187f0,
+	0x08d03807,
+/* 0x0283: timer_done */
+	0xf404bd00,
+	0x80fc1031,
+	0x00f890fc,
+/* 0x028c: send_proc */
+	0x90f980f9,
+	0x9805e898,
+	0x86f004e9,
+	0x0689b804,
+	0xc42a0bf4,
+	0x88940398,
+	0x1880b604,
+	0x98008ebb,
+	0x8a8000fa,
+	0x018d8000,
+	0x80028c80,
+	0x90b6038b,
+	0x0794f001,
+	0xf404e980,
+/* 0x02c6: send_done */
+	0x90fc0231,
+	0x00f880fc,
+/* 0x02cc: find */
+	0x87f080f9,
+	0x0131f458,
+/* 0x02d4: find_loop */
+	0xb8008a98,
+	0x0bf406ae,
+	0x5880b610,
+	0x026886b1,
+	0xf4f01bf4,
+/* 0x02ea: find_done */
+	0x8eb90132,
+	0xf880fc02,
+/* 0x02f1: send */
+	0xcc21f500,
+	0x9701f402,
+/* 0x02fa: recv */
+	0x90f900f8,
+	0xe89880f9,
 	0x04e99805,
-	0xb80486f0,
+	0xb80132f4,
 	0x0bf40689,
-	0x0398c42a,
-	0xb6048894,
-	0x8ebb1880,
-	0x00fa9800,
-	0x80008a80,
-	0x8c80018d,
-	0x038b8002,
-	0xf00190b6,
-	0xe9800794,
-	0x0231f404,
-/* 0x0240: send_done */
-	0x80fc90fc,
-/* 0x0246: find */
-	0x80f900f8,
-	0xf45887f0,
-/* 0x024e: find_loop */
-	0x8a980131,
-	0x06aeb800,
-	0xb6100bf4,
-	0x86b15880,
-	0x1bf40268,
-	0x0132f4f0,
-/* 0x0264: find_done */
-	0xfc028eb9,
-/* 0x026b: send */
-	0xf500f880,
-	0xf4024621,
-	0x00f89701,
-/* 0x0274: recv */
-	0x9805e898,
-	0x32f404e9,
-	0x0689b801,
-	0xc43d0bf4,
-	0x80b60389,
-	0x0784f001,
-	0x9805e880,
-	0xf0f902ea,
-	0xf9018ffe,
-	0x02efb9f0,
-	0xbb049994,
-	0xe0b600e9,
-	0x03eb9818,
-	0x9802ec98,
-	0xee9801ed,
-	0xfca5f900,
-	0x00f8fef0,
-	0xfc0131f4,
-/* 0x02bd: recv_done */
-/* 0x02bf: init */
-	0xf100f8f0,
-	0xcf010817,
-	0x11e70011,
-	0x14b60109,
-	0x0014fe08,
-	0x00e017f1,
-	0xf00013f0,
-	0x01d01c07,
-	0xf004bd00,
-	0x07f0ff17,
-	0x0001d014,
-	0x17f004bd,
-	0x0015f102,
-	0x1007f008,
-	0xbd0001d0,
-	0xe617f104,
-	0x0013f000,
-	0xf40010fe,
-	0x17f01031,
-	0x3807f001,
-	0xbd0001d0,
-	0x58f7f004,
-/* 0x0314: init_proc */
-	0xb001f198,
-	0x0bf40016,
-	0xb615f9fa,
-	0x0ef458f0,
-/* 0x0325: host_send */
-	0xb017f1f2,
-	0x0011cf04,
-	0x04a027f1,
-	0xb80022cf,
-	0x0bf40612,
-	0x071ec42f,
-	0xb704ee94,
-	0x980270e0,
+	0x0389c43d,
+	0xf00180b6,
+	0xe8800784,
+	0x02ea9805,
+	0x8ffef0f9,
+	0xb9f0f901,
+	0x999402ef,
+	0x00e9bb04,
+	0x9818e0b6,
 	0xec9803eb,
 	0x01ed9802,
-	0xf500ee98,
-	0xb6026b21,
-	0x1ec40110,
-	0xb007f10f,
-	0x000ed004,
-	0x0ef404bd,
-/* 0x0365: host_send_done */
-/* 0x0367: host_recv */
-	0xf100f8c3,
-	0xf14e4917,
-	0xb8525413,
-	0x0bf406e1,
-/* 0x0375: host_recv_wait */
-	0xcc17f1b3,
-	0x0011cf04,
-	0x04c827f1,
-	0xf00022cf,
-	0x12b80816,
-	0xec0bf406,
-	0xb60723c4,
-	0x30b70434,
-	0x3b8002f0,
-	0x023c8003,
-	0x80013d80,
-	0x20b6003e,
-	0x0f24f001,
-	0x04c807f1,
-	0xbd0002d0,
-	0x4027f004,
-	0xd00007f0,
-	0x04bd0002,
-/* 0x03be: host_init */
+	0xf900ee98,
+	0xfef0fca5,
+	0x31f400f8,
+/* 0x0347: recv_done */
+	0xfcf0fc01,
+	0xf890fc80,
+/* 0x034d: init */
+	0x0817f100,
+	0x0011cf01,
+	0x010911e7,
+	0xfe0814b6,
+	0x17f10014,
+	0x13f000e0,
+	0x1c07f000,
+	0xbd0001d0,
+	0xff17f004,
+	0xd01407f0,
+	0x04bd0001,
+	0xf10217f0,
+	0xf0080015,
+	0x01d01007,
+	0xf104bd00,
+	0xf000f617,
+	0x10fe0013,
+	0x1031f400,
+	0xf00117f0,
+	0x01d03807,
+	0xf004bd00,
+/* 0x03a2: init_proc */
+	0xf19858f7,
+	0x0016b001,
+	0xf9fa0bf4,
+	0x58f0b615,
+/* 0x03b3: mulu32_32_64 */
+	0xf9f20ef4,
+	0xf920f910,
+	0x9540f930,
+	0xd29510e1,
+	0xbdc4bd10,
+	0xc0edffb4,
+	0xb9301dff,
+	0x34f10234,
+	0x34b6ffff,
+	0x1045b610,
+	0xbb00c3bb,
+	0xe2ff01b4,
+	0x0234b930,
+	0xffff34f1,
+	0xb61034b6,
+	0xc3bb1045,
+	0x01b4bb00,
+	0xbb3012ff,
+	0x40fc00b3,
+	0x20fc30fc,
+	0x00f810fc,
+/* 0x0404: host_send */
+	0x04b017f1,
+	0xf10011cf,
+	0xcf04a027,
+	0x12b80022,
+	0x2f0bf406,
+	0x94071ec4,
+	0xe0b704ee,
+	0xeb980270,
+	0x02ec9803,
+	0x9801ed98,
+	0x21f500ee,
+	0x10b602f1,
+	0x0f1ec401,
+	0x04b007f1,
+	0xbd000ed0,
+	0xc30ef404,
+/* 0x0444: host_send_done */
+/* 0x0446: host_recv */
 	0x17f100f8,
-	0x14b60080,
-	0x7015f110,
-	0xd007f102,
-	0x0001d004,
-	0x17f104bd,
-	0x14b60080,
-	0xf015f110,
-	0xdc07f102,
-	0x0001d004,
-	0x17f004bd,
-	0xc407f101,
-	0x0001d004,
-	0x00f804bd,
-/* 0x03f4: memx_func_enter */
+	0x13f14e49,
+	0xe1b85254,
+	0xb30bf406,
+/* 0x0454: host_recv_wait */
+	0x04cc17f1,
+	0xf10011cf,
+	0xcf04c827,
+	0x16f00022,
+	0x0612b808,
+	0xc4ec0bf4,
+	0x34b60723,
+	0xf030b704,
+	0x033b8002,
+	0x80023c80,
+	0x3e80013d,
+	0x0120b600,
+	0xf10f24f0,
+	0xd004c807,
+	0x04bd0002,
+	0xf04027f0,
+	0x02d00007,
+	0xf804bd00,
+/* 0x049d: host_init */
+	0x8017f100,
+	0x1014b600,
+	0x027015f1,
+	0x04d007f1,
+	0xbd0001d0,
+	0x8017f104,
+	0x1014b600,
+	0x02f015f1,
+	0x04dc07f1,
+	0xbd0001d0,
+	0x0117f004,
+	0x04c407f1,
+	0xbd0001d0,
+/* 0x04d3: memx_func_enter */
+	0xf100f804,
+	0xf1162067,
+	0xf1f55d77,
+	0xb9ffff73,
+	0x21f4026e,
+	0x02d8b904,
+	0xf90487fd,
+	0xfc80f960,
+	0xf4e0fcd0,
+	0x77f13321,
+	0x73f1fffe,
+	0x6eb9ffff,
+	0x0421f402,
+	0xfd02d8b9,
+	0x60f90487,
+	0xd0fc80f9,
+	0x21f4e0fc,
+	0xf067f133,
+	0x026eb926,
+	0xb90421f4,
+	0x87fd02d8,
+	0xf960f904,
+	0xfcd0fc80,
+	0x3321f4e0,
 	0xf10467f0,
 	0xd007e007,
 	0x04bd0006,
-/* 0x0400: memx_func_enter_wait */
+/* 0x053c: memx_func_enter_wait */
 	0x07c067f1,
 	0xf00066cf,
 	0x0bf40464,
-	0x001698f6,
-	0xf80410b6,
-/* 0x0415: memx_func_leave */
-	0x0467f000,
+	0x2c67f0f6,
+	0x800066cf,
+	0x00f8f106,
+/* 0x0554: memx_func_leave */
+	0xcf2c67f0,
+	0x06800066,
+	0x0467f0f2,
 	0x07e407f1,
 	0xbd0006d0,
-/* 0x0421: memx_func_leave_wait */
+/* 0x0569: memx_func_leave_wait */
 	0xc067f104,
 	0x0066cf07,
 	0xf40464f0,
-	0x00f8f61b,
-/* 0x0430: memx_func_wr32 */
+	0x67f1f61b,
+	0x77f126f0,
+	0x73f00001,
+	0x026eb900,
+	0xb90421f4,
+	0x87fd02d8,
+	0xf960f905,
+	0xfcd0fc80,
+	0x3321f4e0,
+	0x162067f1,
+	0xf4026eb9,
+	0xd8b90421,
+	0x0587fd02,
+	0x80f960f9,
+	0xe0fcd0fc,
+	0xf13321f4,
+	0xf00aa277,
+	0x6eb90073,
+	0x0421f402,
+	0xfd02d8b9,
+	0x60f90587,
+	0xd0fc80f9,
+	0x21f4e0fc,
+/* 0x05d3: memx_func_wait_vblank */
+	0xb600f833,
+	0x00f80410,
+/* 0x05d8: memx_func_wr32 */
 	0x98001698,
 	0x10b60115,
 	0xf960f908,
@@ -1102,379 +1347,405 @@ uint32_t nvd0_pwr_code[] = {
 	0x3321f4e0,
 	0xf40242b6,
 	0x00f8e91b,
-/* 0x044c: memx_func_wait */
+/* 0x05f4: memx_func_wait */
 	0xcf2c87f0,
 	0x1e980088,
 	0x011d9800,
 	0x98021c98,
 	0x10b6031b,
-	0x7e21f410,
-/* 0x0466: memx_func_delay */
+	0x8621f410,
+/* 0x060e: memx_func_delay */
 	0x1e9800f8,
 	0x0410b600,
 	0xf86721f4,
-/* 0x0471: memx_exec */
-	0xf9e0f900,
-	0x02c1b9d0,
-/* 0x047b: memx_exec_next */
-	0x9802b2b9,
-	0x10b60013,
-	0x10349504,
-	0x980c30f0,
-	0x55f9de35,
-	0xf40612b8,
-	0xd0fcec1e,
-	0x21f5e0fc,
-	0x00f8026b,
-/* 0x049c: memx_info */
-	0x03acc7f1,
+/* 0x0619: memx_func_train */
+/* 0x061b: memx_exec */
+	0xf900f800,
+	0xb9d0f9e0,
+	0xb2b902c1,
+/* 0x0625: memx_exec_next */
+	0x00139802,
+	0xe70410b6,
+	0xe701f034,
+	0xb601e033,
+	0x30f00132,
+	0xde35980c,
+	0x12b855f9,
+	0xe41ef406,
+	0x98f10b98,
+	0xcbbbf20c,
+	0xc4b7f102,
+	0x00bbcf07,
+	0xe0fcd0fc,
+	0x02f121f5,
+/* 0x065e: memx_info */
+	0xc67000f8,
+	0x0e0bf401,
+/* 0x0664: memx_info_data */
+	0x03ccc7f1,
 	0x0800b7f1,
-	0x026b21f5,
-/* 0x04aa: memx_recv */
-	0xd6b000f8,
-	0xc40bf401,
-	0xf400d6b0,
-	0x00f8e90b,
-/* 0x04b8: memx_init */
-/* 0x04ba: perf_recv */
-	0x00f800f8,
-/* 0x04bc: perf_init */
-/* 0x04be: i2c_drive_scl */
-	0x36b000f8,
-	0x0e0bf400,
-	0x07e007f1,
-	0xbd0001d0,
-/* 0x04cf: i2c_drive_scl_lo */
-	0xf100f804,
-	0xd007e407,
+/* 0x066f: memx_info_train */
+	0xf10b0ef4,
+	0xf10bccc7,
+/* 0x0677: memx_info_send */
+	0xf50100b7,
+	0xf802f121,
+/* 0x067d: memx_recv */
+	0x01d6b000,
+	0xb09b0bf4,
+	0x0bf400d6,
+/* 0x068b: memx_init */
+	0xf800f8d8,
+/* 0x068d: perf_recv */
+/* 0x068f: perf_init */
+	0xf800f800,
+/* 0x0691: i2c_drive_scl */
+	0x0036b000,
+	0xf10e0bf4,
+	0xd007e007,
 	0x04bd0001,
-/* 0x04da: i2c_drive_sda */
-	0x36b000f8,
-	0x0e0bf400,
-	0x07e007f1,
-	0xbd0002d0,
-/* 0x04eb: i2c_drive_sda_lo */
-	0xf100f804,
-	0xd007e407,
+/* 0x06a2: i2c_drive_scl_lo */
+	0x07f100f8,
+	0x01d007e4,
+	0xf804bd00,
+/* 0x06ad: i2c_drive_sda */
+	0x0036b000,
+	0xf10e0bf4,
+	0xd007e007,
 	0x04bd0002,
-/* 0x04f6: i2c_sense_scl */
+/* 0x06be: i2c_drive_sda_lo */
+	0x07f100f8,
+	0x02d007e4,
+	0xf804bd00,
+/* 0x06c9: i2c_sense_scl */
+	0x0132f400,
+	0x07c437f1,
+	0xfd0033cf,
+	0x0bf40431,
+	0x0131f406,
+/* 0x06dc: i2c_sense_scl_done */
+/* 0x06de: i2c_sense_sda */
 	0x32f400f8,
 	0xc437f101,
 	0x0033cf07,
-	0xf40431fd,
+	0xf40432fd,
 	0x31f4060b,
-/* 0x0509: i2c_sense_scl_done */
-/* 0x050b: i2c_sense_sda */
-	0xf400f801,
-	0x37f10132,
-	0x33cf07c4,
-	0x0432fd00,
-	0xf4060bf4,
-/* 0x051e: i2c_sense_sda_done */
-	0x00f80131,
-/* 0x0520: i2c_raise_scl */
-	0x47f140f9,
-	0x37f00898,
-	0xbe21f501,
-/* 0x052d: i2c_raise_scl_wait */
-	0xe8e7f104,
-	0x6721f403,
-	0x04f621f5,
-	0xb60901f4,
-	0x1bf40142,
-/* 0x0541: i2c_raise_scl_done */
-	0xf840fcef,
-/* 0x0545: i2c_start */
-	0xf621f500,
-	0x0d11f404,
-	0x050b21f5,
-	0xf40611f4,
-/* 0x0556: i2c_start_rep */
-	0x37f0300e,
-	0xbe21f500,
-	0x0137f004,
-	0x04da21f5,
-	0xb60076bb,
-	0x50f90465,
-	0xbb046594,
-	0x50bd0256,
-	0xfc0475fd,
-	0x2021f550,
-	0x0464b605,
-/* 0x0583: i2c_start_send */
-	0xf01f11f4,
-	0x21f50037,
-	0xe7f104da,
-	0x21f41388,
-	0x0037f067,
-	0x04be21f5,
-	0x1388e7f1,
-/* 0x059f: i2c_start_out */
-	0xf86721f4,
-/* 0x05a1: i2c_stop */
-	0x0037f000,
-	0x04be21f5,
-	0xf50037f0,
-	0xf104da21,
-	0xf403e8e7,
-	0x37f06721,
-	0xbe21f501,
-	0x88e7f104,
-	0x6721f413,
-	0xf50137f0,
-	0xf104da21,
-	0xf41388e7,
-	0x00f86721,
-/* 0x05d4: i2c_bitw */
-	0x04da21f5,
+/* 0x06f1: i2c_sense_sda_done */
+/* 0x06f3: i2c_raise_scl */
+	0xf900f801,
+	0x9847f140,
+	0x0137f008,
+	0x069121f5,
+/* 0x0700: i2c_raise_scl_wait */
 	0x03e8e7f1,
-	0xbb6721f4,
+	0xf56721f4,
+	0xf406c921,
+	0x42b60901,
+	0xef1bf401,
+/* 0x0714: i2c_raise_scl_done */
+	0x00f840fc,
+/* 0x0718: i2c_start */
+	0x06c921f5,
+	0xf50d11f4,
+	0xf406de21,
+	0x0ef40611,
+/* 0x0729: i2c_start_rep */
+	0x0037f030,
+	0x069121f5,
+	0xf50137f0,
+	0xbb06ad21,
 	0x65b60076,
 	0x9450f904,
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x052021f5,
+	0x06f321f5,
 	0xf40464b6,
-	0xe7f11811,
-	0x21f41388,
-	0x0037f067,
-	0x04be21f5,
+/* 0x0756: i2c_start_send */
+	0x37f01f11,
+	0xad21f500,
+	0x88e7f106,
+	0x6721f413,
+	0xf50037f0,
+	0xf1069121,
+	0xf41388e7,
+/* 0x0772: i2c_start_out */
+	0x00f86721,
+/* 0x0774: i2c_stop */
+	0xf50037f0,
+	0xf0069121,
+	0x21f50037,
+	0xe7f106ad,
+	0x21f403e8,
+	0x0137f067,
+	0x069121f5,
 	0x1388e7f1,
-/* 0x0613: i2c_bitw_out */
-	0xf86721f4,
-/* 0x0615: i2c_bitr */
-	0x0137f000,
-	0x04da21f5,
-	0x03e8e7f1,
-	0xbb6721f4,
-	0x65b60076,
-	0x9450f904,
-	0x56bb0465,
-	0xfd50bd02,
-	0x50fc0475,
-	0x052021f5,
-	0xf40464b6,
-	0x21f51b11,
-	0x37f0050b,
-	0xbe21f500,
-	0x88e7f104,
+	0xf06721f4,
+	0x21f50137,
+	0xe7f106ad,
+	0x21f41388,
+/* 0x07a7: i2c_bitw */
+	0xf500f867,
+	0xf106ad21,
+	0xf403e8e7,
+	0x76bb6721,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb606f321,
+	0x11f40464,
+	0x88e7f118,
 	0x6721f413,
-	0xf4013cf0,
-/* 0x065a: i2c_bitr_done */
-	0x00f80131,
-/* 0x065c: i2c_get_byte */
-	0xf00057f0,
-/* 0x0662: i2c_get_byte_next */
-	0x54b60847,
-	0x0076bb01,
-	0xf90465b6,
-	0x04659450,
-	0xbd0256bb,
-	0x0475fd50,
-	0x21f550fc,
-	0x64b60615,
-	0x2b11f404,
-	0xb60553fd,
-	0x1bf40142,
-	0x0137f0d8,
+	0xf50037f0,
+	0xf1069121,
+	0xf41388e7,
+/* 0x07e6: i2c_bitw_out */
+	0x00f86721,
+/* 0x07e8: i2c_bitr */
+	0xf50137f0,
+	0xf106ad21,
+	0xf403e8e7,
+	0x76bb6721,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb606f321,
+	0x11f40464,
+	0xde21f51b,
+	0x0037f006,
+	0x069121f5,
+	0x1388e7f1,
+	0xf06721f4,
+	0x31f4013c,
+/* 0x082d: i2c_bitr_done */
+/* 0x082f: i2c_get_byte */
+	0xf000f801,
+	0x47f00057,
+/* 0x0835: i2c_get_byte_next */
+	0x0154b608,
 	0xb60076bb,
 	0x50f90465,
 	0xbb046594,
 	0x50bd0256,
 	0xfc0475fd,
-	0xd421f550,
-	0x0464b605,
-/* 0x06ac: i2c_get_byte_done */
-/* 0x06ae: i2c_put_byte */
-	0x47f000f8,
-/* 0x06b1: i2c_put_byte_next */
-	0x0142b608,
-	0xbb3854ff,
+	0xe821f550,
+	0x0464b607,
+	0xfd2b11f4,
+	0x42b60553,
+	0xd81bf401,
+	0xbb0137f0,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x07a721f5,
+/* 0x087f: i2c_get_byte_done */
+	0xf80464b6,
+/* 0x0881: i2c_put_byte */
+	0x0847f000,
+/* 0x0884: i2c_put_byte_next */
+	0xff0142b6,
+	0x76bb3854,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb607a721,
+	0x11f40464,
+	0x0046b034,
+	0xbbd81bf4,
 	0x65b60076,
 	0x9450f904,
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x05d421f5,
+	0x07e821f5,
 	0xf40464b6,
-	0x46b03411,
-	0xd81bf400,
+	0x76bb0f11,
+	0x0136b000,
+	0xf4061bf4,
+/* 0x08da: i2c_put_byte_done */
+	0x00f80132,
+/* 0x08dc: i2c_addr */
 	0xb60076bb,
 	0x50f90465,
 	0xbb046594,
 	0x50bd0256,
 	0xfc0475fd,
-	0x1521f550,
-	0x0464b606,
-	0xbb0f11f4,
-	0x36b00076,
-	0x061bf401,
-/* 0x0707: i2c_put_byte_done */
-	0xf80132f4,
-/* 0x0709: i2c_addr */
-	0x0076bb00,
+	0x1821f550,
+	0x0464b607,
+	0xe72911f4,
+	0xb6012ec3,
+	0x53fd0134,
+	0x0076bb05,
 	0xf90465b6,
 	0x04659450,
 	0xbd0256bb,
 	0x0475fd50,
 	0x21f550fc,
-	0x64b60545,
-	0x2911f404,
-	0x012ec3e7,
-	0xfd0134b6,
-	0x76bb0553,
-	0x0465b600,
-	0x659450f9,
-	0x0256bb04,
-	0x75fd50bd,
-	0xf550fc04,
-	0xb606ae21,
-/* 0x074e: i2c_addr_done */
-	0x00f80464,
-/* 0x0750: i2c_acquire_addr */
-	0xb6f8cec7,
-	0xe0b705e4,
-	0x00f8d014,
-/* 0x075c: i2c_acquire */
-	0x075021f5,
-	0xf00421f4,
-	0x21f403d9,
-/* 0x076b: i2c_release */
-	0xf500f833,
-	0xf4075021,
-	0xdaf00421,
+	0x64b60881,
+/* 0x0921: i2c_addr_done */
+/* 0x0923: i2c_acquire_addr */
+	0xc700f804,
+	0xe4b6f8ce,
+	0x14e0b705,
+/* 0x092f: i2c_acquire */
+	0xf500f8d0,
+	0xf4092321,
+	0xd9f00421,
 	0x3321f403,
-/* 0x077a: i2c_recv */
-	0x32f400f8,
-	0xf8c1c701,
-	0xb00214b6,
-	0x1ff52816,
-	0x13a0013a,
-	0x32980bd4,
-	0xac13a000,
-	0x0031980b,
-	0xf90231f4,
-	0xf9e0f9d0,
-	0x0067f1d0,
-	0x0063f100,
-	0x01679210,
-	0xb60076bb,
-	0x50f90465,
-	0xbb046594,
-	0x50bd0256,
-	0xfc0475fd,
-	0x5c21f550,
-	0x0464b607,
-	0xd6b0d0fc,
-	0xb31bf500,
-	0x0057f000,
-	0xb60076bb,
-	0x50f90465,
-	0xbb046594,
-	0x50bd0256,
-	0xfc0475fd,
-	0x0921f550,
-	0x0464b607,
-	0x00d011f5,
-	0xbbe0c5c7,
+/* 0x093e: i2c_release */
+	0x21f500f8,
+	0x21f40923,
+	0x03daf004,
+	0xf83321f4,
+/* 0x094d: i2c_recv */
+	0x0132f400,
+	0xb6f8c1c7,
+	0x16b00214,
+	0x3a1ff528,
+	0xf413a001,
+	0x0032980c,
+	0x0ccc13a0,
+	0xf4003198,
+	0xd0f90231,
+	0xd0f9e0f9,
+	0x000067f1,
+	0x100063f1,
+	0xbb016792,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x092f21f5,
+	0xfc0464b6,
+	0x00d6b0d0,
+	0x00b31bf5,
+	0xbb0057f0,
 	0x65b60076,
 	0x9450f904,
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x06ae21f5,
+	0x08dc21f5,
 	0xf50464b6,
-	0xf000ad11,
-	0x76bb0157,
+	0xc700d011,
+	0x76bbe0c5,
 	0x0465b600,
 	0x659450f9,
 	0x0256bb04,
 	0x75fd50bd,
 	0xf550fc04,
-	0xb6070921,
+	0xb6088121,
 	0x11f50464,
-	0x76bb008a,
-	0x0465b600,
-	0x659450f9,
-	0x0256bb04,
-	0x75fd50bd,
-	0xf550fc04,
-	0xb6065c21,
-	0x11f40464,
-	0xe05bcb6a,
-	0xb60076bb,
-	0x50f90465,
-	0xbb046594,
-	0x50bd0256,
-	0xfc0475fd,
-	0xa121f550,
-	0x0464b605,
-	0xbd025bb9,
-	0x430ef474,
-/* 0x0880: i2c_recv_not_rd08 */
-	0xf401d6b0,
-	0x57f03d1b,
-	0x0921f500,
-	0x3311f407,
-	0xf5e0c5c7,
-	0xf406ae21,
-	0x57f02911,
-	0x0921f500,
-	0x1f11f407,
-	0xf5e0b5c7,
-	0xf406ae21,
-	0x21f51511,
-	0x74bd05a1,
-	0xf408c5c7,
-	0x32f4091b,
-	0x030ef402,
-/* 0x08c0: i2c_recv_not_wr08 */
-/* 0x08c0: i2c_recv_done */
-	0xf5f8cec7,
-	0xfc076b21,
-	0xf4d0fce0,
-	0x7cb90a12,
-	0x6b21f502,
-/* 0x08d5: i2c_recv_exit */
-/* 0x08d7: i2c_init */
+	0x57f000ad,
+	0x0076bb01,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b608dc,
+	0x8a11f504,
+	0x0076bb00,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b6082f,
+	0x6a11f404,
+	0xbbe05bcb,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x077421f5,
+	0xb90464b6,
+	0x74bd025b,
+/* 0x0a53: i2c_recv_not_rd08 */
+	0xb0430ef4,
+	0x1bf401d6,
+	0x0057f03d,
+	0x08dc21f5,
+	0xc73311f4,
+	0x21f5e0c5,
+	0x11f40881,
+	0x0057f029,
+	0x08dc21f5,
+	0xc71f11f4,
+	0x21f5e0b5,
+	0x11f40881,
+	0x7421f515,
+	0xc774bd07,
+	0x1bf408c5,
+	0x0232f409,
+/* 0x0a93: i2c_recv_not_wr08 */
+/* 0x0a93: i2c_recv_done */
+	0xc7030ef4,
+	0x21f5f8ce,
+	0xe0fc093e,
+	0x12f4d0fc,
+	0x027cb90a,
+	0x02f121f5,
+/* 0x0aa8: i2c_recv_exit */
+/* 0x0aaa: i2c_init */
+	0x00f800f8,
+/* 0x0aac: test_recv */
+	0x05d817f1,
+	0xb60011cf,
+	0x07f10110,
+	0x01d005d8,
+	0xf104bd00,
+	0xf1d900e7,
+	0xf5134fe3,
+	0xf8022321,
+/* 0x0acd: test_init */
+	0x00e7f100,
+	0x2321f508,
+/* 0x0ad7: idle_recv */
 	0xf800f802,
-/* 0x08d9: test_recv */
-	0xd817f100,
-	0x0011cf05,
-	0xf10110b6,
-	0xd005d807,
-	0x04bd0001,
-	0xd900e7f1,
-	0x134fe3f1,
-	0x01b621f5,
-/* 0x08fa: test_init */
-	0xe7f100f8,
-	0x21f50800,
-	0x00f801b6,
-/* 0x0904: idle_recv */
-/* 0x0906: idle */
-	0x31f400f8,
-	0xd417f100,
-	0x0011cf05,
-	0xf10110b6,
-	0xd005d407,
-	0x04bd0001,
-/* 0x091c: idle_loop */
-	0xf45817f0,
-/* 0x0922: idle_proc */
-/* 0x0922: idle_proc_exec */
-	0x10f90232,
-	0xf5021eb9,
-	0xfc027421,
-	0x0911f410,
-	0xf40231f4,
-/* 0x0936: idle_proc_next */
-	0x10b6ef0e,
-	0x061fb858,
-	0xf4e61bf4,
-	0x28f4dd02,
-	0xc10ef400,
+/* 0x0ad9: idle */
+	0x0031f400,
+	0x05d417f1,
+	0xb60011cf,
+	0x07f10110,
+	0x01d005d4,
+/* 0x0aef: idle_loop */
+	0xf004bd00,
+	0x32f45817,
+/* 0x0af5: idle_proc */
+/* 0x0af5: idle_proc_exec */
+	0xb910f902,
+	0x21f5021e,
+	0x10fc02fa,
+	0xf40911f4,
+	0x0ef40231,
+/* 0x0b09: idle_proc_next */
+	0x5810b6ef,
+	0xf4061fb8,
+	0x02f4e61b,
+	0x0028f4dd,
+	0x00c10ef4,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/os.h b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/os.h
index 574acfa44c8c..c8b06cb77e72 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/os.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/os.h
@@ -18,12 +18,18 @@
 #define MEMX_MSG_INFO 0
 #define MEMX_MSG_EXEC 1
 
+/* MEMX: info types */
+#define MEMX_INFO_DATA  0
+#define MEMX_INFO_TRAIN 1
+
 /* MEMX: script opcode definitions */
-#define MEMX_ENTER  0
-#define MEMX_LEAVE  1
-#define MEMX_WR32   2
-#define MEMX_WAIT   3
-#define MEMX_DELAY  4
+#define MEMX_ENTER  1
+#define MEMX_LEAVE  2
+#define MEMX_WR32   3
+#define MEMX_WAIT   4
+#define MEMX_DELAY  5
+#define MEMX_VBLANK 6
+#define MEMX_TRAIN  7
 
 /* I2C_: message identifiers */
 #define I2C__MSG_RD08 0
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/memx.c b/drivers/gpu/drm/nouveau/core/subdev/pwr/memx.c
index def6a9ac68cf..7a9299d7159f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/memx.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/memx.c
@@ -20,10 +20,11 @@ memx_out(struct nouveau_memx *memx)
 	struct nouveau_pwr *ppwr = memx->ppwr;
 	int i;
 
-	if (memx->c.size) {
+	if (memx->c.mthd) {
 		nv_wr32(ppwr, 0x10a1c4, (memx->c.size << 16) | memx->c.mthd);
 		for (i = 0; i < memx->c.size; i++)
 			nv_wr32(ppwr, 0x10a1c4, memx->c.data[i]);
+		memx->c.mthd = 0;
 		memx->c.size = 0;
 	}
 }
@@ -32,7 +33,7 @@ static void
 memx_cmd(struct nouveau_memx *memx, u32 mthd, u32 size, u32 data[])
 {
 	if ((memx->c.size + size >= ARRAY_SIZE(memx->c.data)) ||
-	    (memx->c.size && memx->c.mthd != mthd))
+	    (memx->c.mthd && memx->c.mthd != mthd))
 		memx_out(memx);
 	memcpy(&memx->c.data[memx->c.size], data, size * sizeof(data[0]));
 	memx->c.size += size;
@@ -46,7 +47,8 @@ nouveau_memx_init(struct nouveau_pwr *ppwr, struct nouveau_memx **pmemx)
 	u32 reply[2];
 	int ret;
 
-	ret = ppwr->message(ppwr, reply, PROC_MEMX, MEMX_MSG_INFO, 0, 0);
+	ret = ppwr->message(ppwr, reply, PROC_MEMX, MEMX_MSG_INFO,
+					MEMX_INFO_DATA, 0);
 	if (ret)
 		return ret;
 
@@ -62,8 +64,7 @@ nouveau_memx_init(struct nouveau_pwr *ppwr, struct nouveau_memx **pmemx)
 		nv_wr32(ppwr, 0x10a580, 0x00000003);
 	} while (nv_rd32(ppwr, 0x10a580) != 0x00000003);
 	nv_wr32(ppwr, 0x10a1c0, 0x01000000 | memx->base);
-	nv_wr32(ppwr, 0x10a1c4, 0x00010000 | MEMX_ENTER);
-	nv_wr32(ppwr, 0x10a1c4, 0x00000000);
+
 	return 0;
 }
 
@@ -78,7 +79,6 @@ nouveau_memx_fini(struct nouveau_memx **pmemx, bool exec)
 	memx_out(memx);
 
 	/* release data segment access */
-	nv_wr32(ppwr, 0x10a1c4, 0x00000000 | MEMX_LEAVE);
 	finish = nv_rd32(ppwr, 0x10a1c0) & 0x00ffffff;
 	nv_wr32(ppwr, 0x10a580, 0x00000000);
 
@@ -88,6 +88,8 @@ nouveau_memx_fini(struct nouveau_memx **pmemx, bool exec)
 				 memx->base, finish);
 	}
 
+	nv_debug(memx->ppwr, "Exec took %uns, PPWR_IN %08x\n",
+		 reply[0], reply[1]);
 	kfree(memx);
 	return 0;
 }
@@ -105,7 +107,7 @@ nouveau_memx_wait(struct nouveau_memx *memx,
 {
 	nv_debug(memx->ppwr, "R[%06x] & 0x%08x == 0x%08x, %d us\n",
 				addr, mask, data, nsec);
-	memx_cmd(memx, MEMX_WAIT, 4, (u32[]){ addr, ~mask, data, nsec });
+	memx_cmd(memx, MEMX_WAIT, 4, (u32[]){ addr, mask, data, nsec });
 	memx_out(memx); /* fuc can't handle multiple */
 }
 
@@ -117,4 +119,83 @@ nouveau_memx_nsec(struct nouveau_memx *memx, u32 nsec)
 	memx_out(memx); /* fuc can't handle multiple */
 }
 
+void
+nouveau_memx_wait_vblank(struct nouveau_memx *memx)
+{
+	struct nouveau_pwr *ppwr = memx->ppwr;
+	u32 heads, x, y, px = 0;
+	int i, head_sync;
+
+	if (nv_device(ppwr)->chipset < 0xd0) {
+		heads = nv_rd32(ppwr, 0x610050);
+		for (i = 0; i < 2; i++) {
+			/* Heuristic: sync to head with biggest resolution */
+			if (heads & (2 << (i << 3))) {
+				x = nv_rd32(ppwr, 0x610b40 + (0x540 * i));
+				y = (x & 0xffff0000) >> 16;
+				x &= 0x0000ffff;
+				if ((x * y) > px) {
+					px = (x * y);
+					head_sync = i;
+				}
+			}
+		}
+	}
+
+	if (px == 0) {
+		nv_debug(memx->ppwr, "WAIT VBLANK !NO ACTIVE HEAD\n");
+		return;
+	}
+
+	nv_debug(memx->ppwr, "WAIT VBLANK HEAD%d\n", head_sync);
+	memx_cmd(memx, MEMX_VBLANK, 1, (u32[]){ head_sync });
+	memx_out(memx); /* fuc can't handle multiple */
+}
+
+void
+nouveau_memx_train(struct nouveau_memx *memx)
+{
+	nv_debug(memx->ppwr, "   MEM TRAIN\n");
+	memx_cmd(memx, MEMX_TRAIN, 0, NULL);
+}
+
+int
+nouveau_memx_train_result(struct nouveau_pwr *ppwr, u32 *res, int rsize)
+{
+	u32 reply[2], base, size, i;
+	int ret;
+
+	ret = ppwr->message(ppwr, reply, PROC_MEMX, MEMX_MSG_INFO,
+					MEMX_INFO_TRAIN, 0);
+	if (ret)
+		return ret;
+
+	base = reply[0];
+	size = reply[1] >> 2;
+	if (size > rsize)
+		return -ENOMEM;
+
+	/* read the packet */
+	nv_wr32(ppwr, 0x10a1c0, 0x02000000 | base);
+
+	for (i = 0; i < size; i++)
+		res[i] = nv_rd32(ppwr, 0x10a1c4);
+
+	return 0;
+}
+
+void
+nouveau_memx_block(struct nouveau_memx *memx)
+{
+	nv_debug(memx->ppwr, "   HOST BLOCKED\n");
+	memx_cmd(memx, MEMX_ENTER, 0, NULL);
+}
+
+void
+nouveau_memx_unblock(struct nouveau_memx *memx)
+{
+	nv_debug(memx->ppwr, "   HOST UNBLOCKED\n");
+	memx_cmd(memx, MEMX_LEAVE, 0, NULL);
+}
+
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c b/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c
index 016990a8252c..3656d605168f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c
@@ -31,6 +31,8 @@
 #include <subdev/gpio.h>
 #include <subdev/timer.h>
 
+#include <subdev/bios/fan.h>
+
 static int
 nouveau_fan_update(struct nouveau_fan *fan, bool immediate, int target)
 {
@@ -275,8 +277,11 @@ nouveau_therm_fan_ctor(struct nouveau_therm *therm)
 	/* other random init... */
 	nouveau_therm_fan_set_defaults(therm);
 	nvbios_perf_fan_parse(bios, &priv->fan->perf);
-	if (nvbios_therm_fan_parse(bios, &priv->fan->bios))
-		nv_error(therm, "parsing the thermal table failed\n");
+	if (!nvbios_fan_parse(bios, &priv->fan->bios)) {
+		nv_debug(therm, "parsing the fan table failed\n");
+		if (nvbios_therm_fan_parse(bios, &priv->fan->bios))
+			nv_error(therm, "parsing both fan tables failed\n");
+	}
 	nouveau_therm_fan_safety_checks(therm);
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/fanpwm.c b/drivers/gpu/drm/nouveau/core/subdev/therm/fanpwm.c
index 9a5c07340263..c629d7f2a6a4 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/fanpwm.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/fanpwm.c
@@ -25,6 +25,8 @@
 
 #include <core/option.h>
 #include <subdev/gpio.h>
+#include <subdev/bios.h>
+#include <subdev/bios/fan.h>
 
 #include "priv.h"
 
@@ -86,11 +88,15 @@ nouveau_fanpwm_create(struct nouveau_therm *therm, struct dcb_gpio_func *func)
 {
 	struct nouveau_device *device = nv_device(therm);
 	struct nouveau_therm_priv *tpriv = (void *)therm;
+	struct nouveau_bios *bios = nouveau_bios(therm);
 	struct nouveau_fanpwm_priv *priv;
+	struct nvbios_therm_fan fan;
 	u32 divs, duty;
 
+	nvbios_fan_parse(bios, &fan);
+
 	if (!nouveau_boolopt(device->cfgopt, "NvFanPWM", func->param) ||
-	    !therm->pwm_ctrl ||
+	    !therm->pwm_ctrl || fan.type == NVBIOS_THERM_FAN_TOGGLE ||
 	     therm->pwm_get(therm, func->line, &divs, &duty) == -ENODEV)
 		return -ENODEV;
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/gm107.c b/drivers/gpu/drm/nouveau/core/subdev/therm/gm107.c
new file mode 100644
index 000000000000..668cf3322285
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/gm107.c
@@ -0,0 +1,93 @@
+/*
+ * Copyright 2014 Martin Peres
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Martin Peres
+ */
+
+#include "priv.h"
+
+struct gm107_therm_priv {
+	struct nouveau_therm_priv base;
+};
+
+static int
+gm107_fan_pwm_ctrl(struct nouveau_therm *therm, int line, bool enable)
+{
+	/* nothing to do, it seems hardwired */
+	return 0;
+}
+
+static int
+gm107_fan_pwm_get(struct nouveau_therm *therm, int line, u32 *divs, u32 *duty)
+{
+	*divs = nv_rd32(therm, 0x10eb20) & 0x1fff;
+	*duty = nv_rd32(therm, 0x10eb24) & 0x1fff;
+	return 0;
+}
+
+static int
+gm107_fan_pwm_set(struct nouveau_therm *therm, int line, u32 divs, u32 duty)
+{
+	nv_mask(therm, 0x10eb10, 0x1fff, divs); /* keep the high bits */
+	nv_wr32(therm, 0x10eb14, duty | 0x80000000);
+	return 0;
+}
+
+static int
+gm107_fan_pwm_clock(struct nouveau_therm *therm, int line)
+{
+	return nv_device(therm)->crystal * 1000;
+}
+
+static int
+gm107_therm_ctor(struct nouveau_object *parent,
+		struct nouveau_object *engine,
+		struct nouveau_oclass *oclass, void *data, u32 size,
+		struct nouveau_object **pobject)
+{
+	struct gm107_therm_priv *priv;
+	int ret;
+
+	ret = nouveau_therm_create(parent, engine, oclass, &priv);
+	*pobject = nv_object(priv);
+	if (ret)
+		return ret;
+
+	priv->base.base.pwm_ctrl = gm107_fan_pwm_ctrl;
+	priv->base.base.pwm_get = gm107_fan_pwm_get;
+	priv->base.base.pwm_set = gm107_fan_pwm_set;
+	priv->base.base.pwm_clock = gm107_fan_pwm_clock;
+	priv->base.base.temp_get = nv84_temp_get;
+	priv->base.base.fan_sense = nva3_therm_fan_sense;
+	priv->base.sensor.program_alarms = nouveau_therm_program_alarms_polling;
+	return nouveau_therm_preinit(&priv->base.base);
+}
+
+struct nouveau_oclass
+gm107_therm_oclass = {
+	.handle = NV_SUBDEV(THERM, 0x117),
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = gm107_therm_ctor,
+		.dtor = _nouveau_therm_dtor,
+		.init = nvd0_therm_init,
+		.fini = nv84_therm_fini,
+	},
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/nv84.c b/drivers/gpu/drm/nouveau/core/subdev/therm/nv84.c
index 1d15c52fad0c..14e2e09bfc24 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/nv84.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/nv84.c
@@ -24,6 +24,7 @@
  */
 
 #include "priv.h"
+#include <subdev/fuse.h>
 
 struct nv84_therm_priv {
 	struct nouveau_therm_priv base;
@@ -32,7 +33,25 @@ struct nv84_therm_priv {
 int
 nv84_temp_get(struct nouveau_therm *therm)
 {
-	return nv_rd32(therm, 0x20400);
+	struct nouveau_fuse *fuse = nouveau_fuse(therm);
+
+	if (nv_ro32(fuse, 0x1a8) == 1)
+		return nv_rd32(therm, 0x20400);
+	else
+		return -ENODEV;
+}
+
+void
+nv84_sensor_setup(struct nouveau_therm *therm)
+{
+	struct nouveau_fuse *fuse = nouveau_fuse(therm);
+
+	/* enable temperature reading for cards with insane defaults */
+	if (nv_ro32(fuse, 0x1a8) == 1) {
+		nv_mask(therm, 0x20008, 0x80008000, 0x80000000);
+		nv_mask(therm, 0x2000c, 0x80000003, 0x00000000);
+		mdelay(20); /* wait for the temperature to stabilize */
+	}
 }
 
 static void
@@ -171,6 +190,21 @@ nv84_therm_intr(struct nouveau_subdev *subdev)
 }
 
 static int
+nv84_therm_init(struct nouveau_object *object)
+{
+	struct nv84_therm_priv *priv = (void *)object;
+	int ret;
+
+	ret = nouveau_therm_init(&priv->base.base);
+	if (ret)
+		return ret;
+
+	nv84_sensor_setup(&priv->base.base);
+
+	return 0;
+}
+
+static int
 nv84_therm_ctor(struct nouveau_object *parent,
 		struct nouveau_object *engine,
 		struct nouveau_oclass *oclass, void *data, u32 size,
@@ -228,7 +262,7 @@ nv84_therm_oclass = {
 	.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv84_therm_ctor,
 		.dtor = _nouveau_therm_dtor,
-		.init = _nouveau_therm_init,
+		.init = nv84_therm_init,
 		.fini = nv84_therm_fini,
 	},
 };
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/nva3.c b/drivers/gpu/drm/nouveau/core/subdev/therm/nva3.c
index 0478b2e3fb1d..7893357a7e9f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/nva3.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/nva3.c
@@ -51,6 +51,8 @@ nva3_therm_init(struct nouveau_object *object)
 	if (ret)
 		return ret;
 
+	nv84_sensor_setup(&priv->base.base);
+
 	/* enable fan tach, count revolutions per-second */
 	nv_mask(priv, 0x00e720, 0x00000003, 0x00000002);
 	if (tach->func != DCB_GPIO_UNUSED) {
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/nvd0.c b/drivers/gpu/drm/nouveau/core/subdev/therm/nvd0.c
index bbf117be572f..b70f7cc649b8 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/nvd0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/nvd0.c
@@ -114,7 +114,7 @@ nvd0_fan_pwm_clock(struct nouveau_therm *therm, int line)
 		return nv_device(therm)->crystal * 1000 / 10;
 }
 
-static int
+int
 nvd0_therm_init(struct nouveau_object *object)
 {
 	struct nvd0_therm_priv *priv = (void *)object;
@@ -150,6 +150,8 @@ nvd0_therm_ctor(struct nouveau_object *parent,
 	if (ret)
 		return ret;
 
+	nv84_sensor_setup(&priv->base.base);
+
 	priv->base.base.pwm_ctrl = nvd0_fan_pwm_ctrl;
 	priv->base.base.pwm_get = nvd0_fan_pwm_get;
 	priv->base.base.pwm_set = nvd0_fan_pwm_set;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/priv.h b/drivers/gpu/drm/nouveau/core/subdev/therm/priv.h
index 916fca5c7816..7dba8c281a0b 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/priv.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/priv.h
@@ -145,10 +145,13 @@ int nv50_fan_pwm_get(struct nouveau_therm *, int, u32 *, u32 *);
 int nv50_fan_pwm_set(struct nouveau_therm *, int, u32, u32);
 int nv50_fan_pwm_clock(struct nouveau_therm *, int);
 int nv84_temp_get(struct nouveau_therm *therm);
+void nv84_sensor_setup(struct nouveau_therm *therm);
 int nv84_therm_fini(struct nouveau_object *object, bool suspend);
 
 int nva3_therm_fan_sense(struct nouveau_therm *);
 
+int nvd0_therm_init(struct nouveau_object *object);
+
 int nouveau_fanpwm_create(struct nouveau_therm *, struct dcb_gpio_func *);
 int nouveau_fantog_create(struct nouveau_therm *, struct dcb_gpio_func *);
 int nouveau_fannil_create(struct nouveau_therm *);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/vm/base.c b/drivers/gpu/drm/nouveau/core/subdev/vm/base.c
index 7dd680ff2f6f..f75a683bd47a 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/vm/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/vm/base.c
@@ -296,7 +296,7 @@ nouveau_vm_get(struct nouveau_vm *vm, u64 size, u32 page_shift,
 	int ret;
 
 	mutex_lock(&nv_subdev(vmm)->mutex);
-	ret = nouveau_mm_head(&vm->mm, page_shift, msize, msize, align,
+	ret = nouveau_mm_head(&vm->mm, 0, page_shift, msize, msize, align,
 			     &vma->node);
 	if (unlikely(ret != 0)) {
 		mutex_unlock(&nv_subdev(vmm)->mutex);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/volt/base.c b/drivers/gpu/drm/nouveau/core/subdev/volt/base.c
index 32794a999106..26ccd8df193f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/volt/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/volt/base.c
@@ -101,6 +101,41 @@ nouveau_volt_set_id(struct nouveau_volt *volt, u8 id, int condition)
 	return ret;
 }
 
+static void nouveau_volt_parse_bios(struct nouveau_bios *bios,
+		struct nouveau_volt *volt)
+{
+	struct nvbios_volt_entry ivid;
+	struct nvbios_volt info;
+	u8  ver, hdr, cnt, len;
+	u16 data;
+	int i;
+
+	data = nvbios_volt_parse(bios, &ver, &hdr, &cnt, &len, &info);
+	if (data && info.vidmask && info.base && info.step) {
+		for (i = 0; i < info.vidmask + 1; i++) {
+			if (info.base >= info.min &&
+				info.base <= info.max) {
+				volt->vid[volt->vid_nr].uv = info.base;
+				volt->vid[volt->vid_nr].vid = i;
+				volt->vid_nr++;
+			}
+			info.base += info.step;
+		}
+		volt->vid_mask = info.vidmask;
+	} else if (data && info.vidmask) {
+		for (i = 0; i < cnt; i++) {
+			data = nvbios_volt_entry_parse(bios, i, &ver, &hdr,
+							  &ivid);
+			if (data) {
+				volt->vid[volt->vid_nr].uv = ivid.voltage;
+				volt->vid[volt->vid_nr].vid = ivid.vid;
+				volt->vid_nr++;
+			}
+		}
+		volt->vid_mask = info.vidmask;
+	}
+}
+
 int
 _nouveau_volt_init(struct nouveau_object *object)
 {
@@ -136,10 +171,6 @@ nouveau_volt_create_(struct nouveau_object *parent,
 {
 	struct nouveau_bios *bios = nouveau_bios(parent);
 	struct nouveau_volt *volt;
-	struct nvbios_volt_entry ivid;
-	struct nvbios_volt info;
-	u8  ver, hdr, cnt, len;
-	u16 data;
 	int ret, i;
 
 	ret = nouveau_subdev_create_(parent, engine, oclass, 0, "VOLT",
@@ -152,31 +183,9 @@ nouveau_volt_create_(struct nouveau_object *parent,
 	volt->set = nouveau_volt_set;
 	volt->set_id = nouveau_volt_set_id;
 
-	data = nvbios_volt_parse(bios, &ver, &hdr, &cnt, &len, &info);
-	if (data && info.vidmask && info.base && info.step) {
-		for (i = 0; i < info.vidmask + 1; i++) {
-			if (info.base >= info.min &&
-			    info.base <= info.max) {
-				volt->vid[volt->vid_nr].uv = info.base;
-				volt->vid[volt->vid_nr].vid = i;
-				volt->vid_nr++;
-			}
-			info.base += info.step;
-		}
-		volt->vid_mask = info.vidmask;
-	} else
-	if (data && info.vidmask) {
-		for (i = 0; i < cnt; i++) {
-			data = nvbios_volt_entry_parse(bios, i, &ver, &hdr,
-						      &ivid);
-			if (data) {
-				volt->vid[volt->vid_nr].uv = ivid.voltage;
-				volt->vid[volt->vid_nr].vid = ivid.vid;
-				volt->vid_nr++;
-			}
-		}
-		volt->vid_mask = info.vidmask;
-	}
+	/* Assuming the non-bios device should build the voltage table later */
+	if (bios)
+		nouveau_volt_parse_bios(bios, volt);
 
 	if (volt->vid_nr) {
 		for (i = 0; i < volt->vid_nr; i++) {
diff --git a/drivers/gpu/drm/nouveau/core/subdev/volt/gk20a.c b/drivers/gpu/drm/nouveau/core/subdev/volt/gk20a.c
new file mode 100644
index 000000000000..717368ef31ac
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/volt/gk20a.c
@@ -0,0 +1,199 @@
+/*
+ * Copyright (c) 2014, NVIDIA CORPORATION. All rights reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#ifdef __KERNEL__
+#include <nouveau_platform.h>
+#endif
+#include <subdev/volt.h>
+
+struct cvb_coef {
+	int c0;
+	int c1;
+	int c2;
+	int c3;
+	int c4;
+	int c5;
+};
+
+struct gk20a_volt_priv {
+	struct nouveau_volt base;
+	struct regulator *vdd;
+};
+
+const struct cvb_coef gk20a_cvb_coef[] = {
+	/* MHz,        c0,     c1,   c2,    c3,     c4,   c5 */
+	/*  72 */ { 1209886, -36468,  515,   417, -13123,  203},
+	/* 108 */ { 1130804, -27659,  296,   298, -10834,  221},
+	/* 180 */ { 1162871, -27110,  247,   238, -10681,  268},
+	/* 252 */ { 1220458, -28654,  247,   179, -10376,  298},
+	/* 324 */ { 1280953, -30204,  247,   119,  -9766,  304},
+	/* 396 */ { 1344547, -31777,  247,   119,  -8545,  292},
+	/* 468 */ { 1420168, -34227,  269,    60,  -7172,  256},
+	/* 540 */ { 1490757, -35955,  274,    60,  -5188,  197},
+	/* 612 */ { 1599112, -42583,  398,     0,  -1831,  119},
+	/* 648 */ { 1366986, -16459, -274,     0,  -3204,   72},
+	/* 684 */ { 1391884, -17078, -274,   -60,  -1526,   30},
+	/* 708 */ { 1415522, -17497, -274,   -60,   -458,    0},
+	/* 756 */ { 1464061, -18331, -274,  -119,   1831,  -72},
+	/* 804 */ { 1524225, -20064, -254,  -119,   4272, -155},
+	/* 852 */ { 1608418, -21643, -269,     0,    763,  -48},
+};
+
+/**
+ * cvb_mv = ((c2 * speedo / s_scale + c1) * speedo / s_scale + c0)
+ */
+static inline int
+gk20a_volt_get_cvb_voltage(int speedo, int s_scale,
+		const struct cvb_coef *coef)
+{
+	int mv;
+
+	mv = DIV_ROUND_CLOSEST(coef->c2 * speedo, s_scale);
+	mv = DIV_ROUND_CLOSEST((mv + coef->c1) * speedo, s_scale) + coef->c0;
+	return mv;
+}
+
+/**
+ * cvb_t_mv =
+ * ((c2 * speedo / s_scale + c1) * speedo / s_scale + c0) +
+ * ((c3 * speedo / s_scale + c4 + c5 * T / t_scale) * T / t_scale)
+ */
+static inline int
+gk20a_volt_get_cvb_t_voltage(int speedo, int temp, int s_scale, int t_scale,
+		const struct cvb_coef *coef)
+{
+	int cvb_mv, mv;
+
+	cvb_mv = gk20a_volt_get_cvb_voltage(speedo, s_scale, coef);
+
+	mv = DIV_ROUND_CLOSEST(coef->c3 * speedo, s_scale) + coef->c4 +
+		DIV_ROUND_CLOSEST(coef->c5 * temp, t_scale);
+	mv = DIV_ROUND_CLOSEST(mv * temp, t_scale) + cvb_mv;
+	return mv;
+}
+
+static int
+gk20a_volt_calc_voltage(const struct cvb_coef *coef, int speedo)
+{
+	int mv;
+
+	mv = gk20a_volt_get_cvb_t_voltage(speedo, -10, 100, 10, coef);
+	mv = DIV_ROUND_UP(mv, 1000);
+
+	return mv * 1000;
+}
+
+static int
+gk20a_volt_vid_get(struct nouveau_volt *volt)
+{
+	struct gk20a_volt_priv *priv = (void *)volt;
+	int i, uv;
+
+	uv = regulator_get_voltage(priv->vdd);
+
+	for (i = 0; i < volt->vid_nr; i++)
+		if (volt->vid[i].uv >= uv)
+			return i;
+
+	return -EINVAL;
+}
+
+static int
+gk20a_volt_vid_set(struct nouveau_volt *volt, u8 vid)
+{
+	struct gk20a_volt_priv *priv = (void *)volt;
+
+	nv_debug(volt, "set voltage as %duv\n", volt->vid[vid].uv);
+	return regulator_set_voltage(priv->vdd, volt->vid[vid].uv, 1200000);
+}
+
+static int
+gk20a_volt_set_id(struct nouveau_volt *volt, u8 id, int condition)
+{
+	struct gk20a_volt_priv *priv = (void *)volt;
+	int prev_uv = regulator_get_voltage(priv->vdd);
+	int target_uv = volt->vid[id].uv;
+	int ret;
+
+	nv_debug(volt, "prev=%d, target=%d, condition=%d\n",
+			prev_uv, target_uv, condition);
+	if (!condition ||
+		(condition < 0 && target_uv < prev_uv) ||
+		(condition > 0 && target_uv > prev_uv)) {
+		ret = gk20a_volt_vid_set(volt, volt->vid[id].vid);
+	} else {
+		ret = 0;
+	}
+
+	return ret;
+}
+
+static int
+gk20a_volt_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+	       struct nouveau_oclass *oclass, void *data, u32 size,
+	       struct nouveau_object **pobject)
+{
+	struct gk20a_volt_priv *priv;
+	struct nouveau_volt *volt;
+	struct nouveau_platform_device *plat;
+	int i, ret, uv;
+
+	ret = nouveau_volt_create(parent, engine, oclass, &priv);
+	*pobject = nv_object(priv);
+	if (ret)
+		return ret;
+
+	volt = &priv->base;
+
+	plat = nv_device_to_platform(nv_device(parent));
+
+	uv = regulator_get_voltage(plat->gpu->vdd);
+	nv_info(priv, "The default voltage is %duV\n", uv);
+
+	priv->vdd = plat->gpu->vdd;
+	priv->base.vid_get = gk20a_volt_vid_get;
+	priv->base.vid_set = gk20a_volt_vid_set;
+	priv->base.set_id = gk20a_volt_set_id;
+
+	volt->vid_nr = ARRAY_SIZE(gk20a_cvb_coef);
+	nv_debug(priv, "%s - vid_nr = %d\n", __func__, volt->vid_nr);
+	for (i = 0; i < volt->vid_nr; i++) {
+		volt->vid[i].vid = i;
+		volt->vid[i].uv = gk20a_volt_calc_voltage(&gk20a_cvb_coef[i],
+					plat->gpu_speedo);
+		nv_debug(priv, "%2d: vid=%d, uv=%d\n", i, volt->vid[i].vid,
+					volt->vid[i].uv);
+	}
+
+	return 0;
+}
+
+struct nouveau_oclass
+gk20a_volt_oclass = {
+	.handle = NV_SUBDEV(VOLT, 0xea),
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = gk20a_volt_ctor,
+		.dtor = _nouveau_volt_dtor,
+		.init = _nouveau_volt_init,
+		.fini = _nouveau_volt_fini,
+	},
+};
diff --git a/drivers/gpu/drm/nouveau/dispnv04/crtc.c b/drivers/gpu/drm/nouveau/dispnv04/crtc.c
index b90aa5c1f90a..38402ade6835 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/crtc.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/crtc.c
@@ -26,6 +26,7 @@
 
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_plane_helper.h>
 
 #include "nouveau_drm.h"
 #include "nouveau_reg.h"
@@ -613,7 +614,7 @@ nv_crtc_swap_fbs(struct drm_crtc *crtc, struct drm_framebuffer *old_fb)
 	struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
 	int ret;
 
-	ret = nouveau_bo_pin(nvfb->nvbo, TTM_PL_FLAG_VRAM);
+	ret = nouveau_bo_pin(nvfb->nvbo, TTM_PL_FLAG_VRAM, false);
 	if (ret == 0) {
 		if (disp->image[nv_crtc->index])
 			nouveau_bo_unpin(disp->image[nv_crtc->index]);
@@ -1127,9 +1128,9 @@ nv04_crtc_create(struct drm_device *dev, int crtc_num)
 	drm_mode_crtc_set_gamma_size(&nv_crtc->base, 256);
 
 	ret = nouveau_bo_new(dev, 64*64*4, 0x100, TTM_PL_FLAG_VRAM,
-			     0, 0x0000, NULL, &nv_crtc->cursor.nvbo);
+			     0, 0x0000, NULL, NULL, &nv_crtc->cursor.nvbo);
 	if (!ret) {
-		ret = nouveau_bo_pin(nv_crtc->cursor.nvbo, TTM_PL_FLAG_VRAM);
+		ret = nouveau_bo_pin(nv_crtc->cursor.nvbo, TTM_PL_FLAG_VRAM, false);
 		if (!ret) {
 			ret = nouveau_bo_map(nv_crtc->cursor.nvbo);
 			if (ret)
diff --git a/drivers/gpu/drm/nouveau/dispnv04/overlay.c b/drivers/gpu/drm/nouveau/dispnv04/overlay.c
index b36afcbbc83f..9f2498571d09 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/overlay.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/overlay.c
@@ -97,7 +97,8 @@ nv10_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 		  uint32_t src_w, uint32_t src_h)
 {
 	struct nvif_device *dev = &nouveau_drm(plane->dev)->device;
-	struct nouveau_plane *nv_plane = (struct nouveau_plane *)plane;
+	struct nouveau_plane *nv_plane =
+		container_of(plane, struct nouveau_plane, base);
 	struct nouveau_framebuffer *nv_fb = nouveau_framebuffer(fb);
 	struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
 	struct nouveau_bo *cur = nv_plane->cur;
@@ -125,7 +126,7 @@ nv10_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 			return -ERANGE;
 	}
 
-	ret = nouveau_bo_pin(nv_fb->nvbo, TTM_PL_FLAG_VRAM);
+	ret = nouveau_bo_pin(nv_fb->nvbo, TTM_PL_FLAG_VRAM, false);
 	if (ret)
 		return ret;
 
@@ -173,7 +174,8 @@ static int
 nv10_disable_plane(struct drm_plane *plane)
 {
 	struct nvif_device *dev = &nouveau_drm(plane->dev)->device;
-	struct nouveau_plane *nv_plane = (struct nouveau_plane *)plane;
+	struct nouveau_plane *nv_plane =
+		container_of(plane, struct nouveau_plane, base);
 
 	nvif_wr32(dev, NV_PVIDEO_STOP, 1);
 	if (nv_plane->cur) {
@@ -224,7 +226,8 @@ nv_set_property(struct drm_plane *plane,
 		struct drm_property *property,
 		uint64_t value)
 {
-	struct nouveau_plane *nv_plane = (struct nouveau_plane *)plane;
+	struct nouveau_plane *nv_plane =
+		container_of(plane, struct nouveau_plane, base);
 
 	if (property == nv_plane->props.colorkey)
 		nv_plane->colorkey = value;
@@ -344,7 +347,8 @@ nv04_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 		  uint32_t src_w, uint32_t src_h)
 {
 	struct nvif_device *dev = &nouveau_drm(plane->dev)->device;
-	struct nouveau_plane *nv_plane = (struct nouveau_plane *)plane;
+	struct nouveau_plane *nv_plane =
+		container_of(plane, struct nouveau_plane, base);
 	struct nouveau_framebuffer *nv_fb = nouveau_framebuffer(fb);
 	struct nouveau_bo *cur = nv_plane->cur;
 	uint32_t overlay = 1;
@@ -369,7 +373,7 @@ nv04_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	if (crtc_w < src_w || crtc_h < src_h)
 		return -ERANGE;
 
-	ret = nouveau_bo_pin(nv_fb->nvbo, TTM_PL_FLAG_VRAM);
+	ret = nouveau_bo_pin(nv_fb->nvbo, TTM_PL_FLAG_VRAM, false);
 	if (ret)
 		return ret;
 
@@ -423,7 +427,8 @@ static int
 nv04_disable_plane(struct drm_plane *plane)
 {
 	struct nvif_device *dev = &nouveau_drm(plane->dev)->device;
-	struct nouveau_plane *nv_plane = (struct nouveau_plane *)plane;
+	struct nouveau_plane *nv_plane =
+		container_of(plane, struct nouveau_plane, base);
 
 	nvif_mask(dev, NV_PVIDEO_OVERLAY, 1, 0);
 	nvif_wr32(dev, NV_PVIDEO_OE_STATE, 0);
diff --git a/drivers/gpu/drm/nouveau/nouveau_abi16.c b/drivers/gpu/drm/nouveau/nouveau_abi16.c
index 615714c1727d..d39a15000068 100644
--- a/drivers/gpu/drm/nouveau/nouveau_abi16.c
+++ b/drivers/gpu/drm/nouveau/nouveau_abi16.c
@@ -308,7 +308,7 @@ nouveau_abi16_ioctl_channel_alloc(ABI16_IOCTL_ARGS)
 	ret = nouveau_gem_new(dev, PAGE_SIZE, 0, NOUVEAU_GEM_DOMAIN_GART,
 			      0, 0, &chan->ntfy);
 	if (ret == 0)
-		ret = nouveau_bo_pin(chan->ntfy, TTM_PL_FLAG_TT);
+		ret = nouveau_bo_pin(chan->ntfy, TTM_PL_FLAG_TT, false);
 	if (ret)
 		goto done;
 
@@ -448,7 +448,7 @@ nouveau_abi16_ioctl_notifierobj_alloc(ABI16_IOCTL_ARGS)
 	list_add(&ntfy->head, &chan->notifiers);
 	ntfy->handle = info->handle;
 
-	ret = nouveau_mm_head(&chan->heap, 1, info->size, info->size, 1,
+	ret = nouveau_mm_head(&chan->heap, 0, 1, info->size, info->size, 1,
 			      &ntfy->node);
 	if (ret)
 		goto done;
diff --git a/drivers/gpu/drm/nouveau/nouveau_bios.c b/drivers/gpu/drm/nouveau/nouveau_bios.c
index dae2c96deef8..7df6acc8bb34 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bios.c
+++ b/drivers/gpu/drm/nouveau/nouveau_bios.c
@@ -1258,7 +1258,7 @@ olddcb_table(struct drm_device *dev)
 		return NULL;
 	}
 
-	if (dcb[0] >= 0x41) {
+	if (dcb[0] >= 0x42) {
 		NV_WARN(drm, "DCB version 0x%02x unknown\n", dcb[0]);
 		return NULL;
 	} else
@@ -1481,18 +1481,22 @@ parse_dcb20_entry(struct drm_device *dev, struct dcb_table *dcb,
 			entry->dpconf.link_bw = 540000;
 			break;
 		}
-		switch ((conf & 0x0f000000) >> 24) {
-		case 0xf:
-			entry->dpconf.link_nr = 4;
-			break;
-		case 0x3:
-			entry->dpconf.link_nr = 2;
-			break;
-		default:
-			entry->dpconf.link_nr = 1;
-			break;
+		entry->dpconf.link_nr = (conf & 0x0f000000) >> 24;
+		if (dcb->version < 0x41) {
+			switch (entry->dpconf.link_nr) {
+			case 0xf:
+				entry->dpconf.link_nr = 4;
+				break;
+			case 0x3:
+				entry->dpconf.link_nr = 2;
+				break;
+			default:
+				entry->dpconf.link_nr = 1;
+				break;
+			}
 		}
 		link = entry->dpconf.sor.link;
+		entry->i2c_index += NV_I2C_AUX(0);
 		break;
 	case DCB_OUTPUT_TMDS:
 		if (dcb->version >= 0x40) {
diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c
index 01da508625f2..bba2960d3dfb 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bo.c
+++ b/drivers/gpu/drm/nouveau/nouveau_bo.c
@@ -88,13 +88,13 @@ nv10_bo_get_tile_region(struct drm_device *dev, int i)
 
 static void
 nv10_bo_put_tile_region(struct drm_device *dev, struct nouveau_drm_tile *tile,
-			struct nouveau_fence *fence)
+			struct fence *fence)
 {
 	struct nouveau_drm *drm = nouveau_drm(dev);
 
 	if (tile) {
 		spin_lock(&drm->tile.lock);
-		tile->fence = nouveau_fence_ref(fence);
+		tile->fence = (struct nouveau_fence *)fence_get(fence);
 		tile->used = false;
 		spin_unlock(&drm->tile.lock);
 	}
@@ -181,7 +181,7 @@ nouveau_bo_fixup_align(struct nouveau_bo *nvbo, u32 flags,
 int
 nouveau_bo_new(struct drm_device *dev, int size, int align,
 	       uint32_t flags, uint32_t tile_mode, uint32_t tile_flags,
-	       struct sg_table *sg,
+	       struct sg_table *sg, struct reservation_object *robj,
 	       struct nouveau_bo **pnvbo)
 {
 	struct nouveau_drm *drm = nouveau_drm(dev);
@@ -214,6 +214,9 @@ nouveau_bo_new(struct drm_device *dev, int size, int align,
 	nvbo->tile_flags = tile_flags;
 	nvbo->bo.bdev = &drm->ttm.bdev;
 
+	if (!nv_device_is_cpu_coherent(nvkm_device(&drm->device)))
+		nvbo->force_coherent = flags & TTM_PL_FLAG_UNCACHED;
+
 	nvbo->page_shift = 12;
 	if (drm->client.vm) {
 		if (!(flags & TTM_PL_FLAG_TT) && size > 256 * 1024)
@@ -230,7 +233,7 @@ nouveau_bo_new(struct drm_device *dev, int size, int align,
 	ret = ttm_bo_init(&drm->ttm.bdev, &nvbo->bo, size,
 			  type, &nvbo->placement,
 			  align >> PAGE_SHIFT, false, NULL, acc_size, sg,
-			  nouveau_bo_del_ttm);
+			  robj, nouveau_bo_del_ttm);
 	if (ret) {
 		/* ttm will call nouveau_bo_del_ttm if it fails.. */
 		return ret;
@@ -241,16 +244,16 @@ nouveau_bo_new(struct drm_device *dev, int size, int align,
 }
 
 static void
-set_placement_list(uint32_t *pl, unsigned *n, uint32_t type, uint32_t flags)
+set_placement_list(struct ttm_place *pl, unsigned *n, uint32_t type, uint32_t flags)
 {
 	*n = 0;
 
 	if (type & TTM_PL_FLAG_VRAM)
-		pl[(*n)++] = TTM_PL_FLAG_VRAM | flags;
+		pl[(*n)++].flags = TTM_PL_FLAG_VRAM | flags;
 	if (type & TTM_PL_FLAG_TT)
-		pl[(*n)++] = TTM_PL_FLAG_TT | flags;
+		pl[(*n)++].flags = TTM_PL_FLAG_TT | flags;
 	if (type & TTM_PL_FLAG_SYSTEM)
-		pl[(*n)++] = TTM_PL_FLAG_SYSTEM | flags;
+		pl[(*n)++].flags = TTM_PL_FLAG_SYSTEM | flags;
 }
 
 static void
@@ -258,6 +261,7 @@ set_placement_range(struct nouveau_bo *nvbo, uint32_t type)
 {
 	struct nouveau_drm *drm = nouveau_bdev(nvbo->bo.bdev);
 	u32 vram_pages = drm->device.info.ram_size >> PAGE_SHIFT;
+	unsigned i, fpfn, lpfn;
 
 	if (drm->device.info.family == NV_DEVICE_INFO_V0_CELSIUS &&
 	    nvbo->tile_mode && (type & TTM_PL_FLAG_VRAM) &&
@@ -269,11 +273,19 @@ set_placement_range(struct nouveau_bo *nvbo, uint32_t type)
 		 * at the same time.
 		 */
 		if (nvbo->tile_flags & NOUVEAU_GEM_TILE_ZETA) {
-			nvbo->placement.fpfn = vram_pages / 2;
-			nvbo->placement.lpfn = ~0;
+			fpfn = vram_pages / 2;
+			lpfn = ~0;
 		} else {
-			nvbo->placement.fpfn = 0;
-			nvbo->placement.lpfn = vram_pages / 2;
+			fpfn = 0;
+			lpfn = vram_pages / 2;
+		}
+		for (i = 0; i < nvbo->placement.num_placement; ++i) {
+			nvbo->placements[i].fpfn = fpfn;
+			nvbo->placements[i].lpfn = lpfn;
+		}
+		for (i = 0; i < nvbo->placement.num_busy_placement; ++i) {
+			nvbo->busy_placements[i].fpfn = fpfn;
+			nvbo->busy_placements[i].lpfn = lpfn;
 		}
 	}
 }
@@ -282,8 +294,9 @@ void
 nouveau_bo_placement_set(struct nouveau_bo *nvbo, uint32_t type, uint32_t busy)
 {
 	struct ttm_placement *pl = &nvbo->placement;
-	uint32_t flags = TTM_PL_MASK_CACHING |
-		(nvbo->pin_refcnt ? TTM_PL_FLAG_NO_EVICT : 0);
+	uint32_t flags = (nvbo->force_coherent ? TTM_PL_FLAG_UNCACHED :
+						 TTM_PL_MASK_CACHING) |
+			 (nvbo->pin_refcnt ? TTM_PL_FLAG_NO_EVICT : 0);
 
 	pl->placement = nvbo->placements;
 	set_placement_list(nvbo->placements, &pl->num_placement,
@@ -297,42 +310,75 @@ nouveau_bo_placement_set(struct nouveau_bo *nvbo, uint32_t type, uint32_t busy)
 }
 
 int
-nouveau_bo_pin(struct nouveau_bo *nvbo, uint32_t memtype)
+nouveau_bo_pin(struct nouveau_bo *nvbo, uint32_t memtype, bool contig)
 {
 	struct nouveau_drm *drm = nouveau_bdev(nvbo->bo.bdev);
 	struct ttm_buffer_object *bo = &nvbo->bo;
+	bool force = false, evict = false;
 	int ret;
 
 	ret = ttm_bo_reserve(bo, false, false, false, NULL);
 	if (ret)
-		goto out;
+		return ret;
 
-	if (nvbo->pin_refcnt && !(memtype & (1 << bo->mem.mem_type))) {
-		NV_ERROR(drm, "bo %p pinned elsewhere: 0x%08x vs 0x%08x\n", bo,
-			 1 << bo->mem.mem_type, memtype);
-		ret = -EINVAL;
-		goto out;
+	if (drm->device.info.family >= NV_DEVICE_INFO_V0_TESLA &&
+	    memtype == TTM_PL_FLAG_VRAM && contig) {
+		if (nvbo->tile_flags & NOUVEAU_GEM_TILE_NONCONTIG) {
+			if (bo->mem.mem_type == TTM_PL_VRAM) {
+				struct nouveau_mem *mem = bo->mem.mm_node;
+				if (!list_is_singular(&mem->regions))
+					evict = true;
+			}
+			nvbo->tile_flags &= ~NOUVEAU_GEM_TILE_NONCONTIG;
+			force = true;
+		}
 	}
 
-	if (nvbo->pin_refcnt++)
+	if (nvbo->pin_refcnt) {
+		if (!(memtype & (1 << bo->mem.mem_type)) || evict) {
+			NV_ERROR(drm, "bo %p pinned elsewhere: "
+				      "0x%08x vs 0x%08x\n", bo,
+				 1 << bo->mem.mem_type, memtype);
+			ret = -EBUSY;
+		}
+		nvbo->pin_refcnt++;
 		goto out;
+	}
+
+	if (evict) {
+		nouveau_bo_placement_set(nvbo, TTM_PL_FLAG_TT, 0);
+		ret = nouveau_bo_validate(nvbo, false, false);
+		if (ret)
+			goto out;
+	}
 
+	nvbo->pin_refcnt++;
 	nouveau_bo_placement_set(nvbo, memtype, 0);
 
+	/* drop pin_refcnt temporarily, so we don't trip the assertion
+	 * in nouveau_bo_move() that makes sure we're not trying to
+	 * move a pinned buffer
+	 */
+	nvbo->pin_refcnt--;
 	ret = nouveau_bo_validate(nvbo, false, false);
-	if (ret == 0) {
-		switch (bo->mem.mem_type) {
-		case TTM_PL_VRAM:
-			drm->gem.vram_available -= bo->mem.size;
-			break;
-		case TTM_PL_TT:
-			drm->gem.gart_available -= bo->mem.size;
-			break;
-		default:
-			break;
-		}
+	if (ret)
+		goto out;
+	nvbo->pin_refcnt++;
+
+	switch (bo->mem.mem_type) {
+	case TTM_PL_VRAM:
+		drm->gem.vram_available -= bo->mem.size;
+		break;
+	case TTM_PL_TT:
+		drm->gem.gart_available -= bo->mem.size;
+		break;
+	default:
+		break;
 	}
+
 out:
+	if (force && ret)
+		nvbo->tile_flags |= NOUVEAU_GEM_TILE_NONCONTIG;
 	ttm_bo_unreserve(bo);
 	return ret;
 }
@@ -383,7 +429,14 @@ nouveau_bo_map(struct nouveau_bo *nvbo)
 	if (ret)
 		return ret;
 
-	ret = ttm_bo_kmap(&nvbo->bo, 0, nvbo->bo.mem.num_pages, &nvbo->kmap);
+	/*
+	 * TTM buffers allocated using the DMA API already have a mapping, let's
+	 * use it instead.
+	 */
+	if (!nvbo->force_coherent)
+		ret = ttm_bo_kmap(&nvbo->bo, 0, nvbo->bo.mem.num_pages,
+				  &nvbo->kmap);
+
 	ttm_bo_unreserve(&nvbo->bo);
 	return ret;
 }
@@ -391,10 +444,57 @@ nouveau_bo_map(struct nouveau_bo *nvbo)
 void
 nouveau_bo_unmap(struct nouveau_bo *nvbo)
 {
-	if (nvbo)
+	if (!nvbo)
+		return;
+
+	/*
+	 * TTM buffers allocated using the DMA API already had a coherent
+	 * mapping which we used, no need to unmap.
+	 */
+	if (!nvbo->force_coherent)
 		ttm_bo_kunmap(&nvbo->kmap);
 }
 
+void
+nouveau_bo_sync_for_device(struct nouveau_bo *nvbo)
+{
+	struct nouveau_drm *drm = nouveau_bdev(nvbo->bo.bdev);
+	struct nouveau_device *device = nvkm_device(&drm->device);
+	struct ttm_dma_tt *ttm_dma = (struct ttm_dma_tt *)nvbo->bo.ttm;
+	int i;
+
+	if (!ttm_dma)
+		return;
+
+	/* Don't waste time looping if the object is coherent */
+	if (nvbo->force_coherent)
+		return;
+
+	for (i = 0; i < ttm_dma->ttm.num_pages; i++)
+		dma_sync_single_for_device(nv_device_base(device),
+			ttm_dma->dma_address[i], PAGE_SIZE, DMA_TO_DEVICE);
+}
+
+void
+nouveau_bo_sync_for_cpu(struct nouveau_bo *nvbo)
+{
+	struct nouveau_drm *drm = nouveau_bdev(nvbo->bo.bdev);
+	struct nouveau_device *device = nvkm_device(&drm->device);
+	struct ttm_dma_tt *ttm_dma = (struct ttm_dma_tt *)nvbo->bo.ttm;
+	int i;
+
+	if (!ttm_dma)
+		return;
+
+	/* Don't waste time looping if the object is coherent */
+	if (nvbo->force_coherent)
+		return;
+
+	for (i = 0; i < ttm_dma->ttm.num_pages; i++)
+		dma_sync_single_for_cpu(nv_device_base(device),
+			ttm_dma->dma_address[i], PAGE_SIZE, DMA_FROM_DEVICE);
+}
+
 int
 nouveau_bo_validate(struct nouveau_bo *nvbo, bool interruptible,
 		    bool no_wait_gpu)
@@ -406,15 +506,41 @@ nouveau_bo_validate(struct nouveau_bo *nvbo, bool interruptible,
 	if (ret)
 		return ret;
 
+	nouveau_bo_sync_for_device(nvbo);
+
 	return 0;
 }
 
+static inline void *
+_nouveau_bo_mem_index(struct nouveau_bo *nvbo, unsigned index, void *mem, u8 sz)
+{
+	struct ttm_dma_tt *dma_tt;
+	u8 *m = mem;
+
+	index *= sz;
+
+	if (m) {
+		/* kmap'd address, return the corresponding offset */
+		m += index;
+	} else {
+		/* DMA-API mapping, lookup the right address */
+		dma_tt = (struct ttm_dma_tt *)nvbo->bo.ttm;
+		m = dma_tt->cpu_address[index / PAGE_SIZE];
+		m += index % PAGE_SIZE;
+	}
+
+	return m;
+}
+#define nouveau_bo_mem_index(o, i, m) _nouveau_bo_mem_index(o, i, m, sizeof(*m))
+
 u16
 nouveau_bo_rd16(struct nouveau_bo *nvbo, unsigned index)
 {
 	bool is_iomem;
 	u16 *mem = ttm_kmap_obj_virtual(&nvbo->kmap, &is_iomem);
-	mem = &mem[index];
+
+	mem = nouveau_bo_mem_index(nvbo, index, mem);
+
 	if (is_iomem)
 		return ioread16_native((void __force __iomem *)mem);
 	else
@@ -426,7 +552,9 @@ nouveau_bo_wr16(struct nouveau_bo *nvbo, unsigned index, u16 val)
 {
 	bool is_iomem;
 	u16 *mem = ttm_kmap_obj_virtual(&nvbo->kmap, &is_iomem);
-	mem = &mem[index];
+
+	mem = nouveau_bo_mem_index(nvbo, index, mem);
+
 	if (is_iomem)
 		iowrite16_native(val, (void __force __iomem *)mem);
 	else
@@ -438,7 +566,9 @@ nouveau_bo_rd32(struct nouveau_bo *nvbo, unsigned index)
 {
 	bool is_iomem;
 	u32 *mem = ttm_kmap_obj_virtual(&nvbo->kmap, &is_iomem);
-	mem = &mem[index];
+
+	mem = nouveau_bo_mem_index(nvbo, index, mem);
+
 	if (is_iomem)
 		return ioread32_native((void __force __iomem *)mem);
 	else
@@ -450,7 +580,9 @@ nouveau_bo_wr32(struct nouveau_bo *nvbo, unsigned index, u32 val)
 {
 	bool is_iomem;
 	u32 *mem = ttm_kmap_obj_virtual(&nvbo->kmap, &is_iomem);
-	mem = &mem[index];
+
+	mem = nouveau_bo_mem_index(nvbo, index, mem);
+
 	if (is_iomem)
 		iowrite32_native(val, (void __force __iomem *)mem);
 	else
@@ -961,13 +1093,14 @@ nouveau_bo_move_m2mf(struct ttm_buffer_object *bo, int evict, bool intr,
 	}
 
 	mutex_lock_nested(&cli->mutex, SINGLE_DEPTH_NESTING);
-	ret = nouveau_fence_sync(bo->sync_obj, chan);
+	ret = nouveau_fence_sync(nouveau_bo(bo), chan, true, intr);
 	if (ret == 0) {
 		ret = drm->ttm.move(chan, bo, &bo->mem, new_mem);
 		if (ret == 0) {
 			ret = nouveau_fence_new(chan, false, &fence);
 			if (ret == 0) {
-				ret = ttm_bo_move_accel_cleanup(bo, fence,
+				ret = ttm_bo_move_accel_cleanup(bo,
+								&fence->base,
 								evict,
 								no_wait_gpu,
 								new_mem);
@@ -1041,12 +1174,15 @@ static int
 nouveau_bo_move_flipd(struct ttm_buffer_object *bo, bool evict, bool intr,
 		      bool no_wait_gpu, struct ttm_mem_reg *new_mem)
 {
-	u32 placement_memtype = TTM_PL_FLAG_TT | TTM_PL_MASK_CACHING;
+	struct ttm_place placement_memtype = {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_FLAG_TT | TTM_PL_MASK_CACHING
+	};
 	struct ttm_placement placement;
 	struct ttm_mem_reg tmp_mem;
 	int ret;
 
-	placement.fpfn = placement.lpfn = 0;
 	placement.num_placement = placement.num_busy_placement = 1;
 	placement.placement = placement.busy_placement = &placement_memtype;
 
@@ -1074,12 +1210,15 @@ static int
 nouveau_bo_move_flips(struct ttm_buffer_object *bo, bool evict, bool intr,
 		      bool no_wait_gpu, struct ttm_mem_reg *new_mem)
 {
-	u32 placement_memtype = TTM_PL_FLAG_TT | TTM_PL_MASK_CACHING;
+	struct ttm_place placement_memtype = {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_FLAG_TT | TTM_PL_MASK_CACHING
+	};
 	struct ttm_placement placement;
 	struct ttm_mem_reg tmp_mem;
 	int ret;
 
-	placement.fpfn = placement.lpfn = 0;
 	placement.num_placement = placement.num_busy_placement = 1;
 	placement.placement = placement.busy_placement = &placement_memtype;
 
@@ -1152,8 +1291,9 @@ nouveau_bo_vm_cleanup(struct ttm_buffer_object *bo,
 {
 	struct nouveau_drm *drm = nouveau_bdev(bo->bdev);
 	struct drm_device *dev = drm->dev;
+	struct fence *fence = reservation_object_get_excl(bo->resv);
 
-	nv10_bo_put_tile_region(dev, *old_tile, bo->sync_obj);
+	nv10_bo_put_tile_region(dev, *old_tile, fence);
 	*old_tile = new_tile;
 }
 
@@ -1167,6 +1307,9 @@ nouveau_bo_move(struct ttm_buffer_object *bo, bool evict, bool intr,
 	struct nouveau_drm_tile *new_tile = NULL;
 	int ret = 0;
 
+	if (nvbo->pin_refcnt)
+		NV_WARN(drm, "Moving pinned object %p!\n", nvbo);
+
 	if (drm->device.info.family < NV_DEVICE_INFO_V0_TESLA) {
 		ret = nouveau_bo_vm_bind(bo, new_mem, &new_tile);
 		if (ret)
@@ -1197,9 +1340,7 @@ nouveau_bo_move(struct ttm_buffer_object *bo, bool evict, bool intr,
 	}
 
 	/* Fallback to software copy. */
-	spin_lock(&bo->bdev->fence_lock);
 	ret = ttm_bo_wait(bo, true, intr, no_wait_gpu);
-	spin_unlock(&bo->bdev->fence_lock);
 	if (ret == 0)
 		ret = ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem);
 
@@ -1294,7 +1435,7 @@ nouveau_ttm_fault_reserve_notify(struct ttm_buffer_object *bo)
 	struct nouveau_bo *nvbo = nouveau_bo(bo);
 	struct nvif_device *device = &drm->device;
 	u32 mappable = nv_device_resource_len(nvkm_device(device), 1) >> PAGE_SHIFT;
-	int ret;
+	int i, ret;
 
 	/* as long as the bo isn't in vram, and isn't tiled, we've got
 	 * nothing to do here.
@@ -1319,9 +1460,16 @@ nouveau_ttm_fault_reserve_notify(struct ttm_buffer_object *bo)
 	    bo->mem.start + bo->mem.num_pages < mappable)
 		return 0;
 
+	for (i = 0; i < nvbo->placement.num_placement; ++i) {
+		nvbo->placements[i].fpfn = 0;
+		nvbo->placements[i].lpfn = mappable;
+	}
+
+	for (i = 0; i < nvbo->placement.num_busy_placement; ++i) {
+		nvbo->busy_placements[i].fpfn = 0;
+		nvbo->busy_placements[i].lpfn = mappable;
+	}
 
-	nvbo->placement.fpfn = 0;
-	nvbo->placement.lpfn = mappable;
 	nouveau_bo_placement_set(nvbo, TTM_PL_FLAG_VRAM, 0);
 	return nouveau_bo_validate(nvbo, false, false);
 }
@@ -1354,6 +1502,14 @@ nouveau_ttm_tt_populate(struct ttm_tt *ttm)
 	dev = drm->dev;
 	pdev = nv_device_base(device);
 
+	/*
+	 * Objects matching this condition have been marked as force_coherent,
+	 * so use the DMA API for them.
+	 */
+	if (!nv_device_is_cpu_coherent(device) &&
+	    ttm->caching_state == tt_uncached)
+		return ttm_dma_populate(ttm_dma, dev->dev);
+
 #if __OS_HAS_AGP
 	if (drm->agp.stat == ENABLED) {
 		return ttm_agp_tt_populate(ttm);
@@ -1411,6 +1567,16 @@ nouveau_ttm_tt_unpopulate(struct ttm_tt *ttm)
 	dev = drm->dev;
 	pdev = nv_device_base(device);
 
+	/*
+	 * Objects matching this condition have been marked as force_coherent,
+	 * so use the DMA API for them.
+	 */
+	if (!nv_device_is_cpu_coherent(device) &&
+	    ttm->caching_state == tt_uncached) {
+		ttm_dma_unpopulate(ttm_dma, dev->dev);
+		return;
+	}
+
 #if __OS_HAS_AGP
 	if (drm->agp.stat == ENABLED) {
 		ttm_agp_tt_unpopulate(ttm);
@@ -1436,47 +1602,14 @@ nouveau_ttm_tt_unpopulate(struct ttm_tt *ttm)
 }
 
 void
-nouveau_bo_fence(struct nouveau_bo *nvbo, struct nouveau_fence *fence)
+nouveau_bo_fence(struct nouveau_bo *nvbo, struct nouveau_fence *fence, bool exclusive)
 {
-	struct nouveau_fence *new_fence = nouveau_fence_ref(fence);
-	struct nouveau_fence *old_fence = NULL;
-
-	spin_lock(&nvbo->bo.bdev->fence_lock);
-	old_fence = nvbo->bo.sync_obj;
-	nvbo->bo.sync_obj = new_fence;
-	spin_unlock(&nvbo->bo.bdev->fence_lock);
-
-	nouveau_fence_unref(&old_fence);
-}
-
-static void
-nouveau_bo_fence_unref(void **sync_obj)
-{
-	nouveau_fence_unref((struct nouveau_fence **)sync_obj);
-}
+	struct reservation_object *resv = nvbo->bo.resv;
 
-static void *
-nouveau_bo_fence_ref(void *sync_obj)
-{
-	return nouveau_fence_ref(sync_obj);
-}
-
-static bool
-nouveau_bo_fence_signalled(void *sync_obj)
-{
-	return nouveau_fence_done(sync_obj);
-}
-
-static int
-nouveau_bo_fence_wait(void *sync_obj, bool lazy, bool intr)
-{
-	return nouveau_fence_wait(sync_obj, lazy, intr);
-}
-
-static int
-nouveau_bo_fence_flush(void *sync_obj)
-{
-	return 0;
+	if (exclusive)
+		reservation_object_add_excl_fence(resv, &fence->base);
+	else if (fence)
+		reservation_object_add_shared_fence(resv, &fence->base);
 }
 
 struct ttm_bo_driver nouveau_bo_driver = {
@@ -1489,11 +1622,6 @@ struct ttm_bo_driver nouveau_bo_driver = {
 	.move_notify = nouveau_bo_move_ntfy,
 	.move = nouveau_bo_move,
 	.verify_access = nouveau_bo_verify_access,
-	.sync_obj_signaled = nouveau_bo_fence_signalled,
-	.sync_obj_wait = nouveau_bo_fence_wait,
-	.sync_obj_flush = nouveau_bo_fence_flush,
-	.sync_obj_unref = nouveau_bo_fence_unref,
-	.sync_obj_ref = nouveau_bo_fence_ref,
 	.fault_reserve_notify = &nouveau_ttm_fault_reserve_notify,
 	.io_mem_reserve = &nouveau_ttm_io_mem_reserve,
 	.io_mem_free = &nouveau_ttm_io_mem_free,
diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.h b/drivers/gpu/drm/nouveau/nouveau_bo.h
index ff17c1f432fc..072222efeeb7 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bo.h
+++ b/drivers/gpu/drm/nouveau/nouveau_bo.h
@@ -1,6 +1,8 @@
 #ifndef __NOUVEAU_BO_H__
 #define __NOUVEAU_BO_H__
 
+#include <drm/drm_gem.h>
+
 struct nouveau_channel;
 struct nouveau_fence;
 struct nouveau_vma;
@@ -9,8 +11,9 @@ struct nouveau_bo {
 	struct ttm_buffer_object bo;
 	struct ttm_placement placement;
 	u32 valid_domains;
-	u32 placements[3];
-	u32 busy_placements[3];
+	struct ttm_place placements[3];
+	struct ttm_place busy_placements[3];
+	bool force_coherent;
 	struct ttm_bo_kmap_obj kmap;
 	struct list_head head;
 
@@ -68,8 +71,9 @@ extern struct ttm_bo_driver nouveau_bo_driver;
 void nouveau_bo_move_init(struct nouveau_drm *);
 int  nouveau_bo_new(struct drm_device *, int size, int align, u32 flags,
 		    u32 tile_mode, u32 tile_flags, struct sg_table *sg,
+		    struct reservation_object *robj,
 		    struct nouveau_bo **);
-int  nouveau_bo_pin(struct nouveau_bo *, u32 flags);
+int  nouveau_bo_pin(struct nouveau_bo *, u32 flags, bool contig);
 int  nouveau_bo_unpin(struct nouveau_bo *);
 int  nouveau_bo_map(struct nouveau_bo *);
 void nouveau_bo_unmap(struct nouveau_bo *);
@@ -78,9 +82,11 @@ u16  nouveau_bo_rd16(struct nouveau_bo *, unsigned index);
 void nouveau_bo_wr16(struct nouveau_bo *, unsigned index, u16 val);
 u32  nouveau_bo_rd32(struct nouveau_bo *, unsigned index);
 void nouveau_bo_wr32(struct nouveau_bo *, unsigned index, u32 val);
-void nouveau_bo_fence(struct nouveau_bo *, struct nouveau_fence *);
+void nouveau_bo_fence(struct nouveau_bo *, struct nouveau_fence *, bool exclusive);
 int  nouveau_bo_validate(struct nouveau_bo *, bool interruptible,
 			 bool no_wait_gpu);
+void nouveau_bo_sync_for_device(struct nouveau_bo *nvbo);
+void nouveau_bo_sync_for_cpu(struct nouveau_bo *nvbo);
 
 struct nouveau_vma *
 nouveau_bo_vma_find(struct nouveau_bo *, struct nouveau_vm *);
diff --git a/drivers/gpu/drm/nouveau/nouveau_chan.c b/drivers/gpu/drm/nouveau/nouveau_chan.c
index 3440fc999f2f..aff9099aae6c 100644
--- a/drivers/gpu/drm/nouveau/nouveau_chan.c
+++ b/drivers/gpu/drm/nouveau/nouveau_chan.c
@@ -36,7 +36,7 @@
 #include "nouveau_abi16.h"
 
 MODULE_PARM_DESC(vram_pushbuf, "Create DMA push buffers in VRAM");
-static int nouveau_vram_pushbuf;
+int nouveau_vram_pushbuf;
 module_param_named(vram_pushbuf, nouveau_vram_pushbuf, int, 0400);
 
 int
@@ -102,14 +102,14 @@ nouveau_channel_prep(struct nouveau_drm *drm, struct nvif_device *device,
 	chan->drm = drm;
 
 	/* allocate memory for dma push buffer */
-	target = TTM_PL_FLAG_TT;
+	target = TTM_PL_FLAG_TT | TTM_PL_FLAG_UNCACHED;
 	if (nouveau_vram_pushbuf)
 		target = TTM_PL_FLAG_VRAM;
 
-	ret = nouveau_bo_new(drm->dev, size, 0, target, 0, 0, NULL,
+	ret = nouveau_bo_new(drm->dev, size, 0, target, 0, 0, NULL, NULL,
 			    &chan->push.buffer);
 	if (ret == 0) {
-		ret = nouveau_bo_pin(chan->push.buffer, target);
+		ret = nouveau_bo_pin(chan->push.buffer, target, false);
 		if (ret == 0)
 			ret = nouveau_bo_map(chan->push.buffer);
 	}
@@ -285,7 +285,6 @@ nouveau_channel_init(struct nouveau_channel *chan, u32 vram, u32 gart)
 	struct nouveau_software_chan *swch;
 	struct nv_dma_v0 args = {};
 	int ret, i;
-	bool save;
 
 	nvif_object_map(chan->object);
 
@@ -387,11 +386,7 @@ nouveau_channel_init(struct nouveau_channel *chan, u32 vram, u32 gart)
 	}
 
 	/* initialise synchronisation */
-	save = cli->base.super;
-	cli->base.super = true; /* hack until fencenv50 fixed */
-	ret = nouveau_fence(chan->drm)->context_new(chan);
-	cli->base.super = save;
-	return ret;
+	return nouveau_fence(chan->drm)->context_new(chan);
 }
 
 int
@@ -400,15 +395,20 @@ nouveau_channel_new(struct nouveau_drm *drm, struct nvif_device *device,
 		    struct nouveau_channel **pchan)
 {
 	struct nouveau_cli *cli = (void *)nvif_client(&device->base);
+	bool super;
 	int ret;
 
+	/* hack until fencenv50 is fixed, and agp access relaxed */
+	super = cli->base.super;
+	cli->base.super = true;
+
 	ret = nouveau_channel_ind(drm, device, handle, arg0, pchan);
 	if (ret) {
 		NV_PRINTK(debug, cli, "ib channel create, %d\n", ret);
 		ret = nouveau_channel_dma(drm, device, handle, pchan);
 		if (ret) {
 			NV_PRINTK(debug, cli, "dma channel create, %d\n", ret);
-			return ret;
+			goto done;
 		}
 	}
 
@@ -416,8 +416,9 @@ nouveau_channel_new(struct nouveau_drm *drm, struct nvif_device *device,
 	if (ret) {
 		NV_PRINTK(error, cli, "channel failed to initialise, %d\n", ret);
 		nouveau_channel_del(pchan);
-		return ret;
 	}
 
-	return 0;
+done:
+	cli->base.super = super;
+	return ret;
 }
diff --git a/drivers/gpu/drm/nouveau/nouveau_chan.h b/drivers/gpu/drm/nouveau/nouveau_chan.h
index 20163709d608..8309c24ee698 100644
--- a/drivers/gpu/drm/nouveau/nouveau_chan.h
+++ b/drivers/gpu/drm/nouveau/nouveau_chan.h
@@ -47,4 +47,6 @@ int  nouveau_channel_new(struct nouveau_drm *, struct nvif_device *,
 void nouveau_channel_del(struct nouveau_channel **);
 int  nouveau_channel_idle(struct nouveau_channel *);
 
+extern int nouveau_vram_pushbuf;
+
 #endif
diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.c b/drivers/gpu/drm/nouveau/nouveau_connector.c
index 1ec44c83e919..c8ac9482cf2e 100644
--- a/drivers/gpu/drm/nouveau/nouveau_connector.c
+++ b/drivers/gpu/drm/nouveau/nouveau_connector.c
@@ -45,15 +45,15 @@
 #include <nvif/event.h>
 
 MODULE_PARM_DESC(tv_disable, "Disable TV-out detection");
-static int nouveau_tv_disable = 0;
+int nouveau_tv_disable = 0;
 module_param_named(tv_disable, nouveau_tv_disable, int, 0400);
 
 MODULE_PARM_DESC(ignorelid, "Ignore ACPI lid status");
-static int nouveau_ignorelid = 0;
+int nouveau_ignorelid = 0;
 module_param_named(ignorelid, nouveau_ignorelid, int, 0400);
 
 MODULE_PARM_DESC(duallink, "Allow dual-link TMDS (default: enabled)");
-static int nouveau_duallink = 1;
+int nouveau_duallink = 1;
 module_param_named(duallink, nouveau_duallink, int, 0400);
 
 struct nouveau_encoder *
diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.h b/drivers/gpu/drm/nouveau/nouveau_connector.h
index 68029d041dd2..629a380c7085 100644
--- a/drivers/gpu/drm/nouveau/nouveau_connector.h
+++ b/drivers/gpu/drm/nouveau/nouveau_connector.h
@@ -105,4 +105,8 @@ nouveau_crtc_connector_get(struct nouveau_crtc *nv_crtc)
 struct drm_connector *
 nouveau_connector_create(struct drm_device *, int index);
 
+extern int nouveau_tv_disable;
+extern int nouveau_ignorelid;
+extern int nouveau_duallink;
+
 #endif /* __NOUVEAU_CONNECTOR_H__ */
diff --git a/drivers/gpu/drm/nouveau/nouveau_display.c b/drivers/gpu/drm/nouveau/nouveau_display.c
index 4a21b2b06ce2..f8042433752b 100644
--- a/drivers/gpu/drm/nouveau/nouveau_display.c
+++ b/drivers/gpu/drm/nouveau/nouveau_display.c
@@ -126,7 +126,7 @@ nouveau_display_scanoutpos_head(struct drm_crtc *crtc, int *vpos, int *hpos,
 	if (etime) *etime = ns_to_ktime(args.scan.time[1]);
 
 	if (*vpos < 0)
-		ret |= DRM_SCANOUTPOS_INVBL;
+		ret |= DRM_SCANOUTPOS_IN_VBLANK;
 	return ret;
 }
 
@@ -479,6 +479,7 @@ nouveau_display_create(struct drm_device *dev)
 
 	if (nouveau_modeset != 2 && drm->vbios.dcb.entries) {
 		static const u16 oclass[] = {
+			GM204_DISP,
 			GM107_DISP,
 			GK110_DISP,
 			GK104_DISP,
@@ -568,9 +569,10 @@ nouveau_display_suspend(struct drm_device *dev, bool runtime)
 
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
 		struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
-
-		nouveau_bo_unmap(nv_crtc->cursor.nvbo);
-		nouveau_bo_unpin(nv_crtc->cursor.nvbo);
+		if (nv_crtc->cursor.nvbo) {
+			nouveau_bo_unmap(nv_crtc->cursor.nvbo);
+			nouveau_bo_unpin(nv_crtc->cursor.nvbo);
+		}
 	}
 
 	return 0;
@@ -591,15 +593,17 @@ nouveau_display_resume(struct drm_device *dev, bool runtime)
 		if (!nouveau_fb || !nouveau_fb->nvbo)
 			continue;
 
-		ret = nouveau_bo_pin(nouveau_fb->nvbo, TTM_PL_FLAG_VRAM);
+		ret = nouveau_bo_pin(nouveau_fb->nvbo, TTM_PL_FLAG_VRAM, true);
 		if (ret)
 			NV_ERROR(drm, "Could not pin framebuffer\n");
 	}
 
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
 		struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
+		if (!nv_crtc->cursor.nvbo)
+			continue;
 
-		ret = nouveau_bo_pin(nv_crtc->cursor.nvbo, TTM_PL_FLAG_VRAM);
+		ret = nouveau_bo_pin(nv_crtc->cursor.nvbo, TTM_PL_FLAG_VRAM, true);
 		if (!ret)
 			ret = nouveau_bo_map(nv_crtc->cursor.nvbo);
 		if (ret)
@@ -630,9 +634,10 @@ nouveau_display_resume(struct drm_device *dev, bool runtime)
 
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
 		struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
-		u32 offset = nv_crtc->cursor.nvbo->bo.offset;
 
-		nv_crtc->cursor.set_offset(nv_crtc, offset);
+		if (!nv_crtc->cursor.nvbo)
+			continue;
+		nv_crtc->cursor.set_offset(nv_crtc, nv_crtc->cursor.nvbo->bo.offset);
 		nv_crtc->cursor.set_pos(nv_crtc, nv_crtc->cursor_saved_x,
 						 nv_crtc->cursor_saved_y);
 	}
@@ -657,7 +662,7 @@ nouveau_page_flip_emit(struct nouveau_channel *chan,
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 
 	/* Synchronize with the old framebuffer */
-	ret = nouveau_fence_sync(old_bo->bo.sync_obj, chan);
+	ret = nouveau_fence_sync(old_bo, chan, false, false);
 	if (ret)
 		goto fail;
 
@@ -710,25 +715,30 @@ nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 		return -ENOMEM;
 
 	if (new_bo != old_bo) {
-		ret = nouveau_bo_pin(new_bo, TTM_PL_FLAG_VRAM);
+		ret = nouveau_bo_pin(new_bo, TTM_PL_FLAG_VRAM, true);
 		if (ret)
 			goto fail_free;
 	}
 
 	mutex_lock(&cli->mutex);
-
-	/* synchronise rendering channel with the kernel's channel */
-	spin_lock(&new_bo->bo.bdev->fence_lock);
-	fence = nouveau_fence_ref(new_bo->bo.sync_obj);
-	spin_unlock(&new_bo->bo.bdev->fence_lock);
-	ret = nouveau_fence_sync(fence, chan);
-	nouveau_fence_unref(&fence);
+	ret = ttm_bo_reserve(&new_bo->bo, true, false, false, NULL);
 	if (ret)
 		goto fail_unpin;
 
-	ret = ttm_bo_reserve(&old_bo->bo, true, false, false, NULL);
-	if (ret)
+	/* synchronise rendering channel with the kernel's channel */
+	ret = nouveau_fence_sync(new_bo, chan, false, true);
+	if (ret) {
+		ttm_bo_unreserve(&new_bo->bo);
 		goto fail_unpin;
+	}
+
+	if (new_bo != old_bo) {
+		ttm_bo_unreserve(&new_bo->bo);
+
+		ret = ttm_bo_reserve(&old_bo->bo, true, false, false, NULL);
+		if (ret)
+			goto fail_unpin;
+	}
 
 	/* Initialize a page flip struct */
 	*s = (struct nouveau_page_flip_state)
@@ -774,7 +784,7 @@ nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 	/* Update the crtc struct and cleanup */
 	crtc->primary->fb = fb;
 
-	nouveau_bo_fence(old_bo, fence);
+	nouveau_bo_fence(old_bo, fence, false);
 	ttm_bo_unreserve(&old_bo->bo);
 	if (old_bo != new_bo)
 		nouveau_bo_unpin(old_bo);
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c
index 3ed32dd90303..65910e3aed0c 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.c
@@ -51,6 +51,7 @@
 #include "nouveau_fence.h"
 #include "nouveau_debugfs.h"
 #include "nouveau_usif.h"
+#include "nouveau_connector.h"
 
 MODULE_PARM_DESC(config, "option string to pass to driver core");
 static char *nouveau_config;
@@ -73,7 +74,9 @@ MODULE_PARM_DESC(runpm, "disable (0), force enable (1), optimus only default (-1
 int nouveau_runtime_pm = -1;
 module_param_named(runpm, nouveau_runtime_pm, int, 0400);
 
-static struct drm_driver driver;
+static struct drm_driver driver_stub;
+static struct drm_driver driver_pci;
+static struct drm_driver driver_platform;
 
 static u64
 nouveau_pci_name(struct pci_dev *pdev)
@@ -322,7 +325,7 @@ static int nouveau_drm_probe(struct pci_dev *pdev,
 
 	pci_set_master(pdev);
 
-	ret = drm_get_pci_dev(pdev, pent, &driver);
+	ret = drm_get_pci_dev(pdev, pent, &driver_pci);
 	if (ret) {
 		nouveau_object_ref(NULL, (struct nouveau_object **)&device);
 		return ret;
@@ -610,27 +613,6 @@ fail_display:
 	return ret;
 }
 
-int nouveau_pmops_suspend(struct device *dev)
-{
-	struct pci_dev *pdev = to_pci_dev(dev);
-	struct drm_device *drm_dev = pci_get_drvdata(pdev);
-	int ret;
-
-	if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF ||
-	    drm_dev->switch_power_state == DRM_SWITCH_POWER_DYNAMIC_OFF)
-		return 0;
-
-	ret = nouveau_do_suspend(drm_dev, false);
-	if (ret)
-		return ret;
-
-	pci_save_state(pdev);
-	pci_disable_device(pdev);
-	pci_ignore_hotplug(pdev);
-	pci_set_power_state(pdev, PCI_D3hot);
-	return 0;
-}
-
 static int
 nouveau_do_resume(struct drm_device *dev, bool runtime)
 {
@@ -665,7 +647,29 @@ nouveau_do_resume(struct drm_device *dev, bool runtime)
 	return 0;
 }
 
-int nouveau_pmops_resume(struct device *dev)
+int
+nouveau_pmops_suspend(struct device *dev)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct drm_device *drm_dev = pci_get_drvdata(pdev);
+	int ret;
+
+	if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF ||
+	    drm_dev->switch_power_state == DRM_SWITCH_POWER_DYNAMIC_OFF)
+		return 0;
+
+	ret = nouveau_do_suspend(drm_dev, false);
+	if (ret)
+		return ret;
+
+	pci_save_state(pdev);
+	pci_disable_device(pdev);
+	pci_set_power_state(pdev, PCI_D3hot);
+	return 0;
+}
+
+int
+nouveau_pmops_resume(struct device *dev)
 {
 	struct pci_dev *pdev = to_pci_dev(dev);
 	struct drm_device *drm_dev = pci_get_drvdata(pdev);
@@ -685,20 +689,122 @@ int nouveau_pmops_resume(struct device *dev)
 	return nouveau_do_resume(drm_dev, false);
 }
 
-static int nouveau_pmops_freeze(struct device *dev)
+static int
+nouveau_pmops_freeze(struct device *dev)
 {
 	struct pci_dev *pdev = to_pci_dev(dev);
 	struct drm_device *drm_dev = pci_get_drvdata(pdev);
 	return nouveau_do_suspend(drm_dev, false);
 }
 
-static int nouveau_pmops_thaw(struct device *dev)
+static int
+nouveau_pmops_thaw(struct device *dev)
 {
 	struct pci_dev *pdev = to_pci_dev(dev);
 	struct drm_device *drm_dev = pci_get_drvdata(pdev);
 	return nouveau_do_resume(drm_dev, false);
 }
 
+static int
+nouveau_pmops_runtime_suspend(struct device *dev)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct drm_device *drm_dev = pci_get_drvdata(pdev);
+	int ret;
+
+	if (nouveau_runtime_pm == 0) {
+		pm_runtime_forbid(dev);
+		return -EBUSY;
+	}
+
+	/* are we optimus enabled? */
+	if (nouveau_runtime_pm == -1 && !nouveau_is_optimus() && !nouveau_is_v1_dsm()) {
+		DRM_DEBUG_DRIVER("failing to power off - not optimus\n");
+		pm_runtime_forbid(dev);
+		return -EBUSY;
+	}
+
+	nv_debug_level(SILENT);
+	drm_kms_helper_poll_disable(drm_dev);
+	vga_switcheroo_set_dynamic_switch(pdev, VGA_SWITCHEROO_OFF);
+	nouveau_switcheroo_optimus_dsm();
+	ret = nouveau_do_suspend(drm_dev, true);
+	pci_save_state(pdev);
+	pci_disable_device(pdev);
+	pci_ignore_hotplug(pdev);
+	pci_set_power_state(pdev, PCI_D3cold);
+	drm_dev->switch_power_state = DRM_SWITCH_POWER_DYNAMIC_OFF;
+	return ret;
+}
+
+static int
+nouveau_pmops_runtime_resume(struct device *dev)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct drm_device *drm_dev = pci_get_drvdata(pdev);
+	struct nvif_device *device = &nouveau_drm(drm_dev)->device;
+	int ret;
+
+	if (nouveau_runtime_pm == 0)
+		return -EINVAL;
+
+	pci_set_power_state(pdev, PCI_D0);
+	pci_restore_state(pdev);
+	ret = pci_enable_device(pdev);
+	if (ret)
+		return ret;
+	pci_set_master(pdev);
+
+	ret = nouveau_do_resume(drm_dev, true);
+	drm_kms_helper_poll_enable(drm_dev);
+	/* do magic */
+	nvif_mask(device, 0x88488, (1 << 25), (1 << 25));
+	vga_switcheroo_set_dynamic_switch(pdev, VGA_SWITCHEROO_ON);
+	drm_dev->switch_power_state = DRM_SWITCH_POWER_ON;
+	nv_debug_level(NORMAL);
+	return ret;
+}
+
+static int
+nouveau_pmops_runtime_idle(struct device *dev)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct drm_device *drm_dev = pci_get_drvdata(pdev);
+	struct nouveau_drm *drm = nouveau_drm(drm_dev);
+	struct drm_crtc *crtc;
+
+	if (nouveau_runtime_pm == 0) {
+		pm_runtime_forbid(dev);
+		return -EBUSY;
+	}
+
+	/* are we optimus enabled? */
+	if (nouveau_runtime_pm == -1 && !nouveau_is_optimus() && !nouveau_is_v1_dsm()) {
+		DRM_DEBUG_DRIVER("failing to power off - not optimus\n");
+		pm_runtime_forbid(dev);
+		return -EBUSY;
+	}
+
+	/* if we have a hdmi audio device - make sure it has a driver loaded */
+	if (drm->hdmi_device) {
+		if (!drm->hdmi_device->driver) {
+			DRM_DEBUG_DRIVER("failing to power off - no HDMI audio driver loaded\n");
+			pm_runtime_mark_last_busy(dev);
+			return -EBUSY;
+		}
+	}
+
+	list_for_each_entry(crtc, &drm->dev->mode_config.crtc_list, head) {
+		if (crtc->enabled) {
+			DRM_DEBUG_DRIVER("failing to power off - crtc active\n");
+			return -EBUSY;
+		}
+	}
+	pm_runtime_mark_last_busy(dev);
+	pm_runtime_autosuspend(dev);
+	/* we don't want the main rpm_idle to call suspend - we want to autosuspend */
+	return 1;
+}
 
 static int
 nouveau_drm_open(struct drm_device *dev, struct drm_file *fpriv)
@@ -831,7 +937,7 @@ nouveau_driver_fops = {
 };
 
 static struct drm_driver
-driver = {
+driver_stub = {
 	.driver_features =
 		DRIVER_USE_AGP |
 		DRIVER_GEM | DRIVER_MODESET | DRIVER_PRIME | DRIVER_RENDER,
@@ -905,101 +1011,21 @@ nouveau_drm_pci_table[] = {
 	{}
 };
 
-static int nouveau_pmops_runtime_suspend(struct device *dev)
+static void nouveau_display_options(void)
 {
-	struct pci_dev *pdev = to_pci_dev(dev);
-	struct drm_device *drm_dev = pci_get_drvdata(pdev);
-	int ret;
-
-	if (nouveau_runtime_pm == 0) {
-		pm_runtime_forbid(dev);
-		return -EBUSY;
-	}
-
-	/* are we optimus enabled? */
-	if (nouveau_runtime_pm == -1 && !nouveau_is_optimus() && !nouveau_is_v1_dsm()) {
-		DRM_DEBUG_DRIVER("failing to power off - not optimus\n");
-		pm_runtime_forbid(dev);
-		return -EBUSY;
-	}
-
-	nv_debug_level(SILENT);
-	drm_kms_helper_poll_disable(drm_dev);
-	vga_switcheroo_set_dynamic_switch(pdev, VGA_SWITCHEROO_OFF);
-	nouveau_switcheroo_optimus_dsm();
-	ret = nouveau_do_suspend(drm_dev, true);
-	pci_save_state(pdev);
-	pci_disable_device(pdev);
-	pci_set_power_state(pdev, PCI_D3cold);
-	drm_dev->switch_power_state = DRM_SWITCH_POWER_DYNAMIC_OFF;
-	return ret;
-}
-
-static int nouveau_pmops_runtime_resume(struct device *dev)
-{
-	struct pci_dev *pdev = to_pci_dev(dev);
-	struct drm_device *drm_dev = pci_get_drvdata(pdev);
-	struct nvif_device *device = &nouveau_drm(drm_dev)->device;
-	int ret;
-
-	if (nouveau_runtime_pm == 0)
-		return -EINVAL;
-
-	pci_set_power_state(pdev, PCI_D0);
-	pci_restore_state(pdev);
-	ret = pci_enable_device(pdev);
-	if (ret)
-		return ret;
-	pci_set_master(pdev);
-
-	ret = nouveau_do_resume(drm_dev, true);
-	drm_kms_helper_poll_enable(drm_dev);
-	/* do magic */
-	nvif_mask(device, 0x88488, (1 << 25), (1 << 25));
-	vga_switcheroo_set_dynamic_switch(pdev, VGA_SWITCHEROO_ON);
-	drm_dev->switch_power_state = DRM_SWITCH_POWER_ON;
-	nv_debug_level(NORMAL);
-	return ret;
-}
-
-static int nouveau_pmops_runtime_idle(struct device *dev)
-{
-	struct pci_dev *pdev = to_pci_dev(dev);
-	struct drm_device *drm_dev = pci_get_drvdata(pdev);
-	struct nouveau_drm *drm = nouveau_drm(drm_dev);
-	struct drm_crtc *crtc;
-
-	if (nouveau_runtime_pm == 0) {
-		pm_runtime_forbid(dev);
-		return -EBUSY;
-	}
-
-	/* are we optimus enabled? */
-	if (nouveau_runtime_pm == -1 && !nouveau_is_optimus() && !nouveau_is_v1_dsm()) {
-		DRM_DEBUG_DRIVER("failing to power off - not optimus\n");
-		pm_runtime_forbid(dev);
-		return -EBUSY;
-	}
-
-	/* if we have a hdmi audio device - make sure it has a driver loaded */
-	if (drm->hdmi_device) {
-		if (!drm->hdmi_device->driver) {
-			DRM_DEBUG_DRIVER("failing to power off - no HDMI audio driver loaded\n");
-			pm_runtime_mark_last_busy(dev);
-			return -EBUSY;
-		}
-	}
-
-	list_for_each_entry(crtc, &drm->dev->mode_config.crtc_list, head) {
-		if (crtc->enabled) {
-			DRM_DEBUG_DRIVER("failing to power off - crtc active\n");
-			return -EBUSY;
-		}
-	}
-	pm_runtime_mark_last_busy(dev);
-	pm_runtime_autosuspend(dev);
-	/* we don't want the main rpm_idle to call suspend - we want to autosuspend */
-	return 1;
+	DRM_DEBUG_DRIVER("Loading Nouveau with parameters:\n");
+
+	DRM_DEBUG_DRIVER("... tv_disable   : %d\n", nouveau_tv_disable);
+	DRM_DEBUG_DRIVER("... ignorelid    : %d\n", nouveau_ignorelid);
+	DRM_DEBUG_DRIVER("... duallink     : %d\n", nouveau_duallink);
+	DRM_DEBUG_DRIVER("... nofbaccel    : %d\n", nouveau_nofbaccel);
+	DRM_DEBUG_DRIVER("... config       : %s\n", nouveau_config);
+	DRM_DEBUG_DRIVER("... debug        : %s\n", nouveau_debug);
+	DRM_DEBUG_DRIVER("... noaccel      : %d\n", nouveau_noaccel);
+	DRM_DEBUG_DRIVER("... modeset      : %d\n", nouveau_modeset);
+	DRM_DEBUG_DRIVER("... runpm        : %d\n", nouveau_runtime_pm);
+	DRM_DEBUG_DRIVER("... vram_pushbuf : %d\n", nouveau_vram_pushbuf);
+	DRM_DEBUG_DRIVER("... pstate       : %d\n", nouveau_pstate);
 }
 
 static const struct dev_pm_ops nouveau_pm_ops = {
@@ -1037,7 +1063,7 @@ nouveau_platform_device_create_(struct platform_device *pdev, int size,
 	if (err)
 		return ERR_PTR(err);
 
-	drm = drm_dev_alloc(&driver, &pdev->dev);
+	drm = drm_dev_alloc(&driver_platform, &pdev->dev);
 	if (!drm) {
 		err = -ENOMEM;
 		goto err_free;
@@ -1062,6 +1088,13 @@ EXPORT_SYMBOL(nouveau_platform_device_create_);
 static int __init
 nouveau_drm_init(void)
 {
+	driver_pci = driver_stub;
+	driver_pci.set_busid = drm_pci_set_busid;
+	driver_platform = driver_stub;
+	driver_platform.set_busid = drm_platform_set_busid;
+
+	nouveau_display_options();
+
 	if (nouveau_modeset == -1) {
 #ifdef CONFIG_VGA_CONSOLE
 		if (vgacon_text_force())
@@ -1073,7 +1106,7 @@ nouveau_drm_init(void)
 		return 0;
 
 	nouveau_register_dsm_handler();
-	return drm_pci_init(&driver, &nouveau_drm_pci_driver);
+	return drm_pci_init(&driver_pci, &nouveau_drm_pci_driver);
 }
 
 static void __exit
@@ -1082,7 +1115,7 @@ nouveau_drm_exit(void)
 	if (!nouveau_modeset)
 		return;
 
-	drm_pci_exit(&driver, &nouveau_drm_pci_driver);
+	drm_pci_exit(&driver_pci, &nouveau_drm_pci_driver);
 	nouveau_unregister_dsm_handler();
 }
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.h b/drivers/gpu/drm/nouveau/nouveau_drm.h
index b02b02452c85..8ae36f265fb8 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.h
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.h
@@ -10,7 +10,7 @@
 
 #define DRIVER_MAJOR		1
 #define DRIVER_MINOR		2
-#define DRIVER_PATCHLEVEL	0
+#define DRIVER_PATCHLEVEL	1
 
 /*
  * 1.1.1:
@@ -26,6 +26,8 @@
  * 1.2.0:
  * 	- object api exposed to userspace
  * 	- fermi,kepler,maxwell zbc
+ * 1.2.1:
+ *      - allow concurrent access to bo's mapped read/write.
  */
 
 #include <nvif/client.h>
diff --git a/drivers/gpu/drm/nouveau/nouveau_fbcon.c b/drivers/gpu/drm/nouveau/nouveau_fbcon.c
index 49fe6075cc7c..3ed12a8cfc91 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fbcon.c
+++ b/drivers/gpu/drm/nouveau/nouveau_fbcon.c
@@ -52,7 +52,7 @@
 #include "nouveau_crtc.h"
 
 MODULE_PARM_DESC(nofbaccel, "Disable fbcon acceleration");
-static int nouveau_nofbaccel = 0;
+int nouveau_nofbaccel = 0;
 module_param_named(nofbaccel, nouveau_nofbaccel, int, 0400);
 
 static void
@@ -308,7 +308,8 @@ static int
 nouveau_fbcon_create(struct drm_fb_helper *helper,
 		     struct drm_fb_helper_surface_size *sizes)
 {
-	struct nouveau_fbdev *fbcon = (struct nouveau_fbdev *)helper;
+	struct nouveau_fbdev *fbcon =
+		container_of(helper, struct nouveau_fbdev, helper);
 	struct drm_device *dev = fbcon->dev;
 	struct nouveau_drm *drm = nouveau_drm(dev);
 	struct nvif_device *device = &drm->device;
@@ -340,7 +341,7 @@ nouveau_fbcon_create(struct drm_fb_helper *helper,
 		goto out;
 	}
 
-	ret = nouveau_bo_pin(nvbo, TTM_PL_FLAG_VRAM);
+	ret = nouveau_bo_pin(nvbo, TTM_PL_FLAG_VRAM, false);
 	if (ret) {
 		NV_ERROR(drm, "failed to pin fb: %d\n", ret);
 		goto out_unref;
@@ -497,6 +498,23 @@ nouveau_fbcon_set_suspend_work(struct work_struct *work)
 	console_unlock();
 }
 
+void
+nouveau_fbcon_set_suspend(struct drm_device *dev, int state)
+{
+	struct nouveau_drm *drm = nouveau_drm(dev);
+	if (drm->fbcon) {
+		if (state == FBINFO_STATE_RUNNING) {
+			schedule_work(&drm->fbcon->work);
+			return;
+		}
+		flush_work(&drm->fbcon->work);
+		console_lock();
+		fb_set_suspend(drm->fbcon->helper.fbdev, state);
+		nouveau_fbcon_accel_save_disable(dev);
+		console_unlock();
+	}
+}
+
 int
 nouveau_fbcon_init(struct drm_device *dev)
 {
@@ -556,20 +574,3 @@ nouveau_fbcon_fini(struct drm_device *dev)
 	kfree(drm->fbcon);
 	drm->fbcon = NULL;
 }
-
-void
-nouveau_fbcon_set_suspend(struct drm_device *dev, int state)
-{
-	struct nouveau_drm *drm = nouveau_drm(dev);
-	if (drm->fbcon) {
-		if (state == FBINFO_STATE_RUNNING) {
-			schedule_work(&drm->fbcon->work);
-			return;
-		}
-		flush_work(&drm->fbcon->work);
-		console_lock();
-		fb_set_suspend(drm->fbcon->helper.fbdev, state);
-		nouveau_fbcon_accel_save_disable(dev);
-		console_unlock();
-	}
-}
diff --git a/drivers/gpu/drm/nouveau/nouveau_fbcon.h b/drivers/gpu/drm/nouveau/nouveau_fbcon.h
index 0b465c7d3907..6208e70e4a1c 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fbcon.h
+++ b/drivers/gpu/drm/nouveau/nouveau_fbcon.h
@@ -73,5 +73,8 @@ void nouveau_fbcon_accel_save_disable(struct drm_device *dev);
 void nouveau_fbcon_accel_restore(struct drm_device *dev);
 
 void nouveau_fbcon_output_poll_changed(struct drm_device *dev);
+
+extern int nouveau_nofbaccel;
+
 #endif /* __NV50_FBCON_H__ */
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_fence.c b/drivers/gpu/drm/nouveau/nouveau_fence.c
index 0a93114158cd..f32a434724e3 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fence.c
+++ b/drivers/gpu/drm/nouveau/nouveau_fence.c
@@ -28,6 +28,7 @@
 
 #include <linux/ktime.h>
 #include <linux/hrtimer.h>
+#include <trace/events/fence.h>
 
 #include <nvif/notify.h>
 #include <nvif/event.h>
@@ -36,123 +37,253 @@
 #include "nouveau_dma.h"
 #include "nouveau_fence.h"
 
-struct fence_work {
-	struct work_struct base;
-	struct list_head head;
-	void (*func)(void *);
-	void *data;
-};
+static const struct fence_ops nouveau_fence_ops_uevent;
+static const struct fence_ops nouveau_fence_ops_legacy;
 
-static void
+static inline struct nouveau_fence *
+from_fence(struct fence *fence)
+{
+	return container_of(fence, struct nouveau_fence, base);
+}
+
+static inline struct nouveau_fence_chan *
+nouveau_fctx(struct nouveau_fence *fence)
+{
+	return container_of(fence->base.lock, struct nouveau_fence_chan, lock);
+}
+
+static int
 nouveau_fence_signal(struct nouveau_fence *fence)
 {
-	struct fence_work *work, *temp;
+	int drop = 0;
+
+	fence_signal_locked(&fence->base);
+	list_del(&fence->head);
+	rcu_assign_pointer(fence->channel, NULL);
+
+	if (test_bit(FENCE_FLAG_USER_BITS, &fence->base.flags)) {
+		struct nouveau_fence_chan *fctx = nouveau_fctx(fence);
 
-	list_for_each_entry_safe(work, temp, &fence->work, head) {
-		schedule_work(&work->base);
-		list_del(&work->head);
+		if (!--fctx->notify_ref)
+			drop = 1;
 	}
 
-	fence->channel = NULL;
-	list_del(&fence->head);
+	fence_put(&fence->base);
+	return drop;
+}
+
+static struct nouveau_fence *
+nouveau_local_fence(struct fence *fence, struct nouveau_drm *drm) {
+	struct nouveau_fence_priv *priv = (void*)drm->fence;
+
+	if (fence->ops != &nouveau_fence_ops_legacy &&
+	    fence->ops != &nouveau_fence_ops_uevent)
+		return NULL;
+
+	if (fence->context < priv->context_base ||
+	    fence->context >= priv->context_base + priv->contexts)
+		return NULL;
+
+	return from_fence(fence);
 }
 
 void
 nouveau_fence_context_del(struct nouveau_fence_chan *fctx)
 {
-	struct nouveau_fence *fence, *fnext;
-	spin_lock(&fctx->lock);
-	list_for_each_entry_safe(fence, fnext, &fctx->pending, head) {
-		nouveau_fence_signal(fence);
+	struct nouveau_fence *fence;
+
+	spin_lock_irq(&fctx->lock);
+	while (!list_empty(&fctx->pending)) {
+		fence = list_entry(fctx->pending.next, typeof(*fence), head);
+
+		if (nouveau_fence_signal(fence))
+			nvif_notify_put(&fctx->notify);
+	}
+	spin_unlock_irq(&fctx->lock);
+
+	nvif_notify_fini(&fctx->notify);
+	fctx->dead = 1;
+
+	/*
+	 * Ensure that all accesses to fence->channel complete before freeing
+	 * the channel.
+	 */
+	synchronize_rcu();
+}
+
+static void
+nouveau_fence_context_put(struct kref *fence_ref)
+{
+	kfree(container_of(fence_ref, struct nouveau_fence_chan, fence_ref));
+}
+
+void
+nouveau_fence_context_free(struct nouveau_fence_chan *fctx)
+{
+	kref_put(&fctx->fence_ref, nouveau_fence_context_put);
+}
+
+static int
+nouveau_fence_update(struct nouveau_channel *chan, struct nouveau_fence_chan *fctx)
+{
+	struct nouveau_fence *fence;
+	int drop = 0;
+	u32 seq = fctx->read(chan);
+
+	while (!list_empty(&fctx->pending)) {
+		fence = list_entry(fctx->pending.next, typeof(*fence), head);
+
+		if ((int)(seq - fence->base.seqno) < 0)
+			break;
+
+		drop |= nouveau_fence_signal(fence);
+	}
+
+	return drop;
+}
+
+static int
+nouveau_fence_wait_uevent_handler(struct nvif_notify *notify)
+{
+	struct nouveau_fence_chan *fctx =
+		container_of(notify, typeof(*fctx), notify);
+	unsigned long flags;
+	int ret = NVIF_NOTIFY_KEEP;
+
+	spin_lock_irqsave(&fctx->lock, flags);
+	if (!list_empty(&fctx->pending)) {
+		struct nouveau_fence *fence;
+		struct nouveau_channel *chan;
+
+		fence = list_entry(fctx->pending.next, typeof(*fence), head);
+		chan = rcu_dereference_protected(fence->channel, lockdep_is_held(&fctx->lock));
+		if (nouveau_fence_update(fence->channel, fctx))
+			ret = NVIF_NOTIFY_DROP;
 	}
-	spin_unlock(&fctx->lock);
+	spin_unlock_irqrestore(&fctx->lock, flags);
+
+	return ret;
 }
 
 void
-nouveau_fence_context_new(struct nouveau_fence_chan *fctx)
+nouveau_fence_context_new(struct nouveau_channel *chan, struct nouveau_fence_chan *fctx)
 {
+	struct nouveau_fence_priv *priv = (void*)chan->drm->fence;
+	struct nouveau_cli *cli = (void *)nvif_client(chan->object);
+	int ret;
+
 	INIT_LIST_HEAD(&fctx->flip);
 	INIT_LIST_HEAD(&fctx->pending);
 	spin_lock_init(&fctx->lock);
+	fctx->context = priv->context_base + chan->chid;
+
+	if (chan == chan->drm->cechan)
+		strcpy(fctx->name, "copy engine channel");
+	else if (chan == chan->drm->channel)
+		strcpy(fctx->name, "generic kernel channel");
+	else
+		strcpy(fctx->name, nvkm_client(&cli->base)->name);
+
+	kref_init(&fctx->fence_ref);
+	if (!priv->uevent)
+		return;
+
+	ret = nvif_notify_init(chan->object, NULL,
+			 nouveau_fence_wait_uevent_handler, false,
+			 G82_CHANNEL_DMA_V0_NTFY_UEVENT,
+			 &(struct nvif_notify_uevent_req) { },
+			 sizeof(struct nvif_notify_uevent_req),
+			 sizeof(struct nvif_notify_uevent_rep),
+			 &fctx->notify);
+
+	WARN_ON(ret);
 }
 
+struct nouveau_fence_work {
+	struct work_struct work;
+	struct fence_cb cb;
+	void (*func)(void *);
+	void *data;
+};
+
 static void
 nouveau_fence_work_handler(struct work_struct *kwork)
 {
-	struct fence_work *work = container_of(kwork, typeof(*work), base);
+	struct nouveau_fence_work *work = container_of(kwork, typeof(*work), work);
 	work->func(work->data);
 	kfree(work);
 }
 
+static void nouveau_fence_work_cb(struct fence *fence, struct fence_cb *cb)
+{
+	struct nouveau_fence_work *work = container_of(cb, typeof(*work), cb);
+
+	schedule_work(&work->work);
+}
+
 void
-nouveau_fence_work(struct nouveau_fence *fence,
+nouveau_fence_work(struct fence *fence,
 		   void (*func)(void *), void *data)
 {
-	struct nouveau_channel *chan = fence->channel;
-	struct nouveau_fence_chan *fctx;
-	struct fence_work *work = NULL;
+	struct nouveau_fence_work *work;
 
-	if (nouveau_fence_done(fence)) {
-		func(data);
-		return;
-	}
+	if (fence_is_signaled(fence))
+		goto err;
 
-	fctx = chan->fence;
 	work = kmalloc(sizeof(*work), GFP_KERNEL);
 	if (!work) {
-		WARN_ON(nouveau_fence_wait(fence, false, false));
-		func(data);
-		return;
+		/*
+		 * this might not be a nouveau fence any more,
+		 * so force a lazy wait here
+		 */
+		WARN_ON(nouveau_fence_wait((struct nouveau_fence *)fence,
+					   true, false));
+		goto err;
 	}
 
-	spin_lock(&fctx->lock);
-	if (!fence->channel) {
-		spin_unlock(&fctx->lock);
-		kfree(work);
-		func(data);
-		return;
-	}
-
-	INIT_WORK(&work->base, nouveau_fence_work_handler);
+	INIT_WORK(&work->work, nouveau_fence_work_handler);
 	work->func = func;
 	work->data = data;
-	list_add(&work->head, &fence->work);
-	spin_unlock(&fctx->lock);
-}
-
-static void
-nouveau_fence_update(struct nouveau_channel *chan)
-{
-	struct nouveau_fence_chan *fctx = chan->fence;
-	struct nouveau_fence *fence, *fnext;
 
-	spin_lock(&fctx->lock);
-	list_for_each_entry_safe(fence, fnext, &fctx->pending, head) {
-		if (fctx->read(chan) < fence->sequence)
-			break;
+	if (fence_add_callback(fence, &work->cb, nouveau_fence_work_cb) < 0)
+		goto err_free;
+	return;
 
-		nouveau_fence_signal(fence);
-		nouveau_fence_unref(&fence);
-	}
-	spin_unlock(&fctx->lock);
+err_free:
+	kfree(work);
+err:
+	func(data);
 }
 
 int
 nouveau_fence_emit(struct nouveau_fence *fence, struct nouveau_channel *chan)
 {
 	struct nouveau_fence_chan *fctx = chan->fence;
+	struct nouveau_fence_priv *priv = (void*)chan->drm->fence;
 	int ret;
 
 	fence->channel  = chan;
 	fence->timeout  = jiffies + (15 * HZ);
-	fence->sequence = ++fctx->sequence;
 
+	if (priv->uevent)
+		fence_init(&fence->base, &nouveau_fence_ops_uevent,
+			   &fctx->lock, fctx->context, ++fctx->sequence);
+	else
+		fence_init(&fence->base, &nouveau_fence_ops_legacy,
+			   &fctx->lock, fctx->context, ++fctx->sequence);
+	kref_get(&fctx->fence_ref);
+
+	trace_fence_emit(&fence->base);
 	ret = fctx->emit(fence);
 	if (!ret) {
-		kref_get(&fence->kref);
-		spin_lock(&fctx->lock);
+		fence_get(&fence->base);
+		spin_lock_irq(&fctx->lock);
+
+		if (nouveau_fence_update(chan, fctx))
+			nvif_notify_put(&fctx->notify);
+
 		list_add_tail(&fence->head, &fctx->pending);
-		spin_unlock(&fctx->lock);
+		spin_unlock_irq(&fctx->lock);
 	}
 
 	return ret;
@@ -161,114 +292,73 @@ nouveau_fence_emit(struct nouveau_fence *fence, struct nouveau_channel *chan)
 bool
 nouveau_fence_done(struct nouveau_fence *fence)
 {
-	if (fence->channel)
-		nouveau_fence_update(fence->channel);
-	return !fence->channel;
+	if (fence->base.ops == &nouveau_fence_ops_legacy ||
+	    fence->base.ops == &nouveau_fence_ops_uevent) {
+		struct nouveau_fence_chan *fctx = nouveau_fctx(fence);
+		struct nouveau_channel *chan;
+		unsigned long flags;
+
+		if (test_bit(FENCE_FLAG_SIGNALED_BIT, &fence->base.flags))
+			return true;
+
+		spin_lock_irqsave(&fctx->lock, flags);
+		chan = rcu_dereference_protected(fence->channel, lockdep_is_held(&fctx->lock));
+		if (chan && nouveau_fence_update(chan, fctx))
+			nvif_notify_put(&fctx->notify);
+		spin_unlock_irqrestore(&fctx->lock, flags);
+	}
+	return fence_is_signaled(&fence->base);
 }
 
-struct nouveau_fence_wait {
-	struct nouveau_fence_priv *priv;
-	struct nvif_notify notify;
-};
-
-static int
-nouveau_fence_wait_uevent_handler(struct nvif_notify *notify)
+static long
+nouveau_fence_wait_legacy(struct fence *f, bool intr, long wait)
 {
-	struct nouveau_fence_wait *wait =
-		container_of(notify, typeof(*wait), notify);
-	wake_up_all(&wait->priv->waiting);
-	return NVIF_NOTIFY_KEEP;
-}
-
-static int
-nouveau_fence_wait_uevent(struct nouveau_fence *fence, bool intr)
+	struct nouveau_fence *fence = from_fence(f);
+	unsigned long sleep_time = NSEC_PER_MSEC / 1000;
+	unsigned long t = jiffies, timeout = t + wait;
 
-{
-	struct nouveau_channel *chan = fence->channel;
-	struct nouveau_fence_priv *priv = chan->drm->fence;
-	struct nouveau_fence_wait wait = { .priv = priv };
-	int ret = 0;
+	while (!nouveau_fence_done(fence)) {
+		ktime_t kt;
 
-	ret = nvif_notify_init(chan->object, NULL,
-			       nouveau_fence_wait_uevent_handler, false,
-			       G82_CHANNEL_DMA_V0_NTFY_UEVENT,
-			       &(struct nvif_notify_uevent_req) {
-			       },
-			       sizeof(struct nvif_notify_uevent_req),
-			       sizeof(struct nvif_notify_uevent_rep),
-			       &wait.notify);
-	if (ret)
-		return ret;
+		t = jiffies;
 
-	nvif_notify_get(&wait.notify);
-
-	if (fence->timeout) {
-		unsigned long timeout = fence->timeout - jiffies;
-
-		if (time_before(jiffies, fence->timeout)) {
-			if (intr) {
-				ret = wait_event_interruptible_timeout(
-						priv->waiting,
-						nouveau_fence_done(fence),
-						timeout);
-			} else {
-				ret = wait_event_timeout(priv->waiting,
-						nouveau_fence_done(fence),
-						timeout);
-			}
+		if (wait != MAX_SCHEDULE_TIMEOUT && time_after_eq(t, timeout)) {
+			__set_current_state(TASK_RUNNING);
+			return 0;
 		}
 
-		if (ret >= 0) {
-			fence->timeout = jiffies + ret;
-			if (time_after_eq(jiffies, fence->timeout))
-				ret = -EBUSY;
-		}
-	} else {
-		if (intr) {
-			ret = wait_event_interruptible(priv->waiting,
-					nouveau_fence_done(fence));
-		} else {
-			wait_event(priv->waiting, nouveau_fence_done(fence));
-		}
+		__set_current_state(intr ? TASK_INTERRUPTIBLE :
+					   TASK_UNINTERRUPTIBLE);
+
+		kt = ktime_set(0, sleep_time);
+		schedule_hrtimeout(&kt, HRTIMER_MODE_REL);
+		sleep_time *= 2;
+		if (sleep_time > NSEC_PER_MSEC)
+			sleep_time = NSEC_PER_MSEC;
+
+		if (intr && signal_pending(current))
+			return -ERESTARTSYS;
 	}
 
-	nvif_notify_fini(&wait.notify);
-	if (unlikely(ret < 0))
-		return ret;
+	__set_current_state(TASK_RUNNING);
 
-	return 0;
+	return timeout - t;
 }
 
-int
-nouveau_fence_wait(struct nouveau_fence *fence, bool lazy, bool intr)
+static int
+nouveau_fence_wait_busy(struct nouveau_fence *fence, bool intr)
 {
-	struct nouveau_channel *chan = fence->channel;
-	struct nouveau_fence_priv *priv = chan ? chan->drm->fence : NULL;
-	unsigned long sleep_time = NSEC_PER_MSEC / 1000;
-	ktime_t t;
 	int ret = 0;
 
-	while (priv && priv->uevent && lazy && !nouveau_fence_done(fence)) {
-		ret = nouveau_fence_wait_uevent(fence, intr);
-		if (ret < 0)
-			return ret;
-	}
-
 	while (!nouveau_fence_done(fence)) {
-		if (fence->timeout && time_after_eq(jiffies, fence->timeout)) {
+		if (time_after_eq(jiffies, fence->timeout)) {
 			ret = -EBUSY;
 			break;
 		}
 
-		__set_current_state(intr ? TASK_INTERRUPTIBLE :
-					   TASK_UNINTERRUPTIBLE);
-		if (lazy) {
-			t = ktime_set(0, sleep_time);
-			schedule_hrtimeout(&t, HRTIMER_MODE_REL);
-			sleep_time *= 2;
-			if (sleep_time > NSEC_PER_MSEC)
-				sleep_time = NSEC_PER_MSEC;
-		}
+		__set_current_state(intr ?
+				    TASK_INTERRUPTIBLE :
+				    TASK_UNINTERRUPTIBLE);
 
 		if (intr && signal_pending(current)) {
 			ret = -ERESTARTSYS;
@@ -281,47 +371,95 @@ nouveau_fence_wait(struct nouveau_fence *fence, bool lazy, bool intr)
 }
 
 int
-nouveau_fence_sync(struct nouveau_fence *fence, struct nouveau_channel *chan)
+nouveau_fence_wait(struct nouveau_fence *fence, bool lazy, bool intr)
+{
+	long ret;
+
+	if (!lazy)
+		return nouveau_fence_wait_busy(fence, intr);
+
+	ret = fence_wait_timeout(&fence->base, intr, 15 * HZ);
+	if (ret < 0)
+		return ret;
+	else if (!ret)
+		return -EBUSY;
+	else
+		return 0;
+}
+
+int
+nouveau_fence_sync(struct nouveau_bo *nvbo, struct nouveau_channel *chan, bool exclusive, bool intr)
 {
 	struct nouveau_fence_chan *fctx = chan->fence;
-	struct nouveau_channel *prev;
-	int ret = 0;
+	struct fence *fence;
+	struct reservation_object *resv = nvbo->bo.resv;
+	struct reservation_object_list *fobj;
+	struct nouveau_fence *f;
+	int ret = 0, i;
 
-	prev = fence ? fence->channel : NULL;
-	if (prev) {
-		if (unlikely(prev != chan && !nouveau_fence_done(fence))) {
-			ret = fctx->sync(fence, prev, chan);
-			if (unlikely(ret))
-				ret = nouveau_fence_wait(fence, true, false);
+	if (!exclusive) {
+		ret = reservation_object_reserve_shared(resv);
+
+		if (ret)
+			return ret;
+	}
+
+	fobj = reservation_object_get_list(resv);
+	fence = reservation_object_get_excl(resv);
+
+	if (fence && (!exclusive || !fobj || !fobj->shared_count)) {
+		struct nouveau_channel *prev = NULL;
+		bool must_wait = true;
+
+		f = nouveau_local_fence(fence, chan->drm);
+		if (f) {
+			rcu_read_lock();
+			prev = rcu_dereference(f->channel);
+			if (prev && (prev == chan || fctx->sync(f, prev, chan) == 0))
+				must_wait = false;
+			rcu_read_unlock();
 		}
+
+		if (must_wait)
+			ret = fence_wait(fence, intr);
+
+		return ret;
 	}
 
-	return ret;
-}
+	if (!exclusive || !fobj)
+		return ret;
 
-static void
-nouveau_fence_del(struct kref *kref)
-{
-	struct nouveau_fence *fence = container_of(kref, typeof(*fence), kref);
-	kfree(fence);
+	for (i = 0; i < fobj->shared_count && !ret; ++i) {
+		struct nouveau_channel *prev = NULL;
+		bool must_wait = true;
+
+		fence = rcu_dereference_protected(fobj->shared[i],
+						reservation_object_held(resv));
+
+		f = nouveau_local_fence(fence, chan->drm);
+		if (f) {
+			rcu_read_lock();
+			prev = rcu_dereference(f->channel);
+			if (prev && (prev == chan || fctx->sync(f, prev, chan) == 0))
+				must_wait = false;
+			rcu_read_unlock();
+		}
+
+		if (must_wait)
+			ret = fence_wait(fence, intr);
+	}
+
+	return ret;
 }
 
 void
 nouveau_fence_unref(struct nouveau_fence **pfence)
 {
 	if (*pfence)
-		kref_put(&(*pfence)->kref, nouveau_fence_del);
+		fence_put(&(*pfence)->base);
 	*pfence = NULL;
 }
 
-struct nouveau_fence *
-nouveau_fence_ref(struct nouveau_fence *fence)
-{
-	if (fence)
-		kref_get(&fence->kref);
-	return fence;
-}
-
 int
 nouveau_fence_new(struct nouveau_channel *chan, bool sysmem,
 		  struct nouveau_fence **pfence)
@@ -336,9 +474,7 @@ nouveau_fence_new(struct nouveau_channel *chan, bool sysmem,
 	if (!fence)
 		return -ENOMEM;
 
-	INIT_LIST_HEAD(&fence->work);
 	fence->sysmem = sysmem;
-	kref_init(&fence->kref);
 
 	ret = nouveau_fence_emit(fence, chan);
 	if (ret)
@@ -347,3 +483,108 @@ nouveau_fence_new(struct nouveau_channel *chan, bool sysmem,
 	*pfence = fence;
 	return ret;
 }
+
+static const char *nouveau_fence_get_get_driver_name(struct fence *fence)
+{
+	return "nouveau";
+}
+
+static const char *nouveau_fence_get_timeline_name(struct fence *f)
+{
+	struct nouveau_fence *fence = from_fence(f);
+	struct nouveau_fence_chan *fctx = nouveau_fctx(fence);
+
+	return !fctx->dead ? fctx->name : "dead channel";
+}
+
+/*
+ * In an ideal world, read would not assume the channel context is still alive.
+ * This function may be called from another device, running into free memory as a
+ * result. The drm node should still be there, so we can derive the index from
+ * the fence context.
+ */
+static bool nouveau_fence_is_signaled(struct fence *f)
+{
+	struct nouveau_fence *fence = from_fence(f);
+	struct nouveau_fence_chan *fctx = nouveau_fctx(fence);
+	struct nouveau_channel *chan;
+	bool ret = false;
+
+	rcu_read_lock();
+	chan = rcu_dereference(fence->channel);
+	if (chan)
+		ret = (int)(fctx->read(chan) - fence->base.seqno) >= 0;
+	rcu_read_unlock();
+
+	return ret;
+}
+
+static bool nouveau_fence_no_signaling(struct fence *f)
+{
+	struct nouveau_fence *fence = from_fence(f);
+
+	/*
+	 * caller should have a reference on the fence,
+	 * else fence could get freed here
+	 */
+	WARN_ON(atomic_read(&fence->base.refcount.refcount) <= 1);
+
+	/*
+	 * This needs uevents to work correctly, but fence_add_callback relies on
+	 * being able to enable signaling. It will still get signaled eventually,
+	 * just not right away.
+	 */
+	if (nouveau_fence_is_signaled(f)) {
+		list_del(&fence->head);
+
+		fence_put(&fence->base);
+		return false;
+	}
+
+	return true;
+}
+
+static void nouveau_fence_release(struct fence *f)
+{
+	struct nouveau_fence *fence = from_fence(f);
+	struct nouveau_fence_chan *fctx = nouveau_fctx(fence);
+
+	kref_put(&fctx->fence_ref, nouveau_fence_context_put);
+	fence_free(&fence->base);
+}
+
+static const struct fence_ops nouveau_fence_ops_legacy = {
+	.get_driver_name = nouveau_fence_get_get_driver_name,
+	.get_timeline_name = nouveau_fence_get_timeline_name,
+	.enable_signaling = nouveau_fence_no_signaling,
+	.signaled = nouveau_fence_is_signaled,
+	.wait = nouveau_fence_wait_legacy,
+	.release = nouveau_fence_release
+};
+
+static bool nouveau_fence_enable_signaling(struct fence *f)
+{
+	struct nouveau_fence *fence = from_fence(f);
+	struct nouveau_fence_chan *fctx = nouveau_fctx(fence);
+	bool ret;
+
+	if (!fctx->notify_ref++)
+		nvif_notify_get(&fctx->notify);
+
+	ret = nouveau_fence_no_signaling(f);
+	if (ret)
+		set_bit(FENCE_FLAG_USER_BITS, &fence->base.flags);
+	else if (!--fctx->notify_ref)
+		nvif_notify_put(&fctx->notify);
+
+	return ret;
+}
+
+static const struct fence_ops nouveau_fence_ops_uevent = {
+	.get_driver_name = nouveau_fence_get_get_driver_name,
+	.get_timeline_name = nouveau_fence_get_timeline_name,
+	.enable_signaling = nouveau_fence_enable_signaling,
+	.signaled = nouveau_fence_is_signaled,
+	.wait = fence_default_wait,
+	.release = NULL
+};
diff --git a/drivers/gpu/drm/nouveau/nouveau_fence.h b/drivers/gpu/drm/nouveau/nouveau_fence.h
index c57bb61da58c..96e461c6f68f 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fence.h
+++ b/drivers/gpu/drm/nouveau/nouveau_fence.h
@@ -1,33 +1,37 @@
 #ifndef __NOUVEAU_FENCE_H__
 #define __NOUVEAU_FENCE_H__
 
+#include <linux/fence.h>
+#include <nvif/notify.h>
+
 struct nouveau_drm;
+struct nouveau_bo;
 
 struct nouveau_fence {
+	struct fence base;
+
 	struct list_head head;
-	struct list_head work;
-	struct kref kref;
 
 	bool sysmem;
 
-	struct nouveau_channel *channel;
+	struct nouveau_channel __rcu *channel;
 	unsigned long timeout;
-	u32 sequence;
 };
 
 int  nouveau_fence_new(struct nouveau_channel *, bool sysmem,
 		       struct nouveau_fence **);
-struct nouveau_fence *
-nouveau_fence_ref(struct nouveau_fence *);
 void nouveau_fence_unref(struct nouveau_fence **);
 
 int  nouveau_fence_emit(struct nouveau_fence *, struct nouveau_channel *);
 bool nouveau_fence_done(struct nouveau_fence *);
-void nouveau_fence_work(struct nouveau_fence *, void (*)(void *), void *);
+void nouveau_fence_work(struct fence *, void (*)(void *), void *);
 int  nouveau_fence_wait(struct nouveau_fence *, bool lazy, bool intr);
-int  nouveau_fence_sync(struct nouveau_fence *, struct nouveau_channel *);
+int  nouveau_fence_sync(struct nouveau_bo *, struct nouveau_channel *, bool exclusive, bool intr);
 
 struct nouveau_fence_chan {
+	spinlock_t lock;
+	struct kref fence_ref;
+
 	struct list_head pending;
 	struct list_head flip;
 
@@ -38,8 +42,12 @@ struct nouveau_fence_chan {
 	int  (*emit32)(struct nouveau_channel *, u64, u32);
 	int  (*sync32)(struct nouveau_channel *, u64, u32);
 
-	spinlock_t lock;
 	u32 sequence;
+	u32 context;
+	char name[32];
+
+	struct nvif_notify notify;
+	int notify_ref, dead;
 };
 
 struct nouveau_fence_priv {
@@ -49,14 +57,15 @@ struct nouveau_fence_priv {
 	int  (*context_new)(struct nouveau_channel *);
 	void (*context_del)(struct nouveau_channel *);
 
-	wait_queue_head_t waiting;
+	u32 contexts, context_base;
 	bool uevent;
 };
 
 #define nouveau_fence(drm) ((struct nouveau_fence_priv *)(drm)->fence)
 
-void nouveau_fence_context_new(struct nouveau_fence_chan *);
+void nouveau_fence_context_new(struct nouveau_channel *, struct nouveau_fence_chan *);
 void nouveau_fence_context_del(struct nouveau_fence_chan *);
+void nouveau_fence_context_free(struct nouveau_fence_chan *);
 
 int nv04_fence_create(struct nouveau_drm *);
 int nv04_fence_mthd(struct nouveau_channel *, u32, u32, u32);
diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.c b/drivers/gpu/drm/nouveau/nouveau_gem.c
index 292a677bfed4..bf0f9e21d714 100644
--- a/drivers/gpu/drm/nouveau/nouveau_gem.c
+++ b/drivers/gpu/drm/nouveau/nouveau_gem.c
@@ -36,7 +36,14 @@ void
 nouveau_gem_object_del(struct drm_gem_object *gem)
 {
 	struct nouveau_bo *nvbo = nouveau_gem_object(gem);
+	struct nouveau_drm *drm = nouveau_bdev(nvbo->bo.bdev);
 	struct ttm_buffer_object *bo = &nvbo->bo;
+	struct device *dev = drm->dev->dev;
+	int ret;
+
+	ret = pm_runtime_get_sync(dev);
+	if (WARN_ON(ret < 0 && ret != -EACCES))
+		return;
 
 	if (gem->import_attach)
 		drm_prime_gem_destroy(gem, nvbo->bo.sg);
@@ -46,6 +53,9 @@ nouveau_gem_object_del(struct drm_gem_object *gem)
 	/* reset filp so nouveau_bo_del_ttm() can test for it */
 	gem->filp = NULL;
 	ttm_bo_unref(&bo);
+
+	pm_runtime_mark_last_busy(dev);
+	pm_runtime_put_autosuspend(dev);
 }
 
 int
@@ -53,7 +63,9 @@ nouveau_gem_object_open(struct drm_gem_object *gem, struct drm_file *file_priv)
 {
 	struct nouveau_cli *cli = nouveau_cli(file_priv);
 	struct nouveau_bo *nvbo = nouveau_gem_object(gem);
+	struct nouveau_drm *drm = nouveau_bdev(nvbo->bo.bdev);
 	struct nouveau_vma *vma;
+	struct device *dev = drm->dev->dev;
 	int ret;
 
 	if (!cli->vm)
@@ -71,11 +83,16 @@ nouveau_gem_object_open(struct drm_gem_object *gem, struct drm_file *file_priv)
 			goto out;
 		}
 
+		ret = pm_runtime_get_sync(dev);
+		if (ret < 0 && ret != -EACCES)
+			goto out;
+
 		ret = nouveau_bo_vma_add(nvbo, cli->vm, vma);
-		if (ret) {
+		if (ret)
 			kfree(vma);
-			goto out;
-		}
+
+		pm_runtime_mark_last_busy(dev);
+		pm_runtime_put_autosuspend(dev);
 	} else {
 		vma->refcount++;
 	}
@@ -98,17 +115,23 @@ static void
 nouveau_gem_object_unmap(struct nouveau_bo *nvbo, struct nouveau_vma *vma)
 {
 	const bool mapped = nvbo->bo.mem.mem_type != TTM_PL_SYSTEM;
-	struct nouveau_fence *fence = NULL;
+	struct reservation_object *resv = nvbo->bo.resv;
+	struct reservation_object_list *fobj;
+	struct fence *fence = NULL;
+
+	fobj = reservation_object_get_list(resv);
 
 	list_del(&vma->head);
 
-	if (mapped) {
-		spin_lock(&nvbo->bo.bdev->fence_lock);
-		fence = nouveau_fence_ref(nvbo->bo.sync_obj);
-		spin_unlock(&nvbo->bo.bdev->fence_lock);
-	}
+	if (fobj && fobj->shared_count > 1)
+		ttm_bo_wait(&nvbo->bo, true, false, false);
+	else if (fobj && fobj->shared_count == 1)
+		fence = rcu_dereference_protected(fobj->shared[0],
+						reservation_object_held(resv));
+	else
+		fence = reservation_object_get_excl(nvbo->bo.resv);
 
-	if (fence) {
+	if (fence && mapped) {
 		nouveau_fence_work(fence, nouveau_gem_object_delete, vma);
 	} else {
 		if (mapped)
@@ -116,7 +139,6 @@ nouveau_gem_object_unmap(struct nouveau_bo *nvbo, struct nouveau_vma *vma)
 		nouveau_vm_put(vma);
 		kfree(vma);
 	}
-	nouveau_fence_unref(&fence);
 }
 
 void
@@ -124,6 +146,8 @@ nouveau_gem_object_close(struct drm_gem_object *gem, struct drm_file *file_priv)
 {
 	struct nouveau_cli *cli = nouveau_cli(file_priv);
 	struct nouveau_bo *nvbo = nouveau_gem_object(gem);
+	struct nouveau_drm *drm = nouveau_bdev(nvbo->bo.bdev);
+	struct device *dev = drm->dev->dev;
 	struct nouveau_vma *vma;
 	int ret;
 
@@ -136,8 +160,14 @@ nouveau_gem_object_close(struct drm_gem_object *gem, struct drm_file *file_priv)
 
 	vma = nouveau_bo_vma_find(nvbo, cli->vm);
 	if (vma) {
-		if (--vma->refcount == 0)
-			nouveau_gem_object_unmap(nvbo, vma);
+		if (--vma->refcount == 0) {
+			ret = pm_runtime_get_sync(dev);
+			if (!WARN_ON(ret < 0 && ret != -EACCES)) {
+				nouveau_gem_object_unmap(nvbo, vma);
+				pm_runtime_mark_last_busy(dev);
+				pm_runtime_put_autosuspend(dev);
+			}
+		}
 	}
 	ttm_bo_unreserve(&nvbo->bo);
 }
@@ -160,7 +190,7 @@ nouveau_gem_new(struct drm_device *dev, int size, int align, uint32_t domain,
 		flags |= TTM_PL_FLAG_SYSTEM;
 
 	ret = nouveau_bo_new(dev, size, align, flags, tile_mode,
-			     tile_flags, NULL, pnvbo);
+			     tile_flags, NULL, NULL, pnvbo);
 	if (ret)
 		return ret;
 	nvbo = *pnvbo;
@@ -288,24 +318,23 @@ nouveau_gem_set_domain(struct drm_gem_object *gem, uint32_t read_domains,
 }
 
 struct validate_op {
-	struct list_head vram_list;
-	struct list_head gart_list;
-	struct list_head both_list;
+	struct list_head list;
 	struct ww_acquire_ctx ticket;
 };
 
 static void
-validate_fini_list(struct list_head *list, struct nouveau_fence *fence,
-		   struct ww_acquire_ctx *ticket)
+validate_fini_no_ticket(struct validate_op *op, struct nouveau_fence *fence,
+			struct drm_nouveau_gem_pushbuf_bo *pbbo)
 {
-	struct list_head *entry, *tmp;
 	struct nouveau_bo *nvbo;
+	struct drm_nouveau_gem_pushbuf_bo *b;
 
-	list_for_each_safe(entry, tmp, list) {
-		nvbo = list_entry(entry, struct nouveau_bo, entry);
+	while (!list_empty(&op->list)) {
+		nvbo = list_entry(op->list.next, struct nouveau_bo, entry);
+		b = &pbbo[nvbo->pbbo_index];
 
 		if (likely(fence))
-			nouveau_bo_fence(nvbo, fence);
+			nouveau_bo_fence(nvbo, fence, !!b->write_domains);
 
 		if (unlikely(nvbo->validate_mapped)) {
 			ttm_bo_kunmap(&nvbo->kmap);
@@ -314,23 +343,16 @@ validate_fini_list(struct list_head *list, struct nouveau_fence *fence,
 
 		list_del(&nvbo->entry);
 		nvbo->reserved_by = NULL;
-		ttm_bo_unreserve_ticket(&nvbo->bo, ticket);
+		ttm_bo_unreserve_ticket(&nvbo->bo, &op->ticket);
 		drm_gem_object_unreference_unlocked(&nvbo->gem);
 	}
 }
 
 static void
-validate_fini_no_ticket(struct validate_op *op, struct nouveau_fence *fence)
+validate_fini(struct validate_op *op, struct nouveau_fence *fence,
+	      struct drm_nouveau_gem_pushbuf_bo *pbbo)
 {
-	validate_fini_list(&op->vram_list, fence, &op->ticket);
-	validate_fini_list(&op->gart_list, fence, &op->ticket);
-	validate_fini_list(&op->both_list, fence, &op->ticket);
-}
-
-static void
-validate_fini(struct validate_op *op, struct nouveau_fence *fence)
-{
-	validate_fini_no_ticket(op, fence);
+	validate_fini_no_ticket(op, fence, pbbo);
 	ww_acquire_fini(&op->ticket);
 }
 
@@ -344,6 +366,9 @@ validate_init(struct nouveau_channel *chan, struct drm_file *file_priv,
 	int trycnt = 0;
 	int ret, i;
 	struct nouveau_bo *res_bo = NULL;
+	LIST_HEAD(gart_list);
+	LIST_HEAD(vram_list);
+	LIST_HEAD(both_list);
 
 	ww_acquire_init(&op->ticket, &reservation_ww_class);
 retry:
@@ -360,9 +385,8 @@ retry:
 		gem = drm_gem_object_lookup(dev, file_priv, b->handle);
 		if (!gem) {
 			NV_PRINTK(error, cli, "Unknown handle 0x%08x\n", b->handle);
-			ww_acquire_done(&op->ticket);
-			validate_fini(op, NULL);
-			return -ENOENT;
+			ret = -ENOENT;
+			break;
 		}
 		nvbo = nouveau_gem_object(gem);
 		if (nvbo == res_bo) {
@@ -375,14 +399,16 @@ retry:
 			NV_PRINTK(error, cli, "multiple instances of buffer %d on "
 				      "validation list\n", b->handle);
 			drm_gem_object_unreference_unlocked(gem);
-			ww_acquire_done(&op->ticket);
-			validate_fini(op, NULL);
-			return -EINVAL;
+			ret = -EINVAL;
+			break;
 		}
 
 		ret = ttm_bo_reserve(&nvbo->bo, true, false, true, &op->ticket);
 		if (ret) {
-			validate_fini_no_ticket(op, NULL);
+			list_splice_tail_init(&vram_list, &op->list);
+			list_splice_tail_init(&gart_list, &op->list);
+			list_splice_tail_init(&both_list, &op->list);
+			validate_fini_no_ticket(op, NULL, NULL);
 			if (unlikely(ret == -EDEADLK)) {
 				ret = ttm_bo_reserve_slowpath(&nvbo->bo, true,
 							      &op->ticket);
@@ -390,12 +416,9 @@ retry:
 					res_bo = nvbo;
 			}
 			if (unlikely(ret)) {
-				ww_acquire_done(&op->ticket);
-				ww_acquire_fini(&op->ticket);
-				drm_gem_object_unreference_unlocked(gem);
 				if (ret != -ERESTARTSYS)
 					NV_PRINTK(error, cli, "fail reserve\n");
-				return ret;
+				break;
 			}
 		}
 
@@ -404,45 +427,32 @@ retry:
 		nvbo->pbbo_index = i;
 		if ((b->valid_domains & NOUVEAU_GEM_DOMAIN_VRAM) &&
 		    (b->valid_domains & NOUVEAU_GEM_DOMAIN_GART))
-			list_add_tail(&nvbo->entry, &op->both_list);
+			list_add_tail(&nvbo->entry, &both_list);
 		else
 		if (b->valid_domains & NOUVEAU_GEM_DOMAIN_VRAM)
-			list_add_tail(&nvbo->entry, &op->vram_list);
+			list_add_tail(&nvbo->entry, &vram_list);
 		else
 		if (b->valid_domains & NOUVEAU_GEM_DOMAIN_GART)
-			list_add_tail(&nvbo->entry, &op->gart_list);
+			list_add_tail(&nvbo->entry, &gart_list);
 		else {
 			NV_PRINTK(error, cli, "invalid valid domains: 0x%08x\n",
 				 b->valid_domains);
-			list_add_tail(&nvbo->entry, &op->both_list);
-			ww_acquire_done(&op->ticket);
-			validate_fini(op, NULL);
-			return -EINVAL;
+			list_add_tail(&nvbo->entry, &both_list);
+			ret = -EINVAL;
+			break;
 		}
 		if (nvbo == res_bo)
 			goto retry;
 	}
 
 	ww_acquire_done(&op->ticket);
-	return 0;
-}
-
-static int
-validate_sync(struct nouveau_channel *chan, struct nouveau_bo *nvbo)
-{
-	struct nouveau_fence *fence = NULL;
-	int ret = 0;
-
-	spin_lock(&nvbo->bo.bdev->fence_lock);
-	fence = nouveau_fence_ref(nvbo->bo.sync_obj);
-	spin_unlock(&nvbo->bo.bdev->fence_lock);
-
-	if (fence) {
-		ret = nouveau_fence_sync(fence, chan);
-		nouveau_fence_unref(&fence);
-	}
-
+	list_splice_tail(&vram_list, &op->list);
+	list_splice_tail(&gart_list, &op->list);
+	list_splice_tail(&both_list, &op->list);
+	if (ret)
+		validate_fini(op, NULL, NULL);
 	return ret;
+
 }
 
 static int
@@ -474,9 +484,10 @@ validate_list(struct nouveau_channel *chan, struct nouveau_cli *cli,
 			return ret;
 		}
 
-		ret = validate_sync(chan, nvbo);
+		ret = nouveau_fence_sync(nvbo, chan, !!b->write_domains, true);
 		if (unlikely(ret)) {
-			NV_PRINTK(error, cli, "fail post-validate sync\n");
+			if (ret != -ERESTARTSYS)
+				NV_PRINTK(error, cli, "fail post-validate sync\n");
 			return ret;
 		}
 
@@ -513,11 +524,9 @@ nouveau_gem_pushbuf_validate(struct nouveau_channel *chan,
 			     struct validate_op *op, int *apply_relocs)
 {
 	struct nouveau_cli *cli = nouveau_cli(file_priv);
-	int ret, relocs = 0;
+	int ret;
 
-	INIT_LIST_HEAD(&op->vram_list);
-	INIT_LIST_HEAD(&op->gart_list);
-	INIT_LIST_HEAD(&op->both_list);
+	INIT_LIST_HEAD(&op->list);
 
 	if (nr_buffers == 0)
 		return 0;
@@ -529,34 +538,14 @@ nouveau_gem_pushbuf_validate(struct nouveau_channel *chan,
 		return ret;
 	}
 
-	ret = validate_list(chan, cli, &op->vram_list, pbbo, user_buffers);
+	ret = validate_list(chan, cli, &op->list, pbbo, user_buffers);
 	if (unlikely(ret < 0)) {
 		if (ret != -ERESTARTSYS)
-			NV_PRINTK(error, cli, "validate vram_list\n");
-		validate_fini(op, NULL);
+			NV_PRINTK(error, cli, "validating bo list\n");
+		validate_fini(op, NULL, NULL);
 		return ret;
 	}
-	relocs += ret;
-
-	ret = validate_list(chan, cli, &op->gart_list, pbbo, user_buffers);
-	if (unlikely(ret < 0)) {
-		if (ret != -ERESTARTSYS)
-			NV_PRINTK(error, cli, "validate gart_list\n");
-		validate_fini(op, NULL);
-		return ret;
-	}
-	relocs += ret;
-
-	ret = validate_list(chan, cli, &op->both_list, pbbo, user_buffers);
-	if (unlikely(ret < 0)) {
-		if (ret != -ERESTARTSYS)
-			NV_PRINTK(error, cli, "validate both_list\n");
-		validate_fini(op, NULL);
-		return ret;
-	}
-	relocs += ret;
-
-	*apply_relocs = relocs;
+	*apply_relocs = ret;
 	return 0;
 }
 
@@ -659,9 +648,7 @@ nouveau_gem_pushbuf_reloc_apply(struct nouveau_cli *cli,
 				data |= r->vor;
 		}
 
-		spin_lock(&nvbo->bo.bdev->fence_lock);
-		ret = ttm_bo_wait(&nvbo->bo, false, false, false);
-		spin_unlock(&nvbo->bo.bdev->fence_lock);
+		ret = ttm_bo_wait(&nvbo->bo, true, false, false);
 		if (ret) {
 			NV_PRINTK(error, cli, "reloc wait_idle failed: %d\n", ret);
 			break;
@@ -839,7 +826,7 @@ nouveau_gem_ioctl_pushbuf(struct drm_device *dev, void *data,
 	}
 
 out:
-	validate_fini(&op, fence);
+	validate_fini(&op, fence, bo);
 	nouveau_fence_unref(&fence);
 
 out_prevalid:
@@ -884,17 +871,30 @@ nouveau_gem_ioctl_cpu_prep(struct drm_device *dev, void *data,
 	struct drm_gem_object *gem;
 	struct nouveau_bo *nvbo;
 	bool no_wait = !!(req->flags & NOUVEAU_GEM_CPU_PREP_NOWAIT);
-	int ret = -EINVAL;
+	bool write = !!(req->flags & NOUVEAU_GEM_CPU_PREP_WRITE);
+	int ret;
 
 	gem = drm_gem_object_lookup(dev, file_priv, req->handle);
 	if (!gem)
 		return -ENOENT;
 	nvbo = nouveau_gem_object(gem);
 
-	spin_lock(&nvbo->bo.bdev->fence_lock);
-	ret = ttm_bo_wait(&nvbo->bo, true, true, no_wait);
-	spin_unlock(&nvbo->bo.bdev->fence_lock);
+	if (no_wait)
+		ret = reservation_object_test_signaled_rcu(nvbo->bo.resv, write) ? 0 : -EBUSY;
+	else {
+		long lret;
+
+		lret = reservation_object_wait_timeout_rcu(nvbo->bo.resv, write, true, 30 * HZ);
+		if (!lret)
+			ret = -EBUSY;
+		else if (lret > 0)
+			ret = 0;
+		else
+			ret = lret;
+	}
+	nouveau_bo_sync_for_cpu(nvbo);
 	drm_gem_object_unreference_unlocked(gem);
+
 	return ret;
 }
 
@@ -902,6 +902,17 @@ int
 nouveau_gem_ioctl_cpu_fini(struct drm_device *dev, void *data,
 			   struct drm_file *file_priv)
 {
+	struct drm_nouveau_gem_cpu_fini *req = data;
+	struct drm_gem_object *gem;
+	struct nouveau_bo *nvbo;
+
+	gem = drm_gem_object_lookup(dev, file_priv, req->handle);
+	if (!gem)
+		return -ENOENT;
+	nvbo = nouveau_gem_object(gem);
+
+	nouveau_bo_sync_for_device(nvbo);
+	drm_gem_object_unreference_unlocked(gem);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.h b/drivers/gpu/drm/nouveau/nouveau_gem.h
index ddab762d81fe..e4049faca780 100644
--- a/drivers/gpu/drm/nouveau/nouveau_gem.h
+++ b/drivers/gpu/drm/nouveau/nouveau_gem.h
@@ -39,7 +39,7 @@ struct reservation_object *nouveau_gem_prime_res_obj(struct drm_gem_object *);
 extern void nouveau_gem_prime_unpin(struct drm_gem_object *);
 extern struct sg_table *nouveau_gem_prime_get_sg_table(struct drm_gem_object *);
 extern struct drm_gem_object *nouveau_gem_prime_import_sg_table(
-	struct drm_device *, size_t size, struct sg_table *);
+	struct drm_device *, struct dma_buf_attachment *, struct sg_table *);
 extern void *nouveau_gem_prime_vmap(struct drm_gem_object *);
 extern void nouveau_gem_prime_vunmap(struct drm_gem_object *, void *);
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_nvif.c b/drivers/gpu/drm/nouveau/nouveau_nvif.c
index 47ca88623753..6544b84f0303 100644
--- a/drivers/gpu/drm/nouveau/nouveau_nvif.c
+++ b/drivers/gpu/drm/nouveau/nouveau_nvif.c
@@ -40,12 +40,12 @@
 #include "nouveau_usif.h"
 
 static void
-nvkm_client_unmap(void *priv, void *ptr, u32 size)
+nvkm_client_unmap(void *priv, void __iomem *ptr, u32 size)
 {
 	iounmap(ptr);
 }
 
-static void *
+static void __iomem *
 nvkm_client_map(void *priv, u64 handle, u32 size)
 {
 	return ioremap(handle, size);
diff --git a/drivers/gpu/drm/nouveau/nouveau_platform.c b/drivers/gpu/drm/nouveau/nouveau_platform.c
index 246a824c16ca..b307bbedd4c4 100644
--- a/drivers/gpu/drm/nouveau/nouveau_platform.c
+++ b/drivers/gpu/drm/nouveau/nouveau_platform.c
@@ -27,6 +27,7 @@
 #include <linux/of.h>
 #include <linux/reset.h>
 #include <linux/regulator/consumer.h>
+#include <soc/tegra/fuse.h>
 #include <soc/tegra/pmc.h>
 
 #include "nouveau_drm.h"
@@ -128,6 +129,7 @@ static int nouveau_platform_probe(struct platform_device *pdev)
 	}
 
 	device->gpu = gpu;
+	device->gpu_speedo = tegra_sku_info.gpu_speedo_value;
 
 	err = drm_dev_register(drm, 0);
 	if (err < 0)
diff --git a/drivers/gpu/drm/nouveau/nouveau_platform.h b/drivers/gpu/drm/nouveau/nouveau_platform.h
index 91f66504900e..58c28b5653d5 100644
--- a/drivers/gpu/drm/nouveau/nouveau_platform.h
+++ b/drivers/gpu/drm/nouveau/nouveau_platform.h
@@ -41,6 +41,8 @@ struct nouveau_platform_device {
 	struct nouveau_device device;
 
 	struct nouveau_platform_gpu *gpu;
+
+	int gpu_speedo;
 };
 
 #define nv_device_to_platform(d)                                               \
diff --git a/drivers/gpu/drm/nouveau/nouveau_prime.c b/drivers/gpu/drm/nouveau/nouveau_prime.c
index 1f51008e4d26..dd32ad6db53d 100644
--- a/drivers/gpu/drm/nouveau/nouveau_prime.c
+++ b/drivers/gpu/drm/nouveau/nouveau_prime.c
@@ -23,6 +23,7 @@
  */
 
 #include <drm/drmP.h>
+#include <linux/dma-buf.h>
 
 #include "nouveau_drm.h"
 #include "nouveau_gem.h"
@@ -56,17 +57,20 @@ void nouveau_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr)
 }
 
 struct drm_gem_object *nouveau_gem_prime_import_sg_table(struct drm_device *dev,
-							 size_t size,
+							 struct dma_buf_attachment *attach,
 							 struct sg_table *sg)
 {
 	struct nouveau_bo *nvbo;
+	struct reservation_object *robj = attach->dmabuf->resv;
 	u32 flags = 0;
 	int ret;
 
 	flags = TTM_PL_FLAG_TT;
 
-	ret = nouveau_bo_new(dev, size, 0, flags, 0, 0,
-			     sg, &nvbo);
+	ww_mutex_lock(&robj->lock, NULL);
+	ret = nouveau_bo_new(dev, attach->dmabuf->size, 0, flags, 0, 0,
+			     sg, robj, &nvbo);
+	ww_mutex_unlock(&robj->lock);
 	if (ret)
 		return ERR_PTR(ret);
 
@@ -89,7 +93,7 @@ int nouveau_gem_prime_pin(struct drm_gem_object *obj)
 	int ret;
 
 	/* pin buffer into GTT */
-	ret = nouveau_bo_pin(nvbo, TTM_PL_FLAG_TT);
+	ret = nouveau_bo_pin(nvbo, TTM_PL_FLAG_TT, false);
 	if (ret)
 		return -EINVAL;
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_sysfs.c b/drivers/gpu/drm/nouveau/nouveau_sysfs.c
index 3c6962d15b26..8fbbf3093d86 100644
--- a/drivers/gpu/drm/nouveau/nouveau_sysfs.c
+++ b/drivers/gpu/drm/nouveau/nouveau_sysfs.c
@@ -29,7 +29,7 @@
 #include "nouveau_sysfs.h"
 
 MODULE_PARM_DESC(pstate, "enable sysfs pstate file, which will be moved in the future");
-static int nouveau_pstate;
+int nouveau_pstate;
 module_param_named(pstate, nouveau_pstate, int, 0400);
 
 static inline struct drm_device *
diff --git a/drivers/gpu/drm/nouveau/nouveau_sysfs.h b/drivers/gpu/drm/nouveau/nouveau_sysfs.h
index f973378160f8..4e5ea9241b28 100644
--- a/drivers/gpu/drm/nouveau/nouveau_sysfs.h
+++ b/drivers/gpu/drm/nouveau/nouveau_sysfs.h
@@ -16,4 +16,6 @@ nouveau_sysfs(struct drm_device *dev)
 int  nouveau_sysfs_init(struct drm_device *);
 void nouveau_sysfs_fini(struct drm_device *);
 
+extern int nouveau_pstate;
+
 #endif
diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c b/drivers/gpu/drm/nouveau/nouveau_ttm.c
index 53874b76b031..3d1cfcb96b6b 100644
--- a/drivers/gpu/drm/nouveau/nouveau_ttm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c
@@ -28,6 +28,7 @@
 #include "nouveau_ttm.h"
 #include "nouveau_gem.h"
 
+#include "drm_legacy.h"
 static int
 nouveau_vram_manager_init(struct ttm_mem_type_manager *man, unsigned long psize)
 {
@@ -71,8 +72,7 @@ nouveau_vram_manager_del(struct ttm_mem_type_manager *man,
 static int
 nouveau_vram_manager_new(struct ttm_mem_type_manager *man,
 			 struct ttm_buffer_object *bo,
-			 struct ttm_placement *placement,
-			 uint32_t flags,
+			 const struct ttm_place *place,
 			 struct ttm_mem_reg *mem)
 {
 	struct nouveau_drm *drm = nouveau_bdev(man->bdev);
@@ -158,8 +158,7 @@ nouveau_gart_manager_del(struct ttm_mem_type_manager *man,
 static int
 nouveau_gart_manager_new(struct ttm_mem_type_manager *man,
 			 struct ttm_buffer_object *bo,
-			 struct ttm_placement *placement,
-			 uint32_t flags,
+			 const struct ttm_place *place,
 			 struct ttm_mem_reg *mem)
 {
 	struct nouveau_drm *drm = nouveau_bdev(bo->bdev);
@@ -239,8 +238,7 @@ nv04_gart_manager_del(struct ttm_mem_type_manager *man, struct ttm_mem_reg *mem)
 static int
 nv04_gart_manager_new(struct ttm_mem_type_manager *man,
 		      struct ttm_buffer_object *bo,
-		      struct ttm_placement *placement,
-		      uint32_t flags,
+		      const struct ttm_place *place,
 		      struct ttm_mem_reg *mem)
 {
 	struct nouveau_mem *node;
@@ -284,7 +282,7 @@ nouveau_ttm_mmap(struct file *filp, struct vm_area_struct *vma)
 	struct nouveau_drm *drm = nouveau_drm(file_priv->minor->dev);
 
 	if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET))
-		return drm_mmap(filp, vma);
+		return drm_legacy_mmap(filp, vma);
 
 	return ttm_bo_mmap(filp, vma, &drm->ttm.bdev);
 }
diff --git a/drivers/gpu/drm/nouveau/nv04_fence.c b/drivers/gpu/drm/nouveau/nv04_fence.c
index 239c2c5a9615..f9859deb108a 100644
--- a/drivers/gpu/drm/nouveau/nv04_fence.c
+++ b/drivers/gpu/drm/nouveau/nv04_fence.c
@@ -41,7 +41,7 @@ nv04_fence_emit(struct nouveau_fence *fence)
 	int ret = RING_SPACE(chan, 2);
 	if (ret == 0) {
 		BEGIN_NV04(chan, NvSubSw, 0x0150, 1);
-		OUT_RING  (chan, fence->sequence);
+		OUT_RING  (chan, fence->base.seqno);
 		FIRE_RING (chan);
 	}
 	return ret;
@@ -67,7 +67,7 @@ nv04_fence_context_del(struct nouveau_channel *chan)
 	struct nv04_fence_chan *fctx = chan->fence;
 	nouveau_fence_context_del(&fctx->base);
 	chan->fence = NULL;
-	kfree(fctx);
+	nouveau_fence_context_free(&fctx->base);
 }
 
 static int
@@ -75,7 +75,7 @@ nv04_fence_context_new(struct nouveau_channel *chan)
 {
 	struct nv04_fence_chan *fctx = kzalloc(sizeof(*fctx), GFP_KERNEL);
 	if (fctx) {
-		nouveau_fence_context_new(&fctx->base);
+		nouveau_fence_context_new(chan, &fctx->base);
 		fctx->base.emit = nv04_fence_emit;
 		fctx->base.sync = nv04_fence_sync;
 		fctx->base.read = nv04_fence_read;
@@ -105,5 +105,7 @@ nv04_fence_create(struct nouveau_drm *drm)
 	priv->base.dtor = nv04_fence_destroy;
 	priv->base.context_new = nv04_fence_context_new;
 	priv->base.context_del = nv04_fence_context_del;
+	priv->base.contexts = 15;
+	priv->base.context_base = fence_context_alloc(priv->base.contexts);
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/nv10_fence.c b/drivers/gpu/drm/nouveau/nv10_fence.c
index 4faaf0acf5d7..5e1ea1cdce75 100644
--- a/drivers/gpu/drm/nouveau/nv10_fence.c
+++ b/drivers/gpu/drm/nouveau/nv10_fence.c
@@ -33,7 +33,7 @@ nv10_fence_emit(struct nouveau_fence *fence)
 	int ret = RING_SPACE(chan, 2);
 	if (ret == 0) {
 		BEGIN_NV04(chan, 0, NV10_SUBCHAN_REF_CNT, 1);
-		OUT_RING  (chan, fence->sequence);
+		OUT_RING  (chan, fence->base.seqno);
 		FIRE_RING (chan);
 	}
 	return ret;
@@ -63,7 +63,7 @@ nv10_fence_context_del(struct nouveau_channel *chan)
 		nvif_object_fini(&fctx->head[i]);
 	nvif_object_fini(&fctx->sema);
 	chan->fence = NULL;
-	kfree(fctx);
+	nouveau_fence_context_free(&fctx->base);
 }
 
 int
@@ -75,7 +75,7 @@ nv10_fence_context_new(struct nouveau_channel *chan)
 	if (!fctx)
 		return -ENOMEM;
 
-	nouveau_fence_context_new(&fctx->base);
+	nouveau_fence_context_new(chan, &fctx->base);
 	fctx->base.emit = nv10_fence_emit;
 	fctx->base.read = nv10_fence_read;
 	fctx->base.sync = nv10_fence_sync;
@@ -106,6 +106,8 @@ nv10_fence_create(struct nouveau_drm *drm)
 	priv->base.dtor = nv10_fence_destroy;
 	priv->base.context_new = nv10_fence_context_new;
 	priv->base.context_del = nv10_fence_context_del;
+	priv->base.contexts = 31;
+	priv->base.context_base = fence_context_alloc(priv->base.contexts);
 	spin_lock_init(&priv->lock);
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/nv17_fence.c b/drivers/gpu/drm/nouveau/nv17_fence.c
index ca907479f92f..57860cfa1de5 100644
--- a/drivers/gpu/drm/nouveau/nv17_fence.c
+++ b/drivers/gpu/drm/nouveau/nv17_fence.c
@@ -84,7 +84,7 @@ nv17_fence_context_new(struct nouveau_channel *chan)
 	if (!fctx)
 		return -ENOMEM;
 
-	nouveau_fence_context_new(&fctx->base);
+	nouveau_fence_context_new(chan, &fctx->base);
 	fctx->base.emit = nv10_fence_emit;
 	fctx->base.read = nv10_fence_read;
 	fctx->base.sync = nv17_fence_sync;
@@ -124,12 +124,14 @@ nv17_fence_create(struct nouveau_drm *drm)
 	priv->base.resume = nv17_fence_resume;
 	priv->base.context_new = nv17_fence_context_new;
 	priv->base.context_del = nv10_fence_context_del;
+	priv->base.contexts = 31;
+	priv->base.context_base = fence_context_alloc(priv->base.contexts);
 	spin_lock_init(&priv->lock);
 
 	ret = nouveau_bo_new(drm->dev, 4096, 0x1000, TTM_PL_FLAG_VRAM,
-			     0, 0x0000, NULL, &priv->bo);
+			     0, 0x0000, NULL, NULL, &priv->bo);
 	if (!ret) {
-		ret = nouveau_bo_pin(priv->bo, TTM_PL_FLAG_VRAM);
+		ret = nouveau_bo_pin(priv->bo, TTM_PL_FLAG_VRAM, false);
 		if (!ret) {
 			ret = nouveau_bo_map(priv->bo);
 			if (ret)
diff --git a/drivers/gpu/drm/nouveau/nv50_display.c b/drivers/gpu/drm/nouveau/nv50_display.c
index 03949eaa629f..490b90866baf 100644
--- a/drivers/gpu/drm/nouveau/nv50_display.c
+++ b/drivers/gpu/drm/nouveau/nv50_display.c
@@ -26,6 +26,7 @@
 
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_plane_helper.h>
 #include <drm/drm_dp_helper.h>
 
 #include <nvif/class.h>
@@ -65,15 +66,29 @@ static int
 nv50_chan_create(struct nvif_object *disp, const u32 *oclass, u8 head,
 		 void *data, u32 size, struct nv50_chan *chan)
 {
+	const u32 handle = (oclass[0] << 16) | head;
+	u32 sclass[8];
+	int ret, i;
+
+	ret = nvif_object_sclass(disp, sclass, ARRAY_SIZE(sclass));
+	WARN_ON(ret > ARRAY_SIZE(sclass));
+	if (ret < 0)
+		return ret;
+
 	while (oclass[0]) {
-		int ret = nvif_object_init(disp, NULL, (oclass[0] << 16) | head,
-					   oclass[0], data, size,
-					  &chan->user);
-		if (oclass++, ret == 0) {
-			nvif_object_map(&chan->user);
-			return ret;
+		for (i = 0; i < ARRAY_SIZE(sclass); i++) {
+			if (sclass[i] == oclass[0]) {
+				ret = nvif_object_init(disp, NULL, handle,
+						       oclass[0], data, size,
+						       &chan->user);
+				if (ret == 0)
+					nvif_object_map(&chan->user);
+				return ret;
+			}
 		}
+		oclass++;
 	}
+
 	return -ENOSYS;
 }
 
@@ -110,6 +125,7 @@ nv50_pioc_create(struct nvif_object *disp, const u32 *oclass, u8 head,
 
 struct nv50_curs {
 	struct nv50_pioc base;
+	struct nouveau_bo *image;
 };
 
 static int
@@ -265,6 +281,7 @@ nv50_core_create(struct nvif_object *disp, u64 syncbuf, struct nv50_mast *core)
 		.pushbuf = 0xb0007d00,
 	};
 	static const u32 oclass[] = {
+		GM204_DISP_CORE_CHANNEL_DMA,
 		GM107_DISP_CORE_CHANNEL_DMA,
 		GK110_DISP_CORE_CHANNEL_DMA,
 		GK104_DISP_CORE_CHANNEL_DMA,
@@ -424,8 +441,21 @@ evo_kick(u32 *push, void *evoc)
 	mutex_unlock(&dmac->lock);
 }
 
+#if 1
 #define evo_mthd(p,m,s) *((p)++) = (((s) << 18) | (m))
 #define evo_data(p,d)   *((p)++) = (d)
+#else
+#define evo_mthd(p,m,s) do {                                                   \
+	const u32 _m = (m), _s = (s);                                          \
+	printk(KERN_ERR "%04x %d %s\n", _m, _s, __func__);                     \
+	*((p)++) = ((_s << 18) | _m);                                          \
+} while(0)
+#define evo_data(p,d) do {                                                     \
+	const u32 _d = (d);                                                    \
+	printk(KERN_ERR "\t%08x\n", _d);                                       \
+	*((p)++) = _d;                                                         \
+} while(0)
+#endif
 
 static bool
 evo_sync_wait(void *data)
@@ -791,6 +821,22 @@ nv50_crtc_set_scale(struct nouveau_crtc *nv_crtc, bool update)
 }
 
 static int
+nv50_crtc_set_raster_vblank_dmi(struct nouveau_crtc *nv_crtc, u32 usec)
+{
+	struct nv50_mast *mast = nv50_mast(nv_crtc->base.dev);
+	u32 *push;
+
+	push = evo_wait(mast, 8);
+	if (!push)
+		return -ENOMEM;
+
+	evo_mthd(push, 0x0828 + (nv_crtc->index * 0x400), 1);
+	evo_data(push, usec);
+	evo_kick(push, mast);
+	return 0;
+}
+
+static int
 nv50_crtc_set_color_vibrance(struct nouveau_crtc *nv_crtc, bool update)
 {
 	struct nv50_mast *mast = nv50_mast(nv_crtc->base.dev);
@@ -871,23 +917,24 @@ static void
 nv50_crtc_cursor_show(struct nouveau_crtc *nv_crtc)
 {
 	struct nv50_mast *mast = nv50_mast(nv_crtc->base.dev);
+	struct nv50_curs *curs = nv50_curs(&nv_crtc->base);
 	u32 *push = evo_wait(mast, 16);
 	if (push) {
 		if (nv50_vers(mast) < G82_DISP_CORE_CHANNEL_DMA) {
 			evo_mthd(push, 0x0880 + (nv_crtc->index * 0x400), 2);
 			evo_data(push, 0x85000000);
-			evo_data(push, nv_crtc->cursor.nvbo->bo.offset >> 8);
+			evo_data(push, curs->image->bo.offset >> 8);
 		} else
 		if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA) {
 			evo_mthd(push, 0x0880 + (nv_crtc->index * 0x400), 2);
 			evo_data(push, 0x85000000);
-			evo_data(push, nv_crtc->cursor.nvbo->bo.offset >> 8);
+			evo_data(push, curs->image->bo.offset >> 8);
 			evo_mthd(push, 0x089c + (nv_crtc->index * 0x400), 1);
 			evo_data(push, mast->base.vram.handle);
 		} else {
 			evo_mthd(push, 0x0480 + (nv_crtc->index * 0x300), 2);
 			evo_data(push, 0x85000000);
-			evo_data(push, nv_crtc->cursor.nvbo->bo.offset >> 8);
+			evo_data(push, curs->image->bo.offset >> 8);
 			evo_mthd(push, 0x048c + (nv_crtc->index * 0x300), 1);
 			evo_data(push, mast->base.vram.handle);
 		}
@@ -924,8 +971,9 @@ static void
 nv50_crtc_cursor_show_hide(struct nouveau_crtc *nv_crtc, bool show, bool update)
 {
 	struct nv50_mast *mast = nv50_mast(nv_crtc->base.dev);
+	struct nv50_curs *curs = nv50_curs(&nv_crtc->base);
 
-	if (show)
+	if (show && curs->image)
 		nv50_crtc_cursor_show(nv_crtc);
 	else
 		nv50_crtc_cursor_hide(nv_crtc);
@@ -1025,7 +1073,7 @@ nv50_crtc_commit(struct drm_crtc *crtc)
 		evo_kick(push, mast);
 	}
 
-	nv50_crtc_cursor_show_hide(nv_crtc, nv_crtc->cursor.visible, true);
+	nv50_crtc_cursor_show_hide(nv_crtc, true, true);
 	nv50_display_flip_next(crtc, crtc->primary->fb, NULL, 1);
 }
 
@@ -1044,7 +1092,7 @@ nv50_crtc_swap_fbs(struct drm_crtc *crtc, struct drm_framebuffer *old_fb)
 	struct nv50_head *head = nv50_head(crtc);
 	int ret;
 
-	ret = nouveau_bo_pin(nvfb->nvbo, TTM_PL_FLAG_VRAM);
+	ret = nouveau_bo_pin(nvfb->nvbo, TTM_PL_FLAG_VRAM, true);
 	if (ret == 0) {
 		if (head->image)
 			nouveau_bo_unpin(head->image);
@@ -1066,7 +1114,7 @@ nv50_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *umode,
 	u32 vscan = (mode->flags & DRM_MODE_FLAG_DBLSCAN) ? 2 : 1;
 	u32 hactive, hsynce, hbackp, hfrontp, hblanke, hblanks;
 	u32 vactive, vsynce, vbackp, vfrontp, vblanke, vblanks;
-	u32 vblan2e = 0, vblan2s = 1;
+	u32 vblan2e = 0, vblan2s = 1, vblankus = 0;
 	u32 *push;
 	int ret;
 
@@ -1083,6 +1131,11 @@ nv50_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *umode,
 	vblanke = vsynce + vbackp;
 	vfrontp = (mode->vsync_start - mode->vdisplay) * vscan / ilace;
 	vblanks = vactive - vfrontp - 1;
+	/* XXX: Safe underestimate, even "0" works */
+	vblankus = (vactive - mode->vdisplay - 2) * hactive;
+	vblankus *= 1000;
+	vblankus /= mode->clock;
+
 	if (mode->flags & DRM_MODE_FLAG_INTERLACE) {
 		vblan2e = vactive + vsynce + vbackp;
 		vblan2s = vblan2e + (mode->vdisplay * vscan / ilace);
@@ -1136,6 +1189,11 @@ nv50_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *umode,
 	nv_connector = nouveau_crtc_connector_get(nv_crtc);
 	nv50_crtc_set_dither(nv_crtc, false);
 	nv50_crtc_set_scale(nv_crtc, false);
+
+	/* G94 only accepts this after setting scale */
+	if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA)
+		nv50_crtc_set_raster_vblank_dmi(nv_crtc, vblankus);
+
 	nv50_crtc_set_color_vibrance(nv_crtc, false);
 	nv50_crtc_set_image(nv_crtc, crtc->primary->fb, x, y, false);
 	return 0;
@@ -1215,13 +1273,13 @@ nv50_crtc_cursor_set(struct drm_crtc *crtc, struct drm_file *file_priv,
 		     uint32_t handle, uint32_t width, uint32_t height)
 {
 	struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
+	struct nv50_curs *curs = nv50_curs(crtc);
 	struct drm_device *dev = crtc->dev;
-	struct drm_gem_object *gem;
-	struct nouveau_bo *nvbo;
-	bool visible = (handle != 0);
-	int i, ret = 0;
+	struct drm_gem_object *gem = NULL;
+	struct nouveau_bo *nvbo = NULL;
+	int ret = 0;
 
-	if (visible) {
+	if (handle) {
 		if (width != 64 || height != 64)
 			return -EINVAL;
 
@@ -1230,23 +1288,17 @@ nv50_crtc_cursor_set(struct drm_crtc *crtc, struct drm_file *file_priv,
 			return -ENOENT;
 		nvbo = nouveau_gem_object(gem);
 
-		ret = nouveau_bo_map(nvbo);
-		if (ret == 0) {
-			for (i = 0; i < 64 * 64; i++) {
-				u32 v = nouveau_bo_rd32(nvbo, i);
-				nouveau_bo_wr32(nv_crtc->cursor.nvbo, i, v);
-			}
-			nouveau_bo_unmap(nvbo);
-		}
-
-		drm_gem_object_unreference_unlocked(gem);
+		ret = nouveau_bo_pin(nvbo, TTM_PL_FLAG_VRAM, true);
 	}
 
-	if (visible != nv_crtc->cursor.visible) {
-		nv50_crtc_cursor_show_hide(nv_crtc, visible, true);
-		nv_crtc->cursor.visible = visible;
+	if (ret == 0) {
+		if (curs->image)
+			nouveau_bo_unpin(curs->image);
+		nouveau_bo_ref(nvbo, &curs->image);
 	}
+	drm_gem_object_unreference_unlocked(gem);
 
+	nv50_crtc_cursor_show_hide(nv_crtc, true, true);
 	return ret;
 }
 
@@ -1301,10 +1353,10 @@ nv50_crtc_destroy(struct drm_crtc *crtc)
 		nouveau_bo_unpin(head->image);
 	nouveau_bo_ref(NULL, &head->image);
 
-	nouveau_bo_unmap(nv_crtc->cursor.nvbo);
-	if (nv_crtc->cursor.nvbo)
-		nouveau_bo_unpin(nv_crtc->cursor.nvbo);
-	nouveau_bo_ref(NULL, &nv_crtc->cursor.nvbo);
+	/*XXX: ditto */
+	if (head->curs.image)
+		nouveau_bo_unpin(head->curs.image);
+	nouveau_bo_ref(NULL, &head->curs.image);
 
 	nouveau_bo_unmap(nv_crtc->lut.nvbo);
 	if (nv_crtc->lut.nvbo)
@@ -1336,16 +1388,6 @@ static const struct drm_crtc_funcs nv50_crtc_func = {
 	.page_flip = nouveau_crtc_page_flip,
 };
 
-static void
-nv50_cursor_set_pos(struct nouveau_crtc *nv_crtc, int x, int y)
-{
-}
-
-static void
-nv50_cursor_set_offset(struct nouveau_crtc *nv_crtc, uint32_t offset)
-{
-}
-
 static int
 nv50_crtc_create(struct drm_device *dev, int index)
 {
@@ -1364,8 +1406,6 @@ nv50_crtc_create(struct drm_device *dev, int index)
 	head->base.set_color_vibrance = nv50_crtc_set_color_vibrance;
 	head->base.color_vibrance = 50;
 	head->base.vibrant_hue = 0;
-	head->base.cursor.set_offset = nv50_cursor_set_offset;
-	head->base.cursor.set_pos = nv50_cursor_set_pos;
 	for (i = 0; i < 256; i++) {
 		head->base.lut.r[i] = i << 8;
 		head->base.lut.g[i] = i << 8;
@@ -1378,9 +1418,9 @@ nv50_crtc_create(struct drm_device *dev, int index)
 	drm_mode_crtc_set_gamma_size(crtc, 256);
 
 	ret = nouveau_bo_new(dev, 8192, 0x100, TTM_PL_FLAG_VRAM,
-			     0, 0x0000, NULL, &head->base.lut.nvbo);
+			     0, 0x0000, NULL, NULL, &head->base.lut.nvbo);
 	if (!ret) {
-		ret = nouveau_bo_pin(head->base.lut.nvbo, TTM_PL_FLAG_VRAM);
+		ret = nouveau_bo_pin(head->base.lut.nvbo, TTM_PL_FLAG_VRAM, true);
 		if (!ret) {
 			ret = nouveau_bo_map(head->base.lut.nvbo);
 			if (ret)
@@ -1400,22 +1440,6 @@ nv50_crtc_create(struct drm_device *dev, int index)
 	if (ret)
 		goto out;
 
-	ret = nouveau_bo_new(dev, 64 * 64 * 4, 0x100, TTM_PL_FLAG_VRAM,
-			     0, 0x0000, NULL, &head->base.cursor.nvbo);
-	if (!ret) {
-		ret = nouveau_bo_pin(head->base.cursor.nvbo, TTM_PL_FLAG_VRAM);
-		if (!ret) {
-			ret = nouveau_bo_map(head->base.cursor.nvbo);
-			if (ret)
-				nouveau_bo_unpin(head->base.lut.nvbo);
-		}
-		if (ret)
-			nouveau_bo_ref(NULL, &head->base.cursor.nvbo);
-	}
-
-	if (ret)
-		goto out;
-
 	/* allocate page flip / sync resources */
 	ret = nv50_base_create(disp->disp, index, disp->sync->bo.offset,
 			      &head->sync);
@@ -1651,17 +1675,21 @@ static void
 nv50_audio_mode_set(struct drm_encoder *encoder, struct drm_display_mode *mode)
 {
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
 	struct nouveau_connector *nv_connector;
 	struct nv50_disp *disp = nv50_disp(encoder->dev);
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_sor_hda_eld_v0 eld;
+	struct __packed {
+		struct {
+			struct nv50_disp_mthd_v1 mthd;
+			struct nv50_disp_sor_hda_eld_v0 eld;
+		} base;
 		u8 data[sizeof(nv_connector->base.eld)];
 	} args = {
-		.base.version = 1,
-		.base.method  = NV50_DISP_MTHD_V1_SOR_HDA_ELD,
-		.base.hasht   = nv_encoder->dcb->hasht,
-		.base.hashm   = nv_encoder->dcb->hashm,
+		.base.mthd.version = 1,
+		.base.mthd.method  = NV50_DISP_MTHD_V1_SOR_HDA_ELD,
+		.base.mthd.hasht   = nv_encoder->dcb->hasht,
+		.base.mthd.hashm   = (0xf0ff & nv_encoder->dcb->hashm) |
+				     (0x0100 << nv_crtc->index),
 	};
 
 	nv_connector = nouveau_encoder_connector_get(nv_encoder);
@@ -1671,11 +1699,12 @@ nv50_audio_mode_set(struct drm_encoder *encoder, struct drm_display_mode *mode)
 	drm_edid_to_eld(&nv_connector->base, nv_connector->edid);
 	memcpy(args.data, nv_connector->base.eld, sizeof(args.data));
 
-	nvif_mthd(disp->disp, 0, &args, sizeof(args));
+	nvif_mthd(disp->disp, 0, &args,
+		  sizeof(args.base) + drm_eld_size(args.data));
 }
 
 static void
-nv50_audio_disconnect(struct drm_encoder *encoder)
+nv50_audio_disconnect(struct drm_encoder *encoder, struct nouveau_crtc *nv_crtc)
 {
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
 	struct nv50_disp *disp = nv50_disp(encoder->dev);
@@ -1686,7 +1715,8 @@ nv50_audio_disconnect(struct drm_encoder *encoder)
 		.base.version = 1,
 		.base.method  = NV50_DISP_MTHD_V1_SOR_HDA_ELD,
 		.base.hasht   = nv_encoder->dcb->hasht,
-		.base.hashm   = nv_encoder->dcb->hashm,
+		.base.hashm   = (0xf0ff & nv_encoder->dcb->hashm) |
+				(0x0100 << nv_crtc->index),
 	};
 
 	nvif_mthd(disp->disp, 0, &args, sizeof(args));
@@ -1745,8 +1775,6 @@ nv50_hdmi_disconnect(struct drm_encoder *encoder, struct nouveau_crtc *nv_crtc)
 			       (0x0100 << nv_crtc->index),
 	};
 
-	nv50_audio_disconnect(encoder);
-
 	nvif_mthd(disp->disp, 0, &args, sizeof(args));
 }
 
@@ -1855,6 +1883,7 @@ nv50_sor_disconnect(struct drm_encoder *encoder)
 	if (nv_crtc) {
 		nv50_crtc_prepare(&nv_crtc->base);
 		nv50_sor_ctrl(nv_encoder, 1 << nv_crtc->index, 0);
+		nv50_audio_disconnect(encoder, nv_crtc);
 		nv50_hdmi_disconnect(&nv_encoder->base.base, nv_crtc);
 	}
 }
@@ -1954,6 +1983,7 @@ nv50_sor_mode_set(struct drm_encoder *encoder, struct drm_display_mode *umode,
 			proto = 0x8;
 		else
 			proto = 0x9;
+		nv50_audio_mode_set(encoder, mode);
 		break;
 	default:
 		BUG_ON(1);
@@ -2342,11 +2372,6 @@ nv50_fb_ctor(struct drm_framebuffer *fb)
 	u8 kind = nouveau_bo_tile_layout(nvbo) >> 8;
 	u8 tile = nvbo->tile_mode;
 
-	if (nvbo->tile_flags & NOUVEAU_GEM_TILE_NONCONTIG) {
-		NV_ERROR(drm, "framebuffer requires contiguous bo\n");
-		return -EINVAL;
-	}
-
 	if (drm->device.info.chipset >= 0xc0)
 		tile >>= 4; /* yep.. */
 
@@ -2458,9 +2483,9 @@ nv50_display_create(struct drm_device *dev)
 
 	/* small shared memory area we use for notifiers and semaphores */
 	ret = nouveau_bo_new(dev, 4096, 0x1000, TTM_PL_FLAG_VRAM,
-			     0, 0x0000, NULL, &disp->sync);
+			     0, 0x0000, NULL, NULL, &disp->sync);
 	if (!ret) {
-		ret = nouveau_bo_pin(disp->sync, TTM_PL_FLAG_VRAM);
+		ret = nouveau_bo_pin(disp->sync, TTM_PL_FLAG_VRAM, true);
 		if (!ret) {
 			ret = nouveau_bo_map(disp->sync);
 			if (ret)
diff --git a/drivers/gpu/drm/nouveau/nv50_fence.c b/drivers/gpu/drm/nouveau/nv50_fence.c
index 195cf51a7c31..a82d9ea7c6fd 100644
--- a/drivers/gpu/drm/nouveau/nv50_fence.c
+++ b/drivers/gpu/drm/nouveau/nv50_fence.c
@@ -46,7 +46,7 @@ nv50_fence_context_new(struct nouveau_channel *chan)
 	if (!fctx)
 		return -ENOMEM;
 
-	nouveau_fence_context_new(&fctx->base);
+	nouveau_fence_context_new(chan, &fctx->base);
 	fctx->base.emit = nv10_fence_emit;
 	fctx->base.read = nv10_fence_read;
 	fctx->base.sync = nv17_fence_sync;
@@ -95,12 +95,14 @@ nv50_fence_create(struct nouveau_drm *drm)
 	priv->base.resume = nv17_fence_resume;
 	priv->base.context_new = nv50_fence_context_new;
 	priv->base.context_del = nv10_fence_context_del;
+	priv->base.contexts = 127;
+	priv->base.context_base = fence_context_alloc(priv->base.contexts);
 	spin_lock_init(&priv->lock);
 
 	ret = nouveau_bo_new(drm->dev, 4096, 0x1000, TTM_PL_FLAG_VRAM,
-			     0, 0x0000, NULL, &priv->bo);
+			     0, 0x0000, NULL, NULL, &priv->bo);
 	if (!ret) {
-		ret = nouveau_bo_pin(priv->bo, TTM_PL_FLAG_VRAM);
+		ret = nouveau_bo_pin(priv->bo, TTM_PL_FLAG_VRAM, false);
 		if (!ret) {
 			ret = nouveau_bo_map(priv->bo);
 			if (ret)
diff --git a/drivers/gpu/drm/nouveau/nv84_fence.c b/drivers/gpu/drm/nouveau/nv84_fence.c
index 933a779c93ab..cb5b88938d45 100644
--- a/drivers/gpu/drm/nouveau/nv84_fence.c
+++ b/drivers/gpu/drm/nouveau/nv84_fence.c
@@ -82,7 +82,7 @@ nv84_fence_emit(struct nouveau_fence *fence)
 	else
 		addr += fctx->vma.offset;
 
-	return fctx->base.emit32(chan, addr, fence->sequence);
+	return fctx->base.emit32(chan, addr, fence->base.seqno);
 }
 
 static int
@@ -97,7 +97,7 @@ nv84_fence_sync(struct nouveau_fence *fence,
 	else
 		addr += fctx->vma.offset;
 
-	return fctx->base.sync32(chan, addr, fence->sequence);
+	return fctx->base.sync32(chan, addr, fence->base.seqno);
 }
 
 static u32
@@ -120,11 +120,12 @@ nv84_fence_context_del(struct nouveau_channel *chan)
 		nouveau_bo_vma_del(bo, &fctx->dispc_vma[i]);
 	}
 
+	nouveau_bo_wr32(priv->bo, chan->chid * 16 / 4, fctx->base.sequence);
 	nouveau_bo_vma_del(priv->bo, &fctx->vma_gart);
 	nouveau_bo_vma_del(priv->bo, &fctx->vma);
 	nouveau_fence_context_del(&fctx->base);
 	chan->fence = NULL;
-	kfree(fctx);
+	nouveau_fence_context_free(&fctx->base);
 }
 
 int
@@ -139,12 +140,13 @@ nv84_fence_context_new(struct nouveau_channel *chan)
 	if (!fctx)
 		return -ENOMEM;
 
-	nouveau_fence_context_new(&fctx->base);
+	nouveau_fence_context_new(chan, &fctx->base);
 	fctx->base.emit = nv84_fence_emit;
 	fctx->base.sync = nv84_fence_sync;
 	fctx->base.read = nv84_fence_read;
 	fctx->base.emit32 = nv84_fence_emit32;
 	fctx->base.sync32 = nv84_fence_sync32;
+	fctx->base.sequence = nv84_fence_read(chan);
 
 	ret = nouveau_bo_vma_add(priv->bo, cli->vm, &fctx->vma);
 	if (ret == 0) {
@@ -158,8 +160,6 @@ nv84_fence_context_new(struct nouveau_channel *chan)
 		ret = nouveau_bo_vma_add(bo, cli->vm, &fctx->dispc_vma[i]);
 	}
 
-	nouveau_bo_wr32(priv->bo, chan->chid * 16/4, 0x00000000);
-
 	if (ret)
 		nv84_fence_context_del(chan);
 	return ret;
@@ -168,13 +168,12 @@ nv84_fence_context_new(struct nouveau_channel *chan)
 static bool
 nv84_fence_suspend(struct nouveau_drm *drm)
 {
-	struct nouveau_fifo *pfifo = nvkm_fifo(&drm->device);
 	struct nv84_fence_priv *priv = drm->fence;
 	int i;
 
-	priv->suspend = vmalloc((pfifo->max + 1) * sizeof(u32));
+	priv->suspend = vmalloc(priv->base.contexts * sizeof(u32));
 	if (priv->suspend) {
-		for (i = 0; i <= pfifo->max; i++)
+		for (i = 0; i < priv->base.contexts; i++)
 			priv->suspend[i] = nouveau_bo_rd32(priv->bo, i*4);
 	}
 
@@ -184,12 +183,11 @@ nv84_fence_suspend(struct nouveau_drm *drm)
 static void
 nv84_fence_resume(struct nouveau_drm *drm)
 {
-	struct nouveau_fifo *pfifo = nvkm_fifo(&drm->device);
 	struct nv84_fence_priv *priv = drm->fence;
 	int i;
 
 	if (priv->suspend) {
-		for (i = 0; i <= pfifo->max; i++)
+		for (i = 0; i < priv->base.contexts; i++)
 			nouveau_bo_wr32(priv->bo, i*4, priv->suspend[i]);
 		vfree(priv->suspend);
 		priv->suspend = NULL;
@@ -229,13 +227,14 @@ nv84_fence_create(struct nouveau_drm *drm)
 	priv->base.context_new = nv84_fence_context_new;
 	priv->base.context_del = nv84_fence_context_del;
 
-	init_waitqueue_head(&priv->base.waiting);
+	priv->base.contexts = pfifo->max + 1;
+	priv->base.context_base = fence_context_alloc(priv->base.contexts);
 	priv->base.uevent = true;
 
-	ret = nouveau_bo_new(drm->dev, 16 * (pfifo->max + 1), 0,
-			     TTM_PL_FLAG_VRAM, 0, 0, NULL, &priv->bo);
+	ret = nouveau_bo_new(drm->dev, 16 * priv->base.contexts, 0,
+			     TTM_PL_FLAG_VRAM, 0, 0, NULL, NULL, &priv->bo);
 	if (ret == 0) {
-		ret = nouveau_bo_pin(priv->bo, TTM_PL_FLAG_VRAM);
+		ret = nouveau_bo_pin(priv->bo, TTM_PL_FLAG_VRAM, false);
 		if (ret == 0) {
 			ret = nouveau_bo_map(priv->bo);
 			if (ret)
@@ -246,11 +245,11 @@ nv84_fence_create(struct nouveau_drm *drm)
 	}
 
 	if (ret == 0)
-		ret = nouveau_bo_new(drm->dev, 16 * (pfifo->max + 1), 0,
-				     TTM_PL_FLAG_TT, 0, 0, NULL,
-				     &priv->bo_gart);
+		ret = nouveau_bo_new(drm->dev, 16 * priv->base.contexts, 0,
+				     TTM_PL_FLAG_TT | TTM_PL_FLAG_UNCACHED, 0,
+				     0, NULL, NULL, &priv->bo_gart);
 	if (ret == 0) {
-		ret = nouveau_bo_pin(priv->bo_gart, TTM_PL_FLAG_TT);
+		ret = nouveau_bo_pin(priv->bo_gart, TTM_PL_FLAG_TT, false);
 		if (ret == 0) {
 			ret = nouveau_bo_map(priv->bo_gart);
 			if (ret)
diff --git a/drivers/gpu/drm/nouveau/nvif/class.h b/drivers/gpu/drm/nouveau/nvif/class.h
index 573491f84792..4e308eacb27a 100644
--- a/drivers/gpu/drm/nouveau/nvif/class.h
+++ b/drivers/gpu/drm/nouveau/nvif/class.h
@@ -35,6 +35,7 @@
 #define GK104_DISP                                                   0x00009170
 #define GK110_DISP                                                   0x00009270
 #define GM107_DISP                                                   0x00009470
+#define GM204_DISP                                                   0x00009570
 
 #define NV50_DISP_CURSOR                                             0x0000507a
 #define G82_DISP_CURSOR                                              0x0000827a
@@ -65,6 +66,7 @@
 #define GK104_DISP_CORE_CHANNEL_DMA                                  0x0000917d
 #define GK110_DISP_CORE_CHANNEL_DMA                                  0x0000927d
 #define GM107_DISP_CORE_CHANNEL_DMA                                  0x0000947d
+#define GM204_DISP_CORE_CHANNEL_DMA                                  0x0000957d
 
 #define NV50_DISP_OVERLAY_CHANNEL_DMA                                0x0000507e
 #define G82_DISP_OVERLAY_CHANNEL_DMA                                 0x0000827e
@@ -131,6 +133,7 @@ struct nv_device_v0 {
 #define NV_DEVICE_V0_DISABLE_COPY1                        0x0000010000000000ULL
 #define NV_DEVICE_V0_DISABLE_VIC                          0x0000020000000000ULL
 #define NV_DEVICE_V0_DISABLE_VENC                         0x0000040000000000ULL
+#define NV_DEVICE_V0_DISABLE_COPY2                        0x0000080000000000ULL
 	__u64 disable;	/* disable particular subsystems */
 	__u64 debug0;	/* as above, but *internal* ids, and *NOT* ABI */
 };
@@ -479,6 +482,8 @@ struct nv50_disp_core_channel_dma_v0 {
 	__u32 pushbuf;
 };
 
+#define NV50_DISP_CORE_CHANNEL_DMA_V0_NTFY_UEVENT                          0x00
+
 /* cursor immediate */
 struct nv50_disp_cursor_v0 {
 	__u8  version;
@@ -486,6 +491,8 @@ struct nv50_disp_cursor_v0 {
 	__u8  pad02[6];
 };
 
+#define NV50_DISP_CURSOR_V0_NTFY_UEVENT                                    0x00
+
 /* base */
 struct nv50_disp_base_channel_dma_v0 {
 	__u8  version;
@@ -494,6 +501,8 @@ struct nv50_disp_base_channel_dma_v0 {
 	__u32 pushbuf;
 };
 
+#define NV50_DISP_BASE_CHANNEL_DMA_V0_NTFY_UEVENT                          0x00
+
 /* overlay */
 struct nv50_disp_overlay_channel_dma_v0 {
 	__u8  version;
@@ -502,6 +511,8 @@ struct nv50_disp_overlay_channel_dma_v0 {
 	__u32 pushbuf;
 };
 
+#define NV50_DISP_OVERLAY_CHANNEL_DMA_V0_NTFY_UEVENT                       0x00
+
 /* overlay immediate */
 struct nv50_disp_overlay_v0 {
 	__u8  version;
@@ -509,6 +520,7 @@ struct nv50_disp_overlay_v0 {
 	__u8  pad02[6];
 };
 
+#define NV50_DISP_OVERLAY_V0_NTFY_UEVENT                                   0x00
 
 /*******************************************************************************
  * fermi
diff --git a/drivers/gpu/drm/nouveau/nvif/client.c b/drivers/gpu/drm/nouveau/nvif/client.c
index 3c4df1fc26dc..3f7ac5bc8e03 100644
--- a/drivers/gpu/drm/nouveau/nvif/client.c
+++ b/drivers/gpu/drm/nouveau/nvif/client.c
@@ -62,6 +62,7 @@ nvif_drivers[] = {
 #else
 	&nvif_driver_drm,
 	&nvif_driver_lib,
+	&nvif_driver_null,
 #endif
 	NULL
 };
diff --git a/drivers/gpu/drm/nouveau/nvif/driver.h b/drivers/gpu/drm/nouveau/nvif/driver.h
index b72a8f0c2758..8bd39e69229c 100644
--- a/drivers/gpu/drm/nouveau/nvif/driver.h
+++ b/drivers/gpu/drm/nouveau/nvif/driver.h
@@ -9,13 +9,14 @@ struct nvif_driver {
 	int (*suspend)(void *priv);
 	int (*resume)(void *priv);
 	int (*ioctl)(void *priv, bool super, void *data, u32 size, void **hack);
-	void *(*map)(void *priv, u64 handle, u32 size);
-	void (*unmap)(void *priv, void *ptr, u32 size);
+	void __iomem *(*map)(void *priv, u64 handle, u32 size);
+	void (*unmap)(void *priv, void __iomem *ptr, u32 size);
 	bool keep;
 };
 
 extern const struct nvif_driver nvif_driver_nvkm;
 extern const struct nvif_driver nvif_driver_drm;
 extern const struct nvif_driver nvif_driver_lib;
+extern const struct nvif_driver nvif_driver_null;
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvif/object.h b/drivers/gpu/drm/nouveau/nvif/object.h
index fac3a3bbec44..fe519179b76c 100644
--- a/drivers/gpu/drm/nouveau/nvif/object.h
+++ b/drivers/gpu/drm/nouveau/nvif/object.h
@@ -14,7 +14,7 @@ struct nvif_object {
 	void *priv; /*XXX: hack */
 	void (*dtor)(struct nvif_object *);
 	struct {
-		void *ptr;
+		void __iomem *ptr;
 		u32 size;
 	} map;
 };
@@ -42,7 +42,7 @@ void nvif_object_unmap(struct nvif_object *);
 	struct nvif_object *_object = nvif_object(a);                          \
 	u32 _data;                                                             \
 	if (likely(_object->map.ptr))                                          \
-		_data = ioread##b##_native((u8 *)_object->map.ptr + (c));      \
+		_data = ioread##b##_native((u8 __iomem *)_object->map.ptr + (c));      \
 	else                                                                   \
 		_data = nvif_object_rd(_object, (b) / 8, (c));                 \
 	_data;                                                                 \
@@ -50,7 +50,7 @@ void nvif_object_unmap(struct nvif_object *);
 #define nvif_wr(a,b,c,d) ({                                                    \
 	struct nvif_object *_object = nvif_object(a);                          \
 	if (likely(_object->map.ptr))                                          \
-		iowrite##b##_native((d), (u8 *)_object->map.ptr + (c));        \
+		iowrite##b##_native((d), (u8 __iomem *)_object->map.ptr + (c));        \
 	else                                                                   \
 		nvif_object_wr(_object, (b) / 8, (c), (d));                    \
 })
diff --git a/drivers/gpu/drm/omapdrm/omap_crtc.c b/drivers/gpu/drm/omapdrm/omap_crtc.c
index 2d28dc337cfb..b0566a1ca28f 100644
--- a/drivers/gpu/drm/omapdrm/omap_crtc.c
+++ b/drivers/gpu/drm/omapdrm/omap_crtc.c
@@ -20,6 +20,7 @@
 #include "omap_drv.h"
 
 #include <drm/drm_mode.h>
+#include <drm/drm_plane_helper.h>
 #include "drm_crtc.h"
 #include "drm_crtc_helper.h"
 
diff --git a/drivers/gpu/drm/omapdrm/omap_drv.c b/drivers/gpu/drm/omapdrm/omap_drv.c
index 002b9721e85a..8241ed9b353c 100644
--- a/drivers/gpu/drm/omapdrm/omap_drv.c
+++ b/drivers/gpu/drm/omapdrm/omap_drv.c
@@ -629,6 +629,7 @@ static struct drm_driver omap_drm_driver = {
 		.lastclose = dev_lastclose,
 		.preclose = dev_preclose,
 		.postclose = dev_postclose,
+		.set_busid = drm_platform_set_busid,
 		.get_vblank_counter = drm_vblank_count,
 		.enable_vblank = omap_irq_enable_vblank,
 		.disable_vblank = omap_irq_disable_vblank,
@@ -717,7 +718,6 @@ static const struct dev_pm_ops omapdrm_pm_ops = {
 static struct platform_driver pdev = {
 		.driver = {
 			.name = DRIVER_NAME,
-			.owner = THIS_MODULE,
 #ifdef CONFIG_PM
 			.pm = &omapdrm_pm_ops,
 #endif
diff --git a/drivers/gpu/drm/omapdrm/omap_drv.h b/drivers/gpu/drm/omapdrm/omap_drv.h
index 84d73a61b34b..60e47b33c801 100644
--- a/drivers/gpu/drm/omapdrm/omap_drv.h
+++ b/drivers/gpu/drm/omapdrm/omap_drv.h
@@ -26,6 +26,7 @@
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
 #include <drm/omap_drm.h>
+#include <drm/drm_gem.h>
 #include <linux/platform_data/omap_drm.h>
 
 
diff --git a/drivers/gpu/drm/omapdrm/omap_gem.c b/drivers/gpu/drm/omapdrm/omap_gem.c
index e4849413ee80..aeb91ed653c9 100644
--- a/drivers/gpu/drm/omapdrm/omap_gem.c
+++ b/drivers/gpu/drm/omapdrm/omap_gem.c
@@ -612,8 +612,7 @@ int omap_gem_dumb_create(struct drm_file *file, struct drm_device *dev,
 {
 	union omap_gem_size gsize;
 
-	/* in case someone tries to feed us a completely bogus stride: */
-	args->pitch = align_pitch(args->pitch, args->width, args->bpp);
+	args->pitch = align_pitch(0, args->width, args->bpp);
 	args->size = PAGE_ALIGN(args->pitch * args->height);
 
 	gsize = (union omap_gem_size){
diff --git a/drivers/gpu/drm/omapdrm/omap_plane.c b/drivers/gpu/drm/omapdrm/omap_plane.c
index 891a4dc608af..ee8e2b3a117e 100644
--- a/drivers/gpu/drm/omapdrm/omap_plane.c
+++ b/drivers/gpu/drm/omapdrm/omap_plane.c
@@ -388,20 +388,15 @@ struct drm_plane *omap_plane_init(struct drm_device *dev,
 	struct drm_plane *plane = NULL;
 	struct omap_plane *omap_plane;
 	struct omap_overlay_info *info;
-	int ret;
 
 	DBG("%s: priv=%d", plane_names[id], private_plane);
 
 	omap_plane = kzalloc(sizeof(*omap_plane), GFP_KERNEL);
 	if (!omap_plane)
-		goto fail;
+		return NULL;
 
-	ret = drm_flip_work_init(&omap_plane->unpin_work, 16,
+	drm_flip_work_init(&omap_plane->unpin_work,
 			"unpin", unpin_worker);
-	if (ret) {
-		dev_err(dev->dev, "could not allocate unpin FIFO\n");
-		goto fail;
-	}
 
 	omap_plane->nformats = omap_framebuffer_get_formats(
 			omap_plane->formats, ARRAY_SIZE(omap_plane->formats),
@@ -443,10 +438,4 @@ struct drm_plane *omap_plane_init(struct drm_device *dev,
 		omap_plane->info.zorder = id;
 
 	return plane;
-
-fail:
-	if (plane)
-		omap_plane_destroy(plane);
-
-	return NULL;
 }
diff --git a/drivers/gpu/drm/panel/Kconfig b/drivers/gpu/drm/panel/Kconfig
index bee9f72b3a93..024e98ef8e4d 100644
--- a/drivers/gpu/drm/panel/Kconfig
+++ b/drivers/gpu/drm/panel/Kconfig
@@ -27,4 +27,17 @@ config DRM_PANEL_S6E8AA0
 	select DRM_MIPI_DSI
 	select VIDEOMODE_HELPERS
 
+config DRM_PANEL_SHARP_LQ101R1SX01
+	tristate "Sharp LQ101R1SX01 panel"
+	depends on OF
+	depends on DRM_MIPI_DSI
+	help
+	  Say Y here if you want to enable support for Sharp LQ101R1SX01
+	  TFT-LCD modules. The panel has a 2560x1600 resolution and uses
+	  24 bit RGB per pixel. It provides a dual MIPI DSI interface to
+	  the host and has a built-in LED backlight.
+
+	  To compile this driver as a module, choose M here: the module
+	  will be called panel-sharp-lq101r1sx01.
+
 endmenu
diff --git a/drivers/gpu/drm/panel/Makefile b/drivers/gpu/drm/panel/Makefile
index 8b929212fad7..4b2a0430804b 100644
--- a/drivers/gpu/drm/panel/Makefile
+++ b/drivers/gpu/drm/panel/Makefile
@@ -1,3 +1,4 @@
 obj-$(CONFIG_DRM_PANEL_SIMPLE) += panel-simple.o
 obj-$(CONFIG_DRM_PANEL_LD9040) += panel-ld9040.o
 obj-$(CONFIG_DRM_PANEL_S6E8AA0) += panel-s6e8aa0.o
+obj-$(CONFIG_DRM_PANEL_SHARP_LQ101R1SX01) += panel-sharp-lq101r1sx01.o
diff --git a/drivers/gpu/drm/panel/panel-ld9040.c b/drivers/gpu/drm/panel/panel-ld9040.c
index 42ac67b21e9f..08cf2c588c3d 100644
--- a/drivers/gpu/drm/panel/panel-ld9040.c
+++ b/drivers/gpu/drm/panel/panel-ld9040.c
@@ -145,7 +145,7 @@ static void ld9040_dcs_write(struct ld9040 *ctx, const u8 *data, size_t len)
 	if (ctx->error < 0 || len == 0)
 		return;
 
-	dev_dbg(ctx->dev, "writing dcs seq: %*ph\n", len, data);
+	dev_dbg(ctx->dev, "writing dcs seq: %*ph\n", (int)len, data);
 	ret = ld9040_spi_write_word(ctx, *data);
 
 	while (!ret && --len) {
@@ -154,8 +154,8 @@ static void ld9040_dcs_write(struct ld9040 *ctx, const u8 *data, size_t len)
 	}
 
 	if (ret) {
-		dev_err(ctx->dev, "error %d writing dcs seq: %*ph\n", ret, len,
-			data);
+		dev_err(ctx->dev, "error %d writing dcs seq: %*ph\n", ret,
+			(int)len, data);
 		ctx->error = ret;
 	}
 
@@ -336,17 +336,12 @@ static int ld9040_probe(struct spi_device *spi)
 	if (ret < 0)
 		return ret;
 
-	ctx->reset_gpio = devm_gpiod_get(dev, "reset");
+	ctx->reset_gpio = devm_gpiod_get(dev, "reset", GPIOD_OUT_HIGH);
 	if (IS_ERR(ctx->reset_gpio)) {
 		dev_err(dev, "cannot get reset-gpios %ld\n",
 			PTR_ERR(ctx->reset_gpio));
 		return PTR_ERR(ctx->reset_gpio);
 	}
-	ret = gpiod_direction_output(ctx->reset_gpio, 1);
-	if (ret < 0) {
-		dev_err(dev, "cannot configure reset-gpios %d\n", ret);
-		return ret;
-	}
 
 	spi->bits_per_word = 9;
 	ret = spi_setup(spi);
diff --git a/drivers/gpu/drm/panel/panel-s6e8aa0.c b/drivers/gpu/drm/panel/panel-s6e8aa0.c
index b5217fe37f02..144b2733e3d7 100644
--- a/drivers/gpu/drm/panel/panel-s6e8aa0.c
+++ b/drivers/gpu/drm/panel/panel-s6e8aa0.c
@@ -141,10 +141,10 @@ static void s6e8aa0_dcs_write(struct s6e8aa0 *ctx, const void *data, size_t len)
 	if (ctx->error < 0)
 		return;
 
-	ret = mipi_dsi_dcs_write(dsi, data, len);
+	ret = mipi_dsi_dcs_write_buffer(dsi, data, len);
 	if (ret < 0) {
-		dev_err(ctx->dev, "error %zd writing dcs seq: %*ph\n", ret, len,
-			data);
+		dev_err(ctx->dev, "error %zd writing dcs seq: %*ph\n", ret,
+			(int)len, data);
 		ctx->error = ret;
 	}
 }
@@ -800,27 +800,15 @@ static void s6e8aa0_panel_init(struct s6e8aa0 *ctx)
 }
 
 static void s6e8aa0_set_maximum_return_packet_size(struct s6e8aa0 *ctx,
-						   int size)
+						   u16 size)
 {
 	struct mipi_dsi_device *dsi = to_mipi_dsi_device(ctx->dev);
-	const struct mipi_dsi_host_ops *ops = dsi->host->ops;
-	u8 buf[] = {size, 0};
-	struct mipi_dsi_msg msg = {
-		.channel = dsi->channel,
-		.type = MIPI_DSI_SET_MAXIMUM_RETURN_PACKET_SIZE,
-		.tx_len = sizeof(buf),
-		.tx_buf = buf
-	};
 	int ret;
 
 	if (ctx->error < 0)
 		return;
 
-	if (!ops || !ops->transfer)
-		ret = -EIO;
-	else
-		ret = ops->transfer(dsi->host, &msg);
-
+	ret = mipi_dsi_set_maximum_return_packet_size(dsi, size);
 	if (ret < 0) {
 		dev_err(ctx->dev,
 			"error %d setting maximum return packet size to %d\n",
@@ -1019,17 +1007,12 @@ static int s6e8aa0_probe(struct mipi_dsi_device *dsi)
 		return ret;
 	}
 
-	ctx->reset_gpio = devm_gpiod_get(dev, "reset");
+	ctx->reset_gpio = devm_gpiod_get(dev, "reset", GPIOD_OUT_HIGH);
 	if (IS_ERR(ctx->reset_gpio)) {
 		dev_err(dev, "cannot get reset-gpios %ld\n",
 			PTR_ERR(ctx->reset_gpio));
 		return PTR_ERR(ctx->reset_gpio);
 	}
-	ret = gpiod_direction_output(ctx->reset_gpio, 1);
-	if (ret < 0) {
-		dev_err(dev, "cannot configure reset-gpios %d\n", ret);
-		return ret;
-	}
 
 	ctx->brightness = GAMMA_LEVEL_NUM - 1;
 
@@ -1069,7 +1052,6 @@ static struct mipi_dsi_driver s6e8aa0_driver = {
 	.remove = s6e8aa0_remove,
 	.driver = {
 		.name = "panel_s6e8aa0",
-		.owner = THIS_MODULE,
 		.of_match_table = s6e8aa0_of_match,
 	},
 };
diff --git a/drivers/gpu/drm/panel/panel-sharp-lq101r1sx01.c b/drivers/gpu/drm/panel/panel-sharp-lq101r1sx01.c
new file mode 100644
index 000000000000..9d81759d82fc
--- /dev/null
+++ b/drivers/gpu/drm/panel/panel-sharp-lq101r1sx01.c
@@ -0,0 +1,464 @@
+/*
+ * Copyright (C) 2014 NVIDIA Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/backlight.h>
+#include <linux/gpio/consumer.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/regulator/consumer.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_mipi_dsi.h>
+#include <drm/drm_panel.h>
+
+#include <video/mipi_display.h>
+
+#include <linux/host1x.h>
+
+struct sharp_panel {
+	struct drm_panel base;
+	/* the datasheet refers to them as DSI-LINK1 and DSI-LINK2 */
+	struct mipi_dsi_device *link1;
+	struct mipi_dsi_device *link2;
+
+	struct backlight_device *backlight;
+	struct regulator *supply;
+
+	bool prepared;
+	bool enabled;
+
+	const struct drm_display_mode *mode;
+};
+
+static inline struct sharp_panel *to_sharp_panel(struct drm_panel *panel)
+{
+	return container_of(panel, struct sharp_panel, base);
+}
+
+static int sharp_panel_write(struct sharp_panel *sharp, u16 offset, u8 value)
+{
+	u8 payload[3] = { offset >> 8, offset & 0xff, value };
+	struct mipi_dsi_device *dsi = sharp->link1;
+	ssize_t err;
+
+	err = mipi_dsi_generic_write(dsi, payload, sizeof(payload));
+	if (err < 0) {
+		dev_err(&dsi->dev, "failed to write %02x to %04x: %zd\n",
+			value, offset, err);
+		return err;
+	}
+
+	err = mipi_dsi_dcs_nop(dsi);
+	if (err < 0) {
+		dev_err(&dsi->dev, "failed to send DCS nop: %zd\n", err);
+		return err;
+	}
+
+	usleep_range(10, 20);
+
+	return 0;
+}
+
+static __maybe_unused int sharp_panel_read(struct sharp_panel *sharp,
+					   u16 offset, u8 *value)
+{
+	ssize_t err;
+
+	cpu_to_be16s(&offset);
+
+	err = mipi_dsi_generic_read(sharp->link1, &offset, sizeof(offset),
+				    value, sizeof(*value));
+	if (err < 0)
+		dev_err(&sharp->link1->dev, "failed to read from %04x: %zd\n",
+			offset, err);
+
+	return err;
+}
+
+static int sharp_panel_disable(struct drm_panel *panel)
+{
+	struct sharp_panel *sharp = to_sharp_panel(panel);
+
+	if (!sharp->enabled)
+		return 0;
+
+	if (sharp->backlight) {
+		sharp->backlight->props.power = FB_BLANK_POWERDOWN;
+		backlight_update_status(sharp->backlight);
+	}
+
+	sharp->enabled = false;
+
+	return 0;
+}
+
+static int sharp_panel_unprepare(struct drm_panel *panel)
+{
+	struct sharp_panel *sharp = to_sharp_panel(panel);
+	int err;
+
+	if (!sharp->prepared)
+		return 0;
+
+	err = mipi_dsi_dcs_set_display_off(sharp->link1);
+	if (err < 0)
+		dev_err(panel->dev, "failed to set display off: %d\n", err);
+
+	err = mipi_dsi_dcs_enter_sleep_mode(sharp->link1);
+	if (err < 0)
+		dev_err(panel->dev, "failed to enter sleep mode: %d\n", err);
+
+	msleep(120);
+
+	regulator_disable(sharp->supply);
+
+	sharp->prepared = false;
+
+	return 0;
+}
+
+static int sharp_setup_symmetrical_split(struct mipi_dsi_device *left,
+					 struct mipi_dsi_device *right,
+					 const struct drm_display_mode *mode)
+{
+	int err;
+
+	err = mipi_dsi_dcs_set_column_address(left, 0, mode->hdisplay / 2 - 1);
+	if (err < 0) {
+		dev_err(&left->dev, "failed to set column address: %d\n", err);
+		return err;
+	}
+
+	err = mipi_dsi_dcs_set_page_address(left, 0, mode->vdisplay - 1);
+	if (err < 0) {
+		dev_err(&left->dev, "failed to set page address: %d\n", err);
+		return err;
+	}
+
+	err = mipi_dsi_dcs_set_column_address(right, mode->hdisplay / 2,
+					      mode->hdisplay - 1);
+	if (err < 0) {
+		dev_err(&right->dev, "failed to set column address: %d\n", err);
+		return err;
+	}
+
+	err = mipi_dsi_dcs_set_page_address(right, 0, mode->vdisplay - 1);
+	if (err < 0) {
+		dev_err(&right->dev, "failed to set page address: %d\n", err);
+		return err;
+	}
+
+	return 0;
+}
+
+static int sharp_panel_prepare(struct drm_panel *panel)
+{
+	struct sharp_panel *sharp = to_sharp_panel(panel);
+	u8 format = MIPI_DCS_PIXEL_FMT_24BIT;
+	int err;
+
+	if (sharp->prepared)
+		return 0;
+
+	err = regulator_enable(sharp->supply);
+	if (err < 0)
+		return err;
+
+	usleep_range(10000, 20000);
+
+	err = mipi_dsi_dcs_soft_reset(sharp->link1);
+	if (err < 0) {
+		dev_err(panel->dev, "soft reset failed: %d\n", err);
+		goto poweroff;
+	}
+
+	msleep(120);
+
+	err = mipi_dsi_dcs_exit_sleep_mode(sharp->link1);
+	if (err < 0) {
+		dev_err(panel->dev, "failed to exit sleep mode: %d\n", err);
+		goto poweroff;
+	}
+
+	/*
+	 * The MIPI DCS specification mandates this delay only between the
+	 * exit_sleep_mode and enter_sleep_mode commands, so it isn't strictly
+	 * necessary here.
+	 */
+	/*
+	msleep(120);
+	*/
+
+	/* set left-right mode */
+	err = sharp_panel_write(sharp, 0x1000, 0x2a);
+	if (err < 0) {
+		dev_err(panel->dev, "failed to set left-right mode: %d\n", err);
+		goto poweroff;
+	}
+
+	/* enable command mode */
+	err = sharp_panel_write(sharp, 0x1001, 0x01);
+	if (err < 0) {
+		dev_err(panel->dev, "failed to enable command mode: %d\n", err);
+		goto poweroff;
+	}
+
+	err = mipi_dsi_dcs_set_pixel_format(sharp->link1, format);
+	if (err < 0) {
+		dev_err(panel->dev, "failed to set pixel format: %d\n", err);
+		goto poweroff;
+	}
+
+	/*
+	 * TODO: The device supports both left-right and even-odd split
+	 * configurations, but this driver currently supports only the left-
+	 * right split. To support a different mode a mechanism needs to be
+	 * put in place to communicate the configuration back to the DSI host
+	 * controller.
+	 */
+	err = sharp_setup_symmetrical_split(sharp->link1, sharp->link2,
+					    sharp->mode);
+	if (err < 0) {
+		dev_err(panel->dev, "failed to set up symmetrical split: %d\n",
+			err);
+		goto poweroff;
+	}
+
+	err = mipi_dsi_dcs_set_display_on(sharp->link1);
+	if (err < 0) {
+		dev_err(panel->dev, "failed to set display on: %d\n", err);
+		goto poweroff;
+	}
+
+	sharp->prepared = true;
+
+	return 0;
+
+poweroff:
+	regulator_disable(sharp->supply);
+	return err;
+}
+
+static int sharp_panel_enable(struct drm_panel *panel)
+{
+	struct sharp_panel *sharp = to_sharp_panel(panel);
+
+	if (sharp->enabled)
+		return 0;
+
+	if (sharp->backlight) {
+		sharp->backlight->props.power = FB_BLANK_UNBLANK;
+		backlight_update_status(sharp->backlight);
+	}
+
+	sharp->enabled = true;
+
+	return 0;
+}
+
+static const struct drm_display_mode default_mode = {
+	.clock = 278000,
+	.hdisplay = 2560,
+	.hsync_start = 2560 + 128,
+	.hsync_end = 2560 + 128 + 64,
+	.htotal = 2560 + 128 + 64 + 64,
+	.vdisplay = 1600,
+	.vsync_start = 1600 + 4,
+	.vsync_end = 1600 + 4 + 8,
+	.vtotal = 1600 + 4 + 8 + 32,
+	.vrefresh = 60,
+};
+
+static int sharp_panel_get_modes(struct drm_panel *panel)
+{
+	struct drm_display_mode *mode;
+
+	mode = drm_mode_duplicate(panel->drm, &default_mode);
+	if (!mode) {
+		dev_err(panel->drm->dev, "failed to add mode %ux%ux@%u\n",
+			default_mode.hdisplay, default_mode.vdisplay,
+			default_mode.vrefresh);
+		return -ENOMEM;
+	}
+
+	drm_mode_set_name(mode);
+
+	drm_mode_probed_add(panel->connector, mode);
+
+	panel->connector->display_info.width_mm = 217;
+	panel->connector->display_info.height_mm = 136;
+
+	return 1;
+}
+
+static const struct drm_panel_funcs sharp_panel_funcs = {
+	.disable = sharp_panel_disable,
+	.unprepare = sharp_panel_unprepare,
+	.prepare = sharp_panel_prepare,
+	.enable = sharp_panel_enable,
+	.get_modes = sharp_panel_get_modes,
+};
+
+static const struct of_device_id sharp_of_match[] = {
+	{ .compatible = "sharp,lq101r1sx01", },
+	{ }
+};
+MODULE_DEVICE_TABLE(of, sharp_of_match);
+
+static int sharp_panel_add(struct sharp_panel *sharp)
+{
+	struct device_node *np;
+	int err;
+
+	sharp->mode = &default_mode;
+
+	sharp->supply = devm_regulator_get(&sharp->link1->dev, "power");
+	if (IS_ERR(sharp->supply))
+		return PTR_ERR(sharp->supply);
+
+	np = of_parse_phandle(sharp->link1->dev.of_node, "backlight", 0);
+	if (np) {
+		sharp->backlight = of_find_backlight_by_node(np);
+		of_node_put(np);
+
+		if (!sharp->backlight)
+			return -EPROBE_DEFER;
+	}
+
+	drm_panel_init(&sharp->base);
+	sharp->base.funcs = &sharp_panel_funcs;
+	sharp->base.dev = &sharp->link1->dev;
+
+	err = drm_panel_add(&sharp->base);
+	if (err < 0)
+		goto put_backlight;
+
+	return 0;
+
+put_backlight:
+	if (sharp->backlight)
+		put_device(&sharp->backlight->dev);
+
+	return err;
+}
+
+static void sharp_panel_del(struct sharp_panel *sharp)
+{
+	if (sharp->base.dev)
+		drm_panel_remove(&sharp->base);
+
+	if (sharp->backlight)
+		put_device(&sharp->backlight->dev);
+
+	if (sharp->link2)
+		put_device(&sharp->link2->dev);
+}
+
+static int sharp_panel_probe(struct mipi_dsi_device *dsi)
+{
+	struct mipi_dsi_device *secondary = NULL;
+	struct sharp_panel *sharp;
+	struct device_node *np;
+	int err;
+
+	dsi->lanes = 4;
+	dsi->format = MIPI_DSI_FMT_RGB888;
+	dsi->mode_flags = MIPI_DSI_MODE_LPM;
+
+	/* Find DSI-LINK1 */
+	np = of_parse_phandle(dsi->dev.of_node, "link2", 0);
+	if (np) {
+		secondary = of_find_mipi_dsi_device_by_node(np);
+		of_node_put(np);
+
+		if (!secondary)
+			return -EPROBE_DEFER;
+	}
+
+	/* register a panel for only the DSI-LINK1 interface */
+	if (secondary) {
+		sharp = devm_kzalloc(&dsi->dev, sizeof(*sharp), GFP_KERNEL);
+		if (!sharp) {
+			put_device(&secondary->dev);
+			return -ENOMEM;
+		}
+
+		mipi_dsi_set_drvdata(dsi, sharp);
+
+		sharp->link2 = secondary;
+		sharp->link1 = dsi;
+
+		err = sharp_panel_add(sharp);
+		if (err < 0) {
+			put_device(&secondary->dev);
+			return err;
+		}
+	}
+
+	err = mipi_dsi_attach(dsi);
+	if (err < 0) {
+		if (secondary)
+			sharp_panel_del(sharp);
+
+		return err;
+	}
+
+	return 0;
+}
+
+static int sharp_panel_remove(struct mipi_dsi_device *dsi)
+{
+	struct sharp_panel *sharp = mipi_dsi_get_drvdata(dsi);
+	int err;
+
+	/* only detach from host for the DSI-LINK2 interface */
+	if (!sharp) {
+		mipi_dsi_detach(dsi);
+		return 0;
+	}
+
+	err = sharp_panel_disable(&sharp->base);
+	if (err < 0)
+		dev_err(&dsi->dev, "failed to disable panel: %d\n", err);
+
+	err = mipi_dsi_detach(dsi);
+	if (err < 0)
+		dev_err(&dsi->dev, "failed to detach from DSI host: %d\n", err);
+
+	drm_panel_detach(&sharp->base);
+	sharp_panel_del(sharp);
+
+	return 0;
+}
+
+static void sharp_panel_shutdown(struct mipi_dsi_device *dsi)
+{
+	struct sharp_panel *sharp = mipi_dsi_get_drvdata(dsi);
+
+	/* nothing to do for DSI-LINK2 */
+	if (!sharp)
+		return;
+
+	sharp_panel_disable(&sharp->base);
+}
+
+static struct mipi_dsi_driver sharp_panel_driver = {
+	.driver = {
+		.name = "panel-sharp-lq101r1sx01",
+		.of_match_table = sharp_of_match,
+	},
+	.probe = sharp_panel_probe,
+	.remove = sharp_panel_remove,
+	.shutdown = sharp_panel_shutdown,
+};
+module_mipi_dsi_driver(sharp_panel_driver);
+
+MODULE_AUTHOR("Thierry Reding <treding@nvidia.com>");
+MODULE_DESCRIPTION("Sharp LQ101R1SX01 panel driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpu/drm/panel/panel-simple.c b/drivers/gpu/drm/panel/panel-simple.c
index 4ce1db0a68ff..e95385bf8356 100644
--- a/drivers/gpu/drm/panel/panel-simple.c
+++ b/drivers/gpu/drm/panel/panel-simple.c
@@ -247,21 +247,14 @@ static int panel_simple_probe(struct device *dev, const struct panel_desc *desc)
 	if (IS_ERR(panel->supply))
 		return PTR_ERR(panel->supply);
 
-	panel->enable_gpio = devm_gpiod_get_optional(dev, "enable");
+	panel->enable_gpio = devm_gpiod_get_optional(dev, "enable",
+						     GPIOD_OUT_LOW);
 	if (IS_ERR(panel->enable_gpio)) {
 		err = PTR_ERR(panel->enable_gpio);
 		dev_err(dev, "failed to request GPIO: %d\n", err);
 		return err;
 	}
 
-	if (panel->enable_gpio) {
-		err = gpiod_direction_output(panel->enable_gpio, 0);
-		if (err < 0) {
-			dev_err(dev, "failed to setup GPIO: %d\n", err);
-			return err;
-		}
-	}
-
 	backlight = of_parse_phandle(dev->of_node, "backlight", 0);
 	if (backlight) {
 		panel->backlight = of_find_backlight_by_node(backlight);
@@ -352,6 +345,53 @@ static const struct panel_desc auo_b101aw03 = {
 	},
 };
 
+static const struct drm_display_mode auo_b101xtn01_mode = {
+	.clock = 72000,
+	.hdisplay = 1366,
+	.hsync_start = 1366 + 20,
+	.hsync_end = 1366 + 20 + 70,
+	.htotal = 1366 + 20 + 70,
+	.vdisplay = 768,
+	.vsync_start = 768 + 14,
+	.vsync_end = 768 + 14 + 42,
+	.vtotal = 768 + 14 + 42,
+	.vrefresh = 60,
+	.flags = DRM_MODE_FLAG_NVSYNC | DRM_MODE_FLAG_NHSYNC,
+};
+
+static const struct panel_desc auo_b101xtn01 = {
+	.modes = &auo_b101xtn01_mode,
+	.num_modes = 1,
+	.bpc = 6,
+	.size = {
+		.width = 223,
+		.height = 125,
+	},
+};
+
+static const struct drm_display_mode auo_b116xw03_mode = {
+	.clock = 70589,
+	.hdisplay = 1366,
+	.hsync_start = 1366 + 40,
+	.hsync_end = 1366 + 40 + 40,
+	.htotal = 1366 + 40 + 40 + 32,
+	.vdisplay = 768,
+	.vsync_start = 768 + 10,
+	.vsync_end = 768 + 10 + 12,
+	.vtotal = 768 + 10 + 12 + 6,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc auo_b116xw03 = {
+	.modes = &auo_b116xw03_mode,
+	.num_modes = 1,
+	.bpc = 6,
+	.size = {
+		.width = 256,
+		.height = 144,
+	},
+};
+
 static const struct drm_display_mode auo_b133xtn01_mode = {
 	.clock = 69500,
 	.hdisplay = 1366,
@@ -391,6 +431,7 @@ static const struct drm_display_mode auo_b133htn01_mode = {
 static const struct panel_desc auo_b133htn01 = {
 	.modes = &auo_b133htn01_mode,
 	.num_modes = 1,
+	.bpc = 6,
 	.size = {
 		.width = 293,
 		.height = 165,
@@ -512,22 +553,92 @@ static const struct drm_display_mode foxlink_fl500wvr00_a0t_mode = {
 static const struct panel_desc foxlink_fl500wvr00_a0t = {
 	.modes = &foxlink_fl500wvr00_a0t_mode,
 	.num_modes = 1,
+	.bpc = 8,
 	.size = {
 		.width = 108,
 		.height = 65,
 	},
 };
 
-static const struct drm_display_mode innolux_n116bge_mode = {
+static const struct drm_display_mode hannstar_hsd070pww1_mode = {
+	.clock = 71100,
+	.hdisplay = 1280,
+	.hsync_start = 1280 + 1,
+	.hsync_end = 1280 + 1 + 158,
+	.htotal = 1280 + 1 + 158 + 1,
+	.vdisplay = 800,
+	.vsync_start = 800 + 1,
+	.vsync_end = 800 + 1 + 21,
+	.vtotal = 800 + 1 + 21 + 1,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc hannstar_hsd070pww1 = {
+	.modes = &hannstar_hsd070pww1_mode,
+	.num_modes = 1,
+	.bpc = 6,
+	.size = {
+		.width = 151,
+		.height = 94,
+	},
+};
+
+static const struct drm_display_mode hitachi_tx23d38vm0caa_mode = {
+	.clock = 33333,
+	.hdisplay = 800,
+	.hsync_start = 800 + 85,
+	.hsync_end = 800 + 85 + 86,
+	.htotal = 800 + 85 + 86 + 85,
+	.vdisplay = 480,
+	.vsync_start = 480 + 16,
+	.vsync_end = 480 + 16 + 13,
+	.vtotal = 480 + 16 + 13 + 16,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc hitachi_tx23d38vm0caa = {
+	.modes = &hitachi_tx23d38vm0caa_mode,
+	.num_modes = 1,
+	.bpc = 6,
+	.size = {
+		.width = 195,
+		.height = 117,
+	},
+};
+
+static const struct drm_display_mode innolux_g121i1_l01_mode = {
 	.clock = 71000,
+	.hdisplay = 1280,
+	.hsync_start = 1280 + 64,
+	.hsync_end = 1280 + 64 + 32,
+	.htotal = 1280 + 64 + 32 + 64,
+	.vdisplay = 800,
+	.vsync_start = 800 + 9,
+	.vsync_end = 800 + 9 + 6,
+	.vtotal = 800 + 9 + 6 + 9,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc innolux_g121i1_l01 = {
+	.modes = &innolux_g121i1_l01_mode,
+	.num_modes = 1,
+	.bpc = 6,
+	.size = {
+		.width = 261,
+		.height = 163,
+	},
+};
+
+static const struct drm_display_mode innolux_n116bge_mode = {
+	.clock = 76420,
 	.hdisplay = 1366,
-	.hsync_start = 1366 + 64,
-	.hsync_end = 1366 + 64 + 6,
-	.htotal = 1366 + 64 + 6 + 64,
+	.hsync_start = 1366 + 136,
+	.hsync_end = 1366 + 136 + 30,
+	.htotal = 1366 + 136 + 30 + 60,
 	.vdisplay = 768,
 	.vsync_start = 768 + 8,
-	.vsync_end = 768 + 8 + 4,
-	.vtotal = 768 + 8 + 4 + 8,
+	.vsync_end = 768 + 8 + 12,
+	.vtotal = 768 + 8 + 12 + 12,
 	.vrefresh = 60,
 	.flags = DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC,
 };
@@ -616,6 +727,12 @@ static const struct of_device_id platform_of_match[] = {
 		.compatible = "auo,b101aw03",
 		.data = &auo_b101aw03,
 	}, {
+		.compatible = "auo,b101xtn01",
+		.data = &auo_b101xtn01,
+	}, {
+		.compatible = "auo,b116xw03",
+		.data = &auo_b116xw03,
+	}, {
 		.compatible = "auo,b133htn01",
 		.data = &auo_b133htn01,
 	}, {
@@ -640,6 +757,15 @@ static const struct of_device_id platform_of_match[] = {
 		.compatible = "foxlink,fl500wvr00-a0t",
 		.data = &foxlink_fl500wvr00_a0t,
 	}, {
+		.compatible = "hannstar,hsd070pww1",
+		.data = &hannstar_hsd070pww1,
+	}, {
+		.compatible = "hit,tx23d38vm0caa",
+		.data = &hitachi_tx23d38vm0caa
+	}, {
+		.compatible ="innolux,g121i1-l01",
+		.data = &innolux_g121i1_l01
+	}, {
 		.compatible = "innolux,n116bge",
 		.data = &innolux_n116bge,
 	}, {
@@ -681,7 +807,6 @@ static void panel_simple_platform_shutdown(struct platform_device *pdev)
 static struct platform_driver panel_simple_platform_driver = {
 	.driver = {
 		.name = "panel-simple",
-		.owner = THIS_MODULE,
 		.of_match_table = platform_of_match,
 	},
 	.probe = panel_simple_platform_probe,
@@ -714,6 +839,7 @@ static const struct panel_desc_dsi lg_ld070wx3_sl01 = {
 	.desc = {
 		.modes = &lg_ld070wx3_sl01_mode,
 		.num_modes = 1,
+		.bpc = 8,
 		.size = {
 			.width = 94,
 			.height = 151,
@@ -741,6 +867,7 @@ static const struct panel_desc_dsi lg_lh500wx1_sd03 = {
 	.desc = {
 		.modes = &lg_lh500wx1_sd03_mode,
 		.num_modes = 1,
+		.bpc = 8,
 		.size = {
 			.width = 62,
 			.height = 110,
@@ -768,6 +895,7 @@ static const struct panel_desc_dsi panasonic_vvx10f004b00 = {
 	.desc = {
 		.modes = &panasonic_vvx10f004b00_mode,
 		.num_modes = 1,
+		.bpc = 8,
 		.size = {
 			.width = 217,
 			.height = 136,
@@ -837,7 +965,6 @@ static void panel_simple_dsi_shutdown(struct mipi_dsi_device *dsi)
 static struct mipi_dsi_driver panel_simple_dsi_driver = {
 	.driver = {
 		.name = "panel-simple-dsi",
-		.owner = THIS_MODULE,
 		.of_match_table = dsi_of_match,
 	},
 	.probe = panel_simple_dsi_probe,
diff --git a/drivers/gpu/drm/qxl/Makefile b/drivers/gpu/drm/qxl/Makefile
index ea046ba691d2..bacc4aff1201 100644
--- a/drivers/gpu/drm/qxl/Makefile
+++ b/drivers/gpu/drm/qxl/Makefile
@@ -4,6 +4,6 @@
 
 ccflags-y := -Iinclude/drm
 
-qxl-y := qxl_drv.o qxl_kms.o qxl_display.o qxl_ttm.o qxl_fb.o qxl_object.o qxl_gem.o qxl_cmd.o qxl_image.o qxl_draw.o qxl_debugfs.o qxl_irq.o qxl_dumb.o qxl_ioctl.o qxl_fence.o qxl_release.o
+qxl-y := qxl_drv.o qxl_kms.o qxl_display.o qxl_ttm.o qxl_fb.o qxl_object.o qxl_gem.o qxl_cmd.o qxl_image.o qxl_draw.o qxl_debugfs.o qxl_irq.o qxl_dumb.o qxl_ioctl.o qxl_release.o qxl_prime.o
 
 obj-$(CONFIG_DRM_QXL)+= qxl.o
diff --git a/drivers/gpu/drm/qxl/qxl_cmd.c b/drivers/gpu/drm/qxl/qxl_cmd.c
index eb89653a7a17..97823644d347 100644
--- a/drivers/gpu/drm/qxl/qxl_cmd.c
+++ b/drivers/gpu/drm/qxl/qxl_cmd.c
@@ -620,17 +620,10 @@ static int qxl_reap_surf(struct qxl_device *qdev, struct qxl_bo *surf, bool stal
 	if (ret == -EBUSY)
 		return -EBUSY;
 
-	if (surf->fence.num_active_releases > 0 && stall == false) {
-		qxl_bo_unreserve(surf);
-		return -EBUSY;
-	}
-
 	if (stall)
 		mutex_unlock(&qdev->surf_evict_mutex);
 
-	spin_lock(&surf->tbo.bdev->fence_lock);
 	ret = ttm_bo_wait(&surf->tbo, true, true, !stall);
-	spin_unlock(&surf->tbo.bdev->fence_lock);
 
 	if (stall)
 		mutex_lock(&qdev->surf_evict_mutex);
diff --git a/drivers/gpu/drm/qxl/qxl_debugfs.c b/drivers/gpu/drm/qxl/qxl_debugfs.c
index c3c2bbdc6674..6911b8c44492 100644
--- a/drivers/gpu/drm/qxl/qxl_debugfs.c
+++ b/drivers/gpu/drm/qxl/qxl_debugfs.c
@@ -58,9 +58,17 @@ qxl_debugfs_buffers_info(struct seq_file *m, void *data)
 	struct qxl_bo *bo;
 
 	list_for_each_entry(bo, &qdev->gem.objects, list) {
-		seq_printf(m, "size %ld, pc %d, sync obj %p, num releases %d\n",
-			   (unsigned long)bo->gem_base.size, bo->pin_count,
-			   bo->tbo.sync_obj, bo->fence.num_active_releases);
+		struct reservation_object_list *fobj;
+		int rel;
+
+		rcu_read_lock();
+		fobj = rcu_dereference(bo->tbo.resv->fence);
+		rel = fobj ? fobj->shared_count : 0;
+		rcu_read_unlock();
+
+		seq_printf(m, "size %ld, pc %d, num releases %d\n",
+			   (unsigned long)bo->gem_base.size,
+			   bo->pin_count, rel);
 	}
 	return 0;
 }
diff --git a/drivers/gpu/drm/qxl/qxl_display.c b/drivers/gpu/drm/qxl/qxl_display.c
index b8ced08b6291..4a0a8b29b0a1 100644
--- a/drivers/gpu/drm/qxl/qxl_display.c
+++ b/drivers/gpu/drm/qxl/qxl_display.c
@@ -29,6 +29,7 @@
 #include "qxl_drv.h"
 #include "qxl_object.h"
 #include "drm_crtc_helper.h"
+#include <drm/drm_plane_helper.h>
 
 static bool qxl_head_enabled(struct qxl_head *head)
 {
@@ -100,14 +101,37 @@ static int qxl_display_copy_rom_client_monitors_config(struct qxl_device *qdev)
 	return 0;
 }
 
+static void qxl_update_offset_props(struct qxl_device *qdev)
+{
+	struct drm_device *dev = qdev->ddev;
+	struct drm_connector *connector;
+	struct qxl_output *output;
+	struct qxl_head *head;
+
+	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+		output = drm_connector_to_qxl_output(connector);
+
+		head = &qdev->client_monitors_config->heads[output->index];
+
+		drm_object_property_set_value(&connector->base,
+			dev->mode_config.suggested_x_property, head->x);
+		drm_object_property_set_value(&connector->base,
+			dev->mode_config.suggested_y_property, head->y);
+	}
+}
+
 void qxl_display_read_client_monitors_config(struct qxl_device *qdev)
 {
 
+	struct drm_device *dev = qdev->ddev;
 	while (qxl_display_copy_rom_client_monitors_config(qdev)) {
 		qxl_io_log(qdev, "failed crc check for client_monitors_config,"
 				 " retrying\n");
 	}
 
+	drm_modeset_lock_all(dev);
+	qxl_update_offset_props(qdev);
+	drm_modeset_unlock_all(dev);
 	if (!drm_helper_hpd_irq_event(qdev->ddev)) {
 		/* notify that the monitor configuration changed, to
 		   adjust at the arbitrary resolution */
@@ -187,6 +211,54 @@ static void qxl_crtc_destroy(struct drm_crtc *crtc)
 	kfree(qxl_crtc);
 }
 
+static int qxl_crtc_page_flip(struct drm_crtc *crtc,
+                              struct drm_framebuffer *fb,
+                              struct drm_pending_vblank_event *event,
+                              uint32_t page_flip_flags)
+{
+	struct drm_device *dev = crtc->dev;
+	struct qxl_device *qdev = dev->dev_private;
+	struct qxl_crtc *qcrtc = to_qxl_crtc(crtc);
+	struct qxl_framebuffer *qfb_src = to_qxl_framebuffer(fb);
+	struct qxl_framebuffer *qfb_old = to_qxl_framebuffer(crtc->primary->fb);
+	struct qxl_bo *bo_old = gem_to_qxl_bo(qfb_old->obj);
+	struct qxl_bo *bo = gem_to_qxl_bo(qfb_src->obj);
+	unsigned long flags;
+	struct drm_clip_rect norect = {
+	    .x1 = 0,
+	    .y1 = 0,
+	    .x2 = fb->width,
+	    .y2 = fb->height
+	};
+	int inc = 1;
+	int one_clip_rect = 1;
+	int ret = 0;
+
+	crtc->primary->fb = fb;
+	bo_old->is_primary = false;
+	bo->is_primary = true;
+
+	ret = qxl_bo_reserve(bo, false);
+	if (ret)
+		return ret;
+
+	qxl_draw_dirty_fb(qdev, qfb_src, bo, 0, 0,
+			  &norect, one_clip_rect, inc);
+
+	drm_vblank_get(dev, qcrtc->index);
+
+	if (event) {
+		spin_lock_irqsave(&dev->event_lock, flags);
+		drm_send_vblank_event(dev, qcrtc->index, event);
+		spin_unlock_irqrestore(&dev->event_lock, flags);
+	}
+	drm_vblank_put(dev, qcrtc->index);
+
+	qxl_bo_unreserve(bo);
+
+	return 0;
+}
+
 static int
 qxl_hide_cursor(struct qxl_device *qdev)
 {
@@ -374,6 +446,7 @@ static const struct drm_crtc_funcs qxl_crtc_funcs = {
 	.cursor_move = qxl_crtc_cursor_move,
 	.set_config = drm_crtc_helper_set_config,
 	.destroy = qxl_crtc_destroy,
+	.page_flip = qxl_crtc_page_flip,
 };
 
 static void qxl_user_framebuffer_destroy(struct drm_framebuffer *fb)
@@ -519,11 +592,9 @@ static int qxl_crtc_mode_set(struct drm_crtc *crtc,
 {
 	struct drm_device *dev = crtc->dev;
 	struct qxl_device *qdev = dev->dev_private;
-	struct qxl_mode *m = (void *)mode->private;
 	struct qxl_framebuffer *qfb;
 	struct qxl_bo *bo, *old_bo = NULL;
 	struct qxl_crtc *qcrtc = to_qxl_crtc(crtc);
-	uint32_t width, height, base_offset;
 	bool recreate_primary = false;
 	int ret;
 	int surf_id;
@@ -538,12 +609,6 @@ static int qxl_crtc_mode_set(struct drm_crtc *crtc,
 	}
 	qfb = to_qxl_framebuffer(crtc->primary->fb);
 	bo = gem_to_qxl_bo(qfb->obj);
-	if (!m)
-		/* and do we care? */
-		DRM_DEBUG("%dx%d: not a native mode\n", x, y);
-	else
-		DRM_DEBUG("%dx%d: qxl id %d\n",
-			  mode->hdisplay, mode->vdisplay, m->id);
 	DRM_DEBUG("+%d+%d (%d,%d) => (%d,%d)\n",
 		  x, y,
 		  mode->hdisplay, mode->vdisplay,
@@ -553,9 +618,10 @@ static int qxl_crtc_mode_set(struct drm_crtc *crtc,
 	if (qcrtc->index == 0)
 		recreate_primary = true;
 
-	width = mode->hdisplay;
-	height = mode->vdisplay;
-	base_offset = 0;
+	if (bo->surf.stride * bo->surf.height > qdev->vram_size) {
+		DRM_ERROR("Mode doesn't fit in vram size (vgamem)");
+		return -EINVAL;
+        }
 
 	ret = qxl_bo_reserve(bo, false);
 	if (ret != 0)
@@ -569,10 +635,10 @@ static int qxl_crtc_mode_set(struct drm_crtc *crtc,
 	if (recreate_primary) {
 		qxl_io_destroy_primary(qdev);
 		qxl_io_log(qdev,
-			   "recreate primary: %dx%d (was %dx%d,%d,%d)\n",
-			   width, height, bo->surf.width,
-			   bo->surf.height, bo->surf.stride, bo->surf.format);
-		qxl_io_create_primary(qdev, base_offset, bo);
+			   "recreate primary: %dx%d,%d,%d\n",
+			   bo->surf.width, bo->surf.height,
+			   bo->surf.stride, bo->surf.format);
+		qxl_io_create_primary(qdev, 0, bo);
 		bo->is_primary = true;
 	}
 
@@ -902,6 +968,10 @@ static int qdev_output_init(struct drm_device *dev, int num_output)
 
 	drm_object_attach_property(&connector->base,
 				   qdev->hotplug_mode_update_property, 0);
+	drm_object_attach_property(&connector->base,
+				   dev->mode_config.suggested_x_property, 0);
+	drm_object_attach_property(&connector->base,
+				   dev->mode_config.suggested_y_property, 0);
 	drm_connector_register(connector);
 	return 0;
 }
@@ -1015,6 +1085,7 @@ int qxl_modeset_init(struct qxl_device *qdev)
 
 	qdev->ddev->mode_config.fb_base = qdev->vram_base;
 
+	drm_mode_create_suggested_offset_properties(qdev->ddev);
 	qxl_mode_create_hotplug_mode_update_property(qdev);
 
 	for (i = 0 ; i < qxl_num_crtc; ++i) {
diff --git a/drivers/gpu/drm/qxl/qxl_drv.c b/drivers/gpu/drm/qxl/qxl_drv.c
index a3fd92029a14..1d9b80c91a15 100644
--- a/drivers/gpu/drm/qxl/qxl_drv.c
+++ b/drivers/gpu/drm/qxl/qxl_drv.c
@@ -84,6 +84,7 @@ static const struct file_operations qxl_fops = {
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
 	.poll = drm_poll,
+	.read = drm_read,
 	.mmap = qxl_mmap,
 };
 
@@ -195,6 +196,20 @@ static int qxl_pm_restore(struct device *dev)
 	return qxl_drm_resume(drm_dev, false);
 }
 
+static u32 qxl_noop_get_vblank_counter(struct drm_device *dev, int crtc)
+{
+	return dev->vblank[crtc].count.counter;
+}
+
+static int qxl_noop_enable_vblank(struct drm_device *dev, int crtc)
+{
+	return 0;
+}
+
+static void qxl_noop_disable_vblank(struct drm_device *dev, int crtc)
+{
+}
+
 static const struct dev_pm_ops qxl_pm_ops = {
 	.suspend = qxl_pm_suspend,
 	.resume = qxl_pm_resume,
@@ -212,10 +227,15 @@ static struct pci_driver qxl_pci_driver = {
 };
 
 static struct drm_driver qxl_driver = {
-	.driver_features = DRIVER_GEM | DRIVER_MODESET |
+	.driver_features = DRIVER_GEM | DRIVER_MODESET | DRIVER_PRIME |
 			   DRIVER_HAVE_IRQ | DRIVER_IRQ_SHARED,
 	.load = qxl_driver_load,
 	.unload = qxl_driver_unload,
+	.get_vblank_counter = qxl_noop_get_vblank_counter,
+	.enable_vblank = qxl_noop_enable_vblank,
+	.disable_vblank = qxl_noop_disable_vblank,
+
+	.set_busid = drm_pci_set_busid,
 
 	.dumb_create = qxl_mode_dumb_create,
 	.dumb_map_offset = qxl_mode_dumb_mmap,
@@ -224,6 +244,17 @@ static struct drm_driver qxl_driver = {
 	.debugfs_init = qxl_debugfs_init,
 	.debugfs_cleanup = qxl_debugfs_takedown,
 #endif
+	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
+	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
+	.gem_prime_export = drm_gem_prime_export,
+	.gem_prime_import = drm_gem_prime_import,
+	.gem_prime_pin = qxl_gem_prime_pin,
+	.gem_prime_unpin = qxl_gem_prime_unpin,
+	.gem_prime_get_sg_table = qxl_gem_prime_get_sg_table,
+	.gem_prime_import_sg_table = qxl_gem_prime_import_sg_table,
+	.gem_prime_vmap = qxl_gem_prime_vmap,
+	.gem_prime_vunmap = qxl_gem_prime_vunmap,
+	.gem_prime_mmap = qxl_gem_prime_mmap,
 	.gem_free_object = qxl_gem_object_free,
 	.gem_open_object = qxl_gem_object_open,
 	.gem_close_object = qxl_gem_object_close,
diff --git a/drivers/gpu/drm/qxl/qxl_drv.h b/drivers/gpu/drm/qxl/qxl_drv.h
index 36ed40ba773f..7c6cafe21f5f 100644
--- a/drivers/gpu/drm/qxl/qxl_drv.h
+++ b/drivers/gpu/drm/qxl/qxl_drv.h
@@ -31,6 +31,7 @@
  * Definitions taken from spice-protocol, plus kernel driver specific bits.
  */
 
+#include <linux/fence.h>
 #include <linux/workqueue.h>
 #include <linux/firmware.h>
 #include <linux/platform_device.h>
@@ -42,6 +43,8 @@
 #include <ttm/ttm_placement.h>
 #include <ttm/ttm_module.h>
 
+#include <drm/drm_gem.h>
+
 /* just for ttm_validate_buffer */
 #include <ttm/ttm_execbuf_util.h>
 
@@ -95,31 +98,24 @@ enum {
 	QXL_INTERRUPT_IO_CMD |\
 	QXL_INTERRUPT_CLIENT_MONITORS_CONFIG)
 
-struct qxl_fence {
-	struct qxl_device *qdev;
-	uint32_t num_active_releases;
-	uint32_t *release_ids;
-	struct radix_tree_root tree;
-};
-
 struct qxl_bo {
 	/* Protected by gem.mutex */
 	struct list_head		list;
 	/* Protected by tbo.reserved */
-	u32				placements[3];
+	struct ttm_place		placements[3];
 	struct ttm_placement		placement;
 	struct ttm_buffer_object	tbo;
 	struct ttm_bo_kmap_obj		kmap;
 	unsigned			pin_count;
 	void				*kptr;
 	int                             type;
+
 	/* Constant after initialization */
 	struct drm_gem_object		gem_base;
 	bool is_primary; /* is this now a primary surface */
 	bool hw_surf_alloc;
 	struct qxl_surface surf;
 	uint32_t surface_id;
-	struct qxl_fence fence; /* per bo fence  - list of releases */
 	struct qxl_release *surf_create;
 };
 #define gem_to_qxl_bo(gobj) container_of((gobj), struct qxl_bo, gem_base)
@@ -191,6 +187,8 @@ enum {
  * spice-protocol/qxl_dev.h */
 #define QXL_MAX_RES 96
 struct qxl_release {
+	struct fence base;
+
 	int id;
 	int type;
 	uint32_t release_offset;
@@ -284,7 +282,9 @@ struct qxl_device {
 	uint8_t		slot_gen_bits;
 	uint64_t	va_slot_mask;
 
+	spinlock_t	release_lock;
 	struct idr	release_idr;
+	uint32_t	release_seqno;
 	spinlock_t release_idr_lock;
 	struct mutex	async_io_mutex;
 	unsigned int last_sent_io_cmd;
@@ -532,6 +532,18 @@ int qxl_garbage_collect(struct qxl_device *qdev);
 int qxl_debugfs_init(struct drm_minor *minor);
 void qxl_debugfs_takedown(struct drm_minor *minor);
 
+/* qxl_prime.c */
+int qxl_gem_prime_pin(struct drm_gem_object *obj);
+void qxl_gem_prime_unpin(struct drm_gem_object *obj);
+struct sg_table *qxl_gem_prime_get_sg_table(struct drm_gem_object *obj);
+struct drm_gem_object *qxl_gem_prime_import_sg_table(
+	struct drm_device *dev, struct dma_buf_attachment *attach,
+	struct sg_table *sgt);
+void *qxl_gem_prime_vmap(struct drm_gem_object *obj);
+void qxl_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr);
+int qxl_gem_prime_mmap(struct drm_gem_object *obj,
+				struct vm_area_struct *vma);
+
 /* qxl_irq.c */
 int qxl_irq_init(struct qxl_device *qdev);
 irqreturn_t qxl_irq_handler(int irq, void *arg);
@@ -561,10 +573,4 @@ qxl_surface_lookup(struct drm_device *dev, int surface_id);
 void qxl_surface_evict(struct qxl_device *qdev, struct qxl_bo *surf, bool freeing);
 int qxl_update_surface(struct qxl_device *qdev, struct qxl_bo *surf);
 
-/* qxl_fence.c */
-void qxl_fence_add_release_locked(struct qxl_fence *qfence, uint32_t rel_id);
-int qxl_fence_remove_release(struct qxl_fence *qfence, uint32_t rel_id);
-int qxl_fence_init(struct qxl_device *qdev, struct qxl_fence *qfence);
-void qxl_fence_fini(struct qxl_fence *qfence);
-
 #endif
diff --git a/drivers/gpu/drm/qxl/qxl_fb.c b/drivers/gpu/drm/qxl/qxl_fb.c
index df567888bb1e..3d7c1d00a424 100644
--- a/drivers/gpu/drm/qxl/qxl_fb.c
+++ b/drivers/gpu/drm/qxl/qxl_fb.c
@@ -625,7 +625,8 @@ static int qxl_fb_find_or_create_single(
 		struct drm_fb_helper *helper,
 		struct drm_fb_helper_surface_size *sizes)
 {
-	struct qxl_fbdev *qfbdev = (struct qxl_fbdev *)helper;
+	struct qxl_fbdev *qfbdev =
+		container_of(helper, struct qxl_fbdev, helper);
 	int new_fb = 0;
 	int ret;
 
diff --git a/drivers/gpu/drm/qxl/qxl_fence.c b/drivers/gpu/drm/qxl/qxl_fence.c
deleted file mode 100644
index ae59e91cfb9a..000000000000
--- a/drivers/gpu/drm/qxl/qxl_fence.c
+++ /dev/null
@@ -1,91 +0,0 @@
-/*
- * Copyright 2013 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Dave Airlie
- *          Alon Levy
- */
-
-
-#include "qxl_drv.h"
-
-/* QXL fencing-
-
-   When we submit operations to the GPU we pass a release reference to the GPU
-   with them, the release reference is then added to the release ring when
-   the GPU is finished with that particular operation and has removed it from
-   its tree.
-
-   So we have can have multiple outstanding non linear fences per object.
-
-   From a TTM POV we only care if the object has any outstanding releases on
-   it.
-
-   we wait until all outstanding releases are processeed.
-
-   sync object is just a list of release ids that represent that fence on
-   that buffer.
-
-   we just add new releases onto the sync object attached to the object.
-
-   This currently uses a radix tree to store the list of release ids.
-
-   For some reason every so often qxl hw fails to release, things go wrong.
-*/
-/* must be called with the fence lock held */
-void qxl_fence_add_release_locked(struct qxl_fence *qfence, uint32_t rel_id)
-{
-	radix_tree_insert(&qfence->tree, rel_id, qfence);
-	qfence->num_active_releases++;
-}
-
-int qxl_fence_remove_release(struct qxl_fence *qfence, uint32_t rel_id)
-{
-	void *ret;
-	int retval = 0;
-	struct qxl_bo *bo = container_of(qfence, struct qxl_bo, fence);
-
-	spin_lock(&bo->tbo.bdev->fence_lock);
-
-	ret = radix_tree_delete(&qfence->tree, rel_id);
-	if (ret == qfence)
-		qfence->num_active_releases--;
-	else {
-		DRM_DEBUG("didn't find fence in radix tree for %d\n", rel_id);
-		retval = -ENOENT;
-	}
-	spin_unlock(&bo->tbo.bdev->fence_lock);
-	return retval;
-}
-
-
-int qxl_fence_init(struct qxl_device *qdev, struct qxl_fence *qfence)
-{
-	qfence->qdev = qdev;
-	qfence->num_active_releases = 0;
-	INIT_RADIX_TREE(&qfence->tree, GFP_ATOMIC);
-	return 0;
-}
-
-void qxl_fence_fini(struct qxl_fence *qfence)
-{
-	kfree(qfence->release_ids);
-	qfence->num_active_releases = 0;
-}
diff --git a/drivers/gpu/drm/qxl/qxl_kms.c b/drivers/gpu/drm/qxl/qxl_kms.c
index fd88eb4a3f79..b2977a181935 100644
--- a/drivers/gpu/drm/qxl/qxl_kms.c
+++ b/drivers/gpu/drm/qxl/qxl_kms.c
@@ -223,6 +223,7 @@ static int qxl_device_init(struct qxl_device *qdev,
 
 	idr_init(&qdev->release_idr);
 	spin_lock_init(&qdev->release_idr_lock);
+	spin_lock_init(&qdev->release_lock);
 
 	idr_init(&qdev->surf_id_idr);
 	spin_lock_init(&qdev->surf_id_idr_lock);
@@ -297,6 +298,9 @@ int qxl_driver_unload(struct drm_device *dev)
 
 	if (qdev == NULL)
 		return 0;
+
+	drm_vblank_cleanup(dev);
+
 	qxl_modeset_fini(qdev);
 	qxl_device_fini(qdev);
 
@@ -324,15 +328,20 @@ int qxl_driver_load(struct drm_device *dev, unsigned long flags)
 	if (r)
 		goto out;
 
+	r = drm_vblank_init(dev, 1);
+	if (r)
+		goto unload;
+
 	r = qxl_modeset_init(qdev);
-	if (r) {
-		qxl_driver_unload(dev);
-		goto out;
-	}
+	if (r)
+		goto unload;
 
 	drm_kms_helper_poll_init(qdev->ddev);
 
 	return 0;
+unload:
+	qxl_driver_unload(dev);
+
 out:
 	kfree(qdev);
 	return r;
diff --git a/drivers/gpu/drm/qxl/qxl_object.c b/drivers/gpu/drm/qxl/qxl_object.c
index b95f144f0b49..cdeaf08fdc74 100644
--- a/drivers/gpu/drm/qxl/qxl_object.c
+++ b/drivers/gpu/drm/qxl/qxl_object.c
@@ -36,7 +36,6 @@ static void qxl_ttm_bo_destroy(struct ttm_buffer_object *tbo)
 	qdev = (struct qxl_device *)bo->gem_base.dev->dev_private;
 
 	qxl_surface_evict(qdev, bo, false);
-	qxl_fence_fini(&bo->fence);
 	mutex_lock(&qdev->gem.mutex);
 	list_del_init(&bo->list);
 	mutex_unlock(&qdev->gem.mutex);
@@ -55,21 +54,24 @@ void qxl_ttm_placement_from_domain(struct qxl_bo *qbo, u32 domain, bool pinned)
 {
 	u32 c = 0;
 	u32 pflag = pinned ? TTM_PL_FLAG_NO_EVICT : 0;
+	unsigned i;
 
-	qbo->placement.fpfn = 0;
-	qbo->placement.lpfn = 0;
 	qbo->placement.placement = qbo->placements;
 	qbo->placement.busy_placement = qbo->placements;
 	if (domain == QXL_GEM_DOMAIN_VRAM)
-		qbo->placements[c++] = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_VRAM | pflag;
+		qbo->placements[c++].flags = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_VRAM | pflag;
 	if (domain == QXL_GEM_DOMAIN_SURFACE)
-		qbo->placements[c++] = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_PRIV0 | pflag;
+		qbo->placements[c++].flags = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_PRIV0 | pflag;
 	if (domain == QXL_GEM_DOMAIN_CPU)
-		qbo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM | pflag;
+		qbo->placements[c++].flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM | pflag;
 	if (!c)
-		qbo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		qbo->placements[c++].flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
 	qbo->placement.num_placement = c;
 	qbo->placement.num_busy_placement = c;
+	for (i = 0; i < c; ++i) {
+		qbo->placements[i].fpfn = 0;
+		qbo->placements[i].lpfn = 0;
+	}
 }
 
 
@@ -99,7 +101,6 @@ int qxl_bo_create(struct qxl_device *qdev,
 	bo->type = domain;
 	bo->pin_count = pinned ? 1 : 0;
 	bo->surface_id = 0;
-	qxl_fence_init(qdev, &bo->fence);
 	INIT_LIST_HEAD(&bo->list);
 
 	if (surf)
@@ -109,7 +110,7 @@ int qxl_bo_create(struct qxl_device *qdev,
 
 	r = ttm_bo_init(&qdev->mman.bdev, &bo->tbo, size, type,
 			&bo->placement, 0, !kernel, NULL, size,
-			NULL, &qxl_ttm_bo_destroy);
+			NULL, NULL, &qxl_ttm_bo_destroy);
 	if (unlikely(r != 0)) {
 		if (r != -ERESTARTSYS)
 			dev_err(qdev->dev,
@@ -259,7 +260,7 @@ int qxl_bo_unpin(struct qxl_bo *bo)
 	if (bo->pin_count)
 		return 0;
 	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] &= ~TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags &= ~TTM_PL_FLAG_NO_EVICT;
 	r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
 	if (unlikely(r != 0))
 		dev_err(qdev->dev, "%p validate failed for unpin\n", bo);
diff --git a/drivers/gpu/drm/qxl/qxl_object.h b/drivers/gpu/drm/qxl/qxl_object.h
index 83a423293afd..37af1bc0dd00 100644
--- a/drivers/gpu/drm/qxl/qxl_object.h
+++ b/drivers/gpu/drm/qxl/qxl_object.h
@@ -76,12 +76,10 @@ static inline int qxl_bo_wait(struct qxl_bo *bo, u32 *mem_type,
 		}
 		return r;
 	}
-	spin_lock(&bo->tbo.bdev->fence_lock);
 	if (mem_type)
 		*mem_type = bo->tbo.mem.mem_type;
-	if (bo->tbo.sync_obj)
-		r = ttm_bo_wait(&bo->tbo, true, true, no_wait);
-	spin_unlock(&bo->tbo.bdev->fence_lock);
+
+	r = ttm_bo_wait(&bo->tbo, true, true, no_wait);
 	ttm_bo_unreserve(&bo->tbo);
 	return r;
 }
diff --git a/drivers/gpu/drm/qxl/qxl_prime.c b/drivers/gpu/drm/qxl/qxl_prime.c
new file mode 100644
index 000000000000..3d031b50a8fd
--- /dev/null
+++ b/drivers/gpu/drm/qxl/qxl_prime.c
@@ -0,0 +1,72 @@
+/*
+ * Copyright 2014 Canonical
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Andreas Pokorny
+ */
+
+#include "qxl_drv.h"
+
+/* Empty Implementations as there should not be any other driver for a virtual
+ * device that might share buffers with qxl */
+
+int qxl_gem_prime_pin(struct drm_gem_object *obj)
+{
+	WARN_ONCE(1, "not implemented");
+	return -ENOSYS;
+}
+
+void qxl_gem_prime_unpin(struct drm_gem_object *obj)
+{
+	WARN_ONCE(1, "not implemented");
+}
+
+
+struct sg_table *qxl_gem_prime_get_sg_table(struct drm_gem_object *obj)
+{
+	WARN_ONCE(1, "not implemented");
+	return ERR_PTR(-ENOSYS);
+}
+
+struct drm_gem_object *qxl_gem_prime_import_sg_table(
+	struct drm_device *dev, struct dma_buf_attachment *attach,
+	struct sg_table *table)
+{
+	WARN_ONCE(1, "not implemented");
+	return ERR_PTR(-ENOSYS);
+}
+
+void *qxl_gem_prime_vmap(struct drm_gem_object *obj)
+{
+	WARN_ONCE(1, "not implemented");
+	return ERR_PTR(-ENOSYS);
+}
+
+void qxl_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr)
+{
+	WARN_ONCE(1, "not implemented");
+}
+
+int qxl_gem_prime_mmap(struct drm_gem_object *obj,
+		       struct vm_area_struct *area)
+{
+	WARN_ONCE(1, "not implemented");
+	return ENOSYS;
+}
diff --git a/drivers/gpu/drm/qxl/qxl_release.c b/drivers/gpu/drm/qxl/qxl_release.c
index 14e776f1d14e..d9b25684ac98 100644
--- a/drivers/gpu/drm/qxl/qxl_release.c
+++ b/drivers/gpu/drm/qxl/qxl_release.c
@@ -21,6 +21,7 @@
  */
 #include "qxl_drv.h"
 #include "qxl_object.h"
+#include <trace/events/fence.h>
 
 /*
  * drawable cmd cache - allocate a bunch of VRAM pages, suballocate
@@ -39,6 +40,88 @@
 static const int release_size_per_bo[] = { RELEASE_SIZE, SURFACE_RELEASE_SIZE, RELEASE_SIZE };
 static const int releases_per_bo[] = { RELEASES_PER_BO, SURFACE_RELEASES_PER_BO, RELEASES_PER_BO };
 
+static const char *qxl_get_driver_name(struct fence *fence)
+{
+	return "qxl";
+}
+
+static const char *qxl_get_timeline_name(struct fence *fence)
+{
+	return "release";
+}
+
+static bool qxl_nop_signaling(struct fence *fence)
+{
+	/* fences are always automatically signaled, so just pretend we did this.. */
+	return true;
+}
+
+static long qxl_fence_wait(struct fence *fence, bool intr, signed long timeout)
+{
+	struct qxl_device *qdev;
+	struct qxl_release *release;
+	int count = 0, sc = 0;
+	bool have_drawable_releases;
+	unsigned long cur, end = jiffies + timeout;
+
+	qdev = container_of(fence->lock, struct qxl_device, release_lock);
+	release = container_of(fence, struct qxl_release, base);
+	have_drawable_releases = release->type == QXL_RELEASE_DRAWABLE;
+
+retry:
+	sc++;
+
+	if (fence_is_signaled(fence))
+		goto signaled;
+
+	qxl_io_notify_oom(qdev);
+
+	for (count = 0; count < 11; count++) {
+		if (!qxl_queue_garbage_collect(qdev, true))
+			break;
+
+		if (fence_is_signaled(fence))
+			goto signaled;
+	}
+
+	if (fence_is_signaled(fence))
+		goto signaled;
+
+	if (have_drawable_releases || sc < 4) {
+		if (sc > 2)
+			/* back off */
+			usleep_range(500, 1000);
+
+		if (time_after(jiffies, end))
+			return 0;
+
+		if (have_drawable_releases && sc > 300) {
+			FENCE_WARN(fence, "failed to wait on release %d "
+					  "after spincount %d\n",
+					  fence->context & ~0xf0000000, sc);
+			goto signaled;
+		}
+		goto retry;
+	}
+	/*
+	 * yeah, original sync_obj_wait gave up after 3 spins when
+	 * have_drawable_releases is not set.
+	 */
+
+signaled:
+	cur = jiffies;
+	if (time_after(cur, end))
+		return 0;
+	return end - cur;
+}
+
+static const struct fence_ops qxl_fence_ops = {
+	.get_driver_name = qxl_get_driver_name,
+	.get_timeline_name = qxl_get_timeline_name,
+	.enable_signaling = qxl_nop_signaling,
+	.wait = qxl_fence_wait,
+};
+
 static uint64_t
 qxl_release_alloc(struct qxl_device *qdev, int type,
 		  struct qxl_release **ret)
@@ -46,13 +129,13 @@ qxl_release_alloc(struct qxl_device *qdev, int type,
 	struct qxl_release *release;
 	int handle;
 	size_t size = sizeof(*release);
-	int idr_ret;
 
 	release = kmalloc(size, GFP_KERNEL);
 	if (!release) {
 		DRM_ERROR("Out of memory\n");
 		return 0;
 	}
+	release->base.ops = NULL;
 	release->type = type;
 	release->release_offset = 0;
 	release->surface_release_id = 0;
@@ -60,44 +143,61 @@ qxl_release_alloc(struct qxl_device *qdev, int type,
 
 	idr_preload(GFP_KERNEL);
 	spin_lock(&qdev->release_idr_lock);
-	idr_ret = idr_alloc(&qdev->release_idr, release, 1, 0, GFP_NOWAIT);
+	handle = idr_alloc(&qdev->release_idr, release, 1, 0, GFP_NOWAIT);
+	release->base.seqno = ++qdev->release_seqno;
 	spin_unlock(&qdev->release_idr_lock);
 	idr_preload_end();
-	handle = idr_ret;
-	if (idr_ret < 0)
-		goto release_fail;
+	if (handle < 0) {
+		kfree(release);
+		*ret = NULL;
+		return handle;
+	}
 	*ret = release;
 	QXL_INFO(qdev, "allocated release %lld\n", handle);
 	release->id = handle;
-release_fail:
-
 	return handle;
 }
 
+static void
+qxl_release_free_list(struct qxl_release *release)
+{
+	while (!list_empty(&release->bos)) {
+		struct qxl_bo_list *entry;
+		struct qxl_bo *bo;
+
+		entry = container_of(release->bos.next,
+				     struct qxl_bo_list, tv.head);
+		bo = to_qxl_bo(entry->tv.bo);
+		qxl_bo_unref(&bo);
+		list_del(&entry->tv.head);
+		kfree(entry);
+	}
+}
+
 void
 qxl_release_free(struct qxl_device *qdev,
 		 struct qxl_release *release)
 {
-	struct qxl_bo_list *entry, *tmp;
 	QXL_INFO(qdev, "release %d, type %d\n", release->id,
 		 release->type);
 
 	if (release->surface_release_id)
 		qxl_surface_id_dealloc(qdev, release->surface_release_id);
 
-	list_for_each_entry_safe(entry, tmp, &release->bos, tv.head) {
-		struct qxl_bo *bo = to_qxl_bo(entry->tv.bo);
-		QXL_INFO(qdev, "release %llx\n",
-			drm_vma_node_offset_addr(&entry->tv.bo->vma_node)
-						- DRM_FILE_OFFSET);
-		qxl_fence_remove_release(&bo->fence, release->id);
-		qxl_bo_unref(&bo);
-		kfree(entry);
-	}
 	spin_lock(&qdev->release_idr_lock);
 	idr_remove(&qdev->release_idr, release->id);
 	spin_unlock(&qdev->release_idr_lock);
-	kfree(release);
+
+	if (release->base.ops) {
+		WARN_ON(list_empty(&release->bos));
+		qxl_release_free_list(release);
+
+		fence_signal(&release->base);
+		fence_put(&release->base);
+	} else {
+		qxl_release_free_list(release);
+		kfree(release);
+	}
 }
 
 static int qxl_release_bo_alloc(struct qxl_device *qdev,
@@ -126,6 +226,7 @@ int qxl_release_list_add(struct qxl_release *release, struct qxl_bo *bo)
 
 	qxl_bo_ref(bo);
 	entry->tv.bo = &bo->tbo;
+	entry->tv.shared = false;
 	list_add_tail(&entry->tv.head, &release->bos);
 	return 0;
 }
@@ -142,6 +243,10 @@ static int qxl_release_validate_bo(struct qxl_bo *bo)
 			return ret;
 	}
 
+	ret = reservation_object_reserve_shared(bo->tbo.resv);
+	if (ret)
+		return ret;
+
 	/* allocate a surface for reserved + validated buffers */
 	ret = qxl_bo_check_id(bo->gem_base.dev->dev_private, bo);
 	if (ret)
@@ -159,7 +264,8 @@ int qxl_release_reserve_list(struct qxl_release *release, bool no_intr)
 	if (list_is_singular(&release->bos))
 		return 0;
 
-	ret = ttm_eu_reserve_buffers(&release->ticket, &release->bos);
+	ret = ttm_eu_reserve_buffers(&release->ticket, &release->bos,
+				     !no_intr, NULL);
 	if (ret)
 		return ret;
 
@@ -199,6 +305,8 @@ int qxl_alloc_surface_release_reserved(struct qxl_device *qdev,
 
 		/* stash the release after the create command */
 		idr_ret = qxl_release_alloc(qdev, QXL_RELEASE_SURFACE_CMD, release);
+		if (idr_ret < 0)
+			return idr_ret;
 		bo = qxl_bo_ref(to_qxl_bo(entry->tv.bo));
 
 		(*release)->release_offset = create_rel->release_offset + 64;
@@ -239,6 +347,11 @@ int qxl_alloc_release_reserved(struct qxl_device *qdev, unsigned long size,
 	}
 
 	idr_ret = qxl_release_alloc(qdev, type, release);
+	if (idr_ret < 0) {
+		if (rbo)
+			*rbo = NULL;
+		return idr_ret;
+	}
 
 	mutex_lock(&qdev->release_mutex);
 	if (qdev->current_release_bo_offset[cur_idx] + 1 >= releases_per_bo[cur_idx]) {
@@ -319,40 +432,44 @@ void qxl_release_unmap(struct qxl_device *qdev,
 
 void qxl_release_fence_buffer_objects(struct qxl_release *release)
 {
-	struct ttm_validate_buffer *entry;
 	struct ttm_buffer_object *bo;
 	struct ttm_bo_global *glob;
 	struct ttm_bo_device *bdev;
 	struct ttm_bo_driver *driver;
 	struct qxl_bo *qbo;
+	struct ttm_validate_buffer *entry;
+	struct qxl_device *qdev;
 
 	/* if only one object on the release its the release itself
 	   since these objects are pinned no need to reserve */
-	if (list_is_singular(&release->bos))
+	if (list_is_singular(&release->bos) || list_empty(&release->bos))
 		return;
 
 	bo = list_first_entry(&release->bos, struct ttm_validate_buffer, head)->bo;
 	bdev = bo->bdev;
+	qdev = container_of(bdev, struct qxl_device, mman.bdev);
+
+	/*
+	 * Since we never really allocated a context and we don't want to conflict,
+	 * set the highest bits. This will break if we really allow exporting of dma-bufs.
+	 */
+	fence_init(&release->base, &qxl_fence_ops, &qdev->release_lock,
+		   release->id | 0xf0000000, release->base.seqno);
+	trace_fence_emit(&release->base);
+
 	driver = bdev->driver;
 	glob = bo->glob;
 
 	spin_lock(&glob->lru_lock);
-	spin_lock(&bdev->fence_lock);
 
 	list_for_each_entry(entry, &release->bos, head) {
 		bo = entry->bo;
 		qbo = to_qxl_bo(bo);
 
-		if (!entry->bo->sync_obj)
-			entry->bo->sync_obj = &qbo->fence;
-
-		qxl_fence_add_release_locked(&qbo->fence, release->id);
-
+		reservation_object_add_shared_fence(bo->resv, &release->base);
 		ttm_bo_add_to_lru(bo);
 		__ttm_bo_unreserve(bo);
-		entry->reserved = false;
 	}
-	spin_unlock(&bdev->fence_lock);
 	spin_unlock(&glob->lru_lock);
 	ww_acquire_fini(&release->ticket);
 }
diff --git a/drivers/gpu/drm/qxl/qxl_ttm.c b/drivers/gpu/drm/qxl/qxl_ttm.c
index 71a1baeac14e..0cbc4c987164 100644
--- a/drivers/gpu/drm/qxl/qxl_ttm.c
+++ b/drivers/gpu/drm/qxl/qxl_ttm.c
@@ -127,7 +127,7 @@ int qxl_mmap(struct file *filp, struct vm_area_struct *vma)
 	if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET)) {
 		pr_info("%s: vma->vm_pgoff (%ld) < DRM_FILE_PAGE_OFFSET\n",
 			__func__, vma->vm_pgoff);
-		return drm_mmap(filp, vma);
+		return -EINVAL;
 	}
 
 	file_priv = filp->private_data;
@@ -188,11 +188,13 @@ static void qxl_evict_flags(struct ttm_buffer_object *bo,
 				struct ttm_placement *placement)
 {
 	struct qxl_bo *qbo;
-	static u32 placements = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+	static struct ttm_place placements = {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM
+	};
 
 	if (!qxl_ttm_bo_is_qxl_bo(bo)) {
-		placement->fpfn = 0;
-		placement->lpfn = 0;
 		placement->placement = &placements;
 		placement->busy_placement = &placements;
 		placement->num_placement = 1;
@@ -355,92 +357,6 @@ static int qxl_bo_move(struct ttm_buffer_object *bo,
 	return ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem);
 }
 
-
-static int qxl_sync_obj_wait(void *sync_obj,
-			     bool lazy, bool interruptible)
-{
-	struct qxl_fence *qfence = (struct qxl_fence *)sync_obj;
-	int count = 0, sc = 0;
-	struct qxl_bo *bo = container_of(qfence, struct qxl_bo, fence);
-
-	if (qfence->num_active_releases == 0)
-		return 0;
-
-retry:
-	if (sc == 0) {
-		if (bo->type == QXL_GEM_DOMAIN_SURFACE)
-			qxl_update_surface(qfence->qdev, bo);
-	} else if (sc >= 1) {
-		qxl_io_notify_oom(qfence->qdev);
-	}
-
-	sc++;
-
-	for (count = 0; count < 10; count++) {
-		bool ret;
-		ret = qxl_queue_garbage_collect(qfence->qdev, true);
-		if (ret == false)
-			break;
-
-		if (qfence->num_active_releases == 0)
-			return 0;
-	}
-
-	if (qfence->num_active_releases) {
-		bool have_drawable_releases = false;
-		void **slot;
-		struct radix_tree_iter iter;
-		int release_id;
-
-		radix_tree_for_each_slot(slot, &qfence->tree, &iter, 0) {
-			struct qxl_release *release;
-
-			release_id = iter.index;
-			release = qxl_release_from_id_locked(qfence->qdev, release_id);
-			if (release == NULL)
-				continue;
-
-			if (release->type == QXL_RELEASE_DRAWABLE)
-				have_drawable_releases = true;
-		}
-
-		qxl_queue_garbage_collect(qfence->qdev, true);
-
-		if (have_drawable_releases || sc < 4) {
-			if (sc > 2)
-				/* back off */
-				usleep_range(500, 1000);
-			if (have_drawable_releases && sc > 300) {
-				WARN(1, "sync obj %d still has outstanding releases %d %d %d %ld %d\n", sc, bo->surface_id, bo->is_primary, bo->pin_count, (unsigned long)bo->gem_base.size, qfence->num_active_releases);
-				return -EBUSY;
-			}
-			goto retry;
-		}
-	}
-	return 0;
-}
-
-static int qxl_sync_obj_flush(void *sync_obj)
-{
-	return 0;
-}
-
-static void qxl_sync_obj_unref(void **sync_obj)
-{
-	*sync_obj = NULL;
-}
-
-static void *qxl_sync_obj_ref(void *sync_obj)
-{
-	return sync_obj;
-}
-
-static bool qxl_sync_obj_signaled(void *sync_obj)
-{
-	struct qxl_fence *qfence = (struct qxl_fence *)sync_obj;
-	return (qfence->num_active_releases == 0);
-}
-
 static void qxl_bo_move_notify(struct ttm_buffer_object *bo,
 			       struct ttm_mem_reg *new_mem)
 {
@@ -467,16 +383,9 @@ static struct ttm_bo_driver qxl_bo_driver = {
 	.verify_access = &qxl_verify_access,
 	.io_mem_reserve = &qxl_ttm_io_mem_reserve,
 	.io_mem_free = &qxl_ttm_io_mem_free,
-	.sync_obj_signaled = &qxl_sync_obj_signaled,
-	.sync_obj_wait = &qxl_sync_obj_wait,
-	.sync_obj_flush = &qxl_sync_obj_flush,
-	.sync_obj_unref = &qxl_sync_obj_unref,
-	.sync_obj_ref = &qxl_sync_obj_ref,
 	.move_notify = &qxl_bo_move_notify,
 };
 
-
-
 int qxl_ttm_init(struct qxl_device *qdev)
 {
 	int r;
diff --git a/drivers/gpu/drm/r128/r128_cce.c b/drivers/gpu/drm/r128/r128_cce.c
index 59459fe4e8c5..2c45ac9c1dc3 100644
--- a/drivers/gpu/drm/r128/r128_cce.c
+++ b/drivers/gpu/drm/r128/r128_cce.c
@@ -452,7 +452,7 @@ static int r128_do_init_cce(struct drm_device *dev, drm_r128_init_t *init)
 	dev_priv->span_pitch_offset_c = (((dev_priv->depth_pitch / 8) << 21) |
 					 (dev_priv->span_offset >> 5));
 
-	dev_priv->sarea = drm_getsarea(dev);
+	dev_priv->sarea = drm_legacy_getsarea(dev);
 	if (!dev_priv->sarea) {
 		DRM_ERROR("could not find sarea!\n");
 		dev->dev_private = (void *)dev_priv;
@@ -460,21 +460,21 @@ static int r128_do_init_cce(struct drm_device *dev, drm_r128_init_t *init)
 		return -EINVAL;
 	}
 
-	dev_priv->mmio = drm_core_findmap(dev, init->mmio_offset);
+	dev_priv->mmio = drm_legacy_findmap(dev, init->mmio_offset);
 	if (!dev_priv->mmio) {
 		DRM_ERROR("could not find mmio region!\n");
 		dev->dev_private = (void *)dev_priv;
 		r128_do_cleanup_cce(dev);
 		return -EINVAL;
 	}
-	dev_priv->cce_ring = drm_core_findmap(dev, init->ring_offset);
+	dev_priv->cce_ring = drm_legacy_findmap(dev, init->ring_offset);
 	if (!dev_priv->cce_ring) {
 		DRM_ERROR("could not find cce ring region!\n");
 		dev->dev_private = (void *)dev_priv;
 		r128_do_cleanup_cce(dev);
 		return -EINVAL;
 	}
-	dev_priv->ring_rptr = drm_core_findmap(dev, init->ring_rptr_offset);
+	dev_priv->ring_rptr = drm_legacy_findmap(dev, init->ring_rptr_offset);
 	if (!dev_priv->ring_rptr) {
 		DRM_ERROR("could not find ring read pointer!\n");
 		dev->dev_private = (void *)dev_priv;
@@ -482,7 +482,7 @@ static int r128_do_init_cce(struct drm_device *dev, drm_r128_init_t *init)
 		return -EINVAL;
 	}
 	dev->agp_buffer_token = init->buffers_offset;
-	dev->agp_buffer_map = drm_core_findmap(dev, init->buffers_offset);
+	dev->agp_buffer_map = drm_legacy_findmap(dev, init->buffers_offset);
 	if (!dev->agp_buffer_map) {
 		DRM_ERROR("could not find dma buffer region!\n");
 		dev->dev_private = (void *)dev_priv;
@@ -492,7 +492,7 @@ static int r128_do_init_cce(struct drm_device *dev, drm_r128_init_t *init)
 
 	if (!dev_priv->is_pci) {
 		dev_priv->agp_textures =
-		    drm_core_findmap(dev, init->agp_textures_offset);
+		    drm_legacy_findmap(dev, init->agp_textures_offset);
 		if (!dev_priv->agp_textures) {
 			DRM_ERROR("could not find agp texture region!\n");
 			dev->dev_private = (void *)dev_priv;
@@ -507,9 +507,9 @@ static int r128_do_init_cce(struct drm_device *dev, drm_r128_init_t *init)
 
 #if __OS_HAS_AGP
 	if (!dev_priv->is_pci) {
-		drm_core_ioremap_wc(dev_priv->cce_ring, dev);
-		drm_core_ioremap_wc(dev_priv->ring_rptr, dev);
-		drm_core_ioremap_wc(dev->agp_buffer_map, dev);
+		drm_legacy_ioremap_wc(dev_priv->cce_ring, dev);
+		drm_legacy_ioremap_wc(dev_priv->ring_rptr, dev);
+		drm_legacy_ioremap_wc(dev->agp_buffer_map, dev);
 		if (!dev_priv->cce_ring->handle ||
 		    !dev_priv->ring_rptr->handle ||
 		    !dev->agp_buffer_map->handle) {
@@ -603,11 +603,11 @@ int r128_do_cleanup_cce(struct drm_device *dev)
 #if __OS_HAS_AGP
 		if (!dev_priv->is_pci) {
 			if (dev_priv->cce_ring != NULL)
-				drm_core_ioremapfree(dev_priv->cce_ring, dev);
+				drm_legacy_ioremapfree(dev_priv->cce_ring, dev);
 			if (dev_priv->ring_rptr != NULL)
-				drm_core_ioremapfree(dev_priv->ring_rptr, dev);
+				drm_legacy_ioremapfree(dev_priv->ring_rptr, dev);
 			if (dev->agp_buffer_map != NULL) {
-				drm_core_ioremapfree(dev->agp_buffer_map, dev);
+				drm_legacy_ioremapfree(dev->agp_buffer_map, dev);
 				dev->agp_buffer_map = NULL;
 			}
 		} else
diff --git a/drivers/gpu/drm/r128/r128_drv.c b/drivers/gpu/drm/r128/r128_drv.c
index 5bd307cd8da1..c57b4de63caf 100644
--- a/drivers/gpu/drm/r128/r128_drv.c
+++ b/drivers/gpu/drm/r128/r128_drv.c
@@ -46,7 +46,7 @@ static const struct file_operations r128_driver_fops = {
 	.open = drm_open,
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_mmap,
+	.mmap = drm_legacy_mmap,
 	.poll = drm_poll,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl = r128_compat_ioctl,
@@ -62,6 +62,7 @@ static struct drm_driver driver = {
 	.load = r128_driver_load,
 	.preclose = r128_driver_preclose,
 	.lastclose = r128_driver_lastclose,
+	.set_busid = drm_pci_set_busid,
 	.get_vblank_counter = r128_get_vblank_counter,
 	.enable_vblank = r128_enable_vblank,
 	.disable_vblank = r128_disable_vblank,
diff --git a/drivers/gpu/drm/r128/r128_drv.h b/drivers/gpu/drm/r128/r128_drv.h
index 5bf3f5ff805d..723e5d6f10a4 100644
--- a/drivers/gpu/drm/r128/r128_drv.h
+++ b/drivers/gpu/drm/r128/r128_drv.h
@@ -35,6 +35,9 @@
 #ifndef __R128_DRV_H__
 #define __R128_DRV_H__
 
+#include <drm/ati_pcigart.h>
+#include <drm/drm_legacy.h>
+
 /* General customization:
  */
 #define DRIVER_AUTHOR		"Gareth Hughes, VA Linux Systems Inc."
diff --git a/drivers/gpu/drm/r128/r128_state.c b/drivers/gpu/drm/r128/r128_state.c
index 575e986f82a7..8fd2d9f58f77 100644
--- a/drivers/gpu/drm/r128/r128_state.c
+++ b/drivers/gpu/drm/r128/r128_state.c
@@ -905,7 +905,7 @@ static int r128_cce_dispatch_write_span(struct drm_device *dev,
 	if (IS_ERR(buffer))
 		return PTR_ERR(buffer);
 
-	mask_size = depth->n * sizeof(u8);
+	mask_size = depth->n;
 	if (depth->mask) {
 		mask = memdup_user(depth->mask, mask_size);
 		if (IS_ERR(mask)) {
@@ -1010,7 +1010,7 @@ static int r128_cce_dispatch_write_pixels(struct drm_device *dev,
 	}
 
 	if (depth->mask) {
-		mask_size = depth->n * sizeof(u8);
+		mask_size = depth->n;
 		mask = memdup_user(depth->mask, mask_size);
 		if (IS_ERR(mask)) {
 			kfree(x);
diff --git a/drivers/gpu/drm/radeon/Makefile b/drivers/gpu/drm/radeon/Makefile
index f77b7135ee4c..12bc21219a0e 100644
--- a/drivers/gpu/drm/radeon/Makefile
+++ b/drivers/gpu/drm/radeon/Makefile
@@ -60,7 +60,7 @@ radeon-y := radeon_drv.o
 
 # add UMS driver
 radeon-$(CONFIG_DRM_RADEON_UMS)+= radeon_cp.o radeon_state.o radeon_mem.o \
-	radeon_irq.o r300_cmdbuf.o r600_cp.o r600_blit.o
+	radeon_irq.o r300_cmdbuf.o r600_cp.o r600_blit.o drm_buffer.o
 
 # add KMS driver
 radeon-y += radeon_device.o radeon_asic.o radeon_kms.o \
@@ -72,7 +72,7 @@ radeon-y += radeon_device.o radeon_asic.o radeon_kms.o \
 	radeon_cs.o radeon_bios.o radeon_benchmark.o r100.o r300.o r420.o \
 	rs400.o rs600.o rs690.o rv515.o r520.o r600.o rv770.o radeon_test.o \
 	r200.o radeon_legacy_tv.o r600_cs.o r600_blit_shaders.o \
-	radeon_pm.o atombios_dp.o r600_audio.o r600_hdmi.o dce3_1_afmt.o \
+	radeon_pm.o atombios_dp.o r600_hdmi.o dce3_1_afmt.o \
 	evergreen.o evergreen_cs.o evergreen_blit_shaders.o \
 	evergreen_hdmi.o radeon_trace_points.o ni.o cayman_blit_shaders.o \
 	atombios_encoders.o radeon_semaphore.o radeon_sa.o atombios_i2c.o si.o \
@@ -80,7 +80,8 @@ radeon-y += radeon_device.o radeon_asic.o radeon_kms.o \
 	r600_dpm.o rs780_dpm.o rv6xx_dpm.o rv770_dpm.o rv730_dpm.o rv740_dpm.o \
 	rv770_smc.o cypress_dpm.o btc_dpm.o sumo_dpm.o sumo_smc.o trinity_dpm.o \
 	trinity_smc.o ni_dpm.o si_smc.o si_dpm.o kv_smc.o kv_dpm.o ci_smc.o \
-	ci_dpm.o dce6_afmt.o radeon_vm.o radeon_ucode.o radeon_ib.o
+	ci_dpm.o dce6_afmt.o radeon_vm.o radeon_ucode.o radeon_ib.o radeon_mn.o \
+	radeon_sync.o
 
 # add async DMA block
 radeon-y += \
@@ -104,6 +105,7 @@ radeon-y += \
 	radeon_vce.o \
 	vce_v1_0.o \
 	vce_v2_0.o \
+	radeon_kfd.o
 
 radeon-$(CONFIG_COMPAT) += radeon_ioc32.o
 radeon-$(CONFIG_VGA_SWITCHEROO) += radeon_atpx_handler.o
diff --git a/drivers/gpu/drm/radeon/atom.c b/drivers/gpu/drm/radeon/atom.c
index 15da7ef344a4..ec1593a6a561 100644
--- a/drivers/gpu/drm/radeon/atom.c
+++ b/drivers/gpu/drm/radeon/atom.c
@@ -1217,7 +1217,7 @@ free:
 	return ret;
 }
 
-int atom_execute_table(struct atom_context *ctx, int index, uint32_t * params)
+int atom_execute_table_scratch_unlocked(struct atom_context *ctx, int index, uint32_t * params)
 {
 	int r;
 
@@ -1238,6 +1238,15 @@ int atom_execute_table(struct atom_context *ctx, int index, uint32_t * params)
 	return r;
 }
 
+int atom_execute_table(struct atom_context *ctx, int index, uint32_t * params)
+{
+	int r;
+	mutex_lock(&ctx->scratch_mutex);
+	r = atom_execute_table_scratch_unlocked(ctx, index, params);
+	mutex_unlock(&ctx->scratch_mutex);
+	return r;
+}
+
 static int atom_iio_len[] = { 1, 2, 3, 3, 3, 3, 4, 4, 4, 3 };
 
 static void atom_index_iio(struct atom_context *ctx, int base)
diff --git a/drivers/gpu/drm/radeon/atom.h b/drivers/gpu/drm/radeon/atom.h
index feba6b8d36b3..6d014ddb6b78 100644
--- a/drivers/gpu/drm/radeon/atom.h
+++ b/drivers/gpu/drm/radeon/atom.h
@@ -125,6 +125,7 @@ struct card_info {
 struct atom_context {
 	struct card_info *card;
 	struct mutex mutex;
+	struct mutex scratch_mutex;
 	void *bios;
 	uint32_t cmd_table, data_table;
 	uint16_t *iio;
@@ -145,6 +146,7 @@ extern int atom_debug;
 
 struct atom_context *atom_parse(struct card_info *, void *);
 int atom_execute_table(struct atom_context *, int, uint32_t *);
+int atom_execute_table_scratch_unlocked(struct atom_context *, int, uint32_t *);
 int atom_asic_init(struct atom_context *);
 void atom_destroy(struct atom_context *);
 bool atom_parse_data_header(struct atom_context *ctx, int index, uint16_t *size,
diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c
index 30d242b25078..ed644a4f6f57 100644
--- a/drivers/gpu/drm/radeon/atombios_crtc.c
+++ b/drivers/gpu/drm/radeon/atombios_crtc.c
@@ -1851,10 +1851,9 @@ static int radeon_atom_pick_pll(struct drm_crtc *crtc)
 				return pll;
 		}
 		/* otherwise, pick one of the plls */
-		if ((rdev->family == CHIP_KAVERI) ||
-		    (rdev->family == CHIP_KABINI) ||
+		if ((rdev->family == CHIP_KABINI) ||
 		    (rdev->family == CHIP_MULLINS)) {
-			/* KB/KV/ML has PPLL1 and PPLL2 */
+			/* KB/ML has PPLL1 and PPLL2 */
 			pll_in_use = radeon_get_pll_use_mask(crtc);
 			if (!(pll_in_use & (1 << ATOM_PPLL2)))
 				return ATOM_PPLL2;
@@ -1863,7 +1862,7 @@ static int radeon_atom_pick_pll(struct drm_crtc *crtc)
 			DRM_ERROR("unable to allocate a PPLL\n");
 			return ATOM_PPLL_INVALID;
 		} else {
-			/* CI has PPLL0, PPLL1, and PPLL2 */
+			/* CI/KV has PPLL0, PPLL1, and PPLL2 */
 			pll_in_use = radeon_get_pll_use_mask(crtc);
 			if (!(pll_in_use & (1 << ATOM_PPLL2)))
 				return ATOM_PPLL2;
@@ -2039,6 +2038,7 @@ int atombios_crtc_mode_set(struct drm_crtc *crtc,
 	atombios_crtc_set_base(crtc, x, y, old_fb);
 	atombios_overscan_setup(crtc, mode, adjusted_mode);
 	atombios_scaler_setup(crtc);
+	radeon_cursor_reset(crtc);
 	/* update the hw version fpr dpm */
 	radeon_crtc->hw_mode = *adjusted_mode;
 
@@ -2154,6 +2154,7 @@ static void atombios_crtc_disable(struct drm_crtc *crtc)
 	case ATOM_PPLL0:
 		/* disable the ppll */
 		if ((rdev->family == CHIP_ARUBA) ||
+		    (rdev->family == CHIP_KAVERI) ||
 		    (rdev->family == CHIP_BONAIRE) ||
 		    (rdev->family == CHIP_HAWAII))
 			atombios_crtc_program_pll(crtc, radeon_crtc->crtc_id, radeon_crtc->pll_id,
diff --git a/drivers/gpu/drm/radeon/atombios_dp.c b/drivers/gpu/drm/radeon/atombios_dp.c
index ac14b67621d3..db42a670f995 100644
--- a/drivers/gpu/drm/radeon/atombios_dp.c
+++ b/drivers/gpu/drm/radeon/atombios_dp.c
@@ -100,6 +100,7 @@ static int radeon_process_aux_ch(struct radeon_i2c_chan *chan,
 	memset(&args, 0, sizeof(args));
 
 	mutex_lock(&chan->mutex);
+	mutex_lock(&rdev->mode_info.atom_context->scratch_mutex);
 
 	base = (unsigned char *)(rdev->mode_info.atom_context->scratch + 1);
 
@@ -113,7 +114,7 @@ static int radeon_process_aux_ch(struct radeon_i2c_chan *chan,
 	if (ASIC_IS_DCE4(rdev))
 		args.v2.ucHPD_ID = chan->rec.hpd;
 
-	atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
+	atom_execute_table_scratch_unlocked(rdev->mode_info.atom_context, index, (uint32_t *)&args);
 
 	*ack = args.v1.ucReplyStatus;
 
@@ -147,6 +148,7 @@ static int radeon_process_aux_ch(struct radeon_i2c_chan *chan,
 
 	r = recv_bytes;
 done:
+	mutex_unlock(&rdev->mode_info.atom_context->scratch_mutex);
 	mutex_unlock(&chan->mutex);
 
 	return r;
@@ -232,8 +234,8 @@ void radeon_dp_aux_init(struct radeon_connector *radeon_connector)
 
 /***** general DP utility functions *****/
 
-#define DP_VOLTAGE_MAX         DP_TRAIN_VOLTAGE_SWING_1200
-#define DP_PRE_EMPHASIS_MAX    DP_TRAIN_PRE_EMPHASIS_9_5
+#define DP_VOLTAGE_MAX         DP_TRAIN_VOLTAGE_SWING_LEVEL_3
+#define DP_PRE_EMPHASIS_MAX    DP_TRAIN_PRE_EMPH_LEVEL_3
 
 static void dp_get_adjust_train(u8 link_status[DP_LINK_STATUS_SIZE],
 				int lane_count,
@@ -490,6 +492,10 @@ int radeon_dp_mode_valid_helper(struct drm_connector *connector,
 	struct radeon_connector_atom_dig *dig_connector;
 	int dp_clock;
 
+	if ((mode->clock > 340000) &&
+	    (!radeon_connector_is_dp12_capable(connector)))
+		return MODE_CLOCK_HIGH;
+
 	if (!radeon_connector->con_priv)
 		return MODE_CLOCK_HIGH;
 	dig_connector = radeon_connector->con_priv;
diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c
index a7f2ddf09a9d..b8cd7975f797 100644
--- a/drivers/gpu/drm/radeon/atombios_encoders.c
+++ b/drivers/gpu/drm/radeon/atombios_encoders.c
@@ -291,29 +291,6 @@ static void radeon_atom_backlight_exit(struct radeon_encoder *encoder)
 bool radeon_atom_get_tv_timings(struct radeon_device *rdev, int index,
 				struct drm_display_mode *mode);
 
-
-static inline bool radeon_encoder_is_digital(struct drm_encoder *encoder)
-{
-	struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
-	switch (radeon_encoder->encoder_id) {
-	case ENCODER_OBJECT_ID_INTERNAL_LVDS:
-	case ENCODER_OBJECT_ID_INTERNAL_TMDS1:
-	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_TMDS1:
-	case ENCODER_OBJECT_ID_INTERNAL_LVTM1:
-	case ENCODER_OBJECT_ID_INTERNAL_DVO1:
-	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1:
-	case ENCODER_OBJECT_ID_INTERNAL_DDI:
-	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY:
-	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_LVTMA:
-	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1:
-	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2:
-	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY3:
-		return true;
-	default:
-		return false;
-	}
-}
-
 static bool radeon_atom_mode_fixup(struct drm_encoder *encoder,
 				   const struct drm_display_mode *mode,
 				   struct drm_display_mode *adjusted_mode)
diff --git a/drivers/gpu/drm/radeon/atombios_i2c.c b/drivers/gpu/drm/radeon/atombios_i2c.c
index 9c570fb15b8c..4157780585a0 100644
--- a/drivers/gpu/drm/radeon/atombios_i2c.c
+++ b/drivers/gpu/drm/radeon/atombios_i2c.c
@@ -48,6 +48,7 @@ static int radeon_process_i2c_ch(struct radeon_i2c_chan *chan,
 	memset(&args, 0, sizeof(args));
 
 	mutex_lock(&chan->mutex);
+	mutex_lock(&rdev->mode_info.atom_context->scratch_mutex);
 
 	base = (unsigned char *)rdev->mode_info.atom_context->scratch;
 
@@ -82,7 +83,7 @@ static int radeon_process_i2c_ch(struct radeon_i2c_chan *chan,
 	args.ucSlaveAddr = slave_addr << 1;
 	args.ucLineNumber = chan->rec.i2c_id;
 
-	atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
+	atom_execute_table_scratch_unlocked(rdev->mode_info.atom_context, index, (uint32_t *)&args);
 
 	/* error */
 	if (args.ucStatus != HW_ASSISTED_I2C_STATUS_SUCCESS) {
@@ -95,6 +96,7 @@ static int radeon_process_i2c_ch(struct radeon_i2c_chan *chan,
 		radeon_atom_copy_swap(buf, base, num, false);
 
 done:
+	mutex_unlock(&rdev->mode_info.atom_context->scratch_mutex);
 	mutex_unlock(&chan->mutex);
 
 	return r;
diff --git a/drivers/gpu/drm/radeon/btc_dpm.c b/drivers/gpu/drm/radeon/btc_dpm.c
index f81d7ca134db..0b2929de9f41 100644
--- a/drivers/gpu/drm/radeon/btc_dpm.c
+++ b/drivers/gpu/drm/radeon/btc_dpm.c
@@ -24,6 +24,7 @@
 
 #include "drmP.h"
 #include "radeon.h"
+#include "radeon_asic.h"
 #include "btcd.h"
 #include "r600_dpm.h"
 #include "cypress_dpm.h"
@@ -2099,7 +2100,6 @@ static void btc_apply_state_adjust_rules(struct radeon_device *rdev,
 	bool disable_mclk_switching;
 	u32 mclk, sclk;
 	u16 vddc, vddci;
-	u32 max_sclk_vddc, max_mclk_vddci, max_mclk_vddc;
 
 	if ((rdev->pm.dpm.new_active_crtc_count > 1) ||
 	    btc_dpm_vblank_too_short(rdev))
@@ -2141,39 +2141,6 @@ static void btc_apply_state_adjust_rules(struct radeon_device *rdev,
 			ps->low.vddci = max_limits->vddci;
 	}
 
-	/* limit clocks to max supported clocks based on voltage dependency tables */
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_sclk,
-							&max_sclk_vddc);
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddci_dependency_on_mclk,
-							&max_mclk_vddci);
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_mclk,
-							&max_mclk_vddc);
-
-	if (max_sclk_vddc) {
-		if (ps->low.sclk > max_sclk_vddc)
-			ps->low.sclk = max_sclk_vddc;
-		if (ps->medium.sclk > max_sclk_vddc)
-			ps->medium.sclk = max_sclk_vddc;
-		if (ps->high.sclk > max_sclk_vddc)
-			ps->high.sclk = max_sclk_vddc;
-	}
-	if (max_mclk_vddci) {
-		if (ps->low.mclk > max_mclk_vddci)
-			ps->low.mclk = max_mclk_vddci;
-		if (ps->medium.mclk > max_mclk_vddci)
-			ps->medium.mclk = max_mclk_vddci;
-		if (ps->high.mclk > max_mclk_vddci)
-			ps->high.mclk = max_mclk_vddci;
-	}
-	if (max_mclk_vddc) {
-		if (ps->low.mclk > max_mclk_vddc)
-			ps->low.mclk = max_mclk_vddc;
-		if (ps->medium.mclk > max_mclk_vddc)
-			ps->medium.mclk = max_mclk_vddc;
-		if (ps->high.mclk > max_mclk_vddc)
-			ps->high.mclk = max_mclk_vddc;
-	}
-
 	/* XXX validate the min clocks required for display */
 
 	if (disable_mclk_switching) {
diff --git a/drivers/gpu/drm/radeon/ci_dpm.c b/drivers/gpu/drm/radeon/ci_dpm.c
index d416bb2ff48d..f373a81ba3d5 100644
--- a/drivers/gpu/drm/radeon/ci_dpm.c
+++ b/drivers/gpu/drm/radeon/ci_dpm.c
@@ -24,6 +24,7 @@
 #include <linux/firmware.h>
 #include "drmP.h"
 #include "radeon.h"
+#include "radeon_asic.h"
 #include "radeon_ucode.h"
 #include "cikd.h"
 #include "r600_dpm.h"
@@ -45,15 +46,15 @@
 static const struct ci_pt_defaults defaults_hawaii_xt =
 {
 	1, 0xF, 0xFD, 0x19, 5, 0x14, 0, 0xB0000,
-	{ 0x84,  0x0,   0x0,   0x7F,  0x0,   0x0,   0x5A,  0x60,  0x51,  0x8E,  0x79,  0x6B,  0x5F,  0x90,  0x79  },
-	{ 0x1EA, 0x1EA, 0x1EA, 0x224, 0x224, 0x224, 0x24F, 0x24F, 0x24F, 0x28E, 0x28E, 0x28E, 0x2BC, 0x2BC, 0x2BC }
+	{ 0x2E,  0x00,  0x00,  0x88,  0x00,  0x00,  0x72,  0x60,  0x51,  0xA7,  0x79,  0x6B,  0x90,  0xBD,  0x79  },
+	{ 0x217, 0x217, 0x217, 0x242, 0x242, 0x242, 0x269, 0x269, 0x269, 0x2A1, 0x2A1, 0x2A1, 0x2C9, 0x2C9, 0x2C9 }
 };
 
 static const struct ci_pt_defaults defaults_hawaii_pro =
 {
 	1, 0xF, 0xFD, 0x19, 5, 0x14, 0, 0x65062,
-	{ 0x93,  0x0,   0x0,   0x97,  0x0,   0x0,   0x6B,  0x60,  0x51,  0x95,  0x79,  0x6B,  0x5F,  0x90,  0x79  },
-	{ 0x1EA, 0x1EA, 0x1EA, 0x224, 0x224, 0x224, 0x24F, 0x24F, 0x24F, 0x28E, 0x28E, 0x28E, 0x2BC, 0x2BC, 0x2BC }
+	{ 0x2E,  0x00,  0x00,  0x88,  0x00,  0x00,  0x72,  0x60,  0x51,  0xA7,  0x79,  0x6B,  0x90,  0xBD,  0x79  },
+	{ 0x217, 0x217, 0x217, 0x242, 0x242, 0x242, 0x269, 0x269, 0x269, 0x2A1, 0x2A1, 0x2A1, 0x2C9, 0x2C9, 0x2C9 }
 };
 
 static const struct ci_pt_defaults defaults_bonaire_xt =
@@ -162,8 +163,6 @@ static const struct ci_pt_config_reg didt_config_ci[] =
 };
 
 extern u8 rv770_get_memory_module_index(struct radeon_device *rdev);
-extern void btc_get_max_clock_from_voltage_dependency_table(struct radeon_clock_voltage_dependency_table *table,
-							    u32 *max_clock);
 extern int ni_copy_and_switch_arb_sets(struct radeon_device *rdev,
 				       u32 arb_freq_src, u32 arb_freq_dest);
 extern u8 si_get_ddr3_mclk_frequency_ratio(u32 memory_clock);
@@ -185,6 +184,9 @@ static int ci_set_overdrive_target_tdp(struct radeon_device *rdev,
 				       u32 target_tdp);
 static int ci_update_uvd_dpm(struct radeon_device *rdev, bool gate);
 
+static PPSMC_Result ci_send_msg_to_smc_with_parameter(struct radeon_device *rdev,
+						      PPSMC_Msg msg, u32 parameter);
+
 static struct ci_power_info *ci_get_pi(struct radeon_device *rdev)
 {
         struct ci_power_info *pi = rdev->pm.dpm.priv;
@@ -250,7 +252,10 @@ static void ci_initialize_powertune_defaults(struct radeon_device *rdev)
 
 	if (pi->caps_power_containment) {
 		pi->caps_cac = true;
-		pi->enable_bapm_feature = true;
+		if (rdev->family == CHIP_HAWAII)
+			pi->enable_bapm_feature = false;
+		else
+			pi->enable_bapm_feature = true;
 		pi->enable_tdc_limit_feature = true;
 		pi->enable_pkg_pwr_tracking_feature = true;
 	}
@@ -353,6 +358,21 @@ static int ci_populate_dw8(struct radeon_device *rdev)
 	return 0;
 }
 
+static int ci_populate_fuzzy_fan(struct radeon_device *rdev)
+{
+	struct ci_power_info *pi = ci_get_pi(rdev);
+
+	if ((rdev->pm.dpm.fan.fan_output_sensitivity & (1 << 15)) ||
+	    (rdev->pm.dpm.fan.fan_output_sensitivity == 0))
+		rdev->pm.dpm.fan.fan_output_sensitivity =
+			rdev->pm.dpm.fan.default_fan_output_sensitivity;
+
+	pi->smc_powertune_table.FuzzyFan_PwmSetDelta =
+		cpu_to_be16(rdev->pm.dpm.fan.fan_output_sensitivity);
+
+	return 0;
+}
+
 static int ci_min_max_v_gnbl_pm_lid_from_bapm_vddc(struct radeon_device *rdev)
 {
 	struct ci_power_info *pi = ci_get_pi(rdev);
@@ -478,6 +498,9 @@ static int ci_populate_pm_base(struct radeon_device *rdev)
 		ret = ci_populate_dw8(rdev);
 		if (ret)
 			return ret;
+		ret = ci_populate_fuzzy_fan(rdev);
+		if (ret)
+			return ret;
 		ret = ci_min_max_v_gnbl_pm_lid_from_bapm_vddc(rdev);
 		if (ret)
 			return ret;
@@ -691,6 +714,25 @@ static int ci_enable_smc_cac(struct radeon_device *rdev, bool enable)
 	return ret;
 }
 
+static int ci_enable_thermal_based_sclk_dpm(struct radeon_device *rdev,
+					    bool enable)
+{
+	struct ci_power_info *pi = ci_get_pi(rdev);
+	PPSMC_Result smc_result = PPSMC_Result_OK;
+
+	if (pi->thermal_sclk_dpm_enabled) {
+		if (enable)
+			smc_result = ci_send_msg_to_smc(rdev, PPSMC_MSG_ENABLE_THERMAL_DPM);
+		else
+			smc_result = ci_send_msg_to_smc(rdev, PPSMC_MSG_DISABLE_THERMAL_DPM);
+	}
+
+	if (smc_result == PPSMC_Result_OK)
+		return 0;
+	else
+		return -EINVAL;
+}
+
 static int ci_power_control_set_level(struct radeon_device *rdev)
 {
 	struct ci_power_info *pi = ci_get_pi(rdev);
@@ -701,13 +743,11 @@ static int ci_power_control_set_level(struct radeon_device *rdev)
 	int ret = 0;
 	bool adjust_polarity = false; /* ??? */
 
-	if (pi->caps_power_containment &&
-	    (pi->power_containment_features & POWERCONTAINMENT_FEATURE_BAPM)) {
+	if (pi->caps_power_containment) {
 		adjust_percent = adjust_polarity ?
 			rdev->pm.dpm.tdp_adjustment : (-1 * rdev->pm.dpm.tdp_adjustment);
 		target_tdp = ((100 + adjust_percent) *
 			      (s32)cac_tdp_table->configurable_tdp) / 100;
-		target_tdp *= 256;
 
 		ret = ci_set_overdrive_target_tdp(rdev, (u32)target_tdp);
 	}
@@ -748,7 +788,6 @@ static void ci_apply_state_adjust_rules(struct radeon_device *rdev,
 	struct radeon_clock_and_voltage_limits *max_limits;
 	bool disable_mclk_switching;
 	u32 sclk, mclk;
-	u32 max_sclk_vddc, max_mclk_vddci, max_mclk_vddc;
 	int i;
 
 	if (rps->vce_active) {
@@ -784,29 +823,6 @@ static void ci_apply_state_adjust_rules(struct radeon_device *rdev,
 		}
 	}
 
-	/* limit clocks to max supported clocks based on voltage dependency tables */
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_sclk,
-							&max_sclk_vddc);
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddci_dependency_on_mclk,
-							&max_mclk_vddci);
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_mclk,
-							&max_mclk_vddc);
-
-	for (i = 0; i < ps->performance_level_count; i++) {
-		if (max_sclk_vddc) {
-			if (ps->performance_levels[i].sclk > max_sclk_vddc)
-				ps->performance_levels[i].sclk = max_sclk_vddc;
-		}
-		if (max_mclk_vddci) {
-			if (ps->performance_levels[i].mclk > max_mclk_vddci)
-				ps->performance_levels[i].mclk = max_mclk_vddci;
-		}
-		if (max_mclk_vddc) {
-			if (ps->performance_levels[i].mclk > max_mclk_vddc)
-				ps->performance_levels[i].mclk = max_mclk_vddc;
-		}
-	}
-
 	/* XXX validate the min clocks required for display */
 
 	if (disable_mclk_switching) {
@@ -839,7 +855,7 @@ static void ci_apply_state_adjust_rules(struct radeon_device *rdev,
 	}
 }
 
-static int ci_set_thermal_temperature_range(struct radeon_device *rdev,
+static int ci_thermal_set_temperature_range(struct radeon_device *rdev,
 					    int min_temp, int max_temp)
 {
 	int low_temp = 0 * 1000;
@@ -875,6 +891,350 @@ static int ci_set_thermal_temperature_range(struct radeon_device *rdev,
 	return 0;
 }
 
+static int ci_thermal_enable_alert(struct radeon_device *rdev,
+				   bool enable)
+{
+	u32 thermal_int = RREG32_SMC(CG_THERMAL_INT);
+	PPSMC_Result result;
+
+	if (enable) {
+		thermal_int &= ~(THERM_INT_MASK_HIGH | THERM_INT_MASK_LOW);
+		WREG32_SMC(CG_THERMAL_INT, thermal_int);
+		rdev->irq.dpm_thermal = false;
+		result = ci_send_msg_to_smc(rdev, PPSMC_MSG_Thermal_Cntl_Enable);
+		if (result != PPSMC_Result_OK) {
+			DRM_DEBUG_KMS("Could not enable thermal interrupts.\n");
+			return -EINVAL;
+		}
+	} else {
+		thermal_int |= THERM_INT_MASK_HIGH | THERM_INT_MASK_LOW;
+		WREG32_SMC(CG_THERMAL_INT, thermal_int);
+		rdev->irq.dpm_thermal = true;
+		result = ci_send_msg_to_smc(rdev, PPSMC_MSG_Thermal_Cntl_Disable);
+		if (result != PPSMC_Result_OK) {
+			DRM_DEBUG_KMS("Could not disable thermal interrupts.\n");
+			return -EINVAL;
+		}
+	}
+
+	return 0;
+}
+
+static void ci_fan_ctrl_set_static_mode(struct radeon_device *rdev, u32 mode)
+{
+	struct ci_power_info *pi = ci_get_pi(rdev);
+	u32 tmp;
+
+	if (pi->fan_ctrl_is_in_default_mode) {
+		tmp = (RREG32_SMC(CG_FDO_CTRL2) & FDO_PWM_MODE_MASK) >> FDO_PWM_MODE_SHIFT;
+		pi->fan_ctrl_default_mode = tmp;
+		tmp = (RREG32_SMC(CG_FDO_CTRL2) & TMIN_MASK) >> TMIN_SHIFT;
+		pi->t_min = tmp;
+		pi->fan_ctrl_is_in_default_mode = false;
+	}
+
+	tmp = RREG32_SMC(CG_FDO_CTRL2) & ~TMIN_MASK;
+	tmp |= TMIN(0);
+	WREG32_SMC(CG_FDO_CTRL2, tmp);
+
+	tmp = RREG32_SMC(CG_FDO_CTRL2) & ~FDO_PWM_MODE_MASK;
+	tmp |= FDO_PWM_MODE(mode);
+	WREG32_SMC(CG_FDO_CTRL2, tmp);
+}
+
+static int ci_thermal_setup_fan_table(struct radeon_device *rdev)
+{
+	struct ci_power_info *pi = ci_get_pi(rdev);
+	SMU7_Discrete_FanTable fan_table = { FDO_MODE_HARDWARE };
+	u32 duty100;
+	u32 t_diff1, t_diff2, pwm_diff1, pwm_diff2;
+	u16 fdo_min, slope1, slope2;
+	u32 reference_clock, tmp;
+	int ret;
+	u64 tmp64;
+
+	if (!pi->fan_table_start) {
+		rdev->pm.dpm.fan.ucode_fan_control = false;
+		return 0;
+	}
+
+	duty100 = (RREG32_SMC(CG_FDO_CTRL1) & FMAX_DUTY100_MASK) >> FMAX_DUTY100_SHIFT;
+
+	if (duty100 == 0) {
+		rdev->pm.dpm.fan.ucode_fan_control = false;
+		return 0;
+	}
+
+	tmp64 = (u64)rdev->pm.dpm.fan.pwm_min * duty100;
+	do_div(tmp64, 10000);
+	fdo_min = (u16)tmp64;
+
+	t_diff1 = rdev->pm.dpm.fan.t_med - rdev->pm.dpm.fan.t_min;
+	t_diff2 = rdev->pm.dpm.fan.t_high - rdev->pm.dpm.fan.t_med;
+
+	pwm_diff1 = rdev->pm.dpm.fan.pwm_med - rdev->pm.dpm.fan.pwm_min;
+	pwm_diff2 = rdev->pm.dpm.fan.pwm_high - rdev->pm.dpm.fan.pwm_med;
+
+	slope1 = (u16)((50 + ((16 * duty100 * pwm_diff1) / t_diff1)) / 100);
+	slope2 = (u16)((50 + ((16 * duty100 * pwm_diff2) / t_diff2)) / 100);
+
+	fan_table.TempMin = cpu_to_be16((50 + rdev->pm.dpm.fan.t_min) / 100);
+	fan_table.TempMed = cpu_to_be16((50 + rdev->pm.dpm.fan.t_med) / 100);
+	fan_table.TempMax = cpu_to_be16((50 + rdev->pm.dpm.fan.t_max) / 100);
+
+	fan_table.Slope1 = cpu_to_be16(slope1);
+	fan_table.Slope2 = cpu_to_be16(slope2);
+
+	fan_table.FdoMin = cpu_to_be16(fdo_min);
+
+	fan_table.HystDown = cpu_to_be16(rdev->pm.dpm.fan.t_hyst);
+
+	fan_table.HystUp = cpu_to_be16(1);
+
+	fan_table.HystSlope = cpu_to_be16(1);
+
+	fan_table.TempRespLim = cpu_to_be16(5);
+
+	reference_clock = radeon_get_xclk(rdev);
+
+	fan_table.RefreshPeriod = cpu_to_be32((rdev->pm.dpm.fan.cycle_delay *
+					       reference_clock) / 1600);
+
+	fan_table.FdoMax = cpu_to_be16((u16)duty100);
+
+	tmp = (RREG32_SMC(CG_MULT_THERMAL_CTRL) & TEMP_SEL_MASK) >> TEMP_SEL_SHIFT;
+	fan_table.TempSrc = (uint8_t)tmp;
+
+	ret = ci_copy_bytes_to_smc(rdev,
+				   pi->fan_table_start,
+				   (u8 *)(&fan_table),
+				   sizeof(fan_table),
+				   pi->sram_end);
+
+	if (ret) {
+		DRM_ERROR("Failed to load fan table to the SMC.");
+		rdev->pm.dpm.fan.ucode_fan_control = false;
+	}
+
+	return 0;
+}
+
+static int ci_fan_ctrl_start_smc_fan_control(struct radeon_device *rdev)
+{
+	struct ci_power_info *pi = ci_get_pi(rdev);
+	PPSMC_Result ret;
+
+	if (pi->caps_od_fuzzy_fan_control_support) {
+		ret = ci_send_msg_to_smc_with_parameter(rdev,
+							PPSMC_StartFanControl,
+							FAN_CONTROL_FUZZY);
+		if (ret != PPSMC_Result_OK)
+			return -EINVAL;
+		ret = ci_send_msg_to_smc_with_parameter(rdev,
+							PPSMC_MSG_SetFanPwmMax,
+							rdev->pm.dpm.fan.default_max_fan_pwm);
+		if (ret != PPSMC_Result_OK)
+			return -EINVAL;
+	} else {
+		ret = ci_send_msg_to_smc_with_parameter(rdev,
+							PPSMC_StartFanControl,
+							FAN_CONTROL_TABLE);
+		if (ret != PPSMC_Result_OK)
+			return -EINVAL;
+	}
+
+	return 0;
+}
+
+#if 0
+static int ci_fan_ctrl_stop_smc_fan_control(struct radeon_device *rdev)
+{
+	PPSMC_Result ret;
+
+	ret = ci_send_msg_to_smc(rdev, PPSMC_StopFanControl);
+	if (ret == PPSMC_Result_OK)
+		return 0;
+	else
+		return -EINVAL;
+}
+
+static int ci_fan_ctrl_get_fan_speed_percent(struct radeon_device *rdev,
+					     u32 *speed)
+{
+	u32 duty, duty100;
+	u64 tmp64;
+
+	if (rdev->pm.no_fan)
+		return -ENOENT;
+
+	duty100 = (RREG32_SMC(CG_FDO_CTRL1) & FMAX_DUTY100_MASK) >> FMAX_DUTY100_SHIFT;
+	duty = (RREG32_SMC(CG_THERMAL_STATUS) & FDO_PWM_DUTY_MASK) >> FDO_PWM_DUTY_SHIFT;
+
+	if (duty100 == 0)
+		return -EINVAL;
+
+	tmp64 = (u64)duty * 100;
+	do_div(tmp64, duty100);
+	*speed = (u32)tmp64;
+
+	if (*speed > 100)
+		*speed = 100;
+
+	return 0;
+}
+
+static int ci_fan_ctrl_set_fan_speed_percent(struct radeon_device *rdev,
+					     u32 speed)
+{
+	u32 tmp;
+	u32 duty, duty100;
+	u64 tmp64;
+
+	if (rdev->pm.no_fan)
+		return -ENOENT;
+
+	if (speed > 100)
+		return -EINVAL;
+
+	if (rdev->pm.dpm.fan.ucode_fan_control)
+		ci_fan_ctrl_stop_smc_fan_control(rdev);
+
+	duty100 = (RREG32_SMC(CG_FDO_CTRL1) & FMAX_DUTY100_MASK) >> FMAX_DUTY100_SHIFT;
+
+	if (duty100 == 0)
+		return -EINVAL;
+
+	tmp64 = (u64)speed * duty100;
+	do_div(tmp64, 100);
+	duty = (u32)tmp64;
+
+	tmp = RREG32_SMC(CG_FDO_CTRL0) & ~FDO_STATIC_DUTY_MASK;
+	tmp |= FDO_STATIC_DUTY(duty);
+	WREG32_SMC(CG_FDO_CTRL0, tmp);
+
+	ci_fan_ctrl_set_static_mode(rdev, FDO_PWM_MODE_STATIC);
+
+	return 0;
+}
+
+static int ci_fan_ctrl_get_fan_speed_rpm(struct radeon_device *rdev,
+					 u32 *speed)
+{
+	u32 tach_period;
+	u32 xclk = radeon_get_xclk(rdev);
+
+	if (rdev->pm.no_fan)
+		return -ENOENT;
+
+	if (rdev->pm.fan_pulses_per_revolution == 0)
+		return -ENOENT;
+
+	tach_period = (RREG32_SMC(CG_TACH_STATUS) & TACH_PERIOD_MASK) >> TACH_PERIOD_SHIFT;
+	if (tach_period == 0)
+		return -ENOENT;
+
+	*speed = 60 * xclk * 10000 / tach_period;
+
+	return 0;
+}
+
+static int ci_fan_ctrl_set_fan_speed_rpm(struct radeon_device *rdev,
+					 u32 speed)
+{
+	u32 tach_period, tmp;
+	u32 xclk = radeon_get_xclk(rdev);
+
+	if (rdev->pm.no_fan)
+		return -ENOENT;
+
+	if (rdev->pm.fan_pulses_per_revolution == 0)
+		return -ENOENT;
+
+	if ((speed < rdev->pm.fan_min_rpm) ||
+	    (speed > rdev->pm.fan_max_rpm))
+		return -EINVAL;
+
+	if (rdev->pm.dpm.fan.ucode_fan_control)
+		ci_fan_ctrl_stop_smc_fan_control(rdev);
+
+	tach_period = 60 * xclk * 10000 / (8 * speed);
+	tmp = RREG32_SMC(CG_TACH_CTRL) & ~TARGET_PERIOD_MASK;
+	tmp |= TARGET_PERIOD(tach_period);
+	WREG32_SMC(CG_TACH_CTRL, tmp);
+
+	ci_fan_ctrl_set_static_mode(rdev, FDO_PWM_MODE_STATIC_RPM);
+
+	return 0;
+}
+#endif
+
+static void ci_fan_ctrl_set_default_mode(struct radeon_device *rdev)
+{
+	struct ci_power_info *pi = ci_get_pi(rdev);
+	u32 tmp;
+
+	if (!pi->fan_ctrl_is_in_default_mode) {
+		tmp = RREG32_SMC(CG_FDO_CTRL2) & ~FDO_PWM_MODE_MASK;
+		tmp |= FDO_PWM_MODE(pi->fan_ctrl_default_mode);
+		WREG32_SMC(CG_FDO_CTRL2, tmp);
+
+		tmp = RREG32_SMC(CG_FDO_CTRL2) & ~TMIN_MASK;
+		tmp |= TMIN(pi->t_min);
+		WREG32_SMC(CG_FDO_CTRL2, tmp);
+		pi->fan_ctrl_is_in_default_mode = true;
+	}
+}
+
+static void ci_thermal_start_smc_fan_control(struct radeon_device *rdev)
+{
+	if (rdev->pm.dpm.fan.ucode_fan_control) {
+		ci_fan_ctrl_start_smc_fan_control(rdev);
+		ci_fan_ctrl_set_static_mode(rdev, FDO_PWM_MODE_STATIC);
+	}
+}
+
+static void ci_thermal_initialize(struct radeon_device *rdev)
+{
+	u32 tmp;
+
+	if (rdev->pm.fan_pulses_per_revolution) {
+		tmp = RREG32_SMC(CG_TACH_CTRL) & ~EDGE_PER_REV_MASK;
+		tmp |= EDGE_PER_REV(rdev->pm.fan_pulses_per_revolution -1);
+		WREG32_SMC(CG_TACH_CTRL, tmp);
+	}
+
+	tmp = RREG32_SMC(CG_FDO_CTRL2) & ~TACH_PWM_RESP_RATE_MASK;
+	tmp |= TACH_PWM_RESP_RATE(0x28);
+	WREG32_SMC(CG_FDO_CTRL2, tmp);
+}
+
+static int ci_thermal_start_thermal_controller(struct radeon_device *rdev)
+{
+	int ret;
+
+	ci_thermal_initialize(rdev);
+	ret = ci_thermal_set_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
+	if (ret)
+		return ret;
+	ret = ci_thermal_enable_alert(rdev, true);
+	if (ret)
+		return ret;
+	if (rdev->pm.dpm.fan.ucode_fan_control) {
+		ret = ci_thermal_setup_fan_table(rdev);
+		if (ret)
+			return ret;
+		ci_thermal_start_smc_fan_control(rdev);
+	}
+
+	return 0;
+}
+
+static void ci_thermal_stop_thermal_controller(struct radeon_device *rdev)
+{
+	if (!rdev->pm.no_fan)
+		ci_fan_ctrl_set_default_mode(rdev);
+}
+
 #if 0
 static int ci_read_smc_soft_register(struct radeon_device *rdev,
 				     u16 reg_offset, u32 *value)
@@ -1278,7 +1638,7 @@ static int ci_dpm_force_state_sclk(struct radeon_device *rdev, u32 n)
 
 	if (!pi->sclk_dpm_key_disabled) {
 		PPSMC_Result smc_result =
-			ci_send_msg_to_smc_with_parameter(rdev, PPSMC_MSG_DPM_ForceState, n);
+			ci_send_msg_to_smc_with_parameter(rdev, PPSMC_MSG_SCLKDPM_SetEnabledMask, 1 << n);
 		if (smc_result != PPSMC_Result_OK)
 			return -EINVAL;
 	}
@@ -1292,7 +1652,7 @@ static int ci_dpm_force_state_mclk(struct radeon_device *rdev, u32 n)
 
 	if (!pi->mclk_dpm_key_disabled) {
 		PPSMC_Result smc_result =
-			ci_send_msg_to_smc_with_parameter(rdev, PPSMC_MSG_MCLKDPM_ForceState, n);
+			ci_send_msg_to_smc_with_parameter(rdev, PPSMC_MSG_MCLKDPM_SetEnabledMask, 1 << n);
 		if (smc_result != PPSMC_Result_OK)
 			return -EINVAL;
 	}
@@ -2067,6 +2427,33 @@ static int ci_force_switch_to_arb_f0(struct radeon_device *rdev)
 	return ni_copy_and_switch_arb_sets(rdev, tmp, MC_CG_ARB_FREQ_F0);
 }
 
+static void ci_register_patching_mc_arb(struct radeon_device *rdev,
+					const u32 engine_clock,
+					const u32 memory_clock,
+					u32 *dram_timimg2)
+{
+	bool patch;
+	u32 tmp, tmp2;
+
+	tmp = RREG32(MC_SEQ_MISC0);
+	patch = ((tmp & 0x0000f00) == 0x300) ? true : false;
+
+	if (patch &&
+	    ((rdev->pdev->device == 0x67B0) ||
+	     (rdev->pdev->device == 0x67B1))) {
+		if ((memory_clock > 100000) && (memory_clock <= 125000)) {
+			tmp2 = (((0x31 * engine_clock) / 125000) - 1) & 0xff;
+			*dram_timimg2 &= ~0x00ff0000;
+			*dram_timimg2 |= tmp2 << 16;
+		} else if ((memory_clock > 125000) && (memory_clock <= 137500)) {
+			tmp2 = (((0x36 * engine_clock) / 137500) - 1) & 0xff;
+			*dram_timimg2 &= ~0x00ff0000;
+			*dram_timimg2 |= tmp2 << 16;
+		}
+	}
+}
+
+
 static int ci_populate_memory_timing_parameters(struct radeon_device *rdev,
 						u32 sclk,
 						u32 mclk,
@@ -2082,6 +2469,8 @@ static int ci_populate_memory_timing_parameters(struct radeon_device *rdev,
 	dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2);
 	burst_time = RREG32(MC_ARB_BURST_TIME) & STATE0_MASK;
 
+	ci_register_patching_mc_arb(rdev, sclk, mclk, &dram_timing2);
+
 	arb_regs->McArbDramTiming  = cpu_to_be32(dram_timing);
 	arb_regs->McArbDramTiming2 = cpu_to_be32(dram_timing2);
 	arb_regs->McArbBurstTime = (u8)burst_time;
@@ -2376,10 +2765,10 @@ static int ci_calculate_mclk_params(struct radeon_device *rdev,
 		u32 tmp;
 		u32 reference_clock = rdev->clock.mpll.reference_freq;
 
-		if (pi->mem_gddr5)
-			freq_nom = memory_clock * 4;
+		if (mpll_param.qdr == 1)
+			freq_nom = memory_clock * 4 * (1 << mpll_param.post_div);
 		else
-			freq_nom = memory_clock * 2;
+			freq_nom = memory_clock * 2 * (1 << mpll_param.post_div);
 
 		tmp = (freq_nom / reference_clock);
 		tmp = tmp * tmp;
@@ -2459,7 +2848,6 @@ static int ci_populate_single_memory_level(struct radeon_device *rdev,
 						      &memory_level->MinVddcPhases);
 
 	memory_level->EnabledForThrottle = 1;
-	memory_level->EnabledForActivity = 1;
 	memory_level->UpH = 0;
 	memory_level->DownH = 100;
 	memory_level->VoltageDownH = 0;
@@ -2792,7 +3180,6 @@ static int ci_populate_single_graphic_level(struct radeon_device *rdev,
 
 	graphic_level->CcPwrDynRm = 0;
 	graphic_level->CcPwrDynRm1 = 0;
-	graphic_level->EnabledForActivity = 1;
 	graphic_level->EnabledForThrottle = 1;
 	graphic_level->UpH = 0;
 	graphic_level->DownH = 0;
@@ -2841,10 +3228,13 @@ static int ci_populate_all_graphic_levels(struct radeon_device *rdev)
 						       &pi->smc_state_table.GraphicsLevel[i]);
 		if (ret)
 			return ret;
+		if (i > 1)
+			pi->smc_state_table.GraphicsLevel[i].DeepSleepDivId = 0;
 		if (i == (dpm_table->sclk_table.count - 1))
 			pi->smc_state_table.GraphicsLevel[i].DisplayWatermark =
 				PPSMC_DISPLAY_WATERMARK_HIGH;
 	}
+	pi->smc_state_table.GraphicsLevel[0].EnabledForActivity = 1;
 
 	pi->smc_state_table.GraphicsDpmLevelCount = (u8)dpm_table->sclk_table.count;
 	pi->dpm_level_enable_mask.sclk_dpm_enable_mask =
@@ -2888,6 +3278,16 @@ static int ci_populate_all_memory_levels(struct radeon_device *rdev)
 			return ret;
 	}
 
+	pi->smc_state_table.MemoryLevel[0].EnabledForActivity = 1;
+
+	if ((dpm_table->mclk_table.count >= 2) &&
+	    ((rdev->pdev->device == 0x67B0) || (rdev->pdev->device == 0x67B1))) {
+		pi->smc_state_table.MemoryLevel[1].MinVddc =
+			pi->smc_state_table.MemoryLevel[0].MinVddc;
+		pi->smc_state_table.MemoryLevel[1].MinVddcPhases =
+			pi->smc_state_table.MemoryLevel[0].MinVddcPhases;
+	}
+
 	pi->smc_state_table.MemoryLevel[0].ActivityLevel = cpu_to_be16(0x1F);
 
 	pi->smc_state_table.MemoryDpmLevelCount = (u8)dpm_table->mclk_table.count;
@@ -2944,9 +3344,14 @@ static int ci_setup_default_pcie_tables(struct radeon_device *rdev)
 				  &pi->dpm_table.pcie_speed_table,
 				  SMU7_MAX_LEVELS_LINK);
 
-	ci_setup_pcie_table_entry(&pi->dpm_table.pcie_speed_table, 0,
-				  pi->pcie_gen_powersaving.min,
-				  pi->pcie_lane_powersaving.min);
+	if (rdev->family == CHIP_BONAIRE)
+		ci_setup_pcie_table_entry(&pi->dpm_table.pcie_speed_table, 0,
+					  pi->pcie_gen_powersaving.min,
+					  pi->pcie_lane_powersaving.max);
+	else
+		ci_setup_pcie_table_entry(&pi->dpm_table.pcie_speed_table, 0,
+					  pi->pcie_gen_powersaving.min,
+					  pi->pcie_lane_powersaving.min);
 	ci_setup_pcie_table_entry(&pi->dpm_table.pcie_speed_table, 1,
 				  pi->pcie_gen_performance.min,
 				  pi->pcie_lane_performance.min);
@@ -3013,19 +3418,21 @@ static int ci_setup_default_dpm_tables(struct radeon_device *rdev)
 		     allowed_sclk_vddc_table->entries[i].clk)) {
 			pi->dpm_table.sclk_table.dpm_levels[pi->dpm_table.sclk_table.count].value =
 				allowed_sclk_vddc_table->entries[i].clk;
-			pi->dpm_table.sclk_table.dpm_levels[pi->dpm_table.sclk_table.count].enabled = true;
+			pi->dpm_table.sclk_table.dpm_levels[pi->dpm_table.sclk_table.count].enabled =
+				(i == 0) ? true : false;
 			pi->dpm_table.sclk_table.count++;
 		}
 	}
 
 	pi->dpm_table.mclk_table.count = 0;
 	for (i = 0; i < allowed_mclk_table->count; i++) {
-		if ((i==0) ||
+		if ((i == 0) ||
 		    (pi->dpm_table.mclk_table.dpm_levels[pi->dpm_table.mclk_table.count-1].value !=
 		     allowed_mclk_table->entries[i].clk)) {
 			pi->dpm_table.mclk_table.dpm_levels[pi->dpm_table.mclk_table.count].value =
 				allowed_mclk_table->entries[i].clk;
-			pi->dpm_table.mclk_table.dpm_levels[pi->dpm_table.mclk_table.count].enabled = true;
+			pi->dpm_table.mclk_table.dpm_levels[pi->dpm_table.mclk_table.count].enabled =
+				(i == 0) ? true : false;
 			pi->dpm_table.mclk_table.count++;
 		}
 	}
@@ -3191,7 +3598,7 @@ static int ci_init_smc_table(struct radeon_device *rdev)
 	table->VddcVddciDelta = 4000;
 	table->PhaseResponseTime = 0;
 	table->MemoryThermThrottleEnable = 1;
-	table->PCIeBootLinkLevel = 0;
+	table->PCIeBootLinkLevel = pi->dpm_table.pcie_speed_table.count - 1;
 	table->PCIeGenInterval = 1;
 	if (pi->voltage_control == CISLANDS_VOLTAGE_CONTROL_BY_SVID2)
 		table->SVI2Enable  = 1;
@@ -3345,6 +3752,8 @@ static int ci_upload_dpm_level_enable_mask(struct radeon_device *rdev)
 	struct ci_power_info *pi = ci_get_pi(rdev);
 	PPSMC_Result result;
 
+	ci_apply_disp_minimum_voltage_request(rdev);
+
 	if (!pi->sclk_dpm_key_disabled) {
 		if (pi->dpm_level_enable_mask.sclk_dpm_enable_mask) {
 			result = ci_send_msg_to_smc_with_parameter(rdev,
@@ -3364,7 +3773,7 @@ static int ci_upload_dpm_level_enable_mask(struct radeon_device *rdev)
 				return -EINVAL;
 		}
 	}
-
+#if 0
 	if (!pi->pcie_dpm_key_disabled) {
 		if (pi->dpm_level_enable_mask.pcie_dpm_enable_mask) {
 			result = ci_send_msg_to_smc_with_parameter(rdev,
@@ -3374,9 +3783,7 @@ static int ci_upload_dpm_level_enable_mask(struct radeon_device *rdev)
 				return -EINVAL;
 		}
 	}
-
-	ci_apply_disp_minimum_voltage_request(rdev);
-
+#endif
 	return 0;
 }
 
@@ -3402,7 +3809,7 @@ static void ci_find_dpm_states_clocks_in_dpm_table(struct radeon_device *rdev,
 		pi->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_SCLK;
 	} else {
 		/* XXX check display min clock requirements */
-		if (0 != CISLAND_MINIMUM_ENGINE_CLOCK)
+		if (CISLAND_MINIMUM_ENGINE_CLOCK != CISLAND_MINIMUM_ENGINE_CLOCK)
 			pi->need_update_smu7_dpm_table |= DPMTABLE_UPDATE_SCLK;
 	}
 
@@ -3732,62 +4139,61 @@ int ci_dpm_force_performance_level(struct radeon_device *rdev,
 				   enum radeon_dpm_forced_level level)
 {
 	struct ci_power_info *pi = ci_get_pi(rdev);
-	PPSMC_Result smc_result;
 	u32 tmp, levels, i;
 	int ret;
 
 	if (level == RADEON_DPM_FORCED_LEVEL_HIGH) {
-		if ((!pi->sclk_dpm_key_disabled) &&
-		    pi->dpm_level_enable_mask.sclk_dpm_enable_mask) {
+		if ((!pi->pcie_dpm_key_disabled) &&
+		    pi->dpm_level_enable_mask.pcie_dpm_enable_mask) {
 			levels = 0;
-			tmp = pi->dpm_level_enable_mask.sclk_dpm_enable_mask;
+			tmp = pi->dpm_level_enable_mask.pcie_dpm_enable_mask;
 			while (tmp >>= 1)
 				levels++;
 			if (levels) {
-				ret = ci_dpm_force_state_sclk(rdev, levels);
+				ret = ci_dpm_force_state_pcie(rdev, level);
 				if (ret)
 					return ret;
 				for (i = 0; i < rdev->usec_timeout; i++) {
-					tmp = (RREG32_SMC(TARGET_AND_CURRENT_PROFILE_INDEX) &
-					       CURR_SCLK_INDEX_MASK) >> CURR_SCLK_INDEX_SHIFT;
+					tmp = (RREG32_SMC(TARGET_AND_CURRENT_PROFILE_INDEX_1) &
+					       CURR_PCIE_INDEX_MASK) >> CURR_PCIE_INDEX_SHIFT;
 					if (tmp == levels)
 						break;
 					udelay(1);
 				}
 			}
 		}
-		if ((!pi->mclk_dpm_key_disabled) &&
-		    pi->dpm_level_enable_mask.mclk_dpm_enable_mask) {
+		if ((!pi->sclk_dpm_key_disabled) &&
+		    pi->dpm_level_enable_mask.sclk_dpm_enable_mask) {
 			levels = 0;
-			tmp = pi->dpm_level_enable_mask.mclk_dpm_enable_mask;
+			tmp = pi->dpm_level_enable_mask.sclk_dpm_enable_mask;
 			while (tmp >>= 1)
 				levels++;
 			if (levels) {
-				ret = ci_dpm_force_state_mclk(rdev, levels);
+				ret = ci_dpm_force_state_sclk(rdev, levels);
 				if (ret)
 					return ret;
 				for (i = 0; i < rdev->usec_timeout; i++) {
 					tmp = (RREG32_SMC(TARGET_AND_CURRENT_PROFILE_INDEX) &
-					       CURR_MCLK_INDEX_MASK) >> CURR_MCLK_INDEX_SHIFT;
+					       CURR_SCLK_INDEX_MASK) >> CURR_SCLK_INDEX_SHIFT;
 					if (tmp == levels)
 						break;
 					udelay(1);
 				}
 			}
 		}
-		if ((!pi->pcie_dpm_key_disabled) &&
-		    pi->dpm_level_enable_mask.pcie_dpm_enable_mask) {
+		if ((!pi->mclk_dpm_key_disabled) &&
+		    pi->dpm_level_enable_mask.mclk_dpm_enable_mask) {
 			levels = 0;
-			tmp = pi->dpm_level_enable_mask.pcie_dpm_enable_mask;
+			tmp = pi->dpm_level_enable_mask.mclk_dpm_enable_mask;
 			while (tmp >>= 1)
 				levels++;
 			if (levels) {
-				ret = ci_dpm_force_state_pcie(rdev, level);
+				ret = ci_dpm_force_state_mclk(rdev, levels);
 				if (ret)
 					return ret;
 				for (i = 0; i < rdev->usec_timeout; i++) {
-					tmp = (RREG32_SMC(TARGET_AND_CURRENT_PROFILE_INDEX_1) &
-					       CURR_PCIE_INDEX_MASK) >> CURR_PCIE_INDEX_SHIFT;
+					tmp = (RREG32_SMC(TARGET_AND_CURRENT_PROFILE_INDEX) &
+					       CURR_MCLK_INDEX_MASK) >> CURR_MCLK_INDEX_SHIFT;
 					if (tmp == levels)
 						break;
 					udelay(1);
@@ -3841,21 +4247,17 @@ int ci_dpm_force_performance_level(struct radeon_device *rdev,
 			}
 		}
 	} else if (level == RADEON_DPM_FORCED_LEVEL_AUTO) {
-		if (!pi->sclk_dpm_key_disabled) {
-			smc_result = ci_send_msg_to_smc(rdev, PPSMC_MSG_NoForcedLevel);
-			if (smc_result != PPSMC_Result_OK)
-				return -EINVAL;
-		}
-		if (!pi->mclk_dpm_key_disabled) {
-			smc_result = ci_send_msg_to_smc(rdev, PPSMC_MSG_MCLKDPM_NoForcedLevel);
-			if (smc_result != PPSMC_Result_OK)
-				return -EINVAL;
-		}
 		if (!pi->pcie_dpm_key_disabled) {
-			smc_result = ci_send_msg_to_smc(rdev, PPSMC_MSG_PCIeDPM_UnForceLevel);
+			PPSMC_Result smc_result;
+
+			smc_result = ci_send_msg_to_smc(rdev,
+							PPSMC_MSG_PCIeDPM_UnForceLevel);
 			if (smc_result != PPSMC_Result_OK)
 				return -EINVAL;
 		}
+		ret = ci_upload_dpm_level_enable_mask(rdev);
+		if (ret)
+			return ret;
 	}
 
 	rdev->pm.dpm.forced_level = level;
@@ -4061,6 +4463,96 @@ static int ci_copy_vbios_mc_reg_table(const struct atom_mc_reg_table *table,
 	return 0;
 }
 
+static int ci_register_patching_mc_seq(struct radeon_device *rdev,
+				       struct ci_mc_reg_table *table)
+{
+	u8 i, k;
+	u32 tmp;
+	bool patch;
+
+	tmp = RREG32(MC_SEQ_MISC0);
+	patch = ((tmp & 0x0000f00) == 0x300) ? true : false;
+
+	if (patch &&
+	    ((rdev->pdev->device == 0x67B0) ||
+	     (rdev->pdev->device == 0x67B1))) {
+		for (i = 0; i < table->last; i++) {
+			if (table->last >= SMU7_DISCRETE_MC_REGISTER_ARRAY_SIZE)
+				return -EINVAL;
+			switch(table->mc_reg_address[i].s1 >> 2) {
+			case MC_SEQ_MISC1:
+				for (k = 0; k < table->num_entries; k++) {
+					if ((table->mc_reg_table_entry[k].mclk_max == 125000) ||
+					    (table->mc_reg_table_entry[k].mclk_max == 137500))
+						table->mc_reg_table_entry[k].mc_data[i] =
+							(table->mc_reg_table_entry[k].mc_data[i] & 0xFFFFFFF8) |
+							0x00000007;
+				}
+				break;
+			case MC_SEQ_WR_CTL_D0:
+				for (k = 0; k < table->num_entries; k++) {
+					if ((table->mc_reg_table_entry[k].mclk_max == 125000) ||
+					    (table->mc_reg_table_entry[k].mclk_max == 137500))
+						table->mc_reg_table_entry[k].mc_data[i] =
+							(table->mc_reg_table_entry[k].mc_data[i] & 0xFFFF0F00) |
+							0x0000D0DD;
+				}
+				break;
+			case MC_SEQ_WR_CTL_D1:
+				for (k = 0; k < table->num_entries; k++) {
+					if ((table->mc_reg_table_entry[k].mclk_max == 125000) ||
+					    (table->mc_reg_table_entry[k].mclk_max == 137500))
+						table->mc_reg_table_entry[k].mc_data[i] =
+							(table->mc_reg_table_entry[k].mc_data[i] & 0xFFFF0F00) |
+							0x0000D0DD;
+				}
+				break;
+			case MC_SEQ_WR_CTL_2:
+				for (k = 0; k < table->num_entries; k++) {
+					if ((table->mc_reg_table_entry[k].mclk_max == 125000) ||
+					    (table->mc_reg_table_entry[k].mclk_max == 137500))
+						table->mc_reg_table_entry[k].mc_data[i] = 0;
+				}
+				break;
+			case MC_SEQ_CAS_TIMING:
+				for (k = 0; k < table->num_entries; k++) {
+					if (table->mc_reg_table_entry[k].mclk_max == 125000)
+						table->mc_reg_table_entry[k].mc_data[i] =
+							(table->mc_reg_table_entry[k].mc_data[i] & 0xFFE0FE0F) |
+							0x000C0140;
+					else if (table->mc_reg_table_entry[k].mclk_max == 137500)
+						table->mc_reg_table_entry[k].mc_data[i] =
+							(table->mc_reg_table_entry[k].mc_data[i] & 0xFFE0FE0F) |
+							0x000C0150;
+				}
+				break;
+			case MC_SEQ_MISC_TIMING:
+				for (k = 0; k < table->num_entries; k++) {
+					if (table->mc_reg_table_entry[k].mclk_max == 125000)
+						table->mc_reg_table_entry[k].mc_data[i] =
+							(table->mc_reg_table_entry[k].mc_data[i] & 0xFFFFFFE0) |
+							0x00000030;
+					else if (table->mc_reg_table_entry[k].mclk_max == 137500)
+						table->mc_reg_table_entry[k].mc_data[i] =
+							(table->mc_reg_table_entry[k].mc_data[i] & 0xFFFFFFE0) |
+							0x00000035;
+				}
+				break;
+			default:
+				break;
+			}
+		}
+
+		WREG32(MC_SEQ_IO_DEBUG_INDEX, 3);
+		tmp = RREG32(MC_SEQ_IO_DEBUG_DATA);
+		tmp = (tmp & 0xFFF8FFFF) | (1 << 16);
+		WREG32(MC_SEQ_IO_DEBUG_INDEX, 3);
+		WREG32(MC_SEQ_IO_DEBUG_DATA, tmp);
+	}
+
+	return 0;
+}
+
 static int ci_initialize_mc_reg_table(struct radeon_device *rdev)
 {
 	struct ci_power_info *pi = ci_get_pi(rdev);
@@ -4104,6 +4596,10 @@ static int ci_initialize_mc_reg_table(struct radeon_device *rdev)
 
 	ci_set_s0_mc_reg_index(ci_table);
 
+	ret = ci_register_patching_mc_seq(rdev, ci_table);
+	if (ret)
+		goto init_mc_done;
+
 	ret = ci_set_mc_special_registers(rdev, ci_table);
 	if (ret)
 		goto init_mc_done;
@@ -4700,36 +5196,51 @@ int ci_dpm_enable(struct radeon_device *rdev)
 		return ret;
 	}
 
+	ret = ci_power_control_set_level(rdev);
+	if (ret) {
+		DRM_ERROR("ci_power_control_set_level failed\n");
+		return ret;
+	}
+
 	ci_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
 
+	ret = ci_enable_thermal_based_sclk_dpm(rdev, true);
+	if (ret) {
+		DRM_ERROR("ci_enable_thermal_based_sclk_dpm failed\n");
+		return ret;
+	}
+
+	ci_thermal_start_thermal_controller(rdev);
+
 	ci_update_current_ps(rdev, boot_ps);
 
 	return 0;
 }
 
-int ci_dpm_late_enable(struct radeon_device *rdev)
+static int ci_set_temperature_range(struct radeon_device *rdev)
 {
 	int ret;
 
-	if (rdev->irq.installed &&
-	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
-#if 0
-		PPSMC_Result result;
-#endif
-		ret = ci_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
-		if (ret) {
-			DRM_ERROR("ci_set_thermal_temperature_range failed\n");
-			return ret;
-		}
-		rdev->irq.dpm_thermal = true;
-		radeon_irq_set(rdev);
-#if 0
-		result = ci_send_msg_to_smc(rdev, PPSMC_MSG_EnableThermalInterrupt);
+	ret = ci_thermal_enable_alert(rdev, false);
+	if (ret)
+		return ret;
+	ret = ci_thermal_set_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
+	if (ret)
+		return ret;
+	ret = ci_thermal_enable_alert(rdev, true);
+	if (ret)
+		return ret;
 
-		if (result != PPSMC_Result_OK)
-			DRM_DEBUG_KMS("Could not enable thermal interrupts.\n");
-#endif
-	}
+	return ret;
+}
+
+int ci_dpm_late_enable(struct radeon_device *rdev)
+{
+	int ret;
+
+	ret = ci_set_temperature_range(rdev);
+	if (ret)
+		return ret;
 
 	ci_dpm_powergate_uvd(rdev, true);
 
@@ -4746,6 +5257,8 @@ void ci_dpm_disable(struct radeon_device *rdev)
 	if (!ci_is_smc_running(rdev))
 		return;
 
+	ci_thermal_stop_thermal_controller(rdev);
+
 	if (pi->thermal_protection)
 		ci_enable_thermal_protection(rdev, false);
 	ci_enable_power_containment(rdev, false);
@@ -4754,12 +5267,13 @@ void ci_dpm_disable(struct radeon_device *rdev)
 	ci_enable_spread_spectrum(rdev, false);
 	ci_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, false);
 	ci_stop_dpm(rdev);
-	ci_enable_ds_master_switch(rdev, true);
+	ci_enable_ds_master_switch(rdev, false);
 	ci_enable_ulv(rdev, false);
 	ci_clear_vc(rdev);
 	ci_reset_to_default(rdev);
 	ci_dpm_stop_smc(rdev);
 	ci_force_switch_to_arb_f0(rdev);
+	ci_enable_thermal_based_sclk_dpm(rdev, false);
 
 	ci_update_current_ps(rdev, boot_ps);
 }
@@ -4829,11 +5343,6 @@ int ci_dpm_set_power_state(struct radeon_device *rdev)
 	return 0;
 }
 
-int ci_dpm_power_control_set_level(struct radeon_device *rdev)
-{
-	return ci_power_control_set_level(rdev);
-}
-
 void ci_dpm_reset_asic(struct radeon_device *rdev)
 {
 	ci_set_boot_state(rdev);
@@ -5093,6 +5602,8 @@ void ci_dpm_fini(struct radeon_device *rdev)
 int ci_dpm_init(struct radeon_device *rdev)
 {
 	int index = GetIndexIntoMasterTable(DATA, ASIC_InternalSS_Info);
+	SMU7_Discrete_DpmTable  *dpm_table;
+	struct radeon_gpio_rec gpio;
 	u16 data_offset, size;
 	u8 frev, crev;
 	struct ci_power_info *pi;
@@ -5162,6 +5673,7 @@ int ci_dpm_init(struct radeon_device *rdev)
 	pi->sclk_dpm_key_disabled = 0;
 	pi->mclk_dpm_key_disabled = 0;
 	pi->pcie_dpm_key_disabled = 0;
+	pi->thermal_sclk_dpm_enabled = 0;
 
 	/* mclk dpm is unstable on some R7 260X cards with the old mc ucode */
 	if ((rdev->pdev->device == 0x6658) &&
@@ -5226,6 +5738,55 @@ int ci_dpm_init(struct radeon_device *rdev)
 
 	pi->uvd_enabled = false;
 
+	dpm_table = &pi->smc_state_table;
+
+	gpio = radeon_atombios_lookup_gpio(rdev, VDDC_VRHOT_GPIO_PINID);
+	if (gpio.valid) {
+		dpm_table->VRHotGpio = gpio.shift;
+		rdev->pm.dpm.platform_caps |= ATOM_PP_PLATFORM_CAP_REGULATOR_HOT;
+	} else {
+		dpm_table->VRHotGpio = CISLANDS_UNUSED_GPIO_PIN;
+		rdev->pm.dpm.platform_caps &= ~ATOM_PP_PLATFORM_CAP_REGULATOR_HOT;
+	}
+
+	gpio = radeon_atombios_lookup_gpio(rdev, PP_AC_DC_SWITCH_GPIO_PINID);
+	if (gpio.valid) {
+		dpm_table->AcDcGpio = gpio.shift;
+		rdev->pm.dpm.platform_caps |= ATOM_PP_PLATFORM_CAP_HARDWAREDC;
+	} else {
+		dpm_table->AcDcGpio = CISLANDS_UNUSED_GPIO_PIN;
+		rdev->pm.dpm.platform_caps &= ~ATOM_PP_PLATFORM_CAP_HARDWAREDC;
+	}
+
+	gpio = radeon_atombios_lookup_gpio(rdev, VDDC_PCC_GPIO_PINID);
+	if (gpio.valid) {
+		u32 tmp = RREG32_SMC(CNB_PWRMGT_CNTL);
+
+		switch (gpio.shift) {
+		case 0:
+			tmp &= ~GNB_SLOW_MODE_MASK;
+			tmp |= GNB_SLOW_MODE(1);
+			break;
+		case 1:
+			tmp &= ~GNB_SLOW_MODE_MASK;
+			tmp |= GNB_SLOW_MODE(2);
+			break;
+		case 2:
+			tmp |= GNB_SLOW;
+			break;
+		case 3:
+			tmp |= FORCE_NB_PS1;
+			break;
+		case 4:
+			tmp |= DPM_ENABLED;
+			break;
+		default:
+			DRM_ERROR("Invalid PCC GPIO: %u!\n", gpio.shift);
+			break;
+		}
+		WREG32_SMC(CNB_PWRMGT_CNTL, tmp);
+	}
+
 	pi->voltage_control = CISLANDS_VOLTAGE_CONTROL_NONE;
 	pi->vddci_control = CISLANDS_VOLTAGE_CONTROL_NONE;
 	pi->mvdd_control = CISLANDS_VOLTAGE_CONTROL_NONE;
@@ -5287,15 +5848,21 @@ int ci_dpm_init(struct radeon_device *rdev)
 		rdev->pm.dpm.dyn_state.max_clock_voltage_on_dc =
 			rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac;
 
+	pi->fan_ctrl_is_in_default_mode = true;
+
 	return 0;
 }
 
 void ci_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev,
 						    struct seq_file *m)
 {
+	struct ci_power_info *pi = ci_get_pi(rdev);
+	struct radeon_ps *rps = &pi->current_rps;
 	u32 sclk = ci_get_average_sclk_freq(rdev);
 	u32 mclk = ci_get_average_mclk_freq(rdev);
 
+	seq_printf(m, "uvd    %sabled\n", pi->uvd_enabled ? "en" : "dis");
+	seq_printf(m, "vce    %sabled\n", rps->vce_active ? "en" : "dis");
 	seq_printf(m, "power level avg    sclk: %u mclk: %u\n",
 		   sclk, mclk);
 }
diff --git a/drivers/gpu/drm/radeon/ci_dpm.h b/drivers/gpu/drm/radeon/ci_dpm.h
index 93bbed977ffb..84e3d3bcf9f3 100644
--- a/drivers/gpu/drm/radeon/ci_dpm.h
+++ b/drivers/gpu/drm/radeon/ci_dpm.h
@@ -33,6 +33,8 @@
 
 #define CISLANDS_MAX_HARDWARE_POWERLEVELS 2
 
+#define CISLANDS_UNUSED_GPIO_PIN 0x7F
+
 struct ci_pl {
 	u32 mclk;
 	u32 sclk;
@@ -237,6 +239,7 @@ struct ci_power_info {
 	u32 sclk_dpm_key_disabled;
 	u32 mclk_dpm_key_disabled;
 	u32 pcie_dpm_key_disabled;
+	u32 thermal_sclk_dpm_enabled;
 	struct ci_pcie_perf_range pcie_gen_performance;
 	struct ci_pcie_perf_range pcie_lane_performance;
 	struct ci_pcie_perf_range pcie_gen_powersaving;
@@ -264,6 +267,7 @@ struct ci_power_info {
 	bool caps_automatic_dc_transition;
 	bool caps_sclk_throttle_low_notification;
 	bool caps_dynamic_ac_timing;
+	bool caps_od_fuzzy_fan_control_support;
 	/* flags */
 	bool thermal_protection;
 	bool pcie_performance_request;
@@ -285,6 +289,10 @@ struct ci_power_info {
 	struct ci_ps current_ps;
 	struct radeon_ps requested_rps;
 	struct ci_ps requested_ps;
+	/* fan control */
+	bool fan_ctrl_is_in_default_mode;
+	u32 t_min;
+	u32 fan_ctrl_default_mode;
 };
 
 #define CISLANDS_VOLTAGE_CONTROL_NONE                   0x0
diff --git a/drivers/gpu/drm/radeon/ci_smc.c b/drivers/gpu/drm/radeon/ci_smc.c
index b630edc2fd0c..e78bcad7a43e 100644
--- a/drivers/gpu/drm/radeon/ci_smc.c
+++ b/drivers/gpu/drm/radeon/ci_smc.c
@@ -129,7 +129,7 @@ void ci_reset_smc(struct radeon_device *rdev)
 
 int ci_program_jump_on_start(struct radeon_device *rdev)
 {
-	static u8 data[] = { 0xE0, 0x00, 0x80, 0x40 };
+	static const u8 data[] = { 0xE0, 0x00, 0x80, 0x40 };
 
 	return ci_copy_bytes_to_smc(rdev, 0x0, data, 4, sizeof(data)+1);
 }
diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/radeon/cik.c
index 3d546c606b43..64fdae558d36 100644
--- a/drivers/gpu/drm/radeon/cik.c
+++ b/drivers/gpu/drm/radeon/cik.c
@@ -32,6 +32,7 @@
 #include "cik_blit_shaders.h"
 #include "radeon_ucode.h"
 #include "clearstate_ci.h"
+#include "radeon_kfd.h"
 
 MODULE_FIRMWARE("radeon/BONAIRE_pfp.bin");
 MODULE_FIRMWARE("radeon/BONAIRE_me.bin");
@@ -1563,6 +1564,8 @@ static const u32 godavari_golden_registers[] =
 
 static void cik_init_golden_registers(struct radeon_device *rdev)
 {
+	/* Some of the registers might be dependent on GRBM_GFX_INDEX */
+	mutex_lock(&rdev->grbm_idx_mutex);
 	switch (rdev->family) {
 	case CHIP_BONAIRE:
 		radeon_program_register_sequence(rdev,
@@ -1637,6 +1640,7 @@ static void cik_init_golden_registers(struct radeon_device *rdev)
 	default:
 		break;
 	}
+	mutex_unlock(&rdev->grbm_idx_mutex);
 }
 
 /**
@@ -1806,7 +1810,7 @@ int ci_mc_load_microcode(struct radeon_device *rdev)
 {
 	const __be32 *fw_data = NULL;
 	const __le32 *new_fw_data = NULL;
-	u32 running, blackout = 0;
+	u32 running, blackout = 0, tmp;
 	u32 *io_mc_regs = NULL;
 	const __le32 *new_io_mc_regs = NULL;
 	int i, regs_size, ucode_size;
@@ -1866,6 +1870,15 @@ int ci_mc_load_microcode(struct radeon_device *rdev)
 				WREG32(MC_SEQ_IO_DEBUG_DATA, io_mc_regs[(i << 1) + 1]);
 			}
 		}
+
+		tmp = RREG32(MC_SEQ_MISC0);
+		if ((rdev->pdev->device == 0x6649) && ((tmp & 0xff00) == 0x5600)) {
+			WREG32(MC_SEQ_IO_DEBUG_INDEX, 5);
+			WREG32(MC_SEQ_IO_DEBUG_DATA, 0x00000023);
+			WREG32(MC_SEQ_IO_DEBUG_INDEX, 9);
+			WREG32(MC_SEQ_IO_DEBUG_DATA, 0x000001f0);
+		}
+
 		/* load the MC ucode */
 		for (i = 0; i < ucode_size; i++) {
 			if (rdev->new_fw)
@@ -3419,6 +3432,7 @@ static void cik_setup_rb(struct radeon_device *rdev,
 	u32 disabled_rbs = 0;
 	u32 enabled_rbs = 0;
 
+	mutex_lock(&rdev->grbm_idx_mutex);
 	for (i = 0; i < se_num; i++) {
 		for (j = 0; j < sh_per_se; j++) {
 			cik_select_se_sh(rdev, i, j);
@@ -3430,6 +3444,7 @@ static void cik_setup_rb(struct radeon_device *rdev,
 		}
 	}
 	cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
+	mutex_unlock(&rdev->grbm_idx_mutex);
 
 	mask = 1;
 	for (i = 0; i < max_rb_num_per_se * se_num; i++) {
@@ -3440,6 +3455,7 @@ static void cik_setup_rb(struct radeon_device *rdev,
 
 	rdev->config.cik.backend_enable_mask = enabled_rbs;
 
+	mutex_lock(&rdev->grbm_idx_mutex);
 	for (i = 0; i < se_num; i++) {
 		cik_select_se_sh(rdev, i, 0xffffffff);
 		data = 0;
@@ -3467,6 +3483,7 @@ static void cik_setup_rb(struct radeon_device *rdev,
 		WREG32(PA_SC_RASTER_CONFIG, data);
 	}
 	cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
+	mutex_unlock(&rdev->grbm_idx_mutex);
 }
 
 /**
@@ -3684,6 +3701,12 @@ static void cik_gpu_init(struct radeon_device *rdev)
 	/* set HW defaults for 3D engine */
 	WREG32(CP_MEQ_THRESHOLDS, MEQ1_START(0x30) | MEQ2_START(0x60));
 
+	mutex_lock(&rdev->grbm_idx_mutex);
+	/*
+	 * making sure that the following register writes will be broadcasted
+	 * to all the shaders
+	 */
+	cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
 	WREG32(SX_DEBUG_1, 0x20);
 
 	WREG32(TA_CNTL_AUX, 0x00010000);
@@ -3739,6 +3762,7 @@ static void cik_gpu_init(struct radeon_device *rdev)
 
 	WREG32(PA_CL_ENHANCE, CLIP_VTX_REORDER_ENA | NUM_CLIP_SEQ(3));
 	WREG32(PA_SC_ENHANCE, ENABLE_PA_SC_OUT_OF_ORDER);
+	mutex_unlock(&rdev->grbm_idx_mutex);
 
 	udelay(50);
 }
@@ -3959,41 +3983,38 @@ bool cik_semaphore_ring_emit(struct radeon_device *rdev,
  * @src_offset: src GPU address
  * @dst_offset: dst GPU address
  * @num_gpu_pages: number of GPU pages to xfer
- * @fence: radeon fence object
+ * @resv: reservation object to sync to
  *
  * Copy GPU paging using the CP DMA engine (CIK+).
  * Used by the radeon ttm implementation to move pages if
  * registered as the asic copy callback.
  */
-int cik_copy_cpdma(struct radeon_device *rdev,
-		   uint64_t src_offset, uint64_t dst_offset,
-		   unsigned num_gpu_pages,
-		   struct radeon_fence **fence)
+struct radeon_fence *cik_copy_cpdma(struct radeon_device *rdev,
+				    uint64_t src_offset, uint64_t dst_offset,
+				    unsigned num_gpu_pages,
+				    struct reservation_object *resv)
 {
-	struct radeon_semaphore *sem = NULL;
+	struct radeon_fence *fence;
+	struct radeon_sync sync;
 	int ring_index = rdev->asic->copy.blit_ring_index;
 	struct radeon_ring *ring = &rdev->ring[ring_index];
 	u32 size_in_bytes, cur_size_in_bytes, control;
 	int i, num_loops;
 	int r = 0;
 
-	r = radeon_semaphore_create(rdev, &sem);
-	if (r) {
-		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		return r;
-	}
+	radeon_sync_create(&sync);
 
 	size_in_bytes = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT);
 	num_loops = DIV_ROUND_UP(size_in_bytes, 0x1fffff);
 	r = radeon_ring_lock(rdev, ring, num_loops * 7 + 18);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		radeon_sync_free(rdev, &sync, NULL);
+		return ERR_PTR(r);
 	}
 
-	radeon_semaphore_sync_to(sem, *fence);
-	radeon_semaphore_sync_rings(rdev, sem, ring->idx);
+	radeon_sync_resv(rdev, &sync, resv, false);
+	radeon_sync_rings(rdev, &sync, ring->idx);
 
 	for (i = 0; i < num_loops; i++) {
 		cur_size_in_bytes = size_in_bytes;
@@ -4014,17 +4035,17 @@ int cik_copy_cpdma(struct radeon_device *rdev,
 		dst_offset += cur_size_in_bytes;
 	}
 
-	r = radeon_fence_emit(rdev, fence, ring->idx);
+	r = radeon_fence_emit(rdev, &fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
-		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		radeon_sync_free(rdev, &sync, NULL);
+		return ERR_PTR(r);
 	}
 
 	radeon_ring_unlock_commit(rdev, ring, false);
-	radeon_semaphore_free(rdev, &sem, *fence);
+	radeon_sync_free(rdev, &sync, fence);
 
-	return r;
+	return fence;
 }
 
 /*
@@ -4045,6 +4066,7 @@ int cik_copy_cpdma(struct radeon_device *rdev,
 void cik_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
 {
 	struct radeon_ring *ring = &rdev->ring[ib->ring];
+	unsigned vm_id = ib->vm ? ib->vm->ids[ib->ring].id : 0;
 	u32 header, control = INDIRECT_BUFFER_VALID;
 
 	if (ib->is_const_ib) {
@@ -4073,8 +4095,7 @@ void cik_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
 		header = PACKET3(PACKET3_INDIRECT_BUFFER, 2);
 	}
 
-	control |= ib->length_dw |
-		(ib->vm ? (ib->vm->id << 24) : 0);
+	control |= ib->length_dw | (vm_id << 24);
 
 	radeon_ring_write(ring, header);
 	radeon_ring_write(ring,
@@ -4234,7 +4255,7 @@ static int cik_cp_gfx_load_microcode(struct radeon_device *rdev)
 		WREG32(CP_PFP_UCODE_ADDR, 0);
 		for (i = 0; i < fw_size; i++)
 			WREG32(CP_PFP_UCODE_DATA, le32_to_cpup(fw_data++));
-		WREG32(CP_PFP_UCODE_ADDR, 0);
+		WREG32(CP_PFP_UCODE_ADDR, le32_to_cpu(pfp_hdr->header.ucode_version));
 
 		/* CE */
 		fw_data = (const __le32 *)
@@ -4243,7 +4264,7 @@ static int cik_cp_gfx_load_microcode(struct radeon_device *rdev)
 		WREG32(CP_CE_UCODE_ADDR, 0);
 		for (i = 0; i < fw_size; i++)
 			WREG32(CP_CE_UCODE_DATA, le32_to_cpup(fw_data++));
-		WREG32(CP_CE_UCODE_ADDR, 0);
+		WREG32(CP_CE_UCODE_ADDR, le32_to_cpu(ce_hdr->header.ucode_version));
 
 		/* ME */
 		fw_data = (const __be32 *)
@@ -4252,7 +4273,8 @@ static int cik_cp_gfx_load_microcode(struct radeon_device *rdev)
 		WREG32(CP_ME_RAM_WADDR, 0);
 		for (i = 0; i < fw_size; i++)
 			WREG32(CP_ME_RAM_DATA, le32_to_cpup(fw_data++));
-		WREG32(CP_ME_RAM_WADDR, 0);
+		WREG32(CP_ME_RAM_WADDR, le32_to_cpu(me_hdr->header.ucode_version));
+		WREG32(CP_ME_RAM_RADDR, le32_to_cpu(me_hdr->header.ucode_version));
 	} else {
 		const __be32 *fw_data;
 
@@ -4278,10 +4300,6 @@ static int cik_cp_gfx_load_microcode(struct radeon_device *rdev)
 		WREG32(CP_ME_RAM_WADDR, 0);
 	}
 
-	WREG32(CP_PFP_UCODE_ADDR, 0);
-	WREG32(CP_CE_UCODE_ADDR, 0);
-	WREG32(CP_ME_RAM_WADDR, 0);
-	WREG32(CP_ME_RAM_RADDR, 0);
 	return 0;
 }
 
@@ -4315,8 +4333,8 @@ static int cik_cp_gfx_start(struct radeon_device *rdev)
 	/* init the CE partitions.  CE only used for gfx on CIK */
 	radeon_ring_write(ring, PACKET3(PACKET3_SET_BASE, 2));
 	radeon_ring_write(ring, PACKET3_BASE_INDEX(CE_PARTITION_BASE));
-	radeon_ring_write(ring, 0xc000);
-	radeon_ring_write(ring, 0xc000);
+	radeon_ring_write(ring, 0x8000);
+	radeon_ring_write(ring, 0x8000);
 
 	/* setup clear context state */
 	radeon_ring_write(ring, PACKET3(PACKET3_PREAMBLE_CNTL, 0));
@@ -4563,7 +4581,7 @@ static int cik_cp_compute_load_microcode(struct radeon_device *rdev)
 		WREG32(CP_MEC_ME1_UCODE_ADDR, 0);
 		for (i = 0; i < fw_size; i++)
 			WREG32(CP_MEC_ME1_UCODE_DATA, le32_to_cpup(fw_data++));
-		WREG32(CP_MEC_ME1_UCODE_ADDR, 0);
+		WREG32(CP_MEC_ME1_UCODE_ADDR, le32_to_cpu(mec_hdr->header.ucode_version));
 
 		/* MEC2 */
 		if (rdev->family == CHIP_KAVERI) {
@@ -4577,7 +4595,7 @@ static int cik_cp_compute_load_microcode(struct radeon_device *rdev)
 			WREG32(CP_MEC_ME2_UCODE_ADDR, 0);
 			for (i = 0; i < fw_size; i++)
 				WREG32(CP_MEC_ME2_UCODE_DATA, le32_to_cpup(fw_data++));
-			WREG32(CP_MEC_ME2_UCODE_ADDR, 0);
+			WREG32(CP_MEC_ME2_UCODE_ADDR, le32_to_cpu(mec2_hdr->header.ucode_version));
 		}
 	} else {
 		const __be32 *fw_data;
@@ -4677,19 +4695,18 @@ static int cik_mec_init(struct radeon_device *rdev)
 	/*
 	 * KV:    2 MEC, 4 Pipes/MEC, 8 Queues/Pipe - 64 Queues total
 	 * CI/KB: 1 MEC, 4 Pipes/MEC, 8 Queues/Pipe - 32 Queues total
+	 * Nonetheless, we assign only 1 pipe because all other pipes will
+	 * be handled by KFD
 	 */
-	if (rdev->family == CHIP_KAVERI)
-		rdev->mec.num_mec = 2;
-	else
-		rdev->mec.num_mec = 1;
-	rdev->mec.num_pipe = 4;
+	rdev->mec.num_mec = 1;
+	rdev->mec.num_pipe = 1;
 	rdev->mec.num_queue = rdev->mec.num_mec * rdev->mec.num_pipe * 8;
 
 	if (rdev->mec.hpd_eop_obj == NULL) {
 		r = radeon_bo_create(rdev,
 				     rdev->mec.num_mec *rdev->mec.num_pipe * MEC_HPD_SIZE * 2,
 				     PAGE_SIZE, true,
-				     RADEON_GEM_DOMAIN_GTT, 0, NULL,
+				     RADEON_GEM_DOMAIN_GTT, 0, NULL, NULL,
 				     &rdev->mec.hpd_eop_obj);
 		if (r) {
 			dev_warn(rdev->dev, "(%d) create HDP EOP bo failed\n", r);
@@ -4824,28 +4841,24 @@ static int cik_cp_compute_resume(struct radeon_device *rdev)
 
 	/* init the pipes */
 	mutex_lock(&rdev->srbm_mutex);
-	for (i = 0; i < (rdev->mec.num_pipe * rdev->mec.num_mec); i++) {
-		int me = (i < 4) ? 1 : 2;
-		int pipe = (i < 4) ? i : (i - 4);
 
-		eop_gpu_addr = rdev->mec.hpd_eop_gpu_addr + (i * MEC_HPD_SIZE * 2);
+	eop_gpu_addr = rdev->mec.hpd_eop_gpu_addr;
 
-		cik_srbm_select(rdev, me, pipe, 0, 0);
+	cik_srbm_select(rdev, 0, 0, 0, 0);
 
-		/* write the EOP addr */
-		WREG32(CP_HPD_EOP_BASE_ADDR, eop_gpu_addr >> 8);
-		WREG32(CP_HPD_EOP_BASE_ADDR_HI, upper_32_bits(eop_gpu_addr) >> 8);
+	/* write the EOP addr */
+	WREG32(CP_HPD_EOP_BASE_ADDR, eop_gpu_addr >> 8);
+	WREG32(CP_HPD_EOP_BASE_ADDR_HI, upper_32_bits(eop_gpu_addr) >> 8);
 
-		/* set the VMID assigned */
-		WREG32(CP_HPD_EOP_VMID, 0);
+	/* set the VMID assigned */
+	WREG32(CP_HPD_EOP_VMID, 0);
+
+	/* set the EOP size, register value is 2^(EOP_SIZE+1) dwords */
+	tmp = RREG32(CP_HPD_EOP_CONTROL);
+	tmp &= ~EOP_SIZE_MASK;
+	tmp |= order_base_2(MEC_HPD_SIZE / 8);
+	WREG32(CP_HPD_EOP_CONTROL, tmp);
 
-		/* set the EOP size, register value is 2^(EOP_SIZE+1) dwords */
-		tmp = RREG32(CP_HPD_EOP_CONTROL);
-		tmp &= ~EOP_SIZE_MASK;
-		tmp |= order_base_2(MEC_HPD_SIZE / 8);
-		WREG32(CP_HPD_EOP_CONTROL, tmp);
-	}
-	cik_srbm_select(rdev, 0, 0, 0, 0);
 	mutex_unlock(&rdev->srbm_mutex);
 
 	/* init the queues.  Just two for now. */
@@ -4860,7 +4873,7 @@ static int cik_cp_compute_resume(struct radeon_device *rdev)
 					     sizeof(struct bonaire_mqd),
 					     PAGE_SIZE, true,
 					     RADEON_GEM_DOMAIN_GTT, 0, NULL,
-					     &rdev->ring[idx].mqd_obj);
+					     NULL, &rdev->ring[idx].mqd_obj);
 			if (r) {
 				dev_warn(rdev->dev, "(%d) create MQD bo failed\n", r);
 				return r;
@@ -5899,8 +5912,13 @@ int cik_ib_parse(struct radeon_device *rdev, struct radeon_ib *ib)
  */
 int cik_vm_init(struct radeon_device *rdev)
 {
-	/* number of VMs */
-	rdev->vm_manager.nvm = 16;
+	/*
+	 * number of VMs
+	 * VMID 0 is reserved for System
+	 * radeon graphics/compute will use VMIDs 1-7
+	 * amdkfd will use VMIDs 8-15
+	 */
+	rdev->vm_manager.nvm = RADEON_NUM_OF_VMIDS;
 	/* base offset of vram pages */
 	if (rdev->flags & RADEON_IS_IGP) {
 		u64 tmp = RREG32(MC_VM_FB_OFFSET);
@@ -5960,26 +5978,23 @@ static void cik_vm_decode_fault(struct radeon_device *rdev,
  * Update the page table base and flush the VM TLB
  * using the CP (CIK).
  */
-void cik_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
+void cik_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring,
+		  unsigned vm_id, uint64_t pd_addr)
 {
-	struct radeon_ring *ring = &rdev->ring[ridx];
-	int usepfp = (ridx == RADEON_RING_TYPE_GFX_INDEX);
-
-	if (vm == NULL)
-		return;
+	int usepfp = (ring->idx == RADEON_RING_TYPE_GFX_INDEX);
 
 	radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
 	radeon_ring_write(ring, (WRITE_DATA_ENGINE_SEL(usepfp) |
 				 WRITE_DATA_DST_SEL(0)));
-	if (vm->id < 8) {
+	if (vm_id < 8) {
 		radeon_ring_write(ring,
-				  (VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (vm->id << 2)) >> 2);
+				  (VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (vm_id << 2)) >> 2);
 	} else {
 		radeon_ring_write(ring,
-				  (VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + ((vm->id - 8) << 2)) >> 2);
+				  (VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + ((vm_id - 8) << 2)) >> 2);
 	}
 	radeon_ring_write(ring, 0);
-	radeon_ring_write(ring, vm->pd_gpu_addr >> 12);
+	radeon_ring_write(ring, pd_addr >> 12);
 
 	/* update SH_MEM_* regs */
 	radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
@@ -5987,7 +6002,7 @@ void cik_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
 				 WRITE_DATA_DST_SEL(0)));
 	radeon_ring_write(ring, SRBM_GFX_CNTL >> 2);
 	radeon_ring_write(ring, 0);
-	radeon_ring_write(ring, VMID(vm->id));
+	radeon_ring_write(ring, VMID(vm_id));
 
 	radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 6));
 	radeon_ring_write(ring, (WRITE_DATA_ENGINE_SEL(usepfp) |
@@ -6008,7 +6023,7 @@ void cik_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
 	radeon_ring_write(ring, VMID(0));
 
 	/* HDP flush */
-	cik_hdp_flush_cp_ring_emit(rdev, ridx);
+	cik_hdp_flush_cp_ring_emit(rdev, ring->idx);
 
 	/* bits 0-15 are the VM contexts0-15 */
 	radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
@@ -6016,7 +6031,18 @@ void cik_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
 				 WRITE_DATA_DST_SEL(0)));
 	radeon_ring_write(ring, VM_INVALIDATE_REQUEST >> 2);
 	radeon_ring_write(ring, 0);
-	radeon_ring_write(ring, 1 << vm->id);
+	radeon_ring_write(ring, 1 << vm_id);
+
+	/* wait for the invalidate to complete */
+	radeon_ring_write(ring, PACKET3(PACKET3_WAIT_REG_MEM, 5));
+	radeon_ring_write(ring, (WAIT_REG_MEM_OPERATION(0) | /* wait */
+				 WAIT_REG_MEM_FUNCTION(0) |  /* always */
+				 WAIT_REG_MEM_ENGINE(0))); /* me */
+	radeon_ring_write(ring, VM_INVALIDATE_REQUEST >> 2);
+	radeon_ring_write(ring, 0);
+	radeon_ring_write(ring, 0); /* ref */
+	radeon_ring_write(ring, 0); /* mask */
+	radeon_ring_write(ring, 0x20); /* poll interval */
 
 	/* compute doesn't have PFP */
 	if (usepfp) {
@@ -6061,6 +6087,7 @@ static void cik_wait_for_rlc_serdes(struct radeon_device *rdev)
 	u32 i, j, k;
 	u32 mask;
 
+	mutex_lock(&rdev->grbm_idx_mutex);
 	for (i = 0; i < rdev->config.cik.max_shader_engines; i++) {
 		for (j = 0; j < rdev->config.cik.max_sh_per_se; j++) {
 			cik_select_se_sh(rdev, i, j);
@@ -6072,6 +6099,7 @@ static void cik_wait_for_rlc_serdes(struct radeon_device *rdev)
 		}
 	}
 	cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
+	mutex_unlock(&rdev->grbm_idx_mutex);
 
 	mask = SE_MASTER_BUSY_MASK | GC_MASTER_BUSY | TC0_MASTER_BUSY | TC1_MASTER_BUSY;
 	for (k = 0; k < rdev->usec_timeout; k++) {
@@ -6206,10 +6234,12 @@ static int cik_rlc_resume(struct radeon_device *rdev)
 	WREG32(RLC_LB_CNTR_INIT, 0);
 	WREG32(RLC_LB_CNTR_MAX, 0x00008000);
 
+	mutex_lock(&rdev->grbm_idx_mutex);
 	cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
 	WREG32(RLC_LB_INIT_CU_MASK, 0xffffffff);
 	WREG32(RLC_LB_PARAMS, 0x00600408);
 	WREG32(RLC_LB_CNTL, 0x80000004);
+	mutex_unlock(&rdev->grbm_idx_mutex);
 
 	WREG32(RLC_MC_CNTL, 0);
 	WREG32(RLC_UCODE_CNTL, 0);
@@ -6226,7 +6256,7 @@ static int cik_rlc_resume(struct radeon_device *rdev)
 		WREG32(RLC_GPM_UCODE_ADDR, 0);
 		for (i = 0; i < size; i++)
 			WREG32(RLC_GPM_UCODE_DATA, le32_to_cpup(fw_data++));
-		WREG32(RLC_GPM_UCODE_ADDR, 0);
+		WREG32(RLC_GPM_UCODE_ADDR, le32_to_cpu(hdr->header.ucode_version));
 	} else {
 		const __be32 *fw_data;
 
@@ -6276,11 +6306,13 @@ static void cik_enable_cgcg(struct radeon_device *rdev, bool enable)
 
 		tmp = cik_halt_rlc(rdev);
 
+		mutex_lock(&rdev->grbm_idx_mutex);
 		cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
 		WREG32(RLC_SERDES_WR_CU_MASTER_MASK, 0xffffffff);
 		WREG32(RLC_SERDES_WR_NONCU_MASTER_MASK, 0xffffffff);
 		tmp2 = BPM_ADDR_MASK | CGCG_OVERRIDE_0 | CGLS_ENABLE;
 		WREG32(RLC_SERDES_WR_CTRL, tmp2);
+		mutex_unlock(&rdev->grbm_idx_mutex);
 
 		cik_update_rlc(rdev, tmp);
 
@@ -6316,17 +6348,20 @@ static void cik_enable_mgcg(struct radeon_device *rdev, bool enable)
 		}
 
 		orig = data = RREG32(RLC_CGTT_MGCG_OVERRIDE);
+		data |= 0x00000001;
 		data &= 0xfffffffd;
 		if (orig != data)
 			WREG32(RLC_CGTT_MGCG_OVERRIDE, data);
 
 		tmp = cik_halt_rlc(rdev);
 
+		mutex_lock(&rdev->grbm_idx_mutex);
 		cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
 		WREG32(RLC_SERDES_WR_CU_MASTER_MASK, 0xffffffff);
 		WREG32(RLC_SERDES_WR_NONCU_MASTER_MASK, 0xffffffff);
 		data = BPM_ADDR_MASK | MGCG_OVERRIDE_0;
 		WREG32(RLC_SERDES_WR_CTRL, data);
+		mutex_unlock(&rdev->grbm_idx_mutex);
 
 		cik_update_rlc(rdev, tmp);
 
@@ -6347,7 +6382,7 @@ static void cik_enable_mgcg(struct radeon_device *rdev, bool enable)
 		}
 	} else {
 		orig = data = RREG32(RLC_CGTT_MGCG_OVERRIDE);
-		data |= 0x00000002;
+		data |= 0x00000003;
 		if (orig != data)
 			WREG32(RLC_CGTT_MGCG_OVERRIDE, data);
 
@@ -6370,11 +6405,13 @@ static void cik_enable_mgcg(struct radeon_device *rdev, bool enable)
 
 		tmp = cik_halt_rlc(rdev);
 
+		mutex_lock(&rdev->grbm_idx_mutex);
 		cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
 		WREG32(RLC_SERDES_WR_CU_MASTER_MASK, 0xffffffff);
 		WREG32(RLC_SERDES_WR_NONCU_MASTER_MASK, 0xffffffff);
 		data = BPM_ADDR_MASK | MGCG_OVERRIDE_1;
 		WREG32(RLC_SERDES_WR_CTRL, data);
+		mutex_unlock(&rdev->grbm_idx_mutex);
 
 		cik_update_rlc(rdev, tmp);
 	}
@@ -6803,10 +6840,12 @@ static u32 cik_get_cu_active_bitmap(struct radeon_device *rdev, u32 se, u32 sh)
 	u32 mask = 0, tmp, tmp1;
 	int i;
 
+	mutex_lock(&rdev->grbm_idx_mutex);
 	cik_select_se_sh(rdev, se, sh);
 	tmp = RREG32(CC_GC_SHADER_ARRAY_CONFIG);
 	tmp1 = RREG32(GC_USER_SHADER_ARRAY_CONFIG);
 	cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
+	mutex_unlock(&rdev->grbm_idx_mutex);
 
 	tmp &= 0xffff0000;
 
@@ -7290,8 +7329,7 @@ static int cik_irq_init(struct radeon_device *rdev)
 int cik_irq_set(struct radeon_device *rdev)
 {
 	u32 cp_int_cntl;
-	u32 cp_m1p0, cp_m1p1, cp_m1p2, cp_m1p3;
-	u32 cp_m2p0, cp_m2p1, cp_m2p2, cp_m2p3;
+	u32 cp_m1p0;
 	u32 crtc1 = 0, crtc2 = 0, crtc3 = 0, crtc4 = 0, crtc5 = 0, crtc6 = 0;
 	u32 hpd1, hpd2, hpd3, hpd4, hpd5, hpd6;
 	u32 grbm_int_cntl = 0;
@@ -7325,13 +7363,6 @@ int cik_irq_set(struct radeon_device *rdev)
 	dma_cntl1 = RREG32(SDMA0_CNTL + SDMA1_REGISTER_OFFSET) & ~TRAP_ENABLE;
 
 	cp_m1p0 = RREG32(CP_ME1_PIPE0_INT_CNTL) & ~TIME_STAMP_INT_ENABLE;
-	cp_m1p1 = RREG32(CP_ME1_PIPE1_INT_CNTL) & ~TIME_STAMP_INT_ENABLE;
-	cp_m1p2 = RREG32(CP_ME1_PIPE2_INT_CNTL) & ~TIME_STAMP_INT_ENABLE;
-	cp_m1p3 = RREG32(CP_ME1_PIPE3_INT_CNTL) & ~TIME_STAMP_INT_ENABLE;
-	cp_m2p0 = RREG32(CP_ME2_PIPE0_INT_CNTL) & ~TIME_STAMP_INT_ENABLE;
-	cp_m2p1 = RREG32(CP_ME2_PIPE1_INT_CNTL) & ~TIME_STAMP_INT_ENABLE;
-	cp_m2p2 = RREG32(CP_ME2_PIPE2_INT_CNTL) & ~TIME_STAMP_INT_ENABLE;
-	cp_m2p3 = RREG32(CP_ME2_PIPE3_INT_CNTL) & ~TIME_STAMP_INT_ENABLE;
 
 	if (rdev->flags & RADEON_IS_IGP)
 		thermal_int = RREG32_SMC(CG_THERMAL_INT_CTRL) &
@@ -7353,33 +7384,6 @@ int cik_irq_set(struct radeon_device *rdev)
 			case 0:
 				cp_m1p0 |= TIME_STAMP_INT_ENABLE;
 				break;
-			case 1:
-				cp_m1p1 |= TIME_STAMP_INT_ENABLE;
-				break;
-			case 2:
-				cp_m1p2 |= TIME_STAMP_INT_ENABLE;
-				break;
-			case 3:
-				cp_m1p2 |= TIME_STAMP_INT_ENABLE;
-				break;
-			default:
-				DRM_DEBUG("si_irq_set: sw int cp1 invalid pipe %d\n", ring->pipe);
-				break;
-			}
-		} else if (ring->me == 2) {
-			switch (ring->pipe) {
-			case 0:
-				cp_m2p0 |= TIME_STAMP_INT_ENABLE;
-				break;
-			case 1:
-				cp_m2p1 |= TIME_STAMP_INT_ENABLE;
-				break;
-			case 2:
-				cp_m2p2 |= TIME_STAMP_INT_ENABLE;
-				break;
-			case 3:
-				cp_m2p2 |= TIME_STAMP_INT_ENABLE;
-				break;
 			default:
 				DRM_DEBUG("si_irq_set: sw int cp1 invalid pipe %d\n", ring->pipe);
 				break;
@@ -7396,33 +7400,6 @@ int cik_irq_set(struct radeon_device *rdev)
 			case 0:
 				cp_m1p0 |= TIME_STAMP_INT_ENABLE;
 				break;
-			case 1:
-				cp_m1p1 |= TIME_STAMP_INT_ENABLE;
-				break;
-			case 2:
-				cp_m1p2 |= TIME_STAMP_INT_ENABLE;
-				break;
-			case 3:
-				cp_m1p2 |= TIME_STAMP_INT_ENABLE;
-				break;
-			default:
-				DRM_DEBUG("si_irq_set: sw int cp2 invalid pipe %d\n", ring->pipe);
-				break;
-			}
-		} else if (ring->me == 2) {
-			switch (ring->pipe) {
-			case 0:
-				cp_m2p0 |= TIME_STAMP_INT_ENABLE;
-				break;
-			case 1:
-				cp_m2p1 |= TIME_STAMP_INT_ENABLE;
-				break;
-			case 2:
-				cp_m2p2 |= TIME_STAMP_INT_ENABLE;
-				break;
-			case 3:
-				cp_m2p2 |= TIME_STAMP_INT_ENABLE;
-				break;
 			default:
 				DRM_DEBUG("si_irq_set: sw int cp2 invalid pipe %d\n", ring->pipe);
 				break;
@@ -7511,13 +7488,6 @@ int cik_irq_set(struct radeon_device *rdev)
 	WREG32(SDMA0_CNTL + SDMA1_REGISTER_OFFSET, dma_cntl1);
 
 	WREG32(CP_ME1_PIPE0_INT_CNTL, cp_m1p0);
-	WREG32(CP_ME1_PIPE1_INT_CNTL, cp_m1p1);
-	WREG32(CP_ME1_PIPE2_INT_CNTL, cp_m1p2);
-	WREG32(CP_ME1_PIPE3_INT_CNTL, cp_m1p3);
-	WREG32(CP_ME2_PIPE0_INT_CNTL, cp_m2p0);
-	WREG32(CP_ME2_PIPE1_INT_CNTL, cp_m2p1);
-	WREG32(CP_ME2_PIPE2_INT_CNTL, cp_m2p2);
-	WREG32(CP_ME2_PIPE3_INT_CNTL, cp_m2p3);
 
 	WREG32(GRBM_INT_CNTL, grbm_int_cntl);
 
@@ -7834,6 +7804,10 @@ restart_ih:
 	while (rptr != wptr) {
 		/* wptr/rptr are in bytes! */
 		ring_index = rptr / 4;
+
+		radeon_kfd_interrupt(rdev,
+				(const void *) &rdev->ih.ring[ring_index]);
+
 		src_id =  le32_to_cpu(rdev->ih.ring[ring_index]) & 0xff;
 		src_data = le32_to_cpu(rdev->ih.ring[ring_index + 1]) & 0xfffffff;
 		ring_id = le32_to_cpu(rdev->ih.ring[ring_index + 2]) & 0xff;
@@ -8255,8 +8229,10 @@ restart_ih:
 	}
 	if (queue_hotplug)
 		schedule_work(&rdev->hotplug_work);
-	if (queue_reset)
-		schedule_work(&rdev->reset_work);
+	if (queue_reset) {
+		rdev->needs_reset = true;
+		wake_up_all(&rdev->fence_queue);
+	}
 	if (queue_thermal)
 		schedule_work(&rdev->pm.dpm.thermal.work);
 	rdev->ih.rptr = rptr;
@@ -8521,6 +8497,10 @@ static int cik_startup(struct radeon_device *rdev)
 	if (r)
 		return r;
 
+	r = radeon_kfd_resume(rdev);
+	if (r)
+		return r;
+
 	return 0;
 }
 
@@ -8569,6 +8549,7 @@ int cik_resume(struct radeon_device *rdev)
  */
 int cik_suspend(struct radeon_device *rdev)
 {
+	radeon_kfd_suspend(rdev);
 	radeon_pm_suspend(rdev);
 	dce6_audio_fini(rdev);
 	radeon_vm_manager_fini(rdev);
@@ -9447,6 +9428,9 @@ void dce8_bandwidth_update(struct radeon_device *rdev)
 	u32 num_heads = 0, lb_size;
 	int i;
 
+	if (!rdev->mode_info.mode_config_initialized)
+		return;
+
 	radeon_update_display_priority(rdev);
 
 	for (i = 0; i < rdev->num_crtc; i++) {
diff --git a/drivers/gpu/drm/radeon/cik_reg.h b/drivers/gpu/drm/radeon/cik_reg.h
index ca1bb6133580..79c45e8a536b 100644
--- a/drivers/gpu/drm/radeon/cik_reg.h
+++ b/drivers/gpu/drm/radeon/cik_reg.h
@@ -147,4 +147,140 @@
 
 #define CIK_LB_DESKTOP_HEIGHT                     0x6b0c
 
+#define CP_HQD_IQ_RPTR					0xC970u
+#define AQL_ENABLE					(1U << 0)
+
+#define IDLE					(1 << 2)
+
+struct cik_mqd {
+	uint32_t header;
+	uint32_t compute_dispatch_initiator;
+	uint32_t compute_dim_x;
+	uint32_t compute_dim_y;
+	uint32_t compute_dim_z;
+	uint32_t compute_start_x;
+	uint32_t compute_start_y;
+	uint32_t compute_start_z;
+	uint32_t compute_num_thread_x;
+	uint32_t compute_num_thread_y;
+	uint32_t compute_num_thread_z;
+	uint32_t compute_pipelinestat_enable;
+	uint32_t compute_perfcount_enable;
+	uint32_t compute_pgm_lo;
+	uint32_t compute_pgm_hi;
+	uint32_t compute_tba_lo;
+	uint32_t compute_tba_hi;
+	uint32_t compute_tma_lo;
+	uint32_t compute_tma_hi;
+	uint32_t compute_pgm_rsrc1;
+	uint32_t compute_pgm_rsrc2;
+	uint32_t compute_vmid;
+	uint32_t compute_resource_limits;
+	uint32_t compute_static_thread_mgmt_se0;
+	uint32_t compute_static_thread_mgmt_se1;
+	uint32_t compute_tmpring_size;
+	uint32_t compute_static_thread_mgmt_se2;
+	uint32_t compute_static_thread_mgmt_se3;
+	uint32_t compute_restart_x;
+	uint32_t compute_restart_y;
+	uint32_t compute_restart_z;
+	uint32_t compute_thread_trace_enable;
+	uint32_t compute_misc_reserved;
+	uint32_t compute_user_data_0;
+	uint32_t compute_user_data_1;
+	uint32_t compute_user_data_2;
+	uint32_t compute_user_data_3;
+	uint32_t compute_user_data_4;
+	uint32_t compute_user_data_5;
+	uint32_t compute_user_data_6;
+	uint32_t compute_user_data_7;
+	uint32_t compute_user_data_8;
+	uint32_t compute_user_data_9;
+	uint32_t compute_user_data_10;
+	uint32_t compute_user_data_11;
+	uint32_t compute_user_data_12;
+	uint32_t compute_user_data_13;
+	uint32_t compute_user_data_14;
+	uint32_t compute_user_data_15;
+	uint32_t cp_compute_csinvoc_count_lo;
+	uint32_t cp_compute_csinvoc_count_hi;
+	uint32_t cp_mqd_base_addr_lo;
+	uint32_t cp_mqd_base_addr_hi;
+	uint32_t cp_hqd_active;
+	uint32_t cp_hqd_vmid;
+	uint32_t cp_hqd_persistent_state;
+	uint32_t cp_hqd_pipe_priority;
+	uint32_t cp_hqd_queue_priority;
+	uint32_t cp_hqd_quantum;
+	uint32_t cp_hqd_pq_base_lo;
+	uint32_t cp_hqd_pq_base_hi;
+	uint32_t cp_hqd_pq_rptr;
+	uint32_t cp_hqd_pq_rptr_report_addr_lo;
+	uint32_t cp_hqd_pq_rptr_report_addr_hi;
+	uint32_t cp_hqd_pq_wptr_poll_addr_lo;
+	uint32_t cp_hqd_pq_wptr_poll_addr_hi;
+	uint32_t cp_hqd_pq_doorbell_control;
+	uint32_t cp_hqd_pq_wptr;
+	uint32_t cp_hqd_pq_control;
+	uint32_t cp_hqd_ib_base_addr_lo;
+	uint32_t cp_hqd_ib_base_addr_hi;
+	uint32_t cp_hqd_ib_rptr;
+	uint32_t cp_hqd_ib_control;
+	uint32_t cp_hqd_iq_timer;
+	uint32_t cp_hqd_iq_rptr;
+	uint32_t cp_hqd_dequeue_request;
+	uint32_t cp_hqd_dma_offload;
+	uint32_t cp_hqd_sema_cmd;
+	uint32_t cp_hqd_msg_type;
+	uint32_t cp_hqd_atomic0_preop_lo;
+	uint32_t cp_hqd_atomic0_preop_hi;
+	uint32_t cp_hqd_atomic1_preop_lo;
+	uint32_t cp_hqd_atomic1_preop_hi;
+	uint32_t cp_hqd_hq_status0;
+	uint32_t cp_hqd_hq_control0;
+	uint32_t cp_mqd_control;
+	uint32_t cp_mqd_query_time_lo;
+	uint32_t cp_mqd_query_time_hi;
+	uint32_t cp_mqd_connect_start_time_lo;
+	uint32_t cp_mqd_connect_start_time_hi;
+	uint32_t cp_mqd_connect_end_time_lo;
+	uint32_t cp_mqd_connect_end_time_hi;
+	uint32_t cp_mqd_connect_end_wf_count;
+	uint32_t cp_mqd_connect_end_pq_rptr;
+	uint32_t cp_mqd_connect_end_pq_wptr;
+	uint32_t cp_mqd_connect_end_ib_rptr;
+	uint32_t reserved_96;
+	uint32_t reserved_97;
+	uint32_t reserved_98;
+	uint32_t reserved_99;
+	uint32_t iqtimer_pkt_header;
+	uint32_t iqtimer_pkt_dw0;
+	uint32_t iqtimer_pkt_dw1;
+	uint32_t iqtimer_pkt_dw2;
+	uint32_t iqtimer_pkt_dw3;
+	uint32_t iqtimer_pkt_dw4;
+	uint32_t iqtimer_pkt_dw5;
+	uint32_t iqtimer_pkt_dw6;
+	uint32_t reserved_108;
+	uint32_t reserved_109;
+	uint32_t reserved_110;
+	uint32_t reserved_111;
+	uint32_t queue_doorbell_id0;
+	uint32_t queue_doorbell_id1;
+	uint32_t queue_doorbell_id2;
+	uint32_t queue_doorbell_id3;
+	uint32_t queue_doorbell_id4;
+	uint32_t queue_doorbell_id5;
+	uint32_t queue_doorbell_id6;
+	uint32_t queue_doorbell_id7;
+	uint32_t queue_doorbell_id8;
+	uint32_t queue_doorbell_id9;
+	uint32_t queue_doorbell_id10;
+	uint32_t queue_doorbell_id11;
+	uint32_t queue_doorbell_id12;
+	uint32_t queue_doorbell_id13;
+	uint32_t queue_doorbell_id14;
+	uint32_t queue_doorbell_id15;
+};
+
 #endif
diff --git a/drivers/gpu/drm/radeon/cik_sdma.c b/drivers/gpu/drm/radeon/cik_sdma.c
index c4ffa54b1e3d..42cd0cffe210 100644
--- a/drivers/gpu/drm/radeon/cik_sdma.c
+++ b/drivers/gpu/drm/radeon/cik_sdma.c
@@ -134,7 +134,7 @@ void cik_sdma_ring_ib_execute(struct radeon_device *rdev,
 			      struct radeon_ib *ib)
 {
 	struct radeon_ring *ring = &rdev->ring[ib->ring];
-	u32 extra_bits = (ib->vm ? ib->vm->id : 0) & 0xf;
+	u32 extra_bits = (ib->vm ? ib->vm->ids[ib->ring].id : 0) & 0xf;
 
 	if (rdev->wb.enabled) {
 		u32 next_rptr = ring->wptr + 5;
@@ -530,41 +530,38 @@ void cik_sdma_fini(struct radeon_device *rdev)
  * @src_offset: src GPU address
  * @dst_offset: dst GPU address
  * @num_gpu_pages: number of GPU pages to xfer
- * @fence: radeon fence object
+ * @resv: reservation object to sync to
  *
  * Copy GPU paging using the DMA engine (CIK).
  * Used by the radeon ttm implementation to move pages if
  * registered as the asic copy callback.
  */
-int cik_copy_dma(struct radeon_device *rdev,
-		 uint64_t src_offset, uint64_t dst_offset,
-		 unsigned num_gpu_pages,
-		 struct radeon_fence **fence)
+struct radeon_fence *cik_copy_dma(struct radeon_device *rdev,
+				  uint64_t src_offset, uint64_t dst_offset,
+				  unsigned num_gpu_pages,
+				  struct reservation_object *resv)
 {
-	struct radeon_semaphore *sem = NULL;
+	struct radeon_fence *fence;
+	struct radeon_sync sync;
 	int ring_index = rdev->asic->copy.dma_ring_index;
 	struct radeon_ring *ring = &rdev->ring[ring_index];
 	u32 size_in_bytes, cur_size_in_bytes;
 	int i, num_loops;
 	int r = 0;
 
-	r = radeon_semaphore_create(rdev, &sem);
-	if (r) {
-		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		return r;
-	}
+	radeon_sync_create(&sync);
 
 	size_in_bytes = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT);
 	num_loops = DIV_ROUND_UP(size_in_bytes, 0x1fffff);
 	r = radeon_ring_lock(rdev, ring, num_loops * 7 + 14);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		radeon_sync_free(rdev, &sync, NULL);
+		return ERR_PTR(r);
 	}
 
-	radeon_semaphore_sync_to(sem, *fence);
-	radeon_semaphore_sync_rings(rdev, sem, ring->idx);
+	radeon_sync_resv(rdev, &sync, resv, false);
+	radeon_sync_rings(rdev, &sync, ring->idx);
 
 	for (i = 0; i < num_loops; i++) {
 		cur_size_in_bytes = size_in_bytes;
@@ -582,17 +579,17 @@ int cik_copy_dma(struct radeon_device *rdev,
 		dst_offset += cur_size_in_bytes;
 	}
 
-	r = radeon_fence_emit(rdev, fence, ring->idx);
+	r = radeon_fence_emit(rdev, &fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
-		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		radeon_sync_free(rdev, &sync, NULL);
+		return ERR_PTR(r);
 	}
 
 	radeon_ring_unlock_commit(rdev, ring, false);
-	radeon_semaphore_free(rdev, &sem, *fence);
+	radeon_sync_free(rdev, &sync, fence);
 
-	return r;
+	return fence;
 }
 
 /**
@@ -610,16 +607,19 @@ int cik_sdma_ring_test(struct radeon_device *rdev,
 {
 	unsigned i;
 	int r;
-	void __iomem *ptr = (void *)rdev->vram_scratch.ptr;
+	unsigned index;
 	u32 tmp;
+	u64 gpu_addr;
 
-	if (!ptr) {
-		DRM_ERROR("invalid vram scratch pointer\n");
-		return -EINVAL;
-	}
+	if (ring->idx == R600_RING_TYPE_DMA_INDEX)
+		index = R600_WB_DMA_RING_TEST_OFFSET;
+	else
+		index = CAYMAN_WB_DMA1_RING_TEST_OFFSET;
+
+	gpu_addr = rdev->wb.gpu_addr + index;
 
 	tmp = 0xCAFEDEAD;
-	writel(tmp, ptr);
+	rdev->wb.wb[index/4] = cpu_to_le32(tmp);
 
 	r = radeon_ring_lock(rdev, ring, 5);
 	if (r) {
@@ -627,14 +627,14 @@ int cik_sdma_ring_test(struct radeon_device *rdev,
 		return r;
 	}
 	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_WRITE, SDMA_WRITE_SUB_OPCODE_LINEAR, 0));
-	radeon_ring_write(ring, rdev->vram_scratch.gpu_addr & 0xfffffffc);
-	radeon_ring_write(ring, upper_32_bits(rdev->vram_scratch.gpu_addr));
+	radeon_ring_write(ring, lower_32_bits(gpu_addr));
+	radeon_ring_write(ring, upper_32_bits(gpu_addr));
 	radeon_ring_write(ring, 1); /* number of DWs to follow */
 	radeon_ring_write(ring, 0xDEADBEEF);
 	radeon_ring_unlock_commit(rdev, ring, false);
 
 	for (i = 0; i < rdev->usec_timeout; i++) {
-		tmp = readl(ptr);
+		tmp = le32_to_cpu(rdev->wb.wb[index/4]);
 		if (tmp == 0xDEADBEEF)
 			break;
 		DRM_UDELAY(1);
@@ -663,17 +663,20 @@ int cik_sdma_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
 {
 	struct radeon_ib ib;
 	unsigned i;
+	unsigned index;
 	int r;
-	void __iomem *ptr = (void *)rdev->vram_scratch.ptr;
 	u32 tmp = 0;
+	u64 gpu_addr;
 
-	if (!ptr) {
-		DRM_ERROR("invalid vram scratch pointer\n");
-		return -EINVAL;
-	}
+	if (ring->idx == R600_RING_TYPE_DMA_INDEX)
+		index = R600_WB_DMA_RING_TEST_OFFSET;
+	else
+		index = CAYMAN_WB_DMA1_RING_TEST_OFFSET;
+
+	gpu_addr = rdev->wb.gpu_addr + index;
 
 	tmp = 0xCAFEDEAD;
-	writel(tmp, ptr);
+	rdev->wb.wb[index/4] = cpu_to_le32(tmp);
 
 	r = radeon_ib_get(rdev, ring->idx, &ib, NULL, 256);
 	if (r) {
@@ -682,8 +685,8 @@ int cik_sdma_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
 	}
 
 	ib.ptr[0] = SDMA_PACKET(SDMA_OPCODE_WRITE, SDMA_WRITE_SUB_OPCODE_LINEAR, 0);
-	ib.ptr[1] = rdev->vram_scratch.gpu_addr & 0xfffffffc;
-	ib.ptr[2] = upper_32_bits(rdev->vram_scratch.gpu_addr);
+	ib.ptr[1] = lower_32_bits(gpu_addr);
+	ib.ptr[2] = upper_32_bits(gpu_addr);
 	ib.ptr[3] = 1;
 	ib.ptr[4] = 0xDEADBEEF;
 	ib.length_dw = 5;
@@ -700,7 +703,7 @@ int cik_sdma_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
 		return r;
 	}
 	for (i = 0; i < rdev->usec_timeout; i++) {
-		tmp = readl(ptr);
+		tmp = le32_to_cpu(rdev->wb.wb[index/4]);
 		if (tmp == 0xDEADBEEF)
 			break;
 		DRM_UDELAY(1);
@@ -813,7 +816,6 @@ void cik_sdma_vm_write_pages(struct radeon_device *rdev,
 		for (; ndw > 0; ndw -= 2, --count, pe += 8) {
 			if (flags & R600_PTE_SYSTEM) {
 				value = radeon_vm_map_gart(rdev, addr);
-				value &= 0xFFFFFFFFFFFFF000ULL;
 			} else if (flags & R600_PTE_VALID) {
 				value = addr;
 			} else {
@@ -897,25 +899,24 @@ void cik_sdma_vm_pad_ib(struct radeon_ib *ib)
  * Update the page table base and flush the VM TLB
  * using sDMA (CIK).
  */
-void cik_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
+void cik_dma_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring,
+		      unsigned vm_id, uint64_t pd_addr)
 {
-	struct radeon_ring *ring = &rdev->ring[ridx];
-
-	if (vm == NULL)
-		return;
+	u32 extra_bits = (SDMA_POLL_REG_MEM_EXTRA_OP(0) |
+			  SDMA_POLL_REG_MEM_EXTRA_FUNC(0)); /* always */
 
 	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SRBM_WRITE, 0, 0xf000));
-	if (vm->id < 8) {
-		radeon_ring_write(ring, (VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (vm->id << 2)) >> 2);
+	if (vm_id < 8) {
+		radeon_ring_write(ring, (VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (vm_id << 2)) >> 2);
 	} else {
-		radeon_ring_write(ring, (VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + ((vm->id - 8) << 2)) >> 2);
+		radeon_ring_write(ring, (VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + ((vm_id - 8) << 2)) >> 2);
 	}
-	radeon_ring_write(ring, vm->pd_gpu_addr >> 12);
+	radeon_ring_write(ring, pd_addr >> 12);
 
 	/* update SH_MEM_* regs */
 	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SRBM_WRITE, 0, 0xf000));
 	radeon_ring_write(ring, SRBM_GFX_CNTL >> 2);
-	radeon_ring_write(ring, VMID(vm->id));
+	radeon_ring_write(ring, VMID(vm_id));
 
 	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SRBM_WRITE, 0, 0xf000));
 	radeon_ring_write(ring, SH_MEM_BASES >> 2);
@@ -938,11 +939,18 @@ void cik_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm
 	radeon_ring_write(ring, VMID(0));
 
 	/* flush HDP */
-	cik_sdma_hdp_flush_ring_emit(rdev, ridx);
+	cik_sdma_hdp_flush_ring_emit(rdev, ring->idx);
 
 	/* flush TLB */
 	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SRBM_WRITE, 0, 0xf000));
 	radeon_ring_write(ring, VM_INVALIDATE_REQUEST >> 2);
-	radeon_ring_write(ring, 1 << vm->id);
+	radeon_ring_write(ring, 1 << vm_id);
+
+	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_POLL_REG_MEM, 0, extra_bits));
+	radeon_ring_write(ring, VM_INVALIDATE_REQUEST >> 2);
+	radeon_ring_write(ring, 0);
+	radeon_ring_write(ring, 0); /* reference */
+	radeon_ring_write(ring, 0); /* mask */
+	radeon_ring_write(ring, (0xfff << 16) | 10); /* retry count, poll interval */
 }
 
diff --git a/drivers/gpu/drm/radeon/cikd.h b/drivers/gpu/drm/radeon/cikd.h
index 0c6e1b55d968..03003f8a6de6 100644
--- a/drivers/gpu/drm/radeon/cikd.h
+++ b/drivers/gpu/drm/radeon/cikd.h
@@ -30,6 +30,8 @@
 #define CIK_RB_BITMAP_WIDTH_PER_SH     2
 #define HAWAII_RB_BITMAP_WIDTH_PER_SH  4
 
+#define RADEON_NUM_OF_VMIDS	8
+
 /* DIDT IND registers */
 #define DIDT_SQ_CTRL0                                     0x0
 #       define DIDT_CTRL_EN                               (1 << 0)
@@ -184,7 +186,10 @@
 #define		DIG_THERM_DPM(x)			((x) << 14)
 #define		DIG_THERM_DPM_MASK			0x003FC000
 #define		DIG_THERM_DPM_SHIFT			14
-
+#define	CG_THERMAL_STATUS				0xC0300008
+#define		FDO_PWM_DUTY(x)				((x) << 9)
+#define		FDO_PWM_DUTY_MASK			(0xff << 9)
+#define		FDO_PWM_DUTY_SHIFT			9
 #define	CG_THERMAL_INT					0xC030000C
 #define		CI_DIG_THERM_INTH(x)			((x) << 8)
 #define		CI_DIG_THERM_INTH_MASK			0x0000FF00
@@ -194,7 +199,10 @@
 #define		CI_DIG_THERM_INTL_SHIFT			16
 #define 	THERM_INT_MASK_HIGH			(1 << 24)
 #define 	THERM_INT_MASK_LOW			(1 << 25)
-
+#define	CG_MULT_THERMAL_CTRL				0xC0300010
+#define		TEMP_SEL(x)				((x) << 20)
+#define		TEMP_SEL_MASK				(0xff << 20)
+#define		TEMP_SEL_SHIFT				20
 #define	CG_MULT_THERMAL_STATUS				0xC0300014
 #define		ASIC_MAX_TEMP(x)			((x) << 0)
 #define		ASIC_MAX_TEMP_MASK			0x000001ff
@@ -203,6 +211,36 @@
 #define		CTF_TEMP_MASK				0x0003fe00
 #define		CTF_TEMP_SHIFT				9
 
+#define	CG_FDO_CTRL0					0xC0300064
+#define		FDO_STATIC_DUTY(x)			((x) << 0)
+#define		FDO_STATIC_DUTY_MASK			0x000000FF
+#define		FDO_STATIC_DUTY_SHIFT			0
+#define	CG_FDO_CTRL1					0xC0300068
+#define		FMAX_DUTY100(x)				((x) << 0)
+#define		FMAX_DUTY100_MASK			0x000000FF
+#define		FMAX_DUTY100_SHIFT			0
+#define	CG_FDO_CTRL2					0xC030006C
+#define		TMIN(x)					((x) << 0)
+#define		TMIN_MASK				0x000000FF
+#define		TMIN_SHIFT				0
+#define		FDO_PWM_MODE(x)				((x) << 11)
+#define		FDO_PWM_MODE_MASK			(7 << 11)
+#define		FDO_PWM_MODE_SHIFT			11
+#define		TACH_PWM_RESP_RATE(x)			((x) << 25)
+#define		TACH_PWM_RESP_RATE_MASK			(0x7f << 25)
+#define		TACH_PWM_RESP_RATE_SHIFT		25
+#define CG_TACH_CTRL                                    0xC0300070
+#       define EDGE_PER_REV(x)                          ((x) << 0)
+#       define EDGE_PER_REV_MASK                        (0x7 << 0)
+#       define EDGE_PER_REV_SHIFT                       0
+#       define TARGET_PERIOD(x)                         ((x) << 3)
+#       define TARGET_PERIOD_MASK                       0xfffffff8
+#       define TARGET_PERIOD_SHIFT                      3
+#define CG_TACH_STATUS                                  0xC0300074
+#       define TACH_PERIOD(x)                           ((x) << 0)
+#       define TACH_PERIOD_MASK                         0xffffffff
+#       define TACH_PERIOD_SHIFT                        0
+
 #define CG_ECLK_CNTL                                    0xC05000AC
 #       define ECLK_DIVIDER_MASK                        0x7f
 #       define ECLK_DIR_CNTL_EN                         (1 << 8)
@@ -1137,6 +1175,9 @@
 #define			SH_MEM_ALIGNMENT_MODE_UNALIGNED			3
 #define		DEFAULT_MTYPE(x)				((x) << 4)
 #define		APE1_MTYPE(x)					((x) << 7)
+/* valid for both DEFAULT_MTYPE and APE1_MTYPE */
+#define	MTYPE_CACHED					0
+#define	MTYPE_NONCACHED					3
 
 #define	SX_DEBUG_1					0x9060
 
@@ -1447,6 +1488,16 @@
 #define CP_HQD_ACTIVE                                     0xC91C
 #define CP_HQD_VMID                                       0xC920
 
+#define CP_HQD_PERSISTENT_STATE				0xC924u
+#define	DEFAULT_CP_HQD_PERSISTENT_STATE			(0x33U << 8)
+
+#define CP_HQD_PIPE_PRIORITY				0xC928u
+#define CP_HQD_QUEUE_PRIORITY				0xC92Cu
+#define CP_HQD_QUANTUM					0xC930u
+#define	QUANTUM_EN					1U
+#define	QUANTUM_SCALE_1MS				(1U << 4)
+#define	QUANTUM_DURATION(x)				((x) << 8)
+
 #define CP_HQD_PQ_BASE                                    0xC934
 #define CP_HQD_PQ_BASE_HI                                 0xC938
 #define CP_HQD_PQ_RPTR                                    0xC93C
@@ -1474,12 +1525,32 @@
 #define		PRIV_STATE      			(1 << 30)
 #define		KMD_QUEUE      				(1 << 31)
 
-#define CP_HQD_DEQUEUE_REQUEST                          0xC974
+#define CP_HQD_IB_BASE_ADDR				0xC95Cu
+#define CP_HQD_IB_BASE_ADDR_HI			0xC960u
+#define CP_HQD_IB_RPTR					0xC964u
+#define CP_HQD_IB_CONTROL				0xC968u
+#define	IB_ATC_EN					(1U << 23)
+#define	DEFAULT_MIN_IB_AVAIL_SIZE			(3U << 20)
+
+#define CP_HQD_DEQUEUE_REQUEST			0xC974
+#define	DEQUEUE_REQUEST_DRAIN				1
+#define DEQUEUE_REQUEST_RESET				2
 
 #define CP_MQD_CONTROL                                  0xC99C
 #define		MQD_VMID(x)				((x) << 0)
 #define		MQD_VMID_MASK      			(0xf << 0)
 
+#define CP_HQD_SEMA_CMD					0xC97Cu
+#define CP_HQD_MSG_TYPE					0xC980u
+#define CP_HQD_ATOMIC0_PREOP_LO			0xC984u
+#define CP_HQD_ATOMIC0_PREOP_HI			0xC988u
+#define CP_HQD_ATOMIC1_PREOP_LO			0xC98Cu
+#define CP_HQD_ATOMIC1_PREOP_HI			0xC990u
+#define CP_HQD_HQ_SCHEDULER0			0xC994u
+#define CP_HQD_HQ_SCHEDULER1			0xC998u
+
+#define SH_STATIC_MEM_CONFIG			0x9604u
+
 #define DB_RENDER_CONTROL                               0x28000
 
 #define PA_SC_RASTER_CONFIG                             0x28350
@@ -2069,4 +2140,22 @@
 #define VCE_CMD_IB_AUTO		0x00000005
 #define VCE_CMD_SEMAPHORE	0x00000006
 
+#define ATC_VMID0_PASID_MAPPING					0x339Cu
+#define	ATC_VMID_PASID_MAPPING_UPDATE_STATUS	0x3398u
+#define	ATC_VMID_PASID_MAPPING_VALID				(1U << 31)
+
+#define ATC_VM_APERTURE0_CNTL					0x3310u
+#define	ATS_ACCESS_MODE_NEVER						0
+#define	ATS_ACCESS_MODE_ALWAYS						1
+
+#define ATC_VM_APERTURE0_CNTL2					0x3318u
+#define ATC_VM_APERTURE0_HIGH_ADDR				0x3308u
+#define ATC_VM_APERTURE0_LOW_ADDR				0x3300u
+#define ATC_VM_APERTURE1_CNTL					0x3314u
+#define ATC_VM_APERTURE1_CNTL2					0x331Cu
+#define ATC_VM_APERTURE1_HIGH_ADDR				0x330Cu
+#define ATC_VM_APERTURE1_LOW_ADDR				0x3304u
+
+#define IH_VMID_0_LUT						0x3D40u
+
 #endif
diff --git a/drivers/gpu/drm/radeon/cypress_dpm.c b/drivers/gpu/drm/radeon/cypress_dpm.c
index 47d31e915758..9aad0327e4d1 100644
--- a/drivers/gpu/drm/radeon/cypress_dpm.c
+++ b/drivers/gpu/drm/radeon/cypress_dpm.c
@@ -24,6 +24,7 @@
 
 #include "drmP.h"
 #include "radeon.h"
+#include "radeon_asic.h"
 #include "evergreend.h"
 #include "r600_dpm.h"
 #include "cypress_dpm.h"
diff --git a/drivers/gpu/drm/radeon/dce3_1_afmt.c b/drivers/gpu/drm/radeon/dce3_1_afmt.c
index 51800e340a57..bafdf92a5732 100644
--- a/drivers/gpu/drm/radeon/dce3_1_afmt.c
+++ b/drivers/gpu/drm/radeon/dce3_1_afmt.c
@@ -32,7 +32,7 @@ static void dce3_2_afmt_write_speaker_allocation(struct drm_encoder *encoder)
 	struct drm_connector *connector;
 	struct radeon_connector *radeon_connector = NULL;
 	u32 tmp;
-	u8 *sadb;
+	u8 *sadb = NULL;
 	int sad_count;
 
 	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
@@ -49,8 +49,8 @@ static void dce3_2_afmt_write_speaker_allocation(struct drm_encoder *encoder)
 
 	sad_count = drm_edid_to_speaker_allocation(radeon_connector->edid, &sadb);
 	if (sad_count < 0) {
-		DRM_ERROR("Couldn't read Speaker Allocation Data Block: %d\n", sad_count);
-		return;
+		DRM_DEBUG("Couldn't read Speaker Allocation Data Block: %d\n", sad_count);
+		sad_count = 0;
 	}
 
 	/* program the speaker allocation */
@@ -103,7 +103,7 @@ static void dce3_2_afmt_write_sad_regs(struct drm_encoder *encoder)
 	}
 
 	sad_count = drm_edid_to_sad(radeon_connector->edid, &sads);
-	if (sad_count < 0) {
+	if (sad_count <= 0) {
 		DRM_ERROR("Couldn't read SADs: %d\n", sad_count);
 		return;
 	}
@@ -165,7 +165,7 @@ void dce3_1_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *m
 
 	/* disable audio prior to setting up hw */
 	dig->afmt->pin = r600_audio_get_pin(rdev);
-	r600_audio_enable(rdev, dig->afmt->pin, false);
+	r600_audio_enable(rdev, dig->afmt->pin, 0);
 
 	r600_audio_set_dto(encoder, mode->clock);
 
@@ -240,5 +240,5 @@ void dce3_1_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *m
 	r600_hdmi_audio_workaround(encoder);
 
 	/* enable audio after to setting up hw */
-	r600_audio_enable(rdev, dig->afmt->pin, true);
+	r600_audio_enable(rdev, dig->afmt->pin, 0xf);
 }
diff --git a/drivers/gpu/drm/radeon/dce6_afmt.c b/drivers/gpu/drm/radeon/dce6_afmt.c
index ab29f953a767..f312edf4d50e 100644
--- a/drivers/gpu/drm/radeon/dce6_afmt.c
+++ b/drivers/gpu/drm/radeon/dce6_afmt.c
@@ -155,7 +155,7 @@ void dce6_afmt_write_speaker_allocation(struct drm_encoder *encoder)
 	struct drm_connector *connector;
 	struct radeon_connector *radeon_connector = NULL;
 	u32 offset, tmp;
-	u8 *sadb;
+	u8 *sadb = NULL;
 	int sad_count;
 
 	if (!dig || !dig->afmt || !dig->afmt->pin)
@@ -176,9 +176,9 @@ void dce6_afmt_write_speaker_allocation(struct drm_encoder *encoder)
 	}
 
 	sad_count = drm_edid_to_speaker_allocation(radeon_connector_edid(connector), &sadb);
-	if (sad_count <= 0) {
-		DRM_ERROR("Couldn't read Speaker Allocation Data Block: %d\n", sad_count);
-		return;
+	if (sad_count < 0) {
+		DRM_DEBUG("Couldn't read Speaker Allocation Data Block: %d\n", sad_count);
+		sad_count = 0;
 	}
 
 	/* program the speaker allocation */
@@ -284,13 +284,13 @@ static int dce6_audio_chipset_supported(struct radeon_device *rdev)
 
 void dce6_audio_enable(struct radeon_device *rdev,
 		       struct r600_audio_pin *pin,
-		       bool enable)
+		       u8 enable_mask)
 {
 	if (!pin)
 		return;
 
-	WREG32_ENDPOINT(pin->offset, AZ_F0_CODEC_PIN_CONTROL_HOTPLUG_CONTROL,
-			enable ? AUDIO_ENABLED : 0);
+	WREG32_ENDPOINT(pin->offset, AZ_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL,
+			enable_mask ? AUDIO_ENABLED : 0);
 }
 
 static const u32 pin_offsets[7] =
diff --git a/drivers/gpu/drm/drm_buffer.c b/drivers/gpu/drm/radeon/drm_buffer.c
index 86a4a4a60afc..f4e0f3a3d7b1 100644
--- a/drivers/gpu/drm/drm_buffer.c
+++ b/drivers/gpu/drm/radeon/drm_buffer.c
@@ -33,7 +33,7 @@
  */
 
 #include <linux/export.h>
-#include <drm/drm_buffer.h>
+#include "drm_buffer.h"
 
 /**
  * Allocate the drm buffer object.
@@ -86,7 +86,6 @@ error_out:
 	kfree(*buf);
 	return -ENOMEM;
 }
-EXPORT_SYMBOL(drm_buffer_alloc);
 
 /**
  * Copy the user data to the begin of the buffer and reset the processing
@@ -123,7 +122,6 @@ int drm_buffer_copy_from_user(struct drm_buffer *buf,
 	buf->iterator = 0;
 	return 0;
 }
-EXPORT_SYMBOL(drm_buffer_copy_from_user);
 
 /**
  * Free the drm buffer object
@@ -141,7 +139,6 @@ void drm_buffer_free(struct drm_buffer *buf)
 		kfree(buf);
 	}
 }
-EXPORT_SYMBOL(drm_buffer_free);
 
 /**
  * Read an object from buffer that may be split to multiple parts. If object
@@ -178,4 +175,3 @@ void *drm_buffer_read_object(struct drm_buffer *buf,
 	drm_buffer_advance(buf, objsize);
 	return obj;
 }
-EXPORT_SYMBOL(drm_buffer_read_object);
diff --git a/drivers/gpu/drm/radeon/drm_buffer.h b/drivers/gpu/drm/radeon/drm_buffer.h
new file mode 100644
index 000000000000..c80d3a340b94
--- /dev/null
+++ b/drivers/gpu/drm/radeon/drm_buffer.h
@@ -0,0 +1,148 @@
+/**************************************************************************
+ *
+ * Copyright 2010 Pauli Nieminen.
+ * All Rights Reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
+ * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+ * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+ * USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ *
+ **************************************************************************/
+/*
+ * Multipart buffer for coping data which is larger than the page size.
+ *
+ * Authors:
+ * Pauli Nieminen <suokkos-at-gmail-dot-com>
+ */
+
+#ifndef _DRM_BUFFER_H_
+#define _DRM_BUFFER_H_
+
+#include <drm/drmP.h>
+
+struct drm_buffer {
+	int iterator;
+	int size;
+	char *data[];
+};
+
+
+/**
+ * Return the index of page that buffer is currently pointing at.
+ */
+static inline int drm_buffer_page(struct drm_buffer *buf)
+{
+	return buf->iterator / PAGE_SIZE;
+}
+/**
+ * Return the index of the current byte in the page
+ */
+static inline int drm_buffer_index(struct drm_buffer *buf)
+{
+	return buf->iterator & (PAGE_SIZE - 1);
+}
+/**
+ * Return number of bytes that is left to process
+ */
+static inline int drm_buffer_unprocessed(struct drm_buffer *buf)
+{
+	return buf->size - buf->iterator;
+}
+
+/**
+ * Advance the buffer iterator number of bytes that is given.
+ */
+static inline void drm_buffer_advance(struct drm_buffer *buf, int bytes)
+{
+	buf->iterator += bytes;
+}
+
+/**
+ * Allocate the drm buffer object.
+ *
+ *   buf: A pointer to a pointer where the object is stored.
+ *   size: The number of bytes to allocate.
+ */
+extern int drm_buffer_alloc(struct drm_buffer **buf, int size);
+
+/**
+ * Copy the user data to the begin of the buffer and reset the processing
+ * iterator.
+ *
+ *   user_data: A pointer the data that is copied to the buffer.
+ *   size: The Number of bytes to copy.
+ */
+extern int drm_buffer_copy_from_user(struct drm_buffer *buf,
+		void __user *user_data, int size);
+
+/**
+ * Free the drm buffer object
+ */
+extern void drm_buffer_free(struct drm_buffer *buf);
+
+/**
+ * Read an object from buffer that may be split to multiple parts. If object
+ * is not split function just returns the pointer to object in buffer. But in
+ * case of split object data is copied to given stack object that is suplied
+ * by caller.
+ *
+ * The processing location of the buffer is also advanced to the next byte
+ * after the object.
+ *
+ *   objsize: The size of the objet in bytes.
+ *   stack_obj: A pointer to a memory location where object can be copied.
+ */
+extern void *drm_buffer_read_object(struct drm_buffer *buf,
+		int objsize, void *stack_obj);
+
+/**
+ * Returns the pointer to the dword which is offset number of elements from the
+ * current processing location.
+ *
+ * Caller must make sure that dword is not split in the buffer. This
+ * requirement is easily met if all the sizes of objects in buffer are
+ * multiples of dword and PAGE_SIZE is multiple dword.
+ *
+ * Call to this function doesn't change the processing location.
+ *
+ *   offset: The index of the dword relative to the internat iterator.
+ */
+static inline void *drm_buffer_pointer_to_dword(struct drm_buffer *buffer,
+		int offset)
+{
+	int iter = buffer->iterator + offset * 4;
+	return &buffer->data[iter / PAGE_SIZE][iter & (PAGE_SIZE - 1)];
+}
+/**
+ * Returns the pointer to the dword which is offset number of elements from
+ * the current processing location.
+ *
+ * Call to this function doesn't change the processing location.
+ *
+ *   offset: The index of the byte relative to the internat iterator.
+ */
+static inline void *drm_buffer_pointer_to_byte(struct drm_buffer *buffer,
+		int offset)
+{
+	int iter = buffer->iterator + offset;
+	return &buffer->data[iter / PAGE_SIZE][iter & (PAGE_SIZE - 1)];
+}
+
+#endif
diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c
index e50807c29f69..85995b4e3338 100644
--- a/drivers/gpu/drm/radeon/evergreen.c
+++ b/drivers/gpu/drm/radeon/evergreen.c
@@ -22,7 +22,6 @@
  * Authors: Alex Deucher
  */
 #include <linux/firmware.h>
-#include <linux/platform_device.h>
 #include <linux/slab.h>
 #include <drm/drmP.h>
 #include "radeon.h"
@@ -2346,6 +2345,9 @@ void evergreen_bandwidth_update(struct radeon_device *rdev)
 	u32 num_heads = 0, lb_size;
 	int i;
 
+	if (!rdev->mode_info.mode_config_initialized)
+		return;
+
 	radeon_update_display_priority(rdev);
 
 	for (i = 0; i < rdev->num_crtc; i++) {
@@ -2553,6 +2555,7 @@ void evergreen_mc_stop(struct radeon_device *rdev, struct evergreen_mc_save *sav
 					WREG32(EVERGREEN_CRTC_UPDATE_LOCK + crtc_offsets[i], 1);
 					tmp |= EVERGREEN_CRTC_BLANK_DATA_EN;
 					WREG32(EVERGREEN_CRTC_BLANK_CONTROL + crtc_offsets[i], tmp);
+					WREG32(EVERGREEN_CRTC_UPDATE_LOCK + crtc_offsets[i], 0);
 				}
 			} else {
 				tmp = RREG32(EVERGREEN_CRTC_CONTROL + crtc_offsets[i]);
@@ -3006,7 +3009,7 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
 	u32 vgt_cache_invalidation;
 	u32 hdp_host_path_cntl, tmp;
 	u32 disabled_rb_mask;
-	int i, j, num_shader_engines, ps_thread_count;
+	int i, j, ps_thread_count;
 
 	switch (rdev->family) {
 	case CHIP_CYPRESS:
@@ -3304,8 +3307,6 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
 	rdev->config.evergreen.tile_config |=
 		((gb_addr_config & 0x30000000) >> 28) << 12;
 
-	num_shader_engines = (gb_addr_config & NUM_SHADER_ENGINES(3) >> 12) + 1;
-
 	if ((rdev->family >= CHIP_CEDAR) && (rdev->family <= CHIP_HEMLOCK)) {
 		u32 efuse_straps_4;
 		u32 efuse_straps_3;
@@ -4023,7 +4024,7 @@ int sumo_rlc_init(struct radeon_device *rdev)
 		if (rdev->rlc.save_restore_obj == NULL) {
 			r = radeon_bo_create(rdev, dws * 4, PAGE_SIZE, true,
 					     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
-					     &rdev->rlc.save_restore_obj);
+					     NULL, &rdev->rlc.save_restore_obj);
 			if (r) {
 				dev_warn(rdev->dev, "(%d) create RLC sr bo failed\n", r);
 				return r;
@@ -4102,7 +4103,7 @@ int sumo_rlc_init(struct radeon_device *rdev)
 		if (rdev->rlc.clear_state_obj == NULL) {
 			r = radeon_bo_create(rdev, dws * 4, PAGE_SIZE, true,
 					     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
-					     &rdev->rlc.clear_state_obj);
+					     NULL, &rdev->rlc.clear_state_obj);
 			if (r) {
 				dev_warn(rdev->dev, "(%d) create RLC c bo failed\n", r);
 				sumo_rlc_fini(rdev);
@@ -4179,7 +4180,7 @@ int sumo_rlc_init(struct radeon_device *rdev)
 			r = radeon_bo_create(rdev, rdev->rlc.cp_table_size,
 					     PAGE_SIZE, true,
 					     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
-					     &rdev->rlc.cp_table_obj);
+					     NULL, &rdev->rlc.cp_table_obj);
 			if (r) {
 				dev_warn(rdev->dev, "(%d) create RLC cp table bo failed\n", r);
 				sumo_rlc_fini(rdev);
diff --git a/drivers/gpu/drm/radeon/evergreen_cs.c b/drivers/gpu/drm/radeon/evergreen_cs.c
index 5c8b358f9fba..924b1b7ab455 100644
--- a/drivers/gpu/drm/radeon/evergreen_cs.c
+++ b/drivers/gpu/drm/radeon/evergreen_cs.c
@@ -35,7 +35,7 @@
 #define MIN(a,b)                   (((a)<(b))?(a):(b))
 
 int r600_dma_cs_next_reloc(struct radeon_cs_parser *p,
-			   struct radeon_cs_reloc **cs_reloc);
+			   struct radeon_bo_list **cs_reloc);
 struct evergreen_cs_track {
 	u32			group_size;
 	u32			nbanks;
@@ -1094,7 +1094,7 @@ static int evergreen_cs_parse_packet0(struct radeon_cs_parser *p,
 static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 {
 	struct evergreen_cs_track *track = (struct evergreen_cs_track *)p->track;
-	struct radeon_cs_reloc *reloc;
+	struct radeon_bo_list *reloc;
 	u32 last_reg;
 	u32 m, i, tmp, *ib;
 	int r;
@@ -1792,7 +1792,7 @@ static bool evergreen_is_safe_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 static int evergreen_packet3_check(struct radeon_cs_parser *p,
 				   struct radeon_cs_packet *pkt)
 {
-	struct radeon_cs_reloc *reloc;
+	struct radeon_bo_list *reloc;
 	struct evergreen_cs_track *track;
 	volatile u32 *ib;
 	unsigned idx;
@@ -2661,7 +2661,7 @@ int evergreen_cs_parse(struct radeon_cs_parser *p)
 			p->track = NULL;
 			return r;
 		}
-	} while (p->idx < p->chunks[p->chunk_ib_idx].length_dw);
+	} while (p->idx < p->chunk_ib->length_dw);
 #if 0
 	for (r = 0; r < p->ib.length_dw; r++) {
 		printk(KERN_INFO "%05d  0x%08X\n", r, p->ib.ptr[r]);
@@ -2684,8 +2684,8 @@ int evergreen_cs_parse(struct radeon_cs_parser *p)
  **/
 int evergreen_dma_cs_parse(struct radeon_cs_parser *p)
 {
-	struct radeon_cs_chunk *ib_chunk = &p->chunks[p->chunk_ib_idx];
-	struct radeon_cs_reloc *src_reloc, *dst_reloc, *dst2_reloc;
+	struct radeon_cs_chunk *ib_chunk = p->chunk_ib;
+	struct radeon_bo_list *src_reloc, *dst_reloc, *dst2_reloc;
 	u32 header, cmd, count, sub_cmd;
 	volatile u32 *ib = p->ib.ptr;
 	u32 idx;
@@ -3100,7 +3100,7 @@ int evergreen_dma_cs_parse(struct radeon_cs_parser *p)
 			DRM_ERROR("Unknown packet type %d at %d !\n", cmd, idx);
 			return -EINVAL;
 		}
-	} while (p->idx < p->chunks[p->chunk_ib_idx].length_dw);
+	} while (p->idx < p->chunk_ib->length_dw);
 #if 0
 	for (r = 0; r < p->ib->length_dw; r++) {
 		printk(KERN_INFO "%05d  0x%08X\n", r, p->ib.ptr[r]);
diff --git a/drivers/gpu/drm/radeon/evergreen_dma.c b/drivers/gpu/drm/radeon/evergreen_dma.c
index afaba388c36d..96535aa8659c 100644
--- a/drivers/gpu/drm/radeon/evergreen_dma.c
+++ b/drivers/gpu/drm/radeon/evergreen_dma.c
@@ -104,35 +104,33 @@ void evergreen_dma_ring_ib_execute(struct radeon_device *rdev,
  * Used by the radeon ttm implementation to move pages if
  * registered as the asic copy callback.
  */
-int evergreen_copy_dma(struct radeon_device *rdev,
-		       uint64_t src_offset, uint64_t dst_offset,
-		       unsigned num_gpu_pages,
-		       struct radeon_fence **fence)
+struct radeon_fence *evergreen_copy_dma(struct radeon_device *rdev,
+					uint64_t src_offset,
+					uint64_t dst_offset,
+					unsigned num_gpu_pages,
+					struct reservation_object *resv)
 {
-	struct radeon_semaphore *sem = NULL;
+	struct radeon_fence *fence;
+	struct radeon_sync sync;
 	int ring_index = rdev->asic->copy.dma_ring_index;
 	struct radeon_ring *ring = &rdev->ring[ring_index];
 	u32 size_in_dw, cur_size_in_dw;
 	int i, num_loops;
 	int r = 0;
 
-	r = radeon_semaphore_create(rdev, &sem);
-	if (r) {
-		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		return r;
-	}
+	radeon_sync_create(&sync);
 
 	size_in_dw = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT) / 4;
 	num_loops = DIV_ROUND_UP(size_in_dw, 0xfffff);
 	r = radeon_ring_lock(rdev, ring, num_loops * 5 + 11);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		radeon_sync_free(rdev, &sync, NULL);
+		return ERR_PTR(r);
 	}
 
-	radeon_semaphore_sync_to(sem, *fence);
-	radeon_semaphore_sync_rings(rdev, sem, ring->idx);
+	radeon_sync_resv(rdev, &sync, resv, false);
+	radeon_sync_rings(rdev, &sync, ring->idx);
 
 	for (i = 0; i < num_loops; i++) {
 		cur_size_in_dw = size_in_dw;
@@ -148,17 +146,17 @@ int evergreen_copy_dma(struct radeon_device *rdev,
 		dst_offset += cur_size_in_dw * 4;
 	}
 
-	r = radeon_fence_emit(rdev, fence, ring->idx);
+	r = radeon_fence_emit(rdev, &fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
-		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		radeon_sync_free(rdev, &sync, NULL);
+		return ERR_PTR(r);
 	}
 
 	radeon_ring_unlock_commit(rdev, ring, false);
-	radeon_semaphore_free(rdev, &sem, *fence);
+	radeon_sync_free(rdev, &sync, fence);
 
-	return r;
+	return fence;
 }
 
 /**
diff --git a/drivers/gpu/drm/radeon/evergreen_hdmi.c b/drivers/gpu/drm/radeon/evergreen_hdmi.c
index 278c7a139d74..53abd9b17a50 100644
--- a/drivers/gpu/drm/radeon/evergreen_hdmi.c
+++ b/drivers/gpu/drm/radeon/evergreen_hdmi.c
@@ -38,6 +38,37 @@ extern void dce6_afmt_select_pin(struct drm_encoder *encoder);
 extern void dce6_afmt_write_latency_fields(struct drm_encoder *encoder,
 					   struct drm_display_mode *mode);
 
+/* enable the audio stream */
+static void dce4_audio_enable(struct radeon_device *rdev,
+			      struct r600_audio_pin *pin,
+			      u8 enable_mask)
+{
+	u32 tmp = RREG32(AZ_HOT_PLUG_CONTROL);
+
+	if (!pin)
+		return;
+
+	if (enable_mask) {
+		tmp |= AUDIO_ENABLED;
+		if (enable_mask & 1)
+			tmp |= PIN0_AUDIO_ENABLED;
+		if (enable_mask & 2)
+			tmp |= PIN1_AUDIO_ENABLED;
+		if (enable_mask & 4)
+			tmp |= PIN2_AUDIO_ENABLED;
+		if (enable_mask & 8)
+			tmp |= PIN3_AUDIO_ENABLED;
+	} else {
+		tmp &= ~(AUDIO_ENABLED |
+			 PIN0_AUDIO_ENABLED |
+			 PIN1_AUDIO_ENABLED |
+			 PIN2_AUDIO_ENABLED |
+			 PIN3_AUDIO_ENABLED);
+	}
+
+	WREG32(AZ_HOT_PLUG_CONTROL, tmp);
+}
+
 /*
  * update the N and CTS parameters for a given pixel clock rate
  */
@@ -102,7 +133,7 @@ static void dce4_afmt_write_speaker_allocation(struct drm_encoder *encoder)
 	struct drm_connector *connector;
 	struct radeon_connector *radeon_connector = NULL;
 	u32 tmp;
-	u8 *sadb;
+	u8 *sadb = NULL;
 	int sad_count;
 
 	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
@@ -118,9 +149,9 @@ static void dce4_afmt_write_speaker_allocation(struct drm_encoder *encoder)
 	}
 
 	sad_count = drm_edid_to_speaker_allocation(radeon_connector_edid(connector), &sadb);
-	if (sad_count <= 0) {
-		DRM_ERROR("Couldn't read Speaker Allocation Data Block: %d\n", sad_count);
-		return;
+	if (sad_count < 0) {
+		DRM_DEBUG("Couldn't read Speaker Allocation Data Block: %d\n", sad_count);
+		sad_count = 0;
 	}
 
 	/* program the speaker allocation */
@@ -318,10 +349,10 @@ void evergreen_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode
 	/* disable audio prior to setting up hw */
 	if (ASIC_IS_DCE6(rdev)) {
 		dig->afmt->pin = dce6_audio_get_pin(rdev);
-		dce6_audio_enable(rdev, dig->afmt->pin, false);
+		dce6_audio_enable(rdev, dig->afmt->pin, 0);
 	} else {
 		dig->afmt->pin = r600_audio_get_pin(rdev);
-		r600_audio_enable(rdev, dig->afmt->pin, false);
+		dce4_audio_enable(rdev, dig->afmt->pin, 0);
 	}
 
 	evergreen_audio_set_dto(encoder, mode->clock);
@@ -463,13 +494,15 @@ void evergreen_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode
 
 	/* enable audio after to setting up hw */
 	if (ASIC_IS_DCE6(rdev))
-		dce6_audio_enable(rdev, dig->afmt->pin, true);
+		dce6_audio_enable(rdev, dig->afmt->pin, 1);
 	else
-		r600_audio_enable(rdev, dig->afmt->pin, true);
+		dce4_audio_enable(rdev, dig->afmt->pin, 0xf);
 }
 
 void evergreen_hdmi_enable(struct drm_encoder *encoder, bool enable)
 {
+	struct drm_device *dev = encoder->dev;
+	struct radeon_device *rdev = dev->dev_private;
 	struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
 	struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv;
 
@@ -482,6 +515,14 @@ void evergreen_hdmi_enable(struct drm_encoder *encoder, bool enable)
 	if (!enable && !dig->afmt->enabled)
 		return;
 
+	if (!enable && dig->afmt->pin) {
+		if (ASIC_IS_DCE6(rdev))
+			dce6_audio_enable(rdev, dig->afmt->pin, 0);
+		else
+			dce4_audio_enable(rdev, dig->afmt->pin, 0);
+		dig->afmt->pin = NULL;
+	}
+
 	dig->afmt->enabled = enable;
 
 	DRM_DEBUG("%sabling HDMI interface @ 0x%04X for encoder 0x%x\n",
diff --git a/drivers/gpu/drm/radeon/kv_dpm.c b/drivers/gpu/drm/radeon/kv_dpm.c
index 67cb472d188c..e3e9c10cfba9 100644
--- a/drivers/gpu/drm/radeon/kv_dpm.c
+++ b/drivers/gpu/drm/radeon/kv_dpm.c
@@ -2725,7 +2725,11 @@ int kv_dpm_init(struct radeon_device *rdev)
 
         pi->sram_end = SMC_RAM_END;
 
-	pi->enable_nb_dpm = true;
+	/* Enabling nb dpm on an asrock system prevents dpm from working */
+	if (rdev->pdev->subsystem_vendor == 0x1849)
+		pi->enable_nb_dpm = false;
+	else
+		pi->enable_nb_dpm = true;
 
 	pi->caps_power_containment = true;
 	pi->caps_cac = true;
@@ -2740,10 +2744,17 @@ int kv_dpm_init(struct radeon_device *rdev)
 	pi->caps_sclk_ds = true;
 	pi->enable_auto_thermal_throttling = true;
 	pi->disable_nb_ps3_in_battery = false;
-	if (radeon_bapm == 0)
+	if (radeon_bapm == -1) {
+		/* only enable bapm on KB, ML by default */
+		if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS)
+			pi->bapm_enable = true;
+		else
+			pi->bapm_enable = false;
+	} else if (radeon_bapm == 0) {
 		pi->bapm_enable = false;
-	else
+	} else {
 		pi->bapm_enable = true;
+	}
 	pi->voltage_drop_t = 0;
 	pi->caps_sclk_throttle_low_notification = false;
 	pi->caps_fps = false; /* true? */
@@ -2787,6 +2798,8 @@ void kv_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev,
 		tmp = (RREG32_SMC(SMU_VOLTAGE_STATUS) & SMU_VOLTAGE_CURRENT_LEVEL_MASK) >>
 			SMU_VOLTAGE_CURRENT_LEVEL_SHIFT;
 		vddc = kv_convert_8bit_index_to_voltage(rdev, (u16)tmp);
+		seq_printf(m, "uvd    %sabled\n", pi->uvd_power_gated ? "dis" : "en");
+		seq_printf(m, "vce    %sabled\n", pi->vce_power_gated ? "dis" : "en");
 		seq_printf(m, "power level %d    sclk: %u vddc: %u\n",
 			   current_index, sclk, vddc);
 	}
diff --git a/drivers/gpu/drm/radeon/mkregtable.c b/drivers/gpu/drm/radeon/mkregtable.c
index 4a85bb644e24..b928c17bdeed 100644
--- a/drivers/gpu/drm/radeon/mkregtable.c
+++ b/drivers/gpu/drm/radeon/mkregtable.c
@@ -347,7 +347,7 @@ static inline void list_splice_tail_init(struct list_head *list,
  * list_entry - get the struct for this entry
  * @ptr:	the &struct list_head pointer.
  * @type:	the type of the struct this is embedded in.
- * @member:	the name of the list_struct within the struct.
+ * @member:	the name of the list_head within the struct.
  */
 #define list_entry(ptr, type, member) \
 	container_of(ptr, type, member)
@@ -356,7 +356,7 @@ static inline void list_splice_tail_init(struct list_head *list,
  * list_first_entry - get the first element from a list
  * @ptr:	the list head to take the element from.
  * @type:	the type of the struct this is embedded in.
- * @member:	the name of the list_struct within the struct.
+ * @member:	the name of the list_head within the struct.
  *
  * Note, that list is expected to be not empty.
  */
@@ -406,7 +406,7 @@ static inline void list_splice_tail_init(struct list_head *list,
  * list_for_each_entry	-	iterate over list of given type
  * @pos:	the type * to use as a loop cursor.
  * @head:	the head for your list.
- * @member:	the name of the list_struct within the struct.
+ * @member:	the name of the list_head within the struct.
  */
 #define list_for_each_entry(pos, head, member)				\
 	for (pos = list_entry((head)->next, typeof(*pos), member);	\
@@ -417,7 +417,7 @@ static inline void list_splice_tail_init(struct list_head *list,
  * list_for_each_entry_reverse - iterate backwards over list of given type.
  * @pos:	the type * to use as a loop cursor.
  * @head:	the head for your list.
- * @member:	the name of the list_struct within the struct.
+ * @member:	the name of the list_head within the struct.
  */
 #define list_for_each_entry_reverse(pos, head, member)			\
 	for (pos = list_entry((head)->prev, typeof(*pos), member);	\
@@ -428,7 +428,7 @@ static inline void list_splice_tail_init(struct list_head *list,
  * list_prepare_entry - prepare a pos entry for use in list_for_each_entry_continue()
  * @pos:	the type * to use as a start point
  * @head:	the head of the list
- * @member:	the name of the list_struct within the struct.
+ * @member:	the name of the list_head within the struct.
  *
  * Prepares a pos entry for use as a start point in list_for_each_entry_continue().
  */
@@ -439,7 +439,7 @@ static inline void list_splice_tail_init(struct list_head *list,
  * list_for_each_entry_continue - continue iteration over list of given type
  * @pos:	the type * to use as a loop cursor.
  * @head:	the head for your list.
- * @member:	the name of the list_struct within the struct.
+ * @member:	the name of the list_head within the struct.
  *
  * Continue to iterate over list of given type, continuing after
  * the current position.
@@ -453,7 +453,7 @@ static inline void list_splice_tail_init(struct list_head *list,
  * list_for_each_entry_continue_reverse - iterate backwards from the given point
  * @pos:	the type * to use as a loop cursor.
  * @head:	the head for your list.
- * @member:	the name of the list_struct within the struct.
+ * @member:	the name of the list_head within the struct.
  *
  * Start to iterate over list of given type backwards, continuing after
  * the current position.
@@ -467,7 +467,7 @@ static inline void list_splice_tail_init(struct list_head *list,
  * list_for_each_entry_from - iterate over list of given type from the current point
  * @pos:	the type * to use as a loop cursor.
  * @head:	the head for your list.
- * @member:	the name of the list_struct within the struct.
+ * @member:	the name of the list_head within the struct.
  *
  * Iterate over list of given type, continuing from current position.
  */
@@ -480,7 +480,7 @@ static inline void list_splice_tail_init(struct list_head *list,
  * @pos:	the type * to use as a loop cursor.
  * @n:		another type * to use as temporary storage
  * @head:	the head for your list.
- * @member:	the name of the list_struct within the struct.
+ * @member:	the name of the list_head within the struct.
  */
 #define list_for_each_entry_safe(pos, n, head, member)			\
 	for (pos = list_entry((head)->next, typeof(*pos), member),	\
@@ -493,7 +493,7 @@ static inline void list_splice_tail_init(struct list_head *list,
  * @pos:	the type * to use as a loop cursor.
  * @n:		another type * to use as temporary storage
  * @head:	the head for your list.
- * @member:	the name of the list_struct within the struct.
+ * @member:	the name of the list_head within the struct.
  *
  * Iterate over list of given type, continuing after current point,
  * safe against removal of list entry.
@@ -509,7 +509,7 @@ static inline void list_splice_tail_init(struct list_head *list,
  * @pos:	the type * to use as a loop cursor.
  * @n:		another type * to use as temporary storage
  * @head:	the head for your list.
- * @member:	the name of the list_struct within the struct.
+ * @member:	the name of the list_head within the struct.
  *
  * Iterate over list of given type from current point, safe against
  * removal of list entry.
@@ -524,7 +524,7 @@ static inline void list_splice_tail_init(struct list_head *list,
  * @pos:	the type * to use as a loop cursor.
  * @n:		another type * to use as temporary storage
  * @head:	the head for your list.
- * @member:	the name of the list_struct within the struct.
+ * @member:	the name of the list_head within the struct.
  *
  * Iterate backwards over list of given type, safe against removal
  * of list entry.
diff --git a/drivers/gpu/drm/radeon/ni.c b/drivers/gpu/drm/radeon/ni.c
index 3faee58946dd..aea48c89b241 100644
--- a/drivers/gpu/drm/radeon/ni.c
+++ b/drivers/gpu/drm/radeon/ni.c
@@ -1373,6 +1373,7 @@ void cayman_fence_ring_emit(struct radeon_device *rdev,
 void cayman_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
 {
 	struct radeon_ring *ring = &rdev->ring[ib->ring];
+	unsigned vm_id = ib->vm ? ib->vm->ids[ib->ring].id : 0;
 	u32 cp_coher_cntl = PACKET3_FULL_CACHE_ENA | PACKET3_TC_ACTION_ENA |
 		PACKET3_SH_ACTION_ENA;
 
@@ -1395,15 +1396,14 @@ void cayman_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
 #endif
 			  (ib->gpu_addr & 0xFFFFFFFC));
 	radeon_ring_write(ring, upper_32_bits(ib->gpu_addr) & 0xFF);
-	radeon_ring_write(ring, ib->length_dw | 
-			  (ib->vm ? (ib->vm->id << 24) : 0));
+	radeon_ring_write(ring, ib->length_dw | (vm_id << 24));
 
 	/* flush read cache over gart for this vmid */
 	radeon_ring_write(ring, PACKET3(PACKET3_SURFACE_SYNC, 3));
 	radeon_ring_write(ring, PACKET3_ENGINE_ME | cp_coher_cntl);
 	radeon_ring_write(ring, 0xFFFFFFFF);
 	radeon_ring_write(ring, 0);
-	radeon_ring_write(ring, ((ib->vm ? ib->vm->id : 0) << 24) | 10); /* poll interval */
+	radeon_ring_write(ring, (vm_id << 24) | 10); /* poll interval */
 }
 
 static void cayman_cp_enable(struct radeon_device *rdev, bool enable)
@@ -2502,15 +2502,11 @@ void cayman_vm_decode_fault(struct radeon_device *rdev,
  * Update the page table base and flush the VM TLB
  * using the CP (cayman-si).
  */
-void cayman_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
+void cayman_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring,
+		     unsigned vm_id, uint64_t pd_addr)
 {
-	struct radeon_ring *ring = &rdev->ring[ridx];
-
-	if (vm == NULL)
-		return;
-
-	radeon_ring_write(ring, PACKET0(VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (vm->id << 2), 0));
-	radeon_ring_write(ring, vm->pd_gpu_addr >> 12);
+	radeon_ring_write(ring, PACKET0(VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (vm_id << 2), 0));
+	radeon_ring_write(ring, pd_addr >> 12);
 
 	/* flush hdp cache */
 	radeon_ring_write(ring, PACKET0(HDP_MEM_COHERENCY_FLUSH_CNTL, 0));
@@ -2518,7 +2514,17 @@ void cayman_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
 
 	/* bits 0-7 are the VM contexts0-7 */
 	radeon_ring_write(ring, PACKET0(VM_INVALIDATE_REQUEST, 0));
-	radeon_ring_write(ring, 1 << vm->id);
+	radeon_ring_write(ring, 1 << vm_id);
+
+	/* wait for the invalidate to complete */
+	radeon_ring_write(ring, PACKET3(PACKET3_WAIT_REG_MEM, 5));
+	radeon_ring_write(ring, (WAIT_REG_MEM_FUNCTION(0) |  /* always */
+				 WAIT_REG_MEM_ENGINE(0))); /* me */
+	radeon_ring_write(ring, VM_INVALIDATE_REQUEST >> 2);
+	radeon_ring_write(ring, 0);
+	radeon_ring_write(ring, 0); /* ref */
+	radeon_ring_write(ring, 0); /* mask */
+	radeon_ring_write(ring, 0x20); /* poll interval */
 
 	/* sync PFP to ME, otherwise we might get invalid PFP reads */
 	radeon_ring_write(ring, PACKET3(PACKET3_PFP_SYNC_ME, 0));
diff --git a/drivers/gpu/drm/radeon/ni_dma.c b/drivers/gpu/drm/radeon/ni_dma.c
index f26f0a9fb522..ce787a9f12c0 100644
--- a/drivers/gpu/drm/radeon/ni_dma.c
+++ b/drivers/gpu/drm/radeon/ni_dma.c
@@ -123,6 +123,7 @@ void cayman_dma_ring_ib_execute(struct radeon_device *rdev,
 				struct radeon_ib *ib)
 {
 	struct radeon_ring *ring = &rdev->ring[ib->ring];
+	unsigned vm_id = ib->vm ? ib->vm->ids[ib->ring].id : 0;
 
 	if (rdev->wb.enabled) {
 		u32 next_rptr = ring->wptr + 4;
@@ -140,7 +141,7 @@ void cayman_dma_ring_ib_execute(struct radeon_device *rdev,
 	 */
 	while ((ring->wptr & 7) != 5)
 		radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0));
-	radeon_ring_write(ring, DMA_IB_PACKET(DMA_PACKET_INDIRECT_BUFFER, ib->vm ? ib->vm->id : 0, 0));
+	radeon_ring_write(ring, DMA_IB_PACKET(DMA_PACKET_INDIRECT_BUFFER, vm_id, 0));
 	radeon_ring_write(ring, (ib->gpu_addr & 0xFFFFFFE0));
 	radeon_ring_write(ring, (ib->length_dw << 12) | (upper_32_bits(ib->gpu_addr) & 0xFF));
 
@@ -371,7 +372,6 @@ void cayman_dma_vm_write_pages(struct radeon_device *rdev,
 		for (; ndw > 0; ndw -= 2, --count, pe += 8) {
 			if (flags & R600_PTE_SYSTEM) {
 				value = radeon_vm_map_gart(rdev, addr);
-				value &= 0xFFFFFFFFFFFFF000ULL;
 			} else if (flags & R600_PTE_VALID) {
 				value = addr;
 			} else {
@@ -446,16 +446,12 @@ void cayman_dma_vm_pad_ib(struct radeon_ib *ib)
 		ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0);
 }
 
-void cayman_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
+void cayman_dma_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring,
+			 unsigned vm_id, uint64_t pd_addr)
 {
-	struct radeon_ring *ring = &rdev->ring[ridx];
-
-	if (vm == NULL)
-		return;
-
 	radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0));
-	radeon_ring_write(ring, (0xf << 16) | ((VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (vm->id << 2)) >> 2));
-	radeon_ring_write(ring, vm->pd_gpu_addr >> 12);
+	radeon_ring_write(ring, (0xf << 16) | ((VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (vm_id << 2)) >> 2));
+	radeon_ring_write(ring, pd_addr >> 12);
 
 	/* flush hdp cache */
 	radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0));
@@ -465,6 +461,12 @@ void cayman_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm
 	/* bits 0-7 are the VM contexts0-7 */
 	radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0));
 	radeon_ring_write(ring, (0xf << 16) | (VM_INVALIDATE_REQUEST >> 2));
-	radeon_ring_write(ring, 1 << vm->id);
+	radeon_ring_write(ring, 1 << vm_id);
+
+	/* wait for invalidate to complete */
+	radeon_ring_write(ring, DMA_SRBM_READ_PACKET);
+	radeon_ring_write(ring, (0xff << 20) | (VM_INVALIDATE_REQUEST >> 2));
+	radeon_ring_write(ring, 0); /* mask */
+	radeon_ring_write(ring, 0); /* value */
 }
 
diff --git a/drivers/gpu/drm/radeon/ni_dpm.c b/drivers/gpu/drm/radeon/ni_dpm.c
index 01fc4888e6fe..6d2f16cf2c1c 100644
--- a/drivers/gpu/drm/radeon/ni_dpm.c
+++ b/drivers/gpu/drm/radeon/ni_dpm.c
@@ -23,6 +23,7 @@
 
 #include "drmP.h"
 #include "radeon.h"
+#include "radeon_asic.h"
 #include "nid.h"
 #include "r600_dpm.h"
 #include "ni_dpm.h"
@@ -789,7 +790,6 @@ static void ni_apply_state_adjust_rules(struct radeon_device *rdev,
 	bool disable_mclk_switching;
 	u32 mclk;
 	u16 vddci;
-	u32 max_sclk_vddc, max_mclk_vddci, max_mclk_vddc;
 	int i;
 
 	if ((rdev->pm.dpm.new_active_crtc_count > 1) ||
@@ -816,29 +816,6 @@ static void ni_apply_state_adjust_rules(struct radeon_device *rdev,
 		}
 	}
 
-	/* limit clocks to max supported clocks based on voltage dependency tables */
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_sclk,
-							&max_sclk_vddc);
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddci_dependency_on_mclk,
-							&max_mclk_vddci);
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_mclk,
-							&max_mclk_vddc);
-
-	for (i = 0; i < ps->performance_level_count; i++) {
-		if (max_sclk_vddc) {
-			if (ps->performance_levels[i].sclk > max_sclk_vddc)
-				ps->performance_levels[i].sclk = max_sclk_vddc;
-		}
-		if (max_mclk_vddci) {
-			if (ps->performance_levels[i].mclk > max_mclk_vddci)
-				ps->performance_levels[i].mclk = max_mclk_vddci;
-		}
-		if (max_mclk_vddc) {
-			if (ps->performance_levels[i].mclk > max_mclk_vddc)
-				ps->performance_levels[i].mclk = max_mclk_vddc;
-		}
-	}
-
 	/* XXX validate the min clocks required for display */
 
 	/* adjust low state */
diff --git a/drivers/gpu/drm/radeon/nid.h b/drivers/gpu/drm/radeon/nid.h
index 2e12e4d69253..ad7125486894 100644
--- a/drivers/gpu/drm/radeon/nid.h
+++ b/drivers/gpu/drm/radeon/nid.h
@@ -1133,6 +1133,23 @@
 #define	PACKET3_MEM_SEMAPHORE				0x39
 #define	PACKET3_MPEG_INDEX				0x3A
 #define	PACKET3_WAIT_REG_MEM				0x3C
+#define		WAIT_REG_MEM_FUNCTION(x)                ((x) << 0)
+                /* 0 - always
+		 * 1 - <
+		 * 2 - <=
+		 * 3 - ==
+		 * 4 - !=
+		 * 5 - >=
+		 * 6 - >
+		 */
+#define		WAIT_REG_MEM_MEM_SPACE(x)               ((x) << 4)
+                /* 0 - reg
+		 * 1 - mem
+		 */
+#define		WAIT_REG_MEM_ENGINE(x)                  ((x) << 8)
+                /* 0 - me
+		 * 1 - pfp
+		 */
 #define	PACKET3_MEM_WRITE				0x3D
 #define	PACKET3_PFP_SYNC_ME				0x42
 #define	PACKET3_SURFACE_SYNC				0x43
@@ -1272,6 +1289,13 @@
 					 (1 << 21) |			\
 					 (((n) & 0xFFFFF) << 0))
 
+#define DMA_SRBM_POLL_PACKET		((9 << 28) |			\
+					 (1 << 27) |			\
+					 (1 << 26))
+
+#define DMA_SRBM_READ_PACKET		((9 << 28) |			\
+					 (1 << 27))
+
 /* async DMA Packet types */
 #define	DMA_PACKET_WRITE				  0x2
 #define	DMA_PACKET_COPY					  0x3
diff --git a/drivers/gpu/drm/radeon/ppsmc.h b/drivers/gpu/drm/radeon/ppsmc.h
index 5670b8291285..7e5724a12f8b 100644
--- a/drivers/gpu/drm/radeon/ppsmc.h
+++ b/drivers/gpu/drm/radeon/ppsmc.h
@@ -56,6 +56,14 @@
 #define PPSMC_STATEFLAG_DEEPSLEEP_THROTTLE 0x20
 #define PPSMC_STATEFLAG_DEEPSLEEP_BYPASS   0x40
 
+#define FDO_MODE_HARDWARE 0
+#define FDO_MODE_PIECE_WISE_LINEAR 1
+
+enum FAN_CONTROL {
+	FAN_CONTROL_FUZZY,
+	FAN_CONTROL_TABLE
+};
+
 #define PPSMC_Result_OK             ((uint8_t)0x01)
 #define PPSMC_Result_Failed         ((uint8_t)0xFF)
 
@@ -79,6 +87,8 @@ typedef uint8_t PPSMC_Result;
 #define PPSMC_MSG_DisableCac                ((uint8_t)0x54)
 #define PPSMC_TDPClampingActive             ((uint8_t)0x59)
 #define PPSMC_TDPClampingInactive           ((uint8_t)0x5A)
+#define PPSMC_StartFanControl               ((uint8_t)0x5B)
+#define PPSMC_StopFanControl                ((uint8_t)0x5C)
 #define PPSMC_MSG_NoDisplay                 ((uint8_t)0x5D)
 #define PPSMC_MSG_HasDisplay                ((uint8_t)0x5E)
 #define PPSMC_MSG_UVDPowerOFF               ((uint8_t)0x60)
@@ -106,6 +116,7 @@ typedef uint8_t PPSMC_Result;
 #define PPSMC_MSG_SAMUDPM_SetEnabledMask      ((uint16_t) 0x130)
 #define PPSMC_MSG_MCLKDPM_ForceState          ((uint16_t) 0x131)
 #define PPSMC_MSG_MCLKDPM_NoForcedLevel       ((uint16_t) 0x132)
+#define PPSMC_MSG_Thermal_Cntl_Disable        ((uint16_t) 0x133)
 #define PPSMC_MSG_Voltage_Cntl_Disable        ((uint16_t) 0x135)
 #define PPSMC_MSG_PCIeDPM_Enable              ((uint16_t) 0x136)
 #define PPSMC_MSG_PCIeDPM_Disable             ((uint16_t) 0x13d)
@@ -149,6 +160,10 @@ typedef uint8_t PPSMC_Result;
 #define PPSMC_MSG_MASTER_DeepSleep_ON         ((uint16_t) 0x18F)
 #define PPSMC_MSG_MASTER_DeepSleep_OFF        ((uint16_t) 0x190)
 #define PPSMC_MSG_Remove_DC_Clamp             ((uint16_t) 0x191)
+#define PPSMC_MSG_SetFanPwmMax                ((uint16_t) 0x19A)
+
+#define PPSMC_MSG_ENABLE_THERMAL_DPM          ((uint16_t) 0x19C)
+#define PPSMC_MSG_DISABLE_THERMAL_DPM         ((uint16_t) 0x19D)
 
 #define PPSMC_MSG_API_GetSclkFrequency        ((uint16_t) 0x200)
 #define PPSMC_MSG_API_GetMclkFrequency        ((uint16_t) 0x201)
@@ -157,10 +172,11 @@ typedef uint8_t PPSMC_Result;
 #define PPSMC_MSG_DPM_Config                ((uint32_t) 0x102)
 #define PPSMC_MSG_DPM_ForceState            ((uint32_t) 0x104)
 #define PPSMC_MSG_PG_SIMD_Config            ((uint32_t) 0x108)
-#define PPSMC_MSG_DPM_N_LevelsDisabled      ((uint32_t) 0x112)
+#define PPSMC_MSG_Thermal_Cntl_Enable       ((uint32_t) 0x10a)
 #define PPSMC_MSG_Voltage_Cntl_Enable       ((uint32_t) 0x109)
 #define PPSMC_MSG_VCEPowerOFF               ((uint32_t) 0x10e)
 #define PPSMC_MSG_VCEPowerON                ((uint32_t) 0x10f)
+#define PPSMC_MSG_DPM_N_LevelsDisabled      ((uint32_t) 0x112)
 #define PPSMC_MSG_DCE_RemoveVoltageAdjustment   ((uint32_t) 0x11d)
 #define PPSMC_MSG_DCE_AllowVoltageAdjustment    ((uint32_t) 0x11e)
 #define PPSMC_MSG_EnableBAPM                ((uint32_t) 0x120)
diff --git a/drivers/gpu/drm/radeon/pptable.h b/drivers/gpu/drm/radeon/pptable.h
index 2d532996c697..4c2eec49dadc 100644
--- a/drivers/gpu/drm/radeon/pptable.h
+++ b/drivers/gpu/drm/radeon/pptable.h
@@ -96,6 +96,14 @@ typedef struct _ATOM_PPLIB_FANTABLE2
     USHORT  usTMax;                          // The max temperature
 } ATOM_PPLIB_FANTABLE2;
 
+typedef struct _ATOM_PPLIB_FANTABLE3
+{
+	ATOM_PPLIB_FANTABLE2 basicTable2;
+	UCHAR ucFanControlMode;
+	USHORT usFanPWMMax;
+	USHORT usFanOutputSensitivity;
+} ATOM_PPLIB_FANTABLE3;
+
 typedef struct _ATOM_PPLIB_EXTENDEDHEADER
 {
     USHORT  usSize;
diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c
index b0098e792e62..279801ca5110 100644
--- a/drivers/gpu/drm/radeon/r100.c
+++ b/drivers/gpu/drm/radeon/r100.c
@@ -644,6 +644,7 @@ int r100_pci_gart_init(struct radeon_device *rdev)
 		return r;
 	rdev->gart.table_size = rdev->gart.num_gpu_pages * 4;
 	rdev->asic->gart.tlb_flush = &r100_pci_gart_tlb_flush;
+	rdev->asic->gart.get_page_entry = &r100_pci_gart_get_page_entry;
 	rdev->asic->gart.set_page = &r100_pci_gart_set_page;
 	return radeon_gart_table_ram_alloc(rdev);
 }
@@ -681,11 +682,16 @@ void r100_pci_gart_disable(struct radeon_device *rdev)
 	WREG32(RADEON_AIC_HI_ADDR, 0);
 }
 
+uint64_t r100_pci_gart_get_page_entry(uint64_t addr, uint32_t flags)
+{
+	return addr;
+}
+
 void r100_pci_gart_set_page(struct radeon_device *rdev, unsigned i,
-			    uint64_t addr, uint32_t flags)
+			    uint64_t entry)
 {
 	u32 *gtt = rdev->gart.ptr;
-	gtt[i] = cpu_to_le32(lower_32_bits(addr));
+	gtt[i] = cpu_to_le32(lower_32_bits(entry));
 }
 
 void r100_pci_gart_fini(struct radeon_device *rdev)
@@ -869,13 +875,14 @@ bool r100_semaphore_ring_emit(struct radeon_device *rdev,
 	return false;
 }
 
-int r100_copy_blit(struct radeon_device *rdev,
-		   uint64_t src_offset,
-		   uint64_t dst_offset,
-		   unsigned num_gpu_pages,
-		   struct radeon_fence **fence)
+struct radeon_fence *r100_copy_blit(struct radeon_device *rdev,
+				    uint64_t src_offset,
+				    uint64_t dst_offset,
+				    unsigned num_gpu_pages,
+				    struct reservation_object *resv)
 {
 	struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
+	struct radeon_fence *fence;
 	uint32_t cur_pages;
 	uint32_t stride_bytes = RADEON_GPU_PAGE_SIZE;
 	uint32_t pitch;
@@ -896,7 +903,7 @@ int r100_copy_blit(struct radeon_device *rdev,
 	r = radeon_ring_lock(rdev, ring, ndw);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d) asking for %u dw.\n", r, ndw);
-		return -EINVAL;
+		return ERR_PTR(-EINVAL);
 	}
 	while (num_gpu_pages > 0) {
 		cur_pages = num_gpu_pages;
@@ -936,11 +943,13 @@ int r100_copy_blit(struct radeon_device *rdev,
 			  RADEON_WAIT_2D_IDLECLEAN |
 			  RADEON_WAIT_HOST_IDLECLEAN |
 			  RADEON_WAIT_DMA_GUI_IDLE);
-	if (fence) {
-		r = radeon_fence_emit(rdev, fence, RADEON_RING_TYPE_GFX_INDEX);
+	r = radeon_fence_emit(rdev, &fence, RADEON_RING_TYPE_GFX_INDEX);
+	if (r) {
+		radeon_ring_unlock_undo(rdev, ring);
+		return ERR_PTR(r);
 	}
 	radeon_ring_unlock_commit(rdev, ring, false);
-	return r;
+	return fence;
 }
 
 static int r100_cp_wait_for_idle(struct radeon_device *rdev)
@@ -1251,7 +1260,7 @@ int r100_reloc_pitch_offset(struct radeon_cs_parser *p,
 	int r;
 	u32 tile_flags = 0;
 	u32 tmp;
-	struct radeon_cs_reloc *reloc;
+	struct radeon_bo_list *reloc;
 	u32 value;
 
 	r = radeon_cs_packet_next_reloc(p, &reloc, 0);
@@ -1290,7 +1299,7 @@ int r100_packet3_load_vbpntr(struct radeon_cs_parser *p,
 			     int idx)
 {
 	unsigned c, i;
-	struct radeon_cs_reloc *reloc;
+	struct radeon_bo_list *reloc;
 	struct r100_cs_track *track;
 	int r = 0;
 	volatile uint32_t *ib;
@@ -1539,7 +1548,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p,
 			      struct radeon_cs_packet *pkt,
 			      unsigned idx, unsigned reg)
 {
-	struct radeon_cs_reloc *reloc;
+	struct radeon_bo_list *reloc;
 	struct r100_cs_track *track;
 	volatile uint32_t *ib;
 	uint32_t tmp;
@@ -1898,7 +1907,7 @@ int r100_cs_track_check_pkt3_indx_buffer(struct radeon_cs_parser *p,
 static int r100_packet3_check(struct radeon_cs_parser *p,
 			      struct radeon_cs_packet *pkt)
 {
-	struct radeon_cs_reloc *reloc;
+	struct radeon_bo_list *reloc;
 	struct r100_cs_track *track;
 	unsigned idx;
 	volatile uint32_t *ib;
@@ -2058,7 +2067,7 @@ int r100_cs_parse(struct radeon_cs_parser *p)
 		}
 		if (r)
 			return r;
-	} while (p->idx < p->chunks[p->chunk_ib_idx].length_dw);
+	} while (p->idx < p->chunk_ib->length_dw);
 	return 0;
 }
 
@@ -3204,6 +3213,9 @@ void r100_bandwidth_update(struct radeon_device *rdev)
 	uint32_t pixel_bytes1 = 0;
 	uint32_t pixel_bytes2 = 0;
 
+	if (!rdev->mode_info.mode_config_initialized)
+		return;
+
 	radeon_update_display_priority(rdev);
 
 	if (rdev->mode_info.crtcs[0]->base.enabled) {
diff --git a/drivers/gpu/drm/radeon/r200.c b/drivers/gpu/drm/radeon/r200.c
index 67780374a652..c70e6d5bcd19 100644
--- a/drivers/gpu/drm/radeon/r200.c
+++ b/drivers/gpu/drm/radeon/r200.c
@@ -80,13 +80,14 @@ static int r200_get_vtx_size_0(uint32_t vtx_fmt_0)
 	return vtx_size;
 }
 
-int r200_copy_dma(struct radeon_device *rdev,
-		  uint64_t src_offset,
-		  uint64_t dst_offset,
-		  unsigned num_gpu_pages,
-		  struct radeon_fence **fence)
+struct radeon_fence *r200_copy_dma(struct radeon_device *rdev,
+				   uint64_t src_offset,
+				   uint64_t dst_offset,
+				   unsigned num_gpu_pages,
+				   struct reservation_object *resv)
 {
 	struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
+	struct radeon_fence *fence;
 	uint32_t size;
 	uint32_t cur_size;
 	int i, num_loops;
@@ -98,7 +99,7 @@ int r200_copy_dma(struct radeon_device *rdev,
 	r = radeon_ring_lock(rdev, ring, num_loops * 4 + 64);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		return r;
+		return ERR_PTR(r);
 	}
 	/* Must wait for 2D idle & clean before DMA or hangs might happen */
 	radeon_ring_write(ring, PACKET0(RADEON_WAIT_UNTIL, 0));
@@ -118,11 +119,13 @@ int r200_copy_dma(struct radeon_device *rdev,
 	}
 	radeon_ring_write(ring, PACKET0(RADEON_WAIT_UNTIL, 0));
 	radeon_ring_write(ring, RADEON_WAIT_DMA_GUI_IDLE);
-	if (fence) {
-		r = radeon_fence_emit(rdev, fence, RADEON_RING_TYPE_GFX_INDEX);
+	r = radeon_fence_emit(rdev, &fence, RADEON_RING_TYPE_GFX_INDEX);
+	if (r) {
+		radeon_ring_unlock_undo(rdev, ring);
+		return ERR_PTR(r);
 	}
 	radeon_ring_unlock_commit(rdev, ring, false);
-	return r;
+	return fence;
 }
 
 
@@ -143,7 +146,7 @@ int r200_packet0_check(struct radeon_cs_parser *p,
 		       struct radeon_cs_packet *pkt,
 		       unsigned idx, unsigned reg)
 {
-	struct radeon_cs_reloc *reloc;
+	struct radeon_bo_list *reloc;
 	struct r100_cs_track *track;
 	volatile uint32_t *ib;
 	uint32_t tmp;
diff --git a/drivers/gpu/drm/radeon/r300.c b/drivers/gpu/drm/radeon/r300.c
index 1bc4704034ce..08d68f3e13e9 100644
--- a/drivers/gpu/drm/radeon/r300.c
+++ b/drivers/gpu/drm/radeon/r300.c
@@ -73,11 +73,8 @@ void rv370_pcie_gart_tlb_flush(struct radeon_device *rdev)
 #define R300_PTE_WRITEABLE (1 << 2)
 #define R300_PTE_READABLE  (1 << 3)
 
-void rv370_pcie_gart_set_page(struct radeon_device *rdev, unsigned i,
-			      uint64_t addr, uint32_t flags)
+uint64_t rv370_pcie_gart_get_page_entry(uint64_t addr, uint32_t flags)
 {
-	void __iomem *ptr = rdev->gart.ptr;
-
 	addr = (lower_32_bits(addr) >> 8) |
 		((upper_32_bits(addr) & 0xff) << 24);
 	if (flags & RADEON_GART_PAGE_READ)
@@ -86,10 +83,18 @@ void rv370_pcie_gart_set_page(struct radeon_device *rdev, unsigned i,
 		addr |= R300_PTE_WRITEABLE;
 	if (!(flags & RADEON_GART_PAGE_SNOOP))
 		addr |= R300_PTE_UNSNOOPED;
+	return addr;
+}
+
+void rv370_pcie_gart_set_page(struct radeon_device *rdev, unsigned i,
+			      uint64_t entry)
+{
+	void __iomem *ptr = rdev->gart.ptr;
+
 	/* on x86 we want this to be CPU endian, on powerpc
 	 * on powerpc without HW swappers, it'll get swapped on way
 	 * into VRAM - so no need for cpu_to_le32 on VRAM tables */
-	writel(addr, ((void __iomem *)ptr) + (i * 4));
+	writel(entry, ((void __iomem *)ptr) + (i * 4));
 }
 
 int rv370_pcie_gart_init(struct radeon_device *rdev)
@@ -109,6 +114,7 @@ int rv370_pcie_gart_init(struct radeon_device *rdev)
 		DRM_ERROR("Failed to register debugfs file for PCIE gart !\n");
 	rdev->gart.table_size = rdev->gart.num_gpu_pages * 4;
 	rdev->asic->gart.tlb_flush = &rv370_pcie_gart_tlb_flush;
+	rdev->asic->gart.get_page_entry = &rv370_pcie_gart_get_page_entry;
 	rdev->asic->gart.set_page = &rv370_pcie_gart_set_page;
 	return radeon_gart_table_vram_alloc(rdev);
 }
@@ -598,7 +604,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p,
 		struct radeon_cs_packet *pkt,
 		unsigned idx, unsigned reg)
 {
-	struct radeon_cs_reloc *reloc;
+	struct radeon_bo_list *reloc;
 	struct r100_cs_track *track;
 	volatile uint32_t *ib;
 	uint32_t tmp, tile_flags = 0;
@@ -1142,7 +1148,7 @@ fail:
 static int r300_packet3_check(struct radeon_cs_parser *p,
 			      struct radeon_cs_packet *pkt)
 {
-	struct radeon_cs_reloc *reloc;
+	struct radeon_bo_list *reloc;
 	struct r100_cs_track *track;
 	volatile uint32_t *ib;
 	unsigned idx;
@@ -1283,7 +1289,7 @@ int r300_cs_parse(struct radeon_cs_parser *p)
 		if (r) {
 			return r;
 		}
-	} while (p->idx < p->chunks[p->chunk_ib_idx].length_dw);
+	} while (p->idx < p->chunk_ib->length_dw);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/radeon/r300_cmdbuf.c b/drivers/gpu/drm/radeon/r300_cmdbuf.c
index 84b1d5367a11..9418e388b045 100644
--- a/drivers/gpu/drm/radeon/r300_cmdbuf.c
+++ b/drivers/gpu/drm/radeon/r300_cmdbuf.c
@@ -34,10 +34,10 @@
  */
 
 #include <drm/drmP.h>
-#include <drm/drm_buffer.h>
 #include <drm/radeon_drm.h>
 #include "radeon_drv.h"
 #include "r300_reg.h"
+#include "drm_buffer.h"
 
 #include <asm/unaligned.h>
 
diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c
index ea5c9af722ef..ef5d6066fa5b 100644
--- a/drivers/gpu/drm/radeon/r600.c
+++ b/drivers/gpu/drm/radeon/r600.c
@@ -122,6 +122,94 @@ u32 r600_get_xclk(struct radeon_device *rdev)
 
 int r600_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk)
 {
+	unsigned fb_div = 0, ref_div, vclk_div = 0, dclk_div = 0;
+	int r;
+
+	/* bypass vclk and dclk with bclk */
+	WREG32_P(CG_UPLL_FUNC_CNTL_2,
+		 VCLK_SRC_SEL(1) | DCLK_SRC_SEL(1),
+		 ~(VCLK_SRC_SEL_MASK | DCLK_SRC_SEL_MASK));
+
+	/* assert BYPASS_EN, deassert UPLL_RESET, UPLL_SLEEP and UPLL_CTLREQ */
+	WREG32_P(CG_UPLL_FUNC_CNTL, UPLL_BYPASS_EN_MASK, ~(
+		 UPLL_RESET_MASK | UPLL_SLEEP_MASK | UPLL_CTLREQ_MASK));
+
+	if (rdev->family >= CHIP_RS780)
+		WREG32_P(GFX_MACRO_BYPASS_CNTL, UPLL_BYPASS_CNTL,
+			 ~UPLL_BYPASS_CNTL);
+
+	if (!vclk || !dclk) {
+		/* keep the Bypass mode, put PLL to sleep */
+		WREG32_P(CG_UPLL_FUNC_CNTL, UPLL_SLEEP_MASK, ~UPLL_SLEEP_MASK);
+		return 0;
+	}
+
+	if (rdev->clock.spll.reference_freq == 10000)
+		ref_div = 34;
+	else
+		ref_div = 4;
+
+	r = radeon_uvd_calc_upll_dividers(rdev, vclk, dclk, 50000, 160000,
+					  ref_div + 1, 0xFFF, 2, 30, ~0,
+					  &fb_div, &vclk_div, &dclk_div);
+	if (r)
+		return r;
+
+	if (rdev->family >= CHIP_RV670 && rdev->family < CHIP_RS780)
+		fb_div >>= 1;
+	else
+		fb_div |= 1;
+
+	r = radeon_uvd_send_upll_ctlreq(rdev, CG_UPLL_FUNC_CNTL);
+        if (r)
+                return r;
+
+	/* assert PLL_RESET */
+	WREG32_P(CG_UPLL_FUNC_CNTL, UPLL_RESET_MASK, ~UPLL_RESET_MASK);
+
+	/* For RS780 we have to choose ref clk */
+	if (rdev->family >= CHIP_RS780)
+		WREG32_P(CG_UPLL_FUNC_CNTL, UPLL_REFCLK_SRC_SEL_MASK,
+			 ~UPLL_REFCLK_SRC_SEL_MASK);
+
+	/* set the required fb, ref and post divder values */
+	WREG32_P(CG_UPLL_FUNC_CNTL,
+		 UPLL_FB_DIV(fb_div) |
+		 UPLL_REF_DIV(ref_div),
+		 ~(UPLL_FB_DIV_MASK | UPLL_REF_DIV_MASK));
+	WREG32_P(CG_UPLL_FUNC_CNTL_2,
+		 UPLL_SW_HILEN(vclk_div >> 1) |
+		 UPLL_SW_LOLEN((vclk_div >> 1) + (vclk_div & 1)) |
+		 UPLL_SW_HILEN2(dclk_div >> 1) |
+		 UPLL_SW_LOLEN2((dclk_div >> 1) + (dclk_div & 1)) |
+		 UPLL_DIVEN_MASK | UPLL_DIVEN2_MASK,
+		 ~UPLL_SW_MASK);
+
+	/* give the PLL some time to settle */
+	mdelay(15);
+
+	/* deassert PLL_RESET */
+	WREG32_P(CG_UPLL_FUNC_CNTL, 0, ~UPLL_RESET_MASK);
+
+	mdelay(15);
+
+	/* deassert BYPASS EN */
+	WREG32_P(CG_UPLL_FUNC_CNTL, 0, ~UPLL_BYPASS_EN_MASK);
+
+	if (rdev->family >= CHIP_RS780)
+		WREG32_P(GFX_MACRO_BYPASS_CNTL, 0, ~UPLL_BYPASS_CNTL);
+
+	r = radeon_uvd_send_upll_ctlreq(rdev, CG_UPLL_FUNC_CNTL);
+	if (r)
+		return r;
+
+	/* switch VCLK and DCLK selection */
+	WREG32_P(CG_UPLL_FUNC_CNTL_2,
+		 VCLK_SRC_SEL(2) | DCLK_SRC_SEL(2),
+		 ~(VCLK_SRC_SEL_MASK | DCLK_SRC_SEL_MASK));
+
+	mdelay(100);
+
 	return 0;
 }
 
@@ -992,6 +1080,8 @@ static int r600_pcie_gart_enable(struct radeon_device *rdev)
 	WREG32(MC_VM_L1_TLB_MCB_WR_GFX_CNTL, tmp);
 	WREG32(MC_VM_L1_TLB_MCB_RD_PDMA_CNTL, tmp);
 	WREG32(MC_VM_L1_TLB_MCB_WR_PDMA_CNTL, tmp);
+	WREG32(MC_VM_L1_TLB_MCB_RD_UVD_CNTL, tmp);
+	WREG32(MC_VM_L1_TLB_MCB_WR_UVD_CNTL, tmp);
 	WREG32(MC_VM_L1_TLB_MCB_RD_SEM_CNTL, tmp | ENABLE_SEMAPHORE_MODE);
 	WREG32(MC_VM_L1_TLB_MCB_WR_SEM_CNTL, tmp | ENABLE_SEMAPHORE_MODE);
 	WREG32(VM_CONTEXT0_PAGE_TABLE_START_ADDR, rdev->mc.gtt_start >> 12);
@@ -1042,6 +1132,8 @@ static void r600_pcie_gart_disable(struct radeon_device *rdev)
 	WREG32(MC_VM_L1_TLB_MCB_WR_SYS_CNTL, tmp);
 	WREG32(MC_VM_L1_TLB_MCB_RD_HDP_CNTL, tmp);
 	WREG32(MC_VM_L1_TLB_MCB_WR_HDP_CNTL, tmp);
+	WREG32(MC_VM_L1_TLB_MCB_RD_UVD_CNTL, tmp);
+	WREG32(MC_VM_L1_TLB_MCB_WR_UVD_CNTL, tmp);
 	radeon_gart_table_vram_unpin(rdev);
 }
 
@@ -1338,7 +1430,7 @@ int r600_vram_scratch_init(struct radeon_device *rdev)
 	if (rdev->vram_scratch.robj == NULL) {
 		r = radeon_bo_create(rdev, RADEON_GPU_PAGE_SIZE,
 				     PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM,
-				     0, NULL, &rdev->vram_scratch.robj);
+				     0, NULL, NULL, &rdev->vram_scratch.robj);
 		if (r) {
 			return r;
 		}
@@ -2792,35 +2884,32 @@ bool r600_semaphore_ring_emit(struct radeon_device *rdev,
  * Used by the radeon ttm implementation to move pages if
  * registered as the asic copy callback.
  */
-int r600_copy_cpdma(struct radeon_device *rdev,
-		    uint64_t src_offset, uint64_t dst_offset,
-		    unsigned num_gpu_pages,
-		    struct radeon_fence **fence)
+struct radeon_fence *r600_copy_cpdma(struct radeon_device *rdev,
+				     uint64_t src_offset, uint64_t dst_offset,
+				     unsigned num_gpu_pages,
+				     struct reservation_object *resv)
 {
-	struct radeon_semaphore *sem = NULL;
+	struct radeon_fence *fence;
+	struct radeon_sync sync;
 	int ring_index = rdev->asic->copy.blit_ring_index;
 	struct radeon_ring *ring = &rdev->ring[ring_index];
 	u32 size_in_bytes, cur_size_in_bytes, tmp;
 	int i, num_loops;
 	int r = 0;
 
-	r = radeon_semaphore_create(rdev, &sem);
-	if (r) {
-		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		return r;
-	}
+	radeon_sync_create(&sync);
 
 	size_in_bytes = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT);
 	num_loops = DIV_ROUND_UP(size_in_bytes, 0x1fffff);
 	r = radeon_ring_lock(rdev, ring, num_loops * 6 + 24);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		radeon_sync_free(rdev, &sync, NULL);
+		return ERR_PTR(r);
 	}
 
-	radeon_semaphore_sync_to(sem, *fence);
-	radeon_semaphore_sync_rings(rdev, sem, ring->idx);
+	radeon_sync_resv(rdev, &sync, resv, false);
+	radeon_sync_rings(rdev, &sync, ring->idx);
 
 	radeon_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1));
 	radeon_ring_write(ring, (WAIT_UNTIL - PACKET3_SET_CONFIG_REG_OFFSET) >> 2);
@@ -2846,17 +2935,17 @@ int r600_copy_cpdma(struct radeon_device *rdev,
 	radeon_ring_write(ring, (WAIT_UNTIL - PACKET3_SET_CONFIG_REG_OFFSET) >> 2);
 	radeon_ring_write(ring, WAIT_CP_DMA_IDLE_bit);
 
-	r = radeon_fence_emit(rdev, fence, ring->idx);
+	r = radeon_fence_emit(rdev, &fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
-		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		radeon_sync_free(rdev, &sync, NULL);
+		return ERR_PTR(r);
 	}
 
 	radeon_ring_unlock_commit(rdev, ring, false);
-	radeon_semaphore_free(rdev, &sem, *fence);
+	radeon_sync_free(rdev, &sync, fence);
 
-	return r;
+	return fence;
 }
 
 int r600_set_surface_reg(struct radeon_device *rdev, int reg,
@@ -2907,6 +2996,18 @@ static int r600_startup(struct radeon_device *rdev)
 		return r;
 	}
 
+	if (rdev->has_uvd) {
+		r = uvd_v1_0_resume(rdev);
+		if (!r) {
+			r = radeon_fence_driver_start_ring(rdev, R600_RING_TYPE_UVD_INDEX);
+			if (r) {
+				dev_err(rdev->dev, "failed initializing UVD fences (%d).\n", r);
+			}
+		}
+		if (r)
+			rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size = 0;
+	}
+
 	/* Enable IRQ */
 	if (!rdev->irq.installed) {
 		r = radeon_irq_kms_init(rdev);
@@ -2935,6 +3036,18 @@ static int r600_startup(struct radeon_device *rdev)
 	if (r)
 		return r;
 
+	if (rdev->has_uvd) {
+		ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
+		if (ring->ring_size) {
+			r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
+					     RADEON_CP_PACKET2);
+			if (!r)
+				r = uvd_v1_0_init(rdev);
+			if (r)
+				DRM_ERROR("radeon: failed initializing UVD (%d).\n", r);
+		}
+	}
+
 	r = radeon_ib_pool_init(rdev);
 	if (r) {
 		dev_err(rdev->dev, "IB initialization failed (%d).\n", r);
@@ -2994,6 +3107,10 @@ int r600_suspend(struct radeon_device *rdev)
 	radeon_pm_suspend(rdev);
 	r600_audio_fini(rdev);
 	r600_cp_stop(rdev);
+	if (rdev->has_uvd) {
+		uvd_v1_0_fini(rdev);
+		radeon_uvd_suspend(rdev);
+	}
 	r600_irq_suspend(rdev);
 	radeon_wb_disable(rdev);
 	r600_pcie_gart_disable(rdev);
@@ -3073,6 +3190,14 @@ int r600_init(struct radeon_device *rdev)
 	rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ring_obj = NULL;
 	r600_ring_init(rdev, &rdev->ring[RADEON_RING_TYPE_GFX_INDEX], 1024 * 1024);
 
+	if (rdev->has_uvd) {
+		r = radeon_uvd_init(rdev);
+		if (!r) {
+			rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_obj = NULL;
+			r600_ring_init(rdev, &rdev->ring[R600_RING_TYPE_UVD_INDEX], 4096);
+		}
+	}
+
 	rdev->ih.ring_obj = NULL;
 	r600_ih_ring_init(rdev, 64 * 1024);
 
@@ -3102,6 +3227,10 @@ void r600_fini(struct radeon_device *rdev)
 	r600_audio_fini(rdev);
 	r600_cp_fini(rdev);
 	r600_irq_fini(rdev);
+	if (rdev->has_uvd) {
+		uvd_v1_0_fini(rdev);
+		radeon_uvd_fini(rdev);
+	}
 	radeon_wb_fini(rdev);
 	radeon_ib_pool_fini(rdev);
 	radeon_irq_kms_fini(rdev);
@@ -3235,7 +3364,7 @@ int r600_ih_ring_alloc(struct radeon_device *rdev)
 		r = radeon_bo_create(rdev, rdev->ih.ring_size,
 				     PAGE_SIZE, true,
 				     RADEON_GEM_DOMAIN_GTT, 0,
-				     NULL, &rdev->ih.ring_obj);
+				     NULL, NULL, &rdev->ih.ring_obj);
 		if (r) {
 			DRM_ERROR("radeon: failed to create ih ring buffer (%d).\n", r);
 			return r;
diff --git a/drivers/gpu/drm/radeon/r600_audio.c b/drivers/gpu/drm/radeon/r600_audio.c
deleted file mode 100644
index bffac10c4296..000000000000
--- a/drivers/gpu/drm/radeon/r600_audio.c
+++ /dev/null
@@ -1,207 +0,0 @@
-/*
- * Copyright 2008 Advanced Micro Devices, Inc.
- * Copyright 2008 Red Hat Inc.
- * Copyright 2009 Christian König.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Christian König
- */
-#include <drm/drmP.h>
-#include "radeon.h"
-#include "radeon_reg.h"
-#include "radeon_asic.h"
-#include "atom.h"
-
-/*
- * check if enc_priv stores radeon_encoder_atom_dig
- */
-static bool radeon_dig_encoder(struct drm_encoder *encoder)
-{
-	struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
-	switch (radeon_encoder->encoder_id) {
-	case ENCODER_OBJECT_ID_INTERNAL_LVDS:
-	case ENCODER_OBJECT_ID_INTERNAL_TMDS1:
-	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_TMDS1:
-	case ENCODER_OBJECT_ID_INTERNAL_LVTM1:
-	case ENCODER_OBJECT_ID_INTERNAL_DVO1:
-	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1:
-	case ENCODER_OBJECT_ID_INTERNAL_DDI:
-	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY:
-	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_LVTMA:
-	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1:
-	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2:
-		return true;
-	}
-	return false;
-}
-
-/*
- * check if the chipset is supported
- */
-static int r600_audio_chipset_supported(struct radeon_device *rdev)
-{
-	return ASIC_IS_DCE2(rdev) && !ASIC_IS_NODCE(rdev);
-}
-
-struct r600_audio_pin r600_audio_status(struct radeon_device *rdev)
-{
-	struct r600_audio_pin status;
-	uint32_t value;
-
-	value = RREG32(R600_AUDIO_RATE_BPS_CHANNEL);
-
-	/* number of channels */
-	status.channels = (value & 0x7) + 1;
-
-	/* bits per sample */
-	switch ((value & 0xF0) >> 4) {
-	case 0x0:
-		status.bits_per_sample = 8;
-		break;
-	case 0x1:
-		status.bits_per_sample = 16;
-		break;
-	case 0x2:
-		status.bits_per_sample = 20;
-		break;
-	case 0x3:
-		status.bits_per_sample = 24;
-		break;
-	case 0x4:
-		status.bits_per_sample = 32;
-		break;
-	default:
-		dev_err(rdev->dev, "Unknown bits per sample 0x%x, using 16\n",
-			(int)value);
-		status.bits_per_sample = 16;
-	}
-
-	/* current sampling rate in HZ */
-	if (value & 0x4000)
-		status.rate = 44100;
-	else
-		status.rate = 48000;
-	status.rate *= ((value >> 11) & 0x7) + 1;
-	status.rate /= ((value >> 8) & 0x7) + 1;
-
-	value = RREG32(R600_AUDIO_STATUS_BITS);
-
-	/* iec 60958 status bits */
-	status.status_bits = value & 0xff;
-
-	/* iec 60958 category code */
-	status.category_code = (value >> 8) & 0xff;
-
-	return status;
-}
-
-/*
- * update all hdmi interfaces with current audio parameters
- */
-void r600_audio_update_hdmi(struct work_struct *work)
-{
-	struct radeon_device *rdev = container_of(work, struct radeon_device,
-						  audio_work);
-	struct drm_device *dev = rdev->ddev;
-	struct r600_audio_pin audio_status = r600_audio_status(rdev);
-	struct drm_encoder *encoder;
-	bool changed = false;
-
-	if (rdev->audio.pin[0].channels != audio_status.channels ||
-	    rdev->audio.pin[0].rate != audio_status.rate ||
-	    rdev->audio.pin[0].bits_per_sample != audio_status.bits_per_sample ||
-	    rdev->audio.pin[0].status_bits != audio_status.status_bits ||
-	    rdev->audio.pin[0].category_code != audio_status.category_code) {
-		rdev->audio.pin[0] = audio_status;
-		changed = true;
-	}
-
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
-		if (!radeon_dig_encoder(encoder))
-			continue;
-		if (changed || r600_hdmi_buffer_status_changed(encoder))
-			r600_hdmi_update_audio_settings(encoder);
-	}
-}
-
-/* enable the audio stream */
-void r600_audio_enable(struct radeon_device *rdev,
-		       struct r600_audio_pin *pin,
-		       bool enable)
-{
-	u32 value = 0;
-
-	if (!pin)
-		return;
-
-	if (ASIC_IS_DCE4(rdev)) {
-		if (enable) {
-			value |= 0x81000000; /* Required to enable audio */
-			value |= 0x0e1000f0; /* fglrx sets that too */
-		}
-		WREG32(EVERGREEN_AUDIO_ENABLE, value);
-	} else {
-		WREG32_P(R600_AUDIO_ENABLE,
-			 enable ? 0x81000000 : 0x0, ~0x81000000);
-	}
-}
-
-/*
- * initialize the audio vars
- */
-int r600_audio_init(struct radeon_device *rdev)
-{
-	if (!radeon_audio || !r600_audio_chipset_supported(rdev))
-		return 0;
-
-	rdev->audio.enabled = true;
-
-	rdev->audio.num_pins = 1;
-	rdev->audio.pin[0].channels = -1;
-	rdev->audio.pin[0].rate = -1;
-	rdev->audio.pin[0].bits_per_sample = -1;
-	rdev->audio.pin[0].status_bits = 0;
-	rdev->audio.pin[0].category_code = 0;
-	rdev->audio.pin[0].id = 0;
-	/* disable audio.  it will be set up later */
-	r600_audio_enable(rdev, &rdev->audio.pin[0], false);
-
-	return 0;
-}
-
-/*
- * release the audio timer
- * TODO: How to do this correctly on SMP systems?
- */
-void r600_audio_fini(struct radeon_device *rdev)
-{
-	if (!rdev->audio.enabled)
-		return;
-
-	r600_audio_enable(rdev, &rdev->audio.pin[0], false);
-
-	rdev->audio.enabled = false;
-}
-
-struct r600_audio_pin *r600_audio_get_pin(struct radeon_device *rdev)
-{
-	/* only one pin on 6xx-NI */
-	return &rdev->audio.pin[0];
-}
diff --git a/drivers/gpu/drm/radeon/r600_cp.c b/drivers/gpu/drm/radeon/r600_cp.c
index 8c9b7e26533c..09e3f39925fa 100644
--- a/drivers/gpu/drm/radeon/r600_cp.c
+++ b/drivers/gpu/drm/radeon/r600_cp.c
@@ -1949,15 +1949,15 @@ int r600_do_cleanup_cp(struct drm_device *dev)
 #if __OS_HAS_AGP
 	if (dev_priv->flags & RADEON_IS_AGP) {
 		if (dev_priv->cp_ring != NULL) {
-			drm_core_ioremapfree(dev_priv->cp_ring, dev);
+			drm_legacy_ioremapfree(dev_priv->cp_ring, dev);
 			dev_priv->cp_ring = NULL;
 		}
 		if (dev_priv->ring_rptr != NULL) {
-			drm_core_ioremapfree(dev_priv->ring_rptr, dev);
+			drm_legacy_ioremapfree(dev_priv->ring_rptr, dev);
 			dev_priv->ring_rptr = NULL;
 		}
 		if (dev->agp_buffer_map != NULL) {
-			drm_core_ioremapfree(dev->agp_buffer_map, dev);
+			drm_legacy_ioremapfree(dev->agp_buffer_map, dev);
 			dev->agp_buffer_map = NULL;
 		}
 	} else
@@ -1968,7 +1968,7 @@ int r600_do_cleanup_cp(struct drm_device *dev)
 			r600_page_table_cleanup(dev, &dev_priv->gart_info);
 
 		if (dev_priv->gart_info.gart_table_location == DRM_ATI_GART_FB) {
-			drm_core_ioremapfree(&dev_priv->gart_info.mapping, dev);
+			drm_legacy_ioremapfree(&dev_priv->gart_info.mapping, dev);
 			dev_priv->gart_info.addr = NULL;
 		}
 	}
@@ -2052,27 +2052,27 @@ int r600_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init,
 	dev_priv->buffers_offset = init->buffers_offset;
 	dev_priv->gart_textures_offset = init->gart_textures_offset;
 
-	master_priv->sarea = drm_getsarea(dev);
+	master_priv->sarea = drm_legacy_getsarea(dev);
 	if (!master_priv->sarea) {
 		DRM_ERROR("could not find sarea!\n");
 		r600_do_cleanup_cp(dev);
 		return -EINVAL;
 	}
 
-	dev_priv->cp_ring = drm_core_findmap(dev, init->ring_offset);
+	dev_priv->cp_ring = drm_legacy_findmap(dev, init->ring_offset);
 	if (!dev_priv->cp_ring) {
 		DRM_ERROR("could not find cp ring region!\n");
 		r600_do_cleanup_cp(dev);
 		return -EINVAL;
 	}
-	dev_priv->ring_rptr = drm_core_findmap(dev, init->ring_rptr_offset);
+	dev_priv->ring_rptr = drm_legacy_findmap(dev, init->ring_rptr_offset);
 	if (!dev_priv->ring_rptr) {
 		DRM_ERROR("could not find ring read pointer!\n");
 		r600_do_cleanup_cp(dev);
 		return -EINVAL;
 	}
 	dev->agp_buffer_token = init->buffers_offset;
-	dev->agp_buffer_map = drm_core_findmap(dev, init->buffers_offset);
+	dev->agp_buffer_map = drm_legacy_findmap(dev, init->buffers_offset);
 	if (!dev->agp_buffer_map) {
 		DRM_ERROR("could not find dma buffer region!\n");
 		r600_do_cleanup_cp(dev);
@@ -2081,7 +2081,7 @@ int r600_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init,
 
 	if (init->gart_textures_offset) {
 		dev_priv->gart_textures =
-		    drm_core_findmap(dev, init->gart_textures_offset);
+		    drm_legacy_findmap(dev, init->gart_textures_offset);
 		if (!dev_priv->gart_textures) {
 			DRM_ERROR("could not find GART texture region!\n");
 			r600_do_cleanup_cp(dev);
@@ -2092,9 +2092,9 @@ int r600_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init,
 #if __OS_HAS_AGP
 	/* XXX */
 	if (dev_priv->flags & RADEON_IS_AGP) {
-		drm_core_ioremap_wc(dev_priv->cp_ring, dev);
-		drm_core_ioremap_wc(dev_priv->ring_rptr, dev);
-		drm_core_ioremap_wc(dev->agp_buffer_map, dev);
+		drm_legacy_ioremap_wc(dev_priv->cp_ring, dev);
+		drm_legacy_ioremap_wc(dev_priv->ring_rptr, dev);
+		drm_legacy_ioremap_wc(dev->agp_buffer_map, dev);
 		if (!dev_priv->cp_ring->handle ||
 		    !dev_priv->ring_rptr->handle ||
 		    !dev->agp_buffer_map->handle) {
@@ -2235,7 +2235,7 @@ int r600_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init,
 		dev_priv->gart_info.mapping.size =
 			dev_priv->gart_info.table_size;
 
-		drm_core_ioremap_wc(&dev_priv->gart_info.mapping, dev);
+		drm_legacy_ioremap_wc(&dev_priv->gart_info.mapping, dev);
 		if (!dev_priv->gart_info.mapping.handle) {
 			DRM_ERROR("ioremap failed.\n");
 			r600_do_cleanup_cp(dev);
diff --git a/drivers/gpu/drm/radeon/r600_cs.c b/drivers/gpu/drm/radeon/r600_cs.c
index c47537a1ddba..acc1f99c84d9 100644
--- a/drivers/gpu/drm/radeon/r600_cs.c
+++ b/drivers/gpu/drm/radeon/r600_cs.c
@@ -969,7 +969,7 @@ static int r600_cs_parse_packet0(struct radeon_cs_parser *p,
 static int r600_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 {
 	struct r600_cs_track *track = (struct r600_cs_track *)p->track;
-	struct radeon_cs_reloc *reloc;
+	struct radeon_bo_list *reloc;
 	u32 m, i, tmp, *ib;
 	int r;
 
@@ -1626,7 +1626,7 @@ static bool r600_is_safe_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 static int r600_packet3_check(struct radeon_cs_parser *p,
 				struct radeon_cs_packet *pkt)
 {
-	struct radeon_cs_reloc *reloc;
+	struct radeon_bo_list *reloc;
 	struct r600_cs_track *track;
 	volatile u32 *ib;
 	unsigned idx;
@@ -2316,7 +2316,7 @@ int r600_cs_parse(struct radeon_cs_parser *p)
 			p->track = NULL;
 			return r;
 		}
-	} while (p->idx < p->chunks[p->chunk_ib_idx].length_dw);
+	} while (p->idx < p->chunk_ib->length_dw);
 #if 0
 	for (r = 0; r < p->ib.length_dw; r++) {
 		printk(KERN_INFO "%05d  0x%08X\n", r, p->ib.ptr[r]);
@@ -2351,10 +2351,10 @@ static void r600_cs_parser_fini(struct radeon_cs_parser *parser, int error)
 
 static int r600_cs_parser_relocs_legacy(struct radeon_cs_parser *p)
 {
-	if (p->chunk_relocs_idx == -1) {
+	if (p->chunk_relocs == NULL) {
 		return 0;
 	}
-	p->relocs = kzalloc(sizeof(struct radeon_cs_reloc), GFP_KERNEL);
+	p->relocs = kzalloc(sizeof(struct radeon_bo_list), GFP_KERNEL);
 	if (p->relocs == NULL) {
 		return -ENOMEM;
 	}
@@ -2398,7 +2398,7 @@ int r600_cs_legacy(struct drm_device *dev, void *data, struct drm_file *filp,
 	/* Copy the packet into the IB, the parser will read from the
 	 * input memory (cached) and write to the IB (which can be
 	 * uncached). */
-	ib_chunk = &parser.chunks[parser.chunk_ib_idx];
+	ib_chunk = parser.chunk_ib;
 	parser.ib.length_dw = ib_chunk->length_dw;
 	*l = parser.ib.length_dw;
 	if (copy_from_user(ib, ib_chunk->user_ptr, ib_chunk->length_dw * 4)) {
@@ -2435,24 +2435,24 @@ void r600_cs_legacy_init(void)
  * GPU offset using the provided start.
  **/
 int r600_dma_cs_next_reloc(struct radeon_cs_parser *p,
-			   struct radeon_cs_reloc **cs_reloc)
+			   struct radeon_bo_list **cs_reloc)
 {
 	struct radeon_cs_chunk *relocs_chunk;
 	unsigned idx;
 
 	*cs_reloc = NULL;
-	if (p->chunk_relocs_idx == -1) {
+	if (p->chunk_relocs == NULL) {
 		DRM_ERROR("No relocation chunk !\n");
 		return -EINVAL;
 	}
-	relocs_chunk = &p->chunks[p->chunk_relocs_idx];
+	relocs_chunk = p->chunk_relocs;
 	idx = p->dma_reloc_idx;
 	if (idx >= p->nrelocs) {
 		DRM_ERROR("Relocs at %d after relocations chunk end %d !\n",
 			  idx, p->nrelocs);
 		return -EINVAL;
 	}
-	*cs_reloc = p->relocs_ptr[idx];
+	*cs_reloc = &p->relocs[idx];
 	p->dma_reloc_idx++;
 	return 0;
 }
@@ -2472,8 +2472,8 @@ int r600_dma_cs_next_reloc(struct radeon_cs_parser *p,
  **/
 int r600_dma_cs_parse(struct radeon_cs_parser *p)
 {
-	struct radeon_cs_chunk *ib_chunk = &p->chunks[p->chunk_ib_idx];
-	struct radeon_cs_reloc *src_reloc, *dst_reloc;
+	struct radeon_cs_chunk *ib_chunk = p->chunk_ib;
+	struct radeon_bo_list *src_reloc, *dst_reloc;
 	u32 header, cmd, count, tiled;
 	volatile u32 *ib = p->ib.ptr;
 	u32 idx, idx_value;
@@ -2619,7 +2619,7 @@ int r600_dma_cs_parse(struct radeon_cs_parser *p)
 			DRM_ERROR("Unknown packet type %d at %d !\n", cmd, idx);
 			return -EINVAL;
 		}
-	} while (p->idx < p->chunks[p->chunk_ib_idx].length_dw);
+	} while (p->idx < p->chunk_ib->length_dw);
 #if 0
 	for (r = 0; r < p->ib->length_dw; r++) {
 		printk(KERN_INFO "%05d  0x%08X\n", r, p->ib.ptr[r]);
diff --git a/drivers/gpu/drm/radeon/r600_dma.c b/drivers/gpu/drm/radeon/r600_dma.c
index a908daa006d2..d2dd29ab24fa 100644
--- a/drivers/gpu/drm/radeon/r600_dma.c
+++ b/drivers/gpu/drm/radeon/r600_dma.c
@@ -232,16 +232,19 @@ int r600_dma_ring_test(struct radeon_device *rdev,
 {
 	unsigned i;
 	int r;
-	void __iomem *ptr = (void *)rdev->vram_scratch.ptr;
+	unsigned index;
 	u32 tmp;
+	u64 gpu_addr;
 
-	if (!ptr) {
-		DRM_ERROR("invalid vram scratch pointer\n");
-		return -EINVAL;
-	}
+	if (ring->idx == R600_RING_TYPE_DMA_INDEX)
+		index = R600_WB_DMA_RING_TEST_OFFSET;
+	else
+		index = CAYMAN_WB_DMA1_RING_TEST_OFFSET;
+
+	gpu_addr = rdev->wb.gpu_addr + index;
 
 	tmp = 0xCAFEDEAD;
-	writel(tmp, ptr);
+	rdev->wb.wb[index/4] = cpu_to_le32(tmp);
 
 	r = radeon_ring_lock(rdev, ring, 4);
 	if (r) {
@@ -249,13 +252,13 @@ int r600_dma_ring_test(struct radeon_device *rdev,
 		return r;
 	}
 	radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_WRITE, 0, 0, 1));
-	radeon_ring_write(ring, rdev->vram_scratch.gpu_addr & 0xfffffffc);
-	radeon_ring_write(ring, upper_32_bits(rdev->vram_scratch.gpu_addr) & 0xff);
+	radeon_ring_write(ring, lower_32_bits(gpu_addr));
+	radeon_ring_write(ring, upper_32_bits(gpu_addr) & 0xff);
 	radeon_ring_write(ring, 0xDEADBEEF);
 	radeon_ring_unlock_commit(rdev, ring, false);
 
 	for (i = 0; i < rdev->usec_timeout; i++) {
-		tmp = readl(ptr);
+		tmp = le32_to_cpu(rdev->wb.wb[index/4]);
 		if (tmp == 0xDEADBEEF)
 			break;
 		DRM_UDELAY(1);
@@ -335,17 +338,17 @@ int r600_dma_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
 {
 	struct radeon_ib ib;
 	unsigned i;
+	unsigned index;
 	int r;
-	void __iomem *ptr = (void *)rdev->vram_scratch.ptr;
 	u32 tmp = 0;
+	u64 gpu_addr;
 
-	if (!ptr) {
-		DRM_ERROR("invalid vram scratch pointer\n");
-		return -EINVAL;
-	}
+	if (ring->idx == R600_RING_TYPE_DMA_INDEX)
+		index = R600_WB_DMA_RING_TEST_OFFSET;
+	else
+		index = CAYMAN_WB_DMA1_RING_TEST_OFFSET;
 
-	tmp = 0xCAFEDEAD;
-	writel(tmp, ptr);
+	gpu_addr = rdev->wb.gpu_addr + index;
 
 	r = radeon_ib_get(rdev, ring->idx, &ib, NULL, 256);
 	if (r) {
@@ -354,8 +357,8 @@ int r600_dma_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
 	}
 
 	ib.ptr[0] = DMA_PACKET(DMA_PACKET_WRITE, 0, 0, 1);
-	ib.ptr[1] = rdev->vram_scratch.gpu_addr & 0xfffffffc;
-	ib.ptr[2] = upper_32_bits(rdev->vram_scratch.gpu_addr) & 0xff;
+	ib.ptr[1] = lower_32_bits(gpu_addr);
+	ib.ptr[2] = upper_32_bits(gpu_addr) & 0xff;
 	ib.ptr[3] = 0xDEADBEEF;
 	ib.length_dw = 4;
 
@@ -371,7 +374,7 @@ int r600_dma_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
 		return r;
 	}
 	for (i = 0; i < rdev->usec_timeout; i++) {
-		tmp = readl(ptr);
+		tmp = le32_to_cpu(rdev->wb.wb[index/4]);
 		if (tmp == 0xDEADBEEF)
 			break;
 		DRM_UDELAY(1);
@@ -427,41 +430,38 @@ void r600_dma_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
  * @src_offset: src GPU address
  * @dst_offset: dst GPU address
  * @num_gpu_pages: number of GPU pages to xfer
- * @fence: radeon fence object
+ * @resv: reservation object to sync to
  *
  * Copy GPU paging using the DMA engine (r6xx).
  * Used by the radeon ttm implementation to move pages if
  * registered as the asic copy callback.
  */
-int r600_copy_dma(struct radeon_device *rdev,
-		  uint64_t src_offset, uint64_t dst_offset,
-		  unsigned num_gpu_pages,
-		  struct radeon_fence **fence)
+struct radeon_fence *r600_copy_dma(struct radeon_device *rdev,
+				   uint64_t src_offset, uint64_t dst_offset,
+				   unsigned num_gpu_pages,
+				   struct reservation_object *resv)
 {
-	struct radeon_semaphore *sem = NULL;
+	struct radeon_fence *fence;
+	struct radeon_sync sync;
 	int ring_index = rdev->asic->copy.dma_ring_index;
 	struct radeon_ring *ring = &rdev->ring[ring_index];
 	u32 size_in_dw, cur_size_in_dw;
 	int i, num_loops;
 	int r = 0;
 
-	r = radeon_semaphore_create(rdev, &sem);
-	if (r) {
-		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		return r;
-	}
+	radeon_sync_create(&sync);
 
 	size_in_dw = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT) / 4;
 	num_loops = DIV_ROUND_UP(size_in_dw, 0xFFFE);
 	r = radeon_ring_lock(rdev, ring, num_loops * 4 + 8);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		radeon_sync_free(rdev, &sync, NULL);
+		return ERR_PTR(r);
 	}
 
-	radeon_semaphore_sync_to(sem, *fence);
-	radeon_semaphore_sync_rings(rdev, sem, ring->idx);
+	radeon_sync_resv(rdev, &sync, resv, false);
+	radeon_sync_rings(rdev, &sync, ring->idx);
 
 	for (i = 0; i < num_loops; i++) {
 		cur_size_in_dw = size_in_dw;
@@ -477,15 +477,15 @@ int r600_copy_dma(struct radeon_device *rdev,
 		dst_offset += cur_size_in_dw * 4;
 	}
 
-	r = radeon_fence_emit(rdev, fence, ring->idx);
+	r = radeon_fence_emit(rdev, &fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
-		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		radeon_sync_free(rdev, &sync, NULL);
+		return ERR_PTR(r);
 	}
 
 	radeon_ring_unlock_commit(rdev, ring, false);
-	radeon_semaphore_free(rdev, &sem, *fence);
+	radeon_sync_free(rdev, &sync, fence);
 
-	return r;
+	return fence;
 }
diff --git a/drivers/gpu/drm/radeon/r600_dpm.c b/drivers/gpu/drm/radeon/r600_dpm.c
index 9c61b74ef441..843b65f46ece 100644
--- a/drivers/gpu/drm/radeon/r600_dpm.c
+++ b/drivers/gpu/drm/radeon/r600_dpm.c
@@ -24,6 +24,7 @@
 
 #include "drmP.h"
 #include "radeon.h"
+#include "radeon_asic.h"
 #include "r600d.h"
 #include "r600_dpm.h"
 #include "atom.h"
@@ -810,6 +811,7 @@ union power_info {
 union fan_info {
 	struct _ATOM_PPLIB_FANTABLE fan;
 	struct _ATOM_PPLIB_FANTABLE2 fan2;
+	struct _ATOM_PPLIB_FANTABLE3 fan3;
 };
 
 static int r600_parse_clk_voltage_dep_table(struct radeon_clock_voltage_dependency_table *radeon_table,
@@ -899,6 +901,14 @@ int r600_parse_extended_power_table(struct radeon_device *rdev)
 			else
 				rdev->pm.dpm.fan.t_max = 10900;
 			rdev->pm.dpm.fan.cycle_delay = 100000;
+			if (fan_info->fan.ucFanTableFormat >= 3) {
+				rdev->pm.dpm.fan.control_mode = fan_info->fan3.ucFanControlMode;
+				rdev->pm.dpm.fan.default_max_fan_pwm =
+					le16_to_cpu(fan_info->fan3.usFanPWMMax);
+				rdev->pm.dpm.fan.default_fan_output_sensitivity = 4836;
+				rdev->pm.dpm.fan.fan_output_sensitivity =
+					le16_to_cpu(fan_info->fan3.usFanOutputSensitivity);
+			}
 			rdev->pm.dpm.fan.ucode_fan_control = true;
 		}
 	}
@@ -1255,7 +1265,7 @@ int r600_parse_extended_power_table(struct radeon_device *rdev)
 					(mode_info->atom_context->bios + data_offset +
 					 le16_to_cpu(ext_hdr->usPowerTuneTableOffset));
 				rdev->pm.dpm.dyn_state.cac_tdp_table->maximum_power_delivery_limit =
-					ppt->usMaximumPowerDeliveryLimit;
+					le16_to_cpu(ppt->usMaximumPowerDeliveryLimit);
 				pt = &ppt->power_tune_table;
 			} else {
 				ATOM_PPLIB_POWERTUNE_Table *ppt = (ATOM_PPLIB_POWERTUNE_Table *)
diff --git a/drivers/gpu/drm/radeon/r600_dpm.h b/drivers/gpu/drm/radeon/r600_dpm.h
index 46b9d2a03018..bd499d749bc9 100644
--- a/drivers/gpu/drm/radeon/r600_dpm.h
+++ b/drivers/gpu/drm/radeon/r600_dpm.h
@@ -96,6 +96,9 @@
 #define R600_TEMP_RANGE_MIN (90 * 1000)
 #define R600_TEMP_RANGE_MAX (120 * 1000)
 
+#define FDO_PWM_MODE_STATIC  1
+#define FDO_PWM_MODE_STATIC_RPM 5
+
 enum r600_power_level {
 	R600_POWER_LEVEL_LOW = 0,
 	R600_POWER_LEVEL_MEDIUM = 1,
diff --git a/drivers/gpu/drm/radeon/r600_hdmi.c b/drivers/gpu/drm/radeon/r600_hdmi.c
index 26ef8ced6f89..b90dc0eb08e6 100644
--- a/drivers/gpu/drm/radeon/r600_hdmi.c
+++ b/drivers/gpu/drm/radeon/r600_hdmi.c
@@ -72,6 +72,169 @@ static const struct radeon_hdmi_acr r600_hdmi_predefined_acr[] = {
 
 
 /*
+ * check if the chipset is supported
+ */
+static int r600_audio_chipset_supported(struct radeon_device *rdev)
+{
+	return ASIC_IS_DCE2(rdev) && !ASIC_IS_NODCE(rdev);
+}
+
+static struct r600_audio_pin r600_audio_status(struct radeon_device *rdev)
+{
+	struct r600_audio_pin status;
+	uint32_t value;
+
+	value = RREG32(R600_AUDIO_RATE_BPS_CHANNEL);
+
+	/* number of channels */
+	status.channels = (value & 0x7) + 1;
+
+	/* bits per sample */
+	switch ((value & 0xF0) >> 4) {
+	case 0x0:
+		status.bits_per_sample = 8;
+		break;
+	case 0x1:
+		status.bits_per_sample = 16;
+		break;
+	case 0x2:
+		status.bits_per_sample = 20;
+		break;
+	case 0x3:
+		status.bits_per_sample = 24;
+		break;
+	case 0x4:
+		status.bits_per_sample = 32;
+		break;
+	default:
+		dev_err(rdev->dev, "Unknown bits per sample 0x%x, using 16\n",
+			(int)value);
+		status.bits_per_sample = 16;
+	}
+
+	/* current sampling rate in HZ */
+	if (value & 0x4000)
+		status.rate = 44100;
+	else
+		status.rate = 48000;
+	status.rate *= ((value >> 11) & 0x7) + 1;
+	status.rate /= ((value >> 8) & 0x7) + 1;
+
+	value = RREG32(R600_AUDIO_STATUS_BITS);
+
+	/* iec 60958 status bits */
+	status.status_bits = value & 0xff;
+
+	/* iec 60958 category code */
+	status.category_code = (value >> 8) & 0xff;
+
+	return status;
+}
+
+/*
+ * update all hdmi interfaces with current audio parameters
+ */
+void r600_audio_update_hdmi(struct work_struct *work)
+{
+	struct radeon_device *rdev = container_of(work, struct radeon_device,
+						  audio_work);
+	struct drm_device *dev = rdev->ddev;
+	struct r600_audio_pin audio_status = r600_audio_status(rdev);
+	struct drm_encoder *encoder;
+	bool changed = false;
+
+	if (rdev->audio.pin[0].channels != audio_status.channels ||
+	    rdev->audio.pin[0].rate != audio_status.rate ||
+	    rdev->audio.pin[0].bits_per_sample != audio_status.bits_per_sample ||
+	    rdev->audio.pin[0].status_bits != audio_status.status_bits ||
+	    rdev->audio.pin[0].category_code != audio_status.category_code) {
+		rdev->audio.pin[0] = audio_status;
+		changed = true;
+	}
+
+	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
+		if (!radeon_encoder_is_digital(encoder))
+			continue;
+		if (changed || r600_hdmi_buffer_status_changed(encoder))
+			r600_hdmi_update_audio_settings(encoder);
+	}
+}
+
+/* enable the audio stream */
+void r600_audio_enable(struct radeon_device *rdev,
+		       struct r600_audio_pin *pin,
+		       u8 enable_mask)
+{
+	u32 tmp = RREG32(AZ_HOT_PLUG_CONTROL);
+
+	if (!pin)
+		return;
+
+	if (enable_mask) {
+		tmp |= AUDIO_ENABLED;
+		if (enable_mask & 1)
+			tmp |= PIN0_AUDIO_ENABLED;
+		if (enable_mask & 2)
+			tmp |= PIN1_AUDIO_ENABLED;
+		if (enable_mask & 4)
+			tmp |= PIN2_AUDIO_ENABLED;
+		if (enable_mask & 8)
+			tmp |= PIN3_AUDIO_ENABLED;
+	} else {
+		tmp &= ~(AUDIO_ENABLED |
+			 PIN0_AUDIO_ENABLED |
+			 PIN1_AUDIO_ENABLED |
+			 PIN2_AUDIO_ENABLED |
+			 PIN3_AUDIO_ENABLED);
+	}
+
+	WREG32(AZ_HOT_PLUG_CONTROL, tmp);
+}
+
+/*
+ * initialize the audio vars
+ */
+int r600_audio_init(struct radeon_device *rdev)
+{
+	if (!radeon_audio || !r600_audio_chipset_supported(rdev))
+		return 0;
+
+	rdev->audio.enabled = true;
+
+	rdev->audio.num_pins = 1;
+	rdev->audio.pin[0].channels = -1;
+	rdev->audio.pin[0].rate = -1;
+	rdev->audio.pin[0].bits_per_sample = -1;
+	rdev->audio.pin[0].status_bits = 0;
+	rdev->audio.pin[0].category_code = 0;
+	rdev->audio.pin[0].id = 0;
+	/* disable audio.  it will be set up later */
+	r600_audio_enable(rdev, &rdev->audio.pin[0], 0);
+
+	return 0;
+}
+
+/*
+ * release the audio timer
+ * TODO: How to do this correctly on SMP systems?
+ */
+void r600_audio_fini(struct radeon_device *rdev)
+{
+	if (!rdev->audio.enabled)
+		return;
+
+	r600_audio_enable(rdev, &rdev->audio.pin[0], 0);
+
+	rdev->audio.enabled = false;
+}
+
+struct r600_audio_pin *r600_audio_get_pin(struct radeon_device *rdev)
+{
+	/* only one pin on 6xx-NI */
+	return &rdev->audio.pin[0];
+}
+
+/*
  * calculate CTS and N values if they are not found in the table
  */
 static void r600_hdmi_calc_cts(uint32_t clock, int *CTS, int *N, int freq)
@@ -357,7 +520,7 @@ void r600_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mod
 
 	/* disable audio prior to setting up hw */
 	dig->afmt->pin = r600_audio_get_pin(rdev);
-	r600_audio_enable(rdev, dig->afmt->pin, false);
+	r600_audio_enable(rdev, dig->afmt->pin, 0xf);
 
 	r600_audio_set_dto(encoder, mode->clock);
 
@@ -443,7 +606,7 @@ void r600_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mod
 	WREG32(HDMI0_RAMP_CONTROL3 + offset, 0x00000001);
 
 	/* enable audio after to setting up hw */
-	r600_audio_enable(rdev, dig->afmt->pin, true);
+	r600_audio_enable(rdev, dig->afmt->pin, 0xf);
 }
 
 /**
@@ -528,6 +691,11 @@ void r600_hdmi_enable(struct drm_encoder *encoder, bool enable)
 	if (!enable && !dig->afmt->enabled)
 		return;
 
+	if (!enable && dig->afmt->pin) {
+		r600_audio_enable(rdev, dig->afmt->pin, 0);
+		dig->afmt->pin = NULL;
+	}
+
 	/* Older chipsets require setting HDMI and routing manually */
 	if (!ASIC_IS_DCE3(rdev)) {
 		if (enable)
diff --git a/drivers/gpu/drm/radeon/r600d.h b/drivers/gpu/drm/radeon/r600d.h
index 31e1052ad3e3..1e8495cca41e 100644
--- a/drivers/gpu/drm/radeon/r600d.h
+++ b/drivers/gpu/drm/radeon/r600d.h
@@ -323,11 +323,12 @@
 #define	HDP_TILING_CONFIG				0x2F3C
 #define HDP_DEBUG1                                      0x2F34
 
+#define MC_CONFIG					0x2000
 #define MC_VM_AGP_TOP					0x2184
 #define MC_VM_AGP_BOT					0x2188
 #define	MC_VM_AGP_BASE					0x218C
 #define MC_VM_FB_LOCATION				0x2180
-#define MC_VM_L1_TLB_MCD_RD_A_CNTL			0x219C
+#define MC_VM_L1_TLB_MCB_RD_UVD_CNTL			0x2124
 #define 	ENABLE_L1_TLB					(1 << 0)
 #define		ENABLE_L1_FRAGMENT_PROCESSING			(1 << 1)
 #define		ENABLE_L1_STRICT_ORDERING			(1 << 2)
@@ -347,12 +348,14 @@
 #define		EFFECTIVE_L1_QUEUE_SIZE(x)			(((x) & 7) << 15)
 #define		EFFECTIVE_L1_QUEUE_SIZE_MASK			0x00038000
 #define		EFFECTIVE_L1_QUEUE_SIZE_SHIFT			15
+#define MC_VM_L1_TLB_MCD_RD_A_CNTL			0x219C
 #define MC_VM_L1_TLB_MCD_RD_B_CNTL			0x21A0
 #define MC_VM_L1_TLB_MCB_RD_GFX_CNTL			0x21FC
 #define MC_VM_L1_TLB_MCB_RD_HDP_CNTL			0x2204
 #define MC_VM_L1_TLB_MCB_RD_PDMA_CNTL			0x2208
 #define MC_VM_L1_TLB_MCB_RD_SEM_CNTL			0x220C
 #define	MC_VM_L1_TLB_MCB_RD_SYS_CNTL			0x2200
+#define MC_VM_L1_TLB_MCB_WR_UVD_CNTL			0x212c
 #define MC_VM_L1_TLB_MCD_WR_A_CNTL			0x21A4
 #define MC_VM_L1_TLB_MCD_WR_B_CNTL			0x21A8
 #define MC_VM_L1_TLB_MCB_WR_GFX_CNTL			0x2210
@@ -366,6 +369,8 @@
 #define MC_VM_SYSTEM_APERTURE_HIGH_ADDR			0x2194
 #define MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR		0x2198
 
+#define RS_DQ_RD_RET_CONF				0x2348
+
 #define	PA_CL_ENHANCE					0x8A14
 #define		CLIP_VTX_REORDER_ENA				(1 << 0)
 #define		NUM_CLIP_SEQ(x)					((x) << 1)
@@ -922,6 +927,23 @@
 #       define TARGET_LINK_SPEED_MASK                     (0xf << 0)
 #       define SELECTABLE_DEEMPHASIS                      (1 << 6)
 
+/* Audio */
+#define AZ_HOT_PLUG_CONTROL               0x7300
+#       define AZ_FORCE_CODEC_WAKE        (1 << 0)
+#       define JACK_DETECTION_ENABLE      (1 << 4)
+#       define UNSOLICITED_RESPONSE_ENABLE (1 << 8)
+#       define CODEC_HOT_PLUG_ENABLE      (1 << 12)
+#       define AUDIO_ENABLED              (1 << 31)
+/* DCE3 adds */
+#       define PIN0_JACK_DETECTION_ENABLE (1 << 4)
+#       define PIN1_JACK_DETECTION_ENABLE (1 << 5)
+#       define PIN2_JACK_DETECTION_ENABLE (1 << 6)
+#       define PIN3_JACK_DETECTION_ENABLE (1 << 7)
+#       define PIN0_AUDIO_ENABLED         (1 << 24)
+#       define PIN1_AUDIO_ENABLED         (1 << 25)
+#       define PIN2_AUDIO_ENABLED         (1 << 26)
+#       define PIN3_AUDIO_ENABLED         (1 << 27)
+
 /* Audio clocks DCE 2.0/3.0 */
 #define AUDIO_DTO                         0x7340
 #       define AUDIO_DTO_PHASE(x)         (((x) & 0xffff) << 0)
@@ -1476,6 +1498,7 @@
 #define UVD_CGC_GATE					0xf4a8
 #define UVD_LMI_CTRL2					0xf4f4
 #define UVD_MASTINT_EN					0xf500
+#define UVD_FW_START					0xf51C
 #define UVD_LMI_ADDR_EXT				0xf594
 #define UVD_LMI_CTRL					0xf598
 #define UVD_LMI_SWAP_CNTL				0xf5b4
@@ -1488,6 +1511,13 @@
 #define UVD_MPC_SET_MUX					0xf5f4
 #define UVD_MPC_SET_ALU					0xf5f8
 
+#define UVD_VCPU_CACHE_OFFSET0				0xf608
+#define UVD_VCPU_CACHE_SIZE0				0xf60c
+#define UVD_VCPU_CACHE_OFFSET1				0xf610
+#define UVD_VCPU_CACHE_SIZE1				0xf614
+#define UVD_VCPU_CACHE_OFFSET2				0xf618
+#define UVD_VCPU_CACHE_SIZE2				0xf61c
+
 #define UVD_VCPU_CNTL					0xf660
 #define UVD_SOFT_RESET					0xf680
 #define		RBC_SOFT_RESET					(1<<0)
@@ -1517,9 +1547,35 @@
 
 #define UVD_CONTEXT_ID					0xf6f4
 
+/* rs780 only */
+#define	GFX_MACRO_BYPASS_CNTL				0x30c0
+#define		SPLL_BYPASS_CNTL			(1 << 0)
+#define		UPLL_BYPASS_CNTL			(1 << 1)
+
+#define CG_UPLL_FUNC_CNTL				0x7e0
+#	define UPLL_RESET_MASK				0x00000001
+#	define UPLL_SLEEP_MASK				0x00000002
+#	define UPLL_BYPASS_EN_MASK			0x00000004
 #	define UPLL_CTLREQ_MASK				0x00000008
+#	define UPLL_FB_DIV(x)				((x) << 4)
+#	define UPLL_FB_DIV_MASK				0x0000FFF0
+#	define UPLL_REF_DIV(x)				((x) << 16)
+#	define UPLL_REF_DIV_MASK			0x003F0000
+#	define UPLL_REFCLK_SRC_SEL_MASK			0x20000000
 #	define UPLL_CTLACK_MASK				0x40000000
 #	define UPLL_CTLACK2_MASK			0x80000000
+#define CG_UPLL_FUNC_CNTL_2				0x7e4
+#	define UPLL_SW_HILEN(x)				((x) << 0)
+#	define UPLL_SW_LOLEN(x)				((x) << 4)
+#	define UPLL_SW_HILEN2(x)			((x) << 8)
+#	define UPLL_SW_LOLEN2(x)			((x) << 12)
+#	define UPLL_DIVEN_MASK				0x00010000
+#	define UPLL_DIVEN2_MASK				0x00020000
+#	define UPLL_SW_MASK				0x0003FFFF
+#	define VCLK_SRC_SEL(x)				((x) << 20)
+#	define VCLK_SRC_SEL_MASK			0x01F00000
+#	define DCLK_SRC_SEL(x)				((x) << 25)
+#	define DCLK_SRC_SEL_MASK			0x3E000000
 
 /*
  * PM4
diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h
index 3247bfd14410..3f2a8d3febca 100644
--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@@ -65,6 +65,8 @@
 #include <linux/list.h>
 #include <linux/kref.h>
 #include <linux/interval_tree.h>
+#include <linux/hashtable.h>
+#include <linux/fence.h>
 
 #include <ttm/ttm_bo_api.h>
 #include <ttm/ttm_bo_driver.h>
@@ -72,6 +74,8 @@
 #include <ttm/ttm_module.h>
 #include <ttm/ttm_execbuf_util.h>
 
+#include <drm/drm_gem.h>
+
 #include "radeon_family.h"
 #include "radeon_mode.h"
 #include "radeon_reg.h"
@@ -120,9 +124,6 @@ extern int radeon_backlight;
 #define RADEONFB_CONN_LIMIT			4
 #define RADEON_BIOS_NUM_SCRATCH			8
 
-/* fence seq are set to this number when signaled */
-#define RADEON_FENCE_SIGNALED_SEQ		0LL
-
 /* internal ring indices */
 /* r1xx+ has gfx CP ring */
 #define RADEON_RING_TYPE_GFX_INDEX		0
@@ -149,9 +150,6 @@ extern int radeon_backlight;
 /* number of hw syncs before falling back on blocking */
 #define RADEON_NUM_SYNCS			4
 
-/* number of hw syncs before falling back on blocking */
-#define RADEON_NUM_SYNCS			4
-
 /* hardcode those limit for now */
 #define RADEON_VA_IB_OFFSET			(1 << 20)
 #define RADEON_VA_RESERVED_SIZE			(8 << 20)
@@ -244,6 +242,7 @@ bool radeon_get_bios(struct radeon_device *rdev);
  * Dummy page
  */
 struct radeon_dummy_page {
+	uint64_t	entry;
 	struct page	*page;
 	dma_addr_t	addr;
 };
@@ -350,28 +349,33 @@ extern void evergreen_tiling_fields(unsigned tiling_flags, unsigned *bankw,
  * Fences.
  */
 struct radeon_fence_driver {
+	struct radeon_device		*rdev;
 	uint32_t			scratch_reg;
 	uint64_t			gpu_addr;
 	volatile uint32_t		*cpu_addr;
 	/* sync_seq is protected by ring emission lock */
 	uint64_t			sync_seq[RADEON_NUM_RINGS];
 	atomic64_t			last_seq;
-	bool				initialized;
+	bool				initialized, delayed_irq;
+	struct delayed_work		lockup_work;
 };
 
 struct radeon_fence {
-	struct radeon_device		*rdev;
-	struct kref			kref;
-	/* protected by radeon_fence.lock */
-	uint64_t			seq;
+	struct fence		base;
+
+	struct radeon_device	*rdev;
+	uint64_t		seq;
 	/* RB, DMA, etc. */
-	unsigned			ring;
+	unsigned		ring;
+	bool			is_vm_update;
+
+	wait_queue_t		fence_wake;
 };
 
 int radeon_fence_driver_start_ring(struct radeon_device *rdev, int ring);
 int radeon_fence_driver_init(struct radeon_device *rdev);
 void radeon_fence_driver_fini(struct radeon_device *rdev);
-void radeon_fence_driver_force_completion(struct radeon_device *rdev);
+void radeon_fence_driver_force_completion(struct radeon_device *rdev, int ring);
 int radeon_fence_emit(struct radeon_device *rdev, struct radeon_fence **fence, int ring);
 void radeon_fence_process(struct radeon_device *rdev, int ring);
 bool radeon_fence_signaled(struct radeon_fence *fence);
@@ -447,12 +451,22 @@ struct radeon_mman {
 #endif
 };
 
+struct radeon_bo_list {
+	struct radeon_bo		*robj;
+	struct ttm_validate_buffer	tv;
+	uint64_t			gpu_offset;
+	unsigned			prefered_domains;
+	unsigned			allowed_domains;
+	uint32_t			tiling_flags;
+};
+
 /* bo virtual address in a specific vm */
 struct radeon_bo_va {
 	/* protected by bo being reserved */
 	struct list_head		bo_list;
 	uint32_t			flags;
 	uint64_t			addr;
+	struct radeon_fence		*last_pt_update;
 	unsigned			ref_count;
 
 	/* protected by vm mutex */
@@ -469,7 +483,7 @@ struct radeon_bo {
 	struct list_head		list;
 	/* Protected by tbo.reserved */
 	u32				initial_domain;
-	u32				placements[3];
+	struct ttm_place		placements[4];
 	struct ttm_placement		placement;
 	struct ttm_buffer_object	tbo;
 	struct ttm_bo_kmap_obj		kmap;
@@ -489,6 +503,9 @@ struct radeon_bo {
 
 	struct ttm_bo_kmap_obj		dma_buf_vmap;
 	pid_t				pid;
+
+	struct radeon_mn		*mn;
+	struct interval_tree_node	mn_it;
 };
 #define gem_to_radeon_bo(gobj) container_of((gobj), struct radeon_bo, gem_base)
 
@@ -568,10 +585,9 @@ int radeon_mode_dumb_mmap(struct drm_file *filp,
  * Semaphores.
  */
 struct radeon_semaphore {
-	struct radeon_sa_bo		*sa_bo;
-	signed				waiters;
-	uint64_t			gpu_addr;
-	struct radeon_fence		*sync_to[RADEON_NUM_RINGS];
+	struct radeon_sa_bo	*sa_bo;
+	signed			waiters;
+	uint64_t		gpu_addr;
 };
 
 int radeon_semaphore_create(struct radeon_device *rdev,
@@ -580,16 +596,33 @@ bool radeon_semaphore_emit_signal(struct radeon_device *rdev, int ring,
 				  struct radeon_semaphore *semaphore);
 bool radeon_semaphore_emit_wait(struct radeon_device *rdev, int ring,
 				struct radeon_semaphore *semaphore);
-void radeon_semaphore_sync_to(struct radeon_semaphore *semaphore,
-			      struct radeon_fence *fence);
-int radeon_semaphore_sync_rings(struct radeon_device *rdev,
-				struct radeon_semaphore *semaphore,
-				int waiting_ring);
 void radeon_semaphore_free(struct radeon_device *rdev,
 			   struct radeon_semaphore **semaphore,
 			   struct radeon_fence *fence);
 
 /*
+ * Synchronization
+ */
+struct radeon_sync {
+	struct radeon_semaphore *semaphores[RADEON_NUM_SYNCS];
+	struct radeon_fence	*sync_to[RADEON_NUM_RINGS];
+	struct radeon_fence	*last_vm_update;
+};
+
+void radeon_sync_create(struct radeon_sync *sync);
+void radeon_sync_fence(struct radeon_sync *sync,
+		       struct radeon_fence *fence);
+int radeon_sync_resv(struct radeon_device *rdev,
+		     struct radeon_sync *sync,
+		     struct reservation_object *resv,
+		     bool shared);
+int radeon_sync_rings(struct radeon_device *rdev,
+		      struct radeon_sync *sync,
+		      int waiting_ring);
+void radeon_sync_free(struct radeon_device *rdev, struct radeon_sync *sync,
+		      struct radeon_fence *fence);
+
+/*
  * GART structures, functions & helpers
  */
 struct radeon_mc;
@@ -613,7 +646,7 @@ struct radeon_gart {
 	unsigned			num_cpu_pages;
 	unsigned			table_size;
 	struct page			**pages;
-	dma_addr_t			*pages_addr;
+	uint64_t			*pages_entry;
 	bool				ready;
 };
 
@@ -689,6 +722,10 @@ struct radeon_doorbell {
 
 int radeon_doorbell_get(struct radeon_device *rdev, u32 *page);
 void radeon_doorbell_free(struct radeon_device *rdev, u32 doorbell);
+void radeon_doorbell_get_kfd_info(struct radeon_device *rdev,
+				  phys_addr_t *aperture_base,
+				  size_t *aperture_size,
+				  size_t *start_offset);
 
 /*
  * IRQS.
@@ -702,7 +739,7 @@ struct radeon_flip_work {
 	uint64_t			base;
 	struct drm_pending_vblank_event *event;
 	struct radeon_bo		*old_rbo;
-	struct radeon_fence		*fence;
+	struct fence			*fence;
 };
 
 struct r500_irq_stat_regs {
@@ -780,6 +817,7 @@ struct radeon_irq {
 int radeon_irq_kms_init(struct radeon_device *rdev);
 void radeon_irq_kms_fini(struct radeon_device *rdev);
 void radeon_irq_kms_sw_irq_get(struct radeon_device *rdev, int ring);
+bool radeon_irq_kms_sw_irq_get_delayed(struct radeon_device *rdev, int ring);
 void radeon_irq_kms_sw_irq_put(struct radeon_device *rdev, int ring);
 void radeon_irq_kms_pflip_irq_get(struct radeon_device *rdev, int crtc);
 void radeon_irq_kms_pflip_irq_put(struct radeon_device *rdev, int crtc);
@@ -801,7 +839,7 @@ struct radeon_ib {
 	struct radeon_fence		*fence;
 	struct radeon_vm		*vm;
 	bool				is_const_ib;
-	struct radeon_semaphore		*semaphore;
+	struct radeon_sync		sync;
 };
 
 struct radeon_ring {
@@ -878,33 +916,40 @@ struct radeon_vm_pt {
 	uint64_t			addr;
 };
 
+struct radeon_vm_id {
+	unsigned		id;
+	uint64_t		pd_gpu_addr;
+	/* last flushed PD/PT update */
+	struct radeon_fence	*flushed_updates;
+	/* last use of vmid */
+	struct radeon_fence	*last_id_use;
+};
+
 struct radeon_vm {
-	struct rb_root			va;
-	unsigned			id;
+	struct mutex		mutex;
+
+	struct rb_root		va;
+
+	/* protecting invalidated and freed */
+	spinlock_t		status_lock;
 
 	/* BOs moved, but not yet updated in the PT */
-	struct list_head		invalidated;
+	struct list_head	invalidated;
 
 	/* BOs freed, but not yet updated in the PT */
-	struct list_head		freed;
+	struct list_head	freed;
 
 	/* contains the page directory */
-	struct radeon_bo		*page_directory;
-	uint64_t			pd_gpu_addr;
-	unsigned			max_pde_used;
+	struct radeon_bo	*page_directory;
+	unsigned		max_pde_used;
 
 	/* array of page tables, one for each page directory entry */
-	struct radeon_vm_pt		*page_tables;
+	struct radeon_vm_pt	*page_tables;
 
-	struct radeon_bo_va		*ib_bo_va;
+	struct radeon_bo_va	*ib_bo_va;
 
-	struct mutex			mutex;
-	/* last fence for cs using this vm */
-	struct radeon_fence		*fence;
-	/* last flush or NULL if we still need to flush */
-	struct radeon_fence		*last_flush;
-	/* last use of vmid */
-	struct radeon_fence		*last_id_use;
+	/* for id and flush management per ring */
+	struct radeon_vm_id	ids[RADEON_NUM_RINGS];
 };
 
 struct radeon_vm_manager {
@@ -1012,19 +1057,7 @@ void cayman_dma_fini(struct radeon_device *rdev);
 /*
  * CS.
  */
-struct radeon_cs_reloc {
-	struct drm_gem_object		*gobj;
-	struct radeon_bo		*robj;
-	struct ttm_validate_buffer	tv;
-	uint64_t			gpu_offset;
-	unsigned			prefered_domains;
-	unsigned			allowed_domains;
-	uint32_t			tiling_flags;
-	uint32_t			handle;
-};
-
 struct radeon_cs_chunk {
-	uint32_t		chunk_id;
 	uint32_t		length_dw;
 	uint32_t		*kdata;
 	void __user		*user_ptr;
@@ -1042,16 +1075,15 @@ struct radeon_cs_parser {
 	unsigned		idx;
 	/* relocations */
 	unsigned		nrelocs;
-	struct radeon_cs_reloc	*relocs;
-	struct radeon_cs_reloc	**relocs_ptr;
-	struct radeon_cs_reloc	*vm_bos;
+	struct radeon_bo_list	*relocs;
+	struct radeon_bo_list	*vm_bos;
 	struct list_head	validated;
 	unsigned		dma_reloc_idx;
 	/* indices of various chunks */
-	int			chunk_ib_idx;
-	int			chunk_relocs_idx;
-	int			chunk_flags_idx;
-	int			chunk_const_ib_idx;
+	struct radeon_cs_chunk  *chunk_ib;
+	struct radeon_cs_chunk  *chunk_relocs;
+	struct radeon_cs_chunk  *chunk_flags;
+	struct radeon_cs_chunk  *chunk_const_ib;
 	struct radeon_ib	ib;
 	struct radeon_ib	const_ib;
 	void			*track;
@@ -1065,7 +1097,7 @@ struct radeon_cs_parser {
 
 static inline u32 radeon_get_ib_value(struct radeon_cs_parser *p, int idx)
 {
-	struct radeon_cs_chunk *ibc = &p->chunks[p->chunk_ib_idx];
+	struct radeon_cs_chunk *ibc = p->chunk_ib;
 
 	if (ibc->kdata)
 		return ibc->kdata[idx];
@@ -1120,6 +1152,8 @@ struct radeon_wb {
 #define R600_WB_EVENT_OFFSET     3072
 #define CIK_WB_CP1_WPTR_OFFSET     3328
 #define CIK_WB_CP2_WPTR_OFFSET     3584
+#define R600_WB_DMA_RING_TEST_OFFSET 3588
+#define CAYMAN_WB_DMA1_RING_TEST_OFFSET 3592
 
 /**
  * struct radeon_pm - power management datas
@@ -1475,6 +1509,10 @@ struct radeon_dpm_fan {
 	u8 t_hyst;
 	u32 cycle_delay;
 	u16 t_max;
+	u8 control_mode;
+	u16 default_max_fan_pwm;
+	u16 default_fan_output_sensitivity;
+	u16 fan_output_sensitivity;
 	bool ucode_fan_control;
 };
 
@@ -1608,6 +1646,11 @@ struct radeon_pm {
 	/* internal thermal controller on rv6xx+ */
 	enum radeon_int_thermal_type int_thermal_type;
 	struct device	        *int_hwmon_dev;
+	/* fan control parameters */
+	bool                    no_fan;
+	u8                      fan_pulses_per_revolution;
+	u8                      fan_min_rpm;
+	u8                      fan_max_rpm;
 	/* dpm */
 	bool                    dpm_enabled;
 	struct radeon_dpm       dpm;
@@ -1642,7 +1685,8 @@ int radeon_uvd_get_create_msg(struct radeon_device *rdev, int ring,
 			      uint32_t handle, struct radeon_fence **fence);
 int radeon_uvd_get_destroy_msg(struct radeon_device *rdev, int ring,
 			       uint32_t handle, struct radeon_fence **fence);
-void radeon_uvd_force_into_uvd_segment(struct radeon_bo *rbo);
+void radeon_uvd_force_into_uvd_segment(struct radeon_bo *rbo,
+				       uint32_t allowed_domains);
 void radeon_uvd_free_handles(struct radeon_device *rdev,
 			     struct drm_file *filp);
 int radeon_uvd_cs_parse(struct radeon_cs_parser *parser);
@@ -1731,6 +1775,11 @@ void radeon_test_ring_sync(struct radeon_device *rdev,
 			   struct radeon_ring *cpB);
 void radeon_test_syncing(struct radeon_device *rdev);
 
+/*
+ * MMU Notifier
+ */
+int radeon_mn_register(struct radeon_bo *bo, unsigned long addr);
+void radeon_mn_unregister(struct radeon_bo *bo);
 
 /*
  * Debugfs
@@ -1764,7 +1813,8 @@ struct radeon_asic_ring {
 	void (*hdp_flush)(struct radeon_device *rdev, struct radeon_ring *ring);
 	bool (*emit_semaphore)(struct radeon_device *rdev, struct radeon_ring *cp,
 			       struct radeon_semaphore *semaphore, bool emit_wait);
-	void (*vm_flush)(struct radeon_device *rdev, int ridx, struct radeon_vm *vm);
+	void (*vm_flush)(struct radeon_device *rdev, struct radeon_ring *ring,
+			 unsigned vm_id, uint64_t pd_addr);
 
 	/* testing functions */
 	int (*ring_test)(struct radeon_device *rdev, struct radeon_ring *cp);
@@ -1798,8 +1848,9 @@ struct radeon_asic {
 	/* gart */
 	struct {
 		void (*tlb_flush)(struct radeon_device *rdev);
+		uint64_t (*get_page_entry)(uint64_t addr, uint32_t flags);
 		void (*set_page)(struct radeon_device *rdev, unsigned i,
-				 uint64_t addr, uint32_t flags);
+				 uint64_t entry);
 	} gart;
 	struct {
 		int (*init)(struct radeon_device *rdev);
@@ -1845,24 +1896,24 @@ struct radeon_asic {
 	} display;
 	/* copy functions for bo handling */
 	struct {
-		int (*blit)(struct radeon_device *rdev,
-			    uint64_t src_offset,
-			    uint64_t dst_offset,
-			    unsigned num_gpu_pages,
-			    struct radeon_fence **fence);
+		struct radeon_fence *(*blit)(struct radeon_device *rdev,
+					     uint64_t src_offset,
+					     uint64_t dst_offset,
+					     unsigned num_gpu_pages,
+					     struct reservation_object *resv);
 		u32 blit_ring_index;
-		int (*dma)(struct radeon_device *rdev,
-			   uint64_t src_offset,
-			   uint64_t dst_offset,
-			   unsigned num_gpu_pages,
-			   struct radeon_fence **fence);
+		struct radeon_fence *(*dma)(struct radeon_device *rdev,
+					    uint64_t src_offset,
+					    uint64_t dst_offset,
+					    unsigned num_gpu_pages,
+					    struct reservation_object *resv);
 		u32 dma_ring_index;
 		/* method used for bo copy */
-		int (*copy)(struct radeon_device *rdev,
-			    uint64_t src_offset,
-			    uint64_t dst_offset,
-			    unsigned num_gpu_pages,
-			    struct radeon_fence **fence);
+		struct radeon_fence *(*copy)(struct radeon_device *rdev,
+					     uint64_t src_offset,
+					     uint64_t dst_offset,
+					     unsigned num_gpu_pages,
+					     struct reservation_object *resv);
 		/* ring used for bo copies */
 		u32 copy_ring_index;
 	} copy;
@@ -2144,6 +2195,8 @@ int radeon_gem_info_ioctl(struct drm_device *dev, void *data,
 			  struct drm_file *filp);
 int radeon_gem_create_ioctl(struct drm_device *dev, void *data,
 			    struct drm_file *filp);
+int radeon_gem_userptr_ioctl(struct drm_device *dev, void *data,
+			     struct drm_file *filp);
 int radeon_gem_pin_ioctl(struct drm_device *dev, void *data,
 			 struct drm_file *file_priv);
 int radeon_gem_unpin_ioctl(struct drm_device *dev, void *data,
@@ -2300,6 +2353,7 @@ struct radeon_device {
 	struct radeon_mman		mman;
 	struct radeon_fence_driver	fence_drv[RADEON_NUM_RINGS];
 	wait_queue_head_t		fence_queue;
+	unsigned			fence_context;
 	struct mutex			ring_lock;
 	struct radeon_ring		ring[RADEON_NUM_RINGS];
 	bool				ib_pool_ready;
@@ -2318,7 +2372,7 @@ struct radeon_device {
 	bool				need_dma32;
 	bool				accel_working;
 	bool				fastfb_working; /* IGP feature*/
-	bool				needs_reset;
+	bool				needs_reset, in_reset;
 	struct radeon_surface_reg surface_regs[RADEON_GEM_MAX_SURFACES];
 	const struct firmware *me_fw;	/* all family ME firmware */
 	const struct firmware *pfp_fw;	/* r6/700 PFP firmware */
@@ -2339,7 +2393,6 @@ struct radeon_device {
 	struct radeon_mec mec;
 	struct work_struct hotplug_work;
 	struct work_struct audio_work;
-	struct work_struct reset_work;
 	int num_crtc; /* number of crtcs */
 	struct mutex dc_hw_i2c_mutex; /* display controller hw i2c mutex */
 	bool has_uvd;
@@ -2365,6 +2418,8 @@ struct radeon_device {
 	struct radeon_atcs		atcs;
 	/* srbm instance registers */
 	struct mutex			srbm_mutex;
+	/* GRBM index mutex. Protects concurrents access to GRBM index */
+	struct mutex			grbm_idx_mutex;
 	/* clock, powergating flags */
 	u32 cg_flags;
 	u32 pg_flags;
@@ -2376,6 +2431,13 @@ struct radeon_device {
 	/* tracking pinned memory */
 	u64 vram_pin_size;
 	u64 gart_pin_size;
+
+	/* amdkfd interface */
+	struct kfd_dev		*kfd;
+	struct radeon_sa_manager	kfd_bo;
+
+	struct mutex	mn_lock;
+	DECLARE_HASHTABLE(mn_hash, 7);
 };
 
 bool radeon_is_px(struct drm_device *dev);
@@ -2431,7 +2493,17 @@ void cik_mm_wdoorbell(struct radeon_device *rdev, u32 index, u32 v);
 /*
  * Cast helper
  */
-#define to_radeon_fence(p) ((struct radeon_fence *)(p))
+extern const struct fence_ops radeon_fence_ops;
+
+static inline struct radeon_fence *to_radeon_fence(struct fence *f)
+{
+	struct radeon_fence *__f = container_of(f, struct radeon_fence, base);
+
+	if (__f->base.ops == &radeon_fence_ops)
+		return __f;
+
+	return NULL;
+}
 
 /*
  * Registers read & write functions.
@@ -2751,18 +2823,25 @@ void radeon_atombios_fini(struct radeon_device *rdev);
 /*
  * RING helpers.
  */
-#if DRM_DEBUG_CODE == 0
+
+/**
+ * radeon_ring_write - write a value to the ring
+ *
+ * @ring: radeon_ring structure holding ring information
+ * @v: dword (dw) value to write
+ *
+ * Write a value to the requested ring buffer (all asics).
+ */
 static inline void radeon_ring_write(struct radeon_ring *ring, uint32_t v)
 {
+	if (ring->count_dw <= 0)
+		DRM_ERROR("radeon: writing more dwords to the ring than expected!\n");
+
 	ring->ring[ring->wptr++] = v;
 	ring->wptr &= ring->ptr_mask;
 	ring->count_dw--;
 	ring->ring_free_dw--;
 }
-#else
-/* With debugging this is just too big to inline */
-void radeon_ring_write(struct radeon_ring *ring, uint32_t v);
-#endif
 
 /*
  * ASICs macro.
@@ -2775,7 +2854,8 @@ void radeon_ring_write(struct radeon_ring *ring, uint32_t v);
 #define radeon_vga_set_state(rdev, state) (rdev)->asic->vga_set_state((rdev), (state))
 #define radeon_asic_reset(rdev) (rdev)->asic->asic_reset((rdev))
 #define radeon_gart_tlb_flush(rdev) (rdev)->asic->gart.tlb_flush((rdev))
-#define radeon_gart_set_page(rdev, i, p, f) (rdev)->asic->gart.set_page((rdev), (i), (p), (f))
+#define radeon_gart_get_page_entry(a, f) (rdev)->asic->gart.get_page_entry((a), (f))
+#define radeon_gart_set_page(rdev, i, e) (rdev)->asic->gart.set_page((rdev), (i), (e))
 #define radeon_asic_vm_init(rdev) (rdev)->asic->vm.init((rdev))
 #define radeon_asic_vm_fini(rdev) (rdev)->asic->vm.fini((rdev))
 #define radeon_asic_vm_copy_pages(rdev, ib, pe, src, count) ((rdev)->asic->vm.copy_pages((rdev), (ib), (pe), (src), (count)))
@@ -2788,7 +2868,7 @@ void radeon_ring_write(struct radeon_ring *ring, uint32_t v);
 #define radeon_ring_ib_execute(rdev, r, ib) (rdev)->asic->ring[(r)]->ib_execute((rdev), (ib))
 #define radeon_ring_ib_parse(rdev, r, ib) (rdev)->asic->ring[(r)]->ib_parse((rdev), (ib))
 #define radeon_ring_is_lockup(rdev, r, cp) (rdev)->asic->ring[(r)]->is_lockup((rdev), (cp))
-#define radeon_ring_vm_flush(rdev, r, vm) (rdev)->asic->ring[(r)]->vm_flush((rdev), (r), (vm))
+#define radeon_ring_vm_flush(rdev, r, vm_id, pd_addr) (rdev)->asic->ring[(r)->idx]->vm_flush((rdev), (r), (vm_id), (pd_addr))
 #define radeon_ring_get_rptr(rdev, r) (rdev)->asic->ring[(r)->idx]->get_rptr((rdev), (r))
 #define radeon_ring_get_wptr(rdev, r) (rdev)->asic->ring[(r)->idx]->get_wptr((rdev), (r))
 #define radeon_ring_set_wptr(rdev, r) (rdev)->asic->ring[(r)->idx]->set_wptr((rdev), (r))
@@ -2801,9 +2881,9 @@ void radeon_ring_write(struct radeon_ring *ring, uint32_t v);
 #define radeon_hdmi_setmode(rdev, e, m) (rdev)->asic->display.hdmi_setmode((e), (m))
 #define radeon_fence_ring_emit(rdev, r, fence) (rdev)->asic->ring[(r)]->emit_fence((rdev), (fence))
 #define radeon_semaphore_ring_emit(rdev, r, cp, semaphore, emit_wait) (rdev)->asic->ring[(r)]->emit_semaphore((rdev), (cp), (semaphore), (emit_wait))
-#define radeon_copy_blit(rdev, s, d, np, f) (rdev)->asic->copy.blit((rdev), (s), (d), (np), (f))
-#define radeon_copy_dma(rdev, s, d, np, f) (rdev)->asic->copy.dma((rdev), (s), (d), (np), (f))
-#define radeon_copy(rdev, s, d, np, f) (rdev)->asic->copy.copy((rdev), (s), (d), (np), (f))
+#define radeon_copy_blit(rdev, s, d, np, resv) (rdev)->asic->copy.blit((rdev), (s), (d), (np), (resv))
+#define radeon_copy_dma(rdev, s, d, np, resv) (rdev)->asic->copy.dma((rdev), (s), (d), (np), (resv))
+#define radeon_copy(rdev, s, d, np, resv) (rdev)->asic->copy.copy((rdev), (s), (d), (np), (resv))
 #define radeon_copy_blit_ring_index(rdev) (rdev)->asic->copy.blit_ring_index
 #define radeon_copy_dma_ring_index(rdev) (rdev)->asic->copy.dma_ring_index
 #define radeon_copy_ring_index(rdev) (rdev)->asic->copy.copy_ring_index
@@ -2877,6 +2957,10 @@ extern void radeon_legacy_set_clock_gating(struct radeon_device *rdev, int enabl
 extern void radeon_atom_set_clock_gating(struct radeon_device *rdev, int enable);
 extern void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain);
 extern bool radeon_ttm_bo_is_radeon_bo(struct ttm_buffer_object *bo);
+extern int radeon_ttm_tt_set_userptr(struct ttm_tt *ttm, uint64_t addr,
+				     uint32_t flags);
+extern bool radeon_ttm_tt_has_userptr(struct ttm_tt *ttm);
+extern bool radeon_ttm_tt_is_readonly(struct ttm_tt *ttm);
 extern void radeon_vram_location(struct radeon_device *rdev, struct radeon_mc *mc, u64 base);
 extern void radeon_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc);
 extern int radeon_resume_kms(struct drm_device *dev, bool resume, bool fbcon);
@@ -2893,14 +2977,14 @@ int radeon_vm_manager_init(struct radeon_device *rdev);
 void radeon_vm_manager_fini(struct radeon_device *rdev);
 int radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm);
 void radeon_vm_fini(struct radeon_device *rdev, struct radeon_vm *vm);
-struct radeon_cs_reloc *radeon_vm_get_bos(struct radeon_device *rdev,
+struct radeon_bo_list *radeon_vm_get_bos(struct radeon_device *rdev,
 					  struct radeon_vm *vm,
                                           struct list_head *head);
 struct radeon_fence *radeon_vm_grab_id(struct radeon_device *rdev,
 				       struct radeon_vm *vm, int ring);
 void radeon_vm_flush(struct radeon_device *rdev,
                      struct radeon_vm *vm,
-                     int ring);
+		     int ring, struct radeon_fence *fence);
 void radeon_vm_fence(struct radeon_device *rdev,
 		     struct radeon_vm *vm,
 		     struct radeon_fence *fence);
@@ -2934,10 +3018,10 @@ struct r600_audio_pin *r600_audio_get_pin(struct radeon_device *rdev);
 struct r600_audio_pin *dce6_audio_get_pin(struct radeon_device *rdev);
 void r600_audio_enable(struct radeon_device *rdev,
 		       struct r600_audio_pin *pin,
-		       bool enable);
+		       u8 enable_mask);
 void dce6_audio_enable(struct radeon_device *rdev,
 		       struct r600_audio_pin *pin,
-		       bool enable);
+		       u8 enable_mask);
 
 /*
  * R600 vram scratch functions
@@ -3007,7 +3091,7 @@ bool radeon_cs_packet_next_is_pkt3_nop(struct radeon_cs_parser *p);
 void radeon_cs_dump_packet(struct radeon_cs_parser *p,
 			   struct radeon_cs_packet *pkt);
 int radeon_cs_packet_next_reloc(struct radeon_cs_parser *p,
-				struct radeon_cs_reloc **cs_reloc,
+				struct radeon_bo_list **cs_reloc,
 				int nomm);
 int r600_cs_common_vline_parse(struct radeon_cs_parser *p,
 			       uint32_t *vline_start_end,
diff --git a/drivers/gpu/drm/radeon/radeon_asic.c b/drivers/gpu/drm/radeon/radeon_asic.c
index 2dd5847f9b98..ed0e10eee2dc 100644
--- a/drivers/gpu/drm/radeon/radeon_asic.c
+++ b/drivers/gpu/drm/radeon/radeon_asic.c
@@ -159,11 +159,13 @@ void radeon_agp_disable(struct radeon_device *rdev)
 		DRM_INFO("Forcing AGP to PCIE mode\n");
 		rdev->flags |= RADEON_IS_PCIE;
 		rdev->asic->gart.tlb_flush = &rv370_pcie_gart_tlb_flush;
+		rdev->asic->gart.get_page_entry = &rv370_pcie_gart_get_page_entry;
 		rdev->asic->gart.set_page = &rv370_pcie_gart_set_page;
 	} else {
 		DRM_INFO("Forcing AGP to PCI mode\n");
 		rdev->flags |= RADEON_IS_PCI;
 		rdev->asic->gart.tlb_flush = &r100_pci_gart_tlb_flush;
+		rdev->asic->gart.get_page_entry = &r100_pci_gart_get_page_entry;
 		rdev->asic->gart.set_page = &r100_pci_gart_set_page;
 	}
 	rdev->mc.gtt_size = radeon_gart_size * 1024 * 1024;
@@ -199,6 +201,7 @@ static struct radeon_asic r100_asic = {
 	.mc_wait_for_idle = &r100_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &r100_pci_gart_tlb_flush,
+		.get_page_entry = &r100_pci_gart_get_page_entry,
 		.set_page = &r100_pci_gart_set_page,
 	},
 	.ring = {
@@ -265,6 +268,7 @@ static struct radeon_asic r200_asic = {
 	.mc_wait_for_idle = &r100_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &r100_pci_gart_tlb_flush,
+		.get_page_entry = &r100_pci_gart_get_page_entry,
 		.set_page = &r100_pci_gart_set_page,
 	},
 	.ring = {
@@ -333,6 +337,20 @@ static struct radeon_asic_ring r300_gfx_ring = {
 	.set_wptr = &r100_gfx_set_wptr,
 };
 
+static struct radeon_asic_ring rv515_gfx_ring = {
+	.ib_execute = &r100_ring_ib_execute,
+	.emit_fence = &r300_fence_ring_emit,
+	.emit_semaphore = &r100_semaphore_ring_emit,
+	.cs_parse = &r300_cs_parse,
+	.ring_start = &rv515_ring_start,
+	.ring_test = &r100_ring_test,
+	.ib_test = &r100_ib_test,
+	.is_lockup = &r100_gpu_is_lockup,
+	.get_rptr = &r100_gfx_get_rptr,
+	.get_wptr = &r100_gfx_get_wptr,
+	.set_wptr = &r100_gfx_set_wptr,
+};
+
 static struct radeon_asic r300_asic = {
 	.init = &r300_init,
 	.fini = &r300_fini,
@@ -345,6 +363,7 @@ static struct radeon_asic r300_asic = {
 	.mc_wait_for_idle = &r300_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &r100_pci_gart_tlb_flush,
+		.get_page_entry = &r100_pci_gart_get_page_entry,
 		.set_page = &r100_pci_gart_set_page,
 	},
 	.ring = {
@@ -411,6 +430,7 @@ static struct radeon_asic r300_asic_pcie = {
 	.mc_wait_for_idle = &r300_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &rv370_pcie_gart_tlb_flush,
+		.get_page_entry = &rv370_pcie_gart_get_page_entry,
 		.set_page = &rv370_pcie_gart_set_page,
 	},
 	.ring = {
@@ -477,6 +497,7 @@ static struct radeon_asic r420_asic = {
 	.mc_wait_for_idle = &r300_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &rv370_pcie_gart_tlb_flush,
+		.get_page_entry = &rv370_pcie_gart_get_page_entry,
 		.set_page = &rv370_pcie_gart_set_page,
 	},
 	.ring = {
@@ -543,6 +564,7 @@ static struct radeon_asic rs400_asic = {
 	.mc_wait_for_idle = &rs400_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &rs400_gart_tlb_flush,
+		.get_page_entry = &rs400_gart_get_page_entry,
 		.set_page = &rs400_gart_set_page,
 	},
 	.ring = {
@@ -609,6 +631,7 @@ static struct radeon_asic rs600_asic = {
 	.mc_wait_for_idle = &rs600_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &rs600_gart_tlb_flush,
+		.get_page_entry = &rs600_gart_get_page_entry,
 		.set_page = &rs600_gart_set_page,
 	},
 	.ring = {
@@ -677,6 +700,7 @@ static struct radeon_asic rs690_asic = {
 	.mc_wait_for_idle = &rs690_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &rs400_gart_tlb_flush,
+		.get_page_entry = &rs400_gart_get_page_entry,
 		.set_page = &rs400_gart_set_page,
 	},
 	.ring = {
@@ -745,10 +769,11 @@ static struct radeon_asic rv515_asic = {
 	.mc_wait_for_idle = &rv515_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &rv370_pcie_gart_tlb_flush,
+		.get_page_entry = &rv370_pcie_gart_get_page_entry,
 		.set_page = &rv370_pcie_gart_set_page,
 	},
 	.ring = {
-		[RADEON_RING_TYPE_GFX_INDEX] = &r300_gfx_ring
+		[RADEON_RING_TYPE_GFX_INDEX] = &rv515_gfx_ring
 	},
 	.irq = {
 		.set = &rs600_irq_set,
@@ -811,10 +836,11 @@ static struct radeon_asic r520_asic = {
 	.mc_wait_for_idle = &r520_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &rv370_pcie_gart_tlb_flush,
+		.get_page_entry = &rv370_pcie_gart_get_page_entry,
 		.set_page = &rv370_pcie_gart_set_page,
 	},
 	.ring = {
-		[RADEON_RING_TYPE_GFX_INDEX] = &r300_gfx_ring
+		[RADEON_RING_TYPE_GFX_INDEX] = &rv515_gfx_ring
 	},
 	.irq = {
 		.set = &rs600_irq_set,
@@ -905,6 +931,7 @@ static struct radeon_asic r600_asic = {
 	.get_gpu_clock_counter = &r600_get_gpu_clock_counter,
 	.gart = {
 		.tlb_flush = &r600_pcie_gart_tlb_flush,
+		.get_page_entry = &rs600_gart_get_page_entry,
 		.set_page = &rs600_gart_set_page,
 	},
 	.ring = {
@@ -963,6 +990,19 @@ static struct radeon_asic r600_asic = {
 	},
 };
 
+static struct radeon_asic_ring rv6xx_uvd_ring = {
+	.ib_execute = &uvd_v1_0_ib_execute,
+	.emit_fence = &uvd_v1_0_fence_emit,
+	.emit_semaphore = &uvd_v1_0_semaphore_emit,
+	.cs_parse = &radeon_uvd_cs_parse,
+	.ring_test = &uvd_v1_0_ring_test,
+	.ib_test = &uvd_v1_0_ib_test,
+	.is_lockup = &radeon_ring_test_lockup,
+	.get_rptr = &uvd_v1_0_get_rptr,
+	.get_wptr = &uvd_v1_0_get_wptr,
+	.set_wptr = &uvd_v1_0_set_wptr,
+};
+
 static struct radeon_asic rv6xx_asic = {
 	.init = &r600_init,
 	.fini = &r600_fini,
@@ -977,11 +1017,13 @@ static struct radeon_asic rv6xx_asic = {
 	.get_gpu_clock_counter = &r600_get_gpu_clock_counter,
 	.gart = {
 		.tlb_flush = &r600_pcie_gart_tlb_flush,
+		.get_page_entry = &rs600_gart_get_page_entry,
 		.set_page = &rs600_gart_set_page,
 	},
 	.ring = {
 		[RADEON_RING_TYPE_GFX_INDEX] = &r600_gfx_ring,
 		[R600_RING_TYPE_DMA_INDEX] = &r600_dma_ring,
+		[R600_RING_TYPE_UVD_INDEX] = &rv6xx_uvd_ring,
 	},
 	.irq = {
 		.set = &r600_irq_set,
@@ -1067,11 +1109,13 @@ static struct radeon_asic rs780_asic = {
 	.get_gpu_clock_counter = &r600_get_gpu_clock_counter,
 	.gart = {
 		.tlb_flush = &r600_pcie_gart_tlb_flush,
+		.get_page_entry = &rs600_gart_get_page_entry,
 		.set_page = &rs600_gart_set_page,
 	},
 	.ring = {
 		[RADEON_RING_TYPE_GFX_INDEX] = &r600_gfx_ring,
 		[R600_RING_TYPE_DMA_INDEX] = &r600_dma_ring,
+		[R600_RING_TYPE_UVD_INDEX] = &rv6xx_uvd_ring,
 	},
 	.irq = {
 		.set = &r600_irq_set,
@@ -1170,6 +1214,7 @@ static struct radeon_asic rv770_asic = {
 	.get_gpu_clock_counter = &r600_get_gpu_clock_counter,
 	.gart = {
 		.tlb_flush = &r600_pcie_gart_tlb_flush,
+		.get_page_entry = &rs600_gart_get_page_entry,
 		.set_page = &rs600_gart_set_page,
 	},
 	.ring = {
@@ -1288,6 +1333,7 @@ static struct radeon_asic evergreen_asic = {
 	.get_gpu_clock_counter = &r600_get_gpu_clock_counter,
 	.gart = {
 		.tlb_flush = &evergreen_pcie_gart_tlb_flush,
+		.get_page_entry = &rs600_gart_get_page_entry,
 		.set_page = &rs600_gart_set_page,
 	},
 	.ring = {
@@ -1380,6 +1426,7 @@ static struct radeon_asic sumo_asic = {
 	.get_gpu_clock_counter = &r600_get_gpu_clock_counter,
 	.gart = {
 		.tlb_flush = &evergreen_pcie_gart_tlb_flush,
+		.get_page_entry = &rs600_gart_get_page_entry,
 		.set_page = &rs600_gart_set_page,
 	},
 	.ring = {
@@ -1471,6 +1518,7 @@ static struct radeon_asic btc_asic = {
 	.get_gpu_clock_counter = &r600_get_gpu_clock_counter,
 	.gart = {
 		.tlb_flush = &evergreen_pcie_gart_tlb_flush,
+		.get_page_entry = &rs600_gart_get_page_entry,
 		.set_page = &rs600_gart_set_page,
 	},
 	.ring = {
@@ -1606,6 +1654,7 @@ static struct radeon_asic cayman_asic = {
 	.get_gpu_clock_counter = &r600_get_gpu_clock_counter,
 	.gart = {
 		.tlb_flush = &cayman_pcie_gart_tlb_flush,
+		.get_page_entry = &rs600_gart_get_page_entry,
 		.set_page = &rs600_gart_set_page,
 	},
 	.vm = {
@@ -1709,6 +1758,7 @@ static struct radeon_asic trinity_asic = {
 	.get_gpu_clock_counter = &r600_get_gpu_clock_counter,
 	.gart = {
 		.tlb_flush = &cayman_pcie_gart_tlb_flush,
+		.get_page_entry = &rs600_gart_get_page_entry,
 		.set_page = &rs600_gart_set_page,
 	},
 	.vm = {
@@ -1842,6 +1892,7 @@ static struct radeon_asic si_asic = {
 	.get_gpu_clock_counter = &si_get_gpu_clock_counter,
 	.gart = {
 		.tlb_flush = &si_pcie_gart_tlb_flush,
+		.get_page_entry = &rs600_gart_get_page_entry,
 		.set_page = &rs600_gart_set_page,
 	},
 	.vm = {
@@ -2003,6 +2054,7 @@ static struct radeon_asic ci_asic = {
 	.get_gpu_clock_counter = &cik_get_gpu_clock_counter,
 	.gart = {
 		.tlb_flush = &cik_pcie_gart_tlb_flush,
+		.get_page_entry = &rs600_gart_get_page_entry,
 		.set_page = &rs600_gart_set_page,
 	},
 	.vm = {
@@ -2110,6 +2162,7 @@ static struct radeon_asic kv_asic = {
 	.get_gpu_clock_counter = &cik_get_gpu_clock_counter,
 	.gart = {
 		.tlb_flush = &cik_pcie_gart_tlb_flush,
+		.get_page_entry = &rs600_gart_get_page_entry,
 		.set_page = &rs600_gart_set_page,
 	},
 	.vm = {
@@ -2296,7 +2349,15 @@ int radeon_asic_init(struct radeon_device *rdev)
 	case CHIP_RS780:
 	case CHIP_RS880:
 		rdev->asic = &rs780_asic;
-		rdev->has_uvd = true;
+		/* 760G/780V/880V don't have UVD */
+		if ((rdev->pdev->device == 0x9616)||
+		    (rdev->pdev->device == 0x9611)||
+		    (rdev->pdev->device == 0x9613)||
+		    (rdev->pdev->device == 0x9711)||
+		    (rdev->pdev->device == 0x9713))
+			rdev->has_uvd = false;
+		else
+			rdev->has_uvd = true;
 		break;
 	case CHIP_RV770:
 	case CHIP_RV730:
diff --git a/drivers/gpu/drm/radeon/radeon_asic.h b/drivers/gpu/drm/radeon/radeon_asic.h
index 7756bc1e1cd3..8d787d115653 100644
--- a/drivers/gpu/drm/radeon/radeon_asic.h
+++ b/drivers/gpu/drm/radeon/radeon_asic.h
@@ -67,8 +67,9 @@ bool r100_gpu_is_lockup(struct radeon_device *rdev, struct radeon_ring *cp);
 int r100_asic_reset(struct radeon_device *rdev);
 u32 r100_get_vblank_counter(struct radeon_device *rdev, int crtc);
 void r100_pci_gart_tlb_flush(struct radeon_device *rdev);
+uint64_t r100_pci_gart_get_page_entry(uint64_t addr, uint32_t flags);
 void r100_pci_gart_set_page(struct radeon_device *rdev, unsigned i,
-			    uint64_t addr, uint32_t flags);
+			    uint64_t entry);
 void r100_ring_start(struct radeon_device *rdev, struct radeon_ring *ring);
 int r100_irq_set(struct radeon_device *rdev);
 int r100_irq_process(struct radeon_device *rdev);
@@ -81,11 +82,11 @@ bool r100_semaphore_ring_emit(struct radeon_device *rdev,
 int r100_cs_parse(struct radeon_cs_parser *p);
 void r100_pll_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v);
 uint32_t r100_pll_rreg(struct radeon_device *rdev, uint32_t reg);
-int r100_copy_blit(struct radeon_device *rdev,
-		   uint64_t src_offset,
-		   uint64_t dst_offset,
-		   unsigned num_gpu_pages,
-		   struct radeon_fence **fence);
+struct radeon_fence *r100_copy_blit(struct radeon_device *rdev,
+				    uint64_t src_offset,
+				    uint64_t dst_offset,
+				    unsigned num_gpu_pages,
+				    struct reservation_object *resv);
 int r100_set_surface_reg(struct radeon_device *rdev, int reg,
 			 uint32_t tiling_flags, uint32_t pitch,
 			 uint32_t offset, uint32_t obj_size);
@@ -152,11 +153,11 @@ void r100_gfx_set_wptr(struct radeon_device *rdev,
 /*
  * r200,rv250,rs300,rv280
  */
-extern int r200_copy_dma(struct radeon_device *rdev,
-			 uint64_t src_offset,
-			 uint64_t dst_offset,
-			 unsigned num_gpu_pages,
-			 struct radeon_fence **fence);
+struct radeon_fence *r200_copy_dma(struct radeon_device *rdev,
+				   uint64_t src_offset,
+				   uint64_t dst_offset,
+				   unsigned num_gpu_pages,
+				   struct reservation_object *resv);
 void r200_set_safe_registers(struct radeon_device *rdev);
 
 /*
@@ -172,8 +173,9 @@ extern void r300_fence_ring_emit(struct radeon_device *rdev,
 				struct radeon_fence *fence);
 extern int r300_cs_parse(struct radeon_cs_parser *p);
 extern void rv370_pcie_gart_tlb_flush(struct radeon_device *rdev);
+extern uint64_t rv370_pcie_gart_get_page_entry(uint64_t addr, uint32_t flags);
 extern void rv370_pcie_gart_set_page(struct radeon_device *rdev, unsigned i,
-				     uint64_t addr, uint32_t flags);
+				     uint64_t entry);
 extern void rv370_set_pcie_lanes(struct radeon_device *rdev, int lanes);
 extern int rv370_get_pcie_lanes(struct radeon_device *rdev);
 extern void r300_set_reg_safe(struct radeon_device *rdev);
@@ -208,8 +210,9 @@ extern void rs400_fini(struct radeon_device *rdev);
 extern int rs400_suspend(struct radeon_device *rdev);
 extern int rs400_resume(struct radeon_device *rdev);
 void rs400_gart_tlb_flush(struct radeon_device *rdev);
+uint64_t rs400_gart_get_page_entry(uint64_t addr, uint32_t flags);
 void rs400_gart_set_page(struct radeon_device *rdev, unsigned i,
-			 uint64_t addr, uint32_t flags);
+			 uint64_t entry);
 uint32_t rs400_mc_rreg(struct radeon_device *rdev, uint32_t reg);
 void rs400_mc_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v);
 int rs400_gart_init(struct radeon_device *rdev);
@@ -232,8 +235,9 @@ int rs600_irq_process(struct radeon_device *rdev);
 void rs600_irq_disable(struct radeon_device *rdev);
 u32 rs600_get_vblank_counter(struct radeon_device *rdev, int crtc);
 void rs600_gart_tlb_flush(struct radeon_device *rdev);
+uint64_t rs600_gart_get_page_entry(uint64_t addr, uint32_t flags);
 void rs600_gart_set_page(struct radeon_device *rdev, unsigned i,
-			 uint64_t addr, uint32_t flags);
+			 uint64_t entry);
 uint32_t rs600_mc_rreg(struct radeon_device *rdev, uint32_t reg);
 void rs600_mc_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v);
 void rs600_bandwidth_update(struct radeon_device *rdev);
@@ -340,12 +344,14 @@ int r600_dma_ib_test(struct radeon_device *rdev, struct radeon_ring *ring);
 void r600_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib);
 int r600_ring_test(struct radeon_device *rdev, struct radeon_ring *cp);
 int r600_dma_ring_test(struct radeon_device *rdev, struct radeon_ring *cp);
-int r600_copy_cpdma(struct radeon_device *rdev,
-		    uint64_t src_offset, uint64_t dst_offset,
-		    unsigned num_gpu_pages, struct radeon_fence **fence);
-int r600_copy_dma(struct radeon_device *rdev,
-		  uint64_t src_offset, uint64_t dst_offset,
-		  unsigned num_gpu_pages, struct radeon_fence **fence);
+struct radeon_fence *r600_copy_cpdma(struct radeon_device *rdev,
+				     uint64_t src_offset, uint64_t dst_offset,
+				     unsigned num_gpu_pages,
+				     struct reservation_object *resv);
+struct radeon_fence *r600_copy_dma(struct radeon_device *rdev,
+				   uint64_t src_offset, uint64_t dst_offset,
+				   unsigned num_gpu_pages,
+				   struct reservation_object *resv);
 void r600_hpd_init(struct radeon_device *rdev);
 void r600_hpd_fini(struct radeon_device *rdev);
 bool r600_hpd_sense(struct radeon_device *rdev, enum radeon_hpd_id hpd);
@@ -389,7 +395,6 @@ void r600_disable_interrupts(struct radeon_device *rdev);
 void r600_rlc_stop(struct radeon_device *rdev);
 /* r600 audio */
 int r600_audio_init(struct radeon_device *rdev);
-struct r600_audio_pin r600_audio_status(struct radeon_device *rdev);
 void r600_audio_fini(struct radeon_device *rdev);
 void r600_audio_set_dto(struct drm_encoder *encoder, u32 clock);
 void r600_hdmi_update_avi_infoframe(struct drm_encoder *encoder, void *buffer,
@@ -461,10 +466,10 @@ bool rv770_page_flip_pending(struct radeon_device *rdev, int crtc);
 void r700_vram_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc);
 void r700_cp_stop(struct radeon_device *rdev);
 void r700_cp_fini(struct radeon_device *rdev);
-int rv770_copy_dma(struct radeon_device *rdev,
-		  uint64_t src_offset, uint64_t dst_offset,
-		  unsigned num_gpu_pages,
-		   struct radeon_fence **fence);
+struct radeon_fence *rv770_copy_dma(struct radeon_device *rdev,
+				    uint64_t src_offset, uint64_t dst_offset,
+				    unsigned num_gpu_pages,
+				    struct reservation_object *resv);
 u32 rv770_get_xclk(struct radeon_device *rdev);
 int rv770_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk);
 int rv770_get_temp(struct radeon_device *rdev);
@@ -535,10 +540,10 @@ void evergreen_dma_fence_ring_emit(struct radeon_device *rdev,
 				   struct radeon_fence *fence);
 void evergreen_dma_ring_ib_execute(struct radeon_device *rdev,
 				   struct radeon_ib *ib);
-int evergreen_copy_dma(struct radeon_device *rdev,
-		       uint64_t src_offset, uint64_t dst_offset,
-		       unsigned num_gpu_pages,
-		       struct radeon_fence **fence);
+struct radeon_fence *evergreen_copy_dma(struct radeon_device *rdev,
+					uint64_t src_offset, uint64_t dst_offset,
+					unsigned num_gpu_pages,
+					struct reservation_object *resv);
 void evergreen_hdmi_enable(struct drm_encoder *encoder, bool enable);
 void evergreen_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mode);
 int evergreen_get_temp(struct radeon_device *rdev);
@@ -598,7 +603,8 @@ int cayman_asic_reset(struct radeon_device *rdev);
 void cayman_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib);
 int cayman_vm_init(struct radeon_device *rdev);
 void cayman_vm_fini(struct radeon_device *rdev);
-void cayman_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm);
+void cayman_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring,
+		     unsigned vm_id, uint64_t pd_addr);
 uint32_t cayman_vm_page_flags(struct radeon_device *rdev, uint32_t flags);
 int evergreen_ib_parse(struct radeon_device *rdev, struct radeon_ib *ib);
 int evergreen_dma_ib_parse(struct radeon_device *rdev, struct radeon_ib *ib);
@@ -623,7 +629,8 @@ void cayman_dma_vm_set_pages(struct radeon_device *rdev,
 			     uint32_t incr, uint32_t flags);
 void cayman_dma_vm_pad_ib(struct radeon_ib *ib);
 
-void cayman_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm);
+void cayman_dma_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring,
+			 unsigned vm_id, uint64_t pd_addr);
 
 u32 cayman_gfx_get_rptr(struct radeon_device *rdev,
 			struct radeon_ring *ring);
@@ -698,12 +705,13 @@ int si_irq_set(struct radeon_device *rdev);
 int si_irq_process(struct radeon_device *rdev);
 int si_vm_init(struct radeon_device *rdev);
 void si_vm_fini(struct radeon_device *rdev);
-void si_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm);
+void si_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring,
+		 unsigned vm_id, uint64_t pd_addr);
 int si_ib_parse(struct radeon_device *rdev, struct radeon_ib *ib);
-int si_copy_dma(struct radeon_device *rdev,
-		uint64_t src_offset, uint64_t dst_offset,
-		unsigned num_gpu_pages,
-		struct radeon_fence **fence);
+struct radeon_fence *si_copy_dma(struct radeon_device *rdev,
+				 uint64_t src_offset, uint64_t dst_offset,
+				 unsigned num_gpu_pages,
+				 struct reservation_object *resv);
 
 void si_dma_vm_copy_pages(struct radeon_device *rdev,
 			  struct radeon_ib *ib,
@@ -720,7 +728,8 @@ void si_dma_vm_set_pages(struct radeon_device *rdev,
 			 uint64_t addr, unsigned count,
 			 uint32_t incr, uint32_t flags);
 
-void si_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm);
+void si_dma_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring,
+		     unsigned vm_id, uint64_t pd_addr);
 u32 si_get_xclk(struct radeon_device *rdev);
 uint64_t si_get_gpu_clock_counter(struct radeon_device *rdev);
 int si_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk);
@@ -759,14 +768,14 @@ bool cik_sdma_semaphore_ring_emit(struct radeon_device *rdev,
 				  struct radeon_semaphore *semaphore,
 				  bool emit_wait);
 void cik_sdma_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib);
-int cik_copy_dma(struct radeon_device *rdev,
-		 uint64_t src_offset, uint64_t dst_offset,
-		 unsigned num_gpu_pages,
-		 struct radeon_fence **fence);
-int cik_copy_cpdma(struct radeon_device *rdev,
-		   uint64_t src_offset, uint64_t dst_offset,
-		   unsigned num_gpu_pages,
-		   struct radeon_fence **fence);
+struct radeon_fence *cik_copy_dma(struct radeon_device *rdev,
+				  uint64_t src_offset, uint64_t dst_offset,
+				  unsigned num_gpu_pages,
+				  struct reservation_object *resv);
+struct radeon_fence *cik_copy_cpdma(struct radeon_device *rdev,
+				    uint64_t src_offset, uint64_t dst_offset,
+				    unsigned num_gpu_pages,
+				    struct reservation_object *resv);
 int cik_sdma_ring_test(struct radeon_device *rdev, struct radeon_ring *ring);
 int cik_sdma_ib_test(struct radeon_device *rdev, struct radeon_ring *ring);
 bool cik_sdma_is_lockup(struct radeon_device *rdev, struct radeon_ring *ring);
@@ -792,7 +801,8 @@ int cik_irq_set(struct radeon_device *rdev);
 int cik_irq_process(struct radeon_device *rdev);
 int cik_vm_init(struct radeon_device *rdev);
 void cik_vm_fini(struct radeon_device *rdev);
-void cik_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm);
+void cik_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring,
+		  unsigned vm_id, uint64_t pd_addr);
 
 void cik_sdma_vm_copy_pages(struct radeon_device *rdev,
 			    struct radeon_ib *ib,
@@ -810,7 +820,8 @@ void cik_sdma_vm_set_pages(struct radeon_device *rdev,
 			   uint32_t incr, uint32_t flags);
 void cik_sdma_vm_pad_ib(struct radeon_ib *ib);
 
-void cik_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm);
+void cik_dma_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring,
+		      unsigned vm_id, uint64_t pd_addr);
 int cik_ib_parse(struct radeon_device *rdev, struct radeon_ib *ib);
 u32 cik_gfx_get_rptr(struct radeon_device *rdev,
 		     struct radeon_ring *ring);
@@ -882,6 +893,7 @@ uint32_t uvd_v1_0_get_wptr(struct radeon_device *rdev,
                            struct radeon_ring *ring);
 void uvd_v1_0_set_wptr(struct radeon_device *rdev,
                        struct radeon_ring *ring);
+int uvd_v1_0_resume(struct radeon_device *rdev);
 
 int uvd_v1_0_init(struct radeon_device *rdev);
 void uvd_v1_0_fini(struct radeon_device *rdev);
@@ -889,6 +901,8 @@ int uvd_v1_0_start(struct radeon_device *rdev);
 void uvd_v1_0_stop(struct radeon_device *rdev);
 
 int uvd_v1_0_ring_test(struct radeon_device *rdev, struct radeon_ring *ring);
+void uvd_v1_0_fence_emit(struct radeon_device *rdev,
+			 struct radeon_fence *fence);
 int uvd_v1_0_ib_test(struct radeon_device *rdev, struct radeon_ring *ring);
 bool uvd_v1_0_semaphore_emit(struct radeon_device *rdev,
 			     struct radeon_ring *ring,
diff --git a/drivers/gpu/drm/radeon/radeon_atombios.c b/drivers/gpu/drm/radeon/radeon_atombios.c
index e74c7e387dde..dbc94f300297 100644
--- a/drivers/gpu/drm/radeon/radeon_atombios.c
+++ b/drivers/gpu/drm/radeon/radeon_atombios.c
@@ -196,8 +196,8 @@ void radeon_atombios_i2c_init(struct radeon_device *rdev)
 	}
 }
 
-static struct radeon_gpio_rec radeon_lookup_gpio(struct radeon_device *rdev,
-						 u8 id)
+struct radeon_gpio_rec radeon_atombios_lookup_gpio(struct radeon_device *rdev,
+						   u8 id)
 {
 	struct atom_context *ctx = rdev->mode_info.atom_context;
 	struct radeon_gpio_rec gpio;
@@ -221,6 +221,7 @@ static struct radeon_gpio_rec radeon_lookup_gpio(struct radeon_device *rdev,
 			if (id == pin->ucGPIO_ID) {
 				gpio.id = pin->ucGPIO_ID;
 				gpio.reg = le16_to_cpu(pin->usGpioPin_AIndex) * 4;
+				gpio.shift = pin->ucGpioPinBitShift;
 				gpio.mask = (1 << pin->ucGpioPinBitShift);
 				gpio.valid = true;
 				break;
@@ -458,7 +459,7 @@ static bool radeon_atom_apply_quirks(struct drm_device *dev,
 	return true;
 }
 
-const int supported_devices_connector_convert[] = {
+static const int supported_devices_connector_convert[] = {
 	DRM_MODE_CONNECTOR_Unknown,
 	DRM_MODE_CONNECTOR_VGA,
 	DRM_MODE_CONNECTOR_DVII,
@@ -477,7 +478,7 @@ const int supported_devices_connector_convert[] = {
 	DRM_MODE_CONNECTOR_DisplayPort
 };
 
-const uint16_t supported_devices_connector_object_id_convert[] = {
+static const uint16_t supported_devices_connector_object_id_convert[] = {
 	CONNECTOR_OBJECT_ID_NONE,
 	CONNECTOR_OBJECT_ID_VGA,
 	CONNECTOR_OBJECT_ID_DUAL_LINK_DVI_I, /* not all boards support DL */
@@ -494,7 +495,7 @@ const uint16_t supported_devices_connector_object_id_convert[] = {
 	CONNECTOR_OBJECT_ID_SVIDEO
 };
 
-const int object_connector_convert[] = {
+static const int object_connector_convert[] = {
 	DRM_MODE_CONNECTOR_Unknown,
 	DRM_MODE_CONNECTOR_DVII,
 	DRM_MODE_CONNECTOR_DVII,
@@ -801,7 +802,7 @@ bool radeon_get_atom_connector_info_from_object_table(struct drm_device *dev)
 								hpd_record =
 									(ATOM_HPD_INT_RECORD *)
 									record;
-								gpio = radeon_lookup_gpio(rdev,
+								gpio = radeon_atombios_lookup_gpio(rdev,
 											  hpd_record->ucHPDIntGPIOID);
 								hpd = radeon_atom_get_hpd_info_from_gpio(rdev, &gpio);
 								hpd.plugged_state = hpd_record->ucPlugged_PinState;
@@ -2128,7 +2129,7 @@ static int radeon_atombios_parse_power_table_1_3(struct radeon_device *rdev)
 				rdev->pm.power_state[state_index].clock_info[0].voltage.type =
 					VOLTAGE_GPIO;
 				rdev->pm.power_state[state_index].clock_info[0].voltage.gpio =
-					radeon_lookup_gpio(rdev,
+					radeon_atombios_lookup_gpio(rdev,
 							   power_info->info.asPowerPlayInfo[i].ucVoltageDropIndex);
 				if (misc & ATOM_PM_MISCINFO_VOLTAGE_DROP_ACTIVE_HIGH)
 					rdev->pm.power_state[state_index].clock_info[0].voltage.active_high =
@@ -2164,7 +2165,7 @@ static int radeon_atombios_parse_power_table_1_3(struct radeon_device *rdev)
 				rdev->pm.power_state[state_index].clock_info[0].voltage.type =
 					VOLTAGE_GPIO;
 				rdev->pm.power_state[state_index].clock_info[0].voltage.gpio =
-					radeon_lookup_gpio(rdev,
+					radeon_atombios_lookup_gpio(rdev,
 							   power_info->info_2.asPowerPlayInfo[i].ucVoltageDropIndex);
 				if (misc & ATOM_PM_MISCINFO_VOLTAGE_DROP_ACTIVE_HIGH)
 					rdev->pm.power_state[state_index].clock_info[0].voltage.active_high =
@@ -2200,7 +2201,7 @@ static int radeon_atombios_parse_power_table_1_3(struct radeon_device *rdev)
 				rdev->pm.power_state[state_index].clock_info[0].voltage.type =
 					VOLTAGE_GPIO;
 				rdev->pm.power_state[state_index].clock_info[0].voltage.gpio =
-					radeon_lookup_gpio(rdev,
+					radeon_atombios_lookup_gpio(rdev,
 							   power_info->info_3.asPowerPlayInfo[i].ucVoltageDropIndex);
 				if (misc & ATOM_PM_MISCINFO_VOLTAGE_DROP_ACTIVE_HIGH)
 					rdev->pm.power_state[state_index].clock_info[0].voltage.active_high =
@@ -2248,6 +2249,14 @@ static void radeon_atombios_add_pplib_thermal_controller(struct radeon_device *r
 
 	/* add the i2c bus for thermal/fan chip */
 	if (controller->ucType > 0) {
+		if (controller->ucFanParameters & ATOM_PP_FANPARAMETERS_NOFAN)
+			rdev->pm.no_fan = true;
+		rdev->pm.fan_pulses_per_revolution =
+			controller->ucFanParameters & ATOM_PP_FANPARAMETERS_TACHOMETER_PULSES_PER_REVOLUTION_MASK;
+		if (rdev->pm.fan_pulses_per_revolution) {
+			rdev->pm.fan_min_rpm = controller->ucFanMinRPM;
+			rdev->pm.fan_max_rpm = controller->ucFanMaxRPM;
+		}
 		if (controller->ucType == ATOM_PP_THERMALCONTROLLER_RV6xx) {
 			DRM_INFO("Internal thermal controller %s fan control\n",
 				 (controller->ucFanParameters &
diff --git a/drivers/gpu/drm/radeon/radeon_benchmark.c b/drivers/gpu/drm/radeon/radeon_benchmark.c
index 69f5695bdab9..9e7f23dd14bd 100644
--- a/drivers/gpu/drm/radeon/radeon_benchmark.c
+++ b/drivers/gpu/drm/radeon/radeon_benchmark.c
@@ -45,33 +45,29 @@ static int radeon_benchmark_do_move(struct radeon_device *rdev, unsigned size,
 	for (i = 0; i < n; i++) {
 		switch (flag) {
 		case RADEON_BENCHMARK_COPY_DMA:
-			r = radeon_copy_dma(rdev, saddr, daddr,
-					    size / RADEON_GPU_PAGE_SIZE,
-					    &fence);
+			fence = radeon_copy_dma(rdev, saddr, daddr,
+						size / RADEON_GPU_PAGE_SIZE,
+						NULL);
 			break;
 		case RADEON_BENCHMARK_COPY_BLIT:
-			r = radeon_copy_blit(rdev, saddr, daddr,
-					     size / RADEON_GPU_PAGE_SIZE,
-					     &fence);
+			fence = radeon_copy_blit(rdev, saddr, daddr,
+						 size / RADEON_GPU_PAGE_SIZE,
+						 NULL);
 			break;
 		default:
 			DRM_ERROR("Unknown copy method\n");
-			r = -EINVAL;
+			return -EINVAL;
 		}
-		if (r)
-			goto exit_do_move;
+		if (IS_ERR(fence))
+			return PTR_ERR(fence);
+
 		r = radeon_fence_wait(fence, false);
-		if (r)
-			goto exit_do_move;
 		radeon_fence_unref(&fence);
+		if (r)
+			return r;
 	}
 	end_jiffies = jiffies;
-	r = jiffies_to_msecs(end_jiffies - start_jiffies);
-
-exit_do_move:
-	if (fence)
-		radeon_fence_unref(&fence);
-	return r;
+	return jiffies_to_msecs(end_jiffies - start_jiffies);
 }
 
 
@@ -97,7 +93,7 @@ static void radeon_benchmark_move(struct radeon_device *rdev, unsigned size,
 	int time;
 
 	n = RADEON_BENCHMARK_ITERATIONS;
-	r = radeon_bo_create(rdev, size, PAGE_SIZE, true, sdomain, 0, NULL, &sobj);
+	r = radeon_bo_create(rdev, size, PAGE_SIZE, true, sdomain, 0, NULL, NULL, &sobj);
 	if (r) {
 		goto out_cleanup;
 	}
@@ -109,7 +105,7 @@ static void radeon_benchmark_move(struct radeon_device *rdev, unsigned size,
 	if (r) {
 		goto out_cleanup;
 	}
-	r = radeon_bo_create(rdev, size, PAGE_SIZE, true, ddomain, 0, NULL, &dobj);
+	r = radeon_bo_create(rdev, size, PAGE_SIZE, true, ddomain, 0, NULL, NULL, &dobj);
 	if (r) {
 		goto out_cleanup;
 	}
diff --git a/drivers/gpu/drm/radeon/radeon_bios.c b/drivers/gpu/drm/radeon/radeon_bios.c
index 6a03624fadaa..63ccb8fa799c 100644
--- a/drivers/gpu/drm/radeon/radeon_bios.c
+++ b/drivers/gpu/drm/radeon/radeon_bios.c
@@ -658,12 +658,10 @@ bool radeon_get_bios(struct radeon_device *rdev)
 		r = igp_read_bios_from_vram(rdev);
 	if (r == false)
 		r = radeon_read_bios(rdev);
-	if (r == false) {
+	if (r == false)
 		r = radeon_read_disabled_bios(rdev);
-	}
-	if (r == false) {
+	if (r == false)
 		r = radeon_read_platform_bios(rdev);
-	}
 	if (r == false || rdev->bios == NULL) {
 		DRM_ERROR("Unable to locate a BIOS ROM\n");
 		rdev->bios = NULL;
diff --git a/drivers/gpu/drm/radeon/radeon_combios.c b/drivers/gpu/drm/radeon/radeon_combios.c
index 6651177110f0..3e5f6b71f3ad 100644
--- a/drivers/gpu/drm/radeon/radeon_combios.c
+++ b/drivers/gpu/drm/radeon/radeon_combios.c
@@ -116,7 +116,7 @@ enum radeon_combios_connector {
 	CONNECTOR_UNSUPPORTED_LEGACY
 };
 
-const int legacy_connector_convert[] = {
+static const int legacy_connector_convert[] = {
 	DRM_MODE_CONNECTOR_Unknown,
 	DRM_MODE_CONNECTOR_DVID,
 	DRM_MODE_CONNECTOR_VGA,
diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c
index 300c4b3d4669..26baa9c05f6c 100644
--- a/drivers/gpu/drm/radeon/radeon_connectors.c
+++ b/drivers/gpu/drm/radeon/radeon_connectors.c
@@ -322,6 +322,12 @@ static void radeon_connector_get_edid(struct drm_connector *connector)
 	}
 
 	if (!radeon_connector->edid) {
+		/* don't fetch the edid from the vbios if ddc fails and runpm is
+		 * enabled so we report disconnected.
+		 */
+		if ((rdev->flags & RADEON_IS_PX) && (radeon_runtime_pm != 0))
+			return;
+
 		if (rdev->is_atom_bios) {
 			/* some laptops provide a hardcoded edid in rom for LCDs */
 			if (((connector->connector_type == DRM_MODE_CONNECTOR_LVDS) ||
@@ -826,6 +832,8 @@ static int radeon_lvds_mode_valid(struct drm_connector *connector,
 static enum drm_connector_status
 radeon_lvds_detect(struct drm_connector *connector, bool force)
 {
+	struct drm_device *dev = connector->dev;
+	struct radeon_device *rdev = dev->dev_private;
 	struct radeon_connector *radeon_connector = to_radeon_connector(connector);
 	struct drm_encoder *encoder = radeon_best_single_encoder(connector);
 	enum drm_connector_status ret = connector_status_disconnected;
@@ -842,7 +850,11 @@ radeon_lvds_detect(struct drm_connector *connector, bool force)
 		/* check if panel is valid */
 		if (native_mode->hdisplay >= 320 && native_mode->vdisplay >= 240)
 			ret = connector_status_connected;
-
+		/* don't fetch the edid from the vbios if ddc fails and runpm is
+		 * enabled so we report disconnected.
+		 */
+		if ((rdev->flags & RADEON_IS_PX) && (radeon_runtime_pm != 0))
+			ret = connector_status_disconnected;
 	}
 
 	/* check for edid as well */
@@ -1589,6 +1601,11 @@ radeon_dp_detect(struct drm_connector *connector, bool force)
 			/* check if panel is valid */
 			if (native_mode->hdisplay >= 320 && native_mode->vdisplay >= 240)
 				ret = connector_status_connected;
+			/* don't fetch the edid from the vbios if ddc fails and runpm is
+			 * enabled so we report disconnected.
+			 */
+			if ((rdev->flags & RADEON_IS_PX) && (radeon_runtime_pm != 0))
+				ret = connector_status_disconnected;
 		}
 		/* eDP is always DP */
 		radeon_dig_connector->dp_sink_type = CONNECTOR_OBJECT_ID_DISPLAYPORT;
diff --git a/drivers/gpu/drm/radeon/radeon_cp.c b/drivers/gpu/drm/radeon/radeon_cp.c
index bb0d5c3a8311..ea134a7d51a5 100644
--- a/drivers/gpu/drm/radeon/radeon_cp.c
+++ b/drivers/gpu/drm/radeon/radeon_cp.c
@@ -1298,27 +1298,27 @@ static int radeon_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init,
 	dev_priv->buffers_offset = init->buffers_offset;
 	dev_priv->gart_textures_offset = init->gart_textures_offset;
 
-	master_priv->sarea = drm_getsarea(dev);
+	master_priv->sarea = drm_legacy_getsarea(dev);
 	if (!master_priv->sarea) {
 		DRM_ERROR("could not find sarea!\n");
 		radeon_do_cleanup_cp(dev);
 		return -EINVAL;
 	}
 
-	dev_priv->cp_ring = drm_core_findmap(dev, init->ring_offset);
+	dev_priv->cp_ring = drm_legacy_findmap(dev, init->ring_offset);
 	if (!dev_priv->cp_ring) {
 		DRM_ERROR("could not find cp ring region!\n");
 		radeon_do_cleanup_cp(dev);
 		return -EINVAL;
 	}
-	dev_priv->ring_rptr = drm_core_findmap(dev, init->ring_rptr_offset);
+	dev_priv->ring_rptr = drm_legacy_findmap(dev, init->ring_rptr_offset);
 	if (!dev_priv->ring_rptr) {
 		DRM_ERROR("could not find ring read pointer!\n");
 		radeon_do_cleanup_cp(dev);
 		return -EINVAL;
 	}
 	dev->agp_buffer_token = init->buffers_offset;
-	dev->agp_buffer_map = drm_core_findmap(dev, init->buffers_offset);
+	dev->agp_buffer_map = drm_legacy_findmap(dev, init->buffers_offset);
 	if (!dev->agp_buffer_map) {
 		DRM_ERROR("could not find dma buffer region!\n");
 		radeon_do_cleanup_cp(dev);
@@ -1327,7 +1327,7 @@ static int radeon_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init,
 
 	if (init->gart_textures_offset) {
 		dev_priv->gart_textures =
-		    drm_core_findmap(dev, init->gart_textures_offset);
+		    drm_legacy_findmap(dev, init->gart_textures_offset);
 		if (!dev_priv->gart_textures) {
 			DRM_ERROR("could not find GART texture region!\n");
 			radeon_do_cleanup_cp(dev);
@@ -1337,9 +1337,9 @@ static int radeon_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init,
 
 #if __OS_HAS_AGP
 	if (dev_priv->flags & RADEON_IS_AGP) {
-		drm_core_ioremap_wc(dev_priv->cp_ring, dev);
-		drm_core_ioremap_wc(dev_priv->ring_rptr, dev);
-		drm_core_ioremap_wc(dev->agp_buffer_map, dev);
+		drm_legacy_ioremap_wc(dev_priv->cp_ring, dev);
+		drm_legacy_ioremap_wc(dev_priv->ring_rptr, dev);
+		drm_legacy_ioremap_wc(dev->agp_buffer_map, dev);
 		if (!dev_priv->cp_ring->handle ||
 		    !dev_priv->ring_rptr->handle ||
 		    !dev->agp_buffer_map->handle) {
@@ -1475,7 +1475,7 @@ static int radeon_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init,
 			dev_priv->gart_info.mapping.size =
 			    dev_priv->gart_info.table_size;
 
-			drm_core_ioremap_wc(&dev_priv->gart_info.mapping, dev);
+			drm_legacy_ioremap_wc(&dev_priv->gart_info.mapping, dev);
 			dev_priv->gart_info.addr =
 			    dev_priv->gart_info.mapping.handle;
 
@@ -1569,15 +1569,15 @@ static int radeon_do_cleanup_cp(struct drm_device * dev)
 #if __OS_HAS_AGP
 	if (dev_priv->flags & RADEON_IS_AGP) {
 		if (dev_priv->cp_ring != NULL) {
-			drm_core_ioremapfree(dev_priv->cp_ring, dev);
+			drm_legacy_ioremapfree(dev_priv->cp_ring, dev);
 			dev_priv->cp_ring = NULL;
 		}
 		if (dev_priv->ring_rptr != NULL) {
-			drm_core_ioremapfree(dev_priv->ring_rptr, dev);
+			drm_legacy_ioremapfree(dev_priv->ring_rptr, dev);
 			dev_priv->ring_rptr = NULL;
 		}
 		if (dev->agp_buffer_map != NULL) {
-			drm_core_ioremapfree(dev->agp_buffer_map, dev);
+			drm_legacy_ioremapfree(dev->agp_buffer_map, dev);
 			dev->agp_buffer_map = NULL;
 		}
 	} else
@@ -1597,7 +1597,7 @@ static int radeon_do_cleanup_cp(struct drm_device * dev)
 
 		if (dev_priv->gart_info.gart_table_location == DRM_ATI_GART_FB)
 		{
-			drm_core_ioremapfree(&dev_priv->gart_info.mapping, dev);
+			drm_legacy_ioremapfree(&dev_priv->gart_info.mapping, dev);
 			dev_priv->gart_info.addr = NULL;
 		}
 	}
@@ -2106,9 +2106,9 @@ int radeon_driver_load(struct drm_device *dev, unsigned long flags)
 	else
 		dev_priv->flags |= RADEON_IS_PCI;
 
-	ret = drm_addmap(dev, pci_resource_start(dev->pdev, 2),
-			 pci_resource_len(dev->pdev, 2), _DRM_REGISTERS,
-			 _DRM_READ_ONLY | _DRM_DRIVER, &dev_priv->mmio);
+	ret = drm_legacy_addmap(dev, pci_resource_start(dev->pdev, 2),
+				pci_resource_len(dev->pdev, 2), _DRM_REGISTERS,
+				_DRM_READ_ONLY | _DRM_DRIVER, &dev_priv->mmio);
 	if (ret != 0)
 		return ret;
 
@@ -2135,8 +2135,8 @@ int radeon_master_create(struct drm_device *dev, struct drm_master *master)
 
 	/* prebuild the SAREA */
 	sareapage = max_t(unsigned long, SAREA_MAX, PAGE_SIZE);
-	ret = drm_addmap(dev, 0, sareapage, _DRM_SHM, _DRM_CONTAINS_LOCK,
-			 &master_priv->sarea);
+	ret = drm_legacy_addmap(dev, 0, sareapage, _DRM_SHM, _DRM_CONTAINS_LOCK,
+				&master_priv->sarea);
 	if (ret) {
 		DRM_ERROR("SAREA setup failed\n");
 		kfree(master_priv);
@@ -2162,7 +2162,7 @@ void radeon_master_destroy(struct drm_device *dev, struct drm_master *master)
 
 	master_priv->sarea_priv = NULL;
 	if (master_priv->sarea)
-		drm_rmmap_locked(dev, master_priv->sarea);
+		drm_legacy_rmmap_locked(dev, master_priv->sarea);
 
 	kfree(master_priv);
 
@@ -2181,9 +2181,9 @@ int radeon_driver_firstopen(struct drm_device *dev)
 	dev_priv->gart_info.table_size = RADEON_PCIGART_TABLE_SIZE;
 
 	dev_priv->fb_aper_offset = pci_resource_start(dev->pdev, 0);
-	ret = drm_addmap(dev, dev_priv->fb_aper_offset,
-			 pci_resource_len(dev->pdev, 0), _DRM_FRAME_BUFFER,
-			 _DRM_WRITE_COMBINING, &map);
+	ret = drm_legacy_addmap(dev, dev_priv->fb_aper_offset,
+				pci_resource_len(dev->pdev, 0),
+				_DRM_FRAME_BUFFER, _DRM_WRITE_COMBINING, &map);
 	if (ret != 0)
 		return ret;
 
@@ -2196,7 +2196,7 @@ int radeon_driver_unload(struct drm_device *dev)
 
 	DRM_DEBUG("\n");
 
-	drm_rmmap(dev, dev_priv->mmio);
+	drm_legacy_rmmap(dev, dev_priv->mmio);
 
 	kfree(dev_priv);
 
diff --git a/drivers/gpu/drm/radeon/radeon_cs.c b/drivers/gpu/drm/radeon/radeon_cs.c
index 83f382e8e40e..c830863bc98a 100644
--- a/drivers/gpu/drm/radeon/radeon_cs.c
+++ b/drivers/gpu/drm/radeon/radeon_cs.c
@@ -77,21 +77,18 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 	struct drm_device *ddev = p->rdev->ddev;
 	struct radeon_cs_chunk *chunk;
 	struct radeon_cs_buckets buckets;
-	unsigned i, j;
-	bool duplicate;
+	unsigned i;
+	bool need_mmap_lock = false;
+	int r;
 
-	if (p->chunk_relocs_idx == -1) {
+	if (p->chunk_relocs == NULL) {
 		return 0;
 	}
-	chunk = &p->chunks[p->chunk_relocs_idx];
+	chunk = p->chunk_relocs;
 	p->dma_reloc_idx = 0;
 	/* FIXME: we assume that each relocs use 4 dwords */
 	p->nrelocs = chunk->length_dw / 4;
-	p->relocs_ptr = kcalloc(p->nrelocs, sizeof(void *), GFP_KERNEL);
-	if (p->relocs_ptr == NULL) {
-		return -ENOMEM;
-	}
-	p->relocs = kcalloc(p->nrelocs, sizeof(struct radeon_cs_reloc), GFP_KERNEL);
+	p->relocs = kcalloc(p->nrelocs, sizeof(struct radeon_bo_list), GFP_KERNEL);
 	if (p->relocs == NULL) {
 		return -ENOMEM;
 	}
@@ -100,31 +97,17 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 
 	for (i = 0; i < p->nrelocs; i++) {
 		struct drm_radeon_cs_reloc *r;
+		struct drm_gem_object *gobj;
 		unsigned priority;
 
-		duplicate = false;
 		r = (struct drm_radeon_cs_reloc *)&chunk->kdata[i*4];
-		for (j = 0; j < i; j++) {
-			if (r->handle == p->relocs[j].handle) {
-				p->relocs_ptr[i] = &p->relocs[j];
-				duplicate = true;
-				break;
-			}
-		}
-		if (duplicate) {
-			p->relocs[i].handle = 0;
-			continue;
-		}
-
-		p->relocs[i].gobj = drm_gem_object_lookup(ddev, p->filp,
-							  r->handle);
-		if (p->relocs[i].gobj == NULL) {
+		gobj = drm_gem_object_lookup(ddev, p->filp, r->handle);
+		if (gobj == NULL) {
 			DRM_ERROR("gem object lookup failed 0x%x\n",
 				  r->handle);
 			return -ENOENT;
 		}
-		p->relocs_ptr[i] = &p->relocs[i];
-		p->relocs[i].robj = gem_to_radeon_bo(p->relocs[i].gobj);
+		p->relocs[i].robj = gem_to_radeon_bo(gobj);
 
 		/* The userspace buffer priorities are from 0 to 15. A higher
 		 * number means the buffer is more important.
@@ -136,10 +119,13 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 			   + !!r->write_domain;
 
 		/* the first reloc of an UVD job is the msg and that must be in
-		   VRAM, also but everything into VRAM on AGP cards to avoid
-		   image corruptions */
+		   VRAM, also but everything into VRAM on AGP cards and older
+		   IGP chips to avoid image corruptions */
 		if (p->ring == R600_RING_TYPE_UVD_INDEX &&
-		    (i == 0 || drm_pci_device_is_agp(p->rdev->ddev))) {
+		    (i == 0 || drm_pci_device_is_agp(p->rdev->ddev) ||
+		     p->rdev->family == CHIP_RS780 ||
+		     p->rdev->family == CHIP_RS880)) {
+
 			/* TODO: is this still needed for NI+ ? */
 			p->relocs[i].prefered_domains =
 				RADEON_GEM_DOMAIN_VRAM;
@@ -165,8 +151,21 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 			p->relocs[i].allowed_domains = domain;
 		}
 
+		if (radeon_ttm_tt_has_userptr(p->relocs[i].robj->tbo.ttm)) {
+			uint32_t domain = p->relocs[i].prefered_domains;
+			if (!(domain & RADEON_GEM_DOMAIN_GTT)) {
+				DRM_ERROR("Only RADEON_GEM_DOMAIN_GTT is "
+					  "allowed for userptr BOs\n");
+				return -EINVAL;
+			}
+			need_mmap_lock = true;
+			domain = RADEON_GEM_DOMAIN_GTT;
+			p->relocs[i].prefered_domains = domain;
+			p->relocs[i].allowed_domains = domain;
+		}
+
 		p->relocs[i].tv.bo = &p->relocs[i].robj->tbo;
-		p->relocs[i].handle = r->handle;
+		p->relocs[i].tv.shared = !r->write_domain;
 
 		radeon_cs_buckets_add(&buckets, &p->relocs[i].tv.head,
 				      priority);
@@ -177,8 +176,15 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 	if (p->cs_flags & RADEON_CS_USE_VM)
 		p->vm_bos = radeon_vm_get_bos(p->rdev, p->ib.vm,
 					      &p->validated);
+	if (need_mmap_lock)
+		down_read(&current->mm->mmap_sem);
+
+	r = radeon_bo_list_validate(p->rdev, &p->ticket, &p->validated, p->ring);
 
-	return radeon_bo_list_validate(p->rdev, &p->ticket, &p->validated, p->ring);
+	if (need_mmap_lock)
+		up_read(&current->mm->mmap_sem);
+
+	return r;
 }
 
 static int radeon_cs_get_ring(struct radeon_cs_parser *p, u32 ring, s32 priority)
@@ -224,17 +230,21 @@ static int radeon_cs_get_ring(struct radeon_cs_parser *p, u32 ring, s32 priority
 	return 0;
 }
 
-static void radeon_cs_sync_rings(struct radeon_cs_parser *p)
+static int radeon_cs_sync_rings(struct radeon_cs_parser *p)
 {
-	int i;
+	struct radeon_bo_list *reloc;
+	int r;
 
-	for (i = 0; i < p->nrelocs; i++) {
-		if (!p->relocs[i].robj)
-			continue;
+	list_for_each_entry(reloc, &p->validated, tv.head) {
+		struct reservation_object *resv;
 
-		radeon_semaphore_sync_to(p->ib.semaphore,
-					 p->relocs[i].robj->tbo.sync_obj);
+		resv = reloc->robj->tbo.resv;
+		r = radeon_sync_resv(p->rdev, &p->ib.sync, resv,
+				     reloc->tv.shared);
+		if (r)
+			return r;
 	}
+	return 0;
 }
 
 /* XXX: note that this is called from the legacy UMS CS ioctl as well */
@@ -253,13 +263,11 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
 	INIT_LIST_HEAD(&p->validated);
 	p->idx = 0;
 	p->ib.sa_bo = NULL;
-	p->ib.semaphore = NULL;
 	p->const_ib.sa_bo = NULL;
-	p->const_ib.semaphore = NULL;
-	p->chunk_ib_idx = -1;
-	p->chunk_relocs_idx = -1;
-	p->chunk_flags_idx = -1;
-	p->chunk_const_ib_idx = -1;
+	p->chunk_ib = NULL;
+	p->chunk_relocs = NULL;
+	p->chunk_flags = NULL;
+	p->chunk_const_ib = NULL;
 	p->chunks_array = kcalloc(cs->num_chunks, sizeof(uint64_t), GFP_KERNEL);
 	if (p->chunks_array == NULL) {
 		return -ENOMEM;
@@ -286,24 +294,23 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
 			return -EFAULT;
 		}
 		p->chunks[i].length_dw = user_chunk.length_dw;
-		p->chunks[i].chunk_id = user_chunk.chunk_id;
-		if (p->chunks[i].chunk_id == RADEON_CHUNK_ID_RELOCS) {
-			p->chunk_relocs_idx = i;
+		if (user_chunk.chunk_id == RADEON_CHUNK_ID_RELOCS) {
+			p->chunk_relocs = &p->chunks[i];
 		}
-		if (p->chunks[i].chunk_id == RADEON_CHUNK_ID_IB) {
-			p->chunk_ib_idx = i;
+		if (user_chunk.chunk_id == RADEON_CHUNK_ID_IB) {
+			p->chunk_ib = &p->chunks[i];
 			/* zero length IB isn't useful */
 			if (p->chunks[i].length_dw == 0)
 				return -EINVAL;
 		}
-		if (p->chunks[i].chunk_id == RADEON_CHUNK_ID_CONST_IB) {
-			p->chunk_const_ib_idx = i;
+		if (user_chunk.chunk_id == RADEON_CHUNK_ID_CONST_IB) {
+			p->chunk_const_ib = &p->chunks[i];
 			/* zero length CONST IB isn't useful */
 			if (p->chunks[i].length_dw == 0)
 				return -EINVAL;
 		}
-		if (p->chunks[i].chunk_id == RADEON_CHUNK_ID_FLAGS) {
-			p->chunk_flags_idx = i;
+		if (user_chunk.chunk_id == RADEON_CHUNK_ID_FLAGS) {
+			p->chunk_flags = &p->chunks[i];
 			/* zero length flags aren't useful */
 			if (p->chunks[i].length_dw == 0)
 				return -EINVAL;
@@ -312,10 +319,10 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
 		size = p->chunks[i].length_dw;
 		cdata = (void __user *)(unsigned long)user_chunk.chunk_data;
 		p->chunks[i].user_ptr = cdata;
-		if (p->chunks[i].chunk_id == RADEON_CHUNK_ID_CONST_IB)
+		if (user_chunk.chunk_id == RADEON_CHUNK_ID_CONST_IB)
 			continue;
 
-		if (p->chunks[i].chunk_id == RADEON_CHUNK_ID_IB) {
+		if (user_chunk.chunk_id == RADEON_CHUNK_ID_IB) {
 			if (!p->rdev || !(p->rdev->flags & RADEON_IS_AGP))
 				continue;
 		}
@@ -328,7 +335,7 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
 		if (copy_from_user(p->chunks[i].kdata, cdata, size)) {
 			return -EFAULT;
 		}
-		if (p->chunks[i].chunk_id == RADEON_CHUNK_ID_FLAGS) {
+		if (user_chunk.chunk_id == RADEON_CHUNK_ID_FLAGS) {
 			p->cs_flags = p->chunks[i].kdata[0];
 			if (p->chunks[i].length_dw > 1)
 				ring = p->chunks[i].kdata[1];
@@ -369,8 +376,8 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
 static int cmp_size_smaller_first(void *priv, struct list_head *a,
 				  struct list_head *b)
 {
-	struct radeon_cs_reloc *la = list_entry(a, struct radeon_cs_reloc, tv.head);
-	struct radeon_cs_reloc *lb = list_entry(b, struct radeon_cs_reloc, tv.head);
+	struct radeon_bo_list *la = list_entry(a, struct radeon_bo_list, tv.head);
+	struct radeon_bo_list *lb = list_entry(b, struct radeon_bo_list, tv.head);
 
 	/* Sort A before B if A is smaller. */
 	return (int)la->robj->tbo.num_pages - (int)lb->robj->tbo.num_pages;
@@ -403,7 +410,7 @@ static void radeon_cs_parser_fini(struct radeon_cs_parser *parser, int error, bo
 
 		ttm_eu_fence_buffer_objects(&parser->ticket,
 					    &parser->validated,
-					    parser->ib.fence);
+					    &parser->ib.fence->base);
 	} else if (backoff) {
 		ttm_eu_backoff_reservation(&parser->ticket,
 					   &parser->validated);
@@ -411,14 +418,16 @@ static void radeon_cs_parser_fini(struct radeon_cs_parser *parser, int error, bo
 
 	if (parser->relocs != NULL) {
 		for (i = 0; i < parser->nrelocs; i++) {
-			if (parser->relocs[i].gobj)
-				drm_gem_object_unreference_unlocked(parser->relocs[i].gobj);
+			struct radeon_bo *bo = parser->relocs[i].robj;
+			if (bo == NULL)
+				continue;
+
+			drm_gem_object_unreference_unlocked(&bo->gem_base);
 		}
 	}
 	kfree(parser->track);
 	kfree(parser->relocs);
-	kfree(parser->relocs_ptr);
-	kfree(parser->vm_bos);
+	drm_free_large(parser->vm_bos);
 	for (i = 0; i < parser->nchunks; i++)
 		drm_free_large(parser->chunks[i].kdata);
 	kfree(parser->chunks);
@@ -432,7 +441,7 @@ static int radeon_cs_ib_chunk(struct radeon_device *rdev,
 {
 	int r;
 
-	if (parser->chunk_ib_idx == -1)
+	if (parser->chunk_ib == NULL)
 		return 0;
 
 	if (parser->cs_flags & RADEON_CS_USE_VM)
@@ -444,13 +453,19 @@ static int radeon_cs_ib_chunk(struct radeon_device *rdev,
 		return r;
 	}
 
+	r = radeon_cs_sync_rings(parser);
+	if (r) {
+		if (r != -ERESTARTSYS)
+			DRM_ERROR("Failed to sync rings: %i\n", r);
+		return r;
+	}
+
 	if (parser->ring == R600_RING_TYPE_UVD_INDEX)
 		radeon_uvd_note_usage(rdev);
 	else if ((parser->ring == TN_RING_TYPE_VCE1_INDEX) ||
 		 (parser->ring == TN_RING_TYPE_VCE2_INDEX))
 		radeon_vce_note_usage(rdev);
 
-	radeon_cs_sync_rings(parser);
 	r = radeon_ib_schedule(rdev, &parser->ib, NULL, true);
 	if (r) {
 		DRM_ERROR("Failed to schedule IB !\n");
@@ -486,10 +501,6 @@ static int radeon_bo_vm_update_pte(struct radeon_cs_parser *p,
 	for (i = 0; i < p->nrelocs; i++) {
 		struct radeon_bo *bo;
 
-		/* ignore duplicates */
-		if (p->relocs_ptr[i] != &p->relocs[i])
-			continue;
-
 		bo = p->relocs[i].robj;
 		bo_va = radeon_vm_bo_find(vm, bo);
 		if (bo_va == NULL) {
@@ -500,6 +511,8 @@ static int radeon_bo_vm_update_pte(struct radeon_cs_parser *p,
 		r = radeon_vm_bo_update(rdev, bo_va, &bo->tbo.mem);
 		if (r)
 			return r;
+
+		radeon_sync_fence(&p->ib.sync, bo_va->last_pt_update);
 	}
 
 	return radeon_vm_clear_invalids(rdev, vm);
@@ -512,7 +525,7 @@ static int radeon_cs_ib_vm_chunk(struct radeon_device *rdev,
 	struct radeon_vm *vm = &fpriv->vm;
 	int r;
 
-	if (parser->chunk_ib_idx == -1)
+	if (parser->chunk_ib == NULL)
 		return 0;
 	if ((parser->cs_flags & RADEON_CS_USE_VM) == 0)
 		return 0;
@@ -537,11 +550,16 @@ static int radeon_cs_ib_vm_chunk(struct radeon_device *rdev,
 	if (r) {
 		goto out;
 	}
-	radeon_cs_sync_rings(parser);
-	radeon_semaphore_sync_to(parser->ib.semaphore, vm->fence);
+
+	r = radeon_cs_sync_rings(parser);
+	if (r) {
+		if (r != -ERESTARTSYS)
+			DRM_ERROR("Failed to sync rings: %i\n", r);
+		goto out;
+	}
 
 	if ((rdev->family >= CHIP_TAHITI) &&
-	    (parser->chunk_const_ib_idx != -1)) {
+	    (parser->chunk_const_ib != NULL)) {
 		r = radeon_ib_schedule(rdev, &parser->ib, &parser->const_ib, true);
 	} else {
 		r = radeon_ib_schedule(rdev, &parser->ib, NULL, true);
@@ -568,7 +586,7 @@ static int radeon_cs_ib_fill(struct radeon_device *rdev, struct radeon_cs_parser
 	struct radeon_vm *vm = NULL;
 	int r;
 
-	if (parser->chunk_ib_idx == -1)
+	if (parser->chunk_ib == NULL)
 		return 0;
 
 	if (parser->cs_flags & RADEON_CS_USE_VM) {
@@ -576,8 +594,8 @@ static int radeon_cs_ib_fill(struct radeon_device *rdev, struct radeon_cs_parser
 		vm = &fpriv->vm;
 
 		if ((rdev->family >= CHIP_TAHITI) &&
-		    (parser->chunk_const_ib_idx != -1)) {
-			ib_chunk = &parser->chunks[parser->chunk_const_ib_idx];
+		    (parser->chunk_const_ib != NULL)) {
+			ib_chunk = parser->chunk_const_ib;
 			if (ib_chunk->length_dw > RADEON_IB_VM_MAX_SIZE) {
 				DRM_ERROR("cs IB CONST too big: %d\n", ib_chunk->length_dw);
 				return -EINVAL;
@@ -596,13 +614,13 @@ static int radeon_cs_ib_fill(struct radeon_device *rdev, struct radeon_cs_parser
 				return -EFAULT;
 		}
 
-		ib_chunk = &parser->chunks[parser->chunk_ib_idx];
+		ib_chunk = parser->chunk_ib;
 		if (ib_chunk->length_dw > RADEON_IB_VM_MAX_SIZE) {
 			DRM_ERROR("cs IB too big: %d\n", ib_chunk->length_dw);
 			return -EINVAL;
 		}
 	}
-	ib_chunk = &parser->chunks[parser->chunk_ib_idx];
+	ib_chunk = parser->chunk_ib;
 
 	r =  radeon_ib_get(rdev, parser->ring, &parser->ib,
 			   vm, ib_chunk->length_dw * 4);
@@ -629,6 +647,13 @@ int radeon_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 		up_read(&rdev->exclusive_lock);
 		return -EBUSY;
 	}
+	if (rdev->in_reset) {
+		up_read(&rdev->exclusive_lock);
+		r = radeon_gpu_reset(rdev);
+		if (!r)
+			r = -EAGAIN;
+		return r;
+	}
 	/* initialize parser */
 	memset(&parser, 0, sizeof(struct radeon_cs_parser));
 	parser.filp = filp;
@@ -687,7 +712,7 @@ int radeon_cs_packet_parse(struct radeon_cs_parser *p,
 			   struct radeon_cs_packet *pkt,
 			   unsigned idx)
 {
-	struct radeon_cs_chunk *ib_chunk = &p->chunks[p->chunk_ib_idx];
+	struct radeon_cs_chunk *ib_chunk = p->chunk_ib;
 	struct radeon_device *rdev = p->rdev;
 	uint32_t header;
 
@@ -781,7 +806,7 @@ void radeon_cs_dump_packet(struct radeon_cs_parser *p,
  * GPU offset using the provided start.
  **/
 int radeon_cs_packet_next_reloc(struct radeon_cs_parser *p,
-				struct radeon_cs_reloc **cs_reloc,
+				struct radeon_bo_list **cs_reloc,
 				int nomm)
 {
 	struct radeon_cs_chunk *relocs_chunk;
@@ -789,12 +814,12 @@ int radeon_cs_packet_next_reloc(struct radeon_cs_parser *p,
 	unsigned idx;
 	int r;
 
-	if (p->chunk_relocs_idx == -1) {
+	if (p->chunk_relocs == NULL) {
 		DRM_ERROR("No relocation chunk !\n");
 		return -EINVAL;
 	}
 	*cs_reloc = NULL;
-	relocs_chunk = &p->chunks[p->chunk_relocs_idx];
+	relocs_chunk = p->chunk_relocs;
 	r = radeon_cs_packet_parse(p, &p3reloc, p->idx);
 	if (r)
 		return r;
@@ -820,6 +845,6 @@ int radeon_cs_packet_next_reloc(struct radeon_cs_parser *p,
 			(u64)relocs_chunk->kdata[idx + 3] << 32;
 		(*cs_reloc)->gpu_offset |= relocs_chunk->kdata[idx + 0];
 	} else
-		*cs_reloc = p->relocs_ptr[(idx / 4)];
+		*cs_reloc = &p->relocs[(idx / 4)];
 	return 0;
 }
diff --git a/drivers/gpu/drm/radeon/radeon_cursor.c b/drivers/gpu/drm/radeon/radeon_cursor.c
index 9630e8d95fb4..45e54060ee97 100644
--- a/drivers/gpu/drm/radeon/radeon_cursor.c
+++ b/drivers/gpu/drm/radeon/radeon_cursor.c
@@ -117,106 +117,7 @@ static void radeon_show_cursor(struct drm_crtc *crtc)
 	}
 }
 
-static void radeon_set_cursor(struct drm_crtc *crtc, struct drm_gem_object *obj,
-			      uint64_t gpu_addr)
-{
-	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
-	struct radeon_device *rdev = crtc->dev->dev_private;
-
-	if (ASIC_IS_DCE4(rdev)) {
-		WREG32(EVERGREEN_CUR_SURFACE_ADDRESS_HIGH + radeon_crtc->crtc_offset,
-		       upper_32_bits(gpu_addr));
-		WREG32(EVERGREEN_CUR_SURFACE_ADDRESS + radeon_crtc->crtc_offset,
-		       gpu_addr & 0xffffffff);
-	} else if (ASIC_IS_AVIVO(rdev)) {
-		if (rdev->family >= CHIP_RV770) {
-			if (radeon_crtc->crtc_id)
-				WREG32(R700_D2CUR_SURFACE_ADDRESS_HIGH, upper_32_bits(gpu_addr));
-			else
-				WREG32(R700_D1CUR_SURFACE_ADDRESS_HIGH, upper_32_bits(gpu_addr));
-		}
-		WREG32(AVIVO_D1CUR_SURFACE_ADDRESS + radeon_crtc->crtc_offset,
-		       gpu_addr & 0xffffffff);
-	} else {
-		radeon_crtc->legacy_cursor_offset = gpu_addr - radeon_crtc->legacy_display_base_addr;
-		/* offset is from DISP(2)_BASE_ADDRESS */
-		WREG32(RADEON_CUR_OFFSET + radeon_crtc->crtc_offset, radeon_crtc->legacy_cursor_offset);
-	}
-}
-
-int radeon_crtc_cursor_set(struct drm_crtc *crtc,
-			   struct drm_file *file_priv,
-			   uint32_t handle,
-			   uint32_t width,
-			   uint32_t height)
-{
-	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
-	struct radeon_device *rdev = crtc->dev->dev_private;
-	struct drm_gem_object *obj;
-	struct radeon_bo *robj;
-	uint64_t gpu_addr;
-	int ret;
-
-	if (!handle) {
-		/* turn off cursor */
-		radeon_hide_cursor(crtc);
-		obj = NULL;
-		goto unpin;
-	}
-
-	if ((width > radeon_crtc->max_cursor_width) ||
-	    (height > radeon_crtc->max_cursor_height)) {
-		DRM_ERROR("bad cursor width or height %d x %d\n", width, height);
-		return -EINVAL;
-	}
-
-	obj = drm_gem_object_lookup(crtc->dev, file_priv, handle);
-	if (!obj) {
-		DRM_ERROR("Cannot find cursor object %x for crtc %d\n", handle, radeon_crtc->crtc_id);
-		return -ENOENT;
-	}
-
-	robj = gem_to_radeon_bo(obj);
-	ret = radeon_bo_reserve(robj, false);
-	if (unlikely(ret != 0))
-		goto fail;
-	/* Only 27 bit offset for legacy cursor */
-	ret = radeon_bo_pin_restricted(robj, RADEON_GEM_DOMAIN_VRAM,
-				       ASIC_IS_AVIVO(rdev) ? 0 : 1 << 27,
-				       &gpu_addr);
-	radeon_bo_unreserve(robj);
-	if (ret)
-		goto fail;
-
-	radeon_crtc->cursor_width = width;
-	radeon_crtc->cursor_height = height;
-
-	radeon_lock_cursor(crtc, true);
-	radeon_set_cursor(crtc, obj, gpu_addr);
-	radeon_show_cursor(crtc);
-	radeon_lock_cursor(crtc, false);
-
-unpin:
-	if (radeon_crtc->cursor_bo) {
-		robj = gem_to_radeon_bo(radeon_crtc->cursor_bo);
-		ret = radeon_bo_reserve(robj, false);
-		if (likely(ret == 0)) {
-			radeon_bo_unpin(robj);
-			radeon_bo_unreserve(robj);
-		}
-		drm_gem_object_unreference_unlocked(radeon_crtc->cursor_bo);
-	}
-
-	radeon_crtc->cursor_bo = obj;
-	return 0;
-fail:
-	drm_gem_object_unreference_unlocked(obj);
-
-	return ret;
-}
-
-int radeon_crtc_cursor_move(struct drm_crtc *crtc,
-			    int x, int y)
+static int radeon_cursor_move_locked(struct drm_crtc *crtc, int x, int y)
 {
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
 	struct radeon_device *rdev = crtc->dev->dev_private;
@@ -281,7 +182,6 @@ int radeon_crtc_cursor_move(struct drm_crtc *crtc,
 		}
 	}
 
-	radeon_lock_cursor(crtc, true);
 	if (ASIC_IS_DCE4(rdev)) {
 		WREG32(EVERGREEN_CUR_POSITION + radeon_crtc->crtc_offset, (x << 16) | y);
 		WREG32(EVERGREEN_CUR_HOT_SPOT + radeon_crtc->crtc_offset, (xorigin << 16) | yorigin);
@@ -308,7 +208,173 @@ int radeon_crtc_cursor_move(struct drm_crtc *crtc,
 		WREG32(RADEON_CUR_OFFSET + radeon_crtc->crtc_offset, (radeon_crtc->legacy_cursor_offset +
 								      (yorigin * 256)));
 	}
+
+	radeon_crtc->cursor_x = x;
+	radeon_crtc->cursor_y = y;
+
+	return 0;
+}
+
+int radeon_crtc_cursor_move(struct drm_crtc *crtc,
+			    int x, int y)
+{
+	int ret;
+
+	radeon_lock_cursor(crtc, true);
+	ret = radeon_cursor_move_locked(crtc, x, y);
 	radeon_lock_cursor(crtc, false);
 
+	return ret;
+}
+
+static int radeon_set_cursor(struct drm_crtc *crtc, struct drm_gem_object *obj)
+{
+	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
+	struct radeon_device *rdev = crtc->dev->dev_private;
+	struct radeon_bo *robj = gem_to_radeon_bo(obj);
+	uint64_t gpu_addr;
+	int ret;
+
+	ret = radeon_bo_reserve(robj, false);
+	if (unlikely(ret != 0))
+		goto fail;
+	/* Only 27 bit offset for legacy cursor */
+	ret = radeon_bo_pin_restricted(robj, RADEON_GEM_DOMAIN_VRAM,
+				       ASIC_IS_AVIVO(rdev) ? 0 : 1 << 27,
+				       &gpu_addr);
+	radeon_bo_unreserve(robj);
+	if (ret)
+		goto fail;
+
+	if (ASIC_IS_DCE4(rdev)) {
+		WREG32(EVERGREEN_CUR_SURFACE_ADDRESS_HIGH + radeon_crtc->crtc_offset,
+		       upper_32_bits(gpu_addr));
+		WREG32(EVERGREEN_CUR_SURFACE_ADDRESS + radeon_crtc->crtc_offset,
+		       gpu_addr & 0xffffffff);
+	} else if (ASIC_IS_AVIVO(rdev)) {
+		if (rdev->family >= CHIP_RV770) {
+			if (radeon_crtc->crtc_id)
+				WREG32(R700_D2CUR_SURFACE_ADDRESS_HIGH, upper_32_bits(gpu_addr));
+			else
+				WREG32(R700_D1CUR_SURFACE_ADDRESS_HIGH, upper_32_bits(gpu_addr));
+		}
+		WREG32(AVIVO_D1CUR_SURFACE_ADDRESS + radeon_crtc->crtc_offset,
+		       gpu_addr & 0xffffffff);
+	} else {
+		radeon_crtc->legacy_cursor_offset = gpu_addr - radeon_crtc->legacy_display_base_addr;
+		/* offset is from DISP(2)_BASE_ADDRESS */
+		WREG32(RADEON_CUR_OFFSET + radeon_crtc->crtc_offset, radeon_crtc->legacy_cursor_offset);
+	}
+
 	return 0;
+
+fail:
+	drm_gem_object_unreference_unlocked(obj);
+
+	return ret;
+}
+
+int radeon_crtc_cursor_set2(struct drm_crtc *crtc,
+			    struct drm_file *file_priv,
+			    uint32_t handle,
+			    uint32_t width,
+			    uint32_t height,
+			    int32_t hot_x,
+			    int32_t hot_y)
+{
+	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
+	struct drm_gem_object *obj;
+	int ret;
+
+	if (!handle) {
+		/* turn off cursor */
+		radeon_hide_cursor(crtc);
+		obj = NULL;
+		goto unpin;
+	}
+
+	if ((width > radeon_crtc->max_cursor_width) ||
+	    (height > radeon_crtc->max_cursor_height)) {
+		DRM_ERROR("bad cursor width or height %d x %d\n", width, height);
+		return -EINVAL;
+	}
+
+	obj = drm_gem_object_lookup(crtc->dev, file_priv, handle);
+	if (!obj) {
+		DRM_ERROR("Cannot find cursor object %x for crtc %d\n", handle, radeon_crtc->crtc_id);
+		return -ENOENT;
+	}
+
+	radeon_crtc->cursor_width = width;
+	radeon_crtc->cursor_height = height;
+
+	radeon_lock_cursor(crtc, true);
+
+	if (hot_x != radeon_crtc->cursor_hot_x ||
+	    hot_y != radeon_crtc->cursor_hot_y) {
+		int x, y;
+
+		x = radeon_crtc->cursor_x + radeon_crtc->cursor_hot_x - hot_x;
+		y = radeon_crtc->cursor_y + radeon_crtc->cursor_hot_y - hot_y;
+
+		radeon_cursor_move_locked(crtc, x, y);
+
+		radeon_crtc->cursor_hot_x = hot_x;
+		radeon_crtc->cursor_hot_y = hot_y;
+	}
+
+	ret = radeon_set_cursor(crtc, obj);
+
+	if (ret)
+		DRM_ERROR("radeon_set_cursor returned %d, not changing cursor\n",
+			  ret);
+	else
+		radeon_show_cursor(crtc);
+
+	radeon_lock_cursor(crtc, false);
+
+unpin:
+	if (radeon_crtc->cursor_bo) {
+		struct radeon_bo *robj = gem_to_radeon_bo(radeon_crtc->cursor_bo);
+		ret = radeon_bo_reserve(robj, false);
+		if (likely(ret == 0)) {
+			radeon_bo_unpin(robj);
+			radeon_bo_unreserve(robj);
+		}
+		if (radeon_crtc->cursor_bo != obj)
+			drm_gem_object_unreference_unlocked(radeon_crtc->cursor_bo);
+	}
+
+	radeon_crtc->cursor_bo = obj;
+	return 0;
+}
+
+/**
+ * radeon_cursor_reset - Re-set the current cursor, if any.
+ *
+ * @crtc: drm crtc
+ *
+ * If the CRTC passed in currently has a cursor assigned, this function
+ * makes sure it's visible.
+ */
+void radeon_cursor_reset(struct drm_crtc *crtc)
+{
+	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
+	int ret;
+
+	if (radeon_crtc->cursor_bo) {
+		radeon_lock_cursor(crtc, true);
+
+		radeon_cursor_move_locked(crtc, radeon_crtc->cursor_x,
+					  radeon_crtc->cursor_y);
+
+		ret = radeon_set_cursor(crtc, radeon_crtc->cursor_bo);
+		if (ret)
+			DRM_ERROR("radeon_set_cursor returned %d, not showing "
+				  "cursor\n", ret);
+		else
+			radeon_show_cursor(crtc);
+
+		radeon_lock_cursor(crtc, false);
+	}
 }
diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c
index 12c8329644c4..bd7519fdd3f4 100644
--- a/drivers/gpu/drm/radeon/radeon_device.c
+++ b/drivers/gpu/drm/radeon/radeon_device.c
@@ -377,6 +377,37 @@ void radeon_doorbell_free(struct radeon_device *rdev, u32 doorbell)
 		__clear_bit(doorbell, rdev->doorbell.used);
 }
 
+/**
+ * radeon_doorbell_get_kfd_info - Report doorbell configuration required to
+ *                                setup KFD
+ *
+ * @rdev: radeon_device pointer
+ * @aperture_base: output returning doorbell aperture base physical address
+ * @aperture_size: output returning doorbell aperture size in bytes
+ * @start_offset: output returning # of doorbell bytes reserved for radeon.
+ *
+ * Radeon and the KFD share the doorbell aperture. Radeon sets it up,
+ * takes doorbells required for its own rings and reports the setup to KFD.
+ * Radeon reserved doorbells are at the start of the doorbell aperture.
+ */
+void radeon_doorbell_get_kfd_info(struct radeon_device *rdev,
+				  phys_addr_t *aperture_base,
+				  size_t *aperture_size,
+				  size_t *start_offset)
+{
+	/* The first num_doorbells are used by radeon.
+	 * KFD takes whatever's left in the aperture. */
+	if (rdev->doorbell.size > rdev->doorbell.num_doorbells * sizeof(u32)) {
+		*aperture_base = rdev->doorbell.base;
+		*aperture_size = rdev->doorbell.size;
+		*start_offset = rdev->doorbell.num_doorbells * sizeof(u32);
+	} else {
+		*aperture_base = 0;
+		*aperture_size = 0;
+		*start_offset = 0;
+	}
+}
+
 /*
  * radeon_wb_*()
  * Writeback is the the method by which the the GPU updates special pages
@@ -434,7 +465,7 @@ int radeon_wb_init(struct radeon_device *rdev)
 
 	if (rdev->wb.wb_obj == NULL) {
 		r = radeon_bo_create(rdev, RADEON_GPU_PAGE_SIZE, PAGE_SIZE, true,
-				     RADEON_GEM_DOMAIN_GTT, 0, NULL,
+				     RADEON_GEM_DOMAIN_GTT, 0, NULL, NULL,
 				     &rdev->wb.wb_obj);
 		if (r) {
 			dev_warn(rdev->dev, "(%d) create WB bo failed\n", r);
@@ -743,6 +774,8 @@ int radeon_dummy_page_init(struct radeon_device *rdev)
 		rdev->dummy_page.page = NULL;
 		return -ENOMEM;
 	}
+	rdev->dummy_page.entry = radeon_gart_get_page_entry(rdev->dummy_page.addr,
+							    RADEON_GART_PAGE_DUMMY);
 	return 0;
 }
 
@@ -952,6 +985,7 @@ int radeon_atombios_init(struct radeon_device *rdev)
 	}
 
 	mutex_init(&rdev->mode_info.atom_context->mutex);
+	mutex_init(&rdev->mode_info.atom_context->scratch_mutex);
 	radeon_atom_initialize_bios_scratch_regs(rdev->ddev);
 	atom_allocate_fb_scratch(rdev->mode_info.atom_context);
 	return 0;
@@ -1130,7 +1164,7 @@ static void radeon_check_arguments(struct radeon_device *rdev)
 	if (radeon_vm_block_size == -1) {
 
 		/* Total bits covered by PD + PTs */
-		unsigned bits = ilog2(radeon_vm_size) + 17;
+		unsigned bits = ilog2(radeon_vm_size) + 18;
 
 		/* Make sure the PD is 4K in size up to 8GB address space.
 		   Above that split equal between PD and PTs */
@@ -1257,6 +1291,7 @@ int radeon_device_init(struct radeon_device *rdev,
 	for (i = 0; i < RADEON_NUM_RINGS; i++) {
 		rdev->ring[i].idx = i;
 	}
+	rdev->fence_context = fence_context_alloc(RADEON_NUM_RINGS);
 
 	DRM_INFO("initializing kernel modesetting (%s 0x%04X:0x%04X 0x%04X:0x%04X).\n",
 		radeon_family_name[rdev->family], pdev->vendor, pdev->device,
@@ -1271,9 +1306,12 @@ int radeon_device_init(struct radeon_device *rdev,
 	mutex_init(&rdev->pm.mutex);
 	mutex_init(&rdev->gpu_clock_mutex);
 	mutex_init(&rdev->srbm_mutex);
+	mutex_init(&rdev->grbm_idx_mutex);
 	init_rwsem(&rdev->pm.mclk_lock);
 	init_rwsem(&rdev->exclusive_lock);
 	init_waitqueue_head(&rdev->irq.vblank_queue);
+	mutex_init(&rdev->mn_lock);
+	hash_init(rdev->mn_hash);
 	r = radeon_gem_init(rdev);
 	if (r)
 		return r;
@@ -1399,10 +1437,6 @@ int radeon_device_init(struct radeon_device *rdev,
 	if (r)
 		goto failed;
 
-	r = radeon_ib_ring_tests(rdev);
-	if (r)
-		DRM_ERROR("ib ring test failed (%d).\n", r);
-
 	r = radeon_gem_debugfs_init(rdev);
 	if (r) {
 		DRM_ERROR("registering gem debugfs failed (%d).\n", r);
@@ -1420,6 +1454,10 @@ int radeon_device_init(struct radeon_device *rdev,
 			goto failed;
 	}
 
+	r = radeon_ib_ring_tests(rdev);
+	if (r)
+		DRM_ERROR("ib ring test failed (%d).\n", r);
+
 	if ((radeon_testing & 1)) {
 		if (rdev->accel_working)
 			radeon_test_moves(rdev);
@@ -1497,7 +1535,6 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon)
 	struct drm_crtc *crtc;
 	struct drm_connector *connector;
 	int i, r;
-	bool force_completion = false;
 
 	if (dev == NULL || dev->dev_private == NULL) {
 		return -ENODEV;
@@ -1541,12 +1578,9 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon)
 		r = radeon_fence_wait_empty(rdev, i);
 		if (r) {
 			/* delay GPU reset to resume */
-			force_completion = true;
+			radeon_fence_driver_force_completion(rdev, i);
 		}
 	}
-	if (force_completion) {
-		radeon_fence_driver_force_completion(rdev);
-	}
 
 	radeon_save_bios_scratch_regs(rdev);
 
@@ -1686,8 +1720,6 @@ int radeon_gpu_reset(struct radeon_device *rdev)
 		return 0;
 	}
 
-	rdev->needs_reset = false;
-
 	radeon_save_bios_scratch_regs(rdev);
 	/* block TTM */
 	resched = ttm_bo_lock_delayed_workqueue(&rdev->mman.bdev);
@@ -1704,7 +1736,6 @@ int radeon_gpu_reset(struct radeon_device *rdev)
 		}
 	}
 
-retry:
 	r = radeon_asic_reset(rdev);
 	if (!r) {
 		dev_info(rdev->dev, "GPU reset succeeded, trying to resume\n");
@@ -1713,26 +1744,12 @@ retry:
 
 	radeon_restore_bios_scratch_regs(rdev);
 
-	if (!r) {
-		for (i = 0; i < RADEON_NUM_RINGS; ++i) {
+	for (i = 0; i < RADEON_NUM_RINGS; ++i) {
+		if (!r && ring_data[i]) {
 			radeon_ring_restore(rdev, &rdev->ring[i],
 					    ring_sizes[i], ring_data[i]);
-			ring_sizes[i] = 0;
-			ring_data[i] = NULL;
-		}
-
-		r = radeon_ib_ring_tests(rdev);
-		if (r) {
-			dev_err(rdev->dev, "ib ring test failed (%d).\n", r);
-			if (saved) {
-				saved = false;
-				radeon_suspend(rdev);
-				goto retry;
-			}
-		}
-	} else {
-		radeon_fence_driver_force_completion(rdev);
-		for (i = 0; i < RADEON_NUM_RINGS; ++i) {
+		} else {
+			radeon_fence_driver_force_completion(rdev, i);
 			kfree(ring_data[i]);
 		}
 	}
@@ -1764,19 +1781,32 @@ retry:
 	/* reset hpd state */
 	radeon_hpd_init(rdev);
 
+	ttm_bo_unlock_delayed_workqueue(&rdev->mman.bdev, resched);
+
+	rdev->in_reset = true;
+	rdev->needs_reset = false;
+
+	downgrade_write(&rdev->exclusive_lock);
+
 	drm_helper_resume_force_mode(rdev->ddev);
 
 	/* set the power state here in case we are a PX system or headless */
 	if ((rdev->pm.pm_method == PM_METHOD_DPM) && rdev->pm.dpm_enabled)
 		radeon_pm_compute_clocks(rdev);
 
-	ttm_bo_unlock_delayed_workqueue(&rdev->mman.bdev, resched);
-	if (r) {
+	if (!r) {
+		r = radeon_ib_ring_tests(rdev);
+		if (r && saved)
+			r = -EAGAIN;
+	} else {
 		/* bad news, how to tell it to userspace ? */
 		dev_info(rdev->dev, "GPU reset failed\n");
 	}
 
-	up_write(&rdev->exclusive_lock);
+	rdev->needs_reset = r == -EAGAIN;
+	rdev->in_reset = false;
+
+	up_read(&rdev->exclusive_lock);
 	return r;
 }
 
diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c
index 3fdf87318069..102116902a07 100644
--- a/drivers/gpu/drm/radeon/radeon_display.c
+++ b/drivers/gpu/drm/radeon/radeon_display.c
@@ -32,6 +32,7 @@
 
 #include <linux/pm_runtime.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_plane_helper.h>
 #include <drm/drm_edid.h>
 
 #include <linux/gcd.h>
@@ -402,12 +403,21 @@ static void radeon_flip_work_func(struct work_struct *__work)
 
         down_read(&rdev->exclusive_lock);
 	if (work->fence) {
-		r = radeon_fence_wait(work->fence, false);
-		if (r == -EDEADLK) {
-			up_read(&rdev->exclusive_lock);
-			r = radeon_gpu_reset(rdev);
-			down_read(&rdev->exclusive_lock);
-		}
+		struct radeon_fence *fence;
+
+		fence = to_radeon_fence(work->fence);
+		if (fence && fence->rdev == rdev) {
+			r = radeon_fence_wait(fence, false);
+			if (r == -EDEADLK) {
+				up_read(&rdev->exclusive_lock);
+				do {
+					r = radeon_gpu_reset(rdev);
+				} while (r == -EAGAIN);
+				down_read(&rdev->exclusive_lock);
+			}
+		} else
+			r = fence_wait(work->fence, false);
+
 		if (r)
 			DRM_ERROR("failed to wait on page flip fence (%d)!\n", r);
 
@@ -416,7 +426,8 @@ static void radeon_flip_work_func(struct work_struct *__work)
 		 * confused about which BO the CRTC is scanning out
 		 */
 
-		radeon_fence_unref(&work->fence);
+		fence_put(work->fence);
+		work->fence = NULL;
 	}
 
 	/* We borrow the event spin lock for protecting flip_status */
@@ -474,11 +485,6 @@ static int radeon_crtc_page_flip(struct drm_crtc *crtc,
 	obj = new_radeon_fb->obj;
 	new_rbo = gem_to_radeon_bo(obj);
 
-	spin_lock(&new_rbo->tbo.bdev->fence_lock);
-	if (new_rbo->tbo.sync_obj)
-		work->fence = radeon_fence_ref(new_rbo->tbo.sync_obj);
-	spin_unlock(&new_rbo->tbo.bdev->fence_lock);
-
 	/* pin the new buffer */
 	DRM_DEBUG_DRIVER("flip-ioctl() cur_rbo = %p, new_rbo = %p\n",
 			 work->old_rbo, new_rbo);
@@ -497,6 +503,7 @@ static int radeon_crtc_page_flip(struct drm_crtc *crtc,
 		DRM_ERROR("failed to pin new rbo buffer before flip\n");
 		goto cleanup;
 	}
+	work->fence = fence_get(reservation_object_get_excl(new_rbo->tbo.resv));
 	radeon_bo_get_tiling_flags(new_rbo, &tiling_flags, NULL);
 	radeon_bo_unreserve(new_rbo);
 
@@ -578,9 +585,8 @@ pflip_cleanup:
 
 cleanup:
 	drm_gem_object_unreference_unlocked(&work->old_rbo->gem_base);
-	radeon_fence_unref(&work->fence);
+	fence_put(work->fence);
 	kfree(work);
-
 	return r;
 }
 
@@ -629,7 +635,7 @@ radeon_crtc_set_config(struct drm_mode_set *set)
 	return ret;
 }
 static const struct drm_crtc_funcs radeon_crtc_funcs = {
-	.cursor_set = radeon_crtc_cursor_set,
+	.cursor_set2 = radeon_crtc_cursor_set2,
 	.cursor_move = radeon_crtc_cursor_move,
 	.gamma_set = radeon_crtc_gamma_set,
 	.set_config = radeon_crtc_set_config,
@@ -1917,7 +1923,7 @@ int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc, unsigned int fl
 
 	/* In vblank? */
 	if (in_vbl)
-		ret |= DRM_SCANOUTPOS_INVBL;
+		ret |= DRM_SCANOUTPOS_IN_VBLANK;
 
 	/* Is vpos outside nominal vblank area, but less than
 	 * 1/100 of a frame height away from start of vblank?
diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c
index f9d17b29b343..4f50fb0e3d93 100644
--- a/drivers/gpu/drm/radeon/radeon_drv.c
+++ b/drivers/gpu/drm/radeon/radeon_drv.c
@@ -38,7 +38,11 @@
 #include <linux/module.h>
 #include <linux/pm_runtime.h>
 #include <linux/vga_switcheroo.h>
+#include <drm/drm_gem.h>
+
 #include "drm_crtc_helper.h"
+#include "radeon_kfd.h"
+
 /*
  * KMS wrapper.
  * - 2.0.0 - initial interface
@@ -114,6 +118,9 @@ int radeon_gem_object_open(struct drm_gem_object *obj,
 				struct drm_file *file_priv);
 void radeon_gem_object_close(struct drm_gem_object *obj,
 				struct drm_file *file_priv);
+struct dma_buf *radeon_gem_prime_export(struct drm_device *dev,
+					struct drm_gem_object *gobj,
+					int flags);
 extern int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc,
 				      unsigned int flags,
 				      int *vpos, int *hpos, ktime_t *stime,
@@ -130,7 +137,7 @@ int radeon_mode_dumb_create(struct drm_file *file_priv,
 			    struct drm_mode_create_dumb *args);
 struct sg_table *radeon_gem_prime_get_sg_table(struct drm_gem_object *obj);
 struct drm_gem_object *radeon_gem_prime_import_sg_table(struct drm_device *dev,
-							size_t size,
+							struct dma_buf_attachment *,
 							struct sg_table *sg);
 int radeon_gem_prime_pin(struct drm_gem_object *obj);
 void radeon_gem_prime_unpin(struct drm_gem_object *obj);
@@ -309,7 +316,7 @@ static const struct file_operations radeon_driver_old_fops = {
 	.open = drm_open,
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_mmap,
+	.mmap = drm_legacy_mmap,
 	.poll = drm_poll,
 	.read = drm_read,
 #ifdef CONFIG_COMPAT
@@ -329,6 +336,7 @@ static struct drm_driver driver_old = {
 	.preclose = radeon_driver_preclose,
 	.postclose = radeon_driver_postclose,
 	.lastclose = radeon_driver_lastclose,
+	.set_busid = drm_pci_set_busid,
 	.unload = radeon_driver_unload,
 	.suspend = radeon_suspend,
 	.resume = radeon_resume,
@@ -553,6 +561,7 @@ static struct drm_driver kms_driver = {
 	.preclose = radeon_driver_preclose_kms,
 	.postclose = radeon_driver_postclose_kms,
 	.lastclose = radeon_driver_lastclose_kms,
+	.set_busid = drm_pci_set_busid,
 	.unload = radeon_driver_unload_kms,
 	.get_vblank_counter = radeon_get_vblank_counter_kms,
 	.enable_vblank = radeon_enable_vblank_kms,
@@ -578,7 +587,7 @@ static struct drm_driver kms_driver = {
 
 	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
 	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
-	.gem_prime_export = drm_gem_prime_export,
+	.gem_prime_export = radeon_gem_prime_export,
 	.gem_prime_import = drm_gem_prime_import,
 	.gem_prime_pin = radeon_gem_prime_pin,
 	.gem_prime_unpin = radeon_gem_prime_unpin,
@@ -647,12 +656,15 @@ static int __init radeon_init(void)
 #endif
 	}
 
+	radeon_kfd_init();
+
 	/* let modprobe override vga console setting */
 	return drm_pci_init(driver, pdriver);
 }
 
 static void __exit radeon_exit(void)
 {
+	radeon_kfd_fini();
 	drm_pci_exit(driver, pdriver);
 	radeon_unregister_atpx_handler();
 }
diff --git a/drivers/gpu/drm/radeon/radeon_drv.h b/drivers/gpu/drm/radeon/radeon_drv.h
index dafd812e4571..46bd3938282c 100644
--- a/drivers/gpu/drm/radeon/radeon_drv.h
+++ b/drivers/gpu/drm/radeon/radeon_drv.h
@@ -33,7 +33,9 @@
 
 #include <linux/firmware.h>
 #include <linux/platform_device.h>
+#include <drm/drm_legacy.h>
 
+#include <drm/ati_pcigart.h>
 #include "radeon_family.h"
 
 /* General customization:
diff --git a/drivers/gpu/drm/radeon/radeon_encoders.c b/drivers/gpu/drm/radeon/radeon_encoders.c
index 15edf23b465c..6b670b0bc47b 100644
--- a/drivers/gpu/drm/radeon/radeon_encoders.c
+++ b/drivers/gpu/drm/radeon/radeon_encoders.c
@@ -179,6 +179,9 @@ static void radeon_encoder_add_backlight(struct radeon_encoder *radeon_encoder,
 		    (rdev->pdev->subsystem_vendor == 0x1734) &&
 		    (rdev->pdev->subsystem_device == 0x1107))
 			use_bl = false;
+		/* disable native backlight control on older asics */
+		else if (rdev->family < CHIP_R600)
+			use_bl = false;
 		else
 			use_bl = true;
 	}
@@ -410,3 +413,24 @@ bool radeon_dig_monitor_is_duallink(struct drm_encoder *encoder,
 	}
 }
 
+bool radeon_encoder_is_digital(struct drm_encoder *encoder)
+{
+	struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
+	switch (radeon_encoder->encoder_id) {
+	case ENCODER_OBJECT_ID_INTERNAL_LVDS:
+	case ENCODER_OBJECT_ID_INTERNAL_TMDS1:
+	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_TMDS1:
+	case ENCODER_OBJECT_ID_INTERNAL_LVTM1:
+	case ENCODER_OBJECT_ID_INTERNAL_DVO1:
+	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1:
+	case ENCODER_OBJECT_ID_INTERNAL_DDI:
+	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY:
+	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_LVTMA:
+	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1:
+	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2:
+	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY3:
+		return true;
+	default:
+		return false;
+	}
+}
diff --git a/drivers/gpu/drm/radeon/radeon_fb.c b/drivers/gpu/drm/radeon/radeon_fb.c
index 94b0f2aa3d7c..29b9220ec399 100644
--- a/drivers/gpu/drm/radeon/radeon_fb.c
+++ b/drivers/gpu/drm/radeon/radeon_fb.c
@@ -48,10 +48,40 @@ struct radeon_fbdev {
 	struct radeon_device *rdev;
 };
 
+/**
+ * radeon_fb_helper_set_par - Hide cursor on CRTCs used by fbdev.
+ *
+ * @info: fbdev info
+ *
+ * This function hides the cursor on all CRTCs used by fbdev.
+ */
+static int radeon_fb_helper_set_par(struct fb_info *info)
+{
+	int ret;
+
+	ret = drm_fb_helper_set_par(info);
+
+	/* XXX: with universal plane support fbdev will automatically disable
+	 * all non-primary planes (including the cursor)
+	 */
+	if (ret == 0) {
+		struct drm_fb_helper *fb_helper = info->par;
+		int i;
+
+		for (i = 0; i < fb_helper->crtc_count; i++) {
+			struct drm_crtc *crtc = fb_helper->crtc_info[i].mode_set.crtc;
+
+			radeon_crtc_cursor_set2(crtc, NULL, 0, 0, 0, 0, 0);
+		}
+	}
+
+	return ret;
+}
+
 static struct fb_ops radeonfb_ops = {
 	.owner = THIS_MODULE,
 	.fb_check_var = drm_fb_helper_check_var,
-	.fb_set_par = drm_fb_helper_set_par,
+	.fb_set_par = radeon_fb_helper_set_par,
 	.fb_fillrect = cfb_fillrect,
 	.fb_copyarea = cfb_copyarea,
 	.fb_imageblit = cfb_imageblit,
@@ -189,7 +219,8 @@ out_unref:
 static int radeonfb_create(struct drm_fb_helper *helper,
 			   struct drm_fb_helper_surface_size *sizes)
 {
-	struct radeon_fbdev *rfbdev = (struct radeon_fbdev *)helper;
+	struct radeon_fbdev *rfbdev =
+		container_of(helper, struct radeon_fbdev, helper);
 	struct radeon_device *rdev = rfbdev->rdev;
 	struct fb_info *info;
 	struct drm_framebuffer *fb = NULL;
diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c
index 913787085dfa..d13d1b5a859f 100644
--- a/drivers/gpu/drm/radeon/radeon_fence.c
+++ b/drivers/gpu/drm/radeon/radeon_fence.c
@@ -98,6 +98,25 @@ static u32 radeon_fence_read(struct radeon_device *rdev, int ring)
 }
 
 /**
+ * radeon_fence_schedule_check - schedule lockup check
+ *
+ * @rdev: radeon_device pointer
+ * @ring: ring index we should work with
+ *
+ * Queues a delayed work item to check for lockups.
+ */
+static void radeon_fence_schedule_check(struct radeon_device *rdev, int ring)
+{
+	/*
+	 * Do not reset the timer here with mod_delayed_work,
+	 * this can livelock in an interaction with TTM delayed destroy.
+	 */
+	queue_delayed_work(system_power_efficient_wq,
+			   &rdev->fence_drv[ring].lockup_work,
+			   RADEON_FENCE_JIFFIES_TIMEOUT);
+}
+
+/**
  * radeon_fence_emit - emit a fence on the requested ring
  *
  * @rdev: radeon_device pointer
@@ -111,30 +130,71 @@ int radeon_fence_emit(struct radeon_device *rdev,
 		      struct radeon_fence **fence,
 		      int ring)
 {
+	u64 seq = ++rdev->fence_drv[ring].sync_seq[ring];
+
 	/* we are protected by the ring emission mutex */
 	*fence = kmalloc(sizeof(struct radeon_fence), GFP_KERNEL);
 	if ((*fence) == NULL) {
 		return -ENOMEM;
 	}
-	kref_init(&((*fence)->kref));
 	(*fence)->rdev = rdev;
-	(*fence)->seq = ++rdev->fence_drv[ring].sync_seq[ring];
+	(*fence)->seq = seq;
 	(*fence)->ring = ring;
+	(*fence)->is_vm_update = false;
+	fence_init(&(*fence)->base, &radeon_fence_ops,
+		   &rdev->fence_queue.lock, rdev->fence_context + ring, seq);
 	radeon_fence_ring_emit(rdev, ring, *fence);
 	trace_radeon_fence_emit(rdev->ddev, ring, (*fence)->seq);
+	radeon_fence_schedule_check(rdev, ring);
 	return 0;
 }
 
 /**
- * radeon_fence_process - process a fence
+ * radeon_fence_check_signaled - callback from fence_queue
+ *
+ * this function is called with fence_queue lock held, which is also used
+ * for the fence locking itself, so unlocked variants are used for
+ * fence_signal, and remove_wait_queue.
+ */
+static int radeon_fence_check_signaled(wait_queue_t *wait, unsigned mode, int flags, void *key)
+{
+	struct radeon_fence *fence;
+	u64 seq;
+
+	fence = container_of(wait, struct radeon_fence, fence_wake);
+
+	/*
+	 * We cannot use radeon_fence_process here because we're already
+	 * in the waitqueue, in a call from wake_up_all.
+	 */
+	seq = atomic64_read(&fence->rdev->fence_drv[fence->ring].last_seq);
+	if (seq >= fence->seq) {
+		int ret = fence_signal_locked(&fence->base);
+
+		if (!ret)
+			FENCE_TRACE(&fence->base, "signaled from irq context\n");
+		else
+			FENCE_TRACE(&fence->base, "was already signaled\n");
+
+		radeon_irq_kms_sw_irq_put(fence->rdev, fence->ring);
+		__remove_wait_queue(&fence->rdev->fence_queue, &fence->fence_wake);
+		fence_put(&fence->base);
+	} else
+		FENCE_TRACE(&fence->base, "pending\n");
+	return 0;
+}
+
+/**
+ * radeon_fence_activity - check for fence activity
  *
  * @rdev: radeon_device pointer
  * @ring: ring index the fence is associated with
  *
- * Checks the current fence value and wakes the fence queue
- * if the sequence number has increased (all asics).
+ * Checks the current fence value and calculates the last
+ * signalled fence value. Returns true if activity occured
+ * on the ring, and the fence_queue should be waken up.
  */
-void radeon_fence_process(struct radeon_device *rdev, int ring)
+static bool radeon_fence_activity(struct radeon_device *rdev, int ring)
 {
 	uint64_t seq, last_seq, last_emitted;
 	unsigned count_loop = 0;
@@ -190,23 +250,77 @@ void radeon_fence_process(struct radeon_device *rdev, int ring)
 		}
 	} while (atomic64_xchg(&rdev->fence_drv[ring].last_seq, seq) > seq);
 
-	if (wake)
-		wake_up_all(&rdev->fence_queue);
+	if (seq < last_emitted)
+		radeon_fence_schedule_check(rdev, ring);
+
+	return wake;
 }
 
 /**
- * radeon_fence_destroy - destroy a fence
+ * radeon_fence_check_lockup - check for hardware lockup
  *
- * @kref: fence kref
+ * @work: delayed work item
  *
- * Frees the fence object (all asics).
+ * Checks for fence activity and if there is none probe
+ * the hardware if a lockup occured.
  */
-static void radeon_fence_destroy(struct kref *kref)
+static void radeon_fence_check_lockup(struct work_struct *work)
 {
-	struct radeon_fence *fence;
+	struct radeon_fence_driver *fence_drv;
+	struct radeon_device *rdev;
+	int ring;
+
+	fence_drv = container_of(work, struct radeon_fence_driver,
+				 lockup_work.work);
+	rdev = fence_drv->rdev;
+	ring = fence_drv - &rdev->fence_drv[0];
+
+	if (!down_read_trylock(&rdev->exclusive_lock)) {
+		/* just reschedule the check if a reset is going on */
+		radeon_fence_schedule_check(rdev, ring);
+		return;
+	}
+
+	if (fence_drv->delayed_irq && rdev->ddev->irq_enabled) {
+		unsigned long irqflags;
+
+		fence_drv->delayed_irq = false;
+		spin_lock_irqsave(&rdev->irq.lock, irqflags);
+		radeon_irq_set(rdev);
+		spin_unlock_irqrestore(&rdev->irq.lock, irqflags);
+	}
+
+	if (radeon_fence_activity(rdev, ring))
+		wake_up_all(&rdev->fence_queue);
 
-	fence = container_of(kref, struct radeon_fence, kref);
-	kfree(fence);
+	else if (radeon_ring_is_lockup(rdev, ring, &rdev->ring[ring])) {
+
+		/* good news we believe it's a lockup */
+		dev_warn(rdev->dev, "GPU lockup (current fence id "
+			 "0x%016llx last fence id 0x%016llx on ring %d)\n",
+			 (uint64_t)atomic64_read(&fence_drv->last_seq),
+			 fence_drv->sync_seq[ring], ring);
+
+		/* remember that we need an reset */
+		rdev->needs_reset = true;
+		wake_up_all(&rdev->fence_queue);
+	}
+	up_read(&rdev->exclusive_lock);
+}
+
+/**
+ * radeon_fence_process - process a fence
+ *
+ * @rdev: radeon_device pointer
+ * @ring: ring index the fence is associated with
+ *
+ * Checks the current fence value and wakes the fence queue
+ * if the sequence number has increased (all asics).
+ */
+void radeon_fence_process(struct radeon_device *rdev, int ring)
+{
+	if (radeon_fence_activity(rdev, ring))
+		wake_up_all(&rdev->fence_queue);
 }
 
 /**
@@ -237,6 +351,75 @@ static bool radeon_fence_seq_signaled(struct radeon_device *rdev,
 	return false;
 }
 
+static bool radeon_fence_is_signaled(struct fence *f)
+{
+	struct radeon_fence *fence = to_radeon_fence(f);
+	struct radeon_device *rdev = fence->rdev;
+	unsigned ring = fence->ring;
+	u64 seq = fence->seq;
+
+	if (atomic64_read(&rdev->fence_drv[ring].last_seq) >= seq) {
+		return true;
+	}
+
+	if (down_read_trylock(&rdev->exclusive_lock)) {
+		radeon_fence_process(rdev, ring);
+		up_read(&rdev->exclusive_lock);
+
+		if (atomic64_read(&rdev->fence_drv[ring].last_seq) >= seq) {
+			return true;
+		}
+	}
+	return false;
+}
+
+/**
+ * radeon_fence_enable_signaling - enable signalling on fence
+ * @fence: fence
+ *
+ * This function is called with fence_queue lock held, and adds a callback
+ * to fence_queue that checks if this fence is signaled, and if so it
+ * signals the fence and removes itself.
+ */
+static bool radeon_fence_enable_signaling(struct fence *f)
+{
+	struct radeon_fence *fence = to_radeon_fence(f);
+	struct radeon_device *rdev = fence->rdev;
+
+	if (atomic64_read(&rdev->fence_drv[fence->ring].last_seq) >= fence->seq)
+		return false;
+
+	if (down_read_trylock(&rdev->exclusive_lock)) {
+		radeon_irq_kms_sw_irq_get(rdev, fence->ring);
+
+		if (radeon_fence_activity(rdev, fence->ring))
+			wake_up_all_locked(&rdev->fence_queue);
+
+		/* did fence get signaled after we enabled the sw irq? */
+		if (atomic64_read(&rdev->fence_drv[fence->ring].last_seq) >= fence->seq) {
+			radeon_irq_kms_sw_irq_put(rdev, fence->ring);
+			up_read(&rdev->exclusive_lock);
+			return false;
+		}
+
+		up_read(&rdev->exclusive_lock);
+	} else {
+		/* we're probably in a lockup, lets not fiddle too much */
+		if (radeon_irq_kms_sw_irq_get_delayed(rdev, fence->ring))
+			rdev->fence_drv[fence->ring].delayed_irq = true;
+		radeon_fence_schedule_check(rdev, fence->ring);
+	}
+
+	fence->fence_wake.flags = 0;
+	fence->fence_wake.private = NULL;
+	fence->fence_wake.func = radeon_fence_check_signaled;
+	__add_wait_queue(&rdev->fence_queue, &fence->fence_wake);
+	fence_get(f);
+
+	FENCE_TRACE(&fence->base, "armed on ring %i!\n", fence->ring);
+	return true;
+}
+
 /**
  * radeon_fence_signaled - check if a fence has signaled
  *
@@ -247,14 +430,15 @@ static bool radeon_fence_seq_signaled(struct radeon_device *rdev,
  */
 bool radeon_fence_signaled(struct radeon_fence *fence)
 {
-	if (!fence) {
+	if (!fence)
 		return true;
-	}
-	if (fence->seq == RADEON_FENCE_SIGNALED_SEQ) {
-		return true;
-	}
+
 	if (radeon_fence_seq_signaled(fence->rdev, fence->seq, fence->ring)) {
-		fence->seq = RADEON_FENCE_SIGNALED_SEQ;
+		int ret;
+
+		ret = fence_signal(&fence->base);
+		if (!ret)
+			FENCE_TRACE(&fence->base, "signaled from radeon_fence_signaled\n");
 		return true;
 	}
 	return false;
@@ -283,110 +467,70 @@ static bool radeon_fence_any_seq_signaled(struct radeon_device *rdev, u64 *seq)
 }
 
 /**
- * radeon_fence_wait_seq - wait for a specific sequence numbers
+ * radeon_fence_wait_seq_timeout - wait for a specific sequence numbers
  *
  * @rdev: radeon device pointer
  * @target_seq: sequence number(s) we want to wait for
  * @intr: use interruptable sleep
+ * @timeout: maximum time to wait, or MAX_SCHEDULE_TIMEOUT for infinite wait
  *
  * Wait for the requested sequence number(s) to be written by any ring
  * (all asics).  Sequnce number array is indexed by ring id.
  * @intr selects whether to use interruptable (true) or non-interruptable
  * (false) sleep when waiting for the sequence number.  Helper function
  * for radeon_fence_wait_*().
- * Returns 0 if the sequence number has passed, error for all other cases.
+ * Returns remaining time if the sequence number has passed, 0 when
+ * the wait timeout, or an error for all other cases.
  * -EDEADLK is returned when a GPU lockup has been detected.
  */
-static int radeon_fence_wait_seq(struct radeon_device *rdev, u64 *target_seq,
-				 bool intr)
+static long radeon_fence_wait_seq_timeout(struct radeon_device *rdev,
+					  u64 *target_seq, bool intr,
+					  long timeout)
 {
-	uint64_t last_seq[RADEON_NUM_RINGS];
-	bool signaled;
-	int i, r;
-
-	while (!radeon_fence_any_seq_signaled(rdev, target_seq)) {
+	long r;
+	int i;
 
-		/* Save current sequence values, used to check for GPU lockups */
-		for (i = 0; i < RADEON_NUM_RINGS; ++i) {
-			if (!target_seq[i])
-				continue;
+	if (radeon_fence_any_seq_signaled(rdev, target_seq))
+		return timeout;
 
-			last_seq[i] = atomic64_read(&rdev->fence_drv[i].last_seq);
-			trace_radeon_fence_wait_begin(rdev->ddev, i, target_seq[i]);
-			radeon_irq_kms_sw_irq_get(rdev, i);
-		}
+	/* enable IRQs and tracing */
+	for (i = 0; i < RADEON_NUM_RINGS; ++i) {
+		if (!target_seq[i])
+			continue;
 
-		if (intr) {
-			r = wait_event_interruptible_timeout(rdev->fence_queue, (
-				(signaled = radeon_fence_any_seq_signaled(rdev, target_seq))
-				 || rdev->needs_reset), RADEON_FENCE_JIFFIES_TIMEOUT);
-		} else {
-			r = wait_event_timeout(rdev->fence_queue, (
-				(signaled = radeon_fence_any_seq_signaled(rdev, target_seq))
-				 || rdev->needs_reset), RADEON_FENCE_JIFFIES_TIMEOUT);
-		}
+		trace_radeon_fence_wait_begin(rdev->ddev, i, target_seq[i]);
+		radeon_irq_kms_sw_irq_get(rdev, i);
+	}
 
-		for (i = 0; i < RADEON_NUM_RINGS; ++i) {
-			if (!target_seq[i])
-				continue;
+	if (intr) {
+		r = wait_event_interruptible_timeout(rdev->fence_queue, (
+			radeon_fence_any_seq_signaled(rdev, target_seq)
+			 || rdev->needs_reset), timeout);
+	} else {
+		r = wait_event_timeout(rdev->fence_queue, (
+			radeon_fence_any_seq_signaled(rdev, target_seq)
+			 || rdev->needs_reset), timeout);
+	}
 
-			radeon_irq_kms_sw_irq_put(rdev, i);
-			trace_radeon_fence_wait_end(rdev->ddev, i, target_seq[i]);
-		}
+	if (rdev->needs_reset)
+		r = -EDEADLK;
 
-		if (unlikely(r < 0))
-			return r;
+	for (i = 0; i < RADEON_NUM_RINGS; ++i) {
+		if (!target_seq[i])
+			continue;
 
-		if (unlikely(!signaled)) {
-			if (rdev->needs_reset)
-				return -EDEADLK;
-
-			/* we were interrupted for some reason and fence
-			 * isn't signaled yet, resume waiting */
-			if (r)
-				continue;
-
-			for (i = 0; i < RADEON_NUM_RINGS; ++i) {
-				if (!target_seq[i])
-					continue;
-
-				if (last_seq[i] != atomic64_read(&rdev->fence_drv[i].last_seq))
-					break;
-			}
-
-			if (i != RADEON_NUM_RINGS)
-				continue;
-
-			for (i = 0; i < RADEON_NUM_RINGS; ++i) {
-				if (!target_seq[i])
-					continue;
-
-				if (radeon_ring_is_lockup(rdev, i, &rdev->ring[i]))
-					break;
-			}
-
-			if (i < RADEON_NUM_RINGS) {
-				/* good news we believe it's a lockup */
-				dev_warn(rdev->dev, "GPU lockup (waiting for "
-					 "0x%016llx last fence id 0x%016llx on"
-					 " ring %d)\n",
-					 target_seq[i], last_seq[i], i);
-
-				/* remember that we need an reset */
-				rdev->needs_reset = true;
-				wake_up_all(&rdev->fence_queue);
-				return -EDEADLK;
-			}
-		}
+		radeon_irq_kms_sw_irq_put(rdev, i);
+		trace_radeon_fence_wait_end(rdev->ddev, i, target_seq[i]);
 	}
-	return 0;
+
+	return r;
 }
 
 /**
  * radeon_fence_wait - wait for a fence to signal
  *
  * @fence: radeon fence object
- * @intr: use interruptable sleep
+ * @intr: use interruptible sleep
  *
  * Wait for the requested fence to signal (all asics).
  * @intr selects whether to use interruptable (true) or non-interruptable
@@ -396,22 +540,26 @@ static int radeon_fence_wait_seq(struct radeon_device *rdev, u64 *target_seq,
 int radeon_fence_wait(struct radeon_fence *fence, bool intr)
 {
 	uint64_t seq[RADEON_NUM_RINGS] = {};
-	int r;
+	long r;
 
-	if (fence == NULL) {
-		WARN(1, "Querying an invalid fence : %p !\n", fence);
-		return -EINVAL;
-	}
+	/*
+	 * This function should not be called on !radeon fences.
+	 * If this is the case, it would mean this function can
+	 * also be called on radeon fences belonging to another card.
+	 * exclusive_lock is not held in that case.
+	 */
+	if (WARN_ON_ONCE(!to_radeon_fence(&fence->base)))
+		return fence_wait(&fence->base, intr);
 
 	seq[fence->ring] = fence->seq;
-	if (seq[fence->ring] == RADEON_FENCE_SIGNALED_SEQ)
-		return 0;
-
-	r = radeon_fence_wait_seq(fence->rdev, seq, intr);
-	if (r)
+	r = radeon_fence_wait_seq_timeout(fence->rdev, seq, intr, MAX_SCHEDULE_TIMEOUT);
+	if (r < 0) {
 		return r;
+	}
 
-	fence->seq = RADEON_FENCE_SIGNALED_SEQ;
+	r = fence_signal(&fence->base);
+	if (!r)
+		FENCE_TRACE(&fence->base, "signaled from fence_wait\n");
 	return 0;
 }
 
@@ -434,7 +582,7 @@ int radeon_fence_wait_any(struct radeon_device *rdev,
 {
 	uint64_t seq[RADEON_NUM_RINGS];
 	unsigned i, num_rings = 0;
-	int r;
+	long r;
 
 	for (i = 0; i < RADEON_NUM_RINGS; ++i) {
 		seq[i] = 0;
@@ -445,18 +593,14 @@ int radeon_fence_wait_any(struct radeon_device *rdev,
 
 		seq[i] = fences[i]->seq;
 		++num_rings;
-
-		/* test if something was allready signaled */
-		if (seq[i] == RADEON_FENCE_SIGNALED_SEQ)
-			return 0;
 	}
 
 	/* nothing to wait for ? */
 	if (num_rings == 0)
 		return -ENOENT;
 
-	r = radeon_fence_wait_seq(rdev, seq, intr);
-	if (r) {
+	r = radeon_fence_wait_seq_timeout(rdev, seq, intr, MAX_SCHEDULE_TIMEOUT);
+	if (r < 0) {
 		return r;
 	}
 	return 0;
@@ -475,6 +619,7 @@ int radeon_fence_wait_any(struct radeon_device *rdev,
 int radeon_fence_wait_next(struct radeon_device *rdev, int ring)
 {
 	uint64_t seq[RADEON_NUM_RINGS] = {};
+	long r;
 
 	seq[ring] = atomic64_read(&rdev->fence_drv[ring].last_seq) + 1ULL;
 	if (seq[ring] >= rdev->fence_drv[ring].sync_seq[ring]) {
@@ -482,7 +627,10 @@ int radeon_fence_wait_next(struct radeon_device *rdev, int ring)
 		   already the last emited fence */
 		return -ENOENT;
 	}
-	return radeon_fence_wait_seq(rdev, seq, false);
+	r = radeon_fence_wait_seq_timeout(rdev, seq, false, MAX_SCHEDULE_TIMEOUT);
+	if (r < 0)
+		return r;
+	return 0;
 }
 
 /**
@@ -498,18 +646,18 @@ int radeon_fence_wait_next(struct radeon_device *rdev, int ring)
 int radeon_fence_wait_empty(struct radeon_device *rdev, int ring)
 {
 	uint64_t seq[RADEON_NUM_RINGS] = {};
-	int r;
+	long r;
 
 	seq[ring] = rdev->fence_drv[ring].sync_seq[ring];
 	if (!seq[ring])
 		return 0;
 
-	r = radeon_fence_wait_seq(rdev, seq, false);
-	if (r) {
+	r = radeon_fence_wait_seq_timeout(rdev, seq, false, MAX_SCHEDULE_TIMEOUT);
+	if (r < 0) {
 		if (r == -EDEADLK)
 			return -EDEADLK;
 
-		dev_err(rdev->dev, "error waiting for ring[%d] to become idle (%d)\n",
+		dev_err(rdev->dev, "error waiting for ring[%d] to become idle (%ld)\n",
 			ring, r);
 	}
 	return 0;
@@ -525,7 +673,7 @@ int radeon_fence_wait_empty(struct radeon_device *rdev, int ring)
  */
 struct radeon_fence *radeon_fence_ref(struct radeon_fence *fence)
 {
-	kref_get(&fence->kref);
+	fence_get(&fence->base);
 	return fence;
 }
 
@@ -542,7 +690,7 @@ void radeon_fence_unref(struct radeon_fence **fence)
 
 	*fence = NULL;
 	if (tmp) {
-		kref_put(&tmp->kref, radeon_fence_destroy);
+		fence_put(&tmp->base);
 	}
 }
 
@@ -711,6 +859,9 @@ static void radeon_fence_driver_init_ring(struct radeon_device *rdev, int ring)
 		rdev->fence_drv[ring].sync_seq[i] = 0;
 	atomic64_set(&rdev->fence_drv[ring].last_seq, 0);
 	rdev->fence_drv[ring].initialized = false;
+	INIT_DELAYED_WORK(&rdev->fence_drv[ring].lockup_work,
+			  radeon_fence_check_lockup);
+	rdev->fence_drv[ring].rdev = rdev;
 }
 
 /**
@@ -758,8 +909,9 @@ void radeon_fence_driver_fini(struct radeon_device *rdev)
 		r = radeon_fence_wait_empty(rdev, ring);
 		if (r) {
 			/* no need to trigger GPU reset as we are unloading */
-			radeon_fence_driver_force_completion(rdev);
+			radeon_fence_driver_force_completion(rdev, ring);
 		}
+		cancel_delayed_work_sync(&rdev->fence_drv[ring].lockup_work);
 		wake_up_all(&rdev->fence_queue);
 		radeon_scratch_free(rdev, rdev->fence_drv[ring].scratch_reg);
 		rdev->fence_drv[ring].initialized = false;
@@ -771,18 +923,16 @@ void radeon_fence_driver_fini(struct radeon_device *rdev)
  * radeon_fence_driver_force_completion - force all fence waiter to complete
  *
  * @rdev: radeon device pointer
+ * @ring: the ring to complete
  *
  * In case of GPU reset failure make sure no process keep waiting on fence
  * that will never complete.
  */
-void radeon_fence_driver_force_completion(struct radeon_device *rdev)
+void radeon_fence_driver_force_completion(struct radeon_device *rdev, int ring)
 {
-	int ring;
-
-	for (ring = 0; ring < RADEON_NUM_RINGS; ring++) {
-		if (!rdev->fence_drv[ring].initialized)
-			continue;
+	if (rdev->fence_drv[ring].initialized) {
 		radeon_fence_write(rdev, rdev->fence_drv[ring].sync_seq[ring], ring);
+		cancel_delayed_work_sync(&rdev->fence_drv[ring].lockup_work);
 	}
 }
 
@@ -833,6 +983,7 @@ static int radeon_debugfs_gpu_reset(struct seq_file *m, void *data)
 	down_read(&rdev->exclusive_lock);
 	seq_printf(m, "%d\n", rdev->needs_reset);
 	rdev->needs_reset = true;
+	wake_up_all(&rdev->fence_queue);
 	up_read(&rdev->exclusive_lock);
 
 	return 0;
@@ -852,3 +1003,72 @@ int radeon_debugfs_fence_init(struct radeon_device *rdev)
 	return 0;
 #endif
 }
+
+static const char *radeon_fence_get_driver_name(struct fence *fence)
+{
+	return "radeon";
+}
+
+static const char *radeon_fence_get_timeline_name(struct fence *f)
+{
+	struct radeon_fence *fence = to_radeon_fence(f);
+	switch (fence->ring) {
+	case RADEON_RING_TYPE_GFX_INDEX: return "radeon.gfx";
+	case CAYMAN_RING_TYPE_CP1_INDEX: return "radeon.cp1";
+	case CAYMAN_RING_TYPE_CP2_INDEX: return "radeon.cp2";
+	case R600_RING_TYPE_DMA_INDEX: return "radeon.dma";
+	case CAYMAN_RING_TYPE_DMA1_INDEX: return "radeon.dma1";
+	case R600_RING_TYPE_UVD_INDEX: return "radeon.uvd";
+	case TN_RING_TYPE_VCE1_INDEX: return "radeon.vce1";
+	case TN_RING_TYPE_VCE2_INDEX: return "radeon.vce2";
+	default: WARN_ON_ONCE(1); return "radeon.unk";
+	}
+}
+
+static inline bool radeon_test_signaled(struct radeon_fence *fence)
+{
+	return test_bit(FENCE_FLAG_SIGNALED_BIT, &fence->base.flags);
+}
+
+static signed long radeon_fence_default_wait(struct fence *f, bool intr,
+					     signed long t)
+{
+	struct radeon_fence *fence = to_radeon_fence(f);
+	struct radeon_device *rdev = fence->rdev;
+	bool signaled;
+
+	fence_enable_sw_signaling(&fence->base);
+
+	/*
+	 * This function has to return -EDEADLK, but cannot hold
+	 * exclusive_lock during the wait because some callers
+	 * may already hold it. This means checking needs_reset without
+	 * lock, and not fiddling with any gpu internals.
+	 *
+	 * The callback installed with fence_enable_sw_signaling will
+	 * run before our wait_event_*timeout call, so we will see
+	 * both the signaled fence and the changes to needs_reset.
+	 */
+
+	if (intr)
+		t = wait_event_interruptible_timeout(rdev->fence_queue,
+			((signaled = radeon_test_signaled(fence)) ||
+			 rdev->needs_reset), t);
+	else
+		t = wait_event_timeout(rdev->fence_queue,
+			((signaled = radeon_test_signaled(fence)) ||
+			 rdev->needs_reset), t);
+
+	if (t > 0 && !signaled)
+		return -EDEADLK;
+	return t;
+}
+
+const struct fence_ops radeon_fence_ops = {
+	.get_driver_name = radeon_fence_get_driver_name,
+	.get_timeline_name = radeon_fence_get_timeline_name,
+	.enable_signaling = radeon_fence_enable_signaling,
+	.signaled = radeon_fence_is_signaled,
+	.wait = radeon_fence_default_wait,
+	.release = NULL,
+};
diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c
index a053a0779aac..5450fa95a47e 100644
--- a/drivers/gpu/drm/radeon/radeon_gart.c
+++ b/drivers/gpu/drm/radeon/radeon_gart.c
@@ -128,7 +128,7 @@ int radeon_gart_table_vram_alloc(struct radeon_device *rdev)
 	if (rdev->gart.robj == NULL) {
 		r = radeon_bo_create(rdev, rdev->gart.table_size,
 				     PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM,
-				     0, NULL, &rdev->gart.robj);
+				     0, NULL, NULL, &rdev->gart.robj);
 		if (r) {
 			return r;
 		}
@@ -165,6 +165,19 @@ int radeon_gart_table_vram_pin(struct radeon_device *rdev)
 		radeon_bo_unpin(rdev->gart.robj);
 	radeon_bo_unreserve(rdev->gart.robj);
 	rdev->gart.table_addr = gpu_addr;
+
+	if (!r) {
+		int i;
+
+		/* We might have dropped some GART table updates while it wasn't
+		 * mapped, restore all entries
+		 */
+		for (i = 0; i < rdev->gart.num_gpu_pages; i++)
+			radeon_gart_set_page(rdev, i, rdev->gart.pages_entry[i]);
+		mb();
+		radeon_gart_tlb_flush(rdev);
+	}
+
 	return r;
 }
 
@@ -228,7 +241,6 @@ void radeon_gart_unbind(struct radeon_device *rdev, unsigned offset,
 	unsigned t;
 	unsigned p;
 	int i, j;
-	u64 page_base;
 
 	if (!rdev->gart.ready) {
 		WARN(1, "trying to unbind memory from uninitialized GART !\n");
@@ -239,14 +251,12 @@ void radeon_gart_unbind(struct radeon_device *rdev, unsigned offset,
 	for (i = 0; i < pages; i++, p++) {
 		if (rdev->gart.pages[p]) {
 			rdev->gart.pages[p] = NULL;
-			rdev->gart.pages_addr[p] = rdev->dummy_page.addr;
-			page_base = rdev->gart.pages_addr[p];
 			for (j = 0; j < (PAGE_SIZE / RADEON_GPU_PAGE_SIZE); j++, t++) {
+				rdev->gart.pages_entry[t] = rdev->dummy_page.entry;
 				if (rdev->gart.ptr) {
-					radeon_gart_set_page(rdev, t, page_base,
-							     RADEON_GART_PAGE_DUMMY);
+					radeon_gart_set_page(rdev, t,
+							     rdev->dummy_page.entry);
 				}
-				page_base += RADEON_GPU_PAGE_SIZE;
 			}
 		}
 	}
@@ -274,7 +284,7 @@ int radeon_gart_bind(struct radeon_device *rdev, unsigned offset,
 {
 	unsigned t;
 	unsigned p;
-	uint64_t page_base;
+	uint64_t page_base, page_entry;
 	int i, j;
 
 	if (!rdev->gart.ready) {
@@ -285,14 +295,15 @@ int radeon_gart_bind(struct radeon_device *rdev, unsigned offset,
 	p = t / (PAGE_SIZE / RADEON_GPU_PAGE_SIZE);
 
 	for (i = 0; i < pages; i++, p++) {
-		rdev->gart.pages_addr[p] = dma_addr[i];
 		rdev->gart.pages[p] = pagelist[i];
-		if (rdev->gart.ptr) {
-			page_base = rdev->gart.pages_addr[p];
-			for (j = 0; j < (PAGE_SIZE / RADEON_GPU_PAGE_SIZE); j++, t++) {
-				radeon_gart_set_page(rdev, t, page_base, flags);
-				page_base += RADEON_GPU_PAGE_SIZE;
+		page_base = dma_addr[i];
+		for (j = 0; j < (PAGE_SIZE / RADEON_GPU_PAGE_SIZE); j++, t++) {
+			page_entry = radeon_gart_get_page_entry(page_base, flags);
+			rdev->gart.pages_entry[t] = page_entry;
+			if (rdev->gart.ptr) {
+				radeon_gart_set_page(rdev, t, page_entry);
 			}
+			page_base += RADEON_GPU_PAGE_SIZE;
 		}
 	}
 	mb();
@@ -334,16 +345,15 @@ int radeon_gart_init(struct radeon_device *rdev)
 		radeon_gart_fini(rdev);
 		return -ENOMEM;
 	}
-	rdev->gart.pages_addr = vzalloc(sizeof(dma_addr_t) *
-					rdev->gart.num_cpu_pages);
-	if (rdev->gart.pages_addr == NULL) {
+	rdev->gart.pages_entry = vmalloc(sizeof(uint64_t) *
+					 rdev->gart.num_gpu_pages);
+	if (rdev->gart.pages_entry == NULL) {
 		radeon_gart_fini(rdev);
 		return -ENOMEM;
 	}
 	/* set GART entry to point to the dummy page by default */
-	for (i = 0; i < rdev->gart.num_cpu_pages; i++) {
-		rdev->gart.pages_addr[i] = rdev->dummy_page.addr;
-	}
+	for (i = 0; i < rdev->gart.num_gpu_pages; i++)
+		rdev->gart.pages_entry[i] = rdev->dummy_page.entry;
 	return 0;
 }
 
@@ -356,15 +366,15 @@ int radeon_gart_init(struct radeon_device *rdev)
  */
 void radeon_gart_fini(struct radeon_device *rdev)
 {
-	if (rdev->gart.pages && rdev->gart.pages_addr && rdev->gart.ready) {
+	if (rdev->gart.ready) {
 		/* unbind pages */
 		radeon_gart_unbind(rdev, 0, rdev->gart.num_cpu_pages);
 	}
 	rdev->gart.ready = false;
 	vfree(rdev->gart.pages);
-	vfree(rdev->gart.pages_addr);
+	vfree(rdev->gart.pages_entry);
 	rdev->gart.pages = NULL;
-	rdev->gart.pages_addr = NULL;
+	rdev->gart.pages_entry = NULL;
 
 	radeon_dummy_page_fini(rdev);
 }
diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c
index bfd7e1b0ff3f..d0b4f7d1140d 100644
--- a/drivers/gpu/drm/radeon/radeon_gem.c
+++ b/drivers/gpu/drm/radeon/radeon_gem.c
@@ -67,7 +67,7 @@ int radeon_gem_object_create(struct radeon_device *rdev, unsigned long size,
 
 retry:
 	r = radeon_bo_create(rdev, size, alignment, kernel, initial_domain,
-			     flags, NULL, &robj);
+			     flags, NULL, NULL, &robj);
 	if (r) {
 		if (r != -ERESTARTSYS) {
 			if (initial_domain == RADEON_GEM_DOMAIN_VRAM) {
@@ -94,7 +94,7 @@ static int radeon_gem_set_domain(struct drm_gem_object *gobj,
 {
 	struct radeon_bo *robj;
 	uint32_t domain;
-	int r;
+	long r;
 
 	/* FIXME: reeimplement */
 	robj = gem_to_radeon_bo(gobj);
@@ -110,9 +110,12 @@ static int radeon_gem_set_domain(struct drm_gem_object *gobj,
 	}
 	if (domain == RADEON_GEM_DOMAIN_CPU) {
 		/* Asking for cpu access wait for object idle */
-		r = radeon_bo_wait(robj, NULL, false);
-		if (r) {
-			printk(KERN_ERR "Failed to wait for object !\n");
+		r = reservation_object_wait_timeout_rcu(robj->tbo.resv, true, true, 30 * HZ);
+		if (!r)
+			r = -EBUSY;
+
+		if (r < 0 && r != -EINTR) {
+			printk(KERN_ERR "Failed to wait for object: %li\n", r);
 			return r;
 		}
 	}
@@ -272,6 +275,94 @@ int radeon_gem_create_ioctl(struct drm_device *dev, void *data,
 	return 0;
 }
 
+int radeon_gem_userptr_ioctl(struct drm_device *dev, void *data,
+			     struct drm_file *filp)
+{
+	struct radeon_device *rdev = dev->dev_private;
+	struct drm_radeon_gem_userptr *args = data;
+	struct drm_gem_object *gobj;
+	struct radeon_bo *bo;
+	uint32_t handle;
+	int r;
+
+	if (offset_in_page(args->addr | args->size))
+		return -EINVAL;
+
+	/* reject unknown flag values */
+	if (args->flags & ~(RADEON_GEM_USERPTR_READONLY |
+	    RADEON_GEM_USERPTR_ANONONLY | RADEON_GEM_USERPTR_VALIDATE |
+	    RADEON_GEM_USERPTR_REGISTER))
+		return -EINVAL;
+
+	if (args->flags & RADEON_GEM_USERPTR_READONLY) {
+		/* readonly pages not tested on older hardware */
+		if (rdev->family < CHIP_R600)
+			return -EINVAL;
+
+	} else if (!(args->flags & RADEON_GEM_USERPTR_ANONONLY) ||
+		   !(args->flags & RADEON_GEM_USERPTR_REGISTER)) {
+
+		/* if we want to write to it we must require anonymous
+		   memory and install a MMU notifier */
+		return -EACCES;
+	}
+
+	down_read(&rdev->exclusive_lock);
+
+	/* create a gem object to contain this object in */
+	r = radeon_gem_object_create(rdev, args->size, 0,
+				     RADEON_GEM_DOMAIN_CPU, 0,
+				     false, &gobj);
+	if (r)
+		goto handle_lockup;
+
+	bo = gem_to_radeon_bo(gobj);
+	r = radeon_ttm_tt_set_userptr(bo->tbo.ttm, args->addr, args->flags);
+	if (r)
+		goto release_object;
+
+	if (args->flags & RADEON_GEM_USERPTR_REGISTER) {
+		r = radeon_mn_register(bo, args->addr);
+		if (r)
+			goto release_object;
+	}
+
+	if (args->flags & RADEON_GEM_USERPTR_VALIDATE) {
+		down_read(&current->mm->mmap_sem);
+		r = radeon_bo_reserve(bo, true);
+		if (r) {
+			up_read(&current->mm->mmap_sem);
+			goto release_object;
+		}
+
+		radeon_ttm_placement_from_domain(bo, RADEON_GEM_DOMAIN_GTT);
+		r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
+		radeon_bo_unreserve(bo);
+		up_read(&current->mm->mmap_sem);
+		if (r)
+			goto release_object;
+	}
+
+	r = drm_gem_handle_create(filp, gobj, &handle);
+	/* drop reference from allocate - handle holds it now */
+	drm_gem_object_unreference_unlocked(gobj);
+	if (r)
+		goto handle_lockup;
+
+	args->handle = handle;
+	up_read(&rdev->exclusive_lock);
+	return 0;
+
+release_object:
+	drm_gem_object_unreference_unlocked(gobj);
+
+handle_lockup:
+	up_read(&rdev->exclusive_lock);
+	r = radeon_gem_handle_lockup(rdev, r);
+
+	return r;
+}
+
 int radeon_gem_set_domain_ioctl(struct drm_device *dev, void *data,
 				struct drm_file *filp)
 {
@@ -315,6 +406,10 @@ int radeon_mode_dumb_mmap(struct drm_file *filp,
 		return -ENOENT;
 	}
 	robj = gem_to_radeon_bo(gobj);
+	if (radeon_ttm_tt_has_userptr(robj->tbo.ttm)) {
+		drm_gem_object_unreference_unlocked(gobj);
+		return -EPERM;
+	}
 	*offset_p = radeon_bo_mmap_offset(robj);
 	drm_gem_object_unreference_unlocked(gobj);
 	return 0;
@@ -357,15 +452,22 @@ int radeon_gem_wait_idle_ioctl(struct drm_device *dev, void *data,
 	struct drm_radeon_gem_wait_idle *args = data;
 	struct drm_gem_object *gobj;
 	struct radeon_bo *robj;
-	int r;
+	int r = 0;
 	uint32_t cur_placement = 0;
+	long ret;
 
 	gobj = drm_gem_object_lookup(dev, filp, args->handle);
 	if (gobj == NULL) {
 		return -ENOENT;
 	}
 	robj = gem_to_radeon_bo(gobj);
-	r = radeon_bo_wait(robj, &cur_placement, false);
+
+	ret = reservation_object_wait_timeout_rcu(robj->tbo.resv, true, true, 30 * HZ);
+	if (ret == 0)
+		r = -EBUSY;
+	else if (ret < 0)
+		r = ret;
+
 	/* Flush HDP cache via MMIO if necessary */
 	if (rdev->asic->mmio_hdp_flush &&
 	    radeon_mem_type_to_domain(cur_placement) == RADEON_GEM_DOMAIN_VRAM)
@@ -416,6 +518,68 @@ out:
 	return r;
 }
 
+/**
+ * radeon_gem_va_update_vm -update the bo_va in its VM
+ *
+ * @rdev: radeon_device pointer
+ * @bo_va: bo_va to update
+ *
+ * Update the bo_va directly after setting it's address. Errors are not
+ * vital here, so they are not reported back to userspace.
+ */
+static void radeon_gem_va_update_vm(struct radeon_device *rdev,
+				    struct radeon_bo_va *bo_va)
+{
+	struct ttm_validate_buffer tv, *entry;
+	struct radeon_bo_list *vm_bos;
+	struct ww_acquire_ctx ticket;
+	struct list_head list;
+	unsigned domain;
+	int r;
+
+	INIT_LIST_HEAD(&list);
+
+	tv.bo = &bo_va->bo->tbo;
+	tv.shared = true;
+	list_add(&tv.head, &list);
+
+	vm_bos = radeon_vm_get_bos(rdev, bo_va->vm, &list);
+	if (!vm_bos)
+		return;
+
+	r = ttm_eu_reserve_buffers(&ticket, &list, true, NULL);
+	if (r)
+		goto error_free;
+
+	list_for_each_entry(entry, &list, head) {
+		domain = radeon_mem_type_to_domain(entry->bo->mem.mem_type);
+		/* if anything is swapped out don't swap it in here,
+		   just abort and wait for the next CS */
+		if (domain == RADEON_GEM_DOMAIN_CPU)
+			goto error_unreserve;
+	}
+
+	mutex_lock(&bo_va->vm->mutex);
+	r = radeon_vm_clear_freed(rdev, bo_va->vm);
+	if (r)
+		goto error_unlock;
+
+	if (bo_va->it.start)
+		r = radeon_vm_bo_update(rdev, bo_va, &bo_va->bo->tbo.mem);
+
+error_unlock:
+	mutex_unlock(&bo_va->vm->mutex);
+
+error_unreserve:
+	ttm_eu_backoff_reservation(&ticket, &list);
+
+error_free:
+	drm_free_large(vm_bos);
+
+	if (r && r != -ERESTARTSYS)
+		DRM_ERROR("Couldn't update BO_VA (%d)\n", r);
+}
+
 int radeon_gem_va_ioctl(struct drm_device *dev, void *data,
 			  struct drm_file *filp)
 {
@@ -499,6 +663,7 @@ int radeon_gem_va_ioctl(struct drm_device *dev, void *data,
 		if (bo_va->it.start) {
 			args->operation = RADEON_VA_RESULT_VA_EXIST;
 			args->offset = bo_va->it.start * RADEON_GPU_PAGE_SIZE;
+			radeon_bo_unreserve(rbo);
 			goto out;
 		}
 		r = radeon_vm_bo_set_addr(rdev, bo_va, args->offset, args->flags);
@@ -509,12 +674,13 @@ int radeon_gem_va_ioctl(struct drm_device *dev, void *data,
 	default:
 		break;
 	}
+	if (!r)
+		radeon_gem_va_update_vm(rdev, bo_va);
 	args->operation = RADEON_VA_RESULT_OK;
 	if (r) {
 		args->operation = RADEON_VA_RESULT_ERROR;
 	}
 out:
-	radeon_bo_unreserve(rbo);
 	drm_gem_object_unreference_unlocked(gobj);
 	return r;
 }
@@ -532,6 +698,11 @@ int radeon_gem_op_ioctl(struct drm_device *dev, void *data,
 		return -ENOENT;
 	}
 	robj = gem_to_radeon_bo(gobj);
+
+	r = -EPERM;
+	if (radeon_ttm_tt_has_userptr(robj->tbo.ttm))
+		goto out;
+
 	r = radeon_bo_reserve(robj, false);
 	if (unlikely(r))
 		goto out;
diff --git a/drivers/gpu/drm/radeon/radeon_ib.c b/drivers/gpu/drm/radeon/radeon_ib.c
index 5bf2c0a05827..c39ce1f05703 100644
--- a/drivers/gpu/drm/radeon/radeon_ib.c
+++ b/drivers/gpu/drm/radeon/radeon_ib.c
@@ -64,10 +64,7 @@ int radeon_ib_get(struct radeon_device *rdev, int ring,
 		return r;
 	}
 
-	r = radeon_semaphore_create(rdev, &ib->semaphore);
-	if (r) {
-		return r;
-	}
+	radeon_sync_create(&ib->sync);
 
 	ib->ring = ring;
 	ib->fence = NULL;
@@ -96,7 +93,7 @@ int radeon_ib_get(struct radeon_device *rdev, int ring,
  */
 void radeon_ib_free(struct radeon_device *rdev, struct radeon_ib *ib)
 {
-	radeon_semaphore_free(rdev, &ib->semaphore, ib->fence);
+	radeon_sync_free(rdev, &ib->sync, ib->fence);
 	radeon_sa_bo_free(rdev, &ib->sa_bo, ib->fence);
 	radeon_fence_unref(&ib->fence);
 }
@@ -145,11 +142,11 @@ int radeon_ib_schedule(struct radeon_device *rdev, struct radeon_ib *ib,
 	if (ib->vm) {
 		struct radeon_fence *vm_id_fence;
 		vm_id_fence = radeon_vm_grab_id(rdev, ib->vm, ib->ring);
-        	radeon_semaphore_sync_to(ib->semaphore, vm_id_fence);
+		radeon_sync_fence(&ib->sync, vm_id_fence);
 	}
 
 	/* sync with other rings */
-	r = radeon_semaphore_sync_rings(rdev, ib->semaphore, ib->ring);
+	r = radeon_sync_rings(rdev, &ib->sync, ib->ring);
 	if (r) {
 		dev_err(rdev->dev, "failed to sync rings (%d)\n", r);
 		radeon_ring_unlock_undo(rdev, ring);
@@ -157,11 +154,12 @@ int radeon_ib_schedule(struct radeon_device *rdev, struct radeon_ib *ib,
 	}
 
 	if (ib->vm)
-		radeon_vm_flush(rdev, ib->vm, ib->ring);
+		radeon_vm_flush(rdev, ib->vm, ib->ring,
+				ib->sync.last_vm_update);
 
 	if (const_ib) {
 		radeon_ring_ib_execute(rdev, const_ib->ring, const_ib);
-		radeon_semaphore_free(rdev, &const_ib->semaphore, NULL);
+		radeon_sync_free(rdev, &const_ib->sync, NULL);
 	}
 	radeon_ring_ib_execute(rdev, ib->ring, ib);
 	r = radeon_fence_emit(rdev, &ib->fence, ib->ring);
@@ -269,6 +267,7 @@ int radeon_ib_ring_tests(struct radeon_device *rdev)
 
 		r = radeon_ib_test(rdev, i, ring);
 		if (r) {
+			radeon_fence_driver_force_completion(rdev, i);
 			ring->ready = false;
 			rdev->needs_reset = false;
 
diff --git a/drivers/gpu/drm/radeon/radeon_irq_kms.c b/drivers/gpu/drm/radeon/radeon_irq_kms.c
index 16807afab362..00fc59762e0d 100644
--- a/drivers/gpu/drm/radeon/radeon_irq_kms.c
+++ b/drivers/gpu/drm/radeon/radeon_irq_kms.c
@@ -88,23 +88,6 @@ static void radeon_hotplug_work_func(struct work_struct *work)
 }
 
 /**
- * radeon_irq_reset_work_func - execute gpu reset
- *
- * @work: work struct
- *
- * Execute scheduled gpu reset (cayman+).
- * This function is called when the irq handler
- * thinks we need a gpu reset.
- */
-static void radeon_irq_reset_work_func(struct work_struct *work)
-{
-	struct radeon_device *rdev = container_of(work, struct radeon_device,
-						  reset_work);
-
-	radeon_gpu_reset(rdev);
-}
-
-/**
  * radeon_driver_irq_preinstall_kms - drm irq preinstall callback
  *
  * @dev: drm dev pointer
@@ -202,6 +185,16 @@ static bool radeon_msi_ok(struct radeon_device *rdev)
 	if (rdev->flags & RADEON_IS_AGP)
 		return false;
 
+	/*
+	 * Older chips have a HW limitation, they can only generate 40 bits
+	 * of address for "64-bit" MSIs which breaks on some platforms, notably
+	 * IBM POWER servers, so we limit them
+	 */
+	if (rdev->family < CHIP_BONAIRE) {
+		dev_info(rdev->dev, "radeon: MSI limited to 32-bit\n");
+		rdev->pdev->no_64bit_msi = 1;
+	}
+
 	/* force MSI on */
 	if (radeon_msi == 1)
 		return true;
@@ -284,7 +277,6 @@ int radeon_irq_kms_init(struct radeon_device *rdev)
 
 	INIT_WORK(&rdev->hotplug_work, radeon_hotplug_work_func);
 	INIT_WORK(&rdev->audio_work, r600_audio_update_hdmi);
-	INIT_WORK(&rdev->reset_work, radeon_irq_reset_work_func);
 
 	rdev->irq.installed = true;
 	r = drm_irq_install(rdev->ddev, rdev->ddev->pdev->irq);
@@ -342,6 +334,21 @@ void radeon_irq_kms_sw_irq_get(struct radeon_device *rdev, int ring)
 }
 
 /**
+ * radeon_irq_kms_sw_irq_get_delayed - enable software interrupt
+ *
+ * @rdev: radeon device pointer
+ * @ring: ring whose interrupt you want to enable
+ *
+ * Enables the software interrupt for a specific ring (all asics).
+ * The software interrupt is generally used to signal a fence on
+ * a particular ring.
+ */
+bool radeon_irq_kms_sw_irq_get_delayed(struct radeon_device *rdev, int ring)
+{
+	return atomic_inc_return(&rdev->irq.ring_int[ring]) == 1;
+}
+
+/**
  * radeon_irq_kms_sw_irq_put - disable software interrupt
  *
  * @rdev: radeon device pointer
diff --git a/drivers/gpu/drm/radeon/radeon_kfd.c b/drivers/gpu/drm/radeon/radeon_kfd.c
new file mode 100644
index 000000000000..bef9a0953284
--- /dev/null
+++ b/drivers/gpu/drm/radeon/radeon_kfd.c
@@ -0,0 +1,633 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#include <linux/module.h>
+#include <linux/fdtable.h>
+#include <linux/uaccess.h>
+#include <drm/drmP.h>
+#include "radeon.h"
+#include "cikd.h"
+#include "cik_reg.h"
+#include "radeon_kfd.h"
+#include "radeon_ucode.h"
+#include <linux/firmware.h>
+
+#define CIK_PIPE_PER_MEC	(4)
+
+struct kgd_mem {
+	struct radeon_sa_bo *sa_bo;
+	uint64_t gpu_addr;
+	void *ptr;
+};
+
+static int init_sa_manager(struct kgd_dev *kgd, unsigned int size);
+static void fini_sa_manager(struct kgd_dev *kgd);
+
+static int allocate_mem(struct kgd_dev *kgd, size_t size, size_t alignment,
+		enum kgd_memory_pool pool, struct kgd_mem **mem);
+
+static void free_mem(struct kgd_dev *kgd, struct kgd_mem *mem);
+
+static uint64_t get_vmem_size(struct kgd_dev *kgd);
+static uint64_t get_gpu_clock_counter(struct kgd_dev *kgd);
+
+static uint32_t get_max_engine_clock_in_mhz(struct kgd_dev *kgd);
+static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type);
+
+/*
+ * Register access functions
+ */
+
+static void kgd_program_sh_mem_settings(struct kgd_dev *kgd, uint32_t vmid,
+		uint32_t sh_mem_config,	uint32_t sh_mem_ape1_base,
+		uint32_t sh_mem_ape1_limit, uint32_t sh_mem_bases);
+
+static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
+					unsigned int vmid);
+
+static int kgd_init_memory(struct kgd_dev *kgd);
+
+static int kgd_init_pipeline(struct kgd_dev *kgd, uint32_t pipe_id,
+				uint32_t hpd_size, uint64_t hpd_gpu_addr);
+
+static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
+			uint32_t queue_id, uint32_t __user *wptr);
+
+static bool kgd_hqd_is_occupied(struct kgd_dev *kgd, uint64_t queue_address,
+				uint32_t pipe_id, uint32_t queue_id);
+
+static int kgd_hqd_destroy(struct kgd_dev *kgd, uint32_t reset_type,
+				unsigned int timeout, uint32_t pipe_id,
+				uint32_t queue_id);
+
+static const struct kfd2kgd_calls kfd2kgd = {
+	.init_sa_manager = init_sa_manager,
+	.fini_sa_manager = fini_sa_manager,
+	.allocate_mem = allocate_mem,
+	.free_mem = free_mem,
+	.get_vmem_size = get_vmem_size,
+	.get_gpu_clock_counter = get_gpu_clock_counter,
+	.get_max_engine_clock_in_mhz = get_max_engine_clock_in_mhz,
+	.program_sh_mem_settings = kgd_program_sh_mem_settings,
+	.set_pasid_vmid_mapping = kgd_set_pasid_vmid_mapping,
+	.init_memory = kgd_init_memory,
+	.init_pipeline = kgd_init_pipeline,
+	.hqd_load = kgd_hqd_load,
+	.hqd_is_occupied = kgd_hqd_is_occupied,
+	.hqd_destroy = kgd_hqd_destroy,
+	.get_fw_version = get_fw_version
+};
+
+static const struct kgd2kfd_calls *kgd2kfd;
+
+bool radeon_kfd_init(void)
+{
+#if defined(CONFIG_HSA_AMD_MODULE)
+	bool (*kgd2kfd_init_p)(unsigned, const struct kfd2kgd_calls*,
+				const struct kgd2kfd_calls**);
+
+	kgd2kfd_init_p = symbol_request(kgd2kfd_init);
+
+	if (kgd2kfd_init_p == NULL)
+		return false;
+
+	if (!kgd2kfd_init_p(KFD_INTERFACE_VERSION, &kfd2kgd, &kgd2kfd)) {
+		symbol_put(kgd2kfd_init);
+		kgd2kfd = NULL;
+
+		return false;
+	}
+
+	return true;
+#elif defined(CONFIG_HSA_AMD)
+	if (!kgd2kfd_init(KFD_INTERFACE_VERSION, &kfd2kgd, &kgd2kfd)) {
+		kgd2kfd = NULL;
+
+		return false;
+	}
+
+	return true;
+#else
+	return false;
+#endif
+}
+
+void radeon_kfd_fini(void)
+{
+	if (kgd2kfd) {
+		kgd2kfd->exit();
+		symbol_put(kgd2kfd_init);
+	}
+}
+
+void radeon_kfd_device_probe(struct radeon_device *rdev)
+{
+	if (kgd2kfd)
+		rdev->kfd = kgd2kfd->probe((struct kgd_dev *)rdev, rdev->pdev);
+}
+
+void radeon_kfd_device_init(struct radeon_device *rdev)
+{
+	if (rdev->kfd) {
+		struct kgd2kfd_shared_resources gpu_resources = {
+			.compute_vmid_bitmap = 0xFF00,
+
+			.first_compute_pipe = 1,
+			.compute_pipe_count = 8 - 1,
+		};
+
+		radeon_doorbell_get_kfd_info(rdev,
+				&gpu_resources.doorbell_physical_address,
+				&gpu_resources.doorbell_aperture_size,
+				&gpu_resources.doorbell_start_offset);
+
+		kgd2kfd->device_init(rdev->kfd, &gpu_resources);
+	}
+}
+
+void radeon_kfd_device_fini(struct radeon_device *rdev)
+{
+	if (rdev->kfd) {
+		kgd2kfd->device_exit(rdev->kfd);
+		rdev->kfd = NULL;
+	}
+}
+
+void radeon_kfd_interrupt(struct radeon_device *rdev, const void *ih_ring_entry)
+{
+	if (rdev->kfd)
+		kgd2kfd->interrupt(rdev->kfd, ih_ring_entry);
+}
+
+void radeon_kfd_suspend(struct radeon_device *rdev)
+{
+	if (rdev->kfd)
+		kgd2kfd->suspend(rdev->kfd);
+}
+
+int radeon_kfd_resume(struct radeon_device *rdev)
+{
+	int r = 0;
+
+	if (rdev->kfd)
+		r = kgd2kfd->resume(rdev->kfd);
+
+	return r;
+}
+
+static u32 pool_to_domain(enum kgd_memory_pool p)
+{
+	switch (p) {
+	case KGD_POOL_FRAMEBUFFER: return RADEON_GEM_DOMAIN_VRAM;
+	default: return RADEON_GEM_DOMAIN_GTT;
+	}
+}
+
+static int init_sa_manager(struct kgd_dev *kgd, unsigned int size)
+{
+	struct radeon_device *rdev = (struct radeon_device *)kgd;
+	int r;
+
+	BUG_ON(kgd == NULL);
+
+	r = radeon_sa_bo_manager_init(rdev, &rdev->kfd_bo,
+				      size,
+				      RADEON_GPU_PAGE_SIZE,
+				      RADEON_GEM_DOMAIN_GTT,
+				      RADEON_GEM_GTT_WC);
+
+	if (r)
+		return r;
+
+	r = radeon_sa_bo_manager_start(rdev, &rdev->kfd_bo);
+	if (r)
+		radeon_sa_bo_manager_fini(rdev, &rdev->kfd_bo);
+
+	return r;
+}
+
+static void fini_sa_manager(struct kgd_dev *kgd)
+{
+	struct radeon_device *rdev = (struct radeon_device *)kgd;
+
+	BUG_ON(kgd == NULL);
+
+	radeon_sa_bo_manager_suspend(rdev, &rdev->kfd_bo);
+	radeon_sa_bo_manager_fini(rdev, &rdev->kfd_bo);
+}
+
+static int allocate_mem(struct kgd_dev *kgd, size_t size, size_t alignment,
+		enum kgd_memory_pool pool, struct kgd_mem **mem)
+{
+	struct radeon_device *rdev = (struct radeon_device *)kgd;
+	u32 domain;
+	int r;
+
+	BUG_ON(kgd == NULL);
+
+	domain = pool_to_domain(pool);
+	if (domain != RADEON_GEM_DOMAIN_GTT) {
+		dev_err(rdev->dev,
+			"Only allowed to allocate gart memory for kfd\n");
+		return -EINVAL;
+	}
+
+	*mem = kmalloc(sizeof(struct kgd_mem), GFP_KERNEL);
+	if ((*mem) == NULL)
+		return -ENOMEM;
+
+	r = radeon_sa_bo_new(rdev, &rdev->kfd_bo, &(*mem)->sa_bo, size,
+				alignment);
+	if (r) {
+		dev_err(rdev->dev, "failed to get memory for kfd (%d)\n", r);
+		return r;
+	}
+
+	(*mem)->ptr = radeon_sa_bo_cpu_addr((*mem)->sa_bo);
+	(*mem)->gpu_addr = radeon_sa_bo_gpu_addr((*mem)->sa_bo);
+
+	return 0;
+}
+
+static void free_mem(struct kgd_dev *kgd, struct kgd_mem *mem)
+{
+	struct radeon_device *rdev = (struct radeon_device *)kgd;
+
+	BUG_ON(kgd == NULL);
+
+	radeon_sa_bo_free(rdev, &mem->sa_bo, NULL);
+	kfree(mem);
+}
+
+static uint64_t get_vmem_size(struct kgd_dev *kgd)
+{
+	struct radeon_device *rdev = (struct radeon_device *)kgd;
+
+	BUG_ON(kgd == NULL);
+
+	return rdev->mc.real_vram_size;
+}
+
+static uint64_t get_gpu_clock_counter(struct kgd_dev *kgd)
+{
+	struct radeon_device *rdev = (struct radeon_device *)kgd;
+
+	return rdev->asic->get_gpu_clock_counter(rdev);
+}
+
+static uint32_t get_max_engine_clock_in_mhz(struct kgd_dev *kgd)
+{
+	struct radeon_device *rdev = (struct radeon_device *)kgd;
+
+	/* The sclk is in quantas of 10kHz */
+	return rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.sclk / 100;
+}
+
+static inline struct radeon_device *get_radeon_device(struct kgd_dev *kgd)
+{
+	return (struct radeon_device *)kgd;
+}
+
+static void write_register(struct kgd_dev *kgd, uint32_t offset, uint32_t value)
+{
+	struct radeon_device *rdev = get_radeon_device(kgd);
+
+	writel(value, (void __iomem *)(rdev->rmmio + offset));
+}
+
+static uint32_t read_register(struct kgd_dev *kgd, uint32_t offset)
+{
+	struct radeon_device *rdev = get_radeon_device(kgd);
+
+	return readl((void __iomem *)(rdev->rmmio + offset));
+}
+
+static void lock_srbm(struct kgd_dev *kgd, uint32_t mec, uint32_t pipe,
+			uint32_t queue, uint32_t vmid)
+{
+	struct radeon_device *rdev = get_radeon_device(kgd);
+	uint32_t value = PIPEID(pipe) | MEID(mec) | VMID(vmid) | QUEUEID(queue);
+
+	mutex_lock(&rdev->srbm_mutex);
+	write_register(kgd, SRBM_GFX_CNTL, value);
+}
+
+static void unlock_srbm(struct kgd_dev *kgd)
+{
+	struct radeon_device *rdev = get_radeon_device(kgd);
+
+	write_register(kgd, SRBM_GFX_CNTL, 0);
+	mutex_unlock(&rdev->srbm_mutex);
+}
+
+static void acquire_queue(struct kgd_dev *kgd, uint32_t pipe_id,
+				uint32_t queue_id)
+{
+	uint32_t mec = (++pipe_id / CIK_PIPE_PER_MEC) + 1;
+	uint32_t pipe = (pipe_id % CIK_PIPE_PER_MEC);
+
+	lock_srbm(kgd, mec, pipe, queue_id, 0);
+}
+
+static void release_queue(struct kgd_dev *kgd)
+{
+	unlock_srbm(kgd);
+}
+
+static void kgd_program_sh_mem_settings(struct kgd_dev *kgd, uint32_t vmid,
+					uint32_t sh_mem_config,
+					uint32_t sh_mem_ape1_base,
+					uint32_t sh_mem_ape1_limit,
+					uint32_t sh_mem_bases)
+{
+	lock_srbm(kgd, 0, 0, 0, vmid);
+
+	write_register(kgd, SH_MEM_CONFIG, sh_mem_config);
+	write_register(kgd, SH_MEM_APE1_BASE, sh_mem_ape1_base);
+	write_register(kgd, SH_MEM_APE1_LIMIT, sh_mem_ape1_limit);
+	write_register(kgd, SH_MEM_BASES, sh_mem_bases);
+
+	unlock_srbm(kgd);
+}
+
+static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
+					unsigned int vmid)
+{
+	/*
+	 * We have to assume that there is no outstanding mapping.
+	 * The ATC_VMID_PASID_MAPPING_UPDATE_STATUS bit could be 0
+	 * because a mapping is in progress or because a mapping finished and
+	 * the SW cleared it.
+	 * So the protocol is to always wait & clear.
+	 */
+	uint32_t pasid_mapping = (pasid == 0) ? 0 :
+				(uint32_t)pasid | ATC_VMID_PASID_MAPPING_VALID;
+
+	write_register(kgd, ATC_VMID0_PASID_MAPPING + vmid*sizeof(uint32_t),
+			pasid_mapping);
+
+	while (!(read_register(kgd, ATC_VMID_PASID_MAPPING_UPDATE_STATUS) &
+								(1U << vmid)))
+		cpu_relax();
+	write_register(kgd, ATC_VMID_PASID_MAPPING_UPDATE_STATUS, 1U << vmid);
+
+	/* Mapping vmid to pasid also for IH block */
+	write_register(kgd, IH_VMID_0_LUT + vmid * sizeof(uint32_t),
+			pasid_mapping);
+
+	return 0;
+}
+
+static int kgd_init_memory(struct kgd_dev *kgd)
+{
+	/*
+	 * Configure apertures:
+	 * LDS:         0x60000000'00000000 - 0x60000001'00000000 (4GB)
+	 * Scratch:     0x60000001'00000000 - 0x60000002'00000000 (4GB)
+	 * GPUVM:       0x60010000'00000000 - 0x60020000'00000000 (1TB)
+	 */
+	int i;
+	uint32_t sh_mem_bases = PRIVATE_BASE(0x6000) | SHARED_BASE(0x6000);
+
+	for (i = 8; i < 16; i++) {
+		uint32_t sh_mem_config;
+
+		lock_srbm(kgd, 0, 0, 0, i);
+
+		sh_mem_config = ALIGNMENT_MODE(SH_MEM_ALIGNMENT_MODE_UNALIGNED);
+		sh_mem_config |= DEFAULT_MTYPE(MTYPE_NONCACHED);
+
+		write_register(kgd, SH_MEM_CONFIG, sh_mem_config);
+
+		write_register(kgd, SH_MEM_BASES, sh_mem_bases);
+
+		/* Scratch aperture is not supported for now. */
+		write_register(kgd, SH_STATIC_MEM_CONFIG, 0);
+
+		/* APE1 disabled for now. */
+		write_register(kgd, SH_MEM_APE1_BASE, 1);
+		write_register(kgd, SH_MEM_APE1_LIMIT, 0);
+
+		unlock_srbm(kgd);
+	}
+
+	return 0;
+}
+
+static int kgd_init_pipeline(struct kgd_dev *kgd, uint32_t pipe_id,
+				uint32_t hpd_size, uint64_t hpd_gpu_addr)
+{
+	uint32_t mec = (pipe_id / CIK_PIPE_PER_MEC) + 1;
+	uint32_t pipe = (pipe_id % CIK_PIPE_PER_MEC);
+
+	lock_srbm(kgd, mec, pipe, 0, 0);
+	write_register(kgd, CP_HPD_EOP_BASE_ADDR,
+			lower_32_bits(hpd_gpu_addr >> 8));
+	write_register(kgd, CP_HPD_EOP_BASE_ADDR_HI,
+			upper_32_bits(hpd_gpu_addr >> 8));
+	write_register(kgd, CP_HPD_EOP_VMID, 0);
+	write_register(kgd, CP_HPD_EOP_CONTROL, hpd_size);
+	unlock_srbm(kgd);
+
+	return 0;
+}
+
+static inline struct cik_mqd *get_mqd(void *mqd)
+{
+	return (struct cik_mqd *)mqd;
+}
+
+static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
+			uint32_t queue_id, uint32_t __user *wptr)
+{
+	uint32_t wptr_shadow, is_wptr_shadow_valid;
+	struct cik_mqd *m;
+
+	m = get_mqd(mqd);
+
+	is_wptr_shadow_valid = !get_user(wptr_shadow, wptr);
+
+	acquire_queue(kgd, pipe_id, queue_id);
+	write_register(kgd, CP_MQD_BASE_ADDR, m->cp_mqd_base_addr_lo);
+	write_register(kgd, CP_MQD_BASE_ADDR_HI, m->cp_mqd_base_addr_hi);
+	write_register(kgd, CP_MQD_CONTROL, m->cp_mqd_control);
+
+	write_register(kgd, CP_HQD_PQ_BASE, m->cp_hqd_pq_base_lo);
+	write_register(kgd, CP_HQD_PQ_BASE_HI, m->cp_hqd_pq_base_hi);
+	write_register(kgd, CP_HQD_PQ_CONTROL, m->cp_hqd_pq_control);
+
+	write_register(kgd, CP_HQD_IB_CONTROL, m->cp_hqd_ib_control);
+	write_register(kgd, CP_HQD_IB_BASE_ADDR, m->cp_hqd_ib_base_addr_lo);
+	write_register(kgd, CP_HQD_IB_BASE_ADDR_HI, m->cp_hqd_ib_base_addr_hi);
+
+	write_register(kgd, CP_HQD_IB_RPTR, m->cp_hqd_ib_rptr);
+
+	write_register(kgd, CP_HQD_PERSISTENT_STATE,
+			m->cp_hqd_persistent_state);
+	write_register(kgd, CP_HQD_SEMA_CMD, m->cp_hqd_sema_cmd);
+	write_register(kgd, CP_HQD_MSG_TYPE, m->cp_hqd_msg_type);
+
+	write_register(kgd, CP_HQD_ATOMIC0_PREOP_LO,
+			m->cp_hqd_atomic0_preop_lo);
+
+	write_register(kgd, CP_HQD_ATOMIC0_PREOP_HI,
+			m->cp_hqd_atomic0_preop_hi);
+
+	write_register(kgd, CP_HQD_ATOMIC1_PREOP_LO,
+			m->cp_hqd_atomic1_preop_lo);
+
+	write_register(kgd, CP_HQD_ATOMIC1_PREOP_HI,
+			m->cp_hqd_atomic1_preop_hi);
+
+	write_register(kgd, CP_HQD_PQ_RPTR_REPORT_ADDR,
+			m->cp_hqd_pq_rptr_report_addr_lo);
+
+	write_register(kgd, CP_HQD_PQ_RPTR_REPORT_ADDR_HI,
+			m->cp_hqd_pq_rptr_report_addr_hi);
+
+	write_register(kgd, CP_HQD_PQ_RPTR, m->cp_hqd_pq_rptr);
+
+	write_register(kgd, CP_HQD_PQ_WPTR_POLL_ADDR,
+			m->cp_hqd_pq_wptr_poll_addr_lo);
+
+	write_register(kgd, CP_HQD_PQ_WPTR_POLL_ADDR_HI,
+			m->cp_hqd_pq_wptr_poll_addr_hi);
+
+	write_register(kgd, CP_HQD_PQ_DOORBELL_CONTROL,
+			m->cp_hqd_pq_doorbell_control);
+
+	write_register(kgd, CP_HQD_VMID, m->cp_hqd_vmid);
+
+	write_register(kgd, CP_HQD_QUANTUM, m->cp_hqd_quantum);
+
+	write_register(kgd, CP_HQD_PIPE_PRIORITY, m->cp_hqd_pipe_priority);
+	write_register(kgd, CP_HQD_QUEUE_PRIORITY, m->cp_hqd_queue_priority);
+
+	write_register(kgd, CP_HQD_IQ_RPTR, m->cp_hqd_iq_rptr);
+
+	if (is_wptr_shadow_valid)
+		write_register(kgd, CP_HQD_PQ_WPTR, wptr_shadow);
+
+	write_register(kgd, CP_HQD_ACTIVE, m->cp_hqd_active);
+	release_queue(kgd);
+
+	return 0;
+}
+
+static bool kgd_hqd_is_occupied(struct kgd_dev *kgd, uint64_t queue_address,
+				uint32_t pipe_id, uint32_t queue_id)
+{
+	uint32_t act;
+	bool retval = false;
+	uint32_t low, high;
+
+	acquire_queue(kgd, pipe_id, queue_id);
+	act = read_register(kgd, CP_HQD_ACTIVE);
+	if (act) {
+		low = lower_32_bits(queue_address >> 8);
+		high = upper_32_bits(queue_address >> 8);
+
+		if (low == read_register(kgd, CP_HQD_PQ_BASE) &&
+				high == read_register(kgd, CP_HQD_PQ_BASE_HI))
+			retval = true;
+	}
+	release_queue(kgd);
+	return retval;
+}
+
+static int kgd_hqd_destroy(struct kgd_dev *kgd, uint32_t reset_type,
+				unsigned int timeout, uint32_t pipe_id,
+				uint32_t queue_id)
+{
+	uint32_t temp;
+
+	acquire_queue(kgd, pipe_id, queue_id);
+	write_register(kgd, CP_HQD_PQ_DOORBELL_CONTROL, 0);
+
+	write_register(kgd, CP_HQD_DEQUEUE_REQUEST, reset_type);
+
+	while (true) {
+		temp = read_register(kgd, CP_HQD_ACTIVE);
+		if (temp & 0x1)
+			break;
+		if (timeout == 0) {
+			pr_err("kfd: cp queue preemption time out (%dms)\n",
+				temp);
+			release_queue(kgd);
+			return -ETIME;
+		}
+		msleep(20);
+		timeout -= 20;
+	}
+
+	release_queue(kgd);
+	return 0;
+}
+
+static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type)
+{
+	struct radeon_device *rdev = (struct radeon_device *) kgd;
+	const union radeon_firmware_header *hdr;
+
+	BUG_ON(kgd == NULL || rdev->mec_fw == NULL);
+
+	switch (type) {
+	case KGD_ENGINE_PFP:
+		hdr = (const union radeon_firmware_header *) rdev->pfp_fw->data;
+		break;
+
+	case KGD_ENGINE_ME:
+		hdr = (const union radeon_firmware_header *) rdev->me_fw->data;
+		break;
+
+	case KGD_ENGINE_CE:
+		hdr = (const union radeon_firmware_header *) rdev->ce_fw->data;
+		break;
+
+	case KGD_ENGINE_MEC1:
+		hdr = (const union radeon_firmware_header *) rdev->mec_fw->data;
+		break;
+
+	case KGD_ENGINE_MEC2:
+		hdr = (const union radeon_firmware_header *)
+							rdev->mec2_fw->data;
+		break;
+
+	case KGD_ENGINE_RLC:
+		hdr = (const union radeon_firmware_header *) rdev->rlc_fw->data;
+		break;
+
+	case KGD_ENGINE_SDMA:
+		hdr = (const union radeon_firmware_header *)
+							rdev->sdma_fw->data;
+		break;
+
+	default:
+		return 0;
+	}
+
+	if (hdr == NULL)
+		return 0;
+
+	/* Only 12 bit in use*/
+	return hdr->common.ucode_version;
+}
diff --git a/drivers/gpu/drm/radeon/radeon_kfd.h b/drivers/gpu/drm/radeon/radeon_kfd.h
new file mode 100644
index 000000000000..f90e161ca507
--- /dev/null
+++ b/drivers/gpu/drm/radeon/radeon_kfd.h
@@ -0,0 +1,47 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+/*
+ * radeon_kfd.h defines the private interface between the
+ * AMD kernel graphics drivers and the AMD KFD.
+ */
+
+#ifndef RADEON_KFD_H_INCLUDED
+#define RADEON_KFD_H_INCLUDED
+
+#include <linux/types.h>
+#include "../amd/include/kgd_kfd_interface.h"
+
+struct radeon_device;
+
+bool radeon_kfd_init(void);
+void radeon_kfd_fini(void);
+
+void radeon_kfd_suspend(struct radeon_device *rdev);
+int radeon_kfd_resume(struct radeon_device *rdev);
+void radeon_kfd_interrupt(struct radeon_device *rdev,
+			const void *ih_ring_entry);
+void radeon_kfd_device_probe(struct radeon_device *rdev);
+void radeon_kfd_device_init(struct radeon_device *rdev);
+void radeon_kfd_device_fini(struct radeon_device *rdev);
+
+#endif /* RADEON_KFD_H_INCLUDED */
diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c
index eb7164d07985..3cf9c1fa6475 100644
--- a/drivers/gpu/drm/radeon/radeon_kms.c
+++ b/drivers/gpu/drm/radeon/radeon_kms.c
@@ -34,6 +34,8 @@
 #include <linux/slab.h>
 #include <linux/pm_runtime.h>
 
+#include "radeon_kfd.h"
+
 #if defined(CONFIG_VGA_SWITCHEROO)
 bool radeon_has_atpx(void);
 #else
@@ -63,6 +65,8 @@ int radeon_driver_unload_kms(struct drm_device *dev)
 
 	pm_runtime_get_sync(dev->dev);
 
+	radeon_kfd_device_fini(rdev);
+
 	radeon_acpi_fini(rdev);
 	
 	radeon_modeset_fini(rdev);
@@ -142,6 +146,9 @@ int radeon_driver_load_kms(struct drm_device *dev, unsigned long flags)
 				"Error during ACPI methods call\n");
 	}
 
+	radeon_kfd_device_probe(rdev);
+	radeon_kfd_device_init(rdev);
+
 	if (radeon_is_px(dev)) {
 		pm_runtime_use_autosuspend(dev->dev);
 		pm_runtime_set_autosuspend_delay(dev->dev, 5000);
@@ -621,8 +628,6 @@ int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
 						  RADEON_VA_IB_OFFSET,
 						  RADEON_VM_PAGE_READABLE |
 						  RADEON_VM_PAGE_SNOOPED);
-
-			radeon_bo_unreserve(rdev->ring_tmp_bo.bo);
 			if (r) {
 				radeon_vm_fini(rdev, vm);
 				kfree(fpriv);
@@ -795,6 +800,8 @@ int radeon_get_vblank_timestamp_kms(struct drm_device *dev, int crtc,
 
 	/* Get associated drm_crtc: */
 	drmcrtc = &rdev->mode_info.crtcs[crtc]->base;
+	if (!drmcrtc)
+		return -EINVAL;
 
 	/* Helper routine in DRM core does all the work: */
 	return drm_calc_vbltimestamp_from_scanoutpos(dev, crtc, max_error,
@@ -885,5 +892,6 @@ const struct drm_ioctl_desc radeon_ioctls_kms[] = {
 	DRM_IOCTL_DEF_DRV(RADEON_GEM_BUSY, radeon_gem_busy_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(RADEON_GEM_VA, radeon_gem_va_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(RADEON_GEM_OP, radeon_gem_op_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(RADEON_GEM_USERPTR, radeon_gem_userptr_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
 };
 int radeon_max_kms_ioctl = ARRAY_SIZE(radeon_ioctls_kms);
diff --git a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
index cafb1ccf2ec3..678b4386540d 100644
--- a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
+++ b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
@@ -1054,6 +1054,7 @@ static int radeon_crtc_mode_set(struct drm_crtc *crtc,
 			DRM_ERROR("Mode need scaling but only first crtc can do that.\n");
 		}
 	}
+	radeon_cursor_reset(crtc);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/radeon/radeon_mn.c b/drivers/gpu/drm/radeon/radeon_mn.c
new file mode 100644
index 000000000000..a69bd441dd2d
--- /dev/null
+++ b/drivers/gpu/drm/radeon/radeon_mn.c
@@ -0,0 +1,274 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ * All Rights Reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
+ * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+ * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+ * USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ */
+/*
+ * Authors:
+ *    Christian König <christian.koenig@amd.com>
+ */
+
+#include <linux/firmware.h>
+#include <linux/module.h>
+#include <linux/mmu_notifier.h>
+#include <drm/drmP.h>
+#include <drm/drm.h>
+
+#include "radeon.h"
+
+struct radeon_mn {
+	/* constant after initialisation */
+	struct radeon_device	*rdev;
+	struct mm_struct	*mm;
+	struct mmu_notifier	mn;
+
+	/* only used on destruction */
+	struct work_struct	work;
+
+	/* protected by rdev->mn_lock */
+	struct hlist_node	node;
+
+	/* objects protected by lock */
+	struct mutex		lock;
+	struct rb_root		objects;
+};
+
+/**
+ * radeon_mn_destroy - destroy the rmn
+ *
+ * @work: previously sheduled work item
+ *
+ * Lazy destroys the notifier from a work item
+ */
+static void radeon_mn_destroy(struct work_struct *work)
+{
+	struct radeon_mn *rmn = container_of(work, struct radeon_mn, work);
+	struct radeon_device *rdev = rmn->rdev;
+	struct radeon_bo *bo, *next;
+
+	mutex_lock(&rdev->mn_lock);
+	mutex_lock(&rmn->lock);
+	hash_del(&rmn->node);
+	rbtree_postorder_for_each_entry_safe(bo, next, &rmn->objects, mn_it.rb) {
+		interval_tree_remove(&bo->mn_it, &rmn->objects);
+		bo->mn = NULL;
+	}
+	mutex_unlock(&rmn->lock);
+	mutex_unlock(&rdev->mn_lock);
+	mmu_notifier_unregister(&rmn->mn, rmn->mm);
+	kfree(rmn);
+}
+
+/**
+ * radeon_mn_release - callback to notify about mm destruction
+ *
+ * @mn: our notifier
+ * @mn: the mm this callback is about
+ *
+ * Shedule a work item to lazy destroy our notifier.
+ */
+static void radeon_mn_release(struct mmu_notifier *mn,
+			      struct mm_struct *mm)
+{
+	struct radeon_mn *rmn = container_of(mn, struct radeon_mn, mn);
+	INIT_WORK(&rmn->work, radeon_mn_destroy);
+	schedule_work(&rmn->work);
+}
+
+/**
+ * radeon_mn_invalidate_range_start - callback to notify about mm change
+ *
+ * @mn: our notifier
+ * @mn: the mm this callback is about
+ * @start: start of updated range
+ * @end: end of updated range
+ *
+ * We block for all BOs between start and end to be idle and
+ * unmap them by move them into system domain again.
+ */
+static void radeon_mn_invalidate_range_start(struct mmu_notifier *mn,
+					     struct mm_struct *mm,
+					     unsigned long start,
+					     unsigned long end)
+{
+	struct radeon_mn *rmn = container_of(mn, struct radeon_mn, mn);
+	struct interval_tree_node *it;
+
+	/* notification is exclusive, but interval is inclusive */
+	end -= 1;
+
+	mutex_lock(&rmn->lock);
+
+	it = interval_tree_iter_first(&rmn->objects, start, end);
+	while (it) {
+		struct radeon_bo *bo;
+		struct fence *fence;
+		int r;
+
+		bo = container_of(it, struct radeon_bo, mn_it);
+		it = interval_tree_iter_next(it, start, end);
+
+		r = radeon_bo_reserve(bo, true);
+		if (r) {
+			DRM_ERROR("(%d) failed to reserve user bo\n", r);
+			continue;
+		}
+
+		fence = reservation_object_get_excl(bo->tbo.resv);
+		if (fence) {
+			r = radeon_fence_wait((struct radeon_fence *)fence, false);
+			if (r)
+				DRM_ERROR("(%d) failed to wait for user bo\n", r);
+		}
+
+		radeon_ttm_placement_from_domain(bo, RADEON_GEM_DOMAIN_CPU);
+		r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
+		if (r)
+			DRM_ERROR("(%d) failed to validate user bo\n", r);
+
+		radeon_bo_unreserve(bo);
+	}
+	
+	mutex_unlock(&rmn->lock);
+}
+
+static const struct mmu_notifier_ops radeon_mn_ops = {
+	.release = radeon_mn_release,
+	.invalidate_range_start = radeon_mn_invalidate_range_start,
+};
+
+/**
+ * radeon_mn_get - create notifier context
+ *
+ * @rdev: radeon device pointer
+ *
+ * Creates a notifier context for current->mm.
+ */
+static struct radeon_mn *radeon_mn_get(struct radeon_device *rdev)
+{
+	struct mm_struct *mm = current->mm;
+	struct radeon_mn *rmn;
+	int r;
+
+	down_write(&mm->mmap_sem);
+	mutex_lock(&rdev->mn_lock);
+
+	hash_for_each_possible(rdev->mn_hash, rmn, node, (unsigned long)mm)
+		if (rmn->mm == mm)
+			goto release_locks;
+
+	rmn = kzalloc(sizeof(*rmn), GFP_KERNEL);
+	if (!rmn) {
+		rmn = ERR_PTR(-ENOMEM);
+		goto release_locks;
+	}
+
+	rmn->rdev = rdev;
+	rmn->mm = mm;
+	rmn->mn.ops = &radeon_mn_ops;
+	mutex_init(&rmn->lock);
+	rmn->objects = RB_ROOT;
+	
+	r = __mmu_notifier_register(&rmn->mn, mm);
+	if (r)
+		goto free_rmn;
+
+	hash_add(rdev->mn_hash, &rmn->node, (unsigned long)mm);
+
+release_locks:
+	mutex_unlock(&rdev->mn_lock);
+	up_write(&mm->mmap_sem);
+
+	return rmn;
+
+free_rmn:
+	mutex_unlock(&rdev->mn_lock);
+	up_write(&mm->mmap_sem);
+	kfree(rmn);
+
+	return ERR_PTR(r);
+}
+
+/**
+ * radeon_mn_register - register a BO for notifier updates
+ *
+ * @bo: radeon buffer object
+ * @addr: userptr addr we should monitor
+ *
+ * Registers an MMU notifier for the given BO at the specified address.
+ * Returns 0 on success, -ERRNO if anything goes wrong.
+ */
+int radeon_mn_register(struct radeon_bo *bo, unsigned long addr)
+{
+	unsigned long end = addr + radeon_bo_size(bo) - 1;
+	struct radeon_device *rdev = bo->rdev;
+	struct radeon_mn *rmn;
+	struct interval_tree_node *it;
+
+	rmn = radeon_mn_get(rdev);
+	if (IS_ERR(rmn))
+		return PTR_ERR(rmn);
+
+	mutex_lock(&rmn->lock);
+
+	it = interval_tree_iter_first(&rmn->objects, addr, end);
+	if (it) {
+		mutex_unlock(&rmn->lock);
+		return -EEXIST;
+	}
+
+	bo->mn = rmn;
+	bo->mn_it.start = addr;
+	bo->mn_it.last = end;
+	interval_tree_insert(&bo->mn_it, &rmn->objects);
+
+	mutex_unlock(&rmn->lock);
+
+	return 0;
+}
+
+/**
+ * radeon_mn_unregister - unregister a BO for notifier updates
+ *
+ * @bo: radeon buffer object
+ *
+ * Remove any registration of MMU notifier updates from the buffer object.
+ */
+void radeon_mn_unregister(struct radeon_bo *bo)
+{
+	struct radeon_device *rdev = bo->rdev;
+	struct radeon_mn *rmn;
+
+	mutex_lock(&rdev->mn_lock);
+	rmn = bo->mn;
+	if (rmn == NULL) {
+		mutex_unlock(&rdev->mn_lock);
+		return;
+	}
+
+	mutex_lock(&rmn->lock);
+	interval_tree_remove(&bo->mn_it, &rmn->objects);
+	bo->mn = NULL;
+	mutex_unlock(&rmn->lock);
+	mutex_unlock(&rdev->mn_lock);
+}
diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h
index e27608c29c11..390db897f322 100644
--- a/drivers/gpu/drm/radeon/radeon_mode.h
+++ b/drivers/gpu/drm/radeon/radeon_mode.h
@@ -321,6 +321,10 @@ struct radeon_crtc {
 	uint32_t crtc_offset;
 	struct drm_gem_object *cursor_bo;
 	uint64_t cursor_addr;
+	int cursor_x;
+	int cursor_y;
+	int cursor_hot_x;
+	int cursor_hot_y;
 	int cursor_width;
 	int cursor_height;
 	int max_cursor_width;
@@ -462,6 +466,7 @@ struct radeon_gpio_rec {
 	u8 id;
 	u32 reg;
 	u32 mask;
+	u32 shift;
 };
 
 struct radeon_hpd {
@@ -748,6 +753,8 @@ extern bool radeon_atombios_get_ppll_ss_info(struct radeon_device *rdev,
 extern bool radeon_atombios_get_asic_ss_info(struct radeon_device *rdev,
 					     struct radeon_atom_ss *ss,
 					     int id, u32 clock);
+extern struct radeon_gpio_rec radeon_atombios_lookup_gpio(struct radeon_device *rdev,
+							  u8 id);
 
 extern void radeon_compute_pll_legacy(struct radeon_pll *pll,
 				      uint64_t freq,
@@ -777,6 +784,7 @@ extern void atombios_digital_setup(struct drm_encoder *encoder, int action);
 extern int atombios_get_encoder_mode(struct drm_encoder *encoder);
 extern bool atombios_set_edp_panel_power(struct drm_connector *connector, int action);
 extern void radeon_encoder_set_active_device(struct drm_encoder *encoder);
+extern bool radeon_encoder_is_digital(struct drm_encoder *encoder);
 
 extern void radeon_crtc_load_lut(struct drm_crtc *crtc);
 extern int atombios_crtc_set_base(struct drm_crtc *crtc, int x, int y,
@@ -801,13 +809,16 @@ extern int radeon_crtc_set_base_atomic(struct drm_crtc *crtc,
 extern int radeon_crtc_do_set_base(struct drm_crtc *crtc,
 				   struct drm_framebuffer *fb,
 				   int x, int y, int atomic);
-extern int radeon_crtc_cursor_set(struct drm_crtc *crtc,
-				  struct drm_file *file_priv,
-				  uint32_t handle,
-				  uint32_t width,
-				  uint32_t height);
+extern int radeon_crtc_cursor_set2(struct drm_crtc *crtc,
+				   struct drm_file *file_priv,
+				   uint32_t handle,
+				   uint32_t width,
+				   uint32_t height,
+				   int32_t hot_x,
+				   int32_t hot_y);
 extern int radeon_crtc_cursor_move(struct drm_crtc *crtc,
 				   int x, int y);
+extern void radeon_cursor_reset(struct drm_crtc *crtc);
 
 extern int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc,
 				      unsigned int flags,
diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c
index 480c87d8edc5..86fc56434b28 100644
--- a/drivers/gpu/drm/radeon/radeon_object.c
+++ b/drivers/gpu/drm/radeon/radeon_object.c
@@ -75,6 +75,7 @@ static void radeon_ttm_bo_destroy(struct ttm_buffer_object *tbo)
 	bo = container_of(tbo, struct radeon_bo, tbo);
 
 	radeon_update_memory_usage(bo, bo->tbo.mem.mem_type, -1);
+	radeon_mn_unregister(bo);
 
 	mutex_lock(&bo->rdev->gem.mutex);
 	list_del_init(&bo->list);
@@ -96,40 +97,83 @@ void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain)
 {
 	u32 c = 0, i;
 
-	rbo->placement.fpfn = 0;
-	rbo->placement.lpfn = 0;
 	rbo->placement.placement = rbo->placements;
 	rbo->placement.busy_placement = rbo->placements;
-	if (domain & RADEON_GEM_DOMAIN_VRAM)
-		rbo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED |
-					TTM_PL_FLAG_VRAM;
+	if (domain & RADEON_GEM_DOMAIN_VRAM) {
+		/* Try placing BOs which don't need CPU access outside of the
+		 * CPU accessible part of VRAM
+		 */
+		if ((rbo->flags & RADEON_GEM_NO_CPU_ACCESS) &&
+		    rbo->rdev->mc.visible_vram_size < rbo->rdev->mc.real_vram_size) {
+			rbo->placements[c].fpfn =
+				rbo->rdev->mc.visible_vram_size >> PAGE_SHIFT;
+			rbo->placements[c++].flags = TTM_PL_FLAG_WC |
+						     TTM_PL_FLAG_UNCACHED |
+						     TTM_PL_FLAG_VRAM;
+		}
+
+		rbo->placements[c].fpfn = 0;
+		rbo->placements[c++].flags = TTM_PL_FLAG_WC |
+					     TTM_PL_FLAG_UNCACHED |
+					     TTM_PL_FLAG_VRAM;
+	}
+
 	if (domain & RADEON_GEM_DOMAIN_GTT) {
 		if (rbo->flags & RADEON_GEM_GTT_UC) {
-			rbo->placements[c++] = TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_TT;
+			rbo->placements[c].fpfn = 0;
+			rbo->placements[c++].flags = TTM_PL_FLAG_UNCACHED |
+				TTM_PL_FLAG_TT;
+
 		} else if ((rbo->flags & RADEON_GEM_GTT_WC) ||
 			   (rbo->rdev->flags & RADEON_IS_AGP)) {
-			rbo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED |
+			rbo->placements[c].fpfn = 0;
+			rbo->placements[c++].flags = TTM_PL_FLAG_WC |
+				TTM_PL_FLAG_UNCACHED |
 				TTM_PL_FLAG_TT;
 		} else {
-			rbo->placements[c++] = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_TT;
+			rbo->placements[c].fpfn = 0;
+			rbo->placements[c++].flags = TTM_PL_FLAG_CACHED |
+						     TTM_PL_FLAG_TT;
 		}
 	}
+
 	if (domain & RADEON_GEM_DOMAIN_CPU) {
 		if (rbo->flags & RADEON_GEM_GTT_UC) {
-			rbo->placements[c++] = TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_SYSTEM;
+			rbo->placements[c].fpfn = 0;
+			rbo->placements[c++].flags = TTM_PL_FLAG_UNCACHED |
+				TTM_PL_FLAG_SYSTEM;
+
 		} else if ((rbo->flags & RADEON_GEM_GTT_WC) ||
 		    rbo->rdev->flags & RADEON_IS_AGP) {
-			rbo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED |
+			rbo->placements[c].fpfn = 0;
+			rbo->placements[c++].flags = TTM_PL_FLAG_WC |
+				TTM_PL_FLAG_UNCACHED |
 				TTM_PL_FLAG_SYSTEM;
 		} else {
-			rbo->placements[c++] = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_SYSTEM;
+			rbo->placements[c].fpfn = 0;
+			rbo->placements[c++].flags = TTM_PL_FLAG_CACHED |
+						     TTM_PL_FLAG_SYSTEM;
 		}
 	}
-	if (!c)
-		rbo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+	if (!c) {
+		rbo->placements[c].fpfn = 0;
+		rbo->placements[c++].flags = TTM_PL_MASK_CACHING |
+					     TTM_PL_FLAG_SYSTEM;
+	}
+
 	rbo->placement.num_placement = c;
 	rbo->placement.num_busy_placement = c;
 
+	for (i = 0; i < c; ++i) {
+		if ((rbo->flags & RADEON_GEM_CPU_ACCESS) &&
+		    (rbo->placements[i].flags & TTM_PL_FLAG_VRAM) &&
+		    !rbo->placements[i].fpfn)
+			rbo->placements[i].lpfn =
+				rbo->rdev->mc.visible_vram_size >> PAGE_SHIFT;
+		else
+			rbo->placements[i].lpfn = 0;
+	}
+
 	/*
 	 * Use two-ended allocation depending on the buffer size to
 	 * improve fragmentation quality.
@@ -137,14 +181,16 @@ void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain)
 	 */
 	if (rbo->tbo.mem.size > 512 * 1024) {
 		for (i = 0; i < c; i++) {
-			rbo->placements[i] |= TTM_PL_FLAG_TOPDOWN;
+			rbo->placements[i].flags |= TTM_PL_FLAG_TOPDOWN;
 		}
 	}
 }
 
 int radeon_bo_create(struct radeon_device *rdev,
-		     unsigned long size, int byte_align, bool kernel, u32 domain,
-		     u32 flags, struct sg_table *sg, struct radeon_bo **bo_ptr)
+		     unsigned long size, int byte_align, bool kernel,
+		     u32 domain, u32 flags, struct sg_table *sg,
+		     struct reservation_object *resv,
+		     struct radeon_bo **bo_ptr)
 {
 	struct radeon_bo *bo;
 	enum ttm_bo_type type;
@@ -187,12 +233,19 @@ int radeon_bo_create(struct radeon_device *rdev,
 	if (!(rdev->flags & RADEON_IS_PCIE))
 		bo->flags &= ~(RADEON_GEM_GTT_WC | RADEON_GEM_GTT_UC);
 
+#ifdef CONFIG_X86_32
+	/* XXX: Write-combined CPU mappings of GTT seem broken on 32-bit
+	 * See https://bugs.freedesktop.org/show_bug.cgi?id=84627
+	 */
+	bo->flags &= ~RADEON_GEM_GTT_WC;
+#endif
+
 	radeon_ttm_placement_from_domain(bo, domain);
 	/* Kernel allocation are uninterruptible */
 	down_read(&rdev->pm.mclk_lock);
 	r = ttm_bo_init(&rdev->mman.bdev, &bo->tbo, size, type,
 			&bo->placement, page_align, !kernel, NULL,
-			acc_size, sg, &radeon_ttm_bo_destroy);
+			acc_size, sg, resv, &radeon_ttm_bo_destroy);
 	up_read(&rdev->pm.mclk_lock);
 	if (unlikely(r != 0)) {
 		return r;
@@ -264,6 +317,9 @@ int radeon_bo_pin_restricted(struct radeon_bo *bo, u32 domain, u64 max_offset,
 {
 	int r, i;
 
+	if (radeon_ttm_tt_has_userptr(bo->tbo.ttm))
+		return -EPERM;
+
 	if (bo->pin_count) {
 		bo->pin_count++;
 		if (gpu_addr)
@@ -283,21 +339,19 @@ int radeon_bo_pin_restricted(struct radeon_bo *bo, u32 domain, u64 max_offset,
 		return 0;
 	}
 	radeon_ttm_placement_from_domain(bo, domain);
-	if (domain == RADEON_GEM_DOMAIN_VRAM) {
+	for (i = 0; i < bo->placement.num_placement; i++) {
 		/* force to pin into visible video ram */
-		bo->placement.lpfn = bo->rdev->mc.visible_vram_size >> PAGE_SHIFT;
-	}
-	if (max_offset) {
-		u64 lpfn = max_offset >> PAGE_SHIFT;
-
-		if (!bo->placement.lpfn)
-			bo->placement.lpfn = bo->rdev->mc.gtt_size >> PAGE_SHIFT;
+		if ((bo->placements[i].flags & TTM_PL_FLAG_VRAM) &&
+		    !(bo->flags & RADEON_GEM_NO_CPU_ACCESS) &&
+		    (!max_offset || max_offset > bo->rdev->mc.visible_vram_size))
+			bo->placements[i].lpfn =
+				bo->rdev->mc.visible_vram_size >> PAGE_SHIFT;
+		else
+			bo->placements[i].lpfn = max_offset >> PAGE_SHIFT;
 
-		if (lpfn < bo->placement.lpfn)
-			bo->placement.lpfn = lpfn;
+		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 	}
-	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+
 	r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
 	if (likely(r == 0)) {
 		bo->pin_count = 1;
@@ -329,8 +383,10 @@ int radeon_bo_unpin(struct radeon_bo *bo)
 	bo->pin_count--;
 	if (bo->pin_count)
 		return 0;
-	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] &= ~TTM_PL_FLAG_NO_EVICT;
+	for (i = 0; i < bo->placement.num_placement; i++) {
+		bo->placements[i].lpfn = 0;
+		bo->placements[i].flags &= ~TTM_PL_FLAG_NO_EVICT;
+	}
 	r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
 	if (likely(r == 0)) {
 		if (bo->tbo.mem.mem_type == TTM_PL_VRAM)
@@ -453,21 +509,23 @@ int radeon_bo_list_validate(struct radeon_device *rdev,
 			    struct ww_acquire_ctx *ticket,
 			    struct list_head *head, int ring)
 {
-	struct radeon_cs_reloc *lobj;
-	struct radeon_bo *bo;
+	struct radeon_bo_list *lobj;
+	struct list_head duplicates;
 	int r;
 	u64 bytes_moved = 0, initial_bytes_moved;
 	u64 bytes_moved_threshold = radeon_bo_get_threshold_for_moves(rdev);
 
-	r = ttm_eu_reserve_buffers(ticket, head);
+	INIT_LIST_HEAD(&duplicates);
+	r = ttm_eu_reserve_buffers(ticket, head, true, &duplicates);
 	if (unlikely(r != 0)) {
 		return r;
 	}
 
 	list_for_each_entry(lobj, head, tv.head) {
-		bo = lobj->robj;
+		struct radeon_bo *bo = lobj->robj;
 		if (!bo->pin_count) {
 			u32 domain = lobj->prefered_domains;
+			u32 allowed = lobj->allowed_domains;
 			u32 current_domain =
 				radeon_mem_type_to_domain(bo->tbo.mem.mem_type);
 
@@ -479,7 +537,7 @@ int radeon_bo_list_validate(struct radeon_device *rdev,
 			 * into account. We don't want to disallow buffer moves
 			 * completely.
 			 */
-			if ((lobj->allowed_domains & current_domain) != 0 &&
+			if ((allowed & current_domain) != 0 &&
 			    (domain & current_domain) == 0 && /* will be moved */
 			    bytes_moved > bytes_moved_threshold) {
 				/* don't move it */
@@ -489,7 +547,7 @@ int radeon_bo_list_validate(struct radeon_device *rdev,
 		retry:
 			radeon_ttm_placement_from_domain(bo, domain);
 			if (ring == R600_RING_TYPE_UVD_INDEX)
-				radeon_uvd_force_into_uvd_segment(bo);
+				radeon_uvd_force_into_uvd_segment(bo, allowed);
 
 			initial_bytes_moved = atomic64_read(&rdev->num_bytes_moved);
 			r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
@@ -509,6 +567,12 @@ int radeon_bo_list_validate(struct radeon_device *rdev,
 		lobj->gpu_offset = radeon_bo_gpu_offset(bo);
 		lobj->tiling_flags = bo->tiling_flags;
 	}
+
+	list_for_each_entry(lobj, &duplicates, tv.head) {
+		lobj->gpu_offset = radeon_bo_gpu_offset(lobj->robj);
+		lobj->tiling_flags = lobj->robj->tiling_flags;
+	}
+
 	return 0;
 }
 
@@ -713,8 +777,8 @@ int radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo)
 {
 	struct radeon_device *rdev;
 	struct radeon_bo *rbo;
-	unsigned long offset, size;
-	int r;
+	unsigned long offset, size, lpfn;
+	int i, r;
 
 	if (!radeon_ttm_bo_is_radeon_bo(bo))
 		return 0;
@@ -731,7 +795,13 @@ int radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo)
 
 	/* hurrah the memory is not visible ! */
 	radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_VRAM);
-	rbo->placement.lpfn = rdev->mc.visible_vram_size >> PAGE_SHIFT;
+	lpfn =	rdev->mc.visible_vram_size >> PAGE_SHIFT;
+	for (i = 0; i < rbo->placement.num_placement; i++) {
+		/* Force into visible VRAM */
+		if ((rbo->placements[i].flags & TTM_PL_FLAG_VRAM) &&
+		    (!rbo->placements[i].lpfn || rbo->placements[i].lpfn > lpfn))
+			rbo->placements[i].lpfn = lpfn;
+	}
 	r = ttm_bo_validate(bo, &rbo->placement, false, false);
 	if (unlikely(r == -ENOMEM)) {
 		radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_GTT);
@@ -755,12 +825,29 @@ int radeon_bo_wait(struct radeon_bo *bo, u32 *mem_type, bool no_wait)
 	r = ttm_bo_reserve(&bo->tbo, true, no_wait, false, NULL);
 	if (unlikely(r != 0))
 		return r;
-	spin_lock(&bo->tbo.bdev->fence_lock);
 	if (mem_type)
 		*mem_type = bo->tbo.mem.mem_type;
-	if (bo->tbo.sync_obj)
-		r = ttm_bo_wait(&bo->tbo, true, true, no_wait);
-	spin_unlock(&bo->tbo.bdev->fence_lock);
+
+	r = ttm_bo_wait(&bo->tbo, true, true, no_wait);
 	ttm_bo_unreserve(&bo->tbo);
 	return r;
 }
+
+/**
+ * radeon_bo_fence - add fence to buffer object
+ *
+ * @bo: buffer object in question
+ * @fence: fence to add
+ * @shared: true if fence should be added shared
+ *
+ */
+void radeon_bo_fence(struct radeon_bo *bo, struct radeon_fence *fence,
+                     bool shared)
+{
+	struct reservation_object *resv = bo->tbo.resv;
+
+	if (shared)
+		reservation_object_add_shared_fence(resv, &fence->base);
+	else
+		reservation_object_add_excl_fence(resv, &fence->base);
+}
diff --git a/drivers/gpu/drm/radeon/radeon_object.h b/drivers/gpu/drm/radeon/radeon_object.h
index 98a47fdf3625..3b0b377f76cb 100644
--- a/drivers/gpu/drm/radeon/radeon_object.h
+++ b/drivers/gpu/drm/radeon/radeon_object.h
@@ -126,6 +126,7 @@ extern int radeon_bo_create(struct radeon_device *rdev,
 			    unsigned long size, int byte_align,
 			    bool kernel, u32 domain, u32 flags,
 			    struct sg_table *sg,
+			    struct reservation_object *resv,
 			    struct radeon_bo **bo_ptr);
 extern int radeon_bo_kmap(struct radeon_bo *bo, void **ptr);
 extern void radeon_bo_kunmap(struct radeon_bo *bo);
@@ -154,6 +155,8 @@ extern void radeon_bo_move_notify(struct ttm_buffer_object *bo,
 				  struct ttm_mem_reg *new_mem);
 extern int radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo);
 extern int radeon_bo_get_surface_reg(struct radeon_bo *bo);
+extern void radeon_bo_fence(struct radeon_bo *bo, struct radeon_fence *fence,
+			    bool shared);
 
 /*
  * sub allocation
diff --git a/drivers/gpu/drm/radeon/radeon_pm.c b/drivers/gpu/drm/radeon/radeon_pm.c
index 164898b0010c..f7da8fe96a66 100644
--- a/drivers/gpu/drm/radeon/radeon_pm.c
+++ b/drivers/gpu/drm/radeon/radeon_pm.c
@@ -1287,8 +1287,39 @@ dpm_failed:
 	return ret;
 }
 
+struct radeon_dpm_quirk {
+	u32 chip_vendor;
+	u32 chip_device;
+	u32 subsys_vendor;
+	u32 subsys_device;
+};
+
+/* cards with dpm stability problems */
+static struct radeon_dpm_quirk radeon_dpm_quirk_list[] = {
+	/* TURKS - https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1386534 */
+	{ PCI_VENDOR_ID_ATI, 0x6759, 0x1682, 0x3195 },
+	/* TURKS - https://bugzilla.kernel.org/show_bug.cgi?id=83731 */
+	{ PCI_VENDOR_ID_ATI, 0x6840, 0x1179, 0xfb81 },
+	{ 0, 0, 0, 0 },
+};
+
 int radeon_pm_init(struct radeon_device *rdev)
 {
+	struct radeon_dpm_quirk *p = radeon_dpm_quirk_list;
+	bool disable_dpm = false;
+
+	/* Apply dpm quirks */
+	while (p && p->chip_device != 0) {
+		if (rdev->pdev->vendor == p->chip_vendor &&
+		    rdev->pdev->device == p->chip_device &&
+		    rdev->pdev->subsystem_vendor == p->subsys_vendor &&
+		    rdev->pdev->subsystem_device == p->subsys_device) {
+			disable_dpm = true;
+			break;
+		}
+		++p;
+	}
+
 	/* enable dpm on rv6xx+ */
 	switch (rdev->family) {
 	case CHIP_RV610:
@@ -1344,6 +1375,8 @@ int radeon_pm_init(struct radeon_device *rdev)
 			 (!(rdev->flags & RADEON_IS_IGP)) &&
 			 (!rdev->smc_fw))
 			rdev->pm.pm_method = PM_METHOD_PROFILE;
+		else if (disable_dpm && (radeon_dpm == -1))
+			rdev->pm.pm_method = PM_METHOD_PROFILE;
 		else if (radeon_dpm == 0)
 			rdev->pm.pm_method = PM_METHOD_PROFILE;
 		else
@@ -1556,7 +1589,7 @@ static bool radeon_pm_in_vbl(struct radeon_device *rdev)
 		if (rdev->pm.active_crtcs & (1 << crtc)) {
 			vbl_status = radeon_get_crtc_scanoutpos(rdev->ddev, crtc, 0, &vpos, &hpos, NULL, NULL);
 			if ((vbl_status & DRM_SCANOUTPOS_VALID) &&
-			    !(vbl_status & DRM_SCANOUTPOS_INVBL))
+			    !(vbl_status & DRM_SCANOUTPOS_IN_VBLANK))
 				in_vbl = false;
 		}
 	}
diff --git a/drivers/gpu/drm/radeon/radeon_prime.c b/drivers/gpu/drm/radeon/radeon_prime.c
index 0b16f2cbcf17..f3609c97496b 100644
--- a/drivers/gpu/drm/radeon/radeon_prime.c
+++ b/drivers/gpu/drm/radeon/radeon_prime.c
@@ -27,6 +27,7 @@
 
 #include "radeon.h"
 #include <drm/radeon_drm.h>
+#include <linux/dma-buf.h>
 
 struct sg_table *radeon_gem_prime_get_sg_table(struct drm_gem_object *obj)
 {
@@ -57,15 +58,18 @@ void radeon_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr)
 }
 
 struct drm_gem_object *radeon_gem_prime_import_sg_table(struct drm_device *dev,
-							size_t size,
+							struct dma_buf_attachment *attach,
 							struct sg_table *sg)
 {
+	struct reservation_object *resv = attach->dmabuf->resv;
 	struct radeon_device *rdev = dev->dev_private;
 	struct radeon_bo *bo;
 	int ret;
 
-	ret = radeon_bo_create(rdev, size, PAGE_SIZE, false,
-			       RADEON_GEM_DOMAIN_GTT, 0, sg, &bo);
+	ww_mutex_lock(&resv->lock, NULL);
+	ret = radeon_bo_create(rdev, attach->dmabuf->size, PAGE_SIZE, false,
+			       RADEON_GEM_DOMAIN_GTT, 0, sg, resv, &bo);
+	ww_mutex_unlock(&resv->lock);
 	if (ret)
 		return ERR_PTR(ret);
 
@@ -111,3 +115,13 @@ struct reservation_object *radeon_gem_prime_res_obj(struct drm_gem_object *obj)
 
 	return bo->tbo.resv;
 }
+
+struct dma_buf *radeon_gem_prime_export(struct drm_device *dev,
+					struct drm_gem_object *gobj,
+					int flags)
+{
+	struct radeon_bo *bo = gem_to_radeon_bo(gobj);
+	if (radeon_ttm_tt_has_userptr(bo->tbo.ttm))
+		return ERR_PTR(-EPERM);
+	return drm_gem_prime_export(dev, gobj, flags);
+}
diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c
index d65607902537..2456f69efd23 100644
--- a/drivers/gpu/drm/radeon/radeon_ring.c
+++ b/drivers/gpu/drm/radeon/radeon_ring.c
@@ -45,27 +45,6 @@
 static int radeon_debugfs_ring_init(struct radeon_device *rdev, struct radeon_ring *ring);
 
 /**
- * radeon_ring_write - write a value to the ring
- *
- * @ring: radeon_ring structure holding ring information
- * @v: dword (dw) value to write
- *
- * Write a value to the requested ring buffer (all asics).
- */
-void radeon_ring_write(struct radeon_ring *ring, uint32_t v)
-{
-#if DRM_DEBUG_CODE
-	if (ring->count_dw <= 0) {
-		DRM_ERROR("radeon: writing more dwords to the ring than expected!\n");
-	}
-#endif
-	ring->ring[ring->wptr++] = v;
-	ring->wptr &= ring->ptr_mask;
-	ring->count_dw--;
-	ring->ring_free_dw--;
-}
-
-/**
  * radeon_ring_supports_scratch_reg - check if the ring supports
  * writing to scratch registers
  *
@@ -335,7 +314,7 @@ unsigned radeon_ring_backup(struct radeon_device *rdev, struct radeon_ring *ring
 	}
 
 	/* and then save the content of the ring */
-	*data = kmalloc_array(size, sizeof(uint32_t), GFP_KERNEL);
+	*data = drm_malloc_ab(size, sizeof(uint32_t));
 	if (!*data) {
 		mutex_unlock(&rdev->ring_lock);
 		return 0;
@@ -377,7 +356,7 @@ int radeon_ring_restore(struct radeon_device *rdev, struct radeon_ring *ring,
 	}
 
 	radeon_ring_unlock_commit(rdev, ring, false);
-	kfree(data);
+	drm_free_large(data);
 	return 0;
 }
 
@@ -404,7 +383,7 @@ int radeon_ring_init(struct radeon_device *rdev, struct radeon_ring *ring, unsig
 	/* Allocate ring buffer */
 	if (ring->ring_obj == NULL) {
 		r = radeon_bo_create(rdev, ring->ring_size, PAGE_SIZE, true,
-				     RADEON_GEM_DOMAIN_GTT, 0,
+				     RADEON_GEM_DOMAIN_GTT, 0, NULL,
 				     NULL, &ring->ring_obj);
 		if (r) {
 			dev_err(rdev->dev, "(%d) ring create failed\n", r);
diff --git a/drivers/gpu/drm/radeon/radeon_sa.c b/drivers/gpu/drm/radeon/radeon_sa.c
index b84f97c8718c..c507896aca45 100644
--- a/drivers/gpu/drm/radeon/radeon_sa.c
+++ b/drivers/gpu/drm/radeon/radeon_sa.c
@@ -65,7 +65,7 @@ int radeon_sa_bo_manager_init(struct radeon_device *rdev,
 	}
 
 	r = radeon_bo_create(rdev, size, align, true,
-			     domain, flags, NULL, &sa_manager->bo);
+			     domain, flags, NULL, NULL, &sa_manager->bo);
 	if (r) {
 		dev_err(rdev->dev, "(%d) failed to allocate bo for manager\n", r);
 		return r;
diff --git a/drivers/gpu/drm/radeon/radeon_semaphore.c b/drivers/gpu/drm/radeon/radeon_semaphore.c
index abd6753a570a..e6ad54cdfa62 100644
--- a/drivers/gpu/drm/radeon/radeon_semaphore.c
+++ b/drivers/gpu/drm/radeon/radeon_semaphore.c
@@ -34,15 +34,14 @@
 int radeon_semaphore_create(struct radeon_device *rdev,
 			    struct radeon_semaphore **semaphore)
 {
-	uint64_t *cpu_addr;
-	int i, r;
+	int r;
 
 	*semaphore = kmalloc(sizeof(struct radeon_semaphore), GFP_KERNEL);
 	if (*semaphore == NULL) {
 		return -ENOMEM;
 	}
-	r = radeon_sa_bo_new(rdev, &rdev->ring_tmp_bo, &(*semaphore)->sa_bo,
-			     8 * RADEON_NUM_SYNCS, 8);
+	r = radeon_sa_bo_new(rdev, &rdev->ring_tmp_bo,
+			     &(*semaphore)->sa_bo, 8, 8);
 	if (r) {
 		kfree(*semaphore);
 		*semaphore = NULL;
@@ -51,12 +50,7 @@ int radeon_semaphore_create(struct radeon_device *rdev,
 	(*semaphore)->waiters = 0;
 	(*semaphore)->gpu_addr = radeon_sa_bo_gpu_addr((*semaphore)->sa_bo);
 
-	cpu_addr = radeon_sa_bo_cpu_addr((*semaphore)->sa_bo);
-	for (i = 0; i < RADEON_NUM_SYNCS; ++i)
-		cpu_addr[i] = 0;
-
-	for (i = 0; i < RADEON_NUM_RINGS; ++i)
-		(*semaphore)->sync_to[i] = NULL;
+	*((uint64_t *)radeon_sa_bo_cpu_addr((*semaphore)->sa_bo)) = 0;
 
 	return 0;
 }
@@ -95,99 +89,6 @@ bool radeon_semaphore_emit_wait(struct radeon_device *rdev, int ridx,
 	return false;
 }
 
-/**
- * radeon_semaphore_sync_to - use the semaphore to sync to a fence
- *
- * @semaphore: semaphore object to add fence to
- * @fence: fence to sync to
- *
- * Sync to the fence using this semaphore object
- */
-void radeon_semaphore_sync_to(struct radeon_semaphore *semaphore,
-			      struct radeon_fence *fence)
-{
-        struct radeon_fence *other;
-
-        if (!fence)
-                return;
-
-        other = semaphore->sync_to[fence->ring];
-        semaphore->sync_to[fence->ring] = radeon_fence_later(fence, other);
-}
-
-/**
- * radeon_semaphore_sync_rings - sync ring to all registered fences
- *
- * @rdev: radeon_device pointer
- * @semaphore: semaphore object to use for sync
- * @ring: ring that needs sync
- *
- * Ensure that all registered fences are signaled before letting
- * the ring continue. The caller must hold the ring lock.
- */
-int radeon_semaphore_sync_rings(struct radeon_device *rdev,
-				struct radeon_semaphore *semaphore,
-				int ring)
-{
-	unsigned count = 0;
-	int i, r;
-
-        for (i = 0; i < RADEON_NUM_RINGS; ++i) {
-		struct radeon_fence *fence = semaphore->sync_to[i];
-
-		/* check if we really need to sync */
-                if (!radeon_fence_need_sync(fence, ring))
-			continue;
-
-		/* prevent GPU deadlocks */
-		if (!rdev->ring[i].ready) {
-			dev_err(rdev->dev, "Syncing to a disabled ring!");
-			return -EINVAL;
-		}
-
-		if (++count > RADEON_NUM_SYNCS) {
-			/* not enough room, wait manually */
-			r = radeon_fence_wait(fence, false);
-			if (r)
-				return r;
-			continue;
-		}
-
-		/* allocate enough space for sync command */
-		r = radeon_ring_alloc(rdev, &rdev->ring[i], 16);
-		if (r) {
-			return r;
-		}
-
-		/* emit the signal semaphore */
-		if (!radeon_semaphore_emit_signal(rdev, i, semaphore)) {
-			/* signaling wasn't successful wait manually */
-			radeon_ring_undo(&rdev->ring[i]);
-			r = radeon_fence_wait(fence, false);
-			if (r)
-				return r;
-			continue;
-		}
-
-		/* we assume caller has already allocated space on waiters ring */
-		if (!radeon_semaphore_emit_wait(rdev, ring, semaphore)) {
-			/* waiting wasn't successful wait manually */
-			radeon_ring_undo(&rdev->ring[i]);
-			r = radeon_fence_wait(fence, false);
-			if (r)
-				return r;
-			continue;
-		}
-
-		radeon_ring_commit(rdev, &rdev->ring[i], false);
-		radeon_fence_note_sync(fence, ring);
-
-		semaphore->gpu_addr += 8;
-	}
-
-	return 0;
-}
-
 void radeon_semaphore_free(struct radeon_device *rdev,
 			   struct radeon_semaphore **semaphore,
 			   struct radeon_fence *fence)
diff --git a/drivers/gpu/drm/radeon/radeon_state.c b/drivers/gpu/drm/radeon/radeon_state.c
index 23bb64fd775f..15aee723db77 100644
--- a/drivers/gpu/drm/radeon/radeon_state.c
+++ b/drivers/gpu/drm/radeon/radeon_state.c
@@ -30,9 +30,9 @@
  */
 
 #include <drm/drmP.h>
-#include <drm/drm_buffer.h>
 #include <drm/radeon_drm.h>
 #include "radeon_drv.h"
+#include "drm_buffer.h"
 
 /* ================================================================
  * Helper functions for client state checking and fixup
@@ -1703,7 +1703,7 @@ static int radeon_cp_dispatch_texture(struct drm_device * dev,
 	u32 format;
 	u32 *buffer;
 	const u8 __user *data;
-	int size, dwords, tex_width, blit_width, spitch;
+	unsigned int size, dwords, tex_width, blit_width, spitch;
 	u32 height;
 	int i;
 	u32 texpitch, microtile;
diff --git a/drivers/gpu/drm/radeon/radeon_sync.c b/drivers/gpu/drm/radeon/radeon_sync.c
new file mode 100644
index 000000000000..02ac8a1de4ff
--- /dev/null
+++ b/drivers/gpu/drm/radeon/radeon_sync.c
@@ -0,0 +1,220 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ * All Rights Reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
+ * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+ * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+ * USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ */
+/*
+ * Authors:
+ *    Christian König <christian.koenig@amd.com>
+ */
+
+#include <drm/drmP.h>
+#include "radeon.h"
+#include "radeon_trace.h"
+
+/**
+ * radeon_sync_create - zero init sync object
+ *
+ * @sync: sync object to initialize
+ *
+ * Just clear the sync object for now.
+ */
+void radeon_sync_create(struct radeon_sync *sync)
+{
+	unsigned i;
+
+	for (i = 0; i < RADEON_NUM_SYNCS; ++i)
+		sync->semaphores[i] = NULL;
+
+	for (i = 0; i < RADEON_NUM_RINGS; ++i)
+		sync->sync_to[i] = NULL;
+
+	sync->last_vm_update = NULL;
+}
+
+/**
+ * radeon_sync_fence - use the semaphore to sync to a fence
+ *
+ * @sync: sync object to add fence to
+ * @fence: fence to sync to
+ *
+ * Sync to the fence using the semaphore objects
+ */
+void radeon_sync_fence(struct radeon_sync *sync,
+		       struct radeon_fence *fence)
+{
+	struct radeon_fence *other;
+
+	if (!fence)
+		return;
+
+	other = sync->sync_to[fence->ring];
+	sync->sync_to[fence->ring] = radeon_fence_later(fence, other);
+
+	if (fence->is_vm_update) {
+		other = sync->last_vm_update;
+		sync->last_vm_update = radeon_fence_later(fence, other);
+	}
+}
+
+/**
+ * radeon_sync_resv - use the semaphores to sync to a reservation object
+ *
+ * @sync: sync object to add fences from reservation object to
+ * @resv: reservation object with embedded fence
+ * @shared: true if we should only sync to the exclusive fence
+ *
+ * Sync to the fence using the semaphore objects
+ */
+int radeon_sync_resv(struct radeon_device *rdev,
+		     struct radeon_sync *sync,
+		     struct reservation_object *resv,
+		     bool shared)
+{
+	struct reservation_object_list *flist;
+	struct fence *f;
+	struct radeon_fence *fence;
+	unsigned i;
+	int r = 0;
+
+	/* always sync to the exclusive fence */
+	f = reservation_object_get_excl(resv);
+	fence = f ? to_radeon_fence(f) : NULL;
+	if (fence && fence->rdev == rdev)
+		radeon_sync_fence(sync, fence);
+	else if (f)
+		r = fence_wait(f, true);
+
+	flist = reservation_object_get_list(resv);
+	if (shared || !flist || r)
+		return r;
+
+	for (i = 0; i < flist->shared_count; ++i) {
+		f = rcu_dereference_protected(flist->shared[i],
+					      reservation_object_held(resv));
+		fence = to_radeon_fence(f);
+		if (fence && fence->rdev == rdev)
+			radeon_sync_fence(sync, fence);
+		else
+			r = fence_wait(f, true);
+
+		if (r)
+			break;
+	}
+	return r;
+}
+
+/**
+ * radeon_sync_rings - sync ring to all registered fences
+ *
+ * @rdev: radeon_device pointer
+ * @sync: sync object to use
+ * @ring: ring that needs sync
+ *
+ * Ensure that all registered fences are signaled before letting
+ * the ring continue. The caller must hold the ring lock.
+ */
+int radeon_sync_rings(struct radeon_device *rdev,
+		      struct radeon_sync *sync,
+		      int ring)
+{
+	unsigned count = 0;
+	int i, r;
+
+	for (i = 0; i < RADEON_NUM_RINGS; ++i) {
+		struct radeon_fence *fence = sync->sync_to[i];
+		struct radeon_semaphore *semaphore;
+
+		/* check if we really need to sync */
+		if (!radeon_fence_need_sync(fence, ring))
+			continue;
+
+		/* prevent GPU deadlocks */
+		if (!rdev->ring[i].ready) {
+			dev_err(rdev->dev, "Syncing to a disabled ring!");
+			return -EINVAL;
+		}
+
+		if (count >= RADEON_NUM_SYNCS) {
+			/* not enough room, wait manually */
+			r = radeon_fence_wait(fence, false);
+			if (r)
+				return r;
+			continue;
+		}
+		r = radeon_semaphore_create(rdev, &semaphore);
+		if (r)
+			return r;
+
+		sync->semaphores[count++] = semaphore;
+
+		/* allocate enough space for sync command */
+		r = radeon_ring_alloc(rdev, &rdev->ring[i], 16);
+		if (r)
+			return r;
+
+		/* emit the signal semaphore */
+		if (!radeon_semaphore_emit_signal(rdev, i, semaphore)) {
+			/* signaling wasn't successful wait manually */
+			radeon_ring_undo(&rdev->ring[i]);
+			r = radeon_fence_wait(fence, false);
+			if (r)
+				return r;
+			continue;
+		}
+
+		/* we assume caller has already allocated space on waiters ring */
+		if (!radeon_semaphore_emit_wait(rdev, ring, semaphore)) {
+			/* waiting wasn't successful wait manually */
+			radeon_ring_undo(&rdev->ring[i]);
+			r = radeon_fence_wait(fence, false);
+			if (r)
+				return r;
+			continue;
+		}
+
+		radeon_ring_commit(rdev, &rdev->ring[i], false);
+		radeon_fence_note_sync(fence, ring);
+	}
+
+	return 0;
+}
+
+/**
+ * radeon_sync_free - free the sync object
+ *
+ * @rdev: radeon_device pointer
+ * @sync: sync object to use
+ * @fence: fence to use for the free
+ *
+ * Free the sync object by freeing all semaphores in it.
+ */
+void radeon_sync_free(struct radeon_device *rdev,
+		      struct radeon_sync *sync,
+		      struct radeon_fence *fence)
+{
+	unsigned i;
+
+	for (i = 0; i < RADEON_NUM_SYNCS; ++i)
+		radeon_semaphore_free(rdev, &sync->semaphores[i], fence);
+}
diff --git a/drivers/gpu/drm/radeon/radeon_test.c b/drivers/gpu/drm/radeon/radeon_test.c
index 17bc3dced9f1..07b506b41008 100644
--- a/drivers/gpu/drm/radeon/radeon_test.c
+++ b/drivers/gpu/drm/radeon/radeon_test.c
@@ -67,7 +67,7 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag)
 	}
 
 	r = radeon_bo_create(rdev, size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM,
-			     0, NULL, &vram_obj);
+			     0, NULL, NULL, &vram_obj);
 	if (r) {
 		DRM_ERROR("Failed to create VRAM object\n");
 		goto out_cleanup;
@@ -87,7 +87,8 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag)
 		struct radeon_fence *fence = NULL;
 
 		r = radeon_bo_create(rdev, size, PAGE_SIZE, true,
-				     RADEON_GEM_DOMAIN_GTT, 0, NULL, gtt_obj + i);
+				     RADEON_GEM_DOMAIN_GTT, 0, NULL, NULL,
+				     gtt_obj + i);
 		if (r) {
 			DRM_ERROR("Failed to create GTT object %d\n", i);
 			goto out_lclean;
@@ -116,11 +117,16 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag)
 		radeon_bo_kunmap(gtt_obj[i]);
 
 		if (ring == R600_RING_TYPE_DMA_INDEX)
-			r = radeon_copy_dma(rdev, gtt_addr, vram_addr, size / RADEON_GPU_PAGE_SIZE, &fence);
+			fence = radeon_copy_dma(rdev, gtt_addr, vram_addr,
+						size / RADEON_GPU_PAGE_SIZE,
+						NULL);
 		else
-			r = radeon_copy_blit(rdev, gtt_addr, vram_addr, size / RADEON_GPU_PAGE_SIZE, &fence);
-		if (r) {
+			fence = radeon_copy_blit(rdev, gtt_addr, vram_addr,
+						 size / RADEON_GPU_PAGE_SIZE,
+						 NULL);
+		if (IS_ERR(fence)) {
 			DRM_ERROR("Failed GTT->VRAM copy %d\n", i);
+			r = PTR_ERR(fence);
 			goto out_lclean_unpin;
 		}
 
@@ -162,11 +168,16 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag)
 		radeon_bo_kunmap(vram_obj);
 
 		if (ring == R600_RING_TYPE_DMA_INDEX)
-			r = radeon_copy_dma(rdev, vram_addr, gtt_addr, size / RADEON_GPU_PAGE_SIZE, &fence);
+			fence = radeon_copy_dma(rdev, vram_addr, gtt_addr,
+						size / RADEON_GPU_PAGE_SIZE,
+						NULL);
 		else
-			r = radeon_copy_blit(rdev, vram_addr, gtt_addr, size / RADEON_GPU_PAGE_SIZE, &fence);
-		if (r) {
+			fence = radeon_copy_blit(rdev, vram_addr, gtt_addr,
+						 size / RADEON_GPU_PAGE_SIZE,
+						 NULL);
+		if (IS_ERR(fence)) {
 			DRM_ERROR("Failed VRAM->GTT copy %d\n", i);
+			r = PTR_ERR(fence);
 			goto out_lclean_unpin;
 		}
 
@@ -222,7 +233,7 @@ out_lclean:
 			radeon_bo_unreserve(gtt_obj[i]);
 			radeon_bo_unref(&gtt_obj[i]);
 		}
-		if (fence)
+		if (fence && !IS_ERR(fence))
 			radeon_fence_unref(&fence);
 		break;
 	}
diff --git a/drivers/gpu/drm/radeon/radeon_trace.h b/drivers/gpu/drm/radeon/radeon_trace.h
index 9db74a96ef61..ce075cb08cb2 100644
--- a/drivers/gpu/drm/radeon/radeon_trace.h
+++ b/drivers/gpu/drm/radeon/radeon_trace.h
@@ -38,7 +38,7 @@ TRACE_EVENT(radeon_cs,
 
 	    TP_fast_assign(
 			   __entry->ring = p->ring;
-			   __entry->dw = p->chunks[p->chunk_ib_idx].length_dw;
+			   __entry->dw = p->chunk_ib->length_dw;
 			   __entry->fences = radeon_fence_count_emitted(
 				p->rdev, p->ring);
 			   ),
diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c
index 72afe82a95c9..d02aa1d0f588 100644
--- a/drivers/gpu/drm/radeon/radeon_ttm.c
+++ b/drivers/gpu/drm/radeon/radeon_ttm.c
@@ -39,6 +39,8 @@
 #include <linux/seq_file.h>
 #include <linux/slab.h>
 #include <linux/swiotlb.h>
+#include <linux/swap.h>
+#include <linux/pagemap.h>
 #include <linux/debugfs.h>
 #include "radeon_reg.h"
 #include "radeon.h"
@@ -176,12 +178,15 @@ static int radeon_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 static void radeon_evict_flags(struct ttm_buffer_object *bo,
 				struct ttm_placement *placement)
 {
+	static struct ttm_place placements = {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM
+	};
+
 	struct radeon_bo *rbo;
-	static u32 placements = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
 
 	if (!radeon_ttm_bo_is_radeon_bo(bo)) {
-		placement->fpfn = 0;
-		placement->lpfn = 0;
 		placement->placement = &placements;
 		placement->busy_placement = &placements;
 		placement->num_placement = 1;
@@ -191,9 +196,32 @@ static void radeon_evict_flags(struct ttm_buffer_object *bo,
 	rbo = container_of(bo, struct radeon_bo, tbo);
 	switch (bo->mem.mem_type) {
 	case TTM_PL_VRAM:
-		if (rbo->rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready == false)
+		if (rbo->rdev->ring[radeon_copy_ring_index(rbo->rdev)].ready == false)
 			radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_CPU);
-		else
+		else if (rbo->rdev->mc.visible_vram_size < rbo->rdev->mc.real_vram_size &&
+			 bo->mem.start < (rbo->rdev->mc.visible_vram_size >> PAGE_SHIFT)) {
+			unsigned fpfn = rbo->rdev->mc.visible_vram_size >> PAGE_SHIFT;
+			int i;
+
+			/* Try evicting to the CPU inaccessible part of VRAM
+			 * first, but only set GTT as busy placement, so this
+			 * BO will be evicted to GTT rather than causing other
+			 * BOs to be evicted from VRAM
+			 */
+			radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_VRAM |
+							 RADEON_GEM_DOMAIN_GTT);
+			rbo->placement.num_busy_placement = 0;
+			for (i = 0; i < rbo->placement.num_placement; i++) {
+				if (rbo->placements[i].flags & TTM_PL_FLAG_VRAM) {
+					if (rbo->placements[0].fpfn < fpfn)
+						rbo->placements[0].fpfn = fpfn;
+				} else {
+					rbo->placement.busy_placement =
+						&rbo->placements[i];
+					rbo->placement.num_busy_placement = 1;
+				}
+			}
+		} else
 			radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_GTT);
 		break;
 	case TTM_PL_TT:
@@ -228,6 +256,7 @@ static int radeon_move_blit(struct ttm_buffer_object *bo,
 	struct radeon_device *rdev;
 	uint64_t old_start, new_start;
 	struct radeon_fence *fence;
+	unsigned num_pages;
 	int r, ridx;
 
 	rdev = radeon_get_rdev(bo->bdev);
@@ -264,13 +293,12 @@ static int radeon_move_blit(struct ttm_buffer_object *bo,
 
 	BUILD_BUG_ON((PAGE_SIZE % RADEON_GPU_PAGE_SIZE) != 0);
 
-	/* sync other rings */
-	fence = bo->sync_obj;
-	r = radeon_copy(rdev, old_start, new_start,
-			new_mem->num_pages * (PAGE_SIZE / RADEON_GPU_PAGE_SIZE), /* GPU pages */
-			&fence);
-	/* FIXME: handle copy error */
-	r = ttm_bo_move_accel_cleanup(bo, (void *)fence,
+	num_pages = new_mem->num_pages * (PAGE_SIZE / RADEON_GPU_PAGE_SIZE);
+	fence = radeon_copy(rdev, old_start, new_start, num_pages, bo->resv);
+	if (IS_ERR(fence))
+		return PTR_ERR(fence);
+
+	r = ttm_bo_move_accel_cleanup(bo, &fence->base,
 				      evict, no_wait_gpu, new_mem);
 	radeon_fence_unref(&fence);
 	return r;
@@ -284,20 +312,20 @@ static int radeon_move_vram_ram(struct ttm_buffer_object *bo,
 	struct radeon_device *rdev;
 	struct ttm_mem_reg *old_mem = &bo->mem;
 	struct ttm_mem_reg tmp_mem;
-	u32 placements;
+	struct ttm_place placements;
 	struct ttm_placement placement;
 	int r;
 
 	rdev = radeon_get_rdev(bo->bdev);
 	tmp_mem = *new_mem;
 	tmp_mem.mm_node = NULL;
-	placement.fpfn = 0;
-	placement.lpfn = 0;
 	placement.num_placement = 1;
 	placement.placement = &placements;
 	placement.num_busy_placement = 1;
 	placement.busy_placement = &placements;
-	placements = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT;
+	placements.fpfn = 0;
+	placements.lpfn = 0;
+	placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT;
 	r = ttm_bo_mem_space(bo, &placement, &tmp_mem,
 			     interruptible, no_wait_gpu);
 	if (unlikely(r)) {
@@ -332,19 +360,19 @@ static int radeon_move_ram_vram(struct ttm_buffer_object *bo,
 	struct ttm_mem_reg *old_mem = &bo->mem;
 	struct ttm_mem_reg tmp_mem;
 	struct ttm_placement placement;
-	u32 placements;
+	struct ttm_place placements;
 	int r;
 
 	rdev = radeon_get_rdev(bo->bdev);
 	tmp_mem = *new_mem;
 	tmp_mem.mm_node = NULL;
-	placement.fpfn = 0;
-	placement.lpfn = 0;
 	placement.num_placement = 1;
 	placement.placement = &placements;
 	placement.num_busy_placement = 1;
 	placement.busy_placement = &placements;
-	placements = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT;
+	placements.fpfn = 0;
+	placements.lpfn = 0;
+	placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT;
 	r = ttm_bo_mem_space(bo, &placement, &tmp_mem,
 			     interruptible, no_wait_gpu);
 	if (unlikely(r)) {
@@ -483,39 +511,108 @@ static void radeon_ttm_io_mem_free(struct ttm_bo_device *bdev, struct ttm_mem_re
 {
 }
 
-static int radeon_sync_obj_wait(void *sync_obj, bool lazy, bool interruptible)
-{
-	return radeon_fence_wait((struct radeon_fence *)sync_obj, interruptible);
-}
+/*
+ * TTM backend functions.
+ */
+struct radeon_ttm_tt {
+	struct ttm_dma_tt		ttm;
+	struct radeon_device		*rdev;
+	u64				offset;
 
-static int radeon_sync_obj_flush(void *sync_obj)
+	uint64_t			userptr;
+	struct mm_struct		*usermm;
+	uint32_t			userflags;
+};
+
+/* prepare the sg table with the user pages */
+static int radeon_ttm_tt_pin_userptr(struct ttm_tt *ttm)
 {
+	struct radeon_device *rdev = radeon_get_rdev(ttm->bdev);
+	struct radeon_ttm_tt *gtt = (void *)ttm;
+	unsigned pinned = 0, nents;
+	int r;
+
+	int write = !(gtt->userflags & RADEON_GEM_USERPTR_READONLY);
+	enum dma_data_direction direction = write ?
+		DMA_BIDIRECTIONAL : DMA_TO_DEVICE;
+
+	if (current->mm != gtt->usermm)
+		return -EPERM;
+
+	if (gtt->userflags & RADEON_GEM_USERPTR_ANONONLY) {
+		/* check that we only pin down anonymous memory
+		   to prevent problems with writeback */
+		unsigned long end = gtt->userptr + ttm->num_pages * PAGE_SIZE;
+		struct vm_area_struct *vma;
+		vma = find_vma(gtt->usermm, gtt->userptr);
+		if (!vma || vma->vm_file || vma->vm_end < end)
+			return -EPERM;
+	}
+
+	do {
+		unsigned num_pages = ttm->num_pages - pinned;
+		uint64_t userptr = gtt->userptr + pinned * PAGE_SIZE;
+		struct page **pages = ttm->pages + pinned;
+
+		r = get_user_pages(current, current->mm, userptr, num_pages,
+				   write, 0, pages, NULL);
+		if (r < 0)
+			goto release_pages;
+
+		pinned += r;
+
+	} while (pinned < ttm->num_pages);
+
+	r = sg_alloc_table_from_pages(ttm->sg, ttm->pages, ttm->num_pages, 0,
+				      ttm->num_pages << PAGE_SHIFT,
+				      GFP_KERNEL);
+	if (r)
+		goto release_sg;
+
+	r = -ENOMEM;
+	nents = dma_map_sg(rdev->dev, ttm->sg->sgl, ttm->sg->nents, direction);
+	if (nents != ttm->sg->nents)
+		goto release_sg;
+
+	drm_prime_sg_to_page_addr_arrays(ttm->sg, ttm->pages,
+					 gtt->ttm.dma_address, ttm->num_pages);
+
 	return 0;
-}
 
-static void radeon_sync_obj_unref(void **sync_obj)
-{
-	radeon_fence_unref((struct radeon_fence **)sync_obj);
-}
+release_sg:
+	kfree(ttm->sg);
 
-static void *radeon_sync_obj_ref(void *sync_obj)
-{
-	return radeon_fence_ref((struct radeon_fence *)sync_obj);
+release_pages:
+	release_pages(ttm->pages, pinned, 0);
+	return r;
 }
 
-static bool radeon_sync_obj_signaled(void *sync_obj)
+static void radeon_ttm_tt_unpin_userptr(struct ttm_tt *ttm)
 {
-	return radeon_fence_signaled((struct radeon_fence *)sync_obj);
-}
+	struct radeon_device *rdev = radeon_get_rdev(ttm->bdev);
+	struct radeon_ttm_tt *gtt = (void *)ttm;
+	struct scatterlist *sg;
+	int i;
 
-/*
- * TTM backend functions.
- */
-struct radeon_ttm_tt {
-	struct ttm_dma_tt		ttm;
-	struct radeon_device		*rdev;
-	u64				offset;
-};
+	int write = !(gtt->userflags & RADEON_GEM_USERPTR_READONLY);
+	enum dma_data_direction direction = write ?
+		DMA_BIDIRECTIONAL : DMA_TO_DEVICE;
+
+	/* free the sg table and pages again */
+	dma_unmap_sg(rdev->dev, ttm->sg->sgl, ttm->sg->nents, direction);
+
+	for_each_sg(ttm->sg->sgl, sg, ttm->sg->nents, i) {
+		struct page *page = sg_page(sg);
+
+		if (!(gtt->userflags & RADEON_GEM_USERPTR_READONLY))
+			set_page_dirty(page);
+
+		mark_page_accessed(page);
+		page_cache_release(page);
+	}
+
+	sg_free_table(ttm->sg);
+}
 
 static int radeon_ttm_backend_bind(struct ttm_tt *ttm,
 				   struct ttm_mem_reg *bo_mem)
@@ -525,6 +622,11 @@ static int radeon_ttm_backend_bind(struct ttm_tt *ttm,
 		RADEON_GART_PAGE_WRITE;
 	int r;
 
+	if (gtt->userptr) {
+		radeon_ttm_tt_pin_userptr(ttm);
+		flags &= ~RADEON_GART_PAGE_WRITE;
+	}
+
 	gtt->offset = (unsigned long)(bo_mem->start << PAGE_SHIFT);
 	if (!ttm->num_pages) {
 		WARN(1, "nothing to bind %lu pages for mreg %p back %p!\n",
@@ -547,6 +649,10 @@ static int radeon_ttm_backend_unbind(struct ttm_tt *ttm)
 	struct radeon_ttm_tt *gtt = (void *)ttm;
 
 	radeon_gart_unbind(gtt->rdev, gtt->offset, ttm->num_pages);
+
+	if (gtt->userptr)
+		radeon_ttm_tt_unpin_userptr(ttm);
+
 	return 0;
 }
 
@@ -592,10 +698,17 @@ static struct ttm_tt *radeon_ttm_tt_create(struct ttm_bo_device *bdev,
 	return &gtt->ttm.ttm;
 }
 
+static struct radeon_ttm_tt *radeon_ttm_tt_to_gtt(struct ttm_tt *ttm)
+{
+	if (!ttm || ttm->func != &radeon_backend_func)
+		return NULL;
+	return (struct radeon_ttm_tt *)ttm;
+}
+
 static int radeon_ttm_tt_populate(struct ttm_tt *ttm)
 {
+	struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm);
 	struct radeon_device *rdev;
-	struct radeon_ttm_tt *gtt = (void *)ttm;
 	unsigned i;
 	int r;
 	bool slave = !!(ttm->page_flags & TTM_PAGE_FLAG_SG);
@@ -603,6 +716,16 @@ static int radeon_ttm_tt_populate(struct ttm_tt *ttm)
 	if (ttm->state != tt_unpopulated)
 		return 0;
 
+	if (gtt && gtt->userptr) {
+		ttm->sg = kcalloc(1, sizeof(struct sg_table), GFP_KERNEL);
+		if (!ttm->sg)
+			return -ENOMEM;
+
+		ttm->page_flags |= TTM_PAGE_FLAG_SG;
+		ttm->state = tt_unbound;
+		return 0;
+	}
+
 	if (slave && ttm->sg) {
 		drm_prime_sg_to_page_addr_arrays(ttm->sg, ttm->pages,
 						 gtt->ttm.dma_address, ttm->num_pages);
@@ -648,10 +771,16 @@ static int radeon_ttm_tt_populate(struct ttm_tt *ttm)
 static void radeon_ttm_tt_unpopulate(struct ttm_tt *ttm)
 {
 	struct radeon_device *rdev;
-	struct radeon_ttm_tt *gtt = (void *)ttm;
+	struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm);
 	unsigned i;
 	bool slave = !!(ttm->page_flags & TTM_PAGE_FLAG_SG);
 
+	if (gtt && gtt->userptr) {
+		kfree(ttm->sg);
+		ttm->page_flags &= ~TTM_PAGE_FLAG_SG;
+		return;
+	}
+
 	if (slave)
 		return;
 
@@ -680,6 +809,40 @@ static void radeon_ttm_tt_unpopulate(struct ttm_tt *ttm)
 	ttm_pool_unpopulate(ttm);
 }
 
+int radeon_ttm_tt_set_userptr(struct ttm_tt *ttm, uint64_t addr,
+			      uint32_t flags)
+{
+	struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm);
+
+	if (gtt == NULL)
+		return -EINVAL;
+
+	gtt->userptr = addr;
+	gtt->usermm = current->mm;
+	gtt->userflags = flags;
+	return 0;
+}
+
+bool radeon_ttm_tt_has_userptr(struct ttm_tt *ttm)
+{
+	struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm);
+
+	if (gtt == NULL)
+		return false;
+
+	return !!gtt->userptr;
+}
+
+bool radeon_ttm_tt_is_readonly(struct ttm_tt *ttm)
+{
+	struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm);
+
+	if (gtt == NULL)
+		return false;
+
+	return !!(gtt->userflags & RADEON_GEM_USERPTR_READONLY);
+}
+
 static struct ttm_bo_driver radeon_bo_driver = {
 	.ttm_tt_create = &radeon_ttm_tt_create,
 	.ttm_tt_populate = &radeon_ttm_tt_populate,
@@ -689,11 +852,6 @@ static struct ttm_bo_driver radeon_bo_driver = {
 	.evict_flags = &radeon_evict_flags,
 	.move = &radeon_bo_move,
 	.verify_access = &radeon_verify_access,
-	.sync_obj_signaled = &radeon_sync_obj_signaled,
-	.sync_obj_wait = &radeon_sync_obj_wait,
-	.sync_obj_flush = &radeon_sync_obj_flush,
-	.sync_obj_unref = &radeon_sync_obj_unref,
-	.sync_obj_ref = &radeon_sync_obj_ref,
 	.move_notify = &radeon_bo_move_notify,
 	.fault_reserve_notify = &radeon_bo_fault_reserve_notify,
 	.io_mem_reserve = &radeon_ttm_io_mem_reserve,
@@ -730,7 +888,7 @@ int radeon_ttm_init(struct radeon_device *rdev)
 	radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 
 	r = radeon_bo_create(rdev, 256 * 1024, PAGE_SIZE, true,
-			     RADEON_GEM_DOMAIN_VRAM, 0,
+			     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
 			     NULL, &rdev->stollen_vga_memory);
 	if (r) {
 		return r;
@@ -828,7 +986,7 @@ int radeon_mmap(struct file *filp, struct vm_area_struct *vma)
 	int r;
 
 	if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET)) {
-		return drm_mmap(filp, vma);
+		return -EINVAL;
 	}
 
 	file_priv = filp->private_data;
diff --git a/drivers/gpu/drm/radeon/radeon_uvd.c b/drivers/gpu/drm/radeon/radeon_uvd.c
index 341848a14376..c10b2aec6450 100644
--- a/drivers/gpu/drm/radeon/radeon_uvd.c
+++ b/drivers/gpu/drm/radeon/radeon_uvd.c
@@ -40,12 +40,18 @@
 #define UVD_IDLE_TIMEOUT_MS	1000
 
 /* Firmware Names */
+#define FIRMWARE_R600		"radeon/R600_uvd.bin"
+#define FIRMWARE_RS780		"radeon/RS780_uvd.bin"
+#define FIRMWARE_RV770		"radeon/RV770_uvd.bin"
 #define FIRMWARE_RV710		"radeon/RV710_uvd.bin"
 #define FIRMWARE_CYPRESS	"radeon/CYPRESS_uvd.bin"
 #define FIRMWARE_SUMO		"radeon/SUMO_uvd.bin"
 #define FIRMWARE_TAHITI		"radeon/TAHITI_uvd.bin"
 #define FIRMWARE_BONAIRE	"radeon/BONAIRE_uvd.bin"
 
+MODULE_FIRMWARE(FIRMWARE_R600);
+MODULE_FIRMWARE(FIRMWARE_RS780);
+MODULE_FIRMWARE(FIRMWARE_RV770);
 MODULE_FIRMWARE(FIRMWARE_RV710);
 MODULE_FIRMWARE(FIRMWARE_CYPRESS);
 MODULE_FIRMWARE(FIRMWARE_SUMO);
@@ -63,6 +69,23 @@ int radeon_uvd_init(struct radeon_device *rdev)
 	INIT_DELAYED_WORK(&rdev->uvd.idle_work, radeon_uvd_idle_work_handler);
 
 	switch (rdev->family) {
+	case CHIP_RV610:
+	case CHIP_RV630:
+	case CHIP_RV670:
+	case CHIP_RV620:
+	case CHIP_RV635:
+		fw_name = FIRMWARE_R600;
+		break;
+
+	case CHIP_RS780:
+	case CHIP_RS880:
+		fw_name = FIRMWARE_RS780;
+		break;
+
+	case CHIP_RV770:
+		fw_name = FIRMWARE_RV770;
+		break;
+
 	case CHIP_RV710:
 	case CHIP_RV730:
 	case CHIP_RV740:
@@ -115,9 +138,11 @@ int radeon_uvd_init(struct radeon_device *rdev)
 	}
 
 	bo_size = RADEON_GPU_PAGE_ALIGN(rdev->uvd_fw->size + 8) +
-		  RADEON_UVD_STACK_SIZE + RADEON_UVD_HEAP_SIZE;
+		  RADEON_UVD_STACK_SIZE + RADEON_UVD_HEAP_SIZE +
+		  RADEON_GPU_PAGE_SIZE;
 	r = radeon_bo_create(rdev, bo_size, PAGE_SIZE, true,
-			     RADEON_GEM_DOMAIN_VRAM, 0, NULL, &rdev->uvd.vcpu_bo);
+			     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
+			     NULL, &rdev->uvd.vcpu_bo);
 	if (r) {
 		dev_err(rdev->dev, "(%d) failed to allocate UVD bo\n", r);
 		return r;
@@ -231,10 +256,30 @@ int radeon_uvd_resume(struct radeon_device *rdev)
 	return 0;
 }
 
-void radeon_uvd_force_into_uvd_segment(struct radeon_bo *rbo)
+void radeon_uvd_force_into_uvd_segment(struct radeon_bo *rbo,
+				       uint32_t allowed_domains)
 {
-	rbo->placement.fpfn = 0 >> PAGE_SHIFT;
-	rbo->placement.lpfn = (256 * 1024 * 1024) >> PAGE_SHIFT;
+	int i;
+
+	for (i = 0; i < rbo->placement.num_placement; ++i) {
+		rbo->placements[i].fpfn = 0 >> PAGE_SHIFT;
+		rbo->placements[i].lpfn = (256 * 1024 * 1024) >> PAGE_SHIFT;
+	}
+
+	/* If it must be in VRAM it must be in the first segment as well */
+	if (allowed_domains == RADEON_GEM_DOMAIN_VRAM)
+		return;
+
+	/* abort if we already have more than one placement */
+	if (rbo->placement.num_placement > 1)
+		return;
+
+	/* add another 256MB segment */
+	rbo->placements[1] = rbo->placements[0];
+	rbo->placements[1].fpfn += (256 * 1024 * 1024) >> PAGE_SHIFT;
+	rbo->placements[1].lpfn += (256 * 1024 * 1024) >> PAGE_SHIFT;
+	rbo->placement.num_placement++;
+	rbo->placement.num_busy_placement++;
 }
 
 void radeon_uvd_free_handles(struct radeon_device *rdev, struct drm_file *filp)
@@ -356,6 +401,7 @@ static int radeon_uvd_cs_msg(struct radeon_cs_parser *p, struct radeon_bo *bo,
 {
 	int32_t *msg, msg_type, handle;
 	unsigned img_size = 0;
+	struct fence *f;
 	void *ptr;
 
 	int i, r;
@@ -365,8 +411,9 @@ static int radeon_uvd_cs_msg(struct radeon_cs_parser *p, struct radeon_bo *bo,
 		return -EINVAL;
 	}
 
-	if (bo->tbo.sync_obj) {
-		r = radeon_fence_wait(bo->tbo.sync_obj, false);
+	f = reservation_object_get_excl(bo->tbo.resv);
+	if (f) {
+		r = radeon_fence_wait((struct radeon_fence *)f, false);
 		if (r) {
 			DRM_ERROR("Failed waiting for UVD message (%d)!\n", r);
 			return r;
@@ -441,12 +488,12 @@ static int radeon_uvd_cs_reloc(struct radeon_cs_parser *p,
 			       unsigned buf_sizes[], bool *has_msg_cmd)
 {
 	struct radeon_cs_chunk *relocs_chunk;
-	struct radeon_cs_reloc *reloc;
+	struct radeon_bo_list *reloc;
 	unsigned idx, cmd, offset;
 	uint64_t start, end;
 	int r;
 
-	relocs_chunk = &p->chunks[p->chunk_relocs_idx];
+	relocs_chunk = p->chunk_relocs;
 	offset = radeon_get_ib_value(p, data0);
 	idx = radeon_get_ib_value(p, data1);
 	if (idx >= relocs_chunk->length_dw) {
@@ -455,7 +502,7 @@ static int radeon_uvd_cs_reloc(struct radeon_cs_parser *p,
 		return -EINVAL;
 	}
 
-	reloc = p->relocs_ptr[(idx / 4)];
+	reloc = &p->relocs[(idx / 4)];
 	start = reloc->gpu_offset;
 	end = start + radeon_bo_size(reloc->robj);
 	start += offset;
@@ -563,13 +610,13 @@ int radeon_uvd_cs_parse(struct radeon_cs_parser *p)
 		[0x00000003]	=	2048,
 	};
 
-	if (p->chunks[p->chunk_ib_idx].length_dw % 16) {
+	if (p->chunk_ib->length_dw % 16) {
 		DRM_ERROR("UVD IB length (%d) not 16 dwords aligned!\n",
-			  p->chunks[p->chunk_ib_idx].length_dw);
+			  p->chunk_ib->length_dw);
 		return -EINVAL;
 	}
 
-	if (p->chunk_relocs_idx == -1) {
+	if (p->chunk_relocs == NULL) {
 		DRM_ERROR("No relocation chunk !\n");
 		return -EINVAL;
 	}
@@ -593,7 +640,7 @@ int radeon_uvd_cs_parse(struct radeon_cs_parser *p)
 			DRM_ERROR("Unknown packet type %d !\n", pkt.type);
 			return -EINVAL;
 		}
-	} while (p->idx < p->chunks[p->chunk_ib_idx].length_dw);
+	} while (p->idx < p->chunk_ib->length_dw);
 
 	if (!has_msg_cmd) {
 		DRM_ERROR("UVD-IBs need a msg command!\n");
@@ -604,38 +651,16 @@ int radeon_uvd_cs_parse(struct radeon_cs_parser *p)
 }
 
 static int radeon_uvd_send_msg(struct radeon_device *rdev,
-			       int ring, struct radeon_bo *bo,
+			       int ring, uint64_t addr,
 			       struct radeon_fence **fence)
 {
-	struct ttm_validate_buffer tv;
-	struct ww_acquire_ctx ticket;
-	struct list_head head;
 	struct radeon_ib ib;
-	uint64_t addr;
 	int i, r;
 
-	memset(&tv, 0, sizeof(tv));
-	tv.bo = &bo->tbo;
-
-	INIT_LIST_HEAD(&head);
-	list_add(&tv.head, &head);
-
-	r = ttm_eu_reserve_buffers(&ticket, &head);
-	if (r)
-		return r;
-
-	radeon_ttm_placement_from_domain(bo, RADEON_GEM_DOMAIN_VRAM);
-	radeon_uvd_force_into_uvd_segment(bo);
-
-	r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
-	if (r) 
-		goto err;
-
 	r = radeon_ib_get(rdev, ring, &ib, NULL, 64);
 	if (r)
-		goto err;
+		return r;
 
-	addr = radeon_bo_gpu_offset(bo);
 	ib.ptr[0] = PACKET0(UVD_GPCOM_VCPU_DATA0, 0);
 	ib.ptr[1] = addr;
 	ib.ptr[2] = PACKET0(UVD_GPCOM_VCPU_DATA1, 0);
@@ -647,19 +672,11 @@ static int radeon_uvd_send_msg(struct radeon_device *rdev,
 	ib.length_dw = 16;
 
 	r = radeon_ib_schedule(rdev, &ib, NULL, false);
-	if (r)
-		goto err;
-	ttm_eu_fence_buffer_objects(&ticket, &head, ib.fence);
 
 	if (fence)
 		*fence = radeon_fence_ref(ib.fence);
 
 	radeon_ib_free(rdev, &ib);
-	radeon_bo_unref(&bo);
-	return 0;
-
-err:
-	ttm_eu_backoff_reservation(&ticket, &head);
 	return r;
 }
 
@@ -669,27 +686,18 @@ err:
 int radeon_uvd_get_create_msg(struct radeon_device *rdev, int ring,
 			      uint32_t handle, struct radeon_fence **fence)
 {
-	struct radeon_bo *bo;
-	uint32_t *msg;
-	int r, i;
+	/* we use the last page of the vcpu bo for the UVD message */
+	uint64_t offs = radeon_bo_size(rdev->uvd.vcpu_bo) -
+		RADEON_GPU_PAGE_SIZE;
 
-	r = radeon_bo_create(rdev, 1024, PAGE_SIZE, true,
-			     RADEON_GEM_DOMAIN_VRAM, 0, NULL, &bo);
-	if (r)
-		return r;
+	uint32_t *msg = rdev->uvd.cpu_addr + offs;
+	uint64_t addr = rdev->uvd.gpu_addr + offs;
 
-	r = radeon_bo_reserve(bo, false);
-	if (r) {
-		radeon_bo_unref(&bo);
-		return r;
-	}
+	int r, i;
 
-	r = radeon_bo_kmap(bo, (void **)&msg);
-	if (r) {
-		radeon_bo_unreserve(bo);
-		radeon_bo_unref(&bo);
+	r = radeon_bo_reserve(rdev->uvd.vcpu_bo, true);
+	if (r)
 		return r;
-	}
 
 	/* stitch together an UVD create msg */
 	msg[0] = cpu_to_le32(0x00000de4);
@@ -706,36 +714,26 @@ int radeon_uvd_get_create_msg(struct radeon_device *rdev, int ring,
 	for (i = 11; i < 1024; ++i)
 		msg[i] = cpu_to_le32(0x0);
 
-	radeon_bo_kunmap(bo);
-	radeon_bo_unreserve(bo);
-
-	return radeon_uvd_send_msg(rdev, ring, bo, fence);
+	r = radeon_uvd_send_msg(rdev, ring, addr, fence);
+	radeon_bo_unreserve(rdev->uvd.vcpu_bo);
+	return r;
 }
 
 int radeon_uvd_get_destroy_msg(struct radeon_device *rdev, int ring,
 			       uint32_t handle, struct radeon_fence **fence)
 {
-	struct radeon_bo *bo;
-	uint32_t *msg;
-	int r, i;
+	/* we use the last page of the vcpu bo for the UVD message */
+	uint64_t offs = radeon_bo_size(rdev->uvd.vcpu_bo) -
+		RADEON_GPU_PAGE_SIZE;
 
-	r = radeon_bo_create(rdev, 1024, PAGE_SIZE, true,
-			     RADEON_GEM_DOMAIN_VRAM, 0, NULL, &bo);
-	if (r)
-		return r;
+	uint32_t *msg = rdev->uvd.cpu_addr + offs;
+	uint64_t addr = rdev->uvd.gpu_addr + offs;
 
-	r = radeon_bo_reserve(bo, false);
-	if (r) {
-		radeon_bo_unref(&bo);
-		return r;
-	}
+	int r, i;
 
-	r = radeon_bo_kmap(bo, (void **)&msg);
-	if (r) {
-		radeon_bo_unreserve(bo);
-		radeon_bo_unref(&bo);
+	r = radeon_bo_reserve(rdev->uvd.vcpu_bo, true);
+	if (r)
 		return r;
-	}
 
 	/* stitch together an UVD destroy msg */
 	msg[0] = cpu_to_le32(0x00000de4);
@@ -745,10 +743,9 @@ int radeon_uvd_get_destroy_msg(struct radeon_device *rdev, int ring,
 	for (i = 4; i < 1024; ++i)
 		msg[i] = cpu_to_le32(0x0);
 
-	radeon_bo_kunmap(bo);
-	radeon_bo_unreserve(bo);
-
-	return radeon_uvd_send_msg(rdev, ring, bo, fence);
+	r = radeon_uvd_send_msg(rdev, ring, addr, fence);
+	radeon_bo_unreserve(rdev->uvd.vcpu_bo);
+	return r;
 }
 
 /**
diff --git a/drivers/gpu/drm/radeon/radeon_vce.c b/drivers/gpu/drm/radeon/radeon_vce.c
index c7190aadbd89..976fe432f4e2 100644
--- a/drivers/gpu/drm/radeon/radeon_vce.c
+++ b/drivers/gpu/drm/radeon/radeon_vce.c
@@ -126,7 +126,8 @@ int radeon_vce_init(struct radeon_device *rdev)
 	size = RADEON_GPU_PAGE_ALIGN(rdev->vce_fw->size) +
 	       RADEON_VCE_STACK_SIZE + RADEON_VCE_HEAP_SIZE;
 	r = radeon_bo_create(rdev, size, PAGE_SIZE, true,
-			     RADEON_GEM_DOMAIN_VRAM, 0, NULL, &rdev->vce.vcpu_bo);
+			     RADEON_GEM_DOMAIN_VRAM, 0, NULL, NULL,
+			     &rdev->vce.vcpu_bo);
 	if (r) {
 		dev_err(rdev->dev, "(%d) failed to allocate VCE bo\n", r);
 		return r;
@@ -452,11 +453,11 @@ int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi,
 			unsigned size)
 {
 	struct radeon_cs_chunk *relocs_chunk;
-	struct radeon_cs_reloc *reloc;
+	struct radeon_bo_list *reloc;
 	uint64_t start, end, offset;
 	unsigned idx;
 
-	relocs_chunk = &p->chunks[p->chunk_relocs_idx];
+	relocs_chunk = p->chunk_relocs;
 	offset = radeon_get_ib_value(p, lo);
 	idx = radeon_get_ib_value(p, hi);
 
@@ -466,7 +467,7 @@ int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi,
 		return -EINVAL;
 	}
 
-	reloc = p->relocs_ptr[(idx / 4)];
+	reloc = &p->relocs[(idx / 4)];
 	start = reloc->gpu_offset;
 	end = start + radeon_bo_size(reloc->robj);
 	start += offset;
@@ -533,7 +534,7 @@ int radeon_vce_cs_parse(struct radeon_cs_parser *p)
 	uint32_t *size = &tmp;
 	int i, r;
 
-	while (p->idx < p->chunks[p->chunk_ib_idx].length_dw) {
+	while (p->idx < p->chunk_ib->length_dw) {
 		uint32_t len = radeon_get_ib_value(p, p->idx);
 		uint32_t cmd = radeon_get_ib_value(p, p->idx + 1);
 
diff --git a/drivers/gpu/drm/radeon/radeon_vm.c b/drivers/gpu/drm/radeon/radeon_vm.c
index 088ffdc2f577..06d2246d07f1 100644
--- a/drivers/gpu/drm/radeon/radeon_vm.c
+++ b/drivers/gpu/drm/radeon/radeon_vm.c
@@ -125,39 +125,37 @@ void radeon_vm_manager_fini(struct radeon_device *rdev)
  * Add the page directory to the list of BOs to
  * validate for command submission (cayman+).
  */
-struct radeon_cs_reloc *radeon_vm_get_bos(struct radeon_device *rdev,
+struct radeon_bo_list *radeon_vm_get_bos(struct radeon_device *rdev,
 					  struct radeon_vm *vm,
 					  struct list_head *head)
 {
-	struct radeon_cs_reloc *list;
+	struct radeon_bo_list *list;
 	unsigned i, idx;
 
-	list = kmalloc_array(vm->max_pde_used + 2,
-			     sizeof(struct radeon_cs_reloc), GFP_KERNEL);
+	list = drm_malloc_ab(vm->max_pde_used + 2,
+			     sizeof(struct radeon_bo_list));
 	if (!list)
 		return NULL;
 
 	/* add the vm page table to the list */
-	list[0].gobj = NULL;
 	list[0].robj = vm->page_directory;
 	list[0].prefered_domains = RADEON_GEM_DOMAIN_VRAM;
 	list[0].allowed_domains = RADEON_GEM_DOMAIN_VRAM;
 	list[0].tv.bo = &vm->page_directory->tbo;
+	list[0].tv.shared = true;
 	list[0].tiling_flags = 0;
-	list[0].handle = 0;
 	list_add(&list[0].tv.head, head);
 
 	for (i = 0, idx = 1; i <= vm->max_pde_used; i++) {
 		if (!vm->page_tables[i].bo)
 			continue;
 
-		list[idx].gobj = NULL;
 		list[idx].robj = vm->page_tables[i].bo;
 		list[idx].prefered_domains = RADEON_GEM_DOMAIN_VRAM;
 		list[idx].allowed_domains = RADEON_GEM_DOMAIN_VRAM;
 		list[idx].tv.bo = &list[idx].robj->tbo;
+		list[idx].tv.shared = true;
 		list[idx].tiling_flags = 0;
-		list[idx].handle = 0;
 		list_add(&list[idx++].tv.head, head);
 	}
 
@@ -180,15 +178,18 @@ struct radeon_fence *radeon_vm_grab_id(struct radeon_device *rdev,
 				       struct radeon_vm *vm, int ring)
 {
 	struct radeon_fence *best[RADEON_NUM_RINGS] = {};
+	struct radeon_vm_id *vm_id = &vm->ids[ring];
+
 	unsigned choices[2] = {};
 	unsigned i;
 
 	/* check if the id is still valid */
-	if (vm->last_id_use && vm->last_id_use == rdev->vm_manager.active[vm->id])
+	if (vm_id->id && vm_id->last_id_use &&
+	    vm_id->last_id_use == rdev->vm_manager.active[vm_id->id])
 		return NULL;
 
 	/* we definately need to flush */
-	radeon_fence_unref(&vm->last_flush);
+	vm_id->pd_gpu_addr = ~0ll;
 
 	/* skip over VMID 0, since it is the system VM */
 	for (i = 1; i < rdev->vm_manager.nvm; ++i) {
@@ -196,8 +197,8 @@ struct radeon_fence *radeon_vm_grab_id(struct radeon_device *rdev,
 
 		if (fence == NULL) {
 			/* found a free one */
-			vm->id = i;
-			trace_radeon_vm_grab_id(vm->id, ring);
+			vm_id->id = i;
+			trace_radeon_vm_grab_id(i, ring);
 			return NULL;
 		}
 
@@ -209,8 +210,8 @@ struct radeon_fence *radeon_vm_grab_id(struct radeon_device *rdev,
 
 	for (i = 0; i < 2; ++i) {
 		if (choices[i]) {
-			vm->id = choices[i];
-			trace_radeon_vm_grab_id(vm->id, ring);
+			vm_id->id = choices[i];
+			trace_radeon_vm_grab_id(choices[i], ring);
 			return rdev->vm_manager.active[choices[i]];
 		}
 	}
@@ -226,6 +227,7 @@ struct radeon_fence *radeon_vm_grab_id(struct radeon_device *rdev,
  * @rdev: radeon_device pointer
  * @vm: vm we want to flush
  * @ring: ring to use for flush
+ * @updates: last vm update that is waited for
  *
  * Flush the vm (cayman+).
  *
@@ -233,15 +235,21 @@ struct radeon_fence *radeon_vm_grab_id(struct radeon_device *rdev,
  */
 void radeon_vm_flush(struct radeon_device *rdev,
 		     struct radeon_vm *vm,
-		     int ring)
+		     int ring, struct radeon_fence *updates)
 {
 	uint64_t pd_addr = radeon_bo_gpu_offset(vm->page_directory);
+	struct radeon_vm_id *vm_id = &vm->ids[ring];
+
+	if (pd_addr != vm_id->pd_gpu_addr || !vm_id->flushed_updates ||
+	    radeon_fence_is_earlier(vm_id->flushed_updates, updates)) {
+
+		trace_radeon_vm_flush(pd_addr, ring, vm->ids[ring].id);
+		radeon_fence_unref(&vm_id->flushed_updates);
+		vm_id->flushed_updates = radeon_fence_ref(updates);
+		vm_id->pd_gpu_addr = pd_addr;
+		radeon_ring_vm_flush(rdev, &rdev->ring[ring],
+				     vm_id->id, vm_id->pd_gpu_addr);
 
-	/* if we can't remember our last VM flush then flush now! */
-	if (!vm->last_flush || pd_addr != vm->pd_gpu_addr) {
-		trace_radeon_vm_flush(pd_addr, ring, vm->id);
-		vm->pd_gpu_addr = pd_addr;
-		radeon_ring_vm_flush(rdev, ring, vm);
 	}
 }
 
@@ -261,18 +269,13 @@ void radeon_vm_fence(struct radeon_device *rdev,
 		     struct radeon_vm *vm,
 		     struct radeon_fence *fence)
 {
-	radeon_fence_unref(&vm->fence);
-	vm->fence = radeon_fence_ref(fence);
-
-	radeon_fence_unref(&rdev->vm_manager.active[vm->id]);
-	rdev->vm_manager.active[vm->id] = radeon_fence_ref(fence);
+	unsigned vm_id = vm->ids[fence->ring].id;
 
-	radeon_fence_unref(&vm->last_id_use);
-	vm->last_id_use = radeon_fence_ref(fence);
+	radeon_fence_unref(&rdev->vm_manager.active[vm_id]);
+	rdev->vm_manager.active[vm_id] = radeon_fence_ref(fence);
 
-        /* we just flushed the VM, remember that */
-        if (!vm->last_flush)
-                vm->last_flush = radeon_fence_ref(fence);
+	radeon_fence_unref(&vm->ids[fence->ring].last_id_use);
+	vm->ids[fence->ring].last_id_use = radeon_fence_ref(fence);
 }
 
 /**
@@ -385,34 +388,25 @@ static void radeon_vm_set_pages(struct radeon_device *rdev,
 static int radeon_vm_clear_bo(struct radeon_device *rdev,
 			      struct radeon_bo *bo)
 {
-        struct ttm_validate_buffer tv;
-        struct ww_acquire_ctx ticket;
-        struct list_head head;
 	struct radeon_ib ib;
 	unsigned entries;
 	uint64_t addr;
 	int r;
 
-        memset(&tv, 0, sizeof(tv));
-        tv.bo = &bo->tbo;
-
-        INIT_LIST_HEAD(&head);
-        list_add(&tv.head, &head);
-
-        r = ttm_eu_reserve_buffers(&ticket, &head);
-        if (r)
+	r = radeon_bo_reserve(bo, false);
+	if (r)
 		return r;
 
-        r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
-        if (r)
-                goto error;
+	r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
+	if (r)
+		goto error_unreserve;
 
 	addr = radeon_bo_gpu_offset(bo);
 	entries = radeon_bo_size(bo) / 8;
 
 	r = radeon_ib_get(rdev, R600_RING_TYPE_DMA_INDEX, &ib, NULL, 256);
 	if (r)
-                goto error;
+		goto error_unreserve;
 
 	ib.length_dw = 0;
 
@@ -422,15 +416,16 @@ static int radeon_vm_clear_bo(struct radeon_device *rdev,
 
 	r = radeon_ib_schedule(rdev, &ib, NULL, false);
 	if (r)
-                goto error;
+		goto error_free;
 
-	ttm_eu_fence_buffer_objects(&ticket, &head, ib.fence);
-	radeon_ib_free(rdev, &ib);
+	ib.fence->is_vm_update = true;
+	radeon_bo_fence(bo, ib.fence, false);
 
-	return 0;
+error_free:
+	radeon_ib_free(rdev, &ib);
 
-error:
-	ttm_eu_backoff_reservation(&ticket, &head);
+error_unreserve:
+	radeon_bo_unreserve(bo);
 	return r;
 }
 
@@ -446,7 +441,7 @@ error:
  * Validate and set the offset requested within the vm address space.
  * Returns 0 for success, error for failure.
  *
- * Object has to be reserved!
+ * Object has to be reserved and gets unreserved by this function!
  */
 int radeon_vm_bo_set_addr(struct radeon_device *rdev,
 			  struct radeon_bo_va *bo_va,
@@ -492,7 +487,9 @@ int radeon_vm_bo_set_addr(struct radeon_device *rdev,
 			tmp->vm = vm;
 			tmp->addr = bo_va->addr;
 			tmp->bo = radeon_bo_ref(bo_va->bo);
+			spin_lock(&vm->status_lock);
 			list_add(&tmp->vm_status, &vm->freed);
+			spin_unlock(&vm->status_lock);
 		}
 
 		interval_tree_remove(&bo_va->it, &vm->va);
@@ -545,7 +542,8 @@ int radeon_vm_bo_set_addr(struct radeon_device *rdev,
 
 		r = radeon_bo_create(rdev, RADEON_VM_PTE_COUNT * 8,
 				     RADEON_GPU_PAGE_SIZE, true,
-				     RADEON_GEM_DOMAIN_VRAM, 0, NULL, &pt);
+				     RADEON_GEM_DOMAIN_VRAM, 0,
+				     NULL, NULL, &pt);
 		if (r)
 			return r;
 
@@ -571,7 +569,7 @@ int radeon_vm_bo_set_addr(struct radeon_device *rdev,
 	}
 
 	mutex_unlock(&vm->mutex);
-	return radeon_bo_reserve(bo_va->bo, false);
+	return 0;
 }
 
 /**
@@ -589,10 +587,8 @@ uint64_t radeon_vm_map_gart(struct radeon_device *rdev, uint64_t addr)
 	uint64_t result;
 
 	/* page table offset */
-	result = rdev->gart.pages_addr[addr >> PAGE_SHIFT];
-
-	/* in case cpu page size != gpu page size*/
-	result |= addr & (~PAGE_MASK);
+	result = rdev->gart.pages_entry[addr >> RADEON_GPU_PAGE_SHIFT];
+	result &= ~RADEON_GPU_PAGE_MASK;
 
 	return result;
 }
@@ -694,17 +690,16 @@ int radeon_vm_update_page_directory(struct radeon_device *rdev,
 
 	if (ib.length_dw != 0) {
 		radeon_asic_vm_pad_ib(rdev, &ib);
-		radeon_semaphore_sync_to(ib.semaphore, pd->tbo.sync_obj);
-		radeon_semaphore_sync_to(ib.semaphore, vm->last_id_use);
+
+		radeon_sync_resv(rdev, &ib.sync, pd->tbo.resv, true);
 		WARN_ON(ib.length_dw > ndw);
 		r = radeon_ib_schedule(rdev, &ib, NULL, false);
 		if (r) {
 			radeon_ib_free(rdev, &ib);
 			return r;
 		}
-		radeon_fence_unref(&vm->fence);
-		vm->fence = radeon_fence_ref(ib.fence);
-		radeon_fence_unref(&vm->last_flush);
+		ib.fence->is_vm_update = true;
+		radeon_bo_fence(pd, ib.fence, false);
 	}
 	radeon_ib_free(rdev, &ib);
 
@@ -803,11 +798,11 @@ static void radeon_vm_frag_ptes(struct radeon_device *rdev,
  *
  * Global and local mutex must be locked!
  */
-static void radeon_vm_update_ptes(struct radeon_device *rdev,
-				  struct radeon_vm *vm,
-				  struct radeon_ib *ib,
-				  uint64_t start, uint64_t end,
-				  uint64_t dst, uint32_t flags)
+static int radeon_vm_update_ptes(struct radeon_device *rdev,
+				 struct radeon_vm *vm,
+				 struct radeon_ib *ib,
+				 uint64_t start, uint64_t end,
+				 uint64_t dst, uint32_t flags)
 {
 	uint64_t mask = RADEON_VM_PTE_COUNT - 1;
 	uint64_t last_pte = ~0, last_dst = ~0;
@@ -820,8 +815,12 @@ static void radeon_vm_update_ptes(struct radeon_device *rdev,
 		struct radeon_bo *pt = vm->page_tables[pt_idx].bo;
 		unsigned nptes;
 		uint64_t pte;
+		int r;
 
-		radeon_semaphore_sync_to(ib->semaphore, pt->tbo.sync_obj);
+		radeon_sync_resv(rdev, &ib->sync, pt->tbo.resv, true);
+		r = reservation_object_reserve_shared(pt->tbo.resv);
+		if (r)
+			return r;
 
 		if ((addr & ~mask) == (end & ~mask))
 			nptes = end - addr;
@@ -855,6 +854,33 @@ static void radeon_vm_update_ptes(struct radeon_device *rdev,
 				    last_pte + 8 * count,
 				    last_dst, flags);
 	}
+
+	return 0;
+}
+
+/**
+ * radeon_vm_fence_pts - fence page tables after an update
+ *
+ * @vm: requested vm
+ * @start: start of GPU address range
+ * @end: end of GPU address range
+ * @fence: fence to use
+ *
+ * Fence the page tables in the range @start - @end (cayman+).
+ *
+ * Global and local mutex must be locked!
+ */
+static void radeon_vm_fence_pts(struct radeon_vm *vm,
+				uint64_t start, uint64_t end,
+				struct radeon_fence *fence)
+{
+	unsigned i;
+
+	start >>= radeon_vm_block_size;
+	end >>= radeon_vm_block_size;
+
+	for (i = start; i <= end; ++i)
+		radeon_bo_fence(vm->page_tables[i].bo, fence, true);
 }
 
 /**
@@ -887,11 +913,16 @@ int radeon_vm_bo_update(struct radeon_device *rdev,
 		return -EINVAL;
 	}
 
+	spin_lock(&vm->status_lock);
 	list_del_init(&bo_va->vm_status);
+	spin_unlock(&vm->status_lock);
 
 	bo_va->flags &= ~RADEON_VM_PAGE_VALID;
 	bo_va->flags &= ~RADEON_VM_PAGE_SYSTEM;
 	bo_va->flags &= ~RADEON_VM_PAGE_SNOOPED;
+	if (bo_va->bo && radeon_ttm_tt_is_readonly(bo_va->bo->tbo.ttm))
+		bo_va->flags &= ~RADEON_VM_PAGE_WRITEABLE;
+
 	if (mem) {
 		addr = mem->start << PAGE_SHIFT;
 		if (mem->mem_type != TTM_PL_SYSTEM) {
@@ -953,23 +984,34 @@ int radeon_vm_bo_update(struct radeon_device *rdev,
 		return r;
 	ib.length_dw = 0;
 
-	radeon_vm_update_ptes(rdev, vm, &ib, bo_va->it.start,
-			      bo_va->it.last + 1, addr,
-			      radeon_vm_page_flags(bo_va->flags));
+	if (!(bo_va->flags & RADEON_VM_PAGE_VALID)) {
+		unsigned i;
+
+		for (i = 0; i < RADEON_NUM_RINGS; ++i)
+			radeon_sync_fence(&ib.sync, vm->ids[i].last_id_use);
+	}
+
+	r = radeon_vm_update_ptes(rdev, vm, &ib, bo_va->it.start,
+				  bo_va->it.last + 1, addr,
+				  radeon_vm_page_flags(bo_va->flags));
+	if (r) {
+		radeon_ib_free(rdev, &ib);
+		return r;
+	}
 
 	radeon_asic_vm_pad_ib(rdev, &ib);
 	WARN_ON(ib.length_dw > ndw);
 
-	radeon_semaphore_sync_to(ib.semaphore, vm->fence);
 	r = radeon_ib_schedule(rdev, &ib, NULL, false);
 	if (r) {
 		radeon_ib_free(rdev, &ib);
 		return r;
 	}
-	radeon_fence_unref(&vm->fence);
-	vm->fence = radeon_fence_ref(ib.fence);
+	ib.fence->is_vm_update = true;
+	radeon_vm_fence_pts(vm, bo_va->it.start, bo_va->it.last + 1, ib.fence);
+	radeon_fence_unref(&bo_va->last_pt_update);
+	bo_va->last_pt_update = radeon_fence_ref(ib.fence);
 	radeon_ib_free(rdev, &ib);
-	radeon_fence_unref(&vm->last_flush);
 
 	return 0;
 }
@@ -988,16 +1030,25 @@ int radeon_vm_bo_update(struct radeon_device *rdev,
 int radeon_vm_clear_freed(struct radeon_device *rdev,
 			  struct radeon_vm *vm)
 {
-	struct radeon_bo_va *bo_va, *tmp;
+	struct radeon_bo_va *bo_va;
 	int r;
 
-	list_for_each_entry_safe(bo_va, tmp, &vm->freed, vm_status) {
+	spin_lock(&vm->status_lock);
+	while (!list_empty(&vm->freed)) {
+		bo_va = list_first_entry(&vm->freed,
+			struct radeon_bo_va, vm_status);
+		spin_unlock(&vm->status_lock);
+
 		r = radeon_vm_bo_update(rdev, bo_va, NULL);
 		radeon_bo_unref(&bo_va->bo);
+		radeon_fence_unref(&bo_va->last_pt_update);
 		kfree(bo_va);
 		if (r)
 			return r;
+
+		spin_lock(&vm->status_lock);
 	}
+	spin_unlock(&vm->status_lock);
 	return 0;
 
 }
@@ -1016,14 +1067,23 @@ int radeon_vm_clear_freed(struct radeon_device *rdev,
 int radeon_vm_clear_invalids(struct radeon_device *rdev,
 			     struct radeon_vm *vm)
 {
-	struct radeon_bo_va *bo_va, *tmp;
+	struct radeon_bo_va *bo_va;
 	int r;
 
-	list_for_each_entry_safe(bo_va, tmp, &vm->invalidated, vm_status) {
+	spin_lock(&vm->status_lock);
+	while (!list_empty(&vm->invalidated)) {
+		bo_va = list_first_entry(&vm->invalidated,
+			struct radeon_bo_va, vm_status);
+		spin_unlock(&vm->status_lock);
+
 		r = radeon_vm_bo_update(rdev, bo_va, NULL);
 		if (r)
 			return r;
+
+		spin_lock(&vm->status_lock);
 	}
+	spin_unlock(&vm->status_lock);
+
 	return 0;
 }
 
@@ -1046,14 +1106,17 @@ void radeon_vm_bo_rmv(struct radeon_device *rdev,
 
 	mutex_lock(&vm->mutex);
 	interval_tree_remove(&bo_va->it, &vm->va);
+	spin_lock(&vm->status_lock);
 	list_del(&bo_va->vm_status);
 
 	if (bo_va->addr) {
 		bo_va->bo = radeon_bo_ref(bo_va->bo);
 		list_add(&bo_va->vm_status, &vm->freed);
 	} else {
+		radeon_fence_unref(&bo_va->last_pt_update);
 		kfree(bo_va);
 	}
+	spin_unlock(&vm->status_lock);
 
 	mutex_unlock(&vm->mutex);
 }
@@ -1074,10 +1137,10 @@ void radeon_vm_bo_invalidate(struct radeon_device *rdev,
 
 	list_for_each_entry(bo_va, &bo->va, bo_list) {
 		if (bo_va->addr) {
-			mutex_lock(&bo_va->vm->mutex);
+			spin_lock(&bo_va->vm->status_lock);
 			list_del(&bo_va->vm_status);
 			list_add(&bo_va->vm_status, &bo_va->vm->invalidated);
-			mutex_unlock(&bo_va->vm->mutex);
+			spin_unlock(&bo_va->vm->status_lock);
 		}
 	}
 }
@@ -1095,15 +1158,17 @@ int radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm)
 	const unsigned align = min(RADEON_VM_PTB_ALIGN_SIZE,
 		RADEON_VM_PTE_COUNT * 8);
 	unsigned pd_size, pd_entries, pts_size;
-	int r;
+	int i, r;
 
-	vm->id = 0;
 	vm->ib_bo_va = NULL;
-	vm->fence = NULL;
-	vm->last_flush = NULL;
-	vm->last_id_use = NULL;
+	for (i = 0; i < RADEON_NUM_RINGS; ++i) {
+		vm->ids[i].id = 0;
+		vm->ids[i].flushed_updates = NULL;
+		vm->ids[i].last_id_use = NULL;
+	}
 	mutex_init(&vm->mutex);
 	vm->va = RB_ROOT;
+	spin_lock_init(&vm->status_lock);
 	INIT_LIST_HEAD(&vm->invalidated);
 	INIT_LIST_HEAD(&vm->freed);
 
@@ -1120,7 +1185,7 @@ int radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm)
 
 	r = radeon_bo_create(rdev, pd_size, align, true,
 			     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
-			     &vm->page_directory);
+			     NULL, &vm->page_directory);
 	if (r)
 		return r;
 
@@ -1157,11 +1222,13 @@ void radeon_vm_fini(struct radeon_device *rdev, struct radeon_vm *vm)
 		if (!r) {
 			list_del_init(&bo_va->bo_list);
 			radeon_bo_unreserve(bo_va->bo);
+			radeon_fence_unref(&bo_va->last_pt_update);
 			kfree(bo_va);
 		}
 	}
 	list_for_each_entry_safe(bo_va, tmp, &vm->freed, vm_status) {
 		radeon_bo_unref(&bo_va->bo);
+		radeon_fence_unref(&bo_va->last_pt_update);
 		kfree(bo_va);
 	}
 
@@ -1171,9 +1238,10 @@ void radeon_vm_fini(struct radeon_device *rdev, struct radeon_vm *vm)
 
 	radeon_bo_unref(&vm->page_directory);
 
-	radeon_fence_unref(&vm->fence);
-	radeon_fence_unref(&vm->last_flush);
-	radeon_fence_unref(&vm->last_id_use);
+	for (i = 0; i < RADEON_NUM_RINGS; ++i) {
+		radeon_fence_unref(&vm->ids[i].flushed_updates);
+		radeon_fence_unref(&vm->ids[i].last_id_use);
+	}
 
 	mutex_destroy(&vm->mutex);
 }
diff --git a/drivers/gpu/drm/radeon/rs400.c b/drivers/gpu/drm/radeon/rs400.c
index c5799f16aa4b..34e3235f41d2 100644
--- a/drivers/gpu/drm/radeon/rs400.c
+++ b/drivers/gpu/drm/radeon/rs400.c
@@ -212,11 +212,9 @@ void rs400_gart_fini(struct radeon_device *rdev)
 #define RS400_PTE_WRITEABLE (1 << 2)
 #define RS400_PTE_READABLE  (1 << 3)
 
-void rs400_gart_set_page(struct radeon_device *rdev, unsigned i,
-			 uint64_t addr, uint32_t flags)
+uint64_t rs400_gart_get_page_entry(uint64_t addr, uint32_t flags)
 {
 	uint32_t entry;
-	u32 *gtt = rdev->gart.ptr;
 
 	entry = (lower_32_bits(addr) & PAGE_MASK) |
 		((upper_32_bits(addr) & 0xff) << 4);
@@ -226,8 +224,14 @@ void rs400_gart_set_page(struct radeon_device *rdev, unsigned i,
 		entry |= RS400_PTE_WRITEABLE;
 	if (!(flags & RADEON_GART_PAGE_SNOOP))
 		entry |= RS400_PTE_UNSNOOPED;
-	entry = cpu_to_le32(entry);
-	gtt[i] = entry;
+	return entry;
+}
+
+void rs400_gart_set_page(struct radeon_device *rdev, unsigned i,
+			 uint64_t entry)
+{
+	u32 *gtt = rdev->gart.ptr;
+	gtt[i] = cpu_to_le32(lower_32_bits(entry));
 }
 
 int rs400_mc_wait_for_idle(struct radeon_device *rdev)
diff --git a/drivers/gpu/drm/radeon/rs600.c b/drivers/gpu/drm/radeon/rs600.c
index 5f6db4629aaa..74bce91aecc1 100644
--- a/drivers/gpu/drm/radeon/rs600.c
+++ b/drivers/gpu/drm/radeon/rs600.c
@@ -625,11 +625,8 @@ static void rs600_gart_fini(struct radeon_device *rdev)
 	radeon_gart_table_vram_free(rdev);
 }
 
-void rs600_gart_set_page(struct radeon_device *rdev, unsigned i,
-			 uint64_t addr, uint32_t flags)
+uint64_t rs600_gart_get_page_entry(uint64_t addr, uint32_t flags)
 {
-	void __iomem *ptr = (void *)rdev->gart.ptr;
-
 	addr = addr & 0xFFFFFFFFFFFFF000ULL;
 	addr |= R600_PTE_SYSTEM;
 	if (flags & RADEON_GART_PAGE_VALID)
@@ -640,7 +637,14 @@ void rs600_gart_set_page(struct radeon_device *rdev, unsigned i,
 		addr |= R600_PTE_WRITEABLE;
 	if (flags & RADEON_GART_PAGE_SNOOP)
 		addr |= R600_PTE_SNOOPED;
-	writeq(addr, ptr + (i * 8));
+	return addr;
+}
+
+void rs600_gart_set_page(struct radeon_device *rdev, unsigned i,
+			 uint64_t entry)
+{
+	void __iomem *ptr = (void *)rdev->gart.ptr;
+	writeq(entry, ptr + (i * 8));
 }
 
 int rs600_irq_set(struct radeon_device *rdev)
@@ -879,6 +883,9 @@ void rs600_bandwidth_update(struct radeon_device *rdev)
 	u32 d1mode_priority_a_cnt, d2mode_priority_a_cnt;
 	/* FIXME: implement full support */
 
+	if (!rdev->mode_info.mode_config_initialized)
+		return;
+
 	radeon_update_display_priority(rdev);
 
 	if (rdev->mode_info.crtcs[0]->base.enabled)
diff --git a/drivers/gpu/drm/radeon/rs690.c b/drivers/gpu/drm/radeon/rs690.c
index 3462b64369bf..0a2d36e81108 100644
--- a/drivers/gpu/drm/radeon/rs690.c
+++ b/drivers/gpu/drm/radeon/rs690.c
@@ -579,6 +579,9 @@ void rs690_bandwidth_update(struct radeon_device *rdev)
 	u32 d1mode_priority_a_cnt, d1mode_priority_b_cnt;
 	u32 d2mode_priority_a_cnt, d2mode_priority_b_cnt;
 
+	if (!rdev->mode_info.mode_config_initialized)
+		return;
+
 	radeon_update_display_priority(rdev);
 
 	if (rdev->mode_info.crtcs[0]->base.enabled)
diff --git a/drivers/gpu/drm/radeon/rs780_dpm.c b/drivers/gpu/drm/radeon/rs780_dpm.c
index 02f7710de470..9031f4b69824 100644
--- a/drivers/gpu/drm/radeon/rs780_dpm.c
+++ b/drivers/gpu/drm/radeon/rs780_dpm.c
@@ -24,6 +24,7 @@
 
 #include "drmP.h"
 #include "radeon.h"
+#include "radeon_asic.h"
 #include "rs780d.h"
 #include "r600_dpm.h"
 #include "rs780_dpm.h"
diff --git a/drivers/gpu/drm/radeon/rv515.c b/drivers/gpu/drm/radeon/rv515.c
index 8a477bf1fdb3..c55d653aaf5f 100644
--- a/drivers/gpu/drm/radeon/rv515.c
+++ b/drivers/gpu/drm/radeon/rv515.c
@@ -1277,6 +1277,9 @@ void rv515_bandwidth_update(struct radeon_device *rdev)
 	struct drm_display_mode *mode0 = NULL;
 	struct drm_display_mode *mode1 = NULL;
 
+	if (!rdev->mode_info.mode_config_initialized)
+		return;
+
 	radeon_update_display_priority(rdev);
 
 	if (rdev->mode_info.crtcs[0]->base.enabled)
diff --git a/drivers/gpu/drm/radeon/rv6xx_dpm.c b/drivers/gpu/drm/radeon/rv6xx_dpm.c
index e7045b085715..6a5c233361e9 100644
--- a/drivers/gpu/drm/radeon/rv6xx_dpm.c
+++ b/drivers/gpu/drm/radeon/rv6xx_dpm.c
@@ -24,6 +24,7 @@
 
 #include "drmP.h"
 #include "radeon.h"
+#include "radeon_asic.h"
 #include "rv6xxd.h"
 #include "r600_dpm.h"
 #include "rv6xx_dpm.h"
diff --git a/drivers/gpu/drm/radeon/rv770.c b/drivers/gpu/drm/radeon/rv770.c
index d9f5ce715c9b..372016e266d0 100644
--- a/drivers/gpu/drm/radeon/rv770.c
+++ b/drivers/gpu/drm/radeon/rv770.c
@@ -26,7 +26,6 @@
  *          Jerome Glisse
  */
 #include <linux/firmware.h>
-#include <linux/platform_device.h>
 #include <linux/slab.h>
 #include <drm/drmP.h>
 #include "radeon.h"
diff --git a/drivers/gpu/drm/radeon/rv770_dma.c b/drivers/gpu/drm/radeon/rv770_dma.c
index 74426ac2bb5c..acff6e09cc40 100644
--- a/drivers/gpu/drm/radeon/rv770_dma.c
+++ b/drivers/gpu/drm/radeon/rv770_dma.c
@@ -33,41 +33,38 @@
  * @src_offset: src GPU address
  * @dst_offset: dst GPU address
  * @num_gpu_pages: number of GPU pages to xfer
- * @fence: radeon fence object
+ * @resv: reservation object to sync to
  *
  * Copy GPU paging using the DMA engine (r7xx).
  * Used by the radeon ttm implementation to move pages if
  * registered as the asic copy callback.
  */
-int rv770_copy_dma(struct radeon_device *rdev,
-		  uint64_t src_offset, uint64_t dst_offset,
-		  unsigned num_gpu_pages,
-		  struct radeon_fence **fence)
+struct radeon_fence *rv770_copy_dma(struct radeon_device *rdev,
+				    uint64_t src_offset, uint64_t dst_offset,
+				    unsigned num_gpu_pages,
+				    struct reservation_object *resv)
 {
-	struct radeon_semaphore *sem = NULL;
+	struct radeon_fence *fence;
+	struct radeon_sync sync;
 	int ring_index = rdev->asic->copy.dma_ring_index;
 	struct radeon_ring *ring = &rdev->ring[ring_index];
 	u32 size_in_dw, cur_size_in_dw;
 	int i, num_loops;
 	int r = 0;
 
-	r = radeon_semaphore_create(rdev, &sem);
-	if (r) {
-		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		return r;
-	}
+	radeon_sync_create(&sync);
 
 	size_in_dw = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT) / 4;
 	num_loops = DIV_ROUND_UP(size_in_dw, 0xFFFF);
 	r = radeon_ring_lock(rdev, ring, num_loops * 5 + 8);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		radeon_sync_free(rdev, &sync, NULL);
+		return ERR_PTR(r);
 	}
 
-	radeon_semaphore_sync_to(sem, *fence);
-	radeon_semaphore_sync_rings(rdev, sem, ring->idx);
+	radeon_sync_resv(rdev, &sync, resv, false);
+	radeon_sync_rings(rdev, &sync, ring->idx);
 
 	for (i = 0; i < num_loops; i++) {
 		cur_size_in_dw = size_in_dw;
@@ -83,15 +80,15 @@ int rv770_copy_dma(struct radeon_device *rdev,
 		dst_offset += cur_size_in_dw * 4;
 	}
 
-	r = radeon_fence_emit(rdev, fence, ring->idx);
+	r = radeon_fence_emit(rdev, &fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
-		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		radeon_sync_free(rdev, &sync, NULL);
+		return ERR_PTR(r);
 	}
 
 	radeon_ring_unlock_commit(rdev, ring, false);
-	radeon_semaphore_free(rdev, &sem, *fence);
+	radeon_sync_free(rdev, &sync, fence);
 
-	return r;
+	return fence;
 }
diff --git a/drivers/gpu/drm/radeon/rv770_dpm.c b/drivers/gpu/drm/radeon/rv770_dpm.c
index 3c76e1dcdf04..755a8f96fe46 100644
--- a/drivers/gpu/drm/radeon/rv770_dpm.c
+++ b/drivers/gpu/drm/radeon/rv770_dpm.c
@@ -24,6 +24,7 @@
 
 #include "drmP.h"
 #include "radeon.h"
+#include "radeon_asic.h"
 #include "rv770d.h"
 #include "r600_dpm.h"
 #include "rv770_dpm.h"
diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c
index 3a0b973e8a96..5d89b874a1a2 100644
--- a/drivers/gpu/drm/radeon/si.c
+++ b/drivers/gpu/drm/radeon/si.c
@@ -2384,6 +2384,9 @@ void dce6_bandwidth_update(struct radeon_device *rdev)
 	u32 num_heads = 0, lb_size;
 	int i;
 
+	if (!rdev->mode_info.mode_config_initialized)
+		return;
+
 	radeon_update_display_priority(rdev);
 
 	for (i = 0; i < rdev->num_crtc; i++) {
@@ -3362,6 +3365,7 @@ void si_fence_ring_emit(struct radeon_device *rdev,
 void si_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
 {
 	struct radeon_ring *ring = &rdev->ring[ib->ring];
+	unsigned vm_id = ib->vm ? ib->vm->ids[ib->ring].id : 0;
 	u32 header;
 
 	if (ib->is_const_ib) {
@@ -3397,14 +3401,13 @@ void si_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
 #endif
 			  (ib->gpu_addr & 0xFFFFFFFC));
 	radeon_ring_write(ring, upper_32_bits(ib->gpu_addr) & 0xFFFF);
-	radeon_ring_write(ring, ib->length_dw |
-			  (ib->vm ? (ib->vm->id << 24) : 0));
+	radeon_ring_write(ring, ib->length_dw | (vm_id << 24));
 
 	if (!ib->is_const_ib) {
 		/* flush read cache over gart for this vmid */
 		radeon_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1));
 		radeon_ring_write(ring, (CP_COHER_CNTL2 - PACKET3_SET_CONFIG_REG_START) >> 2);
-		radeon_ring_write(ring, ib->vm ? ib->vm->id : 0);
+		radeon_ring_write(ring, vm_id);
 		radeon_ring_write(ring, PACKET3(PACKET3_SURFACE_SYNC, 3));
 		radeon_ring_write(ring, PACKET3_TCL1_ACTION_ENA |
 				  PACKET3_TC_ACTION_ENA |
@@ -4684,7 +4687,7 @@ static int si_vm_packet3_compute_check(struct radeon_device *rdev,
 int si_ib_parse(struct radeon_device *rdev, struct radeon_ib *ib)
 {
 	int ret = 0;
-	u32 idx = 0;
+	u32 idx = 0, i;
 	struct radeon_cs_packet pkt;
 
 	do {
@@ -4695,6 +4698,12 @@ int si_ib_parse(struct radeon_device *rdev, struct radeon_ib *ib)
 		switch (pkt.type) {
 		case RADEON_PACKET_TYPE0:
 			dev_err(rdev->dev, "Packet0 not allowed!\n");
+			for (i = 0; i < ib->length_dw; i++) {
+				if (i == idx)
+					printk("\t0x%08x <---\n", ib->ptr[i]);
+				else
+					printk("\t0x%08x\n", ib->ptr[i]);
+			}
 			ret = -EINVAL;
 			break;
 		case RADEON_PACKET_TYPE2:
@@ -5014,27 +5023,23 @@ static void si_vm_decode_fault(struct radeon_device *rdev,
 	       block, mc_id);
 }
 
-void si_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
+void si_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring,
+		 unsigned vm_id, uint64_t pd_addr)
 {
-	struct radeon_ring *ring = &rdev->ring[ridx];
-
-	if (vm == NULL)
-		return;
-
 	/* write new base address */
 	radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
 	radeon_ring_write(ring, (WRITE_DATA_ENGINE_SEL(1) |
 				 WRITE_DATA_DST_SEL(0)));
 
-	if (vm->id < 8) {
+	if (vm_id < 8) {
 		radeon_ring_write(ring,
-				  (VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (vm->id << 2)) >> 2);
+				  (VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (vm_id << 2)) >> 2);
 	} else {
 		radeon_ring_write(ring,
-				  (VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + ((vm->id - 8) << 2)) >> 2);
+				  (VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + ((vm_id - 8) << 2)) >> 2);
 	}
 	radeon_ring_write(ring, 0);
-	radeon_ring_write(ring, vm->pd_gpu_addr >> 12);
+	radeon_ring_write(ring, pd_addr >> 12);
 
 	/* flush hdp cache */
 	radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
@@ -5050,7 +5055,17 @@ void si_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
 				 WRITE_DATA_DST_SEL(0)));
 	radeon_ring_write(ring, VM_INVALIDATE_REQUEST >> 2);
 	radeon_ring_write(ring, 0);
-	radeon_ring_write(ring, 1 << vm->id);
+	radeon_ring_write(ring, 1 << vm_id);
+
+	/* wait for the invalidate to complete */
+	radeon_ring_write(ring, PACKET3(PACKET3_WAIT_REG_MEM, 5));
+	radeon_ring_write(ring, (WAIT_REG_MEM_FUNCTION(0) |  /* always */
+				 WAIT_REG_MEM_ENGINE(0))); /* me */
+	radeon_ring_write(ring, VM_INVALIDATE_REQUEST >> 2);
+	radeon_ring_write(ring, 0);
+	radeon_ring_write(ring, 0); /* ref */
+	radeon_ring_write(ring, 0); /* mask */
+	radeon_ring_write(ring, 0x20); /* poll interval */
 
 	/* sync PFP to ME, otherwise we might get invalid PFP reads */
 	radeon_ring_write(ring, PACKET3(PACKET3_PFP_SYNC_ME, 0));
diff --git a/drivers/gpu/drm/radeon/si_dma.c b/drivers/gpu/drm/radeon/si_dma.c
index 7c22baaf94db..83207929fc62 100644
--- a/drivers/gpu/drm/radeon/si_dma.c
+++ b/drivers/gpu/drm/radeon/si_dma.c
@@ -123,7 +123,6 @@ void si_dma_vm_write_pages(struct radeon_device *rdev,
 		for (; ndw > 0; ndw -= 2, --count, pe += 8) {
 			if (flags & R600_PTE_SYSTEM) {
 				value = radeon_vm_map_gart(rdev, addr);
-				value &= 0xFFFFFFFFFFFFF000ULL;
 			} else if (flags & R600_PTE_VALID) {
 				value = addr;
 			} else {
@@ -185,20 +184,17 @@ void si_dma_vm_set_pages(struct radeon_device *rdev,
 	}
 }
 
-void si_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
-{
-	struct radeon_ring *ring = &rdev->ring[ridx];
-
-	if (vm == NULL)
-		return;
+void si_dma_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring,
+		     unsigned vm_id, uint64_t pd_addr)
 
+{
 	radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0));
-	if (vm->id < 8) {
-		radeon_ring_write(ring, (0xf << 16) | ((VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (vm->id << 2)) >> 2));
+	if (vm_id < 8) {
+		radeon_ring_write(ring, (0xf << 16) | ((VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (vm_id << 2)) >> 2));
 	} else {
-		radeon_ring_write(ring, (0xf << 16) | ((VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + ((vm->id - 8) << 2)) >> 2));
+		radeon_ring_write(ring, (0xf << 16) | ((VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + ((vm_id - 8) << 2)) >> 2));
 	}
-	radeon_ring_write(ring, vm->pd_gpu_addr >> 12);
+	radeon_ring_write(ring, pd_addr >> 12);
 
 	/* flush hdp cache */
 	radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0));
@@ -208,7 +204,15 @@ void si_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
 	/* bits 0-7 are the VM contexts0-7 */
 	radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0));
 	radeon_ring_write(ring, (0xf << 16) | (VM_INVALIDATE_REQUEST >> 2));
-	radeon_ring_write(ring, 1 << vm->id);
+	radeon_ring_write(ring, 1 << vm_id);
+
+	/* wait for invalidate to complete */
+	radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_POLL_REG_MEM, 0, 0, 0, 0));
+	radeon_ring_write(ring, VM_INVALIDATE_REQUEST);
+	radeon_ring_write(ring, 0xff << 16); /* retry */
+	radeon_ring_write(ring, 1 << vm_id); /* mask */
+	radeon_ring_write(ring, 0); /* value */
+	radeon_ring_write(ring, (0 << 28) | 0x20); /* func(always) | poll interval */
 }
 
 /**
@@ -218,41 +222,38 @@ void si_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
  * @src_offset: src GPU address
  * @dst_offset: dst GPU address
  * @num_gpu_pages: number of GPU pages to xfer
- * @fence: radeon fence object
+ * @resv: reservation object to sync to
  *
  * Copy GPU paging using the DMA engine (SI).
  * Used by the radeon ttm implementation to move pages if
  * registered as the asic copy callback.
  */
-int si_copy_dma(struct radeon_device *rdev,
-		uint64_t src_offset, uint64_t dst_offset,
-		unsigned num_gpu_pages,
-		struct radeon_fence **fence)
+struct radeon_fence *si_copy_dma(struct radeon_device *rdev,
+				 uint64_t src_offset, uint64_t dst_offset,
+				 unsigned num_gpu_pages,
+				 struct reservation_object *resv)
 {
-	struct radeon_semaphore *sem = NULL;
+	struct radeon_fence *fence;
+	struct radeon_sync sync;
 	int ring_index = rdev->asic->copy.dma_ring_index;
 	struct radeon_ring *ring = &rdev->ring[ring_index];
 	u32 size_in_bytes, cur_size_in_bytes;
 	int i, num_loops;
 	int r = 0;
 
-	r = radeon_semaphore_create(rdev, &sem);
-	if (r) {
-		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		return r;
-	}
+	radeon_sync_create(&sync);
 
 	size_in_bytes = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT);
 	num_loops = DIV_ROUND_UP(size_in_bytes, 0xfffff);
 	r = radeon_ring_lock(rdev, ring, num_loops * 5 + 11);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		radeon_sync_free(rdev, &sync, NULL);
+		return ERR_PTR(r);
 	}
 
-	radeon_semaphore_sync_to(sem, *fence);
-	radeon_semaphore_sync_rings(rdev, sem, ring->idx);
+	radeon_sync_resv(rdev, &sync, resv, false);
+	radeon_sync_rings(rdev, &sync, ring->idx);
 
 	for (i = 0; i < num_loops; i++) {
 		cur_size_in_bytes = size_in_bytes;
@@ -268,16 +269,16 @@ int si_copy_dma(struct radeon_device *rdev,
 		dst_offset += cur_size_in_bytes;
 	}
 
-	r = radeon_fence_emit(rdev, fence, ring->idx);
+	r = radeon_fence_emit(rdev, &fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
-		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		radeon_sync_free(rdev, &sync, NULL);
+		return ERR_PTR(r);
 	}
 
 	radeon_ring_unlock_commit(rdev, ring, false);
-	radeon_semaphore_free(rdev, &sem, *fence);
+	radeon_sync_free(rdev, &sync, fence);
 
-	return r;
+	return fence;
 }
 
diff --git a/drivers/gpu/drm/radeon/si_dpm.c b/drivers/gpu/drm/radeon/si_dpm.c
index 70e61ffeace2..eff8a6444956 100644
--- a/drivers/gpu/drm/radeon/si_dpm.c
+++ b/drivers/gpu/drm/radeon/si_dpm.c
@@ -23,6 +23,7 @@
 
 #include "drmP.h"
 #include "radeon.h"
+#include "radeon_asic.h"
 #include "sid.h"
 #include "r600_dpm.h"
 #include "si_dpm.h"
@@ -2907,6 +2908,22 @@ static int si_init_smc_spll_table(struct radeon_device *rdev)
 	return ret;
 }
 
+struct si_dpm_quirk {
+	u32 chip_vendor;
+	u32 chip_device;
+	u32 subsys_vendor;
+	u32 subsys_device;
+	u32 max_sclk;
+	u32 max_mclk;
+};
+
+/* cards with dpm stability problems */
+static struct si_dpm_quirk si_dpm_quirk_list[] = {
+	/* PITCAIRN - https://bugs.freedesktop.org/show_bug.cgi?id=76490 */
+	{ PCI_VENDOR_ID_ATI, 0x6810, 0x1462, 0x3036, 0, 120000 },
+	{ 0, 0, 0, 0 },
+};
+
 static void si_apply_state_adjust_rules(struct radeon_device *rdev,
 					struct radeon_ps *rps)
 {
@@ -2917,7 +2934,22 @@ static void si_apply_state_adjust_rules(struct radeon_device *rdev,
 	u32 mclk, sclk;
 	u16 vddc, vddci;
 	u32 max_sclk_vddc, max_mclk_vddci, max_mclk_vddc;
+	u32 max_sclk = 0, max_mclk = 0;
 	int i;
+	struct si_dpm_quirk *p = si_dpm_quirk_list;
+
+	/* Apply dpm quirks */
+	while (p && p->chip_device != 0) {
+		if (rdev->pdev->vendor == p->chip_vendor &&
+		    rdev->pdev->device == p->chip_device &&
+		    rdev->pdev->subsystem_vendor == p->subsys_vendor &&
+		    rdev->pdev->subsystem_device == p->subsys_device) {
+			max_sclk = p->max_sclk;
+			max_mclk = p->max_mclk;
+			break;
+		}
+		++p;
+	}
 
 	if ((rdev->pm.dpm.new_active_crtc_count > 1) ||
 	    ni_dpm_vblank_too_short(rdev))
@@ -2971,6 +3003,14 @@ static void si_apply_state_adjust_rules(struct radeon_device *rdev,
 			if (ps->performance_levels[i].mclk > max_mclk_vddc)
 				ps->performance_levels[i].mclk = max_mclk_vddc;
 		}
+		if (max_mclk) {
+			if (ps->performance_levels[i].mclk > max_mclk)
+				ps->performance_levels[i].mclk = max_mclk;
+		}
+		if (max_sclk) {
+			if (ps->performance_levels[i].sclk > max_sclk)
+				ps->performance_levels[i].sclk = max_sclk;
+		}
 	}
 
 	/* XXX validate the min clocks required for display */
@@ -3397,6 +3437,15 @@ static int si_process_firmware_header(struct radeon_device *rdev)
 
 	ret = si_read_smc_sram_dword(rdev,
 				     SISLANDS_SMC_FIRMWARE_HEADER_LOCATION +
+				     SISLANDS_SMC_FIRMWARE_HEADER_fanTable,
+				     &tmp, si_pi->sram_end);
+	if (ret)
+		return ret;
+
+	si_pi->fan_table_start = tmp;
+
+	ret = si_read_smc_sram_dword(rdev,
+				     SISLANDS_SMC_FIRMWARE_HEADER_LOCATION +
 				     SISLANDS_SMC_FIRMWARE_HEADER_mcArbDramAutoRefreshTable,
 				     &tmp, si_pi->sram_end);
 	if (ret)
@@ -5816,8 +5865,33 @@ void si_dpm_setup_asic(struct radeon_device *rdev)
 	si_enable_acpi_power_management(rdev);
 }
 
-static int si_set_thermal_temperature_range(struct radeon_device *rdev,
-					int min_temp, int max_temp)
+static int si_thermal_enable_alert(struct radeon_device *rdev,
+				   bool enable)
+{
+	u32 thermal_int = RREG32(CG_THERMAL_INT);
+
+	if (enable) {
+		PPSMC_Result result;
+
+		thermal_int &= ~(THERM_INT_MASK_HIGH | THERM_INT_MASK_LOW);
+		WREG32(CG_THERMAL_INT, thermal_int);
+		rdev->irq.dpm_thermal = false;
+		result = si_send_msg_to_smc(rdev, PPSMC_MSG_EnableThermalInterrupt);
+		if (result != PPSMC_Result_OK) {
+			DRM_DEBUG_KMS("Could not enable thermal interrupts.\n");
+			return -EINVAL;
+		}
+	} else {
+		thermal_int |= THERM_INT_MASK_HIGH | THERM_INT_MASK_LOW;
+		WREG32(CG_THERMAL_INT, thermal_int);
+		rdev->irq.dpm_thermal = true;
+	}
+
+	return 0;
+}
+
+static int si_thermal_set_temperature_range(struct radeon_device *rdev,
+					    int min_temp, int max_temp)
 {
 	int low_temp = 0 * 1000;
 	int high_temp = 255 * 1000;
@@ -5841,6 +5915,309 @@ static int si_set_thermal_temperature_range(struct radeon_device *rdev,
 	return 0;
 }
 
+static void si_fan_ctrl_set_static_mode(struct radeon_device *rdev, u32 mode)
+{
+	struct si_power_info *si_pi = si_get_pi(rdev);
+	u32 tmp;
+
+	if (si_pi->fan_ctrl_is_in_default_mode) {
+		tmp = (RREG32(CG_FDO_CTRL2) & FDO_PWM_MODE_MASK) >> FDO_PWM_MODE_SHIFT;
+		si_pi->fan_ctrl_default_mode = tmp;
+		tmp = (RREG32(CG_FDO_CTRL2) & TMIN_MASK) >> TMIN_SHIFT;
+		si_pi->t_min = tmp;
+		si_pi->fan_ctrl_is_in_default_mode = false;
+	}
+
+	tmp = RREG32(CG_FDO_CTRL2) & ~TMIN_MASK;
+	tmp |= TMIN(0);
+	WREG32(CG_FDO_CTRL2, tmp);
+
+	tmp = RREG32(CG_FDO_CTRL2) & ~FDO_PWM_MODE_MASK;
+	tmp |= FDO_PWM_MODE(mode);
+	WREG32(CG_FDO_CTRL2, tmp);
+}
+
+static int si_thermal_setup_fan_table(struct radeon_device *rdev)
+{
+	struct si_power_info *si_pi = si_get_pi(rdev);
+	PP_SIslands_FanTable fan_table = { FDO_MODE_HARDWARE };
+	u32 duty100;
+	u32 t_diff1, t_diff2, pwm_diff1, pwm_diff2;
+	u16 fdo_min, slope1, slope2;
+	u32 reference_clock, tmp;
+	int ret;
+	u64 tmp64;
+
+	if (!si_pi->fan_table_start) {
+		rdev->pm.dpm.fan.ucode_fan_control = false;
+		return 0;
+	}
+
+	duty100 = (RREG32(CG_FDO_CTRL1) & FMAX_DUTY100_MASK) >> FMAX_DUTY100_SHIFT;
+
+	if (duty100 == 0) {
+		rdev->pm.dpm.fan.ucode_fan_control = false;
+		return 0;
+	}
+
+	tmp64 = (u64)rdev->pm.dpm.fan.pwm_min * duty100;
+	do_div(tmp64, 10000);
+	fdo_min = (u16)tmp64;
+
+	t_diff1 = rdev->pm.dpm.fan.t_med - rdev->pm.dpm.fan.t_min;
+	t_diff2 = rdev->pm.dpm.fan.t_high - rdev->pm.dpm.fan.t_med;
+
+	pwm_diff1 = rdev->pm.dpm.fan.pwm_med - rdev->pm.dpm.fan.pwm_min;
+	pwm_diff2 = rdev->pm.dpm.fan.pwm_high - rdev->pm.dpm.fan.pwm_med;
+
+	slope1 = (u16)((50 + ((16 * duty100 * pwm_diff1) / t_diff1)) / 100);
+	slope2 = (u16)((50 + ((16 * duty100 * pwm_diff2) / t_diff2)) / 100);
+
+	fan_table.slope1 = cpu_to_be16(slope1);
+	fan_table.slope2 = cpu_to_be16(slope2);
+
+	fan_table.fdo_min = cpu_to_be16(fdo_min);
+
+	fan_table.hys_down = cpu_to_be16(rdev->pm.dpm.fan.t_hyst);
+
+	fan_table.hys_up = cpu_to_be16(1);
+
+	fan_table.hys_slope = cpu_to_be16(1);
+
+	fan_table.temp_resp_lim = cpu_to_be16(5);
+
+	reference_clock = radeon_get_xclk(rdev);
+
+	fan_table.refresh_period = cpu_to_be32((rdev->pm.dpm.fan.cycle_delay *
+						reference_clock) / 1600);
+
+	fan_table.fdo_max = cpu_to_be16((u16)duty100);
+
+	tmp = (RREG32(CG_MULT_THERMAL_CTRL) & TEMP_SEL_MASK) >> TEMP_SEL_SHIFT;
+	fan_table.temp_src = (uint8_t)tmp;
+
+	ret = si_copy_bytes_to_smc(rdev,
+				   si_pi->fan_table_start,
+				   (u8 *)(&fan_table),
+				   sizeof(fan_table),
+				   si_pi->sram_end);
+
+	if (ret) {
+		DRM_ERROR("Failed to load fan table to the SMC.");
+		rdev->pm.dpm.fan.ucode_fan_control = false;
+	}
+
+	return 0;
+}
+
+static int si_fan_ctrl_start_smc_fan_control(struct radeon_device *rdev)
+{
+	PPSMC_Result ret;
+
+	ret = si_send_msg_to_smc(rdev, PPSMC_StartFanControl);
+	if (ret == PPSMC_Result_OK)
+		return 0;
+	else
+		return -EINVAL;
+}
+
+static int si_fan_ctrl_stop_smc_fan_control(struct radeon_device *rdev)
+{
+	PPSMC_Result ret;
+
+	ret = si_send_msg_to_smc(rdev, PPSMC_StopFanControl);
+	if (ret == PPSMC_Result_OK)
+		return 0;
+	else
+		return -EINVAL;
+}
+
+#if 0
+static int si_fan_ctrl_get_fan_speed_percent(struct radeon_device *rdev,
+					     u32 *speed)
+{
+	u32 duty, duty100;
+	u64 tmp64;
+
+	if (rdev->pm.no_fan)
+		return -ENOENT;
+
+	duty100 = (RREG32(CG_FDO_CTRL1) & FMAX_DUTY100_MASK) >> FMAX_DUTY100_SHIFT;
+	duty = (RREG32(CG_THERMAL_STATUS) & FDO_PWM_DUTY_MASK) >> FDO_PWM_DUTY_SHIFT;
+
+	if (duty100 == 0)
+		return -EINVAL;
+
+	tmp64 = (u64)duty * 100;
+	do_div(tmp64, duty100);
+	*speed = (u32)tmp64;
+
+	if (*speed > 100)
+		*speed = 100;
+
+	return 0;
+}
+
+static int si_fan_ctrl_set_fan_speed_percent(struct radeon_device *rdev,
+					     u32 speed)
+{
+	u32 tmp;
+	u32 duty, duty100;
+	u64 tmp64;
+
+	if (rdev->pm.no_fan)
+		return -ENOENT;
+
+	if (speed > 100)
+		return -EINVAL;
+
+	if (rdev->pm.dpm.fan.ucode_fan_control)
+		si_fan_ctrl_stop_smc_fan_control(rdev);
+
+	duty100 = (RREG32(CG_FDO_CTRL1) & FMAX_DUTY100_MASK) >> FMAX_DUTY100_SHIFT;
+
+	if (duty100 == 0)
+		return -EINVAL;
+
+	tmp64 = (u64)speed * duty100;
+	do_div(tmp64, 100);
+	duty = (u32)tmp64;
+
+	tmp = RREG32(CG_FDO_CTRL0) & ~FDO_STATIC_DUTY_MASK;
+	tmp |= FDO_STATIC_DUTY(duty);
+	WREG32(CG_FDO_CTRL0, tmp);
+
+	si_fan_ctrl_set_static_mode(rdev, FDO_PWM_MODE_STATIC);
+
+	return 0;
+}
+
+static int si_fan_ctrl_get_fan_speed_rpm(struct radeon_device *rdev,
+					 u32 *speed)
+{
+	u32 tach_period;
+	u32 xclk = radeon_get_xclk(rdev);
+
+	if (rdev->pm.no_fan)
+		return -ENOENT;
+
+	if (rdev->pm.fan_pulses_per_revolution == 0)
+		return -ENOENT;
+
+	tach_period = (RREG32(CG_TACH_STATUS) & TACH_PERIOD_MASK) >> TACH_PERIOD_SHIFT;
+	if (tach_period == 0)
+		return -ENOENT;
+
+	*speed = 60 * xclk * 10000 / tach_period;
+
+	return 0;
+}
+
+static int si_fan_ctrl_set_fan_speed_rpm(struct radeon_device *rdev,
+					 u32 speed)
+{
+	u32 tach_period, tmp;
+	u32 xclk = radeon_get_xclk(rdev);
+
+	if (rdev->pm.no_fan)
+		return -ENOENT;
+
+	if (rdev->pm.fan_pulses_per_revolution == 0)
+		return -ENOENT;
+
+	if ((speed < rdev->pm.fan_min_rpm) ||
+	    (speed > rdev->pm.fan_max_rpm))
+		return -EINVAL;
+
+	if (rdev->pm.dpm.fan.ucode_fan_control)
+		si_fan_ctrl_stop_smc_fan_control(rdev);
+
+	tach_period = 60 * xclk * 10000 / (8 * speed);
+	tmp = RREG32(CG_TACH_CTRL) & ~TARGET_PERIOD_MASK;
+	tmp |= TARGET_PERIOD(tach_period);
+	WREG32(CG_TACH_CTRL, tmp);
+
+	si_fan_ctrl_set_static_mode(rdev, FDO_PWM_MODE_STATIC_RPM);
+
+	return 0;
+}
+#endif
+
+static void si_fan_ctrl_set_default_mode(struct radeon_device *rdev)
+{
+	struct si_power_info *si_pi = si_get_pi(rdev);
+	u32 tmp;
+
+	if (!si_pi->fan_ctrl_is_in_default_mode) {
+		tmp = RREG32(CG_FDO_CTRL2) & ~FDO_PWM_MODE_MASK;
+		tmp |= FDO_PWM_MODE(si_pi->fan_ctrl_default_mode);
+		WREG32(CG_FDO_CTRL2, tmp);
+
+		tmp = RREG32(CG_FDO_CTRL2) & ~TMIN_MASK;
+		tmp |= TMIN(si_pi->t_min);
+		WREG32(CG_FDO_CTRL2, tmp);
+		si_pi->fan_ctrl_is_in_default_mode = true;
+	}
+}
+
+static void si_thermal_start_smc_fan_control(struct radeon_device *rdev)
+{
+	if (rdev->pm.dpm.fan.ucode_fan_control) {
+		si_fan_ctrl_start_smc_fan_control(rdev);
+		si_fan_ctrl_set_static_mode(rdev, FDO_PWM_MODE_STATIC);
+	}
+}
+
+static void si_thermal_initialize(struct radeon_device *rdev)
+{
+	u32 tmp;
+
+	if (rdev->pm.fan_pulses_per_revolution) {
+		tmp = RREG32(CG_TACH_CTRL) & ~EDGE_PER_REV_MASK;
+		tmp |= EDGE_PER_REV(rdev->pm.fan_pulses_per_revolution -1);
+		WREG32(CG_TACH_CTRL, tmp);
+	}
+
+	tmp = RREG32(CG_FDO_CTRL2) & ~TACH_PWM_RESP_RATE_MASK;
+	tmp |= TACH_PWM_RESP_RATE(0x28);
+	WREG32(CG_FDO_CTRL2, tmp);
+}
+
+static int si_thermal_start_thermal_controller(struct radeon_device *rdev)
+{
+	int ret;
+
+	si_thermal_initialize(rdev);
+	ret = si_thermal_set_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
+	if (ret)
+		return ret;
+	ret = si_thermal_enable_alert(rdev, true);
+	if (ret)
+		return ret;
+	if (rdev->pm.dpm.fan.ucode_fan_control) {
+		ret = si_halt_smc(rdev);
+		if (ret)
+			return ret;
+		ret = si_thermal_setup_fan_table(rdev);
+		if (ret)
+			return ret;
+		ret = si_resume_smc(rdev);
+		if (ret)
+			return ret;
+		si_thermal_start_smc_fan_control(rdev);
+	}
+
+	return 0;
+}
+
+static void si_thermal_stop_thermal_controller(struct radeon_device *rdev)
+{
+	if (!rdev->pm.no_fan) {
+		si_fan_ctrl_set_default_mode(rdev);
+		si_fan_ctrl_stop_smc_fan_control(rdev);
+	}
+}
+
 int si_dpm_enable(struct radeon_device *rdev)
 {
 	struct rv7xx_power_info *pi = rv770_get_pi(rdev);
@@ -5953,31 +6330,39 @@ int si_dpm_enable(struct radeon_device *rdev)
 
 	si_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
 
+	si_thermal_start_thermal_controller(rdev);
+
 	ni_update_current_ps(rdev, boot_ps);
 
 	return 0;
 }
 
-int si_dpm_late_enable(struct radeon_device *rdev)
+static int si_set_temperature_range(struct radeon_device *rdev)
 {
 	int ret;
 
-	if (rdev->irq.installed &&
-	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
-		PPSMC_Result result;
+	ret = si_thermal_enable_alert(rdev, false);
+	if (ret)
+		return ret;
+	ret = si_thermal_set_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
+	if (ret)
+		return ret;
+	ret = si_thermal_enable_alert(rdev, true);
+	if (ret)
+		return ret;
 
-		ret = si_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
-		if (ret)
-			return ret;
-		rdev->irq.dpm_thermal = true;
-		radeon_irq_set(rdev);
-		result = si_send_msg_to_smc(rdev, PPSMC_MSG_EnableThermalInterrupt);
+	return ret;
+}
 
-		if (result != PPSMC_Result_OK)
-			DRM_DEBUG_KMS("Could not enable thermal interrupts.\n");
-	}
+int si_dpm_late_enable(struct radeon_device *rdev)
+{
+	int ret;
 
-	return 0;
+	ret = si_set_temperature_range(rdev);
+	if (ret)
+		return ret;
+
+	return ret;
 }
 
 void si_dpm_disable(struct radeon_device *rdev)
@@ -5987,6 +6372,7 @@ void si_dpm_disable(struct radeon_device *rdev)
 
 	if (!si_is_smc_running(rdev))
 		return;
+	si_thermal_stop_thermal_controller(rdev);
 	si_disable_ulv(rdev);
 	si_clear_vc(rdev);
 	if (pi->thermal_protection)
@@ -6255,7 +6641,7 @@ static void si_parse_pplib_clock_info(struct radeon_device *rdev,
 	if ((rps->class2 & ATOM_PPLIB_CLASSIFICATION2_ULV) &&
 	    index == 0) {
 		/* XXX disable for A0 tahiti */
-		si_pi->ulv.supported = true;
+		si_pi->ulv.supported = false;
 		si_pi->ulv.pl = *pl;
 		si_pi->ulv.one_pcie_lane_in_ulv = false;
 		si_pi->ulv.volt_change_delay = SISLANDS_ULVVOLTAGECHANGEDELAY_DFLT;
@@ -6525,6 +6911,9 @@ int si_dpm_init(struct radeon_device *rdev)
 		rdev->pm.dpm.dyn_state.max_clock_voltage_on_dc =
 			rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac;
 
+	si_pi->fan_ctrl_is_in_default_mode = true;
+	rdev->pm.dpm.fan.ucode_fan_control = false;
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/radeon/si_dpm.h b/drivers/gpu/drm/radeon/si_dpm.h
index 8b5c06a0832d..d16bb1b5f10f 100644
--- a/drivers/gpu/drm/radeon/si_dpm.h
+++ b/drivers/gpu/drm/radeon/si_dpm.h
@@ -182,6 +182,7 @@ struct si_power_info {
 	u32 dte_table_start;
 	u32 spll_table_start;
 	u32 papm_cfg_table_start;
+	u32 fan_table_start;
 	/* CAC stuff */
 	const struct si_cac_config_reg *cac_weights;
 	const struct si_cac_config_reg *lcac_config;
@@ -197,6 +198,10 @@ struct si_power_info {
 	/* SVI2 */
 	u8 svd_gpio_id;
 	u8 svc_gpio_id;
+	/* fan control */
+	bool fan_ctrl_is_in_default_mode;
+	u32 t_min;
+	u32 fan_ctrl_default_mode;
 };
 
 #define SISLANDS_INITIAL_STATE_ARB_INDEX    0
diff --git a/drivers/gpu/drm/radeon/si_smc.c b/drivers/gpu/drm/radeon/si_smc.c
index 73dbc79c959d..e5bb92f16775 100644
--- a/drivers/gpu/drm/radeon/si_smc.c
+++ b/drivers/gpu/drm/radeon/si_smc.c
@@ -135,7 +135,7 @@ void si_reset_smc(struct radeon_device *rdev)
 
 int si_program_jump_on_start(struct radeon_device *rdev)
 {
-	static u8 data[] = { 0x0E, 0x00, 0x40, 0x40 };
+	static const u8 data[] = { 0x0E, 0x00, 0x40, 0x40 };
 
 	return si_copy_bytes_to_smc(rdev, 0x0, data, 4, sizeof(data)+1);
 }
diff --git a/drivers/gpu/drm/radeon/sid.h b/drivers/gpu/drm/radeon/sid.h
index fd414d34d885..84999242c747 100644
--- a/drivers/gpu/drm/radeon/sid.h
+++ b/drivers/gpu/drm/radeon/sid.h
@@ -180,7 +180,10 @@
 #define		DIG_THERM_DPM(x)			((x) << 14)
 #define		DIG_THERM_DPM_MASK			0x003FC000
 #define		DIG_THERM_DPM_SHIFT			14
-
+#define	CG_THERMAL_STATUS				0x704
+#define		FDO_PWM_DUTY(x)				((x) << 9)
+#define		FDO_PWM_DUTY_MASK			(0xff << 9)
+#define		FDO_PWM_DUTY_SHIFT			9
 #define	CG_THERMAL_INT					0x708
 #define		DIG_THERM_INTH(x)			((x) << 8)
 #define		DIG_THERM_INTH_MASK			0x0000FF00
@@ -191,6 +194,10 @@
 #define 	THERM_INT_MASK_HIGH			(1 << 24)
 #define 	THERM_INT_MASK_LOW			(1 << 25)
 
+#define	CG_MULT_THERMAL_CTRL					0x710
+#define		TEMP_SEL(x)					((x) << 20)
+#define		TEMP_SEL_MASK					(0xff << 20)
+#define		TEMP_SEL_SHIFT					20
 #define	CG_MULT_THERMAL_STATUS					0x714
 #define		ASIC_MAX_TEMP(x)				((x) << 0)
 #define		ASIC_MAX_TEMP_MASK				0x000001ff
@@ -199,6 +206,37 @@
 #define		CTF_TEMP_MASK					0x0003fe00
 #define		CTF_TEMP_SHIFT					9
 
+#define	CG_FDO_CTRL0					0x754
+#define		FDO_STATIC_DUTY(x)			((x) << 0)
+#define		FDO_STATIC_DUTY_MASK			0x000000FF
+#define		FDO_STATIC_DUTY_SHIFT			0
+#define	CG_FDO_CTRL1					0x758
+#define		FMAX_DUTY100(x)				((x) << 0)
+#define		FMAX_DUTY100_MASK			0x000000FF
+#define		FMAX_DUTY100_SHIFT			0
+#define	CG_FDO_CTRL2					0x75C
+#define		TMIN(x)					((x) << 0)
+#define		TMIN_MASK				0x000000FF
+#define		TMIN_SHIFT				0
+#define		FDO_PWM_MODE(x)				((x) << 11)
+#define		FDO_PWM_MODE_MASK			(7 << 11)
+#define		FDO_PWM_MODE_SHIFT			11
+#define		TACH_PWM_RESP_RATE(x)			((x) << 25)
+#define		TACH_PWM_RESP_RATE_MASK			(0x7f << 25)
+#define		TACH_PWM_RESP_RATE_SHIFT		25
+
+#define CG_TACH_CTRL                                    0x770
+#       define EDGE_PER_REV(x)                          ((x) << 0)
+#       define EDGE_PER_REV_MASK                        (0x7 << 0)
+#       define EDGE_PER_REV_SHIFT                       0
+#       define TARGET_PERIOD(x)                         ((x) << 3)
+#       define TARGET_PERIOD_MASK                       0xfffffff8
+#       define TARGET_PERIOD_SHIFT                      3
+#define CG_TACH_STATUS                                  0x774
+#       define TACH_PERIOD(x)                           ((x) << 0)
+#       define TACH_PERIOD_MASK                         0xffffffff
+#       define TACH_PERIOD_SHIFT                        0
+
 #define GENERAL_PWRMGT                                  0x780
 #       define GLOBAL_PWRMGT_EN                         (1 << 0)
 #       define STATIC_PM_EN                             (1 << 1)
@@ -736,7 +774,7 @@
 #       define DESCRIPTION16(x)                          (((x) & 0xff) << 0)
 #       define DESCRIPTION17(x)                          (((x) & 0xff) << 8)
 
-#define AZ_F0_CODEC_PIN_CONTROL_HOTPLUG_CONTROL          0x54
+#define AZ_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL         0x54
 #       define AUDIO_ENABLED                             (1 << 31)
 
 #define AZ_F0_CODEC_PIN_CONTROL_RESPONSE_CONFIGURATION_DEFAULT  0x56
@@ -1594,6 +1632,23 @@
 #define	PACKET3_MPEG_INDEX				0x3A
 #define	PACKET3_COPY_DW					0x3B
 #define	PACKET3_WAIT_REG_MEM				0x3C
+#define		WAIT_REG_MEM_FUNCTION(x)                ((x) << 0)
+                /* 0 - always
+		 * 1 - <
+		 * 2 - <=
+		 * 3 - ==
+		 * 4 - !=
+		 * 5 - >=
+		 * 6 - >
+		 */
+#define		WAIT_REG_MEM_MEM_SPACE(x)               ((x) << 4)
+                /* 0 - reg
+		 * 1 - mem
+		 */
+#define		WAIT_REG_MEM_ENGINE(x)                  ((x) << 8)
+                /* 0 - me
+		 * 1 - pfp
+		 */
 #define	PACKET3_MEM_WRITE				0x3D
 #define	PACKET3_COPY_DATA				0x40
 #define	PACKET3_CP_DMA					0x41
@@ -1797,6 +1852,7 @@
 #define	DMA_PACKET_TRAP					  0x7
 #define	DMA_PACKET_SRBM_WRITE				  0x9
 #define	DMA_PACKET_CONSTANT_FILL			  0xd
+#define	DMA_PACKET_POLL_REG_MEM				  0xe
 #define	DMA_PACKET_NOP					  0xf
 
 #define VCE_STATUS					0x20004
diff --git a/drivers/gpu/drm/radeon/sislands_smc.h b/drivers/gpu/drm/radeon/sislands_smc.h
index 623a0b1e2d9d..3c779838d9ab 100644
--- a/drivers/gpu/drm/radeon/sislands_smc.h
+++ b/drivers/gpu/drm/radeon/sislands_smc.h
@@ -245,6 +245,31 @@ typedef struct SISLANDS_SMC_STATETABLE SISLANDS_SMC_STATETABLE;
 #define SI_SMC_SOFT_REGISTER_svi_rework_gpio_id_svd   0x11c
 #define SI_SMC_SOFT_REGISTER_svi_rework_gpio_id_svc   0x120
 
+struct PP_SIslands_FanTable
+{
+	uint8_t  fdo_mode;
+	uint8_t  padding;
+	int16_t  temp_min;
+	int16_t  temp_med;
+	int16_t  temp_max;
+	int16_t  slope1;
+	int16_t  slope2;
+	int16_t  fdo_min;
+	int16_t  hys_up;
+	int16_t  hys_down;
+	int16_t  hys_slope;
+	int16_t  temp_resp_lim;
+	int16_t  temp_curr;
+	int16_t  slope_curr;
+	int16_t  pwm_curr;
+	uint32_t refresh_period;
+	int16_t  fdo_max;
+	uint8_t  temp_src;
+	int8_t  padding2;
+};
+
+typedef struct PP_SIslands_FanTable PP_SIslands_FanTable;
+
 #define SMC_SISLANDS_LKGE_LUT_NUM_OF_TEMP_ENTRIES 16
 #define SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES 32
 
diff --git a/drivers/gpu/drm/radeon/smu7_discrete.h b/drivers/gpu/drm/radeon/smu7_discrete.h
index 82f70c90a9ee..0b0b404ff091 100644
--- a/drivers/gpu/drm/radeon/smu7_discrete.h
+++ b/drivers/gpu/drm/radeon/smu7_discrete.h
@@ -431,6 +431,31 @@ struct SMU7_Discrete_MCRegisters
 
 typedef struct SMU7_Discrete_MCRegisters SMU7_Discrete_MCRegisters;
 
+struct SMU7_Discrete_FanTable
+{
+	uint16_t FdoMode;
+	int16_t  TempMin;
+	int16_t  TempMed;
+	int16_t  TempMax;
+	int16_t  Slope1;
+	int16_t  Slope2;
+	int16_t  FdoMin;
+	int16_t  HystUp;
+	int16_t  HystDown;
+	int16_t  HystSlope;
+	int16_t  TempRespLim;
+	int16_t  TempCurr;
+	int16_t  SlopeCurr;
+	int16_t  PwmCurr;
+	uint32_t RefreshPeriod;
+	int16_t  FdoMax;
+	uint8_t  TempSrc;
+	int8_t   Padding;
+};
+
+typedef struct SMU7_Discrete_FanTable SMU7_Discrete_FanTable;
+
+
 struct SMU7_Discrete_PmFuses {
   // dw0-dw1
   uint8_t BapmVddCVidHiSidd[8];
@@ -462,7 +487,10 @@ struct SMU7_Discrete_PmFuses {
   uint8_t BapmVddCVidHiSidd2[8];
 
   // dw11-dw12
-  uint32_t Reserved6[2];
+  int16_t FuzzyFan_ErrorSetDelta;
+  int16_t FuzzyFan_ErrorRateSetDelta;
+  int16_t FuzzyFan_PwmSetDelta;
+  uint16_t CalcMeasPowerBlend;
 
   // dw13-dw16
   uint8_t GnbLPML[16];
diff --git a/drivers/gpu/drm/radeon/sumo_dpm.c b/drivers/gpu/drm/radeon/sumo_dpm.c
index 3f0e8d7b8dbe..1f8a8833e1be 100644
--- a/drivers/gpu/drm/radeon/sumo_dpm.c
+++ b/drivers/gpu/drm/radeon/sumo_dpm.c
@@ -23,6 +23,7 @@
 
 #include "drmP.h"
 #include "radeon.h"
+#include "radeon_asic.h"
 #include "sumod.h"
 #include "r600_dpm.h"
 #include "cypress_dpm.h"
diff --git a/drivers/gpu/drm/radeon/trinity_dpm.c b/drivers/gpu/drm/radeon/trinity_dpm.c
index 57f780053b3e..b4ec5c4e7969 100644
--- a/drivers/gpu/drm/radeon/trinity_dpm.c
+++ b/drivers/gpu/drm/radeon/trinity_dpm.c
@@ -23,6 +23,7 @@
 
 #include "drmP.h"
 #include "radeon.h"
+#include "radeon_asic.h"
 #include "trinityd.h"
 #include "r600_dpm.h"
 #include "trinity_dpm.h"
diff --git a/drivers/gpu/drm/radeon/uvd_v1_0.c b/drivers/gpu/drm/radeon/uvd_v1_0.c
index cda391347286..e72b3cb59358 100644
--- a/drivers/gpu/drm/radeon/uvd_v1_0.c
+++ b/drivers/gpu/drm/radeon/uvd_v1_0.c
@@ -22,6 +22,7 @@
  * Authors: Christian König <christian.koenig@amd.com>
  */
 
+#include <linux/firmware.h>
 #include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
@@ -70,6 +71,82 @@ void uvd_v1_0_set_wptr(struct radeon_device *rdev,
 }
 
 /**
+ * uvd_v1_0_fence_emit - emit an fence & trap command
+ *
+ * @rdev: radeon_device pointer
+ * @fence: fence to emit
+ *
+ * Write a fence and a trap command to the ring.
+ */
+void uvd_v1_0_fence_emit(struct radeon_device *rdev,
+			 struct radeon_fence *fence)
+{
+	struct radeon_ring *ring = &rdev->ring[fence->ring];
+	uint64_t addr = rdev->fence_drv[fence->ring].gpu_addr;
+
+	radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_DATA0, 0));
+	radeon_ring_write(ring, addr & 0xffffffff);
+	radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_DATA1, 0));
+	radeon_ring_write(ring, fence->seq);
+	radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_CMD, 0));
+	radeon_ring_write(ring, 0);
+
+	radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_DATA0, 0));
+	radeon_ring_write(ring, 0);
+	radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_DATA1, 0));
+	radeon_ring_write(ring, 0);
+	radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_CMD, 0));
+	radeon_ring_write(ring, 2);
+	return;
+}
+
+/**
+ * uvd_v1_0_resume - memory controller programming
+ *
+ * @rdev: radeon_device pointer
+ *
+ * Let the UVD memory controller know it's offsets
+ */
+int uvd_v1_0_resume(struct radeon_device *rdev)
+{
+	uint64_t addr;
+	uint32_t size;
+	int r;
+
+	r = radeon_uvd_resume(rdev);
+	if (r)
+		return r;
+
+	/* programm the VCPU memory controller bits 0-27 */
+	addr = (rdev->uvd.gpu_addr >> 3) + 16;
+	size = RADEON_GPU_PAGE_ALIGN(rdev->uvd_fw->size) >> 3;
+	WREG32(UVD_VCPU_CACHE_OFFSET0, addr);
+	WREG32(UVD_VCPU_CACHE_SIZE0, size);
+
+	addr += size;
+	size = RADEON_UVD_STACK_SIZE >> 3;
+	WREG32(UVD_VCPU_CACHE_OFFSET1, addr);
+	WREG32(UVD_VCPU_CACHE_SIZE1, size);
+
+	addr += size;
+	size = RADEON_UVD_HEAP_SIZE >> 3;
+	WREG32(UVD_VCPU_CACHE_OFFSET2, addr);
+	WREG32(UVD_VCPU_CACHE_SIZE2, size);
+
+	/* bits 28-31 */
+	addr = (rdev->uvd.gpu_addr >> 28) & 0xF;
+	WREG32(UVD_LMI_ADDR_EXT, (addr << 12) | (addr << 0));
+
+	/* bits 32-39 */
+	addr = (rdev->uvd.gpu_addr >> 32) & 0xFF;
+	WREG32(UVD_LMI_EXT40_ADDR, addr | (0x9 << 16) | (0x1 << 31));
+
+	WREG32(UVD_FW_START, *((uint32_t*)rdev->uvd.cpu_addr));
+
+	return 0;
+}
+
+/**
  * uvd_v1_0_init - start and test UVD block
  *
  * @rdev: radeon_device pointer
@@ -130,8 +207,32 @@ done:
 	/* lower clocks again */
 	radeon_set_uvd_clocks(rdev, 0, 0);
 
-	if (!r)
+	if (!r) {
+		switch (rdev->family) {
+		case CHIP_RV610:
+		case CHIP_RV630:
+		case CHIP_RV620:
+			/* 64byte granularity workaround */
+			WREG32(MC_CONFIG, 0);
+			WREG32(MC_CONFIG, 1 << 4);
+			WREG32(RS_DQ_RD_RET_CONF, 0x3f);
+			WREG32(MC_CONFIG, 0x1f);
+
+			/* fall through */
+		case CHIP_RV670:
+		case CHIP_RV635:
+
+			/* write clean workaround */
+			WREG32_P(UVD_VCPU_CNTL, 0x10, ~0x10);
+			break;
+
+		default:
+			/* TODO: Do we need more? */
+			break;
+		}
+
 		DRM_INFO("UVD initialized successfully.\n");
+	}
 
 	return r;
 }
@@ -218,12 +319,12 @@ int uvd_v1_0_start(struct radeon_device *rdev)
 	/* enable UMC */
 	WREG32_P(UVD_LMI_CTRL2, 0, ~(1 << 8));
 
+	WREG32_P(UVD_RB_ARB_CTRL, 0, ~(1 << 3));
+
 	/* boot up the VCPU */
 	WREG32(UVD_SOFT_RESET, 0);
 	mdelay(10);
 
-	WREG32_P(UVD_RB_ARB_CTRL, 0, ~(1 << 3));
-
 	for (i = 0; i < 10; ++i) {
 		uint32_t status;
 		for (j = 0; j < 100; ++j) {
diff --git a/drivers/gpu/drm/radeon/uvd_v2_2.c b/drivers/gpu/drm/radeon/uvd_v2_2.c
index 8bfdadd56598..89193519f8a1 100644
--- a/drivers/gpu/drm/radeon/uvd_v2_2.c
+++ b/drivers/gpu/drm/radeon/uvd_v2_2.c
@@ -72,6 +72,10 @@ int uvd_v2_2_resume(struct radeon_device *rdev)
 	uint32_t chip_id, size;
 	int r;
 
+	/* RV770 uses V1.0 MC */
+	if (rdev->family == CHIP_RV770)
+		return uvd_v1_0_resume(rdev);
+
 	r = radeon_uvd_resume(rdev);
 	if (r)
 		return r;
diff --git a/drivers/gpu/drm/rcar-du/Kconfig b/drivers/gpu/drm/rcar-du/Kconfig
index 2e3d7b5b0ad7..2324a526de65 100644
--- a/drivers/gpu/drm/rcar-du/Kconfig
+++ b/drivers/gpu/drm/rcar-du/Kconfig
@@ -6,14 +6,22 @@ config DRM_RCAR_DU
 	select DRM_KMS_CMA_HELPER
 	select DRM_GEM_CMA_HELPER
 	select DRM_KMS_FB_HELPER
+	select VIDEOMODE_HELPERS
 	help
 	  Choose this option if you have an R-Car chipset.
 	  If M is selected the module will be called rcar-du-drm.
 
+config DRM_RCAR_HDMI
+	bool "R-Car DU HDMI Encoder Support"
+	depends on DRM_RCAR_DU
+	depends on OF
+	help
+	  Enable support for external HDMI encoders.
+
 config DRM_RCAR_LVDS
 	bool "R-Car DU LVDS Encoder Support"
 	depends on DRM_RCAR_DU
 	depends on ARCH_R8A7790 || ARCH_R8A7791 || COMPILE_TEST
 	help
-	  Enable support the R-Car Display Unit embedded LVDS encoders
-	  (currently only on R8A7790).
+	  Enable support for the R-Car Display Unit embedded LVDS encoders
+	  (currently only on R8A7790 and R8A7791).
diff --git a/drivers/gpu/drm/rcar-du/Makefile b/drivers/gpu/drm/rcar-du/Makefile
index 12b8d4477835..05de1c4097af 100644
--- a/drivers/gpu/drm/rcar-du/Makefile
+++ b/drivers/gpu/drm/rcar-du/Makefile
@@ -7,6 +7,8 @@ rcar-du-drm-y := rcar_du_crtc.o \
 		 rcar_du_plane.o \
 		 rcar_du_vgacon.o
 
+rcar-du-drm-$(CONFIG_DRM_RCAR_HDMI)	+= rcar_du_hdmicon.o \
+					   rcar_du_hdmienc.o
 rcar-du-drm-$(CONFIG_DRM_RCAR_LVDS)	+= rcar_du_lvdsenc.o
 
 obj-$(CONFIG_DRM_RCAR_DU)		+= rcar-du-drm.o
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
index 299267db2898..23cc910951f4 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_crtc.c  --  R-Car Display Unit CRTCs
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -19,6 +19,7 @@
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_fb_cma_helper.h>
 #include <drm/drm_gem_cma_helper.h>
+#include <drm/drm_plane_helper.h>
 
 #include "rcar_du_crtc.h"
 #include "rcar_du_drv.h"
@@ -585,7 +586,7 @@ int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index)
 
 	if (irq < 0) {
 		dev_err(rcdu->dev, "no IRQ for CRTC %u\n", index);
-		return ret;
+		return irq;
 	}
 
 	ret = devm_request_irq(rcdu->dev, irq, rcar_du_crtc_irq, irqflags,
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_crtc.h b/drivers/gpu/drm/rcar-du/rcar_du_crtc.h
index 43e7575c700c..984e6083699f 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_crtc.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_crtc.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_crtc.h  --  R-Car Display Unit CRTCs
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -15,7 +15,6 @@
 #define __RCAR_DU_CRTC_H__
 
 #include <linux/mutex.h>
-#include <linux/platform_data/rcar-du.h>
 
 #include <drm/drmP.h>
 #include <drm/drm_crtc.h>
@@ -41,6 +40,15 @@ struct rcar_du_crtc {
 
 #define to_rcar_crtc(c)	container_of(c, struct rcar_du_crtc, crtc)
 
+enum rcar_du_output {
+	RCAR_DU_OUTPUT_DPAD0,
+	RCAR_DU_OUTPUT_DPAD1,
+	RCAR_DU_OUTPUT_LVDS0,
+	RCAR_DU_OUTPUT_LVDS1,
+	RCAR_DU_OUTPUT_TCON,
+	RCAR_DU_OUTPUT_MAX,
+};
+
 int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index);
 void rcar_du_crtc_enable_vblank(struct rcar_du_crtc *rcrtc, bool enable);
 void rcar_du_crtc_cancel_page_flip(struct rcar_du_crtc *rcrtc,
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.c b/drivers/gpu/drm/rcar-du/rcar_du_drv.c
index fda64b7b73e8..7bfa09cf18d5 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_drv.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_drv.c  --  R-Car Display Unit DRM driver
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -15,6 +15,7 @@
 #include <linux/io.h>
 #include <linux/mm.h>
 #include <linux/module.h>
+#include <linux/of_device.h>
 #include <linux/platform_device.h>
 #include <linux/pm.h>
 #include <linux/slab.h>
@@ -30,6 +31,97 @@
 #include "rcar_du_regs.h"
 
 /* -----------------------------------------------------------------------------
+ * Device Information
+ */
+
+static const struct rcar_du_device_info rcar_du_r8a7779_info = {
+	.features = 0,
+	.num_crtcs = 2,
+	.routes = {
+		/* R8A7779 has two RGB outputs and one (currently unsupported)
+		 * TCON output.
+		 */
+		[RCAR_DU_OUTPUT_DPAD0] = {
+			.possible_crtcs = BIT(0),
+			.encoder_type = DRM_MODE_ENCODER_NONE,
+			.port = 0,
+		},
+		[RCAR_DU_OUTPUT_DPAD1] = {
+			.possible_crtcs = BIT(1) | BIT(0),
+			.encoder_type = DRM_MODE_ENCODER_NONE,
+			.port = 1,
+		},
+	},
+	.num_lvds = 0,
+};
+
+static const struct rcar_du_device_info rcar_du_r8a7790_info = {
+	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK | RCAR_DU_FEATURE_DEFR8,
+	.quirks = RCAR_DU_QUIRK_ALIGN_128B | RCAR_DU_QUIRK_LVDS_LANES,
+	.num_crtcs = 3,
+	.routes = {
+		/* R8A7790 has one RGB output, two LVDS outputs and one
+		 * (currently unsupported) TCON output.
+		 */
+		[RCAR_DU_OUTPUT_DPAD0] = {
+			.possible_crtcs = BIT(2) | BIT(1) | BIT(0),
+			.encoder_type = DRM_MODE_ENCODER_NONE,
+			.port = 0,
+		},
+		[RCAR_DU_OUTPUT_LVDS0] = {
+			.possible_crtcs = BIT(0),
+			.encoder_type = DRM_MODE_ENCODER_LVDS,
+			.port = 1,
+		},
+		[RCAR_DU_OUTPUT_LVDS1] = {
+			.possible_crtcs = BIT(2) | BIT(1),
+			.encoder_type = DRM_MODE_ENCODER_LVDS,
+			.port = 2,
+		},
+	},
+	.num_lvds = 2,
+};
+
+static const struct rcar_du_device_info rcar_du_r8a7791_info = {
+	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK | RCAR_DU_FEATURE_DEFR8,
+	.num_crtcs = 2,
+	.routes = {
+		/* R8A7791 has one RGB output, one LVDS output and one
+		 * (currently unsupported) TCON output.
+		 */
+		[RCAR_DU_OUTPUT_DPAD0] = {
+			.possible_crtcs = BIT(1),
+			.encoder_type = DRM_MODE_ENCODER_NONE,
+			.port = 0,
+		},
+		[RCAR_DU_OUTPUT_LVDS0] = {
+			.possible_crtcs = BIT(0),
+			.encoder_type = DRM_MODE_ENCODER_LVDS,
+			.port = 1,
+		},
+	},
+	.num_lvds = 1,
+};
+
+static const struct platform_device_id rcar_du_id_table[] = {
+	{ "rcar-du-r8a7779", (kernel_ulong_t)&rcar_du_r8a7779_info },
+	{ "rcar-du-r8a7790", (kernel_ulong_t)&rcar_du_r8a7790_info },
+	{ "rcar-du-r8a7791", (kernel_ulong_t)&rcar_du_r8a7791_info },
+	{ }
+};
+
+MODULE_DEVICE_TABLE(platform, rcar_du_id_table);
+
+static const struct of_device_id rcar_du_of_table[] = {
+	{ .compatible = "renesas,du-r8a7779", .data = &rcar_du_r8a7779_info },
+	{ .compatible = "renesas,du-r8a7790", .data = &rcar_du_r8a7790_info },
+	{ .compatible = "renesas,du-r8a7791", .data = &rcar_du_r8a7791_info },
+	{ }
+};
+
+MODULE_DEVICE_TABLE(of, rcar_du_of_table);
+
+/* -----------------------------------------------------------------------------
  * DRM operations
  */
 
@@ -53,12 +145,12 @@ static int rcar_du_unload(struct drm_device *dev)
 static int rcar_du_load(struct drm_device *dev, unsigned long flags)
 {
 	struct platform_device *pdev = dev->platformdev;
-	struct rcar_du_platform_data *pdata = pdev->dev.platform_data;
+	struct device_node *np = pdev->dev.of_node;
 	struct rcar_du_device *rcdu;
 	struct resource *mem;
 	int ret;
 
-	if (pdata == NULL) {
+	if (np == NULL) {
 		dev_err(dev->dev, "no platform data\n");
 		return -ENODEV;
 	}
@@ -70,8 +162,8 @@ static int rcar_du_load(struct drm_device *dev, unsigned long flags)
 	}
 
 	rcdu->dev = &pdev->dev;
-	rcdu->pdata = pdata;
-	rcdu->info = (struct rcar_du_device_info *)pdev->id_entry->driver_data;
+	rcdu->info = np ? of_match_device(rcar_du_of_table, rcdu->dev)->data
+		   : (void *)platform_get_device_id(pdev)->driver_data;
 	rcdu->ddev = dev;
 	dev->dev_private = rcdu;
 
@@ -158,6 +250,7 @@ static struct drm_driver rcar_du_driver = {
 	.unload			= rcar_du_unload,
 	.preclose		= rcar_du_preclose,
 	.lastclose		= rcar_du_lastclose,
+	.set_busid		= drm_platform_set_busid,
 	.get_vblank_counter	= drm_vblank_count,
 	.enable_vblank		= rcar_du_enable_vblank,
 	.disable_vblank		= rcar_du_disable_vblank,
@@ -231,84 +324,13 @@ static int rcar_du_remove(struct platform_device *pdev)
 	return 0;
 }
 
-static const struct rcar_du_device_info rcar_du_r8a7779_info = {
-	.features = 0,
-	.num_crtcs = 2,
-	.routes = {
-		/* R8A7779 has two RGB outputs and one (currently unsupported)
-		 * TCON output.
-		 */
-		[RCAR_DU_OUTPUT_DPAD0] = {
-			.possible_crtcs = BIT(0),
-			.encoder_type = DRM_MODE_ENCODER_NONE,
-		},
-		[RCAR_DU_OUTPUT_DPAD1] = {
-			.possible_crtcs = BIT(1) | BIT(0),
-			.encoder_type = DRM_MODE_ENCODER_NONE,
-		},
-	},
-	.num_lvds = 0,
-};
-
-static const struct rcar_du_device_info rcar_du_r8a7790_info = {
-	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK | RCAR_DU_FEATURE_DEFR8,
-	.quirks = RCAR_DU_QUIRK_ALIGN_128B | RCAR_DU_QUIRK_LVDS_LANES,
-	.num_crtcs = 3,
-	.routes = {
-		/* R8A7790 has one RGB output, two LVDS outputs and one
-		 * (currently unsupported) TCON output.
-		 */
-		[RCAR_DU_OUTPUT_DPAD0] = {
-			.possible_crtcs = BIT(2) | BIT(1) | BIT(0),
-			.encoder_type = DRM_MODE_ENCODER_NONE,
-		},
-		[RCAR_DU_OUTPUT_LVDS0] = {
-			.possible_crtcs = BIT(0),
-			.encoder_type = DRM_MODE_ENCODER_LVDS,
-		},
-		[RCAR_DU_OUTPUT_LVDS1] = {
-			.possible_crtcs = BIT(2) | BIT(1),
-			.encoder_type = DRM_MODE_ENCODER_LVDS,
-		},
-	},
-	.num_lvds = 2,
-};
-
-static const struct rcar_du_device_info rcar_du_r8a7791_info = {
-	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK | RCAR_DU_FEATURE_DEFR8,
-	.num_crtcs = 2,
-	.routes = {
-		/* R8A7791 has one RGB output, one LVDS output and one
-		 * (currently unsupported) TCON output.
-		 */
-		[RCAR_DU_OUTPUT_DPAD0] = {
-			.possible_crtcs = BIT(1),
-			.encoder_type = DRM_MODE_ENCODER_NONE,
-		},
-		[RCAR_DU_OUTPUT_LVDS0] = {
-			.possible_crtcs = BIT(0),
-			.encoder_type = DRM_MODE_ENCODER_LVDS,
-		},
-	},
-	.num_lvds = 1,
-};
-
-static const struct platform_device_id rcar_du_id_table[] = {
-	{ "rcar-du-r8a7779", (kernel_ulong_t)&rcar_du_r8a7779_info },
-	{ "rcar-du-r8a7790", (kernel_ulong_t)&rcar_du_r8a7790_info },
-	{ "rcar-du-r8a7791", (kernel_ulong_t)&rcar_du_r8a7791_info },
-	{ }
-};
-
-MODULE_DEVICE_TABLE(platform, rcar_du_id_table);
-
 static struct platform_driver rcar_du_platform_driver = {
 	.probe		= rcar_du_probe,
 	.remove		= rcar_du_remove,
 	.driver		= {
-		.owner	= THIS_MODULE,
 		.name	= "rcar-du",
 		.pm	= &rcar_du_pm_ops,
+		.of_match_table = rcar_du_of_table,
 	},
 	.id_table	= rcar_du_id_table,
 };
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.h b/drivers/gpu/drm/rcar-du/rcar_du_drv.h
index e31b735d3f25..0a724669f02d 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_drv.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_drv.h  --  R-Car Display Unit DRM driver
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -15,7 +15,6 @@
 #define __RCAR_DU_DRV_H__
 
 #include <linux/kernel.h>
-#include <linux/platform_data/rcar-du.h>
 
 #include "rcar_du_crtc.h"
 #include "rcar_du_group.h"
@@ -37,6 +36,7 @@ struct rcar_du_lvdsenc;
  * struct rcar_du_output_routing - Output routing specification
  * @possible_crtcs: bitmask of possible CRTCs for the output
  * @encoder_type: DRM type of the internal encoder associated with the output
+ * @port: device tree port number corresponding to this output route
  *
  * The DU has 5 possible outputs (DPAD0/1, LVDS0/1, TCON). Output routing data
  * specify the valid SoC outputs, which CRTCs can drive the output, and the type
@@ -45,6 +45,7 @@ struct rcar_du_lvdsenc;
 struct rcar_du_output_routing {
 	unsigned int possible_crtcs;
 	unsigned int encoder_type;
+	unsigned int port;
 };
 
 /*
@@ -65,7 +66,6 @@ struct rcar_du_device_info {
 
 struct rcar_du_device {
 	struct device *dev;
-	const struct rcar_du_platform_data *pdata;
 	const struct rcar_du_device_info *info;
 
 	void __iomem *mmio;
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_encoder.c b/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
index 3daa7a168dc6..34a122a39664 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_encoder.c  --  R-Car Display Unit Encoder
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -19,6 +19,8 @@
 
 #include "rcar_du_drv.h"
 #include "rcar_du_encoder.h"
+#include "rcar_du_hdmicon.h"
+#include "rcar_du_hdmienc.h"
 #include "rcar_du_kms.h"
 #include "rcar_du_lvdscon.h"
 #include "rcar_du_lvdsenc.h"
@@ -33,7 +35,7 @@ rcar_du_connector_best_encoder(struct drm_connector *connector)
 {
 	struct rcar_du_connector *rcon = to_rcar_connector(connector);
 
-	return &rcon->encoder->encoder;
+	return rcar_encoder_to_drm_encoder(rcon->encoder);
 }
 
 /* -----------------------------------------------------------------------------
@@ -142,9 +144,11 @@ static const struct drm_encoder_funcs encoder_funcs = {
 int rcar_du_encoder_init(struct rcar_du_device *rcdu,
 			 enum rcar_du_encoder_type type,
 			 enum rcar_du_output output,
-			 const struct rcar_du_encoder_data *data)
+			 struct device_node *enc_node,
+			 struct device_node *con_node)
 {
 	struct rcar_du_encoder *renc;
+	struct drm_encoder *encoder;
 	unsigned int encoder_type;
 	int ret;
 
@@ -153,6 +157,7 @@ int rcar_du_encoder_init(struct rcar_du_device *rcdu,
 		return -ENOMEM;
 
 	renc->output = output;
+	encoder = rcar_encoder_to_drm_encoder(renc);
 
 	switch (output) {
 	case RCAR_DU_OUTPUT_LVDS0:
@@ -174,6 +179,9 @@ int rcar_du_encoder_init(struct rcar_du_device *rcdu,
 	case RCAR_DU_ENCODER_LVDS:
 		encoder_type = DRM_MODE_ENCODER_LVDS;
 		break;
+	case RCAR_DU_ENCODER_HDMI:
+		encoder_type = DRM_MODE_ENCODER_TMDS;
+		break;
 	case RCAR_DU_ENCODER_NONE:
 	default:
 		/* No external encoder, use the internal encoder type. */
@@ -181,21 +189,35 @@ int rcar_du_encoder_init(struct rcar_du_device *rcdu,
 		break;
 	}
 
-	ret = drm_encoder_init(rcdu->ddev, &renc->encoder, &encoder_funcs,
-			       encoder_type);
-	if (ret < 0)
-		return ret;
+	if (type == RCAR_DU_ENCODER_HDMI) {
+		if (renc->lvds) {
+			dev_err(rcdu->dev,
+				"Chaining LVDS and HDMI encoders not supported\n");
+			return -EINVAL;
+		}
 
-	drm_encoder_helper_add(&renc->encoder, &encoder_helper_funcs);
+		ret = rcar_du_hdmienc_init(rcdu, renc, enc_node);
+		if (ret < 0)
+			return ret;
+	} else {
+		ret = drm_encoder_init(rcdu->ddev, encoder, &encoder_funcs,
+				       encoder_type);
+		if (ret < 0)
+			return ret;
+
+		drm_encoder_helper_add(encoder, &encoder_helper_funcs);
+	}
 
 	switch (encoder_type) {
 	case DRM_MODE_ENCODER_LVDS:
-		return rcar_du_lvds_connector_init(rcdu, renc,
-						   &data->connector.lvds.panel);
+		return rcar_du_lvds_connector_init(rcdu, renc, con_node);
 
 	case DRM_MODE_ENCODER_DAC:
 		return rcar_du_vga_connector_init(rcdu, renc);
 
+	case DRM_MODE_ENCODER_TMDS:
+		return rcar_du_hdmi_connector_init(rcdu, renc);
+
 	default:
 		return -EINVAL;
 	}
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_encoder.h b/drivers/gpu/drm/rcar-du/rcar_du_encoder.h
index 0e5a65e45d0e..719b6f2a031c 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_encoder.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_encoder.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_encoder.h  --  R-Car Display Unit Encoder
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -14,21 +14,32 @@
 #ifndef __RCAR_DU_ENCODER_H__
 #define __RCAR_DU_ENCODER_H__
 
-#include <linux/platform_data/rcar-du.h>
-
 #include <drm/drm_crtc.h>
+#include <drm/drm_encoder_slave.h>
 
 struct rcar_du_device;
+struct rcar_du_hdmienc;
 struct rcar_du_lvdsenc;
 
+enum rcar_du_encoder_type {
+	RCAR_DU_ENCODER_UNUSED = 0,
+	RCAR_DU_ENCODER_NONE,
+	RCAR_DU_ENCODER_VGA,
+	RCAR_DU_ENCODER_LVDS,
+	RCAR_DU_ENCODER_HDMI,
+};
+
 struct rcar_du_encoder {
-	struct drm_encoder encoder;
+	struct drm_encoder_slave slave;
 	enum rcar_du_output output;
+	struct rcar_du_hdmienc *hdmi;
 	struct rcar_du_lvdsenc *lvds;
 };
 
 #define to_rcar_encoder(e) \
-	container_of(e, struct rcar_du_encoder, encoder)
+	container_of(e, struct rcar_du_encoder, slave.base)
+
+#define rcar_encoder_to_drm_encoder(e)	(&(e)->slave.base)
 
 struct rcar_du_connector {
 	struct drm_connector connector;
@@ -44,6 +55,7 @@ rcar_du_connector_best_encoder(struct drm_connector *connector);
 int rcar_du_encoder_init(struct rcar_du_device *rcdu,
 			 enum rcar_du_encoder_type type,
 			 enum rcar_du_output output,
-			 const struct rcar_du_encoder_data *data);
+			 struct device_node *enc_node,
+			 struct device_node *con_node);
 
 #endif /* __RCAR_DU_ENCODER_H__ */
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_group.c b/drivers/gpu/drm/rcar-du/rcar_du_group.c
index eb53cd97e8c6..4e7614b145db 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_group.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_group.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_group.c  --  R-Car Display Unit Channels Pair
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_group.h b/drivers/gpu/drm/rcar-du/rcar_du_group.h
index 5025930972ec..0c38cdcda4ca 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_group.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_group.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_group.c  --  R-Car Display Unit Planes and CRTCs Group
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_hdmicon.c b/drivers/gpu/drm/rcar-du/rcar_du_hdmicon.c
new file mode 100644
index 000000000000..4d7d4dd46d26
--- /dev/null
+++ b/drivers/gpu/drm/rcar-du/rcar_du_hdmicon.c
@@ -0,0 +1,121 @@
+/*
+ * R-Car Display Unit HDMI Connector
+ *
+ * Copyright (C) 2014 Renesas Electronics Corporation
+ *
+ * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_encoder_slave.h>
+
+#include "rcar_du_drv.h"
+#include "rcar_du_encoder.h"
+#include "rcar_du_hdmicon.h"
+#include "rcar_du_kms.h"
+
+#define to_slave_funcs(e)	(to_rcar_encoder(e)->slave.slave_funcs)
+
+static int rcar_du_hdmi_connector_get_modes(struct drm_connector *connector)
+{
+	struct rcar_du_connector *con = to_rcar_connector(connector);
+	struct drm_encoder *encoder = rcar_encoder_to_drm_encoder(con->encoder);
+	struct drm_encoder_slave_funcs *sfuncs = to_slave_funcs(encoder);
+
+	if (sfuncs->get_modes == NULL)
+		return 0;
+
+	return sfuncs->get_modes(encoder, connector);
+}
+
+static int rcar_du_hdmi_connector_mode_valid(struct drm_connector *connector,
+					     struct drm_display_mode *mode)
+{
+	struct rcar_du_connector *con = to_rcar_connector(connector);
+	struct drm_encoder *encoder = rcar_encoder_to_drm_encoder(con->encoder);
+	struct drm_encoder_slave_funcs *sfuncs = to_slave_funcs(encoder);
+
+	if (sfuncs->mode_valid == NULL)
+		return MODE_OK;
+
+	return sfuncs->mode_valid(encoder, mode);
+}
+
+static const struct drm_connector_helper_funcs connector_helper_funcs = {
+	.get_modes = rcar_du_hdmi_connector_get_modes,
+	.mode_valid = rcar_du_hdmi_connector_mode_valid,
+	.best_encoder = rcar_du_connector_best_encoder,
+};
+
+static void rcar_du_hdmi_connector_destroy(struct drm_connector *connector)
+{
+	drm_connector_unregister(connector);
+	drm_connector_cleanup(connector);
+}
+
+static enum drm_connector_status
+rcar_du_hdmi_connector_detect(struct drm_connector *connector, bool force)
+{
+	struct rcar_du_connector *con = to_rcar_connector(connector);
+	struct drm_encoder *encoder = rcar_encoder_to_drm_encoder(con->encoder);
+	struct drm_encoder_slave_funcs *sfuncs = to_slave_funcs(encoder);
+
+	if (sfuncs->detect == NULL)
+		return connector_status_unknown;
+
+	return sfuncs->detect(encoder, connector);
+}
+
+static const struct drm_connector_funcs connector_funcs = {
+	.dpms = drm_helper_connector_dpms,
+	.detect = rcar_du_hdmi_connector_detect,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.destroy = rcar_du_hdmi_connector_destroy,
+};
+
+int rcar_du_hdmi_connector_init(struct rcar_du_device *rcdu,
+				struct rcar_du_encoder *renc)
+{
+	struct drm_encoder *encoder = rcar_encoder_to_drm_encoder(renc);
+	struct rcar_du_connector *rcon;
+	struct drm_connector *connector;
+	int ret;
+
+	rcon = devm_kzalloc(rcdu->dev, sizeof(*rcon), GFP_KERNEL);
+	if (rcon == NULL)
+		return -ENOMEM;
+
+	connector = &rcon->connector;
+	connector->display_info.width_mm = 0;
+	connector->display_info.height_mm = 0;
+
+	ret = drm_connector_init(rcdu->ddev, connector, &connector_funcs,
+				 DRM_MODE_CONNECTOR_HDMIA);
+	if (ret < 0)
+		return ret;
+
+	drm_connector_helper_add(connector, &connector_helper_funcs);
+	ret = drm_connector_register(connector);
+	if (ret < 0)
+		return ret;
+
+	drm_helper_connector_dpms(connector, DRM_MODE_DPMS_OFF);
+	drm_object_property_set_value(&connector->base,
+		rcdu->ddev->mode_config.dpms_property, DRM_MODE_DPMS_OFF);
+
+	ret = drm_mode_connector_attach_encoder(connector, encoder);
+	if (ret < 0)
+		return ret;
+
+	connector->encoder = encoder;
+	rcon->encoder = renc;
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_hdmicon.h b/drivers/gpu/drm/rcar-du/rcar_du_hdmicon.h
new file mode 100644
index 000000000000..87daa949227f
--- /dev/null
+++ b/drivers/gpu/drm/rcar-du/rcar_du_hdmicon.h
@@ -0,0 +1,31 @@
+/*
+ * R-Car Display Unit HDMI Connector
+ *
+ * Copyright (C) 2014 Renesas Electronics Corporation
+ *
+ * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#ifndef __RCAR_DU_HDMICON_H__
+#define __RCAR_DU_HDMICON_H__
+
+struct rcar_du_device;
+struct rcar_du_encoder;
+
+#if IS_ENABLED(CONFIG_DRM_RCAR_HDMI)
+int rcar_du_hdmi_connector_init(struct rcar_du_device *rcdu,
+				struct rcar_du_encoder *renc);
+#else
+static inline int rcar_du_hdmi_connector_init(struct rcar_du_device *rcdu,
+					      struct rcar_du_encoder *renc)
+{
+	return -ENOSYS;
+}
+#endif
+
+#endif /* __RCAR_DU_HDMICON_H__ */
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_hdmienc.c b/drivers/gpu/drm/rcar-du/rcar_du_hdmienc.c
new file mode 100644
index 000000000000..359bc999a9c8
--- /dev/null
+++ b/drivers/gpu/drm/rcar-du/rcar_du_hdmienc.c
@@ -0,0 +1,151 @@
+/*
+ * R-Car Display Unit HDMI Encoder
+ *
+ * Copyright (C) 2014 Renesas Electronics Corporation
+ *
+ * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/slab.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_encoder_slave.h>
+
+#include "rcar_du_drv.h"
+#include "rcar_du_encoder.h"
+#include "rcar_du_hdmienc.h"
+
+struct rcar_du_hdmienc {
+	struct rcar_du_encoder *renc;
+	struct device *dev;
+	int dpms;
+};
+
+#define to_rcar_hdmienc(e)	(to_rcar_encoder(e)->hdmi)
+#define to_slave_funcs(e)	(to_rcar_encoder(e)->slave.slave_funcs)
+
+static void rcar_du_hdmienc_dpms(struct drm_encoder *encoder, int mode)
+{
+	struct rcar_du_hdmienc *hdmienc = to_rcar_hdmienc(encoder);
+	struct drm_encoder_slave_funcs *sfuncs = to_slave_funcs(encoder);
+
+	if (hdmienc->dpms == mode)
+		return;
+
+	if (sfuncs->dpms)
+		sfuncs->dpms(encoder, mode);
+
+	hdmienc->dpms = mode;
+}
+
+static bool rcar_du_hdmienc_mode_fixup(struct drm_encoder *encoder,
+				       const struct drm_display_mode *mode,
+				       struct drm_display_mode *adjusted_mode)
+{
+	struct drm_encoder_slave_funcs *sfuncs = to_slave_funcs(encoder);
+
+	if (sfuncs->mode_fixup == NULL)
+		return true;
+
+	return sfuncs->mode_fixup(encoder, mode, adjusted_mode);
+}
+
+static void rcar_du_hdmienc_mode_prepare(struct drm_encoder *encoder)
+{
+	rcar_du_hdmienc_dpms(encoder, DRM_MODE_DPMS_OFF);
+}
+
+static void rcar_du_hdmienc_mode_commit(struct drm_encoder *encoder)
+{
+	rcar_du_hdmienc_dpms(encoder, DRM_MODE_DPMS_ON);
+}
+
+static void rcar_du_hdmienc_mode_set(struct drm_encoder *encoder,
+				     struct drm_display_mode *mode,
+				     struct drm_display_mode *adjusted_mode)
+{
+	struct rcar_du_hdmienc *hdmienc = to_rcar_hdmienc(encoder);
+	struct drm_encoder_slave_funcs *sfuncs = to_slave_funcs(encoder);
+
+	if (sfuncs->mode_set)
+		sfuncs->mode_set(encoder, mode, adjusted_mode);
+
+	rcar_du_crtc_route_output(encoder->crtc, hdmienc->renc->output);
+}
+
+static const struct drm_encoder_helper_funcs encoder_helper_funcs = {
+	.dpms = rcar_du_hdmienc_dpms,
+	.mode_fixup = rcar_du_hdmienc_mode_fixup,
+	.prepare = rcar_du_hdmienc_mode_prepare,
+	.commit = rcar_du_hdmienc_mode_commit,
+	.mode_set = rcar_du_hdmienc_mode_set,
+};
+
+static void rcar_du_hdmienc_cleanup(struct drm_encoder *encoder)
+{
+	struct rcar_du_hdmienc *hdmienc = to_rcar_hdmienc(encoder);
+
+	rcar_du_hdmienc_dpms(encoder, DRM_MODE_DPMS_OFF);
+
+	drm_encoder_cleanup(encoder);
+	put_device(hdmienc->dev);
+}
+
+static const struct drm_encoder_funcs encoder_funcs = {
+	.destroy = rcar_du_hdmienc_cleanup,
+};
+
+int rcar_du_hdmienc_init(struct rcar_du_device *rcdu,
+			 struct rcar_du_encoder *renc, struct device_node *np)
+{
+	struct drm_encoder *encoder = rcar_encoder_to_drm_encoder(renc);
+	struct drm_i2c_encoder_driver *driver;
+	struct i2c_client *i2c_slave;
+	struct rcar_du_hdmienc *hdmienc;
+	int ret;
+
+	hdmienc = devm_kzalloc(rcdu->dev, sizeof(*hdmienc), GFP_KERNEL);
+	if (hdmienc == NULL)
+		return -ENOMEM;
+
+	/* Locate the slave I2C device and driver. */
+	i2c_slave = of_find_i2c_device_by_node(np);
+	if (!i2c_slave || !i2c_get_clientdata(i2c_slave))
+		return -EPROBE_DEFER;
+
+	hdmienc->dev = &i2c_slave->dev;
+
+	if (hdmienc->dev->driver == NULL) {
+		ret = -EPROBE_DEFER;
+		goto error;
+	}
+
+	/* Initialize the slave encoder. */
+	driver = to_drm_i2c_encoder_driver(to_i2c_driver(hdmienc->dev->driver));
+	ret = driver->encoder_init(i2c_slave, rcdu->ddev, &renc->slave);
+	if (ret < 0)
+		goto error;
+
+	ret = drm_encoder_init(rcdu->ddev, encoder, &encoder_funcs,
+			       DRM_MODE_ENCODER_TMDS);
+	if (ret < 0)
+		goto error;
+
+	drm_encoder_helper_add(encoder, &encoder_helper_funcs);
+
+	renc->hdmi = hdmienc;
+	hdmienc->renc = renc;
+
+	return 0;
+
+error:
+	put_device(hdmienc->dev);
+	return ret;
+}
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_hdmienc.h b/drivers/gpu/drm/rcar-du/rcar_du_hdmienc.h
new file mode 100644
index 000000000000..2ff0128ac8e1
--- /dev/null
+++ b/drivers/gpu/drm/rcar-du/rcar_du_hdmienc.h
@@ -0,0 +1,35 @@
+/*
+ * R-Car Display Unit HDMI Encoder
+ *
+ * Copyright (C) 2014 Renesas Electronics Corporation
+ *
+ * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#ifndef __RCAR_DU_HDMIENC_H__
+#define __RCAR_DU_HDMIENC_H__
+
+#include <linux/module.h>
+
+struct device_node;
+struct rcar_du_device;
+struct rcar_du_encoder;
+
+#if IS_ENABLED(CONFIG_DRM_RCAR_HDMI)
+int rcar_du_hdmienc_init(struct rcar_du_device *rcdu,
+			 struct rcar_du_encoder *renc, struct device_node *np);
+#else
+static inline int rcar_du_hdmienc_init(struct rcar_du_device *rcdu,
+				       struct rcar_du_encoder *renc,
+				       struct device_node *np)
+{
+	return -ENOSYS;
+}
+#endif
+
+#endif /* __RCAR_DU_HDMIENC_H__ */
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_kms.c b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
index 76026104d000..0c5ee616b5a3 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_kms.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_kms.c  --  R-Car Display Unit Mode Setting
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -17,6 +17,8 @@
 #include <drm/drm_fb_cma_helper.h>
 #include <drm/drm_gem_cma_helper.h>
 
+#include <linux/of_graph.h>
+
 #include "rcar_du_crtc.h"
 #include "rcar_du_drv.h"
 #include "rcar_du_encoder.h"
@@ -124,9 +126,9 @@ int rcar_du_dumb_create(struct drm_file *file, struct drm_device *dev,
 	else
 		align = 16 * args->bpp / 8;
 
-	args->pitch = roundup(max(args->pitch, min_pitch), align);
+	args->pitch = roundup(min_pitch, align);
 
-	return drm_gem_cma_dumb_create(file, dev, args);
+	return drm_gem_cma_dumb_create_internal(file, dev, args);
 }
 
 static struct drm_framebuffer *
@@ -188,6 +190,172 @@ static const struct drm_mode_config_funcs rcar_du_mode_config_funcs = {
 	.output_poll_changed = rcar_du_output_poll_changed,
 };
 
+static int rcar_du_encoders_init_one(struct rcar_du_device *rcdu,
+				     enum rcar_du_output output,
+				     struct of_endpoint *ep)
+{
+	static const struct {
+		const char *compatible;
+		enum rcar_du_encoder_type type;
+	} encoders[] = {
+		{ "adi,adv7123", RCAR_DU_ENCODER_VGA },
+		{ "adi,adv7511w", RCAR_DU_ENCODER_HDMI },
+		{ "thine,thc63lvdm83d", RCAR_DU_ENCODER_LVDS },
+	};
+
+	enum rcar_du_encoder_type enc_type = RCAR_DU_ENCODER_NONE;
+	struct device_node *connector = NULL;
+	struct device_node *encoder = NULL;
+	struct device_node *prev = NULL;
+	struct device_node *entity_ep_node;
+	struct device_node *entity;
+	int ret;
+
+	/*
+	 * Locate the connected entity and infer its type from the number of
+	 * endpoints.
+	 */
+	entity = of_graph_get_remote_port_parent(ep->local_node);
+	if (!entity) {
+		dev_dbg(rcdu->dev, "unconnected endpoint %s, skipping\n",
+			ep->local_node->full_name);
+		return 0;
+	}
+
+	entity_ep_node = of_parse_phandle(ep->local_node, "remote-endpoint", 0);
+
+	while (1) {
+		struct device_node *ep_node;
+
+		ep_node = of_graph_get_next_endpoint(entity, prev);
+		of_node_put(prev);
+		prev = ep_node;
+
+		if (!ep_node)
+			break;
+
+		if (ep_node == entity_ep_node)
+			continue;
+
+		/*
+		 * We've found one endpoint other than the input, this must
+		 * be an encoder. Locate the connector.
+		 */
+		encoder = entity;
+		connector = of_graph_get_remote_port_parent(ep_node);
+		of_node_put(ep_node);
+
+		if (!connector) {
+			dev_warn(rcdu->dev,
+				 "no connector for encoder %s, skipping\n",
+				 encoder->full_name);
+			of_node_put(entity_ep_node);
+			of_node_put(encoder);
+			return 0;
+		}
+
+		break;
+	}
+
+	of_node_put(entity_ep_node);
+
+	if (encoder) {
+		/*
+		 * If an encoder has been found, get its type based on its
+		 * compatible string.
+		 */
+		unsigned int i;
+
+		for (i = 0; i < ARRAY_SIZE(encoders); ++i) {
+			if (of_device_is_compatible(encoder,
+						    encoders[i].compatible)) {
+				enc_type = encoders[i].type;
+				break;
+			}
+		}
+
+		if (i == ARRAY_SIZE(encoders)) {
+			dev_warn(rcdu->dev,
+				 "unknown encoder type for %s, skipping\n",
+				 encoder->full_name);
+			of_node_put(encoder);
+			of_node_put(connector);
+			return 0;
+		}
+	} else {
+		/*
+		 * If no encoder has been found the entity must be the
+		 * connector.
+		 */
+		connector = entity;
+	}
+
+	ret = rcar_du_encoder_init(rcdu, enc_type, output, encoder, connector);
+	of_node_put(encoder);
+	of_node_put(connector);
+
+	return ret < 0 ? ret : 1;
+}
+
+static int rcar_du_encoders_init(struct rcar_du_device *rcdu)
+{
+	struct device_node *np = rcdu->dev->of_node;
+	struct device_node *prev = NULL;
+	unsigned int num_encoders = 0;
+
+	/*
+	 * Iterate over the endpoints and create one encoder for each output
+	 * pipeline.
+	 */
+	while (1) {
+		struct device_node *ep_node;
+		enum rcar_du_output output;
+		struct of_endpoint ep;
+		unsigned int i;
+		int ret;
+
+		ep_node = of_graph_get_next_endpoint(np, prev);
+		of_node_put(prev);
+		prev = ep_node;
+
+		if (ep_node == NULL)
+			break;
+
+		ret = of_graph_parse_endpoint(ep_node, &ep);
+		if (ret < 0) {
+			of_node_put(ep_node);
+			return ret;
+		}
+
+		/* Find the output route corresponding to the port number. */
+		for (i = 0; i < RCAR_DU_OUTPUT_MAX; ++i) {
+			if (rcdu->info->routes[i].possible_crtcs &&
+			    rcdu->info->routes[i].port == ep.port) {
+				output = i;
+				break;
+			}
+		}
+
+		if (i == RCAR_DU_OUTPUT_MAX) {
+			dev_warn(rcdu->dev,
+				 "port %u references unexisting output, skipping\n",
+				 ep.port);
+			continue;
+		}
+
+		/* Process the output pipeline. */
+		ret = rcar_du_encoders_init_one(rcdu, output, &ep);
+		if (ret < 0) {
+			of_node_put(ep_node);
+			return ret;
+		}
+
+		num_encoders += ret;
+	}
+
+	return num_encoders;
+}
+
 int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 {
 	static const unsigned int mmio_offsets[] = {
@@ -197,6 +365,7 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 	struct drm_device *dev = rcdu->ddev;
 	struct drm_encoder *encoder;
 	struct drm_fbdev_cma *fbdev;
+	unsigned int num_encoders;
 	unsigned int num_groups;
 	unsigned int i;
 	int ret;
@@ -240,28 +409,11 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 	if (ret < 0)
 		return ret;
 
-	for (i = 0; i < rcdu->pdata->num_encoders; ++i) {
-		const struct rcar_du_encoder_data *pdata =
-			&rcdu->pdata->encoders[i];
-		const struct rcar_du_output_routing *route =
-			&rcdu->info->routes[pdata->output];
-
-		if (pdata->type == RCAR_DU_ENCODER_UNUSED)
-			continue;
-
-		if (pdata->output >= RCAR_DU_OUTPUT_MAX ||
-		    route->possible_crtcs == 0) {
-			dev_warn(rcdu->dev,
-				 "encoder %u references unexisting output %u, skipping\n",
-				 i, pdata->output);
-			continue;
-		}
+	ret = rcar_du_encoders_init(rcdu);
+	if (ret < 0)
+		return ret;
 
-		ret = rcar_du_encoder_init(rcdu, pdata->type, pdata->output,
-					   pdata);
-		if (ret < 0)
-			return ret;
-	}
+	num_encoders = ret;
 
 	/* Set the possible CRTCs and possible clones. There's always at least
 	 * one way for all encoders to clone each other, set all bits in the
@@ -273,7 +425,7 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 			&rcdu->info->routes[renc->output];
 
 		encoder->possible_crtcs = route->possible_crtcs;
-		encoder->possible_clones = (1 << rcdu->pdata->num_encoders) - 1;
+		encoder->possible_clones = (1 << num_encoders) - 1;
 	}
 
 	/* Now that the CRTCs have been initialized register the planes. */
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_kms.h b/drivers/gpu/drm/rcar-du/rcar_du_kms.h
index 5750e6af5655..07951d5fe38b 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_kms.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_kms.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_kms.h  --  R-Car Display Unit Mode Setting
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.c b/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.c
index 21426bd234eb..6d9811c052c4 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_lvdscon.c  --  R-Car Display Unit LVDS Connector
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -15,6 +15,10 @@
 #include <drm/drm_crtc.h>
 #include <drm/drm_crtc_helper.h>
 
+#include <video/display_timing.h>
+#include <video/of_display_timing.h>
+#include <video/videomode.h>
+
 #include "rcar_du_drv.h"
 #include "rcar_du_encoder.h"
 #include "rcar_du_kms.h"
@@ -23,7 +27,11 @@
 struct rcar_du_lvds_connector {
 	struct rcar_du_connector connector;
 
-	const struct rcar_du_panel_data *panel;
+	struct {
+		unsigned int width_mm;		/* Panel width in mm */
+		unsigned int height_mm;		/* Panel height in mm */
+		struct videomode mode;
+	} panel;
 };
 
 #define to_rcar_lvds_connector(c) \
@@ -40,18 +48,9 @@ static int rcar_du_lvds_connector_get_modes(struct drm_connector *connector)
 		return 0;
 
 	mode->type = DRM_MODE_TYPE_PREFERRED | DRM_MODE_TYPE_DRIVER;
-	mode->clock = lvdscon->panel->mode.clock;
-	mode->hdisplay = lvdscon->panel->mode.hdisplay;
-	mode->hsync_start = lvdscon->panel->mode.hsync_start;
-	mode->hsync_end = lvdscon->panel->mode.hsync_end;
-	mode->htotal = lvdscon->panel->mode.htotal;
-	mode->vdisplay = lvdscon->panel->mode.vdisplay;
-	mode->vsync_start = lvdscon->panel->mode.vsync_start;
-	mode->vsync_end = lvdscon->panel->mode.vsync_end;
-	mode->vtotal = lvdscon->panel->mode.vtotal;
-	mode->flags = lvdscon->panel->mode.flags;
-
-	drm_mode_set_name(mode);
+
+	drm_display_mode_from_videomode(&lvdscon->panel.mode, mode);
+
 	drm_mode_probed_add(connector, mode);
 
 	return 1;
@@ -83,21 +82,30 @@ static const struct drm_connector_funcs connector_funcs = {
 
 int rcar_du_lvds_connector_init(struct rcar_du_device *rcdu,
 				struct rcar_du_encoder *renc,
-				const struct rcar_du_panel_data *panel)
+				/* TODO const */ struct device_node *np)
 {
+	struct drm_encoder *encoder = rcar_encoder_to_drm_encoder(renc);
 	struct rcar_du_lvds_connector *lvdscon;
 	struct drm_connector *connector;
+	struct display_timing timing;
 	int ret;
 
 	lvdscon = devm_kzalloc(rcdu->dev, sizeof(*lvdscon), GFP_KERNEL);
 	if (lvdscon == NULL)
 		return -ENOMEM;
 
-	lvdscon->panel = panel;
+	ret = of_get_display_timing(np, "panel-timing", &timing);
+	if (ret < 0)
+		return ret;
+
+	videomode_from_timing(&timing, &lvdscon->panel.mode);
+
+	of_property_read_u32(np, "width-mm", &lvdscon->panel.width_mm);
+	of_property_read_u32(np, "height-mm", &lvdscon->panel.height_mm);
 
 	connector = &lvdscon->connector.connector;
-	connector->display_info.width_mm = panel->width_mm;
-	connector->display_info.height_mm = panel->height_mm;
+	connector->display_info.width_mm = lvdscon->panel.width_mm;
+	connector->display_info.height_mm = lvdscon->panel.height_mm;
 
 	ret = drm_connector_init(rcdu->ddev, connector, &connector_funcs,
 				 DRM_MODE_CONNECTOR_LVDS);
@@ -113,11 +121,11 @@ int rcar_du_lvds_connector_init(struct rcar_du_device *rcdu,
 	drm_object_property_set_value(&connector->base,
 		rcdu->ddev->mode_config.dpms_property, DRM_MODE_DPMS_OFF);
 
-	ret = drm_mode_connector_attach_encoder(connector, &renc->encoder);
+	ret = drm_mode_connector_attach_encoder(connector, encoder);
 	if (ret < 0)
 		return ret;
 
-	connector->encoder = &renc->encoder;
+	connector->encoder = encoder;
 	lvdscon->connector.encoder = renc;
 
 	return 0;
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.h b/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.h
index bff8683699ca..d4881ee0be7e 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_lvdscon.h  --  R-Car Display Unit LVDS Connector
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -16,10 +16,9 @@
 
 struct rcar_du_device;
 struct rcar_du_encoder;
-struct rcar_du_panel_data;
 
 int rcar_du_lvds_connector_init(struct rcar_du_device *rcdu,
 				struct rcar_du_encoder *renc,
-				const struct rcar_du_panel_data *panel);
+				struct device_node *np);
 
 #endif /* __RCAR_DU_LVDSCON_H__ */
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c
index df30a075d793..7cfb48ce1791 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_lvdsenc.c  --  R-Car Display Unit LVDS Encoder
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.h b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.h
index 7051c6de19ae..f65aabda0796 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_lvdsenc.h  --  R-Car Display Unit LVDS Encoder
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -16,7 +16,6 @@
 
 #include <linux/io.h>
 #include <linux/module.h>
-#include <linux/platform_data/rcar-du.h>
 
 struct rcar_drm_crtc;
 struct rcar_du_lvdsenc;
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_plane.c b/drivers/gpu/drm/rcar-du/rcar_du_plane.c
index 3fb69d9ae61b..72a7cb47bd9f 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_plane.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_plane.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_plane.c  --  R-Car Display Unit Planes
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_plane.h b/drivers/gpu/drm/rcar-du/rcar_du_plane.h
index f94f9ce84998..3021288b1a89 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_plane.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_plane.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_plane.h  --  R-Car Display Unit Planes
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_vgacon.c b/drivers/gpu/drm/rcar-du/rcar_du_vgacon.c
index 8af3944d31b9..752747a5e920 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_vgacon.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_vgacon.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_vgacon.c  --  R-Car Display Unit VGA Connector
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -52,6 +52,7 @@ static const struct drm_connector_funcs connector_funcs = {
 int rcar_du_vga_connector_init(struct rcar_du_device *rcdu,
 			       struct rcar_du_encoder *renc)
 {
+	struct drm_encoder *encoder = rcar_encoder_to_drm_encoder(renc);
 	struct rcar_du_connector *rcon;
 	struct drm_connector *connector;
 	int ret;
@@ -78,11 +79,11 @@ int rcar_du_vga_connector_init(struct rcar_du_device *rcdu,
 	drm_object_property_set_value(&connector->base,
 		rcdu->ddev->mode_config.dpms_property, DRM_MODE_DPMS_OFF);
 
-	ret = drm_mode_connector_attach_encoder(connector, &renc->encoder);
+	ret = drm_mode_connector_attach_encoder(connector, encoder);
 	if (ret < 0)
 		return ret;
 
-	connector->encoder = &renc->encoder;
+	connector->encoder = encoder;
 	rcon->encoder = renc;
 
 	return 0;
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_vgacon.h b/drivers/gpu/drm/rcar-du/rcar_du_vgacon.h
index b12b0cf7f117..112f50316e01 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_vgacon.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_vgacon.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_vgacon.h  --  R-Car Display Unit VGA Connector
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/rockchip/Kconfig b/drivers/gpu/drm/rockchip/Kconfig
new file mode 100644
index 000000000000..ca9f085efa92
--- /dev/null
+++ b/drivers/gpu/drm/rockchip/Kconfig
@@ -0,0 +1,17 @@
+config DRM_ROCKCHIP
+	tristate "DRM Support for Rockchip"
+	depends on DRM && ROCKCHIP_IOMMU
+	select DRM_KMS_HELPER
+	select DRM_KMS_FB_HELPER
+	select DRM_PANEL
+	select FB_CFB_FILLRECT
+	select FB_CFB_COPYAREA
+	select FB_CFB_IMAGEBLIT
+	select VT_HW_CONSOLE_BINDING if FRAMEBUFFER_CONSOLE
+	select VIDEOMODE_HELPERS
+	help
+	  Choose this option if you have a Rockchip soc chipset.
+	  This driver provides kernel mode setting and buffer
+	  management to userspace. This driver does not provide
+	  2D or 3D acceleration; acceleration is performed by other
+	  IP found on the SoC.
diff --git a/drivers/gpu/drm/rockchip/Makefile b/drivers/gpu/drm/rockchip/Makefile
new file mode 100644
index 000000000000..2cb0672f57ed
--- /dev/null
+++ b/drivers/gpu/drm/rockchip/Makefile
@@ -0,0 +1,8 @@
+#
+# Makefile for the drm device driver.  This driver provides support for the
+# Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.
+
+rockchipdrm-y := rockchip_drm_drv.o rockchip_drm_fb.o rockchip_drm_fbdev.o \
+		rockchip_drm_gem.o
+
+obj-$(CONFIG_DRM_ROCKCHIP) += rockchipdrm.o rockchip_drm_vop.o
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_drv.c b/drivers/gpu/drm/rockchip/rockchip_drm_drv.c
new file mode 100644
index 000000000000..a798c7c71f91
--- /dev/null
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_drv.c
@@ -0,0 +1,551 @@
+/*
+ * Copyright (C) Fuzhou Rockchip Electronics Co.Ltd
+ * Author:Mark Yao <mark.yao@rock-chips.com>
+ *
+ * based on exynos_drm_drv.c
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <asm/dma-iommu.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_fb_helper.h>
+#include <linux/dma-mapping.h>
+#include <linux/pm_runtime.h>
+#include <linux/of_graph.h>
+#include <linux/component.h>
+
+#include "rockchip_drm_drv.h"
+#include "rockchip_drm_fb.h"
+#include "rockchip_drm_fbdev.h"
+#include "rockchip_drm_gem.h"
+
+#define DRIVER_NAME	"rockchip"
+#define DRIVER_DESC	"RockChip Soc DRM"
+#define DRIVER_DATE	"20140818"
+#define DRIVER_MAJOR	1
+#define DRIVER_MINOR	0
+
+/*
+ * Attach a (component) device to the shared drm dma mapping from master drm
+ * device.  This is used by the VOPs to map GEM buffers to a common DMA
+ * mapping.
+ */
+int rockchip_drm_dma_attach_device(struct drm_device *drm_dev,
+				   struct device *dev)
+{
+	struct dma_iommu_mapping *mapping = drm_dev->dev->archdata.mapping;
+	int ret;
+
+	ret = dma_set_coherent_mask(dev, DMA_BIT_MASK(32));
+	if (ret)
+		return ret;
+
+	dma_set_max_seg_size(dev, DMA_BIT_MASK(32));
+
+	return arm_iommu_attach_device(dev, mapping);
+}
+EXPORT_SYMBOL_GPL(rockchip_drm_dma_attach_device);
+
+void rockchip_drm_dma_detach_device(struct drm_device *drm_dev,
+				    struct device *dev)
+{
+	arm_iommu_detach_device(dev);
+}
+EXPORT_SYMBOL_GPL(rockchip_drm_dma_detach_device);
+
+int rockchip_register_crtc_funcs(struct drm_device *dev,
+				 const struct rockchip_crtc_funcs *crtc_funcs,
+				 int pipe)
+{
+	struct rockchip_drm_private *priv = dev->dev_private;
+
+	if (pipe > ROCKCHIP_MAX_CRTC)
+		return -EINVAL;
+
+	priv->crtc_funcs[pipe] = crtc_funcs;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(rockchip_register_crtc_funcs);
+
+void rockchip_unregister_crtc_funcs(struct drm_device *dev, int pipe)
+{
+	struct rockchip_drm_private *priv = dev->dev_private;
+
+	if (pipe > ROCKCHIP_MAX_CRTC)
+		return;
+
+	priv->crtc_funcs[pipe] = NULL;
+}
+EXPORT_SYMBOL_GPL(rockchip_unregister_crtc_funcs);
+
+static struct drm_crtc *rockchip_crtc_from_pipe(struct drm_device *drm,
+						int pipe)
+{
+	struct drm_crtc *crtc;
+	int i = 0;
+
+	list_for_each_entry(crtc, &drm->mode_config.crtc_list, head)
+		if (i++ == pipe)
+			return crtc;
+
+	return NULL;
+}
+
+static int rockchip_drm_crtc_enable_vblank(struct drm_device *dev, int pipe)
+{
+	struct rockchip_drm_private *priv = dev->dev_private;
+	struct drm_crtc *crtc = rockchip_crtc_from_pipe(dev, pipe);
+
+	if (crtc && priv->crtc_funcs[pipe] &&
+	    priv->crtc_funcs[pipe]->enable_vblank)
+		return priv->crtc_funcs[pipe]->enable_vblank(crtc);
+
+	return 0;
+}
+
+static void rockchip_drm_crtc_disable_vblank(struct drm_device *dev, int pipe)
+{
+	struct rockchip_drm_private *priv = dev->dev_private;
+	struct drm_crtc *crtc = rockchip_crtc_from_pipe(dev, pipe);
+
+	if (crtc && priv->crtc_funcs[pipe] &&
+	    priv->crtc_funcs[pipe]->enable_vblank)
+		priv->crtc_funcs[pipe]->disable_vblank(crtc);
+}
+
+static int rockchip_drm_load(struct drm_device *drm_dev, unsigned long flags)
+{
+	struct rockchip_drm_private *private;
+	struct dma_iommu_mapping *mapping;
+	struct device *dev = drm_dev->dev;
+	int ret;
+
+	private = devm_kzalloc(drm_dev->dev, sizeof(*private), GFP_KERNEL);
+	if (!private)
+		return -ENOMEM;
+
+	drm_dev->dev_private = private;
+
+	drm_mode_config_init(drm_dev);
+
+	rockchip_drm_mode_config_init(drm_dev);
+
+	dev->dma_parms = devm_kzalloc(dev, sizeof(*dev->dma_parms),
+				      GFP_KERNEL);
+	if (!dev->dma_parms) {
+		ret = -ENOMEM;
+		goto err_config_cleanup;
+	}
+
+	/* TODO(djkurtz): fetch the mapping start/size from somewhere */
+	mapping = arm_iommu_create_mapping(&platform_bus_type, 0x00000000,
+					   SZ_2G);
+	if (IS_ERR(mapping)) {
+		ret = PTR_ERR(mapping);
+		goto err_config_cleanup;
+	}
+
+	ret = dma_set_mask_and_coherent(dev, DMA_BIT_MASK(32));
+	if (ret)
+		goto err_release_mapping;
+
+	dma_set_max_seg_size(dev, DMA_BIT_MASK(32));
+
+	ret = arm_iommu_attach_device(dev, mapping);
+	if (ret)
+		goto err_release_mapping;
+
+	/* Try to bind all sub drivers. */
+	ret = component_bind_all(dev, drm_dev);
+	if (ret)
+		goto err_detach_device;
+
+	/* init kms poll for handling hpd */
+	drm_kms_helper_poll_init(drm_dev);
+
+	/*
+	 * enable drm irq mode.
+	 * - with irq_enabled = true, we can use the vblank feature.
+	 */
+	drm_dev->irq_enabled = true;
+
+	ret = drm_vblank_init(drm_dev, ROCKCHIP_MAX_CRTC);
+	if (ret)
+		goto err_kms_helper_poll_fini;
+
+	/*
+	 * with vblank_disable_allowed = true, vblank interrupt will be disabled
+	 * by drm timer once a current process gives up ownership of
+	 * vblank event.(after drm_vblank_put function is called)
+	 */
+	drm_dev->vblank_disable_allowed = true;
+
+	ret = rockchip_drm_fbdev_init(drm_dev);
+	if (ret)
+		goto err_vblank_cleanup;
+
+	return 0;
+err_vblank_cleanup:
+	drm_vblank_cleanup(drm_dev);
+err_kms_helper_poll_fini:
+	drm_kms_helper_poll_fini(drm_dev);
+	component_unbind_all(dev, drm_dev);
+err_detach_device:
+	arm_iommu_detach_device(dev);
+err_release_mapping:
+	arm_iommu_release_mapping(dev->archdata.mapping);
+err_config_cleanup:
+	drm_mode_config_cleanup(drm_dev);
+	drm_dev->dev_private = NULL;
+	return ret;
+}
+
+static int rockchip_drm_unload(struct drm_device *drm_dev)
+{
+	struct device *dev = drm_dev->dev;
+
+	rockchip_drm_fbdev_fini(drm_dev);
+	drm_vblank_cleanup(drm_dev);
+	drm_kms_helper_poll_fini(drm_dev);
+	component_unbind_all(dev, drm_dev);
+	arm_iommu_detach_device(dev);
+	arm_iommu_release_mapping(dev->archdata.mapping);
+	drm_mode_config_cleanup(drm_dev);
+	drm_dev->dev_private = NULL;
+
+	return 0;
+}
+
+void rockchip_drm_lastclose(struct drm_device *dev)
+{
+	struct rockchip_drm_private *priv = dev->dev_private;
+
+	drm_fb_helper_restore_fbdev_mode_unlocked(&priv->fbdev_helper);
+}
+
+static const struct file_operations rockchip_drm_driver_fops = {
+	.owner = THIS_MODULE,
+	.open = drm_open,
+	.mmap = rockchip_gem_mmap,
+	.poll = drm_poll,
+	.read = drm_read,
+	.unlocked_ioctl = drm_ioctl,
+#ifdef CONFIG_COMPAT
+	.compat_ioctl = drm_compat_ioctl,
+#endif
+	.release = drm_release,
+};
+
+const struct vm_operations_struct rockchip_drm_vm_ops = {
+	.open = drm_gem_vm_open,
+	.close = drm_gem_vm_close,
+};
+
+static struct drm_driver rockchip_drm_driver = {
+	.driver_features	= DRIVER_MODESET | DRIVER_GEM | DRIVER_PRIME,
+	.load			= rockchip_drm_load,
+	.unload			= rockchip_drm_unload,
+	.lastclose		= rockchip_drm_lastclose,
+	.get_vblank_counter	= drm_vblank_count,
+	.enable_vblank		= rockchip_drm_crtc_enable_vblank,
+	.disable_vblank		= rockchip_drm_crtc_disable_vblank,
+	.gem_vm_ops		= &rockchip_drm_vm_ops,
+	.gem_free_object	= rockchip_gem_free_object,
+	.dumb_create		= rockchip_gem_dumb_create,
+	.dumb_map_offset	= rockchip_gem_dumb_map_offset,
+	.dumb_destroy		= drm_gem_dumb_destroy,
+	.prime_handle_to_fd	= drm_gem_prime_handle_to_fd,
+	.prime_fd_to_handle	= drm_gem_prime_fd_to_handle,
+	.gem_prime_import	= drm_gem_prime_import,
+	.gem_prime_export	= drm_gem_prime_export,
+	.gem_prime_get_sg_table	= rockchip_gem_prime_get_sg_table,
+	.gem_prime_vmap		= rockchip_gem_prime_vmap,
+	.gem_prime_vunmap	= rockchip_gem_prime_vunmap,
+	.gem_prime_mmap		= rockchip_gem_mmap_buf,
+	.fops			= &rockchip_drm_driver_fops,
+	.name	= DRIVER_NAME,
+	.desc	= DRIVER_DESC,
+	.date	= DRIVER_DATE,
+	.major	= DRIVER_MAJOR,
+	.minor	= DRIVER_MINOR,
+};
+
+#ifdef CONFIG_PM_SLEEP
+static int rockchip_drm_sys_suspend(struct device *dev)
+{
+	struct drm_device *drm = dev_get_drvdata(dev);
+	struct drm_connector *connector;
+
+	if (!drm)
+		return 0;
+
+	drm_modeset_lock_all(drm);
+	list_for_each_entry(connector, &drm->mode_config.connector_list, head) {
+		int old_dpms = connector->dpms;
+
+		if (connector->funcs->dpms)
+			connector->funcs->dpms(connector, DRM_MODE_DPMS_OFF);
+
+		/* Set the old mode back to the connector for resume */
+		connector->dpms = old_dpms;
+	}
+	drm_modeset_unlock_all(drm);
+
+	return 0;
+}
+
+static int rockchip_drm_sys_resume(struct device *dev)
+{
+	struct drm_device *drm = dev_get_drvdata(dev);
+	struct drm_connector *connector;
+	enum drm_connector_status status;
+	bool changed = false;
+
+	if (!drm)
+		return 0;
+
+	drm_modeset_lock_all(drm);
+	list_for_each_entry(connector, &drm->mode_config.connector_list, head) {
+		int desired_mode = connector->dpms;
+
+		/*
+		 * at suspend time, we save dpms to connector->dpms,
+		 * restore the old_dpms, and at current time, the connector
+		 * dpms status must be DRM_MODE_DPMS_OFF.
+		 */
+		connector->dpms = DRM_MODE_DPMS_OFF;
+
+		/*
+		 * If the connector has been disconnected during suspend,
+		 * disconnect it from the encoder and leave it off. We'll notify
+		 * userspace at the end.
+		 */
+		if (desired_mode == DRM_MODE_DPMS_ON) {
+			status = connector->funcs->detect(connector, true);
+			if (status == connector_status_disconnected) {
+				connector->encoder = NULL;
+				connector->status = status;
+				changed = true;
+				continue;
+			}
+		}
+		if (connector->funcs->dpms)
+			connector->funcs->dpms(connector, desired_mode);
+	}
+	drm_modeset_unlock_all(drm);
+
+	drm_helper_resume_force_mode(drm);
+
+	if (changed)
+		drm_kms_helper_hotplug_event(drm);
+
+	return 0;
+}
+#endif
+
+static const struct dev_pm_ops rockchip_drm_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(rockchip_drm_sys_suspend,
+				rockchip_drm_sys_resume)
+};
+
+/*
+ * @node: device tree node containing encoder input ports
+ * @encoder: drm_encoder
+ */
+int rockchip_drm_encoder_get_mux_id(struct device_node *node,
+				    struct drm_encoder *encoder)
+{
+	struct device_node *ep = NULL;
+	struct drm_crtc *crtc = encoder->crtc;
+	struct of_endpoint endpoint;
+	struct device_node *port;
+	int ret;
+
+	if (!node || !crtc)
+		return -EINVAL;
+
+	do {
+		ep = of_graph_get_next_endpoint(node, ep);
+		if (!ep)
+			break;
+
+		port = of_graph_get_remote_port(ep);
+		of_node_put(port);
+		if (port == crtc->port) {
+			ret = of_graph_parse_endpoint(ep, &endpoint);
+			return ret ?: endpoint.id;
+		}
+	} while (ep);
+
+	return -EINVAL;
+}
+
+static int compare_of(struct device *dev, void *data)
+{
+	struct device_node *np = data;
+
+	return dev->of_node == np;
+}
+
+static void rockchip_add_endpoints(struct device *dev,
+				   struct component_match **match,
+				   struct device_node *port)
+{
+	struct device_node *ep, *remote;
+
+	for_each_child_of_node(port, ep) {
+		remote = of_graph_get_remote_port_parent(ep);
+		if (!remote || !of_device_is_available(remote)) {
+			of_node_put(remote);
+			continue;
+		} else if (!of_device_is_available(remote->parent)) {
+			dev_warn(dev, "parent device of %s is not available\n",
+				 remote->full_name);
+			of_node_put(remote);
+			continue;
+		}
+
+		component_match_add(dev, match, compare_of, remote);
+		of_node_put(remote);
+	}
+}
+
+static int rockchip_drm_bind(struct device *dev)
+{
+	struct drm_device *drm;
+	int ret;
+
+	drm = drm_dev_alloc(&rockchip_drm_driver, dev);
+	if (!drm)
+		return -ENOMEM;
+
+	ret = drm_dev_set_unique(drm, "%s", dev_name(dev));
+	if (ret)
+		goto err_free;
+
+	ret = drm_dev_register(drm, 0);
+	if (ret)
+		goto err_free;
+
+	dev_set_drvdata(dev, drm);
+
+	return 0;
+
+err_free:
+	drm_dev_unref(drm);
+	return ret;
+}
+
+static void rockchip_drm_unbind(struct device *dev)
+{
+	struct drm_device *drm = dev_get_drvdata(dev);
+
+	drm_dev_unregister(drm);
+	drm_dev_unref(drm);
+	dev_set_drvdata(dev, NULL);
+}
+
+static const struct component_master_ops rockchip_drm_ops = {
+	.bind = rockchip_drm_bind,
+	.unbind = rockchip_drm_unbind,
+};
+
+static int rockchip_drm_platform_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct component_match *match = NULL;
+	struct device_node *np = dev->of_node;
+	struct device_node *port;
+	int i;
+
+	if (!np)
+		return -ENODEV;
+	/*
+	 * Bind the crtc ports first, so that
+	 * drm_of_find_possible_crtcs called from encoder .bind callbacks
+	 * works as expected.
+	 */
+	for (i = 0;; i++) {
+		port = of_parse_phandle(np, "ports", i);
+		if (!port)
+			break;
+
+		if (!of_device_is_available(port->parent)) {
+			of_node_put(port);
+			continue;
+		}
+
+		component_match_add(dev, &match, compare_of, port->parent);
+		of_node_put(port);
+	}
+
+	if (i == 0) {
+		dev_err(dev, "missing 'ports' property\n");
+		return -ENODEV;
+	}
+
+	if (!match) {
+		dev_err(dev, "No available vop found for display-subsystem.\n");
+		return -ENODEV;
+	}
+	/*
+	 * For each bound crtc, bind the encoders attached to its
+	 * remote endpoint.
+	 */
+	for (i = 0;; i++) {
+		port = of_parse_phandle(np, "ports", i);
+		if (!port)
+			break;
+
+		if (!of_device_is_available(port->parent)) {
+			of_node_put(port);
+			continue;
+		}
+
+		rockchip_add_endpoints(dev, &match, port);
+		of_node_put(port);
+	}
+
+	return component_master_add_with_match(dev, &rockchip_drm_ops, match);
+}
+
+static int rockchip_drm_platform_remove(struct platform_device *pdev)
+{
+	component_master_del(&pdev->dev, &rockchip_drm_ops);
+
+	return 0;
+}
+
+static const struct of_device_id rockchip_drm_dt_ids[] = {
+	{ .compatible = "rockchip,display-subsystem", },
+	{ /* sentinel */ },
+};
+MODULE_DEVICE_TABLE(of, rockchip_drm_dt_ids);
+
+static struct platform_driver rockchip_drm_platform_driver = {
+	.probe = rockchip_drm_platform_probe,
+	.remove = rockchip_drm_platform_remove,
+	.driver = {
+		.owner = THIS_MODULE,
+		.name = "rockchip-drm",
+		.of_match_table = rockchip_drm_dt_ids,
+		.pm = &rockchip_drm_pm_ops,
+	},
+};
+
+module_platform_driver(rockchip_drm_platform_driver);
+
+MODULE_AUTHOR("Mark Yao <mark.yao@rock-chips.com>");
+MODULE_DESCRIPTION("ROCKCHIP DRM Driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_drv.h b/drivers/gpu/drm/rockchip/rockchip_drm_drv.h
new file mode 100644
index 000000000000..dc4e5f03ac79
--- /dev/null
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_drv.h
@@ -0,0 +1,68 @@
+/*
+ * Copyright (C) Fuzhou Rockchip Electronics Co.Ltd
+ * Author:Mark Yao <mark.yao@rock-chips.com>
+ *
+ * based on exynos_drm_drv.h
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#ifndef _ROCKCHIP_DRM_DRV_H
+#define _ROCKCHIP_DRM_DRV_H
+
+#include <drm/drm_fb_helper.h>
+#include <drm/drm_gem.h>
+
+#include <linux/module.h>
+#include <linux/component.h>
+
+#define ROCKCHIP_MAX_FB_BUFFER	3
+#define ROCKCHIP_MAX_CONNECTOR	2
+#define ROCKCHIP_MAX_CRTC	2
+
+struct drm_device;
+struct drm_connector;
+
+/*
+ * Rockchip drm private crtc funcs.
+ * @enable_vblank: enable crtc vblank irq.
+ * @disable_vblank: disable crtc vblank irq.
+ */
+struct rockchip_crtc_funcs {
+	int (*enable_vblank)(struct drm_crtc *crtc);
+	void (*disable_vblank)(struct drm_crtc *crtc);
+};
+
+/*
+ * Rockchip drm private structure.
+ *
+ * @crtc: array of enabled CRTCs, used to map from "pipe" to drm_crtc.
+ * @num_pipe: number of pipes for this device.
+ */
+struct rockchip_drm_private {
+	struct drm_fb_helper fbdev_helper;
+	struct drm_gem_object *fbdev_bo;
+	const struct rockchip_crtc_funcs *crtc_funcs[ROCKCHIP_MAX_CRTC];
+};
+
+int rockchip_register_crtc_funcs(struct drm_device *dev,
+				 const struct rockchip_crtc_funcs *crtc_funcs,
+				 int pipe);
+void rockchip_unregister_crtc_funcs(struct drm_device *dev, int pipe);
+int rockchip_drm_encoder_get_mux_id(struct device_node *node,
+				    struct drm_encoder *encoder);
+int rockchip_drm_crtc_mode_config(struct drm_crtc *crtc, int connector_type,
+				  int out_mode);
+int rockchip_drm_dma_attach_device(struct drm_device *drm_dev,
+				   struct device *dev);
+void rockchip_drm_dma_detach_device(struct drm_device *drm_dev,
+				    struct device *dev);
+
+#endif /* _ROCKCHIP_DRM_DRV_H_ */
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_fb.c b/drivers/gpu/drm/rockchip/rockchip_drm_fb.c
new file mode 100644
index 000000000000..77d52893d40f
--- /dev/null
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_fb.c
@@ -0,0 +1,201 @@
+/*
+ * Copyright (C) Fuzhou Rockchip Electronics Co.Ltd
+ * Author:Mark Yao <mark.yao@rock-chips.com>
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/kernel.h>
+#include <drm/drm.h>
+#include <drm/drmP.h>
+#include <drm/drm_fb_helper.h>
+#include <drm/drm_crtc_helper.h>
+
+#include "rockchip_drm_drv.h"
+#include "rockchip_drm_gem.h"
+
+#define to_rockchip_fb(x) container_of(x, struct rockchip_drm_fb, fb)
+
+struct rockchip_drm_fb {
+	struct drm_framebuffer fb;
+	struct drm_gem_object *obj[ROCKCHIP_MAX_FB_BUFFER];
+};
+
+struct drm_gem_object *rockchip_fb_get_gem_obj(struct drm_framebuffer *fb,
+					       unsigned int plane)
+{
+	struct rockchip_drm_fb *rk_fb = to_rockchip_fb(fb);
+
+	if (plane >= ROCKCHIP_MAX_FB_BUFFER)
+		return NULL;
+
+	return rk_fb->obj[plane];
+}
+EXPORT_SYMBOL_GPL(rockchip_fb_get_gem_obj);
+
+static void rockchip_drm_fb_destroy(struct drm_framebuffer *fb)
+{
+	struct rockchip_drm_fb *rockchip_fb = to_rockchip_fb(fb);
+	struct drm_gem_object *obj;
+	int i;
+
+	for (i = 0; i < ROCKCHIP_MAX_FB_BUFFER; i++) {
+		obj = rockchip_fb->obj[i];
+		if (obj)
+			drm_gem_object_unreference_unlocked(obj);
+	}
+
+	drm_framebuffer_cleanup(fb);
+	kfree(rockchip_fb);
+}
+
+static int rockchip_drm_fb_create_handle(struct drm_framebuffer *fb,
+					 struct drm_file *file_priv,
+					 unsigned int *handle)
+{
+	struct rockchip_drm_fb *rockchip_fb = to_rockchip_fb(fb);
+
+	return drm_gem_handle_create(file_priv,
+				     rockchip_fb->obj[0], handle);
+}
+
+static struct drm_framebuffer_funcs rockchip_drm_fb_funcs = {
+	.destroy	= rockchip_drm_fb_destroy,
+	.create_handle	= rockchip_drm_fb_create_handle,
+};
+
+static struct rockchip_drm_fb *
+rockchip_fb_alloc(struct drm_device *dev, struct drm_mode_fb_cmd2 *mode_cmd,
+		  struct drm_gem_object **obj, unsigned int num_planes)
+{
+	struct rockchip_drm_fb *rockchip_fb;
+	int ret;
+	int i;
+
+	rockchip_fb = kzalloc(sizeof(*rockchip_fb), GFP_KERNEL);
+	if (!rockchip_fb)
+		return ERR_PTR(-ENOMEM);
+
+	drm_helper_mode_fill_fb_struct(&rockchip_fb->fb, mode_cmd);
+
+	for (i = 0; i < num_planes; i++)
+		rockchip_fb->obj[i] = obj[i];
+
+	ret = drm_framebuffer_init(dev, &rockchip_fb->fb,
+				   &rockchip_drm_fb_funcs);
+	if (ret) {
+		dev_err(dev->dev, "Failed to initialize framebuffer: %d\n",
+			ret);
+		kfree(rockchip_fb);
+		return ERR_PTR(ret);
+	}
+
+	return rockchip_fb;
+}
+
+static struct drm_framebuffer *
+rockchip_user_fb_create(struct drm_device *dev, struct drm_file *file_priv,
+			struct drm_mode_fb_cmd2 *mode_cmd)
+{
+	struct rockchip_drm_fb *rockchip_fb;
+	struct drm_gem_object *objs[ROCKCHIP_MAX_FB_BUFFER];
+	struct drm_gem_object *obj;
+	unsigned int hsub;
+	unsigned int vsub;
+	int num_planes;
+	int ret;
+	int i;
+
+	hsub = drm_format_horz_chroma_subsampling(mode_cmd->pixel_format);
+	vsub = drm_format_vert_chroma_subsampling(mode_cmd->pixel_format);
+	num_planes = min(drm_format_num_planes(mode_cmd->pixel_format),
+			 ROCKCHIP_MAX_FB_BUFFER);
+
+	for (i = 0; i < num_planes; i++) {
+		unsigned int width = mode_cmd->width / (i ? hsub : 1);
+		unsigned int height = mode_cmd->height / (i ? vsub : 1);
+		unsigned int min_size;
+
+		obj = drm_gem_object_lookup(dev, file_priv,
+					    mode_cmd->handles[i]);
+		if (!obj) {
+			dev_err(dev->dev, "Failed to lookup GEM object\n");
+			ret = -ENXIO;
+			goto err_gem_object_unreference;
+		}
+
+		min_size = (height - 1) * mode_cmd->pitches[i] +
+			mode_cmd->offsets[i] +
+			width * drm_format_plane_cpp(mode_cmd->pixel_format, i);
+
+		if (obj->size < min_size) {
+			drm_gem_object_unreference_unlocked(obj);
+			ret = -EINVAL;
+			goto err_gem_object_unreference;
+		}
+		objs[i] = obj;
+	}
+
+	rockchip_fb = rockchip_fb_alloc(dev, mode_cmd, objs, i);
+	if (IS_ERR(rockchip_fb)) {
+		ret = PTR_ERR(rockchip_fb);
+		goto err_gem_object_unreference;
+	}
+
+	return &rockchip_fb->fb;
+
+err_gem_object_unreference:
+	for (i--; i >= 0; i--)
+		drm_gem_object_unreference_unlocked(objs[i]);
+	return ERR_PTR(ret);
+}
+
+static void rockchip_drm_output_poll_changed(struct drm_device *dev)
+{
+	struct rockchip_drm_private *private = dev->dev_private;
+	struct drm_fb_helper *fb_helper = &private->fbdev_helper;
+
+	drm_fb_helper_hotplug_event(fb_helper);
+}
+
+static const struct drm_mode_config_funcs rockchip_drm_mode_config_funcs = {
+	.fb_create = rockchip_user_fb_create,
+	.output_poll_changed = rockchip_drm_output_poll_changed,
+};
+
+struct drm_framebuffer *
+rockchip_drm_framebuffer_init(struct drm_device *dev,
+			      struct drm_mode_fb_cmd2 *mode_cmd,
+			      struct drm_gem_object *obj)
+{
+	struct rockchip_drm_fb *rockchip_fb;
+
+	rockchip_fb = rockchip_fb_alloc(dev, mode_cmd, &obj, 1);
+	if (IS_ERR(rockchip_fb))
+		return NULL;
+
+	return &rockchip_fb->fb;
+}
+
+void rockchip_drm_mode_config_init(struct drm_device *dev)
+{
+	dev->mode_config.min_width = 0;
+	dev->mode_config.min_height = 0;
+
+	/*
+	 * set max width and height as default value(4096x4096).
+	 * this value would be used to check framebuffer size limitation
+	 * at drm_mode_addfb().
+	 */
+	dev->mode_config.max_width = 4096;
+	dev->mode_config.max_height = 4096;
+
+	dev->mode_config.funcs = &rockchip_drm_mode_config_funcs;
+}
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_fb.h b/drivers/gpu/drm/rockchip/rockchip_drm_fb.h
new file mode 100644
index 000000000000..09574d48226f
--- /dev/null
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_fb.h
@@ -0,0 +1,28 @@
+/*
+ * Copyright (C) Fuzhou Rockchip Electronics Co.Ltd
+ * Author:Mark Yao <mark.yao@rock-chips.com>
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#ifndef _ROCKCHIP_DRM_FB_H
+#define _ROCKCHIP_DRM_FB_H
+
+struct drm_framebuffer *
+rockchip_drm_framebuffer_init(struct drm_device *dev,
+			      struct drm_mode_fb_cmd2 *mode_cmd,
+			      struct drm_gem_object *obj);
+void rockchip_drm_framebuffer_fini(struct drm_framebuffer *fb);
+
+void rockchip_drm_mode_config_init(struct drm_device *dev);
+
+struct drm_gem_object *rockchip_fb_get_gem_obj(struct drm_framebuffer *fb,
+					       unsigned int plane);
+#endif /* _ROCKCHIP_DRM_FB_H */
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_fbdev.c b/drivers/gpu/drm/rockchip/rockchip_drm_fbdev.c
new file mode 100644
index 000000000000..a5d889a8716b
--- /dev/null
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_fbdev.c
@@ -0,0 +1,210 @@
+/*
+ * Copyright (C) Fuzhou Rockchip Electronics Co.Ltd
+ * Author:Mark Yao <mark.yao@rock-chips.com>
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <drm/drm.h>
+#include <drm/drmP.h>
+#include <drm/drm_fb_helper.h>
+#include <drm/drm_crtc_helper.h>
+
+#include "rockchip_drm_drv.h"
+#include "rockchip_drm_gem.h"
+#include "rockchip_drm_fb.h"
+
+#define PREFERRED_BPP		32
+#define to_drm_private(x) \
+		container_of(x, struct rockchip_drm_private, fbdev_helper)
+
+static int rockchip_fbdev_mmap(struct fb_info *info,
+			       struct vm_area_struct *vma)
+{
+	struct drm_fb_helper *helper = info->par;
+	struct rockchip_drm_private *private = to_drm_private(helper);
+
+	return rockchip_gem_mmap_buf(private->fbdev_bo, vma);
+}
+
+static struct fb_ops rockchip_drm_fbdev_ops = {
+	.owner		= THIS_MODULE,
+	.fb_mmap	= rockchip_fbdev_mmap,
+	.fb_fillrect	= cfb_fillrect,
+	.fb_copyarea	= cfb_copyarea,
+	.fb_imageblit	= cfb_imageblit,
+	.fb_check_var	= drm_fb_helper_check_var,
+	.fb_set_par	= drm_fb_helper_set_par,
+	.fb_blank	= drm_fb_helper_blank,
+	.fb_pan_display	= drm_fb_helper_pan_display,
+	.fb_setcmap	= drm_fb_helper_setcmap,
+};
+
+static int rockchip_drm_fbdev_create(struct drm_fb_helper *helper,
+				     struct drm_fb_helper_surface_size *sizes)
+{
+	struct rockchip_drm_private *private = to_drm_private(helper);
+	struct drm_mode_fb_cmd2 mode_cmd = { 0 };
+	struct drm_device *dev = helper->dev;
+	struct rockchip_gem_object *rk_obj;
+	struct drm_framebuffer *fb;
+	unsigned int bytes_per_pixel;
+	unsigned long offset;
+	struct fb_info *fbi;
+	size_t size;
+	int ret;
+
+	bytes_per_pixel = DIV_ROUND_UP(sizes->surface_bpp, 8);
+
+	mode_cmd.width = sizes->surface_width;
+	mode_cmd.height = sizes->surface_height;
+	mode_cmd.pitches[0] = sizes->surface_width * bytes_per_pixel;
+	mode_cmd.pixel_format = drm_mode_legacy_fb_format(sizes->surface_bpp,
+		sizes->surface_depth);
+
+	size = mode_cmd.pitches[0] * mode_cmd.height;
+
+	rk_obj = rockchip_gem_create_object(dev, size);
+	if (IS_ERR(rk_obj))
+		return -ENOMEM;
+
+	private->fbdev_bo = &rk_obj->base;
+
+	fbi = framebuffer_alloc(0, dev->dev);
+	if (!fbi) {
+		dev_err(dev->dev, "Failed to allocate framebuffer info.\n");
+		ret = -ENOMEM;
+		goto err_rockchip_gem_free_object;
+	}
+
+	helper->fb = rockchip_drm_framebuffer_init(dev, &mode_cmd,
+						   private->fbdev_bo);
+	if (IS_ERR(helper->fb)) {
+		dev_err(dev->dev, "Failed to allocate DRM framebuffer.\n");
+		ret = PTR_ERR(helper->fb);
+		goto err_framebuffer_release;
+	}
+
+	helper->fbdev = fbi;
+
+	fbi->par = helper;
+	fbi->flags = FBINFO_FLAG_DEFAULT;
+	fbi->fbops = &rockchip_drm_fbdev_ops;
+
+	ret = fb_alloc_cmap(&fbi->cmap, 256, 0);
+	if (ret) {
+		dev_err(dev->dev, "Failed to allocate color map.\n");
+		goto err_drm_framebuffer_unref;
+	}
+
+	fb = helper->fb;
+	drm_fb_helper_fill_fix(fbi, fb->pitches[0], fb->depth);
+	drm_fb_helper_fill_var(fbi, helper, fb->width, fb->height);
+
+	offset = fbi->var.xoffset * bytes_per_pixel;
+	offset += fbi->var.yoffset * fb->pitches[0];
+
+	dev->mode_config.fb_base = 0;
+	fbi->screen_base = rk_obj->kvaddr + offset;
+	fbi->screen_size = rk_obj->base.size;
+	fbi->fix.smem_len = rk_obj->base.size;
+
+	DRM_DEBUG_KMS("FB [%dx%d]-%d kvaddr=%p offset=%ld size=%d\n",
+		      fb->width, fb->height, fb->depth, rk_obj->kvaddr,
+		      offset, size);
+	return 0;
+
+err_drm_framebuffer_unref:
+	drm_framebuffer_unreference(helper->fb);
+err_framebuffer_release:
+	framebuffer_release(fbi);
+err_rockchip_gem_free_object:
+	rockchip_gem_free_object(&rk_obj->base);
+	return ret;
+}
+
+static const struct drm_fb_helper_funcs rockchip_drm_fb_helper_funcs = {
+	.fb_probe = rockchip_drm_fbdev_create,
+};
+
+int rockchip_drm_fbdev_init(struct drm_device *dev)
+{
+	struct rockchip_drm_private *private = dev->dev_private;
+	struct drm_fb_helper *helper;
+	unsigned int num_crtc;
+	int ret;
+
+	if (!dev->mode_config.num_crtc || !dev->mode_config.num_connector)
+		return -EINVAL;
+
+	num_crtc = dev->mode_config.num_crtc;
+
+	helper = &private->fbdev_helper;
+
+	drm_fb_helper_prepare(dev, helper, &rockchip_drm_fb_helper_funcs);
+
+	ret = drm_fb_helper_init(dev, helper, num_crtc, ROCKCHIP_MAX_CONNECTOR);
+	if (ret < 0) {
+		dev_err(dev->dev, "Failed to initialize drm fb helper - %d.\n",
+			ret);
+		return ret;
+	}
+
+	ret = drm_fb_helper_single_add_all_connectors(helper);
+	if (ret < 0) {
+		dev_err(dev->dev, "Failed to add connectors - %d.\n", ret);
+		goto err_drm_fb_helper_fini;
+	}
+
+	/* disable all the possible outputs/crtcs before entering KMS mode */
+	drm_helper_disable_unused_functions(dev);
+
+	ret = drm_fb_helper_initial_config(helper, PREFERRED_BPP);
+	if (ret < 0) {
+		dev_err(dev->dev, "Failed to set initial hw config - %d.\n",
+			ret);
+		goto err_drm_fb_helper_fini;
+	}
+
+	return 0;
+
+err_drm_fb_helper_fini:
+	drm_fb_helper_fini(helper);
+	return ret;
+}
+
+void rockchip_drm_fbdev_fini(struct drm_device *dev)
+{
+	struct rockchip_drm_private *private = dev->dev_private;
+	struct drm_fb_helper *helper;
+
+	helper = &private->fbdev_helper;
+
+	if (helper->fbdev) {
+		struct fb_info *info;
+		int ret;
+
+		info = helper->fbdev;
+		ret = unregister_framebuffer(info);
+		if (ret < 0)
+			DRM_DEBUG_KMS("failed unregister_framebuffer() - %d\n",
+				      ret);
+
+		if (info->cmap.len)
+			fb_dealloc_cmap(&info->cmap);
+
+		framebuffer_release(info);
+	}
+
+	if (helper->fb)
+		drm_framebuffer_unreference(helper->fb);
+
+	drm_fb_helper_fini(helper);
+}
diff --git a/drivers/staging/android/binder.h b/drivers/gpu/drm/rockchip/rockchip_drm_fbdev.h
index eb0834656dfe..50432e9b5b37 100644
--- a/drivers/staging/android/binder.h
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_fbdev.h
@@ -1,10 +1,6 @@
 /*
- * Copyright (C) 2008 Google, Inc.
- *
- * Based on, but no longer compatible with, the original
- * OpenBinder.org binder driver interface, which is:
- *
- * Copyright (c) 2005 Palmsource, Inc.
+ * Copyright (C) Fuzhou Rockchip Electronics Co.Ltd
+ * Author:Mark Yao <mark.yao@rock-chips.com>
  *
  * This software is licensed under the terms of the GNU General Public
  * License version 2, as published by the Free Software Foundation, and
@@ -14,17 +10,12 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
  */
 
-#ifndef _LINUX_BINDER_H
-#define _LINUX_BINDER_H
-
-#ifdef CONFIG_ANDROID_BINDER_IPC_32BIT
-#define BINDER_IPC_32BIT 1
-#endif
-
-#include "uapi/binder.h"
+#ifndef _ROCKCHIP_DRM_FBDEV_H
+#define _ROCKCHIP_DRM_FBDEV_H
 
-#endif /* _LINUX_BINDER_H */
+int rockchip_drm_fbdev_init(struct drm_device *dev);
+void rockchip_drm_fbdev_fini(struct drm_device *dev);
 
+#endif /* _ROCKCHIP_DRM_FBDEV_H */
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_gem.c b/drivers/gpu/drm/rockchip/rockchip_drm_gem.c
new file mode 100644
index 000000000000..bc98a227dc76
--- /dev/null
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_gem.c
@@ -0,0 +1,294 @@
+/*
+ * Copyright (C) Fuzhou Rockchip Electronics Co.Ltd
+ * Author:Mark Yao <mark.yao@rock-chips.com>
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <drm/drm.h>
+#include <drm/drmP.h>
+#include <drm/drm_gem.h>
+#include <drm/drm_vma_manager.h>
+
+#include <linux/dma-attrs.h>
+
+#include "rockchip_drm_drv.h"
+#include "rockchip_drm_gem.h"
+
+static int rockchip_gem_alloc_buf(struct rockchip_gem_object *rk_obj)
+{
+	struct drm_gem_object *obj = &rk_obj->base;
+	struct drm_device *drm = obj->dev;
+
+	init_dma_attrs(&rk_obj->dma_attrs);
+	dma_set_attr(DMA_ATTR_WRITE_COMBINE, &rk_obj->dma_attrs);
+
+	/* TODO(djkurtz): Use DMA_ATTR_NO_KERNEL_MAPPING except for fbdev */
+	rk_obj->kvaddr = dma_alloc_attrs(drm->dev, obj->size,
+					 &rk_obj->dma_addr, GFP_KERNEL,
+					 &rk_obj->dma_attrs);
+	if (IS_ERR(rk_obj->kvaddr)) {
+		int ret = PTR_ERR(rk_obj->kvaddr);
+
+		DRM_ERROR("failed to allocate %#x byte dma buffer, %d",
+			  obj->size, ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+static void rockchip_gem_free_buf(struct rockchip_gem_object *rk_obj)
+{
+	struct drm_gem_object *obj = &rk_obj->base;
+	struct drm_device *drm = obj->dev;
+
+	dma_free_attrs(drm->dev, obj->size, rk_obj->kvaddr, rk_obj->dma_addr,
+		       &rk_obj->dma_attrs);
+}
+
+int rockchip_gem_mmap_buf(struct drm_gem_object *obj,
+			  struct vm_area_struct *vma)
+{
+	struct rockchip_gem_object *rk_obj = to_rockchip_obj(obj);
+	struct drm_device *drm = obj->dev;
+	unsigned long vm_size;
+
+	vma->vm_flags |= VM_IO | VM_DONTEXPAND | VM_DONTDUMP;
+	vm_size = vma->vm_end - vma->vm_start;
+
+	if (vm_size > obj->size)
+		return -EINVAL;
+
+	return dma_mmap_attrs(drm->dev, vma, rk_obj->kvaddr, rk_obj->dma_addr,
+			     obj->size, &rk_obj->dma_attrs);
+}
+
+/* drm driver mmap file operations */
+int rockchip_gem_mmap(struct file *filp, struct vm_area_struct *vma)
+{
+	struct drm_file *priv = filp->private_data;
+	struct drm_device *dev = priv->minor->dev;
+	struct drm_gem_object *obj;
+	struct drm_vma_offset_node *node;
+	int ret;
+
+	if (drm_device_is_unplugged(dev))
+		return -ENODEV;
+
+	mutex_lock(&dev->struct_mutex);
+
+	node = drm_vma_offset_exact_lookup(dev->vma_offset_manager,
+					   vma->vm_pgoff,
+					   vma_pages(vma));
+	if (!node) {
+		mutex_unlock(&dev->struct_mutex);
+		DRM_ERROR("failed to find vma node.\n");
+		return -EINVAL;
+	} else if (!drm_vma_node_is_allowed(node, filp)) {
+		mutex_unlock(&dev->struct_mutex);
+		return -EACCES;
+	}
+
+	obj = container_of(node, struct drm_gem_object, vma_node);
+	ret = rockchip_gem_mmap_buf(obj, vma);
+
+	mutex_unlock(&dev->struct_mutex);
+
+	return ret;
+}
+
+struct rockchip_gem_object *
+	rockchip_gem_create_object(struct drm_device *drm, unsigned int size)
+{
+	struct rockchip_gem_object *rk_obj;
+	struct drm_gem_object *obj;
+	int ret;
+
+	size = round_up(size, PAGE_SIZE);
+
+	rk_obj = kzalloc(sizeof(*rk_obj), GFP_KERNEL);
+	if (!rk_obj)
+		return ERR_PTR(-ENOMEM);
+
+	obj = &rk_obj->base;
+
+	drm_gem_private_object_init(drm, obj, size);
+
+	ret = rockchip_gem_alloc_buf(rk_obj);
+	if (ret)
+		goto err_free_rk_obj;
+
+	return rk_obj;
+
+err_free_rk_obj:
+	kfree(rk_obj);
+	return ERR_PTR(ret);
+}
+
+/*
+ * rockchip_gem_free_object - (struct drm_driver)->gem_free_object callback
+ * function
+ */
+void rockchip_gem_free_object(struct drm_gem_object *obj)
+{
+	struct rockchip_gem_object *rk_obj;
+
+	drm_gem_free_mmap_offset(obj);
+
+	rk_obj = to_rockchip_obj(obj);
+
+	rockchip_gem_free_buf(rk_obj);
+
+	kfree(rk_obj);
+}
+
+/*
+ * rockchip_gem_create_with_handle - allocate an object with the given
+ * size and create a gem handle on it
+ *
+ * returns a struct rockchip_gem_object* on success or ERR_PTR values
+ * on failure.
+ */
+static struct rockchip_gem_object *
+rockchip_gem_create_with_handle(struct drm_file *file_priv,
+				struct drm_device *drm, unsigned int size,
+				unsigned int *handle)
+{
+	struct rockchip_gem_object *rk_obj;
+	struct drm_gem_object *obj;
+	int ret;
+
+	rk_obj = rockchip_gem_create_object(drm, size);
+	if (IS_ERR(rk_obj))
+		return ERR_CAST(rk_obj);
+
+	obj = &rk_obj->base;
+
+	/*
+	 * allocate a id of idr table where the obj is registered
+	 * and handle has the id what user can see.
+	 */
+	ret = drm_gem_handle_create(file_priv, obj, handle);
+	if (ret)
+		goto err_handle_create;
+
+	/* drop reference from allocate - handle holds it now. */
+	drm_gem_object_unreference_unlocked(obj);
+
+	return rk_obj;
+
+err_handle_create:
+	rockchip_gem_free_object(obj);
+
+	return ERR_PTR(ret);
+}
+
+int rockchip_gem_dumb_map_offset(struct drm_file *file_priv,
+				 struct drm_device *dev, uint32_t handle,
+				 uint64_t *offset)
+{
+	struct drm_gem_object *obj;
+	int ret;
+
+	mutex_lock(&dev->struct_mutex);
+
+	obj = drm_gem_object_lookup(dev, file_priv, handle);
+	if (!obj) {
+		DRM_ERROR("failed to lookup gem object.\n");
+		ret = -EINVAL;
+		goto unlock;
+	}
+
+	ret = drm_gem_create_mmap_offset(obj);
+	if (ret)
+		goto out;
+
+	*offset = drm_vma_node_offset_addr(&obj->vma_node);
+	DRM_DEBUG_KMS("offset = 0x%llx\n", *offset);
+
+out:
+	drm_gem_object_unreference(obj);
+unlock:
+	mutex_unlock(&dev->struct_mutex);
+	return ret;
+}
+
+/*
+ * rockchip_gem_dumb_create - (struct drm_driver)->dumb_create callback
+ * function
+ *
+ * This aligns the pitch and size arguments to the minimum required. wrap
+ * this into your own function if you need bigger alignment.
+ */
+int rockchip_gem_dumb_create(struct drm_file *file_priv,
+			     struct drm_device *dev,
+			     struct drm_mode_create_dumb *args)
+{
+	struct rockchip_gem_object *rk_obj;
+	int min_pitch = DIV_ROUND_UP(args->width * args->bpp, 8);
+
+	/*
+	 * align to 64 bytes since Mali requires it.
+	 */
+	min_pitch = ALIGN(min_pitch, 64);
+
+	if (args->pitch < min_pitch)
+		args->pitch = min_pitch;
+
+	if (args->size < args->pitch * args->height)
+		args->size = args->pitch * args->height;
+
+	rk_obj = rockchip_gem_create_with_handle(file_priv, dev, args->size,
+						 &args->handle);
+
+	return PTR_ERR_OR_ZERO(rk_obj);
+}
+
+/*
+ * Allocate a sg_table for this GEM object.
+ * Note: Both the table's contents, and the sg_table itself must be freed by
+ *       the caller.
+ * Returns a pointer to the newly allocated sg_table, or an ERR_PTR() error.
+ */
+struct sg_table *rockchip_gem_prime_get_sg_table(struct drm_gem_object *obj)
+{
+	struct rockchip_gem_object *rk_obj = to_rockchip_obj(obj);
+	struct drm_device *drm = obj->dev;
+	struct sg_table *sgt;
+	int ret;
+
+	sgt = kzalloc(sizeof(*sgt), GFP_KERNEL);
+	if (!sgt)
+		return ERR_PTR(-ENOMEM);
+
+	ret = dma_get_sgtable_attrs(drm->dev, sgt, rk_obj->kvaddr,
+				    rk_obj->dma_addr, obj->size,
+				    &rk_obj->dma_attrs);
+	if (ret) {
+		DRM_ERROR("failed to allocate sgt, %d\n", ret);
+		kfree(sgt);
+		return ERR_PTR(ret);
+	}
+
+	return sgt;
+}
+
+void *rockchip_gem_prime_vmap(struct drm_gem_object *obj)
+{
+	struct rockchip_gem_object *rk_obj = to_rockchip_obj(obj);
+
+	return rk_obj->kvaddr;
+}
+
+void rockchip_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr)
+{
+	/* Nothing to do */
+}
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_gem.h b/drivers/gpu/drm/rockchip/rockchip_drm_gem.h
new file mode 100644
index 000000000000..67bcebe90003
--- /dev/null
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_gem.h
@@ -0,0 +1,54 @@
+/*
+ * Copyright (C) Fuzhou Rockchip Electronics Co.Ltd
+ * Author:Mark Yao <mark.yao@rock-chips.com>
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#ifndef _ROCKCHIP_DRM_GEM_H
+#define _ROCKCHIP_DRM_GEM_H
+
+#define to_rockchip_obj(x) container_of(x, struct rockchip_gem_object, base)
+
+struct rockchip_gem_object {
+	struct drm_gem_object base;
+	unsigned int flags;
+
+	void *kvaddr;
+	dma_addr_t dma_addr;
+	struct dma_attrs dma_attrs;
+};
+
+struct sg_table *rockchip_gem_prime_get_sg_table(struct drm_gem_object *obj);
+struct drm_gem_object *
+rockchip_gem_prime_import_sg_table(struct drm_device *dev, size_t size,
+				   struct sg_table *sgt);
+void *rockchip_gem_prime_vmap(struct drm_gem_object *obj);
+void rockchip_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr);
+
+/* drm driver mmap file operations */
+int rockchip_gem_mmap(struct file *filp, struct vm_area_struct *vma);
+
+/* mmap a gem object to userspace. */
+int rockchip_gem_mmap_buf(struct drm_gem_object *obj,
+			  struct vm_area_struct *vma);
+
+struct rockchip_gem_object *
+	rockchip_gem_create_object(struct drm_device *drm, unsigned int size);
+
+void rockchip_gem_free_object(struct drm_gem_object *obj);
+
+int rockchip_gem_dumb_create(struct drm_file *file_priv,
+			     struct drm_device *dev,
+			     struct drm_mode_create_dumb *args);
+int rockchip_gem_dumb_map_offset(struct drm_file *file_priv,
+				 struct drm_device *dev, uint32_t handle,
+				 uint64_t *offset);
+#endif /* _ROCKCHIP_DRM_GEM_H */
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
new file mode 100644
index 000000000000..e7ca25b3fb38
--- /dev/null
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
@@ -0,0 +1,1455 @@
+/*
+ * Copyright (C) Fuzhou Rockchip Electronics Co.Ltd
+ * Author:Mark Yao <mark.yao@rock-chips.com>
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <drm/drm.h>
+#include <drm/drmP.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_plane_helper.h>
+
+#include <linux/kernel.h>
+#include <linux/platform_device.h>
+#include <linux/clk.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/component.h>
+
+#include <linux/reset.h>
+#include <linux/delay.h>
+
+#include "rockchip_drm_drv.h"
+#include "rockchip_drm_gem.h"
+#include "rockchip_drm_fb.h"
+#include "rockchip_drm_vop.h"
+
+#define VOP_REG(off, _mask, s) \
+		{.offset = off, \
+		 .mask = _mask, \
+		 .shift = s,}
+
+#define __REG_SET_RELAXED(x, off, mask, shift, v) \
+		vop_mask_write_relaxed(x, off, (mask) << shift, (v) << shift)
+#define __REG_SET_NORMAL(x, off, mask, shift, v) \
+		vop_mask_write(x, off, (mask) << shift, (v) << shift)
+
+#define REG_SET(x, base, reg, v, mode) \
+		__REG_SET_##mode(x, base + reg.offset, reg.mask, reg.shift, v)
+
+#define VOP_WIN_SET(x, win, name, v) \
+		REG_SET(x, win->base, win->phy->name, v, RELAXED)
+#define VOP_CTRL_SET(x, name, v) \
+		REG_SET(x, 0, (x)->data->ctrl->name, v, NORMAL)
+
+#define VOP_WIN_GET(x, win, name) \
+		vop_read_reg(x, win->base, &win->phy->name)
+
+#define VOP_WIN_GET_YRGBADDR(vop, win) \
+		vop_readl(vop, win->base + win->phy->yrgb_mst.offset)
+
+#define to_vop(x) container_of(x, struct vop, crtc)
+#define to_vop_win(x) container_of(x, struct vop_win, base)
+
+struct vop_win_state {
+	struct list_head head;
+	struct drm_framebuffer *fb;
+	dma_addr_t yrgb_mst;
+	struct drm_pending_vblank_event *event;
+};
+
+struct vop_win {
+	struct drm_plane base;
+	const struct vop_win_data *data;
+	struct vop *vop;
+
+	struct list_head pending;
+	struct vop_win_state *active;
+};
+
+struct vop {
+	struct drm_crtc crtc;
+	struct device *dev;
+	struct drm_device *drm_dev;
+	unsigned int dpms;
+
+	int connector_type;
+	int connector_out_mode;
+
+	/* mutex vsync_ work */
+	struct mutex vsync_mutex;
+	bool vsync_work_pending;
+
+	const struct vop_data *data;
+
+	uint32_t *regsbak;
+	void __iomem *regs;
+
+	/* physical map length of vop register */
+	uint32_t len;
+
+	/* one time only one process allowed to config the register */
+	spinlock_t reg_lock;
+	/* lock vop irq reg */
+	spinlock_t irq_lock;
+
+	unsigned int irq;
+
+	/* vop AHP clk */
+	struct clk *hclk;
+	/* vop dclk */
+	struct clk *dclk;
+	/* vop share memory frequency */
+	struct clk *aclk;
+
+	/* vop dclk reset */
+	struct reset_control *dclk_rst;
+
+	int pipe;
+
+	struct vop_win win[];
+};
+
+enum vop_data_format {
+	VOP_FMT_ARGB8888 = 0,
+	VOP_FMT_RGB888,
+	VOP_FMT_RGB565,
+	VOP_FMT_YUV420SP = 4,
+	VOP_FMT_YUV422SP,
+	VOP_FMT_YUV444SP,
+};
+
+struct vop_reg_data {
+	uint32_t offset;
+	uint32_t value;
+};
+
+struct vop_reg {
+	uint32_t offset;
+	uint32_t shift;
+	uint32_t mask;
+};
+
+struct vop_ctrl {
+	struct vop_reg standby;
+	struct vop_reg data_blank;
+	struct vop_reg gate_en;
+	struct vop_reg mmu_en;
+	struct vop_reg rgb_en;
+	struct vop_reg edp_en;
+	struct vop_reg hdmi_en;
+	struct vop_reg mipi_en;
+	struct vop_reg out_mode;
+	struct vop_reg dither_down;
+	struct vop_reg dither_up;
+	struct vop_reg pin_pol;
+
+	struct vop_reg htotal_pw;
+	struct vop_reg hact_st_end;
+	struct vop_reg vtotal_pw;
+	struct vop_reg vact_st_end;
+	struct vop_reg hpost_st_end;
+	struct vop_reg vpost_st_end;
+};
+
+struct vop_win_phy {
+	const uint32_t *data_formats;
+	uint32_t nformats;
+
+	struct vop_reg enable;
+	struct vop_reg format;
+	struct vop_reg act_info;
+	struct vop_reg dsp_info;
+	struct vop_reg dsp_st;
+	struct vop_reg yrgb_mst;
+	struct vop_reg uv_mst;
+	struct vop_reg yrgb_vir;
+	struct vop_reg uv_vir;
+
+	struct vop_reg dst_alpha_ctl;
+	struct vop_reg src_alpha_ctl;
+};
+
+struct vop_win_data {
+	uint32_t base;
+	const struct vop_win_phy *phy;
+	enum drm_plane_type type;
+};
+
+struct vop_data {
+	const struct vop_reg_data *init_table;
+	unsigned int table_size;
+	const struct vop_ctrl *ctrl;
+	const struct vop_win_data *win;
+	unsigned int win_size;
+};
+
+static const uint32_t formats_01[] = {
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_RGB888,
+	DRM_FORMAT_RGB565,
+	DRM_FORMAT_NV12,
+	DRM_FORMAT_NV16,
+	DRM_FORMAT_NV24,
+};
+
+static const uint32_t formats_234[] = {
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_RGB888,
+	DRM_FORMAT_RGB565,
+};
+
+static const struct vop_win_phy win01_data = {
+	.data_formats = formats_01,
+	.nformats = ARRAY_SIZE(formats_01),
+	.enable = VOP_REG(WIN0_CTRL0, 0x1, 0),
+	.format = VOP_REG(WIN0_CTRL0, 0x7, 1),
+	.act_info = VOP_REG(WIN0_ACT_INFO, 0x1fff1fff, 0),
+	.dsp_info = VOP_REG(WIN0_DSP_INFO, 0x0fff0fff, 0),
+	.dsp_st = VOP_REG(WIN0_DSP_ST, 0x1fff1fff, 0),
+	.yrgb_mst = VOP_REG(WIN0_YRGB_MST, 0xffffffff, 0),
+	.uv_mst = VOP_REG(WIN0_CBR_MST, 0xffffffff, 0),
+	.yrgb_vir = VOP_REG(WIN0_VIR, 0x3fff, 0),
+	.uv_vir = VOP_REG(WIN0_VIR, 0x3fff, 16),
+	.src_alpha_ctl = VOP_REG(WIN0_SRC_ALPHA_CTRL, 0xff, 0),
+	.dst_alpha_ctl = VOP_REG(WIN0_DST_ALPHA_CTRL, 0xff, 0),
+};
+
+static const struct vop_win_phy win23_data = {
+	.data_formats = formats_234,
+	.nformats = ARRAY_SIZE(formats_234),
+	.enable = VOP_REG(WIN2_CTRL0, 0x1, 0),
+	.format = VOP_REG(WIN2_CTRL0, 0x7, 1),
+	.dsp_info = VOP_REG(WIN2_DSP_INFO0, 0x0fff0fff, 0),
+	.dsp_st = VOP_REG(WIN2_DSP_ST0, 0x1fff1fff, 0),
+	.yrgb_mst = VOP_REG(WIN2_MST0, 0xffffffff, 0),
+	.yrgb_vir = VOP_REG(WIN2_VIR0_1, 0x1fff, 0),
+	.src_alpha_ctl = VOP_REG(WIN2_SRC_ALPHA_CTRL, 0xff, 0),
+	.dst_alpha_ctl = VOP_REG(WIN2_DST_ALPHA_CTRL, 0xff, 0),
+};
+
+static const struct vop_win_phy cursor_data = {
+	.data_formats = formats_234,
+	.nformats = ARRAY_SIZE(formats_234),
+	.enable = VOP_REG(HWC_CTRL0, 0x1, 0),
+	.format = VOP_REG(HWC_CTRL0, 0x7, 1),
+	.dsp_st = VOP_REG(HWC_DSP_ST, 0x1fff1fff, 0),
+	.yrgb_mst = VOP_REG(HWC_MST, 0xffffffff, 0),
+};
+
+static const struct vop_ctrl ctrl_data = {
+	.standby = VOP_REG(SYS_CTRL, 0x1, 22),
+	.gate_en = VOP_REG(SYS_CTRL, 0x1, 23),
+	.mmu_en = VOP_REG(SYS_CTRL, 0x1, 20),
+	.rgb_en = VOP_REG(SYS_CTRL, 0x1, 12),
+	.hdmi_en = VOP_REG(SYS_CTRL, 0x1, 13),
+	.edp_en = VOP_REG(SYS_CTRL, 0x1, 14),
+	.mipi_en = VOP_REG(SYS_CTRL, 0x1, 15),
+	.dither_down = VOP_REG(DSP_CTRL1, 0xf, 1),
+	.dither_up = VOP_REG(DSP_CTRL1, 0x1, 6),
+	.data_blank = VOP_REG(DSP_CTRL0, 0x1, 19),
+	.out_mode = VOP_REG(DSP_CTRL0, 0xf, 0),
+	.pin_pol = VOP_REG(DSP_CTRL0, 0xf, 4),
+	.htotal_pw = VOP_REG(DSP_HTOTAL_HS_END, 0x1fff1fff, 0),
+	.hact_st_end = VOP_REG(DSP_HACT_ST_END, 0x1fff1fff, 0),
+	.vtotal_pw = VOP_REG(DSP_VTOTAL_VS_END, 0x1fff1fff, 0),
+	.vact_st_end = VOP_REG(DSP_VACT_ST_END, 0x1fff1fff, 0),
+	.hpost_st_end = VOP_REG(POST_DSP_HACT_INFO, 0x1fff1fff, 0),
+	.vpost_st_end = VOP_REG(POST_DSP_VACT_INFO, 0x1fff1fff, 0),
+};
+
+static const struct vop_reg_data vop_init_reg_table[] = {
+	{SYS_CTRL, 0x00c00000},
+	{DSP_CTRL0, 0x00000000},
+	{WIN0_CTRL0, 0x00000080},
+	{WIN1_CTRL0, 0x00000080},
+};
+
+/*
+ * Note: rk3288 has a dedicated 'cursor' window, however, that window requires
+ * special support to get alpha blending working.  For now, just use overlay
+ * window 1 for the drm cursor.
+ */
+static const struct vop_win_data rk3288_vop_win_data[] = {
+	{ .base = 0x00, .phy = &win01_data, .type = DRM_PLANE_TYPE_PRIMARY },
+	{ .base = 0x40, .phy = &win01_data, .type = DRM_PLANE_TYPE_CURSOR },
+	{ .base = 0x00, .phy = &win23_data, .type = DRM_PLANE_TYPE_OVERLAY },
+	{ .base = 0x50, .phy = &win23_data, .type = DRM_PLANE_TYPE_OVERLAY },
+	{ .base = 0x00, .phy = &cursor_data, .type = DRM_PLANE_TYPE_OVERLAY },
+};
+
+static const struct vop_data rk3288_vop = {
+	.init_table = vop_init_reg_table,
+	.table_size = ARRAY_SIZE(vop_init_reg_table),
+	.ctrl = &ctrl_data,
+	.win = rk3288_vop_win_data,
+	.win_size = ARRAY_SIZE(rk3288_vop_win_data),
+};
+
+static const struct of_device_id vop_driver_dt_match[] = {
+	{ .compatible = "rockchip,rk3288-vop",
+	  .data = &rk3288_vop },
+	{},
+};
+
+static inline void vop_writel(struct vop *vop, uint32_t offset, uint32_t v)
+{
+	writel(v, vop->regs + offset);
+	vop->regsbak[offset >> 2] = v;
+}
+
+static inline uint32_t vop_readl(struct vop *vop, uint32_t offset)
+{
+	return readl(vop->regs + offset);
+}
+
+static inline uint32_t vop_read_reg(struct vop *vop, uint32_t base,
+				    const struct vop_reg *reg)
+{
+	return (vop_readl(vop, base + reg->offset) >> reg->shift) & reg->mask;
+}
+
+static inline void vop_cfg_done(struct vop *vop)
+{
+	writel(0x01, vop->regs + REG_CFG_DONE);
+}
+
+static inline void vop_mask_write(struct vop *vop, uint32_t offset,
+				  uint32_t mask, uint32_t v)
+{
+	if (mask) {
+		uint32_t cached_val = vop->regsbak[offset >> 2];
+
+		cached_val = (cached_val & ~mask) | v;
+		writel(cached_val, vop->regs + offset);
+		vop->regsbak[offset >> 2] = cached_val;
+	}
+}
+
+static inline void vop_mask_write_relaxed(struct vop *vop, uint32_t offset,
+					  uint32_t mask, uint32_t v)
+{
+	if (mask) {
+		uint32_t cached_val = vop->regsbak[offset >> 2];
+
+		cached_val = (cached_val & ~mask) | v;
+		writel_relaxed(cached_val, vop->regs + offset);
+		vop->regsbak[offset >> 2] = cached_val;
+	}
+}
+
+static enum vop_data_format vop_convert_format(uint32_t format)
+{
+	switch (format) {
+	case DRM_FORMAT_XRGB8888:
+	case DRM_FORMAT_ARGB8888:
+		return VOP_FMT_ARGB8888;
+	case DRM_FORMAT_RGB888:
+		return VOP_FMT_RGB888;
+	case DRM_FORMAT_RGB565:
+		return VOP_FMT_RGB565;
+	case DRM_FORMAT_NV12:
+		return VOP_FMT_YUV420SP;
+	case DRM_FORMAT_NV16:
+		return VOP_FMT_YUV422SP;
+	case DRM_FORMAT_NV24:
+		return VOP_FMT_YUV444SP;
+	default:
+		DRM_ERROR("unsupport format[%08x]\n", format);
+		return -EINVAL;
+	}
+}
+
+static bool is_alpha_support(uint32_t format)
+{
+	switch (format) {
+	case DRM_FORMAT_ARGB8888:
+		return true;
+	default:
+		return false;
+	}
+}
+
+static void vop_enable(struct drm_crtc *crtc)
+{
+	struct vop *vop = to_vop(crtc);
+	int ret;
+
+	ret = clk_enable(vop->hclk);
+	if (ret < 0) {
+		dev_err(vop->dev, "failed to enable hclk - %d\n", ret);
+		return;
+	}
+
+	ret = clk_enable(vop->dclk);
+	if (ret < 0) {
+		dev_err(vop->dev, "failed to enable dclk - %d\n", ret);
+		goto err_disable_hclk;
+	}
+
+	ret = clk_enable(vop->aclk);
+	if (ret < 0) {
+		dev_err(vop->dev, "failed to enable aclk - %d\n", ret);
+		goto err_disable_dclk;
+	}
+
+	/*
+	 * Slave iommu shares power, irq and clock with vop.  It was associated
+	 * automatically with this master device via common driver code.
+	 * Now that we have enabled the clock we attach it to the shared drm
+	 * mapping.
+	 */
+	ret = rockchip_drm_dma_attach_device(vop->drm_dev, vop->dev);
+	if (ret) {
+		dev_err(vop->dev, "failed to attach dma mapping, %d\n", ret);
+		goto err_disable_aclk;
+	}
+
+	spin_lock(&vop->reg_lock);
+
+	VOP_CTRL_SET(vop, standby, 0);
+
+	spin_unlock(&vop->reg_lock);
+
+	enable_irq(vop->irq);
+
+	drm_vblank_on(vop->drm_dev, vop->pipe);
+
+	return;
+
+err_disable_aclk:
+	clk_disable(vop->aclk);
+err_disable_dclk:
+	clk_disable(vop->dclk);
+err_disable_hclk:
+	clk_disable(vop->hclk);
+}
+
+static void vop_disable(struct drm_crtc *crtc)
+{
+	struct vop *vop = to_vop(crtc);
+
+	drm_vblank_off(crtc->dev, vop->pipe);
+
+	disable_irq(vop->irq);
+
+	/*
+	 * TODO: Since standby doesn't take effect until the next vblank,
+	 * when we turn off dclk below, the vop is probably still active.
+	 */
+	spin_lock(&vop->reg_lock);
+
+	VOP_CTRL_SET(vop, standby, 1);
+
+	spin_unlock(&vop->reg_lock);
+	/*
+	 * disable dclk to stop frame scan, so we can safely detach iommu,
+	 */
+	clk_disable(vop->dclk);
+
+	rockchip_drm_dma_detach_device(vop->drm_dev, vop->dev);
+
+	clk_disable(vop->aclk);
+	clk_disable(vop->hclk);
+}
+
+/*
+ * Caller must hold vsync_mutex.
+ */
+static struct drm_framebuffer *vop_win_last_pending_fb(struct vop_win *vop_win)
+{
+	struct vop_win_state *last;
+	struct vop_win_state *active = vop_win->active;
+
+	if (list_empty(&vop_win->pending))
+		return active ? active->fb : NULL;
+
+	last = list_last_entry(&vop_win->pending, struct vop_win_state, head);
+	return last ? last->fb : NULL;
+}
+
+/*
+ * Caller must hold vsync_mutex.
+ */
+static int vop_win_queue_fb(struct vop_win *vop_win,
+			    struct drm_framebuffer *fb, dma_addr_t yrgb_mst,
+			    struct drm_pending_vblank_event *event)
+{
+	struct vop_win_state *state;
+
+	state = kzalloc(sizeof(*state), GFP_KERNEL);
+	if (!state)
+		return -ENOMEM;
+
+	state->fb = fb;
+	state->yrgb_mst = yrgb_mst;
+	state->event = event;
+
+	list_add_tail(&state->head, &vop_win->pending);
+
+	return 0;
+}
+
+static int vop_update_plane_event(struct drm_plane *plane,
+				  struct drm_crtc *crtc,
+				  struct drm_framebuffer *fb, int crtc_x,
+				  int crtc_y, unsigned int crtc_w,
+				  unsigned int crtc_h, uint32_t src_x,
+				  uint32_t src_y, uint32_t src_w,
+				  uint32_t src_h,
+				  struct drm_pending_vblank_event *event)
+{
+	struct vop_win *vop_win = to_vop_win(plane);
+	const struct vop_win_data *win = vop_win->data;
+	struct vop *vop = to_vop(crtc);
+	struct drm_gem_object *obj;
+	struct rockchip_gem_object *rk_obj;
+	unsigned long offset;
+	unsigned int actual_w;
+	unsigned int actual_h;
+	unsigned int dsp_stx;
+	unsigned int dsp_sty;
+	unsigned int y_vir_stride;
+	dma_addr_t yrgb_mst;
+	enum vop_data_format format;
+	uint32_t val;
+	bool is_alpha;
+	bool visible;
+	int ret;
+	struct drm_rect dest = {
+		.x1 = crtc_x,
+		.y1 = crtc_y,
+		.x2 = crtc_x + crtc_w,
+		.y2 = crtc_y + crtc_h,
+	};
+	struct drm_rect src = {
+		/* 16.16 fixed point */
+		.x1 = src_x,
+		.y1 = src_y,
+		.x2 = src_x + src_w,
+		.y2 = src_y + src_h,
+	};
+	const struct drm_rect clip = {
+		.x2 = crtc->mode.hdisplay,
+		.y2 = crtc->mode.vdisplay,
+	};
+	bool can_position = plane->type != DRM_PLANE_TYPE_PRIMARY;
+
+	ret = drm_plane_helper_check_update(plane, crtc, fb,
+					    &src, &dest, &clip,
+					    DRM_PLANE_HELPER_NO_SCALING,
+					    DRM_PLANE_HELPER_NO_SCALING,
+					    can_position, false, &visible);
+	if (ret)
+		return ret;
+
+	if (!visible)
+		return 0;
+
+	is_alpha = is_alpha_support(fb->pixel_format);
+	format = vop_convert_format(fb->pixel_format);
+	if (format < 0)
+		return format;
+
+	obj = rockchip_fb_get_gem_obj(fb, 0);
+	if (!obj) {
+		DRM_ERROR("fail to get rockchip gem object from framebuffer\n");
+		return -EINVAL;
+	}
+
+	rk_obj = to_rockchip_obj(obj);
+
+	actual_w = (src.x2 - src.x1) >> 16;
+	actual_h = (src.y2 - src.y1) >> 16;
+	crtc_x = max(0, crtc_x);
+	crtc_y = max(0, crtc_y);
+
+	dsp_stx = crtc_x + crtc->mode.htotal - crtc->mode.hsync_start;
+	dsp_sty = crtc_y + crtc->mode.vtotal - crtc->mode.vsync_start;
+
+	offset = (src.x1 >> 16) * (fb->bits_per_pixel >> 3);
+	offset += (src.y1 >> 16) * fb->pitches[0];
+	yrgb_mst = rk_obj->dma_addr + offset;
+
+	y_vir_stride = fb->pitches[0] / (fb->bits_per_pixel >> 3);
+
+	/*
+	 * If this plane update changes the plane's framebuffer, (or more
+	 * precisely, if this update has a different framebuffer than the last
+	 * update), enqueue it so we can track when it completes.
+	 *
+	 * Only when we discover that this update has completed, can we
+	 * unreference any previous framebuffers.
+	 */
+	mutex_lock(&vop->vsync_mutex);
+	if (fb != vop_win_last_pending_fb(vop_win)) {
+		ret = drm_vblank_get(plane->dev, vop->pipe);
+		if (ret) {
+			DRM_ERROR("failed to get vblank, %d\n", ret);
+			mutex_unlock(&vop->vsync_mutex);
+			return ret;
+		}
+
+		drm_framebuffer_reference(fb);
+
+		ret = vop_win_queue_fb(vop_win, fb, yrgb_mst, event);
+		if (ret) {
+			drm_vblank_put(plane->dev, vop->pipe);
+			mutex_unlock(&vop->vsync_mutex);
+			return ret;
+		}
+
+		vop->vsync_work_pending = true;
+	}
+	mutex_unlock(&vop->vsync_mutex);
+
+	spin_lock(&vop->reg_lock);
+
+	VOP_WIN_SET(vop, win, format, format);
+	VOP_WIN_SET(vop, win, yrgb_vir, y_vir_stride);
+	VOP_WIN_SET(vop, win, yrgb_mst, yrgb_mst);
+	val = (actual_h - 1) << 16;
+	val |= (actual_w - 1) & 0xffff;
+	VOP_WIN_SET(vop, win, act_info, val);
+	VOP_WIN_SET(vop, win, dsp_info, val);
+	val = (dsp_sty - 1) << 16;
+	val |= (dsp_stx - 1) & 0xffff;
+	VOP_WIN_SET(vop, win, dsp_st, val);
+
+	if (is_alpha) {
+		VOP_WIN_SET(vop, win, dst_alpha_ctl,
+			    DST_FACTOR_M0(ALPHA_SRC_INVERSE));
+		val = SRC_ALPHA_EN(1) | SRC_COLOR_M0(ALPHA_SRC_PRE_MUL) |
+			SRC_ALPHA_M0(ALPHA_STRAIGHT) |
+			SRC_BLEND_M0(ALPHA_PER_PIX) |
+			SRC_ALPHA_CAL_M0(ALPHA_NO_SATURATION) |
+			SRC_FACTOR_M0(ALPHA_ONE);
+		VOP_WIN_SET(vop, win, src_alpha_ctl, val);
+	} else {
+		VOP_WIN_SET(vop, win, src_alpha_ctl, SRC_ALPHA_EN(0));
+	}
+
+	VOP_WIN_SET(vop, win, enable, 1);
+
+	vop_cfg_done(vop);
+	spin_unlock(&vop->reg_lock);
+
+	return 0;
+}
+
+static int vop_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
+			    struct drm_framebuffer *fb, int crtc_x, int crtc_y,
+			    unsigned int crtc_w, unsigned int crtc_h,
+			    uint32_t src_x, uint32_t src_y, uint32_t src_w,
+			    uint32_t src_h)
+{
+	return vop_update_plane_event(plane, crtc, fb, crtc_x, crtc_y, crtc_w,
+				      crtc_h, src_x, src_y, src_w, src_h,
+				      NULL);
+}
+
+static int vop_update_primary_plane(struct drm_crtc *crtc,
+				    struct drm_pending_vblank_event *event)
+{
+	unsigned int crtc_w, crtc_h;
+
+	crtc_w = crtc->primary->fb->width - crtc->x;
+	crtc_h = crtc->primary->fb->height - crtc->y;
+
+	return vop_update_plane_event(crtc->primary, crtc, crtc->primary->fb,
+				      0, 0, crtc_w, crtc_h, crtc->x << 16,
+				      crtc->y << 16, crtc_w << 16,
+				      crtc_h << 16, event);
+}
+
+static int vop_disable_plane(struct drm_plane *plane)
+{
+	struct vop_win *vop_win = to_vop_win(plane);
+	const struct vop_win_data *win = vop_win->data;
+	struct vop *vop;
+	int ret;
+
+	if (!plane->crtc)
+		return 0;
+
+	vop = to_vop(plane->crtc);
+
+	ret = drm_vblank_get(plane->dev, vop->pipe);
+	if (ret) {
+		DRM_ERROR("failed to get vblank, %d\n", ret);
+		return ret;
+	}
+
+	mutex_lock(&vop->vsync_mutex);
+
+	ret = vop_win_queue_fb(vop_win, NULL, 0, NULL);
+	if (ret) {
+		drm_vblank_put(plane->dev, vop->pipe);
+		mutex_unlock(&vop->vsync_mutex);
+		return ret;
+	}
+
+	vop->vsync_work_pending = true;
+	mutex_unlock(&vop->vsync_mutex);
+
+	spin_lock(&vop->reg_lock);
+	VOP_WIN_SET(vop, win, enable, 0);
+	vop_cfg_done(vop);
+	spin_unlock(&vop->reg_lock);
+
+	return 0;
+}
+
+static void vop_plane_destroy(struct drm_plane *plane)
+{
+	vop_disable_plane(plane);
+	drm_plane_cleanup(plane);
+}
+
+static const struct drm_plane_funcs vop_plane_funcs = {
+	.update_plane = vop_update_plane,
+	.disable_plane = vop_disable_plane,
+	.destroy = vop_plane_destroy,
+};
+
+int rockchip_drm_crtc_mode_config(struct drm_crtc *crtc,
+				  int connector_type,
+				  int out_mode)
+{
+	struct vop *vop = to_vop(crtc);
+
+	vop->connector_type = connector_type;
+	vop->connector_out_mode = out_mode;
+
+	return 0;
+}
+
+static int vop_crtc_enable_vblank(struct drm_crtc *crtc)
+{
+	struct vop *vop = to_vop(crtc);
+	unsigned long flags;
+
+	if (vop->dpms != DRM_MODE_DPMS_ON)
+		return -EPERM;
+
+	spin_lock_irqsave(&vop->irq_lock, flags);
+
+	vop_mask_write(vop, INTR_CTRL0, FS_INTR_MASK, FS_INTR_EN(1));
+
+	spin_unlock_irqrestore(&vop->irq_lock, flags);
+
+	return 0;
+}
+
+static void vop_crtc_disable_vblank(struct drm_crtc *crtc)
+{
+	struct vop *vop = to_vop(crtc);
+	unsigned long flags;
+
+	if (vop->dpms != DRM_MODE_DPMS_ON)
+		return;
+	spin_lock_irqsave(&vop->irq_lock, flags);
+	vop_mask_write(vop, INTR_CTRL0, FS_INTR_MASK, FS_INTR_EN(0));
+	spin_unlock_irqrestore(&vop->irq_lock, flags);
+}
+
+static const struct rockchip_crtc_funcs private_crtc_funcs = {
+	.enable_vblank = vop_crtc_enable_vblank,
+	.disable_vblank = vop_crtc_disable_vblank,
+};
+
+static void vop_crtc_dpms(struct drm_crtc *crtc, int mode)
+{
+	struct vop *vop = to_vop(crtc);
+
+	DRM_DEBUG_KMS("crtc[%d] mode[%d]\n", crtc->base.id, mode);
+
+	if (vop->dpms == mode) {
+		DRM_DEBUG_KMS("desired dpms mode is same as previous one.\n");
+		return;
+	}
+
+	switch (mode) {
+	case DRM_MODE_DPMS_ON:
+		vop_enable(crtc);
+		break;
+	case DRM_MODE_DPMS_STANDBY:
+	case DRM_MODE_DPMS_SUSPEND:
+	case DRM_MODE_DPMS_OFF:
+		vop_disable(crtc);
+		break;
+	default:
+		DRM_DEBUG_KMS("unspecified mode %d\n", mode);
+		break;
+	}
+
+	vop->dpms = mode;
+}
+
+static void vop_crtc_prepare(struct drm_crtc *crtc)
+{
+	vop_crtc_dpms(crtc, DRM_MODE_DPMS_ON);
+}
+
+static bool vop_crtc_mode_fixup(struct drm_crtc *crtc,
+				const struct drm_display_mode *mode,
+				struct drm_display_mode *adjusted_mode)
+{
+	if (adjusted_mode->htotal == 0 || adjusted_mode->vtotal == 0)
+		return false;
+
+	return true;
+}
+
+static int vop_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
+				  struct drm_framebuffer *old_fb)
+{
+	int ret;
+
+	crtc->x = x;
+	crtc->y = y;
+
+	ret = vop_update_primary_plane(crtc, NULL);
+	if (ret < 0) {
+		DRM_ERROR("fail to update plane\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int vop_crtc_mode_set(struct drm_crtc *crtc,
+			     struct drm_display_mode *mode,
+			     struct drm_display_mode *adjusted_mode,
+			     int x, int y, struct drm_framebuffer *fb)
+{
+	struct vop *vop = to_vop(crtc);
+	u16 hsync_len = adjusted_mode->hsync_end - adjusted_mode->hsync_start;
+	u16 hdisplay = adjusted_mode->hdisplay;
+	u16 htotal = adjusted_mode->htotal;
+	u16 hact_st = adjusted_mode->htotal - adjusted_mode->hsync_start;
+	u16 hact_end = hact_st + hdisplay;
+	u16 vdisplay = adjusted_mode->vdisplay;
+	u16 vtotal = adjusted_mode->vtotal;
+	u16 vsync_len = adjusted_mode->vsync_end - adjusted_mode->vsync_start;
+	u16 vact_st = adjusted_mode->vtotal - adjusted_mode->vsync_start;
+	u16 vact_end = vact_st + vdisplay;
+	int ret;
+	uint32_t val;
+
+	/*
+	 * disable dclk to stop frame scan, so that we can safe config mode and
+	 * enable iommu.
+	 */
+	clk_disable(vop->dclk);
+
+	switch (vop->connector_type) {
+	case DRM_MODE_CONNECTOR_LVDS:
+		VOP_CTRL_SET(vop, rgb_en, 1);
+		break;
+	case DRM_MODE_CONNECTOR_eDP:
+		VOP_CTRL_SET(vop, edp_en, 1);
+		break;
+	case DRM_MODE_CONNECTOR_HDMIA:
+		VOP_CTRL_SET(vop, hdmi_en, 1);
+		break;
+	default:
+		DRM_ERROR("unsupport connector_type[%d]\n",
+			  vop->connector_type);
+		return -EINVAL;
+	};
+	VOP_CTRL_SET(vop, out_mode, vop->connector_out_mode);
+
+	val = 0x8;
+	val |= (adjusted_mode->flags & DRM_MODE_FLAG_NHSYNC) ? 1 : 0;
+	val |= (adjusted_mode->flags & DRM_MODE_FLAG_NVSYNC) ? (1 << 1) : 0;
+	VOP_CTRL_SET(vop, pin_pol, val);
+
+	VOP_CTRL_SET(vop, htotal_pw, (htotal << 16) | hsync_len);
+	val = hact_st << 16;
+	val |= hact_end;
+	VOP_CTRL_SET(vop, hact_st_end, val);
+	VOP_CTRL_SET(vop, hpost_st_end, val);
+
+	VOP_CTRL_SET(vop, vtotal_pw, (vtotal << 16) | vsync_len);
+	val = vact_st << 16;
+	val |= vact_end;
+	VOP_CTRL_SET(vop, vact_st_end, val);
+	VOP_CTRL_SET(vop, vpost_st_end, val);
+
+	ret = vop_crtc_mode_set_base(crtc, x, y, fb);
+	if (ret)
+		return ret;
+
+	/*
+	 * reset dclk, take all mode config affect, so the clk would run in
+	 * correct frame.
+	 */
+	reset_control_assert(vop->dclk_rst);
+	usleep_range(10, 20);
+	reset_control_deassert(vop->dclk_rst);
+
+	clk_set_rate(vop->dclk, adjusted_mode->clock * 1000);
+	ret = clk_enable(vop->dclk);
+	if (ret < 0) {
+		dev_err(vop->dev, "failed to enable dclk - %d\n", ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+static void vop_crtc_commit(struct drm_crtc *crtc)
+{
+}
+
+static const struct drm_crtc_helper_funcs vop_crtc_helper_funcs = {
+	.dpms = vop_crtc_dpms,
+	.prepare = vop_crtc_prepare,
+	.mode_fixup = vop_crtc_mode_fixup,
+	.mode_set = vop_crtc_mode_set,
+	.mode_set_base = vop_crtc_mode_set_base,
+	.commit = vop_crtc_commit,
+};
+
+static int vop_crtc_page_flip(struct drm_crtc *crtc,
+			      struct drm_framebuffer *fb,
+			      struct drm_pending_vblank_event *event,
+			      uint32_t page_flip_flags)
+{
+	struct vop *vop = to_vop(crtc);
+	struct drm_framebuffer *old_fb = crtc->primary->fb;
+	int ret;
+
+	/* when the page flip is requested, crtc's dpms should be on */
+	if (vop->dpms > DRM_MODE_DPMS_ON) {
+		DRM_DEBUG("failed page flip request at dpms[%d].\n", vop->dpms);
+		return 0;
+	}
+
+	crtc->primary->fb = fb;
+
+	ret = vop_update_primary_plane(crtc, event);
+	if (ret)
+		crtc->primary->fb = old_fb;
+
+	return ret;
+}
+
+static void vop_win_state_complete(struct vop_win *vop_win,
+				   struct vop_win_state *state)
+{
+	struct vop *vop = vop_win->vop;
+	struct drm_crtc *crtc = &vop->crtc;
+	struct drm_device *drm = crtc->dev;
+	unsigned long flags;
+
+	if (state->event) {
+		spin_lock_irqsave(&drm->event_lock, flags);
+		drm_send_vblank_event(drm, -1, state->event);
+		spin_unlock_irqrestore(&drm->event_lock, flags);
+	}
+
+	list_del(&state->head);
+	drm_vblank_put(crtc->dev, vop->pipe);
+}
+
+static void vop_crtc_destroy(struct drm_crtc *crtc)
+{
+	drm_crtc_cleanup(crtc);
+}
+
+static const struct drm_crtc_funcs vop_crtc_funcs = {
+	.set_config = drm_crtc_helper_set_config,
+	.page_flip = vop_crtc_page_flip,
+	.destroy = vop_crtc_destroy,
+};
+
+static bool vop_win_state_is_active(struct vop_win *vop_win,
+				    struct vop_win_state *state)
+{
+	bool active = false;
+
+	if (state->fb) {
+		dma_addr_t yrgb_mst;
+
+		/* check yrgb_mst to tell if pending_fb is now front */
+		yrgb_mst = VOP_WIN_GET_YRGBADDR(vop_win->vop, vop_win->data);
+
+		active = (yrgb_mst == state->yrgb_mst);
+	} else {
+		bool enabled;
+
+		/* if enable bit is clear, plane is now disabled */
+		enabled = VOP_WIN_GET(vop_win->vop, vop_win->data, enable);
+
+		active = (enabled == 0);
+	}
+
+	return active;
+}
+
+static void vop_win_state_destroy(struct vop_win_state *state)
+{
+	struct drm_framebuffer *fb = state->fb;
+
+	if (fb)
+		drm_framebuffer_unreference(fb);
+
+	kfree(state);
+}
+
+static void vop_win_update_state(struct vop_win *vop_win)
+{
+	struct vop_win_state *state, *n, *new_active = NULL;
+
+	/* Check if any pending states are now active */
+	list_for_each_entry(state, &vop_win->pending, head)
+		if (vop_win_state_is_active(vop_win, state)) {
+			new_active = state;
+			break;
+		}
+
+	if (!new_active)
+		return;
+
+	/*
+	 * Destroy any 'skipped' pending states - states that were queued
+	 * before the newly active state.
+	 */
+	list_for_each_entry_safe(state, n, &vop_win->pending, head) {
+		if (state == new_active)
+			break;
+		vop_win_state_complete(vop_win, state);
+		vop_win_state_destroy(state);
+	}
+
+	vop_win_state_complete(vop_win, new_active);
+
+	if (vop_win->active)
+		vop_win_state_destroy(vop_win->active);
+	vop_win->active = new_active;
+}
+
+static bool vop_win_has_pending_state(struct vop_win *vop_win)
+{
+	return !list_empty(&vop_win->pending);
+}
+
+static irqreturn_t vop_isr_thread(int irq, void *data)
+{
+	struct vop *vop = data;
+	const struct vop_data *vop_data = vop->data;
+	unsigned int i;
+
+	mutex_lock(&vop->vsync_mutex);
+
+	if (!vop->vsync_work_pending)
+		goto done;
+
+	vop->vsync_work_pending = false;
+
+	for (i = 0; i < vop_data->win_size; i++) {
+		struct vop_win *vop_win = &vop->win[i];
+
+		vop_win_update_state(vop_win);
+		if (vop_win_has_pending_state(vop_win))
+			vop->vsync_work_pending = true;
+	}
+
+done:
+	mutex_unlock(&vop->vsync_mutex);
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t vop_isr(int irq, void *data)
+{
+	struct vop *vop = data;
+	uint32_t intr0_reg, active_irqs;
+	unsigned long flags;
+
+	/*
+	 * INTR_CTRL0 register has interrupt status, enable and clear bits, we
+	 * must hold irq_lock to avoid a race with enable/disable_vblank().
+	*/
+	spin_lock_irqsave(&vop->irq_lock, flags);
+	intr0_reg = vop_readl(vop, INTR_CTRL0);
+	active_irqs = intr0_reg & INTR_MASK;
+	/* Clear all active interrupt sources */
+	if (active_irqs)
+		vop_writel(vop, INTR_CTRL0,
+			   intr0_reg | (active_irqs << INTR_CLR_SHIFT));
+	spin_unlock_irqrestore(&vop->irq_lock, flags);
+
+	/* This is expected for vop iommu irqs, since the irq is shared */
+	if (!active_irqs)
+		return IRQ_NONE;
+
+	/* Only Frame Start Interrupt is enabled; other irqs are spurious. */
+	if (!(active_irqs & FS_INTR)) {
+		DRM_ERROR("Unknown VOP IRQs: %#02x\n", active_irqs);
+		return IRQ_NONE;
+	}
+
+	drm_handle_vblank(vop->drm_dev, vop->pipe);
+
+	return (vop->vsync_work_pending) ? IRQ_WAKE_THREAD : IRQ_HANDLED;
+}
+
+static int vop_create_crtc(struct vop *vop)
+{
+	const struct vop_data *vop_data = vop->data;
+	struct device *dev = vop->dev;
+	struct drm_device *drm_dev = vop->drm_dev;
+	struct drm_plane *primary = NULL, *cursor = NULL, *plane;
+	struct drm_crtc *crtc = &vop->crtc;
+	struct device_node *port;
+	int ret;
+	int i;
+
+	/*
+	 * Create drm_plane for primary and cursor planes first, since we need
+	 * to pass them to drm_crtc_init_with_planes, which sets the
+	 * "possible_crtcs" to the newly initialized crtc.
+	 */
+	for (i = 0; i < vop_data->win_size; i++) {
+		struct vop_win *vop_win = &vop->win[i];
+		const struct vop_win_data *win_data = vop_win->data;
+
+		if (win_data->type != DRM_PLANE_TYPE_PRIMARY &&
+		    win_data->type != DRM_PLANE_TYPE_CURSOR)
+			continue;
+
+		ret = drm_universal_plane_init(vop->drm_dev, &vop_win->base,
+					       0, &vop_plane_funcs,
+					       win_data->phy->data_formats,
+					       win_data->phy->nformats,
+					       win_data->type);
+		if (ret) {
+			DRM_ERROR("failed to initialize plane\n");
+			goto err_cleanup_planes;
+		}
+
+		plane = &vop_win->base;
+		if (plane->type == DRM_PLANE_TYPE_PRIMARY)
+			primary = plane;
+		else if (plane->type == DRM_PLANE_TYPE_CURSOR)
+			cursor = plane;
+	}
+
+	ret = drm_crtc_init_with_planes(drm_dev, crtc, primary, cursor,
+					&vop_crtc_funcs);
+	if (ret)
+		return ret;
+
+	drm_crtc_helper_add(crtc, &vop_crtc_helper_funcs);
+
+	/*
+	 * Create drm_planes for overlay windows with possible_crtcs restricted
+	 * to the newly created crtc.
+	 */
+	for (i = 0; i < vop_data->win_size; i++) {
+		struct vop_win *vop_win = &vop->win[i];
+		const struct vop_win_data *win_data = vop_win->data;
+		unsigned long possible_crtcs = 1 << drm_crtc_index(crtc);
+
+		if (win_data->type != DRM_PLANE_TYPE_OVERLAY)
+			continue;
+
+		ret = drm_universal_plane_init(vop->drm_dev, &vop_win->base,
+					       possible_crtcs,
+					       &vop_plane_funcs,
+					       win_data->phy->data_formats,
+					       win_data->phy->nformats,
+					       win_data->type);
+		if (ret) {
+			DRM_ERROR("failed to initialize overlay plane\n");
+			goto err_cleanup_crtc;
+		}
+	}
+
+	port = of_get_child_by_name(dev->of_node, "port");
+	if (!port) {
+		DRM_ERROR("no port node found in %s\n",
+			  dev->of_node->full_name);
+		goto err_cleanup_crtc;
+	}
+
+	crtc->port = port;
+	vop->pipe = drm_crtc_index(crtc);
+	rockchip_register_crtc_funcs(drm_dev, &private_crtc_funcs, vop->pipe);
+
+	return 0;
+
+err_cleanup_crtc:
+	drm_crtc_cleanup(crtc);
+err_cleanup_planes:
+	list_for_each_entry(plane, &drm_dev->mode_config.plane_list, head)
+		drm_plane_cleanup(plane);
+	return ret;
+}
+
+static void vop_destroy_crtc(struct vop *vop)
+{
+	struct drm_crtc *crtc = &vop->crtc;
+
+	rockchip_unregister_crtc_funcs(vop->drm_dev, vop->pipe);
+	of_node_put(crtc->port);
+	drm_crtc_cleanup(crtc);
+}
+
+static int vop_initial(struct vop *vop)
+{
+	const struct vop_data *vop_data = vop->data;
+	const struct vop_reg_data *init_table = vop_data->init_table;
+	struct reset_control *ahb_rst;
+	int i, ret;
+
+	vop->hclk = devm_clk_get(vop->dev, "hclk_vop");
+	if (IS_ERR(vop->hclk)) {
+		dev_err(vop->dev, "failed to get hclk source\n");
+		return PTR_ERR(vop->hclk);
+	}
+	vop->aclk = devm_clk_get(vop->dev, "aclk_vop");
+	if (IS_ERR(vop->aclk)) {
+		dev_err(vop->dev, "failed to get aclk source\n");
+		return PTR_ERR(vop->aclk);
+	}
+	vop->dclk = devm_clk_get(vop->dev, "dclk_vop");
+	if (IS_ERR(vop->dclk)) {
+		dev_err(vop->dev, "failed to get dclk source\n");
+		return PTR_ERR(vop->dclk);
+	}
+
+	ret = clk_prepare(vop->hclk);
+	if (ret < 0) {
+		dev_err(vop->dev, "failed to prepare hclk\n");
+		return ret;
+	}
+
+	ret = clk_prepare(vop->dclk);
+	if (ret < 0) {
+		dev_err(vop->dev, "failed to prepare dclk\n");
+		goto err_unprepare_hclk;
+	}
+
+	ret = clk_prepare(vop->aclk);
+	if (ret < 0) {
+		dev_err(vop->dev, "failed to prepare aclk\n");
+		goto err_unprepare_dclk;
+	}
+
+	/*
+	 * enable hclk, so that we can config vop register.
+	 */
+	ret = clk_enable(vop->hclk);
+	if (ret < 0) {
+		dev_err(vop->dev, "failed to prepare aclk\n");
+		goto err_unprepare_aclk;
+	}
+	/*
+	 * do hclk_reset, reset all vop registers.
+	 */
+	ahb_rst = devm_reset_control_get(vop->dev, "ahb");
+	if (IS_ERR(ahb_rst)) {
+		dev_err(vop->dev, "failed to get ahb reset\n");
+		ret = PTR_ERR(ahb_rst);
+		goto err_disable_hclk;
+	}
+	reset_control_assert(ahb_rst);
+	usleep_range(10, 20);
+	reset_control_deassert(ahb_rst);
+
+	memcpy(vop->regsbak, vop->regs, vop->len);
+
+	for (i = 0; i < vop_data->table_size; i++)
+		vop_writel(vop, init_table[i].offset, init_table[i].value);
+
+	for (i = 0; i < vop_data->win_size; i++) {
+		const struct vop_win_data *win = &vop_data->win[i];
+
+		VOP_WIN_SET(vop, win, enable, 0);
+	}
+
+	vop_cfg_done(vop);
+
+	/*
+	 * do dclk_reset, let all config take affect.
+	 */
+	vop->dclk_rst = devm_reset_control_get(vop->dev, "dclk");
+	if (IS_ERR(vop->dclk_rst)) {
+		dev_err(vop->dev, "failed to get dclk reset\n");
+		ret = PTR_ERR(vop->dclk_rst);
+		goto err_unprepare_aclk;
+	}
+	reset_control_assert(vop->dclk_rst);
+	usleep_range(10, 20);
+	reset_control_deassert(vop->dclk_rst);
+
+	clk_disable(vop->hclk);
+
+	vop->dpms = DRM_MODE_DPMS_OFF;
+
+	return 0;
+
+err_disable_hclk:
+	clk_disable(vop->hclk);
+err_unprepare_aclk:
+	clk_unprepare(vop->aclk);
+err_unprepare_dclk:
+	clk_unprepare(vop->dclk);
+err_unprepare_hclk:
+	clk_unprepare(vop->hclk);
+	return ret;
+}
+
+/*
+ * Initialize the vop->win array elements.
+ */
+static void vop_win_init(struct vop *vop)
+{
+	const struct vop_data *vop_data = vop->data;
+	unsigned int i;
+
+	for (i = 0; i < vop_data->win_size; i++) {
+		struct vop_win *vop_win = &vop->win[i];
+		const struct vop_win_data *win_data = &vop_data->win[i];
+
+		vop_win->data = win_data;
+		vop_win->vop = vop;
+		INIT_LIST_HEAD(&vop_win->pending);
+	}
+}
+
+static int vop_bind(struct device *dev, struct device *master, void *data)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	const struct of_device_id *of_id;
+	const struct vop_data *vop_data;
+	struct drm_device *drm_dev = data;
+	struct vop *vop;
+	struct resource *res;
+	size_t alloc_size;
+	int ret;
+
+	of_id = of_match_device(vop_driver_dt_match, dev);
+	vop_data = of_id->data;
+	if (!vop_data)
+		return -ENODEV;
+
+	/* Allocate vop struct and its vop_win array */
+	alloc_size = sizeof(*vop) + sizeof(*vop->win) * vop_data->win_size;
+	vop = devm_kzalloc(dev, alloc_size, GFP_KERNEL);
+	if (!vop)
+		return -ENOMEM;
+
+	vop->dev = dev;
+	vop->data = vop_data;
+	vop->drm_dev = drm_dev;
+	dev_set_drvdata(dev, vop);
+
+	vop_win_init(vop);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	vop->len = resource_size(res);
+	vop->regs = devm_ioremap_resource(dev, res);
+	if (IS_ERR(vop->regs))
+		return PTR_ERR(vop->regs);
+
+	vop->regsbak = devm_kzalloc(dev, vop->len, GFP_KERNEL);
+	if (!vop->regsbak)
+		return -ENOMEM;
+
+	ret = vop_initial(vop);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "cannot initial vop dev - err %d\n", ret);
+		return ret;
+	}
+
+	vop->irq = platform_get_irq(pdev, 0);
+	if (vop->irq < 0) {
+		dev_err(dev, "cannot find irq for vop\n");
+		return vop->irq;
+	}
+
+	spin_lock_init(&vop->reg_lock);
+	spin_lock_init(&vop->irq_lock);
+
+	mutex_init(&vop->vsync_mutex);
+
+	ret = devm_request_threaded_irq(dev, vop->irq, vop_isr, vop_isr_thread,
+					IRQF_SHARED, dev_name(dev), vop);
+	if (ret)
+		return ret;
+
+	/* IRQ is initially disabled; it gets enabled in power_on */
+	disable_irq(vop->irq);
+
+	ret = vop_create_crtc(vop);
+	if (ret)
+		return ret;
+
+	pm_runtime_enable(&pdev->dev);
+	return 0;
+}
+
+static void vop_unbind(struct device *dev, struct device *master, void *data)
+{
+	struct vop *vop = dev_get_drvdata(dev);
+
+	pm_runtime_disable(dev);
+	vop_destroy_crtc(vop);
+}
+
+static const struct component_ops vop_component_ops = {
+	.bind = vop_bind,
+	.unbind = vop_unbind,
+};
+
+static int vop_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+
+	if (!dev->of_node) {
+		dev_err(dev, "can't find vop devices\n");
+		return -ENODEV;
+	}
+
+	return component_add(dev, &vop_component_ops);
+}
+
+static int vop_remove(struct platform_device *pdev)
+{
+	component_del(&pdev->dev, &vop_component_ops);
+
+	return 0;
+}
+
+struct platform_driver vop_platform_driver = {
+	.probe = vop_probe,
+	.remove = vop_remove,
+	.driver = {
+		.name = "rockchip-vop",
+		.owner = THIS_MODULE,
+		.of_match_table = of_match_ptr(vop_driver_dt_match),
+	},
+};
+
+module_platform_driver(vop_platform_driver);
+
+MODULE_AUTHOR("Mark Yao <mark.yao@rock-chips.com>");
+MODULE_DESCRIPTION("ROCKCHIP VOP Driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_vop.h b/drivers/gpu/drm/rockchip/rockchip_drm_vop.h
new file mode 100644
index 000000000000..63e9b3a084c5
--- /dev/null
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_vop.h
@@ -0,0 +1,201 @@
+/*
+ * Copyright (C) Fuzhou Rockchip Electronics Co.Ltd
+ * Author:Mark Yao <mark.yao@rock-chips.com>
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#ifndef _ROCKCHIP_DRM_VOP_H
+#define _ROCKCHIP_DRM_VOP_H
+
+/* register definition */
+#define REG_CFG_DONE			0x0000
+#define VERSION_INFO			0x0004
+#define SYS_CTRL			0x0008
+#define SYS_CTRL1			0x000c
+#define DSP_CTRL0			0x0010
+#define DSP_CTRL1			0x0014
+#define DSP_BG				0x0018
+#define MCU_CTRL			0x001c
+#define INTR_CTRL0			0x0020
+#define INTR_CTRL1			0x0024
+#define WIN0_CTRL0			0x0030
+#define WIN0_CTRL1			0x0034
+#define WIN0_COLOR_KEY			0x0038
+#define WIN0_VIR			0x003c
+#define WIN0_YRGB_MST			0x0040
+#define WIN0_CBR_MST			0x0044
+#define WIN0_ACT_INFO			0x0048
+#define WIN0_DSP_INFO			0x004c
+#define WIN0_DSP_ST			0x0050
+#define WIN0_SCL_FACTOR_YRGB		0x0054
+#define WIN0_SCL_FACTOR_CBR		0x0058
+#define WIN0_SCL_OFFSET			0x005c
+#define WIN0_SRC_ALPHA_CTRL		0x0060
+#define WIN0_DST_ALPHA_CTRL		0x0064
+#define WIN0_FADING_CTRL		0x0068
+/* win1 register */
+#define WIN1_CTRL0			0x0070
+#define WIN1_CTRL1			0x0074
+#define WIN1_COLOR_KEY			0x0078
+#define WIN1_VIR			0x007c
+#define WIN1_YRGB_MST			0x0080
+#define WIN1_CBR_MST			0x0084
+#define WIN1_ACT_INFO			0x0088
+#define WIN1_DSP_INFO			0x008c
+#define WIN1_DSP_ST			0x0090
+#define WIN1_SCL_FACTOR_YRGB		0x0094
+#define WIN1_SCL_FACTOR_CBR		0x0098
+#define WIN1_SCL_OFFSET			0x009c
+#define WIN1_SRC_ALPHA_CTRL		0x00a0
+#define WIN1_DST_ALPHA_CTRL		0x00a4
+#define WIN1_FADING_CTRL		0x00a8
+/* win2 register */
+#define WIN2_CTRL0			0x00b0
+#define WIN2_CTRL1			0x00b4
+#define WIN2_VIR0_1			0x00b8
+#define WIN2_VIR2_3			0x00bc
+#define WIN2_MST0			0x00c0
+#define WIN2_DSP_INFO0			0x00c4
+#define WIN2_DSP_ST0			0x00c8
+#define WIN2_COLOR_KEY			0x00cc
+#define WIN2_MST1			0x00d0
+#define WIN2_DSP_INFO1			0x00d4
+#define WIN2_DSP_ST1			0x00d8
+#define WIN2_SRC_ALPHA_CTRL		0x00dc
+#define WIN2_MST2			0x00e0
+#define WIN2_DSP_INFO2			0x00e4
+#define WIN2_DSP_ST2			0x00e8
+#define WIN2_DST_ALPHA_CTRL		0x00ec
+#define WIN2_MST3			0x00f0
+#define WIN2_DSP_INFO3			0x00f4
+#define WIN2_DSP_ST3			0x00f8
+#define WIN2_FADING_CTRL		0x00fc
+/* win3 register */
+#define WIN3_CTRL0			0x0100
+#define WIN3_CTRL1			0x0104
+#define WIN3_VIR0_1			0x0108
+#define WIN3_VIR2_3			0x010c
+#define WIN3_MST0			0x0110
+#define WIN3_DSP_INFO0			0x0114
+#define WIN3_DSP_ST0			0x0118
+#define WIN3_COLOR_KEY			0x011c
+#define WIN3_MST1			0x0120
+#define WIN3_DSP_INFO1			0x0124
+#define WIN3_DSP_ST1			0x0128
+#define WIN3_SRC_ALPHA_CTRL		0x012c
+#define WIN3_MST2			0x0130
+#define WIN3_DSP_INFO2			0x0134
+#define WIN3_DSP_ST2			0x0138
+#define WIN3_DST_ALPHA_CTRL		0x013c
+#define WIN3_MST3			0x0140
+#define WIN3_DSP_INFO3			0x0144
+#define WIN3_DSP_ST3			0x0148
+#define WIN3_FADING_CTRL		0x014c
+/* hwc register */
+#define HWC_CTRL0			0x0150
+#define HWC_CTRL1			0x0154
+#define HWC_MST				0x0158
+#define HWC_DSP_ST			0x015c
+#define HWC_SRC_ALPHA_CTRL		0x0160
+#define HWC_DST_ALPHA_CTRL		0x0164
+#define HWC_FADING_CTRL			0x0168
+/* post process register */
+#define POST_DSP_HACT_INFO		0x0170
+#define POST_DSP_VACT_INFO		0x0174
+#define POST_SCL_FACTOR_YRGB		0x0178
+#define POST_SCL_CTRL			0x0180
+#define POST_DSP_VACT_INFO_F1		0x0184
+#define DSP_HTOTAL_HS_END		0x0188
+#define DSP_HACT_ST_END			0x018c
+#define DSP_VTOTAL_VS_END		0x0190
+#define DSP_VACT_ST_END			0x0194
+#define DSP_VS_ST_END_F1		0x0198
+#define DSP_VACT_ST_END_F1		0x019c
+/* register definition end */
+
+/* interrupt define */
+#define DSP_HOLD_VALID_INTR		(1 << 0)
+#define FS_INTR				(1 << 1)
+#define LINE_FLAG_INTR			(1 << 2)
+#define BUS_ERROR_INTR			(1 << 3)
+
+#define INTR_MASK			(DSP_HOLD_VALID_INTR | FS_INTR | \
+					 LINE_FLAG_INTR | BUS_ERROR_INTR)
+
+#define DSP_HOLD_VALID_INTR_EN(x)	((x) << 4)
+#define FS_INTR_EN(x)			((x) << 5)
+#define LINE_FLAG_INTR_EN(x)		((x) << 6)
+#define BUS_ERROR_INTR_EN(x)		((x) << 7)
+#define DSP_HOLD_VALID_INTR_MASK	(1 << 4)
+#define FS_INTR_MASK			(1 << 5)
+#define LINE_FLAG_INTR_MASK		(1 << 6)
+#define BUS_ERROR_INTR_MASK		(1 << 7)
+
+#define INTR_CLR_SHIFT			8
+#define DSP_HOLD_VALID_INTR_CLR		(1 << (INTR_CLR_SHIFT + 0))
+#define FS_INTR_CLR			(1 << (INTR_CLR_SHIFT + 1))
+#define LINE_FLAG_INTR_CLR		(1 << (INTR_CLR_SHIFT + 2))
+#define BUS_ERROR_INTR_CLR		(1 << (INTR_CLR_SHIFT + 3))
+
+#define DSP_LINE_NUM(x)			(((x) & 0x1fff) << 12)
+#define DSP_LINE_NUM_MASK		(0x1fff << 12)
+
+/* src alpha ctrl define */
+#define SRC_FADING_VALUE(x)		(((x) & 0xff) << 24)
+#define SRC_GLOBAL_ALPHA(x)		(((x) & 0xff) << 16)
+#define SRC_FACTOR_M0(x)		(((x) & 0x7) << 6)
+#define SRC_ALPHA_CAL_M0(x)		(((x) & 0x1) << 5)
+#define SRC_BLEND_M0(x)			(((x) & 0x3) << 3)
+#define SRC_ALPHA_M0(x)			(((x) & 0x1) << 2)
+#define SRC_COLOR_M0(x)			(((x) & 0x1) << 1)
+#define SRC_ALPHA_EN(x)			(((x) & 0x1) << 0)
+/* dst alpha ctrl define */
+#define DST_FACTOR_M0(x)		(((x) & 0x7) << 6)
+
+/*
+ * display output interface supported by rockchip lcdc
+ */
+#define ROCKCHIP_OUT_MODE_P888	0
+#define ROCKCHIP_OUT_MODE_P666	1
+#define ROCKCHIP_OUT_MODE_P565	2
+/* for use special outface */
+#define ROCKCHIP_OUT_MODE_AAAA	15
+
+enum alpha_mode {
+	ALPHA_STRAIGHT,
+	ALPHA_INVERSE,
+};
+
+enum global_blend_mode {
+	ALPHA_GLOBAL,
+	ALPHA_PER_PIX,
+	ALPHA_PER_PIX_GLOBAL,
+};
+
+enum alpha_cal_mode {
+	ALPHA_SATURATION,
+	ALPHA_NO_SATURATION,
+};
+
+enum color_mode {
+	ALPHA_SRC_PRE_MUL,
+	ALPHA_SRC_NO_PRE_MUL,
+};
+
+enum factor_mode {
+	ALPHA_ZERO,
+	ALPHA_ONE,
+	ALPHA_SRC,
+	ALPHA_SRC_INVERSE,
+	ALPHA_SRC_GLOBAL,
+};
+
+#endif /* _ROCKCHIP_DRM_VOP_H */
diff --git a/drivers/gpu/drm/savage/savage_bci.c b/drivers/gpu/drm/savage/savage_bci.c
index c97cdc9ab239..d47dff95fe52 100644
--- a/drivers/gpu/drm/savage/savage_bci.c
+++ b/drivers/gpu/drm/savage/savage_bci.c
@@ -556,7 +556,7 @@ int savage_driver_load(struct drm_device *dev, unsigned long chipset)
 /*
  * Initialize mappings. On Savage4 and SavageIX the alignment
  * and size of the aperture is not suitable for automatic MTRR setup
- * in drm_addmap. Therefore we add them manually before the maps are
+ * in drm_legacy_addmap. Therefore we add them manually before the maps are
  * initialized, and tear them down on last close.
  */
 int savage_driver_firstopen(struct drm_device *dev)
@@ -624,19 +624,20 @@ int savage_driver_firstopen(struct drm_device *dev)
 		/* Automatic MTRR setup will do the right thing. */
 	}
 
-	ret = drm_addmap(dev, mmio_base, SAVAGE_MMIO_SIZE, _DRM_REGISTERS,
-			 _DRM_READ_ONLY, &dev_priv->mmio);
+	ret = drm_legacy_addmap(dev, mmio_base, SAVAGE_MMIO_SIZE,
+				_DRM_REGISTERS, _DRM_READ_ONLY,
+				&dev_priv->mmio);
 	if (ret)
 		return ret;
 
-	ret = drm_addmap(dev, fb_base, fb_size, _DRM_FRAME_BUFFER,
-			 _DRM_WRITE_COMBINING, &dev_priv->fb);
+	ret = drm_legacy_addmap(dev, fb_base, fb_size, _DRM_FRAME_BUFFER,
+				_DRM_WRITE_COMBINING, &dev_priv->fb);
 	if (ret)
 		return ret;
 
-	ret = drm_addmap(dev, aperture_base, SAVAGE_APERTURE_SIZE,
-			 _DRM_FRAME_BUFFER, _DRM_WRITE_COMBINING,
-			 &dev_priv->aperture);
+	ret = drm_legacy_addmap(dev, aperture_base, SAVAGE_APERTURE_SIZE,
+				_DRM_FRAME_BUFFER, _DRM_WRITE_COMBINING,
+				&dev_priv->aperture);
 	return ret;
 }
 
@@ -698,14 +699,14 @@ static int savage_do_init_bci(struct drm_device * dev, drm_savage_init_t * init)
 	dev_priv->texture_offset = init->texture_offset;
 	dev_priv->texture_size = init->texture_size;
 
-	dev_priv->sarea = drm_getsarea(dev);
+	dev_priv->sarea = drm_legacy_getsarea(dev);
 	if (!dev_priv->sarea) {
 		DRM_ERROR("could not find sarea!\n");
 		savage_do_cleanup_bci(dev);
 		return -EINVAL;
 	}
 	if (init->status_offset != 0) {
-		dev_priv->status = drm_core_findmap(dev, init->status_offset);
+		dev_priv->status = drm_legacy_findmap(dev, init->status_offset);
 		if (!dev_priv->status) {
 			DRM_ERROR("could not find shadow status region!\n");
 			savage_do_cleanup_bci(dev);
@@ -716,14 +717,14 @@ static int savage_do_init_bci(struct drm_device * dev, drm_savage_init_t * init)
 	}
 	if (dev_priv->dma_type == SAVAGE_DMA_AGP && init->buffers_offset) {
 		dev->agp_buffer_token = init->buffers_offset;
-		dev->agp_buffer_map = drm_core_findmap(dev,
+		dev->agp_buffer_map = drm_legacy_findmap(dev,
 						       init->buffers_offset);
 		if (!dev->agp_buffer_map) {
 			DRM_ERROR("could not find DMA buffer region!\n");
 			savage_do_cleanup_bci(dev);
 			return -EINVAL;
 		}
-		drm_core_ioremap(dev->agp_buffer_map, dev);
+		drm_legacy_ioremap(dev->agp_buffer_map, dev);
 		if (!dev->agp_buffer_map->handle) {
 			DRM_ERROR("failed to ioremap DMA buffer region!\n");
 			savage_do_cleanup_bci(dev);
@@ -732,7 +733,7 @@ static int savage_do_init_bci(struct drm_device * dev, drm_savage_init_t * init)
 	}
 	if (init->agp_textures_offset) {
 		dev_priv->agp_textures =
-		    drm_core_findmap(dev, init->agp_textures_offset);
+		    drm_legacy_findmap(dev, init->agp_textures_offset);
 		if (!dev_priv->agp_textures) {
 			DRM_ERROR("could not find agp texture region!\n");
 			savage_do_cleanup_bci(dev);
@@ -755,7 +756,7 @@ static int savage_do_init_bci(struct drm_device * dev, drm_savage_init_t * init)
 			savage_do_cleanup_bci(dev);
 			return -EINVAL;
 		}
-		dev_priv->cmd_dma = drm_core_findmap(dev, init->cmd_dma_offset);
+		dev_priv->cmd_dma = drm_legacy_findmap(dev, init->cmd_dma_offset);
 		if (!dev_priv->cmd_dma) {
 			DRM_ERROR("could not find command DMA region!\n");
 			savage_do_cleanup_bci(dev);
@@ -768,7 +769,7 @@ static int savage_do_init_bci(struct drm_device * dev, drm_savage_init_t * init)
 				savage_do_cleanup_bci(dev);
 				return -EINVAL;
 			}
-			drm_core_ioremap(dev_priv->cmd_dma, dev);
+			drm_legacy_ioremap(dev_priv->cmd_dma, dev);
 			if (!dev_priv->cmd_dma->handle) {
 				DRM_ERROR("failed to ioremap command "
 					  "DMA region!\n");
@@ -894,11 +895,11 @@ static int savage_do_cleanup_bci(struct drm_device * dev)
 	} else if (dev_priv->cmd_dma && dev_priv->cmd_dma->handle &&
 		   dev_priv->cmd_dma->type == _DRM_AGP &&
 		   dev_priv->dma_type == SAVAGE_DMA_AGP)
-		drm_core_ioremapfree(dev_priv->cmd_dma, dev);
+		drm_legacy_ioremapfree(dev_priv->cmd_dma, dev);
 
 	if (dev_priv->dma_type == SAVAGE_DMA_AGP &&
 	    dev->agp_buffer_map && dev->agp_buffer_map->handle) {
-		drm_core_ioremapfree(dev->agp_buffer_map, dev);
+		drm_legacy_ioremapfree(dev->agp_buffer_map, dev);
 		/* make sure the next instance (which may be running
 		 * in PCI mode) doesn't try to use an old
 		 * agp_buffer_map. */
@@ -1050,7 +1051,7 @@ void savage_reclaim_buffers(struct drm_device *dev, struct drm_file *file_priv)
 		return;
 
 	if (file_priv->master && file_priv->master->lock.hw_lock) {
-		drm_idlelock_take(&file_priv->master->lock);
+		drm_legacy_idlelock_take(&file_priv->master->lock);
 		release_idlelock = 1;
 	}
 
@@ -1069,7 +1070,7 @@ void savage_reclaim_buffers(struct drm_device *dev, struct drm_file *file_priv)
 	}
 
 	if (release_idlelock)
-		drm_idlelock_release(&file_priv->master->lock);
+		drm_legacy_idlelock_release(&file_priv->master->lock);
 }
 
 const struct drm_ioctl_desc savage_ioctls[] = {
diff --git a/drivers/gpu/drm/savage/savage_drv.c b/drivers/gpu/drm/savage/savage_drv.c
index 3c030216e888..21aed1febeb4 100644
--- a/drivers/gpu/drm/savage/savage_drv.c
+++ b/drivers/gpu/drm/savage/savage_drv.c
@@ -40,7 +40,7 @@ static const struct file_operations savage_driver_fops = {
 	.open = drm_open,
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_mmap,
+	.mmap = drm_legacy_mmap,
 	.poll = drm_poll,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl = drm_compat_ioctl,
@@ -57,6 +57,7 @@ static struct drm_driver driver = {
 	.preclose = savage_reclaim_buffers,
 	.lastclose = savage_driver_lastclose,
 	.unload = savage_driver_unload,
+	.set_busid = drm_pci_set_busid,
 	.ioctls = savage_ioctls,
 	.dma_ioctl = savage_bci_buffers,
 	.fops = &savage_driver_fops,
diff --git a/drivers/gpu/drm/savage/savage_drv.h b/drivers/gpu/drm/savage/savage_drv.h
index 335f8fcf1041..37b699571ad0 100644
--- a/drivers/gpu/drm/savage/savage_drv.h
+++ b/drivers/gpu/drm/savage/savage_drv.h
@@ -26,6 +26,8 @@
 #ifndef __SAVAGE_DRV_H__
 #define __SAVAGE_DRV_H__
 
+#include <drm/drm_legacy.h>
+
 #define DRIVER_AUTHOR	"Felix Kuehling"
 
 #define DRIVER_NAME	"savage"
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_backlight.c b/drivers/gpu/drm/shmobile/shmob_drm_backlight.c
index 463aee18f774..33dd41afea0e 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_backlight.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_backlight.c
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_backlight.c  --  SH Mobile DRM Backlight
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_backlight.h b/drivers/gpu/drm/shmobile/shmob_drm_backlight.h
index 9477595d2ff3..bac719ecc301 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_backlight.h
+++ b/drivers/gpu/drm/shmobile/shmob_drm_backlight.h
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_backlight.h  --  SH Mobile DRM Backlight
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
index 47875de89010..859ccb658601 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_crtc.c  --  SH Mobile DRM CRTCs
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -19,6 +19,7 @@
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_fb_cma_helper.h>
 #include <drm/drm_gem_cma_helper.h>
+#include <drm/drm_plane_helper.h>
 
 #include <video/sh_mobile_meram.h>
 
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_crtc.h b/drivers/gpu/drm/shmobile/shmob_drm_crtc.h
index e5bd109c4c38..eddad6dcc88a 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_crtc.h
+++ b/drivers/gpu/drm/shmobile/shmob_drm_crtc.h
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_crtc.h  --  SH Mobile DRM CRTCs
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_drv.c b/drivers/gpu/drm/shmobile/shmob_drm_drv.c
index ff4ba483b602..666321de7b99 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_drv.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_drv.c
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_drv.c  --  SH Mobile DRM driver
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -267,6 +267,7 @@ static struct drm_driver shmob_drm_driver = {
 	.load			= shmob_drm_load,
 	.unload			= shmob_drm_unload,
 	.preclose		= shmob_drm_preclose,
+	.set_busid		= drm_platform_set_busid,
 	.irq_handler		= shmob_drm_irq,
 	.get_vblank_counter	= drm_vblank_count,
 	.enable_vblank		= shmob_drm_enable_vblank,
@@ -347,7 +348,6 @@ static struct platform_driver shmob_drm_platform_driver = {
 	.probe		= shmob_drm_probe,
 	.remove		= shmob_drm_remove,
 	.driver		= {
-		.owner	= THIS_MODULE,
 		.name	= "shmob-drm",
 		.pm	= &shmob_drm_pm_ops,
 	},
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_drv.h b/drivers/gpu/drm/shmobile/shmob_drm_drv.h
index 4d46b811b5a7..02ea315ba69a 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_drv.h
+++ b/drivers/gpu/drm/shmobile/shmob_drm_drv.h
@@ -1,7 +1,7 @@
 /*
  * shmob_drm.h  --  SH Mobile DRM driver
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_kms.c b/drivers/gpu/drm/shmobile/shmob_drm_kms.c
index fc0ef0ca7d04..aaf98ace4a90 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_kms.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_kms.c
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_kms.c  --  SH Mobile DRM Mode Setting
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_kms.h b/drivers/gpu/drm/shmobile/shmob_drm_kms.h
index 9495c9111308..06d5b7caa026 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_kms.h
+++ b/drivers/gpu/drm/shmobile/shmob_drm_kms.h
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_kms.h  --  SH Mobile DRM Mode Setting
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_plane.c b/drivers/gpu/drm/shmobile/shmob_drm_plane.c
index 060ae03e5f9b..1805bb23b113 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_plane.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_plane.c
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_plane.c  --  SH Mobile DRM Planes
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_plane.h b/drivers/gpu/drm/shmobile/shmob_drm_plane.h
index 99623d05e3b0..a58cc1fc3240 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_plane.h
+++ b/drivers/gpu/drm/shmobile/shmob_drm_plane.h
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_plane.h  --  SH Mobile DRM Planes
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_regs.h b/drivers/gpu/drm/shmobile/shmob_drm_regs.h
index 7923cdd6368e..ea17d4415b9e 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_regs.h
+++ b/drivers/gpu/drm/shmobile/shmob_drm_regs.h
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_regs.h  --  SH Mobile DRM registers
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/sis/sis_drv.c b/drivers/gpu/drm/sis/sis_drv.c
index 756f787b7143..79bce76cb8f7 100644
--- a/drivers/gpu/drm/sis/sis_drv.c
+++ b/drivers/gpu/drm/sis/sis_drv.c
@@ -70,7 +70,7 @@ static const struct file_operations sis_driver_fops = {
 	.open = drm_open,
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_mmap,
+	.mmap = drm_legacy_mmap,
 	.poll = drm_poll,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl = drm_compat_ioctl,
@@ -108,6 +108,7 @@ static struct drm_driver driver = {
 	.open = sis_driver_open,
 	.preclose = sis_reclaim_buffers_locked,
 	.postclose = sis_driver_postclose,
+	.set_busid = drm_pci_set_busid,
 	.dma_quiescent = sis_idle,
 	.lastclose = sis_lastclose,
 	.ioctls = sis_ioctls,
diff --git a/drivers/gpu/drm/sis/sis_drv.h b/drivers/gpu/drm/sis/sis_drv.h
index c31c0253054d..16f972b2a76a 100644
--- a/drivers/gpu/drm/sis/sis_drv.h
+++ b/drivers/gpu/drm/sis/sis_drv.h
@@ -28,6 +28,8 @@
 #ifndef _SIS_DRV_H_
 #define _SIS_DRV_H_
 
+#include <drm/drm_legacy.h>
+
 /* General customization:
  */
 
diff --git a/drivers/gpu/drm/sis/sis_mm.c b/drivers/gpu/drm/sis/sis_mm.c
index 77f288e4a0a6..93ad8a5704d1 100644
--- a/drivers/gpu/drm/sis/sis_mm.c
+++ b/drivers/gpu/drm/sis/sis_mm.c
@@ -319,12 +319,12 @@ void sis_reclaim_buffers_locked(struct drm_device *dev,
 	if (!(file->minor->master && file->master->lock.hw_lock))
 		return;
 
-	drm_idlelock_take(&file->master->lock);
+	drm_legacy_idlelock_take(&file->master->lock);
 
 	mutex_lock(&dev->struct_mutex);
 	if (list_empty(&file_priv->obj_list)) {
 		mutex_unlock(&dev->struct_mutex);
-		drm_idlelock_release(&file->master->lock);
+		drm_legacy_idlelock_release(&file->master->lock);
 
 		return;
 	}
@@ -345,7 +345,7 @@ void sis_reclaim_buffers_locked(struct drm_device *dev,
 	}
 	mutex_unlock(&dev->struct_mutex);
 
-	drm_idlelock_release(&file->master->lock);
+	drm_legacy_idlelock_release(&file->master->lock);
 
 	return;
 }
diff --git a/drivers/gpu/drm/sti/Kconfig b/drivers/gpu/drm/sti/Kconfig
index ae8850f3e63b..d6d6b705b8c1 100644
--- a/drivers/gpu/drm/sti/Kconfig
+++ b/drivers/gpu/drm/sti/Kconfig
@@ -5,6 +5,7 @@ config DRM_STI
 	select DRM_KMS_HELPER
 	select DRM_GEM_CMA_HELPER
 	select DRM_KMS_CMA_HELPER
+	select FW_LOADER_USER_HELPER_FALLBACK
 	help
 	  Choose this option to enable DRM on STM stiH41x chipset
 
diff --git a/drivers/gpu/drm/sti/Makefile b/drivers/gpu/drm/sti/Makefile
index 04ac2ceef27f..6ba9d27c1b90 100644
--- a/drivers/gpu/drm/sti/Makefile
+++ b/drivers/gpu/drm/sti/Makefile
@@ -3,6 +3,7 @@ sticompositor-y := \
 	sti_mixer.o \
 	sti_gdp.o \
 	sti_vid.o \
+	sti_cursor.o \
 	sti_compositor.o \
 	sti_drm_crtc.o \
 	sti_drm_plane.o
@@ -18,4 +19,5 @@ obj-$(CONFIG_DRM_STI) = \
 	sti_hda.o \
 	sti_tvout.o \
 	sticompositor.o \
-	sti_drm_drv.o
-\ No newline at end of file
+	sti_hqvdp.o \
+	sti_drm_drv.o
diff --git a/drivers/gpu/drm/sti/sti_compositor.c b/drivers/gpu/drm/sti/sti_compositor.c
index 390d93e9a06c..43215d3020fb 100644
--- a/drivers/gpu/drm/sti/sti_compositor.c
+++ b/drivers/gpu/drm/sti/sti_compositor.c
@@ -24,14 +24,16 @@
  * stiH407 compositor properties
  */
 struct sti_compositor_data stih407_compositor_data = {
-	.nb_subdev = 6,
+	.nb_subdev = 8,
 	.subdev_desc = {
+			{STI_CURSOR_SUBDEV, (int)STI_CURSOR, 0x000},
 			{STI_GPD_SUBDEV, (int)STI_GDP_0, 0x100},
 			{STI_GPD_SUBDEV, (int)STI_GDP_1, 0x200},
 			{STI_GPD_SUBDEV, (int)STI_GDP_2, 0x300},
 			{STI_GPD_SUBDEV, (int)STI_GDP_3, 0x400},
 			{STI_VID_SUBDEV, (int)STI_VID_0, 0x700},
-			{STI_MIXER_MAIN_SUBDEV, STI_MIXER_MAIN, 0xC00}
+			{STI_MIXER_MAIN_SUBDEV, STI_MIXER_MAIN, 0xC00},
+			{STI_MIXER_AUX_SUBDEV, STI_MIXER_AUX, 0xD00},
 	},
 };
 
@@ -67,11 +69,11 @@ static int sti_compositor_init_subdev(struct sti_compositor *compo,
 			break;
 		case STI_GPD_SUBDEV:
 		case STI_VID_SUBDEV:
+		case STI_CURSOR_SUBDEV:
 			compo->layer[layer_id++] =
 			    sti_layer_create(compo->dev, desc[i].id,
 					     compo->regs + desc[i].offset);
 			break;
-			/* case STI_CURSOR_SUBDEV : TODO */
 		default:
 			DRM_ERROR("Unknow subdev compoment type\n");
 			return 1;
@@ -102,33 +104,35 @@ static int sti_compositor_bind(struct device *dev, struct device *master,
 			enum sti_layer_type type = desc & STI_LAYER_TYPE_MASK;
 			enum drm_plane_type plane_type = DRM_PLANE_TYPE_OVERLAY;
 
-			if (compo->mixer[crtc])
+			if (crtc < compo->nb_mixers)
 				plane_type = DRM_PLANE_TYPE_PRIMARY;
 
 			switch (type) {
 			case STI_CUR:
 				cursor = sti_drm_plane_init(drm_dev,
 						compo->layer[i],
-						(1 << crtc) - 1,
-						DRM_PLANE_TYPE_CURSOR);
+						1, DRM_PLANE_TYPE_CURSOR);
 				break;
 			case STI_GDP:
 			case STI_VID:
 				primary = sti_drm_plane_init(drm_dev,
 						compo->layer[i],
-						(1 << crtc) - 1, plane_type);
+						(1 << compo->nb_mixers) - 1,
+						plane_type);
 				plane++;
 				break;
 			case STI_BCK:
+			case STI_VDP:
 				break;
 			}
 
 			/* The first planes are reserved for primary planes*/
-			if (compo->mixer[crtc]) {
+			if (crtc < compo->nb_mixers && primary) {
 				sti_drm_crtc_init(drm_dev, compo->mixer[crtc],
 						primary, cursor);
 				crtc++;
 				cursor = NULL;
+				primary = NULL;
 			}
 		}
 	}
@@ -267,7 +271,6 @@ static int sti_compositor_remove(struct platform_device *pdev)
 static struct platform_driver sti_compositor_driver = {
 	.driver = {
 		.name = "sti-compositor",
-		.owner = THIS_MODULE,
 		.of_match_table = compositor_of_match,
 	},
 	.probe = sti_compositor_probe,
diff --git a/drivers/gpu/drm/sti/sti_compositor.h b/drivers/gpu/drm/sti/sti_compositor.h
index 3ea19db72e0f..019eb44c62cc 100644
--- a/drivers/gpu/drm/sti/sti_compositor.h
+++ b/drivers/gpu/drm/sti/sti_compositor.h
@@ -64,7 +64,6 @@ struct sti_compositor_data {
  * @layer: array of layers
  * @nb_mixers: number of mixers for this compositor
  * @nb_layers: number of layers (GDP,VID,...) for this compositor
- * @enable: true if compositor is enable else false
  * @vtg_vblank_nb: callback for VTG VSYNC notification
  */
 struct sti_compositor {
@@ -83,7 +82,6 @@ struct sti_compositor {
 	struct sti_layer *layer[STI_MAX_LAYER];
 	int nb_mixers;
 	int nb_layers;
-	bool enable;
 	struct notifier_block vtg_vblank_nb;
 };
 
diff --git a/drivers/gpu/drm/sti/sti_cursor.c b/drivers/gpu/drm/sti/sti_cursor.c
new file mode 100644
index 000000000000..010eaee60bf7
--- /dev/null
+++ b/drivers/gpu/drm/sti/sti_cursor.c
@@ -0,0 +1,242 @@
+/*
+ * Copyright (C) STMicroelectronics SA 2014
+ * Authors: Vincent Abriou <vincent.abriou@st.com>
+ *          Fabien Dessenne <fabien.dessenne@st.com>
+ *          for STMicroelectronics.
+ * License terms:  GNU General Public License (GPL), version 2
+ */
+#include <drm/drmP.h>
+
+#include "sti_cursor.h"
+#include "sti_layer.h"
+#include "sti_vtg.h"
+
+/* Registers */
+#define CUR_CTL             0x00
+#define CUR_VPO             0x0C
+#define CUR_PML             0x14
+#define CUR_PMP             0x18
+#define CUR_SIZE            0x1C
+#define CUR_CML             0x20
+#define CUR_AWS             0x28
+#define CUR_AWE             0x2C
+
+#define CUR_CTL_CLUT_UPDATE BIT(1)
+
+#define STI_CURS_MIN_SIZE   1
+#define STI_CURS_MAX_SIZE   128
+
+/*
+ * pixmap dma buffer stucture
+ *
+ * @paddr:  physical address
+ * @size:   buffer size
+ * @base:   virtual address
+ */
+struct dma_pixmap {
+	dma_addr_t paddr;
+	size_t size;
+	void *base;
+};
+
+/**
+ * STI Cursor structure
+ *
+ * @layer:      layer structure
+ * @width:      cursor width
+ * @height:     cursor height
+ * @clut:       color look up table
+ * @clut_paddr: color look up table physical address
+ * @pixmap:     pixmap dma buffer (clut8-format cursor)
+ */
+struct sti_cursor {
+	struct sti_layer layer;
+	unsigned int width;
+	unsigned int height;
+	unsigned short *clut;
+	dma_addr_t clut_paddr;
+	struct dma_pixmap pixmap;
+};
+
+static const uint32_t cursor_supported_formats[] = {
+	DRM_FORMAT_ARGB8888,
+};
+
+#define to_sti_cursor(x) container_of(x, struct sti_cursor, layer)
+
+static const uint32_t *sti_cursor_get_formats(struct sti_layer *layer)
+{
+	return cursor_supported_formats;
+}
+
+static unsigned int sti_cursor_get_nb_formats(struct sti_layer *layer)
+{
+	return ARRAY_SIZE(cursor_supported_formats);
+}
+
+static void sti_cursor_argb8888_to_clut8(struct sti_layer *layer)
+{
+	struct sti_cursor *cursor = to_sti_cursor(layer);
+	u32 *src = layer->vaddr;
+	u8  *dst = cursor->pixmap.base;
+	unsigned int i, j;
+	u32 a, r, g, b;
+
+	for (i = 0; i < cursor->height; i++) {
+		for (j = 0; j < cursor->width; j++) {
+			/* Pick the 2 higher bits of each component */
+			a = (*src >> 30) & 3;
+			r = (*src >> 22) & 3;
+			g = (*src >> 14) & 3;
+			b = (*src >> 6) & 3;
+			*dst = a << 6 | r << 4 | g << 2 | b;
+			src++;
+			dst++;
+		}
+	}
+}
+
+static int sti_cursor_prepare_layer(struct sti_layer *layer, bool first_prepare)
+{
+	struct sti_cursor *cursor = to_sti_cursor(layer);
+	struct drm_display_mode *mode = layer->mode;
+	u32 y, x;
+	u32 val;
+
+	DRM_DEBUG_DRIVER("\n");
+
+	dev_dbg(layer->dev, "%s %s\n", __func__, sti_layer_to_str(layer));
+
+	if (layer->src_w < STI_CURS_MIN_SIZE ||
+	    layer->src_h < STI_CURS_MIN_SIZE ||
+	    layer->src_w > STI_CURS_MAX_SIZE ||
+	    layer->src_h > STI_CURS_MAX_SIZE) {
+		DRM_ERROR("Invalid cursor size (%dx%d)\n",
+				layer->src_w, layer->src_h);
+		return -EINVAL;
+	}
+
+	/* If the cursor size has changed, re-allocated the pixmap */
+	if (!cursor->pixmap.base ||
+	    (cursor->width != layer->src_w) ||
+	    (cursor->height != layer->src_h)) {
+		cursor->width = layer->src_w;
+		cursor->height = layer->src_h;
+
+		if (cursor->pixmap.base)
+			dma_free_writecombine(layer->dev,
+					      cursor->pixmap.size,
+					      cursor->pixmap.base,
+					      cursor->pixmap.paddr);
+
+		cursor->pixmap.size = cursor->width * cursor->height;
+
+		cursor->pixmap.base = dma_alloc_writecombine(layer->dev,
+							cursor->pixmap.size,
+							&cursor->pixmap.paddr,
+							GFP_KERNEL | GFP_DMA);
+		if (!cursor->pixmap.base) {
+			DRM_ERROR("Failed to allocate memory for pixmap\n");
+			return -ENOMEM;
+		}
+	}
+
+	/* Convert ARGB8888 to CLUT8 */
+	sti_cursor_argb8888_to_clut8(layer);
+
+	/* AWS and AWE depend on the mode */
+	y = sti_vtg_get_line_number(*mode, 0);
+	x = sti_vtg_get_pixel_number(*mode, 0);
+	val = y << 16 | x;
+	writel(val, layer->regs + CUR_AWS);
+	y = sti_vtg_get_line_number(*mode, mode->vdisplay - 1);
+	x = sti_vtg_get_pixel_number(*mode, mode->hdisplay - 1);
+	val = y << 16 | x;
+	writel(val, layer->regs + CUR_AWE);
+
+	if (first_prepare) {
+		/* Set and fetch CLUT */
+		writel(cursor->clut_paddr, layer->regs + CUR_CML);
+		writel(CUR_CTL_CLUT_UPDATE, layer->regs + CUR_CTL);
+	}
+
+	return 0;
+}
+
+static int sti_cursor_commit_layer(struct sti_layer *layer)
+{
+	struct sti_cursor *cursor = to_sti_cursor(layer);
+	struct drm_display_mode *mode = layer->mode;
+	u32 ydo, xdo;
+
+	dev_dbg(layer->dev, "%s %s\n", __func__, sti_layer_to_str(layer));
+
+	/* Set memory location, size, and position */
+	writel(cursor->pixmap.paddr, layer->regs + CUR_PML);
+	writel(cursor->width, layer->regs + CUR_PMP);
+	writel(cursor->height << 16 | cursor->width, layer->regs + CUR_SIZE);
+
+	ydo = sti_vtg_get_line_number(*mode, layer->dst_y);
+	xdo = sti_vtg_get_pixel_number(*mode, layer->dst_y);
+	writel((ydo << 16) | xdo, layer->regs + CUR_VPO);
+
+	return 0;
+}
+
+static int sti_cursor_disable_layer(struct sti_layer *layer)
+{
+	return 0;
+}
+
+static void sti_cursor_init(struct sti_layer *layer)
+{
+	struct sti_cursor *cursor = to_sti_cursor(layer);
+	unsigned short *base = cursor->clut;
+	unsigned int a, r, g, b;
+
+	/* Assign CLUT values, ARGB444 format */
+	for (a = 0; a < 4; a++)
+		for (r = 0; r < 4; r++)
+			for (g = 0; g < 4; g++)
+				for (b = 0; b < 4; b++)
+					*base++ = (a * 5) << 12 |
+						  (r * 5) << 8 |
+						  (g * 5) << 4 |
+						  (b * 5);
+}
+
+static const struct sti_layer_funcs cursor_ops = {
+	.get_formats = sti_cursor_get_formats,
+	.get_nb_formats = sti_cursor_get_nb_formats,
+	.init = sti_cursor_init,
+	.prepare = sti_cursor_prepare_layer,
+	.commit = sti_cursor_commit_layer,
+	.disable = sti_cursor_disable_layer,
+};
+
+struct sti_layer *sti_cursor_create(struct device *dev)
+{
+	struct sti_cursor *cursor;
+
+	cursor = devm_kzalloc(dev, sizeof(*cursor), GFP_KERNEL);
+	if (!cursor) {
+		DRM_ERROR("Failed to allocate memory for cursor\n");
+		return NULL;
+	}
+
+	/* Allocate clut buffer */
+	cursor->clut = dma_alloc_writecombine(dev,
+			0x100 * sizeof(unsigned short),
+			&cursor->clut_paddr,
+			GFP_KERNEL | GFP_DMA);
+
+	if (!cursor->clut) {
+		DRM_ERROR("Failed to allocate memory for cursor clut\n");
+		devm_kfree(dev, cursor);
+		return NULL;
+	}
+
+	cursor->layer.ops = &cursor_ops;
+
+	return (struct sti_layer *)cursor;
+}
diff --git a/drivers/gpu/drm/sti/sti_cursor.h b/drivers/gpu/drm/sti/sti_cursor.h
new file mode 100644
index 000000000000..3c9827404f27
--- /dev/null
+++ b/drivers/gpu/drm/sti/sti_cursor.h
@@ -0,0 +1,12 @@
+/*
+ * Copyright (C) STMicroelectronics SA 2013
+ * Authors: Vincent Abriou <vincent.abriou@st.com> for STMicroelectronics.
+ * License terms:  GNU General Public License (GPL), version 2
+ */
+
+#ifndef _STI_CURSOR_H_
+#define _STI_CURSOR_H_
+
+struct sti_layer *sti_cursor_create(struct device *dev);
+
+#endif
diff --git a/drivers/gpu/drm/sti/sti_drm_crtc.c b/drivers/gpu/drm/sti/sti_drm_crtc.c
index d2ae0c0e13be..4c651c200f20 100644
--- a/drivers/gpu/drm/sti/sti_drm_crtc.c
+++ b/drivers/gpu/drm/sti/sti_drm_crtc.c
@@ -10,6 +10,7 @@
 
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_plane_helper.h>
 
 #include "sti_compositor.h"
 #include "sti_drm_drv.h"
@@ -27,7 +28,7 @@ static void sti_drm_crtc_prepare(struct drm_crtc *crtc)
 	struct device *dev = mixer->dev;
 	struct sti_compositor *compo = dev_get_drvdata(dev);
 
-	compo->enable = true;
+	mixer->enabled = true;
 
 	/* Prepare and enable the compo IP clock */
 	if (mixer->id == STI_MIXER_MAIN) {
@@ -37,6 +38,8 @@ static void sti_drm_crtc_prepare(struct drm_crtc *crtc)
 		if (clk_prepare_enable(compo->clk_compo_aux))
 			DRM_INFO("Failed to prepare/enable compo_aux clk\n");
 	}
+
+	sti_mixer_clear_all_layers(mixer);
 }
 
 static void sti_drm_crtc_commit(struct drm_crtc *crtc)
@@ -61,6 +64,8 @@ static void sti_drm_crtc_commit(struct drm_crtc *crtc)
 	/* Enable layer on mixer */
 	if (sti_mixer_set_layer_status(mixer, layer, true))
 		DRM_ERROR("Can not enable layer at mixer\n");
+
+	drm_crtc_vblank_on(crtc);
 }
 
 static bool sti_drm_crtc_mode_fixup(struct drm_crtc *crtc,
@@ -143,7 +148,8 @@ sti_drm_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *mode,
 	w = crtc->primary->fb->width - x;
 	h = crtc->primary->fb->height - y;
 
-	return sti_layer_prepare(layer, crtc->primary->fb, &crtc->mode,
+	return sti_layer_prepare(layer, crtc,
+			crtc->primary->fb, &crtc->mode,
 			mixer->id, 0, 0, w, h, x, y, w, h);
 }
 
@@ -170,7 +176,8 @@ static int sti_drm_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
 	w = crtc->primary->fb->width - crtc->x;
 	h = crtc->primary->fb->height - crtc->y;
 
-	ret = sti_layer_prepare(layer, crtc->primary->fb, &crtc->mode,
+	ret = sti_layer_prepare(layer, crtc,
+				crtc->primary->fb, &crtc->mode,
 				mixer->id, 0, 0, w, h,
 				crtc->x, crtc->y, w, h);
 	if (ret) {
@@ -195,7 +202,7 @@ static void sti_drm_crtc_disable(struct drm_crtc *crtc)
 	struct sti_compositor *compo = dev_get_drvdata(dev);
 	struct sti_layer *layer;
 
-	if (!compo->enable)
+	if (!mixer->enabled)
 		return;
 
 	DRM_DEBUG_KMS("CRTC:%d (%s)\n", crtc->base.id, sti_mixer_to_str(mixer));
@@ -221,7 +228,7 @@ static void sti_drm_crtc_disable(struct drm_crtc *crtc)
 	/* Then disable layer itself */
 	sti_layer_disable(layer);
 
-	drm_vblank_off(crtc->dev, mixer->id);
+	drm_crtc_vblank_off(crtc);
 
 	/* Disable pixel clock and compo IP clocks */
 	if (mixer->id == STI_MIXER_MAIN) {
@@ -232,7 +239,7 @@ static void sti_drm_crtc_disable(struct drm_crtc *crtc)
 		clk_disable_unprepare(compo->clk_compo_aux);
 	}
 
-	compo->enable = false;
+	mixer->enabled = false;
 }
 
 static struct drm_crtc_helper_funcs sti_crtc_helper_funcs = {
@@ -363,7 +370,6 @@ void sti_drm_crtc_disable_vblank(struct drm_device *dev, int crtc)
 	struct sti_drm_private *priv = dev->dev_private;
 	struct sti_compositor *compo = priv->compo;
 	struct notifier_block *vtg_vblank_nb = &compo->vtg_vblank_nb;
-	unsigned long flags;
 
 	DRM_DEBUG_DRIVER("\n");
 
@@ -372,13 +378,10 @@ void sti_drm_crtc_disable_vblank(struct drm_device *dev, int crtc)
 		DRM_DEBUG_DRIVER("Warning: cannot unregister VTG notifier\n");
 
 	/* free the resources of the pending requests */
-	spin_lock_irqsave(&dev->event_lock, flags);
 	if (compo->mixer[crtc]->pending_event) {
 		drm_vblank_put(dev, crtc);
 		compo->mixer[crtc]->pending_event = NULL;
 	}
-	spin_unlock_irqrestore(&dev->event_lock, flags);
-
 }
 EXPORT_SYMBOL(sti_drm_crtc_disable_vblank);
 
@@ -398,6 +401,7 @@ bool sti_drm_crtc_is_main(struct drm_crtc *crtc)
 
 	return false;
 }
+EXPORT_SYMBOL(sti_drm_crtc_is_main);
 
 int sti_drm_crtc_init(struct drm_device *drm_dev, struct sti_mixer *mixer,
 		struct drm_plane *primary, struct drm_plane *cursor)
diff --git a/drivers/gpu/drm/sti/sti_drm_drv.c b/drivers/gpu/drm/sti/sti_drm_drv.c
index 223d93c3a05d..5239fa121726 100644
--- a/drivers/gpu/drm/sti/sti_drm_drv.c
+++ b/drivers/gpu/drm/sti/sti_drm_drv.c
@@ -67,8 +67,12 @@ static int sti_drm_load(struct drm_device *dev, unsigned long flags)
 	sti_drm_mode_config_init(dev);
 
 	ret = component_bind_all(dev->dev, dev);
-	if (ret)
+	if (ret) {
+		drm_kms_helper_poll_fini(dev);
+		drm_mode_config_cleanup(dev);
+		kfree(private);
 		return ret;
+	}
 
 	drm_helper_disable_unused_functions(dev);
 
@@ -184,7 +188,6 @@ static struct platform_driver sti_drm_master_driver = {
 	.probe = sti_drm_master_probe,
 	.remove = sti_drm_master_remove,
 	.driver = {
-		.owner = THIS_MODULE,
 		.name = DRIVER_NAME "__master",
 	},
 };
@@ -228,7 +231,6 @@ static struct platform_driver sti_drm_platform_driver = {
 	.probe = sti_drm_platform_probe,
 	.remove = sti_drm_platform_remove,
 	.driver = {
-		.owner = THIS_MODULE,
 		.name = DRIVER_NAME,
 		.of_match_table = sti_drm_dt_ids,
 	},
diff --git a/drivers/gpu/drm/sti/sti_drm_plane.c b/drivers/gpu/drm/sti/sti_drm_plane.c
index f4118d4cac22..bb6a29339e10 100644
--- a/drivers/gpu/drm/sti/sti_drm_plane.c
+++ b/drivers/gpu/drm/sti/sti_drm_plane.c
@@ -45,7 +45,8 @@ sti_drm_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	}
 
 	/* src_x are in 16.16 format. */
-	res = sti_layer_prepare(layer, fb, &crtc->mode, mixer->id,
+	res = sti_layer_prepare(layer, crtc, fb,
+			&crtc->mode, mixer->id,
 			crtc_x, crtc_y, crtc_w, crtc_h,
 			src_x >> 16, src_y >> 16,
 			src_w >> 16, src_h >> 16);
@@ -193,3 +194,4 @@ struct drm_plane *sti_drm_plane_init(struct drm_device *dev,
 
 	return &layer->plane;
 }
+EXPORT_SYMBOL(sti_drm_plane_init);
diff --git a/drivers/gpu/drm/sti/sti_gdp.c b/drivers/gpu/drm/sti/sti_gdp.c
index 4e30b74559f5..32448d1d1e8f 100644
--- a/drivers/gpu/drm/sti/sti_gdp.c
+++ b/drivers/gpu/drm/sti/sti_gdp.c
@@ -73,7 +73,9 @@ struct sti_gdp_node {
 
 struct sti_gdp_node_list {
 	struct sti_gdp_node *top_field;
+	dma_addr_t top_field_paddr;
 	struct sti_gdp_node *btm_field;
+	dma_addr_t btm_field_paddr;
 };
 
 /**
@@ -81,6 +83,8 @@ struct sti_gdp_node_list {
  *
  * @layer:		layer structure
  * @clk_pix:            pixel clock for the current gdp
+ * @clk_main_parent:    gdp parent clock if main path used
+ * @clk_aux_parent:     gdp parent clock if aux path used
  * @vtg_field_nb:       callback for VTG FIELD (top or bottom) notification
  * @is_curr_top:        true if the current node processed is the top field
  * @node_list:		array of node list
@@ -88,6 +92,8 @@ struct sti_gdp_node_list {
 struct sti_gdp {
 	struct sti_layer layer;
 	struct clk *clk_pix;
+	struct clk *clk_main_parent;
+	struct clk *clk_aux_parent;
 	struct notifier_block vtg_field_nb;
 	bool is_curr_top;
 	struct sti_gdp_node_list node_list[GDP_NODE_NB_BANK];
@@ -168,7 +174,6 @@ static int sti_gdp_get_alpharange(int format)
 static struct sti_gdp_node_list *sti_gdp_get_free_nodes(struct sti_layer *layer)
 {
 	int hw_nvn;
-	void *virt_nvn;
 	struct sti_gdp *gdp = to_sti_gdp(layer);
 	unsigned int i;
 
@@ -176,11 +181,9 @@ static struct sti_gdp_node_list *sti_gdp_get_free_nodes(struct sti_layer *layer)
 	if (!hw_nvn)
 		goto end;
 
-	virt_nvn = dma_to_virt(layer->dev, (dma_addr_t) hw_nvn);
-
 	for (i = 0; i < GDP_NODE_NB_BANK; i++)
-		if ((virt_nvn != gdp->node_list[i].btm_field) &&
-		    (virt_nvn != gdp->node_list[i].top_field))
+		if ((hw_nvn != gdp->node_list[i].btm_field_paddr) &&
+		    (hw_nvn != gdp->node_list[i].top_field_paddr))
 			return &gdp->node_list[i];
 
 	/* in hazardious cases restart with the first node */
@@ -204,7 +207,6 @@ static
 struct sti_gdp_node_list *sti_gdp_get_current_nodes(struct sti_layer *layer)
 {
 	int hw_nvn;
-	void *virt_nvn;
 	struct sti_gdp *gdp = to_sti_gdp(layer);
 	unsigned int i;
 
@@ -212,11 +214,9 @@ struct sti_gdp_node_list *sti_gdp_get_current_nodes(struct sti_layer *layer)
 	if (!hw_nvn)
 		goto end;
 
-	virt_nvn = dma_to_virt(layer->dev, (dma_addr_t) hw_nvn);
-
 	for (i = 0; i < GDP_NODE_NB_BANK; i++)
-		if ((virt_nvn == gdp->node_list[i].btm_field) ||
-				(virt_nvn == gdp->node_list[i].top_field))
+		if ((hw_nvn == gdp->node_list[i].btm_field_paddr) ||
+				(hw_nvn == gdp->node_list[i].top_field_paddr))
 			return &gdp->node_list[i];
 
 end:
@@ -292,8 +292,8 @@ static int sti_gdp_prepare_layer(struct sti_layer *layer, bool first_prepare)
 
 	/* Same content and chained together */
 	memcpy(btm_field, top_field, sizeof(*btm_field));
-	top_field->gam_gdp_nvn = virt_to_dma(dev, btm_field);
-	btm_field->gam_gdp_nvn = virt_to_dma(dev, top_field);
+	top_field->gam_gdp_nvn = list->btm_field_paddr;
+	btm_field->gam_gdp_nvn = list->top_field_paddr;
 
 	/* Interlaced mode */
 	if (layer->mode->flags & DRM_MODE_FLAG_INTERLACE)
@@ -311,6 +311,17 @@ static int sti_gdp_prepare_layer(struct sti_layer *layer, bool first_prepare)
 
 		/* Set and enable gdp clock */
 		if (gdp->clk_pix) {
+			struct clk *clkp;
+			/* According to the mixer used, the gdp pixel clock
+			 * should have a different parent clock. */
+			if (layer->mixer_id == STI_MIXER_MAIN)
+				clkp = gdp->clk_main_parent;
+			else
+				clkp = gdp->clk_aux_parent;
+
+			if (clkp)
+				clk_set_parent(gdp->clk_pix, clkp);
+
 			res = clk_set_rate(gdp->clk_pix, rate);
 			if (res < 0) {
 				DRM_ERROR("Cannot set rate (%dHz) for gdp\n",
@@ -349,8 +360,8 @@ static int sti_gdp_commit_layer(struct sti_layer *layer)
 	struct sti_gdp_node *updated_top_node = updated_list->top_field;
 	struct sti_gdp_node *updated_btm_node = updated_list->btm_field;
 	struct sti_gdp *gdp = to_sti_gdp(layer);
-	u32 dma_updated_top = virt_to_dma(layer->dev, updated_top_node);
-	u32 dma_updated_btm = virt_to_dma(layer->dev, updated_btm_node);
+	u32 dma_updated_top = updated_list->top_field_paddr;
+	u32 dma_updated_btm = updated_list->btm_field_paddr;
 	struct sti_gdp_node_list *curr_list = sti_gdp_get_current_nodes(layer);
 
 	dev_dbg(layer->dev, "%s %s top/btm_node:0x%p/0x%p\n", __func__,
@@ -461,16 +472,16 @@ static void sti_gdp_init(struct sti_layer *layer)
 {
 	struct sti_gdp *gdp = to_sti_gdp(layer);
 	struct device_node *np = layer->dev->of_node;
-	dma_addr_t dma;
+	dma_addr_t dma_addr;
 	void *base;
 	unsigned int i, size;
 
 	/* Allocate all the nodes within a single memory page */
 	size = sizeof(struct sti_gdp_node) *
 	    GDP_NODE_PER_FIELD * GDP_NODE_NB_BANK;
-
 	base = dma_alloc_writecombine(layer->dev,
-			size, &dma, GFP_KERNEL | GFP_DMA);
+			size, &dma_addr, GFP_KERNEL | GFP_DMA);
+
 	if (!base) {
 		DRM_ERROR("Failed to allocate memory for GDP node\n");
 		return;
@@ -478,21 +489,26 @@ static void sti_gdp_init(struct sti_layer *layer)
 	memset(base, 0, size);
 
 	for (i = 0; i < GDP_NODE_NB_BANK; i++) {
-		if (virt_to_dma(layer->dev, base) & 0xF) {
+		if (dma_addr & 0xF) {
 			DRM_ERROR("Mem alignment failed\n");
 			return;
 		}
 		gdp->node_list[i].top_field = base;
+		gdp->node_list[i].top_field_paddr = dma_addr;
+
 		DRM_DEBUG_DRIVER("node[%d].top_field=%p\n", i, base);
 		base += sizeof(struct sti_gdp_node);
+		dma_addr += sizeof(struct sti_gdp_node);
 
-		if (virt_to_dma(layer->dev, base) & 0xF) {
+		if (dma_addr & 0xF) {
 			DRM_ERROR("Mem alignment failed\n");
 			return;
 		}
 		gdp->node_list[i].btm_field = base;
+		gdp->node_list[i].btm_field_paddr = dma_addr;
 		DRM_DEBUG_DRIVER("node[%d].btm_field=%p\n", i, base);
 		base += sizeof(struct sti_gdp_node);
+		dma_addr += sizeof(struct sti_gdp_node);
 	}
 
 	if (of_device_is_compatible(np, "st,stih407-compositor")) {
@@ -520,6 +536,14 @@ static void sti_gdp_init(struct sti_layer *layer)
 		gdp->clk_pix = devm_clk_get(layer->dev, clk_name);
 		if (IS_ERR(gdp->clk_pix))
 			DRM_ERROR("Cannot get %s clock\n", clk_name);
+
+		gdp->clk_main_parent = devm_clk_get(layer->dev, "main_parent");
+		if (IS_ERR(gdp->clk_main_parent))
+			DRM_ERROR("Cannot get main_parent clock\n");
+
+		gdp->clk_aux_parent = devm_clk_get(layer->dev, "aux_parent");
+		if (IS_ERR(gdp->clk_aux_parent))
+			DRM_ERROR("Cannot get aux_parent clock\n");
 	}
 }
 
diff --git a/drivers/gpu/drm/sti/sti_hdmi.c b/drivers/gpu/drm/sti/sti_hdmi.c
index b22968c08d1f..d032e024b0b8 100644
--- a/drivers/gpu/drm/sti/sti_hdmi.c
+++ b/drivers/gpu/drm/sti/sti_hdmi.c
@@ -130,8 +130,7 @@ static irqreturn_t hdmi_irq_thread(int irq, void *arg)
 
 	/* Hot plug/unplug IRQ */
 	if (hdmi->irq_status & HDMI_INT_HOT_PLUG) {
-		/* read gpio to get the status */
-		hdmi->hpd = gpio_get_value(hdmi->hpd_gpio);
+		hdmi->hpd = readl(hdmi->regs + HDMI_STA) & HDMI_STA_HOT_PLUG;
 		if (hdmi->drm_dev)
 			drm_helper_hpd_irq_event(hdmi->drm_dev);
 	}
@@ -273,31 +272,32 @@ static int hdmi_avi_infoframe_config(struct sti_hdmi *hdmi)
 	hdmi_write(hdmi, val, HDMI_SW_DI_CFG);
 
 	/* Infoframe header */
-	val = buffer[0x0];
-	val |= buffer[0x1] << 8;
-	val |= buffer[0x2] << 16;
+	val =  buffer[0];
+	val |= buffer[1] << 8;
+	val |= buffer[2] << 16;
 	hdmi_write(hdmi, val, HDMI_SW_DI_N_HEAD_WORD(HDMI_IFRAME_SLOT_AVI));
 
 	/* Infoframe packet bytes */
-	val = frame[0x0];
-	val |= frame[0x1] << 8;
-	val |= frame[0x2] << 16;
-	val |= frame[0x3] << 24;
+	val =  buffer[3];
+	val |= *(frame++) << 8;
+	val |= *(frame++) << 16;
+	val |= *(frame++) << 24;
 	hdmi_write(hdmi, val, HDMI_SW_DI_N_PKT_WORD0(HDMI_IFRAME_SLOT_AVI));
 
-	val = frame[0x4];
-	val |= frame[0x5] << 8;
-	val |= frame[0x6] << 16;
-	val |= frame[0x7] << 24;
+	val =  *(frame++);
+	val |= *(frame++) << 8;
+	val |= *(frame++) << 16;
+	val |= *(frame++) << 24;
 	hdmi_write(hdmi, val, HDMI_SW_DI_N_PKT_WORD1(HDMI_IFRAME_SLOT_AVI));
 
-	val = frame[0x8];
-	val |= frame[0x9] << 8;
-	val |= frame[0xA] << 16;
-	val |= frame[0xB] << 24;
+	val =  *(frame++);
+	val |= *(frame++) << 8;
+	val |= *(frame++) << 16;
+	val |= *(frame++) << 24;
 	hdmi_write(hdmi, val, HDMI_SW_DI_N_PKT_WORD2(HDMI_IFRAME_SLOT_AVI));
 
-	val = frame[0xC];
+	val = *(frame++);
+	val |= *(frame) << 8;
 	hdmi_write(hdmi, val, HDMI_SW_DI_N_PKT_WORD3(HDMI_IFRAME_SLOT_AVI));
 
 	/* Enable transmission slot for AVI infoframe
@@ -480,17 +480,15 @@ static const struct drm_bridge_funcs sti_hdmi_bridge_funcs = {
 
 static int sti_hdmi_connector_get_modes(struct drm_connector *connector)
 {
-	struct i2c_adapter *i2c_adap;
+	struct sti_hdmi_connector *hdmi_connector
+		= to_sti_hdmi_connector(connector);
+	struct sti_hdmi *hdmi = hdmi_connector->hdmi;
 	struct edid *edid;
 	int count;
 
 	DRM_DEBUG_DRIVER("\n");
 
-	i2c_adap = i2c_get_adapter(1);
-	if (!i2c_adap)
-		goto fail;
-
-	edid = drm_get_edid(connector, i2c_adap);
+	edid = drm_get_edid(connector, hdmi->ddc_adapt);
 	if (!edid)
 		goto fail;
 
@@ -603,29 +601,38 @@ static int sti_hdmi_bind(struct device *dev, struct device *master, void *data)
 	struct sti_hdmi_connector *connector;
 	struct drm_connector *drm_connector;
 	struct drm_bridge *bridge;
-	struct i2c_adapter *i2c_adap;
+	struct device_node *ddc;
 	int err;
 
-	i2c_adap = i2c_get_adapter(1);
-	if (!i2c_adap)
-		return -EPROBE_DEFER;
+	ddc = of_parse_phandle(dev->of_node, "ddc", 0);
+	if (ddc) {
+		hdmi->ddc_adapt = of_find_i2c_adapter_by_node(ddc);
+		if (!hdmi->ddc_adapt) {
+			err = -EPROBE_DEFER;
+			of_node_put(ddc);
+			return err;
+		}
+
+		of_node_put(ddc);
+	}
 
 	/* Set the drm device handle */
 	hdmi->drm_dev = drm_dev;
 
 	encoder = sti_hdmi_find_encoder(drm_dev);
 	if (!encoder)
-		return -ENOMEM;
+		goto err_adapt;
 
 	connector = devm_kzalloc(dev, sizeof(*connector), GFP_KERNEL);
 	if (!connector)
-		return -ENOMEM;
+		goto err_adapt;
+
 
 	connector->hdmi = hdmi;
 
 	bridge = devm_kzalloc(dev, sizeof(*bridge), GFP_KERNEL);
 	if (!bridge)
-		return -ENOMEM;
+		goto err_adapt;
 
 	bridge->driver_private = hdmi;
 	drm_bridge_init(drm_dev, bridge, &sti_hdmi_bridge_funcs);
@@ -662,6 +669,8 @@ err_sysfs:
 err_connector:
 	drm_bridge_cleanup(bridge);
 	drm_connector_cleanup(drm_connector);
+err_adapt:
+	put_device(&hdmi->ddc_adapt->dev);
 	return -EINVAL;
 }
 
@@ -757,13 +766,7 @@ static int sti_hdmi_probe(struct platform_device *pdev)
 		return PTR_ERR(hdmi->clk_audio);
 	}
 
-	hdmi->hpd_gpio = of_get_named_gpio(np, "hdmi,hpd-gpio", 0);
-	if (hdmi->hpd_gpio < 0) {
-		DRM_ERROR("Failed to get hdmi hpd-gpio\n");
-		return -EIO;
-	}
-
-	hdmi->hpd = gpio_get_value(hdmi->hpd_gpio);
+	hdmi->hpd = readl(hdmi->regs + HDMI_STA) & HDMI_STA_HOT_PLUG;
 
 	init_waitqueue_head(&hdmi->wait_event);
 
@@ -788,6 +791,11 @@ static int sti_hdmi_probe(struct platform_device *pdev)
 
 static int sti_hdmi_remove(struct platform_device *pdev)
 {
+	struct sti_hdmi *hdmi = dev_get_drvdata(&pdev->dev);
+
+	if (hdmi->ddc_adapt)
+		put_device(&hdmi->ddc_adapt->dev);
+
 	component_del(&pdev->dev, &sti_hdmi_ops);
 	return 0;
 }
diff --git a/drivers/gpu/drm/sti/sti_hdmi.h b/drivers/gpu/drm/sti/sti_hdmi.h
index 61bec6557ceb..3d22390e1f3b 100644
--- a/drivers/gpu/drm/sti/sti_hdmi.h
+++ b/drivers/gpu/drm/sti/sti_hdmi.h
@@ -14,6 +14,9 @@
 #define HDMI_STA           0x0010
 #define HDMI_STA_DLL_LCK   BIT(5)
 
+#define HDMI_STA_HOT_PLUG_SHIFT 4
+#define HDMI_STA_HOT_PLUG	(1 << HDMI_STA_HOT_PLUG_SHIFT)
+
 struct sti_hdmi;
 
 struct hdmi_phy_ops {
@@ -37,7 +40,6 @@ struct hdmi_phy_ops {
  * @irq_status: interrupt status register
  * @phy_ops: phy start/stop operations
  * @enabled: true if hdmi is enabled else false
- * @hpd_gpio: hdmi hot plug detect gpio number
  * @hpd: hot plug detect status
  * @wait_event: wait event
  * @event_received: wait event status
@@ -57,11 +59,11 @@ struct sti_hdmi {
 	u32 irq_status;
 	struct hdmi_phy_ops *phy_ops;
 	bool enabled;
-	int hpd_gpio;
 	bool hpd;
 	wait_queue_head_t wait_event;
 	bool event_received;
 	struct reset_control *reset;
+	struct i2c_adapter *ddc_adapt;
 };
 
 u32 hdmi_read(struct sti_hdmi *hdmi, int offset);
diff --git a/drivers/gpu/drm/sti/sti_hqvdp.c b/drivers/gpu/drm/sti/sti_hqvdp.c
new file mode 100644
index 000000000000..f3db05dab0ab
--- /dev/null
+++ b/drivers/gpu/drm/sti/sti_hqvdp.c
@@ -0,0 +1,1073 @@
+/*
+ * Copyright (C) STMicroelectronics SA 2014
+ * Authors: Fabien Dessenne <fabien.dessenne@st.com> for STMicroelectronics.
+ * License terms:  GNU General Public License (GPL), version 2
+ */
+
+#include <linux/clk.h>
+#include <linux/component.h>
+#include <linux/firmware.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/reset.h>
+
+#include <drm/drmP.h>
+
+#include "sti_drm_plane.h"
+#include "sti_hqvdp.h"
+#include "sti_hqvdp_lut.h"
+#include "sti_layer.h"
+#include "sti_vtg.h"
+
+/* Firmware name */
+#define HQVDP_FMW_NAME          "hqvdp-stih407.bin"
+
+/* Regs address */
+#define HQVDP_DMEM              0x00000000               /* 0x00000000 */
+#define HQVDP_PMEM              0x00040000               /* 0x00040000 */
+#define HQVDP_RD_PLUG           0x000E0000               /* 0x000E0000 */
+#define HQVDP_RD_PLUG_CONTROL   (HQVDP_RD_PLUG + 0x1000) /* 0x000E1000 */
+#define HQVDP_RD_PLUG_PAGE_SIZE (HQVDP_RD_PLUG + 0x1004) /* 0x000E1004 */
+#define HQVDP_RD_PLUG_MIN_OPC   (HQVDP_RD_PLUG + 0x1008) /* 0x000E1008 */
+#define HQVDP_RD_PLUG_MAX_OPC   (HQVDP_RD_PLUG + 0x100C) /* 0x000E100C */
+#define HQVDP_RD_PLUG_MAX_CHK   (HQVDP_RD_PLUG + 0x1010) /* 0x000E1010 */
+#define HQVDP_RD_PLUG_MAX_MSG   (HQVDP_RD_PLUG + 0x1014) /* 0x000E1014 */
+#define HQVDP_RD_PLUG_MIN_SPACE (HQVDP_RD_PLUG + 0x1018) /* 0x000E1018 */
+#define HQVDP_WR_PLUG           0x000E2000               /* 0x000E2000 */
+#define HQVDP_WR_PLUG_CONTROL   (HQVDP_WR_PLUG + 0x1000) /* 0x000E3000 */
+#define HQVDP_WR_PLUG_PAGE_SIZE (HQVDP_WR_PLUG + 0x1004) /* 0x000E3004 */
+#define HQVDP_WR_PLUG_MIN_OPC   (HQVDP_WR_PLUG + 0x1008) /* 0x000E3008 */
+#define HQVDP_WR_PLUG_MAX_OPC   (HQVDP_WR_PLUG + 0x100C) /* 0x000E300C */
+#define HQVDP_WR_PLUG_MAX_CHK   (HQVDP_WR_PLUG + 0x1010) /* 0x000E3010 */
+#define HQVDP_WR_PLUG_MAX_MSG   (HQVDP_WR_PLUG + 0x1014) /* 0x000E3014 */
+#define HQVDP_WR_PLUG_MIN_SPACE (HQVDP_WR_PLUG + 0x1018) /* 0x000E3018 */
+#define HQVDP_MBX               0x000E4000               /* 0x000E4000 */
+#define HQVDP_MBX_IRQ_TO_XP70   (HQVDP_MBX + 0x0000)     /* 0x000E4000 */
+#define HQVDP_MBX_INFO_HOST     (HQVDP_MBX + 0x0004)     /* 0x000E4004 */
+#define HQVDP_MBX_IRQ_TO_HOST   (HQVDP_MBX + 0x0008)     /* 0x000E4008 */
+#define HQVDP_MBX_INFO_XP70     (HQVDP_MBX + 0x000C)     /* 0x000E400C */
+#define HQVDP_MBX_SW_RESET_CTRL (HQVDP_MBX + 0x0010)     /* 0x000E4010 */
+#define HQVDP_MBX_STARTUP_CTRL1 (HQVDP_MBX + 0x0014)     /* 0x000E4014 */
+#define HQVDP_MBX_STARTUP_CTRL2 (HQVDP_MBX + 0x0018)     /* 0x000E4018 */
+#define HQVDP_MBX_GP_STATUS     (HQVDP_MBX + 0x001C)     /* 0x000E401C */
+#define HQVDP_MBX_NEXT_CMD      (HQVDP_MBX + 0x0020)     /* 0x000E4020 */
+#define HQVDP_MBX_CURRENT_CMD   (HQVDP_MBX + 0x0024)     /* 0x000E4024 */
+#define HQVDP_MBX_SOFT_VSYNC    (HQVDP_MBX + 0x0028)     /* 0x000E4028 */
+
+/* Plugs config */
+#define PLUG_CONTROL_ENABLE     0x00000001
+#define PLUG_PAGE_SIZE_256      0x00000002
+#define PLUG_MIN_OPC_8          0x00000003
+#define PLUG_MAX_OPC_64         0x00000006
+#define PLUG_MAX_CHK_2X         0x00000001
+#define PLUG_MAX_MSG_1X         0x00000000
+#define PLUG_MIN_SPACE_1        0x00000000
+
+/* SW reset CTRL */
+#define SW_RESET_CTRL_FULL      BIT(0)
+#define SW_RESET_CTRL_CORE      BIT(1)
+
+/* Startup ctrl 1 */
+#define STARTUP_CTRL1_RST_DONE  BIT(0)
+#define STARTUP_CTRL1_AUTH_IDLE BIT(2)
+
+/* Startup ctrl 2 */
+#define STARTUP_CTRL2_FETCH_EN  BIT(1)
+
+/* Info xP70 */
+#define INFO_XP70_FW_READY      BIT(15)
+#define INFO_XP70_FW_PROCESSING BIT(14)
+#define INFO_XP70_FW_INITQUEUES BIT(13)
+
+/* SOFT_VSYNC */
+#define SOFT_VSYNC_HW           0x00000000
+#define SOFT_VSYNC_SW_CMD       0x00000001
+#define SOFT_VSYNC_SW_CTRL_IRQ  0x00000003
+
+/* Reset & boot poll config */
+#define POLL_MAX_ATTEMPT        50
+#define POLL_DELAY_MS           20
+
+#define SCALE_FACTOR            8192
+#define SCALE_MAX_FOR_LEG_LUT_F 4096
+#define SCALE_MAX_FOR_LEG_LUT_E 4915
+#define SCALE_MAX_FOR_LEG_LUT_D 6654
+#define SCALE_MAX_FOR_LEG_LUT_C 8192
+
+enum sti_hvsrc_orient {
+	HVSRC_HORI,
+	HVSRC_VERT
+};
+
+/* Command structures */
+struct sti_hqvdp_top {
+	u32 config;
+	u32 mem_format;
+	u32 current_luma;
+	u32 current_enh_luma;
+	u32 current_right_luma;
+	u32 current_enh_right_luma;
+	u32 current_chroma;
+	u32 current_enh_chroma;
+	u32 current_right_chroma;
+	u32 current_enh_right_chroma;
+	u32 output_luma;
+	u32 output_chroma;
+	u32 luma_src_pitch;
+	u32 luma_enh_src_pitch;
+	u32 luma_right_src_pitch;
+	u32 luma_enh_right_src_pitch;
+	u32 chroma_src_pitch;
+	u32 chroma_enh_src_pitch;
+	u32 chroma_right_src_pitch;
+	u32 chroma_enh_right_src_pitch;
+	u32 luma_processed_pitch;
+	u32 chroma_processed_pitch;
+	u32 input_frame_size;
+	u32 input_viewport_ori;
+	u32 input_viewport_ori_right;
+	u32 input_viewport_size;
+	u32 left_view_border_width;
+	u32 right_view_border_width;
+	u32 left_view_3d_offset_width;
+	u32 right_view_3d_offset_width;
+	u32 side_stripe_color;
+	u32 crc_reset_ctrl;
+};
+
+/* Configs for interlaced : no IT, no pass thru, 3 fields */
+#define TOP_CONFIG_INTER_BTM            0x00000000
+#define TOP_CONFIG_INTER_TOP            0x00000002
+
+/* Config for progressive : no IT, no pass thru, 3 fields */
+#define TOP_CONFIG_PROGRESSIVE          0x00000001
+
+/* Default MemFormat: in=420_raster_dual out=444_raster;opaque Mem2Tv mode */
+#define TOP_MEM_FORMAT_DFLT             0x00018060
+
+/* Min/Max size */
+#define MAX_WIDTH                       0x1FFF
+#define MAX_HEIGHT                      0x0FFF
+#define MIN_WIDTH                       0x0030
+#define MIN_HEIGHT                      0x0010
+
+struct sti_hqvdp_vc1re {
+	u32 ctrl_prv_csdi;
+	u32 ctrl_cur_csdi;
+	u32 ctrl_nxt_csdi;
+	u32 ctrl_cur_fmd;
+	u32 ctrl_nxt_fmd;
+};
+
+struct sti_hqvdp_fmd {
+	u32 config;
+	u32 viewport_ori;
+	u32 viewport_size;
+	u32 next_next_luma;
+	u32 next_next_right_luma;
+	u32 next_next_next_luma;
+	u32 next_next_next_right_luma;
+	u32 threshold_scd;
+	u32 threshold_rfd;
+	u32 threshold_move;
+	u32 threshold_cfd;
+};
+
+struct sti_hqvdp_csdi {
+	u32 config;
+	u32 config2;
+	u32 dcdi_config;
+	u32 prev_luma;
+	u32 prev_enh_luma;
+	u32 prev_right_luma;
+	u32 prev_enh_right_luma;
+	u32 next_luma;
+	u32 next_enh_luma;
+	u32 next_right_luma;
+	u32 next_enh_right_luma;
+	u32 prev_chroma;
+	u32 prev_enh_chroma;
+	u32 prev_right_chroma;
+	u32 prev_enh_right_chroma;
+	u32 next_chroma;
+	u32 next_enh_chroma;
+	u32 next_right_chroma;
+	u32 next_enh_right_chroma;
+	u32 prev_motion;
+	u32 prev_right_motion;
+	u32 cur_motion;
+	u32 cur_right_motion;
+	u32 next_motion;
+	u32 next_right_motion;
+};
+
+/* Config for progressive: by pass */
+#define CSDI_CONFIG_PROG                0x00000000
+/* Config for directional deinterlacing without motion */
+#define CSDI_CONFIG_INTER_DIR           0x00000016
+/* Additional configs for fader, blender, motion,... deinterlace algorithms */
+#define CSDI_CONFIG2_DFLT               0x000001B3
+#define CSDI_DCDI_CONFIG_DFLT           0x00203803
+
+struct sti_hqvdp_hvsrc {
+	u32 hor_panoramic_ctrl;
+	u32 output_picture_size;
+	u32 init_horizontal;
+	u32 init_vertical;
+	u32 param_ctrl;
+	u32 yh_coef[NB_COEF];
+	u32 ch_coef[NB_COEF];
+	u32 yv_coef[NB_COEF];
+	u32 cv_coef[NB_COEF];
+	u32 hori_shift;
+	u32 vert_shift;
+};
+
+/* Default ParamCtrl: all controls enabled */
+#define HVSRC_PARAM_CTRL_DFLT           0xFFFFFFFF
+
+struct sti_hqvdp_iqi {
+	u32 config;
+	u32 demo_wind_size;
+	u32 pk_config;
+	u32 coeff0_coeff1;
+	u32 coeff2_coeff3;
+	u32 coeff4;
+	u32 pk_lut;
+	u32 pk_gain;
+	u32 pk_coring_level;
+	u32 cti_config;
+	u32 le_config;
+	u32 le_lut[64];
+	u32 con_bri;
+	u32 sat_gain;
+	u32 pxf_conf;
+	u32 default_color;
+};
+
+/* Default Config : IQI bypassed */
+#define IQI_CONFIG_DFLT                 0x00000001
+/* Default Contrast & Brightness gain = 256 */
+#define IQI_CON_BRI_DFLT                0x00000100
+/* Default Saturation gain = 256 */
+#define IQI_SAT_GAIN_DFLT               0x00000100
+/* Default PxfConf : P2I bypassed */
+#define IQI_PXF_CONF_DFLT               0x00000001
+
+struct sti_hqvdp_top_status {
+	u32 processing_time;
+	u32 input_y_crc;
+	u32 input_uv_crc;
+};
+
+struct sti_hqvdp_fmd_status {
+	u32 fmd_repeat_move_status;
+	u32 fmd_scene_count_status;
+	u32 cfd_sum;
+	u32 field_sum;
+	u32 next_y_fmd_crc;
+	u32 next_next_y_fmd_crc;
+	u32 next_next_next_y_fmd_crc;
+};
+
+struct sti_hqvdp_csdi_status {
+	u32 prev_y_csdi_crc;
+	u32 cur_y_csdi_crc;
+	u32 next_y_csdi_crc;
+	u32 prev_uv_csdi_crc;
+	u32 cur_uv_csdi_crc;
+	u32 next_uv_csdi_crc;
+	u32 y_csdi_crc;
+	u32 uv_csdi_crc;
+	u32 uv_cup_crc;
+	u32 mot_csdi_crc;
+	u32 mot_cur_csdi_crc;
+	u32 mot_prev_csdi_crc;
+};
+
+struct sti_hqvdp_hvsrc_status {
+	u32 y_hvsrc_crc;
+	u32 u_hvsrc_crc;
+	u32 v_hvsrc_crc;
+};
+
+struct sti_hqvdp_iqi_status {
+	u32 pxf_it_status;
+	u32 y_iqi_crc;
+	u32 u_iqi_crc;
+	u32 v_iqi_crc;
+};
+
+/* Main commands. We use 2 commands one being processed by the firmware, one
+ * ready to be fetched upon next Vsync*/
+#define NB_VDP_CMD	2
+
+struct sti_hqvdp_cmd {
+	struct sti_hqvdp_top top;
+	struct sti_hqvdp_vc1re vc1re;
+	struct sti_hqvdp_fmd fmd;
+	struct sti_hqvdp_csdi csdi;
+	struct sti_hqvdp_hvsrc hvsrc;
+	struct sti_hqvdp_iqi iqi;
+	struct sti_hqvdp_top_status top_status;
+	struct sti_hqvdp_fmd_status fmd_status;
+	struct sti_hqvdp_csdi_status csdi_status;
+	struct sti_hqvdp_hvsrc_status hvsrc_status;
+	struct sti_hqvdp_iqi_status iqi_status;
+};
+
+/*
+ * STI HQVDP structure
+ *
+ * @dev:               driver device
+ * @drm_dev:           the drm device
+ * @regs:              registers
+ * @layer:             layer structure for hqvdp it self
+ * @vid_plane:         VID plug used as link with compositor IP
+ * @clk:               IP clock
+ * @clk_pix_main:      pix main clock
+ * @reset:             reset control
+ * @vtg_nb:            notifier to handle VTG Vsync
+ * @btm_field_pending: is there any bottom field (interlaced frame) to display
+ * @curr_field_count:  number of field updates
+ * @last_field_count:  number of field updates since last fps measure
+ * @hqvdp_cmd:         buffer of commands
+ * @hqvdp_cmd_paddr:   physical address of hqvdp_cmd
+ * @vtg:               vtg for main data path
+ */
+struct sti_hqvdp {
+	struct device *dev;
+	struct drm_device *drm_dev;
+	void __iomem *regs;
+	struct sti_layer layer;
+	struct drm_plane *vid_plane;
+	struct clk *clk;
+	struct clk *clk_pix_main;
+	struct reset_control *reset;
+	struct notifier_block vtg_nb;
+	bool btm_field_pending;
+	unsigned int curr_field_count;
+	unsigned int last_field_count;
+	void *hqvdp_cmd;
+	dma_addr_t hqvdp_cmd_paddr;
+	struct sti_vtg *vtg;
+};
+
+#define to_sti_hqvdp(x) container_of(x, struct sti_hqvdp, layer)
+
+static const uint32_t hqvdp_supported_formats[] = {
+	DRM_FORMAT_NV12,
+};
+
+static const uint32_t *sti_hqvdp_get_formats(struct sti_layer *layer)
+{
+	return hqvdp_supported_formats;
+}
+
+static unsigned int sti_hqvdp_get_nb_formats(struct sti_layer *layer)
+{
+	return ARRAY_SIZE(hqvdp_supported_formats);
+}
+
+/**
+ * sti_hqvdp_get_free_cmd
+ * @hqvdp: hqvdp structure
+ *
+ * Look for a hqvdp_cmd that is not being used (or about to be used) by the FW.
+ *
+ * RETURNS:
+ * the offset of the command to be used.
+ * -1 in error cases
+ */
+static int sti_hqvdp_get_free_cmd(struct sti_hqvdp *hqvdp)
+{
+	int curr_cmd, next_cmd;
+	dma_addr_t cmd = hqvdp->hqvdp_cmd_paddr;
+	int i;
+
+	curr_cmd = readl(hqvdp->regs + HQVDP_MBX_CURRENT_CMD);
+	next_cmd = readl(hqvdp->regs + HQVDP_MBX_NEXT_CMD);
+
+	for (i = 0; i < NB_VDP_CMD; i++) {
+		if ((cmd != curr_cmd) && (cmd != next_cmd))
+			return i * sizeof(struct sti_hqvdp_cmd);
+		cmd += sizeof(struct sti_hqvdp_cmd);
+	}
+
+	return -1;
+}
+
+/**
+ * sti_hqvdp_get_curr_cmd
+ * @hqvdp: hqvdp structure
+ *
+ * Look for the hqvdp_cmd that is being used by the FW.
+ *
+ * RETURNS:
+ *  the offset of the command to be used.
+ * -1 in error cases
+ */
+static int sti_hqvdp_get_curr_cmd(struct sti_hqvdp *hqvdp)
+{
+	int curr_cmd;
+	dma_addr_t cmd = hqvdp->hqvdp_cmd_paddr;
+	unsigned int i;
+
+	curr_cmd = readl(hqvdp->regs + HQVDP_MBX_CURRENT_CMD);
+
+	for (i = 0; i < NB_VDP_CMD; i++) {
+		if (cmd == curr_cmd)
+			return i * sizeof(struct sti_hqvdp_cmd);
+
+		cmd += sizeof(struct sti_hqvdp_cmd);
+	}
+
+	return -1;
+}
+
+/**
+ * sti_hqvdp_update_hvsrc
+ * @orient: horizontal or vertical
+ * @scale:  scaling/zoom factor
+ * @hvsrc:  the structure containing the LUT coef
+ *
+ * Update the Y and C Lut coef, as well as the shift param
+ *
+ * RETURNS:
+ * None.
+ */
+static void sti_hqvdp_update_hvsrc(enum sti_hvsrc_orient orient, int scale,
+		struct sti_hqvdp_hvsrc *hvsrc)
+{
+	const int *coef_c, *coef_y;
+	int shift_c, shift_y;
+
+	/* Get the appropriate coef tables */
+	if (scale < SCALE_MAX_FOR_LEG_LUT_F) {
+		coef_y = coef_lut_f_y_legacy;
+		coef_c = coef_lut_f_c_legacy;
+		shift_y = SHIFT_LUT_F_Y_LEGACY;
+		shift_c = SHIFT_LUT_F_C_LEGACY;
+	} else if (scale < SCALE_MAX_FOR_LEG_LUT_E) {
+		coef_y = coef_lut_e_y_legacy;
+		coef_c = coef_lut_e_c_legacy;
+		shift_y = SHIFT_LUT_E_Y_LEGACY;
+		shift_c = SHIFT_LUT_E_C_LEGACY;
+	} else if (scale < SCALE_MAX_FOR_LEG_LUT_D) {
+		coef_y = coef_lut_d_y_legacy;
+		coef_c = coef_lut_d_c_legacy;
+		shift_y = SHIFT_LUT_D_Y_LEGACY;
+		shift_c = SHIFT_LUT_D_C_LEGACY;
+	} else if (scale < SCALE_MAX_FOR_LEG_LUT_C) {
+		coef_y = coef_lut_c_y_legacy;
+		coef_c = coef_lut_c_c_legacy;
+		shift_y = SHIFT_LUT_C_Y_LEGACY;
+		shift_c = SHIFT_LUT_C_C_LEGACY;
+	} else if (scale == SCALE_MAX_FOR_LEG_LUT_C) {
+		coef_y = coef_c = coef_lut_b;
+		shift_y = shift_c = SHIFT_LUT_B;
+	} else {
+		coef_y = coef_c = coef_lut_a_legacy;
+		shift_y = shift_c = SHIFT_LUT_A_LEGACY;
+	}
+
+	if (orient == HVSRC_HORI) {
+		hvsrc->hori_shift = (shift_c << 16) | shift_y;
+		memcpy(hvsrc->yh_coef, coef_y, sizeof(hvsrc->yh_coef));
+		memcpy(hvsrc->ch_coef, coef_c, sizeof(hvsrc->ch_coef));
+	} else {
+		hvsrc->vert_shift = (shift_c << 16) | shift_y;
+		memcpy(hvsrc->yv_coef, coef_y, sizeof(hvsrc->yv_coef));
+		memcpy(hvsrc->cv_coef, coef_c, sizeof(hvsrc->cv_coef));
+	}
+}
+
+/**
+ * sti_hqvdp_check_hw_scaling
+ * @layer: hqvdp layer
+ *
+ * Check if the HW is able to perform the scaling request
+ * The firmware scaling limitation is "CEIL(1/Zy) <= FLOOR(LFW)" where:
+ *   Zy = OutputHeight / InputHeight
+ *   LFW = (Tx * IPClock) / (MaxNbCycles * Cp)
+ *     Tx : Total video mode horizontal resolution
+ *     IPClock : HQVDP IP clock (Mhz)
+ *     MaxNbCycles: max(InputWidth, OutputWidth)
+ *     Cp: Video mode pixel clock (Mhz)
+ *
+ * RETURNS:
+ * True if the HW can scale.
+ */
+static bool sti_hqvdp_check_hw_scaling(struct sti_layer *layer)
+{
+	struct sti_hqvdp *hqvdp = to_sti_hqvdp(layer);
+	unsigned long lfw;
+	unsigned int inv_zy;
+
+	lfw = layer->mode->htotal * (clk_get_rate(hqvdp->clk) / 1000000);
+	lfw /= max(layer->src_w, layer->dst_w) * layer->mode->clock / 1000;
+
+	inv_zy = DIV_ROUND_UP(layer->src_h, layer->dst_h);
+
+	return (inv_zy <= lfw) ? true : false;
+}
+
+/**
+ * sti_hqvdp_prepare_layer
+ * @layer: hqvdp layer
+ * @first_prepare: true if it is the first time this function is called
+ *
+ * Prepares a command for the firmware
+ *
+ * RETURNS:
+ * 0 on success.
+ */
+static int sti_hqvdp_prepare_layer(struct sti_layer *layer, bool first_prepare)
+{
+	struct sti_hqvdp *hqvdp = to_sti_hqvdp(layer);
+	struct sti_hqvdp_cmd *cmd;
+	int scale_h, scale_v;
+	int cmd_offset;
+
+	dev_dbg(hqvdp->dev, "%s %s\n", __func__, sti_layer_to_str(layer));
+
+	/* prepare and commit VID plane */
+	hqvdp->vid_plane->funcs->update_plane(hqvdp->vid_plane,
+					layer->crtc, layer->fb,
+					layer->dst_x, layer->dst_y,
+					layer->dst_w, layer->dst_h,
+					layer->src_x, layer->src_y,
+					layer->src_w, layer->src_h);
+
+	cmd_offset = sti_hqvdp_get_free_cmd(hqvdp);
+	if (cmd_offset == -1) {
+		DRM_ERROR("No available hqvdp_cmd now\n");
+		return -EBUSY;
+	}
+	cmd = hqvdp->hqvdp_cmd + cmd_offset;
+
+	if (!sti_hqvdp_check_hw_scaling(layer)) {
+		DRM_ERROR("Scaling beyond HW capabilities\n");
+		return -EINVAL;
+	}
+
+	/* Static parameters, defaulting to progressive mode */
+	cmd->top.config = TOP_CONFIG_PROGRESSIVE;
+	cmd->top.mem_format = TOP_MEM_FORMAT_DFLT;
+	cmd->hvsrc.param_ctrl = HVSRC_PARAM_CTRL_DFLT;
+	cmd->csdi.config = CSDI_CONFIG_PROG;
+
+	/* VC1RE, FMD bypassed : keep everything set to 0
+	 * IQI/P2I bypassed */
+	cmd->iqi.config = IQI_CONFIG_DFLT;
+	cmd->iqi.con_bri = IQI_CON_BRI_DFLT;
+	cmd->iqi.sat_gain = IQI_SAT_GAIN_DFLT;
+	cmd->iqi.pxf_conf = IQI_PXF_CONF_DFLT;
+
+	/* Buffer planes address */
+	cmd->top.current_luma = (u32) layer->paddr + layer->offsets[0];
+	cmd->top.current_chroma = (u32) layer->paddr + layer->offsets[1];
+
+	/* Pitches */
+	cmd->top.luma_processed_pitch = cmd->top.luma_src_pitch =
+			layer->pitches[0];
+	cmd->top.chroma_processed_pitch = cmd->top.chroma_src_pitch =
+			layer->pitches[1];
+
+	/* Input / output size
+	 * Align to upper even value */
+	layer->dst_w = ALIGN(layer->dst_w, 2);
+	layer->dst_h = ALIGN(layer->dst_h, 2);
+
+	if ((layer->src_w > MAX_WIDTH) || (layer->src_w < MIN_WIDTH) ||
+	    (layer->src_h > MAX_HEIGHT) || (layer->src_h < MIN_HEIGHT) ||
+	    (layer->dst_w > MAX_WIDTH) || (layer->dst_w < MIN_WIDTH) ||
+	    (layer->dst_h > MAX_HEIGHT) || (layer->dst_h < MIN_HEIGHT)) {
+		DRM_ERROR("Invalid in/out size %dx%d -> %dx%d\n",
+				layer->src_w, layer->src_h,
+				layer->dst_w, layer->dst_h);
+		return -EINVAL;
+	}
+	cmd->top.input_viewport_size = cmd->top.input_frame_size =
+			layer->src_h << 16 | layer->src_w;
+	cmd->hvsrc.output_picture_size = layer->dst_h << 16 | layer->dst_w;
+	cmd->top.input_viewport_ori = layer->src_y << 16 | layer->src_x;
+
+	/* Handle interlaced */
+	if (layer->fb->flags & DRM_MODE_FB_INTERLACED) {
+		/* Top field to display */
+		cmd->top.config = TOP_CONFIG_INTER_TOP;
+
+		/* Update pitches and vert size */
+		cmd->top.input_frame_size = (layer->src_h / 2) << 16 |
+					     layer->src_w;
+		cmd->top.luma_processed_pitch *= 2;
+		cmd->top.luma_src_pitch *= 2;
+		cmd->top.chroma_processed_pitch *= 2;
+		cmd->top.chroma_src_pitch *= 2;
+
+		/* Enable directional deinterlacing processing */
+		cmd->csdi.config = CSDI_CONFIG_INTER_DIR;
+		cmd->csdi.config2 = CSDI_CONFIG2_DFLT;
+		cmd->csdi.dcdi_config = CSDI_DCDI_CONFIG_DFLT;
+	}
+
+	/* Update hvsrc lut coef */
+	scale_h = SCALE_FACTOR * layer->dst_w / layer->src_w;
+	sti_hqvdp_update_hvsrc(HVSRC_HORI, scale_h, &cmd->hvsrc);
+
+	scale_v = SCALE_FACTOR * layer->dst_h / layer->src_h;
+	sti_hqvdp_update_hvsrc(HVSRC_VERT, scale_v, &cmd->hvsrc);
+
+	if (first_prepare) {
+		/* Prevent VTG shutdown */
+		if (clk_prepare_enable(hqvdp->clk_pix_main)) {
+			DRM_ERROR("Failed to prepare/enable pix main clk\n");
+			return -ENXIO;
+		}
+
+		/* Register VTG Vsync callback to handle bottom fields */
+		if ((layer->fb->flags & DRM_MODE_FB_INTERLACED) &&
+				sti_vtg_register_client(hqvdp->vtg,
+					&hqvdp->vtg_nb, layer->mixer_id)) {
+			DRM_ERROR("Cannot register VTG notifier\n");
+			return -ENXIO;
+		}
+	}
+
+	return 0;
+}
+
+static int sti_hqvdp_commit_layer(struct sti_layer *layer)
+{
+	struct sti_hqvdp *hqvdp = to_sti_hqvdp(layer);
+	int cmd_offset;
+
+	dev_dbg(hqvdp->dev, "%s %s\n", __func__, sti_layer_to_str(layer));
+
+	cmd_offset = sti_hqvdp_get_free_cmd(hqvdp);
+	if (cmd_offset == -1) {
+		DRM_ERROR("No available hqvdp_cmd now\n");
+		return -EBUSY;
+	}
+
+	writel(hqvdp->hqvdp_cmd_paddr + cmd_offset,
+			hqvdp->regs + HQVDP_MBX_NEXT_CMD);
+
+	hqvdp->curr_field_count++;
+
+	/* Interlaced : get ready to display the bottom field at next Vsync */
+	if (layer->fb->flags & DRM_MODE_FB_INTERLACED)
+		hqvdp->btm_field_pending = true;
+
+	dev_dbg(hqvdp->dev, "%s Posted command:0x%x\n",
+			__func__, hqvdp->hqvdp_cmd_paddr + cmd_offset);
+
+	return 0;
+}
+
+static int sti_hqvdp_disable_layer(struct sti_layer *layer)
+{
+	struct sti_hqvdp *hqvdp = to_sti_hqvdp(layer);
+	int i;
+
+	DRM_DEBUG_DRIVER("%s\n", sti_layer_to_str(layer));
+
+	/* Unregister VTG Vsync callback */
+	if ((layer->fb->flags & DRM_MODE_FB_INTERLACED) &&
+		sti_vtg_unregister_client(hqvdp->vtg, &hqvdp->vtg_nb))
+		DRM_DEBUG_DRIVER("Warning: cannot unregister VTG notifier\n");
+
+	/* Set next cmd to NULL */
+	writel(0, hqvdp->regs + HQVDP_MBX_NEXT_CMD);
+
+	for (i = 0; i < POLL_MAX_ATTEMPT; i++) {
+		if (readl(hqvdp->regs + HQVDP_MBX_INFO_XP70)
+				& INFO_XP70_FW_READY)
+			break;
+		msleep(POLL_DELAY_MS);
+	}
+
+	/* VTG can stop now */
+	clk_disable_unprepare(hqvdp->clk_pix_main);
+
+	if (i == POLL_MAX_ATTEMPT) {
+		DRM_ERROR("XP70 could not revert to idle\n");
+		return -ENXIO;
+	}
+
+	/* disable VID plane */
+	hqvdp->vid_plane->funcs->disable_plane(hqvdp->vid_plane);
+
+	return 0;
+}
+
+/**
+ * sti_vdp_vtg_cb
+ * @nb: notifier block
+ * @evt: event message
+ * @data: private data
+ *
+ * Handle VTG Vsync event, display pending bottom field
+ *
+ * RETURNS:
+ * 0 on success.
+ */
+int sti_hqvdp_vtg_cb(struct notifier_block *nb, unsigned long evt, void *data)
+{
+	struct sti_hqvdp *hqvdp = container_of(nb, struct sti_hqvdp, vtg_nb);
+	int btm_cmd_offset, top_cmd_offest;
+	struct sti_hqvdp_cmd *btm_cmd, *top_cmd;
+
+	if ((evt != VTG_TOP_FIELD_EVENT) && (evt != VTG_BOTTOM_FIELD_EVENT)) {
+		DRM_DEBUG_DRIVER("Unknown event\n");
+		return 0;
+	}
+
+	if (hqvdp->btm_field_pending) {
+		/* Create the btm field command from the current one */
+		btm_cmd_offset = sti_hqvdp_get_free_cmd(hqvdp);
+		top_cmd_offest = sti_hqvdp_get_curr_cmd(hqvdp);
+		if ((btm_cmd_offset == -1) || (top_cmd_offest == -1)) {
+			DRM_ERROR("Cannot get cmds, skip btm field\n");
+			return -EBUSY;
+		}
+
+		btm_cmd = hqvdp->hqvdp_cmd + btm_cmd_offset;
+		top_cmd = hqvdp->hqvdp_cmd + top_cmd_offest;
+
+		memcpy(btm_cmd, top_cmd, sizeof(*btm_cmd));
+
+		btm_cmd->top.config = TOP_CONFIG_INTER_BTM;
+		btm_cmd->top.current_luma +=
+				btm_cmd->top.luma_src_pitch / 2;
+		btm_cmd->top.current_chroma +=
+				btm_cmd->top.chroma_src_pitch / 2;
+
+		/* Post the command to mailbox */
+		writel(hqvdp->hqvdp_cmd_paddr + btm_cmd_offset,
+				hqvdp->regs + HQVDP_MBX_NEXT_CMD);
+
+		hqvdp->curr_field_count++;
+		hqvdp->btm_field_pending = false;
+
+		dev_dbg(hqvdp->dev, "%s Posted command:0x%x\n",
+				__func__, hqvdp->hqvdp_cmd_paddr);
+	}
+
+	return 0;
+}
+
+static struct drm_plane *sti_hqvdp_find_vid(struct drm_device *dev, int id)
+{
+	struct drm_plane *plane;
+
+	list_for_each_entry(plane, &dev->mode_config.plane_list, head) {
+		struct sti_layer *layer = to_sti_layer(plane);
+
+		if (layer->desc == id)
+			return plane;
+	}
+
+	return NULL;
+}
+
+static void sti_hqvd_init(struct sti_layer *layer)
+{
+	struct sti_hqvdp *hqvdp = to_sti_hqvdp(layer);
+	int size;
+
+	/* find the plane macthing with vid 0 */
+	hqvdp->vid_plane = sti_hqvdp_find_vid(hqvdp->drm_dev, STI_VID_0);
+	if (!hqvdp->vid_plane) {
+		DRM_ERROR("Cannot find Main video layer\n");
+		return;
+	}
+
+	hqvdp->vtg_nb.notifier_call = sti_hqvdp_vtg_cb;
+
+	/* Allocate memory for the VDP commands */
+	size = NB_VDP_CMD * sizeof(struct sti_hqvdp_cmd);
+	hqvdp->hqvdp_cmd = dma_alloc_writecombine(hqvdp->dev, size,
+					 &hqvdp->hqvdp_cmd_paddr,
+					 GFP_KERNEL | GFP_DMA);
+	if (!hqvdp->hqvdp_cmd) {
+		DRM_ERROR("Failed to allocate memory for VDP cmd\n");
+		return;
+	}
+
+	memset(hqvdp->hqvdp_cmd, 0, size);
+}
+
+static const struct sti_layer_funcs hqvdp_ops = {
+	.get_formats = sti_hqvdp_get_formats,
+	.get_nb_formats = sti_hqvdp_get_nb_formats,
+	.init = sti_hqvd_init,
+	.prepare = sti_hqvdp_prepare_layer,
+	.commit = sti_hqvdp_commit_layer,
+	.disable = sti_hqvdp_disable_layer,
+};
+
+struct sti_layer *sti_hqvdp_create(struct device *dev)
+{
+	struct sti_hqvdp *hqvdp = dev_get_drvdata(dev);
+
+	hqvdp->layer.ops = &hqvdp_ops;
+
+	return &hqvdp->layer;
+}
+EXPORT_SYMBOL(sti_hqvdp_create);
+
+static void sti_hqvdp_init_plugs(struct sti_hqvdp *hqvdp)
+{
+	/* Configure Plugs (same for RD & WR) */
+	writel(PLUG_PAGE_SIZE_256, hqvdp->regs + HQVDP_RD_PLUG_PAGE_SIZE);
+	writel(PLUG_MIN_OPC_8, hqvdp->regs + HQVDP_RD_PLUG_MIN_OPC);
+	writel(PLUG_MAX_OPC_64, hqvdp->regs + HQVDP_RD_PLUG_MAX_OPC);
+	writel(PLUG_MAX_CHK_2X, hqvdp->regs + HQVDP_RD_PLUG_MAX_CHK);
+	writel(PLUG_MAX_MSG_1X, hqvdp->regs + HQVDP_RD_PLUG_MAX_MSG);
+	writel(PLUG_MIN_SPACE_1, hqvdp->regs + HQVDP_RD_PLUG_MIN_SPACE);
+	writel(PLUG_CONTROL_ENABLE, hqvdp->regs + HQVDP_RD_PLUG_CONTROL);
+
+	writel(PLUG_PAGE_SIZE_256, hqvdp->regs + HQVDP_WR_PLUG_PAGE_SIZE);
+	writel(PLUG_MIN_OPC_8, hqvdp->regs + HQVDP_WR_PLUG_MIN_OPC);
+	writel(PLUG_MAX_OPC_64, hqvdp->regs + HQVDP_WR_PLUG_MAX_OPC);
+	writel(PLUG_MAX_CHK_2X, hqvdp->regs + HQVDP_WR_PLUG_MAX_CHK);
+	writel(PLUG_MAX_MSG_1X, hqvdp->regs + HQVDP_WR_PLUG_MAX_MSG);
+	writel(PLUG_MIN_SPACE_1, hqvdp->regs + HQVDP_WR_PLUG_MIN_SPACE);
+	writel(PLUG_CONTROL_ENABLE, hqvdp->regs + HQVDP_WR_PLUG_CONTROL);
+}
+
+/**
+ * sti_hqvdp_start_xp70
+ * @firmware: firmware found
+ * @ctxt:     hqvdp structure
+ *
+ * Run the xP70 initialization sequence
+ */
+static void sti_hqvdp_start_xp70(const struct firmware *firmware, void *ctxt)
+{
+	struct sti_hqvdp *hqvdp = ctxt;
+	u32 *fw_rd_plug, *fw_wr_plug, *fw_pmem, *fw_dmem;
+	u8 *data;
+	int i;
+	struct fw_header {
+		int rd_size;
+		int wr_size;
+		int pmem_size;
+		int dmem_size;
+	} *header;
+
+	DRM_DEBUG_DRIVER("\n");
+	/* Check firmware parts */
+	if (!firmware) {
+		DRM_ERROR("Firmware not available\n");
+		return;
+	}
+
+	header = (struct fw_header *) firmware->data;
+	if (firmware->size < sizeof(*header)) {
+		DRM_ERROR("Invalid firmware size (%d)\n", firmware->size);
+		goto out;
+	}
+	if ((sizeof(*header) + header->rd_size + header->wr_size +
+		header->pmem_size + header->dmem_size) != firmware->size) {
+		DRM_ERROR("Invalid fmw structure (%d+%d+%d+%d+%d != %d)\n",
+			   sizeof(*header), header->rd_size, header->wr_size,
+			   header->pmem_size, header->dmem_size,
+			   firmware->size);
+		goto out;
+	}
+
+	data = (u8 *) firmware->data;
+	data += sizeof(*header);
+	fw_rd_plug = (void *) data;
+	data += header->rd_size;
+	fw_wr_plug = (void *) data;
+	data += header->wr_size;
+	fw_pmem = (void *) data;
+	data += header->pmem_size;
+	fw_dmem = (void *) data;
+
+	/* Enable clock */
+	if (clk_prepare_enable(hqvdp->clk))
+		DRM_ERROR("Failed to prepare/enable HQVDP clk\n");
+
+	/* Reset */
+	writel(SW_RESET_CTRL_FULL, hqvdp->regs + HQVDP_MBX_SW_RESET_CTRL);
+
+	for (i = 0; i < POLL_MAX_ATTEMPT; i++) {
+		if (readl(hqvdp->regs + HQVDP_MBX_STARTUP_CTRL1)
+				& STARTUP_CTRL1_RST_DONE)
+			break;
+		msleep(POLL_DELAY_MS);
+	}
+	if (i == POLL_MAX_ATTEMPT) {
+		DRM_ERROR("Could not reset\n");
+		goto out;
+	}
+
+	/* Init Read & Write plugs */
+	for (i = 0; i < header->rd_size / 4; i++)
+		writel(fw_rd_plug[i], hqvdp->regs + HQVDP_RD_PLUG + i * 4);
+	for (i = 0; i < header->wr_size / 4; i++)
+		writel(fw_wr_plug[i], hqvdp->regs + HQVDP_WR_PLUG + i * 4);
+
+	sti_hqvdp_init_plugs(hqvdp);
+
+	/* Authorize Idle Mode */
+	writel(STARTUP_CTRL1_AUTH_IDLE, hqvdp->regs + HQVDP_MBX_STARTUP_CTRL1);
+
+	/* Prevent VTG interruption during the boot */
+	writel(SOFT_VSYNC_SW_CTRL_IRQ, hqvdp->regs + HQVDP_MBX_SOFT_VSYNC);
+	writel(0, hqvdp->regs + HQVDP_MBX_NEXT_CMD);
+
+	/* Download PMEM & DMEM */
+	for (i = 0; i < header->pmem_size / 4; i++)
+		writel(fw_pmem[i], hqvdp->regs + HQVDP_PMEM + i * 4);
+	for (i = 0; i < header->dmem_size / 4; i++)
+		writel(fw_dmem[i], hqvdp->regs + HQVDP_DMEM + i * 4);
+
+	/* Enable fetch */
+	writel(STARTUP_CTRL2_FETCH_EN, hqvdp->regs + HQVDP_MBX_STARTUP_CTRL2);
+
+	/* Wait end of boot */
+	for (i = 0; i < POLL_MAX_ATTEMPT; i++) {
+		if (readl(hqvdp->regs + HQVDP_MBX_INFO_XP70)
+				& INFO_XP70_FW_READY)
+			break;
+		msleep(POLL_DELAY_MS);
+	}
+	if (i == POLL_MAX_ATTEMPT) {
+		DRM_ERROR("Could not boot\n");
+		goto out;
+	}
+
+	/* Launch Vsync */
+	writel(SOFT_VSYNC_HW, hqvdp->regs + HQVDP_MBX_SOFT_VSYNC);
+
+	DRM_INFO("HQVDP XP70 started\n");
+out:
+	release_firmware(firmware);
+}
+
+int sti_hqvdp_bind(struct device *dev, struct device *master, void *data)
+{
+	struct sti_hqvdp *hqvdp = dev_get_drvdata(dev);
+	struct drm_device *drm_dev = data;
+	struct sti_layer *layer;
+	int err;
+
+	DRM_DEBUG_DRIVER("\n");
+
+	hqvdp->drm_dev = drm_dev;
+
+	/* Request for firmware */
+	err = request_firmware_nowait(THIS_MODULE, FW_ACTION_HOTPLUG,
+				HQVDP_FMW_NAME,	hqvdp->dev,
+				GFP_KERNEL, hqvdp, sti_hqvdp_start_xp70);
+	if (err) {
+		DRM_ERROR("Can't get HQVDP firmware\n");
+		return err;
+	}
+
+	layer = sti_layer_create(hqvdp->dev, STI_HQVDP_0, hqvdp->regs);
+	if (!layer) {
+		DRM_ERROR("Can't create HQVDP plane\n");
+		return -ENOMEM;
+	}
+
+	sti_drm_plane_init(drm_dev, layer, 1, DRM_PLANE_TYPE_OVERLAY);
+
+	return 0;
+}
+
+static void sti_hqvdp_unbind(struct device *dev,
+		struct device *master, void *data)
+{
+	/* do nothing */
+}
+
+static const struct component_ops sti_hqvdp_ops = {
+	.bind = sti_hqvdp_bind,
+	.unbind = sti_hqvdp_unbind,
+};
+
+static int sti_hqvdp_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct device_node *vtg_np;
+	struct sti_hqvdp *hqvdp;
+	struct resource *res;
+
+	DRM_DEBUG_DRIVER("\n");
+
+	hqvdp = devm_kzalloc(dev, sizeof(*hqvdp), GFP_KERNEL);
+	if (!hqvdp) {
+		DRM_ERROR("Failed to allocate HQVDP context\n");
+		return -ENOMEM;
+	}
+
+	hqvdp->dev = dev;
+
+	/* Get Memory resources */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (res == NULL) {
+		DRM_ERROR("Get memory resource failed\n");
+		return -ENXIO;
+	}
+	hqvdp->regs = devm_ioremap(dev, res->start, resource_size(res));
+	if (hqvdp->regs == NULL) {
+		DRM_ERROR("Register mapping failed\n");
+		return -ENXIO;
+	}
+
+	/* Get clock resources */
+	hqvdp->clk = devm_clk_get(dev, "hqvdp");
+	hqvdp->clk_pix_main = devm_clk_get(dev, "pix_main");
+	if (IS_ERR(hqvdp->clk) || IS_ERR(hqvdp->clk)) {
+		DRM_ERROR("Cannot get clocks\n");
+		return -ENXIO;
+	}
+
+	/* Get reset resources */
+	hqvdp->reset = devm_reset_control_get(dev, "hqvdp");
+	if (!IS_ERR(hqvdp->reset))
+		reset_control_deassert(hqvdp->reset);
+
+	vtg_np = of_parse_phandle(pdev->dev.of_node, "st,vtg", 0);
+	if (vtg_np)
+		hqvdp->vtg = of_vtg_find(vtg_np);
+
+	platform_set_drvdata(pdev, hqvdp);
+
+	return component_add(&pdev->dev, &sti_hqvdp_ops);
+}
+
+static int sti_hqvdp_remove(struct platform_device *pdev)
+{
+	component_del(&pdev->dev, &sti_hqvdp_ops);
+	return 0;
+}
+
+static struct of_device_id hqvdp_of_match[] = {
+	{ .compatible = "st,stih407-hqvdp", },
+	{ /* end node */ }
+};
+MODULE_DEVICE_TABLE(of, hqvdp_of_match);
+
+struct platform_driver sti_hqvdp_driver = {
+	.driver = {
+		.name = "sti-hqvdp",
+		.owner = THIS_MODULE,
+		.of_match_table = hqvdp_of_match,
+	},
+	.probe = sti_hqvdp_probe,
+	.remove = sti_hqvdp_remove,
+};
+
+module_platform_driver(sti_hqvdp_driver);
+
+MODULE_AUTHOR("Benjamin Gaignard <benjamin.gaignard@st.com>");
+MODULE_DESCRIPTION("STMicroelectronics SoC DRM driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/sti/sti_hqvdp.h b/drivers/gpu/drm/sti/sti_hqvdp.h
new file mode 100644
index 000000000000..cd5ecd0a6dea
--- /dev/null
+++ b/drivers/gpu/drm/sti/sti_hqvdp.h
@@ -0,0 +1,12 @@
+/*
+ * Copyright (C) STMicroelectronics SA 2014
+ * Authors: Fabien Dessenne <fabien.dessenne@st.com> for STMicroelectronics.
+ * License terms:  GNU General Public License (GPL), version 2
+ */
+
+#ifndef _STI_HQVDP_H_
+#define _STI_HQVDP_H_
+
+struct sti_layer *sti_hqvdp_create(struct device *dev);
+
+#endif
diff --git a/drivers/gpu/drm/sti/sti_hqvdp_lut.h b/drivers/gpu/drm/sti/sti_hqvdp_lut.h
new file mode 100644
index 000000000000..619af7f4384e
--- /dev/null
+++ b/drivers/gpu/drm/sti/sti_hqvdp_lut.h
@@ -0,0 +1,373 @@
+/*
+ * Copyright (C) STMicroelectronics SA 2014
+ * Authors: Fabien Dessenne <fabien.dessenne@st.com> for STMicroelectronics.
+ * License terms:  GNU General Public License (GPL), version 2
+ */
+
+#ifndef _STI_HQVDP_LUT_H_
+#define _STI_HQVDP_LUT_H_
+
+#define NB_COEF                 128
+
+#define SHIFT_LUT_A_LEGACY      8
+#define SHIFT_LUT_B             8
+#define SHIFT_LUT_C_Y_LEGACY    8
+#define SHIFT_LUT_C_C_LEGACY    8
+#define SHIFT_LUT_D_Y_LEGACY    8
+#define SHIFT_LUT_D_C_LEGACY    8
+#define SHIFT_LUT_E_Y_LEGACY    8
+#define SHIFT_LUT_E_C_LEGACY    8
+#define SHIFT_LUT_F_Y_LEGACY    8
+#define SHIFT_LUT_F_C_LEGACY    8
+
+static const u32 coef_lut_a_legacy[NB_COEF] = {
+	0x0000ffff, 0x00010000, 0x000100ff, 0x00000000,
+	0x00000000, 0x00050000, 0xfffc00ff, 0x00000000,
+	0x00000000, 0x00090000, 0xfff900fe, 0x00000000,
+	0x00000000, 0x0010ffff, 0xfff600fb, 0x00000000,
+	0x00000000, 0x0017fffe, 0xfff400f7, 0x00000000,
+	0x00000000, 0x001ffffd, 0xfff200f2, 0x00000000,
+	0x00000000, 0x0027fffc, 0xfff100ec, 0x00000000,
+	0x00000000, 0x0030fffb, 0xfff000e5, 0x00000000,
+	0x00000000, 0x003afffa, 0xffee00de, 0x00000000,
+	0x00000000, 0x0044fff9, 0xffed00d6, 0x00000000,
+	0x00000000, 0x004efff8, 0xffed00cd, 0x00000000,
+	0x00000000, 0x0059fff6, 0xffed00c4, 0x00000000,
+	0x00000000, 0x0064fff5, 0xffed00ba, 0x00000000,
+	0x00000000, 0x006ffff3, 0xffee00b0, 0x00000000,
+	0x00000000, 0x007afff2, 0xffee00a6, 0x00000000,
+	0x00000000, 0x0085fff1, 0xffef009b, 0x00000000,
+	0x00000000, 0x0090fff0, 0xfff00090, 0x00000000,
+	0x00000000, 0x009bffef, 0xfff10085, 0x00000000,
+	0x00000000, 0x00a6ffee, 0xfff2007a, 0x00000000,
+	0x00000000, 0x00b0ffee, 0xfff3006f, 0x00000000,
+	0x00000000, 0x00baffed, 0xfff50064, 0x00000000,
+	0x00000000, 0x00c4ffed, 0xfff60059, 0x00000000,
+	0x00000000, 0x00cdffed, 0xfff8004e, 0x00000000,
+	0x00000000, 0x00d6ffed, 0xfff90044, 0x00000000,
+	0x00000000, 0x00deffee, 0xfffa003a, 0x00000000,
+	0x00000000, 0x00e5fff0, 0xfffb0030, 0x00000000,
+	0x00000000, 0x00ecfff1, 0xfffc0027, 0x00000000,
+	0x00000000, 0x00f2fff2, 0xfffd001f, 0x00000000,
+	0x00000000, 0x00f7fff4, 0xfffe0017, 0x00000000,
+	0x00000000, 0x00fbfff6, 0xffff0010, 0x00000000,
+	0x00000000, 0x00fefff9, 0x00000009, 0x00000000,
+	0x00000000, 0x00fffffc, 0x00000005, 0x00000000
+};
+
+static const u32 coef_lut_b[NB_COEF] = {
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000,
+	0x00000000, 0x00000000, 0x00000100, 0x00000000
+};
+
+static const u32 coef_lut_c_y_legacy[NB_COEF] = {
+	0x00060004, 0x0038ffe1, 0x003800be, 0x0006ffe1,
+	0x00050005, 0x0042ffe1, 0x003800b3, 0x0007ffe1,
+	0x00040006, 0x0046ffe1, 0x003300b2, 0x0008ffe2,
+	0x00030007, 0x004cffe1, 0x002e00b1, 0x0008ffe2,
+	0x00020006, 0x0051ffe2, 0x002900b0, 0x0009ffe3,
+	0x00010008, 0x0056ffe2, 0x002400ae, 0x0009ffe4,
+	0xffff0008, 0x005cffe3, 0x001f00ad, 0x000affe4,
+	0xfffe0008, 0x0062ffe4, 0x001a00ab, 0x000affe5,
+	0xfffd000a, 0x0066ffe5, 0x001500a8, 0x000bffe6,
+	0xfffc0009, 0x006bffe7, 0x001100a5, 0x000bffe8,
+	0xfffa000a, 0x0070ffe8, 0x000d00a3, 0x000bffe9,
+	0xfff9000b, 0x0076ffea, 0x0008009f, 0x000bffea,
+	0xfff7000b, 0x007affec, 0x0005009b, 0x000cffec,
+	0xfff6000b, 0x007effef, 0x00010098, 0x000cffed,
+	0xfff4000b, 0x0084fff1, 0xfffd0095, 0x000cffee,
+	0xfff3000b, 0x0088fff4, 0xfffa0090, 0x000cfff0,
+	0xfff1000b, 0x008dfff7, 0xfff7008d, 0x000bfff1,
+	0xfff0000c, 0x0090fffa, 0xfff40088, 0x000bfff3,
+	0xffee000c, 0x0095fffd, 0xfff10084, 0x000bfff4,
+	0xffed000c, 0x00980001, 0xffef007e, 0x000bfff6,
+	0xffec000c, 0x009b0005, 0xffec007a, 0x000bfff7,
+	0xffea000b, 0x009f0008, 0xffea0076, 0x000bfff9,
+	0xffe9000b, 0x00a3000d, 0xffe80070, 0x000afffa,
+	0xffe8000b, 0x00a50011, 0xffe7006b, 0x0009fffc,
+	0xffe6000b, 0x00a80015, 0xffe50066, 0x000afffd,
+	0xffe5000a, 0x00ab001a, 0xffe40062, 0x0008fffe,
+	0xffe4000a, 0x00ad001f, 0xffe3005c, 0x0008ffff,
+	0xffe40009, 0x00ae0024, 0xffe20056, 0x00080001,
+	0xffe30009, 0x00b00029, 0xffe20051, 0x00060002,
+	0xffe20008, 0x00b1002e, 0xffe1004c, 0x00070003,
+	0xffe20008, 0x00b20033, 0xffe10046, 0x00060004,
+	0xffe10007, 0x00b30038, 0xffe10042, 0x00050005
+};
+
+static const u32 coef_lut_c_c_legacy[NB_COEF] = {
+	0x0001fff3, 0x003afffb, 0x003a00a1, 0x0001fffb,
+	0x0001fff5, 0x0041fffb, 0x0038009a, 0x0001fffb,
+	0x0001fff5, 0x0046fffb, 0x00340099, 0x0001fffb,
+	0x0001fff7, 0x0049fffb, 0x00300098, 0x0001fffb,
+	0x0001fff9, 0x004cfffb, 0x002d0096, 0x0001fffb,
+	0x0001fffa, 0x004ffffc, 0x00290095, 0x0001fffb,
+	0x0001fff9, 0x0054fffd, 0x00250093, 0x0001fffc,
+	0x0001fffa, 0x0058fffd, 0x00220092, 0x0000fffc,
+	0x0001fffb, 0x005bfffe, 0x001f0090, 0x0000fffc,
+	0x0001fffd, 0x005effff, 0x001c008c, 0x0000fffd,
+	0x0001fffd, 0x00620000, 0x0019008a, 0x0000fffd,
+	0x0001fffe, 0x00660001, 0x00160088, 0xfffffffd,
+	0x0000fffe, 0x006a0003, 0x00130085, 0xfffffffe,
+	0x0000fffe, 0x006e0004, 0x00100083, 0xfffffffe,
+	0x0000fffe, 0x00710006, 0x000e007f, 0xffffffff,
+	0x0000fffe, 0x00750008, 0x000c007c, 0xfffeffff,
+	0xfffffffe, 0x0079000a, 0x000a0079, 0xfffeffff,
+	0xfffffffe, 0x007c000c, 0x00080075, 0xfffe0000,
+	0xffffffff, 0x007f000e, 0x00060071, 0xfffe0000,
+	0xfffeffff, 0x00830010, 0x0004006e, 0xfffe0000,
+	0xfffeffff, 0x00850013, 0x0003006a, 0xfffe0000,
+	0xfffdffff, 0x00880016, 0x00010066, 0xfffe0001,
+	0xfffd0000, 0x008a0019, 0x00000062, 0xfffd0001,
+	0xfffd0000, 0x008c001c, 0xffff005e, 0xfffd0001,
+	0xfffc0000, 0x0090001f, 0xfffe005b, 0xfffb0001,
+	0xfffc0000, 0x00920022, 0xfffd0058, 0xfffa0001,
+	0xfffc0001, 0x00930025, 0xfffd0054, 0xfff90001,
+	0xfffb0001, 0x00950029, 0xfffc004f, 0xfffa0001,
+	0xfffb0001, 0x0096002d, 0xfffb004c, 0xfff90001,
+	0xfffb0001, 0x00980030, 0xfffb0049, 0xfff70001,
+	0xfffb0001, 0x00990034, 0xfffb0046, 0xfff50001,
+	0xfffb0001, 0x009a0038, 0xfffb0041, 0xfff50001
+};
+
+static const u32 coef_lut_d_y_legacy[NB_COEF] = {
+	0xfff80009, 0x0046ffec, 0x004600a3, 0xfff8ffec,
+	0xfff70009, 0x004effed, 0x0044009d, 0xfff9ffeb,
+	0xfff6000a, 0x0052ffee, 0x003f009d, 0xfffaffea,
+	0xfff50009, 0x0057ffef, 0x003b009d, 0xfffbffe9,
+	0xfff50008, 0x005bfff0, 0x0037009c, 0xfffcffe9,
+	0xfff40008, 0x005ffff2, 0x0033009b, 0xfffcffe9,
+	0xfff30007, 0x0064fff3, 0x002f009b, 0xfffdffe8,
+	0xfff20007, 0x0068fff5, 0x002b0099, 0xfffeffe8,
+	0xfff10008, 0x006bfff7, 0x00270097, 0xffffffe8,
+	0xfff00007, 0x006ffff9, 0x00230097, 0xffffffe8,
+	0xffef0006, 0x0073fffb, 0x00200095, 0x0000ffe8,
+	0xffee0005, 0x0077fffe, 0x001c0093, 0x0000ffe9,
+	0xffee0005, 0x007a0000, 0x00180091, 0x0001ffe9,
+	0xffed0005, 0x007d0003, 0x0015008e, 0x0002ffe9,
+	0xffec0005, 0x00800006, 0x0012008b, 0x0002ffea,
+	0xffeb0004, 0x00840008, 0x000e008a, 0x0003ffea,
+	0xffeb0003, 0x0087000b, 0x000b0087, 0x0003ffeb,
+	0xffea0003, 0x008a000e, 0x00080084, 0x0004ffeb,
+	0xffea0002, 0x008b0012, 0x00060080, 0x0005ffec,
+	0xffe90002, 0x008e0015, 0x0003007d, 0x0005ffed,
+	0xffe90001, 0x00910018, 0x0000007a, 0x0005ffee,
+	0xffe90000, 0x0093001c, 0xfffe0077, 0x0005ffee,
+	0xffe80000, 0x00950020, 0xfffb0073, 0x0006ffef,
+	0xffe8ffff, 0x00970023, 0xfff9006f, 0x0007fff0,
+	0xffe8ffff, 0x00970027, 0xfff7006b, 0x0008fff1,
+	0xffe8fffe, 0x0099002b, 0xfff50068, 0x0007fff2,
+	0xffe8fffd, 0x009b002f, 0xfff30064, 0x0007fff3,
+	0xffe9fffc, 0x009b0033, 0xfff2005f, 0x0008fff4,
+	0xffe9fffc, 0x009c0037, 0xfff0005b, 0x0008fff5,
+	0xffe9fffb, 0x009d003b, 0xffef0057, 0x0009fff5,
+	0xffeafffa, 0x009d003f, 0xffee0052, 0x000afff6,
+	0xffebfff9, 0x009d0044, 0xffed004e, 0x0009fff7
+};
+
+static const u32 coef_lut_d_c_legacy[NB_COEF] = {
+	0xfffeffff, 0x003fffff, 0x003f0089, 0xfffeffff,
+	0xfffe0000, 0x00460000, 0x0042007d, 0xfffffffe,
+	0xfffe0000, 0x00490001, 0x003f007d, 0xfffffffd,
+	0xfffd0001, 0x004b0002, 0x003c007d, 0x0000fffc,
+	0xfffd0001, 0x004e0003, 0x0039007c, 0x0000fffc,
+	0xfffc0001, 0x00510005, 0x0036007c, 0x0000fffb,
+	0xfffc0001, 0x00540006, 0x0033007b, 0x0001fffa,
+	0xfffc0003, 0x00550008, 0x00310078, 0x0001fffa,
+	0xfffb0003, 0x00580009, 0x002e0078, 0x0001fffa,
+	0xfffb0002, 0x005b000b, 0x002b0077, 0x0002fff9,
+	0xfffa0003, 0x005e000d, 0x00280075, 0x0002fff9,
+	0xfffa0002, 0x0060000f, 0x00260074, 0x0002fff9,
+	0xfffa0004, 0x00610011, 0x00230072, 0x0002fff9,
+	0xfffa0004, 0x00640013, 0x00200070, 0x0002fff9,
+	0xfff90004, 0x00660015, 0x001e006e, 0x0003fff9,
+	0xfff90004, 0x00680017, 0x001c006c, 0x0003fff9,
+	0xfff90003, 0x006b0019, 0x0019006b, 0x0003fff9,
+	0xfff90003, 0x006c001c, 0x00170068, 0x0004fff9,
+	0xfff90003, 0x006e001e, 0x00150066, 0x0004fff9,
+	0xfff90002, 0x00700020, 0x00130064, 0x0004fffa,
+	0xfff90002, 0x00720023, 0x00110061, 0x0004fffa,
+	0xfff90002, 0x00740026, 0x000f0060, 0x0002fffa,
+	0xfff90002, 0x00750028, 0x000d005e, 0x0003fffa,
+	0xfff90002, 0x0077002b, 0x000b005b, 0x0002fffb,
+	0xfffa0001, 0x0078002e, 0x00090058, 0x0003fffb,
+	0xfffa0001, 0x00780031, 0x00080055, 0x0003fffc,
+	0xfffa0001, 0x007b0033, 0x00060054, 0x0001fffc,
+	0xfffb0000, 0x007c0036, 0x00050051, 0x0001fffc,
+	0xfffc0000, 0x007c0039, 0x0003004e, 0x0001fffd,
+	0xfffc0000, 0x007d003c, 0x0002004b, 0x0001fffd,
+	0xfffdffff, 0x007d003f, 0x00010049, 0x0000fffe,
+	0xfffeffff, 0x007d0042, 0x00000046, 0x0000fffe
+};
+
+static const u32 coef_lut_e_y_legacy[NB_COEF] = {
+	0xfff10001, 0x00490004, 0x00490083, 0xfff10004,
+	0xfff10000, 0x00500006, 0x004b007b, 0xfff10002,
+	0xfff10000, 0x00530007, 0x0048007b, 0xfff10001,
+	0xfff10000, 0x00550009, 0x0046007a, 0xfff10000,
+	0xfff1fffe, 0x0058000b, 0x0043007b, 0xfff2fffe,
+	0xfff1ffff, 0x005a000d, 0x0040007a, 0xfff2fffd,
+	0xfff1fffd, 0x005d000f, 0x003e007a, 0xfff2fffc,
+	0xfff1fffd, 0x005f0011, 0x003b0079, 0xfff3fffb,
+	0xfff1fffc, 0x00610013, 0x00390079, 0xfff3fffa,
+	0xfff1fffb, 0x00640015, 0x00360079, 0xfff3fff9,
+	0xfff1fffa, 0x00660017, 0x00340078, 0xfff4fff8,
+	0xfff1fffb, 0x00680019, 0x00310077, 0xfff4fff7,
+	0xfff2fff9, 0x006a001b, 0x002f0076, 0xfff5fff6,
+	0xfff2fff9, 0x006c001e, 0x002c0075, 0xfff5fff5,
+	0xfff2fff9, 0x006d0020, 0x002a0073, 0xfff6fff5,
+	0xfff3fff7, 0x00700022, 0x00270073, 0xfff6fff4,
+	0xfff3fff7, 0x00710025, 0x00250071, 0xfff7fff3,
+	0xfff4fff6, 0x00730027, 0x00220070, 0xfff7fff3,
+	0xfff5fff6, 0x0073002a, 0x0020006d, 0xfff9fff2,
+	0xfff5fff5, 0x0075002c, 0x001e006c, 0xfff9fff2,
+	0xfff6fff5, 0x0076002f, 0x001b006a, 0xfff9fff2,
+	0xfff7fff4, 0x00770031, 0x00190068, 0xfffbfff1,
+	0xfff8fff4, 0x00780034, 0x00170066, 0xfffafff1,
+	0xfff9fff3, 0x00790036, 0x00150064, 0xfffbfff1,
+	0xfffafff3, 0x00790039, 0x00130061, 0xfffcfff1,
+	0xfffbfff3, 0x0079003b, 0x0011005f, 0xfffdfff1,
+	0xfffcfff2, 0x007a003e, 0x000f005d, 0xfffdfff1,
+	0xfffdfff2, 0x007a0040, 0x000d005a, 0xfffffff1,
+	0xfffefff2, 0x007b0043, 0x000b0058, 0xfffefff1,
+	0x0000fff1, 0x007a0046, 0x00090055, 0x0000fff1,
+	0x0001fff1, 0x007b0048, 0x00070053, 0x0000fff1,
+	0x0002fff1, 0x007b004b, 0x00060050, 0x0000fff1
+};
+
+static const u32 coef_lut_e_c_legacy[NB_COEF] = {
+	0xfffa0001, 0x003f0010, 0x003f006d, 0xfffa0010,
+	0xfffb0002, 0x00440011, 0x00440062, 0xfffa000e,
+	0xfffb0001, 0x00460013, 0x00420062, 0xfffa000d,
+	0xfffb0000, 0x00480014, 0x00410062, 0xfffa000c,
+	0xfffb0001, 0x00490015, 0x003f0061, 0xfffb000b,
+	0xfffb0000, 0x004b0017, 0x003d0061, 0xfffb000a,
+	0xfffb0000, 0x004d0018, 0x003b0062, 0xfffb0008,
+	0xfffcffff, 0x004f001a, 0x00390061, 0xfffb0007,
+	0xfffc0000, 0x004f001c, 0x00380060, 0xfffb0006,
+	0xfffcffff, 0x0052001d, 0x00360060, 0xfffb0005,
+	0xfffdfffe, 0x0053001f, 0x00340060, 0xfffb0004,
+	0xfffdfffe, 0x00540021, 0x0032005e, 0xfffc0004,
+	0xfffeffff, 0x00550022, 0x0030005d, 0xfffc0003,
+	0xfffeffff, 0x00560024, 0x002f005c, 0xfffc0002,
+	0xfffffffd, 0x00580026, 0x002d005c, 0xfffc0001,
+	0xfffffffd, 0x005a0027, 0x002b005c, 0xfffc0000,
+	0x0000fffd, 0x005a0029, 0x0029005a, 0xfffd0000,
+	0x0000fffc, 0x005c002b, 0x0027005a, 0xfffdffff,
+	0x0001fffc, 0x005c002d, 0x00260058, 0xfffdffff,
+	0x0002fffc, 0x005c002f, 0x00240056, 0xfffffffe,
+	0x0003fffc, 0x005d0030, 0x00220055, 0xfffffffe,
+	0x0004fffc, 0x005e0032, 0x00210054, 0xfffefffd,
+	0x0004fffb, 0x00600034, 0x001f0053, 0xfffefffd,
+	0x0005fffb, 0x00600036, 0x001d0052, 0xfffffffc,
+	0x0006fffb, 0x00600038, 0x001c004f, 0x0000fffc,
+	0x0007fffb, 0x00610039, 0x001a004f, 0xfffffffc,
+	0x0008fffb, 0x0062003b, 0x0018004d, 0x0000fffb,
+	0x000afffb, 0x0061003d, 0x0017004b, 0x0000fffb,
+	0x000bfffb, 0x0061003f, 0x00150049, 0x0001fffb,
+	0x000cfffa, 0x00620041, 0x00140048, 0x0000fffb,
+	0x000dfffa, 0x00620042, 0x00130046, 0x0001fffb,
+	0x000efffa, 0x00620044, 0x00110044, 0x0002fffb
+};
+
+static const u32 coef_lut_f_y_legacy[NB_COEF] = {
+	0xfff6fff0, 0x00490012, 0x0049006e, 0xfff60012,
+	0xfff7fff1, 0x004e0013, 0x00490068, 0xfff60010,
+	0xfff7fff2, 0x004f0015, 0x00470067, 0xfff6000f,
+	0xfff7fff5, 0x004f0017, 0x00450065, 0xfff6000e,
+	0xfff8fff5, 0x00500018, 0x00440065, 0xfff6000c,
+	0xfff8fff6, 0x0051001a, 0x00420064, 0xfff6000b,
+	0xfff8fff6, 0x0052001c, 0x00400064, 0xfff6000a,
+	0xfff9fff6, 0x0054001d, 0x003e0064, 0xfff60008,
+	0xfff9fff8, 0x0054001f, 0x003c0063, 0xfff60007,
+	0xfffafff8, 0x00550021, 0x003a0062, 0xfff60006,
+	0xfffbfff7, 0x00560022, 0x00390062, 0xfff60005,
+	0xfffbfff8, 0x00570024, 0x00370061, 0xfff60004,
+	0xfffcfff8, 0x00580026, 0x00350060, 0xfff60003,
+	0xfffdfff8, 0x00590028, 0x0033005f, 0xfff60002,
+	0xfffdfff7, 0x005b002a, 0x0031005f, 0xfff60001,
+	0xfffefff7, 0x005c002c, 0x002f005e, 0xfff60000,
+	0xfffffff6, 0x005e002d, 0x002d005e, 0xfff6ffff,
+	0x0000fff6, 0x005e002f, 0x002c005c, 0xfff7fffe,
+	0x0001fff6, 0x005f0031, 0x002a005b, 0xfff7fffd,
+	0x0002fff6, 0x005f0033, 0x00280059, 0xfff8fffd,
+	0x0003fff6, 0x00600035, 0x00260058, 0xfff8fffc,
+	0x0004fff6, 0x00610037, 0x00240057, 0xfff8fffb,
+	0x0005fff6, 0x00620039, 0x00220056, 0xfff7fffb,
+	0x0006fff6, 0x0062003a, 0x00210055, 0xfff8fffa,
+	0x0007fff6, 0x0063003c, 0x001f0054, 0xfff8fff9,
+	0x0008fff6, 0x0064003e, 0x001d0054, 0xfff6fff9,
+	0x000afff6, 0x00640040, 0x001c0052, 0xfff6fff8,
+	0x000bfff6, 0x00640042, 0x001a0051, 0xfff6fff8,
+	0x000cfff6, 0x00650044, 0x00180050, 0xfff5fff8,
+	0x000efff6, 0x00650045, 0x0017004f, 0xfff5fff7,
+	0x000ffff6, 0x00670047, 0x0015004f, 0xfff2fff7,
+	0x0010fff6, 0x00680049, 0x0013004e, 0xfff1fff7
+};
+
+static const u32 coef_lut_f_c_legacy[NB_COEF] = {
+	0x0000fffb, 0x003a001a, 0x003a005d, 0x0000001a,
+	0x0001fffb, 0x003f001b, 0x00400051, 0x00000019,
+	0x0001fffc, 0x0040001c, 0x003f0051, 0x00000017,
+	0x0002fffb, 0x0042001d, 0x003e0051, 0xffff0016,
+	0x0002fffb, 0x0043001e, 0x003d0051, 0xffff0015,
+	0x0003fffc, 0x00430020, 0x003b0050, 0xffff0014,
+	0x0003fffb, 0x00450021, 0x003a0051, 0xfffe0013,
+	0x0004fffc, 0x00450022, 0x00390050, 0xfffe0012,
+	0x0005fffc, 0x00460023, 0x0038004f, 0xfffe0011,
+	0x0005fffb, 0x00480025, 0x00360050, 0xfffd0010,
+	0x0006fffc, 0x00480026, 0x0035004f, 0xfffd000f,
+	0x0006fffc, 0x00490027, 0x0034004f, 0xfffd000e,
+	0x0007fffd, 0x00490028, 0x0033004e, 0xfffd000d,
+	0x0008fffc, 0x004a002a, 0x0031004d, 0xfffd000d,
+	0x0009fffd, 0x004a002b, 0x0030004d, 0xfffc000c,
+	0x0009fffc, 0x004c002c, 0x002f004d, 0xfffc000b,
+	0x000afffc, 0x004c002e, 0x002e004c, 0xfffc000a,
+	0x000bfffc, 0x004d002f, 0x002c004c, 0xfffc0009,
+	0x000cfffc, 0x004d0030, 0x002b004a, 0xfffd0009,
+	0x000dfffd, 0x004d0031, 0x002a004a, 0xfffc0008,
+	0x000dfffd, 0x004e0033, 0x00280049, 0xfffd0007,
+	0x000efffd, 0x004f0034, 0x00270049, 0xfffc0006,
+	0x000ffffd, 0x004f0035, 0x00260048, 0xfffc0006,
+	0x0010fffd, 0x00500036, 0x00250048, 0xfffb0005,
+	0x0011fffe, 0x004f0038, 0x00230046, 0xfffc0005,
+	0x0012fffe, 0x00500039, 0x00220045, 0xfffc0004,
+	0x0013fffe, 0x0051003a, 0x00210045, 0xfffb0003,
+	0x0014ffff, 0x0050003b, 0x00200043, 0xfffc0003,
+	0x0015ffff, 0x0051003d, 0x001e0043, 0xfffb0002,
+	0x0016ffff, 0x0051003e, 0x001d0042, 0xfffb0002,
+	0x00170000, 0x0051003f, 0x001c0040, 0xfffc0001,
+	0x00190000, 0x00510040, 0x001b003f, 0xfffb0001
+};
+
+#endif
diff --git a/drivers/gpu/drm/sti/sti_layer.c b/drivers/gpu/drm/sti/sti_layer.c
index 06a587c4f1bb..899104f9d4bc 100644
--- a/drivers/gpu/drm/sti/sti_layer.c
+++ b/drivers/gpu/drm/sti/sti_layer.c
@@ -11,7 +11,9 @@
 #include <drm/drm_fb_cma_helper.h>
 
 #include "sti_compositor.h"
+#include "sti_cursor.h"
 #include "sti_gdp.h"
+#include "sti_hqvdp.h"
 #include "sti_layer.h"
 #include "sti_vid.h"
 
@@ -32,10 +34,13 @@ const char *sti_layer_to_str(struct sti_layer *layer)
 		return "VID1";
 	case STI_CURSOR:
 		return "CURSOR";
+	case STI_HQVDP_0:
+		return "HQVDP0";
 	default:
 		return "<UNKNOWN LAYER>";
 	}
 }
+EXPORT_SYMBOL(sti_layer_to_str);
 
 struct sti_layer *sti_layer_create(struct device *dev, int desc,
 				   void __iomem *baseaddr)
@@ -50,6 +55,12 @@ struct sti_layer *sti_layer_create(struct device *dev, int desc,
 	case STI_VID:
 		layer = sti_vid_create(dev);
 		break;
+	case STI_CUR:
+		layer = sti_cursor_create(dev);
+		break;
+	case STI_VDP:
+		layer = sti_hqvdp_create(dev);
+		break;
 	}
 
 	if (!layer) {
@@ -67,8 +78,11 @@ struct sti_layer *sti_layer_create(struct device *dev, int desc,
 
 	return layer;
 }
+EXPORT_SYMBOL(sti_layer_create);
 
-int sti_layer_prepare(struct sti_layer *layer, struct drm_framebuffer *fb,
+int sti_layer_prepare(struct sti_layer *layer,
+		      struct drm_crtc *crtc,
+		      struct drm_framebuffer *fb,
 		      struct drm_display_mode *mode, int mixer_id,
 		      int dest_x, int dest_y, int dest_w, int dest_h,
 		      int src_x, int src_y, int src_w, int src_h)
@@ -88,6 +102,7 @@ int sti_layer_prepare(struct sti_layer *layer, struct drm_framebuffer *fb,
 		return 1;
 	}
 
+	layer->crtc = crtc;
 	layer->fb = fb;
 	layer->mode = mode;
 	layer->mixer_id = mixer_id;
@@ -100,6 +115,7 @@ int sti_layer_prepare(struct sti_layer *layer, struct drm_framebuffer *fb,
 	layer->src_w = src_w;
 	layer->src_h = src_h;
 	layer->format = fb->pixel_format;
+	layer->vaddr = cma_obj->vaddr;
 	layer->paddr = cma_obj->paddr;
 	for (i = 0; i < 4; i++) {
 		layer->pitches[i] = fb->pitches[i];
diff --git a/drivers/gpu/drm/sti/sti_layer.h b/drivers/gpu/drm/sti/sti_layer.h
index 198c3774cc12..ceff497f557e 100644
--- a/drivers/gpu/drm/sti/sti_layer.h
+++ b/drivers/gpu/drm/sti/sti_layer.h
@@ -22,7 +22,8 @@ enum sti_layer_type {
 	STI_GDP = 1 << STI_LAYER_TYPE_SHIFT,
 	STI_VID = 2 << STI_LAYER_TYPE_SHIFT,
 	STI_CUR = 3 << STI_LAYER_TYPE_SHIFT,
-	STI_BCK = 4 << STI_LAYER_TYPE_SHIFT
+	STI_BCK = 4 << STI_LAYER_TYPE_SHIFT,
+	STI_VDP = 5 << STI_LAYER_TYPE_SHIFT
 };
 
 enum sti_layer_id_of_type {
@@ -39,6 +40,7 @@ enum sti_layer_desc {
 	STI_GDP_3       = STI_GDP | STI_ID_3,
 	STI_VID_0       = STI_VID | STI_ID_0,
 	STI_VID_1       = STI_VID | STI_ID_1,
+	STI_HQVDP_0     = STI_VDP | STI_ID_0,
 	STI_CURSOR      = STI_CUR,
 	STI_BACK        = STI_BCK
 };
@@ -67,6 +69,7 @@ struct sti_layer_funcs {
  *
  * @plane:              drm plane it is bound to (if any)
  * @fb:                 drm fb it is bound to
+ * @crtc:               crtc it is bound to
  * @mode:               display mode
  * @desc:               layer type & id
  * @device:		driver device
@@ -82,11 +85,13 @@ struct sti_layer_funcs {
  * @format:             format
  * @pitches:            pitch of 'planes' (eg: Y, U, V)
  * @offsets:            offset of 'planes'
+ * @vaddr:              virtual address of the input buffer
  * @paddr:              physical address of the input buffer
  */
 struct sti_layer {
 	struct drm_plane plane;
 	struct drm_framebuffer *fb;
+	struct drm_crtc *crtc;
 	struct drm_display_mode *mode;
 	enum sti_layer_desc desc;
 	struct device *dev;
@@ -102,12 +107,15 @@ struct sti_layer {
 	uint32_t format;
 	unsigned int pitches[4];
 	unsigned int offsets[4];
+	void *vaddr;
 	dma_addr_t paddr;
 };
 
 struct sti_layer *sti_layer_create(struct device *dev, int desc,
 			void __iomem *baseaddr);
-int sti_layer_prepare(struct sti_layer *layer, struct drm_framebuffer *fb,
+int sti_layer_prepare(struct sti_layer *layer,
+			struct drm_crtc *crtc,
+			struct drm_framebuffer *fb,
 			struct drm_display_mode *mode,
 			int mixer_id,
 			int dest_x, int dest_y,
diff --git a/drivers/gpu/drm/sti/sti_mixer.c b/drivers/gpu/drm/sti/sti_mixer.c
index 79f369db9fb6..13a4b84deab6 100644
--- a/drivers/gpu/drm/sti/sti_mixer.c
+++ b/drivers/gpu/drm/sti/sti_mixer.c
@@ -45,6 +45,7 @@ static const u32 mixerColorSpaceMatIdentity[] = {
 #define GAM_CTL_GDP1_MASK  BIT(4)
 #define GAM_CTL_GDP2_MASK  BIT(5)
 #define GAM_CTL_GDP3_MASK  BIT(6)
+#define GAM_CTL_CURSOR_MASK BIT(9)
 
 const char *sti_mixer_to_str(struct sti_mixer *mixer)
 {
@@ -122,11 +123,15 @@ int sti_mixer_set_layer_depth(struct sti_mixer *mixer, struct sti_layer *layer)
 		layer_id = GAM_DEPTH_GDP3_ID;
 		break;
 	case STI_VID_0:
+	case STI_HQVDP_0:
 		layer_id = GAM_DEPTH_VID0_ID;
 		break;
 	case STI_VID_1:
 		layer_id = GAM_DEPTH_VID1_ID;
 		break;
+	case STI_CURSOR:
+		/* no need to set depth for cursor */
+		return 0;
 	default:
 		DRM_ERROR("Unknown layer %d\n", layer->desc);
 		return 1;
@@ -185,9 +190,12 @@ static u32 sti_mixer_get_layer_mask(struct sti_layer *layer)
 	case STI_GDP_3:
 		return GAM_CTL_GDP3_MASK;
 	case STI_VID_0:
+	case STI_HQVDP_0:
 		return GAM_CTL_VID0_MASK;
 	case STI_VID_1:
 		return GAM_CTL_VID1_MASK;
+	case STI_CURSOR:
+		return GAM_CTL_CURSOR_MASK;
 	default:
 		return 0;
 	}
@@ -215,6 +223,15 @@ int sti_mixer_set_layer_status(struct sti_mixer *mixer,
 	return 0;
 }
 
+void sti_mixer_clear_all_layers(struct sti_mixer *mixer)
+{
+	u32 val;
+
+	DRM_DEBUG_DRIVER("%s clear all layer\n", sti_mixer_to_str(mixer));
+	val = sti_mixer_reg_read(mixer, GAM_MIXER_CTL) & 0xFFFF0000;
+	sti_mixer_reg_write(mixer, GAM_MIXER_CTL, val);
+}
+
 void sti_mixer_set_matrix(struct sti_mixer *mixer)
 {
 	unsigned int i;
diff --git a/drivers/gpu/drm/sti/sti_mixer.h b/drivers/gpu/drm/sti/sti_mixer.h
index 874372102e52..b97282182908 100644
--- a/drivers/gpu/drm/sti/sti_mixer.h
+++ b/drivers/gpu/drm/sti/sti_mixer.h
@@ -23,6 +23,7 @@
  * @id: id of the mixer
  * @drm_crtc: crtc object link to the mixer
  * @pending_event: set if a flip event is pending on crtc
+ * @enabled: to know if the mixer is active or not
  */
 struct sti_mixer {
 	struct device *dev;
@@ -30,6 +31,7 @@ struct sti_mixer {
 	int id;
 	struct drm_crtc	drm_crtc;
 	struct drm_pending_vblank_event *pending_event;
+	bool enabled;
 };
 
 const char *sti_mixer_to_str(struct sti_mixer *mixer);
@@ -39,6 +41,7 @@ struct sti_mixer *sti_mixer_create(struct device *dev, int id,
 
 int sti_mixer_set_layer_status(struct sti_mixer *mixer,
 		struct sti_layer *layer, bool status);
+void sti_mixer_clear_all_layers(struct sti_mixer *mixer);
 int sti_mixer_set_layer_depth(struct sti_mixer *mixer, struct sti_layer *layer);
 int sti_mixer_active_video_area(struct sti_mixer *mixer,
 		struct drm_display_mode *mode);
diff --git a/drivers/gpu/drm/sti/sti_tvout.c b/drivers/gpu/drm/sti/sti_tvout.c
index b8afe490356a..cb924aa2b321 100644
--- a/drivers/gpu/drm/sti/sti_tvout.c
+++ b/drivers/gpu/drm/sti/sti_tvout.c
@@ -16,6 +16,8 @@
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
 
+#include "sti_drm_crtc.h"
+
 /* glue registers */
 #define TVO_CSC_MAIN_M0                  0x000
 #define TVO_CSC_MAIN_M1                  0x004
@@ -96,7 +98,7 @@
 
 #define TVO_SYNC_HD_DCS_SHIFT            8
 
-#define ENCODER_MAIN_CRTC_MASK           BIT(0)
+#define ENCODER_CRTC_MASK                (BIT(0) | BIT(1))
 
 /* enum listing the supported output data format */
 enum sti_tvout_video_out_type {
@@ -149,14 +151,15 @@ static void tvout_write(struct sti_tvout *tvout, u32 val, int offset)
  * Set the clipping mode of a VIP
  *
  * @tvout: tvout structure
+ * @reg: register to set
  * @cr_r:
  * @y_g:
  * @cb_b:
  */
-static void tvout_vip_set_color_order(struct sti_tvout *tvout,
+static void tvout_vip_set_color_order(struct sti_tvout *tvout, int reg,
 				      u32 cr_r, u32 y_g, u32 cb_b)
 {
-	u32 val = tvout_read(tvout, TVO_VIP_HDMI);
+	u32 val = tvout_read(tvout, reg);
 
 	val &= ~(TVO_VIP_REORDER_MASK << TVO_VIP_REORDER_R_SHIFT);
 	val &= ~(TVO_VIP_REORDER_MASK << TVO_VIP_REORDER_G_SHIFT);
@@ -165,52 +168,58 @@ static void tvout_vip_set_color_order(struct sti_tvout *tvout,
 	val |= y_g << TVO_VIP_REORDER_G_SHIFT;
 	val |= cb_b << TVO_VIP_REORDER_B_SHIFT;
 
-	tvout_write(tvout, val, TVO_VIP_HDMI);
+	tvout_write(tvout, val, reg);
 }
 
 /**
  * Set the clipping mode of a VIP
  *
  * @tvout: tvout structure
+ * @reg: register to set
  * @range: clipping range
  */
-static void tvout_vip_set_clip_mode(struct sti_tvout *tvout, u32 range)
+static void tvout_vip_set_clip_mode(struct sti_tvout *tvout, int reg, u32 range)
 {
-	u32 val = tvout_read(tvout, TVO_VIP_HDMI);
+	u32 val = tvout_read(tvout, reg);
 
 	val &= ~(TVO_VIP_CLIP_MASK << TVO_VIP_CLIP_SHIFT);
 	val |= range << TVO_VIP_CLIP_SHIFT;
-	tvout_write(tvout, val, TVO_VIP_HDMI);
+	tvout_write(tvout, val, reg);
 }
 
 /**
  * Set the rounded value of a VIP
  *
  * @tvout: tvout structure
+ * @reg: register to set
  * @rnd: rounded val per component
  */
-static void tvout_vip_set_rnd(struct sti_tvout *tvout, u32 rnd)
+static void tvout_vip_set_rnd(struct sti_tvout *tvout, int reg, u32 rnd)
 {
-	u32 val = tvout_read(tvout, TVO_VIP_HDMI);
+	u32 val = tvout_read(tvout, reg);
 
 	val &= ~(TVO_VIP_RND_MASK << TVO_VIP_RND_SHIFT);
 	val |= rnd << TVO_VIP_RND_SHIFT;
-	tvout_write(tvout, val, TVO_VIP_HDMI);
+	tvout_write(tvout, val, reg);
 }
 
 /**
  * Select the VIP input
  *
  * @tvout: tvout structure
+ * @reg: register to set
+ * @main_path: main or auxiliary path
+ * @sel_input_logic_inverted: need to invert the logic
  * @sel_input: selected_input (main/aux + conv)
  */
 static void tvout_vip_set_sel_input(struct sti_tvout *tvout,
+				    int reg,
 				    bool main_path,
 				    bool sel_input_logic_inverted,
 				    enum sti_tvout_video_out_type video_out)
 {
 	u32 sel_input;
-	u32 val = tvout_read(tvout, TVO_VIP_HDMI);
+	u32 val = tvout_read(tvout, reg);
 
 	if (main_path)
 		sel_input = TVO_VIP_SEL_INPUT_MAIN;
@@ -232,22 +241,24 @@ static void tvout_vip_set_sel_input(struct sti_tvout *tvout,
 
 	val &= ~TVO_VIP_SEL_INPUT_MASK;
 	val |= sel_input;
-	tvout_write(tvout, val, TVO_VIP_HDMI);
+	tvout_write(tvout, val, reg);
 }
 
 /**
  * Select the input video signed or unsigned
  *
  * @tvout: tvout structure
+ * @reg: register to set
  * @in_vid_signed: used video input format
  */
-static void tvout_vip_set_in_vid_fmt(struct sti_tvout *tvout, u32 in_vid_fmt)
+static void tvout_vip_set_in_vid_fmt(struct sti_tvout *tvout,
+		int reg, u32 in_vid_fmt)
 {
-	u32 val = tvout_read(tvout, TVO_VIP_HDMI);
+	u32 val = tvout_read(tvout, reg);
 
 	val &= ~TVO_IN_FMT_SIGNED;
 	val |= in_vid_fmt;
-	tvout_write(tvout, val, TVO_MAIN_IN_VID_FORMAT);
+	tvout_write(tvout, val, reg);
 }
 
 /**
@@ -261,6 +272,7 @@ static void tvout_hdmi_start(struct sti_tvout *tvout, bool main_path)
 {
 	struct device_node *node = tvout->dev->of_node;
 	bool sel_input_logic_inverted = false;
+	u32 tvo_in_vid_format;
 
 	dev_dbg(tvout->dev, "%s\n", __func__);
 
@@ -268,33 +280,36 @@ static void tvout_hdmi_start(struct sti_tvout *tvout, bool main_path)
 		DRM_DEBUG_DRIVER("main vip for hdmi\n");
 		/* select the input sync for hdmi = VTG set 1 */
 		tvout_write(tvout, TVO_SYNC_MAIN_VTG_SET_1, TVO_HDMI_SYNC_SEL);
+		tvo_in_vid_format = TVO_MAIN_IN_VID_FORMAT;
 	} else {
 		DRM_DEBUG_DRIVER("aux vip for hdmi\n");
 		/* select the input sync for hdmi = VTG set 1 */
 		tvout_write(tvout, TVO_SYNC_AUX_VTG_SET_1, TVO_HDMI_SYNC_SEL);
+		tvo_in_vid_format = TVO_AUX_IN_VID_FORMAT;
 	}
 
 	/* set color channel order */
-	tvout_vip_set_color_order(tvout,
+	tvout_vip_set_color_order(tvout, TVO_VIP_HDMI,
 				  TVO_VIP_REORDER_CR_R_SEL,
 				  TVO_VIP_REORDER_Y_G_SEL,
 				  TVO_VIP_REORDER_CB_B_SEL);
 
 	/* set clipping mode (Limited range RGB/Y) */
-	tvout_vip_set_clip_mode(tvout, TVO_VIP_CLIP_LIMITED_RANGE_RGB_Y);
+	tvout_vip_set_clip_mode(tvout, TVO_VIP_HDMI,
+			TVO_VIP_CLIP_LIMITED_RANGE_RGB_Y);
 
 	/* set round mode (rounded to 8-bit per component) */
-	tvout_vip_set_rnd(tvout, TVO_VIP_RND_8BIT_ROUNDED);
+	tvout_vip_set_rnd(tvout, TVO_VIP_HDMI, TVO_VIP_RND_8BIT_ROUNDED);
 
 	if (of_device_is_compatible(node, "st,stih407-tvout")) {
 		/* set input video format */
-		tvout_vip_set_in_vid_fmt(tvout->regs + TVO_MAIN_IN_VID_FORMAT,
-					 TVO_IN_FMT_SIGNED);
+		tvout_vip_set_in_vid_fmt(tvout, tvo_in_vid_format,
+					TVO_IN_FMT_SIGNED);
 		sel_input_logic_inverted = true;
 	}
 
 	/* input selection */
-	tvout_vip_set_sel_input(tvout, main_path,
+	tvout_vip_set_sel_input(tvout, TVO_VIP_HDMI, main_path,
 			sel_input_logic_inverted, STI_TVOUT_VIDEO_OUT_RGB);
 }
 
@@ -309,48 +324,47 @@ static void tvout_hda_start(struct sti_tvout *tvout, bool main_path)
 {
 	struct device_node *node = tvout->dev->of_node;
 	bool sel_input_logic_inverted = false;
+	u32 tvo_in_vid_format;
+	int val;
 
 	dev_dbg(tvout->dev, "%s\n", __func__);
 
-	if (!main_path) {
-		DRM_ERROR("HD Analog on aux not implemented\n");
-		return;
+	if (main_path) {
+		val = TVO_SYNC_MAIN_VTG_SET_2 << TVO_SYNC_HD_DCS_SHIFT;
+		val |= TVO_SYNC_MAIN_VTG_SET_3;
+		tvout_write(tvout, val, TVO_HD_SYNC_SEL);
+		tvo_in_vid_format = TVO_MAIN_IN_VID_FORMAT;
+	} else {
+		val = TVO_SYNC_AUX_VTG_SET_2 << TVO_SYNC_HD_DCS_SHIFT;
+		val |= TVO_SYNC_AUX_VTG_SET_3;
+		tvout_write(tvout, val, TVO_HD_SYNC_SEL);
+		tvo_in_vid_format = TVO_AUX_IN_VID_FORMAT;
 	}
 
-	DRM_DEBUG_DRIVER("main vip for HDF\n");
-
 	/* set color channel order */
-	tvout_vip_set_color_order(tvout->regs + TVO_VIP_HDF,
+	tvout_vip_set_color_order(tvout, TVO_VIP_HDF,
 				  TVO_VIP_REORDER_CR_R_SEL,
 				  TVO_VIP_REORDER_Y_G_SEL,
 				  TVO_VIP_REORDER_CB_B_SEL);
 
-	/* set clipping mode (Limited range RGB/Y) */
-	tvout_vip_set_clip_mode(tvout->regs + TVO_VIP_HDF,
-				TVO_VIP_CLIP_LIMITED_RANGE_CB_CR);
+	/* set clipping mode (EAV/SAV clipping) */
+	tvout_vip_set_clip_mode(tvout, TVO_VIP_HDF, TVO_VIP_CLIP_EAV_SAV);
 
 	/* set round mode (rounded to 10-bit per component) */
-	tvout_vip_set_rnd(tvout->regs + TVO_VIP_HDF, TVO_VIP_RND_10BIT_ROUNDED);
+	tvout_vip_set_rnd(tvout, TVO_VIP_HDF, TVO_VIP_RND_10BIT_ROUNDED);
 
 	if (of_device_is_compatible(node, "st,stih407-tvout")) {
 		/* set input video format */
-		tvout_vip_set_in_vid_fmt(tvout, TVO_IN_FMT_SIGNED);
+		tvout_vip_set_in_vid_fmt(tvout,
+			tvo_in_vid_format, TVO_IN_FMT_SIGNED);
 		sel_input_logic_inverted = true;
 	}
 
 	/* Input selection */
-	tvout_vip_set_sel_input(tvout->regs + TVO_VIP_HDF,
-				main_path,
+	tvout_vip_set_sel_input(tvout, TVO_VIP_HDF, main_path,
 				sel_input_logic_inverted,
 				STI_TVOUT_VIDEO_OUT_YUV);
 
-	/* select the input sync for HD analog = VTG set 3
-	 * and HD DCS = VTG set 2 */
-	tvout_write(tvout,
-		(TVO_SYNC_MAIN_VTG_SET_2 << TVO_SYNC_HD_DCS_SHIFT)
-		| TVO_SYNC_MAIN_VTG_SET_3,
-		TVO_HD_SYNC_SEL);
-
 	/* power up HD DAC */
 	tvout_write(tvout, 0, TVO_HD_DAC_CFG_OFF);
 }
@@ -392,7 +406,7 @@ static void sti_hda_encoder_commit(struct drm_encoder *encoder)
 {
 	struct sti_tvout *tvout = to_sti_tvout(encoder);
 
-	tvout_hda_start(tvout, true);
+	tvout_hda_start(tvout, sti_drm_crtc_is_main(encoder->crtc));
 }
 
 static void sti_hda_encoder_disable(struct drm_encoder *encoder)
@@ -429,7 +443,7 @@ static struct drm_encoder *sti_tvout_create_hda_encoder(struct drm_device *dev,
 
 	drm_encoder = (struct drm_encoder *) encoder;
 
-	drm_encoder->possible_crtcs = ENCODER_MAIN_CRTC_MASK;
+	drm_encoder->possible_crtcs = ENCODER_CRTC_MASK;
 	drm_encoder->possible_clones = 1 << 0;
 
 	drm_encoder_init(dev, drm_encoder,
@@ -444,7 +458,7 @@ static void sti_hdmi_encoder_commit(struct drm_encoder *encoder)
 {
 	struct sti_tvout *tvout = to_sti_tvout(encoder);
 
-	tvout_hdmi_start(tvout, true);
+	tvout_hdmi_start(tvout, sti_drm_crtc_is_main(encoder->crtc));
 }
 
 static void sti_hdmi_encoder_disable(struct drm_encoder *encoder)
@@ -478,7 +492,7 @@ static struct drm_encoder *sti_tvout_create_hdmi_encoder(struct drm_device *dev,
 
 	drm_encoder = (struct drm_encoder *) encoder;
 
-	drm_encoder->possible_crtcs = ENCODER_MAIN_CRTC_MASK;
+	drm_encoder->possible_crtcs = ENCODER_CRTC_MASK;
 	drm_encoder->possible_clones = 1 << 1;
 
 	drm_encoder_init(dev, drm_encoder,
diff --git a/drivers/gpu/drm/sti/sti_vtac.c b/drivers/gpu/drm/sti/sti_vtac.c
index 82a51d488434..97bcdac23ae1 100644
--- a/drivers/gpu/drm/sti/sti_vtac.c
+++ b/drivers/gpu/drm/sti/sti_vtac.c
@@ -56,8 +56,16 @@ struct sti_vtac_mode {
 	u32 phyts_per_pixel;
 };
 
-static const struct sti_vtac_mode vtac_mode_main = {0x2, 0x2, VTAC_5_PPP};
-static const struct sti_vtac_mode vtac_mode_aux = {0x1, 0x0, VTAC_17_PPP};
+static const struct sti_vtac_mode vtac_mode_main = {
+	.vid_in_width = 0x2,
+	.phyts_width = 0x2,
+	.phyts_per_pixel = VTAC_5_PPP,
+};
+static const struct sti_vtac_mode vtac_mode_aux = {
+	.vid_in_width = 0x1,
+	.phyts_width = 0x0,
+	.phyts_per_pixel = VTAC_17_PPP,
+};
 
 /**
  * VTAC structure
diff --git a/drivers/gpu/drm/sti/sti_vtg.c b/drivers/gpu/drm/sti/sti_vtg.c
index 740d6e347a62..9564f2568e2c 100644
--- a/drivers/gpu/drm/sti/sti_vtg.c
+++ b/drivers/gpu/drm/sti/sti_vtg.c
@@ -51,10 +51,19 @@
 #define VTG_TOP_V_HD_3      0x010C
 #define VTG_BOT_V_HD_3      0x0110
 
+#define VTG_H_HD_4          0x0120
+#define VTG_TOP_V_VD_4      0x0124
+#define VTG_BOT_V_VD_4      0x0128
+#define VTG_TOP_V_HD_4      0x012c
+#define VTG_BOT_V_HD_4      0x0130
+
 #define VTG_IRQ_BOTTOM      BIT(0)
 #define VTG_IRQ_TOP         BIT(1)
 #define VTG_IRQ_MASK        (VTG_IRQ_TOP | VTG_IRQ_BOTTOM)
 
+/* Delay introduced by the HDMI in nb of pixel */
+#define HDMI_DELAY          (6)
+
 /* delay introduced by the Arbitrary Waveform Generator in nb of pixels */
 #define AWG_DELAY_HD        (-9)
 #define AWG_DELAY_ED        (-8)
@@ -133,10 +142,10 @@ static void vtg_set_mode(struct sti_vtg *vtg,
 	writel(tmp, vtg->regs + VTG_VID_TFS);
 	writel(tmp, vtg->regs + VTG_VID_BFS);
 
-	/* prepare VTG set 1 and 2 for HDMI and VTG set 3 for HD DAC */
-	tmp = (mode->hsync_end - mode->hsync_start) << 16;
+	/* prepare VTG set 1 for HDMI */
+	tmp = (mode->hsync_end - mode->hsync_start + HDMI_DELAY) << 16;
+	tmp |= HDMI_DELAY;
 	writel(tmp, vtg->regs + VTG_H_HD_1);
-	writel(tmp, vtg->regs + VTG_H_HD_2);
 
 	tmp = (mode->vsync_end - mode->vsync_start + 1) << 16;
 	tmp |= 1;
@@ -146,6 +155,11 @@ static void vtg_set_mode(struct sti_vtg *vtg,
 	writel(0, vtg->regs + VTG_BOT_V_HD_1);
 
 	/* prepare VTG set 2 for for HD DCS */
+	tmp = (mode->hsync_end - mode->hsync_start) << 16;
+	writel(tmp, vtg->regs + VTG_H_HD_2);
+
+	tmp = (mode->vsync_end - mode->vsync_start + 1) << 16;
+	tmp |= 1;
 	writel(tmp, vtg->regs + VTG_TOP_V_VD_2);
 	writel(tmp, vtg->regs + VTG_BOT_V_VD_2);
 	writel(0, vtg->regs + VTG_TOP_V_HD_2);
@@ -166,6 +180,17 @@ static void vtg_set_mode(struct sti_vtg *vtg,
 	writel(tmp, vtg->regs + VTG_TOP_V_HD_3);
 	writel(tmp, vtg->regs + VTG_BOT_V_HD_3);
 
+	/* Prepare VTG set 4 for DVO */
+	tmp = (mode->hsync_end - mode->hsync_start) << 16;
+	writel(tmp, vtg->regs + VTG_H_HD_4);
+
+	tmp = (mode->vsync_end - mode->vsync_start + 1) << 16;
+	tmp |= 1;
+	writel(tmp, vtg->regs + VTG_TOP_V_VD_4);
+	writel(tmp, vtg->regs + VTG_BOT_V_VD_4);
+	writel(0, vtg->regs + VTG_TOP_V_HD_4);
+	writel(0, vtg->regs + VTG_BOT_V_HD_4);
+
 	/* mode */
 	writel(type, vtg->regs + VTG_MODE);
 }
diff --git a/drivers/gpu/drm/tdfx/tdfx_drv.c b/drivers/gpu/drm/tdfx/tdfx_drv.c
index 3492ca5c46d3..fab5ebcb0fef 100644
--- a/drivers/gpu/drm/tdfx/tdfx_drv.c
+++ b/drivers/gpu/drm/tdfx/tdfx_drv.c
@@ -36,6 +36,7 @@
 #include "tdfx_drv.h"
 
 #include <drm/drm_pciids.h>
+#include <drm/drm_legacy.h>
 
 static struct pci_device_id pciidlist[] = {
 	tdfx_PCI_IDS
@@ -46,7 +47,7 @@ static const struct file_operations tdfx_driver_fops = {
 	.open = drm_open,
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_mmap,
+	.mmap = drm_legacy_mmap,
 	.poll = drm_poll,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl = drm_compat_ioctl,
@@ -55,6 +56,7 @@ static const struct file_operations tdfx_driver_fops = {
 };
 
 static struct drm_driver driver = {
+	.set_busid = drm_pci_set_busid,
 	.fops = &tdfx_driver_fops,
 	.name = DRIVER_NAME,
 	.desc = DRIVER_DESC,
diff --git a/drivers/gpu/drm/tegra/Kconfig b/drivers/gpu/drm/tegra/Kconfig
index 354ddb29231f..74d9d621453d 100644
--- a/drivers/gpu/drm/tegra/Kconfig
+++ b/drivers/gpu/drm/tegra/Kconfig
@@ -1,6 +1,7 @@
 config DRM_TEGRA
 	tristate "NVIDIA Tegra DRM"
 	depends on ARCH_TEGRA || (ARM && COMPILE_TEST)
+	depends on COMMON_CLK
 	depends on DRM
 	depends on RESET_CONTROLLER
 	select DRM_KMS_HELPER
diff --git a/drivers/gpu/drm/tegra/dc.c b/drivers/gpu/drm/tegra/dc.c
index 6553fd238685..978993fa3a36 100644
--- a/drivers/gpu/drm/tegra/dc.c
+++ b/drivers/gpu/drm/tegra/dc.c
@@ -9,17 +9,23 @@
 
 #include <linux/clk.h>
 #include <linux/debugfs.h>
+#include <linux/iommu.h>
 #include <linux/reset.h>
 
+#include <soc/tegra/pmc.h>
+
 #include "dc.h"
 #include "drm.h"
 #include "gem.h"
 
+#include <drm/drm_plane_helper.h>
+
 struct tegra_dc_soc_info {
 	bool supports_interlacing;
 	bool supports_cursor;
 	bool supports_block_linear;
 	unsigned int pitch_align;
+	bool has_powergate;
 };
 
 struct tegra_plane {
@@ -32,6 +38,26 @@ static inline struct tegra_plane *to_tegra_plane(struct drm_plane *plane)
 	return container_of(plane, struct tegra_plane, base);
 }
 
+static void tegra_dc_window_commit(struct tegra_dc *dc, unsigned int index)
+{
+	u32 value = WIN_A_ACT_REQ << index;
+
+	tegra_dc_writel(dc, value << 8, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(dc, value, DC_CMD_STATE_CONTROL);
+}
+
+static void tegra_dc_cursor_commit(struct tegra_dc *dc)
+{
+	tegra_dc_writel(dc, CURSOR_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(dc, CURSOR_ACT_REQ, DC_CMD_STATE_CONTROL);
+}
+
+static void tegra_dc_commit(struct tegra_dc *dc)
+{
+	tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
+}
+
 static unsigned int tegra_dc_format(uint32_t format, uint32_t *swap)
 {
 	/* assume no swapping of fetched data */
@@ -142,7 +168,7 @@ static int tegra_dc_setup_window(struct tegra_dc *dc, unsigned int index,
 				 const struct tegra_dc_window *window)
 {
 	unsigned h_offset, v_offset, h_size, v_size, h_dda, v_dda, bpp;
-	unsigned long value;
+	unsigned long value, flags;
 	bool yuv, planar;
 
 	/*
@@ -155,6 +181,8 @@ static int tegra_dc_setup_window(struct tegra_dc *dc, unsigned int index,
 	else
 		bpp = planar ? 1 : 2;
 
+	spin_lock_irqsave(&dc->lock, flags);
+
 	value = WINDOW_A_SELECT << index;
 	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_WINDOW_HEADER);
 
@@ -247,6 +275,7 @@ static int tegra_dc_setup_window(struct tegra_dc *dc, unsigned int index,
 
 		case TEGRA_BO_TILING_MODE_BLOCK:
 			DRM_ERROR("hardware doesn't support block linear mode\n");
+			spin_unlock_irqrestore(&dc->lock, flags);
 			return -EINVAL;
 		}
 
@@ -303,17 +332,267 @@ static int tegra_dc_setup_window(struct tegra_dc *dc, unsigned int index,
 		break;
 	}
 
-	tegra_dc_writel(dc, WIN_A_UPDATE << index, DC_CMD_STATE_CONTROL);
-	tegra_dc_writel(dc, WIN_A_ACT_REQ << index, DC_CMD_STATE_CONTROL);
+	tegra_dc_window_commit(dc, index);
+
+	spin_unlock_irqrestore(&dc->lock, flags);
+
+	return 0;
+}
+
+static int tegra_window_plane_disable(struct drm_plane *plane)
+{
+	struct tegra_dc *dc = to_tegra_dc(plane->crtc);
+	struct tegra_plane *p = to_tegra_plane(plane);
+	unsigned long flags;
+	u32 value;
+
+	if (!plane->crtc)
+		return 0;
+
+	spin_lock_irqsave(&dc->lock, flags);
+
+	value = WINDOW_A_SELECT << p->index;
+	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_WINDOW_HEADER);
+
+	value = tegra_dc_readl(dc, DC_WIN_WIN_OPTIONS);
+	value &= ~WIN_ENABLE;
+	tegra_dc_writel(dc, value, DC_WIN_WIN_OPTIONS);
+
+	tegra_dc_window_commit(dc, p->index);
+
+	spin_unlock_irqrestore(&dc->lock, flags);
 
 	return 0;
 }
 
-static int tegra_plane_update(struct drm_plane *plane, struct drm_crtc *crtc,
-			      struct drm_framebuffer *fb, int crtc_x,
-			      int crtc_y, unsigned int crtc_w,
-			      unsigned int crtc_h, uint32_t src_x,
-			      uint32_t src_y, uint32_t src_w, uint32_t src_h)
+static void tegra_plane_destroy(struct drm_plane *plane)
+{
+	struct tegra_plane *p = to_tegra_plane(plane);
+
+	drm_plane_cleanup(plane);
+	kfree(p);
+}
+
+static const u32 tegra_primary_plane_formats[] = {
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_RGB565,
+};
+
+static int tegra_primary_plane_update(struct drm_plane *plane,
+				      struct drm_crtc *crtc,
+				      struct drm_framebuffer *fb, int crtc_x,
+				      int crtc_y, unsigned int crtc_w,
+				      unsigned int crtc_h, uint32_t src_x,
+				      uint32_t src_y, uint32_t src_w,
+				      uint32_t src_h)
+{
+	struct tegra_bo *bo = tegra_fb_get_plane(fb, 0);
+	struct tegra_plane *p = to_tegra_plane(plane);
+	struct tegra_dc *dc = to_tegra_dc(crtc);
+	struct tegra_dc_window window;
+	int err;
+
+	memset(&window, 0, sizeof(window));
+	window.src.x = src_x >> 16;
+	window.src.y = src_y >> 16;
+	window.src.w = src_w >> 16;
+	window.src.h = src_h >> 16;
+	window.dst.x = crtc_x;
+	window.dst.y = crtc_y;
+	window.dst.w = crtc_w;
+	window.dst.h = crtc_h;
+	window.format = tegra_dc_format(fb->pixel_format, &window.swap);
+	window.bits_per_pixel = fb->bits_per_pixel;
+	window.bottom_up = tegra_fb_is_bottom_up(fb);
+
+	err = tegra_fb_get_tiling(fb, &window.tiling);
+	if (err < 0)
+		return err;
+
+	window.base[0] = bo->paddr + fb->offsets[0];
+	window.stride[0] = fb->pitches[0];
+
+	err = tegra_dc_setup_window(dc, p->index, &window);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+
+static void tegra_primary_plane_destroy(struct drm_plane *plane)
+{
+	tegra_window_plane_disable(plane);
+	tegra_plane_destroy(plane);
+}
+
+static const struct drm_plane_funcs tegra_primary_plane_funcs = {
+	.update_plane = tegra_primary_plane_update,
+	.disable_plane = tegra_window_plane_disable,
+	.destroy = tegra_primary_plane_destroy,
+};
+
+static struct drm_plane *tegra_dc_primary_plane_create(struct drm_device *drm,
+						       struct tegra_dc *dc)
+{
+	struct tegra_plane *plane;
+	unsigned int num_formats;
+	const u32 *formats;
+	int err;
+
+	plane = kzalloc(sizeof(*plane), GFP_KERNEL);
+	if (!plane)
+		return ERR_PTR(-ENOMEM);
+
+	num_formats = ARRAY_SIZE(tegra_primary_plane_formats);
+	formats = tegra_primary_plane_formats;
+
+	err = drm_universal_plane_init(drm, &plane->base, 1 << dc->pipe,
+				       &tegra_primary_plane_funcs, formats,
+				       num_formats, DRM_PLANE_TYPE_PRIMARY);
+	if (err < 0) {
+		kfree(plane);
+		return ERR_PTR(err);
+	}
+
+	return &plane->base;
+}
+
+static const u32 tegra_cursor_plane_formats[] = {
+	DRM_FORMAT_RGBA8888,
+};
+
+static int tegra_cursor_plane_update(struct drm_plane *plane,
+				     struct drm_crtc *crtc,
+				     struct drm_framebuffer *fb, int crtc_x,
+				     int crtc_y, unsigned int crtc_w,
+				     unsigned int crtc_h, uint32_t src_x,
+				     uint32_t src_y, uint32_t src_w,
+				     uint32_t src_h)
+{
+	struct tegra_bo *bo = tegra_fb_get_plane(fb, 0);
+	struct tegra_dc *dc = to_tegra_dc(crtc);
+	u32 value = CURSOR_CLIP_DISPLAY;
+
+	/* scaling not supported for cursor */
+	if ((src_w >> 16 != crtc_w) || (src_h >> 16 != crtc_h))
+		return -EINVAL;
+
+	/* only square cursors supported */
+	if (src_w != src_h)
+		return -EINVAL;
+
+	switch (crtc_w) {
+	case 32:
+		value |= CURSOR_SIZE_32x32;
+		break;
+
+	case 64:
+		value |= CURSOR_SIZE_64x64;
+		break;
+
+	case 128:
+		value |= CURSOR_SIZE_128x128;
+		break;
+
+	case 256:
+		value |= CURSOR_SIZE_256x256;
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	value |= (bo->paddr >> 10) & 0x3fffff;
+	tegra_dc_writel(dc, value, DC_DISP_CURSOR_START_ADDR);
+
+#ifdef CONFIG_ARCH_DMA_ADDR_T_64BIT
+	value = (bo->paddr >> 32) & 0x3;
+	tegra_dc_writel(dc, value, DC_DISP_CURSOR_START_ADDR_HI);
+#endif
+
+	/* enable cursor and set blend mode */
+	value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
+	value |= CURSOR_ENABLE;
+	tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
+
+	value = tegra_dc_readl(dc, DC_DISP_BLEND_CURSOR_CONTROL);
+	value &= ~CURSOR_DST_BLEND_MASK;
+	value &= ~CURSOR_SRC_BLEND_MASK;
+	value |= CURSOR_MODE_NORMAL;
+	value |= CURSOR_DST_BLEND_NEG_K1_TIMES_SRC;
+	value |= CURSOR_SRC_BLEND_K1_TIMES_SRC;
+	value |= CURSOR_ALPHA;
+	tegra_dc_writel(dc, value, DC_DISP_BLEND_CURSOR_CONTROL);
+
+	/* position the cursor */
+	value = (crtc_y & 0x3fff) << 16 | (crtc_x & 0x3fff);
+	tegra_dc_writel(dc, value, DC_DISP_CURSOR_POSITION);
+
+	/* apply changes */
+	tegra_dc_cursor_commit(dc);
+	tegra_dc_commit(dc);
+
+	return 0;
+}
+
+static int tegra_cursor_plane_disable(struct drm_plane *plane)
+{
+	struct tegra_dc *dc = to_tegra_dc(plane->crtc);
+	u32 value;
+
+	if (!plane->crtc)
+		return 0;
+
+	value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
+	value &= ~CURSOR_ENABLE;
+	tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
+
+	tegra_dc_cursor_commit(dc);
+	tegra_dc_commit(dc);
+
+	return 0;
+}
+
+static const struct drm_plane_funcs tegra_cursor_plane_funcs = {
+	.update_plane = tegra_cursor_plane_update,
+	.disable_plane = tegra_cursor_plane_disable,
+	.destroy = tegra_plane_destroy,
+};
+
+static struct drm_plane *tegra_dc_cursor_plane_create(struct drm_device *drm,
+						      struct tegra_dc *dc)
+{
+	struct tegra_plane *plane;
+	unsigned int num_formats;
+	const u32 *formats;
+	int err;
+
+	plane = kzalloc(sizeof(*plane), GFP_KERNEL);
+	if (!plane)
+		return ERR_PTR(-ENOMEM);
+
+	num_formats = ARRAY_SIZE(tegra_cursor_plane_formats);
+	formats = tegra_cursor_plane_formats;
+
+	err = drm_universal_plane_init(drm, &plane->base, 1 << dc->pipe,
+				       &tegra_cursor_plane_funcs, formats,
+				       num_formats, DRM_PLANE_TYPE_CURSOR);
+	if (err < 0) {
+		kfree(plane);
+		return ERR_PTR(err);
+	}
+
+	return &plane->base;
+}
+
+static int tegra_overlay_plane_update(struct drm_plane *plane,
+				      struct drm_crtc *crtc,
+				      struct drm_framebuffer *fb, int crtc_x,
+				      int crtc_y, unsigned int crtc_w,
+				      unsigned int crtc_h, uint32_t src_x,
+				      uint32_t src_y, uint32_t src_w,
+				      uint32_t src_h)
 {
 	struct tegra_plane *p = to_tegra_plane(plane);
 	struct tegra_dc *dc = to_tegra_dc(crtc);
@@ -359,44 +638,19 @@ static int tegra_plane_update(struct drm_plane *plane, struct drm_crtc *crtc,
 	return tegra_dc_setup_window(dc, p->index, &window);
 }
 
-static int tegra_plane_disable(struct drm_plane *plane)
-{
-	struct tegra_dc *dc = to_tegra_dc(plane->crtc);
-	struct tegra_plane *p = to_tegra_plane(plane);
-	unsigned long value;
-
-	if (!plane->crtc)
-		return 0;
-
-	value = WINDOW_A_SELECT << p->index;
-	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_WINDOW_HEADER);
-
-	value = tegra_dc_readl(dc, DC_WIN_WIN_OPTIONS);
-	value &= ~WIN_ENABLE;
-	tegra_dc_writel(dc, value, DC_WIN_WIN_OPTIONS);
-
-	tegra_dc_writel(dc, WIN_A_UPDATE << p->index, DC_CMD_STATE_CONTROL);
-	tegra_dc_writel(dc, WIN_A_ACT_REQ << p->index, DC_CMD_STATE_CONTROL);
-
-	return 0;
-}
-
-static void tegra_plane_destroy(struct drm_plane *plane)
+static void tegra_overlay_plane_destroy(struct drm_plane *plane)
 {
-	struct tegra_plane *p = to_tegra_plane(plane);
-
-	tegra_plane_disable(plane);
-	drm_plane_cleanup(plane);
-	kfree(p);
+	tegra_window_plane_disable(plane);
+	tegra_plane_destroy(plane);
 }
 
-static const struct drm_plane_funcs tegra_plane_funcs = {
-	.update_plane = tegra_plane_update,
-	.disable_plane = tegra_plane_disable,
-	.destroy = tegra_plane_destroy,
+static const struct drm_plane_funcs tegra_overlay_plane_funcs = {
+	.update_plane = tegra_overlay_plane_update,
+	.disable_plane = tegra_window_plane_disable,
+	.destroy = tegra_overlay_plane_destroy,
 };
 
-static const uint32_t plane_formats[] = {
+static const uint32_t tegra_overlay_plane_formats[] = {
 	DRM_FORMAT_XBGR8888,
 	DRM_FORMAT_XRGB8888,
 	DRM_FORMAT_RGB565,
@@ -406,27 +660,44 @@ static const uint32_t plane_formats[] = {
 	DRM_FORMAT_YUV422,
 };
 
-static int tegra_dc_add_planes(struct drm_device *drm, struct tegra_dc *dc)
+static struct drm_plane *tegra_dc_overlay_plane_create(struct drm_device *drm,
+						       struct tegra_dc *dc,
+						       unsigned int index)
 {
-	unsigned int i;
-	int err = 0;
+	struct tegra_plane *plane;
+	unsigned int num_formats;
+	const u32 *formats;
+	int err;
 
-	for (i = 0; i < 2; i++) {
-		struct tegra_plane *plane;
+	plane = kzalloc(sizeof(*plane), GFP_KERNEL);
+	if (!plane)
+		return ERR_PTR(-ENOMEM);
 
-		plane = kzalloc(sizeof(*plane), GFP_KERNEL);
-		if (!plane)
-			return -ENOMEM;
+	plane->index = index;
 
-		plane->index = 1 + i;
+	num_formats = ARRAY_SIZE(tegra_overlay_plane_formats);
+	formats = tegra_overlay_plane_formats;
 
-		err = drm_plane_init(drm, &plane->base, 1 << dc->pipe,
-				     &tegra_plane_funcs, plane_formats,
-				     ARRAY_SIZE(plane_formats), false);
-		if (err < 0) {
-			kfree(plane);
-			return err;
-		}
+	err = drm_universal_plane_init(drm, &plane->base, 1 << dc->pipe,
+				       &tegra_overlay_plane_funcs, formats,
+				       num_formats, DRM_PLANE_TYPE_OVERLAY);
+	if (err < 0) {
+		kfree(plane);
+		return ERR_PTR(err);
+	}
+
+	return &plane->base;
+}
+
+static int tegra_dc_add_planes(struct drm_device *drm, struct tegra_dc *dc)
+{
+	struct drm_plane *plane;
+	unsigned int i;
+
+	for (i = 0; i < 2; i++) {
+		plane = tegra_dc_overlay_plane_create(drm, dc, 1 + i);
+		if (IS_ERR(plane))
+			return PTR_ERR(plane);
 	}
 
 	return 0;
@@ -438,14 +709,16 @@ static int tegra_dc_set_base(struct tegra_dc *dc, int x, int y,
 	struct tegra_bo *bo = tegra_fb_get_plane(fb, 0);
 	unsigned int h_offset = 0, v_offset = 0;
 	struct tegra_bo_tiling tiling;
+	unsigned long value, flags;
 	unsigned int format, swap;
-	unsigned long value;
 	int err;
 
 	err = tegra_fb_get_tiling(fb, &tiling);
 	if (err < 0)
 		return err;
 
+	spin_lock_irqsave(&dc->lock, flags);
+
 	tegra_dc_writel(dc, WINDOW_A_SELECT, DC_CMD_DISPLAY_WINDOW_HEADER);
 
 	value = fb->offsets[0] + y * fb->pitches[0] +
@@ -491,6 +764,7 @@ static int tegra_dc_set_base(struct tegra_dc *dc, int x, int y,
 
 		case TEGRA_BO_TILING_MODE_BLOCK:
 			DRM_ERROR("hardware doesn't support block linear mode\n");
+			spin_unlock_irqrestore(&dc->lock, flags);
 			return -EINVAL;
 		}
 
@@ -513,12 +787,12 @@ static int tegra_dc_set_base(struct tegra_dc *dc, int x, int y,
 	tegra_dc_writel(dc, h_offset, DC_WINBUF_ADDR_H_OFFSET);
 	tegra_dc_writel(dc, v_offset, DC_WINBUF_ADDR_V_OFFSET);
 
-	value = GENERAL_UPDATE | WIN_A_UPDATE;
-	tegra_dc_writel(dc, value, DC_CMD_STATE_CONTROL);
-
 	value = GENERAL_ACT_REQ | WIN_A_ACT_REQ;
+	tegra_dc_writel(dc, value << 8, DC_CMD_STATE_CONTROL);
 	tegra_dc_writel(dc, value, DC_CMD_STATE_CONTROL);
 
+	spin_unlock_irqrestore(&dc->lock, flags);
+
 	return 0;
 }
 
@@ -548,109 +822,6 @@ void tegra_dc_disable_vblank(struct tegra_dc *dc)
 	spin_unlock_irqrestore(&dc->lock, flags);
 }
 
-static int tegra_dc_cursor_set2(struct drm_crtc *crtc, struct drm_file *file,
-				uint32_t handle, uint32_t width,
-				uint32_t height, int32_t hot_x, int32_t hot_y)
-{
-	unsigned long value = CURSOR_CLIP_DISPLAY;
-	struct tegra_dc *dc = to_tegra_dc(crtc);
-	struct drm_gem_object *gem;
-	struct tegra_bo *bo = NULL;
-
-	if (!dc->soc->supports_cursor)
-		return -ENXIO;
-
-	if (width != height)
-		return -EINVAL;
-
-	switch (width) {
-	case 32:
-		value |= CURSOR_SIZE_32x32;
-		break;
-
-	case 64:
-		value |= CURSOR_SIZE_64x64;
-		break;
-
-	case 128:
-		value |= CURSOR_SIZE_128x128;
-
-	case 256:
-		value |= CURSOR_SIZE_256x256;
-		break;
-
-	default:
-		return -EINVAL;
-	}
-
-	if (handle) {
-		gem = drm_gem_object_lookup(crtc->dev, file, handle);
-		if (!gem)
-			return -ENOENT;
-
-		bo = to_tegra_bo(gem);
-	}
-
-	if (bo) {
-		unsigned long addr = (bo->paddr & 0xfffffc00) >> 10;
-#ifdef CONFIG_ARCH_DMA_ADDR_T_64BIT
-		unsigned long high = (bo->paddr & 0xfffffffc) >> 32;
-#endif
-
-		tegra_dc_writel(dc, value | addr, DC_DISP_CURSOR_START_ADDR);
-
-#ifdef CONFIG_ARCH_DMA_ADDR_T_64BIT
-		tegra_dc_writel(dc, high, DC_DISP_CURSOR_START_ADDR_HI);
-#endif
-
-		value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
-		value |= CURSOR_ENABLE;
-		tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
-
-		value = tegra_dc_readl(dc, DC_DISP_BLEND_CURSOR_CONTROL);
-		value &= ~CURSOR_DST_BLEND_MASK;
-		value &= ~CURSOR_SRC_BLEND_MASK;
-		value |= CURSOR_MODE_NORMAL;
-		value |= CURSOR_DST_BLEND_NEG_K1_TIMES_SRC;
-		value |= CURSOR_SRC_BLEND_K1_TIMES_SRC;
-		value |= CURSOR_ALPHA;
-		tegra_dc_writel(dc, value, DC_DISP_BLEND_CURSOR_CONTROL);
-	} else {
-		value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
-		value &= ~CURSOR_ENABLE;
-		tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
-	}
-
-	tegra_dc_writel(dc, CURSOR_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
-	tegra_dc_writel(dc, CURSOR_ACT_REQ, DC_CMD_STATE_CONTROL);
-
-	tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
-	tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
-
-	return 0;
-}
-
-static int tegra_dc_cursor_move(struct drm_crtc *crtc, int x, int y)
-{
-	struct tegra_dc *dc = to_tegra_dc(crtc);
-	unsigned long value;
-
-	if (!dc->soc->supports_cursor)
-		return -ENXIO;
-
-	value = ((y & 0x3fff) << 16) | (x & 0x3fff);
-	tegra_dc_writel(dc, value, DC_DISP_CURSOR_POSITION);
-
-	tegra_dc_writel(dc, CURSOR_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
-	tegra_dc_writel(dc, CURSOR_ACT_REQ, DC_CMD_STATE_CONTROL);
-
-	/* XXX: only required on generations earlier than Tegra124? */
-	tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
-	tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
-
-	return 0;
-}
-
 static void tegra_dc_finish_page_flip(struct tegra_dc *dc)
 {
 	struct drm_device *drm = dc->base.dev;
@@ -658,23 +829,32 @@ static void tegra_dc_finish_page_flip(struct tegra_dc *dc)
 	unsigned long flags, base;
 	struct tegra_bo *bo;
 
-	if (!dc->event)
+	spin_lock_irqsave(&drm->event_lock, flags);
+
+	if (!dc->event) {
+		spin_unlock_irqrestore(&drm->event_lock, flags);
 		return;
+	}
 
 	bo = tegra_fb_get_plane(crtc->primary->fb, 0);
 
+	spin_lock_irqsave(&dc->lock, flags);
+
 	/* check if new start address has been latched */
+	tegra_dc_writel(dc, WINDOW_A_SELECT, DC_CMD_DISPLAY_WINDOW_HEADER);
 	tegra_dc_writel(dc, READ_MUX, DC_CMD_STATE_ACCESS);
 	base = tegra_dc_readl(dc, DC_WINBUF_START_ADDR);
 	tegra_dc_writel(dc, 0, DC_CMD_STATE_ACCESS);
 
+	spin_unlock_irqrestore(&dc->lock, flags);
+
 	if (base == bo->paddr + crtc->primary->fb->offsets[0]) {
-		spin_lock_irqsave(&drm->event_lock, flags);
-		drm_send_vblank_event(drm, dc->pipe, dc->event);
-		drm_vblank_put(drm, dc->pipe);
+		drm_crtc_send_vblank_event(crtc, dc->event);
+		drm_crtc_vblank_put(crtc);
 		dc->event = NULL;
-		spin_unlock_irqrestore(&drm->event_lock, flags);
 	}
+
+	spin_unlock_irqrestore(&drm->event_lock, flags);
 }
 
 void tegra_dc_cancel_page_flip(struct drm_crtc *crtc, struct drm_file *file)
@@ -687,7 +867,7 @@ void tegra_dc_cancel_page_flip(struct drm_crtc *crtc, struct drm_file *file)
 
 	if (dc->event && dc->event->base.file_priv == file) {
 		dc->event->base.destroy(&dc->event->base);
-		drm_vblank_put(drm, dc->pipe);
+		drm_crtc_vblank_put(crtc);
 		dc->event = NULL;
 	}
 
@@ -697,16 +877,16 @@ void tegra_dc_cancel_page_flip(struct drm_crtc *crtc, struct drm_file *file)
 static int tegra_dc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 			      struct drm_pending_vblank_event *event, uint32_t page_flip_flags)
 {
+	unsigned int pipe = drm_crtc_index(crtc);
 	struct tegra_dc *dc = to_tegra_dc(crtc);
-	struct drm_device *drm = crtc->dev;
 
 	if (dc->event)
 		return -EBUSY;
 
 	if (event) {
-		event->pipe = dc->pipe;
+		event->pipe = pipe;
 		dc->event = event;
-		drm_vblank_get(drm, dc->pipe);
+		drm_crtc_vblank_get(crtc);
 	}
 
 	tegra_dc_set_base(dc, 0, 0, fb);
@@ -727,8 +907,6 @@ static void tegra_dc_destroy(struct drm_crtc *crtc)
 }
 
 static const struct drm_crtc_funcs tegra_crtc_funcs = {
-	.cursor_set2 = tegra_dc_cursor_set2,
-	.cursor_move = tegra_dc_cursor_move,
 	.page_flip = tegra_dc_page_flip,
 	.set_config = drm_crtc_helper_set_config,
 	.destroy = tegra_dc_destroy,
@@ -742,7 +920,7 @@ static void tegra_crtc_disable(struct drm_crtc *crtc)
 
 	drm_for_each_legacy_plane(plane, &drm->mode_config.plane_list) {
 		if (plane->crtc == crtc) {
-			tegra_plane_disable(plane);
+			tegra_window_plane_disable(plane);
 			plane->crtc = NULL;
 
 			if (plane->fb) {
@@ -752,7 +930,8 @@ static void tegra_crtc_disable(struct drm_crtc *crtc)
 		}
 	}
 
-	drm_vblank_off(drm, dc->pipe);
+	drm_crtc_vblank_off(crtc);
+	tegra_dc_commit(dc);
 }
 
 static bool tegra_crtc_mode_fixup(struct drm_crtc *crtc,
@@ -841,8 +1020,6 @@ static int tegra_crtc_mode_set(struct drm_crtc *crtc,
 	u32 value;
 	int err;
 
-	drm_vblank_pre_modeset(crtc->dev, dc->pipe);
-
 	err = tegra_crtc_setup_clk(crtc, mode);
 	if (err) {
 		dev_err(dc->dev, "failed to setup clock for CRTC: %d\n", err);
@@ -896,6 +1073,8 @@ static void tegra_crtc_prepare(struct drm_crtc *crtc)
 	unsigned int syncpt;
 	unsigned long value;
 
+	drm_crtc_vblank_off(crtc);
+
 	/* hardware initialization */
 	reset_control_deassert(dc->rst);
 	usleep_range(10000, 20000);
@@ -935,15 +1114,9 @@ static void tegra_crtc_prepare(struct drm_crtc *crtc)
 static void tegra_crtc_commit(struct drm_crtc *crtc)
 {
 	struct tegra_dc *dc = to_tegra_dc(crtc);
-	unsigned long value;
-
-	value = GENERAL_UPDATE | WIN_A_UPDATE;
-	tegra_dc_writel(dc, value, DC_CMD_STATE_CONTROL);
 
-	value = GENERAL_ACT_REQ | WIN_A_ACT_REQ;
-	tegra_dc_writel(dc, value, DC_CMD_STATE_CONTROL);
-
-	drm_vblank_post_modeset(crtc->dev, dc->pipe);
+	drm_crtc_vblank_on(crtc);
+	tegra_dc_commit(dc);
 }
 
 static void tegra_crtc_load_lut(struct drm_crtc *crtc)
@@ -978,7 +1151,7 @@ static irqreturn_t tegra_dc_irq(int irq, void *data)
 		/*
 		dev_dbg(dc->dev, "%s(): vertical blank\n", __func__);
 		*/
-		drm_handle_vblank(dc->base.dev, dc->pipe);
+		drm_crtc_handle_vblank(&dc->base);
 		tegra_dc_finish_page_flip(dc);
 	}
 
@@ -997,7 +1170,7 @@ static int tegra_dc_show_regs(struct seq_file *s, void *data)
 	struct tegra_dc *dc = node->info_ent->data;
 
 #define DUMP_REG(name)						\
-	seq_printf(s, "%-40s %#05x %08lx\n", #name, name,	\
+	seq_printf(s, "%-40s %#05x %08x\n", #name, name,	\
 		   tegra_dc_readl(dc, name))
 
 	DUMP_REG(DC_CMD_GENERAL_INCR_SYNCPT);
@@ -1285,9 +1458,40 @@ static int tegra_dc_init(struct host1x_client *client)
 	struct drm_device *drm = dev_get_drvdata(client->parent);
 	struct tegra_dc *dc = host1x_client_to_dc(client);
 	struct tegra_drm *tegra = drm->dev_private;
+	struct drm_plane *primary = NULL;
+	struct drm_plane *cursor = NULL;
 	int err;
 
-	drm_crtc_init(drm, &dc->base, &tegra_crtc_funcs);
+	if (tegra->domain) {
+		err = iommu_attach_device(tegra->domain, dc->dev);
+		if (err < 0) {
+			dev_err(dc->dev, "failed to attach to domain: %d\n",
+				err);
+			return err;
+		}
+
+		dc->domain = tegra->domain;
+	}
+
+	primary = tegra_dc_primary_plane_create(drm, dc);
+	if (IS_ERR(primary)) {
+		err = PTR_ERR(primary);
+		goto cleanup;
+	}
+
+	if (dc->soc->supports_cursor) {
+		cursor = tegra_dc_cursor_plane_create(drm, dc);
+		if (IS_ERR(cursor)) {
+			err = PTR_ERR(cursor);
+			goto cleanup;
+		}
+	}
+
+	err = drm_crtc_init_with_planes(drm, &dc->base, primary, cursor,
+					&tegra_crtc_funcs);
+	if (err < 0)
+		goto cleanup;
+
 	drm_mode_crtc_set_gamma_size(&dc->base, 256);
 	drm_crtc_helper_add(&dc->base, &tegra_crtc_helper_funcs);
 
@@ -1301,12 +1505,12 @@ static int tegra_dc_init(struct host1x_client *client)
 	err = tegra_dc_rgb_init(drm, dc);
 	if (err < 0 && err != -ENODEV) {
 		dev_err(dc->dev, "failed to initialize RGB output: %d\n", err);
-		return err;
+		goto cleanup;
 	}
 
 	err = tegra_dc_add_planes(drm, dc);
 	if (err < 0)
-		return err;
+		goto cleanup;
 
 	if (IS_ENABLED(CONFIG_DEBUG_FS)) {
 		err = tegra_dc_debugfs_init(dc, drm->primary);
@@ -1319,10 +1523,24 @@ static int tegra_dc_init(struct host1x_client *client)
 	if (err < 0) {
 		dev_err(dc->dev, "failed to request IRQ#%u: %d\n", dc->irq,
 			err);
-		return err;
+		goto cleanup;
 	}
 
 	return 0;
+
+cleanup:
+	if (cursor)
+		drm_plane_cleanup(cursor);
+
+	if (primary)
+		drm_plane_cleanup(primary);
+
+	if (tegra->domain) {
+		iommu_detach_device(tegra->domain, dc->dev);
+		dc->domain = NULL;
+	}
+
+	return err;
 }
 
 static int tegra_dc_exit(struct host1x_client *client)
@@ -1344,6 +1562,11 @@ static int tegra_dc_exit(struct host1x_client *client)
 		return err;
 	}
 
+	if (dc->domain) {
+		iommu_detach_device(dc->domain, dc->dev);
+		dc->domain = NULL;
+	}
+
 	return 0;
 }
 
@@ -1357,6 +1580,7 @@ static const struct tegra_dc_soc_info tegra20_dc_soc_info = {
 	.supports_cursor = false,
 	.supports_block_linear = false,
 	.pitch_align = 8,
+	.has_powergate = false,
 };
 
 static const struct tegra_dc_soc_info tegra30_dc_soc_info = {
@@ -1364,6 +1588,7 @@ static const struct tegra_dc_soc_info tegra30_dc_soc_info = {
 	.supports_cursor = false,
 	.supports_block_linear = false,
 	.pitch_align = 8,
+	.has_powergate = false,
 };
 
 static const struct tegra_dc_soc_info tegra114_dc_soc_info = {
@@ -1371,6 +1596,7 @@ static const struct tegra_dc_soc_info tegra114_dc_soc_info = {
 	.supports_cursor = false,
 	.supports_block_linear = false,
 	.pitch_align = 64,
+	.has_powergate = true,
 };
 
 static const struct tegra_dc_soc_info tegra124_dc_soc_info = {
@@ -1378,6 +1604,7 @@ static const struct tegra_dc_soc_info tegra124_dc_soc_info = {
 	.supports_cursor = true,
 	.supports_block_linear = true,
 	.pitch_align = 64,
+	.has_powergate = true,
 };
 
 static const struct of_device_id tegra_dc_of_match[] = {
@@ -1385,6 +1612,9 @@ static const struct of_device_id tegra_dc_of_match[] = {
 		.compatible = "nvidia,tegra124-dc",
 		.data = &tegra124_dc_soc_info,
 	}, {
+		.compatible = "nvidia,tegra114-dc",
+		.data = &tegra114_dc_soc_info,
+	}, {
 		.compatible = "nvidia,tegra30-dc",
 		.data = &tegra30_dc_soc_info,
 	}, {
@@ -1467,9 +1697,34 @@ static int tegra_dc_probe(struct platform_device *pdev)
 		return PTR_ERR(dc->rst);
 	}
 
-	err = clk_prepare_enable(dc->clk);
-	if (err < 0)
-		return err;
+	if (dc->soc->has_powergate) {
+		if (dc->pipe == 0)
+			dc->powergate = TEGRA_POWERGATE_DIS;
+		else
+			dc->powergate = TEGRA_POWERGATE_DISB;
+
+		err = tegra_powergate_sequence_power_up(dc->powergate, dc->clk,
+							dc->rst);
+		if (err < 0) {
+			dev_err(&pdev->dev, "failed to power partition: %d\n",
+				err);
+			return err;
+		}
+	} else {
+		err = clk_prepare_enable(dc->clk);
+		if (err < 0) {
+			dev_err(&pdev->dev, "failed to enable clock: %d\n",
+				err);
+			return err;
+		}
+
+		err = reset_control_deassert(dc->rst);
+		if (err < 0) {
+			dev_err(&pdev->dev, "failed to deassert reset: %d\n",
+				err);
+			return err;
+		}
+	}
 
 	regs = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	dc->regs = devm_ioremap_resource(&pdev->dev, regs);
@@ -1523,6 +1778,10 @@ static int tegra_dc_remove(struct platform_device *pdev)
 	}
 
 	reset_control_assert(dc->rst);
+
+	if (dc->soc->has_powergate)
+		tegra_powergate_power_off(dc->powergate);
+
 	clk_disable_unprepare(dc->clk);
 
 	return 0;
diff --git a/drivers/gpu/drm/tegra/dpaux.c b/drivers/gpu/drm/tegra/dpaux.c
index 708f783ead47..d6b55e3e3716 100644
--- a/drivers/gpu/drm/tegra/dpaux.c
+++ b/drivers/gpu/drm/tegra/dpaux.c
@@ -533,9 +533,9 @@ int tegra_dpaux_train(struct tegra_dpaux *dpaux, struct drm_dp_link *link,
 
 	for (i = 0; i < link->num_lanes; i++)
 		values[i] = DP_TRAIN_MAX_PRE_EMPHASIS_REACHED |
-			    DP_TRAIN_PRE_EMPHASIS_0 |
+			    DP_TRAIN_PRE_EMPH_LEVEL_0 |
 			    DP_TRAIN_MAX_SWING_REACHED |
-			    DP_TRAIN_VOLTAGE_SWING_400;
+			    DP_TRAIN_VOLTAGE_SWING_LEVEL_0;
 
 	err = drm_dp_dpcd_write(&dpaux->aux, DP_TRAINING_LANE0_SET, values,
 				link->num_lanes);
diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c
index 59736bb810cd..d4f827593dfa 100644
--- a/drivers/gpu/drm/tegra/drm.c
+++ b/drivers/gpu/drm/tegra/drm.c
@@ -8,6 +8,7 @@
  */
 
 #include <linux/host1x.h>
+#include <linux/iommu.h>
 
 #include "drm.h"
 #include "gem.h"
@@ -33,6 +34,17 @@ static int tegra_drm_load(struct drm_device *drm, unsigned long flags)
 	if (!tegra)
 		return -ENOMEM;
 
+	if (iommu_present(&platform_bus_type)) {
+		tegra->domain = iommu_domain_alloc(&platform_bus_type);
+		if (IS_ERR(tegra->domain)) {
+			err = PTR_ERR(tegra->domain);
+			goto free;
+		}
+
+		DRM_DEBUG("IOMMU context initialized\n");
+		drm_mm_init(&tegra->mm, 0, SZ_2G);
+	}
+
 	mutex_init(&tegra->clients_lock);
 	INIT_LIST_HEAD(&tegra->clients);
 	drm->dev_private = tegra;
@@ -42,13 +54,13 @@ static int tegra_drm_load(struct drm_device *drm, unsigned long flags)
 
 	err = tegra_drm_fb_prepare(drm);
 	if (err < 0)
-		return err;
+		goto config;
 
 	drm_kms_helper_poll_init(drm);
 
 	err = host1x_device_init(device);
 	if (err < 0)
-		return err;
+		goto fbdev;
 
 	/*
 	 * We don't use the drm_irq_install() helpers provided by the DRM
@@ -59,18 +71,37 @@ static int tegra_drm_load(struct drm_device *drm, unsigned long flags)
 
 	err = drm_vblank_init(drm, drm->mode_config.num_crtc);
 	if (err < 0)
-		return err;
+		goto device;
 
 	err = tegra_drm_fb_init(drm);
 	if (err < 0)
-		return err;
+		goto vblank;
 
 	return 0;
+
+vblank:
+	drm_vblank_cleanup(drm);
+device:
+	host1x_device_exit(device);
+fbdev:
+	drm_kms_helper_poll_fini(drm);
+	tegra_drm_fb_free(drm);
+config:
+	drm_mode_config_cleanup(drm);
+
+	if (tegra->domain) {
+		iommu_domain_free(tegra->domain);
+		drm_mm_takedown(&tegra->mm);
+	}
+free:
+	kfree(tegra);
+	return err;
 }
 
 static int tegra_drm_unload(struct drm_device *drm)
 {
 	struct host1x_device *device = to_host1x_device(drm->dev);
+	struct tegra_drm *tegra = drm->dev_private;
 	int err;
 
 	drm_kms_helper_poll_fini(drm);
@@ -82,6 +113,13 @@ static int tegra_drm_unload(struct drm_device *drm)
 	if (err < 0)
 		return err;
 
+	if (tegra->domain) {
+		iommu_domain_free(tegra->domain);
+		drm_mm_takedown(&tegra->mm);
+	}
+
+	kfree(tegra);
+
 	return 0;
 }
 
@@ -656,24 +694,28 @@ static const struct file_operations tegra_drm_fops = {
 	.llseek = noop_llseek,
 };
 
-static struct drm_crtc *tegra_crtc_from_pipe(struct drm_device *drm, int pipe)
+static struct drm_crtc *tegra_crtc_from_pipe(struct drm_device *drm,
+					     unsigned int pipe)
 {
 	struct drm_crtc *crtc;
 
 	list_for_each_entry(crtc, &drm->mode_config.crtc_list, head) {
-		struct tegra_dc *dc = to_tegra_dc(crtc);
-
-		if (dc->pipe == pipe)
+		if (pipe == drm_crtc_index(crtc))
 			return crtc;
 	}
 
 	return NULL;
 }
 
-static u32 tegra_drm_get_vblank_counter(struct drm_device *dev, int crtc)
+static u32 tegra_drm_get_vblank_counter(struct drm_device *drm, int pipe)
 {
+	struct drm_crtc *crtc = tegra_crtc_from_pipe(drm, pipe);
+
+	if (!crtc)
+		return 0;
+
 	/* TODO: implement real hardware counter using syncpoints */
-	return drm_vblank_count(dev, crtc);
+	return drm_crtc_vblank_count(crtc);
 }
 
 static int tegra_drm_enable_vblank(struct drm_device *drm, int pipe)
diff --git a/drivers/gpu/drm/tegra/drm.h b/drivers/gpu/drm/tegra/drm.h
index e89c70fa82d5..3a3b2e7b5b3f 100644
--- a/drivers/gpu/drm/tegra/drm.h
+++ b/drivers/gpu/drm/tegra/drm.h
@@ -39,6 +39,9 @@ struct tegra_fbdev {
 struct tegra_drm {
 	struct drm_device *drm;
 
+	struct iommu_domain *domain;
+	struct drm_mm mm;
+
 	struct mutex clients_lock;
 	struct list_head clients;
 
@@ -101,6 +104,7 @@ struct tegra_dc {
 	spinlock_t lock;
 
 	struct drm_crtc base;
+	int powergate;
 	int pipe;
 
 	struct clk *clk;
@@ -120,6 +124,8 @@ struct tegra_dc {
 	struct drm_pending_vblank_event *event;
 
 	const struct tegra_dc_soc_info *soc;
+
+	struct iommu_domain *domain;
 };
 
 static inline struct tegra_dc *
@@ -133,16 +139,15 @@ static inline struct tegra_dc *to_tegra_dc(struct drm_crtc *crtc)
 	return crtc ? container_of(crtc, struct tegra_dc, base) : NULL;
 }
 
-static inline void tegra_dc_writel(struct tegra_dc *dc, unsigned long value,
-				   unsigned long reg)
+static inline void tegra_dc_writel(struct tegra_dc *dc, u32 value,
+				   unsigned long offset)
 {
-	writel(value, dc->regs + (reg << 2));
+	writel(value, dc->regs + (offset << 2));
 }
 
-static inline unsigned long tegra_dc_readl(struct tegra_dc *dc,
-					   unsigned long reg)
+static inline u32 tegra_dc_readl(struct tegra_dc *dc, unsigned long offset)
 {
-	return readl(dc->regs + (reg << 2));
+	return readl(dc->regs + (offset << 2));
 }
 
 struct tegra_dc_window {
@@ -287,6 +292,7 @@ bool tegra_fb_is_bottom_up(struct drm_framebuffer *framebuffer);
 int tegra_fb_get_tiling(struct drm_framebuffer *framebuffer,
 			struct tegra_bo_tiling *tiling);
 int tegra_drm_fb_prepare(struct drm_device *drm);
+void tegra_drm_fb_free(struct drm_device *drm);
 int tegra_drm_fb_init(struct drm_device *drm);
 void tegra_drm_fb_exit(struct drm_device *drm);
 #ifdef CONFIG_DRM_TEGRA_FBDEV
diff --git a/drivers/gpu/drm/tegra/dsi.c b/drivers/gpu/drm/tegra/dsi.c
index f7874458926a..33f67fd601c6 100644
--- a/drivers/gpu/drm/tegra/dsi.c
+++ b/drivers/gpu/drm/tegra/dsi.c
@@ -11,6 +11,7 @@
 #include <linux/host1x.h>
 #include <linux/module.h>
 #include <linux/of.h>
+#include <linux/of_platform.h>
 #include <linux/platform_device.h>
 #include <linux/reset.h>
 
@@ -26,9 +27,6 @@
 #include "dsi.h"
 #include "mipi-phy.h"
 
-#define DSI_VIDEO_FIFO_DEPTH (1920 / 4)
-#define DSI_HOST_FIFO_DEPTH 64
-
 struct tegra_dsi {
 	struct host1x_client client;
 	struct tegra_output output;
@@ -54,6 +52,13 @@ struct tegra_dsi {
 
 	struct regulator *vdd;
 	bool enabled;
+
+	unsigned int video_fifo_depth;
+	unsigned int host_fifo_depth;
+
+	/* for ganged-mode support */
+	struct tegra_dsi *master;
+	struct tegra_dsi *slave;
 };
 
 static inline struct tegra_dsi *
@@ -318,6 +323,21 @@ static const u32 pkt_seq_video_non_burst_sync_events[NUM_PKT_SEQ] = {
 	[11] = PKT_ID0(MIPI_DSI_BLANKING_PACKET) | PKT_LEN0(4),
 };
 
+static const u32 pkt_seq_command_mode[NUM_PKT_SEQ] = {
+	[ 0] = 0,
+	[ 1] = 0,
+	[ 2] = 0,
+	[ 3] = 0,
+	[ 4] = 0,
+	[ 5] = 0,
+	[ 6] = PKT_ID0(MIPI_DSI_DCS_LONG_WRITE) | PKT_LEN0(3) | PKT_LP,
+	[ 7] = 0,
+	[ 8] = 0,
+	[ 9] = 0,
+	[10] = PKT_ID0(MIPI_DSI_DCS_LONG_WRITE) | PKT_LEN0(5) | PKT_LP,
+	[11] = 0,
+};
+
 static int tegra_dsi_set_phy_timing(struct tegra_dsi *dsi)
 {
 	struct mipi_dphy_timing timing;
@@ -329,7 +349,7 @@ static int tegra_dsi_set_phy_timing(struct tegra_dsi *dsi)
 	if (rate < 0)
 		return rate;
 
-	period = DIV_ROUND_CLOSEST(1000000000UL, rate * 2);
+	period = DIV_ROUND_CLOSEST(NSEC_PER_SEC, rate * 2);
 
 	err = mipi_dphy_timing_get_default(&timing, period);
 	if (err < 0)
@@ -369,6 +389,9 @@ static int tegra_dsi_set_phy_timing(struct tegra_dsi *dsi)
 		DSI_TIMING_FIELD(timing.tago, period, 1);
 	tegra_dsi_writel(dsi, value, DSI_BTA_TIMING);
 
+	if (dsi->slave)
+		return tegra_dsi_set_phy_timing(dsi->slave);
+
 	return 0;
 }
 
@@ -426,26 +449,59 @@ static int tegra_dsi_get_format(enum mipi_dsi_pixel_format format,
 	return 0;
 }
 
-static int tegra_output_dsi_enable(struct tegra_output *output)
+static void tegra_dsi_ganged_enable(struct tegra_dsi *dsi, unsigned int start,
+				    unsigned int size)
+{
+	u32 value;
+
+	tegra_dsi_writel(dsi, start, DSI_GANGED_MODE_START);
+	tegra_dsi_writel(dsi, size << 16 | size, DSI_GANGED_MODE_SIZE);
+
+	value = DSI_GANGED_MODE_CONTROL_ENABLE;
+	tegra_dsi_writel(dsi, value, DSI_GANGED_MODE_CONTROL);
+}
+
+static void tegra_dsi_enable(struct tegra_dsi *dsi)
+{
+	u32 value;
+
+	value = tegra_dsi_readl(dsi, DSI_POWER_CONTROL);
+	value |= DSI_POWER_CONTROL_ENABLE;
+	tegra_dsi_writel(dsi, value, DSI_POWER_CONTROL);
+
+	if (dsi->slave)
+		tegra_dsi_enable(dsi->slave);
+}
+
+static unsigned int tegra_dsi_get_lanes(struct tegra_dsi *dsi)
+{
+	if (dsi->master)
+		return dsi->master->lanes + dsi->lanes;
+
+	if (dsi->slave)
+		return dsi->lanes + dsi->slave->lanes;
+
+	return dsi->lanes;
+}
+
+static int tegra_dsi_configure(struct tegra_dsi *dsi, unsigned int pipe,
+			       const struct drm_display_mode *mode)
 {
-	struct tegra_dc *dc = to_tegra_dc(output->encoder.crtc);
-	struct drm_display_mode *mode = &dc->base.mode;
 	unsigned int hact, hsw, hbp, hfp, i, mul, div;
-	struct tegra_dsi *dsi = to_dsi(output);
 	enum tegra_dsi_format format;
-	unsigned long value;
 	const u32 *pkt_seq;
+	u32 value;
 	int err;
 
-	if (dsi->enabled)
-		return 0;
-
 	if (dsi->flags & MIPI_DSI_MODE_VIDEO_SYNC_PULSE) {
 		DRM_DEBUG_KMS("Non-burst video mode with sync pulses\n");
 		pkt_seq = pkt_seq_video_non_burst_sync_pulses;
-	} else {
+	} else if (dsi->flags & MIPI_DSI_MODE_VIDEO) {
 		DRM_DEBUG_KMS("Non-burst video mode with sync events\n");
 		pkt_seq = pkt_seq_video_non_burst_sync_events;
+	} else {
+		DRM_DEBUG_KMS("Command mode\n");
+		pkt_seq = pkt_seq_command_mode;
 	}
 
 	err = tegra_dsi_get_muldiv(dsi->format, &mul, &div);
@@ -456,61 +512,136 @@ static int tegra_output_dsi_enable(struct tegra_output *output)
 	if (err < 0)
 		return err;
 
-	err = clk_enable(dsi->clk);
-	if (err < 0)
-		return err;
-
-	reset_control_deassert(dsi->rst);
-
 	value = DSI_CONTROL_CHANNEL(0) | DSI_CONTROL_FORMAT(format) |
 		DSI_CONTROL_LANES(dsi->lanes - 1) |
-		DSI_CONTROL_SOURCE(dc->pipe);
+		DSI_CONTROL_SOURCE(pipe);
 	tegra_dsi_writel(dsi, value, DSI_CONTROL);
 
-	tegra_dsi_writel(dsi, DSI_VIDEO_FIFO_DEPTH, DSI_MAX_THRESHOLD);
+	tegra_dsi_writel(dsi, dsi->video_fifo_depth, DSI_MAX_THRESHOLD);
 
-	value = DSI_HOST_CONTROL_HS | DSI_HOST_CONTROL_CS |
-		DSI_HOST_CONTROL_ECC;
+	value = DSI_HOST_CONTROL_HS;
 	tegra_dsi_writel(dsi, value, DSI_HOST_CONTROL);
 
 	value = tegra_dsi_readl(dsi, DSI_CONTROL);
+
 	if (dsi->flags & MIPI_DSI_CLOCK_NON_CONTINUOUS)
 		value |= DSI_CONTROL_HS_CLK_CTRL;
+
 	value &= ~DSI_CONTROL_TX_TRIG(3);
-	value &= ~DSI_CONTROL_DCS_ENABLE;
+
+	/* enable DCS commands for command mode */
+	if (dsi->flags & MIPI_DSI_MODE_VIDEO)
+		value &= ~DSI_CONTROL_DCS_ENABLE;
+	else
+		value |= DSI_CONTROL_DCS_ENABLE;
+
 	value |= DSI_CONTROL_VIDEO_ENABLE;
 	value &= ~DSI_CONTROL_HOST_ENABLE;
 	tegra_dsi_writel(dsi, value, DSI_CONTROL);
 
-	err = tegra_dsi_set_phy_timing(dsi);
-	if (err < 0)
-		return err;
-
 	for (i = 0; i < NUM_PKT_SEQ; i++)
 		tegra_dsi_writel(dsi, pkt_seq[i], DSI_PKT_SEQ_0_LO + i);
 
-	/* horizontal active pixels */
-	hact = mode->hdisplay * mul / div;
+	if (dsi->flags & MIPI_DSI_MODE_VIDEO) {
+		/* horizontal active pixels */
+		hact = mode->hdisplay * mul / div;
 
-	/* horizontal sync width */
-	hsw = (mode->hsync_end - mode->hsync_start) * mul / div;
-	hsw -= 10;
+		/* horizontal sync width */
+		hsw = (mode->hsync_end - mode->hsync_start) * mul / div;
+		hsw -= 10;
 
-	/* horizontal back porch */
-	hbp = (mode->htotal - mode->hsync_end) * mul / div;
-	hbp -= 14;
+		/* horizontal back porch */
+		hbp = (mode->htotal - mode->hsync_end) * mul / div;
+		hbp -= 14;
 
-	/* horizontal front porch */
-	hfp = (mode->hsync_start  - mode->hdisplay) * mul / div;
-	hfp -= 8;
+		/* horizontal front porch */
+		hfp = (mode->hsync_start - mode->hdisplay) * mul / div;
+		hfp -= 8;
 
-	tegra_dsi_writel(dsi, hsw << 16 | 0, DSI_PKT_LEN_0_1);
-	tegra_dsi_writel(dsi, hact << 16 | hbp, DSI_PKT_LEN_2_3);
-	tegra_dsi_writel(dsi, hfp, DSI_PKT_LEN_4_5);
-	tegra_dsi_writel(dsi, 0x0f0f << 16, DSI_PKT_LEN_6_7);
+		tegra_dsi_writel(dsi, hsw << 16 | 0, DSI_PKT_LEN_0_1);
+		tegra_dsi_writel(dsi, hact << 16 | hbp, DSI_PKT_LEN_2_3);
+		tegra_dsi_writel(dsi, hfp, DSI_PKT_LEN_4_5);
+		tegra_dsi_writel(dsi, 0x0f0f << 16, DSI_PKT_LEN_6_7);
 
-	/* set SOL delay */
-	tegra_dsi_writel(dsi, 8 * mul / div, DSI_SOL_DELAY);
+		/* set SOL delay (for non-burst mode only) */
+		tegra_dsi_writel(dsi, 8 * mul / div, DSI_SOL_DELAY);
+
+		/* TODO: implement ganged mode */
+	} else {
+		u16 bytes;
+
+		if (dsi->master || dsi->slave) {
+			/*
+			 * For ganged mode, assume symmetric left-right mode.
+			 */
+			bytes = 1 + (mode->hdisplay / 2) * mul / div;
+		} else {
+			/* 1 byte (DCS command) + pixel data */
+			bytes = 1 + mode->hdisplay * mul / div;
+		}
+
+		tegra_dsi_writel(dsi, 0, DSI_PKT_LEN_0_1);
+		tegra_dsi_writel(dsi, bytes << 16, DSI_PKT_LEN_2_3);
+		tegra_dsi_writel(dsi, bytes << 16, DSI_PKT_LEN_4_5);
+		tegra_dsi_writel(dsi, 0, DSI_PKT_LEN_6_7);
+
+		value = MIPI_DCS_WRITE_MEMORY_START << 8 |
+			MIPI_DCS_WRITE_MEMORY_CONTINUE;
+		tegra_dsi_writel(dsi, value, DSI_DCS_CMDS);
+
+		/* set SOL delay */
+		if (dsi->master || dsi->slave) {
+			unsigned int lanes = tegra_dsi_get_lanes(dsi);
+			unsigned long delay, bclk, bclk_ganged;
+
+			/* SOL to valid, valid to FIFO and FIFO write delay */
+			delay = 4 + 4 + 2;
+			delay = DIV_ROUND_UP(delay * mul, div * lanes);
+			/* FIFO read delay */
+			delay = delay + 6;
+
+			bclk = DIV_ROUND_UP(mode->htotal * mul, div * lanes);
+			bclk_ganged = DIV_ROUND_UP(bclk * lanes / 2, lanes);
+			value = bclk - bclk_ganged + delay + 20;
+		} else {
+			/* TODO: revisit for non-ganged mode */
+			value = 8 * mul / div;
+		}
+
+		tegra_dsi_writel(dsi, value, DSI_SOL_DELAY);
+	}
+
+	if (dsi->slave) {
+		err = tegra_dsi_configure(dsi->slave, pipe, mode);
+		if (err < 0)
+			return err;
+
+		/*
+		 * TODO: Support modes other than symmetrical left-right
+		 * split.
+		 */
+		tegra_dsi_ganged_enable(dsi, 0, mode->hdisplay / 2);
+		tegra_dsi_ganged_enable(dsi->slave, mode->hdisplay / 2,
+					mode->hdisplay / 2);
+	}
+
+	return 0;
+}
+
+static int tegra_output_dsi_enable(struct tegra_output *output)
+{
+	struct tegra_dc *dc = to_tegra_dc(output->encoder.crtc);
+	const struct drm_display_mode *mode = &dc->base.mode;
+	struct tegra_dsi *dsi = to_dsi(output);
+	u32 value;
+	int err;
+
+	if (dsi->enabled)
+		return 0;
+
+	err = tegra_dsi_configure(dsi, dc->pipe, mode);
+	if (err < 0)
+		return err;
 
 	/* enable display controller */
 	value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
@@ -531,28 +662,79 @@ static int tegra_output_dsi_enable(struct tegra_output *output)
 	tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
 
 	/* enable DSI controller */
-	value = tegra_dsi_readl(dsi, DSI_POWER_CONTROL);
-	value |= DSI_POWER_CONTROL_ENABLE;
-	tegra_dsi_writel(dsi, value, DSI_POWER_CONTROL);
+	tegra_dsi_enable(dsi);
 
 	dsi->enabled = true;
 
 	return 0;
 }
 
+static int tegra_dsi_wait_idle(struct tegra_dsi *dsi, unsigned long timeout)
+{
+	u32 value;
+
+	timeout = jiffies + msecs_to_jiffies(timeout);
+
+	while (time_before(jiffies, timeout)) {
+		value = tegra_dsi_readl(dsi, DSI_STATUS);
+		if (value & DSI_STATUS_IDLE)
+			return 0;
+
+		usleep_range(1000, 2000);
+	}
+
+	return -ETIMEDOUT;
+}
+
+static void tegra_dsi_video_disable(struct tegra_dsi *dsi)
+{
+	u32 value;
+
+	value = tegra_dsi_readl(dsi, DSI_CONTROL);
+	value &= ~DSI_CONTROL_VIDEO_ENABLE;
+	tegra_dsi_writel(dsi, value, DSI_CONTROL);
+
+	if (dsi->slave)
+		tegra_dsi_video_disable(dsi->slave);
+}
+
+static void tegra_dsi_ganged_disable(struct tegra_dsi *dsi)
+{
+	tegra_dsi_writel(dsi, 0, DSI_GANGED_MODE_START);
+	tegra_dsi_writel(dsi, 0, DSI_GANGED_MODE_SIZE);
+	tegra_dsi_writel(dsi, 0, DSI_GANGED_MODE_CONTROL);
+}
+
+static void tegra_dsi_disable(struct tegra_dsi *dsi)
+{
+	u32 value;
+
+	if (dsi->slave) {
+		tegra_dsi_ganged_disable(dsi->slave);
+		tegra_dsi_ganged_disable(dsi);
+	}
+
+	value = tegra_dsi_readl(dsi, DSI_POWER_CONTROL);
+	value &= ~DSI_POWER_CONTROL_ENABLE;
+	tegra_dsi_writel(dsi, value, DSI_POWER_CONTROL);
+
+	if (dsi->slave)
+		tegra_dsi_disable(dsi->slave);
+
+	usleep_range(5000, 10000);
+}
+
 static int tegra_output_dsi_disable(struct tegra_output *output)
 {
 	struct tegra_dc *dc = to_tegra_dc(output->encoder.crtc);
 	struct tegra_dsi *dsi = to_dsi(output);
 	unsigned long value;
+	int err;
 
 	if (!dsi->enabled)
 		return 0;
 
-	/* disable DSI controller */
-	value = tegra_dsi_readl(dsi, DSI_POWER_CONTROL);
-	value &= ~DSI_POWER_CONTROL_ENABLE;
-	tegra_dsi_writel(dsi, value, DSI_POWER_CONTROL);
+	tegra_dsi_video_disable(dsi);
 
 	/*
 	 * The following accesses registers of the display controller, so make
@@ -576,39 +758,68 @@ static int tegra_output_dsi_disable(struct tegra_output *output)
 		tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
 	}
 
-	clk_disable(dsi->clk);
+	err = tegra_dsi_wait_idle(dsi, 100);
+	if (err < 0)
+		dev_dbg(dsi->dev, "failed to idle DSI: %d\n", err);
+
+	tegra_dsi_disable(dsi);
 
 	dsi->enabled = false;
 
 	return 0;
 }
 
+static void tegra_dsi_set_timeout(struct tegra_dsi *dsi, unsigned long bclk,
+				  unsigned int vrefresh)
+{
+	unsigned int timeout;
+	u32 value;
+
+	/* one frame high-speed transmission timeout */
+	timeout = (bclk / vrefresh) / 512;
+	value = DSI_TIMEOUT_LRX(0x2000) | DSI_TIMEOUT_HTX(timeout);
+	tegra_dsi_writel(dsi, value, DSI_TIMEOUT_0);
+
+	/* 2 ms peripheral timeout for panel */
+	timeout = 2 * bclk / 512 * 1000;
+	value = DSI_TIMEOUT_PR(timeout) | DSI_TIMEOUT_TA(0x2000);
+	tegra_dsi_writel(dsi, value, DSI_TIMEOUT_1);
+
+	value = DSI_TALLY_TA(0) | DSI_TALLY_LRX(0) | DSI_TALLY_HTX(0);
+	tegra_dsi_writel(dsi, value, DSI_TO_TALLY);
+
+	if (dsi->slave)
+		tegra_dsi_set_timeout(dsi->slave, bclk, vrefresh);
+}
+
 static int tegra_output_dsi_setup_clock(struct tegra_output *output,
 					struct clk *clk, unsigned long pclk,
 					unsigned int *divp)
 {
 	struct tegra_dc *dc = to_tegra_dc(output->encoder.crtc);
 	struct drm_display_mode *mode = &dc->base.mode;
-	unsigned int timeout, mul, div, vrefresh;
 	struct tegra_dsi *dsi = to_dsi(output);
-	unsigned long bclk, plld, value;
+	unsigned int mul, div, vrefresh, lanes;
+	unsigned long bclk, plld;
 	int err;
 
+	lanes = tegra_dsi_get_lanes(dsi);
+
 	err = tegra_dsi_get_muldiv(dsi->format, &mul, &div);
 	if (err < 0)
 		return err;
 
-	DRM_DEBUG_KMS("mul: %u, div: %u, lanes: %u\n", mul, div, dsi->lanes);
+	DRM_DEBUG_KMS("mul: %u, div: %u, lanes: %u\n", mul, div, lanes);
 	vrefresh = drm_mode_vrefresh(mode);
 	DRM_DEBUG_KMS("vrefresh: %u\n", vrefresh);
 
 	/* compute byte clock */
-	bclk = (pclk * mul) / (div * dsi->lanes);
+	bclk = (pclk * mul) / (div * lanes);
 
 	/*
 	 * Compute bit clock and round up to the next MHz.
 	 */
-	plld = DIV_ROUND_UP(bclk * 8, 1000000) * 1000000;
+	plld = DIV_ROUND_UP(bclk * 8, USEC_PER_SEC) * USEC_PER_SEC;
 
 	/*
 	 * We divide the frequency by two here, but we make up for that by
@@ -640,25 +851,17 @@ static int tegra_output_dsi_setup_clock(struct tegra_output *output,
 	 * not working properly otherwise. Perhaps the PLLs cannot generate
 	 * frequencies sufficiently high.
 	 */
-	*divp = ((8 * mul) / (div * dsi->lanes)) - 2;
+	*divp = ((8 * mul) / (div * lanes)) - 2;
 
 	/*
 	 * XXX: Move the below somewhere else so that we don't need to have
 	 * access to the vrefresh in this function?
 	 */
+	tegra_dsi_set_timeout(dsi, bclk, vrefresh);
 
-	/* one frame high-speed transmission timeout */
-	timeout = (bclk / vrefresh) / 512;
-	value = DSI_TIMEOUT_LRX(0x2000) | DSI_TIMEOUT_HTX(timeout);
-	tegra_dsi_writel(dsi, value, DSI_TIMEOUT_0);
-
-	/* 2 ms peripheral timeout for panel */
-	timeout = 2 * bclk / 512 * 1000;
-	value = DSI_TIMEOUT_PR(timeout) | DSI_TIMEOUT_TA(0x2000);
-	tegra_dsi_writel(dsi, value, DSI_TIMEOUT_1);
-
-	value = DSI_TALLY_TA(0) | DSI_TALLY_LRX(0) | DSI_TALLY_HTX(0);
-	tegra_dsi_writel(dsi, value, DSI_TO_TALLY);
+	err = tegra_dsi_set_phy_timing(dsi);
+	if (err < 0)
+		return err;
 
 	return 0;
 }
@@ -695,7 +898,7 @@ static int tegra_dsi_pad_enable(struct tegra_dsi *dsi)
 
 static int tegra_dsi_pad_calibrate(struct tegra_dsi *dsi)
 {
-	unsigned long value;
+	u32 value;
 
 	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_0);
 	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_1);
@@ -720,14 +923,17 @@ static int tegra_dsi_init(struct host1x_client *client)
 	struct tegra_dsi *dsi = host1x_client_to_dsi(client);
 	int err;
 
-	dsi->output.type = TEGRA_OUTPUT_DSI;
-	dsi->output.dev = client->dev;
-	dsi->output.ops = &dsi_ops;
-
-	err = tegra_output_init(drm, &dsi->output);
-	if (err < 0) {
-		dev_err(client->dev, "output setup failed: %d\n", err);
-		return err;
+	/* Gangsters must not register their own outputs. */
+	if (!dsi->master) {
+		dsi->output.type = TEGRA_OUTPUT_DSI;
+		dsi->output.dev = client->dev;
+		dsi->output.ops = &dsi_ops;
+
+		err = tegra_output_init(drm, &dsi->output);
+		if (err < 0) {
+			dev_err(client->dev, "output setup failed: %d\n", err);
+			return err;
+		}
 	}
 
 	if (IS_ENABLED(CONFIG_DEBUG_FS)) {
@@ -736,12 +942,6 @@ static int tegra_dsi_init(struct host1x_client *client)
 			dev_err(dsi->dev, "debugfs setup failed: %d\n", err);
 	}
 
-	err = tegra_dsi_pad_calibrate(dsi);
-	if (err < 0) {
-		dev_err(dsi->dev, "MIPI calibration failed: %d\n", err);
-		return err;
-	}
-
 	return 0;
 }
 
@@ -756,16 +956,20 @@ static int tegra_dsi_exit(struct host1x_client *client)
 			dev_err(dsi->dev, "debugfs cleanup failed: %d\n", err);
 	}
 
-	err = tegra_output_disable(&dsi->output);
-	if (err < 0) {
-		dev_err(client->dev, "output failed to disable: %d\n", err);
-		return err;
-	}
-
-	err = tegra_output_exit(&dsi->output);
-	if (err < 0) {
-		dev_err(client->dev, "output cleanup failed: %d\n", err);
-		return err;
+	if (!dsi->master) {
+		err = tegra_output_disable(&dsi->output);
+		if (err < 0) {
+			dev_err(client->dev, "output failed to disable: %d\n",
+				err);
+			return err;
+		}
+
+		err = tegra_output_exit(&dsi->output);
+		if (err < 0) {
+			dev_err(client->dev, "output cleanup failed: %d\n",
+				err);
+			return err;
+		}
 	}
 
 	return 0;
@@ -792,20 +996,324 @@ static int tegra_dsi_setup_clocks(struct tegra_dsi *dsi)
 	return 0;
 }
 
+static const char * const error_report[16] = {
+	"SoT Error",
+	"SoT Sync Error",
+	"EoT Sync Error",
+	"Escape Mode Entry Command Error",
+	"Low-Power Transmit Sync Error",
+	"Peripheral Timeout Error",
+	"False Control Error",
+	"Contention Detected",
+	"ECC Error, single-bit",
+	"ECC Error, multi-bit",
+	"Checksum Error",
+	"DSI Data Type Not Recognized",
+	"DSI VC ID Invalid",
+	"Invalid Transmission Length",
+	"Reserved",
+	"DSI Protocol Violation",
+};
+
+static ssize_t tegra_dsi_read_response(struct tegra_dsi *dsi,
+				       const struct mipi_dsi_msg *msg,
+				       size_t count)
+{
+	u8 *rx = msg->rx_buf;
+	unsigned int i, j, k;
+	size_t size = 0;
+	u16 errors;
+	u32 value;
+
+	/* read and parse packet header */
+	value = tegra_dsi_readl(dsi, DSI_RD_DATA);
+
+	switch (value & 0x3f) {
+	case MIPI_DSI_RX_ACKNOWLEDGE_AND_ERROR_REPORT:
+		errors = (value >> 8) & 0xffff;
+		dev_dbg(dsi->dev, "Acknowledge and error report: %04x\n",
+			errors);
+		for (i = 0; i < ARRAY_SIZE(error_report); i++)
+			if (errors & BIT(i))
+				dev_dbg(dsi->dev, "  %2u: %s\n", i,
+					error_report[i]);
+		break;
+
+	case MIPI_DSI_RX_DCS_SHORT_READ_RESPONSE_1BYTE:
+		rx[0] = (value >> 8) & 0xff;
+		size = 1;
+		break;
+
+	case MIPI_DSI_RX_DCS_SHORT_READ_RESPONSE_2BYTE:
+		rx[0] = (value >>  8) & 0xff;
+		rx[1] = (value >> 16) & 0xff;
+		size = 2;
+		break;
+
+	case MIPI_DSI_RX_DCS_LONG_READ_RESPONSE:
+		size = ((value >> 8) & 0xff00) | ((value >> 8) & 0xff);
+		break;
+
+	case MIPI_DSI_RX_GENERIC_LONG_READ_RESPONSE:
+		size = ((value >> 8) & 0xff00) | ((value >> 8) & 0xff);
+		break;
+
+	default:
+		dev_err(dsi->dev, "unhandled response type: %02x\n",
+			value & 0x3f);
+		return -EPROTO;
+	}
+
+	size = min(size, msg->rx_len);
+
+	if (msg->rx_buf && size > 0) {
+		for (i = 0, j = 0; i < count - 1; i++, j += 4) {
+			u8 *rx = msg->rx_buf + j;
+
+			value = tegra_dsi_readl(dsi, DSI_RD_DATA);
+
+			for (k = 0; k < 4 && (j + k) < msg->rx_len; k++)
+				rx[j + k] = (value >> (k << 3)) & 0xff;
+		}
+	}
+
+	return size;
+}
+
+static int tegra_dsi_transmit(struct tegra_dsi *dsi, unsigned long timeout)
+{
+	tegra_dsi_writel(dsi, DSI_TRIGGER_HOST, DSI_TRIGGER);
+
+	timeout = jiffies + msecs_to_jiffies(timeout);
+
+	while (time_before(jiffies, timeout)) {
+		u32 value = tegra_dsi_readl(dsi, DSI_TRIGGER);
+		if ((value & DSI_TRIGGER_HOST) == 0)
+			return 0;
+
+		usleep_range(1000, 2000);
+	}
+
+	DRM_DEBUG_KMS("timeout waiting for transmission to complete\n");
+	return -ETIMEDOUT;
+}
+
+static int tegra_dsi_wait_for_response(struct tegra_dsi *dsi,
+				       unsigned long timeout)
+{
+	timeout = jiffies + msecs_to_jiffies(250);
+
+	while (time_before(jiffies, timeout)) {
+		u32 value = tegra_dsi_readl(dsi, DSI_STATUS);
+		u8 count = value & 0x1f;
+
+		if (count > 0)
+			return count;
+
+		usleep_range(1000, 2000);
+	}
+
+	DRM_DEBUG_KMS("peripheral returned no data\n");
+	return -ETIMEDOUT;
+}
+
+static void tegra_dsi_writesl(struct tegra_dsi *dsi, unsigned long offset,
+			      const void *buffer, size_t size)
+{
+	const u8 *buf = buffer;
+	size_t i, j;
+	u32 value;
+
+	for (j = 0; j < size; j += 4) {
+		value = 0;
+
+		for (i = 0; i < 4 && j + i < size; i++)
+			value |= buf[j + i] << (i << 3);
+
+		tegra_dsi_writel(dsi, value, DSI_WR_DATA);
+	}
+}
+
+static ssize_t tegra_dsi_host_transfer(struct mipi_dsi_host *host,
+				       const struct mipi_dsi_msg *msg)
+{
+	struct tegra_dsi *dsi = host_to_tegra(host);
+	struct mipi_dsi_packet packet;
+	const u8 *header;
+	size_t count;
+	ssize_t err;
+	u32 value;
+
+	err = mipi_dsi_create_packet(&packet, msg);
+	if (err < 0)
+		return err;
+
+	header = packet.header;
+
+	/* maximum FIFO depth is 1920 words */
+	if (packet.size > dsi->video_fifo_depth * 4)
+		return -ENOSPC;
+
+	/* reset underflow/overflow flags */
+	value = tegra_dsi_readl(dsi, DSI_STATUS);
+	if (value & (DSI_STATUS_UNDERFLOW | DSI_STATUS_OVERFLOW)) {
+		value = DSI_HOST_CONTROL_FIFO_RESET;
+		tegra_dsi_writel(dsi, value, DSI_HOST_CONTROL);
+		usleep_range(10, 20);
+	}
+
+	value = tegra_dsi_readl(dsi, DSI_POWER_CONTROL);
+	value |= DSI_POWER_CONTROL_ENABLE;
+	tegra_dsi_writel(dsi, value, DSI_POWER_CONTROL);
+
+	usleep_range(5000, 10000);
+
+	value = DSI_HOST_CONTROL_CRC_RESET | DSI_HOST_CONTROL_TX_TRIG_HOST |
+		DSI_HOST_CONTROL_CS | DSI_HOST_CONTROL_ECC;
+
+	if ((msg->flags & MIPI_DSI_MSG_USE_LPM) == 0)
+		value |= DSI_HOST_CONTROL_HS;
+
+	/*
+	 * The host FIFO has a maximum of 64 words, so larger transmissions
+	 * need to use the video FIFO.
+	 */
+	if (packet.size > dsi->host_fifo_depth * 4)
+		value |= DSI_HOST_CONTROL_FIFO_SEL;
+
+	tegra_dsi_writel(dsi, value, DSI_HOST_CONTROL);
+
+	/*
+	 * For reads and messages with explicitly requested ACK, generate a
+	 * BTA sequence after the transmission of the packet.
+	 */
+	if ((msg->flags & MIPI_DSI_MSG_REQ_ACK) ||
+	    (msg->rx_buf && msg->rx_len > 0)) {
+		value = tegra_dsi_readl(dsi, DSI_HOST_CONTROL);
+		value |= DSI_HOST_CONTROL_PKT_BTA;
+		tegra_dsi_writel(dsi, value, DSI_HOST_CONTROL);
+	}
+
+	value = DSI_CONTROL_LANES(0) | DSI_CONTROL_HOST_ENABLE;
+	tegra_dsi_writel(dsi, value, DSI_CONTROL);
+
+	/* write packet header, ECC is generated by hardware */
+	value = header[2] << 16 | header[1] << 8 | header[0];
+	tegra_dsi_writel(dsi, value, DSI_WR_DATA);
+
+	/* write payload (if any) */
+	if (packet.payload_length > 0)
+		tegra_dsi_writesl(dsi, DSI_WR_DATA, packet.payload,
+				  packet.payload_length);
+
+	err = tegra_dsi_transmit(dsi, 250);
+	if (err < 0)
+		return err;
+
+	if ((msg->flags & MIPI_DSI_MSG_REQ_ACK) ||
+	    (msg->rx_buf && msg->rx_len > 0)) {
+		err = tegra_dsi_wait_for_response(dsi, 250);
+		if (err < 0)
+			return err;
+
+		count = err;
+
+		value = tegra_dsi_readl(dsi, DSI_RD_DATA);
+		switch (value) {
+		case 0x84:
+			/*
+			dev_dbg(dsi->dev, "ACK\n");
+			*/
+			break;
+
+		case 0x87:
+			/*
+			dev_dbg(dsi->dev, "ESCAPE\n");
+			*/
+			break;
+
+		default:
+			dev_err(dsi->dev, "unknown status: %08x\n", value);
+			break;
+		}
+
+		if (count > 1) {
+			err = tegra_dsi_read_response(dsi, msg, count);
+			if (err < 0)
+				dev_err(dsi->dev,
+					"failed to parse response: %zd\n",
+					err);
+			else {
+				/*
+				 * For read commands, return the number of
+				 * bytes returned by the peripheral.
+				 */
+				count = err;
+			}
+		}
+	} else {
+		/*
+		 * For write commands, we have transmitted the 4-byte header
+		 * plus the variable-length payload.
+		 */
+		count = 4 + packet.payload_length;
+	}
+
+	return count;
+}
+
+static int tegra_dsi_ganged_setup(struct tegra_dsi *dsi)
+{
+	struct clk *parent;
+	int err;
+
+	/* make sure both DSI controllers share the same PLL */
+	parent = clk_get_parent(dsi->slave->clk);
+	if (!parent)
+		return -EINVAL;
+
+	err = clk_set_parent(parent, dsi->clk_parent);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+
 static int tegra_dsi_host_attach(struct mipi_dsi_host *host,
 				 struct mipi_dsi_device *device)
 {
 	struct tegra_dsi *dsi = host_to_tegra(host);
-	struct tegra_output *output = &dsi->output;
 
 	dsi->flags = device->mode_flags;
 	dsi->format = device->format;
 	dsi->lanes = device->lanes;
 
-	output->panel = of_drm_find_panel(device->dev.of_node);
-	if (output->panel) {
-		if (output->connector.dev)
+	if (dsi->slave) {
+		int err;
+
+		dev_dbg(dsi->dev, "attaching dual-channel device %s\n",
+			dev_name(&device->dev));
+
+		err = tegra_dsi_ganged_setup(dsi);
+		if (err < 0) {
+			dev_err(dsi->dev, "failed to set up ganged mode: %d\n",
+				err);
+			return err;
+		}
+	}
+
+	/*
+	 * Slaves don't have a panel associated with them, so they provide
+	 * merely the second channel.
+	 */
+	if (!dsi->master) {
+		struct tegra_output *output = &dsi->output;
+
+		output->panel = of_drm_find_panel(device->dev.of_node);
+		if (output->panel && output->connector.dev) {
+			drm_panel_attach(output->panel, &output->connector);
 			drm_helper_hpd_irq_event(output->connector.dev);
+		}
 	}
 
 	return 0;
@@ -818,10 +1326,10 @@ static int tegra_dsi_host_detach(struct mipi_dsi_host *host,
 	struct tegra_output *output = &dsi->output;
 
 	if (output->panel && &device->dev == output->panel->dev) {
+		output->panel = NULL;
+
 		if (output->connector.dev)
 			drm_helper_hpd_irq_event(output->connector.dev);
-
-		output->panel = NULL;
 	}
 
 	return 0;
@@ -830,8 +1338,29 @@ static int tegra_dsi_host_detach(struct mipi_dsi_host *host,
 static const struct mipi_dsi_host_ops tegra_dsi_host_ops = {
 	.attach = tegra_dsi_host_attach,
 	.detach = tegra_dsi_host_detach,
+	.transfer = tegra_dsi_host_transfer,
 };
 
+static int tegra_dsi_ganged_probe(struct tegra_dsi *dsi)
+{
+	struct device_node *np;
+
+	np = of_parse_phandle(dsi->dev->of_node, "nvidia,ganged-mode", 0);
+	if (np) {
+		struct platform_device *gangster = of_find_device_by_node(np);
+
+		dsi->slave = platform_get_drvdata(gangster);
+		of_node_put(np);
+
+		if (!dsi->slave)
+			return -EPROBE_DEFER;
+
+		dsi->slave->master = dsi;
+	}
+
+	return 0;
+}
+
 static int tegra_dsi_probe(struct platform_device *pdev)
 {
 	struct tegra_dsi *dsi;
@@ -843,11 +1372,19 @@ static int tegra_dsi_probe(struct platform_device *pdev)
 		return -ENOMEM;
 
 	dsi->output.dev = dsi->dev = &pdev->dev;
+	dsi->video_fifo_depth = 1920;
+	dsi->host_fifo_depth = 64;
+
+	err = tegra_dsi_ganged_probe(dsi);
+	if (err < 0)
+		return err;
 
 	err = tegra_output_probe(&dsi->output);
 	if (err < 0)
 		return err;
 
+	dsi->output.connector.polled = DRM_CONNECTOR_POLL_HPD;
+
 	/*
 	 * Assume these values by default. When a DSI peripheral driver
 	 * attaches to the DSI host, the parameters will be taken from
@@ -861,68 +1398,83 @@ static int tegra_dsi_probe(struct platform_device *pdev)
 	if (IS_ERR(dsi->rst))
 		return PTR_ERR(dsi->rst);
 
+	err = reset_control_deassert(dsi->rst);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to bring DSI out of reset: %d\n",
+			err);
+		return err;
+	}
+
 	dsi->clk = devm_clk_get(&pdev->dev, NULL);
 	if (IS_ERR(dsi->clk)) {
 		dev_err(&pdev->dev, "cannot get DSI clock\n");
-		return PTR_ERR(dsi->clk);
+		err = PTR_ERR(dsi->clk);
+		goto reset;
 	}
 
 	err = clk_prepare_enable(dsi->clk);
 	if (err < 0) {
 		dev_err(&pdev->dev, "cannot enable DSI clock\n");
-		return err;
+		goto reset;
 	}
 
 	dsi->clk_lp = devm_clk_get(&pdev->dev, "lp");
 	if (IS_ERR(dsi->clk_lp)) {
 		dev_err(&pdev->dev, "cannot get low-power clock\n");
-		return PTR_ERR(dsi->clk_lp);
+		err = PTR_ERR(dsi->clk_lp);
+		goto disable_clk;
 	}
 
 	err = clk_prepare_enable(dsi->clk_lp);
 	if (err < 0) {
 		dev_err(&pdev->dev, "cannot enable low-power clock\n");
-		return err;
+		goto disable_clk;
 	}
 
 	dsi->clk_parent = devm_clk_get(&pdev->dev, "parent");
 	if (IS_ERR(dsi->clk_parent)) {
 		dev_err(&pdev->dev, "cannot get parent clock\n");
-		return PTR_ERR(dsi->clk_parent);
-	}
-
-	err = clk_prepare_enable(dsi->clk_parent);
-	if (err < 0) {
-		dev_err(&pdev->dev, "cannot enable parent clock\n");
-		return err;
+		err = PTR_ERR(dsi->clk_parent);
+		goto disable_clk_lp;
 	}
 
 	dsi->vdd = devm_regulator_get(&pdev->dev, "avdd-dsi-csi");
 	if (IS_ERR(dsi->vdd)) {
 		dev_err(&pdev->dev, "cannot get VDD supply\n");
-		return PTR_ERR(dsi->vdd);
+		err = PTR_ERR(dsi->vdd);
+		goto disable_clk_lp;
 	}
 
 	err = regulator_enable(dsi->vdd);
 	if (err < 0) {
 		dev_err(&pdev->dev, "cannot enable VDD supply\n");
-		return err;
+		goto disable_clk_lp;
 	}
 
 	err = tegra_dsi_setup_clocks(dsi);
 	if (err < 0) {
 		dev_err(&pdev->dev, "cannot setup clocks\n");
-		return err;
+		goto disable_vdd;
 	}
 
 	regs = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	dsi->regs = devm_ioremap_resource(&pdev->dev, regs);
-	if (IS_ERR(dsi->regs))
-		return PTR_ERR(dsi->regs);
+	if (IS_ERR(dsi->regs)) {
+		err = PTR_ERR(dsi->regs);
+		goto disable_vdd;
+	}
 
 	dsi->mipi = tegra_mipi_request(&pdev->dev);
-	if (IS_ERR(dsi->mipi))
-		return PTR_ERR(dsi->mipi);
+	if (IS_ERR(dsi->mipi)) {
+		err = PTR_ERR(dsi->mipi);
+		goto disable_vdd;
+	}
+
+	err = tegra_dsi_pad_calibrate(dsi);
+	if (err < 0) {
+		dev_err(dsi->dev, "MIPI calibration failed: %d\n", err);
+		goto mipi_free;
+	}
 
 	dsi->host.ops = &tegra_dsi_host_ops;
 	dsi->host.dev = &pdev->dev;
@@ -930,7 +1482,7 @@ static int tegra_dsi_probe(struct platform_device *pdev)
 	err = mipi_dsi_host_register(&dsi->host);
 	if (err < 0) {
 		dev_err(&pdev->dev, "failed to register DSI host: %d\n", err);
-		return err;
+		goto mipi_free;
 	}
 
 	INIT_LIST_HEAD(&dsi->client.list);
@@ -941,12 +1493,26 @@ static int tegra_dsi_probe(struct platform_device *pdev)
 	if (err < 0) {
 		dev_err(&pdev->dev, "failed to register host1x client: %d\n",
 			err);
-		return err;
+		goto unregister;
 	}
 
 	platform_set_drvdata(pdev, dsi);
 
 	return 0;
+
+unregister:
+	mipi_dsi_host_unregister(&dsi->host);
+mipi_free:
+	tegra_mipi_free(dsi->mipi);
+disable_vdd:
+	regulator_disable(dsi->vdd);
+disable_clk_lp:
+	clk_disable_unprepare(dsi->clk_lp);
+disable_clk:
+	clk_disable_unprepare(dsi->clk);
+reset:
+	reset_control_assert(dsi->rst);
+	return err;
 }
 
 static int tegra_dsi_remove(struct platform_device *pdev)
@@ -965,7 +1531,6 @@ static int tegra_dsi_remove(struct platform_device *pdev)
 	tegra_mipi_free(dsi->mipi);
 
 	regulator_disable(dsi->vdd);
-	clk_disable_unprepare(dsi->clk_parent);
 	clk_disable_unprepare(dsi->clk_lp);
 	clk_disable_unprepare(dsi->clk);
 	reset_control_assert(dsi->rst);
diff --git a/drivers/gpu/drm/tegra/dsi.h b/drivers/gpu/drm/tegra/dsi.h
index 5ce610d08d77..bad1006a5150 100644
--- a/drivers/gpu/drm/tegra/dsi.h
+++ b/drivers/gpu/drm/tegra/dsi.h
@@ -21,9 +21,16 @@
 #define DSI_INT_STATUS			0x0d
 #define DSI_INT_MASK			0x0e
 #define DSI_HOST_CONTROL		0x0f
+#define DSI_HOST_CONTROL_FIFO_RESET	(1 << 21)
+#define DSI_HOST_CONTROL_CRC_RESET	(1 << 20)
+#define DSI_HOST_CONTROL_TX_TRIG_SOL	(0 << 12)
+#define DSI_HOST_CONTROL_TX_TRIG_FIFO	(1 << 12)
+#define DSI_HOST_CONTROL_TX_TRIG_HOST	(2 << 12)
 #define DSI_HOST_CONTROL_RAW		(1 << 6)
 #define DSI_HOST_CONTROL_HS		(1 << 5)
-#define DSI_HOST_CONTROL_BTA		(1 << 2)
+#define DSI_HOST_CONTROL_FIFO_SEL	(1 << 4)
+#define DSI_HOST_CONTROL_IMM_BTA	(1 << 3)
+#define DSI_HOST_CONTROL_PKT_BTA	(1 << 2)
 #define DSI_HOST_CONTROL_CS		(1 << 1)
 #define DSI_HOST_CONTROL_ECC		(1 << 0)
 #define DSI_CONTROL			0x10
@@ -39,9 +46,13 @@
 #define DSI_SOL_DELAY			0x11
 #define DSI_MAX_THRESHOLD		0x12
 #define DSI_TRIGGER			0x13
+#define DSI_TRIGGER_HOST		(1 << 1)
+#define DSI_TRIGGER_VIDEO		(1 << 0)
 #define DSI_TX_CRC			0x14
 #define DSI_STATUS			0x15
 #define DSI_STATUS_IDLE			(1 << 10)
+#define DSI_STATUS_UNDERFLOW		(1 <<  9)
+#define DSI_STATUS_OVERFLOW		(1 <<  8)
 #define DSI_INIT_SEQ_CONTROL		0x1a
 #define DSI_INIT_SEQ_DATA_0		0x1b
 #define DSI_INIT_SEQ_DATA_1		0x1c
@@ -104,6 +115,7 @@
 #define DSI_PAD_CONTROL_3		0x51
 #define DSI_PAD_CONTROL_4		0x52
 #define DSI_GANGED_MODE_CONTROL		0x53
+#define DSI_GANGED_MODE_CONTROL_ENABLE	(1 << 0)
 #define DSI_GANGED_MODE_START		0x54
 #define DSI_GANGED_MODE_SIZE		0x55
 #define DSI_RAW_DATA_BYTE_COUNT		0x56
diff --git a/drivers/gpu/drm/tegra/fb.c b/drivers/gpu/drm/tegra/fb.c
index 3513d12d5aa1..e9c715d89261 100644
--- a/drivers/gpu/drm/tegra/fb.c
+++ b/drivers/gpu/drm/tegra/fb.c
@@ -65,8 +65,12 @@ static void tegra_fb_destroy(struct drm_framebuffer *framebuffer)
 	for (i = 0; i < fb->num_planes; i++) {
 		struct tegra_bo *bo = fb->planes[i];
 
-		if (bo)
+		if (bo) {
+			if (bo->pages && bo->vaddr)
+				vunmap(bo->vaddr);
+
 			drm_gem_object_unreference_unlocked(&bo->gem);
+		}
 	}
 
 	drm_framebuffer_cleanup(framebuffer);
@@ -223,14 +227,16 @@ static int tegra_fbdev_probe(struct drm_fb_helper *helper,
 	info = framebuffer_alloc(0, drm->dev);
 	if (!info) {
 		dev_err(drm->dev, "failed to allocate framebuffer info\n");
-		tegra_bo_free_object(&bo->gem);
+		drm_gem_object_unreference_unlocked(&bo->gem);
 		return -ENOMEM;
 	}
 
 	fbdev->fb = tegra_fb_alloc(drm, &cmd, &bo, 1);
 	if (IS_ERR(fbdev->fb)) {
-		dev_err(drm->dev, "failed to allocate DRM framebuffer\n");
 		err = PTR_ERR(fbdev->fb);
+		dev_err(drm->dev, "failed to allocate DRM framebuffer: %d\n",
+			err);
+		drm_gem_object_unreference_unlocked(&bo->gem);
 		goto release;
 	}
 
@@ -254,6 +260,16 @@ static int tegra_fbdev_probe(struct drm_fb_helper *helper,
 	offset = info->var.xoffset * bytes_per_pixel +
 		 info->var.yoffset * fb->pitches[0];
 
+	if (bo->pages) {
+		bo->vaddr = vmap(bo->pages, bo->num_pages, VM_MAP,
+				 pgprot_writecombine(PAGE_KERNEL));
+		if (!bo->vaddr) {
+			dev_err(drm->dev, "failed to vmap() framebuffer\n");
+			err = -ENOMEM;
+			goto destroy;
+		}
+	}
+
 	drm->mode_config.fb_base = (resource_size_t)bo->paddr;
 	info->screen_base = (void __iomem *)bo->vaddr + offset;
 	info->screen_size = size;
@@ -289,6 +305,11 @@ static struct tegra_fbdev *tegra_fbdev_create(struct drm_device *drm)
 	return fbdev;
 }
 
+static void tegra_fbdev_free(struct tegra_fbdev *fbdev)
+{
+	kfree(fbdev);
+}
+
 static int tegra_fbdev_init(struct tegra_fbdev *fbdev,
 			    unsigned int preferred_bpp,
 			    unsigned int num_crtc,
@@ -299,19 +320,21 @@ static int tegra_fbdev_init(struct tegra_fbdev *fbdev,
 
 	err = drm_fb_helper_init(drm, &fbdev->base, num_crtc, max_connectors);
 	if (err < 0) {
-		dev_err(drm->dev, "failed to initialize DRM FB helper\n");
+		dev_err(drm->dev, "failed to initialize DRM FB helper: %d\n",
+			err);
 		return err;
 	}
 
 	err = drm_fb_helper_single_add_all_connectors(&fbdev->base);
 	if (err < 0) {
-		dev_err(drm->dev, "failed to add connectors\n");
+		dev_err(drm->dev, "failed to add connectors: %d\n", err);
 		goto fini;
 	}
 
 	err = drm_fb_helper_initial_config(&fbdev->base, preferred_bpp);
 	if (err < 0) {
-		dev_err(drm->dev, "failed to set initial configuration\n");
+		dev_err(drm->dev, "failed to set initial configuration: %d\n",
+			err);
 		goto fini;
 	}
 
@@ -322,7 +345,7 @@ fini:
 	return err;
 }
 
-static void tegra_fbdev_free(struct tegra_fbdev *fbdev)
+static void tegra_fbdev_exit(struct tegra_fbdev *fbdev)
 {
 	struct fb_info *info = fbdev->base.fbdev;
 
@@ -341,11 +364,11 @@ static void tegra_fbdev_free(struct tegra_fbdev *fbdev)
 
 	if (fbdev->fb) {
 		drm_framebuffer_unregister_private(&fbdev->fb->base);
-		tegra_fb_destroy(&fbdev->fb->base);
+		drm_framebuffer_remove(&fbdev->fb->base);
 	}
 
 	drm_fb_helper_fini(&fbdev->base);
-	kfree(fbdev);
+	tegra_fbdev_free(fbdev);
 }
 
 void tegra_fbdev_restore_mode(struct tegra_fbdev *fbdev)
@@ -393,6 +416,15 @@ int tegra_drm_fb_prepare(struct drm_device *drm)
 	return 0;
 }
 
+void tegra_drm_fb_free(struct drm_device *drm)
+{
+#ifdef CONFIG_DRM_TEGRA_FBDEV
+	struct tegra_drm *tegra = drm->dev_private;
+
+	tegra_fbdev_free(tegra->fbdev);
+#endif
+}
+
 int tegra_drm_fb_init(struct drm_device *drm)
 {
 #ifdef CONFIG_DRM_TEGRA_FBDEV
@@ -413,6 +445,6 @@ void tegra_drm_fb_exit(struct drm_device *drm)
 #ifdef CONFIG_DRM_TEGRA_FBDEV
 	struct tegra_drm *tegra = drm->dev_private;
 
-	tegra_fbdev_free(tegra->fbdev);
+	tegra_fbdev_exit(tegra->fbdev);
 #endif
 }
diff --git a/drivers/gpu/drm/tegra/gem.c b/drivers/gpu/drm/tegra/gem.c
index ce023fa3e8ae..8777b7f75791 100644
--- a/drivers/gpu/drm/tegra/gem.c
+++ b/drivers/gpu/drm/tegra/gem.c
@@ -14,6 +14,7 @@
  */
 
 #include <linux/dma-buf.h>
+#include <linux/iommu.h>
 #include <drm/tegra_drm.h>
 
 #include "drm.h"
@@ -91,13 +92,90 @@ static const struct host1x_bo_ops tegra_bo_ops = {
 	.kunmap = tegra_bo_kunmap,
 };
 
-static void tegra_bo_destroy(struct drm_device *drm, struct tegra_bo *bo)
+/*
+ * A generic iommu_map_sg() function is being reviewed and will hopefully be
+ * merged soon. At that point this function can be dropped in favour of the
+ * one provided by the IOMMU API.
+ */
+static ssize_t __iommu_map_sg(struct iommu_domain *domain, unsigned long iova,
+			      struct scatterlist *sg, unsigned int nents,
+			      int prot)
 {
-	dma_free_writecombine(drm->dev, bo->gem.size, bo->vaddr, bo->paddr);
+	struct scatterlist *s;
+	size_t offset = 0;
+	unsigned int i;
+	int err;
+
+	for_each_sg(sg, s, nents, i) {
+		phys_addr_t phys = page_to_phys(sg_page(s));
+		size_t length = s->offset + s->length;
+
+		err = iommu_map(domain, iova + offset, phys, length, prot);
+		if (err < 0) {
+			iommu_unmap(domain, iova, offset);
+			return err;
+		}
+
+		offset += length;
+	}
+
+	return offset;
 }
 
-struct tegra_bo *tegra_bo_create(struct drm_device *drm, unsigned int size,
-				 unsigned long flags)
+static int tegra_bo_iommu_map(struct tegra_drm *tegra, struct tegra_bo *bo)
+{
+	int prot = IOMMU_READ | IOMMU_WRITE;
+	ssize_t err;
+
+	if (bo->mm)
+		return -EBUSY;
+
+	bo->mm = kzalloc(sizeof(*bo->mm), GFP_KERNEL);
+	if (!bo->mm)
+		return -ENOMEM;
+
+	err = drm_mm_insert_node_generic(&tegra->mm, bo->mm, bo->gem.size,
+					 PAGE_SIZE, 0, 0, 0);
+	if (err < 0) {
+		dev_err(tegra->drm->dev, "out of I/O virtual memory: %zd\n",
+			err);
+		goto free;
+	}
+
+	bo->paddr = bo->mm->start;
+
+	err = __iommu_map_sg(tegra->domain, bo->paddr, bo->sgt->sgl,
+			     bo->sgt->nents, prot);
+	if (err < 0) {
+		dev_err(tegra->drm->dev, "failed to map buffer: %zd\n", err);
+		goto remove;
+	}
+
+	bo->size = err;
+
+	return 0;
+
+remove:
+	drm_mm_remove_node(bo->mm);
+free:
+	kfree(bo->mm);
+	return err;
+}
+
+static int tegra_bo_iommu_unmap(struct tegra_drm *tegra, struct tegra_bo *bo)
+{
+	if (!bo->mm)
+		return 0;
+
+	iommu_unmap(tegra->domain, bo->paddr, bo->size);
+	drm_mm_remove_node(bo->mm);
+	kfree(bo->mm);
+
+	return 0;
+}
+
+static struct tegra_bo *tegra_bo_alloc_object(struct drm_device *drm,
+					      size_t size)
 {
 	struct tegra_bo *bo;
 	int err;
@@ -109,22 +187,124 @@ struct tegra_bo *tegra_bo_create(struct drm_device *drm, unsigned int size,
 	host1x_bo_init(&bo->base, &tegra_bo_ops);
 	size = round_up(size, PAGE_SIZE);
 
-	bo->vaddr = dma_alloc_writecombine(drm->dev, size, &bo->paddr,
-					   GFP_KERNEL | __GFP_NOWARN);
-	if (!bo->vaddr) {
-		dev_err(drm->dev, "failed to allocate buffer with size %u\n",
-			size);
-		err = -ENOMEM;
-		goto err_dma;
-	}
-
 	err = drm_gem_object_init(drm, &bo->gem, size);
-	if (err)
-		goto err_init;
+	if (err < 0)
+		goto free;
 
 	err = drm_gem_create_mmap_offset(&bo->gem);
-	if (err)
-		goto err_mmap;
+	if (err < 0)
+		goto release;
+
+	return bo;
+
+release:
+	drm_gem_object_release(&bo->gem);
+free:
+	kfree(bo);
+	return ERR_PTR(err);
+}
+
+static void tegra_bo_free(struct drm_device *drm, struct tegra_bo *bo)
+{
+	if (bo->pages) {
+		drm_gem_put_pages(&bo->gem, bo->pages, true, true);
+		sg_free_table(bo->sgt);
+		kfree(bo->sgt);
+	} else if (bo->vaddr) {
+		dma_free_writecombine(drm->dev, bo->gem.size, bo->vaddr,
+				      bo->paddr);
+	}
+}
+
+static int tegra_bo_get_pages(struct drm_device *drm, struct tegra_bo *bo)
+{
+	struct scatterlist *s;
+	struct sg_table *sgt;
+	unsigned int i;
+
+	bo->pages = drm_gem_get_pages(&bo->gem);
+	if (IS_ERR(bo->pages))
+		return PTR_ERR(bo->pages);
+
+	bo->num_pages = bo->gem.size >> PAGE_SHIFT;
+
+	sgt = drm_prime_pages_to_sg(bo->pages, bo->num_pages);
+	if (IS_ERR(sgt))
+		goto put_pages;
+
+	/*
+	 * Fake up the SG table so that dma_map_sg() can be used to flush the
+	 * pages associated with it. Note that this relies on the fact that
+	 * the DMA API doesn't hook into IOMMU on Tegra, therefore mapping is
+	 * only cache maintenance.
+	 *
+	 * TODO: Replace this by drm_clflash_sg() once it can be implemented
+	 * without relying on symbols that are not exported.
+	 */
+	for_each_sg(sgt->sgl, s, sgt->nents, i)
+		sg_dma_address(s) = sg_phys(s);
+
+	if (dma_map_sg(drm->dev, sgt->sgl, sgt->nents, DMA_TO_DEVICE) == 0) {
+		sgt = ERR_PTR(-ENOMEM);
+		goto release_sgt;
+	}
+
+	bo->sgt = sgt;
+
+	return 0;
+
+release_sgt:
+	sg_free_table(sgt);
+	kfree(sgt);
+put_pages:
+	drm_gem_put_pages(&bo->gem, bo->pages, false, false);
+	return PTR_ERR(sgt);
+}
+
+static int tegra_bo_alloc(struct drm_device *drm, struct tegra_bo *bo)
+{
+	struct tegra_drm *tegra = drm->dev_private;
+	int err;
+
+	if (tegra->domain) {
+		err = tegra_bo_get_pages(drm, bo);
+		if (err < 0)
+			return err;
+
+		err = tegra_bo_iommu_map(tegra, bo);
+		if (err < 0) {
+			tegra_bo_free(drm, bo);
+			return err;
+		}
+	} else {
+		size_t size = bo->gem.size;
+
+		bo->vaddr = dma_alloc_writecombine(drm->dev, size, &bo->paddr,
+						   GFP_KERNEL | __GFP_NOWARN);
+		if (!bo->vaddr) {
+			dev_err(drm->dev,
+				"failed to allocate buffer of size %zu\n",
+				size);
+			return -ENOMEM;
+		}
+	}
+
+	return 0;
+}
+
+struct tegra_bo *tegra_bo_create(struct drm_device *drm, size_t size,
+				 unsigned long flags)
+{
+	struct tegra_bo *bo;
+	int err;
+
+	bo = tegra_bo_alloc_object(drm, size);
+	if (IS_ERR(bo))
+		return bo;
+
+	err = tegra_bo_alloc(drm, bo);
+	if (err < 0)
+		goto release;
 
 	if (flags & DRM_TEGRA_GEM_CREATE_TILED)
 		bo->tiling.mode = TEGRA_BO_TILING_MODE_TILED;
@@ -134,69 +314,52 @@ struct tegra_bo *tegra_bo_create(struct drm_device *drm, unsigned int size,
 
 	return bo;
 
-err_mmap:
+release:
 	drm_gem_object_release(&bo->gem);
-err_init:
-	tegra_bo_destroy(drm, bo);
-err_dma:
 	kfree(bo);
-
 	return ERR_PTR(err);
 }
 
 struct tegra_bo *tegra_bo_create_with_handle(struct drm_file *file,
 					     struct drm_device *drm,
-					     unsigned int size,
+					     size_t size,
 					     unsigned long flags,
-					     unsigned int *handle)
+					     u32 *handle)
 {
 	struct tegra_bo *bo;
-	int ret;
+	int err;
 
 	bo = tegra_bo_create(drm, size, flags);
 	if (IS_ERR(bo))
 		return bo;
 
-	ret = drm_gem_handle_create(file, &bo->gem, handle);
-	if (ret)
-		goto err;
+	err = drm_gem_handle_create(file, &bo->gem, handle);
+	if (err) {
+		tegra_bo_free_object(&bo->gem);
+		return ERR_PTR(err);
+	}
 
 	drm_gem_object_unreference_unlocked(&bo->gem);
 
 	return bo;
-
-err:
-	tegra_bo_free_object(&bo->gem);
-	return ERR_PTR(ret);
 }
 
 static struct tegra_bo *tegra_bo_import(struct drm_device *drm,
 					struct dma_buf *buf)
 {
+	struct tegra_drm *tegra = drm->dev_private;
 	struct dma_buf_attachment *attach;
 	struct tegra_bo *bo;
-	ssize_t size;
 	int err;
 
-	bo = kzalloc(sizeof(*bo), GFP_KERNEL);
-	if (!bo)
-		return ERR_PTR(-ENOMEM);
-
-	host1x_bo_init(&bo->base, &tegra_bo_ops);
-	size = round_up(buf->size, PAGE_SIZE);
-
-	err = drm_gem_object_init(drm, &bo->gem, size);
-	if (err < 0)
-		goto free;
-
-	err = drm_gem_create_mmap_offset(&bo->gem);
-	if (err < 0)
-		goto release;
+	bo = tegra_bo_alloc_object(drm, buf->size);
+	if (IS_ERR(bo))
+		return bo;
 
 	attach = dma_buf_attach(buf, drm->dev);
 	if (IS_ERR(attach)) {
 		err = PTR_ERR(attach);
-		goto free_mmap;
+		goto free;
 	}
 
 	get_dma_buf(buf);
@@ -212,12 +375,19 @@ static struct tegra_bo *tegra_bo_import(struct drm_device *drm,
 		goto detach;
 	}
 
-	if (bo->sgt->nents > 1) {
-		err = -EINVAL;
-		goto detach;
+	if (tegra->domain) {
+		err = tegra_bo_iommu_map(tegra, bo);
+		if (err < 0)
+			goto detach;
+	} else {
+		if (bo->sgt->nents > 1) {
+			err = -EINVAL;
+			goto detach;
+		}
+
+		bo->paddr = sg_dma_address(bo->sgt->sgl);
 	}
 
-	bo->paddr = sg_dma_address(bo->sgt->sgl);
 	bo->gem.import_attach = attach;
 
 	return bo;
@@ -228,47 +398,41 @@ detach:
 
 	dma_buf_detach(buf, attach);
 	dma_buf_put(buf);
-free_mmap:
-	drm_gem_free_mmap_offset(&bo->gem);
-release:
-	drm_gem_object_release(&bo->gem);
 free:
+	drm_gem_object_release(&bo->gem);
 	kfree(bo);
-
 	return ERR_PTR(err);
 }
 
 void tegra_bo_free_object(struct drm_gem_object *gem)
 {
+	struct tegra_drm *tegra = gem->dev->dev_private;
 	struct tegra_bo *bo = to_tegra_bo(gem);
 
+	if (tegra->domain)
+		tegra_bo_iommu_unmap(tegra, bo);
+
 	if (gem->import_attach) {
 		dma_buf_unmap_attachment(gem->import_attach, bo->sgt,
 					 DMA_TO_DEVICE);
 		drm_prime_gem_destroy(gem, NULL);
 	} else {
-		tegra_bo_destroy(gem->dev, bo);
+		tegra_bo_free(gem->dev, bo);
 	}
 
-	drm_gem_free_mmap_offset(gem);
 	drm_gem_object_release(gem);
-
 	kfree(bo);
 }
 
 int tegra_bo_dumb_create(struct drm_file *file, struct drm_device *drm,
 			 struct drm_mode_create_dumb *args)
 {
-	int min_pitch = DIV_ROUND_UP(args->width * args->bpp, 8);
+	unsigned int min_pitch = DIV_ROUND_UP(args->width * args->bpp, 8);
 	struct tegra_drm *tegra = drm->dev_private;
 	struct tegra_bo *bo;
 
-	min_pitch = round_up(min_pitch, tegra->pitch_align);
-	if (args->pitch < min_pitch)
-		args->pitch = min_pitch;
-
-	if (args->size < args->pitch * args->height)
-		args->size = args->pitch * args->height;
+	args->pitch = round_up(min_pitch, tegra->pitch_align);
+	args->size = args->pitch * args->height;
 
 	bo = tegra_bo_create_with_handle(file, drm, args->size, 0,
 					 &args->handle);
@@ -279,7 +443,7 @@ int tegra_bo_dumb_create(struct drm_file *file, struct drm_device *drm,
 }
 
 int tegra_bo_dumb_map_offset(struct drm_file *file, struct drm_device *drm,
-			     uint32_t handle, uint64_t *offset)
+			     u32 handle, u64 *offset)
 {
 	struct drm_gem_object *gem;
 	struct tegra_bo *bo;
@@ -304,7 +468,38 @@ int tegra_bo_dumb_map_offset(struct drm_file *file, struct drm_device *drm,
 	return 0;
 }
 
+static int tegra_bo_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
+{
+	struct drm_gem_object *gem = vma->vm_private_data;
+	struct tegra_bo *bo = to_tegra_bo(gem);
+	struct page *page;
+	pgoff_t offset;
+	int err;
+
+	if (!bo->pages)
+		return VM_FAULT_SIGBUS;
+
+	offset = ((unsigned long)vmf->virtual_address - vma->vm_start) >> PAGE_SHIFT;
+	page = bo->pages[offset];
+
+	err = vm_insert_page(vma, (unsigned long)vmf->virtual_address, page);
+	switch (err) {
+	case -EAGAIN:
+	case 0:
+	case -ERESTARTSYS:
+	case -EINTR:
+	case -EBUSY:
+		return VM_FAULT_NOPAGE;
+
+	case -ENOMEM:
+		return VM_FAULT_OOM;
+	}
+
+	return VM_FAULT_SIGBUS;
+}
+
 const struct vm_operations_struct tegra_bo_vm_ops = {
+	.fault = tegra_bo_fault,
 	.open = drm_gem_vm_open,
 	.close = drm_gem_vm_close,
 };
@@ -322,12 +517,30 @@ int tegra_drm_mmap(struct file *file, struct vm_area_struct *vma)
 	gem = vma->vm_private_data;
 	bo = to_tegra_bo(gem);
 
-	ret = remap_pfn_range(vma, vma->vm_start, bo->paddr >> PAGE_SHIFT,
-			      vma->vm_end - vma->vm_start, vma->vm_page_prot);
-	if (ret)
-		drm_gem_vm_close(vma);
+	if (!bo->pages) {
+		unsigned long vm_pgoff = vma->vm_pgoff;
+
+		vma->vm_flags &= ~VM_PFNMAP;
+		vma->vm_pgoff = 0;
+
+		ret = dma_mmap_writecombine(gem->dev->dev, vma, bo->vaddr,
+					    bo->paddr, gem->size);
+		if (ret) {
+			drm_gem_vm_close(vma);
+			return ret;
+		}
+
+		vma->vm_pgoff = vm_pgoff;
+	} else {
+		pgprot_t prot = vm_get_page_prot(vma->vm_flags);
+
+		vma->vm_flags |= VM_MIXEDMAP;
+		vma->vm_flags &= ~VM_PFNMAP;
 
-	return ret;
+		vma->vm_page_prot = pgprot_writecombine(prot);
+	}
+
+	return 0;
 }
 
 static struct sg_table *
@@ -342,21 +555,44 @@ tegra_gem_prime_map_dma_buf(struct dma_buf_attachment *attach,
 	if (!sgt)
 		return NULL;
 
-	if (sg_alloc_table(sgt, 1, GFP_KERNEL)) {
-		kfree(sgt);
-		return NULL;
-	}
+	if (bo->pages) {
+		struct scatterlist *sg;
+		unsigned int i;
 
-	sg_dma_address(sgt->sgl) = bo->paddr;
-	sg_dma_len(sgt->sgl) = gem->size;
+		if (sg_alloc_table(sgt, bo->num_pages, GFP_KERNEL))
+			goto free;
+
+		for_each_sg(sgt->sgl, sg, bo->num_pages, i)
+			sg_set_page(sg, bo->pages[i], PAGE_SIZE, 0);
+
+		if (dma_map_sg(attach->dev, sgt->sgl, sgt->nents, dir) == 0)
+			goto free;
+	} else {
+		if (sg_alloc_table(sgt, 1, GFP_KERNEL))
+			goto free;
+
+		sg_dma_address(sgt->sgl) = bo->paddr;
+		sg_dma_len(sgt->sgl) = gem->size;
+	}
 
 	return sgt;
+
+free:
+	sg_free_table(sgt);
+	kfree(sgt);
+	return NULL;
 }
 
 static void tegra_gem_prime_unmap_dma_buf(struct dma_buf_attachment *attach,
 					  struct sg_table *sgt,
 					  enum dma_data_direction dir)
 {
+	struct drm_gem_object *gem = attach->dmabuf->priv;
+	struct tegra_bo *bo = to_tegra_bo(gem);
+
+	if (bo->pages)
+		dma_unmap_sg(attach->dev, sgt->sgl, sgt->nents, dir);
+
 	sg_free_table(sgt);
 	kfree(sgt);
 }
diff --git a/drivers/gpu/drm/tegra/gem.h b/drivers/gpu/drm/tegra/gem.h
index 43a25c853357..6c5f12ac0087 100644
--- a/drivers/gpu/drm/tegra/gem.h
+++ b/drivers/gpu/drm/tegra/gem.h
@@ -15,6 +15,7 @@
 
 #include <drm/drm.h>
 #include <drm/drmP.h>
+#include <drm/drm_gem.h>
 
 #define TEGRA_BO_BOTTOM_UP (1 << 0)
 
@@ -37,6 +38,12 @@ struct tegra_bo {
 	dma_addr_t paddr;
 	void *vaddr;
 
+	struct drm_mm_node *mm;
+	unsigned long num_pages;
+	struct page **pages;
+	/* size of IOMMU mapping */
+	size_t size;
+
 	struct tegra_bo_tiling tiling;
 };
 
@@ -45,18 +52,18 @@ static inline struct tegra_bo *to_tegra_bo(struct drm_gem_object *gem)
 	return container_of(gem, struct tegra_bo, gem);
 }
 
-struct tegra_bo *tegra_bo_create(struct drm_device *drm, unsigned int size,
+struct tegra_bo *tegra_bo_create(struct drm_device *drm, size_t size,
 				 unsigned long flags);
 struct tegra_bo *tegra_bo_create_with_handle(struct drm_file *file,
 					     struct drm_device *drm,
-					     unsigned int size,
+					     size_t size,
 					     unsigned long flags,
-					     unsigned int *handle);
+					     u32 *handle);
 void tegra_bo_free_object(struct drm_gem_object *gem);
 int tegra_bo_dumb_create(struct drm_file *file, struct drm_device *drm,
 			 struct drm_mode_create_dumb *args);
 int tegra_bo_dumb_map_offset(struct drm_file *file, struct drm_device *drm,
-			     uint32_t handle, uint64_t *offset);
+			     u32 handle, u64 *offset);
 
 int tegra_drm_mmap(struct file *file, struct vm_area_struct *vma);
 
diff --git a/drivers/gpu/drm/tegra/output.c b/drivers/gpu/drm/tegra/output.c
index 0c67d7eebc94..6a5c7b81fbc5 100644
--- a/drivers/gpu/drm/tegra/output.c
+++ b/drivers/gpu/drm/tegra/output.c
@@ -157,22 +157,18 @@ static bool tegra_encoder_mode_fixup(struct drm_encoder *encoder,
 
 static void tegra_encoder_prepare(struct drm_encoder *encoder)
 {
+	tegra_encoder_dpms(encoder, DRM_MODE_DPMS_OFF);
 }
 
 static void tegra_encoder_commit(struct drm_encoder *encoder)
 {
+	tegra_encoder_dpms(encoder, DRM_MODE_DPMS_ON);
 }
 
 static void tegra_encoder_mode_set(struct drm_encoder *encoder,
 				   struct drm_display_mode *mode,
 				   struct drm_display_mode *adjusted)
 {
-	struct tegra_output *output = encoder_to_output(encoder);
-	int err;
-
-	err = tegra_output_enable(output);
-	if (err < 0)
-		dev_err(encoder->dev->dev, "tegra_output_enable(): %d\n", err);
 }
 
 static const struct drm_encoder_helper_funcs encoder_helper_funcs = {
@@ -187,7 +183,8 @@ static irqreturn_t hpd_irq(int irq, void *data)
 {
 	struct tegra_output *output = data;
 
-	drm_helper_hpd_irq_event(output->connector.dev);
+	if (output->connector.dev)
+		drm_helper_hpd_irq_event(output->connector.dev);
 
 	return IRQ_HANDLED;
 }
@@ -259,6 +256,13 @@ int tegra_output_probe(struct tegra_output *output)
 		}
 
 		output->connector.polled = DRM_CONNECTOR_POLL_HPD;
+
+		/*
+		 * Disable the interrupt until the connector has been
+		 * initialized to avoid a race in the hotplug interrupt
+		 * handler.
+		 */
+		disable_irq(output->hpd_irq);
 	}
 
 	return 0;
@@ -324,10 +328,27 @@ int tegra_output_init(struct drm_device *drm, struct tegra_output *output)
 
 	output->encoder.possible_crtcs = 0x3;
 
+	/*
+	 * The connector is now registered and ready to receive hotplug events
+	 * so the hotplug interrupt can be enabled.
+	 */
+	if (gpio_is_valid(output->hpd_gpio))
+		enable_irq(output->hpd_irq);
+
 	return 0;
 }
 
 int tegra_output_exit(struct tegra_output *output)
 {
+	/*
+	 * The connector is going away, so the interrupt must be disabled to
+	 * prevent the hotplug interrupt handler from potentially crashing.
+	 */
+	if (gpio_is_valid(output->hpd_gpio))
+		disable_irq(output->hpd_irq);
+
+	if (output->panel)
+		drm_panel_detach(output->panel);
+
 	return 0;
 }
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_crtc.c b/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
index d642d4a02134..c73588483be0 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
@@ -16,6 +16,7 @@
  */
 
 #include "drm_flip_work.h"
+#include <drm/drm_plane_helper.h>
 
 #include "tilcdc_drv.h"
 #include "tilcdc_regs.h"
@@ -664,12 +665,8 @@ struct drm_crtc *tilcdc_crtc_create(struct drm_device *dev)
 	tilcdc_crtc->dpms = DRM_MODE_DPMS_OFF;
 	init_waitqueue_head(&tilcdc_crtc->frame_done_wq);
 
-	ret = drm_flip_work_init(&tilcdc_crtc->unref_work, 16,
+	drm_flip_work_init(&tilcdc_crtc->unref_work,
 			"unref", unref_worker);
-	if (ret) {
-		dev_err(dev->dev, "could not allocate unref FIFO\n");
-		goto fail;
-	}
 
 	ret = drm_crtc_init(dev, crtc, &tilcdc_crtc_funcs);
 	if (ret < 0)
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_drv.c b/drivers/gpu/drm/tilcdc/tilcdc_drv.c
index 6be623b4a86f..095fca91525c 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_drv.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_drv.c
@@ -58,8 +58,7 @@ static struct drm_framebuffer *tilcdc_fb_create(struct drm_device *dev,
 static void tilcdc_fb_output_poll_changed(struct drm_device *dev)
 {
 	struct tilcdc_drm_private *priv = dev->dev_private;
-	if (priv->fbdev)
-		drm_fbdev_cma_hotplug_event(priv->fbdev);
+	drm_fbdev_cma_hotplug_event(priv->fbdev);
 }
 
 static const struct drm_mode_config_funcs mode_config_funcs = {
@@ -84,6 +83,7 @@ static int modeset_init(struct drm_device *dev)
 	if ((priv->num_encoders == 0) || (priv->num_connectors == 0)) {
 		/* oh nos! */
 		dev_err(dev->dev, "no encoders/connectors found\n");
+		drm_mode_config_cleanup(dev);
 		return -ENXIO;
 	}
 
@@ -172,33 +172,37 @@ static int tilcdc_load(struct drm_device *dev, unsigned long flags)
 	dev->dev_private = priv;
 
 	priv->wq = alloc_ordered_workqueue("tilcdc", 0);
+	if (!priv->wq) {
+		ret = -ENOMEM;
+		goto fail_free_priv;
+	}
 
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	if (!res) {
 		dev_err(dev->dev, "failed to get memory resource\n");
 		ret = -EINVAL;
-		goto fail;
+		goto fail_free_wq;
 	}
 
 	priv->mmio = ioremap_nocache(res->start, resource_size(res));
 	if (!priv->mmio) {
 		dev_err(dev->dev, "failed to ioremap\n");
 		ret = -ENOMEM;
-		goto fail;
+		goto fail_free_wq;
 	}
 
 	priv->clk = clk_get(dev->dev, "fck");
 	if (IS_ERR(priv->clk)) {
 		dev_err(dev->dev, "failed to get functional clock\n");
 		ret = -ENODEV;
-		goto fail;
+		goto fail_iounmap;
 	}
 
 	priv->disp_clk = clk_get(dev->dev, "dpll_disp_ck");
 	if (IS_ERR(priv->clk)) {
 		dev_err(dev->dev, "failed to get display clock\n");
 		ret = -ENODEV;
-		goto fail;
+		goto fail_put_clk;
 	}
 
 #ifdef CONFIG_CPU_FREQ
@@ -208,7 +212,7 @@ static int tilcdc_load(struct drm_device *dev, unsigned long flags)
 			CPUFREQ_TRANSITION_NOTIFIER);
 	if (ret) {
 		dev_err(dev->dev, "failed to register cpufreq notifier\n");
-		goto fail;
+		goto fail_put_disp_clk;
 	}
 #endif
 
@@ -253,13 +257,13 @@ static int tilcdc_load(struct drm_device *dev, unsigned long flags)
 	ret = modeset_init(dev);
 	if (ret < 0) {
 		dev_err(dev->dev, "failed to initialize mode setting\n");
-		goto fail;
+		goto fail_cpufreq_unregister;
 	}
 
 	ret = drm_vblank_init(dev, 1);
 	if (ret < 0) {
 		dev_err(dev->dev, "failed to initialize vblank\n");
-		goto fail;
+		goto fail_mode_config_cleanup;
 	}
 
 	pm_runtime_get_sync(dev->dev);
@@ -267,7 +271,7 @@ static int tilcdc_load(struct drm_device *dev, unsigned long flags)
 	pm_runtime_put_sync(dev->dev);
 	if (ret < 0) {
 		dev_err(dev->dev, "failed to install IRQ handler\n");
-		goto fail;
+		goto fail_vblank_cleanup;
 	}
 
 	platform_set_drvdata(pdev, dev);
@@ -283,13 +287,48 @@ static int tilcdc_load(struct drm_device *dev, unsigned long flags)
 	priv->fbdev = drm_fbdev_cma_init(dev, bpp,
 			dev->mode_config.num_crtc,
 			dev->mode_config.num_connector);
+	if (IS_ERR(priv->fbdev)) {
+		ret = PTR_ERR(priv->fbdev);
+		goto fail_irq_uninstall;
+	}
 
 	drm_kms_helper_poll_init(dev);
 
 	return 0;
 
-fail:
-	tilcdc_unload(dev);
+fail_irq_uninstall:
+	pm_runtime_get_sync(dev->dev);
+	drm_irq_uninstall(dev);
+	pm_runtime_put_sync(dev->dev);
+
+fail_vblank_cleanup:
+	drm_vblank_cleanup(dev);
+
+fail_mode_config_cleanup:
+	drm_mode_config_cleanup(dev);
+
+fail_cpufreq_unregister:
+	pm_runtime_disable(dev->dev);
+#ifdef CONFIG_CPU_FREQ
+	cpufreq_unregister_notifier(&priv->freq_transition,
+			CPUFREQ_TRANSITION_NOTIFIER);
+fail_put_disp_clk:
+	clk_put(priv->disp_clk);
+#endif
+
+fail_put_clk:
+	clk_put(priv->clk);
+
+fail_iounmap:
+	iounmap(priv->mmio);
+
+fail_free_wq:
+	flush_workqueue(priv->wq);
+	destroy_workqueue(priv->wq);
+
+fail_free_priv:
+	dev->dev_private = NULL;
+	kfree(priv);
 	return ret;
 }
 
@@ -502,6 +541,7 @@ static struct drm_driver tilcdc_driver = {
 	.unload             = tilcdc_unload,
 	.preclose           = tilcdc_preclose,
 	.lastclose          = tilcdc_lastclose,
+	.set_busid          = drm_platform_set_busid,
 	.irq_handler        = tilcdc_irq,
 	.irq_preinstall     = tilcdc_irq_preinstall,
 	.irq_postinstall    = tilcdc_irq_postinstall,
@@ -604,7 +644,6 @@ static struct platform_driver tilcdc_platform_driver = {
 	.probe      = tilcdc_pdev_probe,
 	.remove     = tilcdc_pdev_remove,
 	.driver     = {
-		.owner  = THIS_MODULE,
 		.name   = "tilcdc",
 		.pm     = &tilcdc_pm_ops,
 		.of_match_table = tilcdc_of_match,
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_panel.c b/drivers/gpu/drm/tilcdc/tilcdc_panel.c
index 4c7aa1d8134f..7a0315855e90 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_panel.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_panel.c
@@ -18,6 +18,7 @@
 #include <linux/pinctrl/pinmux.h>
 #include <linux/pinctrl/consumer.h>
 #include <linux/backlight.h>
+#include <linux/gpio/consumer.h>
 #include <video/display_timing.h>
 #include <video/of_display_timing.h>
 #include <video/videomode.h>
@@ -29,6 +30,7 @@ struct panel_module {
 	struct tilcdc_panel_info *info;
 	struct display_timings *timings;
 	struct backlight_device *backlight;
+	struct gpio_desc *enable_gpio;
 };
 #define to_panel_module(x) container_of(x, struct panel_module, base)
 
@@ -55,13 +57,17 @@ static void panel_encoder_dpms(struct drm_encoder *encoder, int mode)
 {
 	struct panel_encoder *panel_encoder = to_panel_encoder(encoder);
 	struct backlight_device *backlight = panel_encoder->mod->backlight;
+	struct gpio_desc *gpio = panel_encoder->mod->enable_gpio;
 
-	if (!backlight)
-		return;
+	if (backlight) {
+		backlight->props.power = mode == DRM_MODE_DPMS_ON ?
+					 FB_BLANK_UNBLANK : FB_BLANK_POWERDOWN;
+		backlight_update_status(backlight);
+	}
 
-	backlight->props.power = mode == DRM_MODE_DPMS_ON
-				     ? FB_BLANK_UNBLANK : FB_BLANK_POWERDOWN;
-	backlight_update_status(backlight);
+	if (gpio)
+		gpiod_set_value_cansleep(gpio,
+					 mode == DRM_MODE_DPMS_ON ? 1 : 0);
 }
 
 static bool panel_encoder_mode_fixup(struct drm_encoder *encoder,
@@ -311,6 +317,7 @@ static struct tilcdc_panel_info *of_get_panel_info(struct device_node *np)
 	info = kzalloc(sizeof(*info), GFP_KERNEL);
 	if (!info) {
 		pr_err("%s: allocation failed\n", __func__);
+		of_node_put(info_np);
 		return NULL;
 	}
 
@@ -331,22 +338,21 @@ static struct tilcdc_panel_info *of_get_panel_info(struct device_node *np)
 	if (ret) {
 		pr_err("%s: error reading panel-info properties\n", __func__);
 		kfree(info);
+		of_node_put(info_np);
 		return NULL;
 	}
+	of_node_put(info_np);
 
 	return info;
 }
 
-static struct of_device_id panel_of_match[];
-
 static int panel_probe(struct platform_device *pdev)
 {
-	struct device_node *node = pdev->dev.of_node;
+	struct device_node *bl_node, *node = pdev->dev.of_node;
 	struct panel_module *panel_mod;
 	struct tilcdc_module *mod;
 	struct pinctrl *pinctrl;
-	int ret = -EINVAL;
-
+	int ret;
 
 	/* bail out early if no DT data: */
 	if (!node) {
@@ -354,10 +360,40 @@ static int panel_probe(struct platform_device *pdev)
 		return -ENXIO;
 	}
 
-	panel_mod = kzalloc(sizeof(*panel_mod), GFP_KERNEL);
+	panel_mod = devm_kzalloc(&pdev->dev, sizeof(*panel_mod), GFP_KERNEL);
 	if (!panel_mod)
 		return -ENOMEM;
 
+	bl_node = of_parse_phandle(node, "backlight", 0);
+	if (bl_node) {
+		panel_mod->backlight = of_find_backlight_by_node(bl_node);
+		of_node_put(bl_node);
+
+		if (!panel_mod->backlight)
+			return -EPROBE_DEFER;
+
+		dev_info(&pdev->dev, "found backlight\n");
+	}
+
+	panel_mod->enable_gpio = devm_gpiod_get(&pdev->dev, "enable");
+	if (IS_ERR(panel_mod->enable_gpio)) {
+		ret = PTR_ERR(panel_mod->enable_gpio);
+		if (ret != -ENOENT) {
+			dev_err(&pdev->dev, "failed to request enable GPIO\n");
+			goto fail_backlight;
+		}
+
+		/* Optional GPIO is not here, continue silently. */
+		panel_mod->enable_gpio = NULL;
+	} else {
+		ret = gpiod_direction_output(panel_mod->enable_gpio, 0);
+		if (ret < 0) {
+			dev_err(&pdev->dev, "failed to setup GPIO\n");
+			goto fail_backlight;
+		}
+		dev_info(&pdev->dev, "found enable GPIO\n");
+	}
+
 	mod = &panel_mod->base;
 	pdev->dev.platform_data = mod;
 
@@ -370,29 +406,30 @@ static int panel_probe(struct platform_device *pdev)
 	panel_mod->timings = of_get_display_timings(node);
 	if (!panel_mod->timings) {
 		dev_err(&pdev->dev, "could not get panel timings\n");
+		ret = -EINVAL;
 		goto fail_free;
 	}
 
 	panel_mod->info = of_get_panel_info(node);
 	if (!panel_mod->info) {
 		dev_err(&pdev->dev, "could not get panel info\n");
+		ret = -EINVAL;
 		goto fail_timings;
 	}
 
 	mod->preferred_bpp = panel_mod->info->bpp;
 
-	panel_mod->backlight = of_find_backlight_by_node(node);
-	if (panel_mod->backlight)
-		dev_info(&pdev->dev, "found backlight\n");
-
 	return 0;
 
 fail_timings:
 	display_timings_release(panel_mod->timings);
 
 fail_free:
-	kfree(panel_mod);
 	tilcdc_module_cleanup(mod);
+
+fail_backlight:
+	if (panel_mod->backlight)
+		put_device(&panel_mod->backlight->dev);
 	return ret;
 }
 
@@ -400,12 +437,15 @@ static int panel_remove(struct platform_device *pdev)
 {
 	struct tilcdc_module *mod = dev_get_platdata(&pdev->dev);
 	struct panel_module *panel_mod = to_panel_module(mod);
+	struct backlight_device *backlight = panel_mod->backlight;
+
+	if (backlight)
+		put_device(&backlight->dev);
 
 	display_timings_release(panel_mod->timings);
 
 	tilcdc_module_cleanup(mod);
 	kfree(panel_mod->info);
-	kfree(panel_mod);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index 3da89d5dab60..d395b0bef73b 100644
--- a/drivers/gpu/drm/ttm/ttm_bo.c
+++ b/drivers/gpu/drm/ttm/ttm_bo.c
@@ -40,6 +40,7 @@
 #include <linux/file.h>
 #include <linux/module.h>
 #include <linux/atomic.h>
+#include <linux/reservation.h>
 
 #define TTM_ASSERT_LOCKED(param)
 #define TTM_DEBUG(fmt, arg...)
@@ -53,12 +54,13 @@ static struct attribute ttm_bo_count = {
 	.mode = S_IRUGO
 };
 
-static inline int ttm_mem_type_from_flags(uint32_t flags, uint32_t *mem_type)
+static inline int ttm_mem_type_from_place(const struct ttm_place *place,
+					  uint32_t *mem_type)
 {
 	int i;
 
 	for (i = 0; i <= TTM_PL_PRIV5; i++)
-		if (flags & (1 << i)) {
+		if (place->flags & (1 << i)) {
 			*mem_type = i;
 			return 0;
 		}
@@ -89,12 +91,12 @@ static void ttm_bo_mem_space_debug(struct ttm_buffer_object *bo,
 	       bo, bo->mem.num_pages, bo->mem.size >> 10,
 	       bo->mem.size >> 20);
 	for (i = 0; i < placement->num_placement; i++) {
-		ret = ttm_mem_type_from_flags(placement->placement[i],
+		ret = ttm_mem_type_from_place(&placement->placement[i],
 						&mem_type);
 		if (ret)
 			return;
 		pr_err("  placement[%d]=0x%08X (%d)\n",
-		       i, placement->placement[i], mem_type);
+		       i, placement->placement[i].flags, mem_type);
 		ttm_mem_type_debug(bo->bdev, mem_type);
 	}
 }
@@ -141,7 +143,6 @@ static void ttm_bo_release_list(struct kref *list_kref)
 	BUG_ON(atomic_read(&bo->list_kref.refcount));
 	BUG_ON(atomic_read(&bo->kref.refcount));
 	BUG_ON(atomic_read(&bo->cpu_writers));
-	BUG_ON(bo->sync_obj != NULL);
 	BUG_ON(bo->mem.mm_node != NULL);
 	BUG_ON(!list_empty(&bo->lru));
 	BUG_ON(!list_empty(&bo->ddestroy));
@@ -402,36 +403,48 @@ static void ttm_bo_cleanup_memtype_use(struct ttm_buffer_object *bo)
 	ww_mutex_unlock (&bo->resv->lock);
 }
 
+static void ttm_bo_flush_all_fences(struct ttm_buffer_object *bo)
+{
+	struct reservation_object_list *fobj;
+	struct fence *fence;
+	int i;
+
+	fobj = reservation_object_get_list(bo->resv);
+	fence = reservation_object_get_excl(bo->resv);
+	if (fence && !fence->ops->signaled)
+		fence_enable_sw_signaling(fence);
+
+	for (i = 0; fobj && i < fobj->shared_count; ++i) {
+		fence = rcu_dereference_protected(fobj->shared[i],
+					reservation_object_held(bo->resv));
+
+		if (!fence->ops->signaled)
+			fence_enable_sw_signaling(fence);
+	}
+}
+
 static void ttm_bo_cleanup_refs_or_queue(struct ttm_buffer_object *bo)
 {
 	struct ttm_bo_device *bdev = bo->bdev;
 	struct ttm_bo_global *glob = bo->glob;
-	struct ttm_bo_driver *driver = bdev->driver;
-	void *sync_obj = NULL;
 	int put_count;
 	int ret;
 
 	spin_lock(&glob->lru_lock);
 	ret = __ttm_bo_reserve(bo, false, true, false, NULL);
 
-	spin_lock(&bdev->fence_lock);
-	(void) ttm_bo_wait(bo, false, false, true);
-	if (!ret && !bo->sync_obj) {
-		spin_unlock(&bdev->fence_lock);
-		put_count = ttm_bo_del_from_lru(bo);
-
-		spin_unlock(&glob->lru_lock);
-		ttm_bo_cleanup_memtype_use(bo);
+	if (!ret) {
+		if (!ttm_bo_wait(bo, false, false, true)) {
+			put_count = ttm_bo_del_from_lru(bo);
 
-		ttm_bo_list_ref_sub(bo, put_count, true);
+			spin_unlock(&glob->lru_lock);
+			ttm_bo_cleanup_memtype_use(bo);
 
-		return;
-	}
-	if (bo->sync_obj)
-		sync_obj = driver->sync_obj_ref(bo->sync_obj);
-	spin_unlock(&bdev->fence_lock);
+			ttm_bo_list_ref_sub(bo, put_count, true);
 
-	if (!ret) {
+			return;
+		} else
+			ttm_bo_flush_all_fences(bo);
 
 		/*
 		 * Make NO_EVICT bos immediately available to
@@ -450,10 +463,6 @@ static void ttm_bo_cleanup_refs_or_queue(struct ttm_buffer_object *bo)
 	list_add_tail(&bo->ddestroy, &bdev->ddestroy);
 	spin_unlock(&glob->lru_lock);
 
-	if (sync_obj) {
-		driver->sync_obj_flush(sync_obj);
-		driver->sync_obj_unref(&sync_obj);
-	}
 	schedule_delayed_work(&bdev->wq,
 			      ((HZ / 100) < 1) ? 1 : HZ / 100);
 }
@@ -474,44 +483,26 @@ static int ttm_bo_cleanup_refs_and_unlock(struct ttm_buffer_object *bo,
 					  bool interruptible,
 					  bool no_wait_gpu)
 {
-	struct ttm_bo_device *bdev = bo->bdev;
-	struct ttm_bo_driver *driver = bdev->driver;
 	struct ttm_bo_global *glob = bo->glob;
 	int put_count;
 	int ret;
 
-	spin_lock(&bdev->fence_lock);
 	ret = ttm_bo_wait(bo, false, false, true);
 
 	if (ret && !no_wait_gpu) {
-		void *sync_obj;
-
-		/*
-		 * Take a reference to the fence and unreserve,
-		 * at this point the buffer should be dead, so
-		 * no new sync objects can be attached.
-		 */
-		sync_obj = driver->sync_obj_ref(bo->sync_obj);
-		spin_unlock(&bdev->fence_lock);
-
-		__ttm_bo_unreserve(bo);
+		long lret;
+		ww_mutex_unlock(&bo->resv->lock);
 		spin_unlock(&glob->lru_lock);
 
-		ret = driver->sync_obj_wait(sync_obj, false, interruptible);
-		driver->sync_obj_unref(&sync_obj);
-		if (ret)
-			return ret;
+		lret = reservation_object_wait_timeout_rcu(bo->resv,
+							   true,
+							   interruptible,
+							   30 * HZ);
 
-		/*
-		 * remove sync_obj with ttm_bo_wait, the wait should be
-		 * finished, and no new wait object should have been added.
-		 */
-		spin_lock(&bdev->fence_lock);
-		ret = ttm_bo_wait(bo, false, false, true);
-		WARN_ON(ret);
-		spin_unlock(&bdev->fence_lock);
-		if (ret)
-			return ret;
+		if (lret < 0)
+			return lret;
+		else if (lret == 0)
+			return -EBUSY;
 
 		spin_lock(&glob->lru_lock);
 		ret = __ttm_bo_reserve(bo, false, true, false, NULL);
@@ -528,8 +519,14 @@ static int ttm_bo_cleanup_refs_and_unlock(struct ttm_buffer_object *bo,
 			spin_unlock(&glob->lru_lock);
 			return 0;
 		}
-	} else
-		spin_unlock(&bdev->fence_lock);
+
+		/*
+		 * remove sync_obj with ttm_bo_wait, the wait should be
+		 * finished, and no new wait object should have been added.
+		 */
+		ret = ttm_bo_wait(bo, false, false, true);
+		WARN_ON(ret);
+	}
 
 	if (ret || unlikely(list_empty(&bo->ddestroy))) {
 		__ttm_bo_unreserve(bo);
@@ -667,9 +664,7 @@ static int ttm_bo_evict(struct ttm_buffer_object *bo, bool interruptible,
 	struct ttm_placement placement;
 	int ret = 0;
 
-	spin_lock(&bdev->fence_lock);
 	ret = ttm_bo_wait(bo, false, interruptible, no_wait_gpu);
-	spin_unlock(&bdev->fence_lock);
 
 	if (unlikely(ret != 0)) {
 		if (ret != -ERESTARTSYS) {
@@ -685,8 +680,6 @@ static int ttm_bo_evict(struct ttm_buffer_object *bo, bool interruptible,
 	evict_mem.bus.io_reserved_vm = false;
 	evict_mem.bus.io_reserved_count = 0;
 
-	placement.fpfn = 0;
-	placement.lpfn = 0;
 	placement.num_placement = 0;
 	placement.num_busy_placement = 0;
 	bdev->driver->evict_flags(bo, &placement);
@@ -716,6 +709,7 @@ out:
 
 static int ttm_mem_evict_first(struct ttm_bo_device *bdev,
 				uint32_t mem_type,
+				const struct ttm_place *place,
 				bool interruptible,
 				bool no_wait_gpu)
 {
@@ -727,8 +721,21 @@ static int ttm_mem_evict_first(struct ttm_bo_device *bdev,
 	spin_lock(&glob->lru_lock);
 	list_for_each_entry(bo, &man->lru, lru) {
 		ret = __ttm_bo_reserve(bo, false, true, false, NULL);
-		if (!ret)
+		if (!ret) {
+			if (place && (place->fpfn || place->lpfn)) {
+				/* Don't evict this BO if it's outside of the
+				 * requested placement range
+				 */
+				if (place->fpfn >= (bo->mem.start + bo->mem.size) ||
+				    (place->lpfn && place->lpfn <= bo->mem.start)) {
+					__ttm_bo_unreserve(bo);
+					ret = -EBUSY;
+					continue;
+				}
+			}
+
 			break;
+		}
 	}
 
 	if (ret) {
@@ -774,7 +781,7 @@ EXPORT_SYMBOL(ttm_bo_mem_put);
  */
 static int ttm_bo_mem_force_space(struct ttm_buffer_object *bo,
 					uint32_t mem_type,
-					struct ttm_placement *placement,
+					const struct ttm_place *place,
 					struct ttm_mem_reg *mem,
 					bool interruptible,
 					bool no_wait_gpu)
@@ -784,12 +791,12 @@ static int ttm_bo_mem_force_space(struct ttm_buffer_object *bo,
 	int ret;
 
 	do {
-		ret = (*man->func->get_node)(man, bo, placement, 0, mem);
+		ret = (*man->func->get_node)(man, bo, place, mem);
 		if (unlikely(ret != 0))
 			return ret;
 		if (mem->mm_node)
 			break;
-		ret = ttm_mem_evict_first(bdev, mem_type,
+		ret = ttm_mem_evict_first(bdev, mem_type, place,
 					  interruptible, no_wait_gpu);
 		if (unlikely(ret != 0))
 			return ret;
@@ -827,18 +834,18 @@ static uint32_t ttm_bo_select_caching(struct ttm_mem_type_manager *man,
 
 static bool ttm_bo_mt_compatible(struct ttm_mem_type_manager *man,
 				 uint32_t mem_type,
-				 uint32_t proposed_placement,
+				 const struct ttm_place *place,
 				 uint32_t *masked_placement)
 {
 	uint32_t cur_flags = ttm_bo_type_flags(mem_type);
 
-	if ((cur_flags & proposed_placement & TTM_PL_MASK_MEM) == 0)
+	if ((cur_flags & place->flags & TTM_PL_MASK_MEM) == 0)
 		return false;
 
-	if ((proposed_placement & man->available_caching) == 0)
+	if ((place->flags & man->available_caching) == 0)
 		return false;
 
-	cur_flags |= (proposed_placement & man->available_caching);
+	cur_flags |= (place->flags & man->available_caching);
 
 	*masked_placement = cur_flags;
 	return true;
@@ -869,15 +876,14 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 
 	mem->mm_node = NULL;
 	for (i = 0; i < placement->num_placement; ++i) {
-		ret = ttm_mem_type_from_flags(placement->placement[i],
-						&mem_type);
+		const struct ttm_place *place = &placement->placement[i];
+
+		ret = ttm_mem_type_from_place(place, &mem_type);
 		if (ret)
 			return ret;
 		man = &bdev->man[mem_type];
 
-		type_ok = ttm_bo_mt_compatible(man,
-						mem_type,
-						placement->placement[i],
+		type_ok = ttm_bo_mt_compatible(man, mem_type, place,
 						&cur_flags);
 
 		if (!type_ok)
@@ -889,7 +895,7 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 		 * Use the access and other non-mapping-related flag bits from
 		 * the memory placement flags to the current flags
 		 */
-		ttm_flag_masked(&cur_flags, placement->placement[i],
+		ttm_flag_masked(&cur_flags, place->flags,
 				~TTM_PL_MASK_MEMTYPE);
 
 		if (mem_type == TTM_PL_SYSTEM)
@@ -897,8 +903,7 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 
 		if (man->has_type && man->use_type) {
 			type_found = true;
-			ret = (*man->func->get_node)(man, bo, placement,
-						     cur_flags, mem);
+			ret = (*man->func->get_node)(man, bo, place, mem);
 			if (unlikely(ret))
 				return ret;
 		}
@@ -916,17 +921,15 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 		return -EINVAL;
 
 	for (i = 0; i < placement->num_busy_placement; ++i) {
-		ret = ttm_mem_type_from_flags(placement->busy_placement[i],
-						&mem_type);
+		const struct ttm_place *place = &placement->busy_placement[i];
+
+		ret = ttm_mem_type_from_place(place, &mem_type);
 		if (ret)
 			return ret;
 		man = &bdev->man[mem_type];
 		if (!man->has_type)
 			continue;
-		if (!ttm_bo_mt_compatible(man,
-						mem_type,
-						placement->busy_placement[i],
-						&cur_flags))
+		if (!ttm_bo_mt_compatible(man, mem_type, place, &cur_flags))
 			continue;
 
 		cur_flags = ttm_bo_select_caching(man, bo->mem.placement,
@@ -935,7 +938,7 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 		 * Use the access and other non-mapping-related flag bits from
 		 * the memory placement flags to the current flags
 		 */
-		ttm_flag_masked(&cur_flags, placement->busy_placement[i],
+		ttm_flag_masked(&cur_flags, place->flags,
 				~TTM_PL_MASK_MEMTYPE);
 
 		if (mem_type == TTM_PL_SYSTEM) {
@@ -945,7 +948,7 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 			return 0;
 		}
 
-		ret = ttm_bo_mem_force_space(bo, mem_type, placement, mem,
+		ret = ttm_bo_mem_force_space(bo, mem_type, place, mem,
 						interruptible, no_wait_gpu);
 		if (ret == 0 && mem->mm_node) {
 			mem->placement = cur_flags;
@@ -966,7 +969,6 @@ static int ttm_bo_move_buffer(struct ttm_buffer_object *bo,
 {
 	int ret = 0;
 	struct ttm_mem_reg mem;
-	struct ttm_bo_device *bdev = bo->bdev;
 
 	lockdep_assert_held(&bo->resv->lock.base);
 
@@ -975,9 +977,7 @@ static int ttm_bo_move_buffer(struct ttm_buffer_object *bo,
 	 * Have the driver move function wait for idle when necessary,
 	 * instead of doing it here.
 	 */
-	spin_lock(&bdev->fence_lock);
 	ret = ttm_bo_wait(bo, false, interruptible, no_wait_gpu);
-	spin_unlock(&bdev->fence_lock);
 	if (ret)
 		return ret;
 	mem.num_pages = bo->num_pages;
@@ -1006,20 +1006,27 @@ static bool ttm_bo_mem_compat(struct ttm_placement *placement,
 {
 	int i;
 
-	if (mem->mm_node && placement->lpfn != 0 &&
-	    (mem->start < placement->fpfn ||
-	     mem->start + mem->num_pages > placement->lpfn))
-		return false;
-
 	for (i = 0; i < placement->num_placement; i++) {
-		*new_flags = placement->placement[i];
+		const struct ttm_place *heap = &placement->placement[i];
+		if (mem->mm_node &&
+		    (mem->start < heap->fpfn ||
+		     (heap->lpfn != 0 && (mem->start + mem->num_pages) > heap->lpfn)))
+			continue;
+
+		*new_flags = heap->flags;
 		if ((*new_flags & mem->placement & TTM_PL_MASK_CACHING) &&
 		    (*new_flags & mem->placement & TTM_PL_MASK_MEM))
 			return true;
 	}
 
 	for (i = 0; i < placement->num_busy_placement; i++) {
-		*new_flags = placement->busy_placement[i];
+		const struct ttm_place *heap = &placement->busy_placement[i];
+		if (mem->mm_node &&
+		    (mem->start < heap->fpfn ||
+		     (heap->lpfn != 0 && (mem->start + mem->num_pages) > heap->lpfn)))
+			continue;
+
+		*new_flags = heap->flags;
 		if ((*new_flags & mem->placement & TTM_PL_MASK_CACHING) &&
 		    (*new_flags & mem->placement & TTM_PL_MASK_MEM))
 			return true;
@@ -1037,11 +1044,6 @@ int ttm_bo_validate(struct ttm_buffer_object *bo,
 	uint32_t new_flags;
 
 	lockdep_assert_held(&bo->resv->lock.base);
-	/* Check that range is valid */
-	if (placement->lpfn || placement->fpfn)
-		if (placement->fpfn > placement->lpfn ||
-			(placement->lpfn - placement->fpfn) < bo->num_pages)
-			return -EINVAL;
 	/*
 	 * Check whether we need to move buffer.
 	 */
@@ -1070,15 +1072,6 @@ int ttm_bo_validate(struct ttm_buffer_object *bo,
 }
 EXPORT_SYMBOL(ttm_bo_validate);
 
-int ttm_bo_check_placement(struct ttm_buffer_object *bo,
-				struct ttm_placement *placement)
-{
-	BUG_ON((placement->fpfn || placement->lpfn) &&
-	       (bo->mem.num_pages > (placement->lpfn - placement->fpfn)));
-
-	return 0;
-}
-
 int ttm_bo_init(struct ttm_bo_device *bdev,
 		struct ttm_buffer_object *bo,
 		unsigned long size,
@@ -1089,6 +1082,7 @@ int ttm_bo_init(struct ttm_bo_device *bdev,
 		struct file *persistent_swap_storage,
 		size_t acc_size,
 		struct sg_table *sg,
+		struct reservation_object *resv,
 		void (*destroy) (struct ttm_buffer_object *))
 {
 	int ret = 0;
@@ -1142,30 +1136,38 @@ int ttm_bo_init(struct ttm_bo_device *bdev,
 	bo->persistent_swap_storage = persistent_swap_storage;
 	bo->acc_size = acc_size;
 	bo->sg = sg;
-	bo->resv = &bo->ttm_resv;
-	reservation_object_init(bo->resv);
+	if (resv) {
+		bo->resv = resv;
+		lockdep_assert_held(&bo->resv->lock.base);
+	} else {
+		bo->resv = &bo->ttm_resv;
+		reservation_object_init(&bo->ttm_resv);
+	}
 	atomic_inc(&bo->glob->bo_count);
 	drm_vma_node_reset(&bo->vma_node);
 
-	ret = ttm_bo_check_placement(bo, placement);
-
 	/*
 	 * For ttm_bo_type_device buffers, allocate
 	 * address space from the device.
 	 */
-	if (likely(!ret) &&
-	    (bo->type == ttm_bo_type_device ||
-	     bo->type == ttm_bo_type_sg))
+	if (bo->type == ttm_bo_type_device ||
+	    bo->type == ttm_bo_type_sg)
 		ret = drm_vma_offset_add(&bdev->vma_manager, &bo->vma_node,
 					 bo->mem.num_pages);
 
-	locked = ww_mutex_trylock(&bo->resv->lock);
-	WARN_ON(!locked);
+	/* passed reservation objects should already be locked,
+	 * since otherwise lockdep will be angered in radeon.
+	 */
+	if (!resv) {
+		locked = ww_mutex_trylock(&bo->resv->lock);
+		WARN_ON(!locked);
+	}
 
 	if (likely(!ret))
 		ret = ttm_bo_validate(bo, placement, interruptible, false);
 
-	ttm_bo_unreserve(bo);
+	if (!resv)
+		ttm_bo_unreserve(bo);
 
 	if (unlikely(ret))
 		ttm_bo_unref(&bo);
@@ -1223,7 +1225,7 @@ int ttm_bo_create(struct ttm_bo_device *bdev,
 	acc_size = ttm_bo_acc_size(bdev, size, sizeof(struct ttm_buffer_object));
 	ret = ttm_bo_init(bdev, bo, size, type, placement, page_alignment,
 			  interruptible, persistent_swap_storage, acc_size,
-			  NULL, NULL);
+			  NULL, NULL, NULL);
 	if (likely(ret == 0))
 		*p_bo = bo;
 
@@ -1245,7 +1247,7 @@ static int ttm_bo_force_list_clean(struct ttm_bo_device *bdev,
 	spin_lock(&glob->lru_lock);
 	while (!list_empty(&man->lru)) {
 		spin_unlock(&glob->lru_lock);
-		ret = ttm_mem_evict_first(bdev, mem_type, false, false);
+		ret = ttm_mem_evict_first(bdev, mem_type, NULL, false, false);
 		if (ret) {
 			if (allow_errors) {
 				return ret;
@@ -1477,7 +1479,6 @@ int ttm_bo_device_init(struct ttm_bo_device *bdev,
 	bdev->glob = glob;
 	bdev->need_dma32 = need_dma32;
 	bdev->val_seq = 0;
-	spin_lock_init(&bdev->fence_lock);
 	mutex_lock(&glob->device_list_mutex);
 	list_add_tail(&bdev->device_list, &glob->device_list);
 	mutex_unlock(&glob->device_list_mutex);
@@ -1530,65 +1531,56 @@ void ttm_bo_unmap_virtual(struct ttm_buffer_object *bo)
 
 EXPORT_SYMBOL(ttm_bo_unmap_virtual);
 
-
 int ttm_bo_wait(struct ttm_buffer_object *bo,
 		bool lazy, bool interruptible, bool no_wait)
 {
-	struct ttm_bo_driver *driver = bo->bdev->driver;
-	struct ttm_bo_device *bdev = bo->bdev;
-	void *sync_obj;
-	int ret = 0;
-
-	if (likely(bo->sync_obj == NULL))
-		return 0;
+	struct reservation_object_list *fobj;
+	struct reservation_object *resv;
+	struct fence *excl;
+	long timeout = 15 * HZ;
+	int i;
 
-	while (bo->sync_obj) {
+	resv = bo->resv;
+	fobj = reservation_object_get_list(resv);
+	excl = reservation_object_get_excl(resv);
+	if (excl) {
+		if (!fence_is_signaled(excl)) {
+			if (no_wait)
+				return -EBUSY;
 
-		if (driver->sync_obj_signaled(bo->sync_obj)) {
-			void *tmp_obj = bo->sync_obj;
-			bo->sync_obj = NULL;
-			clear_bit(TTM_BO_PRIV_FLAG_MOVING, &bo->priv_flags);
-			spin_unlock(&bdev->fence_lock);
-			driver->sync_obj_unref(&tmp_obj);
-			spin_lock(&bdev->fence_lock);
-			continue;
+			timeout = fence_wait_timeout(excl,
+						     interruptible, timeout);
 		}
+	}
 
-		if (no_wait)
-			return -EBUSY;
+	for (i = 0; fobj && timeout > 0 && i < fobj->shared_count; ++i) {
+		struct fence *fence;
+		fence = rcu_dereference_protected(fobj->shared[i],
+						reservation_object_held(resv));
 
-		sync_obj = driver->sync_obj_ref(bo->sync_obj);
-		spin_unlock(&bdev->fence_lock);
-		ret = driver->sync_obj_wait(sync_obj,
-					    lazy, interruptible);
-		if (unlikely(ret != 0)) {
-			driver->sync_obj_unref(&sync_obj);
-			spin_lock(&bdev->fence_lock);
-			return ret;
-		}
-		spin_lock(&bdev->fence_lock);
-		if (likely(bo->sync_obj == sync_obj)) {
-			void *tmp_obj = bo->sync_obj;
-			bo->sync_obj = NULL;
-			clear_bit(TTM_BO_PRIV_FLAG_MOVING,
-				  &bo->priv_flags);
-			spin_unlock(&bdev->fence_lock);
-			driver->sync_obj_unref(&sync_obj);
-			driver->sync_obj_unref(&tmp_obj);
-			spin_lock(&bdev->fence_lock);
-		} else {
-			spin_unlock(&bdev->fence_lock);
-			driver->sync_obj_unref(&sync_obj);
-			spin_lock(&bdev->fence_lock);
+		if (!fence_is_signaled(fence)) {
+			if (no_wait)
+				return -EBUSY;
+
+			timeout = fence_wait_timeout(fence,
+						     interruptible, timeout);
 		}
 	}
+
+	if (timeout < 0)
+		return timeout;
+
+	if (timeout == 0)
+		return -EBUSY;
+
+	reservation_object_add_excl_fence(resv, NULL);
+	clear_bit(TTM_BO_PRIV_FLAG_MOVING, &bo->priv_flags);
 	return 0;
 }
 EXPORT_SYMBOL(ttm_bo_wait);
 
 int ttm_bo_synccpu_write_grab(struct ttm_buffer_object *bo, bool no_wait)
 {
-	struct ttm_bo_device *bdev = bo->bdev;
 	int ret = 0;
 
 	/*
@@ -1598,9 +1590,7 @@ int ttm_bo_synccpu_write_grab(struct ttm_buffer_object *bo, bool no_wait)
 	ret = ttm_bo_reserve(bo, true, no_wait, false, NULL);
 	if (unlikely(ret != 0))
 		return ret;
-	spin_lock(&bdev->fence_lock);
 	ret = ttm_bo_wait(bo, false, true, no_wait);
-	spin_unlock(&bdev->fence_lock);
 	if (likely(ret == 0))
 		atomic_inc(&bo->cpu_writers);
 	ttm_bo_unreserve(bo);
@@ -1657,9 +1647,7 @@ static int ttm_bo_swapout(struct ttm_mem_shrink *shrink)
 	 * Wait for GPU, then move to system cached.
 	 */
 
-	spin_lock(&bo->bdev->fence_lock);
 	ret = ttm_bo_wait(bo, false, false, false);
-	spin_unlock(&bo->bdev->fence_lock);
 
 	if (unlikely(ret != 0))
 		goto out;
diff --git a/drivers/gpu/drm/ttm/ttm_bo_manager.c b/drivers/gpu/drm/ttm/ttm_bo_manager.c
index 9e103a4875c8..aa0bd054d3e9 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_manager.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_manager.c
@@ -49,18 +49,18 @@ struct ttm_range_manager {
 
 static int ttm_bo_man_get_node(struct ttm_mem_type_manager *man,
 			       struct ttm_buffer_object *bo,
-			       struct ttm_placement *placement,
-			       uint32_t flags,
+			       const struct ttm_place *place,
 			       struct ttm_mem_reg *mem)
 {
 	struct ttm_range_manager *rman = (struct ttm_range_manager *) man->priv;
 	struct drm_mm *mm = &rman->mm;
 	struct drm_mm_node *node = NULL;
+	enum drm_mm_search_flags sflags = DRM_MM_SEARCH_BEST;
 	enum drm_mm_allocator_flags aflags = DRM_MM_CREATE_DEFAULT;
 	unsigned long lpfn;
 	int ret;
 
-	lpfn = placement->lpfn;
+	lpfn = place->lpfn;
 	if (!lpfn)
 		lpfn = man->size;
 
@@ -68,15 +68,16 @@ static int ttm_bo_man_get_node(struct ttm_mem_type_manager *man,
 	if (!node)
 		return -ENOMEM;
 
-	if (flags & TTM_PL_FLAG_TOPDOWN)
+	if (place->flags & TTM_PL_FLAG_TOPDOWN) {
+		sflags = DRM_MM_SEARCH_BELOW;
 		aflags = DRM_MM_CREATE_TOP;
+	}
 
 	spin_lock(&rman->lock);
 	ret = drm_mm_insert_node_in_range_generic(mm, node, mem->num_pages,
 					  mem->page_alignment, 0,
-					  placement->fpfn, lpfn,
-					  DRM_MM_SEARCH_BEST,
-					  aflags);
+					  place->fpfn, lpfn,
+					  sflags, aflags);
 	spin_unlock(&rman->lock);
 
 	if (unlikely(ret)) {
diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c b/drivers/gpu/drm/ttm/ttm_bo_util.c
index 30e5d90cb7bc..882cccdad272 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_util.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_util.c
@@ -37,6 +37,7 @@
 #include <linux/slab.h>
 #include <linux/vmalloc.h>
 #include <linux/module.h>
+#include <linux/reservation.h>
 
 void ttm_bo_free_old_node(struct ttm_buffer_object *bo)
 {
@@ -444,8 +445,6 @@ static int ttm_buffer_object_transfer(struct ttm_buffer_object *bo,
 				      struct ttm_buffer_object **new_obj)
 {
 	struct ttm_buffer_object *fbo;
-	struct ttm_bo_device *bdev = bo->bdev;
-	struct ttm_bo_driver *driver = bdev->driver;
 	int ret;
 
 	fbo = kmalloc(sizeof(*fbo), GFP_KERNEL);
@@ -466,12 +465,6 @@ static int ttm_buffer_object_transfer(struct ttm_buffer_object *bo,
 	drm_vma_node_reset(&fbo->vma_node);
 	atomic_set(&fbo->cpu_writers, 0);
 
-	spin_lock(&bdev->fence_lock);
-	if (bo->sync_obj)
-		fbo->sync_obj = driver->sync_obj_ref(bo->sync_obj);
-	else
-		fbo->sync_obj = NULL;
-	spin_unlock(&bdev->fence_lock);
 	kref_init(&fbo->list_kref);
 	kref_init(&fbo->kref);
 	fbo->destroy = &ttm_transfered_destroy;
@@ -487,28 +480,24 @@ static int ttm_buffer_object_transfer(struct ttm_buffer_object *bo,
 
 pgprot_t ttm_io_prot(uint32_t caching_flags, pgprot_t tmp)
 {
+	/* Cached mappings need no adjustment */
+	if (caching_flags & TTM_PL_FLAG_CACHED)
+		return tmp;
+
 #if defined(__i386__) || defined(__x86_64__)
 	if (caching_flags & TTM_PL_FLAG_WC)
 		tmp = pgprot_writecombine(tmp);
 	else if (boot_cpu_data.x86 > 3)
 		tmp = pgprot_noncached(tmp);
-
-#elif defined(__powerpc__)
-	if (!(caching_flags & TTM_PL_FLAG_CACHED)) {
-		pgprot_val(tmp) |= _PAGE_NO_CACHE;
-		if (caching_flags & TTM_PL_FLAG_UNCACHED)
-			pgprot_val(tmp) |= _PAGE_GUARDED;
-	}
 #endif
-#if defined(__ia64__) || defined(__arm__)
+#if defined(__ia64__) || defined(__arm__) || defined(__powerpc__)
 	if (caching_flags & TTM_PL_FLAG_WC)
 		tmp = pgprot_writecombine(tmp);
 	else
 		tmp = pgprot_noncached(tmp);
 #endif
 #if defined(__sparc__) || defined(__mips__)
-	if (!(caching_flags & TTM_PL_FLAG_CACHED))
-		tmp = pgprot_noncached(tmp);
+	tmp = pgprot_noncached(tmp);
 #endif
 	return tmp;
 }
@@ -567,9 +556,7 @@ static int ttm_bo_kmap_ttm(struct ttm_buffer_object *bo,
 		 * We need to use vmap to get the desired page protection
 		 * or to make the buffer object look contiguous.
 		 */
-		prot = (mem->placement & TTM_PL_FLAG_CACHED) ?
-			PAGE_KERNEL :
-			ttm_io_prot(mem->placement, PAGE_KERNEL);
+		prot = ttm_io_prot(mem->placement, PAGE_KERNEL);
 		map->bo_kmap_type = ttm_bo_map_vmap;
 		map->virtual = vmap(ttm->pages + start_page, num_pages,
 				    0, prot);
@@ -644,30 +631,20 @@ void ttm_bo_kunmap(struct ttm_bo_kmap_obj *map)
 EXPORT_SYMBOL(ttm_bo_kunmap);
 
 int ttm_bo_move_accel_cleanup(struct ttm_buffer_object *bo,
-			      void *sync_obj,
+			      struct fence *fence,
 			      bool evict,
 			      bool no_wait_gpu,
 			      struct ttm_mem_reg *new_mem)
 {
 	struct ttm_bo_device *bdev = bo->bdev;
-	struct ttm_bo_driver *driver = bdev->driver;
 	struct ttm_mem_type_manager *man = &bdev->man[new_mem->mem_type];
 	struct ttm_mem_reg *old_mem = &bo->mem;
 	int ret;
 	struct ttm_buffer_object *ghost_obj;
-	void *tmp_obj = NULL;
 
-	spin_lock(&bdev->fence_lock);
-	if (bo->sync_obj) {
-		tmp_obj = bo->sync_obj;
-		bo->sync_obj = NULL;
-	}
-	bo->sync_obj = driver->sync_obj_ref(sync_obj);
+	reservation_object_add_excl_fence(bo->resv, fence);
 	if (evict) {
 		ret = ttm_bo_wait(bo, false, false, false);
-		spin_unlock(&bdev->fence_lock);
-		if (tmp_obj)
-			driver->sync_obj_unref(&tmp_obj);
 		if (ret)
 			return ret;
 
@@ -688,14 +665,13 @@ int ttm_bo_move_accel_cleanup(struct ttm_buffer_object *bo,
 		 */
 
 		set_bit(TTM_BO_PRIV_FLAG_MOVING, &bo->priv_flags);
-		spin_unlock(&bdev->fence_lock);
-		if (tmp_obj)
-			driver->sync_obj_unref(&tmp_obj);
 
 		ret = ttm_buffer_object_transfer(bo, &ghost_obj);
 		if (ret)
 			return ret;
 
+		reservation_object_add_excl_fence(ghost_obj->resv, fence);
+
 		/**
 		 * If we're not moving to fixed memory, the TTM object
 		 * needs to stay alive. Otherwhise hang it on the ghost
diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c
index 0ce48e5a9cb4..8fb7213277cc 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_vm.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c
@@ -45,10 +45,8 @@ static int ttm_bo_vm_fault_idle(struct ttm_buffer_object *bo,
 				struct vm_area_struct *vma,
 				struct vm_fault *vmf)
 {
-	struct ttm_bo_device *bdev = bo->bdev;
 	int ret = 0;
 
-	spin_lock(&bdev->fence_lock);
 	if (likely(!test_bit(TTM_BO_PRIV_FLAG_MOVING, &bo->priv_flags)))
 		goto out_unlock;
 
@@ -82,7 +80,6 @@ static int ttm_bo_vm_fault_idle(struct ttm_buffer_object *bo,
 			VM_FAULT_NOPAGE;
 
 out_unlock:
-	spin_unlock(&bdev->fence_lock);
 	return ret;
 }
 
@@ -200,9 +197,8 @@ static int ttm_bo_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 						cvma.vm_page_prot);
 	} else {
 		ttm = bo->ttm;
-		if (!(bo->mem.placement & TTM_PL_FLAG_CACHED))
-			cvma.vm_page_prot = ttm_io_prot(bo->mem.placement,
-							cvma.vm_page_prot);
+		cvma.vm_page_prot = ttm_io_prot(bo->mem.placement,
+						cvma.vm_page_prot);
 
 		/* Allocate all page at once, most common usage */
 		if (ttm->bdev->driver->ttm_tt_populate(ttm)) {
diff --git a/drivers/gpu/drm/ttm/ttm_execbuf_util.c b/drivers/gpu/drm/ttm/ttm_execbuf_util.c
index e8dac8758528..3820ae97a030 100644
--- a/drivers/gpu/drm/ttm/ttm_execbuf_util.c
+++ b/drivers/gpu/drm/ttm/ttm_execbuf_util.c
@@ -32,20 +32,12 @@
 #include <linux/sched.h>
 #include <linux/module.h>
 
-static void ttm_eu_backoff_reservation_locked(struct list_head *list)
+static void ttm_eu_backoff_reservation_reverse(struct list_head *list,
+					      struct ttm_validate_buffer *entry)
 {
-	struct ttm_validate_buffer *entry;
-
-	list_for_each_entry(entry, list, head) {
+	list_for_each_entry_continue_reverse(entry, list, head) {
 		struct ttm_buffer_object *bo = entry->bo;
-		if (!entry->reserved)
-			continue;
 
-		entry->reserved = false;
-		if (entry->removed) {
-			ttm_bo_add_to_lru(bo);
-			entry->removed = false;
-		}
 		__ttm_bo_unreserve(bo);
 	}
 }
@@ -56,27 +48,9 @@ static void ttm_eu_del_from_lru_locked(struct list_head *list)
 
 	list_for_each_entry(entry, list, head) {
 		struct ttm_buffer_object *bo = entry->bo;
-		if (!entry->reserved)
-			continue;
-
-		if (!entry->removed) {
-			entry->put_count = ttm_bo_del_from_lru(bo);
-			entry->removed = true;
-		}
-	}
-}
-
-static void ttm_eu_list_ref_sub(struct list_head *list)
-{
-	struct ttm_validate_buffer *entry;
-
-	list_for_each_entry(entry, list, head) {
-		struct ttm_buffer_object *bo = entry->bo;
+		unsigned put_count = ttm_bo_del_from_lru(bo);
 
-		if (entry->put_count) {
-			ttm_bo_list_ref_sub(bo, entry->put_count, true);
-			entry->put_count = 0;
-		}
+		ttm_bo_list_ref_sub(bo, put_count, true);
 	}
 }
 
@@ -91,11 +65,18 @@ void ttm_eu_backoff_reservation(struct ww_acquire_ctx *ticket,
 
 	entry = list_first_entry(list, struct ttm_validate_buffer, head);
 	glob = entry->bo->glob;
+
 	spin_lock(&glob->lru_lock);
-	ttm_eu_backoff_reservation_locked(list);
+	list_for_each_entry(entry, list, head) {
+		struct ttm_buffer_object *bo = entry->bo;
+
+		ttm_bo_add_to_lru(bo);
+		__ttm_bo_unreserve(bo);
+	}
+	spin_unlock(&glob->lru_lock);
+
 	if (ticket)
 		ww_acquire_fini(ticket);
-	spin_unlock(&glob->lru_lock);
 }
 EXPORT_SYMBOL(ttm_eu_backoff_reservation);
 
@@ -112,7 +93,8 @@ EXPORT_SYMBOL(ttm_eu_backoff_reservation);
  */
 
 int ttm_eu_reserve_buffers(struct ww_acquire_ctx *ticket,
-			   struct list_head *list)
+			   struct list_head *list, bool intr,
+			   struct list_head *dups)
 {
 	struct ttm_bo_global *glob;
 	struct ttm_validate_buffer *entry;
@@ -121,60 +103,71 @@ int ttm_eu_reserve_buffers(struct ww_acquire_ctx *ticket,
 	if (list_empty(list))
 		return 0;
 
-	list_for_each_entry(entry, list, head) {
-		entry->reserved = false;
-		entry->put_count = 0;
-		entry->removed = false;
-	}
-
 	entry = list_first_entry(list, struct ttm_validate_buffer, head);
 	glob = entry->bo->glob;
 
 	if (ticket)
 		ww_acquire_init(ticket, &reservation_ww_class);
-retry:
+
 	list_for_each_entry(entry, list, head) {
 		struct ttm_buffer_object *bo = entry->bo;
 
-		/* already slowpath reserved? */
-		if (entry->reserved)
+		ret = __ttm_bo_reserve(bo, intr, (ticket == NULL), true,
+				       ticket);
+		if (!ret && unlikely(atomic_read(&bo->cpu_writers) > 0)) {
+			__ttm_bo_unreserve(bo);
+
+			ret = -EBUSY;
+
+		} else if (ret == -EALREADY && dups) {
+			struct ttm_validate_buffer *safe = entry;
+			entry = list_prev_entry(entry, head);
+			list_del(&safe->head);
+			list_add(&safe->head, dups);
 			continue;
+		}
 
-		ret = __ttm_bo_reserve(bo, true, (ticket == NULL), true,
-				       ticket);
+		if (!ret) {
+			if (!entry->shared)
+				continue;
+
+			ret = reservation_object_reserve_shared(bo->resv);
+			if (!ret)
+				continue;
+		}
 
-		if (ret == -EDEADLK) {
-			/* uh oh, we lost out, drop every reservation and try
-			 * to only reserve this buffer, then start over if
-			 * this succeeds.
-			 */
-			BUG_ON(ticket == NULL);
-			spin_lock(&glob->lru_lock);
-			ttm_eu_backoff_reservation_locked(list);
-			spin_unlock(&glob->lru_lock);
-			ttm_eu_list_ref_sub(list);
+		/* uh oh, we lost out, drop every reservation and try
+		 * to only reserve this buffer, then start over if
+		 * this succeeds.
+		 */
+		ttm_eu_backoff_reservation_reverse(list, entry);
+
+		if (ret == -EDEADLK && intr) {
 			ret = ww_mutex_lock_slow_interruptible(&bo->resv->lock,
 							       ticket);
-			if (unlikely(ret != 0)) {
-				if (ret == -EINTR)
-					ret = -ERESTARTSYS;
-				goto err_fini;
-			}
+		} else if (ret == -EDEADLK) {
+			ww_mutex_lock_slow(&bo->resv->lock, ticket);
+			ret = 0;
+		}
 
-			entry->reserved = true;
-			if (unlikely(atomic_read(&bo->cpu_writers) > 0)) {
-				ret = -EBUSY;
-				goto err;
-			}
-			goto retry;
-		} else if (ret)
-			goto err;
+		if (!ret && entry->shared)
+			ret = reservation_object_reserve_shared(bo->resv);
 
-		entry->reserved = true;
-		if (unlikely(atomic_read(&bo->cpu_writers) > 0)) {
-			ret = -EBUSY;
-			goto err;
+		if (unlikely(ret != 0)) {
+			if (ret == -EINTR)
+				ret = -ERESTARTSYS;
+			if (ticket) {
+				ww_acquire_done(ticket);
+				ww_acquire_fini(ticket);
+			}
+			return ret;
 		}
+
+		/* move this item to the front of the list,
+		 * forces correct iteration of the loop without keeping track
+		 */
+		list_del(&entry->head);
+		list_add(&entry->head, list);
 	}
 
 	if (ticket)
@@ -182,25 +175,12 @@ retry:
 	spin_lock(&glob->lru_lock);
 	ttm_eu_del_from_lru_locked(list);
 	spin_unlock(&glob->lru_lock);
-	ttm_eu_list_ref_sub(list);
 	return 0;
-
-err:
-	spin_lock(&glob->lru_lock);
-	ttm_eu_backoff_reservation_locked(list);
-	spin_unlock(&glob->lru_lock);
-	ttm_eu_list_ref_sub(list);
-err_fini:
-	if (ticket) {
-		ww_acquire_done(ticket);
-		ww_acquire_fini(ticket);
-	}
-	return ret;
 }
 EXPORT_SYMBOL(ttm_eu_reserve_buffers);
 
 void ttm_eu_fence_buffer_objects(struct ww_acquire_ctx *ticket,
-				 struct list_head *list, void *sync_obj)
+				 struct list_head *list, struct fence *fence)
 {
 	struct ttm_validate_buffer *entry;
 	struct ttm_buffer_object *bo;
@@ -217,24 +197,18 @@ void ttm_eu_fence_buffer_objects(struct ww_acquire_ctx *ticket,
 	glob = bo->glob;
 
 	spin_lock(&glob->lru_lock);
-	spin_lock(&bdev->fence_lock);
 
 	list_for_each_entry(entry, list, head) {
 		bo = entry->bo;
-		entry->old_sync_obj = bo->sync_obj;
-		bo->sync_obj = driver->sync_obj_ref(sync_obj);
+		if (entry->shared)
+			reservation_object_add_shared_fence(bo->resv, fence);
+		else
+			reservation_object_add_excl_fence(bo->resv, fence);
 		ttm_bo_add_to_lru(bo);
 		__ttm_bo_unreserve(bo);
-		entry->reserved = false;
 	}
-	spin_unlock(&bdev->fence_lock);
 	spin_unlock(&glob->lru_lock);
 	if (ticket)
 		ww_acquire_fini(ticket);
-
-	list_for_each_entry(entry, list, head) {
-		if (entry->old_sync_obj)
-			driver->sync_obj_unref(&entry->old_sync_obj);
-	}
 }
 EXPORT_SYMBOL(ttm_eu_fence_buffer_objects);
diff --git a/drivers/gpu/drm/ttm/ttm_memory.c b/drivers/gpu/drm/ttm/ttm_memory.c
index dbc2def887cd..a1803fbcc898 100644
--- a/drivers/gpu/drm/ttm/ttm_memory.c
+++ b/drivers/gpu/drm/ttm/ttm_memory.c
@@ -300,7 +300,8 @@ static int ttm_mem_init_highmem_zone(struct ttm_mem_global *glob,
 	zone->glob = glob;
 	glob->zone_highmem = zone;
 	ret = kobject_init_and_add(
-		&zone->kobj, &ttm_mem_zone_kobj_type, &glob->kobj, zone->name);
+		&zone->kobj, &ttm_mem_zone_kobj_type, &glob->kobj, "%s",
+		zone->name);
 	if (unlikely(ret != 0)) {
 		kobject_put(&zone->kobj);
 		return ret;
diff --git a/drivers/gpu/drm/ttm/ttm_page_alloc.c b/drivers/gpu/drm/ttm/ttm_page_alloc.c
index 09874d695188..025c429050c0 100644
--- a/drivers/gpu/drm/ttm/ttm_page_alloc.c
+++ b/drivers/gpu/drm/ttm/ttm_page_alloc.c
@@ -297,11 +297,12 @@ static void ttm_pool_update_free_locked(struct ttm_page_pool *pool,
  *
  * @pool: to free the pages from
  * @free_all: If set to true will free all pages in pool
- * @gfp: GFP flags.
+ * @use_static: Safe to use static buffer
  **/
 static int ttm_page_pool_free(struct ttm_page_pool *pool, unsigned nr_free,
-			      gfp_t gfp)
+			      bool use_static)
 {
+	static struct page *static_buf[NUM_PAGES_TO_ALLOC];
 	unsigned long irq_flags;
 	struct page *p;
 	struct page **pages_to_free;
@@ -311,7 +312,11 @@ static int ttm_page_pool_free(struct ttm_page_pool *pool, unsigned nr_free,
 	if (NUM_PAGES_TO_ALLOC < nr_free)
 		npages_to_free = NUM_PAGES_TO_ALLOC;
 
-	pages_to_free = kmalloc(npages_to_free * sizeof(struct page *), gfp);
+	if (use_static)
+		pages_to_free = static_buf;
+	else
+		pages_to_free = kmalloc(npages_to_free * sizeof(struct page *),
+					GFP_KERNEL);
 	if (!pages_to_free) {
 		pr_err("Failed to allocate memory for pool free operation\n");
 		return 0;
@@ -374,7 +379,8 @@ restart:
 	if (freed_pages)
 		ttm_pages_put(pages_to_free, freed_pages);
 out:
-	kfree(pages_to_free);
+	if (pages_to_free != static_buf)
+		kfree(pages_to_free);
 	return nr_free;
 }
 
@@ -383,8 +389,6 @@ out:
  *
  * XXX: (dchinner) Deadlock warning!
  *
- * We need to pass sc->gfp_mask to ttm_page_pool_free().
- *
  * This code is crying out for a shrinker per pool....
  */
 static unsigned long
@@ -407,8 +411,8 @@ ttm_pool_shrink_scan(struct shrinker *shrink, struct shrink_control *sc)
 		if (shrink_pages == 0)
 			break;
 		pool = &_manager->pools[(i + pool_offset)%NUM_POOLS];
-		shrink_pages = ttm_page_pool_free(pool, nr_free,
-						  sc->gfp_mask);
+		/* OK to use static buffer since global mutex is held. */
+		shrink_pages = ttm_page_pool_free(pool, nr_free, true);
 		freed += nr_free - shrink_pages;
 	}
 	mutex_unlock(&lock);
@@ -710,7 +714,7 @@ static void ttm_put_pages(struct page **pages, unsigned npages, int flags,
 	}
 	spin_unlock_irqrestore(&pool->lock, irq_flags);
 	if (npages)
-		ttm_page_pool_free(pool, npages, GFP_KERNEL);
+		ttm_page_pool_free(pool, npages, false);
 }
 
 /*
@@ -849,9 +853,9 @@ void ttm_page_alloc_fini(void)
 	pr_info("Finalizing pool allocator\n");
 	ttm_pool_mm_shrink_fini(_manager);
 
+	/* OK to use static buffer since global mutex is no longer used. */
 	for (i = 0; i < NUM_POOLS; ++i)
-		ttm_page_pool_free(&_manager->pools[i], FREE_ALL_PAGES,
-				   GFP_KERNEL);
+		ttm_page_pool_free(&_manager->pools[i], FREE_ALL_PAGES, true);
 
 	kobject_put(&_manager->kobj);
 	_manager = NULL;
diff --git a/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c b/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
index c96db433f8af..01e1d27eb078 100644
--- a/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
+++ b/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
@@ -411,11 +411,12 @@ static void ttm_dma_page_put(struct dma_pool *pool, struct dma_page *d_page)
  *
  * @pool: to free the pages from
  * @nr_free: If set to true will free all pages in pool
- * @gfp: GFP flags.
+ * @use_static: Safe to use static buffer
  **/
 static unsigned ttm_dma_page_pool_free(struct dma_pool *pool, unsigned nr_free,
-				       gfp_t gfp)
+				       bool use_static)
 {
+	static struct page *static_buf[NUM_PAGES_TO_ALLOC];
 	unsigned long irq_flags;
 	struct dma_page *dma_p, *tmp;
 	struct page **pages_to_free;
@@ -432,7 +433,11 @@ static unsigned ttm_dma_page_pool_free(struct dma_pool *pool, unsigned nr_free,
 			 npages_to_free, nr_free);
 	}
 #endif
-	pages_to_free = kmalloc(npages_to_free * sizeof(struct page *), gfp);
+	if (use_static)
+		pages_to_free = static_buf;
+	else
+		pages_to_free = kmalloc(npages_to_free * sizeof(struct page *),
+					GFP_KERNEL);
 
 	if (!pages_to_free) {
 		pr_err("%s: Failed to allocate memory for pool free operation\n",
@@ -502,7 +507,8 @@ restart:
 	if (freed_pages)
 		ttm_dma_pages_put(pool, &d_pages, pages_to_free, freed_pages);
 out:
-	kfree(pages_to_free);
+	if (pages_to_free != static_buf)
+		kfree(pages_to_free);
 	return nr_free;
 }
 
@@ -531,7 +537,8 @@ static void ttm_dma_free_pool(struct device *dev, enum pool_type type)
 		if (pool->type != type)
 			continue;
 		/* Takes a spinlock.. */
-		ttm_dma_page_pool_free(pool, FREE_ALL_PAGES, GFP_KERNEL);
+		/* OK to use static buffer since global mutex is held. */
+		ttm_dma_page_pool_free(pool, FREE_ALL_PAGES, true);
 		WARN_ON(((pool->npages_in_use + pool->npages_free) != 0));
 		/* This code path is called after _all_ references to the
 		 * struct device has been dropped - so nobody should be
@@ -986,7 +993,7 @@ void ttm_dma_unpopulate(struct ttm_dma_tt *ttm_dma, struct device *dev)
 
 	/* shrink pool if necessary (only on !is_cached pools)*/
 	if (npages)
-		ttm_dma_page_pool_free(pool, npages, GFP_KERNEL);
+		ttm_dma_page_pool_free(pool, npages, false);
 	ttm->state = tt_unpopulated;
 }
 EXPORT_SYMBOL_GPL(ttm_dma_unpopulate);
@@ -996,8 +1003,6 @@ EXPORT_SYMBOL_GPL(ttm_dma_unpopulate);
  *
  * XXX: (dchinner) Deadlock warning!
  *
- * We need to pass sc->gfp_mask to ttm_dma_page_pool_free().
- *
  * I'm getting sadder as I hear more pathetical whimpers about needing per-pool
  * shrinkers
  */
@@ -1030,8 +1035,8 @@ ttm_dma_pool_shrink_scan(struct shrinker *shrink, struct shrink_control *sc)
 		if (++idx < pool_offset)
 			continue;
 		nr_free = shrink_pages;
-		shrink_pages = ttm_dma_page_pool_free(p->pool, nr_free,
-						      sc->gfp_mask);
+		/* OK to use static buffer since global mutex is held. */
+		shrink_pages = ttm_dma_page_pool_free(p->pool, nr_free, true);
 		freed += nr_free - shrink_pages;
 
 		pr_debug("%s: (%s:%d) Asked to shrink %d, have %d more to go\n",
diff --git a/drivers/gpu/drm/udl/Kconfig b/drivers/gpu/drm/udl/Kconfig
index f02528686cd5..613ab0622d6e 100644
--- a/drivers/gpu/drm/udl/Kconfig
+++ b/drivers/gpu/drm/udl/Kconfig
@@ -1,8 +1,9 @@
 config DRM_UDL
 	tristate "DisplayLink"
 	depends on DRM
+	depends on USB_SUPPORT
 	depends on USB_ARCH_HAS_HCD
-	select DRM_USB
+	select USB
 	select FB_SYS_FILLRECT
 	select FB_SYS_COPYAREA
 	select FB_SYS_IMAGEBLIT
diff --git a/drivers/gpu/drm/udl/Makefile b/drivers/gpu/drm/udl/Makefile
index 05c7481bfd40..195bcac0b6c8 100644
--- a/drivers/gpu/drm/udl/Makefile
+++ b/drivers/gpu/drm/udl/Makefile
@@ -1,6 +1,6 @@
 
 ccflags-y := -Iinclude/drm
 
-udl-y := udl_drv.o udl_modeset.o udl_connector.o udl_encoder.o udl_main.o udl_fb.o udl_transfer.o udl_gem.o
+udl-y := udl_drv.o udl_modeset.o udl_connector.o udl_encoder.o udl_main.o udl_fb.o udl_transfer.o udl_gem.o udl_dmabuf.o
 
 obj-$(CONFIG_DRM_UDL) := udl.o
diff --git a/drivers/gpu/drm/udl/udl_connector.c b/drivers/gpu/drm/udl/udl_connector.c
index e026a9e2942a..0110d95522f3 100644
--- a/drivers/gpu/drm/udl/udl_connector.c
+++ b/drivers/gpu/drm/udl/udl_connector.c
@@ -34,8 +34,8 @@ static u8 *udl_get_edid(struct udl_device *udl)
 		goto error;
 
 	for (i = 0; i < EDID_LENGTH; i++) {
-		ret = usb_control_msg(udl->ddev->usbdev,
-				      usb_rcvctrlpipe(udl->ddev->usbdev, 0), (0x02),
+		ret = usb_control_msg(udl->udev,
+				      usb_rcvctrlpipe(udl->udev, 0), (0x02),
 				      (0x80 | (0x02 << 5)), i << 8, 0xA1, rbuf, 2,
 				      HZ);
 		if (ret < 1) {
diff --git a/drivers/gpu/drm/udl/udl_dmabuf.c b/drivers/gpu/drm/udl/udl_dmabuf.c
new file mode 100644
index 000000000000..ac8a66b4dfc2
--- /dev/null
+++ b/drivers/gpu/drm/udl/udl_dmabuf.c
@@ -0,0 +1,276 @@
+/*
+ * udl_dmabuf.c
+ *
+ * Copyright (c) 2014 The Chromium OS Authors
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program. If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <drm/drmP.h>
+#include "udl_drv.h"
+#include <linux/shmem_fs.h>
+#include <linux/dma-buf.h>
+
+struct udl_drm_dmabuf_attachment {
+	struct sg_table sgt;
+	enum dma_data_direction dir;
+	bool is_mapped;
+};
+
+static int udl_attach_dma_buf(struct dma_buf *dmabuf,
+			      struct device *dev,
+			      struct dma_buf_attachment *attach)
+{
+	struct udl_drm_dmabuf_attachment *udl_attach;
+
+	DRM_DEBUG_PRIME("[DEV:%s] size:%zd\n", dev_name(attach->dev),
+			attach->dmabuf->size);
+
+	udl_attach = kzalloc(sizeof(*udl_attach), GFP_KERNEL);
+	if (!udl_attach)
+		return -ENOMEM;
+
+	udl_attach->dir = DMA_NONE;
+	attach->priv = udl_attach;
+
+	return 0;
+}
+
+static void udl_detach_dma_buf(struct dma_buf *dmabuf,
+			       struct dma_buf_attachment *attach)
+{
+	struct udl_drm_dmabuf_attachment *udl_attach = attach->priv;
+	struct sg_table *sgt;
+
+	if (!udl_attach)
+		return;
+
+	DRM_DEBUG_PRIME("[DEV:%s] size:%zd\n", dev_name(attach->dev),
+			attach->dmabuf->size);
+
+	sgt = &udl_attach->sgt;
+
+	if (udl_attach->dir != DMA_NONE)
+		dma_unmap_sg(attach->dev, sgt->sgl, sgt->nents,
+				udl_attach->dir);
+
+	sg_free_table(sgt);
+	kfree(udl_attach);
+	attach->priv = NULL;
+}
+
+static struct sg_table *udl_map_dma_buf(struct dma_buf_attachment *attach,
+					enum dma_data_direction dir)
+{
+	struct udl_drm_dmabuf_attachment *udl_attach = attach->priv;
+	struct udl_gem_object *obj = to_udl_bo(attach->dmabuf->priv);
+	struct drm_device *dev = obj->base.dev;
+	struct scatterlist *rd, *wr;
+	struct sg_table *sgt = NULL;
+	unsigned int i;
+	int page_count;
+	int nents, ret;
+
+	DRM_DEBUG_PRIME("[DEV:%s] size:%zd dir=%d\n", dev_name(attach->dev),
+			attach->dmabuf->size, dir);
+
+	/* just return current sgt if already requested. */
+	if (udl_attach->dir == dir && udl_attach->is_mapped)
+		return &udl_attach->sgt;
+
+	if (!obj->pages) {
+		ret = udl_gem_get_pages(obj);
+		if (ret) {
+			DRM_ERROR("failed to map pages.\n");
+			return ERR_PTR(ret);
+		}
+	}
+
+	page_count = obj->base.size / PAGE_SIZE;
+	obj->sg = drm_prime_pages_to_sg(obj->pages, page_count);
+	if (IS_ERR(obj->sg)) {
+		DRM_ERROR("failed to allocate sgt.\n");
+		return ERR_CAST(obj->sg);
+	}
+
+	sgt = &udl_attach->sgt;
+
+	ret = sg_alloc_table(sgt, obj->sg->orig_nents, GFP_KERNEL);
+	if (ret) {
+		DRM_ERROR("failed to alloc sgt.\n");
+		return ERR_PTR(-ENOMEM);
+	}
+
+	mutex_lock(&dev->struct_mutex);
+
+	rd = obj->sg->sgl;
+	wr = sgt->sgl;
+	for (i = 0; i < sgt->orig_nents; ++i) {
+		sg_set_page(wr, sg_page(rd), rd->length, rd->offset);
+		rd = sg_next(rd);
+		wr = sg_next(wr);
+	}
+
+	if (dir != DMA_NONE) {
+		nents = dma_map_sg(attach->dev, sgt->sgl, sgt->orig_nents, dir);
+		if (!nents) {
+			DRM_ERROR("failed to map sgl with iommu.\n");
+			sg_free_table(sgt);
+			sgt = ERR_PTR(-EIO);
+			goto err_unlock;
+		}
+	}
+
+	udl_attach->is_mapped = true;
+	udl_attach->dir = dir;
+	attach->priv = udl_attach;
+
+err_unlock:
+	mutex_unlock(&dev->struct_mutex);
+	return sgt;
+}
+
+static void udl_unmap_dma_buf(struct dma_buf_attachment *attach,
+			      struct sg_table *sgt,
+			      enum dma_data_direction dir)
+{
+	/* Nothing to do. */
+	DRM_DEBUG_PRIME("[DEV:%s] size:%zd dir:%d\n", dev_name(attach->dev),
+			attach->dmabuf->size, dir);
+}
+
+static void *udl_dmabuf_kmap(struct dma_buf *dma_buf, unsigned long page_num)
+{
+	/* TODO */
+
+	return NULL;
+}
+
+static void *udl_dmabuf_kmap_atomic(struct dma_buf *dma_buf,
+				    unsigned long page_num)
+{
+	/* TODO */
+
+	return NULL;
+}
+
+static void udl_dmabuf_kunmap(struct dma_buf *dma_buf,
+			      unsigned long page_num, void *addr)
+{
+	/* TODO */
+}
+
+static void udl_dmabuf_kunmap_atomic(struct dma_buf *dma_buf,
+				     unsigned long page_num,
+				     void *addr)
+{
+	/* TODO */
+}
+
+static int udl_dmabuf_mmap(struct dma_buf *dma_buf,
+			   struct vm_area_struct *vma)
+{
+	/* TODO */
+
+	return -EINVAL;
+}
+
+static struct dma_buf_ops udl_dmabuf_ops = {
+	.attach			= udl_attach_dma_buf,
+	.detach			= udl_detach_dma_buf,
+	.map_dma_buf		= udl_map_dma_buf,
+	.unmap_dma_buf		= udl_unmap_dma_buf,
+	.kmap			= udl_dmabuf_kmap,
+	.kmap_atomic		= udl_dmabuf_kmap_atomic,
+	.kunmap			= udl_dmabuf_kunmap,
+	.kunmap_atomic		= udl_dmabuf_kunmap_atomic,
+	.mmap			= udl_dmabuf_mmap,
+	.release		= drm_gem_dmabuf_release,
+};
+
+struct dma_buf *udl_gem_prime_export(struct drm_device *dev,
+				     struct drm_gem_object *obj, int flags)
+{
+	return dma_buf_export(obj, &udl_dmabuf_ops, obj->size, flags, NULL);
+}
+
+static int udl_prime_create(struct drm_device *dev,
+			    size_t size,
+			    struct sg_table *sg,
+			    struct udl_gem_object **obj_p)
+{
+	struct udl_gem_object *obj;
+	int npages;
+
+	npages = size / PAGE_SIZE;
+
+	*obj_p = NULL;
+	obj = udl_gem_alloc_object(dev, npages * PAGE_SIZE);
+	if (!obj)
+		return -ENOMEM;
+
+	obj->sg = sg;
+	obj->pages = drm_malloc_ab(npages, sizeof(struct page *));
+	if (obj->pages == NULL) {
+		DRM_ERROR("obj pages is NULL %d\n", npages);
+		return -ENOMEM;
+	}
+
+	drm_prime_sg_to_page_addr_arrays(sg, obj->pages, NULL, npages);
+
+	*obj_p = obj;
+	return 0;
+}
+
+struct drm_gem_object *udl_gem_prime_import(struct drm_device *dev,
+				struct dma_buf *dma_buf)
+{
+	struct dma_buf_attachment *attach;
+	struct sg_table *sg;
+	struct udl_gem_object *uobj;
+	int ret;
+
+	/* need to attach */
+	get_device(dev->dev);
+	attach = dma_buf_attach(dma_buf, dev->dev);
+	if (IS_ERR(attach)) {
+		put_device(dev->dev);
+		return ERR_CAST(attach);
+	}
+
+	get_dma_buf(dma_buf);
+
+	sg = dma_buf_map_attachment(attach, DMA_BIDIRECTIONAL);
+	if (IS_ERR(sg)) {
+		ret = PTR_ERR(sg);
+		goto fail_detach;
+	}
+
+	ret = udl_prime_create(dev, dma_buf->size, sg, &uobj);
+	if (ret)
+		goto fail_unmap;
+
+	uobj->base.import_attach = attach;
+	uobj->flags = UDL_BO_WC;
+
+	return &uobj->base;
+
+fail_unmap:
+	dma_buf_unmap_attachment(attach, sg, DMA_BIDIRECTIONAL);
+fail_detach:
+	dma_buf_detach(dma_buf, attach);
+	dma_buf_put(dma_buf);
+	put_device(dev->dev);
+	return ERR_PTR(ret);
+}
diff --git a/drivers/gpu/drm/udl/udl_drv.c b/drivers/gpu/drm/udl/udl_drv.c
index 3ddd6cd98ac1..d5728ec85254 100644
--- a/drivers/gpu/drm/udl/udl_drv.c
+++ b/drivers/gpu/drm/udl/udl_drv.c
@@ -7,48 +7,13 @@
  */
 
 #include <linux/module.h>
-#include <drm/drm_usb.h>
+#include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
 #include "udl_drv.h"
 
-static struct drm_driver driver;
-
-/*
- * There are many DisplayLink-based graphics products, all with unique PIDs.
- * So we match on DisplayLink's VID + Vendor-Defined Interface Class (0xff)
- * We also require a match on SubClass (0x00) and Protocol (0x00),
- * which is compatible with all known USB 2.0 era graphics chips and firmware,
- * but allows DisplayLink to increment those for any future incompatible chips
- */
-static struct usb_device_id id_table[] = {
-	{.idVendor = 0x17e9, .bInterfaceClass = 0xff,
-	 .bInterfaceSubClass = 0x00,
-	 .bInterfaceProtocol = 0x00,
-	 .match_flags = USB_DEVICE_ID_MATCH_VENDOR |
-			USB_DEVICE_ID_MATCH_INT_CLASS |
-			USB_DEVICE_ID_MATCH_INT_SUBCLASS |
-			USB_DEVICE_ID_MATCH_INT_PROTOCOL,},
-	{},
-};
-MODULE_DEVICE_TABLE(usb, id_table);
-
-MODULE_LICENSE("GPL");
-
-static int udl_usb_probe(struct usb_interface *interface,
-			 const struct usb_device_id *id)
+static int udl_driver_set_busid(struct drm_device *d, struct drm_master *m)
 {
-	return drm_get_usb_dev(interface, id, &driver);
-}
-
-static void udl_usb_disconnect(struct usb_interface *interface)
-{
-	struct drm_device *dev = usb_get_intfdata(interface);
-
-	drm_kms_helper_poll_disable(dev);
-	drm_connector_unplug_all(dev);
-	udl_fbdev_unplug(dev);
-	udl_drop_usb(dev);
-	drm_unplug_dev(dev);
+	return 0;
 }
 
 static const struct vm_operations_struct udl_gem_vm_ops = {
@@ -75,6 +40,7 @@ static struct drm_driver driver = {
 	.driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_PRIME,
 	.load = udl_driver_load,
 	.unload = udl_driver_unload,
+	.set_busid = udl_driver_set_busid,
 
 	/* gem hooks */
 	.gem_free_object = udl_gem_free_object,
@@ -85,7 +51,9 @@ static struct drm_driver driver = {
 	.dumb_destroy = drm_gem_dumb_destroy,
 	.fops = &udl_driver_fops,
 
+	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
 	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
+	.gem_prime_export = udl_gem_prime_export,
 	.gem_prime_import = udl_gem_prime_import,
 
 	.name = DRIVER_NAME,
@@ -96,6 +64,61 @@ static struct drm_driver driver = {
 	.patchlevel = DRIVER_PATCHLEVEL,
 };
 
+static int udl_usb_probe(struct usb_interface *interface,
+			 const struct usb_device_id *id)
+{
+	struct usb_device *udev = interface_to_usbdev(interface);
+	struct drm_device *dev;
+	int r;
+
+	dev = drm_dev_alloc(&driver, &interface->dev);
+	if (!dev)
+		return -ENOMEM;
+
+	r = drm_dev_register(dev, (unsigned long)udev);
+	if (r)
+		goto err_free;
+
+	usb_set_intfdata(interface, dev);
+	DRM_INFO("Initialized udl on minor %d\n", dev->primary->index);
+
+	return 0;
+
+err_free:
+	drm_dev_unref(dev);
+	return r;
+}
+
+static void udl_usb_disconnect(struct usb_interface *interface)
+{
+	struct drm_device *dev = usb_get_intfdata(interface);
+
+	drm_kms_helper_poll_disable(dev);
+	drm_connector_unplug_all(dev);
+	udl_fbdev_unplug(dev);
+	udl_drop_usb(dev);
+	drm_unplug_dev(dev);
+}
+
+/*
+ * There are many DisplayLink-based graphics products, all with unique PIDs.
+ * So we match on DisplayLink's VID + Vendor-Defined Interface Class (0xff)
+ * We also require a match on SubClass (0x00) and Protocol (0x00),
+ * which is compatible with all known USB 2.0 era graphics chips and firmware,
+ * but allows DisplayLink to increment those for any future incompatible chips
+ */
+static struct usb_device_id id_table[] = {
+	{.idVendor = 0x17e9, .bInterfaceClass = 0xff,
+	 .bInterfaceSubClass = 0x00,
+	 .bInterfaceProtocol = 0x00,
+	 .match_flags = USB_DEVICE_ID_MATCH_VENDOR |
+			USB_DEVICE_ID_MATCH_INT_CLASS |
+			USB_DEVICE_ID_MATCH_INT_SUBCLASS |
+			USB_DEVICE_ID_MATCH_INT_PROTOCOL,},
+	{},
+};
+MODULE_DEVICE_TABLE(usb, id_table);
+
 static struct usb_driver udl_driver = {
 	.name = "udl",
 	.probe = udl_usb_probe,
@@ -105,13 +128,14 @@ static struct usb_driver udl_driver = {
 
 static int __init udl_init(void)
 {
-	return drm_usb_init(&driver, &udl_driver);
+	return usb_register(&udl_driver);
 }
 
 static void __exit udl_exit(void)
 {
-	drm_usb_exit(&driver, &udl_driver);
+	usb_deregister(&udl_driver);
 }
 
 module_init(udl_init);
 module_exit(udl_exit);
+MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/udl/udl_drv.h b/drivers/gpu/drm/udl/udl_drv.h
index 1fbf7b357f16..80adbac82bde 100644
--- a/drivers/gpu/drm/udl/udl_drv.h
+++ b/drivers/gpu/drm/udl/udl_drv.h
@@ -15,6 +15,7 @@
 #define UDL_DRV_H
 
 #include <linux/usb.h>
+#include <drm/drm_gem.h>
 
 #define DRIVER_NAME		"udl"
 #define DRIVER_DESC		"DisplayLink"
@@ -24,6 +25,9 @@
 #define DRIVER_MINOR		0
 #define DRIVER_PATCHLEVEL	1
 
+#define UDL_BO_CACHEABLE		(1 << 0)
+#define UDL_BO_WC		(1 << 1)
+
 struct udl_device;
 
 struct urb_node {
@@ -47,6 +51,7 @@ struct udl_fbdev;
 struct udl_device {
 	struct device *dev;
 	struct drm_device *ddev;
+	struct usb_device *udev;
 
 	int sku_pixel_limit;
 
@@ -67,6 +72,7 @@ struct udl_gem_object {
 	struct page **pages;
 	void *vmapping;
 	struct sg_table *sg;
+	unsigned int flags;
 };
 
 #define to_udl_bo(x) container_of(x, struct udl_gem_object, base)
@@ -118,9 +124,13 @@ int udl_gem_mmap(struct drm_file *file_priv, struct drm_device *dev,
 void udl_gem_free_object(struct drm_gem_object *gem_obj);
 struct udl_gem_object *udl_gem_alloc_object(struct drm_device *dev,
 					    size_t size);
+struct dma_buf *udl_gem_prime_export(struct drm_device *dev,
+				     struct drm_gem_object *obj, int flags);
 struct drm_gem_object *udl_gem_prime_import(struct drm_device *dev,
 				struct dma_buf *dma_buf);
 
+int udl_gem_get_pages(struct udl_gem_object *obj);
+void udl_gem_put_pages(struct udl_gem_object *obj);
 int udl_gem_vmap(struct udl_gem_object *obj);
 void udl_gem_vunmap(struct udl_gem_object *obj);
 int udl_drm_gem_mmap(struct file *filp, struct vm_area_struct *vma);
diff --git a/drivers/gpu/drm/udl/udl_fb.c b/drivers/gpu/drm/udl/udl_fb.c
index d1da339843ca..8cbcb4589bd3 100644
--- a/drivers/gpu/drm/udl/udl_fb.c
+++ b/drivers/gpu/drm/udl/udl_fb.c
@@ -472,7 +472,8 @@ udl_framebuffer_init(struct drm_device *dev,
 static int udlfb_create(struct drm_fb_helper *helper,
 			struct drm_fb_helper_surface_size *sizes)
 {
-	struct udl_fbdev *ufbdev = (struct udl_fbdev *)helper;
+	struct udl_fbdev *ufbdev =
+		container_of(helper, struct udl_fbdev, helper);
 	struct drm_device *dev = ufbdev->helper.dev;
 	struct fb_info *info;
 	struct device *device = dev->dev;
diff --git a/drivers/gpu/drm/udl/udl_gem.c b/drivers/gpu/drm/udl/udl_gem.c
index 8044f5fb7c49..2a0a784ab6ee 100644
--- a/drivers/gpu/drm/udl/udl_gem.c
+++ b/drivers/gpu/drm/udl/udl_gem.c
@@ -25,6 +25,7 @@ struct udl_gem_object *udl_gem_alloc_object(struct drm_device *dev,
 		return NULL;
 	}
 
+	obj->flags = UDL_BO_CACHEABLE;
 	return obj;
 }
 
@@ -56,6 +57,23 @@ udl_gem_create(struct drm_file *file,
 	return 0;
 }
 
+static void update_vm_cache_attr(struct udl_gem_object *obj,
+				 struct vm_area_struct *vma)
+{
+	DRM_DEBUG_KMS("flags = 0x%x\n", obj->flags);
+
+	/* non-cacheable as default. */
+	if (obj->flags & UDL_BO_CACHEABLE) {
+		vma->vm_page_prot = vm_get_page_prot(vma->vm_flags);
+	} else if (obj->flags & UDL_BO_WC) {
+		vma->vm_page_prot =
+			pgprot_writecombine(vm_get_page_prot(vma->vm_flags));
+	} else {
+		vma->vm_page_prot =
+			pgprot_noncached(vm_get_page_prot(vma->vm_flags));
+	}
+}
+
 int udl_dumb_create(struct drm_file *file,
 		    struct drm_device *dev,
 		    struct drm_mode_create_dumb *args)
@@ -77,6 +95,8 @@ int udl_drm_gem_mmap(struct file *filp, struct vm_area_struct *vma)
 	vma->vm_flags &= ~VM_PFNMAP;
 	vma->vm_flags |= VM_MIXEDMAP;
 
+	update_vm_cache_attr(to_udl_bo(vma->vm_private_data), vma);
+
 	return ret;
 }
 
@@ -107,7 +127,7 @@ int udl_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	}
 }
 
-static int udl_gem_get_pages(struct udl_gem_object *obj)
+int udl_gem_get_pages(struct udl_gem_object *obj)
 {
 	struct page **pages;
 
@@ -123,7 +143,7 @@ static int udl_gem_get_pages(struct udl_gem_object *obj)
 	return 0;
 }
 
-static void udl_gem_put_pages(struct udl_gem_object *obj)
+void udl_gem_put_pages(struct udl_gem_object *obj)
 {
 	if (obj->base.import_attach) {
 		drm_free_large(obj->pages);
@@ -164,8 +184,7 @@ void udl_gem_vunmap(struct udl_gem_object *obj)
 		return;
 	}
 
-	if (obj->vmapping)
-		vunmap(obj->vmapping);
+	vunmap(obj->vmapping);
 
 	udl_gem_put_pages(obj);
 }
@@ -220,73 +239,3 @@ unlock:
 	mutex_unlock(&dev->struct_mutex);
 	return ret;
 }
-
-static int udl_prime_create(struct drm_device *dev,
-			    size_t size,
-			    struct sg_table *sg,
-			    struct udl_gem_object **obj_p)
-{
-	struct udl_gem_object *obj;
-	int npages;
-
-	npages = size / PAGE_SIZE;
-
-	*obj_p = NULL;
-	obj = udl_gem_alloc_object(dev, npages * PAGE_SIZE);
-	if (!obj)
-		return -ENOMEM;
-
-	obj->sg = sg;
-	obj->pages = drm_malloc_ab(npages, sizeof(struct page *));
-	if (obj->pages == NULL) {
-		DRM_ERROR("obj pages is NULL %d\n", npages);
-		return -ENOMEM;
-	}
-
-	drm_prime_sg_to_page_addr_arrays(sg, obj->pages, NULL, npages);
-
-	*obj_p = obj;
-	return 0;
-}
-
-struct drm_gem_object *udl_gem_prime_import(struct drm_device *dev,
-				struct dma_buf *dma_buf)
-{
-	struct dma_buf_attachment *attach;
-	struct sg_table *sg;
-	struct udl_gem_object *uobj;
-	int ret;
-
-	/* need to attach */
-	get_device(dev->dev);
-	attach = dma_buf_attach(dma_buf, dev->dev);
-	if (IS_ERR(attach)) {
-		put_device(dev->dev);
-		return ERR_CAST(attach);
-	}
-
-	get_dma_buf(dma_buf);
-
-	sg = dma_buf_map_attachment(attach, DMA_BIDIRECTIONAL);
-	if (IS_ERR(sg)) {
-		ret = PTR_ERR(sg);
-		goto fail_detach;
-	}
-
-	ret = udl_prime_create(dev, dma_buf->size, sg, &uobj);
-	if (ret) {
-		goto fail_unmap;
-	}
-
-	uobj->base.import_attach = attach;
-
-	return &uobj->base;
-
-fail_unmap:
-	dma_buf_unmap_attachment(attach, sg, DMA_BIDIRECTIONAL);
-fail_detach:
-	dma_buf_detach(dma_buf, attach);
-	dma_buf_put(dma_buf);
-	put_device(dev->dev);
-	return ERR_PTR(ret);
-}
diff --git a/drivers/gpu/drm/udl/udl_main.c b/drivers/gpu/drm/udl/udl_main.c
index 42795674bc07..33dbfb2c4748 100644
--- a/drivers/gpu/drm/udl/udl_main.c
+++ b/drivers/gpu/drm/udl/udl_main.c
@@ -202,7 +202,7 @@ static int udl_alloc_urb_list(struct drm_device *dev, int count, size_t size)
 		}
 		unode->urb = urb;
 
-		buf = usb_alloc_coherent(udl->ddev->usbdev, MAX_TRANSFER, GFP_KERNEL,
+		buf = usb_alloc_coherent(udl->udev, MAX_TRANSFER, GFP_KERNEL,
 					 &urb->transfer_dma);
 		if (!buf) {
 			kfree(unode);
@@ -211,7 +211,7 @@ static int udl_alloc_urb_list(struct drm_device *dev, int count, size_t size)
 		}
 
 		/* urb->transfer_buffer_length set to actual before submit */
-		usb_fill_bulk_urb(urb, udl->ddev->usbdev, usb_sndbulkpipe(udl->ddev->usbdev, 1),
+		usb_fill_bulk_urb(urb, udl->udev, usb_sndbulkpipe(udl->udev, 1),
 			buf, size, udl_urb_completion, unode);
 		urb->transfer_flags |= URB_NO_TRANSFER_DMA_MAP;
 
@@ -282,6 +282,7 @@ int udl_submit_urb(struct drm_device *dev, struct urb *urb, size_t len)
 
 int udl_driver_load(struct drm_device *dev, unsigned long flags)
 {
+	struct usb_device *udev = (void*)flags;
 	struct udl_device *udl;
 	int ret = -ENOMEM;
 
@@ -290,10 +291,11 @@ int udl_driver_load(struct drm_device *dev, unsigned long flags)
 	if (!udl)
 		return -ENOMEM;
 
+	udl->udev = udev;
 	udl->ddev = dev;
 	dev->dev_private = udl;
 
-	if (!udl_parse_vendor_descriptor(dev, dev->usbdev)) {
+	if (!udl_parse_vendor_descriptor(dev, udl->udev)) {
 		ret = -ENODEV;
 		DRM_ERROR("firmware not recognized. Assume incompatible device\n");
 		goto err;
diff --git a/drivers/gpu/drm/udl/udl_modeset.c b/drivers/gpu/drm/udl/udl_modeset.c
index dc145d320b25..1701f1dfb23f 100644
--- a/drivers/gpu/drm/udl/udl_modeset.c
+++ b/drivers/gpu/drm/udl/udl_modeset.c
@@ -14,6 +14,7 @@
 #include <drm/drmP.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_plane_helper.h>
 #include "udl_drv.h"
 
 /*
diff --git a/drivers/gpu/drm/via/via_dma.c b/drivers/gpu/drm/via/via_dma.c
index 6fc0648dd37f..d17d8f245c1a 100644
--- a/drivers/gpu/drm/via/via_dma.c
+++ b/drivers/gpu/drm/via/via_dma.c
@@ -161,7 +161,7 @@ int via_dma_cleanup(struct drm_device *dev)
 		if (dev_priv->ring.virtual_start) {
 			via_cmdbuf_reset(dev_priv);
 
-			drm_core_ioremapfree(&dev_priv->ring.map, dev);
+			drm_legacy_ioremapfree(&dev_priv->ring.map, dev);
 			dev_priv->ring.virtual_start = NULL;
 		}
 
@@ -200,7 +200,7 @@ static int via_initialize(struct drm_device *dev,
 	dev_priv->ring.map.flags = 0;
 	dev_priv->ring.map.mtrr = 0;
 
-	drm_core_ioremap(&dev_priv->ring.map, dev);
+	drm_legacy_ioremap(&dev_priv->ring.map, dev);
 
 	if (dev_priv->ring.map.handle == NULL) {
 		via_dma_cleanup(dev);
diff --git a/drivers/gpu/drm/via/via_drv.c b/drivers/gpu/drm/via/via_drv.c
index 50abc2adfaee..ed8aa8ff861a 100644
--- a/drivers/gpu/drm/via/via_drv.c
+++ b/drivers/gpu/drm/via/via_drv.c
@@ -62,7 +62,7 @@ static const struct file_operations via_driver_fops = {
 	.open = drm_open,
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_mmap,
+	.mmap = drm_legacy_mmap,
 	.poll = drm_poll,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl = drm_compat_ioctl,
@@ -79,6 +79,7 @@ static struct drm_driver driver = {
 	.open = via_driver_open,
 	.preclose = via_reclaim_buffers_locked,
 	.postclose = via_driver_postclose,
+	.set_busid = drm_pci_set_busid,
 	.context_dtor = via_final_context,
 	.get_vblank_counter = via_get_vblank_counter,
 	.enable_vblank = via_enable_vblank,
diff --git a/drivers/gpu/drm/via/via_drv.h b/drivers/gpu/drm/via/via_drv.h
index ad0273256beb..ef8c500b4a00 100644
--- a/drivers/gpu/drm/via/via_drv.h
+++ b/drivers/gpu/drm/via/via_drv.h
@@ -25,6 +25,8 @@
 #define _VIA_DRV_H_
 
 #include <drm/drm_mm.h>
+#include <drm/drm_legacy.h>
+
 #define DRIVER_AUTHOR	"Various"
 
 #define DRIVER_NAME		"via"
diff --git a/drivers/gpu/drm/via/via_map.c b/drivers/gpu/drm/via/via_map.c
index d0ab3fb32acd..0b3522dba6e8 100644
--- a/drivers/gpu/drm/via/via_map.c
+++ b/drivers/gpu/drm/via/via_map.c
@@ -31,7 +31,7 @@ static int via_do_init_map(struct drm_device *dev, drm_via_init_t *init)
 
 	DRM_DEBUG("\n");
 
-	dev_priv->sarea = drm_getsarea(dev);
+	dev_priv->sarea = drm_legacy_getsarea(dev);
 	if (!dev_priv->sarea) {
 		DRM_ERROR("could not find sarea!\n");
 		dev->dev_private = (void *)dev_priv;
@@ -39,14 +39,14 @@ static int via_do_init_map(struct drm_device *dev, drm_via_init_t *init)
 		return -EINVAL;
 	}
 
-	dev_priv->fb = drm_core_findmap(dev, init->fb_offset);
+	dev_priv->fb = drm_legacy_findmap(dev, init->fb_offset);
 	if (!dev_priv->fb) {
 		DRM_ERROR("could not find framebuffer!\n");
 		dev->dev_private = (void *)dev_priv;
 		via_do_cleanup_map(dev);
 		return -EINVAL;
 	}
-	dev_priv->mmio = drm_core_findmap(dev, init->mmio_offset);
+	dev_priv->mmio = drm_legacy_findmap(dev, init->mmio_offset);
 	if (!dev_priv->mmio) {
 		DRM_ERROR("could not find mmio region!\n");
 		dev->dev_private = (void *)dev_priv;
diff --git a/drivers/gpu/drm/via/via_mm.c b/drivers/gpu/drm/via/via_mm.c
index d70b1e1544bf..4f20742e7788 100644
--- a/drivers/gpu/drm/via/via_mm.c
+++ b/drivers/gpu/drm/via/via_mm.c
@@ -211,12 +211,12 @@ void via_reclaim_buffers_locked(struct drm_device *dev,
 	if (!(file->minor->master && file->master->lock.hw_lock))
 		return;
 
-	drm_idlelock_take(&file->master->lock);
+	drm_legacy_idlelock_take(&file->master->lock);
 
 	mutex_lock(&dev->struct_mutex);
 	if (list_empty(&file_priv->obj_list)) {
 		mutex_unlock(&dev->struct_mutex);
-		drm_idlelock_release(&file->master->lock);
+		drm_legacy_idlelock_release(&file->master->lock);
 
 		return;
 	}
@@ -231,7 +231,7 @@ void via_reclaim_buffers_locked(struct drm_device *dev,
 	}
 	mutex_unlock(&dev->struct_mutex);
 
-	drm_idlelock_release(&file->master->lock);
+	drm_legacy_idlelock_release(&file->master->lock);
 
 	return;
 }
diff --git a/drivers/gpu/drm/via/via_verifier.c b/drivers/gpu/drm/via/via_verifier.c
index 9dbc92bd1512..0677bbf4ec7e 100644
--- a/drivers/gpu/drm/via/via_verifier.c
+++ b/drivers/gpu/drm/via/via_verifier.c
@@ -31,6 +31,7 @@
 #include "via_3d_reg.h"
 #include <drm/drmP.h>
 #include <drm/via_drm.h>
+#include <drm/drm_legacy.h>
 #include "via_verifier.h"
 #include "via_drv.h"
 
diff --git a/drivers/gpu/drm/vmwgfx/svga_reg.h b/drivers/gpu/drm/vmwgfx/svga_reg.h
index 11323dd5196f..e4259c2c1acc 100644
--- a/drivers/gpu/drm/vmwgfx/svga_reg.h
+++ b/drivers/gpu/drm/vmwgfx/svga_reg.h
@@ -35,7 +35,6 @@
 /*
  * PCI device IDs.
  */
-#define PCI_VENDOR_ID_VMWARE            0x15AD
 #define PCI_DEVICE_ID_VMWARE_SVGA2      0x0405
 
 /*
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c b/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c
index 6327cfc36805..cff2bf9db9d2 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c
@@ -30,66 +30,101 @@
 #include <drm/ttm/ttm_placement.h>
 #include <drm/ttm/ttm_page_alloc.h>
 
-static uint32_t vram_placement_flags = TTM_PL_FLAG_VRAM |
-	TTM_PL_FLAG_CACHED;
-
-static uint32_t vram_ne_placement_flags = TTM_PL_FLAG_VRAM |
-	TTM_PL_FLAG_CACHED |
-	TTM_PL_FLAG_NO_EVICT;
+static struct ttm_place vram_placement_flags = {
+	.fpfn = 0,
+	.lpfn = 0,
+	.flags = TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED
+};
 
-static uint32_t sys_placement_flags = TTM_PL_FLAG_SYSTEM |
-	TTM_PL_FLAG_CACHED;
+static struct ttm_place vram_ne_placement_flags = {
+	.fpfn = 0,
+	.lpfn = 0,
+	.flags = TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED | TTM_PL_FLAG_NO_EVICT
+};
 
-static uint32_t sys_ne_placement_flags = TTM_PL_FLAG_SYSTEM |
-	TTM_PL_FLAG_CACHED |
-	TTM_PL_FLAG_NO_EVICT;
+static struct ttm_place sys_placement_flags = {
+	.fpfn = 0,
+	.lpfn = 0,
+	.flags = TTM_PL_FLAG_SYSTEM | TTM_PL_FLAG_CACHED
+};
 
-static uint32_t gmr_placement_flags = VMW_PL_FLAG_GMR |
-	TTM_PL_FLAG_CACHED;
+static struct ttm_place sys_ne_placement_flags = {
+	.fpfn = 0,
+	.lpfn = 0,
+	.flags = TTM_PL_FLAG_SYSTEM | TTM_PL_FLAG_CACHED | TTM_PL_FLAG_NO_EVICT
+};
 
-static uint32_t gmr_ne_placement_flags = VMW_PL_FLAG_GMR |
-	TTM_PL_FLAG_CACHED |
-	TTM_PL_FLAG_NO_EVICT;
+static struct ttm_place gmr_placement_flags = {
+	.fpfn = 0,
+	.lpfn = 0,
+	.flags = VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED
+};
 
-static uint32_t mob_placement_flags = VMW_PL_FLAG_MOB |
-	TTM_PL_FLAG_CACHED;
+static struct ttm_place gmr_ne_placement_flags = {
+	.fpfn = 0,
+	.lpfn = 0,
+	.flags = VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED | TTM_PL_FLAG_NO_EVICT
+};
 
-struct ttm_placement vmw_vram_placement = {
+static struct ttm_place mob_placement_flags = {
 	.fpfn = 0,
 	.lpfn = 0,
+	.flags = VMW_PL_FLAG_MOB | TTM_PL_FLAG_CACHED
+};
+
+struct ttm_placement vmw_vram_placement = {
 	.num_placement = 1,
 	.placement = &vram_placement_flags,
 	.num_busy_placement = 1,
 	.busy_placement = &vram_placement_flags
 };
 
-static uint32_t vram_gmr_placement_flags[] = {
-	TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED,
-	VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED
+static struct ttm_place vram_gmr_placement_flags[] = {
+	{
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED
+	}, {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED
+	}
 };
 
-static uint32_t gmr_vram_placement_flags[] = {
-	VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED,
-	TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED
+static struct ttm_place gmr_vram_placement_flags[] = {
+	{
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED
+	}, {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED
+	}
 };
 
 struct ttm_placement vmw_vram_gmr_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 2,
 	.placement = vram_gmr_placement_flags,
 	.num_busy_placement = 1,
 	.busy_placement = &gmr_placement_flags
 };
 
-static uint32_t vram_gmr_ne_placement_flags[] = {
-	TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED | TTM_PL_FLAG_NO_EVICT,
-	VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED | TTM_PL_FLAG_NO_EVICT
+static struct ttm_place vram_gmr_ne_placement_flags[] = {
+	{
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED |
+			 TTM_PL_FLAG_NO_EVICT
+	}, {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED |
+			 TTM_PL_FLAG_NO_EVICT
+	}
 };
 
 struct ttm_placement vmw_vram_gmr_ne_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 2,
 	.placement = vram_gmr_ne_placement_flags,
 	.num_busy_placement = 1,
@@ -97,8 +132,6 @@ struct ttm_placement vmw_vram_gmr_ne_placement = {
 };
 
 struct ttm_placement vmw_vram_sys_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 1,
 	.placement = &vram_placement_flags,
 	.num_busy_placement = 1,
@@ -106,8 +139,6 @@ struct ttm_placement vmw_vram_sys_placement = {
 };
 
 struct ttm_placement vmw_vram_ne_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 1,
 	.placement = &vram_ne_placement_flags,
 	.num_busy_placement = 1,
@@ -115,8 +146,6 @@ struct ttm_placement vmw_vram_ne_placement = {
 };
 
 struct ttm_placement vmw_sys_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 1,
 	.placement = &sys_placement_flags,
 	.num_busy_placement = 1,
@@ -124,24 +153,33 @@ struct ttm_placement vmw_sys_placement = {
 };
 
 struct ttm_placement vmw_sys_ne_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 1,
 	.placement = &sys_ne_placement_flags,
 	.num_busy_placement = 1,
 	.busy_placement = &sys_ne_placement_flags
 };
 
-static uint32_t evictable_placement_flags[] = {
-	TTM_PL_FLAG_SYSTEM | TTM_PL_FLAG_CACHED,
-	TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED,
-	VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED,
-	VMW_PL_FLAG_MOB | TTM_PL_FLAG_CACHED
+static struct ttm_place evictable_placement_flags[] = {
+	{
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_FLAG_SYSTEM | TTM_PL_FLAG_CACHED
+	}, {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED
+	}, {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED
+	}, {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = VMW_PL_FLAG_MOB | TTM_PL_FLAG_CACHED
+	}
 };
 
 struct ttm_placement vmw_evictable_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 4,
 	.placement = evictable_placement_flags,
 	.num_busy_placement = 1,
@@ -149,8 +187,6 @@ struct ttm_placement vmw_evictable_placement = {
 };
 
 struct ttm_placement vmw_srf_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 1,
 	.num_busy_placement = 2,
 	.placement = &gmr_placement_flags,
@@ -158,8 +194,6 @@ struct ttm_placement vmw_srf_placement = {
 };
 
 struct ttm_placement vmw_mob_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 1,
 	.num_busy_placement = 1,
 	.placement = &mob_placement_flags,
@@ -768,44 +802,6 @@ static int vmw_ttm_fault_reserve_notify(struct ttm_buffer_object *bo)
 }
 
 /**
- * FIXME: We're using the old vmware polling method to sync.
- * Do this with fences instead.
- */
-
-static void *vmw_sync_obj_ref(void *sync_obj)
-{
-
-	return (void *)
-		vmw_fence_obj_reference((struct vmw_fence_obj *) sync_obj);
-}
-
-static void vmw_sync_obj_unref(void **sync_obj)
-{
-	vmw_fence_obj_unreference((struct vmw_fence_obj **) sync_obj);
-}
-
-static int vmw_sync_obj_flush(void *sync_obj)
-{
-	vmw_fence_obj_flush((struct vmw_fence_obj *) sync_obj);
-	return 0;
-}
-
-static bool vmw_sync_obj_signaled(void *sync_obj)
-{
-	return	vmw_fence_obj_signaled((struct vmw_fence_obj *) sync_obj,
-				       DRM_VMW_FENCE_FLAG_EXEC);
-
-}
-
-static int vmw_sync_obj_wait(void *sync_obj, bool lazy, bool interruptible)
-{
-	return vmw_fence_obj_wait((struct vmw_fence_obj *) sync_obj,
-				  DRM_VMW_FENCE_FLAG_EXEC,
-				  lazy, interruptible,
-				  VMW_FENCE_WAIT_TIMEOUT);
-}
-
-/**
  * vmw_move_notify - TTM move_notify_callback
  *
  * @bo:             The TTM buffer object about to move.
@@ -829,11 +825,7 @@ static void vmw_move_notify(struct ttm_buffer_object *bo,
  */
 static void vmw_swap_notify(struct ttm_buffer_object *bo)
 {
-	struct ttm_bo_device *bdev = bo->bdev;
-
-	spin_lock(&bdev->fence_lock);
 	ttm_bo_wait(bo, false, false, false);
-	spin_unlock(&bdev->fence_lock);
 }
 
 
@@ -846,11 +838,6 @@ struct ttm_bo_driver vmw_bo_driver = {
 	.evict_flags = vmw_evict_flags,
 	.move = NULL,
 	.verify_access = vmw_verify_access,
-	.sync_obj_signaled = vmw_sync_obj_signaled,
-	.sync_obj_wait = vmw_sync_obj_wait,
-	.sync_obj_flush = vmw_sync_obj_flush,
-	.sync_obj_unref = vmw_sync_obj_unref,
-	.sync_obj_ref = vmw_sync_obj_ref,
 	.move_notify = vmw_move_notify,
 	.swap_notify = vmw_swap_notify,
 	.fault_reserve_notify = &vmw_ttm_fault_reserve_notify,
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf_res.c b/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf_res.c
index bfeb4b1f2acf..21e9b7f8dad0 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf_res.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf_res.c
@@ -246,7 +246,8 @@ int vmw_cmdbuf_res_remove(struct vmw_cmdbuf_res_manager *man,
 	struct drm_hash_item *hash;
 	int ret;
 
-	ret = drm_ht_find_item(&man->resources, user_key, &hash);
+	ret = drm_ht_find_item(&man->resources, user_key | (res_type << 24),
+			       &hash);
 	if (likely(ret != 0))
 		return -EINVAL;
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c
index ed1d51006ab1..914b375763dc 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c
@@ -198,13 +198,19 @@ int vmw_dmabuf_to_start_of_vram(struct vmw_private *dev_priv,
 {
 	struct ttm_buffer_object *bo = &buf->base;
 	struct ttm_placement placement;
+	struct ttm_place place;
 	int ret = 0;
 
 	if (pin)
-		placement = vmw_vram_ne_placement;
+		place = vmw_vram_ne_placement.placement[0];
 	else
-		placement = vmw_vram_placement;
-	placement.lpfn = bo->num_pages;
+		place = vmw_vram_placement.placement[0];
+	place.lpfn = bo->num_pages;
+
+	placement.num_placement = 1;
+	placement.placement = &place;
+	placement.num_busy_placement = 1;
+	placement.busy_placement = &place;
 
 	ret = ttm_write_lock(&dev_priv->reservation_sem, interruptible);
 	if (unlikely(ret != 0))
@@ -293,21 +299,23 @@ void vmw_bo_get_guest_ptr(const struct ttm_buffer_object *bo,
  */
 void vmw_bo_pin(struct ttm_buffer_object *bo, bool pin)
 {
-	uint32_t pl_flags;
+	struct ttm_place pl;
 	struct ttm_placement placement;
 	uint32_t old_mem_type = bo->mem.mem_type;
 	int ret;
 
 	lockdep_assert_held(&bo->resv->lock.base);
 
-	pl_flags = TTM_PL_FLAG_VRAM | VMW_PL_FLAG_GMR | VMW_PL_FLAG_MOB
+	pl.fpfn = 0;
+	pl.lpfn = 0;
+	pl.flags = TTM_PL_FLAG_VRAM | VMW_PL_FLAG_GMR | VMW_PL_FLAG_MOB
 		| TTM_PL_FLAG_SYSTEM | TTM_PL_FLAG_CACHED;
 	if (pin)
-		pl_flags |= TTM_PL_FLAG_NO_EVICT;
+		pl.flags |= TTM_PL_FLAG_NO_EVICT;
 
 	memset(&placement, 0, sizeof(placement));
 	placement.num_placement = 1;
-	placement.placement = &pl_flags;
+	placement.placement = &pl;
 
 	ret = ttm_bo_validate(bo, &placement, false, true);
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
index 18b54acacfbb..6c6b655defcf 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
@@ -406,11 +406,9 @@ int vmw_3d_resource_inc(struct vmw_private *dev_priv,
 		if (unlikely(ret != 0))
 			--dev_priv->num_3d_resources;
 	} else if (unhide_svga) {
-		mutex_lock(&dev_priv->hw_mutex);
 		vmw_write(dev_priv, SVGA_REG_ENABLE,
 			  vmw_read(dev_priv, SVGA_REG_ENABLE) &
 			  ~SVGA_REG_ENABLE_HIDE);
-		mutex_unlock(&dev_priv->hw_mutex);
 	}
 
 	mutex_unlock(&dev_priv->release_mutex);
@@ -433,13 +431,10 @@ void vmw_3d_resource_dec(struct vmw_private *dev_priv,
 	mutex_lock(&dev_priv->release_mutex);
 	if (unlikely(--dev_priv->num_3d_resources == 0))
 		vmw_release_device(dev_priv);
-	else if (hide_svga) {
-		mutex_lock(&dev_priv->hw_mutex);
+	else if (hide_svga)
 		vmw_write(dev_priv, SVGA_REG_ENABLE,
 			  vmw_read(dev_priv, SVGA_REG_ENABLE) |
 			  SVGA_REG_ENABLE_HIDE);
-		mutex_unlock(&dev_priv->hw_mutex);
-	}
 
 	n3d = (int32_t) dev_priv->num_3d_resources;
 	mutex_unlock(&dev_priv->release_mutex);
@@ -600,12 +595,14 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 	dev_priv->dev = dev;
 	dev_priv->vmw_chipset = chipset;
 	dev_priv->last_read_seqno = (uint32_t) -100;
-	mutex_init(&dev_priv->hw_mutex);
 	mutex_init(&dev_priv->cmdbuf_mutex);
 	mutex_init(&dev_priv->release_mutex);
 	mutex_init(&dev_priv->binding_mutex);
 	rwlock_init(&dev_priv->resource_lock);
 	ttm_lock_init(&dev_priv->reservation_sem);
+	spin_lock_init(&dev_priv->hw_lock);
+	spin_lock_init(&dev_priv->waiter_lock);
+	spin_lock_init(&dev_priv->cap_lock);
 
 	for (i = vmw_res_context; i < vmw_res_max; ++i) {
 		idr_init(&dev_priv->res_idr[i]);
@@ -626,14 +623,11 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 
 	dev_priv->enable_fb = enable_fbdev;
 
-	mutex_lock(&dev_priv->hw_mutex);
-
 	vmw_write(dev_priv, SVGA_REG_ID, SVGA_ID_2);
 	svga_id = vmw_read(dev_priv, SVGA_REG_ID);
 	if (svga_id != SVGA_ID_2) {
 		ret = -ENOSYS;
 		DRM_ERROR("Unsupported SVGA ID 0x%x\n", svga_id);
-		mutex_unlock(&dev_priv->hw_mutex);
 		goto out_err0;
 	}
 
@@ -683,16 +677,16 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 		dev_priv->prim_bb_mem = dev_priv->vram_size;
 
 	ret = vmw_dma_masks(dev_priv);
-	if (unlikely(ret != 0)) {
-		mutex_unlock(&dev_priv->hw_mutex);
+	if (unlikely(ret != 0))
 		goto out_err0;
-	}
 
-	if (unlikely(dev_priv->prim_bb_mem < dev_priv->vram_size))
+	/*
+	 * Limit back buffer size to VRAM size.  Remove this once
+	 * screen targets are implemented.
+	 */
+	if (dev_priv->prim_bb_mem > dev_priv->vram_size)
 		dev_priv->prim_bb_mem = dev_priv->vram_size;
 
-	mutex_unlock(&dev_priv->hw_mutex);
-
 	vmw_print_capabilities(dev_priv->capabilities);
 
 	if (dev_priv->capabilities & SVGA_CAP_GMR2) {
@@ -885,8 +879,7 @@ static int vmw_driver_unload(struct drm_device *dev)
 
 	if (dev_priv->ctx.res_ht_initialized)
 		drm_ht_remove(&dev_priv->ctx.res_ht);
-	if (dev_priv->ctx.cmd_bounce)
-		vfree(dev_priv->ctx.cmd_bounce);
+	vfree(dev_priv->ctx.cmd_bounce);
 	if (dev_priv->enable_fb) {
 		vmw_fb_close(dev_priv);
 		vmw_kms_restore_vga(dev_priv);
@@ -1059,8 +1052,12 @@ static long vmw_generic_ioctl(struct file *filp, unsigned int cmd,
 
 	vmaster = vmw_master_check(dev, file_priv, flags);
 	if (unlikely(IS_ERR(vmaster))) {
-		DRM_INFO("IOCTL ERROR %d\n", nr);
-		return PTR_ERR(vmaster);
+		ret = PTR_ERR(vmaster);
+
+		if (ret != -ERESTARTSYS)
+			DRM_INFO("IOCTL ERROR Command %d, Error %ld.\n",
+				 nr, ret);
+		return ret;
 	}
 
 	ret = ioctl_func(filp, cmd, arg);
@@ -1153,9 +1150,7 @@ static int vmw_master_set(struct drm_device *dev,
 		if (unlikely(ret != 0))
 			return ret;
 		vmw_kms_save_vga(dev_priv);
-		mutex_lock(&dev_priv->hw_mutex);
 		vmw_write(dev_priv, SVGA_REG_TRACES, 0);
-		mutex_unlock(&dev_priv->hw_mutex);
 	}
 
 	if (active) {
@@ -1189,9 +1184,7 @@ out_no_active_lock:
 	if (!dev_priv->enable_fb) {
 		vmw_kms_restore_vga(dev_priv);
 		vmw_3d_resource_dec(dev_priv, true);
-		mutex_lock(&dev_priv->hw_mutex);
 		vmw_write(dev_priv, SVGA_REG_TRACES, 1);
-		mutex_unlock(&dev_priv->hw_mutex);
 	}
 	return ret;
 }
@@ -1226,9 +1219,7 @@ static void vmw_master_drop(struct drm_device *dev,
 			DRM_ERROR("Unable to clean VRAM on master drop.\n");
 		vmw_kms_restore_vga(dev_priv);
 		vmw_3d_resource_dec(dev_priv, true);
-		mutex_lock(&dev_priv->hw_mutex);
 		vmw_write(dev_priv, SVGA_REG_TRACES, 1);
-		mutex_unlock(&dev_priv->hw_mutex);
 	}
 
 	dev_priv->active_master = &dev_priv->fbdev_master;
@@ -1360,10 +1351,8 @@ static void vmw_pm_complete(struct device *kdev)
 	struct drm_device *dev = pci_get_drvdata(pdev);
 	struct vmw_private *dev_priv = vmw_priv(dev);
 
-	mutex_lock(&dev_priv->hw_mutex);
 	vmw_write(dev_priv, SVGA_REG_ID, SVGA_ID_2);
 	(void) vmw_read(dev_priv, SVGA_REG_ID);
-	mutex_unlock(&dev_priv->hw_mutex);
 
 	/**
 	 * Reclaim 3d reference held by fbdev and potentially
@@ -1418,6 +1407,7 @@ static struct drm_driver driver = {
 	.open = vmw_driver_open,
 	.preclose = vmw_preclose,
 	.postclose = vmw_postclose,
+	.set_busid = drm_pci_set_busid,
 
 	.dumb_create = vmw_dumb_create,
 	.dumb_map_offset = vmw_dumb_map_offset,
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
index 99f731757c4b..d26a6daa9719 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
@@ -342,7 +342,6 @@ struct vmw_sw_context{
 	uint32_t *cmd_bounce;
 	uint32_t cmd_bounce_size;
 	struct list_head resource_list;
-	uint32_t fence_flags;
 	struct ttm_buffer_object *cur_query_bo;
 	struct list_head res_relocations;
 	uint32_t *buf_start;
@@ -400,7 +399,8 @@ struct vmw_private {
 	uint32_t memory_size;
 	bool has_gmr;
 	bool has_mob;
-	struct mutex hw_mutex;
+	spinlock_t hw_lock;
+	spinlock_t cap_lock;
 
 	/*
 	 * VGA registers.
@@ -450,8 +450,9 @@ struct vmw_private {
 	atomic_t marker_seq;
 	wait_queue_head_t fence_queue;
 	wait_queue_head_t fifo_queue;
-	int fence_queue_waiters; /* Protected by hw_mutex */
-	int goal_queue_waiters; /* Protected by hw_mutex */
+	spinlock_t waiter_lock;
+	int fence_queue_waiters; /* Protected by waiter_lock */
+	int goal_queue_waiters; /* Protected by waiter_lock */
 	atomic_t fifo_queue_waiters;
 	uint32_t last_read_seqno;
 	spinlock_t irq_lock;
@@ -554,20 +555,35 @@ static inline struct vmw_master *vmw_master(struct drm_master *master)
 	return (struct vmw_master *) master->driver_priv;
 }
 
+/*
+ * The locking here is fine-grained, so that it is performed once
+ * for every read- and write operation. This is of course costly, but we
+ * don't perform much register access in the timing critical paths anyway.
+ * Instead we have the extra benefit of being sure that we don't forget
+ * the hw lock around register accesses.
+ */
 static inline void vmw_write(struct vmw_private *dev_priv,
 			     unsigned int offset, uint32_t value)
 {
+	unsigned long irq_flags;
+
+	spin_lock_irqsave(&dev_priv->hw_lock, irq_flags);
 	outl(offset, dev_priv->io_start + VMWGFX_INDEX_PORT);
 	outl(value, dev_priv->io_start + VMWGFX_VALUE_PORT);
+	spin_unlock_irqrestore(&dev_priv->hw_lock, irq_flags);
 }
 
 static inline uint32_t vmw_read(struct vmw_private *dev_priv,
 				unsigned int offset)
 {
-	uint32_t val;
+	unsigned long irq_flags;
+	u32 val;
 
+	spin_lock_irqsave(&dev_priv->hw_lock, irq_flags);
 	outl(offset, dev_priv->io_start + VMWGFX_INDEX_PORT);
 	val = inl(dev_priv->io_start + VMWGFX_VALUE_PORT);
+	spin_unlock_irqrestore(&dev_priv->hw_lock, irq_flags);
+
 	return val;
 }
 
@@ -704,6 +720,7 @@ extern void *vmw_fifo_reserve(struct vmw_private *dev_priv, uint32_t bytes);
 extern void vmw_fifo_commit(struct vmw_private *dev_priv, uint32_t bytes);
 extern int vmw_fifo_send_fence(struct vmw_private *dev_priv,
 			       uint32_t *seqno);
+extern void vmw_fifo_ping_host_locked(struct vmw_private *, uint32_t reason);
 extern void vmw_fifo_ping_host(struct vmw_private *dev_priv, uint32_t reason);
 extern bool vmw_fifo_have_3d(struct vmw_private *dev_priv);
 extern bool vmw_fifo_have_pitchlock(struct vmw_private *dev_priv);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
index 36b871686d3c..33176d05db35 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
@@ -346,13 +346,11 @@ static int vmw_bo_to_validate_list(struct vmw_sw_context *sw_context,
 		++sw_context->cur_val_buf;
 		val_buf = &vval_buf->base;
 		val_buf->bo = ttm_bo_reference(bo);
-		val_buf->reserved = false;
+		val_buf->shared = false;
 		list_add_tail(&val_buf->head, &sw_context->validate_nodes);
 		vval_buf->validate_as_mob = validate_as_mob;
 	}
 
-	sw_context->fence_flags |= DRM_VMW_FENCE_FLAG_EXEC;
-
 	if (p_val_node)
 		*p_val_node = val_node;
 
@@ -2337,13 +2335,9 @@ int vmw_execbuf_fence_commands(struct drm_file *file_priv,
 
 	if (p_handle != NULL)
 		ret = vmw_user_fence_create(file_priv, dev_priv->fman,
-					    sequence,
-					    DRM_VMW_FENCE_FLAG_EXEC,
-					    p_fence, p_handle);
+					    sequence, p_fence, p_handle);
 	else
-		ret = vmw_fence_create(dev_priv->fman, sequence,
-				       DRM_VMW_FENCE_FLAG_EXEC,
-				       p_fence);
+		ret = vmw_fence_create(dev_priv->fman, sequence, p_fence);
 
 	if (unlikely(ret != 0 && !synced)) {
 		(void) vmw_fallback_wait(dev_priv, false, false,
@@ -2395,7 +2389,7 @@ vmw_execbuf_copy_fence_user(struct vmw_private *dev_priv,
 		BUG_ON(fence == NULL);
 
 		fence_rep.handle = fence_handle;
-		fence_rep.seqno = fence->seqno;
+		fence_rep.seqno = fence->base.seqno;
 		vmw_update_seqno(dev_priv, &dev_priv->fifo);
 		fence_rep.passed_seqno = dev_priv->last_read_seqno;
 	}
@@ -2416,8 +2410,7 @@ vmw_execbuf_copy_fence_user(struct vmw_private *dev_priv,
 		ttm_ref_object_base_unref(vmw_fp->tfile,
 					  fence_handle, TTM_REF_USAGE);
 		DRM_ERROR("Fence copy error. Syncing.\n");
-		(void) vmw_fence_obj_wait(fence, fence->signal_mask,
-					  false, false,
+		(void) vmw_fence_obj_wait(fence, false, false,
 					  VMW_FENCE_WAIT_TIMEOUT);
 	}
 }
@@ -2469,7 +2462,6 @@ int vmw_execbuf_process(struct drm_file *file_priv,
 	sw_context->fp = vmw_fpriv(file_priv);
 	sw_context->cur_reloc = 0;
 	sw_context->cur_val_buf = 0;
-	sw_context->fence_flags = 0;
 	INIT_LIST_HEAD(&sw_context->resource_list);
 	sw_context->cur_query_bo = dev_priv->pinned_bo;
 	sw_context->last_query_ctx = NULL;
@@ -2495,7 +2487,8 @@ int vmw_execbuf_process(struct drm_file *file_priv,
 	if (unlikely(ret != 0))
 		goto out_err_nores;
 
-	ret = ttm_eu_reserve_buffers(&ticket, &sw_context->validate_nodes);
+	ret = ttm_eu_reserve_buffers(&ticket, &sw_context->validate_nodes,
+				     true, NULL);
 	if (unlikely(ret != 0))
 		goto out_err;
 
@@ -2678,15 +2671,15 @@ void __vmw_execbuf_release_pinned_bo(struct vmw_private *dev_priv,
 	INIT_LIST_HEAD(&validate_list);
 
 	pinned_val.bo = ttm_bo_reference(dev_priv->pinned_bo);
+	pinned_val.shared = false;
 	list_add_tail(&pinned_val.head, &validate_list);
 
 	query_val.bo = ttm_bo_reference(dev_priv->dummy_query_bo);
+	query_val.shared = false;
 	list_add_tail(&query_val.head, &validate_list);
 
-	do {
-		ret = ttm_eu_reserve_buffers(&ticket, &validate_list);
-	} while (ret == -ERESTARTSYS);
-
+	ret = ttm_eu_reserve_buffers(&ticket, &validate_list,
+				     false, NULL);
 	if (unlikely(ret != 0)) {
 		vmw_execbuf_unpin_panic(dev_priv);
 		goto out_no_reserve;
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
index b031b48dbb3c..0a474f391fad 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
@@ -374,10 +374,16 @@ static int vmw_fb_create_bo(struct vmw_private *vmw_priv,
 			    size_t size, struct vmw_dma_buffer **out)
 {
 	struct vmw_dma_buffer *vmw_bo;
-	struct ttm_placement ne_placement = vmw_vram_ne_placement;
+	struct ttm_place ne_place = vmw_vram_ne_placement.placement[0];
+	struct ttm_placement ne_placement;
 	int ret;
 
-	ne_placement.lpfn = (size + PAGE_SIZE - 1) >> PAGE_SHIFT;
+	ne_placement.num_placement = 1;
+	ne_placement.placement = &ne_place;
+	ne_placement.num_busy_placement = 1;
+	ne_placement.busy_placement = &ne_place;
+
+	ne_place.lpfn = (size + PAGE_SIZE - 1) >> PAGE_SHIFT;
 
 	(void) ttm_write_lock(&vmw_priv->reservation_sem, false);
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c
index 436b013b4231..945f1e0dad92 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c
@@ -46,6 +46,7 @@ struct vmw_fence_manager {
 	bool goal_irq_on; /* Protected by @goal_irq_mutex */
 	bool seqno_valid; /* Protected by @lock, and may not be set to true
 			     without the @goal_irq_mutex held. */
+	unsigned ctx;
 };
 
 struct vmw_user_fence {
@@ -80,6 +81,12 @@ struct vmw_event_fence_action {
 	uint32_t *tv_usec;
 };
 
+static struct vmw_fence_manager *
+fman_from_fence(struct vmw_fence_obj *fence)
+{
+	return container_of(fence->base.lock, struct vmw_fence_manager, lock);
+}
+
 /**
  * Note on fencing subsystem usage of irqs:
  * Typically the vmw_fences_update function is called
@@ -102,25 +109,131 @@ struct vmw_event_fence_action {
  * objects with actions attached to them.
  */
 
-static void vmw_fence_obj_destroy_locked(struct kref *kref)
+static void vmw_fence_obj_destroy(struct fence *f)
 {
 	struct vmw_fence_obj *fence =
-		container_of(kref, struct vmw_fence_obj, kref);
+		container_of(f, struct vmw_fence_obj, base);
 
-	struct vmw_fence_manager *fman = fence->fman;
-	unsigned int num_fences;
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
+	unsigned long irq_flags;
 
+	spin_lock_irqsave(&fman->lock, irq_flags);
 	list_del_init(&fence->head);
-	num_fences = --fman->num_fence_objects;
-	spin_unlock_irq(&fman->lock);
-	if (fence->destroy)
-		fence->destroy(fence);
-	else
-		kfree(fence);
+	--fman->num_fence_objects;
+	spin_unlock_irqrestore(&fman->lock, irq_flags);
+	fence->destroy(fence);
+}
 
-	spin_lock_irq(&fman->lock);
+static const char *vmw_fence_get_driver_name(struct fence *f)
+{
+	return "vmwgfx";
+}
+
+static const char *vmw_fence_get_timeline_name(struct fence *f)
+{
+	return "svga";
+}
+
+static bool vmw_fence_enable_signaling(struct fence *f)
+{
+	struct vmw_fence_obj *fence =
+		container_of(f, struct vmw_fence_obj, base);
+
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
+	struct vmw_private *dev_priv = fman->dev_priv;
+
+	__le32 __iomem *fifo_mem = dev_priv->mmio_virt;
+	u32 seqno = ioread32(fifo_mem + SVGA_FIFO_FENCE);
+	if (seqno - fence->base.seqno < VMW_FENCE_WRAP)
+		return false;
+
+	vmw_fifo_ping_host(dev_priv, SVGA_SYNC_GENERIC);
+
+	return true;
+}
+
+struct vmwgfx_wait_cb {
+	struct fence_cb base;
+	struct task_struct *task;
+};
+
+static void
+vmwgfx_wait_cb(struct fence *fence, struct fence_cb *cb)
+{
+	struct vmwgfx_wait_cb *wait =
+		container_of(cb, struct vmwgfx_wait_cb, base);
+
+	wake_up_process(wait->task);
+}
+
+static void __vmw_fences_update(struct vmw_fence_manager *fman);
+
+static long vmw_fence_wait(struct fence *f, bool intr, signed long timeout)
+{
+	struct vmw_fence_obj *fence =
+		container_of(f, struct vmw_fence_obj, base);
+
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
+	struct vmw_private *dev_priv = fman->dev_priv;
+	struct vmwgfx_wait_cb cb;
+	long ret = timeout;
+	unsigned long irq_flags;
+
+	if (likely(vmw_fence_obj_signaled(fence)))
+		return timeout;
+
+	vmw_fifo_ping_host(dev_priv, SVGA_SYNC_GENERIC);
+	vmw_seqno_waiter_add(dev_priv);
+
+	spin_lock_irqsave(f->lock, irq_flags);
+
+	if (intr && signal_pending(current)) {
+		ret = -ERESTARTSYS;
+		goto out;
+	}
+
+	cb.base.func = vmwgfx_wait_cb;
+	cb.task = current;
+	list_add(&cb.base.node, &f->cb_list);
+
+	while (ret > 0) {
+		__vmw_fences_update(fman);
+		if (test_bit(FENCE_FLAG_SIGNALED_BIT, &f->flags))
+			break;
+
+		if (intr)
+			__set_current_state(TASK_INTERRUPTIBLE);
+		else
+			__set_current_state(TASK_UNINTERRUPTIBLE);
+		spin_unlock_irqrestore(f->lock, irq_flags);
+
+		ret = schedule_timeout(ret);
+
+		spin_lock_irqsave(f->lock, irq_flags);
+		if (ret > 0 && intr && signal_pending(current))
+			ret = -ERESTARTSYS;
+	}
+
+	if (!list_empty(&cb.base.node))
+		list_del(&cb.base.node);
+	__set_current_state(TASK_RUNNING);
+
+out:
+	spin_unlock_irqrestore(f->lock, irq_flags);
+
+	vmw_seqno_waiter_remove(dev_priv);
+
+	return ret;
 }
 
+static struct fence_ops vmw_fence_ops = {
+	.get_driver_name = vmw_fence_get_driver_name,
+	.get_timeline_name = vmw_fence_get_timeline_name,
+	.enable_signaling = vmw_fence_enable_signaling,
+	.wait = vmw_fence_wait,
+	.release = vmw_fence_obj_destroy,
+};
+
 
 /**
  * Execute signal actions on fences recently signaled.
@@ -186,6 +299,7 @@ struct vmw_fence_manager *vmw_fence_manager_init(struct vmw_private *dev_priv)
 	fman->event_fence_action_size =
 		ttm_round_pot(sizeof(struct vmw_event_fence_action));
 	mutex_init(&fman->goal_irq_mutex);
+	fman->ctx = fence_context_alloc(1);
 
 	return fman;
 }
@@ -207,23 +321,16 @@ void vmw_fence_manager_takedown(struct vmw_fence_manager *fman)
 }
 
 static int vmw_fence_obj_init(struct vmw_fence_manager *fman,
-			      struct vmw_fence_obj *fence,
-			      u32 seqno,
-			      uint32_t mask,
+			      struct vmw_fence_obj *fence, u32 seqno,
 			      void (*destroy) (struct vmw_fence_obj *fence))
 {
 	unsigned long irq_flags;
-	unsigned int num_fences;
 	int ret = 0;
 
-	fence->seqno = seqno;
+	fence_init(&fence->base, &vmw_fence_ops, &fman->lock,
+		   fman->ctx, seqno);
 	INIT_LIST_HEAD(&fence->seq_passed_actions);
-	fence->fman = fman;
-	fence->signaled = 0;
-	fence->signal_mask = mask;
-	kref_init(&fence->kref);
 	fence->destroy = destroy;
-	init_waitqueue_head(&fence->queue);
 
 	spin_lock_irqsave(&fman->lock, irq_flags);
 	if (unlikely(fman->fifo_down)) {
@@ -231,7 +338,7 @@ static int vmw_fence_obj_init(struct vmw_fence_manager *fman,
 		goto out_unlock;
 	}
 	list_add_tail(&fence->head, &fman->fence_list);
-	num_fences = ++fman->num_fence_objects;
+	++fman->num_fence_objects;
 
 out_unlock:
 	spin_unlock_irqrestore(&fman->lock, irq_flags);
@@ -239,38 +346,6 @@ out_unlock:
 
 }
 
-struct vmw_fence_obj *vmw_fence_obj_reference(struct vmw_fence_obj *fence)
-{
-	if (unlikely(fence == NULL))
-		return NULL;
-
-	kref_get(&fence->kref);
-	return fence;
-}
-
-/**
- * vmw_fence_obj_unreference
- *
- * Note that this function may not be entered with disabled irqs since
- * it may re-enable them in the destroy function.
- *
- */
-void vmw_fence_obj_unreference(struct vmw_fence_obj **fence_p)
-{
-	struct vmw_fence_obj *fence = *fence_p;
-	struct vmw_fence_manager *fman;
-
-	if (unlikely(fence == NULL))
-		return;
-
-	fman = fence->fman;
-	*fence_p = NULL;
-	spin_lock_irq(&fman->lock);
-	BUG_ON(atomic_read(&fence->kref.refcount) == 0);
-	kref_put(&fence->kref, vmw_fence_obj_destroy_locked);
-	spin_unlock_irq(&fman->lock);
-}
-
 static void vmw_fences_perform_actions(struct vmw_fence_manager *fman,
 				struct list_head *list)
 {
@@ -326,7 +401,7 @@ static bool vmw_fence_goal_new_locked(struct vmw_fence_manager *fman,
 	list_for_each_entry(fence, &fman->fence_list, head) {
 		if (!list_empty(&fence->seq_passed_actions)) {
 			fman->seqno_valid = true;
-			iowrite32(fence->seqno,
+			iowrite32(fence->base.seqno,
 				  fifo_mem + SVGA_FIFO_FENCE_GOAL);
 			break;
 		}
@@ -353,27 +428,27 @@ static bool vmw_fence_goal_new_locked(struct vmw_fence_manager *fman,
  */
 static bool vmw_fence_goal_check_locked(struct vmw_fence_obj *fence)
 {
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
 	u32 goal_seqno;
 	__le32 __iomem *fifo_mem;
 
-	if (fence->signaled & DRM_VMW_FENCE_FLAG_EXEC)
+	if (fence_is_signaled_locked(&fence->base))
 		return false;
 
-	fifo_mem = fence->fman->dev_priv->mmio_virt;
+	fifo_mem = fman->dev_priv->mmio_virt;
 	goal_seqno = ioread32(fifo_mem + SVGA_FIFO_FENCE_GOAL);
-	if (likely(fence->fman->seqno_valid &&
-		   goal_seqno - fence->seqno < VMW_FENCE_WRAP))
+	if (likely(fman->seqno_valid &&
+		   goal_seqno - fence->base.seqno < VMW_FENCE_WRAP))
 		return false;
 
-	iowrite32(fence->seqno, fifo_mem + SVGA_FIFO_FENCE_GOAL);
-	fence->fman->seqno_valid = true;
+	iowrite32(fence->base.seqno, fifo_mem + SVGA_FIFO_FENCE_GOAL);
+	fman->seqno_valid = true;
 
 	return true;
 }
 
-void vmw_fences_update(struct vmw_fence_manager *fman)
+static void __vmw_fences_update(struct vmw_fence_manager *fman)
 {
-	unsigned long flags;
 	struct vmw_fence_obj *fence, *next_fence;
 	struct list_head action_list;
 	bool needs_rerun;
@@ -382,32 +457,25 @@ void vmw_fences_update(struct vmw_fence_manager *fman)
 
 	seqno = ioread32(fifo_mem + SVGA_FIFO_FENCE);
 rerun:
-	spin_lock_irqsave(&fman->lock, flags);
 	list_for_each_entry_safe(fence, next_fence, &fman->fence_list, head) {
-		if (seqno - fence->seqno < VMW_FENCE_WRAP) {
+		if (seqno - fence->base.seqno < VMW_FENCE_WRAP) {
 			list_del_init(&fence->head);
-			fence->signaled |= DRM_VMW_FENCE_FLAG_EXEC;
+			fence_signal_locked(&fence->base);
 			INIT_LIST_HEAD(&action_list);
 			list_splice_init(&fence->seq_passed_actions,
 					 &action_list);
 			vmw_fences_perform_actions(fman, &action_list);
-			wake_up_all(&fence->queue);
 		} else
 			break;
 	}
 
-	needs_rerun = vmw_fence_goal_new_locked(fman, seqno);
-
-	if (!list_empty(&fman->cleanup_list))
-		(void) schedule_work(&fman->work);
-	spin_unlock_irqrestore(&fman->lock, flags);
-
 	/*
 	 * Rerun if the fence goal seqno was updated, and the
 	 * hardware might have raced with that update, so that
 	 * we missed a fence_goal irq.
 	 */
 
+	needs_rerun = vmw_fence_goal_new_locked(fman, seqno);
 	if (unlikely(needs_rerun)) {
 		new_seqno = ioread32(fifo_mem + SVGA_FIFO_FENCE);
 		if (new_seqno != seqno) {
@@ -415,107 +483,69 @@ rerun:
 			goto rerun;
 		}
 	}
+
+	if (!list_empty(&fman->cleanup_list))
+		(void) schedule_work(&fman->work);
 }
 
-bool vmw_fence_obj_signaled(struct vmw_fence_obj *fence,
-			    uint32_t flags)
+void vmw_fences_update(struct vmw_fence_manager *fman)
 {
-	struct vmw_fence_manager *fman = fence->fman;
 	unsigned long irq_flags;
-	uint32_t signaled;
 
 	spin_lock_irqsave(&fman->lock, irq_flags);
-	signaled = fence->signaled;
+	__vmw_fences_update(fman);
 	spin_unlock_irqrestore(&fman->lock, irq_flags);
+}
 
-	flags &= fence->signal_mask;
-	if ((signaled & flags) == flags)
-		return 1;
+bool vmw_fence_obj_signaled(struct vmw_fence_obj *fence)
+{
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
 
-	if ((signaled & DRM_VMW_FENCE_FLAG_EXEC) == 0)
-		vmw_fences_update(fman);
+	if (test_bit(FENCE_FLAG_SIGNALED_BIT, &fence->base.flags))
+		return 1;
 
-	spin_lock_irqsave(&fman->lock, irq_flags);
-	signaled = fence->signaled;
-	spin_unlock_irqrestore(&fman->lock, irq_flags);
+	vmw_fences_update(fman);
 
-	return ((signaled & flags) == flags);
+	return fence_is_signaled(&fence->base);
 }
 
-int vmw_fence_obj_wait(struct vmw_fence_obj *fence,
-		       uint32_t flags, bool lazy,
+int vmw_fence_obj_wait(struct vmw_fence_obj *fence, bool lazy,
 		       bool interruptible, unsigned long timeout)
 {
-	struct vmw_private *dev_priv = fence->fman->dev_priv;
-	long ret;
+	long ret = fence_wait_timeout(&fence->base, interruptible, timeout);
 
-	if (likely(vmw_fence_obj_signaled(fence, flags)))
+	if (likely(ret > 0))
 		return 0;
-
-	vmw_fifo_ping_host(dev_priv, SVGA_SYNC_GENERIC);
-	vmw_seqno_waiter_add(dev_priv);
-
-	if (interruptible)
-		ret = wait_event_interruptible_timeout
-			(fence->queue,
-			 vmw_fence_obj_signaled(fence, flags),
-			 timeout);
+	else if (ret == 0)
+		return -EBUSY;
 	else
-		ret = wait_event_timeout
-			(fence->queue,
-			 vmw_fence_obj_signaled(fence, flags),
-			 timeout);
-
-	vmw_seqno_waiter_remove(dev_priv);
-
-	if (unlikely(ret == 0))
-		ret = -EBUSY;
-	else if (likely(ret > 0))
-		ret = 0;
-
-	return ret;
+		return ret;
 }
 
 void vmw_fence_obj_flush(struct vmw_fence_obj *fence)
 {
-	struct vmw_private *dev_priv = fence->fman->dev_priv;
+	struct vmw_private *dev_priv = fman_from_fence(fence)->dev_priv;
 
 	vmw_fifo_ping_host(dev_priv, SVGA_SYNC_GENERIC);
 }
 
 static void vmw_fence_destroy(struct vmw_fence_obj *fence)
 {
-	struct vmw_fence_manager *fman = fence->fman;
-
-	kfree(fence);
-	/*
-	 * Free kernel space accounting.
-	 */
-	ttm_mem_global_free(vmw_mem_glob(fman->dev_priv),
-			    fman->fence_size);
+	fence_free(&fence->base);
 }
 
 int vmw_fence_create(struct vmw_fence_manager *fman,
 		     uint32_t seqno,
-		     uint32_t mask,
 		     struct vmw_fence_obj **p_fence)
 {
-	struct ttm_mem_global *mem_glob = vmw_mem_glob(fman->dev_priv);
 	struct vmw_fence_obj *fence;
 	int ret;
 
-	ret = ttm_mem_global_alloc(mem_glob, fman->fence_size,
-				   false, false);
-	if (unlikely(ret != 0))
-		return ret;
-
 	fence = kzalloc(sizeof(*fence), GFP_KERNEL);
-	if (unlikely(fence == NULL)) {
-		ret = -ENOMEM;
-		goto out_no_object;
-	}
+	if (unlikely(fence == NULL))
+		return -ENOMEM;
 
-	ret = vmw_fence_obj_init(fman, fence, seqno, mask,
+	ret = vmw_fence_obj_init(fman, fence, seqno,
 				 vmw_fence_destroy);
 	if (unlikely(ret != 0))
 		goto out_err_init;
@@ -525,8 +555,6 @@ int vmw_fence_create(struct vmw_fence_manager *fman,
 
 out_err_init:
 	kfree(fence);
-out_no_object:
-	ttm_mem_global_free(mem_glob, fman->fence_size);
 	return ret;
 }
 
@@ -535,7 +563,7 @@ static void vmw_user_fence_destroy(struct vmw_fence_obj *fence)
 {
 	struct vmw_user_fence *ufence =
 		container_of(fence, struct vmw_user_fence, fence);
-	struct vmw_fence_manager *fman = fence->fman;
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
 
 	ttm_base_object_kfree(ufence, base);
 	/*
@@ -559,7 +587,6 @@ static void vmw_user_fence_base_release(struct ttm_base_object **p_base)
 int vmw_user_fence_create(struct drm_file *file_priv,
 			  struct vmw_fence_manager *fman,
 			  uint32_t seqno,
-			  uint32_t mask,
 			  struct vmw_fence_obj **p_fence,
 			  uint32_t *p_handle)
 {
@@ -586,7 +613,7 @@ int vmw_user_fence_create(struct drm_file *file_priv,
 	}
 
 	ret = vmw_fence_obj_init(fman, &ufence->fence, seqno,
-				 mask, vmw_user_fence_destroy);
+				 vmw_user_fence_destroy);
 	if (unlikely(ret != 0)) {
 		kfree(ufence);
 		goto out_no_object;
@@ -629,7 +656,6 @@ out_no_object:
 
 void vmw_fence_fifo_down(struct vmw_fence_manager *fman)
 {
-	unsigned long irq_flags;
 	struct list_head action_list;
 	int ret;
 
@@ -638,35 +664,32 @@ void vmw_fence_fifo_down(struct vmw_fence_manager *fman)
 	 * restart when we've released the fman->lock.
 	 */
 
-	spin_lock_irqsave(&fman->lock, irq_flags);
+	spin_lock_irq(&fman->lock);
 	fman->fifo_down = true;
 	while (!list_empty(&fman->fence_list)) {
 		struct vmw_fence_obj *fence =
 			list_entry(fman->fence_list.prev, struct vmw_fence_obj,
 				   head);
-		kref_get(&fence->kref);
+		fence_get(&fence->base);
 		spin_unlock_irq(&fman->lock);
 
-		ret = vmw_fence_obj_wait(fence, fence->signal_mask,
-					 false, false,
+		ret = vmw_fence_obj_wait(fence, false, false,
 					 VMW_FENCE_WAIT_TIMEOUT);
 
 		if (unlikely(ret != 0)) {
 			list_del_init(&fence->head);
-			fence->signaled |= DRM_VMW_FENCE_FLAG_EXEC;
+			fence_signal(&fence->base);
 			INIT_LIST_HEAD(&action_list);
 			list_splice_init(&fence->seq_passed_actions,
 					 &action_list);
 			vmw_fences_perform_actions(fman, &action_list);
-			wake_up_all(&fence->queue);
 		}
 
-		spin_lock_irq(&fman->lock);
-
 		BUG_ON(!list_empty(&fence->head));
-		kref_put(&fence->kref, vmw_fence_obj_destroy_locked);
+		fence_put(&fence->base);
+		spin_lock_irq(&fman->lock);
 	}
-	spin_unlock_irqrestore(&fman->lock, irq_flags);
+	spin_unlock_irq(&fman->lock);
 }
 
 void vmw_fence_fifo_up(struct vmw_fence_manager *fman)
@@ -716,14 +739,14 @@ int vmw_fence_obj_wait_ioctl(struct drm_device *dev, void *data,
 
 	timeout = jiffies;
 	if (time_after_eq(timeout, (unsigned long)arg->kernel_cookie)) {
-		ret = ((vmw_fence_obj_signaled(fence, arg->flags)) ?
+		ret = ((vmw_fence_obj_signaled(fence)) ?
 		       0 : -EBUSY);
 		goto out;
 	}
 
 	timeout = (unsigned long)arg->kernel_cookie - timeout;
 
-	ret = vmw_fence_obj_wait(fence, arg->flags, arg->lazy, true, timeout);
+	ret = vmw_fence_obj_wait(fence, arg->lazy, true, timeout);
 
 out:
 	ttm_base_object_unref(&base);
@@ -758,12 +781,12 @@ int vmw_fence_obj_signaled_ioctl(struct drm_device *dev, void *data,
 	}
 
 	fence = &(container_of(base, struct vmw_user_fence, base)->fence);
-	fman = fence->fman;
+	fman = fman_from_fence(fence);
 
-	arg->signaled = vmw_fence_obj_signaled(fence, arg->flags);
-	spin_lock_irq(&fman->lock);
+	arg->signaled = vmw_fence_obj_signaled(fence);
 
-	arg->signaled_flags = fence->signaled;
+	arg->signaled_flags = arg->flags;
+	spin_lock_irq(&fman->lock);
 	arg->passed_seqno = dev_priv->last_read_seqno;
 	spin_unlock_irq(&fman->lock);
 
@@ -876,7 +899,7 @@ static void vmw_event_fence_action_cleanup(struct vmw_fence_action *action)
 {
 	struct vmw_event_fence_action *eaction =
 		container_of(action, struct vmw_event_fence_action, action);
-	struct vmw_fence_manager *fman = eaction->fence->fman;
+	struct vmw_fence_manager *fman = fman_from_fence(eaction->fence);
 	unsigned long irq_flags;
 
 	spin_lock_irqsave(&fman->lock, irq_flags);
@@ -900,7 +923,7 @@ static void vmw_event_fence_action_cleanup(struct vmw_fence_action *action)
 static void vmw_fence_obj_add_action(struct vmw_fence_obj *fence,
 			      struct vmw_fence_action *action)
 {
-	struct vmw_fence_manager *fman = fence->fman;
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
 	unsigned long irq_flags;
 	bool run_update = false;
 
@@ -908,7 +931,7 @@ static void vmw_fence_obj_add_action(struct vmw_fence_obj *fence,
 	spin_lock_irqsave(&fman->lock, irq_flags);
 
 	fman->pending_actions[action->type]++;
-	if (fence->signaled & DRM_VMW_FENCE_FLAG_EXEC) {
+	if (fence_is_signaled_locked(&fence->base)) {
 		struct list_head action_list;
 
 		INIT_LIST_HEAD(&action_list);
@@ -960,7 +983,7 @@ int vmw_event_fence_action_queue(struct drm_file *file_priv,
 				 bool interruptible)
 {
 	struct vmw_event_fence_action *eaction;
-	struct vmw_fence_manager *fman = fence->fman;
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
 	struct vmw_fpriv *vmw_fp = vmw_fpriv(file_priv);
 	unsigned long irq_flags;
 
@@ -1000,7 +1023,8 @@ static int vmw_event_fence_action_create(struct drm_file *file_priv,
 				  bool interruptible)
 {
 	struct vmw_event_fence_pending *event;
-	struct drm_device *dev = fence->fman->dev_priv->dev;
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
+	struct drm_device *dev = fman->dev_priv->dev;
 	unsigned long irq_flags;
 	int ret;
 
@@ -1049,6 +1073,8 @@ static int vmw_event_fence_action_create(struct drm_file *file_priv,
 	if (ret != 0)
 		goto out_no_queue;
 
+	return 0;
+
 out_no_queue:
 	event->base.destroy(&event->base);
 out_no_event:
@@ -1124,17 +1150,10 @@ int vmw_fence_event_ioctl(struct drm_device *dev, void *data,
 
 	BUG_ON(fence == NULL);
 
-	if (arg->flags & DRM_VMW_FE_FLAG_REQ_TIME)
-		ret = vmw_event_fence_action_create(file_priv, fence,
-						    arg->flags,
-						    arg->user_data,
-						    true);
-	else
-		ret = vmw_event_fence_action_create(file_priv, fence,
-						    arg->flags,
-						    arg->user_data,
-						    true);
-
+	ret = vmw_event_fence_action_create(file_priv, fence,
+					    arg->flags,
+					    arg->user_data,
+					    true);
 	if (unlikely(ret != 0)) {
 		if (ret != -ERESTARTSYS)
 			DRM_ERROR("Failed to attach event to fence.\n");
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.h b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.h
index faf2e7873860..26a4add39208 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.h
@@ -27,6 +27,8 @@
 
 #ifndef _VMWGFX_FENCE_H_
 
+#include <linux/fence.h>
+
 #define VMW_FENCE_WAIT_TIMEOUT (5*HZ)
 
 struct vmw_private;
@@ -50,16 +52,11 @@ struct vmw_fence_action {
 };
 
 struct vmw_fence_obj {
-	struct kref kref;
-	u32 seqno;
+	struct fence base;
 
-	struct vmw_fence_manager *fman;
 	struct list_head head;
-	uint32_t signaled;
-	uint32_t signal_mask;
 	struct list_head seq_passed_actions;
 	void (*destroy)(struct vmw_fence_obj *fence);
-	wait_queue_head_t queue;
 };
 
 extern struct vmw_fence_manager *
@@ -67,17 +64,29 @@ vmw_fence_manager_init(struct vmw_private *dev_priv);
 
 extern void vmw_fence_manager_takedown(struct vmw_fence_manager *fman);
 
-extern void vmw_fence_obj_unreference(struct vmw_fence_obj **fence_p);
+static inline void
+vmw_fence_obj_unreference(struct vmw_fence_obj **fence_p)
+{
+	struct vmw_fence_obj *fence = *fence_p;
+
+	*fence_p = NULL;
+	if (fence)
+		fence_put(&fence->base);
+}
 
-extern struct vmw_fence_obj *
-vmw_fence_obj_reference(struct vmw_fence_obj *fence);
+static inline struct vmw_fence_obj *
+vmw_fence_obj_reference(struct vmw_fence_obj *fence)
+{
+	if (fence)
+		fence_get(&fence->base);
+	return fence;
+}
 
 extern void vmw_fences_update(struct vmw_fence_manager *fman);
 
-extern bool vmw_fence_obj_signaled(struct vmw_fence_obj *fence,
-				   uint32_t flags);
+extern bool vmw_fence_obj_signaled(struct vmw_fence_obj *fence);
 
-extern int vmw_fence_obj_wait(struct vmw_fence_obj *fence, uint32_t flags,
+extern int vmw_fence_obj_wait(struct vmw_fence_obj *fence,
 			      bool lazy,
 			      bool interruptible, unsigned long timeout);
 
@@ -85,13 +94,11 @@ extern void vmw_fence_obj_flush(struct vmw_fence_obj *fence);
 
 extern int vmw_fence_create(struct vmw_fence_manager *fman,
 			    uint32_t seqno,
-			    uint32_t mask,
 			    struct vmw_fence_obj **p_fence);
 
 extern int vmw_user_fence_create(struct drm_file *file_priv,
 				 struct vmw_fence_manager *fman,
 				 uint32_t sequence,
-				 uint32_t mask,
 				 struct vmw_fence_obj **p_fence,
 				 uint32_t *p_handle);
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c
index 6eae14d2a3f7..39f2b03888e7 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c
@@ -44,10 +44,10 @@ bool vmw_fifo_have_3d(struct vmw_private *dev_priv)
 		if (!dev_priv->has_mob)
 			return false;
 
-		mutex_lock(&dev_priv->hw_mutex);
+		spin_lock(&dev_priv->cap_lock);
 		vmw_write(dev_priv, SVGA_REG_DEV_CAP, SVGA3D_DEVCAP_3D);
 		result = vmw_read(dev_priv, SVGA_REG_DEV_CAP);
-		mutex_unlock(&dev_priv->hw_mutex);
+		spin_unlock(&dev_priv->cap_lock);
 
 		return (result != 0);
 	}
@@ -120,7 +120,6 @@ int vmw_fifo_init(struct vmw_private *dev_priv, struct vmw_fifo_state *fifo)
 	DRM_INFO("height %d\n", vmw_read(dev_priv, SVGA_REG_HEIGHT));
 	DRM_INFO("bpp %d\n", vmw_read(dev_priv, SVGA_REG_BITS_PER_PIXEL));
 
-	mutex_lock(&dev_priv->hw_mutex);
 	dev_priv->enable_state = vmw_read(dev_priv, SVGA_REG_ENABLE);
 	dev_priv->config_done_state = vmw_read(dev_priv, SVGA_REG_CONFIG_DONE);
 	dev_priv->traces_state = vmw_read(dev_priv, SVGA_REG_TRACES);
@@ -143,7 +142,6 @@ int vmw_fifo_init(struct vmw_private *dev_priv, struct vmw_fifo_state *fifo)
 	mb();
 
 	vmw_write(dev_priv, SVGA_REG_CONFIG_DONE, 1);
-	mutex_unlock(&dev_priv->hw_mutex);
 
 	max = ioread32(fifo_mem + SVGA_FIFO_MAX);
 	min = ioread32(fifo_mem  + SVGA_FIFO_MIN);
@@ -163,23 +161,25 @@ int vmw_fifo_init(struct vmw_private *dev_priv, struct vmw_fifo_state *fifo)
 void vmw_fifo_ping_host(struct vmw_private *dev_priv, uint32_t reason)
 {
 	__le32 __iomem *fifo_mem = dev_priv->mmio_virt;
+	static DEFINE_SPINLOCK(ping_lock);
+	unsigned long irq_flags;
 
-	mutex_lock(&dev_priv->hw_mutex);
-
+	/*
+	 * The ping_lock is needed because we don't have an atomic
+	 * test-and-set of the SVGA_FIFO_BUSY register.
+	 */
+	spin_lock_irqsave(&ping_lock, irq_flags);
 	if (unlikely(ioread32(fifo_mem + SVGA_FIFO_BUSY) == 0)) {
 		iowrite32(1, fifo_mem + SVGA_FIFO_BUSY);
 		vmw_write(dev_priv, SVGA_REG_SYNC, reason);
 	}
-
-	mutex_unlock(&dev_priv->hw_mutex);
+	spin_unlock_irqrestore(&ping_lock, irq_flags);
 }
 
 void vmw_fifo_release(struct vmw_private *dev_priv, struct vmw_fifo_state *fifo)
 {
 	__le32 __iomem *fifo_mem = dev_priv->mmio_virt;
 
-	mutex_lock(&dev_priv->hw_mutex);
-
 	vmw_write(dev_priv, SVGA_REG_SYNC, SVGA_SYNC_GENERIC);
 	while (vmw_read(dev_priv, SVGA_REG_BUSY) != 0)
 		;
@@ -193,7 +193,6 @@ void vmw_fifo_release(struct vmw_private *dev_priv, struct vmw_fifo_state *fifo)
 	vmw_write(dev_priv, SVGA_REG_TRACES,
 		  dev_priv->traces_state);
 
-	mutex_unlock(&dev_priv->hw_mutex);
 	vmw_marker_queue_takedown(&fifo->marker_queue);
 
 	if (likely(fifo->static_buffer != NULL)) {
@@ -266,7 +265,7 @@ static int vmw_fifo_wait(struct vmw_private *dev_priv,
 		return vmw_fifo_wait_noirq(dev_priv, bytes,
 					   interruptible, timeout);
 
-	mutex_lock(&dev_priv->hw_mutex);
+	spin_lock(&dev_priv->waiter_lock);
 	if (atomic_add_return(1, &dev_priv->fifo_queue_waiters) > 0) {
 		spin_lock_irqsave(&dev_priv->irq_lock, irq_flags);
 		outl(SVGA_IRQFLAG_FIFO_PROGRESS,
@@ -275,7 +274,7 @@ static int vmw_fifo_wait(struct vmw_private *dev_priv,
 		vmw_write(dev_priv, SVGA_REG_IRQMASK, dev_priv->irq_mask);
 		spin_unlock_irqrestore(&dev_priv->irq_lock, irq_flags);
 	}
-	mutex_unlock(&dev_priv->hw_mutex);
+	spin_unlock(&dev_priv->waiter_lock);
 
 	if (interruptible)
 		ret = wait_event_interruptible_timeout
@@ -291,14 +290,14 @@ static int vmw_fifo_wait(struct vmw_private *dev_priv,
 	else if (likely(ret > 0))
 		ret = 0;
 
-	mutex_lock(&dev_priv->hw_mutex);
+	spin_lock(&dev_priv->waiter_lock);
 	if (atomic_dec_and_test(&dev_priv->fifo_queue_waiters)) {
 		spin_lock_irqsave(&dev_priv->irq_lock, irq_flags);
 		dev_priv->irq_mask &= ~SVGA_IRQFLAG_FIFO_PROGRESS;
 		vmw_write(dev_priv, SVGA_REG_IRQMASK, dev_priv->irq_mask);
 		spin_unlock_irqrestore(&dev_priv->irq_lock, irq_flags);
 	}
-	mutex_unlock(&dev_priv->hw_mutex);
+	spin_unlock(&dev_priv->waiter_lock);
 
 	return ret;
 }
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c b/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c
index 26f8bdde3529..170b61be1e4e 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c
@@ -46,8 +46,7 @@ struct vmwgfx_gmrid_man {
 
 static int vmw_gmrid_man_get_node(struct ttm_mem_type_manager *man,
 				  struct ttm_buffer_object *bo,
-				  struct ttm_placement *placement,
-				  uint32_t flags,
+				  const struct ttm_place *place,
 				  struct ttm_mem_reg *mem)
 {
 	struct vmwgfx_gmrid_man *gman =
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c
index 37881ecf5d7a..69c8ce23123c 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c
@@ -135,13 +135,13 @@ static int vmw_fill_compat_cap(struct vmw_private *dev_priv, void *bounce,
 		(pair_offset + max_size * sizeof(SVGA3dCapPair)) / sizeof(u32);
 	compat_cap->header.type = SVGA3DCAPS_RECORD_DEVCAPS;
 
-	mutex_lock(&dev_priv->hw_mutex);
+	spin_lock(&dev_priv->cap_lock);
 	for (i = 0; i < max_size; ++i) {
 		vmw_write(dev_priv, SVGA_REG_DEV_CAP, i);
 		compat_cap->pairs[i][0] = i;
 		compat_cap->pairs[i][1] = vmw_read(dev_priv, SVGA_REG_DEV_CAP);
 	}
-	mutex_unlock(&dev_priv->hw_mutex);
+	spin_unlock(&dev_priv->cap_lock);
 
 	return 0;
 }
@@ -191,12 +191,12 @@ int vmw_get_cap_3d_ioctl(struct drm_device *dev, void *data,
 		if (num > SVGA3D_DEVCAP_MAX)
 			num = SVGA3D_DEVCAP_MAX;
 
-		mutex_lock(&dev_priv->hw_mutex);
+		spin_lock(&dev_priv->cap_lock);
 		for (i = 0; i < num; ++i) {
 			vmw_write(dev_priv, SVGA_REG_DEV_CAP, i);
 			*bounce32++ = vmw_read(dev_priv, SVGA_REG_DEV_CAP);
 		}
-		mutex_unlock(&dev_priv->hw_mutex);
+		spin_unlock(&dev_priv->cap_lock);
 	} else if (gb_objects) {
 		ret = vmw_fill_compat_cap(dev_priv, bounce, size);
 		if (unlikely(ret != 0))
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c b/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c
index 0c423766c441..9fe9827ee499 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c
@@ -62,13 +62,8 @@ irqreturn_t vmw_irq_handler(int irq, void *arg)
 
 static bool vmw_fifo_idle(struct vmw_private *dev_priv, uint32_t seqno)
 {
-	uint32_t busy;
 
-	mutex_lock(&dev_priv->hw_mutex);
-	busy = vmw_read(dev_priv, SVGA_REG_BUSY);
-	mutex_unlock(&dev_priv->hw_mutex);
-
-	return (busy == 0);
+	return (vmw_read(dev_priv, SVGA_REG_BUSY) == 0);
 }
 
 void vmw_update_seqno(struct vmw_private *dev_priv,
@@ -184,7 +179,7 @@ int vmw_fallback_wait(struct vmw_private *dev_priv,
 
 void vmw_seqno_waiter_add(struct vmw_private *dev_priv)
 {
-	mutex_lock(&dev_priv->hw_mutex);
+	spin_lock(&dev_priv->waiter_lock);
 	if (dev_priv->fence_queue_waiters++ == 0) {
 		unsigned long irq_flags;
 
@@ -195,12 +190,12 @@ void vmw_seqno_waiter_add(struct vmw_private *dev_priv)
 		vmw_write(dev_priv, SVGA_REG_IRQMASK, dev_priv->irq_mask);
 		spin_unlock_irqrestore(&dev_priv->irq_lock, irq_flags);
 	}
-	mutex_unlock(&dev_priv->hw_mutex);
+	spin_unlock(&dev_priv->waiter_lock);
 }
 
 void vmw_seqno_waiter_remove(struct vmw_private *dev_priv)
 {
-	mutex_lock(&dev_priv->hw_mutex);
+	spin_lock(&dev_priv->waiter_lock);
 	if (--dev_priv->fence_queue_waiters == 0) {
 		unsigned long irq_flags;
 
@@ -209,13 +204,13 @@ void vmw_seqno_waiter_remove(struct vmw_private *dev_priv)
 		vmw_write(dev_priv, SVGA_REG_IRQMASK, dev_priv->irq_mask);
 		spin_unlock_irqrestore(&dev_priv->irq_lock, irq_flags);
 	}
-	mutex_unlock(&dev_priv->hw_mutex);
+	spin_unlock(&dev_priv->waiter_lock);
 }
 
 
 void vmw_goal_waiter_add(struct vmw_private *dev_priv)
 {
-	mutex_lock(&dev_priv->hw_mutex);
+	spin_lock(&dev_priv->waiter_lock);
 	if (dev_priv->goal_queue_waiters++ == 0) {
 		unsigned long irq_flags;
 
@@ -226,12 +221,12 @@ void vmw_goal_waiter_add(struct vmw_private *dev_priv)
 		vmw_write(dev_priv, SVGA_REG_IRQMASK, dev_priv->irq_mask);
 		spin_unlock_irqrestore(&dev_priv->irq_lock, irq_flags);
 	}
-	mutex_unlock(&dev_priv->hw_mutex);
+	spin_unlock(&dev_priv->waiter_lock);
 }
 
 void vmw_goal_waiter_remove(struct vmw_private *dev_priv)
 {
-	mutex_lock(&dev_priv->hw_mutex);
+	spin_lock(&dev_priv->waiter_lock);
 	if (--dev_priv->goal_queue_waiters == 0) {
 		unsigned long irq_flags;
 
@@ -240,7 +235,7 @@ void vmw_goal_waiter_remove(struct vmw_private *dev_priv)
 		vmw_write(dev_priv, SVGA_REG_IRQMASK, dev_priv->irq_mask);
 		spin_unlock_irqrestore(&dev_priv->irq_lock, irq_flags);
 	}
-	mutex_unlock(&dev_priv->hw_mutex);
+	spin_unlock(&dev_priv->waiter_lock);
 }
 
 int vmw_wait_seqno(struct vmw_private *dev_priv,
@@ -315,9 +310,7 @@ void vmw_irq_uninstall(struct drm_device *dev)
 	if (!(dev_priv->capabilities & SVGA_CAP_IRQMASK))
 		return;
 
-	mutex_lock(&dev_priv->hw_mutex);
 	vmw_write(dev_priv, SVGA_REG_IRQMASK, 0);
-	mutex_unlock(&dev_priv->hw_mutex);
 
 	status = inl(dev_priv->io_start + VMWGFX_IRQSTATUS_PORT);
 	outl(status, dev_priv->io_start + VMWGFX_IRQSTATUS_PORT);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
index d2bc2b03d4c6..8725b79e7847 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
@@ -187,7 +187,7 @@ int vmw_du_crtc_cursor_set(struct drm_crtc *crtc, struct drm_file *file_priv,
 	 * can do this since the caller in the drm core doesn't check anything
 	 * which is protected by any looks.
 	 */
-	drm_modeset_unlock(&crtc->mutex);
+	drm_modeset_unlock_crtc(crtc);
 	drm_modeset_lock_all(dev_priv->dev);
 
 	/* A lot of the code assumes this */
@@ -252,7 +252,7 @@ int vmw_du_crtc_cursor_set(struct drm_crtc *crtc, struct drm_file *file_priv,
 	ret = 0;
 out:
 	drm_modeset_unlock_all(dev_priv->dev);
-	drm_modeset_lock(&crtc->mutex, NULL);
+	drm_modeset_lock_crtc(crtc, crtc->cursor);
 
 	return ret;
 }
@@ -273,7 +273,7 @@ int vmw_du_crtc_cursor_move(struct drm_crtc *crtc, int x, int y)
 	 * can do this since the caller in the drm core doesn't check anything
 	 * which is protected by any looks.
 	 */
-	drm_modeset_unlock(&crtc->mutex);
+	drm_modeset_unlock_crtc(crtc);
 	drm_modeset_lock_all(dev_priv->dev);
 
 	vmw_cursor_update_position(dev_priv, shown,
@@ -281,7 +281,7 @@ int vmw_du_crtc_cursor_move(struct drm_crtc *crtc, int x, int y)
 				   du->cursor_y + du->hotspot_y);
 
 	drm_modeset_unlock_all(dev_priv->dev);
-	drm_modeset_lock(&crtc->mutex, NULL);
+	drm_modeset_lock_crtc(crtc, crtc->cursor);
 
 	return 0;
 }
@@ -1828,9 +1828,7 @@ vmw_du_connector_detect(struct drm_connector *connector, bool force)
 	struct vmw_private *dev_priv = vmw_priv(dev);
 	struct vmw_display_unit *du = vmw_connector_to_du(connector);
 
-	mutex_lock(&dev_priv->hw_mutex);
 	num_displays = vmw_read(dev_priv, SVGA_REG_NUM_DISPLAYS);
-	mutex_unlock(&dev_priv->hw_mutex);
 
 	return ((vmw_connector_to_du(connector)->unit < num_displays &&
 		 du->pref_active) ?
@@ -1950,6 +1948,14 @@ int vmw_du_connector_fill_modes(struct drm_connector *connector,
 		DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_PVSYNC)
 	};
 	int i;
+	u32 assumed_bpp = 2;
+
+	/*
+	 * If using screen objects, then assume 32-bpp because that's what the
+	 * SVGA device is assuming
+	 */
+	if (dev_priv->sou_priv)
+		assumed_bpp = 4;
 
 	/* Add preferred mode */
 	{
@@ -1960,8 +1966,9 @@ int vmw_du_connector_fill_modes(struct drm_connector *connector,
 		mode->vdisplay = du->pref_height;
 		vmw_guess_mode_timing(mode);
 
-		if (vmw_kms_validate_mode_vram(dev_priv, mode->hdisplay * 2,
-					       mode->vdisplay)) {
+		if (vmw_kms_validate_mode_vram(dev_priv,
+						mode->hdisplay * assumed_bpp,
+						mode->vdisplay)) {
 			drm_mode_probed_add(connector, mode);
 		} else {
 			drm_mode_destroy(dev, mode);
@@ -1983,7 +1990,8 @@ int vmw_du_connector_fill_modes(struct drm_connector *connector,
 		    bmode->vdisplay > max_height)
 			continue;
 
-		if (!vmw_kms_validate_mode_vram(dev_priv, bmode->hdisplay * 2,
+		if (!vmw_kms_validate_mode_vram(dev_priv,
+						bmode->hdisplay * assumed_bpp,
 						bmode->vdisplay))
 			continue;
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c
index 15e185ae4c99..5c289f748ab4 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c
@@ -26,6 +26,7 @@
  **************************************************************************/
 
 #include "vmwgfx_kms.h"
+#include <drm/drm_plane_helper.h>
 
 
 #define vmw_crtc_to_ldu(x) \
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
index a432c0db257c..210ef15b1d09 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
@@ -133,6 +133,7 @@ static void vmw_resource_release(struct kref *kref)
 			struct ttm_validate_buffer val_buf;
 
 			val_buf.bo = bo;
+			val_buf.shared = false;
 			res->func->unbind(res, false, &val_buf);
 		}
 		res->backup_dirty = false;
@@ -429,7 +430,7 @@ int vmw_dmabuf_init(struct vmw_private *dev_priv,
 	ret = ttm_bo_init(bdev, &vmw_bo->base, size,
 			  ttm_bo_type_device, placement,
 			  0, interruptible,
-			  NULL, acc_size, NULL, bo_free);
+			  NULL, acc_size, NULL, NULL, bo_free);
 	return ret;
 }
 
@@ -567,13 +568,18 @@ static int vmw_user_dmabuf_synccpu_grab(struct vmw_user_dma_buffer *user_bo,
 	int ret;
 
 	if (flags & drm_vmw_synccpu_allow_cs) {
-		struct ttm_bo_device *bdev = bo->bdev;
+		bool nonblock = !!(flags & drm_vmw_synccpu_dontblock);
+		long lret;
 
-		spin_lock(&bdev->fence_lock);
-		ret = ttm_bo_wait(bo, false, true,
-				  !!(flags & drm_vmw_synccpu_dontblock));
-		spin_unlock(&bdev->fence_lock);
-		return ret;
+		if (nonblock)
+			return reservation_object_test_signaled_rcu(bo->resv, true) ? 0 : -EBUSY;
+
+		lret = reservation_object_wait_timeout_rcu(bo->resv, true, true, MAX_SCHEDULE_TIMEOUT);
+		if (!lret)
+			return -EBUSY;
+		else if (lret < 0)
+			return lret;
+		return 0;
 	}
 
 	ret = ttm_bo_synccpu_write_grab
@@ -1214,8 +1220,9 @@ vmw_resource_check_buffer(struct vmw_resource *res,
 
 	INIT_LIST_HEAD(&val_list);
 	val_buf->bo = ttm_bo_reference(&res->backup->base);
+	val_buf->shared = false;
 	list_add_tail(&val_buf->head, &val_list);
-	ret = ttm_eu_reserve_buffers(NULL, &val_list);
+	ret = ttm_eu_reserve_buffers(NULL, &val_list, interruptible, NULL);
 	if (unlikely(ret != 0))
 		goto out_no_reserve;
 
@@ -1307,6 +1314,7 @@ int vmw_resource_do_evict(struct vmw_resource *res, bool interruptible)
 	BUG_ON(!func->may_evict);
 
 	val_buf.bo = NULL;
+	val_buf.shared = false;
 	ret = vmw_resource_check_buffer(res, interruptible, &val_buf);
 	if (unlikely(ret != 0))
 		return ret;
@@ -1352,6 +1360,7 @@ int vmw_resource_validate(struct vmw_resource *res)
 		return 0;
 
 	val_buf.bo = NULL;
+	val_buf.shared = false;
 	if (res->backup)
 		val_buf.bo = &res->backup->base;
 	do {
@@ -1419,25 +1428,16 @@ void vmw_fence_single_bo(struct ttm_buffer_object *bo,
 			 struct vmw_fence_obj *fence)
 {
 	struct ttm_bo_device *bdev = bo->bdev;
-	struct ttm_bo_driver *driver = bdev->driver;
-	struct vmw_fence_obj *old_fence_obj;
+
 	struct vmw_private *dev_priv =
 		container_of(bdev, struct vmw_private, bdev);
 
-	if (fence == NULL)
+	if (fence == NULL) {
 		vmw_execbuf_fence_commands(NULL, dev_priv, &fence, NULL);
-	else
-		driver->sync_obj_ref(fence);
-
-	spin_lock(&bdev->fence_lock);
-
-	old_fence_obj = bo->sync_obj;
-	bo->sync_obj = fence;
-
-	spin_unlock(&bdev->fence_lock);
-
-	if (old_fence_obj)
-		vmw_fence_obj_unreference(&old_fence_obj);
+		reservation_object_add_excl_fence(bo->resv, &fence->base);
+		fence_put(&fence->base);
+	} else
+		reservation_object_add_excl_fence(bo->resv, &fence->base);
 }
 
 /**
@@ -1475,10 +1475,10 @@ void vmw_resource_move_notify(struct ttm_buffer_object *bo,
 
 	if (mem->mem_type != VMW_PL_MOB) {
 		struct vmw_resource *res, *n;
-		struct ttm_bo_device *bdev = bo->bdev;
 		struct ttm_validate_buffer val_buf;
 
 		val_buf.bo = bo;
+		val_buf.shared = false;
 
 		list_for_each_entry_safe(res, n, &dma_buf->res_list, mob_head) {
 
@@ -1491,9 +1491,7 @@ void vmw_resource_move_notify(struct ttm_buffer_object *bo,
 			list_del_init(&res->mob_head);
 		}
 
-		spin_lock(&bdev->fence_lock);
 		(void) ttm_bo_wait(bo, false, false, false);
-		spin_unlock(&bdev->fence_lock);
 	}
 }
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c b/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c
index b295463a60b3..7dc591d04d9a 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c
@@ -26,6 +26,7 @@
  **************************************************************************/
 
 #include "vmwgfx_kms.h"
+#include <drm/drm_plane_helper.h>
 
 
 #define vmw_crtc_to_sou(x) \
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c b/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c
index 8719fb3cccc9..6a4584a43aa6 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c
@@ -198,7 +198,7 @@ static int vmw_gb_shader_bind(struct vmw_resource *res,
 	cmd->header.size = sizeof(cmd->body);
 	cmd->body.shid = res->id;
 	cmd->body.mobid = bo->mem.start;
-	cmd->body.offsetInBytes = 0;
+	cmd->body.offsetInBytes = res->backup_offset;
 	res->backup_dirty = false;
 	vmw_fifo_commit(dev_priv, sizeof(*cmd));
 
diff --git a/drivers/gpu/host1x/cdma.c b/drivers/gpu/host1x/cdma.c
index 3995255b16c7..5a8c8d55317a 100644
--- a/drivers/gpu/host1x/cdma.c
+++ b/drivers/gpu/host1x/cdma.c
@@ -97,7 +97,7 @@ fail:
 static void host1x_pushbuffer_push(struct push_buffer *pb, u32 op1, u32 op2)
 {
 	u32 pos = pb->pos;
-	u32 *p = (u32 *)((u32)pb->mapped + pos);
+	u32 *p = (u32 *)((void *)pb->mapped + pos);
 	WARN_ON(pos == pb->fence);
 	*(p++) = op1;
 	*(p++) = op2;
diff --git a/drivers/gpu/host1x/cdma.h b/drivers/gpu/host1x/cdma.h
index 313c4b784348..470087af8fe5 100644
--- a/drivers/gpu/host1x/cdma.h
+++ b/drivers/gpu/host1x/cdma.h
@@ -42,7 +42,7 @@ struct host1x_job;
  */
 
 struct push_buffer {
-	u32 *mapped;			/* mapped pushbuffer memory */
+	void *mapped;			/* mapped pushbuffer memory */
 	dma_addr_t phys;		/* physical address of pushbuffer */
 	u32 fence;			/* index we've written */
 	u32 pos;			/* index to write to */
diff --git a/drivers/gpu/host1x/hw/cdma_hw.c b/drivers/gpu/host1x/hw/cdma_hw.c
index 6b09b71940c2..305ea8f3382d 100644
--- a/drivers/gpu/host1x/hw/cdma_hw.c
+++ b/drivers/gpu/host1x/hw/cdma_hw.c
@@ -26,11 +26,11 @@
 #include "../debug.h"
 
 /*
- * Put the restart at the end of pushbuffer memor
+ * Put the restart at the end of pushbuffer memory
  */
 static void push_buffer_init(struct push_buffer *pb)
 {
-	*(pb->mapped + (pb->size_bytes >> 2)) = host1x_opcode_restart(0);
+	*(u32 *)(pb->mapped + pb->size_bytes) = host1x_opcode_restart(0);
 }
 
 /*
@@ -51,11 +51,11 @@ static void cdma_timeout_cpu_incr(struct host1x_cdma *cdma, u32 getptr,
 
 	/* NOP all the PB slots */
 	while (nr_slots--) {
-		u32 *p = (u32 *)((u32)pb->mapped + getptr);
+		u32 *p = (u32 *)(pb->mapped + getptr);
 		*(p++) = HOST1X_OPCODE_NOP;
 		*(p++) = HOST1X_OPCODE_NOP;
-		dev_dbg(host1x->dev, "%s: NOP at %#llx\n", __func__,
-			(u64)pb->phys + getptr);
+		dev_dbg(host1x->dev, "%s: NOP at %pad+%#x\n", __func__,
+			&pb->phys, getptr);
 		getptr = (getptr + 8) & (pb->size_bytes - 1);
 	}
 	wmb();
diff --git a/drivers/gpu/host1x/hw/channel_hw.c b/drivers/gpu/host1x/hw/channel_hw.c
index 4608257ab656..946c332c3906 100644
--- a/drivers/gpu/host1x/hw/channel_hw.c
+++ b/drivers/gpu/host1x/hw/channel_hw.c
@@ -32,6 +32,7 @@
 static void trace_write_gather(struct host1x_cdma *cdma, struct host1x_bo *bo,
 			       u32 offset, u32 words)
 {
+	struct device *dev = cdma_to_channel(cdma)->dev;
 	void *mem = NULL;
 
 	if (host1x_debug_trace_cmdbuf)
@@ -44,11 +45,14 @@ static void trace_write_gather(struct host1x_cdma *cdma, struct host1x_bo *bo,
 		 * of how much you can output to ftrace at once.
 		 */
 		for (i = 0; i < words; i += TRACE_MAX_LENGTH) {
-			trace_host1x_cdma_push_gather(
-				dev_name(cdma_to_channel(cdma)->dev),
-				(u32)bo, min(words - i, TRACE_MAX_LENGTH),
-				offset + i * sizeof(u32), mem);
+			u32 num_words = min(words - i, TRACE_MAX_LENGTH);
+			offset += i * sizeof(u32);
+
+			trace_host1x_cdma_push_gather(dev_name(dev), bo,
+						      num_words, offset,
+						      mem);
 		}
+
 		host1x_bo_munmap(bo, mem);
 	}
 }
diff --git a/drivers/gpu/host1x/hw/debug_hw.c b/drivers/gpu/host1x/hw/debug_hw.c
index f72c873eff81..791de9351eeb 100644
--- a/drivers/gpu/host1x/hw/debug_hw.c
+++ b/drivers/gpu/host1x/hw/debug_hw.c
@@ -163,8 +163,8 @@ static void show_channel_gathers(struct output *o, struct host1x_cdma *cdma)
 				continue;
 			}
 
-			host1x_debug_output(o, "    GATHER at %#llx+%04x, %d words\n",
-					    (u64)g->base, g->offset, g->words);
+			host1x_debug_output(o, "    GATHER at %pad+%#x, %d words\n",
+					    &g->base, g->offset, g->words);
 
 			show_gather(o, g->base + g->offset, g->words, cdma,
 				    g->base, mapped);
diff --git a/drivers/gpu/host1x/job.h b/drivers/gpu/host1x/job.h
index 33a697d6dcef..8b3c15df0660 100644
--- a/drivers/gpu/host1x/job.h
+++ b/drivers/gpu/host1x/job.h
@@ -23,7 +23,7 @@ struct host1x_job_gather {
 	u32 words;
 	dma_addr_t base;
 	struct host1x_bo *bo;
-	int offset;
+	u32 offset;
 	bool handled;
 };
 
diff --git a/drivers/gpu/host1x/mipi.c b/drivers/gpu/host1x/mipi.c
index 9882ea122024..fbc6ee6ca337 100644
--- a/drivers/gpu/host1x/mipi.c
+++ b/drivers/gpu/host1x/mipi.c
@@ -49,35 +49,47 @@
 #define MIPI_CAL_CONFIG_DSIC		0x10
 #define MIPI_CAL_CONFIG_DSID		0x11
 
+#define MIPI_CAL_CONFIG_DSIAB_CLK	0x19
+#define MIPI_CAL_CONFIG_DSICD_CLK	0x1a
+#define MIPI_CAL_CONFIG_CSIAB_CLK	0x1b
+#define MIPI_CAL_CONFIG_CSICD_CLK	0x1c
+#define MIPI_CAL_CONFIG_CSIE_CLK	0x1d
+
+/* for data and clock lanes */
 #define MIPI_CAL_CONFIG_SELECT		(1 << 21)
+
+/* for data lanes */
 #define MIPI_CAL_CONFIG_HSPDOS(x)	(((x) & 0x1f) << 16)
 #define MIPI_CAL_CONFIG_HSPUOS(x)	(((x) & 0x1f) <<  8)
 #define MIPI_CAL_CONFIG_TERMOS(x)	(((x) & 0x1f) <<  0)
 
+/* for clock lanes */
+#define MIPI_CAL_CONFIG_HSCLKPDOSD(x)	(((x) & 0x1f) <<  8)
+#define MIPI_CAL_CONFIG_HSCLKPUOSD(x)	(((x) & 0x1f) <<  0)
+
 #define MIPI_CAL_BIAS_PAD_CFG0		0x16
 #define MIPI_CAL_BIAS_PAD_PDVCLAMP	(1 << 1)
 #define MIPI_CAL_BIAS_PAD_E_VCLAMP_REF	(1 << 0)
 
 #define MIPI_CAL_BIAS_PAD_CFG1		0x17
+#define MIPI_CAL_BIAS_PAD_DRV_DN_REF(x) (((x) & 0x7) << 16)
 
 #define MIPI_CAL_BIAS_PAD_CFG2		0x18
 #define MIPI_CAL_BIAS_PAD_PDVREG	(1 << 1)
 
-static const struct module {
-	unsigned long reg;
-} modules[] = {
-	{ .reg = MIPI_CAL_CONFIG_CSIA },
-	{ .reg = MIPI_CAL_CONFIG_CSIB },
-	{ .reg = MIPI_CAL_CONFIG_CSIC },
-	{ .reg = MIPI_CAL_CONFIG_CSID },
-	{ .reg = MIPI_CAL_CONFIG_CSIE },
-	{ .reg = MIPI_CAL_CONFIG_DSIA },
-	{ .reg = MIPI_CAL_CONFIG_DSIB },
-	{ .reg = MIPI_CAL_CONFIG_DSIC },
-	{ .reg = MIPI_CAL_CONFIG_DSID },
+struct tegra_mipi_pad {
+	unsigned long data;
+	unsigned long clk;
+};
+
+struct tegra_mipi_soc {
+	bool has_clk_lane;
+	const struct tegra_mipi_pad *pads;
+	unsigned int num_pads;
 };
 
 struct tegra_mipi {
+	const struct tegra_mipi_soc *soc;
 	void __iomem *regs;
 	struct mutex lock;
 	struct clk *clk;
@@ -90,16 +102,16 @@ struct tegra_mipi_device {
 	unsigned long pads;
 };
 
-static inline unsigned long tegra_mipi_readl(struct tegra_mipi *mipi,
-					     unsigned long reg)
+static inline u32 tegra_mipi_readl(struct tegra_mipi *mipi,
+				   unsigned long offset)
 {
-	return readl(mipi->regs + (reg << 2));
+	return readl(mipi->regs + (offset << 2));
 }
 
-static inline void tegra_mipi_writel(struct tegra_mipi *mipi,
-				     unsigned long value, unsigned long reg)
+static inline void tegra_mipi_writel(struct tegra_mipi *mipi, u32 value,
+				     unsigned long offset)
 {
-	writel(value, mipi->regs + (reg << 2));
+	writel(value, mipi->regs + (offset << 2));
 }
 
 struct tegra_mipi_device *tegra_mipi_request(struct device *device)
@@ -117,36 +129,35 @@ struct tegra_mipi_device *tegra_mipi_request(struct device *device)
 
 	dev = kzalloc(sizeof(*dev), GFP_KERNEL);
 	if (!dev) {
-		of_node_put(args.np);
 		err = -ENOMEM;
 		goto out;
 	}
 
 	dev->pdev = of_find_device_by_node(args.np);
 	if (!dev->pdev) {
-		of_node_put(args.np);
 		err = -ENODEV;
 		goto free;
 	}
 
-	of_node_put(args.np);
-
 	dev->mipi = platform_get_drvdata(dev->pdev);
 	if (!dev->mipi) {
 		err = -EPROBE_DEFER;
-		goto pdev_put;
+		goto put;
 	}
 
+	of_node_put(args.np);
+
 	dev->pads = args.args[0];
 	dev->device = device;
 
 	return dev;
 
-pdev_put:
+put:
 	platform_device_put(dev->pdev);
 free:
 	kfree(dev);
 out:
+	of_node_put(args.np);
 	return ERR_PTR(err);
 }
 EXPORT_SYMBOL(tegra_mipi_request);
@@ -161,7 +172,7 @@ EXPORT_SYMBOL(tegra_mipi_free);
 static int tegra_mipi_wait(struct tegra_mipi *mipi)
 {
 	unsigned long timeout = jiffies + msecs_to_jiffies(250);
-	unsigned long value;
+	u32 value;
 
 	while (time_before(jiffies, timeout)) {
 		value = tegra_mipi_readl(mipi, MIPI_CAL_STATUS);
@@ -177,8 +188,9 @@ static int tegra_mipi_wait(struct tegra_mipi *mipi)
 
 int tegra_mipi_calibrate(struct tegra_mipi_device *device)
 {
-	unsigned long value;
+	const struct tegra_mipi_soc *soc = device->mipi->soc;
 	unsigned int i;
+	u32 value;
 	int err;
 
 	err = clk_enable(device->mipi->clk);
@@ -192,23 +204,35 @@ int tegra_mipi_calibrate(struct tegra_mipi_device *device)
 	value |= MIPI_CAL_BIAS_PAD_E_VCLAMP_REF;
 	tegra_mipi_writel(device->mipi, value, MIPI_CAL_BIAS_PAD_CFG0);
 
+	tegra_mipi_writel(device->mipi, MIPI_CAL_BIAS_PAD_DRV_DN_REF(2),
+			  MIPI_CAL_BIAS_PAD_CFG1);
+
 	value = tegra_mipi_readl(device->mipi, MIPI_CAL_BIAS_PAD_CFG2);
 	value &= ~MIPI_CAL_BIAS_PAD_PDVREG;
 	tegra_mipi_writel(device->mipi, value, MIPI_CAL_BIAS_PAD_CFG2);
 
-	for (i = 0; i < ARRAY_SIZE(modules); i++) {
-		if (device->pads & BIT(i))
-			value = MIPI_CAL_CONFIG_SELECT |
-				MIPI_CAL_CONFIG_HSPDOS(0) |
-				MIPI_CAL_CONFIG_HSPUOS(4) |
-				MIPI_CAL_CONFIG_TERMOS(5);
-		else
-			value = 0;
+	for (i = 0; i < soc->num_pads; i++) {
+		u32 clk = 0, data = 0;
+
+		if (device->pads & BIT(i)) {
+			data = MIPI_CAL_CONFIG_SELECT |
+			       MIPI_CAL_CONFIG_HSPDOS(0) |
+			       MIPI_CAL_CONFIG_HSPUOS(4) |
+			       MIPI_CAL_CONFIG_TERMOS(5);
+			clk = MIPI_CAL_CONFIG_SELECT |
+			      MIPI_CAL_CONFIG_HSCLKPDOSD(0) |
+			      MIPI_CAL_CONFIG_HSCLKPUOSD(4);
+		}
 
-		tegra_mipi_writel(device->mipi, value, modules[i].reg);
+		tegra_mipi_writel(device->mipi, data, soc->pads[i].data);
+
+		if (soc->has_clk_lane)
+			tegra_mipi_writel(device->mipi, clk, soc->pads[i].clk);
 	}
 
-	tegra_mipi_writel(device->mipi, MIPI_CAL_CTRL_START, MIPI_CAL_CTRL);
+	value = tegra_mipi_readl(device->mipi, MIPI_CAL_CTRL);
+	value |= MIPI_CAL_CTRL_START;
+	tegra_mipi_writel(device->mipi, value, MIPI_CAL_CTRL);
 
 	err = tegra_mipi_wait(device->mipi);
 
@@ -219,16 +243,63 @@ int tegra_mipi_calibrate(struct tegra_mipi_device *device)
 }
 EXPORT_SYMBOL(tegra_mipi_calibrate);
 
+static const struct tegra_mipi_pad tegra114_mipi_pads[] = {
+	{ .data = MIPI_CAL_CONFIG_CSIA },
+	{ .data = MIPI_CAL_CONFIG_CSIB },
+	{ .data = MIPI_CAL_CONFIG_CSIC },
+	{ .data = MIPI_CAL_CONFIG_CSID },
+	{ .data = MIPI_CAL_CONFIG_CSIE },
+	{ .data = MIPI_CAL_CONFIG_DSIA },
+	{ .data = MIPI_CAL_CONFIG_DSIB },
+	{ .data = MIPI_CAL_CONFIG_DSIC },
+	{ .data = MIPI_CAL_CONFIG_DSID },
+};
+
+static const struct tegra_mipi_soc tegra114_mipi_soc = {
+	.has_clk_lane = false,
+	.pads = tegra114_mipi_pads,
+	.num_pads = ARRAY_SIZE(tegra114_mipi_pads),
+};
+
+static const struct tegra_mipi_pad tegra124_mipi_pads[] = {
+	{ .data = MIPI_CAL_CONFIG_CSIA, .clk = MIPI_CAL_CONFIG_CSIAB_CLK },
+	{ .data = MIPI_CAL_CONFIG_CSIB, .clk = MIPI_CAL_CONFIG_CSIAB_CLK },
+	{ .data = MIPI_CAL_CONFIG_CSIC, .clk = MIPI_CAL_CONFIG_CSICD_CLK },
+	{ .data = MIPI_CAL_CONFIG_CSID, .clk = MIPI_CAL_CONFIG_CSICD_CLK },
+	{ .data = MIPI_CAL_CONFIG_CSIE, .clk = MIPI_CAL_CONFIG_CSIE_CLK },
+	{ .data = MIPI_CAL_CONFIG_DSIA, .clk = MIPI_CAL_CONFIG_DSIAB_CLK },
+	{ .data = MIPI_CAL_CONFIG_DSIB, .clk = MIPI_CAL_CONFIG_DSIAB_CLK },
+};
+
+static const struct tegra_mipi_soc tegra124_mipi_soc = {
+	.has_clk_lane = true,
+	.pads = tegra124_mipi_pads,
+	.num_pads = ARRAY_SIZE(tegra124_mipi_pads),
+};
+
+static struct of_device_id tegra_mipi_of_match[] = {
+	{ .compatible = "nvidia,tegra114-mipi", .data = &tegra114_mipi_soc },
+	{ .compatible = "nvidia,tegra124-mipi", .data = &tegra124_mipi_soc },
+	{ },
+};
+
 static int tegra_mipi_probe(struct platform_device *pdev)
 {
+	const struct of_device_id *match;
 	struct tegra_mipi *mipi;
 	struct resource *res;
 	int err;
 
+	match = of_match_node(tegra_mipi_of_match, pdev->dev.of_node);
+	if (!match)
+		return -ENODEV;
+
 	mipi = devm_kzalloc(&pdev->dev, sizeof(*mipi), GFP_KERNEL);
 	if (!mipi)
 		return -ENOMEM;
 
+	mipi->soc = match->data;
+
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	mipi->regs = devm_ioremap_resource(&pdev->dev, res);
 	if (IS_ERR(mipi->regs))
@@ -260,11 +331,6 @@ static int tegra_mipi_remove(struct platform_device *pdev)
 	return 0;
 }
 
-static struct of_device_id tegra_mipi_of_match[] = {
-	{ .compatible = "nvidia,tegra114-mipi", },
-	{ },
-};
-
 struct platform_driver tegra_mipi_driver = {
 	.driver = {
 		.name = "tegra-mipi",
diff --git a/drivers/gpu/ipu-v3/Kconfig b/drivers/gpu/ipu-v3/Kconfig
index 2f228a2f2a48..aefdff95356d 100644
--- a/drivers/gpu/ipu-v3/Kconfig
+++ b/drivers/gpu/ipu-v3/Kconfig
@@ -1,7 +1,8 @@
 config IMX_IPUV3_CORE
 	tristate "IPUv3 core support"
-	depends on SOC_IMX5 || SOC_IMX6Q || SOC_IMX6SL || ARCH_MULTIPLATFORM
+	depends on SOC_IMX5 || SOC_IMX6Q || ARCH_MULTIPLATFORM
 	depends on RESET_CONTROLLER
+	select GENERIC_IRQ_CHIP
 	help
 	  Choose this if you have a i.MX5/6 system and want to use the Image
 	  Processing Unit. This option only enables IPU base support.
diff --git a/drivers/gpu/ipu-v3/Makefile b/drivers/gpu/ipu-v3/Makefile
index 1887972b4ac2..107ec236a4a6 100644
--- a/drivers/gpu/ipu-v3/Makefile
+++ b/drivers/gpu/ipu-v3/Makefile
@@ -1,3 +1,4 @@
 obj-$(CONFIG_IMX_IPUV3_CORE) += imx-ipu-v3.o
 
-imx-ipu-v3-objs := ipu-common.o ipu-dc.o ipu-di.o ipu-dp.o ipu-dmfc.o ipu-smfc.o
+imx-ipu-v3-objs := ipu-common.o ipu-cpmem.o ipu-csi.o ipu-dc.o ipu-di.o \
+		ipu-dp.o ipu-dmfc.o ipu-ic.o ipu-smfc.o
diff --git a/drivers/gpu/ipu-v3/ipu-common.c b/drivers/gpu/ipu-v3/ipu-common.c
index 04e7b2eafbdd..f707d25ae78f 100644
--- a/drivers/gpu/ipu-v3/ipu-common.c
+++ b/drivers/gpu/ipu-v3/ipu-common.c
@@ -44,17 +44,6 @@ static inline void ipu_cm_write(struct ipu_soc *ipu, u32 value, unsigned offset)
 	writel(value, ipu->cm_reg + offset);
 }
 
-static inline u32 ipu_idmac_read(struct ipu_soc *ipu, unsigned offset)
-{
-	return readl(ipu->idmac_reg + offset);
-}
-
-static inline void ipu_idmac_write(struct ipu_soc *ipu, u32 value,
-		unsigned offset)
-{
-	writel(value, ipu->idmac_reg + offset);
-}
-
 void ipu_srm_dp_sync_update(struct ipu_soc *ipu)
 {
 	u32 val;
@@ -65,457 +54,184 @@ void ipu_srm_dp_sync_update(struct ipu_soc *ipu)
 }
 EXPORT_SYMBOL_GPL(ipu_srm_dp_sync_update);
 
-struct ipu_ch_param __iomem *ipu_get_cpmem(struct ipuv3_channel *channel)
-{
-	struct ipu_soc *ipu = channel->ipu;
-
-	return ipu->cpmem_base + channel->num;
-}
-EXPORT_SYMBOL_GPL(ipu_get_cpmem);
-
-void ipu_cpmem_set_high_priority(struct ipuv3_channel *channel)
-{
-	struct ipu_soc *ipu = channel->ipu;
-	struct ipu_ch_param __iomem *p = ipu_get_cpmem(channel);
-	u32 val;
-
-	if (ipu->ipu_type == IPUV3EX)
-		ipu_ch_param_write_field(p, IPU_FIELD_ID, 1);
-
-	val = ipu_idmac_read(ipu, IDMAC_CHA_PRI(channel->num));
-	val |= 1 << (channel->num % 32);
-	ipu_idmac_write(ipu, val, IDMAC_CHA_PRI(channel->num));
-};
-EXPORT_SYMBOL_GPL(ipu_cpmem_set_high_priority);
-
-void ipu_ch_param_write_field(struct ipu_ch_param __iomem *base, u32 wbs, u32 v)
-{
-	u32 bit = (wbs >> 8) % 160;
-	u32 size = wbs & 0xff;
-	u32 word = (wbs >> 8) / 160;
-	u32 i = bit / 32;
-	u32 ofs = bit % 32;
-	u32 mask = (1 << size) - 1;
-	u32 val;
-
-	pr_debug("%s %d %d %d\n", __func__, word, bit , size);
-
-	val = readl(&base->word[word].data[i]);
-	val &= ~(mask << ofs);
-	val |= v << ofs;
-	writel(val, &base->word[word].data[i]);
-
-	if ((bit + size - 1) / 32 > i) {
-		val = readl(&base->word[word].data[i + 1]);
-		val &= ~(mask >> (ofs ? (32 - ofs) : 0));
-		val |= v >> (ofs ? (32 - ofs) : 0);
-		writel(val, &base->word[word].data[i + 1]);
-	}
-}
-EXPORT_SYMBOL_GPL(ipu_ch_param_write_field);
-
-u32 ipu_ch_param_read_field(struct ipu_ch_param __iomem *base, u32 wbs)
-{
-	u32 bit = (wbs >> 8) % 160;
-	u32 size = wbs & 0xff;
-	u32 word = (wbs >> 8) / 160;
-	u32 i = bit / 32;
-	u32 ofs = bit % 32;
-	u32 mask = (1 << size) - 1;
-	u32 val = 0;
-
-	pr_debug("%s %d %d %d\n", __func__, word, bit , size);
-
-	val = (readl(&base->word[word].data[i]) >> ofs) & mask;
-
-	if ((bit + size - 1) / 32 > i) {
-		u32 tmp;
-		tmp = readl(&base->word[word].data[i + 1]);
-		tmp &= mask >> (ofs ? (32 - ofs) : 0);
-		val |= tmp << (ofs ? (32 - ofs) : 0);
-	}
-
-	return val;
-}
-EXPORT_SYMBOL_GPL(ipu_ch_param_read_field);
-
-int ipu_cpmem_set_format_rgb(struct ipu_ch_param __iomem *p,
-		const struct ipu_rgb *rgb)
-{
-	int bpp = 0, npb = 0, ro, go, bo, to;
-
-	ro = rgb->bits_per_pixel - rgb->red.length - rgb->red.offset;
-	go = rgb->bits_per_pixel - rgb->green.length - rgb->green.offset;
-	bo = rgb->bits_per_pixel - rgb->blue.length - rgb->blue.offset;
-	to = rgb->bits_per_pixel - rgb->transp.length - rgb->transp.offset;
-
-	ipu_ch_param_write_field(p, IPU_FIELD_WID0, rgb->red.length - 1);
-	ipu_ch_param_write_field(p, IPU_FIELD_OFS0, ro);
-	ipu_ch_param_write_field(p, IPU_FIELD_WID1, rgb->green.length - 1);
-	ipu_ch_param_write_field(p, IPU_FIELD_OFS1, go);
-	ipu_ch_param_write_field(p, IPU_FIELD_WID2, rgb->blue.length - 1);
-	ipu_ch_param_write_field(p, IPU_FIELD_OFS2, bo);
-
-	if (rgb->transp.length) {
-		ipu_ch_param_write_field(p, IPU_FIELD_WID3,
-				rgb->transp.length - 1);
-		ipu_ch_param_write_field(p, IPU_FIELD_OFS3, to);
-	} else {
-		ipu_ch_param_write_field(p, IPU_FIELD_WID3, 7);
-		ipu_ch_param_write_field(p, IPU_FIELD_OFS3,
-				rgb->bits_per_pixel);
-	}
-
-	switch (rgb->bits_per_pixel) {
-	case 32:
-		bpp = 0;
-		npb = 15;
-		break;
-	case 24:
-		bpp = 1;
-		npb = 19;
-		break;
-	case 16:
-		bpp = 3;
-		npb = 31;
-		break;
-	case 8:
-		bpp = 5;
-		npb = 63;
-		break;
-	default:
-		return -EINVAL;
-	}
-	ipu_ch_param_write_field(p, IPU_FIELD_BPP, bpp);
-	ipu_ch_param_write_field(p, IPU_FIELD_NPB, npb);
-	ipu_ch_param_write_field(p, IPU_FIELD_PFS, 7); /* rgb mode */
-
-	return 0;
-}
-EXPORT_SYMBOL_GPL(ipu_cpmem_set_format_rgb);
-
-int ipu_cpmem_set_format_passthrough(struct ipu_ch_param __iomem *p,
-		int width)
+enum ipu_color_space ipu_drm_fourcc_to_colorspace(u32 drm_fourcc)
 {
-	int bpp = 0, npb = 0;
-
-	switch (width) {
-	case 32:
-		bpp = 0;
-		npb = 15;
-		break;
-	case 24:
-		bpp = 1;
-		npb = 19;
-		break;
-	case 16:
-		bpp = 3;
-		npb = 31;
-		break;
-	case 8:
-		bpp = 5;
-		npb = 63;
-		break;
+	switch (drm_fourcc) {
+	case DRM_FORMAT_RGB565:
+	case DRM_FORMAT_BGR565:
+	case DRM_FORMAT_RGB888:
+	case DRM_FORMAT_BGR888:
+	case DRM_FORMAT_XRGB8888:
+	case DRM_FORMAT_XBGR8888:
+	case DRM_FORMAT_RGBX8888:
+	case DRM_FORMAT_BGRX8888:
+	case DRM_FORMAT_ARGB8888:
+	case DRM_FORMAT_ABGR8888:
+	case DRM_FORMAT_RGBA8888:
+	case DRM_FORMAT_BGRA8888:
+		return IPUV3_COLORSPACE_RGB;
+	case DRM_FORMAT_YUYV:
+	case DRM_FORMAT_UYVY:
+	case DRM_FORMAT_YUV420:
+	case DRM_FORMAT_YVU420:
+	case DRM_FORMAT_YUV422:
+	case DRM_FORMAT_YVU422:
+	case DRM_FORMAT_NV12:
+	case DRM_FORMAT_NV21:
+	case DRM_FORMAT_NV16:
+	case DRM_FORMAT_NV61:
+		return IPUV3_COLORSPACE_YUV;
 	default:
-		return -EINVAL;
+		return IPUV3_COLORSPACE_UNKNOWN;
 	}
-
-	ipu_ch_param_write_field(p, IPU_FIELD_BPP, bpp);
-	ipu_ch_param_write_field(p, IPU_FIELD_NPB, npb);
-	ipu_ch_param_write_field(p, IPU_FIELD_PFS, 6); /* raw mode */
-
-	return 0;
 }
-EXPORT_SYMBOL_GPL(ipu_cpmem_set_format_passthrough);
+EXPORT_SYMBOL_GPL(ipu_drm_fourcc_to_colorspace);
 
-void ipu_cpmem_set_yuv_interleaved(struct ipu_ch_param __iomem *p,
-				   u32 pixel_format)
+enum ipu_color_space ipu_pixelformat_to_colorspace(u32 pixelformat)
 {
-	switch (pixel_format) {
+	switch (pixelformat) {
+	case V4L2_PIX_FMT_YUV420:
+	case V4L2_PIX_FMT_YVU420:
+	case V4L2_PIX_FMT_YUV422P:
 	case V4L2_PIX_FMT_UYVY:
-		ipu_ch_param_write_field(p, IPU_FIELD_BPP, 3);    /* bits/pixel */
-		ipu_ch_param_write_field(p, IPU_FIELD_PFS, 0xA);  /* pix format */
-		ipu_ch_param_write_field(p, IPU_FIELD_NPB, 31);   /* burst size */
-		break;
 	case V4L2_PIX_FMT_YUYV:
-		ipu_ch_param_write_field(p, IPU_FIELD_BPP, 3);    /* bits/pixel */
-		ipu_ch_param_write_field(p, IPU_FIELD_PFS, 0x8);  /* pix format */
-		ipu_ch_param_write_field(p, IPU_FIELD_NPB, 31);   /* burst size */
-		break;
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_NV21:
+	case V4L2_PIX_FMT_NV16:
+	case V4L2_PIX_FMT_NV61:
+		return IPUV3_COLORSPACE_YUV;
+	case V4L2_PIX_FMT_RGB32:
+	case V4L2_PIX_FMT_BGR32:
+	case V4L2_PIX_FMT_RGB24:
+	case V4L2_PIX_FMT_BGR24:
+	case V4L2_PIX_FMT_RGB565:
+		return IPUV3_COLORSPACE_RGB;
+	default:
+		return IPUV3_COLORSPACE_UNKNOWN;
 	}
 }
-EXPORT_SYMBOL_GPL(ipu_cpmem_set_yuv_interleaved);
+EXPORT_SYMBOL_GPL(ipu_pixelformat_to_colorspace);
 
-void ipu_cpmem_set_yuv_planar_full(struct ipu_ch_param __iomem *p,
-		u32 pixel_format, int stride, int u_offset, int v_offset)
+bool ipu_pixelformat_is_planar(u32 pixelformat)
 {
-	switch (pixel_format) {
+	switch (pixelformat) {
 	case V4L2_PIX_FMT_YUV420:
-		ipu_ch_param_write_field(p, IPU_FIELD_SLUV, (stride / 2) - 1);
-		ipu_ch_param_write_field(p, IPU_FIELD_UBO, u_offset / 8);
-		ipu_ch_param_write_field(p, IPU_FIELD_VBO, v_offset / 8);
-		break;
 	case V4L2_PIX_FMT_YVU420:
-		ipu_ch_param_write_field(p, IPU_FIELD_SLUV, (stride / 2) - 1);
-		ipu_ch_param_write_field(p, IPU_FIELD_UBO, v_offset / 8);
-		ipu_ch_param_write_field(p, IPU_FIELD_VBO, u_offset / 8);
-		break;
+	case V4L2_PIX_FMT_YUV422P:
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_NV21:
+	case V4L2_PIX_FMT_NV16:
+	case V4L2_PIX_FMT_NV61:
+		return true;
 	}
-}
-EXPORT_SYMBOL_GPL(ipu_cpmem_set_yuv_planar_full);
-
-void ipu_cpmem_set_yuv_planar(struct ipu_ch_param __iomem *p, u32 pixel_format,
-		int stride, int height)
-{
-	int u_offset, v_offset;
-	int uv_stride = 0;
 
-	switch (pixel_format) {
-	case V4L2_PIX_FMT_YUV420:
-	case V4L2_PIX_FMT_YVU420:
-		uv_stride = stride / 2;
-		u_offset = stride * height;
-		v_offset = u_offset + (uv_stride * height / 2);
-		ipu_cpmem_set_yuv_planar_full(p, pixel_format, stride,
-				u_offset, v_offset);
-		break;
-	}
+	return false;
 }
-EXPORT_SYMBOL_GPL(ipu_cpmem_set_yuv_planar);
-
-static const struct ipu_rgb def_rgb_32 = {
-	.red	= { .offset = 16, .length = 8, },
-	.green	= { .offset =  8, .length = 8, },
-	.blue	= { .offset =  0, .length = 8, },
-	.transp = { .offset = 24, .length = 8, },
-	.bits_per_pixel = 32,
-};
-
-static const struct ipu_rgb def_bgr_32 = {
-	.red	= { .offset =  0, .length = 8, },
-	.green	= { .offset =  8, .length = 8, },
-	.blue	= { .offset = 16, .length = 8, },
-	.transp = { .offset = 24, .length = 8, },
-	.bits_per_pixel = 32,
-};
-
-static const struct ipu_rgb def_rgb_24 = {
-	.red	= { .offset = 16, .length = 8, },
-	.green	= { .offset =  8, .length = 8, },
-	.blue	= { .offset =  0, .length = 8, },
-	.transp = { .offset =  0, .length = 0, },
-	.bits_per_pixel = 24,
-};
-
-static const struct ipu_rgb def_bgr_24 = {
-	.red	= { .offset =  0, .length = 8, },
-	.green	= { .offset =  8, .length = 8, },
-	.blue	= { .offset = 16, .length = 8, },
-	.transp = { .offset =  0, .length = 0, },
-	.bits_per_pixel = 24,
-};
-
-static const struct ipu_rgb def_rgb_16 = {
-	.red	= { .offset = 11, .length = 5, },
-	.green	= { .offset =  5, .length = 6, },
-	.blue	= { .offset =  0, .length = 5, },
-	.transp = { .offset =  0, .length = 0, },
-	.bits_per_pixel = 16,
-};
-
-static const struct ipu_rgb def_bgr_16 = {
-	.red	= { .offset =  0, .length = 5, },
-	.green	= { .offset =  5, .length = 6, },
-	.blue	= { .offset = 11, .length = 5, },
-	.transp = { .offset =  0, .length = 0, },
-	.bits_per_pixel = 16,
-};
-
-#define Y_OFFSET(pix, x, y)	((x) + pix->width * (y))
-#define U_OFFSET(pix, x, y)	((pix->width * pix->height) + \
-					(pix->width * (y) / 4) + (x) / 2)
-#define V_OFFSET(pix, x, y)	((pix->width * pix->height) + \
-					(pix->width * pix->height / 4) + \
-					(pix->width * (y) / 4) + (x) / 2)
+EXPORT_SYMBOL_GPL(ipu_pixelformat_is_planar);
 
-int ipu_cpmem_set_fmt(struct ipu_ch_param __iomem *cpmem, u32 drm_fourcc)
+enum ipu_color_space ipu_mbus_code_to_colorspace(u32 mbus_code)
 {
-	switch (drm_fourcc) {
-	case DRM_FORMAT_YUV420:
-	case DRM_FORMAT_YVU420:
-		/* pix format */
-		ipu_ch_param_write_field(cpmem, IPU_FIELD_PFS, 2);
-		/* burst size */
-		ipu_ch_param_write_field(cpmem, IPU_FIELD_NPB, 63);
-		break;
-	case DRM_FORMAT_UYVY:
-		/* bits/pixel */
-		ipu_ch_param_write_field(cpmem, IPU_FIELD_BPP, 3);
-		/* pix format */
-		ipu_ch_param_write_field(cpmem, IPU_FIELD_PFS, 0xA);
-		/* burst size */
-		ipu_ch_param_write_field(cpmem, IPU_FIELD_NPB, 31);
-		break;
-	case DRM_FORMAT_YUYV:
-		/* bits/pixel */
-		ipu_ch_param_write_field(cpmem, IPU_FIELD_BPP, 3);
-		/* pix format */
-		ipu_ch_param_write_field(cpmem, IPU_FIELD_PFS, 0x8);
-		/* burst size */
-		ipu_ch_param_write_field(cpmem, IPU_FIELD_NPB, 31);
-		break;
-	case DRM_FORMAT_ABGR8888:
-	case DRM_FORMAT_XBGR8888:
-		ipu_cpmem_set_format_rgb(cpmem, &def_bgr_32);
-		break;
-	case DRM_FORMAT_ARGB8888:
-	case DRM_FORMAT_XRGB8888:
-		ipu_cpmem_set_format_rgb(cpmem, &def_rgb_32);
-		break;
-	case DRM_FORMAT_BGR888:
-		ipu_cpmem_set_format_rgb(cpmem, &def_bgr_24);
-		break;
-	case DRM_FORMAT_RGB888:
-		ipu_cpmem_set_format_rgb(cpmem, &def_rgb_24);
-		break;
-	case DRM_FORMAT_RGB565:
-		ipu_cpmem_set_format_rgb(cpmem, &def_rgb_16);
-		break;
-	case DRM_FORMAT_BGR565:
-		ipu_cpmem_set_format_rgb(cpmem, &def_bgr_16);
-		break;
+	switch (mbus_code & 0xf000) {
+	case 0x1000:
+		return IPUV3_COLORSPACE_RGB;
+	case 0x2000:
+		return IPUV3_COLORSPACE_YUV;
 	default:
-		return -EINVAL;
+		return IPUV3_COLORSPACE_UNKNOWN;
 	}
-
-	return 0;
 }
-EXPORT_SYMBOL_GPL(ipu_cpmem_set_fmt);
+EXPORT_SYMBOL_GPL(ipu_mbus_code_to_colorspace);
 
-/*
- * The V4L2 spec defines packed RGB formats in memory byte order, which from
- * point of view of the IPU corresponds to little-endian words with the first
- * component in the least significant bits.
- * The DRM pixel formats and IPU internal representation are ordered the other
- * way around, with the first named component ordered at the most significant
- * bits. Further, V4L2 formats are not well defined:
- *     http://linuxtv.org/downloads/v4l-dvb-apis/packed-rgb.html
- * We choose the interpretation which matches GStreamer behavior.
- */
-static int v4l2_pix_fmt_to_drm_fourcc(u32 pixelformat)
+int ipu_stride_to_bytes(u32 pixel_stride, u32 pixelformat)
 {
 	switch (pixelformat) {
-	case V4L2_PIX_FMT_RGB565:
+	case V4L2_PIX_FMT_YUV420:
+	case V4L2_PIX_FMT_YVU420:
+	case V4L2_PIX_FMT_YUV422P:
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_NV21:
+	case V4L2_PIX_FMT_NV16:
+	case V4L2_PIX_FMT_NV61:
 		/*
-		 * Here we choose the 'corrected' interpretation of RGBP, a
-		 * little-endian 16-bit word with the red component at the most
-		 * significant bits:
-		 * g[2:0]b[4:0] r[4:0]g[5:3] <=> [16:0] R:G:B
+		 * for the planar YUV formats, the stride passed to
+		 * cpmem must be the stride in bytes of the Y plane.
+		 * And all the planar YUV formats have an 8-bit
+		 * Y component.
 		 */
-		return DRM_FORMAT_RGB565;
+		return (8 * pixel_stride) >> 3;
+	case V4L2_PIX_FMT_RGB565:
+	case V4L2_PIX_FMT_YUYV:
+	case V4L2_PIX_FMT_UYVY:
+		return (16 * pixel_stride) >> 3;
 	case V4L2_PIX_FMT_BGR24:
-		/* B G R <=> [24:0] R:G:B */
-		return DRM_FORMAT_RGB888;
 	case V4L2_PIX_FMT_RGB24:
-		/* R G B <=> [24:0] B:G:R */
-		return DRM_FORMAT_BGR888;
+		return (24 * pixel_stride) >> 3;
 	case V4L2_PIX_FMT_BGR32:
-		/* B G R A <=> [32:0] A:B:G:R */
-		return DRM_FORMAT_XRGB8888;
 	case V4L2_PIX_FMT_RGB32:
-		/* R G B A <=> [32:0] A:B:G:R */
-		return DRM_FORMAT_XBGR8888;
-	case V4L2_PIX_FMT_UYVY:
-		return DRM_FORMAT_UYVY;
-	case V4L2_PIX_FMT_YUYV:
-		return DRM_FORMAT_YUYV;
-	case V4L2_PIX_FMT_YUV420:
-		return DRM_FORMAT_YUV420;
-	case V4L2_PIX_FMT_YVU420:
-		return DRM_FORMAT_YVU420;
+		return (32 * pixel_stride) >> 3;
+	default:
+		break;
 	}
 
 	return -EINVAL;
 }
+EXPORT_SYMBOL_GPL(ipu_stride_to_bytes);
 
-enum ipu_color_space ipu_drm_fourcc_to_colorspace(u32 drm_fourcc)
+int ipu_degrees_to_rot_mode(enum ipu_rotate_mode *mode, int degrees,
+			    bool hflip, bool vflip)
 {
-	switch (drm_fourcc) {
-	case DRM_FORMAT_RGB565:
-	case DRM_FORMAT_BGR565:
-	case DRM_FORMAT_RGB888:
-	case DRM_FORMAT_BGR888:
-	case DRM_FORMAT_XRGB8888:
-	case DRM_FORMAT_XBGR8888:
-	case DRM_FORMAT_RGBX8888:
-	case DRM_FORMAT_BGRX8888:
-	case DRM_FORMAT_ARGB8888:
-	case DRM_FORMAT_ABGR8888:
-	case DRM_FORMAT_RGBA8888:
-	case DRM_FORMAT_BGRA8888:
-		return IPUV3_COLORSPACE_RGB;
-	case DRM_FORMAT_YUYV:
-	case DRM_FORMAT_UYVY:
-	case DRM_FORMAT_YUV420:
-	case DRM_FORMAT_YVU420:
-		return IPUV3_COLORSPACE_YUV;
+	u32 r90, vf, hf;
+
+	switch (degrees) {
+	case 0:
+		vf = hf = r90 = 0;
+		break;
+	case 90:
+		vf = hf = 0;
+		r90 = 1;
+		break;
+	case 180:
+		vf = hf = 1;
+		r90 = 0;
+		break;
+	case 270:
+		vf = hf = r90 = 1;
+		break;
 	default:
-		return IPUV3_COLORSPACE_UNKNOWN;
+		return -EINVAL;
 	}
-}
-EXPORT_SYMBOL_GPL(ipu_drm_fourcc_to_colorspace);
 
-int ipu_cpmem_set_image(struct ipu_ch_param __iomem *cpmem,
-		struct ipu_image *image)
-{
-	struct v4l2_pix_format *pix = &image->pix;
-	int y_offset, u_offset, v_offset;
+	hf ^= (u32)hflip;
+	vf ^= (u32)vflip;
 
-	pr_debug("%s: resolution: %dx%d stride: %d\n",
-			__func__, pix->width, pix->height,
-			pix->bytesperline);
+	*mode = (enum ipu_rotate_mode)((r90 << 2) | (hf << 1) | vf);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_degrees_to_rot_mode);
 
-	ipu_cpmem_set_resolution(cpmem, image->rect.width,
-			image->rect.height);
-	ipu_cpmem_set_stride(cpmem, pix->bytesperline);
+int ipu_rot_mode_to_degrees(int *degrees, enum ipu_rotate_mode mode,
+			    bool hflip, bool vflip)
+{
+	u32 r90, vf, hf;
 
-	ipu_cpmem_set_fmt(cpmem, v4l2_pix_fmt_to_drm_fourcc(pix->pixelformat));
+	r90 = ((u32)mode >> 2) & 0x1;
+	hf = ((u32)mode >> 1) & 0x1;
+	vf = ((u32)mode >> 0) & 0x1;
+	hf ^= (u32)hflip;
+	vf ^= (u32)vflip;
 
-	switch (pix->pixelformat) {
-	case V4L2_PIX_FMT_YUV420:
-	case V4L2_PIX_FMT_YVU420:
-		y_offset = Y_OFFSET(pix, image->rect.left, image->rect.top);
-		u_offset = U_OFFSET(pix, image->rect.left,
-				image->rect.top) - y_offset;
-		v_offset = V_OFFSET(pix, image->rect.left,
-				image->rect.top) - y_offset;
-
-		ipu_cpmem_set_yuv_planar_full(cpmem, pix->pixelformat,
-				pix->bytesperline, u_offset, v_offset);
-		ipu_cpmem_set_buffer(cpmem, 0, image->phys + y_offset);
+	switch ((enum ipu_rotate_mode)((r90 << 2) | (hf << 1) | vf)) {
+	case IPU_ROTATE_NONE:
+		*degrees = 0;
 		break;
-	case V4L2_PIX_FMT_UYVY:
-	case V4L2_PIX_FMT_YUYV:
-		ipu_cpmem_set_buffer(cpmem, 0, image->phys +
-				image->rect.left * 2 +
-				image->rect.top * image->pix.bytesperline);
+	case IPU_ROTATE_90_RIGHT:
+		*degrees = 90;
 		break;
-	case V4L2_PIX_FMT_RGB32:
-	case V4L2_PIX_FMT_BGR32:
-		ipu_cpmem_set_buffer(cpmem, 0, image->phys +
-				image->rect.left * 4 +
-				image->rect.top * image->pix.bytesperline);
+	case IPU_ROTATE_180:
+		*degrees = 180;
 		break;
-	case V4L2_PIX_FMT_RGB565:
-		ipu_cpmem_set_buffer(cpmem, 0, image->phys +
-				image->rect.left * 2 +
-				image->rect.top * image->pix.bytesperline);
-		break;
-	case V4L2_PIX_FMT_RGB24:
-	case V4L2_PIX_FMT_BGR24:
-		ipu_cpmem_set_buffer(cpmem, 0, image->phys +
-				image->rect.left * 3 +
-				image->rect.top * image->pix.bytesperline);
+	case IPU_ROTATE_90_LEFT:
+		*degrees = 270;
 		break;
 	default:
 		return -EINVAL;
@@ -523,27 +239,7 @@ int ipu_cpmem_set_image(struct ipu_ch_param __iomem *cpmem,
 
 	return 0;
 }
-EXPORT_SYMBOL_GPL(ipu_cpmem_set_image);
-
-enum ipu_color_space ipu_pixelformat_to_colorspace(u32 pixelformat)
-{
-	switch (pixelformat) {
-	case V4L2_PIX_FMT_YUV420:
-	case V4L2_PIX_FMT_YVU420:
-	case V4L2_PIX_FMT_UYVY:
-	case V4L2_PIX_FMT_YUYV:
-		return IPUV3_COLORSPACE_YUV;
-	case V4L2_PIX_FMT_RGB32:
-	case V4L2_PIX_FMT_BGR32:
-	case V4L2_PIX_FMT_RGB24:
-	case V4L2_PIX_FMT_BGR24:
-	case V4L2_PIX_FMT_RGB565:
-		return IPUV3_COLORSPACE_RGB;
-	default:
-		return IPUV3_COLORSPACE_UNKNOWN;
-	}
-}
-EXPORT_SYMBOL_GPL(ipu_pixelformat_to_colorspace);
+EXPORT_SYMBOL_GPL(ipu_rot_mode_to_degrees);
 
 struct ipuv3_channel *ipu_idmac_get(struct ipu_soc *ipu, unsigned num)
 {
@@ -587,7 +283,26 @@ void ipu_idmac_put(struct ipuv3_channel *channel)
 }
 EXPORT_SYMBOL_GPL(ipu_idmac_put);
 
-#define idma_mask(ch)			(1 << (ch & 0x1f))
+#define idma_mask(ch)			(1 << ((ch) & 0x1f))
+
+/*
+ * This is an undocumented feature, a write one to a channel bit in
+ * IPU_CHA_CUR_BUF and IPU_CHA_TRIPLE_CUR_BUF will reset the channel's
+ * internal current buffer pointer so that transfers start from buffer
+ * 0 on the next channel enable (that's the theory anyway, the imx6 TRM
+ * only says these are read-only registers). This operation is required
+ * for channel linking to work correctly, for instance video capture
+ * pipelines that carry out image rotations will fail after the first
+ * streaming unless this function is called for each channel before
+ * re-enabling the channels.
+ */
+static void __ipu_idmac_reset_current_buffer(struct ipuv3_channel *channel)
+{
+	struct ipu_soc *ipu = channel->ipu;
+	unsigned int chno = channel->num;
+
+	ipu_cm_write(ipu, idma_mask(chno), IPU_CHA_CUR_BUF(chno));
+}
 
 void ipu_idmac_set_double_buffer(struct ipuv3_channel *channel,
 		bool doublebuffer)
@@ -605,10 +320,81 @@ void ipu_idmac_set_double_buffer(struct ipuv3_channel *channel,
 		reg &= ~idma_mask(channel->num);
 	ipu_cm_write(ipu, reg, IPU_CHA_DB_MODE_SEL(channel->num));
 
+	__ipu_idmac_reset_current_buffer(channel);
+
 	spin_unlock_irqrestore(&ipu->lock, flags);
 }
 EXPORT_SYMBOL_GPL(ipu_idmac_set_double_buffer);
 
+static const struct {
+	int chnum;
+	u32 reg;
+	int shift;
+} idmac_lock_en_info[] = {
+	{ .chnum =  5, .reg = IDMAC_CH_LOCK_EN_1, .shift =  0, },
+	{ .chnum = 11, .reg = IDMAC_CH_LOCK_EN_1, .shift =  2, },
+	{ .chnum = 12, .reg = IDMAC_CH_LOCK_EN_1, .shift =  4, },
+	{ .chnum = 14, .reg = IDMAC_CH_LOCK_EN_1, .shift =  6, },
+	{ .chnum = 15, .reg = IDMAC_CH_LOCK_EN_1, .shift =  8, },
+	{ .chnum = 20, .reg = IDMAC_CH_LOCK_EN_1, .shift = 10, },
+	{ .chnum = 21, .reg = IDMAC_CH_LOCK_EN_1, .shift = 12, },
+	{ .chnum = 22, .reg = IDMAC_CH_LOCK_EN_1, .shift = 14, },
+	{ .chnum = 23, .reg = IDMAC_CH_LOCK_EN_1, .shift = 16, },
+	{ .chnum = 27, .reg = IDMAC_CH_LOCK_EN_1, .shift = 18, },
+	{ .chnum = 28, .reg = IDMAC_CH_LOCK_EN_1, .shift = 20, },
+	{ .chnum = 45, .reg = IDMAC_CH_LOCK_EN_2, .shift =  0, },
+	{ .chnum = 46, .reg = IDMAC_CH_LOCK_EN_2, .shift =  2, },
+	{ .chnum = 47, .reg = IDMAC_CH_LOCK_EN_2, .shift =  4, },
+	{ .chnum = 48, .reg = IDMAC_CH_LOCK_EN_2, .shift =  6, },
+	{ .chnum = 49, .reg = IDMAC_CH_LOCK_EN_2, .shift =  8, },
+	{ .chnum = 50, .reg = IDMAC_CH_LOCK_EN_2, .shift = 10, },
+};
+
+int ipu_idmac_lock_enable(struct ipuv3_channel *channel, int num_bursts)
+{
+	struct ipu_soc *ipu = channel->ipu;
+	unsigned long flags;
+	u32 bursts, regval;
+	int i;
+
+	switch (num_bursts) {
+	case 0:
+	case 1:
+		bursts = 0x00; /* locking disabled */
+		break;
+	case 2:
+		bursts = 0x01;
+		break;
+	case 4:
+		bursts = 0x02;
+		break;
+	case 8:
+		bursts = 0x03;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	for (i = 0; i < ARRAY_SIZE(idmac_lock_en_info); i++) {
+		if (channel->num == idmac_lock_en_info[i].chnum)
+			break;
+	}
+	if (i >= ARRAY_SIZE(idmac_lock_en_info))
+		return -EINVAL;
+
+	spin_lock_irqsave(&ipu->lock, flags);
+
+	regval = ipu_idmac_read(ipu, idmac_lock_en_info[i].reg);
+	regval &= ~(0x03 << idmac_lock_en_info[i].shift);
+	regval |= (bursts << idmac_lock_en_info[i].shift);
+	ipu_idmac_write(ipu, regval, idmac_lock_en_info[i].reg);
+
+	spin_unlock_irqrestore(&ipu->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_idmac_lock_enable);
+
 int ipu_module_enable(struct ipu_soc *ipu, u32 mask)
 {
 	unsigned long lock_flags;
@@ -661,30 +447,6 @@ int ipu_module_disable(struct ipu_soc *ipu, u32 mask)
 }
 EXPORT_SYMBOL_GPL(ipu_module_disable);
 
-int ipu_csi_enable(struct ipu_soc *ipu, int csi)
-{
-	return ipu_module_enable(ipu, csi ? IPU_CONF_CSI1_EN : IPU_CONF_CSI0_EN);
-}
-EXPORT_SYMBOL_GPL(ipu_csi_enable);
-
-int ipu_csi_disable(struct ipu_soc *ipu, int csi)
-{
-	return ipu_module_disable(ipu, csi ? IPU_CONF_CSI1_EN : IPU_CONF_CSI0_EN);
-}
-EXPORT_SYMBOL_GPL(ipu_csi_disable);
-
-int ipu_smfc_enable(struct ipu_soc *ipu)
-{
-	return ipu_module_enable(ipu, IPU_CONF_SMFC_EN);
-}
-EXPORT_SYMBOL_GPL(ipu_smfc_enable);
-
-int ipu_smfc_disable(struct ipu_soc *ipu)
-{
-	return ipu_module_disable(ipu, IPU_CONF_SMFC_EN);
-}
-EXPORT_SYMBOL_GPL(ipu_smfc_disable);
-
 int ipu_idmac_get_current_buffer(struct ipuv3_channel *channel)
 {
 	struct ipu_soc *ipu = channel->ipu;
@@ -694,6 +456,30 @@ int ipu_idmac_get_current_buffer(struct ipuv3_channel *channel)
 }
 EXPORT_SYMBOL_GPL(ipu_idmac_get_current_buffer);
 
+bool ipu_idmac_buffer_is_ready(struct ipuv3_channel *channel, u32 buf_num)
+{
+	struct ipu_soc *ipu = channel->ipu;
+	unsigned long flags;
+	u32 reg = 0;
+
+	spin_lock_irqsave(&ipu->lock, flags);
+	switch (buf_num) {
+	case 0:
+		reg = ipu_cm_read(ipu, IPU_CHA_BUF0_RDY(channel->num));
+		break;
+	case 1:
+		reg = ipu_cm_read(ipu, IPU_CHA_BUF1_RDY(channel->num));
+		break;
+	case 2:
+		reg = ipu_cm_read(ipu, IPU_CHA_BUF2_RDY(channel->num));
+		break;
+	}
+	spin_unlock_irqrestore(&ipu->lock, flags);
+
+	return ((reg & idma_mask(channel->num)) != 0);
+}
+EXPORT_SYMBOL_GPL(ipu_idmac_buffer_is_ready);
+
 void ipu_idmac_select_buffer(struct ipuv3_channel *channel, u32 buf_num)
 {
 	struct ipu_soc *ipu = channel->ipu;
@@ -712,6 +498,34 @@ void ipu_idmac_select_buffer(struct ipuv3_channel *channel, u32 buf_num)
 }
 EXPORT_SYMBOL_GPL(ipu_idmac_select_buffer);
 
+void ipu_idmac_clear_buffer(struct ipuv3_channel *channel, u32 buf_num)
+{
+	struct ipu_soc *ipu = channel->ipu;
+	unsigned int chno = channel->num;
+	unsigned long flags;
+
+	spin_lock_irqsave(&ipu->lock, flags);
+
+	ipu_cm_write(ipu, 0xF0300000, IPU_GPR); /* write one to clear */
+	switch (buf_num) {
+	case 0:
+		ipu_cm_write(ipu, idma_mask(chno), IPU_CHA_BUF0_RDY(chno));
+		break;
+	case 1:
+		ipu_cm_write(ipu, idma_mask(chno), IPU_CHA_BUF1_RDY(chno));
+		break;
+	case 2:
+		ipu_cm_write(ipu, idma_mask(chno), IPU_CHA_BUF2_RDY(chno));
+		break;
+	default:
+		break;
+	}
+	ipu_cm_write(ipu, 0x0, IPU_GPR); /* write one to set */
+
+	spin_unlock_irqrestore(&ipu->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_idmac_clear_buffer);
+
 int ipu_idmac_enable_channel(struct ipuv3_channel *channel)
 {
 	struct ipu_soc *ipu = channel->ipu;
@@ -782,6 +596,8 @@ int ipu_idmac_disable_channel(struct ipuv3_channel *channel)
 	val &= ~idma_mask(channel->num);
 	ipu_idmac_write(ipu, val, IDMAC_CHA_EN(channel->num));
 
+	__ipu_idmac_reset_current_buffer(channel);
+
 	/* Set channel buffers NOT to be ready */
 	ipu_cm_write(ipu, 0xf0000000, IPU_GPR); /* write one to clear */
 
@@ -810,6 +626,31 @@ int ipu_idmac_disable_channel(struct ipuv3_channel *channel)
 }
 EXPORT_SYMBOL_GPL(ipu_idmac_disable_channel);
 
+/*
+ * The imx6 rev. D TRM says that enabling the WM feature will increase
+ * a channel's priority. Refer to Table 36-8 Calculated priority value.
+ * The sub-module that is the sink or source for the channel must enable
+ * watermark signal for this to take effect (SMFC_WM for instance).
+ */
+void ipu_idmac_enable_watermark(struct ipuv3_channel *channel, bool enable)
+{
+	struct ipu_soc *ipu = channel->ipu;
+	unsigned long flags;
+	u32 val;
+
+	spin_lock_irqsave(&ipu->lock, flags);
+
+	val = ipu_idmac_read(ipu, IDMAC_WM_EN(channel->num));
+	if (enable)
+		val |= 1 << (channel->num % 32);
+	else
+		val &= ~(1 << (channel->num % 32));
+	ipu_idmac_write(ipu, val, IDMAC_WM_EN(channel->num));
+
+	spin_unlock_irqrestore(&ipu->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_idmac_enable_watermark);
+
 static int ipu_memory_reset(struct ipu_soc *ipu)
 {
 	unsigned long timeout;
@@ -826,12 +667,66 @@ static int ipu_memory_reset(struct ipu_soc *ipu)
 	return 0;
 }
 
+/*
+ * Set the source mux for the given CSI. Selects either parallel or
+ * MIPI CSI2 sources.
+ */
+void ipu_set_csi_src_mux(struct ipu_soc *ipu, int csi_id, bool mipi_csi2)
+{
+	unsigned long flags;
+	u32 val, mask;
+
+	mask = (csi_id == 1) ? IPU_CONF_CSI1_DATA_SOURCE :
+		IPU_CONF_CSI0_DATA_SOURCE;
+
+	spin_lock_irqsave(&ipu->lock, flags);
+
+	val = ipu_cm_read(ipu, IPU_CONF);
+	if (mipi_csi2)
+		val |= mask;
+	else
+		val &= ~mask;
+	ipu_cm_write(ipu, val, IPU_CONF);
+
+	spin_unlock_irqrestore(&ipu->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_set_csi_src_mux);
+
+/*
+ * Set the source mux for the IC. Selects either CSI[01] or the VDI.
+ */
+void ipu_set_ic_src_mux(struct ipu_soc *ipu, int csi_id, bool vdi)
+{
+	unsigned long flags;
+	u32 val;
+
+	spin_lock_irqsave(&ipu->lock, flags);
+
+	val = ipu_cm_read(ipu, IPU_CONF);
+	if (vdi) {
+		val |= IPU_CONF_IC_INPUT;
+	} else {
+		val &= ~IPU_CONF_IC_INPUT;
+		if (csi_id == 1)
+			val |= IPU_CONF_CSI_SEL;
+		else
+			val &= ~IPU_CONF_CSI_SEL;
+	}
+	ipu_cm_write(ipu, val, IPU_CONF);
+
+	spin_unlock_irqrestore(&ipu->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_set_ic_src_mux);
+
 struct ipu_devtype {
 	const char *name;
 	unsigned long cm_ofs;
 	unsigned long cpmem_ofs;
 	unsigned long srm_ofs;
 	unsigned long tpm_ofs;
+	unsigned long csi0_ofs;
+	unsigned long csi1_ofs;
+	unsigned long ic_ofs;
 	unsigned long disp0_ofs;
 	unsigned long disp1_ofs;
 	unsigned long dc_tmpl_ofs;
@@ -845,6 +740,9 @@ static struct ipu_devtype ipu_type_imx51 = {
 	.cpmem_ofs = 0x1f000000,
 	.srm_ofs = 0x1f040000,
 	.tpm_ofs = 0x1f060000,
+	.csi0_ofs = 0x1f030000,
+	.csi1_ofs = 0x1f038000,
+	.ic_ofs = 0x1f020000,
 	.disp0_ofs = 0x1e040000,
 	.disp1_ofs = 0x1e048000,
 	.dc_tmpl_ofs = 0x1f080000,
@@ -858,6 +756,9 @@ static struct ipu_devtype ipu_type_imx53 = {
 	.cpmem_ofs = 0x07000000,
 	.srm_ofs = 0x07040000,
 	.tpm_ofs = 0x07060000,
+	.csi0_ofs = 0x07030000,
+	.csi1_ofs = 0x07038000,
+	.ic_ofs = 0x07020000,
 	.disp0_ofs = 0x06040000,
 	.disp1_ofs = 0x06048000,
 	.dc_tmpl_ofs = 0x07080000,
@@ -871,6 +772,9 @@ static struct ipu_devtype ipu_type_imx6q = {
 	.cpmem_ofs = 0x00300000,
 	.srm_ofs = 0x00340000,
 	.tpm_ofs = 0x00360000,
+	.csi0_ofs = 0x00230000,
+	.csi1_ofs = 0x00238000,
+	.ic_ofs = 0x00220000,
 	.disp0_ofs = 0x00240000,
 	.disp1_ofs = 0x00248000,
 	.dc_tmpl_ofs = 0x00380000,
@@ -895,8 +799,36 @@ static int ipu_submodules_init(struct ipu_soc *ipu,
 	struct device *dev = &pdev->dev;
 	const struct ipu_devtype *devtype = ipu->devtype;
 
+	ret = ipu_cpmem_init(ipu, dev, ipu_base + devtype->cpmem_ofs);
+	if (ret) {
+		unit = "cpmem";
+		goto err_cpmem;
+	}
+
+	ret = ipu_csi_init(ipu, dev, 0, ipu_base + devtype->csi0_ofs,
+			   IPU_CONF_CSI0_EN, ipu_clk);
+	if (ret) {
+		unit = "csi0";
+		goto err_csi_0;
+	}
+
+	ret = ipu_csi_init(ipu, dev, 1, ipu_base + devtype->csi1_ofs,
+			   IPU_CONF_CSI1_EN, ipu_clk);
+	if (ret) {
+		unit = "csi1";
+		goto err_csi_1;
+	}
+
+	ret = ipu_ic_init(ipu, dev,
+			  ipu_base + devtype->ic_ofs,
+			  ipu_base + devtype->tpm_ofs);
+	if (ret) {
+		unit = "ic";
+		goto err_ic;
+	}
+
 	ret = ipu_di_init(ipu, dev, 0, ipu_base + devtype->disp0_ofs,
-			IPU_CONF_DI0_EN, ipu_clk);
+			  IPU_CONF_DI0_EN, ipu_clk);
 	if (ret) {
 		unit = "di0";
 		goto err_di_0;
@@ -949,6 +881,14 @@ err_dc:
 err_di_1:
 	ipu_di_exit(ipu, 0);
 err_di_0:
+	ipu_ic_exit(ipu);
+err_ic:
+	ipu_csi_exit(ipu, 1);
+err_csi_1:
+	ipu_csi_exit(ipu, 0);
+err_csi_0:
+	ipu_cpmem_exit(ipu);
+err_cpmem:
 	dev_err(&pdev->dev, "init %s failed with %d\n", unit, ret);
 	return ret;
 }
@@ -1025,6 +965,10 @@ static void ipu_submodules_exit(struct ipu_soc *ipu)
 	ipu_dc_exit(ipu);
 	ipu_di_exit(ipu, 1);
 	ipu_di_exit(ipu, 0);
+	ipu_ic_exit(ipu);
+	ipu_csi_exit(ipu, 1);
+	ipu_csi_exit(ipu, 0);
+	ipu_cpmem_exit(ipu);
 }
 
 static int platform_remove_devices_fn(struct device *dev, void *unused)
@@ -1116,8 +1060,10 @@ static int ipu_add_client_devices(struct ipu_soc *ipu, unsigned long ipu_base)
 				id++, &reg->pdata, sizeof(reg->pdata));
 		}
 
-		if (IS_ERR(pdev))
+		if (IS_ERR(pdev)) {
+			ret = PTR_ERR(pdev);
 			goto err_register;
+		}
 	}
 
 	return 0;
@@ -1201,6 +1147,44 @@ static void ipu_irq_exit(struct ipu_soc *ipu)
 	irq_domain_remove(ipu->domain);
 }
 
+void ipu_dump(struct ipu_soc *ipu)
+{
+	int i;
+
+	dev_dbg(ipu->dev, "IPU_CONF = \t0x%08X\n",
+		ipu_cm_read(ipu, IPU_CONF));
+	dev_dbg(ipu->dev, "IDMAC_CONF = \t0x%08X\n",
+		ipu_idmac_read(ipu, IDMAC_CONF));
+	dev_dbg(ipu->dev, "IDMAC_CHA_EN1 = \t0x%08X\n",
+		ipu_idmac_read(ipu, IDMAC_CHA_EN(0)));
+	dev_dbg(ipu->dev, "IDMAC_CHA_EN2 = \t0x%08X\n",
+		ipu_idmac_read(ipu, IDMAC_CHA_EN(32)));
+	dev_dbg(ipu->dev, "IDMAC_CHA_PRI1 = \t0x%08X\n",
+		ipu_idmac_read(ipu, IDMAC_CHA_PRI(0)));
+	dev_dbg(ipu->dev, "IDMAC_CHA_PRI2 = \t0x%08X\n",
+		ipu_idmac_read(ipu, IDMAC_CHA_PRI(32)));
+	dev_dbg(ipu->dev, "IDMAC_BAND_EN1 = \t0x%08X\n",
+		ipu_idmac_read(ipu, IDMAC_BAND_EN(0)));
+	dev_dbg(ipu->dev, "IDMAC_BAND_EN2 = \t0x%08X\n",
+		ipu_idmac_read(ipu, IDMAC_BAND_EN(32)));
+	dev_dbg(ipu->dev, "IPU_CHA_DB_MODE_SEL0 = \t0x%08X\n",
+		ipu_cm_read(ipu, IPU_CHA_DB_MODE_SEL(0)));
+	dev_dbg(ipu->dev, "IPU_CHA_DB_MODE_SEL1 = \t0x%08X\n",
+		ipu_cm_read(ipu, IPU_CHA_DB_MODE_SEL(32)));
+	dev_dbg(ipu->dev, "IPU_FS_PROC_FLOW1 = \t0x%08X\n",
+		ipu_cm_read(ipu, IPU_FS_PROC_FLOW1));
+	dev_dbg(ipu->dev, "IPU_FS_PROC_FLOW2 = \t0x%08X\n",
+		ipu_cm_read(ipu, IPU_FS_PROC_FLOW2));
+	dev_dbg(ipu->dev, "IPU_FS_PROC_FLOW3 = \t0x%08X\n",
+		ipu_cm_read(ipu, IPU_FS_PROC_FLOW3));
+	dev_dbg(ipu->dev, "IPU_FS_DISP_FLOW1 = \t0x%08X\n",
+		ipu_cm_read(ipu, IPU_FS_DISP_FLOW1));
+	for (i = 0; i < 15; i++)
+		dev_dbg(ipu->dev, "IPU_INT_CTRL(%d) = \t%08X\n", i,
+			ipu_cm_read(ipu, IPU_INT_CTRL(i)));
+}
+EXPORT_SYMBOL_GPL(ipu_dump);
+
 static int ipu_probe(struct platform_device *pdev)
 {
 	const struct of_device_id *of_id =
@@ -1243,6 +1227,12 @@ static int ipu_probe(struct platform_device *pdev)
 			ipu_base + devtype->cm_ofs + IPU_CM_IDMAC_REG_OFS);
 	dev_dbg(&pdev->dev, "cpmem:    0x%08lx\n",
 			ipu_base + devtype->cpmem_ofs);
+	dev_dbg(&pdev->dev, "csi0:    0x%08lx\n",
+			ipu_base + devtype->csi0_ofs);
+	dev_dbg(&pdev->dev, "csi1:    0x%08lx\n",
+			ipu_base + devtype->csi1_ofs);
+	dev_dbg(&pdev->dev, "ic:      0x%08lx\n",
+			ipu_base + devtype->ic_ofs);
 	dev_dbg(&pdev->dev, "disp0:    0x%08lx\n",
 			ipu_base + devtype->disp0_ofs);
 	dev_dbg(&pdev->dev, "disp1:    0x%08lx\n",
@@ -1265,10 +1255,8 @@ static int ipu_probe(struct platform_device *pdev)
 	ipu->idmac_reg = devm_ioremap(&pdev->dev,
 			ipu_base + devtype->cm_ofs + IPU_CM_IDMAC_REG_OFS,
 			PAGE_SIZE);
-	ipu->cpmem_base = devm_ioremap(&pdev->dev,
-			ipu_base + devtype->cpmem_ofs, PAGE_SIZE);
 
-	if (!ipu->cm_reg || !ipu->idmac_reg || !ipu->cpmem_base)
+	if (!ipu->cm_reg || !ipu->idmac_reg)
 		return -ENOMEM;
 
 	ipu->clk = devm_clk_get(&pdev->dev, "bus");
diff --git a/drivers/gpu/ipu-v3/ipu-cpmem.c b/drivers/gpu/ipu-v3/ipu-cpmem.c
new file mode 100644
index 000000000000..3bf05bc4ab67
--- /dev/null
+++ b/drivers/gpu/ipu-v3/ipu-cpmem.c
@@ -0,0 +1,764 @@
+/*
+ * Copyright (C) 2012 Mentor Graphics Inc.
+ * Copyright 2005-2012 Freescale Semiconductor, Inc. All Rights Reserved.
+ *
+ * The code contained herein is licensed under the GNU General Public
+ * License. You may obtain a copy of the GNU General Public License
+ * Version 2 or later at the following locations:
+ *
+ * http://www.opensource.org/licenses/gpl-license.html
+ * http://www.gnu.org/copyleft/gpl.html
+ */
+#include <linux/types.h>
+#include <linux/bitrev.h>
+#include <linux/io.h>
+#include <drm/drm_fourcc.h>
+#include "ipu-prv.h"
+
+struct ipu_cpmem_word {
+	u32 data[5];
+	u32 res[3];
+};
+
+struct ipu_ch_param {
+	struct ipu_cpmem_word word[2];
+};
+
+struct ipu_cpmem {
+	struct ipu_ch_param __iomem *base;
+	u32 module;
+	spinlock_t lock;
+	int use_count;
+	struct ipu_soc *ipu;
+};
+
+#define IPU_CPMEM_WORD(word, ofs, size) ((((word) * 160 + (ofs)) << 8) | (size))
+
+#define IPU_FIELD_UBO		IPU_CPMEM_WORD(0, 46, 22)
+#define IPU_FIELD_VBO		IPU_CPMEM_WORD(0, 68, 22)
+#define IPU_FIELD_IOX		IPU_CPMEM_WORD(0, 90, 4)
+#define IPU_FIELD_RDRW		IPU_CPMEM_WORD(0, 94, 1)
+#define IPU_FIELD_SO		IPU_CPMEM_WORD(0, 113, 1)
+#define IPU_FIELD_SLY		IPU_CPMEM_WORD(1, 102, 14)
+#define IPU_FIELD_SLUV		IPU_CPMEM_WORD(1, 128, 14)
+
+#define IPU_FIELD_XV		IPU_CPMEM_WORD(0, 0, 10)
+#define IPU_FIELD_YV		IPU_CPMEM_WORD(0, 10, 9)
+#define IPU_FIELD_XB		IPU_CPMEM_WORD(0, 19, 13)
+#define IPU_FIELD_YB		IPU_CPMEM_WORD(0, 32, 12)
+#define IPU_FIELD_NSB_B		IPU_CPMEM_WORD(0, 44, 1)
+#define IPU_FIELD_CF		IPU_CPMEM_WORD(0, 45, 1)
+#define IPU_FIELD_SX		IPU_CPMEM_WORD(0, 46, 12)
+#define IPU_FIELD_SY		IPU_CPMEM_WORD(0, 58, 11)
+#define IPU_FIELD_NS		IPU_CPMEM_WORD(0, 69, 10)
+#define IPU_FIELD_SDX		IPU_CPMEM_WORD(0, 79, 7)
+#define IPU_FIELD_SM		IPU_CPMEM_WORD(0, 86, 10)
+#define IPU_FIELD_SCC		IPU_CPMEM_WORD(0, 96, 1)
+#define IPU_FIELD_SCE		IPU_CPMEM_WORD(0, 97, 1)
+#define IPU_FIELD_SDY		IPU_CPMEM_WORD(0, 98, 7)
+#define IPU_FIELD_SDRX		IPU_CPMEM_WORD(0, 105, 1)
+#define IPU_FIELD_SDRY		IPU_CPMEM_WORD(0, 106, 1)
+#define IPU_FIELD_BPP		IPU_CPMEM_WORD(0, 107, 3)
+#define IPU_FIELD_DEC_SEL	IPU_CPMEM_WORD(0, 110, 2)
+#define IPU_FIELD_DIM		IPU_CPMEM_WORD(0, 112, 1)
+#define IPU_FIELD_BNDM		IPU_CPMEM_WORD(0, 114, 3)
+#define IPU_FIELD_BM		IPU_CPMEM_WORD(0, 117, 2)
+#define IPU_FIELD_ROT		IPU_CPMEM_WORD(0, 119, 1)
+#define IPU_FIELD_ROT_HF_VF	IPU_CPMEM_WORD(0, 119, 3)
+#define IPU_FIELD_HF		IPU_CPMEM_WORD(0, 120, 1)
+#define IPU_FIELD_VF		IPU_CPMEM_WORD(0, 121, 1)
+#define IPU_FIELD_THE		IPU_CPMEM_WORD(0, 122, 1)
+#define IPU_FIELD_CAP		IPU_CPMEM_WORD(0, 123, 1)
+#define IPU_FIELD_CAE		IPU_CPMEM_WORD(0, 124, 1)
+#define IPU_FIELD_FW		IPU_CPMEM_WORD(0, 125, 13)
+#define IPU_FIELD_FH		IPU_CPMEM_WORD(0, 138, 12)
+#define IPU_FIELD_EBA0		IPU_CPMEM_WORD(1, 0, 29)
+#define IPU_FIELD_EBA1		IPU_CPMEM_WORD(1, 29, 29)
+#define IPU_FIELD_ILO		IPU_CPMEM_WORD(1, 58, 20)
+#define IPU_FIELD_NPB		IPU_CPMEM_WORD(1, 78, 7)
+#define IPU_FIELD_PFS		IPU_CPMEM_WORD(1, 85, 4)
+#define IPU_FIELD_ALU		IPU_CPMEM_WORD(1, 89, 1)
+#define IPU_FIELD_ALBM		IPU_CPMEM_WORD(1, 90, 3)
+#define IPU_FIELD_ID		IPU_CPMEM_WORD(1, 93, 2)
+#define IPU_FIELD_TH		IPU_CPMEM_WORD(1, 95, 7)
+#define IPU_FIELD_SL		IPU_CPMEM_WORD(1, 102, 14)
+#define IPU_FIELD_WID0		IPU_CPMEM_WORD(1, 116, 3)
+#define IPU_FIELD_WID1		IPU_CPMEM_WORD(1, 119, 3)
+#define IPU_FIELD_WID2		IPU_CPMEM_WORD(1, 122, 3)
+#define IPU_FIELD_WID3		IPU_CPMEM_WORD(1, 125, 3)
+#define IPU_FIELD_OFS0		IPU_CPMEM_WORD(1, 128, 5)
+#define IPU_FIELD_OFS1		IPU_CPMEM_WORD(1, 133, 5)
+#define IPU_FIELD_OFS2		IPU_CPMEM_WORD(1, 138, 5)
+#define IPU_FIELD_OFS3		IPU_CPMEM_WORD(1, 143, 5)
+#define IPU_FIELD_SXYS		IPU_CPMEM_WORD(1, 148, 1)
+#define IPU_FIELD_CRE		IPU_CPMEM_WORD(1, 149, 1)
+#define IPU_FIELD_DEC_SEL2	IPU_CPMEM_WORD(1, 150, 1)
+
+static inline struct ipu_ch_param __iomem *
+ipu_get_cpmem(struct ipuv3_channel *ch)
+{
+	struct ipu_cpmem *cpmem = ch->ipu->cpmem_priv;
+
+	return cpmem->base + ch->num;
+}
+
+static void ipu_ch_param_write_field(struct ipuv3_channel *ch, u32 wbs, u32 v)
+{
+	struct ipu_ch_param __iomem *base = ipu_get_cpmem(ch);
+	u32 bit = (wbs >> 8) % 160;
+	u32 size = wbs & 0xff;
+	u32 word = (wbs >> 8) / 160;
+	u32 i = bit / 32;
+	u32 ofs = bit % 32;
+	u32 mask = (1 << size) - 1;
+	u32 val;
+
+	pr_debug("%s %d %d %d\n", __func__, word, bit , size);
+
+	val = readl(&base->word[word].data[i]);
+	val &= ~(mask << ofs);
+	val |= v << ofs;
+	writel(val, &base->word[word].data[i]);
+
+	if ((bit + size - 1) / 32 > i) {
+		val = readl(&base->word[word].data[i + 1]);
+		val &= ~(mask >> (ofs ? (32 - ofs) : 0));
+		val |= v >> (ofs ? (32 - ofs) : 0);
+		writel(val, &base->word[word].data[i + 1]);
+	}
+}
+
+static u32 ipu_ch_param_read_field(struct ipuv3_channel *ch, u32 wbs)
+{
+	struct ipu_ch_param __iomem *base = ipu_get_cpmem(ch);
+	u32 bit = (wbs >> 8) % 160;
+	u32 size = wbs & 0xff;
+	u32 word = (wbs >> 8) / 160;
+	u32 i = bit / 32;
+	u32 ofs = bit % 32;
+	u32 mask = (1 << size) - 1;
+	u32 val = 0;
+
+	pr_debug("%s %d %d %d\n", __func__, word, bit , size);
+
+	val = (readl(&base->word[word].data[i]) >> ofs) & mask;
+
+	if ((bit + size - 1) / 32 > i) {
+		u32 tmp;
+
+		tmp = readl(&base->word[word].data[i + 1]);
+		tmp &= mask >> (ofs ? (32 - ofs) : 0);
+		val |= tmp << (ofs ? (32 - ofs) : 0);
+	}
+
+	return val;
+}
+
+/*
+ * The V4L2 spec defines packed RGB formats in memory byte order, which from
+ * point of view of the IPU corresponds to little-endian words with the first
+ * component in the least significant bits.
+ * The DRM pixel formats and IPU internal representation are ordered the other
+ * way around, with the first named component ordered at the most significant
+ * bits. Further, V4L2 formats are not well defined:
+ *     http://linuxtv.org/downloads/v4l-dvb-apis/packed-rgb.html
+ * We choose the interpretation which matches GStreamer behavior.
+ */
+static int v4l2_pix_fmt_to_drm_fourcc(u32 pixelformat)
+{
+	switch (pixelformat) {
+	case V4L2_PIX_FMT_RGB565:
+		/*
+		 * Here we choose the 'corrected' interpretation of RGBP, a
+		 * little-endian 16-bit word with the red component at the most
+		 * significant bits:
+		 * g[2:0]b[4:0] r[4:0]g[5:3] <=> [16:0] R:G:B
+		 */
+		return DRM_FORMAT_RGB565;
+	case V4L2_PIX_FMT_BGR24:
+		/* B G R <=> [24:0] R:G:B */
+		return DRM_FORMAT_RGB888;
+	case V4L2_PIX_FMT_RGB24:
+		/* R G B <=> [24:0] B:G:R */
+		return DRM_FORMAT_BGR888;
+	case V4L2_PIX_FMT_BGR32:
+		/* B G R A <=> [32:0] A:B:G:R */
+		return DRM_FORMAT_XRGB8888;
+	case V4L2_PIX_FMT_RGB32:
+		/* R G B A <=> [32:0] A:B:G:R */
+		return DRM_FORMAT_XBGR8888;
+	case V4L2_PIX_FMT_UYVY:
+		return DRM_FORMAT_UYVY;
+	case V4L2_PIX_FMT_YUYV:
+		return DRM_FORMAT_YUYV;
+	case V4L2_PIX_FMT_YUV420:
+		return DRM_FORMAT_YUV420;
+	case V4L2_PIX_FMT_YUV422P:
+		return DRM_FORMAT_YUV422;
+	case V4L2_PIX_FMT_YVU420:
+		return DRM_FORMAT_YVU420;
+	case V4L2_PIX_FMT_NV12:
+		return DRM_FORMAT_NV12;
+	case V4L2_PIX_FMT_NV16:
+		return DRM_FORMAT_NV16;
+	}
+
+	return -EINVAL;
+}
+
+void ipu_cpmem_zero(struct ipuv3_channel *ch)
+{
+	struct ipu_ch_param __iomem *p = ipu_get_cpmem(ch);
+	void __iomem *base = p;
+	int i;
+
+	for (i = 0; i < sizeof(*p) / sizeof(u32); i++)
+		writel(0, base + i * sizeof(u32));
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_zero);
+
+void ipu_cpmem_set_resolution(struct ipuv3_channel *ch, int xres, int yres)
+{
+	ipu_ch_param_write_field(ch, IPU_FIELD_FW, xres - 1);
+	ipu_ch_param_write_field(ch, IPU_FIELD_FH, yres - 1);
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_resolution);
+
+void ipu_cpmem_set_stride(struct ipuv3_channel *ch, int stride)
+{
+	ipu_ch_param_write_field(ch, IPU_FIELD_SLY, stride - 1);
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_stride);
+
+void ipu_cpmem_set_high_priority(struct ipuv3_channel *ch)
+{
+	struct ipu_soc *ipu = ch->ipu;
+	u32 val;
+
+	if (ipu->ipu_type == IPUV3EX)
+		ipu_ch_param_write_field(ch, IPU_FIELD_ID, 1);
+
+	val = ipu_idmac_read(ipu, IDMAC_CHA_PRI(ch->num));
+	val |= 1 << (ch->num % 32);
+	ipu_idmac_write(ipu, val, IDMAC_CHA_PRI(ch->num));
+};
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_high_priority);
+
+void ipu_cpmem_set_buffer(struct ipuv3_channel *ch, int bufnum, dma_addr_t buf)
+{
+	if (bufnum)
+		ipu_ch_param_write_field(ch, IPU_FIELD_EBA1, buf >> 3);
+	else
+		ipu_ch_param_write_field(ch, IPU_FIELD_EBA0, buf >> 3);
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_buffer);
+
+void ipu_cpmem_interlaced_scan(struct ipuv3_channel *ch, int stride)
+{
+	ipu_ch_param_write_field(ch, IPU_FIELD_SO, 1);
+	ipu_ch_param_write_field(ch, IPU_FIELD_ILO, stride / 8);
+	ipu_ch_param_write_field(ch, IPU_FIELD_SLY, (stride * 2) - 1);
+};
+EXPORT_SYMBOL_GPL(ipu_cpmem_interlaced_scan);
+
+void ipu_cpmem_set_axi_id(struct ipuv3_channel *ch, u32 id)
+{
+	id &= 0x3;
+	ipu_ch_param_write_field(ch, IPU_FIELD_ID, id);
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_axi_id);
+
+void ipu_cpmem_set_burstsize(struct ipuv3_channel *ch, int burstsize)
+{
+	ipu_ch_param_write_field(ch, IPU_FIELD_NPB, burstsize - 1);
+};
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_burstsize);
+
+void ipu_cpmem_set_block_mode(struct ipuv3_channel *ch)
+{
+	ipu_ch_param_write_field(ch, IPU_FIELD_BM, 1);
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_block_mode);
+
+void ipu_cpmem_set_rotation(struct ipuv3_channel *ch,
+			    enum ipu_rotate_mode rot)
+{
+	u32 temp_rot = bitrev8(rot) >> 5;
+
+	ipu_ch_param_write_field(ch, IPU_FIELD_ROT_HF_VF, temp_rot);
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_rotation);
+
+int ipu_cpmem_set_format_rgb(struct ipuv3_channel *ch,
+			     const struct ipu_rgb *rgb)
+{
+	int bpp = 0, npb = 0, ro, go, bo, to;
+
+	ro = rgb->bits_per_pixel - rgb->red.length - rgb->red.offset;
+	go = rgb->bits_per_pixel - rgb->green.length - rgb->green.offset;
+	bo = rgb->bits_per_pixel - rgb->blue.length - rgb->blue.offset;
+	to = rgb->bits_per_pixel - rgb->transp.length - rgb->transp.offset;
+
+	ipu_ch_param_write_field(ch, IPU_FIELD_WID0, rgb->red.length - 1);
+	ipu_ch_param_write_field(ch, IPU_FIELD_OFS0, ro);
+	ipu_ch_param_write_field(ch, IPU_FIELD_WID1, rgb->green.length - 1);
+	ipu_ch_param_write_field(ch, IPU_FIELD_OFS1, go);
+	ipu_ch_param_write_field(ch, IPU_FIELD_WID2, rgb->blue.length - 1);
+	ipu_ch_param_write_field(ch, IPU_FIELD_OFS2, bo);
+
+	if (rgb->transp.length) {
+		ipu_ch_param_write_field(ch, IPU_FIELD_WID3,
+				rgb->transp.length - 1);
+		ipu_ch_param_write_field(ch, IPU_FIELD_OFS3, to);
+	} else {
+		ipu_ch_param_write_field(ch, IPU_FIELD_WID3, 7);
+		ipu_ch_param_write_field(ch, IPU_FIELD_OFS3,
+				rgb->bits_per_pixel);
+	}
+
+	switch (rgb->bits_per_pixel) {
+	case 32:
+		bpp = 0;
+		npb = 15;
+		break;
+	case 24:
+		bpp = 1;
+		npb = 19;
+		break;
+	case 16:
+		bpp = 3;
+		npb = 31;
+		break;
+	case 8:
+		bpp = 5;
+		npb = 63;
+		break;
+	default:
+		return -EINVAL;
+	}
+	ipu_ch_param_write_field(ch, IPU_FIELD_BPP, bpp);
+	ipu_ch_param_write_field(ch, IPU_FIELD_NPB, npb);
+	ipu_ch_param_write_field(ch, IPU_FIELD_PFS, 7); /* rgb mode */
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_format_rgb);
+
+int ipu_cpmem_set_format_passthrough(struct ipuv3_channel *ch, int width)
+{
+	int bpp = 0, npb = 0;
+
+	switch (width) {
+	case 32:
+		bpp = 0;
+		npb = 15;
+		break;
+	case 24:
+		bpp = 1;
+		npb = 19;
+		break;
+	case 16:
+		bpp = 3;
+		npb = 31;
+		break;
+	case 8:
+		bpp = 5;
+		npb = 63;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	ipu_ch_param_write_field(ch, IPU_FIELD_BPP, bpp);
+	ipu_ch_param_write_field(ch, IPU_FIELD_NPB, npb);
+	ipu_ch_param_write_field(ch, IPU_FIELD_PFS, 6); /* raw mode */
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_format_passthrough);
+
+void ipu_cpmem_set_yuv_interleaved(struct ipuv3_channel *ch, u32 pixel_format)
+{
+	switch (pixel_format) {
+	case V4L2_PIX_FMT_UYVY:
+		ipu_ch_param_write_field(ch, IPU_FIELD_BPP, 3); /* bits/pixel */
+		ipu_ch_param_write_field(ch, IPU_FIELD_PFS, 0xA);/* pix fmt */
+		ipu_ch_param_write_field(ch, IPU_FIELD_NPB, 31);/* burst size */
+		break;
+	case V4L2_PIX_FMT_YUYV:
+		ipu_ch_param_write_field(ch, IPU_FIELD_BPP, 3); /* bits/pixel */
+		ipu_ch_param_write_field(ch, IPU_FIELD_PFS, 0x8);/* pix fmt */
+		ipu_ch_param_write_field(ch, IPU_FIELD_NPB, 31);/* burst size */
+		break;
+	}
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_yuv_interleaved);
+
+void ipu_cpmem_set_yuv_planar_full(struct ipuv3_channel *ch,
+				   u32 pixel_format, int stride,
+				   int u_offset, int v_offset)
+{
+	switch (pixel_format) {
+	case V4L2_PIX_FMT_YUV420:
+	case V4L2_PIX_FMT_YUV422P:
+		ipu_ch_param_write_field(ch, IPU_FIELD_SLUV, (stride / 2) - 1);
+		ipu_ch_param_write_field(ch, IPU_FIELD_UBO, u_offset / 8);
+		ipu_ch_param_write_field(ch, IPU_FIELD_VBO, v_offset / 8);
+		break;
+	case V4L2_PIX_FMT_YVU420:
+		ipu_ch_param_write_field(ch, IPU_FIELD_SLUV, (stride / 2) - 1);
+		ipu_ch_param_write_field(ch, IPU_FIELD_UBO, v_offset / 8);
+		ipu_ch_param_write_field(ch, IPU_FIELD_VBO, u_offset / 8);
+		break;
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_NV16:
+		ipu_ch_param_write_field(ch, IPU_FIELD_SLUV, stride - 1);
+		ipu_ch_param_write_field(ch, IPU_FIELD_UBO, u_offset / 8);
+		ipu_ch_param_write_field(ch, IPU_FIELD_VBO, u_offset / 8);
+		break;
+	}
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_yuv_planar_full);
+
+void ipu_cpmem_set_yuv_planar(struct ipuv3_channel *ch,
+			      u32 pixel_format, int stride, int height)
+{
+	int u_offset, v_offset;
+	int uv_stride = 0;
+
+	switch (pixel_format) {
+	case V4L2_PIX_FMT_YUV420:
+	case V4L2_PIX_FMT_YVU420:
+		uv_stride = stride / 2;
+		u_offset = stride * height;
+		v_offset = u_offset + (uv_stride * height / 2);
+		ipu_cpmem_set_yuv_planar_full(ch, pixel_format, stride,
+					      u_offset, v_offset);
+		break;
+	case V4L2_PIX_FMT_YUV422P:
+		uv_stride = stride / 2;
+		u_offset = stride * height;
+		v_offset = u_offset + (uv_stride * height);
+		ipu_cpmem_set_yuv_planar_full(ch, pixel_format, stride,
+					      u_offset, v_offset);
+		break;
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_NV16:
+		u_offset = stride * height;
+		ipu_cpmem_set_yuv_planar_full(ch, pixel_format, stride,
+					      u_offset, 0);
+		break;
+	}
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_yuv_planar);
+
+static const struct ipu_rgb def_rgb_32 = {
+	.red	= { .offset = 16, .length = 8, },
+	.green	= { .offset =  8, .length = 8, },
+	.blue	= { .offset =  0, .length = 8, },
+	.transp = { .offset = 24, .length = 8, },
+	.bits_per_pixel = 32,
+};
+
+static const struct ipu_rgb def_bgr_32 = {
+	.red	= { .offset =  0, .length = 8, },
+	.green	= { .offset =  8, .length = 8, },
+	.blue	= { .offset = 16, .length = 8, },
+	.transp = { .offset = 24, .length = 8, },
+	.bits_per_pixel = 32,
+};
+
+static const struct ipu_rgb def_rgb_24 = {
+	.red	= { .offset = 16, .length = 8, },
+	.green	= { .offset =  8, .length = 8, },
+	.blue	= { .offset =  0, .length = 8, },
+	.transp = { .offset =  0, .length = 0, },
+	.bits_per_pixel = 24,
+};
+
+static const struct ipu_rgb def_bgr_24 = {
+	.red	= { .offset =  0, .length = 8, },
+	.green	= { .offset =  8, .length = 8, },
+	.blue	= { .offset = 16, .length = 8, },
+	.transp = { .offset =  0, .length = 0, },
+	.bits_per_pixel = 24,
+};
+
+static const struct ipu_rgb def_rgb_16 = {
+	.red	= { .offset = 11, .length = 5, },
+	.green	= { .offset =  5, .length = 6, },
+	.blue	= { .offset =  0, .length = 5, },
+	.transp = { .offset =  0, .length = 0, },
+	.bits_per_pixel = 16,
+};
+
+static const struct ipu_rgb def_bgr_16 = {
+	.red	= { .offset =  0, .length = 5, },
+	.green	= { .offset =  5, .length = 6, },
+	.blue	= { .offset = 11, .length = 5, },
+	.transp = { .offset =  0, .length = 0, },
+	.bits_per_pixel = 16,
+};
+
+#define Y_OFFSET(pix, x, y)	((x) + pix->width * (y))
+#define U_OFFSET(pix, x, y)	((pix->width * pix->height) +		\
+				 (pix->width * (y) / 4) + (x) / 2)
+#define V_OFFSET(pix, x, y)	((pix->width * pix->height) +		\
+				 (pix->width * pix->height / 4) +	\
+				 (pix->width * (y) / 4) + (x) / 2)
+#define U2_OFFSET(pix, x, y)	((pix->width * pix->height) +		\
+				 (pix->width * (y) / 2) + (x) / 2)
+#define V2_OFFSET(pix, x, y)	((pix->width * pix->height) +		\
+				 (pix->width * pix->height / 2) +	\
+				 (pix->width * (y) / 2) + (x) / 2)
+#define UV_OFFSET(pix, x, y)	((pix->width * pix->height) +	\
+				 (pix->width * (y) / 2) + (x))
+#define UV2_OFFSET(pix, x, y)	((pix->width * pix->height) +	\
+				 (pix->width * y) + (x))
+
+int ipu_cpmem_set_fmt(struct ipuv3_channel *ch, u32 drm_fourcc)
+{
+	switch (drm_fourcc) {
+	case DRM_FORMAT_YUV420:
+	case DRM_FORMAT_YVU420:
+		/* pix format */
+		ipu_ch_param_write_field(ch, IPU_FIELD_PFS, 2);
+		/* burst size */
+		ipu_ch_param_write_field(ch, IPU_FIELD_NPB, 31);
+		break;
+	case DRM_FORMAT_YUV422:
+	case DRM_FORMAT_YVU422:
+		/* pix format */
+		ipu_ch_param_write_field(ch, IPU_FIELD_PFS, 1);
+		/* burst size */
+		ipu_ch_param_write_field(ch, IPU_FIELD_NPB, 31);
+		break;
+	case DRM_FORMAT_NV12:
+		/* pix format */
+		ipu_ch_param_write_field(ch, IPU_FIELD_PFS, 4);
+		/* burst size */
+		ipu_ch_param_write_field(ch, IPU_FIELD_NPB, 31);
+		break;
+	case DRM_FORMAT_NV16:
+		/* pix format */
+		ipu_ch_param_write_field(ch, IPU_FIELD_PFS, 3);
+		/* burst size */
+		ipu_ch_param_write_field(ch, IPU_FIELD_NPB, 31);
+		break;
+	case DRM_FORMAT_UYVY:
+		/* bits/pixel */
+		ipu_ch_param_write_field(ch, IPU_FIELD_BPP, 3);
+		/* pix format */
+		ipu_ch_param_write_field(ch, IPU_FIELD_PFS, 0xA);
+		/* burst size */
+		ipu_ch_param_write_field(ch, IPU_FIELD_NPB, 31);
+		break;
+	case DRM_FORMAT_YUYV:
+		/* bits/pixel */
+		ipu_ch_param_write_field(ch, IPU_FIELD_BPP, 3);
+		/* pix format */
+		ipu_ch_param_write_field(ch, IPU_FIELD_PFS, 0x8);
+		/* burst size */
+		ipu_ch_param_write_field(ch, IPU_FIELD_NPB, 31);
+		break;
+	case DRM_FORMAT_ABGR8888:
+	case DRM_FORMAT_XBGR8888:
+		ipu_cpmem_set_format_rgb(ch, &def_bgr_32);
+		break;
+	case DRM_FORMAT_ARGB8888:
+	case DRM_FORMAT_XRGB8888:
+		ipu_cpmem_set_format_rgb(ch, &def_rgb_32);
+		break;
+	case DRM_FORMAT_BGR888:
+		ipu_cpmem_set_format_rgb(ch, &def_bgr_24);
+		break;
+	case DRM_FORMAT_RGB888:
+		ipu_cpmem_set_format_rgb(ch, &def_rgb_24);
+		break;
+	case DRM_FORMAT_RGB565:
+		ipu_cpmem_set_format_rgb(ch, &def_rgb_16);
+		break;
+	case DRM_FORMAT_BGR565:
+		ipu_cpmem_set_format_rgb(ch, &def_bgr_16);
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_fmt);
+
+int ipu_cpmem_set_image(struct ipuv3_channel *ch, struct ipu_image *image)
+{
+	struct v4l2_pix_format *pix = &image->pix;
+	int offset, u_offset, v_offset;
+
+	pr_debug("%s: resolution: %dx%d stride: %d\n",
+		 __func__, pix->width, pix->height,
+		 pix->bytesperline);
+
+	ipu_cpmem_set_resolution(ch, image->rect.width, image->rect.height);
+	ipu_cpmem_set_stride(ch, pix->bytesperline);
+
+	ipu_cpmem_set_fmt(ch, v4l2_pix_fmt_to_drm_fourcc(pix->pixelformat));
+
+	switch (pix->pixelformat) {
+	case V4L2_PIX_FMT_YUV420:
+	case V4L2_PIX_FMT_YVU420:
+		offset = Y_OFFSET(pix, image->rect.left, image->rect.top);
+		u_offset = U_OFFSET(pix, image->rect.left,
+				    image->rect.top) - offset;
+		v_offset = V_OFFSET(pix, image->rect.left,
+				    image->rect.top) - offset;
+
+		ipu_cpmem_set_yuv_planar_full(ch, pix->pixelformat,
+					      pix->bytesperline,
+					      u_offset, v_offset);
+		break;
+	case V4L2_PIX_FMT_YUV422P:
+		offset = Y_OFFSET(pix, image->rect.left, image->rect.top);
+		u_offset = U2_OFFSET(pix, image->rect.left,
+				     image->rect.top) - offset;
+		v_offset = V2_OFFSET(pix, image->rect.left,
+				     image->rect.top) - offset;
+
+		ipu_cpmem_set_yuv_planar_full(ch, pix->pixelformat,
+					      pix->bytesperline,
+					      u_offset, v_offset);
+		break;
+	case V4L2_PIX_FMT_NV12:
+		offset = Y_OFFSET(pix, image->rect.left, image->rect.top);
+		u_offset = UV_OFFSET(pix, image->rect.left,
+				     image->rect.top) - offset;
+		v_offset = 0;
+
+		ipu_cpmem_set_yuv_planar_full(ch, pix->pixelformat,
+					      pix->bytesperline,
+					      u_offset, v_offset);
+		break;
+	case V4L2_PIX_FMT_NV16:
+		offset = Y_OFFSET(pix, image->rect.left, image->rect.top);
+		u_offset = UV2_OFFSET(pix, image->rect.left,
+				      image->rect.top) - offset;
+		v_offset = 0;
+
+		ipu_cpmem_set_yuv_planar_full(ch, pix->pixelformat,
+					      pix->bytesperline,
+					      u_offset, v_offset);
+		break;
+	case V4L2_PIX_FMT_UYVY:
+	case V4L2_PIX_FMT_YUYV:
+	case V4L2_PIX_FMT_RGB565:
+		offset = image->rect.left * 2 +
+			image->rect.top * pix->bytesperline;
+		break;
+	case V4L2_PIX_FMT_RGB32:
+	case V4L2_PIX_FMT_BGR32:
+		offset = image->rect.left * 4 +
+			image->rect.top * pix->bytesperline;
+		break;
+	case V4L2_PIX_FMT_RGB24:
+	case V4L2_PIX_FMT_BGR24:
+		offset = image->rect.left * 3 +
+			image->rect.top * pix->bytesperline;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	ipu_cpmem_set_buffer(ch, 0, image->phys0 + offset);
+	ipu_cpmem_set_buffer(ch, 1, image->phys1 + offset);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_image);
+
+void ipu_cpmem_dump(struct ipuv3_channel *ch)
+{
+	struct ipu_ch_param __iomem *p = ipu_get_cpmem(ch);
+	struct ipu_soc *ipu = ch->ipu;
+	int chno = ch->num;
+
+	dev_dbg(ipu->dev, "ch %d word 0 - %08X %08X %08X %08X %08X\n", chno,
+		readl(&p->word[0].data[0]),
+		readl(&p->word[0].data[1]),
+		readl(&p->word[0].data[2]),
+		readl(&p->word[0].data[3]),
+		readl(&p->word[0].data[4]));
+	dev_dbg(ipu->dev, "ch %d word 1 - %08X %08X %08X %08X %08X\n", chno,
+		readl(&p->word[1].data[0]),
+		readl(&p->word[1].data[1]),
+		readl(&p->word[1].data[2]),
+		readl(&p->word[1].data[3]),
+		readl(&p->word[1].data[4]));
+	dev_dbg(ipu->dev, "PFS 0x%x, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_PFS));
+	dev_dbg(ipu->dev, "BPP 0x%x, ",
+		ipu_ch_param_read_field(ch, IPU_FIELD_BPP));
+	dev_dbg(ipu->dev, "NPB 0x%x\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_NPB));
+
+	dev_dbg(ipu->dev, "FW %d, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_FW));
+	dev_dbg(ipu->dev, "FH %d, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_FH));
+	dev_dbg(ipu->dev, "EBA0 0x%x\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_EBA0) << 3);
+	dev_dbg(ipu->dev, "EBA1 0x%x\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_EBA1) << 3);
+	dev_dbg(ipu->dev, "Stride %d\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_SL));
+	dev_dbg(ipu->dev, "scan_order %d\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_SO));
+	dev_dbg(ipu->dev, "uv_stride %d\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_SLUV));
+	dev_dbg(ipu->dev, "u_offset 0x%x\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_UBO) << 3);
+	dev_dbg(ipu->dev, "v_offset 0x%x\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_VBO) << 3);
+
+	dev_dbg(ipu->dev, "Width0 %d+1, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_WID0));
+	dev_dbg(ipu->dev, "Width1 %d+1, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_WID1));
+	dev_dbg(ipu->dev, "Width2 %d+1, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_WID2));
+	dev_dbg(ipu->dev, "Width3 %d+1, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_WID3));
+	dev_dbg(ipu->dev, "Offset0 %d, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_OFS0));
+	dev_dbg(ipu->dev, "Offset1 %d, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_OFS1));
+	dev_dbg(ipu->dev, "Offset2 %d, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_OFS2));
+	dev_dbg(ipu->dev, "Offset3 %d\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_OFS3));
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_dump);
+
+int ipu_cpmem_init(struct ipu_soc *ipu, struct device *dev, unsigned long base)
+{
+	struct ipu_cpmem *cpmem;
+
+	cpmem = devm_kzalloc(dev, sizeof(*cpmem), GFP_KERNEL);
+	if (!cpmem)
+		return -ENOMEM;
+
+	ipu->cpmem_priv = cpmem;
+
+	spin_lock_init(&cpmem->lock);
+	cpmem->base = devm_ioremap(dev, base, SZ_128K);
+	if (!cpmem->base)
+		return -ENOMEM;
+
+	dev_dbg(dev, "CPMEM base: 0x%08lx remapped to %p\n",
+		base, cpmem->base);
+	cpmem->ipu = ipu;
+
+	return 0;
+}
+
+void ipu_cpmem_exit(struct ipu_soc *ipu)
+{
+}
diff --git a/drivers/gpu/ipu-v3/ipu-csi.c b/drivers/gpu/ipu-v3/ipu-csi.c
new file mode 100644
index 000000000000..752cdd2da89a
--- /dev/null
+++ b/drivers/gpu/ipu-v3/ipu-csi.c
@@ -0,0 +1,741 @@
+/*
+ * Copyright (C) 2012-2014 Mentor Graphics Inc.
+ * Copyright (C) 2005-2009 Freescale Semiconductor, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2 of the License, or (at your
+ * option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
+ * for more details.
+ */
+#include <linux/export.h>
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/errno.h>
+#include <linux/delay.h>
+#include <linux/io.h>
+#include <linux/err.h>
+#include <linux/platform_device.h>
+#include <linux/videodev2.h>
+#include <uapi/linux/v4l2-mediabus.h>
+#include <linux/clk.h>
+#include <linux/clk-provider.h>
+#include <linux/clkdev.h>
+
+#include "ipu-prv.h"
+
+struct ipu_csi {
+	void __iomem *base;
+	int id;
+	u32 module;
+	struct clk *clk_ipu;	/* IPU bus clock */
+	spinlock_t lock;
+	bool inuse;
+	struct ipu_soc *ipu;
+};
+
+/* CSI Register Offsets */
+#define CSI_SENS_CONF		0x0000
+#define CSI_SENS_FRM_SIZE	0x0004
+#define CSI_ACT_FRM_SIZE	0x0008
+#define CSI_OUT_FRM_CTRL	0x000c
+#define CSI_TST_CTRL		0x0010
+#define CSI_CCIR_CODE_1		0x0014
+#define CSI_CCIR_CODE_2		0x0018
+#define CSI_CCIR_CODE_3		0x001c
+#define CSI_MIPI_DI		0x0020
+#define CSI_SKIP		0x0024
+#define CSI_CPD_CTRL		0x0028
+#define CSI_CPD_RC(n)		(0x002c + ((n)*4))
+#define CSI_CPD_RS(n)		(0x004c + ((n)*4))
+#define CSI_CPD_GRC(n)		(0x005c + ((n)*4))
+#define CSI_CPD_GRS(n)		(0x007c + ((n)*4))
+#define CSI_CPD_GBC(n)		(0x008c + ((n)*4))
+#define CSI_CPD_GBS(n)		(0x00Ac + ((n)*4))
+#define CSI_CPD_BC(n)		(0x00Bc + ((n)*4))
+#define CSI_CPD_BS(n)		(0x00Dc + ((n)*4))
+#define CSI_CPD_OFFSET1		0x00ec
+#define CSI_CPD_OFFSET2		0x00f0
+
+/* CSI Register Fields */
+#define CSI_SENS_CONF_DATA_FMT_SHIFT		8
+#define CSI_SENS_CONF_DATA_FMT_MASK		0x00000700
+#define CSI_SENS_CONF_DATA_FMT_RGB_YUV444	0L
+#define CSI_SENS_CONF_DATA_FMT_YUV422_YUYV	1L
+#define CSI_SENS_CONF_DATA_FMT_YUV422_UYVY	2L
+#define CSI_SENS_CONF_DATA_FMT_BAYER		3L
+#define CSI_SENS_CONF_DATA_FMT_RGB565		4L
+#define CSI_SENS_CONF_DATA_FMT_RGB555		5L
+#define CSI_SENS_CONF_DATA_FMT_RGB444		6L
+#define CSI_SENS_CONF_DATA_FMT_JPEG		7L
+
+#define CSI_SENS_CONF_VSYNC_POL_SHIFT		0
+#define CSI_SENS_CONF_HSYNC_POL_SHIFT		1
+#define CSI_SENS_CONF_DATA_POL_SHIFT		2
+#define CSI_SENS_CONF_PIX_CLK_POL_SHIFT		3
+#define CSI_SENS_CONF_SENS_PRTCL_MASK		0x00000070
+#define CSI_SENS_CONF_SENS_PRTCL_SHIFT		4
+#define CSI_SENS_CONF_PACK_TIGHT_SHIFT		7
+#define CSI_SENS_CONF_DATA_WIDTH_SHIFT		11
+#define CSI_SENS_CONF_EXT_VSYNC_SHIFT		15
+#define CSI_SENS_CONF_DIVRATIO_SHIFT		16
+
+#define CSI_SENS_CONF_DIVRATIO_MASK		0x00ff0000
+#define CSI_SENS_CONF_DATA_DEST_SHIFT		24
+#define CSI_SENS_CONF_DATA_DEST_MASK		0x07000000
+#define CSI_SENS_CONF_JPEG8_EN_SHIFT		27
+#define CSI_SENS_CONF_JPEG_EN_SHIFT		28
+#define CSI_SENS_CONF_FORCE_EOF_SHIFT		29
+#define CSI_SENS_CONF_DATA_EN_POL_SHIFT		31
+
+#define CSI_DATA_DEST_IC			2
+#define CSI_DATA_DEST_IDMAC			4
+
+#define CSI_CCIR_ERR_DET_EN			0x01000000
+#define CSI_HORI_DOWNSIZE_EN			0x80000000
+#define CSI_VERT_DOWNSIZE_EN			0x40000000
+#define CSI_TEST_GEN_MODE_EN			0x01000000
+
+#define CSI_HSC_MASK				0x1fff0000
+#define CSI_HSC_SHIFT				16
+#define CSI_VSC_MASK				0x00000fff
+#define CSI_VSC_SHIFT				0
+
+#define CSI_TEST_GEN_R_MASK			0x000000ff
+#define CSI_TEST_GEN_R_SHIFT			0
+#define CSI_TEST_GEN_G_MASK			0x0000ff00
+#define CSI_TEST_GEN_G_SHIFT			8
+#define CSI_TEST_GEN_B_MASK			0x00ff0000
+#define CSI_TEST_GEN_B_SHIFT			16
+
+#define CSI_MAX_RATIO_SKIP_SMFC_MASK		0x00000007
+#define CSI_MAX_RATIO_SKIP_SMFC_SHIFT		0
+#define CSI_SKIP_SMFC_MASK			0x000000f8
+#define CSI_SKIP_SMFC_SHIFT			3
+#define CSI_ID_2_SKIP_MASK			0x00000300
+#define CSI_ID_2_SKIP_SHIFT			8
+
+#define CSI_COLOR_FIRST_ROW_MASK		0x00000002
+#define CSI_COLOR_FIRST_COMP_MASK		0x00000001
+
+/* MIPI CSI-2 data types */
+#define MIPI_DT_YUV420		0x18 /* YYY.../UYVY.... */
+#define MIPI_DT_YUV420_LEGACY	0x1a /* UYY.../VYY...   */
+#define MIPI_DT_YUV422		0x1e /* UYVY...         */
+#define MIPI_DT_RGB444		0x20
+#define MIPI_DT_RGB555		0x21
+#define MIPI_DT_RGB565		0x22
+#define MIPI_DT_RGB666		0x23
+#define MIPI_DT_RGB888		0x24
+#define MIPI_DT_RAW6		0x28
+#define MIPI_DT_RAW7		0x29
+#define MIPI_DT_RAW8		0x2a
+#define MIPI_DT_RAW10		0x2b
+#define MIPI_DT_RAW12		0x2c
+#define MIPI_DT_RAW14		0x2d
+
+/*
+ * Bitfield of CSI bus signal polarities and modes.
+ */
+struct ipu_csi_bus_config {
+	unsigned data_width:4;
+	unsigned clk_mode:3;
+	unsigned ext_vsync:1;
+	unsigned vsync_pol:1;
+	unsigned hsync_pol:1;
+	unsigned pixclk_pol:1;
+	unsigned data_pol:1;
+	unsigned sens_clksrc:1;
+	unsigned pack_tight:1;
+	unsigned force_eof:1;
+	unsigned data_en_pol:1;
+
+	unsigned data_fmt;
+	unsigned mipi_dt;
+};
+
+/*
+ * Enumeration of CSI data bus widths.
+ */
+enum ipu_csi_data_width {
+	IPU_CSI_DATA_WIDTH_4   = 0,
+	IPU_CSI_DATA_WIDTH_8   = 1,
+	IPU_CSI_DATA_WIDTH_10  = 3,
+	IPU_CSI_DATA_WIDTH_12  = 5,
+	IPU_CSI_DATA_WIDTH_16  = 9,
+};
+
+/*
+ * Enumeration of CSI clock modes.
+ */
+enum ipu_csi_clk_mode {
+	IPU_CSI_CLK_MODE_GATED_CLK,
+	IPU_CSI_CLK_MODE_NONGATED_CLK,
+	IPU_CSI_CLK_MODE_CCIR656_PROGRESSIVE,
+	IPU_CSI_CLK_MODE_CCIR656_INTERLACED,
+	IPU_CSI_CLK_MODE_CCIR1120_PROGRESSIVE_DDR,
+	IPU_CSI_CLK_MODE_CCIR1120_PROGRESSIVE_SDR,
+	IPU_CSI_CLK_MODE_CCIR1120_INTERLACED_DDR,
+	IPU_CSI_CLK_MODE_CCIR1120_INTERLACED_SDR,
+};
+
+static inline u32 ipu_csi_read(struct ipu_csi *csi, unsigned offset)
+{
+	return readl(csi->base + offset);
+}
+
+static inline void ipu_csi_write(struct ipu_csi *csi, u32 value,
+				 unsigned offset)
+{
+	writel(value, csi->base + offset);
+}
+
+/*
+ * Set mclk division ratio for generating test mode mclk. Only used
+ * for test generator.
+ */
+static int ipu_csi_set_testgen_mclk(struct ipu_csi *csi, u32 pixel_clk,
+					u32 ipu_clk)
+{
+	u32 temp;
+	u32 div_ratio;
+
+	div_ratio = (ipu_clk / pixel_clk) - 1;
+
+	if (div_ratio > 0xFF || div_ratio < 0) {
+		dev_err(csi->ipu->dev,
+			"value of pixel_clk extends normal range\n");
+		return -EINVAL;
+	}
+
+	temp = ipu_csi_read(csi, CSI_SENS_CONF);
+	temp &= ~CSI_SENS_CONF_DIVRATIO_MASK;
+	ipu_csi_write(csi, temp | (div_ratio << CSI_SENS_CONF_DIVRATIO_SHIFT),
+			  CSI_SENS_CONF);
+
+	return 0;
+}
+
+/*
+ * Find the CSI data format and data width for the given V4L2 media
+ * bus pixel format code.
+ */
+static int mbus_code_to_bus_cfg(struct ipu_csi_bus_config *cfg, u32 mbus_code)
+{
+	switch (mbus_code) {
+	case MEDIA_BUS_FMT_BGR565_2X8_BE:
+	case MEDIA_BUS_FMT_BGR565_2X8_LE:
+	case MEDIA_BUS_FMT_RGB565_2X8_BE:
+	case MEDIA_BUS_FMT_RGB565_2X8_LE:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_RGB565;
+		cfg->mipi_dt = MIPI_DT_RGB565;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_8;
+		break;
+	case MEDIA_BUS_FMT_RGB444_2X8_PADHI_BE:
+	case MEDIA_BUS_FMT_RGB444_2X8_PADHI_LE:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_RGB444;
+		cfg->mipi_dt = MIPI_DT_RGB444;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_8;
+		break;
+	case MEDIA_BUS_FMT_RGB555_2X8_PADHI_BE:
+	case MEDIA_BUS_FMT_RGB555_2X8_PADHI_LE:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_RGB555;
+		cfg->mipi_dt = MIPI_DT_RGB555;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_8;
+		break;
+	case MEDIA_BUS_FMT_UYVY8_2X8:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_YUV422_UYVY;
+		cfg->mipi_dt = MIPI_DT_YUV422;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_8;
+		break;
+	case MEDIA_BUS_FMT_YUYV8_2X8:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_YUV422_YUYV;
+		cfg->mipi_dt = MIPI_DT_YUV422;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_8;
+		break;
+	case MEDIA_BUS_FMT_UYVY8_1X16:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_YUV422_UYVY;
+		cfg->mipi_dt = MIPI_DT_YUV422;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_16;
+		break;
+	case MEDIA_BUS_FMT_YUYV8_1X16:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_YUV422_YUYV;
+		cfg->mipi_dt = MIPI_DT_YUV422;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_16;
+		break;
+	case MEDIA_BUS_FMT_SBGGR8_1X8:
+	case MEDIA_BUS_FMT_SGBRG8_1X8:
+	case MEDIA_BUS_FMT_SGRBG8_1X8:
+	case MEDIA_BUS_FMT_SRGGB8_1X8:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_BAYER;
+		cfg->mipi_dt = MIPI_DT_RAW8;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_8;
+		break;
+	case MEDIA_BUS_FMT_SBGGR10_DPCM8_1X8:
+	case MEDIA_BUS_FMT_SGBRG10_DPCM8_1X8:
+	case MEDIA_BUS_FMT_SGRBG10_DPCM8_1X8:
+	case MEDIA_BUS_FMT_SRGGB10_DPCM8_1X8:
+	case MEDIA_BUS_FMT_SBGGR10_2X8_PADHI_BE:
+	case MEDIA_BUS_FMT_SBGGR10_2X8_PADHI_LE:
+	case MEDIA_BUS_FMT_SBGGR10_2X8_PADLO_BE:
+	case MEDIA_BUS_FMT_SBGGR10_2X8_PADLO_LE:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_BAYER;
+		cfg->mipi_dt = MIPI_DT_RAW10;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_8;
+		break;
+	case MEDIA_BUS_FMT_SBGGR10_1X10:
+	case MEDIA_BUS_FMT_SGBRG10_1X10:
+	case MEDIA_BUS_FMT_SGRBG10_1X10:
+	case MEDIA_BUS_FMT_SRGGB10_1X10:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_BAYER;
+		cfg->mipi_dt = MIPI_DT_RAW10;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_10;
+		break;
+	case MEDIA_BUS_FMT_SBGGR12_1X12:
+	case MEDIA_BUS_FMT_SGBRG12_1X12:
+	case MEDIA_BUS_FMT_SGRBG12_1X12:
+	case MEDIA_BUS_FMT_SRGGB12_1X12:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_BAYER;
+		cfg->mipi_dt = MIPI_DT_RAW12;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_12;
+		break;
+	case MEDIA_BUS_FMT_JPEG_1X8:
+		/* TODO */
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_JPEG;
+		cfg->mipi_dt = MIPI_DT_RAW8;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_8;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+/*
+ * Fill a CSI bus config struct from mbus_config and mbus_framefmt.
+ */
+static void fill_csi_bus_cfg(struct ipu_csi_bus_config *csicfg,
+				 struct v4l2_mbus_config *mbus_cfg,
+				 struct v4l2_mbus_framefmt *mbus_fmt)
+{
+	memset(csicfg, 0, sizeof(*csicfg));
+
+	mbus_code_to_bus_cfg(csicfg, mbus_fmt->code);
+
+	switch (mbus_cfg->type) {
+	case V4L2_MBUS_PARALLEL:
+		csicfg->ext_vsync = 1;
+		csicfg->vsync_pol = (mbus_cfg->flags &
+				     V4L2_MBUS_VSYNC_ACTIVE_LOW) ? 1 : 0;
+		csicfg->hsync_pol = (mbus_cfg->flags &
+				     V4L2_MBUS_HSYNC_ACTIVE_LOW) ? 1 : 0;
+		csicfg->pixclk_pol = (mbus_cfg->flags &
+				      V4L2_MBUS_PCLK_SAMPLE_FALLING) ? 1 : 0;
+		csicfg->clk_mode = IPU_CSI_CLK_MODE_GATED_CLK;
+		break;
+	case V4L2_MBUS_BT656:
+		csicfg->ext_vsync = 0;
+		if (V4L2_FIELD_HAS_BOTH(mbus_fmt->field))
+			csicfg->clk_mode = IPU_CSI_CLK_MODE_CCIR656_INTERLACED;
+		else
+			csicfg->clk_mode = IPU_CSI_CLK_MODE_CCIR656_PROGRESSIVE;
+		break;
+	case V4L2_MBUS_CSI2:
+		/*
+		 * MIPI CSI-2 requires non gated clock mode, all other
+		 * parameters are not applicable for MIPI CSI-2 bus.
+		 */
+		csicfg->clk_mode = IPU_CSI_CLK_MODE_NONGATED_CLK;
+		break;
+	default:
+		/* will never get here, keep compiler quiet */
+		break;
+	}
+}
+
+int ipu_csi_init_interface(struct ipu_csi *csi,
+			   struct v4l2_mbus_config *mbus_cfg,
+			   struct v4l2_mbus_framefmt *mbus_fmt)
+{
+	struct ipu_csi_bus_config cfg;
+	unsigned long flags;
+	u32 data = 0;
+
+	fill_csi_bus_cfg(&cfg, mbus_cfg, mbus_fmt);
+
+	/* Set the CSI_SENS_CONF register remaining fields */
+	data |= cfg.data_width << CSI_SENS_CONF_DATA_WIDTH_SHIFT |
+		cfg.data_fmt << CSI_SENS_CONF_DATA_FMT_SHIFT |
+		cfg.data_pol << CSI_SENS_CONF_DATA_POL_SHIFT |
+		cfg.vsync_pol << CSI_SENS_CONF_VSYNC_POL_SHIFT |
+		cfg.hsync_pol << CSI_SENS_CONF_HSYNC_POL_SHIFT |
+		cfg.pixclk_pol << CSI_SENS_CONF_PIX_CLK_POL_SHIFT |
+		cfg.ext_vsync << CSI_SENS_CONF_EXT_VSYNC_SHIFT |
+		cfg.clk_mode << CSI_SENS_CONF_SENS_PRTCL_SHIFT |
+		cfg.pack_tight << CSI_SENS_CONF_PACK_TIGHT_SHIFT |
+		cfg.force_eof << CSI_SENS_CONF_FORCE_EOF_SHIFT |
+		cfg.data_en_pol << CSI_SENS_CONF_DATA_EN_POL_SHIFT;
+
+	spin_lock_irqsave(&csi->lock, flags);
+
+	ipu_csi_write(csi, data, CSI_SENS_CONF);
+
+	/* Setup sensor frame size */
+	ipu_csi_write(csi,
+		      (mbus_fmt->width - 1) | ((mbus_fmt->height - 1) << 16),
+		      CSI_SENS_FRM_SIZE);
+
+	/* Set CCIR registers */
+
+	switch (cfg.clk_mode) {
+	case IPU_CSI_CLK_MODE_CCIR656_PROGRESSIVE:
+		ipu_csi_write(csi, 0x40030, CSI_CCIR_CODE_1);
+		ipu_csi_write(csi, 0xFF0000, CSI_CCIR_CODE_3);
+		break;
+	case IPU_CSI_CLK_MODE_CCIR656_INTERLACED:
+		if (mbus_fmt->width == 720 && mbus_fmt->height == 576) {
+			/*
+			 * PAL case
+			 *
+			 * Field0BlankEnd = 0x6, Field0BlankStart = 0x2,
+			 * Field0ActiveEnd = 0x4, Field0ActiveStart = 0
+			 * Field1BlankEnd = 0x7, Field1BlankStart = 0x3,
+			 * Field1ActiveEnd = 0x5, Field1ActiveStart = 0x1
+			 */
+			ipu_csi_write(csi, 0x40596 | CSI_CCIR_ERR_DET_EN,
+					  CSI_CCIR_CODE_1);
+			ipu_csi_write(csi, 0xD07DF, CSI_CCIR_CODE_2);
+			ipu_csi_write(csi, 0xFF0000, CSI_CCIR_CODE_3);
+
+		} else if (mbus_fmt->width == 720 && mbus_fmt->height == 480) {
+			/*
+			 * NTSC case
+			 *
+			 * Field0BlankEnd = 0x7, Field0BlankStart = 0x3,
+			 * Field0ActiveEnd = 0x5, Field0ActiveStart = 0x1
+			 * Field1BlankEnd = 0x6, Field1BlankStart = 0x2,
+			 * Field1ActiveEnd = 0x4, Field1ActiveStart = 0
+			 */
+			ipu_csi_write(csi, 0xD07DF | CSI_CCIR_ERR_DET_EN,
+					  CSI_CCIR_CODE_1);
+			ipu_csi_write(csi, 0x40596, CSI_CCIR_CODE_2);
+			ipu_csi_write(csi, 0xFF0000, CSI_CCIR_CODE_3);
+		} else {
+			dev_err(csi->ipu->dev,
+				"Unsupported CCIR656 interlaced video mode\n");
+			spin_unlock_irqrestore(&csi->lock, flags);
+			return -EINVAL;
+		}
+		break;
+	case IPU_CSI_CLK_MODE_CCIR1120_PROGRESSIVE_DDR:
+	case IPU_CSI_CLK_MODE_CCIR1120_PROGRESSIVE_SDR:
+	case IPU_CSI_CLK_MODE_CCIR1120_INTERLACED_DDR:
+	case IPU_CSI_CLK_MODE_CCIR1120_INTERLACED_SDR:
+		ipu_csi_write(csi, 0x40030 | CSI_CCIR_ERR_DET_EN,
+				   CSI_CCIR_CODE_1);
+		ipu_csi_write(csi, 0xFF0000, CSI_CCIR_CODE_3);
+		break;
+	case IPU_CSI_CLK_MODE_GATED_CLK:
+	case IPU_CSI_CLK_MODE_NONGATED_CLK:
+		ipu_csi_write(csi, 0, CSI_CCIR_CODE_1);
+		break;
+	}
+
+	dev_dbg(csi->ipu->dev, "CSI_SENS_CONF = 0x%08X\n",
+		ipu_csi_read(csi, CSI_SENS_CONF));
+	dev_dbg(csi->ipu->dev, "CSI_ACT_FRM_SIZE = 0x%08X\n",
+		ipu_csi_read(csi, CSI_ACT_FRM_SIZE));
+
+	spin_unlock_irqrestore(&csi->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_csi_init_interface);
+
+bool ipu_csi_is_interlaced(struct ipu_csi *csi)
+{
+	unsigned long flags;
+	u32 sensor_protocol;
+
+	spin_lock_irqsave(&csi->lock, flags);
+	sensor_protocol =
+		(ipu_csi_read(csi, CSI_SENS_CONF) &
+		 CSI_SENS_CONF_SENS_PRTCL_MASK) >>
+		CSI_SENS_CONF_SENS_PRTCL_SHIFT;
+	spin_unlock_irqrestore(&csi->lock, flags);
+
+	switch (sensor_protocol) {
+	case IPU_CSI_CLK_MODE_GATED_CLK:
+	case IPU_CSI_CLK_MODE_NONGATED_CLK:
+	case IPU_CSI_CLK_MODE_CCIR656_PROGRESSIVE:
+	case IPU_CSI_CLK_MODE_CCIR1120_PROGRESSIVE_DDR:
+	case IPU_CSI_CLK_MODE_CCIR1120_PROGRESSIVE_SDR:
+		return false;
+	case IPU_CSI_CLK_MODE_CCIR656_INTERLACED:
+	case IPU_CSI_CLK_MODE_CCIR1120_INTERLACED_DDR:
+	case IPU_CSI_CLK_MODE_CCIR1120_INTERLACED_SDR:
+		return true;
+	default:
+		dev_err(csi->ipu->dev,
+			"CSI %d sensor protocol unsupported\n", csi->id);
+		return false;
+	}
+}
+EXPORT_SYMBOL_GPL(ipu_csi_is_interlaced);
+
+void ipu_csi_get_window(struct ipu_csi *csi, struct v4l2_rect *w)
+{
+	unsigned long flags;
+	u32 reg;
+
+	spin_lock_irqsave(&csi->lock, flags);
+
+	reg = ipu_csi_read(csi, CSI_ACT_FRM_SIZE);
+	w->width = (reg & 0xFFFF) + 1;
+	w->height = (reg >> 16 & 0xFFFF) + 1;
+
+	reg = ipu_csi_read(csi, CSI_OUT_FRM_CTRL);
+	w->left = (reg & CSI_HSC_MASK) >> CSI_HSC_SHIFT;
+	w->top = (reg & CSI_VSC_MASK) >> CSI_VSC_SHIFT;
+
+	spin_unlock_irqrestore(&csi->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_csi_get_window);
+
+void ipu_csi_set_window(struct ipu_csi *csi, struct v4l2_rect *w)
+{
+	unsigned long flags;
+	u32 reg;
+
+	spin_lock_irqsave(&csi->lock, flags);
+
+	ipu_csi_write(csi, (w->width - 1) | ((w->height - 1) << 16),
+			  CSI_ACT_FRM_SIZE);
+
+	reg = ipu_csi_read(csi, CSI_OUT_FRM_CTRL);
+	reg &= ~(CSI_HSC_MASK | CSI_VSC_MASK);
+	reg |= ((w->top << CSI_VSC_SHIFT) | (w->left << CSI_HSC_SHIFT));
+	ipu_csi_write(csi, reg, CSI_OUT_FRM_CTRL);
+
+	spin_unlock_irqrestore(&csi->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_csi_set_window);
+
+void ipu_csi_set_test_generator(struct ipu_csi *csi, bool active,
+				u32 r_value, u32 g_value, u32 b_value,
+				u32 pix_clk)
+{
+	unsigned long flags;
+	u32 ipu_clk = clk_get_rate(csi->clk_ipu);
+	u32 temp;
+
+	spin_lock_irqsave(&csi->lock, flags);
+
+	temp = ipu_csi_read(csi, CSI_TST_CTRL);
+
+	if (active == false) {
+		temp &= ~CSI_TEST_GEN_MODE_EN;
+		ipu_csi_write(csi, temp, CSI_TST_CTRL);
+	} else {
+		/* Set sensb_mclk div_ratio */
+		ipu_csi_set_testgen_mclk(csi, pix_clk, ipu_clk);
+
+		temp &= ~(CSI_TEST_GEN_R_MASK | CSI_TEST_GEN_G_MASK |
+			  CSI_TEST_GEN_B_MASK);
+		temp |= CSI_TEST_GEN_MODE_EN;
+		temp |= (r_value << CSI_TEST_GEN_R_SHIFT) |
+			(g_value << CSI_TEST_GEN_G_SHIFT) |
+			(b_value << CSI_TEST_GEN_B_SHIFT);
+		ipu_csi_write(csi, temp, CSI_TST_CTRL);
+	}
+
+	spin_unlock_irqrestore(&csi->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_csi_set_test_generator);
+
+int ipu_csi_set_mipi_datatype(struct ipu_csi *csi, u32 vc,
+			      struct v4l2_mbus_framefmt *mbus_fmt)
+{
+	struct ipu_csi_bus_config cfg;
+	unsigned long flags;
+	u32 temp;
+
+	if (vc > 3)
+		return -EINVAL;
+
+	mbus_code_to_bus_cfg(&cfg, mbus_fmt->code);
+
+	spin_lock_irqsave(&csi->lock, flags);
+
+	temp = ipu_csi_read(csi, CSI_MIPI_DI);
+	temp &= ~(0xff << (vc * 8));
+	temp |= (cfg.mipi_dt << (vc * 8));
+	ipu_csi_write(csi, temp, CSI_MIPI_DI);
+
+	spin_unlock_irqrestore(&csi->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_csi_set_mipi_datatype);
+
+int ipu_csi_set_skip_smfc(struct ipu_csi *csi, u32 skip,
+			  u32 max_ratio, u32 id)
+{
+	unsigned long flags;
+	u32 temp;
+
+	if (max_ratio > 5 || id > 3)
+		return -EINVAL;
+
+	spin_lock_irqsave(&csi->lock, flags);
+
+	temp = ipu_csi_read(csi, CSI_SKIP);
+	temp &= ~(CSI_MAX_RATIO_SKIP_SMFC_MASK | CSI_ID_2_SKIP_MASK |
+		  CSI_SKIP_SMFC_MASK);
+	temp |= (max_ratio << CSI_MAX_RATIO_SKIP_SMFC_SHIFT) |
+		(id << CSI_ID_2_SKIP_SHIFT) |
+		(skip << CSI_SKIP_SMFC_SHIFT);
+	ipu_csi_write(csi, temp, CSI_SKIP);
+
+	spin_unlock_irqrestore(&csi->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_csi_set_skip_smfc);
+
+int ipu_csi_set_dest(struct ipu_csi *csi, enum ipu_csi_dest csi_dest)
+{
+	unsigned long flags;
+	u32 csi_sens_conf, dest;
+
+	if (csi_dest == IPU_CSI_DEST_IDMAC)
+		dest = CSI_DATA_DEST_IDMAC;
+	else
+		dest = CSI_DATA_DEST_IC; /* IC or VDIC */
+
+	spin_lock_irqsave(&csi->lock, flags);
+
+	csi_sens_conf = ipu_csi_read(csi, CSI_SENS_CONF);
+	csi_sens_conf &= ~CSI_SENS_CONF_DATA_DEST_MASK;
+	csi_sens_conf |= (dest << CSI_SENS_CONF_DATA_DEST_SHIFT);
+	ipu_csi_write(csi, csi_sens_conf, CSI_SENS_CONF);
+
+	spin_unlock_irqrestore(&csi->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_csi_set_dest);
+
+int ipu_csi_enable(struct ipu_csi *csi)
+{
+	ipu_module_enable(csi->ipu, csi->module);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_csi_enable);
+
+int ipu_csi_disable(struct ipu_csi *csi)
+{
+	ipu_module_disable(csi->ipu, csi->module);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_csi_disable);
+
+struct ipu_csi *ipu_csi_get(struct ipu_soc *ipu, int id)
+{
+	unsigned long flags;
+	struct ipu_csi *csi, *ret;
+
+	if (id > 1)
+		return ERR_PTR(-EINVAL);
+
+	csi = ipu->csi_priv[id];
+	ret = csi;
+
+	spin_lock_irqsave(&csi->lock, flags);
+
+	if (csi->inuse) {
+		ret = ERR_PTR(-EBUSY);
+		goto unlock;
+	}
+
+	csi->inuse = true;
+unlock:
+	spin_unlock_irqrestore(&csi->lock, flags);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(ipu_csi_get);
+
+void ipu_csi_put(struct ipu_csi *csi)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&csi->lock, flags);
+	csi->inuse = false;
+	spin_unlock_irqrestore(&csi->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_csi_put);
+
+int ipu_csi_init(struct ipu_soc *ipu, struct device *dev, int id,
+		 unsigned long base, u32 module, struct clk *clk_ipu)
+{
+	struct ipu_csi *csi;
+
+	if (id > 1)
+		return -ENODEV;
+
+	csi = devm_kzalloc(dev, sizeof(*csi), GFP_KERNEL);
+	if (!csi)
+		return -ENOMEM;
+
+	ipu->csi_priv[id] = csi;
+
+	spin_lock_init(&csi->lock);
+	csi->module = module;
+	csi->id = id;
+	csi->clk_ipu = clk_ipu;
+	csi->base = devm_ioremap(dev, base, PAGE_SIZE);
+	if (!csi->base)
+		return -ENOMEM;
+
+	dev_dbg(dev, "CSI%d base: 0x%08lx remapped to %p\n",
+		id, base, csi->base);
+	csi->ipu = ipu;
+
+	return 0;
+}
+
+void ipu_csi_exit(struct ipu_soc *ipu, int id)
+{
+}
+
+void ipu_csi_dump(struct ipu_csi *csi)
+{
+	dev_dbg(csi->ipu->dev, "CSI_SENS_CONF:     %08x\n",
+		ipu_csi_read(csi, CSI_SENS_CONF));
+	dev_dbg(csi->ipu->dev, "CSI_SENS_FRM_SIZE: %08x\n",
+		ipu_csi_read(csi, CSI_SENS_FRM_SIZE));
+	dev_dbg(csi->ipu->dev, "CSI_ACT_FRM_SIZE:  %08x\n",
+		ipu_csi_read(csi, CSI_ACT_FRM_SIZE));
+	dev_dbg(csi->ipu->dev, "CSI_OUT_FRM_CTRL:  %08x\n",
+		ipu_csi_read(csi, CSI_OUT_FRM_CTRL));
+	dev_dbg(csi->ipu->dev, "CSI_TST_CTRL:      %08x\n",
+		ipu_csi_read(csi, CSI_TST_CTRL));
+	dev_dbg(csi->ipu->dev, "CSI_CCIR_CODE_1:   %08x\n",
+		ipu_csi_read(csi, CSI_CCIR_CODE_1));
+	dev_dbg(csi->ipu->dev, "CSI_CCIR_CODE_2:   %08x\n",
+		ipu_csi_read(csi, CSI_CCIR_CODE_2));
+	dev_dbg(csi->ipu->dev, "CSI_CCIR_CODE_3:   %08x\n",
+		ipu_csi_read(csi, CSI_CCIR_CODE_3));
+	dev_dbg(csi->ipu->dev, "CSI_MIPI_DI:       %08x\n",
+		ipu_csi_read(csi, CSI_MIPI_DI));
+	dev_dbg(csi->ipu->dev, "CSI_SKIP:          %08x\n",
+		ipu_csi_read(csi, CSI_SKIP));
+}
+EXPORT_SYMBOL_GPL(ipu_csi_dump);
diff --git a/drivers/gpu/ipu-v3/ipu-ic.c b/drivers/gpu/ipu-v3/ipu-ic.c
new file mode 100644
index 000000000000..ad75588e1629
--- /dev/null
+++ b/drivers/gpu/ipu-v3/ipu-ic.c
@@ -0,0 +1,778 @@
+/*
+ * Copyright (C) 2012-2014 Mentor Graphics Inc.
+ * Copyright 2005-2012 Freescale Semiconductor, Inc. All Rights Reserved.
+ *
+ * The code contained herein is licensed under the GNU General Public
+ * License. You may obtain a copy of the GNU General Public License
+ * Version 2 or later at the following locations:
+ *
+ * http://www.opensource.org/licenses/gpl-license.html
+ * http://www.gnu.org/copyleft/gpl.html
+ */
+
+#include <linux/types.h>
+#include <linux/init.h>
+#include <linux/errno.h>
+#include <linux/spinlock.h>
+#include <linux/bitrev.h>
+#include <linux/io.h>
+#include <linux/err.h>
+#include "ipu-prv.h"
+
+/* IC Register Offsets */
+#define IC_CONF                 0x0000
+#define IC_PRP_ENC_RSC          0x0004
+#define IC_PRP_VF_RSC           0x0008
+#define IC_PP_RSC               0x000C
+#define IC_CMBP_1               0x0010
+#define IC_CMBP_2               0x0014
+#define IC_IDMAC_1              0x0018
+#define IC_IDMAC_2              0x001C
+#define IC_IDMAC_3              0x0020
+#define IC_IDMAC_4              0x0024
+
+/* IC Register Fields */
+#define IC_CONF_PRPENC_EN       (1 << 0)
+#define IC_CONF_PRPENC_CSC1     (1 << 1)
+#define IC_CONF_PRPENC_ROT_EN   (1 << 2)
+#define IC_CONF_PRPVF_EN        (1 << 8)
+#define IC_CONF_PRPVF_CSC1      (1 << 9)
+#define IC_CONF_PRPVF_CSC2      (1 << 10)
+#define IC_CONF_PRPVF_CMB       (1 << 11)
+#define IC_CONF_PRPVF_ROT_EN    (1 << 12)
+#define IC_CONF_PP_EN           (1 << 16)
+#define IC_CONF_PP_CSC1         (1 << 17)
+#define IC_CONF_PP_CSC2         (1 << 18)
+#define IC_CONF_PP_CMB          (1 << 19)
+#define IC_CONF_PP_ROT_EN       (1 << 20)
+#define IC_CONF_IC_GLB_LOC_A    (1 << 28)
+#define IC_CONF_KEY_COLOR_EN    (1 << 29)
+#define IC_CONF_RWS_EN          (1 << 30)
+#define IC_CONF_CSI_MEM_WR_EN   (1 << 31)
+
+#define IC_IDMAC_1_CB0_BURST_16         (1 << 0)
+#define IC_IDMAC_1_CB1_BURST_16         (1 << 1)
+#define IC_IDMAC_1_CB2_BURST_16         (1 << 2)
+#define IC_IDMAC_1_CB3_BURST_16         (1 << 3)
+#define IC_IDMAC_1_CB4_BURST_16         (1 << 4)
+#define IC_IDMAC_1_CB5_BURST_16         (1 << 5)
+#define IC_IDMAC_1_CB6_BURST_16         (1 << 6)
+#define IC_IDMAC_1_CB7_BURST_16         (1 << 7)
+#define IC_IDMAC_1_PRPENC_ROT_MASK      (0x7 << 11)
+#define IC_IDMAC_1_PRPENC_ROT_OFFSET    11
+#define IC_IDMAC_1_PRPVF_ROT_MASK       (0x7 << 14)
+#define IC_IDMAC_1_PRPVF_ROT_OFFSET     14
+#define IC_IDMAC_1_PP_ROT_MASK          (0x7 << 17)
+#define IC_IDMAC_1_PP_ROT_OFFSET        17
+#define IC_IDMAC_1_PP_FLIP_RS           (1 << 22)
+#define IC_IDMAC_1_PRPVF_FLIP_RS        (1 << 21)
+#define IC_IDMAC_1_PRPENC_FLIP_RS       (1 << 20)
+
+#define IC_IDMAC_2_PRPENC_HEIGHT_MASK   (0x3ff << 0)
+#define IC_IDMAC_2_PRPENC_HEIGHT_OFFSET 0
+#define IC_IDMAC_2_PRPVF_HEIGHT_MASK    (0x3ff << 10)
+#define IC_IDMAC_2_PRPVF_HEIGHT_OFFSET  10
+#define IC_IDMAC_2_PP_HEIGHT_MASK       (0x3ff << 20)
+#define IC_IDMAC_2_PP_HEIGHT_OFFSET     20
+
+#define IC_IDMAC_3_PRPENC_WIDTH_MASK    (0x3ff << 0)
+#define IC_IDMAC_3_PRPENC_WIDTH_OFFSET  0
+#define IC_IDMAC_3_PRPVF_WIDTH_MASK     (0x3ff << 10)
+#define IC_IDMAC_3_PRPVF_WIDTH_OFFSET   10
+#define IC_IDMAC_3_PP_WIDTH_MASK        (0x3ff << 20)
+#define IC_IDMAC_3_PP_WIDTH_OFFSET      20
+
+struct ic_task_regoffs {
+	u32 rsc;
+	u32 tpmem_csc[2];
+};
+
+struct ic_task_bitfields {
+	u32 ic_conf_en;
+	u32 ic_conf_rot_en;
+	u32 ic_conf_cmb_en;
+	u32 ic_conf_csc1_en;
+	u32 ic_conf_csc2_en;
+	u32 ic_cmb_galpha_bit;
+};
+
+static const struct ic_task_regoffs ic_task_reg[IC_NUM_TASKS] = {
+	[IC_TASK_ENCODER] = {
+		.rsc = IC_PRP_ENC_RSC,
+		.tpmem_csc = {0x2008, 0},
+	},
+	[IC_TASK_VIEWFINDER] = {
+		.rsc = IC_PRP_VF_RSC,
+		.tpmem_csc = {0x4028, 0x4040},
+	},
+	[IC_TASK_POST_PROCESSOR] = {
+		.rsc = IC_PP_RSC,
+		.tpmem_csc = {0x6060, 0x6078},
+	},
+};
+
+static const struct ic_task_bitfields ic_task_bit[IC_NUM_TASKS] = {
+	[IC_TASK_ENCODER] = {
+		.ic_conf_en = IC_CONF_PRPENC_EN,
+		.ic_conf_rot_en = IC_CONF_PRPENC_ROT_EN,
+		.ic_conf_cmb_en = 0,    /* NA */
+		.ic_conf_csc1_en = IC_CONF_PRPENC_CSC1,
+		.ic_conf_csc2_en = 0,   /* NA */
+		.ic_cmb_galpha_bit = 0, /* NA */
+	},
+	[IC_TASK_VIEWFINDER] = {
+		.ic_conf_en = IC_CONF_PRPVF_EN,
+		.ic_conf_rot_en = IC_CONF_PRPVF_ROT_EN,
+		.ic_conf_cmb_en = IC_CONF_PRPVF_CMB,
+		.ic_conf_csc1_en = IC_CONF_PRPVF_CSC1,
+		.ic_conf_csc2_en = IC_CONF_PRPVF_CSC2,
+		.ic_cmb_galpha_bit = 0,
+	},
+	[IC_TASK_POST_PROCESSOR] = {
+		.ic_conf_en = IC_CONF_PP_EN,
+		.ic_conf_rot_en = IC_CONF_PP_ROT_EN,
+		.ic_conf_cmb_en = IC_CONF_PP_CMB,
+		.ic_conf_csc1_en = IC_CONF_PP_CSC1,
+		.ic_conf_csc2_en = IC_CONF_PP_CSC2,
+		.ic_cmb_galpha_bit = 8,
+	},
+};
+
+struct ipu_ic_priv;
+
+struct ipu_ic {
+	enum ipu_ic_task task;
+	const struct ic_task_regoffs *reg;
+	const struct ic_task_bitfields *bit;
+
+	enum ipu_color_space in_cs, g_in_cs;
+	enum ipu_color_space out_cs;
+	bool graphics;
+	bool rotation;
+	bool in_use;
+
+	struct ipu_ic_priv *priv;
+};
+
+struct ipu_ic_priv {
+	void __iomem *base;
+	void __iomem *tpmem_base;
+	spinlock_t lock;
+	struct ipu_soc *ipu;
+	int use_count;
+	struct ipu_ic task[IC_NUM_TASKS];
+};
+
+static inline u32 ipu_ic_read(struct ipu_ic *ic, unsigned offset)
+{
+	return readl(ic->priv->base + offset);
+}
+
+static inline void ipu_ic_write(struct ipu_ic *ic, u32 value, unsigned offset)
+{
+	writel(value, ic->priv->base + offset);
+}
+
+struct ic_csc_params {
+	s16 coeff[3][3];	/* signed 9-bit integer coefficients */
+	s16 offset[3];		/* signed 11+2-bit fixed point offset */
+	u8 scale:2;		/* scale coefficients * 2^(scale-1) */
+	bool sat:1;		/* saturate to (16, 235(Y) / 240(U, V)) */
+};
+
+/*
+ * Y = R *  .299 + G *  .587 + B *  .114;
+ * U = R * -.169 + G * -.332 + B *  .500 + 128.;
+ * V = R *  .500 + G * -.419 + B * -.0813 + 128.;
+ */
+static const struct ic_csc_params ic_csc_rgb2ycbcr = {
+	.coeff = {
+		{ 77, 150, 29 },
+		{ 469, 427, 128 },
+		{ 128, 405, 491 },
+	},
+	.offset = { 0, 512, 512 },
+	.scale = 1,
+};
+
+/* transparent RGB->RGB matrix for graphics combining */
+static const struct ic_csc_params ic_csc_rgb2rgb = {
+	.coeff = {
+		{ 128, 0, 0 },
+		{ 0, 128, 0 },
+		{ 0, 0, 128 },
+	},
+	.scale = 2,
+};
+
+/*
+ * R = (1.164 * (Y - 16)) + (1.596 * (Cr - 128));
+ * G = (1.164 * (Y - 16)) - (0.392 * (Cb - 128)) - (0.813 * (Cr - 128));
+ * B = (1.164 * (Y - 16)) + (2.017 * (Cb - 128);
+ */
+static const struct ic_csc_params ic_csc_ycbcr2rgb = {
+	.coeff = {
+		{ 149, 0, 204 },
+		{ 149, 462, 408 },
+		{ 149, 255, 0 },
+	},
+	.offset = { -446, 266, -554 },
+	.scale = 2,
+};
+
+static int init_csc(struct ipu_ic *ic,
+		    enum ipu_color_space inf,
+		    enum ipu_color_space outf,
+		    int csc_index)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	const struct ic_csc_params *params;
+	u32 __iomem *base;
+	const u16 (*c)[3];
+	const u16 *a;
+	u32 param;
+
+	base = (u32 __iomem *)
+		(priv->tpmem_base + ic->reg->tpmem_csc[csc_index]);
+
+	if (inf == IPUV3_COLORSPACE_YUV && outf == IPUV3_COLORSPACE_RGB)
+		params = &ic_csc_ycbcr2rgb;
+	else if (inf == IPUV3_COLORSPACE_RGB && outf == IPUV3_COLORSPACE_YUV)
+		params = &ic_csc_rgb2ycbcr;
+	else if (inf == IPUV3_COLORSPACE_RGB && outf == IPUV3_COLORSPACE_RGB)
+		params = &ic_csc_rgb2rgb;
+	else {
+		dev_err(priv->ipu->dev, "Unsupported color space conversion\n");
+		return -EINVAL;
+	}
+
+	/* Cast to unsigned */
+	c = (const u16 (*)[3])params->coeff;
+	a = (const u16 *)params->offset;
+
+	param = ((a[0] & 0x1f) << 27) | ((c[0][0] & 0x1ff) << 18) |
+		((c[1][1] & 0x1ff) << 9) | (c[2][2] & 0x1ff);
+	writel(param, base++);
+
+	param = ((a[0] & 0x1fe0) >> 5) | (params->scale << 8) |
+		(params->sat << 9);
+	writel(param, base++);
+
+	param = ((a[1] & 0x1f) << 27) | ((c[0][1] & 0x1ff) << 18) |
+		((c[1][0] & 0x1ff) << 9) | (c[2][0] & 0x1ff);
+	writel(param, base++);
+
+	param = ((a[1] & 0x1fe0) >> 5);
+	writel(param, base++);
+
+	param = ((a[2] & 0x1f) << 27) | ((c[0][2] & 0x1ff) << 18) |
+		((c[1][2] & 0x1ff) << 9) | (c[2][1] & 0x1ff);
+	writel(param, base++);
+
+	param = ((a[2] & 0x1fe0) >> 5);
+	writel(param, base++);
+
+	return 0;
+}
+
+static int calc_resize_coeffs(struct ipu_ic *ic,
+			      u32 in_size, u32 out_size,
+			      u32 *resize_coeff,
+			      u32 *downsize_coeff)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	struct ipu_soc *ipu = priv->ipu;
+	u32 temp_size, temp_downsize;
+
+	/*
+	 * Input size cannot be more than 4096, and output size cannot
+	 * be more than 1024
+	 */
+	if (in_size > 4096) {
+		dev_err(ipu->dev, "Unsupported resize (in_size > 4096)\n");
+		return -EINVAL;
+	}
+	if (out_size > 1024) {
+		dev_err(ipu->dev, "Unsupported resize (out_size > 1024)\n");
+		return -EINVAL;
+	}
+
+	/* Cannot downsize more than 8:1 */
+	if ((out_size << 3) < in_size) {
+		dev_err(ipu->dev, "Unsupported downsize\n");
+		return -EINVAL;
+	}
+
+	/* Compute downsizing coefficient */
+	temp_downsize = 0;
+	temp_size = in_size;
+	while (((temp_size > 1024) || (temp_size >= out_size * 2)) &&
+	       (temp_downsize < 2)) {
+		temp_size >>= 1;
+		temp_downsize++;
+	}
+	*downsize_coeff = temp_downsize;
+
+	/*
+	 * compute resizing coefficient using the following equation:
+	 * resize_coeff = M * (SI - 1) / (SO - 1)
+	 * where M = 2^13, SI = input size, SO = output size
+	 */
+	*resize_coeff = (8192L * (temp_size - 1)) / (out_size - 1);
+	if (*resize_coeff >= 16384L) {
+		dev_err(ipu->dev, "Warning! Overflow on resize coeff.\n");
+		*resize_coeff = 0x3FFF;
+	}
+
+	return 0;
+}
+
+void ipu_ic_task_enable(struct ipu_ic *ic)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	unsigned long flags;
+	u32 ic_conf;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	ic_conf = ipu_ic_read(ic, IC_CONF);
+
+	ic_conf |= ic->bit->ic_conf_en;
+
+	if (ic->rotation)
+		ic_conf |= ic->bit->ic_conf_rot_en;
+
+	if (ic->in_cs != ic->out_cs)
+		ic_conf |= ic->bit->ic_conf_csc1_en;
+
+	if (ic->graphics) {
+		ic_conf |= ic->bit->ic_conf_cmb_en;
+		ic_conf |= ic->bit->ic_conf_csc1_en;
+
+		if (ic->g_in_cs != ic->out_cs)
+			ic_conf |= ic->bit->ic_conf_csc2_en;
+	}
+
+	ipu_ic_write(ic, ic_conf, IC_CONF);
+
+	spin_unlock_irqrestore(&priv->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_ic_task_enable);
+
+void ipu_ic_task_disable(struct ipu_ic *ic)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	unsigned long flags;
+	u32 ic_conf;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	ic_conf = ipu_ic_read(ic, IC_CONF);
+
+	ic_conf &= ~(ic->bit->ic_conf_en |
+		     ic->bit->ic_conf_csc1_en |
+		     ic->bit->ic_conf_rot_en);
+	if (ic->bit->ic_conf_csc2_en)
+		ic_conf &= ~ic->bit->ic_conf_csc2_en;
+	if (ic->bit->ic_conf_cmb_en)
+		ic_conf &= ~ic->bit->ic_conf_cmb_en;
+
+	ipu_ic_write(ic, ic_conf, IC_CONF);
+
+	ic->rotation = ic->graphics = false;
+
+	spin_unlock_irqrestore(&priv->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_ic_task_disable);
+
+int ipu_ic_task_graphics_init(struct ipu_ic *ic,
+			      enum ipu_color_space in_g_cs,
+			      bool galpha_en, u32 galpha,
+			      bool colorkey_en, u32 colorkey)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	unsigned long flags;
+	u32 reg, ic_conf;
+	int ret = 0;
+
+	if (ic->task == IC_TASK_ENCODER)
+		return -EINVAL;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	ic_conf = ipu_ic_read(ic, IC_CONF);
+
+	if (!(ic_conf & ic->bit->ic_conf_csc1_en)) {
+		/* need transparent CSC1 conversion */
+		ret = init_csc(ic, IPUV3_COLORSPACE_RGB,
+			       IPUV3_COLORSPACE_RGB, 0);
+		if (ret)
+			goto unlock;
+	}
+
+	ic->g_in_cs = in_g_cs;
+
+	if (ic->g_in_cs != ic->out_cs) {
+		ret = init_csc(ic, ic->g_in_cs, ic->out_cs, 1);
+		if (ret)
+			goto unlock;
+	}
+
+	if (galpha_en) {
+		ic_conf |= IC_CONF_IC_GLB_LOC_A;
+		reg = ipu_ic_read(ic, IC_CMBP_1);
+		reg &= ~(0xff << ic->bit->ic_cmb_galpha_bit);
+		reg |= (galpha << ic->bit->ic_cmb_galpha_bit);
+		ipu_ic_write(ic, reg, IC_CMBP_1);
+	} else
+		ic_conf &= ~IC_CONF_IC_GLB_LOC_A;
+
+	if (colorkey_en) {
+		ic_conf |= IC_CONF_KEY_COLOR_EN;
+		ipu_ic_write(ic, colorkey, IC_CMBP_2);
+	} else
+		ic_conf &= ~IC_CONF_KEY_COLOR_EN;
+
+	ipu_ic_write(ic, ic_conf, IC_CONF);
+
+	ic->graphics = true;
+unlock:
+	spin_unlock_irqrestore(&priv->lock, flags);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(ipu_ic_task_graphics_init);
+
+int ipu_ic_task_init(struct ipu_ic *ic,
+		     int in_width, int in_height,
+		     int out_width, int out_height,
+		     enum ipu_color_space in_cs,
+		     enum ipu_color_space out_cs)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	u32 reg, downsize_coeff, resize_coeff;
+	unsigned long flags;
+	int ret = 0;
+
+	/* Setup vertical resizing */
+	ret = calc_resize_coeffs(ic, in_height, out_height,
+				 &resize_coeff, &downsize_coeff);
+	if (ret)
+		return ret;
+
+	reg = (downsize_coeff << 30) | (resize_coeff << 16);
+
+	/* Setup horizontal resizing */
+	ret = calc_resize_coeffs(ic, in_width, out_width,
+				 &resize_coeff, &downsize_coeff);
+	if (ret)
+		return ret;
+
+	reg |= (downsize_coeff << 14) | resize_coeff;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	ipu_ic_write(ic, reg, ic->reg->rsc);
+
+	/* Setup color space conversion */
+	ic->in_cs = in_cs;
+	ic->out_cs = out_cs;
+
+	if (ic->in_cs != ic->out_cs) {
+		ret = init_csc(ic, ic->in_cs, ic->out_cs, 0);
+		if (ret)
+			goto unlock;
+	}
+
+unlock:
+	spin_unlock_irqrestore(&priv->lock, flags);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(ipu_ic_task_init);
+
+int ipu_ic_task_idma_init(struct ipu_ic *ic, struct ipuv3_channel *channel,
+			  u32 width, u32 height, int burst_size,
+			  enum ipu_rotate_mode rot)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	struct ipu_soc *ipu = priv->ipu;
+	u32 ic_idmac_1, ic_idmac_2, ic_idmac_3;
+	u32 temp_rot = bitrev8(rot) >> 5;
+	bool need_hor_flip = false;
+	unsigned long flags;
+	int ret = 0;
+
+	if ((burst_size != 8) && (burst_size != 16)) {
+		dev_err(ipu->dev, "Illegal burst length for IC\n");
+		return -EINVAL;
+	}
+
+	width--;
+	height--;
+
+	if (temp_rot & 0x2)	/* Need horizontal flip */
+		need_hor_flip = true;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	ic_idmac_1 = ipu_ic_read(ic, IC_IDMAC_1);
+	ic_idmac_2 = ipu_ic_read(ic, IC_IDMAC_2);
+	ic_idmac_3 = ipu_ic_read(ic, IC_IDMAC_3);
+
+	switch (channel->num) {
+	case IPUV3_CHANNEL_IC_PP_MEM:
+		if (burst_size == 16)
+			ic_idmac_1 |= IC_IDMAC_1_CB2_BURST_16;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_CB2_BURST_16;
+
+		if (need_hor_flip)
+			ic_idmac_1 |= IC_IDMAC_1_PP_FLIP_RS;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_PP_FLIP_RS;
+
+		ic_idmac_2 &= ~IC_IDMAC_2_PP_HEIGHT_MASK;
+		ic_idmac_2 |= height << IC_IDMAC_2_PP_HEIGHT_OFFSET;
+
+		ic_idmac_3 &= ~IC_IDMAC_3_PP_WIDTH_MASK;
+		ic_idmac_3 |= width << IC_IDMAC_3_PP_WIDTH_OFFSET;
+		break;
+	case IPUV3_CHANNEL_MEM_IC_PP:
+		if (burst_size == 16)
+			ic_idmac_1 |= IC_IDMAC_1_CB5_BURST_16;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_CB5_BURST_16;
+		break;
+	case IPUV3_CHANNEL_MEM_ROT_PP:
+		ic_idmac_1 &= ~IC_IDMAC_1_PP_ROT_MASK;
+		ic_idmac_1 |= temp_rot << IC_IDMAC_1_PP_ROT_OFFSET;
+		break;
+	case IPUV3_CHANNEL_MEM_IC_PRP_VF:
+		if (burst_size == 16)
+			ic_idmac_1 |= IC_IDMAC_1_CB6_BURST_16;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_CB6_BURST_16;
+		break;
+	case IPUV3_CHANNEL_IC_PRP_ENC_MEM:
+		if (burst_size == 16)
+			ic_idmac_1 |= IC_IDMAC_1_CB0_BURST_16;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_CB0_BURST_16;
+
+		if (need_hor_flip)
+			ic_idmac_1 |= IC_IDMAC_1_PRPENC_FLIP_RS;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_PRPENC_FLIP_RS;
+
+		ic_idmac_2 &= ~IC_IDMAC_2_PRPENC_HEIGHT_MASK;
+		ic_idmac_2 |= height << IC_IDMAC_2_PRPENC_HEIGHT_OFFSET;
+
+		ic_idmac_3 &= ~IC_IDMAC_3_PRPENC_WIDTH_MASK;
+		ic_idmac_3 |= width << IC_IDMAC_3_PRPENC_WIDTH_OFFSET;
+		break;
+	case IPUV3_CHANNEL_MEM_ROT_ENC:
+		ic_idmac_1 &= ~IC_IDMAC_1_PRPENC_ROT_MASK;
+		ic_idmac_1 |= temp_rot << IC_IDMAC_1_PRPENC_ROT_OFFSET;
+		break;
+	case IPUV3_CHANNEL_IC_PRP_VF_MEM:
+		if (burst_size == 16)
+			ic_idmac_1 |= IC_IDMAC_1_CB1_BURST_16;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_CB1_BURST_16;
+
+		if (need_hor_flip)
+			ic_idmac_1 |= IC_IDMAC_1_PRPVF_FLIP_RS;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_PRPVF_FLIP_RS;
+
+		ic_idmac_2 &= ~IC_IDMAC_2_PRPVF_HEIGHT_MASK;
+		ic_idmac_2 |= height << IC_IDMAC_2_PRPVF_HEIGHT_OFFSET;
+
+		ic_idmac_3 &= ~IC_IDMAC_3_PRPVF_WIDTH_MASK;
+		ic_idmac_3 |= width << IC_IDMAC_3_PRPVF_WIDTH_OFFSET;
+		break;
+	case IPUV3_CHANNEL_MEM_ROT_VF:
+		ic_idmac_1 &= ~IC_IDMAC_1_PRPVF_ROT_MASK;
+		ic_idmac_1 |= temp_rot << IC_IDMAC_1_PRPVF_ROT_OFFSET;
+		break;
+	case IPUV3_CHANNEL_G_MEM_IC_PRP_VF:
+		if (burst_size == 16)
+			ic_idmac_1 |= IC_IDMAC_1_CB3_BURST_16;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_CB3_BURST_16;
+		break;
+	case IPUV3_CHANNEL_G_MEM_IC_PP:
+		if (burst_size == 16)
+			ic_idmac_1 |= IC_IDMAC_1_CB4_BURST_16;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_CB4_BURST_16;
+		break;
+	case IPUV3_CHANNEL_VDI_MEM_IC_VF:
+		if (burst_size == 16)
+			ic_idmac_1 |= IC_IDMAC_1_CB7_BURST_16;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_CB7_BURST_16;
+		break;
+	default:
+		goto unlock;
+	}
+
+	ipu_ic_write(ic, ic_idmac_1, IC_IDMAC_1);
+	ipu_ic_write(ic, ic_idmac_2, IC_IDMAC_2);
+	ipu_ic_write(ic, ic_idmac_3, IC_IDMAC_3);
+
+	if (rot >= IPU_ROTATE_90_RIGHT)
+		ic->rotation = true;
+
+unlock:
+	spin_unlock_irqrestore(&priv->lock, flags);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(ipu_ic_task_idma_init);
+
+int ipu_ic_enable(struct ipu_ic *ic)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	unsigned long flags;
+	u32 module = IPU_CONF_IC_EN;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	if (ic->rotation)
+		module |= IPU_CONF_ROT_EN;
+
+	if (!priv->use_count)
+		ipu_module_enable(priv->ipu, module);
+
+	priv->use_count++;
+
+	spin_unlock_irqrestore(&priv->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_ic_enable);
+
+int ipu_ic_disable(struct ipu_ic *ic)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	unsigned long flags;
+	u32 module = IPU_CONF_IC_EN | IPU_CONF_ROT_EN;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	priv->use_count--;
+
+	if (!priv->use_count)
+		ipu_module_disable(priv->ipu, module);
+
+	if (priv->use_count < 0)
+		priv->use_count = 0;
+
+	spin_unlock_irqrestore(&priv->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_ic_disable);
+
+struct ipu_ic *ipu_ic_get(struct ipu_soc *ipu, enum ipu_ic_task task)
+{
+	struct ipu_ic_priv *priv = ipu->ic_priv;
+	unsigned long flags;
+	struct ipu_ic *ic, *ret;
+
+	if (task >= IC_NUM_TASKS)
+		return ERR_PTR(-EINVAL);
+
+	ic = &priv->task[task];
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	if (ic->in_use) {
+		ret = ERR_PTR(-EBUSY);
+		goto unlock;
+	}
+
+	ic->in_use = true;
+	ret = ic;
+
+unlock:
+	spin_unlock_irqrestore(&priv->lock, flags);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(ipu_ic_get);
+
+void ipu_ic_put(struct ipu_ic *ic)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	unsigned long flags;
+
+	spin_lock_irqsave(&priv->lock, flags);
+	ic->in_use = false;
+	spin_unlock_irqrestore(&priv->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_ic_put);
+
+int ipu_ic_init(struct ipu_soc *ipu, struct device *dev,
+		unsigned long base, unsigned long tpmem_base)
+{
+	struct ipu_ic_priv *priv;
+	int i;
+
+	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	ipu->ic_priv = priv;
+
+	spin_lock_init(&priv->lock);
+	priv->base = devm_ioremap(dev, base, PAGE_SIZE);
+	if (!priv->base)
+		return -ENOMEM;
+	priv->tpmem_base = devm_ioremap(dev, tpmem_base, SZ_64K);
+	if (!priv->tpmem_base)
+		return -ENOMEM;
+
+	dev_dbg(dev, "IC base: 0x%08lx remapped to %p\n", base, priv->base);
+
+	priv->ipu = ipu;
+
+	for (i = 0; i < IC_NUM_TASKS; i++) {
+		priv->task[i].task = i;
+		priv->task[i].priv = priv;
+		priv->task[i].reg = &ic_task_reg[i];
+		priv->task[i].bit = &ic_task_bit[i];
+	}
+
+	return 0;
+}
+
+void ipu_ic_exit(struct ipu_soc *ipu)
+{
+}
+
+void ipu_ic_dump(struct ipu_ic *ic)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	struct ipu_soc *ipu = priv->ipu;
+
+	dev_dbg(ipu->dev, "IC_CONF = \t0x%08X\n",
+		ipu_ic_read(ic, IC_CONF));
+	dev_dbg(ipu->dev, "IC_PRP_ENC_RSC = \t0x%08X\n",
+		ipu_ic_read(ic, IC_PRP_ENC_RSC));
+	dev_dbg(ipu->dev, "IC_PRP_VF_RSC = \t0x%08X\n",
+		ipu_ic_read(ic, IC_PRP_VF_RSC));
+	dev_dbg(ipu->dev, "IC_PP_RSC = \t0x%08X\n",
+		ipu_ic_read(ic, IC_PP_RSC));
+	dev_dbg(ipu->dev, "IC_CMBP_1 = \t0x%08X\n",
+		ipu_ic_read(ic, IC_CMBP_1));
+	dev_dbg(ipu->dev, "IC_CMBP_2 = \t0x%08X\n",
+		ipu_ic_read(ic, IC_CMBP_2));
+	dev_dbg(ipu->dev, "IC_IDMAC_1 = \t0x%08X\n",
+		ipu_ic_read(ic, IC_IDMAC_1));
+	dev_dbg(ipu->dev, "IC_IDMAC_2 = \t0x%08X\n",
+		ipu_ic_read(ic, IC_IDMAC_2));
+	dev_dbg(ipu->dev, "IC_IDMAC_3 = \t0x%08X\n",
+		ipu_ic_read(ic, IC_IDMAC_3));
+	dev_dbg(ipu->dev, "IC_IDMAC_4 = \t0x%08X\n",
+		ipu_ic_read(ic, IC_IDMAC_4));
+}
+EXPORT_SYMBOL_GPL(ipu_ic_dump);
diff --git a/drivers/gpu/ipu-v3/ipu-prv.h b/drivers/gpu/ipu-v3/ipu-prv.h
index c93f50ec04f7..bfb1e8a4483f 100644
--- a/drivers/gpu/ipu-v3/ipu-prv.h
+++ b/drivers/gpu/ipu-v3/ipu-prv.h
@@ -24,23 +24,6 @@ struct ipu_soc;
 
 #include <video/imx-ipu-v3.h>
 
-#define IPUV3_CHANNEL_CSI0			 0
-#define IPUV3_CHANNEL_CSI1			 1
-#define IPUV3_CHANNEL_CSI2			 2
-#define IPUV3_CHANNEL_CSI3			 3
-#define IPUV3_CHANNEL_MEM_BG_SYNC		23
-#define IPUV3_CHANNEL_MEM_FG_SYNC		27
-#define IPUV3_CHANNEL_MEM_DC_SYNC		28
-#define IPUV3_CHANNEL_MEM_FG_SYNC_ALPHA		31
-#define IPUV3_CHANNEL_MEM_DC_ASYNC		41
-#define IPUV3_CHANNEL_ROT_ENC_MEM		45
-#define IPUV3_CHANNEL_ROT_VF_MEM		46
-#define IPUV3_CHANNEL_ROT_PP_MEM		47
-#define IPUV3_CHANNEL_ROT_ENC_MEM_OUT		48
-#define IPUV3_CHANNEL_ROT_VF_MEM_OUT		49
-#define IPUV3_CHANNEL_ROT_PP_MEM_OUT		50
-#define IPUV3_CHANNEL_MEM_BG_SYNC_ALPHA		51
-
 #define IPU_MCU_T_DEFAULT	8
 #define IPU_CM_IDMAC_REG_OFS	0x00008000
 #define IPU_CM_IC_REG_OFS	0x00020000
@@ -85,6 +68,7 @@ struct ipu_soc;
 #define IPU_DISP_TASK_STAT		IPU_CM_REG(0x0254)
 #define IPU_CHA_BUF0_RDY(ch)		IPU_CM_REG(0x0268 + 4 * ((ch) / 32))
 #define IPU_CHA_BUF1_RDY(ch)		IPU_CM_REG(0x0270 + 4 * ((ch) / 32))
+#define IPU_CHA_BUF2_RDY(ch)		IPU_CM_REG(0x0288 + 4 * ((ch) / 32))
 #define IPU_ALT_CHA_BUF0_RDY(ch)	IPU_CM_REG(0x0278 + 4 * ((ch) / 32))
 #define IPU_ALT_CHA_BUF1_RDY(ch)	IPU_CM_REG(0x0280 + 4 * ((ch) / 32))
 
@@ -148,9 +132,12 @@ struct ipuv3_channel {
 	struct ipu_soc *ipu;
 };
 
+struct ipu_cpmem;
+struct ipu_csi;
 struct ipu_dc_priv;
 struct ipu_dmfc_priv;
 struct ipu_di;
+struct ipu_ic_priv;
 struct ipu_smfc_priv;
 
 struct ipu_devtype;
@@ -164,7 +151,6 @@ struct ipu_soc {
 
 	void __iomem		*cm_reg;
 	void __iomem		*idmac_reg;
-	struct ipu_ch_param __iomem	*cpmem_base;
 
 	int			usecount;
 
@@ -176,13 +162,27 @@ struct ipu_soc {
 	int			irq_err;
 	struct irq_domain	*domain;
 
+	struct ipu_cpmem	*cpmem_priv;
 	struct ipu_dc_priv	*dc_priv;
 	struct ipu_dp_priv	*dp_priv;
 	struct ipu_dmfc_priv	*dmfc_priv;
 	struct ipu_di		*di_priv[2];
+	struct ipu_csi		*csi_priv[2];
+	struct ipu_ic_priv	*ic_priv;
 	struct ipu_smfc_priv	*smfc_priv;
 };
 
+static inline u32 ipu_idmac_read(struct ipu_soc *ipu, unsigned offset)
+{
+	return readl(ipu->idmac_reg + offset);
+}
+
+static inline void ipu_idmac_write(struct ipu_soc *ipu, u32 value,
+				   unsigned offset)
+{
+	writel(value, ipu->idmac_reg + offset);
+}
+
 void ipu_srm_dp_sync_update(struct ipu_soc *ipu);
 
 int ipu_module_enable(struct ipu_soc *ipu, u32 mask);
@@ -191,6 +191,14 @@ int ipu_module_disable(struct ipu_soc *ipu, u32 mask);
 bool ipu_idmac_channel_busy(struct ipu_soc *ipu, unsigned int chno);
 int ipu_wait_interrupt(struct ipu_soc *ipu, int irq, int ms);
 
+int ipu_csi_init(struct ipu_soc *ipu, struct device *dev, int id,
+		 unsigned long base, u32 module, struct clk *clk_ipu);
+void ipu_csi_exit(struct ipu_soc *ipu, int id);
+
+int ipu_ic_init(struct ipu_soc *ipu, struct device *dev,
+		unsigned long base, unsigned long tpmem_base);
+void ipu_ic_exit(struct ipu_soc *ipu);
+
 int ipu_di_init(struct ipu_soc *ipu, struct device *dev, int id,
 		unsigned long base, u32 module, struct clk *ipu_clk);
 void ipu_di_exit(struct ipu_soc *ipu, int id);
diff --git a/drivers/gpu/ipu-v3/ipu-smfc.c b/drivers/gpu/ipu-v3/ipu-smfc.c
index e4f85ad286fc..4ef910991413 100644
--- a/drivers/gpu/ipu-v3/ipu-smfc.c
+++ b/drivers/gpu/ipu-v3/ipu-smfc.c
@@ -8,7 +8,6 @@
  * http://www.opensource.org/licenses/gpl-license.html
  * http://www.gnu.org/copyleft/gpl.html
  */
-#define DEBUG
 #include <linux/export.h>
 #include <linux/types.h>
 #include <linux/init.h>
@@ -21,9 +20,18 @@
 
 #include "ipu-prv.h"
 
+struct ipu_smfc {
+	struct ipu_smfc_priv *priv;
+	int chno;
+	bool inuse;
+};
+
 struct ipu_smfc_priv {
 	void __iomem *base;
 	spinlock_t lock;
+	struct ipu_soc *ipu;
+	struct ipu_smfc channel[4];
+	int use_count;
 };
 
 /*SMFC Registers */
@@ -31,63 +39,166 @@ struct ipu_smfc_priv {
 #define SMFC_WMC	0x0004
 #define SMFC_BS		0x0008
 
-int ipu_smfc_set_burstsize(struct ipu_soc *ipu, int channel, int burstsize)
+int ipu_smfc_set_burstsize(struct ipu_smfc *smfc, int burstsize)
 {
-	struct ipu_smfc_priv *smfc = ipu->smfc_priv;
+	struct ipu_smfc_priv *priv = smfc->priv;
 	unsigned long flags;
 	u32 val, shift;
 
-	spin_lock_irqsave(&smfc->lock, flags);
+	spin_lock_irqsave(&priv->lock, flags);
 
-	shift = channel * 4;
-	val = readl(smfc->base + SMFC_BS);
+	shift = smfc->chno * 4;
+	val = readl(priv->base + SMFC_BS);
 	val &= ~(0xf << shift);
 	val |= burstsize << shift;
-	writel(val, smfc->base + SMFC_BS);
+	writel(val, priv->base + SMFC_BS);
 
-	spin_unlock_irqrestore(&smfc->lock, flags);
+	spin_unlock_irqrestore(&priv->lock, flags);
 
 	return 0;
 }
 EXPORT_SYMBOL_GPL(ipu_smfc_set_burstsize);
 
-int ipu_smfc_map_channel(struct ipu_soc *ipu, int channel, int csi_id, int mipi_id)
+int ipu_smfc_map_channel(struct ipu_smfc *smfc, int csi_id, int mipi_id)
 {
-	struct ipu_smfc_priv *smfc = ipu->smfc_priv;
+	struct ipu_smfc_priv *priv = smfc->priv;
 	unsigned long flags;
 	u32 val, shift;
 
-	spin_lock_irqsave(&smfc->lock, flags);
+	spin_lock_irqsave(&priv->lock, flags);
 
-	shift = channel * 3;
-	val = readl(smfc->base + SMFC_MAP);
+	shift = smfc->chno * 3;
+	val = readl(priv->base + SMFC_MAP);
 	val &= ~(0x7 << shift);
 	val |= ((csi_id << 2) | mipi_id) << shift;
-	writel(val, smfc->base + SMFC_MAP);
+	writel(val, priv->base + SMFC_MAP);
 
-	spin_unlock_irqrestore(&smfc->lock, flags);
+	spin_unlock_irqrestore(&priv->lock, flags);
 
 	return 0;
 }
 EXPORT_SYMBOL_GPL(ipu_smfc_map_channel);
 
+int ipu_smfc_set_watermark(struct ipu_smfc *smfc, u32 set_level, u32 clr_level)
+{
+	struct ipu_smfc_priv *priv = smfc->priv;
+	unsigned long flags;
+	u32 val, shift;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	shift = smfc->chno * 6 + (smfc->chno > 1 ? 4 : 0);
+	val = readl(priv->base + SMFC_WMC);
+	val &= ~(0x3f << shift);
+	val |= ((clr_level << 3) | set_level) << shift;
+	writel(val, priv->base + SMFC_WMC);
+
+	spin_unlock_irqrestore(&priv->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_smfc_set_watermark);
+
+int ipu_smfc_enable(struct ipu_smfc *smfc)
+{
+	struct ipu_smfc_priv *priv = smfc->priv;
+	unsigned long flags;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	if (!priv->use_count)
+		ipu_module_enable(priv->ipu, IPU_CONF_SMFC_EN);
+
+	priv->use_count++;
+
+	spin_unlock_irqrestore(&priv->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_smfc_enable);
+
+int ipu_smfc_disable(struct ipu_smfc *smfc)
+{
+	struct ipu_smfc_priv *priv = smfc->priv;
+	unsigned long flags;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	priv->use_count--;
+
+	if (!priv->use_count)
+		ipu_module_disable(priv->ipu, IPU_CONF_SMFC_EN);
+
+	if (priv->use_count < 0)
+		priv->use_count = 0;
+
+	spin_unlock_irqrestore(&priv->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_smfc_disable);
+
+struct ipu_smfc *ipu_smfc_get(struct ipu_soc *ipu, unsigned int chno)
+{
+	struct ipu_smfc_priv *priv = ipu->smfc_priv;
+	struct ipu_smfc *smfc, *ret;
+	unsigned long flags;
+
+	if (chno >= 4)
+		return ERR_PTR(-EINVAL);
+
+	smfc = &priv->channel[chno];
+	ret = smfc;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	if (smfc->inuse) {
+		ret = ERR_PTR(-EBUSY);
+		goto unlock;
+	}
+
+	smfc->inuse = true;
+unlock:
+	spin_unlock_irqrestore(&priv->lock, flags);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(ipu_smfc_get);
+
+void ipu_smfc_put(struct ipu_smfc *smfc)
+{
+	struct ipu_smfc_priv *priv = smfc->priv;
+	unsigned long flags;
+
+	spin_lock_irqsave(&priv->lock, flags);
+	smfc->inuse = false;
+	spin_unlock_irqrestore(&priv->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_smfc_put);
+
 int ipu_smfc_init(struct ipu_soc *ipu, struct device *dev,
 		  unsigned long base)
 {
-	struct ipu_smfc_priv *smfc;
+	struct ipu_smfc_priv *priv;
+	int i;
 
-	smfc = devm_kzalloc(dev, sizeof(*smfc), GFP_KERNEL);
-	if (!smfc)
+	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
 		return -ENOMEM;
 
-	ipu->smfc_priv = smfc;
-	spin_lock_init(&smfc->lock);
+	ipu->smfc_priv = priv;
+	spin_lock_init(&priv->lock);
+	priv->ipu = ipu;
 
-	smfc->base = devm_ioremap(dev, base, PAGE_SIZE);
-	if (!smfc->base)
+	priv->base = devm_ioremap(dev, base, PAGE_SIZE);
+	if (!priv->base)
 		return -ENOMEM;
 
-	pr_debug("%s: ioremap 0x%08lx -> %p\n", __func__, base, smfc->base);
+	for (i = 0; i < 4; i++) {
+		priv->channel[i].priv = priv;
+		priv->channel[i].chno = i;
+	}
+
+	pr_debug("%s: ioremap 0x%08lx -> %p\n", __func__, base, priv->base);
 
 	return 0;
 }
diff --git a/drivers/gpu/vga/vgaarb.c b/drivers/gpu/vga/vgaarb.c
index 77711623b973..7bcbf863656e 100644
--- a/drivers/gpu/vga/vgaarb.c
+++ b/drivers/gpu/vga/vgaarb.c
@@ -400,7 +400,6 @@ int vga_get(struct pci_dev *pdev, unsigned int rsrc, int interruptible)
 		}
 		schedule();
 		remove_wait_queue(&vga_wait_queue, &wait);
-		set_current_state(TASK_RUNNING);
 	}
 	return rc;
 }
diff --git a/drivers/hid/Kconfig b/drivers/hid/Kconfig
index c18d5d71062d..dfdc26970022 100644
--- a/drivers/hid/Kconfig
+++ b/drivers/hid/Kconfig
@@ -27,7 +27,8 @@ if HID
 
 config HID_BATTERY_STRENGTH
 	bool "Battery level reporting for HID devices"
-	depends on HID && POWER_SUPPLY && HID = POWER_SUPPLY
+	depends on HID
+	select POWER_SUPPLY
 	default n
 	---help---
 	This option adds support of reporting battery strength (for HID devices
@@ -371,6 +372,7 @@ config HID_LOGITECH_DJ
 	tristate "Logitech Unifying receivers full support"
 	depends on HIDRAW
 	depends on HID_LOGITECH
+	select HID_LOGITECH_HIDPP
 	---help---
 	Say Y if you want support for Logitech Unifying receivers and devices.
 	Unifying receivers are capable of pairing up to 6 Logitech compliant
@@ -378,6 +380,17 @@ config HID_LOGITECH_DJ
 	generic USB_HID driver and all incoming events will be multiplexed
 	into a single mouse and a single keyboard device.
 
+config HID_LOGITECH_HIDPP
+	tristate "Logitech HID++ devices support"
+	depends on HID_LOGITECH
+	---help---
+	Support for Logitech devices relyingon the HID++ Logitech specification
+
+	Say Y if you want support for Logitech devices relying on the HID++
+	specification. Such devices are the various Logitech Touchpads (T650,
+	T651, TK820), some mice (Zone Touch mouse), or even keyboards (Solar
+	Keayboard).
+
 config LOGITECH_FF
 	bool "Logitech force feedback support"
 	depends on HID_LOGITECH
@@ -530,6 +543,17 @@ config PANTHERLORD_FF
 	  Say Y here if you have a PantherLord/GreenAsia based game controller
 	  or adapter and want to enable force feedback support for it.
 
+config HID_PENMOUNT
+	tristate "Penmount touch device"
+	depends on USB_HID
+	---help---
+	  This selects a driver for the PenMount 6000 touch controller.
+
+	  The driver works around a problem in the report descript allowing
+	  the userspace to touch events instead of mouse events.
+
+	  Say Y here if you have a Penmount based touch controller.
+
 config HID_PETALYNX
 	tristate "Petalynx Maxter remote control"
 	depends on HID
@@ -602,6 +626,13 @@ config HID_PICOLCD_CIR
 	---help---
 	  Provide access to PicoLCD's CIR interface via remote control (LIRC).
 
+config HID_PLANTRONICS
+	tristate "Plantronics USB HID Driver"
+	default !EXPERT
+	depends on HID
+	---help---
+	Provides HID support for Plantronics telephony devices.
+
 config HID_PRIMAX
 	tristate "Primax non-fully HID-compliant devices"
 	depends on HID
@@ -618,7 +649,7 @@ config HID_ROCCAT
 	support for its special functionalities.
 
 config HID_SAITEK
-	tristate "Saitek non-fully HID-compliant devices"
+	tristate "Saitek (Mad Catz) non-fully HID-compliant devices"
 	depends on HID
 	---help---
 	Support for Saitek devices that are not fully compliant with the
@@ -626,6 +657,7 @@ config HID_SAITEK
 
 	Supported devices:
 	- PS1000 Dual Analog Pad
+	- R.A.T.9 Gaming Mouse
 	- R.A.T.7 Gaming Mouse
 	- M.M.O.7 Gaming Mouse
 
diff --git a/drivers/hid/Makefile b/drivers/hid/Makefile
index 4dbac7f8530c..debd15b44b59 100644
--- a/drivers/hid/Makefile
+++ b/drivers/hid/Makefile
@@ -63,6 +63,7 @@ obj-$(CONFIG_HID_LCPOWER)       += hid-lcpower.o
 obj-$(CONFIG_HID_LENOVO)	+= hid-lenovo.o
 obj-$(CONFIG_HID_LOGITECH)	+= hid-logitech.o
 obj-$(CONFIG_HID_LOGITECH_DJ)	+= hid-logitech-dj.o
+obj-$(CONFIG_HID_LOGITECH_HIDPP)	+= hid-logitech-hidpp.o
 obj-$(CONFIG_HID_MAGICMOUSE)    += hid-magicmouse.o
 obj-$(CONFIG_HID_MICROSOFT)	+= hid-microsoft.o
 obj-$(CONFIG_HID_MONTEREY)	+= hid-monterey.o
@@ -71,6 +72,7 @@ obj-$(CONFIG_HID_NTRIG)		+= hid-ntrig.o
 obj-$(CONFIG_HID_ORTEK)		+= hid-ortek.o
 obj-$(CONFIG_HID_PRODIKEYS)	+= hid-prodikeys.o
 obj-$(CONFIG_HID_PANTHERLORD)	+= hid-pl.o
+obj-$(CONFIG_HID_PENMOUNT)	+= hid-penmount.o
 obj-$(CONFIG_HID_PETALYNX)	+= hid-petalynx.o
 obj-$(CONFIG_HID_PICOLCD)	+= hid-picolcd.o
 hid-picolcd-y			+= hid-picolcd_core.o
@@ -93,6 +95,7 @@ ifdef CONFIG_DEBUG_FS
 hid-picolcd-y			+= hid-picolcd_debugfs.o
 endif
 
+obj-$(CONFIG_HID_PLANTRONICS)	+= hid-plantronics.o
 obj-$(CONFIG_HID_PRIMAX)	+= hid-primax.o
 obj-$(CONFIG_HID_ROCCAT)	+= hid-roccat.o hid-roccat-common.o \
 	hid-roccat-arvo.o hid-roccat-isku.o hid-roccat-kone.o \
diff --git a/drivers/hid/hid-core.c b/drivers/hid/hid-core.c
index 12b6e67d9de0..8b638792cb43 100644
--- a/drivers/hid/hid-core.c
+++ b/drivers/hid/hid-core.c
@@ -52,7 +52,7 @@ EXPORT_SYMBOL_GPL(hid_debug);
 
 static int hid_ignore_special_drivers = 0;
 module_param_named(ignore_special_drivers, hid_ignore_special_drivers, int, 0600);
-MODULE_PARM_DESC(debug, "Ignore any special drivers and handle all devices by generic driver");
+MODULE_PARM_DESC(ignore_special_drivers, "Ignore any special drivers and handle all devices by generic driver");
 
 /*
  * Register a new report for a device.
@@ -702,6 +702,11 @@ static void hid_scan_collection(struct hid_parser *parser, unsigned type)
 	if (((parser->global.usage_page << 16) == HID_UP_SENSOR) &&
 	    type == HID_COLLECTION_PHYSICAL)
 		hid->group = HID_GROUP_SENSOR_HUB;
+
+	if (hid->vendor == USB_VENDOR_ID_MICROSOFT &&
+	    hid->product == USB_DEVICE_ID_MS_TYPE_COVER_3 &&
+	    hid->group == HID_GROUP_MULTITOUCH)
+		hid->group = HID_GROUP_GENERIC;
 }
 
 static int hid_scan_main(struct hid_parser *parser, struct hid_item *item)
@@ -780,22 +785,19 @@ static int hid_scan_report(struct hid_device *hid)
 		hid->group = HID_GROUP_MULTITOUCH_WIN_8;
 
 	/*
-	* Vendor specific handlings
-	*/
-	if ((hid->vendor == USB_VENDOR_ID_SYNAPTICS) &&
-	    (hid->group == HID_GROUP_GENERIC) &&
-	    /* only bind to the mouse interface of composite USB devices */
-	    (hid->bus != BUS_USB || hid->type == HID_TYPE_USBMOUSE))
-		/* hid-rmi should take care of them, not hid-generic */
-		hid->group = HID_GROUP_RMI;
-
-	/*
 	 * Vendor specific handlings
 	 */
 	switch (hid->vendor) {
 	case USB_VENDOR_ID_WACOM:
 		hid->group = HID_GROUP_WACOM;
 		break;
+	case USB_VENDOR_ID_SYNAPTICS:
+		if ((hid->group == HID_GROUP_GENERIC) &&
+		    (hid->bus != BUS_USB || hid->type == HID_TYPE_USBMOUSE))
+			/* hid-rmi should only bind to the mouse interface of
+			 * composite USB devices */
+			hid->group = HID_GROUP_RMI;
+		break;
 	}
 
 	vfree(parser);
@@ -1280,12 +1282,6 @@ void hid_output_report(struct hid_report *report, __u8 *data)
 }
 EXPORT_SYMBOL_GPL(hid_output_report);
 
-static int hid_report_len(struct hid_report *report)
-{
-	/* equivalent to DIV_ROUND_UP(report->size, 8) + !!(report->id > 0) */
-	return ((report->size - 1) >> 3) + 1 + (report->id > 0);
-}
-
 /*
  * Allocator for buffer that is going to be passed to hid_output_report()
  */
@@ -1591,6 +1587,9 @@ int hid_connect(struct hid_device *hdev, unsigned int connect_mask)
 	if ((connect_mask & HID_CONNECT_HIDRAW) && !hidraw_connect(hdev))
 		hdev->claimed |= HID_CLAIMED_HIDRAW;
 
+	if (connect_mask & HID_CONNECT_DRIVER)
+		hdev->claimed |= HID_CLAIMED_DRIVER;
+
 	/* Drivers with the ->raw_event callback set are not required to connect
 	 * to any other listener. */
 	if (!hdev->claimed && !hdev->driver->raw_event) {
@@ -1656,6 +1655,7 @@ void hid_disconnect(struct hid_device *hdev)
 		hdev->hiddev_disconnect(hdev);
 	if (hdev->claimed & HID_CLAIMED_HIDRAW)
 		hidraw_disconnect(hdev);
+	hdev->claimed = 0;
 }
 EXPORT_SYMBOL_GPL(hid_disconnect);
 
@@ -1793,6 +1793,7 @@ static const struct hid_device_id hid_have_special_driver[] = {
 	{ HID_USB_DEVICE(USB_VENDOR_ID_HOLTEK_ALT, USB_DEVICE_ID_HOLTEK_ALT_MOUSE_A070) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_HOLTEK_ALT, USB_DEVICE_ID_HOLTEK_ALT_MOUSE_A072) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_HOLTEK_ALT, USB_DEVICE_ID_HOLTEK_ALT_MOUSE_A081) },
+	{ HID_USB_DEVICE(USB_VENDOR_ID_HOLTEK_ALT, USB_DEVICE_ID_HOLTEK_ALT_MOUSE_A0C2) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_HUION, USB_DEVICE_ID_HUION_TABLET) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_JESS2, USB_DEVICE_ID_JESS2_COLOR_RUMBLE_PAD) },
 	{ HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_ION, USB_DEVICE_ID_ICADE) },
@@ -1804,6 +1805,7 @@ static const struct hid_device_id hid_have_special_driver[] = {
 	{ HID_USB_DEVICE(USB_VENDOR_ID_KYE, USB_DEVICE_ID_KYE_ERGO_525V) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_KYE, USB_DEVICE_ID_KYE_EASYPEN_I405X) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_KYE, USB_DEVICE_ID_KYE_MOUSEPEN_I608X) },
+	{ HID_USB_DEVICE(USB_VENDOR_ID_KYE, USB_DEVICE_ID_KYE_MOUSEPEN_I608X_2) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_KYE, USB_DEVICE_ID_KYE_EASYPEN_M610X) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_LABTEC, USB_DEVICE_ID_LABTEC_WIRELESS_KEYBOARD) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_LCPOWER, USB_DEVICE_ID_LCPOWER_LC1000 ) },
@@ -1817,6 +1819,7 @@ static const struct hid_device_id hid_have_special_driver[] = {
 	{ HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH, USB_DEVICE_ID_S510_RECEIVER_2) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH, USB_DEVICE_ID_LOGITECH_RECEIVER) },
 	{ HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_LOGITECH, USB_DEVICE_ID_LOGITECH_HARMONY_PS3) },
+	{ HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_LOGITECH, USB_DEVICE_ID_LOGITECH_T651) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH, USB_DEVICE_ID_DINOVO_DESKTOP) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH, USB_DEVICE_ID_DINOVO_EDGE) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH, USB_DEVICE_ID_DINOVO_MINI) },
@@ -1857,6 +1860,7 @@ static const struct hid_device_id hid_have_special_driver[] = {
 	{ HID_USB_DEVICE(USB_VENDOR_ID_MICROSOFT, USB_DEVICE_ID_MS_DIGITAL_MEDIA_3K) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_MICROSOFT, USB_DEVICE_ID_WIRELESS_OPTICAL_DESKTOP_3_0) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_MICROSOFT, USB_DEVICE_ID_MS_OFFICE_KB) },
+	{ HID_USB_DEVICE(USB_VENDOR_ID_MICROSOFT, USB_DEVICE_ID_MS_TYPE_COVER_3) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_MONTEREY, USB_DEVICE_ID_GENIUS_KB29E) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_MSI, USB_DEVICE_ID_MSI_GT683R_LED_PANEL) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_NTRIG, USB_DEVICE_ID_NTRIG_TOUCH_SCREEN) },
@@ -1880,7 +1884,9 @@ static const struct hid_device_id hid_have_special_driver[] = {
 	{ HID_USB_DEVICE(USB_VENDOR_ID_NTRIG, USB_DEVICE_ID_NTRIG_TOUCH_SCREEN_18) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_ORTEK, USB_DEVICE_ID_ORTEK_PKB1700) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_ORTEK, USB_DEVICE_ID_ORTEK_WKB2000) },
+	{ HID_USB_DEVICE(USB_VENDOR_ID_PENMOUNT, USB_DEVICE_ID_PENMOUNT_6000) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_PETALYNX, USB_DEVICE_ID_PETALYNX_MAXTER_REMOTE) },
+	{ HID_USB_DEVICE(USB_VENDOR_ID_PLANTRONICS, HID_ANY_ID) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_PRIMAX, USB_DEVICE_ID_PRIMAX_KEYBOARD) },
 #if IS_ENABLED(CONFIG_HID_ROCCAT)
 	{ HID_USB_DEVICE(USB_VENDOR_ID_ROCCAT, USB_DEVICE_ID_ROCCAT_ARVO) },
@@ -1904,10 +1910,12 @@ static const struct hid_device_id hid_have_special_driver[] = {
 	{ HID_USB_DEVICE(USB_VENDOR_ID_SAITEK, USB_DEVICE_ID_SAITEK_PS1000) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_SAITEK, USB_DEVICE_ID_SAITEK_RAT7) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_SAITEK, USB_DEVICE_ID_SAITEK_MMO7) },
+	{ HID_USB_DEVICE(USB_VENDOR_ID_MADCATZ, USB_DEVICE_ID_MADCATZ_RAT9) },
 #endif
 	{ HID_USB_DEVICE(USB_VENDOR_ID_SAMSUNG, USB_DEVICE_ID_SAMSUNG_IR_REMOTE) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_SAMSUNG, USB_DEVICE_ID_SAMSUNG_WIRELESS_KBD_MOUSE) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_SKYCABLE, USB_DEVICE_ID_SKYCABLE_WIRELESS_PRESENTER) },
+	{ HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SMK, USB_DEVICE_ID_SMK_PS3_BDREMOTE) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_SONY, USB_DEVICE_ID_SONY_BUZZ_CONTROLLER) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_SONY, USB_DEVICE_ID_SONY_WIRELESS_BUZZ_CONTROLLER) },
 	{ HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SONY, USB_DEVICE_ID_SONY_PS3_BDREMOTE) },
@@ -2533,7 +2541,8 @@ int hid_add_device(struct hid_device *hdev)
 	 * Scan generic devices for group information
 	 */
 	if (hid_ignore_special_drivers ||
-	    !hid_match_id(hdev, hid_have_special_driver)) {
+	    (!hdev->group &&
+	     !hid_match_id(hdev, hid_have_special_driver))) {
 		ret = hid_scan_report(hdev);
 		if (ret)
 			hid_warn(hdev, "bad device descriptor (%d)\n", ret);
diff --git a/drivers/hid/hid-cp2112.c b/drivers/hid/hid-cp2112.c
index a822db5a8338..3318de690e00 100644
--- a/drivers/hid/hid-cp2112.c
+++ b/drivers/hid/hid-cp2112.c
@@ -1069,8 +1069,7 @@ static int cp2112_probe(struct hid_device *hdev, const struct hid_device_id *id)
 	return ret;
 
 err_gpiochip_remove:
-	if (gpiochip_remove(&dev->gc) < 0)
-		hid_err(hdev, "error removing gpio chip\n");
+	gpiochip_remove(&dev->gc);
 err_free_i2c:
 	i2c_del_adapter(&dev->adap);
 err_free_dev:
@@ -1089,8 +1088,7 @@ static void cp2112_remove(struct hid_device *hdev)
 	struct cp2112_device *dev = hid_get_drvdata(hdev);
 
 	sysfs_remove_group(&hdev->dev.kobj, &cp2112_attr_group);
-	if (gpiochip_remove(&dev->gc))
-		hid_err(hdev, "unable to remove gpio chip\n");
+	gpiochip_remove(&dev->gc);
 	i2c_del_adapter(&dev->adap);
 	/* i2c_del_adapter has finished removing all i2c devices from our
 	 * adapter. Well behaved devices should no longer call our cp2112_xfer
diff --git a/drivers/hid/hid-debug.c b/drivers/hid/hid-debug.c
index 84c3cb15ccdd..8bf61d295ffd 100644
--- a/drivers/hid/hid-debug.c
+++ b/drivers/hid/hid-debug.c
@@ -946,6 +946,12 @@ static const char *keys[KEY_MAX + 1] = {
 	[KEY_BRIGHTNESS_MIN] = "BrightnessMin",
 	[KEY_BRIGHTNESS_MAX] = "BrightnessMax",
 	[KEY_BRIGHTNESS_AUTO] = "BrightnessAuto",
+	[KEY_KBDINPUTASSIST_PREV] = "KbdInputAssistPrev",
+	[KEY_KBDINPUTASSIST_NEXT] = "KbdInputAssistNext",
+	[KEY_KBDINPUTASSIST_PREVGROUP] = "KbdInputAssistPrevGroup",
+	[KEY_KBDINPUTASSIST_NEXTGROUP] = "KbdInputAssistNextGroup",
+	[KEY_KBDINPUTASSIST_ACCEPT] = "KbdInputAssistAccept",
+	[KEY_KBDINPUTASSIST_CANCEL] = "KbdInputAssistCancel",
 };
 
 static const char *relatives[REL_MAX + 1] = {
diff --git a/drivers/hid/hid-holtek-mouse.c b/drivers/hid/hid-holtek-mouse.c
index d60fbd0adc0c..78b3a0c76775 100644
--- a/drivers/hid/hid-holtek-mouse.c
+++ b/drivers/hid/hid-holtek-mouse.c
@@ -29,6 +29,7 @@
  *   and Zalman ZM-GM1
  * - USB ID 04d9:a081, sold as SHARKOON DarkGlider Gaming mouse
  * - USB ID 04d9:a072, sold as LEETGION Hellion Gaming Mouse
+ * - USB ID 04d9:a0c2, sold as ETEKCITY Scroll T-140 Gaming Mouse
  */
 
 static __u8 *holtek_mouse_report_fixup(struct hid_device *hdev, __u8 *rdesc,
@@ -42,6 +43,7 @@ static __u8 *holtek_mouse_report_fixup(struct hid_device *hdev, __u8 *rdesc,
 		switch (hdev->product) {
 		case USB_DEVICE_ID_HOLTEK_ALT_MOUSE_A067:
 		case USB_DEVICE_ID_HOLTEK_ALT_MOUSE_A072:
+		case USB_DEVICE_ID_HOLTEK_ALT_MOUSE_A0C2:
 			if (*rsize >= 122 && rdesc[115] == 0xff && rdesc[116] == 0x7f
 					&& rdesc[120] == 0xff && rdesc[121] == 0x7f) {
 				hid_info(hdev, "Fixing up report descriptor\n");
@@ -74,6 +76,8 @@ static const struct hid_device_id holtek_mouse_devices[] = {
 			USB_DEVICE_ID_HOLTEK_ALT_MOUSE_A072) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_HOLTEK_ALT,
 			USB_DEVICE_ID_HOLTEK_ALT_MOUSE_A081) },
+	{ HID_USB_DEVICE(USB_VENDOR_ID_HOLTEK_ALT,
+			USB_DEVICE_ID_HOLTEK_ALT_MOUSE_A0C2) },
 	{ }
 };
 MODULE_DEVICE_TABLE(hid, holtek_mouse_devices);
diff --git a/drivers/hid/hid-ids.h b/drivers/hid/hid-ids.h
index 25cd674d6064..9243359c1821 100644
--- a/drivers/hid/hid-ids.h
+++ b/drivers/hid/hid-ids.h
@@ -296,6 +296,13 @@
 #define USB_DEVICE_ID_DWAV_EGALAX_MULTITOUCH_73F7	0x73f7
 #define USB_DEVICE_ID_DWAV_EGALAX_MULTITOUCH_A001	0xa001
 
+#define USB_VENDOR_ID_ELAN		0x04f3
+#define USB_DEVICE_ID_ELAN_TOUCHSCREEN	0x0089
+#define USB_DEVICE_ID_ELAN_TOUCHSCREEN_009B	0x009b
+#define USB_DEVICE_ID_ELAN_TOUCHSCREEN_0103	0x0103
+#define USB_DEVICE_ID_ELAN_TOUCHSCREEN_010c	0x010c
+#define USB_DEVICE_ID_ELAN_TOUCHSCREEN_016F	0x016f
+
 #define USB_VENDOR_ID_ELECOM		0x056e
 #define USB_DEVICE_ID_ELECOM_BM084	0x0061
 
@@ -479,6 +486,7 @@
 #define USB_DEVICE_ID_HOLTEK_ALT_MOUSE_A070	0xa070
 #define USB_DEVICE_ID_HOLTEK_ALT_MOUSE_A072	0xa072
 #define USB_DEVICE_ID_HOLTEK_ALT_MOUSE_A081	0xa081
+#define USB_DEVICE_ID_HOLTEK_ALT_MOUSE_A0C2	0xa0c2
 #define USB_DEVICE_ID_HOLTEK_ALT_KEYBOARD_A096	0xa096
 
 #define USB_VENDOR_ID_IMATION		0x0718
@@ -518,6 +526,7 @@
 #define USB_DEVICE_ID_KYE_GPEN_560	0x5003
 #define USB_DEVICE_ID_KYE_EASYPEN_I405X	0x5010
 #define USB_DEVICE_ID_KYE_MOUSEPEN_I608X	0x5011
+#define USB_DEVICE_ID_KYE_MOUSEPEN_I608X_2	0x501a
 #define USB_DEVICE_ID_KYE_EASYPEN_M610X	0x5013
 
 #define USB_VENDOR_ID_LABTEC		0x1020
@@ -571,6 +580,7 @@
 
 #define USB_VENDOR_ID_LOGITECH		0x046d
 #define USB_DEVICE_ID_LOGITECH_AUDIOHUB 0x0a0e
+#define USB_DEVICE_ID_LOGITECH_T651	0xb00c
 #define USB_DEVICE_ID_LOGITECH_RECEIVER	0xc101
 #define USB_DEVICE_ID_LOGITECH_HARMONY_FIRST  0xc110
 #define USB_DEVICE_ID_LOGITECH_HARMONY_LAST 0xc14f
@@ -613,6 +623,7 @@
 
 #define USB_VENDOR_ID_MADCATZ		0x0738
 #define USB_DEVICE_ID_MADCATZ_BEATPAD	0x4540
+#define USB_DEVICE_ID_MADCATZ_RAT9	0x1709
 
 #define USB_VENDOR_ID_MCC		0x09db
 #define USB_DEVICE_ID_MCC_PMD1024LS	0x0076
@@ -642,6 +653,7 @@
 #define USB_DEVICE_ID_MS_SURFACE_PRO_2   0x0799
 #define USB_DEVICE_ID_MS_TOUCH_COVER_2   0x07a7
 #define USB_DEVICE_ID_MS_TYPE_COVER_2    0x07a9
+#define USB_DEVICE_ID_MS_TYPE_COVER_3    0x07dc
 
 #define USB_VENDOR_ID_MOJO		0x8282
 #define USB_DEVICE_ID_RETRO_ADAPTER	0x3201
@@ -709,6 +721,8 @@
 #define USB_DEVICE_ID_ORTEK_PKB1700	0x1700
 #define USB_DEVICE_ID_ORTEK_WKB2000	0x2000
 
+#define USB_VENDOR_ID_PLANTRONICS	0x047f
+
 #define USB_VENDOR_ID_PANASONIC		0x04da
 #define USB_DEVICE_ID_PANABOARD_UBT780	0x1044
 #define USB_DEVICE_ID_PANABOARD_UBT880	0x104d
@@ -722,6 +736,7 @@
 #define USB_DEVICE_ID_PENMOUNT_PCI	0x3500
 #define USB_DEVICE_ID_PENMOUNT_1610	0x1610
 #define USB_DEVICE_ID_PENMOUNT_1640	0x1640
+#define USB_DEVICE_ID_PENMOUNT_6000	0x6000
 
 #define USB_VENDOR_ID_PETALYNX		0x18b1
 #define USB_DEVICE_ID_PETALYNX_MAXTER_REMOTE	0x0037
@@ -733,6 +748,8 @@
 #define USB_DEVICE_ID_PI_ENGINEERING_VEC_USB_FOOTPEDAL	0xff
 
 #define USB_VENDOR_ID_PIXART				0x093a
+#define USB_DEVICE_ID_PIXART_USB_OPTICAL_MOUSE_ID2	0x0137
+#define USB_DEVICE_ID_PIXART_USB_OPTICAL_MOUSE		0x2510
 #define USB_DEVICE_ID_PIXART_OPTICAL_TOUCH_SCREEN	0x8001
 #define USB_DEVICE_ID_PIXART_OPTICAL_TOUCH_SCREEN1	0x8002
 #define USB_DEVICE_ID_PIXART_OPTICAL_TOUCH_SCREEN2	0x8003
@@ -803,6 +820,9 @@
 #define USB_VENDOR_ID_SKYCABLE			0x1223
 #define	USB_DEVICE_ID_SKYCABLE_WIRELESS_PRESENTER	0x3F07
 
+#define USB_VENDOR_ID_SMK		0x0609
+#define USB_DEVICE_ID_SMK_PS3_BDREMOTE	0x0306
+
 #define USB_VENDOR_ID_SONY			0x054c
 #define USB_DEVICE_ID_SONY_VAIO_VGX_MOUSE	0x024b
 #define USB_DEVICE_ID_SONY_VAIO_VGP_MOUSE	0x0374
@@ -921,6 +941,9 @@
 #define USB_DEVICE_ID_VERNIER_CYCLOPS	0x0004
 #define USB_DEVICE_ID_VERNIER_LCSPEC	0x0006
 
+#define USB_VENDOR_ID_VTL		0x0306
+#define USB_DEVICE_ID_VTL_MULTITOUCH_FF3F	0xff3f
+
 #define USB_VENDOR_ID_WACOM		0x056a
 #define USB_DEVICE_ID_WACOM_GRAPHIRE_BLUETOOTH	0x81
 #define USB_DEVICE_ID_WACOM_INTUOS4_BLUETOOTH   0x00BD
diff --git a/drivers/hid/hid-input.c b/drivers/hid/hid-input.c
index 2619f7f4517a..9505605b6e22 100644
--- a/drivers/hid/hid-input.c
+++ b/drivers/hid/hid-input.c
@@ -312,6 +312,9 @@ static const struct hid_device_id hid_battery_quirks[] = {
 			       USB_DEVICE_ID_APPLE_ALU_WIRELESS_2011_ANSI),
 	  HID_BATTERY_QUIRK_PERCENT | HID_BATTERY_QUIRK_FEATURE },
 	{ HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_APPLE,
+			       USB_DEVICE_ID_APPLE_ALU_WIRELESS_2011_ISO),
+	  HID_BATTERY_QUIRK_PERCENT | HID_BATTERY_QUIRK_FEATURE },
+	{ HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_APPLE,
 		USB_DEVICE_ID_APPLE_ALU_WIRELESS_ANSI),
 	  HID_BATTERY_QUIRK_PERCENT | HID_BATTERY_QUIRK_FEATURE },
 	{}
@@ -599,6 +602,12 @@ static void hidinput_configure_usage(struct hid_input *hidinput, struct hid_fiel
 		/* These usage IDs map directly to the usage codes. */
 		case HID_GD_X: case HID_GD_Y: case HID_GD_Z:
 		case HID_GD_RX: case HID_GD_RY: case HID_GD_RZ:
+			if (field->flags & HID_MAIN_ITEM_RELATIVE)
+				map_rel(usage->hid & 0xf);
+			else
+				map_abs_clear(usage->hid & 0xf);
+			break;
+
 		case HID_GD_SLIDER: case HID_GD_DIAL: case HID_GD_WHEEL:
 			if (field->flags & HID_MAIN_ITEM_RELATIVE)
 				map_rel(usage->hid & 0xf);
@@ -689,7 +698,10 @@ static void hidinput_configure_usage(struct hid_input *hidinput, struct hid_fiel
 			break;
 
 		case 0x5b: /* TransducerSerialNumber */
-			set_bit(MSC_SERIAL, input->mscbit);
+			usage->type = EV_MSC;
+			usage->code = MSC_SERIAL;
+			bit = input->mscbit;
+			max = MSC_MAX;
 			break;
 
 		default:  goto unknown;
@@ -856,7 +868,14 @@ static void hidinput_configure_usage(struct hid_input *hidinput, struct hid_fiel
 		case 0x28b: map_key_clear(KEY_FORWARDMAIL);	break;
 		case 0x28c: map_key_clear(KEY_SEND);		break;
 
-		default:    goto ignore;
+		case 0x2c7: map_key_clear(KEY_KBDINPUTASSIST_PREV);		break;
+		case 0x2c8: map_key_clear(KEY_KBDINPUTASSIST_NEXT);		break;
+		case 0x2c9: map_key_clear(KEY_KBDINPUTASSIST_PREVGROUP);		break;
+		case 0x2ca: map_key_clear(KEY_KBDINPUTASSIST_NEXTGROUP);		break;
+		case 0x2cb: map_key_clear(KEY_KBDINPUTASSIST_ACCEPT);	break;
+		case 0x2cc: map_key_clear(KEY_KBDINPUTASSIST_CANCEL);	break;
+
+		default: map_key_clear(KEY_UNKNOWN);
 		}
 		break;
 
@@ -1199,7 +1218,7 @@ static void hidinput_led_worker(struct work_struct *work)
 		return hid->ll_driver->request(hid, report, HID_REQ_SET_REPORT);
 
 	/* fall back to generic raw-output-report */
-	len = ((report->size - 1) >> 3) + 1 + (report->id > 0);
+	len = hid_report_len(report);
 	buf = hid_alloc_report_buf(report, GFP_KERNEL);
 	if (!buf)
 		return;
diff --git a/drivers/hid/hid-kye.c b/drivers/hid/hid-kye.c
index b92bf01a1ae8..158fcf577fae 100644
--- a/drivers/hid/hid-kye.c
+++ b/drivers/hid/hid-kye.c
@@ -323,6 +323,7 @@ static __u8 *kye_report_fixup(struct hid_device *hdev, __u8 *rdesc,
 		}
 		break;
 	case USB_DEVICE_ID_KYE_MOUSEPEN_I608X:
+	case USB_DEVICE_ID_KYE_MOUSEPEN_I608X_2:
 		if (*rsize == MOUSEPEN_I608X_RDESC_ORIG_SIZE) {
 			rdesc = mousepen_i608x_rdesc_fixed;
 			*rsize = sizeof(mousepen_i608x_rdesc_fixed);
@@ -415,6 +416,7 @@ static int kye_probe(struct hid_device *hdev, const struct hid_device_id *id)
 	switch (id->product) {
 	case USB_DEVICE_ID_KYE_EASYPEN_I405X:
 	case USB_DEVICE_ID_KYE_MOUSEPEN_I608X:
+	case USB_DEVICE_ID_KYE_MOUSEPEN_I608X_2:
 	case USB_DEVICE_ID_KYE_EASYPEN_M610X:
 		ret = kye_tablet_enable(hdev);
 		if (ret) {
@@ -446,6 +448,8 @@ static const struct hid_device_id kye_devices[] = {
 	{ HID_USB_DEVICE(USB_VENDOR_ID_KYE,
 				USB_DEVICE_ID_KYE_MOUSEPEN_I608X) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_KYE,
+				USB_DEVICE_ID_KYE_MOUSEPEN_I608X_2) },
+	{ HID_USB_DEVICE(USB_VENDOR_ID_KYE,
 				USB_DEVICE_ID_KYE_EASYPEN_M610X) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_KYE,
 				USB_DEVICE_ID_GENIUS_GILA_GAMING_MOUSE) },
diff --git a/drivers/hid/hid-lenovo.c b/drivers/hid/hid-lenovo.c
index bf227f7679af..4c55f4d95798 100644
--- a/drivers/hid/hid-lenovo.c
+++ b/drivers/hid/hid-lenovo.c
@@ -62,7 +62,6 @@ static int lenovo_input_mapping_cptkbd(struct hid_device *hdev,
 	/* HID_UP_LNVENDOR = USB, HID_UP_MSVENDOR = BT */
 	if ((usage->hid & HID_USAGE_PAGE) == HID_UP_MSVENDOR ||
 	    (usage->hid & HID_USAGE_PAGE) == HID_UP_LNVENDOR) {
-		set_bit(EV_REP, hi->input->evbit);
 		switch (usage->hid & HID_USAGE) {
 		case 0x00f1: /* Fn-F4: Mic mute */
 			map_key_clear(KEY_MICMUTE);
@@ -85,13 +84,13 @@ static int lenovo_input_mapping_cptkbd(struct hid_device *hdev,
 		case 0x00f8: /* Fn-F11: View open applications (3 boxes) */
 			map_key_clear(KEY_SCALE);
 			return 1;
-		case 0x00fa: /* Fn-Esc: Fn-lock toggle */
-			map_key_clear(KEY_FN_ESC);
-			return 1;
-		case 0x00fb: /* Fn-F12: Open My computer (6 boxes) USB-only */
+		case 0x00f9: /* Fn-F12: Open My computer (6 boxes) USB-only */
 			/* NB: This mapping is invented in raw_event below */
 			map_key_clear(KEY_FILE);
 			return 1;
+		case 0x00fa: /* Fn-Esc: Fn-lock toggle */
+			map_key_clear(KEY_FN_ESC);
+			return 1;
 		}
 	}
 
@@ -207,8 +206,8 @@ static int lenovo_raw_event(struct hid_device *hdev,
 			&& data[0] == 0x15
 			&& data[1] == 0x94
 			&& data[2] == 0x01)) {
-		data[1] = 0x0;
-		data[2] = 0x4;
+		data[1] = 0x00;
+		data[2] = 0x01;
 	}
 
 	return 0;
diff --git a/drivers/hid/hid-logitech-dj.c b/drivers/hid/hid-logitech-dj.c
index 9bf8637747a5..5bc6d80d5be7 100644
--- a/drivers/hid/hid-logitech-dj.c
+++ b/drivers/hid/hid-logitech-dj.c
@@ -26,9 +26,104 @@
 #include <linux/hid.h>
 #include <linux/module.h>
 #include <linux/usb.h>
+#include <linux/kfifo.h>
 #include <asm/unaligned.h>
 #include "hid-ids.h"
-#include "hid-logitech-dj.h"
+
+#define DJ_MAX_PAIRED_DEVICES			6
+#define DJ_MAX_NUMBER_NOTIFICATIONS		8
+#define DJ_RECEIVER_INDEX			0
+#define DJ_DEVICE_INDEX_MIN			1
+#define DJ_DEVICE_INDEX_MAX			6
+
+#define DJREPORT_SHORT_LENGTH			15
+#define DJREPORT_LONG_LENGTH			32
+
+#define REPORT_ID_DJ_SHORT			0x20
+#define REPORT_ID_DJ_LONG			0x21
+
+#define REPORT_ID_HIDPP_SHORT			0x10
+#define REPORT_ID_HIDPP_LONG			0x11
+
+#define HIDPP_REPORT_SHORT_LENGTH		7
+#define HIDPP_REPORT_LONG_LENGTH		20
+
+#define HIDPP_RECEIVER_INDEX			0xff
+
+#define REPORT_TYPE_RFREPORT_FIRST		0x01
+#define REPORT_TYPE_RFREPORT_LAST		0x1F
+
+/* Command Switch to DJ mode */
+#define REPORT_TYPE_CMD_SWITCH			0x80
+#define CMD_SWITCH_PARAM_DEVBITFIELD		0x00
+#define CMD_SWITCH_PARAM_TIMEOUT_SECONDS	0x01
+#define TIMEOUT_NO_KEEPALIVE			0x00
+
+/* Command to Get the list of Paired devices */
+#define REPORT_TYPE_CMD_GET_PAIRED_DEVICES	0x81
+
+/* Device Paired Notification */
+#define REPORT_TYPE_NOTIF_DEVICE_PAIRED		0x41
+#define SPFUNCTION_MORE_NOTIF_EXPECTED		0x01
+#define SPFUNCTION_DEVICE_LIST_EMPTY		0x02
+#define DEVICE_PAIRED_PARAM_SPFUNCTION		0x00
+#define DEVICE_PAIRED_PARAM_EQUAD_ID_LSB	0x01
+#define DEVICE_PAIRED_PARAM_EQUAD_ID_MSB	0x02
+#define DEVICE_PAIRED_RF_REPORT_TYPE		0x03
+
+/* Device Un-Paired Notification */
+#define REPORT_TYPE_NOTIF_DEVICE_UNPAIRED	0x40
+
+
+/* Connection Status Notification */
+#define REPORT_TYPE_NOTIF_CONNECTION_STATUS	0x42
+#define CONNECTION_STATUS_PARAM_STATUS		0x00
+#define STATUS_LINKLOSS				0x01
+
+/* Error Notification */
+#define REPORT_TYPE_NOTIF_ERROR			0x7F
+#define NOTIF_ERROR_PARAM_ETYPE			0x00
+#define ETYPE_KEEPALIVE_TIMEOUT			0x01
+
+/* supported DJ HID && RF report types */
+#define REPORT_TYPE_KEYBOARD			0x01
+#define REPORT_TYPE_MOUSE			0x02
+#define REPORT_TYPE_CONSUMER_CONTROL		0x03
+#define REPORT_TYPE_SYSTEM_CONTROL		0x04
+#define REPORT_TYPE_MEDIA_CENTER		0x08
+#define REPORT_TYPE_LEDS			0x0E
+
+/* RF Report types bitfield */
+#define STD_KEYBOARD				0x00000002
+#define STD_MOUSE				0x00000004
+#define MULTIMEDIA				0x00000008
+#define POWER_KEYS				0x00000010
+#define MEDIA_CENTER				0x00000100
+#define KBD_LEDS				0x00004000
+
+struct dj_report {
+	u8 report_id;
+	u8 device_index;
+	u8 report_type;
+	u8 report_params[DJREPORT_SHORT_LENGTH - 3];
+};
+
+struct dj_receiver_dev {
+	struct hid_device *hdev;
+	struct dj_device *paired_dj_devices[DJ_MAX_PAIRED_DEVICES +
+					    DJ_DEVICE_INDEX_MIN];
+	struct work_struct work;
+	struct kfifo notif_fifo;
+	spinlock_t lock;
+	bool querying_devices;
+};
+
+struct dj_device {
+	struct hid_device *hdev;
+	struct dj_receiver_dev *dj_receiver_dev;
+	u32 reports_supported;
+	u8 device_index;
+};
 
 /* Keyboard descriptor (1) */
 static const char kbd_descriptor[] = {
@@ -156,6 +251,57 @@ static const char media_descriptor[] = {
 	0xc0,			/* EndCollection                       */
 };				/*                                     */
 
+/* HIDPP descriptor */
+static const char hidpp_descriptor[] = {
+	0x06, 0x00, 0xff,	/* Usage Page (Vendor Defined Page 1)  */
+	0x09, 0x01,		/* Usage (Vendor Usage 1)              */
+	0xa1, 0x01,		/* Collection (Application)            */
+	0x85, 0x10,		/*   Report ID (16)                    */
+	0x75, 0x08,		/*   Report Size (8)                   */
+	0x95, 0x06,		/*   Report Count (6)                  */
+	0x15, 0x00,		/*   Logical Minimum (0)               */
+	0x26, 0xff, 0x00,	/*   Logical Maximum (255)             */
+	0x09, 0x01,		/*   Usage (Vendor Usage 1)            */
+	0x81, 0x00,		/*   Input (Data,Arr,Abs)              */
+	0x09, 0x01,		/*   Usage (Vendor Usage 1)            */
+	0x91, 0x00,		/*   Output (Data,Arr,Abs)             */
+	0xc0,			/* End Collection                      */
+	0x06, 0x00, 0xff,	/* Usage Page (Vendor Defined Page 1)  */
+	0x09, 0x02,		/* Usage (Vendor Usage 2)              */
+	0xa1, 0x01,		/* Collection (Application)            */
+	0x85, 0x11,		/*   Report ID (17)                    */
+	0x75, 0x08,		/*   Report Size (8)                   */
+	0x95, 0x13,		/*   Report Count (19)                 */
+	0x15, 0x00,		/*   Logical Minimum (0)               */
+	0x26, 0xff, 0x00,	/*   Logical Maximum (255)             */
+	0x09, 0x02,		/*   Usage (Vendor Usage 2)            */
+	0x81, 0x00,		/*   Input (Data,Arr,Abs)              */
+	0x09, 0x02,		/*   Usage (Vendor Usage 2)            */
+	0x91, 0x00,		/*   Output (Data,Arr,Abs)             */
+	0xc0,			/* End Collection                      */
+	0x06, 0x00, 0xff,	/* Usage Page (Vendor Defined Page 1)  */
+	0x09, 0x04,		/* Usage (Vendor Usage 0x04)           */
+	0xa1, 0x01,		/* Collection (Application)            */
+	0x85, 0x20,		/*   Report ID (32)                    */
+	0x75, 0x08,		/*   Report Size (8)                   */
+	0x95, 0x0e,		/*   Report Count (14)                 */
+	0x15, 0x00,		/*   Logical Minimum (0)               */
+	0x26, 0xff, 0x00,	/*   Logical Maximum (255)             */
+	0x09, 0x41,		/*   Usage (Vendor Usage 0x41)         */
+	0x81, 0x00,		/*   Input (Data,Arr,Abs)              */
+	0x09, 0x41,		/*   Usage (Vendor Usage 0x41)         */
+	0x91, 0x00,		/*   Output (Data,Arr,Abs)             */
+	0x85, 0x21,		/*   Report ID (33)                    */
+	0x95, 0x1f,		/*   Report Count (31)                 */
+	0x15, 0x00,		/*   Logical Minimum (0)               */
+	0x26, 0xff, 0x00,	/*   Logical Maximum (255)             */
+	0x09, 0x42,		/*   Usage (Vendor Usage 0x42)         */
+	0x81, 0x00,		/*   Input (Data,Arr,Abs)              */
+	0x09, 0x42,		/*   Usage (Vendor Usage 0x42)         */
+	0x91, 0x00,		/*   Output (Data,Arr,Abs)             */
+	0xc0,			/* End Collection                      */
+};
+
 /* Maximum size of all defined hid reports in bytes (including report id) */
 #define MAX_REPORT_SIZE 8
 
@@ -165,7 +311,8 @@ static const char media_descriptor[] = {
 	 sizeof(mse_descriptor) +		\
 	 sizeof(consumer_descriptor) +		\
 	 sizeof(syscontrol_descriptor) +	\
-	 sizeof(media_descriptor))
+	 sizeof(media_descriptor) +	\
+	 sizeof(hidpp_descriptor))
 
 /* Number of possible hid report types that can be created by this driver.
  *
@@ -256,11 +403,15 @@ static void logi_dj_recv_add_djhid_device(struct dj_receiver_dev *djrcv_dev,
 	dj_hiddev->dev.parent = &djrcv_hdev->dev;
 	dj_hiddev->bus = BUS_USB;
 	dj_hiddev->vendor = le16_to_cpu(usbdev->descriptor.idVendor);
-	dj_hiddev->product = le16_to_cpu(usbdev->descriptor.idProduct);
+	dj_hiddev->product =
+		(dj_report->report_params[DEVICE_PAIRED_PARAM_EQUAD_ID_MSB]
+									<< 8) |
+		dj_report->report_params[DEVICE_PAIRED_PARAM_EQUAD_ID_LSB];
 	snprintf(dj_hiddev->name, sizeof(dj_hiddev->name),
-		"Logitech Unifying Device. Wireless PID:%02x%02x",
-		dj_report->report_params[DEVICE_PAIRED_PARAM_EQUAD_ID_MSB],
-		dj_report->report_params[DEVICE_PAIRED_PARAM_EQUAD_ID_LSB]);
+		"Logitech Unifying Device. Wireless PID:%04x",
+		dj_hiddev->product);
+
+	dj_hiddev->group = HID_GROUP_LOGITECH_DJ_DEVICE;
 
 	usb_make_path(usbdev, dj_hiddev->phys, sizeof(dj_hiddev->phys));
 	snprintf(tmpstr, sizeof(tmpstr), ":%d", dj_report->device_index);
@@ -385,18 +536,6 @@ static void logi_dj_recv_forward_null_report(struct dj_receiver_dev *djrcv_dev,
 
 	djdev = djrcv_dev->paired_dj_devices[dj_report->device_index];
 
-	if (!djdev) {
-		dbg_hid("djrcv_dev->paired_dj_devices[dj_report->device_index]"
-			" is NULL, index %d\n", dj_report->device_index);
-		kfifo_in(&djrcv_dev->notif_fifo, dj_report, sizeof(struct dj_report));
-
-		if (schedule_work(&djrcv_dev->work) == 0) {
-			dbg_hid("%s: did not schedule the work item, was already "
-			"queued\n", __func__);
-		}
-		return;
-	}
-
 	memset(reportbuffer, 0, sizeof(reportbuffer));
 
 	for (i = 0; i < NUMBER_OF_HID_REPORTS; i++) {
@@ -421,18 +560,6 @@ static void logi_dj_recv_forward_report(struct dj_receiver_dev *djrcv_dev,
 
 	dj_device = djrcv_dev->paired_dj_devices[dj_report->device_index];
 
-	if (dj_device == NULL) {
-		dbg_hid("djrcv_dev->paired_dj_devices[dj_report->device_index]"
-			" is NULL, index %d\n", dj_report->device_index);
-		kfifo_in(&djrcv_dev->notif_fifo, dj_report, sizeof(struct dj_report));
-
-		if (schedule_work(&djrcv_dev->work) == 0) {
-			dbg_hid("%s: did not schedule the work item, was already "
-			"queued\n", __func__);
-		}
-		return;
-	}
-
 	if ((dj_report->report_type > ARRAY_SIZE(hid_reportid_size_map) - 1) ||
 	    (hid_reportid_size_map[dj_report->report_type] == 0)) {
 		dbg_hid("invalid report type:%x\n", dj_report->report_type);
@@ -446,6 +573,13 @@ static void logi_dj_recv_forward_report(struct dj_receiver_dev *djrcv_dev,
 	}
 }
 
+static void logi_dj_recv_forward_hidpp(struct dj_device *dj_dev, u8 *data,
+				       int size)
+{
+	/* We are called from atomic context (tasklet && djrcv->lock held) */
+	if (hid_input_report(dj_dev->hdev, HID_INPUT_REPORT, data, size, 1))
+		dbg_hid("hid_input_report error\n");
+}
 
 static int logi_dj_recv_send_report(struct dj_receiver_dev *djrcv_dev,
 				    struct dj_report *dj_report)
@@ -496,7 +630,9 @@ static int logi_dj_recv_query_paired_devices(struct dj_receiver_dev *djrcv_dev)
 static int logi_dj_recv_switch_to_dj_mode(struct dj_receiver_dev *djrcv_dev,
 					  unsigned timeout)
 {
+	struct hid_device *hdev = djrcv_dev->hdev;
 	struct dj_report *dj_report;
+	u8 *buf;
 	int retval;
 
 	dj_report = kzalloc(sizeof(struct dj_report), GFP_KERNEL);
@@ -508,7 +644,6 @@ static int logi_dj_recv_switch_to_dj_mode(struct dj_receiver_dev *djrcv_dev,
 	dj_report->report_params[CMD_SWITCH_PARAM_DEVBITFIELD] = 0x3F;
 	dj_report->report_params[CMD_SWITCH_PARAM_TIMEOUT_SECONDS] = (u8)timeout;
 	retval = logi_dj_recv_send_report(djrcv_dev, dj_report);
-	kfree(dj_report);
 
 	/*
 	 * Ugly sleep to work around a USB 3.0 bug when the receiver is still
@@ -517,6 +652,30 @@ static int logi_dj_recv_switch_to_dj_mode(struct dj_receiver_dev *djrcv_dev,
 	 */
 	msleep(50);
 
+	/*
+	 * Magical bits to set up hidpp notifications when the dj devices
+	 * are connected/disconnected.
+	 *
+	 * We can reuse dj_report because HIDPP_REPORT_SHORT_LENGTH is smaller
+	 * than DJREPORT_SHORT_LENGTH.
+	 */
+	buf = (u8 *)dj_report;
+
+	memset(buf, 0, HIDPP_REPORT_SHORT_LENGTH);
+
+	buf[0] = REPORT_ID_HIDPP_SHORT;
+	buf[1] = 0xFF;
+	buf[2] = 0x80;
+	buf[3] = 0x00;
+	buf[4] = 0x00;
+	buf[5] = 0x09;
+	buf[6] = 0x00;
+
+	hid_hw_raw_request(hdev, REPORT_ID_HIDPP_SHORT, buf,
+			HIDPP_REPORT_SHORT_LENGTH, HID_OUTPUT_REPORT,
+			HID_REQ_SET_REPORT);
+
+	kfree(dj_report);
 	return retval;
 }
 
@@ -533,6 +692,9 @@ static void logi_dj_ll_close(struct hid_device *hid)
 	dbg_hid("%s:%s\n", __func__, hid->phys);
 }
 
+static u8 unifying_name_query[]  = {0x10, 0xff, 0x83, 0xb5, 0x40, 0x00, 0x00};
+static u8 unifying_name_answer[] = {0x11, 0xff, 0x83, 0xb5};
+
 static int logi_dj_ll_raw_request(struct hid_device *hid,
 				  unsigned char reportnum, __u8 *buf,
 				  size_t count, unsigned char report_type,
@@ -543,6 +705,22 @@ static int logi_dj_ll_raw_request(struct hid_device *hid,
 	u8 *out_buf;
 	int ret;
 
+	if ((buf[0] == REPORT_ID_HIDPP_SHORT) ||
+	    (buf[0] == REPORT_ID_HIDPP_LONG)) {
+		if (count < 2)
+			return -EINVAL;
+
+		/* special case where we should not overwrite
+		 * the device_index */
+		if (count == 7 && !memcmp(buf, unifying_name_query,
+					  sizeof(unifying_name_query)))
+			buf[4] |= djdev->device_index - 1;
+		else
+			buf[1] = djdev->device_index;
+		return hid_hw_raw_request(djrcv_dev->hdev, reportnum, buf,
+				count, report_type, reqtype);
+	}
+
 	if (buf[0] != REPORT_TYPE_LEDS)
 		return -EINVAL;
 
@@ -621,6 +799,8 @@ static int logi_dj_ll_parse(struct hid_device *hid)
 			__func__, djdev->reports_supported);
 	}
 
+	rdcat(rdesc, &rsize, hidpp_descriptor, sizeof(hidpp_descriptor));
+
 	retval = hid_parse_report(hid, rdesc, rsize);
 	kfree(rdesc);
 
@@ -648,8 +828,7 @@ static struct hid_ll_driver logi_dj_ll_driver = {
 	.raw_request = logi_dj_ll_raw_request,
 };
 
-
-static int logi_dj_raw_event(struct hid_device *hdev,
+static int logi_dj_dj_event(struct hid_device *hdev,
 			     struct hid_report *report, u8 *data,
 			     int size)
 {
@@ -657,36 +836,24 @@ static int logi_dj_raw_event(struct hid_device *hdev,
 	struct dj_report *dj_report = (struct dj_report *) data;
 	unsigned long flags;
 
-	dbg_hid("%s, size:%d\n", __func__, size);
-
-	/* Here we receive all data coming from iface 2, there are 4 cases:
-	 *
-	 * 1) Data should continue its normal processing i.e. data does not
-	 * come from the DJ collection, in which case we do nothing and
-	 * return 0, so hid-core can continue normal processing (will forward
-	 * to associated hidraw device)
+	/*
+	 * Here we receive all data coming from iface 2, there are 3 cases:
 	 *
-	 * 2) Data is from DJ collection, and is intended for this driver i. e.
-	 * data contains arrival, departure, etc notifications, in which case
-	 * we queue them for delayed processing by the work queue. We return 1
-	 * to hid-core as no further processing is required from it.
+	 * 1) Data is intended for this driver i. e. data contains arrival,
+	 * departure, etc notifications, in which case we queue them for delayed
+	 * processing by the work queue. We return 1 to hid-core as no further
+	 * processing is required from it.
 	 *
-	 * 3) Data is from DJ collection, and informs a connection change,
-	 * if the change means rf link loss, then we must send a null report
-	 * to the upper layer to discard potentially pressed keys that may be
-	 * repeated forever by the input layer. Return 1 to hid-core as no
-	 * further processing is required.
+	 * 2) Data informs a connection change, if the change means rf link
+	 * loss, then we must send a null report to the upper layer to discard
+	 * potentially pressed keys that may be repeated forever by the input
+	 * layer. Return 1 to hid-core as no further processing is required.
 	 *
-	 * 4) Data is from DJ collection and is an actual input event from
-	 * a paired DJ device in which case we forward it to the correct hid
-	 * device (via hid_input_report() ) and return 1 so hid-core does not do
-	 * anything else with it.
+	 * 3) Data is an actual input event from a paired DJ device in which
+	 * case we forward it to the correct hid device (via hid_input_report()
+	 * ) and return 1 so hid-core does not anything else with it.
 	 */
 
-	/* case 1) */
-	if (data[0] != REPORT_ID_DJ_SHORT)
-		return false;
-
 	if ((dj_report->device_index < DJ_DEVICE_INDEX_MIN) ||
 	    (dj_report->device_index > DJ_DEVICE_INDEX_MAX)) {
 		/*
@@ -701,8 +868,17 @@ static int logi_dj_raw_event(struct hid_device *hdev,
 	}
 
 	spin_lock_irqsave(&djrcv_dev->lock, flags);
+
+	if (!djrcv_dev->paired_dj_devices[dj_report->device_index]) {
+		/* received an event for an unknown device, bail out */
+		logi_dj_recv_queue_notification(djrcv_dev, dj_report);
+		goto out;
+	}
+
 	switch (dj_report->report_type) {
 	case REPORT_TYPE_NOTIF_DEVICE_PAIRED:
+		/* pairing notifications are handled above the switch */
+		break;
 	case REPORT_TYPE_NOTIF_DEVICE_UNPAIRED:
 		logi_dj_recv_queue_notification(djrcv_dev, dj_report);
 		break;
@@ -715,11 +891,101 @@ static int logi_dj_raw_event(struct hid_device *hdev,
 	default:
 		logi_dj_recv_forward_report(djrcv_dev, dj_report);
 	}
+
+out:
 	spin_unlock_irqrestore(&djrcv_dev->lock, flags);
 
 	return true;
 }
 
+static int logi_dj_hidpp_event(struct hid_device *hdev,
+			     struct hid_report *report, u8 *data,
+			     int size)
+{
+	struct dj_receiver_dev *djrcv_dev = hid_get_drvdata(hdev);
+	struct dj_report *dj_report = (struct dj_report *) data;
+	unsigned long flags;
+	u8 device_index = dj_report->device_index;
+
+	if (device_index == HIDPP_RECEIVER_INDEX) {
+		/* special case were the device wants to know its unifying
+		 * name */
+		if (size == HIDPP_REPORT_LONG_LENGTH &&
+		    !memcmp(data, unifying_name_answer,
+			    sizeof(unifying_name_answer)) &&
+		    ((data[4] & 0xF0) == 0x40))
+			device_index = (data[4] & 0x0F) + 1;
+		else
+			return false;
+	}
+
+	/*
+	 * Data is from the HID++ collection, in this case, we forward the
+	 * data to the corresponding child dj device and return 0 to hid-core
+	 * so he data also goes to the hidraw device of the receiver. This
+	 * allows a user space application to implement the full HID++ routing
+	 * via the receiver.
+	 */
+
+	if ((device_index < DJ_DEVICE_INDEX_MIN) ||
+	    (device_index > DJ_DEVICE_INDEX_MAX)) {
+		/*
+		 * Device index is wrong, bail out.
+		 * This driver can ignore safely the receiver notifications,
+		 * so ignore those reports too.
+		 */
+		dev_err(&hdev->dev, "%s: invalid device index:%d\n",
+				__func__, dj_report->device_index);
+		return false;
+	}
+
+	spin_lock_irqsave(&djrcv_dev->lock, flags);
+
+	if (!djrcv_dev->paired_dj_devices[device_index])
+		/* received an event for an unknown device, bail out */
+		goto out;
+
+	logi_dj_recv_forward_hidpp(djrcv_dev->paired_dj_devices[device_index],
+				   data, size);
+
+out:
+	spin_unlock_irqrestore(&djrcv_dev->lock, flags);
+
+	return false;
+}
+
+static int logi_dj_raw_event(struct hid_device *hdev,
+			     struct hid_report *report, u8 *data,
+			     int size)
+{
+	dbg_hid("%s, size:%d\n", __func__, size);
+
+	switch (data[0]) {
+	case REPORT_ID_DJ_SHORT:
+		if (size != DJREPORT_SHORT_LENGTH) {
+			dev_err(&hdev->dev, "DJ report of bad size (%d)", size);
+			return false;
+		}
+		return logi_dj_dj_event(hdev, report, data, size);
+	case REPORT_ID_HIDPP_SHORT:
+		if (size != HIDPP_REPORT_SHORT_LENGTH) {
+			dev_err(&hdev->dev,
+				"Short HID++ report of bad size (%d)", size);
+			return false;
+		}
+		return logi_dj_hidpp_event(hdev, report, data, size);
+	case REPORT_ID_HIDPP_LONG:
+		if (size != HIDPP_REPORT_LONG_LENGTH) {
+			dev_err(&hdev->dev,
+				"Long HID++ report of bad size (%d)", size);
+			return false;
+		}
+		return logi_dj_hidpp_event(hdev, report, data, size);
+	}
+
+	return false;
+}
+
 static int logi_dj_probe(struct hid_device *hdev,
 			 const struct hid_device_id *id)
 {
@@ -727,9 +993,6 @@ static int logi_dj_probe(struct hid_device *hdev,
 	struct dj_receiver_dev *djrcv_dev;
 	int retval;
 
-	if (is_dj_device((struct dj_device *)hdev->driver_data))
-		return -ENODEV;
-
 	dbg_hid("%s called for ifnum %d\n", __func__,
 		intf->cur_altsetting->desc.bInterfaceNumber);
 
@@ -882,22 +1145,6 @@ static void logi_dj_remove(struct hid_device *hdev)
 	hid_set_drvdata(hdev, NULL);
 }
 
-static int logi_djdevice_probe(struct hid_device *hdev,
-			 const struct hid_device_id *id)
-{
-	int ret;
-	struct dj_device *dj_dev = hdev->driver_data;
-
-	if (!is_dj_device(dj_dev))
-		return -ENODEV;
-
-	ret = hid_parse(hdev);
-	if (!ret)
-		ret = hid_hw_start(hdev, HID_CONNECT_DEFAULT);
-
-	return ret;
-}
-
 static const struct hid_device_id logi_dj_receivers[] = {
 	{HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH,
 		USB_DEVICE_ID_LOGITECH_UNIFYING_RECEIVER)},
@@ -919,51 +1166,8 @@ static struct hid_driver logi_djreceiver_driver = {
 #endif
 };
 
+module_hid_driver(logi_djreceiver_driver);
 
-static const struct hid_device_id logi_dj_devices[] = {
-	{HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH,
-		USB_DEVICE_ID_LOGITECH_UNIFYING_RECEIVER)},
-	{HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH,
-		USB_DEVICE_ID_LOGITECH_UNIFYING_RECEIVER_2)},
-	{}
-};
-
-static struct hid_driver logi_djdevice_driver = {
-	.name = "logitech-djdevice",
-	.id_table = logi_dj_devices,
-	.probe = logi_djdevice_probe,
-};
-
-
-static int __init logi_dj_init(void)
-{
-	int retval;
-
-	dbg_hid("Logitech-DJ:%s\n", __func__);
-
-	retval = hid_register_driver(&logi_djreceiver_driver);
-	if (retval)
-		return retval;
-
-	retval = hid_register_driver(&logi_djdevice_driver);
-	if (retval)
-		hid_unregister_driver(&logi_djreceiver_driver);
-
-	return retval;
-
-}
-
-static void __exit logi_dj_exit(void)
-{
-	dbg_hid("Logitech-DJ:%s\n", __func__);
-
-	hid_unregister_driver(&logi_djdevice_driver);
-	hid_unregister_driver(&logi_djreceiver_driver);
-
-}
-
-module_init(logi_dj_init);
-module_exit(logi_dj_exit);
 MODULE_LICENSE("GPL");
 MODULE_AUTHOR("Logitech");
 MODULE_AUTHOR("Nestor Lopez Casado");
diff --git a/drivers/hid/hid-logitech-dj.h b/drivers/hid/hid-logitech-dj.h
deleted file mode 100644
index daeb0aa4bee9..000000000000
--- a/drivers/hid/hid-logitech-dj.h
+++ /dev/null
@@ -1,125 +0,0 @@
-#ifndef __HID_LOGITECH_DJ_H
-#define __HID_LOGITECH_DJ_H
-
-/*
- *  HID driver for Logitech Unifying receivers
- *
- *  Copyright (c) 2011 Logitech
- */
-
-/*
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
- *
- */
-
-#include <linux/kfifo.h>
-
-#define DJ_MAX_PAIRED_DEVICES			6
-#define DJ_MAX_NUMBER_NOTIFICATIONS		8
-#define DJ_RECEIVER_INDEX			0
-#define DJ_DEVICE_INDEX_MIN 			1
-#define DJ_DEVICE_INDEX_MAX 			6
-
-#define DJREPORT_SHORT_LENGTH			15
-#define DJREPORT_LONG_LENGTH			32
-
-#define REPORT_ID_DJ_SHORT			0x20
-#define REPORT_ID_DJ_LONG			0x21
-
-#define REPORT_TYPE_RFREPORT_FIRST		0x01
-#define REPORT_TYPE_RFREPORT_LAST		0x1F
-
-/* Command Switch to DJ mode */
-#define REPORT_TYPE_CMD_SWITCH			0x80
-#define CMD_SWITCH_PARAM_DEVBITFIELD		0x00
-#define CMD_SWITCH_PARAM_TIMEOUT_SECONDS	0x01
-#define TIMEOUT_NO_KEEPALIVE			0x00
-
-/* Command to Get the list of Paired devices */
-#define REPORT_TYPE_CMD_GET_PAIRED_DEVICES	0x81
-
-/* Device Paired Notification */
-#define REPORT_TYPE_NOTIF_DEVICE_PAIRED		0x41
-#define SPFUNCTION_MORE_NOTIF_EXPECTED		0x01
-#define SPFUNCTION_DEVICE_LIST_EMPTY		0x02
-#define DEVICE_PAIRED_PARAM_SPFUNCTION		0x00
-#define DEVICE_PAIRED_PARAM_EQUAD_ID_LSB	0x01
-#define DEVICE_PAIRED_PARAM_EQUAD_ID_MSB	0x02
-#define DEVICE_PAIRED_RF_REPORT_TYPE		0x03
-
-/* Device Un-Paired Notification */
-#define REPORT_TYPE_NOTIF_DEVICE_UNPAIRED	0x40
-
-
-/* Connection Status Notification */
-#define REPORT_TYPE_NOTIF_CONNECTION_STATUS	0x42
-#define CONNECTION_STATUS_PARAM_STATUS		0x00
-#define STATUS_LINKLOSS				0x01
-
-/* Error Notification */
-#define REPORT_TYPE_NOTIF_ERROR			0x7F
-#define NOTIF_ERROR_PARAM_ETYPE			0x00
-#define ETYPE_KEEPALIVE_TIMEOUT			0x01
-
-/* supported DJ HID && RF report types */
-#define REPORT_TYPE_KEYBOARD			0x01
-#define REPORT_TYPE_MOUSE			0x02
-#define REPORT_TYPE_CONSUMER_CONTROL		0x03
-#define REPORT_TYPE_SYSTEM_CONTROL		0x04
-#define REPORT_TYPE_MEDIA_CENTER		0x08
-#define REPORT_TYPE_LEDS			0x0E
-
-/* RF Report types bitfield */
-#define STD_KEYBOARD				0x00000002
-#define STD_MOUSE				0x00000004
-#define MULTIMEDIA				0x00000008
-#define POWER_KEYS				0x00000010
-#define MEDIA_CENTER				0x00000100
-#define KBD_LEDS				0x00004000
-
-struct dj_report {
-	u8 report_id;
-	u8 device_index;
-	u8 report_type;
-	u8 report_params[DJREPORT_SHORT_LENGTH - 3];
-};
-
-struct dj_receiver_dev {
-	struct hid_device *hdev;
-	struct dj_device *paired_dj_devices[DJ_MAX_PAIRED_DEVICES +
-					    DJ_DEVICE_INDEX_MIN];
-	struct work_struct work;
-	struct kfifo notif_fifo;
-	spinlock_t lock;
-	bool querying_devices;
-};
-
-struct dj_device {
-	struct hid_device *hdev;
-	struct dj_receiver_dev *dj_receiver_dev;
-	u32 reports_supported;
-	u8 device_index;
-};
-
-/**
- * is_dj_device - know if the given dj_device is not the receiver.
- * @dj_dev: the dj device to test
- *
- * This macro tests if a struct dj_device pointer is a device created
- * by the bus enumarator.
- */
-#define is_dj_device(dj_dev) \
-	(&(dj_dev)->dj_receiver_dev->hdev->dev == (dj_dev)->hdev->dev.parent)
-
-#endif
diff --git a/drivers/hid/hid-logitech-hidpp.c b/drivers/hid/hid-logitech-hidpp.c
new file mode 100644
index 000000000000..a93cefe0e522
--- /dev/null
+++ b/drivers/hid/hid-logitech-hidpp.c
@@ -0,0 +1,1282 @@
+/*
+ *  HIDPP protocol for Logitech Unifying receivers
+ *
+ *  Copyright (c) 2011 Logitech (c)
+ *  Copyright (c) 2012-2013 Google (c)
+ *  Copyright (c) 2013-2014 Red Hat Inc.
+ */
+
+/*
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the Free
+ * Software Foundation; version 2 of the License.
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/device.h>
+#include <linux/hid.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/sched.h>
+#include <linux/kfifo.h>
+#include <linux/input/mt.h>
+#include <asm/unaligned.h>
+#include "hid-ids.h"
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Benjamin Tissoires <benjamin.tissoires@gmail.com>");
+MODULE_AUTHOR("Nestor Lopez Casado <nlopezcasad@logitech.com>");
+
+#define REPORT_ID_HIDPP_SHORT			0x10
+#define REPORT_ID_HIDPP_LONG			0x11
+
+#define HIDPP_REPORT_SHORT_LENGTH		7
+#define HIDPP_REPORT_LONG_LENGTH		20
+
+#define HIDPP_QUIRK_CLASS_WTP			BIT(0)
+
+/* bits 1..20 are reserved for classes */
+#define HIDPP_QUIRK_DELAYED_INIT		BIT(21)
+#define HIDPP_QUIRK_WTP_PHYSICAL_BUTTONS	BIT(22)
+#define HIDPP_QUIRK_MULTI_INPUT			BIT(23)
+
+/*
+ * There are two hidpp protocols in use, the first version hidpp10 is known
+ * as register access protocol or RAP, the second version hidpp20 is known as
+ * feature access protocol or FAP
+ *
+ * Most older devices (including the Unifying usb receiver) use the RAP protocol
+ * where as most newer devices use the FAP protocol. Both protocols are
+ * compatible with the underlying transport, which could be usb, Unifiying, or
+ * bluetooth. The message lengths are defined by the hid vendor specific report
+ * descriptor for the HIDPP_SHORT report type (total message lenth 7 bytes) and
+ * the HIDPP_LONG report type (total message length 20 bytes)
+ *
+ * The RAP protocol uses both report types, whereas the FAP only uses HIDPP_LONG
+ * messages. The Unifying receiver itself responds to RAP messages (device index
+ * is 0xFF for the receiver), and all messages (short or long) with a device
+ * index between 1 and 6 are passed untouched to the corresponding paired
+ * Unifying device.
+ *
+ * The paired device can be RAP or FAP, it will receive the message untouched
+ * from the Unifiying receiver.
+ */
+
+struct fap {
+	u8 feature_index;
+	u8 funcindex_clientid;
+	u8 params[HIDPP_REPORT_LONG_LENGTH - 4U];
+};
+
+struct rap {
+	u8 sub_id;
+	u8 reg_address;
+	u8 params[HIDPP_REPORT_LONG_LENGTH - 4U];
+};
+
+struct hidpp_report {
+	u8 report_id;
+	u8 device_index;
+	union {
+		struct fap fap;
+		struct rap rap;
+		u8 rawbytes[sizeof(struct fap)];
+	};
+} __packed;
+
+struct hidpp_device {
+	struct hid_device *hid_dev;
+	struct mutex send_mutex;
+	void *send_receive_buf;
+	wait_queue_head_t wait;
+	bool answer_available;
+	u8 protocol_major;
+	u8 protocol_minor;
+
+	void *private_data;
+
+	struct work_struct work;
+	struct kfifo delayed_work_fifo;
+	atomic_t connected;
+	struct input_dev *delayed_input;
+
+	unsigned long quirks;
+};
+
+
+#define HIDPP_ERROR				0x8f
+#define HIDPP_ERROR_SUCCESS			0x00
+#define HIDPP_ERROR_INVALID_SUBID		0x01
+#define HIDPP_ERROR_INVALID_ADRESS		0x02
+#define HIDPP_ERROR_INVALID_VALUE		0x03
+#define HIDPP_ERROR_CONNECT_FAIL		0x04
+#define HIDPP_ERROR_TOO_MANY_DEVICES		0x05
+#define HIDPP_ERROR_ALREADY_EXISTS		0x06
+#define HIDPP_ERROR_BUSY			0x07
+#define HIDPP_ERROR_UNKNOWN_DEVICE		0x08
+#define HIDPP_ERROR_RESOURCE_ERROR		0x09
+#define HIDPP_ERROR_REQUEST_UNAVAILABLE		0x0a
+#define HIDPP_ERROR_INVALID_PARAM_VALUE		0x0b
+#define HIDPP_ERROR_WRONG_PIN_CODE		0x0c
+
+static void hidpp_connect_event(struct hidpp_device *hidpp_dev);
+
+static int __hidpp_send_report(struct hid_device *hdev,
+				struct hidpp_report *hidpp_report)
+{
+	int fields_count, ret;
+
+	switch (hidpp_report->report_id) {
+	case REPORT_ID_HIDPP_SHORT:
+		fields_count = HIDPP_REPORT_SHORT_LENGTH;
+		break;
+	case REPORT_ID_HIDPP_LONG:
+		fields_count = HIDPP_REPORT_LONG_LENGTH;
+		break;
+	default:
+		return -ENODEV;
+	}
+
+	/*
+	 * set the device_index as the receiver, it will be overwritten by
+	 * hid_hw_request if needed
+	 */
+	hidpp_report->device_index = 0xff;
+
+	ret = hid_hw_raw_request(hdev, hidpp_report->report_id,
+		(u8 *)hidpp_report, fields_count, HID_OUTPUT_REPORT,
+		HID_REQ_SET_REPORT);
+
+	return ret == fields_count ? 0 : -1;
+}
+
+/**
+ * hidpp_send_message_sync() returns 0 in case of success, and something else
+ * in case of a failure.
+ * - If ' something else' is positive, that means that an error has been raised
+ *   by the protocol itself.
+ * - If ' something else' is negative, that means that we had a classic error
+ *   (-ENOMEM, -EPIPE, etc...)
+ */
+static int hidpp_send_message_sync(struct hidpp_device *hidpp,
+	struct hidpp_report *message,
+	struct hidpp_report *response)
+{
+	int ret;
+
+	mutex_lock(&hidpp->send_mutex);
+
+	hidpp->send_receive_buf = response;
+	hidpp->answer_available = false;
+
+	/*
+	 * So that we can later validate the answer when it arrives
+	 * in hidpp_raw_event
+	 */
+	*response = *message;
+
+	ret = __hidpp_send_report(hidpp->hid_dev, message);
+
+	if (ret) {
+		dbg_hid("__hidpp_send_report returned err: %d\n", ret);
+		memset(response, 0, sizeof(struct hidpp_report));
+		goto exit;
+	}
+
+	if (!wait_event_timeout(hidpp->wait, hidpp->answer_available,
+				5*HZ)) {
+		dbg_hid("%s:timeout waiting for response\n", __func__);
+		memset(response, 0, sizeof(struct hidpp_report));
+		ret = -ETIMEDOUT;
+	}
+
+	if (response->report_id == REPORT_ID_HIDPP_SHORT &&
+	    response->fap.feature_index == HIDPP_ERROR) {
+		ret = response->fap.params[1];
+		dbg_hid("__hidpp_send_report got hidpp error %02X\n", ret);
+		goto exit;
+	}
+
+exit:
+	mutex_unlock(&hidpp->send_mutex);
+	return ret;
+
+}
+
+static int hidpp_send_fap_command_sync(struct hidpp_device *hidpp,
+	u8 feat_index, u8 funcindex_clientid, u8 *params, int param_count,
+	struct hidpp_report *response)
+{
+	struct hidpp_report *message;
+	int ret;
+
+	if (param_count > sizeof(message->fap.params))
+		return -EINVAL;
+
+	message = kzalloc(sizeof(struct hidpp_report), GFP_KERNEL);
+	if (!message)
+		return -ENOMEM;
+	message->report_id = REPORT_ID_HIDPP_LONG;
+	message->fap.feature_index = feat_index;
+	message->fap.funcindex_clientid = funcindex_clientid;
+	memcpy(&message->fap.params, params, param_count);
+
+	ret = hidpp_send_message_sync(hidpp, message, response);
+	kfree(message);
+	return ret;
+}
+
+static int hidpp_send_rap_command_sync(struct hidpp_device *hidpp_dev,
+	u8 report_id, u8 sub_id, u8 reg_address, u8 *params, int param_count,
+	struct hidpp_report *response)
+{
+	struct hidpp_report *message;
+	int ret;
+
+	if ((report_id != REPORT_ID_HIDPP_SHORT) &&
+	    (report_id != REPORT_ID_HIDPP_LONG))
+		return -EINVAL;
+
+	if (param_count > sizeof(message->rap.params))
+		return -EINVAL;
+
+	message = kzalloc(sizeof(struct hidpp_report), GFP_KERNEL);
+	if (!message)
+		return -ENOMEM;
+	message->report_id = report_id;
+	message->rap.sub_id = sub_id;
+	message->rap.reg_address = reg_address;
+	memcpy(&message->rap.params, params, param_count);
+
+	ret = hidpp_send_message_sync(hidpp_dev, message, response);
+	kfree(message);
+	return ret;
+}
+
+static void delayed_work_cb(struct work_struct *work)
+{
+	struct hidpp_device *hidpp = container_of(work, struct hidpp_device,
+							work);
+	hidpp_connect_event(hidpp);
+}
+
+static inline bool hidpp_match_answer(struct hidpp_report *question,
+		struct hidpp_report *answer)
+{
+	return (answer->fap.feature_index == question->fap.feature_index) &&
+	   (answer->fap.funcindex_clientid == question->fap.funcindex_clientid);
+}
+
+static inline bool hidpp_match_error(struct hidpp_report *question,
+		struct hidpp_report *answer)
+{
+	return (answer->fap.feature_index == HIDPP_ERROR) &&
+	    (answer->fap.funcindex_clientid == question->fap.feature_index) &&
+	    (answer->fap.params[0] == question->fap.funcindex_clientid);
+}
+
+static inline bool hidpp_report_is_connect_event(struct hidpp_report *report)
+{
+	return (report->report_id == REPORT_ID_HIDPP_SHORT) &&
+		(report->rap.sub_id == 0x41);
+}
+
+/**
+ * hidpp_prefix_name() prefixes the current given name with "Logitech ".
+ */
+static void hidpp_prefix_name(char **name, int name_length)
+{
+#define PREFIX_LENGTH 9 /* "Logitech " */
+
+	int new_length;
+	char *new_name;
+
+	if (name_length > PREFIX_LENGTH &&
+	    strncmp(*name, "Logitech ", PREFIX_LENGTH) == 0)
+		/* The prefix has is already in the name */
+		return;
+
+	new_length = PREFIX_LENGTH + name_length;
+	new_name = kzalloc(new_length, GFP_KERNEL);
+	if (!new_name)
+		return;
+
+	snprintf(new_name, new_length, "Logitech %s", *name);
+
+	kfree(*name);
+
+	*name = new_name;
+}
+
+/* -------------------------------------------------------------------------- */
+/* HIDP++ 1.0 commands                                                        */
+/* -------------------------------------------------------------------------- */
+
+#define HIDPP_SET_REGISTER				0x80
+#define HIDPP_GET_REGISTER				0x81
+#define HIDPP_SET_LONG_REGISTER				0x82
+#define HIDPP_GET_LONG_REGISTER				0x83
+
+#define HIDPP_REG_PAIRING_INFORMATION			0xB5
+#define DEVICE_NAME					0x40
+
+static char *hidpp_get_unifying_name(struct hidpp_device *hidpp_dev)
+{
+	struct hidpp_report response;
+	int ret;
+	/* hid-logitech-dj is in charge of setting the right device index */
+	u8 params[1] = { DEVICE_NAME };
+	char *name;
+	int len;
+
+	ret = hidpp_send_rap_command_sync(hidpp_dev,
+					REPORT_ID_HIDPP_SHORT,
+					HIDPP_GET_LONG_REGISTER,
+					HIDPP_REG_PAIRING_INFORMATION,
+					params, 1, &response);
+	if (ret)
+		return NULL;
+
+	len = response.rap.params[1];
+
+	if (2 + len > sizeof(response.rap.params))
+		return NULL;
+
+	name = kzalloc(len + 1, GFP_KERNEL);
+	if (!name)
+		return NULL;
+
+	memcpy(name, &response.rap.params[2], len);
+
+	/* include the terminating '\0' */
+	hidpp_prefix_name(&name, len + 1);
+
+	return name;
+}
+
+/* -------------------------------------------------------------------------- */
+/* 0x0000: Root                                                               */
+/* -------------------------------------------------------------------------- */
+
+#define HIDPP_PAGE_ROOT					0x0000
+#define HIDPP_PAGE_ROOT_IDX				0x00
+
+#define CMD_ROOT_GET_FEATURE				0x01
+#define CMD_ROOT_GET_PROTOCOL_VERSION			0x11
+
+static int hidpp_root_get_feature(struct hidpp_device *hidpp, u16 feature,
+	u8 *feature_index, u8 *feature_type)
+{
+	struct hidpp_report response;
+	int ret;
+	u8 params[2] = { feature >> 8, feature & 0x00FF };
+
+	ret = hidpp_send_fap_command_sync(hidpp,
+			HIDPP_PAGE_ROOT_IDX,
+			CMD_ROOT_GET_FEATURE,
+			params, 2, &response);
+	if (ret)
+		return ret;
+
+	*feature_index = response.fap.params[0];
+	*feature_type = response.fap.params[1];
+
+	return ret;
+}
+
+static int hidpp_root_get_protocol_version(struct hidpp_device *hidpp)
+{
+	struct hidpp_report response;
+	int ret;
+
+	ret = hidpp_send_fap_command_sync(hidpp,
+			HIDPP_PAGE_ROOT_IDX,
+			CMD_ROOT_GET_PROTOCOL_VERSION,
+			NULL, 0, &response);
+
+	if (ret == HIDPP_ERROR_INVALID_SUBID) {
+		hidpp->protocol_major = 1;
+		hidpp->protocol_minor = 0;
+		return 0;
+	}
+
+	/* the device might not be connected */
+	if (ret == HIDPP_ERROR_RESOURCE_ERROR)
+		return -EIO;
+
+	if (ret > 0) {
+		hid_err(hidpp->hid_dev, "%s: received protocol error 0x%02x\n",
+			__func__, ret);
+		return -EPROTO;
+	}
+	if (ret)
+		return ret;
+
+	hidpp->protocol_major = response.fap.params[0];
+	hidpp->protocol_minor = response.fap.params[1];
+
+	return ret;
+}
+
+static bool hidpp_is_connected(struct hidpp_device *hidpp)
+{
+	int ret;
+
+	ret = hidpp_root_get_protocol_version(hidpp);
+	if (!ret)
+		hid_dbg(hidpp->hid_dev, "HID++ %u.%u device connected.\n",
+			hidpp->protocol_major, hidpp->protocol_minor);
+	return ret == 0;
+}
+
+/* -------------------------------------------------------------------------- */
+/* 0x0005: GetDeviceNameType                                                  */
+/* -------------------------------------------------------------------------- */
+
+#define HIDPP_PAGE_GET_DEVICE_NAME_TYPE			0x0005
+
+#define CMD_GET_DEVICE_NAME_TYPE_GET_COUNT		0x01
+#define CMD_GET_DEVICE_NAME_TYPE_GET_DEVICE_NAME	0x11
+#define CMD_GET_DEVICE_NAME_TYPE_GET_TYPE		0x21
+
+static int hidpp_devicenametype_get_count(struct hidpp_device *hidpp,
+	u8 feature_index, u8 *nameLength)
+{
+	struct hidpp_report response;
+	int ret;
+
+	ret = hidpp_send_fap_command_sync(hidpp, feature_index,
+		CMD_GET_DEVICE_NAME_TYPE_GET_COUNT, NULL, 0, &response);
+
+	if (ret > 0) {
+		hid_err(hidpp->hid_dev, "%s: received protocol error 0x%02x\n",
+			__func__, ret);
+		return -EPROTO;
+	}
+	if (ret)
+		return ret;
+
+	*nameLength = response.fap.params[0];
+
+	return ret;
+}
+
+static int hidpp_devicenametype_get_device_name(struct hidpp_device *hidpp,
+	u8 feature_index, u8 char_index, char *device_name, int len_buf)
+{
+	struct hidpp_report response;
+	int ret, i;
+	int count;
+
+	ret = hidpp_send_fap_command_sync(hidpp, feature_index,
+		CMD_GET_DEVICE_NAME_TYPE_GET_DEVICE_NAME, &char_index, 1,
+		&response);
+
+	if (ret > 0) {
+		hid_err(hidpp->hid_dev, "%s: received protocol error 0x%02x\n",
+			__func__, ret);
+		return -EPROTO;
+	}
+	if (ret)
+		return ret;
+
+	if (response.report_id == REPORT_ID_HIDPP_LONG)
+		count = HIDPP_REPORT_LONG_LENGTH - 4;
+	else
+		count = HIDPP_REPORT_SHORT_LENGTH - 4;
+
+	if (len_buf < count)
+		count = len_buf;
+
+	for (i = 0; i < count; i++)
+		device_name[i] = response.fap.params[i];
+
+	return count;
+}
+
+static char *hidpp_get_device_name(struct hidpp_device *hidpp)
+{
+	u8 feature_type;
+	u8 feature_index;
+	u8 __name_length;
+	char *name;
+	unsigned index = 0;
+	int ret;
+
+	ret = hidpp_root_get_feature(hidpp, HIDPP_PAGE_GET_DEVICE_NAME_TYPE,
+		&feature_index, &feature_type);
+	if (ret)
+		return NULL;
+
+	ret = hidpp_devicenametype_get_count(hidpp, feature_index,
+		&__name_length);
+	if (ret)
+		return NULL;
+
+	name = kzalloc(__name_length + 1, GFP_KERNEL);
+	if (!name)
+		return NULL;
+
+	while (index < __name_length) {
+		ret = hidpp_devicenametype_get_device_name(hidpp,
+			feature_index, index, name + index,
+			__name_length - index);
+		if (ret <= 0) {
+			kfree(name);
+			return NULL;
+		}
+		index += ret;
+	}
+
+	/* include the terminating '\0' */
+	hidpp_prefix_name(&name, __name_length + 1);
+
+	return name;
+}
+
+/* -------------------------------------------------------------------------- */
+/* 0x6100: TouchPadRawXY                                                      */
+/* -------------------------------------------------------------------------- */
+
+#define HIDPP_PAGE_TOUCHPAD_RAW_XY			0x6100
+
+#define CMD_TOUCHPAD_GET_RAW_INFO			0x01
+#define CMD_TOUCHPAD_SET_RAW_REPORT_STATE		0x21
+
+#define EVENT_TOUCHPAD_RAW_XY				0x00
+
+#define TOUCHPAD_RAW_XY_ORIGIN_LOWER_LEFT		0x01
+#define TOUCHPAD_RAW_XY_ORIGIN_UPPER_LEFT		0x03
+
+struct hidpp_touchpad_raw_info {
+	u16 x_size;
+	u16 y_size;
+	u8 z_range;
+	u8 area_range;
+	u8 timestamp_unit;
+	u8 maxcontacts;
+	u8 origin;
+	u16 res;
+};
+
+struct hidpp_touchpad_raw_xy_finger {
+	u8 contact_type;
+	u8 contact_status;
+	u16 x;
+	u16 y;
+	u8 z;
+	u8 area;
+	u8 finger_id;
+};
+
+struct hidpp_touchpad_raw_xy {
+	u16 timestamp;
+	struct hidpp_touchpad_raw_xy_finger fingers[2];
+	u8 spurious_flag;
+	u8 end_of_frame;
+	u8 finger_count;
+	u8 button;
+};
+
+static int hidpp_touchpad_get_raw_info(struct hidpp_device *hidpp,
+	u8 feature_index, struct hidpp_touchpad_raw_info *raw_info)
+{
+	struct hidpp_report response;
+	int ret;
+	u8 *params = (u8 *)response.fap.params;
+
+	ret = hidpp_send_fap_command_sync(hidpp, feature_index,
+		CMD_TOUCHPAD_GET_RAW_INFO, NULL, 0, &response);
+
+	if (ret > 0) {
+		hid_err(hidpp->hid_dev, "%s: received protocol error 0x%02x\n",
+			__func__, ret);
+		return -EPROTO;
+	}
+	if (ret)
+		return ret;
+
+	raw_info->x_size = get_unaligned_be16(&params[0]);
+	raw_info->y_size = get_unaligned_be16(&params[2]);
+	raw_info->z_range = params[4];
+	raw_info->area_range = params[5];
+	raw_info->maxcontacts = params[7];
+	raw_info->origin = params[8];
+	/* res is given in unit per inch */
+	raw_info->res = get_unaligned_be16(&params[13]) * 2 / 51;
+
+	return ret;
+}
+
+static int hidpp_touchpad_set_raw_report_state(struct hidpp_device *hidpp_dev,
+		u8 feature_index, bool send_raw_reports,
+		bool sensor_enhanced_settings)
+{
+	struct hidpp_report response;
+
+	/*
+	 * Params:
+	 *   bit 0 - enable raw
+	 *   bit 1 - 16bit Z, no area
+	 *   bit 2 - enhanced sensitivity
+	 *   bit 3 - width, height (4 bits each) instead of area
+	 *   bit 4 - send raw + gestures (degrades smoothness)
+	 *   remaining bits - reserved
+	 */
+	u8 params = send_raw_reports | (sensor_enhanced_settings << 2);
+
+	return hidpp_send_fap_command_sync(hidpp_dev, feature_index,
+		CMD_TOUCHPAD_SET_RAW_REPORT_STATE, &params, 1, &response);
+}
+
+static void hidpp_touchpad_touch_event(u8 *data,
+	struct hidpp_touchpad_raw_xy_finger *finger)
+{
+	u8 x_m = data[0] << 2;
+	u8 y_m = data[2] << 2;
+
+	finger->x = x_m << 6 | data[1];
+	finger->y = y_m << 6 | data[3];
+
+	finger->contact_type = data[0] >> 6;
+	finger->contact_status = data[2] >> 6;
+
+	finger->z = data[4];
+	finger->area = data[5];
+	finger->finger_id = data[6] >> 4;
+}
+
+static void hidpp_touchpad_raw_xy_event(struct hidpp_device *hidpp_dev,
+		u8 *data, struct hidpp_touchpad_raw_xy *raw_xy)
+{
+	memset(raw_xy, 0, sizeof(struct hidpp_touchpad_raw_xy));
+	raw_xy->end_of_frame = data[8] & 0x01;
+	raw_xy->spurious_flag = (data[8] >> 1) & 0x01;
+	raw_xy->finger_count = data[15] & 0x0f;
+	raw_xy->button = (data[8] >> 2) & 0x01;
+
+	if (raw_xy->finger_count) {
+		hidpp_touchpad_touch_event(&data[2], &raw_xy->fingers[0]);
+		hidpp_touchpad_touch_event(&data[9], &raw_xy->fingers[1]);
+	}
+}
+
+/* ************************************************************************** */
+/*                                                                            */
+/* Device Support                                                             */
+/*                                                                            */
+/* ************************************************************************** */
+
+/* -------------------------------------------------------------------------- */
+/* Touchpad HID++ devices                                                     */
+/* -------------------------------------------------------------------------- */
+
+#define WTP_MANUAL_RESOLUTION				39
+
+struct wtp_data {
+	struct input_dev *input;
+	u16 x_size, y_size;
+	u8 finger_count;
+	u8 mt_feature_index;
+	u8 button_feature_index;
+	u8 maxcontacts;
+	bool flip_y;
+	unsigned int resolution;
+};
+
+static int wtp_input_mapping(struct hid_device *hdev, struct hid_input *hi,
+		struct hid_field *field, struct hid_usage *usage,
+		unsigned long **bit, int *max)
+{
+	struct hidpp_device *hidpp = hid_get_drvdata(hdev);
+
+	if ((hidpp->quirks & HIDPP_QUIRK_MULTI_INPUT) &&
+	    (field->application == HID_GD_KEYBOARD))
+		return 0;
+
+	return -1;
+}
+
+static void wtp_populate_input(struct hidpp_device *hidpp,
+		struct input_dev *input_dev, bool origin_is_hid_core)
+{
+	struct wtp_data *wd = hidpp->private_data;
+
+	if ((hidpp->quirks & HIDPP_QUIRK_MULTI_INPUT) && origin_is_hid_core)
+		/* this is the generic hid-input call */
+		return;
+
+	__set_bit(EV_ABS, input_dev->evbit);
+	__set_bit(EV_KEY, input_dev->evbit);
+	__clear_bit(EV_REL, input_dev->evbit);
+	__clear_bit(EV_LED, input_dev->evbit);
+
+	input_set_abs_params(input_dev, ABS_MT_POSITION_X, 0, wd->x_size, 0, 0);
+	input_abs_set_res(input_dev, ABS_MT_POSITION_X, wd->resolution);
+	input_set_abs_params(input_dev, ABS_MT_POSITION_Y, 0, wd->y_size, 0, 0);
+	input_abs_set_res(input_dev, ABS_MT_POSITION_Y, wd->resolution);
+
+	/* Max pressure is not given by the devices, pick one */
+	input_set_abs_params(input_dev, ABS_MT_PRESSURE, 0, 50, 0, 0);
+
+	input_set_capability(input_dev, EV_KEY, BTN_LEFT);
+
+	if (hidpp->quirks & HIDPP_QUIRK_WTP_PHYSICAL_BUTTONS)
+		input_set_capability(input_dev, EV_KEY, BTN_RIGHT);
+	else
+		__set_bit(INPUT_PROP_BUTTONPAD, input_dev->propbit);
+
+	input_mt_init_slots(input_dev, wd->maxcontacts, INPUT_MT_POINTER |
+		INPUT_MT_DROP_UNUSED);
+
+	wd->input = input_dev;
+}
+
+static void wtp_touch_event(struct wtp_data *wd,
+	struct hidpp_touchpad_raw_xy_finger *touch_report)
+{
+	int slot;
+
+	if (!touch_report->finger_id || touch_report->contact_type)
+		/* no actual data */
+		return;
+
+	slot = input_mt_get_slot_by_key(wd->input, touch_report->finger_id);
+
+	input_mt_slot(wd->input, slot);
+	input_mt_report_slot_state(wd->input, MT_TOOL_FINGER,
+					touch_report->contact_status);
+	if (touch_report->contact_status) {
+		input_event(wd->input, EV_ABS, ABS_MT_POSITION_X,
+				touch_report->x);
+		input_event(wd->input, EV_ABS, ABS_MT_POSITION_Y,
+				wd->flip_y ? wd->y_size - touch_report->y :
+					     touch_report->y);
+		input_event(wd->input, EV_ABS, ABS_MT_PRESSURE,
+				touch_report->area);
+	}
+}
+
+static void wtp_send_raw_xy_event(struct hidpp_device *hidpp,
+		struct hidpp_touchpad_raw_xy *raw)
+{
+	struct wtp_data *wd = hidpp->private_data;
+	int i;
+
+	for (i = 0; i < 2; i++)
+		wtp_touch_event(wd, &(raw->fingers[i]));
+
+	if (raw->end_of_frame &&
+	    !(hidpp->quirks & HIDPP_QUIRK_WTP_PHYSICAL_BUTTONS))
+		input_event(wd->input, EV_KEY, BTN_LEFT, raw->button);
+
+	if (raw->end_of_frame || raw->finger_count <= 2) {
+		input_mt_sync_frame(wd->input);
+		input_sync(wd->input);
+	}
+}
+
+static int wtp_mouse_raw_xy_event(struct hidpp_device *hidpp, u8 *data)
+{
+	struct wtp_data *wd = hidpp->private_data;
+	u8 c1_area = ((data[7] & 0xf) * (data[7] & 0xf) +
+		      (data[7] >> 4) * (data[7] >> 4)) / 2;
+	u8 c2_area = ((data[13] & 0xf) * (data[13] & 0xf) +
+		      (data[13] >> 4) * (data[13] >> 4)) / 2;
+	struct hidpp_touchpad_raw_xy raw = {
+		.timestamp = data[1],
+		.fingers = {
+			{
+				.contact_type = 0,
+				.contact_status = !!data[7],
+				.x = get_unaligned_le16(&data[3]),
+				.y = get_unaligned_le16(&data[5]),
+				.z = c1_area,
+				.area = c1_area,
+				.finger_id = data[2],
+			}, {
+				.contact_type = 0,
+				.contact_status = !!data[13],
+				.x = get_unaligned_le16(&data[9]),
+				.y = get_unaligned_le16(&data[11]),
+				.z = c2_area,
+				.area = c2_area,
+				.finger_id = data[8],
+			}
+		},
+		.finger_count = wd->maxcontacts,
+		.spurious_flag = 0,
+		.end_of_frame = (data[0] >> 7) == 0,
+		.button = data[0] & 0x01,
+	};
+
+	wtp_send_raw_xy_event(hidpp, &raw);
+
+	return 1;
+}
+
+static int wtp_raw_event(struct hid_device *hdev, u8 *data, int size)
+{
+	struct hidpp_device *hidpp = hid_get_drvdata(hdev);
+	struct wtp_data *wd = hidpp->private_data;
+	struct hidpp_report *report = (struct hidpp_report *)data;
+	struct hidpp_touchpad_raw_xy raw;
+
+	if (!wd || !wd->input)
+		return 1;
+
+	switch (data[0]) {
+	case 0x02:
+		if (size < 2) {
+			hid_err(hdev, "Received HID report of bad size (%d)",
+				size);
+			return 1;
+		}
+		if (hidpp->quirks & HIDPP_QUIRK_WTP_PHYSICAL_BUTTONS) {
+			input_event(wd->input, EV_KEY, BTN_LEFT,
+					!!(data[1] & 0x01));
+			input_event(wd->input, EV_KEY, BTN_RIGHT,
+					!!(data[1] & 0x02));
+			input_sync(wd->input);
+			return 0;
+		} else {
+			if (size < 21)
+				return 1;
+			return wtp_mouse_raw_xy_event(hidpp, &data[7]);
+		}
+	case REPORT_ID_HIDPP_LONG:
+		/* size is already checked in hidpp_raw_event. */
+		if ((report->fap.feature_index != wd->mt_feature_index) ||
+		    (report->fap.funcindex_clientid != EVENT_TOUCHPAD_RAW_XY))
+			return 1;
+		hidpp_touchpad_raw_xy_event(hidpp, data + 4, &raw);
+
+		wtp_send_raw_xy_event(hidpp, &raw);
+		return 0;
+	}
+
+	return 0;
+}
+
+static int wtp_get_config(struct hidpp_device *hidpp)
+{
+	struct wtp_data *wd = hidpp->private_data;
+	struct hidpp_touchpad_raw_info raw_info = {0};
+	u8 feature_type;
+	int ret;
+
+	ret = hidpp_root_get_feature(hidpp, HIDPP_PAGE_TOUCHPAD_RAW_XY,
+		&wd->mt_feature_index, &feature_type);
+	if (ret)
+		/* means that the device is not powered up */
+		return ret;
+
+	ret = hidpp_touchpad_get_raw_info(hidpp, wd->mt_feature_index,
+		&raw_info);
+	if (ret)
+		return ret;
+
+	wd->x_size = raw_info.x_size;
+	wd->y_size = raw_info.y_size;
+	wd->maxcontacts = raw_info.maxcontacts;
+	wd->flip_y = raw_info.origin == TOUCHPAD_RAW_XY_ORIGIN_LOWER_LEFT;
+	wd->resolution = raw_info.res;
+	if (!wd->resolution)
+		wd->resolution = WTP_MANUAL_RESOLUTION;
+
+	return 0;
+}
+
+static int wtp_allocate(struct hid_device *hdev, const struct hid_device_id *id)
+{
+	struct hidpp_device *hidpp = hid_get_drvdata(hdev);
+	struct wtp_data *wd;
+
+	wd = devm_kzalloc(&hdev->dev, sizeof(struct wtp_data),
+			GFP_KERNEL);
+	if (!wd)
+		return -ENOMEM;
+
+	hidpp->private_data = wd;
+
+	return 0;
+};
+
+static void wtp_connect(struct hid_device *hdev, bool connected)
+{
+	struct hidpp_device *hidpp = hid_get_drvdata(hdev);
+	struct wtp_data *wd = hidpp->private_data;
+	int ret;
+
+	if (!connected)
+		return;
+
+	if (!wd->x_size) {
+		ret = wtp_get_config(hidpp);
+		if (ret) {
+			hid_err(hdev, "Can not get wtp config: %d\n", ret);
+			return;
+		}
+	}
+
+	hidpp_touchpad_set_raw_report_state(hidpp, wd->mt_feature_index,
+			true, true);
+}
+
+/* -------------------------------------------------------------------------- */
+/* Generic HID++ devices                                                      */
+/* -------------------------------------------------------------------------- */
+
+static int hidpp_input_mapping(struct hid_device *hdev, struct hid_input *hi,
+		struct hid_field *field, struct hid_usage *usage,
+		unsigned long **bit, int *max)
+{
+	struct hidpp_device *hidpp = hid_get_drvdata(hdev);
+
+	if (hidpp->quirks & HIDPP_QUIRK_CLASS_WTP)
+		return wtp_input_mapping(hdev, hi, field, usage, bit, max);
+
+	return 0;
+}
+
+static void hidpp_populate_input(struct hidpp_device *hidpp,
+		struct input_dev *input, bool origin_is_hid_core)
+{
+	if (hidpp->quirks & HIDPP_QUIRK_CLASS_WTP)
+		wtp_populate_input(hidpp, input, origin_is_hid_core);
+}
+
+static void hidpp_input_configured(struct hid_device *hdev,
+				struct hid_input *hidinput)
+{
+	struct hidpp_device *hidpp = hid_get_drvdata(hdev);
+	struct input_dev *input = hidinput->input;
+
+	hidpp_populate_input(hidpp, input, true);
+}
+
+static int hidpp_raw_hidpp_event(struct hidpp_device *hidpp, u8 *data,
+		int size)
+{
+	struct hidpp_report *question = hidpp->send_receive_buf;
+	struct hidpp_report *answer = hidpp->send_receive_buf;
+	struct hidpp_report *report = (struct hidpp_report *)data;
+
+	/*
+	 * If the mutex is locked then we have a pending answer from a
+	 * previoulsly sent command
+	 */
+	if (unlikely(mutex_is_locked(&hidpp->send_mutex))) {
+		/*
+		 * Check for a correct hidpp20 answer or the corresponding
+		 * error
+		 */
+		if (hidpp_match_answer(question, report) ||
+				hidpp_match_error(question, report)) {
+			*answer = *report;
+			hidpp->answer_available = true;
+			wake_up(&hidpp->wait);
+			/*
+			 * This was an answer to a command that this driver sent
+			 * We return 1 to hid-core to avoid forwarding the
+			 * command upstream as it has been treated by the driver
+			 */
+
+			return 1;
+		}
+	}
+
+	if (unlikely(hidpp_report_is_connect_event(report))) {
+		atomic_set(&hidpp->connected,
+				!(report->rap.params[0] & (1 << 6)));
+		if ((hidpp->quirks & HIDPP_QUIRK_DELAYED_INIT) &&
+		    (schedule_work(&hidpp->work) == 0))
+			dbg_hid("%s: connect event already queued\n", __func__);
+		return 1;
+	}
+
+	if (hidpp->quirks & HIDPP_QUIRK_CLASS_WTP)
+		return wtp_raw_event(hidpp->hid_dev, data, size);
+
+	return 0;
+}
+
+static int hidpp_raw_event(struct hid_device *hdev, struct hid_report *report,
+		u8 *data, int size)
+{
+	struct hidpp_device *hidpp = hid_get_drvdata(hdev);
+
+	switch (data[0]) {
+	case REPORT_ID_HIDPP_LONG:
+		if (size != HIDPP_REPORT_LONG_LENGTH) {
+			hid_err(hdev, "received hid++ report of bad size (%d)",
+				size);
+			return 1;
+		}
+		return hidpp_raw_hidpp_event(hidpp, data, size);
+	case REPORT_ID_HIDPP_SHORT:
+		if (size != HIDPP_REPORT_SHORT_LENGTH) {
+			hid_err(hdev, "received hid++ report of bad size (%d)",
+				size);
+			return 1;
+		}
+		return hidpp_raw_hidpp_event(hidpp, data, size);
+	}
+
+	if (hidpp->quirks & HIDPP_QUIRK_CLASS_WTP)
+		return wtp_raw_event(hdev, data, size);
+
+	return 0;
+}
+
+static void hidpp_overwrite_name(struct hid_device *hdev, bool use_unifying)
+{
+	struct hidpp_device *hidpp = hid_get_drvdata(hdev);
+	char *name;
+
+	if (use_unifying)
+		/*
+		 * the device is connected through an Unifying receiver, and
+		 * might not be already connected.
+		 * Ask the receiver for its name.
+		 */
+		name = hidpp_get_unifying_name(hidpp);
+	else
+		name = hidpp_get_device_name(hidpp);
+
+	if (!name)
+		hid_err(hdev, "unable to retrieve the name of the device");
+	else
+		snprintf(hdev->name, sizeof(hdev->name), "%s", name);
+
+	kfree(name);
+}
+
+static int hidpp_input_open(struct input_dev *dev)
+{
+	struct hid_device *hid = input_get_drvdata(dev);
+
+	return hid_hw_open(hid);
+}
+
+static void hidpp_input_close(struct input_dev *dev)
+{
+	struct hid_device *hid = input_get_drvdata(dev);
+
+	hid_hw_close(hid);
+}
+
+static struct input_dev *hidpp_allocate_input(struct hid_device *hdev)
+{
+	struct input_dev *input_dev = devm_input_allocate_device(&hdev->dev);
+
+	if (!input_dev)
+		return NULL;
+
+	input_set_drvdata(input_dev, hdev);
+	input_dev->open = hidpp_input_open;
+	input_dev->close = hidpp_input_close;
+
+	input_dev->name = hdev->name;
+	input_dev->phys = hdev->phys;
+	input_dev->uniq = hdev->uniq;
+	input_dev->id.bustype = hdev->bus;
+	input_dev->id.vendor  = hdev->vendor;
+	input_dev->id.product = hdev->product;
+	input_dev->id.version = hdev->version;
+	input_dev->dev.parent = &hdev->dev;
+
+	return input_dev;
+}
+
+static void hidpp_connect_event(struct hidpp_device *hidpp)
+{
+	struct hid_device *hdev = hidpp->hid_dev;
+	int ret = 0;
+	bool connected = atomic_read(&hidpp->connected);
+	struct input_dev *input;
+	char *name, *devm_name;
+
+	if (hidpp->quirks & HIDPP_QUIRK_CLASS_WTP)
+		wtp_connect(hdev, connected);
+
+	if (!connected || hidpp->delayed_input)
+		return;
+
+	if (!hidpp->protocol_major) {
+		ret = !hidpp_is_connected(hidpp);
+		if (ret) {
+			hid_err(hdev, "Can not get the protocol version.\n");
+			return;
+		}
+	}
+
+	/* the device is already connected, we can ask for its name and
+	 * protocol */
+	hid_info(hdev, "HID++ %u.%u device connected.\n",
+		 hidpp->protocol_major, hidpp->protocol_minor);
+
+	input = hidpp_allocate_input(hdev);
+	if (!input) {
+		hid_err(hdev, "cannot allocate new input device: %d\n", ret);
+		return;
+	}
+
+	name = hidpp_get_device_name(hidpp);
+	if (!name) {
+		hid_err(hdev, "unable to retrieve the name of the device");
+	} else {
+		devm_name = devm_kasprintf(&hdev->dev, GFP_KERNEL, "%s", name);
+		if (devm_name)
+			input->name = devm_name;
+		kfree(name);
+	}
+
+	hidpp_populate_input(hidpp, input, false);
+
+	ret = input_register_device(input);
+	if (ret)
+		input_free_device(input);
+
+	hidpp->delayed_input = input;
+}
+
+static int hidpp_probe(struct hid_device *hdev, const struct hid_device_id *id)
+{
+	struct hidpp_device *hidpp;
+	int ret;
+	bool connected;
+	unsigned int connect_mask = HID_CONNECT_DEFAULT;
+
+	hidpp = devm_kzalloc(&hdev->dev, sizeof(struct hidpp_device),
+			GFP_KERNEL);
+	if (!hidpp)
+		return -ENOMEM;
+
+	hidpp->hid_dev = hdev;
+	hid_set_drvdata(hdev, hidpp);
+
+	hidpp->quirks = id->driver_data;
+
+	if (hidpp->quirks & HIDPP_QUIRK_CLASS_WTP) {
+		ret = wtp_allocate(hdev, id);
+		if (ret)
+			goto wtp_allocate_fail;
+	}
+
+	INIT_WORK(&hidpp->work, delayed_work_cb);
+	mutex_init(&hidpp->send_mutex);
+	init_waitqueue_head(&hidpp->wait);
+
+	ret = hid_parse(hdev);
+	if (ret) {
+		hid_err(hdev, "%s:parse failed\n", __func__);
+		goto hid_parse_fail;
+	}
+
+	/* Allow incoming packets */
+	hid_device_io_start(hdev);
+
+	connected = hidpp_is_connected(hidpp);
+	if (id->group != HID_GROUP_LOGITECH_DJ_DEVICE) {
+		if (!connected) {
+			hid_err(hdev, "Device not connected");
+			hid_device_io_stop(hdev);
+			goto hid_parse_fail;
+		}
+
+		hid_info(hdev, "HID++ %u.%u device connected.\n",
+			 hidpp->protocol_major, hidpp->protocol_minor);
+	}
+
+	hidpp_overwrite_name(hdev, id->group == HID_GROUP_LOGITECH_DJ_DEVICE);
+	atomic_set(&hidpp->connected, connected);
+
+	if (connected && (hidpp->quirks & HIDPP_QUIRK_CLASS_WTP)) {
+		ret = wtp_get_config(hidpp);
+		if (ret)
+			goto hid_parse_fail;
+	}
+
+	/* Block incoming packets */
+	hid_device_io_stop(hdev);
+
+	if (hidpp->quirks & HIDPP_QUIRK_DELAYED_INIT)
+		connect_mask &= ~HID_CONNECT_HIDINPUT;
+
+	/* Re-enable hidinput for multi-input devices */
+	if (hidpp->quirks & HIDPP_QUIRK_MULTI_INPUT)
+		connect_mask |= HID_CONNECT_HIDINPUT;
+
+	ret = hid_hw_start(hdev, connect_mask);
+	if (ret) {
+		hid_err(hdev, "%s:hid_hw_start returned error\n", __func__);
+		goto hid_hw_start_fail;
+	}
+
+	if (hidpp->quirks & HIDPP_QUIRK_DELAYED_INIT) {
+		/* Allow incoming packets */
+		hid_device_io_start(hdev);
+
+		hidpp_connect_event(hidpp);
+	}
+
+	return ret;
+
+hid_hw_start_fail:
+hid_parse_fail:
+	cancel_work_sync(&hidpp->work);
+	mutex_destroy(&hidpp->send_mutex);
+wtp_allocate_fail:
+	hid_set_drvdata(hdev, NULL);
+	return ret;
+}
+
+static void hidpp_remove(struct hid_device *hdev)
+{
+	struct hidpp_device *hidpp = hid_get_drvdata(hdev);
+
+	cancel_work_sync(&hidpp->work);
+	mutex_destroy(&hidpp->send_mutex);
+	hid_hw_stop(hdev);
+}
+
+static const struct hid_device_id hidpp_devices[] = {
+	{ /* wireless touchpad */
+	  HID_DEVICE(BUS_USB, HID_GROUP_LOGITECH_DJ_DEVICE,
+		USB_VENDOR_ID_LOGITECH, 0x4011),
+	  .driver_data = HIDPP_QUIRK_CLASS_WTP | HIDPP_QUIRK_DELAYED_INIT |
+			 HIDPP_QUIRK_WTP_PHYSICAL_BUTTONS },
+	{ /* wireless touchpad T650 */
+	  HID_DEVICE(BUS_USB, HID_GROUP_LOGITECH_DJ_DEVICE,
+		USB_VENDOR_ID_LOGITECH, 0x4101),
+	  .driver_data = HIDPP_QUIRK_CLASS_WTP | HIDPP_QUIRK_DELAYED_INIT },
+	{ /* wireless touchpad T651 */
+	  HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_LOGITECH,
+		USB_DEVICE_ID_LOGITECH_T651),
+	  .driver_data = HIDPP_QUIRK_CLASS_WTP },
+	{ /* Keyboard TK820 */
+	  HID_DEVICE(BUS_USB, HID_GROUP_LOGITECH_DJ_DEVICE,
+		USB_VENDOR_ID_LOGITECH, 0x4102),
+	  .driver_data = HIDPP_QUIRK_DELAYED_INIT | HIDPP_QUIRK_MULTI_INPUT |
+			 HIDPP_QUIRK_CLASS_WTP },
+
+	{ HID_DEVICE(BUS_USB, HID_GROUP_LOGITECH_DJ_DEVICE,
+		USB_VENDOR_ID_LOGITECH, HID_ANY_ID)},
+	{}
+};
+
+MODULE_DEVICE_TABLE(hid, hidpp_devices);
+
+static struct hid_driver hidpp_driver = {
+	.name = "logitech-hidpp-device",
+	.id_table = hidpp_devices,
+	.probe = hidpp_probe,
+	.remove = hidpp_remove,
+	.raw_event = hidpp_raw_event,
+	.input_configured = hidpp_input_configured,
+	.input_mapping = hidpp_input_mapping,
+};
+
+module_hid_driver(hidpp_driver);
diff --git a/drivers/hid/hid-microsoft.c b/drivers/hid/hid-microsoft.c
index 8ba17a946f2a..cacda43f6a6f 100644
--- a/drivers/hid/hid-microsoft.c
+++ b/drivers/hid/hid-microsoft.c
@@ -274,6 +274,8 @@ static const struct hid_device_id ms_devices[] = {
 		.driver_data = MS_NOGET },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_MICROSOFT, USB_DEVICE_ID_MS_COMFORT_MOUSE_4500),
 		.driver_data = MS_DUPLICATE_USAGES },
+	{ HID_USB_DEVICE(USB_VENDOR_ID_MICROSOFT, USB_DEVICE_ID_MS_TYPE_COVER_3),
+		.driver_data = MS_HIDINPUT },
 
 	{ HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_MICROSOFT, USB_DEVICE_ID_MS_PRESENTER_8K_BT),
 		.driver_data = MS_PRESENTER },
diff --git a/drivers/hid/hid-multitouch.c b/drivers/hid/hid-multitouch.c
index 51e25b9407f2..f65e78b46999 100644
--- a/drivers/hid/hid-multitouch.c
+++ b/drivers/hid/hid-multitouch.c
@@ -67,6 +67,7 @@ MODULE_LICENSE("GPL");
 #define MT_QUIRK_IGNORE_DUPLICATES	(1 << 10)
 #define MT_QUIRK_HOVERING		(1 << 11)
 #define MT_QUIRK_CONTACT_CNT_ACCURATE	(1 << 12)
+#define MT_QUIRK_FORCE_GET_FEATURE	(1 << 13)
 
 #define MT_INPUTMODE_TOUCHSCREEN	0x02
 #define MT_INPUTMODE_TOUCHPAD		0x03
@@ -150,6 +151,7 @@ static void mt_post_parse(struct mt_device *td);
 #define MT_CLS_FLATFROG				0x0107
 #define MT_CLS_GENERALTOUCH_TWOFINGERS		0x0108
 #define MT_CLS_GENERALTOUCH_PWT_TENFINGERS	0x0109
+#define MT_CLS_VTL				0x0110
 
 #define MT_DEFAULT_MAXCONTACT	10
 #define MT_MAX_MAXCONTACT	250
@@ -255,6 +257,11 @@ static struct mt_class mt_classes[] = {
 		.sn_move = 2048,
 		.maxcontacts = 40,
 	},
+	{ .name = MT_CLS_VTL,
+		.quirks = MT_QUIRK_ALWAYS_VALID |
+			MT_QUIRK_CONTACT_CNT_ACCURATE |
+			MT_QUIRK_FORCE_GET_FEATURE,
+	},
 	{ }
 };
 
@@ -809,6 +816,9 @@ static void mt_set_input_mode(struct hid_device *hdev)
 	struct mt_device *td = hid_get_drvdata(hdev);
 	struct hid_report *r;
 	struct hid_report_enum *re;
+	struct mt_class *cls = &td->mtclass;
+	char *buf;
+	int report_len;
 
 	if (td->inputmode < 0)
 		return;
@@ -816,6 +826,18 @@ static void mt_set_input_mode(struct hid_device *hdev)
 	re = &(hdev->report_enum[HID_FEATURE_REPORT]);
 	r = re->report_id_hash[td->inputmode];
 	if (r) {
+		if (cls->quirks & MT_QUIRK_FORCE_GET_FEATURE) {
+			report_len = hid_report_len(r);
+			buf = hid_alloc_report_buf(r, GFP_KERNEL);
+			if (!buf) {
+				hid_err(hdev, "failed to allocate buffer for report\n");
+				return;
+			}
+			hid_hw_raw_request(hdev, r->id, buf, report_len,
+					   HID_FEATURE_REPORT,
+					   HID_REQ_GET_REPORT);
+			kfree(buf);
+		}
 		r->field[0]->value[td->inputmode_index] = td->inputmode_value;
 		hid_hw_request(hdev, r, HID_REQ_SET_REPORT);
 	}
@@ -1281,6 +1303,11 @@ static const struct hid_device_id mt_devices[] = {
 		MT_USB_DEVICE(USB_VENDOR_ID_UNITEC,
 			USB_DEVICE_ID_UNITEC_USB_TOUCH_0A19) },
 
+	/* VTL panels */
+	{ .driver_data = MT_CLS_VTL,
+		MT_USB_DEVICE(USB_VENDOR_ID_VTL,
+			USB_DEVICE_ID_VTL_MULTITOUCH_FF3F) },
+
 	/* Wistron panels */
 	{ .driver_data = MT_CLS_NSMU,
 		MT_USB_DEVICE(USB_VENDOR_ID_WISTRON,
diff --git a/drivers/hid/hid-penmount.c b/drivers/hid/hid-penmount.c
new file mode 100644
index 000000000000..c11dce85cd18
--- /dev/null
+++ b/drivers/hid/hid-penmount.c
@@ -0,0 +1,49 @@
+/*
+ *  HID driver for PenMount touchscreens
+ *
+ *  Copyright (c) 2014 Christian Gmeiner <christian.gmeiner <at> gmail.com>
+ *
+ *  based on hid-penmount copyrighted by
+ *    PenMount Touch Solutions <penmount <at> seed.net.tw>
+ */
+
+/*
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the Free
+ * Software Foundation; either version 2 of the License, or (at your option)
+ * any later version.
+ */
+
+#include <linux/module.h>
+#include <linux/hid.h>
+#include "hid-ids.h"
+
+static int penmount_input_mapping(struct hid_device *hdev,
+		struct hid_input *hi, struct hid_field *field,
+		struct hid_usage *usage, unsigned long **bit, int *max)
+{
+	if ((usage->hid & HID_USAGE_PAGE) == HID_UP_BUTTON) {
+		hid_map_usage(hi, usage, bit, max, EV_KEY, BTN_TOUCH);
+		return 1;
+	}
+
+	return 0;
+}
+
+static const struct hid_device_id penmount_devices[] = {
+	{ HID_USB_DEVICE(USB_VENDOR_ID_PENMOUNT, USB_DEVICE_ID_PENMOUNT_6000) },
+	{ }
+};
+MODULE_DEVICE_TABLE(hid, penmount_devices);
+
+static struct hid_driver penmount_driver = {
+	.name = "hid-penmount",
+	.id_table = penmount_devices,
+	.input_mapping = penmount_input_mapping,
+};
+
+module_hid_driver(penmount_driver);
+
+MODULE_AUTHOR("Christian Gmeiner <christian.gmeiner@gmail.com>");
+MODULE_DESCRIPTION("PenMount HID TouchScreen driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/hid/hid-picolcd_core.c b/drivers/hid/hid-picolcd_core.c
index 020df3c2e8b4..c1b29a9eb41a 100644
--- a/drivers/hid/hid-picolcd_core.c
+++ b/drivers/hid/hid-picolcd_core.c
@@ -351,8 +351,8 @@ static int picolcd_raw_event(struct hid_device *hdev,
 		return 1;
 
 	if (size > 64) {
-		hid_warn(hdev, "invalid size value (%d) for picolcd raw event\n",
-				size);
+		hid_warn(hdev, "invalid size value (%d) for picolcd raw event (%d)\n",
+				size, report->id);
 		return 0;
 	}
 
diff --git a/drivers/hid/hid-plantronics.c b/drivers/hid/hid-plantronics.c
new file mode 100644
index 000000000000..2180e0789b76
--- /dev/null
+++ b/drivers/hid/hid-plantronics.c
@@ -0,0 +1,55 @@
+/*
+ *  Plantronics USB HID Driver
+ *
+ *  Copyright (c) 2014 JD Cole <jd.cole@plantronics.com>
+ *  Copyright (c) 2014 Terry Junge <terry.junge@plantronics.com>
+ */
+
+/*
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the Free
+ * Software Foundation; either version 2 of the License, or (at your option)
+ * any later version.
+ */
+
+#include "hid-ids.h"
+
+#include <linux/hid.h>
+#include <linux/module.h>
+
+static int plantronics_input_mapping(struct hid_device *hdev,
+				     struct hid_input *hi,
+				     struct hid_field *field,
+				     struct hid_usage *usage,
+				     unsigned long **bit, int *max)
+{
+	if (field->application == HID_CP_CONSUMERCONTROL
+	    && (usage->hid & HID_USAGE_PAGE) == HID_UP_CONSUMER) {
+		hid_dbg(hdev, "usage: %08x (appl: %08x) - defaulted\n",
+			 usage->hid, field->application);
+		return 0;
+	}
+
+	hid_dbg(hdev, "usage: %08x (appl: %08x) - ignored\n",
+		usage->hid, field->application);
+
+	return -1;
+}
+
+static const struct hid_device_id plantronics_devices[] = {
+	{ HID_USB_DEVICE(USB_VENDOR_ID_PLANTRONICS, HID_ANY_ID) },
+	{ }
+};
+MODULE_DEVICE_TABLE(hid, plantronics_devices);
+
+static struct hid_driver plantronics_driver = {
+	.name = "plantronics",
+	.id_table = plantronics_devices,
+	.input_mapping = plantronics_input_mapping,
+};
+module_hid_driver(plantronics_driver);
+
+MODULE_AUTHOR("JD Cole <jd.cole@plantronics.com>");
+MODULE_AUTHOR("Terry Junge <terry.junge@plantronics.com>");
+MODULE_DESCRIPTION("Plantronics USB HID Driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/hid/hid-rmi.c b/drivers/hid/hid-rmi.c
index 8389e8109218..b51200fe2f33 100644
--- a/drivers/hid/hid-rmi.c
+++ b/drivers/hid/hid-rmi.c
@@ -320,10 +320,7 @@ static int rmi_f11_input_event(struct hid_device *hdev, u8 irq, u8 *data,
 	int offset;
 	int i;
 
-	if (size < hdata->f11.report_size)
-		return 0;
-
-	if (!(irq & hdata->f11.irq_mask))
+	if (!(irq & hdata->f11.irq_mask) || size <= 0)
 		return 0;
 
 	offset = (hdata->max_fingers >> 2) + 1;
@@ -332,9 +329,19 @@ static int rmi_f11_input_event(struct hid_device *hdev, u8 irq, u8 *data,
 		int fs_bit_position = (i & 0x3) << 1;
 		int finger_state = (data[fs_byte_position] >> fs_bit_position) &
 					0x03;
+		int position = offset + 5 * i;
+
+		if (position + 5 > size) {
+			/* partial report, go on with what we received */
+			printk_once(KERN_WARNING
+				"%s %s: Detected incomplete finger report. Finger reports may occasionally get dropped on this platform.\n",
+				 dev_driver_string(&hdev->dev),
+				 dev_name(&hdev->dev));
+			hid_dbg(hdev, "Incomplete finger report\n");
+			break;
+		}
 
-		rmi_f11_process_touch(hdata, i, finger_state,
-				&data[offset + 5 * i]);
+		rmi_f11_process_touch(hdata, i, finger_state, &data[position]);
 	}
 	input_mt_sync_frame(hdata->input);
 	input_sync(hdata->input);
@@ -352,6 +359,11 @@ static int rmi_f30_input_event(struct hid_device *hdev, u8 irq, u8 *data,
 	if (!(irq & hdata->f30.irq_mask))
 		return 0;
 
+	if (size < (int)hdata->f30.report_size) {
+		hid_warn(hdev, "Click Button pressed, but the click data is missing\n");
+		return 0;
+	}
+
 	for (i = 0; i < hdata->gpio_led_count; i++) {
 		if (test_bit(i, &hdata->button_mask)) {
 			value = (data[i / 8] >> (i & 0x07)) & BIT(0);
@@ -412,9 +424,29 @@ static int rmi_read_data_event(struct hid_device *hdev, u8 *data, int size)
 	return 1;
 }
 
+static int rmi_check_sanity(struct hid_device *hdev, u8 *data, int size)
+{
+	int valid_size = size;
+	/*
+	 * On the Dell XPS 13 9333, the bus sometimes get confused and fills
+	 * the report with a sentinel value "ff". Synaptics told us that such
+	 * behavior does not comes from the touchpad itself, so we filter out
+	 * such reports here.
+	 */
+
+	while ((data[valid_size - 1] == 0xff) && valid_size > 0)
+		valid_size--;
+
+	return valid_size;
+}
+
 static int rmi_raw_event(struct hid_device *hdev,
 		struct hid_report *report, u8 *data, int size)
 {
+	size = rmi_check_sanity(hdev, data, size);
+	if (size < 2)
+		return 0;
+
 	switch (data[0]) {
 	case RMI_READ_DATA_REPORT_ID:
 		return rmi_read_data_event(hdev, data, size);
@@ -552,11 +584,15 @@ static int rmi_populate_f11(struct hid_device *hdev)
 	bool has_query10 = false;
 	bool has_query11;
 	bool has_query12;
+	bool has_query27;
+	bool has_query28;
+	bool has_query36 = false;
 	bool has_physical_props;
 	bool has_gestures;
 	bool has_rel;
+	bool has_data40 = false;
 	unsigned x_size, y_size;
-	u16 query12_offset;
+	u16 query_offset;
 
 	if (!data->f11.query_base_addr) {
 		hid_err(hdev, "No 2D sensor found, giving up.\n");
@@ -572,6 +608,8 @@ static int rmi_populate_f11(struct hid_device *hdev)
 	has_query9 = !!(buf[0] & BIT(3));
 	has_query11 = !!(buf[0] & BIT(4));
 	has_query12 = !!(buf[0] & BIT(5));
+	has_query27 = !!(buf[0] & BIT(6));
+	has_query28 = !!(buf[0] & BIT(7));
 
 	/* query 1 to get the max number of fingers */
 	ret = rmi_read(hdev, data->f11.query_base_addr + 1, buf);
@@ -594,43 +632,43 @@ static int rmi_populate_f11(struct hid_device *hdev)
 	has_rel = !!(buf[0] & BIT(3));
 	has_gestures = !!(buf[0] & BIT(5));
 
+	/*
+	 * At least 4 queries are guaranteed to be present in F11
+	 * +1 for query 5 which is present since absolute events are
+	 * reported and +1 for query 12.
+	 */
+	query_offset = 6;
+
+	if (has_rel)
+		++query_offset; /* query 6 is present */
+
 	if (has_gestures) {
 		/* query 8 to find out if query 10 exists */
-		ret = rmi_read(hdev, data->f11.query_base_addr + 8, buf);
+		ret = rmi_read(hdev,
+			data->f11.query_base_addr + query_offset + 1, buf);
 		if (ret) {
 			hid_err(hdev, "can not read gesture information: %d.\n",
 				ret);
 			return ret;
 		}
 		has_query10 = !!(buf[0] & BIT(2));
-	}
-
-	/*
-	 * At least 4 queries are guaranteed to be present in F11
-	 * +1 for query 5 which is present since absolute events are
-	 * reported and +1 for query 12.
-	 */
-	query12_offset = 6;
-
-	if (has_rel)
-		++query12_offset; /* query 6 is present */
 
-	if (has_gestures)
-		query12_offset += 2; /* query 7 and 8 are present */
+		query_offset += 2; /* query 7 and 8 are present */
+	}
 
 	if (has_query9)
-		++query12_offset;
+		++query_offset;
 
 	if (has_query10)
-		++query12_offset;
+		++query_offset;
 
 	if (has_query11)
-		++query12_offset;
+		++query_offset;
 
 	/* query 12 to know if the physical properties are reported */
 	if (has_query12) {
 		ret = rmi_read(hdev, data->f11.query_base_addr
-				+ query12_offset, buf);
+				+ query_offset, buf);
 		if (ret) {
 			hid_err(hdev, "can not get query 12: %d.\n", ret);
 			return ret;
@@ -638,9 +676,10 @@ static int rmi_populate_f11(struct hid_device *hdev)
 		has_physical_props = !!(buf[0] & BIT(5));
 
 		if (has_physical_props) {
+			query_offset += 1;
 			ret = rmi_read_block(hdev,
 					data->f11.query_base_addr
-						+ query12_offset + 1, buf, 4);
+						+ query_offset, buf, 4);
 			if (ret) {
 				hid_err(hdev, "can not read query 15-18: %d.\n",
 					ret);
@@ -655,9 +694,45 @@ static int rmi_populate_f11(struct hid_device *hdev)
 
 			hid_info(hdev, "%s: size in mm: %d x %d\n",
 				 __func__, data->x_size_mm, data->y_size_mm);
+
+			/*
+			 * query 15 - 18 contain the size of the sensor
+			 * and query 19 - 26 contain bezel dimensions
+			 */
+			query_offset += 12;
+		}
+	}
+
+	if (has_query27)
+		++query_offset;
+
+	if (has_query28) {
+		ret = rmi_read(hdev, data->f11.query_base_addr
+				+ query_offset, buf);
+		if (ret) {
+			hid_err(hdev, "can not get query 28: %d.\n", ret);
+			return ret;
 		}
+
+		has_query36 = !!(buf[0] & BIT(6));
 	}
 
+	if (has_query36) {
+		query_offset += 2;
+		ret = rmi_read(hdev, data->f11.query_base_addr
+				+ query_offset, buf);
+		if (ret) {
+			hid_err(hdev, "can not get query 36: %d.\n", ret);
+			return ret;
+		}
+
+		has_data40 = !!(buf[0] & BIT(5));
+	}
+
+
+	if (has_data40)
+		data->f11.report_size += data->max_fingers * 2;
+
 	/*
 	 * retrieve the ctrl registers
 	 * the ctrl register has a size of 20 but a fw bug split it into 16 + 4,
diff --git a/drivers/hid/hid-roccat-kone.c b/drivers/hid/hid-roccat-kone.c
index 6101816a7ddd..c29265055ac1 100644
--- a/drivers/hid/hid-roccat-kone.c
+++ b/drivers/hid/hid-roccat-kone.c
@@ -46,6 +46,7 @@ static void kone_profile_activated(struct kone_device *kone, uint new_profile)
 static void kone_profile_report(struct kone_device *kone, uint new_profile)
 {
 	struct kone_roccat_report roccat_report;
+
 	roccat_report.event = kone_mouse_event_switch_profile;
 	roccat_report.value = new_profile;
 	roccat_report.key = 0;
@@ -163,6 +164,7 @@ static int kone_set_settings(struct usb_device *usb_dev,
 		struct kone_settings const *settings)
 {
 	int retval;
+
 	retval = kone_send(usb_dev, kone_command_settings,
 			settings, sizeof(struct kone_settings));
 	if (retval)
@@ -387,7 +389,7 @@ static struct bin_attribute bin_attr_profile##number = {	\
 	.read = kone_sysfs_read_profilex,			\
 	.write = kone_sysfs_write_profilex,			\
 	.private = &profile_numbers[number-1],			\
-};
+}
 PROFILE_ATTR(1);
 PROFILE_ATTR(2);
 PROFILE_ATTR(3);
@@ -456,6 +458,7 @@ static ssize_t kone_sysfs_show_tcu(struct device *dev,
 static int kone_tcu_command(struct usb_device *usb_dev, int number)
 {
 	unsigned char value;
+
 	value = number;
 	return kone_send(usb_dev, kone_command_calibrate, &value, 1);
 }
@@ -697,10 +700,8 @@ static int kone_init_specials(struct hid_device *hdev)
 			== USB_INTERFACE_PROTOCOL_MOUSE) {
 
 		kone = kzalloc(sizeof(*kone), GFP_KERNEL);
-		if (!kone) {
-			hid_err(hdev, "can't alloc device descriptor\n");
+		if (!kone)
 			return -ENOMEM;
-		}
 		hid_set_drvdata(hdev, kone);
 
 		retval = kone_init_kone_device_struct(usb_dev, kone);
diff --git a/drivers/hid/hid-roccat-pyra.c b/drivers/hid/hid-roccat-pyra.c
index 1a07e07d99a0..47d7e74231e5 100644
--- a/drivers/hid/hid-roccat-pyra.c
+++ b/drivers/hid/hid-roccat-pyra.c
@@ -35,6 +35,8 @@ static struct class *pyra_class;
 static void profile_activated(struct pyra_device *pyra,
 		unsigned int new_profile)
 {
+	if (new_profile >= ARRAY_SIZE(pyra->profile_settings))
+		return;
 	pyra->actual_profile = new_profile;
 	pyra->actual_cpi = pyra->profile_settings[pyra->actual_profile].y_cpi;
 }
@@ -257,9 +259,11 @@ static ssize_t pyra_sysfs_write_settings(struct file *fp,
 	if (off != 0 || count != PYRA_SIZE_SETTINGS)
 		return -EINVAL;
 
-	mutex_lock(&pyra->pyra_lock);
-
 	settings = (struct pyra_settings const *)buf;
+	if (settings->startup_profile >= ARRAY_SIZE(pyra->profile_settings))
+		return -EINVAL;
+
+	mutex_lock(&pyra->pyra_lock);
 
 	retval = pyra_set_settings(usb_dev, settings);
 	if (retval) {
diff --git a/drivers/hid/hid-saitek.c b/drivers/hid/hid-saitek.c
index 69cca1476a0c..5632c54eadf0 100644
--- a/drivers/hid/hid-saitek.c
+++ b/drivers/hid/hid-saitek.c
@@ -7,7 +7,7 @@
  *  (This module is based on "hid-ortek".)
  *  Copyright (c) 2012 Andreas Hübner
  *
- *  R.A.T.7, M.M.O.7 (USB gaming mice):
+ *  R.A.T.7, R.A.T.9, M.M.O.7 (USB gaming mice):
  *  Fixes the mode button which cycles through three constantly pressed
  *  buttons. All three press events are mapped to one button and the
  *  missing release event is generated immediately.
@@ -179,6 +179,8 @@ static const struct hid_device_id saitek_devices[] = {
 		.driver_data = SAITEK_FIX_PS1000 },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_SAITEK, USB_DEVICE_ID_SAITEK_RAT7),
 		.driver_data = SAITEK_RELEASE_MODE_RAT7 },
+	{ HID_USB_DEVICE(USB_VENDOR_ID_MADCATZ, USB_DEVICE_ID_MADCATZ_RAT9),
+		.driver_data = SAITEK_RELEASE_MODE_RAT7 },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_SAITEK, USB_DEVICE_ID_SAITEK_MMO7),
 		.driver_data = SAITEK_RELEASE_MODE_MMO7 },
 	{ }
diff --git a/drivers/hid/hid-sensor-hub.c b/drivers/hid/hid-sensor-hub.c
index 2ac25760a9a9..6a58b6c723aa 100644
--- a/drivers/hid/hid-sensor-hub.c
+++ b/drivers/hid/hid-sensor-hub.c
@@ -641,9 +641,6 @@ static int sensor_hub_probe(struct hid_device *hdev,
 					goto err_stop_hw;
 			}
 			sd->hid_sensor_hub_client_devs[
-				sd->hid_sensor_client_cnt].id =
-							PLATFORM_DEVID_AUTO;
-			sd->hid_sensor_hub_client_devs[
 				sd->hid_sensor_client_cnt].name = name;
 			sd->hid_sensor_hub_client_devs[
 				sd->hid_sensor_client_cnt].platform_data =
@@ -659,8 +656,9 @@ static int sensor_hub_probe(struct hid_device *hdev,
 	if (last_hsdev)
 		last_hsdev->end_collection_index = i;
 
-	ret = mfd_add_devices(&hdev->dev, 0, sd->hid_sensor_hub_client_devs,
-		sd->hid_sensor_client_cnt, NULL, 0, NULL);
+	ret = mfd_add_hotplug_devices(&hdev->dev,
+			sd->hid_sensor_hub_client_devs,
+			sd->hid_sensor_client_cnt);
 	if (ret < 0)
 		goto err_stop_hw;
 
@@ -709,6 +707,9 @@ static const struct hid_device_id sensor_hub_devices[] = {
 			USB_DEVICE_ID_MS_TYPE_COVER_2),
 			.driver_data = HID_SENSOR_HUB_ENUM_QUIRK},
 	{ HID_DEVICE(HID_BUS_ANY, HID_GROUP_SENSOR_HUB, USB_VENDOR_ID_STM_0,
+			USB_DEVICE_ID_STM_HID_SENSOR),
+			.driver_data = HID_SENSOR_HUB_ENUM_QUIRK},
+	{ HID_DEVICE(HID_BUS_ANY, HID_GROUP_SENSOR_HUB, USB_VENDOR_ID_STM_0,
 			USB_DEVICE_ID_STM_HID_SENSOR_1),
 			.driver_data = HID_SENSOR_HUB_ENUM_QUIRK},
 	{ HID_DEVICE(HID_BUS_ANY, HID_GROUP_SENSOR_HUB, USB_VENDOR_ID_TEXAS_INSTRUMENTS,
diff --git a/drivers/hid/hid-sony.c b/drivers/hid/hid-sony.c
index c372368e438c..31e9d2561106 100644
--- a/drivers/hid/hid-sony.c
+++ b/drivers/hid/hid-sony.c
@@ -1,5 +1,5 @@
 /*
- *  HID driver for Sony / PS2 / PS3 BD devices.
+ *  HID driver for Sony / PS2 / PS3 / PS4 BD devices.
  *
  *  Copyright (c) 1999 Andreas Gal
  *  Copyright (c) 2000-2005 Vojtech Pavlik <vojtech@suse.cz>
@@ -8,6 +8,7 @@
  *  Copyright (c) 2012 David Dillow <dave@thedillows.org>
  *  Copyright (c) 2006-2013 Jiri Kosina
  *  Copyright (c) 2013 Colin Leitner <colin.leitner@gmail.com>
+ *  Copyright (c) 2014 Frank Praznik <frank.praznik@gmail.com>
  */
 
 /*
@@ -176,7 +177,7 @@ static u8 dualshock4_usb_rdesc[] = {
 	0x75, 0x06,         /*      Report Size (6),                */
 	0x95, 0x01,         /*      Report Count (1),               */
 	0x15, 0x00,         /*      Logical Minimum (0),            */
-	0x25, 0x7F,         /*      Logical Maximum (127),          */
+	0x25, 0x3F,         /*      Logical Maximum (63),           */
 	0x81, 0x02,         /*      Input (Variable),               */
 	0x05, 0x01,         /*      Usage Page (Desktop),           */
 	0x09, 0x33,         /*      Usage (Rx),                     */
@@ -200,14 +201,14 @@ static u8 dualshock4_usb_rdesc[] = {
 	0x81, 0x02,         /*      Input (Variable),               */
 	0x19, 0x43,         /*      Usage Minimum (43h),            */
 	0x29, 0x45,         /*      Usage Maximum (45h),            */
-	0x16, 0xFF, 0xBF,   /*      Logical Minimum (-16385),       */
-	0x26, 0x00, 0x40,   /*      Logical Maximum (16384),        */
+	0x16, 0x00, 0xE0,   /*      Logical Minimum (-8192),        */
+	0x26, 0xFF, 0x1F,   /*      Logical Maximum (8191),         */
 	0x95, 0x03,         /*      Report Count (3),               */
 	0x81, 0x02,         /*      Input (Variable),               */
 	0x06, 0x00, 0xFF,   /*      Usage Page (FF00h),             */
 	0x09, 0x21,         /*      Usage (21h),                    */
 	0x15, 0x00,         /*      Logical Minimum (0),            */
-	0x25, 0xFF,         /*      Logical Maximum (255),          */
+	0x26, 0xFF, 0x00,   /*      Logical Maximum (255),          */
 	0x75, 0x08,         /*      Report Size (8),                */
 	0x95, 0x27,         /*      Report Count (39),              */
 	0x81, 0x02,         /*      Input (Variable),               */
@@ -395,11 +396,11 @@ static u8 dualshock4_usb_rdesc[] = {
 
 /*
  * The default behavior of the Dualshock 4 is to send reports using report
- * type 1 when running over Bluetooth. However, as soon as it receives a
- * report of type 17 to set the LEDs or rumble it starts returning it's state
- * in report 17 instead of 1.  Since report 17 is undefined in the default HID
+ * type 1 when running over Bluetooth. However, when feature report 2 is
+ * requested during the controller initialization it starts sending input
+ * reports in report 17.  Since report 17 is undefined in the default HID
  * descriptor the button and axis definitions must be moved to report 17 or
- * the HID layer won't process the received input once a report is sent.
+ * the HID layer won't process the received input.
  */
 static u8 dualshock4_bt_rdesc[] = {
 	0x05, 0x01,         /*  Usage Page (Desktop),               */
@@ -509,8 +510,8 @@ static u8 dualshock4_bt_rdesc[] = {
 	0x81, 0x02,         /*      Input (Variable),               */
 	0x19, 0x43,         /*      Usage Minimum (43h),            */
 	0x29, 0x45,         /*      Usage Maximum (45h),            */
-	0x16, 0xFF, 0xBF,   /*      Logical Minimum (-16385),       */
-	0x26, 0x00, 0x40,   /*      Logical Maximum (16384),        */
+	0x16, 0x00, 0xE0,   /*      Logical Minimum (-8192),        */
+	0x26, 0xFF, 0x1F,   /*      Logical Maximum (8191),         */
 	0x95, 0x03,         /*      Report Count (3),               */
 	0x81, 0x02,         /*      Input (Variable),               */
 	0x06, 0x00, 0xFF,   /*      Usage Page (FF00h),             */
@@ -797,6 +798,12 @@ union sixaxis_output_report_01 {
 	__u8 buf[36];
 };
 
+#define DS4_REPORT_0x02_SIZE 37
+#define DS4_REPORT_0x05_SIZE 32
+#define DS4_REPORT_0x11_SIZE 78
+#define DS4_REPORT_0x81_SIZE 7
+#define SIXAXIS_REPORT_0xF2_SIZE 18
+
 static spinlock_t sony_dev_list_lock;
 static LIST_HEAD(sony_device_list);
 static DEFINE_IDA(sony_device_id_allocator);
@@ -810,6 +817,7 @@ struct sony_sc {
 	struct work_struct state_worker;
 	struct power_supply battery;
 	int device_id;
+	__u8 *output_report_dmabuf;
 
 #ifdef CONFIG_SONY_FF
 	__u8 left;
@@ -935,12 +943,13 @@ static void sixaxis_parse_report(struct sony_sc *sc, __u8 *rd, int size)
 	if (rd[30] >= 0xee) {
 		battery_capacity = 100;
 		battery_charging = !(rd[30] & 0x01);
+		cable_state = 1;
 	} else {
 		__u8 index = rd[30] <= 5 ? rd[30] : 5;
 		battery_capacity = sixaxis_battery_capacity[index];
 		battery_charging = 0;
+		cable_state = 0;
 	}
-	cable_state = !(rd[31] & 0x04);
 
 	spin_lock_irqsave(&sc->lock, flags);
 	sc->cable_state = cable_state;
@@ -1082,6 +1091,38 @@ static int sony_mapping(struct hid_device *hdev, struct hid_input *hi,
 	return 0;
 }
 
+static int sony_register_touchpad(struct hid_input *hi, int touch_count,
+					int w, int h)
+{
+	struct input_dev *input_dev = hi->input;
+	int ret;
+
+	ret = input_mt_init_slots(input_dev, touch_count, 0);
+	if (ret < 0)
+		return ret;
+
+	input_set_abs_params(input_dev, ABS_MT_POSITION_X, 0, w, 0, 0);
+	input_set_abs_params(input_dev, ABS_MT_POSITION_Y, 0, h, 0, 0);
+
+	return 0;
+}
+
+static void sony_input_configured(struct hid_device *hdev,
+					struct hid_input *hidinput)
+{
+	struct sony_sc *sc = hid_get_drvdata(hdev);
+
+	/*
+	 * The Dualshock 4 touchpad supports 2 touches and has a
+	 * resolution of 1920x942 (44.86 dots/mm).
+	 */
+	if (sc->quirks & DUALSHOCK4_CONTROLLER) {
+		if (sony_register_touchpad(hidinput, 2, 1920, 942) != 0)
+			hid_err(sc->hdev,
+				"Unable to initialize multi-touch slots\n");
+	}
+}
+
 /*
  * Sending HID_REQ_GET_REPORT changes the operation mode of the ps3 controller
  * to "operational".  Without this, the ps3 controller will not report any
@@ -1108,9 +1149,20 @@ static int sixaxis_set_operational_usb(struct hid_device *hdev)
 
 static int sixaxis_set_operational_bt(struct hid_device *hdev)
 {
-	unsigned char buf[] = { 0xf4,  0x42, 0x03, 0x00, 0x00 };
-	return hid_hw_raw_request(hdev, buf[0], buf, sizeof(buf),
+	static const __u8 report[] = { 0xf4, 0x42, 0x03, 0x00, 0x00 };
+	__u8 *buf;
+	int ret;
+
+	buf = kmemdup(report, sizeof(report), GFP_KERNEL);
+	if (!buf)
+		return -ENOMEM;
+
+	ret = hid_hw_raw_request(hdev, buf[0], buf, sizeof(report),
 				  HID_FEATURE_REPORT, HID_REQ_SET_REPORT);
+
+	kfree(buf);
+
+	return ret;
 }
 
 /*
@@ -1119,10 +1171,19 @@ static int sixaxis_set_operational_bt(struct hid_device *hdev)
  */
 static int dualshock4_set_operational_bt(struct hid_device *hdev)
 {
-	__u8 buf[37] = { 0 };
+	__u8 *buf;
+	int ret;
 
-	return hid_hw_raw_request(hdev, 0x02, buf, sizeof(buf),
+	buf = kmalloc(DS4_REPORT_0x02_SIZE, GFP_KERNEL);
+	if (!buf)
+		return -ENOMEM;
+
+	ret = hid_hw_raw_request(hdev, 0x02, buf, DS4_REPORT_0x02_SIZE,
 				HID_FEATURE_REPORT, HID_REQ_GET_REPORT);
+
+	kfree(buf);
+
+	return ret;
 }
 
 static void sixaxis_set_leds_from_id(int id, __u8 values[MAX_LEDS])
@@ -1437,9 +1498,7 @@ error_leds:
 
 static void sixaxis_state_worker(struct work_struct *work)
 {
-	struct sony_sc *sc = container_of(work, struct sony_sc, state_worker);
-	int n;
-	union sixaxis_output_report_01 report = {
+	static const union sixaxis_output_report_01 default_report = {
 		.buf = {
 			0x01,
 			0x00, 0xff, 0x00, 0xff, 0x00,
@@ -1451,20 +1510,27 @@ static void sixaxis_state_worker(struct work_struct *work)
 			0x00, 0x00, 0x00, 0x00, 0x00
 		}
 	};
+	struct sony_sc *sc = container_of(work, struct sony_sc, state_worker);
+	struct sixaxis_output_report *report =
+		(struct sixaxis_output_report *)sc->output_report_dmabuf;
+	int n;
+
+	/* Initialize the report with default values */
+	memcpy(report, &default_report, sizeof(struct sixaxis_output_report));
 
 #ifdef CONFIG_SONY_FF
-	report.data.rumble.right_motor_on = sc->right ? 1 : 0;
-	report.data.rumble.left_motor_force = sc->left;
+	report->rumble.right_motor_on = sc->right ? 1 : 0;
+	report->rumble.left_motor_force = sc->left;
 #endif
 
-	report.data.leds_bitmap |= sc->led_state[0] << 1;
-	report.data.leds_bitmap |= sc->led_state[1] << 2;
-	report.data.leds_bitmap |= sc->led_state[2] << 3;
-	report.data.leds_bitmap |= sc->led_state[3] << 4;
+	report->leds_bitmap |= sc->led_state[0] << 1;
+	report->leds_bitmap |= sc->led_state[1] << 2;
+	report->leds_bitmap |= sc->led_state[2] << 3;
+	report->leds_bitmap |= sc->led_state[3] << 4;
 
 	/* Set flag for all leds off, required for 3rd party INTEC controller */
-	if ((report.data.leds_bitmap & 0x1E) == 0)
-		report.data.leds_bitmap |= 0x20;
+	if ((report->leds_bitmap & 0x1E) == 0)
+		report->leds_bitmap |= 0x20;
 
 	/*
 	 * The LEDs in the report are indexed in reverse order to their
@@ -1477,28 +1543,30 @@ static void sixaxis_state_worker(struct work_struct *work)
 	 */
 	for (n = 0; n < 4; n++) {
 		if (sc->led_delay_on[n] || sc->led_delay_off[n]) {
-			report.data.led[3 - n].duty_off = sc->led_delay_off[n];
-			report.data.led[3 - n].duty_on = sc->led_delay_on[n];
+			report->led[3 - n].duty_off = sc->led_delay_off[n];
+			report->led[3 - n].duty_on = sc->led_delay_on[n];
 		}
 	}
 
-	hid_hw_raw_request(sc->hdev, report.data.report_id, report.buf,
-			sizeof(report), HID_OUTPUT_REPORT, HID_REQ_SET_REPORT);
+	hid_hw_raw_request(sc->hdev, report->report_id, (__u8 *)report,
+			sizeof(struct sixaxis_output_report),
+			HID_OUTPUT_REPORT, HID_REQ_SET_REPORT);
 }
 
 static void dualshock4_state_worker(struct work_struct *work)
 {
 	struct sony_sc *sc = container_of(work, struct sony_sc, state_worker);
 	struct hid_device *hdev = sc->hdev;
+	__u8 *buf = sc->output_report_dmabuf;
 	int offset;
 
-	__u8 buf[78] = { 0 };
-
 	if (sc->quirks & DUALSHOCK4_CONTROLLER_USB) {
+		memset(buf, 0, DS4_REPORT_0x05_SIZE);
 		buf[0] = 0x05;
 		buf[1] = 0xFF;
 		offset = 4;
 	} else {
+		memset(buf, 0, DS4_REPORT_0x11_SIZE);
 		buf[0] = 0x11;
 		buf[1] = 0xB0;
 		buf[3] = 0x0F;
@@ -1526,12 +1594,33 @@ static void dualshock4_state_worker(struct work_struct *work)
 	buf[offset++] = sc->led_delay_off[3];
 
 	if (sc->quirks & DUALSHOCK4_CONTROLLER_USB)
-		hid_hw_output_report(hdev, buf, 32);
+		hid_hw_output_report(hdev, buf, DS4_REPORT_0x05_SIZE);
 	else
-		hid_hw_raw_request(hdev, 0x11, buf, 78,
+		hid_hw_raw_request(hdev, 0x11, buf, DS4_REPORT_0x11_SIZE,
 				HID_OUTPUT_REPORT, HID_REQ_SET_REPORT);
 }
 
+static int sony_allocate_output_report(struct sony_sc *sc)
+{
+	if (sc->quirks & SIXAXIS_CONTROLLER)
+		sc->output_report_dmabuf =
+			kmalloc(sizeof(union sixaxis_output_report_01),
+				GFP_KERNEL);
+	else if (sc->quirks & DUALSHOCK4_CONTROLLER_BT)
+		sc->output_report_dmabuf = kmalloc(DS4_REPORT_0x11_SIZE,
+						GFP_KERNEL);
+	else if (sc->quirks & DUALSHOCK4_CONTROLLER_USB)
+		sc->output_report_dmabuf = kmalloc(DS4_REPORT_0x05_SIZE,
+						GFP_KERNEL);
+	else
+		return 0;
+
+	if (!sc->output_report_dmabuf)
+		return -ENOMEM;
+
+	return 0;
+}
+
 #ifdef CONFIG_SONY_FF
 static int sony_play_effect(struct input_dev *dev, void *data,
 			    struct ff_effect *effect)
@@ -1654,26 +1743,6 @@ static void sony_battery_remove(struct sony_sc *sc)
 	sc->battery.name = NULL;
 }
 
-static int sony_register_touchpad(struct sony_sc *sc, int touch_count,
-					int w, int h)
-{
-	struct hid_input *hidinput = list_entry(sc->hdev->inputs.next,
-						struct hid_input, list);
-	struct input_dev *input_dev = hidinput->input;
-	int ret;
-
-	ret = input_mt_init_slots(input_dev, touch_count, 0);
-	if (ret < 0) {
-		hid_err(sc->hdev, "Unable to initialize multi-touch slots\n");
-		return ret;
-	}
-
-	input_set_abs_params(input_dev, ABS_MT_POSITION_X, 0, w, 0, 0);
-	input_set_abs_params(input_dev, ABS_MT_POSITION_Y, 0, h, 0, 0);
-
-	return 0;
-}
-
 /*
  * If a controller is plugged in via USB while already connected via Bluetooth
  * it will show up as two devices. A global list of connected controllers and
@@ -1740,6 +1809,7 @@ static int sony_get_bt_devaddr(struct sony_sc *sc)
 
 static int sony_check_add(struct sony_sc *sc)
 {
+	__u8 *buf = NULL;
 	int n, ret;
 
 	if ((sc->quirks & DUALSHOCK4_CONTROLLER_BT) ||
@@ -1755,36 +1825,44 @@ static int sony_check_add(struct sony_sc *sc)
 			return 0;
 		}
 	} else if (sc->quirks & DUALSHOCK4_CONTROLLER_USB) {
-		__u8 buf[7];
+		buf = kmalloc(DS4_REPORT_0x81_SIZE, GFP_KERNEL);
+		if (!buf)
+			return -ENOMEM;
 
 		/*
 		 * The MAC address of a DS4 controller connected via USB can be
 		 * retrieved with feature report 0x81. The address begins at
 		 * offset 1.
 		 */
-		ret = hid_hw_raw_request(sc->hdev, 0x81, buf, sizeof(buf),
-				HID_FEATURE_REPORT, HID_REQ_GET_REPORT);
+		ret = hid_hw_raw_request(sc->hdev, 0x81, buf,
+				DS4_REPORT_0x81_SIZE, HID_FEATURE_REPORT,
+				HID_REQ_GET_REPORT);
 
-		if (ret != 7) {
+		if (ret != DS4_REPORT_0x81_SIZE) {
 			hid_err(sc->hdev, "failed to retrieve feature report 0x81 with the DualShock 4 MAC address\n");
-			return ret < 0 ? ret : -EINVAL;
+			ret = ret < 0 ? ret : -EINVAL;
+			goto out_free;
 		}
 
 		memcpy(sc->mac_address, &buf[1], sizeof(sc->mac_address));
 	} else if (sc->quirks & SIXAXIS_CONTROLLER_USB) {
-		__u8 buf[18];
+		buf = kmalloc(SIXAXIS_REPORT_0xF2_SIZE, GFP_KERNEL);
+		if (!buf)
+			return -ENOMEM;
 
 		/*
 		 * The MAC address of a Sixaxis controller connected via USB can
 		 * be retrieved with feature report 0xf2. The address begins at
 		 * offset 4.
 		 */
-		ret = hid_hw_raw_request(sc->hdev, 0xf2, buf, sizeof(buf),
-				HID_FEATURE_REPORT, HID_REQ_GET_REPORT);
+		ret = hid_hw_raw_request(sc->hdev, 0xf2, buf,
+				SIXAXIS_REPORT_0xF2_SIZE, HID_FEATURE_REPORT,
+				HID_REQ_GET_REPORT);
 
-		if (ret != 18) {
+		if (ret != SIXAXIS_REPORT_0xF2_SIZE) {
 			hid_err(sc->hdev, "failed to retrieve feature report 0xf2 with the Sixaxis MAC address\n");
-			return ret < 0 ? ret : -EINVAL;
+			ret = ret < 0 ? ret : -EINVAL;
+			goto out_free;
 		}
 
 		/*
@@ -1797,7 +1875,13 @@ static int sony_check_add(struct sony_sc *sc)
 		return 0;
 	}
 
-	return sony_check_add_dev_list(sc);
+	ret = sony_check_add_dev_list(sc);
+
+out_free:
+
+	kfree(buf);
+
+	return ret;
 }
 
 static int sony_set_device_id(struct sony_sc *sc)
@@ -1881,6 +1965,12 @@ static int sony_probe(struct hid_device *hdev, const struct hid_device_id *id)
 		return ret;
 	}
 
+	ret = sony_allocate_output_report(sc);
+	if (ret < 0) {
+		hid_err(hdev, "failed to allocate the output report buffer\n");
+		goto err_stop;
+	}
+
 	ret = sony_set_device_id(sc);
 	if (ret < 0) {
 		hid_err(hdev, "failed to allocate the device id\n");
@@ -1923,13 +2013,6 @@ static int sony_probe(struct hid_device *hdev, const struct hid_device_id *id)
 				goto err_stop;
 			}
 		}
-		/*
-		 * The Dualshock 4 touchpad supports 2 touches and has a
-		 * resolution of 1920x940.
-		 */
-		ret = sony_register_touchpad(sc, 2, 1920, 940);
-		if (ret < 0)
-			goto err_stop;
 
 		sony_init_work(sc, dualshock4_state_worker);
 	} else {
@@ -1977,6 +2060,7 @@ err_stop:
 	if (sc->quirks & SONY_BATTERY_SUPPORT)
 		sony_battery_remove(sc);
 	sony_cancel_work_sync(sc);
+	kfree(sc->output_report_dmabuf);
 	sony_remove_dev_list(sc);
 	sony_release_device_id(sc);
 	hid_hw_stop(hdev);
@@ -1997,6 +2081,8 @@ static void sony_remove(struct hid_device *hdev)
 
 	sony_cancel_work_sync(sc);
 
+	kfree(sc->output_report_dmabuf);
+
 	sony_remove_dev_list(sc);
 
 	sony_release_device_id(sc);
@@ -2027,6 +2113,9 @@ static const struct hid_device_id sony_devices[] = {
 	/* Logitech Harmony Adapter for PS3 */
 	{ HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_LOGITECH, USB_DEVICE_ID_LOGITECH_HARMONY_PS3),
 		.driver_data = PS3REMOTE },
+	/* SMK-Link PS3 BD Remote Control */
+	{ HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SMK, USB_DEVICE_ID_SMK_PS3_BDREMOTE),
+		.driver_data = PS3REMOTE },
 	/* Sony Dualshock 4 controllers for PS4 */
 	{ HID_USB_DEVICE(USB_VENDOR_ID_SONY, USB_DEVICE_ID_SONY_PS4_CONTROLLER),
 		.driver_data = DUALSHOCK4_CONTROLLER_USB },
@@ -2037,13 +2126,14 @@ static const struct hid_device_id sony_devices[] = {
 MODULE_DEVICE_TABLE(hid, sony_devices);
 
 static struct hid_driver sony_driver = {
-	.name          = "sony",
-	.id_table      = sony_devices,
-	.input_mapping = sony_mapping,
-	.probe         = sony_probe,
-	.remove        = sony_remove,
-	.report_fixup  = sony_report_fixup,
-	.raw_event     = sony_raw_event
+	.name             = "sony",
+	.id_table         = sony_devices,
+	.input_mapping    = sony_mapping,
+	.input_configured = sony_input_configured,
+	.probe            = sony_probe,
+	.remove           = sony_remove,
+	.report_fixup     = sony_report_fixup,
+	.raw_event        = sony_raw_event
 };
 
 static int __init sony_init(void)
diff --git a/drivers/hid/hid-thingm.c b/drivers/hid/hid-thingm.c
index 134be89b15ea..b95d3978c272 100644
--- a/drivers/hid/hid-thingm.c
+++ b/drivers/hid/hid-thingm.c
@@ -208,10 +208,10 @@ unregister_red:
 
 static void thingm_remove_rgb(struct thingm_rgb *rgb)
 {
-	flush_work(&rgb->work);
 	led_classdev_unregister(&rgb->red.ldev);
 	led_classdev_unregister(&rgb->green.ldev);
 	led_classdev_unregister(&rgb->blue.ldev);
+	flush_work(&rgb->work);
 }
 
 static int thingm_probe(struct hid_device *hdev, const struct hid_device_id *id)
@@ -250,6 +250,7 @@ static int thingm_probe(struct hid_device *hdev, const struct hid_device_id *id)
 
 	if (!tdev->fwinfo) {
 		hid_err(hdev, "unsupported firmware %c\n", tdev->version.major);
+		err = -ENODEV;
 		goto stop;
 	}
 
@@ -286,10 +287,10 @@ static void thingm_remove(struct hid_device *hdev)
 	struct thingm_device *tdev = hid_get_drvdata(hdev);
 	int i;
 
+	hid_hw_stop(hdev);
+
 	for (i = 0; i < tdev->fwinfo->numrgb; ++i)
 		thingm_remove_rgb(tdev->rgb + i);
-
-	hid_hw_stop(hdev);
 }
 
 static const struct hid_device_id thingm_table[] = {
diff --git a/drivers/hid/i2c-hid/i2c-hid.c b/drivers/hid/i2c-hid/i2c-hid.c
index 747d54421e73..d43e967e7533 100644
--- a/drivers/hid/i2c-hid/i2c-hid.c
+++ b/drivers/hid/i2c-hid/i2c-hid.c
@@ -137,6 +137,7 @@ struct i2c_hid {
 						   * descriptor. */
 	unsigned int		bufsize;	/* i2c buffer size */
 	char			*inbuf;		/* Input buffer */
+	char			*rawbuf;	/* Raw Input buffer */
 	char			*cmdbuf;	/* Command buffer */
 	char			*argsbuf;	/* Command arguments buffer */
 
@@ -369,7 +370,7 @@ static int i2c_hid_hwreset(struct i2c_client *client)
 static void i2c_hid_get_input(struct i2c_hid *ihid)
 {
 	int ret, ret_size;
-	int size = le16_to_cpu(ihid->hdesc.wMaxInputLength);
+	int size = ihid->bufsize;
 
 	ret = i2c_master_recv(ihid->client, ihid->inbuf, size);
 	if (ret != size) {
@@ -437,7 +438,7 @@ static void i2c_hid_init_report(struct hid_report *report, u8 *buffer,
 			report->id, buffer, size))
 		return;
 
-	i2c_hid_dbg(ihid, "report (len=%d): %*ph\n", size, size, ihid->inbuf);
+	i2c_hid_dbg(ihid, "report (len=%d): %*ph\n", size, size, buffer);
 
 	ret_size = buffer[0] | (buffer[1] << 8);
 
@@ -504,9 +505,11 @@ static void i2c_hid_find_max_report(struct hid_device *hid, unsigned int type,
 static void i2c_hid_free_buffers(struct i2c_hid *ihid)
 {
 	kfree(ihid->inbuf);
+	kfree(ihid->rawbuf);
 	kfree(ihid->argsbuf);
 	kfree(ihid->cmdbuf);
 	ihid->inbuf = NULL;
+	ihid->rawbuf = NULL;
 	ihid->cmdbuf = NULL;
 	ihid->argsbuf = NULL;
 	ihid->bufsize = 0;
@@ -522,10 +525,11 @@ static int i2c_hid_alloc_buffers(struct i2c_hid *ihid, size_t report_size)
 		       report_size; /* report */
 
 	ihid->inbuf = kzalloc(report_size, GFP_KERNEL);
+	ihid->rawbuf = kzalloc(report_size, GFP_KERNEL);
 	ihid->argsbuf = kzalloc(args_len, GFP_KERNEL);
 	ihid->cmdbuf = kzalloc(sizeof(union command) + args_len, GFP_KERNEL);
 
-	if (!ihid->inbuf || !ihid->argsbuf || !ihid->cmdbuf) {
+	if (!ihid->inbuf || !ihid->rawbuf || !ihid->argsbuf || !ihid->cmdbuf) {
 		i2c_hid_free_buffers(ihid);
 		return -ENOMEM;
 	}
@@ -552,12 +556,12 @@ static int i2c_hid_get_raw_report(struct hid_device *hid,
 
 	ret = i2c_hid_get_report(client,
 			report_type == HID_FEATURE_REPORT ? 0x03 : 0x01,
-			report_number, ihid->inbuf, ask_count);
+			report_number, ihid->rawbuf, ask_count);
 
 	if (ret < 0)
 		return ret;
 
-	ret_count = ihid->inbuf[0] | (ihid->inbuf[1] << 8);
+	ret_count = ihid->rawbuf[0] | (ihid->rawbuf[1] << 8);
 
 	if (ret_count <= 2)
 		return 0;
@@ -566,7 +570,7 @@ static int i2c_hid_get_raw_report(struct hid_device *hid,
 
 	/* The query buffer contains the size, dropping it in the reply */
 	count = min(count, ret_count - 2);
-	memcpy(buf, ihid->inbuf + 2, count);
+	memcpy(buf, ihid->rawbuf + 2, count);
 
 	return count;
 }
@@ -702,12 +706,7 @@ static int i2c_hid_start(struct hid_device *hid)
 
 static void i2c_hid_stop(struct hid_device *hid)
 {
-	struct i2c_client *client = hid->driver_data;
-	struct i2c_hid *ihid = i2c_get_clientdata(client);
-
 	hid->claimed = 0;
-
-	i2c_hid_free_buffers(ihid);
 }
 
 static int i2c_hid_open(struct hid_device *hid)
@@ -1095,7 +1094,7 @@ static int i2c_hid_resume(struct device *dev)
 }
 #endif
 
-#ifdef CONFIG_PM_RUNTIME
+#ifdef CONFIG_PM
 static int i2c_hid_runtime_suspend(struct device *dev)
 {
 	struct i2c_client *client = to_i2c_client(dev);
diff --git a/drivers/hid/uhid.c b/drivers/hid/uhid.c
index 0cb92e347258..e094c572b86e 100644
--- a/drivers/hid/uhid.c
+++ b/drivers/hid/uhid.c
@@ -44,10 +44,12 @@ struct uhid_device {
 	__u8 tail;
 	struct uhid_event *outq[UHID_BUFSIZE];
 
+	/* blocking GET_REPORT support; state changes protected by qlock */
 	struct mutex report_lock;
 	wait_queue_head_t report_wait;
-	atomic_t report_done;
-	atomic_t report_id;
+	bool report_running;
+	u32 report_id;
+	u32 report_type;
 	struct uhid_event report_buf;
 };
 
@@ -90,8 +92,27 @@ static int uhid_queue_event(struct uhid_device *uhid, __u32 event)
 static int uhid_hid_start(struct hid_device *hid)
 {
 	struct uhid_device *uhid = hid->driver_data;
+	struct uhid_event *ev;
+	unsigned long flags;
+
+	ev = kzalloc(sizeof(*ev), GFP_KERNEL);
+	if (!ev)
+		return -ENOMEM;
+
+	ev->type = UHID_START;
 
-	return uhid_queue_event(uhid, UHID_START);
+	if (hid->report_enum[HID_FEATURE_REPORT].numbered)
+		ev->u.start.dev_flags |= UHID_DEV_NUMBERED_FEATURE_REPORTS;
+	if (hid->report_enum[HID_OUTPUT_REPORT].numbered)
+		ev->u.start.dev_flags |= UHID_DEV_NUMBERED_OUTPUT_REPORTS;
+	if (hid->report_enum[HID_INPUT_REPORT].numbered)
+		ev->u.start.dev_flags |= UHID_DEV_NUMBERED_INPUT_REPORTS;
+
+	spin_lock_irqsave(&uhid->qlock, flags);
+	uhid_queue(uhid, ev);
+	spin_unlock_irqrestore(&uhid->qlock, flags);
+
+	return 0;
 }
 
 static void uhid_hid_stop(struct hid_device *hid)
@@ -123,87 +144,169 @@ static int uhid_hid_parse(struct hid_device *hid)
 	return hid_parse_report(hid, uhid->rd_data, uhid->rd_size);
 }
 
-static int uhid_hid_get_raw(struct hid_device *hid, unsigned char rnum,
-			    __u8 *buf, size_t count, unsigned char rtype)
+/* must be called with report_lock held */
+static int __uhid_report_queue_and_wait(struct uhid_device *uhid,
+					struct uhid_event *ev,
+					__u32 *report_id)
+{
+	unsigned long flags;
+	int ret;
+
+	spin_lock_irqsave(&uhid->qlock, flags);
+	*report_id = ++uhid->report_id;
+	uhid->report_type = ev->type + 1;
+	uhid->report_running = true;
+	uhid_queue(uhid, ev);
+	spin_unlock_irqrestore(&uhid->qlock, flags);
+
+	ret = wait_event_interruptible_timeout(uhid->report_wait,
+				!uhid->report_running || !uhid->running,
+				5 * HZ);
+	if (!ret || !uhid->running || uhid->report_running)
+		ret = -EIO;
+	else if (ret < 0)
+		ret = -ERESTARTSYS;
+	else
+		ret = 0;
+
+	uhid->report_running = false;
+
+	return ret;
+}
+
+static void uhid_report_wake_up(struct uhid_device *uhid, u32 id,
+				const struct uhid_event *ev)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&uhid->qlock, flags);
+
+	/* id for old report; drop it silently */
+	if (uhid->report_type != ev->type || uhid->report_id != id)
+		goto unlock;
+	if (!uhid->report_running)
+		goto unlock;
+
+	memcpy(&uhid->report_buf, ev, sizeof(*ev));
+	uhid->report_running = false;
+	wake_up_interruptible(&uhid->report_wait);
+
+unlock:
+	spin_unlock_irqrestore(&uhid->qlock, flags);
+}
+
+static int uhid_hid_get_report(struct hid_device *hid, unsigned char rnum,
+			       u8 *buf, size_t count, u8 rtype)
 {
 	struct uhid_device *uhid = hid->driver_data;
-	__u8 report_type;
+	struct uhid_get_report_reply_req *req;
 	struct uhid_event *ev;
-	unsigned long flags;
 	int ret;
-	size_t uninitialized_var(len);
-	struct uhid_feature_answer_req *req;
 
 	if (!uhid->running)
 		return -EIO;
 
-	switch (rtype) {
-	case HID_FEATURE_REPORT:
-		report_type = UHID_FEATURE_REPORT;
-		break;
-	case HID_OUTPUT_REPORT:
-		report_type = UHID_OUTPUT_REPORT;
-		break;
-	case HID_INPUT_REPORT:
-		report_type = UHID_INPUT_REPORT;
-		break;
-	default:
-		return -EINVAL;
-	}
+	ev = kzalloc(sizeof(*ev), GFP_KERNEL);
+	if (!ev)
+		return -ENOMEM;
+
+	ev->type = UHID_GET_REPORT;
+	ev->u.get_report.rnum = rnum;
+	ev->u.get_report.rtype = rtype;
 
 	ret = mutex_lock_interruptible(&uhid->report_lock);
-	if (ret)
+	if (ret) {
+		kfree(ev);
 		return ret;
+	}
 
-	ev = kzalloc(sizeof(*ev), GFP_KERNEL);
-	if (!ev) {
-		ret = -ENOMEM;
+	/* this _always_ takes ownership of @ev */
+	ret = __uhid_report_queue_and_wait(uhid, ev, &ev->u.get_report.id);
+	if (ret)
 		goto unlock;
+
+	req = &uhid->report_buf.u.get_report_reply;
+	if (req->err) {
+		ret = -EIO;
+	} else {
+		ret = min3(count, (size_t)req->size, (size_t)UHID_DATA_MAX);
+		memcpy(buf, req->data, ret);
 	}
 
-	spin_lock_irqsave(&uhid->qlock, flags);
-	ev->type = UHID_FEATURE;
-	ev->u.feature.id = atomic_inc_return(&uhid->report_id);
-	ev->u.feature.rnum = rnum;
-	ev->u.feature.rtype = report_type;
+unlock:
+	mutex_unlock(&uhid->report_lock);
+	return ret;
+}
 
-	atomic_set(&uhid->report_done, 0);
-	uhid_queue(uhid, ev);
-	spin_unlock_irqrestore(&uhid->qlock, flags);
+static int uhid_hid_set_report(struct hid_device *hid, unsigned char rnum,
+			       const u8 *buf, size_t count, u8 rtype)
+{
+	struct uhid_device *uhid = hid->driver_data;
+	struct uhid_event *ev;
+	int ret;
 
-	ret = wait_event_interruptible_timeout(uhid->report_wait,
-				atomic_read(&uhid->report_done), 5 * HZ);
-
-	/*
-	 * Make sure "uhid->running" is cleared on shutdown before
-	 * "uhid->report_done" is set.
-	 */
-	smp_rmb();
-	if (!ret || !uhid->running) {
-		ret = -EIO;
-	} else if (ret < 0) {
-		ret = -ERESTARTSYS;
-	} else {
-		spin_lock_irqsave(&uhid->qlock, flags);
-		req = &uhid->report_buf.u.feature_answer;
+	if (!uhid->running || count > UHID_DATA_MAX)
+		return -EIO;
 
-		if (req->err) {
-			ret = -EIO;
-		} else {
-			ret = 0;
-			len = min(count,
-				min_t(size_t, req->size, UHID_DATA_MAX));
-			memcpy(buf, req->data, len);
-		}
+	ev = kzalloc(sizeof(*ev), GFP_KERNEL);
+	if (!ev)
+		return -ENOMEM;
+
+	ev->type = UHID_SET_REPORT;
+	ev->u.set_report.rnum = rnum;
+	ev->u.set_report.rtype = rtype;
+	ev->u.set_report.size = count;
+	memcpy(ev->u.set_report.data, buf, count);
 
-		spin_unlock_irqrestore(&uhid->qlock, flags);
+	ret = mutex_lock_interruptible(&uhid->report_lock);
+	if (ret) {
+		kfree(ev);
+		return ret;
 	}
 
-	atomic_set(&uhid->report_done, 1);
+	/* this _always_ takes ownership of @ev */
+	ret = __uhid_report_queue_and_wait(uhid, ev, &ev->u.set_report.id);
+	if (ret)
+		goto unlock;
+
+	if (uhid->report_buf.u.set_report_reply.err)
+		ret = -EIO;
+	else
+		ret = count;
 
 unlock:
 	mutex_unlock(&uhid->report_lock);
-	return ret ? ret : len;
+	return ret;
+}
+
+static int uhid_hid_raw_request(struct hid_device *hid, unsigned char reportnum,
+				__u8 *buf, size_t len, unsigned char rtype,
+				int reqtype)
+{
+	u8 u_rtype;
+
+	switch (rtype) {
+	case HID_FEATURE_REPORT:
+		u_rtype = UHID_FEATURE_REPORT;
+		break;
+	case HID_OUTPUT_REPORT:
+		u_rtype = UHID_OUTPUT_REPORT;
+		break;
+	case HID_INPUT_REPORT:
+		u_rtype = UHID_INPUT_REPORT;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	switch (reqtype) {
+	case HID_REQ_GET_REPORT:
+		return uhid_hid_get_report(hid, reportnum, buf, len, u_rtype);
+	case HID_REQ_SET_REPORT:
+		return uhid_hid_set_report(hid, reportnum, buf, len, u_rtype);
+	default:
+		return -EIO;
+	}
 }
 
 static int uhid_hid_output_raw(struct hid_device *hid, __u8 *buf, size_t count,
@@ -250,29 +353,14 @@ static int uhid_hid_output_report(struct hid_device *hid, __u8 *buf,
 	return uhid_hid_output_raw(hid, buf, count, HID_OUTPUT_REPORT);
 }
 
-static int uhid_raw_request(struct hid_device *hid, unsigned char reportnum,
-			    __u8 *buf, size_t len, unsigned char rtype,
-			    int reqtype)
-{
-	switch (reqtype) {
-	case HID_REQ_GET_REPORT:
-		return uhid_hid_get_raw(hid, reportnum, buf, len, rtype);
-	case HID_REQ_SET_REPORT:
-		/* TODO: implement proper SET_REPORT functionality */
-		return -ENOSYS;
-	default:
-		return -EIO;
-	}
-}
-
 static struct hid_ll_driver uhid_hid_driver = {
 	.start = uhid_hid_start,
 	.stop = uhid_hid_stop,
 	.open = uhid_hid_open,
 	.close = uhid_hid_close,
 	.parse = uhid_hid_parse,
+	.raw_request = uhid_hid_raw_request,
 	.output_report = uhid_hid_output_report,
-	.raw_request = uhid_raw_request,
 };
 
 #ifdef CONFIG_COMPAT
@@ -363,28 +451,27 @@ static int uhid_event_from_user(const char __user *buffer, size_t len,
 }
 #endif
 
-static int uhid_dev_create(struct uhid_device *uhid,
-			   const struct uhid_event *ev)
+static int uhid_dev_create2(struct uhid_device *uhid,
+			    const struct uhid_event *ev)
 {
 	struct hid_device *hid;
+	size_t rd_size, len;
+	void *rd_data;
 	int ret;
 
 	if (uhid->running)
 		return -EALREADY;
 
-	uhid->rd_size = ev->u.create.rd_size;
-	if (uhid->rd_size <= 0 || uhid->rd_size > HID_MAX_DESCRIPTOR_SIZE)
+	rd_size = ev->u.create2.rd_size;
+	if (rd_size <= 0 || rd_size > HID_MAX_DESCRIPTOR_SIZE)
 		return -EINVAL;
 
-	uhid->rd_data = kmalloc(uhid->rd_size, GFP_KERNEL);
-	if (!uhid->rd_data)
+	rd_data = kmemdup(ev->u.create2.rd_data, rd_size, GFP_KERNEL);
+	if (!rd_data)
 		return -ENOMEM;
 
-	if (copy_from_user(uhid->rd_data, ev->u.create.rd_data,
-			   uhid->rd_size)) {
-		ret = -EFAULT;
-		goto err_free;
-	}
+	uhid->rd_size = rd_size;
+	uhid->rd_data = rd_data;
 
 	hid = hid_allocate_device();
 	if (IS_ERR(hid)) {
@@ -392,19 +479,19 @@ static int uhid_dev_create(struct uhid_device *uhid,
 		goto err_free;
 	}
 
-	strncpy(hid->name, ev->u.create.name, 127);
-	hid->name[127] = 0;
-	strncpy(hid->phys, ev->u.create.phys, 63);
-	hid->phys[63] = 0;
-	strncpy(hid->uniq, ev->u.create.uniq, 63);
-	hid->uniq[63] = 0;
+	len = min(sizeof(hid->name), sizeof(ev->u.create2.name)) - 1;
+	strncpy(hid->name, ev->u.create2.name, len);
+	len = min(sizeof(hid->phys), sizeof(ev->u.create2.phys)) - 1;
+	strncpy(hid->phys, ev->u.create2.phys, len);
+	len = min(sizeof(hid->uniq), sizeof(ev->u.create2.uniq)) - 1;
+	strncpy(hid->uniq, ev->u.create2.uniq, len);
 
 	hid->ll_driver = &uhid_hid_driver;
-	hid->bus = ev->u.create.bus;
-	hid->vendor = ev->u.create.vendor;
-	hid->product = ev->u.create.product;
-	hid->version = ev->u.create.version;
-	hid->country = ev->u.create.country;
+	hid->bus = ev->u.create2.bus;
+	hid->vendor = ev->u.create2.vendor;
+	hid->product = ev->u.create2.product;
+	hid->version = ev->u.create2.version;
+	hid->country = ev->u.create2.country;
 	hid->driver_data = uhid;
 	hid->dev.parent = uhid_misc.this_device;
 
@@ -425,67 +512,34 @@ err_hid:
 	uhid->running = false;
 err_free:
 	kfree(uhid->rd_data);
+	uhid->rd_data = NULL;
+	uhid->rd_size = 0;
 	return ret;
 }
 
-static int uhid_dev_create2(struct uhid_device *uhid,
-			    const struct uhid_event *ev)
+static int uhid_dev_create(struct uhid_device *uhid,
+			   struct uhid_event *ev)
 {
-	struct hid_device *hid;
-	int ret;
+	struct uhid_create_req orig;
 
-	if (uhid->running)
-		return -EALREADY;
+	orig = ev->u.create;
 
-	uhid->rd_size = ev->u.create2.rd_size;
-	if (uhid->rd_size <= 0 || uhid->rd_size > HID_MAX_DESCRIPTOR_SIZE)
+	if (orig.rd_size <= 0 || orig.rd_size > HID_MAX_DESCRIPTOR_SIZE)
 		return -EINVAL;
+	if (copy_from_user(&ev->u.create2.rd_data, orig.rd_data, orig.rd_size))
+		return -EFAULT;
 
-	uhid->rd_data = kmemdup(ev->u.create2.rd_data, uhid->rd_size,
-				GFP_KERNEL);
-	if (!uhid->rd_data)
-		return -ENOMEM;
-
-	hid = hid_allocate_device();
-	if (IS_ERR(hid)) {
-		ret = PTR_ERR(hid);
-		goto err_free;
-	}
-
-	strncpy(hid->name, ev->u.create2.name, 127);
-	hid->name[127] = 0;
-	strncpy(hid->phys, ev->u.create2.phys, 63);
-	hid->phys[63] = 0;
-	strncpy(hid->uniq, ev->u.create2.uniq, 63);
-	hid->uniq[63] = 0;
-
-	hid->ll_driver = &uhid_hid_driver;
-	hid->bus = ev->u.create2.bus;
-	hid->vendor = ev->u.create2.vendor;
-	hid->product = ev->u.create2.product;
-	hid->version = ev->u.create2.version;
-	hid->country = ev->u.create2.country;
-	hid->driver_data = uhid;
-	hid->dev.parent = uhid_misc.this_device;
-
-	uhid->hid = hid;
-	uhid->running = true;
-
-	ret = hid_add_device(hid);
-	if (ret) {
-		hid_err(hid, "Cannot register HID device\n");
-		goto err_hid;
-	}
-
-	return 0;
-
-err_hid:
-	hid_destroy_device(hid);
-	uhid->hid = NULL;
-	uhid->running = false;
-err_free:
-	kfree(uhid->rd_data);
-	return ret;
+	memcpy(ev->u.create2.name, orig.name, sizeof(orig.name));
+	memcpy(ev->u.create2.phys, orig.phys, sizeof(orig.phys));
+	memcpy(ev->u.create2.uniq, orig.uniq, sizeof(orig.uniq));
+	ev->u.create2.rd_size = orig.rd_size;
+	ev->u.create2.bus = orig.bus;
+	ev->u.create2.vendor = orig.vendor;
+	ev->u.create2.product = orig.product;
+	ev->u.create2.version = orig.version;
+	ev->u.create2.country = orig.country;
+
+	return uhid_dev_create2(uhid, ev);
 }
 
 static int uhid_dev_destroy(struct uhid_device *uhid)
@@ -493,10 +547,7 @@ static int uhid_dev_destroy(struct uhid_device *uhid)
 	if (!uhid->running)
 		return -EINVAL;
 
-	/* clear "running" before setting "report_done" */
 	uhid->running = false;
-	smp_wmb();
-	atomic_set(&uhid->report_done, 1);
 	wake_up_interruptible(&uhid->report_wait);
 
 	hid_destroy_device(uhid->hid);
@@ -527,28 +578,23 @@ static int uhid_dev_input2(struct uhid_device *uhid, struct uhid_event *ev)
 	return 0;
 }
 
-static int uhid_dev_feature_answer(struct uhid_device *uhid,
-				   struct uhid_event *ev)
+static int uhid_dev_get_report_reply(struct uhid_device *uhid,
+				     struct uhid_event *ev)
 {
-	unsigned long flags;
-
 	if (!uhid->running)
 		return -EINVAL;
 
-	spin_lock_irqsave(&uhid->qlock, flags);
-
-	/* id for old report; drop it silently */
-	if (atomic_read(&uhid->report_id) != ev->u.feature_answer.id)
-		goto unlock;
-	if (atomic_read(&uhid->report_done))
-		goto unlock;
+	uhid_report_wake_up(uhid, ev->u.get_report_reply.id, ev);
+	return 0;
+}
 
-	memcpy(&uhid->report_buf, ev, sizeof(*ev));
-	atomic_set(&uhid->report_done, 1);
-	wake_up_interruptible(&uhid->report_wait);
+static int uhid_dev_set_report_reply(struct uhid_device *uhid,
+				     struct uhid_event *ev)
+{
+	if (!uhid->running)
+		return -EINVAL;
 
-unlock:
-	spin_unlock_irqrestore(&uhid->qlock, flags);
+	uhid_report_wake_up(uhid, ev->u.set_report_reply.id, ev);
 	return 0;
 }
 
@@ -566,7 +612,6 @@ static int uhid_char_open(struct inode *inode, struct file *file)
 	init_waitqueue_head(&uhid->waitq);
 	init_waitqueue_head(&uhid->report_wait);
 	uhid->running = false;
-	atomic_set(&uhid->report_done, 1);
 
 	file->private_data = uhid;
 	nonseekable_open(inode, file);
@@ -675,8 +720,11 @@ static ssize_t uhid_char_write(struct file *file, const char __user *buffer,
 	case UHID_INPUT2:
 		ret = uhid_dev_input2(uhid, &uhid->input_buf);
 		break;
-	case UHID_FEATURE_ANSWER:
-		ret = uhid_dev_feature_answer(uhid, &uhid->input_buf);
+	case UHID_GET_REPORT_REPLY:
+		ret = uhid_dev_get_report_reply(uhid, &uhid->input_buf);
+		break;
+	case UHID_SET_REPORT_REPLY:
+		ret = uhid_dev_set_report_reply(uhid, &uhid->input_buf);
 		break;
 	default:
 		ret = -EOPNOTSUPP;
diff --git a/drivers/hid/usbhid/hid-core.c b/drivers/hid/usbhid/hid-core.c
index 79cf503e37bf..bfbe1bedda7f 100644
--- a/drivers/hid/usbhid/hid-core.c
+++ b/drivers/hid/usbhid/hid-core.c
@@ -82,7 +82,7 @@ static int hid_start_in(struct hid_device *hid)
 	struct usbhid_device *usbhid = hid->driver_data;
 
 	spin_lock_irqsave(&usbhid->lock, flags);
-	if (hid->open > 0 &&
+	if ((hid->open > 0 || hid->quirks & HID_QUIRK_ALWAYS_POLL) &&
 			!test_bit(HID_DISCONNECTED, &usbhid->iofl) &&
 			!test_bit(HID_SUSPENDED, &usbhid->iofl) &&
 			!test_and_set_bit(HID_IN_RUNNING, &usbhid->iofl)) {
@@ -116,40 +116,24 @@ static void hid_reset(struct work_struct *work)
 	struct usbhid_device *usbhid =
 		container_of(work, struct usbhid_device, reset_work);
 	struct hid_device *hid = usbhid->hid;
-	int rc = 0;
+	int rc;
 
 	if (test_bit(HID_CLEAR_HALT, &usbhid->iofl)) {
 		dev_dbg(&usbhid->intf->dev, "clear halt\n");
 		rc = usb_clear_halt(hid_to_usb_dev(hid), usbhid->urbin->pipe);
 		clear_bit(HID_CLEAR_HALT, &usbhid->iofl);
-		hid_start_in(hid);
-	}
-
-	else if (test_bit(HID_RESET_PENDING, &usbhid->iofl)) {
-		dev_dbg(&usbhid->intf->dev, "resetting device\n");
-		rc = usb_lock_device_for_reset(hid_to_usb_dev(hid), usbhid->intf);
 		if (rc == 0) {
-			rc = usb_reset_device(hid_to_usb_dev(hid));
-			usb_unlock_device(hid_to_usb_dev(hid));
+			hid_start_in(hid);
+		} else {
+			dev_dbg(&usbhid->intf->dev,
+					"clear-halt failed: %d\n", rc);
+			set_bit(HID_RESET_PENDING, &usbhid->iofl);
 		}
-		clear_bit(HID_RESET_PENDING, &usbhid->iofl);
 	}
 
-	switch (rc) {
-	case 0:
-		if (!test_bit(HID_IN_RUNNING, &usbhid->iofl))
-			hid_io_error(hid);
-		break;
-	default:
-		hid_err(hid, "can't reset device, %s-%s/input%d, status %d\n",
-			hid_to_usb_dev(hid)->bus->bus_name,
-			hid_to_usb_dev(hid)->devpath,
-			usbhid->ifnum, rc);
-		/* FALLTHROUGH */
-	case -EHOSTUNREACH:
-	case -ENODEV:
-	case -EINTR:
-		break;
+	if (test_bit(HID_RESET_PENDING, &usbhid->iofl)) {
+		dev_dbg(&usbhid->intf->dev, "resetting device\n");
+		usb_queue_reset_device(usbhid->intf);
 	}
 }
 
@@ -292,18 +276,22 @@ static void hid_irq_in(struct urb *urb)
 	case 0:			/* success */
 		usbhid_mark_busy(usbhid);
 		usbhid->retry_delay = 0;
-		hid_input_report(urb->context, HID_INPUT_REPORT,
-				 urb->transfer_buffer,
-				 urb->actual_length, 1);
-		/*
-		 * autosuspend refused while keys are pressed
-		 * because most keyboards don't wake up when
-		 * a key is released
-		 */
-		if (hid_check_keys_pressed(hid))
-			set_bit(HID_KEYS_PRESSED, &usbhid->iofl);
-		else
-			clear_bit(HID_KEYS_PRESSED, &usbhid->iofl);
+		if ((hid->quirks & HID_QUIRK_ALWAYS_POLL) && !hid->open)
+			break;
+		if (!test_bit(HID_RESUME_RUNNING, &usbhid->iofl)) {
+			hid_input_report(urb->context, HID_INPUT_REPORT,
+					 urb->transfer_buffer,
+					 urb->actual_length, 1);
+			/*
+			 * autosuspend refused while keys are pressed
+			 * because most keyboards don't wake up when
+			 * a key is released
+			 */
+			if (hid_check_keys_pressed(hid))
+				set_bit(HID_KEYS_PRESSED, &usbhid->iofl);
+			else
+				clear_bit(HID_KEYS_PRESSED, &usbhid->iofl);
+		}
 		break;
 	case -EPIPE:		/* stall */
 		usbhid_mark_busy(usbhid);
@@ -352,8 +340,7 @@ static int hid_submit_out(struct hid_device *hid)
 	report = usbhid->out[usbhid->outtail].report;
 	raw_report = usbhid->out[usbhid->outtail].raw_report;
 
-	usbhid->urbout->transfer_buffer_length = ((report->size - 1) >> 3) +
-						 1 + (report->id > 0);
+	usbhid->urbout->transfer_buffer_length = hid_report_len(report);
 	usbhid->urbout->dev = hid_to_usb_dev(hid);
 	if (raw_report) {
 		memcpy(usbhid->outbuf, raw_report,
@@ -702,6 +689,7 @@ int usbhid_open(struct hid_device *hid)
 			goto done;
 		}
 		usbhid->intf->needs_remote_wakeup = 1;
+		set_bit(HID_RESUME_RUNNING, &usbhid->iofl);
 		res = hid_start_in(hid);
 		if (res) {
 			if (res != -ENOSPC) {
@@ -715,6 +703,15 @@ int usbhid_open(struct hid_device *hid)
 			}
 		}
 		usb_autopm_put_interface(usbhid->intf);
+
+		/*
+		 * In case events are generated while nobody was listening,
+		 * some are released when the device is re-opened.
+		 * Wait 50 msec for the queue to empty before allowing events
+		 * to go through hid.
+		 */
+		msleep(50);
+		clear_bit(HID_RESUME_RUNNING, &usbhid->iofl);
 	}
 done:
 	mutex_unlock(&hid_open_mut);
@@ -735,8 +732,10 @@ void usbhid_close(struct hid_device *hid)
 	if (!--hid->open) {
 		spin_unlock_irq(&usbhid->lock);
 		hid_cancel_delayed_stuff(usbhid);
-		usb_kill_urb(usbhid->urbin);
-		usbhid->intf->needs_remote_wakeup = 0;
+		if (!(hid->quirks & HID_QUIRK_ALWAYS_POLL)) {
+			usb_kill_urb(usbhid->urbin);
+			usbhid->intf->needs_remote_wakeup = 0;
+		}
 	} else {
 		spin_unlock_irq(&usbhid->lock);
 	}
@@ -1134,6 +1133,19 @@ static int usbhid_start(struct hid_device *hid)
 
 	set_bit(HID_STARTED, &usbhid->iofl);
 
+	if (hid->quirks & HID_QUIRK_ALWAYS_POLL) {
+		ret = usb_autopm_get_interface(usbhid->intf);
+		if (ret)
+			goto fail;
+		usbhid->intf->needs_remote_wakeup = 1;
+		ret = hid_start_in(hid);
+		if (ret) {
+			dev_err(&hid->dev,
+				"failed to start in urb: %d\n", ret);
+		}
+		usb_autopm_put_interface(usbhid->intf);
+	}
+
 	/* Some keyboards don't work until their LEDs have been set.
 	 * Since BIOSes do set the LEDs, it must be safe for any device
 	 * that supports the keyboard boot protocol.
@@ -1166,6 +1178,9 @@ static void usbhid_stop(struct hid_device *hid)
 	if (WARN_ON(!usbhid))
 		return;
 
+	if (hid->quirks & HID_QUIRK_ALWAYS_POLL)
+		usbhid->intf->needs_remote_wakeup = 0;
+
 	clear_bit(HID_STARTED, &usbhid->iofl);
 	spin_lock_irq(&usbhid->lock);	/* Sync with error and led handlers */
 	set_bit(HID_DISCONNECTED, &usbhid->iofl);
diff --git a/drivers/hid/usbhid/hid-quirks.c b/drivers/hid/usbhid/hid-quirks.c
index 15225f3eaed1..b27b3d33ebab 100644
--- a/drivers/hid/usbhid/hid-quirks.c
+++ b/drivers/hid/usbhid/hid-quirks.c
@@ -70,15 +70,23 @@ static const struct hid_blacklist {
 	{ USB_VENDOR_ID_CH, USB_DEVICE_ID_CH_3AXIS_5BUTTON_STICK, HID_QUIRK_NOGET },
 	{ USB_VENDOR_ID_CH, USB_DEVICE_ID_CH_AXIS_295, HID_QUIRK_NOGET },
 	{ USB_VENDOR_ID_DMI, USB_DEVICE_ID_DMI_ENC, HID_QUIRK_NOGET },
+	{ USB_VENDOR_ID_ELAN, USB_DEVICE_ID_ELAN_TOUCHSCREEN, HID_QUIRK_ALWAYS_POLL },
+	{ USB_VENDOR_ID_ELAN, USB_DEVICE_ID_ELAN_TOUCHSCREEN_009B, HID_QUIRK_ALWAYS_POLL },
+	{ USB_VENDOR_ID_ELAN, USB_DEVICE_ID_ELAN_TOUCHSCREEN_0103, HID_QUIRK_ALWAYS_POLL },
+	{ USB_VENDOR_ID_ELAN, USB_DEVICE_ID_ELAN_TOUCHSCREEN_010c, HID_QUIRK_ALWAYS_POLL },
+	{ USB_VENDOR_ID_ELAN, USB_DEVICE_ID_ELAN_TOUCHSCREEN_016F, HID_QUIRK_ALWAYS_POLL },
 	{ USB_VENDOR_ID_ELO, USB_DEVICE_ID_ELO_TS2700, HID_QUIRK_NOGET },
 	{ USB_VENDOR_ID_FORMOSA, USB_DEVICE_ID_FORMOSA_IR_RECEIVER, HID_QUIRK_NO_INIT_REPORTS },
 	{ USB_VENDOR_ID_FREESCALE, USB_DEVICE_ID_FREESCALE_MX28, HID_QUIRK_NOGET },
 	{ USB_VENDOR_ID_MGE, USB_DEVICE_ID_MGE_UPS, HID_QUIRK_NOGET },
+	{ USB_VENDOR_ID_MICROSOFT, USB_DEVICE_ID_MS_TYPE_COVER_3, HID_QUIRK_NO_INIT_REPORTS },
 	{ USB_VENDOR_ID_MSI, USB_DEVICE_ID_MSI_GT683R_LED_PANEL, HID_QUIRK_NO_INIT_REPORTS },
 	{ USB_VENDOR_ID_NEXIO, USB_DEVICE_ID_NEXIO_MULTITOUCH_PTI0750, HID_QUIRK_NO_INIT_REPORTS },
 	{ USB_VENDOR_ID_NOVATEK, USB_DEVICE_ID_NOVATEK_MOUSE, HID_QUIRK_NO_INIT_REPORTS },
 	{ USB_VENDOR_ID_PENMOUNT, USB_DEVICE_ID_PENMOUNT_1610, HID_QUIRK_NOGET },
 	{ USB_VENDOR_ID_PENMOUNT, USB_DEVICE_ID_PENMOUNT_1640, HID_QUIRK_NOGET },
+	{ USB_VENDOR_ID_PIXART, USB_DEVICE_ID_PIXART_USB_OPTICAL_MOUSE, HID_QUIRK_ALWAYS_POLL },
+	{ USB_VENDOR_ID_KYE, USB_DEVICE_ID_PIXART_USB_OPTICAL_MOUSE_ID2, HID_QUIRK_ALWAYS_POLL },
 	{ USB_VENDOR_ID_PIXART, USB_DEVICE_ID_PIXART_OPTICAL_TOUCH_SCREEN, HID_QUIRK_NO_INIT_REPORTS },
 	{ USB_VENDOR_ID_PIXART, USB_DEVICE_ID_PIXART_OPTICAL_TOUCH_SCREEN1, HID_QUIRK_NO_INIT_REPORTS },
 	{ USB_VENDOR_ID_PIXART, USB_DEVICE_ID_PIXART_OPTICAL_TOUCH_SCREEN2, HID_QUIRK_NO_INIT_REPORTS },
@@ -116,6 +124,7 @@ static const struct hid_blacklist {
 	{ USB_VENDOR_ID_CHICONY, USB_DEVICE_ID_CHICONY_WIRELESS, HID_QUIRK_MULTI_INPUT },
 	{ USB_VENDOR_ID_SIGMA_MICRO, USB_DEVICE_ID_SIGMA_MICRO_KEYBOARD, HID_QUIRK_NO_INIT_REPORTS },
 	{ USB_VENDOR_ID_KYE, USB_DEVICE_ID_KYE_MOUSEPEN_I608X, HID_QUIRK_MULTI_INPUT },
+	{ USB_VENDOR_ID_KYE, USB_DEVICE_ID_KYE_MOUSEPEN_I608X_2, HID_QUIRK_MULTI_INPUT },
 	{ USB_VENDOR_ID_KYE, USB_DEVICE_ID_KYE_EASYPEN_M610X, HID_QUIRK_MULTI_INPUT },
 	{ USB_VENDOR_ID_NTRIG, USB_DEVICE_ID_NTRIG_DUOSENSE, HID_QUIRK_NO_INIT_REPORTS },
 	{ USB_VENDOR_ID_SEMICO, USB_DEVICE_ID_SEMICO_USB_KEYKOARD, HID_QUIRK_NO_INIT_REPORTS },
diff --git a/drivers/hid/usbhid/usbhid.h b/drivers/hid/usbhid/usbhid.h
index f633c24ce28b..807922b49aa4 100644
--- a/drivers/hid/usbhid/usbhid.h
+++ b/drivers/hid/usbhid/usbhid.h
@@ -52,6 +52,7 @@ struct usb_interface *usbhid_find_interface(int minor);
 #define HID_STARTED		8
 #define HID_KEYS_PRESSED	10
 #define HID_NO_BANDWIDTH	11
+#define HID_RESUME_RUNNING	12
 
 /*
  * USB-specific HID struct, to be pointed to
diff --git a/drivers/hid/wacom.h b/drivers/hid/wacom.h
index 64bc1b296d91..7db432809e9e 100644
--- a/drivers/hid/wacom.h
+++ b/drivers/hid/wacom.h
@@ -89,6 +89,7 @@
 #include <linux/slab.h>
 #include <linux/module.h>
 #include <linux/mod_devicetable.h>
+#include <linux/hid.h>
 #include <linux/usb/input.h>
 #include <linux/power_supply.h>
 #include <asm/unaligned.h>
@@ -139,8 +140,13 @@ extern const struct hid_device_id wacom_ids[];
 
 void wacom_wac_irq(struct wacom_wac *wacom_wac, size_t len);
 void wacom_setup_device_quirks(struct wacom_features *features);
-int wacom_setup_input_capabilities(struct input_dev *input_dev,
+int wacom_setup_pentouch_input_capabilities(struct input_dev *input_dev,
 				   struct wacom_wac *wacom_wac);
 int wacom_setup_pad_input_capabilities(struct input_dev *input_dev,
 				       struct wacom_wac *wacom_wac);
+void wacom_wac_usage_mapping(struct hid_device *hdev,
+		struct hid_field *field, struct hid_usage *usage);
+int wacom_wac_event(struct hid_device *hdev, struct hid_field *field,
+		struct hid_usage *usage, __s32 value);
+void wacom_wac_report(struct hid_device *hdev, struct hid_report *report);
 #endif
diff --git a/drivers/hid/wacom_sys.c b/drivers/hid/wacom_sys.c
index f0db7eca9023..654202941d30 100644
--- a/drivers/hid/wacom_sys.c
+++ b/drivers/hid/wacom_sys.c
@@ -13,23 +13,27 @@
 
 #include "wacom_wac.h"
 #include "wacom.h"
-#include <linux/hid.h>
+#include <linux/input/mt.h>
 
 #define WAC_MSG_RETRIES		5
 
+#define WAC_CMD_WL_LED_CONTROL	0x03
 #define WAC_CMD_LED_CONTROL	0x20
 #define WAC_CMD_ICON_START	0x21
 #define WAC_CMD_ICON_XFER	0x23
 #define WAC_CMD_ICON_BT_XFER	0x26
 #define WAC_CMD_RETRIES		10
 
-static int wacom_get_report(struct hid_device *hdev, u8 type, u8 id,
-			    void *buf, size_t size, unsigned int retries)
+#define DEV_ATTR_RW_PERM (S_IRUGO | S_IWUSR | S_IWGRP)
+#define DEV_ATTR_WO_PERM (S_IWUSR | S_IWGRP)
+
+static int wacom_get_report(struct hid_device *hdev, u8 type, u8 *buf,
+			    size_t size, unsigned int retries)
 {
 	int retval;
 
 	do {
-		retval = hid_hw_raw_request(hdev, id, buf, size, type,
+		retval = hid_hw_raw_request(hdev, buf[0], buf, size, type,
 				HID_REQ_GET_REPORT);
 	} while ((retval == -ETIMEDOUT || retval == -EPIPE) && --retries);
 
@@ -67,22 +71,15 @@ static int wacom_raw_event(struct hid_device *hdev, struct hid_report *report,
 static int wacom_open(struct input_dev *dev)
 {
 	struct wacom *wacom = input_get_drvdata(dev);
-	int retval;
-
-	mutex_lock(&wacom->lock);
-	retval = hid_hw_open(wacom->hdev);
-	mutex_unlock(&wacom->lock);
 
-	return retval;
+	return hid_hw_open(wacom->hdev);
 }
 
 static void wacom_close(struct input_dev *dev)
 {
 	struct wacom *wacom = input_get_drvdata(dev);
 
-	mutex_lock(&wacom->lock);
 	hid_hw_close(wacom->hdev);
-	mutex_unlock(&wacom->lock);
 }
 
 /*
@@ -106,12 +103,35 @@ static void wacom_feature_mapping(struct hid_device *hdev,
 {
 	struct wacom *wacom = hid_get_drvdata(hdev);
 	struct wacom_features *features = &wacom->wacom_wac.features;
+	struct hid_data *hid_data = &wacom->wacom_wac.hid_data;
+	u8 *data;
+	int ret;
 
 	switch (usage->hid) {
 	case HID_DG_CONTACTMAX:
 		/* leave touch_max as is if predefined */
-		if (!features->touch_max)
-			features->touch_max = field->value[0];
+		if (!features->touch_max) {
+			/* read manually */
+			data = kzalloc(2, GFP_KERNEL);
+			if (!data)
+				break;
+			data[0] = field->report->id;
+			ret = wacom_get_report(hdev, HID_FEATURE_REPORT,
+						data, 2, 0);
+			if (ret == 2)
+				features->touch_max = data[1];
+			kfree(data);
+		}
+		break;
+	case HID_DG_INPUTMODE:
+		/* Ignore if value index is out of bounds. */
+		if (usage->usage_index >= field->report_count) {
+			dev_err(&hdev->dev, "HID_DG_INPUTMODE out of range\n");
+			break;
+		}
+
+		hid_data->inputmode = field->report->id;
+		hid_data->inputmode_index = usage->usage_index;
 		break;
 	}
 }
@@ -166,9 +186,15 @@ static void wacom_usage_mapping(struct hid_device *hdev,
 	if (!pen && !finger)
 		return;
 
-	if (finger && !features->touch_max)
-		/* touch device at least supports one touch point */
-		features->touch_max = 1;
+	/*
+	 * Bamboo models do not support HID_DG_CONTACTMAX.
+	 * And, Bamboo Pen only descriptor contains touch.
+	 */
+	if (features->type != BAMBOO_PT) {
+		/* ISDv4 touch devices at least supports one touch point */
+		if (finger && !features->touch_max)
+			features->touch_max = 1;
+	}
 
 	switch (usage->hid) {
 	case HID_GD_X:
@@ -199,6 +225,24 @@ static void wacom_usage_mapping(struct hid_device *hdev,
 			features->pressure_max = field->logical_maximum;
 		break;
 	}
+
+	if (features->type == HID_GENERIC)
+		wacom_wac_usage_mapping(hdev, field, usage);
+}
+
+static void wacom_post_parse_hid(struct hid_device *hdev,
+				 struct wacom_features *features)
+{
+	struct wacom *wacom = hid_get_drvdata(hdev);
+	struct wacom_wac *wacom_wac = &wacom->wacom_wac;
+
+	if (features->type == HID_GENERIC) {
+		/* Any last-minute generic device setup */
+		if (features->touch_max > 1) {
+			input_mt_init_slots(wacom_wac->input, wacom_wac->features.touch_max,
+				    INPUT_MT_DIRECT);
+		}
+	}
 }
 
 static void wacom_parse_hid(struct hid_device *hdev,
@@ -235,6 +279,27 @@ static void wacom_parse_hid(struct hid_device *hdev,
 				wacom_usage_mapping(hdev, hreport->field[i],
 						hreport->field[i]->usage + j);
 	}
+
+	wacom_post_parse_hid(hdev, features);
+}
+
+static int wacom_hid_set_device_mode(struct hid_device *hdev)
+{
+	struct wacom *wacom = hid_get_drvdata(hdev);
+	struct hid_data *hid_data = &wacom->wacom_wac.hid_data;
+	struct hid_report *r;
+	struct hid_report_enum *re;
+
+	if (hid_data->inputmode < 0)
+		return 0;
+
+	re = &(hdev->report_enum[HID_FEATURE_REPORT]);
+	r = re->report_id_hash[hid_data->inputmode];
+	if (r) {
+		r->field[0]->value[hid_data->inputmode_index] = 2;
+		hid_hw_request(hdev, r, HID_REQ_SET_REPORT);
+	}
+	return 0;
 }
 
 static int wacom_set_device_mode(struct hid_device *hdev, int report_id,
@@ -255,7 +320,7 @@ static int wacom_set_device_mode(struct hid_device *hdev, int report_id,
 					 length, 1);
 		if (error >= 0)
 			error = wacom_get_report(hdev, HID_FEATURE_REPORT,
-			                         report_id, rep_data, length, 1);
+			                         rep_data, length, 1);
 	} while ((error < 0 || rep_data[1] != mode) && limit++ < WAC_MSG_RETRIES);
 
 	kfree(rep_data);
@@ -329,6 +394,9 @@ static int wacom_query_tablet_data(struct hid_device *hdev,
 	if (hdev->bus == BUS_BLUETOOTH)
 		return wacom_bt_query_tablet_data(hdev, 1, features);
 
+	if (features->type == HID_GENERIC)
+		return wacom_hid_set_device_mode(hdev);
+
 	if (features->device_type == BTN_TOOL_FINGER) {
 		if (features->type > TABLETPC) {
 			/* MT Tablet PC touch */
@@ -487,8 +555,14 @@ static int wacom_led_control(struct wacom *wacom)
 {
 	unsigned char *buf;
 	int retval;
+	unsigned char report_id = WAC_CMD_LED_CONTROL;
+	int buf_size = 9;
 
-	buf = kzalloc(9, GFP_KERNEL);
+	if (wacom->wacom_wac.pid) { /* wireless connected */
+		report_id = WAC_CMD_WL_LED_CONTROL;
+		buf_size = 13;
+	}
+	buf = kzalloc(buf_size, GFP_KERNEL);
 	if (!buf)
 		return -ENOMEM;
 
@@ -502,9 +576,16 @@ static int wacom_led_control(struct wacom *wacom)
 		int ring_led = wacom->led.select[0] & 0x03;
 		int ring_lum = (((wacom->led.llv & 0x60) >> 5) - 1) & 0x03;
 		int crop_lum = 0;
-
-		buf[0] = WAC_CMD_LED_CONTROL;
-		buf[1] = (crop_lum << 4) | (ring_lum << 2) | (ring_led);
+		unsigned char led_bits = (crop_lum << 4) | (ring_lum << 2) | (ring_led);
+
+		buf[0] = report_id;
+		if (wacom->wacom_wac.pid) {
+			wacom_get_report(wacom->hdev, HID_FEATURE_REPORT,
+					 buf, buf_size, WAC_CMD_RETRIES);
+			buf[0] = report_id;
+			buf[4] = led_bits;
+		} else
+			buf[1] = led_bits;
 	}
 	else {
 		int led = wacom->led.select[0] | 0x4;
@@ -513,14 +594,14 @@ static int wacom_led_control(struct wacom *wacom)
 		    wacom->wacom_wac.features.type == WACOM_24HD)
 			led |= (wacom->led.select[1] << 4) | 0x40;
 
-		buf[0] = WAC_CMD_LED_CONTROL;
+		buf[0] = report_id;
 		buf[1] = led;
 		buf[2] = wacom->led.llv;
 		buf[3] = wacom->led.hlv;
 		buf[4] = wacom->led.img_lum;
 	}
 
-	retval = wacom_set_report(wacom->hdev, HID_FEATURE_REPORT, buf, 9,
+	retval = wacom_set_report(wacom->hdev, HID_FEATURE_REPORT, buf, buf_size,
 				  WAC_CMD_RETRIES);
 	kfree(buf);
 
@@ -602,9 +683,10 @@ static ssize_t wacom_led##SET_ID##_select_show(struct device *dev,	\
 {									\
 	struct hid_device *hdev = container_of(dev, struct hid_device, dev);\
 	struct wacom *wacom = hid_get_drvdata(hdev);			\
-	return snprintf(buf, 2, "%d\n", wacom->led.select[SET_ID]);	\
+	return scnprintf(buf, PAGE_SIZE, "%d\n",			\
+			 wacom->led.select[SET_ID]);			\
 }									\
-static DEVICE_ATTR(status_led##SET_ID##_select, S_IWUSR | S_IRUSR,	\
+static DEVICE_ATTR(status_led##SET_ID##_select, DEV_ATTR_RW_PERM,	\
 		    wacom_led##SET_ID##_select_show,			\
 		    wacom_led##SET_ID##_select_store)
 
@@ -641,8 +723,15 @@ static ssize_t wacom_##name##_luminance_store(struct device *dev,	\
 	return wacom_luminance_store(wacom, &wacom->led.field,		\
 				     buf, count);			\
 }									\
-static DEVICE_ATTR(name##_luminance, S_IWUSR,				\
-		   NULL, wacom_##name##_luminance_store)
+static ssize_t wacom_##name##_luminance_show(struct device *dev,	\
+	struct device_attribute *attr, char *buf)			\
+{									\
+	struct wacom *wacom = dev_get_drvdata(dev);			\
+	return scnprintf(buf, PAGE_SIZE, "%d\n", wacom->led.field);	\
+}									\
+static DEVICE_ATTR(name##_luminance, DEV_ATTR_RW_PERM,			\
+		   wacom_##name##_luminance_show,			\
+		   wacom_##name##_luminance_store)
 
 DEVICE_LUMINANCE_ATTR(status0, llv);
 DEVICE_LUMINANCE_ATTR(status1, hlv);
@@ -683,7 +772,7 @@ static ssize_t wacom_btnimg##BUTTON_ID##_store(struct device *dev,	\
 {									\
 	return wacom_button_image_store(dev, BUTTON_ID, buf, count);	\
 }									\
-static DEVICE_ATTR(button##BUTTON_ID##_rawimg, S_IWUSR,			\
+static DEVICE_ATTR(button##BUTTON_ID##_rawimg, DEV_ATTR_WO_PERM,	\
 		   NULL, wacom_btnimg##BUTTON_ID##_store)
 
 DEVICE_BTNIMG_ATTR(0);
@@ -989,7 +1078,7 @@ static ssize_t wacom_store_speed(struct device *dev,
 	return count;
 }
 
-static DEVICE_ATTR(speed, S_IRUGO | S_IWUSR | S_IWGRP,
+static DEVICE_ATTR(speed, DEV_ATTR_RW_PERM,
 		wacom_show_speed, wacom_store_speed);
 
 static struct input_dev *wacom_allocate_input(struct wacom *wacom)
@@ -1010,47 +1099,82 @@ static struct input_dev *wacom_allocate_input(struct wacom *wacom)
 	input_dev->uniq = hdev->uniq;
 	input_dev->id.bustype = hdev->bus;
 	input_dev->id.vendor  = hdev->vendor;
-	input_dev->id.product = hdev->product;
+	input_dev->id.product = wacom_wac->pid ? wacom_wac->pid : hdev->product;
 	input_dev->id.version = hdev->version;
 	input_set_drvdata(input_dev, wacom);
 
 	return input_dev;
 }
 
-static void wacom_unregister_inputs(struct wacom *wacom)
+static void wacom_free_inputs(struct wacom *wacom)
 {
-	if (wacom->wacom_wac.input)
-		input_unregister_device(wacom->wacom_wac.input);
-	if (wacom->wacom_wac.pad_input)
-		input_unregister_device(wacom->wacom_wac.pad_input);
-	wacom->wacom_wac.input = NULL;
-	wacom->wacom_wac.pad_input = NULL;
+	struct wacom_wac *wacom_wac = &(wacom->wacom_wac);
+
+	if (wacom_wac->input)
+		input_free_device(wacom_wac->input);
+	if (wacom_wac->pad_input)
+		input_free_device(wacom_wac->pad_input);
+	wacom_wac->input = NULL;
+	wacom_wac->pad_input = NULL;
 }
 
-static int wacom_register_inputs(struct wacom *wacom)
+static int wacom_allocate_inputs(struct wacom *wacom)
 {
 	struct input_dev *input_dev, *pad_input_dev;
 	struct wacom_wac *wacom_wac = &(wacom->wacom_wac);
-	int error;
 
 	input_dev = wacom_allocate_input(wacom);
 	pad_input_dev = wacom_allocate_input(wacom);
 	if (!input_dev || !pad_input_dev) {
-		error = -ENOMEM;
-		goto fail1;
+		wacom_free_inputs(wacom);
+		return -ENOMEM;
 	}
 
 	wacom_wac->input = input_dev;
 	wacom_wac->pad_input = pad_input_dev;
 	wacom_wac->pad_input->name = wacom_wac->pad_name;
 
-	error = wacom_setup_input_capabilities(input_dev, wacom_wac);
-	if (error)
-		goto fail2;
+	return 0;
+}
+
+static void wacom_clean_inputs(struct wacom *wacom)
+{
+	if (wacom->wacom_wac.input) {
+		if (wacom->wacom_wac.input_registered)
+			input_unregister_device(wacom->wacom_wac.input);
+		else
+			input_free_device(wacom->wacom_wac.input);
+	}
+	if (wacom->wacom_wac.pad_input) {
+		if (wacom->wacom_wac.pad_registered)
+			input_unregister_device(wacom->wacom_wac.pad_input);
+		else
+			input_free_device(wacom->wacom_wac.pad_input);
+	}
+	wacom->wacom_wac.input = NULL;
+	wacom->wacom_wac.pad_input = NULL;
+	wacom_destroy_leds(wacom);
+}
+
+static int wacom_register_inputs(struct wacom *wacom)
+{
+	struct input_dev *input_dev, *pad_input_dev;
+	struct wacom_wac *wacom_wac = &(wacom->wacom_wac);
+	int error;
+
+	input_dev = wacom_wac->input;
+	pad_input_dev = wacom_wac->pad_input;
+
+	if (!input_dev || !pad_input_dev)
+		return -EINVAL;
 
-	error = input_register_device(input_dev);
-	if (error)
-		goto fail2;
+	error = wacom_setup_pentouch_input_capabilities(input_dev, wacom_wac);
+	if (!error) {
+		error = input_register_device(input_dev);
+		if (error)
+			return error;
+		wacom_wac->input_registered = true;
+	}
 
 	error = wacom_setup_pad_input_capabilities(pad_input_dev, wacom_wac);
 	if (error) {
@@ -1061,22 +1185,24 @@ static int wacom_register_inputs(struct wacom *wacom)
 	} else {
 		error = input_register_device(pad_input_dev);
 		if (error)
-			goto fail3;
+			goto fail_register_pad_input;
+		wacom_wac->pad_registered = true;
+
+		error = wacom_initialize_leds(wacom);
+		if (error)
+			goto fail_leds;
 	}
 
 	return 0;
 
-fail3:
+fail_leds:
+	input_unregister_device(pad_input_dev);
+	pad_input_dev = NULL;
+	wacom_wac->pad_registered = false;
+fail_register_pad_input:
 	input_unregister_device(input_dev);
-	input_dev = NULL;
-fail2:
 	wacom_wac->input = NULL;
-	wacom_wac->pad_input = NULL;
-fail1:
-	if (input_dev)
-		input_free_device(input_dev);
-	if (pad_input_dev)
-		input_free_device(pad_input_dev);
+	wacom_wac->input_registered = false;
 	return error;
 }
 
@@ -1101,13 +1227,13 @@ static void wacom_wireless_work(struct work_struct *work)
 	hdev1 = usb_get_intfdata(usbdev->config->interface[1]);
 	wacom1 = hid_get_drvdata(hdev1);
 	wacom_wac1 = &(wacom1->wacom_wac);
-	wacom_unregister_inputs(wacom1);
+	wacom_clean_inputs(wacom1);
 
 	/* Touch interface */
 	hdev2 = usb_get_intfdata(usbdev->config->interface[2]);
 	wacom2 = hid_get_drvdata(hdev2);
 	wacom_wac2 = &(wacom2->wacom_wac);
-	wacom_unregister_inputs(wacom2);
+	wacom_clean_inputs(wacom2);
 
 	if (wacom_wac->pid == 0) {
 		hid_info(wacom->hdev, "wireless tablet disconnected\n");
@@ -1140,7 +1266,9 @@ static void wacom_wireless_work(struct work_struct *work)
 			 wacom_wac1->features.name);
 		wacom_wac1->shared->touch_max = wacom_wac1->features.touch_max;
 		wacom_wac1->shared->type = wacom_wac1->features.type;
-		error = wacom_register_inputs(wacom1);
+		wacom_wac1->pid = wacom_wac->pid;
+		error = wacom_allocate_inputs(wacom1) ||
+			wacom_register_inputs(wacom1);
 		if (error)
 			goto fail;
 
@@ -1160,7 +1288,9 @@ static void wacom_wireless_work(struct work_struct *work)
 					 "%s (WL) Pad",wacom_wac2->features.name);
 			snprintf(wacom_wac2->pad_name, WACOM_NAME_MAX,
 				 "%s (WL) Pad", wacom_wac2->features.name);
-			error = wacom_register_inputs(wacom2);
+			wacom_wac2->pid = wacom_wac->pid;
+			error = wacom_allocate_inputs(wacom2) ||
+				wacom_register_inputs(wacom2);
 			if (error)
 				goto fail;
 
@@ -1177,8 +1307,8 @@ static void wacom_wireless_work(struct work_struct *work)
 	return;
 
 fail:
-	wacom_unregister_inputs(wacom1);
-	wacom_unregister_inputs(wacom2);
+	wacom_clean_inputs(wacom1);
+	wacom_clean_inputs(wacom2);
 	return;
 }
 
@@ -1209,12 +1339,6 @@ static void wacom_calculate_res(struct wacom_features *features)
 						    features->unitExpo);
 }
 
-static int wacom_hid_report_len(struct hid_report *report)
-{
-	/* equivalent to DIV_ROUND_UP(report->size, 8) + !!(report->id > 0) */
-	return ((report->size - 1) >> 3) + 1 + (report->id > 0);
-}
-
 static size_t wacom_compute_pktlen(struct hid_device *hdev)
 {
 	struct hid_report_enum *report_enum;
@@ -1224,7 +1348,7 @@ static size_t wacom_compute_pktlen(struct hid_device *hdev)
 	report_enum = hdev->report_enum + HID_INPUT_REPORT;
 
 	list_for_each_entry(report, &report_enum->report_list, list) {
-		size_t report_size = wacom_hid_report_len(report);
+		size_t report_size = hid_report_len(report);
 		if (report_size > size)
 			size = report_size;
 	}
@@ -1241,10 +1365,13 @@ static int wacom_probe(struct hid_device *hdev,
 	struct wacom_wac *wacom_wac;
 	struct wacom_features *features;
 	int error;
+	unsigned int connect_mask = HID_CONNECT_HIDRAW;
 
 	if (!id->driver_data)
 		return -EINVAL;
 
+	hdev->quirks |= HID_QUIRK_NO_INIT_REPORTS;
+
 	wacom = kzalloc(sizeof(struct wacom), GFP_KERNEL);
 	if (!wacom)
 		return -ENOMEM;
@@ -1256,7 +1383,7 @@ static int wacom_probe(struct hid_device *hdev,
 	error = hid_parse(hdev);
 	if (error) {
 		hid_err(hdev, "parse failed\n");
-		goto fail1;
+		goto fail_parse;
 	}
 
 	wacom_wac = &wacom->wacom_wac;
@@ -1265,12 +1392,12 @@ static int wacom_probe(struct hid_device *hdev,
 	features->pktlen = wacom_compute_pktlen(hdev);
 	if (features->pktlen > WACOM_PKGLEN_MAX) {
 		error = -EINVAL;
-		goto fail1;
+		goto fail_pktlen;
 	}
 
 	if (features->check_for_hid_type && features->hid_type != hdev->type) {
 		error = -ENODEV;
-		goto fail1;
+		goto fail_type;
 	}
 
 	wacom->usbdev = dev;
@@ -1278,6 +1405,12 @@ static int wacom_probe(struct hid_device *hdev,
 	mutex_init(&wacom->lock);
 	INIT_WORK(&wacom->work, wacom_wireless_work);
 
+	if (!(features->quirks & WACOM_QUIRK_NO_INPUT)) {
+		error = wacom_allocate_inputs(wacom);
+		if (error)
+			goto fail_allocate_inputs;
+	}
+
 	/* set the default size in case we do not get them from hid */
 	wacom_set_default_phy(features);
 
@@ -1339,24 +1472,20 @@ static int wacom_probe(struct hid_device *hdev,
 
 		error = wacom_add_shared_data(hdev);
 		if (error)
-			goto fail1;
+			goto fail_shared_data;
 	}
 
-	error = wacom_initialize_leds(wacom);
-	if (error)
-		goto fail2;
-
 	if (!(features->quirks & WACOM_QUIRK_MONITOR) &&
 	     (features->quirks & WACOM_QUIRK_BATTERY)) {
 		error = wacom_initialize_battery(wacom);
 		if (error)
-			goto fail3;
+			goto fail_battery;
 	}
 
 	if (!(features->quirks & WACOM_QUIRK_NO_INPUT)) {
 		error = wacom_register_inputs(wacom);
 		if (error)
-			goto fail4;
+			goto fail_register_inputs;
 	}
 
 	if (hdev->bus == BUS_BLUETOOTH) {
@@ -1367,16 +1496,19 @@ static int wacom_probe(struct hid_device *hdev,
 				 error);
 	}
 
-	/* Note that if query fails it is not a hard failure */
-	wacom_query_tablet_data(hdev, features);
+	if (features->type == HID_GENERIC)
+		connect_mask |= HID_CONNECT_DRIVER;
 
 	/* Regular HID work starts now */
-	error = hid_hw_start(hdev, HID_CONNECT_HIDRAW);
+	error = hid_hw_start(hdev, connect_mask);
 	if (error) {
 		hid_err(hdev, "hw start failed\n");
-		goto fail5;
+		goto fail_hw_start;
 	}
 
+	/* Note that if query fails it is not a hard failure */
+	wacom_query_tablet_data(hdev, features);
+
 	if (features->quirks & WACOM_QUIRK_MONITOR)
 		error = hid_hw_open(hdev);
 
@@ -1387,13 +1519,21 @@ static int wacom_probe(struct hid_device *hdev,
 
 	return 0;
 
- fail5:	if (hdev->bus == BUS_BLUETOOTH)
+fail_hw_start:
+	if (hdev->bus == BUS_BLUETOOTH)
 		device_remove_file(&hdev->dev, &dev_attr_speed);
-	wacom_unregister_inputs(wacom);
- fail4:	wacom_destroy_battery(wacom);
- fail3:	wacom_destroy_leds(wacom);
- fail2:	wacom_remove_shared_data(wacom_wac);
- fail1:	kfree(wacom);
+fail_register_inputs:
+	wacom_clean_inputs(wacom);
+	wacom_destroy_battery(wacom);
+fail_battery:
+	wacom_remove_shared_data(wacom_wac);
+fail_shared_data:
+	wacom_clean_inputs(wacom);
+fail_allocate_inputs:
+fail_type:
+fail_pktlen:
+fail_parse:
+	kfree(wacom);
 	hid_set_drvdata(hdev, NULL);
 	return error;
 }
@@ -1405,11 +1545,10 @@ static void wacom_remove(struct hid_device *hdev)
 	hid_hw_stop(hdev);
 
 	cancel_work_sync(&wacom->work);
-	wacom_unregister_inputs(wacom);
+	wacom_clean_inputs(wacom);
 	if (hdev->bus == BUS_BLUETOOTH)
 		device_remove_file(&hdev->dev, &dev_attr_speed);
 	wacom_destroy_battery(wacom);
-	wacom_destroy_leds(wacom);
 	wacom_remove_shared_data(&wacom->wacom_wac);
 
 	hid_set_drvdata(hdev, NULL);
@@ -1444,6 +1583,8 @@ static struct hid_driver wacom_driver = {
 	.id_table =	wacom_ids,
 	.probe =	wacom_probe,
 	.remove =	wacom_remove,
+	.event =	wacom_wac_event,
+	.report =	wacom_wac_report,
 #ifdef CONFIG_PM
 	.resume =	wacom_resume,
 	.reset_resume =	wacom_reset_resume,
diff --git a/drivers/hid/wacom_wac.c b/drivers/hid/wacom_wac.c
index aa6a08eb7ad6..ac7447c7b82e 100644
--- a/drivers/hid/wacom_wac.c
+++ b/drivers/hid/wacom_wac.c
@@ -25,6 +25,10 @@
 #define WACOM_INTUOS_RES	100
 #define WACOM_INTUOS3_RES	200
 
+/* Newer Cintiq and DTU have an offset between tablet and screen areas */
+#define WACOM_DTU_OFFSET	200
+#define WACOM_CINTIQ_OFFSET	400
+
 /*
  * Scale factor relating reported contact size to logical contact area.
  * 2^14/pi is a good approximation on Intuos5 and 3rd-gen Bamboo
@@ -600,8 +604,8 @@ static void wacom_intuos_general(struct wacom_wac *wacom)
 		}
 		input_report_abs(input, ABS_PRESSURE, t);
 		input_report_abs(input, ABS_TILT_X,
-				((data[7] << 1) & 0x7e) | (data[8] >> 7));
-		input_report_abs(input, ABS_TILT_Y, data[8] & 0x7f);
+				 (((data[7] << 1) & 0x7e) | (data[8] >> 7)) - 64);
+		input_report_abs(input, ABS_TILT_Y, (data[8] & 0x7f) - 64);
 		input_report_key(input, BTN_STYLUS, data[1] & 2);
 		input_report_key(input, BTN_STYLUS2, data[1] & 4);
 		input_report_key(input, BTN_TOUCH, t > 10);
@@ -612,8 +616,8 @@ static void wacom_intuos_general(struct wacom_wac *wacom)
 		input_report_abs(input, ABS_WHEEL,
 				(data[6] << 2) | ((data[7] >> 6) & 3));
 		input_report_abs(input, ABS_TILT_X,
-				((data[7] << 1) & 0x7e) | (data[8] >> 7));
-		input_report_abs(input, ABS_TILT_Y, data[8] & 0x7f);
+				 (((data[7] << 1) & 0x7e) | (data[8] >> 7)) - 64);
+		input_report_abs(input, ABS_TILT_Y, (data[8] & 0x7f) - 64);
 	}
 }
 
@@ -915,8 +919,8 @@ static int wacom_intuos_irq(struct wacom_wac *wacom)
 				input_report_key(input, BTN_EXTRA,  data[6] & 0x10);
 
 				input_report_abs(input, ABS_TILT_X,
-					((data[7] << 1) & 0x7e) | (data[8] >> 7));
-				input_report_abs(input, ABS_TILT_Y, data[8] & 0x7f);
+					(((data[7] << 1) & 0x7e) | (data[8] >> 7)) - 64);
+				input_report_abs(input, ABS_TILT_Y, (data[8] & 0x7f) - 64);
 			} else {
 				/* 2D mouse packet */
 				input_report_key(input, BTN_LEFT,   data[8] & 0x04);
@@ -1248,6 +1252,325 @@ static int wacom_tpc_irq(struct wacom_wac *wacom, size_t len)
 	return 0;
 }
 
+static void wacom_map_usage(struct wacom *wacom, struct hid_usage *usage,
+		struct hid_field *field, __u8 type, __u16 code, int fuzz)
+{
+	struct wacom_wac *wacom_wac = &wacom->wacom_wac;
+	struct input_dev *input = wacom_wac->input;
+	int fmin = field->logical_minimum;
+	int fmax = field->logical_maximum;
+
+	usage->type = type;
+	usage->code = code;
+
+	set_bit(type, input->evbit);
+
+	switch (type) {
+	case EV_ABS:
+		input_set_abs_params(input, code, fmin, fmax, fuzz, 0);
+		input_abs_set_res(input, code,
+				  hidinput_calc_abs_res(field, code));
+		break;
+	case EV_KEY:
+		input_set_capability(input, EV_KEY, code);
+		break;
+	case EV_MSC:
+		input_set_capability(input, EV_MSC, code);
+		break;
+	}
+}
+
+static void wacom_wac_pen_usage_mapping(struct hid_device *hdev,
+		struct hid_field *field, struct hid_usage *usage)
+{
+	struct wacom *wacom = hid_get_drvdata(hdev);
+
+	switch (usage->hid) {
+	case HID_GD_X:
+		wacom_map_usage(wacom, usage, field, EV_ABS, ABS_X, 4);
+		break;
+	case HID_GD_Y:
+		wacom_map_usage(wacom, usage, field, EV_ABS, ABS_Y, 4);
+		break;
+	case HID_DG_TIPPRESSURE:
+		wacom_map_usage(wacom, usage, field, EV_ABS, ABS_PRESSURE, 0);
+		break;
+	case HID_DG_INRANGE:
+		wacom_map_usage(wacom, usage, field, EV_KEY, BTN_TOOL_PEN, 0);
+		break;
+	case HID_DG_INVERT:
+		wacom_map_usage(wacom, usage, field, EV_KEY,
+				BTN_TOOL_RUBBER, 0);
+		break;
+	case HID_DG_ERASER:
+	case HID_DG_TIPSWITCH:
+		wacom_map_usage(wacom, usage, field, EV_KEY, BTN_TOUCH, 0);
+		break;
+	case HID_DG_BARRELSWITCH:
+		wacom_map_usage(wacom, usage, field, EV_KEY, BTN_STYLUS, 0);
+		break;
+	case HID_DG_BARRELSWITCH2:
+		wacom_map_usage(wacom, usage, field, EV_KEY, BTN_STYLUS2, 0);
+		break;
+	case HID_DG_TOOLSERIALNUMBER:
+		wacom_map_usage(wacom, usage, field, EV_MSC, MSC_SERIAL, 0);
+		break;
+	}
+}
+
+static int wacom_wac_pen_event(struct hid_device *hdev, struct hid_field *field,
+		struct hid_usage *usage, __s32 value)
+{
+	struct wacom *wacom = hid_get_drvdata(hdev);
+	struct wacom_wac *wacom_wac = &wacom->wacom_wac;
+	struct input_dev *input = wacom_wac->input;
+
+	/* checking which Tool / tip switch to send */
+	switch (usage->hid) {
+	case HID_DG_INRANGE:
+		wacom_wac->hid_data.inrange_state = value;
+		return 0;
+	case HID_DG_INVERT:
+		wacom_wac->hid_data.invert_state = value;
+		return 0;
+	case HID_DG_ERASER:
+	case HID_DG_TIPSWITCH:
+		wacom_wac->hid_data.tipswitch |= value;
+		return 0;
+	}
+
+	/* send pen events only when touch is up or forced out */
+	if (!usage->type || wacom_wac->shared->touch_down)
+		return 0;
+
+	input_event(input, usage->type, usage->code, value);
+
+	return 0;
+}
+
+static void wacom_wac_pen_report(struct hid_device *hdev,
+		struct hid_report *report)
+{
+	struct wacom *wacom = hid_get_drvdata(hdev);
+	struct wacom_wac *wacom_wac = &wacom->wacom_wac;
+	struct input_dev *input = wacom_wac->input;
+	bool prox = wacom_wac->hid_data.inrange_state;
+
+	if (!wacom_wac->shared->stylus_in_proximity) /* first in prox */
+		/* Going into proximity select tool */
+		wacom_wac->tool[0] = wacom_wac->hid_data.invert_state ?
+						BTN_TOOL_RUBBER : BTN_TOOL_PEN;
+
+	/* keep pen state for touch events */
+	wacom_wac->shared->stylus_in_proximity = prox;
+
+	/* send pen events only when touch is up or forced out */
+	if (!wacom_wac->shared->touch_down) {
+		input_report_key(input, BTN_TOUCH,
+				wacom_wac->hid_data.tipswitch);
+		input_report_key(input, wacom_wac->tool[0], prox);
+
+		wacom_wac->hid_data.tipswitch = false;
+
+		input_sync(input);
+	}
+}
+
+static void wacom_wac_finger_usage_mapping(struct hid_device *hdev,
+		struct hid_field *field, struct hid_usage *usage)
+{
+	struct wacom *wacom = hid_get_drvdata(hdev);
+	struct wacom_wac *wacom_wac = &wacom->wacom_wac;
+	struct wacom_features *features = &wacom_wac->features;
+	unsigned touch_max = wacom_wac->features.touch_max;
+
+	switch (usage->hid) {
+	case HID_GD_X:
+		features->last_slot_field = usage->hid;
+		if (touch_max == 1)
+			wacom_map_usage(wacom, usage, field, EV_ABS, ABS_X, 4);
+		else
+			wacom_map_usage(wacom, usage, field, EV_ABS,
+					ABS_MT_POSITION_X, 4);
+		break;
+	case HID_GD_Y:
+		features->last_slot_field = usage->hid;
+		if (touch_max == 1)
+			wacom_map_usage(wacom, usage, field, EV_ABS, ABS_Y, 4);
+		else
+			wacom_map_usage(wacom, usage, field, EV_ABS,
+					ABS_MT_POSITION_Y, 4);
+		break;
+	case HID_DG_CONTACTID:
+		features->last_slot_field = usage->hid;
+		break;
+	case HID_DG_INRANGE:
+		features->last_slot_field = usage->hid;
+		break;
+	case HID_DG_INVERT:
+		features->last_slot_field = usage->hid;
+		break;
+	case HID_DG_TIPSWITCH:
+		features->last_slot_field = usage->hid;
+		wacom_map_usage(wacom, usage, field, EV_KEY, BTN_TOUCH, 0);
+		break;
+	}
+}
+
+static void wacom_wac_finger_slot(struct wacom_wac *wacom_wac,
+		struct input_dev *input)
+{
+	struct hid_data *hid_data = &wacom_wac->hid_data;
+	bool mt = wacom_wac->features.touch_max > 1;
+	bool prox = hid_data->tipswitch &&
+		    !wacom_wac->shared->stylus_in_proximity;
+
+	if (mt) {
+		int slot;
+
+		slot = input_mt_get_slot_by_key(input, hid_data->id);
+		input_mt_slot(input, slot);
+		input_mt_report_slot_state(input, MT_TOOL_FINGER, prox);
+	}
+	else {
+		input_report_key(input, BTN_TOUCH, prox);
+	}
+
+	if (prox) {
+		input_report_abs(input, mt ? ABS_MT_POSITION_X : ABS_X,
+				 hid_data->x);
+		input_report_abs(input, mt ? ABS_MT_POSITION_Y : ABS_Y,
+				 hid_data->y);
+	}
+}
+
+static int wacom_wac_finger_event(struct hid_device *hdev,
+		struct hid_field *field, struct hid_usage *usage, __s32 value)
+{
+	struct wacom *wacom = hid_get_drvdata(hdev);
+	struct wacom_wac *wacom_wac = &wacom->wacom_wac;
+
+	switch (usage->hid) {
+	case HID_GD_X:
+		wacom_wac->hid_data.x = value;
+		break;
+	case HID_GD_Y:
+		wacom_wac->hid_data.y = value;
+		break;
+	case HID_DG_CONTACTID:
+		wacom_wac->hid_data.id = value;
+		break;
+	case HID_DG_TIPSWITCH:
+		wacom_wac->hid_data.tipswitch = value;
+		break;
+	}
+
+
+	if (usage->usage_index + 1 == field->report_count) {
+		if (usage->hid == wacom_wac->features.last_slot_field)
+			wacom_wac_finger_slot(wacom_wac, wacom_wac->input);
+	}
+
+	return 0;
+}
+
+static int wacom_wac_finger_count_touches(struct hid_device *hdev)
+{
+	struct wacom *wacom = hid_get_drvdata(hdev);
+	struct wacom_wac *wacom_wac = &wacom->wacom_wac;
+	struct input_dev *input = wacom_wac->input;
+	unsigned touch_max = wacom_wac->features.touch_max;
+	int count = 0;
+	int i;
+
+	if (touch_max == 1)
+		return wacom_wac->hid_data.tipswitch &&
+		       !wacom_wac->shared->stylus_in_proximity;
+
+	for (i = 0; i < input->mt->num_slots; i++) {
+		struct input_mt_slot *ps = &input->mt->slots[i];
+		int id = input_mt_get_value(ps, ABS_MT_TRACKING_ID);
+		if (id >= 0)
+			count++;
+	}
+
+	return count;
+}
+
+static void wacom_wac_finger_report(struct hid_device *hdev,
+		struct hid_report *report)
+{
+	struct wacom *wacom = hid_get_drvdata(hdev);
+	struct wacom_wac *wacom_wac = &wacom->wacom_wac;
+	struct input_dev *input = wacom_wac->input;
+	unsigned touch_max = wacom_wac->features.touch_max;
+
+	if (touch_max > 1)
+		input_mt_sync_frame(input);
+
+	input_sync(input);
+
+	/* keep touch state for pen event */
+	wacom_wac->shared->touch_down = wacom_wac_finger_count_touches(hdev);
+}
+
+#define WACOM_PEN_FIELD(f)	(((f)->logical == HID_DG_STYLUS) || \
+				 ((f)->physical == HID_DG_STYLUS) || \
+				 ((f)->application == HID_DG_PEN))
+#define WACOM_FINGER_FIELD(f)	(((f)->logical == HID_DG_FINGER) || \
+				 ((f)->physical == HID_DG_FINGER) || \
+				 ((f)->application == HID_DG_TOUCHSCREEN))
+
+void wacom_wac_usage_mapping(struct hid_device *hdev,
+		struct hid_field *field, struct hid_usage *usage)
+{
+	struct wacom *wacom = hid_get_drvdata(hdev);
+	struct wacom_wac *wacom_wac = &wacom->wacom_wac;
+	struct input_dev *input = wacom_wac->input;
+
+	/* currently, only direct devices have proper hid report descriptors */
+	__set_bit(INPUT_PROP_DIRECT, input->propbit);
+
+	if (WACOM_PEN_FIELD(field))
+		return wacom_wac_pen_usage_mapping(hdev, field, usage);
+
+	if (WACOM_FINGER_FIELD(field))
+		return wacom_wac_finger_usage_mapping(hdev, field, usage);
+}
+
+int wacom_wac_event(struct hid_device *hdev, struct hid_field *field,
+		struct hid_usage *usage, __s32 value)
+{
+	struct wacom *wacom = hid_get_drvdata(hdev);
+
+	if (wacom->wacom_wac.features.type != HID_GENERIC)
+		return 0;
+
+	if (WACOM_PEN_FIELD(field))
+		return wacom_wac_pen_event(hdev, field, usage, value);
+
+	if (WACOM_FINGER_FIELD(field))
+		return wacom_wac_finger_event(hdev, field, usage, value);
+
+	return 0;
+}
+
+void wacom_wac_report(struct hid_device *hdev, struct hid_report *report)
+{
+	struct wacom *wacom = hid_get_drvdata(hdev);
+	struct wacom_wac *wacom_wac = &wacom->wacom_wac;
+	struct hid_field *field = report->field[0];
+
+	if (wacom_wac->features.type != HID_GENERIC)
+		return;
+
+	if (WACOM_PEN_FIELD(field))
+		return wacom_wac_pen_report(hdev, report);
+
+	if (WACOM_FINGER_FIELD(field))
+		return wacom_wac_finger_report(hdev, report);
+}
+
 static int wacom_bpt_touch(struct wacom_wac *wacom)
 {
 	struct wacom_features *features = &wacom->features;
@@ -1391,7 +1714,9 @@ static int wacom_bpt_pen(struct wacom_wac *wacom)
 	    return 0;
 
 	if (data[0] == WACOM_REPORT_USB) {
-		if (features->type == INTUOSHT && features->touch_max) {
+		if (features->type == INTUOSHT &&
+		    wacom->shared->touch_input &&
+		    features->touch_max) {
 			input_report_switch(wacom->shared->touch_input,
 					    SW_MUTE_DEVICE, data[8] & 0x40);
 			input_sync(wacom->shared->touch_input);
@@ -1484,7 +1809,8 @@ static int wacom_wireless_irq(struct wacom_wac *wacom, size_t len)
 		int pid, battery, ps_connected;
 
 		if ((wacom->shared->type == INTUOSHT) &&
-				wacom->shared->touch_max) {
+		    wacom->shared->touch_input &&
+		    wacom->shared->touch_max) {
 			input_report_switch(wacom->shared->touch_input,
 					SW_MUTE_DEVICE, data[5] & 0x40);
 			input_sync(wacom->shared->touch_input);
@@ -1548,6 +1874,7 @@ void wacom_wac_irq(struct wacom_wac *wacom_wac, size_t len)
 		break;
 
 	case DTUS:
+	case DTUSX:
 		sync = wacom_dtus_irq(wacom_wac);
 		break;
 
@@ -1636,8 +1963,10 @@ static void wacom_setup_cintiq(struct wacom_wac *wacom_wac)
 	input_set_abs_params(input_dev, ABS_DISTANCE,
 			     0, wacom_wac->features.distance_max, 0, 0);
 	input_set_abs_params(input_dev, ABS_WHEEL, 0, 1023, 0, 0);
-	input_set_abs_params(input_dev, ABS_TILT_X, 0, 127, 0, 0);
-	input_set_abs_params(input_dev, ABS_TILT_Y, 0, 127, 0, 0);
+	input_set_abs_params(input_dev, ABS_TILT_X, -64, 63, 0, 0);
+	input_abs_set_res(input_dev, ABS_TILT_X, 57);
+	input_set_abs_params(input_dev, ABS_TILT_Y, -64, 63, 0, 0);
+	input_abs_set_res(input_dev, ABS_TILT_Y, 57);
 }
 
 static void wacom_setup_intuos(struct wacom_wac *wacom_wac)
@@ -1657,6 +1986,7 @@ static void wacom_setup_intuos(struct wacom_wac *wacom_wac)
 	__set_bit(BTN_TOOL_LENS, input_dev->keybit);
 
 	input_set_abs_params(input_dev, ABS_RZ, -900, 899, 0, 0);
+	input_abs_set_res(input_dev, ABS_RZ, 287);
 	input_set_abs_params(input_dev, ABS_THROTTLE, -1023, 1023, 0, 0);
 }
 
@@ -1739,13 +2069,17 @@ static void wacom_abs_set_axis(struct input_dev *input_dev,
 	}
 }
 
-int wacom_setup_input_capabilities(struct input_dev *input_dev,
+int wacom_setup_pentouch_input_capabilities(struct input_dev *input_dev,
 				   struct wacom_wac *wacom_wac)
 {
 	struct wacom_features *features = &wacom_wac->features;
 
 	input_dev->evbit[0] |= BIT_MASK(EV_KEY) | BIT_MASK(EV_ABS);
 
+	if (features->type == HID_GENERIC)
+		/* setup has already been done */
+		return 0;
+
 	__set_bit(BTN_TOUCH, input_dev->keybit);
 	__set_bit(ABS_MISC, input_dev->absbit);
 
@@ -1753,9 +2087,6 @@ int wacom_setup_input_capabilities(struct input_dev *input_dev,
 
 	switch (features->type) {
 	case WACOM_MO:
-		input_set_abs_params(input_dev, ABS_WHEEL, 0, 71, 0, 0);
-		/* fall through */
-
 	case WACOM_G4:
 		/* fall through */
 
@@ -1798,6 +2129,7 @@ int wacom_setup_input_capabilities(struct input_dev *input_dev,
 
 	case WACOM_24HD:
 		input_set_abs_params(input_dev, ABS_Z, -900, 899, 0, 0);
+		input_abs_set_res(input_dev, ABS_Z, 287);
 		input_set_abs_params(input_dev, ABS_THROTTLE, 0, 71, 0, 0);
 		/* fall through */
 
@@ -1812,6 +2144,7 @@ int wacom_setup_input_capabilities(struct input_dev *input_dev,
 	case WACOM_BEE:
 	case CINTIQ:
 		input_set_abs_params(input_dev, ABS_Z, -900, 899, 0, 0);
+		input_abs_set_res(input_dev, ABS_Z, 287);
 
 		__set_bit(INPUT_PROP_DIRECT, input_dev->propbit);
 
@@ -1820,6 +2153,7 @@ int wacom_setup_input_capabilities(struct input_dev *input_dev,
 
 	case WACOM_13HD:
 		input_set_abs_params(input_dev, ABS_Z, -900, 899, 0, 0);
+		input_abs_set_res(input_dev, ABS_Z, 287);
 		__set_bit(INPUT_PROP_DIRECT, input_dev->propbit);
 		wacom_setup_cintiq(wacom_wac);
 		break;
@@ -1828,6 +2162,7 @@ int wacom_setup_input_capabilities(struct input_dev *input_dev,
 	case INTUOS3L:
 	case INTUOS3S:
 		input_set_abs_params(input_dev, ABS_Z, -900, 899, 0, 0);
+		input_abs_set_res(input_dev, ABS_Z, 287);
 		/* fall through */
 
 	case INTUOS:
@@ -1850,6 +2185,7 @@ int wacom_setup_input_capabilities(struct input_dev *input_dev,
 					      0, 0);
 
 			input_set_abs_params(input_dev, ABS_Z, -900, 899, 0, 0);
+			input_abs_set_res(input_dev, ABS_Z, 287);
 
 			wacom_setup_intuos(wacom_wac);
 		} else if (features->device_type == BTN_TOOL_FINGER) {
@@ -1868,6 +2204,7 @@ int wacom_setup_input_capabilities(struct input_dev *input_dev,
 	case INTUOS4L:
 	case INTUOS4S:
 		input_set_abs_params(input_dev, ABS_Z, -900, 899, 0, 0);
+		input_abs_set_res(input_dev, ABS_Z, 287);
 		wacom_setup_intuos(wacom_wac);
 
 		__set_bit(INPUT_PROP_POINTER, input_dev->propbit);
@@ -1902,6 +2239,7 @@ int wacom_setup_input_capabilities(struct input_dev *input_dev,
 		/* fall through */
 
 	case DTUS:
+	case DTUSX:
 	case PL:
 	case DTU:
 		__set_bit(BTN_TOOL_PEN, input_dev->keybit);
@@ -1952,6 +2290,9 @@ int wacom_setup_input_capabilities(struct input_dev *input_dev,
 				__clear_bit(ABS_X, input_dev->absbit);
 				__clear_bit(ABS_Y, input_dev->absbit);
 				__clear_bit(BTN_TOUCH, input_dev->keybit);
+
+				/* PAD is setup by wacom_setup_pad_input_capabilities later */
+				return 1;
 			}
 		} else if (features->device_type == BTN_TOOL_PEN) {
 			__set_bit(INPUT_PROP_POINTER, input_dev->propbit);
@@ -1967,6 +2308,7 @@ int wacom_setup_input_capabilities(struct input_dev *input_dev,
 
 	case CINTIQ_HYBRID:
 		input_set_abs_params(input_dev, ABS_Z, -900, 899, 0, 0);
+		input_abs_set_res(input_dev, ABS_Z, 287);
 		__set_bit(INPUT_PROP_DIRECT, input_dev->propbit);
 
 		wacom_setup_cintiq(wacom_wac);
@@ -1990,6 +2332,9 @@ int wacom_setup_pad_input_capabilities(struct input_dev *input_dev,
 	input_set_abs_params(input_dev, ABS_X, 0, 1, 0, 0);
 	input_set_abs_params(input_dev, ABS_Y, 0, 1, 0, 0);
 
+	/* kept for making udev and libwacom accepting the pad */
+	__set_bit(BTN_STYLUS, input_dev->keybit);
+
 	switch (features->type) {
 	case GRAPHIRE_BT:
 		__set_bit(BTN_0, input_dev->keybit);
@@ -2006,9 +2351,7 @@ int wacom_setup_pad_input_capabilities(struct input_dev *input_dev,
 
 	case WACOM_G4:
 		__set_bit(BTN_BACK, input_dev->keybit);
-		__set_bit(BTN_LEFT, input_dev->keybit);
 		__set_bit(BTN_FORWARD, input_dev->keybit);
-		__set_bit(BTN_RIGHT, input_dev->keybit);
 		input_set_capability(input_dev, EV_REL, REL_WHEEL);
 		break;
 
@@ -2105,7 +2448,7 @@ int wacom_setup_pad_input_capabilities(struct input_dev *input_dev,
 	case INTUOSPS:
 		/* touch interface does not have the pad device */
 		if (features->device_type != BTN_TOOL_PEN)
-			return 1;
+			return -ENODEV;
 
 		for (i = 0; i < 7; i++)
 			__set_bit(BTN_0 + i, input_dev->keybit);
@@ -2149,8 +2492,10 @@ int wacom_setup_pad_input_capabilities(struct input_dev *input_dev,
 	case INTUOSHT:
 	case BAMBOO_PT:
 		/* pad device is on the touch interface */
-		if (features->device_type != BTN_TOOL_FINGER)
-			return 1;
+		if ((features->device_type != BTN_TOOL_FINGER) ||
+		    /* Bamboo Pen only tablet does not have pad */
+		    ((features->type == BAMBOO_PT) && !features->touch_max))
+			return -ENODEV;
 
 		__clear_bit(ABS_MISC, input_dev->absbit);
 
@@ -2163,7 +2508,7 @@ int wacom_setup_pad_input_capabilities(struct input_dev *input_dev,
 
 	default:
 		/* no pad supported */
-		return 1;
+		return -ENODEV;
 	}
 	return 0;
 }
@@ -2367,11 +2712,13 @@ static const struct wacom_features wacom_features_0x317 =
 	  INTUOSPL, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES, .touch_max = 16,
 	  .check_for_hid_type = true, .hid_type = HID_TYPE_USBNONE };
 static const struct wacom_features wacom_features_0xF4 =
-	{ "Wacom Cintiq 24HD", 104280, 65400, 2047, 63,
-	  WACOM_24HD, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES, 200, 200 };
+	{ "Wacom Cintiq 24HD", 104080, 65200, 2047, 63,
+	  WACOM_24HD, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES,
+	  WACOM_CINTIQ_OFFSET, WACOM_CINTIQ_OFFSET };
 static const struct wacom_features wacom_features_0xF8 =
-	{ "Wacom Cintiq 24HD touch", 104280, 65400, 2047, 63, /* Pen */
-	  WACOM_24HD, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES, 200, 200,
+	{ "Wacom Cintiq 24HD touch", 104080, 65200, 2047, 63, /* Pen */
+	  WACOM_24HD, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES,
+	  WACOM_CINTIQ_OFFSET, WACOM_CINTIQ_OFFSET,
 	  .oVid = USB_VENDOR_ID_WACOM, .oPid = 0xf6 };
 static const struct wacom_features wacom_features_0xF6 =
 	{ "Wacom Cintiq 24HD touch", .type = WACOM_24HDT, /* Touch */
@@ -2387,8 +2734,9 @@ static const struct wacom_features wacom_features_0xC6 =
 	{ "Wacom Cintiq 12WX", 53020, 33440, 1023, 63,
 	  WACOM_BEE, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES };
 static const struct wacom_features wacom_features_0x304 =
-	{ "Wacom Cintiq 13HD", 59352, 33648, 1023, 63,
-	  WACOM_13HD, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES, 200, 200 };
+	{ "Wacom Cintiq 13HD", 59152, 33448, 1023, 63,
+	  WACOM_13HD, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES,
+	  WACOM_CINTIQ_OFFSET, WACOM_CINTIQ_OFFSET };
 static const struct wacom_features wacom_features_0xC7 =
 	{ "Wacom DTU1931", 37832, 30305, 511, 0,
 	  PL, WACOM_INTUOS_RES, WACOM_INTUOS_RES };
@@ -2400,28 +2748,38 @@ static const struct wacom_features wacom_features_0xF0 =
 	{ "Wacom DTU1631", 34623, 19553, 511, 0,
 	  DTU, WACOM_INTUOS_RES, WACOM_INTUOS_RES };
 static const struct wacom_features wacom_features_0xFB =
-	{ "Wacom DTU1031", 22096, 13960, 511, 0,
-	  DTUS, WACOM_INTUOS_RES, WACOM_INTUOS_RES };
+	{ "Wacom DTU1031", 21896, 13760, 511, 0,
+	  DTUS, WACOM_INTUOS_RES, WACOM_INTUOS_RES,
+	  WACOM_DTU_OFFSET, WACOM_DTU_OFFSET };
+static const struct wacom_features wacom_features_0x32F =
+	{ "Wacom DTU1031X", 22472, 12728, 511, 0,
+	  DTUSX, WACOM_INTUOS_RES, WACOM_INTUOS_RES,
+	  WACOM_DTU_OFFSET, WACOM_DTU_OFFSET };
 static const struct wacom_features wacom_features_0x57 =
 	{ "Wacom DTK2241", 95640, 54060, 2047, 63,
-	  DTK, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES, 200, 200 };
+	  DTK, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES,
+	  WACOM_CINTIQ_OFFSET, WACOM_CINTIQ_OFFSET };
 static const struct wacom_features wacom_features_0x59 = /* Pen */
 	{ "Wacom DTH2242", 95640, 54060, 2047, 63,
-	  DTK, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES, 200, 200,
+	  DTK, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES,
+	  WACOM_CINTIQ_OFFSET, WACOM_CINTIQ_OFFSET,
 	  .oVid = USB_VENDOR_ID_WACOM, .oPid = 0x5D };
 static const struct wacom_features wacom_features_0x5D = /* Touch */
 	{ "Wacom DTH2242",       .type = WACOM_24HDT,
 	  .oVid = USB_VENDOR_ID_WACOM, .oPid = 0x59, .touch_max = 10,
 	  .check_for_hid_type = true, .hid_type = HID_TYPE_USBNONE };
 static const struct wacom_features wacom_features_0xCC =
-	{ "Wacom Cintiq 21UX2", 87000, 65400, 2047, 63,
-	  WACOM_21UX2, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES, 200, 200 };
+	{ "Wacom Cintiq 21UX2", 86800, 65200, 2047, 63,
+	  WACOM_21UX2, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES,
+	  WACOM_CINTIQ_OFFSET, WACOM_CINTIQ_OFFSET };
 static const struct wacom_features wacom_features_0xFA =
-	{ "Wacom Cintiq 22HD", 95640, 54060, 2047, 63,
-	  WACOM_22HD, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES, 200, 200 };
+	{ "Wacom Cintiq 22HD", 95440, 53860, 2047, 63,
+	  WACOM_22HD, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES,
+	  WACOM_CINTIQ_OFFSET, WACOM_CINTIQ_OFFSET };
 static const struct wacom_features wacom_features_0x5B =
-	{ "Wacom Cintiq 22HDT", 95640, 54060, 2047, 63,
-	  WACOM_22HD, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES, 200, 200,
+	{ "Wacom Cintiq 22HDT", 95440, 53860, 2047, 63,
+	  WACOM_22HD, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES,
+	  WACOM_CINTIQ_OFFSET, WACOM_CINTIQ_OFFSET,
 	  .oVid = USB_VENDOR_ID_WACOM, .oPid = 0x5e };
 static const struct wacom_features wacom_features_0x5E =
 	{ "Wacom Cintiq 22HDT", .type = WACOM_24HDT,
@@ -2566,13 +2924,30 @@ static const struct wacom_features wacom_features_0x6004 =
 	{ "ISD-V4", 12800, 8000, 255, 0,
 	  TABLETPC, WACOM_INTUOS_RES, WACOM_INTUOS_RES };
 static const struct wacom_features wacom_features_0x307 =
-	{ "Wacom ISDv5 307", 59352, 33648, 2047, 63,
-	  CINTIQ_HYBRID, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES, 200, 200,
+	{ "Wacom ISDv5 307", 59152, 33448, 2047, 63,
+	  CINTIQ_HYBRID, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES,
+	  WACOM_CINTIQ_OFFSET, WACOM_CINTIQ_OFFSET,
 	  .oVid = USB_VENDOR_ID_WACOM, .oPid = 0x309 };
 static const struct wacom_features wacom_features_0x309 =
 	{ "Wacom ISDv5 309", .type = WACOM_24HDT, /* Touch */
 	  .oVid = USB_VENDOR_ID_WACOM, .oPid = 0x0307, .touch_max = 10,
 	  .check_for_hid_type = true, .hid_type = HID_TYPE_USBNONE };
+static const struct wacom_features wacom_features_0x30A =
+	{ "Wacom ISDv5 30A", 59152, 33448, 2047, 63,
+	  CINTIQ_HYBRID, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES,
+	  WACOM_CINTIQ_OFFSET, WACOM_CINTIQ_OFFSET,
+	  .oVid = USB_VENDOR_ID_WACOM, .oPid = 0x30C };
+static const struct wacom_features wacom_features_0x30C =
+	{ "Wacom ISDv5 30C", .type = WACOM_24HDT, /* Touch */
+	  .oVid = USB_VENDOR_ID_WACOM, .oPid = 0x30A, .touch_max = 10,
+	  .check_for_hid_type = true, .hid_type = HID_TYPE_USBNONE };
+static const struct wacom_features wacom_features_0x323 =
+	{ "Wacom Intuos P M", 21600, 13500, 1023, 31,
+	  INTUOSHT, WACOM_INTUOS_RES, WACOM_INTUOS_RES,
+	  .check_for_hid_type = true, .hid_type = HID_TYPE_USBNONE };
+
+static const struct wacom_features wacom_features_HID_ANY_ID =
+	{ "Wacom HID", .type = HID_GENERIC };
 
 #define USB_DEVICE_WACOM(prod)						\
 	HID_DEVICE(BUS_USB, HID_GROUP_WACOM, USB_VENDOR_ID_WACOM, prod),\
@@ -2708,14 +3083,21 @@ const struct hid_device_id wacom_ids[] = {
 	{ USB_DEVICE_WACOM(0x304) },
 	{ USB_DEVICE_WACOM(0x307) },
 	{ USB_DEVICE_WACOM(0x309) },
+	{ USB_DEVICE_WACOM(0x30A) },
+	{ USB_DEVICE_WACOM(0x30C) },
 	{ USB_DEVICE_WACOM(0x30E) },
 	{ USB_DEVICE_WACOM(0x314) },
 	{ USB_DEVICE_WACOM(0x315) },
 	{ USB_DEVICE_WACOM(0x317) },
+	{ USB_DEVICE_WACOM(0x323) },
+	{ USB_DEVICE_WACOM(0x32F) },
 	{ USB_DEVICE_WACOM(0x4001) },
 	{ USB_DEVICE_WACOM(0x4004) },
 	{ USB_DEVICE_WACOM(0x5000) },
 	{ USB_DEVICE_WACOM(0x5002) },
+	{ USB_DEVICE_LENOVO(0x6004) },
+
+	{ USB_DEVICE_WACOM(HID_ANY_ID) },
 	{ }
 };
 MODULE_DEVICE_TABLE(hid, wacom_ids);
diff --git a/drivers/hid/wacom_wac.h b/drivers/hid/wacom_wac.h
index 339ab5d81a2d..bfad815cda8a 100644
--- a/drivers/hid/wacom_wac.h
+++ b/drivers/hid/wacom_wac.h
@@ -80,6 +80,7 @@ enum {
 	PL,
 	DTU,
 	DTUS,
+	DTUSX,
 	INTUOS,
 	INTUOS3S,
 	INTUOS3,
@@ -113,6 +114,7 @@ enum {
 	MTSCREEN,
 	MTTPC,
 	MTTPC_B,
+	HID_GENERIC,
 	MAX_TYPE
 };
 
@@ -143,6 +145,7 @@ struct wacom_features {
 	int pktlen;
 	bool check_for_hid_type;
 	int hid_type;
+	int last_slot_field;
 };
 
 struct wacom_shared {
@@ -154,6 +157,20 @@ struct wacom_shared {
 	struct input_dev *touch_input;
 };
 
+struct hid_data {
+	__s16 inputmode;	/* InputMode HID feature, -1 if non-existent */
+	__s16 inputmode_index;	/* InputMode HID feature index in the report */
+	bool inrange_state;
+	bool invert_state;
+	bool tipswitch;
+	int x;
+	int y;
+	int pressure;
+	int width;
+	int height;
+	int id;
+};
+
 struct wacom_wac {
 	char name[WACOM_NAME_MAX];
 	char pad_name[WACOM_NAME_MAX];
@@ -167,6 +184,8 @@ struct wacom_wac {
 	struct wacom_shared *shared;
 	struct input_dev *input;
 	struct input_dev *pad_input;
+	bool input_registered;
+	bool pad_registered;
 	int pid;
 	int battery_capacity;
 	int num_contacts_left;
@@ -174,6 +193,7 @@ struct wacom_wac {
 	int ps_connected;
 	u8 bt_features;
 	u8 bt_high_speed;
+	struct hid_data hid_data;
 };
 
 #endif
diff --git a/drivers/hsi/clients/nokia-modem.c b/drivers/hsi/clients/nokia-modem.c
index 363b780dacea..f0c21458962c 100644
--- a/drivers/hsi/clients/nokia-modem.c
+++ b/drivers/hsi/clients/nokia-modem.c
@@ -29,7 +29,7 @@
 #include <linux/of_gpio.h>
 #include <linux/hsi/ssi_protocol.h>
 
-static unsigned int pm;
+static unsigned int pm = 1;
 module_param(pm, int, 0400);
 MODULE_PARM_DESC(pm,
 	"Enable power management (0=disabled, 1=userland based [default])");
@@ -164,9 +164,9 @@ static int nokia_modem_probe(struct device *dev)
 	dev_set_drvdata(dev, modem);
 
 	irq = irq_of_parse_and_map(np, 0);
-	if (irq < 0) {
+	if (!irq) {
 		dev_err(dev, "Invalid rst_ind interrupt (%d)\n", irq);
-		return irq;
+		return -EINVAL;
 	}
 	modem->nokia_modem_rst_ind_irq = irq;
 	pflags = irq_get_trigger_type(irq);
@@ -174,7 +174,7 @@ static int nokia_modem_probe(struct device *dev)
 	tasklet_init(&modem->nokia_modem_rst_ind_tasklet,
 			do_nokia_modem_rst_ind_tasklet, (unsigned long)modem);
 	err = devm_request_irq(dev, irq, nokia_modem_rst_ind_isr,
-				IRQF_DISABLED | pflags, "modem_rst_ind", modem);
+				pflags, "modem_rst_ind", modem);
 	if (err < 0) {
 		dev_err(dev, "Request rst_ind irq(%d) failed (flags %d)\n",
 								irq, pflags);
diff --git a/drivers/hsi/controllers/omap_ssi.c b/drivers/hsi/controllers/omap_ssi.c
index bf0eace4cb67..089c6c3feb3e 100644
--- a/drivers/hsi/controllers/omap_ssi.c
+++ b/drivers/hsi/controllers/omap_ssi.c
@@ -555,7 +555,7 @@ static int __exit ssi_remove(struct platform_device *pd)
 	return 0;
 }
 
-#ifdef CONFIG_PM_RUNTIME
+#ifdef CONFIG_PM
 static int omap_ssi_runtime_suspend(struct device *dev)
 {
 	struct hsi_controller *ssi = dev_get_drvdata(dev);
@@ -610,7 +610,6 @@ static struct platform_driver ssi_pdriver = {
 	.remove	= __exit_p(ssi_remove),
 	.driver	= {
 		.name	= "omap_ssi",
-		.owner	= THIS_MODULE,
 		.pm     = DEV_PM_OPS,
 		.of_match_table = omap_ssi_of_match,
 	},
diff --git a/drivers/hsi/controllers/omap_ssi_port.c b/drivers/hsi/controllers/omap_ssi_port.c
index 4c0b5820581e..1f8652b3de06 100644
--- a/drivers/hsi/controllers/omap_ssi_port.c
+++ b/drivers/hsi/controllers/omap_ssi_port.c
@@ -1118,8 +1118,7 @@ static int __init ssi_port_probe(struct platform_device *pd)
 	dev_dbg(&pd->dev, "init ssi port...\n");
 
 	if (!try_module_get(ssi->owner)) {
-		dev_err(&pd->dev, "could not increment parent module refcount (err=%d)\n",
-			err);
+		dev_err(&pd->dev, "could not increment parent module refcount\n");
 		return -ENODEV;
 	}
 
@@ -1260,7 +1259,7 @@ static int __exit ssi_port_remove(struct platform_device *pd)
 	return 0;
 }
 
-#ifdef CONFIG_PM_RUNTIME
+#ifdef CONFIG_PM
 static int ssi_save_port_ctx(struct omap_ssi_port *omap_port)
 {
 	struct hsi_port *port = to_hsi_port(omap_port->dev);
@@ -1385,7 +1384,6 @@ static struct platform_driver ssi_port_pdriver = {
 	.remove	= __exit_p(ssi_port_remove),
 	.driver	= {
 		.name	= "omap_ssi_port",
-		.owner	= THIS_MODULE,
 		.of_match_table = omap_ssi_port_of_match,
 		.pm	= DEV_PM_OPS,
 	},
diff --git a/drivers/hv/channel.c b/drivers/hv/channel.c
index 531a593912ec..433f72a1c006 100644
--- a/drivers/hv/channel.c
+++ b/drivers/hv/channel.c
@@ -165,8 +165,10 @@ int vmbus_open(struct vmbus_channel *newchannel, u32 send_ringbuffer_size,
 	ret = vmbus_post_msg(open_msg,
 			       sizeof(struct vmbus_channel_open_channel));
 
-	if (ret != 0)
+	if (ret != 0) {
+		err = ret;
 		goto error1;
+	}
 
 	t = wait_for_completion_timeout(&open_info->waitevent, 5*HZ);
 	if (t == 0) {
@@ -363,7 +365,6 @@ int vmbus_establish_gpadl(struct vmbus_channel *channel, void *kbuffer,
 	u32 next_gpadl_handle;
 	unsigned long flags;
 	int ret = 0;
-	int t;
 
 	next_gpadl_handle = atomic_read(&vmbus_connection.next_gpadl_handle);
 	atomic_inc(&vmbus_connection.next_gpadl_handle);
@@ -410,9 +411,7 @@ int vmbus_establish_gpadl(struct vmbus_channel *channel, void *kbuffer,
 
 		}
 	}
-	t = wait_for_completion_timeout(&msginfo->waitevent, 5*HZ);
-	BUG_ON(t == 0);
-
+	wait_for_completion(&msginfo->waitevent);
 
 	/* At this point, we received the gpadl created msg */
 	*gpadl_handle = gpadlmsg->gpadl;
@@ -435,7 +434,7 @@ int vmbus_teardown_gpadl(struct vmbus_channel *channel, u32 gpadl_handle)
 	struct vmbus_channel_gpadl_teardown *msg;
 	struct vmbus_channel_msginfo *info;
 	unsigned long flags;
-	int ret, t;
+	int ret;
 
 	info = kmalloc(sizeof(*info) +
 		       sizeof(struct vmbus_channel_gpadl_teardown), GFP_KERNEL);
@@ -457,11 +456,12 @@ int vmbus_teardown_gpadl(struct vmbus_channel *channel, u32 gpadl_handle)
 	ret = vmbus_post_msg(msg,
 			       sizeof(struct vmbus_channel_gpadl_teardown));
 
-	BUG_ON(ret != 0);
-	t = wait_for_completion_timeout(&info->waitevent, 5*HZ);
-	BUG_ON(t == 0);
+	if (ret)
+		goto post_msg_err;
+
+	wait_for_completion(&info->waitevent);
 
-	/* Received a torndown response */
+post_msg_err:
 	spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags);
 	list_del(&info->msglistentry);
 	spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags);
@@ -478,7 +478,7 @@ static void reset_channel_cb(void *arg)
 	channel->onchannel_callback = NULL;
 }
 
-static void vmbus_close_internal(struct vmbus_channel *channel)
+static int vmbus_close_internal(struct vmbus_channel *channel)
 {
 	struct vmbus_channel_close_channel *msg;
 	int ret;
@@ -486,11 +486,14 @@ static void vmbus_close_internal(struct vmbus_channel *channel)
 	channel->state = CHANNEL_OPEN_STATE;
 	channel->sc_creation_callback = NULL;
 	/* Stop callback and cancel the timer asap */
-	if (channel->target_cpu != smp_processor_id())
+	if (channel->target_cpu != get_cpu()) {
+		put_cpu();
 		smp_call_function_single(channel->target_cpu, reset_channel_cb,
 					 channel, true);
-	else
+	} else {
 		reset_channel_cb(channel);
+		put_cpu();
+	}
 
 	/* Send a closing message */
 
@@ -501,11 +504,28 @@ static void vmbus_close_internal(struct vmbus_channel *channel)
 
 	ret = vmbus_post_msg(msg, sizeof(struct vmbus_channel_close_channel));
 
-	BUG_ON(ret != 0);
+	if (ret) {
+		pr_err("Close failed: close post msg return is %d\n", ret);
+		/*
+		 * If we failed to post the close msg,
+		 * it is perhaps better to leak memory.
+		 */
+		return ret;
+	}
+
 	/* Tear down the gpadl for the channel's ring buffer */
-	if (channel->ringbuffer_gpadlhandle)
-		vmbus_teardown_gpadl(channel,
-					  channel->ringbuffer_gpadlhandle);
+	if (channel->ringbuffer_gpadlhandle) {
+		ret = vmbus_teardown_gpadl(channel,
+					   channel->ringbuffer_gpadlhandle);
+		if (ret) {
+			pr_err("Close failed: teardown gpadl return %d\n", ret);
+			/*
+			 * If we failed to teardown gpadl,
+			 * it is perhaps better to leak memory.
+			 */
+			return ret;
+		}
+	}
 
 	/* Cleanup the ring buffers for this channel */
 	hv_ringbuffer_cleanup(&channel->outbound);
@@ -514,7 +534,7 @@ static void vmbus_close_internal(struct vmbus_channel *channel)
 	free_pages((unsigned long)channel->ringbuffer_pages,
 		get_order(channel->ringbuffer_pagecount * PAGE_SIZE));
 
-
+	return ret;
 }
 
 /*
diff --git a/drivers/hv/channel_mgmt.c b/drivers/hv/channel_mgmt.c
index ed9350d42764..2c59f030546b 100644
--- a/drivers/hv/channel_mgmt.c
+++ b/drivers/hv/channel_mgmt.c
@@ -216,19 +216,29 @@ static void vmbus_process_rescind_offer(struct work_struct *work)
 	unsigned long flags;
 	struct vmbus_channel *primary_channel;
 	struct vmbus_channel_relid_released msg;
+	struct device *dev;
+
+	if (channel->device_obj) {
+		dev = get_device(&channel->device_obj->device);
+		if (dev) {
+			vmbus_device_unregister(channel->device_obj);
+			put_device(dev);
+		}
+	}
 
-	if (channel->device_obj)
-		vmbus_device_unregister(channel->device_obj);
 	memset(&msg, 0, sizeof(struct vmbus_channel_relid_released));
 	msg.child_relid = channel->offermsg.child_relid;
 	msg.header.msgtype = CHANNELMSG_RELID_RELEASED;
 	vmbus_post_msg(&msg, sizeof(struct vmbus_channel_relid_released));
 
-	if (channel->target_cpu != smp_processor_id())
+	if (channel->target_cpu != get_cpu()) {
+		put_cpu();
 		smp_call_function_single(channel->target_cpu,
 					 percpu_channel_deq, channel, true);
-	else
+	} else {
 		percpu_channel_deq(channel);
+		put_cpu();
+	}
 
 	if (channel->primary_channel == NULL) {
 		spin_lock_irqsave(&vmbus_connection.channel_lock, flags);
@@ -294,12 +304,15 @@ static void vmbus_process_offer(struct work_struct *work)
 	spin_unlock_irqrestore(&vmbus_connection.channel_lock, flags);
 
 	if (enq) {
-		if (newchannel->target_cpu != smp_processor_id())
+		if (newchannel->target_cpu != get_cpu()) {
+			put_cpu();
 			smp_call_function_single(newchannel->target_cpu,
 						 percpu_channel_enq,
 						 newchannel, true);
-		else
+		} else {
 			percpu_channel_enq(newchannel);
+			put_cpu();
+		}
 	}
 	if (!fnew) {
 		/*
@@ -314,12 +327,15 @@ static void vmbus_process_offer(struct work_struct *work)
 			list_add_tail(&newchannel->sc_list, &channel->sc_list);
 			spin_unlock_irqrestore(&channel->sc_lock, flags);
 
-			if (newchannel->target_cpu != smp_processor_id())
+			if (newchannel->target_cpu != get_cpu()) {
+				put_cpu();
 				smp_call_function_single(newchannel->target_cpu,
 							 percpu_channel_enq,
 							 newchannel, true);
-			else
+			} else {
 				percpu_channel_enq(newchannel);
+				put_cpu();
+			}
 
 			newchannel->state = CHANNEL_OPEN_STATE;
 			if (channel->sc_creation_callback != NULL)
@@ -508,6 +524,8 @@ static void vmbus_onoffer_rescind(struct vmbus_channel_message_header *hdr)
 		/* Just return here, no channel found */
 		return;
 
+	channel->rescind = true;
+
 	/* work is initialized for vmbus_process_rescind_offer() from
 	 * vmbus_process_offer() where the channel got created */
 	queue_work(channel->controlwq, &channel->work);
diff --git a/drivers/hv/connection.c b/drivers/hv/connection.c
index ae22e3c1fc4c..e206619b946e 100644
--- a/drivers/hv/connection.c
+++ b/drivers/hv/connection.c
@@ -427,10 +427,21 @@ int vmbus_post_msg(void *buffer, size_t buflen)
 	 * insufficient resources. Retry the operation a couple of
 	 * times before giving up.
 	 */
-	while (retries < 3) {
-		ret =  hv_post_message(conn_id, 1, buffer, buflen);
-		if (ret != HV_STATUS_INSUFFICIENT_BUFFERS)
+	while (retries < 10) {
+		ret = hv_post_message(conn_id, 1, buffer, buflen);
+
+		switch (ret) {
+		case HV_STATUS_INSUFFICIENT_BUFFERS:
+			ret = -ENOMEM;
+		case -ENOMEM:
+			break;
+		case HV_STATUS_SUCCESS:
 			return ret;
+		default:
+			pr_err("hv_post_msg() failed; error code:%d\n", ret);
+			return -EINVAL;
+		}
+
 		retries++;
 		msleep(100);
 	}
diff --git a/drivers/hv/hv.c b/drivers/hv/hv.c
index edfc8488cb03..3e4235c7a47f 100644
--- a/drivers/hv/hv.c
+++ b/drivers/hv/hv.c
@@ -138,6 +138,8 @@ int hv_init(void)
 	memset(hv_context.synic_event_page, 0, sizeof(void *) * NR_CPUS);
 	memset(hv_context.synic_message_page, 0,
 	       sizeof(void *) * NR_CPUS);
+	memset(hv_context.post_msg_page, 0,
+	       sizeof(void *) * NR_CPUS);
 	memset(hv_context.vp_index, 0,
 	       sizeof(int) * NR_CPUS);
 	memset(hv_context.event_dpc, 0,
@@ -217,26 +219,18 @@ int hv_post_message(union hv_connection_id connection_id,
 		  enum hv_message_type message_type,
 		  void *payload, size_t payload_size)
 {
-	struct aligned_input {
-		u64 alignment8;
-		struct hv_input_post_message msg;
-	};
 
 	struct hv_input_post_message *aligned_msg;
 	u16 status;
-	unsigned long addr;
 
 	if (payload_size > HV_MESSAGE_PAYLOAD_BYTE_COUNT)
 		return -EMSGSIZE;
 
-	addr = (unsigned long)kmalloc(sizeof(struct aligned_input), GFP_ATOMIC);
-	if (!addr)
-		return -ENOMEM;
-
 	aligned_msg = (struct hv_input_post_message *)
-			(ALIGN(addr, HV_HYPERCALL_PARAM_ALIGN));
+			hv_context.post_msg_page[get_cpu()];
 
 	aligned_msg->connectionid = connection_id;
+	aligned_msg->reserved = 0;
 	aligned_msg->message_type = message_type;
 	aligned_msg->payload_size = payload_size;
 	memcpy((void *)aligned_msg->payload, payload, payload_size);
@@ -244,8 +238,7 @@ int hv_post_message(union hv_connection_id connection_id,
 	status = do_hypercall(HVCALL_POST_MESSAGE, aligned_msg, NULL)
 		& 0xFFFF;
 
-	kfree((void *)addr);
-
+	put_cpu();
 	return status;
 }
 
@@ -294,6 +287,14 @@ int hv_synic_alloc(void)
 			pr_err("Unable to allocate SYNIC event page\n");
 			goto err;
 		}
+
+		hv_context.post_msg_page[cpu] =
+			(void *)get_zeroed_page(GFP_ATOMIC);
+
+		if (hv_context.post_msg_page[cpu] == NULL) {
+			pr_err("Unable to allocate post msg page\n");
+			goto err;
+		}
 	}
 
 	return 0;
@@ -308,6 +309,8 @@ static void hv_synic_free_cpu(int cpu)
 		free_page((unsigned long)hv_context.synic_event_page[cpu]);
 	if (hv_context.synic_message_page[cpu])
 		free_page((unsigned long)hv_context.synic_message_page[cpu]);
+	if (hv_context.post_msg_page[cpu])
+		free_page((unsigned long)hv_context.post_msg_page[cpu]);
 }
 
 void hv_synic_free(void)
diff --git a/drivers/hv/hv_balloon.c b/drivers/hv/hv_balloon.c
index 5e90c5d771a7..b958ded8ac7e 100644
--- a/drivers/hv/hv_balloon.c
+++ b/drivers/hv/hv_balloon.c
@@ -1087,10 +1087,12 @@ static void balloon_up(struct work_struct *dummy)
 	struct dm_balloon_response *bl_resp;
 	int alloc_unit;
 	int ret;
-	bool alloc_error = false;
+	bool alloc_error;
 	bool done = false;
 	int i;
 
+	/* The host balloons pages in 2M granularity. */
+	WARN_ON_ONCE(num_pages % PAGES_IN_2M != 0);
 
 	/*
 	 * We will attempt 2M allocations. However, if we fail to
@@ -1107,16 +1109,18 @@ static void balloon_up(struct work_struct *dummy)
 
 
 		num_pages -= num_ballooned;
+		alloc_error = false;
 		num_ballooned = alloc_balloon_pages(&dm_device, num_pages,
 						bl_resp, alloc_unit,
 						 &alloc_error);
 
-		if ((alloc_error) && (alloc_unit != 1)) {
+		if (alloc_unit != 1 && num_ballooned == 0) {
 			alloc_unit = 1;
 			continue;
 		}
 
-		if ((alloc_error) || (num_ballooned == num_pages)) {
+		if ((alloc_unit == 1 && alloc_error) ||
+			(num_ballooned == num_pages)) {
 			bl_resp->more_pages = 0;
 			done = true;
 			dm_device.state = DM_INITIALIZED;
diff --git a/drivers/hv/hv_kvp.c b/drivers/hv/hv_kvp.c
index 521c14625b3a..beb8105c0e7b 100644
--- a/drivers/hv/hv_kvp.c
+++ b/drivers/hv/hv_kvp.c
@@ -350,6 +350,7 @@ kvp_send_key(struct work_struct *dummy)
 	__u8 pool = kvp_transaction.kvp_msg->kvp_hdr.pool;
 	__u32 val32;
 	__u64 val64;
+	int rc;
 
 	msg = kzalloc(sizeof(*msg) + sizeof(struct hv_kvp_msg) , GFP_ATOMIC);
 	if (!msg)
@@ -446,7 +447,13 @@ kvp_send_key(struct work_struct *dummy)
 	}
 
 	msg->len = sizeof(struct hv_kvp_msg);
-	cn_netlink_send(msg, 0, 0, GFP_ATOMIC);
+	rc = cn_netlink_send(msg, 0, 0, GFP_ATOMIC);
+	if (rc) {
+		pr_debug("KVP: failed to communicate to the daemon: %d\n", rc);
+		if (cancel_delayed_work_sync(&kvp_work))
+			kvp_respond_to_host(message, HV_E_FAIL);
+	}
+
 	kfree(msg);
 
 	return;
diff --git a/drivers/hv/hv_snapshot.c b/drivers/hv/hv_snapshot.c
index 34f14fddb666..9d5e0d1efdb5 100644
--- a/drivers/hv/hv_snapshot.c
+++ b/drivers/hv/hv_snapshot.c
@@ -28,7 +28,7 @@
 #define VSS_MINOR  0
 #define VSS_VERSION    (VSS_MAJOR << 16 | VSS_MINOR)
 
-
+#define VSS_USERSPACE_TIMEOUT (msecs_to_jiffies(10 * 1000))
 
 /*
  * Global state maintained for transaction that is being processed.
@@ -55,12 +55,24 @@ static const char vss_name[] = "vss_kernel_module";
 static __u8 *recv_buffer;
 
 static void vss_send_op(struct work_struct *dummy);
+static void vss_timeout_func(struct work_struct *dummy);
+
+static DECLARE_DELAYED_WORK(vss_timeout_work, vss_timeout_func);
 static DECLARE_WORK(vss_send_op_work, vss_send_op);
 
 /*
  * Callback when data is received from user mode.
  */
 
+static void vss_timeout_func(struct work_struct *dummy)
+{
+	/*
+	 * Timeout waiting for userspace component to reply happened.
+	 */
+	pr_warn("VSS: timeout waiting for daemon to reply\n");
+	vss_respond_to_host(HV_E_FAIL);
+}
+
 static void
 vss_cn_callback(struct cn_msg *msg, struct netlink_skb_parms *nsp)
 {
@@ -76,13 +88,15 @@ vss_cn_callback(struct cn_msg *msg, struct netlink_skb_parms *nsp)
 		return;
 
 	}
-	vss_respond_to_host(vss_msg->error);
+	if (cancel_delayed_work_sync(&vss_timeout_work))
+		vss_respond_to_host(vss_msg->error);
 }
 
 
 static void vss_send_op(struct work_struct *dummy)
 {
 	int op = vss_transaction.msg->vss_hdr.operation;
+	int rc;
 	struct cn_msg *msg;
 	struct hv_vss_msg *vss_msg;
 
@@ -98,7 +112,12 @@ static void vss_send_op(struct work_struct *dummy)
 	vss_msg->vss_hdr.operation = op;
 	msg->len = sizeof(struct hv_vss_msg);
 
-	cn_netlink_send(msg, 0, 0, GFP_ATOMIC);
+	rc = cn_netlink_send(msg, 0, 0, GFP_ATOMIC);
+	if (rc) {
+		pr_warn("VSS: failed to communicate to the daemon: %d\n", rc);
+		if (cancel_delayed_work_sync(&vss_timeout_work))
+			vss_respond_to_host(HV_E_FAIL);
+	}
 	kfree(msg);
 
 	return;
@@ -223,6 +242,8 @@ void hv_vss_onchannelcallback(void *context)
 			case VSS_OP_FREEZE:
 			case VSS_OP_THAW:
 				schedule_work(&vss_send_op_work);
+				schedule_delayed_work(&vss_timeout_work,
+						      VSS_USERSPACE_TIMEOUT);
 				return;
 
 			case VSS_OP_HOT_BACKUP:
@@ -277,5 +298,6 @@ hv_vss_init(struct hv_util_service *srv)
 void hv_vss_deinit(void)
 {
 	cn_del_callback(&vss_id);
+	cancel_delayed_work_sync(&vss_timeout_work);
 	cancel_work_sync(&vss_send_op_work);
 }
diff --git a/drivers/hv/hyperv_vmbus.h b/drivers/hv/hyperv_vmbus.h
index 22b750749a39..c386d8dc7223 100644
--- a/drivers/hv/hyperv_vmbus.h
+++ b/drivers/hv/hyperv_vmbus.h
@@ -515,6 +515,10 @@ struct hv_context {
 	 * per-cpu list of the channels based on their CPU affinity.
 	 */
 	struct list_head percpu_list[NR_CPUS];
+	/*
+	 * buffer to post messages to the host.
+	 */
+	void *post_msg_page[NR_CPUS];
 };
 
 extern struct hv_context hv_context;
diff --git a/drivers/hv/ring_buffer.c b/drivers/hv/ring_buffer.c
index 15db66b74141..6361d124f67d 100644
--- a/drivers/hv/ring_buffer.c
+++ b/drivers/hv/ring_buffer.c
@@ -361,6 +361,11 @@ int hv_ringbuffer_init(struct hv_ring_buffer_info *ring_info,
 	ring_info->ring_buffer->read_index =
 		ring_info->ring_buffer->write_index = 0;
 
+	/*
+	 * Set the feature bit for enabling flow control.
+	 */
+	ring_info->ring_buffer->feature_bits.value = 1;
+
 	ring_info->ring_size = buflen;
 	ring_info->ring_datasize = buflen - sizeof(struct hv_ring_buffer);
 
diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig
index f00d048aa583..a7de26d1ac80 100644
--- a/drivers/hwmon/Kconfig
+++ b/drivers/hwmon/Kconfig
@@ -280,8 +280,8 @@ config SENSORS_K10TEMP
 	  If you say yes here you get support for the temperature
 	  sensor(s) inside your CPU. Supported are later revisions of
 	  the AMD Family 10h and all revisions of the AMD Family 11h,
-	  12h (Llano), 14h (Brazos), 15h (Bulldozer/Trinity/Kaveri) and
-	  16h (Kabini/Mullins) microarchitectures.
+	  12h (Llano), 14h (Brazos), 15h (Bulldozer/Trinity/Kaveri/Carrizo)
+	  and 16h (Kabini/Mullins) microarchitectures.
 
 	  This driver can also be built as a module.  If so, the module
 	  will be called k10temp.
@@ -574,6 +574,16 @@ config SENSORS_IIO_HWMON
 	  for those channels specified in the map.  This map can be provided
 	  either via platform data or the device tree bindings.
 
+config SENSORS_I5500
+	tristate "Intel 5500/5520/X58 temperature sensor"
+	depends on X86 && PCI
+	help
+	  If you say yes here you get support for the temperature
+	  sensor inside the Intel 5500, 5520 and X58 chipsets.
+
+	  This driver can also be built as a module. If so, the module
+	  will be called i5500_temp.
+
 config SENSORS_CORETEMP
 	tristate "Intel Core/Core2/Atom temperature sensor"
 	depends on X86
@@ -839,6 +849,16 @@ config SENSORS_MCP3021
 	  This driver can also be built as a module.  If so, the module
 	  will be called mcp3021.
 
+config SENSORS_MENF21BMC_HWMON
+	tristate "MEN 14F021P00 BMC Hardware Monitoring"
+	depends on MFD_MENF21BMC
+	help
+	  Say Y here to include support for the MEN 14F021P00 BMC
+	  hardware monitoring.
+
+	  This driver can also be built as a module. If so the module
+	  will be called menf21bmc_hwmon.
+
 config SENSORS_ADCXX
 	tristate "National Semiconductor ADCxxxSxxx"
 	depends on SPI_MASTER
@@ -1018,11 +1038,11 @@ config SENSORS_LM93
 	  will be called lm93.
 
 config SENSORS_LM95234
-	tristate "National Semiconductor LM95234"
+	tristate "National Semiconductor LM95234 and compatibles"
 	depends on I2C
 	help
-	  If you say yes here you get support for the LM95234 temperature
-	  sensor.
+	  If you say yes here you get support for the LM95233 and LM95234
+	  temperature sensor chips.
 
 	  This driver can also be built as a module.  If so, the module
 	  will be called lm95234.
@@ -1038,10 +1058,11 @@ config SENSORS_LM95241
 	  will be called lm95241.
 
 config SENSORS_LM95245
-	tristate "National Semiconductor LM95245 sensor chip"
+	tristate "National Semiconductor LM95245 and compatibles"
 	depends on I2C
 	help
-	  If you say yes here you get support for LM95245 sensor chip.
+	  If you say yes here you get support for LM95235 and LM95245
+	  temperature sensor chips.
 
 	  This driver can also be built as a module.  If so, the module
 	  will be called lm95245.
@@ -1077,6 +1098,7 @@ config SENSORS_PC87427
 config SENSORS_NTC_THERMISTOR
 	tristate "NTC thermistor support from Murata"
 	depends on !OF || IIO=n || IIO
+	depends on THERMAL || !THERMAL_OF
 	help
 	  This driver supports NTC thermistors sensor reading and its
 	  interpretation. The driver can also monitor the temperature and
@@ -1106,12 +1128,23 @@ config SENSORS_NCT6775
 	help
 	  If you say yes here you get support for the hardware monitoring
 	  functionality of the Nuvoton NCT6106D, NCT6775F, NCT6776F, NCT6779D,
-	  NCT6791D and compatible Super-I/O chips. This driver replaces the
-	  w83627ehf driver for NCT6775F and NCT6776F.
+	  NCT6791D, NCT6792D and compatible Super-I/O chips. This driver
+	  replaces the w83627ehf driver for NCT6775F and NCT6776F.
 
 	  This driver can also be built as a module.  If so, the module
 	  will be called nct6775.
 
+config SENSORS_NCT7802
+	tristate "Nuvoton NCT7802Y"
+	depends on I2C
+	select REGMAP_I2C
+	help
+	  If you say yes here you get support for the Nuvoton NCT7802Y
+	  hardware monitoring chip.
+
+	  This driver can also be built as a module.  If so, the module
+	  will be called nct7802.
+
 config SENSORS_PCF8591
 	tristate "Philips PCF8591 ADC/DAC"
 	depends on I2C
@@ -1443,7 +1476,7 @@ config SENSORS_TMP401
 	depends on I2C
 	help
 	  If you say yes here you get support for Texas Instruments TMP401,
-	  TMP411, TMP431, and TMP432 temperature sensor chips.
+	  TMP411, TMP431, TMP432 and TMP435 temperature sensor chips.
 
 	  This driver can also be built as a module.  If so, the module
 	  will be called tmp401.
diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile
index be28152c9848..6c941472e707 100644
--- a/drivers/hwmon/Makefile
+++ b/drivers/hwmon/Makefile
@@ -68,6 +68,7 @@ obj-$(CONFIG_SENSORS_GPIO_FAN)	+= gpio-fan.o
 obj-$(CONFIG_SENSORS_HIH6130)	+= hih6130.o
 obj-$(CONFIG_SENSORS_HTU21)	+= htu21.o
 obj-$(CONFIG_SENSORS_ULTRA45)	+= ultra45_env.o
+obj-$(CONFIG_SENSORS_I5500)	+= i5500_temp.o
 obj-$(CONFIG_SENSORS_I5K_AMB)	+= i5k_amb.o
 obj-$(CONFIG_SENSORS_IBMAEM)	+= ibmaem.o
 obj-$(CONFIG_SENSORS_IBMPEX)	+= ibmpex.o
@@ -115,8 +116,10 @@ obj-$(CONFIG_SENSORS_MAX6650)	+= max6650.o
 obj-$(CONFIG_SENSORS_MAX6697)	+= max6697.o
 obj-$(CONFIG_SENSORS_MC13783_ADC)+= mc13783-adc.o
 obj-$(CONFIG_SENSORS_MCP3021)	+= mcp3021.o
+obj-$(CONFIG_SENSORS_MENF21BMC_HWMON) += menf21bmc_hwmon.o
 obj-$(CONFIG_SENSORS_NCT6683)	+= nct6683.o
 obj-$(CONFIG_SENSORS_NCT6775)	+= nct6775.o
+obj-$(CONFIG_SENSORS_NCT7802)	+= nct7802.o
 obj-$(CONFIG_SENSORS_NTC_THERMISTOR)	+= ntc_thermistor.o
 obj-$(CONFIG_SENSORS_PC87360)	+= pc87360.o
 obj-$(CONFIG_SENSORS_PC87427)	+= pc87427.o
diff --git a/drivers/hwmon/ab8500.c b/drivers/hwmon/ab8500.c
index d844dc806853..8b6a4f4c8774 100644
--- a/drivers/hwmon/ab8500.c
+++ b/drivers/hwmon/ab8500.c
@@ -6,7 +6,7 @@
  *
  * When the AB8500 thermal warning temperature is reached (threshold cannot
  * be changed by SW), an interrupt is set, and if no further action is taken
- * within a certain time frame, pm_power off will be called.
+ * within a certain time frame, kernel_power_off will be called.
  *
  * When AB8500 thermal shutdown temperature is reached a hardware shutdown of
  * the AB8500 will occur.
@@ -21,6 +21,7 @@
 #include <linux/module.h>
 #include <linux/platform_device.h>
 #include <linux/power/ab8500.h>
+#include <linux/reboot.h>
 #include <linux/slab.h>
 #include <linux/sysfs.h>
 #include "abx500.h"
@@ -106,7 +107,7 @@ static void ab8500_thermal_power_off(struct work_struct *work)
 
 	dev_warn(&abx500_data->pdev->dev, "Power off due to critical temp\n");
 
-	pm_power_off();
+	kernel_power_off();
 }
 
 static ssize_t ab8500_show_name(struct device *dev,
diff --git a/drivers/hwmon/abituguru.c b/drivers/hwmon/abituguru.c
index 9c8a6bab8228..7a09c1615aa9 100644
--- a/drivers/hwmon/abituguru.c
+++ b/drivers/hwmon/abituguru.c
@@ -1547,7 +1547,6 @@ static SIMPLE_DEV_PM_OPS(abituguru_pm, abituguru_suspend, abituguru_resume);
 
 static struct platform_driver abituguru_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= ABIT_UGURU_NAME,
 		.pm	= ABIT_UGURU_PM,
 	},
diff --git a/drivers/hwmon/abituguru3.c b/drivers/hwmon/abituguru3.c
index 4ae74aa8cdc1..3d2a4ae92d1e 100644
--- a/drivers/hwmon/abituguru3.c
+++ b/drivers/hwmon/abituguru3.c
@@ -1167,7 +1167,6 @@ static SIMPLE_DEV_PM_OPS(abituguru3_pm, abituguru3_suspend, abituguru3_resume);
 
 static struct platform_driver abituguru3_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= ABIT_UGURU3_NAME,
 		.pm	= ABIT_UGURU3_PM
 	},
diff --git a/drivers/hwmon/abx500.c b/drivers/hwmon/abx500.c
index 769fe20ec938..13875968c844 100644
--- a/drivers/hwmon/abx500.c
+++ b/drivers/hwmon/abx500.c
@@ -474,7 +474,6 @@ static const struct of_device_id abx500_temp_match[] = {
 
 static struct platform_driver abx500_temp_driver = {
 	.driver = {
-		.owner = THIS_MODULE,
 		.name = "abx500-temp",
 		.of_match_table = of_match_ptr(abx500_temp_match),
 	},
diff --git a/drivers/hwmon/ads1015.c b/drivers/hwmon/ads1015.c
index 126516414c11..f155b8380481 100644
--- a/drivers/hwmon/ads1015.c
+++ b/drivers/hwmon/ads1015.c
@@ -184,20 +184,18 @@ static int ads1015_get_channels_config_of(struct i2c_client *client)
 		return -EINVAL;
 
 	for_each_child_of_node(client->dev.of_node, node) {
-		const __be32 *property;
-		int len;
+		u32 pval;
 		unsigned int channel;
 		unsigned int pga = ADS1015_DEFAULT_PGA;
 		unsigned int data_rate = ADS1015_DEFAULT_DATA_RATE;
 
-		property = of_get_property(node, "reg", &len);
-		if (!property || len != sizeof(int)) {
+		if (of_property_read_u32(node, "reg", &pval)) {
 			dev_err(&client->dev, "invalid reg on %s\n",
 				node->full_name);
 			continue;
 		}
 
-		channel = be32_to_cpup(property);
+		channel = pval;
 		if (channel >= ADS1015_CHANNELS) {
 			dev_err(&client->dev,
 				"invalid channel index %d on %s\n",
@@ -205,20 +203,17 @@ static int ads1015_get_channels_config_of(struct i2c_client *client)
 			continue;
 		}
 
-		property = of_get_property(node, "ti,gain", &len);
-		if (property && len == sizeof(int)) {
-			pga = be32_to_cpup(property);
+		if (!of_property_read_u32(node, "ti,gain", &pval)) {
+			pga = pval;
 			if (pga > 6) {
-				dev_err(&client->dev,
-					"invalid gain on %s\n",
+				dev_err(&client->dev, "invalid gain on %s\n",
 					node->full_name);
 				return -EINVAL;
 			}
 		}
 
-		property = of_get_property(node, "ti,datarate", &len);
-		if (property && len == sizeof(int)) {
-			data_rate = be32_to_cpup(property);
+		if (!of_property_read_u32(node, "ti,datarate", &pval)) {
+			data_rate = pval;
 			if (data_rate > 7) {
 				dev_err(&client->dev,
 					"invalid data_rate on %s\n",
diff --git a/drivers/hwmon/applesmc.c b/drivers/hwmon/applesmc.c
index 3288f13d2d87..0af63da6b603 100644
--- a/drivers/hwmon/applesmc.c
+++ b/drivers/hwmon/applesmc.c
@@ -676,7 +676,6 @@ static struct platform_driver applesmc_driver = {
 	.probe = applesmc_probe,
 	.driver	= {
 		.name = "applesmc",
-		.owner = THIS_MODULE,
 		.pm = &applesmc_pm_ops,
 	},
 };
diff --git a/drivers/hwmon/coretemp.c b/drivers/hwmon/coretemp.c
index d76f0b70c6e0..5b7fec824f10 100644
--- a/drivers/hwmon/coretemp.c
+++ b/drivers/hwmon/coretemp.c
@@ -596,7 +596,6 @@ static int coretemp_remove(struct platform_device *pdev)
 
 static struct platform_driver coretemp_driver = {
 	.driver = {
-		.owner = THIS_MODULE,
 		.name = DRVNAME,
 	},
 	.probe = coretemp_probe,
diff --git a/drivers/hwmon/da9052-hwmon.c b/drivers/hwmon/da9052-hwmon.c
index d14ab3c45daa..c9832bfacfe5 100644
--- a/drivers/hwmon/da9052-hwmon.c
+++ b/drivers/hwmon/da9052-hwmon.c
@@ -26,7 +26,6 @@
 
 struct da9052_hwmon {
 	struct da9052	*da9052;
-	struct device	*class_device;
 	struct mutex	hwmon_lock;
 };
 
@@ -190,13 +189,6 @@ static ssize_t da9052_read_vbbat(struct device *dev,
 	return sprintf(buf, "%d\n", vbbat_reg_to_mv(ret));
 }
 
-static ssize_t da9052_hwmon_show_name(struct device *dev,
-				      struct device_attribute *devattr,
-				      char *buf)
-{
-	return sprintf(buf, "da9052\n");
-}
-
 static ssize_t show_label(struct device *dev,
 			  struct device_attribute *devattr, char *buf)
 {
@@ -243,10 +235,7 @@ static SENSOR_DEVICE_ATTR(temp8_input, S_IRUGO, da9052_read_tjunc, NULL,
 static SENSOR_DEVICE_ATTR(temp8_label, S_IRUGO, show_label, NULL,
 			  DA9052_ADC_TJUNC);
 
-static DEVICE_ATTR(name, S_IRUGO, da9052_hwmon_show_name, NULL);
-
-static struct attribute *da9052_attr[] = {
-	&dev_attr_name.attr,
+static struct attribute *da9052_attrs[] = {
 	&sensor_dev_attr_in0_input.dev_attr.attr,
 	&sensor_dev_attr_in0_label.dev_attr.attr,
 	&sensor_dev_attr_in3_input.dev_attr.attr,
@@ -268,57 +257,31 @@ static struct attribute *da9052_attr[] = {
 	NULL
 };
 
-static const struct attribute_group da9052_attr_group = {.attrs = da9052_attr};
+ATTRIBUTE_GROUPS(da9052);
 
 static int da9052_hwmon_probe(struct platform_device *pdev)
 {
+	struct device *dev = &pdev->dev;
 	struct da9052_hwmon *hwmon;
-	int ret;
+	struct device *hwmon_dev;
 
-	hwmon = devm_kzalloc(&pdev->dev, sizeof(struct da9052_hwmon),
-			     GFP_KERNEL);
+	hwmon = devm_kzalloc(dev, sizeof(struct da9052_hwmon), GFP_KERNEL);
 	if (!hwmon)
 		return -ENOMEM;
 
 	mutex_init(&hwmon->hwmon_lock);
 	hwmon->da9052 = dev_get_drvdata(pdev->dev.parent);
 
-	platform_set_drvdata(pdev, hwmon);
-
-	ret = sysfs_create_group(&pdev->dev.kobj, &da9052_attr_group);
-	if (ret)
-		goto err_mem;
-
-	hwmon->class_device = hwmon_device_register(&pdev->dev);
-	if (IS_ERR(hwmon->class_device)) {
-		ret = PTR_ERR(hwmon->class_device);
-		goto err_sysfs;
-	}
-
-	return 0;
-
-err_sysfs:
-	sysfs_remove_group(&pdev->dev.kobj, &da9052_attr_group);
-err_mem:
-	return ret;
-}
-
-static int da9052_hwmon_remove(struct platform_device *pdev)
-{
-	struct da9052_hwmon *hwmon = platform_get_drvdata(pdev);
-
-	hwmon_device_unregister(hwmon->class_device);
-	sysfs_remove_group(&pdev->dev.kobj, &da9052_attr_group);
-
-	return 0;
+	hwmon_dev = devm_hwmon_device_register_with_groups(dev, "da9052",
+							   hwmon,
+							   da9052_groups);
+	return PTR_ERR_OR_ZERO(hwmon_dev);
 }
 
 static struct platform_driver da9052_hwmon_driver = {
 	.probe = da9052_hwmon_probe,
-	.remove = da9052_hwmon_remove,
 	.driver = {
 		.name = "da9052-hwmon",
-		.owner = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/hwmon/da9055-hwmon.c b/drivers/hwmon/da9055-hwmon.c
index 35eb7738d711..f6e159cabe23 100644
--- a/drivers/hwmon/da9055-hwmon.c
+++ b/drivers/hwmon/da9055-hwmon.c
@@ -36,7 +36,6 @@
 
 struct da9055_hwmon {
 	struct da9055	*da9055;
-	struct device	*class_device;
 	struct mutex	hwmon_lock;
 	struct mutex	irq_lock;
 	struct completion done;
@@ -200,13 +199,6 @@ static ssize_t da9055_read_tjunc(struct device *dev,
 							+ 3076332, 10000));
 }
 
-static ssize_t da9055_hwmon_show_name(struct device *dev,
-				      struct device_attribute *devattr,
-				      char *buf)
-{
-	return sprintf(buf, "da9055\n");
-}
-
 static ssize_t show_label(struct device *dev,
 			  struct device_attribute *devattr, char *buf)
 {
@@ -236,10 +228,7 @@ static SENSOR_DEVICE_ATTR(temp1_input, S_IRUGO, da9055_read_tjunc, NULL,
 static SENSOR_DEVICE_ATTR(temp1_label, S_IRUGO, show_label, NULL,
 			  DA9055_ADC_TJUNC);
 
-static DEVICE_ATTR(name, S_IRUGO, da9055_hwmon_show_name, NULL);
-
-static struct attribute *da9055_attr[] = {
-	&dev_attr_name.attr,
+static struct attribute *da9055_attrs[] = {
 	&sensor_dev_attr_in0_input.dev_attr.attr,
 	&sensor_dev_attr_in0_label.dev_attr.attr,
 	&sensor_dev_attr_in1_input.dev_attr.attr,
@@ -254,15 +243,16 @@ static struct attribute *da9055_attr[] = {
 	NULL
 };
 
-static const struct attribute_group da9055_attr_group = {.attrs = da9055_attr};
+ATTRIBUTE_GROUPS(da9055);
 
 static int da9055_hwmon_probe(struct platform_device *pdev)
 {
+	struct device *dev = &pdev->dev;
 	struct da9055_hwmon *hwmon;
+	struct device *hwmon_dev;
 	int hwmon_irq, ret;
 
-	hwmon = devm_kzalloc(&pdev->dev, sizeof(struct da9055_hwmon),
-			     GFP_KERNEL);
+	hwmon = devm_kzalloc(dev, sizeof(struct da9055_hwmon), GFP_KERNEL);
 	if (!hwmon)
 		return -ENOMEM;
 
@@ -272,8 +262,6 @@ static int da9055_hwmon_probe(struct platform_device *pdev)
 	init_completion(&hwmon->done);
 	hwmon->da9055 = dev_get_drvdata(pdev->dev.parent);
 
-	platform_set_drvdata(pdev, hwmon);
-
 	hwmon_irq = platform_get_irq_byname(pdev, "HWMON");
 	if (hwmon_irq < 0)
 		return hwmon_irq;
@@ -288,39 +276,16 @@ static int da9055_hwmon_probe(struct platform_device *pdev)
 		return ret;
 	}
 
-	ret = sysfs_create_group(&pdev->dev.kobj, &da9055_attr_group);
-	if (ret)
-		return ret;
-
-	hwmon->class_device = hwmon_device_register(&pdev->dev);
-	if (IS_ERR(hwmon->class_device)) {
-		ret = PTR_ERR(hwmon->class_device);
-		goto err;
-	}
-
-	return 0;
-
-err:
-	sysfs_remove_group(&pdev->dev.kobj, &da9055_attr_group);
-	return ret;
-}
-
-static int da9055_hwmon_remove(struct platform_device *pdev)
-{
-	struct da9055_hwmon *hwmon = platform_get_drvdata(pdev);
-
-	sysfs_remove_group(&pdev->dev.kobj, &da9055_attr_group);
-	hwmon_device_unregister(hwmon->class_device);
-
-	return 0;
+	hwmon_dev = devm_hwmon_device_register_with_groups(dev, "da9055",
+							   hwmon,
+							   da9055_groups);
+	return PTR_ERR_OR_ZERO(hwmon_dev);
 }
 
 static struct platform_driver da9055_hwmon_driver = {
 	.probe = da9055_hwmon_probe,
-	.remove = da9055_hwmon_remove,
 	.driver = {
 		.name = "da9055-hwmon",
-		.owner = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/hwmon/dme1737.c b/drivers/hwmon/dme1737.c
index bea0a344fab5..8763c4a8280c 100644
--- a/drivers/hwmon/dme1737.c
+++ b/drivers/hwmon/dme1737.c
@@ -2732,7 +2732,6 @@ static int dme1737_isa_remove(struct platform_device *pdev)
 
 static struct platform_driver dme1737_isa_driver = {
 	.driver = {
-		.owner = THIS_MODULE,
 		.name = "dme1737",
 	},
 	.probe = dme1737_isa_probe,
diff --git a/drivers/hwmon/f71805f.c b/drivers/hwmon/f71805f.c
index 9e57b77ecd34..facd05cda26d 100644
--- a/drivers/hwmon/f71805f.c
+++ b/drivers/hwmon/f71805f.c
@@ -1503,7 +1503,6 @@ static int f71805f_remove(struct platform_device *pdev)
 
 static struct platform_driver f71805f_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= DRVNAME,
 	},
 	.probe		= f71805f_probe,
diff --git a/drivers/hwmon/f71882fg.c b/drivers/hwmon/f71882fg.c
index 03d8592810bf..2e5c6f46e442 100644
--- a/drivers/hwmon/f71882fg.c
+++ b/drivers/hwmon/f71882fg.c
@@ -369,7 +369,6 @@ static int f71882fg_remove(struct platform_device *pdev);
 
 static struct platform_driver f71882fg_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= DRVNAME,
 	},
 	.probe		= f71882fg_probe,
diff --git a/drivers/hwmon/fam15h_power.c b/drivers/hwmon/fam15h_power.c
index fcdbde4ec692..3057dfc7e3bc 100644
--- a/drivers/hwmon/fam15h_power.c
+++ b/drivers/hwmon/fam15h_power.c
@@ -234,7 +234,7 @@ static const struct pci_device_id fam15h_power_id_table[] = {
 	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_NB_F4) },
 	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_M30H_NB_F4) },
 	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_16H_NB_F4) },
-	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_16H_M30H_NB_F3) },
+	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_16H_M30H_NB_F4) },
 	{}
 };
 MODULE_DEVICE_TABLE(pci, fam15h_power_id_table);
diff --git a/drivers/hwmon/g762.c b/drivers/hwmon/g762.c
index 6aac695b1688..9b55e673b67c 100644
--- a/drivers/hwmon/g762.c
+++ b/drivers/hwmon/g762.c
@@ -1084,10 +1084,8 @@ static int g762_probe(struct i2c_client *client, const struct i2c_device_id *id)
 	if (ret)
 		goto clock_dis;
 
-	data->hwmon_dev = devm_hwmon_device_register_with_groups(dev,
-								 client->name,
-								 data,
-								 g762_groups);
+	data->hwmon_dev = hwmon_device_register_with_groups(dev, client->name,
+							    data, g762_groups);
 	if (IS_ERR(data->hwmon_dev)) {
 		ret = PTR_ERR(data->hwmon_dev);
 		goto clock_dis;
diff --git a/drivers/hwmon/gpio-fan.c b/drivers/hwmon/gpio-fan.c
index 4efa1734bdad..36abf814b8c7 100644
--- a/drivers/hwmon/gpio-fan.c
+++ b/drivers/hwmon/gpio-fan.c
@@ -79,7 +79,7 @@ static ssize_t show_fan_alarm(struct device *dev,
 {
 	struct gpio_fan_data *fan_data = dev_get_drvdata(dev);
 	struct gpio_fan_alarm *alarm = fan_data->alarm;
-	int value = gpio_get_value(alarm->gpio);
+	int value = gpio_get_value_cansleep(alarm->gpio);
 
 	if (alarm->active_low)
 		value = !value;
@@ -131,7 +131,7 @@ static void __set_fan_ctrl(struct gpio_fan_data *fan_data, int ctrl_val)
 	int i;
 
 	for (i = 0; i < fan_data->num_ctrl; i++)
-		gpio_set_value(fan_data->ctrl[i], (ctrl_val >> i) & 1);
+		gpio_set_value_cansleep(fan_data->ctrl[i], (ctrl_val >> i) & 1);
 }
 
 static int __get_fan_ctrl(struct gpio_fan_data *fan_data)
@@ -142,7 +142,7 @@ static int __get_fan_ctrl(struct gpio_fan_data *fan_data)
 	for (i = 0; i < fan_data->num_ctrl; i++) {
 		int value;
 
-		value = gpio_get_value(fan_data->ctrl[i]);
+		value = gpio_get_value_cansleep(fan_data->ctrl[i]);
 		ctrl_val |= (value << i);
 	}
 	return ctrl_val;
@@ -369,7 +369,8 @@ static int fan_ctrl_init(struct gpio_fan_data *fan_data,
 		if (err)
 			return err;
 
-		err = gpio_direction_output(ctrl[i], gpio_get_value(ctrl[i]));
+		err = gpio_direction_output(ctrl[i],
+					    gpio_get_value_cansleep(ctrl[i]));
 		if (err)
 			return err;
 	}
@@ -549,6 +550,14 @@ static int gpio_fan_probe(struct platform_device *pdev)
 	return 0;
 }
 
+static void gpio_fan_shutdown(struct platform_device *pdev)
+{
+	struct gpio_fan_data *fan_data = dev_get_drvdata(&pdev->dev);
+
+	if (fan_data->ctrl)
+		set_fan_speed(fan_data, 0);
+}
+
 #ifdef CONFIG_PM_SLEEP
 static int gpio_fan_suspend(struct device *dev)
 {
@@ -580,6 +589,7 @@ static SIMPLE_DEV_PM_OPS(gpio_fan_pm, gpio_fan_suspend, gpio_fan_resume);
 
 static struct platform_driver gpio_fan_driver = {
 	.probe		= gpio_fan_probe,
+	.shutdown	= gpio_fan_shutdown,
 	.driver	= {
 		.name	= "gpio-fan",
 		.pm	= GPIO_FAN_PM,
diff --git a/drivers/hwmon/i5500_temp.c b/drivers/hwmon/i5500_temp.c
new file mode 100644
index 000000000000..3e3ccbf18b4e
--- /dev/null
+++ b/drivers/hwmon/i5500_temp.c
@@ -0,0 +1,149 @@
+/*
+ * i5500_temp - Driver for Intel 5500/5520/X58 chipset thermal sensor
+ *
+ * Copyright (C) 2012, 2014 Jean Delvare <jdelvare@suse.de>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/slab.h>
+#include <linux/jiffies.h>
+#include <linux/device.h>
+#include <linux/pci.h>
+#include <linux/hwmon.h>
+#include <linux/hwmon-sysfs.h>
+#include <linux/err.h>
+#include <linux/mutex.h>
+
+/* Register definitions from datasheet */
+#define REG_TSTHRCATA	0xE2
+#define REG_TSCTRL	0xE8
+#define REG_TSTHRRPEX	0xEB
+#define REG_TSTHRLO	0xEC
+#define REG_TSTHRHI	0xEE
+#define REG_CTHINT	0xF0
+#define REG_TSFSC	0xF3
+#define REG_CTSTS	0xF4
+#define REG_TSTHRRQPI	0xF5
+#define REG_CTCTRL	0xF7
+#define REG_TSTIMER	0xF8
+
+/*
+ * Sysfs stuff
+ */
+
+/* Sensor resolution : 0.5 degree C */
+static ssize_t show_temp(struct device *dev,
+			 struct device_attribute *devattr, char *buf)
+{
+	struct pci_dev *pdev = to_pci_dev(dev->parent);
+	long temp;
+	u16 tsthrhi;
+	s8 tsfsc;
+
+	pci_read_config_word(pdev, REG_TSTHRHI, &tsthrhi);
+	pci_read_config_byte(pdev, REG_TSFSC, &tsfsc);
+	temp = ((long)tsthrhi - tsfsc) * 500;
+
+	return sprintf(buf, "%ld\n", temp);
+}
+
+static ssize_t show_thresh(struct device *dev,
+			   struct device_attribute *devattr, char *buf)
+{
+	struct pci_dev *pdev = to_pci_dev(dev->parent);
+	int reg = to_sensor_dev_attr(devattr)->index;
+	long temp;
+	u16 tsthr;
+
+	pci_read_config_word(pdev, reg, &tsthr);
+	temp = tsthr * 500;
+
+	return sprintf(buf, "%ld\n", temp);
+}
+
+static ssize_t show_alarm(struct device *dev,
+			  struct device_attribute *devattr, char *buf)
+{
+	struct pci_dev *pdev = to_pci_dev(dev->parent);
+	int nr = to_sensor_dev_attr(devattr)->index;
+	u8 ctsts;
+
+	pci_read_config_byte(pdev, REG_CTSTS, &ctsts);
+	return sprintf(buf, "%u\n", (unsigned int)ctsts & (1 << nr));
+}
+
+static DEVICE_ATTR(temp1_input, S_IRUGO, show_temp, NULL);
+static SENSOR_DEVICE_ATTR(temp1_crit, S_IRUGO, show_thresh, NULL, 0xE2);
+static SENSOR_DEVICE_ATTR(temp1_max_hyst, S_IRUGO, show_thresh, NULL, 0xEC);
+static SENSOR_DEVICE_ATTR(temp1_max, S_IRUGO, show_thresh, NULL, 0xEE);
+static SENSOR_DEVICE_ATTR(temp1_crit_alarm, S_IRUGO, show_alarm, NULL, 0);
+static SENSOR_DEVICE_ATTR(temp1_max_alarm, S_IRUGO, show_alarm, NULL, 1);
+
+static struct attribute *i5500_temp_attrs[] = {
+	&dev_attr_temp1_input.attr,
+	&sensor_dev_attr_temp1_crit.dev_attr.attr,
+	&sensor_dev_attr_temp1_max_hyst.dev_attr.attr,
+	&sensor_dev_attr_temp1_max.dev_attr.attr,
+	&sensor_dev_attr_temp1_crit_alarm.dev_attr.attr,
+	&sensor_dev_attr_temp1_max_alarm.dev_attr.attr,
+	NULL
+};
+
+ATTRIBUTE_GROUPS(i5500_temp);
+
+static const struct pci_device_id i5500_temp_ids[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x3438) },
+	{ 0 },
+};
+
+MODULE_DEVICE_TABLE(pci, i5500_temp_ids);
+
+static int i5500_temp_probe(struct pci_dev *pdev,
+			    const struct pci_device_id *id)
+{
+	int err;
+	struct device *hwmon_dev;
+	u32 tstimer;
+	s8 tsfsc;
+
+	err = pci_enable_device(pdev);
+	if (err) {
+		dev_err(&pdev->dev, "Failed to enable device\n");
+		return err;
+	}
+
+	pci_read_config_byte(pdev, REG_TSFSC, &tsfsc);
+	pci_read_config_dword(pdev, REG_TSTIMER, &tstimer);
+	if (tsfsc == 0x7F && tstimer == 0x07D30D40) {
+		dev_notice(&pdev->dev, "Sensor seems to be disabled\n");
+		return -ENODEV;
+	}
+
+	hwmon_dev = devm_hwmon_device_register_with_groups(&pdev->dev,
+							   "intel5500", NULL,
+							   i5500_temp_groups);
+	return PTR_ERR_OR_ZERO(hwmon_dev);
+}
+
+static struct pci_driver i5500_temp_driver = {
+	.name = "i5500_temp",
+	.id_table = i5500_temp_ids,
+	.probe = i5500_temp_probe,
+};
+
+module_pci_driver(i5500_temp_driver);
+
+MODULE_AUTHOR("Jean Delvare <jdelvare@suse.de>");
+MODULE_DESCRIPTION("Intel 5500/5520/X58 chipset thermal sensor driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/hwmon/i5k_amb.c b/drivers/hwmon/i5k_amb.c
index 6c0080a3b902..6b3d1972cef7 100644
--- a/drivers/hwmon/i5k_amb.c
+++ b/drivers/hwmon/i5k_amb.c
@@ -581,7 +581,6 @@ static int i5k_amb_remove(struct platform_device *pdev)
 
 static struct platform_driver i5k_amb_driver = {
 	.driver = {
-		.owner = THIS_MODULE,
 		.name = DRVNAME,
 	},
 	.probe = i5k_amb_probe,
diff --git a/drivers/hwmon/ibmpowernv.c b/drivers/hwmon/ibmpowernv.c
index d2bf2c97ae70..febe8175d36c 100644
--- a/drivers/hwmon/ibmpowernv.c
+++ b/drivers/hwmon/ibmpowernv.c
@@ -74,9 +74,6 @@ struct platform_data {
 	u32 sensors_count; /* Total count of sensors from each group */
 };
 
-/* Platform device representing all the ibmpowernv sensors */
-static struct platform_device *pdevice;
-
 static ssize_t show_sensor(struct device *dev, struct device_attribute *devattr,
 			   char *buf)
 {
@@ -99,7 +96,7 @@ static ssize_t show_sensor(struct device *dev, struct device_attribute *devattr,
 	return sprintf(buf, "%u\n", x);
 }
 
-static int __init get_sensor_index_attr(const char *name, u32 *index,
+static int get_sensor_index_attr(const char *name, u32 *index,
 					char *attr)
 {
 	char *hash_pos = strchr(name, '#');
@@ -136,7 +133,7 @@ static int __init get_sensor_index_attr(const char *name, u32 *index,
  * which need to be mapped as fan2_input, temp1_max respectively before
  * populating them inside hwmon device class.
  */
-static int __init create_hwmon_attr_name(struct device *dev, enum sensors type,
+static int create_hwmon_attr_name(struct device *dev, enum sensors type,
 					 const char *node_name,
 					 char *hwmon_attr_name)
 {
@@ -172,7 +169,7 @@ static int __init create_hwmon_attr_name(struct device *dev, enum sensors type,
 	return 0;
 }
 
-static int __init populate_attr_groups(struct platform_device *pdev)
+static int populate_attr_groups(struct platform_device *pdev)
 {
 	struct platform_data *pdata = platform_get_drvdata(pdev);
 	const struct attribute_group **pgroups = pdata->attr_groups;
@@ -180,11 +177,6 @@ static int __init populate_attr_groups(struct platform_device *pdev)
 	enum sensors type;
 
 	opal = of_find_node_by_path("/ibm,opal/sensors");
-	if (!opal) {
-		dev_err(&pdev->dev, "Opal node 'sensors' not found\n");
-		return -ENODEV;
-	}
-
 	for_each_child_of_node(opal, np) {
 		if (np->name == NULL)
 			continue;
@@ -221,7 +213,7 @@ static int __init populate_attr_groups(struct platform_device *pdev)
  * to the name required by the higher 'hwmon' driver like fan1_input, temp1_max
  * etc..
  */
-static int __init create_device_attrs(struct platform_device *pdev)
+static int create_device_attrs(struct platform_device *pdev)
 {
 	struct platform_data *pdata = platform_get_drvdata(pdev);
 	const struct attribute_group **pgroups = pdata->attr_groups;
@@ -280,7 +272,7 @@ exit_put_node:
 	return err;
 }
 
-static int __init ibmpowernv_probe(struct platform_device *pdev)
+static int ibmpowernv_probe(struct platform_device *pdev)
 {
 	struct platform_data *pdata;
 	struct device *hwmon_dev;
@@ -309,55 +301,24 @@ static int __init ibmpowernv_probe(struct platform_device *pdev)
 	return PTR_ERR_OR_ZERO(hwmon_dev);
 }
 
-static struct platform_driver ibmpowernv_driver = {
-	.driver = {
-		.owner = THIS_MODULE,
-		.name = DRVNAME,
+static const struct platform_device_id opal_sensor_driver_ids[] = {
+	{
+		.name = "opal-sensor",
 	},
+	{ }
 };
+MODULE_DEVICE_TABLE(platform, opal_sensor_driver_ids);
 
-static int __init ibmpowernv_init(void)
-{
-	int err;
-
-	pdevice = platform_device_alloc(DRVNAME, 0);
-	if (!pdevice) {
-		pr_err("Device allocation failed\n");
-		err = -ENOMEM;
-		goto exit;
-	}
-
-	err = platform_device_add(pdevice);
-	if (err) {
-		pr_err("Device addition failed (%d)\n", err);
-		goto exit_device_put;
-	}
-
-	err = platform_driver_probe(&ibmpowernv_driver, ibmpowernv_probe);
-	if (err) {
-		pr_err("Platfrom driver probe failed\n");
-		goto exit_device_del;
-	}
-
-	return 0;
-
-exit_device_del:
-	platform_device_del(pdevice);
-exit_device_put:
-	platform_device_put(pdevice);
-exit:
-	return err;
-}
+static struct platform_driver ibmpowernv_driver = {
+	.probe		= ibmpowernv_probe,
+	.id_table	= opal_sensor_driver_ids,
+	.driver		= {
+		.name	= DRVNAME,
+	},
+};
 
-static void __exit ibmpowernv_exit(void)
-{
-	platform_driver_unregister(&ibmpowernv_driver);
-	platform_device_unregister(pdevice);
-}
+module_platform_driver(ibmpowernv_driver);
 
 MODULE_AUTHOR("Neelesh Gupta <neelegup@linux.vnet.ibm.com>");
 MODULE_DESCRIPTION("IBM POWERNV platform sensors");
 MODULE_LICENSE("GPL");
-
-module_init(ibmpowernv_init);
-module_exit(ibmpowernv_exit);
diff --git a/drivers/hwmon/iio_hwmon.c b/drivers/hwmon/iio_hwmon.c
index 14c82daab019..17ae2eb26ce2 100644
--- a/drivers/hwmon/iio_hwmon.c
+++ b/drivers/hwmon/iio_hwmon.c
@@ -63,7 +63,7 @@ static int iio_hwmon_probe(struct platform_device *pdev)
 	struct iio_hwmon_state *st;
 	struct sensor_device_attribute *a;
 	int ret, i;
-	int in_i = 1, temp_i = 1, curr_i = 1;
+	int in_i = 1, temp_i = 1, curr_i = 1, humidity_i = 1;
 	enum iio_chan_type type;
 	struct iio_channel *channels;
 	const char *name = "iio_hwmon";
@@ -123,6 +123,11 @@ static int iio_hwmon_probe(struct platform_device *pdev)
 							  "curr%d_input",
 							  curr_i++);
 			break;
+		case IIO_HUMIDITYRELATIVE:
+			a->dev_attr.attr.name = kasprintf(GFP_KERNEL,
+							  "humidity%d_input",
+							  humidity_i++);
+			break;
 		default:
 			ret = -EINVAL;
 			goto error_release_channels;
@@ -172,7 +177,6 @@ MODULE_DEVICE_TABLE(of, iio_hwmon_of_match);
 static struct platform_driver __refdata iio_hwmon_driver = {
 	.driver = {
 		.name = "iio_hwmon",
-		.owner = THIS_MODULE,
 		.of_match_table = iio_hwmon_of_match,
 	},
 	.probe = iio_hwmon_probe,
diff --git a/drivers/hwmon/ina2xx.c b/drivers/hwmon/ina2xx.c
index bfd3f3eeabcd..e01feba909c3 100644
--- a/drivers/hwmon/ina2xx.c
+++ b/drivers/hwmon/ina2xx.c
@@ -223,6 +223,7 @@ static int ina2xx_probe(struct i2c_client *client,
 	struct device *hwmon_dev;
 	long shunt = 10000; /* default shunt value 10mOhms */
 	u32 val;
+	int ret;
 
 	if (!i2c_check_functionality(adapter, I2C_FUNC_SMBUS_WORD_DATA))
 		return -ENODEV;
@@ -247,12 +248,25 @@ static int ina2xx_probe(struct i2c_client *client,
 	data->config = &ina2xx_config[data->kind];
 
 	/* device configuration */
-	i2c_smbus_write_word_swapped(client, INA2XX_CONFIG,
-				     data->config->config_default);
-	/* set current LSB to 1mA, shunt is in uOhms */
-	/* (equation 13 in datasheet) */
-	i2c_smbus_write_word_swapped(client, INA2XX_CALIBRATION,
-				     data->config->calibration_factor / shunt);
+	ret = i2c_smbus_write_word_swapped(client, INA2XX_CONFIG,
+					   data->config->config_default);
+	if (ret < 0) {
+		dev_err(dev,
+			"error writing to the config register: %d", ret);
+		return -ENODEV;
+	}
+
+	/*
+	 * Set current LSB to 1mA, shunt is in uOhms
+	 * (equation 13 in datasheet).
+	 */
+	ret = i2c_smbus_write_word_swapped(client, INA2XX_CALIBRATION,
+				data->config->calibration_factor / shunt);
+	if (ret < 0) {
+		dev_err(dev,
+			"error writing to the calibration register: %d", ret);
+		return -ENODEV;
+	}
 
 	data->client = client;
 	mutex_init(&data->update_lock);
diff --git a/drivers/hwmon/it87.c b/drivers/hwmon/it87.c
index a327fd3402a7..409116c52cc5 100644
--- a/drivers/hwmon/it87.c
+++ b/drivers/hwmon/it87.c
@@ -498,7 +498,6 @@ static void it87_init_device(struct platform_device *pdev);
 
 static struct platform_driver it87_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= DRVNAME,
 	},
 	.probe	= it87_probe,
diff --git a/drivers/hwmon/jz4740-hwmon.c b/drivers/hwmon/jz4740-hwmon.c
index 7488e36809c8..df9b3447f2a8 100644
--- a/drivers/hwmon/jz4740-hwmon.c
+++ b/drivers/hwmon/jz4740-hwmon.c
@@ -172,7 +172,6 @@ static struct platform_driver jz4740_hwmon_driver = {
 	.remove = jz4740_hwmon_remove,
 	.driver = {
 		.name = "jz4740-hwmon",
-		.owner = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/hwmon/k10temp.c b/drivers/hwmon/k10temp.c
index f7b46f68ef43..1e7bdcdcb295 100644
--- a/drivers/hwmon/k10temp.c
+++ b/drivers/hwmon/k10temp.c
@@ -33,6 +33,9 @@ static bool force;
 module_param(force, bool, 0444);
 MODULE_PARM_DESC(force, "force loading on processors with erratum 319");
 
+/* Provide lock for writing to NB_SMU_IND_ADDR */
+static DEFINE_MUTEX(nb_smu_ind_mutex);
+
 /* CPUID function 0x80000001, ebx */
 #define CPUID_PKGTYPE_MASK	0xf0000000
 #define CPUID_PKGTYPE_F		0x00000000
@@ -51,13 +54,38 @@ MODULE_PARM_DESC(force, "force loading on processors with erratum 319");
 #define REG_NORTHBRIDGE_CAPABILITIES	0xe8
 #define  NB_CAP_HTC			0x00000400
 
+/*
+ * For F15h M60h, functionality of REG_REPORTED_TEMPERATURE
+ * has been moved to D0F0xBC_xD820_0CA4 [Reported Temperature
+ * Control]
+ */
+#define F15H_M60H_REPORTED_TEMP_CTRL_OFFSET	0xd8200ca4
+#define PCI_DEVICE_ID_AMD_15H_M60H_NB_F3	0x1573
+
+static void amd_nb_smu_index_read(struct pci_dev *pdev, unsigned int devfn,
+				  int offset, u32 *val)
+{
+	mutex_lock(&nb_smu_ind_mutex);
+	pci_bus_write_config_dword(pdev->bus, devfn,
+				   0xb8, offset);
+	pci_bus_read_config_dword(pdev->bus, devfn,
+				  0xbc, val);
+	mutex_unlock(&nb_smu_ind_mutex);
+}
+
 static ssize_t show_temp(struct device *dev,
 			 struct device_attribute *attr, char *buf)
 {
 	u32 regval;
-
-	pci_read_config_dword(to_pci_dev(dev),
-			      REG_REPORTED_TEMPERATURE, &regval);
+	struct pci_dev *pdev = dev_get_drvdata(dev);
+
+	if (boot_cpu_data.x86 == 0x15 && boot_cpu_data.x86_model == 0x60) {
+		amd_nb_smu_index_read(pdev, PCI_DEVFN(0, 0),
+				      F15H_M60H_REPORTED_TEMP_CTRL_OFFSET,
+				      &regval);
+	} else {
+		pci_read_config_dword(pdev, REG_REPORTED_TEMPERATURE, &regval);
+	}
 	return sprintf(buf, "%u\n", (regval >> 21) * 125);
 }
 
@@ -75,7 +103,7 @@ static ssize_t show_temp_crit(struct device *dev,
 	u32 regval;
 	int value;
 
-	pci_read_config_dword(to_pci_dev(dev),
+	pci_read_config_dword(dev_get_drvdata(dev),
 			      REG_HARDWARE_THERMAL_CONTROL, &regval);
 	value = ((regval >> 16) & 0x7f) * 500 + 52000;
 	if (show_hyst)
@@ -83,17 +111,43 @@ static ssize_t show_temp_crit(struct device *dev,
 	return sprintf(buf, "%d\n", value);
 }
 
-static ssize_t show_name(struct device *dev,
-			 struct device_attribute *attr, char *buf)
-{
-	return sprintf(buf, "k10temp\n");
-}
-
 static DEVICE_ATTR(temp1_input, S_IRUGO, show_temp, NULL);
 static DEVICE_ATTR(temp1_max, S_IRUGO, show_temp_max, NULL);
 static SENSOR_DEVICE_ATTR(temp1_crit, S_IRUGO, show_temp_crit, NULL, 0);
 static SENSOR_DEVICE_ATTR(temp1_crit_hyst, S_IRUGO, show_temp_crit, NULL, 1);
-static DEVICE_ATTR(name, S_IRUGO, show_name, NULL);
+
+static umode_t k10temp_is_visible(struct kobject *kobj,
+				  struct attribute *attr, int index)
+{
+	struct device *dev = container_of(kobj, struct device, kobj);
+	struct pci_dev *pdev = dev_get_drvdata(dev);
+
+	if (index >= 2) {
+		u32 reg_caps, reg_htc;
+
+		pci_read_config_dword(pdev, REG_NORTHBRIDGE_CAPABILITIES,
+				      &reg_caps);
+		pci_read_config_dword(pdev, REG_HARDWARE_THERMAL_CONTROL,
+				      &reg_htc);
+		if (!(reg_caps & NB_CAP_HTC) || !(reg_htc & HTC_ENABLE))
+			return 0;
+	}
+	return attr->mode;
+}
+
+static struct attribute *k10temp_attrs[] = {
+	&dev_attr_temp1_input.attr,
+	&dev_attr_temp1_max.attr,
+	&sensor_dev_attr_temp1_crit.dev_attr.attr,
+	&sensor_dev_attr_temp1_crit_hyst.dev_attr.attr,
+	NULL
+};
+
+static const struct attribute_group k10temp_group = {
+	.attrs = k10temp_attrs,
+	.is_visible = k10temp_is_visible,
+};
+__ATTRIBUTE_GROUPS(k10temp);
 
 static bool has_erratum_319(struct pci_dev *pdev)
 {
@@ -132,76 +186,23 @@ static bool has_erratum_319(struct pci_dev *pdev)
 static int k10temp_probe(struct pci_dev *pdev,
 				   const struct pci_device_id *id)
 {
-	struct device *hwmon_dev;
-	u32 reg_caps, reg_htc;
 	int unreliable = has_erratum_319(pdev);
-	int err;
-
-	if (unreliable && !force) {
-		dev_err(&pdev->dev,
-			"unreliable CPU thermal sensor; monitoring disabled\n");
-		err = -ENODEV;
-		goto exit;
-	}
-
-	err = device_create_file(&pdev->dev, &dev_attr_temp1_input);
-	if (err)
-		goto exit;
-	err = device_create_file(&pdev->dev, &dev_attr_temp1_max);
-	if (err)
-		goto exit_remove;
-
-	pci_read_config_dword(pdev, REG_NORTHBRIDGE_CAPABILITIES, &reg_caps);
-	pci_read_config_dword(pdev, REG_HARDWARE_THERMAL_CONTROL, &reg_htc);
-	if ((reg_caps & NB_CAP_HTC) && (reg_htc & HTC_ENABLE)) {
-		err = device_create_file(&pdev->dev,
-				&sensor_dev_attr_temp1_crit.dev_attr);
-		if (err)
-			goto exit_remove;
-		err = device_create_file(&pdev->dev,
-				&sensor_dev_attr_temp1_crit_hyst.dev_attr);
-		if (err)
-			goto exit_remove;
-	}
-
-	err = device_create_file(&pdev->dev, &dev_attr_name);
-	if (err)
-		goto exit_remove;
-
-	hwmon_dev = hwmon_device_register(&pdev->dev);
-	if (IS_ERR(hwmon_dev)) {
-		err = PTR_ERR(hwmon_dev);
-		goto exit_remove;
-	}
-	pci_set_drvdata(pdev, hwmon_dev);
+	struct device *dev = &pdev->dev;
+	struct device *hwmon_dev;
 
-	if (unreliable && force)
-		dev_warn(&pdev->dev,
+	if (unreliable) {
+		if (!force) {
+			dev_err(dev,
+				"unreliable CPU thermal sensor; monitoring disabled\n");
+			return -ENODEV;
+		}
+		dev_warn(dev,
 			 "unreliable CPU thermal sensor; check erratum 319\n");
-	return 0;
-
-exit_remove:
-	device_remove_file(&pdev->dev, &dev_attr_name);
-	device_remove_file(&pdev->dev, &dev_attr_temp1_input);
-	device_remove_file(&pdev->dev, &dev_attr_temp1_max);
-	device_remove_file(&pdev->dev,
-			   &sensor_dev_attr_temp1_crit.dev_attr);
-	device_remove_file(&pdev->dev,
-			   &sensor_dev_attr_temp1_crit_hyst.dev_attr);
-exit:
-	return err;
-}
+	}
 
-static void k10temp_remove(struct pci_dev *pdev)
-{
-	hwmon_device_unregister(pci_get_drvdata(pdev));
-	device_remove_file(&pdev->dev, &dev_attr_name);
-	device_remove_file(&pdev->dev, &dev_attr_temp1_input);
-	device_remove_file(&pdev->dev, &dev_attr_temp1_max);
-	device_remove_file(&pdev->dev,
-			   &sensor_dev_attr_temp1_crit.dev_attr);
-	device_remove_file(&pdev->dev,
-			   &sensor_dev_attr_temp1_crit_hyst.dev_attr);
+	hwmon_dev = devm_hwmon_device_register_with_groups(dev, "k10temp", pdev,
+							   k10temp_groups);
+	return PTR_ERR_OR_ZERO(hwmon_dev);
 }
 
 static const struct pci_device_id k10temp_id_table[] = {
@@ -211,6 +212,7 @@ static const struct pci_device_id k10temp_id_table[] = {
 	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_NB_F3) },
 	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_M10H_F3) },
 	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_M30H_NB_F3) },
+	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_M60H_NB_F3) },
 	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_16H_NB_F3) },
 	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_16H_M30H_NB_F3) },
 	{}
@@ -221,7 +223,6 @@ static struct pci_driver k10temp_driver = {
 	.name = "k10temp",
 	.id_table = k10temp_id_table,
 	.probe = k10temp_probe,
-	.remove = k10temp_remove,
 };
 
 module_pci_driver(k10temp_driver);
diff --git a/drivers/hwmon/lm75.c b/drivers/hwmon/lm75.c
index d16dbb33a531..fe41d5ae7cb2 100644
--- a/drivers/hwmon/lm75.c
+++ b/drivers/hwmon/lm75.c
@@ -44,6 +44,7 @@ enum lm75_type {		/* keep sorted in alphabetical order */
 	g751,
 	lm75,
 	lm75a,
+	lm75b,
 	max6625,
 	max6626,
 	mcp980x,
@@ -176,6 +177,10 @@ static struct attribute *lm75_attrs[] = {
 };
 ATTRIBUTE_GROUPS(lm75);
 
+static const struct thermal_zone_of_device_ops lm75_of_thermal_ops = {
+	.get_temp = lm75_read_temp,
+};
+
 /*-----------------------------------------------------------------------*/
 
 /* device probe and removal */
@@ -233,6 +238,10 @@ lm75_probe(struct i2c_client *client, const struct i2c_device_id *id)
 		data->resolution = 9;
 		data->sample_time = HZ / 2;
 		break;
+	case lm75b:
+		data->resolution = 11;
+		data->sample_time = HZ / 4;
+		break;
 	case max6625:
 		data->resolution = 9;
 		data->sample_time = HZ / 4;
@@ -291,10 +300,9 @@ lm75_probe(struct i2c_client *client, const struct i2c_device_id *id)
 	if (IS_ERR(data->hwmon_dev))
 		return PTR_ERR(data->hwmon_dev);
 
-	data->tz = thermal_zone_of_sensor_register(data->hwmon_dev,
-						   0,
+	data->tz = thermal_zone_of_sensor_register(data->hwmon_dev, 0,
 						   data->hwmon_dev,
-						   lm75_read_temp, NULL);
+						   &lm75_of_thermal_ops);
 	if (IS_ERR(data->tz))
 		data->tz = NULL;
 
@@ -322,6 +330,7 @@ static const struct i2c_device_id lm75_ids[] = {
 	{ "g751", g751, },
 	{ "lm75", lm75, },
 	{ "lm75a", lm75a, },
+	{ "lm75b", lm75b, },
 	{ "max6625", max6625, },
 	{ "max6626", max6626, },
 	{ "mcp980x", mcp980x, },
@@ -409,6 +418,12 @@ static int lm75_detect(struct i2c_client *new_client,
 		 || i2c_smbus_read_byte_data(new_client, 7) != os)
 			return -ENODEV;
 	}
+	/*
+	 * It is very unlikely that this is a LM75 if both
+	 * hysteresis and temperature limit registers are 0.
+	 */
+	if (hyst == 0 && os == 0)
+		return -ENODEV;
 
 	/* Addresses cycling */
 	for (i = 8; i <= 248; i += 40) {
diff --git a/drivers/hwmon/lm78.c b/drivers/hwmon/lm78.c
index 759661c7d480..539efe4ad991 100644
--- a/drivers/hwmon/lm78.c
+++ b/drivers/hwmon/lm78.c
@@ -836,7 +836,6 @@ static int lm78_isa_probe(struct platform_device *pdev)
 
 static struct platform_driver lm78_isa_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= "lm78",
 	},
 	.probe		= lm78_isa_probe,
diff --git a/drivers/hwmon/lm95234.c b/drivers/hwmon/lm95234.c
index 411202bdaf6b..8796de39ff9b 100644
--- a/drivers/hwmon/lm95234.c
+++ b/drivers/hwmon/lm95234.c
@@ -1,7 +1,7 @@
 /*
  * Driver for Texas Instruments / National Semiconductor LM95234
  *
- * Copyright (c) 2013 Guenter Roeck <linux@roeck-us.net>
+ * Copyright (c) 2013, 2014 Guenter Roeck <linux@roeck-us.net>
  *
  * Derived from lm95241.c
  * Copyright (C) 2008, 2010 Davide Rizzo <elpa.rizzo@gmail.com>
@@ -30,7 +30,10 @@
 
 #define DRVNAME "lm95234"
 
-static const unsigned short normal_i2c[] = { 0x18, 0x4d, 0x4e, I2C_CLIENT_END };
+enum chips { lm95233, lm95234 };
+
+static const unsigned short normal_i2c[] = {
+	0x18, 0x2a, 0x2b, 0x4d, 0x4e, I2C_CLIENT_END };
 
 /* LM95234 registers */
 #define LM95234_REG_MAN_ID		0xFE
@@ -53,11 +56,13 @@ static const unsigned short normal_i2c[] = { 0x18, 0x4d, 0x4e, I2C_CLIENT_END };
 #define LM95234_REG_TCRIT_HYST		0x5a
 
 #define NATSEMI_MAN_ID			0x01
+#define LM95233_CHIP_ID			0x89
 #define LM95234_CHIP_ID			0x79
 
 /* Client data (each client gets its own) */
 struct lm95234_data {
 	struct i2c_client *client;
+	const struct attribute_group *groups[3];
 	struct mutex update_lock;
 	unsigned long last_updated, interval;	/* in jiffies */
 	bool valid;		/* false until following fields are valid */
@@ -564,35 +569,23 @@ static SENSOR_DEVICE_ATTR(temp5_offset, S_IWUSR | S_IRUGO, show_offset,
 static DEVICE_ATTR(update_interval, S_IWUSR | S_IRUGO, show_interval,
 		   set_interval);
 
-static struct attribute *lm95234_attrs[] = {
+static struct attribute *lm95234_common_attrs[] = {
 	&sensor_dev_attr_temp1_input.dev_attr.attr,
 	&sensor_dev_attr_temp2_input.dev_attr.attr,
 	&sensor_dev_attr_temp3_input.dev_attr.attr,
-	&sensor_dev_attr_temp4_input.dev_attr.attr,
-	&sensor_dev_attr_temp5_input.dev_attr.attr,
 	&sensor_dev_attr_temp2_fault.dev_attr.attr,
 	&sensor_dev_attr_temp3_fault.dev_attr.attr,
-	&sensor_dev_attr_temp4_fault.dev_attr.attr,
-	&sensor_dev_attr_temp5_fault.dev_attr.attr,
 	&sensor_dev_attr_temp2_type.dev_attr.attr,
 	&sensor_dev_attr_temp3_type.dev_attr.attr,
-	&sensor_dev_attr_temp4_type.dev_attr.attr,
-	&sensor_dev_attr_temp5_type.dev_attr.attr,
 	&sensor_dev_attr_temp1_max.dev_attr.attr,
 	&sensor_dev_attr_temp2_max.dev_attr.attr,
 	&sensor_dev_attr_temp3_max.dev_attr.attr,
-	&sensor_dev_attr_temp4_max.dev_attr.attr,
-	&sensor_dev_attr_temp5_max.dev_attr.attr,
 	&sensor_dev_attr_temp1_max_hyst.dev_attr.attr,
 	&sensor_dev_attr_temp2_max_hyst.dev_attr.attr,
 	&sensor_dev_attr_temp3_max_hyst.dev_attr.attr,
-	&sensor_dev_attr_temp4_max_hyst.dev_attr.attr,
-	&sensor_dev_attr_temp5_max_hyst.dev_attr.attr,
 	&sensor_dev_attr_temp1_max_alarm.dev_attr.attr,
 	&sensor_dev_attr_temp2_max_alarm.dev_attr.attr,
 	&sensor_dev_attr_temp3_max_alarm.dev_attr.attr,
-	&sensor_dev_attr_temp4_max_alarm.dev_attr.attr,
-	&sensor_dev_attr_temp5_max_alarm.dev_attr.attr,
 	&sensor_dev_attr_temp2_crit.dev_attr.attr,
 	&sensor_dev_attr_temp3_crit.dev_attr.attr,
 	&sensor_dev_attr_temp2_crit_hyst.dev_attr.attr,
@@ -601,18 +594,44 @@ static struct attribute *lm95234_attrs[] = {
 	&sensor_dev_attr_temp3_crit_alarm.dev_attr.attr,
 	&sensor_dev_attr_temp2_offset.dev_attr.attr,
 	&sensor_dev_attr_temp3_offset.dev_attr.attr,
+	&dev_attr_update_interval.attr,
+	NULL
+};
+
+static const struct attribute_group lm95234_common_group = {
+	.attrs = lm95234_common_attrs,
+};
+
+static struct attribute *lm95234_attrs[] = {
+	&sensor_dev_attr_temp4_input.dev_attr.attr,
+	&sensor_dev_attr_temp5_input.dev_attr.attr,
+	&sensor_dev_attr_temp4_fault.dev_attr.attr,
+	&sensor_dev_attr_temp5_fault.dev_attr.attr,
+	&sensor_dev_attr_temp4_type.dev_attr.attr,
+	&sensor_dev_attr_temp5_type.dev_attr.attr,
+	&sensor_dev_attr_temp4_max.dev_attr.attr,
+	&sensor_dev_attr_temp5_max.dev_attr.attr,
+	&sensor_dev_attr_temp4_max_hyst.dev_attr.attr,
+	&sensor_dev_attr_temp5_max_hyst.dev_attr.attr,
+	&sensor_dev_attr_temp4_max_alarm.dev_attr.attr,
+	&sensor_dev_attr_temp5_max_alarm.dev_attr.attr,
 	&sensor_dev_attr_temp4_offset.dev_attr.attr,
 	&sensor_dev_attr_temp5_offset.dev_attr.attr,
-	&dev_attr_update_interval.attr,
 	NULL
 };
-ATTRIBUTE_GROUPS(lm95234);
+
+static const struct attribute_group lm95234_group = {
+	.attrs = lm95234_attrs,
+};
 
 static int lm95234_detect(struct i2c_client *client,
 			  struct i2c_board_info *info)
 {
 	struct i2c_adapter *adapter = client->adapter;
+	int address = client->addr;
+	u8 config_mask, model_mask;
 	int mfg_id, chip_id, val;
+	const char *name;
 
 	if (!i2c_check_functionality(adapter, I2C_FUNC_SMBUS_BYTE_DATA))
 		return -ENODEV;
@@ -622,15 +641,31 @@ static int lm95234_detect(struct i2c_client *client,
 		return -ENODEV;
 
 	chip_id = i2c_smbus_read_byte_data(client, LM95234_REG_CHIP_ID);
-	if (chip_id != LM95234_CHIP_ID)
+	switch (chip_id) {
+	case LM95233_CHIP_ID:
+		if (address != 0x18 && address != 0x2a && address != 0x2b)
+			return -ENODEV;
+		config_mask = 0xbf;
+		model_mask = 0xf9;
+		name = "lm95233";
+		break;
+	case LM95234_CHIP_ID:
+		if (address != 0x18 && address != 0x4d && address != 0x4e)
+			return -ENODEV;
+		config_mask = 0xbc;
+		model_mask = 0xe1;
+		name = "lm95234";
+		break;
+	default:
 		return -ENODEV;
+	}
 
 	val = i2c_smbus_read_byte_data(client, LM95234_REG_STATUS);
 	if (val & 0x30)
 		return -ENODEV;
 
 	val = i2c_smbus_read_byte_data(client, LM95234_REG_CONFIG);
-	if (val & 0xbc)
+	if (val & config_mask)
 		return -ENODEV;
 
 	val = i2c_smbus_read_byte_data(client, LM95234_REG_CONVRATE);
@@ -638,14 +673,14 @@ static int lm95234_detect(struct i2c_client *client,
 		return -ENODEV;
 
 	val = i2c_smbus_read_byte_data(client, LM95234_REG_REM_MODEL);
-	if (val & 0xe1)
+	if (val & model_mask)
 		return -ENODEV;
 
 	val = i2c_smbus_read_byte_data(client, LM95234_REG_REM_MODEL_STS);
-	if (val & 0xe1)
+	if (val & model_mask)
 		return -ENODEV;
 
-	strlcpy(info->type, "lm95234", I2C_NAME_SIZE);
+	strlcpy(info->type, name, I2C_NAME_SIZE);
 	return 0;
 }
 
@@ -698,15 +733,19 @@ static int lm95234_probe(struct i2c_client *client,
 	if (err < 0)
 		return err;
 
+	data->groups[0] = &lm95234_common_group;
+	if (id->driver_data == lm95234)
+		data->groups[1] = &lm95234_group;
+
 	hwmon_dev = devm_hwmon_device_register_with_groups(dev, client->name,
-							   data,
-							   lm95234_groups);
+							   data, data->groups);
 	return PTR_ERR_OR_ZERO(hwmon_dev);
 }
 
 /* Driver data (common to all clients) */
 static const struct i2c_device_id lm95234_id[] = {
-	{ "lm95234", 0 },
+	{ "lm95233", lm95233 },
+	{ "lm95234", lm95234 },
 	{ }
 };
 MODULE_DEVICE_TABLE(i2c, lm95234_id);
@@ -725,5 +764,5 @@ static struct i2c_driver lm95234_driver = {
 module_i2c_driver(lm95234_driver);
 
 MODULE_AUTHOR("Guenter Roeck <linux@roeck-us.net>");
-MODULE_DESCRIPTION("LM95234 sensor driver");
+MODULE_DESCRIPTION("LM95233/LM95234 sensor driver");
 MODULE_LICENSE("GPL");
diff --git a/drivers/hwmon/lm95245.c b/drivers/hwmon/lm95245.c
index 0ae0dfdafdff..e7aef4561c83 100644
--- a/drivers/hwmon/lm95245.c
+++ b/drivers/hwmon/lm95245.c
@@ -1,10 +1,8 @@
 /*
  * Copyright (C) 2011 Alexander Stein <alexander.stein@systec-electronic.com>
  *
- * The LM95245 is a sensor chip made by National Semiconductors.
+ * The LM95245 is a sensor chip made by TI / National Semiconductor.
  * It reports up to two temperatures (its own plus an external one).
- * Complete datasheet can be obtained from National's website at:
- *   http://www.national.com/ds.cgi/LM/LM95245.pdf
  *
  * This driver is based on lm95241.c
  *
@@ -34,8 +32,6 @@
 #include <linux/mutex.h>
 #include <linux/sysfs.h>
 
-#define DEVNAME "lm95245"
-
 static const unsigned short normal_i2c[] = {
 	0x18, 0x19, 0x29, 0x4c, 0x4d, I2C_CLIENT_END };
 
@@ -98,7 +94,8 @@ static const unsigned short normal_i2c[] = {
 #define STATUS1_LOC		0x01
 
 #define MANUFACTURER_ID		0x01
-#define DEFAULT_REVISION	0xB3
+#define LM95235_REVISION	0xB1
+#define LM95245_REVISION	0xB3
 
 static const u8 lm95245_reg_address[] = {
 	LM95245_REG_R_LOCAL_TEMPH_S,
@@ -427,17 +424,32 @@ static int lm95245_detect(struct i2c_client *new_client,
 			  struct i2c_board_info *info)
 {
 	struct i2c_adapter *adapter = new_client->adapter;
+	int address = new_client->addr;
+	const char *name;
+	int rev, id;
 
 	if (!i2c_check_functionality(adapter, I2C_FUNC_SMBUS_BYTE_DATA))
 		return -ENODEV;
 
-	if (i2c_smbus_read_byte_data(new_client, LM95245_REG_R_MAN_ID)
-			!= MANUFACTURER_ID
-		|| i2c_smbus_read_byte_data(new_client, LM95245_REG_R_CHIP_ID)
-			!= DEFAULT_REVISION)
+	id = i2c_smbus_read_byte_data(new_client, LM95245_REG_R_MAN_ID);
+	if (id != MANUFACTURER_ID)
 		return -ENODEV;
 
-	strlcpy(info->type, DEVNAME, I2C_NAME_SIZE);
+	rev = i2c_smbus_read_byte_data(new_client, LM95245_REG_R_CHIP_ID);
+	switch (rev) {
+	case LM95235_REVISION:
+		if (address != 0x18 && address != 0x29 && address != 0x4c)
+			return -ENODEV;
+		name = "lm95235";
+		break;
+	case LM95245_REVISION:
+		name = "lm95245";
+		break;
+	default:
+		return -ENODEV;
+	}
+
+	strlcpy(info->type, name, I2C_NAME_SIZE);
 	return 0;
 }
 
@@ -484,7 +496,8 @@ static int lm95245_probe(struct i2c_client *client,
 
 /* Driver data (common to all clients) */
 static const struct i2c_device_id lm95245_id[] = {
-	{ DEVNAME, 0 },
+	{ "lm95235", 0 },
+	{ "lm95245", 0 },
 	{ }
 };
 MODULE_DEVICE_TABLE(i2c, lm95245_id);
@@ -492,7 +505,7 @@ MODULE_DEVICE_TABLE(i2c, lm95245_id);
 static struct i2c_driver lm95245_driver = {
 	.class		= I2C_CLASS_HWMON,
 	.driver = {
-		.name	= DEVNAME,
+		.name	= "lm95245",
 	},
 	.probe		= lm95245_probe,
 	.id_table	= lm95245_id,
@@ -503,5 +516,5 @@ static struct i2c_driver lm95245_driver = {
 module_i2c_driver(lm95245_driver);
 
 MODULE_AUTHOR("Alexander Stein <alexander.stein@systec-electronic.com>");
-MODULE_DESCRIPTION("LM95245 sensor driver");
+MODULE_DESCRIPTION("LM95235/LM95245 sensor driver");
 MODULE_LICENSE("GPL");
diff --git a/drivers/hwmon/max197.c b/drivers/hwmon/max197.c
index 82128ad79a91..cb0dcfda958c 100644
--- a/drivers/hwmon/max197.c
+++ b/drivers/hwmon/max197.c
@@ -334,7 +334,6 @@ MODULE_DEVICE_TABLE(platform, max197_device_ids);
 static struct platform_driver max197_driver = {
 	.driver = {
 		.name = "max197",
-		.owner = THIS_MODULE,
 	},
 	.probe = max197_probe,
 	.remove = max197_remove,
diff --git a/drivers/hwmon/mc13783-adc.c b/drivers/hwmon/mc13783-adc.c
index ae00e60d856c..0c02f40eb0c1 100644
--- a/drivers/hwmon/mc13783-adc.c
+++ b/drivers/hwmon/mc13783-adc.c
@@ -267,7 +267,6 @@ MODULE_DEVICE_TABLE(platform, mc13783_adc_idtable);
 static struct platform_driver mc13783_adc_driver = {
 	.remove		= mc13783_adc_remove,
 	.driver		= {
-		.owner	= THIS_MODULE,
 		.name	= DRIVER_NAME,
 	},
 	.id_table	= mc13783_adc_idtable,
diff --git a/drivers/hwmon/menf21bmc_hwmon.c b/drivers/hwmon/menf21bmc_hwmon.c
new file mode 100644
index 000000000000..c29a4c3c6b9e
--- /dev/null
+++ b/drivers/hwmon/menf21bmc_hwmon.c
@@ -0,0 +1,230 @@
+/*
+ *  MEN 14F021P00 Board Management Controller (BMC) hwmon driver.
+ *
+ *  This is the core hwmon driver of the MEN 14F021P00 BMC.
+ *  The BMC monitors the board voltages which can be access with this
+ *  driver through sysfs.
+ *
+ *  Copyright (C) 2014 MEN Mikro Elektronik Nuernberg GmbH
+ *
+ *  This program is free software; you can redistribute  it and/or modify it
+ *  under  the terms of  the GNU General  Public License as published by the
+ *  Free Software Foundation;  either version 2 of the  License, or (at your
+ *  option) any later version.
+ */
+
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/platform_device.h>
+#include <linux/hwmon.h>
+#include <linux/hwmon-sysfs.h>
+#include <linux/jiffies.h>
+#include <linux/slab.h>
+#include <linux/i2c.h>
+#include <linux/err.h>
+
+#define DRV_NAME  "menf21bmc_hwmon"
+
+#define BMC_VOLT_COUNT	5
+#define MENF21BMC_V33	0
+#define MENF21BMC_V5	1
+#define MENF21BMC_V12	2
+#define MENF21BMC_V5_SB	3
+#define MENF21BMC_VBAT	4
+
+#define IDX_TO_VOLT_MIN_CMD(idx) (0x40 + idx)
+#define IDX_TO_VOLT_MAX_CMD(idx) (0x50 + idx)
+#define IDX_TO_VOLT_INP_CMD(idx) (0x60 + idx)
+
+struct menf21bmc_hwmon {
+	bool valid;
+	struct i2c_client *i2c_client;
+	unsigned long last_update;
+	int in_val[BMC_VOLT_COUNT];
+	int in_min[BMC_VOLT_COUNT];
+	int in_max[BMC_VOLT_COUNT];
+};
+
+static const char *const input_names[] = {
+	[MENF21BMC_V33]		= "MON_3_3V",
+	[MENF21BMC_V5]		= "MON_5V",
+	[MENF21BMC_V12]		= "MON_12V",
+	[MENF21BMC_V5_SB]	= "5V_STANDBY",
+	[MENF21BMC_VBAT]	= "VBAT"
+};
+
+static struct menf21bmc_hwmon *menf21bmc_hwmon_update(struct device *dev)
+{
+	int i;
+	int val;
+	struct menf21bmc_hwmon *drv_data = dev_get_drvdata(dev);
+	struct menf21bmc_hwmon *data_ret = drv_data;
+
+	if (time_after(jiffies, drv_data->last_update + HZ)
+	    || !drv_data->valid) {
+		for (i = 0; i < BMC_VOLT_COUNT; i++) {
+			val = i2c_smbus_read_word_data(drv_data->i2c_client,
+						       IDX_TO_VOLT_INP_CMD(i));
+			if (val < 0) {
+				data_ret = ERR_PTR(val);
+				goto abort;
+			}
+			drv_data->in_val[i] = val;
+		}
+		drv_data->last_update = jiffies;
+		drv_data->valid = true;
+	}
+abort:
+	return data_ret;
+}
+
+static int menf21bmc_hwmon_get_volt_limits(struct menf21bmc_hwmon *drv_data)
+{
+	int i, val;
+
+	for (i = 0; i < BMC_VOLT_COUNT; i++) {
+		val = i2c_smbus_read_word_data(drv_data->i2c_client,
+					       IDX_TO_VOLT_MIN_CMD(i));
+		if (val < 0)
+			return val;
+
+		drv_data->in_min[i] = val;
+
+		val = i2c_smbus_read_word_data(drv_data->i2c_client,
+					       IDX_TO_VOLT_MAX_CMD(i));
+		if (val < 0)
+			return val;
+
+		drv_data->in_max[i] = val;
+	}
+	return 0;
+}
+
+static ssize_t
+show_label(struct device *dev, struct device_attribute *devattr, char *buf)
+{
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+
+	return sprintf(buf, "%s\n", input_names[attr->index]);
+}
+
+static ssize_t
+show_in(struct device *dev, struct device_attribute *devattr, char *buf)
+{
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct menf21bmc_hwmon *drv_data = menf21bmc_hwmon_update(dev);
+
+	if (IS_ERR(drv_data))
+		return PTR_ERR(drv_data);
+
+	return sprintf(buf, "%d\n", drv_data->in_val[attr->index]);
+}
+
+static ssize_t
+show_min(struct device *dev, struct device_attribute *devattr, char *buf)
+{
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct menf21bmc_hwmon *drv_data = dev_get_drvdata(dev);
+
+	return sprintf(buf, "%d\n", drv_data->in_min[attr->index]);
+}
+
+static ssize_t
+show_max(struct device *dev, struct device_attribute *devattr, char *buf)
+{
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct menf21bmc_hwmon *drv_data = dev_get_drvdata(dev);
+
+	return sprintf(buf, "%d\n", drv_data->in_max[attr->index]);
+}
+
+#define create_voltage_sysfs(idx)			\
+static SENSOR_DEVICE_ATTR(in##idx##_input, S_IRUGO,	\
+			show_in, NULL, idx);		\
+static SENSOR_DEVICE_ATTR(in##idx##_min, S_IRUGO,	\
+			show_min, NULL, idx);		\
+static SENSOR_DEVICE_ATTR(in##idx##_max, S_IRUGO,	\
+			show_max, NULL, idx);		\
+static SENSOR_DEVICE_ATTR(in##idx##_label, S_IRUGO,	\
+			show_label, NULL, idx);
+
+create_voltage_sysfs(0);
+create_voltage_sysfs(1);
+create_voltage_sysfs(2);
+create_voltage_sysfs(3);
+create_voltage_sysfs(4);
+
+static struct attribute *menf21bmc_hwmon_attrs[] = {
+	&sensor_dev_attr_in0_input.dev_attr.attr,
+	&sensor_dev_attr_in0_min.dev_attr.attr,
+	&sensor_dev_attr_in0_max.dev_attr.attr,
+	&sensor_dev_attr_in0_label.dev_attr.attr,
+
+	&sensor_dev_attr_in1_input.dev_attr.attr,
+	&sensor_dev_attr_in1_min.dev_attr.attr,
+	&sensor_dev_attr_in1_max.dev_attr.attr,
+	&sensor_dev_attr_in1_label.dev_attr.attr,
+
+	&sensor_dev_attr_in2_input.dev_attr.attr,
+	&sensor_dev_attr_in2_min.dev_attr.attr,
+	&sensor_dev_attr_in2_max.dev_attr.attr,
+	&sensor_dev_attr_in2_label.dev_attr.attr,
+
+	&sensor_dev_attr_in3_input.dev_attr.attr,
+	&sensor_dev_attr_in3_min.dev_attr.attr,
+	&sensor_dev_attr_in3_max.dev_attr.attr,
+	&sensor_dev_attr_in3_label.dev_attr.attr,
+
+	&sensor_dev_attr_in4_input.dev_attr.attr,
+	&sensor_dev_attr_in4_min.dev_attr.attr,
+	&sensor_dev_attr_in4_max.dev_attr.attr,
+	&sensor_dev_attr_in4_label.dev_attr.attr,
+	NULL
+};
+
+ATTRIBUTE_GROUPS(menf21bmc_hwmon);
+
+static int menf21bmc_hwmon_probe(struct platform_device *pdev)
+{
+	int ret;
+	struct menf21bmc_hwmon *drv_data;
+	struct i2c_client *i2c_client = to_i2c_client(pdev->dev.parent);
+	struct device *hwmon_dev;
+
+	drv_data = devm_kzalloc(&pdev->dev, sizeof(struct menf21bmc_hwmon),
+				GFP_KERNEL);
+	if (!drv_data)
+		return -ENOMEM;
+
+	drv_data->i2c_client = i2c_client;
+
+	ret = menf21bmc_hwmon_get_volt_limits(drv_data);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to read sensor limits");
+		return ret;
+	}
+
+	hwmon_dev = devm_hwmon_device_register_with_groups(&pdev->dev,
+						   "menf21bmc", drv_data,
+						   menf21bmc_hwmon_groups);
+	if (IS_ERR(hwmon_dev))
+		return PTR_ERR(hwmon_dev);
+
+	dev_info(&pdev->dev, "MEN 14F021P00 BMC hwmon device enabled");
+
+	return 0;
+}
+
+static struct platform_driver menf21bmc_hwmon = {
+	.probe		= menf21bmc_hwmon_probe,
+	.driver		= {
+		.name		= DRV_NAME,
+	},
+};
+
+module_platform_driver(menf21bmc_hwmon);
+
+MODULE_AUTHOR("Andreas Werner <andreas.werner@men.de>");
+MODULE_DESCRIPTION("MEN 14F021P00 BMC hwmon");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS("platform:menf21bmc_hwmon");
diff --git a/drivers/hwmon/nct6683.c b/drivers/hwmon/nct6683.c
index 7710f4694ba1..f3830db02d46 100644
--- a/drivers/hwmon/nct6683.c
+++ b/drivers/hwmon/nct6683.c
@@ -1282,7 +1282,6 @@ static const struct dev_pm_ops nct6683_dev_pm_ops = {
 
 static struct platform_driver nct6683_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= DRVNAME,
 		.pm	= NCT6683_DEV_PM_OPS,
 	},
diff --git a/drivers/hwmon/nct6775.c b/drivers/hwmon/nct6775.c
index 504cbddbdd90..1be41177b620 100644
--- a/drivers/hwmon/nct6775.c
+++ b/drivers/hwmon/nct6775.c
@@ -38,6 +38,7 @@
  * nct6776f     9      5       3       6+3    0xc330 0xc1    0x5ca3
  * nct6779d    15      5       5       2+6    0xc560 0xc1    0x5ca3
  * nct6791d    15      6       6       2+6    0xc800 0xc1    0x5ca3
+ * nct6792d    15      6       6       2+6    0xc910 0xc1    0x5ca3
  *
  * #temp lists the number of monitored temperature sources (first value) plus
  * the number of directly connectable temperature sensors (second value).
@@ -61,7 +62,7 @@
 
 #define USE_ALTERNATE
 
-enum kinds { nct6106, nct6775, nct6776, nct6779, nct6791 };
+enum kinds { nct6106, nct6775, nct6776, nct6779, nct6791, nct6792 };
 
 /* used to set data->name = nct6775_device_names[data->sio_kind] */
 static const char * const nct6775_device_names[] = {
@@ -70,6 +71,7 @@ static const char * const nct6775_device_names[] = {
 	"nct6776",
 	"nct6779",
 	"nct6791",
+	"nct6792",
 };
 
 static unsigned short force_id;
@@ -100,6 +102,7 @@ MODULE_PARM_DESC(fan_debounce, "Enable debouncing for fan RPM signal");
 #define SIO_NCT6776_ID		0xc330
 #define SIO_NCT6779_ID		0xc560
 #define SIO_NCT6791_ID		0xc800
+#define SIO_NCT6792_ID		0xc910
 #define SIO_ID_MASK		0xFFF0
 
 enum pwm_enable { off, manual, thermal_cruise, speed_cruise, sf3, sf4 };
@@ -529,6 +532,12 @@ static const s8 NCT6791_ALARM_BITS[] = {
 	4, 5, 13, -1, -1, -1,		/* temp1..temp6 */
 	12, 9 };			/* intrusion0, intrusion1 */
 
+/* NCT6792 specific data */
+
+static const u16 NCT6792_REG_TEMP_MON[] = {
+	0x73, 0x75, 0x77, 0x79, 0x7b, 0x7d };
+static const u16 NCT6792_REG_BEEP[NUM_REG_BEEP] = {
+	0xb2, 0xb3, 0xb4, 0xb5, 0xbf };
 
 /* NCT6102D/NCT6106D specific data */
 
@@ -1043,13 +1052,14 @@ static bool is_word_sized(struct nct6775_data *data, u16 reg)
 		  reg == 0x73 || reg == 0x75 || reg == 0x77;
 	case nct6779:
 	case nct6791:
+	case nct6792:
 		return reg == 0x150 || reg == 0x153 || reg == 0x155 ||
 		  ((reg & 0xfff0) == 0x4b0 && (reg & 0x000f) < 0x0b) ||
 		  reg == 0x402 ||
 		  reg == 0x63a || reg == 0x63c || reg == 0x63e ||
 		  reg == 0x640 || reg == 0x642 ||
 		  reg == 0x73 || reg == 0x75 || reg == 0x77 || reg == 0x79 ||
-		  reg == 0x7b;
+		  reg == 0x7b || reg == 0x7d;
 	}
 	return false;
 }
@@ -1063,6 +1073,7 @@ static bool is_word_sized(struct nct6775_data *data, u16 reg)
 static inline void nct6775_set_bank(struct nct6775_data *data, u16 reg)
 {
 	u8 bank = reg >> 8;
+
 	if (data->bank != bank) {
 		outb_p(NCT6775_REG_BANK, data->addr + ADDR_REG_OFFSET);
 		outb_p(bank, data->addr + DATA_REG_OFFSET);
@@ -1300,6 +1311,7 @@ static void nct6775_update_pwm(struct device *dev)
 		if (!data->target_speed_tolerance[i] ||
 		    data->pwm_enable[i] == speed_cruise) {
 			u8 t = fanmodecfg & 0x0f;
+
 			if (data->REG_TOLERANCE_H) {
 				t |= (nct6775_read_value(data,
 				      data->REG_TOLERANCE_H[i]) & 0x70) >> 1;
@@ -1391,6 +1403,7 @@ static void nct6775_update_pwm_limits(struct device *dev)
 		case nct6106:
 		case nct6779:
 		case nct6791:
+		case nct6792:
 			reg = nct6775_read_value(data,
 					data->REG_CRITICAL_PWM_ENABLE[i]);
 			if (reg & data->CRITICAL_PWM_ENABLE_MASK)
@@ -1473,6 +1486,7 @@ static struct nct6775_data *nct6775_update_device(struct device *dev)
 		data->alarms = 0;
 		for (i = 0; i < NUM_REG_ALARM; i++) {
 			u8 alarm;
+
 			if (!data->REG_ALARM[i])
 				continue;
 			alarm = nct6775_read_value(data, data->REG_ALARM[i]);
@@ -1482,6 +1496,7 @@ static struct nct6775_data *nct6775_update_device(struct device *dev)
 		data->beeps = 0;
 		for (i = 0; i < NUM_REG_BEEP; i++) {
 			u8 beep;
+
 			if (!data->REG_BEEP[i])
 				continue;
 			beep = nct6775_read_value(data, data->REG_BEEP[i]);
@@ -1504,8 +1519,9 @@ show_in_reg(struct device *dev, struct device_attribute *attr, char *buf)
 {
 	struct nct6775_data *data = nct6775_update_device(dev);
 	struct sensor_device_attribute_2 *sattr = to_sensor_dev_attr_2(attr);
-	int nr = sattr->nr;
 	int index = sattr->index;
+	int nr = sattr->nr;
+
 	return sprintf(buf, "%ld\n", in_from_reg(data->in[nr][index], nr));
 }
 
@@ -1515,10 +1531,12 @@ store_in_reg(struct device *dev, struct device_attribute *attr, const char *buf,
 {
 	struct nct6775_data *data = dev_get_drvdata(dev);
 	struct sensor_device_attribute_2 *sattr = to_sensor_dev_attr_2(attr);
-	int nr = sattr->nr;
 	int index = sattr->index;
+	int nr = sattr->nr;
 	unsigned long val;
-	int err = kstrtoul(buf, 10, &val);
+	int err;
+
+	err = kstrtoul(buf, 10, &val);
 	if (err < 0)
 		return err;
 	mutex_lock(&data->update_lock);
@@ -1535,6 +1553,7 @@ show_alarm(struct device *dev, struct device_attribute *attr, char *buf)
 	struct nct6775_data *data = nct6775_update_device(dev);
 	struct sensor_device_attribute *sattr = to_sensor_dev_attr(attr);
 	int nr = data->ALARM_BITS[sattr->index];
+
 	return sprintf(buf, "%u\n",
 		       (unsigned int)((data->alarms >> nr) & 0x01));
 }
@@ -1570,6 +1589,7 @@ show_temp_alarm(struct device *dev, struct device_attribute *attr, char *buf)
 	nr = find_temp_source(data, sattr->index, data->num_temp_alarms);
 	if (nr >= 0) {
 		int bit = data->ALARM_BITS[nr + TEMP_ALARM_BASE];
+
 		alarm = (data->alarms >> bit) & 0x01;
 	}
 	return sprintf(buf, "%u\n", alarm);
@@ -1595,8 +1615,9 @@ store_beep(struct device *dev, struct device_attribute *attr, const char *buf,
 	int nr = data->BEEP_BITS[sattr->index];
 	int regindex = nr >> 3;
 	unsigned long val;
+	int err;
 
-	int err = kstrtoul(buf, 10, &val);
+	err = kstrtoul(buf, 10, &val);
 	if (err < 0)
 		return err;
 	if (val > 1)
@@ -1629,6 +1650,7 @@ show_temp_beep(struct device *dev, struct device_attribute *attr, char *buf)
 	nr = find_temp_source(data, sattr->index, data->num_temp_beeps);
 	if (nr >= 0) {
 		int bit = data->BEEP_BITS[nr + TEMP_ALARM_BASE];
+
 		beep = (data->beeps >> bit) & 0x01;
 	}
 	return sprintf(buf, "%u\n", beep);
@@ -1642,8 +1664,9 @@ store_temp_beep(struct device *dev, struct device_attribute *attr,
 	struct nct6775_data *data = dev_get_drvdata(dev);
 	int nr, bit, regindex;
 	unsigned long val;
+	int err;
 
-	int err = kstrtoul(buf, 10, &val);
+	err = kstrtoul(buf, 10, &val);
 	if (err < 0)
 		return err;
 	if (val > 1)
@@ -1715,6 +1738,7 @@ show_fan(struct device *dev, struct device_attribute *attr, char *buf)
 	struct nct6775_data *data = nct6775_update_device(dev);
 	struct sensor_device_attribute *sattr = to_sensor_dev_attr(attr);
 	int nr = sattr->index;
+
 	return sprintf(buf, "%d\n", data->rpm[nr]);
 }
 
@@ -1724,6 +1748,7 @@ show_fan_min(struct device *dev, struct device_attribute *attr, char *buf)
 	struct nct6775_data *data = nct6775_update_device(dev);
 	struct sensor_device_attribute *sattr = to_sensor_dev_attr(attr);
 	int nr = sattr->index;
+
 	return sprintf(buf, "%d\n",
 		       data->fan_from_reg_min(data->fan_min[nr],
 					      data->fan_div[nr]));
@@ -1735,6 +1760,7 @@ show_fan_div(struct device *dev, struct device_attribute *attr, char *buf)
 	struct nct6775_data *data = nct6775_update_device(dev);
 	struct sensor_device_attribute *sattr = to_sensor_dev_attr(attr);
 	int nr = sattr->index;
+
 	return sprintf(buf, "%u\n", div_from_reg(data->fan_div[nr]));
 }
 
@@ -1746,9 +1772,9 @@ store_fan_min(struct device *dev, struct device_attribute *attr,
 	struct sensor_device_attribute *sattr = to_sensor_dev_attr(attr);
 	int nr = sattr->index;
 	unsigned long val;
-	int err;
 	unsigned int reg;
 	u8 new_div;
+	int err;
 
 	err = kstrtoul(buf, 10, &val);
 	if (err < 0)
@@ -1932,6 +1958,7 @@ show_temp_label(struct device *dev, struct device_attribute *attr, char *buf)
 	struct nct6775_data *data = nct6775_update_device(dev);
 	struct sensor_device_attribute *sattr = to_sensor_dev_attr(attr);
 	int nr = sattr->index;
+
 	return sprintf(buf, "%s\n", data->temp_label[data->temp_src[nr]]);
 }
 
@@ -2008,6 +2035,7 @@ show_temp_type(struct device *dev, struct device_attribute *attr, char *buf)
 	struct nct6775_data *data = nct6775_update_device(dev);
 	struct sensor_device_attribute *sattr = to_sensor_dev_attr(attr);
 	int nr = sattr->index;
+
 	return sprintf(buf, "%d\n", (int)data->temp_type[nr]);
 }
 
@@ -2790,6 +2818,7 @@ store_auto_pwm(struct device *dev, struct device_attribute *attr,
 		case nct6106:
 		case nct6779:
 		case nct6791:
+		case nct6792:
 			nct6775_write_value(data, data->REG_CRITICAL_PWM[nr],
 					    val);
 			reg = nct6775_read_value(data,
@@ -2997,6 +3026,7 @@ static ssize_t
 show_vid(struct device *dev, struct device_attribute *attr, char *buf)
 {
 	struct nct6775_data *data = dev_get_drvdata(dev);
+
 	return sprintf(buf, "%d\n", vid_from_reg(data->vid, data->vrm));
 }
 
@@ -3202,7 +3232,7 @@ nct6775_check_fan_inputs(struct nct6775_data *data)
 		pwm4pin = false;
 		pwm5pin = false;
 		pwm6pin = false;
-	} else {	/* NCT6779D or NCT6791D */
+	} else {	/* NCT6779D, NCT6791D, or NCT6792D */
 		regval = superio_inb(sioreg, 0x1c);
 
 		fan3pin = !(regval & (1 << 5));
@@ -3215,7 +3245,7 @@ nct6775_check_fan_inputs(struct nct6775_data *data)
 
 		fan4min = fan4pin;
 
-		if (data->kind == nct6791) {
+		if (data->kind == nct6791 || data->kind == nct6792) {
 			regval = superio_inb(sioreg, 0x2d);
 			fan6pin = (regval & (1 << 1));
 			pwm6pin = (regval & (1 << 0));
@@ -3588,6 +3618,7 @@ static int nct6775_probe(struct platform_device *pdev)
 
 		break;
 	case nct6791:
+	case nct6792:
 		data->in_num = 15;
 		data->pwm_num = 6;
 		data->auto_pwm_num = 4;
@@ -3650,12 +3681,20 @@ static int nct6775_probe(struct platform_device *pdev)
 		data->REG_WEIGHT_TEMP[1] = NCT6791_REG_WEIGHT_TEMP_STEP_TOL;
 		data->REG_WEIGHT_TEMP[2] = NCT6791_REG_WEIGHT_TEMP_BASE;
 		data->REG_ALARM = NCT6791_REG_ALARM;
-		data->REG_BEEP = NCT6776_REG_BEEP;
+		if (data->kind == nct6791)
+			data->REG_BEEP = NCT6776_REG_BEEP;
+		else
+			data->REG_BEEP = NCT6792_REG_BEEP;
 
 		reg_temp = NCT6779_REG_TEMP;
-		reg_temp_mon = NCT6779_REG_TEMP_MON;
 		num_reg_temp = ARRAY_SIZE(NCT6779_REG_TEMP);
-		num_reg_temp_mon = ARRAY_SIZE(NCT6779_REG_TEMP_MON);
+		if (data->kind == nct6791) {
+			reg_temp_mon = NCT6779_REG_TEMP_MON;
+			num_reg_temp_mon = ARRAY_SIZE(NCT6779_REG_TEMP_MON);
+		} else {
+			reg_temp_mon = NCT6792_REG_TEMP_MON;
+			num_reg_temp_mon = ARRAY_SIZE(NCT6792_REG_TEMP_MON);
+		}
 		reg_temp_over = NCT6779_REG_TEMP_OVER;
 		reg_temp_hyst = NCT6779_REG_TEMP_HYST;
 		reg_temp_config = NCT6779_REG_TEMP_CONFIG;
@@ -3854,6 +3893,7 @@ static int nct6775_probe(struct platform_device *pdev)
 	case nct6106:
 	case nct6779:
 	case nct6791:
+	case nct6792:
 		break;
 	}
 
@@ -3885,6 +3925,7 @@ static int nct6775_probe(struct platform_device *pdev)
 			tmp |= 0x3e;
 			break;
 		case nct6791:
+		case nct6792:
 			tmp |= 0x7e;
 			break;
 		}
@@ -3972,7 +4013,7 @@ static int nct6775_resume(struct device *dev)
 	mutex_lock(&data->update_lock);
 	data->bank = 0xff;		/* Force initial bank selection */
 
-	if (data->kind == nct6791) {
+	if (data->kind == nct6791 || data->kind == nct6792) {
 		err = superio_enter(data->sioreg);
 		if (err)
 			goto abort;
@@ -4039,7 +4080,6 @@ static const struct dev_pm_ops nct6775_dev_pm_ops = {
 
 static struct platform_driver nct6775_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= DRVNAME,
 		.pm	= NCT6775_DEV_PM_OPS,
 	},
@@ -4052,6 +4092,7 @@ static const char * const nct6775_sio_names[] __initconst = {
 	"NCT6776D/F",
 	"NCT6779D",
 	"NCT6791D",
+	"NCT6792D",
 };
 
 /* nct6775_find() looks for a '627 in the Super-I/O config space */
@@ -4086,6 +4127,9 @@ static int __init nct6775_find(int sioaddr, struct nct6775_sio_data *sio_data)
 	case SIO_NCT6791_ID:
 		sio_data->kind = nct6791;
 		break;
+	case SIO_NCT6792_ID:
+		sio_data->kind = nct6792;
+		break;
 	default:
 		if (val != 0xffff)
 			pr_debug("unsupported chip ID: 0x%04x\n", val);
@@ -4111,7 +4155,7 @@ static int __init nct6775_find(int sioaddr, struct nct6775_sio_data *sio_data)
 		superio_outb(sioaddr, SIO_REG_ENABLE, val | 0x01);
 	}
 
-	if (sio_data->kind == nct6791)
+	if (sio_data->kind == nct6791 || sio_data->kind == nct6792)
 		nct6791_enable_io_mapping(sioaddr);
 
 	superio_exit(sioaddr);
@@ -4221,7 +4265,7 @@ static void __exit sensors_nct6775_exit(void)
 }
 
 MODULE_AUTHOR("Guenter Roeck <linux@roeck-us.net>");
-MODULE_DESCRIPTION("NCT6106D/NCT6775F/NCT6776F/NCT6779D/NCT6791D driver");
+MODULE_DESCRIPTION("NCT6106D/NCT6775F/NCT6776F/NCT6779D/NCT6791D/NCT6792D driver");
 MODULE_LICENSE("GPL");
 
 module_init(sensors_nct6775_init);
diff --git a/drivers/hwmon/nct7802.c b/drivers/hwmon/nct7802.c
new file mode 100644
index 000000000000..ec5678289e4a
--- /dev/null
+++ b/drivers/hwmon/nct7802.c
@@ -0,0 +1,860 @@
+/*
+ * nct7802 - Driver for Nuvoton NCT7802Y
+ *
+ * Copyright (C) 2014  Guenter Roeck <linux@roeck-us.net>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/err.h>
+#include <linux/i2c.h>
+#include <linux/init.h>
+#include <linux/hwmon.h>
+#include <linux/hwmon-sysfs.h>
+#include <linux/jiffies.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/regmap.h>
+#include <linux/slab.h>
+
+#define DRVNAME "nct7802"
+
+static const u8 REG_VOLTAGE[5] = { 0x09, 0x0a, 0x0c, 0x0d, 0x0e };
+
+static const u8 REG_VOLTAGE_LIMIT_LSB[2][5] = {
+	{ 0x40, 0x00, 0x42, 0x44, 0x46 },
+	{ 0x3f, 0x00, 0x41, 0x43, 0x45 },
+};
+
+static const u8 REG_VOLTAGE_LIMIT_MSB[5] = { 0x48, 0x00, 0x47, 0x47, 0x48 };
+
+static const u8 REG_VOLTAGE_LIMIT_MSB_SHIFT[2][5] = {
+	{ 0, 0, 4, 0, 4 },
+	{ 2, 0, 6, 2, 6 },
+};
+
+#define REG_BANK		0x00
+#define REG_TEMP_LSB		0x05
+#define REG_TEMP_PECI_LSB	0x08
+#define REG_VOLTAGE_LOW		0x0f
+#define REG_FANCOUNT_LOW	0x13
+#define REG_START		0x21
+#define REG_MODE		0x22
+#define REG_PECI_ENABLE		0x23
+#define REG_FAN_ENABLE		0x24
+#define REG_VMON_ENABLE		0x25
+#define REG_VENDOR_ID		0xfd
+#define REG_CHIP_ID		0xfe
+#define REG_VERSION_ID		0xff
+
+/*
+ * Data structures and manipulation thereof
+ */
+
+struct nct7802_data {
+	struct regmap *regmap;
+	struct mutex access_lock; /* for multi-byte read and write operations */
+};
+
+static int nct7802_read_temp(struct nct7802_data *data,
+			     u8 reg_temp, u8 reg_temp_low, int *temp)
+{
+	unsigned int t1, t2 = 0;
+	int err;
+
+	*temp = 0;
+
+	mutex_lock(&data->access_lock);
+	err = regmap_read(data->regmap, reg_temp, &t1);
+	if (err < 0)
+		goto abort;
+	t1 <<= 8;
+	if (reg_temp_low) {	/* 11 bit data */
+		err = regmap_read(data->regmap, reg_temp_low, &t2);
+		if (err < 0)
+			goto abort;
+	}
+	t1 |= t2 & 0xe0;
+	*temp = (s16)t1 / 32 * 125;
+abort:
+	mutex_unlock(&data->access_lock);
+	return err;
+}
+
+static int nct7802_read_fan(struct nct7802_data *data, u8 reg_fan)
+{
+	unsigned int f1, f2;
+	int ret;
+
+	mutex_lock(&data->access_lock);
+	ret = regmap_read(data->regmap, reg_fan, &f1);
+	if (ret < 0)
+		goto abort;
+	ret = regmap_read(data->regmap, REG_FANCOUNT_LOW, &f2);
+	if (ret < 0)
+		goto abort;
+	ret = (f1 << 5) | (f2 >> 3);
+	/* convert fan count to rpm */
+	if (ret == 0x1fff)	/* maximum value, assume fan is stopped */
+		ret = 0;
+	else if (ret)
+		ret = DIV_ROUND_CLOSEST(1350000U, ret);
+abort:
+	mutex_unlock(&data->access_lock);
+	return ret;
+}
+
+static int nct7802_read_fan_min(struct nct7802_data *data, u8 reg_fan_low,
+				u8 reg_fan_high)
+{
+	unsigned int f1, f2;
+	int ret;
+
+	mutex_lock(&data->access_lock);
+	ret = regmap_read(data->regmap, reg_fan_low, &f1);
+	if (ret < 0)
+		goto abort;
+	ret = regmap_read(data->regmap, reg_fan_high, &f2);
+	if (ret < 0)
+		goto abort;
+	ret = f1 | ((f2 & 0xf8) << 5);
+	/* convert fan count to rpm */
+	if (ret == 0x1fff)	/* maximum value, assume no limit */
+		ret = 0;
+	else if (ret)
+		ret = DIV_ROUND_CLOSEST(1350000U, ret);
+abort:
+	mutex_unlock(&data->access_lock);
+	return ret;
+}
+
+static int nct7802_write_fan_min(struct nct7802_data *data, u8 reg_fan_low,
+				 u8 reg_fan_high, unsigned int limit)
+{
+	int err;
+
+	if (limit)
+		limit = DIV_ROUND_CLOSEST(1350000U, limit);
+	else
+		limit = 0x1fff;
+	limit = clamp_val(limit, 0, 0x1fff);
+
+	mutex_lock(&data->access_lock);
+	err = regmap_write(data->regmap, reg_fan_low, limit & 0xff);
+	if (err < 0)
+		goto abort;
+
+	err = regmap_write(data->regmap, reg_fan_high, (limit & 0x1f00) >> 5);
+abort:
+	mutex_unlock(&data->access_lock);
+	return err;
+}
+
+static u8 nct7802_vmul[] = { 4, 2, 2, 2, 2 };
+
+static int nct7802_read_voltage(struct nct7802_data *data, int nr, int index)
+{
+	unsigned int v1, v2;
+	int ret;
+
+	mutex_lock(&data->access_lock);
+	if (index == 0) {	/* voltage */
+		ret = regmap_read(data->regmap, REG_VOLTAGE[nr], &v1);
+		if (ret < 0)
+			goto abort;
+		ret = regmap_read(data->regmap, REG_VOLTAGE_LOW, &v2);
+		if (ret < 0)
+			goto abort;
+		ret = ((v1 << 2) | (v2 >> 6)) * nct7802_vmul[nr];
+	}  else {	/* limit */
+		int shift = 8 - REG_VOLTAGE_LIMIT_MSB_SHIFT[index - 1][nr];
+
+		ret = regmap_read(data->regmap,
+				  REG_VOLTAGE_LIMIT_LSB[index - 1][nr], &v1);
+		if (ret < 0)
+			goto abort;
+		ret = regmap_read(data->regmap, REG_VOLTAGE_LIMIT_MSB[nr],
+				  &v2);
+		if (ret < 0)
+			goto abort;
+		ret = (v1 | ((v2 << shift) & 0x300)) * nct7802_vmul[nr];
+	}
+abort:
+	mutex_unlock(&data->access_lock);
+	return ret;
+}
+
+static int nct7802_write_voltage(struct nct7802_data *data, int nr, int index,
+				 unsigned int voltage)
+{
+	int shift = 8 - REG_VOLTAGE_LIMIT_MSB_SHIFT[index - 1][nr];
+	int err;
+
+	voltage = DIV_ROUND_CLOSEST(voltage, nct7802_vmul[nr]);
+	voltage = clamp_val(voltage, 0, 0x3ff);
+
+	mutex_lock(&data->access_lock);
+	err = regmap_write(data->regmap,
+			   REG_VOLTAGE_LIMIT_LSB[index - 1][nr],
+			   voltage & 0xff);
+	if (err < 0)
+		goto abort;
+
+	err = regmap_update_bits(data->regmap, REG_VOLTAGE_LIMIT_MSB[nr],
+				 0x0300 >> shift, (voltage & 0x0300) >> shift);
+abort:
+	mutex_unlock(&data->access_lock);
+	return err;
+}
+
+static ssize_t show_in(struct device *dev, struct device_attribute *attr,
+		       char *buf)
+{
+	struct sensor_device_attribute_2 *sattr = to_sensor_dev_attr_2(attr);
+	struct nct7802_data *data = dev_get_drvdata(dev);
+	int voltage;
+
+	voltage = nct7802_read_voltage(data, sattr->nr, sattr->index);
+	if (voltage < 0)
+		return voltage;
+
+	return sprintf(buf, "%d\n", voltage);
+}
+
+static ssize_t store_in(struct device *dev, struct device_attribute *attr,
+			const char *buf, size_t count)
+{
+	struct sensor_device_attribute_2 *sattr = to_sensor_dev_attr_2(attr);
+	struct nct7802_data *data = dev_get_drvdata(dev);
+	int index = sattr->index;
+	int nr = sattr->nr;
+	unsigned long val;
+	int err;
+
+	err = kstrtoul(buf, 10, &val);
+	if (err < 0)
+		return err;
+
+	err = nct7802_write_voltage(data, nr, index, val);
+	return err ? : count;
+}
+
+static ssize_t show_temp(struct device *dev, struct device_attribute *attr,
+			 char *buf)
+{
+	struct nct7802_data *data = dev_get_drvdata(dev);
+	struct sensor_device_attribute_2 *sattr = to_sensor_dev_attr_2(attr);
+	int err, temp;
+
+	err = nct7802_read_temp(data, sattr->nr, sattr->index, &temp);
+	if (err < 0)
+		return err;
+
+	return sprintf(buf, "%d\n", temp);
+}
+
+static ssize_t store_temp(struct device *dev, struct device_attribute *attr,
+			  const char *buf, size_t count)
+{
+	struct sensor_device_attribute_2 *sattr = to_sensor_dev_attr_2(attr);
+	struct nct7802_data *data = dev_get_drvdata(dev);
+	int nr = sattr->nr;
+	long val;
+	int err;
+
+	err = kstrtol(buf, 10, &val);
+	if (err < 0)
+		return err;
+
+	val = clamp_val(DIV_ROUND_CLOSEST(val, 1000), -128, 127);
+
+	err = regmap_write(data->regmap, nr, val & 0xff);
+	return err ? : count;
+}
+
+static ssize_t show_fan(struct device *dev, struct device_attribute *attr,
+			char *buf)
+{
+	struct sensor_device_attribute *sattr = to_sensor_dev_attr(attr);
+	struct nct7802_data *data = dev_get_drvdata(dev);
+	int speed;
+
+	speed = nct7802_read_fan(data, sattr->index);
+	if (speed < 0)
+		return speed;
+
+	return sprintf(buf, "%d\n", speed);
+}
+
+static ssize_t show_fan_min(struct device *dev, struct device_attribute *attr,
+			    char *buf)
+{
+	struct sensor_device_attribute_2 *sattr = to_sensor_dev_attr_2(attr);
+	struct nct7802_data *data = dev_get_drvdata(dev);
+	int speed;
+
+	speed = nct7802_read_fan_min(data, sattr->nr, sattr->index);
+	if (speed < 0)
+		return speed;
+
+	return sprintf(buf, "%d\n", speed);
+}
+
+static ssize_t store_fan_min(struct device *dev, struct device_attribute *attr,
+			     const char *buf, size_t count)
+{
+	struct sensor_device_attribute_2 *sattr = to_sensor_dev_attr_2(attr);
+	struct nct7802_data *data = dev_get_drvdata(dev);
+	unsigned long val;
+	int err;
+
+	err = kstrtoul(buf, 10, &val);
+	if (err < 0)
+		return err;
+
+	err = nct7802_write_fan_min(data, sattr->nr, sattr->index, val);
+	return err ? : count;
+}
+
+static ssize_t show_alarm(struct device *dev, struct device_attribute *attr,
+			  char *buf)
+{
+	struct nct7802_data *data = dev_get_drvdata(dev);
+	struct sensor_device_attribute_2 *sattr = to_sensor_dev_attr_2(attr);
+	int bit = sattr->index;
+	unsigned int val;
+	int ret;
+
+	ret = regmap_read(data->regmap, sattr->nr, &val);
+	if (ret < 0)
+		return ret;
+
+	return sprintf(buf, "%u\n", !!(val & (1 << bit)));
+}
+
+static ssize_t
+show_beep(struct device *dev, struct device_attribute *attr, char *buf)
+{
+	struct sensor_device_attribute_2 *sattr = to_sensor_dev_attr_2(attr);
+	struct nct7802_data *data = dev_get_drvdata(dev);
+	unsigned int regval;
+	int err;
+
+	err = regmap_read(data->regmap, sattr->nr, &regval);
+	if (err)
+		return err;
+
+	return sprintf(buf, "%u\n", !!(regval & (1 << sattr->index)));
+}
+
+static ssize_t
+store_beep(struct device *dev, struct device_attribute *attr, const char *buf,
+	   size_t count)
+{
+	struct sensor_device_attribute_2 *sattr = to_sensor_dev_attr_2(attr);
+	struct nct7802_data *data = dev_get_drvdata(dev);
+	unsigned long val;
+	int err;
+
+	err = kstrtoul(buf, 10, &val);
+	if (err < 0)
+		return err;
+	if (val > 1)
+		return -EINVAL;
+
+	err = regmap_update_bits(data->regmap, sattr->nr, 1 << sattr->index,
+				 val ? 1 << sattr->index : 0);
+	return err ? : count;
+}
+
+static SENSOR_DEVICE_ATTR_2(temp1_input, S_IRUGO, show_temp, NULL, 0x01,
+			    REG_TEMP_LSB);
+static SENSOR_DEVICE_ATTR_2(temp1_min, S_IRUGO | S_IWUSR, show_temp,
+			    store_temp, 0x31, 0);
+static SENSOR_DEVICE_ATTR_2(temp1_max, S_IRUGO | S_IWUSR, show_temp,
+			    store_temp, 0x30, 0);
+static SENSOR_DEVICE_ATTR_2(temp1_crit, S_IRUGO | S_IWUSR, show_temp,
+			    store_temp, 0x3a, 0);
+
+static SENSOR_DEVICE_ATTR_2(temp2_input, S_IRUGO, show_temp, NULL, 0x02,
+			    REG_TEMP_LSB);
+static SENSOR_DEVICE_ATTR_2(temp2_min, S_IRUGO | S_IWUSR, show_temp,
+			    store_temp, 0x33, 0);
+static SENSOR_DEVICE_ATTR_2(temp2_max, S_IRUGO | S_IWUSR, show_temp,
+			    store_temp, 0x32, 0);
+static SENSOR_DEVICE_ATTR_2(temp2_crit, S_IRUGO | S_IWUSR, show_temp,
+			    store_temp, 0x3b, 0);
+
+static SENSOR_DEVICE_ATTR_2(temp3_input, S_IRUGO, show_temp, NULL, 0x03,
+			    REG_TEMP_LSB);
+static SENSOR_DEVICE_ATTR_2(temp3_min, S_IRUGO | S_IWUSR, show_temp,
+			    store_temp, 0x35, 0);
+static SENSOR_DEVICE_ATTR_2(temp3_max, S_IRUGO | S_IWUSR, show_temp,
+			    store_temp, 0x34, 0);
+static SENSOR_DEVICE_ATTR_2(temp3_crit, S_IRUGO | S_IWUSR, show_temp,
+			    store_temp, 0x3c, 0);
+
+static SENSOR_DEVICE_ATTR_2(temp4_input, S_IRUGO, show_temp, NULL, 0x04, 0);
+static SENSOR_DEVICE_ATTR_2(temp4_min, S_IRUGO | S_IWUSR, show_temp,
+			    store_temp, 0x37, 0);
+static SENSOR_DEVICE_ATTR_2(temp4_max, S_IRUGO | S_IWUSR, show_temp,
+			    store_temp, 0x36, 0);
+static SENSOR_DEVICE_ATTR_2(temp4_crit, S_IRUGO | S_IWUSR, show_temp,
+			    store_temp, 0x3d, 0);
+
+static SENSOR_DEVICE_ATTR_2(temp5_input, S_IRUGO, show_temp, NULL, 0x06,
+			    REG_TEMP_PECI_LSB);
+static SENSOR_DEVICE_ATTR_2(temp5_min, S_IRUGO | S_IWUSR, show_temp,
+			    store_temp, 0x39, 0);
+static SENSOR_DEVICE_ATTR_2(temp5_max, S_IRUGO | S_IWUSR, show_temp,
+			    store_temp, 0x38, 0);
+static SENSOR_DEVICE_ATTR_2(temp5_crit, S_IRUGO | S_IWUSR, show_temp,
+			    store_temp, 0x3e, 0);
+
+static SENSOR_DEVICE_ATTR_2(temp6_input, S_IRUGO, show_temp, NULL, 0x07,
+			    REG_TEMP_PECI_LSB);
+
+static SENSOR_DEVICE_ATTR_2(temp1_min_alarm, S_IRUGO, show_alarm, NULL,
+			    0x18, 0);
+static SENSOR_DEVICE_ATTR_2(temp2_min_alarm, S_IRUGO, show_alarm, NULL,
+			    0x18, 1);
+static SENSOR_DEVICE_ATTR_2(temp3_min_alarm, S_IRUGO, show_alarm, NULL,
+			    0x18, 2);
+static SENSOR_DEVICE_ATTR_2(temp4_min_alarm, S_IRUGO, show_alarm, NULL,
+			    0x18, 3);
+static SENSOR_DEVICE_ATTR_2(temp5_min_alarm, S_IRUGO, show_alarm, NULL,
+			    0x18, 4);
+
+static SENSOR_DEVICE_ATTR_2(temp1_max_alarm, S_IRUGO, show_alarm, NULL,
+			    0x19, 0);
+static SENSOR_DEVICE_ATTR_2(temp2_max_alarm, S_IRUGO, show_alarm, NULL,
+			    0x19, 1);
+static SENSOR_DEVICE_ATTR_2(temp3_max_alarm, S_IRUGO, show_alarm, NULL,
+			    0x19, 2);
+static SENSOR_DEVICE_ATTR_2(temp4_max_alarm, S_IRUGO, show_alarm, NULL,
+			    0x19, 3);
+static SENSOR_DEVICE_ATTR_2(temp5_max_alarm, S_IRUGO, show_alarm, NULL,
+			    0x19, 4);
+
+static SENSOR_DEVICE_ATTR_2(temp1_crit_alarm, S_IRUGO, show_alarm, NULL,
+			    0x1b, 0);
+static SENSOR_DEVICE_ATTR_2(temp2_crit_alarm, S_IRUGO, show_alarm, NULL,
+			    0x1b, 1);
+static SENSOR_DEVICE_ATTR_2(temp3_crit_alarm, S_IRUGO, show_alarm, NULL,
+			    0x1b, 2);
+static SENSOR_DEVICE_ATTR_2(temp4_crit_alarm, S_IRUGO, show_alarm, NULL,
+			    0x1b, 3);
+static SENSOR_DEVICE_ATTR_2(temp5_crit_alarm, S_IRUGO, show_alarm, NULL,
+			    0x1b, 4);
+
+static SENSOR_DEVICE_ATTR_2(temp1_fault, S_IRUGO, show_alarm, NULL, 0x17, 0);
+static SENSOR_DEVICE_ATTR_2(temp2_fault, S_IRUGO, show_alarm, NULL, 0x17, 1);
+static SENSOR_DEVICE_ATTR_2(temp3_fault, S_IRUGO, show_alarm, NULL, 0x17, 2);
+
+static SENSOR_DEVICE_ATTR_2(temp1_beep, S_IRUGO | S_IWUSR, show_beep,
+			    store_beep, 0x5c, 0);
+static SENSOR_DEVICE_ATTR_2(temp2_beep, S_IRUGO | S_IWUSR, show_beep,
+			    store_beep, 0x5c, 1);
+static SENSOR_DEVICE_ATTR_2(temp3_beep, S_IRUGO | S_IWUSR, show_beep,
+			    store_beep, 0x5c, 2);
+static SENSOR_DEVICE_ATTR_2(temp4_beep, S_IRUGO | S_IWUSR, show_beep,
+			    store_beep, 0x5c, 3);
+static SENSOR_DEVICE_ATTR_2(temp5_beep, S_IRUGO | S_IWUSR, show_beep,
+			    store_beep, 0x5c, 4);
+static SENSOR_DEVICE_ATTR_2(temp6_beep, S_IRUGO | S_IWUSR, show_beep,
+			    store_beep, 0x5c, 5);
+
+static struct attribute *nct7802_temp_attrs[] = {
+	&sensor_dev_attr_temp1_input.dev_attr.attr,
+	&sensor_dev_attr_temp1_min.dev_attr.attr,
+	&sensor_dev_attr_temp1_max.dev_attr.attr,
+	&sensor_dev_attr_temp1_crit.dev_attr.attr,
+	&sensor_dev_attr_temp1_min_alarm.dev_attr.attr,
+	&sensor_dev_attr_temp1_max_alarm.dev_attr.attr,
+	&sensor_dev_attr_temp1_crit_alarm.dev_attr.attr,
+	&sensor_dev_attr_temp1_fault.dev_attr.attr,
+	&sensor_dev_attr_temp1_beep.dev_attr.attr,
+
+	&sensor_dev_attr_temp2_input.dev_attr.attr,		/* 9 */
+	&sensor_dev_attr_temp2_min.dev_attr.attr,
+	&sensor_dev_attr_temp2_max.dev_attr.attr,
+	&sensor_dev_attr_temp2_crit.dev_attr.attr,
+	&sensor_dev_attr_temp2_min_alarm.dev_attr.attr,
+	&sensor_dev_attr_temp2_max_alarm.dev_attr.attr,
+	&sensor_dev_attr_temp2_crit_alarm.dev_attr.attr,
+	&sensor_dev_attr_temp2_fault.dev_attr.attr,
+	&sensor_dev_attr_temp2_beep.dev_attr.attr,
+
+	&sensor_dev_attr_temp3_input.dev_attr.attr,		/* 18 */
+	&sensor_dev_attr_temp3_min.dev_attr.attr,
+	&sensor_dev_attr_temp3_max.dev_attr.attr,
+	&sensor_dev_attr_temp3_crit.dev_attr.attr,
+	&sensor_dev_attr_temp3_min_alarm.dev_attr.attr,
+	&sensor_dev_attr_temp3_max_alarm.dev_attr.attr,
+	&sensor_dev_attr_temp3_crit_alarm.dev_attr.attr,
+	&sensor_dev_attr_temp3_fault.dev_attr.attr,
+	&sensor_dev_attr_temp3_beep.dev_attr.attr,
+
+	&sensor_dev_attr_temp4_input.dev_attr.attr,		/* 27 */
+	&sensor_dev_attr_temp4_min.dev_attr.attr,
+	&sensor_dev_attr_temp4_max.dev_attr.attr,
+	&sensor_dev_attr_temp4_crit.dev_attr.attr,
+	&sensor_dev_attr_temp4_min_alarm.dev_attr.attr,
+	&sensor_dev_attr_temp4_max_alarm.dev_attr.attr,
+	&sensor_dev_attr_temp4_crit_alarm.dev_attr.attr,
+	&sensor_dev_attr_temp4_beep.dev_attr.attr,
+
+	&sensor_dev_attr_temp5_input.dev_attr.attr,		/* 35 */
+	&sensor_dev_attr_temp5_min.dev_attr.attr,
+	&sensor_dev_attr_temp5_max.dev_attr.attr,
+	&sensor_dev_attr_temp5_crit.dev_attr.attr,
+	&sensor_dev_attr_temp5_min_alarm.dev_attr.attr,
+	&sensor_dev_attr_temp5_max_alarm.dev_attr.attr,
+	&sensor_dev_attr_temp5_crit_alarm.dev_attr.attr,
+	&sensor_dev_attr_temp5_beep.dev_attr.attr,
+
+	&sensor_dev_attr_temp6_input.dev_attr.attr,		/* 43 */
+	&sensor_dev_attr_temp6_beep.dev_attr.attr,
+
+	NULL
+};
+
+static umode_t nct7802_temp_is_visible(struct kobject *kobj,
+				       struct attribute *attr, int index)
+{
+	struct device *dev = container_of(kobj, struct device, kobj);
+	struct nct7802_data *data = dev_get_drvdata(dev);
+	unsigned int reg;
+	int err;
+
+	err = regmap_read(data->regmap, REG_MODE, &reg);
+	if (err < 0)
+		return 0;
+
+	if (index < 9 &&
+	    (reg & 03) != 0x01 && (reg & 0x03) != 0x02)		/* RD1 */
+		return 0;
+	if (index >= 9 && index < 18 &&
+	    (reg & 0x0c) != 0x04 && (reg & 0x0c) != 0x08)	/* RD2 */
+		return 0;
+	if (index >= 18 && index < 27 && (reg & 0x30) != 0x10)	/* RD3 */
+		return 0;
+	if (index >= 27 && index < 35)				/* local */
+		return attr->mode;
+
+	err = regmap_read(data->regmap, REG_PECI_ENABLE, &reg);
+	if (err < 0)
+		return 0;
+
+	if (index >= 35 && index < 43 && !(reg & 0x01))		/* PECI 0 */
+		return 0;
+
+	if (index >= 0x43 && (!(reg & 0x02)))			/* PECI 1 */
+		return 0;
+
+	return attr->mode;
+}
+
+static struct attribute_group nct7802_temp_group = {
+	.attrs = nct7802_temp_attrs,
+	.is_visible = nct7802_temp_is_visible,
+};
+
+static SENSOR_DEVICE_ATTR_2(in0_input, S_IRUGO, show_in, NULL, 0, 0);
+static SENSOR_DEVICE_ATTR_2(in0_min, S_IRUGO | S_IWUSR, show_in, store_in,
+			    0, 1);
+static SENSOR_DEVICE_ATTR_2(in0_max, S_IRUGO | S_IWUSR, show_in, store_in,
+			    0, 2);
+static SENSOR_DEVICE_ATTR_2(in0_alarm, S_IRUGO, show_alarm, NULL, 0x1e, 3);
+static SENSOR_DEVICE_ATTR_2(in0_beep, S_IRUGO | S_IWUSR, show_beep, store_beep,
+			    0x5a, 3);
+
+static SENSOR_DEVICE_ATTR_2(in1_input, S_IRUGO, show_in, NULL, 1, 0);
+
+static SENSOR_DEVICE_ATTR_2(in2_input, S_IRUGO, show_in, NULL, 2, 0);
+static SENSOR_DEVICE_ATTR_2(in2_min, S_IRUGO | S_IWUSR, show_in, store_in,
+			    2, 1);
+static SENSOR_DEVICE_ATTR_2(in2_max, S_IRUGO | S_IWUSR, show_in, store_in,
+			    2, 2);
+static SENSOR_DEVICE_ATTR_2(in2_alarm, S_IRUGO, show_alarm, NULL, 0x1e, 0);
+static SENSOR_DEVICE_ATTR_2(in2_beep, S_IRUGO | S_IWUSR, show_beep, store_beep,
+			    0x5a, 0);
+
+static SENSOR_DEVICE_ATTR_2(in3_input, S_IRUGO, show_in, NULL, 3, 0);
+static SENSOR_DEVICE_ATTR_2(in3_min, S_IRUGO | S_IWUSR, show_in, store_in,
+			    3, 1);
+static SENSOR_DEVICE_ATTR_2(in3_max, S_IRUGO | S_IWUSR, show_in, store_in,
+			    3, 2);
+static SENSOR_DEVICE_ATTR_2(in3_alarm, S_IRUGO, show_alarm, NULL, 0x1e, 1);
+static SENSOR_DEVICE_ATTR_2(in3_beep, S_IRUGO | S_IWUSR, show_beep, store_beep,
+			    0x5a, 1);
+
+static SENSOR_DEVICE_ATTR_2(in4_input, S_IRUGO, show_in, NULL, 4, 0);
+static SENSOR_DEVICE_ATTR_2(in4_min, S_IRUGO | S_IWUSR, show_in, store_in,
+			    4, 1);
+static SENSOR_DEVICE_ATTR_2(in4_max, S_IRUGO | S_IWUSR, show_in, store_in,
+			    4, 2);
+static SENSOR_DEVICE_ATTR_2(in4_alarm, S_IRUGO, show_alarm, NULL, 0x1e, 2);
+static SENSOR_DEVICE_ATTR_2(in4_beep, S_IRUGO | S_IWUSR, show_beep, store_beep,
+			    0x5a, 2);
+
+static struct attribute *nct7802_in_attrs[] = {
+	&sensor_dev_attr_in0_input.dev_attr.attr,
+	&sensor_dev_attr_in0_min.dev_attr.attr,
+	&sensor_dev_attr_in0_max.dev_attr.attr,
+	&sensor_dev_attr_in0_alarm.dev_attr.attr,
+	&sensor_dev_attr_in0_beep.dev_attr.attr,
+
+	&sensor_dev_attr_in1_input.dev_attr.attr,	/* 5 */
+
+	&sensor_dev_attr_in2_input.dev_attr.attr,	/* 6 */
+	&sensor_dev_attr_in2_min.dev_attr.attr,
+	&sensor_dev_attr_in2_max.dev_attr.attr,
+	&sensor_dev_attr_in2_alarm.dev_attr.attr,
+	&sensor_dev_attr_in2_beep.dev_attr.attr,
+
+	&sensor_dev_attr_in3_input.dev_attr.attr,	/* 11 */
+	&sensor_dev_attr_in3_min.dev_attr.attr,
+	&sensor_dev_attr_in3_max.dev_attr.attr,
+	&sensor_dev_attr_in3_alarm.dev_attr.attr,
+	&sensor_dev_attr_in3_beep.dev_attr.attr,
+
+	&sensor_dev_attr_in4_input.dev_attr.attr,	/* 17 */
+	&sensor_dev_attr_in4_min.dev_attr.attr,
+	&sensor_dev_attr_in4_max.dev_attr.attr,
+	&sensor_dev_attr_in4_alarm.dev_attr.attr,
+	&sensor_dev_attr_in4_beep.dev_attr.attr,
+
+	NULL,
+};
+
+static umode_t nct7802_in_is_visible(struct kobject *kobj,
+				     struct attribute *attr, int index)
+{
+	struct device *dev = container_of(kobj, struct device, kobj);
+	struct nct7802_data *data = dev_get_drvdata(dev);
+	unsigned int reg;
+	int err;
+
+	if (index < 6)						/* VCC, VCORE */
+		return attr->mode;
+
+	err = regmap_read(data->regmap, REG_MODE, &reg);
+	if (err < 0)
+		return 0;
+
+	if (index >= 6 && index < 11 && (reg & 0x03) != 0x03)	/* VSEN1 */
+		return 0;
+	if (index >= 11 && index < 17 && (reg & 0x0c) != 0x0c)	/* VSEN2 */
+		return 0;
+	if (index >= 17 && (reg & 0x30) != 0x30)		/* VSEN3 */
+		return 0;
+
+	return attr->mode;
+}
+
+static struct attribute_group nct7802_in_group = {
+	.attrs = nct7802_in_attrs,
+	.is_visible = nct7802_in_is_visible,
+};
+
+static SENSOR_DEVICE_ATTR(fan1_input, S_IRUGO, show_fan, NULL, 0x10);
+static SENSOR_DEVICE_ATTR_2(fan1_min, S_IRUGO | S_IWUSR, show_fan_min,
+			    store_fan_min, 0x49, 0x4c);
+static SENSOR_DEVICE_ATTR_2(fan1_alarm, S_IRUGO, show_alarm, NULL, 0x1a, 0);
+static SENSOR_DEVICE_ATTR_2(fan1_beep, S_IRUGO | S_IWUSR, show_beep, store_beep,
+			    0x5b, 0);
+static SENSOR_DEVICE_ATTR(fan2_input, S_IRUGO, show_fan, NULL, 0x11);
+static SENSOR_DEVICE_ATTR_2(fan2_min, S_IRUGO | S_IWUSR, show_fan_min,
+			    store_fan_min, 0x4a, 0x4d);
+static SENSOR_DEVICE_ATTR_2(fan2_alarm, S_IRUGO, show_alarm, NULL, 0x1a, 1);
+static SENSOR_DEVICE_ATTR_2(fan2_beep, S_IRUGO | S_IWUSR, show_beep, store_beep,
+			    0x5b, 1);
+static SENSOR_DEVICE_ATTR(fan3_input, S_IRUGO, show_fan, NULL, 0x12);
+static SENSOR_DEVICE_ATTR_2(fan3_min, S_IRUGO | S_IWUSR, show_fan_min,
+			    store_fan_min, 0x4b, 0x4e);
+static SENSOR_DEVICE_ATTR_2(fan3_alarm, S_IRUGO, show_alarm, NULL, 0x1a, 2);
+static SENSOR_DEVICE_ATTR_2(fan3_beep, S_IRUGO | S_IWUSR, show_beep, store_beep,
+			    0x5b, 2);
+
+static struct attribute *nct7802_fan_attrs[] = {
+	&sensor_dev_attr_fan1_input.dev_attr.attr,
+	&sensor_dev_attr_fan1_min.dev_attr.attr,
+	&sensor_dev_attr_fan1_alarm.dev_attr.attr,
+	&sensor_dev_attr_fan1_beep.dev_attr.attr,
+	&sensor_dev_attr_fan2_input.dev_attr.attr,
+	&sensor_dev_attr_fan2_min.dev_attr.attr,
+	&sensor_dev_attr_fan2_alarm.dev_attr.attr,
+	&sensor_dev_attr_fan2_beep.dev_attr.attr,
+	&sensor_dev_attr_fan3_input.dev_attr.attr,
+	&sensor_dev_attr_fan3_min.dev_attr.attr,
+	&sensor_dev_attr_fan3_alarm.dev_attr.attr,
+	&sensor_dev_attr_fan3_beep.dev_attr.attr,
+
+	NULL
+};
+
+static umode_t nct7802_fan_is_visible(struct kobject *kobj,
+				      struct attribute *attr, int index)
+{
+	struct device *dev = container_of(kobj, struct device, kobj);
+	struct nct7802_data *data = dev_get_drvdata(dev);
+	int fan = index / 4;	/* 4 attributes per fan */
+	unsigned int reg;
+	int err;
+
+	err = regmap_read(data->regmap, REG_FAN_ENABLE, &reg);
+	if (err < 0 || !(reg & (1 << fan)))
+		return 0;
+
+	return attr->mode;
+}
+
+static struct attribute_group nct7802_fan_group = {
+	.attrs = nct7802_fan_attrs,
+	.is_visible = nct7802_fan_is_visible,
+};
+
+static const struct attribute_group *nct7802_groups[] = {
+	&nct7802_temp_group,
+	&nct7802_in_group,
+	&nct7802_fan_group,
+	NULL
+};
+
+static int nct7802_detect(struct i2c_client *client,
+			  struct i2c_board_info *info)
+{
+	int reg;
+
+	/*
+	 * Chip identification registers are only available in bank 0,
+	 * so only attempt chip detection if bank 0 is selected
+	 */
+	reg = i2c_smbus_read_byte_data(client, REG_BANK);
+	if (reg != 0x00)
+		return -ENODEV;
+
+	reg = i2c_smbus_read_byte_data(client, REG_VENDOR_ID);
+	if (reg != 0x50)
+		return -ENODEV;
+
+	reg = i2c_smbus_read_byte_data(client, REG_CHIP_ID);
+	if (reg != 0xc3)
+		return -ENODEV;
+
+	reg = i2c_smbus_read_byte_data(client, REG_VERSION_ID);
+	if (reg < 0 || (reg & 0xf0) != 0x20)
+		return -ENODEV;
+
+	/* Also validate lower bits of voltage and temperature registers */
+	reg = i2c_smbus_read_byte_data(client, REG_TEMP_LSB);
+	if (reg < 0 || (reg & 0x1f))
+		return -ENODEV;
+
+	reg = i2c_smbus_read_byte_data(client, REG_TEMP_PECI_LSB);
+	if (reg < 0 || (reg & 0x3f))
+		return -ENODEV;
+
+	reg = i2c_smbus_read_byte_data(client, REG_VOLTAGE_LOW);
+	if (reg < 0 || (reg & 0x3f))
+		return -ENODEV;
+
+	strlcpy(info->type, "nct7802", I2C_NAME_SIZE);
+	return 0;
+}
+
+static bool nct7802_regmap_is_volatile(struct device *dev, unsigned int reg)
+{
+	return reg != REG_BANK && reg <= 0x20;
+}
+
+static struct regmap_config nct7802_regmap_config = {
+	.reg_bits = 8,
+	.val_bits = 8,
+	.cache_type = REGCACHE_RBTREE,
+	.volatile_reg = nct7802_regmap_is_volatile,
+};
+
+static int nct7802_init_chip(struct nct7802_data *data)
+{
+	int err;
+
+	/* Enable ADC */
+	err = regmap_update_bits(data->regmap, REG_START, 0x01, 0x01);
+	if (err)
+		return err;
+
+	/* Enable local temperature sensor */
+	err = regmap_update_bits(data->regmap, REG_MODE, 0x40, 0x40);
+	if (err)
+		return err;
+
+	/* Enable Vcore and VCC voltage monitoring */
+	return regmap_update_bits(data->regmap, REG_VMON_ENABLE, 0x03, 0x03);
+}
+
+static int nct7802_probe(struct i2c_client *client,
+			 const struct i2c_device_id *id)
+{
+	struct device *dev = &client->dev;
+	struct nct7802_data *data;
+	struct device *hwmon_dev;
+	int ret;
+
+	data = devm_kzalloc(dev, sizeof(*data), GFP_KERNEL);
+	if (data == NULL)
+		return -ENOMEM;
+
+	data->regmap = devm_regmap_init_i2c(client, &nct7802_regmap_config);
+	if (IS_ERR(data->regmap))
+		return PTR_ERR(data->regmap);
+
+	mutex_init(&data->access_lock);
+
+	ret = nct7802_init_chip(data);
+	if (ret < 0)
+		return ret;
+
+	hwmon_dev = devm_hwmon_device_register_with_groups(dev, client->name,
+							   data,
+							   nct7802_groups);
+	return PTR_ERR_OR_ZERO(hwmon_dev);
+}
+
+static const unsigned short nct7802_address_list[] = {
+	0x28, 0x29, 0x2a, 0x2b, 0x2c, 0x2d, 0x2e, 0x2f, I2C_CLIENT_END
+};
+
+static const struct i2c_device_id nct7802_idtable[] = {
+	{ "nct7802", 0 },
+	{ }
+};
+MODULE_DEVICE_TABLE(i2c, nct7802_idtable);
+
+static struct i2c_driver nct7802_driver = {
+	.class = I2C_CLASS_HWMON,
+	.driver = {
+		.name = DRVNAME,
+	},
+	.detect = nct7802_detect,
+	.probe = nct7802_probe,
+	.id_table = nct7802_idtable,
+	.address_list = nct7802_address_list,
+};
+
+module_i2c_driver(nct7802_driver);
+
+MODULE_AUTHOR("Guenter Roeck <linux@roeck-us.net>");
+MODULE_DESCRIPTION("NCT7802Y Hardware Monitoring Driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/hwmon/ntc_thermistor.c b/drivers/hwmon/ntc_thermistor.c
index bd410722cd4b..112e4d45e4a0 100644
--- a/drivers/hwmon/ntc_thermistor.c
+++ b/drivers/hwmon/ntc_thermistor.c
@@ -38,6 +38,7 @@
 
 #include <linux/hwmon.h>
 #include <linux/hwmon-sysfs.h>
+#include <linux/thermal.h>
 
 struct ntc_compensation {
 	int		temp_c;
@@ -182,6 +183,7 @@ struct ntc_data {
 	struct device *dev;
 	int n_comp;
 	char name[PLATFORM_NAME_SIZE];
+	struct thermal_zone_device *tz;
 };
 
 #if defined(CONFIG_OF) && IS_ENABLED(CONFIG_IIO)
@@ -428,6 +430,20 @@ static int ntc_thermistor_get_ohm(struct ntc_data *data)
 	return -EINVAL;
 }
 
+static int ntc_read_temp(void *dev, long *temp)
+{
+	struct ntc_data *data = dev_get_drvdata(dev);
+	int ohm;
+
+	ohm = ntc_thermistor_get_ohm(data);
+	if (ohm < 0)
+		return ohm;
+
+	*temp = get_temp_mc(data, ohm);
+
+	return 0;
+}
+
 static ssize_t ntc_show_name(struct device *dev,
 		struct device_attribute *attr, char *buf)
 {
@@ -470,6 +486,10 @@ static const struct attribute_group ntc_attr_group = {
 	.attrs = ntc_attributes,
 };
 
+static const struct thermal_zone_of_device_ops ntc_of_thermal_ops = {
+	.get_temp = ntc_read_temp,
+};
+
 static int ntc_thermistor_probe(struct platform_device *pdev)
 {
 	const struct of_device_id *of_id =
@@ -562,6 +582,13 @@ static int ntc_thermistor_probe(struct platform_device *pdev)
 	dev_info(&pdev->dev, "Thermistor type: %s successfully probed.\n",
 								pdev_id->name);
 
+	data->tz = thermal_zone_of_sensor_register(data->dev, 0, data->dev,
+						   &ntc_of_thermal_ops);
+	if (IS_ERR(data->tz)) {
+		dev_dbg(&pdev->dev, "Failed to register to thermal fw.\n");
+		data->tz = NULL;
+	}
+
 	return 0;
 err_after_sysfs:
 	sysfs_remove_group(&data->dev->kobj, &ntc_attr_group);
@@ -578,13 +605,14 @@ static int ntc_thermistor_remove(struct platform_device *pdev)
 	sysfs_remove_group(&data->dev->kobj, &ntc_attr_group);
 	ntc_iio_channel_release(pdata);
 
+	thermal_zone_of_sensor_unregister(data->dev, data->tz);
+
 	return 0;
 }
 
 static struct platform_driver ntc_thermistor_driver = {
 	.driver = {
 		.name = "ntc-thermistor",
-		.owner = THIS_MODULE,
 		.of_match_table = of_match_ptr(ntc_match),
 	},
 	.probe = ntc_thermistor_probe,
diff --git a/drivers/hwmon/pc87360.c b/drivers/hwmon/pc87360.c
index 145f674c1d87..d50fbf93a737 100644
--- a/drivers/hwmon/pc87360.c
+++ b/drivers/hwmon/pc87360.c
@@ -244,7 +244,6 @@ static struct pc87360_data *pc87360_update_device(struct device *dev);
 
 static struct platform_driver pc87360_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= "pc87360",
 	},
 	.probe		= pc87360_probe,
diff --git a/drivers/hwmon/pc87427.c b/drivers/hwmon/pc87427.c
index 9e4684e747ea..cb9fdd37bd0d 100644
--- a/drivers/hwmon/pc87427.c
+++ b/drivers/hwmon/pc87427.c
@@ -1153,7 +1153,6 @@ static int pc87427_remove(struct platform_device *pdev)
 
 static struct platform_driver pc87427_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= DRVNAME,
 	},
 	.probe		= pc87427_probe,
diff --git a/drivers/hwmon/pmbus/Kconfig b/drivers/hwmon/pmbus/Kconfig
index 6e1e4935fc62..a674cd83a4e2 100644
--- a/drivers/hwmon/pmbus/Kconfig
+++ b/drivers/hwmon/pmbus/Kconfig
@@ -47,15 +47,22 @@ config SENSORS_LM25066
 	  be called lm25066.
 
 config SENSORS_LTC2978
-	tristate "Linear Technologies LTC2974, LTC2978, LTC3880, and LTC3883"
+	tristate "Linear Technologies LTC2978 and compatibles"
 	default n
 	help
 	  If you say yes here you get hardware monitoring support for Linear
-	  Technology LTC2974, LTC2978, LTC3880, and LTC3883.
+	  Technology LTC2974, LTC2977, LTC2978, LTC3880, LTC3883, and LTM4676.
 
 	  This driver can also be built as a module. If so, the module will
 	  be called ltc2978.
 
+config SENSORS_LTC2978_REGULATOR
+	boolean "Regulator support for LTC2978 and compatibles"
+	depends on SENSORS_LTC2978 && REGULATOR
+	help
+	  If you say yes here you get regulator support for Linear
+	  Technology LTC2974, LTC2977, LTC2978, LTC3880, LTC3883, and LTM4676.
+
 config SENSORS_MAX16064
 	tristate "Maxim MAX16064"
 	default n
diff --git a/drivers/hwmon/pmbus/ltc2978.c b/drivers/hwmon/pmbus/ltc2978.c
index e24ed521051a..0835050ec245 100644
--- a/drivers/hwmon/pmbus/ltc2978.c
+++ b/drivers/hwmon/pmbus/ltc2978.c
@@ -22,6 +22,7 @@
 #include <linux/err.h>
 #include <linux/slab.h>
 #include <linux/i2c.h>
+#include <linux/regulator/driver.h>
 #include "pmbus.h"
 
 enum chips { ltc2974, ltc2977, ltc2978, ltc3880, ltc3883, ltm4676 };
@@ -374,6 +375,19 @@ static const struct i2c_device_id ltc2978_id[] = {
 };
 MODULE_DEVICE_TABLE(i2c, ltc2978_id);
 
+#if IS_ENABLED(CONFIG_SENSORS_LTC2978_REGULATOR)
+static const struct regulator_desc ltc2978_reg_desc[] = {
+	PMBUS_REGULATOR("vout", 0),
+	PMBUS_REGULATOR("vout", 1),
+	PMBUS_REGULATOR("vout", 2),
+	PMBUS_REGULATOR("vout", 3),
+	PMBUS_REGULATOR("vout", 4),
+	PMBUS_REGULATOR("vout", 5),
+	PMBUS_REGULATOR("vout", 6),
+	PMBUS_REGULATOR("vout", 7),
+};
+#endif /* CONFIG_SENSORS_LTC2978_REGULATOR */
+
 static int ltc2978_probe(struct i2c_client *client,
 			 const struct i2c_device_id *id)
 {
@@ -487,13 +501,36 @@ static int ltc2978_probe(struct i2c_client *client,
 	default:
 		return -ENODEV;
 	}
+
+#if IS_ENABLED(CONFIG_SENSORS_LTC2978_REGULATOR)
+	info->num_regulators = info->pages;
+	info->reg_desc = ltc2978_reg_desc;
+	if (info->num_regulators > ARRAY_SIZE(ltc2978_reg_desc)) {
+		dev_err(&client->dev, "num_regulators too large!");
+		info->num_regulators = ARRAY_SIZE(ltc2978_reg_desc);
+	}
+#endif
+
 	return pmbus_do_probe(client, id, info);
 }
 
-/* This is the driver that will be inserted */
+#ifdef CONFIG_OF
+static const struct of_device_id ltc2978_of_match[] = {
+	{ .compatible = "lltc,ltc2974" },
+	{ .compatible = "lltc,ltc2977" },
+	{ .compatible = "lltc,ltc2978" },
+	{ .compatible = "lltc,ltc3880" },
+	{ .compatible = "lltc,ltc3883" },
+	{ .compatible = "lltc,ltm4676" },
+	{ }
+};
+MODULE_DEVICE_TABLE(of, ltc2978_of_match);
+#endif
+
 static struct i2c_driver ltc2978_driver = {
 	.driver = {
 		   .name = "ltc2978",
+		   .of_match_table = of_match_ptr(ltc2978_of_match),
 		   },
 	.probe = ltc2978_probe,
 	.remove = pmbus_do_remove,
diff --git a/drivers/hwmon/pmbus/pmbus.h b/drivers/hwmon/pmbus/pmbus.h
index fa9beb3eb60c..89a23ff836e7 100644
--- a/drivers/hwmon/pmbus/pmbus.h
+++ b/drivers/hwmon/pmbus/pmbus.h
@@ -19,6 +19,8 @@
  * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  */
 
+#include <linux/regulator/driver.h>
+
 #ifndef PMBUS_H
 #define PMBUS_H
 
@@ -186,6 +188,11 @@
 #define PMBUS_VIRT_STATUS_VMON		(PMBUS_VIRT_BASE + 35)
 
 /*
+ * OPERATION
+ */
+#define PB_OPERATION_CONTROL_ON		(1<<7)
+
+/*
  * CAPABILITY
  */
 #define PB_CAPABILITY_SMBALERT		(1<<4)
@@ -365,8 +372,27 @@ struct pmbus_driver_info {
 	 */
 	int (*identify)(struct i2c_client *client,
 			struct pmbus_driver_info *info);
+
+	/* Regulator functionality, if supported by this chip driver. */
+	int num_regulators;
+	const struct regulator_desc *reg_desc;
 };
 
+/* Regulator ops */
+
+extern struct regulator_ops pmbus_regulator_ops;
+
+/* Macro for filling in array of struct regulator_desc */
+#define PMBUS_REGULATOR(_name, _id)				\
+	[_id] = {						\
+		.name = (_name # _id),				\
+		.id = (_id),					\
+		.of_match = of_match_ptr(_name # _id),		\
+		.regulators_node = of_match_ptr("regulators"),	\
+		.ops = &pmbus_regulator_ops,			\
+		.owner = THIS_MODULE,				\
+	}
+
 /* Function declarations */
 
 void pmbus_clear_cache(struct i2c_client *client);
@@ -375,6 +401,10 @@ int pmbus_read_word_data(struct i2c_client *client, u8 page, u8 reg);
 int pmbus_write_word_data(struct i2c_client *client, u8 page, u8 reg, u16 word);
 int pmbus_read_byte_data(struct i2c_client *client, int page, u8 reg);
 int pmbus_write_byte(struct i2c_client *client, int page, u8 value);
+int pmbus_write_byte_data(struct i2c_client *client, int page, u8 reg,
+			  u8 value);
+int pmbus_update_byte_data(struct i2c_client *client, int page, u8 reg,
+			   u8 mask, u8 value);
 void pmbus_clear_faults(struct i2c_client *client);
 bool pmbus_check_byte_register(struct i2c_client *client, int page, int reg);
 bool pmbus_check_word_register(struct i2c_client *client, int page, int reg);
diff --git a/drivers/hwmon/pmbus/pmbus_core.c b/drivers/hwmon/pmbus/pmbus_core.c
index 291d11fe93e7..f2e47c7dd808 100644
--- a/drivers/hwmon/pmbus/pmbus_core.c
+++ b/drivers/hwmon/pmbus/pmbus_core.c
@@ -29,6 +29,8 @@
 #include <linux/hwmon-sysfs.h>
 #include <linux/jiffies.h>
 #include <linux/i2c/pmbus.h>
+#include <linux/regulator/driver.h>
+#include <linux/regulator/machine.h>
 #include "pmbus.h"
 
 /*
@@ -253,6 +255,37 @@ int pmbus_read_byte_data(struct i2c_client *client, int page, u8 reg)
 }
 EXPORT_SYMBOL_GPL(pmbus_read_byte_data);
 
+int pmbus_write_byte_data(struct i2c_client *client, int page, u8 reg, u8 value)
+{
+	int rv;
+
+	rv = pmbus_set_page(client, page);
+	if (rv < 0)
+		return rv;
+
+	return i2c_smbus_write_byte_data(client, reg, value);
+}
+EXPORT_SYMBOL_GPL(pmbus_write_byte_data);
+
+int pmbus_update_byte_data(struct i2c_client *client, int page, u8 reg,
+			   u8 mask, u8 value)
+{
+	unsigned int tmp;
+	int rv;
+
+	rv = pmbus_read_byte_data(client, page, reg);
+	if (rv < 0)
+		return rv;
+
+	tmp = (rv & ~mask) | (value & mask);
+
+	if (tmp != rv)
+		rv = pmbus_write_byte_data(client, page, reg, tmp);
+
+	return rv;
+}
+EXPORT_SYMBOL_GPL(pmbus_update_byte_data);
+
 /*
  * _pmbus_read_byte_data() is similar to pmbus_read_byte_data(), but checks if
  * a device specific mapping function exists and calls it if necessary.
@@ -1727,6 +1760,84 @@ static int pmbus_init_common(struct i2c_client *client, struct pmbus_data *data,
 	return 0;
 }
 
+#if IS_ENABLED(CONFIG_REGULATOR)
+static int pmbus_regulator_is_enabled(struct regulator_dev *rdev)
+{
+	struct device *dev = rdev_get_dev(rdev);
+	struct i2c_client *client = to_i2c_client(dev->parent);
+	u8 page = rdev_get_id(rdev);
+	int ret;
+
+	ret = pmbus_read_byte_data(client, page, PMBUS_OPERATION);
+	if (ret < 0)
+		return ret;
+
+	return !!(ret & PB_OPERATION_CONTROL_ON);
+}
+
+static int _pmbus_regulator_on_off(struct regulator_dev *rdev, bool enable)
+{
+	struct device *dev = rdev_get_dev(rdev);
+	struct i2c_client *client = to_i2c_client(dev->parent);
+	u8 page = rdev_get_id(rdev);
+
+	return pmbus_update_byte_data(client, page, PMBUS_OPERATION,
+				      PB_OPERATION_CONTROL_ON,
+				      enable ? PB_OPERATION_CONTROL_ON : 0);
+}
+
+static int pmbus_regulator_enable(struct regulator_dev *rdev)
+{
+	return _pmbus_regulator_on_off(rdev, 1);
+}
+
+static int pmbus_regulator_disable(struct regulator_dev *rdev)
+{
+	return _pmbus_regulator_on_off(rdev, 0);
+}
+
+struct regulator_ops pmbus_regulator_ops = {
+	.enable = pmbus_regulator_enable,
+	.disable = pmbus_regulator_disable,
+	.is_enabled = pmbus_regulator_is_enabled,
+};
+EXPORT_SYMBOL_GPL(pmbus_regulator_ops);
+
+static int pmbus_regulator_register(struct pmbus_data *data)
+{
+	struct device *dev = data->dev;
+	const struct pmbus_driver_info *info = data->info;
+	const struct pmbus_platform_data *pdata = dev_get_platdata(dev);
+	struct regulator_dev *rdev;
+	int i;
+
+	for (i = 0; i < info->num_regulators; i++) {
+		struct regulator_config config = { };
+
+		config.dev = dev;
+		config.driver_data = data;
+
+		if (pdata && pdata->reg_init_data)
+			config.init_data = &pdata->reg_init_data[i];
+
+		rdev = devm_regulator_register(dev, &info->reg_desc[i],
+					       &config);
+		if (IS_ERR(rdev)) {
+			dev_err(dev, "Failed to register %s regulator\n",
+				info->reg_desc[i].name);
+			return PTR_ERR(rdev);
+		}
+	}
+
+	return 0;
+}
+#else
+static int pmbus_regulator_register(struct pmbus_data *data)
+{
+	return 0;
+}
+#endif
+
 int pmbus_do_probe(struct i2c_client *client, const struct i2c_device_id *id,
 		   struct pmbus_driver_info *info)
 {
@@ -1781,8 +1892,15 @@ int pmbus_do_probe(struct i2c_client *client, const struct i2c_device_id *id,
 		dev_err(dev, "Failed to register hwmon device\n");
 		goto out_kfree;
 	}
+
+	ret = pmbus_regulator_register(data);
+	if (ret)
+		goto out_unregister;
+
 	return 0;
 
+out_unregister:
+	hwmon_device_unregister(data->hwmon_dev);
 out_kfree:
 	kfree(data->group.attrs);
 	return ret;
diff --git a/drivers/hwmon/pwm-fan.c b/drivers/hwmon/pwm-fan.c
index 823c877a1ec0..1991d9032c38 100644
--- a/drivers/hwmon/pwm-fan.c
+++ b/drivers/hwmon/pwm-fan.c
@@ -161,10 +161,17 @@ static int pwm_fan_suspend(struct device *dev)
 static int pwm_fan_resume(struct device *dev)
 {
 	struct pwm_fan_ctx *ctx = dev_get_drvdata(dev);
+	unsigned long duty;
+	int ret;
 
-	if (ctx->pwm_value)
-		return pwm_enable(ctx->pwm);
-	return 0;
+	if (ctx->pwm_value == 0)
+		return 0;
+
+	duty = DIV_ROUND_UP(ctx->pwm_value * (ctx->pwm->period - 1), MAX_PWM);
+	ret = pwm_config(ctx->pwm, duty, ctx->pwm->period);
+	if (ret)
+		return ret;
+	return pwm_enable(ctx->pwm);
 }
 #endif
 
diff --git a/drivers/hwmon/s3c-hwmon.c b/drivers/hwmon/s3c-hwmon.c
index 0674c13bbd4b..0c4710d35d16 100644
--- a/drivers/hwmon/s3c-hwmon.c
+++ b/drivers/hwmon/s3c-hwmon.c
@@ -378,7 +378,6 @@ static int s3c_hwmon_remove(struct platform_device *dev)
 static struct platform_driver s3c_hwmon_driver = {
 	.driver	= {
 		.name		= "s3c-hwmon",
-		.owner		= THIS_MODULE,
 	},
 	.probe		= s3c_hwmon_probe,
 	.remove		= s3c_hwmon_remove,
diff --git a/drivers/hwmon/sch5627.c b/drivers/hwmon/sch5627.c
index 0cc99fd83e8e..19f85c0da270 100644
--- a/drivers/hwmon/sch5627.c
+++ b/drivers/hwmon/sch5627.c
@@ -591,7 +591,6 @@ error:
 
 static struct platform_driver sch5627_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= DRVNAME,
 	},
 	.probe		= sch5627_probe,
diff --git a/drivers/hwmon/sch5636.c b/drivers/hwmon/sch5636.c
index 547b5c952eff..131a2815dbda 100644
--- a/drivers/hwmon/sch5636.c
+++ b/drivers/hwmon/sch5636.c
@@ -521,7 +521,6 @@ error:
 
 static struct platform_driver sch5636_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= DRVNAME,
 	},
 	.probe		= sch5636_probe,
diff --git a/drivers/hwmon/sht15.c b/drivers/hwmon/sht15.c
index 97cd45a8432c..d4f0935daaa1 100644
--- a/drivers/hwmon/sht15.c
+++ b/drivers/hwmon/sht15.c
@@ -1087,7 +1087,6 @@ MODULE_DEVICE_TABLE(platform, sht15_device_ids);
 static struct platform_driver sht15_driver = {
 	.driver = {
 		.name = "sht15",
-		.owner = THIS_MODULE,
 	},
 	.probe = sht15_probe,
 	.remove = sht15_remove,
diff --git a/drivers/hwmon/sis5595.c b/drivers/hwmon/sis5595.c
index bf1d7893d51c..45a028fb8851 100644
--- a/drivers/hwmon/sis5595.c
+++ b/drivers/hwmon/sis5595.c
@@ -215,7 +215,6 @@ static void sis5595_init_device(struct sis5595_data *data);
 
 static struct platform_driver sis5595_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= "sis5595",
 	},
 	.probe		= sis5595_probe,
diff --git a/drivers/hwmon/smsc47b397.c b/drivers/hwmon/smsc47b397.c
index bd89e87bd6ae..6bd200756560 100644
--- a/drivers/hwmon/smsc47b397.c
+++ b/drivers/hwmon/smsc47b397.c
@@ -100,8 +100,6 @@ static u8 smsc47b397_reg_temp[] = {0x25, 0x26, 0x27, 0x80};
 
 struct smsc47b397_data {
 	unsigned short addr;
-	const char *name;
-	struct device *hwmon_dev;
 	struct mutex lock;
 
 	struct mutex update_lock;
@@ -202,15 +200,7 @@ static SENSOR_DEVICE_ATTR(fan2_input, S_IRUGO, show_fan, NULL, 1);
 static SENSOR_DEVICE_ATTR(fan3_input, S_IRUGO, show_fan, NULL, 2);
 static SENSOR_DEVICE_ATTR(fan4_input, S_IRUGO, show_fan, NULL, 3);
 
-static ssize_t show_name(struct device *dev, struct device_attribute
-			 *devattr, char *buf)
-{
-	struct smsc47b397_data *data = dev_get_drvdata(dev);
-	return sprintf(buf, "%s\n", data->name);
-}
-static DEVICE_ATTR(name, S_IRUGO, show_name, NULL);
-
-static struct attribute *smsc47b397_attributes[] = {
+static struct attribute *smsc47b397_attrs[] = {
 	&sensor_dev_attr_temp1_input.dev_attr.attr,
 	&sensor_dev_attr_temp2_input.dev_attr.attr,
 	&sensor_dev_attr_temp3_input.dev_attr.attr,
@@ -220,41 +210,26 @@ static struct attribute *smsc47b397_attributes[] = {
 	&sensor_dev_attr_fan3_input.dev_attr.attr,
 	&sensor_dev_attr_fan4_input.dev_attr.attr,
 
-	&dev_attr_name.attr,
 	NULL
 };
 
-static const struct attribute_group smsc47b397_group = {
-	.attrs = smsc47b397_attributes,
-};
-
-static int smsc47b397_remove(struct platform_device *pdev)
-{
-	struct smsc47b397_data *data = platform_get_drvdata(pdev);
-
-	hwmon_device_unregister(data->hwmon_dev);
-	sysfs_remove_group(&pdev->dev.kobj, &smsc47b397_group);
-
-	return 0;
-}
+ATTRIBUTE_GROUPS(smsc47b397);
 
 static int smsc47b397_probe(struct platform_device *pdev);
 
 static struct platform_driver smsc47b397_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= DRVNAME,
 	},
 	.probe		= smsc47b397_probe,
-	.remove		= smsc47b397_remove,
 };
 
 static int smsc47b397_probe(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
 	struct smsc47b397_data *data;
+	struct device *hwmon_dev;
 	struct resource *res;
-	int err = 0;
 
 	res = platform_get_resource(pdev, IORESOURCE_IO, 0);
 	if (!devm_request_region(dev, res->start, SMSC_EXTENT,
@@ -270,26 +245,13 @@ static int smsc47b397_probe(struct platform_device *pdev)
 		return -ENOMEM;
 
 	data->addr = res->start;
-	data->name = "smsc47b397";
 	mutex_init(&data->lock);
 	mutex_init(&data->update_lock);
-	platform_set_drvdata(pdev, data);
-
-	err = sysfs_create_group(&dev->kobj, &smsc47b397_group);
-	if (err)
-		return err;
 
-	data->hwmon_dev = hwmon_device_register(dev);
-	if (IS_ERR(data->hwmon_dev)) {
-		err = PTR_ERR(data->hwmon_dev);
-		goto error_remove;
-	}
-
-	return 0;
-
-error_remove:
-	sysfs_remove_group(&dev->kobj, &smsc47b397_group);
-	return err;
+	hwmon_dev = devm_hwmon_device_register_with_groups(dev, "smsc47b397",
+							   data,
+							   smsc47b397_groups);
+	return PTR_ERR_OR_ZERO(hwmon_dev);
 }
 
 static int __init smsc47b397_device_add(unsigned short address)
diff --git a/drivers/hwmon/smsc47m1.c b/drivers/hwmon/smsc47m1.c
index d7485659acc5..5d323186d2c1 100644
--- a/drivers/hwmon/smsc47m1.c
+++ b/drivers/hwmon/smsc47m1.c
@@ -849,7 +849,6 @@ static int __exit smsc47m1_remove(struct platform_device *pdev)
 
 static struct platform_driver smsc47m1_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= DRVNAME,
 	},
 	.remove		= __exit_p(smsc47m1_remove),
diff --git a/drivers/hwmon/tmp102.c b/drivers/hwmon/tmp102.c
index 51719956cc03..ba9f478f64ee 100644
--- a/drivers/hwmon/tmp102.c
+++ b/drivers/hwmon/tmp102.c
@@ -158,6 +158,10 @@ ATTRIBUTE_GROUPS(tmp102);
 #define TMP102_CONFIG  (TMP102_CONF_TM | TMP102_CONF_EM | TMP102_CONF_CR1)
 #define TMP102_CONFIG_RD_ONLY (TMP102_CONF_R0 | TMP102_CONF_R1 | TMP102_CONF_AL)
 
+static const struct thermal_zone_of_device_ops tmp102_of_thermal_ops = {
+	.get_temp = tmp102_read_temp,
+};
+
 static int tmp102_probe(struct i2c_client *client,
 				  const struct i2c_device_id *id)
 {
@@ -215,7 +219,7 @@ static int tmp102_probe(struct i2c_client *client,
 	}
 	tmp102->hwmon_dev = hwmon_dev;
 	tmp102->tz = thermal_zone_of_sensor_register(hwmon_dev, 0, hwmon_dev,
-						     tmp102_read_temp, NULL);
+						     &tmp102_of_thermal_ops);
 	if (IS_ERR(tmp102->tz))
 		tmp102->tz = NULL;
 
diff --git a/drivers/hwmon/tmp401.c b/drivers/hwmon/tmp401.c
index 7fa6e7d0b9b6..99664ebc738d 100644
--- a/drivers/hwmon/tmp401.c
+++ b/drivers/hwmon/tmp401.c
@@ -44,9 +44,10 @@
 #include <linux/sysfs.h>
 
 /* Addresses to scan */
-static const unsigned short normal_i2c[] = { 0x4c, 0x4d, 0x4e, I2C_CLIENT_END };
+static const unsigned short normal_i2c[] = { 0x37, 0x48, 0x49, 0x4a, 0x4c, 0x4d,
+	0x4e, 0x4f, I2C_CLIENT_END };
 
-enum chips { tmp401, tmp411, tmp431, tmp432 };
+enum chips { tmp401, tmp411, tmp431, tmp432, tmp435 };
 
 /*
  * The TMP401 registers, note some registers have different addresses for
@@ -136,6 +137,7 @@ static const u8 TMP432_STATUS_REG[] = {
 #define TMP411C_DEVICE_ID			0x10
 #define TMP431_DEVICE_ID			0x31
 #define TMP432_DEVICE_ID			0x32
+#define TMP435_DEVICE_ID			0x35
 
 /*
  * Driver data (common to all clients)
@@ -146,6 +148,7 @@ static const struct i2c_device_id tmp401_id[] = {
 	{ "tmp411", tmp411 },
 	{ "tmp431", tmp431 },
 	{ "tmp432", tmp432 },
+	{ "tmp435", tmp435 },
 	{ }
 };
 MODULE_DEVICE_TABLE(i2c, tmp401_id);
@@ -613,10 +616,10 @@ static const struct attribute_group tmp432_group = {
  * Begin non sysfs callback code (aka Real code)
  */
 
-static void tmp401_init_client(struct tmp401_data *data,
-			       struct i2c_client *client)
+static int tmp401_init_client(struct tmp401_data *data,
+			      struct i2c_client *client)
 {
-	int config, config_orig;
+	int config, config_orig, status = 0;
 
 	/* Set the conversion rate to 2 Hz */
 	i2c_smbus_write_byte_data(client, TMP401_CONVERSION_RATE_WRITE, 5);
@@ -624,16 +627,18 @@ static void tmp401_init_client(struct tmp401_data *data,
 
 	/* Start conversions (disable shutdown if necessary) */
 	config = i2c_smbus_read_byte_data(client, TMP401_CONFIG_READ);
-	if (config < 0) {
-		dev_warn(&client->dev, "Initialization failed!\n");
-		return;
-	}
+	if (config < 0)
+		return config;
 
 	config_orig = config;
 	config &= ~TMP401_CONFIG_SHUTDOWN;
 
 	if (config != config_orig)
-		i2c_smbus_write_byte_data(client, TMP401_CONFIG_WRITE, config);
+		status = i2c_smbus_write_byte_data(client,
+						   TMP401_CONFIG_WRITE,
+						   config);
+
+	return status;
 }
 
 static int tmp401_detect(struct i2c_client *client,
@@ -675,15 +680,18 @@ static int tmp401_detect(struct i2c_client *client,
 		kind = tmp411;
 		break;
 	case TMP431_DEVICE_ID:
-		if (client->addr == 0x4e)
+		if (client->addr != 0x4c && client->addr != 0x4d)
 			return -ENODEV;
 		kind = tmp431;
 		break;
 	case TMP432_DEVICE_ID:
-		if (client->addr == 0x4e)
+		if (client->addr != 0x4c && client->addr != 0x4d)
 			return -ENODEV;
 		kind = tmp432;
 		break;
+	case TMP435_DEVICE_ID:
+		kind = tmp435;
+		break;
 	default:
 		return -ENODEV;
 	}
@@ -705,11 +713,13 @@ static int tmp401_detect(struct i2c_client *client,
 static int tmp401_probe(struct i2c_client *client,
 			const struct i2c_device_id *id)
 {
-	const char *names[] = { "TMP401", "TMP411", "TMP431", "TMP432" };
+	static const char * const names[] = {
+		"TMP401", "TMP411", "TMP431", "TMP432", "TMP435"
+	};
 	struct device *dev = &client->dev;
 	struct device *hwmon_dev;
 	struct tmp401_data *data;
-	int groups = 0;
+	int groups = 0, status;
 
 	data = devm_kzalloc(dev, sizeof(struct tmp401_data), GFP_KERNEL);
 	if (!data)
@@ -720,7 +730,9 @@ static int tmp401_probe(struct i2c_client *client,
 	data->kind = id->driver_data;
 
 	/* Initialize the TMP401 chip */
-	tmp401_init_client(data, client);
+	status = tmp401_init_client(data, client);
+	if (status < 0)
+		return status;
 
 	/* Register sysfs hooks */
 	data->groups[groups++] = &tmp401_group;
diff --git a/drivers/hwmon/twl4030-madc-hwmon.c b/drivers/hwmon/twl4030-madc-hwmon.c
index 9a0e2b8e8b94..b5caf7fdb487 100644
--- a/drivers/hwmon/twl4030-madc-hwmon.c
+++ b/drivers/hwmon/twl4030-madc-hwmon.c
@@ -107,7 +107,6 @@ static struct platform_driver twl4030_madc_hwmon_driver = {
 	.probe = twl4030_madc_hwmon_probe,
 	.driver = {
 		   .name = "twl4030_madc_hwmon",
-		   .owner = THIS_MODULE,
 		   },
 };
 
diff --git a/drivers/hwmon/ultra45_env.c b/drivers/hwmon/ultra45_env.c
index 7d4658636064..f2816c7c918f 100644
--- a/drivers/hwmon/ultra45_env.c
+++ b/drivers/hwmon/ultra45_env.c
@@ -314,7 +314,6 @@ MODULE_DEVICE_TABLE(of, env_match);
 static struct platform_driver env_driver = {
 	.driver = {
 		.name = "ultra45_env",
-		.owner = THIS_MODULE,
 		.of_match_table = env_match,
 	},
 	.probe		= env_probe,
diff --git a/drivers/hwmon/vexpress.c b/drivers/hwmon/vexpress.c
index c53619086f33..cf1848b8fb32 100644
--- a/drivers/hwmon/vexpress.c
+++ b/drivers/hwmon/vexpress.c
@@ -247,7 +247,6 @@ static struct platform_driver vexpress_hwmon_driver = {
 	.probe = vexpress_hwmon_probe,
 	.driver	= {
 		.name = DRVNAME,
-		.owner = THIS_MODULE,
 		.of_match_table = vexpress_hwmon_of_match,
 	},
 };
diff --git a/drivers/hwmon/via-cputemp.c b/drivers/hwmon/via-cputemp.c
index 8df43c51de2c..ac91c07e3f90 100644
--- a/drivers/hwmon/via-cputemp.c
+++ b/drivers/hwmon/via-cputemp.c
@@ -205,7 +205,6 @@ static int via_cputemp_remove(struct platform_device *pdev)
 
 static struct platform_driver via_cputemp_driver = {
 	.driver = {
-		.owner = THIS_MODULE,
 		.name = DRVNAME,
 	},
 	.probe = via_cputemp_probe,
diff --git a/drivers/hwmon/via686a.c b/drivers/hwmon/via686a.c
index babd732b4e18..40dd93c8f9f4 100644
--- a/drivers/hwmon/via686a.c
+++ b/drivers/hwmon/via686a.c
@@ -670,7 +670,6 @@ static const struct attribute_group via686a_group = {
 
 static struct platform_driver via686a_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= "via686a",
 	},
 	.probe		= via686a_probe,
diff --git a/drivers/hwmon/vt1211.c b/drivers/hwmon/vt1211.c
index 3ea57c3504e2..3a6bfa51cb94 100644
--- a/drivers/hwmon/vt1211.c
+++ b/drivers/hwmon/vt1211.c
@@ -1233,7 +1233,6 @@ static int vt1211_remove(struct platform_device *pdev)
 
 static struct platform_driver vt1211_driver = {
 	.driver = {
-		.owner = THIS_MODULE,
 		.name  = DRVNAME,
 	},
 	.probe  = vt1211_probe,
diff --git a/drivers/hwmon/vt8231.c b/drivers/hwmon/vt8231.c
index b3babe3326fb..cb69a8c2ed5b 100644
--- a/drivers/hwmon/vt8231.c
+++ b/drivers/hwmon/vt8231.c
@@ -759,7 +759,6 @@ static const struct attribute_group vt8231_group = {
 
 static struct platform_driver vt8231_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= "vt8231",
 	},
 	.probe	= vt8231_probe,
diff --git a/drivers/hwmon/w83627ehf.c b/drivers/hwmon/w83627ehf.c
index f0ab61db7a0d..b10353b31806 100644
--- a/drivers/hwmon/w83627ehf.c
+++ b/drivers/hwmon/w83627ehf.c
@@ -2705,7 +2705,6 @@ static const struct dev_pm_ops w83627ehf_dev_pm_ops = {
 
 static struct platform_driver w83627ehf_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= DRVNAME,
 		.pm	= W83627EHF_DEV_PM_OPS,
 	},
diff --git a/drivers/hwmon/w83627hf.c b/drivers/hwmon/w83627hf.c
index 2f55973a8c4c..721295b9a051 100644
--- a/drivers/hwmon/w83627hf.c
+++ b/drivers/hwmon/w83627hf.c
@@ -474,7 +474,6 @@ static const struct dev_pm_ops w83627hf_dev_pm_ops = {
 
 static struct platform_driver w83627hf_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= DRVNAME,
 		.pm	= W83627HF_DEV_PM_OPS,
 	},
diff --git a/drivers/hwmon/w83781d.c b/drivers/hwmon/w83781d.c
index 84911616d8c0..54848fdd181e 100644
--- a/drivers/hwmon/w83781d.c
+++ b/drivers/hwmon/w83781d.c
@@ -1839,7 +1839,6 @@ w83781d_isa_remove(struct platform_device *pdev)
 
 static struct platform_driver w83781d_isa_driver = {
 	.driver = {
-		.owner = THIS_MODULE,
 		.name = "w83781d",
 	},
 	.probe = w83781d_isa_probe,
diff --git a/drivers/hwmon/wm831x-hwmon.c b/drivers/hwmon/wm831x-hwmon.c
index 3e6a3195cd11..a16cce72e4e2 100644
--- a/drivers/hwmon/wm831x-hwmon.c
+++ b/drivers/hwmon/wm831x-hwmon.c
@@ -154,7 +154,6 @@ static struct platform_driver wm831x_hwmon_driver = {
 	.probe = wm831x_hwmon_probe,
 	.driver = {
 		.name = "wm831x-hwmon",
-		.owner = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/hwmon/wm8350-hwmon.c b/drivers/hwmon/wm8350-hwmon.c
index 90e3d918e597..31af438ffa88 100644
--- a/drivers/hwmon/wm8350-hwmon.c
+++ b/drivers/hwmon/wm8350-hwmon.c
@@ -93,7 +93,6 @@ static struct platform_driver wm8350_hwmon_driver = {
 	.probe = wm8350_hwmon_probe,
 	.driver = {
 		.name = "wm8350-hwmon",
-		.owner = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/hwspinlock/omap_hwspinlock.c b/drivers/hwspinlock/omap_hwspinlock.c
index c1e2cd4d85fe..47a275c6ece1 100644
--- a/drivers/hwspinlock/omap_hwspinlock.c
+++ b/drivers/hwspinlock/omap_hwspinlock.c
@@ -179,7 +179,6 @@ static struct platform_driver omap_hwspinlock_driver = {
 	.remove		= omap_hwspinlock_remove,
 	.driver		= {
 		.name	= "omap_hwspinlock",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/hwspinlock/u8500_hsem.c b/drivers/hwspinlock/u8500_hsem.c
index 401c33bcdb45..e93eabbd660f 100644
--- a/drivers/hwspinlock/u8500_hsem.c
+++ b/drivers/hwspinlock/u8500_hsem.c
@@ -175,7 +175,6 @@ static struct platform_driver u8500_hsem_driver = {
 	.remove		= u8500_hsem_remove,
 	.driver		= {
 		.name	= "u8500_hsem",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/i2c/Kconfig b/drivers/i2c/Kconfig
index b51a402752c4..8c9e619f3026 100644
--- a/drivers/i2c/Kconfig
+++ b/drivers/i2c/Kconfig
@@ -110,6 +110,16 @@ config I2C_STUB
 
 	  If you don't know what to do here, definitely say N.
 
+config I2C_SLAVE
+	bool "I2C slave support"
+
+if I2C_SLAVE
+
+config I2C_SLAVE_EEPROM
+	tristate "I2C eeprom slave driver"
+
+endif
+
 config I2C_DEBUG_CORE
 	bool "I2C Core debugging messages"
 	help
diff --git a/drivers/i2c/Makefile b/drivers/i2c/Makefile
index 1722f50f2473..45095b3d16a9 100644
--- a/drivers/i2c/Makefile
+++ b/drivers/i2c/Makefile
@@ -9,6 +9,7 @@ obj-$(CONFIG_I2C_CHARDEV)	+= i2c-dev.o
 obj-$(CONFIG_I2C_MUX)		+= i2c-mux.o
 obj-y				+= algos/ busses/ muxes/
 obj-$(CONFIG_I2C_STUB)		+= i2c-stub.o
+obj-$(CONFIG_I2C_SLAVE_EEPROM)	+= i2c-slave-eeprom.o
 
 ccflags-$(CONFIG_I2C_DEBUG_CORE) := -DDEBUG
 CFLAGS_i2c-core.o := -Wno-deprecated-declarations
diff --git a/drivers/i2c/algos/i2c-algo-bit.c b/drivers/i2c/algos/i2c-algo-bit.c
index 65ef9664d5da..899bede81b31 100644
--- a/drivers/i2c/algos/i2c-algo-bit.c
+++ b/drivers/i2c/algos/i2c-algo-bit.c
@@ -12,11 +12,6 @@
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
     GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
-    MA 02110-1301 USA.
  * ------------------------------------------------------------------------- */
 
 /* With some changes from Frodo Looijaard <frodol@dds.nl>, Kyösti Mälkki
diff --git a/drivers/i2c/algos/i2c-algo-pca.c b/drivers/i2c/algos/i2c-algo-pca.c
index 8b10f88b13d9..580dbf05c148 100644
--- a/drivers/i2c/algos/i2c-algo-pca.c
+++ b/drivers/i2c/algos/i2c-algo-pca.c
@@ -12,11 +12,6 @@
  *  but WITHOUT ANY WARRANTY; without even the implied warranty of
  *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  *  GNU General Public License for more details.
- *
- *  You should have received a copy of the GNU General Public License
- *  along with this program; if not, write to the Free Software
- *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- *  MA 02110-1301 USA.
  */
 
 #include <linux/kernel.h>
diff --git a/drivers/i2c/algos/i2c-algo-pcf.c b/drivers/i2c/algos/i2c-algo-pcf.c
index 34370090b753..270d84bfc2c6 100644
--- a/drivers/i2c/algos/i2c-algo-pcf.c
+++ b/drivers/i2c/algos/i2c-algo-pcf.c
@@ -14,11 +14,6 @@
  *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  *  GNU General Public License for more details.
  *
- *  You should have received a copy of the GNU General Public License
- *  along with this program; if not, write to the Free Software
- *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- *  MA 02110-1301 USA.
- *
  * With some changes from Kyösti Mälkki <kmalkki@cc.hut.fi> and
  * Frodo Looijaard <frodol@dds.nl>, and also from Martin Bailey
  * <mbailey@littlefeet-inc.com>
diff --git a/drivers/i2c/algos/i2c-algo-pcf.h b/drivers/i2c/algos/i2c-algo-pcf.h
index 1ec703ee788d..262ee801975b 100644
--- a/drivers/i2c/algos/i2c-algo-pcf.h
+++ b/drivers/i2c/algos/i2c-algo-pcf.h
@@ -12,12 +12,7 @@
     This program is distributed in the hope that it will be useful,
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-    GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
-    MA 02110-1301 USA.							*/
+    GNU General Public License for more details.			*/
 /* --------------------------------------------------------------------	*/
 
 /* With some changes from Frodo Looijaard <frodol@dds.nl> */
diff --git a/drivers/i2c/busses/Kconfig b/drivers/i2c/busses/Kconfig
index 2ac87fa3058d..ab838d9e28b6 100644
--- a/drivers/i2c/busses/Kconfig
+++ b/drivers/i2c/busses/Kconfig
@@ -77,6 +77,16 @@ config I2C_AMD8111
 	  This driver can also be built as a module.  If so, the module
 	  will be called i2c-amd8111.
 
+config I2C_HIX5HD2
+	tristate "Hix5hd2 high-speed I2C driver"
+	depends on ARCH_HIX5HD2
+	help
+	  Say Y here to include support for high-speed I2C controller in the
+	  Hisilicon based hix5hd2 SoCs.
+
+	  This driver can also be built as a module.  If so, the module
+	  will be called i2c-hix5hd2.
+
 config I2C_I801
 	tristate "Intel 82801 (ICH/PCH)"
 	depends on PCI
@@ -112,6 +122,8 @@ config I2C_I801
 	    Wildcat Point (PCH)
 	    Wildcat Point-LP (PCH)
 	    BayTrail (SOC)
+	    Sunrise Point-H (PCH)
+	    Sunrise Point-LP (PCH)
 
 	  This driver can also be built as a module.  If so, the module
 	  will be called i2c-i801.
@@ -337,6 +349,17 @@ config I2C_AU1550
 	  This driver can also be built as a module.  If so, the module
 	  will be called i2c-au1550.
 
+config I2C_AXXIA
+	tristate "Axxia I2C controller"
+	depends on ARCH_AXXIA || COMPILE_TEST
+	default ARCH_AXXIA
+	help
+	  Say yes if you want to support the I2C bus on Axxia platforms.
+
+	  Please note that this controller is limited to transfers of maximum
+	  255 bytes in length. Any attempt to to a larger transfer will return
+	  an error.
+
 config I2C_BCM2835
 	tristate "Broadcom BCM2835 I2C controller"
 	depends on ARCH_BCM2835
@@ -357,7 +380,7 @@ config I2C_BCM_KONA
 	  If you say yes to this option, support will be included for the
 	  I2C interface on the Broadcom Kona family of processors.
 
-	  If you do not need KONA I2C inteface, say N.
+	  If you do not need KONA I2C interface, say N.
 
 config I2C_BLACKFIN_TWI
 	tristate "Blackfin TWI I2C support"
@@ -423,6 +446,7 @@ config I2C_DESIGNWARE_CORE
 config I2C_DESIGNWARE_PLATFORM
 	tristate "Synopsys DesignWare Platform"
 	select I2C_DESIGNWARE_CORE
+	depends on (ACPI && COMMON_CLK) || !ACPI
 	help
 	  If you say yes to this option, support will be included for the
 	  Synopsys DesignWare I2C adapter. Only master mode is supported.
@@ -465,7 +489,7 @@ config I2C_EG20T
 
 config I2C_EXYNOS5
 	tristate "Exynos5 high-speed I2C driver"
-	depends on ARCH_EXYNOS5 && OF
+	depends on ARCH_EXYNOS && OF
 	default y
 	help
 	  High-speed I2C controller on Exynos5 based Samsung SoCs.
@@ -500,6 +524,16 @@ config I2C_IBM_IIC
 	  This driver can also be built as a module.  If so, the module
 	  will be called i2c-ibm_iic.
 
+config I2C_IMG
+	tristate "Imagination Technologies I2C SCB Controller"
+	depends on MIPS || METAG || COMPILE_TEST
+	help
+	  Say Y here if you want to use the IMG I2C SCB controller,
+	  available on the TZ1090 and other IMG SoCs.
+
+	  This driver can also be built as a module.  If so, the module
+	  will be called i2c-img-scb.
+
 config I2C_IMX
 	tristate "IMX I2C interface"
 	depends on ARCH_MXC
@@ -530,6 +564,13 @@ config I2C_KEMPLD
 	  This driver can also be built as a module. If so, the module
 	  will be called i2c-kempld.
 
+config I2C_MESON
+	tristate "Amlogic Meson I2C controller"
+	depends on ARCH_MESON
+	help
+	  If you say yes to this option, support will be included for the
+	  I2C interface on the Amlogic Meson family of SoCs.
+
 config I2C_MPC
 	tristate "MPC107/824x/85xx/512x/52xx/83xx/86xx"
 	depends on PPC
@@ -679,7 +720,7 @@ config I2C_RIIC
 
 config I2C_RK3X
 	tristate "Rockchip RK3xxx I2C adapter"
-	depends on OF
+	depends on OF && COMMON_CLK
 	help
 	  Say Y here to include support for the I2C adapter in Rockchip RK3xxx
 	  SoCs.
@@ -712,6 +753,7 @@ config I2C_SH7760
 
 config I2C_SH_MOBILE
 	tristate "SuperH Mobile I2C Controller"
+	depends on HAS_DMA
 	depends on SUPERH || ARCH_SHMOBILE || COMPILE_TEST
 	help
 	  If you say yes to this option, support will be included for the
@@ -839,6 +881,7 @@ config I2C_XLR
 config I2C_RCAR
 	tristate "Renesas R-Car I2C Controller"
 	depends on ARCH_SHMOBILE || COMPILE_TEST
+	select I2C_SLAVE
 	help
 	  If you say yes to this option, support will be included for the
 	  R-Car I2C controller.
@@ -858,6 +901,16 @@ config I2C_DIOLAN_U2C
 	  This driver can also be built as a module.  If so, the module
 	  will be called i2c-diolan-u2c.
 
+config I2C_DLN2
+       tristate "Diolan DLN-2 USB I2C adapter"
+       depends on MFD_DLN2
+       help
+	 If you say yes to this option, support will be included for Diolan
+	 DLN2, a USB to I2C interface.
+
+	 This driver can also be built as a module.  If so, the module
+	 will be called i2c-dln2.
+
 config I2C_PARPORT
 	tristate "Parallel port adapter"
 	depends on PARPORT
@@ -1021,4 +1074,15 @@ config SCx200_ACB
 	  This support is also available as a module.  If so, the module
 	  will be called scx200_acb.
 
+config I2C_OPAL
+	tristate "IBM OPAL I2C driver"
+	depends on PPC_POWERNV
+	default y
+	help
+	  This exposes the PowerNV platform i2c busses to the linux i2c layer,
+	  the driver is based on the OPAL interfaces.
+
+	  This driver can also be built as a module. If so, the module will be
+	  called as i2c-opal.
+
 endmenu
diff --git a/drivers/i2c/busses/Makefile b/drivers/i2c/busses/Makefile
index 49bf07e5ef4d..56388f658d2f 100644
--- a/drivers/i2c/busses/Makefile
+++ b/drivers/i2c/busses/Makefile
@@ -31,6 +31,7 @@ obj-$(CONFIG_I2C_POWERMAC)	+= i2c-powermac.o
 # Embedded system I2C/SMBus host controller drivers
 obj-$(CONFIG_I2C_AT91)		+= i2c-at91.o
 obj-$(CONFIG_I2C_AU1550)	+= i2c-au1550.o
+obj-$(CONFIG_I2C_AXXIA)		+= i2c-axxia.o
 obj-$(CONFIG_I2C_BCM2835)	+= i2c-bcm2835.o
 obj-$(CONFIG_I2C_BLACKFIN_TWI)	+= i2c-bfin-twi.o
 obj-$(CONFIG_I2C_CADENCE)	+= i2c-cadence.o
@@ -47,10 +48,13 @@ obj-$(CONFIG_I2C_EG20T)		+= i2c-eg20t.o
 obj-$(CONFIG_I2C_EXYNOS5)	+= i2c-exynos5.o
 obj-$(CONFIG_I2C_GPIO)		+= i2c-gpio.o
 obj-$(CONFIG_I2C_HIGHLANDER)	+= i2c-highlander.o
+obj-$(CONFIG_I2C_HIX5HD2)	+= i2c-hix5hd2.o
 obj-$(CONFIG_I2C_IBM_IIC)	+= i2c-ibm_iic.o
+obj-$(CONFIG_I2C_IMG)		+= i2c-img-scb.o
 obj-$(CONFIG_I2C_IMX)		+= i2c-imx.o
 obj-$(CONFIG_I2C_IOP3XX)	+= i2c-iop3xx.o
 obj-$(CONFIG_I2C_KEMPLD)	+= i2c-kempld.o
+obj-$(CONFIG_I2C_MESON)		+= i2c-meson.o
 obj-$(CONFIG_I2C_MPC)		+= i2c-mpc.o
 obj-$(CONFIG_I2C_MV64XXX)	+= i2c-mv64xxx.o
 obj-$(CONFIG_I2C_MXS)		+= i2c-mxs.o
@@ -85,6 +89,7 @@ obj-$(CONFIG_I2C_RCAR)		+= i2c-rcar.o
 
 # External I2C/SMBus adapter drivers
 obj-$(CONFIG_I2C_DIOLAN_U2C)	+= i2c-diolan-u2c.o
+obj-$(CONFIG_I2C_DLN2)		+= i2c-dln2.o
 obj-$(CONFIG_I2C_PARPORT)	+= i2c-parport.o
 obj-$(CONFIG_I2C_PARPORT_LIGHT)	+= i2c-parport-light.o
 obj-$(CONFIG_I2C_ROBOTFUZZ_OSIF)	+= i2c-robotfuzz-osif.o
@@ -97,6 +102,7 @@ obj-$(CONFIG_I2C_ACORN)		+= i2c-acorn.o
 obj-$(CONFIG_I2C_BCM_KONA)	+= i2c-bcm-kona.o
 obj-$(CONFIG_I2C_CROS_EC_TUNNEL)	+= i2c-cros-ec-tunnel.o
 obj-$(CONFIG_I2C_ELEKTOR)	+= i2c-elektor.o
+obj-$(CONFIG_I2C_OPAL)		+= i2c-opal.o
 obj-$(CONFIG_I2C_PCA_ISA)	+= i2c-pca-isa.o
 obj-$(CONFIG_I2C_SIBYTE)	+= i2c-sibyte.o
 obj-$(CONFIG_SCx200_ACB)	+= scx200_acb.o
diff --git a/drivers/i2c/busses/i2c-ali1535.c b/drivers/i2c/busses/i2c-ali1535.c
index 451e305f7971..4f2d78868281 100644
--- a/drivers/i2c/busses/i2c-ali1535.c
+++ b/drivers/i2c/busses/i2c-ali1535.c
@@ -14,10 +14,6 @@
  *  but WITHOUT ANY WARRANTY; without even the implied warranty of
  *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  *  GNU General Public License for more details.
- *
- *  You should have received a copy of the GNU General Public License
- *  along with this program; if not, write to the Free Software
- *  Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 /*
diff --git a/drivers/i2c/busses/i2c-ali15x3.c b/drivers/i2c/busses/i2c-ali15x3.c
index 2fa21ce9682b..45c5c4883022 100644
--- a/drivers/i2c/busses/i2c-ali15x3.c
+++ b/drivers/i2c/busses/i2c-ali15x3.c
@@ -12,10 +12,6 @@
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
     GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 /*
diff --git a/drivers/i2c/busses/i2c-amd756-s4882.c b/drivers/i2c/busses/i2c-amd756-s4882.c
index 41fc6837fb8b..65e324054970 100644
--- a/drivers/i2c/busses/i2c-amd756-s4882.c
+++ b/drivers/i2c/busses/i2c-amd756-s4882.c
@@ -12,10 +12,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  */
  
 /*
diff --git a/drivers/i2c/busses/i2c-amd756.c b/drivers/i2c/busses/i2c-amd756.c
index a16f72891358..6c7113d990f8 100644
--- a/drivers/i2c/busses/i2c-amd756.c
+++ b/drivers/i2c/busses/i2c-amd756.c
@@ -15,10 +15,6 @@
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
     GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 /*
diff --git a/drivers/i2c/busses/i2c-at91.c b/drivers/i2c/busses/i2c-at91.c
index 917d54588d95..636fd2efad88 100644
--- a/drivers/i2c/busses/i2c-at91.c
+++ b/drivers/i2c/busses/i2c-at91.c
@@ -31,10 +31,13 @@
 #include <linux/platform_device.h>
 #include <linux/slab.h>
 #include <linux/platform_data/dma-atmel.h>
+#include <linux/pm_runtime.h>
+#include <linux/pinctrl/consumer.h>
 
 #define DEFAULT_TWI_CLK_HZ		100000		/* max 400 Kbits/s */
 #define AT91_I2C_TIMEOUT	msecs_to_jiffies(100)	/* transfer timeout */
 #define AT91_I2C_DMA_THRESHOLD	8			/* enable DMA if transfer size is bigger than this threshold */
+#define AUTOSUSPEND_TIMEOUT		2000
 
 /* AT91 TWI register definitions */
 #define	AT91_TWI_CR		0x0000	/* Control Register */
@@ -72,7 +75,6 @@ struct at91_twi_pdata {
 	unsigned clk_max_div;
 	unsigned clk_offset;
 	bool has_unre_flag;
-	bool has_dma_support;
 	struct at_dma_slave dma_slave;
 };
 
@@ -434,7 +436,7 @@ static int at91_do_twi_transfer(struct at91_twi_dev *dev)
 		}
 	}
 
-	ret = wait_for_completion_io_timeout(&dev->cmd_complete,
+	ret = wait_for_completion_timeout(&dev->cmd_complete,
 					     dev->adapter.timeout);
 	if (ret == 0) {
 		dev_err(dev->dev, "controller timed out\n");
@@ -481,6 +483,10 @@ static int at91_twi_xfer(struct i2c_adapter *adap, struct i2c_msg *msg, int num)
 
 	dev_dbg(&adap->dev, "at91_xfer: processing %d messages:\n", num);
 
+	ret = pm_runtime_get_sync(dev->dev);
+	if (ret < 0)
+		goto out;
+
 	/*
 	 * The hardware can handle at most two messages concatenated by a
 	 * repeated start via it's internal address feature.
@@ -488,18 +494,21 @@ static int at91_twi_xfer(struct i2c_adapter *adap, struct i2c_msg *msg, int num)
 	if (num > 2) {
 		dev_err(dev->dev,
 			"cannot handle more than two concatenated messages.\n");
-		return 0;
+		ret = 0;
+		goto out;
 	} else if (num == 2) {
 		int internal_address = 0;
 		int i;
 
 		if (msg->flags & I2C_M_RD) {
 			dev_err(dev->dev, "first transfer must be write.\n");
-			return -EINVAL;
+			ret = -EINVAL;
+			goto out;
 		}
 		if (msg->len > 3) {
 			dev_err(dev->dev, "first message size must be <= 3.\n");
-			return -EINVAL;
+			ret = -EINVAL;
+			goto out;
 		}
 
 		/* 1st msg is put into the internal address, start with 2nd */
@@ -523,7 +532,12 @@ static int at91_twi_xfer(struct i2c_adapter *adap, struct i2c_msg *msg, int num)
 
 	ret = at91_do_twi_transfer(dev);
 
-	return (ret < 0) ? ret : num;
+	ret = (ret < 0) ? ret : num;
+out:
+	pm_runtime_mark_last_busy(dev->dev);
+	pm_runtime_put_autosuspend(dev->dev);
+
+	return ret;
 }
 
 static u32 at91_twi_func(struct i2c_adapter *adapter)
@@ -541,35 +555,30 @@ static struct at91_twi_pdata at91rm9200_config = {
 	.clk_max_div = 5,
 	.clk_offset = 3,
 	.has_unre_flag = true,
-	.has_dma_support = false,
 };
 
 static struct at91_twi_pdata at91sam9261_config = {
 	.clk_max_div = 5,
 	.clk_offset = 4,
 	.has_unre_flag = false,
-	.has_dma_support = false,
 };
 
 static struct at91_twi_pdata at91sam9260_config = {
 	.clk_max_div = 7,
 	.clk_offset = 4,
 	.has_unre_flag = false,
-	.has_dma_support = false,
 };
 
 static struct at91_twi_pdata at91sam9g20_config = {
 	.clk_max_div = 7,
 	.clk_offset = 4,
 	.has_unre_flag = false,
-	.has_dma_support = false,
 };
 
 static struct at91_twi_pdata at91sam9g10_config = {
 	.clk_max_div = 7,
 	.clk_offset = 4,
 	.has_unre_flag = false,
-	.has_dma_support = false,
 };
 
 static const struct platform_device_id at91_twi_devtypes[] = {
@@ -598,7 +607,6 @@ static struct at91_twi_pdata at91sam9x5_config = {
 	.clk_max_div = 7,
 	.clk_offset = 4,
 	.has_unre_flag = false,
-	.has_dma_support = true,
 };
 
 static const struct of_device_id atmel_twi_dt_ids[] = {
@@ -627,30 +635,11 @@ static const struct of_device_id atmel_twi_dt_ids[] = {
 MODULE_DEVICE_TABLE(of, atmel_twi_dt_ids);
 #endif
 
-static bool filter(struct dma_chan *chan, void *pdata)
-{
-	struct at91_twi_pdata *sl_pdata = pdata;
-	struct at_dma_slave *sl;
-
-	if (!sl_pdata)
-		return false;
-
-	sl = &sl_pdata->dma_slave;
-	if (sl && (sl->dma_dev == chan->device->dev)) {
-		chan->private = sl;
-		return true;
-	} else {
-		return false;
-	}
-}
-
 static int at91_twi_configure_dma(struct at91_twi_dev *dev, u32 phy_addr)
 {
 	int ret = 0;
-	struct at91_twi_pdata *pdata = dev->pdata;
 	struct dma_slave_config slave_config;
 	struct at91_twi_dma *dma = &dev->dma;
-	dma_cap_mask_t mask;
 
 	memset(&slave_config, 0, sizeof(slave_config));
 	slave_config.src_addr = (dma_addr_t)phy_addr + AT91_TWI_RHR;
@@ -661,22 +650,17 @@ static int at91_twi_configure_dma(struct at91_twi_dev *dev, u32 phy_addr)
 	slave_config.dst_maxburst = 1;
 	slave_config.device_fc = false;
 
-	dma_cap_zero(mask);
-	dma_cap_set(DMA_SLAVE, mask);
-
-	dma->chan_tx = dma_request_slave_channel_compat(mask, filter, pdata,
-							dev->dev, "tx");
-	if (!dma->chan_tx) {
-		dev_err(dev->dev, "can't get a DMA channel for tx\n");
-		ret = -EBUSY;
+	dma->chan_tx = dma_request_slave_channel_reason(dev->dev, "tx");
+	if (IS_ERR(dma->chan_tx)) {
+		ret = PTR_ERR(dma->chan_tx);
+		dma->chan_tx = NULL;
 		goto error;
 	}
 
-	dma->chan_rx = dma_request_slave_channel_compat(mask, filter, pdata,
-							dev->dev, "rx");
-	if (!dma->chan_rx) {
-		dev_err(dev->dev, "can't get a DMA channel for rx\n");
-		ret = -EBUSY;
+	dma->chan_rx = dma_request_slave_channel_reason(dev->dev, "rx");
+	if (IS_ERR(dma->chan_rx)) {
+		ret = PTR_ERR(dma->chan_rx);
+		dma->chan_rx = NULL;
 		goto error;
 	}
 
@@ -697,6 +681,7 @@ static int at91_twi_configure_dma(struct at91_twi_dev *dev, u32 phy_addr)
 	sg_init_table(&dma->sg, 1);
 	dma->buf_mapped = false;
 	dma->xfer_in_progress = false;
+	dev->use_dma = true;
 
 	dev_info(dev->dev, "using %s (tx) and %s (rx) for DMA transfers\n",
 		 dma_chan_name(dma->chan_tx), dma_chan_name(dma->chan_rx));
@@ -704,7 +689,8 @@ static int at91_twi_configure_dma(struct at91_twi_dev *dev, u32 phy_addr)
 	return ret;
 
 error:
-	dev_info(dev->dev, "can't use DMA\n");
+	if (ret != -EPROBE_DEFER)
+		dev_info(dev->dev, "can't use DMA, error %d\n", ret);
 	if (dma->chan_rx)
 		dma_release_channel(dma->chan_rx);
 	if (dma->chan_tx)
@@ -772,9 +758,10 @@ static int at91_twi_probe(struct platform_device *pdev)
 	}
 	clk_prepare_enable(dev->clk);
 
-	if (dev->pdata->has_dma_support) {
-		if (at91_twi_configure_dma(dev, phy_addr) == 0)
-			dev->use_dma = true;
+	if (dev->dev->of_node) {
+		rc = at91_twi_configure_dma(dev, phy_addr);
+		if (rc == -EPROBE_DEFER)
+			return rc;
 	}
 
 	rc = of_property_read_u32(dev->dev->of_node, "clock-frequency",
@@ -795,11 +782,20 @@ static int at91_twi_probe(struct platform_device *pdev)
 	dev->adapter.timeout = AT91_I2C_TIMEOUT;
 	dev->adapter.dev.of_node = pdev->dev.of_node;
 
+	pm_runtime_set_autosuspend_delay(dev->dev, AUTOSUSPEND_TIMEOUT);
+	pm_runtime_use_autosuspend(dev->dev);
+	pm_runtime_set_active(dev->dev);
+	pm_runtime_enable(dev->dev);
+
 	rc = i2c_add_numbered_adapter(&dev->adapter);
 	if (rc) {
 		dev_err(dev->dev, "Adapter %s registration failed\n",
 			dev->adapter.name);
 		clk_disable_unprepare(dev->clk);
+
+		pm_runtime_disable(dev->dev);
+		pm_runtime_set_suspended(dev->dev);
+
 		return rc;
 	}
 
@@ -814,6 +810,9 @@ static int at91_twi_remove(struct platform_device *pdev)
 	i2c_del_adapter(&dev->adapter);
 	clk_disable_unprepare(dev->clk);
 
+	pm_runtime_disable(dev->dev);
+	pm_runtime_set_suspended(dev->dev);
+
 	return 0;
 }
 
@@ -823,7 +822,9 @@ static int at91_twi_runtime_suspend(struct device *dev)
 {
 	struct at91_twi_dev *twi_dev = dev_get_drvdata(dev);
 
-	clk_disable(twi_dev->clk);
+	clk_disable_unprepare(twi_dev->clk);
+
+	pinctrl_pm_select_sleep_state(dev);
 
 	return 0;
 }
@@ -832,10 +833,38 @@ static int at91_twi_runtime_resume(struct device *dev)
 {
 	struct at91_twi_dev *twi_dev = dev_get_drvdata(dev);
 
-	return clk_enable(twi_dev->clk);
+	pinctrl_pm_select_default_state(dev);
+
+	return clk_prepare_enable(twi_dev->clk);
+}
+
+static int at91_twi_suspend_noirq(struct device *dev)
+{
+	if (!pm_runtime_status_suspended(dev))
+		at91_twi_runtime_suspend(dev);
+
+	return 0;
+}
+
+static int at91_twi_resume_noirq(struct device *dev)
+{
+	int ret;
+
+	if (!pm_runtime_status_suspended(dev)) {
+		ret = at91_twi_runtime_resume(dev);
+		if (ret)
+			return ret;
+	}
+
+	pm_runtime_mark_last_busy(dev);
+	pm_request_autosuspend(dev);
+
+	return 0;
 }
 
 static const struct dev_pm_ops at91_twi_pm = {
+	.suspend_noirq	= at91_twi_suspend_noirq,
+	.resume_noirq	= at91_twi_resume_noirq,
 	.runtime_suspend	= at91_twi_runtime_suspend,
 	.runtime_resume		= at91_twi_runtime_resume,
 };
@@ -851,7 +880,6 @@ static struct platform_driver at91_twi_driver = {
 	.id_table	= at91_twi_devtypes,
 	.driver		= {
 		.name	= "at91_i2c",
-		.owner	= THIS_MODULE,
 		.of_match_table = of_match_ptr(atmel_twi_dt_ids),
 		.pm	= at91_twi_pm_ops,
 	},
diff --git a/drivers/i2c/busses/i2c-au1550.c b/drivers/i2c/busses/i2c-au1550.c
index 8762458ca7da..a6aae84e5706 100644
--- a/drivers/i2c/busses/i2c-au1550.c
+++ b/drivers/i2c/busses/i2c-au1550.c
@@ -21,10 +21,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
  */
 
 #include <linux/delay.h>
@@ -417,7 +413,6 @@ static const struct dev_pm_ops i2c_au1550_pmops = {
 static struct platform_driver au1xpsc_smbus_driver = {
 	.driver = {
 		.name	= "au1xpsc_smbus",
-		.owner	= THIS_MODULE,
 		.pm	= AU1XPSC_SMBUS_PMOPS,
 	},
 	.probe		= i2c_au1550_probe,
diff --git a/drivers/i2c/busses/i2c-axxia.c b/drivers/i2c/busses/i2c-axxia.c
new file mode 100644
index 000000000000..768a598d8d03
--- /dev/null
+++ b/drivers/i2c/busses/i2c-axxia.c
@@ -0,0 +1,559 @@
+/*
+ * This driver implements I2C master functionality using the LSI API2C
+ * controller.
+ *
+ * NOTE: The controller has a limitation in that it can only do transfers of
+ * maximum 255 bytes at a time. If a larger transfer is attempted, error code
+ * (-EINVAL) is returned.
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ */
+#include <linux/clk.h>
+#include <linux/clkdev.h>
+#include <linux/err.h>
+#include <linux/i2c.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/platform_device.h>
+
+#define SCL_WAIT_TIMEOUT_NS 25000000
+#define I2C_XFER_TIMEOUT    (msecs_to_jiffies(250))
+#define I2C_STOP_TIMEOUT    (msecs_to_jiffies(100))
+#define FIFO_SIZE           8
+
+#define GLOBAL_CONTROL		0x00
+#define   GLOBAL_MST_EN         BIT(0)
+#define   GLOBAL_SLV_EN         BIT(1)
+#define   GLOBAL_IBML_EN        BIT(2)
+#define INTERRUPT_STATUS	0x04
+#define INTERRUPT_ENABLE	0x08
+#define   INT_SLV               BIT(1)
+#define   INT_MST               BIT(0)
+#define WAIT_TIMER_CONTROL	0x0c
+#define   WT_EN			BIT(15)
+#define   WT_VALUE(_x)		((_x) & 0x7fff)
+#define IBML_TIMEOUT		0x10
+#define IBML_LOW_MEXT		0x14
+#define IBML_LOW_SEXT		0x18
+#define TIMER_CLOCK_DIV		0x1c
+#define I2C_BUS_MONITOR		0x20
+#define SOFT_RESET		0x24
+#define MST_COMMAND		0x28
+#define   CMD_BUSY		(1<<3)
+#define   CMD_MANUAL		(0x00 | CMD_BUSY)
+#define   CMD_AUTO		(0x01 | CMD_BUSY)
+#define MST_RX_XFER		0x2c
+#define MST_TX_XFER		0x30
+#define MST_ADDR_1		0x34
+#define MST_ADDR_2		0x38
+#define MST_DATA		0x3c
+#define MST_TX_FIFO		0x40
+#define MST_RX_FIFO		0x44
+#define MST_INT_ENABLE		0x48
+#define MST_INT_STATUS		0x4c
+#define   MST_STATUS_RFL	(1 << 13) /* RX FIFO serivce */
+#define   MST_STATUS_TFL	(1 << 12) /* TX FIFO service */
+#define   MST_STATUS_SNS	(1 << 11) /* Manual mode done */
+#define   MST_STATUS_SS		(1 << 10) /* Automatic mode done */
+#define   MST_STATUS_SCC	(1 << 9)  /* Stop complete */
+#define   MST_STATUS_IP		(1 << 8)  /* Invalid parameter */
+#define   MST_STATUS_TSS	(1 << 7)  /* Timeout */
+#define   MST_STATUS_AL		(1 << 6)  /* Arbitration lost */
+#define   MST_STATUS_ND		(1 << 5)  /* NAK on data phase */
+#define   MST_STATUS_NA		(1 << 4)  /* NAK on address phase */
+#define   MST_STATUS_NAK	(MST_STATUS_NA | \
+				 MST_STATUS_ND)
+#define   MST_STATUS_ERR	(MST_STATUS_NAK | \
+				 MST_STATUS_AL  | \
+				 MST_STATUS_IP  | \
+				 MST_STATUS_TSS)
+#define MST_TX_BYTES_XFRD	0x50
+#define MST_RX_BYTES_XFRD	0x54
+#define SCL_HIGH_PERIOD		0x80
+#define SCL_LOW_PERIOD		0x84
+#define SPIKE_FLTR_LEN		0x88
+#define SDA_SETUP_TIME		0x8c
+#define SDA_HOLD_TIME		0x90
+
+/**
+ * axxia_i2c_dev - I2C device context
+ * @base: pointer to register struct
+ * @msg: pointer to current message
+ * @msg_xfrd: number of bytes transferred in msg
+ * @msg_err: error code for completed message
+ * @msg_complete: xfer completion object
+ * @dev: device reference
+ * @adapter: core i2c abstraction
+ * @i2c_clk: clock reference for i2c input clock
+ * @bus_clk_rate: current i2c bus clock rate
+ */
+struct axxia_i2c_dev {
+	void __iomem *base;
+	struct i2c_msg *msg;
+	size_t msg_xfrd;
+	int msg_err;
+	struct completion msg_complete;
+	struct device *dev;
+	struct i2c_adapter adapter;
+	struct clk *i2c_clk;
+	u32 bus_clk_rate;
+};
+
+static void i2c_int_disable(struct axxia_i2c_dev *idev, u32 mask)
+{
+	u32 int_en;
+
+	int_en = readl(idev->base + MST_INT_ENABLE);
+	writel(int_en & ~mask, idev->base + MST_INT_ENABLE);
+}
+
+static void i2c_int_enable(struct axxia_i2c_dev *idev, u32 mask)
+{
+	u32 int_en;
+
+	int_en = readl(idev->base + MST_INT_ENABLE);
+	writel(int_en | mask, idev->base + MST_INT_ENABLE);
+}
+
+/**
+ * ns_to_clk - Convert time (ns) to clock cycles for the given clock frequency.
+ */
+static u32 ns_to_clk(u64 ns, u32 clk_mhz)
+{
+	return div_u64(ns * clk_mhz, 1000);
+}
+
+static int axxia_i2c_init(struct axxia_i2c_dev *idev)
+{
+	u32 divisor = clk_get_rate(idev->i2c_clk) / idev->bus_clk_rate;
+	u32 clk_mhz = clk_get_rate(idev->i2c_clk) / 1000000;
+	u32 t_setup;
+	u32 t_high, t_low;
+	u32 tmo_clk;
+	u32 prescale;
+	unsigned long timeout;
+
+	dev_dbg(idev->dev, "rate=%uHz per_clk=%uMHz -> ratio=1:%u\n",
+		idev->bus_clk_rate, clk_mhz, divisor);
+
+	/* Reset controller */
+	writel(0x01, idev->base + SOFT_RESET);
+	timeout = jiffies + msecs_to_jiffies(100);
+	while (readl(idev->base + SOFT_RESET) & 1) {
+		if (time_after(jiffies, timeout)) {
+			dev_warn(idev->dev, "Soft reset failed\n");
+			break;
+		}
+	}
+
+	/* Enable Master Mode */
+	writel(0x1, idev->base + GLOBAL_CONTROL);
+
+	if (idev->bus_clk_rate <= 100000) {
+		/* Standard mode SCL 50/50, tSU:DAT = 250 ns */
+		t_high = divisor * 1 / 2;
+		t_low = divisor * 1 / 2;
+		t_setup = ns_to_clk(250, clk_mhz);
+	} else {
+		/* Fast mode SCL 33/66, tSU:DAT = 100 ns */
+		t_high = divisor * 1 / 3;
+		t_low = divisor * 2 / 3;
+		t_setup = ns_to_clk(100, clk_mhz);
+	}
+
+	/* SCL High Time */
+	writel(t_high, idev->base + SCL_HIGH_PERIOD);
+	/* SCL Low Time */
+	writel(t_low, idev->base + SCL_LOW_PERIOD);
+	/* SDA Setup Time */
+	writel(t_setup, idev->base + SDA_SETUP_TIME);
+	/* SDA Hold Time, 300ns */
+	writel(ns_to_clk(300, clk_mhz), idev->base + SDA_HOLD_TIME);
+	/* Filter <50ns spikes */
+	writel(ns_to_clk(50, clk_mhz), idev->base + SPIKE_FLTR_LEN);
+
+	/* Configure Time-Out Registers */
+	tmo_clk = ns_to_clk(SCL_WAIT_TIMEOUT_NS, clk_mhz);
+
+	/* Find prescaler value that makes tmo_clk fit in 15-bits counter. */
+	for (prescale = 0; prescale < 15; ++prescale) {
+		if (tmo_clk <= 0x7fff)
+			break;
+		tmo_clk >>= 1;
+	}
+	if (tmo_clk > 0x7fff)
+		tmo_clk = 0x7fff;
+
+	/* Prescale divider (log2) */
+	writel(prescale, idev->base + TIMER_CLOCK_DIV);
+	/* Timeout in divided clocks */
+	writel(WT_EN | WT_VALUE(tmo_clk), idev->base + WAIT_TIMER_CONTROL);
+
+	/* Mask all master interrupt bits */
+	i2c_int_disable(idev, ~0);
+
+	/* Interrupt enable */
+	writel(0x01, idev->base + INTERRUPT_ENABLE);
+
+	return 0;
+}
+
+static int i2c_m_rd(const struct i2c_msg *msg)
+{
+	return (msg->flags & I2C_M_RD) != 0;
+}
+
+static int i2c_m_ten(const struct i2c_msg *msg)
+{
+	return (msg->flags & I2C_M_TEN) != 0;
+}
+
+static int i2c_m_recv_len(const struct i2c_msg *msg)
+{
+	return (msg->flags & I2C_M_RECV_LEN) != 0;
+}
+
+/**
+ * axxia_i2c_empty_rx_fifo - Fetch data from RX FIFO and update SMBus block
+ * transfer length if this is the first byte of such a transfer.
+ */
+static int axxia_i2c_empty_rx_fifo(struct axxia_i2c_dev *idev)
+{
+	struct i2c_msg *msg = idev->msg;
+	size_t rx_fifo_avail = readl(idev->base + MST_RX_FIFO);
+	int bytes_to_transfer = min(rx_fifo_avail, msg->len - idev->msg_xfrd);
+
+	while (bytes_to_transfer-- > 0) {
+		int c = readl(idev->base + MST_DATA);
+
+		if (idev->msg_xfrd == 0 && i2c_m_recv_len(msg)) {
+			/*
+			 * Check length byte for SMBus block read
+			 */
+			if (c <= 0 || c > I2C_SMBUS_BLOCK_MAX) {
+				idev->msg_err = -EPROTO;
+				i2c_int_disable(idev, ~0);
+				complete(&idev->msg_complete);
+				break;
+			}
+			msg->len = 1 + c;
+			writel(msg->len, idev->base + MST_RX_XFER);
+		}
+		msg->buf[idev->msg_xfrd++] = c;
+	}
+
+	return 0;
+}
+
+/**
+ * axxia_i2c_fill_tx_fifo - Fill TX FIFO from current message buffer.
+ * @return: Number of bytes left to transfer.
+ */
+static int axxia_i2c_fill_tx_fifo(struct axxia_i2c_dev *idev)
+{
+	struct i2c_msg *msg = idev->msg;
+	size_t tx_fifo_avail = FIFO_SIZE - readl(idev->base + MST_TX_FIFO);
+	int bytes_to_transfer = min(tx_fifo_avail, msg->len - idev->msg_xfrd);
+	int ret = msg->len - idev->msg_xfrd - bytes_to_transfer;
+
+	while (bytes_to_transfer-- > 0)
+		writel(msg->buf[idev->msg_xfrd++], idev->base + MST_DATA);
+
+	return ret;
+}
+
+static irqreturn_t axxia_i2c_isr(int irq, void *_dev)
+{
+	struct axxia_i2c_dev *idev = _dev;
+	u32 status;
+
+	if (!(readl(idev->base + INTERRUPT_STATUS) & INT_MST))
+		return IRQ_NONE;
+
+	/* Read interrupt status bits */
+	status = readl(idev->base + MST_INT_STATUS);
+
+	if (!idev->msg) {
+		dev_warn(idev->dev, "unexpected interrupt\n");
+		goto out;
+	}
+
+	/* RX FIFO needs service? */
+	if (i2c_m_rd(idev->msg) && (status & MST_STATUS_RFL))
+		axxia_i2c_empty_rx_fifo(idev);
+
+	/* TX FIFO needs service? */
+	if (!i2c_m_rd(idev->msg) && (status & MST_STATUS_TFL)) {
+		if (axxia_i2c_fill_tx_fifo(idev) == 0)
+			i2c_int_disable(idev, MST_STATUS_TFL);
+	}
+
+	if (status & MST_STATUS_SCC) {
+		/* Stop completed */
+		i2c_int_disable(idev, ~0);
+		complete(&idev->msg_complete);
+	} else if (status & MST_STATUS_SNS) {
+		/* Transfer done */
+		i2c_int_disable(idev, ~0);
+		if (i2c_m_rd(idev->msg) && idev->msg_xfrd < idev->msg->len)
+			axxia_i2c_empty_rx_fifo(idev);
+		complete(&idev->msg_complete);
+	} else if (unlikely(status & MST_STATUS_ERR)) {
+		/* Transfer error */
+		i2c_int_disable(idev, ~0);
+		if (status & MST_STATUS_AL)
+			idev->msg_err = -EAGAIN;
+		else if (status & MST_STATUS_NAK)
+			idev->msg_err = -ENXIO;
+		else
+			idev->msg_err = -EIO;
+		dev_dbg(idev->dev, "error %#x, addr=%#x rx=%u/%u tx=%u/%u\n",
+			status,
+			idev->msg->addr,
+			readl(idev->base + MST_RX_BYTES_XFRD),
+			readl(idev->base + MST_RX_XFER),
+			readl(idev->base + MST_TX_BYTES_XFRD),
+			readl(idev->base + MST_TX_XFER));
+		complete(&idev->msg_complete);
+	}
+
+out:
+	/* Clear interrupt */
+	writel(INT_MST, idev->base + INTERRUPT_STATUS);
+
+	return IRQ_HANDLED;
+}
+
+static int axxia_i2c_xfer_msg(struct axxia_i2c_dev *idev, struct i2c_msg *msg)
+{
+	u32 int_mask = MST_STATUS_ERR | MST_STATUS_SNS;
+	u32 rx_xfer, tx_xfer;
+	u32 addr_1, addr_2;
+	int ret;
+
+	if (msg->len > 255) {
+		dev_warn(idev->dev, "unsupported length %u\n", msg->len);
+		return -EINVAL;
+	}
+
+	idev->msg = msg;
+	idev->msg_xfrd = 0;
+	idev->msg_err = 0;
+	reinit_completion(&idev->msg_complete);
+
+	if (i2c_m_ten(msg)) {
+		/* 10-bit address
+		 *   addr_1: 5'b11110 | addr[9:8] | (R/nW)
+		 *   addr_2: addr[7:0]
+		 */
+		addr_1 = 0xF0 | ((msg->addr >> 7) & 0x06);
+		addr_2 = msg->addr & 0xFF;
+	} else {
+		/* 7-bit address
+		 *   addr_1: addr[6:0] | (R/nW)
+		 *   addr_2: dont care
+		 */
+		addr_1 = (msg->addr << 1) & 0xFF;
+		addr_2 = 0;
+	}
+
+	if (i2c_m_rd(msg)) {
+		/* I2C read transfer */
+		rx_xfer = i2c_m_recv_len(msg) ? I2C_SMBUS_BLOCK_MAX : msg->len;
+		tx_xfer = 0;
+		addr_1 |= 1;	/* Set the R/nW bit of the address */
+	} else {
+		/* I2C write transfer */
+		rx_xfer = 0;
+		tx_xfer = msg->len;
+	}
+
+	writel(rx_xfer, idev->base + MST_RX_XFER);
+	writel(tx_xfer, idev->base + MST_TX_XFER);
+	writel(addr_1, idev->base + MST_ADDR_1);
+	writel(addr_2, idev->base + MST_ADDR_2);
+
+	if (i2c_m_rd(msg))
+		int_mask |= MST_STATUS_RFL;
+	else if (axxia_i2c_fill_tx_fifo(idev) != 0)
+		int_mask |= MST_STATUS_TFL;
+
+	/* Start manual mode */
+	writel(CMD_MANUAL, idev->base + MST_COMMAND);
+
+	i2c_int_enable(idev, int_mask);
+
+	ret = wait_for_completion_timeout(&idev->msg_complete,
+					  I2C_XFER_TIMEOUT);
+
+	i2c_int_disable(idev, int_mask);
+
+	if (readl(idev->base + MST_COMMAND) & CMD_BUSY)
+		dev_warn(idev->dev, "busy after xfer\n");
+
+	if (ret == 0)
+		idev->msg_err = -ETIMEDOUT;
+
+	if (unlikely(idev->msg_err) && idev->msg_err != -ENXIO)
+		axxia_i2c_init(idev);
+
+	return idev->msg_err;
+}
+
+static int axxia_i2c_stop(struct axxia_i2c_dev *idev)
+{
+	u32 int_mask = MST_STATUS_ERR | MST_STATUS_SCC;
+	int ret;
+
+	reinit_completion(&idev->msg_complete);
+
+	/* Issue stop */
+	writel(0xb, idev->base + MST_COMMAND);
+	i2c_int_enable(idev, int_mask);
+	ret = wait_for_completion_timeout(&idev->msg_complete,
+					  I2C_STOP_TIMEOUT);
+	i2c_int_disable(idev, int_mask);
+	if (ret == 0)
+		return -ETIMEDOUT;
+
+	if (readl(idev->base + MST_COMMAND) & CMD_BUSY)
+		dev_warn(idev->dev, "busy after stop\n");
+
+	return 0;
+}
+
+static int
+axxia_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg msgs[], int num)
+{
+	struct axxia_i2c_dev *idev = i2c_get_adapdata(adap);
+	int i;
+	int ret = 0;
+
+	for (i = 0; ret == 0 && i < num; ++i)
+		ret = axxia_i2c_xfer_msg(idev, &msgs[i]);
+
+	axxia_i2c_stop(idev);
+
+	return ret ? : i;
+}
+
+static u32 axxia_i2c_func(struct i2c_adapter *adap)
+{
+	u32 caps = (I2C_FUNC_I2C | I2C_FUNC_10BIT_ADDR |
+		    I2C_FUNC_SMBUS_EMUL | I2C_FUNC_SMBUS_BLOCK_DATA);
+	return caps;
+}
+
+static const struct i2c_algorithm axxia_i2c_algo = {
+	.master_xfer = axxia_i2c_xfer,
+	.functionality = axxia_i2c_func,
+};
+
+static int axxia_i2c_probe(struct platform_device *pdev)
+{
+	struct device_node *np = pdev->dev.of_node;
+	struct axxia_i2c_dev *idev = NULL;
+	struct resource *res;
+	void __iomem *base;
+	int irq;
+	int ret = 0;
+
+	idev = devm_kzalloc(&pdev->dev, sizeof(*idev), GFP_KERNEL);
+	if (!idev)
+		return -ENOMEM;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(base))
+		return PTR_ERR(base);
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(&pdev->dev, "missing interrupt resource\n");
+		return irq;
+	}
+
+	idev->i2c_clk = devm_clk_get(&pdev->dev, "i2c");
+	if (IS_ERR(idev->i2c_clk)) {
+		dev_err(&pdev->dev, "missing clock\n");
+		return PTR_ERR(idev->i2c_clk);
+	}
+
+	idev->base = base;
+	idev->dev = &pdev->dev;
+	init_completion(&idev->msg_complete);
+
+	of_property_read_u32(np, "clock-frequency", &idev->bus_clk_rate);
+	if (idev->bus_clk_rate == 0)
+		idev->bus_clk_rate = 100000;	/* default clock rate */
+
+	ret = axxia_i2c_init(idev);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to initialize\n");
+		return ret;
+	}
+
+	ret = devm_request_irq(&pdev->dev, irq, axxia_i2c_isr, 0,
+			       pdev->name, idev);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to claim IRQ%d\n", irq);
+		return ret;
+	}
+
+	clk_prepare_enable(idev->i2c_clk);
+
+	i2c_set_adapdata(&idev->adapter, idev);
+	strlcpy(idev->adapter.name, pdev->name, sizeof(idev->adapter.name));
+	idev->adapter.owner = THIS_MODULE;
+	idev->adapter.algo = &axxia_i2c_algo;
+	idev->adapter.dev.parent = &pdev->dev;
+	idev->adapter.dev.of_node = pdev->dev.of_node;
+
+	platform_set_drvdata(pdev, idev);
+
+	ret = i2c_add_adapter(&idev->adapter);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to add adapter\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int axxia_i2c_remove(struct platform_device *pdev)
+{
+	struct axxia_i2c_dev *idev = platform_get_drvdata(pdev);
+
+	clk_disable_unprepare(idev->i2c_clk);
+	i2c_del_adapter(&idev->adapter);
+
+	return 0;
+}
+
+/* Match table for of_platform binding */
+static const struct of_device_id axxia_i2c_of_match[] = {
+	{ .compatible = "lsi,api2c", },
+	{},
+};
+
+MODULE_DEVICE_TABLE(of, axxia_i2c_of_match);
+
+static struct platform_driver axxia_i2c_driver = {
+	.probe = axxia_i2c_probe,
+	.remove = axxia_i2c_remove,
+	.driver = {
+		.name = "axxia-i2c",
+		.of_match_table = axxia_i2c_of_match,
+	},
+};
+
+module_platform_driver(axxia_i2c_driver);
+
+MODULE_DESCRIPTION("Axxia I2C Bus driver");
+MODULE_AUTHOR("Anders Berg <anders.berg@lsi.com>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/i2c/busses/i2c-bcm-kona.c b/drivers/i2c/busses/i2c-bcm-kona.c
index 18a74a6751a9..2c9d9b1c8e64 100644
--- a/drivers/i2c/busses/i2c-bcm-kona.c
+++ b/drivers/i2c/busses/i2c-bcm-kona.c
@@ -895,7 +895,6 @@ MODULE_DEVICE_TABLE(of, bcm_kona_i2c_of_match);
 static struct platform_driver bcm_kona_i2c_driver = {
 	.driver = {
 		   .name = "bcm-kona-i2c",
-		   .owner = THIS_MODULE,
 		   .of_match_table = bcm_kona_i2c_of_match,
 		   },
 	.probe = bcm_kona_i2c_probe,
diff --git a/drivers/i2c/busses/i2c-bcm2835.c b/drivers/i2c/busses/i2c-bcm2835.c
index 4b8ecd0b3661..5d6feb937b9d 100644
--- a/drivers/i2c/busses/i2c-bcm2835.c
+++ b/drivers/i2c/busses/i2c-bcm2835.c
@@ -313,7 +313,6 @@ static struct platform_driver bcm2835_i2c_driver = {
 	.remove		= bcm2835_i2c_remove,
 	.driver		= {
 		.name	= "i2c-bcm2835",
-		.owner	= THIS_MODULE,
 		.of_match_table = bcm2835_i2c_of_match,
 	},
 };
diff --git a/drivers/i2c/busses/i2c-bfin-twi.c b/drivers/i2c/busses/i2c-bfin-twi.c
index 067c1615e968..af162b4c7a6d 100644
--- a/drivers/i2c/busses/i2c-bfin-twi.c
+++ b/drivers/i2c/busses/i2c-bfin-twi.c
@@ -717,7 +717,6 @@ static struct platform_driver i2c_bfin_twi_driver = {
 	.remove		= i2c_bfin_twi_remove,
 	.driver		= {
 		.name	= "i2c-bfin-twi",
-		.owner	= THIS_MODULE,
 		.pm	= I2C_BFIN_TWI_PM_OPS,
 	},
 };
diff --git a/drivers/i2c/busses/i2c-cadence.c b/drivers/i2c/busses/i2c-cadence.c
index 63f3f03ecc9b..626f74ecd4be 100644
--- a/drivers/i2c/busses/i2c-cadence.c
+++ b/drivers/i2c/busses/i2c-cadence.c
@@ -111,6 +111,8 @@
 #define CDNS_I2C_DIVA_MAX	4
 #define CDNS_I2C_DIVB_MAX	64
 
+#define CDNS_I2C_TIMEOUT_MAX	0xFF
+
 #define cdns_i2c_readreg(offset)       readl_relaxed(id->membase + offset)
 #define cdns_i2c_writereg(val, offset) writel_relaxed(val, id->membase + offset)
 
@@ -852,6 +854,15 @@ static int cdns_i2c_probe(struct platform_device *pdev)
 		goto err_clk_dis;
 	}
 
+	/*
+	 * Cadence I2C controller has a bug wherein it generates
+	 * invalid read transaction after HW timeout in master receiver mode.
+	 * HW timeout is not used by this driver and the interrupt is disabled.
+	 * But the feature itself cannot be disabled. Hence maximum value
+	 * is written to this register to reduce the chances of error.
+	 */
+	cdns_i2c_writereg(CDNS_I2C_TIMEOUT_MAX, CDNS_I2C_TIME_OUT_OFFSET);
+
 	dev_info(&pdev->dev, "%u kHz mmio %08lx irq %d\n",
 		 id->i2c_clk / 1000, (unsigned long)r_mem->start, id->irq);
 
@@ -890,7 +901,6 @@ MODULE_DEVICE_TABLE(of, cdns_i2c_of_match);
 static struct platform_driver cdns_i2c_drv = {
 	.driver = {
 		.name  = DRIVER_NAME,
-		.owner = THIS_MODULE,
 		.of_match_table = cdns_i2c_of_match,
 		.pm = &cdns_i2c_dev_pm_ops,
 	},
diff --git a/drivers/i2c/busses/i2c-cbus-gpio.c b/drivers/i2c/busses/i2c-cbus-gpio.c
index bdf040fd8675..b4f91e48948a 100644
--- a/drivers/i2c/busses/i2c-cbus-gpio.c
+++ b/drivers/i2c/busses/i2c-cbus-gpio.c
@@ -287,7 +287,6 @@ static struct platform_driver cbus_i2c_driver = {
 	.probe	= cbus_i2c_probe,
 	.remove	= cbus_i2c_remove,
 	.driver	= {
-		.owner	= THIS_MODULE,
 		.name	= "i2c-cbus-gpio",
 		.of_match_table = of_match_ptr(i2c_cbus_dt_ids),
 	},
diff --git a/drivers/i2c/busses/i2c-cpm.c b/drivers/i2c/busses/i2c-cpm.c
index f3b89a4698b6..2d466538b2e2 100644
--- a/drivers/i2c/busses/i2c-cpm.c
+++ b/drivers/i2c/busses/i2c-cpm.c
@@ -23,10 +23,6 @@
  *  but WITHOUT ANY WARRANTY; without even the implied warranty of
  *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  *  GNU General Public License for more details.
- *
- *  You should have received a copy of the GNU General Public License
- *  along with this program; if not, write to the Free Software
- *  Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  */
 
 #include <linux/kernel.h>
@@ -720,7 +716,6 @@ static struct platform_driver cpm_i2c_driver = {
 	.remove		= cpm_i2c_remove,
 	.driver = {
 		.name = "fsl-i2c-cpm",
-		.owner = THIS_MODULE,
 		.of_match_table = cpm_i2c_match,
 	},
 };
diff --git a/drivers/i2c/busses/i2c-cros-ec-tunnel.c b/drivers/i2c/busses/i2c-cros-ec-tunnel.c
index 05e033c98115..875c22ae5400 100644
--- a/drivers/i2c/busses/i2c-cros-ec-tunnel.c
+++ b/drivers/i2c/busses/i2c-cros-ec-tunnel.c
@@ -16,6 +16,8 @@
 #include <linux/platform_device.h>
 #include <linux/slab.h>
 
+#define I2C_MAX_RETRIES 3
+
 /**
  * struct ec_i2c_device - Driver data for I2C tunnel
  *
@@ -94,7 +96,7 @@ static int ec_i2c_construct_message(u8 *buf, const struct i2c_msg i2c_msgs[],
 		msg->addr_flags = i2c_msg->addr;
 
 		if (i2c_msg->flags & I2C_M_TEN)
-			msg->addr_flags |= EC_I2C_FLAG_10BIT;
+			return -EINVAL;
 
 		if (i2c_msg->flags & I2C_M_RD) {
 			msg->addr_flags |= EC_I2C_FLAG_READ;
@@ -218,7 +220,9 @@ static int ec_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg i2c_msgs[],
 		}
 	}
 
-	ec_i2c_construct_message(request, i2c_msgs, num, bus_num);
+	result = ec_i2c_construct_message(request, i2c_msgs, num, bus_num);
+	if (result)
+		goto exit;
 
 	msg.version = 0;
 	msg.command = EC_CMD_I2C_PASSTHRU;
@@ -227,7 +231,7 @@ static int ec_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg i2c_msgs[],
 	msg.indata = response;
 	msg.insize = response_len;
 
-	result = bus->ec->cmd_xfer(bus->ec, &msg);
+	result = cros_ec_cmd_xfer(bus->ec, &msg);
 	if (result < 0)
 		goto exit;
 
@@ -290,6 +294,7 @@ static int ec_i2c_probe(struct platform_device *pdev)
 	bus->adap.algo_data = bus;
 	bus->adap.dev.parent = &pdev->dev;
 	bus->adap.dev.of_node = np;
+	bus->adap.retries = I2C_MAX_RETRIES;
 
 	err = i2c_add_adapter(&bus->adap);
 	if (err) {
@@ -310,11 +315,20 @@ static int ec_i2c_remove(struct platform_device *dev)
 	return 0;
 }
 
+#ifdef CONFIG_OF
+static const struct of_device_id cros_ec_i2c_of_match[] = {
+	{ .compatible = "google,cros-ec-i2c-tunnel" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, cros_ec_i2c_of_match);
+#endif
+
 static struct platform_driver ec_i2c_tunnel_driver = {
 	.probe = ec_i2c_probe,
 	.remove = ec_i2c_remove,
 	.driver = {
 		.name = "cros-ec-i2c-tunnel",
+		.of_match_table = of_match_ptr(cros_ec_i2c_of_match),
 	},
 };
 
diff --git a/drivers/i2c/busses/i2c-davinci.c b/drivers/i2c/busses/i2c-davinci.c
index 4d9614719128..6dc7ff5d3d9a 100644
--- a/drivers/i2c/busses/i2c-davinci.c
+++ b/drivers/i2c/busses/i2c-davinci.c
@@ -17,10 +17,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  * ----------------------------------------------------------------------------
  *
  */
@@ -372,8 +368,7 @@ i2c_davinci_xfer_msg(struct i2c_adapter *adap, struct i2c_msg *msg, int stop)
 		flag |= DAVINCI_I2C_MDR_STP;
 	davinci_i2c_write_reg(dev, DAVINCI_I2C_MDR_REG, flag);
 
-	r = wait_for_completion_interruptible_timeout(&dev->cmd_complete,
-						      dev->adapter.timeout);
+	r = wait_for_completion_timeout(&dev->cmd_complete, dev->adapter.timeout);
 	if (r == 0) {
 		dev_err(dev->dev, "controller timed out\n");
 		davinci_i2c_recover_bus(dev);
@@ -384,7 +379,6 @@ i2c_davinci_xfer_msg(struct i2c_adapter *adap, struct i2c_msg *msg, int stop)
 	if (dev->buf_len) {
 		/* This should be 0 if all bytes were transferred
 		 * or dev->cmd_err denotes an error.
-		 * A signal may have aborted the transfer.
 		 */
 		if (r >= 0) {
 			dev_err(dev->dev, "abnormal termination buf_len=%i\n",
@@ -411,11 +405,9 @@ i2c_davinci_xfer_msg(struct i2c_adapter *adap, struct i2c_msg *msg, int stop)
 	if (dev->cmd_err & DAVINCI_I2C_STR_NACK) {
 		if (msg->flags & I2C_M_IGNORE_NAK)
 			return msg->len;
-		if (stop) {
-			w = davinci_i2c_read_reg(dev, DAVINCI_I2C_MDR_REG);
-			w |= DAVINCI_I2C_MDR_STP;
-			davinci_i2c_write_reg(dev, DAVINCI_I2C_MDR_REG, w);
-		}
+		w = davinci_i2c_read_reg(dev, DAVINCI_I2C_MDR_REG);
+		w |= DAVINCI_I2C_MDR_STP;
+		davinci_i2c_write_reg(dev, DAVINCI_I2C_MDR_REG, w);
 		return -EREMOTEIO;
 	}
 	return -EIO;
@@ -640,13 +632,17 @@ static int davinci_i2c_probe(struct platform_device *pdev)
 {
 	struct davinci_i2c_dev *dev;
 	struct i2c_adapter *adap;
-	struct resource *mem, *irq;
-	int r;
-
-	irq = platform_get_resource(pdev, IORESOURCE_IRQ, 0);
-	if (!irq) {
-		dev_err(&pdev->dev, "no irq resource?\n");
-		return -ENODEV;
+	struct resource *mem;
+	int r, irq;
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq <= 0) {
+		if (!irq)
+			irq = -ENXIO;
+		if (irq != -EPROBE_DEFER)
+			dev_err(&pdev->dev,
+				"can't get irq resource ret=%d\n", irq);
+		return irq;
 	}
 
 	dev = devm_kzalloc(&pdev->dev, sizeof(struct davinci_i2c_dev),
@@ -661,7 +657,7 @@ static int davinci_i2c_probe(struct platform_device *pdev)
 	init_completion(&dev->xfr_complete);
 #endif
 	dev->dev = &pdev->dev;
-	dev->irq = irq->start;
+	dev->irq = irq;
 	dev->pdata = dev_get_platdata(&pdev->dev);
 	platform_set_drvdata(pdev, dev);
 
@@ -793,7 +789,6 @@ static struct platform_driver davinci_i2c_driver = {
 	.remove		= davinci_i2c_remove,
 	.driver		= {
 		.name	= "i2c_davinci",
-		.owner	= THIS_MODULE,
 		.pm	= davinci_i2c_pm_ops,
 		.of_match_table = davinci_i2c_of_match,
 	},
diff --git a/drivers/i2c/busses/i2c-designware-core.c b/drivers/i2c/busses/i2c-designware-core.c
index 3c20e4bd6dd1..23628b7bfb8d 100644
--- a/drivers/i2c/busses/i2c-designware-core.c
+++ b/drivers/i2c/busses/i2c-designware-core.c
@@ -18,10 +18,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  * ----------------------------------------------------------------------------
  *
  */
@@ -363,7 +359,7 @@ int i2c_dw_init(struct dw_i2c_dev *dev)
 	}
 
 	/* Configure Tx/Rx FIFO threshold levels */
-	dw_writel(dev, dev->tx_fifo_depth - 1, DW_IC_TX_TL);
+	dw_writel(dev, dev->tx_fifo_depth / 2, DW_IC_TX_TL);
 	dw_writel(dev, 0, DW_IC_RX_TL);
 
 	/* configure the i2c master */
diff --git a/drivers/i2c/busses/i2c-designware-core.h b/drivers/i2c/busses/i2c-designware-core.h
index d66b6cbc9edc..5a410ef17abd 100644
--- a/drivers/i2c/busses/i2c-designware-core.h
+++ b/drivers/i2c/busses/i2c-designware-core.h
@@ -18,10 +18,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  * ----------------------------------------------------------------------------
  *
  */
diff --git a/drivers/i2c/busses/i2c-designware-pcidrv.c b/drivers/i2c/busses/i2c-designware-pcidrv.c
index d31d313ab4f7..acb40f95db78 100644
--- a/drivers/i2c/busses/i2c-designware-pcidrv.c
+++ b/drivers/i2c/busses/i2c-designware-pcidrv.c
@@ -19,10 +19,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  * ----------------------------------------------------------------------------
  *
  */
diff --git a/drivers/i2c/busses/i2c-designware-platdrv.c b/drivers/i2c/busses/i2c-designware-platdrv.c
index bc8773333155..2b463c313e4e 100644
--- a/drivers/i2c/busses/i2c-designware-platdrv.c
+++ b/drivers/i2c/busses/i2c-designware-platdrv.c
@@ -18,10 +18,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  * ----------------------------------------------------------------------------
  *
  */
@@ -30,6 +26,7 @@
 #include <linux/delay.h>
 #include <linux/i2c.h>
 #include <linux/clk.h>
+#include <linux/clk-provider.h>
 #include <linux/errno.h>
 #include <linux/sched.h>
 #include <linux/err.h>
@@ -41,6 +38,7 @@
 #include <linux/io.h>
 #include <linux/slab.h>
 #include <linux/acpi.h>
+#include <linux/platform_data/i2c-designware.h>
 #include "i2c-designware-core.h"
 
 static struct i2c_algorithm i2c_dw_algo = {
@@ -79,10 +77,7 @@ static void dw_i2c_acpi_params(struct platform_device *pdev, char method[],
 static int dw_i2c_acpi_configure(struct platform_device *pdev)
 {
 	struct dw_i2c_dev *dev = platform_get_drvdata(pdev);
-	bool fs_mode = dev->master_cfg & DW_IC_CON_SPEED_FAST;
-
-	if (!ACPI_HANDLE(&pdev->dev))
-		return -ENODEV;
+	const struct acpi_device_id *id;
 
 	dev->adapter.nr = -1;
 	dev->tx_fifo_depth = 32;
@@ -92,14 +87,33 @@ static int dw_i2c_acpi_configure(struct platform_device *pdev)
 	 * Try to get SDA hold time and *CNT values from an ACPI method if
 	 * it exists for both supported speed modes.
 	 */
-	dw_i2c_acpi_params(pdev, "SSCN", &dev->ss_hcnt, &dev->ss_lcnt,
-			   fs_mode ? NULL : &dev->sda_hold_time);
+	dw_i2c_acpi_params(pdev, "SSCN", &dev->ss_hcnt, &dev->ss_lcnt, NULL);
 	dw_i2c_acpi_params(pdev, "FMCN", &dev->fs_hcnt, &dev->fs_lcnt,
-			   fs_mode ? &dev->sda_hold_time : NULL);
+			   &dev->sda_hold_time);
+
+	/*
+	 * Provide a way for Designware I2C host controllers that are not
+	 * based on Intel LPSS to specify their input clock frequency via
+	 * id->driver_data.
+	 */
+	id = acpi_match_device(pdev->dev.driver->acpi_match_table, &pdev->dev);
+	if (id && id->driver_data)
+		clk_register_fixed_rate(&pdev->dev, dev_name(&pdev->dev), NULL,
+					CLK_IS_ROOT, id->driver_data);
 
 	return 0;
 }
 
+static void dw_i2c_acpi_unconfigure(struct platform_device *pdev)
+{
+	struct dw_i2c_dev *dev = platform_get_drvdata(pdev);
+	const struct acpi_device_id *id;
+
+	id = acpi_match_device(pdev->dev.driver->acpi_match_table, &pdev->dev);
+	if (id && id->driver_data)
+		clk_unregister(dev->clk);
+}
+
 static const struct acpi_device_id dw_i2c_acpi_match[] = {
 	{ "INT33C2", 0 },
 	{ "INT33C3", 0 },
@@ -107,6 +121,7 @@ static const struct acpi_device_id dw_i2c_acpi_match[] = {
 	{ "INT3433", 0 },
 	{ "80860F41", 0 },
 	{ "808622C1", 0 },
+	{ "AMD0010", 133 * 1000 * 1000 },
 	{ }
 };
 MODULE_DEVICE_TABLE(acpi, dw_i2c_acpi_match);
@@ -115,6 +130,7 @@ static inline int dw_i2c_acpi_configure(struct platform_device *pdev)
 {
 	return -ENODEV;
 }
+static inline void dw_i2c_acpi_unconfigure(struct platform_device *pdev) { }
 #endif
 
 static int dw_i2c_probe(struct platform_device *pdev)
@@ -122,7 +138,9 @@ static int dw_i2c_probe(struct platform_device *pdev)
 	struct dw_i2c_dev *dev;
 	struct i2c_adapter *adap;
 	struct resource *mem;
+	struct dw_i2c_platform_data *pdata;
 	int irq, r;
+	u32 clk_freq, ht = 0;
 
 	irq = platform_get_irq(pdev, 0);
 	if (irq < 0) {
@@ -145,21 +163,14 @@ static int dw_i2c_probe(struct platform_device *pdev)
 	dev->irq = irq;
 	platform_set_drvdata(pdev, dev);
 
-	dev->clk = devm_clk_get(&pdev->dev, NULL);
-	dev->get_clk_rate_khz = i2c_dw_get_clk_rate_khz;
-
-	if (IS_ERR(dev->clk))
-		return PTR_ERR(dev->clk);
-	clk_prepare_enable(dev->clk);
-
-	if (pdev->dev.of_node) {
-		u32 ht = 0;
-		u32 ic_clk = dev->get_clk_rate_khz(dev);
+	/* fast mode by default because of legacy reasons */
+	clk_freq = 400000;
 
+	if (ACPI_COMPANION(&pdev->dev)) {
+		dw_i2c_acpi_configure(pdev);
+	} else if (pdev->dev.of_node) {
 		of_property_read_u32(pdev->dev.of_node,
 					"i2c-sda-hold-time-ns", &ht);
-		dev->sda_hold_time = div_u64((u64)ic_clk * ht + 500000,
-					     1000000);
 
 		of_property_read_u32(pdev->dev.of_node,
 				     "i2c-sda-falling-time-ns",
@@ -167,6 +178,21 @@ static int dw_i2c_probe(struct platform_device *pdev)
 		of_property_read_u32(pdev->dev.of_node,
 				     "i2c-scl-falling-time-ns",
 				     &dev->scl_falling_time);
+
+		of_property_read_u32(pdev->dev.of_node, "clock-frequency",
+				     &clk_freq);
+
+		/* Only standard mode at 100kHz and fast mode at 400kHz
+		 * are supported.
+		 */
+		if (clk_freq != 100000 && clk_freq != 400000) {
+			dev_err(&pdev->dev, "Only 100kHz and 400kHz supported");
+			return -EINVAL;
+		}
+	} else {
+		pdata = dev_get_platdata(&pdev->dev);
+		if (pdata)
+			clk_freq = pdata->i2c_scl_freq;
 	}
 
 	dev->functionality =
@@ -176,12 +202,27 @@ static int dw_i2c_probe(struct platform_device *pdev)
 		I2C_FUNC_SMBUS_BYTE_DATA |
 		I2C_FUNC_SMBUS_WORD_DATA |
 		I2C_FUNC_SMBUS_I2C_BLOCK;
-	dev->master_cfg =  DW_IC_CON_MASTER | DW_IC_CON_SLAVE_DISABLE |
-		DW_IC_CON_RESTART_EN | DW_IC_CON_SPEED_FAST;
+	if (clk_freq == 100000)
+		dev->master_cfg =  DW_IC_CON_MASTER | DW_IC_CON_SLAVE_DISABLE |
+			DW_IC_CON_RESTART_EN | DW_IC_CON_SPEED_STD;
+	else
+		dev->master_cfg =  DW_IC_CON_MASTER | DW_IC_CON_SLAVE_DISABLE |
+			DW_IC_CON_RESTART_EN | DW_IC_CON_SPEED_FAST;
 
-	/* Try first if we can configure the device from ACPI */
-	r = dw_i2c_acpi_configure(pdev);
-	if (r) {
+	dev->clk = devm_clk_get(&pdev->dev, NULL);
+	dev->get_clk_rate_khz = i2c_dw_get_clk_rate_khz;
+	if (IS_ERR(dev->clk))
+		return PTR_ERR(dev->clk);
+	clk_prepare_enable(dev->clk);
+
+	if (!dev->sda_hold_time && ht) {
+		u32 ic_clk = dev->get_clk_rate_khz(dev);
+
+		dev->sda_hold_time = div_u64((u64)ic_clk * ht + 500000,
+					     1000000);
+	}
+
+	if (!dev->tx_fifo_depth) {
 		u32 param1 = i2c_dw_read_comp_param(dev);
 
 		dev->tx_fifo_depth = ((param1 >> 16) & 0xff) + 1;
@@ -237,6 +278,9 @@ static int dw_i2c_remove(struct platform_device *pdev)
 	pm_runtime_put(&pdev->dev);
 	pm_runtime_disable(&pdev->dev);
 
+	if (ACPI_COMPANION(&pdev->dev))
+		dw_i2c_acpi_unconfigure(pdev);
+
 	return 0;
 }
 
@@ -283,7 +327,6 @@ static struct platform_driver dw_i2c_driver = {
 	.remove = dw_i2c_remove,
 	.driver		= {
 		.name	= "i2c_designware",
-		.owner	= THIS_MODULE,
 		.of_match_table = of_match_ptr(dw_i2c_of_match),
 		.acpi_match_table = ACPI_PTR(dw_i2c_acpi_match),
 		.pm	= &dw_i2c_dev_pm_ops,
diff --git a/drivers/i2c/busses/i2c-dln2.c b/drivers/i2c/busses/i2c-dln2.c
new file mode 100644
index 000000000000..b3fb86af4cbb
--- /dev/null
+++ b/drivers/i2c/busses/i2c-dln2.c
@@ -0,0 +1,262 @@
+/*
+ * Driver for the Diolan DLN-2 USB-I2C adapter
+ *
+ * Copyright (c) 2014 Intel Corporation
+ *
+ * Derived from:
+ *  i2c-diolan-u2c.c
+ *  Copyright (c) 2010-2011 Ericsson AB
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation, version 2.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/slab.h>
+#include <linux/i2c.h>
+#include <linux/platform_device.h>
+#include <linux/mfd/dln2.h>
+
+#define DLN2_I2C_MODULE_ID		0x03
+#define DLN2_I2C_CMD(cmd)		DLN2_CMD(cmd, DLN2_I2C_MODULE_ID)
+
+/* I2C commands */
+#define DLN2_I2C_GET_PORT_COUNT		DLN2_I2C_CMD(0x00)
+#define DLN2_I2C_ENABLE			DLN2_I2C_CMD(0x01)
+#define DLN2_I2C_DISABLE		DLN2_I2C_CMD(0x02)
+#define DLN2_I2C_IS_ENABLED		DLN2_I2C_CMD(0x03)
+#define DLN2_I2C_WRITE			DLN2_I2C_CMD(0x06)
+#define DLN2_I2C_READ			DLN2_I2C_CMD(0x07)
+#define DLN2_I2C_SCAN_DEVICES		DLN2_I2C_CMD(0x08)
+#define DLN2_I2C_PULLUP_ENABLE		DLN2_I2C_CMD(0x09)
+#define DLN2_I2C_PULLUP_DISABLE		DLN2_I2C_CMD(0x0A)
+#define DLN2_I2C_PULLUP_IS_ENABLED	DLN2_I2C_CMD(0x0B)
+#define DLN2_I2C_TRANSFER		DLN2_I2C_CMD(0x0C)
+#define DLN2_I2C_SET_MAX_REPLY_COUNT	DLN2_I2C_CMD(0x0D)
+#define DLN2_I2C_GET_MAX_REPLY_COUNT	DLN2_I2C_CMD(0x0E)
+
+#define DLN2_I2C_MAX_XFER_SIZE		256
+#define DLN2_I2C_BUF_SIZE		(DLN2_I2C_MAX_XFER_SIZE + 16)
+
+struct dln2_i2c {
+	struct platform_device *pdev;
+	struct i2c_adapter adapter;
+	u8 port;
+	/*
+	 * Buffer to hold the packet for read or write transfers. One is enough
+	 * since we can't have multiple transfers in parallel on the i2c bus.
+	 */
+	void *buf;
+};
+
+static int dln2_i2c_enable(struct dln2_i2c *dln2, bool enable)
+{
+	u16 cmd;
+	struct {
+		u8 port;
+	} tx;
+
+	tx.port = dln2->port;
+
+	if (enable)
+		cmd = DLN2_I2C_ENABLE;
+	else
+		cmd = DLN2_I2C_DISABLE;
+
+	return dln2_transfer_tx(dln2->pdev, cmd, &tx, sizeof(tx));
+}
+
+static int dln2_i2c_write(struct dln2_i2c *dln2, u8 addr,
+			  u8 *data, u16 data_len)
+{
+	int ret;
+	struct {
+		u8 port;
+		u8 addr;
+		u8 mem_addr_len;
+		__le32 mem_addr;
+		__le16 buf_len;
+		u8 buf[DLN2_I2C_MAX_XFER_SIZE];
+	} __packed *tx = dln2->buf;
+	unsigned len;
+
+	BUILD_BUG_ON(sizeof(*tx) > DLN2_I2C_BUF_SIZE);
+
+	tx->port = dln2->port;
+	tx->addr = addr;
+	tx->mem_addr_len = 0;
+	tx->mem_addr = 0;
+	tx->buf_len = cpu_to_le16(data_len);
+	memcpy(tx->buf, data, data_len);
+
+	len = sizeof(*tx) + data_len - DLN2_I2C_MAX_XFER_SIZE;
+	ret = dln2_transfer_tx(dln2->pdev, DLN2_I2C_WRITE, tx, len);
+	if (ret < 0)
+		return ret;
+
+	return data_len;
+}
+
+static int dln2_i2c_read(struct dln2_i2c *dln2, u16 addr, u8 *data,
+			 u16 data_len)
+{
+	int ret;
+	struct {
+		u8 port;
+		u8 addr;
+		u8 mem_addr_len;
+		__le32 mem_addr;
+		__le16 buf_len;
+	} __packed tx;
+	struct {
+		__le16 buf_len;
+		u8 buf[DLN2_I2C_MAX_XFER_SIZE];
+	} __packed *rx = dln2->buf;
+	unsigned rx_len = sizeof(*rx);
+
+	BUILD_BUG_ON(sizeof(*rx) > DLN2_I2C_BUF_SIZE);
+
+	tx.port = dln2->port;
+	tx.addr = addr;
+	tx.mem_addr_len = 0;
+	tx.mem_addr = 0;
+	tx.buf_len = cpu_to_le16(data_len);
+
+	ret = dln2_transfer(dln2->pdev, DLN2_I2C_READ, &tx, sizeof(tx),
+			    rx, &rx_len);
+	if (ret < 0)
+		return ret;
+	if (rx_len < sizeof(rx->buf_len) + data_len)
+		return -EPROTO;
+	if (le16_to_cpu(rx->buf_len) != data_len)
+		return -EPROTO;
+
+	memcpy(data, rx->buf, data_len);
+
+	return data_len;
+}
+
+static int dln2_i2c_xfer(struct i2c_adapter *adapter,
+			 struct i2c_msg *msgs, int num)
+{
+	struct dln2_i2c *dln2 = i2c_get_adapdata(adapter);
+	struct i2c_msg *pmsg;
+	struct device *dev = &dln2->adapter.dev;
+	int i;
+
+	for (i = 0; i < num; i++) {
+		int ret;
+
+		pmsg = &msgs[i];
+
+		if (pmsg->len > DLN2_I2C_MAX_XFER_SIZE) {
+			dev_warn(dev, "maximum transfer size exceeded\n");
+			return -EOPNOTSUPP;
+		}
+
+		if (pmsg->flags & I2C_M_RD) {
+			ret = dln2_i2c_read(dln2, pmsg->addr, pmsg->buf,
+					    pmsg->len);
+			if (ret < 0)
+				return ret;
+
+			pmsg->len = ret;
+		} else {
+			ret = dln2_i2c_write(dln2, pmsg->addr, pmsg->buf,
+					     pmsg->len);
+			if (ret != pmsg->len)
+				return -EPROTO;
+		}
+	}
+
+	return num;
+}
+
+static u32 dln2_i2c_func(struct i2c_adapter *a)
+{
+	return I2C_FUNC_I2C | I2C_FUNC_SMBUS_BYTE | I2C_FUNC_SMBUS_BYTE_DATA |
+		I2C_FUNC_SMBUS_WORD_DATA | I2C_FUNC_SMBUS_BLOCK_PROC_CALL |
+		I2C_FUNC_SMBUS_I2C_BLOCK;
+}
+
+static const struct i2c_algorithm dln2_i2c_usb_algorithm = {
+	.master_xfer = dln2_i2c_xfer,
+	.functionality = dln2_i2c_func,
+};
+
+static int dln2_i2c_probe(struct platform_device *pdev)
+{
+	int ret;
+	struct dln2_i2c *dln2;
+	struct device *dev = &pdev->dev;
+	struct dln2_platform_data *pdata = dev_get_platdata(&pdev->dev);
+
+	dln2 = devm_kzalloc(dev, sizeof(*dln2), GFP_KERNEL);
+	if (!dln2)
+		return -ENOMEM;
+
+	dln2->buf = devm_kmalloc(dev, DLN2_I2C_BUF_SIZE, GFP_KERNEL);
+	if (!dln2->buf)
+		return -ENOMEM;
+
+	dln2->pdev = pdev;
+	dln2->port = pdata->port;
+
+	/* setup i2c adapter description */
+	dln2->adapter.owner = THIS_MODULE;
+	dln2->adapter.class = I2C_CLASS_HWMON;
+	dln2->adapter.algo = &dln2_i2c_usb_algorithm;
+	dln2->adapter.dev.parent = dev;
+	i2c_set_adapdata(&dln2->adapter, dln2);
+	snprintf(dln2->adapter.name, sizeof(dln2->adapter.name), "%s-%s-%d",
+		 "dln2-i2c", dev_name(pdev->dev.parent), dln2->port);
+
+	platform_set_drvdata(pdev, dln2);
+
+	/* initialize the i2c interface */
+	ret = dln2_i2c_enable(dln2, true);
+	if (ret < 0) {
+		dev_err(dev, "failed to initialize adapter: %d\n", ret);
+		return ret;
+	}
+
+	/* and finally attach to i2c layer */
+	ret = i2c_add_adapter(&dln2->adapter);
+	if (ret < 0) {
+		dev_err(dev, "failed to add I2C adapter: %d\n", ret);
+		goto out_disable;
+	}
+
+	return 0;
+
+out_disable:
+	dln2_i2c_enable(dln2, false);
+
+	return ret;
+}
+
+static int dln2_i2c_remove(struct platform_device *pdev)
+{
+	struct dln2_i2c *dln2 = platform_get_drvdata(pdev);
+
+	i2c_del_adapter(&dln2->adapter);
+	dln2_i2c_enable(dln2, false);
+
+	return 0;
+}
+
+static struct platform_driver dln2_i2c_driver = {
+	.driver.name	= "dln2-i2c",
+	.probe		= dln2_i2c_probe,
+	.remove		= dln2_i2c_remove,
+};
+
+module_platform_driver(dln2_i2c_driver);
+
+MODULE_AUTHOR("Laurentiu Palcu <laurentiu.palcu@intel.com>");
+MODULE_DESCRIPTION("Driver for the Diolan DLN2 I2C master interface");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS("platform:dln2-i2c");
diff --git a/drivers/i2c/busses/i2c-efm32.c b/drivers/i2c/busses/i2c-efm32.c
index 10b8323b08d4..8eff62738877 100644
--- a/drivers/i2c/busses/i2c-efm32.c
+++ b/drivers/i2c/busses/i2c-efm32.c
@@ -473,7 +473,6 @@ static struct platform_driver efm32_i2c_driver = {
 
 	.driver = {
 		.name = DRIVER_NAME,
-		.owner = THIS_MODULE,
 		.of_match_table = efm32_i2c_dt_ids,
 	},
 };
diff --git a/drivers/i2c/busses/i2c-eg20t.c b/drivers/i2c/busses/i2c-eg20t.c
index a44ea13d1434..76e699f9ed97 100644
--- a/drivers/i2c/busses/i2c-eg20t.c
+++ b/drivers/i2c/busses/i2c-eg20t.c
@@ -9,10 +9,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307, USA.
  */
 
 #include <linux/module.h>
diff --git a/drivers/i2c/busses/i2c-elektor.c b/drivers/i2c/busses/i2c-elektor.c
index 485497066ed7..92e8c0ce1625 100644
--- a/drivers/i2c/busses/i2c-elektor.c
+++ b/drivers/i2c/busses/i2c-elektor.c
@@ -12,11 +12,7 @@
     This program is distributed in the hope that it will be useful,
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-    GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.		     */
+    GNU General Public License for more details.			     */
 /* ------------------------------------------------------------------------- */
 
 /* With some changes from Kyösti Mälkki <kmalkki@cc.hut.fi> and even
diff --git a/drivers/i2c/busses/i2c-exynos5.c b/drivers/i2c/busses/i2c-exynos5.c
index 28073f1d6d47..b29c7500461a 100644
--- a/drivers/i2c/busses/i2c-exynos5.c
+++ b/drivers/i2c/busses/i2c-exynos5.c
@@ -83,7 +83,6 @@
 #define HSI2C_INT_TX_ALMOSTEMPTY_EN		(1u << 0)
 #define HSI2C_INT_RX_ALMOSTFULL_EN		(1u << 1)
 #define HSI2C_INT_TRAILING_EN			(1u << 6)
-#define HSI2C_INT_I2C_EN			(1u << 9)
 
 /* I2C_INT_STAT Register bits */
 #define HSI2C_INT_TX_ALMOSTEMPTY		(1u << 0)
@@ -95,6 +94,17 @@
 #define HSI2C_INT_TRAILING			(1u << 6)
 #define HSI2C_INT_I2C				(1u << 9)
 
+#define HSI2C_INT_TRANS_DONE			(1u << 7)
+#define HSI2C_INT_TRANS_ABORT			(1u << 8)
+#define HSI2C_INT_NO_DEV_ACK			(1u << 9)
+#define HSI2C_INT_NO_DEV			(1u << 10)
+#define HSI2C_INT_TIMEOUT			(1u << 11)
+#define HSI2C_INT_I2C_TRANS			(HSI2C_INT_TRANS_DONE |	\
+						HSI2C_INT_TRANS_ABORT |	\
+						HSI2C_INT_NO_DEV_ACK |	\
+						HSI2C_INT_NO_DEV |	\
+						HSI2C_INT_TIMEOUT)
+
 /* I2C_FIFO_STAT Register bits */
 #define HSI2C_RX_FIFO_EMPTY			(1u << 24)
 #define HSI2C_RX_FIFO_FULL			(1u << 23)
@@ -143,6 +153,8 @@
 
 #define EXYNOS5_I2C_TIMEOUT (msecs_to_jiffies(1000))
 
+#define HSI2C_EXYNOS7	BIT(0)
+
 struct exynos5_i2c {
 	struct i2c_adapter	adap;
 	unsigned int		suspended:1;
@@ -192,6 +204,7 @@ struct exynos5_i2c {
  */
 struct exynos_hsi2c_variant {
 	unsigned int	fifo_depth;
+	unsigned int	hw;
 };
 
 static const struct exynos_hsi2c_variant exynos5250_hsi2c_data = {
@@ -202,6 +215,11 @@ static const struct exynos_hsi2c_variant exynos5260_hsi2c_data = {
 	.fifo_depth	= 16,
 };
 
+static const struct exynos_hsi2c_variant exynos7_hsi2c_data = {
+	.fifo_depth	= 16,
+	.hw		= HSI2C_EXYNOS7,
+};
+
 static const struct of_device_id exynos5_i2c_match[] = {
 	{
 		.compatible = "samsung,exynos5-hsi2c",
@@ -212,6 +230,9 @@ static const struct of_device_id exynos5_i2c_match[] = {
 	}, {
 		.compatible = "samsung,exynos5260-hsi2c",
 		.data = &exynos5260_hsi2c_data
+	}, {
+		.compatible = "samsung,exynos7-hsi2c",
+		.data = &exynos7_hsi2c_data
 	}, {},
 };
 MODULE_DEVICE_TABLE(of, exynos5_i2c_match);
@@ -256,13 +277,24 @@ static int exynos5_i2c_set_timing(struct exynos5_i2c *i2c, int mode)
 				i2c->hs_clock : i2c->fs_clock;
 
 	/*
+	 * In case of HSI2C controller in Exynos5 series
 	 * FPCLK / FI2C =
 	 * (CLK_DIV + 1) * (TSCLK_L + TSCLK_H + 2) + 8 + 2 * FLT_CYCLE
+	 *
+	 * In case of HSI2C controllers in Exynos7 series
+	 * FPCLK / FI2C =
+	 * (CLK_DIV + 1) * (TSCLK_L + TSCLK_H + 2) + 8 + FLT_CYCLE
+	 *
 	 * utemp0 = (CLK_DIV + 1) * (TSCLK_L + TSCLK_H + 2)
 	 * utemp1 = (TSCLK_L + TSCLK_H + 2)
 	 */
 	t_ftl_cycle = (readl(i2c->regs + HSI2C_CONF) >> 16) & 0x7;
-	utemp0 = (clkin / op_clk) - 8 - 2 * t_ftl_cycle;
+	utemp0 = (clkin / op_clk) - 8;
+
+	if (i2c->variant->hw == HSI2C_EXYNOS7)
+		utemp0 -= t_ftl_cycle;
+	else
+		utemp0 -= 2 * t_ftl_cycle;
 
 	/* CLK_DIV max is 256 */
 	for (div = 0; div < 256; div++) {
@@ -407,7 +439,28 @@ static irqreturn_t exynos5_i2c_irq(int irqno, void *dev_id)
 	writel(int_status, i2c->regs + HSI2C_INT_STATUS);
 
 	/* handle interrupt related to the transfer status */
-	if (int_status & HSI2C_INT_I2C) {
+	if (i2c->variant->hw == HSI2C_EXYNOS7) {
+		if (int_status & HSI2C_INT_TRANS_DONE) {
+			i2c->trans_done = 1;
+			i2c->state = 0;
+		} else if (int_status & HSI2C_INT_TRANS_ABORT) {
+			dev_dbg(i2c->dev, "Deal with arbitration lose\n");
+			i2c->state = -EAGAIN;
+			goto stop;
+		} else if (int_status & HSI2C_INT_NO_DEV_ACK) {
+			dev_dbg(i2c->dev, "No ACK from device\n");
+			i2c->state = -ENXIO;
+			goto stop;
+		} else if (int_status & HSI2C_INT_NO_DEV) {
+			dev_dbg(i2c->dev, "No device\n");
+			i2c->state = -ENXIO;
+			goto stop;
+		} else if (int_status & HSI2C_INT_TIMEOUT) {
+			dev_dbg(i2c->dev, "Accessing device timed out\n");
+			i2c->state = -ETIMEDOUT;
+			goto stop;
+		}
+	} else if (int_status & HSI2C_INT_I2C) {
 		trans_status = readl(i2c->regs + HSI2C_TRANS_STATUS);
 		if (trans_status & HSI2C_NO_DEV_ACK) {
 			dev_dbg(i2c->dev, "No ACK from device\n");
@@ -423,7 +476,7 @@ static irqreturn_t exynos5_i2c_irq(int irqno, void *dev_id)
 			goto stop;
 		} else if (trans_status & HSI2C_TIMEOUT_AUTO) {
 			dev_dbg(i2c->dev, "Accessing device timed out\n");
-			i2c->state = -EAGAIN;
+			i2c->state = -ETIMEDOUT;
 			goto stop;
 		} else if (trans_status & HSI2C_TRANS_DONE) {
 			i2c->trans_done = 1;
@@ -512,12 +565,17 @@ static int exynos5_i2c_wait_bus_idle(struct exynos5_i2c *i2c)
 static void exynos5_i2c_message_start(struct exynos5_i2c *i2c, int stop)
 {
 	u32 i2c_ctl;
-	u32 int_en = HSI2C_INT_I2C_EN;
+	u32 int_en = 0;
 	u32 i2c_auto_conf = 0;
 	u32 fifo_ctl;
 	unsigned long flags;
 	unsigned short trig_lvl;
 
+	if (i2c->variant->hw == HSI2C_EXYNOS7)
+		int_en |= HSI2C_INT_I2C_TRANS;
+	else
+		int_en |= HSI2C_INT_I2C;
+
 	i2c_ctl = readl(i2c->regs + HSI2C_CTL);
 	i2c_ctl &= ~(HSI2C_TXCHON | HSI2C_RXCHON);
 	fifo_ctl = HSI2C_RXFIFO_EN | HSI2C_TXFIFO_EN;
@@ -724,12 +782,13 @@ static int exynos5_i2c_probe(struct platform_device *pdev)
 		goto err_clk;
 	}
 
+	/* Need to check the variant before setting up. */
+	i2c->variant = exynos5_i2c_get_variant(pdev);
+
 	ret = exynos5_hsi2c_clock_setup(i2c);
 	if (ret)
 		goto err_clk;
 
-	i2c->variant = exynos5_i2c_get_variant(pdev);
-
 	exynos5_i2c_reset(i2c);
 
 	ret = i2c_add_adapter(&i2c->adap);
@@ -802,7 +861,6 @@ static struct platform_driver exynos5_i2c_driver = {
 	.probe		= exynos5_i2c_probe,
 	.remove		= exynos5_i2c_remove,
 	.driver		= {
-		.owner	= THIS_MODULE,
 		.name	= "exynos5-hsi2c",
 		.pm	= &exynos5_i2c_dev_pm_ops,
 		.of_match_table = exynos5_i2c_match,
diff --git a/drivers/i2c/busses/i2c-gpio.c b/drivers/i2c/busses/i2c-gpio.c
index 933f1e453e41..34cfc0ebdcb9 100644
--- a/drivers/i2c/busses/i2c-gpio.c
+++ b/drivers/i2c/busses/i2c-gpio.c
@@ -260,7 +260,6 @@ MODULE_DEVICE_TABLE(of, i2c_gpio_dt_ids);
 static struct platform_driver i2c_gpio_driver = {
 	.driver		= {
 		.name	= "i2c-gpio",
-		.owner	= THIS_MODULE,
 		.of_match_table	= of_match_ptr(i2c_gpio_dt_ids),
 	},
 	.probe		= i2c_gpio_probe,
diff --git a/drivers/i2c/busses/i2c-highlander.c b/drivers/i2c/busses/i2c-highlander.c
index 512fcfabc18e..56dc69e7349f 100644
--- a/drivers/i2c/busses/i2c-highlander.c
+++ b/drivers/i2c/busses/i2c-highlander.c
@@ -456,7 +456,6 @@ static int highlander_i2c_remove(struct platform_device *pdev)
 static struct platform_driver highlander_i2c_driver = {
 	.driver		= {
 		.name	= "i2c-highlander",
-		.owner	= THIS_MODULE,
 	},
 
 	.probe		= highlander_i2c_probe,
diff --git a/drivers/i2c/busses/i2c-hix5hd2.c b/drivers/i2c/busses/i2c-hix5hd2.c
new file mode 100644
index 000000000000..8fe78d08e01c
--- /dev/null
+++ b/drivers/i2c/busses/i2c-hix5hd2.c
@@ -0,0 +1,557 @@
+/*
+ * Copyright (c) 2014 Linaro Ltd.
+ * Copyright (c) 2014 Hisilicon Limited.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * Now only support 7 bit address.
+ */
+
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/i2c.h>
+#include <linux/io.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+
+/* Register Map */
+#define HIX5I2C_CTRL		0x00
+#define HIX5I2C_COM		0x04
+#define HIX5I2C_ICR		0x08
+#define HIX5I2C_SR		0x0c
+#define HIX5I2C_SCL_H		0x10
+#define HIX5I2C_SCL_L		0x14
+#define HIX5I2C_TXR		0x18
+#define HIX5I2C_RXR		0x1c
+
+/* I2C_CTRL_REG */
+#define I2C_ENABLE		BIT(8)
+#define I2C_UNMASK_TOTAL	BIT(7)
+#define I2C_UNMASK_START	BIT(6)
+#define I2C_UNMASK_END		BIT(5)
+#define I2C_UNMASK_SEND		BIT(4)
+#define I2C_UNMASK_RECEIVE	BIT(3)
+#define I2C_UNMASK_ACK		BIT(2)
+#define I2C_UNMASK_ARBITRATE	BIT(1)
+#define I2C_UNMASK_OVER		BIT(0)
+#define I2C_UNMASK_ALL		(I2C_UNMASK_ACK | I2C_UNMASK_OVER)
+
+/* I2C_COM_REG */
+#define I2C_NO_ACK		BIT(4)
+#define I2C_START		BIT(3)
+#define I2C_READ		BIT(2)
+#define I2C_WRITE		BIT(1)
+#define I2C_STOP		BIT(0)
+
+/* I2C_ICR_REG */
+#define I2C_CLEAR_START		BIT(6)
+#define I2C_CLEAR_END		BIT(5)
+#define I2C_CLEAR_SEND		BIT(4)
+#define I2C_CLEAR_RECEIVE	BIT(3)
+#define I2C_CLEAR_ACK		BIT(2)
+#define I2C_CLEAR_ARBITRATE	BIT(1)
+#define I2C_CLEAR_OVER		BIT(0)
+#define I2C_CLEAR_ALL		(I2C_CLEAR_START | I2C_CLEAR_END | \
+				I2C_CLEAR_SEND | I2C_CLEAR_RECEIVE | \
+				I2C_CLEAR_ACK | I2C_CLEAR_ARBITRATE | \
+				I2C_CLEAR_OVER)
+
+/* I2C_SR_REG */
+#define I2C_BUSY		BIT(7)
+#define I2C_START_INTR		BIT(6)
+#define I2C_END_INTR		BIT(5)
+#define I2C_SEND_INTR		BIT(4)
+#define I2C_RECEIVE_INTR	BIT(3)
+#define I2C_ACK_INTR		BIT(2)
+#define I2C_ARBITRATE_INTR	BIT(1)
+#define I2C_OVER_INTR		BIT(0)
+
+#define HIX5I2C_MAX_FREQ	400000		/* 400k */
+#define HIX5I2C_READ_OPERATION	0x01
+
+enum hix5hd2_i2c_state {
+	HIX5I2C_STAT_RW_ERR = -1,
+	HIX5I2C_STAT_INIT,
+	HIX5I2C_STAT_RW,
+	HIX5I2C_STAT_SND_STOP,
+	HIX5I2C_STAT_RW_SUCCESS,
+};
+
+struct hix5hd2_i2c_priv {
+	struct i2c_adapter adap;
+	struct i2c_msg *msg;
+	struct completion msg_complete;
+	unsigned int msg_idx;
+	unsigned int msg_len;
+	int stop;
+	void __iomem *regs;
+	struct clk *clk;
+	struct device *dev;
+	spinlock_t lock;	/* IRQ synchronization */
+	int err;
+	unsigned int freq;
+	enum hix5hd2_i2c_state state;
+};
+
+static u32 hix5hd2_i2c_clr_pend_irq(struct hix5hd2_i2c_priv *priv)
+{
+	u32 val = readl_relaxed(priv->regs + HIX5I2C_SR);
+
+	writel_relaxed(val, priv->regs + HIX5I2C_ICR);
+
+	return val;
+}
+
+static void hix5hd2_i2c_clr_all_irq(struct hix5hd2_i2c_priv *priv)
+{
+	writel_relaxed(I2C_CLEAR_ALL, priv->regs + HIX5I2C_ICR);
+}
+
+static void hix5hd2_i2c_disable_irq(struct hix5hd2_i2c_priv *priv)
+{
+	writel_relaxed(0, priv->regs + HIX5I2C_CTRL);
+}
+
+static void hix5hd2_i2c_enable_irq(struct hix5hd2_i2c_priv *priv)
+{
+	writel_relaxed(I2C_ENABLE | I2C_UNMASK_TOTAL | I2C_UNMASK_ALL,
+		       priv->regs + HIX5I2C_CTRL);
+}
+
+static void hix5hd2_i2c_drv_setrate(struct hix5hd2_i2c_priv *priv)
+{
+	u32 rate, val;
+	u32 scl, sysclock;
+
+	/* close all i2c interrupt */
+	val = readl_relaxed(priv->regs + HIX5I2C_CTRL);
+	writel_relaxed(val & (~I2C_UNMASK_TOTAL), priv->regs + HIX5I2C_CTRL);
+
+	rate = priv->freq;
+	sysclock = clk_get_rate(priv->clk);
+	scl = (sysclock / (rate * 2)) / 2 - 1;
+	writel_relaxed(scl, priv->regs + HIX5I2C_SCL_H);
+	writel_relaxed(scl, priv->regs + HIX5I2C_SCL_L);
+
+	/* restore original interrupt*/
+	writel_relaxed(val, priv->regs + HIX5I2C_CTRL);
+
+	dev_dbg(priv->dev, "%s: sysclock=%d, rate=%d, scl=%d\n",
+		__func__, sysclock, rate, scl);
+}
+
+static void hix5hd2_i2c_init(struct hix5hd2_i2c_priv *priv)
+{
+	hix5hd2_i2c_disable_irq(priv);
+	hix5hd2_i2c_drv_setrate(priv);
+	hix5hd2_i2c_clr_all_irq(priv);
+	hix5hd2_i2c_enable_irq(priv);
+}
+
+static void hix5hd2_i2c_reset(struct hix5hd2_i2c_priv *priv)
+{
+	clk_disable_unprepare(priv->clk);
+	msleep(20);
+	clk_prepare_enable(priv->clk);
+	hix5hd2_i2c_init(priv);
+}
+
+static int hix5hd2_i2c_wait_bus_idle(struct hix5hd2_i2c_priv *priv)
+{
+	unsigned long stop_time;
+	u32 int_status;
+
+	/* wait for 100 milli seconds for the bus to be idle */
+	stop_time = jiffies + msecs_to_jiffies(100);
+	do {
+		int_status = hix5hd2_i2c_clr_pend_irq(priv);
+		if (!(int_status & I2C_BUSY))
+			return 0;
+
+		usleep_range(50, 200);
+	} while (time_before(jiffies, stop_time));
+
+	return -EBUSY;
+}
+
+static void hix5hd2_rw_over(struct hix5hd2_i2c_priv *priv)
+{
+	if (priv->state == HIX5I2C_STAT_SND_STOP)
+		dev_dbg(priv->dev, "%s: rw and send stop over\n", __func__);
+	else
+		dev_dbg(priv->dev, "%s: have not data to send\n", __func__);
+
+	priv->state = HIX5I2C_STAT_RW_SUCCESS;
+	priv->err = 0;
+}
+
+static void hix5hd2_rw_handle_stop(struct hix5hd2_i2c_priv *priv)
+{
+	if (priv->stop) {
+		priv->state = HIX5I2C_STAT_SND_STOP;
+		writel_relaxed(I2C_STOP, priv->regs + HIX5I2C_COM);
+	} else {
+		hix5hd2_rw_over(priv);
+	}
+}
+
+static void hix5hd2_read_handle(struct hix5hd2_i2c_priv *priv)
+{
+	if (priv->msg_len == 1) {
+		/* the last byte don't need send ACK */
+		writel_relaxed(I2C_READ | I2C_NO_ACK, priv->regs + HIX5I2C_COM);
+	} else if (priv->msg_len > 1) {
+		/* if i2c master receive data will send ACK */
+		writel_relaxed(I2C_READ, priv->regs + HIX5I2C_COM);
+	} else {
+		hix5hd2_rw_handle_stop(priv);
+	}
+}
+
+static void hix5hd2_write_handle(struct hix5hd2_i2c_priv *priv)
+{
+	u8 data;
+
+	if (priv->msg_len > 0) {
+		data = priv->msg->buf[priv->msg_idx++];
+		writel_relaxed(data, priv->regs + HIX5I2C_TXR);
+		writel_relaxed(I2C_WRITE, priv->regs + HIX5I2C_COM);
+	} else {
+		hix5hd2_rw_handle_stop(priv);
+	}
+}
+
+static int hix5hd2_rw_preprocess(struct hix5hd2_i2c_priv *priv)
+{
+	u8 data;
+
+	if (priv->state == HIX5I2C_STAT_INIT) {
+		priv->state = HIX5I2C_STAT_RW;
+	} else if (priv->state == HIX5I2C_STAT_RW) {
+		if (priv->msg->flags & I2C_M_RD) {
+			data = readl_relaxed(priv->regs + HIX5I2C_RXR);
+			priv->msg->buf[priv->msg_idx++] = data;
+		}
+		priv->msg_len--;
+	} else {
+		dev_dbg(priv->dev, "%s: error: priv->state = %d, msg_len = %d\n",
+			__func__, priv->state, priv->msg_len);
+		return -EAGAIN;
+	}
+	return 0;
+}
+
+static irqreturn_t hix5hd2_i2c_irq(int irqno, void *dev_id)
+{
+	struct hix5hd2_i2c_priv *priv = dev_id;
+	u32 int_status;
+	int ret;
+
+	spin_lock(&priv->lock);
+
+	int_status = hix5hd2_i2c_clr_pend_irq(priv);
+
+	/* handle error */
+	if (int_status & I2C_ARBITRATE_INTR) {
+		/* bus error */
+		dev_dbg(priv->dev, "ARB bus loss\n");
+		priv->err = -EAGAIN;
+		priv->state = HIX5I2C_STAT_RW_ERR;
+		goto stop;
+	} else if (int_status & I2C_ACK_INTR) {
+		/* ack error */
+		dev_dbg(priv->dev, "No ACK from device\n");
+		priv->err = -ENXIO;
+		priv->state = HIX5I2C_STAT_RW_ERR;
+		goto stop;
+	}
+
+	if (int_status & I2C_OVER_INTR) {
+		if (priv->msg_len > 0) {
+			ret = hix5hd2_rw_preprocess(priv);
+			if (ret) {
+				priv->err = ret;
+				priv->state = HIX5I2C_STAT_RW_ERR;
+				goto stop;
+			}
+			if (priv->msg->flags & I2C_M_RD)
+				hix5hd2_read_handle(priv);
+			else
+				hix5hd2_write_handle(priv);
+		} else {
+			hix5hd2_rw_over(priv);
+		}
+	}
+
+stop:
+	if ((priv->state == HIX5I2C_STAT_RW_SUCCESS &&
+	     priv->msg->len == priv->msg_idx) ||
+	    (priv->state == HIX5I2C_STAT_RW_ERR)) {
+		hix5hd2_i2c_disable_irq(priv);
+		hix5hd2_i2c_clr_pend_irq(priv);
+		complete(&priv->msg_complete);
+	}
+
+	spin_unlock(&priv->lock);
+
+	return IRQ_HANDLED;
+}
+
+static void hix5hd2_i2c_message_start(struct hix5hd2_i2c_priv *priv, int stop)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&priv->lock, flags);
+	hix5hd2_i2c_clr_all_irq(priv);
+	hix5hd2_i2c_enable_irq(priv);
+
+	if (priv->msg->flags & I2C_M_RD)
+		writel_relaxed((priv->msg->addr << 1) | HIX5I2C_READ_OPERATION,
+			       priv->regs + HIX5I2C_TXR);
+	else
+		writel_relaxed(priv->msg->addr << 1,
+			       priv->regs + HIX5I2C_TXR);
+
+	writel_relaxed(I2C_WRITE | I2C_START, priv->regs + HIX5I2C_COM);
+	spin_unlock_irqrestore(&priv->lock, flags);
+}
+
+static int hix5hd2_i2c_xfer_msg(struct hix5hd2_i2c_priv *priv,
+				struct i2c_msg *msgs, int stop)
+{
+	unsigned long timeout;
+	int ret;
+
+	priv->msg = msgs;
+	priv->msg_idx = 0;
+	priv->msg_len = priv->msg->len;
+	priv->stop = stop;
+	priv->err = 0;
+	priv->state = HIX5I2C_STAT_INIT;
+
+	reinit_completion(&priv->msg_complete);
+	hix5hd2_i2c_message_start(priv, stop);
+
+	timeout = wait_for_completion_timeout(&priv->msg_complete,
+					      priv->adap.timeout);
+	if (timeout == 0) {
+		priv->state = HIX5I2C_STAT_RW_ERR;
+		priv->err = -ETIMEDOUT;
+		dev_warn(priv->dev, "%s timeout=%d\n",
+			 msgs->flags & I2C_M_RD ? "rx" : "tx",
+			 priv->adap.timeout);
+	}
+	ret = priv->state;
+
+	/*
+	 * If this is the last message to be transfered (stop == 1)
+	 * Then check if the bus can be brought back to idle.
+	 */
+	if (priv->state == HIX5I2C_STAT_RW_SUCCESS && stop)
+		ret = hix5hd2_i2c_wait_bus_idle(priv);
+
+	if (ret < 0)
+		hix5hd2_i2c_reset(priv);
+
+	return priv->err;
+}
+
+static int hix5hd2_i2c_xfer(struct i2c_adapter *adap,
+			    struct i2c_msg *msgs, int num)
+{
+	struct hix5hd2_i2c_priv *priv = i2c_get_adapdata(adap);
+	int i, ret, stop;
+
+	pm_runtime_get_sync(priv->dev);
+
+	for (i = 0; i < num; i++, msgs++) {
+		stop = (i == num - 1);
+		ret = hix5hd2_i2c_xfer_msg(priv, msgs, stop);
+		if (ret < 0)
+			goto out;
+	}
+
+	if (i == num) {
+		ret = num;
+	} else {
+		/* Only one message, cannot access the device */
+		if (i == 1)
+			ret = -EREMOTEIO;
+		else
+			ret = i;
+
+		dev_warn(priv->dev, "xfer message failed\n");
+	}
+
+out:
+	pm_runtime_mark_last_busy(priv->dev);
+	pm_runtime_put_autosuspend(priv->dev);
+	return ret;
+}
+
+static u32 hix5hd2_i2c_func(struct i2c_adapter *adap)
+{
+	return I2C_FUNC_I2C | (I2C_FUNC_SMBUS_EMUL & ~I2C_FUNC_SMBUS_QUICK);
+}
+
+static const struct i2c_algorithm hix5hd2_i2c_algorithm = {
+	.master_xfer		= hix5hd2_i2c_xfer,
+	.functionality		= hix5hd2_i2c_func,
+};
+
+static int hix5hd2_i2c_probe(struct platform_device *pdev)
+{
+	struct device_node *np = pdev->dev.of_node;
+	struct hix5hd2_i2c_priv *priv;
+	struct resource *mem;
+	unsigned int freq;
+	int irq, ret;
+
+	priv = devm_kzalloc(&pdev->dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	if (of_property_read_u32(np, "clock-frequency", &freq)) {
+		/* use 100k as default value */
+		priv->freq = 100000;
+	} else {
+		if (freq > HIX5I2C_MAX_FREQ) {
+			priv->freq = HIX5I2C_MAX_FREQ;
+			dev_warn(priv->dev, "use max freq %d instead\n",
+				 HIX5I2C_MAX_FREQ);
+		} else {
+			priv->freq = freq;
+		}
+	}
+
+	mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	priv->regs = devm_ioremap_resource(&pdev->dev, mem);
+	if (IS_ERR(priv->regs))
+		return PTR_ERR(priv->regs);
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq <= 0) {
+		dev_err(&pdev->dev, "cannot find HS-I2C IRQ\n");
+		return irq;
+	}
+
+	priv->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(priv->clk)) {
+		dev_err(&pdev->dev, "cannot get clock\n");
+		return PTR_ERR(priv->clk);
+	}
+	clk_prepare_enable(priv->clk);
+
+	strlcpy(priv->adap.name, "hix5hd2-i2c", sizeof(priv->adap.name));
+	priv->dev = &pdev->dev;
+	priv->adap.owner = THIS_MODULE;
+	priv->adap.algo = &hix5hd2_i2c_algorithm;
+	priv->adap.retries = 3;
+	priv->adap.dev.of_node = np;
+	priv->adap.algo_data = priv;
+	priv->adap.dev.parent = &pdev->dev;
+	i2c_set_adapdata(&priv->adap, priv);
+	platform_set_drvdata(pdev, priv);
+	spin_lock_init(&priv->lock);
+	init_completion(&priv->msg_complete);
+
+	hix5hd2_i2c_init(priv);
+
+	ret = devm_request_irq(&pdev->dev, irq, hix5hd2_i2c_irq,
+			       IRQF_NO_SUSPEND | IRQF_ONESHOT,
+			       dev_name(&pdev->dev), priv);
+	if (ret != 0) {
+		dev_err(&pdev->dev, "cannot request HS-I2C IRQ %d\n", irq);
+		goto err_clk;
+	}
+
+	pm_suspend_ignore_children(&pdev->dev, true);
+	pm_runtime_set_autosuspend_delay(priv->dev, MSEC_PER_SEC);
+	pm_runtime_use_autosuspend(priv->dev);
+	pm_runtime_set_active(priv->dev);
+	pm_runtime_enable(priv->dev);
+
+	ret = i2c_add_adapter(&priv->adap);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to add bus to i2c core\n");
+		goto err_runtime;
+	}
+
+	return ret;
+
+err_runtime:
+	pm_runtime_disable(priv->dev);
+	pm_runtime_set_suspended(priv->dev);
+err_clk:
+	clk_disable_unprepare(priv->clk);
+	return ret;
+}
+
+static int hix5hd2_i2c_remove(struct platform_device *pdev)
+{
+	struct hix5hd2_i2c_priv *priv = platform_get_drvdata(pdev);
+
+	i2c_del_adapter(&priv->adap);
+	pm_runtime_disable(priv->dev);
+	pm_runtime_set_suspended(priv->dev);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM
+static int hix5hd2_i2c_runtime_suspend(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct hix5hd2_i2c_priv *priv = platform_get_drvdata(pdev);
+
+	clk_disable_unprepare(priv->clk);
+
+	return 0;
+}
+
+static int hix5hd2_i2c_runtime_resume(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct hix5hd2_i2c_priv *priv = platform_get_drvdata(pdev);
+
+	clk_prepare_enable(priv->clk);
+	hix5hd2_i2c_init(priv);
+
+	return 0;
+}
+#endif
+
+static const struct dev_pm_ops hix5hd2_i2c_pm_ops = {
+	SET_RUNTIME_PM_OPS(hix5hd2_i2c_runtime_suspend,
+			      hix5hd2_i2c_runtime_resume,
+			      NULL)
+};
+
+static const struct of_device_id hix5hd2_i2c_match[] = {
+	{ .compatible = "hisilicon,hix5hd2-i2c" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, hix5hd2_i2c_match);
+
+static struct platform_driver hix5hd2_i2c_driver = {
+	.probe		= hix5hd2_i2c_probe,
+	.remove		= hix5hd2_i2c_remove,
+	.driver		= {
+		.name	= "hix5hd2-i2c",
+		.pm	= &hix5hd2_i2c_pm_ops,
+		.of_match_table = hix5hd2_i2c_match,
+	},
+};
+
+module_platform_driver(hix5hd2_i2c_driver);
+
+MODULE_DESCRIPTION("Hix5hd2 I2C Bus driver");
+MODULE_AUTHOR("Wei Yan <sledge.yanwei@huawei.com>");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:i2c-hix5hd2");
diff --git a/drivers/i2c/busses/i2c-hydra.c b/drivers/i2c/busses/i2c-hydra.c
index 14d2b76de25f..b7864cf42a72 100644
--- a/drivers/i2c/busses/i2c-hydra.c
+++ b/drivers/i2c/busses/i2c-hydra.c
@@ -15,10 +15,6 @@
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
     GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 #include <linux/kernel.h>
diff --git a/drivers/i2c/busses/i2c-i801.c b/drivers/i2c/busses/i2c-i801.c
index 10467a327749..8fafb254e42a 100644
--- a/drivers/i2c/busses/i2c-i801.c
+++ b/drivers/i2c/busses/i2c-i801.c
@@ -2,7 +2,7 @@
     Copyright (c) 1998 - 2002  Frodo Looijaard <frodol@dds.nl>,
     Philip Edelbrock <phil@netroedge.com>, and Mark D. Studebaker
     <mdsxyz123@yahoo.com>
-    Copyright (C) 2007 - 2012  Jean Delvare <jdelvare@suse.de>
+    Copyright (C) 2007 - 2014  Jean Delvare <jdelvare@suse.de>
     Copyright (C) 2010         Intel Corporation,
                                David Woodhouse <dwmw2@infradead.org>
 
@@ -15,10 +15,6 @@
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
     GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 /*
@@ -62,6 +58,8 @@
  * Wildcat Point (PCH)		0x8ca2	32	hard	yes	yes	yes
  * Wildcat Point-LP (PCH)	0x9ca2	32	hard	yes	yes	yes
  * BayTrail (SOC)		0x0f12	32	hard	yes	yes	yes
+ * Sunrise Point-H (PCH) 	0xa123  32	hard	yes	yes	yes
+ * Sunrise Point-LP (PCH)	0x9d23	32	hard	yes	yes	yes
  *
  * Features supported by this driver:
  * Software PEC				no
@@ -112,12 +110,16 @@
 
 /* PCI Address Constants */
 #define SMBBAR		4
+#define SMBPCICTL	0x004
 #define SMBPCISTS	0x006
 #define SMBHSTCFG	0x040
 
 /* Host status bits for SMBPCISTS */
 #define SMBPCISTS_INTS		0x08
 
+/* Control bits for SMBPCICTL */
+#define SMBPCICTL_INTDIS	0x0400
+
 /* Host configuration bits for SMBHSTCFG */
 #define SMBHSTCFG_HST_EN	1
 #define SMBHSTCFG_SMB_SMI_EN	2
@@ -184,6 +186,8 @@
 #define PCI_DEVICE_ID_INTEL_WELLSBURG_SMBUS_MS2		0x8d7f
 #define PCI_DEVICE_ID_INTEL_LYNXPOINT_LP_SMBUS		0x9c22
 #define PCI_DEVICE_ID_INTEL_WILDCATPOINT_LP_SMBUS	0x9ca2
+#define PCI_DEVICE_ID_INTEL_SUNRISEPOINT_H_SMBUS	0xa123
+#define PCI_DEVICE_ID_INTEL_SUNRISEPOINT_LP_SMBUS	0x9d23
 
 struct i801_mux_config {
 	char *gpio_chip;
@@ -373,6 +377,7 @@ static int i801_transaction(struct i801_priv *priv, int xact)
 {
 	int status;
 	int result;
+	const struct i2c_adapter *adap = &priv->adapter;
 
 	result = i801_check_pre(priv);
 	if (result < 0)
@@ -381,7 +386,14 @@ static int i801_transaction(struct i801_priv *priv, int xact)
 	if (priv->features & FEATURE_IRQ) {
 		outb_p(xact | SMBHSTCNT_INTREN | SMBHSTCNT_START,
 		       SMBHSTCNT(priv));
-		wait_event(priv->waitq, (status = priv->status));
+		result = wait_event_timeout(priv->waitq,
+					    (status = priv->status),
+					    adap->timeout);
+		if (!result) {
+			status = -ETIMEDOUT;
+			dev_warn(&priv->pci_dev->dev,
+				 "Timeout waiting for interrupt!\n");
+		}
 		priv->status = 0;
 		return i801_check_post(priv, status);
 	}
@@ -495,9 +507,6 @@ static irqreturn_t i801_isr(int irq, void *dev_id)
 		return IRQ_NONE;
 
 	status = inb_p(SMBHSTSTS(priv));
-	if (status != 0x42)
-		dev_dbg(&priv->pci_dev->dev, "irq: status = %02x\n", status);
-
 	if (status & SMBHSTSTS_BYTE_DONE)
 		i801_isr_byte_done(priv);
 
@@ -529,6 +538,7 @@ static int i801_block_transaction_byte_by_byte(struct i801_priv *priv,
 	int smbcmd;
 	int status;
 	int result;
+	const struct i2c_adapter *adap = &priv->adapter;
 
 	result = i801_check_pre(priv);
 	if (result < 0)
@@ -557,7 +567,14 @@ static int i801_block_transaction_byte_by_byte(struct i801_priv *priv,
 		priv->data = &data->block[1];
 
 		outb_p(priv->cmd | SMBHSTCNT_START, SMBHSTCNT(priv));
-		wait_event(priv->waitq, (status = priv->status));
+		result = wait_event_timeout(priv->waitq,
+					    (status = priv->status),
+					    adap->timeout);
+		if (!result) {
+			status = -ETIMEDOUT;
+			dev_warn(&priv->pci_dev->dev,
+				 "Timeout waiting for interrupt!\n");
+		}
 		priv->status = 0;
 		return i801_check_post(priv, status);
 	}
@@ -830,6 +847,8 @@ static const struct pci_device_id i801_ids[] = {
 	{ PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_WILDCATPOINT_LP_SMBUS) },
 	{ PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_BAYTRAIL_SMBUS) },
 	{ PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_BRASWELL_SMBUS) },
+	{ PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_SUNRISEPOINT_H_SMBUS) },
+	{ PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_SUNRISEPOINT_LP_SMBUS) },
 	{ 0, }
 };
 
@@ -1213,6 +1232,25 @@ static int i801_probe(struct pci_dev *dev, const struct pci_device_id *id)
 		outb_p(inb_p(SMBAUXCTL(priv)) &
 		       ~(SMBAUXCTL_CRC | SMBAUXCTL_E32B), SMBAUXCTL(priv));
 
+	/* Default timeout in interrupt mode: 200 ms */
+	priv->adapter.timeout = HZ / 5;
+
+	if (priv->features & FEATURE_IRQ) {
+		u16 pcictl, pcists;
+
+		/* Complain if an interrupt is already pending */
+		pci_read_config_word(priv->pci_dev, SMBPCISTS, &pcists);
+		if (pcists & SMBPCISTS_INTS)
+			dev_warn(&dev->dev, "An interrupt is pending!\n");
+
+		/* Check if interrupts have been disabled */
+		pci_read_config_word(priv->pci_dev, SMBPCICTL, &pcictl);
+		if (pcictl & SMBPCICTL_INTDIS) {
+			dev_info(&dev->dev, "Interrupts are disabled\n");
+			priv->features &= ~FEATURE_IRQ;
+		}
+	}
+
 	if (priv->features & FEATURE_IRQ) {
 		init_waitqueue_head(&priv->waitq);
 
@@ -1221,10 +1259,11 @@ static int i801_probe(struct pci_dev *dev, const struct pci_device_id *id)
 		if (err) {
 			dev_err(&dev->dev, "Failed to allocate irq %d: %d\n",
 				dev->irq, err);
-			goto exit_release;
+			priv->features &= ~FEATURE_IRQ;
 		}
-		dev_info(&dev->dev, "SMBus using PCI Interrupt\n");
 	}
+	dev_info(&dev->dev, "SMBus using %s\n",
+		 priv->features & FEATURE_IRQ ? "PCI interrupt" : "polling");
 
 	/* set up the sysfs linkage to our parent device */
 	priv->adapter.dev.parent = &dev->dev;
@@ -1251,7 +1290,6 @@ static int i801_probe(struct pci_dev *dev, const struct pci_device_id *id)
 exit_free_irq:
 	if (priv->features & FEATURE_IRQ)
 		free_irq(dev->irq, priv);
-exit_release:
 	pci_release_region(dev, SMBBAR);
 exit:
 	kfree(priv);
diff --git a/drivers/i2c/busses/i2c-ibm_iic.c b/drivers/i2c/busses/i2c-ibm_iic.c
index 274312c96b12..722f839cfa3c 100644
--- a/drivers/i2c/busses/i2c-ibm_iic.c
+++ b/drivers/i2c/busses/i2c-ibm_iic.c
@@ -802,7 +802,6 @@ static const struct of_device_id ibm_iic_match[] = {
 static struct platform_driver ibm_iic_driver = {
 	.driver = {
 		.name = "ibm-iic",
-		.owner = THIS_MODULE,
 		.of_match_table = ibm_iic_match,
 	},
 	.probe	= iic_probe,
diff --git a/drivers/i2c/busses/i2c-img-scb.c b/drivers/i2c/busses/i2c-img-scb.c
new file mode 100644
index 000000000000..0fcc1694c607
--- /dev/null
+++ b/drivers/i2c/busses/i2c-img-scb.c
@@ -0,0 +1,1412 @@
+/*
+ * I2C adapter for the IMG Serial Control Bus (SCB) IP block.
+ *
+ * Copyright (C) 2009, 2010, 2012, 2014 Imagination Technologies Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * There are three ways that this I2C controller can be driven:
+ *
+ * - Raw control of the SDA and SCK signals.
+ *
+ *   This corresponds to MODE_RAW, which takes control of the signals
+ *   directly for a certain number of clock cycles (the INT_TIMING
+ *   interrupt can be used for timing).
+ *
+ * - Atomic commands. A low level I2C symbol (such as generate
+ *   start/stop/ack/nack bit, generate byte, receive byte, and receive
+ *   ACK) is given to the hardware, with detection of completion by bits
+ *   in the LINESTAT register.
+ *
+ *   This mode of operation is used by MODE_ATOMIC, which uses an I2C
+ *   state machine in the interrupt handler to compose/react to I2C
+ *   transactions using atomic mode commands, and also by MODE_SEQUENCE,
+ *   which emits a simple fixed sequence of atomic mode commands.
+ *
+ *   Due to software control, the use of atomic commands usually results
+ *   in suboptimal use of the bus, with gaps between the I2C symbols while
+ *   the driver decides what to do next.
+ *
+ * - Automatic mode. A bus address, and whether to read/write is
+ *   specified, and the hardware takes care of the I2C state machine,
+ *   using a FIFO to send/receive bytes of data to an I2C slave. The
+ *   driver just has to keep the FIFO drained or filled in response to the
+ *   appropriate FIFO interrupts.
+ *
+ *   This corresponds to MODE_AUTOMATIC, which manages the FIFOs and deals
+ *   with control of repeated start bits between I2C messages.
+ *
+ *   Use of automatic mode and the FIFO can make much more efficient use
+ *   of the bus compared to individual atomic commands, with potentially
+ *   no wasted time between I2C symbols or I2C messages.
+ *
+ * In most cases MODE_AUTOMATIC is used, however if any of the messages in
+ * a transaction are zero byte writes (e.g. used by i2cdetect for probing
+ * the bus), MODE_ATOMIC must be used since automatic mode is normally
+ * started by the writing of data into the FIFO.
+ *
+ * The other modes are used in specific circumstances where MODE_ATOMIC and
+ * MODE_AUTOMATIC aren't appropriate. MODE_RAW is used to implement a bus
+ * recovery routine. MODE_SEQUENCE is used to reset the bus and make sure
+ * it is in a sane state.
+ *
+ * Notice that the driver implements a timer-based timeout mechanism.
+ * The reason for this mechanism is to reduce the number of interrupts
+ * received in automatic mode.
+ *
+ * The driver would get a slave event and transaction done interrupts for
+ * each atomic mode command that gets completed. However, these events are
+ * not needed in automatic mode, becase those atomic mode commands are
+ * managed automatically by the hardware.
+ *
+ * In practice, normal I2C transactions will be complete well before you
+ * get the timer interrupt, as the timer is re-scheduled during FIFO
+ * maintenance and disabled after the transaction is complete.
+ *
+ * In this way normal automatic mode operation isn't impacted by
+ * unnecessary interrupts, but the exceptional abort condition can still be
+ * detected (with a slight delay).
+ */
+
+#include <linux/bitops.h>
+#include <linux/clk.h>
+#include <linux/completion.h>
+#include <linux/err.h>
+#include <linux/i2c.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of_platform.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+#include <linux/timer.h>
+
+/* Register offsets */
+
+#define SCB_STATUS_REG			0x00
+#define SCB_OVERRIDE_REG		0x04
+#define SCB_READ_ADDR_REG		0x08
+#define SCB_READ_COUNT_REG		0x0c
+#define SCB_WRITE_ADDR_REG		0x10
+#define SCB_READ_DATA_REG		0x14
+#define SCB_WRITE_DATA_REG		0x18
+#define SCB_FIFO_STATUS_REG		0x1c
+#define SCB_CONTROL_SOFT_RESET		0x1f
+#define SCB_CLK_SET_REG			0x3c
+#define SCB_INT_STATUS_REG		0x40
+#define SCB_INT_CLEAR_REG		0x44
+#define SCB_INT_MASK_REG		0x48
+#define SCB_CONTROL_REG			0x4c
+#define SCB_TIME_TPL_REG		0x50
+#define SCB_TIME_TPH_REG		0x54
+#define SCB_TIME_TP2S_REG		0x58
+#define SCB_TIME_TBI_REG		0x60
+#define SCB_TIME_TSL_REG		0x64
+#define SCB_TIME_TDL_REG		0x68
+#define SCB_TIME_TSDL_REG		0x6c
+#define SCB_TIME_TSDH_REG		0x70
+#define SCB_READ_XADDR_REG		0x74
+#define SCB_WRITE_XADDR_REG		0x78
+#define SCB_WRITE_COUNT_REG		0x7c
+#define SCB_CORE_REV_REG		0x80
+#define SCB_TIME_TCKH_REG		0x84
+#define SCB_TIME_TCKL_REG		0x88
+#define SCB_FIFO_FLUSH_REG		0x8c
+#define SCB_READ_FIFO_REG		0x94
+#define SCB_CLEAR_REG			0x98
+
+/* SCB_CONTROL_REG bits */
+
+#define SCB_CONTROL_CLK_ENABLE		0x1e0
+#define SCB_CONTROL_TRANSACTION_HALT	0x200
+
+#define FIFO_READ_FULL			BIT(0)
+#define FIFO_READ_EMPTY			BIT(1)
+#define FIFO_WRITE_FULL			BIT(2)
+#define FIFO_WRITE_EMPTY		BIT(3)
+
+/* SCB_CLK_SET_REG bits */
+#define SCB_FILT_DISABLE		BIT(31)
+#define SCB_FILT_BYPASS			BIT(30)
+#define SCB_FILT_INC_MASK		0x7f
+#define SCB_FILT_INC_SHIFT		16
+#define SCB_INC_MASK			0x7f
+#define SCB_INC_SHIFT			8
+
+/* SCB_INT_*_REG bits */
+
+#define INT_BUS_INACTIVE		BIT(0)
+#define INT_UNEXPECTED_START		BIT(1)
+#define INT_SCLK_LOW_TIMEOUT		BIT(2)
+#define INT_SDAT_LOW_TIMEOUT		BIT(3)
+#define INT_WRITE_ACK_ERR		BIT(4)
+#define INT_ADDR_ACK_ERR		BIT(5)
+#define INT_FIFO_FULL			BIT(9)
+#define INT_FIFO_FILLING		BIT(10)
+#define INT_FIFO_EMPTY			BIT(11)
+#define INT_FIFO_EMPTYING		BIT(12)
+#define INT_TRANSACTION_DONE		BIT(15)
+#define INT_SLAVE_EVENT			BIT(16)
+#define INT_TIMING			BIT(18)
+
+#define INT_FIFO_FULL_FILLING	(INT_FIFO_FULL  | INT_FIFO_FILLING)
+#define INT_FIFO_EMPTY_EMPTYING	(INT_FIFO_EMPTY | INT_FIFO_EMPTYING)
+
+/* Level interrupts need clearing after handling instead of before */
+#define INT_LEVEL			0x01e00
+
+/* Don't allow any interrupts while the clock may be off */
+#define INT_ENABLE_MASK_INACTIVE	0x00000
+
+/* Interrupt masks for the different driver modes */
+
+#define INT_ENABLE_MASK_RAW		INT_TIMING
+
+#define INT_ENABLE_MASK_ATOMIC		(INT_TRANSACTION_DONE | \
+					 INT_SLAVE_EVENT      | \
+					 INT_ADDR_ACK_ERR     | \
+					 INT_WRITE_ACK_ERR)
+
+#define INT_ENABLE_MASK_AUTOMATIC	(INT_SCLK_LOW_TIMEOUT | \
+					 INT_ADDR_ACK_ERR     | \
+					 INT_WRITE_ACK_ERR    | \
+					 INT_FIFO_FULL        | \
+					 INT_FIFO_FILLING     | \
+					 INT_FIFO_EMPTY       | \
+					 INT_FIFO_EMPTYING)
+
+#define INT_ENABLE_MASK_WAITSTOP	(INT_SLAVE_EVENT      | \
+					 INT_ADDR_ACK_ERR     | \
+					 INT_WRITE_ACK_ERR)
+
+/* SCB_STATUS_REG fields */
+
+#define LINESTAT_SCLK_LINE_STATUS	BIT(0)
+#define LINESTAT_SCLK_EN		BIT(1)
+#define LINESTAT_SDAT_LINE_STATUS	BIT(2)
+#define LINESTAT_SDAT_EN		BIT(3)
+#define LINESTAT_DET_START_STATUS	BIT(4)
+#define LINESTAT_DET_STOP_STATUS	BIT(5)
+#define LINESTAT_DET_ACK_STATUS		BIT(6)
+#define LINESTAT_DET_NACK_STATUS	BIT(7)
+#define LINESTAT_BUS_IDLE		BIT(8)
+#define LINESTAT_T_DONE_STATUS		BIT(9)
+#define LINESTAT_SCLK_OUT_STATUS	BIT(10)
+#define LINESTAT_SDAT_OUT_STATUS	BIT(11)
+#define LINESTAT_GEN_LINE_MASK_STATUS	BIT(12)
+#define LINESTAT_START_BIT_DET		BIT(13)
+#define LINESTAT_STOP_BIT_DET		BIT(14)
+#define LINESTAT_ACK_DET		BIT(15)
+#define LINESTAT_NACK_DET		BIT(16)
+#define LINESTAT_INPUT_HELD_V		BIT(17)
+#define LINESTAT_ABORT_DET		BIT(18)
+#define LINESTAT_ACK_OR_NACK_DET	(LINESTAT_ACK_DET | LINESTAT_NACK_DET)
+#define LINESTAT_INPUT_DATA		0xff000000
+#define LINESTAT_INPUT_DATA_SHIFT	24
+
+#define LINESTAT_CLEAR_SHIFT		13
+#define LINESTAT_LATCHED		(0x3f << LINESTAT_CLEAR_SHIFT)
+
+/* SCB_OVERRIDE_REG fields */
+
+#define OVERRIDE_SCLK_OVR		BIT(0)
+#define OVERRIDE_SCLKEN_OVR		BIT(1)
+#define OVERRIDE_SDAT_OVR		BIT(2)
+#define OVERRIDE_SDATEN_OVR		BIT(3)
+#define OVERRIDE_MASTER			BIT(9)
+#define OVERRIDE_LINE_OVR_EN		BIT(10)
+#define OVERRIDE_DIRECT			BIT(11)
+#define OVERRIDE_CMD_SHIFT		4
+#define OVERRIDE_CMD_MASK		0x1f
+#define OVERRIDE_DATA_SHIFT		24
+
+#define OVERRIDE_SCLK_DOWN		(OVERRIDE_LINE_OVR_EN | \
+					 OVERRIDE_SCLKEN_OVR)
+#define OVERRIDE_SCLK_UP		(OVERRIDE_LINE_OVR_EN | \
+					 OVERRIDE_SCLKEN_OVR | \
+					 OVERRIDE_SCLK_OVR)
+#define OVERRIDE_SDAT_DOWN		(OVERRIDE_LINE_OVR_EN | \
+					 OVERRIDE_SDATEN_OVR)
+#define OVERRIDE_SDAT_UP		(OVERRIDE_LINE_OVR_EN | \
+					 OVERRIDE_SDATEN_OVR | \
+					 OVERRIDE_SDAT_OVR)
+
+/* OVERRIDE_CMD values */
+
+#define CMD_PAUSE			0x00
+#define CMD_GEN_DATA			0x01
+#define CMD_GEN_START			0x02
+#define CMD_GEN_STOP			0x03
+#define CMD_GEN_ACK			0x04
+#define CMD_GEN_NACK			0x05
+#define CMD_RET_DATA			0x08
+#define CMD_RET_ACK			0x09
+
+/* Fixed timing values */
+
+#define TIMEOUT_TBI			0x0
+#define TIMEOUT_TSL			0xffff
+#define TIMEOUT_TDL			0x0
+
+/* Transaction timeout */
+
+#define IMG_I2C_TIMEOUT			(msecs_to_jiffies(1000))
+
+/*
+ * Worst incs are 1 (innacurate) and 16*256 (irregular).
+ * So a sensible inc is the logarithmic mean: 64 (2^6), which is
+ * in the middle of the valid range (0-127).
+ */
+#define SCB_OPT_INC		64
+
+/* Setup the clock enable filtering for 25 ns */
+#define SCB_FILT_GLITCH		25
+
+/*
+ * Bits to return from interrupt handler functions for different modes.
+ * This delays completion until we've finished with the registers, so that the
+ * function waiting for completion can safely disable the clock to save power.
+ */
+#define ISR_COMPLETE_M		BIT(31)
+#define ISR_FATAL_M		BIT(30)
+#define ISR_WAITSTOP		BIT(29)
+#define ISR_STATUS_M		0x0000ffff	/* contains +ve errno */
+#define ISR_COMPLETE(err)	(ISR_COMPLETE_M | (ISR_STATUS_M & (err)))
+#define ISR_FATAL(err)		(ISR_COMPLETE(err) | ISR_FATAL_M)
+
+#define REL_SOC_IP_SCB_2_2_1	0x00020201
+
+enum img_i2c_mode {
+	MODE_INACTIVE,
+	MODE_RAW,
+	MODE_ATOMIC,
+	MODE_AUTOMATIC,
+	MODE_SEQUENCE,
+	MODE_FATAL,
+	MODE_WAITSTOP,
+	MODE_SUSPEND,
+};
+
+/* Timing parameters for i2c modes (in ns) */
+struct img_i2c_timings {
+	const char *name;
+	unsigned int max_bitrate;
+	unsigned int tckh, tckl, tsdh, tsdl;
+	unsigned int tp2s, tpl, tph;
+};
+
+/* The timings array must be ordered from slower to faster */
+static struct img_i2c_timings timings[] = {
+	/* Standard mode */
+	{
+		.name = "standard",
+		.max_bitrate = 100000,
+		.tckh = 4000,
+		.tckl = 4700,
+		.tsdh = 4700,
+		.tsdl = 8700,
+		.tp2s = 4700,
+		.tpl = 4700,
+		.tph = 4000,
+	},
+	/* Fast mode */
+	{
+		.name = "fast",
+		.max_bitrate = 400000,
+		.tckh = 600,
+		.tckl = 1300,
+		.tsdh = 600,
+		.tsdl = 1200,
+		.tp2s = 1300,
+		.tpl = 600,
+		.tph = 600,
+	},
+};
+
+/* Reset dance */
+static u8 img_i2c_reset_seq[] = { CMD_GEN_START,
+				  CMD_GEN_DATA, 0xff,
+				  CMD_RET_ACK,
+				  CMD_GEN_START,
+				  CMD_GEN_STOP,
+				  0 };
+/* Just issue a stop (after an abort condition) */
+static u8 img_i2c_stop_seq[] = {  CMD_GEN_STOP,
+				  0 };
+
+/* We're interested in different interrupts depending on the mode */
+static unsigned int img_i2c_int_enable_by_mode[] = {
+	[MODE_INACTIVE]  = INT_ENABLE_MASK_INACTIVE,
+	[MODE_RAW]       = INT_ENABLE_MASK_RAW,
+	[MODE_ATOMIC]    = INT_ENABLE_MASK_ATOMIC,
+	[MODE_AUTOMATIC] = INT_ENABLE_MASK_AUTOMATIC,
+	[MODE_SEQUENCE]  = INT_ENABLE_MASK_ATOMIC,
+	[MODE_FATAL]     = 0,
+	[MODE_WAITSTOP]  = INT_ENABLE_MASK_WAITSTOP,
+	[MODE_SUSPEND]   = 0,
+};
+
+/* Atomic command names */
+static const char * const img_i2c_atomic_cmd_names[] = {
+	[CMD_PAUSE]	= "PAUSE",
+	[CMD_GEN_DATA]	= "GEN_DATA",
+	[CMD_GEN_START]	= "GEN_START",
+	[CMD_GEN_STOP]	= "GEN_STOP",
+	[CMD_GEN_ACK]	= "GEN_ACK",
+	[CMD_GEN_NACK]	= "GEN_NACK",
+	[CMD_RET_DATA]	= "RET_DATA",
+	[CMD_RET_ACK]	= "RET_ACK",
+};
+
+struct img_i2c {
+	struct i2c_adapter adap;
+
+	void __iomem *base;
+
+	/*
+	 * The scb core clock is used to get the input frequency, and to disable
+	 * it after every set of transactions to save some power.
+	 */
+	struct clk *scb_clk, *sys_clk;
+	unsigned int bitrate;
+	bool need_wr_rd_fence;
+
+	/* state */
+	struct completion msg_complete;
+	spinlock_t lock;	/* lock before doing anything with the state */
+	struct i2c_msg msg;
+
+	/* After the last transaction, wait for a stop bit */
+	bool last_msg;
+	int msg_status;
+
+	enum img_i2c_mode mode;
+	u32 int_enable;		/* depends on mode */
+	u32 line_status;	/* line status over command */
+
+	/*
+	 * To avoid slave event interrupts in automatic mode, use a timer to
+	 * poll the abort condition if we don't get an interrupt for too long.
+	 */
+	struct timer_list check_timer;
+	bool t_halt;
+
+	/* atomic mode state */
+	bool at_t_done;
+	bool at_slave_event;
+	int at_cur_cmd;
+	u8 at_cur_data;
+
+	/* Sequence: either reset or stop. See img_i2c_sequence. */
+	u8 *seq;
+
+	/* raw mode */
+	unsigned int raw_timeout;
+};
+
+static void img_i2c_writel(struct img_i2c *i2c, u32 offset, u32 value)
+{
+	writel(value, i2c->base + offset);
+}
+
+static u32 img_i2c_readl(struct img_i2c *i2c, u32 offset)
+{
+	return readl(i2c->base + offset);
+}
+
+/*
+ * The code to read from the master read fifo, and write to the master
+ * write fifo, checks a bit in an SCB register before every byte to
+ * ensure that the fifo is not full (write fifo) or empty (read fifo).
+ * Due to clock domain crossing inside the SCB block the updated value
+ * of this bit is only visible after 2 cycles.
+ *
+ * The scb_wr_rd_fence() function does 2 dummy writes (to the read-only
+ * revision register), and it's called after reading from or writing to the
+ * fifos to ensure that subsequent reads of the fifo status bits do not read
+ * stale values.
+ */
+static void img_i2c_wr_rd_fence(struct img_i2c *i2c)
+{
+	if (i2c->need_wr_rd_fence) {
+		img_i2c_writel(i2c, SCB_CORE_REV_REG, 0);
+		img_i2c_writel(i2c, SCB_CORE_REV_REG, 0);
+	}
+}
+
+static void img_i2c_switch_mode(struct img_i2c *i2c, enum img_i2c_mode mode)
+{
+	i2c->mode = mode;
+	i2c->int_enable = img_i2c_int_enable_by_mode[mode];
+	i2c->line_status = 0;
+}
+
+static void img_i2c_raw_op(struct img_i2c *i2c)
+{
+	i2c->raw_timeout = 0;
+	img_i2c_writel(i2c, SCB_OVERRIDE_REG,
+		OVERRIDE_SCLKEN_OVR |
+		OVERRIDE_SDATEN_OVR |
+		OVERRIDE_MASTER |
+		OVERRIDE_LINE_OVR_EN |
+		OVERRIDE_DIRECT |
+		((i2c->at_cur_cmd & OVERRIDE_CMD_MASK) << OVERRIDE_CMD_SHIFT) |
+		(i2c->at_cur_data << OVERRIDE_DATA_SHIFT));
+}
+
+static const char *img_i2c_atomic_op_name(unsigned int cmd)
+{
+	if (unlikely(cmd >= ARRAY_SIZE(img_i2c_atomic_cmd_names)))
+		return "UNKNOWN";
+	return img_i2c_atomic_cmd_names[cmd];
+}
+
+/* Send a single atomic mode command to the hardware */
+static void img_i2c_atomic_op(struct img_i2c *i2c, int cmd, u8 data)
+{
+	i2c->at_cur_cmd = cmd;
+	i2c->at_cur_data = data;
+
+	/* work around lack of data setup time when generating data */
+	if (cmd == CMD_GEN_DATA && i2c->mode == MODE_ATOMIC) {
+		u32 line_status = img_i2c_readl(i2c, SCB_STATUS_REG);
+
+		if (line_status & LINESTAT_SDAT_LINE_STATUS && !(data & 0x80)) {
+			/* hold the data line down for a moment */
+			img_i2c_switch_mode(i2c, MODE_RAW);
+			img_i2c_raw_op(i2c);
+			return;
+		}
+	}
+
+	dev_dbg(i2c->adap.dev.parent,
+		"atomic cmd=%s (%d) data=%#x\n",
+		img_i2c_atomic_op_name(cmd), cmd, data);
+	i2c->at_t_done = (cmd == CMD_RET_DATA || cmd == CMD_RET_ACK);
+	i2c->at_slave_event = false;
+	i2c->line_status = 0;
+
+	img_i2c_writel(i2c, SCB_OVERRIDE_REG,
+		((cmd & OVERRIDE_CMD_MASK) << OVERRIDE_CMD_SHIFT) |
+		OVERRIDE_MASTER |
+		OVERRIDE_DIRECT |
+		(data << OVERRIDE_DATA_SHIFT));
+}
+
+/* Start a transaction in atomic mode */
+static void img_i2c_atomic_start(struct img_i2c *i2c)
+{
+	img_i2c_switch_mode(i2c, MODE_ATOMIC);
+	img_i2c_writel(i2c, SCB_INT_MASK_REG, i2c->int_enable);
+	img_i2c_atomic_op(i2c, CMD_GEN_START, 0x00);
+}
+
+static void img_i2c_soft_reset(struct img_i2c *i2c)
+{
+	i2c->t_halt = false;
+	img_i2c_writel(i2c, SCB_CONTROL_REG, 0);
+	img_i2c_writel(i2c, SCB_CONTROL_REG,
+		       SCB_CONTROL_CLK_ENABLE | SCB_CONTROL_SOFT_RESET);
+}
+
+/* enable or release transaction halt for control of repeated starts */
+static void img_i2c_transaction_halt(struct img_i2c *i2c, bool t_halt)
+{
+	u32 val;
+
+	if (i2c->t_halt == t_halt)
+		return;
+	i2c->t_halt = t_halt;
+	val = img_i2c_readl(i2c, SCB_CONTROL_REG);
+	if (t_halt)
+		val |= SCB_CONTROL_TRANSACTION_HALT;
+	else
+		val &= ~SCB_CONTROL_TRANSACTION_HALT;
+	img_i2c_writel(i2c, SCB_CONTROL_REG, val);
+}
+
+/* Drain data from the FIFO into the buffer (automatic mode) */
+static void img_i2c_read_fifo(struct img_i2c *i2c)
+{
+	while (i2c->msg.len) {
+		u32 fifo_status;
+		u8 data;
+
+		fifo_status = img_i2c_readl(i2c, SCB_FIFO_STATUS_REG);
+		if (fifo_status & FIFO_READ_EMPTY)
+			break;
+
+		data = img_i2c_readl(i2c, SCB_READ_DATA_REG);
+		*i2c->msg.buf = data;
+
+		img_i2c_writel(i2c, SCB_READ_FIFO_REG, 0xff);
+		img_i2c_wr_rd_fence(i2c);
+		i2c->msg.len--;
+		i2c->msg.buf++;
+	}
+}
+
+/* Fill the FIFO with data from the buffer (automatic mode) */
+static void img_i2c_write_fifo(struct img_i2c *i2c)
+{
+	while (i2c->msg.len) {
+		u32 fifo_status;
+
+		fifo_status = img_i2c_readl(i2c, SCB_FIFO_STATUS_REG);
+		if (fifo_status & FIFO_WRITE_FULL)
+			break;
+
+		img_i2c_writel(i2c, SCB_WRITE_DATA_REG, *i2c->msg.buf);
+		img_i2c_wr_rd_fence(i2c);
+		i2c->msg.len--;
+		i2c->msg.buf++;
+	}
+
+	/* Disable fifo emptying interrupt if nothing more to write */
+	if (!i2c->msg.len)
+		i2c->int_enable &= ~INT_FIFO_EMPTYING;
+}
+
+/* Start a read transaction in automatic mode */
+static void img_i2c_read(struct img_i2c *i2c)
+{
+	img_i2c_switch_mode(i2c, MODE_AUTOMATIC);
+	if (!i2c->last_msg)
+		i2c->int_enable |= INT_SLAVE_EVENT;
+
+	img_i2c_writel(i2c, SCB_INT_MASK_REG, i2c->int_enable);
+	img_i2c_writel(i2c, SCB_READ_ADDR_REG, i2c->msg.addr);
+	img_i2c_writel(i2c, SCB_READ_COUNT_REG, i2c->msg.len);
+
+	img_i2c_transaction_halt(i2c, false);
+	mod_timer(&i2c->check_timer, jiffies + msecs_to_jiffies(1));
+}
+
+/* Start a write transaction in automatic mode */
+static void img_i2c_write(struct img_i2c *i2c)
+{
+	img_i2c_switch_mode(i2c, MODE_AUTOMATIC);
+	if (!i2c->last_msg)
+		i2c->int_enable |= INT_SLAVE_EVENT;
+
+	img_i2c_writel(i2c, SCB_WRITE_ADDR_REG, i2c->msg.addr);
+	img_i2c_writel(i2c, SCB_WRITE_COUNT_REG, i2c->msg.len);
+
+	img_i2c_transaction_halt(i2c, false);
+	mod_timer(&i2c->check_timer, jiffies + msecs_to_jiffies(1));
+	img_i2c_write_fifo(i2c);
+
+	/* img_i2c_write_fifo() may modify int_enable */
+	img_i2c_writel(i2c, SCB_INT_MASK_REG, i2c->int_enable);
+}
+
+/*
+ * Indicate that the transaction is complete. This is called from the
+ * ISR to wake up the waiting thread, after which the ISR must not
+ * access any more SCB registers.
+ */
+static void img_i2c_complete_transaction(struct img_i2c *i2c, int status)
+{
+	img_i2c_switch_mode(i2c, MODE_INACTIVE);
+	if (status) {
+		i2c->msg_status = status;
+		img_i2c_transaction_halt(i2c, false);
+	}
+	complete(&i2c->msg_complete);
+}
+
+static unsigned int img_i2c_raw_atomic_delay_handler(struct img_i2c *i2c,
+					u32 int_status, u32 line_status)
+{
+	/* Stay in raw mode for this, so we don't just loop infinitely */
+	img_i2c_atomic_op(i2c, i2c->at_cur_cmd, i2c->at_cur_data);
+	img_i2c_switch_mode(i2c, MODE_ATOMIC);
+	return 0;
+}
+
+static unsigned int img_i2c_raw(struct img_i2c *i2c, u32 int_status,
+				u32 line_status)
+{
+	if (int_status & INT_TIMING) {
+		if (i2c->raw_timeout == 0)
+			return img_i2c_raw_atomic_delay_handler(i2c,
+				int_status, line_status);
+		--i2c->raw_timeout;
+	}
+	return 0;
+}
+
+static unsigned int img_i2c_sequence(struct img_i2c *i2c, u32 int_status)
+{
+	static const unsigned int continue_bits[] = {
+		[CMD_GEN_START] = LINESTAT_START_BIT_DET,
+		[CMD_GEN_DATA]  = LINESTAT_INPUT_HELD_V,
+		[CMD_RET_ACK]   = LINESTAT_ACK_DET | LINESTAT_NACK_DET,
+		[CMD_RET_DATA]  = LINESTAT_INPUT_HELD_V,
+		[CMD_GEN_STOP]  = LINESTAT_STOP_BIT_DET,
+	};
+	int next_cmd = -1;
+	u8 next_data = 0x00;
+
+	if (int_status & INT_SLAVE_EVENT)
+		i2c->at_slave_event = true;
+	if (int_status & INT_TRANSACTION_DONE)
+		i2c->at_t_done = true;
+
+	if (!i2c->at_slave_event || !i2c->at_t_done)
+		return 0;
+
+	/* wait if no continue bits are set */
+	if (i2c->at_cur_cmd >= 0 &&
+	    i2c->at_cur_cmd < ARRAY_SIZE(continue_bits)) {
+		unsigned int cont_bits = continue_bits[i2c->at_cur_cmd];
+
+		if (cont_bits) {
+			cont_bits |= LINESTAT_ABORT_DET;
+			if (!(i2c->line_status & cont_bits))
+				return 0;
+		}
+	}
+
+	/* follow the sequence of commands in i2c->seq */
+	next_cmd = *i2c->seq;
+	/* stop on a nil */
+	if (!next_cmd) {
+		img_i2c_writel(i2c, SCB_OVERRIDE_REG, 0);
+		return ISR_COMPLETE(0);
+	}
+	/* when generating data, the next byte is the data */
+	if (next_cmd == CMD_GEN_DATA) {
+		++i2c->seq;
+		next_data = *i2c->seq;
+	}
+	++i2c->seq;
+	img_i2c_atomic_op(i2c, next_cmd, next_data);
+
+	return 0;
+}
+
+static void img_i2c_reset_start(struct img_i2c *i2c)
+{
+	/* Initiate the magic dance */
+	img_i2c_switch_mode(i2c, MODE_SEQUENCE);
+	img_i2c_writel(i2c, SCB_INT_MASK_REG, i2c->int_enable);
+	i2c->seq = img_i2c_reset_seq;
+	i2c->at_slave_event = true;
+	i2c->at_t_done = true;
+	i2c->at_cur_cmd = -1;
+
+	/* img_i2c_reset_seq isn't empty so the following won't fail */
+	img_i2c_sequence(i2c, 0);
+}
+
+static void img_i2c_stop_start(struct img_i2c *i2c)
+{
+	/* Initiate a stop bit sequence */
+	img_i2c_switch_mode(i2c, MODE_SEQUENCE);
+	img_i2c_writel(i2c, SCB_INT_MASK_REG, i2c->int_enable);
+	i2c->seq = img_i2c_stop_seq;
+	i2c->at_slave_event = true;
+	i2c->at_t_done = true;
+	i2c->at_cur_cmd = -1;
+
+	/* img_i2c_stop_seq isn't empty so the following won't fail */
+	img_i2c_sequence(i2c, 0);
+}
+
+static unsigned int img_i2c_atomic(struct img_i2c *i2c,
+				   u32 int_status,
+				   u32 line_status)
+{
+	int next_cmd = -1;
+	u8 next_data = 0x00;
+
+	if (int_status & INT_SLAVE_EVENT)
+		i2c->at_slave_event = true;
+	if (int_status & INT_TRANSACTION_DONE)
+		i2c->at_t_done = true;
+
+	if (!i2c->at_slave_event || !i2c->at_t_done)
+		goto next_atomic_cmd;
+	if (i2c->line_status & LINESTAT_ABORT_DET) {
+		dev_dbg(i2c->adap.dev.parent, "abort condition detected\n");
+		next_cmd = CMD_GEN_STOP;
+		i2c->msg_status = -EIO;
+		goto next_atomic_cmd;
+	}
+
+	/* i2c->at_cur_cmd may have completed */
+	switch (i2c->at_cur_cmd) {
+	case CMD_GEN_START:
+		next_cmd = CMD_GEN_DATA;
+		next_data = (i2c->msg.addr << 1);
+		if (i2c->msg.flags & I2C_M_RD)
+			next_data |= 0x1;
+		break;
+	case CMD_GEN_DATA:
+		if (i2c->line_status & LINESTAT_INPUT_HELD_V)
+			next_cmd = CMD_RET_ACK;
+		break;
+	case CMD_RET_ACK:
+		if (i2c->line_status & LINESTAT_ACK_DET) {
+			if (i2c->msg.len == 0) {
+				next_cmd = CMD_GEN_STOP;
+			} else if (i2c->msg.flags & I2C_M_RD) {
+				next_cmd = CMD_RET_DATA;
+			} else {
+				next_cmd = CMD_GEN_DATA;
+				next_data = *i2c->msg.buf;
+				--i2c->msg.len;
+				++i2c->msg.buf;
+			}
+		} else if (i2c->line_status & LINESTAT_NACK_DET) {
+			i2c->msg_status = -EIO;
+			next_cmd = CMD_GEN_STOP;
+		}
+		break;
+	case CMD_RET_DATA:
+		if (i2c->line_status & LINESTAT_INPUT_HELD_V) {
+			*i2c->msg.buf = (i2c->line_status &
+						LINESTAT_INPUT_DATA)
+					>> LINESTAT_INPUT_DATA_SHIFT;
+			--i2c->msg.len;
+			++i2c->msg.buf;
+			if (i2c->msg.len)
+				next_cmd = CMD_GEN_ACK;
+			else
+				next_cmd = CMD_GEN_NACK;
+		}
+		break;
+	case CMD_GEN_ACK:
+		if (i2c->line_status & LINESTAT_ACK_DET) {
+			next_cmd = CMD_RET_DATA;
+		} else {
+			i2c->msg_status = -EIO;
+			next_cmd = CMD_GEN_STOP;
+		}
+		break;
+	case CMD_GEN_NACK:
+		next_cmd = CMD_GEN_STOP;
+		break;
+	case CMD_GEN_STOP:
+		img_i2c_writel(i2c, SCB_OVERRIDE_REG, 0);
+		return ISR_COMPLETE(0);
+	default:
+		dev_err(i2c->adap.dev.parent, "bad atomic command %d\n",
+			i2c->at_cur_cmd);
+		i2c->msg_status = -EIO;
+		next_cmd = CMD_GEN_STOP;
+		break;
+	}
+
+next_atomic_cmd:
+	if (next_cmd != -1) {
+		/* don't actually stop unless we're the last transaction */
+		if (next_cmd == CMD_GEN_STOP && !i2c->msg_status &&
+						!i2c->last_msg)
+			return ISR_COMPLETE(0);
+		img_i2c_atomic_op(i2c, next_cmd, next_data);
+	}
+	return 0;
+}
+
+/*
+ * Timer function to check if something has gone wrong in automatic mode (so we
+ * don't have to handle so many interrupts just to catch an exception).
+ */
+static void img_i2c_check_timer(unsigned long arg)
+{
+	struct img_i2c *i2c = (struct img_i2c *)arg;
+	unsigned long flags;
+	unsigned int line_status;
+
+	spin_lock_irqsave(&i2c->lock, flags);
+	line_status = img_i2c_readl(i2c, SCB_STATUS_REG);
+
+	/* check for an abort condition */
+	if (line_status & LINESTAT_ABORT_DET) {
+		dev_dbg(i2c->adap.dev.parent,
+			"abort condition detected by check timer\n");
+		/* enable slave event interrupt mask to trigger irq */
+		img_i2c_writel(i2c, SCB_INT_MASK_REG,
+			       i2c->int_enable | INT_SLAVE_EVENT);
+	}
+
+	spin_unlock_irqrestore(&i2c->lock, flags);
+}
+
+static unsigned int img_i2c_auto(struct img_i2c *i2c,
+				 unsigned int int_status,
+				 unsigned int line_status)
+{
+	if (int_status & (INT_WRITE_ACK_ERR | INT_ADDR_ACK_ERR))
+		return ISR_COMPLETE(EIO);
+
+	if (line_status & LINESTAT_ABORT_DET) {
+		dev_dbg(i2c->adap.dev.parent, "abort condition detected\n");
+		/* empty the read fifo */
+		if ((i2c->msg.flags & I2C_M_RD) &&
+		    (int_status & INT_FIFO_FULL_FILLING))
+			img_i2c_read_fifo(i2c);
+		/* use atomic mode and try to force a stop bit */
+		i2c->msg_status = -EIO;
+		img_i2c_stop_start(i2c);
+		return 0;
+	}
+
+	/* Enable transaction halt on start bit */
+	if (!i2c->last_msg && i2c->line_status & LINESTAT_START_BIT_DET) {
+		img_i2c_transaction_halt(i2c, true);
+		/* we're no longer interested in the slave event */
+		i2c->int_enable &= ~INT_SLAVE_EVENT;
+	}
+
+	mod_timer(&i2c->check_timer, jiffies + msecs_to_jiffies(1));
+
+	if (i2c->msg.flags & I2C_M_RD) {
+		if (int_status & INT_FIFO_FULL_FILLING) {
+			img_i2c_read_fifo(i2c);
+			if (i2c->msg.len == 0)
+				return ISR_WAITSTOP;
+		}
+	} else {
+		if (int_status & INT_FIFO_EMPTY_EMPTYING) {
+			/*
+			 * The write fifo empty indicates that we're in the
+			 * last byte so it's safe to start a new write
+			 * transaction without losing any bytes from the
+			 * previous one.
+			 * see 2.3.7 Repeated Start Transactions.
+			 */
+			if ((int_status & INT_FIFO_EMPTY) &&
+			    i2c->msg.len == 0)
+				return ISR_WAITSTOP;
+			img_i2c_write_fifo(i2c);
+		}
+	}
+
+	return 0;
+}
+
+static irqreturn_t img_i2c_isr(int irq, void *dev_id)
+{
+	struct img_i2c *i2c = (struct img_i2c *)dev_id;
+	u32 int_status, line_status;
+	/* We handle transaction completion AFTER accessing registers */
+	unsigned int hret;
+
+	/* Read interrupt status register. */
+	int_status = img_i2c_readl(i2c, SCB_INT_STATUS_REG);
+	/* Clear detected interrupts. */
+	img_i2c_writel(i2c, SCB_INT_CLEAR_REG, int_status);
+
+	/*
+	 * Read line status and clear it until it actually is clear.  We have
+	 * to be careful not to lose any line status bits that get latched.
+	 */
+	line_status = img_i2c_readl(i2c, SCB_STATUS_REG);
+	if (line_status & LINESTAT_LATCHED) {
+		img_i2c_writel(i2c, SCB_CLEAR_REG,
+			      (line_status & LINESTAT_LATCHED)
+				>> LINESTAT_CLEAR_SHIFT);
+		img_i2c_wr_rd_fence(i2c);
+	}
+
+	spin_lock(&i2c->lock);
+
+	/* Keep track of line status bits received */
+	i2c->line_status &= ~LINESTAT_INPUT_DATA;
+	i2c->line_status |= line_status;
+
+	/*
+	 * Certain interrupts indicate that sclk low timeout is not
+	 * a problem. If any of these are set, just continue.
+	 */
+	if ((int_status & INT_SCLK_LOW_TIMEOUT) &&
+	    !(int_status & (INT_SLAVE_EVENT |
+			    INT_FIFO_EMPTY |
+			    INT_FIFO_FULL))) {
+		dev_crit(i2c->adap.dev.parent,
+			 "fatal: clock low timeout occurred %s addr 0x%02x\n",
+			 (i2c->msg.flags & I2C_M_RD) ? "reading" : "writing",
+			 i2c->msg.addr);
+		hret = ISR_FATAL(EIO);
+		goto out;
+	}
+
+	if (i2c->mode == MODE_ATOMIC)
+		hret = img_i2c_atomic(i2c, int_status, line_status);
+	else if (i2c->mode == MODE_AUTOMATIC)
+		hret = img_i2c_auto(i2c, int_status, line_status);
+	else if (i2c->mode == MODE_SEQUENCE)
+		hret = img_i2c_sequence(i2c, int_status);
+	else if (i2c->mode == MODE_WAITSTOP && (int_status & INT_SLAVE_EVENT) &&
+			 (line_status & LINESTAT_STOP_BIT_DET))
+		hret = ISR_COMPLETE(0);
+	else if (i2c->mode == MODE_RAW)
+		hret = img_i2c_raw(i2c, int_status, line_status);
+	else
+		hret = 0;
+
+	/* Clear detected level interrupts. */
+	img_i2c_writel(i2c, SCB_INT_CLEAR_REG, int_status & INT_LEVEL);
+
+out:
+	if (hret & ISR_WAITSTOP) {
+		/*
+		 * Only wait for stop on last message.
+		 * Also we may already have detected the stop bit.
+		 */
+		if (!i2c->last_msg || i2c->line_status & LINESTAT_STOP_BIT_DET)
+			hret = ISR_COMPLETE(0);
+		else
+			img_i2c_switch_mode(i2c, MODE_WAITSTOP);
+	}
+
+	/* now we've finished using regs, handle transaction completion */
+	if (hret & ISR_COMPLETE_M) {
+		int status = -(hret & ISR_STATUS_M);
+
+		img_i2c_complete_transaction(i2c, status);
+		if (hret & ISR_FATAL_M)
+			img_i2c_switch_mode(i2c, MODE_FATAL);
+	}
+
+	/* Enable interrupts (int_enable may be altered by changing mode) */
+	img_i2c_writel(i2c, SCB_INT_MASK_REG, i2c->int_enable);
+
+	spin_unlock(&i2c->lock);
+
+	return IRQ_HANDLED;
+}
+
+/* Force a bus reset sequence and wait for it to complete */
+static int img_i2c_reset_bus(struct img_i2c *i2c)
+{
+	unsigned long flags;
+	int ret;
+
+	spin_lock_irqsave(&i2c->lock, flags);
+	reinit_completion(&i2c->msg_complete);
+	img_i2c_reset_start(i2c);
+	spin_unlock_irqrestore(&i2c->lock, flags);
+
+	ret = wait_for_completion_timeout(&i2c->msg_complete, IMG_I2C_TIMEOUT);
+	if (ret == 0)
+		return -ETIMEDOUT;
+	return 0;
+}
+
+static int img_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg *msgs,
+			int num)
+{
+	struct img_i2c *i2c = i2c_get_adapdata(adap);
+	bool atomic = false;
+	int i, ret;
+
+	if (i2c->mode == MODE_SUSPEND) {
+		WARN(1, "refusing to service transaction in suspended state\n");
+		return -EIO;
+	}
+
+	if (i2c->mode == MODE_FATAL)
+		return -EIO;
+
+	for (i = 0; i < num; i++) {
+		if (likely(msgs[i].len))
+			continue;
+		/*
+		 * 0 byte reads are not possible because the slave could try
+		 * and pull the data line low, preventing a stop bit.
+		 */
+		if (unlikely(msgs[i].flags & I2C_M_RD))
+			return -EIO;
+		/*
+		 * 0 byte writes are possible and used for probing, but we
+		 * cannot do them in automatic mode, so use atomic mode
+		 * instead.
+		 */
+		atomic = true;
+	}
+
+	ret = clk_prepare_enable(i2c->scb_clk);
+	if (ret)
+		return ret;
+
+	for (i = 0; i < num; i++) {
+		struct i2c_msg *msg = &msgs[i];
+		unsigned long flags;
+
+		spin_lock_irqsave(&i2c->lock, flags);
+
+		/*
+		 * Make a copy of the message struct. We mustn't modify the
+		 * original or we'll confuse drivers and i2c-dev.
+		 */
+		i2c->msg = *msg;
+		i2c->msg_status = 0;
+
+		/*
+		 * After the last message we must have waited for a stop bit.
+		 * Not waiting can cause problems when the clock is disabled
+		 * before the stop bit is sent, and the linux I2C interface
+		 * requires separate transfers not to joined with repeated
+		 * start.
+		 */
+		i2c->last_msg = (i == num - 1);
+		reinit_completion(&i2c->msg_complete);
+
+		if (atomic)
+			img_i2c_atomic_start(i2c);
+		else if (msg->flags & I2C_M_RD)
+			img_i2c_read(i2c);
+		else
+			img_i2c_write(i2c);
+		spin_unlock_irqrestore(&i2c->lock, flags);
+
+		ret = wait_for_completion_timeout(&i2c->msg_complete,
+						  IMG_I2C_TIMEOUT);
+		del_timer_sync(&i2c->check_timer);
+
+		if (ret == 0) {
+			dev_err(adap->dev.parent, "i2c transfer timed out\n");
+			i2c->msg_status = -ETIMEDOUT;
+			break;
+		}
+
+		if (i2c->msg_status)
+			break;
+	}
+
+	clk_disable_unprepare(i2c->scb_clk);
+
+	return i2c->msg_status ? i2c->msg_status : num;
+}
+
+static u32 img_i2c_func(struct i2c_adapter *adap)
+{
+	return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL;
+}
+
+static const struct i2c_algorithm img_i2c_algo = {
+	.master_xfer = img_i2c_xfer,
+	.functionality = img_i2c_func,
+};
+
+static int img_i2c_init(struct img_i2c *i2c)
+{
+	unsigned int clk_khz, bitrate_khz, clk_period, tckh, tckl, tsdh;
+	unsigned int i, ret, data, prescale, inc, int_bitrate, filt;
+	struct img_i2c_timings timing;
+	u32 rev;
+
+	ret = clk_prepare_enable(i2c->scb_clk);
+	if (ret)
+		return ret;
+
+	rev = img_i2c_readl(i2c, SCB_CORE_REV_REG);
+	if ((rev & 0x00ffffff) < 0x00020200) {
+		dev_info(i2c->adap.dev.parent,
+			 "Unknown hardware revision (%d.%d.%d.%d)\n",
+			 (rev >> 24) & 0xff, (rev >> 16) & 0xff,
+			 (rev >> 8) & 0xff, rev & 0xff);
+		clk_disable_unprepare(i2c->scb_clk);
+		return -EINVAL;
+	}
+
+	if (rev == REL_SOC_IP_SCB_2_2_1) {
+		i2c->need_wr_rd_fence = true;
+		dev_info(i2c->adap.dev.parent, "fence quirk enabled");
+	}
+
+	bitrate_khz = i2c->bitrate / 1000;
+	clk_khz = clk_get_rate(i2c->scb_clk) / 1000;
+
+	/* Determine what mode we're in from the bitrate */
+	timing = timings[0];
+	for (i = 0; i < ARRAY_SIZE(timings); i++) {
+		if (i2c->bitrate <= timings[i].max_bitrate) {
+			timing = timings[i];
+			break;
+		}
+	}
+
+	/* Find the prescale that would give us that inc (approx delay = 0) */
+	prescale = SCB_OPT_INC * clk_khz / (256 * 16 * bitrate_khz);
+	prescale = clamp_t(unsigned int, prescale, 1, 8);
+	clk_khz /= prescale;
+
+	/* Setup the clock increment value */
+	inc = (256 * 16 * bitrate_khz) / clk_khz;
+
+	/*
+	 * The clock generation logic allows to filter glitches on the bus.
+	 * This filter is able to remove bus glitches shorter than 50ns.
+	 * If the clock enable rate is greater than 20 MHz, no filtering
+	 * is required, so we need to disable it.
+	 * If it's between the 20-40 MHz range, there's no need to divide
+	 * the clock to get a filter.
+	 */
+	if (clk_khz < 20000) {
+		filt = SCB_FILT_DISABLE;
+	} else if (clk_khz < 40000) {
+		filt = SCB_FILT_BYPASS;
+	} else {
+		/* Calculate filter clock */
+		filt = (64000 / ((clk_khz / 1000) * SCB_FILT_GLITCH));
+
+		/* Scale up if needed */
+		if (64000 % ((clk_khz / 1000) * SCB_FILT_GLITCH))
+			inc++;
+
+		if (filt > SCB_FILT_INC_MASK)
+			filt = SCB_FILT_INC_MASK;
+
+		filt = (filt & SCB_FILT_INC_MASK) << SCB_FILT_INC_SHIFT;
+	}
+	data = filt | ((inc & SCB_INC_MASK) << SCB_INC_SHIFT) | (prescale - 1);
+	img_i2c_writel(i2c, SCB_CLK_SET_REG, data);
+
+	/* Obtain the clock period of the fx16 clock in ns */
+	clk_period = (256 * 1000000) / (clk_khz * inc);
+
+	/* Calculate the bitrate in terms of internal clock pulses */
+	int_bitrate = 1000000 / (bitrate_khz * clk_period);
+	if ((1000000 % (bitrate_khz * clk_period)) >=
+	    ((bitrate_khz * clk_period) / 2))
+		int_bitrate++;
+
+	/* Setup TCKH value */
+	tckh = timing.tckh / clk_period;
+	if (timing.tckh % clk_period)
+		tckh++;
+
+	if (tckh > 0)
+		data = tckh - 1;
+	else
+		data = 0;
+
+	img_i2c_writel(i2c, SCB_TIME_TCKH_REG, data);
+
+	/* Setup TCKL value */
+	tckl = int_bitrate - tckh;
+
+	if (tckl > 0)
+		data = tckl - 1;
+	else
+		data = 0;
+
+	img_i2c_writel(i2c, SCB_TIME_TCKL_REG, data);
+
+	/* Setup TSDH value */
+	tsdh = timing.tsdh / clk_period;
+	if (timing.tsdh % clk_period)
+		tsdh++;
+
+	if (tsdh > 1)
+		data = tsdh - 1;
+	else
+		data = 0x01;
+	img_i2c_writel(i2c, SCB_TIME_TSDH_REG, data);
+
+	/* This value is used later */
+	tsdh = data;
+
+	/* Setup TPL value */
+	data = timing.tpl / clk_period;
+	if (data > 0)
+		--data;
+	img_i2c_writel(i2c, SCB_TIME_TPL_REG, data);
+
+	/* Setup TPH value */
+	data = timing.tph / clk_period;
+	if (data > 0)
+		--data;
+	img_i2c_writel(i2c, SCB_TIME_TPH_REG, data);
+
+	/* Setup TSDL value to TPL + TSDH + 2 */
+	img_i2c_writel(i2c, SCB_TIME_TSDL_REG, data + tsdh + 2);
+
+	/* Setup TP2S value */
+	data = timing.tp2s / clk_period;
+	if (data > 0)
+		--data;
+	img_i2c_writel(i2c, SCB_TIME_TP2S_REG, data);
+
+	img_i2c_writel(i2c, SCB_TIME_TBI_REG, TIMEOUT_TBI);
+	img_i2c_writel(i2c, SCB_TIME_TSL_REG, TIMEOUT_TSL);
+	img_i2c_writel(i2c, SCB_TIME_TDL_REG, TIMEOUT_TDL);
+
+	/* Take module out of soft reset and enable clocks */
+	img_i2c_soft_reset(i2c);
+
+	/* Disable all interrupts */
+	img_i2c_writel(i2c, SCB_INT_MASK_REG, 0);
+
+	/* Clear all interrupts */
+	img_i2c_writel(i2c, SCB_INT_CLEAR_REG, ~0);
+
+	/* Clear the scb_line_status events */
+	img_i2c_writel(i2c, SCB_CLEAR_REG, ~0);
+
+	/* Enable interrupts */
+	img_i2c_writel(i2c, SCB_INT_MASK_REG, i2c->int_enable);
+
+	/* Perform a synchronous sequence to reset the bus */
+	ret = img_i2c_reset_bus(i2c);
+
+	clk_disable_unprepare(i2c->scb_clk);
+
+	return ret;
+}
+
+static int img_i2c_probe(struct platform_device *pdev)
+{
+	struct device_node *node = pdev->dev.of_node;
+	struct img_i2c *i2c;
+	struct resource *res;
+	int irq, ret;
+	u32 val;
+
+	i2c = devm_kzalloc(&pdev->dev, sizeof(struct img_i2c), GFP_KERNEL);
+	if (!i2c)
+		return -ENOMEM;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	i2c->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(i2c->base))
+		return PTR_ERR(i2c->base);
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(&pdev->dev, "can't get irq number\n");
+		return irq;
+	}
+
+	i2c->sys_clk = devm_clk_get(&pdev->dev, "sys");
+	if (IS_ERR(i2c->sys_clk)) {
+		dev_err(&pdev->dev, "can't get system clock\n");
+		return PTR_ERR(i2c->sys_clk);
+	}
+
+	i2c->scb_clk = devm_clk_get(&pdev->dev, "scb");
+	if (IS_ERR(i2c->scb_clk)) {
+		dev_err(&pdev->dev, "can't get core clock\n");
+		return PTR_ERR(i2c->scb_clk);
+	}
+
+	ret = devm_request_irq(&pdev->dev, irq, img_i2c_isr, 0,
+			       pdev->name, i2c);
+	if (ret) {
+		dev_err(&pdev->dev, "can't request irq %d\n", irq);
+		return ret;
+	}
+
+	/* Set up the exception check timer */
+	init_timer(&i2c->check_timer);
+	i2c->check_timer.function = img_i2c_check_timer;
+	i2c->check_timer.data = (unsigned long)i2c;
+
+	i2c->bitrate = timings[0].max_bitrate;
+	if (!of_property_read_u32(node, "clock-frequency", &val))
+		i2c->bitrate = val;
+
+	i2c_set_adapdata(&i2c->adap, i2c);
+	i2c->adap.dev.parent = &pdev->dev;
+	i2c->adap.dev.of_node = node;
+	i2c->adap.owner = THIS_MODULE;
+	i2c->adap.algo = &img_i2c_algo;
+	i2c->adap.retries = 5;
+	i2c->adap.nr = pdev->id;
+	snprintf(i2c->adap.name, sizeof(i2c->adap.name), "IMG SCB I2C");
+
+	img_i2c_switch_mode(i2c, MODE_INACTIVE);
+	spin_lock_init(&i2c->lock);
+	init_completion(&i2c->msg_complete);
+
+	platform_set_drvdata(pdev, i2c);
+
+	ret = clk_prepare_enable(i2c->sys_clk);
+	if (ret)
+		return ret;
+
+	ret = img_i2c_init(i2c);
+	if (ret)
+		goto disable_clk;
+
+	ret = i2c_add_numbered_adapter(&i2c->adap);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to add adapter\n");
+		goto disable_clk;
+	}
+
+	return 0;
+
+disable_clk:
+	clk_disable_unprepare(i2c->sys_clk);
+	return ret;
+}
+
+static int img_i2c_remove(struct platform_device *dev)
+{
+	struct img_i2c *i2c = platform_get_drvdata(dev);
+
+	i2c_del_adapter(&i2c->adap);
+	clk_disable_unprepare(i2c->sys_clk);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int img_i2c_suspend(struct device *dev)
+{
+	struct img_i2c *i2c = dev_get_drvdata(dev);
+
+	img_i2c_switch_mode(i2c, MODE_SUSPEND);
+
+	clk_disable_unprepare(i2c->sys_clk);
+
+	return 0;
+}
+
+static int img_i2c_resume(struct device *dev)
+{
+	struct img_i2c *i2c = dev_get_drvdata(dev);
+	int ret;
+
+	ret = clk_prepare_enable(i2c->sys_clk);
+	if (ret)
+		return ret;
+
+	img_i2c_init(i2c);
+
+	return 0;
+}
+#endif /* CONFIG_PM_SLEEP */
+
+static SIMPLE_DEV_PM_OPS(img_i2c_pm, img_i2c_suspend, img_i2c_resume);
+
+static const struct of_device_id img_scb_i2c_match[] = {
+	{ .compatible = "img,scb-i2c" },
+	{ }
+};
+MODULE_DEVICE_TABLE(of, img_scb_i2c_match);
+
+static struct platform_driver img_scb_i2c_driver = {
+	.driver = {
+		.name		= "img-i2c-scb",
+		.of_match_table	= img_scb_i2c_match,
+		.pm		= &img_i2c_pm,
+	},
+	.probe = img_i2c_probe,
+	.remove = img_i2c_remove,
+};
+module_platform_driver(img_scb_i2c_driver);
+
+MODULE_AUTHOR("James Hogan <james.hogan@imgtec.com>");
+MODULE_DESCRIPTION("IMG host I2C driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/i2c/busses/i2c-imx.c b/drivers/i2c/busses/i2c-imx.c
index 613069bc561a..7f3a9fe9bf4e 100644
--- a/drivers/i2c/busses/i2c-imx.c
+++ b/drivers/i2c/busses/i2c-imx.c
@@ -11,11 +11,6 @@
  *	MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  *	GNU General Public License for more details.
  *
- *	You should have received a copy of the GNU General Public License
- *	along with this program; if not, write to the Free Software
- *	Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307,
- *	USA.
- *
  * Author:
  *	Darius Augulis, Teltonika Inc.
  *
@@ -37,22 +32,27 @@
 /** Includes *******************************************************************
 *******************************************************************************/
 
-#include <linux/init.h>
-#include <linux/kernel.h>
-#include <linux/module.h>
-#include <linux/errno.h>
-#include <linux/err.h>
-#include <linux/interrupt.h>
+#include <linux/clk.h>
+#include <linux/completion.h>
 #include <linux/delay.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/dmapool.h>
+#include <linux/err.h>
+#include <linux/errno.h>
 #include <linux/i2c.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
 #include <linux/io.h>
-#include <linux/sched.h>
-#include <linux/platform_device.h>
-#include <linux/clk.h>
-#include <linux/slab.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
 #include <linux/of.h>
 #include <linux/of_device.h>
+#include <linux/of_dma.h>
 #include <linux/platform_data/i2c-imx.h>
+#include <linux/platform_device.h>
+#include <linux/sched.h>
+#include <linux/slab.h>
 
 /** Defines ********************************************************************
 *******************************************************************************/
@@ -63,6 +63,15 @@
 /* Default value */
 #define IMX_I2C_BIT_RATE	100000	/* 100kHz */
 
+/*
+ * Enable DMA if transfer byte size is bigger than this threshold.
+ * As the hardware request, it must bigger than 4 bytes.\
+ * I have set '16' here, maybe it's not the best but I think it's
+ * the appropriate.
+ */
+#define DMA_THRESHOLD	16
+#define DMA_TIMEOUT	1000
+
 /* IMX I2C registers:
  * the I2C register offset is different between SoCs,
  * to provid support for all these chips, split the
@@ -88,6 +97,7 @@
 #define I2SR_IBB	0x20
 #define I2SR_IAAS	0x40
 #define I2SR_ICF	0x80
+#define I2CR_DMAEN	0x02
 #define I2CR_RSTA	0x04
 #define I2CR_TXAK	0x08
 #define I2CR_MTX	0x10
@@ -174,6 +184,17 @@ struct imx_i2c_hwdata {
 	unsigned		i2cr_ien_opcode;
 };
 
+struct imx_i2c_dma {
+	struct dma_chan		*chan_tx;
+	struct dma_chan		*chan_rx;
+	struct dma_chan		*chan_using;
+	struct completion	cmd_complete;
+	dma_addr_t		dma_buf;
+	unsigned int		dma_len;
+	enum dma_transfer_direction dma_transfer_dir;
+	enum dma_data_direction dma_data_dir;
+};
+
 struct imx_i2c_struct {
 	struct i2c_adapter	adapter;
 	struct clk		*clk;
@@ -186,6 +207,8 @@ struct imx_i2c_struct {
 	unsigned int		cur_clk;
 	unsigned int		bitrate;
 	const struct imx_i2c_hwdata	*hwdata;
+
+	struct imx_i2c_dma	*dma;
 };
 
 static const struct imx_i2c_hwdata imx1_i2c_hwdata  = {
@@ -256,6 +279,138 @@ static inline unsigned char imx_i2c_read_reg(struct imx_i2c_struct *i2c_imx,
 	return readb(i2c_imx->base + (reg << i2c_imx->hwdata->regshift));
 }
 
+/* Functions for DMA support */
+static void i2c_imx_dma_request(struct imx_i2c_struct *i2c_imx,
+						dma_addr_t phy_addr)
+{
+	struct imx_i2c_dma *dma;
+	struct dma_slave_config dma_sconfig;
+	struct device *dev = &i2c_imx->adapter.dev;
+	int ret;
+
+	dma = devm_kzalloc(dev, sizeof(*dma), GFP_KERNEL);
+	if (!dma)
+		return;
+
+	dma->chan_tx = dma_request_slave_channel(dev, "tx");
+	if (!dma->chan_tx) {
+		dev_dbg(dev, "can't request DMA tx channel\n");
+		ret = -ENODEV;
+		goto fail_al;
+	}
+
+	dma_sconfig.dst_addr = phy_addr +
+				(IMX_I2C_I2DR << i2c_imx->hwdata->regshift);
+	dma_sconfig.dst_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+	dma_sconfig.dst_maxburst = 1;
+	dma_sconfig.direction = DMA_MEM_TO_DEV;
+	ret = dmaengine_slave_config(dma->chan_tx, &dma_sconfig);
+	if (ret < 0) {
+		dev_dbg(dev, "can't configure tx channel\n");
+		goto fail_tx;
+	}
+
+	dma->chan_rx = dma_request_slave_channel(dev, "rx");
+	if (!dma->chan_rx) {
+		dev_dbg(dev, "can't request DMA rx channel\n");
+		ret = -ENODEV;
+		goto fail_tx;
+	}
+
+	dma_sconfig.src_addr = phy_addr +
+				(IMX_I2C_I2DR << i2c_imx->hwdata->regshift);
+	dma_sconfig.src_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+	dma_sconfig.src_maxburst = 1;
+	dma_sconfig.direction = DMA_DEV_TO_MEM;
+	ret = dmaengine_slave_config(dma->chan_rx, &dma_sconfig);
+	if (ret < 0) {
+		dev_dbg(dev, "can't configure rx channel\n");
+		goto fail_rx;
+	}
+
+	i2c_imx->dma = dma;
+	init_completion(&dma->cmd_complete);
+	dev_info(dev, "using %s (tx) and %s (rx) for DMA transfers\n",
+		dma_chan_name(dma->chan_tx), dma_chan_name(dma->chan_rx));
+
+	return;
+
+fail_rx:
+	dma_release_channel(dma->chan_rx);
+fail_tx:
+	dma_release_channel(dma->chan_tx);
+fail_al:
+	devm_kfree(dev, dma);
+	dev_info(dev, "can't use DMA\n");
+}
+
+static void i2c_imx_dma_callback(void *arg)
+{
+	struct imx_i2c_struct *i2c_imx = (struct imx_i2c_struct *)arg;
+	struct imx_i2c_dma *dma = i2c_imx->dma;
+
+	dma_unmap_single(dma->chan_using->device->dev, dma->dma_buf,
+			dma->dma_len, dma->dma_data_dir);
+	complete(&dma->cmd_complete);
+}
+
+static int i2c_imx_dma_xfer(struct imx_i2c_struct *i2c_imx,
+					struct i2c_msg *msgs)
+{
+	struct imx_i2c_dma *dma = i2c_imx->dma;
+	struct dma_async_tx_descriptor *txdesc;
+	struct device *dev = &i2c_imx->adapter.dev;
+	struct device *chan_dev = dma->chan_using->device->dev;
+
+	dma->dma_buf = dma_map_single(chan_dev, msgs->buf,
+					dma->dma_len, dma->dma_data_dir);
+	if (dma_mapping_error(chan_dev, dma->dma_buf)) {
+		dev_err(dev, "DMA mapping failed\n");
+		goto err_map;
+	}
+
+	txdesc = dmaengine_prep_slave_single(dma->chan_using, dma->dma_buf,
+					dma->dma_len, dma->dma_transfer_dir,
+					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!txdesc) {
+		dev_err(dev, "Not able to get desc for DMA xfer\n");
+		goto err_desc;
+	}
+
+	txdesc->callback = i2c_imx_dma_callback;
+	txdesc->callback_param = i2c_imx;
+	if (dma_submit_error(dmaengine_submit(txdesc))) {
+		dev_err(dev, "DMA submit failed\n");
+		goto err_submit;
+	}
+
+	dma_async_issue_pending(dma->chan_using);
+	return 0;
+
+err_submit:
+err_desc:
+	dma_unmap_single(chan_dev, dma->dma_buf,
+			dma->dma_len, dma->dma_data_dir);
+err_map:
+	return -EINVAL;
+}
+
+static void i2c_imx_dma_free(struct imx_i2c_struct *i2c_imx)
+{
+	struct imx_i2c_dma *dma = i2c_imx->dma;
+
+	dma->dma_buf = 0;
+	dma->dma_len = 0;
+
+	dma_release_channel(dma->chan_tx);
+	dma->chan_tx = NULL;
+
+	dma_release_channel(dma->chan_rx);
+	dma->chan_rx = NULL;
+
+	dma->chan_using = NULL;
+}
+
 /** Functions for IMX I2C adapter driver ***************************************
 *******************************************************************************/
 
@@ -268,6 +423,14 @@ static int i2c_imx_bus_busy(struct imx_i2c_struct *i2c_imx, int for_busy)
 
 	while (1) {
 		temp = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2SR);
+
+		/* check for arbitration lost */
+		if (temp & I2SR_IAL) {
+			temp &= ~I2SR_IAL;
+			imx_i2c_write_reg(temp, i2c_imx, IMX_I2C_I2SR);
+			return -EAGAIN;
+		}
+
 		if (for_busy && (temp & I2SR_IBB))
 			break;
 		if (!for_busy && !(temp & I2SR_IBB))
@@ -379,6 +542,7 @@ static int i2c_imx_start(struct imx_i2c_struct *i2c_imx)
 	i2c_imx->stopped = 0;
 
 	temp |= I2CR_IIEN | I2CR_MTX | I2CR_TXAK;
+	temp &= ~I2CR_DMAEN;
 	imx_i2c_write_reg(temp, i2c_imx, IMX_I2C_I2CR);
 	return result;
 }
@@ -392,6 +556,8 @@ static void i2c_imx_stop(struct imx_i2c_struct *i2c_imx)
 		dev_dbg(&i2c_imx->adapter.dev, "<%s>\n", __func__);
 		temp = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2CR);
 		temp &= ~(I2CR_MSTA | I2CR_MTX);
+		if (i2c_imx->dma)
+			temp &= ~I2CR_DMAEN;
 		imx_i2c_write_reg(temp, i2c_imx, IMX_I2C_I2CR);
 	}
 	if (is_imx1_i2c(i2c_imx)) {
@@ -432,6 +598,155 @@ static irqreturn_t i2c_imx_isr(int irq, void *dev_id)
 	return IRQ_NONE;
 }
 
+static int i2c_imx_dma_write(struct imx_i2c_struct *i2c_imx,
+					struct i2c_msg *msgs)
+{
+	int result;
+	unsigned int temp = 0;
+	unsigned long orig_jiffies = jiffies;
+	struct imx_i2c_dma *dma = i2c_imx->dma;
+	struct device *dev = &i2c_imx->adapter.dev;
+
+	dma->chan_using = dma->chan_tx;
+	dma->dma_transfer_dir = DMA_MEM_TO_DEV;
+	dma->dma_data_dir = DMA_TO_DEVICE;
+	dma->dma_len = msgs->len - 1;
+	result = i2c_imx_dma_xfer(i2c_imx, msgs);
+	if (result)
+		return result;
+
+	temp = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2CR);
+	temp |= I2CR_DMAEN;
+	imx_i2c_write_reg(temp, i2c_imx, IMX_I2C_I2CR);
+
+	/*
+	 * Write slave address.
+	 * The first byte must be transmitted by the CPU.
+	 */
+	imx_i2c_write_reg(msgs->addr << 1, i2c_imx, IMX_I2C_I2DR);
+	reinit_completion(&i2c_imx->dma->cmd_complete);
+	result = wait_for_completion_timeout(
+				&i2c_imx->dma->cmd_complete,
+				msecs_to_jiffies(DMA_TIMEOUT));
+	if (result <= 0) {
+		dmaengine_terminate_all(dma->chan_using);
+		return result ?: -ETIMEDOUT;
+	}
+
+	/* Waiting for transfer complete. */
+	while (1) {
+		temp = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2SR);
+		if (temp & I2SR_ICF)
+			break;
+		if (time_after(jiffies, orig_jiffies +
+				msecs_to_jiffies(DMA_TIMEOUT))) {
+			dev_dbg(dev, "<%s> Timeout\n", __func__);
+			return -ETIMEDOUT;
+		}
+		schedule();
+	}
+
+	temp = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2CR);
+	temp &= ~I2CR_DMAEN;
+	imx_i2c_write_reg(temp, i2c_imx, IMX_I2C_I2CR);
+
+	/* The last data byte must be transferred by the CPU. */
+	imx_i2c_write_reg(msgs->buf[msgs->len-1],
+				i2c_imx, IMX_I2C_I2DR);
+	result = i2c_imx_trx_complete(i2c_imx);
+	if (result)
+		return result;
+
+	return i2c_imx_acked(i2c_imx);
+}
+
+static int i2c_imx_dma_read(struct imx_i2c_struct *i2c_imx,
+			struct i2c_msg *msgs, bool is_lastmsg)
+{
+	int result;
+	unsigned int temp;
+	unsigned long orig_jiffies = jiffies;
+	struct imx_i2c_dma *dma = i2c_imx->dma;
+	struct device *dev = &i2c_imx->adapter.dev;
+
+	temp = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2CR);
+	temp |= I2CR_DMAEN;
+	imx_i2c_write_reg(temp, i2c_imx, IMX_I2C_I2CR);
+
+	dma->chan_using = dma->chan_rx;
+	dma->dma_transfer_dir = DMA_DEV_TO_MEM;
+	dma->dma_data_dir = DMA_FROM_DEVICE;
+	/* The last two data bytes must be transferred by the CPU. */
+	dma->dma_len = msgs->len - 2;
+	result = i2c_imx_dma_xfer(i2c_imx, msgs);
+	if (result)
+		return result;
+
+	reinit_completion(&i2c_imx->dma->cmd_complete);
+	result = wait_for_completion_timeout(
+				&i2c_imx->dma->cmd_complete,
+				msecs_to_jiffies(DMA_TIMEOUT));
+	if (result <= 0) {
+		dmaengine_terminate_all(dma->chan_using);
+		return result ?: -ETIMEDOUT;
+	}
+
+	/* waiting for transfer complete. */
+	while (1) {
+		temp = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2SR);
+		if (temp & I2SR_ICF)
+			break;
+		if (time_after(jiffies, orig_jiffies +
+				msecs_to_jiffies(DMA_TIMEOUT))) {
+			dev_dbg(dev, "<%s> Timeout\n", __func__);
+			return -ETIMEDOUT;
+		}
+		schedule();
+	}
+
+	temp = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2CR);
+	temp &= ~I2CR_DMAEN;
+	imx_i2c_write_reg(temp, i2c_imx, IMX_I2C_I2CR);
+
+	/* read n-1 byte data */
+	temp = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2CR);
+	temp |= I2CR_TXAK;
+	imx_i2c_write_reg(temp, i2c_imx, IMX_I2C_I2CR);
+
+	msgs->buf[msgs->len-2] = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2DR);
+	/* read n byte data */
+	result = i2c_imx_trx_complete(i2c_imx);
+	if (result)
+		return result;
+
+	if (is_lastmsg) {
+		/*
+		 * It must generate STOP before read I2DR to prevent
+		 * controller from generating another clock cycle
+		 */
+		dev_dbg(dev, "<%s> clear MSTA\n", __func__);
+		temp = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2CR);
+		temp &= ~(I2CR_MSTA | I2CR_MTX);
+		imx_i2c_write_reg(temp, i2c_imx, IMX_I2C_I2CR);
+		i2c_imx_bus_busy(i2c_imx, 0);
+		i2c_imx->stopped = 1;
+	} else {
+		/*
+		 * For i2c master receiver repeat restart operation like:
+		 * read -> repeat MSTA -> read/write
+		 * The controller must set MTX before read the last byte in
+		 * the first read operation, otherwise the first read cost
+		 * one extra clock cycle.
+		 */
+		temp = readb(i2c_imx->base + IMX_I2C_I2CR);
+		temp |= I2CR_MTX;
+		writeb(temp, i2c_imx->base + IMX_I2C_I2CR);
+	}
+	msgs->buf[msgs->len-1] = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2DR);
+
+	return 0;
+}
+
 static int i2c_imx_write(struct imx_i2c_struct *i2c_imx, struct i2c_msg *msgs)
 {
 	int i, result;
@@ -501,6 +816,9 @@ static int i2c_imx_read(struct imx_i2c_struct *i2c_imx, struct i2c_msg *msgs, bo
 
 	dev_dbg(&i2c_imx->adapter.dev, "<%s> read data\n", __func__);
 
+	if (i2c_imx->dma && msgs->len >= DMA_THRESHOLD && !block_data)
+		return i2c_imx_dma_read(i2c_imx, msgs, is_lastmsg);
+
 	/* read data */
 	for (i = 0; i < msgs->len; i++) {
 		u8 len = 0;
@@ -615,8 +933,12 @@ static int i2c_imx_xfer(struct i2c_adapter *adapter,
 #endif
 		if (msgs[i].flags & I2C_M_RD)
 			result = i2c_imx_read(i2c_imx, &msgs[i], is_lastmsg);
-		else
-			result = i2c_imx_write(i2c_imx, &msgs[i]);
+		else {
+			if (i2c_imx->dma && msgs[i].len >= DMA_THRESHOLD)
+				result = i2c_imx_dma_write(i2c_imx, &msgs[i]);
+			else
+				result = i2c_imx_write(i2c_imx, &msgs[i]);
+		}
 		if (result)
 			goto fail0;
 	}
@@ -651,6 +973,7 @@ static int i2c_imx_probe(struct platform_device *pdev)
 	struct imxi2c_platform_data *pdata = dev_get_platdata(&pdev->dev);
 	void __iomem *base;
 	int irq, ret;
+	dma_addr_t phy_addr;
 
 	dev_dbg(&pdev->dev, "<%s>\n", __func__);
 
@@ -665,8 +988,8 @@ static int i2c_imx_probe(struct platform_device *pdev)
 	if (IS_ERR(base))
 		return PTR_ERR(base);
 
-	i2c_imx = devm_kzalloc(&pdev->dev, sizeof(struct imx_i2c_struct),
-				GFP_KERNEL);
+	phy_addr = (dma_addr_t)res->start;
+	i2c_imx = devm_kzalloc(&pdev->dev, sizeof(*i2c_imx), GFP_KERNEL);
 	if (!i2c_imx)
 		return -ENOMEM;
 
@@ -702,7 +1025,7 @@ static int i2c_imx_probe(struct platform_device *pdev)
 				pdev->name, i2c_imx);
 	if (ret) {
 		dev_err(&pdev->dev, "can't claim irq %d\n", irq);
-		return ret;
+		goto clk_disable;
 	}
 
 	/* Init queue */
@@ -727,7 +1050,7 @@ static int i2c_imx_probe(struct platform_device *pdev)
 	ret = i2c_add_numbered_adapter(&i2c_imx->adapter);
 	if (ret < 0) {
 		dev_err(&pdev->dev, "registration failed\n");
-		return ret;
+		goto clk_disable;
 	}
 
 	/* Set up platform driver data */
@@ -740,7 +1063,14 @@ static int i2c_imx_probe(struct platform_device *pdev)
 		i2c_imx->adapter.name);
 	dev_info(&i2c_imx->adapter.dev, "IMX I2C adapter registered\n");
 
+	/* Init DMA config if support*/
+	i2c_imx_dma_request(i2c_imx, phy_addr);
+
 	return 0;   /* Return OK */
+
+clk_disable:
+	clk_disable_unprepare(i2c_imx->clk);
+	return ret;
 }
 
 static int i2c_imx_remove(struct platform_device *pdev)
@@ -751,6 +1081,9 @@ static int i2c_imx_remove(struct platform_device *pdev)
 	dev_dbg(&i2c_imx->adapter.dev, "adapter removed\n");
 	i2c_del_adapter(&i2c_imx->adapter);
 
+	if (i2c_imx->dma)
+		i2c_imx_dma_free(i2c_imx);
+
 	/* setup chip registers to defaults */
 	imx_i2c_write_reg(0, i2c_imx, IMX_I2C_IADR);
 	imx_i2c_write_reg(0, i2c_imx, IMX_I2C_IFDR);
@@ -765,7 +1098,6 @@ static struct platform_driver i2c_imx_driver = {
 	.remove = i2c_imx_remove,
 	.driver	= {
 		.name	= DRIVER_NAME,
-		.owner	= THIS_MODULE,
 		.of_match_table = i2c_imx_dt_ids,
 	},
 	.id_table	= imx_i2c_devtype,
diff --git a/drivers/i2c/busses/i2c-iop3xx.c b/drivers/i2c/busses/i2c-iop3xx.c
index 3d16c2f60a5e..72d6161cf77c 100644
--- a/drivers/i2c/busses/i2c-iop3xx.c
+++ b/drivers/i2c/busses/i2c-iop3xx.c
@@ -516,7 +516,6 @@ static struct platform_driver iop3xx_i2c_driver = {
 	.probe		= iop3xx_i2c_probe,
 	.remove		= iop3xx_i2c_remove,
 	.driver		= {
-		.owner	= THIS_MODULE,
 		.name	= "IOP3xx-I2C",
 	},
 };
diff --git a/drivers/i2c/busses/i2c-iop3xx.h b/drivers/i2c/busses/i2c-iop3xx.h
index 097e270955d0..2d6929c2bd92 100644
--- a/drivers/i2c/busses/i2c-iop3xx.h
+++ b/drivers/i2c/busses/i2c-iop3xx.h
@@ -11,11 +11,7 @@
     This program is distributed in the hope that it will be useful,
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-    GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.                */
+    GNU General Public License for more details.			     */
 /* ------------------------------------------------------------------------- */
 
 
diff --git a/drivers/i2c/busses/i2c-isch.c b/drivers/i2c/busses/i2c-isch.c
index cf99dbf21fd1..c2f25f19d76f 100644
--- a/drivers/i2c/busses/i2c-isch.c
+++ b/drivers/i2c/busses/i2c-isch.c
@@ -14,10 +14,6 @@
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
     GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 /*
@@ -313,7 +309,6 @@ static int smbus_sch_remove(struct platform_device *pdev)
 static struct platform_driver smbus_sch_driver = {
 	.driver = {
 		.name = "isch_smbus",
-		.owner = THIS_MODULE,
 	},
 	.probe		= smbus_sch_probe,
 	.remove		= smbus_sch_remove,
diff --git a/drivers/i2c/busses/i2c-ismt.c b/drivers/i2c/busses/i2c-ismt.c
index d9ee43c80cde..f2b0ff011631 100644
--- a/drivers/i2c/busses/i2c-ismt.c
+++ b/drivers/i2c/busses/i2c-ismt.c
@@ -14,10 +14,6 @@
  * WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  * General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
  * The full GNU General Public License is included in this distribution
  * in the file called LICENSE.GPL.
  *
@@ -81,7 +77,7 @@
 #define PCI_DEVICE_ID_INTEL_S1200_SMT1	0x0c5a
 #define PCI_DEVICE_ID_INTEL_AVOTON_SMT	0x1f15
 
-#define ISMT_DESC_ENTRIES	32	/* number of descriptor entries */
+#define ISMT_DESC_ENTRIES	2	/* number of descriptor entries */
 #define ISMT_MAX_RETRIES	3	/* number of SMBus retries to attempt */
 
 /* Hardware Descriptor Constants - Control Field */
diff --git a/drivers/i2c/busses/i2c-kempld.c b/drivers/i2c/busses/i2c-kempld.c
index af8f65fb1c05..25993d2e64bf 100644
--- a/drivers/i2c/busses/i2c-kempld.c
+++ b/drivers/i2c/busses/i2c-kempld.c
@@ -394,7 +394,6 @@ static int kempld_i2c_resume(struct platform_device *pdev)
 static struct platform_driver kempld_i2c_driver = {
 	.driver = {
 		.name = "kempld-i2c",
-		.owner = THIS_MODULE,
 	},
 	.probe		= kempld_i2c_probe,
 	.remove		= kempld_i2c_remove,
diff --git a/drivers/i2c/busses/i2c-meson.c b/drivers/i2c/busses/i2c-meson.c
new file mode 100644
index 000000000000..5e176adca8e8
--- /dev/null
+++ b/drivers/i2c/busses/i2c-meson.c
@@ -0,0 +1,492 @@
+/*
+ * I2C bus driver for Amlogic Meson SoCs
+ *
+ * Copyright (C) 2014 Beniamino Galvani <b.galvani@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/clk.h>
+#include <linux/completion.h>
+#include <linux/i2c.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/types.h>
+
+/* Meson I2C register map */
+#define REG_CTRL		0x00
+#define REG_SLAVE_ADDR		0x04
+#define REG_TOK_LIST0		0x08
+#define REG_TOK_LIST1		0x0c
+#define REG_TOK_WDATA0		0x10
+#define REG_TOK_WDATA1		0x14
+#define REG_TOK_RDATA0		0x18
+#define REG_TOK_RDATA1		0x1c
+
+/* Control register fields */
+#define REG_CTRL_START		BIT(0)
+#define REG_CTRL_ACK_IGNORE	BIT(1)
+#define REG_CTRL_STATUS		BIT(2)
+#define REG_CTRL_ERROR		BIT(3)
+#define REG_CTRL_CLKDIV_SHIFT	12
+#define REG_CTRL_CLKDIV_MASK	((BIT(10) - 1) << REG_CTRL_CLKDIV_SHIFT)
+
+#define I2C_TIMEOUT_MS		500
+#define DEFAULT_FREQ		100000
+
+enum {
+	TOKEN_END = 0,
+	TOKEN_START,
+	TOKEN_SLAVE_ADDR_WRITE,
+	TOKEN_SLAVE_ADDR_READ,
+	TOKEN_DATA,
+	TOKEN_DATA_LAST,
+	TOKEN_STOP,
+};
+
+enum {
+	STATE_IDLE,
+	STATE_READ,
+	STATE_WRITE,
+	STATE_STOP,
+};
+
+/**
+ * struct meson_i2c - Meson I2C device private data
+ *
+ * @adap:	I2C adapter instance
+ * @dev:	Pointer to device structure
+ * @regs:	Base address of the device memory mapped registers
+ * @clk:	Pointer to clock structure
+ * @irq:	IRQ number
+ * @msg:	Pointer to the current I2C message
+ * @state:	Current state in the driver state machine
+ * @last:	Flag set for the last message in the transfer
+ * @count:	Number of bytes to be sent/received in current transfer
+ * @pos:	Current position in the send/receive buffer
+ * @error:	Flag set when an error is received
+ * @lock:	To avoid race conditions between irq handler and xfer code
+ * @done:	Completion used to wait for transfer termination
+ * @frequency:	Operating frequency of I2C bus clock
+ * @tokens:	Sequence of tokens to be written to the device
+ * @num_tokens:	Number of tokens
+ */
+struct meson_i2c {
+	struct i2c_adapter	adap;
+	struct device		*dev;
+	void __iomem		*regs;
+	struct clk		*clk;
+	int			irq;
+
+	struct i2c_msg		*msg;
+	int			state;
+	bool			last;
+	int			count;
+	int			pos;
+	int			error;
+
+	spinlock_t		lock;
+	struct completion	done;
+	unsigned int		frequency;
+	u32			tokens[2];
+	int			num_tokens;
+};
+
+static void meson_i2c_set_mask(struct meson_i2c *i2c, int reg, u32 mask,
+			       u32 val)
+{
+	u32 data;
+
+	data = readl(i2c->regs + reg);
+	data &= ~mask;
+	data |= val & mask;
+	writel(data, i2c->regs + reg);
+}
+
+static void meson_i2c_reset_tokens(struct meson_i2c *i2c)
+{
+	i2c->tokens[0] = 0;
+	i2c->tokens[1] = 0;
+	i2c->num_tokens = 0;
+}
+
+static void meson_i2c_add_token(struct meson_i2c *i2c, int token)
+{
+	if (i2c->num_tokens < 8)
+		i2c->tokens[0] |= (token & 0xf) << (i2c->num_tokens * 4);
+	else
+		i2c->tokens[1] |= (token & 0xf) << ((i2c->num_tokens % 8) * 4);
+
+	i2c->num_tokens++;
+}
+
+static void meson_i2c_write_tokens(struct meson_i2c *i2c)
+{
+	writel(i2c->tokens[0], i2c->regs + REG_TOK_LIST0);
+	writel(i2c->tokens[1], i2c->regs + REG_TOK_LIST1);
+}
+
+static void meson_i2c_set_clk_div(struct meson_i2c *i2c)
+{
+	unsigned long clk_rate = clk_get_rate(i2c->clk);
+	unsigned int div;
+
+	div = DIV_ROUND_UP(clk_rate, i2c->frequency * 4);
+	meson_i2c_set_mask(i2c, REG_CTRL, REG_CTRL_CLKDIV_MASK,
+			   div << REG_CTRL_CLKDIV_SHIFT);
+
+	dev_dbg(i2c->dev, "%s: clk %lu, freq %u, div %u\n", __func__,
+		clk_rate, i2c->frequency, div);
+}
+
+static void meson_i2c_get_data(struct meson_i2c *i2c, char *buf, int len)
+{
+	u32 rdata0, rdata1;
+	int i;
+
+	rdata0 = readl(i2c->regs + REG_TOK_RDATA0);
+	rdata1 = readl(i2c->regs + REG_TOK_RDATA1);
+
+	dev_dbg(i2c->dev, "%s: data %08x %08x len %d\n", __func__,
+		rdata0, rdata1, len);
+
+	for (i = 0; i < min_t(int, 4, len); i++)
+		*buf++ = (rdata0 >> i * 8) & 0xff;
+
+	for (i = 4; i < min_t(int, 8, len); i++)
+		*buf++ = (rdata1 >> (i - 4) * 8) & 0xff;
+}
+
+static void meson_i2c_put_data(struct meson_i2c *i2c, char *buf, int len)
+{
+	u32 wdata0 = 0, wdata1 = 0;
+	int i;
+
+	for (i = 0; i < min_t(int, 4, len); i++)
+		wdata0 |= *buf++ << (i * 8);
+
+	for (i = 4; i < min_t(int, 8, len); i++)
+		wdata1 |= *buf++ << ((i - 4) * 8);
+
+	writel(wdata0, i2c->regs + REG_TOK_WDATA0);
+	writel(wdata0, i2c->regs + REG_TOK_WDATA1);
+
+	dev_dbg(i2c->dev, "%s: data %08x %08x len %d\n", __func__,
+		wdata0, wdata1, len);
+}
+
+static void meson_i2c_prepare_xfer(struct meson_i2c *i2c)
+{
+	bool write = !(i2c->msg->flags & I2C_M_RD);
+	int i;
+
+	i2c->count = min_t(int, i2c->msg->len - i2c->pos, 8);
+
+	for (i = 0; i < i2c->count - 1; i++)
+		meson_i2c_add_token(i2c, TOKEN_DATA);
+
+	if (i2c->count) {
+		if (write || i2c->pos + i2c->count < i2c->msg->len)
+			meson_i2c_add_token(i2c, TOKEN_DATA);
+		else
+			meson_i2c_add_token(i2c, TOKEN_DATA_LAST);
+	}
+
+	if (write)
+		meson_i2c_put_data(i2c, i2c->msg->buf + i2c->pos, i2c->count);
+}
+
+static void meson_i2c_stop(struct meson_i2c *i2c)
+{
+	dev_dbg(i2c->dev, "%s: last %d\n", __func__, i2c->last);
+
+	if (i2c->last) {
+		i2c->state = STATE_STOP;
+		meson_i2c_add_token(i2c, TOKEN_STOP);
+	} else {
+		i2c->state = STATE_IDLE;
+		complete_all(&i2c->done);
+	}
+}
+
+static irqreturn_t meson_i2c_irq(int irqno, void *dev_id)
+{
+	struct meson_i2c *i2c = dev_id;
+	unsigned int ctrl;
+
+	spin_lock(&i2c->lock);
+
+	meson_i2c_reset_tokens(i2c);
+	ctrl = readl(i2c->regs + REG_CTRL);
+
+	dev_dbg(i2c->dev, "irq: state %d, pos %d, count %d, ctrl %08x\n",
+		i2c->state, i2c->pos, i2c->count, ctrl);
+
+	if (ctrl & REG_CTRL_ERROR && i2c->state != STATE_IDLE) {
+		/*
+		 * The bit is set when the IGNORE_NAK bit is cleared
+		 * and the device didn't respond. In this case, the
+		 * I2C controller automatically generates a STOP
+		 * condition.
+		 */
+		dev_dbg(i2c->dev, "error bit set\n");
+		i2c->error = -ENXIO;
+		i2c->state = STATE_IDLE;
+		complete_all(&i2c->done);
+		goto out;
+	}
+
+	switch (i2c->state) {
+	case STATE_READ:
+		if (i2c->count > 0) {
+			meson_i2c_get_data(i2c, i2c->msg->buf + i2c->pos,
+					   i2c->count);
+			i2c->pos += i2c->count;
+		}
+
+		if (i2c->pos >= i2c->msg->len) {
+			meson_i2c_stop(i2c);
+			break;
+		}
+
+		meson_i2c_prepare_xfer(i2c);
+		break;
+	case STATE_WRITE:
+		i2c->pos += i2c->count;
+
+		if (i2c->pos >= i2c->msg->len) {
+			meson_i2c_stop(i2c);
+			break;
+		}
+
+		meson_i2c_prepare_xfer(i2c);
+		break;
+	case STATE_STOP:
+		i2c->state = STATE_IDLE;
+		complete_all(&i2c->done);
+		break;
+	case STATE_IDLE:
+		break;
+	}
+
+out:
+	if (i2c->state != STATE_IDLE) {
+		/* Restart the processing */
+		meson_i2c_write_tokens(i2c);
+		meson_i2c_set_mask(i2c, REG_CTRL, REG_CTRL_START, 0);
+		meson_i2c_set_mask(i2c, REG_CTRL, REG_CTRL_START,
+				   REG_CTRL_START);
+	}
+
+	spin_unlock(&i2c->lock);
+
+	return IRQ_HANDLED;
+}
+
+static void meson_i2c_do_start(struct meson_i2c *i2c, struct i2c_msg *msg)
+{
+	int token;
+
+	token = (msg->flags & I2C_M_RD) ? TOKEN_SLAVE_ADDR_READ :
+		TOKEN_SLAVE_ADDR_WRITE;
+
+	writel(msg->addr << 1, i2c->regs + REG_SLAVE_ADDR);
+	meson_i2c_add_token(i2c, TOKEN_START);
+	meson_i2c_add_token(i2c, token);
+}
+
+static int meson_i2c_xfer_msg(struct meson_i2c *i2c, struct i2c_msg *msg,
+			      int last)
+{
+	unsigned long time_left, flags;
+	int ret = 0;
+
+	i2c->msg = msg;
+	i2c->last = last;
+	i2c->pos = 0;
+	i2c->count = 0;
+	i2c->error = 0;
+
+	meson_i2c_reset_tokens(i2c);
+
+	flags = (msg->flags & I2C_M_IGNORE_NAK) ? REG_CTRL_ACK_IGNORE : 0;
+	meson_i2c_set_mask(i2c, REG_CTRL, REG_CTRL_ACK_IGNORE, flags);
+
+	if (!(msg->flags & I2C_M_NOSTART))
+		meson_i2c_do_start(i2c, msg);
+
+	i2c->state = (msg->flags & I2C_M_RD) ? STATE_READ : STATE_WRITE;
+	meson_i2c_prepare_xfer(i2c);
+	meson_i2c_write_tokens(i2c);
+	reinit_completion(&i2c->done);
+
+	/* Start the transfer */
+	meson_i2c_set_mask(i2c, REG_CTRL, REG_CTRL_START, REG_CTRL_START);
+
+	time_left = msecs_to_jiffies(I2C_TIMEOUT_MS);
+	time_left = wait_for_completion_timeout(&i2c->done, time_left);
+
+	/*
+	 * Protect access to i2c struct and registers from interrupt
+	 * handlers triggered by a transfer terminated after the
+	 * timeout period
+	 */
+	spin_lock_irqsave(&i2c->lock, flags);
+
+	/* Abort any active operation */
+	meson_i2c_set_mask(i2c, REG_CTRL, REG_CTRL_START, 0);
+
+	if (!time_left) {
+		i2c->state = STATE_IDLE;
+		ret = -ETIMEDOUT;
+	}
+
+	if (i2c->error)
+		ret = i2c->error;
+
+	spin_unlock_irqrestore(&i2c->lock, flags);
+
+	return ret;
+}
+
+static int meson_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg *msgs,
+			  int num)
+{
+	struct meson_i2c *i2c = adap->algo_data;
+	int i, ret = 0, count = 0;
+
+	clk_enable(i2c->clk);
+	meson_i2c_set_clk_div(i2c);
+
+	for (i = 0; i < num; i++) {
+		ret = meson_i2c_xfer_msg(i2c, msgs + i, i == num - 1);
+		if (ret)
+			break;
+		count++;
+	}
+
+	clk_disable(i2c->clk);
+
+	return ret ? ret : count;
+}
+
+static u32 meson_i2c_func(struct i2c_adapter *adap)
+{
+	return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL;
+}
+
+static const struct i2c_algorithm meson_i2c_algorithm = {
+	.master_xfer	= meson_i2c_xfer,
+	.functionality	= meson_i2c_func,
+};
+
+static int meson_i2c_probe(struct platform_device *pdev)
+{
+	struct device_node *np = pdev->dev.of_node;
+	struct meson_i2c *i2c;
+	struct resource *mem;
+	int ret = 0;
+
+	i2c = devm_kzalloc(&pdev->dev, sizeof(struct meson_i2c), GFP_KERNEL);
+	if (!i2c)
+		return -ENOMEM;
+
+	if (of_property_read_u32(pdev->dev.of_node, "clock-frequency",
+				 &i2c->frequency))
+		i2c->frequency = DEFAULT_FREQ;
+
+	i2c->dev = &pdev->dev;
+	platform_set_drvdata(pdev, i2c);
+
+	spin_lock_init(&i2c->lock);
+	init_completion(&i2c->done);
+
+	i2c->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(i2c->clk)) {
+		dev_err(&pdev->dev, "can't get device clock\n");
+		return PTR_ERR(i2c->clk);
+	}
+
+	mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	i2c->regs = devm_ioremap_resource(&pdev->dev, mem);
+	if (IS_ERR(i2c->regs))
+		return PTR_ERR(i2c->regs);
+
+	i2c->irq = platform_get_irq(pdev, 0);
+	if (i2c->irq < 0) {
+		dev_err(&pdev->dev, "can't find IRQ\n");
+		return i2c->irq;
+	}
+
+	ret = devm_request_irq(&pdev->dev, i2c->irq, meson_i2c_irq,
+			       0, dev_name(&pdev->dev), i2c);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "can't request IRQ\n");
+		return ret;
+	}
+
+	ret = clk_prepare(i2c->clk);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "can't prepare clock\n");
+		return ret;
+	}
+
+	strlcpy(i2c->adap.name, "Meson I2C adapter",
+		sizeof(i2c->adap.name));
+	i2c->adap.owner = THIS_MODULE;
+	i2c->adap.algo = &meson_i2c_algorithm;
+	i2c->adap.dev.parent = &pdev->dev;
+	i2c->adap.dev.of_node = np;
+	i2c->adap.algo_data = i2c;
+
+	/*
+	 * A transfer is triggered when START bit changes from 0 to 1.
+	 * Ensure that the bit is set to 0 after probe
+	 */
+	meson_i2c_set_mask(i2c, REG_CTRL, REG_CTRL_START, 0);
+
+	ret = i2c_add_adapter(&i2c->adap);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "can't register adapter\n");
+		clk_unprepare(i2c->clk);
+		return ret;
+	}
+
+	return 0;
+}
+
+static int meson_i2c_remove(struct platform_device *pdev)
+{
+	struct meson_i2c *i2c = platform_get_drvdata(pdev);
+
+	i2c_del_adapter(&i2c->adap);
+	clk_unprepare(i2c->clk);
+
+	return 0;
+}
+
+static const struct of_device_id meson_i2c_match[] = {
+	{ .compatible = "amlogic,meson6-i2c" },
+	{ },
+};
+
+static struct platform_driver meson_i2c_driver = {
+	.probe   = meson_i2c_probe,
+	.remove  = meson_i2c_remove,
+	.driver  = {
+		.name  = "meson-i2c",
+		.of_match_table = meson_i2c_match,
+	},
+};
+
+module_platform_driver(meson_i2c_driver);
+
+MODULE_DESCRIPTION("Amlogic Meson I2C Bus driver");
+MODULE_AUTHOR("Beniamino Galvani <b.galvani@gmail.com>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/i2c/busses/i2c-mpc.c b/drivers/i2c/busses/i2c-mpc.c
index 0edf630b099a..c74cc2be613b 100644
--- a/drivers/i2c/busses/i2c-mpc.c
+++ b/drivers/i2c/busses/i2c-mpc.c
@@ -124,7 +124,7 @@ static void mpc_i2c_fixup(struct mpc_i2c *i2c)
 static int i2c_wait(struct mpc_i2c *i2c, unsigned timeout, int writing)
 {
 	unsigned long orig_jiffies = jiffies;
-	u32 x;
+	u32 cmd_err;
 	int result = 0;
 
 	if (!i2c->irq) {
@@ -133,11 +133,11 @@ static int i2c_wait(struct mpc_i2c *i2c, unsigned timeout, int writing)
 			if (time_after(jiffies, orig_jiffies + timeout)) {
 				dev_dbg(i2c->dev, "timeout\n");
 				writeccr(i2c, 0);
-				result = -EIO;
+				result = -ETIMEDOUT;
 				break;
 			}
 		}
-		x = readb(i2c->base + MPC_I2C_SR);
+		cmd_err = readb(i2c->base + MPC_I2C_SR);
 		writeb(0, i2c->base + MPC_I2C_SR);
 	} else {
 		/* Interrupt mode */
@@ -150,28 +150,28 @@ static int i2c_wait(struct mpc_i2c *i2c, unsigned timeout, int writing)
 			result = -ETIMEDOUT;
 		}
 
-		x = i2c->interrupt;
+		cmd_err = i2c->interrupt;
 		i2c->interrupt = 0;
 	}
 
 	if (result < 0)
 		return result;
 
-	if (!(x & CSR_MCF)) {
+	if (!(cmd_err & CSR_MCF)) {
 		dev_dbg(i2c->dev, "unfinished\n");
 		return -EIO;
 	}
 
-	if (x & CSR_MAL) {
+	if (cmd_err & CSR_MAL) {
 		dev_dbg(i2c->dev, "MAL\n");
-		return -EIO;
+		return -EAGAIN;
 	}
 
-	if (writing && (x & CSR_RXAK)) {
+	if (writing && (cmd_err & CSR_RXAK)) {
 		dev_dbg(i2c->dev, "No RXAK\n");
 		/* generate stop */
 		writeccr(i2c, CCR_MEN);
-		return -EIO;
+		return -ENXIO;
 	}
 	return 0;
 }
@@ -813,7 +813,6 @@ static struct platform_driver mpc_i2c_driver = {
 	.probe		= fsl_i2c_probe,
 	.remove		= fsl_i2c_remove,
 	.driver = {
-		.owner = THIS_MODULE,
 		.name = DRV_NAME,
 		.of_match_table = mpc_i2c_of_match,
 		.pm = MPC_I2C_PM_OPS,
diff --git a/drivers/i2c/busses/i2c-mv64xxx.c b/drivers/i2c/busses/i2c-mv64xxx.c
index 2f64273d3f2b..30059c1df2a3 100644
--- a/drivers/i2c/busses/i2c-mv64xxx.c
+++ b/drivers/i2c/busses/i2c-mv64xxx.c
@@ -30,12 +30,12 @@
 #define MV64XXX_I2C_BAUD_DIV_N(val)			(val & 0x7)
 #define MV64XXX_I2C_BAUD_DIV_M(val)			((val & 0xf) << 3)
 
-#define	MV64XXX_I2C_REG_CONTROL_ACK			0x00000004
-#define	MV64XXX_I2C_REG_CONTROL_IFLG			0x00000008
-#define	MV64XXX_I2C_REG_CONTROL_STOP			0x00000010
-#define	MV64XXX_I2C_REG_CONTROL_START			0x00000020
-#define	MV64XXX_I2C_REG_CONTROL_TWSIEN			0x00000040
-#define	MV64XXX_I2C_REG_CONTROL_INTEN			0x00000080
+#define	MV64XXX_I2C_REG_CONTROL_ACK			BIT(2)
+#define	MV64XXX_I2C_REG_CONTROL_IFLG			BIT(3)
+#define	MV64XXX_I2C_REG_CONTROL_STOP			BIT(4)
+#define	MV64XXX_I2C_REG_CONTROL_START			BIT(5)
+#define	MV64XXX_I2C_REG_CONTROL_TWSIEN			BIT(6)
+#define	MV64XXX_I2C_REG_CONTROL_INTEN			BIT(7)
 
 /* Ctlr status values */
 #define	MV64XXX_I2C_STATUS_BUS_ERR			0x00
@@ -68,19 +68,17 @@
 #define	MV64XXX_I2C_REG_BRIDGE_TIMING			0xe0
 
 /* Bridge Control values */
-#define	MV64XXX_I2C_BRIDGE_CONTROL_WR			0x00000001
-#define	MV64XXX_I2C_BRIDGE_CONTROL_RD			0x00000002
+#define	MV64XXX_I2C_BRIDGE_CONTROL_WR			BIT(0)
+#define	MV64XXX_I2C_BRIDGE_CONTROL_RD			BIT(1)
 #define	MV64XXX_I2C_BRIDGE_CONTROL_ADDR_SHIFT		2
-#define	MV64XXX_I2C_BRIDGE_CONTROL_ADDR_EXT		0x00001000
+#define	MV64XXX_I2C_BRIDGE_CONTROL_ADDR_EXT		BIT(12)
 #define	MV64XXX_I2C_BRIDGE_CONTROL_TX_SIZE_SHIFT	13
 #define	MV64XXX_I2C_BRIDGE_CONTROL_RX_SIZE_SHIFT	16
-#define	MV64XXX_I2C_BRIDGE_CONTROL_ENABLE		0x00080000
+#define	MV64XXX_I2C_BRIDGE_CONTROL_ENABLE		BIT(19)
+#define	MV64XXX_I2C_BRIDGE_CONTROL_REPEATED_START	BIT(20)
 
 /* Bridge Status values */
-#define	MV64XXX_I2C_BRIDGE_STATUS_ERROR			0x00000001
-#define	MV64XXX_I2C_STATUS_OFFLOAD_ERROR		0xf0000001
-#define	MV64XXX_I2C_STATUS_OFFLOAD_OK			0xf0000000
-
+#define	MV64XXX_I2C_BRIDGE_STATUS_ERROR			BIT(0)
 
 /* Driver states */
 enum {
@@ -99,14 +97,12 @@ enum {
 	MV64XXX_I2C_ACTION_INVALID,
 	MV64XXX_I2C_ACTION_CONTINUE,
 	MV64XXX_I2C_ACTION_SEND_RESTART,
-	MV64XXX_I2C_ACTION_OFFLOAD_RESTART,
 	MV64XXX_I2C_ACTION_SEND_ADDR_1,
 	MV64XXX_I2C_ACTION_SEND_ADDR_2,
 	MV64XXX_I2C_ACTION_SEND_DATA,
 	MV64XXX_I2C_ACTION_RCV_DATA,
 	MV64XXX_I2C_ACTION_RCV_DATA_STOP,
 	MV64XXX_I2C_ACTION_SEND_STOP,
-	MV64XXX_I2C_ACTION_OFFLOAD_SEND_STOP,
 };
 
 struct mv64xxx_i2c_regs {
@@ -193,75 +189,6 @@ mv64xxx_i2c_prepare_for_io(struct mv64xxx_i2c_data *drv_data,
 	}
 }
 
-static int mv64xxx_i2c_offload_msg(struct mv64xxx_i2c_data *drv_data)
-{
-	unsigned long data_reg_hi = 0;
-	unsigned long data_reg_lo = 0;
-	unsigned long ctrl_reg;
-	struct i2c_msg *msg = drv_data->msgs;
-
-	if (!drv_data->offload_enabled)
-		return -EOPNOTSUPP;
-
-	/* Only regular transactions can be offloaded */
-	if ((msg->flags & ~(I2C_M_TEN | I2C_M_RD)) != 0)
-		return -EINVAL;
-
-	/* Only 1-8 byte transfers can be offloaded */
-	if (msg->len < 1 || msg->len > 8)
-		return -EINVAL;
-
-	/* Build transaction */
-	ctrl_reg = MV64XXX_I2C_BRIDGE_CONTROL_ENABLE |
-		   (msg->addr << MV64XXX_I2C_BRIDGE_CONTROL_ADDR_SHIFT);
-
-	if ((msg->flags & I2C_M_TEN) != 0)
-		ctrl_reg |=  MV64XXX_I2C_BRIDGE_CONTROL_ADDR_EXT;
-
-	if ((msg->flags & I2C_M_RD) == 0) {
-		u8 local_buf[8] = { 0 };
-
-		memcpy(local_buf, msg->buf, msg->len);
-		data_reg_lo = cpu_to_le32(*((u32 *)local_buf));
-		data_reg_hi = cpu_to_le32(*((u32 *)(local_buf+4)));
-
-		ctrl_reg |= MV64XXX_I2C_BRIDGE_CONTROL_WR |
-		    (msg->len - 1) << MV64XXX_I2C_BRIDGE_CONTROL_TX_SIZE_SHIFT;
-
-		writel(data_reg_lo,
-			drv_data->reg_base + MV64XXX_I2C_REG_TX_DATA_LO);
-		writel(data_reg_hi,
-			drv_data->reg_base + MV64XXX_I2C_REG_TX_DATA_HI);
-
-	} else {
-		ctrl_reg |= MV64XXX_I2C_BRIDGE_CONTROL_RD |
-		    (msg->len - 1) << MV64XXX_I2C_BRIDGE_CONTROL_RX_SIZE_SHIFT;
-	}
-
-	/* Execute transaction */
-	writel(ctrl_reg, drv_data->reg_base + MV64XXX_I2C_REG_BRIDGE_CONTROL);
-
-	return 0;
-}
-
-static void
-mv64xxx_i2c_update_offload_data(struct mv64xxx_i2c_data *drv_data)
-{
-	struct i2c_msg *msg = drv_data->msg;
-
-	if (msg->flags & I2C_M_RD) {
-		u32 data_reg_lo = readl(drv_data->reg_base +
-				MV64XXX_I2C_REG_RX_DATA_LO);
-		u32 data_reg_hi = readl(drv_data->reg_base +
-				MV64XXX_I2C_REG_RX_DATA_HI);
-		u8 local_buf[8] = { 0 };
-
-		*((u32 *)local_buf) = le32_to_cpu(data_reg_lo);
-		*((u32 *)(local_buf+4)) = le32_to_cpu(data_reg_hi);
-		memcpy(msg->buf, local_buf, msg->len);
-	}
-
-}
 /*
  *****************************************************************************
  *
@@ -389,16 +316,6 @@ mv64xxx_i2c_fsm(struct mv64xxx_i2c_data *drv_data, u32 status)
 		drv_data->rc = -ENXIO;
 		break;
 
-	case MV64XXX_I2C_STATUS_OFFLOAD_OK:
-		if (drv_data->send_stop || drv_data->aborting) {
-			drv_data->action = MV64XXX_I2C_ACTION_OFFLOAD_SEND_STOP;
-			drv_data->state = MV64XXX_I2C_STATE_IDLE;
-		} else {
-			drv_data->action = MV64XXX_I2C_ACTION_OFFLOAD_RESTART;
-			drv_data->state = MV64XXX_I2C_STATE_WAITING_FOR_RESTART;
-		}
-		break;
-
 	default:
 		dev_err(&drv_data->adapter.dev,
 			"mv64xxx_i2c_fsm: Ctlr Error -- state: 0x%x, "
@@ -419,25 +336,15 @@ static void mv64xxx_i2c_send_start(struct mv64xxx_i2c_data *drv_data)
 	drv_data->aborting = 0;
 	drv_data->rc = 0;
 
-	/* Can we offload this msg ? */
-	if (mv64xxx_i2c_offload_msg(drv_data) < 0) {
-		/* No, switch to standard path */
-		mv64xxx_i2c_prepare_for_io(drv_data, drv_data->msgs);
-		writel(drv_data->cntl_bits | MV64XXX_I2C_REG_CONTROL_START,
-			drv_data->reg_base + drv_data->reg_offsets.control);
-	}
+	mv64xxx_i2c_prepare_for_io(drv_data, drv_data->msgs);
+	writel(drv_data->cntl_bits | MV64XXX_I2C_REG_CONTROL_START,
+	       drv_data->reg_base + drv_data->reg_offsets.control);
 }
 
 static void
 mv64xxx_i2c_do_action(struct mv64xxx_i2c_data *drv_data)
 {
 	switch(drv_data->action) {
-	case MV64XXX_I2C_ACTION_OFFLOAD_RESTART:
-		mv64xxx_i2c_update_offload_data(drv_data);
-		writel(0, drv_data->reg_base +	MV64XXX_I2C_REG_BRIDGE_CONTROL);
-		writel(0, drv_data->reg_base +
-			MV64XXX_I2C_REG_BRIDGE_INTR_CAUSE);
-		/* FALLTHRU */
 	case MV64XXX_I2C_ACTION_SEND_RESTART:
 		/* We should only get here if we have further messages */
 		BUG_ON(drv_data->num_msgs == 0);
@@ -518,16 +425,71 @@ mv64xxx_i2c_do_action(struct mv64xxx_i2c_data *drv_data)
 		drv_data->block = 0;
 		wake_up(&drv_data->waitq);
 		break;
+	}
+}
 
-	case MV64XXX_I2C_ACTION_OFFLOAD_SEND_STOP:
-		mv64xxx_i2c_update_offload_data(drv_data);
-		writel(0, drv_data->reg_base +	MV64XXX_I2C_REG_BRIDGE_CONTROL);
-		writel(0, drv_data->reg_base +
-			MV64XXX_I2C_REG_BRIDGE_INTR_CAUSE);
-		drv_data->block = 0;
-		wake_up(&drv_data->waitq);
-		break;
+static void
+mv64xxx_i2c_read_offload_rx_data(struct mv64xxx_i2c_data *drv_data,
+				 struct i2c_msg *msg)
+{
+	u32 buf[2];
+
+	buf[0] = readl(drv_data->reg_base + MV64XXX_I2C_REG_RX_DATA_LO);
+	buf[1] = readl(drv_data->reg_base + MV64XXX_I2C_REG_RX_DATA_HI);
+
+	memcpy(msg->buf, buf, msg->len);
+}
+
+static int
+mv64xxx_i2c_intr_offload(struct mv64xxx_i2c_data *drv_data)
+{
+	u32 cause, status;
+
+	cause = readl(drv_data->reg_base +
+		      MV64XXX_I2C_REG_BRIDGE_INTR_CAUSE);
+	if (!cause)
+		return IRQ_NONE;
+
+	status = readl(drv_data->reg_base +
+		       MV64XXX_I2C_REG_BRIDGE_STATUS);
+
+	if (status & MV64XXX_I2C_BRIDGE_STATUS_ERROR) {
+		drv_data->rc = -EIO;
+		goto out;
+	}
+
+	drv_data->rc = 0;
+
+	/*
+	 * Transaction is a one message read transaction, read data
+	 * for this message.
+	 */
+	if (drv_data->num_msgs == 1 && drv_data->msgs[0].flags & I2C_M_RD) {
+		mv64xxx_i2c_read_offload_rx_data(drv_data, drv_data->msgs);
+		drv_data->msgs++;
+		drv_data->num_msgs--;
+	}
+	/*
+	 * Transaction is a two messages write/read transaction, read
+	 * data for the second (read) message.
+	 */
+	else if (drv_data->num_msgs == 2 &&
+		 !(drv_data->msgs[0].flags & I2C_M_RD) &&
+		 drv_data->msgs[1].flags & I2C_M_RD) {
+		mv64xxx_i2c_read_offload_rx_data(drv_data, drv_data->msgs + 1);
+		drv_data->msgs += 2;
+		drv_data->num_msgs -= 2;
 	}
+
+out:
+	writel(0, drv_data->reg_base +	MV64XXX_I2C_REG_BRIDGE_CONTROL);
+	writel(0, drv_data->reg_base +
+	       MV64XXX_I2C_REG_BRIDGE_INTR_CAUSE);
+	drv_data->block = 0;
+
+	wake_up(&drv_data->waitq);
+
+	return IRQ_HANDLED;
 }
 
 static irqreturn_t
@@ -540,20 +502,9 @@ mv64xxx_i2c_intr(int irq, void *dev_id)
 
 	spin_lock_irqsave(&drv_data->lock, flags);
 
-	if (drv_data->offload_enabled) {
-		while (readl(drv_data->reg_base +
-				MV64XXX_I2C_REG_BRIDGE_INTR_CAUSE)) {
-			int reg_status = readl(drv_data->reg_base +
-					MV64XXX_I2C_REG_BRIDGE_STATUS);
-			if (reg_status & MV64XXX_I2C_BRIDGE_STATUS_ERROR)
-				status = MV64XXX_I2C_STATUS_OFFLOAD_ERROR;
-			else
-				status = MV64XXX_I2C_STATUS_OFFLOAD_OK;
-			mv64xxx_i2c_fsm(drv_data, status);
-			mv64xxx_i2c_do_action(drv_data);
-			rc = IRQ_HANDLED;
-		}
-	}
+	if (drv_data->offload_enabled)
+		rc = mv64xxx_i2c_intr_offload(drv_data);
+
 	while (readl(drv_data->reg_base + drv_data->reg_offsets.control) &
 						MV64XXX_I2C_REG_CONTROL_IFLG) {
 		status = readl(drv_data->reg_base + drv_data->reg_offsets.status);
@@ -635,6 +586,117 @@ mv64xxx_i2c_execute_msg(struct mv64xxx_i2c_data *drv_data, struct i2c_msg *msg,
 	return drv_data->rc;
 }
 
+static void
+mv64xxx_i2c_prepare_tx(struct mv64xxx_i2c_data *drv_data)
+{
+	struct i2c_msg *msg = drv_data->msgs;
+	u32 buf[2];
+
+	memcpy(buf, msg->buf, msg->len);
+
+	writel(buf[0], drv_data->reg_base + MV64XXX_I2C_REG_TX_DATA_LO);
+	writel(buf[1], drv_data->reg_base + MV64XXX_I2C_REG_TX_DATA_HI);
+}
+
+static int
+mv64xxx_i2c_offload_xfer(struct mv64xxx_i2c_data *drv_data)
+{
+	struct i2c_msg *msgs = drv_data->msgs;
+	int num = drv_data->num_msgs;
+	unsigned long ctrl_reg;
+	unsigned long flags;
+
+	spin_lock_irqsave(&drv_data->lock, flags);
+
+	/* Build transaction */
+	ctrl_reg = MV64XXX_I2C_BRIDGE_CONTROL_ENABLE |
+		(msgs[0].addr << MV64XXX_I2C_BRIDGE_CONTROL_ADDR_SHIFT);
+
+	if (msgs[0].flags & I2C_M_TEN)
+		ctrl_reg |= MV64XXX_I2C_BRIDGE_CONTROL_ADDR_EXT;
+
+	/* Single write message transaction */
+	if (num == 1 && !(msgs[0].flags & I2C_M_RD)) {
+		size_t len = msgs[0].len - 1;
+
+		ctrl_reg |= MV64XXX_I2C_BRIDGE_CONTROL_WR |
+			(len << MV64XXX_I2C_BRIDGE_CONTROL_TX_SIZE_SHIFT);
+		mv64xxx_i2c_prepare_tx(drv_data);
+	}
+	/* Single read message transaction */
+	else if (num == 1 && msgs[0].flags & I2C_M_RD) {
+		size_t len = msgs[0].len - 1;
+
+		ctrl_reg |= MV64XXX_I2C_BRIDGE_CONTROL_RD |
+			(len << MV64XXX_I2C_BRIDGE_CONTROL_RX_SIZE_SHIFT);
+	}
+	/*
+	 * Transaction with one write and one read message. This is
+	 * guaranteed by the mv64xx_i2c_can_offload() checks.
+	 */
+	else if (num == 2) {
+		size_t lentx = msgs[0].len - 1;
+		size_t lenrx = msgs[1].len - 1;
+
+		ctrl_reg |=
+			MV64XXX_I2C_BRIDGE_CONTROL_RD |
+			MV64XXX_I2C_BRIDGE_CONTROL_WR |
+			(lentx << MV64XXX_I2C_BRIDGE_CONTROL_TX_SIZE_SHIFT) |
+			(lenrx << MV64XXX_I2C_BRIDGE_CONTROL_RX_SIZE_SHIFT) |
+			MV64XXX_I2C_BRIDGE_CONTROL_REPEATED_START;
+		mv64xxx_i2c_prepare_tx(drv_data);
+	}
+
+	/* Execute transaction */
+	drv_data->block = 1;
+	writel(ctrl_reg, drv_data->reg_base + MV64XXX_I2C_REG_BRIDGE_CONTROL);
+	spin_unlock_irqrestore(&drv_data->lock, flags);
+
+	mv64xxx_i2c_wait_for_completion(drv_data);
+
+	return drv_data->rc;
+}
+
+static bool
+mv64xxx_i2c_valid_offload_sz(struct i2c_msg *msg)
+{
+	return msg->len <= 8 && msg->len >= 1;
+}
+
+static bool
+mv64xxx_i2c_can_offload(struct mv64xxx_i2c_data *drv_data)
+{
+	struct i2c_msg *msgs = drv_data->msgs;
+	int num = drv_data->num_msgs;
+
+	return false;
+
+	if (!drv_data->offload_enabled)
+		return false;
+
+	/*
+	 * We can offload a transaction consisting of a single
+	 * message, as long as the message has a length between 1 and
+	 * 8 bytes.
+	 */
+	if (num == 1 && mv64xxx_i2c_valid_offload_sz(msgs))
+		return true;
+
+	/*
+	 * We can offload a transaction consisting of two messages, if
+	 * the first is a write and a second is a read, and both have
+	 * a length between 1 and 8 bytes.
+	 */
+	if (num == 2 &&
+	    mv64xxx_i2c_valid_offload_sz(msgs) &&
+	    mv64xxx_i2c_valid_offload_sz(msgs + 1) &&
+	    !(msgs[0].flags & I2C_M_RD) &&
+	    msgs[1].flags & I2C_M_RD)
+		return true;
+
+	return false;
+}
+
 /*
  *****************************************************************************
  *
@@ -658,7 +720,11 @@ mv64xxx_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg msgs[], int num)
 	drv_data->msgs = msgs;
 	drv_data->num_msgs = num;
 
-	rc = mv64xxx_i2c_execute_msg(drv_data, &msgs[0], num == 1);
+	if (mv64xxx_i2c_can_offload(drv_data))
+		rc = mv64xxx_i2c_offload_xfer(drv_data);
+	else
+		rc = mv64xxx_i2c_execute_msg(drv_data, &msgs[0], num == 1);
+
 	if (rc < 0)
 		ret = rc;
 
@@ -925,7 +991,6 @@ static struct platform_driver mv64xxx_i2c_driver = {
 	.probe	= mv64xxx_i2c_probe,
 	.remove	= mv64xxx_i2c_remove,
 	.driver	= {
-		.owner	= THIS_MODULE,
 		.name	= MV64XXX_I2C_CTLR_NAME,
 		.of_match_table = mv64xxx_i2c_of_match_table,
 	},
diff --git a/drivers/i2c/busses/i2c-mxs.c b/drivers/i2c/busses/i2c-mxs.c
index 65a21fed08b5..ff8b12c8d25f 100644
--- a/drivers/i2c/busses/i2c-mxs.c
+++ b/drivers/i2c/busses/i2c-mxs.c
@@ -307,6 +307,9 @@ static int mxs_i2c_pio_wait_xfer_end(struct mxs_i2c_dev *i2c)
 	unsigned long timeout = jiffies + msecs_to_jiffies(1000);
 
 	while (readl(i2c->regs + MXS_I2C_CTRL0) & MXS_I2C_CTRL0_RUN) {
+		if (readl(i2c->regs + MXS_I2C_CTRL1) &
+				MXS_I2C_CTRL1_NO_SLAVE_ACK_IRQ)
+			return -ENXIO;
 		if (time_after(jiffies, timeout))
 			return -ETIMEDOUT;
 		cond_resched();
@@ -808,7 +811,7 @@ static int mxs_i2c_probe(struct platform_device *pdev)
 	struct resource *res;
 	int err, irq;
 
-	i2c = devm_kzalloc(dev, sizeof(struct mxs_i2c_dev), GFP_KERNEL);
+	i2c = devm_kzalloc(dev, sizeof(*i2c), GFP_KERNEL);
 	if (!i2c)
 		return -ENOMEM;
 
@@ -890,7 +893,6 @@ static int mxs_i2c_remove(struct platform_device *pdev)
 static struct platform_driver mxs_i2c_driver = {
 	.driver = {
 		   .name = DRIVER_NAME,
-		   .owner = THIS_MODULE,
 		   .of_match_table = mxs_i2c_dt_ids,
 		   },
 	.probe = mxs_i2c_probe,
diff --git a/drivers/i2c/busses/i2c-nforce2-s4985.c b/drivers/i2c/busses/i2c-nforce2-s4985.c
index b170bdffb5de..88eda09e73c0 100644
--- a/drivers/i2c/busses/i2c-nforce2-s4985.c
+++ b/drivers/i2c/busses/i2c-nforce2-s4985.c
@@ -12,10 +12,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  */
 
 /*
diff --git a/drivers/i2c/busses/i2c-nforce2.c b/drivers/i2c/busses/i2c-nforce2.c
index ee3a76c7ae97..70b3c9158509 100644
--- a/drivers/i2c/busses/i2c-nforce2.c
+++ b/drivers/i2c/busses/i2c-nforce2.c
@@ -17,10 +17,6 @@
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
     GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 /*
diff --git a/drivers/i2c/busses/i2c-nomadik.c b/drivers/i2c/busses/i2c-nomadik.c
index 9ad038d223c4..97998946c4f6 100644
--- a/drivers/i2c/busses/i2c-nomadik.c
+++ b/drivers/i2c/busses/i2c-nomadik.c
@@ -932,7 +932,7 @@ static int nmk_i2c_runtime_resume(struct device *dev)
 
 static const struct dev_pm_ops nmk_i2c_pm = {
 	SET_LATE_SYSTEM_SLEEP_PM_OPS(nmk_i2c_suspend_late, nmk_i2c_resume_early)
-	SET_PM_RUNTIME_PM_OPS(nmk_i2c_runtime_suspend,
+	SET_RUNTIME_PM_OPS(nmk_i2c_runtime_suspend,
 			nmk_i2c_runtime_resume,
 			NULL)
 };
diff --git a/drivers/i2c/busses/i2c-ocores.c b/drivers/i2c/busses/i2c-ocores.c
index 2a4fe0b7cfb7..7249b5b1e5d0 100644
--- a/drivers/i2c/busses/i2c-ocores.c
+++ b/drivers/i2c/busses/i2c-ocores.c
@@ -480,7 +480,6 @@ static struct platform_driver ocores_i2c_driver = {
 	.probe   = ocores_i2c_probe,
 	.remove  = ocores_i2c_remove,
 	.driver  = {
-		.owner = THIS_MODULE,
 		.name = "ocores-i2c",
 		.of_match_table = ocores_i2c_match,
 		.pm = OCORES_I2C_PM,
diff --git a/drivers/i2c/busses/i2c-octeon.c b/drivers/i2c/busses/i2c-octeon.c
index 81042b08a947..6e75e016bffc 100644
--- a/drivers/i2c/busses/i2c-octeon.c
+++ b/drivers/i2c/busses/i2c-octeon.c
@@ -623,7 +623,6 @@ static struct platform_driver octeon_i2c_driver = {
 	.probe		= octeon_i2c_probe,
 	.remove		= octeon_i2c_remove,
 	.driver		= {
-		.owner	= THIS_MODULE,
 		.name	= DRV_NAME,
 		.of_match_table = octeon_i2c_match,
 	},
diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
index 0dffb0e62c3b..0e894193accf 100644
--- a/drivers/i2c/busses/i2c-omap.c
+++ b/drivers/i2c/busses/i2c-omap.c
@@ -22,10 +22,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  */
 
 #include <linux/module.h>
@@ -58,6 +54,9 @@
 /* timeout for pm runtime autosuspend */
 #define OMAP_I2C_PM_TIMEOUT		1000	/* ms */
 
+/* timeout for making decision on bus free status */
+#define OMAP_I2C_BUS_FREE_TIMEOUT (msecs_to_jiffies(10))
+
 /* For OMAP3 I2C_IV has changed to I2C_WE (wakeup enable) */
 enum {
 	OMAP_I2C_REV_REG = 0,
@@ -102,7 +101,7 @@ enum {
 #define OMAP_I2C_STAT_ROVR	(1 << 11)	/* Receive overrun */
 #define OMAP_I2C_STAT_XUDF	(1 << 10)	/* Transmit underflow */
 #define OMAP_I2C_STAT_AAS	(1 << 9)	/* Address as slave */
-#define OMAP_I2C_STAT_AD0	(1 << 8)	/* Address zero */
+#define OMAP_I2C_STAT_BF	(1 << 8)	/* Bus Free */
 #define OMAP_I2C_STAT_XRDY	(1 << 4)	/* Transmit data ready */
 #define OMAP_I2C_STAT_RRDY	(1 << 3)	/* Receive data ready */
 #define OMAP_I2C_STAT_ARDY	(1 << 2)	/* Register access ready */
@@ -150,16 +149,20 @@ enum {
 #define OMAP_I2C_SCLH_HSSCLH	8
 
 /* I2C System Test Register (OMAP_I2C_SYSTEST): */
-#ifdef DEBUG
 #define OMAP_I2C_SYSTEST_ST_EN		(1 << 15)	/* System test enable */
 #define OMAP_I2C_SYSTEST_FREE		(1 << 14)	/* Free running mode */
 #define OMAP_I2C_SYSTEST_TMODE_MASK	(3 << 12)	/* Test mode select */
 #define OMAP_I2C_SYSTEST_TMODE_SHIFT	(12)		/* Test mode select */
+/* Functional mode */
+#define OMAP_I2C_SYSTEST_SCL_I_FUNC	(1 << 8)	/* SCL line input value */
+#define OMAP_I2C_SYSTEST_SCL_O_FUNC	(1 << 7)	/* SCL line output value */
+#define OMAP_I2C_SYSTEST_SDA_I_FUNC	(1 << 6)	/* SDA line input value */
+#define OMAP_I2C_SYSTEST_SDA_O_FUNC	(1 << 5)	/* SDA line output value */
+/* SDA/SCL IO mode */
 #define OMAP_I2C_SYSTEST_SCL_I		(1 << 3)	/* SCL line sense in */
 #define OMAP_I2C_SYSTEST_SCL_O		(1 << 2)	/* SCL line drive out */
 #define OMAP_I2C_SYSTEST_SDA_I		(1 << 1)	/* SDA line sense in */
 #define OMAP_I2C_SYSTEST_SDA_O		(1 << 0)	/* SDA line drive out */
-#endif
 
 /* OCP_SYSSTATUS bit definitions */
 #define SYSS_RESETDONE_MASK		(1 << 0)
@@ -206,6 +209,9 @@ struct omap_i2c_dev {
 						 */
 	u32			rev;
 	unsigned		b_hw:1;		/* bad h/w fixes */
+	unsigned		bb_valid:1;	/* true when BB-bit reflects
+						 * the I2C bus state
+						 */
 	unsigned		receiver:1;	/* true when we're in receiver mode */
 	u16			iestate;	/* Saved interrupt register */
 	u16			pscstate;
@@ -294,6 +300,12 @@ static void __omap_i2c_init(struct omap_i2c_dev *dev)
 	omap_i2c_write_reg(dev, OMAP_I2C_CON_REG, OMAP_I2C_CON_EN);
 
 	/*
+	 * NOTE: right after setting CON_EN, STAT_BB could be 0 while the
+	 * bus is busy. It will be changed to 1 on the next IP FCLK clock.
+	 * udelay(1) will be enough to fix that.
+	 */
+
+	/*
 	 * Don't write to this register if the IE state is 0 as it can
 	 * cause deadlock.
 	 */
@@ -332,7 +344,12 @@ static int omap_i2c_reset(struct omap_i2c_dev *dev)
 		/* SYSC register is cleared by the reset; rewrite it */
 		omap_i2c_write_reg(dev, OMAP_I2C_SYSC_REG, sysc);
 
+		if (dev->rev > OMAP_I2C_REV_ON_3430_3530) {
+			/* Schedule I2C-bus monitoring on the next transfer */
+			dev->bb_valid = 0;
+		}
 	}
+
 	return 0;
 }
 
@@ -445,6 +462,11 @@ static int omap_i2c_init(struct omap_i2c_dev *dev)
 	dev->scllstate = scll;
 	dev->sclhstate = sclh;
 
+	if (dev->rev <= OMAP_I2C_REV_ON_3430_3530) {
+		/* Not implemented */
+		dev->bb_valid = 1;
+	}
+
 	__omap_i2c_init(dev);
 
 	return 0;
@@ -469,6 +491,91 @@ static int omap_i2c_wait_for_bb(struct omap_i2c_dev *dev)
 	return 0;
 }
 
+/*
+ * Wait while BB-bit doesn't reflect the I2C bus state
+ *
+ * In a multimaster environment, after IP software reset, BB-bit value doesn't
+ * correspond to the current bus state. It may happen what BB-bit will be 0,
+ * while the bus is busy due to another I2C master activity.
+ * Here are BB-bit values after reset:
+ *     SDA   SCL   BB   NOTES
+ *       0     0    0   1, 2
+ *       1     0    0   1, 2
+ *       0     1    1
+ *       1     1    0   3
+ * Later, if IP detect SDA=0 and SCL=1 (ACK) or SDA 1->0 while SCL=1 (START)
+ * combinations on the bus, it set BB-bit to 1.
+ * If IP detect SDA 0->1 while SCL=1 (STOP) combination on the bus,
+ * it set BB-bit to 0 and BF to 1.
+ * BB and BF bits correctly tracks the bus state while IP is suspended
+ * BB bit became valid on the next FCLK clock after CON_EN bit set
+ *
+ * NOTES:
+ * 1. Any transfer started when BB=0 and bus is busy wouldn't be
+ *    completed by IP and results in controller timeout.
+ * 2. Any transfer started when BB=0 and SCL=0 results in IP
+ *    starting to drive SDA low. In that case IP corrupt data
+ *    on the bus.
+ * 3. Any transfer started in the middle of another master's transfer
+ *    results in unpredictable results and data corruption
+ */
+static int omap_i2c_wait_for_bb_valid(struct omap_i2c_dev *dev)
+{
+	unsigned long bus_free_timeout = 0;
+	unsigned long timeout;
+	int bus_free = 0;
+	u16 stat, systest;
+
+	if (dev->bb_valid)
+		return 0;
+
+	timeout = jiffies + OMAP_I2C_TIMEOUT;
+	while (1) {
+		stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
+		/*
+		 * We will see BB or BF event in a case IP had detected any
+		 * activity on the I2C bus. Now IP correctly tracks the bus
+		 * state. BB-bit value is valid.
+		 */
+		if (stat & (OMAP_I2C_STAT_BB | OMAP_I2C_STAT_BF))
+			break;
+
+		/*
+		 * Otherwise, we must look signals on the bus to make
+		 * the right decision.
+		 */
+		systest = omap_i2c_read_reg(dev, OMAP_I2C_SYSTEST_REG);
+		if ((systest & OMAP_I2C_SYSTEST_SCL_I_FUNC) &&
+		    (systest & OMAP_I2C_SYSTEST_SDA_I_FUNC)) {
+			if (!bus_free) {
+				bus_free_timeout = jiffies +
+					OMAP_I2C_BUS_FREE_TIMEOUT;
+				bus_free = 1;
+			}
+
+			/*
+			 * SDA and SCL lines was high for 10 ms without bus
+			 * activity detected. The bus is free. Consider
+			 * BB-bit value is valid.
+			 */
+			if (time_after(jiffies, bus_free_timeout))
+				break;
+		} else {
+			bus_free = 0;
+		}
+
+		if (time_after(jiffies, timeout)) {
+			dev_warn(dev->dev, "timeout waiting for bus ready\n");
+			return -ETIMEDOUT;
+		}
+
+		msleep(1);
+	}
+
+	dev->bb_valid = 1;
+	return 0;
+}
+
 static void omap_i2c_resize_fifo(struct omap_i2c_dev *dev, u8 size, bool is_rx)
 {
 	u16		buf;
@@ -561,7 +668,11 @@ static int omap_i2c_xfer_msg(struct i2c_adapter *adap,
 
 	if (!dev->b_hw && stop)
 		w |= OMAP_I2C_CON_STP;
-
+	/*
+	 * NOTE: STAT_BB bit could became 1 here if another master occupy
+	 * the bus. IP successfully complete transfer when the bus will be
+	 * free again (BB reset to 0).
+	 */
 	omap_i2c_write_reg(dev, OMAP_I2C_CON_REG, w);
 
 	/*
@@ -604,13 +715,15 @@ static int omap_i2c_xfer_msg(struct i2c_adapter *adap,
 		return 0;
 
 	/* We have an error */
-	if (dev->cmd_err & (OMAP_I2C_STAT_AL | OMAP_I2C_STAT_ROVR |
-			    OMAP_I2C_STAT_XUDF)) {
+	if (dev->cmd_err & (OMAP_I2C_STAT_ROVR | OMAP_I2C_STAT_XUDF)) {
 		omap_i2c_reset(dev);
 		__omap_i2c_init(dev);
 		return -EIO;
 	}
 
+	if (dev->cmd_err & OMAP_I2C_STAT_AL)
+		return -EAGAIN;
+
 	if (dev->cmd_err & OMAP_I2C_STAT_NACK) {
 		if (msg->flags & I2C_M_IGNORE_NAK)
 			return 0;
@@ -639,6 +752,10 @@ omap_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg msgs[], int num)
 	if (r < 0)
 		goto out;
 
+	r = omap_i2c_wait_for_bb_valid(dev);
+	if (r < 0)
+		goto out;
+
 	r = omap_i2c_wait_for_bb(dev);
 	if (r < 0)
 		goto out;
@@ -926,14 +1043,12 @@ omap_i2c_isr_thread(int this_irq, void *dev_id)
 		if (stat & OMAP_I2C_STAT_NACK) {
 			err |= OMAP_I2C_STAT_NACK;
 			omap_i2c_ack_stat(dev, OMAP_I2C_STAT_NACK);
-			break;
 		}
 
 		if (stat & OMAP_I2C_STAT_AL) {
 			dev_err(dev->dev, "Arbitration lost\n");
 			err |= OMAP_I2C_STAT_AL;
 			omap_i2c_ack_stat(dev, OMAP_I2C_STAT_AL);
-			break;
 		}
 
 		/*
@@ -958,11 +1073,13 @@ omap_i2c_isr_thread(int this_irq, void *dev_id)
 			if (dev->fifo_size)
 				num_bytes = dev->buf_len;
 
-			omap_i2c_receive_data(dev, num_bytes, true);
-
-			if (dev->errata & I2C_OMAP_ERRATA_I207)
+			if (dev->errata & I2C_OMAP_ERRATA_I207) {
 				i2c_omap_errata_i207(dev, stat);
+				num_bytes = (omap_i2c_read_reg(dev,
+					OMAP_I2C_BUFSTAT_REG) >> 8) & 0x3F;
+			}
 
+			omap_i2c_receive_data(dev, num_bytes, true);
 			omap_i2c_ack_stat(dev, OMAP_I2C_STAT_RDR);
 			continue;
 		}
@@ -1284,7 +1401,6 @@ static int omap_i2c_remove(struct platform_device *pdev)
 }
 
 #ifdef CONFIG_PM
-#ifdef CONFIG_PM_RUNTIME
 static int omap_i2c_runtime_suspend(struct device *dev)
 {
 	struct platform_device *pdev = to_platform_device(dev);
@@ -1322,7 +1438,6 @@ static int omap_i2c_runtime_resume(struct device *dev)
 
 	return 0;
 }
-#endif /* CONFIG_PM_RUNTIME */
 
 static struct dev_pm_ops omap_i2c_pm_ops = {
 	SET_RUNTIME_PM_OPS(omap_i2c_runtime_suspend,
@@ -1338,7 +1453,6 @@ static struct platform_driver omap_i2c_driver = {
 	.remove		= omap_i2c_remove,
 	.driver		= {
 		.name	= "omap_i2c",
-		.owner	= THIS_MODULE,
 		.pm	= OMAP_I2C_PM_OPS,
 		.of_match_table = of_match_ptr(omap_i2c_of_match),
 	},
diff --git a/drivers/i2c/busses/i2c-opal.c b/drivers/i2c/busses/i2c-opal.c
new file mode 100644
index 000000000000..16f90b1a7508
--- /dev/null
+++ b/drivers/i2c/busses/i2c-opal.c
@@ -0,0 +1,294 @@
+/*
+ * IBM OPAL I2C driver
+ * Copyright (C) 2014 IBM
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.
+ */
+
+#include <linux/device.h>
+#include <linux/i2c.h>
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+
+#include <asm/firmware.h>
+#include <asm/opal.h>
+
+static int i2c_opal_translate_error(int rc)
+{
+	switch (rc) {
+	case OPAL_NO_MEM:
+		return -ENOMEM;
+	case OPAL_PARAMETER:
+		return -EINVAL;
+	case OPAL_I2C_ARBT_LOST:
+		return -EAGAIN;
+	case OPAL_I2C_TIMEOUT:
+		return -ETIMEDOUT;
+	case OPAL_I2C_NACK_RCVD:
+		return -ENXIO;
+	case OPAL_I2C_STOP_ERR:
+		return -EBUSY;
+	default:
+		return -EIO;
+	}
+}
+
+static int i2c_opal_send_request(u32 bus_id, struct opal_i2c_request *req)
+{
+	struct opal_msg msg;
+	int token, rc;
+
+	token = opal_async_get_token_interruptible();
+	if (token < 0) {
+		if (token != -ERESTARTSYS)
+			pr_err("Failed to get the async token\n");
+
+		return token;
+	}
+
+	rc = opal_i2c_request(token, bus_id, req);
+	if (rc != OPAL_ASYNC_COMPLETION) {
+		rc = i2c_opal_translate_error(rc);
+		goto exit;
+	}
+
+	rc = opal_async_wait_response(token, &msg);
+	if (rc)
+		goto exit;
+
+	rc = be64_to_cpu(msg.params[1]);
+	if (rc != OPAL_SUCCESS) {
+		rc = i2c_opal_translate_error(rc);
+		goto exit;
+	}
+
+exit:
+	opal_async_release_token(token);
+	return rc;
+}
+
+static int i2c_opal_master_xfer(struct i2c_adapter *adap, struct i2c_msg *msgs,
+				int num)
+{
+	unsigned long opal_id = (unsigned long)adap->algo_data;
+	struct opal_i2c_request req;
+	int rc, i;
+
+	/* We only support fairly simple combinations here of one
+	 * or two messages
+	 */
+	memset(&req, 0, sizeof(req));
+	switch(num) {
+	case 0:
+		return 0;
+	case 1:
+		req.type = (msgs[0].flags & I2C_M_RD) ?
+			OPAL_I2C_RAW_READ : OPAL_I2C_RAW_WRITE;
+		req.addr = cpu_to_be16(msgs[0].addr);
+		req.size = cpu_to_be32(msgs[0].len);
+		req.buffer_ra = cpu_to_be64(__pa(msgs[0].buf));
+		break;
+	case 2:
+		/* For two messages, we basically support only simple
+		 * smbus transactions of a write plus a read. We might
+		 * want to allow also two writes but we'd have to bounce
+		 * the data into a single buffer.
+		 */
+		if ((msgs[0].flags & I2C_M_RD) || !(msgs[1].flags & I2C_M_RD))
+			return -EOPNOTSUPP;
+		if (msgs[0].len > 4)
+			return -EOPNOTSUPP;
+		if (msgs[0].addr != msgs[1].addr)
+			return -EOPNOTSUPP;
+		req.type = OPAL_I2C_SM_READ;
+		req.addr = cpu_to_be16(msgs[0].addr);
+		req.subaddr_sz = msgs[0].len;
+		for (i = 0; i < msgs[0].len; i++)
+			req.subaddr = (req.subaddr << 8) | msgs[0].buf[i];
+		req.subaddr = cpu_to_be32(req.subaddr);
+		req.size = cpu_to_be32(msgs[1].len);
+		req.buffer_ra = cpu_to_be64(__pa(msgs[1].buf));
+		break;
+	default:
+		return -EOPNOTSUPP;
+	}
+
+	rc = i2c_opal_send_request(opal_id, &req);
+	if (rc)
+		return rc;
+
+	return num;
+}
+
+static int i2c_opal_smbus_xfer(struct i2c_adapter *adap, u16 addr,
+			       unsigned short flags, char read_write,
+			       u8 command, int size, union i2c_smbus_data *data)
+{
+	unsigned long opal_id = (unsigned long)adap->algo_data;
+	struct opal_i2c_request req;
+	u8 local[2];
+	int rc;
+
+	memset(&req, 0, sizeof(req));
+
+	req.addr = cpu_to_be16(addr);
+	switch (size) {
+	case I2C_SMBUS_BYTE:
+		req.buffer_ra = cpu_to_be64(__pa(&data->byte));
+		req.size = cpu_to_be32(1);
+		/* Fall through */
+	case I2C_SMBUS_QUICK:
+		req.type = (read_write == I2C_SMBUS_READ) ?
+			OPAL_I2C_RAW_READ : OPAL_I2C_RAW_WRITE;
+		break;
+	case I2C_SMBUS_BYTE_DATA:
+		req.buffer_ra = cpu_to_be64(__pa(&data->byte));
+		req.size = cpu_to_be32(1);
+		req.subaddr = cpu_to_be32(command);
+		req.subaddr_sz = 1;
+		req.type = (read_write == I2C_SMBUS_READ) ?
+			OPAL_I2C_SM_READ : OPAL_I2C_SM_WRITE;
+		break;
+	case I2C_SMBUS_WORD_DATA:
+		if (!read_write) {
+			local[0] = data->word & 0xff;
+			local[1] = (data->word >> 8) & 0xff;
+		}
+		req.buffer_ra = cpu_to_be64(__pa(local));
+		req.size = cpu_to_be32(2);
+		req.subaddr = cpu_to_be32(command);
+		req.subaddr_sz = 1;
+		req.type = (read_write == I2C_SMBUS_READ) ?
+			OPAL_I2C_SM_READ : OPAL_I2C_SM_WRITE;
+		break;
+	case I2C_SMBUS_I2C_BLOCK_DATA:
+		req.buffer_ra = cpu_to_be64(__pa(&data->block[1]));
+		req.size = cpu_to_be32(data->block[0]);
+		req.subaddr = cpu_to_be32(command);
+		req.subaddr_sz = 1;
+		req.type = (read_write == I2C_SMBUS_READ) ?
+			OPAL_I2C_SM_READ : OPAL_I2C_SM_WRITE;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	rc = i2c_opal_send_request(opal_id, &req);
+	if (!rc && read_write && size == I2C_SMBUS_WORD_DATA) {
+		data->word = ((u16)local[1]) << 8;
+		data->word |= local[0];
+	}
+
+	return rc;
+}
+
+static u32 i2c_opal_func(struct i2c_adapter *adapter)
+{
+	return I2C_FUNC_I2C | I2C_FUNC_SMBUS_QUICK | I2C_FUNC_SMBUS_BYTE |
+	       I2C_FUNC_SMBUS_BYTE_DATA | I2C_FUNC_SMBUS_WORD_DATA |
+	       I2C_FUNC_SMBUS_I2C_BLOCK;
+}
+
+static const struct i2c_algorithm i2c_opal_algo = {
+	.master_xfer	= i2c_opal_master_xfer,
+	.smbus_xfer	= i2c_opal_smbus_xfer,
+	.functionality	= i2c_opal_func,
+};
+
+static int i2c_opal_probe(struct platform_device *pdev)
+{
+	struct i2c_adapter	*adapter;
+	const char		*pname;
+	u32			opal_id;
+	int			rc;
+
+	if (!pdev->dev.of_node)
+		return -ENODEV;
+
+	rc = of_property_read_u32(pdev->dev.of_node, "ibm,opal-id", &opal_id);
+	if (rc) {
+		dev_err(&pdev->dev, "Missing ibm,opal-id property !\n");
+		return -EIO;
+	}
+
+	adapter = devm_kzalloc(&pdev->dev, sizeof(*adapter), GFP_KERNEL);
+	if (!adapter)
+		return -ENOMEM;
+
+	adapter->algo = &i2c_opal_algo;
+	adapter->algo_data = (void *)(unsigned long)opal_id;
+	adapter->dev.parent = &pdev->dev;
+	adapter->dev.of_node = of_node_get(pdev->dev.of_node);
+	pname = of_get_property(pdev->dev.of_node, "ibm,port-name", NULL);
+	if (pname)
+		strlcpy(adapter->name, pname, sizeof(adapter->name));
+	else
+		strlcpy(adapter->name, "opal", sizeof(adapter->name));
+
+	platform_set_drvdata(pdev, adapter);
+	rc = i2c_add_adapter(adapter);
+	if (rc)
+		dev_err(&pdev->dev, "Failed to register the i2c adapter\n");
+
+	return rc;
+}
+
+static int i2c_opal_remove(struct platform_device *pdev)
+{
+	struct i2c_adapter *adapter = platform_get_drvdata(pdev);
+
+	i2c_del_adapter(adapter);
+
+	return 0;
+}
+
+static const struct of_device_id i2c_opal_of_match[] = {
+	{
+		.compatible = "ibm,opal-i2c",
+	},
+	{ }
+};
+MODULE_DEVICE_TABLE(of, i2c_opal_of_match);
+
+static struct platform_driver i2c_opal_driver = {
+	.probe	= i2c_opal_probe,
+	.remove	= i2c_opal_remove,
+	.driver	= {
+		.name		= "i2c-opal",
+		.of_match_table	= i2c_opal_of_match,
+	},
+};
+
+static int __init i2c_opal_init(void)
+{
+	if (!firmware_has_feature(FW_FEATURE_OPAL))
+		return -ENODEV;
+
+	return platform_driver_register(&i2c_opal_driver);
+}
+module_init(i2c_opal_init);
+
+static void __exit i2c_opal_exit(void)
+{
+	return platform_driver_unregister(&i2c_opal_driver);
+}
+module_exit(i2c_opal_exit);
+
+MODULE_AUTHOR("Neelesh Gupta <neelegup@linux.vnet.ibm.com>");
+MODULE_DESCRIPTION("IBM OPAL I2C driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/i2c/busses/i2c-parport-light.c b/drivers/i2c/busses/i2c-parport-light.c
index 62f55fe624cb..1bcdd10b68b9 100644
--- a/drivers/i2c/busses/i2c-parport-light.c
+++ b/drivers/i2c/busses/i2c-parport-light.c
@@ -18,10 +18,6 @@
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    GNU General Public License for more details.
-
-   You should have received a copy of the GNU General Public License
-   along with this program; if not, write to the Free Software
-   Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  * ------------------------------------------------------------------------ */
 
 #include <linux/kernel.h>
@@ -187,7 +183,6 @@ static int i2c_parport_remove(struct platform_device *pdev)
 
 static struct platform_driver i2c_parport_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= DRVNAME,
 	},
 	.probe		= i2c_parport_probe,
diff --git a/drivers/i2c/busses/i2c-parport.c b/drivers/i2c/busses/i2c-parport.c
index a27aae2d6757..a1fac5aa9bae 100644
--- a/drivers/i2c/busses/i2c-parport.c
+++ b/drivers/i2c/busses/i2c-parport.c
@@ -18,10 +18,6 @@
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    GNU General Public License for more details.
-
-   You should have received a copy of the GNU General Public License
-   along with this program; if not, write to the Free Software
-   Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  * ------------------------------------------------------------------------ */
 
 #include <linux/kernel.h>
diff --git a/drivers/i2c/busses/i2c-parport.h b/drivers/i2c/busses/i2c-parport.h
index e572f3aac0f7..4e1294536805 100644
--- a/drivers/i2c/busses/i2c-parport.h
+++ b/drivers/i2c/busses/i2c-parport.h
@@ -12,10 +12,6 @@
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    GNU General Public License for more details.
-
-   You should have received a copy of the GNU General Public License
-   along with this program; if not, write to the Free Software
-   Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  * ------------------------------------------------------------------------ */
 
 #define PORT_DATA	0
diff --git a/drivers/i2c/busses/i2c-pasemi.c b/drivers/i2c/busses/i2c-pasemi.c
index 7a9dce43e115..df1dbc92a024 100644
--- a/drivers/i2c/busses/i2c-pasemi.c
+++ b/drivers/i2c/busses/i2c-pasemi.c
@@ -11,10 +11,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
  */
 
 #include <linux/module.h>
diff --git a/drivers/i2c/busses/i2c-pca-isa.c b/drivers/i2c/busses/i2c-pca-isa.c
index 323f061a3163..e0eb4ca0102e 100644
--- a/drivers/i2c/busses/i2c-pca-isa.c
+++ b/drivers/i2c/busses/i2c-pca-isa.c
@@ -12,10 +12,6 @@
  *  but WITHOUT ANY WARRANTY; without even the implied warranty of
  *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  *  GNU General Public License for more details.
- *
- *  You should have received a copy of the GNU General Public License
- *  along with this program; if not, write to the Free Software
- *  Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  */
 
 #include <linux/kernel.h>
diff --git a/drivers/i2c/busses/i2c-pca-platform.c b/drivers/i2c/busses/i2c-pca-platform.c
index 845f12598e79..6336f02ec566 100644
--- a/drivers/i2c/busses/i2c-pca-platform.c
+++ b/drivers/i2c/busses/i2c-pca-platform.c
@@ -280,7 +280,6 @@ static struct platform_driver i2c_pca_pf_driver = {
 	.remove = i2c_pca_pf_remove,
 	.driver = {
 		.name = "i2c-pca-platform",
-		.owner = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/i2c/busses/i2c-piix4.c b/drivers/i2c/busses/i2c-piix4.c
index a6f54ba27e2a..67cbec6796a0 100644
--- a/drivers/i2c/busses/i2c-piix4.c
+++ b/drivers/i2c/busses/i2c-piix4.c
@@ -11,10 +11,6 @@
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
     GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 /*
diff --git a/drivers/i2c/busses/i2c-pmcmsp.c b/drivers/i2c/busses/i2c-pmcmsp.c
index 8564768fee32..44f03eed00dd 100644
--- a/drivers/i2c/busses/i2c-pmcmsp.c
+++ b/drivers/i2c/busses/i2c-pmcmsp.c
@@ -18,10 +18,6 @@
  *  ANY THEORY OF LIABILITY, WHETHER IN  CONTRACT, STRICT LIABILITY, OR TORT
  *  (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
  *  THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
- *
- *  You should have received a copy of the  GNU General Public License along
- *  with this program; if not, write  to the Free Software Foundation, Inc.,
- *  675 Mass Ave, Cambridge, MA 02139, USA.
  */
 
 #include <linux/kernel.h>
@@ -628,7 +624,6 @@ static struct platform_driver pmcmsptwi_driver = {
 	.remove	= pmcmsptwi_remove,
 	.driver = {
 		.name	= DRV_NAME,
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/i2c/busses/i2c-pnx.c b/drivers/i2c/busses/i2c-pnx.c
index dc7ff829ad78..e814a36d9b78 100644
--- a/drivers/i2c/busses/i2c-pnx.c
+++ b/drivers/i2c/busses/i2c-pnx.c
@@ -751,7 +751,6 @@ MODULE_DEVICE_TABLE(of, i2c_pnx_of_match);
 static struct platform_driver i2c_pnx_driver = {
 	.driver = {
 		.name = "pnx-i2c",
-		.owner = THIS_MODULE,
 		.of_match_table = of_match_ptr(i2c_pnx_of_match),
 		.pm = PNX_I2C_PM,
 	},
diff --git a/drivers/i2c/busses/i2c-powermac.c b/drivers/i2c/busses/i2c-powermac.c
index 01e967763c2a..60a53c169ed2 100644
--- a/drivers/i2c/busses/i2c-powermac.c
+++ b/drivers/i2c/busses/i2c-powermac.c
@@ -14,10 +14,6 @@
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
     GNU General Public License for more details.
 
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
-
 */
 
 #include <linux/module.h>
diff --git a/drivers/i2c/busses/i2c-puv3.c b/drivers/i2c/busses/i2c-puv3.c
index c83fc3ccdd2b..82b6f02544da 100644
--- a/drivers/i2c/busses/i2c-puv3.c
+++ b/drivers/i2c/busses/i2c-puv3.c
@@ -270,7 +270,6 @@ static struct platform_driver puv3_i2c_driver = {
 	.remove		= puv3_i2c_remove,
 	.driver		= {
 		.name	= "PKUnity-v3-I2C",
-		.owner	= THIS_MODULE,
 		.pm	= PUV3_I2C_PM,
 	}
 };
diff --git a/drivers/i2c/busses/i2c-pxa.c b/drivers/i2c/busses/i2c-pxa.c
index be671f7a0e06..d9c0d6a17ad6 100644
--- a/drivers/i2c/busses/i2c-pxa.c
+++ b/drivers/i2c/busses/i2c-pxa.c
@@ -885,7 +885,9 @@ static void i2c_pxa_irq_txempty(struct pxa_i2c *i2c, u32 isr)
 		return; /* ignore */
 	}
 
-	if (isr & ISR_BED) {
+	if ((isr & ISR_BED) &&
+		(!((i2c->msg->flags & I2C_M_IGNORE_NAK) &&
+			(isr & ISR_ACKNAK)))) {
 		int ret = BUS_ERROR;
 
 		/*
@@ -919,12 +921,14 @@ static void i2c_pxa_irq_txempty(struct pxa_i2c *i2c, u32 isr)
 		icr |= ICR_ALDIE | ICR_TB;
 
 		/*
-		 * If this is the last byte of the last message, send
-		 * a STOP.
+		 * If this is the last byte of the last message or last byte
+		 * of any message with I2C_M_STOP (e.g. SCCB), send a STOP.
 		 */
-		if (i2c->msg_ptr == i2c->msg->len &&
-		    i2c->msg_idx == i2c->msg_num - 1)
-			icr |= ICR_STOP;
+		if ((i2c->msg_ptr == i2c->msg->len) &&
+			((i2c->msg->flags & I2C_M_STOP) ||
+			(i2c->msg_idx == i2c->msg_num - 1)))
+				icr |= ICR_STOP;
+
 	} else if (i2c->msg_idx < i2c->msg_num - 1) {
 		/*
 		 * Next segment of the message.
@@ -1071,7 +1075,8 @@ static int i2c_pxa_xfer(struct i2c_adapter *adap, struct i2c_msg msgs[], int num
 
 static u32 i2c_pxa_functionality(struct i2c_adapter *adap)
 {
-	return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL;
+	return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL |
+		I2C_FUNC_PROTOCOL_MANGLING | I2C_FUNC_NOSTART;
 }
 
 static const struct i2c_algorithm i2c_pxa_algorithm = {
@@ -1328,7 +1333,6 @@ static struct platform_driver i2c_pxa_driver = {
 	.remove		= i2c_pxa_remove,
 	.driver		= {
 		.name	= "pxa2xx-i2c",
-		.owner	= THIS_MODULE,
 		.pm	= I2C_PXA_DEV_PM_OPS,
 		.of_match_table = i2c_pxa_dt_ids,
 	},
diff --git a/drivers/i2c/busses/i2c-qup.c b/drivers/i2c/busses/i2c-qup.c
index 092d89bd3224..4dad23bdffbe 100644
--- a/drivers/i2c/busses/i2c-qup.c
+++ b/drivers/i2c/busses/i2c-qup.c
@@ -764,7 +764,6 @@ static struct platform_driver qup_i2c_driver = {
 	.remove = qup_i2c_remove,
 	.driver = {
 		.name = "i2c_qup",
-		.owner = THIS_MODULE,
 		.pm = &qup_i2c_qup_pm_ops,
 		.of_match_table = qup_i2c_dt_match,
 	},
diff --git a/drivers/i2c/busses/i2c-rcar.c b/drivers/i2c/busses/i2c-rcar.c
index e506fcd3ca04..71a6e07eb7ab 100644
--- a/drivers/i2c/busses/i2c-rcar.c
+++ b/drivers/i2c/busses/i2c-rcar.c
@@ -48,6 +48,12 @@
 #define ICMAR	0x20	/* master address */
 #define ICRXTX	0x24	/* data port */
 
+/* ICSCR */
+#define SDBS	(1 << 3)	/* slave data buffer select */
+#define SIE	(1 << 2)	/* slave interface enable */
+#define GCAE	(1 << 1)	/* general call address enable */
+#define FNA	(1 << 0)	/* forced non acknowledgment */
+
 /* ICMCR */
 #define MDBS	(1 << 7)	/* non-fifo mode switch */
 #define FSCL	(1 << 6)	/* override SCL pin */
@@ -58,6 +64,15 @@
 #define FSB	(1 << 1)	/* force stop bit */
 #define ESG	(1 << 0)	/* en startbit gen */
 
+/* ICSSR (also for ICSIER) */
+#define GCAR	(1 << 6)	/* general call received */
+#define STM	(1 << 5)	/* slave transmit mode */
+#define SSR	(1 << 4)	/* stop received */
+#define SDE	(1 << 3)	/* slave data empty */
+#define SDT	(1 << 2)	/* slave data transmitted */
+#define SDR	(1 << 1)	/* slave data received */
+#define SAR	(1 << 0)	/* slave addr received */
+
 /* ICMSR (also for ICMIE) */
 #define MNR	(1 << 6)	/* nack received */
 #define MAL	(1 << 5)	/* arbitration lost */
@@ -103,6 +118,7 @@ struct rcar_i2c_priv {
 	u32 icccr;
 	u32 flags;
 	enum rcar_i2c_type devtype;
+	struct i2c_client *slave;
 };
 
 #define rcar_i2c_priv_to_dev(p)		((p)->adap.dev.parent)
@@ -126,15 +142,6 @@ static u32 rcar_i2c_read(struct rcar_i2c_priv *priv, int reg)
 
 static void rcar_i2c_init(struct rcar_i2c_priv *priv)
 {
-	/*
-	 * reset slave mode.
-	 * slave mode is not used on this driver
-	 */
-	rcar_i2c_write(priv, ICSIER, 0);
-	rcar_i2c_write(priv, ICSAR, 0);
-	rcar_i2c_write(priv, ICSCR, 0);
-	rcar_i2c_write(priv, ICSSR, 0);
-
 	/* reset master mode */
 	rcar_i2c_write(priv, ICMIER, 0);
 	rcar_i2c_write(priv, ICMCR, 0);
@@ -195,7 +202,7 @@ static int rcar_i2c_clock_calculate(struct rcar_i2c_priv *priv,
 	 */
 	rate = clk_get_rate(priv->clk);
 	cdf = rate / 20000000;
-	if (cdf >= 1 << cdf_width) {
+	if (cdf >= 1U << cdf_width) {
 		dev_err(dev, "Input clock %lu too high\n", rate);
 		return -EIO;
 	}
@@ -245,7 +252,7 @@ scgd_find:
 	return 0;
 }
 
-static int rcar_i2c_prepare_msg(struct rcar_i2c_priv *priv)
+static void rcar_i2c_prepare_msg(struct rcar_i2c_priv *priv)
 {
 	int read = !!rcar_i2c_is_recv(priv);
 
@@ -253,8 +260,6 @@ static int rcar_i2c_prepare_msg(struct rcar_i2c_priv *priv)
 	rcar_i2c_write(priv, ICMSR, 0);
 	rcar_i2c_write(priv, ICMCR, RCAR_BUS_PHASE_START);
 	rcar_i2c_write(priv, ICMIER, read ? RCAR_IRQ_RECV : RCAR_IRQ_SEND);
-
-	return 0;
 }
 
 /*
@@ -362,18 +367,83 @@ static int rcar_i2c_irq_recv(struct rcar_i2c_priv *priv, u32 msr)
 	return 0;
 }
 
+static bool rcar_i2c_slave_irq(struct rcar_i2c_priv *priv)
+{
+	u32 ssr_raw, ssr_filtered;
+	u8 value;
+
+	ssr_raw = rcar_i2c_read(priv, ICSSR) & 0xff;
+	ssr_filtered = ssr_raw & rcar_i2c_read(priv, ICSIER);
+
+	if (!ssr_filtered)
+		return false;
+
+	/* address detected */
+	if (ssr_filtered & SAR) {
+		/* read or write request */
+		if (ssr_raw & STM) {
+			i2c_slave_event(priv->slave, I2C_SLAVE_REQ_READ_START, &value);
+			rcar_i2c_write(priv, ICRXTX, value);
+			rcar_i2c_write(priv, ICSIER, SDE | SSR | SAR);
+		} else {
+			i2c_slave_event(priv->slave, I2C_SLAVE_REQ_WRITE_START, &value);
+			rcar_i2c_read(priv, ICRXTX);	/* dummy read */
+			rcar_i2c_write(priv, ICSIER, SDR | SSR | SAR);
+		}
+
+		rcar_i2c_write(priv, ICSSR, ~SAR & 0xff);
+	}
+
+	/* master sent stop */
+	if (ssr_filtered & SSR) {
+		i2c_slave_event(priv->slave, I2C_SLAVE_STOP, &value);
+		rcar_i2c_write(priv, ICSIER, SAR | SSR);
+		rcar_i2c_write(priv, ICSSR, ~SSR & 0xff);
+	}
+
+	/* master wants to write to us */
+	if (ssr_filtered & SDR) {
+		int ret;
+
+		value = rcar_i2c_read(priv, ICRXTX);
+		ret = i2c_slave_event(priv->slave, I2C_SLAVE_REQ_WRITE_END, &value);
+		/* Send NACK in case of error */
+		rcar_i2c_write(priv, ICSCR, SIE | SDBS | (ret < 0 ? FNA : 0));
+		i2c_slave_event(priv->slave, I2C_SLAVE_REQ_WRITE_START, &value);
+		rcar_i2c_write(priv, ICSSR, ~SDR & 0xff);
+	}
+
+	/* master wants to read from us */
+	if (ssr_filtered & SDE) {
+		i2c_slave_event(priv->slave, I2C_SLAVE_REQ_READ_END, &value);
+		i2c_slave_event(priv->slave, I2C_SLAVE_REQ_READ_START, &value);
+		rcar_i2c_write(priv, ICRXTX, value);
+		rcar_i2c_write(priv, ICSSR, ~SDE & 0xff);
+	}
+
+	return true;
+}
+
 static irqreturn_t rcar_i2c_irq(int irq, void *ptr)
 {
 	struct rcar_i2c_priv *priv = ptr;
+	irqreturn_t result = IRQ_HANDLED;
 	u32 msr;
 
 	/*-------------- spin lock -----------------*/
 	spin_lock(&priv->lock);
 
+	if (rcar_i2c_slave_irq(priv))
+		goto exit;
+
 	msr = rcar_i2c_read(priv, ICMSR);
 
 	/* Only handle interrupts that are currently enabled */
 	msr &= rcar_i2c_read(priv, ICMIER);
+	if (!msr) {
+		result = IRQ_NONE;
+		goto exit;
+	}
 
 	/* Arbitration lost */
 	if (msr & MAL) {
@@ -408,10 +478,11 @@ out:
 		wake_up(&priv->wait);
 	}
 
+exit:
 	spin_unlock(&priv->lock);
 	/*-------------- spin unlock -----------------*/
 
-	return IRQ_HANDLED;
+	return result;
 }
 
 static int rcar_i2c_master_xfer(struct i2c_adapter *adap,
@@ -453,17 +524,14 @@ static int rcar_i2c_master_xfer(struct i2c_adapter *adap,
 		priv->msg	= &msgs[i];
 		priv->pos	= 0;
 		priv->flags	= 0;
-		if (priv->msg == &msgs[num - 1])
+		if (i == num - 1)
 			rcar_i2c_flags_set(priv, ID_LAST_MSG);
 
-		ret = rcar_i2c_prepare_msg(priv);
+		rcar_i2c_prepare_msg(priv);
 
 		spin_unlock_irqrestore(&priv->lock, flags);
 		/*-------------- spin unlock -----------------*/
 
-		if (ret < 0)
-			break;
-
 		timeout = wait_event_timeout(priv->wait,
 					     rcar_i2c_flags_has(priv, ID_DONE),
 					     5 * HZ);
@@ -498,6 +566,43 @@ out:
 	return ret;
 }
 
+static int rcar_reg_slave(struct i2c_client *slave)
+{
+	struct rcar_i2c_priv *priv = i2c_get_adapdata(slave->adapter);
+
+	if (priv->slave)
+		return -EBUSY;
+
+	if (slave->flags & I2C_CLIENT_TEN)
+		return -EAFNOSUPPORT;
+
+	pm_runtime_forbid(rcar_i2c_priv_to_dev(priv));
+
+	priv->slave = slave;
+	rcar_i2c_write(priv, ICSAR, slave->addr);
+	rcar_i2c_write(priv, ICSSR, 0);
+	rcar_i2c_write(priv, ICSIER, SAR | SSR);
+	rcar_i2c_write(priv, ICSCR, SIE | SDBS);
+
+	return 0;
+}
+
+static int rcar_unreg_slave(struct i2c_client *slave)
+{
+	struct rcar_i2c_priv *priv = i2c_get_adapdata(slave->adapter);
+
+	WARN_ON(!priv->slave);
+
+	rcar_i2c_write(priv, ICSIER, 0);
+	rcar_i2c_write(priv, ICSCR, 0);
+
+	priv->slave = NULL;
+
+	pm_runtime_allow(rcar_i2c_priv_to_dev(priv));
+
+	return 0;
+}
+
 static u32 rcar_i2c_func(struct i2c_adapter *adap)
 {
 	/* This HW can't do SMBUS_QUICK and NOSTART */
@@ -507,6 +612,8 @@ static u32 rcar_i2c_func(struct i2c_adapter *adap)
 static const struct i2c_algorithm rcar_i2c_algo = {
 	.master_xfer	= rcar_i2c_master_xfer,
 	.functionality	= rcar_i2c_func,
+	.reg_slave	= rcar_reg_slave,
+	.unreg_slave	= rcar_unreg_slave,
 };
 
 static const struct of_device_id rcar_i2c_dt_ids[] = {
@@ -619,7 +726,6 @@ MODULE_DEVICE_TABLE(platform, rcar_i2c_id_table);
 static struct platform_driver rcar_i2c_driver = {
 	.driver	= {
 		.name	= "i2c-rcar",
-		.owner	= THIS_MODULE,
 		.of_match_table = rcar_i2c_dt_ids,
 	},
 	.probe		= rcar_i2c_probe,
diff --git a/drivers/i2c/busses/i2c-riic.c b/drivers/i2c/busses/i2c-riic.c
index af3b3d032a9f..d7e3af671543 100644
--- a/drivers/i2c/busses/i2c-riic.c
+++ b/drivers/i2c/busses/i2c-riic.c
@@ -414,7 +414,6 @@ static struct platform_driver riic_i2c_driver = {
 	.remove		= riic_i2c_remove,
 	.driver		= {
 		.name	= "i2c-riic",
-		.owner	= THIS_MODULE,
 		.of_match_table = riic_i2c_dt_ids,
 	},
 };
diff --git a/drivers/i2c/busses/i2c-rk3x.c b/drivers/i2c/busses/i2c-rk3x.c
index b38b0529946a..92462843db66 100644
--- a/drivers/i2c/busses/i2c-rk3x.c
+++ b/drivers/i2c/busses/i2c-rk3x.c
@@ -24,6 +24,7 @@
 #include <linux/wait.h>
 #include <linux/mfd/syscon.h>
 #include <linux/regmap.h>
+#include <linux/math64.h>
 
 
 /* Register Map */
@@ -97,6 +98,7 @@ struct rk3x_i2c {
 	/* Hardware resources */
 	void __iomem *regs;
 	struct clk *clk;
+	struct notifier_block clk_rate_nb;
 
 	/* Settings */
 	unsigned int scl_frequency;
@@ -208,7 +210,7 @@ static void rk3x_i2c_prepare_read(struct rk3x_i2c *i2c)
 	 * The hw can read up to 32 bytes at a time. If we need more than one
 	 * chunk, send an ACK after the last byte of the current chunk.
 	 */
-	if (unlikely(len > 32)) {
+	if (len > 32) {
 		len = 32;
 		con &= ~REG_CON_LASTACK;
 	} else {
@@ -403,7 +405,7 @@ static irqreturn_t rk3x_i2c_irq(int irqno, void *dev_id)
 	}
 
 	/* is there anything left to handle? */
-	if (unlikely((ipd & REG_INT_ALL) == 0))
+	if ((ipd & REG_INT_ALL) == 0)
 		goto out;
 
 	switch (i2c->state) {
@@ -428,18 +430,231 @@ out:
 	return IRQ_HANDLED;
 }
 
-static void rk3x_i2c_set_scl_rate(struct rk3x_i2c *i2c, unsigned long scl_rate)
+/**
+ * Calculate divider values for desired SCL frequency
+ *
+ * @clk_rate: I2C input clock rate
+ * @scl_rate: Desired SCL rate
+ * @div_low: Divider output for low
+ * @div_high: Divider output for high
+ *
+ * Returns: 0 on success, -EINVAL if the goal SCL rate is too slow. In that case
+ * a best-effort divider value is returned in divs. If the target rate is
+ * too high, we silently use the highest possible rate.
+ */
+static int rk3x_i2c_calc_divs(unsigned long clk_rate, unsigned long scl_rate,
+			      unsigned long *div_low, unsigned long *div_high)
 {
-	unsigned long i2c_rate = clk_get_rate(i2c->clk);
-	unsigned int div;
+	unsigned long min_low_ns, min_high_ns;
+	unsigned long max_data_hold_ns;
+	unsigned long data_hold_buffer_ns;
+	unsigned long max_low_ns, min_total_ns;
+
+	unsigned long clk_rate_khz, scl_rate_khz;
+
+	unsigned long min_low_div, min_high_div;
+	unsigned long max_low_div;
+
+	unsigned long min_div_for_hold, min_total_div;
+	unsigned long extra_div, extra_low_div, ideal_low_div;
+
+	int ret = 0;
+
+	/* Only support standard-mode and fast-mode */
+	if (WARN_ON(scl_rate > 400000))
+		scl_rate = 400000;
+
+	/* prevent scl_rate_khz from becoming 0 */
+	if (WARN_ON(scl_rate < 1000))
+		scl_rate = 1000;
+
+	/*
+	 * min_low_ns:  The minimum number of ns we need to hold low
+	 *		to meet i2c spec
+	 * min_high_ns: The minimum number of ns we need to hold high
+	 *		to meet i2c spec
+	 * max_low_ns:  The maximum number of ns we can hold low
+	 *		to meet i2c spec
+	 *
+	 * Note: max_low_ns should be (max data hold time * 2 - buffer)
+	 *	 This is because the i2c host on Rockchip holds the data line
+	 *	 for half the low time.
+	 */
+	if (scl_rate <= 100000) {
+		min_low_ns = 4700;
+		min_high_ns = 4000;
+		max_data_hold_ns = 3450;
+		data_hold_buffer_ns = 50;
+	} else {
+		min_low_ns = 1300;
+		min_high_ns = 600;
+		max_data_hold_ns = 900;
+		data_hold_buffer_ns = 50;
+	}
+	max_low_ns = max_data_hold_ns * 2 - data_hold_buffer_ns;
+	min_total_ns = min_low_ns + min_high_ns;
+
+	/* Adjust to avoid overflow */
+	clk_rate_khz = DIV_ROUND_UP(clk_rate, 1000);
+	scl_rate_khz = scl_rate / 1000;
+
+	/*
+	 * We need the total div to be >= this number
+	 * so we don't clock too fast.
+	 */
+	min_total_div = DIV_ROUND_UP(clk_rate_khz, scl_rate_khz * 8);
+
+	/* These are the min dividers needed for min hold times. */
+	min_low_div = DIV_ROUND_UP(clk_rate_khz * min_low_ns, 8 * 1000000);
+	min_high_div = DIV_ROUND_UP(clk_rate_khz * min_high_ns, 8 * 1000000);
+	min_div_for_hold = (min_low_div + min_high_div);
+
+	/*
+	 * This is the maximum divider so we don't go over the max.
+	 * We don't round up here (we round down) since this is a max.
+	 */
+	max_low_div = clk_rate_khz * max_low_ns / (8 * 1000000);
+
+	if (min_low_div > max_low_div) {
+		WARN_ONCE(true,
+			  "Conflicting, min_low_div %lu, max_low_div %lu\n",
+			  min_low_div, max_low_div);
+		max_low_div = min_low_div;
+	}
+
+	if (min_div_for_hold > min_total_div) {
+		/*
+		 * Time needed to meet hold requirements is important.
+		 * Just use that.
+		 */
+		*div_low = min_low_div;
+		*div_high = min_high_div;
+	} else {
+		/*
+		 * We've got to distribute some time among the low and high
+		 * so we don't run too fast.
+		 */
+		extra_div = min_total_div - min_div_for_hold;
+
+		/*
+		 * We'll try to split things up perfectly evenly,
+		 * biasing slightly towards having a higher div
+		 * for low (spend more time low).
+		 */
+		ideal_low_div = DIV_ROUND_UP(clk_rate_khz * min_low_ns,
+					     scl_rate_khz * 8 * min_total_ns);
+
+		/* Don't allow it to go over the max */
+		if (ideal_low_div > max_low_div)
+			ideal_low_div = max_low_div;
+
+		/*
+		 * Handle when the ideal low div is going to take up
+		 * more than we have.
+		 */
+		if (ideal_low_div > min_low_div + extra_div)
+			ideal_low_div = min_low_div + extra_div;
 
-	/* set DIV = DIVH = DIVL
-	 * SCL rate = (clk rate) / (8 * (DIVH + 1 + DIVL + 1))
-	 *          = (clk rate) / (16 * (DIV + 1))
+		/* Give low the "ideal" and give high whatever extra is left */
+		extra_low_div = ideal_low_div - min_low_div;
+		*div_low = ideal_low_div;
+		*div_high = min_high_div + (extra_div - extra_low_div);
+	}
+
+	/*
+	 * Adjust to the fact that the hardware has an implicit "+1".
+	 * NOTE: Above calculations always produce div_low > 0 and div_high > 0.
 	 */
-	div = DIV_ROUND_UP(i2c_rate, scl_rate * 16) - 1;
+	*div_low = *div_low - 1;
+	*div_high = *div_high - 1;
+
+	/* Maximum divider supported by hw is 0xffff */
+	if (*div_low > 0xffff) {
+		*div_low = 0xffff;
+		ret = -EINVAL;
+	}
+
+	if (*div_high > 0xffff) {
+		*div_high = 0xffff;
+		ret = -EINVAL;
+	}
+
+	return ret;
+}
+
+static void rk3x_i2c_adapt_div(struct rk3x_i2c *i2c, unsigned long clk_rate)
+{
+	unsigned long div_low, div_high;
+	u64 t_low_ns, t_high_ns;
+	int ret;
+
+	ret = rk3x_i2c_calc_divs(clk_rate, i2c->scl_frequency, &div_low,
+				 &div_high);
+
+	WARN_ONCE(ret != 0, "Could not reach SCL freq %u", i2c->scl_frequency);
+
+	clk_enable(i2c->clk);
+	i2c_writel(i2c, (div_high << 16) | (div_low & 0xffff), REG_CLKDIV);
+	clk_disable(i2c->clk);
 
-	i2c_writel(i2c, (div << 16) | (div & 0xffff), REG_CLKDIV);
+	t_low_ns = div_u64(((u64)div_low + 1) * 8 * 1000000000, clk_rate);
+	t_high_ns = div_u64(((u64)div_high + 1) * 8 * 1000000000, clk_rate);
+	dev_dbg(i2c->dev,
+		"CLK %lukhz, Req %uns, Act low %lluns high %lluns\n",
+		clk_rate / 1000,
+		1000000000 / i2c->scl_frequency,
+		t_low_ns, t_high_ns);
+}
+
+/**
+ * rk3x_i2c_clk_notifier_cb - Clock rate change callback
+ * @nb:		Pointer to notifier block
+ * @event:	Notification reason
+ * @data:	Pointer to notification data object
+ *
+ * The callback checks whether a valid bus frequency can be generated after the
+ * change. If so, the change is acknowledged, otherwise the change is aborted.
+ * New dividers are written to the HW in the pre- or post change notification
+ * depending on the scaling direction.
+ *
+ * Code adapted from i2c-cadence.c.
+ *
+ * Return:	NOTIFY_STOP if the rate change should be aborted, NOTIFY_OK
+ *		to acknowedge the change, NOTIFY_DONE if the notification is
+ *		considered irrelevant.
+ */
+static int rk3x_i2c_clk_notifier_cb(struct notifier_block *nb, unsigned long
+				    event, void *data)
+{
+	struct clk_notifier_data *ndata = data;
+	struct rk3x_i2c *i2c = container_of(nb, struct rk3x_i2c, clk_rate_nb);
+	unsigned long div_low, div_high;
+
+	switch (event) {
+	case PRE_RATE_CHANGE:
+		if (rk3x_i2c_calc_divs(ndata->new_rate, i2c->scl_frequency,
+				      &div_low, &div_high) != 0) {
+			return NOTIFY_STOP;
+		}
+
+		/* scale up */
+		if (ndata->new_rate > ndata->old_rate)
+			rk3x_i2c_adapt_div(i2c, ndata->new_rate);
+
+		return NOTIFY_OK;
+	case POST_RATE_CHANGE:
+		/* scale down */
+		if (ndata->new_rate < ndata->old_rate)
+			rk3x_i2c_adapt_div(i2c, ndata->new_rate);
+		return NOTIFY_OK;
+	case ABORT_RATE_CHANGE:
+		/* scale up */
+		if (ndata->new_rate > ndata->old_rate)
+			rk3x_i2c_adapt_div(i2c, ndata->old_rate);
+		return NOTIFY_OK;
+	default:
+		return NOTIFY_DONE;
+	}
 }
 
 /**
@@ -536,9 +751,6 @@ static int rk3x_i2c_xfer(struct i2c_adapter *adap,
 
 	clk_enable(i2c->clk);
 
-	/* The clock rate might have changed, so setup the divider again */
-	rk3x_i2c_set_scl_rate(i2c, i2c->scl_frequency);
-
 	i2c->is_last_msg = false;
 
 	/*
@@ -624,6 +836,7 @@ static int rk3x_i2c_probe(struct platform_device *pdev)
 	int bus_nr;
 	u32 value;
 	int irq;
+	unsigned long clk_rate;
 
 	i2c = devm_kzalloc(&pdev->dev, sizeof(struct rk3x_i2c), GFP_KERNEL);
 	if (!i2c)
@@ -724,16 +937,28 @@ static int rk3x_i2c_probe(struct platform_device *pdev)
 		return ret;
 	}
 
+	i2c->clk_rate_nb.notifier_call = rk3x_i2c_clk_notifier_cb;
+	ret = clk_notifier_register(i2c->clk, &i2c->clk_rate_nb);
+	if (ret != 0) {
+		dev_err(&pdev->dev, "Unable to register clock notifier\n");
+		goto err_clk;
+	}
+
+	clk_rate = clk_get_rate(i2c->clk);
+	rk3x_i2c_adapt_div(i2c, clk_rate);
+
 	ret = i2c_add_adapter(&i2c->adap);
 	if (ret < 0) {
 		dev_err(&pdev->dev, "Could not register adapter\n");
-		goto err_clk;
+		goto err_clk_notifier;
 	}
 
 	dev_info(&pdev->dev, "Initialized RK3xxx I2C bus at %p\n", i2c->regs);
 
 	return 0;
 
+err_clk_notifier:
+	clk_notifier_unregister(i2c->clk, &i2c->clk_rate_nb);
 err_clk:
 	clk_unprepare(i2c->clk);
 	return ret;
@@ -744,6 +969,8 @@ static int rk3x_i2c_remove(struct platform_device *pdev)
 	struct rk3x_i2c *i2c = platform_get_drvdata(pdev);
 
 	i2c_del_adapter(&i2c->adap);
+
+	clk_notifier_unregister(i2c->clk, &i2c->clk_rate_nb);
 	clk_unprepare(i2c->clk);
 
 	return 0;
@@ -753,7 +980,6 @@ static struct platform_driver rk3x_i2c_driver = {
 	.probe   = rk3x_i2c_probe,
 	.remove  = rk3x_i2c_remove,
 	.driver  = {
-		.owner = THIS_MODULE,
 		.name  = "rk3x-i2c",
 		.of_match_table = rk3x_i2c_match,
 	},
diff --git a/drivers/i2c/busses/i2c-s3c2410.c b/drivers/i2c/busses/i2c-s3c2410.c
index e086fb075f2b..958c8db4ec30 100644
--- a/drivers/i2c/busses/i2c-s3c2410.c
+++ b/drivers/i2c/busses/i2c-s3c2410.c
@@ -14,10 +14,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 */
 
 #include <linux/kernel.h>
@@ -39,6 +35,8 @@
 #include <linux/of.h>
 #include <linux/of_gpio.h>
 #include <linux/pinctrl/consumer.h>
+#include <linux/mfd/syscon.h>
+#include <linux/regmap.h>
 
 #include <asm/irq.h>
 
@@ -91,6 +89,9 @@
 /* Max time to wait for bus to become idle after a xfer (in us) */
 #define S3C2410_IDLE_TIMEOUT	5000
 
+/* Exynos5 Sysreg offset */
+#define EXYNOS5_SYS_I2C_CFG	0x0234
+
 /* i2c controller state */
 enum s3c24xx_i2c_state {
 	STATE_IDLE,
@@ -127,6 +128,8 @@ struct s3c24xx_i2c {
 #if defined(CONFIG_ARM_S3C24XX_CPUFREQ)
 	struct notifier_block	freq_transition;
 #endif
+	struct regmap		*sysreg;
+	unsigned int		sys_i2c_cfg;
 };
 
 static struct platform_device_id s3c24xx_driver_ids[] = {
@@ -244,7 +247,7 @@ static bool is_ack(struct s3c24xx_i2c *i2c)
 		}
 		usleep_range(1000, 2000);
 	}
-	dev_err(i2c->dev, "ack was not recieved\n");
+	dev_err(i2c->dev, "ack was not received\n");
 	return false;
 }
 
@@ -782,14 +785,16 @@ static int s3c24xx_i2c_xfer(struct i2c_adapter *adap,
 	int ret;
 
 	pm_runtime_get_sync(&adap->dev);
-	clk_prepare_enable(i2c->clk);
+	ret = clk_enable(i2c->clk);
+	if (ret)
+		return ret;
 
 	for (retry = 0; retry < adap->retries; retry++) {
 
 		ret = s3c24xx_i2c_doxfer(i2c, msgs, num);
 
 		if (ret != -EAGAIN) {
-			clk_disable_unprepare(i2c->clk);
+			clk_disable(i2c->clk);
 			pm_runtime_put(&adap->dev);
 			return ret;
 		}
@@ -799,7 +804,7 @@ static int s3c24xx_i2c_xfer(struct i2c_adapter *adap,
 		udelay(100);
 	}
 
-	clk_disable_unprepare(i2c->clk);
+	clk_disable(i2c->clk);
 	pm_runtime_put(&adap->dev);
 	return -EREMOTEIO;
 }
@@ -1075,6 +1080,7 @@ static void
 s3c24xx_i2c_parse_dt(struct device_node *np, struct s3c24xx_i2c *i2c)
 {
 	struct s3c2410_platform_i2c *pdata = i2c->pdata;
+	int id;
 
 	if (!np)
 		return;
@@ -1084,6 +1090,21 @@ s3c24xx_i2c_parse_dt(struct device_node *np, struct s3c24xx_i2c *i2c)
 	of_property_read_u32(np, "samsung,i2c-slave-addr", &pdata->slave_addr);
 	of_property_read_u32(np, "samsung,i2c-max-bus-freq",
 				(u32 *)&pdata->frequency);
+	/*
+	 * Exynos5's legacy i2c controller and new high speed i2c
+	 * controller have muxed interrupt sources. By default the
+	 * interrupts for 4-channel HS-I2C controller are enabled.
+	 * If nodes for first four channels of legacy i2c controller
+	 * are available then re-configure the interrupts via the
+	 * system register.
+	 */
+	id = of_alias_get_id(np, "i2c");
+	i2c->sysreg = syscon_regmap_lookup_by_phandle(np,
+			"samsung,sysreg-phandle");
+	if (IS_ERR(i2c->sysreg))
+		return;
+
+	regmap_update_bits(i2c->sysreg, EXYNOS5_SYS_I2C_CFG, BIT(id), 0);
 }
 #else
 static void
@@ -1178,7 +1199,7 @@ static int s3c24xx_i2c_probe(struct platform_device *pdev)
 
 	clk_prepare_enable(i2c->clk);
 	ret = s3c24xx_i2c_init(i2c);
-	clk_disable_unprepare(i2c->clk);
+	clk_disable(i2c->clk);
 	if (ret != 0) {
 		dev_err(&pdev->dev, "I2C controller init failed\n");
 		return ret;
@@ -1191,6 +1212,7 @@ static int s3c24xx_i2c_probe(struct platform_device *pdev)
 		i2c->irq = ret = platform_get_irq(pdev, 0);
 		if (ret <= 0) {
 			dev_err(&pdev->dev, "cannot find IRQ\n");
+			clk_unprepare(i2c->clk);
 			return ret;
 		}
 
@@ -1199,6 +1221,7 @@ static int s3c24xx_i2c_probe(struct platform_device *pdev)
 
 		if (ret != 0) {
 			dev_err(&pdev->dev, "cannot claim IRQ %d\n", i2c->irq);
+			clk_unprepare(i2c->clk);
 			return ret;
 		}
 	}
@@ -1206,6 +1229,7 @@ static int s3c24xx_i2c_probe(struct platform_device *pdev)
 	ret = s3c24xx_i2c_register_cpufreq(i2c);
 	if (ret < 0) {
 		dev_err(&pdev->dev, "failed to register cpufreq notifier\n");
+		clk_unprepare(i2c->clk);
 		return ret;
 	}
 
@@ -1222,6 +1246,7 @@ static int s3c24xx_i2c_probe(struct platform_device *pdev)
 	if (ret < 0) {
 		dev_err(&pdev->dev, "failed to add bus to i2c core\n");
 		s3c24xx_i2c_deregister_cpufreq(i2c);
+		clk_unprepare(i2c->clk);
 		return ret;
 	}
 
@@ -1243,6 +1268,8 @@ static int s3c24xx_i2c_remove(struct platform_device *pdev)
 {
 	struct s3c24xx_i2c *i2c = platform_get_drvdata(pdev);
 
+	clk_unprepare(i2c->clk);
+
 	pm_runtime_disable(&i2c->adap.dev);
 	pm_runtime_disable(&pdev->dev);
 
@@ -1264,6 +1291,9 @@ static int s3c24xx_i2c_suspend_noirq(struct device *dev)
 
 	i2c->suspended = 1;
 
+	if (!IS_ERR(i2c->sysreg))
+		regmap_read(i2c->sysreg, EXYNOS5_SYS_I2C_CFG, &i2c->sys_i2c_cfg);
+
 	return 0;
 }
 
@@ -1271,10 +1301,16 @@ static int s3c24xx_i2c_resume_noirq(struct device *dev)
 {
 	struct platform_device *pdev = to_platform_device(dev);
 	struct s3c24xx_i2c *i2c = platform_get_drvdata(pdev);
+	int ret;
 
-	clk_prepare_enable(i2c->clk);
+	if (!IS_ERR(i2c->sysreg))
+		regmap_write(i2c->sysreg, EXYNOS5_SYS_I2C_CFG, i2c->sys_i2c_cfg);
+
+	ret = clk_enable(i2c->clk);
+	if (ret)
+		return ret;
 	s3c24xx_i2c_init(i2c);
-	clk_disable_unprepare(i2c->clk);
+	clk_disable(i2c->clk);
 	i2c->suspended = 0;
 
 	return 0;
@@ -1305,7 +1341,6 @@ static struct platform_driver s3c24xx_i2c_driver = {
 	.remove		= s3c24xx_i2c_remove,
 	.id_table	= s3c24xx_driver_ids,
 	.driver		= {
-		.owner	= THIS_MODULE,
 		.name	= "s3c-i2c",
 		.pm	= S3C24XX_DEV_PM_OPS,
 		.of_match_table = of_match_ptr(s3c24xx_i2c_match),
diff --git a/drivers/i2c/busses/i2c-sh7760.c b/drivers/i2c/busses/i2c-sh7760.c
index d76f3d9737ec..24968384b401 100644
--- a/drivers/i2c/busses/i2c-sh7760.c
+++ b/drivers/i2c/busses/i2c-sh7760.c
@@ -552,7 +552,6 @@ static int sh7760_i2c_remove(struct platform_device *pdev)
 static struct platform_driver sh7760_i2c_drv = {
 	.driver	= {
 		.name	= SH7760_I2C_DEVNAME,
-		.owner	= THIS_MODULE,
 	},
 	.probe		= sh7760_i2c_probe,
 	.remove		= sh7760_i2c_remove,
diff --git a/drivers/i2c/busses/i2c-sh_mobile.c b/drivers/i2c/busses/i2c-sh_mobile.c
index 8b5e79cb4468..007818b3e174 100644
--- a/drivers/i2c/busses/i2c-sh_mobile.c
+++ b/drivers/i2c/busses/i2c-sh_mobile.c
@@ -1,6 +1,8 @@
 /*
  * SuperH Mobile I2C Controller
  *
+ * Copyright (C) 2014 Wolfram Sang <wsa@sang-engineering.com>
+ *
  * Copyright (C) 2008 Magnus Damm
  *
  * Portions of the code based on out-of-tree driver i2c-sh7343.c
@@ -14,26 +16,24 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  */
 
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+#include <linux/err.h>
+#include <linux/i2c.h>
+#include <linux/i2c/i2c-sh_mobile.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
 #include <linux/kernel.h>
 #include <linux/module.h>
-#include <linux/init.h>
-#include <linux/delay.h>
+#include <linux/of_device.h>
 #include <linux/platform_device.h>
-#include <linux/interrupt.h>
-#include <linux/i2c.h>
-#include <linux/err.h>
 #include <linux/pm_runtime.h>
-#include <linux/clk.h>
-#include <linux/io.h>
 #include <linux/slab.h>
-#include <linux/of_device.h>
-#include <linux/i2c/i2c-sh_mobile.h>
 
 /* Transmit operation:                                                      */
 /*                                                                          */
@@ -114,6 +114,7 @@ enum sh_mobile_i2c_op {
 	OP_TX_FIRST,
 	OP_TX,
 	OP_TX_STOP,
+	OP_TX_STOP_DATA,
 	OP_TX_TO_RX,
 	OP_RX,
 	OP_RX_STOP,
@@ -138,6 +139,13 @@ struct sh_mobile_i2c_data {
 	int pos;
 	int sr;
 	bool send_stop;
+	bool stop_after_dma;
+
+	struct resource *res;
+	struct dma_chan *dma_tx;
+	struct dma_chan *dma_rx;
+	struct scatterlist sg;
+	enum dma_data_direction dma_direction;
 };
 
 struct sh_mobile_dt_config {
@@ -175,6 +183,8 @@ struct sh_mobile_dt_config {
 
 #define ICIC_ICCLB8		0x80
 #define ICIC_ICCHB8		0x40
+#define ICIC_TDMAE		0x20
+#define ICIC_RDMAE		0x10
 #define ICIC_ALE		0x08
 #define ICIC_TACKE		0x04
 #define ICIC_WAITE		0x02
@@ -281,6 +291,7 @@ static int sh_mobile_i2c_init(struct sh_mobile_i2c_data *pd)
 	else
 		pd->icic &= ~ICIC_ICCHB8;
 
+	dev_dbg(pd->dev, "timing values: L/H=0x%x/0x%x\n", pd->iccl, pd->icch);
 	return 0;
 }
 
@@ -336,8 +347,10 @@ static unsigned char i2c_op(struct sh_mobile_i2c_data *pd,
 	case OP_TX: /* write data */
 		iic_wr(pd, ICDR, data);
 		break;
-	case OP_TX_STOP: /* write data and issue a stop afterwards */
+	case OP_TX_STOP_DATA: /* write data and issue a stop afterwards */
 		iic_wr(pd, ICDR, data);
+		/* fallthrough */
+	case OP_TX_STOP: /* issue a stop */
 		iic_wr(pd, ICCR, pd->send_stop ? ICCR_ICE | ICCR_TRS
 					       : ICCR_ICE | ICCR_TRS | ICCR_BBSY);
 		break;
@@ -393,13 +406,17 @@ static int sh_mobile_i2c_isr_tx(struct sh_mobile_i2c_data *pd)
 {
 	unsigned char data;
 
-	if (pd->pos == pd->msg->len)
+	if (pd->pos == pd->msg->len) {
+		/* Send stop if we haven't yet (DMA case) */
+		if (pd->send_stop && pd->stop_after_dma)
+			i2c_op(pd, OP_TX_STOP, 0);
 		return 1;
+	}
 
 	sh_mobile_i2c_get_data(pd, &data);
 
 	if (sh_mobile_i2c_is_last_byte(pd))
-		i2c_op(pd, OP_TX_STOP, data);
+		i2c_op(pd, OP_TX_STOP_DATA, data);
 	else if (sh_mobile_i2c_is_first_byte(pd))
 		i2c_op(pd, OP_TX_FIRST, data);
 	else
@@ -433,6 +450,13 @@ static int sh_mobile_i2c_isr_rx(struct sh_mobile_i2c_data *pd)
 		real_pos = pd->pos - 2;
 
 		if (pd->pos == pd->msg->len) {
+			if (pd->stop_after_dma) {
+				/* Simulate PIO end condition after DMA transfer */
+				i2c_op(pd, OP_RX_STOP, 0);
+				pd->pos++;
+				break;
+			}
+
 			if (real_pos < 0) {
 				i2c_op(pd, OP_RX_STOP, 0);
 				break;
@@ -451,10 +475,9 @@ static int sh_mobile_i2c_isr_rx(struct sh_mobile_i2c_data *pd)
 
 static irqreturn_t sh_mobile_i2c_isr(int irq, void *dev_id)
 {
-	struct platform_device *dev = dev_id;
-	struct sh_mobile_i2c_data *pd = platform_get_drvdata(dev);
+	struct sh_mobile_i2c_data *pd = dev_id;
 	unsigned char sr;
-	int wakeup;
+	int wakeup = 0;
 
 	sr = iic_rd(pd, ICSR);
 	pd->sr |= sr; /* remember state */
@@ -463,15 +486,21 @@ static irqreturn_t sh_mobile_i2c_isr(int irq, void *dev_id)
 	       (pd->msg->flags & I2C_M_RD) ? "read" : "write",
 	       pd->pos, pd->msg->len);
 
-	if (sr & (ICSR_AL | ICSR_TACK)) {
+	/* Kick off TxDMA after preface was done */
+	if (pd->dma_direction == DMA_TO_DEVICE && pd->pos == 0)
+		iic_set_clr(pd, ICIC, ICIC_TDMAE, 0);
+	else if (sr & (ICSR_AL | ICSR_TACK))
 		/* don't interrupt transaction - continue to issue stop */
 		iic_wr(pd, ICSR, sr & ~(ICSR_AL | ICSR_TACK));
-		wakeup = 0;
-	} else if (pd->msg->flags & I2C_M_RD)
+	else if (pd->msg->flags & I2C_M_RD)
 		wakeup = sh_mobile_i2c_isr_rx(pd);
 	else
 		wakeup = sh_mobile_i2c_isr_tx(pd);
 
+	/* Kick off RxDMA after preface was done */
+	if (pd->dma_direction == DMA_FROM_DEVICE && pd->pos == 1)
+		iic_set_clr(pd, ICIC, ICIC_RDMAE, 0);
+
 	if (sr & ICSR_WAIT) /* TODO: add delay here to support slow acks */
 		iic_wr(pd, ICSR, sr & ~ICSR_WAIT);
 
@@ -486,6 +515,130 @@ static irqreturn_t sh_mobile_i2c_isr(int irq, void *dev_id)
 	return IRQ_HANDLED;
 }
 
+static void sh_mobile_i2c_dma_unmap(struct sh_mobile_i2c_data *pd)
+{
+	struct dma_chan *chan = pd->dma_direction == DMA_FROM_DEVICE
+				? pd->dma_rx : pd->dma_tx;
+
+	dma_unmap_single(chan->device->dev, sg_dma_address(&pd->sg),
+			 pd->msg->len, pd->dma_direction);
+
+	pd->dma_direction = DMA_NONE;
+}
+
+static void sh_mobile_i2c_cleanup_dma(struct sh_mobile_i2c_data *pd)
+{
+	if (pd->dma_direction == DMA_NONE)
+		return;
+	else if (pd->dma_direction == DMA_FROM_DEVICE)
+		dmaengine_terminate_all(pd->dma_rx);
+	else if (pd->dma_direction == DMA_TO_DEVICE)
+		dmaengine_terminate_all(pd->dma_tx);
+
+	sh_mobile_i2c_dma_unmap(pd);
+}
+
+static void sh_mobile_i2c_dma_callback(void *data)
+{
+	struct sh_mobile_i2c_data *pd = data;
+
+	sh_mobile_i2c_dma_unmap(pd);
+	pd->pos = pd->msg->len;
+	pd->stop_after_dma = true;
+
+	iic_set_clr(pd, ICIC, 0, ICIC_TDMAE | ICIC_RDMAE);
+}
+
+static struct dma_chan *sh_mobile_i2c_request_dma_chan(struct device *dev,
+				enum dma_transfer_direction dir, dma_addr_t port_addr)
+{
+	struct dma_chan *chan;
+	struct dma_slave_config cfg;
+	char *chan_name = dir == DMA_MEM_TO_DEV ? "tx" : "rx";
+	int ret;
+
+	chan = dma_request_slave_channel_reason(dev, chan_name);
+	if (IS_ERR(chan)) {
+		ret = PTR_ERR(chan);
+		dev_dbg(dev, "request_channel failed for %s (%d)\n", chan_name, ret);
+		return chan;
+	}
+
+	memset(&cfg, 0, sizeof(cfg));
+	cfg.direction = dir;
+	if (dir == DMA_MEM_TO_DEV) {
+		cfg.dst_addr = port_addr;
+		cfg.dst_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+	} else {
+		cfg.src_addr = port_addr;
+		cfg.src_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+	}
+
+	ret = dmaengine_slave_config(chan, &cfg);
+	if (ret) {
+		dev_dbg(dev, "slave_config failed for %s (%d)\n", chan_name, ret);
+		dma_release_channel(chan);
+		return ERR_PTR(ret);
+	}
+
+	dev_dbg(dev, "got DMA channel for %s\n", chan_name);
+	return chan;
+}
+
+static void sh_mobile_i2c_xfer_dma(struct sh_mobile_i2c_data *pd)
+{
+	bool read = pd->msg->flags & I2C_M_RD;
+	enum dma_data_direction dir = read ? DMA_FROM_DEVICE : DMA_TO_DEVICE;
+	struct dma_chan *chan = read ? pd->dma_rx : pd->dma_tx;
+	struct dma_async_tx_descriptor *txdesc;
+	dma_addr_t dma_addr;
+	dma_cookie_t cookie;
+
+	if (PTR_ERR(chan) == -EPROBE_DEFER) {
+		if (read)
+			chan = pd->dma_rx = sh_mobile_i2c_request_dma_chan(pd->dev, DMA_DEV_TO_MEM,
+									   pd->res->start + ICDR);
+		else
+			chan = pd->dma_tx = sh_mobile_i2c_request_dma_chan(pd->dev, DMA_MEM_TO_DEV,
+									   pd->res->start + ICDR);
+	}
+
+	if (IS_ERR(chan))
+		return;
+
+	dma_addr = dma_map_single(chan->device->dev, pd->msg->buf, pd->msg->len, dir);
+	if (dma_mapping_error(pd->dev, dma_addr)) {
+		dev_dbg(pd->dev, "dma map failed, using PIO\n");
+		return;
+	}
+
+	sg_dma_len(&pd->sg) = pd->msg->len;
+	sg_dma_address(&pd->sg) = dma_addr;
+
+	pd->dma_direction = dir;
+
+	txdesc = dmaengine_prep_slave_sg(chan, &pd->sg, 1,
+					 read ? DMA_DEV_TO_MEM : DMA_MEM_TO_DEV,
+					 DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!txdesc) {
+		dev_dbg(pd->dev, "dma prep slave sg failed, using PIO\n");
+		sh_mobile_i2c_cleanup_dma(pd);
+		return;
+	}
+
+	txdesc->callback = sh_mobile_i2c_dma_callback;
+	txdesc->callback_param = pd;
+
+	cookie = dmaengine_submit(txdesc);
+	if (dma_submit_error(cookie)) {
+		dev_dbg(pd->dev, "submitting dma failed, using PIO\n");
+		sh_mobile_i2c_cleanup_dma(pd);
+		return;
+	}
+
+	dma_async_issue_pending(chan);
+}
+
 static int start_ch(struct sh_mobile_i2c_data *pd, struct i2c_msg *usr_msg,
 		    bool do_init)
 {
@@ -510,6 +663,9 @@ static int start_ch(struct sh_mobile_i2c_data *pd, struct i2c_msg *usr_msg,
 	pd->pos = -1;
 	pd->sr = 0;
 
+	if (pd->msg->len > 8)
+		sh_mobile_i2c_xfer_dma(pd);
+
 	/* Enable all interrupts to begin with */
 	iic_wr(pd, ICIC, ICIC_DTEE | ICIC_WAITE | ICIC_ALE | ICIC_TACKE);
 	return 0;
@@ -579,6 +735,7 @@ static int sh_mobile_i2c_xfer(struct i2c_adapter *adapter,
 		bool do_start = pd->send_stop || !i;
 		msg = &msgs[i];
 		pd->send_stop = i == num - 1 || msg->flags & I2C_M_STOP;
+		pd->stop_after_dma = false;
 
 		err = start_ch(pd, msg, do_start);
 		if (err)
@@ -593,6 +750,9 @@ static int sh_mobile_i2c_xfer(struct i2c_adapter *adapter,
 				       5 * HZ);
 		if (!k) {
 			dev_err(pd->dev, "Transfer request timed out\n");
+			if (pd->dma_direction != DMA_NONE)
+				sh_mobile_i2c_cleanup_dma(pd);
+
 			err = -ETIMEDOUT;
 			break;
 		}
@@ -626,22 +786,37 @@ static const struct sh_mobile_dt_config default_dt_config = {
 	.clks_per_count = 1,
 };
 
-static const struct sh_mobile_dt_config rcar_gen2_dt_config = {
+static const struct sh_mobile_dt_config fast_clock_dt_config = {
 	.clks_per_count = 2,
 };
 
 static const struct of_device_id sh_mobile_i2c_dt_ids[] = {
 	{ .compatible = "renesas,rmobile-iic", .data = &default_dt_config },
-	{ .compatible = "renesas,iic-r8a7790", .data = &rcar_gen2_dt_config },
-	{ .compatible = "renesas,iic-r8a7791", .data = &rcar_gen2_dt_config },
-	{ .compatible = "renesas,iic-r8a7792", .data = &rcar_gen2_dt_config },
-	{ .compatible = "renesas,iic-r8a7793", .data = &rcar_gen2_dt_config },
-	{ .compatible = "renesas,iic-r8a7794", .data = &rcar_gen2_dt_config },
+	{ .compatible = "renesas,iic-r8a73a4", .data = &fast_clock_dt_config },
+	{ .compatible = "renesas,iic-r8a7790", .data = &fast_clock_dt_config },
+	{ .compatible = "renesas,iic-r8a7791", .data = &fast_clock_dt_config },
+	{ .compatible = "renesas,iic-r8a7792", .data = &fast_clock_dt_config },
+	{ .compatible = "renesas,iic-r8a7793", .data = &fast_clock_dt_config },
+	{ .compatible = "renesas,iic-r8a7794", .data = &fast_clock_dt_config },
+	{ .compatible = "renesas,iic-sh73a0", .data = &fast_clock_dt_config },
 	{},
 };
 MODULE_DEVICE_TABLE(of, sh_mobile_i2c_dt_ids);
 
-static int sh_mobile_i2c_hook_irqs(struct platform_device *dev)
+static void sh_mobile_i2c_release_dma(struct sh_mobile_i2c_data *pd)
+{
+	if (!IS_ERR(pd->dma_tx)) {
+		dma_release_channel(pd->dma_tx);
+		pd->dma_tx = ERR_PTR(-EPROBE_DEFER);
+	}
+
+	if (!IS_ERR(pd->dma_rx)) {
+		dma_release_channel(pd->dma_rx);
+		pd->dma_rx = ERR_PTR(-EPROBE_DEFER);
+	}
+}
+
+static int sh_mobile_i2c_hook_irqs(struct platform_device *dev, struct sh_mobile_i2c_data *pd)
 {
 	struct resource *res;
 	resource_size_t n;
@@ -650,7 +825,7 @@ static int sh_mobile_i2c_hook_irqs(struct platform_device *dev)
 	while ((res = platform_get_resource(dev, IORESOURCE_IRQ, k))) {
 		for (n = res->start; n <= res->end; n++) {
 			ret = devm_request_irq(&dev->dev, n, sh_mobile_i2c_isr,
-					  0, dev_name(&dev->dev), dev);
+					  0, dev_name(&dev->dev), pd);
 			if (ret) {
 				dev_err(&dev->dev, "cannot request IRQ %pa\n", &n);
 				return ret;
@@ -681,7 +856,7 @@ static int sh_mobile_i2c_probe(struct platform_device *dev)
 		return PTR_ERR(pd->clk);
 	}
 
-	ret = sh_mobile_i2c_hook_irqs(dev);
+	ret = sh_mobile_i2c_hook_irqs(dev, pd);
 	if (ret)
 		return ret;
 
@@ -690,6 +865,7 @@ static int sh_mobile_i2c_probe(struct platform_device *dev)
 
 	res = platform_get_resource(dev, IORESOURCE_MEM, 0);
 
+	pd->res = res;
 	pd->reg = devm_ioremap_resource(&dev->dev, res);
 	if (IS_ERR(pd->reg))
 		return PTR_ERR(pd->reg);
@@ -727,6 +903,11 @@ static int sh_mobile_i2c_probe(struct platform_device *dev)
 	if (ret)
 		return ret;
 
+	/* Init DMA */
+	sg_init_table(&pd->sg, 1);
+	pd->dma_direction = DMA_NONE;
+	pd->dma_rx = pd->dma_tx = ERR_PTR(-EPROBE_DEFER);
+
 	/* Enable Runtime PM for this device.
 	 *
 	 * Also tell the Runtime PM core to ignore children
@@ -758,13 +939,12 @@ static int sh_mobile_i2c_probe(struct platform_device *dev)
 
 	ret = i2c_add_numbered_adapter(adap);
 	if (ret < 0) {
+		sh_mobile_i2c_release_dma(pd);
 		dev_err(&dev->dev, "cannot add numbered adapter\n");
 		return ret;
 	}
 
-	dev_info(&dev->dev,
-		 "I2C adapter %d with bus speed %lu Hz (L/H=0x%x/0x%x)\n",
-		 adap->nr, pd->bus_speed, pd->iccl, pd->icch);
+	dev_info(&dev->dev, "I2C adapter %d, bus speed %lu Hz\n", adap->nr, pd->bus_speed);
 
 	return 0;
 }
@@ -774,6 +954,7 @@ static int sh_mobile_i2c_remove(struct platform_device *dev)
 	struct sh_mobile_i2c_data *pd = platform_get_drvdata(dev);
 
 	i2c_del_adapter(&pd->adap);
+	sh_mobile_i2c_release_dma(pd);
 	pm_runtime_disable(&dev->dev);
 	return 0;
 }
@@ -798,7 +979,6 @@ static const struct dev_pm_ops sh_mobile_i2c_dev_pm_ops = {
 static struct platform_driver sh_mobile_i2c_driver = {
 	.driver		= {
 		.name		= "i2c-sh_mobile",
-		.owner		= THIS_MODULE,
 		.pm		= &sh_mobile_i2c_dev_pm_ops,
 		.of_match_table = sh_mobile_i2c_dt_ids,
 	},
@@ -810,16 +990,15 @@ static int __init sh_mobile_i2c_adap_init(void)
 {
 	return platform_driver_register(&sh_mobile_i2c_driver);
 }
+subsys_initcall(sh_mobile_i2c_adap_init);
 
 static void __exit sh_mobile_i2c_adap_exit(void)
 {
 	platform_driver_unregister(&sh_mobile_i2c_driver);
 }
-
-subsys_initcall(sh_mobile_i2c_adap_init);
 module_exit(sh_mobile_i2c_adap_exit);
 
 MODULE_DESCRIPTION("SuperH Mobile I2C Bus Controller driver");
-MODULE_AUTHOR("Magnus Damm");
+MODULE_AUTHOR("Magnus Damm and Wolfram Sang");
 MODULE_LICENSE("GPL v2");
 MODULE_ALIAS("platform:i2c-sh_mobile");
diff --git a/drivers/i2c/busses/i2c-sibyte.c b/drivers/i2c/busses/i2c-sibyte.c
index 0fe505d7abe9..2b6219d86b0f 100644
--- a/drivers/i2c/busses/i2c-sibyte.c
+++ b/drivers/i2c/busses/i2c-sibyte.c
@@ -12,10 +12,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
  */
 
 #include <linux/kernel.h>
diff --git a/drivers/i2c/busses/i2c-simtec.c b/drivers/i2c/busses/i2c-simtec.c
index 964e5c6f84ab..b4685bb9b5d7 100644
--- a/drivers/i2c/busses/i2c-simtec.c
+++ b/drivers/i2c/busses/i2c-simtec.c
@@ -12,10 +12,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 */
 
 #include <linux/kernel.h>
@@ -158,7 +154,6 @@ static int simtec_i2c_remove(struct platform_device *dev)
 static struct platform_driver simtec_i2c_driver = {
 	.driver		= {
 		.name		= "simtec-i2c",
-		.owner		= THIS_MODULE,
 	},
 	.probe		= simtec_i2c_probe,
 	.remove		= simtec_i2c_remove,
diff --git a/drivers/i2c/busses/i2c-sirf.c b/drivers/i2c/busses/i2c-sirf.c
index b1336d5f0531..1092d4eeeb54 100644
--- a/drivers/i2c/busses/i2c-sirf.c
+++ b/drivers/i2c/busses/i2c-sirf.c
@@ -451,7 +451,6 @@ MODULE_DEVICE_TABLE(of, sirfsoc_i2c_of_match);
 static struct platform_driver i2c_sirfsoc_driver = {
 	.driver = {
 		.name = "sirfsoc_i2c",
-		.owner = THIS_MODULE,
 #ifdef CONFIG_PM
 		.pm = &i2c_sirfsoc_pm_ops,
 #endif
diff --git a/drivers/i2c/busses/i2c-sis5595.c b/drivers/i2c/busses/i2c-sis5595.c
index ac9bc33acef4..7d58a40faf2d 100644
--- a/drivers/i2c/busses/i2c-sis5595.c
+++ b/drivers/i2c/busses/i2c-sis5595.c
@@ -11,10 +11,6 @@
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
     GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 /* Note: we assume there can only be one SIS5595 with one SMBus interface */
diff --git a/drivers/i2c/busses/i2c-sis630.c b/drivers/i2c/busses/i2c-sis630.c
index c6366733008d..1e6805b5cef2 100644
--- a/drivers/i2c/busses/i2c-sis630.c
+++ b/drivers/i2c/busses/i2c-sis630.c
@@ -10,10 +10,6 @@
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
     GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 /*
diff --git a/drivers/i2c/busses/i2c-sis96x.c b/drivers/i2c/busses/i2c-sis96x.c
index 8dc2fc5f74ff..44b904426073 100644
--- a/drivers/i2c/busses/i2c-sis96x.c
+++ b/drivers/i2c/busses/i2c-sis96x.c
@@ -10,10 +10,6 @@
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
     GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 /*
diff --git a/drivers/i2c/busses/i2c-st.c b/drivers/i2c/busses/i2c-st.c
index 2e4eccd6599a..88057fad9dfe 100644
--- a/drivers/i2c/busses/i2c-st.c
+++ b/drivers/i2c/busses/i2c-st.c
@@ -863,7 +863,6 @@ MODULE_DEVICE_TABLE(of, st_i2c_match);
 static struct platform_driver st_i2c_driver = {
 	.driver = {
 		.name = "st-i2c",
-		.owner = THIS_MODULE,
 		.of_match_table = st_i2c_match,
 		.pm = ST_I2C_PM,
 	},
diff --git a/drivers/i2c/busses/i2c-stu300.c b/drivers/i2c/busses/i2c-stu300.c
index 6a44f37798c8..4885da9e9298 100644
--- a/drivers/i2c/busses/i2c-stu300.c
+++ b/drivers/i2c/busses/i2c-stu300.c
@@ -981,7 +981,6 @@ static const struct of_device_id stu300_dt_match[] = {
 static struct platform_driver stu300_i2c_driver = {
 	.driver = {
 		.name	= NAME,
-		.owner	= THIS_MODULE,
 		.pm	= STU300_I2C_PM,
 		.of_match_table = stu300_dt_match,
 	},
diff --git a/drivers/i2c/busses/i2c-sun6i-p2wi.c b/drivers/i2c/busses/i2c-sun6i-p2wi.c
index 4d75d4759709..7668e2e9b8fd 100644
--- a/drivers/i2c/busses/i2c-sun6i-p2wi.c
+++ b/drivers/i2c/busses/i2c-sun6i-p2wi.c
@@ -332,7 +332,6 @@ static struct platform_driver p2wi_driver = {
 	.probe	= p2wi_probe,
 	.remove	= p2wi_remove,
 	.driver	= {
-		.owner = THIS_MODULE,
 		.name = "i2c-sunxi-p2wi",
 		.of_match_table = p2wi_of_match_table,
 	},
diff --git a/drivers/i2c/busses/i2c-taos-evm.c b/drivers/i2c/busses/i2c-taos-evm.c
index 10855a0b7e7f..4c7fc2d47014 100644
--- a/drivers/i2c/busses/i2c-taos-evm.c
+++ b/drivers/i2c/busses/i2c-taos-evm.c
@@ -13,10 +13,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
  */
 
 #include <linux/delay.h>
diff --git a/drivers/i2c/busses/i2c-tegra.c b/drivers/i2c/busses/i2c-tegra.c
index efba1ebe16ba..28b87e683503 100644
--- a/drivers/i2c/busses/i2c-tegra.c
+++ b/drivers/i2c/busses/i2c-tegra.c
@@ -893,7 +893,6 @@ static struct platform_driver tegra_i2c_driver = {
 	.remove  = tegra_i2c_remove,
 	.driver  = {
 		.name  = "tegra-i2c",
-		.owner = THIS_MODULE,
 		.of_match_table = tegra_i2c_of_match,
 		.pm    = TEGRA_I2C_PM,
 	},
diff --git a/drivers/i2c/busses/i2c-versatile.c b/drivers/i2c/busses/i2c-versatile.c
index 6bb3a89a440f..240637f01d11 100644
--- a/drivers/i2c/busses/i2c-versatile.c
+++ b/drivers/i2c/busses/i2c-versatile.c
@@ -138,7 +138,6 @@ static struct platform_driver i2c_versatile_driver = {
 	.remove		= i2c_versatile_remove,
 	.driver		= {
 		.name	= "versatile-i2c",
-		.owner	= THIS_MODULE,
 		.of_match_table = i2c_versatile_match,
 	},
 };
diff --git a/drivers/i2c/busses/i2c-via.c b/drivers/i2c/busses/i2c-via.c
index f4a1ed757612..59b1d233ca7b 100644
--- a/drivers/i2c/busses/i2c-via.c
+++ b/drivers/i2c/busses/i2c-via.c
@@ -12,10 +12,6 @@
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
     GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 #include <linux/kernel.h>
diff --git a/drivers/i2c/busses/i2c-viapro.c b/drivers/i2c/busses/i2c-viapro.c
index 6841200b6e50..0ee2646f3b00 100644
--- a/drivers/i2c/busses/i2c-viapro.c
+++ b/drivers/i2c/busses/i2c-viapro.c
@@ -13,10 +13,6 @@
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
     GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 /*
diff --git a/drivers/i2c/busses/i2c-wmt.c b/drivers/i2c/busses/i2c-wmt.c
index f80a38c2072c..82ea34925489 100644
--- a/drivers/i2c/busses/i2c-wmt.c
+++ b/drivers/i2c/busses/i2c-wmt.c
@@ -462,7 +462,6 @@ static struct platform_driver wmt_i2c_driver = {
 	.remove		= wmt_i2c_remove,
 	.driver		= {
 		.name	= "wmt-i2c",
-		.owner	= THIS_MODULE,
 		.of_match_table = wmt_i2c_dt_ids,
 	},
 };
diff --git a/drivers/i2c/busses/i2c-xiic.c b/drivers/i2c/busses/i2c-xiic.c
index ade9223912d3..e8400042b358 100644
--- a/drivers/i2c/busses/i2c-xiic.c
+++ b/drivers/i2c/busses/i2c-xiic.c
@@ -12,10 +12,6 @@
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
  *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
- *
  *
  * This code was implemented by Mocean Laboratories AB when porting linux
  * to the automotive development board Russellville. The copyright holder
@@ -50,6 +46,11 @@ enum xilinx_i2c_state {
 	STATE_START
 };
 
+enum xiic_endian {
+	LITTLE,
+	BIG
+};
+
 /**
  * struct xiic_i2c - Internal representation of the XIIC I2C bus
  * @base:	Memory base of the HW registers
@@ -74,6 +75,7 @@ struct xiic_i2c {
 	enum xilinx_i2c_state	state;
 	struct i2c_msg		*rx_msg;
 	int			rx_pos;
+	enum xiic_endian	endianness;
 };
 
 
@@ -174,29 +176,58 @@ struct xiic_i2c {
 static void xiic_start_xfer(struct xiic_i2c *i2c);
 static void __xiic_start_xfer(struct xiic_i2c *i2c);
 
+/*
+ * For the register read and write functions, a little-endian and big-endian
+ * version are necessary. Endianness is detected during the probe function.
+ * Only the least significant byte [doublet] of the register are ever
+ * accessed. This requires an offset of 3 [2] from the base address for
+ * big-endian systems.
+ */
+
 static inline void xiic_setreg8(struct xiic_i2c *i2c, int reg, u8 value)
 {
-	iowrite8(value, i2c->base + reg);
+	if (i2c->endianness == LITTLE)
+		iowrite8(value, i2c->base + reg);
+	else
+		iowrite8(value, i2c->base + reg + 3);
 }
 
 static inline u8 xiic_getreg8(struct xiic_i2c *i2c, int reg)
 {
-	return ioread8(i2c->base + reg);
+	u8 ret;
+
+	if (i2c->endianness == LITTLE)
+		ret = ioread8(i2c->base + reg);
+	else
+		ret = ioread8(i2c->base + reg + 3);
+	return ret;
 }
 
 static inline void xiic_setreg16(struct xiic_i2c *i2c, int reg, u16 value)
 {
-	iowrite16(value, i2c->base + reg);
+	if (i2c->endianness == LITTLE)
+		iowrite16(value, i2c->base + reg);
+	else
+		iowrite16be(value, i2c->base + reg + 2);
 }
 
 static inline void xiic_setreg32(struct xiic_i2c *i2c, int reg, int value)
 {
-	iowrite32(value, i2c->base + reg);
+	if (i2c->endianness == LITTLE)
+		iowrite32(value, i2c->base + reg);
+	else
+		iowrite32be(value, i2c->base + reg);
 }
 
 static inline int xiic_getreg32(struct xiic_i2c *i2c, int reg)
 {
-	return ioread32(i2c->base + reg);
+	u32 ret;
+
+	if (i2c->endianness == LITTLE)
+		ret = ioread32(i2c->base + reg);
+	else
+		ret = ioread32be(i2c->base + reg);
+	return ret;
 }
 
 static inline void xiic_irq_dis(struct xiic_i2c *i2c, u32 mask)
@@ -696,6 +727,7 @@ static int xiic_i2c_probe(struct platform_device *pdev)
 	struct resource *res;
 	int ret, irq;
 	u8 i;
+	u32 sr;
 
 	i2c = devm_kzalloc(&pdev->dev, sizeof(*i2c), GFP_KERNEL);
 	if (!i2c)
@@ -728,6 +760,18 @@ static int xiic_i2c_probe(struct platform_device *pdev)
 		return ret;
 	}
 
+	/*
+	 * Detect endianness
+	 * Try to reset the TX FIFO. Then check the EMPTY flag. If it is not
+	 * set, assume that the endianness was wrong and swap.
+	 */
+	i2c->endianness = LITTLE;
+	xiic_setreg32(i2c, XIIC_CR_REG_OFFSET, XIIC_CR_TX_FIFO_RESET_MASK);
+	/* Reset is cleared in xiic_reinit */
+	sr = xiic_getreg32(i2c, XIIC_SR_REG_OFFSET);
+	if (!(sr & XIIC_SR_TX_FIFO_EMPTY_MASK))
+		i2c->endianness = BIG;
+
 	xiic_reinit(i2c);
 
 	/* add i2c adapter to i2c tree */
@@ -771,7 +815,6 @@ static struct platform_driver xiic_i2c_driver = {
 	.probe   = xiic_i2c_probe,
 	.remove  = xiic_i2c_remove,
 	.driver  = {
-		.owner = THIS_MODULE,
 		.name = DRIVER_NAME,
 		.of_match_table = of_match_ptr(xiic_of_match),
 	},
diff --git a/drivers/i2c/busses/i2c-xlr.c b/drivers/i2c/busses/i2c-xlr.c
index 17f7352eca6b..8b36bcfd952d 100644
--- a/drivers/i2c/busses/i2c-xlr.c
+++ b/drivers/i2c/busses/i2c-xlr.c
@@ -263,7 +263,6 @@ static struct platform_driver xlr_i2c_driver = {
 	.remove = xlr_i2c_remove,
 	.driver = {
 		.name   = "xlr-i2cbus",
-		.owner  = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/i2c/busses/scx200_acb.c b/drivers/i2c/busses/scx200_acb.c
index ff3f5747e43b..0a7e410b6195 100644
--- a/drivers/i2c/busses/scx200_acb.c
+++ b/drivers/i2c/busses/scx200_acb.c
@@ -17,10 +17,6 @@
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
     General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
@@ -548,7 +544,6 @@ static int scx200_remove(struct platform_device *pdev)
 static struct platform_driver scx200_pci_driver = {
 	.driver = {
 		.name = "cs5535-smb",
-		.owner = THIS_MODULE,
 	},
 	.probe = scx200_probe,
 	.remove = scx200_remove,
diff --git a/drivers/i2c/i2c-boardinfo.c b/drivers/i2c/i2c-boardinfo.c
index f24cc64e2e8c..90e322959303 100644
--- a/drivers/i2c/i2c-boardinfo.c
+++ b/drivers/i2c/i2c-boardinfo.c
@@ -10,11 +10,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- * MA 02110-1301 USA.
  */
 
 #include <linux/kernel.h>
diff --git a/drivers/i2c/i2c-core.c b/drivers/i2c/i2c-core.c
index ccfbbab82a15..e9eae57a2b50 100644
--- a/drivers/i2c/i2c-core.c
+++ b/drivers/i2c/i2c-core.c
@@ -10,12 +10,7 @@
     This program is distributed in the hope that it will be useful,
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-    GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
-    MA 02110-1301 USA.							     */
+    GNU General Public License for more details.			     */
 /* ------------------------------------------------------------------------- */
 
 /* With some changes from Kyösti Mälkki <kmalkki@cc.hut.fi>.
@@ -29,6 +24,7 @@
    (c) 2013  Wolfram Sang <wsa@the-dreams.de>
    I2C ACPI code Copyright (C) 2014 Intel Corp
    Author: Lan Tianyu <tianyu.lan@intel.com>
+   I2C slave support (c) 2014 by Wolfram Sang <wsa@sang-engineering.com>
  */
 
 #include <linux/module.h>
@@ -50,9 +46,11 @@
 #include <linux/irqflags.h>
 #include <linux/rwsem.h>
 #include <linux/pm_runtime.h>
+#include <linux/pm_domain.h>
 #include <linux/acpi.h>
 #include <linux/jump_label.h>
 #include <asm/uaccess.h>
+#include <linux/err.h>
 
 #include "i2c-core.h"
 
@@ -264,7 +262,7 @@ acpi_i2c_space_handler(u32 function, acpi_physical_address command,
 	struct acpi_resource *ares;
 	u32 accessor_type = function >> 16;
 	u8 action = function & ACPI_IO_MASK;
-	acpi_status ret = AE_OK;
+	acpi_status ret;
 	int status;
 
 	ret = acpi_buffer_to_resource(info->connection, info->length, &ares);
@@ -407,6 +405,7 @@ static int acpi_i2c_install_space_handler(struct i2c_adapter *adapter)
 		return -ENOMEM;
 	}
 
+	acpi_walk_dep_device_list(handle);
 	return 0;
 }
 
@@ -630,6 +629,17 @@ static int i2c_device_probe(struct device *dev)
 	if (!client)
 		return 0;
 
+	if (!client->irq && dev->of_node) {
+		int irq = of_irq_get(dev->of_node, 0);
+
+		if (irq == -EPROBE_DEFER)
+			return irq;
+		if (irq < 0)
+			irq = 0;
+
+		client->irq = irq;
+	}
+
 	driver = to_i2c_driver(dev->driver);
 	if (!driver->probe || !driver->id_table)
 		return -ENODEV;
@@ -643,10 +653,13 @@ static int i2c_device_probe(struct device *dev)
 	if (status < 0)
 		return status;
 
-	acpi_dev_pm_attach(&client->dev, true);
-	status = driver->probe(client, i2c_match_id(driver->id_table, client));
-	if (status)
-		acpi_dev_pm_detach(&client->dev, true);
+	status = dev_pm_domain_attach(&client->dev, true);
+	if (status != -EPROBE_DEFER) {
+		status = driver->probe(client, i2c_match_id(driver->id_table,
+					client));
+		if (status)
+			dev_pm_domain_detach(&client->dev, true);
+	}
 
 	return status;
 }
@@ -666,7 +679,10 @@ static int i2c_device_remove(struct device *dev)
 		status = driver->remove(client);
 	}
 
-	acpi_dev_pm_detach(&client->dev, true);
+	if (dev->of_node)
+		irq_dispose_mapping(client->irq);
+
+	dev_pm_domain_detach(&client->dev, true);
 	return status;
 }
 
@@ -1366,63 +1382,67 @@ static void i2c_scan_static_board_info(struct i2c_adapter *adapter)
 /* OF support code */
 
 #if IS_ENABLED(CONFIG_OF)
-static void of_i2c_register_devices(struct i2c_adapter *adap)
+static struct i2c_client *of_i2c_register_device(struct i2c_adapter *adap,
+						 struct device_node *node)
 {
-	void *result;
-	struct device_node *node;
+	struct i2c_client *result;
+	struct i2c_board_info info = {};
+	struct dev_archdata dev_ad = {};
+	const __be32 *addr;
+	int len;
 
-	/* Only register child devices if the adapter has a node pointer set */
-	if (!adap->dev.of_node)
-		return;
+	dev_dbg(&adap->dev, "of_i2c: register %s\n", node->full_name);
 
-	dev_dbg(&adap->dev, "of_i2c: walking child nodes\n");
+	if (of_modalias_node(node, info.type, sizeof(info.type)) < 0) {
+		dev_err(&adap->dev, "of_i2c: modalias failure on %s\n",
+			node->full_name);
+		return ERR_PTR(-EINVAL);
+	}
 
-	for_each_available_child_of_node(adap->dev.of_node, node) {
-		struct i2c_board_info info = {};
-		struct dev_archdata dev_ad = {};
-		const __be32 *addr;
-		int len;
+	addr = of_get_property(node, "reg", &len);
+	if (!addr || (len < sizeof(int))) {
+		dev_err(&adap->dev, "of_i2c: invalid reg on %s\n",
+			node->full_name);
+		return ERR_PTR(-EINVAL);
+	}
 
-		dev_dbg(&adap->dev, "of_i2c: register %s\n", node->full_name);
+	info.addr = be32_to_cpup(addr);
+	if (info.addr > (1 << 10) - 1) {
+		dev_err(&adap->dev, "of_i2c: invalid addr=%x on %s\n",
+			info.addr, node->full_name);
+		return ERR_PTR(-EINVAL);
+	}
 
-		if (of_modalias_node(node, info.type, sizeof(info.type)) < 0) {
-			dev_err(&adap->dev, "of_i2c: modalias failure on %s\n",
-				node->full_name);
-			continue;
-		}
+	info.of_node = of_node_get(node);
+	info.archdata = &dev_ad;
 
-		addr = of_get_property(node, "reg", &len);
-		if (!addr || (len < sizeof(int))) {
-			dev_err(&adap->dev, "of_i2c: invalid reg on %s\n",
-				node->full_name);
-			continue;
-		}
+	if (of_get_property(node, "wakeup-source", NULL))
+		info.flags |= I2C_CLIENT_WAKE;
 
-		info.addr = be32_to_cpup(addr);
-		if (info.addr > (1 << 10) - 1) {
-			dev_err(&adap->dev, "of_i2c: invalid addr=%x on %s\n",
-				info.addr, node->full_name);
-			continue;
-		}
+	request_module("%s%s", I2C_MODULE_PREFIX, info.type);
 
-		info.irq = irq_of_parse_and_map(node, 0);
-		info.of_node = of_node_get(node);
-		info.archdata = &dev_ad;
+	result = i2c_new_device(adap, &info);
+	if (result == NULL) {
+		dev_err(&adap->dev, "of_i2c: Failure registering %s\n",
+			node->full_name);
+		of_node_put(node);
+		return ERR_PTR(-EINVAL);
+	}
+	return result;
+}
 
-		if (of_get_property(node, "wakeup-source", NULL))
-			info.flags |= I2C_CLIENT_WAKE;
+static void of_i2c_register_devices(struct i2c_adapter *adap)
+{
+	struct device_node *node;
 
-		request_module("%s%s", I2C_MODULE_PREFIX, info.type);
+	/* Only register child devices if the adapter has a node pointer set */
+	if (!adap->dev.of_node)
+		return;
 
-		result = i2c_new_device(adap, &info);
-		if (result == NULL) {
-			dev_err(&adap->dev, "of_i2c: Failure registering %s\n",
-				node->full_name);
-			of_node_put(node);
-			irq_dispose_mapping(info.irq);
-			continue;
-		}
-	}
+	dev_dbg(&adap->dev, "of_i2c: walking child nodes\n");
+
+	for_each_available_child_of_node(adap->dev.of_node, node)
+		of_i2c_register_device(adap, node);
 }
 
 static int of_dev_node_match(struct device *dev, void *data)
@@ -1942,6 +1962,52 @@ void i2c_clients_command(struct i2c_adapter *adap, unsigned int cmd, void *arg)
 }
 EXPORT_SYMBOL(i2c_clients_command);
 
+#if IS_ENABLED(CONFIG_OF_DYNAMIC)
+static int of_i2c_notify(struct notifier_block *nb, unsigned long action,
+			 void *arg)
+{
+	struct of_reconfig_data *rd = arg;
+	struct i2c_adapter *adap;
+	struct i2c_client *client;
+
+	switch (of_reconfig_get_state_change(action, rd)) {
+	case OF_RECONFIG_CHANGE_ADD:
+		adap = of_find_i2c_adapter_by_node(rd->dn->parent);
+		if (adap == NULL)
+			return NOTIFY_OK;	/* not for us */
+
+		client = of_i2c_register_device(adap, rd->dn);
+		put_device(&adap->dev);
+
+		if (IS_ERR(client)) {
+			pr_err("%s: failed to create for '%s'\n",
+					__func__, rd->dn->full_name);
+			return notifier_from_errno(PTR_ERR(client));
+		}
+		break;
+	case OF_RECONFIG_CHANGE_REMOVE:
+		/* find our device by node */
+		client = of_find_i2c_device_by_node(rd->dn);
+		if (client == NULL)
+			return NOTIFY_OK;	/* no? not meant for us */
+
+		/* unregister takes one ref away */
+		i2c_unregister_device(client);
+
+		/* and put the reference of the find */
+		put_device(&client->dev);
+		break;
+	}
+
+	return NOTIFY_OK;
+}
+static struct notifier_block i2c_of_notifier = {
+	.notifier_call = of_i2c_notify,
+};
+#else
+extern struct notifier_block i2c_of_notifier;
+#endif /* CONFIG_OF_DYNAMIC */
+
 static int __init i2c_init(void)
 {
 	int retval;
@@ -1959,6 +2025,10 @@ static int __init i2c_init(void)
 	retval = i2c_add_driver(&dummy_driver);
 	if (retval)
 		goto class_err;
+
+	if (IS_ENABLED(CONFIG_OF_DYNAMIC))
+		WARN_ON(of_reconfig_notifier_register(&i2c_of_notifier));
+
 	return 0;
 
 class_err:
@@ -1972,6 +2042,8 @@ bus_err:
 
 static void __exit i2c_exit(void)
 {
+	if (IS_ENABLED(CONFIG_OF_DYNAMIC))
+		WARN_ON(of_reconfig_notifier_unregister(&i2c_of_notifier));
 	i2c_del_driver(&dummy_driver);
 #ifdef CONFIG_I2C_COMPAT
 	class_compat_unregister(i2c_adapter_compat_class);
@@ -2900,6 +2972,56 @@ trace:
 }
 EXPORT_SYMBOL(i2c_smbus_xfer);
 
+#if IS_ENABLED(CONFIG_I2C_SLAVE)
+int i2c_slave_register(struct i2c_client *client, i2c_slave_cb_t slave_cb)
+{
+	int ret;
+
+	if (!client || !slave_cb)
+		return -EINVAL;
+
+	if (!(client->flags & I2C_CLIENT_TEN)) {
+		/* Enforce stricter address checking */
+		ret = i2c_check_addr_validity(client->addr);
+		if (ret)
+			return ret;
+	}
+
+	if (!client->adapter->algo->reg_slave)
+		return -EOPNOTSUPP;
+
+	client->slave_cb = slave_cb;
+
+	i2c_lock_adapter(client->adapter);
+	ret = client->adapter->algo->reg_slave(client);
+	i2c_unlock_adapter(client->adapter);
+
+	if (ret)
+		client->slave_cb = NULL;
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(i2c_slave_register);
+
+int i2c_slave_unregister(struct i2c_client *client)
+{
+	int ret;
+
+	if (!client->adapter->algo->unreg_slave)
+		return -EOPNOTSUPP;
+
+	i2c_lock_adapter(client->adapter);
+	ret = client->adapter->algo->unreg_slave(client);
+	i2c_unlock_adapter(client->adapter);
+
+	if (ret == 0)
+		client->slave_cb = NULL;
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(i2c_slave_unregister);
+#endif
+
 MODULE_AUTHOR("Simon G. Vogl <simon@tk.uni-linz.ac.at>");
 MODULE_DESCRIPTION("I2C-Bus main module");
 MODULE_LICENSE("GPL");
diff --git a/drivers/i2c/i2c-core.h b/drivers/i2c/i2c-core.h
index 18a8fd21d2c2..17700bfddcf5 100644
--- a/drivers/i2c/i2c-core.h
+++ b/drivers/i2c/i2c-core.h
@@ -10,11 +10,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- * MA 02110-1301 USA.
  */
 
 #include <linux/rwsem.h>
diff --git a/drivers/i2c/i2c-dev.c b/drivers/i2c/i2c-dev.c
index 80b47e8ce030..71c7a3975b62 100644
--- a/drivers/i2c/i2c-dev.c
+++ b/drivers/i2c/i2c-dev.c
@@ -14,11 +14,6 @@
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
     GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
-    MA 02110-1301 USA.
 */
 
 /* Note that this is a complete rewrite of Simon Vogl's i2c-dev module.
diff --git a/drivers/i2c/i2c-mux.c b/drivers/i2c/i2c-mux.c
index 2d0847b6be62..593f7ca9adc7 100644
--- a/drivers/i2c/i2c-mux.c
+++ b/drivers/i2c/i2c-mux.c
@@ -110,6 +110,7 @@ struct i2c_adapter *i2c_add_mux_adapter(struct i2c_adapter *parent,
 						 void *, u32))
 {
 	struct i2c_mux_priv *priv;
+	char symlink_name[20];
 	int ret;
 
 	priv = kzalloc(sizeof(struct i2c_mux_priv), GFP_KERNEL);
@@ -183,6 +184,12 @@ struct i2c_adapter *i2c_add_mux_adapter(struct i2c_adapter *parent,
 		return NULL;
 	}
 
+	WARN(sysfs_create_link(&priv->adap.dev.kobj, &mux_dev->kobj, "mux_device"),
+			       "can't create symlink to mux device\n");
+
+	snprintf(symlink_name, sizeof(symlink_name), "channel-%u", chan_id);
+	WARN(sysfs_create_link(&mux_dev->kobj, &priv->adap.dev.kobj, symlink_name),
+			       "can't create symlink for channel %u\n", chan_id);
 	dev_info(&parent->dev, "Added multiplexed i2c bus %d\n",
 		 i2c_adapter_id(&priv->adap));
 
@@ -193,7 +200,12 @@ EXPORT_SYMBOL_GPL(i2c_add_mux_adapter);
 void i2c_del_mux_adapter(struct i2c_adapter *adap)
 {
 	struct i2c_mux_priv *priv = adap->algo_data;
+	char symlink_name[20];
+
+	snprintf(symlink_name, sizeof(symlink_name), "channel-%u", priv->chan_id);
+	sysfs_remove_link(&adap->dev.parent->kobj, symlink_name);
 
+	sysfs_remove_link(&priv->adap.dev.kobj, "mux_device");
 	i2c_del_adapter(adap);
 	kfree(priv);
 }
diff --git a/drivers/i2c/i2c-slave-eeprom.c b/drivers/i2c/i2c-slave-eeprom.c
new file mode 100644
index 000000000000..cf9b09db092f
--- /dev/null
+++ b/drivers/i2c/i2c-slave-eeprom.c
@@ -0,0 +1,170 @@
+/*
+ * I2C slave mode EEPROM simulator
+ *
+ * Copyright (C) 2014 by Wolfram Sang, Sang Engineering <wsa@sang-engineering.com>
+ * Copyright (C) 2014 by Renesas Electronics Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; version 2 of the License.
+ *
+ * Because most IP blocks can only detect one I2C slave address anyhow, this
+ * driver does not support simulating EEPROM types which take more than one
+ * address. It is prepared to simulate bigger EEPROMs with an internal 16 bit
+ * pointer, yet implementation is deferred until the need actually arises.
+ */
+
+#include <linux/i2c.h>
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+#include <linux/sysfs.h>
+
+struct eeprom_data {
+	struct bin_attribute bin;
+	bool first_write;
+	spinlock_t buffer_lock;
+	u8 buffer_idx;
+	u8 buffer[];
+};
+
+static int i2c_slave_eeprom_slave_cb(struct i2c_client *client,
+				     enum i2c_slave_event event, u8 *val)
+{
+	struct eeprom_data *eeprom = i2c_get_clientdata(client);
+
+	switch (event) {
+	case I2C_SLAVE_REQ_WRITE_END:
+		if (eeprom->first_write) {
+			eeprom->buffer_idx = *val;
+			eeprom->first_write = false;
+		} else {
+			spin_lock(&eeprom->buffer_lock);
+			eeprom->buffer[eeprom->buffer_idx++] = *val;
+			spin_unlock(&eeprom->buffer_lock);
+		}
+		break;
+
+	case I2C_SLAVE_REQ_READ_START:
+		spin_lock(&eeprom->buffer_lock);
+		*val = eeprom->buffer[eeprom->buffer_idx];
+		spin_unlock(&eeprom->buffer_lock);
+		break;
+
+	case I2C_SLAVE_REQ_READ_END:
+		eeprom->buffer_idx++;
+		break;
+
+	case I2C_SLAVE_STOP:
+		eeprom->first_write = true;
+		break;
+
+	default:
+		break;
+	}
+
+	return 0;
+}
+
+static ssize_t i2c_slave_eeprom_bin_read(struct file *filp, struct kobject *kobj,
+		struct bin_attribute *attr, char *buf, loff_t off, size_t count)
+{
+	struct eeprom_data *eeprom;
+	unsigned long flags;
+
+	if (off + count > attr->size)
+		return -EFBIG;
+
+	eeprom = dev_get_drvdata(container_of(kobj, struct device, kobj));
+
+	spin_lock_irqsave(&eeprom->buffer_lock, flags);
+	memcpy(buf, &eeprom->buffer[off], count);
+	spin_unlock_irqrestore(&eeprom->buffer_lock, flags);
+
+	return count;
+}
+
+static ssize_t i2c_slave_eeprom_bin_write(struct file *filp, struct kobject *kobj,
+		struct bin_attribute *attr, char *buf, loff_t off, size_t count)
+{
+	struct eeprom_data *eeprom;
+	unsigned long flags;
+
+	if (off + count > attr->size)
+		return -EFBIG;
+
+	eeprom = dev_get_drvdata(container_of(kobj, struct device, kobj));
+
+	spin_lock_irqsave(&eeprom->buffer_lock, flags);
+	memcpy(&eeprom->buffer[off], buf, count);
+	spin_unlock_irqrestore(&eeprom->buffer_lock, flags);
+
+	return count;
+}
+
+static int i2c_slave_eeprom_probe(struct i2c_client *client, const struct i2c_device_id *id)
+{
+	struct eeprom_data *eeprom;
+	int ret;
+	unsigned size = id->driver_data;
+
+	eeprom = devm_kzalloc(&client->dev, sizeof(struct eeprom_data) + size, GFP_KERNEL);
+	if (!eeprom)
+		return -ENOMEM;
+
+	eeprom->first_write = true;
+	spin_lock_init(&eeprom->buffer_lock);
+	i2c_set_clientdata(client, eeprom);
+
+	sysfs_bin_attr_init(&eeprom->bin);
+	eeprom->bin.attr.name = "slave-eeprom";
+	eeprom->bin.attr.mode = S_IRUSR | S_IWUSR;
+	eeprom->bin.read = i2c_slave_eeprom_bin_read;
+	eeprom->bin.write = i2c_slave_eeprom_bin_write;
+	eeprom->bin.size = size;
+
+	ret = sysfs_create_bin_file(&client->dev.kobj, &eeprom->bin);
+	if (ret)
+		return ret;
+
+	ret = i2c_slave_register(client, i2c_slave_eeprom_slave_cb);
+	if (ret) {
+		sysfs_remove_bin_file(&client->dev.kobj, &eeprom->bin);
+		return ret;
+	}
+
+	return 0;
+};
+
+static int i2c_slave_eeprom_remove(struct i2c_client *client)
+{
+	struct eeprom_data *eeprom = i2c_get_clientdata(client);
+
+	i2c_slave_unregister(client);
+	sysfs_remove_bin_file(&client->dev.kobj, &eeprom->bin);
+
+	return 0;
+}
+
+static const struct i2c_device_id i2c_slave_eeprom_id[] = {
+	{ "slave-24c02", 2048 / 8 },
+	{ }
+};
+MODULE_DEVICE_TABLE(i2c, i2c_slave_eeprom_id);
+
+static struct i2c_driver i2c_slave_eeprom_driver = {
+	.driver = {
+		.name = "i2c-slave-eeprom",
+		.owner = THIS_MODULE,
+	},
+	.probe = i2c_slave_eeprom_probe,
+	.remove = i2c_slave_eeprom_remove,
+	.id_table = i2c_slave_eeprom_id,
+};
+module_i2c_driver(i2c_slave_eeprom_driver);
+
+MODULE_AUTHOR("Wolfram Sang <wsa@sang-engineering.com>");
+MODULE_DESCRIPTION("I2C slave mode EEPROM simulator");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/i2c/i2c-smbus.c b/drivers/i2c/i2c-smbus.c
index fc99f0d6b4a5..9ebf9cb4ad7a 100644
--- a/drivers/i2c/i2c-smbus.c
+++ b/drivers/i2c/i2c-smbus.c
@@ -13,11 +13,6 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- * MA 02110-1301 USA.
  */
 
 #include <linux/kernel.h>
diff --git a/drivers/i2c/i2c-stub.c b/drivers/i2c/i2c-stub.c
index d241aa295d96..af2a94e1140b 100644
--- a/drivers/i2c/i2c-stub.c
+++ b/drivers/i2c/i2c-stub.c
@@ -13,10 +13,6 @@
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
     GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License
-    along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 #define DEBUG 1
diff --git a/drivers/i2c/muxes/i2c-arb-gpio-challenge.c b/drivers/i2c/muxes/i2c-arb-gpio-challenge.c
index 69afffa8f427..5cf1b60b69e2 100644
--- a/drivers/i2c/muxes/i2c-arb-gpio-challenge.c
+++ b/drivers/i2c/muxes/i2c-arb-gpio-challenge.c
@@ -235,7 +235,6 @@ static struct platform_driver i2c_arbitrator_driver = {
 	.probe	= i2c_arbitrator_probe,
 	.remove	= i2c_arbitrator_remove,
 	.driver	= {
-		.owner	= THIS_MODULE,
 		.name	= "i2c-arb-gpio-challenge",
 		.of_match_table = i2c_arbitrator_of_match,
 	},
diff --git a/drivers/i2c/muxes/i2c-mux-gpio.c b/drivers/i2c/muxes/i2c-mux-gpio.c
index d8989c823f50..f5798eb4076b 100644
--- a/drivers/i2c/muxes/i2c-mux-gpio.c
+++ b/drivers/i2c/muxes/i2c-mux-gpio.c
@@ -280,7 +280,6 @@ static struct platform_driver i2c_mux_gpio_driver = {
 	.probe	= i2c_mux_gpio_probe,
 	.remove	= i2c_mux_gpio_remove,
 	.driver	= {
-		.owner	= THIS_MODULE,
 		.name	= "i2c-mux-gpio",
 		.of_match_table = i2c_mux_gpio_of_match,
 	},
diff --git a/drivers/i2c/muxes/i2c-mux-pinctrl.c b/drivers/i2c/muxes/i2c-mux-pinctrl.c
index 4ff0ef3e07a6..b48378c4b40d 100644
--- a/drivers/i2c/muxes/i2c-mux-pinctrl.c
+++ b/drivers/i2c/muxes/i2c-mux-pinctrl.c
@@ -264,7 +264,6 @@ MODULE_DEVICE_TABLE(of, i2c_mux_pinctrl_of_match);
 static struct platform_driver i2c_mux_pinctrl_driver = {
 	.driver	= {
 		.name	= "i2c-mux-pinctrl",
-		.owner	= THIS_MODULE,
 		.of_match_table = of_match_ptr(i2c_mux_pinctrl_of_match),
 	},
 	.probe	= i2c_mux_pinctrl_probe,
diff --git a/drivers/ide/atiixp.c b/drivers/ide/atiixp.c
index dbd0f242ec18..76650e92db41 100644
--- a/drivers/ide/atiixp.c
+++ b/drivers/ide/atiixp.c
@@ -19,12 +19,12 @@
 #define ATIIXP_IDE_UDMA_CONTROL		0x54
 #define ATIIXP_IDE_UDMA_MODE		0x56
 
-typedef struct {
+struct atiixp_ide_timing {
 	u8 command_width;
 	u8 recover_width;
-} atiixp_ide_timing;
+};
 
-static atiixp_ide_timing pio_timing[] = {
+static struct atiixp_ide_timing pio_timing[] = {
 	{ 0x05, 0x0d },
 	{ 0x04, 0x07 },
 	{ 0x03, 0x04 },
@@ -32,7 +32,7 @@ static atiixp_ide_timing pio_timing[] = {
 	{ 0x02, 0x00 },
 };
 
-static atiixp_ide_timing mdma_timing[] = {
+static struct atiixp_ide_timing mdma_timing[] = {
 	{ 0x07, 0x07 },
 	{ 0x02, 0x01 },
 	{ 0x02, 0x00 },
diff --git a/drivers/ide/au1xxx-ide.c b/drivers/ide/au1xxx-ide.c
index 07ea58084068..4d181a918d72 100644
--- a/drivers/ide/au1xxx-ide.c
+++ b/drivers/ide/au1xxx-ide.c
@@ -586,7 +586,6 @@ static int au_ide_remove(struct platform_device *dev)
 static struct platform_driver au1200_ide_driver = {
 	.driver = {
 		.name		= "au1200-ide",
-		.owner		= THIS_MODULE,
 	},
 	.probe 		= au_ide_probe,
 	.remove		= au_ide_remove,
diff --git a/drivers/ide/gayle.c b/drivers/ide/gayle.c
index 97a2f9dc75d8..901e6ebfeb96 100644
--- a/drivers/ide/gayle.c
+++ b/drivers/ide/gayle.c
@@ -179,7 +179,6 @@ static struct platform_driver amiga_gayle_ide_driver = {
 	.remove = __exit_p(amiga_gayle_ide_remove),
 	.driver   = {
 		.name	= "amiga-gayle-ide",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/ide/ide-disk.c b/drivers/ide/ide-disk.c
index ee880382e3bc..56b9708894a5 100644
--- a/drivers/ide/ide-disk.c
+++ b/drivers/ide/ide-disk.c
@@ -685,8 +685,10 @@ static void ide_disk_setup(ide_drive_t *drive)
 	printk(KERN_INFO "%s: max request size: %dKiB\n", drive->name,
 	       queue_max_sectors(q) / 2);
 
-	if (ata_id_is_ssd(id))
+	if (ata_id_is_ssd(id)) {
 		queue_flag_set_unlocked(QUEUE_FLAG_NONROT, q);
+		queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, q);
+	}
 
 	/* calculate drive capacity, and select LBA if possible */
 	ide_disk_get_capacity(drive);
diff --git a/drivers/ide/ide-floppy.c b/drivers/ide/ide-floppy.c
index 3d42043fec51..8c6363cdd208 100644
--- a/drivers/ide/ide-floppy.c
+++ b/drivers/ide/ide-floppy.c
@@ -487,7 +487,7 @@ static void ide_floppy_setup(ide_drive_t *drive)
 	 * it. It should be fixed as of version 1.9, but to be on the safe side
 	 * we'll leave the limitation below for the 2.2.x tree.
 	 */
-	if (!strncmp((char *)&id[ATA_ID_PROD], "IOMEGA ZIP 100 ATAPI", 20)) {
+	if (strstarts((char *)&id[ATA_ID_PROD], "IOMEGA ZIP 100 ATAPI")) {
 		drive->atapi_flags |= IDE_AFLAG_ZIP_DRIVE;
 		/* This value will be visible in the /proc/ide/hdx/settings */
 		drive->pc_delay = IDEFLOPPY_PC_DELAY;
@@ -498,7 +498,7 @@ static void ide_floppy_setup(ide_drive_t *drive)
 	 * Guess what? The IOMEGA Clik! drive also needs the above fix. It makes
 	 * nasty clicking noises without it, so please don't remove this.
 	 */
-	if (strncmp((char *)&id[ATA_ID_PROD], "IOMEGA Clik!", 11) == 0) {
+	if (strstarts((char *)&id[ATA_ID_PROD], "IOMEGA Clik!")) {
 		blk_queue_max_hw_sectors(drive->queue, 64);
 		drive->atapi_flags |= IDE_AFLAG_CLIK_DRIVE;
 		/* IOMEGA Clik! drives do not support lock/unlock commands */
diff --git a/drivers/ide/ide-park.c b/drivers/ide/ide-park.c
index f41558a0bcd1..ca958604cda2 100644
--- a/drivers/ide/ide-park.c
+++ b/drivers/ide/ide-park.c
@@ -46,7 +46,7 @@ static void issue_park_cmd(ide_drive_t *drive, unsigned long timeout)
 	 * timeout has expired, so power management will be reenabled.
 	 */
 	rq = blk_get_request(q, READ, GFP_NOWAIT);
-	if (unlikely(!rq))
+	if (IS_ERR(rq))
 		goto out;
 
 	rq->cmd[0] = REQ_UNPARK_HEADS;
diff --git a/drivers/ide/ide.c b/drivers/ide/ide.c
index 2ce6268a2734..e29b02ca9e91 100644
--- a/drivers/ide/ide.c
+++ b/drivers/ide/ide.c
@@ -101,8 +101,7 @@ void ide_device_put(ide_drive_t *drive)
 	struct device *host_dev = drive->hwif->host->dev[0];
 	struct module *module = host_dev ? host_dev->driver->owner : NULL;
 
-	if (module)
-		module_put(module);
+	module_put(module);
 #endif
 	put_device(&drive->gendev);
 }
diff --git a/drivers/ide/ide_platform.c b/drivers/ide/ide_platform.c
index d48de6de503e..2b43be3bca0a 100644
--- a/drivers/ide/ide_platform.c
+++ b/drivers/ide/ide_platform.c
@@ -125,7 +125,6 @@ static int plat_ide_remove(struct platform_device *pdev)
 static struct platform_driver platform_ide_driver = {
 	.driver = {
 		.name = "pata_platform",
-		.owner = THIS_MODULE,
 	},
 	.probe = plat_ide_probe,
 	.remove = plat_ide_remove,
diff --git a/drivers/ide/palm_bk3710.c b/drivers/ide/palm_bk3710.c
index ba20d18c0373..8012e43bf8f6 100644
--- a/drivers/ide/palm_bk3710.c
+++ b/drivers/ide/palm_bk3710.c
@@ -386,7 +386,6 @@ MODULE_ALIAS("platform:palm_bk3710");
 static struct platform_driver platform_bk_driver = {
 	.driver = {
 		.name = "palm_bk3710",
-		.owner = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/ide/tx4938ide.c b/drivers/ide/tx4938ide.c
index 68edd4f58a28..40a3f55b08dd 100644
--- a/drivers/ide/tx4938ide.c
+++ b/drivers/ide/tx4938ide.c
@@ -198,7 +198,6 @@ static int __exit tx4938ide_remove(struct platform_device *pdev)
 static struct platform_driver tx4938ide_driver = {
 	.driver		= {
 		.name	= "tx4938ide",
-		.owner	= THIS_MODULE,
 	},
 	.remove = __exit_p(tx4938ide_remove),
 };
diff --git a/drivers/ide/tx4939ide.c b/drivers/ide/tx4939ide.c
index 4ecdee5eca83..67d4a7d4acc8 100644
--- a/drivers/ide/tx4939ide.c
+++ b/drivers/ide/tx4939ide.c
@@ -618,7 +618,6 @@ static int tx4939ide_resume(struct platform_device *dev)
 static struct platform_driver tx4939ide_driver = {
 	.driver = {
 		.name = MODNAME,
-		.owner = THIS_MODULE,
 	},
 	.remove = __exit_p(tx4939ide_remove),
 	.resume = tx4939ide_resume,
diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c
index 9b7ee7e427df..9cceacb92f9d 100644
--- a/drivers/idle/intel_idle.c
+++ b/drivers/idle/intel_idle.c
@@ -128,28 +128,28 @@ static struct cpuidle_state nehalem_cstates[] = {
 	{
 		.name = "C1-NHM",
 		.desc = "MWAIT 0x00",
-		.flags = MWAIT2flg(0x00) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x00),
 		.exit_latency = 3,
 		.target_residency = 6,
 		.enter = &intel_idle },
 	{
 		.name = "C1E-NHM",
 		.desc = "MWAIT 0x01",
-		.flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x01),
 		.exit_latency = 10,
 		.target_residency = 20,
 		.enter = &intel_idle },
 	{
 		.name = "C3-NHM",
 		.desc = "MWAIT 0x10",
-		.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 20,
 		.target_residency = 80,
 		.enter = &intel_idle },
 	{
 		.name = "C6-NHM",
 		.desc = "MWAIT 0x20",
-		.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 200,
 		.target_residency = 800,
 		.enter = &intel_idle },
@@ -161,35 +161,35 @@ static struct cpuidle_state snb_cstates[] = {
 	{
 		.name = "C1-SNB",
 		.desc = "MWAIT 0x00",
-		.flags = MWAIT2flg(0x00) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x00),
 		.exit_latency = 2,
 		.target_residency = 2,
 		.enter = &intel_idle },
 	{
 		.name = "C1E-SNB",
 		.desc = "MWAIT 0x01",
-		.flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x01),
 		.exit_latency = 10,
 		.target_residency = 20,
 		.enter = &intel_idle },
 	{
 		.name = "C3-SNB",
 		.desc = "MWAIT 0x10",
-		.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 80,
 		.target_residency = 211,
 		.enter = &intel_idle },
 	{
 		.name = "C6-SNB",
 		.desc = "MWAIT 0x20",
-		.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 104,
 		.target_residency = 345,
 		.enter = &intel_idle },
 	{
 		.name = "C7-SNB",
 		.desc = "MWAIT 0x30",
-		.flags = MWAIT2flg(0x30) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x30) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 109,
 		.target_residency = 345,
 		.enter = &intel_idle },
@@ -201,42 +201,42 @@ static struct cpuidle_state byt_cstates[] = {
 	{
 		.name = "C1-BYT",
 		.desc = "MWAIT 0x00",
-		.flags = MWAIT2flg(0x00) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x00),
 		.exit_latency = 1,
 		.target_residency = 1,
 		.enter = &intel_idle },
 	{
 		.name = "C1E-BYT",
 		.desc = "MWAIT 0x01",
-		.flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x01),
 		.exit_latency = 15,
 		.target_residency = 30,
 		.enter = &intel_idle },
 	{
 		.name = "C6N-BYT",
 		.desc = "MWAIT 0x58",
-		.flags = MWAIT2flg(0x58) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x58) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 40,
 		.target_residency = 275,
 		.enter = &intel_idle },
 	{
 		.name = "C6S-BYT",
 		.desc = "MWAIT 0x52",
-		.flags = MWAIT2flg(0x52) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x52) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 140,
 		.target_residency = 560,
 		.enter = &intel_idle },
 	{
 		.name = "C7-BYT",
 		.desc = "MWAIT 0x60",
-		.flags = MWAIT2flg(0x60) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x60) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 1200,
 		.target_residency = 1500,
 		.enter = &intel_idle },
 	{
 		.name = "C7S-BYT",
 		.desc = "MWAIT 0x64",
-		.flags = MWAIT2flg(0x64) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x64) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 10000,
 		.target_residency = 20000,
 		.enter = &intel_idle },
@@ -248,35 +248,35 @@ static struct cpuidle_state ivb_cstates[] = {
 	{
 		.name = "C1-IVB",
 		.desc = "MWAIT 0x00",
-		.flags = MWAIT2flg(0x00) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x00),
 		.exit_latency = 1,
 		.target_residency = 1,
 		.enter = &intel_idle },
 	{
 		.name = "C1E-IVB",
 		.desc = "MWAIT 0x01",
-		.flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x01),
 		.exit_latency = 10,
 		.target_residency = 20,
 		.enter = &intel_idle },
 	{
 		.name = "C3-IVB",
 		.desc = "MWAIT 0x10",
-		.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 59,
 		.target_residency = 156,
 		.enter = &intel_idle },
 	{
 		.name = "C6-IVB",
 		.desc = "MWAIT 0x20",
-		.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 80,
 		.target_residency = 300,
 		.enter = &intel_idle },
 	{
 		.name = "C7-IVB",
 		.desc = "MWAIT 0x30",
-		.flags = MWAIT2flg(0x30) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x30) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 87,
 		.target_residency = 300,
 		.enter = &intel_idle },
@@ -288,28 +288,28 @@ static struct cpuidle_state ivt_cstates[] = {
 	{
 		.name = "C1-IVT",
 		.desc = "MWAIT 0x00",
-		.flags = MWAIT2flg(0x00) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x00),
 		.exit_latency = 1,
 		.target_residency = 1,
 		.enter = &intel_idle },
 	{
 		.name = "C1E-IVT",
 		.desc = "MWAIT 0x01",
-		.flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x01),
 		.exit_latency = 10,
 		.target_residency = 80,
 		.enter = &intel_idle },
 	{
 		.name = "C3-IVT",
 		.desc = "MWAIT 0x10",
-		.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 59,
 		.target_residency = 156,
 		.enter = &intel_idle },
 	{
 		.name = "C6-IVT",
 		.desc = "MWAIT 0x20",
-		.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 82,
 		.target_residency = 300,
 		.enter = &intel_idle },
@@ -321,28 +321,28 @@ static struct cpuidle_state ivt_cstates_4s[] = {
 	{
 		.name = "C1-IVT-4S",
 		.desc = "MWAIT 0x00",
-		.flags = MWAIT2flg(0x00) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x00),
 		.exit_latency = 1,
 		.target_residency = 1,
 		.enter = &intel_idle },
 	{
 		.name = "C1E-IVT-4S",
 		.desc = "MWAIT 0x01",
-		.flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x01),
 		.exit_latency = 10,
 		.target_residency = 250,
 		.enter = &intel_idle },
 	{
 		.name = "C3-IVT-4S",
 		.desc = "MWAIT 0x10",
-		.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 59,
 		.target_residency = 300,
 		.enter = &intel_idle },
 	{
 		.name = "C6-IVT-4S",
 		.desc = "MWAIT 0x20",
-		.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 84,
 		.target_residency = 400,
 		.enter = &intel_idle },
@@ -354,28 +354,28 @@ static struct cpuidle_state ivt_cstates_8s[] = {
 	{
 		.name = "C1-IVT-8S",
 		.desc = "MWAIT 0x00",
-		.flags = MWAIT2flg(0x00) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x00),
 		.exit_latency = 1,
 		.target_residency = 1,
 		.enter = &intel_idle },
 	{
 		.name = "C1E-IVT-8S",
 		.desc = "MWAIT 0x01",
-		.flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x01),
 		.exit_latency = 10,
 		.target_residency = 500,
 		.enter = &intel_idle },
 	{
 		.name = "C3-IVT-8S",
 		.desc = "MWAIT 0x10",
-		.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 59,
 		.target_residency = 600,
 		.enter = &intel_idle },
 	{
 		.name = "C6-IVT-8S",
 		.desc = "MWAIT 0x20",
-		.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 88,
 		.target_residency = 700,
 		.enter = &intel_idle },
@@ -387,56 +387,56 @@ static struct cpuidle_state hsw_cstates[] = {
 	{
 		.name = "C1-HSW",
 		.desc = "MWAIT 0x00",
-		.flags = MWAIT2flg(0x00) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x00),
 		.exit_latency = 2,
 		.target_residency = 2,
 		.enter = &intel_idle },
 	{
 		.name = "C1E-HSW",
 		.desc = "MWAIT 0x01",
-		.flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x01),
 		.exit_latency = 10,
 		.target_residency = 20,
 		.enter = &intel_idle },
 	{
 		.name = "C3-HSW",
 		.desc = "MWAIT 0x10",
-		.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 33,
 		.target_residency = 100,
 		.enter = &intel_idle },
 	{
 		.name = "C6-HSW",
 		.desc = "MWAIT 0x20",
-		.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 133,
 		.target_residency = 400,
 		.enter = &intel_idle },
 	{
 		.name = "C7s-HSW",
 		.desc = "MWAIT 0x32",
-		.flags = MWAIT2flg(0x32) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x32) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 166,
 		.target_residency = 500,
 		.enter = &intel_idle },
 	{
 		.name = "C8-HSW",
 		.desc = "MWAIT 0x40",
-		.flags = MWAIT2flg(0x40) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x40) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 300,
 		.target_residency = 900,
 		.enter = &intel_idle },
 	{
 		.name = "C9-HSW",
 		.desc = "MWAIT 0x50",
-		.flags = MWAIT2flg(0x50) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x50) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 600,
 		.target_residency = 1800,
 		.enter = &intel_idle },
 	{
 		.name = "C10-HSW",
 		.desc = "MWAIT 0x60",
-		.flags = MWAIT2flg(0x60) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x60) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 2600,
 		.target_residency = 7700,
 		.enter = &intel_idle },
@@ -447,56 +447,56 @@ static struct cpuidle_state bdw_cstates[] = {
 	{
 		.name = "C1-BDW",
 		.desc = "MWAIT 0x00",
-		.flags = MWAIT2flg(0x00) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x00),
 		.exit_latency = 2,
 		.target_residency = 2,
 		.enter = &intel_idle },
 	{
 		.name = "C1E-BDW",
 		.desc = "MWAIT 0x01",
-		.flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x01),
 		.exit_latency = 10,
 		.target_residency = 20,
 		.enter = &intel_idle },
 	{
 		.name = "C3-BDW",
 		.desc = "MWAIT 0x10",
-		.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 40,
 		.target_residency = 100,
 		.enter = &intel_idle },
 	{
 		.name = "C6-BDW",
 		.desc = "MWAIT 0x20",
-		.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 133,
 		.target_residency = 400,
 		.enter = &intel_idle },
 	{
 		.name = "C7s-BDW",
 		.desc = "MWAIT 0x32",
-		.flags = MWAIT2flg(0x32) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x32) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 166,
 		.target_residency = 500,
 		.enter = &intel_idle },
 	{
 		.name = "C8-BDW",
 		.desc = "MWAIT 0x40",
-		.flags = MWAIT2flg(0x40) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x40) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 300,
 		.target_residency = 900,
 		.enter = &intel_idle },
 	{
 		.name = "C9-BDW",
 		.desc = "MWAIT 0x50",
-		.flags = MWAIT2flg(0x50) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x50) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 600,
 		.target_residency = 1800,
 		.enter = &intel_idle },
 	{
 		.name = "C10-BDW",
 		.desc = "MWAIT 0x60",
-		.flags = MWAIT2flg(0x60) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x60) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 2600,
 		.target_residency = 7700,
 		.enter = &intel_idle },
@@ -508,28 +508,28 @@ static struct cpuidle_state atom_cstates[] = {
 	{
 		.name = "C1E-ATM",
 		.desc = "MWAIT 0x00",
-		.flags = MWAIT2flg(0x00) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x00),
 		.exit_latency = 10,
 		.target_residency = 20,
 		.enter = &intel_idle },
 	{
 		.name = "C2-ATM",
 		.desc = "MWAIT 0x10",
-		.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x10),
 		.exit_latency = 20,
 		.target_residency = 80,
 		.enter = &intel_idle },
 	{
 		.name = "C4-ATM",
 		.desc = "MWAIT 0x30",
-		.flags = MWAIT2flg(0x30) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x30) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 100,
 		.target_residency = 400,
 		.enter = &intel_idle },
 	{
 		.name = "C6-ATM",
 		.desc = "MWAIT 0x52",
-		.flags = MWAIT2flg(0x52) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x52) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 140,
 		.target_residency = 560,
 		.enter = &intel_idle },
@@ -540,14 +540,14 @@ static struct cpuidle_state avn_cstates[] = {
 	{
 		.name = "C1-AVN",
 		.desc = "MWAIT 0x00",
-		.flags = MWAIT2flg(0x00) | CPUIDLE_FLAG_TIME_VALID,
+		.flags = MWAIT2flg(0x00),
 		.exit_latency = 2,
 		.target_residency = 2,
 		.enter = &intel_idle },
 	{
 		.name = "C6-AVN",
 		.desc = "MWAIT 0x51",
-		.flags = MWAIT2flg(0x51) | CPUIDLE_FLAG_TIME_VALID | CPUIDLE_FLAG_TLB_FLUSHED,
+		.flags = MWAIT2flg(0x51) | CPUIDLE_FLAG_TLB_FLUSHED,
 		.exit_latency = 15,
 		.target_residency = 45,
 		.enter = &intel_idle },
diff --git a/drivers/iio/accel/Kconfig b/drivers/iio/accel/Kconfig
index 12addf272a61..9b9be8725e9d 100644
--- a/drivers/iio/accel/Kconfig
+++ b/drivers/iio/accel/Kconfig
@@ -6,17 +6,32 @@
 menu "Accelerometers"
 
 config BMA180
-	tristate "Bosch BMA180 3-Axis Accelerometer Driver"
+	tristate "Bosch BMA180/BMA250 3-Axis Accelerometer Driver"
 	depends on I2C
 	select IIO_BUFFER
 	select IIO_TRIGGERED_BUFFER
 	help
-	  Say Y here if you want to build a driver for the Bosch BMA180
-	  triaxial acceleration sensor.
+	  Say Y here if you want to build a driver for the Bosch BMA180 or
+	  BMA250 triaxial acceleration sensor.
 
 	  To compile this driver as a module, choose M here: the
 	  module will be called bma180.
 
+config BMC150_ACCEL
+	tristate "Bosch BMC150 Accelerometer Driver"
+	depends on I2C
+	select IIO_BUFFER
+	select IIO_TRIGGERED_BUFFER
+	help
+	  Say yes here to build support for the following Bosch accelerometers:
+	  BMC150, BMI055, BMA250E, BMA222E, BMA255, BMA280.
+
+	  Currently this only supports the device via an i2c interface.
+
+	  This is a combo module with both accelerometer and magnetometer.
+	  This driver is only implementing accelerometer part, which has
+	  its own address and register map.
+
 config HID_SENSOR_ACCEL_3D
 	depends on HID_SENSOR_HUB
 	select IIO_BUFFER
@@ -84,7 +99,8 @@ config KXCJK1013
 	select IIO_TRIGGERED_BUFFER
 	help
 	  Say Y here if you want to build a driver for the Kionix KXCJK-1013
-	  triaxial acceleration sensor.
+	  triaxial acceleration sensor. This driver also supports KXCJ9-1008
+	  and KXTJ2-1009.
 
 	  To compile this driver as a module, choose M here: the module will
 	  be called kxcjk-1013.
diff --git a/drivers/iio/accel/Makefile b/drivers/iio/accel/Makefile
index 6578ca1a8e09..a593996c6539 100644
--- a/drivers/iio/accel/Makefile
+++ b/drivers/iio/accel/Makefile
@@ -4,6 +4,7 @@
 
 # When adding new entries keep the list in alphabetical order
 obj-$(CONFIG_BMA180) += bma180.o
+obj-$(CONFIG_BMC150_ACCEL) += bmc150-accel.o
 obj-$(CONFIG_HID_SENSOR_ACCEL_3D) += hid-sensor-accel-3d.o
 obj-$(CONFIG_KXCJK1013) += kxcjk-1013.o
 obj-$(CONFIG_KXSD9)	+= kxsd9.o
diff --git a/drivers/iio/accel/bma180.c b/drivers/iio/accel/bma180.c
index 19100fddd2ed..1096da327130 100644
--- a/drivers/iio/accel/bma180.c
+++ b/drivers/iio/accel/bma180.c
@@ -3,9 +3,15 @@
  *
  * Copyright 2013 Oleksandr Kravchenko <x0199363@ti.com>
  *
+ * Support for BMA250 (c) Peter Meerwald <pmeerw@pmeerw.net>
+ *
  * This file is subject to the terms and conditions of version 2 of
  * the GNU General Public License.  See the file COPYING in the main
  * directory of this archive for more details.
+ *
+ * SPI is not supported by driver
+ * BMA180: 7-bit I2C slave address 0x40 or 0x41
+ * BMA250: 7-bit I2C slave address 0x18 or 0x19
  */
 
 #include <linux/module.h>
@@ -26,9 +32,37 @@
 #define BMA180_DRV_NAME "bma180"
 #define BMA180_IRQ_NAME "bma180_event"
 
+enum {
+	BMA180,
+	BMA250,
+};
+
+struct bma180_data;
+
+struct bma180_part_info {
+	const struct iio_chan_spec *channels;
+	unsigned num_channels;
+	const int *scale_table;
+	unsigned num_scales;
+	const int *bw_table;
+	unsigned num_bw;
+
+	u8 int_reset_reg, int_reset_mask;
+	u8 sleep_reg, sleep_mask;
+	u8 bw_reg, bw_mask;
+	u8 scale_reg, scale_mask;
+	u8 power_reg, power_mask, lowpower_val;
+	u8 int_enable_reg, int_enable_mask;
+	u8 softreset_reg;
+
+	int (*chip_config)(struct bma180_data *data);
+	void (*chip_disable)(struct bma180_data *data);
+};
+
 /* Register set */
 #define BMA180_CHIP_ID		0x00 /* Need to distinguish BMA180 from other */
 #define BMA180_ACC_X_LSB	0x02 /* First of 6 registers of accel data */
+#define BMA180_TEMP		0x08
 #define BMA180_CTRL_REG0	0x0d
 #define BMA180_RESET		0x10
 #define BMA180_BW_TCS		0x20
@@ -49,65 +83,81 @@
 #define BMA180_SMP_SKIP		BIT(0)
 
 /* Bit masks for registers bit fields */
-#define BMA180_RANGE		0x0e /* Range of measured accel values*/
+#define BMA180_RANGE		0x0e /* Range of measured accel values */
 #define BMA180_BW		0xf0 /* Accel bandwidth */
 #define BMA180_MODE_CONFIG	0x03 /* Config operation modes */
 
 /* We have to write this value in reset register to do soft reset */
 #define BMA180_RESET_VAL	0xb6
 
-#define BMA_180_ID_REG_VAL	0x03
+#define BMA180_ID_REG_VAL	0x03
 
 /* Chip power modes */
-#define BMA180_LOW_NOISE	0x00
 #define BMA180_LOW_POWER	0x03
 
-#define BMA180_LOW_NOISE_STR	"low_noise"
-#define BMA180_LOW_POWER_STR	"low_power"
-
-/* Defaults values */
-#define BMA180_DEF_PMODE	0
-#define BMA180_DEF_BW		20
-#define BMA180_DEF_SCALE	2452
-
-/* Available values for sysfs */
-#define BMA180_FLP_FREQ_AVAILABLE \
-	"10 20 40 75 150 300"
-#define BMA180_SCALE_AVAILABLE \
-	"0.001275 0.001863 0.002452 0.003727 0.004903 0.009709 0.019417"
+#define BMA250_RANGE_REG	0x0f
+#define BMA250_BW_REG		0x10
+#define BMA250_POWER_REG	0x11
+#define BMA250_RESET_REG	0x14
+#define BMA250_INT_ENABLE_REG	0x17
+#define BMA250_INT_MAP_REG	0x1a
+#define BMA250_INT_RESET_REG	0x21
+
+#define BMA250_RANGE_MASK	GENMASK(3, 0) /* Range of accel values */
+#define BMA250_BW_MASK		GENMASK(4, 0) /* Accel bandwidth */
+#define BMA250_SUSPEND_MASK	BIT(7) /* chip will sleep */
+#define BMA250_LOWPOWER_MASK	BIT(6)
+#define BMA250_DATA_INTEN_MASK	BIT(4)
+#define BMA250_INT1_DATA_MASK	BIT(0)
+#define BMA250_INT_RESET_MASK	BIT(7) /* Reset pending interrupts */
 
 struct bma180_data {
 	struct i2c_client *client;
 	struct iio_trigger *trig;
+	const struct bma180_part_info *part_info;
 	struct mutex mutex;
-	int sleep_state;
+	bool sleep_state;
 	int scale;
 	int bw;
-	int pmode;
-	char *buff;
+	bool pmode;
+	u8 buff[16]; /* 3x 16-bit + 8-bit + padding + timestamp */
 };
 
-enum bma180_axis {
+enum bma180_chan {
 	AXIS_X,
 	AXIS_Y,
 	AXIS_Z,
+	TEMP
 };
 
-static int bw_table[] = { 10, 20, 40, 75, 150, 300 }; /* Hz */
-static int scale_table[] = { 1275, 1863, 2452, 3727, 4903, 9709, 19417 };
+static int bma180_bw_table[] = { 10, 20, 40, 75, 150, 300 }; /* Hz */
+static int bma180_scale_table[] = { 1275, 1863, 2452, 3727, 4903, 9709, 19417 };
+
+static int bma250_bw_table[] = { 8, 16, 31, 63, 125, 250 }; /* Hz */
+static int bma250_scale_table[] = { 0, 0, 0, 38344, 0, 76590, 0, 0, 153180, 0,
+	0, 0, 306458 };
 
-static int bma180_get_acc_reg(struct bma180_data *data, enum bma180_axis axis)
+static int bma180_get_data_reg(struct bma180_data *data, enum bma180_chan chan)
 {
-	u8 reg = BMA180_ACC_X_LSB + axis * 2;
 	int ret;
 
 	if (data->sleep_state)
 		return -EBUSY;
 
-	ret = i2c_smbus_read_word_data(data->client, reg);
-	if (ret < 0)
-		dev_err(&data->client->dev,
-			"failed to read accel_%c registers\n", 'x' + axis);
+	switch (chan) {
+	case TEMP:
+		ret = i2c_smbus_read_byte_data(data->client, BMA180_TEMP);
+		if (ret < 0)
+			dev_err(&data->client->dev, "failed to read temp register\n");
+		break;
+	default:
+		ret = i2c_smbus_read_word_data(data->client,
+			BMA180_ACC_X_LSB + chan * 2);
+		if (ret < 0)
+			dev_err(&data->client->dev,
+				"failed to read accel_%c register\n",
+				'x' + chan);
+	}
 
 	return ret;
 }
@@ -125,7 +175,8 @@ static int bma180_set_bits(struct bma180_data *data, u8 reg, u8 mask, u8 val)
 
 static int bma180_reset_intr(struct bma180_data *data)
 {
-	int ret = bma180_set_bits(data, BMA180_CTRL_REG0, BMA180_RESET_INT, 1);
+	int ret = bma180_set_bits(data, data->part_info->int_reset_reg,
+		data->part_info->int_reset_mask, 1);
 
 	if (ret)
 		dev_err(&data->client->dev, "failed to reset interrupt\n");
@@ -133,12 +184,10 @@ static int bma180_reset_intr(struct bma180_data *data)
 	return ret;
 }
 
-static int bma180_set_new_data_intr_state(struct bma180_data *data, int state)
+static int bma180_set_new_data_intr_state(struct bma180_data *data, bool state)
 {
-	u8 reg_val = state ? BMA180_NEW_DATA_INT : 0x00;
-	int ret = i2c_smbus_write_byte_data(data->client, BMA180_CTRL_REG3,
-			reg_val);
-
+	int ret = bma180_set_bits(data, data->part_info->int_enable_reg,
+			data->part_info->int_enable_mask, state);
 	if (ret)
 		goto err;
 	ret = bma180_reset_intr(data);
@@ -153,9 +202,10 @@ err:
 	return ret;
 }
 
-static int bma180_set_sleep_state(struct bma180_data *data, int state)
+static int bma180_set_sleep_state(struct bma180_data *data, bool state)
 {
-	int ret = bma180_set_bits(data, BMA180_CTRL_REG0, BMA180_SLEEP, state);
+	int ret = bma180_set_bits(data, data->part_info->sleep_reg,
+		data->part_info->sleep_mask, state);
 
 	if (ret) {
 		dev_err(&data->client->dev,
@@ -167,7 +217,7 @@ static int bma180_set_sleep_state(struct bma180_data *data, int state)
 	return 0;
 }
 
-static int bma180_set_ee_writing_state(struct bma180_data *data, int state)
+static int bma180_set_ee_writing_state(struct bma180_data *data, bool state)
 {
 	int ret = bma180_set_bits(data, BMA180_CTRL_REG0, BMA180_EE_W, state);
 
@@ -185,10 +235,10 @@ static int bma180_set_bw(struct bma180_data *data, int val)
 	if (data->sleep_state)
 		return -EBUSY;
 
-	for (i = 0; i < ARRAY_SIZE(bw_table); ++i) {
-		if (bw_table[i] == val) {
-			ret = bma180_set_bits(data,
-					BMA180_BW_TCS, BMA180_BW, i);
+	for (i = 0; i < data->part_info->num_bw; ++i) {
+		if (data->part_info->bw_table[i] == val) {
+			ret = bma180_set_bits(data, data->part_info->bw_reg,
+				data->part_info->bw_mask, i);
 			if (ret) {
 				dev_err(&data->client->dev,
 					"failed to set bandwidth\n");
@@ -209,10 +259,10 @@ static int bma180_set_scale(struct bma180_data *data, int val)
 	if (data->sleep_state)
 		return -EBUSY;
 
-	for (i = 0; i < ARRAY_SIZE(scale_table); ++i)
-		if (scale_table[i] == val) {
-			ret = bma180_set_bits(data,
-					BMA180_OFFSET_LSB1, BMA180_RANGE, i);
+	for (i = 0; i < data->part_info->num_scales; ++i)
+		if (data->part_info->scale_table[i] == val) {
+			ret = bma180_set_bits(data, data->part_info->scale_reg,
+				data->part_info->scale_mask, i);
 			if (ret) {
 				dev_err(&data->client->dev,
 					"failed to set scale\n");
@@ -225,11 +275,11 @@ static int bma180_set_scale(struct bma180_data *data, int val)
 	return -EINVAL;
 }
 
-static int bma180_set_pmode(struct bma180_data *data, int mode)
+static int bma180_set_pmode(struct bma180_data *data, bool mode)
 {
-	u8 reg_val = mode ? BMA180_LOW_POWER : BMA180_LOW_NOISE;
-	int ret = bma180_set_bits(data, BMA180_TCO_Z, BMA180_MODE_CONFIG,
-			reg_val);
+	u8 reg_val = mode ? data->part_info->lowpower_val : 0;
+	int ret = bma180_set_bits(data, data->part_info->power_reg,
+		data->part_info->power_mask, reg_val);
 
 	if (ret) {
 		dev_err(&data->client->dev, "failed to set power mode\n");
@@ -243,7 +293,7 @@ static int bma180_set_pmode(struct bma180_data *data, int mode)
 static int bma180_soft_reset(struct bma180_data *data)
 {
 	int ret = i2c_smbus_write_byte_data(data->client,
-			BMA180_RESET, BMA180_RESET_VAL);
+		data->part_info->softreset_reg, BMA180_RESET_VAL);
 
 	if (ret)
 		dev_err(&data->client->dev, "failed to reset the chip\n");
@@ -257,57 +307,99 @@ static int bma180_chip_init(struct bma180_data *data)
 	int ret = i2c_smbus_read_byte_data(data->client, BMA180_CHIP_ID);
 
 	if (ret < 0)
-		goto err;
-	if (ret != BMA_180_ID_REG_VAL) {
-		ret = -ENODEV;
-		goto err;
-	}
+		return ret;
+	if (ret != BMA180_ID_REG_VAL)
+		return -ENODEV;
 
 	ret = bma180_soft_reset(data);
 	if (ret)
-		goto err;
+		return ret;
 	/*
 	 * No serial transaction should occur within minimum 10 us
 	 * after soft_reset command
 	 */
 	msleep(20);
 
-	ret = bma180_set_bits(data, BMA180_CTRL_REG0, BMA180_DIS_WAKE_UP, 1);
+	ret = bma180_set_new_data_intr_state(data, false);
+	if (ret)
+		return ret;
+
+	return bma180_set_pmode(data, false);
+}
+
+static int bma180_chip_config(struct bma180_data *data)
+{
+	int ret = bma180_chip_init(data);
+
 	if (ret)
 		goto err;
-	ret = bma180_set_ee_writing_state(data, 1);
+	ret = bma180_set_bits(data, BMA180_CTRL_REG0, BMA180_DIS_WAKE_UP, 1);
 	if (ret)
 		goto err;
-	ret = bma180_set_new_data_intr_state(data, 0);
+	ret = bma180_set_ee_writing_state(data, true);
 	if (ret)
 		goto err;
 	ret = bma180_set_bits(data, BMA180_OFFSET_LSB1, BMA180_SMP_SKIP, 1);
 	if (ret)
 		goto err;
-	ret = bma180_set_pmode(data, BMA180_DEF_PMODE);
+	ret = bma180_set_bw(data, 20); /* 20 Hz */
 	if (ret)
 		goto err;
-	ret = bma180_set_bw(data, BMA180_DEF_BW);
+	ret = bma180_set_scale(data, 2452); /* 2 G */
+	if (ret)
+		goto err;
+
+	return 0;
+
+err:
+	dev_err(&data->client->dev, "failed to config the chip\n");
+	return ret;
+}
+
+static int bma250_chip_config(struct bma180_data *data)
+{
+	int ret = bma180_chip_init(data);
+
 	if (ret)
 		goto err;
-	ret = bma180_set_scale(data, BMA180_DEF_SCALE);
+	ret = bma180_set_bw(data, 16); /* 16 Hz */
+	if (ret)
+		goto err;
+	ret = bma180_set_scale(data, 38344); /* 2 G */
+	if (ret)
+		goto err;
+	ret = bma180_set_bits(data, BMA250_INT_MAP_REG,
+		BMA250_INT1_DATA_MASK, 1);
 	if (ret)
 		goto err;
 
 	return 0;
 
 err:
-	dev_err(&data->client->dev, "failed to init the chip\n");
+	dev_err(&data->client->dev, "failed to config the chip\n");
 	return ret;
 }
 
 static void bma180_chip_disable(struct bma180_data *data)
 {
-	if (bma180_set_new_data_intr_state(data, 0))
+	if (bma180_set_new_data_intr_state(data, false))
 		goto err;
-	if (bma180_set_ee_writing_state(data, 0))
+	if (bma180_set_ee_writing_state(data, false))
+		goto err;
+	if (bma180_set_sleep_state(data, true))
+		goto err;
+
+	return;
+
+err:
+	dev_err(&data->client->dev, "failed to disable the chip\n");
+}
+
+static void bma250_chip_disable(struct bma180_data *data)
+{
+	if (bma180_set_new_data_intr_state(data, false))
 		goto err;
-	if (bma180_set_sleep_state(data, 1))
+	if (bma180_set_sleep_state(data, true))
 		goto err;
 
 	return;
@@ -316,13 +408,51 @@ err:
 	dev_err(&data->client->dev, "failed to disable the chip\n");
 }
 
-static IIO_CONST_ATTR(in_accel_filter_low_pass_3db_frequency_available,
-		BMA180_FLP_FREQ_AVAILABLE);
-static IIO_CONST_ATTR(in_accel_scale_available, BMA180_SCALE_AVAILABLE);
+static ssize_t bma180_show_avail(char *buf, const int *vals, unsigned n,
+				 bool micros)
+{
+	size_t len = 0;
+	int i;
+
+	for (i = 0; i < n; i++) {
+		if (!vals[i])
+			continue;
+		len += scnprintf(buf + len, PAGE_SIZE - len,
+			micros ? "0.%06d " : "%d ", vals[i]);
+	}
+	buf[len - 1] = '\n';
+
+	return len;
+}
+
+static ssize_t bma180_show_filter_freq_avail(struct device *dev,
+				struct device_attribute *attr, char *buf)
+{
+	struct bma180_data *data = iio_priv(dev_to_iio_dev(dev));
+
+	return bma180_show_avail(buf, data->part_info->bw_table,
+		data->part_info->num_bw, false);
+}
+
+static ssize_t bma180_show_scale_avail(struct device *dev,
+				struct device_attribute *attr, char *buf)
+{
+	struct bma180_data *data = iio_priv(dev_to_iio_dev(dev));
+
+	return bma180_show_avail(buf, data->part_info->scale_table,
+		data->part_info->num_scales, true);
+}
+
+static IIO_DEVICE_ATTR(in_accel_filter_low_pass_3db_frequency_available,
+	S_IRUGO, bma180_show_filter_freq_avail, NULL, 0);
+
+static IIO_DEVICE_ATTR(in_accel_scale_available,
+	S_IRUGO, bma180_show_scale_avail, NULL, 0);
 
 static struct attribute *bma180_attributes[] = {
-	&iio_const_attr_in_accel_filter_low_pass_3db_frequency_available.dev_attr.attr,
-	&iio_const_attr_in_accel_scale_available.dev_attr.attr,
+	&iio_dev_attr_in_accel_filter_low_pass_3db_frequency_available.
+		dev_attr.attr,
+	&iio_dev_attr_in_accel_scale_available.dev_attr.attr,
 	NULL,
 };
 
@@ -340,22 +470,35 @@ static int bma180_read_raw(struct iio_dev *indio_dev,
 	switch (mask) {
 	case IIO_CHAN_INFO_RAW:
 		mutex_lock(&data->mutex);
-		if (iio_buffer_enabled(indio_dev))
-			ret = -EBUSY;
-		else
-			ret = bma180_get_acc_reg(data, chan->scan_index);
+		if (iio_buffer_enabled(indio_dev)) {
+			mutex_unlock(&data->mutex);
+			return -EBUSY;
+		}
+		ret = bma180_get_data_reg(data, chan->scan_index);
 		mutex_unlock(&data->mutex);
 		if (ret < 0)
 			return ret;
-		*val = (s16)ret >> chan->scan_type.shift;
+		*val = sign_extend32(ret >> chan->scan_type.shift,
+			chan->scan_type.realbits - 1);
 		return IIO_VAL_INT;
 	case IIO_CHAN_INFO_LOW_PASS_FILTER_3DB_FREQUENCY:
 		*val = data->bw;
 		return IIO_VAL_INT;
 	case IIO_CHAN_INFO_SCALE:
-		*val = 0;
-		*val2 = data->scale;
-		return IIO_VAL_INT_PLUS_MICRO;
+		switch (chan->type) {
+		case IIO_ACCEL:
+			*val = 0;
+			*val2 = data->scale;
+			return IIO_VAL_INT_PLUS_MICRO;
+		case IIO_TEMP:
+			*val = 500;
+			return IIO_VAL_INT;
+		default:
+			return -EINVAL;
+		}
+	case IIO_CHAN_INFO_OFFSET:
+		*val = 48; /* 0 LSB @ 24 degree C */
+		return IIO_VAL_INT;
 	default:
 		return -EINVAL;
 	}
@@ -387,33 +530,14 @@ static int bma180_write_raw(struct iio_dev *indio_dev,
 	}
 }
 
-static int bma180_update_scan_mode(struct iio_dev *indio_dev,
-		const unsigned long *scan_mask)
-{
-	struct bma180_data *data = iio_priv(indio_dev);
-
-	if (data->buff)
-		devm_kfree(&indio_dev->dev, data->buff);
-	data->buff = devm_kzalloc(&indio_dev->dev,
-			indio_dev->scan_bytes, GFP_KERNEL);
-	if (!data->buff)
-		return -ENOMEM;
-
-	return 0;
-}
-
 static const struct iio_info bma180_info = {
 	.attrs			= &bma180_attrs_group,
 	.read_raw		= bma180_read_raw,
 	.write_raw		= bma180_write_raw,
-	.update_scan_mode	= bma180_update_scan_mode,
 	.driver_module		= THIS_MODULE,
 };
 
-static const char * const bma180_power_modes[] = {
-	BMA180_LOW_NOISE_STR,
-	BMA180_LOW_POWER_STR,
-};
+static const char * const bma180_power_modes[] = { "low_noise", "low_power" };
 
 static int bma180_get_power_mode(struct iio_dev *indio_dev,
 		const struct iio_chan_spec *chan)
@@ -449,7 +573,7 @@ static const struct iio_chan_spec_ext_info bma180_ext_info[] = {
 	{ },
 };
 
-#define BMA180_CHANNEL(_axis) {					\
+#define BMA180_ACC_CHANNEL(_axis, _bits) {				\
 	.type = IIO_ACCEL,						\
 	.modified = 1,							\
 	.channel2 = IIO_MOD_##_axis,					\
@@ -459,18 +583,70 @@ static const struct iio_chan_spec_ext_info bma180_ext_info[] = {
 	.scan_index = AXIS_##_axis,					\
 	.scan_type = {							\
 		.sign = 's',						\
-		.realbits = 14,						\
+		.realbits = _bits,					\
 		.storagebits = 16,					\
-		.shift = 2,						\
+		.shift = 16 - _bits,					\
 	},								\
 	.ext_info = bma180_ext_info,					\
 }
 
+#define BMA180_TEMP_CHANNEL {						\
+	.type = IIO_TEMP,						\
+	.info_mask_separate = BIT(IIO_CHAN_INFO_RAW) |			\
+		BIT(IIO_CHAN_INFO_SCALE) | BIT(IIO_CHAN_INFO_OFFSET),	\
+	.scan_index = TEMP,						\
+	.scan_type = {							\
+		.sign = 's',						\
+		.realbits = 8,						\
+		.storagebits = 16,					\
+	},								\
+}
+
 static const struct iio_chan_spec bma180_channels[] = {
-	BMA180_CHANNEL(X),
-	BMA180_CHANNEL(Y),
-	BMA180_CHANNEL(Z),
-	IIO_CHAN_SOFT_TIMESTAMP(3),
+	BMA180_ACC_CHANNEL(X, 14),
+	BMA180_ACC_CHANNEL(Y, 14),
+	BMA180_ACC_CHANNEL(Z, 14),
+	BMA180_TEMP_CHANNEL,
+	IIO_CHAN_SOFT_TIMESTAMP(4),
+};
+
+static const struct iio_chan_spec bma250_channels[] = {
+	BMA180_ACC_CHANNEL(X, 10),
+	BMA180_ACC_CHANNEL(Y, 10),
+	BMA180_ACC_CHANNEL(Z, 10),
+	BMA180_TEMP_CHANNEL,
+	IIO_CHAN_SOFT_TIMESTAMP(4),
+};
+
+static const struct bma180_part_info bma180_part_info[] = {
+	[BMA180] = {
+		bma180_channels, ARRAY_SIZE(bma180_channels),
+		bma180_scale_table, ARRAY_SIZE(bma180_scale_table),
+		bma180_bw_table, ARRAY_SIZE(bma180_bw_table),
+		BMA180_CTRL_REG0, BMA180_RESET_INT,
+		BMA180_CTRL_REG0, BMA180_SLEEP,
+		BMA180_BW_TCS, BMA180_BW,
+		BMA180_OFFSET_LSB1, BMA180_RANGE,
+		BMA180_TCO_Z, BMA180_MODE_CONFIG, BMA180_LOW_POWER,
+		BMA180_CTRL_REG3, BMA180_NEW_DATA_INT,
+		BMA180_RESET,
+		bma180_chip_config,
+		bma180_chip_disable,
+	},
+	[BMA250] = {
+		bma250_channels, ARRAY_SIZE(bma250_channels),
+		bma250_scale_table, ARRAY_SIZE(bma250_scale_table),
+		bma250_bw_table, ARRAY_SIZE(bma250_bw_table),
+		BMA250_INT_RESET_REG, BMA250_INT_RESET_MASK,
+		BMA250_POWER_REG, BMA250_SUSPEND_MASK,
+		BMA250_BW_REG, BMA250_BW_MASK,
+		BMA250_RANGE_REG, BMA250_RANGE_MASK,
+		BMA250_POWER_REG, BMA250_LOWPOWER_MASK, 1,
+		BMA250_INT_ENABLE_REG, BMA250_DATA_INTEN_MASK,
+		BMA250_RESET_REG,
+		bma250_chip_config,
+		bma250_chip_disable,
+	},
 };
 
 static irqreturn_t bma180_trigger_handler(int irq, void *p)
@@ -485,13 +661,14 @@ static irqreturn_t bma180_trigger_handler(int irq, void *p)
 
 	for_each_set_bit(bit, indio_dev->buffer->scan_mask,
 			 indio_dev->masklength) {
-		ret = bma180_get_acc_reg(data, bit);
+		ret = bma180_get_data_reg(data, bit);
 		if (ret < 0) {
 			mutex_unlock(&data->mutex);
 			goto err;
 		}
 		((s16 *)data->buff)[i++] = ret;
 	}
+
 	mutex_unlock(&data->mutex);
 
 	iio_push_to_buffers_with_timestamp(indio_dev, data->buff, time_ns);
@@ -529,7 +706,6 @@ static int bma180_probe(struct i2c_client *client,
 {
 	struct bma180_data *data;
 	struct iio_dev *indio_dev;
-	struct iio_trigger *trig;
 	int ret;
 
 	indio_dev = devm_iio_device_alloc(&client->dev, sizeof(*data));
@@ -539,43 +715,45 @@ static int bma180_probe(struct i2c_client *client,
 	data = iio_priv(indio_dev);
 	i2c_set_clientdata(client, indio_dev);
 	data->client = client;
+	data->part_info = &bma180_part_info[id->driver_data];
 
-	ret = bma180_chip_init(data);
+	ret = data->part_info->chip_config(data);
 	if (ret < 0)
 		goto err_chip_disable;
 
 	mutex_init(&data->mutex);
-
 	indio_dev->dev.parent = &client->dev;
-	indio_dev->channels = bma180_channels;
-	indio_dev->num_channels = ARRAY_SIZE(bma180_channels);
-	indio_dev->name = BMA180_DRV_NAME;
+	indio_dev->channels = data->part_info->channels;
+	indio_dev->num_channels = data->part_info->num_channels;
+	indio_dev->name = id->name;
 	indio_dev->modes = INDIO_DIRECT_MODE;
 	indio_dev->info = &bma180_info;
 
-	trig = iio_trigger_alloc("%s-dev%d", indio_dev->name, indio_dev->id);
-	if (!trig) {
-		ret = -ENOMEM;
-		goto err_chip_disable;
-	}
+	if (client->irq > 0) {
+		data->trig = iio_trigger_alloc("%s-dev%d", indio_dev->name,
+			indio_dev->id);
+		if (!data->trig) {
+			ret = -ENOMEM;
+			goto err_chip_disable;
+		}
 
-	ret = devm_request_irq(&client->dev, client->irq,
-			iio_trigger_generic_data_rdy_poll,
-			IRQF_TRIGGER_RISING, BMA180_IRQ_NAME, trig);
-	if (ret) {
-		dev_err(&client->dev, "unable to request IRQ\n");
-		goto err_trigger_free;
-	}
+		ret = devm_request_irq(&client->dev, client->irq,
+			iio_trigger_generic_data_rdy_poll, IRQF_TRIGGER_RISING,
+			"bma180_event", data->trig);
+		if (ret) {
+			dev_err(&client->dev, "unable to request IRQ\n");
+			goto err_trigger_free;
+		}
 
-	trig->dev.parent = &client->dev;
-	trig->ops = &bma180_trigger_ops;
-	iio_trigger_set_drvdata(trig, indio_dev);
-	data->trig = trig;
-	indio_dev->trig = iio_trigger_get(trig);
+		data->trig->dev.parent = &client->dev;
+		data->trig->ops = &bma180_trigger_ops;
+		iio_trigger_set_drvdata(data->trig, indio_dev);
+		indio_dev->trig = iio_trigger_get(data->trig);
 
-	ret = iio_trigger_register(trig);
-	if (ret)
-		goto err_trigger_free;
+		ret = iio_trigger_register(data->trig);
+		if (ret)
+			goto err_trigger_free;
+	}
 
 	ret = iio_triggered_buffer_setup(indio_dev, NULL,
 			bma180_trigger_handler, NULL);
@@ -595,11 +773,12 @@ static int bma180_probe(struct i2c_client *client,
 err_buffer_cleanup:
 	iio_triggered_buffer_cleanup(indio_dev);
 err_trigger_unregister:
-	iio_trigger_unregister(trig);
+	if (data->trig)
+		iio_trigger_unregister(data->trig);
 err_trigger_free:
-	iio_trigger_free(trig);
+	iio_trigger_free(data->trig);
 err_chip_disable:
-	bma180_chip_disable(data);
+	data->part_info->chip_disable(data);
 
 	return ret;
 }
@@ -611,11 +790,13 @@ static int bma180_remove(struct i2c_client *client)
 
 	iio_device_unregister(indio_dev);
 	iio_triggered_buffer_cleanup(indio_dev);
-	iio_trigger_unregister(data->trig);
-	iio_trigger_free(data->trig);
+	if (data->trig) {
+		iio_trigger_unregister(data->trig);
+		iio_trigger_free(data->trig);
+	}
 
 	mutex_lock(&data->mutex);
-	bma180_chip_disable(data);
+	data->part_info->chip_disable(data);
 	mutex_unlock(&data->mutex);
 
 	return 0;
@@ -629,7 +810,7 @@ static int bma180_suspend(struct device *dev)
 	int ret;
 
 	mutex_lock(&data->mutex);
-	ret = bma180_set_sleep_state(data, 1);
+	ret = bma180_set_sleep_state(data, true);
 	mutex_unlock(&data->mutex);
 
 	return ret;
@@ -642,7 +823,7 @@ static int bma180_resume(struct device *dev)
 	int ret;
 
 	mutex_lock(&data->mutex);
-	ret = bma180_set_sleep_state(data, 0);
+	ret = bma180_set_sleep_state(data, false);
 	mutex_unlock(&data->mutex);
 
 	return ret;
@@ -654,27 +835,28 @@ static SIMPLE_DEV_PM_OPS(bma180_pm_ops, bma180_suspend, bma180_resume);
 #define BMA180_PM_OPS NULL
 #endif
 
-static struct i2c_device_id bma180_id[] = {
-	{ BMA180_DRV_NAME, 0 },
+static struct i2c_device_id bma180_ids[] = {
+	{ "bma180", BMA180 },
+	{ "bma250", BMA250 },
 	{ }
 };
 
-MODULE_DEVICE_TABLE(i2c, bma180_id);
+MODULE_DEVICE_TABLE(i2c, bma180_ids);
 
 static struct i2c_driver bma180_driver = {
 	.driver = {
-		.name	= BMA180_DRV_NAME,
+		.name	= "bma180",
 		.owner	= THIS_MODULE,
 		.pm	= BMA180_PM_OPS,
 	},
 	.probe		= bma180_probe,
 	.remove		= bma180_remove,
-	.id_table	= bma180_id,
+	.id_table	= bma180_ids,
 };
 
 module_i2c_driver(bma180_driver);
 
 MODULE_AUTHOR("Kravchenko Oleksandr <x0199363@ti.com>");
 MODULE_AUTHOR("Texas Instruments, Inc.");
-MODULE_DESCRIPTION("Bosch BMA180 triaxial acceleration sensor");
+MODULE_DESCRIPTION("Bosch BMA180/BMA250 triaxial acceleration sensor");
 MODULE_LICENSE("GPL");
diff --git a/drivers/iio/accel/bmc150-accel.c b/drivers/iio/accel/bmc150-accel.c
new file mode 100644
index 000000000000..066d0c04072c
--- /dev/null
+++ b/drivers/iio/accel/bmc150-accel.c
@@ -0,0 +1,1456 @@
+/*
+ * 3-axis accelerometer driver supporting following Bosch-Sensortec chips:
+ *  - BMC150
+ *  - BMI055
+ *  - BMA255
+ *  - BMA250E
+ *  - BMA222E
+ *  - BMA280
+ *
+ * Copyright (c) 2014, Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ */
+
+#include <linux/module.h>
+#include <linux/i2c.h>
+#include <linux/interrupt.h>
+#include <linux/delay.h>
+#include <linux/slab.h>
+#include <linux/acpi.h>
+#include <linux/gpio/consumer.h>
+#include <linux/pm.h>
+#include <linux/pm_runtime.h>
+#include <linux/iio/iio.h>
+#include <linux/iio/sysfs.h>
+#include <linux/iio/buffer.h>
+#include <linux/iio/events.h>
+#include <linux/iio/trigger.h>
+#include <linux/iio/trigger_consumer.h>
+#include <linux/iio/triggered_buffer.h>
+
+#define BMC150_ACCEL_DRV_NAME			"bmc150_accel"
+#define BMC150_ACCEL_IRQ_NAME			"bmc150_accel_event"
+#define BMC150_ACCEL_GPIO_NAME			"bmc150_accel_int"
+
+#define BMC150_ACCEL_REG_CHIP_ID		0x00
+
+#define BMC150_ACCEL_REG_INT_STATUS_2		0x0B
+#define BMC150_ACCEL_ANY_MOTION_MASK		0x07
+#define BMC150_ACCEL_ANY_MOTION_BIT_X		BIT(0)
+#define BMC150_ACCEL_ANY_MOTION_BIT_Y		BIT(1)
+#define BMC150_ACCEL_ANY_MOTION_BIT_Z		BIT(2)
+#define BMC150_ACCEL_ANY_MOTION_BIT_SIGN	BIT(3)
+
+#define BMC150_ACCEL_REG_PMU_LPW		0x11
+#define BMC150_ACCEL_PMU_MODE_MASK		0xE0
+#define BMC150_ACCEL_PMU_MODE_SHIFT		5
+#define BMC150_ACCEL_PMU_BIT_SLEEP_DUR_MASK	0x17
+#define BMC150_ACCEL_PMU_BIT_SLEEP_DUR_SHIFT	1
+
+#define BMC150_ACCEL_REG_PMU_RANGE		0x0F
+
+#define BMC150_ACCEL_DEF_RANGE_2G		0x03
+#define BMC150_ACCEL_DEF_RANGE_4G		0x05
+#define BMC150_ACCEL_DEF_RANGE_8G		0x08
+#define BMC150_ACCEL_DEF_RANGE_16G		0x0C
+
+/* Default BW: 125Hz */
+#define BMC150_ACCEL_REG_PMU_BW		0x10
+#define BMC150_ACCEL_DEF_BW			125
+
+#define BMC150_ACCEL_REG_INT_MAP_0		0x19
+#define BMC150_ACCEL_INT_MAP_0_BIT_SLOPE	BIT(2)
+
+#define BMC150_ACCEL_REG_INT_MAP_1		0x1A
+#define BMC150_ACCEL_INT_MAP_1_BIT_DATA	BIT(0)
+
+#define BMC150_ACCEL_REG_INT_RST_LATCH		0x21
+#define BMC150_ACCEL_INT_MODE_LATCH_RESET	0x80
+#define BMC150_ACCEL_INT_MODE_LATCH_INT	0x0F
+#define BMC150_ACCEL_INT_MODE_NON_LATCH_INT	0x00
+
+#define BMC150_ACCEL_REG_INT_EN_0		0x16
+#define BMC150_ACCEL_INT_EN_BIT_SLP_X		BIT(0)
+#define BMC150_ACCEL_INT_EN_BIT_SLP_Y		BIT(1)
+#define BMC150_ACCEL_INT_EN_BIT_SLP_Z		BIT(2)
+
+#define BMC150_ACCEL_REG_INT_EN_1		0x17
+#define BMC150_ACCEL_INT_EN_BIT_DATA_EN	BIT(4)
+
+#define BMC150_ACCEL_REG_INT_OUT_CTRL		0x20
+#define BMC150_ACCEL_INT_OUT_CTRL_INT1_LVL	BIT(0)
+
+#define BMC150_ACCEL_REG_INT_5			0x27
+#define BMC150_ACCEL_SLOPE_DUR_MASK		0x03
+
+#define BMC150_ACCEL_REG_INT_6			0x28
+#define BMC150_ACCEL_SLOPE_THRES_MASK		0xFF
+
+/* Slope duration in terms of number of samples */
+#define BMC150_ACCEL_DEF_SLOPE_DURATION		1
+/* in terms of multiples of g's/LSB, based on range */
+#define BMC150_ACCEL_DEF_SLOPE_THRESHOLD	1
+
+#define BMC150_ACCEL_REG_XOUT_L		0x02
+
+#define BMC150_ACCEL_MAX_STARTUP_TIME_MS	100
+
+/* Sleep Duration values */
+#define BMC150_ACCEL_SLEEP_500_MICRO		0x05
+#define BMC150_ACCEL_SLEEP_1_MS		0x06
+#define BMC150_ACCEL_SLEEP_2_MS		0x07
+#define BMC150_ACCEL_SLEEP_4_MS		0x08
+#define BMC150_ACCEL_SLEEP_6_MS		0x09
+#define BMC150_ACCEL_SLEEP_10_MS		0x0A
+#define BMC150_ACCEL_SLEEP_25_MS		0x0B
+#define BMC150_ACCEL_SLEEP_50_MS		0x0C
+#define BMC150_ACCEL_SLEEP_100_MS		0x0D
+#define BMC150_ACCEL_SLEEP_500_MS		0x0E
+#define BMC150_ACCEL_SLEEP_1_SEC		0x0F
+
+#define BMC150_ACCEL_REG_TEMP			0x08
+#define BMC150_ACCEL_TEMP_CENTER_VAL		24
+
+#define BMC150_ACCEL_AXIS_TO_REG(axis)	(BMC150_ACCEL_REG_XOUT_L + (axis * 2))
+#define BMC150_AUTO_SUSPEND_DELAY_MS		2000
+
+enum bmc150_accel_axis {
+	AXIS_X,
+	AXIS_Y,
+	AXIS_Z,
+};
+
+enum bmc150_power_modes {
+	BMC150_ACCEL_SLEEP_MODE_NORMAL,
+	BMC150_ACCEL_SLEEP_MODE_DEEP_SUSPEND,
+	BMC150_ACCEL_SLEEP_MODE_LPM,
+	BMC150_ACCEL_SLEEP_MODE_SUSPEND = 0x04,
+};
+
+struct bmc150_scale_info {
+	int scale;
+	u8 reg_range;
+};
+
+struct bmc150_accel_chip_info {
+	u8 chip_id;
+	const struct iio_chan_spec *channels;
+	int num_channels;
+	const struct bmc150_scale_info scale_table[4];
+};
+
+struct bmc150_accel_data {
+	struct i2c_client *client;
+	struct iio_trigger *dready_trig;
+	struct iio_trigger *motion_trig;
+	struct mutex mutex;
+	s16 buffer[8];
+	u8 bw_bits;
+	u32 slope_dur;
+	u32 slope_thres;
+	u32 range;
+	int ev_enable_state;
+	bool dready_trigger_on;
+	bool motion_trigger_on;
+	int64_t timestamp;
+	const struct bmc150_accel_chip_info *chip_info;
+};
+
+static const struct {
+	int val;
+	int val2;
+	u8 bw_bits;
+} bmc150_accel_samp_freq_table[] = { {7, 810000, 0x08},
+				     {15, 630000, 0x09},
+				     {31, 250000, 0x0A},
+				     {62, 500000, 0x0B},
+				     {125, 0, 0x0C},
+				     {250, 0, 0x0D},
+				     {500, 0, 0x0E},
+				     {1000, 0, 0x0F} };
+
+static const struct {
+	int bw_bits;
+	int msec;
+} bmc150_accel_sample_upd_time[] = { {0x08, 64},
+				     {0x09, 32},
+				     {0x0A, 16},
+				     {0x0B, 8},
+				     {0x0C, 4},
+				     {0x0D, 2},
+				     {0x0E, 1},
+				     {0x0F, 1} };
+
+static const struct {
+	int sleep_dur;
+	u8 reg_value;
+} bmc150_accel_sleep_value_table[] = { {0, 0},
+				       {500, BMC150_ACCEL_SLEEP_500_MICRO},
+				       {1000, BMC150_ACCEL_SLEEP_1_MS},
+				       {2000, BMC150_ACCEL_SLEEP_2_MS},
+				       {4000, BMC150_ACCEL_SLEEP_4_MS},
+				       {6000, BMC150_ACCEL_SLEEP_6_MS},
+				       {10000, BMC150_ACCEL_SLEEP_10_MS},
+				       {25000, BMC150_ACCEL_SLEEP_25_MS},
+				       {50000, BMC150_ACCEL_SLEEP_50_MS},
+				       {100000, BMC150_ACCEL_SLEEP_100_MS},
+				       {500000, BMC150_ACCEL_SLEEP_500_MS},
+				       {1000000, BMC150_ACCEL_SLEEP_1_SEC} };
+
+
+static int bmc150_accel_set_mode(struct bmc150_accel_data *data,
+				 enum bmc150_power_modes mode,
+				 int dur_us)
+{
+	int i;
+	int ret;
+	u8 lpw_bits;
+	int dur_val = -1;
+
+	if (dur_us > 0) {
+		for (i = 0; i < ARRAY_SIZE(bmc150_accel_sleep_value_table);
+									 ++i) {
+			if (bmc150_accel_sleep_value_table[i].sleep_dur ==
+									dur_us)
+				dur_val =
+				bmc150_accel_sleep_value_table[i].reg_value;
+		}
+	} else
+		dur_val = 0;
+
+	if (dur_val < 0)
+		return -EINVAL;
+
+	lpw_bits = mode << BMC150_ACCEL_PMU_MODE_SHIFT;
+	lpw_bits |= (dur_val << BMC150_ACCEL_PMU_BIT_SLEEP_DUR_SHIFT);
+
+	dev_dbg(&data->client->dev, "Set Mode bits %x\n", lpw_bits);
+
+	ret = i2c_smbus_write_byte_data(data->client,
+					BMC150_ACCEL_REG_PMU_LPW, lpw_bits);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_pmu_lpw\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int bmc150_accel_set_bw(struct bmc150_accel_data *data, int val,
+			       int val2)
+{
+	int i;
+	int ret;
+
+	for (i = 0; i < ARRAY_SIZE(bmc150_accel_samp_freq_table); ++i) {
+		if (bmc150_accel_samp_freq_table[i].val == val &&
+				bmc150_accel_samp_freq_table[i].val2 == val2) {
+			ret = i2c_smbus_write_byte_data(
+				data->client,
+				BMC150_ACCEL_REG_PMU_BW,
+				bmc150_accel_samp_freq_table[i].bw_bits);
+			if (ret < 0)
+				return ret;
+
+			data->bw_bits =
+				bmc150_accel_samp_freq_table[i].bw_bits;
+			return 0;
+		}
+	}
+
+	return -EINVAL;
+}
+
+static int bmc150_accel_chip_init(struct bmc150_accel_data *data)
+{
+	int ret;
+
+	ret = i2c_smbus_read_byte_data(data->client, BMC150_ACCEL_REG_CHIP_ID);
+	if (ret < 0) {
+		dev_err(&data->client->dev,
+			"Error: Reading chip id\n");
+		return ret;
+	}
+
+	dev_dbg(&data->client->dev, "Chip Id %x\n", ret);
+	if (ret != data->chip_info->chip_id) {
+		dev_err(&data->client->dev, "Invalid chip %x\n", ret);
+		return -ENODEV;
+	}
+
+	ret = bmc150_accel_set_mode(data, BMC150_ACCEL_SLEEP_MODE_NORMAL, 0);
+	if (ret < 0)
+		return ret;
+
+	/* Set Bandwidth */
+	ret = bmc150_accel_set_bw(data, BMC150_ACCEL_DEF_BW, 0);
+	if (ret < 0)
+		return ret;
+
+	/* Set Default Range */
+	ret = i2c_smbus_write_byte_data(data->client,
+					BMC150_ACCEL_REG_PMU_RANGE,
+					BMC150_ACCEL_DEF_RANGE_4G);
+	if (ret < 0) {
+		dev_err(&data->client->dev,
+					"Error writing reg_pmu_range\n");
+		return ret;
+	}
+
+	data->range = BMC150_ACCEL_DEF_RANGE_4G;
+
+	/* Set default slope duration */
+	ret = i2c_smbus_read_byte_data(data->client, BMC150_ACCEL_REG_INT_5);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading reg_int_5\n");
+		return ret;
+	}
+	data->slope_dur |= BMC150_ACCEL_DEF_SLOPE_DURATION;
+	ret = i2c_smbus_write_byte_data(data->client,
+					BMC150_ACCEL_REG_INT_5,
+					data->slope_dur);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_int_5\n");
+		return ret;
+	}
+	dev_dbg(&data->client->dev, "slope_dur %x\n", data->slope_dur);
+
+	/* Set default slope thresholds */
+	ret = i2c_smbus_write_byte_data(data->client,
+					BMC150_ACCEL_REG_INT_6,
+					BMC150_ACCEL_DEF_SLOPE_THRESHOLD);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_int_6\n");
+		return ret;
+	}
+	data->slope_thres = BMC150_ACCEL_DEF_SLOPE_THRESHOLD;
+	dev_dbg(&data->client->dev, "slope_thres %x\n", data->slope_thres);
+
+	/* Set default as latched interrupts */
+	ret = i2c_smbus_write_byte_data(data->client,
+					BMC150_ACCEL_REG_INT_RST_LATCH,
+					BMC150_ACCEL_INT_MODE_LATCH_INT |
+					BMC150_ACCEL_INT_MODE_LATCH_RESET);
+	if (ret < 0) {
+		dev_err(&data->client->dev,
+			"Error writing reg_int_rst_latch\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int bmc150_accel_setup_any_motion_interrupt(
+					struct bmc150_accel_data *data,
+					bool status)
+{
+	int ret;
+
+	/* Enable/Disable INT1 mapping */
+	ret = i2c_smbus_read_byte_data(data->client,
+				       BMC150_ACCEL_REG_INT_MAP_0);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading reg_int_map_0\n");
+		return ret;
+	}
+	if (status)
+		ret |= BMC150_ACCEL_INT_MAP_0_BIT_SLOPE;
+	else
+		ret &= ~BMC150_ACCEL_INT_MAP_0_BIT_SLOPE;
+
+	ret = i2c_smbus_write_byte_data(data->client,
+					BMC150_ACCEL_REG_INT_MAP_0,
+					ret);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_int_map_0\n");
+		return ret;
+	}
+
+	if (status) {
+		/* Set slope duration (no of samples) */
+		ret = i2c_smbus_write_byte_data(data->client,
+						BMC150_ACCEL_REG_INT_5,
+						data->slope_dur);
+		if (ret < 0) {
+			dev_err(&data->client->dev, "Error write reg_int_5\n");
+			return ret;
+		}
+
+		/* Set slope thresholds */
+		ret = i2c_smbus_write_byte_data(data->client,
+						BMC150_ACCEL_REG_INT_6,
+						data->slope_thres);
+		if (ret < 0) {
+			dev_err(&data->client->dev, "Error write reg_int_6\n");
+			return ret;
+		}
+
+		/*
+		 * New data interrupt is always non-latched,
+		 * which will have higher priority, so no need
+		 * to set latched mode, we will be flooded anyway with INTR
+		 */
+		if (!data->dready_trigger_on) {
+			ret = i2c_smbus_write_byte_data(data->client,
+					BMC150_ACCEL_REG_INT_RST_LATCH,
+					BMC150_ACCEL_INT_MODE_LATCH_INT |
+					BMC150_ACCEL_INT_MODE_LATCH_RESET);
+			if (ret < 0) {
+				dev_err(&data->client->dev,
+					"Error writing reg_int_rst_latch\n");
+				return ret;
+			}
+		}
+
+		ret = i2c_smbus_write_byte_data(data->client,
+						BMC150_ACCEL_REG_INT_EN_0,
+						BMC150_ACCEL_INT_EN_BIT_SLP_X |
+						BMC150_ACCEL_INT_EN_BIT_SLP_Y |
+						BMC150_ACCEL_INT_EN_BIT_SLP_Z);
+	} else
+		ret = i2c_smbus_write_byte_data(data->client,
+						BMC150_ACCEL_REG_INT_EN_0,
+						0);
+
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_int_en_0\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int bmc150_accel_setup_new_data_interrupt(struct bmc150_accel_data *data,
+					   bool status)
+{
+	int ret;
+
+	/* Enable/Disable INT1 mapping */
+	ret = i2c_smbus_read_byte_data(data->client,
+				       BMC150_ACCEL_REG_INT_MAP_1);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading reg_int_map_1\n");
+		return ret;
+	}
+	if (status)
+		ret |= BMC150_ACCEL_INT_MAP_1_BIT_DATA;
+	else
+		ret &= ~BMC150_ACCEL_INT_MAP_1_BIT_DATA;
+
+	ret = i2c_smbus_write_byte_data(data->client,
+					BMC150_ACCEL_REG_INT_MAP_1,
+					ret);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_int_map_1\n");
+		return ret;
+	}
+
+	if (status) {
+		/*
+		 * Set non latched mode interrupt and clear any latched
+		 * interrupt
+		 */
+		ret = i2c_smbus_write_byte_data(data->client,
+					BMC150_ACCEL_REG_INT_RST_LATCH,
+					BMC150_ACCEL_INT_MODE_NON_LATCH_INT |
+					BMC150_ACCEL_INT_MODE_LATCH_RESET);
+		if (ret < 0) {
+			dev_err(&data->client->dev,
+				"Error writing reg_int_rst_latch\n");
+			return ret;
+		}
+
+		ret = i2c_smbus_write_byte_data(data->client,
+					BMC150_ACCEL_REG_INT_EN_1,
+					BMC150_ACCEL_INT_EN_BIT_DATA_EN);
+
+	} else {
+		/* Restore default interrupt mode */
+		ret = i2c_smbus_write_byte_data(data->client,
+					BMC150_ACCEL_REG_INT_RST_LATCH,
+					BMC150_ACCEL_INT_MODE_LATCH_INT |
+					BMC150_ACCEL_INT_MODE_LATCH_RESET);
+		if (ret < 0) {
+			dev_err(&data->client->dev,
+				"Error writing reg_int_rst_latch\n");
+			return ret;
+		}
+
+		ret = i2c_smbus_write_byte_data(data->client,
+						BMC150_ACCEL_REG_INT_EN_1,
+						0);
+	}
+
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_int_en_1\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int bmc150_accel_get_bw(struct bmc150_accel_data *data, int *val,
+			       int *val2)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(bmc150_accel_samp_freq_table); ++i) {
+		if (bmc150_accel_samp_freq_table[i].bw_bits == data->bw_bits) {
+			*val = bmc150_accel_samp_freq_table[i].val;
+			*val2 = bmc150_accel_samp_freq_table[i].val2;
+			return IIO_VAL_INT_PLUS_MICRO;
+		}
+	}
+
+	return -EINVAL;
+}
+
+#ifdef CONFIG_PM
+static int bmc150_accel_get_startup_times(struct bmc150_accel_data *data)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(bmc150_accel_sample_upd_time); ++i) {
+		if (bmc150_accel_sample_upd_time[i].bw_bits == data->bw_bits)
+			return bmc150_accel_sample_upd_time[i].msec;
+	}
+
+	return BMC150_ACCEL_MAX_STARTUP_TIME_MS;
+}
+
+static int bmc150_accel_set_power_state(struct bmc150_accel_data *data, bool on)
+{
+	int ret;
+
+	if (on)
+		ret = pm_runtime_get_sync(&data->client->dev);
+	else {
+		pm_runtime_mark_last_busy(&data->client->dev);
+		ret = pm_runtime_put_autosuspend(&data->client->dev);
+	}
+	if (ret < 0) {
+		dev_err(&data->client->dev,
+			"Failed: bmc150_accel_set_power_state for %d\n", on);
+		if (on)
+			pm_runtime_put_noidle(&data->client->dev);
+
+		return ret;
+	}
+
+	return 0;
+}
+#else
+static int bmc150_accel_set_power_state(struct bmc150_accel_data *data, bool on)
+{
+	return 0;
+}
+#endif
+
+static int bmc150_accel_set_scale(struct bmc150_accel_data *data, int val)
+{
+	int ret, i;
+
+	for (i = 0; i < ARRAY_SIZE(data->chip_info->scale_table); ++i) {
+		if (data->chip_info->scale_table[i].scale == val) {
+			ret = i2c_smbus_write_byte_data(
+				     data->client,
+				     BMC150_ACCEL_REG_PMU_RANGE,
+				     data->chip_info->scale_table[i].reg_range);
+			if (ret < 0) {
+				dev_err(&data->client->dev,
+					"Error writing pmu_range\n");
+				return ret;
+			}
+
+			data->range = data->chip_info->scale_table[i].reg_range;
+			return 0;
+		}
+	}
+
+	return -EINVAL;
+}
+
+static int bmc150_accel_get_temp(struct bmc150_accel_data *data, int *val)
+{
+	int ret;
+
+	mutex_lock(&data->mutex);
+
+	ret = i2c_smbus_read_byte_data(data->client, BMC150_ACCEL_REG_TEMP);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading reg_temp\n");
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+	*val = sign_extend32(ret, 7);
+
+	mutex_unlock(&data->mutex);
+
+	return IIO_VAL_INT;
+}
+
+static int bmc150_accel_get_axis(struct bmc150_accel_data *data,
+				 struct iio_chan_spec const *chan,
+				 int *val)
+{
+	int ret;
+	int axis = chan->scan_index;
+
+	mutex_lock(&data->mutex);
+	ret = bmc150_accel_set_power_state(data, true);
+	if (ret < 0) {
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+
+	ret = i2c_smbus_read_word_data(data->client,
+				       BMC150_ACCEL_AXIS_TO_REG(axis));
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading axis %d\n", axis);
+		bmc150_accel_set_power_state(data, false);
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+	*val = sign_extend32(ret >> chan->scan_type.shift,
+			     chan->scan_type.realbits - 1);
+	ret = bmc150_accel_set_power_state(data, false);
+	mutex_unlock(&data->mutex);
+	if (ret < 0)
+		return ret;
+
+	return IIO_VAL_INT;
+}
+
+static int bmc150_accel_read_raw(struct iio_dev *indio_dev,
+				 struct iio_chan_spec const *chan,
+				 int *val, int *val2, long mask)
+{
+	struct bmc150_accel_data *data = iio_priv(indio_dev);
+	int ret;
+
+	switch (mask) {
+	case IIO_CHAN_INFO_RAW:
+		switch (chan->type) {
+		case IIO_TEMP:
+			return bmc150_accel_get_temp(data, val);
+		case IIO_ACCEL:
+			if (iio_buffer_enabled(indio_dev))
+				return -EBUSY;
+			else
+				return bmc150_accel_get_axis(data, chan, val);
+		default:
+			return -EINVAL;
+		}
+	case IIO_CHAN_INFO_OFFSET:
+		if (chan->type == IIO_TEMP) {
+			*val = BMC150_ACCEL_TEMP_CENTER_VAL;
+			return IIO_VAL_INT;
+		} else
+			return -EINVAL;
+	case IIO_CHAN_INFO_SCALE:
+		*val = 0;
+		switch (chan->type) {
+		case IIO_TEMP:
+			*val2 = 500000;
+			return IIO_VAL_INT_PLUS_MICRO;
+		case IIO_ACCEL:
+		{
+			int i;
+			const struct bmc150_scale_info *si;
+			int st_size = ARRAY_SIZE(data->chip_info->scale_table);
+
+			for (i = 0; i < st_size; ++i) {
+				si = &data->chip_info->scale_table[i];
+				if (si->reg_range == data->range) {
+					*val2 = si->scale;
+					return IIO_VAL_INT_PLUS_MICRO;
+				}
+			}
+			return -EINVAL;
+		}
+		default:
+			return -EINVAL;
+		}
+	case IIO_CHAN_INFO_SAMP_FREQ:
+		mutex_lock(&data->mutex);
+		ret = bmc150_accel_get_bw(data, val, val2);
+		mutex_unlock(&data->mutex);
+		return ret;
+	default:
+		return -EINVAL;
+	}
+}
+
+static int bmc150_accel_write_raw(struct iio_dev *indio_dev,
+				  struct iio_chan_spec const *chan,
+				  int val, int val2, long mask)
+{
+	struct bmc150_accel_data *data = iio_priv(indio_dev);
+	int ret;
+
+	switch (mask) {
+	case IIO_CHAN_INFO_SAMP_FREQ:
+		mutex_lock(&data->mutex);
+		ret = bmc150_accel_set_bw(data, val, val2);
+		mutex_unlock(&data->mutex);
+		break;
+	case IIO_CHAN_INFO_SCALE:
+		if (val)
+			return -EINVAL;
+
+		mutex_lock(&data->mutex);
+		ret = bmc150_accel_set_scale(data, val2);
+		mutex_unlock(&data->mutex);
+		return ret;
+	default:
+		ret = -EINVAL;
+	}
+
+	return ret;
+}
+
+static int bmc150_accel_read_event(struct iio_dev *indio_dev,
+				   const struct iio_chan_spec *chan,
+				   enum iio_event_type type,
+				   enum iio_event_direction dir,
+				   enum iio_event_info info,
+				   int *val, int *val2)
+{
+	struct bmc150_accel_data *data = iio_priv(indio_dev);
+
+	*val2 = 0;
+	switch (info) {
+	case IIO_EV_INFO_VALUE:
+		*val = data->slope_thres;
+		break;
+	case IIO_EV_INFO_PERIOD:
+		*val = data->slope_dur & BMC150_ACCEL_SLOPE_DUR_MASK;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return IIO_VAL_INT;
+}
+
+static int bmc150_accel_write_event(struct iio_dev *indio_dev,
+				    const struct iio_chan_spec *chan,
+				    enum iio_event_type type,
+				    enum iio_event_direction dir,
+				    enum iio_event_info info,
+				    int val, int val2)
+{
+	struct bmc150_accel_data *data = iio_priv(indio_dev);
+
+	if (data->ev_enable_state)
+		return -EBUSY;
+
+	switch (info) {
+	case IIO_EV_INFO_VALUE:
+		data->slope_thres = val;
+		break;
+	case IIO_EV_INFO_PERIOD:
+		data->slope_dur &= ~BMC150_ACCEL_SLOPE_DUR_MASK;
+		data->slope_dur |= val & BMC150_ACCEL_SLOPE_DUR_MASK;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int bmc150_accel_read_event_config(struct iio_dev *indio_dev,
+					  const struct iio_chan_spec *chan,
+					  enum iio_event_type type,
+					  enum iio_event_direction dir)
+{
+
+	struct bmc150_accel_data *data = iio_priv(indio_dev);
+
+	return data->ev_enable_state;
+}
+
+static int bmc150_accel_write_event_config(struct iio_dev *indio_dev,
+					   const struct iio_chan_spec *chan,
+					   enum iio_event_type type,
+					   enum iio_event_direction dir,
+					   int state)
+{
+	struct bmc150_accel_data *data = iio_priv(indio_dev);
+	int ret;
+
+	if (state && data->ev_enable_state)
+		return 0;
+
+	mutex_lock(&data->mutex);
+
+	if (!state && data->motion_trigger_on) {
+		data->ev_enable_state = 0;
+		mutex_unlock(&data->mutex);
+		return 0;
+	}
+
+	/*
+	 * We will expect the enable and disable to do operation in
+	 * in reverse order. This will happen here anyway as our
+	 * resume operation uses sync mode runtime pm calls, the
+	 * suspend operation will be delayed by autosuspend delay
+	 * So the disable operation will still happen in reverse of
+	 * enable operation. When runtime pm is disabled the mode
+	 * is always on so sequence doesn't matter
+	 */
+
+	ret = bmc150_accel_set_power_state(data, state);
+	if (ret < 0) {
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+
+	ret =  bmc150_accel_setup_any_motion_interrupt(data, state);
+	if (ret < 0) {
+		bmc150_accel_set_power_state(data, false);
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+
+	data->ev_enable_state = state;
+	mutex_unlock(&data->mutex);
+
+	return 0;
+}
+
+static int bmc150_accel_validate_trigger(struct iio_dev *indio_dev,
+				   struct iio_trigger *trig)
+{
+	struct bmc150_accel_data *data = iio_priv(indio_dev);
+
+	if (data->dready_trig != trig && data->motion_trig != trig)
+		return -EINVAL;
+
+	return 0;
+}
+
+static IIO_CONST_ATTR_SAMP_FREQ_AVAIL(
+		"7.810000 15.630000 31.250000 62.500000 125 250 500 1000");
+
+static struct attribute *bmc150_accel_attributes[] = {
+	&iio_const_attr_sampling_frequency_available.dev_attr.attr,
+	NULL,
+};
+
+static const struct attribute_group bmc150_accel_attrs_group = {
+	.attrs = bmc150_accel_attributes,
+};
+
+static const struct iio_event_spec bmc150_accel_event = {
+		.type = IIO_EV_TYPE_ROC,
+		.dir = IIO_EV_DIR_EITHER,
+		.mask_separate = BIT(IIO_EV_INFO_VALUE) |
+				 BIT(IIO_EV_INFO_ENABLE) |
+				 BIT(IIO_EV_INFO_PERIOD)
+};
+
+#define BMC150_ACCEL_CHANNEL(_axis, bits) {				\
+	.type = IIO_ACCEL,						\
+	.modified = 1,							\
+	.channel2 = IIO_MOD_##_axis,					\
+	.info_mask_separate = BIT(IIO_CHAN_INFO_RAW),			\
+	.info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE) |		\
+				BIT(IIO_CHAN_INFO_SAMP_FREQ),		\
+	.scan_index = AXIS_##_axis,					\
+	.scan_type = {							\
+		.sign = 's',						\
+		.realbits = (bits),					\
+		.storagebits = 16,					\
+		.shift = 16 - (bits),					\
+	},								\
+	.event_spec = &bmc150_accel_event,				\
+	.num_event_specs = 1						\
+}
+
+#define BMC150_ACCEL_CHANNELS(bits) {					\
+	{								\
+		.type = IIO_TEMP,					\
+		.info_mask_separate = BIT(IIO_CHAN_INFO_RAW) |		\
+				      BIT(IIO_CHAN_INFO_SCALE) |	\
+				      BIT(IIO_CHAN_INFO_OFFSET),	\
+		.scan_index = -1,					\
+	},								\
+	BMC150_ACCEL_CHANNEL(X, bits),					\
+	BMC150_ACCEL_CHANNEL(Y, bits),					\
+	BMC150_ACCEL_CHANNEL(Z, bits),					\
+	IIO_CHAN_SOFT_TIMESTAMP(3),					\
+}
+
+static const struct iio_chan_spec bma222e_accel_channels[] =
+	BMC150_ACCEL_CHANNELS(8);
+static const struct iio_chan_spec bma250e_accel_channels[] =
+	BMC150_ACCEL_CHANNELS(10);
+static const struct iio_chan_spec bmc150_accel_channels[] =
+	BMC150_ACCEL_CHANNELS(12);
+static const struct iio_chan_spec bma280_accel_channels[] =
+	BMC150_ACCEL_CHANNELS(14);
+
+enum {
+	bmc150,
+	bmi055,
+	bma255,
+	bma250e,
+	bma222e,
+	bma280,
+};
+
+static const struct bmc150_accel_chip_info bmc150_accel_chip_info_tbl[] = {
+	[bmc150] = {
+		.chip_id = 0xFA,
+		.channels = bmc150_accel_channels,
+		.num_channels = ARRAY_SIZE(bmc150_accel_channels),
+		.scale_table = { {9610, BMC150_ACCEL_DEF_RANGE_2G},
+				 {19122, BMC150_ACCEL_DEF_RANGE_4G},
+				 {38344, BMC150_ACCEL_DEF_RANGE_8G},
+				 {76590, BMC150_ACCEL_DEF_RANGE_16G} },
+	},
+	[bmi055] = {
+		.chip_id = 0xFA,
+		.channels = bmc150_accel_channels,
+		.num_channels = ARRAY_SIZE(bmc150_accel_channels),
+		.scale_table = { {9610, BMC150_ACCEL_DEF_RANGE_2G},
+				 {19122, BMC150_ACCEL_DEF_RANGE_4G},
+				 {38344, BMC150_ACCEL_DEF_RANGE_8G},
+				 {76590, BMC150_ACCEL_DEF_RANGE_16G} },
+	},
+	[bma255] = {
+		.chip_id = 0xFA,
+		.channels = bmc150_accel_channels,
+		.num_channels = ARRAY_SIZE(bmc150_accel_channels),
+		.scale_table = { {9610, BMC150_ACCEL_DEF_RANGE_2G},
+				 {19122, BMC150_ACCEL_DEF_RANGE_4G},
+				 {38344, BMC150_ACCEL_DEF_RANGE_8G},
+				 {76590, BMC150_ACCEL_DEF_RANGE_16G} },
+	},
+	[bma250e] = {
+		.chip_id = 0xF9,
+		.channels = bma250e_accel_channels,
+		.num_channels = ARRAY_SIZE(bma250e_accel_channels),
+		.scale_table = { {38344, BMC150_ACCEL_DEF_RANGE_2G},
+				 {76590, BMC150_ACCEL_DEF_RANGE_4G},
+				 {153277, BMC150_ACCEL_DEF_RANGE_8G},
+				 {306457, BMC150_ACCEL_DEF_RANGE_16G} },
+	},
+	[bma222e] = {
+		.chip_id = 0xF8,
+		.channels = bma222e_accel_channels,
+		.num_channels = ARRAY_SIZE(bma222e_accel_channels),
+		.scale_table = { {153277, BMC150_ACCEL_DEF_RANGE_2G},
+				 {306457, BMC150_ACCEL_DEF_RANGE_4G},
+				 {612915, BMC150_ACCEL_DEF_RANGE_8G},
+				 {1225831, BMC150_ACCEL_DEF_RANGE_16G} },
+	},
+	[bma280] = {
+		.chip_id = 0xFB,
+		.channels = bma280_accel_channels,
+		.num_channels = ARRAY_SIZE(bma280_accel_channels),
+		.scale_table = { {2392, BMC150_ACCEL_DEF_RANGE_2G},
+				 {4785, BMC150_ACCEL_DEF_RANGE_4G},
+				 {9581, BMC150_ACCEL_DEF_RANGE_8G},
+				 {19152, BMC150_ACCEL_DEF_RANGE_16G} },
+	},
+};
+
+static const struct iio_info bmc150_accel_info = {
+	.attrs			= &bmc150_accel_attrs_group,
+	.read_raw		= bmc150_accel_read_raw,
+	.write_raw		= bmc150_accel_write_raw,
+	.read_event_value	= bmc150_accel_read_event,
+	.write_event_value	= bmc150_accel_write_event,
+	.write_event_config	= bmc150_accel_write_event_config,
+	.read_event_config	= bmc150_accel_read_event_config,
+	.validate_trigger	= bmc150_accel_validate_trigger,
+	.driver_module		= THIS_MODULE,
+};
+
+static irqreturn_t bmc150_accel_trigger_handler(int irq, void *p)
+{
+	struct iio_poll_func *pf = p;
+	struct iio_dev *indio_dev = pf->indio_dev;
+	struct bmc150_accel_data *data = iio_priv(indio_dev);
+	int bit, ret, i = 0;
+
+	mutex_lock(&data->mutex);
+	for_each_set_bit(bit, indio_dev->buffer->scan_mask,
+			 indio_dev->masklength) {
+		ret = i2c_smbus_read_word_data(data->client,
+					       BMC150_ACCEL_AXIS_TO_REG(bit));
+		if (ret < 0) {
+			mutex_unlock(&data->mutex);
+			goto err_read;
+		}
+		data->buffer[i++] = ret;
+	}
+	mutex_unlock(&data->mutex);
+
+	iio_push_to_buffers_with_timestamp(indio_dev, data->buffer,
+					   data->timestamp);
+err_read:
+	iio_trigger_notify_done(indio_dev->trig);
+
+	return IRQ_HANDLED;
+}
+
+static int bmc150_accel_trig_try_reen(struct iio_trigger *trig)
+{
+	struct iio_dev *indio_dev = iio_trigger_get_drvdata(trig);
+	struct bmc150_accel_data *data = iio_priv(indio_dev);
+	int ret;
+
+	/* new data interrupts don't need ack */
+	if (data->dready_trigger_on)
+		return 0;
+
+	mutex_lock(&data->mutex);
+	/* clear any latched interrupt */
+	ret = i2c_smbus_write_byte_data(data->client,
+					BMC150_ACCEL_REG_INT_RST_LATCH,
+					BMC150_ACCEL_INT_MODE_LATCH_INT |
+					BMC150_ACCEL_INT_MODE_LATCH_RESET);
+	mutex_unlock(&data->mutex);
+	if (ret < 0) {
+		dev_err(&data->client->dev,
+			"Error writing reg_int_rst_latch\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int bmc150_accel_data_rdy_trigger_set_state(struct iio_trigger *trig,
+						   bool state)
+{
+	struct iio_dev *indio_dev = iio_trigger_get_drvdata(trig);
+	struct bmc150_accel_data *data = iio_priv(indio_dev);
+	int ret;
+
+	mutex_lock(&data->mutex);
+
+	if (!state && data->ev_enable_state && data->motion_trigger_on) {
+		data->motion_trigger_on = false;
+		mutex_unlock(&data->mutex);
+		return 0;
+	}
+
+	/*
+	 * Refer to comment in bmc150_accel_write_event_config for
+	 * enable/disable operation order
+	 */
+	ret = bmc150_accel_set_power_state(data, state);
+	if (ret < 0) {
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+	if (data->motion_trig == trig)
+		ret =  bmc150_accel_setup_any_motion_interrupt(data, state);
+	else
+		ret = bmc150_accel_setup_new_data_interrupt(data, state);
+	if (ret < 0) {
+		bmc150_accel_set_power_state(data, false);
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+	if (data->motion_trig == trig)
+		data->motion_trigger_on = state;
+	else
+		data->dready_trigger_on = state;
+
+	mutex_unlock(&data->mutex);
+
+	return ret;
+}
+
+static const struct iio_trigger_ops bmc150_accel_trigger_ops = {
+	.set_trigger_state = bmc150_accel_data_rdy_trigger_set_state,
+	.try_reenable = bmc150_accel_trig_try_reen,
+	.owner = THIS_MODULE,
+};
+
+static irqreturn_t bmc150_accel_event_handler(int irq, void *private)
+{
+	struct iio_dev *indio_dev = private;
+	struct bmc150_accel_data *data = iio_priv(indio_dev);
+	int ret;
+	int dir;
+
+	ret = i2c_smbus_read_byte_data(data->client,
+				       BMC150_ACCEL_REG_INT_STATUS_2);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading reg_int_status_2\n");
+		goto ack_intr_status;
+	}
+
+	if (ret & BMC150_ACCEL_ANY_MOTION_BIT_SIGN)
+		dir = IIO_EV_DIR_FALLING;
+	else
+		dir = IIO_EV_DIR_RISING;
+
+	if (ret & BMC150_ACCEL_ANY_MOTION_BIT_X)
+		iio_push_event(indio_dev, IIO_MOD_EVENT_CODE(IIO_ACCEL,
+							0,
+							IIO_MOD_X,
+							IIO_EV_TYPE_ROC,
+							dir),
+							data->timestamp);
+	if (ret & BMC150_ACCEL_ANY_MOTION_BIT_Y)
+		iio_push_event(indio_dev, IIO_MOD_EVENT_CODE(IIO_ACCEL,
+							0,
+							IIO_MOD_Y,
+							IIO_EV_TYPE_ROC,
+							dir),
+							data->timestamp);
+	if (ret & BMC150_ACCEL_ANY_MOTION_BIT_Z)
+		iio_push_event(indio_dev, IIO_MOD_EVENT_CODE(IIO_ACCEL,
+							0,
+							IIO_MOD_Z,
+							IIO_EV_TYPE_ROC,
+							dir),
+							data->timestamp);
+ack_intr_status:
+	if (!data->dready_trigger_on)
+		ret = i2c_smbus_write_byte_data(data->client,
+					BMC150_ACCEL_REG_INT_RST_LATCH,
+					BMC150_ACCEL_INT_MODE_LATCH_INT |
+					BMC150_ACCEL_INT_MODE_LATCH_RESET);
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t bmc150_accel_data_rdy_trig_poll(int irq, void *private)
+{
+	struct iio_dev *indio_dev = private;
+	struct bmc150_accel_data *data = iio_priv(indio_dev);
+
+	data->timestamp = iio_get_time_ns();
+
+	if (data->dready_trigger_on)
+		iio_trigger_poll(data->dready_trig);
+	else if (data->motion_trigger_on)
+		iio_trigger_poll(data->motion_trig);
+
+	if (data->ev_enable_state)
+		return IRQ_WAKE_THREAD;
+	else
+		return IRQ_HANDLED;
+}
+
+static const char *bmc150_accel_match_acpi_device(struct device *dev, int *data)
+{
+	const struct acpi_device_id *id;
+
+	id = acpi_match_device(dev->driver->acpi_match_table, dev);
+
+	if (!id)
+		return NULL;
+
+	*data = (int) id->driver_data;
+
+	return dev_name(dev);
+}
+
+static int bmc150_accel_gpio_probe(struct i2c_client *client,
+					struct bmc150_accel_data *data)
+{
+	struct device *dev;
+	struct gpio_desc *gpio;
+	int ret;
+
+	if (!client)
+		return -EINVAL;
+
+	dev = &client->dev;
+
+	/* data ready gpio interrupt pin */
+	gpio = devm_gpiod_get_index(dev, BMC150_ACCEL_GPIO_NAME, 0);
+	if (IS_ERR(gpio)) {
+		dev_err(dev, "Failed: gpio get index\n");
+		return PTR_ERR(gpio);
+	}
+
+	ret = gpiod_direction_input(gpio);
+	if (ret)
+		return ret;
+
+	ret = gpiod_to_irq(gpio);
+
+	dev_dbg(dev, "GPIO resource, no:%d irq:%d\n", desc_to_gpio(gpio), ret);
+
+	return ret;
+}
+
+static int bmc150_accel_probe(struct i2c_client *client,
+			      const struct i2c_device_id *id)
+{
+	struct bmc150_accel_data *data;
+	struct iio_dev *indio_dev;
+	int ret;
+	const char *name = NULL;
+	int chip_id = 0;
+
+	indio_dev = devm_iio_device_alloc(&client->dev, sizeof(*data));
+	if (!indio_dev)
+		return -ENOMEM;
+
+	data = iio_priv(indio_dev);
+	i2c_set_clientdata(client, indio_dev);
+	data->client = client;
+
+	if (id) {
+		name = id->name;
+		chip_id = id->driver_data;
+	}
+
+	if (ACPI_HANDLE(&client->dev))
+		name = bmc150_accel_match_acpi_device(&client->dev, &chip_id);
+
+	data->chip_info = &bmc150_accel_chip_info_tbl[chip_id];
+
+	ret = bmc150_accel_chip_init(data);
+	if (ret < 0)
+		return ret;
+
+	mutex_init(&data->mutex);
+
+	indio_dev->dev.parent = &client->dev;
+	indio_dev->channels = data->chip_info->channels;
+	indio_dev->num_channels = data->chip_info->num_channels;
+	indio_dev->name = name;
+	indio_dev->modes = INDIO_DIRECT_MODE;
+	indio_dev->info = &bmc150_accel_info;
+
+	if (client->irq < 0)
+		client->irq = bmc150_accel_gpio_probe(client, data);
+
+	if (client->irq >= 0) {
+		ret = devm_request_threaded_irq(
+						&client->dev, client->irq,
+						bmc150_accel_data_rdy_trig_poll,
+						bmc150_accel_event_handler,
+						IRQF_TRIGGER_RISING,
+						BMC150_ACCEL_IRQ_NAME,
+						indio_dev);
+		if (ret)
+			return ret;
+
+		data->dready_trig = devm_iio_trigger_alloc(&client->dev,
+							   "%s-dev%d",
+							   indio_dev->name,
+							   indio_dev->id);
+		if (!data->dready_trig)
+			return -ENOMEM;
+
+		data->motion_trig = devm_iio_trigger_alloc(&client->dev,
+							  "%s-any-motion-dev%d",
+							  indio_dev->name,
+							  indio_dev->id);
+		if (!data->motion_trig)
+			return -ENOMEM;
+
+		data->dready_trig->dev.parent = &client->dev;
+		data->dready_trig->ops = &bmc150_accel_trigger_ops;
+		iio_trigger_set_drvdata(data->dready_trig, indio_dev);
+		ret = iio_trigger_register(data->dready_trig);
+		if (ret)
+			return ret;
+
+		data->motion_trig->dev.parent = &client->dev;
+		data->motion_trig->ops = &bmc150_accel_trigger_ops;
+		iio_trigger_set_drvdata(data->motion_trig, indio_dev);
+		ret = iio_trigger_register(data->motion_trig);
+		if (ret) {
+			data->motion_trig = NULL;
+			goto err_trigger_unregister;
+		}
+
+		ret = iio_triggered_buffer_setup(indio_dev,
+						 &iio_pollfunc_store_time,
+						 bmc150_accel_trigger_handler,
+						 NULL);
+		if (ret < 0) {
+			dev_err(&client->dev,
+				"Failed: iio triggered buffer setup\n");
+			goto err_trigger_unregister;
+		}
+	}
+
+	ret = iio_device_register(indio_dev);
+	if (ret < 0) {
+		dev_err(&client->dev, "Unable to register iio device\n");
+		goto err_buffer_cleanup;
+	}
+
+	ret = pm_runtime_set_active(&client->dev);
+	if (ret)
+		goto err_iio_unregister;
+
+	pm_runtime_enable(&client->dev);
+	pm_runtime_set_autosuspend_delay(&client->dev,
+					 BMC150_AUTO_SUSPEND_DELAY_MS);
+	pm_runtime_use_autosuspend(&client->dev);
+
+	return 0;
+
+err_iio_unregister:
+	iio_device_unregister(indio_dev);
+err_buffer_cleanup:
+	if (data->dready_trig)
+		iio_triggered_buffer_cleanup(indio_dev);
+err_trigger_unregister:
+	if (data->dready_trig)
+		iio_trigger_unregister(data->dready_trig);
+	if (data->motion_trig)
+		iio_trigger_unregister(data->motion_trig);
+
+	return ret;
+}
+
+static int bmc150_accel_remove(struct i2c_client *client)
+{
+	struct iio_dev *indio_dev = i2c_get_clientdata(client);
+	struct bmc150_accel_data *data = iio_priv(indio_dev);
+
+	pm_runtime_disable(&client->dev);
+	pm_runtime_set_suspended(&client->dev);
+	pm_runtime_put_noidle(&client->dev);
+
+	iio_device_unregister(indio_dev);
+
+	if (data->dready_trig) {
+		iio_triggered_buffer_cleanup(indio_dev);
+		iio_trigger_unregister(data->dready_trig);
+		iio_trigger_unregister(data->motion_trig);
+	}
+
+	mutex_lock(&data->mutex);
+	bmc150_accel_set_mode(data, BMC150_ACCEL_SLEEP_MODE_DEEP_SUSPEND, 0);
+	mutex_unlock(&data->mutex);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int bmc150_accel_suspend(struct device *dev)
+{
+	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
+	struct bmc150_accel_data *data = iio_priv(indio_dev);
+
+	mutex_lock(&data->mutex);
+	bmc150_accel_set_mode(data, BMC150_ACCEL_SLEEP_MODE_SUSPEND, 0);
+	mutex_unlock(&data->mutex);
+
+	return 0;
+}
+
+static int bmc150_accel_resume(struct device *dev)
+{
+	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
+	struct bmc150_accel_data *data = iio_priv(indio_dev);
+
+	mutex_lock(&data->mutex);
+	if (data->dready_trigger_on || data->motion_trigger_on ||
+							data->ev_enable_state)
+		bmc150_accel_set_mode(data, BMC150_ACCEL_SLEEP_MODE_NORMAL, 0);
+	mutex_unlock(&data->mutex);
+
+	return 0;
+}
+#endif
+
+#ifdef CONFIG_PM
+static int bmc150_accel_runtime_suspend(struct device *dev)
+{
+	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
+	struct bmc150_accel_data *data = iio_priv(indio_dev);
+	int ret;
+
+	dev_dbg(&data->client->dev,  __func__);
+	ret = bmc150_accel_set_mode(data, BMC150_ACCEL_SLEEP_MODE_SUSPEND, 0);
+	if (ret < 0)
+		return -EAGAIN;
+
+	return 0;
+}
+
+static int bmc150_accel_runtime_resume(struct device *dev)
+{
+	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
+	struct bmc150_accel_data *data = iio_priv(indio_dev);
+	int ret;
+	int sleep_val;
+
+	dev_dbg(&data->client->dev,  __func__);
+
+	ret = bmc150_accel_set_mode(data, BMC150_ACCEL_SLEEP_MODE_NORMAL, 0);
+	if (ret < 0)
+		return ret;
+
+	sleep_val = bmc150_accel_get_startup_times(data);
+	if (sleep_val < 20)
+		usleep_range(sleep_val * 1000, 20000);
+	else
+		msleep_interruptible(sleep_val);
+
+	return 0;
+}
+#endif
+
+static const struct dev_pm_ops bmc150_accel_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(bmc150_accel_suspend, bmc150_accel_resume)
+	SET_RUNTIME_PM_OPS(bmc150_accel_runtime_suspend,
+			   bmc150_accel_runtime_resume, NULL)
+};
+
+static const struct acpi_device_id bmc150_accel_acpi_match[] = {
+	{"BSBA0150",	bmc150},
+	{"BMC150A",	bmc150},
+	{"BMI055A",	bmi055},
+	{"BMA0255",	bma255},
+	{"BMA250E",	bma250e},
+	{"BMA222E",	bma222e},
+	{"BMA0280",	bma280},
+	{ },
+};
+MODULE_DEVICE_TABLE(acpi, bmc150_accel_acpi_match);
+
+static const struct i2c_device_id bmc150_accel_id[] = {
+	{"bmc150_accel",	bmc150},
+	{"bmi055_accel",	bmi055},
+	{"bma255",		bma255},
+	{"bma250e",		bma250e},
+	{"bma222e",		bma222e},
+	{"bma280",		bma280},
+	{}
+};
+
+MODULE_DEVICE_TABLE(i2c, bmc150_accel_id);
+
+static struct i2c_driver bmc150_accel_driver = {
+	.driver = {
+		.name	= BMC150_ACCEL_DRV_NAME,
+		.acpi_match_table = ACPI_PTR(bmc150_accel_acpi_match),
+		.pm	= &bmc150_accel_pm_ops,
+	},
+	.probe		= bmc150_accel_probe,
+	.remove		= bmc150_accel_remove,
+	.id_table	= bmc150_accel_id,
+};
+module_i2c_driver(bmc150_accel_driver);
+
+MODULE_AUTHOR("Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>");
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("BMC150 accelerometer driver");
diff --git a/drivers/iio/accel/hid-sensor-accel-3d.c b/drivers/iio/accel/hid-sensor-accel-3d.c
index 54e464e4bb72..d5d95317003a 100644
--- a/drivers/iio/accel/hid-sensor-accel-3d.c
+++ b/drivers/iio/accel/hid-sensor-accel-3d.c
@@ -419,7 +419,6 @@ static struct platform_driver hid_accel_3d_platform_driver = {
 	.id_table = hid_accel_3d_ids,
 	.driver = {
 		.name	= KBUILD_MODNAME,
-		.owner	= THIS_MODULE,
 	},
 	.probe		= hid_accel_3d_probe,
 	.remove		= hid_accel_3d_remove,
diff --git a/drivers/iio/accel/kxcjk-1013.c b/drivers/iio/accel/kxcjk-1013.c
index 7941cf2d31ee..da2fe93739a2 100644
--- a/drivers/iio/accel/kxcjk-1013.c
+++ b/drivers/iio/accel/kxcjk-1013.c
@@ -21,10 +21,13 @@
 #include <linux/string.h>
 #include <linux/acpi.h>
 #include <linux/gpio/consumer.h>
+#include <linux/pm.h>
+#include <linux/pm_runtime.h>
 #include <linux/iio/iio.h>
 #include <linux/iio/sysfs.h>
 #include <linux/iio/buffer.h>
 #include <linux/iio/trigger.h>
+#include <linux/iio/events.h>
 #include <linux/iio/trigger_consumer.h>
 #include <linux/iio/triggered_buffer.h>
 #include <linux/iio/accel/kxcjk_1013.h>
@@ -71,15 +74,40 @@
 #define KXCJK1013_DATA_MASK_12_BIT	0x0FFF
 #define KXCJK1013_MAX_STARTUP_TIME_US	100000
 
+#define KXCJK1013_SLEEP_DELAY_MS	2000
+
+#define KXCJK1013_REG_INT_SRC2_BIT_ZP	BIT(0)
+#define KXCJK1013_REG_INT_SRC2_BIT_ZN	BIT(1)
+#define KXCJK1013_REG_INT_SRC2_BIT_YP	BIT(2)
+#define KXCJK1013_REG_INT_SRC2_BIT_YN	BIT(3)
+#define KXCJK1013_REG_INT_SRC2_BIT_XP	BIT(4)
+#define KXCJK1013_REG_INT_SRC2_BIT_XN	BIT(5)
+
+#define KXCJK1013_DEFAULT_WAKE_THRES	1
+
+enum kx_chipset {
+	KXCJK1013,
+	KXCJ91008,
+	KXTJ21009,
+	KX_MAX_CHIPS /* this must be last */
+};
+
 struct kxcjk1013_data {
 	struct i2c_client *client;
-	struct iio_trigger *trig;
-	bool trig_mode;
+	struct iio_trigger *dready_trig;
+	struct iio_trigger *motion_trig;
 	struct mutex mutex;
 	s16 buffer[8];
-	int power_state;
 	u8 odr_bits;
+	u8 range;
+	int wake_thres;
+	int wake_dur;
 	bool active_high_intr;
+	bool dready_trigger_on;
+	int ev_enable_state;
+	bool motion_trigger_on;
+	int64_t timestamp;
+	enum kx_chipset chipset;
 };
 
 enum kxcjk1013_axis {
@@ -93,6 +121,12 @@ enum kxcjk1013_mode {
 	OPERATION,
 };
 
+enum kxcjk1013_range {
+	KXCJK1013_RANGE_2G,
+	KXCJK1013_RANGE_4G,
+	KXCJK1013_RANGE_8G,
+};
+
 static const struct {
 	int val;
 	int val2;
@@ -107,10 +141,78 @@ static const struct {
 static const struct {
 	int odr_bits;
 	int usec;
-} odr_start_up_times[] = { {0x08, 100000}, {0x09, 100000}, {0x0A, 100000},
-			   {0x0B, 100000}, { 0, 80000}, {0x01, 41000},
-			   {0x02, 21000}, {0x03, 11000}, {0x04, 6400},
-			   {0x05, 3900}, {0x06, 2700}, {0x07, 2100} };
+} odr_start_up_times[KX_MAX_CHIPS][12] = {
+	/* KXCJK-1013 */
+	{
+		{0x08, 100000},
+		{0x09, 100000},
+		{0x0A, 100000},
+		{0x0B, 100000},
+		{0, 80000},
+		{0x01, 41000},
+		{0x02, 21000},
+		{0x03, 11000},
+		{0x04, 6400},
+		{0x05, 3900},
+		{0x06, 2700},
+		{0x07, 2100},
+	},
+	/* KXCJ9-1008 */
+	{
+		{0x08, 100000},
+		{0x09, 100000},
+		{0x0A, 100000},
+		{0x0B, 100000},
+		{0, 80000},
+		{0x01, 41000},
+		{0x02, 21000},
+		{0x03, 11000},
+		{0x04, 6400},
+		{0x05, 3900},
+		{0x06, 2700},
+		{0x07, 2100},
+	},
+	/* KXCTJ2-1009 */
+	{
+		{0x08, 1240000},
+		{0x09, 621000},
+		{0x0A, 309000},
+		{0x0B, 151000},
+		{0, 80000},
+		{0x01, 41000},
+		{0x02, 21000},
+		{0x03, 11000},
+		{0x04, 6000},
+		{0x05, 4000},
+		{0x06, 3000},
+		{0x07, 2000},
+	},
+};
+
+static const struct {
+	u16 scale;
+	u8 gsel_0;
+	u8 gsel_1;
+} KXCJK1013_scale_table[] = { {9582, 0, 0},
+			      {19163, 1, 0},
+			      {38326, 0, 1} };
+
+static const struct {
+	int val;
+	int val2;
+	int odr_bits;
+} wake_odr_data_rate_table[] = { {0, 781000, 0x00},
+				 {1, 563000, 0x01},
+				 {3, 125000, 0x02},
+				 {6, 250000, 0x03},
+				 {12, 500000, 0x04},
+				 {25, 0, 0x05},
+				 {50, 0, 0x06},
+				 {100, 0, 0x06},
+				 {200, 0, 0x06},
+				 {400, 0, 0x06},
+				 {800, 0, 0x06},
+				 {1600, 0, 0x06} };
 
 static int kxcjk1013_set_mode(struct kxcjk1013_data *data,
 			      enum kxcjk1013_mode mode)
@@ -138,6 +240,53 @@ static int kxcjk1013_set_mode(struct kxcjk1013_data *data,
 	return 0;
 }
 
+static int kxcjk1013_get_mode(struct kxcjk1013_data *data,
+			      enum kxcjk1013_mode *mode)
+{
+	int ret;
+
+	ret = i2c_smbus_read_byte_data(data->client, KXCJK1013_REG_CTRL1);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading reg_ctrl1\n");
+		return ret;
+	}
+
+	if (ret & KXCJK1013_REG_CTRL1_BIT_PC1)
+		*mode = OPERATION;
+	else
+		*mode = STANDBY;
+
+	return 0;
+}
+
+static int kxcjk1013_set_range(struct kxcjk1013_data *data, int range_index)
+{
+	int ret;
+
+	ret = i2c_smbus_read_byte_data(data->client, KXCJK1013_REG_CTRL1);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading reg_ctrl1\n");
+		return ret;
+	}
+
+	ret &= ~(KXCJK1013_REG_CTRL1_BIT_GSEL0 |
+		 KXCJK1013_REG_CTRL1_BIT_GSEL1);
+	ret |= (KXCJK1013_scale_table[range_index].gsel_0 << 3);
+	ret |= (KXCJK1013_scale_table[range_index].gsel_1 << 4);
+
+	ret = i2c_smbus_write_byte_data(data->client,
+					KXCJK1013_REG_CTRL1,
+					ret);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_ctrl1\n");
+		return ret;
+	}
+
+	data->range = range_index;
+
+	return 0;
+}
+
 static int kxcjk1013_chip_init(struct kxcjk1013_data *data)
 {
 	int ret;
@@ -160,10 +309,6 @@ static int kxcjk1013_chip_init(struct kxcjk1013_data *data)
 		return ret;
 	}
 
-	/* Setting range to 4G */
-	ret |= KXCJK1013_REG_CTRL1_BIT_GSEL0;
-	ret &= ~KXCJK1013_REG_CTRL1_BIT_GSEL1;
-
 	/* Set 12 bit mode */
 	ret |= KXCJK1013_REG_CTRL1_BIT_RES;
 
@@ -174,6 +319,11 @@ static int kxcjk1013_chip_init(struct kxcjk1013_data *data)
 		return ret;
 	}
 
+	/* Setting range to 4G */
+	ret = kxcjk1013_set_range(data, KXCJK1013_RANGE_4G);
+	if (ret < 0)
+		return ret;
+
 	ret = i2c_smbus_read_byte_data(data->client, KXCJK1013_REG_DATA_CTRL);
 	if (ret < 0) {
 		dev_err(&data->client->dev, "Error reading reg_data_ctrl\n");
@@ -201,14 +351,147 @@ static int kxcjk1013_chip_init(struct kxcjk1013_data *data)
 		return ret;
 	}
 
+	ret = kxcjk1013_set_mode(data, OPERATION);
+	if (ret < 0)
+		return ret;
+
+	data->wake_thres = KXCJK1013_DEFAULT_WAKE_THRES;
+
 	return 0;
 }
 
-static int kxcjk1013_chip_setup_interrupt(struct kxcjk1013_data *data,
-					  bool status)
+#ifdef CONFIG_PM
+static int kxcjk1013_get_startup_times(struct kxcjk1013_data *data)
+{
+	int i;
+	int idx = data->chipset;
+
+	for (i = 0; i < ARRAY_SIZE(odr_start_up_times[idx]); ++i) {
+		if (odr_start_up_times[idx][i].odr_bits == data->odr_bits)
+			return odr_start_up_times[idx][i].usec;
+	}
+
+	return KXCJK1013_MAX_STARTUP_TIME_US;
+}
+#endif
+
+static int kxcjk1013_set_power_state(struct kxcjk1013_data *data, bool on)
 {
 	int ret;
 
+	if (on)
+		ret = pm_runtime_get_sync(&data->client->dev);
+	else {
+		pm_runtime_mark_last_busy(&data->client->dev);
+		ret = pm_runtime_put_autosuspend(&data->client->dev);
+	}
+	if (ret < 0) {
+		dev_err(&data->client->dev,
+			"Failed: kxcjk1013_set_power_state for %d\n", on);
+		return ret;
+	}
+
+	return 0;
+}
+
+static int kxcjk1013_chip_update_thresholds(struct kxcjk1013_data *data)
+{
+	int ret;
+
+	ret = i2c_smbus_write_byte_data(data->client,
+					KXCJK1013_REG_WAKE_TIMER,
+					data->wake_dur);
+	if (ret < 0) {
+		dev_err(&data->client->dev,
+			"Error writing reg_wake_timer\n");
+		return ret;
+	}
+
+	ret = i2c_smbus_write_byte_data(data->client,
+					KXCJK1013_REG_WAKE_THRES,
+					data->wake_thres);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_wake_thres\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int kxcjk1013_setup_any_motion_interrupt(struct kxcjk1013_data *data,
+						bool status)
+{
+	int ret;
+	enum kxcjk1013_mode store_mode;
+
+	ret = kxcjk1013_get_mode(data, &store_mode);
+	if (ret < 0)
+		return ret;
+
+	/* This is requirement by spec to change state to STANDBY */
+	ret = kxcjk1013_set_mode(data, STANDBY);
+	if (ret < 0)
+		return ret;
+
+	ret = kxcjk1013_chip_update_thresholds(data);
+	if (ret < 0)
+		return ret;
+
+	ret = i2c_smbus_read_byte_data(data->client, KXCJK1013_REG_INT_CTRL1);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading reg_int_ctrl1\n");
+		return ret;
+	}
+
+	if (status)
+		ret |= KXCJK1013_REG_INT_REG1_BIT_IEN;
+	else
+		ret &= ~KXCJK1013_REG_INT_REG1_BIT_IEN;
+
+	ret = i2c_smbus_write_byte_data(data->client, KXCJK1013_REG_INT_CTRL1,
+					ret);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_int_ctrl1\n");
+		return ret;
+	}
+
+	ret = i2c_smbus_read_byte_data(data->client, KXCJK1013_REG_CTRL1);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading reg_ctrl1\n");
+		return ret;
+	}
+
+	if (status)
+		ret |= KXCJK1013_REG_CTRL1_BIT_WUFE;
+	else
+		ret &= ~KXCJK1013_REG_CTRL1_BIT_WUFE;
+
+	ret = i2c_smbus_write_byte_data(data->client,
+					KXCJK1013_REG_CTRL1, ret);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_ctrl1\n");
+		return ret;
+	}
+
+	if (store_mode == OPERATION) {
+		ret = kxcjk1013_set_mode(data, OPERATION);
+		if (ret < 0)
+			return ret;
+	}
+
+	return 0;
+}
+
+static int kxcjk1013_setup_new_data_interrupt(struct kxcjk1013_data *data,
+					      bool status)
+{
+	int ret;
+	enum kxcjk1013_mode store_mode;
+
+	ret = kxcjk1013_get_mode(data, &store_mode);
+	if (ret < 0)
+		return ret;
+
 	/* This is requirement by spec to change state to STANDBY */
 	ret = kxcjk1013_set_mode(data, STANDBY);
 	if (ret < 0)
@@ -250,7 +533,13 @@ static int kxcjk1013_chip_setup_interrupt(struct kxcjk1013_data *data,
 		return ret;
 	}
 
-	return ret;
+	if (store_mode == OPERATION) {
+		ret = kxcjk1013_set_mode(data, OPERATION);
+		if (ret < 0)
+			return ret;
+	}
+
+	return 0;
 }
 
 static int kxcjk1013_convert_freq_to_bit(int val, int val2)
@@ -267,10 +556,29 @@ static int kxcjk1013_convert_freq_to_bit(int val, int val2)
 	return -EINVAL;
 }
 
+static int kxcjk1013_convert_wake_odr_to_bit(int val, int val2)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(wake_odr_data_rate_table); ++i) {
+		if (wake_odr_data_rate_table[i].val == val &&
+			wake_odr_data_rate_table[i].val2 == val2) {
+			return wake_odr_data_rate_table[i].odr_bits;
+		}
+	}
+
+	return -EINVAL;
+}
+
 static int kxcjk1013_set_odr(struct kxcjk1013_data *data, int val, int val2)
 {
 	int ret;
 	int odr_bits;
+	enum kxcjk1013_mode store_mode;
+
+	ret = kxcjk1013_get_mode(data, &store_mode);
+	if (ret < 0)
+		return ret;
 
 	odr_bits = kxcjk1013_convert_freq_to_bit(val, val2);
 	if (odr_bits < 0)
@@ -290,9 +598,18 @@ static int kxcjk1013_set_odr(struct kxcjk1013_data *data, int val, int val2)
 
 	data->odr_bits = odr_bits;
 
-	/* Check, if the ODR is changed after data enable */
-	if (data->power_state) {
-		/* Set the state back to operation */
+	odr_bits = kxcjk1013_convert_wake_odr_to_bit(val, val2);
+	if (odr_bits < 0)
+		return odr_bits;
+
+	ret = i2c_smbus_write_byte_data(data->client, KXCJK1013_REG_CTRL2,
+					odr_bits);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_ctrl2\n");
+		return ret;
+	}
+
+	if (store_mode == OPERATION) {
 		ret = kxcjk1013_set_mode(data, OPERATION);
 		if (ret < 0)
 			return ret;
@@ -331,16 +648,38 @@ static int kxcjk1013_get_acc_reg(struct kxcjk1013_data *data, int axis)
 	return ret;
 }
 
-static int kxcjk1013_get_startup_times(struct kxcjk1013_data *data)
+static int kxcjk1013_set_scale(struct kxcjk1013_data *data, int val)
 {
-	int i;
+	int ret, i;
+	enum kxcjk1013_mode store_mode;
+
+
+	for (i = 0; i < ARRAY_SIZE(KXCJK1013_scale_table); ++i) {
+		if (KXCJK1013_scale_table[i].scale == val) {
+
+			ret = kxcjk1013_get_mode(data, &store_mode);
+			if (ret < 0)
+				return ret;
+
+			ret = kxcjk1013_set_mode(data, STANDBY);
+			if (ret < 0)
+				return ret;
+
+			ret = kxcjk1013_set_range(data, i);
+			if (ret < 0)
+				return ret;
 
-	for (i = 0; i < ARRAY_SIZE(odr_start_up_times); ++i) {
-		if (odr_start_up_times[i].odr_bits == data->odr_bits)
-			return odr_start_up_times[i].usec;
+			if (store_mode == OPERATION) {
+				ret = kxcjk1013_set_mode(data, OPERATION);
+				if (ret)
+					return ret;
+			}
+
+			return 0;
+		}
 	}
 
-	return KXCJK1013_MAX_STARTUP_TIME_US;
+	return -EINVAL;
 }
 
 static int kxcjk1013_read_raw(struct iio_dev *indio_dev,
@@ -356,34 +695,30 @@ static int kxcjk1013_read_raw(struct iio_dev *indio_dev,
 		if (iio_buffer_enabled(indio_dev))
 			ret = -EBUSY;
 		else {
-			int sleep_val;
-
-			ret = kxcjk1013_set_mode(data, OPERATION);
+			ret = kxcjk1013_set_power_state(data, true);
 			if (ret < 0) {
 				mutex_unlock(&data->mutex);
 				return ret;
 			}
-			++data->power_state;
-			sleep_val = kxcjk1013_get_startup_times(data);
-			if (sleep_val < 20000)
-				usleep_range(sleep_val, 20000);
-			else
-				msleep_interruptible(sleep_val/1000);
 			ret = kxcjk1013_get_acc_reg(data, chan->scan_index);
-			if (--data->power_state == 0)
-				kxcjk1013_set_mode(data, STANDBY);
+			if (ret < 0) {
+				kxcjk1013_set_power_state(data, false);
+				mutex_unlock(&data->mutex);
+				return ret;
+			}
+			*val = sign_extend32(ret >> 4, 11);
+			ret = kxcjk1013_set_power_state(data, false);
 		}
 		mutex_unlock(&data->mutex);
 
 		if (ret < 0)
 			return ret;
 
-		*val = sign_extend32(ret >> 4, 11);
 		return IIO_VAL_INT;
 
 	case IIO_CHAN_INFO_SCALE:
 		*val = 0;
-		*val2 = 19163; /* range +-4g (4/2047*9.806650) */
+		*val2 = KXCJK1013_scale_table[data->range].scale;
 		return IIO_VAL_INT_PLUS_MICRO;
 
 	case IIO_CHAN_INFO_SAMP_FREQ:
@@ -410,6 +745,14 @@ static int kxcjk1013_write_raw(struct iio_dev *indio_dev,
 		ret = kxcjk1013_set_odr(data, val, val2);
 		mutex_unlock(&data->mutex);
 		break;
+	case IIO_CHAN_INFO_SCALE:
+		if (val)
+			return -EINVAL;
+
+		mutex_lock(&data->mutex);
+		ret = kxcjk1013_set_scale(data, val2);
+		mutex_unlock(&data->mutex);
+		break;
 	default:
 		ret = -EINVAL;
 	}
@@ -417,12 +760,120 @@ static int kxcjk1013_write_raw(struct iio_dev *indio_dev,
 	return ret;
 }
 
+static int kxcjk1013_read_event(struct iio_dev *indio_dev,
+				   const struct iio_chan_spec *chan,
+				   enum iio_event_type type,
+				   enum iio_event_direction dir,
+				   enum iio_event_info info,
+				   int *val, int *val2)
+{
+	struct kxcjk1013_data *data = iio_priv(indio_dev);
+
+	*val2 = 0;
+	switch (info) {
+	case IIO_EV_INFO_VALUE:
+		*val = data->wake_thres;
+		break;
+	case IIO_EV_INFO_PERIOD:
+		*val = data->wake_dur;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return IIO_VAL_INT;
+}
+
+static int kxcjk1013_write_event(struct iio_dev *indio_dev,
+				    const struct iio_chan_spec *chan,
+				    enum iio_event_type type,
+				    enum iio_event_direction dir,
+				    enum iio_event_info info,
+				    int val, int val2)
+{
+	struct kxcjk1013_data *data = iio_priv(indio_dev);
+
+	if (data->ev_enable_state)
+		return -EBUSY;
+
+	switch (info) {
+	case IIO_EV_INFO_VALUE:
+		data->wake_thres = val;
+		break;
+	case IIO_EV_INFO_PERIOD:
+		data->wake_dur = val;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int kxcjk1013_read_event_config(struct iio_dev *indio_dev,
+					  const struct iio_chan_spec *chan,
+					  enum iio_event_type type,
+					  enum iio_event_direction dir)
+{
+
+	struct kxcjk1013_data *data = iio_priv(indio_dev);
+
+	return data->ev_enable_state;
+}
+
+static int kxcjk1013_write_event_config(struct iio_dev *indio_dev,
+					   const struct iio_chan_spec *chan,
+					   enum iio_event_type type,
+					   enum iio_event_direction dir,
+					   int state)
+{
+	struct kxcjk1013_data *data = iio_priv(indio_dev);
+	int ret;
+
+	if (state && data->ev_enable_state)
+		return 0;
+
+	mutex_lock(&data->mutex);
+
+	if (!state && data->motion_trigger_on) {
+		data->ev_enable_state = 0;
+		mutex_unlock(&data->mutex);
+		return 0;
+	}
+
+	/*
+	 * We will expect the enable and disable to do operation in
+	 * in reverse order. This will happen here anyway as our
+	 * resume operation uses sync mode runtime pm calls, the
+	 * suspend operation will be delayed by autosuspend delay
+	 * So the disable operation will still happen in reverse of
+	 * enable operation. When runtime pm is disabled the mode
+	 * is always on so sequence doesn't matter
+	 */
+	ret = kxcjk1013_set_power_state(data, state);
+	if (ret < 0) {
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+
+	ret =  kxcjk1013_setup_any_motion_interrupt(data, state);
+	if (ret < 0) {
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+
+	data->ev_enable_state = state;
+	mutex_unlock(&data->mutex);
+
+	return 0;
+}
+
 static int kxcjk1013_validate_trigger(struct iio_dev *indio_dev,
 				      struct iio_trigger *trig)
 {
 	struct kxcjk1013_data *data = iio_priv(indio_dev);
 
-	if (data->trig != trig)
+	if (data->dready_trig != trig && data->motion_trig != trig)
 		return -EINVAL;
 
 	return 0;
@@ -431,8 +882,11 @@ static int kxcjk1013_validate_trigger(struct iio_dev *indio_dev,
 static IIO_CONST_ATTR_SAMP_FREQ_AVAIL(
 	"0.781000 1.563000 3.125000 6.250000 12.500000 25 50 100 200 400 800 1600");
 
+static IIO_CONST_ATTR(in_accel_scale_available, "0.009582 0.019163 0.038326");
+
 static struct attribute *kxcjk1013_attributes[] = {
 	&iio_const_attr_sampling_frequency_available.dev_attr.attr,
+	&iio_const_attr_in_accel_scale_available.dev_attr.attr,
 	NULL,
 };
 
@@ -440,6 +894,14 @@ static const struct attribute_group kxcjk1013_attrs_group = {
 	.attrs = kxcjk1013_attributes,
 };
 
+static const struct iio_event_spec kxcjk1013_event = {
+		.type = IIO_EV_TYPE_THRESH,
+		.dir = IIO_EV_DIR_EITHER,
+		.mask_separate = BIT(IIO_EV_INFO_VALUE) |
+				 BIT(IIO_EV_INFO_ENABLE) |
+				 BIT(IIO_EV_INFO_PERIOD)
+};
+
 #define KXCJK1013_CHANNEL(_axis) {					\
 	.type = IIO_ACCEL,						\
 	.modified = 1,							\
@@ -455,6 +917,8 @@ static const struct attribute_group kxcjk1013_attrs_group = {
 		.shift = 4,						\
 		.endianness = IIO_CPU,					\
 	},								\
+	.event_spec = &kxcjk1013_event,				\
+	.num_event_specs = 1						\
 }
 
 static const struct iio_chan_spec kxcjk1013_channels[] = {
@@ -468,6 +932,10 @@ static const struct iio_info kxcjk1013_info = {
 	.attrs			= &kxcjk1013_attrs_group,
 	.read_raw		= kxcjk1013_read_raw,
 	.write_raw		= kxcjk1013_write_raw,
+	.read_event_value	= kxcjk1013_read_event,
+	.write_event_value	= kxcjk1013_write_event,
+	.write_event_config	= kxcjk1013_write_event_config,
+	.read_event_config	= kxcjk1013_read_event_config,
 	.validate_trigger	= kxcjk1013_validate_trigger,
 	.driver_module		= THIS_MODULE,
 };
@@ -493,7 +961,7 @@ static irqreturn_t kxcjk1013_trigger_handler(int irq, void *p)
 	mutex_unlock(&data->mutex);
 
 	iio_push_to_buffers_with_timestamp(indio_dev, data->buffer,
-					   pf->timestamp);
+					   data->timestamp);
 err:
 	iio_trigger_notify_done(indio_dev->trig);
 
@@ -520,20 +988,34 @@ static int kxcjk1013_data_rdy_trigger_set_state(struct iio_trigger *trig,
 {
 	struct iio_dev *indio_dev = iio_trigger_get_drvdata(trig);
 	struct kxcjk1013_data *data = iio_priv(indio_dev);
+	int ret;
 
 	mutex_lock(&data->mutex);
-	if (state) {
-		kxcjk1013_chip_setup_interrupt(data, true);
-		kxcjk1013_set_mode(data, OPERATION);
-		++data->power_state;
-	} else {
-		if (--data->power_state) {
-			mutex_unlock(&data->mutex);
-			return 0;
-		}
-		kxcjk1013_chip_setup_interrupt(data, false);
-		kxcjk1013_set_mode(data, STANDBY);
+
+	if (!state && data->ev_enable_state && data->motion_trigger_on) {
+		data->motion_trigger_on = false;
+		mutex_unlock(&data->mutex);
+		return 0;
 	}
+
+	ret = kxcjk1013_set_power_state(data, state);
+	if (ret < 0) {
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+	if (data->motion_trig == trig)
+		ret = kxcjk1013_setup_any_motion_interrupt(data, state);
+	else
+		ret = kxcjk1013_setup_new_data_interrupt(data, state);
+	if (ret < 0) {
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+	if (data->motion_trig == trig)
+		data->motion_trigger_on = state;
+	else
+		data->dready_trigger_on = state;
+
 	mutex_unlock(&data->mutex);
 
 	return 0;
@@ -545,10 +1027,124 @@ static const struct iio_trigger_ops kxcjk1013_trigger_ops = {
 	.owner = THIS_MODULE,
 };
 
-static int kxcjk1013_acpi_gpio_probe(struct i2c_client *client,
-				     struct kxcjk1013_data *data)
+static irqreturn_t kxcjk1013_event_handler(int irq, void *private)
+{
+	struct iio_dev *indio_dev = private;
+	struct kxcjk1013_data *data = iio_priv(indio_dev);
+	int ret;
+
+	ret = i2c_smbus_read_byte_data(data->client, KXCJK1013_REG_INT_SRC1);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading reg_int_src1\n");
+		goto ack_intr;
+	}
+
+	if (ret & 0x02) {
+		ret = i2c_smbus_read_byte_data(data->client,
+					       KXCJK1013_REG_INT_SRC2);
+		if (ret < 0) {
+			dev_err(&data->client->dev,
+				"Error reading reg_int_src2\n");
+			goto ack_intr;
+		}
+
+		if (ret & KXCJK1013_REG_INT_SRC2_BIT_XN)
+			iio_push_event(indio_dev,
+				       IIO_MOD_EVENT_CODE(IIO_ACCEL,
+				       0,
+				       IIO_MOD_X,
+				       IIO_EV_TYPE_THRESH,
+				       IIO_EV_DIR_FALLING),
+				       data->timestamp);
+		if (ret & KXCJK1013_REG_INT_SRC2_BIT_XP)
+			iio_push_event(indio_dev,
+				       IIO_MOD_EVENT_CODE(IIO_ACCEL,
+				       0,
+				       IIO_MOD_X,
+				       IIO_EV_TYPE_THRESH,
+				       IIO_EV_DIR_RISING),
+				       data->timestamp);
+
+
+		if (ret & KXCJK1013_REG_INT_SRC2_BIT_YN)
+			iio_push_event(indio_dev,
+				       IIO_MOD_EVENT_CODE(IIO_ACCEL,
+				       0,
+				       IIO_MOD_Y,
+				       IIO_EV_TYPE_THRESH,
+				       IIO_EV_DIR_FALLING),
+				       data->timestamp);
+		if (ret & KXCJK1013_REG_INT_SRC2_BIT_YP)
+			iio_push_event(indio_dev,
+				       IIO_MOD_EVENT_CODE(IIO_ACCEL,
+				       0,
+				       IIO_MOD_Y,
+				       IIO_EV_TYPE_THRESH,
+				       IIO_EV_DIR_RISING),
+				       data->timestamp);
+
+		if (ret & KXCJK1013_REG_INT_SRC2_BIT_ZN)
+			iio_push_event(indio_dev,
+				       IIO_MOD_EVENT_CODE(IIO_ACCEL,
+				       0,
+				       IIO_MOD_Z,
+				       IIO_EV_TYPE_THRESH,
+				       IIO_EV_DIR_FALLING),
+				       data->timestamp);
+		if (ret & KXCJK1013_REG_INT_SRC2_BIT_ZP)
+			iio_push_event(indio_dev,
+				       IIO_MOD_EVENT_CODE(IIO_ACCEL,
+				       0,
+				       IIO_MOD_Z,
+				       IIO_EV_TYPE_THRESH,
+				       IIO_EV_DIR_RISING),
+				       data->timestamp);
+	}
+
+ack_intr:
+	if (data->dready_trigger_on)
+		return IRQ_HANDLED;
+
+	ret = i2c_smbus_read_byte_data(data->client, KXCJK1013_REG_INT_REL);
+	if (ret < 0)
+		dev_err(&data->client->dev, "Error reading reg_int_rel\n");
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t kxcjk1013_data_rdy_trig_poll(int irq, void *private)
+{
+	struct iio_dev *indio_dev = private;
+	struct kxcjk1013_data *data = iio_priv(indio_dev);
+
+	data->timestamp = iio_get_time_ns();
+
+	if (data->dready_trigger_on)
+		iio_trigger_poll(data->dready_trig);
+	else if (data->motion_trigger_on)
+		iio_trigger_poll(data->motion_trig);
+
+	if (data->ev_enable_state)
+		return IRQ_WAKE_THREAD;
+	else
+		return IRQ_HANDLED;
+}
+
+static const char *kxcjk1013_match_acpi_device(struct device *dev,
+					       enum kx_chipset *chipset)
 {
 	const struct acpi_device_id *id;
+	id = acpi_match_device(dev->driver->acpi_match_table, dev);
+	if (!id)
+		return NULL;
+	*chipset = (enum kx_chipset)id->driver_data;
+
+	return dev_name(dev);
+}
+
+static int kxcjk1013_gpio_probe(struct i2c_client *client,
+				struct kxcjk1013_data *data)
+{
 	struct device *dev;
 	struct gpio_desc *gpio;
 	int ret;
@@ -557,12 +1153,6 @@ static int kxcjk1013_acpi_gpio_probe(struct i2c_client *client,
 		return -EINVAL;
 
 	dev = &client->dev;
-	if (!ACPI_HANDLE(dev))
-		return -ENODEV;
-
-	id = acpi_match_device(dev->driver->acpi_match_table, dev);
-	if (!id)
-		return -ENODEV;
 
 	/* data ready gpio interrupt pin */
 	gpio = devm_gpiod_get_index(dev, "kxcjk1013_int", 0);
@@ -587,8 +1177,8 @@ static int kxcjk1013_probe(struct i2c_client *client,
 {
 	struct kxcjk1013_data *data;
 	struct iio_dev *indio_dev;
-	struct iio_trigger *trig = NULL;
 	struct kxcjk_1013_platform_data *pdata;
+	const char *name;
 	int ret;
 
 	indio_dev = devm_iio_device_alloc(&client->dev, sizeof(*data));
@@ -605,6 +1195,15 @@ static int kxcjk1013_probe(struct i2c_client *client,
 	else
 		data->active_high_intr = true; /* default polarity */
 
+	if (id) {
+		data->chipset = (enum kx_chipset)(id->driver_data);
+		name = id->name;
+	} else if (ACPI_HANDLE(&client->dev)) {
+		name = kxcjk1013_match_acpi_device(&client->dev,
+						   &data->chipset);
+	} else
+		return -ENODEV;
+
 	ret = kxcjk1013_chip_init(data);
 	if (ret < 0)
 		return ret;
@@ -614,41 +1213,54 @@ static int kxcjk1013_probe(struct i2c_client *client,
 	indio_dev->dev.parent = &client->dev;
 	indio_dev->channels = kxcjk1013_channels;
 	indio_dev->num_channels = ARRAY_SIZE(kxcjk1013_channels);
-	indio_dev->name = KXCJK1013_DRV_NAME;
+	indio_dev->name = name;
 	indio_dev->modes = INDIO_DIRECT_MODE;
 	indio_dev->info = &kxcjk1013_info;
 
 	if (client->irq < 0)
-		client->irq = kxcjk1013_acpi_gpio_probe(client, data);
+		client->irq = kxcjk1013_gpio_probe(client, data);
 
 	if (client->irq >= 0) {
-		trig = iio_trigger_alloc("%s-dev%d", indio_dev->name,
-					 indio_dev->id);
-		if (!trig)
-			return -ENOMEM;
+		ret = devm_request_threaded_irq(&client->dev, client->irq,
+						kxcjk1013_data_rdy_trig_poll,
+						kxcjk1013_event_handler,
+						IRQF_TRIGGER_RISING,
+						KXCJK1013_IRQ_NAME,
+						indio_dev);
+		if (ret)
+			return ret;
 
-		data->trig_mode = true;
+		data->dready_trig = devm_iio_trigger_alloc(&client->dev,
+							   "%s-dev%d",
+							   indio_dev->name,
+							   indio_dev->id);
+		if (!data->dready_trig)
+			return -ENOMEM;
 
-		ret = devm_request_irq(&client->dev, client->irq,
-					iio_trigger_generic_data_rdy_poll,
-					IRQF_TRIGGER_RISING,
-					KXCJK1013_IRQ_NAME,
-					trig);
-		if (ret) {
-			dev_err(&client->dev, "unable to request IRQ\n");
-			goto err_trigger_free;
-		}
+		data->motion_trig = devm_iio_trigger_alloc(&client->dev,
+							  "%s-any-motion-dev%d",
+							  indio_dev->name,
+							  indio_dev->id);
+		if (!data->motion_trig)
+			return -ENOMEM;
 
-		trig->dev.parent = &client->dev;
-		trig->ops = &kxcjk1013_trigger_ops;
-		iio_trigger_set_drvdata(trig, indio_dev);
-		data->trig = trig;
-		indio_dev->trig = trig;
+		data->dready_trig->dev.parent = &client->dev;
+		data->dready_trig->ops = &kxcjk1013_trigger_ops;
+		iio_trigger_set_drvdata(data->dready_trig, indio_dev);
+		indio_dev->trig = data->dready_trig;
 		iio_trigger_get(indio_dev->trig);
-
-		ret = iio_trigger_register(trig);
+		ret = iio_trigger_register(data->dready_trig);
 		if (ret)
-			goto err_trigger_free;
+			return ret;
+
+		data->motion_trig->dev.parent = &client->dev;
+		data->motion_trig->ops = &kxcjk1013_trigger_ops;
+		iio_trigger_set_drvdata(data->motion_trig, indio_dev);
+		ret = iio_trigger_register(data->motion_trig);
+		if (ret) {
+			data->motion_trig = NULL;
+			goto err_trigger_unregister;
+		}
 
 		ret = iio_triggered_buffer_setup(indio_dev,
 						&iio_pollfunc_store_time,
@@ -661,23 +1273,33 @@ static int kxcjk1013_probe(struct i2c_client *client,
 		}
 	}
 
-	ret = devm_iio_device_register(&client->dev, indio_dev);
+	ret = iio_device_register(indio_dev);
 	if (ret < 0) {
 		dev_err(&client->dev, "unable to register iio device\n");
 		goto err_buffer_cleanup;
 	}
 
+	ret = pm_runtime_set_active(&client->dev);
+	if (ret)
+		goto err_iio_unregister;
+
+	pm_runtime_enable(&client->dev);
+	pm_runtime_set_autosuspend_delay(&client->dev,
+					 KXCJK1013_SLEEP_DELAY_MS);
+	pm_runtime_use_autosuspend(&client->dev);
+
 	return 0;
 
+err_iio_unregister:
+	iio_device_unregister(indio_dev);
 err_buffer_cleanup:
-	if (data->trig_mode)
+	if (data->dready_trig)
 		iio_triggered_buffer_cleanup(indio_dev);
 err_trigger_unregister:
-	if (data->trig_mode)
-		iio_trigger_unregister(trig);
-err_trigger_free:
-	if (data->trig_mode)
-		iio_trigger_free(trig);
+	if (data->dready_trig)
+		iio_trigger_unregister(data->dready_trig);
+	if (data->motion_trig)
+		iio_trigger_unregister(data->motion_trig);
 
 	return ret;
 }
@@ -687,10 +1309,16 @@ static int kxcjk1013_remove(struct i2c_client *client)
 	struct iio_dev *indio_dev = i2c_get_clientdata(client);
 	struct kxcjk1013_data *data = iio_priv(indio_dev);
 
-	if (data->trig_mode) {
+	pm_runtime_disable(&client->dev);
+	pm_runtime_set_suspended(&client->dev);
+	pm_runtime_put_noidle(&client->dev);
+
+	iio_device_unregister(indio_dev);
+
+	if (data->dready_trig) {
 		iio_triggered_buffer_cleanup(indio_dev);
-		iio_trigger_unregister(data->trig);
-		iio_trigger_free(data->trig);
+		iio_trigger_unregister(data->dready_trig);
+		iio_trigger_unregister(data->motion_trig);
 	}
 
 	mutex_lock(&data->mutex);
@@ -705,43 +1333,80 @@ static int kxcjk1013_suspend(struct device *dev)
 {
 	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
 	struct kxcjk1013_data *data = iio_priv(indio_dev);
+	int ret;
 
 	mutex_lock(&data->mutex);
-	kxcjk1013_set_mode(data, STANDBY);
+	ret = kxcjk1013_set_mode(data, STANDBY);
 	mutex_unlock(&data->mutex);
 
-	return 0;
+	return ret;
 }
 
 static int kxcjk1013_resume(struct device *dev)
 {
 	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
 	struct kxcjk1013_data *data = iio_priv(indio_dev);
+	int ret = 0;
 
 	mutex_lock(&data->mutex);
+	/* Check, if the suspend occured while active */
+	if (data->dready_trigger_on || data->motion_trigger_on ||
+							data->ev_enable_state)
+		ret = kxcjk1013_set_mode(data, OPERATION);
+	mutex_unlock(&data->mutex);
 
-	if (data->power_state)
-		kxcjk1013_set_mode(data, OPERATION);
+	return ret;
+}
+#endif
 
-	mutex_unlock(&data->mutex);
+#ifdef CONFIG_PM
+static int kxcjk1013_runtime_suspend(struct device *dev)
+{
+	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
+	struct kxcjk1013_data *data = iio_priv(indio_dev);
 
-	return 0;
+	return kxcjk1013_set_mode(data, STANDBY);
 }
 
-static SIMPLE_DEV_PM_OPS(kxcjk1013_pm_ops, kxcjk1013_suspend, kxcjk1013_resume);
-#define KXCJK1013_PM_OPS (&kxcjk1013_pm_ops)
-#else
-#define KXCJK1013_PM_OPS NULL
+static int kxcjk1013_runtime_resume(struct device *dev)
+{
+	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
+	struct kxcjk1013_data *data = iio_priv(indio_dev);
+	int ret;
+	int sleep_val;
+
+	ret = kxcjk1013_set_mode(data, OPERATION);
+	if (ret < 0)
+		return ret;
+
+	sleep_val = kxcjk1013_get_startup_times(data);
+	if (sleep_val < 20000)
+		usleep_range(sleep_val, 20000);
+	else
+		msleep_interruptible(sleep_val/1000);
+
+	return 0;
+}
 #endif
 
+static const struct dev_pm_ops kxcjk1013_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(kxcjk1013_suspend, kxcjk1013_resume)
+	SET_RUNTIME_PM_OPS(kxcjk1013_runtime_suspend,
+			   kxcjk1013_runtime_resume, NULL)
+};
+
 static const struct acpi_device_id kx_acpi_match[] = {
-	{"KXCJ1013", 0},
+	{"KXCJ1013", KXCJK1013},
+	{"KXCJ1008", KXCJ91008},
+	{"KXTJ1009", KXTJ21009},
 	{ },
 };
 MODULE_DEVICE_TABLE(acpi, kx_acpi_match);
 
 static const struct i2c_device_id kxcjk1013_id[] = {
-	{"kxcjk1013", 0},
+	{"kxcjk1013", KXCJK1013},
+	{"kxcj91008", KXCJ91008},
+	{"kxtj21009", KXTJ21009},
 	{}
 };
 
@@ -751,7 +1416,7 @@ static struct i2c_driver kxcjk1013_driver = {
 	.driver = {
 		.name	= KXCJK1013_DRV_NAME,
 		.acpi_match_table = ACPI_PTR(kx_acpi_match),
-		.pm	= KXCJK1013_PM_OPS,
+		.pm	= &kxcjk1013_pm_ops,
 	},
 	.probe		= kxcjk1013_probe,
 	.remove		= kxcjk1013_remove,
diff --git a/drivers/iio/accel/st_accel.h b/drivers/iio/accel/st_accel.h
index c3877630b2e4..fa9646034305 100644
--- a/drivers/iio/accel/st_accel.h
+++ b/drivers/iio/accel/st_accel.h
@@ -33,8 +33,7 @@ static const struct st_sensors_platform_data default_accel_pdata = {
 	.drdy_int_pin = 1,
 };
 
-int st_accel_common_probe(struct iio_dev *indio_dev,
-					struct st_sensors_platform_data *pdata);
+int st_accel_common_probe(struct iio_dev *indio_dev);
 void st_accel_common_remove(struct iio_dev *indio_dev);
 
 #ifdef CONFIG_IIO_BUFFER
diff --git a/drivers/iio/accel/st_accel_core.c b/drivers/iio/accel/st_accel_core.c
index 087864854c61..53f32629283a 100644
--- a/drivers/iio/accel/st_accel_core.c
+++ b/drivers/iio/accel/st_accel_core.c
@@ -161,7 +161,7 @@ static const struct iio_chan_spec st_accel_16bit_channels[] = {
 	IIO_CHAN_SOFT_TIMESTAMP(3)
 };
 
-static const struct st_sensors st_accel_sensors[] = {
+static const struct st_sensor_settings st_accel_sensors_settings[] = {
 	{
 		.wai = ST_ACCEL_1_WAI_EXP,
 		.sensors_supported = {
@@ -457,8 +457,7 @@ static const struct iio_trigger_ops st_accel_trigger_ops = {
 #define ST_ACCEL_TRIGGER_OPS NULL
 #endif
 
-int st_accel_common_probe(struct iio_dev *indio_dev,
-				struct st_sensors_platform_data *plat_data)
+int st_accel_common_probe(struct iio_dev *indio_dev)
 {
 	struct st_sensor_data *adata = iio_priv(indio_dev);
 	int irq = adata->get_irq_data_ready(indio_dev);
@@ -470,24 +469,25 @@ int st_accel_common_probe(struct iio_dev *indio_dev,
 	st_sensors_power_enable(indio_dev);
 
 	err = st_sensors_check_device_support(indio_dev,
-				ARRAY_SIZE(st_accel_sensors), st_accel_sensors);
+					ARRAY_SIZE(st_accel_sensors_settings),
+					st_accel_sensors_settings);
 	if (err < 0)
 		return err;
 
 	adata->num_data_channels = ST_ACCEL_NUMBER_DATA_CHANNELS;
-	adata->multiread_bit = adata->sensor->multi_read_bit;
-	indio_dev->channels = adata->sensor->ch;
+	adata->multiread_bit = adata->sensor_settings->multi_read_bit;
+	indio_dev->channels = adata->sensor_settings->ch;
 	indio_dev->num_channels = ST_SENSORS_NUMBER_ALL_CHANNELS;
 
 	adata->current_fullscale = (struct st_sensor_fullscale_avl *)
-						&adata->sensor->fs.fs_avl[0];
-	adata->odr = adata->sensor->odr.odr_avl[0].hz;
+					&adata->sensor_settings->fs.fs_avl[0];
+	adata->odr = adata->sensor_settings->odr.odr_avl[0].hz;
 
-	if (!plat_data)
-		plat_data =
+	if (!adata->dev->platform_data)
+		adata->dev->platform_data =
 			(struct st_sensors_platform_data *)&default_accel_pdata;
 
-	err = st_sensors_init_sensor(indio_dev, plat_data);
+	err = st_sensors_init_sensor(indio_dev, adata->dev->platform_data);
 	if (err < 0)
 		return err;
 
diff --git a/drivers/iio/accel/st_accel_i2c.c b/drivers/iio/accel/st_accel_i2c.c
index 7164aeff3ab1..c7246bdd30b9 100644
--- a/drivers/iio/accel/st_accel_i2c.c
+++ b/drivers/iio/accel/st_accel_i2c.c
@@ -79,12 +79,11 @@ static int st_accel_i2c_probe(struct i2c_client *client,
 		return -ENOMEM;
 
 	adata = iio_priv(indio_dev);
-	adata->dev = &client->dev;
 	st_sensors_of_i2c_probe(client, st_accel_of_match);
 
 	st_sensors_i2c_configure(indio_dev, client, adata);
 
-	err = st_accel_common_probe(indio_dev, client->dev.platform_data);
+	err = st_accel_common_probe(indio_dev);
 	if (err < 0)
 		return err;
 
diff --git a/drivers/iio/accel/st_accel_spi.c b/drivers/iio/accel/st_accel_spi.c
index 195639646e34..12ec29389e4b 100644
--- a/drivers/iio/accel/st_accel_spi.c
+++ b/drivers/iio/accel/st_accel_spi.c
@@ -29,11 +29,10 @@ static int st_accel_spi_probe(struct spi_device *spi)
 		return -ENOMEM;
 
 	adata = iio_priv(indio_dev);
-	adata->dev = &spi->dev;
 
 	st_sensors_spi_configure(indio_dev, spi, adata);
 
-	err = st_accel_common_probe(indio_dev, spi->dev.platform_data);
+	err = st_accel_common_probe(indio_dev);
 	if (err < 0)
 		return err;
 
diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
index 11b048a59fde..0f79e4725763 100644
--- a/drivers/iio/adc/Kconfig
+++ b/drivers/iio/adc/Kconfig
@@ -127,9 +127,17 @@ config AT91_ADC
 	help
 	  Say yes here to build support for Atmel AT91 ADC.
 
+config AXP288_ADC
+	tristate "X-Powers AXP288 ADC driver"
+	depends on MFD_AXP20X
+	help
+	  Say yes here to have support for X-Powers power management IC (PMIC) ADC
+	  device. Depending on platform configuration, this general purpose ADC can
+	  be used for sampling sensors such as thermal resistors.
+
 config EXYNOS_ADC
 	tristate "Exynos ADC driver support"
-	depends on ARCH_EXYNOS || (OF && COMPILE_TEST)
+	depends on ARCH_EXYNOS || ARCH_S3C24XX || ARCH_S3C64XX || (OF && COMPILE_TEST)
 	help
 	  Core support for the ADC block found in the Samsung EXYNOS series
 	  of SoCs for drivers such as the touchscreen and hwmon to use to share
@@ -206,6 +214,30 @@ config NAU7802
 	  To compile this driver as a module, choose M here: the
 	  module will be called nau7802.
 
+config QCOM_SPMI_IADC
+	tristate "Qualcomm SPMI PMIC current ADC"
+	depends on SPMI
+	select REGMAP_SPMI
+	help
+	  This is the IIO Current ADC driver for Qualcomm QPNP IADC Chip.
+
+	  The driver supports single mode operation to read from one of two
+	  channels (external or internal). Hardware have additional
+	  channels internally used for gain and offset calibration.
+
+	  To compile this driver as a module, choose M here: the module will
+	  be called qcom-spmi-iadc.
+
+config ROCKCHIP_SARADC
+	tristate "Rockchip SARADC driver"
+	depends on ARCH_ROCKCHIP || (ARM && COMPILE_TEST)
+	help
+	  Say yes here to build support for the SARADC found in SoCs from
+	  Rockchip.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called rockchip_saradc.
+
 config TI_ADC081C
 	tristate "Texas Instruments ADC081C021/027"
 	depends on I2C
@@ -216,6 +248,16 @@ config TI_ADC081C
 	  This driver can also be built as a module. If so, the module will be
 	  called ti-adc081c.
 
+config TI_ADC128S052
+	tristate "Texas Instruments ADC128S052"
+	depends on SPI
+	help
+	  If you say yes here you get support for Texas Instruments ADC128S052
+	  chip.
+
+	  This driver can also be built as a module. If so, the module will be
+	  called ti-adc128s052.
+
 config TI_AM335X_ADC
 	tristate "TI's AM335X ADC driver"
 	depends on MFD_TI_AM335X_TSCADC
diff --git a/drivers/iio/adc/Makefile b/drivers/iio/adc/Makefile
index ad81b512aa3d..701fdb7c96aa 100644
--- a/drivers/iio/adc/Makefile
+++ b/drivers/iio/adc/Makefile
@@ -14,6 +14,7 @@ obj-$(CONFIG_AD7793) += ad7793.o
 obj-$(CONFIG_AD7887) += ad7887.o
 obj-$(CONFIG_AD799X) += ad799x.o
 obj-$(CONFIG_AT91_ADC) += at91_adc.o
+obj-$(CONFIG_AXP288_ADC) += axp288_adc.o
 obj-$(CONFIG_EXYNOS_ADC) += exynos_adc.o
 obj-$(CONFIG_LP8788_ADC) += lp8788_adc.o
 obj-$(CONFIG_MAX1027) += max1027.o
@@ -22,7 +23,10 @@ obj-$(CONFIG_MCP320X) += mcp320x.o
 obj-$(CONFIG_MCP3422) += mcp3422.o
 obj-$(CONFIG_MEN_Z188_ADC) += men_z188_adc.o
 obj-$(CONFIG_NAU7802) += nau7802.o
+obj-$(CONFIG_QCOM_SPMI_IADC) += qcom-spmi-iadc.o
+obj-$(CONFIG_ROCKCHIP_SARADC) += rockchip_saradc.o
 obj-$(CONFIG_TI_ADC081C) += ti-adc081c.o
+obj-$(CONFIG_TI_ADC128S052) += ti-adc128s052.o
 obj-$(CONFIG_TI_AM335X_ADC) += ti_am335x_adc.o
 obj-$(CONFIG_TWL4030_MADC) += twl4030-madc.o
 obj-$(CONFIG_TWL6030_GPADC) += twl6030-gpadc.o
diff --git a/drivers/iio/adc/ad799x.c b/drivers/iio/adc/ad799x.c
index e37412da15f5..b99de00e57b8 100644
--- a/drivers/iio/adc/ad799x.c
+++ b/drivers/iio/adc/ad799x.c
@@ -143,9 +143,15 @@ static int ad799x_write_config(struct ad799x_state *st, u16 val)
 	case ad7998:
 		return i2c_smbus_write_word_swapped(st->client, AD7998_CONF_REG,
 			val);
-	default:
+	case ad7992:
+	case ad7993:
+	case ad7994:
 		return i2c_smbus_write_byte_data(st->client, AD7998_CONF_REG,
 			val);
+	default:
+		/* Will be written when doing a conversion */
+		st->config = val;
+		return 0;
 	}
 }
 
@@ -155,8 +161,13 @@ static int ad799x_read_config(struct ad799x_state *st)
 	case ad7997:
 	case ad7998:
 		return i2c_smbus_read_word_swapped(st->client, AD7998_CONF_REG);
-	default:
+	case ad7992:
+	case ad7993:
+	case ad7994:
 		return i2c_smbus_read_byte_data(st->client, AD7998_CONF_REG);
+	default:
+		/* No readback support */
+		return st->config;
 	}
 }
 
diff --git a/drivers/iio/adc/at91_adc.c b/drivers/iio/adc/at91_adc.c
index 7eadaf16adc1..ff61ae55dd3f 100644
--- a/drivers/iio/adc/at91_adc.c
+++ b/drivers/iio/adc/at91_adc.c
@@ -267,7 +267,7 @@ static irqreturn_t at91_adc_trigger_handler(int irq, void *p)
 }
 
 /* Handler for classic adc channel eoc trigger */
-void handle_adc_eoc_trigger(int irq, struct iio_dev *idev)
+static void handle_adc_eoc_trigger(int irq, struct iio_dev *idev)
 {
 	struct at91_adc_state *st = iio_priv(idev);
 
diff --git a/drivers/iio/adc/axp288_adc.c b/drivers/iio/adc/axp288_adc.c
new file mode 100644
index 000000000000..08bcfb061ca5
--- /dev/null
+++ b/drivers/iio/adc/axp288_adc.c
@@ -0,0 +1,261 @@
+/*
+ * axp288_adc.c - X-Powers AXP288 PMIC ADC Driver
+ *
+ * Copyright (C) 2014 Intel Corporation
+ *
+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.	See the GNU
+ * General Public License for more details.
+ *
+ */
+
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/device.h>
+#include <linux/regmap.h>
+#include <linux/mfd/axp20x.h>
+#include <linux/platform_device.h>
+
+#include <linux/iio/iio.h>
+#include <linux/iio/machine.h>
+#include <linux/iio/driver.h>
+
+#define AXP288_ADC_EN_MASK		0xF1
+#define AXP288_ADC_TS_PIN_GPADC		0xF2
+#define AXP288_ADC_TS_PIN_ON		0xF3
+
+enum axp288_adc_id {
+	AXP288_ADC_TS,
+	AXP288_ADC_PMIC,
+	AXP288_ADC_GP,
+	AXP288_ADC_BATT_CHRG_I,
+	AXP288_ADC_BATT_DISCHRG_I,
+	AXP288_ADC_BATT_V,
+	AXP288_ADC_NR_CHAN,
+};
+
+struct axp288_adc_info {
+	int irq;
+	struct regmap *regmap;
+};
+
+static const struct iio_chan_spec const axp288_adc_channels[] = {
+	{
+		.indexed = 1,
+		.type = IIO_TEMP,
+		.channel = 0,
+		.address = AXP288_TS_ADC_H,
+		.datasheet_name = "TS_PIN",
+	}, {
+		.indexed = 1,
+		.type = IIO_TEMP,
+		.channel = 1,
+		.address = AXP288_PMIC_ADC_H,
+		.datasheet_name = "PMIC_TEMP",
+	}, {
+		.indexed = 1,
+		.type = IIO_TEMP,
+		.channel = 2,
+		.address = AXP288_GP_ADC_H,
+		.datasheet_name = "GPADC",
+	}, {
+		.indexed = 1,
+		.type = IIO_CURRENT,
+		.channel = 3,
+		.address = AXP20X_BATT_CHRG_I_H,
+		.datasheet_name = "BATT_CHG_I",
+		.info_mask_separate = BIT(IIO_CHAN_INFO_PROCESSED),
+	}, {
+		.indexed = 1,
+		.type = IIO_CURRENT,
+		.channel = 4,
+		.address = AXP20X_BATT_DISCHRG_I_H,
+		.datasheet_name = "BATT_DISCHRG_I",
+		.info_mask_separate = BIT(IIO_CHAN_INFO_PROCESSED),
+	}, {
+		.indexed = 1,
+		.type = IIO_VOLTAGE,
+		.channel = 5,
+		.address = AXP20X_BATT_V_H,
+		.datasheet_name = "BATT_V",
+		.info_mask_separate = BIT(IIO_CHAN_INFO_PROCESSED),
+	},
+};
+
+#define AXP288_ADC_MAP(_adc_channel_label, _consumer_dev_name,	\
+		_consumer_channel)				\
+	{							\
+		.adc_channel_label = _adc_channel_label,	\
+		.consumer_dev_name = _consumer_dev_name,	\
+		.consumer_channel = _consumer_channel,		\
+	}
+
+/* for consumer drivers */
+static struct iio_map axp288_adc_default_maps[] = {
+	AXP288_ADC_MAP("TS_PIN", "axp288-batt", "axp288-batt-temp"),
+	AXP288_ADC_MAP("PMIC_TEMP", "axp288-pmic", "axp288-pmic-temp"),
+	AXP288_ADC_MAP("GPADC", "axp288-gpadc", "axp288-system-temp"),
+	AXP288_ADC_MAP("BATT_CHG_I", "axp288-chrg", "axp288-chrg-curr"),
+	AXP288_ADC_MAP("BATT_DISCHRG_I", "axp288-chrg", "axp288-chrg-d-curr"),
+	AXP288_ADC_MAP("BATT_V", "axp288-batt", "axp288-batt-volt"),
+	{},
+};
+
+static int axp288_adc_read_channel(int *val, unsigned long address,
+				struct regmap *regmap)
+{
+	u8 buf[2];
+
+	if (regmap_bulk_read(regmap, address, buf, 2))
+		return -EIO;
+	*val = (buf[0] << 4) + ((buf[1] >> 4) & 0x0F);
+
+	return IIO_VAL_INT;
+}
+
+static int axp288_adc_set_ts(struct regmap *regmap, unsigned int mode,
+				unsigned long address)
+{
+	/* channels other than GPADC do not need to switch TS pin */
+	if (address != AXP288_GP_ADC_H)
+		return 0;
+
+	return regmap_write(regmap, AXP288_ADC_TS_PIN_CTRL, mode);
+}
+
+static int axp288_adc_read_raw(struct iio_dev *indio_dev,
+			struct iio_chan_spec const *chan,
+			int *val, int *val2, long mask)
+{
+	int ret;
+	struct axp288_adc_info *info = iio_priv(indio_dev);
+
+	mutex_lock(&indio_dev->mlock);
+	switch (mask) {
+	case IIO_CHAN_INFO_RAW:
+		if (axp288_adc_set_ts(info->regmap, AXP288_ADC_TS_PIN_GPADC,
+					chan->address)) {
+			dev_err(&indio_dev->dev, "GPADC mode\n");
+			ret = -EINVAL;
+			break;
+		}
+		ret = axp288_adc_read_channel(val, chan->address, info->regmap);
+		if (axp288_adc_set_ts(info->regmap, AXP288_ADC_TS_PIN_ON,
+						chan->address))
+			dev_err(&indio_dev->dev, "TS pin restore\n");
+		break;
+	case IIO_CHAN_INFO_PROCESSED:
+		ret = axp288_adc_read_channel(val, chan->address, info->regmap);
+		break;
+	default:
+		ret = -EINVAL;
+	}
+	mutex_unlock(&indio_dev->mlock);
+
+	return ret;
+}
+
+static int axp288_adc_set_state(struct regmap *regmap)
+{
+	/* ADC should be always enabled for internal FG to function */
+	if (regmap_write(regmap, AXP288_ADC_TS_PIN_CTRL, AXP288_ADC_TS_PIN_ON))
+		return -EIO;
+
+	return regmap_write(regmap, AXP20X_ADC_EN1, AXP288_ADC_EN_MASK);
+}
+
+static const struct iio_info axp288_adc_iio_info = {
+	.read_raw = &axp288_adc_read_raw,
+	.driver_module = THIS_MODULE,
+};
+
+static int axp288_adc_probe(struct platform_device *pdev)
+{
+	int ret;
+	struct axp288_adc_info *info;
+	struct iio_dev *indio_dev;
+	struct axp20x_dev *axp20x = dev_get_drvdata(pdev->dev.parent);
+
+	indio_dev = devm_iio_device_alloc(&pdev->dev, sizeof(*info));
+	if (!indio_dev)
+		return -ENOMEM;
+
+	info = iio_priv(indio_dev);
+	info->irq = platform_get_irq(pdev, 0);
+	if (info->irq < 0) {
+		dev_err(&pdev->dev, "no irq resource?\n");
+		return info->irq;
+	}
+	platform_set_drvdata(pdev, indio_dev);
+	info->regmap = axp20x->regmap;
+	/*
+	 * Set ADC to enabled state at all time, including system suspend.
+	 * otherwise internal fuel gauge functionality may be affected.
+	 */
+	ret = axp288_adc_set_state(axp20x->regmap);
+	if (ret) {
+		dev_err(&pdev->dev, "unable to enable ADC device\n");
+		return ret;
+	}
+
+	indio_dev->dev.parent = &pdev->dev;
+	indio_dev->name = pdev->name;
+	indio_dev->channels = axp288_adc_channels;
+	indio_dev->num_channels = ARRAY_SIZE(axp288_adc_channels);
+	indio_dev->info = &axp288_adc_iio_info;
+	indio_dev->modes = INDIO_DIRECT_MODE;
+	ret = iio_map_array_register(indio_dev, axp288_adc_default_maps);
+	if (ret < 0)
+		return ret;
+
+	ret = iio_device_register(indio_dev);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "unable to register iio device\n");
+		goto err_array_unregister;
+	}
+	return 0;
+
+err_array_unregister:
+	iio_map_array_unregister(indio_dev);
+
+	return ret;
+}
+
+static int axp288_adc_remove(struct platform_device *pdev)
+{
+	struct iio_dev *indio_dev = platform_get_drvdata(pdev);
+
+	iio_device_unregister(indio_dev);
+	iio_map_array_unregister(indio_dev);
+
+	return 0;
+}
+
+static struct platform_device_id axp288_adc_id_table[] = {
+	{ .name = "axp288_adc" },
+	{},
+};
+
+static struct platform_driver axp288_adc_driver = {
+	.probe = axp288_adc_probe,
+	.remove = axp288_adc_remove,
+	.id_table = axp288_adc_id_table,
+	.driver = {
+		.name = "axp288_adc",
+	},
+};
+
+MODULE_DEVICE_TABLE(platform, axp288_adc_id_table);
+
+module_platform_driver(axp288_adc_driver);
+
+MODULE_AUTHOR("Jacob Pan <jacob.jun.pan@linux.intel.com>");
+MODULE_DESCRIPTION("X-Powers AXP288 ADC Driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/iio/adc/exynos_adc.c b/drivers/iio/adc/exynos_adc.c
index fc9dfc23ecb7..3a2dbb3b4926 100644
--- a/drivers/iio/adc/exynos_adc.c
+++ b/drivers/iio/adc/exynos_adc.c
@@ -39,14 +39,19 @@
 #include <linux/iio/iio.h>
 #include <linux/iio/machine.h>
 #include <linux/iio/driver.h>
+#include <linux/mfd/syscon.h>
+#include <linux/regmap.h>
 
-/* EXYNOS4412/5250 ADC_V1 registers definitions */
+/* S3C/EXYNOS4412/5250 ADC_V1 registers definitions */
 #define ADC_V1_CON(x)		((x) + 0x00)
 #define ADC_V1_DLY(x)		((x) + 0x08)
 #define ADC_V1_DATX(x)		((x) + 0x0C)
 #define ADC_V1_INTCLR(x)	((x) + 0x18)
 #define ADC_V1_MUX(x)		((x) + 0x1c)
 
+/* S3C2410 ADC registers definitions */
+#define ADC_S3C2410_MUX(x)	((x) + 0x18)
+
 /* Future ADC_V2 registers definitions */
 #define ADC_V2_CON1(x)		((x) + 0x00)
 #define ADC_V2_CON2(x)		((x) + 0x04)
@@ -61,6 +66,11 @@
 #define ADC_V1_CON_PRSCLV(x)	(((x) & 0xFF) << 6)
 #define ADC_V1_CON_STANDBY	(1u << 2)
 
+/* Bit definitions for S3C2410 ADC */
+#define ADC_S3C2410_CON_SELMUX(x) (((x) & 7) << 3)
+#define ADC_S3C2410_DATX_MASK	0x3FF
+#define ADC_S3C2416_CON_RES_SEL	(1u << 3)
+
 /* Bit definitions for ADC_V2 */
 #define ADC_V2_CON1_SOFT_RESET	(1u << 2)
 
@@ -77,15 +87,19 @@
 
 /* Bit definitions common for ADC_V1 and ADC_V2 */
 #define ADC_CON_EN_START	(1u << 0)
+#define ADC_CON_EN_START_MASK	(0x3 << 0)
 #define ADC_DATX_MASK		0xFFF
 
 #define EXYNOS_ADC_TIMEOUT	(msecs_to_jiffies(100))
 
+#define EXYNOS_ADCV1_PHY_OFFSET	0x0718
+#define EXYNOS_ADCV2_PHY_OFFSET	0x0720
+
 struct exynos_adc {
 	struct exynos_adc_data	*data;
 	struct device		*dev;
 	void __iomem		*regs;
-	void __iomem		*enable_reg;
+	struct regmap		*pmu_map;
 	struct clk		*clk;
 	struct clk		*sclk;
 	unsigned int		irq;
@@ -100,6 +114,9 @@ struct exynos_adc {
 struct exynos_adc_data {
 	int num_channels;
 	bool needs_sclk;
+	bool needs_adc_phy;
+	int phy_offset;
+	u32 mask;
 
 	void (*init_hw)(struct exynos_adc *info);
 	void (*exit_hw)(struct exynos_adc *info);
@@ -171,7 +188,8 @@ static void exynos_adc_v1_init_hw(struct exynos_adc *info)
 {
 	u32 con1;
 
-	writel(1, info->enable_reg);
+	if (info->data->needs_adc_phy)
+		regmap_write(info->pmu_map, info->data->phy_offset, 1);
 
 	/* set default prescaler values and Enable prescaler */
 	con1 =  ADC_V1_CON_PRSCLV(49) | ADC_V1_CON_PRSCEN;
@@ -185,7 +203,8 @@ static void exynos_adc_v1_exit_hw(struct exynos_adc *info)
 {
 	u32 con;
 
-	writel(0, info->enable_reg);
+	if (info->data->needs_adc_phy)
+		regmap_write(info->pmu_map, info->data->phy_offset, 0);
 
 	con = readl(ADC_V1_CON(info->regs));
 	con |= ADC_V1_CON_STANDBY;
@@ -210,6 +229,9 @@ static void exynos_adc_v1_start_conv(struct exynos_adc *info,
 
 static const struct exynos_adc_data exynos_adc_v1_data = {
 	.num_channels	= MAX_ADC_V1_CHANNELS,
+	.mask		= ADC_DATX_MASK,	/* 12 bit ADC resolution */
+	.needs_adc_phy	= true,
+	.phy_offset	= EXYNOS_ADCV1_PHY_OFFSET,
 
 	.init_hw	= exynos_adc_v1_init_hw,
 	.exit_hw	= exynos_adc_v1_exit_hw,
@@ -217,11 +239,89 @@ static const struct exynos_adc_data exynos_adc_v1_data = {
 	.start_conv	= exynos_adc_v1_start_conv,
 };
 
+static void exynos_adc_s3c2416_start_conv(struct exynos_adc *info,
+					  unsigned long addr)
+{
+	u32 con1;
+
+	/* Enable 12 bit ADC resolution */
+	con1 = readl(ADC_V1_CON(info->regs));
+	con1 |= ADC_S3C2416_CON_RES_SEL;
+	writel(con1, ADC_V1_CON(info->regs));
+
+	/* Select channel for S3C2416 */
+	writel(addr, ADC_S3C2410_MUX(info->regs));
+
+	con1 = readl(ADC_V1_CON(info->regs));
+	writel(con1 | ADC_CON_EN_START, ADC_V1_CON(info->regs));
+}
+
+static struct exynos_adc_data const exynos_adc_s3c2416_data = {
+	.num_channels	= MAX_ADC_V1_CHANNELS,
+	.mask		= ADC_DATX_MASK,	/* 12 bit ADC resolution */
+
+	.init_hw	= exynos_adc_v1_init_hw,
+	.exit_hw	= exynos_adc_v1_exit_hw,
+	.start_conv	= exynos_adc_s3c2416_start_conv,
+};
+
+static void exynos_adc_s3c2443_start_conv(struct exynos_adc *info,
+					  unsigned long addr)
+{
+	u32 con1;
+
+	/* Select channel for S3C2433 */
+	writel(addr, ADC_S3C2410_MUX(info->regs));
+
+	con1 = readl(ADC_V1_CON(info->regs));
+	writel(con1 | ADC_CON_EN_START, ADC_V1_CON(info->regs));
+}
+
+static struct exynos_adc_data const exynos_adc_s3c2443_data = {
+	.num_channels	= MAX_ADC_V1_CHANNELS,
+	.mask		= ADC_S3C2410_DATX_MASK, /* 10 bit ADC resolution */
+
+	.init_hw	= exynos_adc_v1_init_hw,
+	.exit_hw	= exynos_adc_v1_exit_hw,
+	.start_conv	= exynos_adc_s3c2443_start_conv,
+};
+
+static void exynos_adc_s3c64xx_start_conv(struct exynos_adc *info,
+					  unsigned long addr)
+{
+	u32 con1;
+
+	con1 = readl(ADC_V1_CON(info->regs));
+	con1 &= ~ADC_S3C2410_CON_SELMUX(0x7);
+	con1 |= ADC_S3C2410_CON_SELMUX(addr);
+	writel(con1 | ADC_CON_EN_START, ADC_V1_CON(info->regs));
+}
+
+static struct exynos_adc_data const exynos_adc_s3c24xx_data = {
+	.num_channels	= MAX_ADC_V1_CHANNELS,
+	.mask		= ADC_S3C2410_DATX_MASK, /* 10 bit ADC resolution */
+
+	.init_hw	= exynos_adc_v1_init_hw,
+	.exit_hw	= exynos_adc_v1_exit_hw,
+	.start_conv	= exynos_adc_s3c64xx_start_conv,
+};
+
+static struct exynos_adc_data const exynos_adc_s3c64xx_data = {
+	.num_channels	= MAX_ADC_V1_CHANNELS,
+	.mask		= ADC_DATX_MASK,	/* 12 bit ADC resolution */
+
+	.init_hw	= exynos_adc_v1_init_hw,
+	.exit_hw	= exynos_adc_v1_exit_hw,
+	.clear_irq	= exynos_adc_v1_clear_irq,
+	.start_conv	= exynos_adc_s3c64xx_start_conv,
+};
+
 static void exynos_adc_v2_init_hw(struct exynos_adc *info)
 {
 	u32 con1, con2;
 
-	writel(1, info->enable_reg);
+	if (info->data->needs_adc_phy)
+		regmap_write(info->pmu_map, info->data->phy_offset, 1);
 
 	con1 = ADC_V2_CON1_SOFT_RESET;
 	writel(con1, ADC_V2_CON1(info->regs));
@@ -238,7 +338,8 @@ static void exynos_adc_v2_exit_hw(struct exynos_adc *info)
 {
 	u32 con;
 
-	writel(0, info->enable_reg);
+	if (info->data->needs_adc_phy)
+		regmap_write(info->pmu_map, info->data->phy_offset, 0);
 
 	con = readl(ADC_V2_CON1(info->regs));
 	con &= ~ADC_CON_EN_START;
@@ -266,6 +367,9 @@ static void exynos_adc_v2_start_conv(struct exynos_adc *info,
 
 static const struct exynos_adc_data exynos_adc_v2_data = {
 	.num_channels	= MAX_ADC_V2_CHANNELS,
+	.mask		= ADC_DATX_MASK, /* 12 bit ADC resolution */
+	.needs_adc_phy	= true,
+	.phy_offset	= EXYNOS_ADCV2_PHY_OFFSET,
 
 	.init_hw	= exynos_adc_v2_init_hw,
 	.exit_hw	= exynos_adc_v2_exit_hw,
@@ -275,7 +379,10 @@ static const struct exynos_adc_data exynos_adc_v2_data = {
 
 static const struct exynos_adc_data exynos3250_adc_data = {
 	.num_channels	= MAX_EXYNOS3250_ADC_CHANNELS,
+	.mask		= ADC_DATX_MASK, /* 12 bit ADC resolution */
 	.needs_sclk	= true,
+	.needs_adc_phy	= true,
+	.phy_offset	= EXYNOS_ADCV1_PHY_OFFSET,
 
 	.init_hw	= exynos_adc_v2_init_hw,
 	.exit_hw	= exynos_adc_v2_exit_hw,
@@ -283,8 +390,52 @@ static const struct exynos_adc_data exynos3250_adc_data = {
 	.start_conv	= exynos_adc_v2_start_conv,
 };
 
+static void exynos_adc_exynos7_init_hw(struct exynos_adc *info)
+{
+	u32 con1, con2;
+
+	if (info->data->needs_adc_phy)
+		regmap_write(info->pmu_map, info->data->phy_offset, 1);
+
+	con1 = ADC_V2_CON1_SOFT_RESET;
+	writel(con1, ADC_V2_CON1(info->regs));
+
+	con2 = readl(ADC_V2_CON2(info->regs));
+	con2 &= ~ADC_V2_CON2_C_TIME(7);
+	con2 |= ADC_V2_CON2_C_TIME(0);
+	writel(con2, ADC_V2_CON2(info->regs));
+
+	/* Enable interrupts */
+	writel(1, ADC_V2_INT_EN(info->regs));
+}
+
+static const struct exynos_adc_data exynos7_adc_data = {
+	.num_channels	= MAX_ADC_V1_CHANNELS,
+	.mask		= ADC_DATX_MASK, /* 12 bit ADC resolution */
+
+	.init_hw	= exynos_adc_exynos7_init_hw,
+	.exit_hw	= exynos_adc_v2_exit_hw,
+	.clear_irq	= exynos_adc_v2_clear_irq,
+	.start_conv	= exynos_adc_v2_start_conv,
+};
+
 static const struct of_device_id exynos_adc_match[] = {
 	{
+		.compatible = "samsung,s3c2410-adc",
+		.data = &exynos_adc_s3c24xx_data,
+	}, {
+		.compatible = "samsung,s3c2416-adc",
+		.data = &exynos_adc_s3c2416_data,
+	}, {
+		.compatible = "samsung,s3c2440-adc",
+		.data = &exynos_adc_s3c24xx_data,
+	}, {
+		.compatible = "samsung,s3c2443-adc",
+		.data = &exynos_adc_s3c2443_data,
+	}, {
+		.compatible = "samsung,s3c6410-adc",
+		.data = &exynos_adc_s3c64xx_data,
+	}, {
 		.compatible = "samsung,exynos-adc-v1",
 		.data = &exynos_adc_v1_data,
 	}, {
@@ -293,6 +444,9 @@ static const struct of_device_id exynos_adc_match[] = {
 	}, {
 		.compatible = "samsung,exynos3250-adc",
 		.data = &exynos3250_adc_data,
+	}, {
+		.compatible = "samsung,exynos7-adc",
+		.data = &exynos7_adc_data,
 	},
 	{},
 };
@@ -347,9 +501,10 @@ static int exynos_read_raw(struct iio_dev *indio_dev,
 static irqreturn_t exynos_adc_isr(int irq, void *dev_id)
 {
 	struct exynos_adc *info = (struct exynos_adc *)dev_id;
+	u32 mask = info->data->mask;
 
 	/* Read value */
-	info->value = readl(ADC_V1_DATX(info->regs)) & ADC_DATX_MASK;
+	info->value = readl(ADC_V1_DATX(info->regs)) & mask;
 
 	/* clear irq */
 	if (info->data->clear_irq)
@@ -442,10 +597,16 @@ static int exynos_adc_probe(struct platform_device *pdev)
 	if (IS_ERR(info->regs))
 		return PTR_ERR(info->regs);
 
-	mem = platform_get_resource(pdev, IORESOURCE_MEM, 1);
-	info->enable_reg = devm_ioremap_resource(&pdev->dev, mem);
-	if (IS_ERR(info->enable_reg))
-		return PTR_ERR(info->enable_reg);
+
+	if (info->data->needs_adc_phy) {
+		info->pmu_map = syscon_regmap_lookup_by_phandle(
+					pdev->dev.of_node,
+					"samsung,syscon-phandle");
+		if (IS_ERR(info->pmu_map)) {
+			dev_err(&pdev->dev, "syscon regmap lookup failed.\n");
+			return PTR_ERR(info->pmu_map);
+		}
+	}
 
 	irq = platform_get_irq(pdev, 0);
 	if (irq < 0) {
@@ -606,7 +767,6 @@ static struct platform_driver exynos_adc_driver = {
 	.remove		= exynos_adc_remove,
 	.driver		= {
 		.name	= "exynos-adc",
-		.owner	= THIS_MODULE,
 		.of_match_table = exynos_adc_match,
 		.pm	= &exynos_adc_pm_ops,
 	},
diff --git a/drivers/iio/adc/lp8788_adc.c b/drivers/iio/adc/lp8788_adc.c
index 5c8c91595f47..152cfc8e1c7b 100644
--- a/drivers/iio/adc/lp8788_adc.c
+++ b/drivers/iio/adc/lp8788_adc.c
@@ -244,7 +244,6 @@ static struct platform_driver lp8788_adc_driver = {
 	.remove = lp8788_adc_remove,
 	.driver = {
 		.name = LP8788_DEV_ADC,
-		.owner = THIS_MODULE,
 	},
 };
 module_platform_driver(lp8788_adc_driver);
diff --git a/drivers/iio/adc/mcp320x.c b/drivers/iio/adc/mcp320x.c
index 28a086e48776..efbfd12a4bfd 100644
--- a/drivers/iio/adc/mcp320x.c
+++ b/drivers/iio/adc/mcp320x.c
@@ -1,9 +1,30 @@
 /*
  * Copyright (C) 2013 Oskar Andero <oskar.andero@gmail.com>
+ * Copyright (C) 2014 Rose Technology
+ * 	   Allan Bendorff Jensen <abj@rosetechnology.dk>
+ *	   Soren Andersen <san@rosetechnology.dk>
+ *
+ * Driver for following ADC chips from Microchip Technology's:
+ * 10 Bit converter
+ * MCP3001
+ * MCP3002
+ * MCP3004
+ * MCP3008
+ * ------------
+ * 12 bit converter
+ * MCP3201
+ * MCP3202
+ * MCP3204
+ * MCP3208
+ * ------------
  *
- * Driver for Microchip Technology's MCP3204 and MCP3208 ADC chips.
  * Datasheet can be found here:
- * http://ww1.microchip.com/downloads/en/devicedoc/21298c.pdf
+ * http://ww1.microchip.com/downloads/en/DeviceDoc/21293C.pdf  mcp3001
+ * http://ww1.microchip.com/downloads/en/DeviceDoc/21294E.pdf  mcp3002
+ * http://ww1.microchip.com/downloads/en/DeviceDoc/21295d.pdf  mcp3004/08
+ * http://ww1.microchip.com/downloads/en/DeviceDoc/21290D.pdf  mcp3201
+ * http://ww1.microchip.com/downloads/en/DeviceDoc/21034D.pdf  mcp3202
+ * http://ww1.microchip.com/downloads/en/DeviceDoc/21298c.pdf  mcp3204/08
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License version 2 as
@@ -11,19 +32,29 @@
  */
 
 #include <linux/err.h>
+#include <linux/delay.h>
 #include <linux/spi/spi.h>
 #include <linux/module.h>
 #include <linux/iio/iio.h>
 #include <linux/regulator/consumer.h>
 
-#define MCP_SINGLE_ENDED	(1 << 3)
-#define MCP_START_BIT		(1 << 4)
-
 enum {
+	mcp3001,
+	mcp3002,
+	mcp3004,
+	mcp3008,
+	mcp3201,
+	mcp3202,
 	mcp3204,
 	mcp3208,
 };
 
+struct mcp320x_chip_info {
+	const struct iio_chan_spec *channels;
+	unsigned int num_channels;
+	unsigned int resolution;
+};
+
 struct mcp320x {
 	struct spi_device *spi;
 	struct spi_message msg;
@@ -34,19 +65,69 @@ struct mcp320x {
 
 	struct regulator *reg;
 	struct mutex lock;
+	const struct mcp320x_chip_info *chip_info;
 };
 
-static int mcp320x_adc_conversion(struct mcp320x *adc, u8 msg)
+static int mcp320x_channel_to_tx_data(int device_index,
+			const unsigned int channel, bool differential)
+{
+	int start_bit = 1;
+
+	switch (device_index) {
+	case mcp3001:
+	case mcp3201:
+		return 0;
+	case mcp3002:
+	case mcp3202:
+		return ((start_bit << 4) | (!differential << 3) |
+							(channel << 2));
+	case mcp3004:
+	case mcp3204:
+	case mcp3008:
+	case mcp3208:
+		return ((start_bit << 6) | (!differential << 5) |
+							(channel << 2));
+	default:
+		return -EINVAL;
+	}
+}
+
+static int mcp320x_adc_conversion(struct mcp320x *adc, u8 channel,
+				  bool differential, int device_index)
 {
 	int ret;
 
-	adc->tx_buf = msg;
-	ret = spi_sync(adc->spi, &adc->msg);
-	if (ret < 0)
-		return ret;
+	adc->rx_buf[0] = 0;
+	adc->rx_buf[1] = 0;
+	adc->tx_buf = mcp320x_channel_to_tx_data(device_index,
+						channel, differential);
+
+	if (device_index != mcp3001 && device_index != mcp3201) {
+		ret = spi_sync(adc->spi, &adc->msg);
+		if (ret < 0)
+			return ret;
+	} else {
+		ret = spi_read(adc->spi, &adc->rx_buf, sizeof(adc->rx_buf));
+		if (ret < 0)
+			return ret;
+	}
 
-	return ((adc->rx_buf[0] & 0x3f) << 6)  |
-		(adc->rx_buf[1] >> 2);
+	switch (device_index) {
+	case mcp3001:
+		return (adc->rx_buf[0] << 5 | adc->rx_buf[1] >> 3);
+	case mcp3002:
+	case mcp3004:
+	case mcp3008:
+		return (adc->rx_buf[0] << 2 | adc->rx_buf[1] >> 6);
+	case mcp3201:
+		return (adc->rx_buf[0] << 7 | adc->rx_buf[1] >> 1);
+	case mcp3202:
+	case mcp3204:
+	case mcp3208:
+		return (adc->rx_buf[0] << 4 | adc->rx_buf[1] >> 4);
+	default:
+		return -EINVAL;
+	}
 }
 
 static int mcp320x_read_raw(struct iio_dev *indio_dev,
@@ -55,18 +136,17 @@ static int mcp320x_read_raw(struct iio_dev *indio_dev,
 {
 	struct mcp320x *adc = iio_priv(indio_dev);
 	int ret = -EINVAL;
+	int device_index = 0;
 
 	mutex_lock(&adc->lock);
 
+	device_index = spi_get_device_id(adc->spi)->driver_data;
+
 	switch (mask) {
 	case IIO_CHAN_INFO_RAW:
-		if (channel->differential)
-			ret = mcp320x_adc_conversion(adc,
-				MCP_START_BIT | channel->address);
-		else
-			ret = mcp320x_adc_conversion(adc,
-				MCP_START_BIT | MCP_SINGLE_ENDED |
-				channel->address);
+		ret = mcp320x_adc_conversion(adc, channel->address,
+			channel->differential, device_index);
+
 		if (ret < 0)
 			goto out;
 
@@ -75,18 +155,15 @@ static int mcp320x_read_raw(struct iio_dev *indio_dev,
 		break;
 
 	case IIO_CHAN_INFO_SCALE:
-		/* Digital output code = (4096 * Vin) / Vref */
 		ret = regulator_get_voltage(adc->reg);
 		if (ret < 0)
 			goto out;
 
+		/* convert regulator output voltage to mV */
 		*val = ret / 1000;
-		*val2 = 12;
+		*val2 = adc->chip_info->resolution;
 		ret = IIO_VAL_FRACTIONAL_LOG2;
 		break;
-
-	default:
-		break;
 	}
 
 out:
@@ -117,6 +194,16 @@ out:
 		.info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE) \
 	}
 
+static const struct iio_chan_spec mcp3201_channels[] = {
+	MCP320X_VOLTAGE_CHANNEL_DIFF(0),
+};
+
+static const struct iio_chan_spec mcp3202_channels[] = {
+	MCP320X_VOLTAGE_CHANNEL(0),
+	MCP320X_VOLTAGE_CHANNEL(1),
+	MCP320X_VOLTAGE_CHANNEL_DIFF(0),
+};
+
 static const struct iio_chan_spec mcp3204_channels[] = {
 	MCP320X_VOLTAGE_CHANNEL(0),
 	MCP320X_VOLTAGE_CHANNEL(1),
@@ -146,19 +233,46 @@ static const struct iio_info mcp320x_info = {
 	.driver_module = THIS_MODULE,
 };
 
-struct mcp3208_chip_info {
-	const struct iio_chan_spec *channels;
-	unsigned int num_channels;
-};
-
-static const struct mcp3208_chip_info mcp3208_chip_infos[] = {
+static const struct mcp320x_chip_info mcp320x_chip_infos[] = {
+	[mcp3001] = {
+		.channels = mcp3201_channels,
+		.num_channels = ARRAY_SIZE(mcp3201_channels),
+		.resolution = 10
+	},
+	[mcp3002] = {
+		.channels = mcp3202_channels,
+		.num_channels = ARRAY_SIZE(mcp3202_channels),
+		.resolution = 10
+	},
+	[mcp3004] = {
+		.channels = mcp3204_channels,
+		.num_channels = ARRAY_SIZE(mcp3204_channels),
+		.resolution = 10
+	},
+	[mcp3008] = {
+		.channels = mcp3208_channels,
+		.num_channels = ARRAY_SIZE(mcp3208_channels),
+		.resolution = 10
+	},
+	[mcp3201] = {
+		.channels = mcp3201_channels,
+		.num_channels = ARRAY_SIZE(mcp3201_channels),
+		.resolution = 12
+	},
+	[mcp3202] = {
+		.channels = mcp3202_channels,
+		.num_channels = ARRAY_SIZE(mcp3202_channels),
+		.resolution = 12
+	},
 	[mcp3204] = {
 		.channels = mcp3204_channels,
-		.num_channels = ARRAY_SIZE(mcp3204_channels)
+		.num_channels = ARRAY_SIZE(mcp3204_channels),
+		.resolution = 12
 	},
 	[mcp3208] = {
 		.channels = mcp3208_channels,
-		.num_channels = ARRAY_SIZE(mcp3208_channels)
+		.num_channels = ARRAY_SIZE(mcp3208_channels),
+		.resolution = 12
 	},
 };
 
@@ -166,7 +280,7 @@ static int mcp320x_probe(struct spi_device *spi)
 {
 	struct iio_dev *indio_dev;
 	struct mcp320x *adc;
-	const struct mcp3208_chip_info *chip_info;
+	const struct mcp320x_chip_info *chip_info;
 	int ret;
 
 	indio_dev = devm_iio_device_alloc(&spi->dev, sizeof(*adc));
@@ -181,7 +295,7 @@ static int mcp320x_probe(struct spi_device *spi)
 	indio_dev->modes = INDIO_DIRECT_MODE;
 	indio_dev->info = &mcp320x_info;
 
-	chip_info = &mcp3208_chip_infos[spi_get_device_id(spi)->driver_data];
+	chip_info = &mcp320x_chip_infos[spi_get_device_id(spi)->driver_data];
 	indio_dev->channels = chip_info->channels;
 	indio_dev->num_channels = chip_info->num_channels;
 
@@ -226,7 +340,45 @@ static int mcp320x_remove(struct spi_device *spi)
 	return 0;
 }
 
+#if defined(CONFIG_OF)
+static const struct of_device_id mcp320x_dt_ids[] = {
+	{
+		.compatible = "mcp3001",
+		.data = &mcp320x_chip_infos[mcp3001],
+	}, {
+		.compatible = "mcp3002",
+		.data = &mcp320x_chip_infos[mcp3002],
+	}, {
+		.compatible = "mcp3004",
+		.data = &mcp320x_chip_infos[mcp3004],
+	}, {
+		.compatible = "mcp3008",
+		.data = &mcp320x_chip_infos[mcp3008],
+	}, {
+		.compatible = "mcp3201",
+		.data = &mcp320x_chip_infos[mcp3201],
+	}, {
+		.compatible = "mcp3202",
+		.data = &mcp320x_chip_infos[mcp3202],
+	}, {
+		.compatible = "mcp3204",
+		.data = &mcp320x_chip_infos[mcp3204],
+	}, {
+		.compatible = "mcp3208",
+		.data = &mcp320x_chip_infos[mcp3208],
+	}, {
+	}
+};
+MODULE_DEVICE_TABLE(of, mcp320x_dt_ids);
+#endif
+
 static const struct spi_device_id mcp320x_id[] = {
+	{ "mcp3001", mcp3001 },
+	{ "mcp3002", mcp3002 },
+	{ "mcp3004", mcp3004 },
+	{ "mcp3008", mcp3008 },
+	{ "mcp3201", mcp3201 },
+	{ "mcp3202", mcp3202 },
 	{ "mcp3204", mcp3204 },
 	{ "mcp3208", mcp3208 },
 	{ }
@@ -245,5 +397,5 @@ static struct spi_driver mcp320x_driver = {
 module_spi_driver(mcp320x_driver);
 
 MODULE_AUTHOR("Oskar Andero <oskar.andero@gmail.com>");
-MODULE_DESCRIPTION("Microchip Technology MCP3204/08");
+MODULE_DESCRIPTION("Microchip Technology MCP3x01/02/04/08");
 MODULE_LICENSE("GPL v2");
diff --git a/drivers/iio/adc/men_z188_adc.c b/drivers/iio/adc/men_z188_adc.c
index b58d6302521f..d095efe1ba14 100644
--- a/drivers/iio/adc/men_z188_adc.c
+++ b/drivers/iio/adc/men_z188_adc.c
@@ -152,6 +152,7 @@ static void men_z188_remove(struct mcb_device *dev)
 
 static const struct mcb_device_id men_z188_ids[] = {
 	{ .device = 0xbc },
+	{ }
 };
 MODULE_DEVICE_TABLE(mcb, men_z188_ids);
 
diff --git a/drivers/iio/adc/qcom-spmi-iadc.c b/drivers/iio/adc/qcom-spmi-iadc.c
new file mode 100644
index 000000000000..b9666f2f5e51
--- /dev/null
+++ b/drivers/iio/adc/qcom-spmi-iadc.c
@@ -0,0 +1,595 @@
+/*
+ * Copyright (c) 2012-2014, The Linux Foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/bitops.h>
+#include <linux/completion.h>
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/iio/iio.h>
+#include <linux/interrupt.h>
+#include <linux/kernel.h>
+#include <linux/mutex.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/platform_device.h>
+#include <linux/regmap.h>
+#include <linux/slab.h>
+
+/* IADC register and bit definition */
+#define IADC_REVISION2				0x1
+#define IADC_REVISION2_SUPPORTED_IADC		1
+
+#define IADC_PERPH_TYPE				0x4
+#define IADC_PERPH_TYPE_ADC			8
+
+#define IADC_PERPH_SUBTYPE			0x5
+#define IADC_PERPH_SUBTYPE_IADC			3
+
+#define IADC_STATUS1				0x8
+#define IADC_STATUS1_OP_MODE			4
+#define IADC_STATUS1_REQ_STS			BIT(1)
+#define IADC_STATUS1_EOC			BIT(0)
+#define IADC_STATUS1_REQ_STS_EOC_MASK		0x3
+
+#define IADC_MODE_CTL				0x40
+#define IADC_OP_MODE_SHIFT			3
+#define IADC_OP_MODE_NORMAL			0
+#define IADC_TRIM_EN				BIT(0)
+
+#define IADC_EN_CTL1				0x46
+#define IADC_EN_CTL1_SET			BIT(7)
+
+#define IADC_CH_SEL_CTL				0x48
+
+#define IADC_DIG_PARAM				0x50
+#define IADC_DIG_DEC_RATIO_SEL_SHIFT		2
+
+#define IADC_HW_SETTLE_DELAY			0x51
+
+#define IADC_CONV_REQ				0x52
+#define IADC_CONV_REQ_SET			BIT(7)
+
+#define IADC_FAST_AVG_CTL			0x5a
+#define IADC_FAST_AVG_EN			0x5b
+#define IADC_FAST_AVG_EN_SET			BIT(7)
+
+#define IADC_PERH_RESET_CTL3			0xda
+#define IADC_FOLLOW_WARM_RB			BIT(2)
+
+#define IADC_DATA				0x60	/* 16 bits */
+
+#define IADC_SEC_ACCESS				0xd0
+#define IADC_SEC_ACCESS_DATA			0xa5
+
+#define IADC_NOMINAL_RSENSE			0xf4
+#define IADC_NOMINAL_RSENSE_SIGN_MASK		BIT(7)
+
+#define IADC_REF_GAIN_MICRO_VOLTS		17857
+
+#define IADC_INT_RSENSE_DEVIATION		15625	/* nano Ohms per bit */
+
+#define IADC_INT_RSENSE_IDEAL_VALUE		10000	/* micro Ohms */
+#define IADC_INT_RSENSE_DEFAULT_VALUE		7800	/* micro Ohms */
+#define IADC_INT_RSENSE_DEFAULT_GF		9000	/* micro Ohms */
+#define IADC_INT_RSENSE_DEFAULT_SMIC		9700	/* micro Ohms */
+
+#define IADC_CONV_TIME_MIN_US			2000
+#define IADC_CONV_TIME_MAX_US			2100
+
+#define IADC_DEF_PRESCALING			0 /* 1:1 */
+#define IADC_DEF_DECIMATION			0 /* 512 */
+#define IADC_DEF_HW_SETTLE_TIME			0 /* 0 us */
+#define IADC_DEF_AVG_SAMPLES			0 /* 1 sample */
+
+/* IADC channel list */
+#define IADC_INT_RSENSE				0
+#define IADC_EXT_RSENSE				1
+#define IADC_GAIN_17P857MV			3
+#define IADC_EXT_OFFSET_CSP_CSN			5
+#define IADC_INT_OFFSET_CSP2_CSN2		6
+
+/**
+ * struct iadc_chip - IADC Current ADC device structure.
+ * @regmap: regmap for register read/write.
+ * @dev: This device pointer.
+ * @base: base offset for the ADC peripheral.
+ * @rsense: Values of the internal and external sense resister in micro Ohms.
+ * @poll_eoc: Poll for end of conversion instead of waiting for IRQ.
+ * @offset: Raw offset values for the internal and external channels.
+ * @gain: Raw gain of the channels.
+ * @lock: ADC lock for access to the peripheral.
+ * @complete: ADC notification after end of conversion interrupt is received.
+ */
+struct iadc_chip {
+	struct regmap	*regmap;
+	struct device	*dev;
+	u16		base;
+	bool		poll_eoc;
+	u32		rsense[2];
+	u16		offset[2];
+	u16		gain;
+	struct mutex	lock;
+	struct completion complete;
+};
+
+static int iadc_read(struct iadc_chip *iadc, u16 offset, u8 *data)
+{
+	unsigned int val;
+	int ret;
+
+	ret = regmap_read(iadc->regmap, iadc->base + offset, &val);
+	if (ret < 0)
+		return ret;
+
+	*data = val;
+	return 0;
+}
+
+static int iadc_write(struct iadc_chip *iadc, u16 offset, u8 data)
+{
+	return regmap_write(iadc->regmap, iadc->base + offset, data);
+}
+
+static int iadc_reset(struct iadc_chip *iadc)
+{
+	u8 data;
+	int ret;
+
+	ret = iadc_write(iadc, IADC_SEC_ACCESS, IADC_SEC_ACCESS_DATA);
+	if (ret < 0)
+		return ret;
+
+	ret = iadc_read(iadc, IADC_PERH_RESET_CTL3, &data);
+	if (ret < 0)
+		return ret;
+
+	ret = iadc_write(iadc, IADC_SEC_ACCESS, IADC_SEC_ACCESS_DATA);
+	if (ret < 0)
+		return ret;
+
+	data |= IADC_FOLLOW_WARM_RB;
+
+	return iadc_write(iadc, IADC_PERH_RESET_CTL3, data);
+}
+
+static int iadc_set_state(struct iadc_chip *iadc, bool state)
+{
+	return iadc_write(iadc, IADC_EN_CTL1, state ? IADC_EN_CTL1_SET : 0);
+}
+
+static void iadc_status_show(struct iadc_chip *iadc)
+{
+	u8 mode, sta1, chan, dig, en, req;
+	int ret;
+
+	ret = iadc_read(iadc, IADC_MODE_CTL, &mode);
+	if (ret < 0)
+		return;
+
+	ret = iadc_read(iadc, IADC_DIG_PARAM, &dig);
+	if (ret < 0)
+		return;
+
+	ret = iadc_read(iadc, IADC_CH_SEL_CTL, &chan);
+	if (ret < 0)
+		return;
+
+	ret = iadc_read(iadc, IADC_CONV_REQ, &req);
+	if (ret < 0)
+		return;
+
+	ret = iadc_read(iadc, IADC_STATUS1, &sta1);
+	if (ret < 0)
+		return;
+
+	ret = iadc_read(iadc, IADC_EN_CTL1, &en);
+	if (ret < 0)
+		return;
+
+	dev_err(iadc->dev,
+		"mode:%02x en:%02x chan:%02x dig:%02x req:%02x sta1:%02x\n",
+		mode, en, chan, dig, req, sta1);
+}
+
+static int iadc_configure(struct iadc_chip *iadc, int channel)
+{
+	u8 decim, mode;
+	int ret;
+
+	/* Mode selection */
+	mode = (IADC_OP_MODE_NORMAL << IADC_OP_MODE_SHIFT) | IADC_TRIM_EN;
+	ret = iadc_write(iadc, IADC_MODE_CTL, mode);
+	if (ret < 0)
+		return ret;
+
+	/* Channel selection */
+	ret = iadc_write(iadc, IADC_CH_SEL_CTL, channel);
+	if (ret < 0)
+		return ret;
+
+	/* Digital parameter setup */
+	decim = IADC_DEF_DECIMATION << IADC_DIG_DEC_RATIO_SEL_SHIFT;
+	ret = iadc_write(iadc, IADC_DIG_PARAM, decim);
+	if (ret < 0)
+		return ret;
+
+	/* HW settle time delay */
+	ret = iadc_write(iadc, IADC_HW_SETTLE_DELAY, IADC_DEF_HW_SETTLE_TIME);
+	if (ret < 0)
+		return ret;
+
+	ret = iadc_write(iadc, IADC_FAST_AVG_CTL, IADC_DEF_AVG_SAMPLES);
+	if (ret < 0)
+		return ret;
+
+	if (IADC_DEF_AVG_SAMPLES)
+		ret = iadc_write(iadc, IADC_FAST_AVG_EN, IADC_FAST_AVG_EN_SET);
+	else
+		ret = iadc_write(iadc, IADC_FAST_AVG_EN, 0);
+
+	if (ret < 0)
+		return ret;
+
+	if (!iadc->poll_eoc)
+		reinit_completion(&iadc->complete);
+
+	ret = iadc_set_state(iadc, true);
+	if (ret < 0)
+		return ret;
+
+	/* Request conversion */
+	return iadc_write(iadc, IADC_CONV_REQ, IADC_CONV_REQ_SET);
+}
+
+static int iadc_poll_wait_eoc(struct iadc_chip *iadc, unsigned int interval_us)
+{
+	unsigned int count, retry;
+	int ret;
+	u8 sta1;
+
+	retry = interval_us / IADC_CONV_TIME_MIN_US;
+
+	for (count = 0; count < retry; count++) {
+		ret = iadc_read(iadc, IADC_STATUS1, &sta1);
+		if (ret < 0)
+			return ret;
+
+		sta1 &= IADC_STATUS1_REQ_STS_EOC_MASK;
+		if (sta1 == IADC_STATUS1_EOC)
+			return 0;
+
+		usleep_range(IADC_CONV_TIME_MIN_US, IADC_CONV_TIME_MAX_US);
+	}
+
+	iadc_status_show(iadc);
+
+	return -ETIMEDOUT;
+}
+
+static int iadc_read_result(struct iadc_chip *iadc, u16 *data)
+{
+	return regmap_bulk_read(iadc->regmap, iadc->base + IADC_DATA, data, 2);
+}
+
+static int iadc_do_conversion(struct iadc_chip *iadc, int chan, u16 *data)
+{
+	unsigned int wait;
+	int ret;
+
+	ret = iadc_configure(iadc, chan);
+	if (ret < 0)
+		goto exit;
+
+	wait = BIT(IADC_DEF_AVG_SAMPLES) * IADC_CONV_TIME_MIN_US * 2;
+
+	if (iadc->poll_eoc) {
+		ret = iadc_poll_wait_eoc(iadc, wait);
+	} else {
+		ret = wait_for_completion_timeout(&iadc->complete, wait);
+		if (!ret)
+			ret = -ETIMEDOUT;
+		else
+			/* double check conversion status */
+			ret = iadc_poll_wait_eoc(iadc, IADC_CONV_TIME_MIN_US);
+	}
+
+	if (!ret)
+		ret = iadc_read_result(iadc, data);
+exit:
+	iadc_set_state(iadc, false);
+	if (ret < 0)
+		dev_err(iadc->dev, "conversion failed\n");
+
+	return ret;
+}
+
+static int iadc_read_raw(struct iio_dev *indio_dev,
+			 struct iio_chan_spec const *chan,
+			 int *val, int *val2, long mask)
+{
+	struct iadc_chip *iadc = iio_priv(indio_dev);
+	s32 isense_ua, vsense_uv;
+	u16 adc_raw, vsense_raw;
+	int ret;
+
+	switch (mask) {
+	case IIO_CHAN_INFO_RAW:
+		mutex_lock(&iadc->lock);
+		ret = iadc_do_conversion(iadc, chan->channel, &adc_raw);
+		mutex_unlock(&iadc->lock);
+		if (ret < 0)
+			return ret;
+
+		vsense_raw = adc_raw - iadc->offset[chan->channel];
+
+		vsense_uv = vsense_raw * IADC_REF_GAIN_MICRO_VOLTS;
+		vsense_uv /= (s32)iadc->gain - iadc->offset[chan->channel];
+
+		isense_ua = vsense_uv / iadc->rsense[chan->channel];
+
+		dev_dbg(iadc->dev, "off %d gain %d adc %d %duV I %duA\n",
+			iadc->offset[chan->channel], iadc->gain,
+			adc_raw, vsense_uv, isense_ua);
+
+		*val = isense_ua;
+		return IIO_VAL_INT;
+	case IIO_CHAN_INFO_SCALE:
+		*val = 0;
+		*val2 = 1000;
+		return IIO_VAL_INT_PLUS_MICRO;
+	}
+
+	return -EINVAL;
+}
+
+static const struct iio_info iadc_info = {
+	.read_raw = iadc_read_raw,
+	.driver_module = THIS_MODULE,
+};
+
+static irqreturn_t iadc_isr(int irq, void *dev_id)
+{
+	struct iadc_chip *iadc = dev_id;
+
+	complete(&iadc->complete);
+
+	return IRQ_HANDLED;
+}
+
+static int iadc_update_offset(struct iadc_chip *iadc)
+{
+	int ret;
+
+	ret = iadc_do_conversion(iadc, IADC_GAIN_17P857MV, &iadc->gain);
+	if (ret < 0)
+		return ret;
+
+	ret = iadc_do_conversion(iadc, IADC_INT_OFFSET_CSP2_CSN2,
+				 &iadc->offset[IADC_INT_RSENSE]);
+	if (ret < 0)
+		return ret;
+
+	if (iadc->gain == iadc->offset[IADC_INT_RSENSE]) {
+		dev_err(iadc->dev, "error: internal offset == gain %d\n",
+			iadc->gain);
+		return -EINVAL;
+	}
+
+	ret = iadc_do_conversion(iadc, IADC_EXT_OFFSET_CSP_CSN,
+				 &iadc->offset[IADC_EXT_RSENSE]);
+	if (ret < 0)
+		return ret;
+
+	if (iadc->gain == iadc->offset[IADC_EXT_RSENSE]) {
+		dev_err(iadc->dev, "error: external offset == gain %d\n",
+			iadc->gain);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int iadc_version_check(struct iadc_chip *iadc)
+{
+	u8 val;
+	int ret;
+
+	ret = iadc_read(iadc, IADC_PERPH_TYPE, &val);
+	if (ret < 0)
+		return ret;
+
+	if (val < IADC_PERPH_TYPE_ADC) {
+		dev_err(iadc->dev, "%d is not ADC\n", val);
+		return -EINVAL;
+	}
+
+	ret = iadc_read(iadc, IADC_PERPH_SUBTYPE, &val);
+	if (ret < 0)
+		return ret;
+
+	if (val < IADC_PERPH_SUBTYPE_IADC) {
+		dev_err(iadc->dev, "%d is not IADC\n", val);
+		return -EINVAL;
+	}
+
+	ret = iadc_read(iadc, IADC_REVISION2, &val);
+	if (ret < 0)
+		return ret;
+
+	if (val < IADC_REVISION2_SUPPORTED_IADC) {
+		dev_err(iadc->dev, "revision %d not supported\n", val);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int iadc_rsense_read(struct iadc_chip *iadc, struct device_node *node)
+{
+	int ret, sign, int_sense;
+	u8 deviation;
+
+	ret = of_property_read_u32(node, "qcom,external-resistor-micro-ohms",
+				   &iadc->rsense[IADC_EXT_RSENSE]);
+	if (ret < 0)
+		iadc->rsense[IADC_EXT_RSENSE] = IADC_INT_RSENSE_IDEAL_VALUE;
+
+	if (!iadc->rsense[IADC_EXT_RSENSE]) {
+		dev_err(iadc->dev, "external resistor can't be zero Ohms");
+		return -EINVAL;
+	}
+
+	ret = iadc_read(iadc, IADC_NOMINAL_RSENSE, &deviation);
+	if (ret < 0)
+		return ret;
+
+	/*
+	 * Deviation value stored is an offset from 10 mili Ohms, bit 7 is
+	 * the sign, the remaining bits have an LSB of 15625 nano Ohms.
+	 */
+	sign = (deviation & IADC_NOMINAL_RSENSE_SIGN_MASK) ? -1 : 1;
+
+	deviation &= ~IADC_NOMINAL_RSENSE_SIGN_MASK;
+
+	/* Scale it to nono Ohms */
+	int_sense = IADC_INT_RSENSE_IDEAL_VALUE * 1000;
+	int_sense += sign * deviation * IADC_INT_RSENSE_DEVIATION;
+	int_sense /= 1000; /* micro Ohms */
+
+	iadc->rsense[IADC_INT_RSENSE] = int_sense;
+	return 0;
+}
+
+static const struct iio_chan_spec iadc_channels[] = {
+	{
+		.type = IIO_CURRENT,
+		.datasheet_name	= "INTERNAL_RSENSE",
+		.channel = 0,
+		.info_mask_separate = BIT(IIO_CHAN_INFO_RAW) |
+				      BIT(IIO_CHAN_INFO_SCALE),
+		.indexed = 1,
+	},
+	{
+		.type = IIO_CURRENT,
+		.datasheet_name	= "EXTERNAL_RSENSE",
+		.channel = 1,
+		.info_mask_separate = BIT(IIO_CHAN_INFO_RAW) |
+				      BIT(IIO_CHAN_INFO_SCALE),
+		.indexed = 1,
+	},
+};
+
+static int iadc_probe(struct platform_device *pdev)
+{
+	struct device_node *node = pdev->dev.of_node;
+	struct device *dev = &pdev->dev;
+	struct iio_dev *indio_dev;
+	struct iadc_chip *iadc;
+	int ret, irq_eoc;
+	u32 res;
+
+	indio_dev = devm_iio_device_alloc(dev, sizeof(*iadc));
+	if (!indio_dev)
+		return -ENOMEM;
+
+	iadc = iio_priv(indio_dev);
+	iadc->dev = dev;
+
+	iadc->regmap = dev_get_regmap(dev->parent, NULL);
+	if (!iadc->regmap)
+		return -ENODEV;
+
+	init_completion(&iadc->complete);
+	mutex_init(&iadc->lock);
+
+	ret = of_property_read_u32(node, "reg", &res);
+	if (ret < 0)
+		return -ENODEV;
+
+	iadc->base = res;
+
+	ret = iadc_version_check(iadc);
+	if (ret < 0)
+		return -ENODEV;
+
+	ret = iadc_rsense_read(iadc, node);
+	if (ret < 0)
+		return -ENODEV;
+
+	dev_dbg(iadc->dev, "sense resistors %d and %d micro Ohm\n",
+		iadc->rsense[IADC_INT_RSENSE],
+		iadc->rsense[IADC_EXT_RSENSE]);
+
+	irq_eoc = platform_get_irq(pdev, 0);
+	if (irq_eoc == -EPROBE_DEFER)
+		return irq_eoc;
+
+	if (irq_eoc < 0)
+		iadc->poll_eoc = true;
+
+	ret = iadc_reset(iadc);
+	if (ret < 0) {
+		dev_err(dev, "reset failed\n");
+		return ret;
+	}
+
+	if (!iadc->poll_eoc) {
+		ret = devm_request_irq(dev, irq_eoc, iadc_isr, 0,
+					"spmi-iadc", iadc);
+		if (!ret)
+			enable_irq_wake(irq_eoc);
+		else
+			return ret;
+	} else {
+		device_init_wakeup(iadc->dev, 1);
+	}
+
+	ret = iadc_update_offset(iadc);
+	if (ret < 0) {
+		dev_err(dev, "failed offset calibration\n");
+		return ret;
+	}
+
+	indio_dev->dev.parent = dev;
+	indio_dev->dev.of_node = node;
+	indio_dev->name = pdev->name;
+	indio_dev->modes = INDIO_DIRECT_MODE;
+	indio_dev->info = &iadc_info;
+	indio_dev->channels = iadc_channels;
+	indio_dev->num_channels = ARRAY_SIZE(iadc_channels);
+
+	return devm_iio_device_register(dev, indio_dev);
+}
+
+static const struct of_device_id iadc_match_table[] = {
+	{ .compatible = "qcom,spmi-iadc" },
+	{ }
+};
+
+MODULE_DEVICE_TABLE(of, iadc_match_table);
+
+static struct platform_driver iadc_driver = {
+	.driver = {
+		   .name = "qcom-spmi-iadc",
+		   .of_match_table = iadc_match_table,
+	},
+	.probe = iadc_probe,
+};
+
+module_platform_driver(iadc_driver);
+
+MODULE_ALIAS("platform:qcom-spmi-iadc");
+MODULE_DESCRIPTION("Qualcomm SPMI PMIC current ADC driver");
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR("Ivan T. Ivanov <iivanov@mm-sol.com>");
diff --git a/drivers/iio/adc/rockchip_saradc.c b/drivers/iio/adc/rockchip_saradc.c
new file mode 100644
index 000000000000..8d4e019ea4ca
--- /dev/null
+++ b/drivers/iio/adc/rockchip_saradc.c
@@ -0,0 +1,351 @@
+/*
+ * Rockchip Successive Approximation Register (SAR) A/D Converter
+ * Copyright (C) 2014 ROCKCHIP, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/clk.h>
+#include <linux/completion.h>
+#include <linux/regulator/consumer.h>
+#include <linux/iio/iio.h>
+
+#define SARADC_DATA			0x00
+
+#define SARADC_STAS			0x04
+#define SARADC_STAS_BUSY		BIT(0)
+
+#define SARADC_CTRL			0x08
+#define SARADC_CTRL_IRQ_STATUS		BIT(6)
+#define SARADC_CTRL_IRQ_ENABLE		BIT(5)
+#define SARADC_CTRL_POWER_CTRL		BIT(3)
+#define SARADC_CTRL_CHN_MASK		0x7
+
+#define SARADC_DLY_PU_SOC		0x0c
+#define SARADC_DLY_PU_SOC_MASK		0x3f
+
+#define SARADC_TIMEOUT			msecs_to_jiffies(100)
+
+struct rockchip_saradc_data {
+	int				num_bits;
+	const struct iio_chan_spec	*channels;
+	int				num_channels;
+	unsigned long			clk_rate;
+};
+
+struct rockchip_saradc {
+	void __iomem		*regs;
+	struct clk		*pclk;
+	struct clk		*clk;
+	struct completion	completion;
+	struct regulator	*vref;
+	const struct rockchip_saradc_data *data;
+	u16			last_val;
+};
+
+static int rockchip_saradc_read_raw(struct iio_dev *indio_dev,
+				    struct iio_chan_spec const *chan,
+				    int *val, int *val2, long mask)
+{
+	struct rockchip_saradc *info = iio_priv(indio_dev);
+	int ret;
+
+	switch (mask) {
+	case IIO_CHAN_INFO_RAW:
+		mutex_lock(&indio_dev->mlock);
+
+		reinit_completion(&info->completion);
+
+		/* 8 clock periods as delay between power up and start cmd */
+		writel_relaxed(8, info->regs + SARADC_DLY_PU_SOC);
+
+		/* Select the channel to be used and trigger conversion */
+		writel(SARADC_CTRL_POWER_CTRL
+				| (chan->channel & SARADC_CTRL_CHN_MASK)
+				| SARADC_CTRL_IRQ_ENABLE,
+		       info->regs + SARADC_CTRL);
+
+		if (!wait_for_completion_timeout(&info->completion,
+						 SARADC_TIMEOUT)) {
+			writel_relaxed(0, info->regs + SARADC_CTRL);
+			mutex_unlock(&indio_dev->mlock);
+			return -ETIMEDOUT;
+		}
+
+		*val = info->last_val;
+		mutex_unlock(&indio_dev->mlock);
+		return IIO_VAL_INT;
+	case IIO_CHAN_INFO_SCALE:
+		ret = regulator_get_voltage(info->vref);
+		if (ret < 0) {
+			dev_err(&indio_dev->dev, "failed to get voltage\n");
+			return ret;
+		}
+
+		*val = ret / 1000;
+		*val2 = info->data->num_bits;
+		return IIO_VAL_FRACTIONAL_LOG2;
+	default:
+		return -EINVAL;
+	}
+}
+
+static irqreturn_t rockchip_saradc_isr(int irq, void *dev_id)
+{
+	struct rockchip_saradc *info = (struct rockchip_saradc *)dev_id;
+
+	/* Read value */
+	info->last_val = readl_relaxed(info->regs + SARADC_DATA);
+	info->last_val &= GENMASK(info->data->num_bits - 1, 0);
+
+	/* Clear irq & power down adc */
+	writel_relaxed(0, info->regs + SARADC_CTRL);
+
+	complete(&info->completion);
+
+	return IRQ_HANDLED;
+}
+
+static const struct iio_info rockchip_saradc_iio_info = {
+	.read_raw = rockchip_saradc_read_raw,
+	.driver_module = THIS_MODULE,
+};
+
+#define ADC_CHANNEL(_index, _id) {				\
+	.type = IIO_VOLTAGE,					\
+	.indexed = 1,						\
+	.channel = _index,					\
+	.info_mask_separate = BIT(IIO_CHAN_INFO_RAW),		\
+	.info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE),	\
+	.datasheet_name = _id,					\
+}
+
+static const struct iio_chan_spec rockchip_saradc_iio_channels[] = {
+	ADC_CHANNEL(0, "adc0"),
+	ADC_CHANNEL(1, "adc1"),
+	ADC_CHANNEL(2, "adc2"),
+};
+
+static const struct rockchip_saradc_data saradc_data = {
+	.num_bits = 10,
+	.channels = rockchip_saradc_iio_channels,
+	.num_channels = ARRAY_SIZE(rockchip_saradc_iio_channels),
+	.clk_rate = 1000000,
+};
+
+static const struct iio_chan_spec rockchip_rk3066_tsadc_iio_channels[] = {
+	ADC_CHANNEL(0, "adc0"),
+	ADC_CHANNEL(1, "adc1"),
+};
+
+static const struct rockchip_saradc_data rk3066_tsadc_data = {
+	.num_bits = 12,
+	.channels = rockchip_rk3066_tsadc_iio_channels,
+	.num_channels = ARRAY_SIZE(rockchip_rk3066_tsadc_iio_channels),
+	.clk_rate = 50000,
+};
+
+static const struct of_device_id rockchip_saradc_match[] = {
+	{
+		.compatible = "rockchip,saradc",
+		.data = &saradc_data,
+	}, {
+		.compatible = "rockchip,rk3066-tsadc",
+		.data = &rk3066_tsadc_data,
+	},
+	{},
+};
+MODULE_DEVICE_TABLE(of, rockchip_saradc_match);
+
+static int rockchip_saradc_probe(struct platform_device *pdev)
+{
+	struct rockchip_saradc *info = NULL;
+	struct device_node *np = pdev->dev.of_node;
+	struct iio_dev *indio_dev = NULL;
+	struct resource	*mem;
+	const struct of_device_id *match;
+	int ret;
+	int irq;
+
+	if (!np)
+		return -ENODEV;
+
+	indio_dev = devm_iio_device_alloc(&pdev->dev, sizeof(*info));
+	if (!indio_dev) {
+		dev_err(&pdev->dev, "failed allocating iio device\n");
+		return -ENOMEM;
+	}
+	info = iio_priv(indio_dev);
+
+	match = of_match_device(rockchip_saradc_match, &pdev->dev);
+	info->data = match->data;
+
+	mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	info->regs = devm_ioremap_resource(&pdev->dev, mem);
+	if (IS_ERR(info->regs))
+		return PTR_ERR(info->regs);
+
+	init_completion(&info->completion);
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(&pdev->dev, "no irq resource?\n");
+		return irq;
+	}
+
+	ret = devm_request_irq(&pdev->dev, irq, rockchip_saradc_isr,
+			       0, dev_name(&pdev->dev), info);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed requesting irq %d\n", irq);
+		return ret;
+	}
+
+	info->pclk = devm_clk_get(&pdev->dev, "apb_pclk");
+	if (IS_ERR(info->pclk)) {
+		dev_err(&pdev->dev, "failed to get pclk\n");
+		return PTR_ERR(info->pclk);
+	}
+
+	info->clk = devm_clk_get(&pdev->dev, "saradc");
+	if (IS_ERR(info->clk)) {
+		dev_err(&pdev->dev, "failed to get adc clock\n");
+		return PTR_ERR(info->clk);
+	}
+
+	info->vref = devm_regulator_get(&pdev->dev, "vref");
+	if (IS_ERR(info->vref)) {
+		dev_err(&pdev->dev, "failed to get regulator, %ld\n",
+			PTR_ERR(info->vref));
+		return PTR_ERR(info->vref);
+	}
+
+	/*
+	 * Use a default value for the converter clock.
+	 * This may become user-configurable in the future.
+	 */
+	ret = clk_set_rate(info->clk, info->data->clk_rate);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to set adc clk rate, %d\n", ret);
+		return ret;
+	}
+
+	ret = regulator_enable(info->vref);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to enable vref regulator\n");
+		return ret;
+	}
+
+	ret = clk_prepare_enable(info->pclk);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to enable pclk\n");
+		goto err_reg_voltage;
+	}
+
+	ret = clk_prepare_enable(info->clk);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to enable converter clock\n");
+		goto err_pclk;
+	}
+
+	platform_set_drvdata(pdev, indio_dev);
+
+	indio_dev->name = dev_name(&pdev->dev);
+	indio_dev->dev.parent = &pdev->dev;
+	indio_dev->dev.of_node = pdev->dev.of_node;
+	indio_dev->info = &rockchip_saradc_iio_info;
+	indio_dev->modes = INDIO_DIRECT_MODE;
+
+	indio_dev->channels = info->data->channels;
+	indio_dev->num_channels = info->data->num_channels;
+
+	ret = iio_device_register(indio_dev);
+	if (ret)
+		goto err_clk;
+
+	return 0;
+
+err_clk:
+	clk_disable_unprepare(info->clk);
+err_pclk:
+	clk_disable_unprepare(info->pclk);
+err_reg_voltage:
+	regulator_disable(info->vref);
+	return ret;
+}
+
+static int rockchip_saradc_remove(struct platform_device *pdev)
+{
+	struct iio_dev *indio_dev = platform_get_drvdata(pdev);
+	struct rockchip_saradc *info = iio_priv(indio_dev);
+
+	iio_device_unregister(indio_dev);
+	clk_disable_unprepare(info->clk);
+	clk_disable_unprepare(info->pclk);
+	regulator_disable(info->vref);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int rockchip_saradc_suspend(struct device *dev)
+{
+	struct iio_dev *indio_dev = dev_get_drvdata(dev);
+	struct rockchip_saradc *info = iio_priv(indio_dev);
+
+	clk_disable_unprepare(info->clk);
+	clk_disable_unprepare(info->pclk);
+	regulator_disable(info->vref);
+
+	return 0;
+}
+
+static int rockchip_saradc_resume(struct device *dev)
+{
+	struct iio_dev *indio_dev = dev_get_drvdata(dev);
+	struct rockchip_saradc *info = iio_priv(indio_dev);
+	int ret;
+
+	ret = regulator_enable(info->vref);
+	if (ret)
+		return ret;
+
+	ret = clk_prepare_enable(info->pclk);
+	if (ret)
+		return ret;
+
+	ret = clk_prepare_enable(info->clk);
+	if (ret)
+		return ret;
+
+	return ret;
+}
+#endif
+
+static SIMPLE_DEV_PM_OPS(rockchip_saradc_pm_ops,
+			 rockchip_saradc_suspend, rockchip_saradc_resume);
+
+static struct platform_driver rockchip_saradc_driver = {
+	.probe		= rockchip_saradc_probe,
+	.remove		= rockchip_saradc_remove,
+	.driver		= {
+		.name	= "rockchip-saradc",
+		.of_match_table = rockchip_saradc_match,
+		.pm	= &rockchip_saradc_pm_ops,
+	},
+};
+
+module_platform_driver(rockchip_saradc_driver);
diff --git a/drivers/iio/adc/ti-adc128s052.c b/drivers/iio/adc/ti-adc128s052.c
new file mode 100644
index 000000000000..655cb564ec54
--- /dev/null
+++ b/drivers/iio/adc/ti-adc128s052.c
@@ -0,0 +1,179 @@
+/*
+ * Copyright (C) 2014 Angelo Compagnucci <angelo.compagnucci@gmail.com>
+ *
+ * Driver for Texas Instruments' ADC128S052 ADC chip.
+ * Datasheet can be found here:
+ * http://www.ti.com/lit/ds/symlink/adc128s052.pdf
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/err.h>
+#include <linux/spi/spi.h>
+#include <linux/module.h>
+#include <linux/iio/iio.h>
+#include <linux/regulator/consumer.h>
+
+struct adc128 {
+	struct spi_device *spi;
+
+	struct regulator *reg;
+	struct mutex lock;
+
+	u8 buffer[2] ____cacheline_aligned;
+};
+
+static int adc128_adc_conversion(struct adc128 *adc, u8 channel)
+{
+	int ret;
+
+	mutex_lock(&adc->lock);
+
+	adc->buffer[0] = channel << 3;
+	adc->buffer[1] = 0;
+
+	ret = spi_write(adc->spi, &adc->buffer, 2);
+	if (ret < 0) {
+		mutex_unlock(&adc->lock);
+		return ret;
+	}
+
+	ret = spi_read(adc->spi, &adc->buffer, 2);
+
+	mutex_unlock(&adc->lock);
+
+	if (ret < 0)
+		return ret;
+
+	return ((adc->buffer[0] << 8 | adc->buffer[1]) & 0xFFF);
+}
+
+static int adc128_read_raw(struct iio_dev *indio_dev,
+			   struct iio_chan_spec const *channel, int *val,
+			   int *val2, long mask)
+{
+	struct adc128 *adc = iio_priv(indio_dev);
+	int ret;
+
+	switch (mask) {
+	case IIO_CHAN_INFO_RAW:
+
+		ret = adc128_adc_conversion(adc, channel->channel);
+		if (ret < 0)
+			return ret;
+
+		*val = ret;
+		return IIO_VAL_INT;
+
+	case IIO_CHAN_INFO_SCALE:
+
+		ret = regulator_get_voltage(adc->reg);
+		if (ret < 0)
+			return ret;
+
+		*val = ret / 1000;
+		*val2 = 12;
+		return IIO_VAL_FRACTIONAL_LOG2;
+
+	default:
+		return -EINVAL;
+	}
+
+}
+
+#define ADC128_VOLTAGE_CHANNEL(num)	\
+	{ \
+		.type = IIO_VOLTAGE, \
+		.indexed = 1, \
+		.channel = (num), \
+		.info_mask_separate = BIT(IIO_CHAN_INFO_RAW), \
+		.info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE) \
+	}
+
+static const struct iio_chan_spec adc128_channels[] = {
+	ADC128_VOLTAGE_CHANNEL(0),
+	ADC128_VOLTAGE_CHANNEL(1),
+	ADC128_VOLTAGE_CHANNEL(2),
+	ADC128_VOLTAGE_CHANNEL(3),
+	ADC128_VOLTAGE_CHANNEL(4),
+	ADC128_VOLTAGE_CHANNEL(5),
+	ADC128_VOLTAGE_CHANNEL(6),
+	ADC128_VOLTAGE_CHANNEL(7),
+};
+
+static const struct iio_info adc128_info = {
+	.read_raw = adc128_read_raw,
+	.driver_module = THIS_MODULE,
+};
+
+static int adc128_probe(struct spi_device *spi)
+{
+	struct iio_dev *indio_dev;
+	struct adc128 *adc;
+	int ret;
+
+	indio_dev = devm_iio_device_alloc(&spi->dev, sizeof(*adc));
+	if (!indio_dev)
+		return -ENOMEM;
+
+	adc = iio_priv(indio_dev);
+	adc->spi = spi;
+
+	spi_set_drvdata(spi, indio_dev);
+
+	indio_dev->dev.parent = &spi->dev;
+	indio_dev->name = spi_get_device_id(spi)->name;
+	indio_dev->modes = INDIO_DIRECT_MODE;
+	indio_dev->info = &adc128_info;
+
+	indio_dev->channels = adc128_channels;
+	indio_dev->num_channels = ARRAY_SIZE(adc128_channels);
+
+	adc->reg = devm_regulator_get(&spi->dev, "vref");
+	if (IS_ERR(adc->reg))
+		return PTR_ERR(adc->reg);
+
+	ret = regulator_enable(adc->reg);
+	if (ret < 0)
+		return ret;
+
+	mutex_init(&adc->lock);
+
+	ret = iio_device_register(indio_dev);
+
+	return ret;
+}
+
+static int adc128_remove(struct spi_device *spi)
+{
+	struct iio_dev *indio_dev = spi_get_drvdata(spi);
+	struct adc128 *adc = iio_priv(indio_dev);
+
+	iio_device_unregister(indio_dev);
+	regulator_disable(adc->reg);
+
+	return 0;
+}
+
+static const struct spi_device_id adc128_id[] = {
+	{ "adc128s052", 0},
+	{ }
+};
+MODULE_DEVICE_TABLE(spi, adc128_id);
+
+static struct spi_driver adc128_driver = {
+	.driver = {
+		.name = "adc128s052",
+		.owner = THIS_MODULE,
+	},
+	.probe = adc128_probe,
+	.remove = adc128_remove,
+	.id_table = adc128_id,
+};
+module_spi_driver(adc128_driver);
+
+MODULE_AUTHOR("Angelo Compagnucci <angelo.compagnucci@gmail.com>");
+MODULE_DESCRIPTION("Texas Instruments ADC128S052");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/iio/adc/ti_am335x_adc.c b/drivers/iio/adc/ti_am335x_adc.c
index d5dc4c6ce86c..b730864731e8 100644
--- a/drivers/iio/adc/ti_am335x_adc.c
+++ b/drivers/iio/adc/ti_am335x_adc.c
@@ -545,7 +545,6 @@ MODULE_DEVICE_TABLE(of, ti_adc_dt_ids);
 static struct platform_driver tiadc_driver = {
 	.driver = {
 		.name   = "TI-am335x-adc",
-		.owner	= THIS_MODULE,
 		.pm	= TIADC_PM_OPS,
 		.of_match_table = ti_adc_dt_ids,
 	},
diff --git a/drivers/iio/adc/twl4030-madc.c b/drivers/iio/adc/twl4030-madc.c
index eb86786e698e..94c5f05b4bc1 100644
--- a/drivers/iio/adc/twl4030-madc.c
+++ b/drivers/iio/adc/twl4030-madc.c
@@ -883,7 +883,6 @@ static struct platform_driver twl4030_madc_driver = {
 	.remove = twl4030_madc_remove,
 	.driver = {
 		   .name = "twl4030_madc",
-		   .owner = THIS_MODULE,
 		   .of_match_table = of_match_ptr(twl_madc_of_match),
 	},
 };
diff --git a/drivers/iio/adc/twl6030-gpadc.c b/drivers/iio/adc/twl6030-gpadc.c
index 15282f148b3b..89d8aa1d2818 100644
--- a/drivers/iio/adc/twl6030-gpadc.c
+++ b/drivers/iio/adc/twl6030-gpadc.c
@@ -994,7 +994,6 @@ static struct platform_driver twl6030_gpadc_driver = {
 	.remove		= twl6030_gpadc_remove,
 	.driver		= {
 		.name	= DRIVER_NAME,
-		.owner	= THIS_MODULE,
 		.pm	= &twl6030_gpadc_pm_ops,
 		.of_match_table = of_twl6030_match_tbl,
 	},
diff --git a/drivers/iio/adc/vf610_adc.c b/drivers/iio/adc/vf610_adc.c
index 44799eb5930e..8ec353c01d98 100644
--- a/drivers/iio/adc/vf610_adc.c
+++ b/drivers/iio/adc/vf610_adc.c
@@ -91,7 +91,7 @@
 #define VF610_ADC_CAL			0x80
 
 /* Other field define */
-#define VF610_ADC_ADCHC(x)		((x) & 0xF)
+#define VF610_ADC_ADCHC(x)		((x) & 0x1F)
 #define VF610_ADC_AIEN			(0x1 << 7)
 #define VF610_ADC_CONV_DISABLE		0x1F
 #define VF610_ADC_HS_COCO0		0x1
@@ -153,6 +153,12 @@ struct vf610_adc {
 				BIT(IIO_CHAN_INFO_SAMP_FREQ),	\
 }
 
+#define VF610_ADC_TEMPERATURE_CHAN(_idx, _chan_type) {	\
+	.type = (_chan_type),	\
+	.channel = (_idx),		\
+	.info_mask_separate = BIT(IIO_CHAN_INFO_PROCESSED),	\
+}
+
 static const struct iio_chan_spec vf610_adc_iio_channels[] = {
 	VF610_ADC_CHAN(0, IIO_VOLTAGE),
 	VF610_ADC_CHAN(1, IIO_VOLTAGE),
@@ -170,6 +176,7 @@ static const struct iio_chan_spec vf610_adc_iio_channels[] = {
 	VF610_ADC_CHAN(13, IIO_VOLTAGE),
 	VF610_ADC_CHAN(14, IIO_VOLTAGE),
 	VF610_ADC_CHAN(15, IIO_VOLTAGE),
+	VF610_ADC_TEMPERATURE_CHAN(26, IIO_TEMP),
 	/* sentinel */
 };
 
@@ -451,6 +458,7 @@ static int vf610_read_raw(struct iio_dev *indio_dev,
 
 	switch (mask) {
 	case IIO_CHAN_INFO_RAW:
+	case IIO_CHAN_INFO_PROCESSED:
 		mutex_lock(&indio_dev->mlock);
 		reinit_completion(&info->completion);
 
@@ -468,7 +476,23 @@ static int vf610_read_raw(struct iio_dev *indio_dev,
 			return ret;
 		}
 
-		*val = info->value;
+		switch (chan->type) {
+		case IIO_VOLTAGE:
+			*val = info->value;
+			break;
+		case IIO_TEMP:
+			/*
+			* Calculate in degree Celsius times 1000
+			* Using sensor slope of 1.84 mV/°C and
+			* V at 25°C of 696 mV
+			*/
+			*val = 25000 - ((int)info->value - 864) * 1000000 / 1840;
+			break;
+		default:
+			mutex_unlock(&indio_dev->mlock);
+			return -EINVAL;
+		}
+
 		mutex_unlock(&indio_dev->mlock);
 		return IIO_VAL_INT;
 
@@ -569,9 +593,9 @@ static int vf610_adc_probe(struct platform_device *pdev)
 		return PTR_ERR(info->regs);
 
 	irq = platform_get_irq(pdev, 0);
-	if (irq <= 0) {
+	if (irq < 0) {
 		dev_err(&pdev->dev, "no irq resource?\n");
-		return -EINVAL;
+		return irq;
 	}
 
 	ret = devm_request_irq(info->dev, irq,
@@ -586,8 +610,7 @@ static int vf610_adc_probe(struct platform_device *pdev)
 	if (IS_ERR(info->clk)) {
 		dev_err(&pdev->dev, "failed getting clock, err = %ld\n",
 						PTR_ERR(info->clk));
-		ret = PTR_ERR(info->clk);
-		return ret;
+		return PTR_ERR(info->clk);
 	}
 
 	info->vref = devm_regulator_get(&pdev->dev, "vref");
@@ -681,24 +704,25 @@ static int vf610_adc_resume(struct device *dev)
 
 	ret = clk_prepare_enable(info->clk);
 	if (ret)
-		return ret;
+		goto disable_reg;
 
 	vf610_adc_hw_init(info);
 
 	return 0;
+
+disable_reg:
+	regulator_disable(info->vref);
+	return ret;
 }
 #endif
 
-static SIMPLE_DEV_PM_OPS(vf610_adc_pm_ops,
-			vf610_adc_suspend,
-			vf610_adc_resume);
+static SIMPLE_DEV_PM_OPS(vf610_adc_pm_ops, vf610_adc_suspend, vf610_adc_resume);
 
 static struct platform_driver vf610_adc_driver = {
 	.probe          = vf610_adc_probe,
 	.remove         = vf610_adc_remove,
 	.driver         = {
 		.name   = DRIVER_NAME,
-		.owner  = THIS_MODULE,
 		.of_match_table = vf610_adc_match,
 		.pm     = &vf610_adc_pm_ops,
 	},
diff --git a/drivers/iio/adc/viperboard_adc.c b/drivers/iio/adc/viperboard_adc.c
index 9acf6b6d705b..3be2e35721cc 100644
--- a/drivers/iio/adc/viperboard_adc.c
+++ b/drivers/iio/adc/viperboard_adc.c
@@ -145,7 +145,6 @@ static int vprbrd_adc_probe(struct platform_device *pdev)
 static struct platform_driver vprbrd_adc_driver = {
 	.driver = {
 		.name	= "viperboard-adc",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= vprbrd_adc_probe,
 };
diff --git a/drivers/iio/adc/xilinx-xadc-core.c b/drivers/iio/adc/xilinx-xadc-core.c
index 626b39749767..a221f7329b79 100644
--- a/drivers/iio/adc/xilinx-xadc-core.c
+++ b/drivers/iio/adc/xilinx-xadc-core.c
@@ -1201,12 +1201,16 @@ static int xadc_probe(struct platform_device *pdev)
 			goto err_device_free;
 
 		xadc->convst_trigger = xadc_alloc_trigger(indio_dev, "convst");
-		if (IS_ERR(xadc->convst_trigger))
+		if (IS_ERR(xadc->convst_trigger)) {
+			ret = PTR_ERR(xadc->convst_trigger);
 			goto err_triggered_buffer_cleanup;
+		}
 		xadc->samplerate_trigger = xadc_alloc_trigger(indio_dev,
 			"samplerate");
-		if (IS_ERR(xadc->samplerate_trigger))
+		if (IS_ERR(xadc->samplerate_trigger)) {
+			ret = PTR_ERR(xadc->samplerate_trigger);
 			goto err_free_convst_trigger;
+		}
 	}
 
 	xadc->clk = devm_clk_get(&pdev->dev, NULL);
@@ -1322,7 +1326,6 @@ static struct platform_driver xadc_driver = {
 	.remove = xadc_remove,
 	.driver = {
 		.name = "xadc",
-		.owner = THIS_MODULE,
 		.of_match_table = xadc_of_match_table,
 	},
 };
diff --git a/drivers/iio/common/st_sensors/st_sensors_buffer.c b/drivers/iio/common/st_sensors/st_sensors_buffer.c
index 1665c8e4b62b..e18bc6782256 100644
--- a/drivers/iio/common/st_sensors/st_sensors_buffer.c
+++ b/drivers/iio/common/st_sensors/st_sensors_buffer.c
@@ -71,7 +71,7 @@ int st_sensors_get_buffer_element(struct iio_dev *indio_dev, u8 *buf)
 				goto st_sensors_free_memory;
 			}
 
-			for (i = 0; i < n * num_data_channels; i++) {
+			for (i = 0; i < n * byte_for_channel; i++) {
 				if (i < n)
 					buf[i] = rx_array[i];
 				else
diff --git a/drivers/iio/common/st_sensors/st_sensors_core.c b/drivers/iio/common/st_sensors/st_sensors_core.c
index 8a4ec00a91a0..edd13d2b4121 100644
--- a/drivers/iio/common/st_sensors/st_sensors_core.c
+++ b/drivers/iio/common/st_sensors/st_sensors_core.c
@@ -44,18 +44,18 @@ st_sensors_write_data_with_mask_error:
 	return err;
 }
 
-static int st_sensors_match_odr(struct st_sensors *sensor,
+static int st_sensors_match_odr(struct st_sensor_settings *sensor_settings,
 			unsigned int odr, struct st_sensor_odr_avl *odr_out)
 {
 	int i, ret = -EINVAL;
 
 	for (i = 0; i < ST_SENSORS_ODR_LIST_MAX; i++) {
-		if (sensor->odr.odr_avl[i].hz == 0)
+		if (sensor_settings->odr.odr_avl[i].hz == 0)
 			goto st_sensors_match_odr_error;
 
-		if (sensor->odr.odr_avl[i].hz == odr) {
-			odr_out->hz = sensor->odr.odr_avl[i].hz;
-			odr_out->value = sensor->odr.odr_avl[i].value;
+		if (sensor_settings->odr.odr_avl[i].hz == odr) {
+			odr_out->hz = sensor_settings->odr.odr_avl[i].hz;
+			odr_out->value = sensor_settings->odr.odr_avl[i].value;
 			ret = 0;
 			break;
 		}
@@ -71,23 +71,26 @@ int st_sensors_set_odr(struct iio_dev *indio_dev, unsigned int odr)
 	struct st_sensor_odr_avl odr_out = {0, 0};
 	struct st_sensor_data *sdata = iio_priv(indio_dev);
 
-	err = st_sensors_match_odr(sdata->sensor, odr, &odr_out);
+	err = st_sensors_match_odr(sdata->sensor_settings, odr, &odr_out);
 	if (err < 0)
 		goto st_sensors_match_odr_error;
 
-	if ((sdata->sensor->odr.addr == sdata->sensor->pw.addr) &&
-			(sdata->sensor->odr.mask == sdata->sensor->pw.mask)) {
+	if ((sdata->sensor_settings->odr.addr ==
+					sdata->sensor_settings->pw.addr) &&
+				(sdata->sensor_settings->odr.mask ==
+					sdata->sensor_settings->pw.mask)) {
 		if (sdata->enabled == true) {
 			err = st_sensors_write_data_with_mask(indio_dev,
-				sdata->sensor->odr.addr,
-				sdata->sensor->odr.mask,
+				sdata->sensor_settings->odr.addr,
+				sdata->sensor_settings->odr.mask,
 				odr_out.value);
 		} else {
 			err = 0;
 		}
 	} else {
 		err = st_sensors_write_data_with_mask(indio_dev,
-			sdata->sensor->odr.addr, sdata->sensor->odr.mask,
+			sdata->sensor_settings->odr.addr,
+			sdata->sensor_settings->odr.mask,
 			odr_out.value);
 	}
 	if (err >= 0)
@@ -98,16 +101,16 @@ st_sensors_match_odr_error:
 }
 EXPORT_SYMBOL(st_sensors_set_odr);
 
-static int st_sensors_match_fs(struct st_sensors *sensor,
+static int st_sensors_match_fs(struct st_sensor_settings *sensor_settings,
 					unsigned int fs, int *index_fs_avl)
 {
 	int i, ret = -EINVAL;
 
 	for (i = 0; i < ST_SENSORS_FULLSCALE_AVL_MAX; i++) {
-		if (sensor->fs.fs_avl[i].num == 0)
+		if (sensor_settings->fs.fs_avl[i].num == 0)
 			goto st_sensors_match_odr_error;
 
-		if (sensor->fs.fs_avl[i].num == fs) {
+		if (sensor_settings->fs.fs_avl[i].num == fs) {
 			*index_fs_avl = i;
 			ret = 0;
 			break;
@@ -118,25 +121,24 @@ st_sensors_match_odr_error:
 	return ret;
 }
 
-static int st_sensors_set_fullscale(struct iio_dev *indio_dev,
-								unsigned int fs)
+static int st_sensors_set_fullscale(struct iio_dev *indio_dev, unsigned int fs)
 {
 	int err, i = 0;
 	struct st_sensor_data *sdata = iio_priv(indio_dev);
 
-	err = st_sensors_match_fs(sdata->sensor, fs, &i);
+	err = st_sensors_match_fs(sdata->sensor_settings, fs, &i);
 	if (err < 0)
 		goto st_accel_set_fullscale_error;
 
 	err = st_sensors_write_data_with_mask(indio_dev,
-				sdata->sensor->fs.addr,
-				sdata->sensor->fs.mask,
-				sdata->sensor->fs.fs_avl[i].value);
+				sdata->sensor_settings->fs.addr,
+				sdata->sensor_settings->fs.mask,
+				sdata->sensor_settings->fs.fs_avl[i].value);
 	if (err < 0)
 		goto st_accel_set_fullscale_error;
 
 	sdata->current_fullscale = (struct st_sensor_fullscale_avl *)
-						&sdata->sensor->fs.fs_avl[i];
+					&sdata->sensor_settings->fs.fs_avl[i];
 	return err;
 
 st_accel_set_fullscale_error:
@@ -153,10 +155,12 @@ int st_sensors_set_enable(struct iio_dev *indio_dev, bool enable)
 	struct st_sensor_data *sdata = iio_priv(indio_dev);
 
 	if (enable) {
-		tmp_value = sdata->sensor->pw.value_on;
-		if ((sdata->sensor->odr.addr == sdata->sensor->pw.addr) &&
-			(sdata->sensor->odr.mask == sdata->sensor->pw.mask)) {
-			err = st_sensors_match_odr(sdata->sensor,
+		tmp_value = sdata->sensor_settings->pw.value_on;
+		if ((sdata->sensor_settings->odr.addr ==
+					sdata->sensor_settings->pw.addr) &&
+				(sdata->sensor_settings->odr.mask ==
+					sdata->sensor_settings->pw.mask)) {
+			err = st_sensors_match_odr(sdata->sensor_settings,
 							sdata->odr, &odr_out);
 			if (err < 0)
 				goto set_enable_error;
@@ -164,8 +168,8 @@ int st_sensors_set_enable(struct iio_dev *indio_dev, bool enable)
 			found = true;
 		}
 		err = st_sensors_write_data_with_mask(indio_dev,
-				sdata->sensor->pw.addr,
-				sdata->sensor->pw.mask, tmp_value);
+				sdata->sensor_settings->pw.addr,
+				sdata->sensor_settings->pw.mask, tmp_value);
 		if (err < 0)
 			goto set_enable_error;
 
@@ -175,9 +179,9 @@ int st_sensors_set_enable(struct iio_dev *indio_dev, bool enable)
 			sdata->odr = odr_out.hz;
 	} else {
 		err = st_sensors_write_data_with_mask(indio_dev,
-				sdata->sensor->pw.addr,
-				sdata->sensor->pw.mask,
-				sdata->sensor->pw.value_off);
+				sdata->sensor_settings->pw.addr,
+				sdata->sensor_settings->pw.mask,
+				sdata->sensor_settings->pw.value_off);
 		if (err < 0)
 			goto set_enable_error;
 
@@ -194,8 +198,9 @@ int st_sensors_set_axis_enable(struct iio_dev *indio_dev, u8 axis_enable)
 	struct st_sensor_data *sdata = iio_priv(indio_dev);
 
 	return st_sensors_write_data_with_mask(indio_dev,
-				sdata->sensor->enable_axis.addr,
-				sdata->sensor->enable_axis.mask, axis_enable);
+				sdata->sensor_settings->enable_axis.addr,
+				sdata->sensor_settings->enable_axis.mask,
+				axis_enable);
 }
 EXPORT_SYMBOL(st_sensors_set_axis_enable);
 
@@ -236,13 +241,13 @@ void st_sensors_power_disable(struct iio_dev *indio_dev)
 EXPORT_SYMBOL(st_sensors_power_disable);
 
 static int st_sensors_set_drdy_int_pin(struct iio_dev *indio_dev,
-				       struct st_sensors_platform_data *pdata)
+					struct st_sensors_platform_data *pdata)
 {
 	struct st_sensor_data *sdata = iio_priv(indio_dev);
 
 	switch (pdata->drdy_int_pin) {
 	case 1:
-		if (sdata->sensor->drdy_irq.mask_int1 == 0) {
+		if (sdata->sensor_settings->drdy_irq.mask_int1 == 0) {
 			dev_err(&indio_dev->dev,
 					"DRDY on INT1 not available.\n");
 			return -EINVAL;
@@ -250,7 +255,7 @@ static int st_sensors_set_drdy_int_pin(struct iio_dev *indio_dev,
 		sdata->drdy_int_pin = 1;
 		break;
 	case 2:
-		if (sdata->sensor->drdy_irq.mask_int2 == 0) {
+		if (sdata->sensor_settings->drdy_irq.mask_int2 == 0) {
 			dev_err(&indio_dev->dev,
 					"DRDY on INT2 not available.\n");
 			return -EINVAL;
@@ -306,8 +311,11 @@ int st_sensors_init_sensor(struct iio_dev *indio_dev,
 	if (of_pdata)
 		pdata = of_pdata;
 
-	if (pdata)
+	if (pdata) {
 		err = st_sensors_set_drdy_int_pin(indio_dev, pdata);
+		if (err < 0)
+			return err;
+	}
 
 	err = st_sensors_set_enable(indio_dev, false);
 	if (err < 0)
@@ -315,7 +323,7 @@ int st_sensors_init_sensor(struct iio_dev *indio_dev,
 
 	if (sdata->current_fullscale) {
 		err = st_sensors_set_fullscale(indio_dev,
-					       sdata->current_fullscale->num);
+						sdata->current_fullscale->num);
 		if (err < 0)
 			return err;
 	} else
@@ -327,7 +335,8 @@ int st_sensors_init_sensor(struct iio_dev *indio_dev,
 
 	/* set BDU */
 	err = st_sensors_write_data_with_mask(indio_dev,
-			sdata->sensor->bdu.addr, sdata->sensor->bdu.mask, true);
+					sdata->sensor_settings->bdu.addr,
+					sdata->sensor_settings->bdu.mask, true);
 	if (err < 0)
 		return err;
 
@@ -343,26 +352,28 @@ int st_sensors_set_dataready_irq(struct iio_dev *indio_dev, bool enable)
 	u8 drdy_mask;
 	struct st_sensor_data *sdata = iio_priv(indio_dev);
 
-	if (!sdata->sensor->drdy_irq.addr)
+	if (!sdata->sensor_settings->drdy_irq.addr)
 		return 0;
 
 	/* Enable/Disable the interrupt generator 1. */
-	if (sdata->sensor->drdy_irq.ig1.en_addr > 0) {
+	if (sdata->sensor_settings->drdy_irq.ig1.en_addr > 0) {
 		err = st_sensors_write_data_with_mask(indio_dev,
-			sdata->sensor->drdy_irq.ig1.en_addr,
-			sdata->sensor->drdy_irq.ig1.en_mask, (int)enable);
+				sdata->sensor_settings->drdy_irq.ig1.en_addr,
+				sdata->sensor_settings->drdy_irq.ig1.en_mask,
+				(int)enable);
 		if (err < 0)
 			goto st_accel_set_dataready_irq_error;
 	}
 
 	if (sdata->drdy_int_pin == 1)
-		drdy_mask = sdata->sensor->drdy_irq.mask_int1;
+		drdy_mask = sdata->sensor_settings->drdy_irq.mask_int1;
 	else
-		drdy_mask = sdata->sensor->drdy_irq.mask_int2;
+		drdy_mask = sdata->sensor_settings->drdy_irq.mask_int2;
 
 	/* Enable/Disable the interrupt generator for data ready. */
 	err = st_sensors_write_data_with_mask(indio_dev,
-			sdata->sensor->drdy_irq.addr, drdy_mask, (int)enable);
+					sdata->sensor_settings->drdy_irq.addr,
+					drdy_mask, (int)enable);
 
 st_accel_set_dataready_irq_error:
 	return err;
@@ -375,8 +386,8 @@ int st_sensors_set_fullscale_by_gain(struct iio_dev *indio_dev, int scale)
 	struct st_sensor_data *sdata = iio_priv(indio_dev);
 
 	for (i = 0; i < ST_SENSORS_FULLSCALE_AVL_MAX; i++) {
-		if ((sdata->sensor->fs.fs_avl[i].gain == scale) &&
-				(sdata->sensor->fs.fs_avl[i].gain != 0)) {
+		if ((sdata->sensor_settings->fs.fs_avl[i].gain == scale) &&
+				(sdata->sensor_settings->fs.fs_avl[i].gain != 0)) {
 			err = 0;
 			break;
 		}
@@ -385,7 +396,7 @@ int st_sensors_set_fullscale_by_gain(struct iio_dev *indio_dev, int scale)
 		goto st_sensors_match_scale_error;
 
 	err = st_sensors_set_fullscale(indio_dev,
-					sdata->sensor->fs.fs_avl[i].num);
+				sdata->sensor_settings->fs.fs_avl[i].num);
 
 st_sensors_match_scale_error:
 	return err;
@@ -436,7 +447,7 @@ int st_sensors_read_info_raw(struct iio_dev *indio_dev,
 		if (err < 0)
 			goto out;
 
-		msleep((sdata->sensor->bootime * 1000) / sdata->odr);
+		msleep((sdata->sensor_settings->bootime * 1000) / sdata->odr);
 		err = st_sensors_read_axis_data(indio_dev, ch, val);
 		if (err < 0)
 			goto out;
@@ -453,7 +464,8 @@ out:
 EXPORT_SYMBOL(st_sensors_read_info_raw);
 
 int st_sensors_check_device_support(struct iio_dev *indio_dev,
-			int num_sensors_list, const struct st_sensors *sensors)
+			int num_sensors_list,
+			const struct st_sensor_settings *sensor_settings)
 {
 	u8 wai;
 	int i, n, err;
@@ -467,23 +479,24 @@ int st_sensors_check_device_support(struct iio_dev *indio_dev,
 	}
 
 	for (i = 0; i < num_sensors_list; i++) {
-		if (sensors[i].wai == wai)
+		if (sensor_settings[i].wai == wai)
 			break;
 	}
 	if (i == num_sensors_list)
 		goto device_not_supported;
 
-	for (n = 0; n < ARRAY_SIZE(sensors[i].sensors_supported); n++) {
+	for (n = 0; n < ARRAY_SIZE(sensor_settings[i].sensors_supported); n++) {
 		if (strcmp(indio_dev->name,
-				&sensors[i].sensors_supported[n][0]) == 0)
+				&sensor_settings[i].sensors_supported[n][0]) == 0)
 			break;
 	}
-	if (n == ARRAY_SIZE(sensors[i].sensors_supported)) {
+	if (n == ARRAY_SIZE(sensor_settings[i].sensors_supported)) {
 		dev_err(&indio_dev->dev, "device name and WhoAmI mismatch.\n");
 		goto sensor_name_mismatch;
 	}
 
-	sdata->sensor = (struct st_sensors *)&sensors[i];
+	sdata->sensor_settings =
+			(struct st_sensor_settings *)&sensor_settings[i];
 
 	return i;
 
@@ -505,11 +518,11 @@ ssize_t st_sensors_sysfs_sampling_frequency_avail(struct device *dev,
 
 	mutex_lock(&indio_dev->mlock);
 	for (i = 0; i < ST_SENSORS_ODR_LIST_MAX; i++) {
-		if (sdata->sensor->odr.odr_avl[i].hz == 0)
+		if (sdata->sensor_settings->odr.odr_avl[i].hz == 0)
 			break;
 
 		len += scnprintf(buf + len, PAGE_SIZE - len, "%d ",
-					sdata->sensor->odr.odr_avl[i].hz);
+				sdata->sensor_settings->odr.odr_avl[i].hz);
 	}
 	mutex_unlock(&indio_dev->mlock);
 	buf[len - 1] = '\n';
@@ -527,11 +540,11 @@ ssize_t st_sensors_sysfs_scale_avail(struct device *dev,
 
 	mutex_lock(&indio_dev->mlock);
 	for (i = 0; i < ST_SENSORS_FULLSCALE_AVL_MAX; i++) {
-		if (sdata->sensor->fs.fs_avl[i].num == 0)
+		if (sdata->sensor_settings->fs.fs_avl[i].num == 0)
 			break;
 
 		len += scnprintf(buf + len, PAGE_SIZE - len, "0.%06u ",
-					sdata->sensor->fs.fs_avl[i].gain);
+				sdata->sensor_settings->fs.fs_avl[i].gain);
 	}
 	mutex_unlock(&indio_dev->mlock);
 	buf[len - 1] = '\n';
diff --git a/drivers/iio/common/st_sensors/st_sensors_i2c.c b/drivers/iio/common/st_sensors/st_sensors_i2c.c
index bb6f3085f57b..98cfee296d46 100644
--- a/drivers/iio/common/st_sensors/st_sensors_i2c.c
+++ b/drivers/iio/common/st_sensors/st_sensors_i2c.c
@@ -72,6 +72,7 @@ void st_sensors_i2c_configure(struct iio_dev *indio_dev,
 	indio_dev->dev.parent = &client->dev;
 	indio_dev->name = client->name;
 
+	sdata->dev = &client->dev;
 	sdata->tf = &st_sensors_tf_i2c;
 	sdata->get_irq_data_ready = st_sensors_i2c_get_irq;
 }
diff --git a/drivers/iio/common/st_sensors/st_sensors_spi.c b/drivers/iio/common/st_sensors/st_sensors_spi.c
index 251baf6abc25..78a6a1ab3ece 100644
--- a/drivers/iio/common/st_sensors/st_sensors_spi.c
+++ b/drivers/iio/common/st_sensors/st_sensors_spi.c
@@ -111,6 +111,7 @@ void st_sensors_spi_configure(struct iio_dev *indio_dev,
 	indio_dev->dev.parent = &spi->dev;
 	indio_dev->name = spi->modalias;
 
+	sdata->dev = &spi->dev;
 	sdata->tf = &st_sensors_tf_spi;
 	sdata->get_irq_data_ready = st_sensors_spi_get_irq;
 }
diff --git a/drivers/iio/dac/Kconfig b/drivers/iio/dac/Kconfig
index f278eff42a4c..2236ea22f98a 100644
--- a/drivers/iio/dac/Kconfig
+++ b/drivers/iio/dac/Kconfig
@@ -152,6 +152,14 @@ config MAX517
 	  This driver can also be built as a module.  If so, the module
 	  will be called max517.
 
+config MAX5821
+	tristate "Maxim MAX5821 DAC driver"
+	depends on I2C
+	depends on OF
+	help
+	  Say yes here to build support for Maxim MAX5821
+	  10 bits DAC.
+
 config MCP4725
 	tristate "MCP4725 DAC driver"
 	depends on I2C
diff --git a/drivers/iio/dac/Makefile b/drivers/iio/dac/Makefile
index 10107640bb46..52be7e1acf16 100644
--- a/drivers/iio/dac/Makefile
+++ b/drivers/iio/dac/Makefile
@@ -17,5 +17,6 @@ obj-$(CONFIG_AD5791) += ad5791.o
 obj-$(CONFIG_AD5686) += ad5686.o
 obj-$(CONFIG_AD7303) += ad7303.o
 obj-$(CONFIG_MAX517) += max517.o
+obj-$(CONFIG_MAX5821) += max5821.o
 obj-$(CONFIG_MCP4725) += mcp4725.o
 obj-$(CONFIG_MCP4922) += mcp4922.o
diff --git a/drivers/iio/dac/max5821.c b/drivers/iio/dac/max5821.c
new file mode 100644
index 000000000000..6e914495b346
--- /dev/null
+++ b/drivers/iio/dac/max5821.c
@@ -0,0 +1,405 @@
+ /*
+  * iio/dac/max5821.c
+  * Copyright (C) 2014 Philippe Reynes
+  *
+  * This program is free software; you can redistribute it and/or modify
+  * it under the terms of the GNU General Public License version 2 as
+  * published by the Free Software Foundation.
+  */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/i2c.h>
+#include <linux/iio/iio.h>
+#include <linux/regulator/consumer.h>
+
+#define MAX5821_MAX_DAC_CHANNELS		2
+
+/* command bytes */
+#define MAX5821_LOAD_DAC_A_IN_REG_B		0x00
+#define MAX5821_LOAD_DAC_B_IN_REG_A		0x10
+#define MAX5821_EXTENDED_COMMAND_MODE		0xf0
+#define MAX5821_READ_DAC_A_COMMAND		0xf1
+#define MAX5821_READ_DAC_B_COMMAND		0xf2
+
+#define MAX5821_EXTENDED_POWER_UP		0x00
+#define MAX5821_EXTENDED_POWER_DOWN_MODE0	0x01
+#define MAX5821_EXTENDED_POWER_DOWN_MODE1	0x02
+#define MAX5821_EXTENDED_POWER_DOWN_MODE2	0x03
+#define MAX5821_EXTENDED_DAC_A			0x04
+#define MAX5821_EXTENDED_DAC_B			0x08
+
+enum max5821_device_ids {
+	ID_MAX5821,
+};
+
+struct max5821_data {
+	struct i2c_client	*client;
+	struct regulator	*vref_reg;
+	unsigned short		vref_mv;
+	bool			powerdown[MAX5821_MAX_DAC_CHANNELS];
+	u8			powerdown_mode[MAX5821_MAX_DAC_CHANNELS];
+	struct mutex		lock;
+};
+
+static const char * const max5821_powerdown_modes[] = {
+	"three_state",
+	"1kohm_to_gnd",
+	"100kohm_to_gnd",
+};
+
+enum {
+	MAX5821_THREE_STATE,
+	MAX5821_1KOHM_TO_GND,
+	MAX5821_100KOHM_TO_GND
+};
+
+static int max5821_get_powerdown_mode(struct iio_dev *indio_dev,
+				      const struct iio_chan_spec *chan)
+{
+	struct max5821_data *st = iio_priv(indio_dev);
+
+	return st->powerdown_mode[chan->channel];
+}
+
+static int max5821_set_powerdown_mode(struct iio_dev *indio_dev,
+				      const struct iio_chan_spec *chan,
+				      unsigned int mode)
+{
+	struct max5821_data *st = iio_priv(indio_dev);
+
+	st->powerdown_mode[chan->channel] = mode;
+
+	return 0;
+}
+
+static const struct iio_enum max5821_powerdown_mode_enum = {
+	.items = max5821_powerdown_modes,
+	.num_items = ARRAY_SIZE(max5821_powerdown_modes),
+	.get = max5821_get_powerdown_mode,
+	.set = max5821_set_powerdown_mode,
+};
+
+static ssize_t max5821_read_dac_powerdown(struct iio_dev *indio_dev,
+					  uintptr_t private,
+					  const struct iio_chan_spec *chan,
+					  char *buf)
+{
+	struct max5821_data *st = iio_priv(indio_dev);
+
+	return sprintf(buf, "%d\n", st->powerdown[chan->channel]);
+}
+
+static int max5821_sync_powerdown_mode(struct max5821_data *data,
+				       const struct iio_chan_spec *chan)
+{
+	u8 outbuf[2];
+
+	outbuf[0] = MAX5821_EXTENDED_COMMAND_MODE;
+
+	if (chan->channel == 0)
+		outbuf[1] = MAX5821_EXTENDED_DAC_A;
+	else
+		outbuf[1] = MAX5821_EXTENDED_DAC_B;
+
+	if (data->powerdown[chan->channel])
+		outbuf[1] |= data->powerdown_mode[chan->channel] + 1;
+	else
+		outbuf[1] |= MAX5821_EXTENDED_POWER_UP;
+
+	return i2c_master_send(data->client, outbuf, 2);
+}
+
+static ssize_t max5821_write_dac_powerdown(struct iio_dev *indio_dev,
+					   uintptr_t private,
+					   const struct iio_chan_spec *chan,
+					   const char *buf, size_t len)
+{
+	struct max5821_data *data = iio_priv(indio_dev);
+	bool powerdown;
+	int ret;
+
+	ret = strtobool(buf, &powerdown);
+	if (ret)
+		return ret;
+
+	data->powerdown[chan->channel] = powerdown;
+
+	ret = max5821_sync_powerdown_mode(data, chan);
+	if (ret < 0)
+		return ret;
+
+	return len;
+}
+
+static const struct iio_chan_spec_ext_info max5821_ext_info[] = {
+	{
+		.name = "powerdown",
+		.read = max5821_read_dac_powerdown,
+		.write = max5821_write_dac_powerdown,
+		.shared = IIO_SEPARATE,
+	},
+	IIO_ENUM("powerdown_mode", IIO_SEPARATE, &max5821_powerdown_mode_enum),
+	IIO_ENUM_AVAILABLE("powerdown_mode", &max5821_powerdown_mode_enum),
+	{ },
+};
+
+#define MAX5821_CHANNEL(chan) {					\
+	.type = IIO_VOLTAGE,					\
+	.indexed = 1,						\
+	.output = 1,						\
+	.channel = (chan),					\
+	.info_mask_separate = BIT(IIO_CHAN_INFO_RAW),		\
+	.info_mask_shared_by_all = BIT(IIO_CHAN_INFO_SCALE),	\
+	.ext_info = max5821_ext_info,				\
+}
+
+static const struct iio_chan_spec max5821_channels[] = {
+	MAX5821_CHANNEL(0),
+	MAX5821_CHANNEL(1)
+};
+
+static const u8 max5821_read_dac_command[] = {
+	MAX5821_READ_DAC_A_COMMAND,
+	MAX5821_READ_DAC_B_COMMAND
+};
+
+static const u8 max5821_load_dac_command[] = {
+	MAX5821_LOAD_DAC_A_IN_REG_B,
+	MAX5821_LOAD_DAC_B_IN_REG_A
+};
+
+static int max5821_get_value(struct iio_dev *indio_dev,
+			     int *val, int channel)
+{
+	struct max5821_data *data = iio_priv(indio_dev);
+	struct i2c_client *client = data->client;
+	u8 outbuf[1];
+	u8 inbuf[2];
+	int ret;
+
+	if ((channel != 0) && (channel != 1))
+		return -EINVAL;
+
+	outbuf[0] = max5821_read_dac_command[channel];
+
+	mutex_lock(&data->lock);
+
+	ret = i2c_master_send(client, outbuf, 1);
+	if (ret < 0) {
+		mutex_unlock(&data->lock);
+		return ret;
+	} else if (ret != 1) {
+		mutex_unlock(&data->lock);
+		return -EIO;
+	}
+
+	ret = i2c_master_recv(client, inbuf, 2);
+	if (ret < 0) {
+		mutex_unlock(&data->lock);
+		return ret;
+	} else if (ret != 2) {
+		mutex_unlock(&data->lock);
+		return -EIO;
+	}
+
+	mutex_unlock(&data->lock);
+
+	*val = ((inbuf[0] & 0x0f) << 6) | (inbuf[1] >> 2);
+
+	return IIO_VAL_INT;
+}
+
+static int max5821_set_value(struct iio_dev *indio_dev,
+			     int val, int channel)
+{
+	struct max5821_data *data = iio_priv(indio_dev);
+	struct i2c_client *client = data->client;
+	u8 outbuf[2];
+	int ret;
+
+	if ((val < 0) || (val > 1023))
+		return -EINVAL;
+
+	if ((channel != 0) && (channel != 1))
+		return -EINVAL;
+
+	outbuf[0] = max5821_load_dac_command[channel];
+	outbuf[0] |= val >> 6;
+	outbuf[1] = (val & 0x3f) << 2;
+
+	ret = i2c_master_send(client, outbuf, 2);
+	if (ret < 0)
+		return ret;
+	else if (ret != 2)
+		return -EIO;
+	else
+		return 0;
+}
+
+static int max5821_read_raw(struct iio_dev *indio_dev,
+			   struct iio_chan_spec const *chan,
+			   int *val, int *val2, long mask)
+{
+	struct max5821_data *data = iio_priv(indio_dev);
+
+	switch (mask) {
+	case IIO_CHAN_INFO_RAW:
+		return max5821_get_value(indio_dev, val, chan->channel);
+	case IIO_CHAN_INFO_SCALE:
+		*val = data->vref_mv;
+		*val2 = 10;
+		return IIO_VAL_FRACTIONAL_LOG2;
+	default:
+		return -EINVAL;
+	}
+}
+
+static int max5821_write_raw(struct iio_dev *indio_dev,
+			     struct iio_chan_spec const *chan,
+			     int val, int val2, long mask)
+{
+	if (val2 != 0)
+		return -EINVAL;
+
+	switch (mask) {
+	case IIO_CHAN_INFO_RAW:
+		return max5821_set_value(indio_dev, val, chan->channel);
+	default:
+		return -EINVAL;
+	}
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int max5821_suspend(struct device *dev)
+{
+	u8 outbuf[2] = { MAX5821_EXTENDED_COMMAND_MODE,
+			 MAX5821_EXTENDED_DAC_A |
+			 MAX5821_EXTENDED_DAC_B |
+			 MAX5821_EXTENDED_POWER_DOWN_MODE2 };
+
+	return i2c_master_send(to_i2c_client(dev), outbuf, 2);
+}
+
+static int max5821_resume(struct device *dev)
+{
+	u8 outbuf[2] = { MAX5821_EXTENDED_COMMAND_MODE,
+			 MAX5821_EXTENDED_DAC_A |
+			 MAX5821_EXTENDED_DAC_B |
+			 MAX5821_EXTENDED_POWER_UP };
+
+	return i2c_master_send(to_i2c_client(dev), outbuf, 2);
+}
+
+static SIMPLE_DEV_PM_OPS(max5821_pm_ops, max5821_suspend, max5821_resume);
+#define MAX5821_PM_OPS (&max5821_pm_ops)
+#else
+#define MAX5821_PM_OPS NULL
+#endif /* CONFIG_PM_SLEEP */
+
+static const struct iio_info max5821_info = {
+	.read_raw = max5821_read_raw,
+	.write_raw = max5821_write_raw,
+	.driver_module = THIS_MODULE,
+};
+
+static int max5821_probe(struct i2c_client *client,
+			const struct i2c_device_id *id)
+{
+	struct max5821_data *data;
+	struct iio_dev *indio_dev;
+	u32 tmp;
+	int ret;
+
+	indio_dev = devm_iio_device_alloc(&client->dev, sizeof(*data));
+	if (!indio_dev)
+		return -ENOMEM;
+	data = iio_priv(indio_dev);
+	i2c_set_clientdata(client, indio_dev);
+	data->client = client;
+	mutex_init(&data->lock);
+
+	/* max5821 start in powerdown mode 100Kohm to ground */
+	for (tmp = 0; tmp < MAX5821_MAX_DAC_CHANNELS; tmp++) {
+		data->powerdown[tmp] = true;
+		data->powerdown_mode[tmp] = MAX5821_100KOHM_TO_GND;
+	}
+
+	data->vref_reg = devm_regulator_get(&client->dev, "vref");
+	if (IS_ERR(data->vref_reg)) {
+		ret = PTR_ERR(data->vref_reg);
+		dev_err(&client->dev,
+			"Failed to get vref regulator: %d\n", ret);
+		goto error_free_reg;
+	}
+
+	ret = regulator_enable(data->vref_reg);
+	if (ret) {
+		dev_err(&client->dev,
+			"Failed to enable vref regulator: %d\n", ret);
+		goto error_free_reg;
+	}
+
+	ret = regulator_get_voltage(data->vref_reg);
+	if (ret < 0) {
+		dev_err(&client->dev,
+			"Failed to get voltage on regulator: %d\n", ret);
+		goto error_disable_reg;
+	}
+
+	data->vref_mv = ret / 1000;
+
+	indio_dev->name = id->name;
+	indio_dev->dev.parent = &client->dev;
+	indio_dev->num_channels = ARRAY_SIZE(max5821_channels);
+	indio_dev->channels = max5821_channels;
+	indio_dev->modes = INDIO_DIRECT_MODE;
+	indio_dev->info = &max5821_info;
+
+	return iio_device_register(indio_dev);
+
+error_disable_reg:
+	regulator_disable(data->vref_reg);
+
+error_free_reg:
+
+	return ret;
+}
+
+static int max5821_remove(struct i2c_client *client)
+{
+	struct iio_dev *indio_dev = i2c_get_clientdata(client);
+	struct max5821_data *data = iio_priv(indio_dev);
+
+	iio_device_unregister(indio_dev);
+	regulator_disable(data->vref_reg);
+
+	return 0;
+}
+
+static const struct i2c_device_id max5821_id[] = {
+	{ "max5821", ID_MAX5821 },
+	{ }
+};
+MODULE_DEVICE_TABLE(i2c, max5821_id);
+
+static const struct of_device_id max5821_of_match[] = {
+	{ .compatible = "maxim,max5821" },
+	{ }
+};
+
+static struct i2c_driver max5821_driver = {
+	.driver = {
+		.name	= "max5821",
+		.pm     = MAX5821_PM_OPS,
+		.owner	= THIS_MODULE,
+	},
+	.probe		= max5821_probe,
+	.remove		= max5821_remove,
+	.id_table	= max5821_id,
+};
+module_i2c_driver(max5821_driver);
+
+MODULE_AUTHOR("Philippe Reynes <tremyfr@yahoo.fr>");
+MODULE_DESCRIPTION("MAX5821 DAC");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/iio/gyro/Kconfig b/drivers/iio/gyro/Kconfig
index ac2d69e34c8c..b3d0e94f72eb 100644
--- a/drivers/iio/gyro/Kconfig
+++ b/drivers/iio/gyro/Kconfig
@@ -50,6 +50,17 @@ config ADXRS450
 	  This driver can also be built as a module.  If so, the module
 	  will be called adxrs450.
 
+config BMG160
+	tristate "BOSCH BMG160 Gyro Sensor"
+	depends on I2C
+	select IIO_TRIGGERED_BUFFER if IIO_BUFFER
+	help
+	  Say yes here to build support for Bosch BMG160 Tri-axis Gyro Sensor
+	  driver. This driver also supports BMI055 gyroscope.
+
+	  This driver can also be built as a module.  If so, the module
+	  will be called bmg160.
+
 config HID_SENSOR_GYRO_3D
 	depends on HID_SENSOR_HUB
 	select IIO_BUFFER
diff --git a/drivers/iio/gyro/Makefile b/drivers/iio/gyro/Makefile
index 2f2752a4ea83..36a38776f739 100644
--- a/drivers/iio/gyro/Makefile
+++ b/drivers/iio/gyro/Makefile
@@ -8,6 +8,7 @@ obj-$(CONFIG_ADIS16130) += adis16130.o
 obj-$(CONFIG_ADIS16136) += adis16136.o
 obj-$(CONFIG_ADIS16260) += adis16260.o
 obj-$(CONFIG_ADXRS450) += adxrs450.o
+obj-$(CONFIG_BMG160) += bmg160.o
 
 obj-$(CONFIG_HID_SENSOR_GYRO_3D) += hid-sensor-gyro-3d.o
 
diff --git a/drivers/iio/gyro/bmg160.c b/drivers/iio/gyro/bmg160.c
new file mode 100644
index 000000000000..60451b328242
--- /dev/null
+++ b/drivers/iio/gyro/bmg160.c
@@ -0,0 +1,1273 @@
+/*
+ * BMG160 Gyro Sensor driver
+ * Copyright (c) 2014, Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ */
+
+#include <linux/module.h>
+#include <linux/i2c.h>
+#include <linux/interrupt.h>
+#include <linux/delay.h>
+#include <linux/slab.h>
+#include <linux/acpi.h>
+#include <linux/gpio/consumer.h>
+#include <linux/pm.h>
+#include <linux/pm_runtime.h>
+#include <linux/iio/iio.h>
+#include <linux/iio/sysfs.h>
+#include <linux/iio/buffer.h>
+#include <linux/iio/trigger.h>
+#include <linux/iio/events.h>
+#include <linux/iio/trigger_consumer.h>
+#include <linux/iio/triggered_buffer.h>
+
+#define BMG160_DRV_NAME		"bmg160"
+#define BMG160_IRQ_NAME		"bmg160_event"
+#define BMG160_GPIO_NAME		"gpio_int"
+
+#define BMG160_REG_CHIP_ID		0x00
+#define BMG160_CHIP_ID_VAL		0x0F
+
+#define BMG160_REG_PMU_LPW		0x11
+#define BMG160_MODE_NORMAL		0x00
+#define BMG160_MODE_DEEP_SUSPEND	0x20
+#define BMG160_MODE_SUSPEND		0x80
+
+#define BMG160_REG_RANGE		0x0F
+
+#define BMG160_RANGE_2000DPS		0
+#define BMG160_RANGE_1000DPS		1
+#define BMG160_RANGE_500DPS		2
+#define BMG160_RANGE_250DPS		3
+#define BMG160_RANGE_125DPS		4
+
+#define BMG160_REG_PMU_BW		0x10
+#define BMG160_NO_FILTER		0
+#define BMG160_DEF_BW			100
+
+#define BMG160_REG_INT_MAP_0		0x17
+#define BMG160_INT_MAP_0_BIT_ANY	BIT(1)
+
+#define BMG160_REG_INT_MAP_1		0x18
+#define BMG160_INT_MAP_1_BIT_NEW_DATA	BIT(0)
+
+#define BMG160_REG_INT_RST_LATCH	0x21
+#define BMG160_INT_MODE_LATCH_RESET	0x80
+#define BMG160_INT_MODE_LATCH_INT	0x0F
+#define BMG160_INT_MODE_NON_LATCH_INT	0x00
+
+#define BMG160_REG_INT_EN_0		0x15
+#define BMG160_DATA_ENABLE_INT		BIT(7)
+
+#define BMG160_REG_INT_EN_1		0x16
+#define BMG160_INT1_BIT_OD		BIT(1)
+
+#define BMG160_REG_XOUT_L		0x02
+#define BMG160_AXIS_TO_REG(axis)	(BMG160_REG_XOUT_L + (axis * 2))
+
+#define BMG160_REG_SLOPE_THRES		0x1B
+#define BMG160_SLOPE_THRES_MASK	0x0F
+
+#define BMG160_REG_MOTION_INTR		0x1C
+#define BMG160_INT_MOTION_X		BIT(0)
+#define BMG160_INT_MOTION_Y		BIT(1)
+#define BMG160_INT_MOTION_Z		BIT(2)
+#define BMG160_ANY_DUR_MASK		0x30
+#define BMG160_ANY_DUR_SHIFT		4
+
+#define BMG160_REG_INT_STATUS_2	0x0B
+#define BMG160_ANY_MOTION_MASK		0x07
+#define BMG160_ANY_MOTION_BIT_X		BIT(0)
+#define BMG160_ANY_MOTION_BIT_Y		BIT(1)
+#define BMG160_ANY_MOTION_BIT_Z		BIT(2)
+
+#define BMG160_REG_TEMP		0x08
+#define BMG160_TEMP_CENTER_VAL		23
+
+#define BMG160_MAX_STARTUP_TIME_MS	80
+
+#define BMG160_AUTO_SUSPEND_DELAY_MS	2000
+
+struct bmg160_data {
+	struct i2c_client *client;
+	struct iio_trigger *dready_trig;
+	struct iio_trigger *motion_trig;
+	struct mutex mutex;
+	s16 buffer[8];
+	u8 bw_bits;
+	u32 dps_range;
+	int ev_enable_state;
+	int slope_thres;
+	bool dready_trigger_on;
+	bool motion_trigger_on;
+	int64_t timestamp;
+};
+
+enum bmg160_axis {
+	AXIS_X,
+	AXIS_Y,
+	AXIS_Z,
+};
+
+static const struct {
+	int val;
+	int bw_bits;
+} bmg160_samp_freq_table[] = { {100, 0x07},
+			       {200, 0x06},
+			       {400, 0x03},
+			       {1000, 0x02},
+			       {2000, 0x01} };
+
+static const struct {
+	int scale;
+	int dps_range;
+} bmg160_scale_table[] = { { 1065, BMG160_RANGE_2000DPS},
+			   { 532, BMG160_RANGE_1000DPS},
+			   { 266, BMG160_RANGE_500DPS},
+			   { 133, BMG160_RANGE_250DPS},
+			   { 66, BMG160_RANGE_125DPS} };
+
+static int bmg160_set_mode(struct bmg160_data *data, u8 mode)
+{
+	int ret;
+
+	ret = i2c_smbus_write_byte_data(data->client,
+					BMG160_REG_PMU_LPW, mode);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_pmu_lpw\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int bmg160_convert_freq_to_bit(int val)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(bmg160_samp_freq_table); ++i) {
+		if (bmg160_samp_freq_table[i].val == val)
+			return bmg160_samp_freq_table[i].bw_bits;
+	}
+
+	return -EINVAL;
+}
+
+static int bmg160_set_bw(struct bmg160_data *data, int val)
+{
+	int ret;
+	int bw_bits;
+
+	bw_bits = bmg160_convert_freq_to_bit(val);
+	if (bw_bits < 0)
+		return bw_bits;
+
+	ret = i2c_smbus_write_byte_data(data->client, BMG160_REG_PMU_BW,
+					bw_bits);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_pmu_bw\n");
+		return ret;
+	}
+
+	data->bw_bits = bw_bits;
+
+	return 0;
+}
+
+static int bmg160_chip_init(struct bmg160_data *data)
+{
+	int ret;
+
+	ret = i2c_smbus_read_byte_data(data->client, BMG160_REG_CHIP_ID);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading reg_chip_id\n");
+		return ret;
+	}
+
+	dev_dbg(&data->client->dev, "Chip Id %x\n", ret);
+	if (ret != BMG160_CHIP_ID_VAL) {
+		dev_err(&data->client->dev, "invalid chip %x\n", ret);
+		return -ENODEV;
+	}
+
+	ret = bmg160_set_mode(data, BMG160_MODE_NORMAL);
+	if (ret < 0)
+		return ret;
+
+	/* Wait upto 500 ms to be ready after changing mode */
+	usleep_range(500, 1000);
+
+	/* Set Bandwidth */
+	ret = bmg160_set_bw(data, BMG160_DEF_BW);
+	if (ret < 0)
+		return ret;
+
+	/* Set Default Range */
+	ret = i2c_smbus_write_byte_data(data->client,
+					BMG160_REG_RANGE,
+					BMG160_RANGE_500DPS);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_range\n");
+		return ret;
+	}
+	data->dps_range = BMG160_RANGE_500DPS;
+
+	ret = i2c_smbus_read_byte_data(data->client, BMG160_REG_SLOPE_THRES);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading reg_slope_thres\n");
+		return ret;
+	}
+	data->slope_thres = ret;
+
+	/* Set default interrupt mode */
+	ret = i2c_smbus_read_byte_data(data->client, BMG160_REG_INT_EN_1);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading reg_int_en_1\n");
+		return ret;
+	}
+	ret &= ~BMG160_INT1_BIT_OD;
+	ret = i2c_smbus_write_byte_data(data->client,
+					BMG160_REG_INT_EN_1, ret);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_int_en_1\n");
+		return ret;
+	}
+
+	ret = i2c_smbus_write_byte_data(data->client,
+					BMG160_REG_INT_RST_LATCH,
+					BMG160_INT_MODE_LATCH_INT |
+					BMG160_INT_MODE_LATCH_RESET);
+	if (ret < 0) {
+		dev_err(&data->client->dev,
+			"Error writing reg_motion_intr\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int bmg160_set_power_state(struct bmg160_data *data, bool on)
+{
+#ifdef CONFIG_PM
+	int ret;
+
+	if (on)
+		ret = pm_runtime_get_sync(&data->client->dev);
+	else {
+		pm_runtime_mark_last_busy(&data->client->dev);
+		ret = pm_runtime_put_autosuspend(&data->client->dev);
+	}
+
+	if (ret < 0) {
+		dev_err(&data->client->dev,
+			"Failed: bmg160_set_power_state for %d\n", on);
+		if (on)
+			pm_runtime_put_noidle(&data->client->dev);
+
+		return ret;
+	}
+#endif
+
+	return 0;
+}
+
+static int bmg160_setup_any_motion_interrupt(struct bmg160_data *data,
+					     bool status)
+{
+	int ret;
+
+	/* Enable/Disable INT_MAP0 mapping */
+	ret = i2c_smbus_read_byte_data(data->client,  BMG160_REG_INT_MAP_0);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading reg_int_map0\n");
+		return ret;
+	}
+	if (status)
+		ret |= BMG160_INT_MAP_0_BIT_ANY;
+	else
+		ret &= ~BMG160_INT_MAP_0_BIT_ANY;
+
+	ret = i2c_smbus_write_byte_data(data->client,
+					BMG160_REG_INT_MAP_0,
+					ret);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_int_map0\n");
+		return ret;
+	}
+
+	/* Enable/Disable slope interrupts */
+	if (status) {
+		/* Update slope thres */
+		ret = i2c_smbus_write_byte_data(data->client,
+						BMG160_REG_SLOPE_THRES,
+						data->slope_thres);
+		if (ret < 0) {
+			dev_err(&data->client->dev,
+				"Error writing reg_slope_thres\n");
+			return ret;
+		}
+
+		ret = i2c_smbus_write_byte_data(data->client,
+						BMG160_REG_MOTION_INTR,
+						BMG160_INT_MOTION_X |
+						BMG160_INT_MOTION_Y |
+						BMG160_INT_MOTION_Z);
+		if (ret < 0) {
+			dev_err(&data->client->dev,
+				"Error writing reg_motion_intr\n");
+			return ret;
+		}
+
+		/*
+		 * New data interrupt is always non-latched,
+		 * which will have higher priority, so no need
+		 * to set latched mode, we will be flooded anyway with INTR
+		 */
+		if (!data->dready_trigger_on) {
+			ret = i2c_smbus_write_byte_data(data->client,
+						BMG160_REG_INT_RST_LATCH,
+						BMG160_INT_MODE_LATCH_INT |
+						BMG160_INT_MODE_LATCH_RESET);
+			if (ret < 0) {
+				dev_err(&data->client->dev,
+					"Error writing reg_rst_latch\n");
+				return ret;
+			}
+		}
+
+		ret = i2c_smbus_write_byte_data(data->client,
+						BMG160_REG_INT_EN_0,
+						BMG160_DATA_ENABLE_INT);
+
+	} else
+		ret = i2c_smbus_write_byte_data(data->client,
+						BMG160_REG_INT_EN_0,
+						0);
+
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_int_en0\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int bmg160_setup_new_data_interrupt(struct bmg160_data *data,
+					   bool status)
+{
+	int ret;
+
+	/* Enable/Disable INT_MAP1 mapping */
+	ret = i2c_smbus_read_byte_data(data->client, BMG160_REG_INT_MAP_1);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading reg_int_map1\n");
+		return ret;
+	}
+
+	if (status)
+		ret |= BMG160_INT_MAP_1_BIT_NEW_DATA;
+	else
+		ret &= ~BMG160_INT_MAP_1_BIT_NEW_DATA;
+
+	ret = i2c_smbus_write_byte_data(data->client,
+					BMG160_REG_INT_MAP_1,
+					ret);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_int_map1\n");
+		return ret;
+	}
+
+	if (status) {
+		ret = i2c_smbus_write_byte_data(data->client,
+						BMG160_REG_INT_RST_LATCH,
+						BMG160_INT_MODE_NON_LATCH_INT |
+						BMG160_INT_MODE_LATCH_RESET);
+		if (ret < 0) {
+			dev_err(&data->client->dev,
+				"Error writing reg_rst_latch\n");
+				return ret;
+		}
+
+		ret = i2c_smbus_write_byte_data(data->client,
+						BMG160_REG_INT_EN_0,
+						BMG160_DATA_ENABLE_INT);
+
+	} else {
+		/* Restore interrupt mode */
+		ret = i2c_smbus_write_byte_data(data->client,
+						BMG160_REG_INT_RST_LATCH,
+						BMG160_INT_MODE_LATCH_INT |
+						BMG160_INT_MODE_LATCH_RESET);
+		if (ret < 0) {
+			dev_err(&data->client->dev,
+				"Error writing reg_rst_latch\n");
+				return ret;
+		}
+
+		ret = i2c_smbus_write_byte_data(data->client,
+						BMG160_REG_INT_EN_0,
+						0);
+	}
+
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_int_en0\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int bmg160_get_bw(struct bmg160_data *data, int *val)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(bmg160_samp_freq_table); ++i) {
+		if (bmg160_samp_freq_table[i].bw_bits == data->bw_bits) {
+			*val = bmg160_samp_freq_table[i].val;
+			return IIO_VAL_INT;
+		}
+	}
+
+	return -EINVAL;
+}
+
+static int bmg160_set_scale(struct bmg160_data *data, int val)
+{
+	int ret, i;
+
+	for (i = 0; i < ARRAY_SIZE(bmg160_scale_table); ++i) {
+		if (bmg160_scale_table[i].scale == val) {
+			ret = i2c_smbus_write_byte_data(
+					data->client,
+					BMG160_REG_RANGE,
+					bmg160_scale_table[i].dps_range);
+			if (ret < 0) {
+				dev_err(&data->client->dev,
+					"Error writing reg_range\n");
+				return ret;
+			}
+			data->dps_range = bmg160_scale_table[i].dps_range;
+			return 0;
+		}
+	}
+
+	return -EINVAL;
+}
+
+static int bmg160_get_temp(struct bmg160_data *data, int *val)
+{
+	int ret;
+
+	mutex_lock(&data->mutex);
+	ret = bmg160_set_power_state(data, true);
+	if (ret < 0) {
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+
+	ret = i2c_smbus_read_byte_data(data->client, BMG160_REG_TEMP);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading reg_temp\n");
+		bmg160_set_power_state(data, false);
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+
+	*val = sign_extend32(ret, 7);
+	ret = bmg160_set_power_state(data, false);
+	mutex_unlock(&data->mutex);
+	if (ret < 0)
+		return ret;
+
+	return IIO_VAL_INT;
+}
+
+static int bmg160_get_axis(struct bmg160_data *data, int axis, int *val)
+{
+	int ret;
+
+	mutex_lock(&data->mutex);
+	ret = bmg160_set_power_state(data, true);
+	if (ret < 0) {
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+
+	ret = i2c_smbus_read_word_data(data->client, BMG160_AXIS_TO_REG(axis));
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading axis %d\n", axis);
+		bmg160_set_power_state(data, false);
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+
+	*val = sign_extend32(ret, 15);
+	ret = bmg160_set_power_state(data, false);
+	mutex_unlock(&data->mutex);
+	if (ret < 0)
+		return ret;
+
+	return IIO_VAL_INT;
+}
+
+static int bmg160_read_raw(struct iio_dev *indio_dev,
+			   struct iio_chan_spec const *chan,
+			   int *val, int *val2, long mask)
+{
+	struct bmg160_data *data = iio_priv(indio_dev);
+	int ret;
+
+	switch (mask) {
+	case IIO_CHAN_INFO_RAW:
+		switch (chan->type) {
+		case IIO_TEMP:
+			return bmg160_get_temp(data, val);
+		case IIO_ANGL_VEL:
+			if (iio_buffer_enabled(indio_dev))
+				return -EBUSY;
+			else
+				return bmg160_get_axis(data, chan->scan_index,
+						       val);
+		default:
+			return -EINVAL;
+		}
+	case IIO_CHAN_INFO_OFFSET:
+		if (chan->type == IIO_TEMP) {
+			*val = BMG160_TEMP_CENTER_VAL;
+			return IIO_VAL_INT;
+		} else
+			return -EINVAL;
+	case IIO_CHAN_INFO_SCALE:
+		*val = 0;
+		switch (chan->type) {
+		case IIO_TEMP:
+			*val2 = 500000;
+			return IIO_VAL_INT_PLUS_MICRO;
+		case IIO_ANGL_VEL:
+		{
+			int i;
+
+			for (i = 0; i < ARRAY_SIZE(bmg160_scale_table); ++i) {
+				if (bmg160_scale_table[i].dps_range ==
+							data->dps_range) {
+					*val2 = bmg160_scale_table[i].scale;
+					return IIO_VAL_INT_PLUS_MICRO;
+				}
+			}
+			return -EINVAL;
+		}
+		default:
+			return -EINVAL;
+		}
+	case IIO_CHAN_INFO_SAMP_FREQ:
+		*val2 = 0;
+		mutex_lock(&data->mutex);
+		ret = bmg160_get_bw(data, val);
+		mutex_unlock(&data->mutex);
+		return ret;
+	default:
+		return -EINVAL;
+	}
+}
+
+static int bmg160_write_raw(struct iio_dev *indio_dev,
+			    struct iio_chan_spec const *chan,
+			    int val, int val2, long mask)
+{
+	struct bmg160_data *data = iio_priv(indio_dev);
+	int ret;
+
+	switch (mask) {
+	case IIO_CHAN_INFO_SAMP_FREQ:
+		mutex_lock(&data->mutex);
+		/*
+		 * Section 4.2 of spec
+		 * In suspend mode, the only supported operations are reading
+		 * registers as well as writing to the (0x14) softreset
+		 * register. Since we will be in suspend mode by default, change
+		 * mode to power on for other writes.
+		 */
+		ret = bmg160_set_power_state(data, true);
+		if (ret < 0) {
+			mutex_unlock(&data->mutex);
+			return ret;
+		}
+		ret = bmg160_set_bw(data, val);
+		if (ret < 0) {
+			bmg160_set_power_state(data, false);
+			mutex_unlock(&data->mutex);
+			return ret;
+		}
+		ret = bmg160_set_power_state(data, false);
+		mutex_unlock(&data->mutex);
+		return ret;
+	case IIO_CHAN_INFO_SCALE:
+		if (val)
+			return -EINVAL;
+
+		mutex_lock(&data->mutex);
+		/* Refer to comments above for the suspend mode ops */
+		ret = bmg160_set_power_state(data, true);
+		if (ret < 0) {
+			mutex_unlock(&data->mutex);
+			return ret;
+		}
+		ret = bmg160_set_scale(data, val2);
+		if (ret < 0) {
+			bmg160_set_power_state(data, false);
+			mutex_unlock(&data->mutex);
+			return ret;
+		}
+		ret = bmg160_set_power_state(data, false);
+		mutex_unlock(&data->mutex);
+		return ret;
+	default:
+		return -EINVAL;
+	}
+
+	return -EINVAL;
+}
+
+static int bmg160_read_event(struct iio_dev *indio_dev,
+			     const struct iio_chan_spec *chan,
+			     enum iio_event_type type,
+			     enum iio_event_direction dir,
+			     enum iio_event_info info,
+			     int *val, int *val2)
+{
+	struct bmg160_data *data = iio_priv(indio_dev);
+
+	*val2 = 0;
+	switch (info) {
+	case IIO_EV_INFO_VALUE:
+		*val = data->slope_thres & BMG160_SLOPE_THRES_MASK;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return IIO_VAL_INT;
+}
+
+static int bmg160_write_event(struct iio_dev *indio_dev,
+			      const struct iio_chan_spec *chan,
+			      enum iio_event_type type,
+			      enum iio_event_direction dir,
+			      enum iio_event_info info,
+			      int val, int val2)
+{
+	struct bmg160_data *data = iio_priv(indio_dev);
+
+	switch (info) {
+	case IIO_EV_INFO_VALUE:
+		if (data->ev_enable_state)
+			return -EBUSY;
+		data->slope_thres &= ~BMG160_SLOPE_THRES_MASK;
+		data->slope_thres |= (val & BMG160_SLOPE_THRES_MASK);
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int bmg160_read_event_config(struct iio_dev *indio_dev,
+				    const struct iio_chan_spec *chan,
+				    enum iio_event_type type,
+				    enum iio_event_direction dir)
+{
+
+	struct bmg160_data *data = iio_priv(indio_dev);
+
+	return data->ev_enable_state;
+}
+
+static int bmg160_write_event_config(struct iio_dev *indio_dev,
+				     const struct iio_chan_spec *chan,
+				     enum iio_event_type type,
+				     enum iio_event_direction dir,
+				     int state)
+{
+	struct bmg160_data *data = iio_priv(indio_dev);
+	int ret;
+
+	if (state && data->ev_enable_state)
+		return 0;
+
+	mutex_lock(&data->mutex);
+
+	if (!state && data->motion_trigger_on) {
+		data->ev_enable_state = 0;
+		mutex_unlock(&data->mutex);
+		return 0;
+	}
+	/*
+	 * We will expect the enable and disable to do operation in
+	 * in reverse order. This will happen here anyway as our
+	 * resume operation uses sync mode runtime pm calls, the
+	 * suspend operation will be delayed by autosuspend delay
+	 * So the disable operation will still happen in reverse of
+	 * enable operation. When runtime pm is disabled the mode
+	 * is always on so sequence doesn't matter
+	 */
+	ret = bmg160_set_power_state(data, state);
+	if (ret < 0) {
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+
+	ret =  bmg160_setup_any_motion_interrupt(data, state);
+	if (ret < 0) {
+		bmg160_set_power_state(data, false);
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+
+	data->ev_enable_state = state;
+	mutex_unlock(&data->mutex);
+
+	return 0;
+}
+
+static int bmg160_validate_trigger(struct iio_dev *indio_dev,
+				   struct iio_trigger *trig)
+{
+	struct bmg160_data *data = iio_priv(indio_dev);
+
+	if (data->dready_trig != trig && data->motion_trig != trig)
+		return -EINVAL;
+
+	return 0;
+}
+
+static IIO_CONST_ATTR_SAMP_FREQ_AVAIL("100 200 400 1000 2000");
+
+static IIO_CONST_ATTR(in_anglvel_scale_available,
+		      "0.001065 0.000532 0.000266 0.000133 0.000066");
+
+static struct attribute *bmg160_attributes[] = {
+	&iio_const_attr_sampling_frequency_available.dev_attr.attr,
+	&iio_const_attr_in_anglvel_scale_available.dev_attr.attr,
+	NULL,
+};
+
+static const struct attribute_group bmg160_attrs_group = {
+	.attrs = bmg160_attributes,
+};
+
+static const struct iio_event_spec bmg160_event = {
+		.type = IIO_EV_TYPE_ROC,
+		.dir = IIO_EV_DIR_EITHER,
+		.mask_shared_by_type = BIT(IIO_EV_INFO_VALUE) |
+				       BIT(IIO_EV_INFO_ENABLE)
+};
+
+#define BMG160_CHANNEL(_axis) {					\
+	.type = IIO_ANGL_VEL,						\
+	.modified = 1,							\
+	.channel2 = IIO_MOD_##_axis,					\
+	.info_mask_separate = BIT(IIO_CHAN_INFO_RAW),			\
+	.info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE) |		\
+				    BIT(IIO_CHAN_INFO_SAMP_FREQ),	\
+	.scan_index = AXIS_##_axis,					\
+	.scan_type = {							\
+		.sign = 's',						\
+		.realbits = 16,					\
+		.storagebits = 16,					\
+	},								\
+	.event_spec = &bmg160_event,					\
+	.num_event_specs = 1						\
+}
+
+static const struct iio_chan_spec bmg160_channels[] = {
+	{
+		.type = IIO_TEMP,
+		.info_mask_separate = BIT(IIO_CHAN_INFO_RAW) |
+				      BIT(IIO_CHAN_INFO_SCALE) |
+				      BIT(IIO_CHAN_INFO_OFFSET),
+		.scan_index = -1,
+	},
+	BMG160_CHANNEL(X),
+	BMG160_CHANNEL(Y),
+	BMG160_CHANNEL(Z),
+	IIO_CHAN_SOFT_TIMESTAMP(3),
+};
+
+static const struct iio_info bmg160_info = {
+	.attrs			= &bmg160_attrs_group,
+	.read_raw		= bmg160_read_raw,
+	.write_raw		= bmg160_write_raw,
+	.read_event_value	= bmg160_read_event,
+	.write_event_value	= bmg160_write_event,
+	.write_event_config	= bmg160_write_event_config,
+	.read_event_config	= bmg160_read_event_config,
+	.validate_trigger	= bmg160_validate_trigger,
+	.driver_module		= THIS_MODULE,
+};
+
+static irqreturn_t bmg160_trigger_handler(int irq, void *p)
+{
+	struct iio_poll_func *pf = p;
+	struct iio_dev *indio_dev = pf->indio_dev;
+	struct bmg160_data *data = iio_priv(indio_dev);
+	int bit, ret, i = 0;
+
+	mutex_lock(&data->mutex);
+	for_each_set_bit(bit, indio_dev->buffer->scan_mask,
+			 indio_dev->masklength) {
+		ret = i2c_smbus_read_word_data(data->client,
+					       BMG160_AXIS_TO_REG(bit));
+		if (ret < 0) {
+			mutex_unlock(&data->mutex);
+			goto err;
+		}
+		data->buffer[i++] = ret;
+	}
+	mutex_unlock(&data->mutex);
+
+	iio_push_to_buffers_with_timestamp(indio_dev, data->buffer,
+					   data->timestamp);
+err:
+	iio_trigger_notify_done(indio_dev->trig);
+
+	return IRQ_HANDLED;
+}
+
+static int bmg160_trig_try_reen(struct iio_trigger *trig)
+{
+	struct iio_dev *indio_dev = iio_trigger_get_drvdata(trig);
+	struct bmg160_data *data = iio_priv(indio_dev);
+	int ret;
+
+	/* new data interrupts don't need ack */
+	if (data->dready_trigger_on)
+		return 0;
+
+	/* Set latched mode interrupt and clear any latched interrupt */
+	ret = i2c_smbus_write_byte_data(data->client,
+					BMG160_REG_INT_RST_LATCH,
+					BMG160_INT_MODE_LATCH_INT |
+					BMG160_INT_MODE_LATCH_RESET);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error writing reg_rst_latch\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int bmg160_data_rdy_trigger_set_state(struct iio_trigger *trig,
+					     bool state)
+{
+	struct iio_dev *indio_dev = iio_trigger_get_drvdata(trig);
+	struct bmg160_data *data = iio_priv(indio_dev);
+	int ret;
+
+	mutex_lock(&data->mutex);
+
+	if (!state && data->ev_enable_state && data->motion_trigger_on) {
+		data->motion_trigger_on = false;
+		mutex_unlock(&data->mutex);
+		return 0;
+	}
+
+	/*
+	 * Refer to comment in bmg160_write_event_config for
+	 * enable/disable operation order
+	 */
+	ret = bmg160_set_power_state(data, state);
+	if (ret < 0) {
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+	if (data->motion_trig == trig)
+		ret =  bmg160_setup_any_motion_interrupt(data, state);
+	else
+		ret = bmg160_setup_new_data_interrupt(data, state);
+	if (ret < 0) {
+		bmg160_set_power_state(data, false);
+		mutex_unlock(&data->mutex);
+		return ret;
+	}
+	if (data->motion_trig == trig)
+		data->motion_trigger_on = state;
+	else
+		data->dready_trigger_on = state;
+
+	mutex_unlock(&data->mutex);
+
+	return 0;
+}
+
+static const struct iio_trigger_ops bmg160_trigger_ops = {
+	.set_trigger_state = bmg160_data_rdy_trigger_set_state,
+	.try_reenable = bmg160_trig_try_reen,
+	.owner = THIS_MODULE,
+};
+
+static irqreturn_t bmg160_event_handler(int irq, void *private)
+{
+	struct iio_dev *indio_dev = private;
+	struct bmg160_data *data = iio_priv(indio_dev);
+	int ret;
+	int dir;
+
+	ret = i2c_smbus_read_byte_data(data->client, BMG160_REG_INT_STATUS_2);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "Error reading reg_int_status2\n");
+		goto ack_intr_status;
+	}
+
+	if (ret & 0x08)
+		dir = IIO_EV_DIR_RISING;
+	else
+		dir = IIO_EV_DIR_FALLING;
+
+	if (ret & BMG160_ANY_MOTION_BIT_X)
+		iio_push_event(indio_dev, IIO_MOD_EVENT_CODE(IIO_ANGL_VEL,
+							0,
+							IIO_MOD_X,
+							IIO_EV_TYPE_ROC,
+							dir),
+							data->timestamp);
+	if (ret & BMG160_ANY_MOTION_BIT_Y)
+		iio_push_event(indio_dev, IIO_MOD_EVENT_CODE(IIO_ANGL_VEL,
+							0,
+							IIO_MOD_Y,
+							IIO_EV_TYPE_ROC,
+							dir),
+							data->timestamp);
+	if (ret & BMG160_ANY_MOTION_BIT_Z)
+		iio_push_event(indio_dev, IIO_MOD_EVENT_CODE(IIO_ANGL_VEL,
+							0,
+							IIO_MOD_Z,
+							IIO_EV_TYPE_ROC,
+							dir),
+							data->timestamp);
+
+ack_intr_status:
+	if (!data->dready_trigger_on) {
+		ret = i2c_smbus_write_byte_data(data->client,
+					BMG160_REG_INT_RST_LATCH,
+					BMG160_INT_MODE_LATCH_INT |
+					BMG160_INT_MODE_LATCH_RESET);
+		if (ret < 0)
+			dev_err(&data->client->dev,
+				"Error writing reg_rst_latch\n");
+	}
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t bmg160_data_rdy_trig_poll(int irq, void *private)
+{
+	struct iio_dev *indio_dev = private;
+	struct bmg160_data *data = iio_priv(indio_dev);
+
+	data->timestamp = iio_get_time_ns();
+
+	if (data->dready_trigger_on)
+		iio_trigger_poll(data->dready_trig);
+	else if (data->motion_trigger_on)
+		iio_trigger_poll(data->motion_trig);
+
+	if (data->ev_enable_state)
+		return IRQ_WAKE_THREAD;
+	else
+		return IRQ_HANDLED;
+
+}
+
+static int bmg160_gpio_probe(struct i2c_client *client,
+			     struct bmg160_data *data)
+
+{
+	struct device *dev;
+	struct gpio_desc *gpio;
+	int ret;
+
+	if (!client)
+		return -EINVAL;
+
+	dev = &client->dev;
+
+	/* data ready gpio interrupt pin */
+	gpio = devm_gpiod_get_index(dev, BMG160_GPIO_NAME, 0);
+	if (IS_ERR(gpio)) {
+		dev_err(dev, "acpi gpio get index failed\n");
+		return PTR_ERR(gpio);
+	}
+
+	ret = gpiod_direction_input(gpio);
+	if (ret)
+		return ret;
+
+	ret = gpiod_to_irq(gpio);
+
+	dev_dbg(dev, "GPIO resource, no:%d irq:%d\n", desc_to_gpio(gpio), ret);
+
+	return ret;
+}
+
+static const char *bmg160_match_acpi_device(struct device *dev)
+{
+	const struct acpi_device_id *id;
+
+	id = acpi_match_device(dev->driver->acpi_match_table, dev);
+	if (!id)
+		return NULL;
+
+	return dev_name(dev);
+}
+
+static int bmg160_probe(struct i2c_client *client,
+			const struct i2c_device_id *id)
+{
+	struct bmg160_data *data;
+	struct iio_dev *indio_dev;
+	int ret;
+	const char *name = NULL;
+
+	indio_dev = devm_iio_device_alloc(&client->dev, sizeof(*data));
+	if (!indio_dev)
+		return -ENOMEM;
+
+	data = iio_priv(indio_dev);
+	i2c_set_clientdata(client, indio_dev);
+	data->client = client;
+
+	ret = bmg160_chip_init(data);
+	if (ret < 0)
+		return ret;
+
+	mutex_init(&data->mutex);
+
+	if (id)
+		name = id->name;
+
+	if (ACPI_HANDLE(&client->dev))
+		name = bmg160_match_acpi_device(&client->dev);
+
+	indio_dev->dev.parent = &client->dev;
+	indio_dev->channels = bmg160_channels;
+	indio_dev->num_channels = ARRAY_SIZE(bmg160_channels);
+	indio_dev->name = name;
+	indio_dev->modes = INDIO_DIRECT_MODE;
+	indio_dev->info = &bmg160_info;
+
+	if (client->irq <= 0)
+		client->irq = bmg160_gpio_probe(client, data);
+
+	if (client->irq > 0) {
+		ret = devm_request_threaded_irq(&client->dev,
+						client->irq,
+						bmg160_data_rdy_trig_poll,
+						bmg160_event_handler,
+						IRQF_TRIGGER_RISING,
+						BMG160_IRQ_NAME,
+						indio_dev);
+		if (ret)
+			return ret;
+
+		data->dready_trig = devm_iio_trigger_alloc(&client->dev,
+							   "%s-dev%d",
+							   indio_dev->name,
+							   indio_dev->id);
+		if (!data->dready_trig)
+			return -ENOMEM;
+
+		data->motion_trig = devm_iio_trigger_alloc(&client->dev,
+							  "%s-any-motion-dev%d",
+							  indio_dev->name,
+							  indio_dev->id);
+		if (!data->motion_trig)
+			return -ENOMEM;
+
+		data->dready_trig->dev.parent = &client->dev;
+		data->dready_trig->ops = &bmg160_trigger_ops;
+		iio_trigger_set_drvdata(data->dready_trig, indio_dev);
+		ret = iio_trigger_register(data->dready_trig);
+		if (ret)
+			return ret;
+
+		data->motion_trig->dev.parent = &client->dev;
+		data->motion_trig->ops = &bmg160_trigger_ops;
+		iio_trigger_set_drvdata(data->motion_trig, indio_dev);
+		ret = iio_trigger_register(data->motion_trig);
+		if (ret) {
+			data->motion_trig = NULL;
+			goto err_trigger_unregister;
+		}
+
+		ret = iio_triggered_buffer_setup(indio_dev,
+						 NULL,
+						 bmg160_trigger_handler,
+						 NULL);
+		if (ret < 0) {
+			dev_err(&client->dev,
+				"iio triggered buffer setup failed\n");
+			goto err_trigger_unregister;
+		}
+	}
+
+	ret = iio_device_register(indio_dev);
+	if (ret < 0) {
+		dev_err(&client->dev, "unable to register iio device\n");
+		goto err_buffer_cleanup;
+	}
+
+	ret = pm_runtime_set_active(&client->dev);
+	if (ret)
+		goto err_iio_unregister;
+
+	pm_runtime_enable(&client->dev);
+	pm_runtime_set_autosuspend_delay(&client->dev,
+					 BMG160_AUTO_SUSPEND_DELAY_MS);
+	pm_runtime_use_autosuspend(&client->dev);
+
+	return 0;
+
+err_iio_unregister:
+	iio_device_unregister(indio_dev);
+err_buffer_cleanup:
+	if (data->dready_trig)
+		iio_triggered_buffer_cleanup(indio_dev);
+err_trigger_unregister:
+	if (data->dready_trig)
+		iio_trigger_unregister(data->dready_trig);
+	if (data->motion_trig)
+		iio_trigger_unregister(data->motion_trig);
+
+	return ret;
+}
+
+static int bmg160_remove(struct i2c_client *client)
+{
+	struct iio_dev *indio_dev = i2c_get_clientdata(client);
+	struct bmg160_data *data = iio_priv(indio_dev);
+
+	pm_runtime_disable(&client->dev);
+	pm_runtime_set_suspended(&client->dev);
+	pm_runtime_put_noidle(&client->dev);
+
+	iio_device_unregister(indio_dev);
+
+	if (data->dready_trig) {
+		iio_triggered_buffer_cleanup(indio_dev);
+		iio_trigger_unregister(data->dready_trig);
+		iio_trigger_unregister(data->motion_trig);
+	}
+
+	mutex_lock(&data->mutex);
+	bmg160_set_mode(data, BMG160_MODE_DEEP_SUSPEND);
+	mutex_unlock(&data->mutex);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int bmg160_suspend(struct device *dev)
+{
+	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
+	struct bmg160_data *data = iio_priv(indio_dev);
+
+	mutex_lock(&data->mutex);
+	bmg160_set_mode(data, BMG160_MODE_SUSPEND);
+	mutex_unlock(&data->mutex);
+
+	return 0;
+}
+
+static int bmg160_resume(struct device *dev)
+{
+	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
+	struct bmg160_data *data = iio_priv(indio_dev);
+
+	mutex_lock(&data->mutex);
+	if (data->dready_trigger_on || data->motion_trigger_on ||
+							data->ev_enable_state)
+		bmg160_set_mode(data, BMG160_MODE_NORMAL);
+	mutex_unlock(&data->mutex);
+
+	return 0;
+}
+#endif
+
+#ifdef CONFIG_PM
+static int bmg160_runtime_suspend(struct device *dev)
+{
+	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
+	struct bmg160_data *data = iio_priv(indio_dev);
+	int ret;
+
+	ret = bmg160_set_mode(data, BMG160_MODE_SUSPEND);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "set mode failed\n");
+		return -EAGAIN;
+	}
+
+	return 0;
+}
+
+static int bmg160_runtime_resume(struct device *dev)
+{
+	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
+	struct bmg160_data *data = iio_priv(indio_dev);
+	int ret;
+
+	ret = bmg160_set_mode(data, BMG160_MODE_NORMAL);
+	if (ret < 0)
+		return ret;
+
+	msleep_interruptible(BMG160_MAX_STARTUP_TIME_MS);
+
+	return 0;
+}
+#endif
+
+static const struct dev_pm_ops bmg160_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(bmg160_suspend, bmg160_resume)
+	SET_RUNTIME_PM_OPS(bmg160_runtime_suspend,
+			   bmg160_runtime_resume, NULL)
+};
+
+static const struct acpi_device_id bmg160_acpi_match[] = {
+	{"BMG0160", 0},
+	{"BMI055B", 0},
+	{},
+};
+
+MODULE_DEVICE_TABLE(acpi, bmg160_acpi_match);
+
+static const struct i2c_device_id bmg160_id[] = {
+	{"bmg160", 0},
+	{"bmi055_gyro", 0},
+	{}
+};
+
+MODULE_DEVICE_TABLE(i2c, bmg160_id);
+
+static struct i2c_driver bmg160_driver = {
+	.driver = {
+		.name	= BMG160_DRV_NAME,
+		.acpi_match_table = ACPI_PTR(bmg160_acpi_match),
+		.pm	= &bmg160_pm_ops,
+	},
+	.probe		= bmg160_probe,
+	.remove		= bmg160_remove,
+	.id_table	= bmg160_id,
+};
+module_i2c_driver(bmg160_driver);
+
+MODULE_AUTHOR("Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>");
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("BMG160 Gyro driver");
diff --git a/drivers/iio/gyro/hid-sensor-gyro-3d.c b/drivers/iio/gyro/hid-sensor-gyro-3d.c
index fa034a3dad78..a3ea1e8785d7 100644
--- a/drivers/iio/gyro/hid-sensor-gyro-3d.c
+++ b/drivers/iio/gyro/hid-sensor-gyro-3d.c
@@ -416,7 +416,6 @@ static struct platform_driver hid_gyro_3d_platform_driver = {
 	.id_table = hid_gyro_3d_ids,
 	.driver = {
 		.name	= KBUILD_MODNAME,
-		.owner	= THIS_MODULE,
 	},
 	.probe		= hid_gyro_3d_probe,
 	.remove		= hid_gyro_3d_remove,
diff --git a/drivers/iio/gyro/st_gyro.h b/drivers/iio/gyro/st_gyro.h
index c197360c450b..5353d6328c54 100644
--- a/drivers/iio/gyro/st_gyro.h
+++ b/drivers/iio/gyro/st_gyro.h
@@ -30,8 +30,7 @@ static const struct st_sensors_platform_data gyro_pdata = {
 	.drdy_int_pin = 2,
 };
 
-int st_gyro_common_probe(struct iio_dev *indio_dev,
-					struct st_sensors_platform_data *pdata);
+int st_gyro_common_probe(struct iio_dev *indio_dev);
 void st_gyro_common_remove(struct iio_dev *indio_dev);
 
 #ifdef CONFIG_IIO_BUFFER
diff --git a/drivers/iio/gyro/st_gyro_core.c b/drivers/iio/gyro/st_gyro_core.c
index f156fc6c5c6c..f07a2336f7dc 100644
--- a/drivers/iio/gyro/st_gyro_core.c
+++ b/drivers/iio/gyro/st_gyro_core.c
@@ -103,7 +103,7 @@ static const struct iio_chan_spec st_gyro_16bit_channels[] = {
 	IIO_CHAN_SOFT_TIMESTAMP(3)
 };
 
-static const struct st_sensors st_gyro_sensors[] = {
+static const struct st_sensor_settings st_gyro_sensors_settings[] = {
 	{
 		.wai = ST_GYRO_1_WAI_EXP,
 		.sensors_supported = {
@@ -309,8 +309,7 @@ static const struct iio_trigger_ops st_gyro_trigger_ops = {
 #define ST_GYRO_TRIGGER_OPS NULL
 #endif
 
-int st_gyro_common_probe(struct iio_dev *indio_dev,
-					struct st_sensors_platform_data *pdata)
+int st_gyro_common_probe(struct iio_dev *indio_dev)
 {
 	struct st_sensor_data *gdata = iio_priv(indio_dev);
 	int irq = gdata->get_irq_data_ready(indio_dev);
@@ -322,20 +321,22 @@ int st_gyro_common_probe(struct iio_dev *indio_dev,
 	st_sensors_power_enable(indio_dev);
 
 	err = st_sensors_check_device_support(indio_dev,
-				ARRAY_SIZE(st_gyro_sensors), st_gyro_sensors);
+					ARRAY_SIZE(st_gyro_sensors_settings),
+					st_gyro_sensors_settings);
 	if (err < 0)
 		return err;
 
 	gdata->num_data_channels = ST_GYRO_NUMBER_DATA_CHANNELS;
-	gdata->multiread_bit = gdata->sensor->multi_read_bit;
-	indio_dev->channels = gdata->sensor->ch;
+	gdata->multiread_bit = gdata->sensor_settings->multi_read_bit;
+	indio_dev->channels = gdata->sensor_settings->ch;
 	indio_dev->num_channels = ST_SENSORS_NUMBER_ALL_CHANNELS;
 
 	gdata->current_fullscale = (struct st_sensor_fullscale_avl *)
-						&gdata->sensor->fs.fs_avl[0];
-	gdata->odr = gdata->sensor->odr.odr_avl[0].hz;
+					&gdata->sensor_settings->fs.fs_avl[0];
+	gdata->odr = gdata->sensor_settings->odr.odr_avl[0].hz;
 
-	err = st_sensors_init_sensor(indio_dev, pdata);
+	err = st_sensors_init_sensor(indio_dev,
+				(struct st_sensors_platform_data *)&gyro_pdata);
 	if (err < 0)
 		return err;
 
diff --git a/drivers/iio/gyro/st_gyro_i2c.c b/drivers/iio/gyro/st_gyro_i2c.c
index 8fa0ad2ef4ef..64480b16c689 100644
--- a/drivers/iio/gyro/st_gyro_i2c.c
+++ b/drivers/iio/gyro/st_gyro_i2c.c
@@ -67,13 +67,11 @@ static int st_gyro_i2c_probe(struct i2c_client *client,
 		return -ENOMEM;
 
 	gdata = iio_priv(indio_dev);
-	gdata->dev = &client->dev;
 	st_sensors_of_i2c_probe(client, st_gyro_of_match);
 
 	st_sensors_i2c_configure(indio_dev, client, gdata);
 
-	err = st_gyro_common_probe(indio_dev,
-				(struct st_sensors_platform_data *)&gyro_pdata);
+	err = st_gyro_common_probe(indio_dev);
 	if (err < 0)
 		return err;
 
diff --git a/drivers/iio/gyro/st_gyro_spi.c b/drivers/iio/gyro/st_gyro_spi.c
index b4ad3be26687..e59bead6bc3c 100644
--- a/drivers/iio/gyro/st_gyro_spi.c
+++ b/drivers/iio/gyro/st_gyro_spi.c
@@ -29,12 +29,10 @@ static int st_gyro_spi_probe(struct spi_device *spi)
 		return -ENOMEM;
 
 	gdata = iio_priv(indio_dev);
-	gdata->dev = &spi->dev;
 
 	st_sensors_spi_configure(indio_dev, spi, gdata);
 
-	err = st_gyro_common_probe(indio_dev,
-				(struct st_sensors_platform_data *)&gyro_pdata);
+	err = st_gyro_common_probe(indio_dev);
 	if (err < 0)
 		return err;
 
diff --git a/drivers/iio/humidity/Kconfig b/drivers/iio/humidity/Kconfig
index e116bd8dd0e4..4813b793b9f7 100644
--- a/drivers/iio/humidity/Kconfig
+++ b/drivers/iio/humidity/Kconfig
@@ -22,4 +22,14 @@ config SI7005
 	  To compile this driver as a module, choose M here: the module
 	  will be called si7005.
 
+config SI7020
+	tristate "Si7013/20/21 Relative Humidity and Temperature Sensors"
+	depends on I2C
+	help
+	  Say yes here to build support for the Silicon Labs Si7013/20/21
+	  Relative Humidity and Temperature Sensors.
+
+	  To compile this driver as a module, choose M here: the module
+	  will be called si7020.
+
 endmenu
diff --git a/drivers/iio/humidity/Makefile b/drivers/iio/humidity/Makefile
index e3f3d942e646..86e2d26e9f4d 100644
--- a/drivers/iio/humidity/Makefile
+++ b/drivers/iio/humidity/Makefile
@@ -4,3 +4,4 @@
 
 obj-$(CONFIG_DHT11) += dht11.o
 obj-$(CONFIG_SI7005) += si7005.o
+obj-$(CONFIG_SI7020) += si7020.o
diff --git a/drivers/iio/humidity/dht11.c b/drivers/iio/humidity/dht11.c
index d8771f546bf2..623c145d8a97 100644
--- a/drivers/iio/humidity/dht11.c
+++ b/drivers/iio/humidity/dht11.c
@@ -281,7 +281,6 @@ static int dht11_probe(struct platform_device *pdev)
 static struct platform_driver dht11_driver = {
 	.driver = {
 		.name	= DRIVER_NAME,
-		.owner	= THIS_MODULE,
 		.of_match_table = dht11_dt_ids,
 	},
 	.probe  = dht11_probe,
diff --git a/drivers/iio/humidity/si7020.c b/drivers/iio/humidity/si7020.c
new file mode 100644
index 000000000000..b54164677b89
--- /dev/null
+++ b/drivers/iio/humidity/si7020.c
@@ -0,0 +1,161 @@
+/*
+ * si7020.c - Silicon Labs Si7013/20/21 Relative Humidity and Temp Sensors
+ * Copyright (c) 2013,2014  Uplogix, Inc.
+ * David Barksdale <dbarksdale@uplogix.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+/*
+ * The Silicon Labs Si7013/20/21 Relative Humidity and Temperature Sensors
+ * are i2c devices which have an identical programming interface for
+ * measuring relative humidity and temperature. The Si7013 has an additional
+ * temperature input which this driver does not support.
+ *
+ * Data Sheets:
+ *   Si7013: http://www.silabs.com/Support%20Documents/TechnicalDocs/Si7013.pdf
+ *   Si7020: http://www.silabs.com/Support%20Documents/TechnicalDocs/Si7020.pdf
+ *   Si7021: http://www.silabs.com/Support%20Documents/TechnicalDocs/Si7021.pdf
+ */
+
+#include <linux/delay.h>
+#include <linux/i2c.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/sysfs.h>
+
+#include <linux/iio/iio.h>
+#include <linux/iio/sysfs.h>
+
+/* Measure Relative Humidity, Hold Master Mode */
+#define SI7020CMD_RH_HOLD	0xE5
+/* Measure Temperature, Hold Master Mode */
+#define SI7020CMD_TEMP_HOLD	0xE3
+/* Software Reset */
+#define SI7020CMD_RESET		0xFE
+
+static int si7020_read_raw(struct iio_dev *indio_dev,
+			   struct iio_chan_spec const *chan, int *val,
+			   int *val2, long mask)
+{
+	struct i2c_client *client = iio_priv(indio_dev);
+	int ret;
+
+	switch (mask) {
+	case IIO_CHAN_INFO_RAW:
+		ret = i2c_smbus_read_word_data(client,
+					       chan->type == IIO_TEMP ?
+					       SI7020CMD_TEMP_HOLD :
+					       SI7020CMD_RH_HOLD);
+		if (ret < 0)
+			return ret;
+		*val = ret >> 2;
+		if (chan->type == IIO_HUMIDITYRELATIVE)
+			*val &= GENMASK(11, 0);
+		return IIO_VAL_INT;
+	case IIO_CHAN_INFO_SCALE:
+		if (chan->type == IIO_TEMP)
+			*val = 175720; /* = 175.72 * 1000 */
+		else
+			*val = 125 * 1000;
+		*val2 = 65536 >> 2;
+		return IIO_VAL_FRACTIONAL;
+	case IIO_CHAN_INFO_OFFSET:
+		/*
+		 * Since iio_convert_raw_to_processed_unlocked assumes offset
+		 * is an integer we have to round these values and lose
+		 * accuracy.
+		 * Relative humidity will be 0.0032959% too high and
+		 * temperature will be 0.00277344 degrees too high.
+		 * This is no big deal because it's within the accuracy of the
+		 * sensor.
+		 */
+		if (chan->type == IIO_TEMP)
+			*val = -4368; /* = -46.85 * (65536 >> 2) / 175.72 */
+		else
+			*val = -786; /* = -6 * (65536 >> 2) / 125 */
+		return IIO_VAL_INT;
+	default:
+		break;
+	}
+
+	return -EINVAL;
+}
+
+static const struct iio_chan_spec si7020_channels[] = {
+	{
+		.type = IIO_HUMIDITYRELATIVE,
+		.info_mask_separate = BIT(IIO_CHAN_INFO_RAW) |
+			BIT(IIO_CHAN_INFO_SCALE) | BIT(IIO_CHAN_INFO_OFFSET),
+	},
+	{
+		.type = IIO_TEMP,
+		.info_mask_separate = BIT(IIO_CHAN_INFO_RAW) |
+			BIT(IIO_CHAN_INFO_SCALE) | BIT(IIO_CHAN_INFO_OFFSET),
+	}
+};
+
+static const struct iio_info si7020_info = {
+	.read_raw = si7020_read_raw,
+	.driver_module = THIS_MODULE,
+};
+
+static int si7020_probe(struct i2c_client *client,
+			const struct i2c_device_id *id)
+{
+	struct iio_dev *indio_dev;
+	struct i2c_client **data;
+	int ret;
+
+	if (!i2c_check_functionality(client->adapter,
+				     I2C_FUNC_SMBUS_WRITE_BYTE |
+				     I2C_FUNC_SMBUS_READ_WORD_DATA))
+		return -ENODEV;
+
+	/* Reset device, loads default settings. */
+	ret = i2c_smbus_write_byte(client, SI7020CMD_RESET);
+	if (ret < 0)
+		return ret;
+	/* Wait the maximum power-up time after software reset. */
+	msleep(15);
+
+	indio_dev = devm_iio_device_alloc(&client->dev, sizeof(*client));
+	if (!indio_dev)
+		return -ENOMEM;
+
+	data = iio_priv(indio_dev);
+	*data = client;
+
+	indio_dev->dev.parent = &client->dev;
+	indio_dev->name = dev_name(&client->dev);
+	indio_dev->modes = INDIO_DIRECT_MODE;
+	indio_dev->info = &si7020_info;
+	indio_dev->channels = si7020_channels;
+	indio_dev->num_channels = ARRAY_SIZE(si7020_channels);
+
+	return devm_iio_device_register(&client->dev, indio_dev);
+}
+
+static const struct i2c_device_id si7020_id[] = {
+	{ "si7020", 0 },
+	{ }
+};
+MODULE_DEVICE_TABLE(i2c, si7020_id);
+
+static struct i2c_driver si7020_driver = {
+	.driver.name	= "si7020",
+	.probe		= si7020_probe,
+	.id_table	= si7020_id,
+};
+
+module_i2c_driver(si7020_driver);
+MODULE_DESCRIPTION("Silicon Labs Si7013/20/21 Relative Humidity and Temperature Sensors");
+MODULE_AUTHOR("David Barksdale <dbarksdale@uplogix.com>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/iio/imu/inv_mpu6050/inv_mpu_core.c b/drivers/iio/imu/inv_mpu6050/inv_mpu_core.c
index 0c6517c94a9d..b75519deac1a 100644
--- a/drivers/iio/imu/inv_mpu6050/inv_mpu_core.c
+++ b/drivers/iio/imu/inv_mpu6050/inv_mpu_core.c
@@ -673,8 +673,7 @@ static int inv_mpu_probe(struct i2c_client *client,
 
 	st = iio_priv(indio_dev);
 	st->client = client;
-	pdata = (struct inv_mpu6050_platform_data
-			*)dev_get_platdata(&client->dev);
+	pdata = dev_get_platdata(&client->dev);
 	if (pdata)
 		st->plat_data = *pdata;
 	/* power is turned on inside check chip type*/
diff --git a/drivers/iio/industrialio-buffer.c b/drivers/iio/industrialio-buffer.c
index 0472ee268271..f971f79103ec 100644
--- a/drivers/iio/industrialio-buffer.c
+++ b/drivers/iio/industrialio-buffer.c
@@ -942,13 +942,34 @@ int iio_push_to_buffers(struct iio_dev *indio_dev, const void *data)
 }
 EXPORT_SYMBOL_GPL(iio_push_to_buffers);
 
+static int iio_buffer_add_demux(struct iio_buffer *buffer,
+	struct iio_demux_table **p, unsigned int in_loc, unsigned int out_loc,
+	unsigned int length)
+{
+
+	if (*p && (*p)->from + (*p)->length == in_loc &&
+		(*p)->to + (*p)->length == out_loc) {
+		(*p)->length += length;
+	} else {
+		*p = kmalloc(sizeof(**p), GFP_KERNEL);
+		if (*p == NULL)
+			return -ENOMEM;
+		(*p)->from = in_loc;
+		(*p)->to = out_loc;
+		(*p)->length = length;
+		list_add_tail(&(*p)->l, &buffer->demux_list);
+	}
+
+	return 0;
+}
+
 static int iio_buffer_update_demux(struct iio_dev *indio_dev,
 				   struct iio_buffer *buffer)
 {
 	const struct iio_chan_spec *ch;
 	int ret, in_ind = -1, out_ind, length;
 	unsigned in_loc = 0, out_loc = 0;
-	struct iio_demux_table *p;
+	struct iio_demux_table *p = NULL;
 
 	/* Clear out any old demux */
 	iio_buffer_demux_free(buffer);
@@ -979,14 +1000,7 @@ static int iio_buffer_update_demux(struct iio_dev *indio_dev,
 			else
 				length = ch->scan_type.storagebits / 8;
 			/* Make sure we are aligned */
-			in_loc += length;
-			if (in_loc % length)
-				in_loc += length - in_loc % length;
-		}
-		p = kmalloc(sizeof(*p), GFP_KERNEL);
-		if (p == NULL) {
-			ret = -ENOMEM;
-			goto error_clear_mux_table;
+			in_loc = roundup(in_loc, length) + length;
 		}
 		ch = iio_find_channel_from_si(indio_dev, in_ind);
 		if (ch->scan_type.repeat > 1)
@@ -994,24 +1008,16 @@ static int iio_buffer_update_demux(struct iio_dev *indio_dev,
 				ch->scan_type.repeat;
 		else
 			length = ch->scan_type.storagebits / 8;
-		if (out_loc % length)
-			out_loc += length - out_loc % length;
-		if (in_loc % length)
-			in_loc += length - in_loc % length;
-		p->from = in_loc;
-		p->to = out_loc;
-		p->length = length;
-		list_add_tail(&p->l, &buffer->demux_list);
+		out_loc = roundup(out_loc, length);
+		in_loc = roundup(in_loc, length);
+		ret = iio_buffer_add_demux(buffer, &p, in_loc, out_loc, length);
+		if (ret)
+			goto error_clear_mux_table;
 		out_loc += length;
 		in_loc += length;
 	}
 	/* Relies on scan_timestamp being last */
 	if (buffer->scan_timestamp) {
-		p = kmalloc(sizeof(*p), GFP_KERNEL);
-		if (p == NULL) {
-			ret = -ENOMEM;
-			goto error_clear_mux_table;
-		}
 		ch = iio_find_channel_from_si(indio_dev,
 			indio_dev->scan_index_timestamp);
 		if (ch->scan_type.repeat > 1)
@@ -1019,14 +1025,11 @@ static int iio_buffer_update_demux(struct iio_dev *indio_dev,
 				ch->scan_type.repeat;
 		else
 			length = ch->scan_type.storagebits / 8;
-		if (out_loc % length)
-			out_loc += length - out_loc % length;
-		if (in_loc % length)
-			in_loc += length - in_loc % length;
-		p->from = in_loc;
-		p->to = out_loc;
-		p->length = length;
-		list_add_tail(&p->l, &buffer->demux_list);
+		out_loc = roundup(out_loc, length);
+		in_loc = roundup(in_loc, length);
+		ret = iio_buffer_add_demux(buffer, &p, in_loc, out_loc, length);
+		if (ret)
+			goto error_clear_mux_table;
 		out_loc += length;
 		in_loc += length;
 	}
diff --git a/drivers/iio/inkern.c b/drivers/iio/inkern.c
index f0846108d006..90c8cb727cc7 100644
--- a/drivers/iio/inkern.c
+++ b/drivers/iio/inkern.c
@@ -100,6 +100,28 @@ static int iio_dev_node_match(struct device *dev, void *data)
 	return dev->of_node == data && dev->type == &iio_device_type;
 }
 
+/**
+ * __of_iio_simple_xlate - translate iiospec to the IIO channel index
+ * @indio_dev:	pointer to the iio_dev structure
+ * @iiospec:	IIO specifier as found in the device tree
+ *
+ * This is simple translation function, suitable for the most 1:1 mapped
+ * channels in IIO chips. This function performs only one sanity check:
+ * whether IIO index is less than num_channels (that is specified in the
+ * iio_dev).
+ */
+static int __of_iio_simple_xlate(struct iio_dev *indio_dev,
+				const struct of_phandle_args *iiospec)
+{
+	if (!iiospec->args_count)
+		return 0;
+
+	if (iiospec->args[0] >= indio_dev->num_channels)
+		return -EINVAL;
+
+	return iiospec->args[0];
+}
+
 static int __of_iio_channel_get(struct iio_channel *channel,
 				struct device_node *np, int index)
 {
@@ -122,18 +144,19 @@ static int __of_iio_channel_get(struct iio_channel *channel,
 
 	indio_dev = dev_to_iio_dev(idev);
 	channel->indio_dev = indio_dev;
-	index = iiospec.args_count ? iiospec.args[0] : 0;
-	if (index >= indio_dev->num_channels) {
-		err = -EINVAL;
+	if (indio_dev->info->of_xlate)
+		index = indio_dev->info->of_xlate(indio_dev, &iiospec);
+	else
+		index = __of_iio_simple_xlate(indio_dev, &iiospec);
+	if (index < 0)
 		goto err_put;
-	}
 	channel->channel = &indio_dev->channels[index];
 
 	return 0;
 
 err_put:
 	iio_device_put(indio_dev);
-	return err;
+	return index;
 }
 
 static struct iio_channel *of_iio_channel_get(struct device_node *np, int index)
@@ -426,6 +449,9 @@ static int iio_channel_read(struct iio_channel *chan, int *val, int *val2,
 	if (val2 == NULL)
 		val2 = &unused;
 
+	if(!iio_channel_has_info(chan->channel, info))
+		return -EINVAL;
+
 	if (chan->indio_dev->info->read_raw_multi) {
 		ret = chan->indio_dev->info->read_raw_multi(chan->indio_dev,
 					chan->channel, INDIO_MAX_RAW_ELEMENTS,
diff --git a/drivers/iio/light/Kconfig b/drivers/iio/light/Kconfig
index bf05ca5b0a57..5bea821adcae 100644
--- a/drivers/iio/light/Kconfig
+++ b/drivers/iio/light/Kconfig
@@ -17,6 +17,16 @@ config ADJD_S311
 	 This driver can also be built as a module.  If so, the module
 	 will be called adjd_s311.
 
+config AL3320A
+	tristate "AL3320A ambient light sensor"
+	depends on I2C
+	help
+	 Say Y here if you want to build a driver for the Dyna Image AL3320A
+	 ambient light sensor.
+
+	 To compile this driver as a module, choose M here: the
+	 module will be called al3320a.
+
 config APDS9300
 	tristate "APDS9300 ambient light sensor"
 	depends on I2C
diff --git a/drivers/iio/light/Makefile b/drivers/iio/light/Makefile
index 8b8c09f9c1f8..47877a36cc12 100644
--- a/drivers/iio/light/Makefile
+++ b/drivers/iio/light/Makefile
@@ -4,6 +4,7 @@
 
 # When adding new entries keep the list in alphabetical order
 obj-$(CONFIG_ADJD_S311)		+= adjd_s311.o
+obj-$(CONFIG_AL3320A)		+= al3320a.o
 obj-$(CONFIG_APDS9300)		+= apds9300.o
 obj-$(CONFIG_CM32181)		+= cm32181.o
 obj-$(CONFIG_CM36651)		+= cm36651.o
diff --git a/drivers/iio/light/al3320a.c b/drivers/iio/light/al3320a.c
new file mode 100644
index 000000000000..6aac6513fd41
--- /dev/null
+++ b/drivers/iio/light/al3320a.c
@@ -0,0 +1,232 @@
+/*
+ * AL3320A - Dyna Image Ambient Light Sensor
+ *
+ * Copyright (c) 2014, Intel Corporation.
+ *
+ * This file is subject to the terms and conditions of version 2 of
+ * the GNU General Public License.  See the file COPYING in the main
+ * directory of this archive for more details.
+ *
+ * IIO driver for AL3320A (7-bit I2C slave address 0x1C).
+ *
+ * TODO: interrupt support, thresholds
+ *
+ */
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/i2c.h>
+
+#include <linux/iio/iio.h>
+#include <linux/iio/sysfs.h>
+
+#define AL3320A_DRV_NAME "al3320a"
+
+#define AL3320A_REG_CONFIG		0x00
+#define AL3320A_REG_STATUS		0x01
+#define AL3320A_REG_INT			0x02
+#define AL3320A_REG_WAIT		0x06
+#define AL3320A_REG_CONFIG_RANGE	0x07
+#define AL3320A_REG_PERSIST		0x08
+#define AL3320A_REG_MEAN_TIME		0x09
+#define AL3320A_REG_ADUMMY		0x0A
+#define AL3320A_REG_DATA_LOW		0x22
+
+#define AL3320A_REG_LOW_THRESH_LOW	0x30
+#define AL3320A_REG_LOW_THRESH_HIGH	0x31
+#define AL3320A_REG_HIGH_THRESH_LOW	0x32
+#define AL3320A_REG_HIGH_THRESH_HIGH	0x33
+
+#define AL3320A_CONFIG_DISABLE		0x00
+#define AL3320A_CONFIG_ENABLE		0x01
+
+#define AL3320A_GAIN_SHIFT		1
+#define AL3320A_GAIN_MASK		(BIT(2) | BIT(1))
+
+/* chip params default values */
+#define AL3320A_DEFAULT_MEAN_TIME	4
+#define AL3320A_DEFAULT_WAIT_TIME	0 /* no waiting */
+
+#define AL3320A_SCALE_AVAILABLE "0.512 0.128 0.032 0.01"
+
+enum al3320a_range {
+	AL3320A_RANGE_1, /* 33.28 Klx */
+	AL3320A_RANGE_2, /* 8.32 Klx  */
+	AL3320A_RANGE_3, /* 2.08 Klx  */
+	AL3320A_RANGE_4  /* 0.65 Klx  */
+};
+
+static const int al3320a_scales[][2] = {
+	{0, 512000}, {0, 128000}, {0, 32000}, {0, 10000}
+};
+
+struct al3320a_data {
+	struct i2c_client *client;
+};
+
+static const struct iio_chan_spec al3320a_channels[] = {
+	{
+		.type	= IIO_LIGHT,
+		.info_mask_separate = BIT(IIO_CHAN_INFO_RAW) |
+				      BIT(IIO_CHAN_INFO_SCALE),
+	}
+};
+
+static IIO_CONST_ATTR(in_illuminance_scale_available, AL3320A_SCALE_AVAILABLE);
+
+static struct attribute *al3320a_attributes[] = {
+	&iio_const_attr_in_illuminance_scale_available.dev_attr.attr,
+	NULL,
+};
+
+static const struct attribute_group al3320a_attribute_group = {
+	.attrs = al3320a_attributes,
+};
+
+static int al3320a_init(struct al3320a_data *data)
+{
+	int ret;
+
+	/* power on */
+	ret = i2c_smbus_write_byte_data(data->client, AL3320A_REG_CONFIG,
+					AL3320A_CONFIG_ENABLE);
+	if (ret < 0)
+		return ret;
+
+	ret = i2c_smbus_write_byte_data(data->client, AL3320A_REG_CONFIG_RANGE,
+					AL3320A_RANGE_3 << AL3320A_GAIN_SHIFT);
+	if (ret < 0)
+		return ret;
+
+	ret = i2c_smbus_write_byte_data(data->client, AL3320A_REG_MEAN_TIME,
+					AL3320A_DEFAULT_MEAN_TIME);
+	if (ret < 0)
+		return ret;
+
+	ret = i2c_smbus_write_byte_data(data->client, AL3320A_REG_WAIT,
+					AL3320A_DEFAULT_WAIT_TIME);
+	if (ret < 0)
+		return ret;
+
+	return 0;
+}
+
+static int al3320a_read_raw(struct iio_dev *indio_dev,
+			    struct iio_chan_spec const *chan, int *val,
+			    int *val2, long mask)
+{
+	struct al3320a_data *data = iio_priv(indio_dev);
+	int ret;
+
+	switch (mask) {
+	case IIO_CHAN_INFO_RAW:
+		/*
+		 * ALS ADC value is stored in two adjacent registers:
+		 * - low byte of output is stored at AL3320A_REG_DATA_LOW
+		 * - high byte of output is stored at AL3320A_REG_DATA_LOW + 1
+		 */
+		ret = i2c_smbus_read_word_data(data->client,
+					       AL3320A_REG_DATA_LOW);
+		if (ret < 0)
+			return ret;
+		*val = ret;
+		return IIO_VAL_INT;
+	case IIO_CHAN_INFO_SCALE:
+		ret = i2c_smbus_read_byte_data(data->client,
+					       AL3320A_REG_CONFIG_RANGE);
+		if (ret < 0)
+			return ret;
+
+		ret = (ret & AL3320A_GAIN_MASK) >> AL3320A_GAIN_SHIFT;
+		*val = al3320a_scales[ret][0];
+		*val2 = al3320a_scales[ret][1];
+
+		return IIO_VAL_INT_PLUS_MICRO;
+	}
+	return -EINVAL;
+}
+
+static int al3320a_write_raw(struct iio_dev *indio_dev,
+			     struct iio_chan_spec const *chan, int val,
+			     int val2, long mask)
+{
+	struct al3320a_data *data = iio_priv(indio_dev);
+	int i;
+
+	switch (mask) {
+	case IIO_CHAN_INFO_SCALE:
+		for (i = 0; i < ARRAY_SIZE(al3320a_scales); i++) {
+			if (val == al3320a_scales[i][0] &&
+			    val2 == al3320a_scales[i][1])
+				return i2c_smbus_write_byte_data(data->client,
+					AL3320A_REG_CONFIG_RANGE,
+					i << AL3320A_GAIN_SHIFT);
+		}
+		break;
+	}
+	return -EINVAL;
+}
+
+static const struct iio_info al3320a_info = {
+	.driver_module	= THIS_MODULE,
+	.read_raw	= al3320a_read_raw,
+	.write_raw	= al3320a_write_raw,
+	.attrs		= &al3320a_attribute_group,
+};
+
+static int al3320a_probe(struct i2c_client *client,
+			 const struct i2c_device_id *id)
+{
+	struct al3320a_data *data;
+	struct iio_dev *indio_dev;
+	int ret;
+
+	indio_dev = devm_iio_device_alloc(&client->dev, sizeof(*data));
+	if (!indio_dev)
+		return -ENOMEM;
+
+	data = iio_priv(indio_dev);
+	i2c_set_clientdata(client, indio_dev);
+	data->client = client;
+
+	indio_dev->dev.parent = &client->dev;
+	indio_dev->info = &al3320a_info;
+	indio_dev->name = AL3320A_DRV_NAME;
+	indio_dev->channels = al3320a_channels;
+	indio_dev->num_channels = ARRAY_SIZE(al3320a_channels);
+	indio_dev->modes = INDIO_DIRECT_MODE;
+
+	ret = al3320a_init(data);
+	if (ret < 0) {
+		dev_err(&client->dev, "al3320a chip init failed\n");
+		return ret;
+	}
+	return devm_iio_device_register(&client->dev, indio_dev);
+}
+
+static int al3320a_remove(struct i2c_client *client)
+{
+	return i2c_smbus_write_byte_data(client, AL3320A_REG_CONFIG,
+					 AL3320A_CONFIG_DISABLE);
+}
+
+static const struct i2c_device_id al3320a_id[] = {
+	{"al3320a", 0},
+	{}
+};
+MODULE_DEVICE_TABLE(i2c, al3320a_id);
+
+static struct i2c_driver al3320a_driver = {
+	.driver = {
+		.name = AL3320A_DRV_NAME,
+	},
+	.probe		= al3320a_probe,
+	.remove		= al3320a_remove,
+	.id_table	= al3320a_id,
+};
+
+module_i2c_driver(al3320a_driver);
+
+MODULE_AUTHOR("Daniel Baluta <daniel.baluta@intel.com>");
+MODULE_DESCRIPTION("AL3320A Ambient Light Sensor driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/iio/light/hid-sensor-als.c b/drivers/iio/light/hid-sensor-als.c
index 96e71e103ea7..a5283d75c096 100644
--- a/drivers/iio/light/hid-sensor-als.c
+++ b/drivers/iio/light/hid-sensor-als.c
@@ -381,7 +381,6 @@ static struct platform_driver hid_als_platform_driver = {
 	.id_table = hid_als_ids,
 	.driver = {
 		.name	= KBUILD_MODNAME,
-		.owner	= THIS_MODULE,
 	},
 	.probe		= hid_als_probe,
 	.remove		= hid_als_remove,
diff --git a/drivers/iio/light/hid-sensor-prox.c b/drivers/iio/light/hid-sensor-prox.c
index 412bae86d6ae..f5a514698fd8 100644
--- a/drivers/iio/light/hid-sensor-prox.c
+++ b/drivers/iio/light/hid-sensor-prox.c
@@ -373,7 +373,6 @@ static struct platform_driver hid_prox_platform_driver = {
 	.id_table = hid_prox_ids,
 	.driver = {
 		.name	= KBUILD_MODNAME,
-		.owner	= THIS_MODULE,
 	},
 	.probe		= hid_prox_probe,
 	.remove		= hid_prox_remove,
diff --git a/drivers/iio/light/lm3533-als.c b/drivers/iio/light/lm3533-als.c
index c1aadc6b865a..ae3c71bdd6c6 100644
--- a/drivers/iio/light/lm3533-als.c
+++ b/drivers/iio/light/lm3533-als.c
@@ -915,7 +915,6 @@ static int lm3533_als_remove(struct platform_device *pdev)
 static struct platform_driver lm3533_als_driver = {
 	.driver	= {
 		.name	= "lm3533-als",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= lm3533_als_probe,
 	.remove		= lm3533_als_remove,
diff --git a/drivers/iio/light/tsl4531.c b/drivers/iio/light/tsl4531.c
index a15006efa137..0763b8632573 100644
--- a/drivers/iio/light/tsl4531.c
+++ b/drivers/iio/light/tsl4531.c
@@ -230,9 +230,12 @@ static int tsl4531_resume(struct device *dev)
 	return i2c_smbus_write_byte_data(to_i2c_client(dev), TSL4531_CONTROL,
 		TSL4531_MODE_NORMAL);
 }
-#endif
 
 static SIMPLE_DEV_PM_OPS(tsl4531_pm_ops, tsl4531_suspend, tsl4531_resume);
+#define TSL4531_PM_OPS (&tsl4531_pm_ops)
+#else
+#define TSL4531_PM_OPS NULL
+#endif
 
 static const struct i2c_device_id tsl4531_id[] = {
 	{ "tsl4531", 0 },
@@ -243,7 +246,7 @@ MODULE_DEVICE_TABLE(i2c, tsl4531_id);
 static struct i2c_driver tsl4531_driver = {
 	.driver = {
 		.name   = TSL4531_DRV_NAME,
-		.pm	= &tsl4531_pm_ops,
+		.pm	= TSL4531_PM_OPS,
 		.owner  = THIS_MODULE,
 	},
 	.probe  = tsl4531_probe,
diff --git a/drivers/iio/magnetometer/ak8975.c b/drivers/iio/magnetometer/ak8975.c
index a2357921d761..bf5ef077e791 100644
--- a/drivers/iio/magnetometer/ak8975.c
+++ b/drivers/iio/magnetometer/ak8975.c
@@ -477,8 +477,8 @@ static const struct acpi_device_id ak_acpi_match[] = {
 };
 MODULE_DEVICE_TABLE(acpi, ak_acpi_match);
 
-static char *ak8975_match_acpi_device(struct device *dev,
-				enum asahi_compass_chipset *chipset)
+static const char *ak8975_match_acpi_device(struct device *dev,
+					    enum asahi_compass_chipset *chipset)
 {
 	const struct acpi_device_id *id;
 
@@ -487,7 +487,7 @@ static char *ak8975_match_acpi_device(struct device *dev,
 		return NULL;
 	*chipset = (int)id->driver_data;
 
-	return (char *)dev_name(dev);
+	return dev_name(dev);
 }
 
 static int ak8975_probe(struct i2c_client *client,
@@ -497,7 +497,7 @@ static int ak8975_probe(struct i2c_client *client,
 	struct iio_dev *indio_dev;
 	int eoc_gpio;
 	int err;
-	char *name = NULL;
+	const char *name = NULL;
 
 	/* Grab and set up the supplied GPIO. */
 	if (client->dev.platform_data)
@@ -539,7 +539,7 @@ static int ak8975_probe(struct i2c_client *client,
 	if (id) {
 		data->chipset =
 			(enum asahi_compass_chipset)(id->driver_data);
-		name = (char *) id->name;
+		name = id->name;
 	} else if (ACPI_HANDLE(&client->dev))
 		name = ak8975_match_acpi_device(&client->dev, &data->chipset);
 	else
diff --git a/drivers/iio/magnetometer/hid-sensor-magn-3d.c b/drivers/iio/magnetometer/hid-sensor-magn-3d.c
index 3ec777a8f64e..6294575d2777 100644
--- a/drivers/iio/magnetometer/hid-sensor-magn-3d.c
+++ b/drivers/iio/magnetometer/hid-sensor-magn-3d.c
@@ -246,8 +246,7 @@ static const struct iio_info magn_3d_info = {
 };
 
 /* Function to push data to buffer */
-static void hid_sensor_push_data(struct iio_dev *indio_dev, const void *data,
-	int len)
+static void hid_sensor_push_data(struct iio_dev *indio_dev, const void *data)
 {
 	dev_dbg(&indio_dev->dev, "hid_sensor_push_data\n");
 	iio_push_to_buffers(indio_dev, data);
@@ -263,9 +262,7 @@ static int magn_3d_proc_event(struct hid_sensor_hub_device *hsdev,
 
 	dev_dbg(&indio_dev->dev, "magn_3d_proc_event\n");
 	if (atomic_read(&magn_state->common_attributes.data_ready))
-		hid_sensor_push_data(indio_dev,
-				magn_state->iio_vals,
-				sizeof(magn_state->iio_vals));
+		hid_sensor_push_data(indio_dev, magn_state->iio_vals);
 
 	return 0;
 }
@@ -533,7 +530,6 @@ static struct platform_driver hid_magn_3d_platform_driver = {
 	.id_table = hid_magn_3d_ids,
 	.driver = {
 		.name	= KBUILD_MODNAME,
-		.owner	= THIS_MODULE,
 	},
 	.probe		= hid_magn_3d_probe,
 	.remove		= hid_magn_3d_remove,
diff --git a/drivers/iio/magnetometer/st_magn.h b/drivers/iio/magnetometer/st_magn.h
index 694e33e0fb72..7e81d00ef0c3 100644
--- a/drivers/iio/magnetometer/st_magn.h
+++ b/drivers/iio/magnetometer/st_magn.h
@@ -18,8 +18,7 @@
 #define LSM303DLM_MAGN_DEV_NAME		"lsm303dlm_magn"
 #define LIS3MDL_MAGN_DEV_NAME		"lis3mdl"
 
-int st_magn_common_probe(struct iio_dev *indio_dev,
-					struct st_sensors_platform_data *pdata);
+int st_magn_common_probe(struct iio_dev *indio_dev);
 void st_magn_common_remove(struct iio_dev *indio_dev);
 
 #ifdef CONFIG_IIO_BUFFER
diff --git a/drivers/iio/magnetometer/st_magn_core.c b/drivers/iio/magnetometer/st_magn_core.c
index 68cae86dbd29..8ade473f99fe 100644
--- a/drivers/iio/magnetometer/st_magn_core.c
+++ b/drivers/iio/magnetometer/st_magn_core.c
@@ -149,7 +149,7 @@ static const struct iio_chan_spec st_magn_2_16bit_channels[] = {
 	IIO_CHAN_SOFT_TIMESTAMP(3)
 };
 
-static const struct st_sensors st_magn_sensors[] = {
+static const struct st_sensor_settings st_magn_sensors_settings[] = {
 	{
 		.wai = ST_MAGN_1_WAI_EXP,
 		.sensors_supported = {
@@ -361,8 +361,7 @@ static const struct iio_info magn_info = {
 	.write_raw = &st_magn_write_raw,
 };
 
-int st_magn_common_probe(struct iio_dev *indio_dev,
-					struct st_sensors_platform_data *pdata)
+int st_magn_common_probe(struct iio_dev *indio_dev)
 {
 	struct st_sensor_data *mdata = iio_priv(indio_dev);
 	int irq = mdata->get_irq_data_ready(indio_dev);
@@ -374,20 +373,21 @@ int st_magn_common_probe(struct iio_dev *indio_dev,
 	st_sensors_power_enable(indio_dev);
 
 	err = st_sensors_check_device_support(indio_dev,
-				ARRAY_SIZE(st_magn_sensors), st_magn_sensors);
+					ARRAY_SIZE(st_magn_sensors_settings),
+					st_magn_sensors_settings);
 	if (err < 0)
 		return err;
 
 	mdata->num_data_channels = ST_MAGN_NUMBER_DATA_CHANNELS;
-	mdata->multiread_bit = mdata->sensor->multi_read_bit;
-	indio_dev->channels = mdata->sensor->ch;
+	mdata->multiread_bit = mdata->sensor_settings->multi_read_bit;
+	indio_dev->channels = mdata->sensor_settings->ch;
 	indio_dev->num_channels = ST_SENSORS_NUMBER_ALL_CHANNELS;
 
 	mdata->current_fullscale = (struct st_sensor_fullscale_avl *)
-						&mdata->sensor->fs.fs_avl[0];
-	mdata->odr = mdata->sensor->odr.odr_avl[0].hz;
+					&mdata->sensor_settings->fs.fs_avl[0];
+	mdata->odr = mdata->sensor_settings->odr.odr_avl[0].hz;
 
-	err = st_sensors_init_sensor(indio_dev, pdata);
+	err = st_sensors_init_sensor(indio_dev, NULL);
 	if (err < 0)
 		return err;
 
diff --git a/drivers/iio/magnetometer/st_magn_i2c.c b/drivers/iio/magnetometer/st_magn_i2c.c
index 689250058442..92e5c15452a3 100644
--- a/drivers/iio/magnetometer/st_magn_i2c.c
+++ b/drivers/iio/magnetometer/st_magn_i2c.c
@@ -51,12 +51,11 @@ static int st_magn_i2c_probe(struct i2c_client *client,
 		return -ENOMEM;
 
 	mdata = iio_priv(indio_dev);
-	mdata->dev = &client->dev;
 	st_sensors_of_i2c_probe(client, st_magn_of_match);
 
 	st_sensors_i2c_configure(indio_dev, client, mdata);
 
-	err = st_magn_common_probe(indio_dev, NULL);
+	err = st_magn_common_probe(indio_dev);
 	if (err < 0)
 		return err;
 
diff --git a/drivers/iio/magnetometer/st_magn_spi.c b/drivers/iio/magnetometer/st_magn_spi.c
index a6143ea51dfc..7adacf160146 100644
--- a/drivers/iio/magnetometer/st_magn_spi.c
+++ b/drivers/iio/magnetometer/st_magn_spi.c
@@ -29,11 +29,10 @@ static int st_magn_spi_probe(struct spi_device *spi)
 		return -ENOMEM;
 
 	mdata = iio_priv(indio_dev);
-	mdata->dev = &spi->dev;
 
 	st_sensors_spi_configure(indio_dev, spi, mdata);
 
-	err = st_magn_common_probe(indio_dev, NULL);
+	err = st_magn_common_probe(indio_dev);
 	if (err < 0)
 		return err;
 
diff --git a/drivers/iio/orientation/hid-sensor-incl-3d.c b/drivers/iio/orientation/hid-sensor-incl-3d.c
index 2478f6c2ef25..1ff181bbbcef 100644
--- a/drivers/iio/orientation/hid-sensor-incl-3d.c
+++ b/drivers/iio/orientation/hid-sensor-incl-3d.c
@@ -437,7 +437,6 @@ static struct platform_driver hid_incl_3d_platform_driver = {
 	.id_table = hid_incl_3d_ids,
 	.driver = {
 		.name	= KBUILD_MODNAME,
-		.owner	= THIS_MODULE,
 	},
 	.probe		= hid_incl_3d_probe,
 	.remove		= hid_incl_3d_remove,
diff --git a/drivers/iio/orientation/hid-sensor-rotation.c b/drivers/iio/orientation/hid-sensor-rotation.c
index dccf848e8b0f..4afb6c79ccbc 100644
--- a/drivers/iio/orientation/hid-sensor-rotation.c
+++ b/drivers/iio/orientation/hid-sensor-rotation.c
@@ -334,7 +334,6 @@ static struct platform_driver hid_dev_rot_platform_driver = {
 	.id_table = hid_dev_rot_ids,
 	.driver = {
 		.name	= KBUILD_MODNAME,
-		.owner	= THIS_MODULE,
 	},
 	.probe		= hid_dev_rot_probe,
 	.remove		= hid_dev_rot_remove,
diff --git a/drivers/iio/pressure/Kconfig b/drivers/iio/pressure/Kconfig
index 15afbc919521..a3be53792072 100644
--- a/drivers/iio/pressure/Kconfig
+++ b/drivers/iio/pressure/Kconfig
@@ -5,6 +5,17 @@
 
 menu "Pressure sensors"
 
+config BMP280
+	tristate "Bosch Sensortec BMP280 pressure sensor driver"
+	depends on I2C
+	select REGMAP_I2C
+	help
+	 Say yes here to build support for Bosch Sensortec BMP280
+	 pressure and temperature sensor.
+
+	 To compile this driver as a module, choose M here: the module
+	 will be called bmp280.
+
 config HID_SENSOR_PRESS
 	depends on HID_SENSOR_HUB
 	select IIO_BUFFER
diff --git a/drivers/iio/pressure/Makefile b/drivers/iio/pressure/Makefile
index 90a37e85cf21..88011f2ae00e 100644
--- a/drivers/iio/pressure/Makefile
+++ b/drivers/iio/pressure/Makefile
@@ -3,6 +3,7 @@
 #
 
 # When adding new entries keep the list in alphabetical order
+obj-$(CONFIG_BMP280) += bmp280.o
 obj-$(CONFIG_HID_SENSOR_PRESS)   += hid-sensor-press.o
 obj-$(CONFIG_MPL115) += mpl115.o
 obj-$(CONFIG_MPL3115) += mpl3115.o
diff --git a/drivers/iio/pressure/bmp280.c b/drivers/iio/pressure/bmp280.c
new file mode 100644
index 000000000000..75038dacfff1
--- /dev/null
+++ b/drivers/iio/pressure/bmp280.c
@@ -0,0 +1,455 @@
+/*
+ * Copyright (c) 2014 Intel Corporation
+ *
+ * Driver for Bosch Sensortec BMP280 digital pressure sensor.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+
+#define pr_fmt(fmt) "bmp280: " fmt
+
+#include <linux/module.h>
+#include <linux/i2c.h>
+#include <linux/acpi.h>
+#include <linux/regmap.h>
+#include <linux/iio/iio.h>
+#include <linux/iio/sysfs.h>
+
+#define BMP280_REG_TEMP_XLSB		0xFC
+#define BMP280_REG_TEMP_LSB		0xFB
+#define BMP280_REG_TEMP_MSB		0xFA
+#define BMP280_REG_PRESS_XLSB		0xF9
+#define BMP280_REG_PRESS_LSB		0xF8
+#define BMP280_REG_PRESS_MSB		0xF7
+
+#define BMP280_REG_CONFIG		0xF5
+#define BMP280_REG_CTRL_MEAS		0xF4
+#define BMP280_REG_STATUS		0xF3
+#define BMP280_REG_RESET		0xE0
+#define BMP280_REG_ID			0xD0
+
+#define BMP280_REG_COMP_TEMP_START	0x88
+#define BMP280_COMP_TEMP_REG_COUNT	6
+
+#define BMP280_REG_COMP_PRESS_START	0x8E
+#define BMP280_COMP_PRESS_REG_COUNT	18
+
+#define BMP280_FILTER_MASK		(BIT(4) | BIT(3) | BIT(2))
+#define BMP280_FILTER_OFF		0
+#define BMP280_FILTER_2X		BIT(2)
+#define BMP280_FILTER_4X		BIT(3)
+#define BMP280_FILTER_8X		(BIT(3) | BIT(2))
+#define BMP280_FILTER_16X		BIT(4)
+
+#define BMP280_OSRS_TEMP_MASK		(BIT(7) | BIT(6) | BIT(5))
+#define BMP280_OSRS_TEMP_SKIP		0
+#define BMP280_OSRS_TEMP_1X		BIT(5)
+#define BMP280_OSRS_TEMP_2X		BIT(6)
+#define BMP280_OSRS_TEMP_4X		(BIT(6) | BIT(5))
+#define BMP280_OSRS_TEMP_8X		BIT(7)
+#define BMP280_OSRS_TEMP_16X		(BIT(7) | BIT(5))
+
+#define BMP280_OSRS_PRESS_MASK		(BIT(4) | BIT(3) | BIT(2))
+#define BMP280_OSRS_PRESS_SKIP		0
+#define BMP280_OSRS_PRESS_1X		BIT(2)
+#define BMP280_OSRS_PRESS_2X		BIT(3)
+#define BMP280_OSRS_PRESS_4X		(BIT(3) | BIT(2))
+#define BMP280_OSRS_PRESS_8X		BIT(4)
+#define BMP280_OSRS_PRESS_16X		(BIT(4) | BIT(2))
+
+#define BMP280_MODE_MASK		(BIT(1) | BIT(0))
+#define BMP280_MODE_SLEEP		0
+#define BMP280_MODE_FORCED		BIT(0)
+#define BMP280_MODE_NORMAL		(BIT(1) | BIT(0))
+
+#define BMP280_CHIP_ID			0x58
+#define BMP280_SOFT_RESET_VAL		0xB6
+
+struct bmp280_data {
+	struct i2c_client *client;
+	struct mutex lock;
+	struct regmap *regmap;
+
+	/*
+	 * Carryover value from temperature conversion, used in pressure
+	 * calculation.
+	 */
+	s32 t_fine;
+};
+
+/* Compensation parameters. */
+struct bmp280_comp_temp {
+	u16 dig_t1;
+	s16 dig_t2, dig_t3;
+};
+
+struct bmp280_comp_press {
+	u16 dig_p1;
+	s16 dig_p2, dig_p3, dig_p4, dig_p5, dig_p6, dig_p7, dig_p8, dig_p9;
+};
+
+static const struct iio_chan_spec bmp280_channels[] = {
+	{
+		.type = IIO_PRESSURE,
+		.info_mask_separate = BIT(IIO_CHAN_INFO_PROCESSED),
+	},
+	{
+		.type = IIO_TEMP,
+		.info_mask_separate = BIT(IIO_CHAN_INFO_PROCESSED),
+	},
+};
+
+static bool bmp280_is_writeable_reg(struct device *dev, unsigned int reg)
+{
+	switch (reg) {
+	case BMP280_REG_CONFIG:
+	case BMP280_REG_CTRL_MEAS:
+	case BMP280_REG_RESET:
+		return true;
+	default:
+		return false;
+	};
+}
+
+static bool bmp280_is_volatile_reg(struct device *dev, unsigned int reg)
+{
+	switch (reg) {
+	case BMP280_REG_TEMP_XLSB:
+	case BMP280_REG_TEMP_LSB:
+	case BMP280_REG_TEMP_MSB:
+	case BMP280_REG_PRESS_XLSB:
+	case BMP280_REG_PRESS_LSB:
+	case BMP280_REG_PRESS_MSB:
+	case BMP280_REG_STATUS:
+		return true;
+	default:
+		return false;
+	}
+}
+
+static const struct regmap_config bmp280_regmap_config = {
+	.reg_bits = 8,
+	.val_bits = 8,
+
+	.max_register = BMP280_REG_TEMP_XLSB,
+	.cache_type = REGCACHE_RBTREE,
+
+	.writeable_reg = bmp280_is_writeable_reg,
+	.volatile_reg = bmp280_is_volatile_reg,
+};
+
+static int bmp280_read_compensation_temp(struct bmp280_data *data,
+					 struct bmp280_comp_temp *comp)
+{
+	int ret;
+	__le16 buf[BMP280_COMP_TEMP_REG_COUNT / 2];
+
+	ret = regmap_bulk_read(data->regmap, BMP280_REG_COMP_TEMP_START,
+			       buf, BMP280_COMP_TEMP_REG_COUNT);
+	if (ret < 0) {
+		dev_err(&data->client->dev,
+			"failed to read temperature calibration parameters\n");
+		return ret;
+	}
+
+	comp->dig_t1 = (u16) le16_to_cpu(buf[0]);
+	comp->dig_t2 = (s16) le16_to_cpu(buf[1]);
+	comp->dig_t3 = (s16) le16_to_cpu(buf[2]);
+
+	return 0;
+}
+
+static int bmp280_read_compensation_press(struct bmp280_data *data,
+					  struct bmp280_comp_press *comp)
+{
+	int ret;
+	__le16 buf[BMP280_COMP_PRESS_REG_COUNT / 2];
+
+	ret = regmap_bulk_read(data->regmap, BMP280_REG_COMP_PRESS_START,
+			       buf, BMP280_COMP_PRESS_REG_COUNT);
+	if (ret < 0) {
+		dev_err(&data->client->dev,
+			"failed to read pressure calibration parameters\n");
+		return ret;
+	}
+
+	comp->dig_p1 = (u16) le16_to_cpu(buf[0]);
+	comp->dig_p2 = (s16) le16_to_cpu(buf[1]);
+	comp->dig_p3 = (s16) le16_to_cpu(buf[2]);
+	comp->dig_p4 = (s16) le16_to_cpu(buf[3]);
+	comp->dig_p5 = (s16) le16_to_cpu(buf[4]);
+	comp->dig_p6 = (s16) le16_to_cpu(buf[5]);
+	comp->dig_p7 = (s16) le16_to_cpu(buf[6]);
+	comp->dig_p8 = (s16) le16_to_cpu(buf[7]);
+	comp->dig_p9 = (s16) le16_to_cpu(buf[8]);
+
+	return 0;
+}
+
+/*
+ * Returns temperature in DegC, resolution is 0.01 DegC.  Output value of
+ * "5123" equals 51.23 DegC.  t_fine carries fine temperature as global
+ * value.
+ *
+ * Taken from datasheet, Section 3.11.3, "Compensation formula".
+ */
+static s32 bmp280_compensate_temp(struct bmp280_data *data,
+				  struct bmp280_comp_temp *comp,
+				  s32 adc_temp)
+{
+	s32 var1, var2, t;
+
+	var1 = (((adc_temp >> 3) - ((s32) comp->dig_t1 << 1)) *
+		((s32) comp->dig_t2)) >> 11;
+	var2 = (((((adc_temp >> 4) - ((s32) comp->dig_t1)) *
+		  ((adc_temp >> 4) - ((s32) comp->dig_t1))) >> 12) *
+		((s32) comp->dig_t3)) >> 14;
+
+	data->t_fine = var1 + var2;
+	t = (data->t_fine * 5 + 128) >> 8;
+
+	return t;
+}
+
+/*
+ * Returns pressure in Pa as unsigned 32 bit integer in Q24.8 format (24
+ * integer bits and 8 fractional bits).  Output value of "24674867"
+ * represents 24674867/256 = 96386.2 Pa = 963.862 hPa
+ *
+ * Taken from datasheet, Section 3.11.3, "Compensation formula".
+ */
+static u32 bmp280_compensate_press(struct bmp280_data *data,
+				   struct bmp280_comp_press *comp,
+				   s32 adc_press)
+{
+	s64 var1, var2, p;
+
+	var1 = ((s64) data->t_fine) - 128000;
+	var2 = var1 * var1 * (s64) comp->dig_p6;
+	var2 = var2 + ((var1 * (s64) comp->dig_p5) << 17);
+	var2 = var2 + (((s64) comp->dig_p4) << 35);
+	var1 = ((var1 * var1 * (s64) comp->dig_p3) >> 8) +
+		((var1 * (s64) comp->dig_p2) << 12);
+	var1 = (((((s64) 1) << 47) + var1)) * ((s64) comp->dig_p1) >> 33;
+
+	if (var1 == 0)
+		return 0;
+
+	p = ((((s64) 1048576 - adc_press) << 31) - var2) * 3125;
+	p = div64_s64(p, var1);
+	var1 = (((s64) comp->dig_p9) * (p >> 13) * (p >> 13)) >> 25;
+	var2 = (((s64) comp->dig_p8) * p) >> 19;
+	p = ((p + var1 + var2) >> 8) + (((s64) comp->dig_p7) << 4);
+
+	return (u32) p;
+}
+
+static int bmp280_read_temp(struct bmp280_data *data,
+			    int *val)
+{
+	int ret;
+	__be32 tmp = 0;
+	s32 adc_temp, comp_temp;
+	struct bmp280_comp_temp comp;
+
+	ret = bmp280_read_compensation_temp(data, &comp);
+	if (ret < 0)
+		return ret;
+
+	ret = regmap_bulk_read(data->regmap, BMP280_REG_TEMP_MSB,
+			       (u8 *) &tmp, 3);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "failed to read temperature\n");
+		return ret;
+	}
+
+	adc_temp = be32_to_cpu(tmp) >> 12;
+	comp_temp = bmp280_compensate_temp(data, &comp, adc_temp);
+
+	/*
+	 * val might be NULL if we're called by the read_press routine,
+	 * who only cares about the carry over t_fine value.
+	 */
+	if (val) {
+		*val = comp_temp * 10;
+		return IIO_VAL_INT;
+	}
+
+	return 0;
+}
+
+static int bmp280_read_press(struct bmp280_data *data,
+			     int *val, int *val2)
+{
+	int ret;
+	__be32 tmp = 0;
+	s32 adc_press;
+	u32 comp_press;
+	struct bmp280_comp_press comp;
+
+	ret = bmp280_read_compensation_press(data, &comp);
+	if (ret < 0)
+		return ret;
+
+	/* Read and compensate temperature so we get a reading of t_fine. */
+	ret = bmp280_read_temp(data, NULL);
+	if (ret < 0)
+		return ret;
+
+	ret = regmap_bulk_read(data->regmap, BMP280_REG_PRESS_MSB,
+			       (u8 *) &tmp, 3);
+	if (ret < 0) {
+		dev_err(&data->client->dev, "failed to read pressure\n");
+		return ret;
+	}
+
+	adc_press = be32_to_cpu(tmp) >> 12;
+	comp_press = bmp280_compensate_press(data, &comp, adc_press);
+
+	*val = comp_press;
+	*val2 = 256000;
+
+	return IIO_VAL_FRACTIONAL;
+}
+
+static int bmp280_read_raw(struct iio_dev *indio_dev,
+			   struct iio_chan_spec const *chan,
+			   int *val, int *val2, long mask)
+{
+	int ret;
+	struct bmp280_data *data = iio_priv(indio_dev);
+
+	mutex_lock(&data->lock);
+
+	switch (mask) {
+	case IIO_CHAN_INFO_PROCESSED:
+		switch (chan->type) {
+		case IIO_PRESSURE:
+			ret = bmp280_read_press(data, val, val2);
+			break;
+		case IIO_TEMP:
+			ret = bmp280_read_temp(data, val);
+			break;
+		default:
+			ret = -EINVAL;
+			break;
+		}
+		break;
+	default:
+		ret = -EINVAL;
+		break;
+	}
+
+	mutex_unlock(&data->lock);
+
+	return ret;
+}
+
+static const struct iio_info bmp280_info = {
+	.driver_module = THIS_MODULE,
+	.read_raw = &bmp280_read_raw,
+};
+
+static int bmp280_chip_init(struct bmp280_data *data)
+{
+	int ret;
+
+	ret = regmap_update_bits(data->regmap, BMP280_REG_CTRL_MEAS,
+				 BMP280_OSRS_TEMP_MASK |
+				 BMP280_OSRS_PRESS_MASK |
+				 BMP280_MODE_MASK,
+				 BMP280_OSRS_TEMP_2X |
+				 BMP280_OSRS_PRESS_16X |
+				 BMP280_MODE_NORMAL);
+	if (ret < 0) {
+		dev_err(&data->client->dev,
+			"failed to write config register\n");
+		return ret;
+	}
+
+	ret = regmap_update_bits(data->regmap, BMP280_REG_CONFIG,
+				 BMP280_FILTER_MASK,
+				 BMP280_FILTER_4X);
+	if (ret < 0) {
+		dev_err(&data->client->dev,
+			"failed to write config register\n");
+		return ret;
+	}
+
+	return ret;
+}
+
+static int bmp280_probe(struct i2c_client *client,
+			const struct i2c_device_id *id)
+{
+	int ret;
+	struct iio_dev *indio_dev;
+	struct bmp280_data *data;
+	unsigned int chip_id;
+
+	indio_dev = devm_iio_device_alloc(&client->dev, sizeof(*data));
+	if (!indio_dev)
+		return -ENOMEM;
+
+	i2c_set_clientdata(client, indio_dev);
+	data = iio_priv(indio_dev);
+	mutex_init(&data->lock);
+	data->client = client;
+
+	indio_dev->dev.parent = &client->dev;
+	indio_dev->name = id->name;
+	indio_dev->channels = bmp280_channels;
+	indio_dev->num_channels = ARRAY_SIZE(bmp280_channels);
+	indio_dev->info = &bmp280_info;
+	indio_dev->modes = INDIO_DIRECT_MODE;
+
+	data->regmap = devm_regmap_init_i2c(client, &bmp280_regmap_config);
+	if (IS_ERR(data->regmap)) {
+		dev_err(&client->dev, "failed to allocate register map\n");
+		return PTR_ERR(data->regmap);
+	}
+
+	ret = regmap_read(data->regmap, BMP280_REG_ID, &chip_id);
+	if (ret < 0)
+		return ret;
+	if (chip_id != BMP280_CHIP_ID) {
+		dev_err(&client->dev, "bad chip id.  expected %x got %x\n",
+			BMP280_CHIP_ID, chip_id);
+		return -EINVAL;
+	}
+
+	ret = bmp280_chip_init(data);
+	if (ret < 0)
+		return ret;
+
+	return devm_iio_device_register(&client->dev, indio_dev);
+}
+
+static const struct acpi_device_id bmp280_acpi_match[] = {
+	{"BMP0280", 0},
+	{ },
+};
+MODULE_DEVICE_TABLE(acpi, bmp280_acpi_match);
+
+static const struct i2c_device_id bmp280_id[] = {
+	{"bmp280", 0},
+	{ },
+};
+MODULE_DEVICE_TABLE(i2c, bmp280_id);
+
+static struct i2c_driver bmp280_driver = {
+	.driver = {
+		.name	= "bmp280",
+		.acpi_match_table = ACPI_PTR(bmp280_acpi_match),
+	},
+	.probe		= bmp280_probe,
+	.id_table	= bmp280_id,
+};
+module_i2c_driver(bmp280_driver);
+
+MODULE_AUTHOR("Vlad Dogaru <vlad.dogaru@intel.com>");
+MODULE_DESCRIPTION("Driver for Bosch Sensortec BMP280 pressure and temperature sensor");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/iio/pressure/hid-sensor-press.c b/drivers/iio/pressure/hid-sensor-press.c
index 2c0d2a4fed8c..764928682df2 100644
--- a/drivers/iio/pressure/hid-sensor-press.c
+++ b/drivers/iio/pressure/hid-sensor-press.c
@@ -382,7 +382,6 @@ static struct platform_driver hid_press_platform_driver = {
 	.id_table = hid_press_ids,
 	.driver = {
 		.name	= KBUILD_MODNAME,
-		.owner	= THIS_MODULE,
 	},
 	.probe		= hid_press_probe,
 	.remove		= hid_press_remove,
diff --git a/drivers/iio/pressure/st_pressure.h b/drivers/iio/pressure/st_pressure.h
index 242943c0c4e4..f5f41490060b 100644
--- a/drivers/iio/pressure/st_pressure.h
+++ b/drivers/iio/pressure/st_pressure.h
@@ -26,8 +26,7 @@ static const struct st_sensors_platform_data default_press_pdata = {
 	.drdy_int_pin = 1,
 };
 
-int st_press_common_probe(struct iio_dev *indio_dev,
-					struct st_sensors_platform_data *pdata);
+int st_press_common_probe(struct iio_dev *indio_dev);
 void st_press_common_remove(struct iio_dev *indio_dev);
 
 #ifdef CONFIG_IIO_BUFFER
diff --git a/drivers/iio/pressure/st_pressure_buffer.c b/drivers/iio/pressure/st_pressure_buffer.c
index b37b1c9ac932..2ff53f222352 100644
--- a/drivers/iio/pressure/st_pressure_buffer.c
+++ b/drivers/iio/pressure/st_pressure_buffer.c
@@ -38,10 +38,10 @@ static int st_press_buffer_preenable(struct iio_dev *indio_dev)
 static int st_press_buffer_postenable(struct iio_dev *indio_dev)
 {
 	int err;
-	struct st_sensor_data *pdata = iio_priv(indio_dev);
+	struct st_sensor_data *press_data = iio_priv(indio_dev);
 
-	pdata->buffer_data = kmalloc(indio_dev->scan_bytes, GFP_KERNEL);
-	if (pdata->buffer_data == NULL) {
+	press_data->buffer_data = kmalloc(indio_dev->scan_bytes, GFP_KERNEL);
+	if (press_data->buffer_data == NULL) {
 		err = -ENOMEM;
 		goto allocate_memory_error;
 	}
@@ -53,7 +53,7 @@ static int st_press_buffer_postenable(struct iio_dev *indio_dev)
 	return err;
 
 st_press_buffer_postenable_error:
-	kfree(pdata->buffer_data);
+	kfree(press_data->buffer_data);
 allocate_memory_error:
 	return err;
 }
@@ -61,7 +61,7 @@ allocate_memory_error:
 static int st_press_buffer_predisable(struct iio_dev *indio_dev)
 {
 	int err;
-	struct st_sensor_data *pdata = iio_priv(indio_dev);
+	struct st_sensor_data *press_data = iio_priv(indio_dev);
 
 	err = iio_triggered_buffer_predisable(indio_dev);
 	if (err < 0)
@@ -70,7 +70,7 @@ static int st_press_buffer_predisable(struct iio_dev *indio_dev)
 	err = st_sensors_set_enable(indio_dev, false);
 
 st_press_buffer_predisable_error:
-	kfree(pdata->buffer_data);
+	kfree(press_data->buffer_data);
 	return err;
 }
 
diff --git a/drivers/iio/pressure/st_pressure_core.c b/drivers/iio/pressure/st_pressure_core.c
index 473d914ef470..97baf40d424b 100644
--- a/drivers/iio/pressure/st_pressure_core.c
+++ b/drivers/iio/pressure/st_pressure_core.c
@@ -175,7 +175,7 @@ static const struct iio_chan_spec st_press_lps001wp_channels[] = {
 	IIO_CHAN_SOFT_TIMESTAMP(1)
 };
 
-static const struct st_sensors st_press_sensors[] = {
+static const struct st_sensor_settings st_press_sensors_settings[] = {
 	{
 		.wai = ST_PRESS_LPS331AP_WAI_EXP,
 		.sensors_supported = {
@@ -333,7 +333,7 @@ static int st_press_read_raw(struct iio_dev *indio_dev,
 							int *val2, long mask)
 {
 	int err;
-	struct st_sensor_data *pdata = iio_priv(indio_dev);
+	struct st_sensor_data *press_data = iio_priv(indio_dev);
 
 	switch (mask) {
 	case IIO_CHAN_INFO_RAW:
@@ -347,10 +347,10 @@ static int st_press_read_raw(struct iio_dev *indio_dev,
 
 		switch (ch->type) {
 		case IIO_PRESSURE:
-			*val2 = pdata->current_fullscale->gain;
+			*val2 = press_data->current_fullscale->gain;
 			break;
 		case IIO_TEMP:
-			*val2 = pdata->current_fullscale->gain2;
+			*val2 = press_data->current_fullscale->gain2;
 			break;
 		default:
 			err = -EINVAL;
@@ -371,7 +371,7 @@ static int st_press_read_raw(struct iio_dev *indio_dev,
 
 		return IIO_VAL_FRACTIONAL;
 	case IIO_CHAN_INFO_SAMP_FREQ:
-		*val = pdata->odr;
+		*val = press_data->odr;
 		return IIO_VAL_INT;
 	default:
 		return -EINVAL;
@@ -409,11 +409,10 @@ static const struct iio_trigger_ops st_press_trigger_ops = {
 #define ST_PRESS_TRIGGER_OPS NULL
 #endif
 
-int st_press_common_probe(struct iio_dev *indio_dev,
-				struct st_sensors_platform_data *plat_data)
+int st_press_common_probe(struct iio_dev *indio_dev)
 {
-	struct st_sensor_data *pdata = iio_priv(indio_dev);
-	int irq = pdata->get_irq_data_ready(indio_dev);
+	struct st_sensor_data *press_data = iio_priv(indio_dev);
+	int irq = press_data->get_irq_data_ready(indio_dev);
 	int err;
 
 	indio_dev->modes = INDIO_DIRECT_MODE;
@@ -422,28 +421,30 @@ int st_press_common_probe(struct iio_dev *indio_dev,
 	st_sensors_power_enable(indio_dev);
 
 	err = st_sensors_check_device_support(indio_dev,
-					      ARRAY_SIZE(st_press_sensors),
-					      st_press_sensors);
+					ARRAY_SIZE(st_press_sensors_settings),
+					st_press_sensors_settings);
 	if (err < 0)
 		return err;
 
-	pdata->num_data_channels = ST_PRESS_NUMBER_DATA_CHANNELS;
-	pdata->multiread_bit     = pdata->sensor->multi_read_bit;
-	indio_dev->channels      = pdata->sensor->ch;
-	indio_dev->num_channels  = pdata->sensor->num_ch;
+	press_data->num_data_channels = ST_PRESS_NUMBER_DATA_CHANNELS;
+	press_data->multiread_bit = press_data->sensor_settings->multi_read_bit;
+	indio_dev->channels = press_data->sensor_settings->ch;
+	indio_dev->num_channels = press_data->sensor_settings->num_ch;
 
-	if (pdata->sensor->fs.addr != 0)
-		pdata->current_fullscale = (struct st_sensor_fullscale_avl *)
-			&pdata->sensor->fs.fs_avl[0];
+	if (press_data->sensor_settings->fs.addr != 0)
+		press_data->current_fullscale =
+			(struct st_sensor_fullscale_avl *)
+				&press_data->sensor_settings->fs.fs_avl[0];
 
-	pdata->odr = pdata->sensor->odr.odr_avl[0].hz;
+	press_data->odr = press_data->sensor_settings->odr.odr_avl[0].hz;
 
 	/* Some devices don't support a data ready pin. */
-	if (!plat_data && pdata->sensor->drdy_irq.addr)
-		plat_data =
+	if (!press_data->dev->platform_data &&
+				press_data->sensor_settings->drdy_irq.addr)
+		press_data->dev->platform_data =
 			(struct st_sensors_platform_data *)&default_press_pdata;
 
-	err = st_sensors_init_sensor(indio_dev, plat_data);
+	err = st_sensors_init_sensor(indio_dev, press_data->dev->platform_data);
 	if (err < 0)
 		return err;
 
@@ -479,12 +480,12 @@ EXPORT_SYMBOL(st_press_common_probe);
 
 void st_press_common_remove(struct iio_dev *indio_dev)
 {
-	struct st_sensor_data *pdata = iio_priv(indio_dev);
+	struct st_sensor_data *press_data = iio_priv(indio_dev);
 
 	st_sensors_power_disable(indio_dev);
 
 	iio_device_unregister(indio_dev);
-	if (pdata->get_irq_data_ready(indio_dev) > 0)
+	if (press_data->get_irq_data_ready(indio_dev) > 0)
 		st_sensors_deallocate_trigger(indio_dev);
 
 	st_press_deallocate_ring(indio_dev);
diff --git a/drivers/iio/pressure/st_pressure_i2c.c b/drivers/iio/pressure/st_pressure_i2c.c
index acaf165260bb..137788bba4a3 100644
--- a/drivers/iio/pressure/st_pressure_i2c.c
+++ b/drivers/iio/pressure/st_pressure_i2c.c
@@ -43,20 +43,19 @@ static int st_press_i2c_probe(struct i2c_client *client,
 						const struct i2c_device_id *id)
 {
 	struct iio_dev *indio_dev;
-	struct st_sensor_data *pdata;
+	struct st_sensor_data *press_data;
 	int err;
 
-	indio_dev = devm_iio_device_alloc(&client->dev, sizeof(*pdata));
+	indio_dev = devm_iio_device_alloc(&client->dev, sizeof(*press_data));
 	if (!indio_dev)
 		return -ENOMEM;
 
-	pdata = iio_priv(indio_dev);
-	pdata->dev = &client->dev;
+	press_data = iio_priv(indio_dev);
 	st_sensors_of_i2c_probe(client, st_press_of_match);
 
-	st_sensors_i2c_configure(indio_dev, client, pdata);
+	st_sensors_i2c_configure(indio_dev, client, press_data);
 
-	err = st_press_common_probe(indio_dev, client->dev.platform_data);
+	err = st_press_common_probe(indio_dev);
 	if (err < 0)
 		return err;
 
diff --git a/drivers/iio/pressure/st_pressure_spi.c b/drivers/iio/pressure/st_pressure_spi.c
index f45d430ec529..1ffa6d4d349c 100644
--- a/drivers/iio/pressure/st_pressure_spi.c
+++ b/drivers/iio/pressure/st_pressure_spi.c
@@ -21,19 +21,18 @@
 static int st_press_spi_probe(struct spi_device *spi)
 {
 	struct iio_dev *indio_dev;
-	struct st_sensor_data *pdata;
+	struct st_sensor_data *press_data;
 	int err;
 
-	indio_dev = devm_iio_device_alloc(&spi->dev, sizeof(*pdata));
+	indio_dev = devm_iio_device_alloc(&spi->dev, sizeof(*press_data));
 	if (indio_dev == NULL)
 		return -ENOMEM;
 
-	pdata = iio_priv(indio_dev);
-	pdata->dev = &spi->dev;
+	press_data = iio_priv(indio_dev);
 
-	st_sensors_spi_configure(indio_dev, spi, pdata);
+	st_sensors_spi_configure(indio_dev, spi, press_data);
 
-	err = st_press_common_probe(indio_dev, spi->dev.platform_data);
+	err = st_press_common_probe(indio_dev);
 	if (err < 0)
 		return err;
 
diff --git a/drivers/iio/proximity/as3935.c b/drivers/iio/proximity/as3935.c
index 5e780ef206f3..466aa4314667 100644
--- a/drivers/iio/proximity/as3935.c
+++ b/drivers/iio/proximity/as3935.c
@@ -95,7 +95,7 @@ static int as3935_read(struct as3935_state *st, unsigned int reg, int *val)
 	*val = ret;
 
 	return 0;
-};
+}
 
 static int as3935_write(struct as3935_state *st,
 				unsigned int reg,
@@ -107,7 +107,7 @@ static int as3935_write(struct as3935_state *st,
 	buf[1] = val;
 
 	return spi_write(st->spi, buf, 2);
-};
+}
 
 static ssize_t as3935_sensor_sensitivity_show(struct device *dev,
 					struct device_attribute *attr,
@@ -122,7 +122,7 @@ static ssize_t as3935_sensor_sensitivity_show(struct device *dev,
 	val = (val & AS3935_AFE_MASK) >> 1;
 
 	return sprintf(buf, "%d\n", val);
-};
+}
 
 static ssize_t as3935_sensor_sensitivity_store(struct device *dev,
 					struct device_attribute *attr,
@@ -142,7 +142,7 @@ static ssize_t as3935_sensor_sensitivity_store(struct device *dev,
 	as3935_write(st, AS3935_AFE_GAIN, val << 1);
 
 	return len;
-};
+}
 
 static IIO_DEVICE_ATTR(sensor_sensitivity, S_IRUGO | S_IWUSR,
 	as3935_sensor_sensitivity_show, as3935_sensor_sensitivity_store, 0);
@@ -214,7 +214,7 @@ err_read:
 	iio_trigger_notify_done(indio_dev->trig);
 
 	return IRQ_HANDLED;
-};
+}
 
 static const struct iio_trigger_ops iio_interrupt_trigger_ops = {
 	.owner = THIS_MODULE,
@@ -238,7 +238,7 @@ static void as3935_event_work(struct work_struct *work)
 		dev_warn(&st->spi->dev, "noise level is too high");
 		break;
 	}
-};
+}
 
 static irqreturn_t as3935_interrupt_handler(int irq, void *private)
 {
@@ -330,7 +330,7 @@ static int as3935_probe(struct spi_device *spi)
 		return -EINVAL;
 	}
 
-	indio_dev = devm_iio_device_alloc(&spi->dev, sizeof(st));
+	indio_dev = devm_iio_device_alloc(&spi->dev, sizeof(*st));
 	if (!indio_dev)
 		return -ENOMEM;
 
@@ -417,7 +417,7 @@ unregister_trigger:
 	iio_trigger_unregister(st->trig);
 
 	return ret;
-};
+}
 
 static int as3935_remove(struct spi_device *spi)
 {
@@ -429,7 +429,7 @@ static int as3935_remove(struct spi_device *spi)
 	iio_trigger_unregister(st->trig);
 
 	return 0;
-};
+}
 
 static const struct spi_device_id as3935_id[] = {
 	{"as3935", 0},
diff --git a/drivers/iio/trigger/iio-trig-interrupt.c b/drivers/iio/trigger/iio-trig-interrupt.c
index 7a149a7822bc..572bc6f02ca8 100644
--- a/drivers/iio/trigger/iio-trig-interrupt.c
+++ b/drivers/iio/trigger/iio-trig-interrupt.c
@@ -109,7 +109,6 @@ static struct platform_driver iio_interrupt_trigger_driver = {
 	.remove = iio_interrupt_trigger_remove,
 	.driver = {
 		.name = "iio_interrupt_trigger",
-		.owner = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/infiniband/Kconfig b/drivers/infiniband/Kconfig
index 77089399359b..b899531498eb 100644
--- a/drivers/infiniband/Kconfig
+++ b/drivers/infiniband/Kconfig
@@ -38,6 +38,17 @@ config INFINIBAND_USER_MEM
 	depends on INFINIBAND_USER_ACCESS != n
 	default y
 
+config INFINIBAND_ON_DEMAND_PAGING
+	bool "InfiniBand on-demand paging support"
+	depends on INFINIBAND_USER_MEM
+	select MMU_NOTIFIER
+	default y
+	---help---
+	  On demand paging support for the InfiniBand subsystem.
+	  Together with driver support this allows registration of
+	  memory regions without pinning their pages, fetching the
+	  pages on demand instead.
+
 config INFINIBAND_ADDR_TRANS
 	bool
 	depends on INFINIBAND
diff --git a/drivers/infiniband/core/Makefile b/drivers/infiniband/core/Makefile
index ffd0af6734af..acf736764445 100644
--- a/drivers/infiniband/core/Makefile
+++ b/drivers/infiniband/core/Makefile
@@ -11,6 +11,7 @@ obj-$(CONFIG_INFINIBAND_USER_ACCESS) +=	ib_uverbs.o ib_ucm.o \
 ib_core-y :=			packer.o ud_header.o verbs.o sysfs.o \
 				device.o fmr_pool.o cache.o netlink.o
 ib_core-$(CONFIG_INFINIBAND_USER_MEM) += umem.o
+ib_core-$(CONFIG_INFINIBAND_ON_DEMAND_PAGING) += umem_odp.o umem_rbtree.o
 
 ib_mad-y :=			mad.o smi.o agent.o mad_rmpp.o
 
diff --git a/drivers/infiniband/core/addr.c b/drivers/infiniband/core/addr.c
index 8172d37f9add..f80da50d84a5 100644
--- a/drivers/infiniband/core/addr.c
+++ b/drivers/infiniband/core/addr.c
@@ -176,8 +176,8 @@ static void set_timeout(unsigned long time)
 	unsigned long delay;
 
 	delay = time - jiffies;
-	if ((long)delay <= 0)
-		delay = 1;
+	if ((long)delay < 0)
+		delay = 0;
 
 	mod_delayed_work(addr_wq, &work, delay);
 }
diff --git a/drivers/infiniband/core/multicast.c b/drivers/infiniband/core/multicast.c
index d2360a8ef0b2..fa17b552ff78 100644
--- a/drivers/infiniband/core/multicast.c
+++ b/drivers/infiniband/core/multicast.c
@@ -525,17 +525,22 @@ static void join_handler(int status, struct ib_sa_mcmember_rec *rec,
 	if (status)
 		process_join_error(group, status);
 	else {
+		int mgids_changed, is_mgid0;
 		ib_find_pkey(group->port->dev->device, group->port->port_num,
 			     be16_to_cpu(rec->pkey), &pkey_index);
 
 		spin_lock_irq(&group->port->lock);
-		group->rec = *rec;
 		if (group->state == MCAST_BUSY &&
 		    group->pkey_index == MCAST_INVALID_PKEY_INDEX)
 			group->pkey_index = pkey_index;
-		if (!memcmp(&mgid0, &group->rec.mgid, sizeof mgid0)) {
+		mgids_changed = memcmp(&rec->mgid, &group->rec.mgid,
+				       sizeof(group->rec.mgid));
+		group->rec = *rec;
+		if (mgids_changed) {
 			rb_erase(&group->node, &group->port->table);
-			mcast_insert(group->port, group, 1);
+			is_mgid0 = !memcmp(&mgid0, &group->rec.mgid,
+					   sizeof(mgid0));
+			mcast_insert(group->port, group, is_mgid0);
 		}
 		spin_unlock_irq(&group->port->lock);
 	}
diff --git a/drivers/infiniband/core/umem.c b/drivers/infiniband/core/umem.c
index df0c4f605a21..aec7a6aa2951 100644
--- a/drivers/infiniband/core/umem.c
+++ b/drivers/infiniband/core/umem.c
@@ -39,6 +39,7 @@
 #include <linux/hugetlb.h>
 #include <linux/dma-attrs.h>
 #include <linux/slab.h>
+#include <rdma/ib_umem_odp.h>
 
 #include "uverbs.h"
 
@@ -69,6 +70,10 @@ static void __ib_umem_release(struct ib_device *dev, struct ib_umem *umem, int d
 
 /**
  * ib_umem_get - Pin and DMA map userspace memory.
+ *
+ * If access flags indicate ODP memory, avoid pinning. Instead, stores
+ * the mm for future page fault handling in conjunction with MMU notifiers.
+ *
  * @context: userspace context to pin memory for
  * @addr: userspace virtual address to start at
  * @size: length of region to pin
@@ -103,17 +108,30 @@ struct ib_umem *ib_umem_get(struct ib_ucontext *context, unsigned long addr,
 
 	umem->context   = context;
 	umem->length    = size;
-	umem->offset    = addr & ~PAGE_MASK;
+	umem->address   = addr;
 	umem->page_size = PAGE_SIZE;
 	umem->pid       = get_task_pid(current, PIDTYPE_PID);
 	/*
-	 * We ask for writable memory if any access flags other than
-	 * "remote read" are set.  "Local write" and "remote write"
+	 * We ask for writable memory if any of the following
+	 * access flags are set.  "Local write" and "remote write"
 	 * obviously require write access.  "Remote atomic" can do
 	 * things like fetch and add, which will modify memory, and
 	 * "MW bind" can change permissions by binding a window.
 	 */
-	umem->writable  = !!(access & ~IB_ACCESS_REMOTE_READ);
+	umem->writable  = !!(access &
+		(IB_ACCESS_LOCAL_WRITE   | IB_ACCESS_REMOTE_WRITE |
+		 IB_ACCESS_REMOTE_ATOMIC | IB_ACCESS_MW_BIND));
+
+	if (access & IB_ACCESS_ON_DEMAND) {
+		ret = ib_umem_odp_get(context, umem);
+		if (ret) {
+			kfree(umem);
+			return ERR_PTR(ret);
+		}
+		return umem;
+	}
+
+	umem->odp_data = NULL;
 
 	/* We assume the memory is from hugetlb until proved otherwise */
 	umem->hugetlb   = 1;
@@ -132,7 +150,7 @@ struct ib_umem *ib_umem_get(struct ib_ucontext *context, unsigned long addr,
 	if (!vma_list)
 		umem->hugetlb = 0;
 
-	npages = PAGE_ALIGN(size + umem->offset) >> PAGE_SHIFT;
+	npages = ib_umem_num_pages(umem);
 
 	down_write(&current->mm->mmap_sem);
 
@@ -235,6 +253,11 @@ void ib_umem_release(struct ib_umem *umem)
 	struct task_struct *task;
 	unsigned long diff;
 
+	if (umem->odp_data) {
+		ib_umem_odp_release(umem);
+		return;
+	}
+
 	__ib_umem_release(umem->context->device, umem, 1);
 
 	task = get_pid_task(umem->pid, PIDTYPE_PID);
@@ -246,7 +269,7 @@ void ib_umem_release(struct ib_umem *umem)
 	if (!mm)
 		goto out;
 
-	diff = PAGE_ALIGN(umem->length + umem->offset) >> PAGE_SHIFT;
+	diff = ib_umem_num_pages(umem);
 
 	/*
 	 * We may be called with the mm's mmap_sem already held.  This
@@ -283,6 +306,9 @@ int ib_umem_page_count(struct ib_umem *umem)
 	int n;
 	struct scatterlist *sg;
 
+	if (umem->odp_data)
+		return ib_umem_num_pages(umem);
+
 	shift = ilog2(umem->page_size);
 
 	n = 0;
@@ -292,3 +318,37 @@ int ib_umem_page_count(struct ib_umem *umem)
 	return n;
 }
 EXPORT_SYMBOL(ib_umem_page_count);
+
+/*
+ * Copy from the given ib_umem's pages to the given buffer.
+ *
+ * umem - the umem to copy from
+ * offset - offset to start copying from
+ * dst - destination buffer
+ * length - buffer length
+ *
+ * Returns 0 on success, or an error code.
+ */
+int ib_umem_copy_from(void *dst, struct ib_umem *umem, size_t offset,
+		      size_t length)
+{
+	size_t end = offset + length;
+	int ret;
+
+	if (offset > umem->length || length > umem->length - offset) {
+		pr_err("ib_umem_copy_from not in range. offset: %zd umem length: %zd end: %zd\n",
+		       offset, umem->length, end);
+		return -EINVAL;
+	}
+
+	ret = sg_pcopy_to_buffer(umem->sg_head.sgl, umem->nmap, dst, length,
+				 offset + ib_umem_offset(umem));
+
+	if (ret < 0)
+		return ret;
+	else if (ret != length)
+		return -EINVAL;
+	else
+		return 0;
+}
+EXPORT_SYMBOL(ib_umem_copy_from);
diff --git a/drivers/infiniband/core/umem_odp.c b/drivers/infiniband/core/umem_odp.c
new file mode 100644
index 000000000000..6095872549e7
--- /dev/null
+++ b/drivers/infiniband/core/umem_odp.c
@@ -0,0 +1,668 @@
+/*
+ * Copyright (c) 2014 Mellanox Technologies. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include <linux/types.h>
+#include <linux/sched.h>
+#include <linux/pid.h>
+#include <linux/slab.h>
+#include <linux/export.h>
+#include <linux/vmalloc.h>
+
+#include <rdma/ib_verbs.h>
+#include <rdma/ib_umem.h>
+#include <rdma/ib_umem_odp.h>
+
+static void ib_umem_notifier_start_account(struct ib_umem *item)
+{
+	mutex_lock(&item->odp_data->umem_mutex);
+
+	/* Only update private counters for this umem if it has them.
+	 * Otherwise skip it. All page faults will be delayed for this umem. */
+	if (item->odp_data->mn_counters_active) {
+		int notifiers_count = item->odp_data->notifiers_count++;
+
+		if (notifiers_count == 0)
+			/* Initialize the completion object for waiting on
+			 * notifiers. Since notifier_count is zero, no one
+			 * should be waiting right now. */
+			reinit_completion(&item->odp_data->notifier_completion);
+	}
+	mutex_unlock(&item->odp_data->umem_mutex);
+}
+
+static void ib_umem_notifier_end_account(struct ib_umem *item)
+{
+	mutex_lock(&item->odp_data->umem_mutex);
+
+	/* Only update private counters for this umem if it has them.
+	 * Otherwise skip it. All page faults will be delayed for this umem. */
+	if (item->odp_data->mn_counters_active) {
+		/*
+		 * This sequence increase will notify the QP page fault that
+		 * the page that is going to be mapped in the spte could have
+		 * been freed.
+		 */
+		++item->odp_data->notifiers_seq;
+		if (--item->odp_data->notifiers_count == 0)
+			complete_all(&item->odp_data->notifier_completion);
+	}
+	mutex_unlock(&item->odp_data->umem_mutex);
+}
+
+/* Account for a new mmu notifier in an ib_ucontext. */
+static void ib_ucontext_notifier_start_account(struct ib_ucontext *context)
+{
+	atomic_inc(&context->notifier_count);
+}
+
+/* Account for a terminating mmu notifier in an ib_ucontext.
+ *
+ * Must be called with the ib_ucontext->umem_rwsem semaphore unlocked, since
+ * the function takes the semaphore itself. */
+static void ib_ucontext_notifier_end_account(struct ib_ucontext *context)
+{
+	int zero_notifiers = atomic_dec_and_test(&context->notifier_count);
+
+	if (zero_notifiers &&
+	    !list_empty(&context->no_private_counters)) {
+		/* No currently running mmu notifiers. Now is the chance to
+		 * add private accounting to all previously added umems. */
+		struct ib_umem_odp *odp_data, *next;
+
+		/* Prevent concurrent mmu notifiers from working on the
+		 * no_private_counters list. */
+		down_write(&context->umem_rwsem);
+
+		/* Read the notifier_count again, with the umem_rwsem
+		 * semaphore taken for write. */
+		if (!atomic_read(&context->notifier_count)) {
+			list_for_each_entry_safe(odp_data, next,
+						 &context->no_private_counters,
+						 no_private_counters) {
+				mutex_lock(&odp_data->umem_mutex);
+				odp_data->mn_counters_active = true;
+				list_del(&odp_data->no_private_counters);
+				complete_all(&odp_data->notifier_completion);
+				mutex_unlock(&odp_data->umem_mutex);
+			}
+		}
+
+		up_write(&context->umem_rwsem);
+	}
+}
+
+static int ib_umem_notifier_release_trampoline(struct ib_umem *item, u64 start,
+					       u64 end, void *cookie) {
+	/*
+	 * Increase the number of notifiers running, to
+	 * prevent any further fault handling on this MR.
+	 */
+	ib_umem_notifier_start_account(item);
+	item->odp_data->dying = 1;
+	/* Make sure that the fact the umem is dying is out before we release
+	 * all pending page faults. */
+	smp_wmb();
+	complete_all(&item->odp_data->notifier_completion);
+	item->context->invalidate_range(item, ib_umem_start(item),
+					ib_umem_end(item));
+	return 0;
+}
+
+static void ib_umem_notifier_release(struct mmu_notifier *mn,
+				     struct mm_struct *mm)
+{
+	struct ib_ucontext *context = container_of(mn, struct ib_ucontext, mn);
+
+	if (!context->invalidate_range)
+		return;
+
+	ib_ucontext_notifier_start_account(context);
+	down_read(&context->umem_rwsem);
+	rbt_ib_umem_for_each_in_range(&context->umem_tree, 0,
+				      ULLONG_MAX,
+				      ib_umem_notifier_release_trampoline,
+				      NULL);
+	up_read(&context->umem_rwsem);
+}
+
+static int invalidate_page_trampoline(struct ib_umem *item, u64 start,
+				      u64 end, void *cookie)
+{
+	ib_umem_notifier_start_account(item);
+	item->context->invalidate_range(item, start, start + PAGE_SIZE);
+	ib_umem_notifier_end_account(item);
+	return 0;
+}
+
+static void ib_umem_notifier_invalidate_page(struct mmu_notifier *mn,
+					     struct mm_struct *mm,
+					     unsigned long address)
+{
+	struct ib_ucontext *context = container_of(mn, struct ib_ucontext, mn);
+
+	if (!context->invalidate_range)
+		return;
+
+	ib_ucontext_notifier_start_account(context);
+	down_read(&context->umem_rwsem);
+	rbt_ib_umem_for_each_in_range(&context->umem_tree, address,
+				      address + PAGE_SIZE,
+				      invalidate_page_trampoline, NULL);
+	up_read(&context->umem_rwsem);
+	ib_ucontext_notifier_end_account(context);
+}
+
+static int invalidate_range_start_trampoline(struct ib_umem *item, u64 start,
+					     u64 end, void *cookie)
+{
+	ib_umem_notifier_start_account(item);
+	item->context->invalidate_range(item, start, end);
+	return 0;
+}
+
+static void ib_umem_notifier_invalidate_range_start(struct mmu_notifier *mn,
+						    struct mm_struct *mm,
+						    unsigned long start,
+						    unsigned long end)
+{
+	struct ib_ucontext *context = container_of(mn, struct ib_ucontext, mn);
+
+	if (!context->invalidate_range)
+		return;
+
+	ib_ucontext_notifier_start_account(context);
+	down_read(&context->umem_rwsem);
+	rbt_ib_umem_for_each_in_range(&context->umem_tree, start,
+				      end,
+				      invalidate_range_start_trampoline, NULL);
+	up_read(&context->umem_rwsem);
+}
+
+static int invalidate_range_end_trampoline(struct ib_umem *item, u64 start,
+					   u64 end, void *cookie)
+{
+	ib_umem_notifier_end_account(item);
+	return 0;
+}
+
+static void ib_umem_notifier_invalidate_range_end(struct mmu_notifier *mn,
+						  struct mm_struct *mm,
+						  unsigned long start,
+						  unsigned long end)
+{
+	struct ib_ucontext *context = container_of(mn, struct ib_ucontext, mn);
+
+	if (!context->invalidate_range)
+		return;
+
+	down_read(&context->umem_rwsem);
+	rbt_ib_umem_for_each_in_range(&context->umem_tree, start,
+				      end,
+				      invalidate_range_end_trampoline, NULL);
+	up_read(&context->umem_rwsem);
+	ib_ucontext_notifier_end_account(context);
+}
+
+static struct mmu_notifier_ops ib_umem_notifiers = {
+	.release                    = ib_umem_notifier_release,
+	.invalidate_page            = ib_umem_notifier_invalidate_page,
+	.invalidate_range_start     = ib_umem_notifier_invalidate_range_start,
+	.invalidate_range_end       = ib_umem_notifier_invalidate_range_end,
+};
+
+int ib_umem_odp_get(struct ib_ucontext *context, struct ib_umem *umem)
+{
+	int ret_val;
+	struct pid *our_pid;
+	struct mm_struct *mm = get_task_mm(current);
+
+	if (!mm)
+		return -EINVAL;
+
+	/* Prevent creating ODP MRs in child processes */
+	rcu_read_lock();
+	our_pid = get_task_pid(current->group_leader, PIDTYPE_PID);
+	rcu_read_unlock();
+	put_pid(our_pid);
+	if (context->tgid != our_pid) {
+		ret_val = -EINVAL;
+		goto out_mm;
+	}
+
+	umem->hugetlb = 0;
+	umem->odp_data = kzalloc(sizeof(*umem->odp_data), GFP_KERNEL);
+	if (!umem->odp_data) {
+		ret_val = -ENOMEM;
+		goto out_mm;
+	}
+	umem->odp_data->umem = umem;
+
+	mutex_init(&umem->odp_data->umem_mutex);
+
+	init_completion(&umem->odp_data->notifier_completion);
+
+	umem->odp_data->page_list = vzalloc(ib_umem_num_pages(umem) *
+					    sizeof(*umem->odp_data->page_list));
+	if (!umem->odp_data->page_list) {
+		ret_val = -ENOMEM;
+		goto out_odp_data;
+	}
+
+	umem->odp_data->dma_list = vzalloc(ib_umem_num_pages(umem) *
+					  sizeof(*umem->odp_data->dma_list));
+	if (!umem->odp_data->dma_list) {
+		ret_val = -ENOMEM;
+		goto out_page_list;
+	}
+
+	/*
+	 * When using MMU notifiers, we will get a
+	 * notification before the "current" task (and MM) is
+	 * destroyed. We use the umem_rwsem semaphore to synchronize.
+	 */
+	down_write(&context->umem_rwsem);
+	context->odp_mrs_count++;
+	if (likely(ib_umem_start(umem) != ib_umem_end(umem)))
+		rbt_ib_umem_insert(&umem->odp_data->interval_tree,
+				   &context->umem_tree);
+	if (likely(!atomic_read(&context->notifier_count)))
+		umem->odp_data->mn_counters_active = true;
+	else
+		list_add(&umem->odp_data->no_private_counters,
+			 &context->no_private_counters);
+	downgrade_write(&context->umem_rwsem);
+
+	if (context->odp_mrs_count == 1) {
+		/*
+		 * Note that at this point, no MMU notifier is running
+		 * for this context!
+		 */
+		atomic_set(&context->notifier_count, 0);
+		INIT_HLIST_NODE(&context->mn.hlist);
+		context->mn.ops = &ib_umem_notifiers;
+		/*
+		 * Lock-dep detects a false positive for mmap_sem vs.
+		 * umem_rwsem, due to not grasping downgrade_write correctly.
+		 */
+		lockdep_off();
+		ret_val = mmu_notifier_register(&context->mn, mm);
+		lockdep_on();
+		if (ret_val) {
+			pr_err("Failed to register mmu_notifier %d\n", ret_val);
+			ret_val = -EBUSY;
+			goto out_mutex;
+		}
+	}
+
+	up_read(&context->umem_rwsem);
+
+	/*
+	 * Note that doing an mmput can cause a notifier for the relevant mm.
+	 * If the notifier is called while we hold the umem_rwsem, this will
+	 * cause a deadlock. Therefore, we release the reference only after we
+	 * released the semaphore.
+	 */
+	mmput(mm);
+	return 0;
+
+out_mutex:
+	up_read(&context->umem_rwsem);
+	vfree(umem->odp_data->dma_list);
+out_page_list:
+	vfree(umem->odp_data->page_list);
+out_odp_data:
+	kfree(umem->odp_data);
+out_mm:
+	mmput(mm);
+	return ret_val;
+}
+
+void ib_umem_odp_release(struct ib_umem *umem)
+{
+	struct ib_ucontext *context = umem->context;
+
+	/*
+	 * Ensure that no more pages are mapped in the umem.
+	 *
+	 * It is the driver's responsibility to ensure, before calling us,
+	 * that the hardware will not attempt to access the MR any more.
+	 */
+	ib_umem_odp_unmap_dma_pages(umem, ib_umem_start(umem),
+				    ib_umem_end(umem));
+
+	down_write(&context->umem_rwsem);
+	if (likely(ib_umem_start(umem) != ib_umem_end(umem)))
+		rbt_ib_umem_remove(&umem->odp_data->interval_tree,
+				   &context->umem_tree);
+	context->odp_mrs_count--;
+	if (!umem->odp_data->mn_counters_active) {
+		list_del(&umem->odp_data->no_private_counters);
+		complete_all(&umem->odp_data->notifier_completion);
+	}
+
+	/*
+	 * Downgrade the lock to a read lock. This ensures that the notifiers
+	 * (who lock the mutex for reading) will be able to finish, and we
+	 * will be able to enventually obtain the mmu notifiers SRCU. Note
+	 * that since we are doing it atomically, no other user could register
+	 * and unregister while we do the check.
+	 */
+	downgrade_write(&context->umem_rwsem);
+	if (!context->odp_mrs_count) {
+		struct task_struct *owning_process = NULL;
+		struct mm_struct *owning_mm        = NULL;
+
+		owning_process = get_pid_task(context->tgid,
+					      PIDTYPE_PID);
+		if (owning_process == NULL)
+			/*
+			 * The process is already dead, notifier were removed
+			 * already.
+			 */
+			goto out;
+
+		owning_mm = get_task_mm(owning_process);
+		if (owning_mm == NULL)
+			/*
+			 * The process' mm is already dead, notifier were
+			 * removed already.
+			 */
+			goto out_put_task;
+		mmu_notifier_unregister(&context->mn, owning_mm);
+
+		mmput(owning_mm);
+
+out_put_task:
+		put_task_struct(owning_process);
+	}
+out:
+	up_read(&context->umem_rwsem);
+
+	vfree(umem->odp_data->dma_list);
+	vfree(umem->odp_data->page_list);
+	kfree(umem->odp_data);
+	kfree(umem);
+}
+
+/*
+ * Map for DMA and insert a single page into the on-demand paging page tables.
+ *
+ * @umem: the umem to insert the page to.
+ * @page_index: index in the umem to add the page to.
+ * @page: the page struct to map and add.
+ * @access_mask: access permissions needed for this page.
+ * @current_seq: sequence number for synchronization with invalidations.
+ *               the sequence number is taken from
+ *               umem->odp_data->notifiers_seq.
+ *
+ * The function returns -EFAULT if the DMA mapping operation fails. It returns
+ * -EAGAIN if a concurrent invalidation prevents us from updating the page.
+ *
+ * The page is released via put_page even if the operation failed. For
+ * on-demand pinning, the page is released whenever it isn't stored in the
+ * umem.
+ */
+static int ib_umem_odp_map_dma_single_page(
+		struct ib_umem *umem,
+		int page_index,
+		u64 base_virt_addr,
+		struct page *page,
+		u64 access_mask,
+		unsigned long current_seq)
+{
+	struct ib_device *dev = umem->context->device;
+	dma_addr_t dma_addr;
+	int stored_page = 0;
+	int remove_existing_mapping = 0;
+	int ret = 0;
+
+	mutex_lock(&umem->odp_data->umem_mutex);
+	/*
+	 * Note: we avoid writing if seq is different from the initial seq, to
+	 * handle case of a racing notifier. This check also allows us to bail
+	 * early if we have a notifier running in parallel with us.
+	 */
+	if (ib_umem_mmu_notifier_retry(umem, current_seq)) {
+		ret = -EAGAIN;
+		goto out;
+	}
+	if (!(umem->odp_data->dma_list[page_index])) {
+		dma_addr = ib_dma_map_page(dev,
+					   page,
+					   0, PAGE_SIZE,
+					   DMA_BIDIRECTIONAL);
+		if (ib_dma_mapping_error(dev, dma_addr)) {
+			ret = -EFAULT;
+			goto out;
+		}
+		umem->odp_data->dma_list[page_index] = dma_addr | access_mask;
+		umem->odp_data->page_list[page_index] = page;
+		stored_page = 1;
+	} else if (umem->odp_data->page_list[page_index] == page) {
+		umem->odp_data->dma_list[page_index] |= access_mask;
+	} else {
+		pr_err("error: got different pages in IB device and from get_user_pages. IB device page: %p, gup page: %p\n",
+		       umem->odp_data->page_list[page_index], page);
+		/* Better remove the mapping now, to prevent any further
+		 * damage. */
+		remove_existing_mapping = 1;
+	}
+
+out:
+	mutex_unlock(&umem->odp_data->umem_mutex);
+
+	/* On Demand Paging - avoid pinning the page */
+	if (umem->context->invalidate_range || !stored_page)
+		put_page(page);
+
+	if (remove_existing_mapping && umem->context->invalidate_range) {
+		invalidate_page_trampoline(
+			umem,
+			base_virt_addr + (page_index * PAGE_SIZE),
+			base_virt_addr + ((page_index+1)*PAGE_SIZE),
+			NULL);
+		ret = -EAGAIN;
+	}
+
+	return ret;
+}
+
+/**
+ * ib_umem_odp_map_dma_pages - Pin and DMA map userspace memory in an ODP MR.
+ *
+ * Pins the range of pages passed in the argument, and maps them to
+ * DMA addresses. The DMA addresses of the mapped pages is updated in
+ * umem->odp_data->dma_list.
+ *
+ * Returns the number of pages mapped in success, negative error code
+ * for failure.
+ * An -EAGAIN error code is returned when a concurrent mmu notifier prevents
+ * the function from completing its task.
+ *
+ * @umem: the umem to map and pin
+ * @user_virt: the address from which we need to map.
+ * @bcnt: the minimal number of bytes to pin and map. The mapping might be
+ *        bigger due to alignment, and may also be smaller in case of an error
+ *        pinning or mapping a page. The actual pages mapped is returned in
+ *        the return value.
+ * @access_mask: bit mask of the requested access permissions for the given
+ *               range.
+ * @current_seq: the MMU notifiers sequance value for synchronization with
+ *               invalidations. the sequance number is read from
+ *               umem->odp_data->notifiers_seq before calling this function
+ */
+int ib_umem_odp_map_dma_pages(struct ib_umem *umem, u64 user_virt, u64 bcnt,
+			      u64 access_mask, unsigned long current_seq)
+{
+	struct task_struct *owning_process  = NULL;
+	struct mm_struct   *owning_mm       = NULL;
+	struct page       **local_page_list = NULL;
+	u64 off;
+	int j, k, ret = 0, start_idx, npages = 0;
+	u64 base_virt_addr;
+
+	if (access_mask == 0)
+		return -EINVAL;
+
+	if (user_virt < ib_umem_start(umem) ||
+	    user_virt + bcnt > ib_umem_end(umem))
+		return -EFAULT;
+
+	local_page_list = (struct page **)__get_free_page(GFP_KERNEL);
+	if (!local_page_list)
+		return -ENOMEM;
+
+	off = user_virt & (~PAGE_MASK);
+	user_virt = user_virt & PAGE_MASK;
+	base_virt_addr = user_virt;
+	bcnt += off; /* Charge for the first page offset as well. */
+
+	owning_process = get_pid_task(umem->context->tgid, PIDTYPE_PID);
+	if (owning_process == NULL) {
+		ret = -EINVAL;
+		goto out_no_task;
+	}
+
+	owning_mm = get_task_mm(owning_process);
+	if (owning_mm == NULL) {
+		ret = -EINVAL;
+		goto out_put_task;
+	}
+
+	start_idx = (user_virt - ib_umem_start(umem)) >> PAGE_SHIFT;
+	k = start_idx;
+
+	while (bcnt > 0) {
+		const size_t gup_num_pages =
+			min_t(size_t, ALIGN(bcnt, PAGE_SIZE) / PAGE_SIZE,
+			      PAGE_SIZE / sizeof(struct page *));
+
+		down_read(&owning_mm->mmap_sem);
+		/*
+		 * Note: this might result in redundent page getting. We can
+		 * avoid this by checking dma_list to be 0 before calling
+		 * get_user_pages. However, this make the code much more
+		 * complex (and doesn't gain us much performance in most use
+		 * cases).
+		 */
+		npages = get_user_pages(owning_process, owning_mm, user_virt,
+					gup_num_pages,
+					access_mask & ODP_WRITE_ALLOWED_BIT, 0,
+					local_page_list, NULL);
+		up_read(&owning_mm->mmap_sem);
+
+		if (npages < 0)
+			break;
+
+		bcnt -= min_t(size_t, npages << PAGE_SHIFT, bcnt);
+		user_virt += npages << PAGE_SHIFT;
+		for (j = 0; j < npages; ++j) {
+			ret = ib_umem_odp_map_dma_single_page(
+				umem, k, base_virt_addr, local_page_list[j],
+				access_mask, current_seq);
+			if (ret < 0)
+				break;
+			k++;
+		}
+
+		if (ret < 0) {
+			/* Release left over pages when handling errors. */
+			for (++j; j < npages; ++j)
+				put_page(local_page_list[j]);
+			break;
+		}
+	}
+
+	if (ret >= 0) {
+		if (npages < 0 && k == start_idx)
+			ret = npages;
+		else
+			ret = k - start_idx;
+	}
+
+	mmput(owning_mm);
+out_put_task:
+	put_task_struct(owning_process);
+out_no_task:
+	free_page((unsigned long)local_page_list);
+	return ret;
+}
+EXPORT_SYMBOL(ib_umem_odp_map_dma_pages);
+
+void ib_umem_odp_unmap_dma_pages(struct ib_umem *umem, u64 virt,
+				 u64 bound)
+{
+	int idx;
+	u64 addr;
+	struct ib_device *dev = umem->context->device;
+
+	virt  = max_t(u64, virt,  ib_umem_start(umem));
+	bound = min_t(u64, bound, ib_umem_end(umem));
+	/* Note that during the run of this function, the
+	 * notifiers_count of the MR is > 0, preventing any racing
+	 * faults from completion. We might be racing with other
+	 * invalidations, so we must make sure we free each page only
+	 * once. */
+	for (addr = virt; addr < bound; addr += (u64)umem->page_size) {
+		idx = (addr - ib_umem_start(umem)) / PAGE_SIZE;
+		mutex_lock(&umem->odp_data->umem_mutex);
+		if (umem->odp_data->page_list[idx]) {
+			struct page *page = umem->odp_data->page_list[idx];
+			struct page *head_page = compound_head(page);
+			dma_addr_t dma = umem->odp_data->dma_list[idx];
+			dma_addr_t dma_addr = dma & ODP_DMA_ADDR_MASK;
+
+			WARN_ON(!dma_addr);
+
+			ib_dma_unmap_page(dev, dma_addr, PAGE_SIZE,
+					  DMA_BIDIRECTIONAL);
+			if (dma & ODP_WRITE_ALLOWED_BIT)
+				/*
+				 * set_page_dirty prefers being called with
+				 * the page lock. However, MMU notifiers are
+				 * called sometimes with and sometimes without
+				 * the lock. We rely on the umem_mutex instead
+				 * to prevent other mmu notifiers from
+				 * continuing and allowing the page mapping to
+				 * be removed.
+				 */
+				set_page_dirty(head_page);
+			/* on demand pinning support */
+			if (!umem->context->invalidate_range)
+				put_page(page);
+			umem->odp_data->page_list[idx] = NULL;
+			umem->odp_data->dma_list[idx] = 0;
+		}
+		mutex_unlock(&umem->odp_data->umem_mutex);
+	}
+}
+EXPORT_SYMBOL(ib_umem_odp_unmap_dma_pages);
diff --git a/drivers/infiniband/core/umem_rbtree.c b/drivers/infiniband/core/umem_rbtree.c
new file mode 100644
index 000000000000..727d788448f5
--- /dev/null
+++ b/drivers/infiniband/core/umem_rbtree.c
@@ -0,0 +1,94 @@
+/*
+ * Copyright (c) 2014 Mellanox Technologies. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/interval_tree_generic.h>
+#include <linux/sched.h>
+#include <linux/gfp.h>
+#include <rdma/ib_umem_odp.h>
+
+/*
+ * The ib_umem list keeps track of memory regions for which the HW
+ * device request to receive notification when the related memory
+ * mapping is changed.
+ *
+ * ib_umem_lock protects the list.
+ */
+
+static inline u64 node_start(struct umem_odp_node *n)
+{
+	struct ib_umem_odp *umem_odp =
+			container_of(n, struct ib_umem_odp, interval_tree);
+
+	return ib_umem_start(umem_odp->umem);
+}
+
+/* Note that the representation of the intervals in the interval tree
+ * considers the ending point as contained in the interval, while the
+ * function ib_umem_end returns the first address which is not contained
+ * in the umem.
+ */
+static inline u64 node_last(struct umem_odp_node *n)
+{
+	struct ib_umem_odp *umem_odp =
+			container_of(n, struct ib_umem_odp, interval_tree);
+
+	return ib_umem_end(umem_odp->umem) - 1;
+}
+
+INTERVAL_TREE_DEFINE(struct umem_odp_node, rb, u64, __subtree_last,
+		     node_start, node_last, , rbt_ib_umem)
+
+/* @last is not a part of the interval. See comment for function
+ * node_last.
+ */
+int rbt_ib_umem_for_each_in_range(struct rb_root *root,
+				  u64 start, u64 last,
+				  umem_call_back cb,
+				  void *cookie)
+{
+	int ret_val = 0;
+	struct umem_odp_node *node;
+	struct ib_umem_odp *umem;
+
+	if (unlikely(start == last))
+		return ret_val;
+
+	for (node = rbt_ib_umem_iter_first(root, start, last - 1); node;
+			node = rbt_ib_umem_iter_next(node, start, last - 1)) {
+		umem = container_of(node, struct ib_umem_odp, interval_tree);
+		ret_val = cb(umem->umem, start, last, cookie) || ret_val;
+	}
+
+	return ret_val;
+}
diff --git a/drivers/infiniband/core/uverbs.h b/drivers/infiniband/core/uverbs.h
index 643c08a025a5..b716b0815644 100644
--- a/drivers/infiniband/core/uverbs.h
+++ b/drivers/infiniband/core/uverbs.h
@@ -258,5 +258,6 @@ IB_UVERBS_DECLARE_CMD(close_xrcd);
 
 IB_UVERBS_DECLARE_EX_CMD(create_flow);
 IB_UVERBS_DECLARE_EX_CMD(destroy_flow);
+IB_UVERBS_DECLARE_EX_CMD(query_device);
 
 #endif /* UVERBS_H */
diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index 0600c50e6215..532d8eba8b02 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -36,6 +36,7 @@
 #include <linux/file.h>
 #include <linux/fs.h>
 #include <linux/slab.h>
+#include <linux/sched.h>
 
 #include <asm/uaccess.h>
 
@@ -288,6 +289,9 @@ ssize_t ib_uverbs_get_context(struct ib_uverbs_file *file,
 	struct ib_uverbs_get_context_resp resp;
 	struct ib_udata                   udata;
 	struct ib_device                 *ibdev = file->device->ib_dev;
+#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
+	struct ib_device_attr		  dev_attr;
+#endif
 	struct ib_ucontext		 *ucontext;
 	struct file			 *filp;
 	int ret;
@@ -325,8 +329,25 @@ ssize_t ib_uverbs_get_context(struct ib_uverbs_file *file,
 	INIT_LIST_HEAD(&ucontext->ah_list);
 	INIT_LIST_HEAD(&ucontext->xrcd_list);
 	INIT_LIST_HEAD(&ucontext->rule_list);
+	rcu_read_lock();
+	ucontext->tgid = get_task_pid(current->group_leader, PIDTYPE_PID);
+	rcu_read_unlock();
 	ucontext->closing = 0;
 
+#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
+	ucontext->umem_tree = RB_ROOT;
+	init_rwsem(&ucontext->umem_rwsem);
+	ucontext->odp_mrs_count = 0;
+	INIT_LIST_HEAD(&ucontext->no_private_counters);
+
+	ret = ib_query_device(ibdev, &dev_attr);
+	if (ret)
+		goto err_free;
+	if (!(dev_attr.device_cap_flags & IB_DEVICE_ON_DEMAND_PAGING))
+		ucontext->invalidate_range = NULL;
+
+#endif
+
 	resp.num_comp_vectors = file->device->num_comp_vectors;
 
 	ret = get_unused_fd_flags(O_CLOEXEC);
@@ -371,6 +392,7 @@ err_fd:
 	put_unused_fd(resp.async_fd);
 
 err_free:
+	put_pid(ucontext->tgid);
 	ibdev->dealloc_ucontext(ucontext);
 
 err:
@@ -378,6 +400,52 @@ err:
 	return ret;
 }
 
+static void copy_query_dev_fields(struct ib_uverbs_file *file,
+				  struct ib_uverbs_query_device_resp *resp,
+				  struct ib_device_attr *attr)
+{
+	resp->fw_ver		= attr->fw_ver;
+	resp->node_guid		= file->device->ib_dev->node_guid;
+	resp->sys_image_guid	= attr->sys_image_guid;
+	resp->max_mr_size	= attr->max_mr_size;
+	resp->page_size_cap	= attr->page_size_cap;
+	resp->vendor_id		= attr->vendor_id;
+	resp->vendor_part_id	= attr->vendor_part_id;
+	resp->hw_ver		= attr->hw_ver;
+	resp->max_qp		= attr->max_qp;
+	resp->max_qp_wr		= attr->max_qp_wr;
+	resp->device_cap_flags	= attr->device_cap_flags;
+	resp->max_sge		= attr->max_sge;
+	resp->max_sge_rd	= attr->max_sge_rd;
+	resp->max_cq		= attr->max_cq;
+	resp->max_cqe		= attr->max_cqe;
+	resp->max_mr		= attr->max_mr;
+	resp->max_pd		= attr->max_pd;
+	resp->max_qp_rd_atom	= attr->max_qp_rd_atom;
+	resp->max_ee_rd_atom	= attr->max_ee_rd_atom;
+	resp->max_res_rd_atom	= attr->max_res_rd_atom;
+	resp->max_qp_init_rd_atom	= attr->max_qp_init_rd_atom;
+	resp->max_ee_init_rd_atom	= attr->max_ee_init_rd_atom;
+	resp->atomic_cap		= attr->atomic_cap;
+	resp->max_ee			= attr->max_ee;
+	resp->max_rdd			= attr->max_rdd;
+	resp->max_mw			= attr->max_mw;
+	resp->max_raw_ipv6_qp		= attr->max_raw_ipv6_qp;
+	resp->max_raw_ethy_qp		= attr->max_raw_ethy_qp;
+	resp->max_mcast_grp		= attr->max_mcast_grp;
+	resp->max_mcast_qp_attach	= attr->max_mcast_qp_attach;
+	resp->max_total_mcast_qp_attach	= attr->max_total_mcast_qp_attach;
+	resp->max_ah			= attr->max_ah;
+	resp->max_fmr			= attr->max_fmr;
+	resp->max_map_per_fmr		= attr->max_map_per_fmr;
+	resp->max_srq			= attr->max_srq;
+	resp->max_srq_wr		= attr->max_srq_wr;
+	resp->max_srq_sge		= attr->max_srq_sge;
+	resp->max_pkeys			= attr->max_pkeys;
+	resp->local_ca_ack_delay	= attr->local_ca_ack_delay;
+	resp->phys_port_cnt		= file->device->ib_dev->phys_port_cnt;
+}
+
 ssize_t ib_uverbs_query_device(struct ib_uverbs_file *file,
 			       const char __user *buf,
 			       int in_len, int out_len)
@@ -398,47 +466,7 @@ ssize_t ib_uverbs_query_device(struct ib_uverbs_file *file,
 		return ret;
 
 	memset(&resp, 0, sizeof resp);
-
-	resp.fw_ver 		       = attr.fw_ver;
-	resp.node_guid 		       = file->device->ib_dev->node_guid;
-	resp.sys_image_guid 	       = attr.sys_image_guid;
-	resp.max_mr_size 	       = attr.max_mr_size;
-	resp.page_size_cap 	       = attr.page_size_cap;
-	resp.vendor_id 		       = attr.vendor_id;
-	resp.vendor_part_id 	       = attr.vendor_part_id;
-	resp.hw_ver 		       = attr.hw_ver;
-	resp.max_qp 		       = attr.max_qp;
-	resp.max_qp_wr 		       = attr.max_qp_wr;
-	resp.device_cap_flags 	       = attr.device_cap_flags;
-	resp.max_sge 		       = attr.max_sge;
-	resp.max_sge_rd 	       = attr.max_sge_rd;
-	resp.max_cq 		       = attr.max_cq;
-	resp.max_cqe 		       = attr.max_cqe;
-	resp.max_mr 		       = attr.max_mr;
-	resp.max_pd 		       = attr.max_pd;
-	resp.max_qp_rd_atom 	       = attr.max_qp_rd_atom;
-	resp.max_ee_rd_atom 	       = attr.max_ee_rd_atom;
-	resp.max_res_rd_atom 	       = attr.max_res_rd_atom;
-	resp.max_qp_init_rd_atom       = attr.max_qp_init_rd_atom;
-	resp.max_ee_init_rd_atom       = attr.max_ee_init_rd_atom;
-	resp.atomic_cap 	       = attr.atomic_cap;
-	resp.max_ee 		       = attr.max_ee;
-	resp.max_rdd 		       = attr.max_rdd;
-	resp.max_mw 		       = attr.max_mw;
-	resp.max_raw_ipv6_qp 	       = attr.max_raw_ipv6_qp;
-	resp.max_raw_ethy_qp 	       = attr.max_raw_ethy_qp;
-	resp.max_mcast_grp 	       = attr.max_mcast_grp;
-	resp.max_mcast_qp_attach       = attr.max_mcast_qp_attach;
-	resp.max_total_mcast_qp_attach = attr.max_total_mcast_qp_attach;
-	resp.max_ah 		       = attr.max_ah;
-	resp.max_fmr 		       = attr.max_fmr;
-	resp.max_map_per_fmr 	       = attr.max_map_per_fmr;
-	resp.max_srq 		       = attr.max_srq;
-	resp.max_srq_wr 	       = attr.max_srq_wr;
-	resp.max_srq_sge 	       = attr.max_srq_sge;
-	resp.max_pkeys 		       = attr.max_pkeys;
-	resp.local_ca_ack_delay        = attr.local_ca_ack_delay;
-	resp.phys_port_cnt	       = file->device->ib_dev->phys_port_cnt;
+	copy_query_dev_fields(file, &resp, &attr);
 
 	if (copy_to_user((void __user *) (unsigned long) cmd.response,
 			 &resp, sizeof resp))
@@ -947,6 +975,18 @@ ssize_t ib_uverbs_reg_mr(struct ib_uverbs_file *file,
 		goto err_free;
 	}
 
+	if (cmd.access_flags & IB_ACCESS_ON_DEMAND) {
+		struct ib_device_attr attr;
+
+		ret = ib_query_device(pd->device, &attr);
+		if (ret || !(attr.device_cap_flags &
+				IB_DEVICE_ON_DEMAND_PAGING)) {
+			pr_debug("ODP support not available\n");
+			ret = -EINVAL;
+			goto err_put;
+		}
+	}
+
 	mr = pd->device->reg_user_mr(pd, cmd.start, cmd.length, cmd.hca_va,
 				     cmd.access_flags, &udata);
 	if (IS_ERR(mr)) {
@@ -2518,6 +2558,8 @@ ssize_t ib_uverbs_create_ah(struct ib_uverbs_file *file,
 	attr.grh.sgid_index    = cmd.attr.grh.sgid_index;
 	attr.grh.hop_limit     = cmd.attr.grh.hop_limit;
 	attr.grh.traffic_class = cmd.attr.grh.traffic_class;
+	attr.vlan_id           = 0;
+	memset(&attr.dmac, 0, sizeof(attr.dmac));
 	memcpy(attr.grh.dgid.raw, cmd.attr.grh.dgid, 16);
 
 	ah = ib_create_ah(pd, &attr);
@@ -3251,3 +3293,52 @@ ssize_t ib_uverbs_destroy_srq(struct ib_uverbs_file *file,
 
 	return ret ? ret : in_len;
 }
+
+int ib_uverbs_ex_query_device(struct ib_uverbs_file *file,
+			      struct ib_udata *ucore,
+			      struct ib_udata *uhw)
+{
+	struct ib_uverbs_ex_query_device_resp resp;
+	struct ib_uverbs_ex_query_device  cmd;
+	struct ib_device_attr attr;
+	struct ib_device *device;
+	int err;
+
+	device = file->device->ib_dev;
+	if (ucore->inlen < sizeof(cmd))
+		return -EINVAL;
+
+	err = ib_copy_from_udata(&cmd, ucore, sizeof(cmd));
+	if (err)
+		return err;
+
+	if (cmd.reserved)
+		return -EINVAL;
+
+	err = device->query_device(device, &attr);
+	if (err)
+		return err;
+
+	memset(&resp, 0, sizeof(resp));
+	copy_query_dev_fields(file, &resp.base, &attr);
+	resp.comp_mask = 0;
+
+#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
+	if (cmd.comp_mask & IB_USER_VERBS_EX_QUERY_DEVICE_ODP) {
+		resp.odp_caps.general_caps = attr.odp_caps.general_caps;
+		resp.odp_caps.per_transport_caps.rc_odp_caps =
+			attr.odp_caps.per_transport_caps.rc_odp_caps;
+		resp.odp_caps.per_transport_caps.uc_odp_caps =
+			attr.odp_caps.per_transport_caps.uc_odp_caps;
+		resp.odp_caps.per_transport_caps.ud_odp_caps =
+			attr.odp_caps.per_transport_caps.ud_odp_caps;
+		resp.comp_mask |= IB_USER_VERBS_EX_QUERY_DEVICE_ODP;
+	}
+#endif
+
+	err = ib_copy_to_udata(ucore, &resp, sizeof(resp));
+	if (err)
+		return err;
+
+	return 0;
+}
diff --git a/drivers/infiniband/core/uverbs_main.c b/drivers/infiniband/core/uverbs_main.c
index c73b22a257fe..e6c23b9eab33 100644
--- a/drivers/infiniband/core/uverbs_main.c
+++ b/drivers/infiniband/core/uverbs_main.c
@@ -122,7 +122,8 @@ static int (*uverbs_ex_cmd_table[])(struct ib_uverbs_file *file,
 				    struct ib_udata *ucore,
 				    struct ib_udata *uhw) = {
 	[IB_USER_VERBS_EX_CMD_CREATE_FLOW]	= ib_uverbs_ex_create_flow,
-	[IB_USER_VERBS_EX_CMD_DESTROY_FLOW]	= ib_uverbs_ex_destroy_flow
+	[IB_USER_VERBS_EX_CMD_DESTROY_FLOW]	= ib_uverbs_ex_destroy_flow,
+	[IB_USER_VERBS_EX_CMD_QUERY_DEVICE]	= ib_uverbs_ex_query_device
 };
 
 static void ib_uverbs_add_one(struct ib_device *device);
@@ -296,6 +297,8 @@ static int ib_uverbs_cleanup_ucontext(struct ib_uverbs_file *file,
 		kfree(uobj);
 	}
 
+	put_pid(context->tgid);
+
 	return context->device->dealloc_ucontext(context);
 }
 
@@ -477,6 +480,7 @@ static void ib_uverbs_async_handler(struct ib_uverbs_file *file,
 
 	entry->desc.async.element    = element;
 	entry->desc.async.event_type = event;
+	entry->desc.async.reserved   = 0;
 	entry->counter               = counter;
 
 	list_add_tail(&entry->list, &file->async_file->event_list);
@@ -502,6 +506,10 @@ void ib_uverbs_qp_event_handler(struct ib_event *event, void *context_ptr)
 {
 	struct ib_uevent_object *uobj;
 
+	/* for XRC target qp's, check that qp is live */
+	if (!event->element.qp->uobject || !event->element.qp->uobject->live)
+		return;
+
 	uobj = container_of(event->element.qp->uobject,
 			    struct ib_uevent_object, uobject);
 
diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
index c2b89cc5dbca..f93eb8da7b5a 100644
--- a/drivers/infiniband/core/verbs.c
+++ b/drivers/infiniband/core/verbs.c
@@ -879,7 +879,8 @@ int ib_resolve_eth_l2_attrs(struct ib_qp *qp,
 		if (rdma_link_local_addr((struct in6_addr *)qp_attr->ah_attr.grh.dgid.raw)) {
 			rdma_get_ll_mac((struct in6_addr *)qp_attr->ah_attr.grh.dgid.raw, qp_attr->ah_attr.dmac);
 			rdma_get_ll_mac((struct in6_addr *)sgid.raw, qp_attr->smac);
-			qp_attr->vlan_id = rdma_get_vlan_id(&sgid);
+			if (!(*qp_attr_mask & IB_QP_VID))
+				qp_attr->vlan_id = rdma_get_vlan_id(&sgid);
 		} else {
 			ret = rdma_addr_find_dmac_by_grh(&sgid, &qp_attr->ah_attr.grh.dgid,
 					qp_attr->ah_attr.dmac, &qp_attr->vlan_id);
diff --git a/drivers/infiniband/hw/amso1100/c2_provider.c b/drivers/infiniband/hw/amso1100/c2_provider.c
index 2d5cbf4363e4..bdf3507810cb 100644
--- a/drivers/infiniband/hw/amso1100/c2_provider.c
+++ b/drivers/infiniband/hw/amso1100/c2_provider.c
@@ -476,7 +476,7 @@ static struct ib_mr *c2_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 					 c2mr->umem->page_size,
 					 i,
 					 length,
-					 c2mr->umem->offset,
+					 ib_umem_offset(c2mr->umem),
 					 &kva,
 					 c2_convert_access(acc),
 					 c2mr);
diff --git a/drivers/infiniband/hw/cxgb4/cm.c b/drivers/infiniband/hw/cxgb4/cm.c
index c2fb71c182a8..9edc200b311d 100644
--- a/drivers/infiniband/hw/cxgb4/cm.c
+++ b/drivers/infiniband/hw/cxgb4/cm.c
@@ -236,10 +236,12 @@ static void release_tid(struct c4iw_rdev *rdev, u32 hwtid, struct sk_buff *skb)
 static void set_emss(struct c4iw_ep *ep, u16 opt)
 {
 	ep->emss = ep->com.dev->rdev.lldi.mtus[GET_TCPOPT_MSS(opt)] -
-		   sizeof(struct iphdr) - sizeof(struct tcphdr);
+		   ((AF_INET == ep->com.remote_addr.ss_family) ?
+		    sizeof(struct iphdr) : sizeof(struct ipv6hdr)) -
+		   sizeof(struct tcphdr);
 	ep->mss = ep->emss;
 	if (GET_TCPOPT_TSTAMP(opt))
-		ep->emss -= 12;
+		ep->emss -= round_up(TCPOLEN_TIMESTAMP, 4);
 	if (ep->emss < 128)
 		ep->emss = 128;
 	if (ep->emss & 7)
@@ -415,6 +417,7 @@ static struct dst_entry *find_route(struct c4iw_dev *dev, __be32 local_ip,
 		return NULL;
 	if (!our_interface(dev, n->dev) &&
 	    !(n->dev->flags & IFF_LOOPBACK)) {
+		neigh_release(n);
 		dst_release(&rt->dst);
 		return NULL;
 	}
@@ -469,13 +472,13 @@ static void send_flowc(struct c4iw_ep *ep, struct sk_buff *skb)
 	skb = get_skb(skb, flowclen, GFP_KERNEL);
 	flowc = (struct fw_flowc_wr *)__skb_put(skb, flowclen);
 
-	flowc->op_to_nparams = cpu_to_be32(FW_WR_OP(FW_FLOWC_WR) |
-					   FW_FLOWC_WR_NPARAMS(8));
-	flowc->flowid_len16 = cpu_to_be32(FW_WR_LEN16(DIV_ROUND_UP(flowclen,
-					  16)) | FW_WR_FLOWID(ep->hwtid));
+	flowc->op_to_nparams = cpu_to_be32(FW_WR_OP_V(FW_FLOWC_WR) |
+					   FW_FLOWC_WR_NPARAMS_V(8));
+	flowc->flowid_len16 = cpu_to_be32(FW_WR_LEN16_V(DIV_ROUND_UP(flowclen,
+					  16)) | FW_WR_FLOWID_V(ep->hwtid));
 
 	flowc->mnemval[0].mnemonic = FW_FLOWC_MNEM_PFNVFN;
-	flowc->mnemval[0].val = cpu_to_be32(FW_PFVF_CMD_PFN
+	flowc->mnemval[0].val = cpu_to_be32(FW_PFVF_CMD_PFN_V
 					    (ep->com.dev->rdev.lldi.pf));
 	flowc->mnemval[1].mnemonic = FW_FLOWC_MNEM_CH;
 	flowc->mnemval[1].val = cpu_to_be32(ep->tx_chan);
@@ -581,11 +584,14 @@ static void c4iw_record_pm_msg(struct c4iw_ep *ep,
 }
 
 static void best_mtu(const unsigned short *mtus, unsigned short mtu,
-		     unsigned int *idx, int use_ts)
+		     unsigned int *idx, int use_ts, int ipv6)
 {
-	unsigned short hdr_size = sizeof(struct iphdr) +
+	unsigned short hdr_size = (ipv6 ?
+				   sizeof(struct ipv6hdr) :
+				   sizeof(struct iphdr)) +
 				  sizeof(struct tcphdr) +
-				  (use_ts ? 12 : 0);
+				  (use_ts ?
+				   round_up(TCPOLEN_TIMESTAMP, 4) : 0);
 	unsigned short data_size = mtu - hdr_size;
 
 	cxgb4_best_aligned_mtu(mtus, hdr_size, data_size, 8, idx);
@@ -634,7 +640,8 @@ static int send_connect(struct c4iw_ep *ep)
 	set_wr_txq(skb, CPL_PRIORITY_SETUP, ep->ctrlq_idx);
 
 	best_mtu(ep->com.dev->rdev.lldi.mtus, ep->mtu, &mtu_idx,
-		 enable_tcp_timestamps);
+		 enable_tcp_timestamps,
+		 (AF_INET == ep->com.remote_addr.ss_family) ? 0 : 1);
 	wscale = compute_wscale(rcv_win);
 
 	/*
@@ -642,32 +649,33 @@ static int send_connect(struct c4iw_ep *ep)
 	 * remainder will be specified in the rx_data_ack.
 	 */
 	win = ep->rcv_win >> 10;
-	if (win > RCV_BUFSIZ_MASK)
-		win = RCV_BUFSIZ_MASK;
+	if (win > RCV_BUFSIZ_M)
+		win = RCV_BUFSIZ_M;
 
 	opt0 = (nocong ? NO_CONG(1) : 0) |
-	       KEEP_ALIVE(1) |
+	       KEEP_ALIVE_F |
 	       DELACK(1) |
-	       WND_SCALE(wscale) |
-	       MSS_IDX(mtu_idx) |
-	       L2T_IDX(ep->l2t->idx) |
-	       TX_CHAN(ep->tx_chan) |
-	       SMAC_SEL(ep->smac_idx) |
+	       WND_SCALE_V(wscale) |
+	       MSS_IDX_V(mtu_idx) |
+	       L2T_IDX_V(ep->l2t->idx) |
+	       TX_CHAN_V(ep->tx_chan) |
+	       SMAC_SEL_V(ep->smac_idx) |
 	       DSCP(ep->tos) |
-	       ULP_MODE(ULP_MODE_TCPDDP) |
-	       RCV_BUFSIZ(win);
-	opt2 = RX_CHANNEL(0) |
+	       ULP_MODE_V(ULP_MODE_TCPDDP) |
+	       RCV_BUFSIZ_V(win);
+	opt2 = RX_CHANNEL_V(0) |
 	       CCTRL_ECN(enable_ecn) |
-	       RSS_QUEUE_VALID | RSS_QUEUE(ep->rss_qid);
+	       RSS_QUEUE_VALID_F | RSS_QUEUE_V(ep->rss_qid);
 	if (enable_tcp_timestamps)
 		opt2 |= TSTAMPS_EN(1);
 	if (enable_tcp_sack)
 		opt2 |= SACK_EN(1);
 	if (wscale && enable_tcp_window_scaling)
-		opt2 |= WND_SCALE_EN(1);
+		opt2 |= WND_SCALE_EN_F;
 	if (is_t5(ep->com.dev->rdev.lldi.adapter_type)) {
-		opt2 |= T5_OPT_2_VALID;
+		opt2 |= T5_OPT_2_VALID_F;
 		opt2 |= V_CONG_CNTRL(CONG_ALG_TAHOE);
+		opt2 |= CONG_CNTRL_VALID; /* OPT_2_ISS for T5 */
 	}
 	t4_set_arp_err_handler(skb, ep, act_open_req_arp_failure);
 
@@ -713,8 +721,6 @@ static int send_connect(struct c4iw_ep *ep)
 	} else {
 		u32 isn = (prandom_u32() & ~7UL) - 1;
 
-		opt2 |= T5_OPT_2_VALID;
-		opt2 |= CONG_CNTRL_VALID; /* OPT_2_ISS for T5 */
 		if (peer2peer)
 			isn += 4;
 
@@ -730,7 +736,7 @@ static int send_connect(struct c4iw_ep *ep)
 			t5_req->local_ip = la->sin_addr.s_addr;
 			t5_req->peer_ip = ra->sin_addr.s_addr;
 			t5_req->opt0 = cpu_to_be64(opt0);
-			t5_req->params = cpu_to_be64(V_FILTER_TUPLE(
+			t5_req->params = cpu_to_be64(FILTER_TUPLE_V(
 						     cxgb4_select_ntuple(
 					     ep->com.dev->rdev.lldi.ports[0],
 					     ep->l2t)));
@@ -756,10 +762,10 @@ static int send_connect(struct c4iw_ep *ep)
 			t5_req6->peer_ip_lo = *((__be64 *)
 						(ra6->sin6_addr.s6_addr + 8));
 			t5_req6->opt0 = cpu_to_be64(opt0);
-			t5_req6->params = (__force __be64)cpu_to_be32(
+			t5_req6->params = cpu_to_be64(FILTER_TUPLE_V(
 							cxgb4_select_ntuple(
 						ep->com.dev->rdev.lldi.ports[0],
-						ep->l2t));
+						ep->l2t)));
 			t5_req6->rsvd = cpu_to_be32(isn);
 			PDBG("%s snd_isn %u\n", __func__,
 			     be32_to_cpu(t5_req6->rsvd));
@@ -797,16 +803,16 @@ static void send_mpa_req(struct c4iw_ep *ep, struct sk_buff *skb,
 	req = (struct fw_ofld_tx_data_wr *)skb_put(skb, wrlen);
 	memset(req, 0, wrlen);
 	req->op_to_immdlen = cpu_to_be32(
-		FW_WR_OP(FW_OFLD_TX_DATA_WR) |
-		FW_WR_COMPL(1) |
-		FW_WR_IMMDLEN(mpalen));
+		FW_WR_OP_V(FW_OFLD_TX_DATA_WR) |
+		FW_WR_COMPL_F |
+		FW_WR_IMMDLEN_V(mpalen));
 	req->flowid_len16 = cpu_to_be32(
-		FW_WR_FLOWID(ep->hwtid) |
-		FW_WR_LEN16(wrlen >> 4));
+		FW_WR_FLOWID_V(ep->hwtid) |
+		FW_WR_LEN16_V(wrlen >> 4));
 	req->plen = cpu_to_be32(mpalen);
 	req->tunnel_to_proxy = cpu_to_be32(
-		FW_OFLD_TX_DATA_WR_FLUSH(1) |
-		FW_OFLD_TX_DATA_WR_SHOVE(1));
+		FW_OFLD_TX_DATA_WR_FLUSH_F |
+		FW_OFLD_TX_DATA_WR_SHOVE_F);
 
 	mpa = (struct mpa_message *)(req + 1);
 	memcpy(mpa->key, MPA_KEY_REQ, sizeof(mpa->key));
@@ -891,16 +897,16 @@ static int send_mpa_reject(struct c4iw_ep *ep, const void *pdata, u8 plen)
 	req = (struct fw_ofld_tx_data_wr *)skb_put(skb, wrlen);
 	memset(req, 0, wrlen);
 	req->op_to_immdlen = cpu_to_be32(
-		FW_WR_OP(FW_OFLD_TX_DATA_WR) |
-		FW_WR_COMPL(1) |
-		FW_WR_IMMDLEN(mpalen));
+		FW_WR_OP_V(FW_OFLD_TX_DATA_WR) |
+		FW_WR_COMPL_F |
+		FW_WR_IMMDLEN_V(mpalen));
 	req->flowid_len16 = cpu_to_be32(
-		FW_WR_FLOWID(ep->hwtid) |
-		FW_WR_LEN16(wrlen >> 4));
+		FW_WR_FLOWID_V(ep->hwtid) |
+		FW_WR_LEN16_V(wrlen >> 4));
 	req->plen = cpu_to_be32(mpalen);
 	req->tunnel_to_proxy = cpu_to_be32(
-		FW_OFLD_TX_DATA_WR_FLUSH(1) |
-		FW_OFLD_TX_DATA_WR_SHOVE(1));
+		FW_OFLD_TX_DATA_WR_FLUSH_F |
+		FW_OFLD_TX_DATA_WR_SHOVE_F);
 
 	mpa = (struct mpa_message *)(req + 1);
 	memset(mpa, 0, sizeof(*mpa));
@@ -971,16 +977,16 @@ static int send_mpa_reply(struct c4iw_ep *ep, const void *pdata, u8 plen)
 	req = (struct fw_ofld_tx_data_wr *) skb_put(skb, wrlen);
 	memset(req, 0, wrlen);
 	req->op_to_immdlen = cpu_to_be32(
-		FW_WR_OP(FW_OFLD_TX_DATA_WR) |
-		FW_WR_COMPL(1) |
-		FW_WR_IMMDLEN(mpalen));
+		FW_WR_OP_V(FW_OFLD_TX_DATA_WR) |
+		FW_WR_COMPL_F |
+		FW_WR_IMMDLEN_V(mpalen));
 	req->flowid_len16 = cpu_to_be32(
-		FW_WR_FLOWID(ep->hwtid) |
-		FW_WR_LEN16(wrlen >> 4));
+		FW_WR_FLOWID_V(ep->hwtid) |
+		FW_WR_LEN16_V(wrlen >> 4));
 	req->plen = cpu_to_be32(mpalen);
 	req->tunnel_to_proxy = cpu_to_be32(
-		FW_OFLD_TX_DATA_WR_FLUSH(1) |
-		FW_OFLD_TX_DATA_WR_SHOVE(1));
+		FW_OFLD_TX_DATA_WR_FLUSH_F |
+		FW_OFLD_TX_DATA_WR_SHOVE_F);
 
 	mpa = (struct mpa_message *)(req + 1);
 	memset(mpa, 0, sizeof(*mpa));
@@ -1243,15 +1249,15 @@ static int update_rx_credits(struct c4iw_ep *ep, u32 credits)
 	 * due to the limit in the number of bits in the RCV_BUFSIZ field,
 	 * then add the overage in to the credits returned.
 	 */
-	if (ep->rcv_win > RCV_BUFSIZ_MASK * 1024)
-		credits += ep->rcv_win - RCV_BUFSIZ_MASK * 1024;
+	if (ep->rcv_win > RCV_BUFSIZ_M * 1024)
+		credits += ep->rcv_win - RCV_BUFSIZ_M * 1024;
 
 	req = (struct cpl_rx_data_ack *) skb_put(skb, wrlen);
 	memset(req, 0, wrlen);
 	INIT_TP_WR(req, ep->hwtid);
 	OPCODE_TID(req) = cpu_to_be32(MK_OPCODE_TID(CPL_RX_DATA_ACK,
 						    ep->hwtid));
-	req->credit_dack = cpu_to_be32(credits | RX_FORCE_ACK(1) |
+	req->credit_dack = cpu_to_be32(credits | RX_FORCE_ACK_F |
 				       F_RX_DACK_CHANGE |
 				       V_RX_DACK_MODE(dack_mode));
 	set_wr_txq(skb, CPL_PRIORITY_ACK, ep->ctrlq_idx);
@@ -1634,7 +1640,8 @@ static void process_mpa_request(struct c4iw_ep *ep, struct sk_buff *skb)
 		__state_set(&ep->com, MPA_REQ_RCVD);
 
 		/* drive upcall */
-		mutex_lock(&ep->parent_ep->com.mutex);
+		mutex_lock_nested(&ep->parent_ep->com.mutex,
+				  SINGLE_DEPTH_NESTING);
 		if (ep->parent_ep->com.state != DEAD) {
 			if (connect_request_upcall(ep))
 				abort_connection(ep, skb, GFP_KERNEL);
@@ -1745,7 +1752,7 @@ static void send_fw_act_open_req(struct c4iw_ep *ep, unsigned int atid)
 	req = (struct fw_ofld_connection_wr *)__skb_put(skb, sizeof(*req));
 	memset(req, 0, sizeof(*req));
 	req->op_compl = htonl(V_WR_OP(FW_OFLD_CONNECTION_WR));
-	req->len16_pkd = htonl(FW_WR_LEN16(DIV_ROUND_UP(sizeof(*req), 16)));
+	req->len16_pkd = htonl(FW_WR_LEN16_V(DIV_ROUND_UP(sizeof(*req), 16)));
 	req->le.filter = cpu_to_be32(cxgb4_select_ntuple(
 				     ep->com.dev->rdev.lldi.ports[0],
 				     ep->l2t));
@@ -1756,14 +1763,15 @@ static void send_fw_act_open_req(struct c4iw_ep *ep, unsigned int atid)
 	req->le.pport = sin->sin_port;
 	req->le.u.ipv4.pip = sin->sin_addr.s_addr;
 	req->tcb.t_state_to_astid =
-			htonl(V_FW_OFLD_CONNECTION_WR_T_STATE(TCP_SYN_SENT) |
-			V_FW_OFLD_CONNECTION_WR_ASTID(atid));
+			htonl(FW_OFLD_CONNECTION_WR_T_STATE_V(TCP_SYN_SENT) |
+			FW_OFLD_CONNECTION_WR_ASTID_V(atid));
 	req->tcb.cplrxdataack_cplpassacceptrpl =
-			htons(F_FW_OFLD_CONNECTION_WR_CPLRXDATAACK);
+			htons(FW_OFLD_CONNECTION_WR_CPLRXDATAACK_F);
 	req->tcb.tx_max = (__force __be32) jiffies;
 	req->tcb.rcv_adv = htons(1);
 	best_mtu(ep->com.dev->rdev.lldi.mtus, ep->mtu, &mtu_idx,
-		 enable_tcp_timestamps);
+		 enable_tcp_timestamps,
+		 (AF_INET == ep->com.remote_addr.ss_family) ? 0 : 1);
 	wscale = compute_wscale(rcv_win);
 
 	/*
@@ -1771,34 +1779,34 @@ static void send_fw_act_open_req(struct c4iw_ep *ep, unsigned int atid)
 	 * remainder will be specified in the rx_data_ack.
 	 */
 	win = ep->rcv_win >> 10;
-	if (win > RCV_BUFSIZ_MASK)
-		win = RCV_BUFSIZ_MASK;
+	if (win > RCV_BUFSIZ_M)
+		win = RCV_BUFSIZ_M;
 
 	req->tcb.opt0 = (__force __be64) (TCAM_BYPASS(1) |
 		(nocong ? NO_CONG(1) : 0) |
-		KEEP_ALIVE(1) |
+		KEEP_ALIVE_F |
 		DELACK(1) |
-		WND_SCALE(wscale) |
-		MSS_IDX(mtu_idx) |
-		L2T_IDX(ep->l2t->idx) |
-		TX_CHAN(ep->tx_chan) |
-		SMAC_SEL(ep->smac_idx) |
+		WND_SCALE_V(wscale) |
+		MSS_IDX_V(mtu_idx) |
+		L2T_IDX_V(ep->l2t->idx) |
+		TX_CHAN_V(ep->tx_chan) |
+		SMAC_SEL_V(ep->smac_idx) |
 		DSCP(ep->tos) |
-		ULP_MODE(ULP_MODE_TCPDDP) |
-		RCV_BUFSIZ(win));
+		ULP_MODE_V(ULP_MODE_TCPDDP) |
+		RCV_BUFSIZ_V(win));
 	req->tcb.opt2 = (__force __be32) (PACE(1) |
 		TX_QUEUE(ep->com.dev->rdev.lldi.tx_modq[ep->tx_chan]) |
-		RX_CHANNEL(0) |
+		RX_CHANNEL_V(0) |
 		CCTRL_ECN(enable_ecn) |
-		RSS_QUEUE_VALID | RSS_QUEUE(ep->rss_qid));
+		RSS_QUEUE_VALID_F | RSS_QUEUE_V(ep->rss_qid));
 	if (enable_tcp_timestamps)
-		req->tcb.opt2 |= (__force __be32) TSTAMPS_EN(1);
+		req->tcb.opt2 |= (__force __be32)TSTAMPS_EN(1);
 	if (enable_tcp_sack)
-		req->tcb.opt2 |= (__force __be32) SACK_EN(1);
+		req->tcb.opt2 |= (__force __be32)SACK_EN(1);
 	if (wscale && enable_tcp_window_scaling)
-		req->tcb.opt2 |= (__force __be32) WND_SCALE_EN(1);
-	req->tcb.opt0 = cpu_to_be64((__force u64) req->tcb.opt0);
-	req->tcb.opt2 = cpu_to_be32((__force u32) req->tcb.opt2);
+		req->tcb.opt2 |= (__force __be32)WND_SCALE_EN_F;
+	req->tcb.opt0 = cpu_to_be64((__force u64)req->tcb.opt0);
+	req->tcb.opt2 = cpu_to_be32((__force u32)req->tcb.opt2);
 	set_wr_txq(skb, CPL_PRIORITY_CONTROL, ep->ctrlq_idx);
 	set_bit(ACT_OFLD_CONN, &ep->com.history);
 	c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t);
@@ -2162,7 +2170,8 @@ static void accept_cr(struct c4iw_ep *ep, struct sk_buff *skb,
 						    ep->hwtid));
 
 	best_mtu(ep->com.dev->rdev.lldi.mtus, ep->mtu, &mtu_idx,
-		 enable_tcp_timestamps && req->tcpopt.tstamp);
+		 enable_tcp_timestamps && req->tcpopt.tstamp,
+		 (AF_INET == ep->com.remote_addr.ss_family) ? 0 : 1);
 	wscale = compute_wscale(rcv_win);
 
 	/*
@@ -2170,28 +2179,28 @@ static void accept_cr(struct c4iw_ep *ep, struct sk_buff *skb,
 	 * remainder will be specified in the rx_data_ack.
 	 */
 	win = ep->rcv_win >> 10;
-	if (win > RCV_BUFSIZ_MASK)
-		win = RCV_BUFSIZ_MASK;
+	if (win > RCV_BUFSIZ_M)
+		win = RCV_BUFSIZ_M;
 	opt0 = (nocong ? NO_CONG(1) : 0) |
-	       KEEP_ALIVE(1) |
+	       KEEP_ALIVE_F |
 	       DELACK(1) |
-	       WND_SCALE(wscale) |
-	       MSS_IDX(mtu_idx) |
-	       L2T_IDX(ep->l2t->idx) |
-	       TX_CHAN(ep->tx_chan) |
-	       SMAC_SEL(ep->smac_idx) |
+	       WND_SCALE_V(wscale) |
+	       MSS_IDX_V(mtu_idx) |
+	       L2T_IDX_V(ep->l2t->idx) |
+	       TX_CHAN_V(ep->tx_chan) |
+	       SMAC_SEL_V(ep->smac_idx) |
 	       DSCP(ep->tos >> 2) |
-	       ULP_MODE(ULP_MODE_TCPDDP) |
-	       RCV_BUFSIZ(win);
-	opt2 = RX_CHANNEL(0) |
-	       RSS_QUEUE_VALID | RSS_QUEUE(ep->rss_qid);
+	       ULP_MODE_V(ULP_MODE_TCPDDP) |
+	       RCV_BUFSIZ_V(win);
+	opt2 = RX_CHANNEL_V(0) |
+	       RSS_QUEUE_VALID_F | RSS_QUEUE_V(ep->rss_qid);
 
 	if (enable_tcp_timestamps && req->tcpopt.tstamp)
 		opt2 |= TSTAMPS_EN(1);
 	if (enable_tcp_sack && req->tcpopt.sack)
 		opt2 |= SACK_EN(1);
 	if (wscale && enable_tcp_window_scaling)
-		opt2 |= WND_SCALE_EN(1);
+		opt2 |= WND_SCALE_EN_F;
 	if (enable_ecn) {
 		const struct tcphdr *tcph;
 		u32 hlen = ntohl(req->hdr_len);
@@ -2203,7 +2212,7 @@ static void accept_cr(struct c4iw_ep *ep, struct sk_buff *skb,
 	}
 	if (is_t5(ep->com.dev->rdev.lldi.adapter_type)) {
 		u32 isn = (prandom_u32() & ~7UL) - 1;
-		opt2 |= T5_OPT_2_VALID;
+		opt2 |= T5_OPT_2_VALID_F;
 		opt2 |= V_CONG_CNTRL(CONG_ALG_TAHOE);
 		opt2 |= CONG_CNTRL_VALID; /* OPT_2_ISS for T5 */
 		rpl5 = (void *)rpl;
@@ -3118,6 +3127,8 @@ static int create_server6(struct c4iw_dev *dev, struct c4iw_listen_ep *ep)
 		err = c4iw_wait_for_reply(&ep->com.dev->rdev,
 					  &ep->com.wr_wait,
 					  0, 0, __func__);
+	else if (err > 0)
+		err = net_xmit_errno(err);
 	if (err)
 		pr_err("cxgb4_create_server6/filter failed err %d stid %d laddr %pI6 lport %d\n",
 		       err, ep->stid,
@@ -3151,6 +3162,8 @@ static int create_server4(struct c4iw_dev *dev, struct c4iw_listen_ep *ep)
 			err = c4iw_wait_for_reply(&ep->com.dev->rdev,
 						  &ep->com.wr_wait,
 						  0, 0, __func__);
+		else if (err > 0)
+			err = net_xmit_errno(err);
 	}
 	if (err)
 		pr_err("cxgb4_create_server/filter failed err %d stid %d laddr %pI4 lport %d\n"
@@ -3529,9 +3542,9 @@ static void send_fw_pass_open_req(struct c4iw_dev *dev, struct sk_buff *skb,
 	req_skb = alloc_skb(sizeof(struct fw_ofld_connection_wr), GFP_KERNEL);
 	req = (struct fw_ofld_connection_wr *)__skb_put(req_skb, sizeof(*req));
 	memset(req, 0, sizeof(*req));
-	req->op_compl = htonl(V_WR_OP(FW_OFLD_CONNECTION_WR) | FW_WR_COMPL(1));
-	req->len16_pkd = htonl(FW_WR_LEN16(DIV_ROUND_UP(sizeof(*req), 16)));
-	req->le.version_cpl = htonl(F_FW_OFLD_CONNECTION_WR_CPL);
+	req->op_compl = htonl(V_WR_OP(FW_OFLD_CONNECTION_WR) | FW_WR_COMPL_F);
+	req->len16_pkd = htonl(FW_WR_LEN16_V(DIV_ROUND_UP(sizeof(*req), 16)));
+	req->le.version_cpl = htonl(FW_OFLD_CONNECTION_WR_CPL_F);
 	req->le.filter = (__force __be32) filter;
 	req->le.lport = lport;
 	req->le.pport = rport;
@@ -3540,16 +3553,16 @@ static void send_fw_pass_open_req(struct c4iw_dev *dev, struct sk_buff *skb,
 	req->tcb.rcv_nxt = htonl(rcv_isn + 1);
 	req->tcb.rcv_adv = htons(window);
 	req->tcb.t_state_to_astid =
-		 htonl(V_FW_OFLD_CONNECTION_WR_T_STATE(TCP_SYN_RECV) |
-			V_FW_OFLD_CONNECTION_WR_RCV_SCALE(cpl->tcpopt.wsf) |
-			V_FW_OFLD_CONNECTION_WR_ASTID(
+		 htonl(FW_OFLD_CONNECTION_WR_T_STATE_V(TCP_SYN_RECV) |
+			FW_OFLD_CONNECTION_WR_RCV_SCALE_V(cpl->tcpopt.wsf) |
+			FW_OFLD_CONNECTION_WR_ASTID_V(
 			GET_PASS_OPEN_TID(ntohl(cpl->tos_stid))));
 
 	/*
 	 * We store the qid in opt2 which will be used by the firmware
 	 * to send us the wr response.
 	 */
-	req->tcb.opt2 = htonl(V_RSS_QUEUE(rss_qid));
+	req->tcb.opt2 = htonl(RSS_QUEUE_V(rss_qid));
 
 	/*
 	 * We initialize the MSS index in TCB to 0xF.
@@ -3557,7 +3570,7 @@ static void send_fw_pass_open_req(struct c4iw_dev *dev, struct sk_buff *skb,
 	 * TCB picks up the correct value. If this was 0
 	 * TP will ignore any value > 0 for MSS index.
 	 */
-	req->tcb.opt0 = cpu_to_be64(V_MSS_IDX(0xF));
+	req->tcb.opt0 = cpu_to_be64(MSS_IDX_V(0xF));
 	req->cookie = (unsigned long)skb;
 
 	set_wr_txq(req_skb, CPL_PRIORITY_CONTROL, port_id);
diff --git a/drivers/infiniband/hw/cxgb4/cq.c b/drivers/infiniband/hw/cxgb4/cq.c
index 0f773e78e080..e9fd3a029296 100644
--- a/drivers/infiniband/hw/cxgb4/cq.c
+++ b/drivers/infiniband/hw/cxgb4/cq.c
@@ -51,9 +51,9 @@ static int destroy_cq(struct c4iw_rdev *rdev, struct t4_cq *cq,
 	res_wr = (struct fw_ri_res_wr *)__skb_put(skb, wr_len);
 	memset(res_wr, 0, wr_len);
 	res_wr->op_nres = cpu_to_be32(
-			FW_WR_OP(FW_RI_RES_WR) |
+			FW_WR_OP_V(FW_RI_RES_WR) |
 			V_FW_RI_RES_WR_NRES(1) |
-			FW_WR_COMPL(1));
+			FW_WR_COMPL_F);
 	res_wr->len16_pkd = cpu_to_be32(DIV_ROUND_UP(wr_len, 16));
 	res_wr->cookie = (unsigned long) &wr_wait;
 	res = res_wr->res;
@@ -121,9 +121,9 @@ static int create_cq(struct c4iw_rdev *rdev, struct t4_cq *cq,
 	res_wr = (struct fw_ri_res_wr *)__skb_put(skb, wr_len);
 	memset(res_wr, 0, wr_len);
 	res_wr->op_nres = cpu_to_be32(
-			FW_WR_OP(FW_RI_RES_WR) |
+			FW_WR_OP_V(FW_RI_RES_WR) |
 			V_FW_RI_RES_WR_NRES(1) |
-			FW_WR_COMPL(1));
+			FW_WR_COMPL_F);
 	res_wr->len16_pkd = cpu_to_be32(DIV_ROUND_UP(wr_len, 16));
 	res_wr->cookie = (unsigned long) &wr_wait;
 	res = res_wr->res;
diff --git a/drivers/infiniband/hw/cxgb4/device.c b/drivers/infiniband/hw/cxgb4/device.c
index f25df5276c22..eb5df4e62703 100644
--- a/drivers/infiniband/hw/cxgb4/device.c
+++ b/drivers/infiniband/hw/cxgb4/device.c
@@ -60,7 +60,7 @@ int c4iw_wr_log = 0;
 module_param(c4iw_wr_log, int, 0444);
 MODULE_PARM_DESC(c4iw_wr_log, "Enables logging of work request timing data.");
 
-int c4iw_wr_log_size_order = 12;
+static int c4iw_wr_log_size_order = 12;
 module_param(c4iw_wr_log_size_order, int, 0444);
 MODULE_PARM_DESC(c4iw_wr_log_size_order,
 		 "Number of entries (log2) in the work request timing log.");
@@ -670,7 +670,7 @@ static int ep_open(struct inode *inode, struct file *file)
 	idr_for_each(&epd->devp->stid_idr, count_idrs, &count);
 	spin_unlock_irq(&epd->devp->lock);
 
-	epd->bufsize = count * 160;
+	epd->bufsize = count * 240;
 	epd->buf = vmalloc(epd->bufsize);
 	if (!epd->buf) {
 		ret = -ENOMEM;
diff --git a/drivers/infiniband/hw/cxgb4/mem.c b/drivers/infiniband/hw/cxgb4/mem.c
index ec7a2988a703..cb43c2299ac0 100644
--- a/drivers/infiniband/hw/cxgb4/mem.c
+++ b/drivers/infiniband/hw/cxgb4/mem.c
@@ -50,6 +50,13 @@ static int inline_threshold = C4IW_INLINE_THRESHOLD;
 module_param(inline_threshold, int, 0644);
 MODULE_PARM_DESC(inline_threshold, "inline vs dsgl threshold (default=128)");
 
+static int mr_exceeds_hw_limits(struct c4iw_dev *dev, u64 length)
+{
+	return (is_t4(dev->rdev.lldi.adapter_type) ||
+		is_t5(dev->rdev.lldi.adapter_type)) &&
+		length >= 8*1024*1024*1024ULL;
+}
+
 static int _c4iw_write_mem_dma_aligned(struct c4iw_rdev *rdev, u32 addr,
 				       u32 len, dma_addr_t data, int wait)
 {
@@ -74,18 +81,18 @@ static int _c4iw_write_mem_dma_aligned(struct c4iw_rdev *rdev, u32 addr,
 	req = (struct ulp_mem_io *)__skb_put(skb, wr_len);
 	memset(req, 0, wr_len);
 	INIT_ULPTX_WR(req, wr_len, 0, 0);
-	req->wr.wr_hi = cpu_to_be32(FW_WR_OP(FW_ULPTX_WR) |
-			(wait ? FW_WR_COMPL(1) : 0));
+	req->wr.wr_hi = cpu_to_be32(FW_WR_OP_V(FW_ULPTX_WR) |
+			(wait ? FW_WR_COMPL_F : 0));
 	req->wr.wr_lo = wait ? (__force __be64)(unsigned long) &wr_wait : 0L;
-	req->wr.wr_mid = cpu_to_be32(FW_WR_LEN16(DIV_ROUND_UP(wr_len, 16)));
-	req->cmd = cpu_to_be32(ULPTX_CMD(ULP_TX_MEM_WRITE));
+	req->wr.wr_mid = cpu_to_be32(FW_WR_LEN16_V(DIV_ROUND_UP(wr_len, 16)));
+	req->cmd = cpu_to_be32(ULPTX_CMD_V(ULP_TX_MEM_WRITE));
 	req->cmd |= cpu_to_be32(V_T5_ULP_MEMIO_ORDER(1));
-	req->dlen = cpu_to_be32(ULP_MEMIO_DATA_LEN(len>>5));
+	req->dlen = cpu_to_be32(ULP_MEMIO_DATA_LEN_V(len>>5));
 	req->len16 = cpu_to_be32(DIV_ROUND_UP(wr_len-sizeof(req->wr), 16));
-	req->lock_addr = cpu_to_be32(ULP_MEMIO_ADDR(addr));
+	req->lock_addr = cpu_to_be32(ULP_MEMIO_ADDR_V(addr));
 
 	sgl = (struct ulptx_sgl *)(req + 1);
-	sgl->cmd_nsge = cpu_to_be32(ULPTX_CMD(ULP_TX_SC_DSGL) |
+	sgl->cmd_nsge = cpu_to_be32(ULPTX_CMD_V(ULP_TX_SC_DSGL) |
 				    ULPTX_NSGE(1));
 	sgl->len0 = cpu_to_be32(len);
 	sgl->addr0 = cpu_to_be64(data);
@@ -107,12 +114,12 @@ static int _c4iw_write_mem_inline(struct c4iw_rdev *rdev, u32 addr, u32 len,
 	u8 wr_len, *to_dp, *from_dp;
 	int copy_len, num_wqe, i, ret = 0;
 	struct c4iw_wr_wait wr_wait;
-	__be32 cmd = cpu_to_be32(ULPTX_CMD(ULP_TX_MEM_WRITE));
+	__be32 cmd = cpu_to_be32(ULPTX_CMD_V(ULP_TX_MEM_WRITE));
 
 	if (is_t4(rdev->lldi.adapter_type))
-		cmd |= cpu_to_be32(ULP_MEMIO_ORDER(1));
+		cmd |= cpu_to_be32(ULP_MEMIO_ORDER_F);
 	else
-		cmd |= cpu_to_be32(V_T5_ULP_MEMIO_IMM(1));
+		cmd |= cpu_to_be32(T5_ULP_MEMIO_IMM_F);
 
 	addr &= 0x7FFFFFF;
 	PDBG("%s addr 0x%x len %u\n", __func__, addr, len);
@@ -135,23 +142,23 @@ static int _c4iw_write_mem_inline(struct c4iw_rdev *rdev, u32 addr, u32 len,
 		INIT_ULPTX_WR(req, wr_len, 0, 0);
 
 		if (i == (num_wqe-1)) {
-			req->wr.wr_hi = cpu_to_be32(FW_WR_OP(FW_ULPTX_WR) |
-						    FW_WR_COMPL(1));
+			req->wr.wr_hi = cpu_to_be32(FW_WR_OP_V(FW_ULPTX_WR) |
+						    FW_WR_COMPL_F);
 			req->wr.wr_lo = (__force __be64)(unsigned long) &wr_wait;
 		} else
-			req->wr.wr_hi = cpu_to_be32(FW_WR_OP(FW_ULPTX_WR));
+			req->wr.wr_hi = cpu_to_be32(FW_WR_OP_V(FW_ULPTX_WR));
 		req->wr.wr_mid = cpu_to_be32(
-				       FW_WR_LEN16(DIV_ROUND_UP(wr_len, 16)));
+				       FW_WR_LEN16_V(DIV_ROUND_UP(wr_len, 16)));
 
 		req->cmd = cmd;
-		req->dlen = cpu_to_be32(ULP_MEMIO_DATA_LEN(
+		req->dlen = cpu_to_be32(ULP_MEMIO_DATA_LEN_V(
 				DIV_ROUND_UP(copy_len, T4_ULPTX_MIN_IO)));
 		req->len16 = cpu_to_be32(DIV_ROUND_UP(wr_len-sizeof(req->wr),
 						      16));
-		req->lock_addr = cpu_to_be32(ULP_MEMIO_ADDR(addr + i * 3));
+		req->lock_addr = cpu_to_be32(ULP_MEMIO_ADDR_V(addr + i * 3));
 
 		sc = (struct ulptx_idata *)(req + 1);
-		sc->cmd_more = cpu_to_be32(ULPTX_CMD(ULP_TX_SC_IMM));
+		sc->cmd_more = cpu_to_be32(ULPTX_CMD_V(ULP_TX_SC_IMM));
 		sc->len = cpu_to_be32(roundup(copy_len, T4_ULPTX_MIN_IO));
 
 		to_dp = (u8 *)(sc + 1);
@@ -369,9 +376,11 @@ static int register_mem(struct c4iw_dev *rhp, struct c4iw_pd *php,
 	int ret;
 
 	ret = write_tpt_entry(&rhp->rdev, 0, &stag, 1, mhp->attr.pdid,
-			      FW_RI_STAG_NSMR, mhp->attr.perms,
+			      FW_RI_STAG_NSMR, mhp->attr.len ?
+			      mhp->attr.perms : 0,
 			      mhp->attr.mw_bind_enable, mhp->attr.zbva,
-			      mhp->attr.va_fbo, mhp->attr.len, shift - 12,
+			      mhp->attr.va_fbo, mhp->attr.len ?
+			      mhp->attr.len : -1, shift - 12,
 			      mhp->attr.pbl_size, mhp->attr.pbl_addr);
 	if (ret)
 		return ret;
@@ -536,6 +545,11 @@ int c4iw_reregister_phys_mem(struct ib_mr *mr, int mr_rereg_mask,
 			return ret;
 	}
 
+	if (mr_exceeds_hw_limits(rhp, total_size)) {
+		kfree(page_list);
+		return -EINVAL;
+	}
+
 	ret = reregister_mem(rhp, php, &mh, shift, npages);
 	kfree(page_list);
 	if (ret)
@@ -596,6 +610,12 @@ struct ib_mr *c4iw_register_phys_mem(struct ib_pd *pd,
 	if (ret)
 		goto err;
 
+	if (mr_exceeds_hw_limits(rhp, total_size)) {
+		kfree(page_list);
+		ret = -EINVAL;
+		goto err;
+	}
+
 	ret = alloc_pbl(mhp, npages);
 	if (ret) {
 		kfree(page_list);
@@ -699,6 +719,10 @@ struct ib_mr *c4iw_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 
 	php = to_c4iw_pd(pd);
 	rhp = php->rhp;
+
+	if (mr_exceeds_hw_limits(rhp, length))
+		return ERR_PTR(-EINVAL);
+
 	mhp = kzalloc(sizeof(*mhp), GFP_KERNEL);
 	if (!mhp)
 		return ERR_PTR(-ENOMEM);
diff --git a/drivers/infiniband/hw/cxgb4/provider.c b/drivers/infiniband/hw/cxgb4/provider.c
index 72e3b69d1b76..66bd6a2ad83b 100644
--- a/drivers/infiniband/hw/cxgb4/provider.c
+++ b/drivers/infiniband/hw/cxgb4/provider.c
@@ -408,10 +408,10 @@ static ssize_t show_fw_ver(struct device *dev, struct device_attribute *attr,
 	PDBG("%s dev 0x%p\n", __func__, dev);
 
 	return sprintf(buf, "%u.%u.%u.%u\n",
-			FW_HDR_FW_VER_MAJOR_GET(c4iw_dev->rdev.lldi.fw_vers),
-			FW_HDR_FW_VER_MINOR_GET(c4iw_dev->rdev.lldi.fw_vers),
-			FW_HDR_FW_VER_MICRO_GET(c4iw_dev->rdev.lldi.fw_vers),
-			FW_HDR_FW_VER_BUILD_GET(c4iw_dev->rdev.lldi.fw_vers));
+			FW_HDR_FW_VER_MAJOR_G(c4iw_dev->rdev.lldi.fw_vers),
+			FW_HDR_FW_VER_MINOR_G(c4iw_dev->rdev.lldi.fw_vers),
+			FW_HDR_FW_VER_MICRO_G(c4iw_dev->rdev.lldi.fw_vers),
+			FW_HDR_FW_VER_BUILD_G(c4iw_dev->rdev.lldi.fw_vers));
 }
 
 static ssize_t show_hca(struct device *dev, struct device_attribute *attr,
diff --git a/drivers/infiniband/hw/cxgb4/qp.c b/drivers/infiniband/hw/cxgb4/qp.c
index 41cd6882b648..bb85d479e66e 100644
--- a/drivers/infiniband/hw/cxgb4/qp.c
+++ b/drivers/infiniband/hw/cxgb4/qp.c
@@ -271,9 +271,9 @@ static int create_qp(struct c4iw_rdev *rdev, struct t4_wq *wq,
 	res_wr = (struct fw_ri_res_wr *)__skb_put(skb, wr_len);
 	memset(res_wr, 0, wr_len);
 	res_wr->op_nres = cpu_to_be32(
-			FW_WR_OP(FW_RI_RES_WR) |
+			FW_WR_OP_V(FW_RI_RES_WR) |
 			V_FW_RI_RES_WR_NRES(2) |
-			FW_WR_COMPL(1));
+			FW_WR_COMPL_F);
 	res_wr->len16_pkd = cpu_to_be32(DIV_ROUND_UP(wr_len, 16));
 	res_wr->cookie = (unsigned long) &wr_wait;
 	res = res_wr->res;
@@ -1082,10 +1082,10 @@ static void post_terminate(struct c4iw_qp *qhp, struct t4_cqe *err_cqe,
 
 	wqe = (struct fw_ri_wr *)__skb_put(skb, sizeof(*wqe));
 	memset(wqe, 0, sizeof *wqe);
-	wqe->op_compl = cpu_to_be32(FW_WR_OP(FW_RI_INIT_WR));
+	wqe->op_compl = cpu_to_be32(FW_WR_OP_V(FW_RI_INIT_WR));
 	wqe->flowid_len16 = cpu_to_be32(
-		FW_WR_FLOWID(qhp->ep->hwtid) |
-		FW_WR_LEN16(DIV_ROUND_UP(sizeof *wqe, 16)));
+		FW_WR_FLOWID_V(qhp->ep->hwtid) |
+		FW_WR_LEN16_V(DIV_ROUND_UP(sizeof(*wqe), 16)));
 
 	wqe->u.terminate.type = FW_RI_TYPE_TERMINATE;
 	wqe->u.terminate.immdlen = cpu_to_be32(sizeof *term);
@@ -1204,11 +1204,11 @@ static int rdma_fini(struct c4iw_dev *rhp, struct c4iw_qp *qhp,
 	wqe = (struct fw_ri_wr *)__skb_put(skb, sizeof(*wqe));
 	memset(wqe, 0, sizeof *wqe);
 	wqe->op_compl = cpu_to_be32(
-		FW_WR_OP(FW_RI_INIT_WR) |
-		FW_WR_COMPL(1));
+		FW_WR_OP_V(FW_RI_INIT_WR) |
+		FW_WR_COMPL_F);
 	wqe->flowid_len16 = cpu_to_be32(
-		FW_WR_FLOWID(ep->hwtid) |
-		FW_WR_LEN16(DIV_ROUND_UP(sizeof *wqe, 16)));
+		FW_WR_FLOWID_V(ep->hwtid) |
+		FW_WR_LEN16_V(DIV_ROUND_UP(sizeof(*wqe), 16)));
 	wqe->cookie = (unsigned long) &ep->com.wr_wait;
 
 	wqe->u.fini.type = FW_RI_TYPE_FINI;
@@ -1273,11 +1273,11 @@ static int rdma_init(struct c4iw_dev *rhp, struct c4iw_qp *qhp)
 	wqe = (struct fw_ri_wr *)__skb_put(skb, sizeof(*wqe));
 	memset(wqe, 0, sizeof *wqe);
 	wqe->op_compl = cpu_to_be32(
-		FW_WR_OP(FW_RI_INIT_WR) |
-		FW_WR_COMPL(1));
+		FW_WR_OP_V(FW_RI_INIT_WR) |
+		FW_WR_COMPL_F);
 	wqe->flowid_len16 = cpu_to_be32(
-		FW_WR_FLOWID(qhp->ep->hwtid) |
-		FW_WR_LEN16(DIV_ROUND_UP(sizeof *wqe, 16)));
+		FW_WR_FLOWID_V(qhp->ep->hwtid) |
+		FW_WR_LEN16_V(DIV_ROUND_UP(sizeof(*wqe), 16)));
 
 	wqe->cookie = (unsigned long) &qhp->ep->com.wr_wait;
 
@@ -1538,9 +1538,9 @@ err:
 	set_state(qhp, C4IW_QP_STATE_ERROR);
 	free = 1;
 	abort = 1;
-	wake_up(&qhp->wait);
 	BUG_ON(!ep);
 	flush_qp(qhp);
+	wake_up(&qhp->wait);
 out:
 	mutex_unlock(&qhp->mutex);
 
diff --git a/drivers/infiniband/hw/ehca/ehca_mrmw.c b/drivers/infiniband/hw/ehca/ehca_mrmw.c
index 3488e8c9fcb4..f914b30999f8 100644
--- a/drivers/infiniband/hw/ehca/ehca_mrmw.c
+++ b/drivers/infiniband/hw/ehca/ehca_mrmw.c
@@ -399,7 +399,7 @@ reg_user_mr_fallback:
 	pginfo.num_kpages = num_kpages;
 	pginfo.num_hwpages = num_hwpages;
 	pginfo.u.usr.region = e_mr->umem;
-	pginfo.next_hwpage = e_mr->umem->offset / hwpage_size;
+	pginfo.next_hwpage = ib_umem_offset(e_mr->umem) / hwpage_size;
 	pginfo.u.usr.next_sg = pginfo.u.usr.region->sg_head.sgl;
 	ret = ehca_reg_mr(shca, e_mr, (u64 *)virt, length, mr_access_flags,
 			  e_pd, &pginfo, &e_mr->ib.ib_mr.lkey,
diff --git a/drivers/infiniband/hw/ipath/ipath_fs.c b/drivers/infiniband/hw/ipath/ipath_fs.c
index e0c404bdc4a8..4977082e081f 100644
--- a/drivers/infiniband/hw/ipath/ipath_fs.c
+++ b/drivers/infiniband/hw/ipath/ipath_fs.c
@@ -82,7 +82,6 @@ static int create_file(const char *name, umode_t mode,
 {
 	int error;
 
-	*dentry = NULL;
 	mutex_lock(&parent->d_inode->i_mutex);
 	*dentry = lookup_one_len(name, parent, strlen(name));
 	if (!IS_ERR(*dentry))
diff --git a/drivers/infiniband/hw/ipath/ipath_mr.c b/drivers/infiniband/hw/ipath/ipath_mr.c
index 5e61e9bff697..c7278f6a8217 100644
--- a/drivers/infiniband/hw/ipath/ipath_mr.c
+++ b/drivers/infiniband/hw/ipath/ipath_mr.c
@@ -214,7 +214,7 @@ struct ib_mr *ipath_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 	mr->mr.user_base = start;
 	mr->mr.iova = virt_addr;
 	mr->mr.length = length;
-	mr->mr.offset = umem->offset;
+	mr->mr.offset = ib_umem_offset(umem);
 	mr->mr.access_flags = mr_access_flags;
 	mr->mr.max_segs = n;
 	mr->umem = umem;
diff --git a/drivers/infiniband/hw/mlx4/cq.c b/drivers/infiniband/hw/mlx4/cq.c
index 1066eec854a9..a3b70f6c4035 100644
--- a/drivers/infiniband/hw/mlx4/cq.c
+++ b/drivers/infiniband/hw/mlx4/cq.c
@@ -233,7 +233,10 @@ struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, int entries, int vector
 	if (err)
 		goto err_dbmap;
 
-	cq->mcq.comp  = mlx4_ib_cq_comp;
+	if (context)
+		cq->mcq.tasklet_ctx.comp = mlx4_ib_cq_comp;
+	else
+		cq->mcq.comp = mlx4_ib_cq_comp;
 	cq->mcq.event = mlx4_ib_cq_event;
 
 	if (context)
diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index bda5994ceb68..9117b7a2d5f8 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -1114,7 +1114,8 @@ static int mlx4_ib_tunnel_steer_add(struct ib_qp *qp, struct ib_flow_attr *flow_
 	struct mlx4_dev	*dev = to_mdev(qp->device)->dev;
 	int err = 0;
 
-	if (dev->caps.tunnel_offload_mode != MLX4_TUNNEL_OFFLOAD_MODE_VXLAN)
+	if (dev->caps.tunnel_offload_mode != MLX4_TUNNEL_OFFLOAD_MODE_VXLAN ||
+	    dev->caps.dmfs_high_steer_mode == MLX4_STEERING_DMFS_A0_STATIC)
 		return 0; /* do nothing */
 
 	ib_flow = flow_attr + 1;
@@ -1173,18 +1174,24 @@ static struct ib_flow *mlx4_ib_create_flow(struct ib_qp *qp,
 		err = __mlx4_ib_create_flow(qp, flow_attr, domain, type[i],
 					    &mflow->reg_id[i]);
 		if (err)
-			goto err_free;
+			goto err_create_flow;
 		i++;
 	}
 
 	if (i < ARRAY_SIZE(type) && flow_attr->type == IB_FLOW_ATTR_NORMAL) {
 		err = mlx4_ib_tunnel_steer_add(qp, flow_attr, &mflow->reg_id[i]);
 		if (err)
-			goto err_free;
+			goto err_create_flow;
+		i++;
 	}
 
 	return &mflow->ibflow;
 
+err_create_flow:
+	while (i) {
+		(void)__mlx4_ib_destroy_flow(to_mdev(qp->device)->dev, mflow->reg_id[i]);
+		i--;
+	}
 err_free:
 	kfree(mflow);
 	return ERR_PTR(err);
@@ -1969,8 +1976,7 @@ static void mlx4_ib_alloc_eqs(struct mlx4_dev *dev, struct mlx4_ib_dev *ibdev)
 	    dev->caps.num_ports > dev->caps.comp_pool)
 		return;
 
-	eq_per_port = rounddown_pow_of_two(dev->caps.comp_pool/
-					dev->caps.num_ports);
+	eq_per_port = dev->caps.comp_pool / dev->caps.num_ports;
 
 	/* Init eq table */
 	added_eqs = 0;
@@ -2222,7 +2228,7 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
 		ibdev->steer_qpn_count = MLX4_IB_UC_MAX_NUM_QPS;
 		err = mlx4_qp_reserve_range(dev, ibdev->steer_qpn_count,
 					    MLX4_IB_UC_STEER_QPN_ALIGN,
-					    &ibdev->steer_qpn_base);
+					    &ibdev->steer_qpn_base, 0);
 		if (err)
 			goto err_counter;
 
diff --git a/drivers/infiniband/hw/mlx4/mr.c b/drivers/infiniband/hw/mlx4/mr.c
index 8f9325cfc85d..c36ccbd9a644 100644
--- a/drivers/infiniband/hw/mlx4/mr.c
+++ b/drivers/infiniband/hw/mlx4/mr.c
@@ -223,7 +223,6 @@ int mlx4_ib_rereg_user_mr(struct ib_mr *mr, int flags,
 
 	if (flags & IB_MR_REREG_TRANS) {
 		int shift;
-		int err;
 		int n;
 
 		mlx4_mr_rereg_mem_cleanup(dev->dev, &mmr->mmr);
diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index 9c5150c3cb31..cf000b7ad64f 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -802,16 +802,21 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd,
 			}
 		}
 	} else {
-		/* Raw packet QPNs must be aligned to 8 bits. If not, the WQE
-		 * BlueFlame setup flow wrongly causes VLAN insertion. */
+		/* Raw packet QPNs may not have bits 6,7 set in their qp_num;
+		 * otherwise, the WQE BlueFlame setup flow wrongly causes
+		 * VLAN insertion. */
 		if (init_attr->qp_type == IB_QPT_RAW_PACKET)
-			err = mlx4_qp_reserve_range(dev->dev, 1, 1 << 8, &qpn);
+			err = mlx4_qp_reserve_range(dev->dev, 1, 1, &qpn,
+						    (init_attr->cap.max_send_wr ?
+						     MLX4_RESERVE_ETH_BF_QP : 0) |
+						    (init_attr->cap.max_recv_wr ?
+						     MLX4_RESERVE_A0_QP : 0));
 		else
 			if (qp->flags & MLX4_IB_QP_NETIF)
 				err = mlx4_ib_steer_qp_alloc(dev, 1, &qpn);
 			else
 				err = mlx4_qp_reserve_range(dev->dev, 1, 1,
-							    &qpn);
+							    &qpn, 0);
 		if (err)
 			goto err_proxy;
 	}
diff --git a/drivers/infiniband/hw/mlx5/Makefile b/drivers/infiniband/hw/mlx5/Makefile
index 4ea0135af484..27a70159e2ea 100644
--- a/drivers/infiniband/hw/mlx5/Makefile
+++ b/drivers/infiniband/hw/mlx5/Makefile
@@ -1,3 +1,4 @@
 obj-$(CONFIG_MLX5_INFINIBAND)	+= mlx5_ib.o
 
 mlx5_ib-y :=	main.o cq.o doorbell.o qp.o mem.o srq.o mr.o ah.o mad.o
+mlx5_ib-$(CONFIG_INFINIBAND_ON_DEMAND_PAGING) += odp.o
diff --git a/drivers/infiniband/hw/mlx5/cq.c b/drivers/infiniband/hw/mlx5/cq.c
index e4056279166d..c463e7bba5f4 100644
--- a/drivers/infiniband/hw/mlx5/cq.c
+++ b/drivers/infiniband/hw/mlx5/cq.c
@@ -752,7 +752,7 @@ struct ib_cq *mlx5_ib_create_cq(struct ib_device *ibdev, int entries,
 		return ERR_PTR(-EINVAL);
 
 	entries = roundup_pow_of_two(entries + 1);
-	if (entries > dev->mdev->caps.max_cqes)
+	if (entries > dev->mdev->caps.gen.max_cqes)
 		return ERR_PTR(-EINVAL);
 
 	cq = kzalloc(sizeof(*cq), GFP_KERNEL);
@@ -805,14 +805,14 @@ struct ib_cq *mlx5_ib_create_cq(struct ib_device *ibdev, int entries,
 		}
 
 
-	mlx5_vfree(cqb);
+	kvfree(cqb);
 	return &cq->ibcq;
 
 err_cmd:
 	mlx5_core_destroy_cq(dev->mdev, &cq->mcq);
 
 err_cqb:
-	mlx5_vfree(cqb);
+	kvfree(cqb);
 	if (context)
 		destroy_cq_user(cq, context);
 	else
@@ -919,7 +919,7 @@ int mlx5_ib_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period)
 	int err;
 	u32 fsel;
 
-	if (!(dev->mdev->caps.flags & MLX5_DEV_CAP_FLAG_CQ_MODER))
+	if (!(dev->mdev->caps.gen.flags & MLX5_DEV_CAP_FLAG_CQ_MODER))
 		return -ENOSYS;
 
 	in = kzalloc(sizeof(*in), GFP_KERNEL);
@@ -1074,7 +1074,7 @@ int mlx5_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata)
 	int uninitialized_var(cqe_size);
 	unsigned long flags;
 
-	if (!(dev->mdev->caps.flags & MLX5_DEV_CAP_FLAG_RESIZE_CQ)) {
+	if (!(dev->mdev->caps.gen.flags & MLX5_DEV_CAP_FLAG_RESIZE_CQ)) {
 		pr_info("Firmware does not support resize CQ\n");
 		return -ENOSYS;
 	}
@@ -1083,7 +1083,7 @@ int mlx5_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata)
 		return -EINVAL;
 
 	entries = roundup_pow_of_two(entries + 1);
-	if (entries > dev->mdev->caps.max_cqes + 1)
+	if (entries > dev->mdev->caps.gen.max_cqes + 1)
 		return -EINVAL;
 
 	if (entries == ibcq->cqe + 1)
@@ -1159,11 +1159,11 @@ int mlx5_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata)
 	}
 	mutex_unlock(&cq->resize_mutex);
 
-	mlx5_vfree(in);
+	kvfree(in);
 	return 0;
 
 ex_alloc:
-	mlx5_vfree(in);
+	kvfree(in);
 
 ex_resize:
 	if (udata)
diff --git a/drivers/infiniband/hw/mlx5/mad.c b/drivers/infiniband/hw/mlx5/mad.c
index b514bbb5610f..657af9a1167c 100644
--- a/drivers/infiniband/hw/mlx5/mad.c
+++ b/drivers/infiniband/hw/mlx5/mad.c
@@ -129,7 +129,7 @@ int mlx5_query_ext_port_caps(struct mlx5_ib_dev *dev, u8 port)
 
 	packet_error = be16_to_cpu(out_mad->status);
 
-	dev->mdev->caps.ext_port_cap[port - 1] = (!err && !packet_error) ?
+	dev->mdev->caps.gen.ext_port_cap[port - 1] = (!err && !packet_error) ?
 		MLX_EXT_PORT_CAP_FLAG_EXTENDED_PORT_INFO : 0;
 
 out:
diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index d8907b20522a..8a87404e9c76 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -157,11 +157,13 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
 	struct ib_smp *in_mad  = NULL;
 	struct ib_smp *out_mad = NULL;
+	struct mlx5_general_caps *gen;
 	int err = -ENOMEM;
 	int max_rq_sg;
 	int max_sq_sg;
 	u64 flags;
 
+	gen = &dev->mdev->caps.gen;
 	in_mad  = kzalloc(sizeof(*in_mad), GFP_KERNEL);
 	out_mad = kmalloc(sizeof(*out_mad), GFP_KERNEL);
 	if (!in_mad || !out_mad)
@@ -183,7 +185,7 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
 		IB_DEVICE_PORT_ACTIVE_EVENT		|
 		IB_DEVICE_SYS_IMAGE_GUID		|
 		IB_DEVICE_RC_RNR_NAK_GEN;
-	flags = dev->mdev->caps.flags;
+	flags = gen->flags;
 	if (flags & MLX5_DEV_CAP_FLAG_BAD_PKEY_CNTR)
 		props->device_cap_flags |= IB_DEVICE_BAD_PKEY_CNTR;
 	if (flags & MLX5_DEV_CAP_FLAG_BAD_QKEY_CNTR)
@@ -213,34 +215,41 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
 	memcpy(&props->sys_image_guid, out_mad->data +	4, 8);
 
 	props->max_mr_size	   = ~0ull;
-	props->page_size_cap	   = dev->mdev->caps.min_page_sz;
-	props->max_qp		   = 1 << dev->mdev->caps.log_max_qp;
-	props->max_qp_wr	   = dev->mdev->caps.max_wqes;
-	max_rq_sg = dev->mdev->caps.max_rq_desc_sz / sizeof(struct mlx5_wqe_data_seg);
-	max_sq_sg = (dev->mdev->caps.max_sq_desc_sz - sizeof(struct mlx5_wqe_ctrl_seg)) /
+	props->page_size_cap	   = gen->min_page_sz;
+	props->max_qp		   = 1 << gen->log_max_qp;
+	props->max_qp_wr	   = gen->max_wqes;
+	max_rq_sg = gen->max_rq_desc_sz / sizeof(struct mlx5_wqe_data_seg);
+	max_sq_sg = (gen->max_sq_desc_sz - sizeof(struct mlx5_wqe_ctrl_seg)) /
 		sizeof(struct mlx5_wqe_data_seg);
 	props->max_sge = min(max_rq_sg, max_sq_sg);
-	props->max_cq		   = 1 << dev->mdev->caps.log_max_cq;
-	props->max_cqe		   = dev->mdev->caps.max_cqes - 1;
-	props->max_mr		   = 1 << dev->mdev->caps.log_max_mkey;
-	props->max_pd		   = 1 << dev->mdev->caps.log_max_pd;
-	props->max_qp_rd_atom	   = dev->mdev->caps.max_ra_req_qp;
-	props->max_qp_init_rd_atom = dev->mdev->caps.max_ra_res_qp;
+	props->max_cq		   = 1 << gen->log_max_cq;
+	props->max_cqe		   = gen->max_cqes - 1;
+	props->max_mr		   = 1 << gen->log_max_mkey;
+	props->max_pd		   = 1 << gen->log_max_pd;
+	props->max_qp_rd_atom	   = 1 << gen->log_max_ra_req_qp;
+	props->max_qp_init_rd_atom = 1 << gen->log_max_ra_res_qp;
+	props->max_srq		   = 1 << gen->log_max_srq;
+	props->max_srq_wr	   = gen->max_srq_wqes - 1;
+	props->local_ca_ack_delay  = gen->local_ca_ack_delay;
 	props->max_res_rd_atom	   = props->max_qp_rd_atom * props->max_qp;
-	props->max_srq		   = 1 << dev->mdev->caps.log_max_srq;
-	props->max_srq_wr	   = dev->mdev->caps.max_srq_wqes - 1;
 	props->max_srq_sge	   = max_rq_sg - 1;
 	props->max_fast_reg_page_list_len = (unsigned int)-1;
-	props->local_ca_ack_delay  = dev->mdev->caps.local_ca_ack_delay;
+	props->local_ca_ack_delay  = gen->local_ca_ack_delay;
 	props->atomic_cap	   = IB_ATOMIC_NONE;
 	props->masked_atomic_cap   = IB_ATOMIC_NONE;
 	props->max_pkeys	   = be16_to_cpup((__be16 *)(out_mad->data + 28));
-	props->max_mcast_grp	   = 1 << dev->mdev->caps.log_max_mcg;
-	props->max_mcast_qp_attach = dev->mdev->caps.max_qp_mcg;
+	props->max_mcast_grp	   = 1 << gen->log_max_mcg;
+	props->max_mcast_qp_attach = gen->max_qp_mcg;
 	props->max_total_mcast_qp_attach = props->max_mcast_qp_attach *
 					   props->max_mcast_grp;
 	props->max_map_per_fmr = INT_MAX; /* no limit in ConnectIB */
 
+#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
+	if (dev->mdev->caps.gen.flags & MLX5_DEV_CAP_FLAG_ON_DMND_PG)
+		props->device_cap_flags |= IB_DEVICE_ON_DEMAND_PAGING;
+	props->odp_caps = dev->odp_caps;
+#endif
+
 out:
 	kfree(in_mad);
 	kfree(out_mad);
@@ -254,10 +263,12 @@ int mlx5_ib_query_port(struct ib_device *ibdev, u8 port,
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
 	struct ib_smp *in_mad  = NULL;
 	struct ib_smp *out_mad = NULL;
+	struct mlx5_general_caps *gen;
 	int ext_active_speed;
 	int err = -ENOMEM;
 
-	if (port < 1 || port > dev->mdev->caps.num_ports) {
+	gen = &dev->mdev->caps.gen;
+	if (port < 1 || port > gen->num_ports) {
 		mlx5_ib_warn(dev, "invalid port number %d\n", port);
 		return -EINVAL;
 	}
@@ -288,8 +299,8 @@ int mlx5_ib_query_port(struct ib_device *ibdev, u8 port,
 	props->phys_state	= out_mad->data[33] >> 4;
 	props->port_cap_flags	= be32_to_cpup((__be32 *)(out_mad->data + 20));
 	props->gid_tbl_len	= out_mad->data[50];
-	props->max_msg_sz	= 1 << to_mdev(ibdev)->mdev->caps.log_max_msg;
-	props->pkey_tbl_len	= to_mdev(ibdev)->mdev->caps.port[port - 1].pkey_table_len;
+	props->max_msg_sz	= 1 << gen->log_max_msg;
+	props->pkey_tbl_len	= gen->port[port - 1].pkey_table_len;
 	props->bad_pkey_cntr	= be16_to_cpup((__be16 *)(out_mad->data + 46));
 	props->qkey_viol_cntr	= be16_to_cpup((__be16 *)(out_mad->data + 48));
 	props->active_width	= out_mad->data[31] & 0xf;
@@ -316,7 +327,7 @@ int mlx5_ib_query_port(struct ib_device *ibdev, u8 port,
 
 	/* If reported active speed is QDR, check if is FDR-10 */
 	if (props->active_speed == 4) {
-		if (dev->mdev->caps.ext_port_cap[port - 1] &
+		if (gen->ext_port_cap[port - 1] &
 		    MLX_EXT_PORT_CAP_FLAG_EXTENDED_PORT_INFO) {
 			init_query_mad(in_mad);
 			in_mad->attr_id = MLX5_ATTR_EXTENDED_PORT_INFO;
@@ -470,6 +481,7 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev,
 	struct mlx5_ib_alloc_ucontext_req_v2 req;
 	struct mlx5_ib_alloc_ucontext_resp resp;
 	struct mlx5_ib_ucontext *context;
+	struct mlx5_general_caps *gen;
 	struct mlx5_uuar_info *uuari;
 	struct mlx5_uar *uars;
 	int gross_uuars;
@@ -480,6 +492,7 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev,
 	int i;
 	size_t reqlen;
 
+	gen = &dev->mdev->caps.gen;
 	if (!dev->ib_active)
 		return ERR_PTR(-EAGAIN);
 
@@ -512,14 +525,14 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev,
 
 	num_uars = req.total_num_uuars / MLX5_NON_FP_BF_REGS_PER_PAGE;
 	gross_uuars = num_uars * MLX5_BF_REGS_PER_PAGE;
-	resp.qp_tab_size      = 1 << dev->mdev->caps.log_max_qp;
-	resp.bf_reg_size      = dev->mdev->caps.bf_reg_size;
+	resp.qp_tab_size      = 1 << gen->log_max_qp;
+	resp.bf_reg_size      = gen->bf_reg_size;
 	resp.cache_line_size  = L1_CACHE_BYTES;
-	resp.max_sq_desc_sz = dev->mdev->caps.max_sq_desc_sz;
-	resp.max_rq_desc_sz = dev->mdev->caps.max_rq_desc_sz;
-	resp.max_send_wqebb = dev->mdev->caps.max_wqes;
-	resp.max_recv_wr = dev->mdev->caps.max_wqes;
-	resp.max_srq_recv_wr = dev->mdev->caps.max_srq_wqes;
+	resp.max_sq_desc_sz = gen->max_sq_desc_sz;
+	resp.max_rq_desc_sz = gen->max_rq_desc_sz;
+	resp.max_send_wqebb = gen->max_wqes;
+	resp.max_recv_wr = gen->max_wqes;
+	resp.max_srq_recv_wr = gen->max_srq_wqes;
 
 	context = kzalloc(sizeof(*context), GFP_KERNEL);
 	if (!context)
@@ -561,11 +574,15 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev,
 			goto out_count;
 	}
 
+#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
+	context->ibucontext.invalidate_range = &mlx5_ib_invalidate_range;
+#endif
+
 	INIT_LIST_HEAD(&context->db_page_list);
 	mutex_init(&context->db_page_mutex);
 
 	resp.tot_uuars = req.total_num_uuars;
-	resp.num_ports = dev->mdev->caps.num_ports;
+	resp.num_ports = gen->num_ports;
 	err = ib_copy_to_udata(udata, &resp,
 			       sizeof(resp) - sizeof(resp.reserved));
 	if (err)
@@ -650,13 +667,13 @@ static int mlx5_ib_mmap(struct ib_ucontext *ibcontext, struct vm_area_struct *vm
 			return -EINVAL;
 
 		idx = get_index(vma->vm_pgoff);
+		if (idx >= uuari->num_uars)
+			return -EINVAL;
+
 		pfn = uar_index2pfn(dev, uuari->uars[idx].index);
 		mlx5_ib_dbg(dev, "uar idx 0x%lx, pfn 0x%llx\n", idx,
 			    (unsigned long long)pfn);
 
-		if (idx >= uuari->num_uars)
-			return -EINVAL;
-
 		vma->vm_page_prot = pgprot_writecombine(vma->vm_page_prot);
 		if (io_remap_pfn_range(vma, vma->vm_start, pfn,
 				       PAGE_SIZE, vma->vm_page_prot))
@@ -851,7 +868,7 @@ static ssize_t show_reg_pages(struct device *device,
 	struct mlx5_ib_dev *dev =
 		container_of(device, struct mlx5_ib_dev, ib_dev.dev);
 
-	return sprintf(buf, "%d\n", dev->mdev->priv.reg_pages);
+	return sprintf(buf, "%d\n", atomic_read(&dev->mdev->priv.reg_pages));
 }
 
 static ssize_t show_hca(struct device *device, struct device_attribute *attr,
@@ -967,9 +984,11 @@ static void mlx5_ib_event(struct mlx5_core_dev *dev, void *context,
 
 static void get_ext_port_caps(struct mlx5_ib_dev *dev)
 {
+	struct mlx5_general_caps *gen;
 	int port;
 
-	for (port = 1; port <= dev->mdev->caps.num_ports; port++)
+	gen = &dev->mdev->caps.gen;
+	for (port = 1; port <= gen->num_ports; port++)
 		mlx5_query_ext_port_caps(dev, port);
 }
 
@@ -977,9 +996,11 @@ static int get_port_caps(struct mlx5_ib_dev *dev)
 {
 	struct ib_device_attr *dprops = NULL;
 	struct ib_port_attr *pprops = NULL;
+	struct mlx5_general_caps *gen;
 	int err = 0;
 	int port;
 
+	gen = &dev->mdev->caps.gen;
 	pprops = kmalloc(sizeof(*pprops), GFP_KERNEL);
 	if (!pprops)
 		goto out;
@@ -994,14 +1015,14 @@ static int get_port_caps(struct mlx5_ib_dev *dev)
 		goto out;
 	}
 
-	for (port = 1; port <= dev->mdev->caps.num_ports; port++) {
+	for (port = 1; port <= gen->num_ports; port++) {
 		err = mlx5_ib_query_port(&dev->ib_dev, port, pprops);
 		if (err) {
 			mlx5_ib_warn(dev, "query_port %d failed %d\n", port, err);
 			break;
 		}
-		dev->mdev->caps.port[port - 1].pkey_table_len = dprops->max_pkeys;
-		dev->mdev->caps.port[port - 1].gid_table_len = pprops->gid_tbl_len;
+		gen->port[port - 1].pkey_table_len = dprops->max_pkeys;
+		gen->port[port - 1].gid_table_len = pprops->gid_tbl_len;
 		mlx5_ib_dbg(dev, "pkey_table_len %d, gid_table_len %d\n",
 			    dprops->max_pkeys, pprops->gid_tbl_len);
 	}
@@ -1279,8 +1300,8 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev)
 	strlcpy(dev->ib_dev.name, "mlx5_%d", IB_DEVICE_NAME_MAX);
 	dev->ib_dev.owner		= THIS_MODULE;
 	dev->ib_dev.node_type		= RDMA_NODE_IB_CA;
-	dev->ib_dev.local_dma_lkey	= mdev->caps.reserved_lkey;
-	dev->num_ports		= mdev->caps.num_ports;
+	dev->ib_dev.local_dma_lkey	= mdev->caps.gen.reserved_lkey;
+	dev->num_ports		= mdev->caps.gen.num_ports;
 	dev->ib_dev.phys_port_cnt     = dev->num_ports;
 	dev->ib_dev.num_comp_vectors	= dev->num_comp_vectors;
 	dev->ib_dev.dma_device	= &mdev->pdev->dev;
@@ -1310,6 +1331,8 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev)
 		(1ull << IB_USER_VERBS_CMD_DESTROY_SRQ)		|
 		(1ull << IB_USER_VERBS_CMD_CREATE_XSRQ)		|
 		(1ull << IB_USER_VERBS_CMD_OPEN_QP);
+	dev->ib_dev.uverbs_ex_cmd_mask =
+		(1ull << IB_USER_VERBS_EX_CMD_QUERY_DEVICE);
 
 	dev->ib_dev.query_device	= mlx5_ib_query_device;
 	dev->ib_dev.query_port		= mlx5_ib_query_port;
@@ -1355,7 +1378,9 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev)
 	dev->ib_dev.free_fast_reg_page_list  = mlx5_ib_free_fast_reg_page_list;
 	dev->ib_dev.check_mr_status	= mlx5_ib_check_mr_status;
 
-	if (mdev->caps.flags & MLX5_DEV_CAP_FLAG_XRC) {
+	mlx5_ib_internal_query_odp_caps(dev);
+
+	if (mdev->caps.gen.flags & MLX5_DEV_CAP_FLAG_XRC) {
 		dev->ib_dev.alloc_xrcd = mlx5_ib_alloc_xrcd;
 		dev->ib_dev.dealloc_xrcd = mlx5_ib_dealloc_xrcd;
 		dev->ib_dev.uverbs_cmd_mask |=
@@ -1368,16 +1393,19 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev)
 		goto err_eqs;
 
 	mutex_init(&dev->cap_mask_mutex);
-	spin_lock_init(&dev->mr_lock);
 
 	err = create_dev_resources(&dev->devr);
 	if (err)
 		goto err_eqs;
 
-	err = ib_register_device(&dev->ib_dev, NULL);
+	err = mlx5_ib_odp_init_one(dev);
 	if (err)
 		goto err_rsrc;
 
+	err = ib_register_device(&dev->ib_dev, NULL);
+	if (err)
+		goto err_odp;
+
 	err = create_umr_res(dev);
 	if (err)
 		goto err_dev;
@@ -1399,6 +1427,9 @@ err_umrc:
 err_dev:
 	ib_unregister_device(&dev->ib_dev);
 
+err_odp:
+	mlx5_ib_odp_remove_one(dev);
+
 err_rsrc:
 	destroy_dev_resources(&dev->devr);
 
@@ -1414,8 +1445,10 @@ err_dealloc:
 static void mlx5_ib_remove(struct mlx5_core_dev *mdev, void *context)
 {
 	struct mlx5_ib_dev *dev = context;
-	destroy_umrc_res(dev);
+
 	ib_unregister_device(&dev->ib_dev);
+	destroy_umrc_res(dev);
+	mlx5_ib_odp_remove_one(dev);
 	destroy_dev_resources(&dev->devr);
 	free_comp_eqs(dev);
 	ib_dealloc_device(&dev->ib_dev);
@@ -1429,15 +1462,30 @@ static struct mlx5_interface mlx5_ib_interface = {
 
 static int __init mlx5_ib_init(void)
 {
+	int err;
+
 	if (deprecated_prof_sel != 2)
 		pr_warn("prof_sel is deprecated for mlx5_ib, set it for mlx5_core\n");
 
-	return mlx5_register_interface(&mlx5_ib_interface);
+	err = mlx5_ib_odp_init();
+	if (err)
+		return err;
+
+	err = mlx5_register_interface(&mlx5_ib_interface);
+	if (err)
+		goto clean_odp;
+
+	return err;
+
+clean_odp:
+	mlx5_ib_odp_cleanup();
+	return err;
 }
 
 static void __exit mlx5_ib_cleanup(void)
 {
 	mlx5_unregister_interface(&mlx5_ib_interface);
+	mlx5_ib_odp_cleanup();
 }
 
 module_init(mlx5_ib_init);
diff --git a/drivers/infiniband/hw/mlx5/mem.c b/drivers/infiniband/hw/mlx5/mem.c
index a3e81444c825..b56e4c5593ee 100644
--- a/drivers/infiniband/hw/mlx5/mem.c
+++ b/drivers/infiniband/hw/mlx5/mem.c
@@ -32,6 +32,7 @@
 
 #include <linux/module.h>
 #include <rdma/ib_umem.h>
+#include <rdma/ib_umem_odp.h>
 #include "mlx5_ib.h"
 
 /* @umem: umem object to scan
@@ -55,16 +56,28 @@ void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr, int *count, int *shift,
 	u64 pfn;
 	struct scatterlist *sg;
 	int entry;
+	unsigned long page_shift = ilog2(umem->page_size);
 
-	addr = addr >> PAGE_SHIFT;
+	/* With ODP we must always match OS page size. */
+	if (umem->odp_data) {
+		*count = ib_umem_page_count(umem);
+		*shift = PAGE_SHIFT;
+		*ncont = *count;
+		if (order)
+			*order = ilog2(roundup_pow_of_two(*count));
+
+		return;
+	}
+
+	addr = addr >> page_shift;
 	tmp = (unsigned long)addr;
 	m = find_first_bit(&tmp, sizeof(tmp));
 	skip = 1 << m;
 	mask = skip - 1;
 	i = 0;
 	for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) {
-		len = sg_dma_len(sg) >> PAGE_SHIFT;
-		pfn = sg_dma_address(sg) >> PAGE_SHIFT;
+		len = sg_dma_len(sg) >> page_shift;
+		pfn = sg_dma_address(sg) >> page_shift;
 		for (k = 0; k < len; k++) {
 			if (!(i & mask)) {
 				tmp = (unsigned long)pfn;
@@ -103,14 +116,43 @@ void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr, int *count, int *shift,
 
 		*ncont = 0;
 	}
-	*shift = PAGE_SHIFT + m;
+	*shift = page_shift + m;
 	*count = i;
 }
 
-void mlx5_ib_populate_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem,
-			  int page_shift, __be64 *pas, int umr)
+#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
+static u64 umem_dma_to_mtt(dma_addr_t umem_dma)
 {
-	int shift = page_shift - PAGE_SHIFT;
+	u64 mtt_entry = umem_dma & ODP_DMA_ADDR_MASK;
+
+	if (umem_dma & ODP_READ_ALLOWED_BIT)
+		mtt_entry |= MLX5_IB_MTT_READ;
+	if (umem_dma & ODP_WRITE_ALLOWED_BIT)
+		mtt_entry |= MLX5_IB_MTT_WRITE;
+
+	return mtt_entry;
+}
+#endif
+
+/*
+ * Populate the given array with bus addresses from the umem.
+ *
+ * dev - mlx5_ib device
+ * umem - umem to use to fill the pages
+ * page_shift - determines the page size used in the resulting array
+ * offset - offset into the umem to start from,
+ *          only implemented for ODP umems
+ * num_pages - total number of pages to fill
+ * pas - bus addresses array to fill
+ * access_flags - access flags to set on all present pages.
+		  use enum mlx5_ib_mtt_access_flags for this.
+ */
+void __mlx5_ib_populate_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem,
+			    int page_shift, size_t offset, size_t num_pages,
+			    __be64 *pas, int access_flags)
+{
+	unsigned long umem_page_shift = ilog2(umem->page_size);
+	int shift = page_shift - umem_page_shift;
 	int mask = (1 << shift) - 1;
 	int i, k;
 	u64 cur = 0;
@@ -118,28 +160,49 @@ void mlx5_ib_populate_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem,
 	int len;
 	struct scatterlist *sg;
 	int entry;
+#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
+	const bool odp = umem->odp_data != NULL;
+
+	if (odp) {
+		WARN_ON(shift != 0);
+		WARN_ON(access_flags != (MLX5_IB_MTT_READ | MLX5_IB_MTT_WRITE));
+
+		for (i = 0; i < num_pages; ++i) {
+			dma_addr_t pa = umem->odp_data->dma_list[offset + i];
+
+			pas[i] = cpu_to_be64(umem_dma_to_mtt(pa));
+		}
+		return;
+	}
+#endif
 
 	i = 0;
 	for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) {
-		len = sg_dma_len(sg) >> PAGE_SHIFT;
+		len = sg_dma_len(sg) >> umem_page_shift;
 		base = sg_dma_address(sg);
 		for (k = 0; k < len; k++) {
 			if (!(i & mask)) {
-				cur = base + (k << PAGE_SHIFT);
-				if (umr)
-					cur |= 3;
+				cur = base + (k << umem_page_shift);
+				cur |= access_flags;
 
 				pas[i >> shift] = cpu_to_be64(cur);
 				mlx5_ib_dbg(dev, "pas[%d] 0x%llx\n",
 					    i >> shift, be64_to_cpu(pas[i >> shift]));
 			}  else
 				mlx5_ib_dbg(dev, "=====> 0x%llx\n",
-					    base + (k << PAGE_SHIFT));
+					    base + (k << umem_page_shift));
 			i++;
 		}
 	}
 }
 
+void mlx5_ib_populate_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem,
+			  int page_shift, __be64 *pas, int access_flags)
+{
+	return __mlx5_ib_populate_pas(dev, umem, page_shift, 0,
+				      ib_umem_num_pages(umem), pas,
+				      access_flags);
+}
 int mlx5_ib_get_buf_offset(u64 addr, int page_shift, u32 *offset)
 {
 	u64 page_size;
diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h
index 386780f0d1e1..83f22fe297c8 100644
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -111,6 +111,8 @@ struct mlx5_ib_pd {
  */
 
 #define MLX5_IB_SEND_UMR_UNREG	IB_SEND_RESERVED_START
+#define MLX5_IB_SEND_UMR_FAIL_IF_FREE (IB_SEND_RESERVED_START << 1)
+#define MLX5_IB_SEND_UMR_UPDATE_MTT (IB_SEND_RESERVED_START << 2)
 #define MLX5_IB_QPT_REG_UMR	IB_QPT_RESERVED1
 #define MLX5_IB_WR_UMR		IB_WR_RESERVED1
 
@@ -147,6 +149,29 @@ enum {
 	MLX5_QP_EMPTY
 };
 
+/*
+ * Connect-IB can trigger up to four concurrent pagefaults
+ * per-QP.
+ */
+enum mlx5_ib_pagefault_context {
+	MLX5_IB_PAGEFAULT_RESPONDER_READ,
+	MLX5_IB_PAGEFAULT_REQUESTOR_READ,
+	MLX5_IB_PAGEFAULT_RESPONDER_WRITE,
+	MLX5_IB_PAGEFAULT_REQUESTOR_WRITE,
+	MLX5_IB_PAGEFAULT_CONTEXTS
+};
+
+static inline enum mlx5_ib_pagefault_context
+	mlx5_ib_get_pagefault_context(struct mlx5_pagefault *pagefault)
+{
+	return pagefault->flags & (MLX5_PFAULT_REQUESTOR | MLX5_PFAULT_WRITE);
+}
+
+struct mlx5_ib_pfault {
+	struct work_struct	work;
+	struct mlx5_pagefault	mpfault;
+};
+
 struct mlx5_ib_qp {
 	struct ib_qp		ibqp;
 	struct mlx5_core_qp	mqp;
@@ -192,6 +217,21 @@ struct mlx5_ib_qp {
 
 	/* Store signature errors */
 	bool			signature_en;
+
+#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
+	/*
+	 * A flag that is true for QP's that are in a state that doesn't
+	 * allow page faults, and shouldn't schedule any more faults.
+	 */
+	int                     disable_page_faults;
+	/*
+	 * The disable_page_faults_lock protects a QP's disable_page_faults
+	 * field, allowing for a thread to atomically check whether the QP
+	 * allows page faults, and if so schedule a page fault.
+	 */
+	spinlock_t              disable_page_faults_lock;
+	struct mlx5_ib_pfault	pagefaults[MLX5_IB_PAGEFAULT_CONTEXTS];
+#endif
 };
 
 struct mlx5_ib_cq_buf {
@@ -206,6 +246,19 @@ enum mlx5_ib_qp_flags {
 	MLX5_IB_QP_SIGNATURE_HANDLING           = 1 << 1,
 };
 
+struct mlx5_umr_wr {
+	union {
+		u64			virt_addr;
+		u64			offset;
+	} target;
+	struct ib_pd		       *pd;
+	unsigned int			page_shift;
+	unsigned int			npages;
+	u32				length;
+	int				access_flags;
+	u32				mkey;
+};
+
 struct mlx5_shared_mr_info {
 	int mr_id;
 	struct ib_umem		*umem;
@@ -253,6 +306,13 @@ struct mlx5_ib_xrcd {
 	u32			xrcdn;
 };
 
+enum mlx5_ib_mtt_access_flags {
+	MLX5_IB_MTT_READ  = (1 << 0),
+	MLX5_IB_MTT_WRITE = (1 << 1),
+};
+
+#define MLX5_IB_MTT_PRESENT (MLX5_IB_MTT_READ | MLX5_IB_MTT_WRITE)
+
 struct mlx5_ib_mr {
 	struct ib_mr		ibmr;
 	struct mlx5_core_mr	mmr;
@@ -261,12 +321,11 @@ struct mlx5_ib_mr {
 	struct list_head	list;
 	int			order;
 	int			umred;
-	__be64			*pas;
-	dma_addr_t		dma;
 	int			npages;
 	struct mlx5_ib_dev     *dev;
 	struct mlx5_create_mkey_mbox_out out;
 	struct mlx5_core_sig_ctx    *sig;
+	int			live;
 };
 
 struct mlx5_ib_fast_reg_page_list {
@@ -372,11 +431,18 @@ struct mlx5_ib_dev {
 	struct umr_common		umrc;
 	/* sync used page count stats
 	 */
-	spinlock_t			mr_lock;
 	struct mlx5_ib_resources	devr;
 	struct mlx5_mr_cache		cache;
 	struct timer_list		delay_timer;
 	int				fill_delay;
+#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
+	struct ib_odp_caps	odp_caps;
+	/*
+	 * Sleepable RCU that prevents destruction of MRs while they are still
+	 * being used by a page fault handler.
+	 */
+	struct srcu_struct      mr_srcu;
+#endif
 };
 
 static inline struct mlx5_ib_cq *to_mibcq(struct mlx5_core_cq *mcq)
@@ -490,6 +556,8 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 int mlx5_ib_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr,
 		      struct ib_recv_wr **bad_wr);
 void *mlx5_get_send_wqe(struct mlx5_ib_qp *qp, int n);
+int mlx5_ib_read_user_wqe(struct mlx5_ib_qp *qp, int send, int wqe_index,
+			  void *buffer, u32 length);
 struct ib_cq *mlx5_ib_create_cq(struct ib_device *ibdev, int entries,
 				int vector, struct ib_ucontext *context,
 				struct ib_udata *udata);
@@ -502,6 +570,8 @@ struct ib_mr *mlx5_ib_get_dma_mr(struct ib_pd *pd, int acc);
 struct ib_mr *mlx5_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 				  u64 virt_addr, int access_flags,
 				  struct ib_udata *udata);
+int mlx5_ib_update_mtt(struct mlx5_ib_mr *mr, u64 start_page_index,
+		       int npages, int zap);
 int mlx5_ib_dereg_mr(struct ib_mr *ibmr);
 int mlx5_ib_destroy_mr(struct ib_mr *ibmr);
 struct ib_mr *mlx5_ib_create_mr(struct ib_pd *pd,
@@ -533,8 +603,11 @@ int mlx5_ib_init_fmr(struct mlx5_ib_dev *dev);
 void mlx5_ib_cleanup_fmr(struct mlx5_ib_dev *dev);
 void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr, int *count, int *shift,
 			int *ncont, int *order);
+void __mlx5_ib_populate_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem,
+			    int page_shift, size_t offset, size_t num_pages,
+			    __be64 *pas, int access_flags);
 void mlx5_ib_populate_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem,
-			  int page_shift, __be64 *pas, int umr);
+			  int page_shift, __be64 *pas, int access_flags);
 void mlx5_ib_copy_pas(u64 *old, u64 *new, int step, int num);
 int mlx5_ib_get_cqe_size(struct mlx5_ib_dev *dev, struct ib_cq *ibcq);
 int mlx5_mr_cache_init(struct mlx5_ib_dev *dev);
@@ -544,6 +617,38 @@ void mlx5_umr_cq_handler(struct ib_cq *cq, void *cq_context);
 int mlx5_ib_check_mr_status(struct ib_mr *ibmr, u32 check_mask,
 			    struct ib_mr_status *mr_status);
 
+#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
+extern struct workqueue_struct *mlx5_ib_page_fault_wq;
+
+int mlx5_ib_internal_query_odp_caps(struct mlx5_ib_dev *dev);
+void mlx5_ib_mr_pfault_handler(struct mlx5_ib_qp *qp,
+			       struct mlx5_ib_pfault *pfault);
+void mlx5_ib_odp_create_qp(struct mlx5_ib_qp *qp);
+int mlx5_ib_odp_init_one(struct mlx5_ib_dev *ibdev);
+void mlx5_ib_odp_remove_one(struct mlx5_ib_dev *ibdev);
+int __init mlx5_ib_odp_init(void);
+void mlx5_ib_odp_cleanup(void);
+void mlx5_ib_qp_disable_pagefaults(struct mlx5_ib_qp *qp);
+void mlx5_ib_qp_enable_pagefaults(struct mlx5_ib_qp *qp);
+void mlx5_ib_invalidate_range(struct ib_umem *umem, unsigned long start,
+			      unsigned long end);
+
+#else /* CONFIG_INFINIBAND_ON_DEMAND_PAGING */
+static inline int mlx5_ib_internal_query_odp_caps(struct mlx5_ib_dev *dev)
+{
+	return 0;
+}
+
+static inline void mlx5_ib_odp_create_qp(struct mlx5_ib_qp *qp)		{}
+static inline int mlx5_ib_odp_init_one(struct mlx5_ib_dev *ibdev) { return 0; }
+static inline void mlx5_ib_odp_remove_one(struct mlx5_ib_dev *ibdev)	{}
+static inline int mlx5_ib_odp_init(void) { return 0; }
+static inline void mlx5_ib_odp_cleanup(void)				{}
+static inline void mlx5_ib_qp_disable_pagefaults(struct mlx5_ib_qp *qp) {}
+static inline void mlx5_ib_qp_enable_pagefaults(struct mlx5_ib_qp *qp)  {}
+
+#endif /* CONFIG_INFINIBAND_ON_DEMAND_PAGING */
+
 static inline void init_query_mad(struct ib_smp *mad)
 {
 	mad->base_version  = 1;
@@ -561,4 +666,7 @@ static inline u8 convert_access(int acc)
 	       MLX5_PERM_LOCAL_READ;
 }
 
+#define MLX5_MAX_UMR_SHIFT 16
+#define MLX5_MAX_UMR_PAGES (1 << MLX5_MAX_UMR_SHIFT)
+
 #endif /* MLX5_IB_H */
diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c
index 80b3c63eab5d..32a28bd50b20 100644
--- a/drivers/infiniband/hw/mlx5/mr.c
+++ b/drivers/infiniband/hw/mlx5/mr.c
@@ -37,21 +37,34 @@
 #include <linux/export.h>
 #include <linux/delay.h>
 #include <rdma/ib_umem.h>
+#include <rdma/ib_umem_odp.h>
+#include <rdma/ib_verbs.h>
 #include "mlx5_ib.h"
 
 enum {
 	MAX_PENDING_REG_MR = 8,
 };
 
-enum {
-	MLX5_UMR_ALIGN	= 2048
-};
+#define MLX5_UMR_ALIGN 2048
+#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
+static __be64 mlx5_ib_update_mtt_emergency_buffer[
+		MLX5_UMR_MTT_MIN_CHUNK_SIZE/sizeof(__be64)]
+	__aligned(MLX5_UMR_ALIGN);
+static DEFINE_MUTEX(mlx5_ib_update_mtt_emergency_buffer_mutex);
+#endif
+
+static int clean_mr(struct mlx5_ib_mr *mr);
 
-static __be64 *mr_align(__be64 *ptr, int align)
+static int destroy_mkey(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr)
 {
-	unsigned long mask = align - 1;
+	int err = mlx5_core_destroy_mkey(dev->mdev, &mr->mmr);
 
-	return (__be64 *)(((unsigned long)ptr + mask) & ~mask);
+#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
+	/* Wait until all page fault handlers using the mr complete. */
+	synchronize_srcu(&dev->mr_srcu);
+#endif
+
+	return err;
 }
 
 static int order2idx(struct mlx5_ib_dev *dev, int order)
@@ -146,7 +159,7 @@ static int add_keys(struct mlx5_ib_dev *dev, int c, int num)
 		mr->order = ent->order;
 		mr->umred = 1;
 		mr->dev = dev;
-		in->seg.status = 1 << 6;
+		in->seg.status = MLX5_MKEY_STATUS_FREE;
 		in->seg.xlt_oct_size = cpu_to_be32((npages + 1) / 2);
 		in->seg.qpn_mkey7_0 = cpu_to_be32(0xffffff << 8);
 		in->seg.flags = MLX5_ACCESS_MODE_MTT | MLX5_PERM_UMR_EN;
@@ -159,6 +172,9 @@ static int add_keys(struct mlx5_ib_dev *dev, int c, int num)
 					    sizeof(*in), reg_mr_callback,
 					    mr, &mr->out);
 		if (err) {
+			spin_lock_irq(&ent->lock);
+			ent->pending--;
+			spin_unlock_irq(&ent->lock);
 			mlx5_ib_warn(dev, "create mkey failed %d\n", err);
 			kfree(mr);
 			break;
@@ -188,7 +204,7 @@ static void remove_keys(struct mlx5_ib_dev *dev, int c, int num)
 		ent->cur--;
 		ent->size--;
 		spin_unlock_irq(&ent->lock);
-		err = mlx5_core_destroy_mkey(dev->mdev, &mr->mmr);
+		err = destroy_mkey(dev, mr);
 		if (err)
 			mlx5_ib_warn(dev, "failed destroy mkey\n");
 		else
@@ -479,7 +495,7 @@ static void clean_keys(struct mlx5_ib_dev *dev, int c)
 		ent->cur--;
 		ent->size--;
 		spin_unlock_irq(&ent->lock);
-		err = mlx5_core_destroy_mkey(dev->mdev, &mr->mmr);
+		err = destroy_mkey(dev, mr);
 		if (err)
 			mlx5_ib_warn(dev, "failed destroy mkey\n");
 		else
@@ -665,7 +681,7 @@ static int get_octo_len(u64 addr, u64 len, int page_size)
 
 static int use_umr(int order)
 {
-	return order <= 17;
+	return order <= MLX5_MAX_UMR_SHIFT;
 }
 
 static void prep_umr_reg_wqe(struct ib_pd *pd, struct ib_send_wr *wr,
@@ -675,6 +691,7 @@ static void prep_umr_reg_wqe(struct ib_pd *pd, struct ib_send_wr *wr,
 {
 	struct mlx5_ib_dev *dev = to_mdev(pd->device);
 	struct ib_mr *mr = dev->umrc.mr;
+	struct mlx5_umr_wr *umrwr = (struct mlx5_umr_wr *)&wr->wr.fast_reg;
 
 	sg->addr = dma;
 	sg->length = ALIGN(sizeof(u64) * n, 64);
@@ -689,21 +706,24 @@ static void prep_umr_reg_wqe(struct ib_pd *pd, struct ib_send_wr *wr,
 		wr->num_sge = 0;
 
 	wr->opcode = MLX5_IB_WR_UMR;
-	wr->wr.fast_reg.page_list_len = n;
-	wr->wr.fast_reg.page_shift = page_shift;
-	wr->wr.fast_reg.rkey = key;
-	wr->wr.fast_reg.iova_start = virt_addr;
-	wr->wr.fast_reg.length = len;
-	wr->wr.fast_reg.access_flags = access_flags;
-	wr->wr.fast_reg.page_list = (struct ib_fast_reg_page_list *)pd;
+
+	umrwr->npages = n;
+	umrwr->page_shift = page_shift;
+	umrwr->mkey = key;
+	umrwr->target.virt_addr = virt_addr;
+	umrwr->length = len;
+	umrwr->access_flags = access_flags;
+	umrwr->pd = pd;
 }
 
 static void prep_umr_unreg_wqe(struct mlx5_ib_dev *dev,
 			       struct ib_send_wr *wr, u32 key)
 {
-	wr->send_flags = MLX5_IB_SEND_UMR_UNREG;
+	struct mlx5_umr_wr *umrwr = (struct mlx5_umr_wr *)&wr->wr.fast_reg;
+
+	wr->send_flags = MLX5_IB_SEND_UMR_UNREG | MLX5_IB_SEND_UMR_FAIL_IF_FREE;
 	wr->opcode = MLX5_IB_WR_UMR;
-	wr->wr.fast_reg.rkey = key;
+	umrwr->mkey = key;
 }
 
 void mlx5_umr_cq_handler(struct ib_cq *cq, void *cq_context)
@@ -739,7 +759,10 @@ static struct mlx5_ib_mr *reg_umr(struct ib_pd *pd, struct ib_umem *umem,
 	struct ib_send_wr wr, *bad;
 	struct mlx5_ib_mr *mr;
 	struct ib_sge sg;
-	int size = sizeof(u64) * npages;
+	int size;
+	__be64 *mr_pas;
+	__be64 *pas;
+	dma_addr_t dma;
 	int err = 0;
 	int i;
 
@@ -758,25 +781,31 @@ static struct mlx5_ib_mr *reg_umr(struct ib_pd *pd, struct ib_umem *umem,
 	if (!mr)
 		return ERR_PTR(-EAGAIN);
 
-	mr->pas = kmalloc(size + MLX5_UMR_ALIGN - 1, GFP_KERNEL);
-	if (!mr->pas) {
+	/* UMR copies MTTs in units of MLX5_UMR_MTT_ALIGNMENT bytes.
+	 * To avoid copying garbage after the pas array, we allocate
+	 * a little more. */
+	size = ALIGN(sizeof(u64) * npages, MLX5_UMR_MTT_ALIGNMENT);
+	mr_pas = kmalloc(size + MLX5_UMR_ALIGN - 1, GFP_KERNEL);
+	if (!mr_pas) {
 		err = -ENOMEM;
 		goto free_mr;
 	}
 
-	mlx5_ib_populate_pas(dev, umem, page_shift,
-			     mr_align(mr->pas, MLX5_UMR_ALIGN), 1);
+	pas = PTR_ALIGN(mr_pas, MLX5_UMR_ALIGN);
+	mlx5_ib_populate_pas(dev, umem, page_shift, pas, MLX5_IB_MTT_PRESENT);
+	/* Clear padding after the actual pages. */
+	memset(pas + npages, 0, size - npages * sizeof(u64));
 
-	mr->dma = dma_map_single(ddev, mr_align(mr->pas, MLX5_UMR_ALIGN), size,
-				 DMA_TO_DEVICE);
-	if (dma_mapping_error(ddev, mr->dma)) {
+	dma = dma_map_single(ddev, pas, size, DMA_TO_DEVICE);
+	if (dma_mapping_error(ddev, dma)) {
 		err = -ENOMEM;
 		goto free_pas;
 	}
 
 	memset(&wr, 0, sizeof(wr));
 	wr.wr_id = (u64)(unsigned long)&umr_context;
-	prep_umr_reg_wqe(pd, &wr, &sg, mr->dma, npages, mr->mmr.key, page_shift, virt_addr, len, access_flags);
+	prep_umr_reg_wqe(pd, &wr, &sg, dma, npages, mr->mmr.key, page_shift,
+			 virt_addr, len, access_flags);
 
 	mlx5_ib_init_umr_context(&umr_context);
 	down(&umrc->sem);
@@ -796,12 +825,14 @@ static struct mlx5_ib_mr *reg_umr(struct ib_pd *pd, struct ib_umem *umem,
 	mr->mmr.size = len;
 	mr->mmr.pd = to_mpd(pd)->pdn;
 
+	mr->live = 1;
+
 unmap_dma:
 	up(&umrc->sem);
-	dma_unmap_single(ddev, mr->dma, size, DMA_TO_DEVICE);
+	dma_unmap_single(ddev, dma, size, DMA_TO_DEVICE);
 
 free_pas:
-	kfree(mr->pas);
+	kfree(mr_pas);
 
 free_mr:
 	if (err) {
@@ -812,6 +843,128 @@ free_mr:
 	return mr;
 }
 
+#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
+int mlx5_ib_update_mtt(struct mlx5_ib_mr *mr, u64 start_page_index, int npages,
+		       int zap)
+{
+	struct mlx5_ib_dev *dev = mr->dev;
+	struct device *ddev = dev->ib_dev.dma_device;
+	struct umr_common *umrc = &dev->umrc;
+	struct mlx5_ib_umr_context umr_context;
+	struct ib_umem *umem = mr->umem;
+	int size;
+	__be64 *pas;
+	dma_addr_t dma;
+	struct ib_send_wr wr, *bad;
+	struct mlx5_umr_wr *umrwr = (struct mlx5_umr_wr *)&wr.wr.fast_reg;
+	struct ib_sge sg;
+	int err = 0;
+	const int page_index_alignment = MLX5_UMR_MTT_ALIGNMENT / sizeof(u64);
+	const int page_index_mask = page_index_alignment - 1;
+	size_t pages_mapped = 0;
+	size_t pages_to_map = 0;
+	size_t pages_iter = 0;
+	int use_emergency_buf = 0;
+
+	/* UMR copies MTTs in units of MLX5_UMR_MTT_ALIGNMENT bytes,
+	 * so we need to align the offset and length accordingly */
+	if (start_page_index & page_index_mask) {
+		npages += start_page_index & page_index_mask;
+		start_page_index &= ~page_index_mask;
+	}
+
+	pages_to_map = ALIGN(npages, page_index_alignment);
+
+	if (start_page_index + pages_to_map > MLX5_MAX_UMR_PAGES)
+		return -EINVAL;
+
+	size = sizeof(u64) * pages_to_map;
+	size = min_t(int, PAGE_SIZE, size);
+	/* We allocate with GFP_ATOMIC to avoid recursion into page-reclaim
+	 * code, when we are called from an invalidation. The pas buffer must
+	 * be 2k-aligned for Connect-IB. */
+	pas = (__be64 *)get_zeroed_page(GFP_ATOMIC);
+	if (!pas) {
+		mlx5_ib_warn(dev, "unable to allocate memory during MTT update, falling back to slower chunked mechanism.\n");
+		pas = mlx5_ib_update_mtt_emergency_buffer;
+		size = MLX5_UMR_MTT_MIN_CHUNK_SIZE;
+		use_emergency_buf = 1;
+		mutex_lock(&mlx5_ib_update_mtt_emergency_buffer_mutex);
+		memset(pas, 0, size);
+	}
+	pages_iter = size / sizeof(u64);
+	dma = dma_map_single(ddev, pas, size, DMA_TO_DEVICE);
+	if (dma_mapping_error(ddev, dma)) {
+		mlx5_ib_err(dev, "unable to map DMA during MTT update.\n");
+		err = -ENOMEM;
+		goto free_pas;
+	}
+
+	for (pages_mapped = 0;
+	     pages_mapped < pages_to_map && !err;
+	     pages_mapped += pages_iter, start_page_index += pages_iter) {
+		dma_sync_single_for_cpu(ddev, dma, size, DMA_TO_DEVICE);
+
+		npages = min_t(size_t,
+			       pages_iter,
+			       ib_umem_num_pages(umem) - start_page_index);
+
+		if (!zap) {
+			__mlx5_ib_populate_pas(dev, umem, PAGE_SHIFT,
+					       start_page_index, npages, pas,
+					       MLX5_IB_MTT_PRESENT);
+			/* Clear padding after the pages brought from the
+			 * umem. */
+			memset(pas + npages, 0, size - npages * sizeof(u64));
+		}
+
+		dma_sync_single_for_device(ddev, dma, size, DMA_TO_DEVICE);
+
+		memset(&wr, 0, sizeof(wr));
+		wr.wr_id = (u64)(unsigned long)&umr_context;
+
+		sg.addr = dma;
+		sg.length = ALIGN(npages * sizeof(u64),
+				MLX5_UMR_MTT_ALIGNMENT);
+		sg.lkey = dev->umrc.mr->lkey;
+
+		wr.send_flags = MLX5_IB_SEND_UMR_FAIL_IF_FREE |
+				MLX5_IB_SEND_UMR_UPDATE_MTT;
+		wr.sg_list = &sg;
+		wr.num_sge = 1;
+		wr.opcode = MLX5_IB_WR_UMR;
+		umrwr->npages = sg.length / sizeof(u64);
+		umrwr->page_shift = PAGE_SHIFT;
+		umrwr->mkey = mr->mmr.key;
+		umrwr->target.offset = start_page_index;
+
+		mlx5_ib_init_umr_context(&umr_context);
+		down(&umrc->sem);
+		err = ib_post_send(umrc->qp, &wr, &bad);
+		if (err) {
+			mlx5_ib_err(dev, "UMR post send failed, err %d\n", err);
+		} else {
+			wait_for_completion(&umr_context.done);
+			if (umr_context.status != IB_WC_SUCCESS) {
+				mlx5_ib_err(dev, "UMR completion failed, code %d\n",
+					    umr_context.status);
+				err = -EFAULT;
+			}
+		}
+		up(&umrc->sem);
+	}
+	dma_unmap_single(ddev, dma, size, DMA_TO_DEVICE);
+
+free_pas:
+	if (!use_emergency_buf)
+		free_page((unsigned long)pas);
+	else
+		mutex_unlock(&mlx5_ib_update_mtt_emergency_buffer_mutex);
+
+	return err;
+}
+#endif
+
 static struct mlx5_ib_mr *reg_create(struct ib_pd *pd, u64 virt_addr,
 				     u64 length, struct ib_umem *umem,
 				     int npages, int page_shift,
@@ -822,6 +975,8 @@ static struct mlx5_ib_mr *reg_create(struct ib_pd *pd, u64 virt_addr,
 	struct mlx5_ib_mr *mr;
 	int inlen;
 	int err;
+	bool pg_cap = !!(dev->mdev->caps.gen.flags &
+			 MLX5_DEV_CAP_FLAG_ON_DMND_PG);
 
 	mr = kzalloc(sizeof(*mr), GFP_KERNEL);
 	if (!mr)
@@ -833,8 +988,12 @@ static struct mlx5_ib_mr *reg_create(struct ib_pd *pd, u64 virt_addr,
 		err = -ENOMEM;
 		goto err_1;
 	}
-	mlx5_ib_populate_pas(dev, umem, page_shift, in->pas, 0);
+	mlx5_ib_populate_pas(dev, umem, page_shift, in->pas,
+			     pg_cap ? MLX5_IB_MTT_PRESENT : 0);
 
+	/* The MLX5_MKEY_INBOX_PG_ACCESS bit allows setting the access flags
+	 * in the page list submitted with the command. */
+	in->flags = pg_cap ? cpu_to_be32(MLX5_MKEY_INBOX_PG_ACCESS) : 0;
 	in->seg.flags = convert_access(access_flags) |
 		MLX5_ACCESS_MODE_MTT;
 	in->seg.flags_pd = cpu_to_be32(to_mpd(pd)->pdn);
@@ -853,14 +1012,15 @@ static struct mlx5_ib_mr *reg_create(struct ib_pd *pd, u64 virt_addr,
 		goto err_2;
 	}
 	mr->umem = umem;
-	mlx5_vfree(in);
+	mr->live = 1;
+	kvfree(in);
 
 	mlx5_ib_dbg(dev, "mkey = 0x%x\n", mr->mmr.key);
 
 	return mr;
 
 err_2:
-	mlx5_vfree(in);
+	kvfree(in);
 
 err_1:
 	kfree(mr);
@@ -881,12 +1041,12 @@ struct ib_mr *mlx5_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 	int order;
 	int err;
 
-	mlx5_ib_dbg(dev, "start 0x%llx, virt_addr 0x%llx, length 0x%llx\n",
-		    start, virt_addr, length);
+	mlx5_ib_dbg(dev, "start 0x%llx, virt_addr 0x%llx, length 0x%llx, access_flags 0x%x\n",
+		    start, virt_addr, length, access_flags);
 	umem = ib_umem_get(pd->uobject->context, start, length, access_flags,
 			   0);
 	if (IS_ERR(umem)) {
-		mlx5_ib_dbg(dev, "umem get failed\n");
+		mlx5_ib_dbg(dev, "umem get failed (%ld)\n", PTR_ERR(umem));
 		return (void *)umem;
 	}
 
@@ -907,6 +1067,10 @@ struct ib_mr *mlx5_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 			mlx5_ib_dbg(dev, "cache empty for order %d", order);
 			mr = NULL;
 		}
+	} else if (access_flags & IB_ACCESS_ON_DEMAND) {
+		err = -EINVAL;
+		pr_err("Got MR registration for ODP MR > 512MB, not supported for Connect-IB");
+		goto error;
 	}
 
 	if (!mr)
@@ -922,16 +1086,51 @@ struct ib_mr *mlx5_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 
 	mr->umem = umem;
 	mr->npages = npages;
-	spin_lock(&dev->mr_lock);
-	dev->mdev->priv.reg_pages += npages;
-	spin_unlock(&dev->mr_lock);
+	atomic_add(npages, &dev->mdev->priv.reg_pages);
 	mr->ibmr.lkey = mr->mmr.key;
 	mr->ibmr.rkey = mr->mmr.key;
 
+#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
+	if (umem->odp_data) {
+		/*
+		 * This barrier prevents the compiler from moving the
+		 * setting of umem->odp_data->private to point to our
+		 * MR, before reg_umr finished, to ensure that the MR
+		 * initialization have finished before starting to
+		 * handle invalidations.
+		 */
+		smp_wmb();
+		mr->umem->odp_data->private = mr;
+		/*
+		 * Make sure we will see the new
+		 * umem->odp_data->private value in the invalidation
+		 * routines, before we can get page faults on the
+		 * MR. Page faults can happen once we put the MR in
+		 * the tree, below this line. Without the barrier,
+		 * there can be a fault handling and an invalidation
+		 * before umem->odp_data->private == mr is visible to
+		 * the invalidation handler.
+		 */
+		smp_wmb();
+	}
+#endif
+
 	return &mr->ibmr;
 
 error:
+	/*
+	 * Destroy the umem *before* destroying the MR, to ensure we
+	 * will not have any in-flight notifiers when destroying the
+	 * MR.
+	 *
+	 * As the MR is completely invalid to begin with, and this
+	 * error path is only taken if we can't push the mr entry into
+	 * the pagefault tree, this is safe.
+	 */
+
 	ib_umem_release(umem);
+	/* Kill the MR, and return an error code. */
+	clean_mr(mr);
 	return ERR_PTR(err);
 }
 
@@ -968,17 +1167,14 @@ error:
 	return err;
 }
 
-int mlx5_ib_dereg_mr(struct ib_mr *ibmr)
+static int clean_mr(struct mlx5_ib_mr *mr)
 {
-	struct mlx5_ib_dev *dev = to_mdev(ibmr->device);
-	struct mlx5_ib_mr *mr = to_mmr(ibmr);
-	struct ib_umem *umem = mr->umem;
-	int npages = mr->npages;
+	struct mlx5_ib_dev *dev = to_mdev(mr->ibmr.device);
 	int umred = mr->umred;
 	int err;
 
 	if (!umred) {
-		err = mlx5_core_destroy_mkey(dev->mdev, &mr->mmr);
+		err = destroy_mkey(dev, mr);
 		if (err) {
 			mlx5_ib_warn(dev, "failed to destroy mkey 0x%x (%d)\n",
 				     mr->mmr.key, err);
@@ -993,15 +1189,47 @@ int mlx5_ib_dereg_mr(struct ib_mr *ibmr)
 		free_cached_mr(dev, mr);
 	}
 
-	if (umem) {
+	if (!umred)
+		kfree(mr);
+
+	return 0;
+}
+
+int mlx5_ib_dereg_mr(struct ib_mr *ibmr)
+{
+	struct mlx5_ib_dev *dev = to_mdev(ibmr->device);
+	struct mlx5_ib_mr *mr = to_mmr(ibmr);
+	int npages = mr->npages;
+	struct ib_umem *umem = mr->umem;
+
+#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
+	if (umem && umem->odp_data) {
+		/* Prevent new page faults from succeeding */
+		mr->live = 0;
+		/* Wait for all running page-fault handlers to finish. */
+		synchronize_srcu(&dev->mr_srcu);
+		/* Destroy all page mappings */
+		mlx5_ib_invalidate_range(umem, ib_umem_start(umem),
+					 ib_umem_end(umem));
+		/*
+		 * We kill the umem before the MR for ODP,
+		 * so that there will not be any invalidations in
+		 * flight, looking at the *mr struct.
+		 */
 		ib_umem_release(umem);
-		spin_lock(&dev->mr_lock);
-		dev->mdev->priv.reg_pages -= npages;
-		spin_unlock(&dev->mr_lock);
+		atomic_sub(npages, &dev->mdev->priv.reg_pages);
+
+		/* Avoid double-freeing the umem. */
+		umem = NULL;
 	}
+#endif
 
-	if (!umred)
-		kfree(mr);
+	clean_mr(mr);
+
+	if (umem) {
+		ib_umem_release(umem);
+		atomic_sub(npages, &dev->mdev->priv.reg_pages);
+	}
 
 	return 0;
 }
@@ -1025,7 +1253,7 @@ struct ib_mr *mlx5_ib_create_mr(struct ib_pd *pd,
 		goto err_free;
 	}
 
-	in->seg.status = 1 << 6; /* free */
+	in->seg.status = MLX5_MKEY_STATUS_FREE;
 	in->seg.xlt_oct_size = cpu_to_be32(ndescs);
 	in->seg.qpn_mkey7_0 = cpu_to_be32(0xffffff << 8);
 	in->seg.flags_pd = cpu_to_be32(to_mpd(pd)->pdn);
@@ -1110,7 +1338,7 @@ int mlx5_ib_destroy_mr(struct ib_mr *ibmr)
 		kfree(mr->sig);
 	}
 
-	err = mlx5_core_destroy_mkey(dev->mdev, &mr->mmr);
+	err = destroy_mkey(dev, mr);
 	if (err) {
 		mlx5_ib_warn(dev, "failed to destroy mkey 0x%x (%d)\n",
 			     mr->mmr.key, err);
@@ -1140,7 +1368,7 @@ struct ib_mr *mlx5_ib_alloc_fast_reg_mr(struct ib_pd *pd,
 		goto err_free;
 	}
 
-	in->seg.status = 1 << 6; /* free */
+	in->seg.status = MLX5_MKEY_STATUS_FREE;
 	in->seg.xlt_oct_size = cpu_to_be32((max_page_list_len + 1) / 2);
 	in->seg.qpn_mkey7_0 = cpu_to_be32(0xffffff << 8);
 	in->seg.flags = MLX5_PERM_UMR_EN | MLX5_ACCESS_MODE_MTT;
diff --git a/drivers/infiniband/hw/mlx5/odp.c b/drivers/infiniband/hw/mlx5/odp.c
new file mode 100644
index 000000000000..a2c541c4809a
--- /dev/null
+++ b/drivers/infiniband/hw/mlx5/odp.c
@@ -0,0 +1,798 @@
+/*
+ * Copyright (c) 2014 Mellanox Technologies. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include <rdma/ib_umem.h>
+#include <rdma/ib_umem_odp.h>
+
+#include "mlx5_ib.h"
+
+#define MAX_PREFETCH_LEN (4*1024*1024U)
+
+/* Timeout in ms to wait for an active mmu notifier to complete when handling
+ * a pagefault. */
+#define MMU_NOTIFIER_TIMEOUT 1000
+
+struct workqueue_struct *mlx5_ib_page_fault_wq;
+
+void mlx5_ib_invalidate_range(struct ib_umem *umem, unsigned long start,
+			      unsigned long end)
+{
+	struct mlx5_ib_mr *mr;
+	const u64 umr_block_mask = (MLX5_UMR_MTT_ALIGNMENT / sizeof(u64)) - 1;
+	u64 idx = 0, blk_start_idx = 0;
+	int in_block = 0;
+	u64 addr;
+
+	if (!umem || !umem->odp_data) {
+		pr_err("invalidation called on NULL umem or non-ODP umem\n");
+		return;
+	}
+
+	mr = umem->odp_data->private;
+
+	if (!mr || !mr->ibmr.pd)
+		return;
+
+	start = max_t(u64, ib_umem_start(umem), start);
+	end = min_t(u64, ib_umem_end(umem), end);
+
+	/*
+	 * Iteration one - zap the HW's MTTs. The notifiers_count ensures that
+	 * while we are doing the invalidation, no page fault will attempt to
+	 * overwrite the same MTTs.  Concurent invalidations might race us,
+	 * but they will write 0s as well, so no difference in the end result.
+	 */
+
+	for (addr = start; addr < end; addr += (u64)umem->page_size) {
+		idx = (addr - ib_umem_start(umem)) / PAGE_SIZE;
+		/*
+		 * Strive to write the MTTs in chunks, but avoid overwriting
+		 * non-existing MTTs. The huristic here can be improved to
+		 * estimate the cost of another UMR vs. the cost of bigger
+		 * UMR.
+		 */
+		if (umem->odp_data->dma_list[idx] &
+		    (ODP_READ_ALLOWED_BIT | ODP_WRITE_ALLOWED_BIT)) {
+			if (!in_block) {
+				blk_start_idx = idx;
+				in_block = 1;
+			}
+		} else {
+			u64 umr_offset = idx & umr_block_mask;
+
+			if (in_block && umr_offset == 0) {
+				mlx5_ib_update_mtt(mr, blk_start_idx,
+						   idx - blk_start_idx, 1);
+				in_block = 0;
+			}
+		}
+	}
+	if (in_block)
+		mlx5_ib_update_mtt(mr, blk_start_idx, idx - blk_start_idx + 1,
+				   1);
+
+	/*
+	 * We are now sure that the device will not access the
+	 * memory. We can safely unmap it, and mark it as dirty if
+	 * needed.
+	 */
+
+	ib_umem_odp_unmap_dma_pages(umem, start, end);
+}
+
+#define COPY_ODP_BIT_MLX_TO_IB(reg, ib_caps, field_name, bit_name) do {	\
+	if (be32_to_cpu(reg.field_name) & MLX5_ODP_SUPPORT_##bit_name)	\
+		ib_caps->field_name |= IB_ODP_SUPPORT_##bit_name;	\
+} while (0)
+
+int mlx5_ib_internal_query_odp_caps(struct mlx5_ib_dev *dev)
+{
+	int err;
+	struct mlx5_odp_caps hw_caps;
+	struct ib_odp_caps *caps = &dev->odp_caps;
+
+	memset(caps, 0, sizeof(*caps));
+
+	if (!(dev->mdev->caps.gen.flags & MLX5_DEV_CAP_FLAG_ON_DMND_PG))
+		return 0;
+
+	err = mlx5_query_odp_caps(dev->mdev, &hw_caps);
+	if (err)
+		goto out;
+
+	caps->general_caps = IB_ODP_SUPPORT;
+	COPY_ODP_BIT_MLX_TO_IB(hw_caps, caps, per_transport_caps.ud_odp_caps,
+			       SEND);
+	COPY_ODP_BIT_MLX_TO_IB(hw_caps, caps, per_transport_caps.rc_odp_caps,
+			       SEND);
+	COPY_ODP_BIT_MLX_TO_IB(hw_caps, caps, per_transport_caps.rc_odp_caps,
+			       RECV);
+	COPY_ODP_BIT_MLX_TO_IB(hw_caps, caps, per_transport_caps.rc_odp_caps,
+			       WRITE);
+	COPY_ODP_BIT_MLX_TO_IB(hw_caps, caps, per_transport_caps.rc_odp_caps,
+			       READ);
+
+out:
+	return err;
+}
+
+static struct mlx5_ib_mr *mlx5_ib_odp_find_mr_lkey(struct mlx5_ib_dev *dev,
+						   u32 key)
+{
+	u32 base_key = mlx5_base_mkey(key);
+	struct mlx5_core_mr *mmr = __mlx5_mr_lookup(dev->mdev, base_key);
+	struct mlx5_ib_mr *mr = container_of(mmr, struct mlx5_ib_mr, mmr);
+
+	if (!mmr || mmr->key != key || !mr->live)
+		return NULL;
+
+	return container_of(mmr, struct mlx5_ib_mr, mmr);
+}
+
+static void mlx5_ib_page_fault_resume(struct mlx5_ib_qp *qp,
+				      struct mlx5_ib_pfault *pfault,
+				      int error) {
+	struct mlx5_ib_dev *dev = to_mdev(qp->ibqp.pd->device);
+	int ret = mlx5_core_page_fault_resume(dev->mdev, qp->mqp.qpn,
+					      pfault->mpfault.flags,
+					      error);
+	if (ret)
+		pr_err("Failed to resolve the page fault on QP 0x%x\n",
+		       qp->mqp.qpn);
+}
+
+/*
+ * Handle a single data segment in a page-fault WQE.
+ *
+ * Returns number of pages retrieved on success. The caller will continue to
+ * the next data segment.
+ * Can return the following error codes:
+ * -EAGAIN to designate a temporary error. The caller will abort handling the
+ *  page fault and resolve it.
+ * -EFAULT when there's an error mapping the requested pages. The caller will
+ *  abort the page fault handling and possibly move the QP to an error state.
+ * On other errors the QP should also be closed with an error.
+ */
+static int pagefault_single_data_segment(struct mlx5_ib_qp *qp,
+					 struct mlx5_ib_pfault *pfault,
+					 u32 key, u64 io_virt, size_t bcnt,
+					 u32 *bytes_mapped)
+{
+	struct mlx5_ib_dev *mib_dev = to_mdev(qp->ibqp.pd->device);
+	int srcu_key;
+	unsigned int current_seq;
+	u64 start_idx;
+	int npages = 0, ret = 0;
+	struct mlx5_ib_mr *mr;
+	u64 access_mask = ODP_READ_ALLOWED_BIT;
+
+	srcu_key = srcu_read_lock(&mib_dev->mr_srcu);
+	mr = mlx5_ib_odp_find_mr_lkey(mib_dev, key);
+	/*
+	 * If we didn't find the MR, it means the MR was closed while we were
+	 * handling the ODP event. In this case we return -EFAULT so that the
+	 * QP will be closed.
+	 */
+	if (!mr || !mr->ibmr.pd) {
+		pr_err("Failed to find relevant mr for lkey=0x%06x, probably the MR was destroyed\n",
+		       key);
+		ret = -EFAULT;
+		goto srcu_unlock;
+	}
+	if (!mr->umem->odp_data) {
+		pr_debug("skipping non ODP MR (lkey=0x%06x) in page fault handler.\n",
+			 key);
+		if (bytes_mapped)
+			*bytes_mapped +=
+				(bcnt - pfault->mpfault.bytes_committed);
+		goto srcu_unlock;
+	}
+	if (mr->ibmr.pd != qp->ibqp.pd) {
+		pr_err("Page-fault with different PDs for QP and MR.\n");
+		ret = -EFAULT;
+		goto srcu_unlock;
+	}
+
+	current_seq = ACCESS_ONCE(mr->umem->odp_data->notifiers_seq);
+	/*
+	 * Ensure the sequence number is valid for some time before we call
+	 * gup.
+	 */
+	smp_rmb();
+
+	/*
+	 * Avoid branches - this code will perform correctly
+	 * in all iterations (in iteration 2 and above,
+	 * bytes_committed == 0).
+	 */
+	io_virt += pfault->mpfault.bytes_committed;
+	bcnt -= pfault->mpfault.bytes_committed;
+
+	start_idx = (io_virt - (mr->mmr.iova & PAGE_MASK)) >> PAGE_SHIFT;
+
+	if (mr->umem->writable)
+		access_mask |= ODP_WRITE_ALLOWED_BIT;
+	npages = ib_umem_odp_map_dma_pages(mr->umem, io_virt, bcnt,
+					   access_mask, current_seq);
+	if (npages < 0) {
+		ret = npages;
+		goto srcu_unlock;
+	}
+
+	if (npages > 0) {
+		mutex_lock(&mr->umem->odp_data->umem_mutex);
+		if (!ib_umem_mmu_notifier_retry(mr->umem, current_seq)) {
+			/*
+			 * No need to check whether the MTTs really belong to
+			 * this MR, since ib_umem_odp_map_dma_pages already
+			 * checks this.
+			 */
+			ret = mlx5_ib_update_mtt(mr, start_idx, npages, 0);
+		} else {
+			ret = -EAGAIN;
+		}
+		mutex_unlock(&mr->umem->odp_data->umem_mutex);
+		if (ret < 0) {
+			if (ret != -EAGAIN)
+				pr_err("Failed to update mkey page tables\n");
+			goto srcu_unlock;
+		}
+
+		if (bytes_mapped) {
+			u32 new_mappings = npages * PAGE_SIZE -
+				(io_virt - round_down(io_virt, PAGE_SIZE));
+			*bytes_mapped += min_t(u32, new_mappings, bcnt);
+		}
+	}
+
+srcu_unlock:
+	if (ret == -EAGAIN) {
+		if (!mr->umem->odp_data->dying) {
+			struct ib_umem_odp *odp_data = mr->umem->odp_data;
+			unsigned long timeout =
+				msecs_to_jiffies(MMU_NOTIFIER_TIMEOUT);
+
+			if (!wait_for_completion_timeout(
+					&odp_data->notifier_completion,
+					timeout)) {
+				pr_warn("timeout waiting for mmu notifier completion\n");
+			}
+		} else {
+			/* The MR is being killed, kill the QP as well. */
+			ret = -EFAULT;
+		}
+	}
+	srcu_read_unlock(&mib_dev->mr_srcu, srcu_key);
+	pfault->mpfault.bytes_committed = 0;
+	return ret ? ret : npages;
+}
+
+/**
+ * Parse a series of data segments for page fault handling.
+ *
+ * @qp the QP on which the fault occurred.
+ * @pfault contains page fault information.
+ * @wqe points at the first data segment in the WQE.
+ * @wqe_end points after the end of the WQE.
+ * @bytes_mapped receives the number of bytes that the function was able to
+ *               map. This allows the caller to decide intelligently whether
+ *               enough memory was mapped to resolve the page fault
+ *               successfully (e.g. enough for the next MTU, or the entire
+ *               WQE).
+ * @total_wqe_bytes receives the total data size of this WQE in bytes (minus
+ *                  the committed bytes).
+ *
+ * Returns the number of pages loaded if positive, zero for an empty WQE, or a
+ * negative error code.
+ */
+static int pagefault_data_segments(struct mlx5_ib_qp *qp,
+				   struct mlx5_ib_pfault *pfault, void *wqe,
+				   void *wqe_end, u32 *bytes_mapped,
+				   u32 *total_wqe_bytes, int receive_queue)
+{
+	int ret = 0, npages = 0;
+	u64 io_virt;
+	u32 key;
+	u32 byte_count;
+	size_t bcnt;
+	int inline_segment;
+
+	/* Skip SRQ next-WQE segment. */
+	if (receive_queue && qp->ibqp.srq)
+		wqe += sizeof(struct mlx5_wqe_srq_next_seg);
+
+	if (bytes_mapped)
+		*bytes_mapped = 0;
+	if (total_wqe_bytes)
+		*total_wqe_bytes = 0;
+
+	while (wqe < wqe_end) {
+		struct mlx5_wqe_data_seg *dseg = wqe;
+
+		io_virt = be64_to_cpu(dseg->addr);
+		key = be32_to_cpu(dseg->lkey);
+		byte_count = be32_to_cpu(dseg->byte_count);
+		inline_segment = !!(byte_count &  MLX5_INLINE_SEG);
+		bcnt	       = byte_count & ~MLX5_INLINE_SEG;
+
+		if (inline_segment) {
+			bcnt = bcnt & MLX5_WQE_INLINE_SEG_BYTE_COUNT_MASK;
+			wqe += ALIGN(sizeof(struct mlx5_wqe_inline_seg) + bcnt,
+				     16);
+		} else {
+			wqe += sizeof(*dseg);
+		}
+
+		/* receive WQE end of sg list. */
+		if (receive_queue && bcnt == 0 && key == MLX5_INVALID_LKEY &&
+		    io_virt == 0)
+			break;
+
+		if (!inline_segment && total_wqe_bytes) {
+			*total_wqe_bytes += bcnt - min_t(size_t, bcnt,
+					pfault->mpfault.bytes_committed);
+		}
+
+		/* A zero length data segment designates a length of 2GB. */
+		if (bcnt == 0)
+			bcnt = 1U << 31;
+
+		if (inline_segment || bcnt <= pfault->mpfault.bytes_committed) {
+			pfault->mpfault.bytes_committed -=
+				min_t(size_t, bcnt,
+				      pfault->mpfault.bytes_committed);
+			continue;
+		}
+
+		ret = pagefault_single_data_segment(qp, pfault, key, io_virt,
+						    bcnt, bytes_mapped);
+		if (ret < 0)
+			break;
+		npages += ret;
+	}
+
+	return ret < 0 ? ret : npages;
+}
+
+/*
+ * Parse initiator WQE. Advances the wqe pointer to point at the
+ * scatter-gather list, and set wqe_end to the end of the WQE.
+ */
+static int mlx5_ib_mr_initiator_pfault_handler(
+	struct mlx5_ib_qp *qp, struct mlx5_ib_pfault *pfault,
+	void **wqe, void **wqe_end, int wqe_length)
+{
+	struct mlx5_ib_dev *dev = to_mdev(qp->ibqp.pd->device);
+	struct mlx5_wqe_ctrl_seg *ctrl = *wqe;
+	u16 wqe_index = pfault->mpfault.wqe.wqe_index;
+	unsigned ds, opcode;
+#if defined(DEBUG)
+	u32 ctrl_wqe_index, ctrl_qpn;
+#endif
+
+	ds = be32_to_cpu(ctrl->qpn_ds) & MLX5_WQE_CTRL_DS_MASK;
+	if (ds * MLX5_WQE_DS_UNITS > wqe_length) {
+		mlx5_ib_err(dev, "Unable to read the complete WQE. ds = 0x%x, ret = 0x%x\n",
+			    ds, wqe_length);
+		return -EFAULT;
+	}
+
+	if (ds == 0) {
+		mlx5_ib_err(dev, "Got WQE with zero DS. wqe_index=%x, qpn=%x\n",
+			    wqe_index, qp->mqp.qpn);
+		return -EFAULT;
+	}
+
+#if defined(DEBUG)
+	ctrl_wqe_index = (be32_to_cpu(ctrl->opmod_idx_opcode) &
+			MLX5_WQE_CTRL_WQE_INDEX_MASK) >>
+			MLX5_WQE_CTRL_WQE_INDEX_SHIFT;
+	if (wqe_index != ctrl_wqe_index) {
+		mlx5_ib_err(dev, "Got WQE with invalid wqe_index. wqe_index=0x%x, qpn=0x%x ctrl->wqe_index=0x%x\n",
+			    wqe_index, qp->mqp.qpn,
+			    ctrl_wqe_index);
+		return -EFAULT;
+	}
+
+	ctrl_qpn = (be32_to_cpu(ctrl->qpn_ds) & MLX5_WQE_CTRL_QPN_MASK) >>
+		MLX5_WQE_CTRL_QPN_SHIFT;
+	if (qp->mqp.qpn != ctrl_qpn) {
+		mlx5_ib_err(dev, "Got WQE with incorrect QP number. wqe_index=0x%x, qpn=0x%x ctrl->qpn=0x%x\n",
+			    wqe_index, qp->mqp.qpn,
+			    ctrl_qpn);
+		return -EFAULT;
+	}
+#endif /* DEBUG */
+
+	*wqe_end = *wqe + ds * MLX5_WQE_DS_UNITS;
+	*wqe += sizeof(*ctrl);
+
+	opcode = be32_to_cpu(ctrl->opmod_idx_opcode) &
+		 MLX5_WQE_CTRL_OPCODE_MASK;
+	switch (qp->ibqp.qp_type) {
+	case IB_QPT_RC:
+		switch (opcode) {
+		case MLX5_OPCODE_SEND:
+		case MLX5_OPCODE_SEND_IMM:
+		case MLX5_OPCODE_SEND_INVAL:
+			if (!(dev->odp_caps.per_transport_caps.rc_odp_caps &
+			      IB_ODP_SUPPORT_SEND))
+				goto invalid_transport_or_opcode;
+			break;
+		case MLX5_OPCODE_RDMA_WRITE:
+		case MLX5_OPCODE_RDMA_WRITE_IMM:
+			if (!(dev->odp_caps.per_transport_caps.rc_odp_caps &
+			      IB_ODP_SUPPORT_WRITE))
+				goto invalid_transport_or_opcode;
+			*wqe += sizeof(struct mlx5_wqe_raddr_seg);
+			break;
+		case MLX5_OPCODE_RDMA_READ:
+			if (!(dev->odp_caps.per_transport_caps.rc_odp_caps &
+			      IB_ODP_SUPPORT_READ))
+				goto invalid_transport_or_opcode;
+			*wqe += sizeof(struct mlx5_wqe_raddr_seg);
+			break;
+		default:
+			goto invalid_transport_or_opcode;
+		}
+		break;
+	case IB_QPT_UD:
+		switch (opcode) {
+		case MLX5_OPCODE_SEND:
+		case MLX5_OPCODE_SEND_IMM:
+			if (!(dev->odp_caps.per_transport_caps.ud_odp_caps &
+			      IB_ODP_SUPPORT_SEND))
+				goto invalid_transport_or_opcode;
+			*wqe += sizeof(struct mlx5_wqe_datagram_seg);
+			break;
+		default:
+			goto invalid_transport_or_opcode;
+		}
+		break;
+	default:
+invalid_transport_or_opcode:
+		mlx5_ib_err(dev, "ODP fault on QP of an unsupported opcode or transport. transport: 0x%x opcode: 0x%x.\n",
+			    qp->ibqp.qp_type, opcode);
+		return -EFAULT;
+	}
+
+	return 0;
+}
+
+/*
+ * Parse responder WQE. Advances the wqe pointer to point at the
+ * scatter-gather list, and set wqe_end to the end of the WQE.
+ */
+static int mlx5_ib_mr_responder_pfault_handler(
+	struct mlx5_ib_qp *qp, struct mlx5_ib_pfault *pfault,
+	void **wqe, void **wqe_end, int wqe_length)
+{
+	struct mlx5_ib_dev *dev = to_mdev(qp->ibqp.pd->device);
+	struct mlx5_ib_wq *wq = &qp->rq;
+	int wqe_size = 1 << wq->wqe_shift;
+
+	if (qp->ibqp.srq) {
+		mlx5_ib_err(dev, "ODP fault on SRQ is not supported\n");
+		return -EFAULT;
+	}
+
+	if (qp->wq_sig) {
+		mlx5_ib_err(dev, "ODP fault with WQE signatures is not supported\n");
+		return -EFAULT;
+	}
+
+	if (wqe_size > wqe_length) {
+		mlx5_ib_err(dev, "Couldn't read all of the receive WQE's content\n");
+		return -EFAULT;
+	}
+
+	switch (qp->ibqp.qp_type) {
+	case IB_QPT_RC:
+		if (!(dev->odp_caps.per_transport_caps.rc_odp_caps &
+		      IB_ODP_SUPPORT_RECV))
+			goto invalid_transport_or_opcode;
+		break;
+	default:
+invalid_transport_or_opcode:
+		mlx5_ib_err(dev, "ODP fault on QP of an unsupported transport. transport: 0x%x\n",
+			    qp->ibqp.qp_type);
+		return -EFAULT;
+	}
+
+	*wqe_end = *wqe + wqe_size;
+
+	return 0;
+}
+
+static void mlx5_ib_mr_wqe_pfault_handler(struct mlx5_ib_qp *qp,
+					  struct mlx5_ib_pfault *pfault)
+{
+	struct mlx5_ib_dev *dev = to_mdev(qp->ibqp.pd->device);
+	int ret;
+	void *wqe, *wqe_end;
+	u32 bytes_mapped, total_wqe_bytes;
+	char *buffer = NULL;
+	int resume_with_error = 0;
+	u16 wqe_index = pfault->mpfault.wqe.wqe_index;
+	int requestor = pfault->mpfault.flags & MLX5_PFAULT_REQUESTOR;
+
+	buffer = (char *)__get_free_page(GFP_KERNEL);
+	if (!buffer) {
+		mlx5_ib_err(dev, "Error allocating memory for IO page fault handling.\n");
+		resume_with_error = 1;
+		goto resolve_page_fault;
+	}
+
+	ret = mlx5_ib_read_user_wqe(qp, requestor, wqe_index, buffer,
+				    PAGE_SIZE);
+	if (ret < 0) {
+		mlx5_ib_err(dev, "Failed reading a WQE following page fault, error=%x, wqe_index=%x, qpn=%x\n",
+			    -ret, wqe_index, qp->mqp.qpn);
+		resume_with_error = 1;
+		goto resolve_page_fault;
+	}
+
+	wqe = buffer;
+	if (requestor)
+		ret = mlx5_ib_mr_initiator_pfault_handler(qp, pfault, &wqe,
+							  &wqe_end, ret);
+	else
+		ret = mlx5_ib_mr_responder_pfault_handler(qp, pfault, &wqe,
+							  &wqe_end, ret);
+	if (ret < 0) {
+		resume_with_error = 1;
+		goto resolve_page_fault;
+	}
+
+	if (wqe >= wqe_end) {
+		mlx5_ib_err(dev, "ODP fault on invalid WQE.\n");
+		resume_with_error = 1;
+		goto resolve_page_fault;
+	}
+
+	ret = pagefault_data_segments(qp, pfault, wqe, wqe_end, &bytes_mapped,
+				      &total_wqe_bytes, !requestor);
+	if (ret == -EAGAIN) {
+		goto resolve_page_fault;
+	} else if (ret < 0 || total_wqe_bytes > bytes_mapped) {
+		mlx5_ib_err(dev, "Error getting user pages for page fault. Error: 0x%x\n",
+			    -ret);
+		resume_with_error = 1;
+		goto resolve_page_fault;
+	}
+
+resolve_page_fault:
+	mlx5_ib_page_fault_resume(qp, pfault, resume_with_error);
+	mlx5_ib_dbg(dev, "PAGE FAULT completed. QP 0x%x resume_with_error=%d, flags: 0x%x\n",
+		    qp->mqp.qpn, resume_with_error, pfault->mpfault.flags);
+
+	free_page((unsigned long)buffer);
+}
+
+static int pages_in_range(u64 address, u32 length)
+{
+	return (ALIGN(address + length, PAGE_SIZE) -
+		(address & PAGE_MASK)) >> PAGE_SHIFT;
+}
+
+static void mlx5_ib_mr_rdma_pfault_handler(struct mlx5_ib_qp *qp,
+					   struct mlx5_ib_pfault *pfault)
+{
+	struct mlx5_pagefault *mpfault = &pfault->mpfault;
+	u64 address;
+	u32 length;
+	u32 prefetch_len = mpfault->bytes_committed;
+	int prefetch_activated = 0;
+	u32 rkey = mpfault->rdma.r_key;
+	int ret;
+
+	/* The RDMA responder handler handles the page fault in two parts.
+	 * First it brings the necessary pages for the current packet
+	 * (and uses the pfault context), and then (after resuming the QP)
+	 * prefetches more pages. The second operation cannot use the pfault
+	 * context and therefore uses the dummy_pfault context allocated on
+	 * the stack */
+	struct mlx5_ib_pfault dummy_pfault = {};
+
+	dummy_pfault.mpfault.bytes_committed = 0;
+
+	mpfault->rdma.rdma_va += mpfault->bytes_committed;
+	mpfault->rdma.rdma_op_len -= min(mpfault->bytes_committed,
+					 mpfault->rdma.rdma_op_len);
+	mpfault->bytes_committed = 0;
+
+	address = mpfault->rdma.rdma_va;
+	length  = mpfault->rdma.rdma_op_len;
+
+	/* For some operations, the hardware cannot tell the exact message
+	 * length, and in those cases it reports zero. Use prefetch
+	 * logic. */
+	if (length == 0) {
+		prefetch_activated = 1;
+		length = mpfault->rdma.packet_size;
+		prefetch_len = min(MAX_PREFETCH_LEN, prefetch_len);
+	}
+
+	ret = pagefault_single_data_segment(qp, pfault, rkey, address, length,
+					    NULL);
+	if (ret == -EAGAIN) {
+		/* We're racing with an invalidation, don't prefetch */
+		prefetch_activated = 0;
+	} else if (ret < 0 || pages_in_range(address, length) > ret) {
+		mlx5_ib_page_fault_resume(qp, pfault, 1);
+		return;
+	}
+
+	mlx5_ib_page_fault_resume(qp, pfault, 0);
+
+	/* At this point, there might be a new pagefault already arriving in
+	 * the eq, switch to the dummy pagefault for the rest of the
+	 * processing. We're still OK with the objects being alive as the
+	 * work-queue is being fenced. */
+
+	if (prefetch_activated) {
+		ret = pagefault_single_data_segment(qp, &dummy_pfault, rkey,
+						    address,
+						    prefetch_len,
+						    NULL);
+		if (ret < 0) {
+			pr_warn("Prefetch failed (ret = %d, prefetch_activated = %d) for QPN %d, address: 0x%.16llx, length = 0x%.16x\n",
+				ret, prefetch_activated,
+				qp->ibqp.qp_num, address, prefetch_len);
+		}
+	}
+}
+
+void mlx5_ib_mr_pfault_handler(struct mlx5_ib_qp *qp,
+			       struct mlx5_ib_pfault *pfault)
+{
+	u8 event_subtype = pfault->mpfault.event_subtype;
+
+	switch (event_subtype) {
+	case MLX5_PFAULT_SUBTYPE_WQE:
+		mlx5_ib_mr_wqe_pfault_handler(qp, pfault);
+		break;
+	case MLX5_PFAULT_SUBTYPE_RDMA:
+		mlx5_ib_mr_rdma_pfault_handler(qp, pfault);
+		break;
+	default:
+		pr_warn("Invalid page fault event subtype: 0x%x\n",
+			event_subtype);
+		mlx5_ib_page_fault_resume(qp, pfault, 1);
+		break;
+	}
+}
+
+static void mlx5_ib_qp_pfault_action(struct work_struct *work)
+{
+	struct mlx5_ib_pfault *pfault = container_of(work,
+						     struct mlx5_ib_pfault,
+						     work);
+	enum mlx5_ib_pagefault_context context =
+		mlx5_ib_get_pagefault_context(&pfault->mpfault);
+	struct mlx5_ib_qp *qp = container_of(pfault, struct mlx5_ib_qp,
+					     pagefaults[context]);
+	mlx5_ib_mr_pfault_handler(qp, pfault);
+}
+
+void mlx5_ib_qp_disable_pagefaults(struct mlx5_ib_qp *qp)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&qp->disable_page_faults_lock, flags);
+	qp->disable_page_faults = 1;
+	spin_unlock_irqrestore(&qp->disable_page_faults_lock, flags);
+
+	/*
+	 * Note that at this point, we are guarenteed that no more
+	 * work queue elements will be posted to the work queue with
+	 * the QP we are closing.
+	 */
+	flush_workqueue(mlx5_ib_page_fault_wq);
+}
+
+void mlx5_ib_qp_enable_pagefaults(struct mlx5_ib_qp *qp)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&qp->disable_page_faults_lock, flags);
+	qp->disable_page_faults = 0;
+	spin_unlock_irqrestore(&qp->disable_page_faults_lock, flags);
+}
+
+static void mlx5_ib_pfault_handler(struct mlx5_core_qp *qp,
+				   struct mlx5_pagefault *pfault)
+{
+	/*
+	 * Note that we will only get one fault event per QP per context
+	 * (responder/initiator, read/write), until we resolve the page fault
+	 * with the mlx5_ib_page_fault_resume command. Since this function is
+	 * called from within the work element, there is no risk of missing
+	 * events.
+	 */
+	struct mlx5_ib_qp *mibqp = to_mibqp(qp);
+	enum mlx5_ib_pagefault_context context =
+		mlx5_ib_get_pagefault_context(pfault);
+	struct mlx5_ib_pfault *qp_pfault = &mibqp->pagefaults[context];
+
+	qp_pfault->mpfault = *pfault;
+
+	/* No need to stop interrupts here since we are in an interrupt */
+	spin_lock(&mibqp->disable_page_faults_lock);
+	if (!mibqp->disable_page_faults)
+		queue_work(mlx5_ib_page_fault_wq, &qp_pfault->work);
+	spin_unlock(&mibqp->disable_page_faults_lock);
+}
+
+void mlx5_ib_odp_create_qp(struct mlx5_ib_qp *qp)
+{
+	int i;
+
+	qp->disable_page_faults = 1;
+	spin_lock_init(&qp->disable_page_faults_lock);
+
+	qp->mqp.pfault_handler	= mlx5_ib_pfault_handler;
+
+	for (i = 0; i < MLX5_IB_PAGEFAULT_CONTEXTS; ++i)
+		INIT_WORK(&qp->pagefaults[i].work, mlx5_ib_qp_pfault_action);
+}
+
+int mlx5_ib_odp_init_one(struct mlx5_ib_dev *ibdev)
+{
+	int ret;
+
+	ret = init_srcu_struct(&ibdev->mr_srcu);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+void mlx5_ib_odp_remove_one(struct mlx5_ib_dev *ibdev)
+{
+	cleanup_srcu_struct(&ibdev->mr_srcu);
+}
+
+int __init mlx5_ib_odp_init(void)
+{
+	mlx5_ib_page_fault_wq =
+		create_singlethread_workqueue("mlx5_ib_page_faults");
+	if (!mlx5_ib_page_fault_wq)
+		return -ENOMEM;
+
+	return 0;
+}
+
+void mlx5_ib_odp_cleanup(void)
+{
+	destroy_workqueue(mlx5_ib_page_fault_wq);
+}
diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index 8c574b63d77b..be0cd358b080 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -70,15 +70,6 @@ static const u32 mlx5_ib_opcode[] = {
 	[MLX5_IB_WR_UMR]			= MLX5_OPCODE_UMR,
 };
 
-struct umr_wr {
-	u64				virt_addr;
-	struct ib_pd		       *pd;
-	unsigned int			page_shift;
-	unsigned int			npages;
-	u32				length;
-	int				access_flags;
-	u32				mkey;
-};
 
 static int is_qp0(enum ib_qp_type qp_type)
 {
@@ -110,6 +101,77 @@ void *mlx5_get_send_wqe(struct mlx5_ib_qp *qp, int n)
 	return get_wqe(qp, qp->sq.offset + (n << MLX5_IB_SQ_STRIDE));
 }
 
+/**
+ * mlx5_ib_read_user_wqe() - Copy a user-space WQE to kernel space.
+ *
+ * @qp: QP to copy from.
+ * @send: copy from the send queue when non-zero, use the receive queue
+ *	  otherwise.
+ * @wqe_index:  index to start copying from. For send work queues, the
+ *		wqe_index is in units of MLX5_SEND_WQE_BB.
+ *		For receive work queue, it is the number of work queue
+ *		element in the queue.
+ * @buffer: destination buffer.
+ * @length: maximum number of bytes to copy.
+ *
+ * Copies at least a single WQE, but may copy more data.
+ *
+ * Return: the number of bytes copied, or an error code.
+ */
+int mlx5_ib_read_user_wqe(struct mlx5_ib_qp *qp, int send, int wqe_index,
+			  void *buffer, u32 length)
+{
+	struct ib_device *ibdev = qp->ibqp.device;
+	struct mlx5_ib_dev *dev = to_mdev(ibdev);
+	struct mlx5_ib_wq *wq = send ? &qp->sq : &qp->rq;
+	size_t offset;
+	size_t wq_end;
+	struct ib_umem *umem = qp->umem;
+	u32 first_copy_length;
+	int wqe_length;
+	int ret;
+
+	if (wq->wqe_cnt == 0) {
+		mlx5_ib_dbg(dev, "mlx5_ib_read_user_wqe for a QP with wqe_cnt == 0. qp_type: 0x%x\n",
+			    qp->ibqp.qp_type);
+		return -EINVAL;
+	}
+
+	offset = wq->offset + ((wqe_index % wq->wqe_cnt) << wq->wqe_shift);
+	wq_end = wq->offset + (wq->wqe_cnt << wq->wqe_shift);
+
+	if (send && length < sizeof(struct mlx5_wqe_ctrl_seg))
+		return -EINVAL;
+
+	if (offset > umem->length ||
+	    (send && offset + sizeof(struct mlx5_wqe_ctrl_seg) > umem->length))
+		return -EINVAL;
+
+	first_copy_length = min_t(u32, offset + length, wq_end) - offset;
+	ret = ib_umem_copy_from(buffer, umem, offset, first_copy_length);
+	if (ret)
+		return ret;
+
+	if (send) {
+		struct mlx5_wqe_ctrl_seg *ctrl = buffer;
+		int ds = be32_to_cpu(ctrl->qpn_ds) & MLX5_WQE_CTRL_DS_MASK;
+
+		wqe_length = ds * MLX5_WQE_DS_UNITS;
+	} else {
+		wqe_length = 1 << wq->wqe_shift;
+	}
+
+	if (wqe_length <= first_copy_length)
+		return first_copy_length;
+
+	ret = ib_umem_copy_from(buffer + first_copy_length, umem, wq->offset,
+				wqe_length - first_copy_length);
+	if (ret)
+		return ret;
+
+	return wqe_length;
+}
+
 static void mlx5_ib_qp_event(struct mlx5_core_qp *qp, int type)
 {
 	struct ib_qp *ibqp = &to_mibqp(qp)->ibqp;
@@ -158,11 +220,13 @@ static void mlx5_ib_qp_event(struct mlx5_core_qp *qp, int type)
 static int set_rq_size(struct mlx5_ib_dev *dev, struct ib_qp_cap *cap,
 		       int has_rq, struct mlx5_ib_qp *qp, struct mlx5_ib_create_qp *ucmd)
 {
+	struct mlx5_general_caps *gen;
 	int wqe_size;
 	int wq_size;
 
+	gen = &dev->mdev->caps.gen;
 	/* Sanity check RQ size before proceeding */
-	if (cap->max_recv_wr  > dev->mdev->caps.max_wqes)
+	if (cap->max_recv_wr  > gen->max_wqes)
 		return -EINVAL;
 
 	if (!has_rq) {
@@ -182,10 +246,10 @@ static int set_rq_size(struct mlx5_ib_dev *dev, struct ib_qp_cap *cap,
 			wq_size = roundup_pow_of_two(cap->max_recv_wr) * wqe_size;
 			wq_size = max_t(int, wq_size, MLX5_SEND_WQE_BB);
 			qp->rq.wqe_cnt = wq_size / wqe_size;
-			if (wqe_size > dev->mdev->caps.max_rq_desc_sz) {
+			if (wqe_size > gen->max_rq_desc_sz) {
 				mlx5_ib_dbg(dev, "wqe_size %d, max %d\n",
 					    wqe_size,
-					    dev->mdev->caps.max_rq_desc_sz);
+					    gen->max_rq_desc_sz);
 				return -EINVAL;
 			}
 			qp->rq.wqe_shift = ilog2(wqe_size);
@@ -266,9 +330,11 @@ static int calc_send_wqe(struct ib_qp_init_attr *attr)
 static int calc_sq_size(struct mlx5_ib_dev *dev, struct ib_qp_init_attr *attr,
 			struct mlx5_ib_qp *qp)
 {
+	struct mlx5_general_caps *gen;
 	int wqe_size;
 	int wq_size;
 
+	gen = &dev->mdev->caps.gen;
 	if (!attr->cap.max_send_wr)
 		return 0;
 
@@ -277,9 +343,9 @@ static int calc_sq_size(struct mlx5_ib_dev *dev, struct ib_qp_init_attr *attr,
 	if (wqe_size < 0)
 		return wqe_size;
 
-	if (wqe_size > dev->mdev->caps.max_sq_desc_sz) {
+	if (wqe_size > gen->max_sq_desc_sz) {
 		mlx5_ib_dbg(dev, "wqe_size(%d) > max_sq_desc_sz(%d)\n",
-			    wqe_size, dev->mdev->caps.max_sq_desc_sz);
+			    wqe_size, gen->max_sq_desc_sz);
 		return -EINVAL;
 	}
 
@@ -292,9 +358,9 @@ static int calc_sq_size(struct mlx5_ib_dev *dev, struct ib_qp_init_attr *attr,
 
 	wq_size = roundup_pow_of_two(attr->cap.max_send_wr * wqe_size);
 	qp->sq.wqe_cnt = wq_size / MLX5_SEND_WQE_BB;
-	if (qp->sq.wqe_cnt > dev->mdev->caps.max_wqes) {
+	if (qp->sq.wqe_cnt > gen->max_wqes) {
 		mlx5_ib_dbg(dev, "wqe count(%d) exceeds limits(%d)\n",
-			    qp->sq.wqe_cnt, dev->mdev->caps.max_wqes);
+			    qp->sq.wqe_cnt, gen->max_wqes);
 		return -ENOMEM;
 	}
 	qp->sq.wqe_shift = ilog2(MLX5_SEND_WQE_BB);
@@ -309,11 +375,13 @@ static int set_user_buf_size(struct mlx5_ib_dev *dev,
 			    struct mlx5_ib_qp *qp,
 			    struct mlx5_ib_create_qp *ucmd)
 {
+	struct mlx5_general_caps *gen;
 	int desc_sz = 1 << qp->sq.wqe_shift;
 
-	if (desc_sz > dev->mdev->caps.max_sq_desc_sz) {
+	gen = &dev->mdev->caps.gen;
+	if (desc_sz > gen->max_sq_desc_sz) {
 		mlx5_ib_warn(dev, "desc_sz %d, max_sq_desc_sz %d\n",
-			     desc_sz, dev->mdev->caps.max_sq_desc_sz);
+			     desc_sz, gen->max_sq_desc_sz);
 		return -EINVAL;
 	}
 
@@ -325,9 +393,9 @@ static int set_user_buf_size(struct mlx5_ib_dev *dev,
 
 	qp->sq.wqe_cnt = ucmd->sq_wqe_count;
 
-	if (qp->sq.wqe_cnt > dev->mdev->caps.max_wqes) {
+	if (qp->sq.wqe_cnt > gen->max_wqes) {
 		mlx5_ib_warn(dev, "wqe_cnt %d, max_wqes %d\n",
-			     qp->sq.wqe_cnt, dev->mdev->caps.max_wqes);
+			     qp->sq.wqe_cnt, gen->max_wqes);
 		return -EINVAL;
 	}
 
@@ -641,7 +709,7 @@ err_unmap:
 	mlx5_ib_db_unmap_user(context, &qp->db);
 
 err_free:
-	mlx5_vfree(*in);
+	kvfree(*in);
 
 err_umem:
 	if (qp->umem)
@@ -755,7 +823,7 @@ err_wrid:
 	kfree(qp->rq.wrid);
 
 err_free:
-	mlx5_vfree(*in);
+	kvfree(*in);
 
 err_buf:
 	mlx5_buf_free(dev->mdev, &qp->buf);
@@ -803,16 +871,20 @@ static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd,
 	struct mlx5_ib_resources *devr = &dev->devr;
 	struct mlx5_ib_create_qp_resp resp;
 	struct mlx5_create_qp_mbox_in *in;
+	struct mlx5_general_caps *gen;
 	struct mlx5_ib_create_qp ucmd;
 	int inlen = sizeof(*in);
 	int err;
 
+	mlx5_ib_odp_create_qp(qp);
+
+	gen = &dev->mdev->caps.gen;
 	mutex_init(&qp->mutex);
 	spin_lock_init(&qp->sq.lock);
 	spin_lock_init(&qp->rq.lock);
 
 	if (init_attr->create_flags & IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK) {
-		if (!(dev->mdev->caps.flags & MLX5_DEV_CAP_FLAG_BLOCK_MCAST)) {
+		if (!(gen->flags & MLX5_DEV_CAP_FLAG_BLOCK_MCAST)) {
 			mlx5_ib_dbg(dev, "block multicast loopback isn't supported\n");
 			return -EINVAL;
 		} else {
@@ -851,9 +923,9 @@ static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd,
 				mlx5_ib_dbg(dev, "invalid rq params\n");
 				return -EINVAL;
 			}
-			if (ucmd.sq_wqe_count > dev->mdev->caps.max_wqes) {
+			if (ucmd.sq_wqe_count > gen->max_wqes) {
 				mlx5_ib_dbg(dev, "requested sq_wqe_count (%d) > max allowed (%d)\n",
-					    ucmd.sq_wqe_count, dev->mdev->caps.max_wqes);
+					    ucmd.sq_wqe_count, gen->max_wqes);
 				return -EINVAL;
 			}
 			err = create_user_qp(dev, pd, qp, udata, &in, &resp, &inlen);
@@ -963,7 +1035,7 @@ static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd,
 		goto err_create;
 	}
 
-	mlx5_vfree(in);
+	kvfree(in);
 	/* Hardware wants QPN written in big-endian order (after
 	 * shifting) for send doorbell.  Precompute this value to save
 	 * a little bit when posting sends.
@@ -980,7 +1052,7 @@ err_create:
 	else if (qp->create_type == MLX5_QP_KERNEL)
 		destroy_qp_kernel(dev, qp);
 
-	mlx5_vfree(in);
+	kvfree(in);
 	return err;
 }
 
@@ -1003,9 +1075,14 @@ static void mlx5_ib_lock_cqs(struct mlx5_ib_cq *send_cq, struct mlx5_ib_cq *recv
 			}
 		} else {
 			spin_lock_irq(&send_cq->lock);
+			__acquire(&recv_cq->lock);
 		}
 	} else if (recv_cq) {
 		spin_lock_irq(&recv_cq->lock);
+		__acquire(&send_cq->lock);
+	} else {
+		__acquire(&send_cq->lock);
+		__acquire(&recv_cq->lock);
 	}
 }
 
@@ -1025,10 +1102,15 @@ static void mlx5_ib_unlock_cqs(struct mlx5_ib_cq *send_cq, struct mlx5_ib_cq *re
 				spin_unlock_irq(&recv_cq->lock);
 			}
 		} else {
+			__release(&recv_cq->lock);
 			spin_unlock_irq(&send_cq->lock);
 		}
 	} else if (recv_cq) {
+		__release(&send_cq->lock);
 		spin_unlock_irq(&recv_cq->lock);
+	} else {
+		__release(&recv_cq->lock);
+		__release(&send_cq->lock);
 	}
 }
 
@@ -1080,11 +1162,13 @@ static void destroy_qp_common(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp)
 	in = kzalloc(sizeof(*in), GFP_KERNEL);
 	if (!in)
 		return;
-	if (qp->state != IB_QPS_RESET)
+	if (qp->state != IB_QPS_RESET) {
+		mlx5_ib_qp_disable_pagefaults(qp);
 		if (mlx5_core_qp_modify(dev->mdev, to_mlx5_state(qp->state),
 					MLX5_QP_STATE_RST, in, sizeof(*in), &qp->mqp))
 			mlx5_ib_warn(dev, "mlx5_ib: modify QP %06x to RESET failed\n",
 				     qp->mqp.qpn);
+	}
 
 	get_cqs(qp, &send_cq, &recv_cq);
 
@@ -1144,6 +1228,7 @@ struct ib_qp *mlx5_ib_create_qp(struct ib_pd *pd,
 				struct ib_qp_init_attr *init_attr,
 				struct ib_udata *udata)
 {
+	struct mlx5_general_caps *gen;
 	struct mlx5_ib_dev *dev;
 	struct mlx5_ib_qp *qp;
 	u16 xrcdn = 0;
@@ -1161,11 +1246,12 @@ struct ib_qp *mlx5_ib_create_qp(struct ib_pd *pd,
 		}
 		dev = to_mdev(to_mxrcd(init_attr->xrcd)->ibxrcd.device);
 	}
+	gen = &dev->mdev->caps.gen;
 
 	switch (init_attr->qp_type) {
 	case IB_QPT_XRC_TGT:
 	case IB_QPT_XRC_INI:
-		if (!(dev->mdev->caps.flags & MLX5_DEV_CAP_FLAG_XRC)) {
+		if (!(gen->flags & MLX5_DEV_CAP_FLAG_XRC)) {
 			mlx5_ib_dbg(dev, "XRC not supported\n");
 			return ERR_PTR(-ENOSYS);
 		}
@@ -1272,6 +1358,9 @@ enum {
 
 static int ib_rate_to_mlx5(struct mlx5_ib_dev *dev, u8 rate)
 {
+	struct mlx5_general_caps *gen;
+
+	gen = &dev->mdev->caps.gen;
 	if (rate == IB_RATE_PORT_CURRENT) {
 		return 0;
 	} else if (rate < IB_RATE_2_5_GBPS || rate > IB_RATE_300_GBPS) {
@@ -1279,7 +1368,7 @@ static int ib_rate_to_mlx5(struct mlx5_ib_dev *dev, u8 rate)
 	} else {
 		while (rate != IB_RATE_2_5_GBPS &&
 		       !(1 << (rate + MLX5_STAT_RATE_OFFSET) &
-			 dev->mdev->caps.stat_rate_support))
+			 gen->stat_rate_support))
 			--rate;
 	}
 
@@ -1290,8 +1379,10 @@ static int mlx5_set_path(struct mlx5_ib_dev *dev, const struct ib_ah_attr *ah,
 			 struct mlx5_qp_path *path, u8 port, int attr_mask,
 			 u32 path_flags, const struct ib_qp_attr *attr)
 {
+	struct mlx5_general_caps *gen;
 	int err;
 
+	gen = &dev->mdev->caps.gen;
 	path->fl = (path_flags & MLX5_PATH_FLAG_FL) ? 0x80 : 0;
 	path->free_ar = (path_flags & MLX5_PATH_FLAG_FREE_AR) ? 0x80 : 0;
 
@@ -1302,6 +1393,11 @@ static int mlx5_set_path(struct mlx5_ib_dev *dev, const struct ib_ah_attr *ah,
 	path->rlid	= cpu_to_be16(ah->dlid);
 
 	if (ah->ah_flags & IB_AH_GRH) {
+		if (ah->grh.sgid_index >= gen->port[port - 1].gid_table_len) {
+			pr_err(KERN_ERR "sgid_index (%u) too large. max is %d\n",
+			       ah->grh.sgid_index, gen->port[port - 1].gid_table_len);
+			return -EINVAL;
+		}
 		path->grh_mlid |= 1 << 7;
 		path->mgid_index = ah->grh.sgid_index;
 		path->hop_limit  = ah->grh.hop_limit;
@@ -1317,22 +1413,6 @@ static int mlx5_set_path(struct mlx5_ib_dev *dev, const struct ib_ah_attr *ah,
 	path->static_rate = err;
 	path->port = port;
 
-	if (ah->ah_flags & IB_AH_GRH) {
-		if (ah->grh.sgid_index >= dev->mdev->caps.port[port - 1].gid_table_len) {
-			pr_err(KERN_ERR "sgid_index (%u) too large. max is %d\n",
-			       ah->grh.sgid_index, dev->mdev->caps.port[port - 1].gid_table_len);
-			return -EINVAL;
-		}
-
-		path->grh_mlid |= 1 << 7;
-		path->mgid_index = ah->grh.sgid_index;
-		path->hop_limit  = ah->grh.hop_limit;
-		path->tclass_flowlabel =
-			cpu_to_be32((ah->grh.traffic_class << 20) |
-				    (ah->grh.flow_label));
-		memcpy(path->rgid, ah->grh.dgid.raw, 16);
-	}
-
 	if (attr_mask & IB_QP_TIMEOUT)
 		path->ackto_lt = attr->timeout << 3;
 
@@ -1492,6 +1572,7 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
 	struct mlx5_ib_qp *qp = to_mqp(ibqp);
 	struct mlx5_ib_cq *send_cq, *recv_cq;
 	struct mlx5_qp_context *context;
+	struct mlx5_general_caps *gen;
 	struct mlx5_modify_qp_mbox_in *in;
 	struct mlx5_ib_pd *pd;
 	enum mlx5_qp_state mlx5_cur, mlx5_new;
@@ -1500,6 +1581,7 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
 	int mlx5_st;
 	int err;
 
+	gen = &dev->mdev->caps.gen;
 	in = kzalloc(sizeof(*in), GFP_KERNEL);
 	if (!in)
 		return -ENOMEM;
@@ -1539,7 +1621,7 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
 			err = -EINVAL;
 			goto out;
 		}
-		context->mtu_msgmax = (attr->path_mtu << 5) | dev->mdev->caps.log_max_msg;
+		context->mtu_msgmax = (attr->path_mtu << 5) | gen->log_max_msg;
 	}
 
 	if (attr_mask & IB_QP_DEST_QPN)
@@ -1634,6 +1716,15 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
 	if (mlx5_st < 0)
 		goto out;
 
+	/* If moving to a reset or error state, we must disable page faults on
+	 * this QP and flush all current page faults. Otherwise a stale page
+	 * fault may attempt to work on this QP after it is reset and moved
+	 * again to RTS, and may cause the driver and the device to get out of
+	 * sync. */
+	if (cur_state != IB_QPS_RESET && cur_state != IB_QPS_ERR &&
+	    (new_state == IB_QPS_RESET || new_state == IB_QPS_ERR))
+		mlx5_ib_qp_disable_pagefaults(qp);
+
 	optpar = ib_mask_to_mlx5_opt(attr_mask);
 	optpar &= opt_mask[mlx5_cur][mlx5_new][mlx5_st];
 	in->optparam = cpu_to_be32(optpar);
@@ -1643,6 +1734,9 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
 	if (err)
 		goto out;
 
+	if (cur_state == IB_QPS_RESET && new_state == IB_QPS_INIT)
+		mlx5_ib_qp_enable_pagefaults(qp);
+
 	qp->state = new_state;
 
 	if (attr_mask & IB_QP_ACCESS_FLAGS)
@@ -1685,9 +1779,11 @@ int mlx5_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 	struct mlx5_ib_dev *dev = to_mdev(ibqp->device);
 	struct mlx5_ib_qp *qp = to_mqp(ibqp);
 	enum ib_qp_state cur_state, new_state;
+	struct mlx5_general_caps *gen;
 	int err = -EINVAL;
 	int port;
 
+	gen = &dev->mdev->caps.gen;
 	mutex_lock(&qp->mutex);
 
 	cur_state = attr_mask & IB_QP_CUR_STATE ? attr->cur_qp_state : qp->state;
@@ -1699,21 +1795,21 @@ int mlx5_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 		goto out;
 
 	if ((attr_mask & IB_QP_PORT) &&
-	    (attr->port_num == 0 || attr->port_num > dev->mdev->caps.num_ports))
+	    (attr->port_num == 0 || attr->port_num > gen->num_ports))
 		goto out;
 
 	if (attr_mask & IB_QP_PKEY_INDEX) {
 		port = attr_mask & IB_QP_PORT ? attr->port_num : qp->port;
-		if (attr->pkey_index >= dev->mdev->caps.port[port - 1].pkey_table_len)
+		if (attr->pkey_index >= gen->port[port - 1].pkey_table_len)
 			goto out;
 	}
 
 	if (attr_mask & IB_QP_MAX_QP_RD_ATOMIC &&
-	    attr->max_rd_atomic > dev->mdev->caps.max_ra_res_qp)
+	    attr->max_rd_atomic > (1 << gen->log_max_ra_res_qp))
 		goto out;
 
 	if (attr_mask & IB_QP_MAX_DEST_RD_ATOMIC &&
-	    attr->max_dest_rd_atomic > dev->mdev->caps.max_ra_req_qp)
+	    attr->max_dest_rd_atomic > (1 << gen->log_max_ra_req_qp))
 		goto out;
 
 	if (cur_state == new_state && cur_state == IB_QPS_RESET) {
@@ -1830,37 +1926,70 @@ static void set_frwr_umr_segment(struct mlx5_wqe_umr_ctrl_seg *umr,
 	umr->mkey_mask = frwr_mkey_mask();
 }
 
+static __be64 get_umr_reg_mr_mask(void)
+{
+	u64 result;
+
+	result = MLX5_MKEY_MASK_LEN		|
+		 MLX5_MKEY_MASK_PAGE_SIZE	|
+		 MLX5_MKEY_MASK_START_ADDR	|
+		 MLX5_MKEY_MASK_PD		|
+		 MLX5_MKEY_MASK_LR		|
+		 MLX5_MKEY_MASK_LW		|
+		 MLX5_MKEY_MASK_KEY		|
+		 MLX5_MKEY_MASK_RR		|
+		 MLX5_MKEY_MASK_RW		|
+		 MLX5_MKEY_MASK_A		|
+		 MLX5_MKEY_MASK_FREE;
+
+	return cpu_to_be64(result);
+}
+
+static __be64 get_umr_unreg_mr_mask(void)
+{
+	u64 result;
+
+	result = MLX5_MKEY_MASK_FREE;
+
+	return cpu_to_be64(result);
+}
+
+static __be64 get_umr_update_mtt_mask(void)
+{
+	u64 result;
+
+	result = MLX5_MKEY_MASK_FREE;
+
+	return cpu_to_be64(result);
+}
+
 static void set_reg_umr_segment(struct mlx5_wqe_umr_ctrl_seg *umr,
 				struct ib_send_wr *wr)
 {
-	struct umr_wr *umrwr = (struct umr_wr *)&wr->wr.fast_reg;
-	u64 mask;
+	struct mlx5_umr_wr *umrwr = (struct mlx5_umr_wr *)&wr->wr.fast_reg;
 
 	memset(umr, 0, sizeof(*umr));
 
+	if (wr->send_flags & MLX5_IB_SEND_UMR_FAIL_IF_FREE)
+		umr->flags = MLX5_UMR_CHECK_FREE; /* fail if free */
+	else
+		umr->flags = MLX5_UMR_CHECK_NOT_FREE; /* fail if not free */
+
 	if (!(wr->send_flags & MLX5_IB_SEND_UMR_UNREG)) {
-		umr->flags = 1 << 5; /* fail if not free */
 		umr->klm_octowords = get_klm_octo(umrwr->npages);
-		mask =  MLX5_MKEY_MASK_LEN		|
-			MLX5_MKEY_MASK_PAGE_SIZE	|
-			MLX5_MKEY_MASK_START_ADDR	|
-			MLX5_MKEY_MASK_PD		|
-			MLX5_MKEY_MASK_LR		|
-			MLX5_MKEY_MASK_LW		|
-			MLX5_MKEY_MASK_KEY		|
-			MLX5_MKEY_MASK_RR		|
-			MLX5_MKEY_MASK_RW		|
-			MLX5_MKEY_MASK_A		|
-			MLX5_MKEY_MASK_FREE;
-		umr->mkey_mask = cpu_to_be64(mask);
+		if (wr->send_flags & MLX5_IB_SEND_UMR_UPDATE_MTT) {
+			umr->mkey_mask = get_umr_update_mtt_mask();
+			umr->bsf_octowords = get_klm_octo(umrwr->target.offset);
+			umr->flags |= MLX5_UMR_TRANSLATION_OFFSET_EN;
+		} else {
+			umr->mkey_mask = get_umr_reg_mr_mask();
+		}
 	} else {
-		umr->flags = 2 << 5; /* fail if free */
-		mask = MLX5_MKEY_MASK_FREE;
-		umr->mkey_mask = cpu_to_be64(mask);
+		umr->mkey_mask = get_umr_unreg_mr_mask();
 	}
 
 	if (!wr->num_sge)
-		umr->flags |= (1 << 7); /* inline */
+		umr->flags |= MLX5_UMR_INLINE;
 }
 
 static u8 get_umr_flags(int acc)
@@ -1877,7 +2006,7 @@ static void set_mkey_segment(struct mlx5_mkey_seg *seg, struct ib_send_wr *wr,
 {
 	memset(seg, 0, sizeof(*seg));
 	if (li) {
-		seg->status = 1 << 6;
+		seg->status = MLX5_MKEY_STATUS_FREE;
 		return;
 	}
 
@@ -1894,19 +2023,23 @@ static void set_mkey_segment(struct mlx5_mkey_seg *seg, struct ib_send_wr *wr,
 
 static void set_reg_mkey_segment(struct mlx5_mkey_seg *seg, struct ib_send_wr *wr)
 {
+	struct mlx5_umr_wr *umrwr = (struct mlx5_umr_wr *)&wr->wr.fast_reg;
+
 	memset(seg, 0, sizeof(*seg));
 	if (wr->send_flags & MLX5_IB_SEND_UMR_UNREG) {
-		seg->status = 1 << 6;
+		seg->status = MLX5_MKEY_STATUS_FREE;
 		return;
 	}
 
-	seg->flags = convert_access(wr->wr.fast_reg.access_flags);
-	seg->flags_pd = cpu_to_be32(to_mpd((struct ib_pd *)wr->wr.fast_reg.page_list)->pdn);
-	seg->start_addr = cpu_to_be64(wr->wr.fast_reg.iova_start);
-	seg->len = cpu_to_be64(wr->wr.fast_reg.length);
-	seg->log2_page_size = wr->wr.fast_reg.page_shift;
+	seg->flags = convert_access(umrwr->access_flags);
+	if (!(wr->send_flags & MLX5_IB_SEND_UMR_UPDATE_MTT)) {
+		seg->flags_pd = cpu_to_be32(to_mpd(umrwr->pd)->pdn);
+		seg->start_addr = cpu_to_be64(umrwr->target.virt_addr);
+	}
+	seg->len = cpu_to_be64(umrwr->length);
+	seg->log2_page_size = umrwr->page_shift;
 	seg->qpn_mkey7_0 = cpu_to_be32(0xffffff00 |
-				       mlx5_mkey_variant(wr->wr.fast_reg.rkey));
+				       mlx5_mkey_variant(umrwr->mkey));
 }
 
 static void set_frwr_pages(struct mlx5_wqe_data_seg *dseg,
@@ -2020,56 +2153,31 @@ static u8 bs_selector(int block_size)
 	}
 }
 
-static int format_selector(struct ib_sig_attrs *attr,
-			   struct ib_sig_domain *domain,
-			   int *selector)
+static void mlx5_fill_inl_bsf(struct ib_sig_domain *domain,
+			      struct mlx5_bsf_inl *inl)
 {
+	/* Valid inline section and allow BSF refresh */
+	inl->vld_refresh = cpu_to_be16(MLX5_BSF_INL_VALID |
+				       MLX5_BSF_REFRESH_DIF);
+	inl->dif_apptag = cpu_to_be16(domain->sig.dif.app_tag);
+	inl->dif_reftag = cpu_to_be32(domain->sig.dif.ref_tag);
+	/* repeating block */
+	inl->rp_inv_seed = MLX5_BSF_REPEAT_BLOCK;
+	inl->sig_type = domain->sig.dif.bg_type == IB_T10DIF_CRC ?
+			MLX5_DIF_CRC : MLX5_DIF_IPCS;
 
-#define FORMAT_DIF_NONE		0
-#define FORMAT_DIF_CRC_INC	8
-#define FORMAT_DIF_CRC_NO_INC	12
-#define FORMAT_DIF_CSUM_INC	13
-#define FORMAT_DIF_CSUM_NO_INC	14
+	if (domain->sig.dif.ref_remap)
+		inl->dif_inc_ref_guard_check |= MLX5_BSF_INC_REFTAG;
 
-	switch (domain->sig.dif.type) {
-	case IB_T10DIF_NONE:
-		/* No DIF */
-		*selector = FORMAT_DIF_NONE;
-		break;
-	case IB_T10DIF_TYPE1: /* Fall through */
-	case IB_T10DIF_TYPE2:
-		switch (domain->sig.dif.bg_type) {
-		case IB_T10DIF_CRC:
-			*selector = FORMAT_DIF_CRC_INC;
-			break;
-		case IB_T10DIF_CSUM:
-			*selector = FORMAT_DIF_CSUM_INC;
-			break;
-		default:
-			return 1;
-		}
-		break;
-	case IB_T10DIF_TYPE3:
-		switch (domain->sig.dif.bg_type) {
-		case IB_T10DIF_CRC:
-			*selector = domain->sig.dif.type3_inc_reftag ?
-					   FORMAT_DIF_CRC_INC :
-					   FORMAT_DIF_CRC_NO_INC;
-			break;
-		case IB_T10DIF_CSUM:
-			*selector = domain->sig.dif.type3_inc_reftag ?
-					   FORMAT_DIF_CSUM_INC :
-					   FORMAT_DIF_CSUM_NO_INC;
-			break;
-		default:
-			return 1;
-		}
-		break;
-	default:
-		return 1;
+	if (domain->sig.dif.app_escape) {
+		if (domain->sig.dif.ref_escape)
+			inl->dif_inc_ref_guard_check |= MLX5_BSF_APPREF_ESCAPE;
+		else
+			inl->dif_inc_ref_guard_check |= MLX5_BSF_APPTAG_ESCAPE;
 	}
 
-	return 0;
+	inl->dif_app_bitmask_check =
+		cpu_to_be16(domain->sig.dif.apptag_check_mask);
 }
 
 static int mlx5_set_bsf(struct ib_mr *sig_mr,
@@ -2080,45 +2188,49 @@ static int mlx5_set_bsf(struct ib_mr *sig_mr,
 	struct mlx5_bsf_basic *basic = &bsf->basic;
 	struct ib_sig_domain *mem = &sig_attrs->mem;
 	struct ib_sig_domain *wire = &sig_attrs->wire;
-	int ret, selector;
 
 	memset(bsf, 0, sizeof(*bsf));
+
+	/* Basic + Extended + Inline */
+	basic->bsf_size_sbs = 1 << 7;
+	/* Input domain check byte mask */
+	basic->check_byte_mask = sig_attrs->check_mask;
+	basic->raw_data_size = cpu_to_be32(data_size);
+
+	/* Memory domain */
 	switch (sig_attrs->mem.sig_type) {
+	case IB_SIG_TYPE_NONE:
+		break;
 	case IB_SIG_TYPE_T10_DIF:
-		if (sig_attrs->wire.sig_type != IB_SIG_TYPE_T10_DIF)
-			return -EINVAL;
+		basic->mem.bs_selector = bs_selector(mem->sig.dif.pi_interval);
+		basic->m_bfs_psv = cpu_to_be32(msig->psv_memory.psv_idx);
+		mlx5_fill_inl_bsf(mem, &bsf->m_inl);
+		break;
+	default:
+		return -EINVAL;
+	}
 
-		/* Input domain check byte mask */
-		basic->check_byte_mask = sig_attrs->check_mask;
+	/* Wire domain */
+	switch (sig_attrs->wire.sig_type) {
+	case IB_SIG_TYPE_NONE:
+		break;
+	case IB_SIG_TYPE_T10_DIF:
 		if (mem->sig.dif.pi_interval == wire->sig.dif.pi_interval &&
-		    mem->sig.dif.type == wire->sig.dif.type) {
+		    mem->sig_type == wire->sig_type) {
 			/* Same block structure */
-			basic->bsf_size_sbs = 1 << 4;
+			basic->bsf_size_sbs |= 1 << 4;
 			if (mem->sig.dif.bg_type == wire->sig.dif.bg_type)
-				basic->wire.copy_byte_mask |= 0xc0;
+				basic->wire.copy_byte_mask |= MLX5_CPY_GRD_MASK;
 			if (mem->sig.dif.app_tag == wire->sig.dif.app_tag)
-				basic->wire.copy_byte_mask |= 0x30;
+				basic->wire.copy_byte_mask |= MLX5_CPY_APP_MASK;
 			if (mem->sig.dif.ref_tag == wire->sig.dif.ref_tag)
-				basic->wire.copy_byte_mask |= 0x0f;
+				basic->wire.copy_byte_mask |= MLX5_CPY_REF_MASK;
 		} else
 			basic->wire.bs_selector = bs_selector(wire->sig.dif.pi_interval);
 
-		basic->mem.bs_selector = bs_selector(mem->sig.dif.pi_interval);
-		basic->raw_data_size = cpu_to_be32(data_size);
-
-		ret = format_selector(sig_attrs, mem, &selector);
-		if (ret)
-			return -EINVAL;
-		basic->m_bfs_psv = cpu_to_be32(selector << 24 |
-					       msig->psv_memory.psv_idx);
-
-		ret = format_selector(sig_attrs, wire, &selector);
-		if (ret)
-			return -EINVAL;
-		basic->w_bfs_psv = cpu_to_be32(selector << 24 |
-					       msig->psv_wire.psv_idx);
+		basic->w_bfs_psv = cpu_to_be32(msig->psv_wire.psv_idx);
+		mlx5_fill_inl_bsf(wire, &bsf->w_inl);
 		break;
-
 	default:
 		return -EINVAL;
 	}
@@ -2317,20 +2429,21 @@ static int set_psv_wr(struct ib_sig_domain *domain,
 	memset(psv_seg, 0, sizeof(*psv_seg));
 	psv_seg->psv_num = cpu_to_be32(psv_idx);
 	switch (domain->sig_type) {
+	case IB_SIG_TYPE_NONE:
+		break;
 	case IB_SIG_TYPE_T10_DIF:
 		psv_seg->transient_sig = cpu_to_be32(domain->sig.dif.bg << 16 |
 						     domain->sig.dif.app_tag);
 		psv_seg->ref_tag = cpu_to_be32(domain->sig.dif.ref_tag);
-
-		*seg += sizeof(*psv_seg);
-		*size += sizeof(*psv_seg) / 16;
 		break;
-
 	default:
 		pr_err("Bad signature type given.\n");
 		return 1;
 	}
 
+	*seg += sizeof(*psv_seg);
+	*size += sizeof(*psv_seg) / 16;
+
 	return 0;
 }
 
@@ -2423,7 +2536,7 @@ static u8 get_fence(u8 fence, struct ib_send_wr *wr)
 
 static int begin_wqe(struct mlx5_ib_qp *qp, void **seg,
 		     struct mlx5_wqe_ctrl_seg **ctrl,
-		     struct ib_send_wr *wr, int *idx,
+		     struct ib_send_wr *wr, unsigned *idx,
 		     int *size, int nreq)
 {
 	int err = 0;
@@ -2749,6 +2862,8 @@ out:
 
 		if (bf->need_lock)
 			spin_lock(&bf->lock);
+		else
+			__acquire(&bf->lock);
 
 		/* TBD enable WC */
 		if (0 && nreq == 1 && bf->uuarn && inl && size > 1 && size <= bf->buf_size / 16) {
@@ -2765,6 +2880,8 @@ out:
 		bf->offset ^= bf->buf_size;
 		if (bf->need_lock)
 			spin_unlock(&bf->lock);
+		else
+			__release(&bf->lock);
 	}
 
 	spin_unlock_irqrestore(&qp->sq.lock, flags);
@@ -2893,7 +3010,8 @@ static void to_ib_ah_attr(struct mlx5_ib_dev *ibdev, struct ib_ah_attr *ib_ah_at
 	memset(ib_ah_attr, 0, sizeof(*ib_ah_attr));
 	ib_ah_attr->port_num	  = path->port;
 
-	if (ib_ah_attr->port_num == 0 || ib_ah_attr->port_num > dev->caps.num_ports)
+	if (ib_ah_attr->port_num == 0 ||
+	    ib_ah_attr->port_num > dev->caps.gen.num_ports)
 		return;
 
 	ib_ah_attr->sl = path->sl & 0xf;
@@ -2924,6 +3042,14 @@ int mlx5_ib_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, int qp_attr
 	int mlx5_state;
 	int err = 0;
 
+#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
+	/*
+	 * Wait for any outstanding page faults, in case the user frees memory
+	 * based upon this query's result.
+	 */
+	flush_workqueue(mlx5_ib_page_fault_wq);
+#endif
+
 	mutex_lock(&qp->mutex);
 	outb = kzalloc(sizeof(*outb), GFP_KERNEL);
 	if (!outb) {
@@ -3011,10 +3137,12 @@ struct ib_xrcd *mlx5_ib_alloc_xrcd(struct ib_device *ibdev,
 					  struct ib_udata *udata)
 {
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
+	struct mlx5_general_caps *gen;
 	struct mlx5_ib_xrcd *xrcd;
 	int err;
 
-	if (!(dev->mdev->caps.flags & MLX5_DEV_CAP_FLAG_XRC))
+	gen = &dev->mdev->caps.gen;
+	if (!(gen->flags & MLX5_DEV_CAP_FLAG_XRC))
 		return ERR_PTR(-ENOSYS);
 
 	xrcd = kmalloc(sizeof(*xrcd), GFP_KERNEL);
diff --git a/drivers/infiniband/hw/mlx5/srq.c b/drivers/infiniband/hw/mlx5/srq.c
index 70bd131ba646..41fec66217dd 100644
--- a/drivers/infiniband/hw/mlx5/srq.c
+++ b/drivers/infiniband/hw/mlx5/srq.c
@@ -141,7 +141,7 @@ static int create_srq_user(struct ib_pd *pd, struct mlx5_ib_srq *srq,
 	return 0;
 
 err_in:
-	mlx5_vfree(*in);
+	kvfree(*in);
 
 err_umem:
 	ib_umem_release(srq->umem);
@@ -209,7 +209,7 @@ static int create_srq_kernel(struct mlx5_ib_dev *dev, struct mlx5_ib_srq *srq,
 	return 0;
 
 err_in:
-	mlx5_vfree(*in);
+	kvfree(*in);
 
 err_buf:
 	mlx5_buf_free(dev->mdev, &srq->buf);
@@ -238,6 +238,7 @@ struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd,
 				  struct ib_udata *udata)
 {
 	struct mlx5_ib_dev *dev = to_mdev(pd->device);
+	struct mlx5_general_caps *gen;
 	struct mlx5_ib_srq *srq;
 	int desc_size;
 	int buf_size;
@@ -247,11 +248,12 @@ struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd,
 	int is_xrc;
 	u32 flgs, xrcdn;
 
+	gen = &dev->mdev->caps.gen;
 	/* Sanity check SRQ size before proceeding */
-	if (init_attr->attr.max_wr >= dev->mdev->caps.max_srq_wqes) {
+	if (init_attr->attr.max_wr >= gen->max_srq_wqes) {
 		mlx5_ib_dbg(dev, "max_wr %d, cap %d\n",
 			    init_attr->attr.max_wr,
-			    dev->mdev->caps.max_srq_wqes);
+			    gen->max_srq_wqes);
 		return ERR_PTR(-EINVAL);
 	}
 
@@ -304,7 +306,7 @@ struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd,
 	in->ctx.pd = cpu_to_be32(to_mpd(pd)->pdn);
 	in->ctx.db_record = cpu_to_be64(srq->db.dma);
 	err = mlx5_core_create_srq(dev->mdev, &srq->msrq, in, inlen);
-	mlx5_vfree(in);
+	kvfree(in);
 	if (err) {
 		mlx5_ib_dbg(dev, "create SRQ failed, err %d\n", err);
 		goto err_usr_kern_srq;
diff --git a/drivers/infiniband/hw/nes/nes_verbs.c b/drivers/infiniband/hw/nes/nes_verbs.c
index fef067c959fc..c0d0296e7a00 100644
--- a/drivers/infiniband/hw/nes/nes_verbs.c
+++ b/drivers/infiniband/hw/nes/nes_verbs.c
@@ -2341,9 +2341,9 @@ static struct ib_mr *nes_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 	nes_debug(NES_DBG_MR, "User base = 0x%lX, Virt base = 0x%lX, length = %u,"
 			" offset = %u, page size = %u.\n",
 			(unsigned long int)start, (unsigned long int)virt, (u32)length,
-			region->offset, region->page_size);
+			ib_umem_offset(region), region->page_size);
 
-	skip_pages = ((u32)region->offset) >> 12;
+	skip_pages = ((u32)ib_umem_offset(region)) >> 12;
 
 	if (ib_copy_from_udata(&req, udata, sizeof(req))) {
 		ib_umem_release(region);
@@ -2408,7 +2408,7 @@ static struct ib_mr *nes_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 				region_length -= skip_pages << 12;
 				for (page_index = skip_pages; page_index < chunk_pages; page_index++) {
 					skip_pages = 0;
-					if ((page_count != 0) && (page_count<<12)-(region->offset&(4096-1)) >= region->length)
+					if ((page_count != 0) && (page_count << 12) - (ib_umem_offset(region) & (4096 - 1)) >= region->length)
 						goto enough_pages;
 					if ((page_count&0x01FF) == 0) {
 						if (page_count >= 1024 * 512) {
diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_ah.c b/drivers/infiniband/hw/ocrdma/ocrdma_ah.c
index ac02ce4e8040..f3cc8c9e65ae 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma_ah.c
+++ b/drivers/infiniband/hw/ocrdma/ocrdma_ah.c
@@ -96,7 +96,6 @@ struct ib_ah *ocrdma_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr)
 	struct ocrdma_pd *pd = get_ocrdma_pd(ibpd);
 	struct ocrdma_dev *dev = get_ocrdma_dev(ibpd->device);
 	union ib_gid sgid;
-	u8 zmac[ETH_ALEN];
 
 	if (!(attr->ah_flags & IB_AH_GRH))
 		return ERR_PTR(-EINVAL);
@@ -118,9 +117,7 @@ struct ib_ah *ocrdma_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr)
 		goto av_conf_err;
 	}
 
-	memset(&zmac, 0, ETH_ALEN);
-	if (pd->uctx &&
-	    memcmp(attr->dmac, &zmac, ETH_ALEN)) {
+	if (pd->uctx) {
 		status = rdma_addr_find_dmac_by_grh(&sgid, &attr->grh.dgid,
                                         attr->dmac, &attr->vlan_id);
 		if (status) {
diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_hw.c b/drivers/infiniband/hw/ocrdma/ocrdma_hw.c
index dd35ae558ae1..638bff1ffc6c 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma_hw.c
+++ b/drivers/infiniband/hw/ocrdma/ocrdma_hw.c
@@ -348,11 +348,6 @@ static void *ocrdma_init_emb_mqe(u8 opcode, u32 cmd_len)
 	return mqe;
 }
 
-static void *ocrdma_alloc_mqe(void)
-{
-	return kzalloc(sizeof(struct ocrdma_mqe), GFP_KERNEL);
-}
-
 static void ocrdma_free_q(struct ocrdma_dev *dev, struct ocrdma_queue_info *q)
 {
 	dma_free_coherent(&dev->nic_info.pdev->dev, q->size, q->va, q->dma);
@@ -566,8 +561,8 @@ static int ocrdma_mbx_create_mq(struct ocrdma_dev *dev,
 	cmd->cqid_pages |= (cq->id << OCRDMA_CREATE_MQ_CQ_ID_SHIFT);
 	cmd->async_cqid_valid = OCRDMA_CREATE_MQ_ASYNC_CQ_VALID;
 
-	cmd->async_event_bitmap = Bit(OCRDMA_ASYNC_GRP5_EVE_CODE);
-	cmd->async_event_bitmap |= Bit(OCRDMA_ASYNC_RDMA_EVE_CODE);
+	cmd->async_event_bitmap = BIT(OCRDMA_ASYNC_GRP5_EVE_CODE);
+	cmd->async_event_bitmap |= BIT(OCRDMA_ASYNC_RDMA_EVE_CODE);
 
 	cmd->async_cqid_ringsize = cq->id;
 	cmd->async_cqid_ringsize |= (ocrdma_encoded_q_len(mq->len) <<
@@ -1189,10 +1184,10 @@ int ocrdma_mbx_rdma_stats(struct ocrdma_dev *dev, bool reset)
 {
 	struct ocrdma_rdma_stats_req *req = dev->stats_mem.va;
 	struct ocrdma_mqe *mqe = &dev->stats_mem.mqe;
-	struct ocrdma_rdma_stats_resp *old_stats = NULL;
+	struct ocrdma_rdma_stats_resp *old_stats;
 	int status;
 
-	old_stats = kzalloc(sizeof(*old_stats), GFP_KERNEL);
+	old_stats = kmalloc(sizeof(*old_stats), GFP_KERNEL);
 	if (old_stats == NULL)
 		return -ENOMEM;
 
@@ -1235,10 +1230,9 @@ static int ocrdma_mbx_get_ctrl_attribs(struct ocrdma_dev *dev)
 	struct ocrdma_get_ctrl_attribs_rsp *ctrl_attr_rsp;
 	struct mgmt_hba_attribs *hba_attribs;
 
-	mqe = ocrdma_alloc_mqe();
+	mqe = kzalloc(sizeof(struct ocrdma_mqe), GFP_KERNEL);
 	if (!mqe)
 		return status;
-	memset(mqe, 0, sizeof(*mqe));
 
 	dma.size = sizeof(struct ocrdma_get_ctrl_attribs_rsp);
 	dma.va	 = dma_alloc_coherent(&dev->nic_info.pdev->dev,
@@ -2279,7 +2273,8 @@ mbx_err:
 
 static int ocrdma_set_av_params(struct ocrdma_qp *qp,
 				struct ocrdma_modify_qp *cmd,
-				struct ib_qp_attr *attrs)
+				struct ib_qp_attr *attrs,
+				int attr_mask)
 {
 	int status;
 	struct ib_ah_attr *ah_attr = &attrs->ah_attr;
@@ -2319,8 +2314,8 @@ static int ocrdma_set_av_params(struct ocrdma_qp *qp,
 	ocrdma_cpu_to_le32(&cmd->params.dgid[0], sizeof(cmd->params.dgid));
 	ocrdma_cpu_to_le32(&cmd->params.sgid[0], sizeof(cmd->params.sgid));
 	cmd->params.vlan_dmac_b4_to_b5 = mac_addr[4] | (mac_addr[5] << 8);
-	vlan_id = ah_attr->vlan_id;
-	if (vlan_id && (vlan_id < 0x1000)) {
+	if (attr_mask & IB_QP_VID) {
+		vlan_id = attrs->vlan_id;
 		cmd->params.vlan_dmac_b4_to_b5 |=
 		    vlan_id << OCRDMA_QP_PARAMS_VLAN_SHIFT;
 		cmd->flags |= OCRDMA_QP_PARA_VLAN_EN_VALID;
@@ -2347,7 +2342,7 @@ static int ocrdma_set_qp_params(struct ocrdma_qp *qp,
 		cmd->flags |= OCRDMA_QP_PARA_QKEY_VALID;
 	}
 	if (attr_mask & IB_QP_AV) {
-		status = ocrdma_set_av_params(qp, cmd, attrs);
+		status = ocrdma_set_av_params(qp, cmd, attrs, attr_mask);
 		if (status)
 			return status;
 	} else if (qp->qp_type == IB_QPT_GSI || qp->qp_type == IB_QPT_UD) {
diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_main.c b/drivers/infiniband/hw/ocrdma/ocrdma_main.c
index 256a06bc0b68..b0b2257b8e04 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma_main.c
+++ b/drivers/infiniband/hw/ocrdma/ocrdma_main.c
@@ -388,6 +388,15 @@ static void ocrdma_remove_sysfiles(struct ocrdma_dev *dev)
 		device_remove_file(&dev->ibdev.dev, ocrdma_attributes[i]);
 }
 
+static void ocrdma_add_default_sgid(struct ocrdma_dev *dev)
+{
+	/* GID Index 0 - Invariant manufacturer-assigned EUI-64 */
+	union ib_gid *sgid = &dev->sgid_tbl[0];
+
+	sgid->global.subnet_prefix = cpu_to_be64(0xfe80000000000000LL);
+	ocrdma_get_guid(dev, &sgid->raw[8]);
+}
+
 static void ocrdma_init_ipv4_gids(struct ocrdma_dev *dev,
 				  struct net_device *net)
 {
@@ -434,6 +443,7 @@ static void ocrdma_init_gid_table(struct ocrdma_dev *dev)
 				rdma_vlan_dev_real_dev(net_dev) : net_dev;
 
 		if (real_dev == dev->nic_info.netdev) {
+			ocrdma_add_default_sgid(dev);
 			ocrdma_init_ipv4_gids(dev, net_dev);
 			ocrdma_init_ipv6_gids(dev, net_dev);
 		}
@@ -646,8 +656,10 @@ static int __init ocrdma_init_module(void)
 	return 0;
 
 err_be_reg:
+#if IS_ENABLED(CONFIG_IPV6)
 	ocrdma_unregister_inet6addr_notifier();
 err_notifier6:
+#endif
 	ocrdma_unregister_inetaddr_notifier();
 	return status;
 }
diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_sli.h b/drivers/infiniband/hw/ocrdma/ocrdma_sli.h
index 904989ec5eaa..4e036480c1a8 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma_sli.h
+++ b/drivers/infiniband/hw/ocrdma/ocrdma_sli.h
@@ -28,8 +28,6 @@
 #ifndef __OCRDMA_SLI_H__
 #define __OCRDMA_SLI_H__
 
-#define Bit(_b) (1 << (_b))
-
 enum {
 	OCRDMA_ASIC_GEN_SKH_R = 0x04,
 	OCRDMA_ASIC_GEN_LANCER = 0x0B
@@ -103,7 +101,7 @@ enum {
 	QTYPE_MCCQ	= 3
 };
 
-#define OCRDMA_MAX_SGID (8)
+#define OCRDMA_MAX_SGID		8
 
 #define OCRDMA_MAX_QP    2048
 #define OCRDMA_MAX_CQ    2048
@@ -128,33 +126,33 @@ enum {
 #define OCRDMA_DB_CQ_RING_ID_EXT_MASK  0x0C00	/* bits 10-11 of qid at 12-11 */
 /* qid #2 msbits at 12-11 */
 #define OCRDMA_DB_CQ_RING_ID_EXT_MASK_SHIFT  0x1
-#define OCRDMA_DB_CQ_NUM_POPPED_SHIFT       (16)	/* bits 16 - 28 */
+#define OCRDMA_DB_CQ_NUM_POPPED_SHIFT	16	/* bits 16 - 28 */
 /* Rearm bit */
-#define OCRDMA_DB_CQ_REARM_SHIFT        (29)	/* bit 29 */
+#define OCRDMA_DB_CQ_REARM_SHIFT	29	/* bit 29 */
 /* solicited bit */
-#define OCRDMA_DB_CQ_SOLICIT_SHIFT   (31)	/* bit 31 */
+#define OCRDMA_DB_CQ_SOLICIT_SHIFT	31	/* bit 31 */
 
 #define OCRDMA_EQ_ID_MASK		0x1FF	/* bits 0 - 8 */
 #define OCRDMA_EQ_ID_EXT_MASK		0x3e00	/* bits 9-13 */
-#define OCRDMA_EQ_ID_EXT_MASK_SHIFT	(2)	/* qid bits 9-13 at 11-15 */
+#define OCRDMA_EQ_ID_EXT_MASK_SHIFT	2	/* qid bits 9-13 at 11-15 */
 
 /* Clear the interrupt for this eq */
-#define OCRDMA_EQ_CLR_SHIFT			(9)	/* bit 9 */
+#define OCRDMA_EQ_CLR_SHIFT		9	/* bit 9 */
 /* Must be 1 */
-#define OCRDMA_EQ_TYPE_SHIFT		(10)	/* bit 10 */
+#define OCRDMA_EQ_TYPE_SHIFT		10	/* bit 10 */
 /* Number of event entries processed */
-#define OCRDMA_NUM_EQE_SHIFT		(16)	/* bits 16 - 28 */
+#define OCRDMA_NUM_EQE_SHIFT		16	/* bits 16 - 28 */
 /* Rearm bit */
-#define OCRDMA_REARM_SHIFT		(29)	/* bit 29 */
+#define OCRDMA_REARM_SHIFT		29	/* bit 29 */
 
 #define OCRDMA_MQ_ID_MASK		0x7FF	/* bits 0 - 10 */
 /* Number of entries posted */
-#define OCRDMA_MQ_NUM_MQE_SHIFT	(16)	/* bits 16 - 29 */
+#define OCRDMA_MQ_NUM_MQE_SHIFT	16	/* bits 16 - 29 */
 
-#define OCRDMA_MIN_HPAGE_SIZE (4096)
+#define OCRDMA_MIN_HPAGE_SIZE	4096
 
-#define OCRDMA_MIN_Q_PAGE_SIZE (4096)
-#define OCRDMA_MAX_Q_PAGES     (8)
+#define OCRDMA_MIN_Q_PAGE_SIZE	4096
+#define OCRDMA_MAX_Q_PAGES	8
 
 #define OCRDMA_SLI_ASIC_ID_OFFSET	0x9C
 #define OCRDMA_SLI_ASIC_REV_MASK	0x000000FF
@@ -170,14 +168,14 @@ enum {
 # 6: 256K Bytes
 # 7: 512K Bytes
 */
-#define OCRDMA_MAX_Q_PAGE_SIZE_CNT (8)
+#define OCRDMA_MAX_Q_PAGE_SIZE_CNT	8
 #define OCRDMA_Q_PAGE_BASE_SIZE (OCRDMA_MIN_Q_PAGE_SIZE * OCRDMA_MAX_Q_PAGES)
 
-#define MAX_OCRDMA_QP_PAGES      (8)
+#define MAX_OCRDMA_QP_PAGES		8
 #define OCRDMA_MAX_WQE_MEM_SIZE (MAX_OCRDMA_QP_PAGES * OCRDMA_MIN_HQ_PAGE_SIZE)
 
-#define OCRDMA_CREATE_CQ_MAX_PAGES (4)
-#define OCRDMA_DPP_CQE_SIZE (4)
+#define OCRDMA_CREATE_CQ_MAX_PAGES	4
+#define OCRDMA_DPP_CQE_SIZE		4
 
 #define OCRDMA_GEN2_MAX_CQE 1024
 #define OCRDMA_GEN2_CQ_PAGE_SIZE 4096
@@ -238,7 +236,7 @@ struct ocrdma_mqe_sge {
 
 enum {
 	OCRDMA_MQE_HDR_EMB_SHIFT	= 0,
-	OCRDMA_MQE_HDR_EMB_MASK		= Bit(0),
+	OCRDMA_MQE_HDR_EMB_MASK		= BIT(0),
 	OCRDMA_MQE_HDR_SGE_CNT_SHIFT	= 3,
 	OCRDMA_MQE_HDR_SGE_CNT_MASK	= 0x1F << OCRDMA_MQE_HDR_SGE_CNT_SHIFT,
 	OCRDMA_MQE_HDR_SPECIAL_SHIFT	= 24,
@@ -292,7 +290,7 @@ struct ocrdma_pa {
 	u32 hi;
 };
 
-#define MAX_OCRDMA_EQ_PAGES (8)
+#define MAX_OCRDMA_EQ_PAGES	8
 struct ocrdma_create_eq_req {
 	struct ocrdma_mbx_hdr req;
 	u32 num_pages;
@@ -304,7 +302,7 @@ struct ocrdma_create_eq_req {
 };
 
 enum {
-	OCRDMA_CREATE_EQ_VALID	= Bit(29),
+	OCRDMA_CREATE_EQ_VALID	= BIT(29),
 	OCRDMA_CREATE_EQ_CNT_SHIFT	= 26,
 	OCRDMA_CREATE_CQ_DELAY_SHIFT	= 13,
 };
@@ -314,7 +312,7 @@ struct ocrdma_create_eq_rsp {
 	u32 vector_eqid;
 };
 
-#define OCRDMA_EQ_MINOR_OTHER (0x1)
+#define OCRDMA_EQ_MINOR_OTHER	0x1
 
 enum {
 	OCRDMA_MCQE_STATUS_SHIFT	= 0,
@@ -322,13 +320,13 @@ enum {
 	OCRDMA_MCQE_ESTATUS_SHIFT	= 16,
 	OCRDMA_MCQE_ESTATUS_MASK	= 0xFFFF << OCRDMA_MCQE_ESTATUS_SHIFT,
 	OCRDMA_MCQE_CONS_SHIFT		= 27,
-	OCRDMA_MCQE_CONS_MASK		= Bit(27),
+	OCRDMA_MCQE_CONS_MASK		= BIT(27),
 	OCRDMA_MCQE_CMPL_SHIFT		= 28,
-	OCRDMA_MCQE_CMPL_MASK		= Bit(28),
+	OCRDMA_MCQE_CMPL_MASK		= BIT(28),
 	OCRDMA_MCQE_AE_SHIFT		= 30,
-	OCRDMA_MCQE_AE_MASK		= Bit(30),
+	OCRDMA_MCQE_AE_MASK		= BIT(30),
 	OCRDMA_MCQE_VALID_SHIFT		= 31,
-	OCRDMA_MCQE_VALID_MASK		= Bit(31)
+	OCRDMA_MCQE_VALID_MASK		= BIT(31)
 };
 
 struct ocrdma_mcqe {
@@ -339,13 +337,13 @@ struct ocrdma_mcqe {
 };
 
 enum {
-	OCRDMA_AE_MCQE_QPVALID		= Bit(31),
+	OCRDMA_AE_MCQE_QPVALID		= BIT(31),
 	OCRDMA_AE_MCQE_QPID_MASK	= 0xFFFF,
 
-	OCRDMA_AE_MCQE_CQVALID		= Bit(31),
+	OCRDMA_AE_MCQE_CQVALID		= BIT(31),
 	OCRDMA_AE_MCQE_CQID_MASK	= 0xFFFF,
-	OCRDMA_AE_MCQE_VALID		= Bit(31),
-	OCRDMA_AE_MCQE_AE		= Bit(30),
+	OCRDMA_AE_MCQE_VALID		= BIT(31),
+	OCRDMA_AE_MCQE_AE		= BIT(30),
 	OCRDMA_AE_MCQE_EVENT_TYPE_SHIFT	= 16,
 	OCRDMA_AE_MCQE_EVENT_TYPE_MASK	=
 					0xFF << OCRDMA_AE_MCQE_EVENT_TYPE_SHIFT,
@@ -386,9 +384,9 @@ enum {
 	OCRDMA_AE_MPA_MCQE_EVENT_TYPE_MASK	= 0xFF <<
 					OCRDMA_AE_MPA_MCQE_EVENT_TYPE_SHIFT,
 	OCRDMA_AE_MPA_MCQE_EVENT_AE_SHIFT	= 30,
-	OCRDMA_AE_MPA_MCQE_EVENT_AE_MASK	= Bit(30),
+	OCRDMA_AE_MPA_MCQE_EVENT_AE_MASK	= BIT(30),
 	OCRDMA_AE_MPA_MCQE_EVENT_VALID_SHIFT	= 31,
-	OCRDMA_AE_MPA_MCQE_EVENT_VALID_MASK	= Bit(31)
+	OCRDMA_AE_MPA_MCQE_EVENT_VALID_MASK	= BIT(31)
 };
 
 struct ocrdma_ae_mpa_mcqe {
@@ -412,9 +410,9 @@ enum {
 	OCRDMA_AE_QP_MCQE_EVENT_TYPE_MASK	= 0xFF <<
 				OCRDMA_AE_QP_MCQE_EVENT_TYPE_SHIFT,
 	OCRDMA_AE_QP_MCQE_EVENT_AE_SHIFT	= 30,
-	OCRDMA_AE_QP_MCQE_EVENT_AE_MASK		= Bit(30),
+	OCRDMA_AE_QP_MCQE_EVENT_AE_MASK		= BIT(30),
 	OCRDMA_AE_QP_MCQE_EVENT_VALID_SHIFT	= 31,
-	OCRDMA_AE_QP_MCQE_EVENT_VALID_MASK	= Bit(31)
+	OCRDMA_AE_QP_MCQE_EVENT_VALID_MASK	= BIT(31)
 };
 
 struct ocrdma_ae_qp_mcqe {
@@ -449,9 +447,9 @@ enum OCRDMA_ASYNC_EVENT_TYPE {
 /* mailbox command request and responses */
 enum {
 	OCRDMA_MBX_QUERY_CFG_CQ_OVERFLOW_SHIFT		= 2,
-	OCRDMA_MBX_QUERY_CFG_CQ_OVERFLOW_MASK		= Bit(2),
+	OCRDMA_MBX_QUERY_CFG_CQ_OVERFLOW_MASK		= BIT(2),
 	OCRDMA_MBX_QUERY_CFG_SRQ_SUPPORTED_SHIFT	= 3,
-	OCRDMA_MBX_QUERY_CFG_SRQ_SUPPORTED_MASK		= Bit(3),
+	OCRDMA_MBX_QUERY_CFG_SRQ_SUPPORTED_MASK		= BIT(3),
 	OCRDMA_MBX_QUERY_CFG_MAX_QP_SHIFT		= 8,
 	OCRDMA_MBX_QUERY_CFG_MAX_QP_MASK		= 0xFFFFFF <<
 				OCRDMA_MBX_QUERY_CFG_MAX_QP_SHIFT,
@@ -672,9 +670,9 @@ enum {
 	OCRDMA_CREATE_CQ_PAGE_SIZE_MASK		= 0xFF,
 
 	OCRDMA_CREATE_CQ_COALESCWM_SHIFT	= 12,
-	OCRDMA_CREATE_CQ_COALESCWM_MASK		= Bit(13) | Bit(12),
-	OCRDMA_CREATE_CQ_FLAGS_NODELAY		= Bit(14),
-	OCRDMA_CREATE_CQ_FLAGS_AUTO_VALID	= Bit(15),
+	OCRDMA_CREATE_CQ_COALESCWM_MASK		= BIT(13) | BIT(12),
+	OCRDMA_CREATE_CQ_FLAGS_NODELAY		= BIT(14),
+	OCRDMA_CREATE_CQ_FLAGS_AUTO_VALID	= BIT(15),
 
 	OCRDMA_CREATE_CQ_EQ_ID_MASK		= 0xFFFF,
 	OCRDMA_CREATE_CQ_CQE_COUNT_MASK		= 0xFFFF
@@ -687,8 +685,8 @@ enum {
 	OCRDMA_CREATE_CQ_EQID_SHIFT		= 22,
 
 	OCRDMA_CREATE_CQ_CNT_SHIFT		= 27,
-	OCRDMA_CREATE_CQ_FLAGS_VALID		= Bit(29),
-	OCRDMA_CREATE_CQ_FLAGS_EVENTABLE	= Bit(31),
+	OCRDMA_CREATE_CQ_FLAGS_VALID		= BIT(29),
+	OCRDMA_CREATE_CQ_FLAGS_EVENTABLE	= BIT(31),
 	OCRDMA_CREATE_CQ_DEF_FLAGS		= OCRDMA_CREATE_CQ_FLAGS_VALID |
 					OCRDMA_CREATE_CQ_FLAGS_EVENTABLE |
 					OCRDMA_CREATE_CQ_FLAGS_NODELAY
@@ -731,8 +729,8 @@ enum {
 	OCRDMA_CREATE_MQ_V0_CQ_ID_SHIFT		= 22,
 	OCRDMA_CREATE_MQ_CQ_ID_SHIFT		= 16,
 	OCRDMA_CREATE_MQ_RING_SIZE_SHIFT	= 16,
-	OCRDMA_CREATE_MQ_VALID			= Bit(31),
-	OCRDMA_CREATE_MQ_ASYNC_CQ_VALID		= Bit(0)
+	OCRDMA_CREATE_MQ_VALID			= BIT(31),
+	OCRDMA_CREATE_MQ_ASYNC_CQ_VALID		= BIT(0)
 };
 
 struct ocrdma_create_mq_req {
@@ -783,7 +781,7 @@ enum {
 	OCRDMA_CREATE_QP_REQ_SQ_PAGE_SIZE_SHIFT	= 16,
 	OCRDMA_CREATE_QP_REQ_RQ_PAGE_SIZE_SHIFT	= 19,
 	OCRDMA_CREATE_QP_REQ_QPT_SHIFT		= 29,
-	OCRDMA_CREATE_QP_REQ_QPT_MASK		= Bit(31) | Bit(30) | Bit(29),
+	OCRDMA_CREATE_QP_REQ_QPT_MASK		= BIT(31) | BIT(30) | BIT(29),
 
 	OCRDMA_CREATE_QP_REQ_MAX_RQE_SHIFT	= 0,
 	OCRDMA_CREATE_QP_REQ_MAX_RQE_MASK	= 0xFFFF,
@@ -798,23 +796,23 @@ enum {
 					OCRDMA_CREATE_QP_REQ_MAX_SGE_SEND_SHIFT,
 
 	OCRDMA_CREATE_QP_REQ_FMR_EN_SHIFT		= 0,
-	OCRDMA_CREATE_QP_REQ_FMR_EN_MASK		= Bit(0),
+	OCRDMA_CREATE_QP_REQ_FMR_EN_MASK		= BIT(0),
 	OCRDMA_CREATE_QP_REQ_ZERO_LKEYEN_SHIFT		= 1,
-	OCRDMA_CREATE_QP_REQ_ZERO_LKEYEN_MASK		= Bit(1),
+	OCRDMA_CREATE_QP_REQ_ZERO_LKEYEN_MASK		= BIT(1),
 	OCRDMA_CREATE_QP_REQ_BIND_MEMWIN_SHIFT		= 2,
-	OCRDMA_CREATE_QP_REQ_BIND_MEMWIN_MASK		= Bit(2),
+	OCRDMA_CREATE_QP_REQ_BIND_MEMWIN_MASK		= BIT(2),
 	OCRDMA_CREATE_QP_REQ_INB_WREN_SHIFT		= 3,
-	OCRDMA_CREATE_QP_REQ_INB_WREN_MASK		= Bit(3),
+	OCRDMA_CREATE_QP_REQ_INB_WREN_MASK		= BIT(3),
 	OCRDMA_CREATE_QP_REQ_INB_RDEN_SHIFT		= 4,
-	OCRDMA_CREATE_QP_REQ_INB_RDEN_MASK		= Bit(4),
+	OCRDMA_CREATE_QP_REQ_INB_RDEN_MASK		= BIT(4),
 	OCRDMA_CREATE_QP_REQ_USE_SRQ_SHIFT		= 5,
-	OCRDMA_CREATE_QP_REQ_USE_SRQ_MASK		= Bit(5),
+	OCRDMA_CREATE_QP_REQ_USE_SRQ_MASK		= BIT(5),
 	OCRDMA_CREATE_QP_REQ_ENABLE_RPIR_SHIFT		= 6,
-	OCRDMA_CREATE_QP_REQ_ENABLE_RPIR_MASK		= Bit(6),
+	OCRDMA_CREATE_QP_REQ_ENABLE_RPIR_MASK		= BIT(6),
 	OCRDMA_CREATE_QP_REQ_ENABLE_DPP_SHIFT		= 7,
-	OCRDMA_CREATE_QP_REQ_ENABLE_DPP_MASK		= Bit(7),
+	OCRDMA_CREATE_QP_REQ_ENABLE_DPP_MASK		= BIT(7),
 	OCRDMA_CREATE_QP_REQ_ENABLE_DPP_CQ_SHIFT	= 8,
-	OCRDMA_CREATE_QP_REQ_ENABLE_DPP_CQ_MASK		= Bit(8),
+	OCRDMA_CREATE_QP_REQ_ENABLE_DPP_CQ_MASK		= BIT(8),
 	OCRDMA_CREATE_QP_REQ_MAX_SGE_RECV_SHIFT		= 16,
 	OCRDMA_CREATE_QP_REQ_MAX_SGE_RECV_MASK		= 0xFFFF <<
 				OCRDMA_CREATE_QP_REQ_MAX_SGE_RECV_SHIFT,
@@ -927,7 +925,7 @@ enum {
 	OCRDMA_CREATE_QP_RSP_SQ_ID_MASK			= 0xFFFF <<
 				OCRDMA_CREATE_QP_RSP_SQ_ID_SHIFT,
 
-	OCRDMA_CREATE_QP_RSP_DPP_ENABLED_MASK		= Bit(0),
+	OCRDMA_CREATE_QP_RSP_DPP_ENABLED_MASK		= BIT(0),
 	OCRDMA_CREATE_QP_RSP_DPP_PAGE_OFFSET_SHIFT	= 1,
 	OCRDMA_CREATE_QP_RSP_DPP_PAGE_OFFSET_MASK	= 0x7FFF <<
 				OCRDMA_CREATE_QP_RSP_DPP_PAGE_OFFSET_SHIFT,
@@ -964,38 +962,38 @@ enum {
 	OCRDMA_MODIFY_QP_ID_SHIFT	= 0,
 	OCRDMA_MODIFY_QP_ID_MASK	= 0xFFFF,
 
-	OCRDMA_QP_PARA_QPS_VALID	= Bit(0),
-	OCRDMA_QP_PARA_SQD_ASYNC_VALID	= Bit(1),
-	OCRDMA_QP_PARA_PKEY_VALID	= Bit(2),
-	OCRDMA_QP_PARA_QKEY_VALID	= Bit(3),
-	OCRDMA_QP_PARA_PMTU_VALID	= Bit(4),
-	OCRDMA_QP_PARA_ACK_TO_VALID	= Bit(5),
-	OCRDMA_QP_PARA_RETRY_CNT_VALID	= Bit(6),
-	OCRDMA_QP_PARA_RRC_VALID	= Bit(7),
-	OCRDMA_QP_PARA_RQPSN_VALID	= Bit(8),
-	OCRDMA_QP_PARA_MAX_IRD_VALID	= Bit(9),
-	OCRDMA_QP_PARA_MAX_ORD_VALID	= Bit(10),
-	OCRDMA_QP_PARA_RNT_VALID	= Bit(11),
-	OCRDMA_QP_PARA_SQPSN_VALID	= Bit(12),
-	OCRDMA_QP_PARA_DST_QPN_VALID	= Bit(13),
-	OCRDMA_QP_PARA_MAX_WQE_VALID	= Bit(14),
-	OCRDMA_QP_PARA_MAX_RQE_VALID	= Bit(15),
-	OCRDMA_QP_PARA_SGE_SEND_VALID	= Bit(16),
-	OCRDMA_QP_PARA_SGE_RECV_VALID	= Bit(17),
-	OCRDMA_QP_PARA_SGE_WR_VALID	= Bit(18),
-	OCRDMA_QP_PARA_INB_RDEN_VALID	= Bit(19),
-	OCRDMA_QP_PARA_INB_WREN_VALID	= Bit(20),
-	OCRDMA_QP_PARA_FLOW_LBL_VALID	= Bit(21),
-	OCRDMA_QP_PARA_BIND_EN_VALID	= Bit(22),
-	OCRDMA_QP_PARA_ZLKEY_EN_VALID	= Bit(23),
-	OCRDMA_QP_PARA_FMR_EN_VALID	= Bit(24),
-	OCRDMA_QP_PARA_INBAT_EN_VALID	= Bit(25),
-	OCRDMA_QP_PARA_VLAN_EN_VALID	= Bit(26),
-
-	OCRDMA_MODIFY_QP_FLAGS_RD	= Bit(0),
-	OCRDMA_MODIFY_QP_FLAGS_WR	= Bit(1),
-	OCRDMA_MODIFY_QP_FLAGS_SEND	= Bit(2),
-	OCRDMA_MODIFY_QP_FLAGS_ATOMIC	= Bit(3)
+	OCRDMA_QP_PARA_QPS_VALID	= BIT(0),
+	OCRDMA_QP_PARA_SQD_ASYNC_VALID	= BIT(1),
+	OCRDMA_QP_PARA_PKEY_VALID	= BIT(2),
+	OCRDMA_QP_PARA_QKEY_VALID	= BIT(3),
+	OCRDMA_QP_PARA_PMTU_VALID	= BIT(4),
+	OCRDMA_QP_PARA_ACK_TO_VALID	= BIT(5),
+	OCRDMA_QP_PARA_RETRY_CNT_VALID	= BIT(6),
+	OCRDMA_QP_PARA_RRC_VALID	= BIT(7),
+	OCRDMA_QP_PARA_RQPSN_VALID	= BIT(8),
+	OCRDMA_QP_PARA_MAX_IRD_VALID	= BIT(9),
+	OCRDMA_QP_PARA_MAX_ORD_VALID	= BIT(10),
+	OCRDMA_QP_PARA_RNT_VALID	= BIT(11),
+	OCRDMA_QP_PARA_SQPSN_VALID	= BIT(12),
+	OCRDMA_QP_PARA_DST_QPN_VALID	= BIT(13),
+	OCRDMA_QP_PARA_MAX_WQE_VALID	= BIT(14),
+	OCRDMA_QP_PARA_MAX_RQE_VALID	= BIT(15),
+	OCRDMA_QP_PARA_SGE_SEND_VALID	= BIT(16),
+	OCRDMA_QP_PARA_SGE_RECV_VALID	= BIT(17),
+	OCRDMA_QP_PARA_SGE_WR_VALID	= BIT(18),
+	OCRDMA_QP_PARA_INB_RDEN_VALID	= BIT(19),
+	OCRDMA_QP_PARA_INB_WREN_VALID	= BIT(20),
+	OCRDMA_QP_PARA_FLOW_LBL_VALID	= BIT(21),
+	OCRDMA_QP_PARA_BIND_EN_VALID	= BIT(22),
+	OCRDMA_QP_PARA_ZLKEY_EN_VALID	= BIT(23),
+	OCRDMA_QP_PARA_FMR_EN_VALID	= BIT(24),
+	OCRDMA_QP_PARA_INBAT_EN_VALID	= BIT(25),
+	OCRDMA_QP_PARA_VLAN_EN_VALID	= BIT(26),
+
+	OCRDMA_MODIFY_QP_FLAGS_RD	= BIT(0),
+	OCRDMA_MODIFY_QP_FLAGS_WR	= BIT(1),
+	OCRDMA_MODIFY_QP_FLAGS_SEND	= BIT(2),
+	OCRDMA_MODIFY_QP_FLAGS_ATOMIC	= BIT(3)
 };
 
 enum {
@@ -1014,15 +1012,15 @@ enum {
 	OCRDMA_QP_PARAMS_MAX_SGE_SEND_MASK	= 0xFFFF <<
 					OCRDMA_QP_PARAMS_MAX_SGE_SEND_SHIFT,
 
-	OCRDMA_QP_PARAMS_FLAGS_FMR_EN		= Bit(0),
-	OCRDMA_QP_PARAMS_FLAGS_LKEY_0_EN	= Bit(1),
-	OCRDMA_QP_PARAMS_FLAGS_BIND_MW_EN	= Bit(2),
-	OCRDMA_QP_PARAMS_FLAGS_INBWR_EN		= Bit(3),
-	OCRDMA_QP_PARAMS_FLAGS_INBRD_EN		= Bit(4),
+	OCRDMA_QP_PARAMS_FLAGS_FMR_EN		= BIT(0),
+	OCRDMA_QP_PARAMS_FLAGS_LKEY_0_EN	= BIT(1),
+	OCRDMA_QP_PARAMS_FLAGS_BIND_MW_EN	= BIT(2),
+	OCRDMA_QP_PARAMS_FLAGS_INBWR_EN		= BIT(3),
+	OCRDMA_QP_PARAMS_FLAGS_INBRD_EN		= BIT(4),
 	OCRDMA_QP_PARAMS_STATE_SHIFT		= 5,
-	OCRDMA_QP_PARAMS_STATE_MASK		= Bit(5) | Bit(6) | Bit(7),
-	OCRDMA_QP_PARAMS_FLAGS_SQD_ASYNC	= Bit(8),
-	OCRDMA_QP_PARAMS_FLAGS_INB_ATEN		= Bit(9),
+	OCRDMA_QP_PARAMS_STATE_MASK		= BIT(5) | BIT(6) | BIT(7),
+	OCRDMA_QP_PARAMS_FLAGS_SQD_ASYNC	= BIT(8),
+	OCRDMA_QP_PARAMS_FLAGS_INB_ATEN		= BIT(9),
 	OCRDMA_QP_PARAMS_MAX_SGE_RECV_SHIFT	= 16,
 	OCRDMA_QP_PARAMS_MAX_SGE_RECV_MASK	= 0xFFFF <<
 					OCRDMA_QP_PARAMS_MAX_SGE_RECV_SHIFT,
@@ -1277,7 +1275,7 @@ struct ocrdma_alloc_pd {
 };
 
 enum {
-	OCRDMA_ALLOC_PD_RSP_DPP			= Bit(16),
+	OCRDMA_ALLOC_PD_RSP_DPP			= BIT(16),
 	OCRDMA_ALLOC_PD_RSP_DPP_PAGE_SHIFT	= 20,
 	OCRDMA_ALLOC_PD_RSP_PDID_MASK		= 0xFFFF,
 };
@@ -1309,18 +1307,18 @@ enum {
 	OCRDMA_ALLOC_LKEY_PD_ID_MASK		= 0xFFFF,
 
 	OCRDMA_ALLOC_LKEY_ADDR_CHECK_SHIFT	= 0,
-	OCRDMA_ALLOC_LKEY_ADDR_CHECK_MASK	= Bit(0),
+	OCRDMA_ALLOC_LKEY_ADDR_CHECK_MASK	= BIT(0),
 	OCRDMA_ALLOC_LKEY_FMR_SHIFT		= 1,
-	OCRDMA_ALLOC_LKEY_FMR_MASK		= Bit(1),
+	OCRDMA_ALLOC_LKEY_FMR_MASK		= BIT(1),
 	OCRDMA_ALLOC_LKEY_REMOTE_INV_SHIFT	= 2,
-	OCRDMA_ALLOC_LKEY_REMOTE_INV_MASK	= Bit(2),
+	OCRDMA_ALLOC_LKEY_REMOTE_INV_MASK	= BIT(2),
 	OCRDMA_ALLOC_LKEY_REMOTE_WR_SHIFT	= 3,
-	OCRDMA_ALLOC_LKEY_REMOTE_WR_MASK	= Bit(3),
+	OCRDMA_ALLOC_LKEY_REMOTE_WR_MASK	= BIT(3),
 	OCRDMA_ALLOC_LKEY_REMOTE_RD_SHIFT	= 4,
-	OCRDMA_ALLOC_LKEY_REMOTE_RD_MASK	= Bit(4),
+	OCRDMA_ALLOC_LKEY_REMOTE_RD_MASK	= BIT(4),
 	OCRDMA_ALLOC_LKEY_LOCAL_WR_SHIFT	= 5,
-	OCRDMA_ALLOC_LKEY_LOCAL_WR_MASK		= Bit(5),
-	OCRDMA_ALLOC_LKEY_REMOTE_ATOMIC_MASK	= Bit(6),
+	OCRDMA_ALLOC_LKEY_LOCAL_WR_MASK		= BIT(5),
+	OCRDMA_ALLOC_LKEY_REMOTE_ATOMIC_MASK	= BIT(6),
 	OCRDMA_ALLOC_LKEY_REMOTE_ATOMIC_SHIFT	= 6,
 	OCRDMA_ALLOC_LKEY_PBL_SIZE_SHIFT	= 16,
 	OCRDMA_ALLOC_LKEY_PBL_SIZE_MASK		= 0xFFFF <<
@@ -1379,21 +1377,21 @@ enum {
 	OCRDMA_REG_NSMR_HPAGE_SIZE_MASK		= 0xFF <<
 					OCRDMA_REG_NSMR_HPAGE_SIZE_SHIFT,
 	OCRDMA_REG_NSMR_BIND_MEMWIN_SHIFT	= 24,
-	OCRDMA_REG_NSMR_BIND_MEMWIN_MASK	= Bit(24),
+	OCRDMA_REG_NSMR_BIND_MEMWIN_MASK	= BIT(24),
 	OCRDMA_REG_NSMR_ZB_SHIFT		= 25,
-	OCRDMA_REG_NSMR_ZB_SHIFT_MASK		= Bit(25),
+	OCRDMA_REG_NSMR_ZB_SHIFT_MASK		= BIT(25),
 	OCRDMA_REG_NSMR_REMOTE_INV_SHIFT	= 26,
-	OCRDMA_REG_NSMR_REMOTE_INV_MASK		= Bit(26),
+	OCRDMA_REG_NSMR_REMOTE_INV_MASK		= BIT(26),
 	OCRDMA_REG_NSMR_REMOTE_WR_SHIFT		= 27,
-	OCRDMA_REG_NSMR_REMOTE_WR_MASK		= Bit(27),
+	OCRDMA_REG_NSMR_REMOTE_WR_MASK		= BIT(27),
 	OCRDMA_REG_NSMR_REMOTE_RD_SHIFT		= 28,
-	OCRDMA_REG_NSMR_REMOTE_RD_MASK		= Bit(28),
+	OCRDMA_REG_NSMR_REMOTE_RD_MASK		= BIT(28),
 	OCRDMA_REG_NSMR_LOCAL_WR_SHIFT		= 29,
-	OCRDMA_REG_NSMR_LOCAL_WR_MASK		= Bit(29),
+	OCRDMA_REG_NSMR_LOCAL_WR_MASK		= BIT(29),
 	OCRDMA_REG_NSMR_REMOTE_ATOMIC_SHIFT	= 30,
-	OCRDMA_REG_NSMR_REMOTE_ATOMIC_MASK	= Bit(30),
+	OCRDMA_REG_NSMR_REMOTE_ATOMIC_MASK	= BIT(30),
 	OCRDMA_REG_NSMR_LAST_SHIFT		= 31,
-	OCRDMA_REG_NSMR_LAST_MASK		= Bit(31)
+	OCRDMA_REG_NSMR_LAST_MASK		= BIT(31)
 };
 
 struct ocrdma_reg_nsmr {
@@ -1420,7 +1418,7 @@ enum {
 					OCRDMA_REG_NSMR_CONT_NUM_PBL_SHIFT,
 
 	OCRDMA_REG_NSMR_CONT_LAST_SHIFT		= 31,
-	OCRDMA_REG_NSMR_CONT_LAST_MASK		= Bit(31)
+	OCRDMA_REG_NSMR_CONT_LAST_MASK		= BIT(31)
 };
 
 struct ocrdma_reg_nsmr_cont {
@@ -1566,7 +1564,7 @@ struct ocrdma_delete_ah_tbl_rsp {
 
 enum {
 	OCRDMA_EQE_VALID_SHIFT		= 0,
-	OCRDMA_EQE_VALID_MASK		= Bit(0),
+	OCRDMA_EQE_VALID_MASK		= BIT(0),
 	OCRDMA_EQE_FOR_CQE_MASK		= 0xFFFE,
 	OCRDMA_EQE_RESOURCE_ID_SHIFT	= 16,
 	OCRDMA_EQE_RESOURCE_ID_MASK	= 0xFFFF <<
@@ -1624,11 +1622,11 @@ enum {
 	OCRDMA_CQE_UD_STATUS_MASK	= 0x7 << OCRDMA_CQE_UD_STATUS_SHIFT,
 	OCRDMA_CQE_STATUS_SHIFT		= 16,
 	OCRDMA_CQE_STATUS_MASK		= 0xFF << OCRDMA_CQE_STATUS_SHIFT,
-	OCRDMA_CQE_VALID		= Bit(31),
-	OCRDMA_CQE_INVALIDATE		= Bit(30),
-	OCRDMA_CQE_QTYPE		= Bit(29),
-	OCRDMA_CQE_IMM			= Bit(28),
-	OCRDMA_CQE_WRITE_IMM		= Bit(27),
+	OCRDMA_CQE_VALID		= BIT(31),
+	OCRDMA_CQE_INVALIDATE		= BIT(30),
+	OCRDMA_CQE_QTYPE		= BIT(29),
+	OCRDMA_CQE_IMM			= BIT(28),
+	OCRDMA_CQE_WRITE_IMM		= BIT(27),
 	OCRDMA_CQE_QTYPE_SQ		= 0,
 	OCRDMA_CQE_QTYPE_RQ		= 1,
 	OCRDMA_CQE_SRCQP_MASK		= 0xFFFFFF
@@ -1772,8 +1770,8 @@ struct ocrdma_grh {
 	u16	rsvd;
 } __packed;
 
-#define OCRDMA_AV_VALID		Bit(7)
-#define OCRDMA_AV_VLAN_VALID	Bit(1)
+#define OCRDMA_AV_VALID		BIT(7)
+#define OCRDMA_AV_VLAN_VALID	BIT(1)
 
 struct ocrdma_av {
 	struct ocrdma_eth_vlan eth_hdr;
diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c
index 8f5f2577f288..fb8d8c4dfbb9 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c
+++ b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c
@@ -388,7 +388,7 @@ struct ib_ucontext *ocrdma_alloc_ucontext(struct ib_device *ibdev,
 
 	memset(&resp, 0, sizeof(resp));
 	resp.ah_tbl_len = ctx->ah_tbl.len;
-	resp.ah_tbl_page = ctx->ah_tbl.pa;
+	resp.ah_tbl_page = virt_to_phys(ctx->ah_tbl.va);
 
 	status = ocrdma_add_mmap(ctx, resp.ah_tbl_page, resp.ah_tbl_len);
 	if (status)
@@ -805,7 +805,7 @@ struct ib_mr *ocrdma_reg_user_mr(struct ib_pd *ibpd, u64 start, u64 len,
 		goto umem_err;
 
 	mr->hwmr.pbe_size = mr->umem->page_size;
-	mr->hwmr.fbo = mr->umem->offset;
+	mr->hwmr.fbo = ib_umem_offset(mr->umem);
 	mr->hwmr.va = usr_addr;
 	mr->hwmr.len = len;
 	mr->hwmr.remote_wr = (acc & IB_ACCESS_REMOTE_WRITE) ? 1 : 0;
@@ -870,7 +870,7 @@ static int ocrdma_copy_cq_uresp(struct ocrdma_dev *dev, struct ocrdma_cq *cq,
 	uresp.page_size = PAGE_ALIGN(cq->len);
 	uresp.num_pages = 1;
 	uresp.max_hw_cqe = cq->max_hw_cqe;
-	uresp.page_addr[0] = cq->pa;
+	uresp.page_addr[0] = virt_to_phys(cq->va);
 	uresp.db_page_addr =  ocrdma_get_db_addr(dev, uctx->cntxt_pd->id);
 	uresp.db_page_size = dev->nic_info.db_page_size;
 	uresp.phase_change = cq->phase_change ? 1 : 0;
@@ -1123,13 +1123,13 @@ static int ocrdma_copy_qp_uresp(struct ocrdma_qp *qp,
 	uresp.sq_dbid = qp->sq.dbid;
 	uresp.num_sq_pages = 1;
 	uresp.sq_page_size = PAGE_ALIGN(qp->sq.len);
-	uresp.sq_page_addr[0] = qp->sq.pa;
+	uresp.sq_page_addr[0] = virt_to_phys(qp->sq.va);
 	uresp.num_wqe_allocated = qp->sq.max_cnt;
 	if (!srq) {
 		uresp.rq_dbid = qp->rq.dbid;
 		uresp.num_rq_pages = 1;
 		uresp.rq_page_size = PAGE_ALIGN(qp->rq.len);
-		uresp.rq_page_addr[0] = qp->rq.pa;
+		uresp.rq_page_addr[0] = virt_to_phys(qp->rq.va);
 		uresp.num_rqe_allocated = qp->rq.max_cnt;
 	}
 	uresp.db_page_addr = usr_db;
@@ -1410,6 +1410,8 @@ int ocrdma_query_qp(struct ib_qp *ibqp,
 	mutex_unlock(&dev->dev_lock);
 	if (status)
 		goto mbx_err;
+	if (qp->qp_type == IB_QPT_UD)
+		qp_attr->qkey = params.qkey;
 	qp_attr->qp_state = get_ibqp_state(IB_QPS_INIT);
 	qp_attr->cur_qp_state = get_ibqp_state(IB_QPS_INIT);
 	qp_attr->path_mtu =
@@ -1680,7 +1682,7 @@ static int ocrdma_copy_srq_uresp(struct ocrdma_dev *dev, struct ocrdma_srq *srq,
 	memset(&uresp, 0, sizeof(uresp));
 	uresp.rq_dbid = srq->rq.dbid;
 	uresp.num_rq_pages = 1;
-	uresp.rq_page_addr[0] = srq->rq.pa;
+	uresp.rq_page_addr[0] = virt_to_phys(srq->rq.va);
 	uresp.rq_page_size = srq->rq.len;
 	uresp.db_page_addr = dev->nic_info.unmapped_db +
 	    (srq->pd->id * dev->nic_info.db_page_size);
@@ -1870,7 +1872,7 @@ static int ocrdma_build_inline_sges(struct ocrdma_qp *qp,
 		hdr->total_len = ocrdma_sglist_len(wr->sg_list, wr->num_sge);
 		if (unlikely(hdr->total_len > qp->max_inline_data)) {
 			pr_err("%s() supported_len=0x%x,\n"
-			       " unspported len req=0x%x\n", __func__,
+			       " unsupported len req=0x%x\n", __func__,
 				qp->max_inline_data, hdr->total_len);
 			return -EINVAL;
 		}
diff --git a/drivers/infiniband/hw/qib/qib_fs.c b/drivers/infiniband/hw/qib/qib_fs.c
index cab610ccd50e..81854586c081 100644
--- a/drivers/infiniband/hw/qib/qib_fs.c
+++ b/drivers/infiniband/hw/qib/qib_fs.c
@@ -89,7 +89,6 @@ static int create_file(const char *name, umode_t mode,
 {
 	int error;
 
-	*dentry = NULL;
 	mutex_lock(&parent->d_inode->i_mutex);
 	*dentry = lookup_one_len(name, parent, strlen(name));
 	if (!IS_ERR(*dentry))
diff --git a/drivers/infiniband/hw/qib/qib_mr.c b/drivers/infiniband/hw/qib/qib_mr.c
index 9bbb55347cc1..a77fb4fb14e4 100644
--- a/drivers/infiniband/hw/qib/qib_mr.c
+++ b/drivers/infiniband/hw/qib/qib_mr.c
@@ -258,7 +258,7 @@ struct ib_mr *qib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 	mr->mr.user_base = start;
 	mr->mr.iova = virt_addr;
 	mr->mr.length = length;
-	mr->mr.offset = umem->offset;
+	mr->mr.offset = ib_umem_offset(umem);
 	mr->mr.access_flags = mr_access_flags;
 	mr->umem = umem;
 
diff --git a/drivers/infiniband/hw/usnic/usnic_ib_qp_grp.c b/drivers/infiniband/hw/usnic/usnic_ib_qp_grp.c
index f8dfd76be89f..db3588df3546 100644
--- a/drivers/infiniband/hw/usnic/usnic_ib_qp_grp.c
+++ b/drivers/infiniband/hw/usnic/usnic_ib_qp_grp.c
@@ -511,7 +511,7 @@ int usnic_ib_qp_grp_modify(struct usnic_ib_qp_grp *qp_grp,
 		usnic_ib_qp_grp_state_to_string(old_state),
 		usnic_ib_qp_grp_state_to_string(new_state));
 	} else {
-		usnic_err("Failed to transistion %u from %s to %s",
+		usnic_err("Failed to transition %u from %s to %s",
 		qp_grp->grp_id,
 		usnic_ib_qp_grp_state_to_string(old_state),
 		usnic_ib_qp_grp_state_to_string(new_state));
diff --git a/drivers/infiniband/hw/usnic/usnic_uiom.c b/drivers/infiniband/hw/usnic/usnic_uiom.c
index 801a1d6937e4..417de1f32960 100644
--- a/drivers/infiniband/hw/usnic/usnic_uiom.c
+++ b/drivers/infiniband/hw/usnic/usnic_uiom.c
@@ -507,7 +507,7 @@ int usnic_uiom_attach_dev_to_pd(struct usnic_uiom_pd *pd, struct device *dev)
 	if (err)
 		goto out_free_dev;
 
-	if (!iommu_domain_has_cap(pd->domain, IOMMU_CAP_CACHE_COHERENCY)) {
+	if (!iommu_capable(dev->bus, IOMMU_CAP_CACHE_COHERENCY)) {
 		usnic_err("IOMMU of %s does not support cache coherency\n",
 				dev_name(dev));
 		err = -EINVAL;
diff --git a/drivers/infiniband/ulp/ipoib/ipoib.h b/drivers/infiniband/ulp/ipoib/ipoib.h
index d7562beb5423..8ba80a6d3a46 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib.h
+++ b/drivers/infiniband/ulp/ipoib/ipoib.h
@@ -98,9 +98,15 @@ enum {
 
 	IPOIB_MCAST_FLAG_FOUND	  = 0,	/* used in set_multicast_list */
 	IPOIB_MCAST_FLAG_SENDONLY = 1,
-	IPOIB_MCAST_FLAG_BUSY	  = 2,	/* joining or already joined */
+	/*
+	 * For IPOIB_MCAST_FLAG_BUSY
+	 * When set, in flight join and mcast->mc is unreliable
+	 * When clear and mcast->mc IS_ERR_OR_NULL, need to restart or
+	 *   haven't started yet
+	 * When clear and mcast->mc is valid pointer, join was successful
+	 */
+	IPOIB_MCAST_FLAG_BUSY	  = 2,
 	IPOIB_MCAST_FLAG_ATTACHED = 3,
-	IPOIB_MCAST_JOIN_STARTED  = 4,
 
 	MAX_SEND_CQE		  = 16,
 	IPOIB_CM_COPYBREAK	  = 256,
@@ -317,6 +323,7 @@ struct ipoib_dev_priv {
 	struct list_head multicast_list;
 	struct rb_root multicast_tree;
 
+	struct workqueue_struct *wq;
 	struct delayed_work mcast_task;
 	struct work_struct carrier_on_task;
 	struct work_struct flush_light;
@@ -477,10 +484,10 @@ void ipoib_ib_dev_flush_heavy(struct work_struct *work);
 void ipoib_pkey_event(struct work_struct *work);
 void ipoib_ib_dev_cleanup(struct net_device *dev);
 
-int ipoib_ib_dev_open(struct net_device *dev, int flush);
+int ipoib_ib_dev_open(struct net_device *dev);
 int ipoib_ib_dev_up(struct net_device *dev);
-int ipoib_ib_dev_down(struct net_device *dev, int flush);
-int ipoib_ib_dev_stop(struct net_device *dev, int flush);
+int ipoib_ib_dev_down(struct net_device *dev);
+int ipoib_ib_dev_stop(struct net_device *dev);
 void ipoib_pkey_dev_check_presence(struct net_device *dev);
 
 int ipoib_dev_init(struct net_device *dev, struct ib_device *ca, int port);
@@ -492,7 +499,7 @@ void ipoib_mcast_send(struct net_device *dev, u8 *daddr, struct sk_buff *skb);
 
 void ipoib_mcast_restart_task(struct work_struct *work);
 int ipoib_mcast_start_thread(struct net_device *dev);
-int ipoib_mcast_stop_thread(struct net_device *dev, int flush);
+int ipoib_mcast_stop_thread(struct net_device *dev);
 
 void ipoib_mcast_dev_down(struct net_device *dev);
 void ipoib_mcast_dev_flush(struct net_device *dev);
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c
index 933efcea0d03..56959adb6c7d 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c
@@ -474,7 +474,7 @@ static int ipoib_cm_req_handler(struct ib_cm_id *cm_id, struct ib_cm_event *even
 	}
 
 	spin_lock_irq(&priv->lock);
-	queue_delayed_work(ipoib_workqueue,
+	queue_delayed_work(priv->wq,
 			   &priv->cm.stale_task, IPOIB_CM_RX_DELAY);
 	/* Add this entry to passive ids list head, but do not re-add it
 	 * if IB_EVENT_QP_LAST_WQE_REACHED has moved it to flush list. */
@@ -576,7 +576,7 @@ void ipoib_cm_handle_rx_wc(struct net_device *dev, struct ib_wc *wc)
 			spin_lock_irqsave(&priv->lock, flags);
 			list_splice_init(&priv->cm.rx_drain_list, &priv->cm.rx_reap_list);
 			ipoib_cm_start_rx_drain(priv);
-			queue_work(ipoib_workqueue, &priv->cm.rx_reap_task);
+			queue_work(priv->wq, &priv->cm.rx_reap_task);
 			spin_unlock_irqrestore(&priv->lock, flags);
 		} else
 			ipoib_warn(priv, "cm recv completion event with wrid %d (> %d)\n",
@@ -603,7 +603,7 @@ void ipoib_cm_handle_rx_wc(struct net_device *dev, struct ib_wc *wc)
 				spin_lock_irqsave(&priv->lock, flags);
 				list_move(&p->list, &priv->cm.rx_reap_list);
 				spin_unlock_irqrestore(&priv->lock, flags);
-				queue_work(ipoib_workqueue, &priv->cm.rx_reap_task);
+				queue_work(priv->wq, &priv->cm.rx_reap_task);
 			}
 			return;
 		}
@@ -827,7 +827,7 @@ void ipoib_cm_handle_tx_wc(struct net_device *dev, struct ib_wc *wc)
 
 		if (test_and_clear_bit(IPOIB_FLAG_INITIALIZED, &tx->flags)) {
 			list_move(&tx->list, &priv->cm.reap_list);
-			queue_work(ipoib_workqueue, &priv->cm.reap_task);
+			queue_work(priv->wq, &priv->cm.reap_task);
 		}
 
 		clear_bit(IPOIB_FLAG_OPER_UP, &tx->flags);
@@ -1255,7 +1255,7 @@ static int ipoib_cm_tx_handler(struct ib_cm_id *cm_id,
 
 		if (test_and_clear_bit(IPOIB_FLAG_INITIALIZED, &tx->flags)) {
 			list_move(&tx->list, &priv->cm.reap_list);
-			queue_work(ipoib_workqueue, &priv->cm.reap_task);
+			queue_work(priv->wq, &priv->cm.reap_task);
 		}
 
 		spin_unlock_irqrestore(&priv->lock, flags);
@@ -1284,7 +1284,7 @@ struct ipoib_cm_tx *ipoib_cm_create_tx(struct net_device *dev, struct ipoib_path
 	tx->dev = dev;
 	list_add(&tx->list, &priv->cm.start_list);
 	set_bit(IPOIB_FLAG_INITIALIZED, &tx->flags);
-	queue_work(ipoib_workqueue, &priv->cm.start_task);
+	queue_work(priv->wq, &priv->cm.start_task);
 	return tx;
 }
 
@@ -1295,7 +1295,7 @@ void ipoib_cm_destroy_tx(struct ipoib_cm_tx *tx)
 	if (test_and_clear_bit(IPOIB_FLAG_INITIALIZED, &tx->flags)) {
 		spin_lock_irqsave(&priv->lock, flags);
 		list_move(&tx->list, &priv->cm.reap_list);
-		queue_work(ipoib_workqueue, &priv->cm.reap_task);
+		queue_work(priv->wq, &priv->cm.reap_task);
 		ipoib_dbg(priv, "Reap connection for gid %pI6\n",
 			  tx->neigh->daddr + 4);
 		tx->neigh = NULL;
@@ -1417,7 +1417,7 @@ void ipoib_cm_skb_too_long(struct net_device *dev, struct sk_buff *skb,
 
 	skb_queue_tail(&priv->cm.skb_queue, skb);
 	if (e)
-		queue_work(ipoib_workqueue, &priv->cm.skb_task);
+		queue_work(priv->wq, &priv->cm.skb_task);
 }
 
 static void ipoib_cm_rx_reap(struct work_struct *work)
@@ -1450,7 +1450,7 @@ static void ipoib_cm_stale_task(struct work_struct *work)
 	}
 
 	if (!list_empty(&priv->cm.passive_ids))
-		queue_delayed_work(ipoib_workqueue,
+		queue_delayed_work(priv->wq,
 				   &priv->cm.stale_task, IPOIB_CM_RX_DELAY);
 	spin_unlock_irq(&priv->lock);
 }
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ib.c b/drivers/infiniband/ulp/ipoib/ipoib_ib.c
index 72626c348174..fe65abb5150c 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c
@@ -655,7 +655,7 @@ void ipoib_reap_ah(struct work_struct *work)
 	__ipoib_reap_ah(dev);
 
 	if (!test_bit(IPOIB_STOP_REAPER, &priv->flags))
-		queue_delayed_work(ipoib_workqueue, &priv->ah_reap_task,
+		queue_delayed_work(priv->wq, &priv->ah_reap_task,
 				   round_jiffies_relative(HZ));
 }
 
@@ -664,7 +664,7 @@ static void ipoib_ib_tx_timer_func(unsigned long ctx)
 	drain_tx_cq((struct net_device *)ctx);
 }
 
-int ipoib_ib_dev_open(struct net_device *dev, int flush)
+int ipoib_ib_dev_open(struct net_device *dev)
 {
 	struct ipoib_dev_priv *priv = netdev_priv(dev);
 	int ret;
@@ -696,7 +696,7 @@ int ipoib_ib_dev_open(struct net_device *dev, int flush)
 	}
 
 	clear_bit(IPOIB_STOP_REAPER, &priv->flags);
-	queue_delayed_work(ipoib_workqueue, &priv->ah_reap_task,
+	queue_delayed_work(priv->wq, &priv->ah_reap_task,
 			   round_jiffies_relative(HZ));
 
 	if (!test_and_set_bit(IPOIB_FLAG_INITIALIZED, &priv->flags))
@@ -706,7 +706,7 @@ int ipoib_ib_dev_open(struct net_device *dev, int flush)
 dev_stop:
 	if (!test_and_set_bit(IPOIB_FLAG_INITIALIZED, &priv->flags))
 		napi_enable(&priv->napi);
-	ipoib_ib_dev_stop(dev, flush);
+	ipoib_ib_dev_stop(dev);
 	return -1;
 }
 
@@ -738,7 +738,7 @@ int ipoib_ib_dev_up(struct net_device *dev)
 	return ipoib_mcast_start_thread(dev);
 }
 
-int ipoib_ib_dev_down(struct net_device *dev, int flush)
+int ipoib_ib_dev_down(struct net_device *dev)
 {
 	struct ipoib_dev_priv *priv = netdev_priv(dev);
 
@@ -747,7 +747,7 @@ int ipoib_ib_dev_down(struct net_device *dev, int flush)
 	clear_bit(IPOIB_FLAG_OPER_UP, &priv->flags);
 	netif_carrier_off(dev);
 
-	ipoib_mcast_stop_thread(dev, flush);
+	ipoib_mcast_stop_thread(dev);
 	ipoib_mcast_dev_flush(dev);
 
 	ipoib_flush_paths(dev);
@@ -807,7 +807,7 @@ void ipoib_drain_cq(struct net_device *dev)
 	local_bh_enable();
 }
 
-int ipoib_ib_dev_stop(struct net_device *dev, int flush)
+int ipoib_ib_dev_stop(struct net_device *dev)
 {
 	struct ipoib_dev_priv *priv = netdev_priv(dev);
 	struct ib_qp_attr qp_attr;
@@ -880,8 +880,7 @@ timeout:
 	/* Wait for all AHs to be reaped */
 	set_bit(IPOIB_STOP_REAPER, &priv->flags);
 	cancel_delayed_work(&priv->ah_reap_task);
-	if (flush)
-		flush_workqueue(ipoib_workqueue);
+	flush_workqueue(priv->wq);
 
 	begin = jiffies;
 
@@ -918,7 +917,7 @@ int ipoib_ib_dev_init(struct net_device *dev, struct ib_device *ca, int port)
 		    (unsigned long) dev);
 
 	if (dev->flags & IFF_UP) {
-		if (ipoib_ib_dev_open(dev, 1)) {
+		if (ipoib_ib_dev_open(dev)) {
 			ipoib_transport_dev_cleanup(dev);
 			return -ENODEV;
 		}
@@ -1040,12 +1039,12 @@ static void __ipoib_ib_dev_flush(struct ipoib_dev_priv *priv,
 	}
 
 	if (level >= IPOIB_FLUSH_NORMAL)
-		ipoib_ib_dev_down(dev, 0);
+		ipoib_ib_dev_down(dev);
 
 	if (level == IPOIB_FLUSH_HEAVY) {
 		if (test_bit(IPOIB_FLAG_INITIALIZED, &priv->flags))
-			ipoib_ib_dev_stop(dev, 0);
-		if (ipoib_ib_dev_open(dev, 0) != 0)
+			ipoib_ib_dev_stop(dev);
+		if (ipoib_ib_dev_open(dev) != 0)
 			return;
 		if (netif_queue_stopped(dev))
 			netif_start_queue(dev);
@@ -1097,7 +1096,7 @@ void ipoib_ib_dev_cleanup(struct net_device *dev)
 	 */
 	ipoib_flush_paths(dev);
 
-	ipoib_mcast_stop_thread(dev, 1);
+	ipoib_mcast_stop_thread(dev);
 	ipoib_mcast_dev_flush(dev);
 
 	ipoib_transport_dev_cleanup(dev);
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c
index 13e6e0431592..6bad17d4d588 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_main.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c
@@ -108,7 +108,7 @@ int ipoib_open(struct net_device *dev)
 
 	set_bit(IPOIB_FLAG_ADMIN_UP, &priv->flags);
 
-	if (ipoib_ib_dev_open(dev, 1)) {
+	if (ipoib_ib_dev_open(dev)) {
 		if (!test_bit(IPOIB_PKEY_ASSIGNED, &priv->flags))
 			return 0;
 		goto err_disable;
@@ -139,7 +139,7 @@ int ipoib_open(struct net_device *dev)
 	return 0;
 
 err_stop:
-	ipoib_ib_dev_stop(dev, 1);
+	ipoib_ib_dev_stop(dev);
 
 err_disable:
 	clear_bit(IPOIB_FLAG_ADMIN_UP, &priv->flags);
@@ -157,8 +157,8 @@ static int ipoib_stop(struct net_device *dev)
 
 	netif_stop_queue(dev);
 
-	ipoib_ib_dev_down(dev, 1);
-	ipoib_ib_dev_stop(dev, 0);
+	ipoib_ib_dev_down(dev);
+	ipoib_ib_dev_stop(dev);
 
 	if (!test_bit(IPOIB_FLAG_SUBINTERFACE, &priv->flags)) {
 		struct ipoib_dev_priv *cpriv;
@@ -839,7 +839,7 @@ static void ipoib_set_mcast_list(struct net_device *dev)
 		return;
 	}
 
-	queue_work(ipoib_workqueue, &priv->restart_task);
+	queue_work(priv->wq, &priv->restart_task);
 }
 
 static u32 ipoib_addr_hash(struct ipoib_neigh_hash *htbl, u8 *daddr)
@@ -954,7 +954,7 @@ static void ipoib_reap_neigh(struct work_struct *work)
 	__ipoib_reap_neigh(priv);
 
 	if (!test_bit(IPOIB_STOP_NEIGH_GC, &priv->flags))
-		queue_delayed_work(ipoib_workqueue, &priv->neigh_reap_task,
+		queue_delayed_work(priv->wq, &priv->neigh_reap_task,
 				   arp_tbl.gc_interval);
 }
 
@@ -1133,7 +1133,7 @@ static int ipoib_neigh_hash_init(struct ipoib_dev_priv *priv)
 
 	/* start garbage collection */
 	clear_bit(IPOIB_STOP_NEIGH_GC, &priv->flags);
-	queue_delayed_work(ipoib_workqueue, &priv->neigh_reap_task,
+	queue_delayed_work(priv->wq, &priv->neigh_reap_task,
 			   arp_tbl.gc_interval);
 
 	return 0;
@@ -1262,15 +1262,13 @@ int ipoib_dev_init(struct net_device *dev, struct ib_device *ca, int port)
 {
 	struct ipoib_dev_priv *priv = netdev_priv(dev);
 
-	if (ipoib_neigh_hash_init(priv) < 0)
-		goto out;
 	/* Allocate RX/TX "rings" to hold queued skbs */
 	priv->rx_ring =	kzalloc(ipoib_recvq_size * sizeof *priv->rx_ring,
 				GFP_KERNEL);
 	if (!priv->rx_ring) {
 		printk(KERN_WARNING "%s: failed to allocate RX ring (%d entries)\n",
 		       ca->name, ipoib_recvq_size);
-		goto out_neigh_hash_cleanup;
+		goto out;
 	}
 
 	priv->tx_ring = vzalloc(ipoib_sendq_size * sizeof *priv->tx_ring);
@@ -1285,16 +1283,24 @@ int ipoib_dev_init(struct net_device *dev, struct ib_device *ca, int port)
 	if (ipoib_ib_dev_init(dev, ca, port))
 		goto out_tx_ring_cleanup;
 
+	/*
+	 * Must be after ipoib_ib_dev_init so we can allocate a per
+	 * device wq there and use it here
+	 */
+	if (ipoib_neigh_hash_init(priv) < 0)
+		goto out_dev_uninit;
+
 	return 0;
 
+out_dev_uninit:
+	ipoib_ib_dev_cleanup(dev);
+
 out_tx_ring_cleanup:
 	vfree(priv->tx_ring);
 
 out_rx_ring_cleanup:
 	kfree(priv->rx_ring);
 
-out_neigh_hash_cleanup:
-	ipoib_neigh_hash_uninit(dev);
 out:
 	return -ENOMEM;
 }
@@ -1317,6 +1323,12 @@ void ipoib_dev_cleanup(struct net_device *dev)
 	}
 	unregister_netdevice_many(&head);
 
+	/*
+	 * Must be before ipoib_ib_dev_cleanup or we delete an in use
+	 * work queue
+	 */
+	ipoib_neigh_hash_uninit(dev);
+
 	ipoib_ib_dev_cleanup(dev);
 
 	kfree(priv->rx_ring);
@@ -1324,8 +1336,6 @@ void ipoib_dev_cleanup(struct net_device *dev)
 
 	priv->rx_ring = NULL;
 	priv->tx_ring = NULL;
-
-	ipoib_neigh_hash_uninit(dev);
 }
 
 static const struct header_ops ipoib_header_ops = {
@@ -1364,7 +1374,7 @@ void ipoib_setup(struct net_device *dev)
 	dev->tx_queue_len	 = ipoib_sendq_size * 2;
 	dev->features		 = (NETIF_F_VLAN_CHALLENGED	|
 				    NETIF_F_HIGHDMA);
-	dev->priv_flags		&= ~IFF_XMIT_DST_RELEASE;
+	netif_keep_dst(dev);
 
 	memcpy(dev->broadcast, ipv4_bcast_addr, INFINIBAND_ALEN);
 
@@ -1636,7 +1646,7 @@ register_failed:
 	/* Stop GC if started before flush */
 	set_bit(IPOIB_STOP_NEIGH_GC, &priv->flags);
 	cancel_delayed_work(&priv->neigh_reap_task);
-	flush_workqueue(ipoib_workqueue);
+	flush_workqueue(priv->wq);
 
 event_failed:
 	ipoib_dev_cleanup(priv->dev);
@@ -1707,7 +1717,7 @@ static void ipoib_remove_one(struct ib_device *device)
 		/* Stop GC */
 		set_bit(IPOIB_STOP_NEIGH_GC, &priv->flags);
 		cancel_delayed_work(&priv->neigh_reap_task);
-		flush_workqueue(ipoib_workqueue);
+		flush_workqueue(priv->wq);
 
 		unregister_netdev(priv->dev);
 		free_netdev(priv->dev);
@@ -1748,8 +1758,13 @@ static int __init ipoib_init_module(void)
 	 * unregister_netdev() and linkwatch_event take the rtnl lock,
 	 * so flush_scheduled_work() can deadlock during device
 	 * removal.
+	 *
+	 * In addition, bringing one device up and another down at the
+	 * same time can deadlock a single workqueue, so we have this
+	 * global fallback workqueue, but we also attempt to open a
+	 * per device workqueue each time we bring an interface up
 	 */
-	ipoib_workqueue = create_singlethread_workqueue("ipoib");
+	ipoib_workqueue = create_singlethread_workqueue("ipoib_flush");
 	if (!ipoib_workqueue) {
 		ret = -ENOMEM;
 		goto err_fs;
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
index ffb83b5f7e80..bc50dd0d0e4d 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
@@ -190,12 +190,6 @@ static int ipoib_mcast_join_finish(struct ipoib_mcast *mcast,
 		spin_unlock_irq(&priv->lock);
 		priv->tx_wr.wr.ud.remote_qkey = priv->qkey;
 		set_qkey = 1;
-
-		if (!ipoib_cm_admin_enabled(dev)) {
-			rtnl_lock();
-			dev_set_mtu(dev, min(priv->mcast_mtu, priv->admin_mtu));
-			rtnl_unlock();
-		}
 	}
 
 	if (!test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags)) {
@@ -277,16 +271,27 @@ ipoib_mcast_sendonly_join_complete(int status,
 	struct ipoib_mcast *mcast = multicast->context;
 	struct net_device *dev = mcast->dev;
 
+	/*
+	 * We have to take the mutex to force mcast_sendonly_join to
+	 * return from ib_sa_multicast_join and set mcast->mc to a
+	 * valid value.  Otherwise we were racing with ourselves in
+	 * that we might fail here, but get a valid return from
+	 * ib_sa_multicast_join after we had cleared mcast->mc here,
+	 * resulting in mis-matched joins and leaves and a deadlock
+	 */
+	mutex_lock(&mcast_mutex);
+
 	/* We trap for port events ourselves. */
 	if (status == -ENETRESET)
-		return 0;
+		goto out;
 
 	if (!status)
 		status = ipoib_mcast_join_finish(mcast, &multicast->rec);
 
 	if (status) {
 		if (mcast->logcount++ < 20)
-			ipoib_dbg_mcast(netdev_priv(dev), "multicast join failed for %pI6, status %d\n",
+			ipoib_dbg_mcast(netdev_priv(dev), "sendonly multicast "
+					"join failed for %pI6, status %d\n",
 					mcast->mcmember.mgid.raw, status);
 
 		/* Flush out any queued packets */
@@ -296,11 +301,15 @@ ipoib_mcast_sendonly_join_complete(int status,
 			dev_kfree_skb_any(skb_dequeue(&mcast->pkt_queue));
 		}
 		netif_tx_unlock_bh(dev);
-
-		/* Clear the busy flag so we try again */
-		status = test_and_clear_bit(IPOIB_MCAST_FLAG_BUSY,
-					    &mcast->flags);
 	}
+out:
+	clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
+	if (status)
+		mcast->mc = NULL;
+	complete(&mcast->done);
+	if (status == -ENETRESET)
+		status = 0;
+	mutex_unlock(&mcast_mutex);
 	return status;
 }
 
@@ -318,12 +327,14 @@ static int ipoib_mcast_sendonly_join(struct ipoib_mcast *mcast)
 	int ret = 0;
 
 	if (!test_bit(IPOIB_FLAG_OPER_UP, &priv->flags)) {
-		ipoib_dbg_mcast(priv, "device shutting down, no multicast joins\n");
+		ipoib_dbg_mcast(priv, "device shutting down, no sendonly "
+				"multicast joins\n");
 		return -ENODEV;
 	}
 
-	if (test_and_set_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags)) {
-		ipoib_dbg_mcast(priv, "multicast entry busy, skipping\n");
+	if (test_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags)) {
+		ipoib_dbg_mcast(priv, "multicast entry busy, skipping "
+				"sendonly join\n");
 		return -EBUSY;
 	}
 
@@ -331,6 +342,9 @@ static int ipoib_mcast_sendonly_join(struct ipoib_mcast *mcast)
 	rec.port_gid = priv->local_gid;
 	rec.pkey     = cpu_to_be16(priv->pkey);
 
+	mutex_lock(&mcast_mutex);
+	init_completion(&mcast->done);
+	set_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
 	mcast->mc = ib_sa_join_multicast(&ipoib_sa_client, priv->ca,
 					 priv->port, &rec,
 					 IB_SA_MCMEMBER_REC_MGID	|
@@ -343,12 +357,14 @@ static int ipoib_mcast_sendonly_join(struct ipoib_mcast *mcast)
 	if (IS_ERR(mcast->mc)) {
 		ret = PTR_ERR(mcast->mc);
 		clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
-		ipoib_warn(priv, "ib_sa_join_multicast failed (ret = %d)\n",
-			   ret);
+		complete(&mcast->done);
+		ipoib_warn(priv, "ib_sa_join_multicast for sendonly join "
+			   "failed (ret = %d)\n", ret);
 	} else {
-		ipoib_dbg_mcast(priv, "no multicast record for %pI6, starting join\n",
-				mcast->mcmember.mgid.raw);
+		ipoib_dbg_mcast(priv, "no multicast record for %pI6, starting "
+				"sendonly join\n", mcast->mcmember.mgid.raw);
 	}
+	mutex_unlock(&mcast_mutex);
 
 	return ret;
 }
@@ -359,18 +375,29 @@ void ipoib_mcast_carrier_on_task(struct work_struct *work)
 						   carrier_on_task);
 	struct ib_port_attr attr;
 
-	/*
-	 * Take rtnl_lock to avoid racing with ipoib_stop() and
-	 * turning the carrier back on while a device is being
-	 * removed.
-	 */
 	if (ib_query_port(priv->ca, priv->port, &attr) ||
 	    attr.state != IB_PORT_ACTIVE) {
 		ipoib_dbg(priv, "Keeping carrier off until IB port is active\n");
 		return;
 	}
 
-	rtnl_lock();
+	/*
+	 * Take rtnl_lock to avoid racing with ipoib_stop() and
+	 * turning the carrier back on while a device is being
+	 * removed.  However, ipoib_stop() will attempt to flush
+	 * the workqueue while holding the rtnl lock, so loop
+	 * on trylock until either we get the lock or we see
+	 * FLAG_ADMIN_UP go away as that signals that we are bailing
+	 * and can safely ignore the carrier on work.
+	 */
+	while (!rtnl_trylock()) {
+		if (!test_bit(IPOIB_FLAG_ADMIN_UP, &priv->flags))
+			return;
+		else
+			msleep(20);
+	}
+	if (!ipoib_cm_admin_enabled(priv->dev))
+		dev_set_mtu(priv->dev, min(priv->mcast_mtu, priv->admin_mtu));
 	netif_carrier_on(priv->dev);
 	rtnl_unlock();
 }
@@ -385,60 +412,63 @@ static int ipoib_mcast_join_complete(int status,
 	ipoib_dbg_mcast(priv, "join completion for %pI6 (status %d)\n",
 			mcast->mcmember.mgid.raw, status);
 
+	/*
+	 * We have to take the mutex to force mcast_join to
+	 * return from ib_sa_multicast_join and set mcast->mc to a
+	 * valid value.  Otherwise we were racing with ourselves in
+	 * that we might fail here, but get a valid return from
+	 * ib_sa_multicast_join after we had cleared mcast->mc here,
+	 * resulting in mis-matched joins and leaves and a deadlock
+	 */
+	mutex_lock(&mcast_mutex);
+
 	/* We trap for port events ourselves. */
-	if (status == -ENETRESET) {
-		status = 0;
+	if (status == -ENETRESET)
 		goto out;
-	}
 
 	if (!status)
 		status = ipoib_mcast_join_finish(mcast, &multicast->rec);
 
 	if (!status) {
 		mcast->backoff = 1;
-		mutex_lock(&mcast_mutex);
 		if (test_bit(IPOIB_MCAST_RUN, &priv->flags))
-			queue_delayed_work(ipoib_workqueue,
-					   &priv->mcast_task, 0);
-		mutex_unlock(&mcast_mutex);
+			queue_delayed_work(priv->wq, &priv->mcast_task, 0);
 
 		/*
-		 * Defer carrier on work to ipoib_workqueue to avoid a
+		 * Defer carrier on work to priv->wq to avoid a
 		 * deadlock on rtnl_lock here.
 		 */
 		if (mcast == priv->broadcast)
-			queue_work(ipoib_workqueue, &priv->carrier_on_task);
-
-		status = 0;
-		goto out;
-	}
-
-	if (mcast->logcount++ < 20) {
-		if (status == -ETIMEDOUT || status == -EAGAIN) {
-			ipoib_dbg_mcast(priv, "multicast join failed for %pI6, status %d\n",
-					mcast->mcmember.mgid.raw, status);
-		} else {
-			ipoib_warn(priv, "multicast join failed for %pI6, status %d\n",
-				   mcast->mcmember.mgid.raw, status);
+			queue_work(priv->wq, &priv->carrier_on_task);
+	} else {
+		if (mcast->logcount++ < 20) {
+			if (status == -ETIMEDOUT || status == -EAGAIN) {
+				ipoib_dbg_mcast(priv, "multicast join failed for %pI6, status %d\n",
+						mcast->mcmember.mgid.raw, status);
+			} else {
+				ipoib_warn(priv, "multicast join failed for %pI6, status %d\n",
+					   mcast->mcmember.mgid.raw, status);
+			}
 		}
-	}
-
-	mcast->backoff *= 2;
-	if (mcast->backoff > IPOIB_MAX_BACKOFF_SECONDS)
-		mcast->backoff = IPOIB_MAX_BACKOFF_SECONDS;
 
-	/* Clear the busy flag so we try again */
-	status = test_and_clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
-
-	mutex_lock(&mcast_mutex);
+		mcast->backoff *= 2;
+		if (mcast->backoff > IPOIB_MAX_BACKOFF_SECONDS)
+			mcast->backoff = IPOIB_MAX_BACKOFF_SECONDS;
+	}
+out:
 	spin_lock_irq(&priv->lock);
-	if (test_bit(IPOIB_MCAST_RUN, &priv->flags))
-		queue_delayed_work(ipoib_workqueue, &priv->mcast_task,
+	clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
+	if (status)
+		mcast->mc = NULL;
+	complete(&mcast->done);
+	if (status == -ENETRESET)
+		status = 0;
+	if (status && test_bit(IPOIB_MCAST_RUN, &priv->flags))
+		queue_delayed_work(priv->wq, &priv->mcast_task,
 				   mcast->backoff * HZ);
 	spin_unlock_irq(&priv->lock);
 	mutex_unlock(&mcast_mutex);
-out:
-	complete(&mcast->done);
+
 	return status;
 }
 
@@ -487,10 +517,9 @@ static void ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast,
 		rec.hop_limit	  = priv->broadcast->mcmember.hop_limit;
 	}
 
-	set_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
+	mutex_lock(&mcast_mutex);
 	init_completion(&mcast->done);
-	set_bit(IPOIB_MCAST_JOIN_STARTED, &mcast->flags);
-
+	set_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
 	mcast->mc = ib_sa_join_multicast(&ipoib_sa_client, priv->ca, priv->port,
 					 &rec, comp_mask, GFP_KERNEL,
 					 ipoib_mcast_join_complete, mcast);
@@ -504,13 +533,11 @@ static void ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast,
 		if (mcast->backoff > IPOIB_MAX_BACKOFF_SECONDS)
 			mcast->backoff = IPOIB_MAX_BACKOFF_SECONDS;
 
-		mutex_lock(&mcast_mutex);
 		if (test_bit(IPOIB_MCAST_RUN, &priv->flags))
-			queue_delayed_work(ipoib_workqueue,
-					   &priv->mcast_task,
+			queue_delayed_work(priv->wq, &priv->mcast_task,
 					   mcast->backoff * HZ);
-		mutex_unlock(&mcast_mutex);
 	}
+	mutex_unlock(&mcast_mutex);
 }
 
 void ipoib_mcast_join_task(struct work_struct *work)
@@ -547,8 +574,8 @@ void ipoib_mcast_join_task(struct work_struct *work)
 			ipoib_warn(priv, "failed to allocate broadcast group\n");
 			mutex_lock(&mcast_mutex);
 			if (test_bit(IPOIB_MCAST_RUN, &priv->flags))
-				queue_delayed_work(ipoib_workqueue,
-						   &priv->mcast_task, HZ);
+				queue_delayed_work(priv->wq, &priv->mcast_task,
+						   HZ);
 			mutex_unlock(&mcast_mutex);
 			return;
 		}
@@ -563,7 +590,8 @@ void ipoib_mcast_join_task(struct work_struct *work)
 	}
 
 	if (!test_bit(IPOIB_MCAST_FLAG_ATTACHED, &priv->broadcast->flags)) {
-		if (!test_bit(IPOIB_MCAST_FLAG_BUSY, &priv->broadcast->flags))
+		if (IS_ERR_OR_NULL(priv->broadcast->mc) &&
+		    !test_bit(IPOIB_MCAST_FLAG_BUSY, &priv->broadcast->flags))
 			ipoib_mcast_join(dev, priv->broadcast, 0);
 		return;
 	}
@@ -571,23 +599,33 @@ void ipoib_mcast_join_task(struct work_struct *work)
 	while (1) {
 		struct ipoib_mcast *mcast = NULL;
 
+		/*
+		 * Need the mutex so our flags are consistent, need the
+		 * priv->lock so we don't race with list removals in either
+		 * mcast_dev_flush or mcast_restart_task
+		 */
+		mutex_lock(&mcast_mutex);
 		spin_lock_irq(&priv->lock);
 		list_for_each_entry(mcast, &priv->multicast_list, list) {
-			if (!test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags)
-			    && !test_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags)
-			    && !test_bit(IPOIB_MCAST_FLAG_ATTACHED, &mcast->flags)) {
+			if (IS_ERR_OR_NULL(mcast->mc) &&
+			    !test_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags) &&
+			    !test_bit(IPOIB_MCAST_FLAG_ATTACHED, &mcast->flags)) {
 				/* Found the next unjoined group */
 				break;
 			}
 		}
 		spin_unlock_irq(&priv->lock);
+		mutex_unlock(&mcast_mutex);
 
 		if (&mcast->list == &priv->multicast_list) {
 			/* All done */
 			break;
 		}
 
-		ipoib_mcast_join(dev, mcast, 1);
+		if (test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags))
+			ipoib_mcast_sendonly_join(mcast);
+		else
+			ipoib_mcast_join(dev, mcast, 1);
 		return;
 	}
 
@@ -604,13 +642,13 @@ int ipoib_mcast_start_thread(struct net_device *dev)
 
 	mutex_lock(&mcast_mutex);
 	if (!test_and_set_bit(IPOIB_MCAST_RUN, &priv->flags))
-		queue_delayed_work(ipoib_workqueue, &priv->mcast_task, 0);
+		queue_delayed_work(priv->wq, &priv->mcast_task, 0);
 	mutex_unlock(&mcast_mutex);
 
 	return 0;
 }
 
-int ipoib_mcast_stop_thread(struct net_device *dev, int flush)
+int ipoib_mcast_stop_thread(struct net_device *dev)
 {
 	struct ipoib_dev_priv *priv = netdev_priv(dev);
 
@@ -621,8 +659,7 @@ int ipoib_mcast_stop_thread(struct net_device *dev, int flush)
 	cancel_delayed_work(&priv->mcast_task);
 	mutex_unlock(&mcast_mutex);
 
-	if (flush)
-		flush_workqueue(ipoib_workqueue);
+	flush_workqueue(priv->wq);
 
 	return 0;
 }
@@ -633,6 +670,9 @@ static int ipoib_mcast_leave(struct net_device *dev, struct ipoib_mcast *mcast)
 	int ret = 0;
 
 	if (test_and_clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags))
+		ipoib_warn(priv, "ipoib_mcast_leave on an in-flight join\n");
+
+	if (!IS_ERR_OR_NULL(mcast->mc))
 		ib_sa_free_multicast(mcast->mc);
 
 	if (test_and_clear_bit(IPOIB_MCAST_FLAG_ATTACHED, &mcast->flags)) {
@@ -685,6 +725,8 @@ void ipoib_mcast_send(struct net_device *dev, u8 *daddr, struct sk_buff *skb)
 		memcpy(mcast->mcmember.mgid.raw, mgid, sizeof (union ib_gid));
 		__ipoib_mcast_add(dev, mcast);
 		list_add_tail(&mcast->list, &priv->multicast_list);
+		if (!test_and_set_bit(IPOIB_MCAST_RUN, &priv->flags))
+			queue_delayed_work(priv->wq, &priv->mcast_task, 0);
 	}
 
 	if (!mcast->ah) {
@@ -698,8 +740,6 @@ void ipoib_mcast_send(struct net_device *dev, u8 *daddr, struct sk_buff *skb)
 		if (test_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags))
 			ipoib_dbg_mcast(priv, "no address vector, "
 					"but multicast join already started\n");
-		else if (test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags))
-			ipoib_mcast_sendonly_join(mcast);
 
 		/*
 		 * If lookup completes between here and out:, don't
@@ -759,9 +799,12 @@ void ipoib_mcast_dev_flush(struct net_device *dev)
 
 	spin_unlock_irqrestore(&priv->lock, flags);
 
-	/* seperate between the wait to the leave*/
+	/*
+	 * make sure the in-flight joins have finished before we attempt
+	 * to leave
+	 */
 	list_for_each_entry_safe(mcast, tmcast, &remove_list, list)
-		if (test_bit(IPOIB_MCAST_JOIN_STARTED, &mcast->flags))
+		if (test_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags))
 			wait_for_completion(&mcast->done);
 
 	list_for_each_entry_safe(mcast, tmcast, &remove_list, list) {
@@ -794,8 +837,6 @@ void ipoib_mcast_restart_task(struct work_struct *work)
 
 	ipoib_dbg_mcast(priv, "restarting multicast task\n");
 
-	ipoib_mcast_stop_thread(dev, 0);
-
 	local_irq_save(flags);
 	netif_addr_lock(dev);
 	spin_lock(&priv->lock);
@@ -880,14 +921,38 @@ void ipoib_mcast_restart_task(struct work_struct *work)
 	netif_addr_unlock(dev);
 	local_irq_restore(flags);
 
-	/* We have to cancel outside of the spinlock */
+	/*
+	 * make sure the in-flight joins have finished before we attempt
+	 * to leave
+	 */
+	list_for_each_entry_safe(mcast, tmcast, &remove_list, list)
+		if (test_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags))
+			wait_for_completion(&mcast->done);
+
+	/*
+	 * We have to cancel outside of the spinlock, but we have to
+	 * take the rtnl lock or else we race with the removal of
+	 * entries from the remove list in mcast_dev_flush as part
+	 * of ipoib_stop().  We detect the drop of the ADMIN_UP flag
+	 * to signal that we have hit this particular race, and we
+	 * return since we know we don't need to do anything else
+	 * anyway.
+	 */
+	while (!rtnl_trylock()) {
+		if (!test_bit(IPOIB_FLAG_ADMIN_UP, &priv->flags))
+			return;
+		else
+			msleep(20);
+	}
 	list_for_each_entry_safe(mcast, tmcast, &remove_list, list) {
 		ipoib_mcast_leave(mcast->dev, mcast);
 		ipoib_mcast_free(mcast);
 	}
-
-	if (test_bit(IPOIB_FLAG_ADMIN_UP, &priv->flags))
-		ipoib_mcast_start_thread(dev);
+	/*
+	 * Restart our join task if needed
+	 */
+	ipoib_mcast_start_thread(dev);
+	rtnl_unlock();
 }
 
 #ifdef CONFIG_INFINIBAND_IPOIB_DEBUG
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_verbs.c b/drivers/infiniband/ulp/ipoib/ipoib_verbs.c
index c56d5d44c53b..b72a753eb41d 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_verbs.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_verbs.c
@@ -145,10 +145,20 @@ int ipoib_transport_dev_init(struct net_device *dev, struct ib_device *ca)
 	int ret, size;
 	int i;
 
+	/*
+	 * the various IPoIB tasks assume they will never race against
+	 * themselves, so always use a single thread workqueue
+	 */
+	priv->wq = create_singlethread_workqueue("ipoib_wq");
+	if (!priv->wq) {
+		printk(KERN_WARNING "ipoib: failed to allocate device WQ\n");
+		return -ENODEV;
+	}
+
 	priv->pd = ib_alloc_pd(priv->ca);
 	if (IS_ERR(priv->pd)) {
 		printk(KERN_WARNING "%s: failed to allocate PD\n", ca->name);
-		return -ENODEV;
+		goto out_free_wq;
 	}
 
 	priv->mr = ib_get_dma_mr(priv->pd, IB_ACCESS_LOCAL_WRITE);
@@ -242,6 +252,10 @@ out_free_mr:
 
 out_free_pd:
 	ib_dealloc_pd(priv->pd);
+
+out_free_wq:
+	destroy_workqueue(priv->wq);
+	priv->wq = NULL;
 	return -ENODEV;
 }
 
@@ -270,6 +284,12 @@ void ipoib_transport_dev_cleanup(struct net_device *dev)
 
 	if (ib_dealloc_pd(priv->pd))
 		ipoib_warn(priv, "ib_dealloc_pd failed\n");
+
+	if (priv->wq) {
+		flush_workqueue(priv->wq);
+		destroy_workqueue(priv->wq);
+		priv->wq = NULL;
+	}
 }
 
 void ipoib_event(struct ib_event_handler *handler,
diff --git a/drivers/infiniband/ulp/iser/iscsi_iser.c b/drivers/infiniband/ulp/iser/iscsi_iser.c
index 93ce62fe1594..6a594aac2290 100644
--- a/drivers/infiniband/ulp/iser/iscsi_iser.c
+++ b/drivers/infiniband/ulp/iser/iscsi_iser.c
@@ -83,7 +83,7 @@ module_param_named(max_lun, iscsi_max_lun, uint, S_IRUGO);
 
 int iser_debug_level = 0;
 bool iser_pi_enable = false;
-int iser_pi_guard = 0;
+int iser_pi_guard = 1;
 
 MODULE_DESCRIPTION("iSER (iSCSI Extensions for RDMA) Datamover");
 MODULE_LICENSE("Dual BSD/GPL");
@@ -97,14 +97,24 @@ module_param_named(pi_enable, iser_pi_enable, bool, 0644);
 MODULE_PARM_DESC(pi_enable, "Enable T10-PI offload support (default:disabled)");
 
 module_param_named(pi_guard, iser_pi_guard, int, 0644);
-MODULE_PARM_DESC(pi_guard, "T10-PI guard_type, 0:CRC|1:IP_CSUM (default:CRC)");
+MODULE_PARM_DESC(pi_guard, "T10-PI guard_type [deprecated]");
 
 static struct workqueue_struct *release_wq;
 struct iser_global ig;
 
+/*
+ * iscsi_iser_recv() - Process a successfull recv completion
+ * @conn:         iscsi connection
+ * @hdr:          iscsi header
+ * @rx_data:      buffer containing receive data payload
+ * @rx_data_len:  length of rx_data
+ *
+ * Notes: In case of data length errors or iscsi PDU completion failures
+ *        this routine will signal iscsi layer of connection failure.
+ */
 void
-iscsi_iser_recv(struct iscsi_conn *conn,
-		struct iscsi_hdr *hdr, char *rx_data, int rx_data_len)
+iscsi_iser_recv(struct iscsi_conn *conn, struct iscsi_hdr *hdr,
+		char *rx_data, int rx_data_len)
 {
 	int rc = 0;
 	int datalen;
@@ -135,49 +145,96 @@ error:
 	iscsi_conn_failure(conn, rc);
 }
 
-static int iscsi_iser_pdu_alloc(struct iscsi_task *task, uint8_t opcode)
+/**
+ * iscsi_iser_pdu_alloc() - allocate an iscsi-iser PDU
+ * @task:     iscsi task
+ * @opcode:   iscsi command opcode
+ *
+ * Netes: This routine can't fail, just assign iscsi task
+ *        hdr and max hdr size.
+ */
+static int
+iscsi_iser_pdu_alloc(struct iscsi_task *task, uint8_t opcode)
 {
 	struct iscsi_iser_task *iser_task = task->dd_data;
 
 	task->hdr = (struct iscsi_hdr *)&iser_task->desc.iscsi_header;
 	task->hdr_max = sizeof(iser_task->desc.iscsi_header);
+
 	return 0;
 }
 
-int iser_initialize_task_headers(struct iscsi_task *task,
-						struct iser_tx_desc *tx_desc)
+/**
+ * iser_initialize_task_headers() - Initialize task headers
+ * @task:       iscsi task
+ * @tx_desc:    iser tx descriptor
+ *
+ * Notes:
+ * This routine may race with iser teardown flow for scsi
+ * error handling TMFs. So for TMF we should acquire the
+ * state mutex to avoid dereferencing the IB device which
+ * may have already been terminated.
+ */
+int
+iser_initialize_task_headers(struct iscsi_task *task,
+			     struct iser_tx_desc *tx_desc)
 {
-	struct iser_conn       *ib_conn   = task->conn->dd_data;
-	struct iser_device     *device    = ib_conn->device;
+	struct iser_conn *iser_conn = task->conn->dd_data;
+	struct iser_device *device = iser_conn->ib_conn.device;
 	struct iscsi_iser_task *iser_task = task->dd_data;
 	u64 dma_addr;
+	const bool mgmt_task = !task->sc && !in_interrupt();
+	int ret = 0;
+
+	if (unlikely(mgmt_task))
+		mutex_lock(&iser_conn->state_mutex);
+
+	if (unlikely(iser_conn->state != ISER_CONN_UP)) {
+		ret = -ENODEV;
+		goto out;
+	}
 
 	dma_addr = ib_dma_map_single(device->ib_device, (void *)tx_desc,
 				ISER_HEADERS_LEN, DMA_TO_DEVICE);
-	if (ib_dma_mapping_error(device->ib_device, dma_addr))
-		return -ENOMEM;
+	if (ib_dma_mapping_error(device->ib_device, dma_addr)) {
+		ret = -ENOMEM;
+		goto out;
+	}
 
 	tx_desc->dma_addr = dma_addr;
 	tx_desc->tx_sg[0].addr   = tx_desc->dma_addr;
 	tx_desc->tx_sg[0].length = ISER_HEADERS_LEN;
 	tx_desc->tx_sg[0].lkey   = device->mr->lkey;
 
-	iser_task->ib_conn = ib_conn;
-	return 0;
+	iser_task->iser_conn = iser_conn;
+out:
+	if (unlikely(mgmt_task))
+		mutex_unlock(&iser_conn->state_mutex);
+
+	return ret;
 }
+
 /**
- * iscsi_iser_task_init - Initialize task
+ * iscsi_iser_task_init() - Initialize iscsi-iser task
  * @task: iscsi task
  *
  * Initialize the task for the scsi command or mgmt command.
+ *
+ * Return: Returns zero on success or -ENOMEM when failing
+ *         to init task headers (dma mapping error).
  */
 static int
 iscsi_iser_task_init(struct iscsi_task *task)
 {
 	struct iscsi_iser_task *iser_task = task->dd_data;
+	int ret;
 
-	if (iser_initialize_task_headers(task, &iser_task->desc))
-			return -ENOMEM;
+	ret = iser_initialize_task_headers(task, &iser_task->desc);
+	if (ret) {
+		iser_err("Failed to init task %p, err = %d\n",
+			 iser_task, ret);
+		return ret;
+	}
 
 	/* mgmt task */
 	if (!task->sc)
@@ -191,7 +248,7 @@ iscsi_iser_task_init(struct iscsi_task *task)
 }
 
 /**
- * iscsi_iser_mtask_xmit - xmit management(immediate) task
+ * iscsi_iser_mtask_xmit() - xmit management (immediate) task
  * @conn: iscsi connection
  * @task: task management task
  *
@@ -249,6 +306,12 @@ iscsi_iser_task_xmit_unsol_data_exit:
 	return error;
 }
 
+/**
+ * iscsi_iser_task_xmit() - xmit iscsi-iser task
+ * @task: iscsi task
+ *
+ * Return: zero on success or escalates $error on failure.
+ */
 static int
 iscsi_iser_task_xmit(struct iscsi_task *task)
 {
@@ -286,12 +349,24 @@ iscsi_iser_task_xmit(struct iscsi_task *task)
 	return error;
 }
 
+/**
+ * iscsi_iser_cleanup_task() - cleanup an iscsi-iser task
+ * @task: iscsi task
+ *
+ * Notes: In case the RDMA device is already NULL (might have
+ *        been removed in DEVICE_REMOVAL CM event it will bail-out
+ *        without doing dma unmapping.
+ */
 static void iscsi_iser_cleanup_task(struct iscsi_task *task)
 {
 	struct iscsi_iser_task *iser_task = task->dd_data;
 	struct iser_tx_desc    *tx_desc   = &iser_task->desc;
-	struct iser_conn       *ib_conn	  = task->conn->dd_data;
-	struct iser_device     *device	  = ib_conn->device;
+	struct iser_conn       *iser_conn	  = task->conn->dd_data;
+	struct iser_device *device = iser_conn->ib_conn.device;
+
+	/* DEVICE_REMOVAL event might have already released the device */
+	if (!device)
+		return;
 
 	ib_dma_unmap_single(device->ib_device,
 		tx_desc->dma_addr, ISER_HEADERS_LEN, DMA_TO_DEVICE);
@@ -306,7 +381,20 @@ static void iscsi_iser_cleanup_task(struct iscsi_task *task)
 	}
 }
 
-static u8 iscsi_iser_check_protection(struct iscsi_task *task, sector_t *sector)
+/**
+ * iscsi_iser_check_protection() - check protection information status of task.
+ * @task:     iscsi task
+ * @sector:   error sector if exsists (output)
+ *
+ * Return: zero if no data-integrity errors have occured
+ *         0x1: data-integrity error occured in the guard-block
+ *         0x2: data-integrity error occured in the reference tag
+ *         0x3: data-integrity error occured in the application tag
+ *
+ *         In addition the error sector is marked.
+ */
+static u8
+iscsi_iser_check_protection(struct iscsi_task *task, sector_t *sector)
 {
 	struct iscsi_iser_task *iser_task = task->dd_data;
 
@@ -318,8 +406,17 @@ static u8 iscsi_iser_check_protection(struct iscsi_task *task, sector_t *sector)
 						 sector);
 }
 
+/**
+ * iscsi_iser_conn_create() - create a new iscsi-iser connection
+ * @cls_session: iscsi class connection
+ * @conn_idx:    connection index within the session (for MCS)
+ *
+ * Return: iscsi_cls_conn when iscsi_conn_setup succeeds or NULL
+ *         otherwise.
+ */
 static struct iscsi_cls_conn *
-iscsi_iser_conn_create(struct iscsi_cls_session *cls_session, uint32_t conn_idx)
+iscsi_iser_conn_create(struct iscsi_cls_session *cls_session,
+		       uint32_t conn_idx)
 {
 	struct iscsi_conn *conn;
 	struct iscsi_cls_conn *cls_conn;
@@ -338,13 +435,25 @@ iscsi_iser_conn_create(struct iscsi_cls_session *cls_session, uint32_t conn_idx)
 	return cls_conn;
 }
 
+/**
+ * iscsi_iser_conn_bind() - bind iscsi and iser connection structures
+ * @cls_session:     iscsi class session
+ * @cls_conn:        iscsi class connection
+ * @transport_eph:   transport end-point handle
+ * @is_leading:      indicate if this is the session leading connection (MCS)
+ *
+ * Return: zero on success, $error if iscsi_conn_bind fails and
+ *         -EINVAL in case end-point doesn't exsits anymore or iser connection
+ *         state is not UP (teardown already started).
+ */
 static int
 iscsi_iser_conn_bind(struct iscsi_cls_session *cls_session,
-		     struct iscsi_cls_conn *cls_conn, uint64_t transport_eph,
+		     struct iscsi_cls_conn *cls_conn,
+		     uint64_t transport_eph,
 		     int is_leading)
 {
 	struct iscsi_conn *conn = cls_conn->dd_data;
-	struct iser_conn *ib_conn;
+	struct iser_conn *iser_conn;
 	struct iscsi_endpoint *ep;
 	int error;
 
@@ -360,66 +469,100 @@ iscsi_iser_conn_bind(struct iscsi_cls_session *cls_session,
 			 (unsigned long long)transport_eph);
 		return -EINVAL;
 	}
-	ib_conn = ep->dd_data;
+	iser_conn = ep->dd_data;
 
-	mutex_lock(&ib_conn->state_mutex);
-	if (ib_conn->state != ISER_CONN_UP) {
+	mutex_lock(&iser_conn->state_mutex);
+	if (iser_conn->state != ISER_CONN_UP) {
 		error = -EINVAL;
 		iser_err("iser_conn %p state is %d, teardown started\n",
-			 ib_conn, ib_conn->state);
+			 iser_conn, iser_conn->state);
 		goto out;
 	}
 
-	error = iser_alloc_rx_descriptors(ib_conn, conn->session);
+	error = iser_alloc_rx_descriptors(iser_conn, conn->session);
 	if (error)
 		goto out;
 
 	/* binds the iSER connection retrieved from the previously
 	 * connected ep_handle to the iSCSI layer connection. exchanges
 	 * connection pointers */
-	iser_info("binding iscsi conn %p to ib_conn %p\n", conn, ib_conn);
+	iser_info("binding iscsi conn %p to iser_conn %p\n", conn, iser_conn);
 
-	conn->dd_data = ib_conn;
-	ib_conn->iscsi_conn = conn;
+	conn->dd_data = iser_conn;
+	iser_conn->iscsi_conn = conn;
 
 out:
-	mutex_unlock(&ib_conn->state_mutex);
+	mutex_unlock(&iser_conn->state_mutex);
 	return error;
 }
 
+/**
+ * iscsi_iser_conn_start() - start iscsi-iser connection
+ * @cls_conn: iscsi class connection
+ *
+ * Notes: Here iser intialize (or re-initialize) stop_completion as
+ *        from this point iscsi must call conn_stop in session/connection
+ *        teardown so iser transport must wait for it.
+ */
 static int
 iscsi_iser_conn_start(struct iscsi_cls_conn *cls_conn)
 {
 	struct iscsi_conn *iscsi_conn;
-	struct iser_conn *ib_conn;
+	struct iser_conn *iser_conn;
 
 	iscsi_conn = cls_conn->dd_data;
-	ib_conn = iscsi_conn->dd_data;
-	reinit_completion(&ib_conn->stop_completion);
+	iser_conn = iscsi_conn->dd_data;
+	reinit_completion(&iser_conn->stop_completion);
 
 	return iscsi_conn_start(cls_conn);
 }
 
+/**
+ * iscsi_iser_conn_stop() - stop iscsi-iser connection
+ * @cls_conn:  iscsi class connection
+ * @flag:      indicate if recover or terminate (passed as is)
+ *
+ * Notes: Calling iscsi_conn_stop might theoretically race with
+ *        DEVICE_REMOVAL event and dereference a previously freed RDMA device
+ *        handle, so we call it under iser the state lock to protect against
+ *        this kind of race.
+ */
 static void
 iscsi_iser_conn_stop(struct iscsi_cls_conn *cls_conn, int flag)
 {
 	struct iscsi_conn *conn = cls_conn->dd_data;
-	struct iser_conn *ib_conn = conn->dd_data;
+	struct iser_conn *iser_conn = conn->dd_data;
 
-	iser_dbg("stopping iscsi_conn: %p, ib_conn: %p\n", conn, ib_conn);
-	iscsi_conn_stop(cls_conn, flag);
+	iser_info("stopping iscsi_conn: %p, iser_conn: %p\n", conn, iser_conn);
 
 	/*
 	 * Userspace may have goofed up and not bound the connection or
 	 * might have only partially setup the connection.
 	 */
-	if (ib_conn) {
+	if (iser_conn) {
+		mutex_lock(&iser_conn->state_mutex);
+		iser_conn_terminate(iser_conn);
+		iscsi_conn_stop(cls_conn, flag);
+
+		/* unbind */
+		iser_conn->iscsi_conn = NULL;
 		conn->dd_data = NULL;
-		complete(&ib_conn->stop_completion);
+
+		complete(&iser_conn->stop_completion);
+		mutex_unlock(&iser_conn->state_mutex);
+	} else {
+		iscsi_conn_stop(cls_conn, flag);
 	}
 }
 
-static void iscsi_iser_session_destroy(struct iscsi_cls_session *cls_session)
+/**
+ * iscsi_iser_session_destroy() - destroy iscsi-iser session
+ * @cls_session: iscsi class session
+ *
+ * Removes and free iscsi host.
+ */
+static void
+iscsi_iser_session_destroy(struct iscsi_cls_session *cls_session)
 {
 	struct Scsi_Host *shost = iscsi_session_to_shost(cls_session);
 
@@ -431,14 +574,25 @@ static void iscsi_iser_session_destroy(struct iscsi_cls_session *cls_session)
 static inline unsigned int
 iser_dif_prot_caps(int prot_caps)
 {
-	return ((prot_caps & IB_PROT_T10DIF_TYPE_1) ? SHOST_DIF_TYPE1_PROTECTION |
-						      SHOST_DIX_TYPE1_PROTECTION : 0) |
-	       ((prot_caps & IB_PROT_T10DIF_TYPE_2) ? SHOST_DIF_TYPE2_PROTECTION |
-						      SHOST_DIX_TYPE2_PROTECTION : 0) |
-	       ((prot_caps & IB_PROT_T10DIF_TYPE_3) ? SHOST_DIF_TYPE3_PROTECTION |
-						      SHOST_DIX_TYPE3_PROTECTION : 0);
+	return ((prot_caps & IB_PROT_T10DIF_TYPE_1) ?
+		SHOST_DIF_TYPE1_PROTECTION | SHOST_DIX_TYPE0_PROTECTION |
+		SHOST_DIX_TYPE1_PROTECTION : 0) |
+	       ((prot_caps & IB_PROT_T10DIF_TYPE_2) ?
+		SHOST_DIF_TYPE2_PROTECTION | SHOST_DIX_TYPE2_PROTECTION : 0) |
+	       ((prot_caps & IB_PROT_T10DIF_TYPE_3) ?
+		SHOST_DIF_TYPE3_PROTECTION | SHOST_DIX_TYPE3_PROTECTION : 0);
 }
 
+/**
+ * iscsi_iser_session_create() - create an iscsi-iser session
+ * @ep:             iscsi end-point handle
+ * @cmds_max:       maximum commands in this session
+ * @qdepth:         session command queue depth
+ * @initial_cmdsn:  initiator command sequnce number
+ *
+ * Allocates and adds a scsi host, expose DIF supprot if
+ * exists, and sets up an iscsi session.
+ */
 static struct iscsi_cls_session *
 iscsi_iser_session_create(struct iscsi_endpoint *ep,
 			  uint16_t cmds_max, uint16_t qdepth,
@@ -447,7 +601,9 @@ iscsi_iser_session_create(struct iscsi_endpoint *ep,
 	struct iscsi_cls_session *cls_session;
 	struct iscsi_session *session;
 	struct Scsi_Host *shost;
-	struct iser_conn *ib_conn = NULL;
+	struct iser_conn *iser_conn = NULL;
+	struct ib_conn *ib_conn;
+	u16 max_cmds;
 
 	shost = iscsi_host_alloc(&iscsi_iser_sht, 0, 0);
 	if (!shost)
@@ -464,26 +620,42 @@ iscsi_iser_session_create(struct iscsi_endpoint *ep,
 	 * the leading conn's ep so this will be NULL;
 	 */
 	if (ep) {
-		ib_conn = ep->dd_data;
+		iser_conn = ep->dd_data;
+		max_cmds = iser_conn->max_cmds;
+
+		mutex_lock(&iser_conn->state_mutex);
+		if (iser_conn->state != ISER_CONN_UP) {
+			iser_err("iser conn %p already started teardown\n",
+				 iser_conn);
+			mutex_unlock(&iser_conn->state_mutex);
+			goto free_host;
+		}
+
+		ib_conn = &iser_conn->ib_conn;
 		if (ib_conn->pi_support) {
 			u32 sig_caps = ib_conn->device->dev_attr.sig_prot_cap;
 
 			scsi_host_set_prot(shost, iser_dif_prot_caps(sig_caps));
-			if (iser_pi_guard)
-				scsi_host_set_guard(shost, SHOST_DIX_GUARD_IP);
-			else
-				scsi_host_set_guard(shost, SHOST_DIX_GUARD_CRC);
+			scsi_host_set_guard(shost, SHOST_DIX_GUARD_IP |
+						   SHOST_DIX_GUARD_CRC);
 		}
-	}
 
-	if (iscsi_host_add(shost,
-			   ep ? ib_conn->device->ib_device->dma_device : NULL))
-		goto free_host;
+		if (iscsi_host_add(shost,
+				   ib_conn->device->ib_device->dma_device)) {
+			mutex_unlock(&iser_conn->state_mutex);
+			goto free_host;
+		}
+		mutex_unlock(&iser_conn->state_mutex);
+	} else {
+		max_cmds = ISER_DEF_XMIT_CMDS_MAX;
+		if (iscsi_host_add(shost, NULL))
+			goto free_host;
+	}
 
-	if (cmds_max > ISER_DEF_XMIT_CMDS_MAX) {
+	if (cmds_max > max_cmds) {
 		iser_info("cmds_max changed from %u to %u\n",
-			  cmds_max, ISER_DEF_XMIT_CMDS_MAX);
-		cmds_max = ISER_DEF_XMIT_CMDS_MAX;
+			  cmds_max, max_cmds);
+		cmds_max = max_cmds;
 	}
 
 	cls_session = iscsi_session_setup(&iscsi_iser_transport, shost,
@@ -549,6 +721,13 @@ iscsi_iser_set_param(struct iscsi_cls_conn *cls_conn,
 	return 0;
 }
 
+/**
+ * iscsi_iser_set_param() - set class connection parameter
+ * @cls_conn:    iscsi class connection
+ * @stats:       iscsi stats to output
+ *
+ * Output connection statistics.
+ */
 static void
 iscsi_iser_conn_get_stats(struct iscsi_cls_conn *cls_conn, struct iscsi_stats *stats)
 {
@@ -577,18 +756,18 @@ iscsi_iser_conn_get_stats(struct iscsi_cls_conn *cls_conn, struct iscsi_stats *s
 static int iscsi_iser_get_ep_param(struct iscsi_endpoint *ep,
 				   enum iscsi_param param, char *buf)
 {
-	struct iser_conn *ib_conn = ep->dd_data;
+	struct iser_conn *iser_conn = ep->dd_data;
 	int len;
 
 	switch (param) {
 	case ISCSI_PARAM_CONN_PORT:
 	case ISCSI_PARAM_CONN_ADDRESS:
-		if (!ib_conn || !ib_conn->cma_id)
+		if (!iser_conn || !iser_conn->ib_conn.cma_id)
 			return -ENOTCONN;
 
 		return iscsi_conn_get_addr_param((struct sockaddr_storage *)
-					&ib_conn->cma_id->route.addr.dst_addr,
-					param, buf);
+				&iser_conn->ib_conn.cma_id->route.addr.dst_addr,
+				param, buf);
 		break;
 	default:
 		return -ENOSYS;
@@ -597,29 +776,44 @@ static int iscsi_iser_get_ep_param(struct iscsi_endpoint *ep,
 	return len;
 }
 
+/**
+ * iscsi_iser_ep_connect() - Initiate iSER connection establishment
+ * @shost:          scsi_host
+ * @dst_addr:       destination address
+ * @non-blocking:   indicate if routine can block
+ *
+ * Allocate an iscsi endpoint, an iser_conn structure and bind them.
+ * After that start RDMA connection establishment via rdma_cm. We
+ * don't allocate iser_conn embedded in iscsi_endpoint since in teardown
+ * the endpoint will be destroyed at ep_disconnect while iser_conn will
+ * cleanup its resources asynchronuously.
+ *
+ * Return: iscsi_endpoint created by iscsi layer or ERR_PTR(error)
+ *         if fails.
+ */
 static struct iscsi_endpoint *
 iscsi_iser_ep_connect(struct Scsi_Host *shost, struct sockaddr *dst_addr,
 		      int non_blocking)
 {
 	int err;
-	struct iser_conn *ib_conn;
+	struct iser_conn *iser_conn;
 	struct iscsi_endpoint *ep;
 
 	ep = iscsi_create_endpoint(0);
 	if (!ep)
 		return ERR_PTR(-ENOMEM);
 
-	ib_conn = kzalloc(sizeof(*ib_conn), GFP_KERNEL);
-	if (!ib_conn) {
+	iser_conn = kzalloc(sizeof(*iser_conn), GFP_KERNEL);
+	if (!iser_conn) {
 		err = -ENOMEM;
 		goto failure;
 	}
 
-	ep->dd_data = ib_conn;
-	ib_conn->ep = ep;
-	iser_conn_init(ib_conn);
+	ep->dd_data = iser_conn;
+	iser_conn->ep = ep;
+	iser_conn_init(iser_conn);
 
-	err = iser_connect(ib_conn, NULL, dst_addr, non_blocking);
+	err = iser_connect(iser_conn, NULL, dst_addr, non_blocking);
 	if (err)
 		goto failure;
 
@@ -629,25 +823,38 @@ failure:
 	return ERR_PTR(err);
 }
 
+/**
+ * iscsi_iser_ep_poll() - poll for iser connection establishment to complete
+ * @ep:            iscsi endpoint (created at ep_connect)
+ * @timeout_ms:    polling timeout allowed in ms.
+ *
+ * This routine boils down to waiting for up_completion signaling
+ * that cma_id got CONNECTED event.
+ *
+ * Return: 1 if succeeded in connection establishment, 0 if timeout expired
+ *         (libiscsi will retry will kick in) or -1 if interrupted by signal
+ *         or more likely iser connection state transitioned to TEMINATING or
+ *         DOWN during the wait period.
+ */
 static int
 iscsi_iser_ep_poll(struct iscsi_endpoint *ep, int timeout_ms)
 {
-	struct iser_conn *ib_conn;
+	struct iser_conn *iser_conn;
 	int rc;
 
-	ib_conn = ep->dd_data;
-	rc = wait_for_completion_interruptible_timeout(&ib_conn->up_completion,
+	iser_conn = ep->dd_data;
+	rc = wait_for_completion_interruptible_timeout(&iser_conn->up_completion,
 						       msecs_to_jiffies(timeout_ms));
 	/* if conn establishment failed, return error code to iscsi */
 	if (rc == 0) {
-		mutex_lock(&ib_conn->state_mutex);
-		if (ib_conn->state == ISER_CONN_TERMINATING ||
-		    ib_conn->state == ISER_CONN_DOWN)
+		mutex_lock(&iser_conn->state_mutex);
+		if (iser_conn->state == ISER_CONN_TERMINATING ||
+		    iser_conn->state == ISER_CONN_DOWN)
 			rc = -1;
-		mutex_unlock(&ib_conn->state_mutex);
+		mutex_unlock(&iser_conn->state_mutex);
 	}
 
-	iser_info("ib conn %p rc = %d\n", ib_conn, rc);
+	iser_info("ib conn %p rc = %d\n", iser_conn, rc);
 
 	if (rc > 0)
 		return 1; /* success, this is the equivalent of POLLOUT */
@@ -657,15 +864,26 @@ iscsi_iser_ep_poll(struct iscsi_endpoint *ep, int timeout_ms)
 		return rc; /* signal */
 }
 
+/**
+ * iscsi_iser_ep_disconnect() - Initiate connection teardown process
+ * @ep:    iscsi endpoint handle
+ *
+ * This routine is not blocked by iser and RDMA termination process
+ * completion as we queue a deffered work for iser/RDMA destruction
+ * and cleanup or actually call it immediately in case we didn't pass
+ * iscsi conn bind/start stage, thus it is safe.
+ */
 static void
 iscsi_iser_ep_disconnect(struct iscsi_endpoint *ep)
 {
-	struct iser_conn *ib_conn;
+	struct iser_conn *iser_conn;
+
+	iser_conn = ep->dd_data;
+	iser_info("ep %p iser conn %p state %d\n",
+		  ep, iser_conn, iser_conn->state);
 
-	ib_conn = ep->dd_data;
-	iser_info("ep %p ib conn %p state %d\n", ep, ib_conn, ib_conn->state);
-	mutex_lock(&ib_conn->state_mutex);
-	iser_conn_terminate(ib_conn);
+	mutex_lock(&iser_conn->state_mutex);
+	iser_conn_terminate(iser_conn);
 
 	/*
 	 * if iser_conn and iscsi_conn are bound, we must wait for
@@ -673,14 +891,14 @@ iscsi_iser_ep_disconnect(struct iscsi_endpoint *ep)
 	 * the iser resources. Otherwise we are safe to free resources
 	 * immediately.
 	 */
-	if (ib_conn->iscsi_conn) {
-		INIT_WORK(&ib_conn->release_work, iser_release_work);
-		queue_work(release_wq, &ib_conn->release_work);
-		mutex_unlock(&ib_conn->state_mutex);
+	if (iser_conn->iscsi_conn) {
+		INIT_WORK(&iser_conn->release_work, iser_release_work);
+		queue_work(release_wq, &iser_conn->release_work);
+		mutex_unlock(&iser_conn->state_mutex);
 	} else {
-		ib_conn->state = ISER_CONN_DOWN;
-		mutex_unlock(&ib_conn->state_mutex);
-		iser_conn_release(ib_conn);
+		iser_conn->state = ISER_CONN_DOWN;
+		mutex_unlock(&iser_conn->state_mutex);
+		iser_conn_release(iser_conn);
 	}
 	iscsi_destroy_endpoint(ep);
 }
@@ -743,7 +961,7 @@ static struct scsi_host_template iscsi_iser_sht = {
 	.module                 = THIS_MODULE,
 	.name                   = "iSCSI Initiator over iSER",
 	.queuecommand           = iscsi_queuecommand,
-	.change_queue_depth	= iscsi_change_queue_depth,
+	.change_queue_depth	= scsi_change_queue_depth,
 	.sg_tablesize           = ISCSI_ISER_SG_TABLESIZE,
 	.max_sectors		= 1024,
 	.cmd_per_lun            = ISER_DEF_CMD_PER_LUN,
@@ -754,6 +972,7 @@ static struct scsi_host_template iscsi_iser_sht = {
 	.use_clustering         = DISABLE_CLUSTERING,
 	.proc_name              = "iscsi_iser",
 	.this_id                = -1,
+	.track_queue_depth	= 1,
 };
 
 static struct iscsi_transport iscsi_iser_transport = {
@@ -843,7 +1062,7 @@ register_transport_failure:
 
 static void __exit iser_exit(void)
 {
-	struct iser_conn *ib_conn, *n;
+	struct iser_conn *iser_conn, *n;
 	int connlist_empty;
 
 	iser_dbg("Removing iSER datamover...\n");
@@ -856,8 +1075,9 @@ static void __exit iser_exit(void)
 	if (!connlist_empty) {
 		iser_err("Error cleanup stage completed but we still have iser "
 			 "connections, destroying them anyway.\n");
-		list_for_each_entry_safe(ib_conn, n, &ig.connlist, conn_list) {
-			iser_conn_release(ib_conn);
+		list_for_each_entry_safe(iser_conn, n, &ig.connlist,
+					 conn_list) {
+			iser_conn_release(iser_conn);
 		}
 	}
 
diff --git a/drivers/infiniband/ulp/iser/iscsi_iser.h b/drivers/infiniband/ulp/iser/iscsi_iser.h
index 9f0e0e34d6ca..5ce26817e7e1 100644
--- a/drivers/infiniband/ulp/iser/iscsi_iser.h
+++ b/drivers/infiniband/ulp/iser/iscsi_iser.h
@@ -69,39 +69,35 @@
 
 #define DRV_NAME	"iser"
 #define PFX		DRV_NAME ": "
-#define DRV_VER		"1.4.1"
+#define DRV_VER		"1.5"
 
-#define iser_dbg(fmt, arg...)				\
-	do {						\
-		if (iser_debug_level > 2)		\
-			printk(KERN_DEBUG PFX "%s:" fmt,\
-				__func__ , ## arg);	\
+#define iser_dbg(fmt, arg...)				 \
+	do {						 \
+		if (unlikely(iser_debug_level > 2))	 \
+			printk(KERN_DEBUG PFX "%s: " fmt,\
+				__func__ , ## arg);	 \
 	} while (0)
 
 #define iser_warn(fmt, arg...)				\
 	do {						\
-		if (iser_debug_level > 0)		\
-			pr_warn(PFX "%s:" fmt,          \
+		if (unlikely(iser_debug_level > 0))	\
+			pr_warn(PFX "%s: " fmt,		\
 				__func__ , ## arg);	\
 	} while (0)
 
 #define iser_info(fmt, arg...)				\
 	do {						\
-		if (iser_debug_level > 1)		\
-			pr_info(PFX "%s:" fmt,          \
+		if (unlikely(iser_debug_level > 1))	\
+			pr_info(PFX "%s: " fmt,		\
 				__func__ , ## arg);	\
 	} while (0)
 
-#define iser_err(fmt, arg...)				\
-	do {						\
-		printk(KERN_ERR PFX "%s:" fmt,          \
-		       __func__ , ## arg);		\
-	} while (0)
+#define iser_err(fmt, arg...) \
+	pr_err(PFX "%s: " fmt, __func__ , ## arg)
 
 #define SHIFT_4K	12
 #define SIZE_4K	(1ULL << SHIFT_4K)
 #define MASK_4K	(~(SIZE_4K-1))
-
 					/* support up to 512KB in one RDMA */
 #define ISCSI_ISER_SG_TABLESIZE         (0x80000 >> SHIFT_4K)
 #define ISER_DEF_XMIT_CMDS_DEFAULT		512
@@ -145,18 +141,37 @@
 					ISER_MAX_TX_MISC_PDUS         + \
 					ISER_MAX_RX_MISC_PDUS)
 
+#define ISER_GET_MAX_XMIT_CMDS(send_wr) ((send_wr			\
+					 - ISER_MAX_TX_MISC_PDUS	\
+					 - ISER_MAX_RX_MISC_PDUS) /	\
+					 (1 + ISER_INFLIGHT_DATAOUTS))
+
+#define ISER_WC_BATCH_COUNT   16
+#define ISER_SIGNAL_CMD_COUNT 32
+
 #define ISER_VER			0x10
 #define ISER_WSV			0x08
 #define ISER_RSV			0x04
 
 #define ISER_FASTREG_LI_WRID		0xffffffffffffffffULL
+#define ISER_BEACON_WRID		0xfffffffffffffffeULL
 
+/**
+ * struct iser_hdr - iSER header
+ *
+ * @flags:        flags support (zbva, remote_inv)
+ * @rsvd:         reserved
+ * @write_stag:   write rkey
+ * @write_va:     write virtual address
+ * @reaf_stag:    read rkey
+ * @read_va:      read virtual address
+ */
 struct iser_hdr {
 	u8      flags;
 	u8      rsvd[3];
-	__be32  write_stag; /* write rkey */
+	__be32  write_stag;
 	__be64  write_va;
-	__be32  read_stag;  /* read rkey */
+	__be32  read_stag;
 	__be64  read_va;
 } __attribute__((packed));
 
@@ -179,7 +194,7 @@ struct iser_cm_hdr {
 /* Length of an object name string */
 #define ISER_OBJECT_NAME_SIZE		    64
 
-enum iser_ib_conn_state {
+enum iser_conn_state {
 	ISER_CONN_INIT,		   /* descriptor allocd, no conn          */
 	ISER_CONN_PENDING,	   /* in the process of being established */
 	ISER_CONN_UP,		   /* up and running                      */
@@ -200,37 +215,63 @@ enum iser_data_dir {
 	ISER_DIRS_NUM
 };
 
+/**
+ * struct iser_data_buf - iSER data buffer
+ *
+ * @buf:          pointer to the sg list
+ * @size:         num entries of this sg
+ * @data_len:     total beffer byte len
+ * @dma_nents:    returned by dma_map_sg
+ * @copy_buf:     allocated copy buf for SGs unaligned
+ *                for rdma which are copied
+ * @sg_single:    SG-ified clone of a non SG SC or
+ *                unaligned SG
+ */
 struct iser_data_buf {
-	void               *buf;      /* pointer to the sg list               */
-	unsigned int       size;      /* num entries of this sg               */
-	unsigned long      data_len;  /* total data len                       */
-	unsigned int       dma_nents; /* returned by dma_map_sg               */
-	char       	   *copy_buf; /* allocated copy buf for SGs unaligned *
-	                               * for rdma which are copied            */
-	struct scatterlist sg_single; /* SG-ified clone of a non SG SC or     *
-				       * unaligned SG                         */
+	void               *buf;
+	unsigned int       size;
+	unsigned long      data_len;
+	unsigned int       dma_nents;
+	char               *copy_buf;
+	struct scatterlist sg_single;
   };
 
 /* fwd declarations */
 struct iser_device;
-struct iser_cq_desc;
 struct iscsi_iser_task;
 struct iscsi_endpoint;
 
+/**
+ * struct iser_mem_reg - iSER memory registration info
+ *
+ * @lkey:         MR local key
+ * @rkey:         MR remote key
+ * @va:           MR start address (buffer va)
+ * @len:          MR length
+ * @mem_h:        pointer to registration context (FMR/Fastreg)
+ */
 struct iser_mem_reg {
 	u32  lkey;
 	u32  rkey;
 	u64  va;
 	u64  len;
 	void *mem_h;
-	int  is_mr;
 };
 
+/**
+ * struct iser_regd_buf - iSER buffer registration desc
+ *
+ * @reg:          memory registration info
+ * @virt_addr:    virtual address of buffer
+ * @device:       reference to iser device
+ * @direction:    dma direction (for dma_unmap)
+ * @data_size:    data buffer size in bytes
+ */
 struct iser_regd_buf {
-	struct iser_mem_reg     reg;        /* memory registration info        */
+	struct iser_mem_reg     reg;
 	void                    *virt_addr;
-	struct iser_device      *device;    /* device->device for dma_unmap    */
-	enum dma_data_direction direction;  /* direction for dma_unmap	       */
+	struct iser_device      *device;
+	enum dma_data_direction direction;
 	unsigned int            data_size;
 };
 
@@ -240,19 +281,39 @@ enum iser_desc_type {
 	ISCSI_TX_DATAOUT
 };
 
+/**
+ * struct iser_tx_desc - iSER TX descriptor (for send wr_id)
+ *
+ * @iser_header:   iser header
+ * @iscsi_header:  iscsi header
+ * @type:          command/control/dataout
+ * @dam_addr:      header buffer dma_address
+ * @tx_sg:         sg[0] points to iser/iscsi headers
+ *                 sg[1] optionally points to either of immediate data
+ *                 unsolicited data-out or control
+ * @num_sge:       number sges used on this TX task
+ */
 struct iser_tx_desc {
 	struct iser_hdr              iser_header;
 	struct iscsi_hdr             iscsi_header;
 	enum   iser_desc_type        type;
 	u64		             dma_addr;
-	/* sg[0] points to iser/iscsi headers, sg[1] optionally points to either
-	of immediate data, unsolicited data-out or control (login,text) */
 	struct ib_sge		     tx_sg[2];
 	int                          num_sge;
 };
 
 #define ISER_RX_PAD_SIZE	(256 - (ISER_RX_PAYLOAD_SIZE + \
 					sizeof(u64) + sizeof(struct ib_sge)))
+/**
+ * struct iser_rx_desc - iSER RX descriptor (for recv wr_id)
+ *
+ * @iser_header:   iser header
+ * @iscsi_header:  iscsi header
+ * @data:          received data segment
+ * @dma_addr:      receive buffer dma address
+ * @rx_sg:         ib_sge of receive buffer
+ * @pad:           for sense data TODO: Modify to maximum sense length supported
+ */
 struct iser_rx_desc {
 	struct iser_hdr              iser_header;
 	struct iscsi_hdr             iscsi_header;
@@ -262,28 +323,60 @@ struct iser_rx_desc {
 	char		             pad[ISER_RX_PAD_SIZE];
 } __attribute__((packed));
 
-#define ISER_MAX_CQ 4
-
 struct iser_conn;
+struct ib_conn;
 struct iscsi_iser_task;
 
+/**
+ * struct iser_comp - iSER completion context
+ *
+ * @device:     pointer to device handle
+ * @cq:         completion queue
+ * @wcs:        work completion array
+ * @tasklet:    Tasklet handle
+ * @active_qps: Number of active QPs attached
+ *              to completion context
+ */
+struct iser_comp {
+	struct iser_device      *device;
+	struct ib_cq		*cq;
+	struct ib_wc		 wcs[ISER_WC_BATCH_COUNT];
+	struct tasklet_struct	 tasklet;
+	int                      active_qps;
+};
+
+/**
+ * struct iser_device - iSER device handle
+ *
+ * @ib_device:     RDMA device
+ * @pd:            Protection Domain for this device
+ * @dev_attr:      Device attributes container
+ * @mr:            Global DMA memory region
+ * @event_handler: IB events handle routine
+ * @ig_list:	   entry in devices list
+ * @refcount:      Reference counter, dominated by open iser connections
+ * @comps_used:    Number of completion contexts used, Min between online
+ *                 cpus and device max completion vectors
+ * @comps:         Dinamically allocated array of completion handlers
+ * Memory registration pool Function pointers (FMR or Fastreg):
+ *     @iser_alloc_rdma_reg_res: Allocation of memory regions pool
+ *     @iser_free_rdma_reg_res:  Free of memory regions pool
+ *     @iser_reg_rdma_mem:       Memory registration routine
+ *     @iser_unreg_rdma_mem:     Memory deregistration routine
+ */
 struct iser_device {
 	struct ib_device             *ib_device;
 	struct ib_pd	             *pd;
 	struct ib_device_attr	     dev_attr;
-	struct ib_cq	             *rx_cq[ISER_MAX_CQ];
-	struct ib_cq	             *tx_cq[ISER_MAX_CQ];
 	struct ib_mr	             *mr;
-	struct tasklet_struct	     cq_tasklet[ISER_MAX_CQ];
 	struct ib_event_handler      event_handler;
-	struct list_head             ig_list; /* entry in ig devices list */
+	struct list_head             ig_list;
 	int                          refcount;
-	int                          cq_active_qps[ISER_MAX_CQ];
-	int			     cqs_used;
-	struct iser_cq_desc	     *cq_desc;
-	int                          (*iser_alloc_rdma_reg_res)(struct iser_conn *ib_conn,
+	int			     comps_used;
+	struct iser_comp	     *comps;
+	int                          (*iser_alloc_rdma_reg_res)(struct ib_conn *ib_conn,
 								unsigned cmds_max);
-	void                         (*iser_free_rdma_reg_res)(struct iser_conn *ib_conn);
+	void                         (*iser_free_rdma_reg_res)(struct ib_conn *ib_conn);
 	int                          (*iser_reg_rdma_mem)(struct iscsi_iser_task *iser_task,
 							  enum iser_data_dir cmd_dir);
 	void                         (*iser_unreg_rdma_mem)(struct iscsi_iser_task *iser_task,
@@ -301,78 +394,164 @@ enum iser_reg_indicator {
 	ISER_FASTREG_PROTECTED	= 1 << 3,
 };
 
+/**
+ * struct iser_pi_context - Protection information context
+ *
+ * @prot_mr:        protection memory region
+ * @prot_frpl:      protection fastreg page list
+ * @sig_mr:         signature feature enabled memory region
+ */
 struct iser_pi_context {
 	struct ib_mr                   *prot_mr;
 	struct ib_fast_reg_page_list   *prot_frpl;
 	struct ib_mr                   *sig_mr;
 };
 
+/**
+ * struct fast_reg_descriptor - Fast registration descriptor
+ *
+ * @list:           entry in connection fastreg pool
+ * @data_mr:        data memory region
+ * @data_frpl:      data fastreg page list
+ * @pi_ctx:         protection information context
+ * @reg_indicators: fast registration indicators
+ */
 struct fast_reg_descriptor {
 	struct list_head		  list;
-	/* For fast registration - FRWR */
 	struct ib_mr			 *data_mr;
 	struct ib_fast_reg_page_list     *data_frpl;
 	struct iser_pi_context		 *pi_ctx;
-	/* registration indicators container */
 	u8				  reg_indicators;
 };
 
+/**
+ * struct ib_conn - Infiniband related objects
+ *
+ * @cma_id:              rdma_cm connection maneger handle
+ * @qp:                  Connection Queue-pair
+ * @post_recv_buf_count: post receive counter
+ * @sig_count:           send work request signal count
+ * @rx_wr:               receive work request for batch posts
+ * @device:              reference to iser device
+ * @comp:                iser completion context
+ * @pi_support:          Indicate device T10-PI support
+ * @beacon:              beacon send wr to signal all flush errors were drained
+ * @flush_comp:          completes when all connection completions consumed
+ * @lock:                protects fmr/fastreg pool
+ * @union.fmr:
+ *     @pool:            FMR pool for fast registrations
+ *     @page_vec:        page vector to hold mapped commands pages
+ *                       used for registration
+ * @union.fastreg:
+ *     @pool:            Fast registration descriptors pool for fast
+ *                       registrations
+ *     @pool_size:       Size of pool
+ */
+struct ib_conn {
+	struct rdma_cm_id           *cma_id;
+	struct ib_qp	            *qp;
+	int                          post_recv_buf_count;
+	u8                           sig_count;
+	struct ib_recv_wr	     rx_wr[ISER_MIN_POSTED_RX];
+	struct iser_device          *device;
+	struct iser_comp	    *comp;
+	bool			     pi_support;
+	struct ib_send_wr	     beacon;
+	struct completion	     flush_comp;
+	spinlock_t		     lock;
+	union {
+		struct {
+			struct ib_fmr_pool      *pool;
+			struct iser_page_vec	*page_vec;
+		} fmr;
+		struct {
+			struct list_head	 pool;
+			int			 pool_size;
+		} fastreg;
+	};
+};
+
+/**
+ * struct iser_conn - iSER connection context
+ *
+ * @ib_conn:          connection RDMA resources
+ * @iscsi_conn:       link to matching iscsi connection
+ * @ep:               transport handle
+ * @state:            connection logical state
+ * @qp_max_recv_dtos: maximum number of data outs, corresponds
+ *                    to max number of post recvs
+ * @qp_max_recv_dtos_mask: (qp_max_recv_dtos - 1)
+ * @min_posted_rx:    (qp_max_recv_dtos >> 2)
+ * @max_cmds:         maximum cmds allowed for this connection
+ * @name:             connection peer portal
+ * @release_work:     deffered work for release job
+ * @state_mutex:      protects iser onnection state
+ * @stop_completion:  conn_stop completion
+ * @ib_completion:    RDMA cleanup completion
+ * @up_completion:    connection establishment completed
+ *                    (state is ISER_CONN_UP)
+ * @conn_list:        entry in ig conn list
+ * @login_buf:        login data buffer (stores login parameters)
+ * @login_req_buf:    login request buffer
+ * @login_req_dma:    login request buffer dma address
+ * @login_resp_buf:   login response buffer
+ * @login_resp_dma:   login response buffer dma address
+ * @rx_desc_head:     head of rx_descs cyclic buffer
+ * @rx_descs:         rx buffers array (cyclic buffer)
+ * @num_rx_descs:     number of rx descriptors
+ */
 struct iser_conn {
+	struct ib_conn		     ib_conn;
 	struct iscsi_conn	     *iscsi_conn;
 	struct iscsi_endpoint	     *ep;
-	enum iser_ib_conn_state	     state;	    /* rdma connection state   */
-	atomic_t		     refcount;
-	spinlock_t		     lock;	    /* used for state changes  */
-	struct iser_device           *device;       /* device context          */
-	struct rdma_cm_id            *cma_id;       /* CMA ID		       */
-	struct ib_qp	             *qp;           /* QP 		       */
-	unsigned		     qp_max_recv_dtos; /* num of rx buffers */
-	unsigned		     qp_max_recv_dtos_mask; /* above minus 1 */
-	unsigned		     min_posted_rx; /* qp_max_recv_dtos >> 2 */
-	int                          post_recv_buf_count; /* posted rx count  */
-	atomic_t                     post_send_buf_count; /* posted tx count   */
+	enum iser_conn_state	     state;
+	unsigned		     qp_max_recv_dtos;
+	unsigned		     qp_max_recv_dtos_mask;
+	unsigned		     min_posted_rx;
+	u16                          max_cmds;
 	char 			     name[ISER_OBJECT_NAME_SIZE];
 	struct work_struct	     release_work;
-	struct completion	     stop_completion;
 	struct mutex		     state_mutex;
-	struct completion	     flush_completion;
+	struct completion	     stop_completion;
+	struct completion	     ib_completion;
 	struct completion	     up_completion;
-	struct list_head	     conn_list;       /* entry in ig conn list */
+	struct list_head	     conn_list;
 
 	char  			     *login_buf;
 	char			     *login_req_buf, *login_resp_buf;
 	u64			     login_req_dma, login_resp_dma;
 	unsigned int 		     rx_desc_head;
 	struct iser_rx_desc	     *rx_descs;
-	struct ib_recv_wr	     rx_wr[ISER_MIN_POSTED_RX];
-	bool			     pi_support;
-
-	/* Connection memory registration pool */
-	union {
-		struct {
-			struct ib_fmr_pool      *pool;	   /* pool of IB FMRs         */
-			struct iser_page_vec	*page_vec; /* represents SG to fmr maps*
-							    * maps serialized as tx is*/
-		} fmr;
-		struct {
-			struct list_head	pool;
-			int			pool_size;
-		} fastreg;
-	};
+	u32                          num_rx_descs;
 };
 
+/**
+ * struct iscsi_iser_task - iser task context
+ *
+ * @desc:     TX descriptor
+ * @iser_conn:        link to iser connection
+ * @status:           current task status
+ * @sc:               link to scsi command
+ * @command_sent:     indicate if command was sent
+ * @dir:              iser data direction
+ * @rdma_regd:        task rdma registration desc
+ * @data:             iser data buffer desc
+ * @data_copy:        iser data copy buffer desc (bounce buffer)
+ * @prot:             iser protection buffer desc
+ * @prot_copy:        iser protection copy buffer desc (bounce buffer)
+ */
 struct iscsi_iser_task {
 	struct iser_tx_desc          desc;
-	struct iser_conn	     *ib_conn;
+	struct iser_conn	     *iser_conn;
 	enum iser_task_status 	     status;
 	struct scsi_cmnd	     *sc;
-	int                          command_sent;  /* set if command  sent  */
-	int                          dir[ISER_DIRS_NUM];      /* set if dir use*/
-	struct iser_regd_buf         rdma_regd[ISER_DIRS_NUM];/* regd rdma buf */
-	struct iser_data_buf         data[ISER_DIRS_NUM];     /* orig. data des*/
-	struct iser_data_buf         data_copy[ISER_DIRS_NUM];/* contig. copy  */
-	struct iser_data_buf         prot[ISER_DIRS_NUM];     /* prot desc     */
-	struct iser_data_buf         prot_copy[ISER_DIRS_NUM];/* prot copy     */
+	int                          command_sent;
+	int                          dir[ISER_DIRS_NUM];
+	struct iser_regd_buf         rdma_regd[ISER_DIRS_NUM];
+	struct iser_data_buf         data[ISER_DIRS_NUM];
+	struct iser_data_buf         data_copy[ISER_DIRS_NUM];
+	struct iser_data_buf         prot[ISER_DIRS_NUM];
+	struct iser_data_buf         prot_copy[ISER_DIRS_NUM];
 };
 
 struct iser_page_vec {
@@ -382,17 +561,20 @@ struct iser_page_vec {
 	int data_size;
 };
 
-struct iser_cq_desc {
-	struct iser_device           *device;
-	int                          cq_index;
-};
-
+/**
+ * struct iser_global: iSER global context
+ *
+ * @device_list_mutex:    protects device_list
+ * @device_list:          iser devices global list
+ * @connlist_mutex:       protects connlist
+ * @connlist:             iser connections global list
+ * @desc_cache:           kmem cache for tx dataout
+ */
 struct iser_global {
-	struct mutex      device_list_mutex;/*                   */
-	struct list_head  device_list;	     /* all iSER devices */
+	struct mutex      device_list_mutex;
+	struct list_head  device_list;
 	struct mutex      connlist_mutex;
-	struct list_head  connlist;		/* all iSER IB connections */
-
+	struct list_head  connlist;
 	struct kmem_cache *desc_cache;
 };
 
@@ -401,9 +583,6 @@ extern int iser_debug_level;
 extern bool iser_pi_enable;
 extern int iser_pi_guard;
 
-/* allocate connection resources needed for rdma functionality */
-int iser_conn_set_full_featured_mode(struct iscsi_conn *conn);
-
 int iser_send_control(struct iscsi_conn *conn,
 		      struct iscsi_task *task);
 
@@ -415,29 +594,30 @@ int iser_send_data_out(struct iscsi_conn *conn,
 		       struct iscsi_data *hdr);
 
 void iscsi_iser_recv(struct iscsi_conn *conn,
-		     struct iscsi_hdr       *hdr,
-		     char                   *rx_data,
-		     int                    rx_data_len);
+		     struct iscsi_hdr *hdr,
+		     char *rx_data,
+		     int rx_data_len);
 
-void iser_conn_init(struct iser_conn *ib_conn);
+void iser_conn_init(struct iser_conn *iser_conn);
 
-void iser_conn_release(struct iser_conn *ib_conn);
+void iser_conn_release(struct iser_conn *iser_conn);
 
-void iser_conn_terminate(struct iser_conn *ib_conn);
+int iser_conn_terminate(struct iser_conn *iser_conn);
 
 void iser_release_work(struct work_struct *work);
 
 void iser_rcv_completion(struct iser_rx_desc *desc,
-			 unsigned long    dto_xfer_len,
-			struct iser_conn *ib_conn);
+			 unsigned long dto_xfer_len,
+			 struct ib_conn *ib_conn);
 
-void iser_snd_completion(struct iser_tx_desc *desc, struct iser_conn *ib_conn);
+void iser_snd_completion(struct iser_tx_desc *desc,
+			 struct ib_conn *ib_conn);
 
 void iser_task_rdma_init(struct iscsi_iser_task *task);
 
 void iser_task_rdma_finalize(struct iscsi_iser_task *task);
 
-void iser_free_rx_descriptors(struct iser_conn *ib_conn);
+void iser_free_rx_descriptors(struct iser_conn *iser_conn);
 
 void iser_finalize_rdma_unaligned_sg(struct iscsi_iser_task *iser_task,
 				     struct iser_data_buf *mem,
@@ -449,38 +629,40 @@ int  iser_reg_rdma_mem_fmr(struct iscsi_iser_task *task,
 int  iser_reg_rdma_mem_fastreg(struct iscsi_iser_task *task,
 			       enum iser_data_dir cmd_dir);
 
-int  iser_connect(struct iser_conn   *ib_conn,
-		  struct sockaddr    *src_addr,
-		  struct sockaddr    *dst_addr,
-		  int                non_blocking);
+int  iser_connect(struct iser_conn *iser_conn,
+		  struct sockaddr *src_addr,
+		  struct sockaddr *dst_addr,
+		  int non_blocking);
 
-int  iser_reg_page_vec(struct iser_conn     *ib_conn,
+int  iser_reg_page_vec(struct ib_conn *ib_conn,
 		       struct iser_page_vec *page_vec,
-		       struct iser_mem_reg  *mem_reg);
+		       struct iser_mem_reg *mem_reg);
 
 void iser_unreg_mem_fmr(struct iscsi_iser_task *iser_task,
 			enum iser_data_dir cmd_dir);
 void iser_unreg_mem_fastreg(struct iscsi_iser_task *iser_task,
 			    enum iser_data_dir cmd_dir);
 
-int  iser_post_recvl(struct iser_conn *ib_conn);
-int  iser_post_recvm(struct iser_conn *ib_conn, int count);
-int  iser_post_send(struct iser_conn *ib_conn, struct iser_tx_desc *tx_desc);
+int  iser_post_recvl(struct iser_conn *iser_conn);
+int  iser_post_recvm(struct iser_conn *iser_conn, int count);
+int  iser_post_send(struct ib_conn *ib_conn, struct iser_tx_desc *tx_desc,
+		    bool signal);
 
 int iser_dma_map_task_data(struct iscsi_iser_task *iser_task,
-			    struct iser_data_buf       *data,
-			    enum   iser_data_dir       iser_dir,
-			    enum   dma_data_direction  dma_dir);
+			   struct iser_data_buf *data,
+			   enum iser_data_dir iser_dir,
+			   enum dma_data_direction dma_dir);
 
 void iser_dma_unmap_task_data(struct iscsi_iser_task *iser_task,
 			      struct iser_data_buf *data);
 int  iser_initialize_task_headers(struct iscsi_task *task,
 			struct iser_tx_desc *tx_desc);
-int iser_alloc_rx_descriptors(struct iser_conn *ib_conn, struct iscsi_session *session);
-int iser_create_fmr_pool(struct iser_conn *ib_conn, unsigned cmds_max);
-void iser_free_fmr_pool(struct iser_conn *ib_conn);
-int iser_create_fastreg_pool(struct iser_conn *ib_conn, unsigned cmds_max);
-void iser_free_fastreg_pool(struct iser_conn *ib_conn);
+int iser_alloc_rx_descriptors(struct iser_conn *iser_conn,
+			      struct iscsi_session *session);
+int iser_create_fmr_pool(struct ib_conn *ib_conn, unsigned cmds_max);
+void iser_free_fmr_pool(struct ib_conn *ib_conn);
+int iser_create_fastreg_pool(struct ib_conn *ib_conn, unsigned cmds_max);
+void iser_free_fastreg_pool(struct ib_conn *ib_conn);
 u8 iser_check_task_pi_status(struct iscsi_iser_task *iser_task,
 			     enum iser_data_dir cmd_dir, sector_t *sector);
 #endif
diff --git a/drivers/infiniband/ulp/iser/iser_initiator.c b/drivers/infiniband/ulp/iser/iser_initiator.c
index 8d44a4060634..3821633f1065 100644
--- a/drivers/infiniband/ulp/iser/iser_initiator.c
+++ b/drivers/infiniband/ulp/iser/iser_initiator.c
@@ -49,7 +49,7 @@ static int iser_prepare_read_cmd(struct iscsi_task *task)
 
 {
 	struct iscsi_iser_task *iser_task = task->dd_data;
-	struct iser_device  *device = iser_task->ib_conn->device;
+	struct iser_device  *device = iser_task->iser_conn->ib_conn.device;
 	struct iser_regd_buf *regd_buf;
 	int err;
 	struct iser_hdr *hdr = &iser_task->desc.iser_header;
@@ -103,7 +103,7 @@ iser_prepare_write_cmd(struct iscsi_task *task,
 		       unsigned int edtl)
 {
 	struct iscsi_iser_task *iser_task = task->dd_data;
-	struct iser_device  *device = iser_task->ib_conn->device;
+	struct iser_device  *device = iser_task->iser_conn->ib_conn.device;
 	struct iser_regd_buf *regd_buf;
 	int err;
 	struct iser_hdr *hdr = &iser_task->desc.iser_header;
@@ -160,10 +160,10 @@ iser_prepare_write_cmd(struct iscsi_task *task,
 }
 
 /* creates a new tx descriptor and adds header regd buffer */
-static void iser_create_send_desc(struct iser_conn	*ib_conn,
+static void iser_create_send_desc(struct iser_conn	*iser_conn,
 				  struct iser_tx_desc	*tx_desc)
 {
-	struct iser_device *device = ib_conn->device;
+	struct iser_device *device = iser_conn->ib_conn.device;
 
 	ib_dma_sync_single_for_cpu(device->ib_device,
 		tx_desc->dma_addr, ISER_HEADERS_LEN, DMA_TO_DEVICE);
@@ -179,103 +179,108 @@ static void iser_create_send_desc(struct iser_conn	*ib_conn,
 	}
 }
 
-static void iser_free_login_buf(struct iser_conn *ib_conn)
+static void iser_free_login_buf(struct iser_conn *iser_conn)
 {
-	if (!ib_conn->login_buf)
+	struct iser_device *device = iser_conn->ib_conn.device;
+
+	if (!iser_conn->login_buf)
 		return;
 
-	if (ib_conn->login_req_dma)
-		ib_dma_unmap_single(ib_conn->device->ib_device,
-				    ib_conn->login_req_dma,
+	if (iser_conn->login_req_dma)
+		ib_dma_unmap_single(device->ib_device,
+				    iser_conn->login_req_dma,
 				    ISCSI_DEF_MAX_RECV_SEG_LEN, DMA_TO_DEVICE);
 
-	if (ib_conn->login_resp_dma)
-		ib_dma_unmap_single(ib_conn->device->ib_device,
-				    ib_conn->login_resp_dma,
+	if (iser_conn->login_resp_dma)
+		ib_dma_unmap_single(device->ib_device,
+				    iser_conn->login_resp_dma,
 				    ISER_RX_LOGIN_SIZE, DMA_FROM_DEVICE);
 
-	kfree(ib_conn->login_buf);
+	kfree(iser_conn->login_buf);
 
 	/* make sure we never redo any unmapping */
-	ib_conn->login_req_dma = 0;
-	ib_conn->login_resp_dma = 0;
-	ib_conn->login_buf = NULL;
+	iser_conn->login_req_dma = 0;
+	iser_conn->login_resp_dma = 0;
+	iser_conn->login_buf = NULL;
 }
 
-static int iser_alloc_login_buf(struct iser_conn *ib_conn)
+static int iser_alloc_login_buf(struct iser_conn *iser_conn)
 {
-	struct iser_device	*device;
+	struct iser_device *device = iser_conn->ib_conn.device;
 	int			req_err, resp_err;
 
-	BUG_ON(ib_conn->device == NULL);
+	BUG_ON(device == NULL);
 
-	device = ib_conn->device;
-
-	ib_conn->login_buf = kmalloc(ISCSI_DEF_MAX_RECV_SEG_LEN +
+	iser_conn->login_buf = kmalloc(ISCSI_DEF_MAX_RECV_SEG_LEN +
 				     ISER_RX_LOGIN_SIZE, GFP_KERNEL);
-	if (!ib_conn->login_buf)
+	if (!iser_conn->login_buf)
 		goto out_err;
 
-	ib_conn->login_req_buf  = ib_conn->login_buf;
-	ib_conn->login_resp_buf = ib_conn->login_buf +
+	iser_conn->login_req_buf  = iser_conn->login_buf;
+	iser_conn->login_resp_buf = iser_conn->login_buf +
 						ISCSI_DEF_MAX_RECV_SEG_LEN;
 
-	ib_conn->login_req_dma = ib_dma_map_single(ib_conn->device->ib_device,
-				(void *)ib_conn->login_req_buf,
-				ISCSI_DEF_MAX_RECV_SEG_LEN, DMA_TO_DEVICE);
+	iser_conn->login_req_dma = ib_dma_map_single(device->ib_device,
+						     iser_conn->login_req_buf,
+						     ISCSI_DEF_MAX_RECV_SEG_LEN,
+						     DMA_TO_DEVICE);
 
-	ib_conn->login_resp_dma = ib_dma_map_single(ib_conn->device->ib_device,
-				(void *)ib_conn->login_resp_buf,
-				ISER_RX_LOGIN_SIZE, DMA_FROM_DEVICE);
+	iser_conn->login_resp_dma = ib_dma_map_single(device->ib_device,
+						      iser_conn->login_resp_buf,
+						      ISER_RX_LOGIN_SIZE,
+						      DMA_FROM_DEVICE);
 
 	req_err  = ib_dma_mapping_error(device->ib_device,
-					ib_conn->login_req_dma);
+					iser_conn->login_req_dma);
 	resp_err = ib_dma_mapping_error(device->ib_device,
-					ib_conn->login_resp_dma);
+					iser_conn->login_resp_dma);
 
 	if (req_err || resp_err) {
 		if (req_err)
-			ib_conn->login_req_dma = 0;
+			iser_conn->login_req_dma = 0;
 		if (resp_err)
-			ib_conn->login_resp_dma = 0;
+			iser_conn->login_resp_dma = 0;
 		goto free_login_buf;
 	}
 	return 0;
 
 free_login_buf:
-	iser_free_login_buf(ib_conn);
+	iser_free_login_buf(iser_conn);
 
 out_err:
 	iser_err("unable to alloc or map login buf\n");
 	return -ENOMEM;
 }
 
-int iser_alloc_rx_descriptors(struct iser_conn *ib_conn, struct iscsi_session *session)
+int iser_alloc_rx_descriptors(struct iser_conn *iser_conn,
+			      struct iscsi_session *session)
 {
 	int i, j;
 	u64 dma_addr;
 	struct iser_rx_desc *rx_desc;
 	struct ib_sge       *rx_sg;
-	struct iser_device  *device = ib_conn->device;
+	struct ib_conn *ib_conn = &iser_conn->ib_conn;
+	struct iser_device *device = ib_conn->device;
 
-	ib_conn->qp_max_recv_dtos = session->cmds_max;
-	ib_conn->qp_max_recv_dtos_mask = session->cmds_max - 1; /* cmds_max is 2^N */
-	ib_conn->min_posted_rx = ib_conn->qp_max_recv_dtos >> 2;
+	iser_conn->qp_max_recv_dtos = session->cmds_max;
+	iser_conn->qp_max_recv_dtos_mask = session->cmds_max - 1; /* cmds_max is 2^N */
+	iser_conn->min_posted_rx = iser_conn->qp_max_recv_dtos >> 2;
 
 	if (device->iser_alloc_rdma_reg_res(ib_conn, session->scsi_cmds_max))
 		goto create_rdma_reg_res_failed;
 
-	if (iser_alloc_login_buf(ib_conn))
+	if (iser_alloc_login_buf(iser_conn))
 		goto alloc_login_buf_fail;
 
-	ib_conn->rx_descs = kmalloc(session->cmds_max *
+	iser_conn->num_rx_descs = session->cmds_max;
+	iser_conn->rx_descs = kmalloc(iser_conn->num_rx_descs *
 				sizeof(struct iser_rx_desc), GFP_KERNEL);
-	if (!ib_conn->rx_descs)
+	if (!iser_conn->rx_descs)
 		goto rx_desc_alloc_fail;
 
-	rx_desc = ib_conn->rx_descs;
+	rx_desc = iser_conn->rx_descs;
 
-	for (i = 0; i < ib_conn->qp_max_recv_dtos; i++, rx_desc++)  {
+	for (i = 0; i < iser_conn->qp_max_recv_dtos; i++, rx_desc++)  {
 		dma_addr = ib_dma_map_single(device->ib_device, (void *)rx_desc,
 					ISER_RX_PAYLOAD_SIZE, DMA_FROM_DEVICE);
 		if (ib_dma_mapping_error(device->ib_device, dma_addr))
@@ -289,18 +294,18 @@ int iser_alloc_rx_descriptors(struct iser_conn *ib_conn, struct iscsi_session *s
 		rx_sg->lkey   = device->mr->lkey;
 	}
 
-	ib_conn->rx_desc_head = 0;
+	iser_conn->rx_desc_head = 0;
 	return 0;
 
 rx_desc_dma_map_failed:
-	rx_desc = ib_conn->rx_descs;
+	rx_desc = iser_conn->rx_descs;
 	for (j = 0; j < i; j++, rx_desc++)
 		ib_dma_unmap_single(device->ib_device, rx_desc->dma_addr,
 				    ISER_RX_PAYLOAD_SIZE, DMA_FROM_DEVICE);
-	kfree(ib_conn->rx_descs);
-	ib_conn->rx_descs = NULL;
+	kfree(iser_conn->rx_descs);
+	iser_conn->rx_descs = NULL;
 rx_desc_alloc_fail:
-	iser_free_login_buf(ib_conn);
+	iser_free_login_buf(iser_conn);
 alloc_login_buf_fail:
 	device->iser_free_rdma_reg_res(ib_conn);
 create_rdma_reg_res_failed:
@@ -308,33 +313,35 @@ create_rdma_reg_res_failed:
 	return -ENOMEM;
 }
 
-void iser_free_rx_descriptors(struct iser_conn *ib_conn)
+void iser_free_rx_descriptors(struct iser_conn *iser_conn)
 {
 	int i;
 	struct iser_rx_desc *rx_desc;
+	struct ib_conn *ib_conn = &iser_conn->ib_conn;
 	struct iser_device *device = ib_conn->device;
 
-	if (!ib_conn->rx_descs)
+	if (!iser_conn->rx_descs)
 		goto free_login_buf;
 
 	if (device->iser_free_rdma_reg_res)
 		device->iser_free_rdma_reg_res(ib_conn);
 
-	rx_desc = ib_conn->rx_descs;
-	for (i = 0; i < ib_conn->qp_max_recv_dtos; i++, rx_desc++)
+	rx_desc = iser_conn->rx_descs;
+	for (i = 0; i < iser_conn->qp_max_recv_dtos; i++, rx_desc++)
 		ib_dma_unmap_single(device->ib_device, rx_desc->dma_addr,
 				    ISER_RX_PAYLOAD_SIZE, DMA_FROM_DEVICE);
-	kfree(ib_conn->rx_descs);
+	kfree(iser_conn->rx_descs);
 	/* make sure we never redo any unmapping */
-	ib_conn->rx_descs = NULL;
+	iser_conn->rx_descs = NULL;
 
 free_login_buf:
-	iser_free_login_buf(ib_conn);
+	iser_free_login_buf(iser_conn);
 }
 
 static int iser_post_rx_bufs(struct iscsi_conn *conn, struct iscsi_hdr *req)
 {
-	struct iser_conn *ib_conn = conn->dd_data;
+	struct iser_conn *iser_conn = conn->dd_data;
+	struct ib_conn *ib_conn = &iser_conn->ib_conn;
 	struct iscsi_session *session = conn->session;
 
 	iser_dbg("req op %x flags %x\n", req->opcode, req->flags);
@@ -343,34 +350,37 @@ static int iser_post_rx_bufs(struct iscsi_conn *conn, struct iscsi_hdr *req)
 		return 0;
 
 	/*
-	 * Check that there is one posted recv buffer (for the last login
-	 * response) and no posted send buffers left - they must have been
-	 * consumed during previous login phases.
+	 * Check that there is one posted recv buffer
+	 * (for the last login response).
 	 */
 	WARN_ON(ib_conn->post_recv_buf_count != 1);
-	WARN_ON(atomic_read(&ib_conn->post_send_buf_count) != 0);
 
 	if (session->discovery_sess) {
 		iser_info("Discovery session, re-using login RX buffer\n");
 		return 0;
 	} else
 		iser_info("Normal session, posting batch of RX %d buffers\n",
-			  ib_conn->min_posted_rx);
+			  iser_conn->min_posted_rx);
 
 	/* Initial post receive buffers */
-	if (iser_post_recvm(ib_conn, ib_conn->min_posted_rx))
+	if (iser_post_recvm(iser_conn, iser_conn->min_posted_rx))
 		return -ENOMEM;
 
 	return 0;
 }
 
+static inline bool iser_signal_comp(u8 sig_count)
+{
+	return ((sig_count % ISER_SIGNAL_CMD_COUNT) == 0);
+}
+
 /**
  * iser_send_command - send command PDU
  */
 int iser_send_command(struct iscsi_conn *conn,
 		      struct iscsi_task *task)
 {
-	struct iser_conn *ib_conn = conn->dd_data;
+	struct iser_conn *iser_conn = conn->dd_data;
 	struct iscsi_iser_task *iser_task = task->dd_data;
 	unsigned long edtl;
 	int err;
@@ -378,12 +388,13 @@ int iser_send_command(struct iscsi_conn *conn,
 	struct iscsi_scsi_req *hdr = (struct iscsi_scsi_req *)task->hdr;
 	struct scsi_cmnd *sc  =  task->sc;
 	struct iser_tx_desc *tx_desc = &iser_task->desc;
+	u8 sig_count = ++iser_conn->ib_conn.sig_count;
 
 	edtl = ntohl(hdr->data_length);
 
 	/* build the tx desc regd header and add it to the tx desc dto */
 	tx_desc->type = ISCSI_TX_SCSI_COMMAND;
-	iser_create_send_desc(ib_conn, tx_desc);
+	iser_create_send_desc(iser_conn, tx_desc);
 
 	if (hdr->flags & ISCSI_FLAG_CMD_READ) {
 		data_buf = &iser_task->data[ISER_DIR_IN];
@@ -423,7 +434,8 @@ int iser_send_command(struct iscsi_conn *conn,
 
 	iser_task->status = ISER_TASK_STATUS_STARTED;
 
-	err = iser_post_send(ib_conn, tx_desc);
+	err = iser_post_send(&iser_conn->ib_conn, tx_desc,
+			     iser_signal_comp(sig_count));
 	if (!err)
 		return 0;
 
@@ -439,7 +451,7 @@ int iser_send_data_out(struct iscsi_conn *conn,
 		       struct iscsi_task *task,
 		       struct iscsi_data *hdr)
 {
-	struct iser_conn *ib_conn = conn->dd_data;
+	struct iser_conn *iser_conn = conn->dd_data;
 	struct iscsi_iser_task *iser_task = task->dd_data;
 	struct iser_tx_desc *tx_desc = NULL;
 	struct iser_regd_buf *regd_buf;
@@ -488,7 +500,7 @@ int iser_send_data_out(struct iscsi_conn *conn,
 		 itt, buf_offset, data_seg_len);
 
 
-	err = iser_post_send(ib_conn, tx_desc);
+	err = iser_post_send(&iser_conn->ib_conn, tx_desc, true);
 	if (!err)
 		return 0;
 
@@ -501,7 +513,7 @@ send_data_out_error:
 int iser_send_control(struct iscsi_conn *conn,
 		      struct iscsi_task *task)
 {
-	struct iser_conn *ib_conn = conn->dd_data;
+	struct iser_conn *iser_conn = conn->dd_data;
 	struct iscsi_iser_task *iser_task = task->dd_data;
 	struct iser_tx_desc *mdesc = &iser_task->desc;
 	unsigned long data_seg_len;
@@ -510,9 +522,9 @@ int iser_send_control(struct iscsi_conn *conn,
 
 	/* build the tx desc regd header and add it to the tx desc dto */
 	mdesc->type = ISCSI_TX_CONTROL;
-	iser_create_send_desc(ib_conn, mdesc);
+	iser_create_send_desc(iser_conn, mdesc);
 
-	device = ib_conn->device;
+	device = iser_conn->ib_conn.device;
 
 	data_seg_len = ntoh24(task->hdr->dlength);
 
@@ -524,16 +536,16 @@ int iser_send_control(struct iscsi_conn *conn,
 		}
 
 		ib_dma_sync_single_for_cpu(device->ib_device,
-			ib_conn->login_req_dma, task->data_count,
+			iser_conn->login_req_dma, task->data_count,
 			DMA_TO_DEVICE);
 
-		memcpy(ib_conn->login_req_buf, task->data, task->data_count);
+		memcpy(iser_conn->login_req_buf, task->data, task->data_count);
 
 		ib_dma_sync_single_for_device(device->ib_device,
-			ib_conn->login_req_dma, task->data_count,
+			iser_conn->login_req_dma, task->data_count,
 			DMA_TO_DEVICE);
 
-		tx_dsg->addr    = ib_conn->login_req_dma;
+		tx_dsg->addr    = iser_conn->login_req_dma;
 		tx_dsg->length  = task->data_count;
 		tx_dsg->lkey    = device->mr->lkey;
 		mdesc->num_sge = 2;
@@ -542,7 +554,7 @@ int iser_send_control(struct iscsi_conn *conn,
 	if (task == conn->login_task) {
 		iser_dbg("op %x dsl %lx, posting login rx buffer\n",
 			 task->hdr->opcode, data_seg_len);
-		err = iser_post_recvl(ib_conn);
+		err = iser_post_recvl(iser_conn);
 		if (err)
 			goto send_control_error;
 		err = iser_post_rx_bufs(conn, task->hdr);
@@ -550,7 +562,7 @@ int iser_send_control(struct iscsi_conn *conn,
 			goto send_control_error;
 	}
 
-	err = iser_post_send(ib_conn, mdesc);
+	err = iser_post_send(&iser_conn->ib_conn, mdesc, true);
 	if (!err)
 		return 0;
 
@@ -564,15 +576,17 @@ send_control_error:
  */
 void iser_rcv_completion(struct iser_rx_desc *rx_desc,
 			 unsigned long rx_xfer_len,
-			 struct iser_conn *ib_conn)
+			 struct ib_conn *ib_conn)
 {
+	struct iser_conn *iser_conn = container_of(ib_conn, struct iser_conn,
+						   ib_conn);
 	struct iscsi_hdr *hdr;
 	u64 rx_dma;
 	int rx_buflen, outstanding, count, err;
 
 	/* differentiate between login to all other PDUs */
-	if ((char *)rx_desc == ib_conn->login_resp_buf) {
-		rx_dma = ib_conn->login_resp_dma;
+	if ((char *)rx_desc == iser_conn->login_resp_buf) {
+		rx_dma = iser_conn->login_resp_dma;
 		rx_buflen = ISER_RX_LOGIN_SIZE;
 	} else {
 		rx_dma = rx_desc->dma_addr;
@@ -580,14 +594,14 @@ void iser_rcv_completion(struct iser_rx_desc *rx_desc,
 	}
 
 	ib_dma_sync_single_for_cpu(ib_conn->device->ib_device, rx_dma,
-			rx_buflen, DMA_FROM_DEVICE);
+				   rx_buflen, DMA_FROM_DEVICE);
 
 	hdr = &rx_desc->iscsi_header;
 
 	iser_dbg("op 0x%x itt 0x%x dlen %d\n", hdr->opcode,
 			hdr->itt, (int)(rx_xfer_len - ISER_HEADERS_LEN));
 
-	iscsi_iser_recv(ib_conn->iscsi_conn, hdr, rx_desc->data,
+	iscsi_iser_recv(iser_conn->iscsi_conn, hdr, rx_desc->data,
 			rx_xfer_len - ISER_HEADERS_LEN);
 
 	ib_dma_sync_single_for_device(ib_conn->device->ib_device, rx_dma,
@@ -599,21 +613,21 @@ void iser_rcv_completion(struct iser_rx_desc *rx_desc,
 	 * for the posted rx bufs refcount to become zero handles everything   */
 	ib_conn->post_recv_buf_count--;
 
-	if (rx_dma == ib_conn->login_resp_dma)
+	if (rx_dma == iser_conn->login_resp_dma)
 		return;
 
 	outstanding = ib_conn->post_recv_buf_count;
-	if (outstanding + ib_conn->min_posted_rx <= ib_conn->qp_max_recv_dtos) {
-		count = min(ib_conn->qp_max_recv_dtos - outstanding,
-						ib_conn->min_posted_rx);
-		err = iser_post_recvm(ib_conn, count);
+	if (outstanding + iser_conn->min_posted_rx <= iser_conn->qp_max_recv_dtos) {
+		count = min(iser_conn->qp_max_recv_dtos - outstanding,
+			    iser_conn->min_posted_rx);
+		err = iser_post_recvm(iser_conn, count);
 		if (err)
 			iser_err("posting %d rx bufs err %d\n", count, err);
 	}
 }
 
 void iser_snd_completion(struct iser_tx_desc *tx_desc,
-			struct iser_conn *ib_conn)
+			struct ib_conn *ib_conn)
 {
 	struct iscsi_task *task;
 	struct iser_device *device = ib_conn->device;
@@ -625,8 +639,6 @@ void iser_snd_completion(struct iser_tx_desc *tx_desc,
 		tx_desc = NULL;
 	}
 
-	atomic_dec(&ib_conn->post_send_buf_count);
-
 	if (tx_desc && tx_desc->type == ISCSI_TX_CONTROL) {
 		/* this arithmetic is legal by libiscsi dd_data allocation */
 		task = (void *) ((long)(void *)tx_desc -
@@ -658,7 +670,7 @@ void iser_task_rdma_init(struct iscsi_iser_task *iser_task)
 
 void iser_task_rdma_finalize(struct iscsi_iser_task *iser_task)
 {
-	struct iser_device *device = iser_task->ib_conn->device;
+	struct iser_device *device = iser_task->iser_conn->ib_conn.device;
 	int is_rdma_data_aligned = 1;
 	int is_rdma_prot_aligned = 1;
 	int prot_count = scsi_prot_sg_count(iser_task->sc);
diff --git a/drivers/infiniband/ulp/iser/iser_memory.c b/drivers/infiniband/ulp/iser/iser_memory.c
index 47acd3ad3a17..abce9339333f 100644
--- a/drivers/infiniband/ulp/iser/iser_memory.c
+++ b/drivers/infiniband/ulp/iser/iser_memory.c
@@ -49,7 +49,7 @@ static int iser_start_rdma_unaligned_sg(struct iscsi_iser_task *iser_task,
 					struct iser_data_buf *data_copy,
 					enum iser_data_dir cmd_dir)
 {
-	struct ib_device *dev = iser_task->ib_conn->device->ib_device;
+	struct ib_device *dev = iser_task->iser_conn->ib_conn.device->ib_device;
 	struct scatterlist *sgl = (struct scatterlist *)data->buf;
 	struct scatterlist *sg;
 	char *mem = NULL;
@@ -73,7 +73,6 @@ static int iser_start_rdma_unaligned_sg(struct iscsi_iser_task *iser_task,
 
 	if (cmd_dir == ISER_DIR_OUT) {
 		/* copy the unaligned sg the buffer which is used for RDMA */
-		int i;
 		char *p, *from;
 
 		sgl = (struct scatterlist *)data->buf;
@@ -116,7 +115,7 @@ void iser_finalize_rdma_unaligned_sg(struct iscsi_iser_task *iser_task,
 	struct ib_device *dev;
 	unsigned long  cmd_data_len;
 
-	dev = iser_task->ib_conn->device->ib_device;
+	dev = iser_task->iser_conn->ib_conn.device->ib_device;
 
 	ib_dma_unmap_sg(dev, &data_copy->sg_single, 1,
 			(cmd_dir == ISER_DIR_OUT) ?
@@ -322,7 +321,7 @@ int iser_dma_map_task_data(struct iscsi_iser_task *iser_task,
 	struct ib_device *dev;
 
 	iser_task->dir[iser_dir] = 1;
-	dev = iser_task->ib_conn->device->ib_device;
+	dev = iser_task->iser_conn->ib_conn.device->ib_device;
 
 	data->dma_nents = ib_dma_map_sg(dev, data->buf, data->size, dma_dir);
 	if (data->dma_nents == 0) {
@@ -337,7 +336,7 @@ void iser_dma_unmap_task_data(struct iscsi_iser_task *iser_task,
 {
 	struct ib_device *dev;
 
-	dev = iser_task->ib_conn->device->ib_device;
+	dev = iser_task->iser_conn->ib_conn.device->ib_device;
 	ib_dma_unmap_sg(dev, data->buf, data->size, DMA_FROM_DEVICE);
 }
 
@@ -348,7 +347,7 @@ static int fall_to_bounce_buf(struct iscsi_iser_task *iser_task,
 			      enum iser_data_dir cmd_dir,
 			      int aligned_len)
 {
-	struct iscsi_conn    *iscsi_conn = iser_task->ib_conn->iscsi_conn;
+	struct iscsi_conn    *iscsi_conn = iser_task->iser_conn->iscsi_conn;
 
 	iscsi_conn->fmr_unalign_cnt++;
 	iser_warn("rdma alignment violation (%d/%d aligned) or FMR not supported\n",
@@ -377,7 +376,7 @@ static int fall_to_bounce_buf(struct iscsi_iser_task *iser_task,
 int iser_reg_rdma_mem_fmr(struct iscsi_iser_task *iser_task,
 			  enum iser_data_dir cmd_dir)
 {
-	struct iser_conn     *ib_conn = iser_task->ib_conn;
+	struct ib_conn *ib_conn = &iser_task->iser_conn->ib_conn;
 	struct iser_device   *device = ib_conn->device;
 	struct ib_device     *ibdev = device->ib_device;
 	struct iser_data_buf *mem = &iser_task->data[cmd_dir];
@@ -409,7 +408,6 @@ int iser_reg_rdma_mem_fmr(struct iscsi_iser_task *iser_task,
 		regd_buf->reg.rkey = device->mr->rkey;
 		regd_buf->reg.len  = ib_sg_dma_len(ibdev, &sg[0]);
 		regd_buf->reg.va   = ib_sg_dma_address(ibdev, &sg[0]);
-		regd_buf->reg.is_mr = 0;
 
 		iser_dbg("PHYSICAL Mem.register: lkey: 0x%08X rkey: 0x%08X  "
 			 "va: 0x%08lX sz: %ld]\n",
@@ -432,7 +430,7 @@ int iser_reg_rdma_mem_fmr(struct iscsi_iser_task *iser_task,
 				 ib_conn->fmr.page_vec->offset);
 			for (i = 0; i < ib_conn->fmr.page_vec->length; i++)
 				iser_err("page_vec[%d] = 0x%llx\n", i,
-					 (unsigned long long) ib_conn->fmr.page_vec->pages[i]);
+					 (unsigned long long)ib_conn->fmr.page_vec->pages[i]);
 		}
 		if (err)
 			return err;
@@ -440,92 +438,80 @@ int iser_reg_rdma_mem_fmr(struct iscsi_iser_task *iser_task,
 	return 0;
 }
 
-static inline enum ib_t10_dif_type
-scsi2ib_prot_type(unsigned char prot_type)
+static void
+iser_set_dif_domain(struct scsi_cmnd *sc, struct ib_sig_attrs *sig_attrs,
+		    struct ib_sig_domain *domain)
 {
-	switch (prot_type) {
-	case SCSI_PROT_DIF_TYPE0:
-		return IB_T10DIF_NONE;
-	case SCSI_PROT_DIF_TYPE1:
-		return IB_T10DIF_TYPE1;
-	case SCSI_PROT_DIF_TYPE2:
-		return IB_T10DIF_TYPE2;
-	case SCSI_PROT_DIF_TYPE3:
-		return IB_T10DIF_TYPE3;
-	default:
-		return IB_T10DIF_NONE;
-	}
-}
-
+	domain->sig_type = IB_SIG_TYPE_T10_DIF;
+	domain->sig.dif.pi_interval = scsi_prot_interval(sc);
+	domain->sig.dif.ref_tag = scsi_prot_ref_tag(sc);
+	/*
+	 * At the moment we hard code those, but in the future
+	 * we will take them from sc.
+	 */
+	domain->sig.dif.apptag_check_mask = 0xffff;
+	domain->sig.dif.app_escape = true;
+	domain->sig.dif.ref_escape = true;
+	if (sc->prot_flags & SCSI_PROT_REF_INCREMENT)
+		domain->sig.dif.ref_remap = true;
+};
 
 static int
 iser_set_sig_attrs(struct scsi_cmnd *sc, struct ib_sig_attrs *sig_attrs)
 {
-	unsigned char scsi_ptype = scsi_get_prot_type(sc);
-
-	sig_attrs->mem.sig_type = IB_SIG_TYPE_T10_DIF;
-	sig_attrs->wire.sig_type = IB_SIG_TYPE_T10_DIF;
-	sig_attrs->mem.sig.dif.pi_interval = sc->device->sector_size;
-	sig_attrs->wire.sig.dif.pi_interval = sc->device->sector_size;
-
 	switch (scsi_get_prot_op(sc)) {
 	case SCSI_PROT_WRITE_INSERT:
 	case SCSI_PROT_READ_STRIP:
-		sig_attrs->mem.sig.dif.type = IB_T10DIF_NONE;
-		sig_attrs->wire.sig.dif.type = scsi2ib_prot_type(scsi_ptype);
+		sig_attrs->mem.sig_type = IB_SIG_TYPE_NONE;
+		iser_set_dif_domain(sc, sig_attrs, &sig_attrs->wire);
 		sig_attrs->wire.sig.dif.bg_type = IB_T10DIF_CRC;
-		sig_attrs->wire.sig.dif.ref_tag = scsi_get_lba(sc) &
-						  0xffffffff;
 		break;
 	case SCSI_PROT_READ_INSERT:
 	case SCSI_PROT_WRITE_STRIP:
-		sig_attrs->mem.sig.dif.type = scsi2ib_prot_type(scsi_ptype);
-		sig_attrs->mem.sig.dif.bg_type = IB_T10DIF_CRC;
-		sig_attrs->mem.sig.dif.ref_tag = scsi_get_lba(sc) &
-						 0xffffffff;
-		sig_attrs->wire.sig.dif.type = IB_T10DIF_NONE;
+		sig_attrs->wire.sig_type = IB_SIG_TYPE_NONE;
+		iser_set_dif_domain(sc, sig_attrs, &sig_attrs->mem);
+		sig_attrs->mem.sig.dif.bg_type = sc->prot_flags & SCSI_PROT_IP_CHECKSUM ?
+						IB_T10DIF_CSUM : IB_T10DIF_CRC;
 		break;
 	case SCSI_PROT_READ_PASS:
 	case SCSI_PROT_WRITE_PASS:
-		sig_attrs->mem.sig.dif.type = scsi2ib_prot_type(scsi_ptype);
-		sig_attrs->mem.sig.dif.bg_type = IB_T10DIF_CRC;
-		sig_attrs->mem.sig.dif.ref_tag = scsi_get_lba(sc) &
-						 0xffffffff;
-		sig_attrs->wire.sig.dif.type = scsi2ib_prot_type(scsi_ptype);
+		iser_set_dif_domain(sc, sig_attrs, &sig_attrs->wire);
 		sig_attrs->wire.sig.dif.bg_type = IB_T10DIF_CRC;
-		sig_attrs->wire.sig.dif.ref_tag = scsi_get_lba(sc) &
-						  0xffffffff;
+		iser_set_dif_domain(sc, sig_attrs, &sig_attrs->mem);
+		sig_attrs->mem.sig.dif.bg_type = sc->prot_flags & SCSI_PROT_IP_CHECKSUM ?
+						IB_T10DIF_CSUM : IB_T10DIF_CRC;
 		break;
 	default:
 		iser_err("Unsupported PI operation %d\n",
 			 scsi_get_prot_op(sc));
 		return -EINVAL;
 	}
+
 	return 0;
 }
 
-
-static int
+static inline void
 iser_set_prot_checks(struct scsi_cmnd *sc, u8 *mask)
 {
-	switch (scsi_get_prot_type(sc)) {
-	case SCSI_PROT_DIF_TYPE0:
-		*mask = 0x0;
-		break;
-	case SCSI_PROT_DIF_TYPE1:
-	case SCSI_PROT_DIF_TYPE2:
-		*mask = ISER_CHECK_GUARD | ISER_CHECK_REFTAG;
-		break;
-	case SCSI_PROT_DIF_TYPE3:
-		*mask = ISER_CHECK_GUARD;
-		break;
-	default:
-		iser_err("Unsupported protection type %d\n",
-			 scsi_get_prot_type(sc));
-		return -EINVAL;
-	}
+	*mask = 0;
+	if (sc->prot_flags & SCSI_PROT_REF_CHECK)
+		*mask |= ISER_CHECK_REFTAG;
+	if (sc->prot_flags & SCSI_PROT_GUARD_CHECK)
+		*mask |= ISER_CHECK_GUARD;
+}
 
-	return 0;
+static void
+iser_inv_rkey(struct ib_send_wr *inv_wr, struct ib_mr *mr)
+{
+	u32 rkey;
+
+	memset(inv_wr, 0, sizeof(*inv_wr));
+	inv_wr->opcode = IB_WR_LOCAL_INV;
+	inv_wr->wr_id = ISER_FASTREG_LI_WRID;
+	inv_wr->ex.invalidate_rkey = mr->rkey;
+
+	rkey = ib_inc_rkey(mr->rkey);
+	ib_update_fast_reg_key(mr, rkey);
 }
 
 static int
@@ -533,32 +519,23 @@ iser_reg_sig_mr(struct iscsi_iser_task *iser_task,
 		struct fast_reg_descriptor *desc, struct ib_sge *data_sge,
 		struct ib_sge *prot_sge, struct ib_sge *sig_sge)
 {
-	struct iser_conn *ib_conn = iser_task->ib_conn;
+	struct ib_conn *ib_conn = &iser_task->iser_conn->ib_conn;
 	struct iser_pi_context *pi_ctx = desc->pi_ctx;
 	struct ib_send_wr sig_wr, inv_wr;
 	struct ib_send_wr *bad_wr, *wr = NULL;
 	struct ib_sig_attrs sig_attrs;
 	int ret;
-	u32 key;
 
 	memset(&sig_attrs, 0, sizeof(sig_attrs));
 	ret = iser_set_sig_attrs(iser_task->sc, &sig_attrs);
 	if (ret)
 		goto err;
 
-	ret = iser_set_prot_checks(iser_task->sc, &sig_attrs.check_mask);
-	if (ret)
-		goto err;
+	iser_set_prot_checks(iser_task->sc, &sig_attrs.check_mask);
 
 	if (!(desc->reg_indicators & ISER_SIG_KEY_VALID)) {
-		memset(&inv_wr, 0, sizeof(inv_wr));
-		inv_wr.opcode = IB_WR_LOCAL_INV;
-		inv_wr.wr_id = ISER_FASTREG_LI_WRID;
-		inv_wr.ex.invalidate_rkey = pi_ctx->sig_mr->rkey;
+		iser_inv_rkey(&inv_wr, pi_ctx->sig_mr);
 		wr = &inv_wr;
-		/* Bump the key */
-		key = (u8)(pi_ctx->sig_mr->rkey & 0x000000FF);
-		ib_update_fast_reg_key(pi_ctx->sig_mr, ++key);
 	}
 
 	memset(&sig_wr, 0, sizeof(sig_wr));
@@ -588,12 +565,7 @@ iser_reg_sig_mr(struct iscsi_iser_task *iser_task,
 
 	sig_sge->lkey = pi_ctx->sig_mr->lkey;
 	sig_sge->addr = 0;
-	sig_sge->length = data_sge->length + prot_sge->length;
-	if (scsi_get_prot_op(iser_task->sc) == SCSI_PROT_WRITE_INSERT ||
-	    scsi_get_prot_op(iser_task->sc) == SCSI_PROT_READ_STRIP) {
-		sig_sge->length += (data_sge->length /
-				   iser_task->sc->device->sector_size) * 8;
-	}
+	sig_sge->length = scsi_transfer_length(iser_task->sc);
 
 	iser_dbg("sig_sge: addr: 0x%llx  length: %u lkey: 0x%x\n",
 		 sig_sge->addr, sig_sge->length,
@@ -609,14 +581,13 @@ static int iser_fast_reg_mr(struct iscsi_iser_task *iser_task,
 			    struct ib_sge *sge)
 {
 	struct fast_reg_descriptor *desc = regd_buf->reg.mem_h;
-	struct iser_conn *ib_conn = iser_task->ib_conn;
+	struct ib_conn *ib_conn = &iser_task->iser_conn->ib_conn;
 	struct iser_device *device = ib_conn->device;
 	struct ib_device *ibdev = device->ib_device;
 	struct ib_mr *mr;
 	struct ib_fast_reg_page_list *frpl;
 	struct ib_send_wr fastreg_wr, inv_wr;
 	struct ib_send_wr *bad_wr, *wr = NULL;
-	u8 key;
 	int ret, offset, size, plen;
 
 	/* if there a single dma entry, dma mr suffices */
@@ -648,14 +619,8 @@ static int iser_fast_reg_mr(struct iscsi_iser_task *iser_task,
 	}
 
 	if (!(desc->reg_indicators & ind)) {
-		memset(&inv_wr, 0, sizeof(inv_wr));
-		inv_wr.wr_id = ISER_FASTREG_LI_WRID;
-		inv_wr.opcode = IB_WR_LOCAL_INV;
-		inv_wr.ex.invalidate_rkey = mr->rkey;
+		iser_inv_rkey(&inv_wr, mr);
 		wr = &inv_wr;
-		/* Bump the key */
-		key = (u8)(mr->rkey & 0x000000FF);
-		ib_update_fast_reg_key(mr, ++key);
 	}
 
 	/* Prepare FASTREG WR */
@@ -700,7 +665,7 @@ static int iser_fast_reg_mr(struct iscsi_iser_task *iser_task,
 int iser_reg_rdma_mem_fastreg(struct iscsi_iser_task *iser_task,
 			      enum iser_data_dir cmd_dir)
 {
-	struct iser_conn *ib_conn = iser_task->ib_conn;
+	struct ib_conn *ib_conn = &iser_task->iser_conn->ib_conn;
 	struct iser_device *device = ib_conn->device;
 	struct ib_device *ibdev = device->ib_device;
 	struct iser_data_buf *mem = &iser_task->data[cmd_dir];
@@ -773,15 +738,11 @@ int iser_reg_rdma_mem_fastreg(struct iscsi_iser_task *iser_task,
 		regd_buf->reg.rkey = desc->pi_ctx->sig_mr->rkey;
 		regd_buf->reg.va = sig_sge.addr;
 		regd_buf->reg.len = sig_sge.length;
-		regd_buf->reg.is_mr = 1;
 	} else {
-		if (desc) {
+		if (desc)
 			regd_buf->reg.rkey = desc->data_mr->rkey;
-			regd_buf->reg.is_mr = 1;
-		} else {
+		else
 			regd_buf->reg.rkey = device->mr->rkey;
-			regd_buf->reg.is_mr = 0;
-		}
 
 		regd_buf->reg.lkey = data_sge.lkey;
 		regd_buf->reg.va = data_sge.addr;
diff --git a/drivers/infiniband/ulp/iser/iser_verbs.c b/drivers/infiniband/ulp/iser/iser_verbs.c
index 3bfec4bbda52..695a2704bd43 100644
--- a/drivers/infiniband/ulp/iser/iser_verbs.c
+++ b/drivers/infiniband/ulp/iser/iser_verbs.c
@@ -39,8 +39,12 @@
 #include "iscsi_iser.h"
 
 #define ISCSI_ISER_MAX_CONN	8
-#define ISER_MAX_RX_CQ_LEN	(ISER_QP_MAX_RECV_DTOS * ISCSI_ISER_MAX_CONN)
-#define ISER_MAX_TX_CQ_LEN	(ISER_QP_MAX_REQ_DTOS  * ISCSI_ISER_MAX_CONN)
+#define ISER_MAX_RX_LEN		(ISER_QP_MAX_RECV_DTOS * ISCSI_ISER_MAX_CONN)
+#define ISER_MAX_TX_LEN		(ISER_QP_MAX_REQ_DTOS  * ISCSI_ISER_MAX_CONN)
+#define ISER_MAX_CQ_LEN		(ISER_MAX_RX_LEN + ISER_MAX_TX_LEN + \
+				 ISCSI_ISER_MAX_CONN)
+
+static int iser_cq_poll_limit = 512;
 
 static void iser_cq_tasklet_fn(unsigned long data);
 static void iser_cq_callback(struct ib_cq *cq, void *cq_context);
@@ -71,9 +75,8 @@ static void iser_event_handler(struct ib_event_handler *handler,
  */
 static int iser_create_device_ib_res(struct iser_device *device)
 {
-	struct iser_cq_desc *cq_desc;
 	struct ib_device_attr *dev_attr = &device->dev_attr;
-	int ret, i;
+	int ret, i, max_cqe;
 
 	ret = ib_query_device(device->ib_device, dev_attr);
 	if (ret) {
@@ -101,51 +104,43 @@ static int iser_create_device_ib_res(struct iser_device *device)
 		return -1;
 	}
 
-	device->cqs_used = min(ISER_MAX_CQ, device->ib_device->num_comp_vectors);
-	iser_info("using %d CQs, device %s supports %d vectors\n",
-		  device->cqs_used, device->ib_device->name,
-		  device->ib_device->num_comp_vectors);
+	device->comps_used = min_t(int, num_online_cpus(),
+				 device->ib_device->num_comp_vectors);
+
+	device->comps = kcalloc(device->comps_used, sizeof(*device->comps),
+				GFP_KERNEL);
+	if (!device->comps)
+		goto comps_err;
 
-	device->cq_desc = kmalloc(sizeof(struct iser_cq_desc) * device->cqs_used,
-				  GFP_KERNEL);
-	if (device->cq_desc == NULL)
-		goto cq_desc_err;
-	cq_desc = device->cq_desc;
+	max_cqe = min(ISER_MAX_CQ_LEN, dev_attr->max_cqe);
+
+	iser_info("using %d CQs, device %s supports %d vectors max_cqe %d\n",
+		  device->comps_used, device->ib_device->name,
+		  device->ib_device->num_comp_vectors, max_cqe);
 
 	device->pd = ib_alloc_pd(device->ib_device);
 	if (IS_ERR(device->pd))
 		goto pd_err;
 
-	for (i = 0; i < device->cqs_used; i++) {
-		cq_desc[i].device   = device;
-		cq_desc[i].cq_index = i;
-
-		device->rx_cq[i] = ib_create_cq(device->ib_device,
-					  iser_cq_callback,
-					  iser_cq_event_callback,
-					  (void *)&cq_desc[i],
-					  ISER_MAX_RX_CQ_LEN, i);
-		if (IS_ERR(device->rx_cq[i])) {
-			device->rx_cq[i] = NULL;
-			goto cq_err;
-		}
-
-		device->tx_cq[i] = ib_create_cq(device->ib_device,
-					  NULL, iser_cq_event_callback,
-					  (void *)&cq_desc[i],
-					  ISER_MAX_TX_CQ_LEN, i);
-
-		if (IS_ERR(device->tx_cq[i])) {
-			device->tx_cq[i] = NULL;
+	for (i = 0; i < device->comps_used; i++) {
+		struct iser_comp *comp = &device->comps[i];
+
+		comp->device = device;
+		comp->cq = ib_create_cq(device->ib_device,
+					iser_cq_callback,
+					iser_cq_event_callback,
+					(void *)comp,
+					max_cqe, i);
+		if (IS_ERR(comp->cq)) {
+			comp->cq = NULL;
 			goto cq_err;
 		}
 
-		if (ib_req_notify_cq(device->rx_cq[i], IB_CQ_NEXT_COMP))
+		if (ib_req_notify_cq(comp->cq, IB_CQ_NEXT_COMP))
 			goto cq_err;
 
-		tasklet_init(&device->cq_tasklet[i],
-			     iser_cq_tasklet_fn,
-			(unsigned long)&cq_desc[i]);
+		tasklet_init(&comp->tasklet, iser_cq_tasklet_fn,
+			     (unsigned long)comp);
 	}
 
 	device->mr = ib_get_dma_mr(device->pd, IB_ACCESS_LOCAL_WRITE |
@@ -164,19 +159,19 @@ static int iser_create_device_ib_res(struct iser_device *device)
 handler_err:
 	ib_dereg_mr(device->mr);
 dma_mr_err:
-	for (i = 0; i < device->cqs_used; i++)
-		tasklet_kill(&device->cq_tasklet[i]);
+	for (i = 0; i < device->comps_used; i++)
+		tasklet_kill(&device->comps[i].tasklet);
 cq_err:
-	for (i = 0; i < device->cqs_used; i++) {
-		if (device->tx_cq[i])
-			ib_destroy_cq(device->tx_cq[i]);
-		if (device->rx_cq[i])
-			ib_destroy_cq(device->rx_cq[i]);
+	for (i = 0; i < device->comps_used; i++) {
+		struct iser_comp *comp = &device->comps[i];
+
+		if (comp->cq)
+			ib_destroy_cq(comp->cq);
 	}
 	ib_dealloc_pd(device->pd);
 pd_err:
-	kfree(device->cq_desc);
-cq_desc_err:
+	kfree(device->comps);
+comps_err:
 	iser_err("failed to allocate an IB resource\n");
 	return -1;
 }
@@ -190,19 +185,20 @@ static void iser_free_device_ib_res(struct iser_device *device)
 	int i;
 	BUG_ON(device->mr == NULL);
 
-	for (i = 0; i < device->cqs_used; i++) {
-		tasklet_kill(&device->cq_tasklet[i]);
-		(void)ib_destroy_cq(device->tx_cq[i]);
-		(void)ib_destroy_cq(device->rx_cq[i]);
-		device->tx_cq[i] = NULL;
-		device->rx_cq[i] = NULL;
+	for (i = 0; i < device->comps_used; i++) {
+		struct iser_comp *comp = &device->comps[i];
+
+		tasklet_kill(&comp->tasklet);
+		ib_destroy_cq(comp->cq);
+		comp->cq = NULL;
 	}
 
 	(void)ib_unregister_event_handler(&device->event_handler);
 	(void)ib_dereg_mr(device->mr);
 	(void)ib_dealloc_pd(device->pd);
 
-	kfree(device->cq_desc);
+	kfree(device->comps);
+	device->comps = NULL;
 
 	device->mr = NULL;
 	device->pd = NULL;
@@ -213,7 +209,7 @@ static void iser_free_device_ib_res(struct iser_device *device)
  *
  * returns 0 on success, or errno code on failure
  */
-int iser_create_fmr_pool(struct iser_conn *ib_conn, unsigned cmds_max)
+int iser_create_fmr_pool(struct ib_conn *ib_conn, unsigned cmds_max)
 {
 	struct iser_device *device = ib_conn->device;
 	struct ib_fmr_pool_param params;
@@ -263,7 +259,7 @@ int iser_create_fmr_pool(struct iser_conn *ib_conn, unsigned cmds_max)
 /**
  * iser_free_fmr_pool - releases the FMR pool and page vec
  */
-void iser_free_fmr_pool(struct iser_conn *ib_conn)
+void iser_free_fmr_pool(struct ib_conn *ib_conn)
 {
 	iser_info("freeing conn %p fmr pool %p\n",
 		  ib_conn, ib_conn->fmr.pool);
@@ -367,10 +363,10 @@ fast_reg_mr_failure:
  * for fast registration work requests.
  * returns 0 on success, or errno code on failure
  */
-int iser_create_fastreg_pool(struct iser_conn *ib_conn, unsigned cmds_max)
+int iser_create_fastreg_pool(struct ib_conn *ib_conn, unsigned cmds_max)
 {
-	struct iser_device	*device = ib_conn->device;
-	struct fast_reg_descriptor	*desc;
+	struct iser_device *device = ib_conn->device;
+	struct fast_reg_descriptor *desc;
 	int i, ret;
 
 	INIT_LIST_HEAD(&ib_conn->fastreg.pool);
@@ -406,7 +402,7 @@ err:
 /**
  * iser_free_fastreg_pool - releases the pool of fast_reg descriptors
  */
-void iser_free_fastreg_pool(struct iser_conn *ib_conn)
+void iser_free_fastreg_pool(struct ib_conn *ib_conn)
 {
 	struct fast_reg_descriptor *desc, *tmp;
 	int i = 0;
@@ -440,9 +436,12 @@ void iser_free_fastreg_pool(struct iser_conn *ib_conn)
  *
  * returns 0 on success, -1 on failure
  */
-static int iser_create_ib_conn_res(struct iser_conn *ib_conn)
+static int iser_create_ib_conn_res(struct ib_conn *ib_conn)
 {
+	struct iser_conn *iser_conn = container_of(ib_conn, struct iser_conn,
+						   ib_conn);
 	struct iser_device	*device;
+	struct ib_device_attr *dev_attr;
 	struct ib_qp_init_attr	init_attr;
 	int			ret = -ENOMEM;
 	int index, min_index = 0;
@@ -450,33 +449,48 @@ static int iser_create_ib_conn_res(struct iser_conn *ib_conn)
 	BUG_ON(ib_conn->device == NULL);
 
 	device = ib_conn->device;
+	dev_attr = &device->dev_attr;
 
 	memset(&init_attr, 0, sizeof init_attr);
 
 	mutex_lock(&ig.connlist_mutex);
 	/* select the CQ with the minimal number of usages */
-	for (index = 0; index < device->cqs_used; index++)
-		if (device->cq_active_qps[index] <
-		    device->cq_active_qps[min_index])
+	for (index = 0; index < device->comps_used; index++) {
+		if (device->comps[index].active_qps <
+		    device->comps[min_index].active_qps)
 			min_index = index;
-	device->cq_active_qps[min_index]++;
+	}
+	ib_conn->comp = &device->comps[min_index];
+	ib_conn->comp->active_qps++;
 	mutex_unlock(&ig.connlist_mutex);
 	iser_info("cq index %d used for ib_conn %p\n", min_index, ib_conn);
 
 	init_attr.event_handler = iser_qp_event_callback;
 	init_attr.qp_context	= (void *)ib_conn;
-	init_attr.send_cq	= device->tx_cq[min_index];
-	init_attr.recv_cq	= device->rx_cq[min_index];
+	init_attr.send_cq	= ib_conn->comp->cq;
+	init_attr.recv_cq	= ib_conn->comp->cq;
 	init_attr.cap.max_recv_wr  = ISER_QP_MAX_RECV_DTOS;
 	init_attr.cap.max_send_sge = 2;
 	init_attr.cap.max_recv_sge = 1;
 	init_attr.sq_sig_type	= IB_SIGNAL_REQ_WR;
 	init_attr.qp_type	= IB_QPT_RC;
 	if (ib_conn->pi_support) {
-		init_attr.cap.max_send_wr = ISER_QP_SIG_MAX_REQ_DTOS;
+		init_attr.cap.max_send_wr = ISER_QP_SIG_MAX_REQ_DTOS + 1;
 		init_attr.create_flags |= IB_QP_CREATE_SIGNATURE_EN;
+		iser_conn->max_cmds =
+			ISER_GET_MAX_XMIT_CMDS(ISER_QP_SIG_MAX_REQ_DTOS);
 	} else {
-		init_attr.cap.max_send_wr  = ISER_QP_MAX_REQ_DTOS;
+		if (dev_attr->max_qp_wr > ISER_QP_MAX_REQ_DTOS) {
+			init_attr.cap.max_send_wr  = ISER_QP_MAX_REQ_DTOS + 1;
+			iser_conn->max_cmds =
+				ISER_GET_MAX_XMIT_CMDS(ISER_QP_MAX_REQ_DTOS);
+		} else {
+			init_attr.cap.max_send_wr = dev_attr->max_qp_wr;
+			iser_conn->max_cmds =
+				ISER_GET_MAX_XMIT_CMDS(dev_attr->max_qp_wr);
+			iser_dbg("device %s supports max_send_wr %d\n",
+				 device->ib_device->name, dev_attr->max_qp_wr);
+		}
 	}
 
 	ret = rdma_create_qp(ib_conn->cma_id, device->pd, &init_attr);
@@ -490,32 +504,12 @@ static int iser_create_ib_conn_res(struct iser_conn *ib_conn)
 	return ret;
 
 out_err:
+	mutex_lock(&ig.connlist_mutex);
+	ib_conn->comp->active_qps--;
+	mutex_unlock(&ig.connlist_mutex);
 	iser_err("unable to alloc mem or create resource, err %d\n", ret);
-	return ret;
-}
-
-/**
- * releases the QP object
- */
-static void iser_free_ib_conn_res(struct iser_conn *ib_conn)
-{
-	int cq_index;
-	BUG_ON(ib_conn == NULL);
-
-	iser_info("freeing conn %p cma_id %p qp %p\n",
-		  ib_conn, ib_conn->cma_id,
-		  ib_conn->qp);
-
-	/* qp is created only once both addr & route are resolved */
-
-	if (ib_conn->qp != NULL) {
-		cq_index = ((struct iser_cq_desc *)ib_conn->qp->recv_cq->cq_context)->cq_index;
-		ib_conn->device->cq_active_qps[cq_index]--;
-
-		rdma_destroy_qp(ib_conn->cma_id);
-	}
 
-	ib_conn->qp	  = NULL;
+	return ret;
 }
 
 /**
@@ -572,88 +566,146 @@ static void iser_device_try_release(struct iser_device *device)
 /**
  * Called with state mutex held
  **/
-static int iser_conn_state_comp_exch(struct iser_conn *ib_conn,
-				     enum iser_ib_conn_state comp,
-				     enum iser_ib_conn_state exch)
+static int iser_conn_state_comp_exch(struct iser_conn *iser_conn,
+				     enum iser_conn_state comp,
+				     enum iser_conn_state exch)
 {
 	int ret;
 
-	if ((ret = (ib_conn->state == comp)))
-		ib_conn->state = exch;
+	ret = (iser_conn->state == comp);
+	if (ret)
+		iser_conn->state = exch;
+
 	return ret;
 }
 
 void iser_release_work(struct work_struct *work)
 {
-	struct iser_conn *ib_conn;
-	int rc;
+	struct iser_conn *iser_conn;
 
-	ib_conn = container_of(work, struct iser_conn, release_work);
+	iser_conn = container_of(work, struct iser_conn, release_work);
 
-	/* wait for .conn_stop callback */
-	rc = wait_for_completion_timeout(&ib_conn->stop_completion, 30 * HZ);
-	WARN_ON(rc == 0);
+	/* Wait for conn_stop to complete */
+	wait_for_completion(&iser_conn->stop_completion);
+	/* Wait for IB resouces cleanup to complete */
+	wait_for_completion(&iser_conn->ib_completion);
 
-	/* wait for the qp`s post send and post receive buffers to empty */
-	rc = wait_for_completion_timeout(&ib_conn->flush_completion, 30 * HZ);
-	WARN_ON(rc == 0);
+	mutex_lock(&iser_conn->state_mutex);
+	iser_conn->state = ISER_CONN_DOWN;
+	mutex_unlock(&iser_conn->state_mutex);
 
-	ib_conn->state = ISER_CONN_DOWN;
+	iser_conn_release(iser_conn);
+}
 
-	mutex_lock(&ib_conn->state_mutex);
-	ib_conn->state = ISER_CONN_DOWN;
-	mutex_unlock(&ib_conn->state_mutex);
+/**
+ * iser_free_ib_conn_res - release IB related resources
+ * @iser_conn: iser connection struct
+ * @destroy_device: indicator if we need to try to release
+ *     the iser device (only iscsi shutdown and DEVICE_REMOVAL
+ *     will use this.
+ *
+ * This routine is called with the iser state mutex held
+ * so the cm_id removal is out of here. It is Safe to
+ * be invoked multiple times.
+ */
+static void iser_free_ib_conn_res(struct iser_conn *iser_conn,
+				  bool destroy_device)
+{
+	struct ib_conn *ib_conn = &iser_conn->ib_conn;
+	struct iser_device *device = ib_conn->device;
+
+	iser_info("freeing conn %p cma_id %p qp %p\n",
+		  iser_conn, ib_conn->cma_id, ib_conn->qp);
 
-	iser_conn_release(ib_conn);
+	iser_free_rx_descriptors(iser_conn);
+
+	if (ib_conn->qp != NULL) {
+		ib_conn->comp->active_qps--;
+		rdma_destroy_qp(ib_conn->cma_id);
+		ib_conn->qp = NULL;
+	}
+
+	if (destroy_device && device != NULL) {
+		iser_device_try_release(device);
+		ib_conn->device = NULL;
+	}
 }
 
 /**
  * Frees all conn objects and deallocs conn descriptor
  */
-void iser_conn_release(struct iser_conn *ib_conn)
+void iser_conn_release(struct iser_conn *iser_conn)
 {
-	struct iser_device  *device = ib_conn->device;
+	struct ib_conn *ib_conn = &iser_conn->ib_conn;
 
 	mutex_lock(&ig.connlist_mutex);
-	list_del(&ib_conn->conn_list);
+	list_del(&iser_conn->conn_list);
 	mutex_unlock(&ig.connlist_mutex);
 
-	mutex_lock(&ib_conn->state_mutex);
-	BUG_ON(ib_conn->state != ISER_CONN_DOWN);
-
-	iser_free_rx_descriptors(ib_conn);
-	iser_free_ib_conn_res(ib_conn);
-	ib_conn->device = NULL;
-	/* on EVENT_ADDR_ERROR there's no device yet for this conn */
-	if (device != NULL)
-		iser_device_try_release(device);
-	mutex_unlock(&ib_conn->state_mutex);
+	mutex_lock(&iser_conn->state_mutex);
+	if (iser_conn->state != ISER_CONN_DOWN) {
+		iser_warn("iser conn %p state %d, expected state down.\n",
+			  iser_conn, iser_conn->state);
+		iser_conn->state = ISER_CONN_DOWN;
+	}
+	/*
+	 * In case we never got to bind stage, we still need to
+	 * release IB resources (which is safe to call more than once).
+	 */
+	iser_free_ib_conn_res(iser_conn, true);
+	mutex_unlock(&iser_conn->state_mutex);
 
-	/* if cma handler context, the caller actually destroy the id */
 	if (ib_conn->cma_id != NULL) {
 		rdma_destroy_id(ib_conn->cma_id);
 		ib_conn->cma_id = NULL;
 	}
-	kfree(ib_conn);
+
+	kfree(iser_conn);
 }
 
 /**
  * triggers start of the disconnect procedures and wait for them to be done
+ * Called with state mutex held
  */
-void iser_conn_terminate(struct iser_conn *ib_conn)
+int iser_conn_terminate(struct iser_conn *iser_conn)
 {
+	struct ib_conn *ib_conn = &iser_conn->ib_conn;
+	struct ib_send_wr *bad_wr;
 	int err = 0;
 
-	/* change the ib conn state only if the conn is UP, however always call
-	 * rdma_disconnect since this is the only way to cause the CMA to change
-	 * the QP state to ERROR
+	/* terminate the iser conn only if the conn state is UP */
+	if (!iser_conn_state_comp_exch(iser_conn, ISER_CONN_UP,
+				       ISER_CONN_TERMINATING))
+		return 0;
+
+	iser_info("iser_conn %p state %d\n", iser_conn, iser_conn->state);
+
+	/* suspend queuing of new iscsi commands */
+	if (iser_conn->iscsi_conn)
+		iscsi_suspend_queue(iser_conn->iscsi_conn);
+
+	/*
+	 * In case we didn't already clean up the cma_id (peer initiated
+	 * a disconnection), we need to Cause the CMA to change the QP
+	 * state to ERROR.
 	 */
+	if (ib_conn->cma_id) {
+		err = rdma_disconnect(ib_conn->cma_id);
+		if (err)
+			iser_err("Failed to disconnect, conn: 0x%p err %d\n",
+				 iser_conn, err);
+
+		/* post an indication that all flush errors were consumed */
+		err = ib_post_send(ib_conn->qp, &ib_conn->beacon, &bad_wr);
+		if (err) {
+			iser_err("conn %p failed to post beacon", ib_conn);
+			return 1;
+		}
+
+		wait_for_completion(&ib_conn->flush_comp);
+	}
 
-	iser_conn_state_comp_exch(ib_conn, ISER_CONN_UP, ISER_CONN_TERMINATING);
-	err = rdma_disconnect(ib_conn->cma_id);
-	if (err)
-		iser_err("Failed to disconnect, conn: 0x%p err %d\n",
-			 ib_conn,err);
+	return 1;
 }
 
 /**
@@ -661,10 +713,10 @@ void iser_conn_terminate(struct iser_conn *ib_conn)
  **/
 static void iser_connect_error(struct rdma_cm_id *cma_id)
 {
-	struct iser_conn *ib_conn;
+	struct iser_conn *iser_conn;
 
-	ib_conn = (struct iser_conn *)cma_id->context;
-	ib_conn->state = ISER_CONN_DOWN;
+	iser_conn = (struct iser_conn *)cma_id->context;
+	iser_conn->state = ISER_CONN_DOWN;
 }
 
 /**
@@ -673,14 +725,16 @@ static void iser_connect_error(struct rdma_cm_id *cma_id)
 static void iser_addr_handler(struct rdma_cm_id *cma_id)
 {
 	struct iser_device *device;
-	struct iser_conn   *ib_conn;
+	struct iser_conn   *iser_conn;
+	struct ib_conn   *ib_conn;
 	int    ret;
 
-	ib_conn = (struct iser_conn *)cma_id->context;
-	if (ib_conn->state != ISER_CONN_PENDING)
+	iser_conn = (struct iser_conn *)cma_id->context;
+	if (iser_conn->state != ISER_CONN_PENDING)
 		/* bailout */
 		return;
 
+	ib_conn = &iser_conn->ib_conn;
 	device = iser_device_find_by_ib_device(cma_id);
 	if (!device) {
 		iser_err("device lookup/creation failed\n");
@@ -719,14 +773,15 @@ static void iser_route_handler(struct rdma_cm_id *cma_id)
 	struct rdma_conn_param conn_param;
 	int    ret;
 	struct iser_cm_hdr req_hdr;
-	struct iser_conn *ib_conn = (struct iser_conn *)cma_id->context;
+	struct iser_conn *iser_conn = (struct iser_conn *)cma_id->context;
+	struct ib_conn *ib_conn = &iser_conn->ib_conn;
 	struct iser_device *device = ib_conn->device;
 
-	if (ib_conn->state != ISER_CONN_PENDING)
+	if (iser_conn->state != ISER_CONN_PENDING)
 		/* bailout */
 		return;
 
-	ret = iser_create_ib_conn_res((struct iser_conn *)cma_id->context);
+	ret = iser_create_ib_conn_res(ib_conn);
 	if (ret)
 		goto failure;
 
@@ -755,57 +810,60 @@ failure:
 
 static void iser_connected_handler(struct rdma_cm_id *cma_id)
 {
-	struct iser_conn *ib_conn;
+	struct iser_conn *iser_conn;
 	struct ib_qp_attr attr;
 	struct ib_qp_init_attr init_attr;
 
-	ib_conn = (struct iser_conn *)cma_id->context;
-	if (ib_conn->state != ISER_CONN_PENDING)
+	iser_conn = (struct iser_conn *)cma_id->context;
+	if (iser_conn->state != ISER_CONN_PENDING)
 		/* bailout */
 		return;
 
 	(void)ib_query_qp(cma_id->qp, &attr, ~0, &init_attr);
 	iser_info("remote qpn:%x my qpn:%x\n", attr.dest_qp_num, cma_id->qp->qp_num);
 
-	ib_conn->state = ISER_CONN_UP;
-	complete(&ib_conn->up_completion);
+	iser_conn->state = ISER_CONN_UP;
+	complete(&iser_conn->up_completion);
 }
 
 static void iser_disconnected_handler(struct rdma_cm_id *cma_id)
 {
-	struct iser_conn *ib_conn;
+	struct iser_conn *iser_conn = (struct iser_conn *)cma_id->context;
 
-	ib_conn = (struct iser_conn *)cma_id->context;
-
-	/* getting here when the state is UP means that the conn is being *
-	 * terminated asynchronously from the iSCSI layer's perspective.  */
-	if (iser_conn_state_comp_exch(ib_conn, ISER_CONN_UP,
-					ISER_CONN_TERMINATING)){
-		if (ib_conn->iscsi_conn)
-			iscsi_conn_failure(ib_conn->iscsi_conn, ISCSI_ERR_CONN_FAILED);
+	if (iser_conn_terminate(iser_conn)) {
+		if (iser_conn->iscsi_conn)
+			iscsi_conn_failure(iser_conn->iscsi_conn,
+					   ISCSI_ERR_CONN_FAILED);
 		else
 			iser_err("iscsi_iser connection isn't bound\n");
 	}
+}
 
-	/* Complete the termination process if no posts are pending. This code
-	 * block also exists in iser_handle_comp_error(), but it is needed here
-	 * for cases of no flushes at all, e.g. discovery over rdma.
+static void iser_cleanup_handler(struct rdma_cm_id *cma_id,
+				 bool destroy_device)
+{
+	struct iser_conn *iser_conn = (struct iser_conn *)cma_id->context;
+
+	/*
+	 * We are not guaranteed that we visited disconnected_handler
+	 * by now, call it here to be safe that we handle CM drep
+	 * and flush errors.
 	 */
-	if (ib_conn->post_recv_buf_count == 0 &&
-	    (atomic_read(&ib_conn->post_send_buf_count) == 0)) {
-		complete(&ib_conn->flush_completion);
-	}
-}
+	iser_disconnected_handler(cma_id);
+	iser_free_ib_conn_res(iser_conn, destroy_device);
+	complete(&iser_conn->ib_completion);
+};
 
 static int iser_cma_handler(struct rdma_cm_id *cma_id, struct rdma_cm_event *event)
 {
-	struct iser_conn *ib_conn;
+	struct iser_conn *iser_conn;
+	int ret = 0;
 
-	ib_conn = (struct iser_conn *)cma_id->context;
+	iser_conn = (struct iser_conn *)cma_id->context;
 	iser_info("event %d status %d conn %p id %p\n",
 		  event->event, event->status, cma_id->context, cma_id);
 
-	mutex_lock(&ib_conn->state_mutex);
+	mutex_lock(&iser_conn->state_mutex);
 	switch (event->event) {
 	case RDMA_CM_EVENT_ADDR_RESOLVED:
 		iser_addr_handler(cma_id);
@@ -824,57 +882,74 @@ static int iser_cma_handler(struct rdma_cm_id *cma_id, struct rdma_cm_event *eve
 		iser_connect_error(cma_id);
 		break;
 	case RDMA_CM_EVENT_DISCONNECTED:
-	case RDMA_CM_EVENT_DEVICE_REMOVAL:
 	case RDMA_CM_EVENT_ADDR_CHANGE:
 	case RDMA_CM_EVENT_TIMEWAIT_EXIT:
-		iser_disconnected_handler(cma_id);
+		iser_cleanup_handler(cma_id, false);
+		break;
+	case RDMA_CM_EVENT_DEVICE_REMOVAL:
+		/*
+		 * we *must* destroy the device as we cannot rely
+		 * on iscsid to be around to initiate error handling.
+		 * also if we are not in state DOWN implicitly destroy
+		 * the cma_id.
+		 */
+		iser_cleanup_handler(cma_id, true);
+		if (iser_conn->state != ISER_CONN_DOWN) {
+			iser_conn->ib_conn.cma_id = NULL;
+			ret = 1;
+		}
 		break;
 	default:
 		iser_err("Unexpected RDMA CM event (%d)\n", event->event);
 		break;
 	}
-	mutex_unlock(&ib_conn->state_mutex);
-	return 0;
+	mutex_unlock(&iser_conn->state_mutex);
+
+	return ret;
 }
 
-void iser_conn_init(struct iser_conn *ib_conn)
+void iser_conn_init(struct iser_conn *iser_conn)
 {
-	ib_conn->state = ISER_CONN_INIT;
-	ib_conn->post_recv_buf_count = 0;
-	atomic_set(&ib_conn->post_send_buf_count, 0);
-	init_completion(&ib_conn->stop_completion);
-	init_completion(&ib_conn->flush_completion);
-	init_completion(&ib_conn->up_completion);
-	INIT_LIST_HEAD(&ib_conn->conn_list);
-	spin_lock_init(&ib_conn->lock);
-	mutex_init(&ib_conn->state_mutex);
+	iser_conn->state = ISER_CONN_INIT;
+	iser_conn->ib_conn.post_recv_buf_count = 0;
+	init_completion(&iser_conn->ib_conn.flush_comp);
+	init_completion(&iser_conn->stop_completion);
+	init_completion(&iser_conn->ib_completion);
+	init_completion(&iser_conn->up_completion);
+	INIT_LIST_HEAD(&iser_conn->conn_list);
+	spin_lock_init(&iser_conn->ib_conn.lock);
+	mutex_init(&iser_conn->state_mutex);
 }
 
  /**
  * starts the process of connecting to the target
  * sleeps until the connection is established or rejected
  */
-int iser_connect(struct iser_conn   *ib_conn,
+int iser_connect(struct iser_conn   *iser_conn,
 		 struct sockaddr    *src_addr,
 		 struct sockaddr    *dst_addr,
 		 int                 non_blocking)
 {
+	struct ib_conn *ib_conn = &iser_conn->ib_conn;
 	int err = 0;
 
-	mutex_lock(&ib_conn->state_mutex);
+	mutex_lock(&iser_conn->state_mutex);
 
-	sprintf(ib_conn->name, "%pISp", dst_addr);
+	sprintf(iser_conn->name, "%pISp", dst_addr);
 
-	iser_info("connecting to: %s\n", ib_conn->name);
+	iser_info("connecting to: %s\n", iser_conn->name);
 
 	/* the device is known only --after-- address resolution */
 	ib_conn->device = NULL;
 
-	ib_conn->state = ISER_CONN_PENDING;
+	iser_conn->state = ISER_CONN_PENDING;
+
+	ib_conn->beacon.wr_id = ISER_BEACON_WRID;
+	ib_conn->beacon.opcode = IB_WR_SEND;
 
 	ib_conn->cma_id = rdma_create_id(iser_cma_handler,
-					     (void *)ib_conn,
-					     RDMA_PS_TCP, IB_QPT_RC);
+					 (void *)iser_conn,
+					 RDMA_PS_TCP, IB_QPT_RC);
 	if (IS_ERR(ib_conn->cma_id)) {
 		err = PTR_ERR(ib_conn->cma_id);
 		iser_err("rdma_create_id failed: %d\n", err);
@@ -888,27 +963,27 @@ int iser_connect(struct iser_conn   *ib_conn,
 	}
 
 	if (!non_blocking) {
-		wait_for_completion_interruptible(&ib_conn->up_completion);
+		wait_for_completion_interruptible(&iser_conn->up_completion);
 
-		if (ib_conn->state != ISER_CONN_UP) {
+		if (iser_conn->state != ISER_CONN_UP) {
 			err =  -EIO;
 			goto connect_failure;
 		}
 	}
-	mutex_unlock(&ib_conn->state_mutex);
+	mutex_unlock(&iser_conn->state_mutex);
 
 	mutex_lock(&ig.connlist_mutex);
-	list_add(&ib_conn->conn_list, &ig.connlist);
+	list_add(&iser_conn->conn_list, &ig.connlist);
 	mutex_unlock(&ig.connlist_mutex);
 	return 0;
 
 id_failure:
 	ib_conn->cma_id = NULL;
 addr_failure:
-	ib_conn->state = ISER_CONN_DOWN;
+	iser_conn->state = ISER_CONN_DOWN;
 connect_failure:
-	mutex_unlock(&ib_conn->state_mutex);
-	iser_conn_release(ib_conn);
+	mutex_unlock(&iser_conn->state_mutex);
+	iser_conn_release(iser_conn);
 	return err;
 }
 
@@ -917,7 +992,7 @@ connect_failure:
  *
  * returns: 0 on success, errno code on failure
  */
-int iser_reg_page_vec(struct iser_conn     *ib_conn,
+int iser_reg_page_vec(struct ib_conn *ib_conn,
 		      struct iser_page_vec *page_vec,
 		      struct iser_mem_reg  *mem_reg)
 {
@@ -944,7 +1019,6 @@ int iser_reg_page_vec(struct iser_conn     *ib_conn,
 	mem_reg->rkey  = mem->fmr->rkey;
 	mem_reg->len   = page_vec->length * SIZE_4K;
 	mem_reg->va    = io_addr;
-	mem_reg->is_mr = 1;
 	mem_reg->mem_h = (void *)mem;
 
 	mem_reg->va   += page_vec->offset;
@@ -971,7 +1045,7 @@ void iser_unreg_mem_fmr(struct iscsi_iser_task *iser_task,
 	struct iser_mem_reg *reg = &iser_task->rdma_regd[cmd_dir].reg;
 	int ret;
 
-	if (!reg->is_mr)
+	if (!reg->mem_h)
 		return;
 
 	iser_dbg("PHYSICAL Mem.Unregister mem_h %p\n",reg->mem_h);
@@ -987,30 +1061,31 @@ void iser_unreg_mem_fastreg(struct iscsi_iser_task *iser_task,
 			    enum iser_data_dir cmd_dir)
 {
 	struct iser_mem_reg *reg = &iser_task->rdma_regd[cmd_dir].reg;
-	struct iser_conn *ib_conn = iser_task->ib_conn;
+	struct iser_conn *iser_conn = iser_task->iser_conn;
+	struct ib_conn *ib_conn = &iser_conn->ib_conn;
 	struct fast_reg_descriptor *desc = reg->mem_h;
 
-	if (!reg->is_mr)
+	if (!desc)
 		return;
 
 	reg->mem_h = NULL;
-	reg->is_mr = 0;
 	spin_lock_bh(&ib_conn->lock);
 	list_add_tail(&desc->list, &ib_conn->fastreg.pool);
 	spin_unlock_bh(&ib_conn->lock);
 }
 
-int iser_post_recvl(struct iser_conn *ib_conn)
+int iser_post_recvl(struct iser_conn *iser_conn)
 {
 	struct ib_recv_wr rx_wr, *rx_wr_failed;
+	struct ib_conn *ib_conn = &iser_conn->ib_conn;
 	struct ib_sge	  sge;
 	int ib_ret;
 
-	sge.addr   = ib_conn->login_resp_dma;
+	sge.addr   = iser_conn->login_resp_dma;
 	sge.length = ISER_RX_LOGIN_SIZE;
 	sge.lkey   = ib_conn->device->mr->lkey;
 
-	rx_wr.wr_id   = (unsigned long)ib_conn->login_resp_buf;
+	rx_wr.wr_id   = (uintptr_t)iser_conn->login_resp_buf;
 	rx_wr.sg_list = &sge;
 	rx_wr.num_sge = 1;
 	rx_wr.next    = NULL;
@@ -1024,20 +1099,21 @@ int iser_post_recvl(struct iser_conn *ib_conn)
 	return ib_ret;
 }
 
-int iser_post_recvm(struct iser_conn *ib_conn, int count)
+int iser_post_recvm(struct iser_conn *iser_conn, int count)
 {
 	struct ib_recv_wr *rx_wr, *rx_wr_failed;
 	int i, ib_ret;
-	unsigned int my_rx_head = ib_conn->rx_desc_head;
+	struct ib_conn *ib_conn = &iser_conn->ib_conn;
+	unsigned int my_rx_head = iser_conn->rx_desc_head;
 	struct iser_rx_desc *rx_desc;
 
 	for (rx_wr = ib_conn->rx_wr, i = 0; i < count; i++, rx_wr++) {
-		rx_desc		= &ib_conn->rx_descs[my_rx_head];
-		rx_wr->wr_id	= (unsigned long)rx_desc;
+		rx_desc		= &iser_conn->rx_descs[my_rx_head];
+		rx_wr->wr_id	= (uintptr_t)rx_desc;
 		rx_wr->sg_list	= &rx_desc->rx_sg;
 		rx_wr->num_sge	= 1;
 		rx_wr->next	= rx_wr + 1;
-		my_rx_head = (my_rx_head + 1) & ib_conn->qp_max_recv_dtos_mask;
+		my_rx_head = (my_rx_head + 1) & iser_conn->qp_max_recv_dtos_mask;
 	}
 
 	rx_wr--;
@@ -1049,7 +1125,7 @@ int iser_post_recvm(struct iser_conn *ib_conn, int count)
 		iser_err("ib_post_recv failed ret=%d\n", ib_ret);
 		ib_conn->post_recv_buf_count -= count;
 	} else
-		ib_conn->rx_desc_head = my_rx_head;
+		iser_conn->rx_desc_head = my_rx_head;
 	return ib_ret;
 }
 
@@ -1059,139 +1135,167 @@ int iser_post_recvm(struct iser_conn *ib_conn, int count)
  *
  * returns 0 on success, -1 on failure
  */
-int iser_post_send(struct iser_conn *ib_conn, struct iser_tx_desc *tx_desc)
+int iser_post_send(struct ib_conn *ib_conn, struct iser_tx_desc *tx_desc,
+		   bool signal)
 {
 	int		  ib_ret;
 	struct ib_send_wr send_wr, *send_wr_failed;
 
 	ib_dma_sync_single_for_device(ib_conn->device->ib_device,
-		tx_desc->dma_addr, ISER_HEADERS_LEN, DMA_TO_DEVICE);
+				      tx_desc->dma_addr, ISER_HEADERS_LEN,
+				      DMA_TO_DEVICE);
 
 	send_wr.next	   = NULL;
-	send_wr.wr_id	   = (unsigned long)tx_desc;
+	send_wr.wr_id	   = (uintptr_t)tx_desc;
 	send_wr.sg_list	   = tx_desc->tx_sg;
 	send_wr.num_sge	   = tx_desc->num_sge;
 	send_wr.opcode	   = IB_WR_SEND;
-	send_wr.send_flags = IB_SEND_SIGNALED;
-
-	atomic_inc(&ib_conn->post_send_buf_count);
+	send_wr.send_flags = signal ? IB_SEND_SIGNALED : 0;
 
 	ib_ret = ib_post_send(ib_conn->qp, &send_wr, &send_wr_failed);
-	if (ib_ret) {
+	if (ib_ret)
 		iser_err("ib_post_send failed, ret:%d\n", ib_ret);
-		atomic_dec(&ib_conn->post_send_buf_count);
-	}
+
 	return ib_ret;
 }
 
-static void iser_handle_comp_error(struct iser_tx_desc *desc,
-				struct iser_conn *ib_conn)
+/**
+ * is_iser_tx_desc - Indicate if the completion wr_id
+ *     is a TX descriptor or not.
+ * @iser_conn: iser connection
+ * @wr_id: completion WR identifier
+ *
+ * Since we cannot rely on wc opcode in FLUSH errors
+ * we must work around it by checking if the wr_id address
+ * falls in the iser connection rx_descs buffer. If so
+ * it is an RX descriptor, otherwize it is a TX.
+ */
+static inline bool
+is_iser_tx_desc(struct iser_conn *iser_conn, void *wr_id)
 {
-	if (desc && desc->type == ISCSI_TX_DATAOUT)
-		kmem_cache_free(ig.desc_cache, desc);
-
-	if (ib_conn->post_recv_buf_count == 0 &&
-	    atomic_read(&ib_conn->post_send_buf_count) == 0) {
-		/**
-		 * getting here when the state is UP means that the conn is
-		 * being terminated asynchronously from the iSCSI layer's
-		 * perspective. It is safe to peek at the connection state
-		 * since iscsi_conn_failure is allowed to be called twice.
-		 **/
-		if (ib_conn->state == ISER_CONN_UP)
-			iscsi_conn_failure(ib_conn->iscsi_conn,
+	void *start = iser_conn->rx_descs;
+	int len = iser_conn->num_rx_descs * sizeof(*iser_conn->rx_descs);
+
+	if (wr_id >= start && wr_id < start + len)
+		return false;
+
+	return true;
+}
+
+/**
+ * iser_handle_comp_error() - Handle error completion
+ * @ib_conn:   connection RDMA resources
+ * @wc:        work completion
+ *
+ * Notes: We may handle a FLUSH error completion and in this case
+ *        we only cleanup in case TX type was DATAOUT. For non-FLUSH
+ *        error completion we should also notify iscsi layer that
+ *        connection is failed (in case we passed bind stage).
+ */
+static void
+iser_handle_comp_error(struct ib_conn *ib_conn,
+		       struct ib_wc *wc)
+{
+	void *wr_id = (void *)(uintptr_t)wc->wr_id;
+	struct iser_conn *iser_conn = container_of(ib_conn, struct iser_conn,
+						   ib_conn);
+
+	if (wc->status != IB_WC_WR_FLUSH_ERR)
+		if (iser_conn->iscsi_conn)
+			iscsi_conn_failure(iser_conn->iscsi_conn,
 					   ISCSI_ERR_CONN_FAILED);
 
-		/* no more non completed posts to the QP, complete the
-		 * termination process w.o worrying on disconnect event */
-		complete(&ib_conn->flush_completion);
+	if (is_iser_tx_desc(iser_conn, wr_id)) {
+		struct iser_tx_desc *desc = wr_id;
+
+		if (desc->type == ISCSI_TX_DATAOUT)
+			kmem_cache_free(ig.desc_cache, desc);
+	} else {
+		ib_conn->post_recv_buf_count--;
 	}
 }
 
-static int iser_drain_tx_cq(struct iser_device  *device, int cq_index)
+/**
+ * iser_handle_wc - handle a single work completion
+ * @wc: work completion
+ *
+ * Soft-IRQ context, work completion can be either
+ * SEND or RECV, and can turn out successful or
+ * with error (or flush error).
+ */
+static void iser_handle_wc(struct ib_wc *wc)
 {
-	struct ib_cq  *cq = device->tx_cq[cq_index];
-	struct ib_wc  wc;
+	struct ib_conn *ib_conn;
 	struct iser_tx_desc *tx_desc;
-	struct iser_conn *ib_conn;
-	int completed_tx = 0;
-
-	while (ib_poll_cq(cq, 1, &wc) == 1) {
-		tx_desc	= (struct iser_tx_desc *) (unsigned long) wc.wr_id;
-		ib_conn = wc.qp->qp_context;
-		if (wc.status == IB_WC_SUCCESS) {
-			if (wc.opcode == IB_WC_SEND)
-				iser_snd_completion(tx_desc, ib_conn);
-			else
-				iser_err("expected opcode %d got %d\n",
-					IB_WC_SEND, wc.opcode);
+	struct iser_rx_desc *rx_desc;
+
+	ib_conn = wc->qp->qp_context;
+	if (likely(wc->status == IB_WC_SUCCESS)) {
+		if (wc->opcode == IB_WC_RECV) {
+			rx_desc = (struct iser_rx_desc *)(uintptr_t)wc->wr_id;
+			iser_rcv_completion(rx_desc, wc->byte_len,
+					    ib_conn);
+		} else
+		if (wc->opcode == IB_WC_SEND) {
+			tx_desc = (struct iser_tx_desc *)(uintptr_t)wc->wr_id;
+			iser_snd_completion(tx_desc, ib_conn);
 		} else {
-			iser_err("tx id %llx status %d vend_err %x\n",
-				 wc.wr_id, wc.status, wc.vendor_err);
-			if (wc.wr_id != ISER_FASTREG_LI_WRID) {
-				atomic_dec(&ib_conn->post_send_buf_count);
-				iser_handle_comp_error(tx_desc, ib_conn);
-			}
+			iser_err("Unknown wc opcode %d\n", wc->opcode);
 		}
-		completed_tx++;
+	} else {
+		if (wc->status != IB_WC_WR_FLUSH_ERR)
+			iser_err("wr id %llx status %d vend_err %x\n",
+				 wc->wr_id, wc->status, wc->vendor_err);
+		else
+			iser_dbg("flush error: wr id %llx\n", wc->wr_id);
+
+		if (wc->wr_id != ISER_FASTREG_LI_WRID &&
+		    wc->wr_id != ISER_BEACON_WRID)
+			iser_handle_comp_error(ib_conn, wc);
+
+		/* complete in case all flush errors were consumed */
+		if (wc->wr_id == ISER_BEACON_WRID)
+			complete(&ib_conn->flush_comp);
 	}
-	return completed_tx;
 }
 
-
+/**
+ * iser_cq_tasklet_fn - iSER completion polling loop
+ * @data: iSER completion context
+ *
+ * Soft-IRQ context, polling connection CQ until
+ * either CQ was empty or we exausted polling budget
+ */
 static void iser_cq_tasklet_fn(unsigned long data)
 {
-	struct iser_cq_desc *cq_desc = (struct iser_cq_desc *)data;
-	struct iser_device  *device = cq_desc->device;
-	int cq_index = cq_desc->cq_index;
-	struct ib_cq	     *cq = device->rx_cq[cq_index];
-	 struct ib_wc	     wc;
-	 struct iser_rx_desc *desc;
-	 unsigned long	     xfer_len;
-	struct iser_conn *ib_conn;
-	int completed_tx, completed_rx = 0;
-
-	/* First do tx drain, so in a case where we have rx flushes and a successful
-	 * tx completion we will still go through completion error handling.
-	 */
-	completed_tx = iser_drain_tx_cq(device, cq_index);
-
-	while (ib_poll_cq(cq, 1, &wc) == 1) {
-		desc	 = (struct iser_rx_desc *) (unsigned long) wc.wr_id;
-		BUG_ON(desc == NULL);
-		ib_conn = wc.qp->qp_context;
-		if (wc.status == IB_WC_SUCCESS) {
-			if (wc.opcode == IB_WC_RECV) {
-				xfer_len = (unsigned long)wc.byte_len;
-				iser_rcv_completion(desc, xfer_len, ib_conn);
-			} else
-				iser_err("expected opcode %d got %d\n",
-					IB_WC_RECV, wc.opcode);
-		} else {
-			if (wc.status != IB_WC_WR_FLUSH_ERR)
-				iser_err("rx id %llx status %d vend_err %x\n",
-					wc.wr_id, wc.status, wc.vendor_err);
-			ib_conn->post_recv_buf_count--;
-			iser_handle_comp_error(NULL, ib_conn);
-		}
-		completed_rx++;
-		if (!(completed_rx & 63))
-			completed_tx += iser_drain_tx_cq(device, cq_index);
+	struct iser_comp *comp = (struct iser_comp *)data;
+	struct ib_cq *cq = comp->cq;
+	struct ib_wc *const wcs = comp->wcs;
+	int i, n, completed = 0;
+
+	while ((n = ib_poll_cq(cq, ARRAY_SIZE(comp->wcs), wcs)) > 0) {
+		for (i = 0; i < n; i++)
+			iser_handle_wc(&wcs[i]);
+
+		completed += n;
+		if (completed >= iser_cq_poll_limit)
+			break;
 	}
-	/* #warning "it is assumed here that arming CQ only once its empty" *
-	 * " would not cause interrupts to be missed"                       */
+
+	/*
+	 * It is assumed here that arming CQ only once its empty
+	 * would not cause interrupts to be missed.
+	 */
 	ib_req_notify_cq(cq, IB_CQ_NEXT_COMP);
 
-	iser_dbg("got %d rx %d tx completions\n", completed_rx, completed_tx);
+	iser_dbg("got %d completions\n", completed);
 }
 
 static void iser_cq_callback(struct ib_cq *cq, void *cq_context)
 {
-	struct iser_cq_desc *cq_desc = (struct iser_cq_desc *)cq_context;
-	struct iser_device  *device = cq_desc->device;
-	int cq_index = cq_desc->cq_index;
+	struct iser_comp *comp = cq_context;
 
-	tasklet_schedule(&device->cq_tasklet[cq_index]);
+	tasklet_schedule(&comp->tasklet);
 }
 
 u8 iser_check_task_pi_status(struct iscsi_iser_task *iser_task,
diff --git a/drivers/infiniband/ulp/isert/ib_isert.c b/drivers/infiniband/ulp/isert/ib_isert.c
index da8ff124762a..dafb3c531f96 100644
--- a/drivers/infiniband/ulp/isert/ib_isert.c
+++ b/drivers/infiniband/ulp/isert/ib_isert.c
@@ -22,7 +22,6 @@
 #include <linux/socket.h>
 #include <linux/in.h>
 #include <linux/in6.h>
-#include <linux/llist.h>
 #include <rdma/ib_verbs.h>
 #include <rdma/rdma_cm.h>
 #include <target/target_core_base.h>
@@ -36,11 +35,17 @@
 #define	ISERT_MAX_CONN		8
 #define ISER_MAX_RX_CQ_LEN	(ISERT_QP_MAX_RECV_DTOS * ISERT_MAX_CONN)
 #define ISER_MAX_TX_CQ_LEN	(ISERT_QP_MAX_REQ_DTOS  * ISERT_MAX_CONN)
+#define ISER_MAX_CQ_LEN		(ISER_MAX_RX_CQ_LEN + ISER_MAX_TX_CQ_LEN + \
+				 ISERT_MAX_CONN)
+
+int isert_debug_level = 0;
+module_param_named(debug_level, isert_debug_level, int, 0644);
+MODULE_PARM_DESC(debug_level, "Enable debug tracing if > 0 (default:0)");
 
 static DEFINE_MUTEX(device_list_mutex);
 static LIST_HEAD(device_list);
-static struct workqueue_struct *isert_rx_wq;
 static struct workqueue_struct *isert_comp_wq;
+static struct workqueue_struct *isert_release_wq;
 
 static void
 isert_unmap_cmd(struct isert_cmd *isert_cmd, struct isert_conn *isert_conn);
@@ -54,19 +59,32 @@ isert_reg_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd,
 	       struct isert_rdma_wr *wr);
 static int
 isert_put_response(struct iscsi_conn *conn, struct iscsi_cmd *cmd);
+static int
+isert_rdma_post_recvl(struct isert_conn *isert_conn);
+static int
+isert_rdma_accept(struct isert_conn *isert_conn);
+struct rdma_cm_id *isert_setup_id(struct isert_np *isert_np);
+
+static inline bool
+isert_prot_cmd(struct isert_conn *conn, struct se_cmd *cmd)
+{
+	return (conn->pi_support &&
+		cmd->prot_op != TARGET_PROT_NORMAL);
+}
+
 
 static void
 isert_qp_event_callback(struct ib_event *e, void *context)
 {
 	struct isert_conn *isert_conn = (struct isert_conn *)context;
 
-	pr_err("isert_qp_event_callback event: %d\n", e->event);
+	isert_err("conn %p event: %d\n", isert_conn, e->event);
 	switch (e->event) {
 	case IB_EVENT_COMM_EST:
 		rdma_notify(isert_conn->conn_cm_id, IB_EVENT_COMM_EST);
 		break;
 	case IB_EVENT_QP_LAST_WQE_REACHED:
-		pr_warn("Reached TX IB_EVENT_QP_LAST_WQE_REACHED:\n");
+		isert_warn("Reached TX IB_EVENT_QP_LAST_WQE_REACHED\n");
 		break;
 	default:
 		break;
@@ -80,72 +98,77 @@ isert_query_device(struct ib_device *ib_dev, struct ib_device_attr *devattr)
 
 	ret = ib_query_device(ib_dev, devattr);
 	if (ret) {
-		pr_err("ib_query_device() failed: %d\n", ret);
+		isert_err("ib_query_device() failed: %d\n", ret);
 		return ret;
 	}
-	pr_debug("devattr->max_sge: %d\n", devattr->max_sge);
-	pr_debug("devattr->max_sge_rd: %d\n", devattr->max_sge_rd);
+	isert_dbg("devattr->max_sge: %d\n", devattr->max_sge);
+	isert_dbg("devattr->max_sge_rd: %d\n", devattr->max_sge_rd);
 
 	return 0;
 }
 
 static int
-isert_conn_setup_qp(struct isert_conn *isert_conn, struct rdma_cm_id *cma_id,
-		    u8 protection)
+isert_conn_setup_qp(struct isert_conn *isert_conn, struct rdma_cm_id *cma_id)
 {
 	struct isert_device *device = isert_conn->conn_device;
 	struct ib_qp_init_attr attr;
-	int ret, index, min_index = 0;
+	struct isert_comp *comp;
+	int ret, i, min = 0;
 
 	mutex_lock(&device_list_mutex);
-	for (index = 0; index < device->cqs_used; index++)
-		if (device->cq_active_qps[index] <
-		    device->cq_active_qps[min_index])
-			min_index = index;
-	device->cq_active_qps[min_index]++;
-	pr_debug("isert_conn_setup_qp: Using min_index: %d\n", min_index);
+	for (i = 0; i < device->comps_used; i++)
+		if (device->comps[i].active_qps <
+		    device->comps[min].active_qps)
+			min = i;
+	comp = &device->comps[min];
+	comp->active_qps++;
+	isert_info("conn %p, using comp %p min_index: %d\n",
+		   isert_conn, comp, min);
 	mutex_unlock(&device_list_mutex);
 
 	memset(&attr, 0, sizeof(struct ib_qp_init_attr));
 	attr.event_handler = isert_qp_event_callback;
 	attr.qp_context = isert_conn;
-	attr.send_cq = device->dev_tx_cq[min_index];
-	attr.recv_cq = device->dev_rx_cq[min_index];
+	attr.send_cq = comp->cq;
+	attr.recv_cq = comp->cq;
 	attr.cap.max_send_wr = ISERT_QP_MAX_REQ_DTOS;
-	attr.cap.max_recv_wr = ISERT_QP_MAX_RECV_DTOS;
+	attr.cap.max_recv_wr = ISERT_QP_MAX_RECV_DTOS + 1;
 	/*
 	 * FIXME: Use devattr.max_sge - 2 for max_send_sge as
-	 * work-around for RDMA_READ..
+	 * work-around for RDMA_READs with ConnectX-2.
+	 *
+	 * Also, still make sure to have at least two SGEs for
+	 * outgoing control PDU responses.
 	 */
-	attr.cap.max_send_sge = device->dev_attr.max_sge - 2;
+	attr.cap.max_send_sge = max(2, device->dev_attr.max_sge - 2);
 	isert_conn->max_sge = attr.cap.max_send_sge;
 
 	attr.cap.max_recv_sge = 1;
 	attr.sq_sig_type = IB_SIGNAL_REQ_WR;
 	attr.qp_type = IB_QPT_RC;
-	if (protection)
+	if (device->pi_capable)
 		attr.create_flags |= IB_QP_CREATE_SIGNATURE_EN;
 
-	pr_debug("isert_conn_setup_qp cma_id->device: %p\n",
-		 cma_id->device);
-	pr_debug("isert_conn_setup_qp conn_pd->device: %p\n",
-		 isert_conn->conn_pd->device);
-
 	ret = rdma_create_qp(cma_id, isert_conn->conn_pd, &attr);
 	if (ret) {
-		pr_err("rdma_create_qp failed for cma_id %d\n", ret);
-		return ret;
+		isert_err("rdma_create_qp failed for cma_id %d\n", ret);
+		goto err;
 	}
 	isert_conn->conn_qp = cma_id->qp;
-	pr_debug("rdma_create_qp() returned success >>>>>>>>>>>>>>>>>>>>>>>>>.\n");
 
 	return 0;
+err:
+	mutex_lock(&device_list_mutex);
+	comp->active_qps--;
+	mutex_unlock(&device_list_mutex);
+
+	return ret;
 }
 
 static void
 isert_cq_event_callback(struct ib_event *e, void *context)
 {
-	pr_debug("isert_cq_event_callback event: %d\n", e->event);
+	isert_dbg("event: %d\n", e->event);
 }
 
 static int
@@ -179,6 +202,7 @@ isert_alloc_rx_descriptors(struct isert_conn *isert_conn)
 	}
 
 	isert_conn->conn_rx_desc_head = 0;
+
 	return 0;
 
 dma_map_fail:
@@ -190,6 +214,8 @@ dma_map_fail:
 	kfree(isert_conn->conn_rx_descs);
 	isert_conn->conn_rx_descs = NULL;
 fail:
+	isert_err("conn %p failed to allocate rx descriptors\n", isert_conn);
+
 	return -ENOMEM;
 }
 
@@ -213,24 +239,24 @@ isert_free_rx_descriptors(struct isert_conn *isert_conn)
 	isert_conn->conn_rx_descs = NULL;
 }
 
-static void isert_cq_tx_work(struct work_struct *);
-static void isert_cq_tx_callback(struct ib_cq *, void *);
-static void isert_cq_rx_work(struct work_struct *);
-static void isert_cq_rx_callback(struct ib_cq *, void *);
+static void isert_cq_work(struct work_struct *);
+static void isert_cq_callback(struct ib_cq *, void *);
 
 static int
 isert_create_device_ib_res(struct isert_device *device)
 {
 	struct ib_device *ib_dev = device->ib_device;
-	struct isert_cq_desc *cq_desc;
 	struct ib_device_attr *dev_attr;
-	int ret = 0, i, j;
+	int ret = 0, i;
+	int max_cqe;
 
 	dev_attr = &device->dev_attr;
 	ret = isert_query_device(ib_dev, dev_attr);
 	if (ret)
 		return ret;
 
+	max_cqe = min(ISER_MAX_CQ_LEN, dev_attr->max_cqe);
+
 	/* asign function handlers */
 	if (dev_attr->device_cap_flags & IB_DEVICE_MEM_MGT_EXTENSIONS &&
 	    dev_attr->device_cap_flags & IB_DEVICE_SIGNATURE_HANDOVER) {
@@ -247,55 +273,38 @@ isert_create_device_ib_res(struct isert_device *device)
 	device->pi_capable = dev_attr->device_cap_flags &
 			     IB_DEVICE_SIGNATURE_HANDOVER ? true : false;
 
-	device->cqs_used = min_t(int, num_online_cpus(),
-				 device->ib_device->num_comp_vectors);
-	device->cqs_used = min(ISERT_MAX_CQ, device->cqs_used);
-	pr_debug("Using %d CQs, device %s supports %d vectors support "
-		 "Fast registration %d pi_capable %d\n",
-		 device->cqs_used, device->ib_device->name,
-		 device->ib_device->num_comp_vectors, device->use_fastreg,
-		 device->pi_capable);
-	device->cq_desc = kzalloc(sizeof(struct isert_cq_desc) *
-				device->cqs_used, GFP_KERNEL);
-	if (!device->cq_desc) {
-		pr_err("Unable to allocate device->cq_desc\n");
+	device->comps_used = min(ISERT_MAX_CQ, min_t(int, num_online_cpus(),
+					device->ib_device->num_comp_vectors));
+	isert_info("Using %d CQs, %s supports %d vectors support "
+		   "Fast registration %d pi_capable %d\n",
+		   device->comps_used, device->ib_device->name,
+		   device->ib_device->num_comp_vectors, device->use_fastreg,
+		   device->pi_capable);
+
+	device->comps = kcalloc(device->comps_used, sizeof(struct isert_comp),
+				GFP_KERNEL);
+	if (!device->comps) {
+		isert_err("Unable to allocate completion contexts\n");
 		return -ENOMEM;
 	}
-	cq_desc = device->cq_desc;
-
-	for (i = 0; i < device->cqs_used; i++) {
-		cq_desc[i].device = device;
-		cq_desc[i].cq_index = i;
-
-		INIT_WORK(&cq_desc[i].cq_rx_work, isert_cq_rx_work);
-		device->dev_rx_cq[i] = ib_create_cq(device->ib_device,
-						isert_cq_rx_callback,
-						isert_cq_event_callback,
-						(void *)&cq_desc[i],
-						ISER_MAX_RX_CQ_LEN, i);
-		if (IS_ERR(device->dev_rx_cq[i])) {
-			ret = PTR_ERR(device->dev_rx_cq[i]);
-			device->dev_rx_cq[i] = NULL;
-			goto out_cq;
-		}
 
-		INIT_WORK(&cq_desc[i].cq_tx_work, isert_cq_tx_work);
-		device->dev_tx_cq[i] = ib_create_cq(device->ib_device,
-						isert_cq_tx_callback,
-						isert_cq_event_callback,
-						(void *)&cq_desc[i],
-						ISER_MAX_TX_CQ_LEN, i);
-		if (IS_ERR(device->dev_tx_cq[i])) {
-			ret = PTR_ERR(device->dev_tx_cq[i]);
-			device->dev_tx_cq[i] = NULL;
-			goto out_cq;
-		}
+	for (i = 0; i < device->comps_used; i++) {
+		struct isert_comp *comp = &device->comps[i];
 
-		ret = ib_req_notify_cq(device->dev_rx_cq[i], IB_CQ_NEXT_COMP);
-		if (ret)
+		comp->device = device;
+		INIT_WORK(&comp->work, isert_cq_work);
+		comp->cq = ib_create_cq(device->ib_device,
+					isert_cq_callback,
+					isert_cq_event_callback,
+					(void *)comp,
+					max_cqe, i);
+		if (IS_ERR(comp->cq)) {
+			ret = PTR_ERR(comp->cq);
+			comp->cq = NULL;
 			goto out_cq;
+		}
 
-		ret = ib_req_notify_cq(device->dev_tx_cq[i], IB_CQ_NEXT_COMP);
+		ret = ib_req_notify_cq(comp->cq, IB_CQ_NEXT_COMP);
 		if (ret)
 			goto out_cq;
 	}
@@ -303,19 +312,15 @@ isert_create_device_ib_res(struct isert_device *device)
 	return 0;
 
 out_cq:
-	for (j = 0; j < i; j++) {
-		cq_desc = &device->cq_desc[j];
+	for (i = 0; i < device->comps_used; i++) {
+		struct isert_comp *comp = &device->comps[i];
 
-		if (device->dev_rx_cq[j]) {
-			cancel_work_sync(&cq_desc->cq_rx_work);
-			ib_destroy_cq(device->dev_rx_cq[j]);
-		}
-		if (device->dev_tx_cq[j]) {
-			cancel_work_sync(&cq_desc->cq_tx_work);
-			ib_destroy_cq(device->dev_tx_cq[j]);
+		if (comp->cq) {
+			cancel_work_sync(&comp->work);
+			ib_destroy_cq(comp->cq);
 		}
 	}
-	kfree(device->cq_desc);
+	kfree(device->comps);
 
 	return ret;
 }
@@ -323,21 +328,18 @@ out_cq:
 static void
 isert_free_device_ib_res(struct isert_device *device)
 {
-	struct isert_cq_desc *cq_desc;
 	int i;
 
-	for (i = 0; i < device->cqs_used; i++) {
-		cq_desc = &device->cq_desc[i];
+	isert_info("device %p\n", device);
 
-		cancel_work_sync(&cq_desc->cq_rx_work);
-		cancel_work_sync(&cq_desc->cq_tx_work);
-		ib_destroy_cq(device->dev_rx_cq[i]);
-		ib_destroy_cq(device->dev_tx_cq[i]);
-		device->dev_rx_cq[i] = NULL;
-		device->dev_tx_cq[i] = NULL;
-	}
+	for (i = 0; i < device->comps_used; i++) {
+		struct isert_comp *comp = &device->comps[i];
 
-	kfree(device->cq_desc);
+		cancel_work_sync(&comp->work);
+		ib_destroy_cq(comp->cq);
+		comp->cq = NULL;
+	}
+	kfree(device->comps);
 }
 
 static void
@@ -345,6 +347,7 @@ isert_device_try_release(struct isert_device *device)
 {
 	mutex_lock(&device_list_mutex);
 	device->refcount--;
+	isert_info("device %p refcount %d\n", device, device->refcount);
 	if (!device->refcount) {
 		isert_free_device_ib_res(device);
 		list_del(&device->dev_node);
@@ -363,6 +366,8 @@ isert_device_find_by_ib_dev(struct rdma_cm_id *cma_id)
 	list_for_each_entry(device, &device_list, dev_node) {
 		if (device->ib_device->node_guid == cma_id->device->node_guid) {
 			device->refcount++;
+			isert_info("Found iser device %p refcount %d\n",
+				   device, device->refcount);
 			mutex_unlock(&device_list_mutex);
 			return device;
 		}
@@ -386,6 +391,8 @@ isert_device_find_by_ib_dev(struct rdma_cm_id *cma_id)
 
 	device->refcount++;
 	list_add_tail(&device->dev_node, &device_list);
+	isert_info("Created a new iser device %p refcount %d\n",
+		   device, device->refcount);
 	mutex_unlock(&device_list_mutex);
 
 	return device;
@@ -400,7 +407,7 @@ isert_conn_free_fastreg_pool(struct isert_conn *isert_conn)
 	if (list_empty(&isert_conn->conn_fr_pool))
 		return;
 
-	pr_debug("Freeing conn %p fastreg pool", isert_conn);
+	isert_info("Freeing conn %p fastreg pool", isert_conn);
 
 	list_for_each_entry_safe(fr_desc, tmp,
 				 &isert_conn->conn_fr_pool, list) {
@@ -418,87 +425,97 @@ isert_conn_free_fastreg_pool(struct isert_conn *isert_conn)
 	}
 
 	if (i < isert_conn->conn_fr_pool_size)
-		pr_warn("Pool still has %d regions registered\n",
+		isert_warn("Pool still has %d regions registered\n",
 			isert_conn->conn_fr_pool_size - i);
 }
 
 static int
+isert_create_pi_ctx(struct fast_reg_descriptor *desc,
+		    struct ib_device *device,
+		    struct ib_pd *pd)
+{
+	struct ib_mr_init_attr mr_init_attr;
+	struct pi_context *pi_ctx;
+	int ret;
+
+	pi_ctx = kzalloc(sizeof(*desc->pi_ctx), GFP_KERNEL);
+	if (!pi_ctx) {
+		isert_err("Failed to allocate pi context\n");
+		return -ENOMEM;
+	}
+
+	pi_ctx->prot_frpl = ib_alloc_fast_reg_page_list(device,
+					    ISCSI_ISER_SG_TABLESIZE);
+	if (IS_ERR(pi_ctx->prot_frpl)) {
+		isert_err("Failed to allocate prot frpl err=%ld\n",
+			  PTR_ERR(pi_ctx->prot_frpl));
+		ret = PTR_ERR(pi_ctx->prot_frpl);
+		goto err_pi_ctx;
+	}
+
+	pi_ctx->prot_mr = ib_alloc_fast_reg_mr(pd, ISCSI_ISER_SG_TABLESIZE);
+	if (IS_ERR(pi_ctx->prot_mr)) {
+		isert_err("Failed to allocate prot frmr err=%ld\n",
+			  PTR_ERR(pi_ctx->prot_mr));
+		ret = PTR_ERR(pi_ctx->prot_mr);
+		goto err_prot_frpl;
+	}
+	desc->ind |= ISERT_PROT_KEY_VALID;
+
+	memset(&mr_init_attr, 0, sizeof(mr_init_attr));
+	mr_init_attr.max_reg_descriptors = 2;
+	mr_init_attr.flags |= IB_MR_SIGNATURE_EN;
+	pi_ctx->sig_mr = ib_create_mr(pd, &mr_init_attr);
+	if (IS_ERR(pi_ctx->sig_mr)) {
+		isert_err("Failed to allocate signature enabled mr err=%ld\n",
+			  PTR_ERR(pi_ctx->sig_mr));
+		ret = PTR_ERR(pi_ctx->sig_mr);
+		goto err_prot_mr;
+	}
+
+	desc->pi_ctx = pi_ctx;
+	desc->ind |= ISERT_SIG_KEY_VALID;
+	desc->ind &= ~ISERT_PROTECTED;
+
+	return 0;
+
+err_prot_mr:
+	ib_dereg_mr(desc->pi_ctx->prot_mr);
+err_prot_frpl:
+	ib_free_fast_reg_page_list(desc->pi_ctx->prot_frpl);
+err_pi_ctx:
+	kfree(desc->pi_ctx);
+
+	return ret;
+}
+
+static int
 isert_create_fr_desc(struct ib_device *ib_device, struct ib_pd *pd,
-		     struct fast_reg_descriptor *fr_desc, u8 protection)
+		     struct fast_reg_descriptor *fr_desc)
 {
 	int ret;
 
 	fr_desc->data_frpl = ib_alloc_fast_reg_page_list(ib_device,
 							 ISCSI_ISER_SG_TABLESIZE);
 	if (IS_ERR(fr_desc->data_frpl)) {
-		pr_err("Failed to allocate data frpl err=%ld\n",
-		       PTR_ERR(fr_desc->data_frpl));
+		isert_err("Failed to allocate data frpl err=%ld\n",
+			  PTR_ERR(fr_desc->data_frpl));
 		return PTR_ERR(fr_desc->data_frpl);
 	}
 
 	fr_desc->data_mr = ib_alloc_fast_reg_mr(pd, ISCSI_ISER_SG_TABLESIZE);
 	if (IS_ERR(fr_desc->data_mr)) {
-		pr_err("Failed to allocate data frmr err=%ld\n",
-		       PTR_ERR(fr_desc->data_mr));
+		isert_err("Failed to allocate data frmr err=%ld\n",
+			  PTR_ERR(fr_desc->data_mr));
 		ret = PTR_ERR(fr_desc->data_mr);
 		goto err_data_frpl;
 	}
-	pr_debug("Create fr_desc %p page_list %p\n",
-		 fr_desc, fr_desc->data_frpl->page_list);
 	fr_desc->ind |= ISERT_DATA_KEY_VALID;
 
-	if (protection) {
-		struct ib_mr_init_attr mr_init_attr = {0};
-		struct pi_context *pi_ctx;
-
-		fr_desc->pi_ctx = kzalloc(sizeof(*fr_desc->pi_ctx), GFP_KERNEL);
-		if (!fr_desc->pi_ctx) {
-			pr_err("Failed to allocate pi context\n");
-			ret = -ENOMEM;
-			goto err_data_mr;
-		}
-		pi_ctx = fr_desc->pi_ctx;
-
-		pi_ctx->prot_frpl = ib_alloc_fast_reg_page_list(ib_device,
-						    ISCSI_ISER_SG_TABLESIZE);
-		if (IS_ERR(pi_ctx->prot_frpl)) {
-			pr_err("Failed to allocate prot frpl err=%ld\n",
-			       PTR_ERR(pi_ctx->prot_frpl));
-			ret = PTR_ERR(pi_ctx->prot_frpl);
-			goto err_pi_ctx;
-		}
-
-		pi_ctx->prot_mr = ib_alloc_fast_reg_mr(pd, ISCSI_ISER_SG_TABLESIZE);
-		if (IS_ERR(pi_ctx->prot_mr)) {
-			pr_err("Failed to allocate prot frmr err=%ld\n",
-			       PTR_ERR(pi_ctx->prot_mr));
-			ret = PTR_ERR(pi_ctx->prot_mr);
-			goto err_prot_frpl;
-		}
-		fr_desc->ind |= ISERT_PROT_KEY_VALID;
-
-		mr_init_attr.max_reg_descriptors = 2;
-		mr_init_attr.flags |= IB_MR_SIGNATURE_EN;
-		pi_ctx->sig_mr = ib_create_mr(pd, &mr_init_attr);
-		if (IS_ERR(pi_ctx->sig_mr)) {
-			pr_err("Failed to allocate signature enabled mr err=%ld\n",
-			       PTR_ERR(pi_ctx->sig_mr));
-			ret = PTR_ERR(pi_ctx->sig_mr);
-			goto err_prot_mr;
-		}
-		fr_desc->ind |= ISERT_SIG_KEY_VALID;
-	}
-	fr_desc->ind &= ~ISERT_PROTECTED;
+	isert_dbg("Created fr_desc %p\n", fr_desc);
 
 	return 0;
-err_prot_mr:
-	ib_dereg_mr(fr_desc->pi_ctx->prot_mr);
-err_prot_frpl:
-	ib_free_fast_reg_page_list(fr_desc->pi_ctx->prot_frpl);
-err_pi_ctx:
-	kfree(fr_desc->pi_ctx);
-err_data_mr:
-	ib_dereg_mr(fr_desc->data_mr);
+
 err_data_frpl:
 	ib_free_fast_reg_page_list(fr_desc->data_frpl);
 
@@ -506,7 +523,7 @@ err_data_frpl:
 }
 
 static int
-isert_conn_create_fastreg_pool(struct isert_conn *isert_conn, u8 pi_support)
+isert_conn_create_fastreg_pool(struct isert_conn *isert_conn)
 {
 	struct fast_reg_descriptor *fr_desc;
 	struct isert_device *device = isert_conn->conn_device;
@@ -524,16 +541,15 @@ isert_conn_create_fastreg_pool(struct isert_conn *isert_conn, u8 pi_support)
 	for (i = 0; i < tag_num; i++) {
 		fr_desc = kzalloc(sizeof(*fr_desc), GFP_KERNEL);
 		if (!fr_desc) {
-			pr_err("Failed to allocate fast_reg descriptor\n");
+			isert_err("Failed to allocate fast_reg descriptor\n");
 			ret = -ENOMEM;
 			goto err;
 		}
 
 		ret = isert_create_fr_desc(device->ib_device,
-					   isert_conn->conn_pd, fr_desc,
-					   pi_support);
+					   isert_conn->conn_pd, fr_desc);
 		if (ret) {
-			pr_err("Failed to create fastreg descriptor err=%d\n",
+			isert_err("Failed to create fastreg descriptor err=%d\n",
 			       ret);
 			kfree(fr_desc);
 			goto err;
@@ -543,7 +559,7 @@ isert_conn_create_fastreg_pool(struct isert_conn *isert_conn, u8 pi_support)
 		isert_conn->conn_fr_pool_size++;
 	}
 
-	pr_debug("Creating conn %p fastreg pool size=%d",
+	isert_dbg("Creating conn %p fastreg pool size=%d",
 		 isert_conn, isert_conn->conn_fr_pool_size);
 
 	return 0;
@@ -556,47 +572,45 @@ err:
 static int
 isert_connect_request(struct rdma_cm_id *cma_id, struct rdma_cm_event *event)
 {
-	struct iscsi_np *np = cma_id->context;
-	struct isert_np *isert_np = np->np_context;
+	struct isert_np *isert_np = cma_id->context;
+	struct iscsi_np *np = isert_np->np;
 	struct isert_conn *isert_conn;
 	struct isert_device *device;
 	struct ib_device *ib_dev = cma_id->device;
 	int ret = 0;
-	u8 pi_support;
 
 	spin_lock_bh(&np->np_thread_lock);
 	if (!np->enabled) {
 		spin_unlock_bh(&np->np_thread_lock);
-		pr_debug("iscsi_np is not enabled, reject connect request\n");
+		isert_dbg("iscsi_np is not enabled, reject connect request\n");
 		return rdma_reject(cma_id, NULL, 0);
 	}
 	spin_unlock_bh(&np->np_thread_lock);
 
-	pr_debug("Entering isert_connect_request cma_id: %p, context: %p\n",
+	isert_dbg("cma_id: %p, portal: %p\n",
 		 cma_id, cma_id->context);
 
 	isert_conn = kzalloc(sizeof(struct isert_conn), GFP_KERNEL);
 	if (!isert_conn) {
-		pr_err("Unable to allocate isert_conn\n");
+		isert_err("Unable to allocate isert_conn\n");
 		return -ENOMEM;
 	}
 	isert_conn->state = ISER_CONN_INIT;
 	INIT_LIST_HEAD(&isert_conn->conn_accept_node);
 	init_completion(&isert_conn->conn_login_comp);
+	init_completion(&isert_conn->login_req_comp);
 	init_completion(&isert_conn->conn_wait);
-	init_completion(&isert_conn->conn_wait_comp_err);
 	kref_init(&isert_conn->conn_kref);
 	mutex_init(&isert_conn->conn_mutex);
 	spin_lock_init(&isert_conn->conn_lock);
 	INIT_LIST_HEAD(&isert_conn->conn_fr_pool);
 
-	cma_id->context = isert_conn;
 	isert_conn->conn_cm_id = cma_id;
 
 	isert_conn->login_buf = kzalloc(ISCSI_DEF_MAX_RECV_SEG_LEN +
 					ISER_RX_LOGIN_SIZE, GFP_KERNEL);
 	if (!isert_conn->login_buf) {
-		pr_err("Unable to allocate isert_conn->login_buf\n");
+		isert_err("Unable to allocate isert_conn->login_buf\n");
 		ret = -ENOMEM;
 		goto out;
 	}
@@ -604,7 +618,7 @@ isert_connect_request(struct rdma_cm_id *cma_id, struct rdma_cm_event *event)
 	isert_conn->login_req_buf = isert_conn->login_buf;
 	isert_conn->login_rsp_buf = isert_conn->login_buf +
 				    ISCSI_DEF_MAX_RECV_SEG_LEN;
-	pr_debug("Set login_buf: %p login_req_buf: %p login_rsp_buf: %p\n",
+	isert_dbg("Set login_buf: %p login_req_buf: %p login_rsp_buf: %p\n",
 		 isert_conn->login_buf, isert_conn->login_req_buf,
 		 isert_conn->login_rsp_buf);
 
@@ -614,7 +628,7 @@ isert_connect_request(struct rdma_cm_id *cma_id, struct rdma_cm_event *event)
 
 	ret = ib_dma_mapping_error(ib_dev, isert_conn->login_req_dma);
 	if (ret) {
-		pr_err("ib_dma_mapping_error failed for login_req_dma: %d\n",
+		isert_err("ib_dma_mapping_error failed for login_req_dma: %d\n",
 		       ret);
 		isert_conn->login_req_dma = 0;
 		goto out_login_buf;
@@ -626,7 +640,7 @@ isert_connect_request(struct rdma_cm_id *cma_id, struct rdma_cm_event *event)
 
 	ret = ib_dma_mapping_error(ib_dev, isert_conn->login_rsp_dma);
 	if (ret) {
-		pr_err("ib_dma_mapping_error failed for login_rsp_dma: %d\n",
+		isert_err("ib_dma_mapping_error failed for login_rsp_dma: %d\n",
 		       ret);
 		isert_conn->login_rsp_dma = 0;
 		goto out_req_dma_map;
@@ -642,13 +656,13 @@ isert_connect_request(struct rdma_cm_id *cma_id, struct rdma_cm_event *event)
 	isert_conn->initiator_depth = min_t(u8,
 				event->param.conn.initiator_depth,
 				device->dev_attr.max_qp_init_rd_atom);
-	pr_debug("Using initiator_depth: %u\n", isert_conn->initiator_depth);
+	isert_dbg("Using initiator_depth: %u\n", isert_conn->initiator_depth);
 
 	isert_conn->conn_device = device;
 	isert_conn->conn_pd = ib_alloc_pd(isert_conn->conn_device->ib_device);
 	if (IS_ERR(isert_conn->conn_pd)) {
 		ret = PTR_ERR(isert_conn->conn_pd);
-		pr_err("ib_alloc_pd failed for conn %p: ret=%d\n",
+		isert_err("ib_alloc_pd failed for conn %p: ret=%d\n",
 		       isert_conn, ret);
 		goto out_pd;
 	}
@@ -657,20 +671,20 @@ isert_connect_request(struct rdma_cm_id *cma_id, struct rdma_cm_event *event)
 					   IB_ACCESS_LOCAL_WRITE);
 	if (IS_ERR(isert_conn->conn_mr)) {
 		ret = PTR_ERR(isert_conn->conn_mr);
-		pr_err("ib_get_dma_mr failed for conn %p: ret=%d\n",
+		isert_err("ib_get_dma_mr failed for conn %p: ret=%d\n",
 		       isert_conn, ret);
 		goto out_mr;
 	}
 
-	pi_support = np->tpg_np->tpg->tpg_attrib.t10_pi;
-	if (pi_support && !device->pi_capable) {
-		pr_err("Protection information requested but not supported, "
-		       "rejecting connect request\n");
-		ret = rdma_reject(cma_id, NULL, 0);
-		goto out_mr;
-	}
+	ret = isert_conn_setup_qp(isert_conn, cma_id);
+	if (ret)
+		goto out_conn_dev;
 
-	ret = isert_conn_setup_qp(isert_conn, cma_id, pi_support);
+	ret = isert_rdma_post_recvl(isert_conn);
+	if (ret)
+		goto out_conn_dev;
+
+	ret = isert_rdma_accept(isert_conn);
 	if (ret)
 		goto out_conn_dev;
 
@@ -678,7 +692,7 @@ isert_connect_request(struct rdma_cm_id *cma_id, struct rdma_cm_event *event)
 	list_add_tail(&isert_conn->conn_accept_node, &isert_np->np_accept_list);
 	mutex_unlock(&isert_np->np_accept_mutex);
 
-	pr_debug("isert_connect_request() up np_sem np: %p\n", np);
+	isert_info("np %p: Allow accept_np to continue\n", np);
 	up(&isert_np->np_sem);
 	return 0;
 
@@ -698,6 +712,7 @@ out_login_buf:
 	kfree(isert_conn->login_buf);
 out:
 	kfree(isert_conn);
+	rdma_reject(cma_id, NULL, 0);
 	return ret;
 }
 
@@ -706,24 +721,25 @@ isert_connect_release(struct isert_conn *isert_conn)
 {
 	struct ib_device *ib_dev = isert_conn->conn_cm_id->device;
 	struct isert_device *device = isert_conn->conn_device;
-	int cq_index;
 
-	pr_debug("Entering isert_connect_release(): >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\n");
+	isert_dbg("conn %p\n", isert_conn);
 
 	if (device && device->use_fastreg)
 		isert_conn_free_fastreg_pool(isert_conn);
 
+	isert_free_rx_descriptors(isert_conn);
+	rdma_destroy_id(isert_conn->conn_cm_id);
+
 	if (isert_conn->conn_qp) {
-		cq_index = ((struct isert_cq_desc *)
-			isert_conn->conn_qp->recv_cq->cq_context)->cq_index;
-		pr_debug("isert_connect_release: cq_index: %d\n", cq_index);
-		isert_conn->conn_device->cq_active_qps[cq_index]--;
+		struct isert_comp *comp = isert_conn->conn_qp->recv_cq->cq_context;
 
-		rdma_destroy_qp(isert_conn->conn_cm_id);
-	}
+		isert_dbg("dec completion context %p active_qps\n", comp);
+		mutex_lock(&device_list_mutex);
+		comp->active_qps--;
+		mutex_unlock(&device_list_mutex);
 
-	isert_free_rx_descriptors(isert_conn);
-	rdma_destroy_id(isert_conn->conn_cm_id);
+		ib_destroy_qp(isert_conn->conn_qp);
+	}
 
 	ib_dereg_mr(isert_conn->conn_mr);
 	ib_dealloc_pd(isert_conn->conn_pd);
@@ -740,16 +756,24 @@ isert_connect_release(struct isert_conn *isert_conn)
 
 	if (device)
 		isert_device_try_release(device);
-
-	pr_debug("Leaving isert_connect_release >>>>>>>>>>>>\n");
 }
 
 static void
 isert_connected_handler(struct rdma_cm_id *cma_id)
 {
-	struct isert_conn *isert_conn = cma_id->context;
+	struct isert_conn *isert_conn = cma_id->qp->qp_context;
+
+	isert_info("conn %p\n", isert_conn);
+
+	if (!kref_get_unless_zero(&isert_conn->conn_kref)) {
+		isert_warn("conn %p connect_release is running\n", isert_conn);
+		return;
+	}
 
-	kref_get(&isert_conn->conn_kref);
+	mutex_lock(&isert_conn->conn_mutex);
+	if (isert_conn->state != ISER_CONN_FULL_FEATURE)
+		isert_conn->state = ISER_CONN_UP;
+	mutex_unlock(&isert_conn->conn_mutex);
 }
 
 static void
@@ -758,8 +782,8 @@ isert_release_conn_kref(struct kref *kref)
 	struct isert_conn *isert_conn = container_of(kref,
 				struct isert_conn, conn_kref);
 
-	pr_debug("Calling isert_connect_release for final kref %s/%d\n",
-		 current->comm, current->pid);
+	isert_info("conn %p final kref %s/%d\n", isert_conn, current->comm,
+		   current->pid);
 
 	isert_connect_release(isert_conn);
 }
@@ -770,61 +794,111 @@ isert_put_conn(struct isert_conn *isert_conn)
 	kref_put(&isert_conn->conn_kref, isert_release_conn_kref);
 }
 
+/**
+ * isert_conn_terminate() - Initiate connection termination
+ * @isert_conn: isert connection struct
+ *
+ * Notes:
+ * In case the connection state is FULL_FEATURE, move state
+ * to TEMINATING and start teardown sequence (rdma_disconnect).
+ * In case the connection state is UP, complete flush as well.
+ *
+ * This routine must be called with conn_mutex held. Thus it is
+ * safe to call multiple times.
+ */
 static void
-isert_disconnect_work(struct work_struct *work)
+isert_conn_terminate(struct isert_conn *isert_conn)
 {
-	struct isert_conn *isert_conn = container_of(work,
-				struct isert_conn, conn_logout_work);
+	int err;
 
-	pr_debug("isert_disconnect_work(): >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\n");
-	mutex_lock(&isert_conn->conn_mutex);
-	if (isert_conn->state == ISER_CONN_UP)
+	switch (isert_conn->state) {
+	case ISER_CONN_TERMINATING:
+		break;
+	case ISER_CONN_UP:
+	case ISER_CONN_FULL_FEATURE: /* FALLTHRU */
+		isert_info("Terminating conn %p state %d\n",
+			   isert_conn, isert_conn->state);
 		isert_conn->state = ISER_CONN_TERMINATING;
-
-	if (isert_conn->post_recv_buf_count == 0 &&
-	    atomic_read(&isert_conn->post_send_buf_count) == 0) {
-		mutex_unlock(&isert_conn->conn_mutex);
-		goto wake_up;
-	}
-	if (!isert_conn->conn_cm_id) {
-		mutex_unlock(&isert_conn->conn_mutex);
-		isert_put_conn(isert_conn);
-		return;
+		err = rdma_disconnect(isert_conn->conn_cm_id);
+		if (err)
+			isert_warn("Failed rdma_disconnect isert_conn %p\n",
+				   isert_conn);
+		break;
+	default:
+		isert_warn("conn %p teminating in state %d\n",
+			   isert_conn, isert_conn->state);
 	}
+}
+
+static int
+isert_np_cma_handler(struct isert_np *isert_np,
+		     enum rdma_cm_event_type event)
+{
+	isert_dbg("isert np %p, handling event %d\n", isert_np, event);
 
-	if (isert_conn->disconnect) {
-		/* Send DREQ/DREP towards our initiator */
-		rdma_disconnect(isert_conn->conn_cm_id);
+	switch (event) {
+	case RDMA_CM_EVENT_DEVICE_REMOVAL:
+		isert_np->np_cm_id = NULL;
+		break;
+	case RDMA_CM_EVENT_ADDR_CHANGE:
+		isert_np->np_cm_id = isert_setup_id(isert_np);
+		if (IS_ERR(isert_np->np_cm_id)) {
+			isert_err("isert np %p setup id failed: %ld\n",
+				  isert_np, PTR_ERR(isert_np->np_cm_id));
+			isert_np->np_cm_id = NULL;
+		}
+		break;
+	default:
+		isert_err("isert np %p Unexpected event %d\n",
+			  isert_np, event);
 	}
 
+	return -1;
+}
+
+static int
+isert_disconnected_handler(struct rdma_cm_id *cma_id,
+			   enum rdma_cm_event_type event)
+{
+	struct isert_np *isert_np = cma_id->context;
+	struct isert_conn *isert_conn;
+
+	if (isert_np->np_cm_id == cma_id)
+		return isert_np_cma_handler(cma_id->context, event);
+
+	isert_conn = cma_id->qp->qp_context;
+
+	mutex_lock(&isert_conn->conn_mutex);
+	isert_conn_terminate(isert_conn);
 	mutex_unlock(&isert_conn->conn_mutex);
 
-wake_up:
+	isert_info("conn %p completing conn_wait\n", isert_conn);
 	complete(&isert_conn->conn_wait);
+
+	return 0;
 }
 
 static void
-isert_disconnected_handler(struct rdma_cm_id *cma_id, bool disconnect)
+isert_connect_error(struct rdma_cm_id *cma_id)
 {
-	struct isert_conn *isert_conn = (struct isert_conn *)cma_id->context;
+	struct isert_conn *isert_conn = cma_id->qp->qp_context;
 
-	isert_conn->disconnect = disconnect;
-	INIT_WORK(&isert_conn->conn_logout_work, isert_disconnect_work);
-	schedule_work(&isert_conn->conn_logout_work);
+	isert_put_conn(isert_conn);
 }
 
 static int
 isert_cma_handler(struct rdma_cm_id *cma_id, struct rdma_cm_event *event)
 {
 	int ret = 0;
-	bool disconnect = false;
 
-	pr_debug("isert_cma_handler: event %d status %d conn %p id %p\n",
-		 event->event, event->status, cma_id->context, cma_id);
+	isert_info("event %d status %d id %p np %p\n", event->event,
+		   event->status, cma_id, cma_id->context);
 
 	switch (event->event) {
 	case RDMA_CM_EVENT_CONNECT_REQUEST:
 		ret = isert_connect_request(cma_id, event);
+		if (ret)
+			isert_err("failed handle connect request %d\n", ret);
 		break;
 	case RDMA_CM_EVENT_ESTABLISHED:
 		isert_connected_handler(cma_id);
@@ -832,22 +906,19 @@ isert_cma_handler(struct rdma_cm_id *cma_id, struct rdma_cm_event *event)
 	case RDMA_CM_EVENT_ADDR_CHANGE:    /* FALLTHRU */
 	case RDMA_CM_EVENT_DISCONNECTED:   /* FALLTHRU */
 	case RDMA_CM_EVENT_DEVICE_REMOVAL: /* FALLTHRU */
-		disconnect = true;
 	case RDMA_CM_EVENT_TIMEWAIT_EXIT:  /* FALLTHRU */
-		isert_disconnected_handler(cma_id, disconnect);
+		ret = isert_disconnected_handler(cma_id, event->event);
 		break;
+	case RDMA_CM_EVENT_REJECTED:       /* FALLTHRU */
+	case RDMA_CM_EVENT_UNREACHABLE:    /* FALLTHRU */
 	case RDMA_CM_EVENT_CONNECT_ERROR:
+		isert_connect_error(cma_id);
+		break;
 	default:
-		pr_err("Unhandled RDMA CMA event: %d\n", event->event);
+		isert_err("Unhandled RDMA CMA event: %d\n", event->event);
 		break;
 	}
 
-	if (ret != 0) {
-		pr_err("isert_cma_handler failed RDMA_CM_EVENT: 0x%08x %d\n",
-		       event->event, ret);
-		dump_stack();
-	}
-
 	return ret;
 }
 
@@ -861,7 +932,7 @@ isert_post_recv(struct isert_conn *isert_conn, u32 count)
 
 	for (rx_wr = isert_conn->conn_rx_wr, i = 0; i < count; i++, rx_wr++) {
 		rx_desc		= &isert_conn->conn_rx_descs[rx_head];
-		rx_wr->wr_id	= (unsigned long)rx_desc;
+		rx_wr->wr_id	= (uintptr_t)rx_desc;
 		rx_wr->sg_list	= &rx_desc->rx_sg;
 		rx_wr->num_sge	= 1;
 		rx_wr->next	= rx_wr + 1;
@@ -875,10 +946,10 @@ isert_post_recv(struct isert_conn *isert_conn, u32 count)
 	ret = ib_post_recv(isert_conn->conn_qp, isert_conn->conn_rx_wr,
 				&rx_wr_failed);
 	if (ret) {
-		pr_err("ib_post_recv() failed with ret: %d\n", ret);
+		isert_err("ib_post_recv() failed with ret: %d\n", ret);
 		isert_conn->post_recv_buf_count -= count;
 	} else {
-		pr_debug("isert_post_recv(): Posted %d RX buffers\n", count);
+		isert_dbg("isert_post_recv(): Posted %d RX buffers\n", count);
 		isert_conn->conn_rx_desc_head = rx_head;
 	}
 	return ret;
@@ -895,19 +966,15 @@ isert_post_send(struct isert_conn *isert_conn, struct iser_tx_desc *tx_desc)
 				      ISER_HEADERS_LEN, DMA_TO_DEVICE);
 
 	send_wr.next	= NULL;
-	send_wr.wr_id	= (unsigned long)tx_desc;
+	send_wr.wr_id	= (uintptr_t)tx_desc;
 	send_wr.sg_list	= tx_desc->tx_sg;
 	send_wr.num_sge	= tx_desc->num_sge;
 	send_wr.opcode	= IB_WR_SEND;
 	send_wr.send_flags = IB_SEND_SIGNALED;
 
-	atomic_inc(&isert_conn->post_send_buf_count);
-
 	ret = ib_post_send(isert_conn->conn_qp, &send_wr, &send_wr_failed);
-	if (ret) {
-		pr_err("ib_post_send() failed, ret: %d\n", ret);
-		atomic_dec(&isert_conn->post_send_buf_count);
-	}
+	if (ret)
+		isert_err("ib_post_send() failed, ret: %d\n", ret);
 
 	return ret;
 }
@@ -930,7 +997,7 @@ isert_create_send_desc(struct isert_conn *isert_conn,
 
 	if (tx_desc->tx_sg[0].lkey != isert_conn->conn_mr->lkey) {
 		tx_desc->tx_sg[0].lkey = isert_conn->conn_mr->lkey;
-		pr_debug("tx_desc %p lkey mismatch, fixing\n", tx_desc);
+		isert_dbg("tx_desc %p lkey mismatch, fixing\n", tx_desc);
 	}
 }
 
@@ -944,7 +1011,7 @@ isert_init_tx_hdrs(struct isert_conn *isert_conn,
 	dma_addr = ib_dma_map_single(ib_dev, (void *)tx_desc,
 			ISER_HEADERS_LEN, DMA_TO_DEVICE);
 	if (ib_dma_mapping_error(ib_dev, dma_addr)) {
-		pr_err("ib_dma_mapping_error() failed\n");
+		isert_err("ib_dma_mapping_error() failed\n");
 		return -ENOMEM;
 	}
 
@@ -953,40 +1020,24 @@ isert_init_tx_hdrs(struct isert_conn *isert_conn,
 	tx_desc->tx_sg[0].length = ISER_HEADERS_LEN;
 	tx_desc->tx_sg[0].lkey = isert_conn->conn_mr->lkey;
 
-	pr_debug("isert_init_tx_hdrs: Setup tx_sg[0].addr: 0x%llx length: %u"
-		 " lkey: 0x%08x\n", tx_desc->tx_sg[0].addr,
-		 tx_desc->tx_sg[0].length, tx_desc->tx_sg[0].lkey);
+	isert_dbg("Setup tx_sg[0].addr: 0x%llx length: %u lkey: 0x%x\n",
+		  tx_desc->tx_sg[0].addr, tx_desc->tx_sg[0].length,
+		  tx_desc->tx_sg[0].lkey);
 
 	return 0;
 }
 
 static void
 isert_init_send_wr(struct isert_conn *isert_conn, struct isert_cmd *isert_cmd,
-		   struct ib_send_wr *send_wr, bool coalesce)
+		   struct ib_send_wr *send_wr)
 {
 	struct iser_tx_desc *tx_desc = &isert_cmd->tx_desc;
 
 	isert_cmd->rdma_wr.iser_ib_op = ISER_IB_SEND;
-	send_wr->wr_id = (unsigned long)&isert_cmd->tx_desc;
+	send_wr->wr_id = (uintptr_t)&isert_cmd->tx_desc;
 	send_wr->opcode = IB_WR_SEND;
 	send_wr->sg_list = &tx_desc->tx_sg[0];
 	send_wr->num_sge = isert_cmd->tx_desc.num_sge;
-	/*
-	 * Coalesce send completion interrupts by only setting IB_SEND_SIGNALED
-	 * bit for every ISERT_COMP_BATCH_COUNT number of ib_post_send() calls.
-	 */
-	mutex_lock(&isert_conn->conn_mutex);
-	if (coalesce && isert_conn->state == ISER_CONN_UP &&
-	    ++isert_conn->conn_comp_batch < ISERT_COMP_BATCH_COUNT) {
-		tx_desc->llnode_active = true;
-		llist_add(&tx_desc->comp_llnode, &isert_conn->conn_comp_llist);
-		mutex_unlock(&isert_conn->conn_mutex);
-		return;
-	}
-	isert_conn->conn_comp_batch = 0;
-	tx_desc->comp_llnode_batch = llist_del_all(&isert_conn->conn_comp_llist);
-	mutex_unlock(&isert_conn->conn_mutex);
-
 	send_wr->send_flags = IB_SEND_SIGNALED;
 }
 
@@ -1002,22 +1053,21 @@ isert_rdma_post_recvl(struct isert_conn *isert_conn)
 	sge.length = ISER_RX_LOGIN_SIZE;
 	sge.lkey = isert_conn->conn_mr->lkey;
 
-	pr_debug("Setup sge: addr: %llx length: %d 0x%08x\n",
+	isert_dbg("Setup sge: addr: %llx length: %d 0x%08x\n",
 		sge.addr, sge.length, sge.lkey);
 
 	memset(&rx_wr, 0, sizeof(struct ib_recv_wr));
-	rx_wr.wr_id = (unsigned long)isert_conn->login_req_buf;
+	rx_wr.wr_id = (uintptr_t)isert_conn->login_req_buf;
 	rx_wr.sg_list = &sge;
 	rx_wr.num_sge = 1;
 
 	isert_conn->post_recv_buf_count++;
 	ret = ib_post_recv(isert_conn->conn_qp, &rx_wr, &rx_wr_fail);
 	if (ret) {
-		pr_err("ib_post_recv() failed: %d\n", ret);
+		isert_err("ib_post_recv() failed: %d\n", ret);
 		isert_conn->post_recv_buf_count--;
 	}
 
-	pr_debug("ib_post_recv(): returned success >>>>>>>>>>>>>>>>>>>>>>>>\n");
 	return ret;
 }
 
@@ -1057,13 +1107,9 @@ isert_put_login_tx(struct iscsi_conn *conn, struct iscsi_login *login,
 		if (login->login_complete) {
 			if (!conn->sess->sess_ops->SessionType &&
 			    isert_conn->conn_device->use_fastreg) {
-				/* Normal Session and fastreg is used */
-				u8 pi_support = login->np->tpg_np->tpg->tpg_attrib.t10_pi;
-
-				ret = isert_conn_create_fastreg_pool(isert_conn,
-								     pi_support);
+				ret = isert_conn_create_fastreg_pool(isert_conn);
 				if (ret) {
-					pr_err("Conn: %p failed to create"
+					isert_err("Conn: %p failed to create"
 					       " fastreg pool\n", isert_conn);
 					return ret;
 				}
@@ -1077,7 +1123,10 @@ isert_put_login_tx(struct iscsi_conn *conn, struct iscsi_login *login,
 			if (ret)
 				return ret;
 
-			isert_conn->state = ISER_CONN_UP;
+			/* Now we are in FULL_FEATURE phase */
+			mutex_lock(&isert_conn->conn_mutex);
+			isert_conn->state = ISER_CONN_FULL_FEATURE;
+			mutex_unlock(&isert_conn->conn_mutex);
 			goto post_send;
 		}
 
@@ -1094,18 +1143,17 @@ post_send:
 }
 
 static void
-isert_rx_login_req(struct iser_rx_desc *rx_desc, int rx_buflen,
-		   struct isert_conn *isert_conn)
+isert_rx_login_req(struct isert_conn *isert_conn)
 {
+	struct iser_rx_desc *rx_desc = (void *)isert_conn->login_req_buf;
+	int rx_buflen = isert_conn->login_req_len;
 	struct iscsi_conn *conn = isert_conn->conn;
 	struct iscsi_login *login = conn->conn_login;
 	int size;
 
-	if (!login) {
-		pr_err("conn->conn_login is NULL\n");
-		dump_stack();
-		return;
-	}
+	isert_info("conn %p\n", isert_conn);
+
+	WARN_ON_ONCE(!login);
 
 	if (login->first_request) {
 		struct iscsi_login_req *login_req =
@@ -1131,8 +1179,9 @@ isert_rx_login_req(struct iser_rx_desc *rx_desc, int rx_buflen,
 	memcpy(&login->req[0], (void *)&rx_desc->iscsi_header, ISCSI_HDR_LEN);
 
 	size = min(rx_buflen, MAX_KEY_VALUE_PAIRS);
-	pr_debug("Using login payload size: %d, rx_buflen: %d MAX_KEY_VALUE_PAIRS: %d\n",
-		 size, rx_buflen, MAX_KEY_VALUE_PAIRS);
+	isert_dbg("Using login payload size: %d, rx_buflen: %d "
+		  "MAX_KEY_VALUE_PAIRS: %d\n", size, rx_buflen,
+		  MAX_KEY_VALUE_PAIRS);
 	memcpy(login->req_buf, &rx_desc->data[0], size);
 
 	if (login->first_request) {
@@ -1151,7 +1200,7 @@ static struct iscsi_cmd
 
 	cmd = iscsit_allocate_cmd(conn, TASK_INTERRUPTIBLE);
 	if (!cmd) {
-		pr_err("Unable to allocate iscsi_cmd + isert_cmd\n");
+		isert_err("Unable to allocate iscsi_cmd + isert_cmd\n");
 		return NULL;
 	}
 	isert_cmd = iscsit_priv_cmd(cmd);
@@ -1194,8 +1243,8 @@ isert_handle_scsi_cmd(struct isert_conn *isert_conn,
 	sg = &cmd->se_cmd.t_data_sg[0];
 	sg_nents = max(1UL, DIV_ROUND_UP(imm_data_len, PAGE_SIZE));
 
-	pr_debug("Copying Immediate SG: %p sg_nents: %u from %p imm_data_len: %d\n",
-		 sg, sg_nents, &rx_desc->data[0], imm_data_len);
+	isert_dbg("Copying Immediate SG: %p sg_nents: %u from %p imm_data_len: %d\n",
+		  sg, sg_nents, &rx_desc->data[0], imm_data_len);
 
 	sg_copy_from_buffer(sg, sg_nents, &rx_desc->data[0], imm_data_len);
 
@@ -1239,13 +1288,15 @@ isert_handle_iscsi_dataout(struct isert_conn *isert_conn,
 	 * FIXME: Unexpected unsolicited_data out
 	 */
 	if (!cmd->unsolicited_data) {
-		pr_err("Received unexpected solicited data payload\n");
+		isert_err("Received unexpected solicited data payload\n");
 		dump_stack();
 		return -1;
 	}
 
-	pr_debug("Unsolicited DataOut unsol_data_len: %u, write_data_done: %u, data_length: %u\n",
-		 unsol_data_len, cmd->write_data_done, cmd->se_cmd.data_length);
+	isert_dbg("Unsolicited DataOut unsol_data_len: %u, "
+		  "write_data_done: %u, data_length: %u\n",
+		  unsol_data_len,  cmd->write_data_done,
+		  cmd->se_cmd.data_length);
 
 	sg_off = cmd->write_data_done / PAGE_SIZE;
 	sg_start = &cmd->se_cmd.t_data_sg[sg_off];
@@ -1255,12 +1306,13 @@ isert_handle_iscsi_dataout(struct isert_conn *isert_conn,
 	 * FIXME: Non page-aligned unsolicited_data out
 	 */
 	if (page_off) {
-		pr_err("Received unexpected non-page aligned data payload\n");
+		isert_err("unexpected non-page aligned data payload\n");
 		dump_stack();
 		return -1;
 	}
-	pr_debug("Copying DataOut: sg_start: %p, sg_off: %u sg_nents: %u from %p %u\n",
-		 sg_start, sg_off, sg_nents, &rx_desc->data[0], unsol_data_len);
+	isert_dbg("Copying DataOut: sg_start: %p, sg_off: %u "
+		  "sg_nents: %u from %p %u\n", sg_start, sg_off,
+		  sg_nents, &rx_desc->data[0], unsol_data_len);
 
 	sg_copy_from_buffer(sg_start, sg_nents, &rx_desc->data[0],
 			    unsol_data_len);
@@ -1307,8 +1359,8 @@ isert_handle_text_cmd(struct isert_conn *isert_conn, struct isert_cmd *isert_cmd
 
 	text_in = kzalloc(payload_length, GFP_KERNEL);
 	if (!text_in) {
-		pr_err("Unable to allocate text_in of payload_length: %u\n",
-		       payload_length);
+		isert_err("Unable to allocate text_in of payload_length: %u\n",
+			  payload_length);
 		return -ENOMEM;
 	}
 	cmd->text_in_ptr = text_in;
@@ -1333,8 +1385,8 @@ isert_rx_opcode(struct isert_conn *isert_conn, struct iser_rx_desc *rx_desc,
 
 	if (sess->sess_ops->SessionType &&
 	   (!(opcode & ISCSI_OP_TEXT) || !(opcode & ISCSI_OP_LOGOUT))) {
-		pr_err("Got illegal opcode: 0x%02x in SessionType=Discovery,"
-		       " ignoring\n", opcode);
+		isert_err("Got illegal opcode: 0x%02x in SessionType=Discovery,"
+			  " ignoring\n", opcode);
 		return 0;
 	}
 
@@ -1380,10 +1432,6 @@ isert_rx_opcode(struct isert_conn *isert_conn, struct iser_rx_desc *rx_desc,
 			break;
 
 		ret = iscsit_handle_logout_cmd(conn, cmd, (unsigned char *)hdr);
-		if (ret > 0)
-			wait_for_completion_timeout(&conn->conn_logout_comp,
-						    SECONDS_FOR_LOGOUT_COMP *
-						    HZ);
 		break;
 	case ISCSI_OP_TEXT:
 		cmd = isert_allocate_cmd(conn);
@@ -1395,7 +1443,7 @@ isert_rx_opcode(struct isert_conn *isert_conn, struct iser_rx_desc *rx_desc,
 					    rx_desc, (struct iscsi_text *)hdr);
 		break;
 	default:
-		pr_err("Got unknown iSCSI OpCode: 0x%02x\n", opcode);
+		isert_err("Got unknown iSCSI OpCode: 0x%02x\n", opcode);
 		dump_stack();
 		break;
 	}
@@ -1416,23 +1464,23 @@ isert_rx_do_work(struct iser_rx_desc *rx_desc, struct isert_conn *isert_conn)
 		if (iser_hdr->flags & ISER_RSV) {
 			read_stag = be32_to_cpu(iser_hdr->read_stag);
 			read_va = be64_to_cpu(iser_hdr->read_va);
-			pr_debug("ISER_RSV: read_stag: 0x%08x read_va: 0x%16llx\n",
-				 read_stag, (unsigned long long)read_va);
+			isert_dbg("ISER_RSV: read_stag: 0x%x read_va: 0x%llx\n",
+				  read_stag, (unsigned long long)read_va);
 		}
 		if (iser_hdr->flags & ISER_WSV) {
 			write_stag = be32_to_cpu(iser_hdr->write_stag);
 			write_va = be64_to_cpu(iser_hdr->write_va);
-			pr_debug("ISER_WSV: write__stag: 0x%08x write_va: 0x%16llx\n",
-				 write_stag, (unsigned long long)write_va);
+			isert_dbg("ISER_WSV: write_stag: 0x%x write_va: 0x%llx\n",
+				  write_stag, (unsigned long long)write_va);
 		}
 
-		pr_debug("ISER ISCSI_CTRL PDU\n");
+		isert_dbg("ISER ISCSI_CTRL PDU\n");
 		break;
 	case ISER_HELLO:
-		pr_err("iSER Hello message\n");
+		isert_err("iSER Hello message\n");
 		break;
 	default:
-		pr_warn("Unknown iSER hdr flags: 0x%02x\n", iser_hdr->flags);
+		isert_warn("Unknown iSER hdr flags: 0x%02x\n", iser_hdr->flags);
 		break;
 	}
 
@@ -1442,7 +1490,7 @@ isert_rx_do_work(struct iser_rx_desc *rx_desc, struct isert_conn *isert_conn)
 
 static void
 isert_rx_completion(struct iser_rx_desc *desc, struct isert_conn *isert_conn,
-		    unsigned long xfer_len)
+		    u32 xfer_len)
 {
 	struct ib_device *ib_dev = isert_conn->conn_cm_id->device;
 	struct iscsi_hdr *hdr;
@@ -1452,34 +1500,43 @@ isert_rx_completion(struct iser_rx_desc *desc, struct isert_conn *isert_conn,
 	if ((char *)desc == isert_conn->login_req_buf) {
 		rx_dma = isert_conn->login_req_dma;
 		rx_buflen = ISER_RX_LOGIN_SIZE;
-		pr_debug("ISER login_buf: Using rx_dma: 0x%llx, rx_buflen: %d\n",
+		isert_dbg("login_buf: Using rx_dma: 0x%llx, rx_buflen: %d\n",
 			 rx_dma, rx_buflen);
 	} else {
 		rx_dma = desc->dma_addr;
 		rx_buflen = ISER_RX_PAYLOAD_SIZE;
-		pr_debug("ISER req_buf: Using rx_dma: 0x%llx, rx_buflen: %d\n",
+		isert_dbg("req_buf: Using rx_dma: 0x%llx, rx_buflen: %d\n",
 			 rx_dma, rx_buflen);
 	}
 
 	ib_dma_sync_single_for_cpu(ib_dev, rx_dma, rx_buflen, DMA_FROM_DEVICE);
 
 	hdr = &desc->iscsi_header;
-	pr_debug("iSCSI opcode: 0x%02x, ITT: 0x%08x, flags: 0x%02x dlen: %d\n",
+	isert_dbg("iSCSI opcode: 0x%02x, ITT: 0x%08x, flags: 0x%02x dlen: %d\n",
 		 hdr->opcode, hdr->itt, hdr->flags,
 		 (int)(xfer_len - ISER_HEADERS_LEN));
 
-	if ((char *)desc == isert_conn->login_req_buf)
-		isert_rx_login_req(desc, xfer_len - ISER_HEADERS_LEN,
-				   isert_conn);
-	else
+	if ((char *)desc == isert_conn->login_req_buf) {
+		isert_conn->login_req_len = xfer_len - ISER_HEADERS_LEN;
+		if (isert_conn->conn) {
+			struct iscsi_login *login = isert_conn->conn->conn_login;
+
+			if (login && !login->first_request)
+				isert_rx_login_req(isert_conn);
+		}
+		mutex_lock(&isert_conn->conn_mutex);
+		complete(&isert_conn->login_req_comp);
+		mutex_unlock(&isert_conn->conn_mutex);
+	} else {
 		isert_rx_do_work(desc, isert_conn);
+	}
 
 	ib_dma_sync_single_for_device(ib_dev, rx_dma, rx_buflen,
 				      DMA_FROM_DEVICE);
 
 	isert_conn->post_recv_buf_count--;
-	pr_debug("iSERT: Decremented post_recv_buf_count: %d\n",
-		 isert_conn->post_recv_buf_count);
+	isert_dbg("Decremented post_recv_buf_count: %d\n",
+		  isert_conn->post_recv_buf_count);
 
 	if ((char *)desc == isert_conn->login_req_buf)
 		return;
@@ -1490,7 +1547,7 @@ isert_rx_completion(struct iser_rx_desc *desc, struct isert_conn *isert_conn,
 				ISERT_MIN_POSTED_RX);
 		err = isert_post_recv(isert_conn, count);
 		if (err) {
-			pr_err("isert_post_recv() count: %d failed, %d\n",
+			isert_err("isert_post_recv() count: %d failed, %d\n",
 			       count, err);
 		}
 	}
@@ -1519,12 +1576,12 @@ isert_map_data_buf(struct isert_conn *isert_conn, struct isert_cmd *isert_cmd,
 	data->dma_nents = ib_dma_map_sg(ib_dev, data->sg, data->nents,
 					data->dma_dir);
 	if (unlikely(!data->dma_nents)) {
-		pr_err("Cmd: unable to dma map SGs %p\n", sg);
+		isert_err("Cmd: unable to dma map SGs %p\n", sg);
 		return -EINVAL;
 	}
 
-	pr_debug("Mapped cmd: %p count: %u sg: %p sg_nents: %u rdma_len %d\n",
-		 isert_cmd, data->dma_nents, data->sg, data->nents, data->len);
+	isert_dbg("Mapped cmd: %p count: %u sg: %p sg_nents: %u rdma_len %d\n",
+		  isert_cmd, data->dma_nents, data->sg, data->nents, data->len);
 
 	return 0;
 }
@@ -1545,21 +1602,21 @@ isert_unmap_cmd(struct isert_cmd *isert_cmd, struct isert_conn *isert_conn)
 {
 	struct isert_rdma_wr *wr = &isert_cmd->rdma_wr;
 
-	pr_debug("isert_unmap_cmd: %p\n", isert_cmd);
+	isert_dbg("Cmd %p\n", isert_cmd);
 
 	if (wr->data.sg) {
-		pr_debug("isert_unmap_cmd: %p unmap_sg op\n", isert_cmd);
+		isert_dbg("Cmd %p unmap_sg op\n", isert_cmd);
 		isert_unmap_data_buf(isert_conn, &wr->data);
 	}
 
 	if (wr->send_wr) {
-		pr_debug("isert_unmap_cmd: %p free send_wr\n", isert_cmd);
+		isert_dbg("Cmd %p free send_wr\n", isert_cmd);
 		kfree(wr->send_wr);
 		wr->send_wr = NULL;
 	}
 
 	if (wr->ib_sge) {
-		pr_debug("isert_unmap_cmd: %p free ib_sge\n", isert_cmd);
+		isert_dbg("Cmd %p free ib_sge\n", isert_cmd);
 		kfree(wr->ib_sge);
 		wr->ib_sge = NULL;
 	}
@@ -1571,11 +1628,10 @@ isert_unreg_rdma(struct isert_cmd *isert_cmd, struct isert_conn *isert_conn)
 	struct isert_rdma_wr *wr = &isert_cmd->rdma_wr;
 	LIST_HEAD(unmap_list);
 
-	pr_debug("unreg_fastreg_cmd: %p\n", isert_cmd);
+	isert_dbg("Cmd %p\n", isert_cmd);
 
 	if (wr->fr_desc) {
-		pr_debug("unreg_fastreg_cmd: %p free fr_desc %p\n",
-			 isert_cmd, wr->fr_desc);
+		isert_dbg("Cmd %p free fr_desc %p\n", isert_cmd, wr->fr_desc);
 		if (wr->fr_desc->ind & ISERT_PROTECTED) {
 			isert_unmap_data_buf(isert_conn, &wr->prot);
 			wr->fr_desc->ind &= ~ISERT_PROTECTED;
@@ -1587,7 +1643,7 @@ isert_unreg_rdma(struct isert_cmd *isert_cmd, struct isert_conn *isert_conn)
 	}
 
 	if (wr->data.sg) {
-		pr_debug("unreg_fastreg_cmd: %p unmap_sg op\n", isert_cmd);
+		isert_dbg("Cmd %p unmap_sg op\n", isert_cmd);
 		isert_unmap_data_buf(isert_conn, &wr->data);
 	}
 
@@ -1603,7 +1659,7 @@ isert_put_cmd(struct isert_cmd *isert_cmd, bool comp_err)
 	struct iscsi_conn *conn = isert_conn->conn;
 	struct isert_device *device = isert_conn->conn_device;
 
-	pr_debug("Entering isert_put_cmd: %p\n", isert_cmd);
+	isert_dbg("Cmd %p\n", isert_cmd);
 
 	switch (cmd->iscsi_opcode) {
 	case ISCSI_OP_SCSI_CMD:
@@ -1653,7 +1709,7 @@ isert_put_cmd(struct isert_cmd *isert_cmd, bool comp_err)
 		 * associated cmd->se_cmd needs to be released.
 		 */
 		if (cmd->se_cmd.se_tfo != NULL) {
-			pr_debug("Calling transport_generic_free_cmd from"
+			isert_dbg("Calling transport_generic_free_cmd from"
 				 " isert_put_cmd for 0x%02x\n",
 				 cmd->iscsi_opcode);
 			transport_generic_free_cmd(&cmd->se_cmd, 0);
@@ -1672,7 +1728,7 @@ static void
 isert_unmap_tx_desc(struct iser_tx_desc *tx_desc, struct ib_device *ib_dev)
 {
 	if (tx_desc->dma_addr != 0) {
-		pr_debug("Calling ib_dma_unmap_single for tx_desc->dma_addr\n");
+		isert_dbg("unmap single for tx_desc->dma_addr\n");
 		ib_dma_unmap_single(ib_dev, tx_desc->dma_addr,
 				    ISER_HEADERS_LEN, DMA_TO_DEVICE);
 		tx_desc->dma_addr = 0;
@@ -1684,7 +1740,7 @@ isert_completion_put(struct iser_tx_desc *tx_desc, struct isert_cmd *isert_cmd,
 		     struct ib_device *ib_dev, bool comp_err)
 {
 	if (isert_cmd->pdu_buf_dma != 0) {
-		pr_debug("Calling ib_dma_unmap_single for isert_cmd->pdu_buf_dma\n");
+		isert_dbg("unmap single for isert_cmd->pdu_buf_dma\n");
 		ib_dma_unmap_single(ib_dev, isert_cmd->pdu_buf_dma,
 				    isert_cmd->pdu_buf_len, DMA_TO_DEVICE);
 		isert_cmd->pdu_buf_dma = 0;
@@ -1702,7 +1758,7 @@ isert_check_pi_status(struct se_cmd *se_cmd, struct ib_mr *sig_mr)
 
 	ret = ib_check_mr_status(sig_mr, IB_MR_CHECK_SIG_STATUS, &mr_status);
 	if (ret) {
-		pr_err("ib_check_mr_status failed, ret %d\n", ret);
+		isert_err("ib_check_mr_status failed, ret %d\n", ret);
 		goto fail_mr_status;
 	}
 
@@ -1725,12 +1781,12 @@ isert_check_pi_status(struct se_cmd *se_cmd, struct ib_mr *sig_mr)
 		do_div(sec_offset_err, block_size);
 		se_cmd->bad_sector = sec_offset_err + se_cmd->t_task_lba;
 
-		pr_err("isert: PI error found type %d at sector 0x%llx "
-		       "expected 0x%x vs actual 0x%x\n",
-		       mr_status.sig_err.err_type,
-		       (unsigned long long)se_cmd->bad_sector,
-		       mr_status.sig_err.expected,
-		       mr_status.sig_err.actual);
+		isert_err("PI error found type %d at sector 0x%llx "
+			  "expected 0x%x vs actual 0x%x\n",
+			  mr_status.sig_err.err_type,
+			  (unsigned long long)se_cmd->bad_sector,
+			  mr_status.sig_err.expected,
+			  mr_status.sig_err.actual);
 		ret = 1;
 	}
 
@@ -1786,7 +1842,7 @@ isert_completion_rdma_read(struct iser_tx_desc *tx_desc,
 	cmd->write_data_done = wr->data.len;
 	wr->send_wr_num = 0;
 
-	pr_debug("Cmd: %p RDMA_READ comp calling execute_cmd\n", isert_cmd);
+	isert_dbg("Cmd: %p RDMA_READ comp calling execute_cmd\n", isert_cmd);
 	spin_lock_bh(&cmd->istate_lock);
 	cmd->cmd_flags |= ICF_GOT_LAST_DATAOUT;
 	cmd->i_state = ISTATE_RECEIVED_LAST_DATAOUT;
@@ -1808,36 +1864,22 @@ isert_do_control_comp(struct work_struct *work)
 	struct ib_device *ib_dev = isert_conn->conn_cm_id->device;
 	struct iscsi_cmd *cmd = isert_cmd->iscsi_cmd;
 
+	isert_dbg("Cmd %p i_state %d\n", isert_cmd, cmd->i_state);
+
 	switch (cmd->i_state) {
 	case ISTATE_SEND_TASKMGTRSP:
-		pr_debug("Calling iscsit_tmr_post_handler >>>>>>>>>>>>>>>>>\n");
-
-		atomic_dec(&isert_conn->post_send_buf_count);
 		iscsit_tmr_post_handler(cmd, cmd->conn);
-
-		cmd->i_state = ISTATE_SENT_STATUS;
-		isert_completion_put(&isert_cmd->tx_desc, isert_cmd, ib_dev, false);
-		break;
-	case ISTATE_SEND_REJECT:
-		pr_debug("Got isert_do_control_comp ISTATE_SEND_REJECT: >>>\n");
-		atomic_dec(&isert_conn->post_send_buf_count);
-
+	case ISTATE_SEND_REJECT:   /* FALLTHRU */
+	case ISTATE_SEND_TEXTRSP:  /* FALLTHRU */
 		cmd->i_state = ISTATE_SENT_STATUS;
-		isert_completion_put(&isert_cmd->tx_desc, isert_cmd, ib_dev, false);
+		isert_completion_put(&isert_cmd->tx_desc, isert_cmd,
+				     ib_dev, false);
 		break;
 	case ISTATE_SEND_LOGOUTRSP:
-		pr_debug("Calling iscsit_logout_post_handler >>>>>>>>>>>>>>\n");
-
-		atomic_dec(&isert_conn->post_send_buf_count);
 		iscsit_logout_post_handler(cmd, cmd->conn);
 		break;
-	case ISTATE_SEND_TEXTRSP:
-		atomic_dec(&isert_conn->post_send_buf_count);
-		cmd->i_state = ISTATE_SENT_STATUS;
-		isert_completion_put(&isert_cmd->tx_desc, isert_cmd, ib_dev, false);
-		break;
 	default:
-		pr_err("Unknown do_control_comp i_state %d\n", cmd->i_state);
+		isert_err("Unknown i_state %d\n", cmd->i_state);
 		dump_stack();
 		break;
 	}
@@ -1850,7 +1892,6 @@ isert_response_completion(struct iser_tx_desc *tx_desc,
 			  struct ib_device *ib_dev)
 {
 	struct iscsi_cmd *cmd = isert_cmd->iscsi_cmd;
-	struct isert_rdma_wr *wr = &isert_cmd->rdma_wr;
 
 	if (cmd->i_state == ISTATE_SEND_TASKMGTRSP ||
 	    cmd->i_state == ISTATE_SEND_LOGOUTRSP ||
@@ -1863,267 +1904,151 @@ isert_response_completion(struct iser_tx_desc *tx_desc,
 		return;
 	}
 
-	/**
-	 * If send_wr_num is 0 this means that we got
-	 * RDMA completion and we cleared it and we should
-	 * simply decrement the response post. else the
-	 * response is incorporated in send_wr_num, just
-	 * sub it.
-	 **/
-	if (wr->send_wr_num)
-		atomic_sub(wr->send_wr_num, &isert_conn->post_send_buf_count);
-	else
-		atomic_dec(&isert_conn->post_send_buf_count);
-
 	cmd->i_state = ISTATE_SENT_STATUS;
 	isert_completion_put(tx_desc, isert_cmd, ib_dev, false);
 }
 
 static void
-__isert_send_completion(struct iser_tx_desc *tx_desc,
-		        struct isert_conn *isert_conn)
+isert_send_completion(struct iser_tx_desc *tx_desc,
+		      struct isert_conn *isert_conn)
 {
 	struct ib_device *ib_dev = isert_conn->conn_cm_id->device;
 	struct isert_cmd *isert_cmd = tx_desc->isert_cmd;
 	struct isert_rdma_wr *wr;
 
 	if (!isert_cmd) {
-		atomic_dec(&isert_conn->post_send_buf_count);
 		isert_unmap_tx_desc(tx_desc, ib_dev);
 		return;
 	}
 	wr = &isert_cmd->rdma_wr;
 
+	isert_dbg("Cmd %p iser_ib_op %d\n", isert_cmd, wr->iser_ib_op);
+
 	switch (wr->iser_ib_op) {
 	case ISER_IB_RECV:
-		pr_err("isert_send_completion: Got ISER_IB_RECV\n");
+		isert_err("Got ISER_IB_RECV\n");
 		dump_stack();
 		break;
 	case ISER_IB_SEND:
-		pr_debug("isert_send_completion: Got ISER_IB_SEND\n");
 		isert_response_completion(tx_desc, isert_cmd,
 					  isert_conn, ib_dev);
 		break;
 	case ISER_IB_RDMA_WRITE:
-		pr_debug("isert_send_completion: Got ISER_IB_RDMA_WRITE\n");
-		atomic_sub(wr->send_wr_num, &isert_conn->post_send_buf_count);
 		isert_completion_rdma_write(tx_desc, isert_cmd);
 		break;
 	case ISER_IB_RDMA_READ:
-		pr_debug("isert_send_completion: Got ISER_IB_RDMA_READ:\n");
-
-		atomic_sub(wr->send_wr_num, &isert_conn->post_send_buf_count);
 		isert_completion_rdma_read(tx_desc, isert_cmd);
 		break;
 	default:
-		pr_err("Unknown wr->iser_ib_op: 0x%02x\n", wr->iser_ib_op);
+		isert_err("Unknown wr->iser_ib_op: 0x%x\n", wr->iser_ib_op);
 		dump_stack();
 		break;
 	}
 }
 
-static void
-isert_send_completion(struct iser_tx_desc *tx_desc,
-		      struct isert_conn *isert_conn)
-{
-	struct llist_node *llnode = tx_desc->comp_llnode_batch;
-	struct iser_tx_desc *t;
-	/*
-	 * Drain coalesced completion llist starting from comp_llnode_batch
-	 * setup in isert_init_send_wr(), and then complete trailing tx_desc.
-	 */
-	while (llnode) {
-		t = llist_entry(llnode, struct iser_tx_desc, comp_llnode);
-		llnode = llist_next(llnode);
-		__isert_send_completion(t, isert_conn);
-	}
-	__isert_send_completion(tx_desc, isert_conn);
-}
-
-static void
-isert_cq_drain_comp_llist(struct isert_conn *isert_conn, struct ib_device *ib_dev)
+/**
+ * is_isert_tx_desc() - Indicate if the completion wr_id
+ *     is a TX descriptor or not.
+ * @isert_conn: iser connection
+ * @wr_id: completion WR identifier
+ *
+ * Since we cannot rely on wc opcode in FLUSH errors
+ * we must work around it by checking if the wr_id address
+ * falls in the iser connection rx_descs buffer. If so
+ * it is an RX descriptor, otherwize it is a TX.
+ */
+static inline bool
+is_isert_tx_desc(struct isert_conn *isert_conn, void *wr_id)
 {
-	struct llist_node *llnode;
-	struct isert_rdma_wr *wr;
-	struct iser_tx_desc *t;
-
-	mutex_lock(&isert_conn->conn_mutex);
-	llnode = llist_del_all(&isert_conn->conn_comp_llist);
-	isert_conn->conn_comp_batch = 0;
-	mutex_unlock(&isert_conn->conn_mutex);
+	void *start = isert_conn->conn_rx_descs;
+	int len = ISERT_QP_MAX_RECV_DTOS * sizeof(*isert_conn->conn_rx_descs);
 
-	while (llnode) {
-		t = llist_entry(llnode, struct iser_tx_desc, comp_llnode);
-		llnode = llist_next(llnode);
-		wr = &t->isert_cmd->rdma_wr;
+	if (wr_id >= start && wr_id < start + len)
+		return false;
 
-		/**
-		 * If send_wr_num is 0 this means that we got
-		 * RDMA completion and we cleared it and we should
-		 * simply decrement the response post. else the
-		 * response is incorporated in send_wr_num, just
-		 * sub it.
-		 **/
-		if (wr->send_wr_num)
-			atomic_sub(wr->send_wr_num,
-				   &isert_conn->post_send_buf_count);
-		else
-			atomic_dec(&isert_conn->post_send_buf_count);
-
-		isert_completion_put(t, t->isert_cmd, ib_dev, true);
-	}
+	return true;
 }
 
 static void
-isert_cq_tx_comp_err(struct iser_tx_desc *tx_desc, struct isert_conn *isert_conn)
+isert_cq_comp_err(struct isert_conn *isert_conn, struct ib_wc *wc)
 {
-	struct ib_device *ib_dev = isert_conn->conn_cm_id->device;
-	struct isert_cmd *isert_cmd = tx_desc->isert_cmd;
-	struct llist_node *llnode = tx_desc->comp_llnode_batch;
-	struct isert_rdma_wr *wr;
-	struct iser_tx_desc *t;
-
-	while (llnode) {
-		t = llist_entry(llnode, struct iser_tx_desc, comp_llnode);
-		llnode = llist_next(llnode);
-		wr = &t->isert_cmd->rdma_wr;
+	if (wc->wr_id == ISER_BEACON_WRID) {
+		isert_info("conn %p completing conn_wait_comp_err\n",
+			   isert_conn);
+		complete(&isert_conn->conn_wait_comp_err);
+	} else if (is_isert_tx_desc(isert_conn, (void *)(uintptr_t)wc->wr_id)) {
+		struct ib_device *ib_dev = isert_conn->conn_cm_id->device;
+		struct isert_cmd *isert_cmd;
+		struct iser_tx_desc *desc;
 
-		/**
-		 * If send_wr_num is 0 this means that we got
-		 * RDMA completion and we cleared it and we should
-		 * simply decrement the response post. else the
-		 * response is incorporated in send_wr_num, just
-		 * sub it.
-		 **/
-		if (wr->send_wr_num)
-			atomic_sub(wr->send_wr_num,
-				   &isert_conn->post_send_buf_count);
+		desc = (struct iser_tx_desc *)(uintptr_t)wc->wr_id;
+		isert_cmd = desc->isert_cmd;
+		if (!isert_cmd)
+			isert_unmap_tx_desc(desc, ib_dev);
 		else
-			atomic_dec(&isert_conn->post_send_buf_count);
-
-		isert_completion_put(t, t->isert_cmd, ib_dev, true);
-	}
-	tx_desc->comp_llnode_batch = NULL;
-
-	if (!isert_cmd)
-		isert_unmap_tx_desc(tx_desc, ib_dev);
-	else
-		isert_completion_put(tx_desc, isert_cmd, ib_dev, true);
-}
-
-static void
-isert_cq_rx_comp_err(struct isert_conn *isert_conn)
-{
-	struct ib_device *ib_dev = isert_conn->conn_cm_id->device;
-	struct iscsi_conn *conn = isert_conn->conn;
-
-	if (isert_conn->post_recv_buf_count)
-		return;
-
-	isert_cq_drain_comp_llist(isert_conn, ib_dev);
-
-	if (conn->sess) {
-		target_sess_cmd_list_set_waiting(conn->sess->se_sess);
-		target_wait_for_sess_cmds(conn->sess->se_sess);
+			isert_completion_put(desc, isert_cmd, ib_dev, true);
+	} else {
+		isert_conn->post_recv_buf_count--;
+		if (!isert_conn->post_recv_buf_count)
+			iscsit_cause_connection_reinstatement(isert_conn->conn, 0);
 	}
-
-	while (atomic_read(&isert_conn->post_send_buf_count))
-		msleep(3000);
-
-	mutex_lock(&isert_conn->conn_mutex);
-	isert_conn->state = ISER_CONN_DOWN;
-	mutex_unlock(&isert_conn->conn_mutex);
-
-	iscsit_cause_connection_reinstatement(isert_conn->conn, 0);
-
-	complete(&isert_conn->conn_wait_comp_err);
 }
 
 static void
-isert_cq_tx_work(struct work_struct *work)
+isert_handle_wc(struct ib_wc *wc)
 {
-	struct isert_cq_desc *cq_desc = container_of(work,
-				struct isert_cq_desc, cq_tx_work);
-	struct isert_device *device = cq_desc->device;
-	int cq_index = cq_desc->cq_index;
-	struct ib_cq *tx_cq = device->dev_tx_cq[cq_index];
 	struct isert_conn *isert_conn;
 	struct iser_tx_desc *tx_desc;
-	struct ib_wc wc;
-
-	while (ib_poll_cq(tx_cq, 1, &wc) == 1) {
-		tx_desc = (struct iser_tx_desc *)(unsigned long)wc.wr_id;
-		isert_conn = wc.qp->qp_context;
+	struct iser_rx_desc *rx_desc;
 
-		if (wc.status == IB_WC_SUCCESS) {
-			isert_send_completion(tx_desc, isert_conn);
+	isert_conn = wc->qp->qp_context;
+	if (likely(wc->status == IB_WC_SUCCESS)) {
+		if (wc->opcode == IB_WC_RECV) {
+			rx_desc = (struct iser_rx_desc *)(uintptr_t)wc->wr_id;
+			isert_rx_completion(rx_desc, isert_conn, wc->byte_len);
 		} else {
-			pr_debug("TX wc.status != IB_WC_SUCCESS >>>>>>>>>>>>>>\n");
-			pr_debug("TX wc.status: 0x%08x\n", wc.status);
-			pr_debug("TX wc.vendor_err: 0x%08x\n", wc.vendor_err);
-
-			if (wc.wr_id != ISER_FASTREG_LI_WRID) {
-				if (tx_desc->llnode_active)
-					continue;
-
-				atomic_dec(&isert_conn->post_send_buf_count);
-				isert_cq_tx_comp_err(tx_desc, isert_conn);
-			}
+			tx_desc = (struct iser_tx_desc *)(uintptr_t)wc->wr_id;
+			isert_send_completion(tx_desc, isert_conn);
 		}
-	}
-
-	ib_req_notify_cq(tx_cq, IB_CQ_NEXT_COMP);
-}
-
-static void
-isert_cq_tx_callback(struct ib_cq *cq, void *context)
-{
-	struct isert_cq_desc *cq_desc = (struct isert_cq_desc *)context;
+	} else {
+		if (wc->status != IB_WC_WR_FLUSH_ERR)
+			isert_err("wr id %llx status %d vend_err %x\n",
+				  wc->wr_id, wc->status, wc->vendor_err);
+		else
+			isert_dbg("flush error: wr id %llx\n", wc->wr_id);
 
-	queue_work(isert_comp_wq, &cq_desc->cq_tx_work);
+		if (wc->wr_id != ISER_FASTREG_LI_WRID)
+			isert_cq_comp_err(isert_conn, wc);
+	}
 }
 
 static void
-isert_cq_rx_work(struct work_struct *work)
+isert_cq_work(struct work_struct *work)
 {
-	struct isert_cq_desc *cq_desc = container_of(work,
-			struct isert_cq_desc, cq_rx_work);
-	struct isert_device *device = cq_desc->device;
-	int cq_index = cq_desc->cq_index;
-	struct ib_cq *rx_cq = device->dev_rx_cq[cq_index];
-	struct isert_conn *isert_conn;
-	struct iser_rx_desc *rx_desc;
-	struct ib_wc wc;
-	unsigned long xfer_len;
+	enum { isert_poll_budget = 65536 };
+	struct isert_comp *comp = container_of(work, struct isert_comp,
+					       work);
+	struct ib_wc *const wcs = comp->wcs;
+	int i, n, completed = 0;
 
-	while (ib_poll_cq(rx_cq, 1, &wc) == 1) {
-		rx_desc = (struct iser_rx_desc *)(unsigned long)wc.wr_id;
-		isert_conn = wc.qp->qp_context;
+	while ((n = ib_poll_cq(comp->cq, ARRAY_SIZE(comp->wcs), wcs)) > 0) {
+		for (i = 0; i < n; i++)
+			isert_handle_wc(&wcs[i]);
 
-		if (wc.status == IB_WC_SUCCESS) {
-			xfer_len = (unsigned long)wc.byte_len;
-			isert_rx_completion(rx_desc, isert_conn, xfer_len);
-		} else {
-			pr_debug("RX wc.status != IB_WC_SUCCESS >>>>>>>>>>>>>>\n");
-			if (wc.status != IB_WC_WR_FLUSH_ERR) {
-				pr_debug("RX wc.status: 0x%08x\n", wc.status);
-				pr_debug("RX wc.vendor_err: 0x%08x\n",
-					 wc.vendor_err);
-			}
-			isert_conn->post_recv_buf_count--;
-			isert_cq_rx_comp_err(isert_conn);
-		}
+		completed += n;
+		if (completed >= isert_poll_budget)
+			break;
 	}
 
-	ib_req_notify_cq(rx_cq, IB_CQ_NEXT_COMP);
+	ib_req_notify_cq(comp->cq, IB_CQ_NEXT_COMP);
 }
 
 static void
-isert_cq_rx_callback(struct ib_cq *cq, void *context)
+isert_cq_callback(struct ib_cq *cq, void *context)
 {
-	struct isert_cq_desc *cq_desc = (struct isert_cq_desc *)context;
+	struct isert_comp *comp = context;
 
-	queue_work(isert_rx_wq, &cq_desc->cq_rx_work);
+	queue_work(isert_comp_wq, &comp->work);
 }
 
 static int
@@ -2132,13 +2057,10 @@ isert_post_response(struct isert_conn *isert_conn, struct isert_cmd *isert_cmd)
 	struct ib_send_wr *wr_failed;
 	int ret;
 
-	atomic_inc(&isert_conn->post_send_buf_count);
-
 	ret = ib_post_send(isert_conn->conn_qp, &isert_cmd->tx_desc.send_wr,
 			   &wr_failed);
 	if (ret) {
-		pr_err("ib_post_send failed with %d\n", ret);
-		atomic_dec(&isert_conn->post_send_buf_count);
+		isert_err("ib_post_send failed with %d\n", ret);
 		return ret;
 	}
 	return ret;
@@ -2185,9 +2107,9 @@ isert_put_response(struct iscsi_conn *conn, struct iscsi_cmd *cmd)
 		isert_cmd->tx_desc.num_sge = 2;
 	}
 
-	isert_init_send_wr(isert_conn, isert_cmd, send_wr, true);
+	isert_init_send_wr(isert_conn, isert_cmd, send_wr);
 
-	pr_debug("Posting SCSI Response IB_WR_SEND >>>>>>>>>>>>>>>>>>>>>>\n");
+	isert_dbg("Posting SCSI Response\n");
 
 	return isert_post_response(isert_conn, isert_cmd);
 }
@@ -2216,8 +2138,16 @@ isert_get_sup_prot_ops(struct iscsi_conn *conn)
 	struct isert_conn *isert_conn = (struct isert_conn *)conn->context;
 	struct isert_device *device = isert_conn->conn_device;
 
-	if (device->pi_capable)
-		return TARGET_PROT_ALL;
+	if (conn->tpg->tpg_attrib.t10_pi) {
+		if (device->pi_capable) {
+			isert_info("conn %p PI offload enabled\n", isert_conn);
+			isert_conn->pi_support = true;
+			return TARGET_PROT_ALL;
+		}
+	}
+
+	isert_info("conn %p PI offload disabled\n", isert_conn);
+	isert_conn->pi_support = false;
 
 	return TARGET_PROT_NORMAL;
 }
@@ -2235,9 +2165,9 @@ isert_put_nopin(struct iscsi_cmd *cmd, struct iscsi_conn *conn,
 			       &isert_cmd->tx_desc.iscsi_header,
 			       nopout_response);
 	isert_init_tx_hdrs(isert_conn, &isert_cmd->tx_desc);
-	isert_init_send_wr(isert_conn, isert_cmd, send_wr, false);
+	isert_init_send_wr(isert_conn, isert_cmd, send_wr);
 
-	pr_debug("Posting NOPIN Response IB_WR_SEND >>>>>>>>>>>>>>>>>>>>>>\n");
+	isert_dbg("conn %p Posting NOPIN Response\n", isert_conn);
 
 	return isert_post_response(isert_conn, isert_cmd);
 }
@@ -2253,9 +2183,9 @@ isert_put_logout_rsp(struct iscsi_cmd *cmd, struct iscsi_conn *conn)
 	iscsit_build_logout_rsp(cmd, conn, (struct iscsi_logout_rsp *)
 				&isert_cmd->tx_desc.iscsi_header);
 	isert_init_tx_hdrs(isert_conn, &isert_cmd->tx_desc);
-	isert_init_send_wr(isert_conn, isert_cmd, send_wr, false);
+	isert_init_send_wr(isert_conn, isert_cmd, send_wr);
 
-	pr_debug("Posting Logout Response IB_WR_SEND >>>>>>>>>>>>>>>>>>>>>>\n");
+	isert_dbg("conn %p Posting Logout Response\n", isert_conn);
 
 	return isert_post_response(isert_conn, isert_cmd);
 }
@@ -2271,9 +2201,9 @@ isert_put_tm_rsp(struct iscsi_cmd *cmd, struct iscsi_conn *conn)
 	iscsit_build_task_mgt_rsp(cmd, conn, (struct iscsi_tm_rsp *)
 				  &isert_cmd->tx_desc.iscsi_header);
 	isert_init_tx_hdrs(isert_conn, &isert_cmd->tx_desc);
-	isert_init_send_wr(isert_conn, isert_cmd, send_wr, false);
+	isert_init_send_wr(isert_conn, isert_cmd, send_wr);
 
-	pr_debug("Posting Task Management Response IB_WR_SEND >>>>>>>>>>>>>>>>>>>>>>\n");
+	isert_dbg("conn %p Posting Task Management Response\n", isert_conn);
 
 	return isert_post_response(isert_conn, isert_cmd);
 }
@@ -2303,9 +2233,9 @@ isert_put_reject(struct iscsi_cmd *cmd, struct iscsi_conn *conn)
 	tx_dsg->lkey	= isert_conn->conn_mr->lkey;
 	isert_cmd->tx_desc.num_sge = 2;
 
-	isert_init_send_wr(isert_conn, isert_cmd, send_wr, false);
+	isert_init_send_wr(isert_conn, isert_cmd, send_wr);
 
-	pr_debug("Posting Reject IB_WR_SEND >>>>>>>>>>>>>>>>>>>>>>\n");
+	isert_dbg("conn %p Posting Reject\n", isert_conn);
 
 	return isert_post_response(isert_conn, isert_cmd);
 }
@@ -2343,9 +2273,9 @@ isert_put_text_rsp(struct iscsi_cmd *cmd, struct iscsi_conn *conn)
 		tx_dsg->lkey	= isert_conn->conn_mr->lkey;
 		isert_cmd->tx_desc.num_sge = 2;
 	}
-	isert_init_send_wr(isert_conn, isert_cmd, send_wr, false);
+	isert_init_send_wr(isert_conn, isert_cmd, send_wr);
 
-	pr_debug("Posting Text Response IB_WR_SEND >>>>>>>>>>>>>>>>>>>>>>\n");
+	isert_dbg("conn %p Text Reject\n", isert_conn);
 
 	return isert_post_response(isert_conn, isert_cmd);
 }
@@ -2368,30 +2298,31 @@ isert_build_rdma_wr(struct isert_conn *isert_conn, struct isert_cmd *isert_cmd,
 
 	send_wr->sg_list = ib_sge;
 	send_wr->num_sge = sg_nents;
-	send_wr->wr_id = (unsigned long)&isert_cmd->tx_desc;
+	send_wr->wr_id = (uintptr_t)&isert_cmd->tx_desc;
 	/*
 	 * Perform mapping of TCM scatterlist memory ib_sge dma_addr.
 	 */
 	for_each_sg(sg_start, tmp_sg, sg_nents, i) {
-		pr_debug("ISER RDMA from SGL dma_addr: 0x%16llx dma_len: %u, page_off: %u\n",
-			 (unsigned long long)tmp_sg->dma_address,
-			 tmp_sg->length, page_off);
+		isert_dbg("RDMA from SGL dma_addr: 0x%llx dma_len: %u, "
+			  "page_off: %u\n",
+			  (unsigned long long)tmp_sg->dma_address,
+			  tmp_sg->length, page_off);
 
 		ib_sge->addr = ib_sg_dma_address(ib_dev, tmp_sg) + page_off;
 		ib_sge->length = min_t(u32, data_left,
 				ib_sg_dma_len(ib_dev, tmp_sg) - page_off);
 		ib_sge->lkey = isert_conn->conn_mr->lkey;
 
-		pr_debug("RDMA ib_sge: addr: 0x%16llx  length: %u lkey: %08x\n",
-			 ib_sge->addr, ib_sge->length, ib_sge->lkey);
+		isert_dbg("RDMA ib_sge: addr: 0x%llx  length: %u lkey: %x\n",
+			  ib_sge->addr, ib_sge->length, ib_sge->lkey);
 		page_off = 0;
 		data_left -= ib_sge->length;
 		ib_sge++;
-		pr_debug("Incrementing ib_sge pointer to %p\n", ib_sge);
+		isert_dbg("Incrementing ib_sge pointer to %p\n", ib_sge);
 	}
 
-	pr_debug("Set outgoing sg_list: %p num_sg: %u from TCM SGLs\n",
-		 send_wr->sg_list, send_wr->num_sge);
+	isert_dbg("Set outgoing sg_list: %p num_sg: %u from TCM SGLs\n",
+		  send_wr->sg_list, send_wr->num_sge);
 
 	return sg_nents;
 }
@@ -2423,7 +2354,7 @@ isert_map_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd,
 
 	ib_sge = kzalloc(sizeof(struct ib_sge) * data->nents, GFP_KERNEL);
 	if (!ib_sge) {
-		pr_warn("Unable to allocate ib_sge\n");
+		isert_warn("Unable to allocate ib_sge\n");
 		ret = -ENOMEM;
 		goto unmap_cmd;
 	}
@@ -2433,7 +2364,7 @@ isert_map_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd,
 	wr->send_wr = kzalloc(sizeof(struct ib_send_wr) * wr->send_wr_num,
 				GFP_KERNEL);
 	if (!wr->send_wr) {
-		pr_debug("Unable to allocate wr->send_wr\n");
+		isert_dbg("Unable to allocate wr->send_wr\n");
 		ret = -ENOMEM;
 		goto unmap_cmd;
 	}
@@ -2497,9 +2428,9 @@ isert_map_fr_pagelist(struct ib_device *ib_dev,
 			chunk_start = start_addr;
 		end_addr = start_addr + ib_sg_dma_len(ib_dev, tmp_sg);
 
-		pr_debug("SGL[%d] dma_addr: 0x%16llx len: %u\n",
-			 i, (unsigned long long)tmp_sg->dma_address,
-			 tmp_sg->length);
+		isert_dbg("SGL[%d] dma_addr: 0x%llx len: %u\n",
+			  i, (unsigned long long)tmp_sg->dma_address,
+			  tmp_sg->length);
 
 		if ((end_addr & ~PAGE_MASK) && i < last_ent) {
 			new_chunk = 0;
@@ -2510,8 +2441,8 @@ isert_map_fr_pagelist(struct ib_device *ib_dev,
 		page = chunk_start & PAGE_MASK;
 		do {
 			fr_pl[n_pages++] = page;
-			pr_debug("Mapped page_list[%d] page_addr: 0x%16llx\n",
-				 n_pages - 1, page);
+			isert_dbg("Mapped page_list[%d] page_addr: 0x%llx\n",
+				  n_pages - 1, page);
 			page += PAGE_SIZE;
 		} while (page < end_addr);
 	}
@@ -2519,6 +2450,21 @@ isert_map_fr_pagelist(struct ib_device *ib_dev,
 	return n_pages;
 }
 
+static inline void
+isert_inv_rkey(struct ib_send_wr *inv_wr, struct ib_mr *mr)
+{
+	u32 rkey;
+
+	memset(inv_wr, 0, sizeof(*inv_wr));
+	inv_wr->wr_id = ISER_FASTREG_LI_WRID;
+	inv_wr->opcode = IB_WR_LOCAL_INV;
+	inv_wr->ex.invalidate_rkey = mr->rkey;
+
+	/* Bump the key */
+	rkey = ib_inc_rkey(mr->rkey);
+	ib_update_fast_reg_key(mr, rkey);
+}
+
 static int
 isert_fast_reg_mr(struct isert_conn *isert_conn,
 		  struct fast_reg_descriptor *fr_desc,
@@ -2533,15 +2479,13 @@ isert_fast_reg_mr(struct isert_conn *isert_conn,
 	struct ib_send_wr *bad_wr, *wr = NULL;
 	int ret, pagelist_len;
 	u32 page_off;
-	u8 key;
 
 	if (mem->dma_nents == 1) {
 		sge->lkey = isert_conn->conn_mr->lkey;
 		sge->addr = ib_sg_dma_address(ib_dev, &mem->sg[0]);
 		sge->length = ib_sg_dma_len(ib_dev, &mem->sg[0]);
-		pr_debug("%s:%d sge: addr: 0x%llx  length: %u lkey: %x\n",
-			 __func__, __LINE__, sge->addr, sge->length,
-			 sge->lkey);
+		isert_dbg("sge: addr: 0x%llx  length: %u lkey: %x\n",
+			 sge->addr, sge->length, sge->lkey);
 		return 0;
 	}
 
@@ -2557,21 +2501,15 @@ isert_fast_reg_mr(struct isert_conn *isert_conn,
 
 	page_off = mem->offset % PAGE_SIZE;
 
-	pr_debug("Use fr_desc %p sg_nents %d offset %u\n",
-		 fr_desc, mem->nents, mem->offset);
+	isert_dbg("Use fr_desc %p sg_nents %d offset %u\n",
+		  fr_desc, mem->nents, mem->offset);
 
 	pagelist_len = isert_map_fr_pagelist(ib_dev, mem->sg, mem->nents,
 					     &frpl->page_list[0]);
 
-	if (!(fr_desc->ind & ISERT_DATA_KEY_VALID)) {
-		memset(&inv_wr, 0, sizeof(inv_wr));
-		inv_wr.wr_id = ISER_FASTREG_LI_WRID;
-		inv_wr.opcode = IB_WR_LOCAL_INV;
-		inv_wr.ex.invalidate_rkey = mr->rkey;
+	if (!(fr_desc->ind & ind)) {
+		isert_inv_rkey(&inv_wr, mr);
 		wr = &inv_wr;
-		/* Bump the key */
-		key = (u8)(mr->rkey & 0x000000FF);
-		ib_update_fast_reg_key(mr, ++key);
 	}
 
 	/* Prepare FASTREG WR */
@@ -2593,7 +2531,7 @@ isert_fast_reg_mr(struct isert_conn *isert_conn,
 
 	ret = ib_post_send(isert_conn->conn_qp, wr, &bad_wr);
 	if (ret) {
-		pr_err("fast registration failed, ret:%d\n", ret);
+		isert_err("fast registration failed, ret:%d\n", ret);
 		return ret;
 	}
 	fr_desc->ind &= ~ind;
@@ -2602,68 +2540,54 @@ isert_fast_reg_mr(struct isert_conn *isert_conn,
 	sge->addr = frpl->page_list[0] + page_off;
 	sge->length = mem->len;
 
-	pr_debug("%s:%d sge: addr: 0x%llx  length: %u lkey: %x\n",
-		 __func__, __LINE__, sge->addr, sge->length,
-		 sge->lkey);
+	isert_dbg("sge: addr: 0x%llx  length: %u lkey: %x\n",
+		  sge->addr, sge->length, sge->lkey);
 
 	return ret;
 }
 
-static inline enum ib_t10_dif_type
-se2ib_prot_type(enum target_prot_type prot_type)
+static inline void
+isert_set_dif_domain(struct se_cmd *se_cmd, struct ib_sig_attrs *sig_attrs,
+		     struct ib_sig_domain *domain)
 {
-	switch (prot_type) {
-	case TARGET_DIF_TYPE0_PROT:
-		return IB_T10DIF_NONE;
-	case TARGET_DIF_TYPE1_PROT:
-		return IB_T10DIF_TYPE1;
-	case TARGET_DIF_TYPE2_PROT:
-		return IB_T10DIF_TYPE2;
-	case TARGET_DIF_TYPE3_PROT:
-		return IB_T10DIF_TYPE3;
-	default:
-		return IB_T10DIF_NONE;
-	}
-}
+	domain->sig_type = IB_SIG_TYPE_T10_DIF;
+	domain->sig.dif.bg_type = IB_T10DIF_CRC;
+	domain->sig.dif.pi_interval = se_cmd->se_dev->dev_attrib.block_size;
+	domain->sig.dif.ref_tag = se_cmd->reftag_seed;
+	/*
+	 * At the moment we hard code those, but if in the future
+	 * the target core would like to use it, we will take it
+	 * from se_cmd.
+	 */
+	domain->sig.dif.apptag_check_mask = 0xffff;
+	domain->sig.dif.app_escape = true;
+	domain->sig.dif.ref_escape = true;
+	if (se_cmd->prot_type == TARGET_DIF_TYPE1_PROT ||
+	    se_cmd->prot_type == TARGET_DIF_TYPE2_PROT)
+		domain->sig.dif.ref_remap = true;
+};
 
 static int
 isert_set_sig_attrs(struct se_cmd *se_cmd, struct ib_sig_attrs *sig_attrs)
 {
-	enum ib_t10_dif_type ib_prot_type = se2ib_prot_type(se_cmd->prot_type);
-
-	sig_attrs->mem.sig_type = IB_SIG_TYPE_T10_DIF;
-	sig_attrs->wire.sig_type = IB_SIG_TYPE_T10_DIF;
-	sig_attrs->mem.sig.dif.pi_interval =
-				se_cmd->se_dev->dev_attrib.block_size;
-	sig_attrs->wire.sig.dif.pi_interval =
-				se_cmd->se_dev->dev_attrib.block_size;
-
 	switch (se_cmd->prot_op) {
 	case TARGET_PROT_DIN_INSERT:
 	case TARGET_PROT_DOUT_STRIP:
-		sig_attrs->mem.sig.dif.type = IB_T10DIF_NONE;
-		sig_attrs->wire.sig.dif.type = ib_prot_type;
-		sig_attrs->wire.sig.dif.bg_type = IB_T10DIF_CRC;
-		sig_attrs->wire.sig.dif.ref_tag = se_cmd->reftag_seed;
+		sig_attrs->mem.sig_type = IB_SIG_TYPE_NONE;
+		isert_set_dif_domain(se_cmd, sig_attrs, &sig_attrs->wire);
 		break;
 	case TARGET_PROT_DOUT_INSERT:
 	case TARGET_PROT_DIN_STRIP:
-		sig_attrs->mem.sig.dif.type = ib_prot_type;
-		sig_attrs->mem.sig.dif.bg_type = IB_T10DIF_CRC;
-		sig_attrs->mem.sig.dif.ref_tag = se_cmd->reftag_seed;
-		sig_attrs->wire.sig.dif.type = IB_T10DIF_NONE;
+		sig_attrs->wire.sig_type = IB_SIG_TYPE_NONE;
+		isert_set_dif_domain(se_cmd, sig_attrs, &sig_attrs->mem);
 		break;
 	case TARGET_PROT_DIN_PASS:
 	case TARGET_PROT_DOUT_PASS:
-		sig_attrs->mem.sig.dif.type = ib_prot_type;
-		sig_attrs->mem.sig.dif.bg_type = IB_T10DIF_CRC;
-		sig_attrs->mem.sig.dif.ref_tag = se_cmd->reftag_seed;
-		sig_attrs->wire.sig.dif.type = ib_prot_type;
-		sig_attrs->wire.sig.dif.bg_type = IB_T10DIF_CRC;
-		sig_attrs->wire.sig.dif.ref_tag = se_cmd->reftag_seed;
+		isert_set_dif_domain(se_cmd, sig_attrs, &sig_attrs->wire);
+		isert_set_dif_domain(se_cmd, sig_attrs, &sig_attrs->mem);
 		break;
 	default:
-		pr_err("Unsupported PI operation %d\n", se_cmd->prot_op);
+		isert_err("Unsupported PI operation %d\n", se_cmd->prot_op);
 		return -EINVAL;
 	}
 
@@ -2679,17 +2603,16 @@ isert_set_prot_checks(u8 prot_checks)
 }
 
 static int
-isert_reg_sig_mr(struct isert_conn *isert_conn, struct se_cmd *se_cmd,
-		 struct fast_reg_descriptor *fr_desc,
-		 struct ib_sge *data_sge, struct ib_sge *prot_sge,
-		 struct ib_sge *sig_sge)
+isert_reg_sig_mr(struct isert_conn *isert_conn,
+		 struct se_cmd *se_cmd,
+		 struct isert_rdma_wr *rdma_wr,
+		 struct fast_reg_descriptor *fr_desc)
 {
 	struct ib_send_wr sig_wr, inv_wr;
 	struct ib_send_wr *bad_wr, *wr = NULL;
 	struct pi_context *pi_ctx = fr_desc->pi_ctx;
 	struct ib_sig_attrs sig_attrs;
 	int ret;
-	u32 key;
 
 	memset(&sig_attrs, 0, sizeof(sig_attrs));
 	ret = isert_set_sig_attrs(se_cmd, &sig_attrs);
@@ -2699,26 +2622,20 @@ isert_reg_sig_mr(struct isert_conn *isert_conn, struct se_cmd *se_cmd,
 	sig_attrs.check_mask = isert_set_prot_checks(se_cmd->prot_checks);
 
 	if (!(fr_desc->ind & ISERT_SIG_KEY_VALID)) {
-		memset(&inv_wr, 0, sizeof(inv_wr));
-		inv_wr.opcode = IB_WR_LOCAL_INV;
-		inv_wr.wr_id = ISER_FASTREG_LI_WRID;
-		inv_wr.ex.invalidate_rkey = pi_ctx->sig_mr->rkey;
+		isert_inv_rkey(&inv_wr, pi_ctx->sig_mr);
 		wr = &inv_wr;
-		/* Bump the key */
-		key = (u8)(pi_ctx->sig_mr->rkey & 0x000000FF);
-		ib_update_fast_reg_key(pi_ctx->sig_mr, ++key);
 	}
 
 	memset(&sig_wr, 0, sizeof(sig_wr));
 	sig_wr.opcode = IB_WR_REG_SIG_MR;
 	sig_wr.wr_id = ISER_FASTREG_LI_WRID;
-	sig_wr.sg_list = data_sge;
+	sig_wr.sg_list = &rdma_wr->ib_sg[DATA];
 	sig_wr.num_sge = 1;
 	sig_wr.wr.sig_handover.access_flags = IB_ACCESS_LOCAL_WRITE;
 	sig_wr.wr.sig_handover.sig_attrs = &sig_attrs;
 	sig_wr.wr.sig_handover.sig_mr = pi_ctx->sig_mr;
 	if (se_cmd->t_prot_sg)
-		sig_wr.wr.sig_handover.prot = prot_sge;
+		sig_wr.wr.sig_handover.prot = &rdma_wr->ib_sg[PROT];
 
 	if (!wr)
 		wr = &sig_wr;
@@ -2727,39 +2644,98 @@ isert_reg_sig_mr(struct isert_conn *isert_conn, struct se_cmd *se_cmd,
 
 	ret = ib_post_send(isert_conn->conn_qp, wr, &bad_wr);
 	if (ret) {
-		pr_err("fast registration failed, ret:%d\n", ret);
+		isert_err("fast registration failed, ret:%d\n", ret);
 		goto err;
 	}
 	fr_desc->ind &= ~ISERT_SIG_KEY_VALID;
 
-	sig_sge->lkey = pi_ctx->sig_mr->lkey;
-	sig_sge->addr = 0;
-	sig_sge->length = se_cmd->data_length;
+	rdma_wr->ib_sg[SIG].lkey = pi_ctx->sig_mr->lkey;
+	rdma_wr->ib_sg[SIG].addr = 0;
+	rdma_wr->ib_sg[SIG].length = se_cmd->data_length;
 	if (se_cmd->prot_op != TARGET_PROT_DIN_STRIP &&
 	    se_cmd->prot_op != TARGET_PROT_DOUT_INSERT)
 		/*
 		 * We have protection guards on the wire
 		 * so we need to set a larget transfer
 		 */
-		sig_sge->length += se_cmd->prot_length;
+		rdma_wr->ib_sg[SIG].length += se_cmd->prot_length;
 
-	pr_debug("sig_sge: addr: 0x%llx  length: %u lkey: %x\n",
-		 sig_sge->addr, sig_sge->length,
-		 sig_sge->lkey);
+	isert_dbg("sig_sge: addr: 0x%llx  length: %u lkey: %x\n",
+		  rdma_wr->ib_sg[SIG].addr, rdma_wr->ib_sg[SIG].length,
+		  rdma_wr->ib_sg[SIG].lkey);
 err:
 	return ret;
 }
 
 static int
+isert_handle_prot_cmd(struct isert_conn *isert_conn,
+		      struct isert_cmd *isert_cmd,
+		      struct isert_rdma_wr *wr)
+{
+	struct isert_device *device = isert_conn->conn_device;
+	struct se_cmd *se_cmd = &isert_cmd->iscsi_cmd->se_cmd;
+	int ret;
+
+	if (!wr->fr_desc->pi_ctx) {
+		ret = isert_create_pi_ctx(wr->fr_desc,
+					  device->ib_device,
+					  isert_conn->conn_pd);
+		if (ret) {
+			isert_err("conn %p failed to allocate pi_ctx\n",
+				  isert_conn);
+			return ret;
+		}
+	}
+
+	if (se_cmd->t_prot_sg) {
+		ret = isert_map_data_buf(isert_conn, isert_cmd,
+					 se_cmd->t_prot_sg,
+					 se_cmd->t_prot_nents,
+					 se_cmd->prot_length,
+					 0, wr->iser_ib_op, &wr->prot);
+		if (ret) {
+			isert_err("conn %p failed to map protection buffer\n",
+				  isert_conn);
+			return ret;
+		}
+
+		memset(&wr->ib_sg[PROT], 0, sizeof(wr->ib_sg[PROT]));
+		ret = isert_fast_reg_mr(isert_conn, wr->fr_desc, &wr->prot,
+					ISERT_PROT_KEY_VALID, &wr->ib_sg[PROT]);
+		if (ret) {
+			isert_err("conn %p failed to fast reg mr\n",
+				  isert_conn);
+			goto unmap_prot_cmd;
+		}
+	}
+
+	ret = isert_reg_sig_mr(isert_conn, se_cmd, wr, wr->fr_desc);
+	if (ret) {
+		isert_err("conn %p failed to fast reg mr\n",
+			  isert_conn);
+		goto unmap_prot_cmd;
+	}
+	wr->fr_desc->ind |= ISERT_PROTECTED;
+
+	return 0;
+
+unmap_prot_cmd:
+	if (se_cmd->t_prot_sg)
+		isert_unmap_data_buf(isert_conn, &wr->prot);
+
+	return ret;
+}
+
+static int
 isert_reg_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd,
 	       struct isert_rdma_wr *wr)
 {
 	struct se_cmd *se_cmd = &cmd->se_cmd;
 	struct isert_cmd *isert_cmd = iscsit_priv_cmd(cmd);
 	struct isert_conn *isert_conn = conn->context;
-	struct ib_sge data_sge;
-	struct ib_send_wr *send_wr;
 	struct fast_reg_descriptor *fr_desc = NULL;
+	struct ib_send_wr *send_wr;
+	struct ib_sge *ib_sg;
 	u32 offset;
 	int ret = 0;
 	unsigned long flags;
@@ -2773,8 +2749,7 @@ isert_reg_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd,
 	if (ret)
 		return ret;
 
-	if (wr->data.dma_nents != 1 ||
-	    se_cmd->prot_op != TARGET_PROT_NORMAL) {
+	if (wr->data.dma_nents != 1 || isert_prot_cmd(isert_conn, se_cmd)) {
 		spin_lock_irqsave(&isert_conn->conn_lock, flags);
 		fr_desc = list_first_entry(&isert_conn->conn_fr_pool,
 					   struct fast_reg_descriptor, list);
@@ -2784,38 +2759,21 @@ isert_reg_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd,
 	}
 
 	ret = isert_fast_reg_mr(isert_conn, fr_desc, &wr->data,
-				ISERT_DATA_KEY_VALID, &data_sge);
+				ISERT_DATA_KEY_VALID, &wr->ib_sg[DATA]);
 	if (ret)
 		goto unmap_cmd;
 
-	if (se_cmd->prot_op != TARGET_PROT_NORMAL) {
-		struct ib_sge prot_sge, sig_sge;
-
-		if (se_cmd->t_prot_sg) {
-			ret = isert_map_data_buf(isert_conn, isert_cmd,
-						 se_cmd->t_prot_sg,
-						 se_cmd->t_prot_nents,
-						 se_cmd->prot_length,
-						 0, wr->iser_ib_op, &wr->prot);
-			if (ret)
-				goto unmap_cmd;
-
-			ret = isert_fast_reg_mr(isert_conn, fr_desc, &wr->prot,
-						ISERT_PROT_KEY_VALID, &prot_sge);
-			if (ret)
-				goto unmap_prot_cmd;
-		}
-
-		ret = isert_reg_sig_mr(isert_conn, se_cmd, fr_desc,
-				       &data_sge, &prot_sge, &sig_sge);
+	if (isert_prot_cmd(isert_conn, se_cmd)) {
+		ret = isert_handle_prot_cmd(isert_conn, isert_cmd, wr);
 		if (ret)
-			goto unmap_prot_cmd;
+			goto unmap_cmd;
 
-		fr_desc->ind |= ISERT_PROTECTED;
-		memcpy(&wr->s_ib_sge, &sig_sge, sizeof(sig_sge));
-	} else
-		memcpy(&wr->s_ib_sge, &data_sge, sizeof(data_sge));
+		ib_sg = &wr->ib_sg[SIG];
+	} else {
+		ib_sg = &wr->ib_sg[DATA];
+	}
 
+	memcpy(&wr->s_ib_sge, ib_sg, sizeof(*ib_sg));
 	wr->ib_sge = &wr->s_ib_sge;
 	wr->send_wr_num = 1;
 	memset(&wr->s_send_wr, 0, sizeof(*send_wr));
@@ -2825,12 +2783,12 @@ isert_reg_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd,
 	send_wr = &isert_cmd->rdma_wr.s_send_wr;
 	send_wr->sg_list = &wr->s_ib_sge;
 	send_wr->num_sge = 1;
-	send_wr->wr_id = (unsigned long)&isert_cmd->tx_desc;
+	send_wr->wr_id = (uintptr_t)&isert_cmd->tx_desc;
 	if (wr->iser_ib_op == ISER_IB_RDMA_WRITE) {
 		send_wr->opcode = IB_WR_RDMA_WRITE;
 		send_wr->wr.rdma.remote_addr = isert_cmd->read_va;
 		send_wr->wr.rdma.rkey = isert_cmd->read_stag;
-		send_wr->send_flags = se_cmd->prot_op == TARGET_PROT_NORMAL ?
+		send_wr->send_flags = !isert_prot_cmd(isert_conn, se_cmd) ?
 				      0 : IB_SEND_SIGNALED;
 	} else {
 		send_wr->opcode = IB_WR_RDMA_READ;
@@ -2840,9 +2798,7 @@ isert_reg_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd,
 	}
 
 	return 0;
-unmap_prot_cmd:
-	if (se_cmd->t_prot_sg)
-		isert_unmap_data_buf(isert_conn, &wr->prot);
+
 unmap_cmd:
 	if (fr_desc) {
 		spin_lock_irqsave(&isert_conn->conn_lock, flags);
@@ -2865,16 +2821,17 @@ isert_put_datain(struct iscsi_conn *conn, struct iscsi_cmd *cmd)
 	struct ib_send_wr *wr_failed;
 	int rc;
 
-	pr_debug("Cmd: %p RDMA_WRITE data_length: %u\n",
+	isert_dbg("Cmd: %p RDMA_WRITE data_length: %u\n",
 		 isert_cmd, se_cmd->data_length);
+
 	wr->iser_ib_op = ISER_IB_RDMA_WRITE;
 	rc = device->reg_rdma_mem(conn, cmd, wr);
 	if (rc) {
-		pr_err("Cmd: %p failed to prepare RDMA res\n", isert_cmd);
+		isert_err("Cmd: %p failed to prepare RDMA res\n", isert_cmd);
 		return rc;
 	}
 
-	if (se_cmd->prot_op == TARGET_PROT_NORMAL) {
+	if (!isert_prot_cmd(isert_conn, se_cmd)) {
 		/*
 		 * Build isert_conn->tx_desc for iSCSI response PDU and attach
 		 */
@@ -2884,24 +2841,20 @@ isert_put_datain(struct iscsi_conn *conn, struct iscsi_cmd *cmd)
 				     &isert_cmd->tx_desc.iscsi_header);
 		isert_init_tx_hdrs(isert_conn, &isert_cmd->tx_desc);
 		isert_init_send_wr(isert_conn, isert_cmd,
-				   &isert_cmd->tx_desc.send_wr, true);
+				   &isert_cmd->tx_desc.send_wr);
 		isert_cmd->rdma_wr.s_send_wr.next = &isert_cmd->tx_desc.send_wr;
 		wr->send_wr_num += 1;
 	}
 
-	atomic_add(wr->send_wr_num, &isert_conn->post_send_buf_count);
-
 	rc = ib_post_send(isert_conn->conn_qp, wr->send_wr, &wr_failed);
-	if (rc) {
-		pr_warn("ib_post_send() failed for IB_WR_RDMA_WRITE\n");
-		atomic_sub(wr->send_wr_num, &isert_conn->post_send_buf_count);
-	}
+	if (rc)
+		isert_warn("ib_post_send() failed for IB_WR_RDMA_WRITE\n");
 
-	if (se_cmd->prot_op == TARGET_PROT_NORMAL)
-		pr_debug("Cmd: %p posted RDMA_WRITE + Response for iSER Data "
+	if (!isert_prot_cmd(isert_conn, se_cmd))
+		isert_dbg("Cmd: %p posted RDMA_WRITE + Response for iSER Data "
 			 "READ\n", isert_cmd);
 	else
-		pr_debug("Cmd: %p posted RDMA_WRITE for iSER Data READ\n",
+		isert_dbg("Cmd: %p posted RDMA_WRITE for iSER Data READ\n",
 			 isert_cmd);
 
 	return 1;
@@ -2918,23 +2871,20 @@ isert_get_dataout(struct iscsi_conn *conn, struct iscsi_cmd *cmd, bool recovery)
 	struct ib_send_wr *wr_failed;
 	int rc;
 
-	pr_debug("Cmd: %p RDMA_READ data_length: %u write_data_done: %u\n",
+	isert_dbg("Cmd: %p RDMA_READ data_length: %u write_data_done: %u\n",
 		 isert_cmd, se_cmd->data_length, cmd->write_data_done);
 	wr->iser_ib_op = ISER_IB_RDMA_READ;
 	rc = device->reg_rdma_mem(conn, cmd, wr);
 	if (rc) {
-		pr_err("Cmd: %p failed to prepare RDMA res\n", isert_cmd);
+		isert_err("Cmd: %p failed to prepare RDMA res\n", isert_cmd);
 		return rc;
 	}
 
-	atomic_add(wr->send_wr_num, &isert_conn->post_send_buf_count);
-
 	rc = ib_post_send(isert_conn->conn_qp, wr->send_wr, &wr_failed);
-	if (rc) {
-		pr_warn("ib_post_send() failed for IB_WR_RDMA_READ\n");
-		atomic_sub(wr->send_wr_num, &isert_conn->post_send_buf_count);
-	}
-	pr_debug("Cmd: %p posted RDMA_READ memory for ISER Data WRITE\n",
+	if (rc)
+		isert_warn("ib_post_send() failed for IB_WR_RDMA_READ\n");
+
+	isert_dbg("Cmd: %p posted RDMA_READ memory for ISER Data WRITE\n",
 		 isert_cmd);
 
 	return 0;
@@ -2950,7 +2900,7 @@ isert_immediate_queue(struct iscsi_conn *conn, struct iscsi_cmd *cmd, int state)
 		ret = isert_put_nopin(cmd, conn, false);
 		break;
 	default:
-		pr_err("Unknown immediate state: 0x%02x\n", state);
+		isert_err("Unknown immediate state: 0x%02x\n", state);
 		ret = -EINVAL;
 		break;
 	}
@@ -2961,15 +2911,14 @@ isert_immediate_queue(struct iscsi_conn *conn, struct iscsi_cmd *cmd, int state)
 static int
 isert_response_queue(struct iscsi_conn *conn, struct iscsi_cmd *cmd, int state)
 {
+	struct isert_conn *isert_conn = conn->context;
 	int ret;
 
 	switch (state) {
 	case ISTATE_SEND_LOGOUTRSP:
 		ret = isert_put_logout_rsp(cmd, conn);
-		if (!ret) {
-			pr_debug("Returning iSER Logout -EAGAIN\n");
-			ret = -EAGAIN;
-		}
+		if (!ret)
+			isert_conn->logout_posted = true;
 		break;
 	case ISTATE_SEND_NOPIN:
 		ret = isert_put_nopin(cmd, conn, true);
@@ -2991,7 +2940,7 @@ isert_response_queue(struct iscsi_conn *conn, struct iscsi_cmd *cmd, int state)
 		ret = isert_put_response(conn, cmd);
 		break;
 	default:
-		pr_err("Unknown response state: 0x%02x\n", state);
+		isert_err("Unknown response state: 0x%02x\n", state);
 		ret = -EINVAL;
 		break;
 	}
@@ -2999,27 +2948,64 @@ isert_response_queue(struct iscsi_conn *conn, struct iscsi_cmd *cmd, int state)
 	return ret;
 }
 
+struct rdma_cm_id *
+isert_setup_id(struct isert_np *isert_np)
+{
+	struct iscsi_np *np = isert_np->np;
+	struct rdma_cm_id *id;
+	struct sockaddr *sa;
+	int ret;
+
+	sa = (struct sockaddr *)&np->np_sockaddr;
+	isert_dbg("ksockaddr: %p, sa: %p\n", &np->np_sockaddr, sa);
+
+	id = rdma_create_id(isert_cma_handler, isert_np,
+			    RDMA_PS_TCP, IB_QPT_RC);
+	if (IS_ERR(id)) {
+		isert_err("rdma_create_id() failed: %ld\n", PTR_ERR(id));
+		ret = PTR_ERR(id);
+		goto out;
+	}
+	isert_dbg("id %p context %p\n", id, id->context);
+
+	ret = rdma_bind_addr(id, sa);
+	if (ret) {
+		isert_err("rdma_bind_addr() failed: %d\n", ret);
+		goto out_id;
+	}
+
+	ret = rdma_listen(id, ISERT_RDMA_LISTEN_BACKLOG);
+	if (ret) {
+		isert_err("rdma_listen() failed: %d\n", ret);
+		goto out_id;
+	}
+
+	return id;
+out_id:
+	rdma_destroy_id(id);
+out:
+	return ERR_PTR(ret);
+}
+
 static int
 isert_setup_np(struct iscsi_np *np,
 	       struct __kernel_sockaddr_storage *ksockaddr)
 {
 	struct isert_np *isert_np;
 	struct rdma_cm_id *isert_lid;
-	struct sockaddr *sa;
 	int ret;
 
 	isert_np = kzalloc(sizeof(struct isert_np), GFP_KERNEL);
 	if (!isert_np) {
-		pr_err("Unable to allocate struct isert_np\n");
+		isert_err("Unable to allocate struct isert_np\n");
 		return -ENOMEM;
 	}
 	sema_init(&isert_np->np_sem, 0);
 	mutex_init(&isert_np->np_accept_mutex);
 	INIT_LIST_HEAD(&isert_np->np_accept_list);
 	init_completion(&isert_np->np_login_comp);
+	isert_np->np = np;
 
-	sa = (struct sockaddr *)ksockaddr;
-	pr_debug("ksockaddr: %p, sa: %p\n", ksockaddr, sa);
 	/*
 	 * Setup the np->np_sockaddr from the passed sockaddr setup
 	 * in iscsi_target_configfs.c code..
@@ -3027,37 +3013,20 @@ isert_setup_np(struct iscsi_np *np,
 	memcpy(&np->np_sockaddr, ksockaddr,
 	       sizeof(struct __kernel_sockaddr_storage));
 
-	isert_lid = rdma_create_id(isert_cma_handler, np, RDMA_PS_TCP,
-				IB_QPT_RC);
+	isert_lid = isert_setup_id(isert_np);
 	if (IS_ERR(isert_lid)) {
-		pr_err("rdma_create_id() for isert_listen_handler failed: %ld\n",
-		       PTR_ERR(isert_lid));
 		ret = PTR_ERR(isert_lid);
 		goto out;
 	}
 
-	ret = rdma_bind_addr(isert_lid, sa);
-	if (ret) {
-		pr_err("rdma_bind_addr() for isert_lid failed: %d\n", ret);
-		goto out_lid;
-	}
-
-	ret = rdma_listen(isert_lid, ISERT_RDMA_LISTEN_BACKLOG);
-	if (ret) {
-		pr_err("rdma_listen() for isert_lid failed: %d\n", ret);
-		goto out_lid;
-	}
-
 	isert_np->np_cm_id = isert_lid;
 	np->np_context = isert_np;
-	pr_debug("Setup isert_lid->context: %p\n", isert_lid->context);
 
 	return 0;
 
-out_lid:
-	rdma_destroy_id(isert_lid);
 out:
 	kfree(isert_np);
+
 	return ret;
 }
 
@@ -3073,16 +3042,12 @@ isert_rdma_accept(struct isert_conn *isert_conn)
 	cp.retry_count = 7;
 	cp.rnr_retry_count = 7;
 
-	pr_debug("Before rdma_accept >>>>>>>>>>>>>>>>>>>>.\n");
-
 	ret = rdma_accept(cm_id, &cp);
 	if (ret) {
-		pr_err("rdma_accept() failed with: %d\n", ret);
+		isert_err("rdma_accept() failed with: %d\n", ret);
 		return ret;
 	}
 
-	pr_debug("After rdma_accept >>>>>>>>>>>>>>>>>>>>>.\n");
-
 	return 0;
 }
 
@@ -3092,7 +3057,15 @@ isert_get_login_rx(struct iscsi_conn *conn, struct iscsi_login *login)
 	struct isert_conn *isert_conn = (struct isert_conn *)conn->context;
 	int ret;
 
-	pr_debug("isert_get_login_rx before conn_login_comp conn: %p\n", conn);
+	isert_info("before login_req comp conn: %p\n", isert_conn);
+	ret = wait_for_completion_interruptible(&isert_conn->login_req_comp);
+	if (ret) {
+		isert_err("isert_conn %p interrupted before got login req\n",
+			  isert_conn);
+		return ret;
+	}
+	reinit_completion(&isert_conn->login_req_comp);
+
 	/*
 	 * For login requests after the first PDU, isert_rx_login_req() will
 	 * kick schedule_delayed_work(&conn->login_work) as the packet is
@@ -3102,11 +3075,15 @@ isert_get_login_rx(struct iscsi_conn *conn, struct iscsi_login *login)
 	if (!login->first_request)
 		return 0;
 
+	isert_rx_login_req(isert_conn);
+
+	isert_info("before conn_login_comp conn: %p\n", conn);
 	ret = wait_for_completion_interruptible(&isert_conn->conn_login_comp);
 	if (ret)
 		return ret;
 
-	pr_debug("isert_get_login_rx processing login->req: %p\n", login->req);
+	isert_info("processing login->req: %p\n", login->req);
+
 	return 0;
 }
 
@@ -3153,13 +3130,13 @@ isert_accept_np(struct iscsi_np *np, struct iscsi_conn *conn)
 
 accept_wait:
 	ret = down_interruptible(&isert_np->np_sem);
-	if (max_accept > 5)
+	if (ret || max_accept > 5)
 		return -ENODEV;
 
 	spin_lock_bh(&np->np_thread_lock);
 	if (np->np_thread_state >= ISCSI_NP_THREAD_RESET) {
 		spin_unlock_bh(&np->np_thread_lock);
-		pr_debug("np_thread_state %d for isert_accept_np\n",
+		isert_dbg("np_thread_state %d for isert_accept_np\n",
 			 np->np_thread_state);
 		/**
 		 * No point in stalling here when np_thread
@@ -3184,17 +3161,10 @@ accept_wait:
 	isert_conn->conn = conn;
 	max_accept = 0;
 
-	ret = isert_rdma_post_recvl(isert_conn);
-	if (ret)
-		return ret;
-
-	ret = isert_rdma_accept(isert_conn);
-	if (ret)
-		return ret;
-
 	isert_set_conn_info(np, conn, isert_conn);
 
-	pr_debug("Processing isert_accept_np: isert_conn: %p\n", isert_conn);
+	isert_dbg("Processing isert_conn: %p\n", isert_conn);
+
 	return 0;
 }
 
@@ -3202,24 +3172,103 @@ static void
 isert_free_np(struct iscsi_np *np)
 {
 	struct isert_np *isert_np = (struct isert_np *)np->np_context;
+	struct isert_conn *isert_conn, *n;
 
-	rdma_destroy_id(isert_np->np_cm_id);
+	if (isert_np->np_cm_id)
+		rdma_destroy_id(isert_np->np_cm_id);
+
+	/*
+	 * FIXME: At this point we don't have a good way to insure
+	 * that at this point we don't have hanging connections that
+	 * completed RDMA establishment but didn't start iscsi login
+	 * process. So work-around this by cleaning up what ever piled
+	 * up in np_accept_list.
+	 */
+	mutex_lock(&isert_np->np_accept_mutex);
+	if (!list_empty(&isert_np->np_accept_list)) {
+		isert_info("Still have isert connections, cleaning up...\n");
+		list_for_each_entry_safe(isert_conn, n,
+					 &isert_np->np_accept_list,
+					 conn_accept_node) {
+			isert_info("cleaning isert_conn %p state (%d)\n",
+				   isert_conn, isert_conn->state);
+			isert_connect_release(isert_conn);
+		}
+	}
+	mutex_unlock(&isert_np->np_accept_mutex);
 
 	np->np_context = NULL;
 	kfree(isert_np);
 }
 
+static void isert_release_work(struct work_struct *work)
+{
+	struct isert_conn *isert_conn = container_of(work,
+						     struct isert_conn,
+						     release_work);
+
+	isert_info("Starting release conn %p\n", isert_conn);
+
+	wait_for_completion(&isert_conn->conn_wait);
+
+	mutex_lock(&isert_conn->conn_mutex);
+	isert_conn->state = ISER_CONN_DOWN;
+	mutex_unlock(&isert_conn->conn_mutex);
+
+	isert_info("Destroying conn %p\n", isert_conn);
+	isert_put_conn(isert_conn);
+}
+
+static void
+isert_wait4logout(struct isert_conn *isert_conn)
+{
+	struct iscsi_conn *conn = isert_conn->conn;
+
+	isert_info("conn %p\n", isert_conn);
+
+	if (isert_conn->logout_posted) {
+		isert_info("conn %p wait for conn_logout_comp\n", isert_conn);
+		wait_for_completion_timeout(&conn->conn_logout_comp,
+					    SECONDS_FOR_LOGOUT_COMP * HZ);
+	}
+}
+
+static void
+isert_wait4cmds(struct iscsi_conn *conn)
+{
+	isert_info("iscsi_conn %p\n", conn);
+
+	if (conn->sess) {
+		target_sess_cmd_list_set_waiting(conn->sess->se_sess);
+		target_wait_for_sess_cmds(conn->sess->se_sess);
+	}
+}
+
+static void
+isert_wait4flush(struct isert_conn *isert_conn)
+{
+	struct ib_recv_wr *bad_wr;
+
+	isert_info("conn %p\n", isert_conn);
+
+	init_completion(&isert_conn->conn_wait_comp_err);
+	isert_conn->beacon.wr_id = ISER_BEACON_WRID;
+	/* post an indication that all flush errors were consumed */
+	if (ib_post_recv(isert_conn->conn_qp, &isert_conn->beacon, &bad_wr)) {
+		isert_err("conn %p failed to post beacon", isert_conn);
+		return;
+	}
+
+	wait_for_completion(&isert_conn->conn_wait_comp_err);
+}
+
 static void isert_wait_conn(struct iscsi_conn *conn)
 {
 	struct isert_conn *isert_conn = conn->context;
 
-	pr_debug("isert_wait_conn: Starting \n");
+	isert_info("Starting conn %p\n", isert_conn);
 
 	mutex_lock(&isert_conn->conn_mutex);
-	if (isert_conn->conn_cm_id && !isert_conn->disconnect) {
-		pr_debug("Calling rdma_disconnect from isert_wait_conn\n");
-		rdma_disconnect(isert_conn->conn_cm_id);
-	}
 	/*
 	 * Only wait for conn_wait_comp_err if the isert_conn made it
 	 * into full feature phase..
@@ -3228,14 +3277,15 @@ static void isert_wait_conn(struct iscsi_conn *conn)
 		mutex_unlock(&isert_conn->conn_mutex);
 		return;
 	}
-	if (isert_conn->state == ISER_CONN_UP)
-		isert_conn->state = ISER_CONN_TERMINATING;
+	isert_conn_terminate(isert_conn);
 	mutex_unlock(&isert_conn->conn_mutex);
 
-	wait_for_completion(&isert_conn->conn_wait_comp_err);
+	isert_wait4cmds(conn);
+	isert_wait4flush(isert_conn);
+	isert_wait4logout(isert_conn);
 
-	wait_for_completion(&isert_conn->conn_wait);
-	isert_put_conn(isert_conn);
+	INIT_WORK(&isert_conn->release_work, isert_release_work);
+	queue_work(isert_release_wq, &isert_conn->release_work);
 }
 
 static void isert_free_conn(struct iscsi_conn *conn)
@@ -3270,35 +3320,39 @@ static int __init isert_init(void)
 {
 	int ret;
 
-	isert_rx_wq = alloc_workqueue("isert_rx_wq", 0, 0);
-	if (!isert_rx_wq) {
-		pr_err("Unable to allocate isert_rx_wq\n");
+	isert_comp_wq = alloc_workqueue("isert_comp_wq", 0, 0);
+	if (!isert_comp_wq) {
+		isert_err("Unable to allocate isert_comp_wq\n");
+		ret = -ENOMEM;
 		return -ENOMEM;
 	}
 
-	isert_comp_wq = alloc_workqueue("isert_comp_wq", 0, 0);
-	if (!isert_comp_wq) {
-		pr_err("Unable to allocate isert_comp_wq\n");
+	isert_release_wq = alloc_workqueue("isert_release_wq", WQ_UNBOUND,
+					WQ_UNBOUND_MAX_ACTIVE);
+	if (!isert_release_wq) {
+		isert_err("Unable to allocate isert_release_wq\n");
 		ret = -ENOMEM;
-		goto destroy_rx_wq;
+		goto destroy_comp_wq;
 	}
 
 	iscsit_register_transport(&iser_target_transport);
-	pr_debug("iSER_TARGET[0] - Loaded iser_target_transport\n");
+	isert_info("iSER_TARGET[0] - Loaded iser_target_transport\n");
+
 	return 0;
 
-destroy_rx_wq:
-	destroy_workqueue(isert_rx_wq);
+destroy_comp_wq:
+	destroy_workqueue(isert_comp_wq);
+
 	return ret;
 }
 
 static void __exit isert_exit(void)
 {
 	flush_scheduled_work();
+	destroy_workqueue(isert_release_wq);
 	destroy_workqueue(isert_comp_wq);
-	destroy_workqueue(isert_rx_wq);
 	iscsit_unregister_transport(&iser_target_transport);
-	pr_debug("iSER_TARGET[0] - Released iser_target_transport\n");
+	isert_info("iSER_TARGET[0] - Released iser_target_transport\n");
 }
 
 MODULE_DESCRIPTION("iSER-Target for mainline target infrastructure");
diff --git a/drivers/infiniband/ulp/isert/ib_isert.h b/drivers/infiniband/ulp/isert/ib_isert.h
index 04f51f7bf614..8dc8415d152d 100644
--- a/drivers/infiniband/ulp/isert/ib_isert.h
+++ b/drivers/infiniband/ulp/isert/ib_isert.h
@@ -4,9 +4,37 @@
 #include <rdma/ib_verbs.h>
 #include <rdma/rdma_cm.h>
 
+#define DRV_NAME	"isert"
+#define PFX		DRV_NAME ": "
+
+#define isert_dbg(fmt, arg...)				 \
+	do {						 \
+		if (unlikely(isert_debug_level > 2))	 \
+			printk(KERN_DEBUG PFX "%s: " fmt,\
+				__func__ , ## arg);	 \
+	} while (0)
+
+#define isert_warn(fmt, arg...)				\
+	do {						\
+		if (unlikely(isert_debug_level > 0))	\
+			pr_warn(PFX "%s: " fmt,         \
+				__func__ , ## arg);	\
+	} while (0)
+
+#define isert_info(fmt, arg...)				\
+	do {						\
+		if (unlikely(isert_debug_level > 1))	\
+			pr_info(PFX "%s: " fmt,         \
+				__func__ , ## arg);	\
+	} while (0)
+
+#define isert_err(fmt, arg...) \
+	pr_err(PFX "%s: " fmt, __func__ , ## arg)
+
 #define ISERT_RDMA_LISTEN_BACKLOG	10
 #define ISCSI_ISER_SG_TABLESIZE		256
 #define ISER_FASTREG_LI_WRID		0xffffffffffffffffULL
+#define ISER_BEACON_WRID               0xfffffffffffffffeULL
 
 enum isert_desc_type {
 	ISCSI_TX_CONTROL,
@@ -23,6 +51,7 @@ enum iser_ib_op_code {
 enum iser_conn_state {
 	ISER_CONN_INIT,
 	ISER_CONN_UP,
+	ISER_CONN_FULL_FEATURE,
 	ISER_CONN_TERMINATING,
 	ISER_CONN_DOWN,
 };
@@ -44,9 +73,6 @@ struct iser_tx_desc {
 	struct ib_sge	tx_sg[2];
 	int		num_sge;
 	struct isert_cmd *isert_cmd;
-	struct llist_node *comp_llnode_batch;
-	struct llist_node comp_llnode;
-	bool		llnode_active;
 	struct ib_send_wr send_wr;
 } __packed;
 
@@ -81,6 +107,12 @@ struct isert_data_buf {
 	enum dma_data_direction dma_dir;
 };
 
+enum {
+	DATA = 0,
+	PROT = 1,
+	SIG = 2,
+};
+
 struct isert_rdma_wr {
 	struct list_head	wr_list;
 	struct isert_cmd	*isert_cmd;
@@ -90,6 +122,7 @@ struct isert_rdma_wr {
 	int			send_wr_num;
 	struct ib_send_wr	*send_wr;
 	struct ib_send_wr	s_send_wr;
+	struct ib_sge		ib_sg[3];
 	struct isert_data_buf	data;
 	struct isert_data_buf	prot;
 	struct fast_reg_descriptor *fr_desc;
@@ -117,14 +150,15 @@ struct isert_device;
 struct isert_conn {
 	enum iser_conn_state	state;
 	int			post_recv_buf_count;
-	atomic_t		post_send_buf_count;
 	u32			responder_resources;
 	u32			initiator_depth;
+	bool			pi_support;
 	u32			max_sge;
 	char			*login_buf;
 	char			*login_req_buf;
 	char			*login_rsp_buf;
 	u64			login_req_dma;
+	int			login_req_len;
 	u64			login_rsp_dma;
 	unsigned int		conn_rx_desc_head;
 	struct iser_rx_desc	*conn_rx_descs;
@@ -132,13 +166,13 @@ struct isert_conn {
 	struct iscsi_conn	*conn;
 	struct list_head	conn_accept_node;
 	struct completion	conn_login_comp;
+	struct completion	login_req_comp;
 	struct iser_tx_desc	conn_login_tx_desc;
 	struct rdma_cm_id	*conn_cm_id;
 	struct ib_pd		*conn_pd;
 	struct ib_mr		*conn_mr;
 	struct ib_qp		*conn_qp;
 	struct isert_device	*conn_device;
-	struct work_struct	conn_logout_work;
 	struct mutex		conn_mutex;
 	struct completion	conn_wait;
 	struct completion	conn_wait_comp_err;
@@ -147,31 +181,38 @@ struct isert_conn {
 	int			conn_fr_pool_size;
 	/* lock to protect fastreg pool */
 	spinlock_t		conn_lock;
-#define ISERT_COMP_BATCH_COUNT	8
-	int			conn_comp_batch;
-	struct llist_head	conn_comp_llist;
-	bool                    disconnect;
+	struct work_struct	release_work;
+	struct ib_recv_wr       beacon;
+	bool                    logout_posted;
 };
 
 #define ISERT_MAX_CQ 64
 
-struct isert_cq_desc {
-	struct isert_device	*device;
-	int			cq_index;
-	struct work_struct	cq_rx_work;
-	struct work_struct	cq_tx_work;
+/**
+ * struct isert_comp - iSER completion context
+ *
+ * @device:     pointer to device handle
+ * @cq:         completion queue
+ * @wcs:        work completion array
+ * @active_qps: Number of active QPs attached
+ *              to completion context
+ * @work:       completion work handle
+ */
+struct isert_comp {
+	struct isert_device     *device;
+	struct ib_cq		*cq;
+	struct ib_wc		 wcs[16];
+	int                      active_qps;
+	struct work_struct	 work;
 };
 
 struct isert_device {
 	int			use_fastreg;
 	bool			pi_capable;
-	int			cqs_used;
 	int			refcount;
-	int			cq_active_qps[ISERT_MAX_CQ];
 	struct ib_device	*ib_device;
-	struct ib_cq		*dev_rx_cq[ISERT_MAX_CQ];
-	struct ib_cq		*dev_tx_cq[ISERT_MAX_CQ];
-	struct isert_cq_desc	*cq_desc;
+	struct isert_comp	*comps;
+	int                     comps_used;
 	struct list_head	dev_node;
 	struct ib_device_attr	dev_attr;
 	int			(*reg_rdma_mem)(struct iscsi_conn *conn,
@@ -182,6 +223,7 @@ struct isert_device {
 };
 
 struct isert_np {
+	struct iscsi_np         *np;
 	struct semaphore	np_sem;
 	struct rdma_cm_id	*np_cm_id;
 	struct mutex		np_accept_mutex;
diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c
index 62d2a18e1b41..0747c0595a9d 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.c
+++ b/drivers/infiniband/ulp/srp/ib_srp.c
@@ -123,10 +123,15 @@ MODULE_PARM_DESC(dev_loss_tmo,
 		 " if fast_io_fail_tmo has not been set. \"off\" means that"
 		 " this functionality is disabled.");
 
+static unsigned ch_count;
+module_param(ch_count, uint, 0444);
+MODULE_PARM_DESC(ch_count,
+		 "Number of RDMA channels to use for communication with an SRP target. Using more than one channel improves performance if the HCA supports multiple completion vectors. The default value is the minimum of four times the number of online CPU sockets and the number of completion vectors supported by the HCA.");
+
 static void srp_add_one(struct ib_device *device);
 static void srp_remove_one(struct ib_device *device);
-static void srp_recv_completion(struct ib_cq *cq, void *target_ptr);
-static void srp_send_completion(struct ib_cq *cq, void *target_ptr);
+static void srp_recv_completion(struct ib_cq *cq, void *ch_ptr);
+static void srp_send_completion(struct ib_cq *cq, void *ch_ptr);
 static int srp_cm_handler(struct ib_cm_id *cm_id, struct ib_cm_event *event);
 
 static struct scsi_transport_template *ib_srp_transport_template;
@@ -262,7 +267,7 @@ static int srp_init_qp(struct srp_target_port *target,
 
 	ret = ib_find_pkey(target->srp_host->srp_dev->dev,
 			   target->srp_host->port,
-			   be16_to_cpu(target->path.pkey),
+			   be16_to_cpu(target->pkey),
 			   &attr->pkey_index);
 	if (ret)
 		goto out;
@@ -283,18 +288,23 @@ out:
 	return ret;
 }
 
-static int srp_new_cm_id(struct srp_target_port *target)
+static int srp_new_cm_id(struct srp_rdma_ch *ch)
 {
+	struct srp_target_port *target = ch->target;
 	struct ib_cm_id *new_cm_id;
 
 	new_cm_id = ib_create_cm_id(target->srp_host->srp_dev->dev,
-				    srp_cm_handler, target);
+				    srp_cm_handler, ch);
 	if (IS_ERR(new_cm_id))
 		return PTR_ERR(new_cm_id);
 
-	if (target->cm_id)
-		ib_destroy_cm_id(target->cm_id);
-	target->cm_id = new_cm_id;
+	if (ch->cm_id)
+		ib_destroy_cm_id(ch->cm_id);
+	ch->cm_id = new_cm_id;
+	ch->path.sgid = target->sgid;
+	ch->path.dgid = target->orig_dgid;
+	ch->path.pkey = target->pkey;
+	ch->path.service_id = target->service_id;
 
 	return 0;
 }
@@ -443,8 +453,44 @@ static struct srp_fr_pool *srp_alloc_fr_pool(struct srp_target_port *target)
 				  dev->max_pages_per_mr);
 }
 
-static int srp_create_target_ib(struct srp_target_port *target)
+/**
+ * srp_destroy_qp() - destroy an RDMA queue pair
+ * @ch: SRP RDMA channel.
+ *
+ * Change a queue pair into the error state and wait until all receive
+ * completions have been processed before destroying it. This avoids that
+ * the receive completion handler can access the queue pair while it is
+ * being destroyed.
+ */
+static void srp_destroy_qp(struct srp_rdma_ch *ch)
+{
+	struct srp_target_port *target = ch->target;
+	static struct ib_qp_attr attr = { .qp_state = IB_QPS_ERR };
+	static struct ib_recv_wr wr = { .wr_id = SRP_LAST_WR_ID };
+	struct ib_recv_wr *bad_wr;
+	int ret;
+
+	/* Destroying a QP and reusing ch->done is only safe if not connected */
+	WARN_ON_ONCE(target->connected);
+
+	ret = ib_modify_qp(ch->qp, &attr, IB_QP_STATE);
+	WARN_ONCE(ret, "ib_cm_init_qp_attr() returned %d\n", ret);
+	if (ret)
+		goto out;
+
+	init_completion(&ch->done);
+	ret = ib_post_recv(ch->qp, &wr, &bad_wr);
+	WARN_ONCE(ret, "ib_post_recv() returned %d\n", ret);
+	if (ret == 0)
+		wait_for_completion(&ch->done);
+
+out:
+	ib_destroy_qp(ch->qp);
+}
+
+static int srp_create_ch_ib(struct srp_rdma_ch *ch)
 {
+	struct srp_target_port *target = ch->target;
 	struct srp_device *dev = target->srp_host->srp_dev;
 	struct ib_qp_init_attr *init_attr;
 	struct ib_cq *recv_cq, *send_cq;
@@ -458,15 +504,16 @@ static int srp_create_target_ib(struct srp_target_port *target)
 	if (!init_attr)
 		return -ENOMEM;
 
-	recv_cq = ib_create_cq(dev->dev, srp_recv_completion, NULL, target,
-			       target->queue_size, target->comp_vector);
+	/* + 1 for SRP_LAST_WR_ID */
+	recv_cq = ib_create_cq(dev->dev, srp_recv_completion, NULL, ch,
+			       target->queue_size + 1, ch->comp_vector);
 	if (IS_ERR(recv_cq)) {
 		ret = PTR_ERR(recv_cq);
 		goto err;
 	}
 
-	send_cq = ib_create_cq(dev->dev, srp_send_completion, NULL, target,
-			       m * target->queue_size, target->comp_vector);
+	send_cq = ib_create_cq(dev->dev, srp_send_completion, NULL, ch,
+			       m * target->queue_size, ch->comp_vector);
 	if (IS_ERR(send_cq)) {
 		ret = PTR_ERR(send_cq);
 		goto err_recv_cq;
@@ -476,7 +523,7 @@ static int srp_create_target_ib(struct srp_target_port *target)
 
 	init_attr->event_handler       = srp_qp_event;
 	init_attr->cap.max_send_wr     = m * target->queue_size;
-	init_attr->cap.max_recv_wr     = target->queue_size;
+	init_attr->cap.max_recv_wr     = target->queue_size + 1;
 	init_attr->cap.max_recv_sge    = 1;
 	init_attr->cap.max_send_sge    = 1;
 	init_attr->sq_sig_type         = IB_SIGNAL_REQ_WR;
@@ -502,9 +549,9 @@ static int srp_create_target_ib(struct srp_target_port *target)
 				     "FR pool allocation failed (%d)\n", ret);
 			goto err_qp;
 		}
-		if (target->fr_pool)
-			srp_destroy_fr_pool(target->fr_pool);
-		target->fr_pool = fr_pool;
+		if (ch->fr_pool)
+			srp_destroy_fr_pool(ch->fr_pool);
+		ch->fr_pool = fr_pool;
 	} else if (!dev->use_fast_reg && dev->has_fmr) {
 		fmr_pool = srp_alloc_fmr_pool(target);
 		if (IS_ERR(fmr_pool)) {
@@ -513,21 +560,21 @@ static int srp_create_target_ib(struct srp_target_port *target)
 				     "FMR pool allocation failed (%d)\n", ret);
 			goto err_qp;
 		}
-		if (target->fmr_pool)
-			ib_destroy_fmr_pool(target->fmr_pool);
-		target->fmr_pool = fmr_pool;
+		if (ch->fmr_pool)
+			ib_destroy_fmr_pool(ch->fmr_pool);
+		ch->fmr_pool = fmr_pool;
 	}
 
-	if (target->qp)
-		ib_destroy_qp(target->qp);
-	if (target->recv_cq)
-		ib_destroy_cq(target->recv_cq);
-	if (target->send_cq)
-		ib_destroy_cq(target->send_cq);
+	if (ch->qp)
+		srp_destroy_qp(ch);
+	if (ch->recv_cq)
+		ib_destroy_cq(ch->recv_cq);
+	if (ch->send_cq)
+		ib_destroy_cq(ch->send_cq);
 
-	target->qp = qp;
-	target->recv_cq = recv_cq;
-	target->send_cq = send_cq;
+	ch->qp = qp;
+	ch->recv_cq = recv_cq;
+	ch->send_cq = send_cq;
 
 	kfree(init_attr);
 	return 0;
@@ -548,93 +595,117 @@ err:
 
 /*
  * Note: this function may be called without srp_alloc_iu_bufs() having been
- * invoked. Hence the target->[rt]x_ring checks.
+ * invoked. Hence the ch->[rt]x_ring checks.
  */
-static void srp_free_target_ib(struct srp_target_port *target)
+static void srp_free_ch_ib(struct srp_target_port *target,
+			   struct srp_rdma_ch *ch)
 {
 	struct srp_device *dev = target->srp_host->srp_dev;
 	int i;
 
+	if (!ch->target)
+		return;
+
+	if (ch->cm_id) {
+		ib_destroy_cm_id(ch->cm_id);
+		ch->cm_id = NULL;
+	}
+
+	/* If srp_new_cm_id() succeeded but srp_create_ch_ib() not, return. */
+	if (!ch->qp)
+		return;
+
 	if (dev->use_fast_reg) {
-		if (target->fr_pool)
-			srp_destroy_fr_pool(target->fr_pool);
+		if (ch->fr_pool)
+			srp_destroy_fr_pool(ch->fr_pool);
 	} else {
-		if (target->fmr_pool)
-			ib_destroy_fmr_pool(target->fmr_pool);
+		if (ch->fmr_pool)
+			ib_destroy_fmr_pool(ch->fmr_pool);
 	}
-	ib_destroy_qp(target->qp);
-	ib_destroy_cq(target->send_cq);
-	ib_destroy_cq(target->recv_cq);
+	srp_destroy_qp(ch);
+	ib_destroy_cq(ch->send_cq);
+	ib_destroy_cq(ch->recv_cq);
+
+	/*
+	 * Avoid that the SCSI error handler tries to use this channel after
+	 * it has been freed. The SCSI error handler can namely continue
+	 * trying to perform recovery actions after scsi_remove_host()
+	 * returned.
+	 */
+	ch->target = NULL;
 
-	target->qp = NULL;
-	target->send_cq = target->recv_cq = NULL;
+	ch->qp = NULL;
+	ch->send_cq = ch->recv_cq = NULL;
 
-	if (target->rx_ring) {
+	if (ch->rx_ring) {
 		for (i = 0; i < target->queue_size; ++i)
-			srp_free_iu(target->srp_host, target->rx_ring[i]);
-		kfree(target->rx_ring);
-		target->rx_ring = NULL;
+			srp_free_iu(target->srp_host, ch->rx_ring[i]);
+		kfree(ch->rx_ring);
+		ch->rx_ring = NULL;
 	}
-	if (target->tx_ring) {
+	if (ch->tx_ring) {
 		for (i = 0; i < target->queue_size; ++i)
-			srp_free_iu(target->srp_host, target->tx_ring[i]);
-		kfree(target->tx_ring);
-		target->tx_ring = NULL;
+			srp_free_iu(target->srp_host, ch->tx_ring[i]);
+		kfree(ch->tx_ring);
+		ch->tx_ring = NULL;
 	}
 }
 
 static void srp_path_rec_completion(int status,
 				    struct ib_sa_path_rec *pathrec,
-				    void *target_ptr)
+				    void *ch_ptr)
 {
-	struct srp_target_port *target = target_ptr;
+	struct srp_rdma_ch *ch = ch_ptr;
+	struct srp_target_port *target = ch->target;
 
-	target->status = status;
+	ch->status = status;
 	if (status)
 		shost_printk(KERN_ERR, target->scsi_host,
 			     PFX "Got failed path rec status %d\n", status);
 	else
-		target->path = *pathrec;
-	complete(&target->done);
+		ch->path = *pathrec;
+	complete(&ch->done);
 }
 
-static int srp_lookup_path(struct srp_target_port *target)
+static int srp_lookup_path(struct srp_rdma_ch *ch)
 {
+	struct srp_target_port *target = ch->target;
 	int ret;
 
-	target->path.numb_path = 1;
-
-	init_completion(&target->done);
-
-	target->path_query_id = ib_sa_path_rec_get(&srp_sa_client,
-						   target->srp_host->srp_dev->dev,
-						   target->srp_host->port,
-						   &target->path,
-						   IB_SA_PATH_REC_SERVICE_ID	|
-						   IB_SA_PATH_REC_DGID		|
-						   IB_SA_PATH_REC_SGID		|
-						   IB_SA_PATH_REC_NUMB_PATH	|
-						   IB_SA_PATH_REC_PKEY,
-						   SRP_PATH_REC_TIMEOUT_MS,
-						   GFP_KERNEL,
-						   srp_path_rec_completion,
-						   target, &target->path_query);
-	if (target->path_query_id < 0)
-		return target->path_query_id;
-
-	ret = wait_for_completion_interruptible(&target->done);
+	ch->path.numb_path = 1;
+
+	init_completion(&ch->done);
+
+	ch->path_query_id = ib_sa_path_rec_get(&srp_sa_client,
+					       target->srp_host->srp_dev->dev,
+					       target->srp_host->port,
+					       &ch->path,
+					       IB_SA_PATH_REC_SERVICE_ID |
+					       IB_SA_PATH_REC_DGID	 |
+					       IB_SA_PATH_REC_SGID	 |
+					       IB_SA_PATH_REC_NUMB_PATH	 |
+					       IB_SA_PATH_REC_PKEY,
+					       SRP_PATH_REC_TIMEOUT_MS,
+					       GFP_KERNEL,
+					       srp_path_rec_completion,
+					       ch, &ch->path_query);
+	if (ch->path_query_id < 0)
+		return ch->path_query_id;
+
+	ret = wait_for_completion_interruptible(&ch->done);
 	if (ret < 0)
 		return ret;
 
-	if (target->status < 0)
+	if (ch->status < 0)
 		shost_printk(KERN_WARNING, target->scsi_host,
 			     PFX "Path record query failed\n");
 
-	return target->status;
+	return ch->status;
 }
 
-static int srp_send_req(struct srp_target_port *target)
+static int srp_send_req(struct srp_rdma_ch *ch, bool multich)
 {
+	struct srp_target_port *target = ch->target;
 	struct {
 		struct ib_cm_req_param param;
 		struct srp_login_req   priv;
@@ -645,11 +716,11 @@ static int srp_send_req(struct srp_target_port *target)
 	if (!req)
 		return -ENOMEM;
 
-	req->param.primary_path 	      = &target->path;
+	req->param.primary_path		      = &ch->path;
 	req->param.alternate_path 	      = NULL;
 	req->param.service_id 		      = target->service_id;
-	req->param.qp_num 		      = target->qp->qp_num;
-	req->param.qp_type 		      = target->qp->qp_type;
+	req->param.qp_num		      = ch->qp->qp_num;
+	req->param.qp_type		      = ch->qp->qp_type;
 	req->param.private_data 	      = &req->priv;
 	req->param.private_data_len 	      = sizeof req->priv;
 	req->param.flow_control 	      = 1;
@@ -673,6 +744,8 @@ static int srp_send_req(struct srp_target_port *target)
 	req->priv.req_it_iu_len = cpu_to_be32(target->max_iu_len);
 	req->priv.req_buf_fmt 	= cpu_to_be16(SRP_BUF_FORMAT_DIRECT |
 					      SRP_BUF_FORMAT_INDIRECT);
+	req->priv.req_flags	= (multich ? SRP_MULTICHAN_MULTI :
+				   SRP_MULTICHAN_SINGLE);
 	/*
 	 * In the published SRP specification (draft rev. 16a), the
 	 * port identifier format is 8 bytes of ID extension followed
@@ -684,7 +757,7 @@ static int srp_send_req(struct srp_target_port *target)
 	 */
 	if (target->io_class == SRP_REV10_IB_IO_CLASS) {
 		memcpy(req->priv.initiator_port_id,
-		       &target->path.sgid.global.interface_id, 8);
+		       &target->sgid.global.interface_id, 8);
 		memcpy(req->priv.initiator_port_id + 8,
 		       &target->initiator_ext, 8);
 		memcpy(req->priv.target_port_id,     &target->ioc_guid, 8);
@@ -693,7 +766,7 @@ static int srp_send_req(struct srp_target_port *target)
 		memcpy(req->priv.initiator_port_id,
 		       &target->initiator_ext, 8);
 		memcpy(req->priv.initiator_port_id + 8,
-		       &target->path.sgid.global.interface_id, 8);
+		       &target->sgid.global.interface_id, 8);
 		memcpy(req->priv.target_port_id,     &target->id_ext, 8);
 		memcpy(req->priv.target_port_id + 8, &target->ioc_guid, 8);
 	}
@@ -713,7 +786,7 @@ static int srp_send_req(struct srp_target_port *target)
 		       &target->srp_host->srp_dev->dev->node_guid, 8);
 	}
 
-	status = ib_send_cm_req(target->cm_id, &req->param);
+	status = ib_send_cm_req(ch->cm_id, &req->param);
 
 	kfree(req);
 
@@ -754,28 +827,35 @@ static bool srp_change_conn_state(struct srp_target_port *target,
 
 static void srp_disconnect_target(struct srp_target_port *target)
 {
+	struct srp_rdma_ch *ch;
+	int i;
+
 	if (srp_change_conn_state(target, false)) {
 		/* XXX should send SRP_I_LOGOUT request */
 
-		if (ib_send_cm_dreq(target->cm_id, NULL, 0)) {
-			shost_printk(KERN_DEBUG, target->scsi_host,
-				     PFX "Sending CM DREQ failed\n");
+		for (i = 0; i < target->ch_count; i++) {
+			ch = &target->ch[i];
+			if (ch->cm_id && ib_send_cm_dreq(ch->cm_id, NULL, 0)) {
+				shost_printk(KERN_DEBUG, target->scsi_host,
+					     PFX "Sending CM DREQ failed\n");
+			}
 		}
 	}
 }
 
-static void srp_free_req_data(struct srp_target_port *target)
+static void srp_free_req_data(struct srp_target_port *target,
+			      struct srp_rdma_ch *ch)
 {
 	struct srp_device *dev = target->srp_host->srp_dev;
 	struct ib_device *ibdev = dev->dev;
 	struct srp_request *req;
 	int i;
 
-	if (!target->req_ring)
+	if (!ch->target || !ch->req_ring)
 		return;
 
 	for (i = 0; i < target->req_ring_size; ++i) {
-		req = &target->req_ring[i];
+		req = &ch->req_ring[i];
 		if (dev->use_fast_reg)
 			kfree(req->fr_list);
 		else
@@ -789,12 +869,13 @@ static void srp_free_req_data(struct srp_target_port *target)
 		kfree(req->indirect_desc);
 	}
 
-	kfree(target->req_ring);
-	target->req_ring = NULL;
+	kfree(ch->req_ring);
+	ch->req_ring = NULL;
 }
 
-static int srp_alloc_req_data(struct srp_target_port *target)
+static int srp_alloc_req_data(struct srp_rdma_ch *ch)
 {
+	struct srp_target_port *target = ch->target;
 	struct srp_device *srp_dev = target->srp_host->srp_dev;
 	struct ib_device *ibdev = srp_dev->dev;
 	struct srp_request *req;
@@ -802,15 +883,13 @@ static int srp_alloc_req_data(struct srp_target_port *target)
 	dma_addr_t dma_addr;
 	int i, ret = -ENOMEM;
 
-	INIT_LIST_HEAD(&target->free_reqs);
-
-	target->req_ring = kzalloc(target->req_ring_size *
-				   sizeof(*target->req_ring), GFP_KERNEL);
-	if (!target->req_ring)
+	ch->req_ring = kcalloc(target->req_ring_size, sizeof(*ch->req_ring),
+			       GFP_KERNEL);
+	if (!ch->req_ring)
 		goto out;
 
 	for (i = 0; i < target->req_ring_size; ++i) {
-		req = &target->req_ring[i];
+		req = &ch->req_ring[i];
 		mr_list = kmalloc(target->cmd_sg_cnt * sizeof(void *),
 				  GFP_KERNEL);
 		if (!mr_list)
@@ -834,8 +913,6 @@ static int srp_alloc_req_data(struct srp_target_port *target)
 			goto out;
 
 		req->indirect_dma_addr = dma_addr;
-		req->index = i;
-		list_add_tail(&req->list, &target->free_reqs);
 	}
 	ret = 0;
 
@@ -860,6 +937,9 @@ static void srp_del_scsi_host_attr(struct Scsi_Host *shost)
 
 static void srp_remove_target(struct srp_target_port *target)
 {
+	struct srp_rdma_ch *ch;
+	int i;
+
 	WARN_ON_ONCE(target->state != SRP_TARGET_REMOVED);
 
 	srp_del_scsi_host_attr(target->scsi_host);
@@ -868,11 +948,18 @@ static void srp_remove_target(struct srp_target_port *target)
 	scsi_remove_host(target->scsi_host);
 	srp_stop_rport_timers(target->rport);
 	srp_disconnect_target(target);
-	ib_destroy_cm_id(target->cm_id);
-	srp_free_target_ib(target);
+	for (i = 0; i < target->ch_count; i++) {
+		ch = &target->ch[i];
+		srp_free_ch_ib(target, ch);
+	}
 	cancel_work_sync(&target->tl_err_work);
 	srp_rport_put(target->rport);
-	srp_free_req_data(target);
+	for (i = 0; i < target->ch_count; i++) {
+		ch = &target->ch[i];
+		srp_free_req_data(target, ch);
+	}
+	kfree(target->ch);
+	target->ch = NULL;
 
 	spin_lock(&target->srp_host->target_lock);
 	list_del(&target->list);
@@ -898,25 +985,25 @@ static void srp_rport_delete(struct srp_rport *rport)
 	srp_queue_remove_work(target);
 }
 
-static int srp_connect_target(struct srp_target_port *target)
+static int srp_connect_ch(struct srp_rdma_ch *ch, bool multich)
 {
-	int retries = 3;
+	struct srp_target_port *target = ch->target;
 	int ret;
 
-	WARN_ON_ONCE(target->connected);
+	WARN_ON_ONCE(!multich && target->connected);
 
 	target->qp_in_error = false;
 
-	ret = srp_lookup_path(target);
+	ret = srp_lookup_path(ch);
 	if (ret)
 		return ret;
 
 	while (1) {
-		init_completion(&target->done);
-		ret = srp_send_req(target);
+		init_completion(&ch->done);
+		ret = srp_send_req(ch, multich);
 		if (ret)
 			return ret;
-		ret = wait_for_completion_interruptible(&target->done);
+		ret = wait_for_completion_interruptible(&ch->done);
 		if (ret < 0)
 			return ret;
 
@@ -926,13 +1013,13 @@ static int srp_connect_target(struct srp_target_port *target)
 		 * back, or SRP_DLID_REDIRECT if we get a lid/qp
 		 * redirect REJ back.
 		 */
-		switch (target->status) {
+		switch (ch->status) {
 		case 0:
 			srp_change_conn_state(target, true);
 			return 0;
 
 		case SRP_PORT_REDIRECT:
-			ret = srp_lookup_path(target);
+			ret = srp_lookup_path(ch);
 			if (ret)
 				return ret;
 			break;
@@ -941,27 +1028,18 @@ static int srp_connect_target(struct srp_target_port *target)
 			break;
 
 		case SRP_STALE_CONN:
-			/* Our current CM id was stale, and is now in timewait.
-			 * Try to reconnect with a new one.
-			 */
-			if (!retries-- || srp_new_cm_id(target)) {
-				shost_printk(KERN_ERR, target->scsi_host, PFX
-					     "giving up on stale connection\n");
-				target->status = -ECONNRESET;
-				return target->status;
-			}
-
 			shost_printk(KERN_ERR, target->scsi_host, PFX
-				     "retrying stale connection\n");
-			break;
+				     "giving up on stale connection\n");
+			ch->status = -ECONNRESET;
+			return ch->status;
 
 		default:
-			return target->status;
+			return ch->status;
 		}
 	}
 }
 
-static int srp_inv_rkey(struct srp_target_port *target, u32 rkey)
+static int srp_inv_rkey(struct srp_rdma_ch *ch, u32 rkey)
 {
 	struct ib_send_wr *bad_wr;
 	struct ib_send_wr wr = {
@@ -973,13 +1051,14 @@ static int srp_inv_rkey(struct srp_target_port *target, u32 rkey)
 		.ex.invalidate_rkey = rkey,
 	};
 
-	return ib_post_send(target->qp, &wr, &bad_wr);
+	return ib_post_send(ch->qp, &wr, &bad_wr);
 }
 
 static void srp_unmap_data(struct scsi_cmnd *scmnd,
-			   struct srp_target_port *target,
+			   struct srp_rdma_ch *ch,
 			   struct srp_request *req)
 {
+	struct srp_target_port *target = ch->target;
 	struct srp_device *dev = target->srp_host->srp_dev;
 	struct ib_device *ibdev = dev->dev;
 	int i, res;
@@ -993,7 +1072,7 @@ static void srp_unmap_data(struct scsi_cmnd *scmnd,
 		struct srp_fr_desc **pfr;
 
 		for (i = req->nmdesc, pfr = req->fr_list; i > 0; i--, pfr++) {
-			res = srp_inv_rkey(target, (*pfr)->mr->rkey);
+			res = srp_inv_rkey(ch, (*pfr)->mr->rkey);
 			if (res < 0) {
 				shost_printk(KERN_ERR, target->scsi_host, PFX
 				  "Queueing INV WR for rkey %#x failed (%d)\n",
@@ -1003,7 +1082,7 @@ static void srp_unmap_data(struct scsi_cmnd *scmnd,
 			}
 		}
 		if (req->nmdesc)
-			srp_fr_pool_put(target->fr_pool, req->fr_list,
+			srp_fr_pool_put(ch->fr_pool, req->fr_list,
 					req->nmdesc);
 	} else {
 		struct ib_pool_fmr **pfmr;
@@ -1018,7 +1097,7 @@ static void srp_unmap_data(struct scsi_cmnd *scmnd,
 
 /**
  * srp_claim_req - Take ownership of the scmnd associated with a request.
- * @target: SRP target port.
+ * @ch: SRP RDMA channel.
  * @req: SRP request.
  * @sdev: If not NULL, only take ownership for this SCSI device.
  * @scmnd: If NULL, take ownership of @req->scmnd. If not NULL, only take
@@ -1027,14 +1106,14 @@ static void srp_unmap_data(struct scsi_cmnd *scmnd,
  * Return value:
  * Either NULL or a pointer to the SCSI command the caller became owner of.
  */
-static struct scsi_cmnd *srp_claim_req(struct srp_target_port *target,
+static struct scsi_cmnd *srp_claim_req(struct srp_rdma_ch *ch,
 				       struct srp_request *req,
 				       struct scsi_device *sdev,
 				       struct scsi_cmnd *scmnd)
 {
 	unsigned long flags;
 
-	spin_lock_irqsave(&target->lock, flags);
+	spin_lock_irqsave(&ch->lock, flags);
 	if (req->scmnd &&
 	    (!sdev || req->scmnd->device == sdev) &&
 	    (!scmnd || req->scmnd == scmnd)) {
@@ -1043,40 +1122,37 @@ static struct scsi_cmnd *srp_claim_req(struct srp_target_port *target,
 	} else {
 		scmnd = NULL;
 	}
-	spin_unlock_irqrestore(&target->lock, flags);
+	spin_unlock_irqrestore(&ch->lock, flags);
 
 	return scmnd;
 }
 
 /**
  * srp_free_req() - Unmap data and add request to the free request list.
- * @target: SRP target port.
+ * @ch:     SRP RDMA channel.
  * @req:    Request to be freed.
  * @scmnd:  SCSI command associated with @req.
  * @req_lim_delta: Amount to be added to @target->req_lim.
  */
-static void srp_free_req(struct srp_target_port *target,
-			 struct srp_request *req, struct scsi_cmnd *scmnd,
-			 s32 req_lim_delta)
+static void srp_free_req(struct srp_rdma_ch *ch, struct srp_request *req,
+			 struct scsi_cmnd *scmnd, s32 req_lim_delta)
 {
 	unsigned long flags;
 
-	srp_unmap_data(scmnd, target, req);
+	srp_unmap_data(scmnd, ch, req);
 
-	spin_lock_irqsave(&target->lock, flags);
-	target->req_lim += req_lim_delta;
-	list_add_tail(&req->list, &target->free_reqs);
-	spin_unlock_irqrestore(&target->lock, flags);
+	spin_lock_irqsave(&ch->lock, flags);
+	ch->req_lim += req_lim_delta;
+	spin_unlock_irqrestore(&ch->lock, flags);
 }
 
-static void srp_finish_req(struct srp_target_port *target,
-			   struct srp_request *req, struct scsi_device *sdev,
-			   int result)
+static void srp_finish_req(struct srp_rdma_ch *ch, struct srp_request *req,
+			   struct scsi_device *sdev, int result)
 {
-	struct scsi_cmnd *scmnd = srp_claim_req(target, req, sdev, NULL);
+	struct scsi_cmnd *scmnd = srp_claim_req(ch, req, sdev, NULL);
 
 	if (scmnd) {
-		srp_free_req(target, req, scmnd, 0);
+		srp_free_req(ch, req, scmnd, 0);
 		scmnd->result = result;
 		scmnd->scsi_done(scmnd);
 	}
@@ -1085,9 +1161,10 @@ static void srp_finish_req(struct srp_target_port *target,
 static void srp_terminate_io(struct srp_rport *rport)
 {
 	struct srp_target_port *target = rport->lld_data;
+	struct srp_rdma_ch *ch;
 	struct Scsi_Host *shost = target->scsi_host;
 	struct scsi_device *sdev;
-	int i;
+	int i, j;
 
 	/*
 	 * Invoking srp_terminate_io() while srp_queuecommand() is running
@@ -1096,9 +1173,15 @@ static void srp_terminate_io(struct srp_rport *rport)
 	shost_for_each_device(sdev, shost)
 		WARN_ON_ONCE(sdev->request_queue->request_fn_active);
 
-	for (i = 0; i < target->req_ring_size; ++i) {
-		struct srp_request *req = &target->req_ring[i];
-		srp_finish_req(target, req, NULL, DID_TRANSPORT_FAILFAST << 16);
+	for (i = 0; i < target->ch_count; i++) {
+		ch = &target->ch[i];
+
+		for (j = 0; j < target->req_ring_size; ++j) {
+			struct srp_request *req = &ch->req_ring[j];
+
+			srp_finish_req(ch, req, NULL,
+				       DID_TRANSPORT_FAILFAST << 16);
+		}
 	}
 }
 
@@ -1114,34 +1197,61 @@ static void srp_terminate_io(struct srp_rport *rport)
 static int srp_rport_reconnect(struct srp_rport *rport)
 {
 	struct srp_target_port *target = rport->lld_data;
-	int i, ret;
+	struct srp_rdma_ch *ch;
+	int i, j, ret = 0;
+	bool multich = false;
 
 	srp_disconnect_target(target);
+
+	if (target->state == SRP_TARGET_SCANNING)
+		return -ENODEV;
+
 	/*
 	 * Now get a new local CM ID so that we avoid confusing the target in
 	 * case things are really fouled up. Doing so also ensures that all CM
 	 * callbacks will have finished before a new QP is allocated.
 	 */
-	ret = srp_new_cm_id(target);
-
-	for (i = 0; i < target->req_ring_size; ++i) {
-		struct srp_request *req = &target->req_ring[i];
-		srp_finish_req(target, req, NULL, DID_RESET << 16);
+	for (i = 0; i < target->ch_count; i++) {
+		ch = &target->ch[i];
+		if (!ch->target)
+			break;
+		ret += srp_new_cm_id(ch);
 	}
+	for (i = 0; i < target->ch_count; i++) {
+		ch = &target->ch[i];
+		if (!ch->target)
+			break;
+		for (j = 0; j < target->req_ring_size; ++j) {
+			struct srp_request *req = &ch->req_ring[j];
 
-	/*
-	 * Whether or not creating a new CM ID succeeded, create a new
-	 * QP. This guarantees that all callback functions for the old QP have
-	 * finished before any send requests are posted on the new QP.
-	 */
-	ret += srp_create_target_ib(target);
-
-	INIT_LIST_HEAD(&target->free_tx);
-	for (i = 0; i < target->queue_size; ++i)
-		list_add(&target->tx_ring[i]->list, &target->free_tx);
+			srp_finish_req(ch, req, NULL, DID_RESET << 16);
+		}
+	}
+	for (i = 0; i < target->ch_count; i++) {
+		ch = &target->ch[i];
+		if (!ch->target)
+			break;
+		/*
+		 * Whether or not creating a new CM ID succeeded, create a new
+		 * QP. This guarantees that all completion callback function
+		 * invocations have finished before request resetting starts.
+		 */
+		ret += srp_create_ch_ib(ch);
 
-	if (ret == 0)
-		ret = srp_connect_target(target);
+		INIT_LIST_HEAD(&ch->free_tx);
+		for (j = 0; j < target->queue_size; ++j)
+			list_add(&ch->tx_ring[j]->list, &ch->free_tx);
+	}
+	for (i = 0; i < target->ch_count; i++) {
+		ch = &target->ch[i];
+		if (ret || !ch->target) {
+			if (i > 1)
+				ret = 0;
+			break;
+		}
+		ret = srp_connect_ch(ch, multich);
+		multich = true;
+	}
 
 	if (ret == 0)
 		shost_printk(KERN_INFO, target->scsi_host,
@@ -1165,12 +1275,12 @@ static void srp_map_desc(struct srp_map_state *state, dma_addr_t dma_addr,
 }
 
 static int srp_map_finish_fmr(struct srp_map_state *state,
-			      struct srp_target_port *target)
+			      struct srp_rdma_ch *ch)
 {
 	struct ib_pool_fmr *fmr;
 	u64 io_addr = 0;
 
-	fmr = ib_fmr_pool_map_phys(target->fmr_pool, state->pages,
+	fmr = ib_fmr_pool_map_phys(ch->fmr_pool, state->pages,
 				   state->npages, io_addr);
 	if (IS_ERR(fmr))
 		return PTR_ERR(fmr);
@@ -1184,15 +1294,16 @@ static int srp_map_finish_fmr(struct srp_map_state *state,
 }
 
 static int srp_map_finish_fr(struct srp_map_state *state,
-			     struct srp_target_port *target)
+			     struct srp_rdma_ch *ch)
 {
+	struct srp_target_port *target = ch->target;
 	struct srp_device *dev = target->srp_host->srp_dev;
 	struct ib_send_wr *bad_wr;
 	struct ib_send_wr wr;
 	struct srp_fr_desc *desc;
 	u32 rkey;
 
-	desc = srp_fr_pool_get(target->fr_pool);
+	desc = srp_fr_pool_get(ch->fr_pool);
 	if (!desc)
 		return -ENOMEM;
 
@@ -1221,12 +1332,13 @@ static int srp_map_finish_fr(struct srp_map_state *state,
 	srp_map_desc(state, state->base_dma_addr, state->dma_len,
 		     desc->mr->rkey);
 
-	return ib_post_send(target->qp, &wr, &bad_wr);
+	return ib_post_send(ch->qp, &wr, &bad_wr);
 }
 
 static int srp_finish_mapping(struct srp_map_state *state,
-			      struct srp_target_port *target)
+			      struct srp_rdma_ch *ch)
 {
+	struct srp_target_port *target = ch->target;
 	int ret = 0;
 
 	if (state->npages == 0)
@@ -1237,8 +1349,8 @@ static int srp_finish_mapping(struct srp_map_state *state,
 			     target->rkey);
 	else
 		ret = target->srp_host->srp_dev->use_fast_reg ?
-			srp_map_finish_fr(state, target) :
-			srp_map_finish_fmr(state, target);
+			srp_map_finish_fr(state, ch) :
+			srp_map_finish_fmr(state, ch);
 
 	if (ret == 0) {
 		state->npages = 0;
@@ -1258,10 +1370,11 @@ static void srp_map_update_start(struct srp_map_state *state,
 }
 
 static int srp_map_sg_entry(struct srp_map_state *state,
-			    struct srp_target_port *target,
+			    struct srp_rdma_ch *ch,
 			    struct scatterlist *sg, int sg_index,
 			    bool use_mr)
 {
+	struct srp_target_port *target = ch->target;
 	struct srp_device *dev = target->srp_host->srp_dev;
 	struct ib_device *ibdev = dev->dev;
 	dma_addr_t dma_addr = ib_sg_dma_address(ibdev, sg);
@@ -1290,7 +1403,7 @@ static int srp_map_sg_entry(struct srp_map_state *state,
 	 */
 	if ((!dev->use_fast_reg && dma_addr & ~dev->mr_page_mask) ||
 	    dma_len > dev->mr_max_size) {
-		ret = srp_finish_mapping(state, target);
+		ret = srp_finish_mapping(state, ch);
 		if (ret)
 			return ret;
 
@@ -1311,7 +1424,7 @@ static int srp_map_sg_entry(struct srp_map_state *state,
 	while (dma_len) {
 		unsigned offset = dma_addr & ~dev->mr_page_mask;
 		if (state->npages == dev->max_pages_per_mr || offset != 0) {
-			ret = srp_finish_mapping(state, target);
+			ret = srp_finish_mapping(state, ch);
 			if (ret)
 				return ret;
 
@@ -1335,17 +1448,18 @@ static int srp_map_sg_entry(struct srp_map_state *state,
 	 */
 	ret = 0;
 	if (len != dev->mr_page_size) {
-		ret = srp_finish_mapping(state, target);
+		ret = srp_finish_mapping(state, ch);
 		if (!ret)
 			srp_map_update_start(state, NULL, 0, 0);
 	}
 	return ret;
 }
 
-static int srp_map_sg(struct srp_map_state *state,
-		      struct srp_target_port *target, struct srp_request *req,
-		      struct scatterlist *scat, int count)
+static int srp_map_sg(struct srp_map_state *state, struct srp_rdma_ch *ch,
+		      struct srp_request *req, struct scatterlist *scat,
+		      int count)
 {
+	struct srp_target_port *target = ch->target;
 	struct srp_device *dev = target->srp_host->srp_dev;
 	struct ib_device *ibdev = dev->dev;
 	struct scatterlist *sg;
@@ -1356,14 +1470,14 @@ static int srp_map_sg(struct srp_map_state *state,
 	state->pages	= req->map_page;
 	if (dev->use_fast_reg) {
 		state->next_fr = req->fr_list;
-		use_mr = !!target->fr_pool;
+		use_mr = !!ch->fr_pool;
 	} else {
 		state->next_fmr = req->fmr_list;
-		use_mr = !!target->fmr_pool;
+		use_mr = !!ch->fmr_pool;
 	}
 
 	for_each_sg(scat, sg, count, i) {
-		if (srp_map_sg_entry(state, target, sg, i, use_mr)) {
+		if (srp_map_sg_entry(state, ch, sg, i, use_mr)) {
 			/*
 			 * Memory registration failed, so backtrack to the
 			 * first unmapped entry and continue on without using
@@ -1385,7 +1499,7 @@ backtrack:
 		}
 	}
 
-	if (use_mr && srp_finish_mapping(state, target))
+	if (use_mr && srp_finish_mapping(state, ch))
 		goto backtrack;
 
 	req->nmdesc = state->nmdesc;
@@ -1393,9 +1507,10 @@ backtrack:
 	return 0;
 }
 
-static int srp_map_data(struct scsi_cmnd *scmnd, struct srp_target_port *target,
+static int srp_map_data(struct scsi_cmnd *scmnd, struct srp_rdma_ch *ch,
 			struct srp_request *req)
 {
+	struct srp_target_port *target = ch->target;
 	struct scatterlist *scat;
 	struct srp_cmd *cmd = req->cmd->buf;
 	int len, nents, count;
@@ -1457,7 +1572,7 @@ static int srp_map_data(struct scsi_cmnd *scmnd, struct srp_target_port *target,
 				   target->indirect_size, DMA_TO_DEVICE);
 
 	memset(&state, 0, sizeof(state));
-	srp_map_sg(&state, target, req, scat, count);
+	srp_map_sg(&state, ch, req, scat, count);
 
 	/* We've mapped the request, now pull as much of the indirect
 	 * descriptor table as we can into the command buffer. If this
@@ -1518,20 +1633,20 @@ map_complete:
 /*
  * Return an IU and possible credit to the free pool
  */
-static void srp_put_tx_iu(struct srp_target_port *target, struct srp_iu *iu,
+static void srp_put_tx_iu(struct srp_rdma_ch *ch, struct srp_iu *iu,
 			  enum srp_iu_type iu_type)
 {
 	unsigned long flags;
 
-	spin_lock_irqsave(&target->lock, flags);
-	list_add(&iu->list, &target->free_tx);
+	spin_lock_irqsave(&ch->lock, flags);
+	list_add(&iu->list, &ch->free_tx);
 	if (iu_type != SRP_IU_RSP)
-		++target->req_lim;
-	spin_unlock_irqrestore(&target->lock, flags);
+		++ch->req_lim;
+	spin_unlock_irqrestore(&ch->lock, flags);
 }
 
 /*
- * Must be called with target->lock held to protect req_lim and free_tx.
+ * Must be called with ch->lock held to protect req_lim and free_tx.
  * If IU is not sent, it must be returned using srp_put_tx_iu().
  *
  * Note:
@@ -1543,35 +1658,36 @@ static void srp_put_tx_iu(struct srp_target_port *target, struct srp_iu *iu,
  * - SRP_IU_RSP: 1, since a conforming SRP target never sends more than
  *   one unanswered SRP request to an initiator.
  */
-static struct srp_iu *__srp_get_tx_iu(struct srp_target_port *target,
+static struct srp_iu *__srp_get_tx_iu(struct srp_rdma_ch *ch,
 				      enum srp_iu_type iu_type)
 {
+	struct srp_target_port *target = ch->target;
 	s32 rsv = (iu_type == SRP_IU_TSK_MGMT) ? 0 : SRP_TSK_MGMT_SQ_SIZE;
 	struct srp_iu *iu;
 
-	srp_send_completion(target->send_cq, target);
+	srp_send_completion(ch->send_cq, ch);
 
-	if (list_empty(&target->free_tx))
+	if (list_empty(&ch->free_tx))
 		return NULL;
 
 	/* Initiator responses to target requests do not consume credits */
 	if (iu_type != SRP_IU_RSP) {
-		if (target->req_lim <= rsv) {
+		if (ch->req_lim <= rsv) {
 			++target->zero_req_lim;
 			return NULL;
 		}
 
-		--target->req_lim;
+		--ch->req_lim;
 	}
 
-	iu = list_first_entry(&target->free_tx, struct srp_iu, list);
+	iu = list_first_entry(&ch->free_tx, struct srp_iu, list);
 	list_del(&iu->list);
 	return iu;
 }
 
-static int srp_post_send(struct srp_target_port *target,
-			 struct srp_iu *iu, int len)
+static int srp_post_send(struct srp_rdma_ch *ch, struct srp_iu *iu, int len)
 {
+	struct srp_target_port *target = ch->target;
 	struct ib_sge list;
 	struct ib_send_wr wr, *bad_wr;
 
@@ -1586,11 +1702,12 @@ static int srp_post_send(struct srp_target_port *target,
 	wr.opcode     = IB_WR_SEND;
 	wr.send_flags = IB_SEND_SIGNALED;
 
-	return ib_post_send(target->qp, &wr, &bad_wr);
+	return ib_post_send(ch->qp, &wr, &bad_wr);
 }
 
-static int srp_post_recv(struct srp_target_port *target, struct srp_iu *iu)
+static int srp_post_recv(struct srp_rdma_ch *ch, struct srp_iu *iu)
 {
+	struct srp_target_port *target = ch->target;
 	struct ib_recv_wr wr, *bad_wr;
 	struct ib_sge list;
 
@@ -1603,35 +1720,39 @@ static int srp_post_recv(struct srp_target_port *target, struct srp_iu *iu)
 	wr.sg_list  = &list;
 	wr.num_sge  = 1;
 
-	return ib_post_recv(target->qp, &wr, &bad_wr);
+	return ib_post_recv(ch->qp, &wr, &bad_wr);
 }
 
-static void srp_process_rsp(struct srp_target_port *target, struct srp_rsp *rsp)
+static void srp_process_rsp(struct srp_rdma_ch *ch, struct srp_rsp *rsp)
 {
+	struct srp_target_port *target = ch->target;
 	struct srp_request *req;
 	struct scsi_cmnd *scmnd;
 	unsigned long flags;
 
 	if (unlikely(rsp->tag & SRP_TAG_TSK_MGMT)) {
-		spin_lock_irqsave(&target->lock, flags);
-		target->req_lim += be32_to_cpu(rsp->req_lim_delta);
-		spin_unlock_irqrestore(&target->lock, flags);
+		spin_lock_irqsave(&ch->lock, flags);
+		ch->req_lim += be32_to_cpu(rsp->req_lim_delta);
+		spin_unlock_irqrestore(&ch->lock, flags);
 
-		target->tsk_mgmt_status = -1;
+		ch->tsk_mgmt_status = -1;
 		if (be32_to_cpu(rsp->resp_data_len) >= 4)
-			target->tsk_mgmt_status = rsp->data[3];
-		complete(&target->tsk_mgmt_done);
+			ch->tsk_mgmt_status = rsp->data[3];
+		complete(&ch->tsk_mgmt_done);
 	} else {
-		req = &target->req_ring[rsp->tag];
-		scmnd = srp_claim_req(target, req, NULL, NULL);
+		scmnd = scsi_host_find_tag(target->scsi_host, rsp->tag);
+		if (scmnd) {
+			req = (void *)scmnd->host_scribble;
+			scmnd = srp_claim_req(ch, req, NULL, scmnd);
+		}
 		if (!scmnd) {
 			shost_printk(KERN_ERR, target->scsi_host,
-				     "Null scmnd for RSP w/tag %016llx\n",
-				     (unsigned long long) rsp->tag);
+				     "Null scmnd for RSP w/tag %#016llx received on ch %td / QP %#x\n",
+				     rsp->tag, ch - target->ch, ch->qp->qp_num);
 
-			spin_lock_irqsave(&target->lock, flags);
-			target->req_lim += be32_to_cpu(rsp->req_lim_delta);
-			spin_unlock_irqrestore(&target->lock, flags);
+			spin_lock_irqsave(&ch->lock, flags);
+			ch->req_lim += be32_to_cpu(rsp->req_lim_delta);
+			spin_unlock_irqrestore(&ch->lock, flags);
 
 			return;
 		}
@@ -1653,7 +1774,7 @@ static void srp_process_rsp(struct srp_target_port *target, struct srp_rsp *rsp)
 		else if (unlikely(rsp->flags & SRP_RSP_FLAG_DOOVER))
 			scsi_set_resid(scmnd, -be32_to_cpu(rsp->data_out_res_cnt));
 
-		srp_free_req(target, req, scmnd,
+		srp_free_req(ch, req, scmnd,
 			     be32_to_cpu(rsp->req_lim_delta));
 
 		scmnd->host_scribble = NULL;
@@ -1661,18 +1782,19 @@ static void srp_process_rsp(struct srp_target_port *target, struct srp_rsp *rsp)
 	}
 }
 
-static int srp_response_common(struct srp_target_port *target, s32 req_delta,
+static int srp_response_common(struct srp_rdma_ch *ch, s32 req_delta,
 			       void *rsp, int len)
 {
+	struct srp_target_port *target = ch->target;
 	struct ib_device *dev = target->srp_host->srp_dev->dev;
 	unsigned long flags;
 	struct srp_iu *iu;
 	int err;
 
-	spin_lock_irqsave(&target->lock, flags);
-	target->req_lim += req_delta;
-	iu = __srp_get_tx_iu(target, SRP_IU_RSP);
-	spin_unlock_irqrestore(&target->lock, flags);
+	spin_lock_irqsave(&ch->lock, flags);
+	ch->req_lim += req_delta;
+	iu = __srp_get_tx_iu(ch, SRP_IU_RSP);
+	spin_unlock_irqrestore(&ch->lock, flags);
 
 	if (!iu) {
 		shost_printk(KERN_ERR, target->scsi_host, PFX
@@ -1684,17 +1806,17 @@ static int srp_response_common(struct srp_target_port *target, s32 req_delta,
 	memcpy(iu->buf, rsp, len);
 	ib_dma_sync_single_for_device(dev, iu->dma, len, DMA_TO_DEVICE);
 
-	err = srp_post_send(target, iu, len);
+	err = srp_post_send(ch, iu, len);
 	if (err) {
 		shost_printk(KERN_ERR, target->scsi_host, PFX
 			     "unable to post response: %d\n", err);
-		srp_put_tx_iu(target, iu, SRP_IU_RSP);
+		srp_put_tx_iu(ch, iu, SRP_IU_RSP);
 	}
 
 	return err;
 }
 
-static void srp_process_cred_req(struct srp_target_port *target,
+static void srp_process_cred_req(struct srp_rdma_ch *ch,
 				 struct srp_cred_req *req)
 {
 	struct srp_cred_rsp rsp = {
@@ -1703,14 +1825,15 @@ static void srp_process_cred_req(struct srp_target_port *target,
 	};
 	s32 delta = be32_to_cpu(req->req_lim_delta);
 
-	if (srp_response_common(target, delta, &rsp, sizeof rsp))
-		shost_printk(KERN_ERR, target->scsi_host, PFX
+	if (srp_response_common(ch, delta, &rsp, sizeof(rsp)))
+		shost_printk(KERN_ERR, ch->target->scsi_host, PFX
 			     "problems processing SRP_CRED_REQ\n");
 }
 
-static void srp_process_aer_req(struct srp_target_port *target,
+static void srp_process_aer_req(struct srp_rdma_ch *ch,
 				struct srp_aer_req *req)
 {
+	struct srp_target_port *target = ch->target;
 	struct srp_aer_rsp rsp = {
 		.opcode = SRP_AER_RSP,
 		.tag = req->tag,
@@ -1720,19 +1843,20 @@ static void srp_process_aer_req(struct srp_target_port *target,
 	shost_printk(KERN_ERR, target->scsi_host, PFX
 		     "ignoring AER for LUN %llu\n", be64_to_cpu(req->lun));
 
-	if (srp_response_common(target, delta, &rsp, sizeof rsp))
+	if (srp_response_common(ch, delta, &rsp, sizeof(rsp)))
 		shost_printk(KERN_ERR, target->scsi_host, PFX
 			     "problems processing SRP_AER_REQ\n");
 }
 
-static void srp_handle_recv(struct srp_target_port *target, struct ib_wc *wc)
+static void srp_handle_recv(struct srp_rdma_ch *ch, struct ib_wc *wc)
 {
+	struct srp_target_port *target = ch->target;
 	struct ib_device *dev = target->srp_host->srp_dev->dev;
 	struct srp_iu *iu = (struct srp_iu *) (uintptr_t) wc->wr_id;
 	int res;
 	u8 opcode;
 
-	ib_dma_sync_single_for_cpu(dev, iu->dma, target->max_ti_iu_len,
+	ib_dma_sync_single_for_cpu(dev, iu->dma, ch->max_ti_iu_len,
 				   DMA_FROM_DEVICE);
 
 	opcode = *(u8 *) iu->buf;
@@ -1746,15 +1870,15 @@ static void srp_handle_recv(struct srp_target_port *target, struct ib_wc *wc)
 
 	switch (opcode) {
 	case SRP_RSP:
-		srp_process_rsp(target, iu->buf);
+		srp_process_rsp(ch, iu->buf);
 		break;
 
 	case SRP_CRED_REQ:
-		srp_process_cred_req(target, iu->buf);
+		srp_process_cred_req(ch, iu->buf);
 		break;
 
 	case SRP_AER_REQ:
-		srp_process_aer_req(target, iu->buf);
+		srp_process_aer_req(ch, iu->buf);
 		break;
 
 	case SRP_T_LOGOUT:
@@ -1769,10 +1893,10 @@ static void srp_handle_recv(struct srp_target_port *target, struct ib_wc *wc)
 		break;
 	}
 
-	ib_dma_sync_single_for_device(dev, iu->dma, target->max_ti_iu_len,
+	ib_dma_sync_single_for_device(dev, iu->dma, ch->max_ti_iu_len,
 				      DMA_FROM_DEVICE);
 
-	res = srp_post_recv(target, iu);
+	res = srp_post_recv(ch, iu);
 	if (res != 0)
 		shost_printk(KERN_ERR, target->scsi_host,
 			     PFX "Recv failed with error code %d\n", res);
@@ -1795,8 +1919,15 @@ static void srp_tl_err_work(struct work_struct *work)
 }
 
 static void srp_handle_qp_err(u64 wr_id, enum ib_wc_status wc_status,
-			      bool send_err, struct srp_target_port *target)
+			      bool send_err, struct srp_rdma_ch *ch)
 {
+	struct srp_target_port *target = ch->target;
+
+	if (wr_id == SRP_LAST_WR_ID) {
+		complete(&ch->done);
+		return;
+	}
+
 	if (target->connected && !target->qp_in_error) {
 		if (wr_id & LOCAL_INV_WR_ID_MASK) {
 			shost_printk(KERN_ERR, target->scsi_host, PFX
@@ -1817,33 +1948,33 @@ static void srp_handle_qp_err(u64 wr_id, enum ib_wc_status wc_status,
 	target->qp_in_error = true;
 }
 
-static void srp_recv_completion(struct ib_cq *cq, void *target_ptr)
+static void srp_recv_completion(struct ib_cq *cq, void *ch_ptr)
 {
-	struct srp_target_port *target = target_ptr;
+	struct srp_rdma_ch *ch = ch_ptr;
 	struct ib_wc wc;
 
 	ib_req_notify_cq(cq, IB_CQ_NEXT_COMP);
 	while (ib_poll_cq(cq, 1, &wc) > 0) {
 		if (likely(wc.status == IB_WC_SUCCESS)) {
-			srp_handle_recv(target, &wc);
+			srp_handle_recv(ch, &wc);
 		} else {
-			srp_handle_qp_err(wc.wr_id, wc.status, false, target);
+			srp_handle_qp_err(wc.wr_id, wc.status, false, ch);
 		}
 	}
 }
 
-static void srp_send_completion(struct ib_cq *cq, void *target_ptr)
+static void srp_send_completion(struct ib_cq *cq, void *ch_ptr)
 {
-	struct srp_target_port *target = target_ptr;
+	struct srp_rdma_ch *ch = ch_ptr;
 	struct ib_wc wc;
 	struct srp_iu *iu;
 
 	while (ib_poll_cq(cq, 1, &wc) > 0) {
 		if (likely(wc.status == IB_WC_SUCCESS)) {
 			iu = (struct srp_iu *) (uintptr_t) wc.wr_id;
-			list_add(&iu->list, &target->free_tx);
+			list_add(&iu->list, &ch->free_tx);
 		} else {
-			srp_handle_qp_err(wc.wr_id, wc.status, true, target);
+			srp_handle_qp_err(wc.wr_id, wc.status, true, ch);
 		}
 	}
 }
@@ -1852,11 +1983,14 @@ static int srp_queuecommand(struct Scsi_Host *shost, struct scsi_cmnd *scmnd)
 {
 	struct srp_target_port *target = host_to_target(shost);
 	struct srp_rport *rport = target->rport;
+	struct srp_rdma_ch *ch;
 	struct srp_request *req;
 	struct srp_iu *iu;
 	struct srp_cmd *cmd;
 	struct ib_device *dev;
 	unsigned long flags;
+	u32 tag;
+	u16 idx;
 	int len, ret;
 	const bool in_scsi_eh = !in_interrupt() && current == shost->ehandler;
 
@@ -1873,15 +2007,22 @@ static int srp_queuecommand(struct Scsi_Host *shost, struct scsi_cmnd *scmnd)
 	if (unlikely(scmnd->result))
 		goto err;
 
-	spin_lock_irqsave(&target->lock, flags);
-	iu = __srp_get_tx_iu(target, SRP_IU_CMD);
-	if (!iu)
-		goto err_unlock;
+	WARN_ON_ONCE(scmnd->request->tag < 0);
+	tag = blk_mq_unique_tag(scmnd->request);
+	ch = &target->ch[blk_mq_unique_tag_to_hwq(tag)];
+	idx = blk_mq_unique_tag_to_tag(tag);
+	WARN_ONCE(idx >= target->req_ring_size, "%s: tag %#x: idx %d >= %d\n",
+		  dev_name(&shost->shost_gendev), tag, idx,
+		  target->req_ring_size);
 
-	req = list_first_entry(&target->free_reqs, struct srp_request, list);
-	list_del(&req->list);
-	spin_unlock_irqrestore(&target->lock, flags);
+	spin_lock_irqsave(&ch->lock, flags);
+	iu = __srp_get_tx_iu(ch, SRP_IU_CMD);
+	spin_unlock_irqrestore(&ch->lock, flags);
 
+	if (!iu)
+		goto err;
+
+	req = &ch->req_ring[idx];
 	dev = target->srp_host->srp_dev->dev;
 	ib_dma_sync_single_for_cpu(dev, iu->dma, target->max_iu_len,
 				   DMA_TO_DEVICE);
@@ -1893,13 +2034,13 @@ static int srp_queuecommand(struct Scsi_Host *shost, struct scsi_cmnd *scmnd)
 
 	cmd->opcode = SRP_CMD;
 	cmd->lun    = cpu_to_be64((u64) scmnd->device->lun << 48);
-	cmd->tag    = req->index;
+	cmd->tag    = tag;
 	memcpy(cmd->cdb, scmnd->cmnd, scmnd->cmd_len);
 
 	req->scmnd    = scmnd;
 	req->cmd      = iu;
 
-	len = srp_map_data(scmnd, target, req);
+	len = srp_map_data(scmnd, ch, req);
 	if (len < 0) {
 		shost_printk(KERN_ERR, target->scsi_host,
 			     PFX "Failed to map data (%d)\n", len);
@@ -1917,7 +2058,7 @@ static int srp_queuecommand(struct Scsi_Host *shost, struct scsi_cmnd *scmnd)
 	ib_dma_sync_single_for_device(dev, iu->dma, target->max_iu_len,
 				      DMA_TO_DEVICE);
 
-	if (srp_post_send(target, iu, len)) {
+	if (srp_post_send(ch, iu, len)) {
 		shost_printk(KERN_ERR, target->scsi_host, PFX "Send failed\n");
 		goto err_unmap;
 	}
@@ -1931,10 +2072,10 @@ unlock_rport:
 	return ret;
 
 err_unmap:
-	srp_unmap_data(scmnd, target, req);
+	srp_unmap_data(scmnd, ch, req);
 
 err_iu:
-	srp_put_tx_iu(target, iu, SRP_IU_CMD);
+	srp_put_tx_iu(ch, iu, SRP_IU_CMD);
 
 	/*
 	 * Avoid that the loops that iterate over the request ring can
@@ -1942,12 +2083,6 @@ err_iu:
 	 */
 	req->scmnd = NULL;
 
-	spin_lock_irqsave(&target->lock, flags);
-	list_add(&req->list, &target->free_reqs);
-
-err_unlock:
-	spin_unlock_irqrestore(&target->lock, flags);
-
 err:
 	if (scmnd->result) {
 		scmnd->scsi_done(scmnd);
@@ -1961,53 +2096,54 @@ err:
 
 /*
  * Note: the resources allocated in this function are freed in
- * srp_free_target_ib().
+ * srp_free_ch_ib().
  */
-static int srp_alloc_iu_bufs(struct srp_target_port *target)
+static int srp_alloc_iu_bufs(struct srp_rdma_ch *ch)
 {
+	struct srp_target_port *target = ch->target;
 	int i;
 
-	target->rx_ring = kzalloc(target->queue_size * sizeof(*target->rx_ring),
-				  GFP_KERNEL);
-	if (!target->rx_ring)
+	ch->rx_ring = kcalloc(target->queue_size, sizeof(*ch->rx_ring),
+			      GFP_KERNEL);
+	if (!ch->rx_ring)
 		goto err_no_ring;
-	target->tx_ring = kzalloc(target->queue_size * sizeof(*target->tx_ring),
-				  GFP_KERNEL);
-	if (!target->tx_ring)
+	ch->tx_ring = kcalloc(target->queue_size, sizeof(*ch->tx_ring),
+			      GFP_KERNEL);
+	if (!ch->tx_ring)
 		goto err_no_ring;
 
 	for (i = 0; i < target->queue_size; ++i) {
-		target->rx_ring[i] = srp_alloc_iu(target->srp_host,
-						  target->max_ti_iu_len,
-						  GFP_KERNEL, DMA_FROM_DEVICE);
-		if (!target->rx_ring[i])
+		ch->rx_ring[i] = srp_alloc_iu(target->srp_host,
+					      ch->max_ti_iu_len,
+					      GFP_KERNEL, DMA_FROM_DEVICE);
+		if (!ch->rx_ring[i])
 			goto err;
 	}
 
 	for (i = 0; i < target->queue_size; ++i) {
-		target->tx_ring[i] = srp_alloc_iu(target->srp_host,
-						  target->max_iu_len,
-						  GFP_KERNEL, DMA_TO_DEVICE);
-		if (!target->tx_ring[i])
+		ch->tx_ring[i] = srp_alloc_iu(target->srp_host,
+					      target->max_iu_len,
+					      GFP_KERNEL, DMA_TO_DEVICE);
+		if (!ch->tx_ring[i])
 			goto err;
 
-		list_add(&target->tx_ring[i]->list, &target->free_tx);
+		list_add(&ch->tx_ring[i]->list, &ch->free_tx);
 	}
 
 	return 0;
 
 err:
 	for (i = 0; i < target->queue_size; ++i) {
-		srp_free_iu(target->srp_host, target->rx_ring[i]);
-		srp_free_iu(target->srp_host, target->tx_ring[i]);
+		srp_free_iu(target->srp_host, ch->rx_ring[i]);
+		srp_free_iu(target->srp_host, ch->tx_ring[i]);
 	}
 
 
 err_no_ring:
-	kfree(target->tx_ring);
-	target->tx_ring = NULL;
-	kfree(target->rx_ring);
-	target->rx_ring = NULL;
+	kfree(ch->tx_ring);
+	ch->tx_ring = NULL;
+	kfree(ch->rx_ring);
+	ch->rx_ring = NULL;
 
 	return -ENOMEM;
 }
@@ -2041,23 +2177,24 @@ static uint32_t srp_compute_rq_tmo(struct ib_qp_attr *qp_attr, int attr_mask)
 
 static void srp_cm_rep_handler(struct ib_cm_id *cm_id,
 			       struct srp_login_rsp *lrsp,
-			       struct srp_target_port *target)
+			       struct srp_rdma_ch *ch)
 {
+	struct srp_target_port *target = ch->target;
 	struct ib_qp_attr *qp_attr = NULL;
 	int attr_mask = 0;
 	int ret;
 	int i;
 
 	if (lrsp->opcode == SRP_LOGIN_RSP) {
-		target->max_ti_iu_len = be32_to_cpu(lrsp->max_ti_iu_len);
-		target->req_lim       = be32_to_cpu(lrsp->req_lim_delta);
+		ch->max_ti_iu_len = be32_to_cpu(lrsp->max_ti_iu_len);
+		ch->req_lim       = be32_to_cpu(lrsp->req_lim_delta);
 
 		/*
 		 * Reserve credits for task management so we don't
 		 * bounce requests back to the SCSI mid-layer.
 		 */
 		target->scsi_host->can_queue
-			= min(target->req_lim - SRP_TSK_MGMT_SQ_SIZE,
+			= min(ch->req_lim - SRP_TSK_MGMT_SQ_SIZE,
 			      target->scsi_host->can_queue);
 		target->scsi_host->cmd_per_lun
 			= min_t(int, target->scsi_host->can_queue,
@@ -2069,8 +2206,8 @@ static void srp_cm_rep_handler(struct ib_cm_id *cm_id,
 		goto error;
 	}
 
-	if (!target->rx_ring) {
-		ret = srp_alloc_iu_bufs(target);
+	if (!ch->rx_ring) {
+		ret = srp_alloc_iu_bufs(ch);
 		if (ret)
 			goto error;
 	}
@@ -2085,13 +2222,14 @@ static void srp_cm_rep_handler(struct ib_cm_id *cm_id,
 	if (ret)
 		goto error_free;
 
-	ret = ib_modify_qp(target->qp, qp_attr, attr_mask);
+	ret = ib_modify_qp(ch->qp, qp_attr, attr_mask);
 	if (ret)
 		goto error_free;
 
 	for (i = 0; i < target->queue_size; i++) {
-		struct srp_iu *iu = target->rx_ring[i];
-		ret = srp_post_recv(target, iu);
+		struct srp_iu *iu = ch->rx_ring[i];
+
+		ret = srp_post_recv(ch, iu);
 		if (ret)
 			goto error_free;
 	}
@@ -2103,7 +2241,7 @@ static void srp_cm_rep_handler(struct ib_cm_id *cm_id,
 
 	target->rq_tmo_jiffies = srp_compute_rq_tmo(qp_attr, attr_mask);
 
-	ret = ib_modify_qp(target->qp, qp_attr, attr_mask);
+	ret = ib_modify_qp(ch->qp, qp_attr, attr_mask);
 	if (ret)
 		goto error_free;
 
@@ -2113,13 +2251,14 @@ error_free:
 	kfree(qp_attr);
 
 error:
-	target->status = ret;
+	ch->status = ret;
 }
 
 static void srp_cm_rej_handler(struct ib_cm_id *cm_id,
 			       struct ib_cm_event *event,
-			       struct srp_target_port *target)
+			       struct srp_rdma_ch *ch)
 {
+	struct srp_target_port *target = ch->target;
 	struct Scsi_Host *shost = target->scsi_host;
 	struct ib_class_port_info *cpi;
 	int opcode;
@@ -2127,12 +2266,12 @@ static void srp_cm_rej_handler(struct ib_cm_id *cm_id,
 	switch (event->param.rej_rcvd.reason) {
 	case IB_CM_REJ_PORT_CM_REDIRECT:
 		cpi = event->param.rej_rcvd.ari;
-		target->path.dlid = cpi->redirect_lid;
-		target->path.pkey = cpi->redirect_pkey;
+		ch->path.dlid = cpi->redirect_lid;
+		ch->path.pkey = cpi->redirect_pkey;
 		cm_id->remote_cm_qpn = be32_to_cpu(cpi->redirect_qp) & 0x00ffffff;
-		memcpy(target->path.dgid.raw, cpi->redirect_gid, 16);
+		memcpy(ch->path.dgid.raw, cpi->redirect_gid, 16);
 
-		target->status = target->path.dlid ?
+		ch->status = ch->path.dlid ?
 			SRP_DLID_REDIRECT : SRP_PORT_REDIRECT;
 		break;
 
@@ -2143,26 +2282,26 @@ static void srp_cm_rej_handler(struct ib_cm_id *cm_id,
 			 * reject reason code 25 when they mean 24
 			 * (port redirect).
 			 */
-			memcpy(target->path.dgid.raw,
+			memcpy(ch->path.dgid.raw,
 			       event->param.rej_rcvd.ari, 16);
 
 			shost_printk(KERN_DEBUG, shost,
 				     PFX "Topspin/Cisco redirect to target port GID %016llx%016llx\n",
-				     (unsigned long long) be64_to_cpu(target->path.dgid.global.subnet_prefix),
-				     (unsigned long long) be64_to_cpu(target->path.dgid.global.interface_id));
+				     be64_to_cpu(ch->path.dgid.global.subnet_prefix),
+				     be64_to_cpu(ch->path.dgid.global.interface_id));
 
-			target->status = SRP_PORT_REDIRECT;
+			ch->status = SRP_PORT_REDIRECT;
 		} else {
 			shost_printk(KERN_WARNING, shost,
 				     "  REJ reason: IB_CM_REJ_PORT_REDIRECT\n");
-			target->status = -ECONNRESET;
+			ch->status = -ECONNRESET;
 		}
 		break;
 
 	case IB_CM_REJ_DUPLICATE_LOCAL_COMM_ID:
 		shost_printk(KERN_WARNING, shost,
 			    "  REJ reason: IB_CM_REJ_DUPLICATE_LOCAL_COMM_ID\n");
-		target->status = -ECONNRESET;
+		ch->status = -ECONNRESET;
 		break;
 
 	case IB_CM_REJ_CONSUMER_DEFINED:
@@ -2177,30 +2316,31 @@ static void srp_cm_rej_handler(struct ib_cm_id *cm_id,
 			else
 				shost_printk(KERN_WARNING, shost, PFX
 					     "SRP LOGIN from %pI6 to %pI6 REJECTED, reason 0x%08x\n",
-					     target->path.sgid.raw,
-					     target->orig_dgid, reason);
+					     target->sgid.raw,
+					     target->orig_dgid.raw, reason);
 		} else
 			shost_printk(KERN_WARNING, shost,
 				     "  REJ reason: IB_CM_REJ_CONSUMER_DEFINED,"
 				     " opcode 0x%02x\n", opcode);
-		target->status = -ECONNRESET;
+		ch->status = -ECONNRESET;
 		break;
 
 	case IB_CM_REJ_STALE_CONN:
 		shost_printk(KERN_WARNING, shost, "  REJ reason: stale connection\n");
-		target->status = SRP_STALE_CONN;
+		ch->status = SRP_STALE_CONN;
 		break;
 
 	default:
 		shost_printk(KERN_WARNING, shost, "  REJ reason 0x%x\n",
 			     event->param.rej_rcvd.reason);
-		target->status = -ECONNRESET;
+		ch->status = -ECONNRESET;
 	}
 }
 
 static int srp_cm_handler(struct ib_cm_id *cm_id, struct ib_cm_event *event)
 {
-	struct srp_target_port *target = cm_id->context;
+	struct srp_rdma_ch *ch = cm_id->context;
+	struct srp_target_port *target = ch->target;
 	int comp = 0;
 
 	switch (event->event) {
@@ -2208,19 +2348,19 @@ static int srp_cm_handler(struct ib_cm_id *cm_id, struct ib_cm_event *event)
 		shost_printk(KERN_DEBUG, target->scsi_host,
 			     PFX "Sending CM REQ failed\n");
 		comp = 1;
-		target->status = -ECONNRESET;
+		ch->status = -ECONNRESET;
 		break;
 
 	case IB_CM_REP_RECEIVED:
 		comp = 1;
-		srp_cm_rep_handler(cm_id, event->private_data, target);
+		srp_cm_rep_handler(cm_id, event->private_data, ch);
 		break;
 
 	case IB_CM_REJ_RECEIVED:
 		shost_printk(KERN_DEBUG, target->scsi_host, PFX "REJ received\n");
 		comp = 1;
 
-		srp_cm_rej_handler(cm_id, event, target);
+		srp_cm_rej_handler(cm_id, event, ch);
 		break;
 
 	case IB_CM_DREQ_RECEIVED:
@@ -2238,7 +2378,7 @@ static int srp_cm_handler(struct ib_cm_id *cm_id, struct ib_cm_event *event)
 			     PFX "connection closed\n");
 		comp = 1;
 
-		target->status = 0;
+		ch->status = 0;
 		break;
 
 	case IB_CM_MRA_RECEIVED:
@@ -2253,65 +2393,30 @@ static int srp_cm_handler(struct ib_cm_id *cm_id, struct ib_cm_event *event)
 	}
 
 	if (comp)
-		complete(&target->done);
+		complete(&ch->done);
 
 	return 0;
 }
 
 /**
- * srp_change_queue_type - changing device queue tag type
- * @sdev: scsi device struct
- * @tag_type: requested tag type
- *
- * Returns queue tag type.
- */
-static int
-srp_change_queue_type(struct scsi_device *sdev, int tag_type)
-{
-	if (sdev->tagged_supported) {
-		scsi_set_tag_type(sdev, tag_type);
-		if (tag_type)
-			scsi_activate_tcq(sdev, sdev->queue_depth);
-		else
-			scsi_deactivate_tcq(sdev, sdev->queue_depth);
-	} else
-		tag_type = 0;
-
-	return tag_type;
-}
-
-/**
  * srp_change_queue_depth - setting device queue depth
  * @sdev: scsi device struct
  * @qdepth: requested queue depth
- * @reason: SCSI_QDEPTH_DEFAULT/SCSI_QDEPTH_QFULL/SCSI_QDEPTH_RAMP_UP
- * (see include/scsi/scsi_host.h for definition)
  *
  * Returns queue depth.
  */
 static int
-srp_change_queue_depth(struct scsi_device *sdev, int qdepth, int reason)
+srp_change_queue_depth(struct scsi_device *sdev, int qdepth)
 {
-	struct Scsi_Host *shost = sdev->host;
-	int max_depth;
-	if (reason == SCSI_QDEPTH_DEFAULT || reason == SCSI_QDEPTH_RAMP_UP) {
-		max_depth = shost->can_queue;
-		if (!sdev->tagged_supported)
-			max_depth = 1;
-		if (qdepth > max_depth)
-			qdepth = max_depth;
-		scsi_adjust_queue_depth(sdev, scsi_get_tag_type(sdev), qdepth);
-	} else if (reason == SCSI_QDEPTH_QFULL)
-		scsi_track_queue_full(sdev, qdepth);
-	else
-		return -EOPNOTSUPP;
-
-	return sdev->queue_depth;
+	if (!sdev->tagged_supported)
+		qdepth = 1;
+	return scsi_change_queue_depth(sdev, qdepth);
 }
 
-static int srp_send_tsk_mgmt(struct srp_target_port *target,
-			     u64 req_tag, unsigned int lun, u8 func)
+static int srp_send_tsk_mgmt(struct srp_rdma_ch *ch, u64 req_tag,
+			     unsigned int lun, u8 func)
 {
+	struct srp_target_port *target = ch->target;
 	struct srp_rport *rport = target->rport;
 	struct ib_device *dev = target->srp_host->srp_dev->dev;
 	struct srp_iu *iu;
@@ -2320,16 +2425,16 @@ static int srp_send_tsk_mgmt(struct srp_target_port *target,
 	if (!target->connected || target->qp_in_error)
 		return -1;
 
-	init_completion(&target->tsk_mgmt_done);
+	init_completion(&ch->tsk_mgmt_done);
 
 	/*
-	 * Lock the rport mutex to avoid that srp_create_target_ib() is
+	 * Lock the rport mutex to avoid that srp_create_ch_ib() is
 	 * invoked while a task management function is being sent.
 	 */
 	mutex_lock(&rport->mutex);
-	spin_lock_irq(&target->lock);
-	iu = __srp_get_tx_iu(target, SRP_IU_TSK_MGMT);
-	spin_unlock_irq(&target->lock);
+	spin_lock_irq(&ch->lock);
+	iu = __srp_get_tx_iu(ch, SRP_IU_TSK_MGMT);
+	spin_unlock_irq(&ch->lock);
 
 	if (!iu) {
 		mutex_unlock(&rport->mutex);
@@ -2350,15 +2455,15 @@ static int srp_send_tsk_mgmt(struct srp_target_port *target,
 
 	ib_dma_sync_single_for_device(dev, iu->dma, sizeof *tsk_mgmt,
 				      DMA_TO_DEVICE);
-	if (srp_post_send(target, iu, sizeof *tsk_mgmt)) {
-		srp_put_tx_iu(target, iu, SRP_IU_TSK_MGMT);
+	if (srp_post_send(ch, iu, sizeof(*tsk_mgmt))) {
+		srp_put_tx_iu(ch, iu, SRP_IU_TSK_MGMT);
 		mutex_unlock(&rport->mutex);
 
 		return -1;
 	}
 	mutex_unlock(&rport->mutex);
 
-	if (!wait_for_completion_timeout(&target->tsk_mgmt_done,
+	if (!wait_for_completion_timeout(&ch->tsk_mgmt_done,
 					 msecs_to_jiffies(SRP_ABORT_TIMEOUT_MS)))
 		return -1;
 
@@ -2369,20 +2474,32 @@ static int srp_abort(struct scsi_cmnd *scmnd)
 {
 	struct srp_target_port *target = host_to_target(scmnd->device->host);
 	struct srp_request *req = (struct srp_request *) scmnd->host_scribble;
+	u32 tag;
+	u16 ch_idx;
+	struct srp_rdma_ch *ch;
 	int ret;
 
 	shost_printk(KERN_ERR, target->scsi_host, "SRP abort called\n");
 
-	if (!req || !srp_claim_req(target, req, NULL, scmnd))
+	if (!req)
+		return SUCCESS;
+	tag = blk_mq_unique_tag(scmnd->request);
+	ch_idx = blk_mq_unique_tag_to_hwq(tag);
+	if (WARN_ON_ONCE(ch_idx >= target->ch_count))
 		return SUCCESS;
-	if (srp_send_tsk_mgmt(target, req->index, scmnd->device->lun,
+	ch = &target->ch[ch_idx];
+	if (!srp_claim_req(ch, req, NULL, scmnd))
+		return SUCCESS;
+	shost_printk(KERN_ERR, target->scsi_host,
+		     "Sending SRP abort for tag %#x\n", tag);
+	if (srp_send_tsk_mgmt(ch, tag, scmnd->device->lun,
 			      SRP_TSK_ABORT_TASK) == 0)
 		ret = SUCCESS;
 	else if (target->rport->state == SRP_RPORT_LOST)
 		ret = FAST_IO_FAIL;
 	else
 		ret = FAILED;
-	srp_free_req(target, req, scmnd, 0);
+	srp_free_req(ch, req, scmnd, 0);
 	scmnd->result = DID_ABORT << 16;
 	scmnd->scsi_done(scmnd);
 
@@ -2392,19 +2509,25 @@ static int srp_abort(struct scsi_cmnd *scmnd)
 static int srp_reset_device(struct scsi_cmnd *scmnd)
 {
 	struct srp_target_port *target = host_to_target(scmnd->device->host);
+	struct srp_rdma_ch *ch;
 	int i;
 
 	shost_printk(KERN_ERR, target->scsi_host, "SRP reset_device called\n");
 
-	if (srp_send_tsk_mgmt(target, SRP_TAG_NO_REQ, scmnd->device->lun,
+	ch = &target->ch[0];
+	if (srp_send_tsk_mgmt(ch, SRP_TAG_NO_REQ, scmnd->device->lun,
 			      SRP_TSK_LUN_RESET))
 		return FAILED;
-	if (target->tsk_mgmt_status)
+	if (ch->tsk_mgmt_status)
 		return FAILED;
 
-	for (i = 0; i < target->req_ring_size; ++i) {
-		struct srp_request *req = &target->req_ring[i];
-		srp_finish_req(target, req, scmnd->device, DID_RESET << 16);
+	for (i = 0; i < target->ch_count; i++) {
+		ch = &target->ch[i];
+		for (i = 0; i < target->req_ring_size; ++i) {
+			struct srp_request *req = &ch->req_ring[i];
+
+			srp_finish_req(ch, req, scmnd->device, DID_RESET << 16);
+		}
 	}
 
 	return SUCCESS;
@@ -2466,7 +2589,7 @@ static ssize_t show_pkey(struct device *dev, struct device_attribute *attr,
 {
 	struct srp_target_port *target = host_to_target(class_to_shost(dev));
 
-	return sprintf(buf, "0x%04x\n", be16_to_cpu(target->path.pkey));
+	return sprintf(buf, "0x%04x\n", be16_to_cpu(target->pkey));
 }
 
 static ssize_t show_sgid(struct device *dev, struct device_attribute *attr,
@@ -2474,15 +2597,16 @@ static ssize_t show_sgid(struct device *dev, struct device_attribute *attr,
 {
 	struct srp_target_port *target = host_to_target(class_to_shost(dev));
 
-	return sprintf(buf, "%pI6\n", target->path.sgid.raw);
+	return sprintf(buf, "%pI6\n", target->sgid.raw);
 }
 
 static ssize_t show_dgid(struct device *dev, struct device_attribute *attr,
 			 char *buf)
 {
 	struct srp_target_port *target = host_to_target(class_to_shost(dev));
+	struct srp_rdma_ch *ch = &target->ch[0];
 
-	return sprintf(buf, "%pI6\n", target->path.dgid.raw);
+	return sprintf(buf, "%pI6\n", ch->path.dgid.raw);
 }
 
 static ssize_t show_orig_dgid(struct device *dev,
@@ -2490,15 +2614,21 @@ static ssize_t show_orig_dgid(struct device *dev,
 {
 	struct srp_target_port *target = host_to_target(class_to_shost(dev));
 
-	return sprintf(buf, "%pI6\n", target->orig_dgid);
+	return sprintf(buf, "%pI6\n", target->orig_dgid.raw);
 }
 
 static ssize_t show_req_lim(struct device *dev,
 			    struct device_attribute *attr, char *buf)
 {
 	struct srp_target_port *target = host_to_target(class_to_shost(dev));
+	struct srp_rdma_ch *ch;
+	int i, req_lim = INT_MAX;
 
-	return sprintf(buf, "%d\n", target->req_lim);
+	for (i = 0; i < target->ch_count; i++) {
+		ch = &target->ch[i];
+		req_lim = min(req_lim, ch->req_lim);
+	}
+	return sprintf(buf, "%d\n", req_lim);
 }
 
 static ssize_t show_zero_req_lim(struct device *dev,
@@ -2525,6 +2655,14 @@ static ssize_t show_local_ib_device(struct device *dev,
 	return sprintf(buf, "%s\n", target->srp_host->srp_dev->dev->name);
 }
 
+static ssize_t show_ch_count(struct device *dev, struct device_attribute *attr,
+			     char *buf)
+{
+	struct srp_target_port *target = host_to_target(class_to_shost(dev));
+
+	return sprintf(buf, "%d\n", target->ch_count);
+}
+
 static ssize_t show_comp_vector(struct device *dev,
 				struct device_attribute *attr, char *buf)
 {
@@ -2568,6 +2706,7 @@ static DEVICE_ATTR(req_lim,         S_IRUGO, show_req_lim,         NULL);
 static DEVICE_ATTR(zero_req_lim,    S_IRUGO, show_zero_req_lim,	   NULL);
 static DEVICE_ATTR(local_ib_port,   S_IRUGO, show_local_ib_port,   NULL);
 static DEVICE_ATTR(local_ib_device, S_IRUGO, show_local_ib_device, NULL);
+static DEVICE_ATTR(ch_count,        S_IRUGO, show_ch_count,        NULL);
 static DEVICE_ATTR(comp_vector,     S_IRUGO, show_comp_vector,     NULL);
 static DEVICE_ATTR(tl_retry_count,  S_IRUGO, show_tl_retry_count,  NULL);
 static DEVICE_ATTR(cmd_sg_entries,  S_IRUGO, show_cmd_sg_entries,  NULL);
@@ -2585,6 +2724,7 @@ static struct device_attribute *srp_host_attrs[] = {
 	&dev_attr_zero_req_lim,
 	&dev_attr_local_ib_port,
 	&dev_attr_local_ib_device,
+	&dev_attr_ch_count,
 	&dev_attr_comp_vector,
 	&dev_attr_tl_retry_count,
 	&dev_attr_cmd_sg_entries,
@@ -2600,7 +2740,6 @@ static struct scsi_host_template srp_template = {
 	.info				= srp_target_info,
 	.queuecommand			= srp_queuecommand,
 	.change_queue_depth             = srp_change_queue_depth,
-	.change_queue_type              = srp_change_queue_type,
 	.eh_abort_handler		= srp_abort,
 	.eh_device_reset_handler	= srp_reset_device,
 	.eh_host_reset_handler		= srp_reset_host,
@@ -2610,14 +2749,28 @@ static struct scsi_host_template srp_template = {
 	.this_id			= -1,
 	.cmd_per_lun			= SRP_DEFAULT_CMD_SQ_SIZE,
 	.use_clustering			= ENABLE_CLUSTERING,
-	.shost_attrs			= srp_host_attrs
+	.shost_attrs			= srp_host_attrs,
+	.use_blk_tags			= 1,
+	.track_queue_depth		= 1,
 };
 
+static int srp_sdev_count(struct Scsi_Host *host)
+{
+	struct scsi_device *sdev;
+	int c = 0;
+
+	shost_for_each_device(sdev, host)
+		c++;
+
+	return c;
+}
+
 static int srp_add_target(struct srp_host *host, struct srp_target_port *target)
 {
 	struct srp_rport_identifiers ids;
 	struct srp_rport *rport;
 
+	target->state = SRP_TARGET_SCANNING;
 	sprintf(target->target_name, "SRP.T10:%016llX",
 		 (unsigned long long) be64_to_cpu(target->id_ext));
 
@@ -2640,11 +2793,26 @@ static int srp_add_target(struct srp_host *host, struct srp_target_port *target)
 	list_add_tail(&target->list, &host->target_list);
 	spin_unlock(&host->target_lock);
 
-	target->state = SRP_TARGET_LIVE;
-
 	scsi_scan_target(&target->scsi_host->shost_gendev,
 			 0, target->scsi_id, SCAN_WILD_CARD, 0);
 
+	if (!target->connected || target->qp_in_error) {
+		shost_printk(KERN_INFO, target->scsi_host,
+			     PFX "SCSI scan failed - removing SCSI host\n");
+		srp_queue_remove_work(target);
+		goto out;
+	}
+
+	pr_debug(PFX "%s: SCSI scan succeeded - detected %d LUNs\n",
+		 dev_name(&target->scsi_host->shost_gendev),
+		 srp_sdev_count(target->scsi_host));
+
+	spin_lock_irq(&target->lock);
+	if (target->state == SRP_TARGET_SCANNING)
+		target->state = SRP_TARGET_LIVE;
+	spin_unlock_irq(&target->lock);
+
+out:
 	return 0;
 }
 
@@ -2760,7 +2928,7 @@ static int srp_parse_options(const char *buf, struct srp_target_port *target)
 		return -ENOMEM;
 
 	sep_opt = options;
-	while ((p = strsep(&sep_opt, ",")) != NULL) {
+	while ((p = strsep(&sep_opt, ",\n")) != NULL) {
 		if (!*p)
 			continue;
 
@@ -2801,11 +2969,15 @@ static int srp_parse_options(const char *buf, struct srp_target_port *target)
 			}
 
 			for (i = 0; i < 16; ++i) {
-				strlcpy(dgid, p + i * 2, 3);
-				target->path.dgid.raw[i] = simple_strtoul(dgid, NULL, 16);
+				strlcpy(dgid, p + i * 2, sizeof(dgid));
+				if (sscanf(dgid, "%hhx",
+					   &target->orig_dgid.raw[i]) < 1) {
+					ret = -EINVAL;
+					kfree(p);
+					goto out;
+				}
 			}
 			kfree(p);
-			memcpy(target->orig_dgid, target->path.dgid.raw, 16);
 			break;
 
 		case SRP_OPT_PKEY:
@@ -2813,7 +2985,7 @@ static int srp_parse_options(const char *buf, struct srp_target_port *target)
 				pr_warn("bad P_Key parameter '%s'\n", p);
 				goto out;
 			}
-			target->path.pkey = cpu_to_be16(token);
+			target->pkey = cpu_to_be16(token);
 			break;
 
 		case SRP_OPT_SERVICE_ID:
@@ -2823,7 +2995,6 @@ static int srp_parse_options(const char *buf, struct srp_target_port *target)
 				goto out;
 			}
 			target->service_id = cpu_to_be64(simple_strtoull(p, NULL, 16));
-			target->path.service_id = target->service_id;
 			kfree(p);
 			break;
 
@@ -2960,9 +3131,11 @@ static ssize_t srp_create_target(struct device *dev,
 		container_of(dev, struct srp_host, dev);
 	struct Scsi_Host *target_host;
 	struct srp_target_port *target;
+	struct srp_rdma_ch *ch;
 	struct srp_device *srp_dev = host->srp_dev;
 	struct ib_device *ibdev = srp_dev->dev;
-	int ret;
+	int ret, node_idx, node, cpu, i;
+	bool multich = false;
 
 	target_host = scsi_host_alloc(&srp_template,
 				      sizeof (struct srp_target_port));
@@ -2988,12 +3161,22 @@ static ssize_t srp_create_target(struct device *dev,
 	target->tl_retry_count	= 7;
 	target->queue_size	= SRP_DEFAULT_QUEUE_SIZE;
 
+	/*
+	 * Avoid that the SCSI host can be removed by srp_remove_target()
+	 * before this function returns.
+	 */
+	scsi_host_get(target->scsi_host);
+
 	mutex_lock(&host->add_target_mutex);
 
 	ret = srp_parse_options(buf, target);
 	if (ret)
 		goto err;
 
+	ret = scsi_init_shared_tag_map(target_host, target_host->can_queue);
+	if (ret)
+		goto err;
+
 	target->req_ring_size = target->queue_size - SRP_TSK_MGMT_SQ_SIZE;
 
 	if (!srp_conn_unique(target->srp_host, target)) {
@@ -3022,59 +3205,115 @@ static ssize_t srp_create_target(struct device *dev,
 	INIT_WORK(&target->tl_err_work, srp_tl_err_work);
 	INIT_WORK(&target->remove_work, srp_remove_work);
 	spin_lock_init(&target->lock);
-	INIT_LIST_HEAD(&target->free_tx);
-	ret = srp_alloc_req_data(target);
+	ret = ib_query_gid(ibdev, host->port, 0, &target->sgid);
 	if (ret)
-		goto err_free_mem;
+		goto err;
 
-	ret = ib_query_gid(ibdev, host->port, 0, &target->path.sgid);
-	if (ret)
-		goto err_free_mem;
+	ret = -ENOMEM;
+	target->ch_count = max_t(unsigned, num_online_nodes(),
+				 min(ch_count ? :
+				     min(4 * num_online_nodes(),
+					 ibdev->num_comp_vectors),
+				     num_online_cpus()));
+	target->ch = kcalloc(target->ch_count, sizeof(*target->ch),
+			     GFP_KERNEL);
+	if (!target->ch)
+		goto err;
 
-	ret = srp_create_target_ib(target);
-	if (ret)
-		goto err_free_mem;
+	node_idx = 0;
+	for_each_online_node(node) {
+		const int ch_start = (node_idx * target->ch_count /
+				      num_online_nodes());
+		const int ch_end = ((node_idx + 1) * target->ch_count /
+				    num_online_nodes());
+		const int cv_start = (node_idx * ibdev->num_comp_vectors /
+				      num_online_nodes() + target->comp_vector)
+				     % ibdev->num_comp_vectors;
+		const int cv_end = ((node_idx + 1) * ibdev->num_comp_vectors /
+				    num_online_nodes() + target->comp_vector)
+				   % ibdev->num_comp_vectors;
+		int cpu_idx = 0;
+
+		for_each_online_cpu(cpu) {
+			if (cpu_to_node(cpu) != node)
+				continue;
+			if (ch_start + cpu_idx >= ch_end)
+				continue;
+			ch = &target->ch[ch_start + cpu_idx];
+			ch->target = target;
+			ch->comp_vector = cv_start == cv_end ? cv_start :
+				cv_start + cpu_idx % (cv_end - cv_start);
+			spin_lock_init(&ch->lock);
+			INIT_LIST_HEAD(&ch->free_tx);
+			ret = srp_new_cm_id(ch);
+			if (ret)
+				goto err_disconnect;
 
-	ret = srp_new_cm_id(target);
-	if (ret)
-		goto err_free_ib;
+			ret = srp_create_ch_ib(ch);
+			if (ret)
+				goto err_disconnect;
 
-	ret = srp_connect_target(target);
-	if (ret) {
-		shost_printk(KERN_ERR, target->scsi_host,
-			     PFX "Connection failed\n");
-		goto err_cm_id;
+			ret = srp_alloc_req_data(ch);
+			if (ret)
+				goto err_disconnect;
+
+			ret = srp_connect_ch(ch, multich);
+			if (ret) {
+				shost_printk(KERN_ERR, target->scsi_host,
+					     PFX "Connection %d/%d failed\n",
+					     ch_start + cpu_idx,
+					     target->ch_count);
+				if (node_idx == 0 && cpu_idx == 0) {
+					goto err_disconnect;
+				} else {
+					srp_free_ch_ib(target, ch);
+					srp_free_req_data(target, ch);
+					target->ch_count = ch - target->ch;
+					break;
+				}
+			}
+
+			multich = true;
+			cpu_idx++;
+		}
+		node_idx++;
 	}
 
+	target->scsi_host->nr_hw_queues = target->ch_count;
+
 	ret = srp_add_target(host, target);
 	if (ret)
 		goto err_disconnect;
 
-	shost_printk(KERN_DEBUG, target->scsi_host, PFX
-		     "new target: id_ext %016llx ioc_guid %016llx pkey %04x service_id %016llx sgid %pI6 dgid %pI6\n",
-		     be64_to_cpu(target->id_ext),
-		     be64_to_cpu(target->ioc_guid),
-		     be16_to_cpu(target->path.pkey),
-		     be64_to_cpu(target->service_id),
-		     target->path.sgid.raw, target->path.dgid.raw);
+	if (target->state != SRP_TARGET_REMOVED) {
+		shost_printk(KERN_DEBUG, target->scsi_host, PFX
+			     "new target: id_ext %016llx ioc_guid %016llx pkey %04x service_id %016llx sgid %pI6 dgid %pI6\n",
+			     be64_to_cpu(target->id_ext),
+			     be64_to_cpu(target->ioc_guid),
+			     be16_to_cpu(target->pkey),
+			     be64_to_cpu(target->service_id),
+			     target->sgid.raw, target->orig_dgid.raw);
+	}
 
 	ret = count;
 
 out:
 	mutex_unlock(&host->add_target_mutex);
+
+	scsi_host_put(target->scsi_host);
+
 	return ret;
 
 err_disconnect:
 	srp_disconnect_target(target);
 
-err_cm_id:
-	ib_destroy_cm_id(target->cm_id);
-
-err_free_ib:
-	srp_free_target_ib(target);
+	for (i = 0; i < target->ch_count; i++) {
+		ch = &target->ch[i];
+		srp_free_ch_ib(target, ch);
+		srp_free_req_data(target, ch);
+	}
 
-err_free_mem:
-	srp_free_req_data(target);
+	kfree(target->ch);
 
 err:
 	scsi_host_put(target_host);
diff --git a/drivers/infiniband/ulp/srp/ib_srp.h b/drivers/infiniband/ulp/srp/ib_srp.h
index e46ecb15aa0d..a611556406ac 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.h
+++ b/drivers/infiniband/ulp/srp/ib_srp.h
@@ -70,9 +70,12 @@ enum {
 
 	LOCAL_INV_WR_ID_MASK	= 1,
 	FAST_REG_WR_ID_MASK	= 2,
+
+	SRP_LAST_WR_ID		= 0xfffffffcU,
 };
 
 enum srp_target_state {
+	SRP_TARGET_SCANNING,
 	SRP_TARGET_LIVE,
 	SRP_TARGET_REMOVED,
 };
@@ -115,7 +118,6 @@ struct srp_host {
 };
 
 struct srp_request {
-	struct list_head	list;
 	struct scsi_cmnd       *scmnd;
 	struct srp_iu	       *cmd;
 	union {
@@ -126,24 +128,62 @@ struct srp_request {
 	struct srp_direct_buf  *indirect_desc;
 	dma_addr_t		indirect_dma_addr;
 	short			nmdesc;
-	short			index;
 };
 
-struct srp_target_port {
+/**
+ * struct srp_rdma_ch
+ * @comp_vector: Completion vector used by this RDMA channel.
+ */
+struct srp_rdma_ch {
 	/* These are RW in the hot path, and commonly used together */
 	struct list_head	free_tx;
-	struct list_head	free_reqs;
 	spinlock_t		lock;
 	s32			req_lim;
 
 	/* These are read-only in the hot path */
-	struct ib_cq	       *send_cq ____cacheline_aligned_in_smp;
+	struct srp_target_port *target ____cacheline_aligned_in_smp;
+	struct ib_cq	       *send_cq;
 	struct ib_cq	       *recv_cq;
 	struct ib_qp	       *qp;
 	union {
 		struct ib_fmr_pool     *fmr_pool;
 		struct srp_fr_pool     *fr_pool;
 	};
+
+	/* Everything above this point is used in the hot path of
+	 * command processing. Try to keep them packed into cachelines.
+	 */
+
+	struct completion	done;
+	int			status;
+
+	struct ib_sa_path_rec	path;
+	struct ib_sa_query     *path_query;
+	int			path_query_id;
+
+	struct ib_cm_id	       *cm_id;
+	struct srp_iu	      **tx_ring;
+	struct srp_iu	      **rx_ring;
+	struct srp_request     *req_ring;
+	int			max_ti_iu_len;
+	int			comp_vector;
+
+	struct completion	tsk_mgmt_done;
+	u8			tsk_mgmt_status;
+};
+
+/**
+ * struct srp_target_port
+ * @comp_vector: Completion vector used by the first RDMA channel created for
+ *   this target port.
+ */
+struct srp_target_port {
+	/* read and written in the hot path */
+	spinlock_t		lock;
+
+	/* read only in the hot path */
+	struct srp_rdma_ch	*ch;
+	u32			ch_count;
 	u32			lkey;
 	u32			rkey;
 	enum srp_target_state	state;
@@ -152,10 +192,8 @@ struct srp_target_port {
 	unsigned int		indirect_size;
 	bool			allow_ext_sg;
 
-	/* Everything above this point is used in the hot path of
-	 * command processing. Try to keep them packed into cachelines.
-	 */
-
+	/* other member variables */
+	union ib_gid		sgid;
 	__be64			id_ext;
 	__be64			ioc_guid;
 	__be64			service_id;
@@ -172,34 +210,19 @@ struct srp_target_port {
 	int			comp_vector;
 	int			tl_retry_count;
 
-	struct ib_sa_path_rec	path;
-	__be16			orig_dgid[8];
-	struct ib_sa_query     *path_query;
-	int			path_query_id;
+	union ib_gid		orig_dgid;
+	__be16			pkey;
 
 	u32			rq_tmo_jiffies;
 	bool			connected;
 
-	struct ib_cm_id	       *cm_id;
-
-	int			max_ti_iu_len;
-
 	int			zero_req_lim;
 
-	struct srp_iu	       **tx_ring;
-	struct srp_iu	       **rx_ring;
-	struct srp_request	*req_ring;
-
 	struct work_struct	tl_err_work;
 	struct work_struct	remove_work;
 
 	struct list_head	list;
-	struct completion	done;
-	int			status;
 	bool			qp_in_error;
-
-	struct completion	tsk_mgmt_done;
-	u8			tsk_mgmt_status;
 };
 
 struct srp_iu {
diff --git a/drivers/infiniband/ulp/srpt/ib_srpt.c b/drivers/infiniband/ulp/srpt/ib_srpt.c
index d28a8c284da9..eb694ddad79f 100644
--- a/drivers/infiniband/ulp/srpt/ib_srpt.c
+++ b/drivers/infiniband/ulp/srpt/ib_srpt.c
@@ -1708,17 +1708,17 @@ static int srpt_handle_cmd(struct srpt_rdma_ch *ch,
 
 	switch (srp_cmd->task_attr) {
 	case SRP_CMD_SIMPLE_Q:
-		cmd->sam_task_attr = MSG_SIMPLE_TAG;
+		cmd->sam_task_attr = TCM_SIMPLE_TAG;
 		break;
 	case SRP_CMD_ORDERED_Q:
 	default:
-		cmd->sam_task_attr = MSG_ORDERED_TAG;
+		cmd->sam_task_attr = TCM_ORDERED_TAG;
 		break;
 	case SRP_CMD_HEAD_OF_Q:
-		cmd->sam_task_attr = MSG_HEAD_TAG;
+		cmd->sam_task_attr = TCM_HEAD_TAG;
 		break;
 	case SRP_CMD_ACA:
-		cmd->sam_task_attr = MSG_ACA_TAG;
+		cmd->sam_task_attr = TCM_ACA_TAG;
 		break;
 	}
 
@@ -1733,7 +1733,7 @@ static int srpt_handle_cmd(struct srpt_rdma_ch *ch,
 				       sizeof(srp_cmd->lun));
 	rc = target_submit_cmd(cmd, ch->sess, srp_cmd->cdb,
 			&send_ioctx->sense_data[0], unpacked_lun, data_len,
-			MSG_SIMPLE_TAG, dir, TARGET_SCF_ACK_KREF);
+			TCM_SIMPLE_TAG, dir, TARGET_SCF_ACK_KREF);
 	if (rc != 0) {
 		ret = TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE;
 		goto send_sense;
@@ -2092,6 +2092,7 @@ static int srpt_create_ch_ib(struct srpt_rdma_ch *ch)
 	if (!qp_init)
 		goto out;
 
+retry:
 	ch->cq = ib_create_cq(sdev->device, srpt_completion, NULL, ch,
 			      ch->rq_size + srp_sq_size, 0);
 	if (IS_ERR(ch->cq)) {
@@ -2115,6 +2116,13 @@ static int srpt_create_ch_ib(struct srpt_rdma_ch *ch)
 	ch->qp = ib_create_qp(sdev->pd, qp_init);
 	if (IS_ERR(ch->qp)) {
 		ret = PTR_ERR(ch->qp);
+		if (ret == -ENOMEM) {
+			srp_sq_size /= 2;
+			if (srp_sq_size >= MIN_SRPT_SQ_SIZE) {
+				ib_destroy_cq(ch->cq);
+				goto retry;
+			}
+		}
 		printk(KERN_ERR "failed to create_qp ret= %d\n", ret);
 		goto err_destroy_cq;
 	}
@@ -3574,7 +3582,7 @@ static int srpt_parse_i_port_id(u8 i_port_id[16], const char *name)
 	int ret, rc;
 
 	p = name;
-	if (strnicmp(p, "0x", 2) == 0)
+	if (strncasecmp(p, "0x", 2) == 0)
 		p += 2;
 	ret = -EINVAL;
 	len = strlen(p);
diff --git a/drivers/input/evdev.c b/drivers/input/evdev.c
index b1a52abc58df..18d4b2c8fe55 100644
--- a/drivers/input/evdev.c
+++ b/drivers/input/evdev.c
@@ -451,7 +451,7 @@ static int evdev_open(struct inode *inode, struct file *file)
 
  err_free_client:
 	evdev_detach_client(evdev, client);
-	kfree(client);
+	kvfree(client);
 	return error;
 }
 
diff --git a/drivers/input/keyboard/adp5520-keys.c b/drivers/input/keyboard/adp5520-keys.c
index 7f4a8b58efc1..db1004dad108 100644
--- a/drivers/input/keyboard/adp5520-keys.c
+++ b/drivers/input/keyboard/adp5520-keys.c
@@ -184,7 +184,6 @@ static int adp5520_keys_remove(struct platform_device *pdev)
 static struct platform_driver adp5520_keys_driver = {
 	.driver	= {
 		.name	= "adp5520-keys",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= adp5520_keys_probe,
 	.remove		= adp5520_keys_remove,
diff --git a/drivers/input/keyboard/adp5588-keys.c b/drivers/input/keyboard/adp5588-keys.c
index 728133ec6d7b..21a62d0fa764 100644
--- a/drivers/input/keyboard/adp5588-keys.c
+++ b/drivers/input/keyboard/adp5588-keys.c
@@ -251,9 +251,7 @@ static void adp5588_gpio_remove(struct adp5588_kpad *kpad)
 			dev_warn(dev, "teardown failed %d\n", error);
 	}
 
-	error = gpiochip_remove(&kpad->gc);
-	if (error)
-		dev_warn(dev, "gpiochip_remove failed %d\n", error);
+	gpiochip_remove(&kpad->gc);
 }
 #else
 static inline int adp5588_gpio_add(struct adp5588_kpad *kpad)
diff --git a/drivers/input/keyboard/adp5589-keys.c b/drivers/input/keyboard/adp5589-keys.c
index 6329549bf6ad..a45267729dfc 100644
--- a/drivers/input/keyboard/adp5589-keys.c
+++ b/drivers/input/keyboard/adp5589-keys.c
@@ -567,9 +567,7 @@ static void adp5589_gpio_remove(struct adp5589_kpad *kpad)
 			dev_warn(dev, "teardown failed %d\n", error);
 	}
 
-	error = gpiochip_remove(&kpad->gc);
-	if (error)
-		dev_warn(dev, "gpiochip_remove failed %d\n", error);
+	gpiochip_remove(&kpad->gc);
 }
 #else
 static inline int adp5589_gpio_add(struct adp5589_kpad *kpad)
diff --git a/drivers/input/keyboard/amikbd.c b/drivers/input/keyboard/amikbd.c
index 3196f2d29ade..e04a3b4e55d6 100644
--- a/drivers/input/keyboard/amikbd.c
+++ b/drivers/input/keyboard/amikbd.c
@@ -268,7 +268,6 @@ static struct platform_driver amikbd_driver = {
 	.remove = __exit_p(amikbd_remove),
 	.driver   = {
 		.name	= "amiga-keyboard",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/input/keyboard/bf54x-keys.c b/drivers/input/keyboard/bf54x-keys.c
index e6d46c5994d7..81b07dddae86 100644
--- a/drivers/input/keyboard/bf54x-keys.c
+++ b/drivers/input/keyboard/bf54x-keys.c
@@ -385,7 +385,6 @@ static int bfin_kpad_resume(struct platform_device *pdev)
 static struct platform_driver bfin_kpad_device_driver = {
 	.driver		= {
 		.name	= DRV_NAME,
-		.owner	= THIS_MODULE,
 	},
 	.probe		= bfin_kpad_probe,
 	.remove		= bfin_kpad_remove,
diff --git a/drivers/input/keyboard/clps711x-keypad.c b/drivers/input/keyboard/clps711x-keypad.c
index 552b65c6e6b0..27ef29f8fe6a 100644
--- a/drivers/input/keyboard/clps711x-keypad.c
+++ b/drivers/input/keyboard/clps711x-keypad.c
@@ -194,7 +194,6 @@ MODULE_DEVICE_TABLE(of, clps711x_keypad_of_match);
 static struct platform_driver clps711x_keypad_driver = {
 	.driver	= {
 		.name		= "clps711x-keypad",
-		.owner		= THIS_MODULE,
 		.of_match_table	= clps711x_keypad_of_match,
 	},
 	.probe	= clps711x_keypad_probe,
diff --git a/drivers/input/keyboard/cros_ec_keyb.c b/drivers/input/keyboard/cros_ec_keyb.c
index 1e83ef99098e..ffa989f2c785 100644
--- a/drivers/input/keyboard/cros_ec_keyb.c
+++ b/drivers/input/keyboard/cros_ec_keyb.c
@@ -157,7 +157,7 @@ static int cros_ec_keyb_get_state(struct cros_ec_keyb *ckdev, uint8_t *kb_state)
 		.insize = ckdev->cols,
 	};
 
-	return ckdev->ec->cmd_xfer(ckdev->ec, &msg);
+	return cros_ec_cmd_xfer(ckdev->ec, &msg);
 }
 
 static irqreturn_t cros_ec_keyb_irq(int irq, void *data)
diff --git a/drivers/input/keyboard/davinci_keyscan.c b/drivers/input/keyboard/davinci_keyscan.c
index 1559dc1cf951..f363d1d2907a 100644
--- a/drivers/input/keyboard/davinci_keyscan.c
+++ b/drivers/input/keyboard/davinci_keyscan.c
@@ -322,7 +322,6 @@ static int davinci_ks_remove(struct platform_device *pdev)
 static struct platform_driver davinci_ks_driver = {
 	.driver	= {
 		.name = "davinci_keyscan",
-		.owner = THIS_MODULE,
 	},
 	.remove	= davinci_ks_remove,
 };
diff --git a/drivers/input/keyboard/ep93xx_keypad.c b/drivers/input/keyboard/ep93xx_keypad.c
index e59876212b8c..f77b295e0123 100644
--- a/drivers/input/keyboard/ep93xx_keypad.c
+++ b/drivers/input/keyboard/ep93xx_keypad.c
@@ -373,7 +373,6 @@ static int ep93xx_keypad_remove(struct platform_device *pdev)
 static struct platform_driver ep93xx_keypad_driver = {
 	.driver		= {
 		.name	= "ep93xx-keypad",
-		.owner	= THIS_MODULE,
 		.pm	= &ep93xx_keypad_pm_ops,
 	},
 	.probe		= ep93xx_keypad_probe,
diff --git a/drivers/input/keyboard/goldfish_events.c b/drivers/input/keyboard/goldfish_events.c
index 69e854763370..907e4e278fce 100644
--- a/drivers/input/keyboard/goldfish_events.c
+++ b/drivers/input/keyboard/goldfish_events.c
@@ -181,7 +181,6 @@ static int events_probe(struct platform_device *pdev)
 static struct platform_driver events_driver = {
 	.probe	= events_probe,
 	.driver	= {
-		.owner	= THIS_MODULE,
 		.name	= "goldfish_events",
 	},
 };
diff --git a/drivers/input/keyboard/gpio_keys.c b/drivers/input/keyboard/gpio_keys.c
index eefd976ab4eb..883d6aed5b9a 100644
--- a/drivers/input/keyboard/gpio_keys.c
+++ b/drivers/input/keyboard/gpio_keys.c
@@ -840,7 +840,6 @@ static struct platform_driver gpio_keys_device_driver = {
 	.remove		= gpio_keys_remove,
 	.driver		= {
 		.name	= "gpio-keys",
-		.owner	= THIS_MODULE,
 		.pm	= &gpio_keys_pm_ops,
 		.of_match_table = of_match_ptr(gpio_keys_of_match),
 	}
diff --git a/drivers/input/keyboard/gpio_keys_polled.c b/drivers/input/keyboard/gpio_keys_polled.c
index 432d36395f35..90df4df58b07 100644
--- a/drivers/input/keyboard/gpio_keys_polled.c
+++ b/drivers/input/keyboard/gpio_keys_polled.c
@@ -23,10 +23,9 @@
 #include <linux/ioport.h>
 #include <linux/platform_device.h>
 #include <linux/gpio.h>
+#include <linux/gpio/consumer.h>
 #include <linux/gpio_keys.h>
-#include <linux/of.h>
-#include <linux/of_platform.h>
-#include <linux/of_gpio.h>
+#include <linux/property.h>
 
 #define DRV_NAME	"gpio-keys-polled"
 
@@ -51,15 +50,14 @@ static void gpio_keys_polled_check_state(struct input_dev *input,
 	int state;
 
 	if (bdata->can_sleep)
-		state = !!gpio_get_value_cansleep(button->gpio);
+		state = !!gpiod_get_value_cansleep(button->gpiod);
 	else
-		state = !!gpio_get_value(button->gpio);
+		state = !!gpiod_get_value(button->gpiod);
 
 	if (state != bdata->last_state) {
 		unsigned int type = button->type ?: EV_KEY;
 
-		input_event(input, type, button->code,
-			    !!(state ^ button->active_low));
+		input_event(input, type, button->code, state);
 		input_sync(input);
 		bdata->count = 0;
 		bdata->last_state = state;
@@ -102,21 +100,15 @@ static void gpio_keys_polled_close(struct input_polled_dev *dev)
 		pdata->disable(bdev->dev);
 }
 
-#ifdef CONFIG_OF
 static struct gpio_keys_platform_data *gpio_keys_polled_get_devtree_pdata(struct device *dev)
 {
-	struct device_node *node, *pp;
 	struct gpio_keys_platform_data *pdata;
 	struct gpio_keys_button *button;
+	struct fwnode_handle *child;
 	int error;
 	int nbuttons;
-	int i;
-
-	node = dev->of_node;
-	if (!node)
-		return NULL;
 
-	nbuttons = of_get_child_count(node);
+	nbuttons = device_get_child_node_count(dev);
 	if (nbuttons == 0)
 		return NULL;
 
@@ -126,52 +118,44 @@ static struct gpio_keys_platform_data *gpio_keys_polled_get_devtree_pdata(struct
 		return ERR_PTR(-ENOMEM);
 
 	pdata->buttons = (struct gpio_keys_button *)(pdata + 1);
-	pdata->nbuttons = nbuttons;
 
-	pdata->rep = !!of_get_property(node, "autorepeat", NULL);
-	of_property_read_u32(node, "poll-interval", &pdata->poll_interval);
+	pdata->rep = device_property_present(dev, "autorepeat");
+	device_property_read_u32(dev, "poll-interval", &pdata->poll_interval);
 
-	i = 0;
-	for_each_child_of_node(node, pp) {
-		int gpio;
-		enum of_gpio_flags flags;
+	device_for_each_child_node(dev, child) {
+		struct gpio_desc *desc;
 
-		if (!of_find_property(pp, "gpios", NULL)) {
-			pdata->nbuttons--;
-			dev_warn(dev, "Found button without gpios\n");
-			continue;
-		}
-
-		gpio = of_get_gpio_flags(pp, 0, &flags);
-		if (gpio < 0) {
-			error = gpio;
+		desc = devm_get_gpiod_from_child(dev, child);
+		if (IS_ERR(desc)) {
+			error = PTR_ERR(desc);
 			if (error != -EPROBE_DEFER)
 				dev_err(dev,
 					"Failed to get gpio flags, error: %d\n",
 					error);
+			fwnode_handle_put(child);
 			return ERR_PTR(error);
 		}
 
-		button = &pdata->buttons[i++];
-
-		button->gpio = gpio;
-		button->active_low = flags & OF_GPIO_ACTIVE_LOW;
+		button = &pdata->buttons[pdata->nbuttons++];
+		button->gpiod = desc;
 
-		if (of_property_read_u32(pp, "linux,code", &button->code)) {
-			dev_err(dev, "Button without keycode: 0x%x\n",
-				button->gpio);
+		if (fwnode_property_read_u32(child, "linux,code", &button->code)) {
+			dev_err(dev, "Button without keycode: %d\n",
+				pdata->nbuttons - 1);
+			fwnode_handle_put(child);
 			return ERR_PTR(-EINVAL);
 		}
 
-		button->desc = of_get_property(pp, "label", NULL);
+		fwnode_property_read_string(child, "label", &button->desc);
 
-		if (of_property_read_u32(pp, "linux,input-type", &button->type))
+		if (fwnode_property_read_u32(child, "linux,input-type",
+					     &button->type))
 			button->type = EV_KEY;
 
-		button->wakeup = !!of_get_property(pp, "gpio-key,wakeup", NULL);
+		button->wakeup = fwnode_property_present(child, "gpio-key,wakeup");
 
-		if (of_property_read_u32(pp, "debounce-interval",
-					 &button->debounce_interval))
+		if (fwnode_property_read_u32(child, "debounce-interval",
+					     &button->debounce_interval))
 			button->debounce_interval = 5;
 	}
 
@@ -187,15 +171,6 @@ static const struct of_device_id gpio_keys_polled_of_match[] = {
 };
 MODULE_DEVICE_TABLE(of, gpio_keys_polled_of_match);
 
-#else
-
-static inline struct gpio_keys_platform_data *
-gpio_keys_polled_get_devtree_pdata(struct device *dev)
-{
-	return NULL;
-}
-#endif
-
 static int gpio_keys_polled_probe(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
@@ -259,7 +234,6 @@ static int gpio_keys_polled_probe(struct platform_device *pdev)
 	for (i = 0; i < pdata->nbuttons; i++) {
 		struct gpio_keys_button *button = &pdata->buttons[i];
 		struct gpio_keys_button_data *bdata = &bdev->data[i];
-		unsigned int gpio = button->gpio;
 		unsigned int type = button->type ?: EV_KEY;
 
 		if (button->wakeup) {
@@ -267,15 +241,31 @@ static int gpio_keys_polled_probe(struct platform_device *pdev)
 			return -EINVAL;
 		}
 
-		error = devm_gpio_request_one(&pdev->dev, gpio, GPIOF_IN,
-					      button->desc ? : DRV_NAME);
-		if (error) {
-			dev_err(dev, "unable to claim gpio %u, err=%d\n",
-				gpio, error);
-			return error;
+		/*
+		 * Legacy GPIO number so request the GPIO here and
+		 * convert it to descriptor.
+		 */
+		if (!button->gpiod && gpio_is_valid(button->gpio)) {
+			unsigned flags = 0;
+
+			if (button->active_low)
+				flags |= GPIOF_ACTIVE_LOW;
+
+			error = devm_gpio_request_one(&pdev->dev, button->gpio,
+					flags, button->desc ? : DRV_NAME);
+			if (error) {
+				dev_err(dev, "unable to claim gpio %u, err=%d\n",
+					button->gpio, error);
+				return error;
+			}
+
+			button->gpiod = gpio_to_desc(button->gpio);
 		}
 
-		bdata->can_sleep = gpio_cansleep(gpio);
+		if (IS_ERR(button->gpiod))
+			return PTR_ERR(button->gpiod);
+
+		bdata->can_sleep = gpiod_cansleep(button->gpiod);
 		bdata->last_state = -1;
 		bdata->threshold = DIV_ROUND_UP(button->debounce_interval,
 						pdata->poll_interval);
@@ -307,8 +297,7 @@ static struct platform_driver gpio_keys_polled_driver = {
 	.probe	= gpio_keys_polled_probe,
 	.driver	= {
 		.name	= DRV_NAME,
-		.owner	= THIS_MODULE,
-		.of_match_table = of_match_ptr(gpio_keys_polled_of_match),
+		.of_match_table = gpio_keys_polled_of_match,
 	},
 };
 module_platform_driver(gpio_keys_polled_driver);
diff --git a/drivers/input/keyboard/imx_keypad.c b/drivers/input/keyboard/imx_keypad.c
index 20a99c368d16..e53f232eda0e 100644
--- a/drivers/input/keyboard/imx_keypad.c
+++ b/drivers/input/keyboard/imx_keypad.c
@@ -580,7 +580,6 @@ static SIMPLE_DEV_PM_OPS(imx_kbd_pm_ops, imx_kbd_suspend, imx_kbd_resume);
 static struct platform_driver imx_keypad_driver = {
 	.driver		= {
 		.name	= "imx-keypad",
-		.owner	= THIS_MODULE,
 		.pm	= &imx_kbd_pm_ops,
 		.of_match_table = of_match_ptr(imx_keypad_of_match),
 	},
diff --git a/drivers/input/keyboard/jornada680_kbd.c b/drivers/input/keyboard/jornada680_kbd.c
index 0ba4428da24a..80c81278ad2c 100644
--- a/drivers/input/keyboard/jornada680_kbd.c
+++ b/drivers/input/keyboard/jornada680_kbd.c
@@ -237,7 +237,6 @@ static int jornada680kbd_probe(struct platform_device *pdev)
 static struct platform_driver jornada680kbd_driver = {
 	.driver	= {
 		.name	= "jornada680_kbd",
-		.owner	= THIS_MODULE,
 	},
 	.probe	= jornada680kbd_probe,
 };
diff --git a/drivers/input/keyboard/jornada720_kbd.c b/drivers/input/keyboard/jornada720_kbd.c
index cd729d485e98..421d9c55b0e8 100644
--- a/drivers/input/keyboard/jornada720_kbd.c
+++ b/drivers/input/keyboard/jornada720_kbd.c
@@ -167,7 +167,6 @@ MODULE_ALIAS("platform:jornada720_kbd");
 static struct platform_driver jornada720_kbd_driver = {
 	.driver  = {
 		.name    = "jornada720_kbd",
-		.owner	= THIS_MODULE,
 	 },
 	.probe   = jornada720_kbd_probe,
 	.remove  = jornada720_kbd_remove,
diff --git a/drivers/input/keyboard/lpc32xx-keys.c b/drivers/input/keyboard/lpc32xx-keys.c
index 9e9786d576cb..265d641c40e2 100644
--- a/drivers/input/keyboard/lpc32xx-keys.c
+++ b/drivers/input/keyboard/lpc32xx-keys.c
@@ -329,7 +329,6 @@ static struct platform_driver lpc32xx_kscan_driver = {
 	.probe		= lpc32xx_kscan_probe,
 	.driver		= {
 		.name	= DRV_NAME,
-		.owner	= THIS_MODULE,
 		.pm	= &lpc32xx_kscan_pm_ops,
 		.of_match_table = lpc32xx_kscan_match,
 	}
diff --git a/drivers/input/keyboard/matrix_keypad.c b/drivers/input/keyboard/matrix_keypad.c
index e651fa692afe..b370a59cb759 100644
--- a/drivers/input/keyboard/matrix_keypad.c
+++ b/drivers/input/keyboard/matrix_keypad.c
@@ -565,7 +565,6 @@ static struct platform_driver matrix_keypad_driver = {
 	.remove		= matrix_keypad_remove,
 	.driver		= {
 		.name	= "matrix-keypad",
-		.owner	= THIS_MODULE,
 		.pm	= &matrix_keypad_pm_ops,
 		.of_match_table = of_match_ptr(matrix_keypad_dt_match),
 	},
diff --git a/drivers/input/keyboard/nomadik-ske-keypad.c b/drivers/input/keyboard/nomadik-ske-keypad.c
index 63332e2f8628..c7d5b1666fc3 100644
--- a/drivers/input/keyboard/nomadik-ske-keypad.c
+++ b/drivers/input/keyboard/nomadik-ske-keypad.c
@@ -425,7 +425,6 @@ static SIMPLE_DEV_PM_OPS(ske_keypad_dev_pm_ops,
 static struct platform_driver ske_keypad_driver = {
 	.driver = {
 		.name = "nmk-ske-keypad",
-		.owner  = THIS_MODULE,
 		.pm = &ske_keypad_dev_pm_ops,
 	},
 	.remove = ske_keypad_remove,
diff --git a/drivers/input/keyboard/nspire-keypad.c b/drivers/input/keyboard/nspire-keypad.c
index b31064981e96..7abfd34eb87e 100644
--- a/drivers/input/keyboard/nspire-keypad.c
+++ b/drivers/input/keyboard/nspire-keypad.c
@@ -268,7 +268,6 @@ MODULE_DEVICE_TABLE(of, nspire_keypad_dt_match);
 static struct platform_driver nspire_keypad_driver = {
 	.driver = {
 		.name = "nspire-keypad",
-		.owner = THIS_MODULE,
 		.of_match_table = nspire_keypad_dt_match,
 	},
 	.probe = nspire_keypad_probe,
diff --git a/drivers/input/keyboard/omap-keypad.c b/drivers/input/keyboard/omap-keypad.c
index b1acc9852eb7..7502e46165fa 100644
--- a/drivers/input/keyboard/omap-keypad.c
+++ b/drivers/input/keyboard/omap-keypad.c
@@ -383,7 +383,6 @@ static struct platform_driver omap_kp_driver = {
 	.resume		= omap_kp_resume,
 	.driver		= {
 		.name	= "omap-keypad",
-		.owner	= THIS_MODULE,
 	},
 };
 module_platform_driver(omap_kp_driver);
diff --git a/drivers/input/keyboard/omap4-keypad.c b/drivers/input/keyboard/omap4-keypad.c
index 024b7bdffe5b..b052afec9a11 100644
--- a/drivers/input/keyboard/omap4-keypad.c
+++ b/drivers/input/keyboard/omap4-keypad.c
@@ -460,7 +460,6 @@ static struct platform_driver omap4_keypad_driver = {
 	.remove		= omap4_keypad_remove,
 	.driver		= {
 		.name	= "omap4-keypad",
-		.owner	= THIS_MODULE,
 		.pm	= &omap4_keypad_pm_ops,
 		.of_match_table = omap_keypad_dt_match,
 	},
diff --git a/drivers/input/keyboard/pmic8xxx-keypad.c b/drivers/input/keyboard/pmic8xxx-keypad.c
index 80c6b0ef3fc8..32580afecc26 100644
--- a/drivers/input/keyboard/pmic8xxx-keypad.c
+++ b/drivers/input/keyboard/pmic8xxx-keypad.c
@@ -687,7 +687,6 @@ static struct platform_driver pmic8xxx_kp_driver = {
 	.probe		= pmic8xxx_kp_probe,
 	.driver		= {
 		.name = "pm8xxx-keypad",
-		.owner = THIS_MODULE,
 		.pm = &pm8xxx_kp_pm_ops,
 		.of_match_table = pm8xxx_match_table,
 	},
diff --git a/drivers/input/keyboard/pxa27x_keypad.c b/drivers/input/keyboard/pxa27x_keypad.c
index e08fc0ae913d..a90d6bdc499e 100644
--- a/drivers/input/keyboard/pxa27x_keypad.c
+++ b/drivers/input/keyboard/pxa27x_keypad.c
@@ -835,7 +835,6 @@ static struct platform_driver pxa27x_keypad_driver = {
 	.driver		= {
 		.name	= "pxa27x-keypad",
 		.of_match_table = of_match_ptr(pxa27x_keypad_dt_match),
-		.owner	= THIS_MODULE,
 		.pm	= &pxa27x_keypad_pm_ops,
 	},
 };
diff --git a/drivers/input/keyboard/pxa930_rotary.c b/drivers/input/keyboard/pxa930_rotary.c
index 374ca0246c8f..1cf5211fddaa 100644
--- a/drivers/input/keyboard/pxa930_rotary.c
+++ b/drivers/input/keyboard/pxa930_rotary.c
@@ -189,7 +189,6 @@ static int pxa930_rotary_remove(struct platform_device *pdev)
 static struct platform_driver pxa930_rotary_driver = {
 	.driver		= {
 		.name	= "pxa930-rotary",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= pxa930_rotary_probe,
 	.remove		= pxa930_rotary_remove,
diff --git a/drivers/input/keyboard/samsung-keypad.c b/drivers/input/keyboard/samsung-keypad.c
index 5e80fbf7b5ed..6b9fdf6cf8e8 100644
--- a/drivers/input/keyboard/samsung-keypad.c
+++ b/drivers/input/keyboard/samsung-keypad.c
@@ -463,7 +463,7 @@ static int samsung_keypad_remove(struct platform_device *pdev)
 	return 0;
 }
 
-#ifdef CONFIG_PM_RUNTIME
+#ifdef CONFIG_PM
 static int samsung_keypad_runtime_suspend(struct device *dev)
 {
 	struct platform_device *pdev = to_platform_device(dev);
@@ -602,7 +602,6 @@ static struct platform_driver samsung_keypad_driver = {
 	.remove		= samsung_keypad_remove,
 	.driver		= {
 		.name	= "samsung-keypad",
-		.owner	= THIS_MODULE,
 		.of_match_table = of_match_ptr(samsung_keypad_dt_match),
 		.pm	= &samsung_keypad_pm_ops,
 	},
diff --git a/drivers/input/keyboard/spear-keyboard.c b/drivers/input/keyboard/spear-keyboard.c
index 258af10e5811..f42a543db043 100644
--- a/drivers/input/keyboard/spear-keyboard.c
+++ b/drivers/input/keyboard/spear-keyboard.c
@@ -385,7 +385,6 @@ static struct platform_driver spear_kbd_driver = {
 	.remove		= spear_kbd_remove,
 	.driver		= {
 		.name	= "keyboard",
-		.owner	= THIS_MODULE,
 		.pm	= &spear_kbd_pm_ops,
 		.of_match_table = of_match_ptr(spear_kbd_id_table),
 	},
diff --git a/drivers/input/keyboard/tc3589x-keypad.c b/drivers/input/keyboard/tc3589x-keypad.c
index ad7abae69078..8ff612d160b0 100644
--- a/drivers/input/keyboard/tc3589x-keypad.c
+++ b/drivers/input/keyboard/tc3589x-keypad.c
@@ -505,7 +505,6 @@ static SIMPLE_DEV_PM_OPS(tc3589x_keypad_dev_pm_ops,
 static struct platform_driver tc3589x_keypad_driver = {
 	.driver	= {
 		.name	= "tc3589x-keypad",
-		.owner	= THIS_MODULE,
 		.pm	= &tc3589x_keypad_dev_pm_ops,
 	},
 	.probe	= tc3589x_keypad_probe,
diff --git a/drivers/input/keyboard/tegra-kbc.c b/drivers/input/keyboard/tegra-kbc.c
index 9757a58bc897..f97c73bd14f8 100644
--- a/drivers/input/keyboard/tegra-kbc.c
+++ b/drivers/input/keyboard/tegra-kbc.c
@@ -822,7 +822,6 @@ static struct platform_driver tegra_kbc_driver = {
 	.probe		= tegra_kbc_probe,
 	.driver	= {
 		.name	= "tegra-kbc",
-		.owner  = THIS_MODULE,
 		.pm	= &tegra_kbc_pm_ops,
 		.of_match_table = tegra_kbc_of_match,
 	},
diff --git a/drivers/input/keyboard/twl4030_keypad.c b/drivers/input/keyboard/twl4030_keypad.c
index c5a11700a1bf..bbcccd67247d 100644
--- a/drivers/input/keyboard/twl4030_keypad.c
+++ b/drivers/input/keyboard/twl4030_keypad.c
@@ -459,7 +459,6 @@ static struct platform_driver twl4030_kp_driver = {
 	.probe		= twl4030_kp_probe,
 	.driver		= {
 		.name	= "twl4030_keypad",
-		.owner	= THIS_MODULE,
 		.of_match_table = of_match_ptr(twl4030_keypad_dt_match_table),
 	},
 };
diff --git a/drivers/input/keyboard/w90p910_keypad.c b/drivers/input/keyboard/w90p910_keypad.c
index e8b9d94daae7..a1ff69c53102 100644
--- a/drivers/input/keyboard/w90p910_keypad.c
+++ b/drivers/input/keyboard/w90p910_keypad.c
@@ -258,7 +258,6 @@ static struct platform_driver w90p910_keypad_driver = {
 	.remove		= w90p910_keypad_remove,
 	.driver		= {
 		.name	= "nuc900-kpi",
-		.owner	= THIS_MODULE,
 	},
 };
 module_platform_driver(w90p910_keypad_driver);
diff --git a/drivers/input/misc/88pm80x_onkey.c b/drivers/input/misc/88pm80x_onkey.c
index ee43e5b7c881..cf9908f1e5d5 100644
--- a/drivers/input/misc/88pm80x_onkey.c
+++ b/drivers/input/misc/88pm80x_onkey.c
@@ -153,7 +153,6 @@ static int pm80x_onkey_remove(struct platform_device *pdev)
 static struct platform_driver pm80x_onkey_driver = {
 	.driver = {
 		   .name = "88pm80x-onkey",
-		   .owner = THIS_MODULE,
 		   .pm = &pm80x_onkey_pm_ops,
 		   },
 	.probe = pm80x_onkey_probe,
diff --git a/drivers/input/misc/88pm860x_onkey.c b/drivers/input/misc/88pm860x_onkey.c
index 3cbd1b309220..cc87443aa2ee 100644
--- a/drivers/input/misc/88pm860x_onkey.c
+++ b/drivers/input/misc/88pm860x_onkey.c
@@ -136,7 +136,6 @@ static SIMPLE_DEV_PM_OPS(pm860x_onkey_pm_ops, pm860x_onkey_suspend, pm860x_onkey
 static struct platform_driver pm860x_onkey_driver = {
 	.driver		= {
 		.name	= "88pm860x-onkey",
-		.owner	= THIS_MODULE,
 		.pm	= &pm860x_onkey_pm_ops,
 	},
 	.probe		= pm860x_onkey_probe,
diff --git a/drivers/input/misc/ab8500-ponkey.c b/drivers/input/misc/ab8500-ponkey.c
index 95ef7dd6442d..1f7e15ca5fbe 100644
--- a/drivers/input/misc/ab8500-ponkey.c
+++ b/drivers/input/misc/ab8500-ponkey.c
@@ -123,7 +123,6 @@ static const struct of_device_id ab8500_ponkey_match[] = {
 static struct platform_driver ab8500_ponkey_driver = {
 	.driver		= {
 		.name	= "ab8500-poweron-key",
-		.owner	= THIS_MODULE,
 		.of_match_table = of_match_ptr(ab8500_ponkey_match),
 	},
 	.probe		= ab8500_ponkey_probe,
diff --git a/drivers/input/misc/arizona-haptics.c b/drivers/input/misc/arizona-haptics.c
index ef2e281b0a43..4dbbed74c9e4 100644
--- a/drivers/input/misc/arizona-haptics.c
+++ b/drivers/input/misc/arizona-haptics.c
@@ -225,7 +225,6 @@ static struct platform_driver arizona_haptics_driver = {
 	.remove		= arizona_haptics_remove,
 	.driver		= {
 		.name	= "arizona-haptics",
-		.owner	= THIS_MODULE,
 	},
 };
 module_platform_driver(arizona_haptics_driver);
diff --git a/drivers/input/misc/bfin_rotary.c b/drivers/input/misc/bfin_rotary.c
index e69d9bcb37e1..3f4351579372 100644
--- a/drivers/input/misc/bfin_rotary.c
+++ b/drivers/input/misc/bfin_rotary.c
@@ -256,7 +256,6 @@ static struct platform_driver bfin_rotary_device_driver = {
 	.remove		= bfin_rotary_remove,
 	.driver		= {
 		.name	= "bfin-rotary",
-		.owner	= THIS_MODULE,
 #ifdef CONFIG_PM
 		.pm	= &bfin_rotary_pm_ops,
 #endif
diff --git a/drivers/input/misc/cobalt_btns.c b/drivers/input/misc/cobalt_btns.c
index 3e11510ff82d..fbe72afc9347 100644
--- a/drivers/input/misc/cobalt_btns.c
+++ b/drivers/input/misc/cobalt_btns.c
@@ -157,7 +157,6 @@ static struct platform_driver cobalt_buttons_driver = {
 	.remove	= cobalt_buttons_remove,
 	.driver	= {
 		.name	= "Cobalt buttons",
-		.owner	= THIS_MODULE,
 	},
 };
 module_platform_driver(cobalt_buttons_driver);
diff --git a/drivers/input/misc/da9052_onkey.c b/drivers/input/misc/da9052_onkey.c
index 184c8f21ab59..266e07fdc182 100644
--- a/drivers/input/misc/da9052_onkey.c
+++ b/drivers/input/misc/da9052_onkey.c
@@ -149,7 +149,6 @@ static struct platform_driver da9052_onkey_driver = {
 	.remove	= da9052_onkey_remove,
 	.driver = {
 		.name	= "da9052-onkey",
-		.owner	= THIS_MODULE,
 	},
 };
 module_platform_driver(da9052_onkey_driver);
diff --git a/drivers/input/misc/da9055_onkey.c b/drivers/input/misc/da9055_onkey.c
index 4765799fef74..3251a9693f45 100644
--- a/drivers/input/misc/da9055_onkey.c
+++ b/drivers/input/misc/da9055_onkey.c
@@ -157,7 +157,6 @@ static struct platform_driver da9055_onkey_driver = {
 	.remove	= da9055_onkey_remove,
 	.driver = {
 		.name	= "da9055-onkey",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/input/misc/dm355evm_keys.c b/drivers/input/misc/dm355evm_keys.c
index 0eba94f581df..b6b7bd4e5462 100644
--- a/drivers/input/misc/dm355evm_keys.c
+++ b/drivers/input/misc/dm355evm_keys.c
@@ -263,7 +263,6 @@ static struct platform_driver dm355evm_keys_driver = {
 	.probe		= dm355evm_keys_probe,
 	.remove		= dm355evm_keys_remove,
 	.driver		= {
-		.owner	= THIS_MODULE,
 		.name	= "dm355evm_keys",
 	},
 };
diff --git a/drivers/input/misc/gpio-beeper.c b/drivers/input/misc/gpio-beeper.c
index 8886af63eae3..4817c5f0c3e4 100644
--- a/drivers/input/misc/gpio-beeper.c
+++ b/drivers/input/misc/gpio-beeper.c
@@ -112,7 +112,6 @@ MODULE_DEVICE_TABLE(of, gpio_beeper_of_match);
 static struct platform_driver gpio_beeper_platform_driver = {
 	.driver	= {
 		.name		= BEEPER_MODNAME,
-		.owner		= THIS_MODULE,
 		.of_match_table	= of_match_ptr(gpio_beeper_of_match),
 	},
 	.probe	= gpio_beeper_probe,
diff --git a/drivers/input/misc/gpio_tilt_polled.c b/drivers/input/misc/gpio_tilt_polled.c
index 1a81d9115226..f103b99d1852 100644
--- a/drivers/input/misc/gpio_tilt_polled.c
+++ b/drivers/input/misc/gpio_tilt_polled.c
@@ -199,7 +199,6 @@ static struct platform_driver gpio_tilt_polled_driver = {
 	.remove	= gpio_tilt_polled_remove,
 	.driver	= {
 		.name	= DRV_NAME,
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/input/misc/ideapad_slidebar.c b/drivers/input/misc/ideapad_slidebar.c
index edfd6239f131..b0acb878d1cf 100644
--- a/drivers/input/misc/ideapad_slidebar.c
+++ b/drivers/input/misc/ideapad_slidebar.c
@@ -272,7 +272,6 @@ static int ideapad_remove(struct platform_device *pdev)
 static struct platform_driver slidebar_drv = {
 	.driver = {
 		.name = "ideapad_slidebar",
-		.owner = THIS_MODULE,
 	},
 	.remove = ideapad_remove,
 };
diff --git a/drivers/input/misc/ixp4xx-beeper.c b/drivers/input/misc/ixp4xx-beeper.c
index ed8e5e8449d3..1fe149f3def2 100644
--- a/drivers/input/misc/ixp4xx-beeper.c
+++ b/drivers/input/misc/ixp4xx-beeper.c
@@ -168,7 +168,6 @@ static void ixp4xx_spkr_shutdown(struct platform_device *dev)
 static struct platform_driver ixp4xx_spkr_platform_driver = {
 	.driver		= {
 		.name	= "ixp4xx-beeper",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= ixp4xx_spkr_probe,
 	.remove		= ixp4xx_spkr_remove,
diff --git a/drivers/input/misc/m68kspkr.c b/drivers/input/misc/m68kspkr.c
index def21dc84522..312d63623038 100644
--- a/drivers/input/misc/m68kspkr.c
+++ b/drivers/input/misc/m68kspkr.c
@@ -100,7 +100,6 @@ static void m68kspkr_shutdown(struct platform_device *dev)
 static struct platform_driver m68kspkr_platform_driver = {
 	.driver		= {
 		.name	= "m68kspkr",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= m68kspkr_probe,
 	.remove		= m68kspkr_remove,
diff --git a/drivers/input/misc/max77693-haptic.c b/drivers/input/misc/max77693-haptic.c
index 157a1f50c015..39e930c10ebb 100644
--- a/drivers/input/misc/max77693-haptic.c
+++ b/drivers/input/misc/max77693-haptic.c
@@ -341,7 +341,6 @@ static SIMPLE_DEV_PM_OPS(max77693_haptic_pm_ops,
 static struct platform_driver max77693_haptic_driver = {
 	.driver		= {
 		.name	= "max77693-haptic",
-		.owner	= THIS_MODULE,
 		.pm	= &max77693_haptic_pm_ops,
 	},
 	.probe		= max77693_haptic_probe,
diff --git a/drivers/input/misc/max8925_onkey.c b/drivers/input/misc/max8925_onkey.c
index 11d0d3770239..7c49b8d23894 100644
--- a/drivers/input/misc/max8925_onkey.c
+++ b/drivers/input/misc/max8925_onkey.c
@@ -166,7 +166,6 @@ static SIMPLE_DEV_PM_OPS(max8925_onkey_pm_ops, max8925_onkey_suspend, max8925_on
 static struct platform_driver max8925_onkey_driver = {
 	.driver		= {
 		.name	= "max8925-onkey",
-		.owner	= THIS_MODULE,
 		.pm	= &max8925_onkey_pm_ops,
 	},
 	.probe		= max8925_onkey_probe,
diff --git a/drivers/input/misc/max8997_haptic.c b/drivers/input/misc/max8997_haptic.c
index 980437ac314d..d0f687281339 100644
--- a/drivers/input/misc/max8997_haptic.c
+++ b/drivers/input/misc/max8997_haptic.c
@@ -399,7 +399,6 @@ MODULE_DEVICE_TABLE(i2c, max8997_haptic_id);
 static struct platform_driver max8997_haptic_driver = {
 	.driver	= {
 		.name	= "max8997-haptic",
-		.owner	= THIS_MODULE,
 		.pm	= &max8997_haptic_pm_ops,
 	},
 	.probe		= max8997_haptic_probe,
diff --git a/drivers/input/misc/mc13783-pwrbutton.c b/drivers/input/misc/mc13783-pwrbutton.c
index 0df6e8d8bd03..afdf8ef25ee8 100644
--- a/drivers/input/misc/mc13783-pwrbutton.c
+++ b/drivers/input/misc/mc13783-pwrbutton.c
@@ -258,7 +258,6 @@ static struct platform_driver mc13783_pwrbutton_driver = {
 	.remove		= mc13783_pwrbutton_remove,
 	.driver		= {
 		.name	= "mc13783-pwrbutton",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/input/misc/palmas-pwrbutton.c b/drivers/input/misc/palmas-pwrbutton.c
index 2c4d6ca5faab..1f9b5ee92746 100644
--- a/drivers/input/misc/palmas-pwrbutton.c
+++ b/drivers/input/misc/palmas-pwrbutton.c
@@ -317,7 +317,6 @@ static struct platform_driver palmas_pwron_driver = {
 	.remove	= palmas_pwron_remove,
 	.driver	= {
 		.name	= "palmas_pwrbutton",
-		.owner	= THIS_MODULE,
 		.of_match_table = of_match_ptr(of_palmas_pwr_match),
 		.pm	= &palmas_pwron_pm,
 	},
diff --git a/drivers/input/misc/pcap_keys.c b/drivers/input/misc/pcap_keys.c
index cd230365166e..3b81daf67726 100644
--- a/drivers/input/misc/pcap_keys.c
+++ b/drivers/input/misc/pcap_keys.c
@@ -121,7 +121,6 @@ static struct platform_driver pcap_keys_device_driver = {
 	.remove		= pcap_keys_remove,
 	.driver		= {
 		.name	= "pcap-keys",
-		.owner	= THIS_MODULE,
 	}
 };
 module_platform_driver(pcap_keys_device_driver);
diff --git a/drivers/input/misc/pcspkr.c b/drivers/input/misc/pcspkr.c
index 674a2cfc3c0e..72b1fc3ab910 100644
--- a/drivers/input/misc/pcspkr.c
+++ b/drivers/input/misc/pcspkr.c
@@ -125,7 +125,6 @@ static const struct dev_pm_ops pcspkr_pm_ops = {
 static struct platform_driver pcspkr_platform_driver = {
 	.driver		= {
 		.name	= "pcspkr",
-		.owner	= THIS_MODULE,
 		.pm	= &pcspkr_pm_ops,
 	},
 	.probe		= pcspkr_probe,
diff --git a/drivers/input/misc/pm8xxx-vibrator.c b/drivers/input/misc/pm8xxx-vibrator.c
index e5266cd9acc0..5113877153d7 100644
--- a/drivers/input/misc/pm8xxx-vibrator.c
+++ b/drivers/input/misc/pm8xxx-vibrator.c
@@ -222,7 +222,6 @@ static struct platform_driver pm8xxx_vib_driver = {
 	.probe		= pm8xxx_vib_probe,
 	.driver		= {
 		.name	= "pm8xxx-vib",
-		.owner	= THIS_MODULE,
 		.pm	= &pm8xxx_vib_pm_ops,
 		.of_match_table = pm8xxx_vib_id_table,
 	},
diff --git a/drivers/input/misc/pmic8xxx-pwrkey.c b/drivers/input/misc/pmic8xxx-pwrkey.c
index 28999455c752..c4ca20e63221 100644
--- a/drivers/input/misc/pmic8xxx-pwrkey.c
+++ b/drivers/input/misc/pmic8xxx-pwrkey.c
@@ -193,7 +193,6 @@ static struct platform_driver pmic8xxx_pwrkey_driver = {
 	.remove		= pmic8xxx_pwrkey_remove,
 	.driver		= {
 		.name	= "pm8xxx-pwrkey",
-		.owner	= THIS_MODULE,
 		.pm	= &pm8xxx_pwr_key_pm_ops,
 		.of_match_table = pm8xxx_pwr_key_id_table,
 	},
diff --git a/drivers/input/misc/pwm-beeper.c b/drivers/input/misc/pwm-beeper.c
index 93f640a38d94..a28ee70ff158 100644
--- a/drivers/input/misc/pwm-beeper.c
+++ b/drivers/input/misc/pwm-beeper.c
@@ -187,7 +187,6 @@ static struct platform_driver pwm_beeper_driver = {
 	.remove = pwm_beeper_remove,
 	.driver = {
 		.name	= "pwm-beeper",
-		.owner	= THIS_MODULE,
 		.pm	= PWM_BEEPER_PM_OPS,
 		.of_match_table = of_match_ptr(pwm_beeper_match),
 	},
diff --git a/drivers/input/misc/rb532_button.c b/drivers/input/misc/rb532_button.c
index 83fff38b86b3..e956e81cd4e6 100644
--- a/drivers/input/misc/rb532_button.c
+++ b/drivers/input/misc/rb532_button.c
@@ -96,7 +96,6 @@ static struct platform_driver rb532_button_driver = {
 	.remove = rb532_button_remove,
 	.driver = {
 		.name = DRV_NAME,
-		.owner = THIS_MODULE,
 	},
 };
 module_platform_driver(rb532_button_driver);
diff --git a/drivers/input/misc/retu-pwrbutton.c b/drivers/input/misc/retu-pwrbutton.c
index 4bff1aa9b0db..0c8ac60e2639 100644
--- a/drivers/input/misc/retu-pwrbutton.c
+++ b/drivers/input/misc/retu-pwrbutton.c
@@ -85,7 +85,6 @@ static struct platform_driver retu_pwrbutton_driver = {
 	.remove		= retu_pwrbutton_remove,
 	.driver		= {
 		.name	= "retu-pwrbutton",
-		.owner	= THIS_MODULE,
 	},
 };
 module_platform_driver(retu_pwrbutton_driver);
diff --git a/drivers/input/misc/rotary_encoder.c b/drivers/input/misc/rotary_encoder.c
index 93558a1c7f70..f27f81ee84ed 100644
--- a/drivers/input/misc/rotary_encoder.c
+++ b/drivers/input/misc/rotary_encoder.c
@@ -325,7 +325,6 @@ static struct platform_driver rotary_encoder_driver = {
 	.remove		= rotary_encoder_remove,
 	.driver		= {
 		.name	= DRV_NAME,
-		.owner	= THIS_MODULE,
 		.of_match_table = of_match_ptr(rotary_encoder_of_match),
 	}
 };
diff --git a/drivers/input/misc/sgi_btns.c b/drivers/input/misc/sgi_btns.c
index f10474937a64..7bbe79d89f5c 100644
--- a/drivers/input/misc/sgi_btns.c
+++ b/drivers/input/misc/sgi_btns.c
@@ -157,7 +157,6 @@ static struct platform_driver sgi_buttons_driver = {
 	.remove	= sgi_buttons_remove,
 	.driver	= {
 		.name	= "sgibtns",
-		.owner	= THIS_MODULE,
 	},
 };
 module_platform_driver(sgi_buttons_driver);
diff --git a/drivers/input/misc/sirfsoc-onkey.c b/drivers/input/misc/sirfsoc-onkey.c
index 151f169afb7f..9d5b89befe6f 100644
--- a/drivers/input/misc/sirfsoc-onkey.c
+++ b/drivers/input/misc/sirfsoc-onkey.c
@@ -203,7 +203,6 @@ static struct platform_driver sirfsoc_pwrc_driver = {
 	.remove		= sirfsoc_pwrc_remove,
 	.driver		= {
 		.name	= "sirfsoc-pwrc",
-		.owner	= THIS_MODULE,
 		.pm	= &sirfsoc_pwrc_pm_ops,
 		.of_match_table = sirfsoc_pwrc_of_match,
 	}
diff --git a/drivers/input/misc/soc_button_array.c b/drivers/input/misc/soc_button_array.c
index e097f1ab427f..79cc0f79896f 100644
--- a/drivers/input/misc/soc_button_array.c
+++ b/drivers/input/misc/soc_button_array.c
@@ -214,7 +214,6 @@ static struct platform_driver soc_button_driver = {
 	.remove		= soc_button_remove,
 	.driver		= {
 		.name = KBUILD_MODNAME,
-		.owner = THIS_MODULE,
 		.acpi_match_table = ACPI_PTR(soc_button_acpi_match),
 	},
 };
diff --git a/drivers/input/misc/sparcspkr.c b/drivers/input/misc/sparcspkr.c
index 179ff1cd6f6b..54116e544c96 100644
--- a/drivers/input/misc/sparcspkr.c
+++ b/drivers/input/misc/sparcspkr.c
@@ -257,7 +257,6 @@ static const struct of_device_id bbc_beep_match[] = {
 static struct platform_driver bbc_beep_driver = {
 	.driver = {
 		.name = "bbcbeep",
-		.owner = THIS_MODULE,
 		.of_match_table = bbc_beep_match,
 	},
 	.probe		= bbc_beep_probe,
@@ -337,7 +336,6 @@ static const struct of_device_id grover_beep_match[] = {
 static struct platform_driver grover_beep_driver = {
 	.driver = {
 		.name = "groverbeep",
-		.owner = THIS_MODULE,
 		.of_match_table = grover_beep_match,
 	},
 	.probe		= grover_beep_probe,
diff --git a/drivers/input/misc/twl4030-pwrbutton.c b/drivers/input/misc/twl4030-pwrbutton.c
index 8400a1a34d87..e98cc81a84c6 100644
--- a/drivers/input/misc/twl4030-pwrbutton.c
+++ b/drivers/input/misc/twl4030-pwrbutton.c
@@ -102,7 +102,6 @@ static struct platform_driver twl4030_pwrbutton_driver = {
 	.probe		= twl4030_pwrbutton_probe,
 	.driver		= {
 		.name	= "twl4030_pwrbutton",
-		.owner	= THIS_MODULE,
 		.of_match_table = of_match_ptr(twl4030_pwrbutton_dt_match_table),
 	},
 };
diff --git a/drivers/input/misc/twl4030-vibra.c b/drivers/input/misc/twl4030-vibra.c
index c802c4af480c..fc17b9592f54 100644
--- a/drivers/input/misc/twl4030-vibra.c
+++ b/drivers/input/misc/twl4030-vibra.c
@@ -251,7 +251,6 @@ static struct platform_driver twl4030_vibra_driver = {
 	.probe		= twl4030_vibra_probe,
 	.driver		= {
 		.name	= "twl4030-vibra",
-		.owner	= THIS_MODULE,
 		.pm	= &twl4030_vibra_pm_ops,
 	},
 };
diff --git a/drivers/input/misc/twl6040-vibra.c b/drivers/input/misc/twl6040-vibra.c
index 190fdef06e8f..0e0d094df2e6 100644
--- a/drivers/input/misc/twl6040-vibra.c
+++ b/drivers/input/misc/twl6040-vibra.c
@@ -386,7 +386,6 @@ static struct platform_driver twl6040_vibra_driver = {
 	.probe		= twl6040_vibra_probe,
 	.driver		= {
 		.name	= "twl6040-vibra",
-		.owner	= THIS_MODULE,
 		.pm	= &twl6040_vibra_pm_ops,
 	},
 };
diff --git a/drivers/input/misc/wistron_btns.c b/drivers/input/misc/wistron_btns.c
index 7b7add5061a5..e25f87ba19f6 100644
--- a/drivers/input/misc/wistron_btns.c
+++ b/drivers/input/misc/wistron_btns.c
@@ -1347,7 +1347,6 @@ static const struct dev_pm_ops wistron_pm_ops = {
 static struct platform_driver wistron_driver = {
 	.driver		= {
 		.name	= "wistron-bios",
-		.owner	= THIS_MODULE,
 #ifdef CONFIG_PM
 		.pm	= &wistron_pm_ops,
 #endif
diff --git a/drivers/input/misc/wm831x-on.c b/drivers/input/misc/wm831x-on.c
index 173b6dcca0da..59d4f7bcb4a3 100644
--- a/drivers/input/misc/wm831x-on.c
+++ b/drivers/input/misc/wm831x-on.c
@@ -138,7 +138,6 @@ static struct platform_driver wm831x_on_driver = {
 	.remove		= wm831x_on_remove,
 	.driver		= {
 		.name	= "wm831x-on",
-		.owner	= THIS_MODULE,
 	},
 };
 module_platform_driver(wm831x_on_driver);
diff --git a/drivers/input/misc/xen-kbdfront.c b/drivers/input/misc/xen-kbdfront.c
index e2ecfc6e633d..95599e478e19 100644
--- a/drivers/input/misc/xen-kbdfront.c
+++ b/drivers/input/misc/xen-kbdfront.c
@@ -365,12 +365,13 @@ static const struct xenbus_device_id xenkbd_ids[] = {
 	{ "" }
 };
 
-static DEFINE_XENBUS_DRIVER(xenkbd, ,
+static struct xenbus_driver xenkbd_driver = {
+	.ids = xenkbd_ids,
 	.probe = xenkbd_probe,
 	.remove = xenkbd_remove,
 	.resume = xenkbd_resume,
 	.otherend_changed = xenkbd_backend_changed,
-);
+};
 
 static int __init xenkbd_init(void)
 {
diff --git a/drivers/input/mouse/amimouse.c b/drivers/input/mouse/amimouse.c
index 62ec52b2e347..a7fd8f22ba56 100644
--- a/drivers/input/mouse/amimouse.c
+++ b/drivers/input/mouse/amimouse.c
@@ -141,7 +141,6 @@ static struct platform_driver amimouse_driver = {
 	.remove = __exit_p(amimouse_remove),
 	.driver   = {
 		.name	= "amiga-mouse",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/input/mouse/gpio_mouse.c b/drivers/input/mouse/gpio_mouse.c
index 8c7d94200bdb..ced07391304b 100644
--- a/drivers/input/mouse/gpio_mouse.c
+++ b/drivers/input/mouse/gpio_mouse.c
@@ -171,7 +171,6 @@ static struct platform_driver gpio_mouse_device_driver = {
 	.remove		= gpio_mouse_remove,
 	.driver		= {
 		.name	= "gpio_mouse",
-		.owner	= THIS_MODULE,
 	}
 };
 module_platform_driver(gpio_mouse_device_driver);
diff --git a/drivers/input/mouse/navpoint.c b/drivers/input/mouse/navpoint.c
index 69017279e8d7..d6e8f58a1de3 100644
--- a/drivers/input/mouse/navpoint.c
+++ b/drivers/input/mouse/navpoint.c
@@ -353,7 +353,6 @@ static struct platform_driver navpoint_driver = {
 	.remove		= navpoint_remove,
 	.driver = {
 		.name	= "navpoint",
-		.owner	= THIS_MODULE,
 		.pm	= &navpoint_pm_ops,
 	},
 };
diff --git a/drivers/input/serio/altera_ps2.c b/drivers/input/serio/altera_ps2.c
index f50325a03d4a..131d7826dc6b 100644
--- a/drivers/input/serio/altera_ps2.c
+++ b/drivers/input/serio/altera_ps2.c
@@ -156,7 +156,6 @@ static struct platform_driver altera_ps2_driver = {
 	.remove		= altera_ps2_remove,
 	.driver	= {
 		.name	= DRV_NAME,
-		.owner	= THIS_MODULE,
 		.of_match_table = of_match_ptr(altera_ps2_match),
 	},
 };
diff --git a/drivers/input/serio/apbps2.c b/drivers/input/serio/apbps2.c
index 98be824544a5..45d4e08ca4f8 100644
--- a/drivers/input/serio/apbps2.c
+++ b/drivers/input/serio/apbps2.c
@@ -214,7 +214,6 @@ MODULE_DEVICE_TABLE(of, apbps2_of_match);
 static struct platform_driver apbps2_of_driver = {
 	.driver = {
 		.name = "grlib-apbps2",
-		.owner = THIS_MODULE,
 		.of_match_table = apbps2_of_match,
 	},
 	.probe = apbps2_of_probe,
diff --git a/drivers/input/serio/arc_ps2.c b/drivers/input/serio/arc_ps2.c
index 8024a6d7fccb..99e57a418753 100644
--- a/drivers/input/serio/arc_ps2.c
+++ b/drivers/input/serio/arc_ps2.c
@@ -266,7 +266,6 @@ MODULE_DEVICE_TABLE(of, arc_ps2_match);
 static struct platform_driver arc_ps2_driver = {
 	.driver	= {
 		.name		= "arc_ps2",
-		.owner		= THIS_MODULE,
 		.of_match_table	= of_match_ptr(arc_ps2_match),
 	},
 	.probe	= arc_ps2_probe,
diff --git a/drivers/input/serio/at32psif.c b/drivers/input/serio/at32psif.c
index 3290b287ac4b..2e4ff5bac754 100644
--- a/drivers/input/serio/at32psif.c
+++ b/drivers/input/serio/at32psif.c
@@ -352,7 +352,6 @@ static struct platform_driver psif_driver = {
 	.remove		= __exit_p(psif_remove),
 	.driver		= {
 		.name	= "atmel_psif",
-		.owner	= THIS_MODULE,
 		.pm	= &psif_pm_ops,
 	},
 };
diff --git a/drivers/input/serio/ct82c710.c b/drivers/input/serio/ct82c710.c
index cfe549d4eaa5..9c54c43c9749 100644
--- a/drivers/input/serio/ct82c710.c
+++ b/drivers/input/serio/ct82c710.c
@@ -209,7 +209,6 @@ static int ct82c710_remove(struct platform_device *dev)
 static struct platform_driver ct82c710_driver = {
 	.driver		= {
 		.name	= "ct82c710",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= ct82c710_probe,
 	.remove		= ct82c710_remove,
diff --git a/drivers/input/serio/i8042-sparcio.h b/drivers/input/serio/i8042-sparcio.h
index 93cb7912703c..afcd1c1a05b2 100644
--- a/drivers/input/serio/i8042-sparcio.h
+++ b/drivers/input/serio/i8042-sparcio.h
@@ -99,7 +99,6 @@ MODULE_DEVICE_TABLE(of, sparc_i8042_match);
 static struct platform_driver sparc_i8042_driver = {
 	.driver = {
 		.name = "i8042",
-		.owner = THIS_MODULE,
 		.of_match_table = sparc_i8042_match,
 	},
 	.probe		= sparc_i8042_probe,
diff --git a/drivers/input/serio/i8042.c b/drivers/input/serio/i8042.c
index 804d2e02010a..986a71c614b0 100644
--- a/drivers/input/serio/i8042.c
+++ b/drivers/input/serio/i8042.c
@@ -1477,7 +1477,6 @@ static int i8042_remove(struct platform_device *dev)
 static struct platform_driver i8042_driver = {
 	.driver		= {
 		.name	= "i8042",
-		.owner	= THIS_MODULE,
 #ifdef CONFIG_PM
 		.pm	= &i8042_pm_ops,
 #endif
diff --git a/drivers/input/serio/maceps2.c b/drivers/input/serio/maceps2.c
index bc85e1cc66d8..e365c5f4cbc9 100644
--- a/drivers/input/serio/maceps2.c
+++ b/drivers/input/serio/maceps2.c
@@ -162,7 +162,6 @@ static int maceps2_remove(struct platform_device *dev)
 static struct platform_driver maceps2_driver = {
 	.driver		= {
 		.name	= "maceps2",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= maceps2_probe,
 	.remove		= maceps2_remove,
diff --git a/drivers/input/serio/olpc_apsp.c b/drivers/input/serio/olpc_apsp.c
index d906f3ebc8c8..8e9a4209fcad 100644
--- a/drivers/input/serio/olpc_apsp.c
+++ b/drivers/input/serio/olpc_apsp.c
@@ -273,7 +273,6 @@ static struct platform_driver olpc_apsp_driver = {
 	.remove		= olpc_apsp_remove,
 	.driver		= {
 		.name	= "olpc-apsp",
-		.owner	= THIS_MODULE,
 		.of_match_table = olpc_apsp_dt_ids,
 	},
 };
diff --git a/drivers/input/serio/q40kbd.c b/drivers/input/serio/q40kbd.c
index 594256c38554..5a9d521510bf 100644
--- a/drivers/input/serio/q40kbd.c
+++ b/drivers/input/serio/q40kbd.c
@@ -186,7 +186,6 @@ static int q40kbd_remove(struct platform_device *pdev)
 static struct platform_driver q40kbd_driver = {
 	.driver		= {
 		.name	= "q40kbd",
-		.owner	= THIS_MODULE,
 	},
 	.remove		= q40kbd_remove,
 };
diff --git a/drivers/input/serio/rpckbd.c b/drivers/input/serio/rpckbd.c
index e462e7791bb8..8cf964736902 100644
--- a/drivers/input/serio/rpckbd.c
+++ b/drivers/input/serio/rpckbd.c
@@ -168,7 +168,6 @@ static struct platform_driver rpckbd_driver = {
 	.remove		= rpckbd_remove,
 	.driver		= {
 		.name	= "kart",
-		.owner	= THIS_MODULE,
 	},
 };
 module_platform_driver(rpckbd_driver);
diff --git a/drivers/input/serio/xilinx_ps2.c b/drivers/input/serio/xilinx_ps2.c
index e6cf52ebad87..5223cbf94262 100644
--- a/drivers/input/serio/xilinx_ps2.c
+++ b/drivers/input/serio/xilinx_ps2.c
@@ -363,7 +363,6 @@ MODULE_DEVICE_TABLE(of, xps2_of_match);
 static struct platform_driver xps2_of_driver = {
 	.driver = {
 		.name = DRIVER_NAME,
-		.owner = THIS_MODULE,
 		.of_match_table = xps2_of_match,
 	},
 	.probe		= xps2_of_probe,
diff --git a/drivers/input/touchscreen/88pm860x-ts.c b/drivers/input/touchscreen/88pm860x-ts.c
index 0d4a9fad4a78..251ff2aa0633 100644
--- a/drivers/input/touchscreen/88pm860x-ts.c
+++ b/drivers/input/touchscreen/88pm860x-ts.c
@@ -292,7 +292,6 @@ static int pm860x_touch_probe(struct platform_device *pdev)
 static struct platform_driver pm860x_touch_driver = {
 	.driver	= {
 		.name	= "88pm860x-touch",
-		.owner	= THIS_MODULE,
 	},
 	.probe	= pm860x_touch_probe,
 };
diff --git a/drivers/input/touchscreen/ad7879.c b/drivers/input/touchscreen/ad7879.c
index 927feec1e856..fec66ad80513 100644
--- a/drivers/input/touchscreen/ad7879.c
+++ b/drivers/input/touchscreen/ad7879.c
@@ -468,14 +468,10 @@ static int ad7879_gpio_add(struct ad7879 *ts,
 static void ad7879_gpio_remove(struct ad7879 *ts)
 {
 	const struct ad7879_platform_data *pdata = dev_get_platdata(ts->dev);
-	int ret;
 
-	if (pdata->gpio_export) {
-		ret = gpiochip_remove(&ts->gc);
-		if (ret)
-			dev_err(ts->dev, "failed to remove gpio %d\n",
-				ts->gc.base);
-	}
+	if (pdata->gpio_export)
+		gpiochip_remove(&ts->gc);
+
 }
 #else
 static inline int ad7879_gpio_add(struct ad7879 *ts,
diff --git a/drivers/input/touchscreen/atmel-wm97xx.c b/drivers/input/touchscreen/atmel-wm97xx.c
index 279c0e42b8a7..7ec0421c0dd8 100644
--- a/drivers/input/touchscreen/atmel-wm97xx.c
+++ b/drivers/input/touchscreen/atmel-wm97xx.c
@@ -425,7 +425,6 @@ static struct platform_driver atmel_wm97xx_driver = {
 	.remove		= __exit_p(atmel_wm97xx_remove),
 	.driver		= {
 		.name	= "wm97xx-touch",
-		.owner	= THIS_MODULE,
 		.pm	= &atmel_wm97xx_pm_ops,
 	},
 };
diff --git a/drivers/input/touchscreen/atmel_mxt_ts.c b/drivers/input/touchscreen/atmel_mxt_ts.c
index bb070206223c..95ee92a91bd2 100644
--- a/drivers/input/touchscreen/atmel_mxt_ts.c
+++ b/drivers/input/touchscreen/atmel_mxt_ts.c
@@ -99,13 +99,9 @@
 #define MXT_T6_STATUS_COMSERR	(1 << 2)
 
 /* MXT_GEN_POWER_T7 field */
-struct t7_config {
-	u8 idle;
-	u8 active;
-} __packed;
-
-#define MXT_POWER_CFG_RUN		0
-#define MXT_POWER_CFG_DEEPSLEEP		1
+#define MXT_POWER_IDLEACQINT	0
+#define MXT_POWER_ACTVACQINT	1
+#define MXT_POWER_ACTV2IDLETO	2
 
 /* MXT_GEN_ACQUIRE_T8 field */
 #define MXT_ACQUIRE_CHRGTIME	0
@@ -117,6 +113,7 @@ struct t7_config {
 #define MXT_ACQUIRE_ATCHCALSTHR	7
 
 /* MXT_TOUCH_MULTI_T9 field */
+#define MXT_TOUCH_CTRL		0
 #define MXT_T9_ORIENT		9
 #define MXT_T9_RANGE		18
 
@@ -256,7 +253,6 @@ struct mxt_data {
 	bool update_input;
 	u8 last_message_count;
 	u8 num_touchids;
-	struct t7_config t7_cfg;
 
 	/* Cached parameters from object table */
 	u16 T5_address;
@@ -672,6 +668,20 @@ static void mxt_proc_t6_messages(struct mxt_data *data, u8 *msg)
 	data->t6_status = status;
 }
 
+static int mxt_write_object(struct mxt_data *data,
+				 u8 type, u8 offset, u8 val)
+{
+	struct mxt_object *object;
+	u16 reg;
+
+	object = mxt_get_object(data, type);
+	if (!object || offset >= mxt_obj_size(object))
+		return -EINVAL;
+
+	reg = object->start_address;
+	return mxt_write_reg(data->client, reg + offset, val);
+}
+
 static void mxt_input_button(struct mxt_data *data, u8 *message)
 {
 	struct input_dev *input = data->input_dev;
@@ -1742,60 +1752,6 @@ err_free_object_table:
 	return error;
 }
 
-static int mxt_set_t7_power_cfg(struct mxt_data *data, u8 sleep)
-{
-	struct device *dev = &data->client->dev;
-	int error;
-	struct t7_config *new_config;
-	struct t7_config deepsleep = { .active = 0, .idle = 0 };
-
-	if (sleep == MXT_POWER_CFG_DEEPSLEEP)
-		new_config = &deepsleep;
-	else
-		new_config = &data->t7_cfg;
-
-	error = __mxt_write_reg(data->client, data->T7_address,
-				sizeof(data->t7_cfg), new_config);
-	if (error)
-		return error;
-
-	dev_dbg(dev, "Set T7 ACTV:%d IDLE:%d\n",
-		new_config->active, new_config->idle);
-
-	return 0;
-}
-
-static int mxt_init_t7_power_cfg(struct mxt_data *data)
-{
-	struct device *dev = &data->client->dev;
-	int error;
-	bool retry = false;
-
-recheck:
-	error = __mxt_read_reg(data->client, data->T7_address,
-				sizeof(data->t7_cfg), &data->t7_cfg);
-	if (error)
-		return error;
-
-	if (data->t7_cfg.active == 0 || data->t7_cfg.idle == 0) {
-		if (!retry) {
-			dev_dbg(dev, "T7 cfg zero, resetting\n");
-			mxt_soft_reset(data);
-			retry = true;
-			goto recheck;
-		} else {
-			dev_dbg(dev, "T7 cfg zero after reset, overriding\n");
-			data->t7_cfg.active = 20;
-			data->t7_cfg.idle = 100;
-			return mxt_set_t7_power_cfg(data, MXT_POWER_CFG_RUN);
-		}
-	}
-
-	dev_dbg(dev, "Initialized power cfg: ACTV %d, IDLE %d\n",
-		data->t7_cfg.active, data->t7_cfg.idle);
-	return 0;
-}
-
 static int mxt_configure_objects(struct mxt_data *data,
 				 const struct firmware *cfg)
 {
@@ -1809,12 +1765,6 @@ static int mxt_configure_objects(struct mxt_data *data,
 			dev_warn(dev, "Error %d updating config\n", error);
 	}
 
-	error = mxt_init_t7_power_cfg(data);
-	if (error) {
-		dev_err(dev, "Failed to initialize power cfg\n");
-		return error;
-	}
-
 	error = mxt_initialize_t9_input_device(data);
 	if (error)
 		return error;
@@ -2093,15 +2043,16 @@ static const struct attribute_group mxt_attr_group = {
 
 static void mxt_start(struct mxt_data *data)
 {
-	mxt_set_t7_power_cfg(data, MXT_POWER_CFG_RUN);
-
-	/* Recalibrate since chip has been in deep sleep */
-	mxt_t6_command(data, MXT_COMMAND_CALIBRATE, 1, false);
+	/* Touch enable */
+	mxt_write_object(data,
+			MXT_TOUCH_MULTI_T9, MXT_TOUCH_CTRL, 0x83);
 }
 
 static void mxt_stop(struct mxt_data *data)
 {
-	mxt_set_t7_power_cfg(data, MXT_POWER_CFG_DEEPSLEEP);
+	/* Touch disable */
+	mxt_write_object(data,
+			MXT_TOUCH_MULTI_T9, MXT_TOUCH_CTRL, 0);
 }
 
 static int mxt_input_open(struct input_dev *dev)
@@ -2266,6 +2217,8 @@ static int __maybe_unused mxt_resume(struct device *dev)
 	struct mxt_data *data = i2c_get_clientdata(client);
 	struct input_dev *input_dev = data->input_dev;
 
+	mxt_soft_reset(data);
+
 	mutex_lock(&input_dev->mutex);
 
 	if (input_dev->users)
diff --git a/drivers/input/touchscreen/cyttsp4_core.c b/drivers/input/touchscreen/cyttsp4_core.c
index a035a390f8e2..568a3d340c8a 100644
--- a/drivers/input/touchscreen/cyttsp4_core.c
+++ b/drivers/input/touchscreen/cyttsp4_core.c
@@ -1716,7 +1716,7 @@ static void cyttsp4_free_si_ptrs(struct cyttsp4 *cd)
 	kfree(si->btn_rec_data);
 }
 
-#if defined(CONFIG_PM_SLEEP) || defined(CONFIG_PM_RUNTIME)
+#ifdef CONFIG_PM
 static int cyttsp4_core_sleep(struct cyttsp4 *cd)
 {
 	int rc;
diff --git a/drivers/input/touchscreen/da9034-ts.c b/drivers/input/touchscreen/da9034-ts.c
index cf6f4b31db4d..8264822dc4b9 100644
--- a/drivers/input/touchscreen/da9034-ts.c
+++ b/drivers/input/touchscreen/da9034-ts.c
@@ -357,7 +357,6 @@ static int da9034_touch_probe(struct platform_device *pdev)
 static struct platform_driver da9034_touch_driver = {
 	.driver	= {
 		.name	= "da9034-touch",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= da9034_touch_probe,
 };
diff --git a/drivers/input/touchscreen/da9052_tsi.c b/drivers/input/touchscreen/da9052_tsi.c
index ab64d58c3ac0..5a013bb7bcad 100644
--- a/drivers/input/touchscreen/da9052_tsi.c
+++ b/drivers/input/touchscreen/da9052_tsi.c
@@ -337,7 +337,6 @@ static struct platform_driver da9052_tsi_driver = {
 	.remove	= da9052_ts_remove,
 	.driver	= {
 		.name	= "da9052-tsi",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/input/touchscreen/edt-ft5x06.c b/drivers/input/touchscreen/edt-ft5x06.c
index dcc68efd0d7f..d4c24fb7704f 100644
--- a/drivers/input/touchscreen/edt-ft5x06.c
+++ b/drivers/input/touchscreen/edt-ft5x06.c
@@ -812,7 +812,7 @@ static int edt_ft5x06_ts_identify(struct i2c_client *client,
 	/* if we find something consistent, stay with that assumption
 	 * at least M09 won't send 3 bytes here
 	 */
-	if (!(strnicmp(rdbuf + 1, "EP0", 3))) {
+	if (!(strncasecmp(rdbuf + 1, "EP0", 3))) {
 		tsdata->version = M06;
 
 		/* remove last '$' end marker */
diff --git a/drivers/input/touchscreen/intel-mid-touch.c b/drivers/input/touchscreen/intel-mid-touch.c
index c38ca4a7e386..b4f0725a1c3d 100644
--- a/drivers/input/touchscreen/intel-mid-touch.c
+++ b/drivers/input/touchscreen/intel-mid-touch.c
@@ -644,7 +644,6 @@ static int mrstouch_probe(struct platform_device *pdev)
 static struct platform_driver mrstouch_driver = {
 	.driver = {
 		.name	= "pmic_touch",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= mrstouch_probe,
 };
diff --git a/drivers/input/touchscreen/jornada720_ts.c b/drivers/input/touchscreen/jornada720_ts.c
index 651ec71a5c68..ea3b6a5b83e6 100644
--- a/drivers/input/touchscreen/jornada720_ts.c
+++ b/drivers/input/touchscreen/jornada720_ts.c
@@ -148,7 +148,6 @@ static struct platform_driver jornada720_ts_driver = {
 	.probe		= jornada720_ts_probe,
 	.driver		= {
 		.name	= "jornada_ts",
-		.owner	= THIS_MODULE,
 	},
 };
 module_platform_driver(jornada720_ts_driver);
diff --git a/drivers/input/touchscreen/lpc32xx_ts.c b/drivers/input/touchscreen/lpc32xx_ts.c
index bb47d3442a35..24d704cd9f88 100644
--- a/drivers/input/touchscreen/lpc32xx_ts.c
+++ b/drivers/input/touchscreen/lpc32xx_ts.c
@@ -396,7 +396,6 @@ static struct platform_driver lpc32xx_ts_driver = {
 	.remove		= lpc32xx_ts_remove,
 	.driver		= {
 		.name	= MOD_NAME,
-		.owner	= THIS_MODULE,
 		.pm	= LPC32XX_TS_PM_OPS,
 		.of_match_table = of_match_ptr(lpc32xx_tsc_of_match),
 	},
diff --git a/drivers/input/touchscreen/mc13783_ts.c b/drivers/input/touchscreen/mc13783_ts.c
index d6f099c47f84..913e25a994b4 100644
--- a/drivers/input/touchscreen/mc13783_ts.c
+++ b/drivers/input/touchscreen/mc13783_ts.c
@@ -243,7 +243,6 @@ static int mc13783_ts_remove(struct platform_device *pdev)
 static struct platform_driver mc13783_ts_driver = {
 	.remove		= mc13783_ts_remove,
 	.driver		= {
-		.owner	= THIS_MODULE,
 		.name	= MC13783_TS_NAME,
 	},
 };
diff --git a/drivers/input/touchscreen/pcap_ts.c b/drivers/input/touchscreen/pcap_ts.c
index cff2376817e5..23a354a392ae 100644
--- a/drivers/input/touchscreen/pcap_ts.c
+++ b/drivers/input/touchscreen/pcap_ts.c
@@ -247,7 +247,6 @@ static struct platform_driver pcap_ts_driver = {
 	.remove		= pcap_ts_remove,
 	.driver		= {
 		.name	= "pcap-ts",
-		.owner	= THIS_MODULE,
 		.pm	= PCAP_TS_PM_OPS,
 	},
 };
diff --git a/drivers/input/touchscreen/s3c2410_ts.c b/drivers/input/touchscreen/s3c2410_ts.c
index 5a69ded9b53c..bdfa27dc097b 100644
--- a/drivers/input/touchscreen/s3c2410_ts.c
+++ b/drivers/input/touchscreen/s3c2410_ts.c
@@ -422,7 +422,6 @@ MODULE_DEVICE_TABLE(platform, s3cts_driver_ids);
 static struct platform_driver s3c_ts_driver = {
 	.driver         = {
 		.name   = "samsung-ts",
-		.owner  = THIS_MODULE,
 #ifdef CONFIG_PM
 		.pm	= &s3c_ts_pmops,
 #endif
diff --git a/drivers/input/touchscreen/stmpe-ts.c b/drivers/input/touchscreen/stmpe-ts.c
index 42ce31afa259..2d5ff86b343f 100644
--- a/drivers/input/touchscreen/stmpe-ts.c
+++ b/drivers/input/touchscreen/stmpe-ts.c
@@ -384,7 +384,6 @@ static int stmpe_ts_remove(struct platform_device *pdev)
 static struct platform_driver stmpe_ts_driver = {
 	.driver = {
 		   .name = STMPE_TS_NAME,
-		   .owner = THIS_MODULE,
 		   },
 	.probe = stmpe_input_probe,
 	.remove = stmpe_ts_remove,
diff --git a/drivers/input/touchscreen/sun4i-ts.c b/drivers/input/touchscreen/sun4i-ts.c
index 2ba826024954..28a06749ae42 100644
--- a/drivers/input/touchscreen/sun4i-ts.c
+++ b/drivers/input/touchscreen/sun4i-ts.c
@@ -324,7 +324,6 @@ MODULE_DEVICE_TABLE(of, sun4i_ts_of_match);
 
 static struct platform_driver sun4i_ts_driver = {
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= "sun4i-ts",
 		.of_match_table = of_match_ptr(sun4i_ts_of_match),
 	},
diff --git a/drivers/input/touchscreen/ti_am335x_tsc.c b/drivers/input/touchscreen/ti_am335x_tsc.c
index 2ce649520fe0..004f1346a957 100644
--- a/drivers/input/touchscreen/ti_am335x_tsc.c
+++ b/drivers/input/touchscreen/ti_am335x_tsc.c
@@ -518,7 +518,6 @@ static struct platform_driver ti_tsc_driver = {
 	.remove	= titsc_remove,
 	.driver	= {
 		.name   = "TI-am335x-tsc",
-		.owner	= THIS_MODULE,
 		.pm	= TITSC_PM_OPS,
 		.of_match_table = ti_tsc_dt_ids,
 	},
diff --git a/drivers/input/touchscreen/tps6507x-ts.c b/drivers/input/touchscreen/tps6507x-ts.c
index 94cde2cb1491..4ffd829d1990 100644
--- a/drivers/input/touchscreen/tps6507x-ts.c
+++ b/drivers/input/touchscreen/tps6507x-ts.c
@@ -314,7 +314,6 @@ static int tps6507x_ts_remove(struct platform_device *pdev)
 static struct platform_driver tps6507x_ts_driver = {
 	.driver = {
 		.name = "tps6507x-ts",
-		.owner = THIS_MODULE,
 	},
 	.probe = tps6507x_ts_probe,
 	.remove = tps6507x_ts_remove,
diff --git a/drivers/input/touchscreen/ucb1400_ts.c b/drivers/input/touchscreen/ucb1400_ts.c
index bf132c45af1a..c1e23cfc6155 100644
--- a/drivers/input/touchscreen/ucb1400_ts.c
+++ b/drivers/input/touchscreen/ucb1400_ts.c
@@ -442,7 +442,6 @@ static struct platform_driver ucb1400_ts_driver = {
 	.remove	= ucb1400_ts_remove,
 	.driver	= {
 		.name	= "ucb1400_ts",
-		.owner	= THIS_MODULE,
 		.pm	= &ucb1400_ts_pm_ops,
 	},
 };
diff --git a/drivers/input/touchscreen/w90p910_ts.c b/drivers/input/touchscreen/w90p910_ts.c
index 003d0c3b5d08..da6004e97753 100644
--- a/drivers/input/touchscreen/w90p910_ts.c
+++ b/drivers/input/touchscreen/w90p910_ts.c
@@ -326,7 +326,6 @@ static struct platform_driver w90x900ts_driver = {
 	.remove		= w90x900ts_remove,
 	.driver		= {
 		.name	= "nuc900-ts",
-		.owner	= THIS_MODULE,
 	},
 };
 module_platform_driver(w90x900ts_driver);
diff --git a/drivers/input/touchscreen/wm831x-ts.c b/drivers/input/touchscreen/wm831x-ts.c
index 1b953a066b2c..1db0a1410929 100644
--- a/drivers/input/touchscreen/wm831x-ts.c
+++ b/drivers/input/touchscreen/wm831x-ts.c
@@ -392,7 +392,6 @@ static int wm831x_ts_remove(struct platform_device *pdev)
 static struct platform_driver wm831x_ts_driver = {
 	.driver = {
 		.name = "wm831x-touch",
-		.owner = THIS_MODULE,
 	},
 	.probe = wm831x_ts_probe,
 	.remove = wm831x_ts_remove,
diff --git a/drivers/iommu/Kconfig b/drivers/iommu/Kconfig
index dd5112265cc9..325188eef1c1 100644
--- a/drivers/iommu/Kconfig
+++ b/drivers/iommu/Kconfig
@@ -15,7 +15,7 @@ if IOMMU_SUPPORT
 
 config OF_IOMMU
        def_bool y
-       depends on OF
+       depends on OF && IOMMU_API
 
 config FSL_PAMU
 	bool "Freescale IOMMU support"
@@ -144,13 +144,26 @@ config OMAP_IOMMU
 	select IOMMU_API
 
 config OMAP_IOMMU_DEBUG
-       tristate "Export OMAP IOMMU internals in DebugFS"
-       depends on OMAP_IOMMU && DEBUG_FS
-       help
-         Select this to see extensive information about
-         the internal state of OMAP IOMMU in debugfs.
+	bool "Export OMAP IOMMU internals in DebugFS"
+	depends on OMAP_IOMMU && DEBUG_FS
+	---help---
+	  Select this to see extensive information about
+	  the internal state of OMAP IOMMU in debugfs.
 
-         Say N unless you know you need this.
+	  Say N unless you know you need this.
+
+config ROCKCHIP_IOMMU
+	bool "Rockchip IOMMU Support"
+	depends on ARM
+	depends on ARCH_ROCKCHIP || COMPILE_TEST
+	select IOMMU_API
+	select ARM_DMA_USE_IOMMU
+	help
+	  Support for IOMMUs found on Rockchip rk32xx SOCs.
+	  These IOMMUs allow virtualization of the address space used by most
+	  cores within the multimedia subsystem.
+	  Say Y here if you are using a Rockchip SoC that includes an IOMMU
+	  device.
 
 config TEGRA_IOMMU_GART
 	bool "Tegra GART IOMMU Support"
@@ -163,18 +176,18 @@ config TEGRA_IOMMU_GART
 	  hardware included on Tegra SoCs.
 
 config TEGRA_IOMMU_SMMU
-	bool "Tegra SMMU IOMMU Support"
-	depends on ARCH_TEGRA && TEGRA_AHB
+	bool "NVIDIA Tegra SMMU Support"
+	depends on ARCH_TEGRA
+	depends on TEGRA_AHB
+	depends on TEGRA_MC
 	select IOMMU_API
 	help
-	  Enables support for remapping discontiguous physical memory
-	  shared with the operating system into contiguous I/O virtual
-	  space through the SMMU (System Memory Management Unit)
-	  hardware included on Tegra SoCs.
+	  This driver supports the IOMMU hardware (SMMU) found on NVIDIA Tegra
+	  SoCs (Tegra30 up to Tegra124).
 
 config EXYNOS_IOMMU
 	bool "Exynos IOMMU Support"
-	depends on ARCH_EXYNOS
+	depends on ARCH_EXYNOS && ARM
 	select IOMMU_API
 	select ARM_DMA_USE_IOMMU
 	help
diff --git a/drivers/iommu/Makefile b/drivers/iommu/Makefile
index 16edef74b8ee..7b976f294a69 100644
--- a/drivers/iommu/Makefile
+++ b/drivers/iommu/Makefile
@@ -11,8 +11,8 @@ obj-$(CONFIG_INTEL_IOMMU) += iova.o intel-iommu.o
 obj-$(CONFIG_IPMMU_VMSA) += ipmmu-vmsa.o
 obj-$(CONFIG_IRQ_REMAP) += intel_irq_remapping.o irq_remapping.o
 obj-$(CONFIG_OMAP_IOMMU) += omap-iommu.o
-obj-$(CONFIG_OMAP_IOMMU) += omap-iommu2.o
 obj-$(CONFIG_OMAP_IOMMU_DEBUG) += omap-iommu-debug.o
+obj-$(CONFIG_ROCKCHIP_IOMMU) += rockchip-iommu.o
 obj-$(CONFIG_TEGRA_IOMMU_GART) += tegra-gart.o
 obj-$(CONFIG_TEGRA_IOMMU_SMMU) += tegra-smmu.o
 obj-$(CONFIG_EXYNOS_IOMMU) += exynos-iommu.o
diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c
index ecb0109a5360..98024856df07 100644
--- a/drivers/iommu/amd_iommu.c
+++ b/drivers/iommu/amd_iommu.c
@@ -88,6 +88,27 @@ int amd_iommu_max_glx_val = -1;
 static struct dma_map_ops amd_iommu_dma_ops;
 
 /*
+ * This struct contains device specific data for the IOMMU
+ */
+struct iommu_dev_data {
+	struct list_head list;		  /* For domain->dev_list */
+	struct list_head dev_data_list;	  /* For global dev_data_list */
+	struct list_head alias_list;      /* Link alias-groups together */
+	struct iommu_dev_data *alias_data;/* The alias dev_data */
+	struct protection_domain *domain; /* Domain the device is bound to */
+	u16 devid;			  /* PCI Device ID */
+	bool iommu_v2;			  /* Device can make use of IOMMUv2 */
+	bool passthrough;		  /* Default for device is pt_domain */
+	struct {
+		bool enabled;
+		int qdep;
+	} ats;				  /* ATS state */
+	bool pri_tlp;			  /* PASID TLB required for
+					     PPR completions */
+	u32 errata;			  /* Bitmap for errata to apply */
+};
+
+/*
  * general struct to manage commands send to an IOMMU
  */
 struct iommu_cmd {
@@ -114,8 +135,9 @@ static struct iommu_dev_data *alloc_dev_data(u16 devid)
 	if (!dev_data)
 		return NULL;
 
+	INIT_LIST_HEAD(&dev_data->alias_list);
+
 	dev_data->devid = devid;
-	atomic_set(&dev_data->bind, 0);
 
 	spin_lock_irqsave(&dev_data_list_lock, flags);
 	list_add_tail(&dev_data->dev_data_list, &dev_data_list);
@@ -260,17 +282,13 @@ static bool check_device(struct device *dev)
 	return true;
 }
 
-static int init_iommu_group(struct device *dev)
+static void init_iommu_group(struct device *dev)
 {
 	struct iommu_group *group;
 
 	group = iommu_group_get_for_dev(dev);
-
-	if (IS_ERR(group))
-		return PTR_ERR(group);
-
-	iommu_group_put(group);
-	return 0;
+	if (!IS_ERR(group))
+		iommu_group_put(group);
 }
 
 static int __last_alias(struct pci_dev *pdev, u16 alias, void *data)
@@ -340,7 +358,6 @@ static int iommu_init_device(struct device *dev)
 	struct pci_dev *pdev = to_pci_dev(dev);
 	struct iommu_dev_data *dev_data;
 	u16 alias;
-	int ret;
 
 	if (dev->archdata.iommu)
 		return 0;
@@ -362,12 +379,9 @@ static int iommu_init_device(struct device *dev)
 			return -ENOTSUPP;
 		}
 		dev_data->alias_data = alias_data;
-	}
 
-	ret = init_iommu_group(dev);
-	if (ret) {
-		free_dev_data(dev_data);
-		return ret;
+		/* Add device to the alias_list */
+		list_add(&dev_data->alias_list, &alias_data->alias_list);
 	}
 
 	if (pci_iommuv2_capable(pdev)) {
@@ -455,6 +469,15 @@ int __init amd_iommu_init_devices(void)
 			goto out_free;
 	}
 
+	/*
+	 * Initialize IOMMU groups only after iommu_init_device() has
+	 * had a chance to populate any IVRS defined aliases.
+	 */
+	for_each_pci_dev(pdev) {
+		if (check_device(&pdev->dev))
+			init_iommu_group(&pdev->dev);
+	}
+
 	return 0;
 
 out_free:
@@ -1368,6 +1391,9 @@ static int iommu_map_page(struct protection_domain *dom,
 	count     = PAGE_SIZE_PTE_COUNT(page_size);
 	pte       = alloc_pte(dom, bus_addr, page_size, NULL, GFP_KERNEL);
 
+	if (!pte)
+		return -ENOMEM;
+
 	for (i = 0; i < count; ++i)
 		if (IOMMU_PTE_PRESENT(pte[i]))
 			return -EBUSY;
@@ -2122,35 +2148,29 @@ static void do_detach(struct iommu_dev_data *dev_data)
 static int __attach_device(struct iommu_dev_data *dev_data,
 			   struct protection_domain *domain)
 {
+	struct iommu_dev_data *head, *entry;
 	int ret;
 
 	/* lock domain */
 	spin_lock(&domain->lock);
 
-	if (dev_data->alias_data != NULL) {
-		struct iommu_dev_data *alias_data = dev_data->alias_data;
-
-		/* Some sanity checks */
-		ret = -EBUSY;
-		if (alias_data->domain != NULL &&
-				alias_data->domain != domain)
-			goto out_unlock;
+	head = dev_data;
 
-		if (dev_data->domain != NULL &&
-				dev_data->domain != domain)
-			goto out_unlock;
+	if (head->alias_data != NULL)
+		head = head->alias_data;
 
-		/* Do real assignment */
-		if (alias_data->domain == NULL)
-			do_attach(alias_data, domain);
+	/* Now we have the root of the alias group, if any */
 
-		atomic_inc(&alias_data->bind);
-	}
+	ret = -EBUSY;
+	if (head->domain != NULL)
+		goto out_unlock;
 
-	if (dev_data->domain == NULL)
-		do_attach(dev_data, domain);
+	/* Attach alias group root */
+	do_attach(head, domain);
 
-	atomic_inc(&dev_data->bind);
+	/* Attach other devices in the alias group */
+	list_for_each_entry(entry, &head->alias_list, alias_list)
+		do_attach(entry, domain);
 
 	ret = 0;
 
@@ -2298,6 +2318,7 @@ static int attach_device(struct device *dev,
  */
 static void __detach_device(struct iommu_dev_data *dev_data)
 {
+	struct iommu_dev_data *head, *entry;
 	struct protection_domain *domain;
 	unsigned long flags;
 
@@ -2307,15 +2328,14 @@ static void __detach_device(struct iommu_dev_data *dev_data)
 
 	spin_lock_irqsave(&domain->lock, flags);
 
-	if (dev_data->alias_data != NULL) {
-		struct iommu_dev_data *alias_data = dev_data->alias_data;
+	head = dev_data;
+	if (head->alias_data != NULL)
+		head = head->alias_data;
 
-		if (atomic_dec_and_test(&alias_data->bind))
-			do_detach(alias_data);
-	}
+	list_for_each_entry(entry, &head->alias_list, alias_list)
+		do_detach(entry);
 
-	if (atomic_dec_and_test(&dev_data->bind))
-		do_detach(dev_data);
+	do_detach(head);
 
 	spin_unlock_irqrestore(&domain->lock, flags);
 
@@ -2415,6 +2435,7 @@ static int device_change_notifier(struct notifier_block *nb,
 	case BUS_NOTIFY_ADD_DEVICE:
 
 		iommu_init_device(dev);
+		init_iommu_group(dev);
 
 		/*
 		 * dev_data is still NULL and
@@ -3158,7 +3179,6 @@ static void cleanup_domain(struct protection_domain *domain)
 		entry = list_first_entry(&domain->dev_list,
 					 struct iommu_dev_data, list);
 		__detach_device(entry);
-		atomic_set(&entry->bind, 0);
 	}
 
 	write_unlock_irqrestore(&amd_iommu_devtable_lock, flags);
@@ -3384,28 +3404,30 @@ static phys_addr_t amd_iommu_iova_to_phys(struct iommu_domain *dom,
 	return paddr;
 }
 
-static int amd_iommu_domain_has_cap(struct iommu_domain *domain,
-				    unsigned long cap)
+static bool amd_iommu_capable(enum iommu_cap cap)
 {
 	switch (cap) {
 	case IOMMU_CAP_CACHE_COHERENCY:
-		return 1;
+		return true;
 	case IOMMU_CAP_INTR_REMAP:
-		return irq_remapping_enabled;
+		return (irq_remapping_enabled == 1);
+	case IOMMU_CAP_NOEXEC:
+		return false;
 	}
 
-	return 0;
+	return false;
 }
 
 static const struct iommu_ops amd_iommu_ops = {
+	.capable = amd_iommu_capable,
 	.domain_init = amd_iommu_domain_init,
 	.domain_destroy = amd_iommu_domain_destroy,
 	.attach_dev = amd_iommu_attach_device,
 	.detach_dev = amd_iommu_detach_device,
 	.map = amd_iommu_map,
 	.unmap = amd_iommu_unmap,
+	.map_sg = default_iommu_map_sg,
 	.iova_to_phys = amd_iommu_iova_to_phys,
-	.domain_has_cap = amd_iommu_domain_has_cap,
 	.pgsize_bitmap	= AMD_IOMMU_PGSIZES,
 };
 
@@ -4049,7 +4071,7 @@ static int setup_ioapic_entry(int irq, struct IO_APIC_route_entry *entry,
 	int devid;
 	int ret;
 
-	cfg = irq_get_chip_data(irq);
+	cfg = irq_cfg(irq);
 	if (!cfg)
 		return -EINVAL;
 
@@ -4112,7 +4134,7 @@ static int set_affinity(struct irq_data *data, const struct cpumask *mask,
 	if (!config_enabled(CONFIG_SMP))
 		return -1;
 
-	cfg       = data->chip_data;
+	cfg       = irqd_cfg(data);
 	irq       = data->irq;
 	irte_info = &cfg->irq_2_irte;
 
@@ -4150,7 +4172,7 @@ static int free_irq(int irq)
 	struct irq_2_irte *irte_info;
 	struct irq_cfg *cfg;
 
-	cfg = irq_get_chip_data(irq);
+	cfg = irq_cfg(irq);
 	if (!cfg)
 		return -EINVAL;
 
@@ -4169,7 +4191,7 @@ static void compose_msi_msg(struct pci_dev *pdev,
 	struct irq_cfg *cfg;
 	union irte irte;
 
-	cfg = irq_get_chip_data(irq);
+	cfg = irq_cfg(irq);
 	if (!cfg)
 		return;
 
@@ -4198,7 +4220,7 @@ static int msi_alloc_irq(struct pci_dev *pdev, int irq, int nvec)
 	if (!pdev)
 		return -EINVAL;
 
-	cfg = irq_get_chip_data(irq);
+	cfg = irq_cfg(irq);
 	if (!cfg)
 		return -EINVAL;
 
@@ -4218,7 +4240,7 @@ static int msi_setup_irq(struct pci_dev *pdev, unsigned int irq,
 	if (!pdev)
 		return -EINVAL;
 
-	cfg = irq_get_chip_data(irq);
+	cfg = irq_cfg(irq);
 	if (!cfg)
 		return -EINVAL;
 
@@ -4235,13 +4257,13 @@ static int msi_setup_irq(struct pci_dev *pdev, unsigned int irq,
 	return 0;
 }
 
-static int setup_hpet_msi(unsigned int irq, unsigned int id)
+static int alloc_hpet_msi(unsigned int irq, unsigned int id)
 {
 	struct irq_2_irte *irte_info;
 	struct irq_cfg *cfg;
 	int index, devid;
 
-	cfg = irq_get_chip_data(irq);
+	cfg = irq_cfg(irq);
 	if (!cfg)
 		return -EINVAL;
 
@@ -4274,6 +4296,6 @@ struct irq_remap_ops amd_iommu_irq_ops = {
 	.compose_msi_msg	= compose_msi_msg,
 	.msi_alloc_irq		= msi_alloc_irq,
 	.msi_setup_irq		= msi_setup_irq,
-	.setup_hpet_msi		= setup_hpet_msi,
+	.alloc_hpet_msi		= alloc_hpet_msi,
 };
 #endif
diff --git a/drivers/iommu/amd_iommu_init.c b/drivers/iommu/amd_iommu_init.c
index 3783e0b44df6..b0522f15730f 100644
--- a/drivers/iommu/amd_iommu_init.c
+++ b/drivers/iommu/amd_iommu_init.c
@@ -712,7 +712,7 @@ static void __init set_dev_entry_from_acpi(struct amd_iommu *iommu,
 	set_iommu_for_device(iommu, devid);
 }
 
-static int __init add_special_device(u8 type, u8 id, u16 devid, bool cmd_line)
+static int __init add_special_device(u8 type, u8 id, u16 *devid, bool cmd_line)
 {
 	struct devid_map *entry;
 	struct list_head *list;
@@ -731,6 +731,8 @@ static int __init add_special_device(u8 type, u8 id, u16 devid, bool cmd_line)
 		pr_info("AMD-Vi: Command-line override present for %s id %d - ignoring\n",
 			type == IVHD_SPECIAL_IOAPIC ? "IOAPIC" : "HPET", id);
 
+		*devid = entry->devid;
+
 		return 0;
 	}
 
@@ -739,7 +741,7 @@ static int __init add_special_device(u8 type, u8 id, u16 devid, bool cmd_line)
 		return -ENOMEM;
 
 	entry->id	= id;
-	entry->devid	= devid;
+	entry->devid	= *devid;
 	entry->cmd_line	= cmd_line;
 
 	list_add_tail(&entry->list, list);
@@ -754,7 +756,7 @@ static int __init add_early_maps(void)
 	for (i = 0; i < early_ioapic_map_size; ++i) {
 		ret = add_special_device(IVHD_SPECIAL_IOAPIC,
 					 early_ioapic_map[i].id,
-					 early_ioapic_map[i].devid,
+					 &early_ioapic_map[i].devid,
 					 early_ioapic_map[i].cmd_line);
 		if (ret)
 			return ret;
@@ -763,7 +765,7 @@ static int __init add_early_maps(void)
 	for (i = 0; i < early_hpet_map_size; ++i) {
 		ret = add_special_device(IVHD_SPECIAL_HPET,
 					 early_hpet_map[i].id,
-					 early_hpet_map[i].devid,
+					 &early_hpet_map[i].devid,
 					 early_hpet_map[i].cmd_line);
 		if (ret)
 			return ret;
@@ -978,10 +980,17 @@ static int __init init_iommu_from_acpi(struct amd_iommu *iommu,
 				    PCI_SLOT(devid),
 				    PCI_FUNC(devid));
 
-			set_dev_entry_from_acpi(iommu, devid, e->flags, 0);
-			ret = add_special_device(type, handle, devid, false);
+			ret = add_special_device(type, handle, &devid, false);
 			if (ret)
 				return ret;
+
+			/*
+			 * add_special_device might update the devid in case a
+			 * command-line override is present. So call
+			 * set_dev_entry_from_acpi after add_special_device.
+			 */
+			set_dev_entry_from_acpi(iommu, devid, e->flags, 0);
+
 			break;
 		}
 		default:
diff --git a/drivers/iommu/amd_iommu_types.h b/drivers/iommu/amd_iommu_types.h
index 8e43b7cba133..cec51a8ba844 100644
--- a/drivers/iommu/amd_iommu_types.h
+++ b/drivers/iommu/amd_iommu_types.h
@@ -418,27 +418,6 @@ struct protection_domain {
 };
 
 /*
- * This struct contains device specific data for the IOMMU
- */
-struct iommu_dev_data {
-	struct list_head list;		  /* For domain->dev_list */
-	struct list_head dev_data_list;	  /* For global dev_data_list */
-	struct iommu_dev_data *alias_data;/* The alias dev_data */
-	struct protection_domain *domain; /* Domain the device is bound to */
-	atomic_t bind;			  /* Domain attach reference count */
-	u16 devid;			  /* PCI Device ID */
-	bool iommu_v2;			  /* Device can make use of IOMMUv2 */
-	bool passthrough;		  /* Default for device is pt_domain */
-	struct {
-		bool enabled;
-		int qdep;
-	} ats;				  /* ATS state */
-	bool pri_tlp;			  /* PASID TLB required for
-					     PPR completions */
-	u32 errata;			  /* Bitmap for errata to apply */
-};
-
-/*
  * For dynamic growth the aperture size is split into ranges of 128MB of
  * DMA address space each. This struct represents one such range.
  */
diff --git a/drivers/iommu/amd_iommu_v2.c b/drivers/iommu/amd_iommu_v2.c
index 5f578e850fc5..90f70d0e1141 100644
--- a/drivers/iommu/amd_iommu_v2.c
+++ b/drivers/iommu/amd_iommu_v2.c
@@ -92,13 +92,6 @@ static spinlock_t state_lock;
 
 static struct workqueue_struct *iommu_wq;
 
-/*
- * Empty page table - Used between
- * mmu_notifier_invalidate_range_start and
- * mmu_notifier_invalidate_range_end
- */
-static u64 *empty_page_table;
-
 static void free_pasid_states(struct device_state *dev_state);
 
 static u16 device_id(struct pci_dev *pdev)
@@ -279,10 +272,8 @@ static void free_pasid_state(struct pasid_state *pasid_state)
 
 static void put_pasid_state(struct pasid_state *pasid_state)
 {
-	if (atomic_dec_and_test(&pasid_state->count)) {
-		put_device_state(pasid_state->device_state);
+	if (atomic_dec_and_test(&pasid_state->count))
 		wake_up(&pasid_state->wq);
-	}
 }
 
 static void put_pasid_state_wait(struct pasid_state *pasid_state)
@@ -291,9 +282,7 @@ static void put_pasid_state_wait(struct pasid_state *pasid_state)
 
 	prepare_to_wait(&pasid_state->wq, &wait, TASK_UNINTERRUPTIBLE);
 
-	if (atomic_dec_and_test(&pasid_state->count))
-		put_device_state(pasid_state->device_state);
-	else
+	if (!atomic_dec_and_test(&pasid_state->count))
 		schedule();
 
 	finish_wait(&pasid_state->wq, &wait);
@@ -402,9 +391,11 @@ static void __mn_flush_page(struct mmu_notifier *mn,
 
 static int mn_clear_flush_young(struct mmu_notifier *mn,
 				struct mm_struct *mm,
-				unsigned long address)
+				unsigned long start,
+				unsigned long end)
 {
-	__mn_flush_page(mn, address);
+	for (; start < end; start += PAGE_SIZE)
+		__mn_flush_page(mn, start);
 
 	return 0;
 }
@@ -416,46 +407,21 @@ static void mn_invalidate_page(struct mmu_notifier *mn,
 	__mn_flush_page(mn, address);
 }
 
-static void mn_invalidate_range_start(struct mmu_notifier *mn,
-				      struct mm_struct *mm,
-				      unsigned long start, unsigned long end)
-{
-	struct pasid_state *pasid_state;
-	struct device_state *dev_state;
-	unsigned long flags;
-
-	pasid_state = mn_to_state(mn);
-	dev_state   = pasid_state->device_state;
-
-	spin_lock_irqsave(&pasid_state->lock, flags);
-	if (pasid_state->mmu_notifier_count == 0) {
-		amd_iommu_domain_set_gcr3(dev_state->domain,
-					  pasid_state->pasid,
-					  __pa(empty_page_table));
-	}
-	pasid_state->mmu_notifier_count += 1;
-	spin_unlock_irqrestore(&pasid_state->lock, flags);
-}
-
-static void mn_invalidate_range_end(struct mmu_notifier *mn,
-				    struct mm_struct *mm,
-				    unsigned long start, unsigned long end)
+static void mn_invalidate_range(struct mmu_notifier *mn,
+				struct mm_struct *mm,
+				unsigned long start, unsigned long end)
 {
 	struct pasid_state *pasid_state;
 	struct device_state *dev_state;
-	unsigned long flags;
 
 	pasid_state = mn_to_state(mn);
 	dev_state   = pasid_state->device_state;
 
-	spin_lock_irqsave(&pasid_state->lock, flags);
-	pasid_state->mmu_notifier_count -= 1;
-	if (pasid_state->mmu_notifier_count == 0) {
-		amd_iommu_domain_set_gcr3(dev_state->domain,
-					  pasid_state->pasid,
-					  __pa(pasid_state->mm->pgd));
-	}
-	spin_unlock_irqrestore(&pasid_state->lock, flags);
+	if ((start ^ (end - 1)) < PAGE_SIZE)
+		amd_iommu_flush_page(dev_state->domain, pasid_state->pasid,
+				     start);
+	else
+		amd_iommu_flush_tlb(dev_state->domain, pasid_state->pasid);
 }
 
 static void mn_release(struct mmu_notifier *mn, struct mm_struct *mm)
@@ -480,8 +446,7 @@ static struct mmu_notifier_ops iommu_mn = {
 	.release		= mn_release,
 	.clear_flush_young      = mn_clear_flush_young,
 	.invalidate_page        = mn_invalidate_page,
-	.invalidate_range_start = mn_invalidate_range_start,
-	.invalidate_range_end   = mn_invalidate_range_end,
+	.invalidate_range       = mn_invalidate_range,
 };
 
 static void set_pri_tag_status(struct pasid_state *pasid_state,
@@ -511,45 +476,67 @@ static void finish_pri_tag(struct device_state *dev_state,
 	spin_unlock_irqrestore(&pasid_state->lock, flags);
 }
 
+static void handle_fault_error(struct fault *fault)
+{
+	int status;
+
+	if (!fault->dev_state->inv_ppr_cb) {
+		set_pri_tag_status(fault->state, fault->tag, PPR_INVALID);
+		return;
+	}
+
+	status = fault->dev_state->inv_ppr_cb(fault->dev_state->pdev,
+					      fault->pasid,
+					      fault->address,
+					      fault->flags);
+	switch (status) {
+	case AMD_IOMMU_INV_PRI_RSP_SUCCESS:
+		set_pri_tag_status(fault->state, fault->tag, PPR_SUCCESS);
+		break;
+	case AMD_IOMMU_INV_PRI_RSP_INVALID:
+		set_pri_tag_status(fault->state, fault->tag, PPR_INVALID);
+		break;
+	case AMD_IOMMU_INV_PRI_RSP_FAIL:
+		set_pri_tag_status(fault->state, fault->tag, PPR_FAILURE);
+		break;
+	default:
+		BUG();
+	}
+}
+
 static void do_fault(struct work_struct *work)
 {
 	struct fault *fault = container_of(work, struct fault, work);
-	int npages, write;
-	struct page *page;
+	struct mm_struct *mm;
+	struct vm_area_struct *vma;
+	u64 address;
+	int ret, write;
 
 	write = !!(fault->flags & PPR_FAULT_WRITE);
 
-	down_read(&fault->state->mm->mmap_sem);
-	npages = get_user_pages(NULL, fault->state->mm,
-				fault->address, 1, write, 0, &page, NULL);
-	up_read(&fault->state->mm->mmap_sem);
-
-	if (npages == 1) {
-		put_page(page);
-	} else if (fault->dev_state->inv_ppr_cb) {
-		int status;
-
-		status = fault->dev_state->inv_ppr_cb(fault->dev_state->pdev,
-						      fault->pasid,
-						      fault->address,
-						      fault->flags);
-		switch (status) {
-		case AMD_IOMMU_INV_PRI_RSP_SUCCESS:
-			set_pri_tag_status(fault->state, fault->tag, PPR_SUCCESS);
-			break;
-		case AMD_IOMMU_INV_PRI_RSP_INVALID:
-			set_pri_tag_status(fault->state, fault->tag, PPR_INVALID);
-			break;
-		case AMD_IOMMU_INV_PRI_RSP_FAIL:
-			set_pri_tag_status(fault->state, fault->tag, PPR_FAILURE);
-			break;
-		default:
-			BUG();
-		}
-	} else {
-		set_pri_tag_status(fault->state, fault->tag, PPR_INVALID);
+	mm = fault->state->mm;
+	address = fault->address;
+
+	down_read(&mm->mmap_sem);
+	vma = find_extend_vma(mm, address);
+	if (!vma || address < vma->vm_start) {
+		/* failed to get a vma in the right range */
+		up_read(&mm->mmap_sem);
+		handle_fault_error(fault);
+		goto out;
+	}
+
+	ret = handle_mm_fault(mm, vma, address, write);
+	if (ret & VM_FAULT_ERROR) {
+		/* failed to service fault */
+		up_read(&mm->mmap_sem);
+		handle_fault_error(fault);
+		goto out;
 	}
 
+	up_read(&mm->mmap_sem);
+
+out:
 	finish_pri_tag(fault->dev_state, fault->state, fault->tag);
 
 	put_pasid_state(fault->state);
@@ -952,18 +939,10 @@ static int __init amd_iommu_v2_init(void)
 	if (iommu_wq == NULL)
 		goto out;
 
-	ret = -ENOMEM;
-	empty_page_table = (u64 *)get_zeroed_page(GFP_KERNEL);
-	if (empty_page_table == NULL)
-		goto out_destroy_wq;
-
 	amd_iommu_register_ppr_notifier(&ppr_nb);
 
 	return 0;
 
-out_destroy_wq:
-	destroy_workqueue(iommu_wq);
-
 out:
 	return ret;
 }
@@ -997,8 +976,6 @@ static void __exit amd_iommu_v2_exit(void)
 	}
 
 	destroy_workqueue(iommu_wq);
-
-	free_page((unsigned long)empty_page_table);
 }
 
 module_init(amd_iommu_v2_init);
diff --git a/drivers/iommu/arm-smmu.c b/drivers/iommu/arm-smmu.c
index a83cc2a2a2ca..6cd47b75286f 100644
--- a/drivers/iommu/arm-smmu.c
+++ b/drivers/iommu/arm-smmu.c
@@ -24,7 +24,7 @@
  *	- v7/v8 long-descriptor format
  *	- Non-secure access to the SMMU
  *	- 4k and 64k pages, with contiguous pte hints.
- *	- Up to 42-bit addressing (dependent on VA_BITS)
+ *	- Up to 48-bit addressing (dependent on VA_BITS)
  *	- Context fault reporting
  */
 
@@ -59,7 +59,7 @@
 
 /* SMMU global address space */
 #define ARM_SMMU_GR0(smmu)		((smmu)->base)
-#define ARM_SMMU_GR1(smmu)		((smmu)->base + (smmu)->pagesize)
+#define ARM_SMMU_GR1(smmu)		((smmu)->base + (1 << (smmu)->pgshift))
 
 /*
  * SMMU global address space with conditional offset to access secure
@@ -224,7 +224,7 @@
 
 /* Translation context bank */
 #define ARM_SMMU_CB_BASE(smmu)		((smmu)->base + ((smmu)->size >> 1))
-#define ARM_SMMU_CB(smmu, n)		((n) * (smmu)->pagesize)
+#define ARM_SMMU_CB(smmu, n)		((n) * (1 << (smmu)->pgshift))
 
 #define ARM_SMMU_CB_SCTLR		0x0
 #define ARM_SMMU_CB_RESUME		0x8
@@ -326,6 +326,16 @@
 
 #define FSYNR0_WNR			(1 << 4)
 
+static int force_stage;
+module_param_named(force_stage, force_stage, int, S_IRUGO | S_IWUSR);
+MODULE_PARM_DESC(force_stage,
+	"Force SMMU mappings to be installed at a particular stage of translation. A value of '1' or '2' forces the corresponding stage. All other values are ignored (i.e. no stage is forced). Note that selecting a specific stage will disable support for nested translation.");
+
+enum arm_smmu_arch_version {
+	ARM_SMMU_V1 = 1,
+	ARM_SMMU_V2,
+};
+
 struct arm_smmu_smr {
 	u8				idx;
 	u16				mask;
@@ -349,7 +359,7 @@ struct arm_smmu_device {
 
 	void __iomem			*base;
 	unsigned long			size;
-	unsigned long			pagesize;
+	unsigned long			pgshift;
 
 #define ARM_SMMU_FEAT_COHERENT_WALK	(1 << 0)
 #define ARM_SMMU_FEAT_STREAM_MATCH	(1 << 1)
@@ -360,7 +370,7 @@ struct arm_smmu_device {
 
 #define ARM_SMMU_OPT_SECURE_CFG_ACCESS (1 << 0)
 	u32				options;
-	int				version;
+	enum arm_smmu_arch_version	version;
 
 	u32				num_context_banks;
 	u32				num_s2_context_banks;
@@ -370,8 +380,9 @@ struct arm_smmu_device {
 	u32				num_mapping_groups;
 	DECLARE_BITMAP(smr_map, ARM_SMMU_MAX_SMRS);
 
-	unsigned long			input_size;
+	unsigned long			s1_input_size;
 	unsigned long			s1_output_size;
+	unsigned long			s2_input_size;
 	unsigned long			s2_output_size;
 
 	u32				num_global_irqs;
@@ -393,9 +404,16 @@ struct arm_smmu_cfg {
 #define ARM_SMMU_CB_ASID(cfg)		((cfg)->cbndx)
 #define ARM_SMMU_CB_VMID(cfg)		((cfg)->cbndx + 1)
 
+enum arm_smmu_domain_stage {
+	ARM_SMMU_DOMAIN_S1 = 0,
+	ARM_SMMU_DOMAIN_S2,
+	ARM_SMMU_DOMAIN_NESTED,
+};
+
 struct arm_smmu_domain {
 	struct arm_smmu_device		*smmu;
 	struct arm_smmu_cfg		cfg;
+	enum arm_smmu_domain_stage	stage;
 	spinlock_t			lock;
 };
 
@@ -426,17 +444,17 @@ static void parse_driver_options(struct arm_smmu_device *smmu)
 	} while (arm_smmu_options[++i].opt);
 }
 
-static struct device *dev_get_master_dev(struct device *dev)
+static struct device_node *dev_get_dev_node(struct device *dev)
 {
 	if (dev_is_pci(dev)) {
 		struct pci_bus *bus = to_pci_dev(dev)->bus;
 
 		while (!pci_is_root_bus(bus))
 			bus = bus->parent;
-		return bus->bridge->parent;
+		return bus->bridge->parent->of_node;
 	}
 
-	return dev;
+	return dev->of_node;
 }
 
 static struct arm_smmu_master *find_smmu_master(struct arm_smmu_device *smmu,
@@ -461,15 +479,17 @@ static struct arm_smmu_master *find_smmu_master(struct arm_smmu_device *smmu,
 }
 
 static struct arm_smmu_master_cfg *
-find_smmu_master_cfg(struct arm_smmu_device *smmu, struct device *dev)
+find_smmu_master_cfg(struct device *dev)
 {
-	struct arm_smmu_master *master;
+	struct arm_smmu_master_cfg *cfg = NULL;
+	struct iommu_group *group = iommu_group_get(dev);
 
-	if (dev_is_pci(dev))
-		return dev->archdata.iommu;
+	if (group) {
+		cfg = iommu_group_get_iommudata(group);
+		iommu_group_put(group);
+	}
 
-	master = find_smmu_master(smmu, dev->of_node);
-	return master ? &master->cfg : NULL;
+	return cfg;
 }
 
 static int insert_smmu_master(struct arm_smmu_device *smmu,
@@ -545,7 +565,7 @@ static struct arm_smmu_device *find_smmu_for_device(struct device *dev)
 {
 	struct arm_smmu_device *smmu;
 	struct arm_smmu_master *master = NULL;
-	struct device_node *dev_node = dev_get_master_dev(dev)->of_node;
+	struct device_node *dev_node = dev_get_dev_node(dev);
 
 	spin_lock(&arm_smmu_devices_lock);
 	list_for_each_entry(smmu, &arm_smmu_devices, list) {
@@ -729,7 +749,7 @@ static void arm_smmu_init_context_bank(struct arm_smmu_domain *smmu_domain)
 
 	/* CBAR */
 	reg = cfg->cbar;
-	if (smmu->version == 1)
+	if (smmu->version == ARM_SMMU_V1)
 		reg |= cfg->irptndx << CBAR_IRPTNDX_SHIFT;
 
 	/*
@@ -744,7 +764,7 @@ static void arm_smmu_init_context_bank(struct arm_smmu_domain *smmu_domain)
 	}
 	writel_relaxed(reg, gr1_base + ARM_SMMU_GR1_CBAR(cfg->cbndx));
 
-	if (smmu->version > 1) {
+	if (smmu->version > ARM_SMMU_V1) {
 		/* CBA2R */
 #ifdef CONFIG_64BIT
 		reg = CBA2R_RW64_64BIT;
@@ -755,7 +775,7 @@ static void arm_smmu_init_context_bank(struct arm_smmu_domain *smmu_domain)
 			       gr1_base + ARM_SMMU_GR1_CBA2R(cfg->cbndx));
 
 		/* TTBCR2 */
-		switch (smmu->input_size) {
+		switch (smmu->s1_input_size) {
 		case 32:
 			reg = (TTBCR2_ADDR_32 << TTBCR2_SEP_SHIFT);
 			break;
@@ -817,14 +837,14 @@ static void arm_smmu_init_context_bank(struct arm_smmu_domain *smmu_domain)
 	 * TTBCR
 	 * We use long descriptor, with inner-shareable WBWA tables in TTBR0.
 	 */
-	if (smmu->version > 1) {
+	if (smmu->version > ARM_SMMU_V1) {
 		if (PAGE_SIZE == SZ_4K)
 			reg = TTBCR_TG0_4K;
 		else
 			reg = TTBCR_TG0_64K;
 
 		if (!stage1) {
-			reg |= (64 - smmu->s1_output_size) << TTBCR_T0SZ_SHIFT;
+			reg |= (64 - smmu->s2_input_size) << TTBCR_T0SZ_SHIFT;
 
 			switch (smmu->s2_output_size) {
 			case 32:
@@ -847,7 +867,7 @@ static void arm_smmu_init_context_bank(struct arm_smmu_domain *smmu_domain)
 				break;
 			}
 		} else {
-			reg |= (64 - smmu->input_size) << TTBCR_T0SZ_SHIFT;
+			reg |= (64 - smmu->s1_input_size) << TTBCR_T0SZ_SHIFT;
 		}
 	} else {
 		reg = 0;
@@ -893,19 +913,46 @@ static int arm_smmu_init_domain_context(struct iommu_domain *domain,
 	if (smmu_domain->smmu)
 		goto out_unlock;
 
-	if (smmu->features & ARM_SMMU_FEAT_TRANS_NESTED) {
+	/*
+	 * Mapping the requested stage onto what we support is surprisingly
+	 * complicated, mainly because the spec allows S1+S2 SMMUs without
+	 * support for nested translation. That means we end up with the
+	 * following table:
+	 *
+	 * Requested        Supported        Actual
+	 *     S1               N              S1
+	 *     S1             S1+S2            S1
+	 *     S1               S2             S2
+	 *     S1               S1             S1
+	 *     N                N              N
+	 *     N              S1+S2            S2
+	 *     N                S2             S2
+	 *     N                S1             S1
+	 *
+	 * Note that you can't actually request stage-2 mappings.
+	 */
+	if (!(smmu->features & ARM_SMMU_FEAT_TRANS_S1))
+		smmu_domain->stage = ARM_SMMU_DOMAIN_S2;
+	if (!(smmu->features & ARM_SMMU_FEAT_TRANS_S2))
+		smmu_domain->stage = ARM_SMMU_DOMAIN_S1;
+
+	switch (smmu_domain->stage) {
+	case ARM_SMMU_DOMAIN_S1:
+		cfg->cbar = CBAR_TYPE_S1_TRANS_S2_BYPASS;
+		start = smmu->num_s2_context_banks;
+		break;
+	case ARM_SMMU_DOMAIN_NESTED:
 		/*
 		 * We will likely want to change this if/when KVM gets
 		 * involved.
 		 */
-		cfg->cbar = CBAR_TYPE_S1_TRANS_S2_BYPASS;
-		start = smmu->num_s2_context_banks;
-	} else if (smmu->features & ARM_SMMU_FEAT_TRANS_S1) {
-		cfg->cbar = CBAR_TYPE_S1_TRANS_S2_BYPASS;
-		start = smmu->num_s2_context_banks;
-	} else {
+	case ARM_SMMU_DOMAIN_S2:
 		cfg->cbar = CBAR_TYPE_S2_TRANS;
 		start = 0;
+		break;
+	default:
+		ret = -EINVAL;
+		goto out_unlock;
 	}
 
 	ret = __arm_smmu_alloc_bitmap(smmu->context_map, start,
@@ -914,7 +961,7 @@ static int arm_smmu_init_domain_context(struct iommu_domain *domain,
 		goto out_unlock;
 
 	cfg->cbndx = ret;
-	if (smmu->version == 1) {
+	if (smmu->version == ARM_SMMU_V1) {
 		cfg->irptndx = atomic_inc_return(&smmu->irptndx);
 		cfg->irptndx %= smmu->num_context_irqs;
 	} else {
@@ -1151,9 +1198,10 @@ static int arm_smmu_domain_add_master(struct arm_smmu_domain *smmu_domain,
 	struct arm_smmu_device *smmu = smmu_domain->smmu;
 	void __iomem *gr0_base = ARM_SMMU_GR0(smmu);
 
+	/* Devices in an IOMMU group may already be configured */
 	ret = arm_smmu_master_configure_smrs(smmu, cfg);
 	if (ret)
-		return ret;
+		return ret == -EEXIST ? 0 : ret;
 
 	for (i = 0; i < cfg->num_streamids; ++i) {
 		u32 idx, s2cr;
@@ -1174,6 +1222,10 @@ static void arm_smmu_domain_remove_master(struct arm_smmu_domain *smmu_domain,
 	struct arm_smmu_device *smmu = smmu_domain->smmu;
 	void __iomem *gr0_base = ARM_SMMU_GR0(smmu);
 
+	/* An IOMMU group is torn down by the first device to be removed */
+	if ((smmu->features & ARM_SMMU_FEAT_STREAM_MATCH) && !cfg->smrs)
+		return;
+
 	/*
 	 * We *must* clear the S2CR first, because freeing the SMR means
 	 * that it can be re-allocated immediately.
@@ -1195,12 +1247,17 @@ static int arm_smmu_attach_dev(struct iommu_domain *domain, struct device *dev)
 	struct arm_smmu_device *smmu, *dom_smmu;
 	struct arm_smmu_master_cfg *cfg;
 
-	smmu = dev_get_master_dev(dev)->archdata.iommu;
+	smmu = find_smmu_for_device(dev);
 	if (!smmu) {
 		dev_err(dev, "cannot attach to SMMU, is it on the same bus?\n");
 		return -ENXIO;
 	}
 
+	if (dev->archdata.iommu) {
+		dev_err(dev, "already attached to IOMMU domain\n");
+		return -EEXIST;
+	}
+
 	/*
 	 * Sanity check the domain. We don't support domains across
 	 * different SMMUs.
@@ -1223,11 +1280,14 @@ static int arm_smmu_attach_dev(struct iommu_domain *domain, struct device *dev)
 	}
 
 	/* Looks ok, so add the device to the domain */
-	cfg = find_smmu_master_cfg(smmu_domain->smmu, dev);
+	cfg = find_smmu_master_cfg(dev);
 	if (!cfg)
 		return -ENODEV;
 
-	return arm_smmu_domain_add_master(smmu_domain, cfg);
+	ret = arm_smmu_domain_add_master(smmu_domain, cfg);
+	if (!ret)
+		dev->archdata.iommu = domain;
+	return ret;
 }
 
 static void arm_smmu_detach_dev(struct iommu_domain *domain, struct device *dev)
@@ -1235,9 +1295,12 @@ static void arm_smmu_detach_dev(struct iommu_domain *domain, struct device *dev)
 	struct arm_smmu_domain *smmu_domain = domain->priv;
 	struct arm_smmu_master_cfg *cfg;
 
-	cfg = find_smmu_master_cfg(smmu_domain->smmu, dev);
-	if (cfg)
-		arm_smmu_domain_remove_master(smmu_domain, cfg);
+	cfg = find_smmu_master_cfg(dev);
+	if (!cfg)
+		return;
+
+	dev->archdata.iommu = NULL;
+	arm_smmu_domain_remove_master(smmu_domain, cfg);
 }
 
 static bool arm_smmu_pte_is_contiguous_range(unsigned long addr,
@@ -1252,7 +1315,7 @@ static int arm_smmu_alloc_init_pte(struct arm_smmu_device *smmu, pmd_t *pmd,
 				   unsigned long pfn, int prot, int stage)
 {
 	pte_t *pte, *start;
-	pteval_t pteval = ARM_SMMU_PTE_PAGE | ARM_SMMU_PTE_AF | ARM_SMMU_PTE_XN;
+	pteval_t pteval = ARM_SMMU_PTE_PAGE | ARM_SMMU_PTE_AF;
 
 	if (pmd_none(*pmd)) {
 		/* Allocate a new set of tables */
@@ -1286,10 +1349,11 @@ static int arm_smmu_alloc_init_pte(struct arm_smmu_device *smmu, pmd_t *pmd,
 			pteval |= ARM_SMMU_PTE_MEMATTR_NC;
 	}
 
+	if (prot & IOMMU_NOEXEC)
+		pteval |= ARM_SMMU_PTE_XN;
+
 	/* If no access, create a faulting entry to avoid TLB fills */
-	if (prot & IOMMU_EXEC)
-		pteval &= ~ARM_SMMU_PTE_XN;
-	else if (!(prot & (IOMMU_READ | IOMMU_WRITE)))
+	if (!(prot & (IOMMU_READ | IOMMU_WRITE)))
 		pteval &= ~ARM_SMMU_PTE_PAGE;
 
 	pteval |= ARM_SMMU_PTE_SH_IS;
@@ -1379,6 +1443,7 @@ static int arm_smmu_alloc_init_pmd(struct arm_smmu_device *smmu, pud_t *pud,
 		ret = arm_smmu_alloc_init_pte(smmu, pmd, addr, next, pfn,
 					      prot, stage);
 		phys += next - addr;
+		pfn = __phys_to_pfn(phys);
 	} while (pmd++, addr = next, addr < end);
 
 	return ret;
@@ -1431,9 +1496,11 @@ static int arm_smmu_handle_mapping(struct arm_smmu_domain *smmu_domain,
 
 	if (cfg->cbar == CBAR_TYPE_S2_TRANS) {
 		stage = 2;
+		input_mask = (1ULL << smmu->s2_input_size) - 1;
 		output_mask = (1ULL << smmu->s2_output_size) - 1;
 	} else {
 		stage = 1;
+		input_mask = (1ULL << smmu->s1_input_size) - 1;
 		output_mask = (1ULL << smmu->s1_output_size) - 1;
 	}
 
@@ -1443,7 +1510,6 @@ static int arm_smmu_handle_mapping(struct arm_smmu_domain *smmu_domain,
 	if (size & ~PAGE_MASK)
 		return -EINVAL;
 
-	input_mask = (1ULL << smmu->input_size) - 1;
 	if ((phys_addr_t)iova & ~input_mask)
 		return -ERANGE;
 
@@ -1526,20 +1592,21 @@ static phys_addr_t arm_smmu_iova_to_phys(struct iommu_domain *domain,
 	return __pfn_to_phys(pte_pfn(pte)) | (iova & ~PAGE_MASK);
 }
 
-static int arm_smmu_domain_has_cap(struct iommu_domain *domain,
-				   unsigned long cap)
+static bool arm_smmu_capable(enum iommu_cap cap)
 {
-	struct arm_smmu_domain *smmu_domain = domain->priv;
-	struct arm_smmu_device *smmu = smmu_domain->smmu;
-	u32 features = smmu ? smmu->features : 0;
-
 	switch (cap) {
 	case IOMMU_CAP_CACHE_COHERENCY:
-		return features & ARM_SMMU_FEAT_COHERENT_WALK;
+		/*
+		 * Return true here as the SMMU can always send out coherent
+		 * requests.
+		 */
+		return true;
 	case IOMMU_CAP_INTR_REMAP:
-		return 1; /* MSIs are just memory writes */
+		return true; /* MSIs are just memory writes */
+	case IOMMU_CAP_NOEXEC:
+		return true;
 	default:
-		return 0;
+		return false;
 	}
 }
 
@@ -1549,17 +1616,19 @@ static int __arm_smmu_get_pci_sid(struct pci_dev *pdev, u16 alias, void *data)
 	return 0; /* Continue walking */
 }
 
+static void __arm_smmu_release_pci_iommudata(void *data)
+{
+	kfree(data);
+}
+
 static int arm_smmu_add_device(struct device *dev)
 {
 	struct arm_smmu_device *smmu;
+	struct arm_smmu_master_cfg *cfg;
 	struct iommu_group *group;
+	void (*releasefn)(void *) = NULL;
 	int ret;
 
-	if (dev->archdata.iommu) {
-		dev_warn(dev, "IOMMU driver already assigned to device\n");
-		return -EINVAL;
-	}
-
 	smmu = find_smmu_for_device(dev);
 	if (!smmu)
 		return -ENODEV;
@@ -1571,7 +1640,6 @@ static int arm_smmu_add_device(struct device *dev)
 	}
 
 	if (dev_is_pci(dev)) {
-		struct arm_smmu_master_cfg *cfg;
 		struct pci_dev *pdev = to_pci_dev(dev);
 
 		cfg = kzalloc(sizeof(*cfg), GFP_KERNEL);
@@ -1587,11 +1655,20 @@ static int arm_smmu_add_device(struct device *dev)
 		 */
 		pci_for_each_dma_alias(pdev, __arm_smmu_get_pci_sid,
 				       &cfg->streamids[0]);
-		dev->archdata.iommu = cfg;
+		releasefn = __arm_smmu_release_pci_iommudata;
 	} else {
-		dev->archdata.iommu = smmu;
+		struct arm_smmu_master *master;
+
+		master = find_smmu_master(smmu, dev->of_node);
+		if (!master) {
+			ret = -ENODEV;
+			goto out_put_group;
+		}
+
+		cfg = &master->cfg;
 	}
 
+	iommu_group_set_iommudata(group, cfg, releasefn);
 	ret = iommu_group_add_device(group, dev);
 
 out_put_group:
@@ -1601,27 +1678,60 @@ out_put_group:
 
 static void arm_smmu_remove_device(struct device *dev)
 {
-	if (dev_is_pci(dev))
-		kfree(dev->archdata.iommu);
-
-	dev->archdata.iommu = NULL;
 	iommu_group_remove_device(dev);
 }
 
+static int arm_smmu_domain_get_attr(struct iommu_domain *domain,
+				    enum iommu_attr attr, void *data)
+{
+	struct arm_smmu_domain *smmu_domain = domain->priv;
+
+	switch (attr) {
+	case DOMAIN_ATTR_NESTING:
+		*(int *)data = (smmu_domain->stage == ARM_SMMU_DOMAIN_NESTED);
+		return 0;
+	default:
+		return -ENODEV;
+	}
+}
+
+static int arm_smmu_domain_set_attr(struct iommu_domain *domain,
+				    enum iommu_attr attr, void *data)
+{
+	struct arm_smmu_domain *smmu_domain = domain->priv;
+
+	switch (attr) {
+	case DOMAIN_ATTR_NESTING:
+		if (smmu_domain->smmu)
+			return -EPERM;
+		if (*(int *)data)
+			smmu_domain->stage = ARM_SMMU_DOMAIN_NESTED;
+		else
+			smmu_domain->stage = ARM_SMMU_DOMAIN_S1;
+
+		return 0;
+	default:
+		return -ENODEV;
+	}
+}
+
 static const struct iommu_ops arm_smmu_ops = {
-	.domain_init	= arm_smmu_domain_init,
-	.domain_destroy	= arm_smmu_domain_destroy,
-	.attach_dev	= arm_smmu_attach_dev,
-	.detach_dev	= arm_smmu_detach_dev,
-	.map		= arm_smmu_map,
-	.unmap		= arm_smmu_unmap,
-	.iova_to_phys	= arm_smmu_iova_to_phys,
-	.domain_has_cap	= arm_smmu_domain_has_cap,
-	.add_device	= arm_smmu_add_device,
-	.remove_device	= arm_smmu_remove_device,
-	.pgsize_bitmap	= (SECTION_SIZE |
-			   ARM_SMMU_PTE_CONT_SIZE |
-			   PAGE_SIZE),
+	.capable		= arm_smmu_capable,
+	.domain_init		= arm_smmu_domain_init,
+	.domain_destroy		= arm_smmu_domain_destroy,
+	.attach_dev		= arm_smmu_attach_dev,
+	.detach_dev		= arm_smmu_detach_dev,
+	.map			= arm_smmu_map,
+	.unmap			= arm_smmu_unmap,
+	.map_sg			= default_iommu_map_sg,
+	.iova_to_phys		= arm_smmu_iova_to_phys,
+	.add_device		= arm_smmu_add_device,
+	.remove_device		= arm_smmu_remove_device,
+	.domain_get_attr	= arm_smmu_domain_get_attr,
+	.domain_set_attr	= arm_smmu_domain_set_attr,
+	.pgsize_bitmap		= (SECTION_SIZE |
+				   ARM_SMMU_PTE_CONT_SIZE |
+				   PAGE_SIZE),
 };
 
 static void arm_smmu_device_reset(struct arm_smmu_device *smmu)
@@ -1702,10 +1812,6 @@ static int arm_smmu_device_cfg_probe(struct arm_smmu_device *smmu)
 	u32 id;
 
 	dev_notice(smmu->dev, "probing hardware configuration...\n");
-
-	/* Primecell ID */
-	id = readl_relaxed(gr0_base + ARM_SMMU_GR0_PIDR2);
-	smmu->version = ((id >> PIDR2_ARCH_SHIFT) & PIDR2_ARCH_MASK) + 1;
 	dev_notice(smmu->dev, "SMMUv%d with:\n", smmu->version);
 
 	/* ID0 */
@@ -1716,6 +1822,13 @@ static int arm_smmu_device_cfg_probe(struct arm_smmu_device *smmu)
 		return -ENODEV;
 	}
 #endif
+
+	/* Restrict available stages based on module parameter */
+	if (force_stage == 1)
+		id &= ~(ID0_S2TS | ID0_NTS);
+	else if (force_stage == 2)
+		id &= ~(ID0_S1TS | ID0_NTS);
+
 	if (id & ID0_S1TS) {
 		smmu->features |= ARM_SMMU_FEAT_TRANS_S1;
 		dev_notice(smmu->dev, "\tstage 1 translation\n");
@@ -1732,8 +1845,7 @@ static int arm_smmu_device_cfg_probe(struct arm_smmu_device *smmu)
 	}
 
 	if (!(smmu->features &
-		(ARM_SMMU_FEAT_TRANS_S1 | ARM_SMMU_FEAT_TRANS_S2 |
-		 ARM_SMMU_FEAT_TRANS_NESTED))) {
+		(ARM_SMMU_FEAT_TRANS_S1 | ARM_SMMU_FEAT_TRANS_S2))) {
 		dev_err(smmu->dev, "\tno translation support!\n");
 		return -ENODEV;
 	}
@@ -1779,12 +1891,12 @@ static int arm_smmu_device_cfg_probe(struct arm_smmu_device *smmu)
 
 	/* ID1 */
 	id = readl_relaxed(gr0_base + ARM_SMMU_GR0_ID1);
-	smmu->pagesize = (id & ID1_PAGESIZE) ? SZ_64K : SZ_4K;
+	smmu->pgshift = (id & ID1_PAGESIZE) ? 16 : 12;
 
 	/* Check for size mismatch of SMMU address space from mapped region */
 	size = 1 <<
 		(((id >> ID1_NUMPAGENDXB_SHIFT) & ID1_NUMPAGENDXB_MASK) + 1);
-	size *= (smmu->pagesize << 1);
+	size *= 2 << smmu->pgshift;
 	if (smmu->size != size)
 		dev_warn(smmu->dev,
 			"SMMU address space size (0x%lx) differs from mapped region size (0x%lx)!\n",
@@ -1803,28 +1915,21 @@ static int arm_smmu_device_cfg_probe(struct arm_smmu_device *smmu)
 	/* ID2 */
 	id = readl_relaxed(gr0_base + ARM_SMMU_GR0_ID2);
 	size = arm_smmu_id_size_to_bits((id >> ID2_IAS_SHIFT) & ID2_IAS_MASK);
+	smmu->s1_output_size = min_t(unsigned long, PHYS_MASK_SHIFT, size);
 
-	/*
-	 * Stage-1 output limited by stage-2 input size due to pgd
-	 * allocation (PTRS_PER_PGD).
-	 */
-	if (smmu->features & ARM_SMMU_FEAT_TRANS_NESTED) {
+	/* Stage-2 input size limited due to pgd allocation (PTRS_PER_PGD) */
 #ifdef CONFIG_64BIT
-		smmu->s1_output_size = min_t(unsigned long, VA_BITS, size);
+	smmu->s2_input_size = min_t(unsigned long, VA_BITS, size);
 #else
-		smmu->s1_output_size = min(32UL, size);
+	smmu->s2_input_size = min(32UL, size);
 #endif
-	} else {
-		smmu->s1_output_size = min_t(unsigned long, PHYS_MASK_SHIFT,
-					     size);
-	}
 
 	/* The stage-2 output mask is also applied for bypass */
 	size = arm_smmu_id_size_to_bits((id >> ID2_OAS_SHIFT) & ID2_OAS_MASK);
 	smmu->s2_output_size = min_t(unsigned long, PHYS_MASK_SHIFT, size);
 
-	if (smmu->version == 1) {
-		smmu->input_size = 32;
+	if (smmu->version == ARM_SMMU_V1) {
+		smmu->s1_input_size = 32;
 	} else {
 #ifdef CONFIG_64BIT
 		size = (id >> ID2_UBS_SHIFT) & ID2_UBS_MASK;
@@ -1832,7 +1937,7 @@ static int arm_smmu_device_cfg_probe(struct arm_smmu_device *smmu)
 #else
 		size = 32;
 #endif
-		smmu->input_size = size;
+		smmu->s1_input_size = size;
 
 		if ((PAGE_SIZE == SZ_4K && !(id & ID2_PTFS_4K)) ||
 		    (PAGE_SIZE == SZ_64K && !(id & ID2_PTFS_64K)) ||
@@ -1843,15 +1948,30 @@ static int arm_smmu_device_cfg_probe(struct arm_smmu_device *smmu)
 		}
 	}
 
-	dev_notice(smmu->dev,
-		   "\t%lu-bit VA, %lu-bit IPA, %lu-bit PA\n",
-		   smmu->input_size, smmu->s1_output_size,
-		   smmu->s2_output_size);
+	if (smmu->features & ARM_SMMU_FEAT_TRANS_S1)
+		dev_notice(smmu->dev, "\tStage-1: %lu-bit VA -> %lu-bit IPA\n",
+			   smmu->s1_input_size, smmu->s1_output_size);
+
+	if (smmu->features & ARM_SMMU_FEAT_TRANS_S2)
+		dev_notice(smmu->dev, "\tStage-2: %lu-bit IPA -> %lu-bit PA\n",
+			   smmu->s2_input_size, smmu->s2_output_size);
+
 	return 0;
 }
 
+static const struct of_device_id arm_smmu_of_match[] = {
+	{ .compatible = "arm,smmu-v1", .data = (void *)ARM_SMMU_V1 },
+	{ .compatible = "arm,smmu-v2", .data = (void *)ARM_SMMU_V2 },
+	{ .compatible = "arm,mmu-400", .data = (void *)ARM_SMMU_V1 },
+	{ .compatible = "arm,mmu-401", .data = (void *)ARM_SMMU_V1 },
+	{ .compatible = "arm,mmu-500", .data = (void *)ARM_SMMU_V2 },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, arm_smmu_of_match);
+
 static int arm_smmu_device_dt_probe(struct platform_device *pdev)
 {
+	const struct of_device_id *of_id;
 	struct resource *res;
 	struct arm_smmu_device *smmu;
 	struct device *dev = &pdev->dev;
@@ -1866,6 +1986,9 @@ static int arm_smmu_device_dt_probe(struct platform_device *pdev)
 	}
 	smmu->dev = dev;
 
+	of_id = of_match_node(arm_smmu_of_match, dev->of_node);
+	smmu->version = (enum arm_smmu_arch_version)of_id->data;
+
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	smmu->base = devm_ioremap_resource(dev, res);
 	if (IS_ERR(smmu->base))
@@ -1930,7 +2053,7 @@ static int arm_smmu_device_dt_probe(struct platform_device *pdev)
 
 	parse_driver_options(smmu);
 
-	if (smmu->version > 1 &&
+	if (smmu->version > ARM_SMMU_V1 &&
 	    smmu->num_context_banks != smmu->num_context_irqs) {
 		dev_err(dev,
 			"found only %d context interrupt(s) but %d required\n",
@@ -2011,20 +2134,8 @@ static int arm_smmu_device_remove(struct platform_device *pdev)
 	return 0;
 }
 
-#ifdef CONFIG_OF
-static struct of_device_id arm_smmu_of_match[] = {
-	{ .compatible = "arm,smmu-v1", },
-	{ .compatible = "arm,smmu-v2", },
-	{ .compatible = "arm,mmu-400", },
-	{ .compatible = "arm,mmu-500", },
-	{ },
-};
-MODULE_DEVICE_TABLE(of, arm_smmu_of_match);
-#endif
-
 static struct platform_driver arm_smmu_driver = {
 	.driver	= {
-		.owner		= THIS_MODULE,
 		.name		= "arm-smmu",
 		.of_match_table	= of_match_ptr(arm_smmu_of_match),
 	},
@@ -2034,8 +2145,20 @@ static struct platform_driver arm_smmu_driver = {
 
 static int __init arm_smmu_init(void)
 {
+	struct device_node *np;
 	int ret;
 
+	/*
+	 * Play nice with systems that don't have an ARM SMMU by checking that
+	 * an ARM SMMU exists in the system before proceeding with the driver
+	 * and IOMMU bus operation registration.
+	 */
+	np = of_find_matching_node(NULL, arm_smmu_of_match);
+	if (!np)
+		return 0;
+
+	of_node_put(np);
+
 	ret = platform_driver_register(&arm_smmu_driver);
 	if (ret)
 		return ret;
diff --git a/drivers/iommu/dmar.c b/drivers/iommu/dmar.c
index 06d268abe951..9847613085e1 100644
--- a/drivers/iommu/dmar.c
+++ b/drivers/iommu/dmar.c
@@ -44,6 +44,14 @@
 
 #include "irq_remapping.h"
 
+typedef int (*dmar_res_handler_t)(struct acpi_dmar_header *, void *);
+struct dmar_res_callback {
+	dmar_res_handler_t	cb[ACPI_DMAR_TYPE_RESERVED];
+	void			*arg[ACPI_DMAR_TYPE_RESERVED];
+	bool			ignore_unhandled;
+	bool			print_entry;
+};
+
 /*
  * Assumptions:
  * 1) The hotplug framework guarentees that DMAR unit will be hot-added
@@ -62,11 +70,12 @@ LIST_HEAD(dmar_drhd_units);
 struct acpi_table_header * __initdata dmar_tbl;
 static acpi_size dmar_tbl_size;
 static int dmar_dev_scope_status = 1;
+static unsigned long dmar_seq_ids[BITS_TO_LONGS(DMAR_UNITS_SUPPORTED)];
 
 static int alloc_iommu(struct dmar_drhd_unit *drhd);
 static void free_iommu(struct intel_iommu *iommu);
 
-static void __init dmar_register_drhd_unit(struct dmar_drhd_unit *drhd)
+static void dmar_register_drhd_unit(struct dmar_drhd_unit *drhd)
 {
 	/*
 	 * add INCLUDE_ALL at the tail, so scan the list will find it at
@@ -155,6 +164,7 @@ dmar_alloc_pci_notify_info(struct pci_dev *dev, unsigned long event)
 	if (event == BUS_NOTIFY_ADD_DEVICE) {
 		for (tmp = dev; tmp; tmp = tmp->bus->self) {
 			level--;
+			info->path[level].bus = tmp->bus->number;
 			info->path[level].device = PCI_SLOT(tmp->devfn);
 			info->path[level].function = PCI_FUNC(tmp->devfn);
 			if (pci_is_root_bus(tmp->bus))
@@ -177,17 +187,33 @@ static bool dmar_match_pci_path(struct dmar_pci_notify_info *info, int bus,
 	int i;
 
 	if (info->bus != bus)
-		return false;
+		goto fallback;
 	if (info->level != count)
-		return false;
+		goto fallback;
 
 	for (i = 0; i < count; i++) {
 		if (path[i].device != info->path[i].device ||
 		    path[i].function != info->path[i].function)
-			return false;
+			goto fallback;
 	}
 
 	return true;
+
+fallback:
+
+	if (count != 1)
+		return false;
+
+	i = info->level - 1;
+	if (bus              == info->path[i].bus &&
+	    path[0].device   == info->path[i].device &&
+	    path[0].function == info->path[i].function) {
+		pr_info(FW_BUG "RMRR entry for device %02x:%02x.%x is broken - applying workaround\n",
+			bus, path[0].device, path[0].function);
+		return true;
+	}
+
+	return false;
 }
 
 /* Return: > 0 if match found, 0 if no match found, < 0 if error happens */
@@ -247,7 +273,7 @@ int dmar_remove_dev_scope(struct dmar_pci_notify_info *info, u16 segment,
 
 	for_each_active_dev_scope(devices, count, index, tmp)
 		if (tmp == &info->dev->dev) {
-			rcu_assign_pointer(devices[index].dev, NULL);
+			RCU_INIT_POINTER(devices[index].dev, NULL);
 			synchronize_rcu();
 			put_device(tmp);
 			return 1;
@@ -327,24 +353,45 @@ static struct notifier_block dmar_pci_bus_nb = {
 	.priority = INT_MIN,
 };
 
+static struct dmar_drhd_unit *
+dmar_find_dmaru(struct acpi_dmar_hardware_unit *drhd)
+{
+	struct dmar_drhd_unit *dmaru;
+
+	list_for_each_entry_rcu(dmaru, &dmar_drhd_units, list)
+		if (dmaru->segment == drhd->segment &&
+		    dmaru->reg_base_addr == drhd->address)
+			return dmaru;
+
+	return NULL;
+}
+
 /**
  * dmar_parse_one_drhd - parses exactly one DMA remapping hardware definition
  * structure which uniquely represent one DMA remapping hardware unit
  * present in the platform
  */
-static int __init
-dmar_parse_one_drhd(struct acpi_dmar_header *header)
+static int dmar_parse_one_drhd(struct acpi_dmar_header *header, void *arg)
 {
 	struct acpi_dmar_hardware_unit *drhd;
 	struct dmar_drhd_unit *dmaru;
 	int ret = 0;
 
 	drhd = (struct acpi_dmar_hardware_unit *)header;
-	dmaru = kzalloc(sizeof(*dmaru), GFP_KERNEL);
+	dmaru = dmar_find_dmaru(drhd);
+	if (dmaru)
+		goto out;
+
+	dmaru = kzalloc(sizeof(*dmaru) + header->length, GFP_KERNEL);
 	if (!dmaru)
 		return -ENOMEM;
 
-	dmaru->hdr = header;
+	/*
+	 * If header is allocated from slab by ACPI _DSM method, we need to
+	 * copy the content because the memory buffer will be freed on return.
+	 */
+	dmaru->hdr = (void *)(dmaru + 1);
+	memcpy(dmaru->hdr, header, header->length);
 	dmaru->reg_base_addr = drhd->address;
 	dmaru->segment = drhd->segment;
 	dmaru->include_all = drhd->flags & 0x1; /* BIT0: INCLUDE_ALL */
@@ -364,6 +411,11 @@ dmar_parse_one_drhd(struct acpi_dmar_header *header)
 		return ret;
 	}
 	dmar_register_drhd_unit(dmaru);
+
+out:
+	if (arg)
+		(*(int *)arg)++;
+
 	return 0;
 }
 
@@ -376,7 +428,8 @@ static void dmar_free_drhd(struct dmar_drhd_unit *dmaru)
 	kfree(dmaru);
 }
 
-static int __init dmar_parse_one_andd(struct acpi_dmar_header *header)
+static int __init dmar_parse_one_andd(struct acpi_dmar_header *header,
+				      void *arg)
 {
 	struct acpi_dmar_andd *andd = (void *)header;
 
@@ -397,8 +450,7 @@ static int __init dmar_parse_one_andd(struct acpi_dmar_header *header)
 }
 
 #ifdef CONFIG_ACPI_NUMA
-static int __init
-dmar_parse_one_rhsa(struct acpi_dmar_header *header)
+static int dmar_parse_one_rhsa(struct acpi_dmar_header *header, void *arg)
 {
 	struct acpi_dmar_rhsa *rhsa;
 	struct dmar_drhd_unit *drhd;
@@ -425,6 +477,8 @@ dmar_parse_one_rhsa(struct acpi_dmar_header *header)
 
 	return 0;
 }
+#else
+#define	dmar_parse_one_rhsa		dmar_res_noop
 #endif
 
 static void __init
@@ -486,6 +540,52 @@ static int __init dmar_table_detect(void)
 	return (ACPI_SUCCESS(status) ? 1 : 0);
 }
 
+static int dmar_walk_remapping_entries(struct acpi_dmar_header *start,
+				       size_t len, struct dmar_res_callback *cb)
+{
+	int ret = 0;
+	struct acpi_dmar_header *iter, *next;
+	struct acpi_dmar_header *end = ((void *)start) + len;
+
+	for (iter = start; iter < end && ret == 0; iter = next) {
+		next = (void *)iter + iter->length;
+		if (iter->length == 0) {
+			/* Avoid looping forever on bad ACPI tables */
+			pr_debug(FW_BUG "Invalid 0-length structure\n");
+			break;
+		} else if (next > end) {
+			/* Avoid passing table end */
+			pr_warn(FW_BUG "record passes table end\n");
+			ret = -EINVAL;
+			break;
+		}
+
+		if (cb->print_entry)
+			dmar_table_print_dmar_entry(iter);
+
+		if (iter->type >= ACPI_DMAR_TYPE_RESERVED) {
+			/* continue for forward compatibility */
+			pr_debug("Unknown DMAR structure type %d\n",
+				 iter->type);
+		} else if (cb->cb[iter->type]) {
+			ret = cb->cb[iter->type](iter, cb->arg[iter->type]);
+		} else if (!cb->ignore_unhandled) {
+			pr_warn("No handler for DMAR structure type %d\n",
+				iter->type);
+			ret = -EINVAL;
+		}
+	}
+
+	return ret;
+}
+
+static inline int dmar_walk_dmar_table(struct acpi_table_dmar *dmar,
+				       struct dmar_res_callback *cb)
+{
+	return dmar_walk_remapping_entries((void *)(dmar + 1),
+			dmar->header.length - sizeof(*dmar), cb);
+}
+
 /**
  * parse_dmar_table - parses the DMA reporting table
  */
@@ -493,9 +593,18 @@ static int __init
 parse_dmar_table(void)
 {
 	struct acpi_table_dmar *dmar;
-	struct acpi_dmar_header *entry_header;
 	int ret = 0;
 	int drhd_count = 0;
+	struct dmar_res_callback cb = {
+		.print_entry = true,
+		.ignore_unhandled = true,
+		.arg[ACPI_DMAR_TYPE_HARDWARE_UNIT] = &drhd_count,
+		.cb[ACPI_DMAR_TYPE_HARDWARE_UNIT] = &dmar_parse_one_drhd,
+		.cb[ACPI_DMAR_TYPE_RESERVED_MEMORY] = &dmar_parse_one_rmrr,
+		.cb[ACPI_DMAR_TYPE_ROOT_ATS] = &dmar_parse_one_atsr,
+		.cb[ACPI_DMAR_TYPE_HARDWARE_AFFINITY] = &dmar_parse_one_rhsa,
+		.cb[ACPI_DMAR_TYPE_NAMESPACE] = &dmar_parse_one_andd,
+	};
 
 	/*
 	 * Do it again, earlier dmar_tbl mapping could be mapped with
@@ -519,51 +628,10 @@ parse_dmar_table(void)
 	}
 
 	pr_info("Host address width %d\n", dmar->width + 1);
-
-	entry_header = (struct acpi_dmar_header *)(dmar + 1);
-	while (((unsigned long)entry_header) <
-			(((unsigned long)dmar) + dmar_tbl->length)) {
-		/* Avoid looping forever on bad ACPI tables */
-		if (entry_header->length == 0) {
-			pr_warn("Invalid 0-length structure\n");
-			ret = -EINVAL;
-			break;
-		}
-
-		dmar_table_print_dmar_entry(entry_header);
-
-		switch (entry_header->type) {
-		case ACPI_DMAR_TYPE_HARDWARE_UNIT:
-			drhd_count++;
-			ret = dmar_parse_one_drhd(entry_header);
-			break;
-		case ACPI_DMAR_TYPE_RESERVED_MEMORY:
-			ret = dmar_parse_one_rmrr(entry_header);
-			break;
-		case ACPI_DMAR_TYPE_ROOT_ATS:
-			ret = dmar_parse_one_atsr(entry_header);
-			break;
-		case ACPI_DMAR_TYPE_HARDWARE_AFFINITY:
-#ifdef CONFIG_ACPI_NUMA
-			ret = dmar_parse_one_rhsa(entry_header);
-#endif
-			break;
-		case ACPI_DMAR_TYPE_NAMESPACE:
-			ret = dmar_parse_one_andd(entry_header);
-			break;
-		default:
-			pr_warn("Unknown DMAR structure type %d\n",
-				entry_header->type);
-			ret = 0; /* for forward compatibility */
-			break;
-		}
-		if (ret)
-			break;
-
-		entry_header = ((void *)entry_header + entry_header->length);
-	}
-	if (drhd_count == 0)
+	ret = dmar_walk_dmar_table(dmar, &cb);
+	if (ret == 0 && drhd_count == 0)
 		pr_warn(FW_BUG "No DRHD structure found in DMAR table\n");
+
 	return ret;
 }
 
@@ -761,76 +829,68 @@ static void warn_invalid_dmar(u64 addr, const char *message)
 		dmi_get_system_info(DMI_PRODUCT_VERSION));
 }
 
-static int __init check_zero_address(void)
+static int __ref
+dmar_validate_one_drhd(struct acpi_dmar_header *entry, void *arg)
 {
-	struct acpi_table_dmar *dmar;
-	struct acpi_dmar_header *entry_header;
 	struct acpi_dmar_hardware_unit *drhd;
+	void __iomem *addr;
+	u64 cap, ecap;
 
-	dmar = (struct acpi_table_dmar *)dmar_tbl;
-	entry_header = (struct acpi_dmar_header *)(dmar + 1);
-
-	while (((unsigned long)entry_header) <
-			(((unsigned long)dmar) + dmar_tbl->length)) {
-		/* Avoid looping forever on bad ACPI tables */
-		if (entry_header->length == 0) {
-			pr_warn("Invalid 0-length structure\n");
-			return 0;
-		}
+	drhd = (void *)entry;
+	if (!drhd->address) {
+		warn_invalid_dmar(0, "");
+		return -EINVAL;
+	}
 
-		if (entry_header->type == ACPI_DMAR_TYPE_HARDWARE_UNIT) {
-			void __iomem *addr;
-			u64 cap, ecap;
+	if (arg)
+		addr = ioremap(drhd->address, VTD_PAGE_SIZE);
+	else
+		addr = early_ioremap(drhd->address, VTD_PAGE_SIZE);
+	if (!addr) {
+		pr_warn("IOMMU: can't validate: %llx\n", drhd->address);
+		return -EINVAL;
+	}
 
-			drhd = (void *)entry_header;
-			if (!drhd->address) {
-				warn_invalid_dmar(0, "");
-				goto failed;
-			}
+	cap = dmar_readq(addr + DMAR_CAP_REG);
+	ecap = dmar_readq(addr + DMAR_ECAP_REG);
 
-			addr = early_ioremap(drhd->address, VTD_PAGE_SIZE);
-			if (!addr ) {
-				printk("IOMMU: can't validate: %llx\n", drhd->address);
-				goto failed;
-			}
-			cap = dmar_readq(addr + DMAR_CAP_REG);
-			ecap = dmar_readq(addr + DMAR_ECAP_REG);
-			early_iounmap(addr, VTD_PAGE_SIZE);
-			if (cap == (uint64_t)-1 && ecap == (uint64_t)-1) {
-				warn_invalid_dmar(drhd->address,
-						  " returns all ones");
-				goto failed;
-			}
-		}
+	if (arg)
+		iounmap(addr);
+	else
+		early_iounmap(addr, VTD_PAGE_SIZE);
 
-		entry_header = ((void *)entry_header + entry_header->length);
+	if (cap == (uint64_t)-1 && ecap == (uint64_t)-1) {
+		warn_invalid_dmar(drhd->address, " returns all ones");
+		return -EINVAL;
 	}
-	return 1;
 
-failed:
 	return 0;
 }
 
 int __init detect_intel_iommu(void)
 {
 	int ret;
+	struct dmar_res_callback validate_drhd_cb = {
+		.cb[ACPI_DMAR_TYPE_HARDWARE_UNIT] = &dmar_validate_one_drhd,
+		.ignore_unhandled = true,
+	};
 
 	down_write(&dmar_global_lock);
 	ret = dmar_table_detect();
 	if (ret)
-		ret = check_zero_address();
-	{
-		if (ret && !no_iommu && !iommu_detected && !dmar_disabled) {
-			iommu_detected = 1;
-			/* Make sure ACS will be enabled */
-			pci_request_acs();
-		}
+		ret = !dmar_walk_dmar_table((struct acpi_table_dmar *)dmar_tbl,
+					    &validate_drhd_cb);
+	if (ret && !no_iommu && !iommu_detected && !dmar_disabled) {
+		iommu_detected = 1;
+		/* Make sure ACS will be enabled */
+		pci_request_acs();
+	}
 
 #ifdef CONFIG_X86
-		if (ret)
-			x86_init.iommu.iommu_init = intel_iommu_init;
+	if (ret)
+		x86_init.iommu.iommu_init = intel_iommu_init;
 #endif
-	}
+
 	early_acpi_os_unmap_memory((void __iomem *)dmar_tbl, dmar_tbl_size);
 	dmar_tbl = NULL;
 	up_write(&dmar_global_lock);
@@ -914,11 +974,32 @@ out:
 	return err;
 }
 
+static int dmar_alloc_seq_id(struct intel_iommu *iommu)
+{
+	iommu->seq_id = find_first_zero_bit(dmar_seq_ids,
+					    DMAR_UNITS_SUPPORTED);
+	if (iommu->seq_id >= DMAR_UNITS_SUPPORTED) {
+		iommu->seq_id = -1;
+	} else {
+		set_bit(iommu->seq_id, dmar_seq_ids);
+		sprintf(iommu->name, "dmar%d", iommu->seq_id);
+	}
+
+	return iommu->seq_id;
+}
+
+static void dmar_free_seq_id(struct intel_iommu *iommu)
+{
+	if (iommu->seq_id >= 0) {
+		clear_bit(iommu->seq_id, dmar_seq_ids);
+		iommu->seq_id = -1;
+	}
+}
+
 static int alloc_iommu(struct dmar_drhd_unit *drhd)
 {
 	struct intel_iommu *iommu;
 	u32 ver, sts;
-	static int iommu_allocated = 0;
 	int agaw = 0;
 	int msagaw = 0;
 	int err;
@@ -932,13 +1013,16 @@ static int alloc_iommu(struct dmar_drhd_unit *drhd)
 	if (!iommu)
 		return -ENOMEM;
 
-	iommu->seq_id = iommu_allocated++;
-	sprintf (iommu->name, "dmar%d", iommu->seq_id);
+	if (dmar_alloc_seq_id(iommu) < 0) {
+		pr_err("IOMMU: failed to allocate seq_id\n");
+		err = -ENOSPC;
+		goto error;
+	}
 
 	err = map_iommu(iommu, drhd->reg_base_addr);
 	if (err) {
 		pr_err("IOMMU: failed to map %s\n", iommu->name);
-		goto error;
+		goto error_free_seq_id;
 	}
 
 	err = -EINVAL;
@@ -988,9 +1072,11 @@ static int alloc_iommu(struct dmar_drhd_unit *drhd)
 
 	return 0;
 
- err_unmap:
+err_unmap:
 	unmap_iommu(iommu);
- error:
+error_free_seq_id:
+	dmar_free_seq_id(iommu);
+error:
 	kfree(iommu);
 	return err;
 }
@@ -1014,6 +1100,7 @@ static void free_iommu(struct intel_iommu *iommu)
 	if (iommu->reg)
 		unmap_iommu(iommu);
 
+	dmar_free_seq_id(iommu);
 	kfree(iommu);
 }
 
@@ -1644,12 +1731,17 @@ int __init dmar_ir_support(void)
 	return dmar->flags & 0x1;
 }
 
+/* Check whether DMAR units are in use */
+static inline bool dmar_in_use(void)
+{
+	return irq_remapping_enabled || intel_iommu_enabled;
+}
+
 static int __init dmar_free_unused_resources(void)
 {
 	struct dmar_drhd_unit *dmaru, *dmaru_n;
 
-	/* DMAR units are in use */
-	if (irq_remapping_enabled || intel_iommu_enabled)
+	if (dmar_in_use())
 		return 0;
 
 	if (dmar_dev_scope_status != 1 && !list_empty(&dmar_drhd_units))
@@ -1667,3 +1759,242 @@ static int __init dmar_free_unused_resources(void)
 
 late_initcall(dmar_free_unused_resources);
 IOMMU_INIT_POST(detect_intel_iommu);
+
+/*
+ * DMAR Hotplug Support
+ * For more details, please refer to Intel(R) Virtualization Technology
+ * for Directed-IO Architecture Specifiction, Rev 2.2, Section 8.8
+ * "Remapping Hardware Unit Hot Plug".
+ */
+static u8 dmar_hp_uuid[] = {
+	/* 0000 */    0xA6, 0xA3, 0xC1, 0xD8, 0x9B, 0xBE, 0x9B, 0x4C,
+	/* 0008 */    0x91, 0xBF, 0xC3, 0xCB, 0x81, 0xFC, 0x5D, 0xAF
+};
+
+/*
+ * Currently there's only one revision and BIOS will not check the revision id,
+ * so use 0 for safety.
+ */
+#define	DMAR_DSM_REV_ID			0
+#define	DMAR_DSM_FUNC_DRHD		1
+#define	DMAR_DSM_FUNC_ATSR		2
+#define	DMAR_DSM_FUNC_RHSA		3
+
+static inline bool dmar_detect_dsm(acpi_handle handle, int func)
+{
+	return acpi_check_dsm(handle, dmar_hp_uuid, DMAR_DSM_REV_ID, 1 << func);
+}
+
+static int dmar_walk_dsm_resource(acpi_handle handle, int func,
+				  dmar_res_handler_t handler, void *arg)
+{
+	int ret = -ENODEV;
+	union acpi_object *obj;
+	struct acpi_dmar_header *start;
+	struct dmar_res_callback callback;
+	static int res_type[] = {
+		[DMAR_DSM_FUNC_DRHD] = ACPI_DMAR_TYPE_HARDWARE_UNIT,
+		[DMAR_DSM_FUNC_ATSR] = ACPI_DMAR_TYPE_ROOT_ATS,
+		[DMAR_DSM_FUNC_RHSA] = ACPI_DMAR_TYPE_HARDWARE_AFFINITY,
+	};
+
+	if (!dmar_detect_dsm(handle, func))
+		return 0;
+
+	obj = acpi_evaluate_dsm_typed(handle, dmar_hp_uuid, DMAR_DSM_REV_ID,
+				      func, NULL, ACPI_TYPE_BUFFER);
+	if (!obj)
+		return -ENODEV;
+
+	memset(&callback, 0, sizeof(callback));
+	callback.cb[res_type[func]] = handler;
+	callback.arg[res_type[func]] = arg;
+	start = (struct acpi_dmar_header *)obj->buffer.pointer;
+	ret = dmar_walk_remapping_entries(start, obj->buffer.length, &callback);
+
+	ACPI_FREE(obj);
+
+	return ret;
+}
+
+static int dmar_hp_add_drhd(struct acpi_dmar_header *header, void *arg)
+{
+	int ret;
+	struct dmar_drhd_unit *dmaru;
+
+	dmaru = dmar_find_dmaru((struct acpi_dmar_hardware_unit *)header);
+	if (!dmaru)
+		return -ENODEV;
+
+	ret = dmar_ir_hotplug(dmaru, true);
+	if (ret == 0)
+		ret = dmar_iommu_hotplug(dmaru, true);
+
+	return ret;
+}
+
+static int dmar_hp_remove_drhd(struct acpi_dmar_header *header, void *arg)
+{
+	int i, ret;
+	struct device *dev;
+	struct dmar_drhd_unit *dmaru;
+
+	dmaru = dmar_find_dmaru((struct acpi_dmar_hardware_unit *)header);
+	if (!dmaru)
+		return 0;
+
+	/*
+	 * All PCI devices managed by this unit should have been destroyed.
+	 */
+	if (!dmaru->include_all && dmaru->devices && dmaru->devices_cnt)
+		for_each_active_dev_scope(dmaru->devices,
+					  dmaru->devices_cnt, i, dev)
+			return -EBUSY;
+
+	ret = dmar_ir_hotplug(dmaru, false);
+	if (ret == 0)
+		ret = dmar_iommu_hotplug(dmaru, false);
+
+	return ret;
+}
+
+static int dmar_hp_release_drhd(struct acpi_dmar_header *header, void *arg)
+{
+	struct dmar_drhd_unit *dmaru;
+
+	dmaru = dmar_find_dmaru((struct acpi_dmar_hardware_unit *)header);
+	if (dmaru) {
+		list_del_rcu(&dmaru->list);
+		synchronize_rcu();
+		dmar_free_drhd(dmaru);
+	}
+
+	return 0;
+}
+
+static int dmar_hotplug_insert(acpi_handle handle)
+{
+	int ret;
+	int drhd_count = 0;
+
+	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
+				     &dmar_validate_one_drhd, (void *)1);
+	if (ret)
+		goto out;
+
+	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
+				     &dmar_parse_one_drhd, (void *)&drhd_count);
+	if (ret == 0 && drhd_count == 0) {
+		pr_warn(FW_BUG "No DRHD structures in buffer returned by _DSM method\n");
+		goto out;
+	} else if (ret) {
+		goto release_drhd;
+	}
+
+	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_RHSA,
+				     &dmar_parse_one_rhsa, NULL);
+	if (ret)
+		goto release_drhd;
+
+	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_ATSR,
+				     &dmar_parse_one_atsr, NULL);
+	if (ret)
+		goto release_atsr;
+
+	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
+				     &dmar_hp_add_drhd, NULL);
+	if (!ret)
+		return 0;
+
+	dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
+			       &dmar_hp_remove_drhd, NULL);
+release_atsr:
+	dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_ATSR,
+			       &dmar_release_one_atsr, NULL);
+release_drhd:
+	dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
+			       &dmar_hp_release_drhd, NULL);
+out:
+	return ret;
+}
+
+static int dmar_hotplug_remove(acpi_handle handle)
+{
+	int ret;
+
+	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_ATSR,
+				     &dmar_check_one_atsr, NULL);
+	if (ret)
+		return ret;
+
+	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
+				     &dmar_hp_remove_drhd, NULL);
+	if (ret == 0) {
+		WARN_ON(dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_ATSR,
+					       &dmar_release_one_atsr, NULL));
+		WARN_ON(dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
+					       &dmar_hp_release_drhd, NULL));
+	} else {
+		dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
+				       &dmar_hp_add_drhd, NULL);
+	}
+
+	return ret;
+}
+
+static acpi_status dmar_get_dsm_handle(acpi_handle handle, u32 lvl,
+				       void *context, void **retval)
+{
+	acpi_handle *phdl = retval;
+
+	if (dmar_detect_dsm(handle, DMAR_DSM_FUNC_DRHD)) {
+		*phdl = handle;
+		return AE_CTRL_TERMINATE;
+	}
+
+	return AE_OK;
+}
+
+static int dmar_device_hotplug(acpi_handle handle, bool insert)
+{
+	int ret;
+	acpi_handle tmp = NULL;
+	acpi_status status;
+
+	if (!dmar_in_use())
+		return 0;
+
+	if (dmar_detect_dsm(handle, DMAR_DSM_FUNC_DRHD)) {
+		tmp = handle;
+	} else {
+		status = acpi_walk_namespace(ACPI_TYPE_DEVICE, handle,
+					     ACPI_UINT32_MAX,
+					     dmar_get_dsm_handle,
+					     NULL, NULL, &tmp);
+		if (ACPI_FAILURE(status)) {
+			pr_warn("Failed to locate _DSM method.\n");
+			return -ENXIO;
+		}
+	}
+	if (tmp == NULL)
+		return 0;
+
+	down_write(&dmar_global_lock);
+	if (insert)
+		ret = dmar_hotplug_insert(tmp);
+	else
+		ret = dmar_hotplug_remove(tmp);
+	up_write(&dmar_global_lock);
+
+	return ret;
+}
+
+int dmar_device_add(acpi_handle handle)
+{
+	return dmar_device_hotplug(handle, true);
+}
+
+int dmar_device_remove(acpi_handle handle)
+{
+	return dmar_device_hotplug(handle, false);
+}
diff --git a/drivers/iommu/exynos-iommu.c b/drivers/iommu/exynos-iommu.c
index d037e87a1fe5..7ce52737c7a1 100644
--- a/drivers/iommu/exynos-iommu.c
+++ b/drivers/iommu/exynos-iommu.c
@@ -32,7 +32,7 @@
 typedef u32 sysmmu_iova_t;
 typedef u32 sysmmu_pte_t;
 
-/* We does not consider super section mapping (16MB) */
+/* We do not consider super section mapping (16MB) */
 #define SECT_ORDER 20
 #define LPAGE_ORDER 16
 #define SPAGE_ORDER 12
@@ -307,7 +307,7 @@ static void show_fault_information(const char *name,
 
 static irqreturn_t exynos_sysmmu_irq(int irq, void *dev_id)
 {
-	/* SYSMMU is in blocked when interrupt occurred. */
+	/* SYSMMU is in blocked state when interrupt occurred. */
 	struct sysmmu_drvdata *data = dev_id;
 	enum exynos_sysmmu_inttype itype;
 	sysmmu_iova_t addr = -1;
@@ -567,8 +567,8 @@ static void sysmmu_tlb_invalidate_entry(struct device *dev, sysmmu_iova_t iova,
 		/*
 		 * L2TLB invalidation required
 		 * 4KB page: 1 invalidation
-		 * 64KB page: 16 invalidation
-		 * 1MB page: 64 invalidation
+		 * 64KB page: 16 invalidations
+		 * 1MB page: 64 invalidations
 		 * because it is set-associative TLB
 		 * with 8-way and 64 sets.
 		 * 1MB page can be cached in one of all sets.
@@ -684,7 +684,6 @@ static const struct of_device_id sysmmu_of_match[] __initconst = {
 static struct platform_driver exynos_sysmmu_driver __refdata = {
 	.probe	= exynos_sysmmu_probe,
 	.driver	= {
-		.owner		= THIS_MODULE,
 		.name		= "exynos-sysmmu",
 		.of_match_table	= sysmmu_of_match,
 	}
@@ -714,7 +713,7 @@ static int exynos_iommu_domain_init(struct iommu_domain *domain)
 	if (!priv->lv2entcnt)
 		goto err_counter;
 
-	/* w/a of System MMU v3.3 to prevent caching 1MiB mapping */
+	/* Workaround for System MMU v3.3 to prevent caching 1MiB mapping */
 	for (i = 0; i < NUM_LV1ENTRIES; i += 8) {
 		priv->pgtable[i + 0] = ZERO_LV2LINK;
 		priv->pgtable[i + 1] = ZERO_LV2LINK;
@@ -861,14 +860,14 @@ static sysmmu_pte_t *alloc_lv2entry(struct exynos_iommu_domain *priv,
 		pgtable_flush(sent, sent + 1);
 
 		/*
-		 * If pretched SLPD is a fault SLPD in zero_l2_table, FLPD cache
-		 * may caches the address of zero_l2_table. This function
-		 * replaces the zero_l2_table with new L2 page table to write
-		 * valid mappings.
+		 * If pre-fetched SLPD is a faulty SLPD in zero_l2_table,
+		 * FLPD cache may cache the address of zero_l2_table. This
+		 * function replaces the zero_l2_table with new L2 page table
+		 * to write valid mappings.
 		 * Accessing the valid area may cause page fault since FLPD
-		 * cache may still caches zero_l2_table for the valid area
-		 * instead of new L2 page table that have the mapping
-		 * information of the valid area
+		 * cache may still cache zero_l2_table for the valid area
+		 * instead of new L2 page table that has the mapping
+		 * information of the valid area.
 		 * Thus any replacement of zero_l2_table with other valid L2
 		 * page table must involve FLPD cache invalidation for System
 		 * MMU v3.3.
@@ -963,27 +962,27 @@ static int lv2set_page(sysmmu_pte_t *pent, phys_addr_t paddr, size_t size,
 /*
  * *CAUTION* to the I/O virtual memory managers that support exynos-iommu:
  *
- * System MMU v3.x have an advanced logic to improve address translation
+ * System MMU v3.x has advanced logic to improve address translation
  * performance with caching more page table entries by a page table walk.
- * However, the logic has a bug that caching fault page table entries and System
- * MMU reports page fault if the cached fault entry is hit even though the fault
- * entry is updated to a valid entry after the entry is cached.
- * To prevent caching fault page table entries which may be updated to valid
- * entries later, the virtual memory manager should care about the w/a about the
- * problem. The followings describe w/a.
+ * However, the logic has a bug that while caching faulty page table entries,
+ * System MMU reports page fault if the cached fault entry is hit even though
+ * the fault entry is updated to a valid entry after the entry is cached.
+ * To prevent caching faulty page table entries which may be updated to valid
+ * entries later, the virtual memory manager should care about the workaround
+ * for the problem. The following describes the workaround.
  *
  * Any two consecutive I/O virtual address regions must have a hole of 128KiB
- * in maximum to prevent misbehavior of System MMU 3.x. (w/a of h/w bug)
+ * at maximum to prevent misbehavior of System MMU 3.x (workaround for h/w bug).
  *
- * Precisely, any start address of I/O virtual region must be aligned by
+ * Precisely, any start address of I/O virtual region must be aligned with
  * the following sizes for System MMU v3.1 and v3.2.
  * System MMU v3.1: 128KiB
  * System MMU v3.2: 256KiB
  *
  * Because System MMU v3.3 caches page table entries more aggressively, it needs
- * more w/a.
- * - Any two consecutive I/O virtual regions must be have a hole of larger size
- *   than or equal size to 128KiB.
+ * more workarounds.
+ * - Any two consecutive I/O virtual regions must have a hole of size larger
+ *   than or equal to 128KiB.
  * - Start address of an I/O virtual region must be aligned by 128KiB.
  */
 static int exynos_iommu_map(struct iommu_domain *domain, unsigned long l_iova,
@@ -1061,7 +1060,8 @@ static size_t exynos_iommu_unmap(struct iommu_domain *domain,
 			goto err;
 		}
 
-		*ent = ZERO_LV2LINK; /* w/a for h/w bug in Sysmem MMU v3.3 */
+		/* workaround for h/w bug in System MMU v3.3 */
+		*ent = ZERO_LV2LINK;
 		pgtable_flush(ent, ent + 1);
 		size = SECT_SIZE;
 		goto done;
@@ -1177,6 +1177,7 @@ static const struct iommu_ops exynos_iommu_ops = {
 	.detach_dev = exynos_iommu_detach_device,
 	.map = exynos_iommu_map,
 	.unmap = exynos_iommu_unmap,
+	.map_sg = default_iommu_map_sg,
 	.iova_to_phys = exynos_iommu_iova_to_phys,
 	.add_device = exynos_iommu_add_device,
 	.remove_device = exynos_iommu_remove_device,
diff --git a/drivers/iommu/fsl_pamu.c b/drivers/iommu/fsl_pamu.c
index 2b6ce9387af1..80ac68d884c5 100644
--- a/drivers/iommu/fsl_pamu.c
+++ b/drivers/iommu/fsl_pamu.c
@@ -1227,7 +1227,6 @@ static const struct of_device_id fsl_of_pamu_ids[] = {
 static struct platform_driver fsl_of_pamu_driver = {
 	.driver = {
 		.name = "fsl-of-pamu",
-		.owner = THIS_MODULE,
 	},
 	.probe = fsl_pamu_probe,
 };
diff --git a/drivers/iommu/fsl_pamu_domain.c b/drivers/iommu/fsl_pamu_domain.c
index 56feed7cec15..c828f80d48b0 100644
--- a/drivers/iommu/fsl_pamu_domain.c
+++ b/drivers/iommu/fsl_pamu_domain.c
@@ -411,8 +411,7 @@ static phys_addr_t fsl_pamu_iova_to_phys(struct iommu_domain *domain,
 	return get_phys_addr(dma_domain, iova);
 }
 
-static int fsl_pamu_domain_has_cap(struct iommu_domain *domain,
-				      unsigned long cap)
+static bool fsl_pamu_capable(enum iommu_cap cap)
 {
 	return cap == IOMMU_CAP_CACHE_COHERENCY;
 }
@@ -1080,6 +1079,7 @@ static u32 fsl_pamu_get_windows(struct iommu_domain *domain)
 }
 
 static const struct iommu_ops fsl_pamu_ops = {
+	.capable	= fsl_pamu_capable,
 	.domain_init	= fsl_pamu_domain_init,
 	.domain_destroy = fsl_pamu_domain_destroy,
 	.attach_dev	= fsl_pamu_attach_device,
@@ -1089,7 +1089,6 @@ static const struct iommu_ops fsl_pamu_ops = {
 	.domain_get_windows = fsl_pamu_get_windows,
 	.domain_set_windows = fsl_pamu_set_windows,
 	.iova_to_phys	= fsl_pamu_iova_to_phys,
-	.domain_has_cap = fsl_pamu_domain_has_cap,
 	.domain_set_attr = fsl_pamu_set_domain_attr,
 	.domain_get_attr = fsl_pamu_get_domain_attr,
 	.add_device	= fsl_pamu_add_device,
diff --git a/drivers/iommu/intel-iommu.c b/drivers/iommu/intel-iommu.c
index 5619f264862d..40dfbc0444c0 100644
--- a/drivers/iommu/intel-iommu.c
+++ b/drivers/iommu/intel-iommu.c
@@ -195,6 +195,7 @@ static inline void set_root_present(struct root_entry *root)
 }
 static inline void set_root_value(struct root_entry *root, unsigned long value)
 {
+	root->val &= ~VTD_PAGE_MASK;
 	root->val |= value & VTD_PAGE_MASK;
 }
 
@@ -247,6 +248,7 @@ static inline void context_set_translation_type(struct context_entry *context,
 static inline void context_set_address_root(struct context_entry *context,
 					    unsigned long value)
 {
+	context->lo &= ~VTD_PAGE_MASK;
 	context->lo |= value & VTD_PAGE_MASK;
 }
 
@@ -328,17 +330,10 @@ static int hw_pass_through = 1;
 /* si_domain contains mulitple devices */
 #define DOMAIN_FLAG_STATIC_IDENTITY	(1 << 1)
 
-/* define the limit of IOMMUs supported in each domain */
-#ifdef	CONFIG_X86
-# define	IOMMU_UNITS_SUPPORTED	MAX_IO_APICS
-#else
-# define	IOMMU_UNITS_SUPPORTED	64
-#endif
-
 struct dmar_domain {
 	int	id;			/* domain id */
 	int	nid;			/* node id */
-	DECLARE_BITMAP(iommu_bmp, IOMMU_UNITS_SUPPORTED);
+	DECLARE_BITMAP(iommu_bmp, DMAR_UNITS_SUPPORTED);
 					/* bitmap of iommus this domain uses*/
 
 	struct list_head devices; 	/* all devices' list */
@@ -1132,8 +1127,11 @@ static int iommu_alloc_root_entry(struct intel_iommu *iommu)
 	unsigned long flags;
 
 	root = (struct root_entry *)alloc_pgtable_page(iommu->node);
-	if (!root)
+	if (!root) {
+		pr_err("IOMMU: allocating root entry for %s failed\n",
+			iommu->name);
 		return -ENOMEM;
+	}
 
 	__iommu_flush_cache(iommu, root, ROOT_SIZE);
 
@@ -1473,7 +1471,7 @@ static int iommu_init_domains(struct intel_iommu *iommu)
 	return 0;
 }
 
-static void free_dmar_iommu(struct intel_iommu *iommu)
+static void disable_dmar_iommu(struct intel_iommu *iommu)
 {
 	struct dmar_domain *domain;
 	int i;
@@ -1497,11 +1495,16 @@ static void free_dmar_iommu(struct intel_iommu *iommu)
 
 	if (iommu->gcmd & DMA_GCMD_TE)
 		iommu_disable_translation(iommu);
+}
 
-	kfree(iommu->domains);
-	kfree(iommu->domain_ids);
-	iommu->domains = NULL;
-	iommu->domain_ids = NULL;
+static void free_dmar_iommu(struct intel_iommu *iommu)
+{
+	if ((iommu->domains) && (iommu->domain_ids)) {
+		kfree(iommu->domains);
+		kfree(iommu->domain_ids);
+		iommu->domains = NULL;
+		iommu->domain_ids = NULL;
+	}
 
 	g_iommus[iommu->seq_id] = NULL;
 
@@ -1983,7 +1986,7 @@ static int __domain_mapping(struct dmar_domain *domain, unsigned long iov_pfn,
 {
 	struct dma_pte *first_pte = NULL, *pte = NULL;
 	phys_addr_t uninitialized_var(pteval);
-	unsigned long sg_res;
+	unsigned long sg_res = 0;
 	unsigned int largepage_lvl = 0;
 	unsigned long lvl_pages = 0;
 
@@ -1994,10 +1997,8 @@ static int __domain_mapping(struct dmar_domain *domain, unsigned long iov_pfn,
 
 	prot &= DMA_PTE_READ | DMA_PTE_WRITE | DMA_PTE_SNP;
 
-	if (sg)
-		sg_res = 0;
-	else {
-		sg_res = nr_pages + 1;
+	if (!sg) {
+		sg_res = nr_pages;
 		pteval = ((phys_addr_t)phys_pfn << VTD_PAGE_SHIFT) | prot;
 	}
 
@@ -2708,6 +2709,41 @@ static int __init iommu_prepare_static_identity_mapping(int hw)
 	return 0;
 }
 
+static void intel_iommu_init_qi(struct intel_iommu *iommu)
+{
+	/*
+	 * Start from the sane iommu hardware state.
+	 * If the queued invalidation is already initialized by us
+	 * (for example, while enabling interrupt-remapping) then
+	 * we got the things already rolling from a sane state.
+	 */
+	if (!iommu->qi) {
+		/*
+		 * Clear any previous faults.
+		 */
+		dmar_fault(-1, iommu);
+		/*
+		 * Disable queued invalidation if supported and already enabled
+		 * before OS handover.
+		 */
+		dmar_disable_qi(iommu);
+	}
+
+	if (dmar_enable_qi(iommu)) {
+		/*
+		 * Queued Invalidate not enabled, use Register Based Invalidate
+		 */
+		iommu->flush.flush_context = __iommu_flush_context;
+		iommu->flush.flush_iotlb = __iommu_flush_iotlb;
+		pr_info("IOMMU: %s using Register based invalidation\n",
+			iommu->name);
+	} else {
+		iommu->flush.flush_context = qi_flush_context;
+		iommu->flush.flush_iotlb = qi_flush_iotlb;
+		pr_info("IOMMU: %s using Queued invalidation\n", iommu->name);
+	}
+}
+
 static int __init init_dmars(void)
 {
 	struct dmar_drhd_unit *drhd;
@@ -2728,14 +2764,18 @@ static int __init init_dmars(void)
 		 * threaded kernel __init code path all other access are read
 		 * only
 		 */
-		if (g_num_of_iommus < IOMMU_UNITS_SUPPORTED) {
+		if (g_num_of_iommus < DMAR_UNITS_SUPPORTED) {
 			g_num_of_iommus++;
 			continue;
 		}
 		printk_once(KERN_ERR "intel-iommu: exceeded %d IOMMUs\n",
-			  IOMMU_UNITS_SUPPORTED);
+			  DMAR_UNITS_SUPPORTED);
 	}
 
+	/* Preallocate enough resources for IOMMU hot-addition */
+	if (g_num_of_iommus < DMAR_UNITS_SUPPORTED)
+		g_num_of_iommus = DMAR_UNITS_SUPPORTED;
+
 	g_iommus = kcalloc(g_num_of_iommus, sizeof(struct intel_iommu *),
 			GFP_KERNEL);
 	if (!g_iommus) {
@@ -2764,58 +2804,14 @@ static int __init init_dmars(void)
 		 * among all IOMMU's. Need to Split it later.
 		 */
 		ret = iommu_alloc_root_entry(iommu);
-		if (ret) {
-			printk(KERN_ERR "IOMMU: allocate root entry failed\n");
+		if (ret)
 			goto free_iommu;
-		}
 		if (!ecap_pass_through(iommu->ecap))
 			hw_pass_through = 0;
 	}
 
-	/*
-	 * Start from the sane iommu hardware state.
-	 */
-	for_each_active_iommu(iommu, drhd) {
-		/*
-		 * If the queued invalidation is already initialized by us
-		 * (for example, while enabling interrupt-remapping) then
-		 * we got the things already rolling from a sane state.
-		 */
-		if (iommu->qi)
-			continue;
-
-		/*
-		 * Clear any previous faults.
-		 */
-		dmar_fault(-1, iommu);
-		/*
-		 * Disable queued invalidation if supported and already enabled
-		 * before OS handover.
-		 */
-		dmar_disable_qi(iommu);
-	}
-
-	for_each_active_iommu(iommu, drhd) {
-		if (dmar_enable_qi(iommu)) {
-			/*
-			 * Queued Invalidate not enabled, use Register Based
-			 * Invalidate
-			 */
-			iommu->flush.flush_context = __iommu_flush_context;
-			iommu->flush.flush_iotlb = __iommu_flush_iotlb;
-			printk(KERN_INFO "IOMMU %d 0x%Lx: using Register based "
-			       "invalidation\n",
-				iommu->seq_id,
-			       (unsigned long long)drhd->reg_base_addr);
-		} else {
-			iommu->flush.flush_context = qi_flush_context;
-			iommu->flush.flush_iotlb = qi_flush_iotlb;
-			printk(KERN_INFO "IOMMU %d 0x%Lx: using Queued "
-			       "invalidation\n",
-				iommu->seq_id,
-			       (unsigned long long)drhd->reg_base_addr);
-		}
-	}
+	for_each_active_iommu(iommu, drhd)
+		intel_iommu_init_qi(iommu);
 
 	if (iommu_pass_through)
 		iommu_identity_mapping |= IDENTMAP_ALL;
@@ -2901,8 +2897,10 @@ static int __init init_dmars(void)
 	return 0;
 
 free_iommu:
-	for_each_active_iommu(iommu, drhd)
+	for_each_active_iommu(iommu, drhd) {
+		disable_dmar_iommu(iommu);
 		free_dmar_iommu(iommu);
+	}
 	kfree(deferred_flush);
 free_g_iommus:
 	kfree(g_iommus);
@@ -3682,7 +3680,7 @@ static inline void init_iommu_pm_ops(void) {}
 #endif	/* CONFIG_PM */
 
 
-int __init dmar_parse_one_rmrr(struct acpi_dmar_header *header)
+int __init dmar_parse_one_rmrr(struct acpi_dmar_header *header, void *arg)
 {
 	struct acpi_dmar_reserved_memory *rmrr;
 	struct dmar_rmrr_unit *rmrru;
@@ -3708,17 +3706,48 @@ int __init dmar_parse_one_rmrr(struct acpi_dmar_header *header)
 	return 0;
 }
 
-int __init dmar_parse_one_atsr(struct acpi_dmar_header *hdr)
+static struct dmar_atsr_unit *dmar_find_atsr(struct acpi_dmar_atsr *atsr)
+{
+	struct dmar_atsr_unit *atsru;
+	struct acpi_dmar_atsr *tmp;
+
+	list_for_each_entry_rcu(atsru, &dmar_atsr_units, list) {
+		tmp = (struct acpi_dmar_atsr *)atsru->hdr;
+		if (atsr->segment != tmp->segment)
+			continue;
+		if (atsr->header.length != tmp->header.length)
+			continue;
+		if (memcmp(atsr, tmp, atsr->header.length) == 0)
+			return atsru;
+	}
+
+	return NULL;
+}
+
+int dmar_parse_one_atsr(struct acpi_dmar_header *hdr, void *arg)
 {
 	struct acpi_dmar_atsr *atsr;
 	struct dmar_atsr_unit *atsru;
 
+	if (system_state != SYSTEM_BOOTING && !intel_iommu_enabled)
+		return 0;
+
 	atsr = container_of(hdr, struct acpi_dmar_atsr, header);
-	atsru = kzalloc(sizeof(*atsru), GFP_KERNEL);
+	atsru = dmar_find_atsr(atsr);
+	if (atsru)
+		return 0;
+
+	atsru = kzalloc(sizeof(*atsru) + hdr->length, GFP_KERNEL);
 	if (!atsru)
 		return -ENOMEM;
 
-	atsru->hdr = hdr;
+	/*
+	 * If memory is allocated from slab by ACPI _DSM method, we need to
+	 * copy the memory content because the memory buffer will be freed
+	 * on return.
+	 */
+	atsru->hdr = (void *)(atsru + 1);
+	memcpy(atsru->hdr, hdr, hdr->length);
 	atsru->include_all = atsr->flags & 0x1;
 	if (!atsru->include_all) {
 		atsru->devices = dmar_alloc_dev_scope((void *)(atsr + 1),
@@ -3741,6 +3770,138 @@ static void intel_iommu_free_atsr(struct dmar_atsr_unit *atsru)
 	kfree(atsru);
 }
 
+int dmar_release_one_atsr(struct acpi_dmar_header *hdr, void *arg)
+{
+	struct acpi_dmar_atsr *atsr;
+	struct dmar_atsr_unit *atsru;
+
+	atsr = container_of(hdr, struct acpi_dmar_atsr, header);
+	atsru = dmar_find_atsr(atsr);
+	if (atsru) {
+		list_del_rcu(&atsru->list);
+		synchronize_rcu();
+		intel_iommu_free_atsr(atsru);
+	}
+
+	return 0;
+}
+
+int dmar_check_one_atsr(struct acpi_dmar_header *hdr, void *arg)
+{
+	int i;
+	struct device *dev;
+	struct acpi_dmar_atsr *atsr;
+	struct dmar_atsr_unit *atsru;
+
+	atsr = container_of(hdr, struct acpi_dmar_atsr, header);
+	atsru = dmar_find_atsr(atsr);
+	if (!atsru)
+		return 0;
+
+	if (!atsru->include_all && atsru->devices && atsru->devices_cnt)
+		for_each_active_dev_scope(atsru->devices, atsru->devices_cnt,
+					  i, dev)
+			return -EBUSY;
+
+	return 0;
+}
+
+static int intel_iommu_add(struct dmar_drhd_unit *dmaru)
+{
+	int sp, ret = 0;
+	struct intel_iommu *iommu = dmaru->iommu;
+
+	if (g_iommus[iommu->seq_id])
+		return 0;
+
+	if (hw_pass_through && !ecap_pass_through(iommu->ecap)) {
+		pr_warn("IOMMU: %s doesn't support hardware pass through.\n",
+			iommu->name);
+		return -ENXIO;
+	}
+	if (!ecap_sc_support(iommu->ecap) &&
+	    domain_update_iommu_snooping(iommu)) {
+		pr_warn("IOMMU: %s doesn't support snooping.\n",
+			iommu->name);
+		return -ENXIO;
+	}
+	sp = domain_update_iommu_superpage(iommu) - 1;
+	if (sp >= 0 && !(cap_super_page_val(iommu->cap) & (1 << sp))) {
+		pr_warn("IOMMU: %s doesn't support large page.\n",
+			iommu->name);
+		return -ENXIO;
+	}
+
+	/*
+	 * Disable translation if already enabled prior to OS handover.
+	 */
+	if (iommu->gcmd & DMA_GCMD_TE)
+		iommu_disable_translation(iommu);
+
+	g_iommus[iommu->seq_id] = iommu;
+	ret = iommu_init_domains(iommu);
+	if (ret == 0)
+		ret = iommu_alloc_root_entry(iommu);
+	if (ret)
+		goto out;
+
+	if (dmaru->ignored) {
+		/*
+		 * we always have to disable PMRs or DMA may fail on this device
+		 */
+		if (force_on)
+			iommu_disable_protect_mem_regions(iommu);
+		return 0;
+	}
+
+	intel_iommu_init_qi(iommu);
+	iommu_flush_write_buffer(iommu);
+	ret = dmar_set_interrupt(iommu);
+	if (ret)
+		goto disable_iommu;
+
+	iommu_set_root_entry(iommu);
+	iommu->flush.flush_context(iommu, 0, 0, 0, DMA_CCMD_GLOBAL_INVL);
+	iommu->flush.flush_iotlb(iommu, 0, 0, 0, DMA_TLB_GLOBAL_FLUSH);
+	iommu_enable_translation(iommu);
+
+	if (si_domain) {
+		ret = iommu_attach_domain(si_domain, iommu);
+		if (ret < 0 || si_domain->id != ret)
+			goto disable_iommu;
+		domain_attach_iommu(si_domain, iommu);
+	}
+
+	iommu_disable_protect_mem_regions(iommu);
+	return 0;
+
+disable_iommu:
+	disable_dmar_iommu(iommu);
+out:
+	free_dmar_iommu(iommu);
+	return ret;
+}
+
+int dmar_iommu_hotplug(struct dmar_drhd_unit *dmaru, bool insert)
+{
+	int ret = 0;
+	struct intel_iommu *iommu = dmaru->iommu;
+
+	if (!intel_iommu_enabled)
+		return 0;
+	if (iommu == NULL)
+		return -EINVAL;
+
+	if (insert) {
+		ret = intel_iommu_add(dmaru);
+	} else {
+		disable_dmar_iommu(iommu);
+		free_dmar_iommu(iommu);
+	}
+
+	return ret;
+}
+
 static void intel_iommu_free_dmars(void)
 {
 	struct dmar_rmrr_unit *rmrru, *rmrr_n;
@@ -3865,16 +4026,7 @@ static int device_notifier(struct notifier_block *nb,
 	if (iommu_dummy(dev))
 		return 0;
 
-	if (action != BUS_NOTIFY_UNBOUND_DRIVER &&
-	    action != BUS_NOTIFY_DEL_DEVICE)
-		return 0;
-
-	/*
-	 * If the device is still attached to a device driver we can't
-	 * tear down the domain yet as DMA mappings may still be in use.
-	 * Wait for the BUS_NOTIFY_UNBOUND_DRIVER event to do that.
-	 */
-	if (action == BUS_NOTIFY_DEL_DEVICE && dev->driver != NULL)
+	if (action != BUS_NOTIFY_REMOVED_DEVICE)
 		return 0;
 
 	domain = find_domain(dev);
@@ -4268,6 +4420,10 @@ static int intel_iommu_attach_device(struct iommu_domain *domain,
 				domain_remove_one_dev_info(old_domain, dev);
 			else
 				domain_remove_dev_info(old_domain);
+
+			if (!domain_type_is_vm_or_si(old_domain) &&
+			     list_empty(&old_domain->devices))
+				domain_exit(old_domain);
 		}
 	}
 
@@ -4415,17 +4571,14 @@ static phys_addr_t intel_iommu_iova_to_phys(struct iommu_domain *domain,
 	return phys;
 }
 
-static int intel_iommu_domain_has_cap(struct iommu_domain *domain,
-				      unsigned long cap)
+static bool intel_iommu_capable(enum iommu_cap cap)
 {
-	struct dmar_domain *dmar_domain = domain->priv;
-
 	if (cap == IOMMU_CAP_CACHE_COHERENCY)
-		return dmar_domain->iommu_snooping;
+		return domain_update_iommu_snooping(NULL) == 1;
 	if (cap == IOMMU_CAP_INTR_REMAP)
-		return irq_remapping_enabled;
+		return irq_remapping_enabled == 1;
 
-	return 0;
+	return false;
 }
 
 static int intel_iommu_add_device(struct device *dev)
@@ -4464,14 +4617,15 @@ static void intel_iommu_remove_device(struct device *dev)
 }
 
 static const struct iommu_ops intel_iommu_ops = {
+	.capable	= intel_iommu_capable,
 	.domain_init	= intel_iommu_domain_init,
 	.domain_destroy = intel_iommu_domain_destroy,
 	.attach_dev	= intel_iommu_attach_device,
 	.detach_dev	= intel_iommu_detach_device,
 	.map		= intel_iommu_map,
 	.unmap		= intel_iommu_unmap,
+	.map_sg		= default_iommu_map_sg,
 	.iova_to_phys	= intel_iommu_iova_to_phys,
-	.domain_has_cap = intel_iommu_domain_has_cap,
 	.add_device	= intel_iommu_add_device,
 	.remove_device	= intel_iommu_remove_device,
 	.pgsize_bitmap	= INTEL_IOMMU_PGSIZES,
diff --git a/drivers/iommu/intel_irq_remapping.c b/drivers/iommu/intel_irq_remapping.c
index 0df41f6264f5..a55b207b9425 100644
--- a/drivers/iommu/intel_irq_remapping.c
+++ b/drivers/iommu/intel_irq_remapping.c
@@ -36,7 +36,6 @@ struct hpet_scope {
 
 static struct ioapic_scope ir_ioapic[MAX_IO_APICS];
 static struct hpet_scope ir_hpet[MAX_HPET_TBS];
-static int ir_ioapic_num, ir_hpet_num;
 
 /*
  * Lock ordering:
@@ -55,7 +54,7 @@ static int __init parse_ioapics_under_ir(void);
 
 static struct irq_2_iommu *irq_2_iommu(unsigned int irq)
 {
-	struct irq_cfg *cfg = irq_get_chip_data(irq);
+	struct irq_cfg *cfg = irq_cfg(irq);
 	return cfg ? &cfg->irq_2_iommu : NULL;
 }
 
@@ -86,7 +85,7 @@ static int alloc_irte(struct intel_iommu *iommu, int irq, u16 count)
 {
 	struct ir_table *table = iommu->ir_table;
 	struct irq_2_iommu *irq_iommu = irq_2_iommu(irq);
-	struct irq_cfg *cfg = irq_get_chip_data(irq);
+	struct irq_cfg *cfg = irq_cfg(irq);
 	unsigned int mask = 0;
 	unsigned long flags;
 	int index;
@@ -154,7 +153,7 @@ static int map_irq_to_irte_handle(int irq, u16 *sub_handle)
 static int set_irte_irq(int irq, struct intel_iommu *iommu, u16 index, u16 subhandle)
 {
 	struct irq_2_iommu *irq_iommu = irq_2_iommu(irq);
-	struct irq_cfg *cfg = irq_get_chip_data(irq);
+	struct irq_cfg *cfg = irq_cfg(irq);
 	unsigned long flags;
 
 	if (!irq_iommu)
@@ -206,7 +205,7 @@ static struct intel_iommu *map_hpet_to_ir(u8 hpet_id)
 	int i;
 
 	for (i = 0; i < MAX_HPET_TBS; i++)
-		if (ir_hpet[i].id == hpet_id)
+		if (ir_hpet[i].id == hpet_id && ir_hpet[i].iommu)
 			return ir_hpet[i].iommu;
 	return NULL;
 }
@@ -216,7 +215,7 @@ static struct intel_iommu *map_ioapic_to_ir(int apic)
 	int i;
 
 	for (i = 0; i < MAX_IO_APICS; i++)
-		if (ir_ioapic[i].id == apic)
+		if (ir_ioapic[i].id == apic && ir_ioapic[i].iommu)
 			return ir_ioapic[i].iommu;
 	return NULL;
 }
@@ -325,7 +324,7 @@ static int set_ioapic_sid(struct irte *irte, int apic)
 
 	down_read(&dmar_global_lock);
 	for (i = 0; i < MAX_IO_APICS; i++) {
-		if (ir_ioapic[i].id == apic) {
+		if (ir_ioapic[i].iommu && ir_ioapic[i].id == apic) {
 			sid = (ir_ioapic[i].bus << 8) | ir_ioapic[i].devfn;
 			break;
 		}
@@ -352,7 +351,7 @@ static int set_hpet_sid(struct irte *irte, u8 id)
 
 	down_read(&dmar_global_lock);
 	for (i = 0; i < MAX_HPET_TBS; i++) {
-		if (ir_hpet[i].id == id) {
+		if (ir_hpet[i].iommu && ir_hpet[i].id == id) {
 			sid = (ir_hpet[i].bus << 8) | ir_hpet[i].devfn;
 			break;
 		}
@@ -438,8 +437,7 @@ static void iommu_set_irq_remapping(struct intel_iommu *iommu, int mode)
 		    (addr) | IR_X2APIC_MODE(mode) | INTR_REMAP_TABLE_REG_SIZE);
 
 	/* Set interrupt-remapping table pointer */
-	iommu->gcmd |= DMA_GCMD_SIRTP;
-	writel(iommu->gcmd, iommu->reg + DMAR_GCMD_REG);
+	writel(iommu->gcmd | DMA_GCMD_SIRTP, iommu->reg + DMAR_GCMD_REG);
 
 	IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG,
 		      readl, (sts & DMA_GSTS_IRTPS), sts);
@@ -474,17 +472,17 @@ static void iommu_set_irq_remapping(struct intel_iommu *iommu, int mode)
 	raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
 }
 
-
-static int intel_setup_irq_remapping(struct intel_iommu *iommu, int mode)
+static int intel_setup_irq_remapping(struct intel_iommu *iommu)
 {
 	struct ir_table *ir_table;
 	struct page *pages;
 	unsigned long *bitmap;
 
-	ir_table = iommu->ir_table = kzalloc(sizeof(struct ir_table),
-					     GFP_ATOMIC);
+	if (iommu->ir_table)
+		return 0;
 
-	if (!iommu->ir_table)
+	ir_table = kzalloc(sizeof(struct ir_table), GFP_ATOMIC);
+	if (!ir_table)
 		return -ENOMEM;
 
 	pages = alloc_pages_node(iommu->node, GFP_ATOMIC | __GFP_ZERO,
@@ -493,24 +491,37 @@ static int intel_setup_irq_remapping(struct intel_iommu *iommu, int mode)
 	if (!pages) {
 		pr_err("IR%d: failed to allocate pages of order %d\n",
 		       iommu->seq_id, INTR_REMAP_PAGE_ORDER);
-		kfree(iommu->ir_table);
-		return -ENOMEM;
+		goto out_free_table;
 	}
 
 	bitmap = kcalloc(BITS_TO_LONGS(INTR_REMAP_TABLE_ENTRIES),
 			 sizeof(long), GFP_ATOMIC);
 	if (bitmap == NULL) {
 		pr_err("IR%d: failed to allocate bitmap\n", iommu->seq_id);
-		__free_pages(pages, INTR_REMAP_PAGE_ORDER);
-		kfree(ir_table);
-		return -ENOMEM;
+		goto out_free_pages;
 	}
 
 	ir_table->base = page_address(pages);
 	ir_table->bitmap = bitmap;
-
-	iommu_set_irq_remapping(iommu, mode);
+	iommu->ir_table = ir_table;
 	return 0;
+
+out_free_pages:
+	__free_pages(pages, INTR_REMAP_PAGE_ORDER);
+out_free_table:
+	kfree(ir_table);
+	return -ENOMEM;
+}
+
+static void intel_teardown_irq_remapping(struct intel_iommu *iommu)
+{
+	if (iommu && iommu->ir_table) {
+		free_pages((unsigned long)iommu->ir_table->base,
+			   INTR_REMAP_PAGE_ORDER);
+		kfree(iommu->ir_table->bitmap);
+		kfree(iommu->ir_table);
+		iommu->ir_table = NULL;
+	}
 }
 
 /*
@@ -667,9 +678,10 @@ static int __init intel_enable_irq_remapping(void)
 		if (!ecap_ir_support(iommu->ecap))
 			continue;
 
-		if (intel_setup_irq_remapping(iommu, eim))
+		if (intel_setup_irq_remapping(iommu))
 			goto error;
 
+		iommu_set_irq_remapping(iommu, eim);
 		setup = 1;
 	}
 
@@ -690,9 +702,11 @@ static int __init intel_enable_irq_remapping(void)
 	return eim ? IRQ_REMAP_X2APIC_MODE : IRQ_REMAP_XAPIC_MODE;
 
 error:
-	/*
-	 * handle error condition gracefully here!
-	 */
+	for_each_iommu(iommu, drhd)
+		if (ecap_ir_support(iommu->ecap)) {
+			iommu_disable_irq_remapping(iommu);
+			intel_teardown_irq_remapping(iommu);
+		}
 
 	if (x2apic_present)
 		pr_warn("Failed to enable irq remapping.  You are vulnerable to irq-injection attacks.\n");
@@ -700,12 +714,13 @@ error:
 	return -1;
 }
 
-static void ir_parse_one_hpet_scope(struct acpi_dmar_device_scope *scope,
-				      struct intel_iommu *iommu)
+static int ir_parse_one_hpet_scope(struct acpi_dmar_device_scope *scope,
+				   struct intel_iommu *iommu,
+				   struct acpi_dmar_hardware_unit *drhd)
 {
 	struct acpi_dmar_pci_path *path;
 	u8 bus;
-	int count;
+	int count, free = -1;
 
 	bus = scope->bus;
 	path = (struct acpi_dmar_pci_path *)(scope + 1);
@@ -721,19 +736,36 @@ static void ir_parse_one_hpet_scope(struct acpi_dmar_device_scope *scope,
 					   PCI_SECONDARY_BUS);
 		path++;
 	}
-	ir_hpet[ir_hpet_num].bus   = bus;
-	ir_hpet[ir_hpet_num].devfn = PCI_DEVFN(path->device, path->function);
-	ir_hpet[ir_hpet_num].iommu = iommu;
-	ir_hpet[ir_hpet_num].id    = scope->enumeration_id;
-	ir_hpet_num++;
+
+	for (count = 0; count < MAX_HPET_TBS; count++) {
+		if (ir_hpet[count].iommu == iommu &&
+		    ir_hpet[count].id == scope->enumeration_id)
+			return 0;
+		else if (ir_hpet[count].iommu == NULL && free == -1)
+			free = count;
+	}
+	if (free == -1) {
+		pr_warn("Exceeded Max HPET blocks\n");
+		return -ENOSPC;
+	}
+
+	ir_hpet[free].iommu = iommu;
+	ir_hpet[free].id    = scope->enumeration_id;
+	ir_hpet[free].bus   = bus;
+	ir_hpet[free].devfn = PCI_DEVFN(path->device, path->function);
+	pr_info("HPET id %d under DRHD base 0x%Lx\n",
+		scope->enumeration_id, drhd->address);
+
+	return 0;
 }
 
-static void ir_parse_one_ioapic_scope(struct acpi_dmar_device_scope *scope,
-				      struct intel_iommu *iommu)
+static int ir_parse_one_ioapic_scope(struct acpi_dmar_device_scope *scope,
+				     struct intel_iommu *iommu,
+				     struct acpi_dmar_hardware_unit *drhd)
 {
 	struct acpi_dmar_pci_path *path;
 	u8 bus;
-	int count;
+	int count, free = -1;
 
 	bus = scope->bus;
 	path = (struct acpi_dmar_pci_path *)(scope + 1);
@@ -750,54 +782,63 @@ static void ir_parse_one_ioapic_scope(struct acpi_dmar_device_scope *scope,
 		path++;
 	}
 
-	ir_ioapic[ir_ioapic_num].bus   = bus;
-	ir_ioapic[ir_ioapic_num].devfn = PCI_DEVFN(path->device, path->function);
-	ir_ioapic[ir_ioapic_num].iommu = iommu;
-	ir_ioapic[ir_ioapic_num].id    = scope->enumeration_id;
-	ir_ioapic_num++;
+	for (count = 0; count < MAX_IO_APICS; count++) {
+		if (ir_ioapic[count].iommu == iommu &&
+		    ir_ioapic[count].id == scope->enumeration_id)
+			return 0;
+		else if (ir_ioapic[count].iommu == NULL && free == -1)
+			free = count;
+	}
+	if (free == -1) {
+		pr_warn("Exceeded Max IO APICS\n");
+		return -ENOSPC;
+	}
+
+	ir_ioapic[free].bus   = bus;
+	ir_ioapic[free].devfn = PCI_DEVFN(path->device, path->function);
+	ir_ioapic[free].iommu = iommu;
+	ir_ioapic[free].id    = scope->enumeration_id;
+	pr_info("IOAPIC id %d under DRHD base  0x%Lx IOMMU %d\n",
+		scope->enumeration_id, drhd->address, iommu->seq_id);
+
+	return 0;
 }
 
 static int ir_parse_ioapic_hpet_scope(struct acpi_dmar_header *header,
 				      struct intel_iommu *iommu)
 {
+	int ret = 0;
 	struct acpi_dmar_hardware_unit *drhd;
 	struct acpi_dmar_device_scope *scope;
 	void *start, *end;
 
 	drhd = (struct acpi_dmar_hardware_unit *)header;
-
 	start = (void *)(drhd + 1);
 	end = ((void *)drhd) + header->length;
 
-	while (start < end) {
+	while (start < end && ret == 0) {
 		scope = start;
-		if (scope->entry_type == ACPI_DMAR_SCOPE_TYPE_IOAPIC) {
-			if (ir_ioapic_num == MAX_IO_APICS) {
-				printk(KERN_WARNING "Exceeded Max IO APICS\n");
-				return -1;
-			}
-
-			printk(KERN_INFO "IOAPIC id %d under DRHD base "
-			       " 0x%Lx IOMMU %d\n", scope->enumeration_id,
-			       drhd->address, iommu->seq_id);
+		if (scope->entry_type == ACPI_DMAR_SCOPE_TYPE_IOAPIC)
+			ret = ir_parse_one_ioapic_scope(scope, iommu, drhd);
+		else if (scope->entry_type == ACPI_DMAR_SCOPE_TYPE_HPET)
+			ret = ir_parse_one_hpet_scope(scope, iommu, drhd);
+		start += scope->length;
+	}
 
-			ir_parse_one_ioapic_scope(scope, iommu);
-		} else if (scope->entry_type == ACPI_DMAR_SCOPE_TYPE_HPET) {
-			if (ir_hpet_num == MAX_HPET_TBS) {
-				printk(KERN_WARNING "Exceeded Max HPET blocks\n");
-				return -1;
-			}
+	return ret;
+}
 
-			printk(KERN_INFO "HPET id %d under DRHD base"
-			       " 0x%Lx\n", scope->enumeration_id,
-			       drhd->address);
+static void ir_remove_ioapic_hpet_scope(struct intel_iommu *iommu)
+{
+	int i;
 
-			ir_parse_one_hpet_scope(scope, iommu);
-		}
-		start += scope->length;
-	}
+	for (i = 0; i < MAX_HPET_TBS; i++)
+		if (ir_hpet[i].iommu == iommu)
+			ir_hpet[i].iommu = NULL;
 
-	return 0;
+	for (i = 0; i < MAX_IO_APICS; i++)
+		if (ir_ioapic[i].iommu == iommu)
+			ir_ioapic[i].iommu = NULL;
 }
 
 /*
@@ -1009,7 +1050,7 @@ static int
 intel_ioapic_set_affinity(struct irq_data *data, const struct cpumask *mask,
 			  bool force)
 {
-	struct irq_cfg *cfg = data->chip_data;
+	struct irq_cfg *cfg = irqd_cfg(data);
 	unsigned int dest, irq = data->irq;
 	struct irte irte;
 	int err;
@@ -1064,7 +1105,7 @@ static void intel_compose_msi_msg(struct pci_dev *pdev,
 	u16 sub_handle = 0;
 	int ir_index;
 
-	cfg = irq_get_chip_data(irq);
+	cfg = irq_cfg(irq);
 
 	ir_index = map_irq_to_irte_handle(irq, &sub_handle);
 	BUG_ON(ir_index == -1);
@@ -1139,7 +1180,7 @@ static int intel_msi_setup_irq(struct pci_dev *pdev, unsigned int irq,
 	return ret;
 }
 
-static int intel_setup_hpet_msi(unsigned int irq, unsigned int id)
+static int intel_alloc_hpet_msi(unsigned int irq, unsigned int id)
 {
 	int ret = -1;
 	struct intel_iommu *iommu;
@@ -1170,5 +1211,88 @@ struct irq_remap_ops intel_irq_remap_ops = {
 	.compose_msi_msg	= intel_compose_msi_msg,
 	.msi_alloc_irq		= intel_msi_alloc_irq,
 	.msi_setup_irq		= intel_msi_setup_irq,
-	.setup_hpet_msi		= intel_setup_hpet_msi,
+	.alloc_hpet_msi		= intel_alloc_hpet_msi,
 };
+
+/*
+ * Support of Interrupt Remapping Unit Hotplug
+ */
+static int dmar_ir_add(struct dmar_drhd_unit *dmaru, struct intel_iommu *iommu)
+{
+	int ret;
+	int eim = x2apic_enabled();
+
+	if (eim && !ecap_eim_support(iommu->ecap)) {
+		pr_info("DRHD %Lx: EIM not supported by DRHD, ecap %Lx\n",
+			iommu->reg_phys, iommu->ecap);
+		return -ENODEV;
+	}
+
+	if (ir_parse_ioapic_hpet_scope(dmaru->hdr, iommu)) {
+		pr_warn("DRHD %Lx: failed to parse managed IOAPIC/HPET\n",
+			iommu->reg_phys);
+		return -ENODEV;
+	}
+
+	/* TODO: check all IOAPICs are covered by IOMMU */
+
+	/* Setup Interrupt-remapping now. */
+	ret = intel_setup_irq_remapping(iommu);
+	if (ret) {
+		pr_err("DRHD %Lx: failed to allocate resource\n",
+		       iommu->reg_phys);
+		ir_remove_ioapic_hpet_scope(iommu);
+		return ret;
+	}
+
+	if (!iommu->qi) {
+		/* Clear previous faults. */
+		dmar_fault(-1, iommu);
+		iommu_disable_irq_remapping(iommu);
+		dmar_disable_qi(iommu);
+	}
+
+	/* Enable queued invalidation */
+	ret = dmar_enable_qi(iommu);
+	if (!ret) {
+		iommu_set_irq_remapping(iommu, eim);
+	} else {
+		pr_err("DRHD %Lx: failed to enable queued invalidation, ecap %Lx, ret %d\n",
+		       iommu->reg_phys, iommu->ecap, ret);
+		intel_teardown_irq_remapping(iommu);
+		ir_remove_ioapic_hpet_scope(iommu);
+	}
+
+	return ret;
+}
+
+int dmar_ir_hotplug(struct dmar_drhd_unit *dmaru, bool insert)
+{
+	int ret = 0;
+	struct intel_iommu *iommu = dmaru->iommu;
+
+	if (!irq_remapping_enabled)
+		return 0;
+	if (iommu == NULL)
+		return -EINVAL;
+	if (!ecap_ir_support(iommu->ecap))
+		return 0;
+
+	if (insert) {
+		if (!iommu->ir_table)
+			ret = dmar_ir_add(dmaru, iommu);
+	} else {
+		if (iommu->ir_table) {
+			if (!bitmap_empty(iommu->ir_table->bitmap,
+					  INTR_REMAP_TABLE_ENTRIES)) {
+				ret = -EBUSY;
+			} else {
+				iommu_disable_irq_remapping(iommu);
+				intel_teardown_irq_remapping(iommu);
+				ir_remove_ioapic_hpet_scope(iommu);
+			}
+		}
+	}
+
+	return ret;
+}
diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c
index 0639b9274b11..f7718d73e984 100644
--- a/drivers/iommu/iommu.c
+++ b/drivers/iommu/iommu.c
@@ -30,6 +30,7 @@
 #include <linux/notifier.h>
 #include <linux/err.h>
 #include <linux/pci.h>
+#include <linux/bitops.h>
 #include <trace/events/iommu.h>
 
 static struct kset *iommu_group_kset;
@@ -519,6 +520,9 @@ int iommu_group_id(struct iommu_group *group)
 }
 EXPORT_SYMBOL_GPL(iommu_group_id);
 
+static struct iommu_group *get_pci_alias_group(struct pci_dev *pdev,
+					       unsigned long *devfns);
+
 /*
  * To consider a PCI device isolated, we require ACS to support Source
  * Validation, Request Redirection, Completer Redirection, and Upstream
@@ -529,6 +533,86 @@ EXPORT_SYMBOL_GPL(iommu_group_id);
  */
 #define REQ_ACS_FLAGS   (PCI_ACS_SV | PCI_ACS_RR | PCI_ACS_CR | PCI_ACS_UF)
 
+/*
+ * For multifunction devices which are not isolated from each other, find
+ * all the other non-isolated functions and look for existing groups.  For
+ * each function, we also need to look for aliases to or from other devices
+ * that may already have a group.
+ */
+static struct iommu_group *get_pci_function_alias_group(struct pci_dev *pdev,
+							unsigned long *devfns)
+{
+	struct pci_dev *tmp = NULL;
+	struct iommu_group *group;
+
+	if (!pdev->multifunction || pci_acs_enabled(pdev, REQ_ACS_FLAGS))
+		return NULL;
+
+	for_each_pci_dev(tmp) {
+		if (tmp == pdev || tmp->bus != pdev->bus ||
+		    PCI_SLOT(tmp->devfn) != PCI_SLOT(pdev->devfn) ||
+		    pci_acs_enabled(tmp, REQ_ACS_FLAGS))
+			continue;
+
+		group = get_pci_alias_group(tmp, devfns);
+		if (group) {
+			pci_dev_put(tmp);
+			return group;
+		}
+	}
+
+	return NULL;
+}
+
+/*
+ * Look for aliases to or from the given device for exisiting groups.  The
+ * dma_alias_devfn only supports aliases on the same bus, therefore the search
+ * space is quite small (especially since we're really only looking at pcie
+ * device, and therefore only expect multiple slots on the root complex or
+ * downstream switch ports).  It's conceivable though that a pair of
+ * multifunction devices could have aliases between them that would cause a
+ * loop.  To prevent this, we use a bitmap to track where we've been.
+ */
+static struct iommu_group *get_pci_alias_group(struct pci_dev *pdev,
+					       unsigned long *devfns)
+{
+	struct pci_dev *tmp = NULL;
+	struct iommu_group *group;
+
+	if (test_and_set_bit(pdev->devfn & 0xff, devfns))
+		return NULL;
+
+	group = iommu_group_get(&pdev->dev);
+	if (group)
+		return group;
+
+	for_each_pci_dev(tmp) {
+		if (tmp == pdev || tmp->bus != pdev->bus)
+			continue;
+
+		/* We alias them or they alias us */
+		if (((pdev->dev_flags & PCI_DEV_FLAGS_DMA_ALIAS_DEVFN) &&
+		     pdev->dma_alias_devfn == tmp->devfn) ||
+		    ((tmp->dev_flags & PCI_DEV_FLAGS_DMA_ALIAS_DEVFN) &&
+		     tmp->dma_alias_devfn == pdev->devfn)) {
+
+			group = get_pci_alias_group(tmp, devfns);
+			if (group) {
+				pci_dev_put(tmp);
+				return group;
+			}
+
+			group = get_pci_function_alias_group(tmp, devfns);
+			if (group) {
+				pci_dev_put(tmp);
+				return group;
+			}
+		}
+	}
+
+	return NULL;
+}
+
 struct group_for_pci_data {
 	struct pci_dev *pdev;
 	struct iommu_group *group;
@@ -557,7 +641,7 @@ static struct iommu_group *iommu_group_get_for_pci_dev(struct pci_dev *pdev)
 	struct group_for_pci_data data;
 	struct pci_bus *bus;
 	struct iommu_group *group = NULL;
-	struct pci_dev *tmp;
+	u64 devfns[4] = { 0 };
 
 	/*
 	 * Find the upstream DMA alias for the device.  A device must not
@@ -591,76 +675,21 @@ static struct iommu_group *iommu_group_get_for_pci_dev(struct pci_dev *pdev)
 	}
 
 	/*
-	 * Next we need to consider DMA alias quirks.  If one device aliases
-	 * to another, they should be grouped together.  It's theoretically
-	 * possible that aliases could create chains of devices where each
-	 * device aliases another device.  If we then factor in multifunction
-	 * ACS grouping requirements, each alias could incorporate a new slot
-	 * with multiple functions, each with aliases.  This is all extremely
-	 * unlikely as DMA alias quirks are typically only used for PCIe
-	 * devices where we usually have a single slot per bus.  Furthermore,
-	 * the alias quirk is usually to another function within the slot
-	 * (and ACS multifunction is not supported) or to a different slot
-	 * that doesn't physically exist.  The likely scenario is therefore
-	 * that everything on the bus gets grouped together.  To reduce the
-	 * problem space, share the IOMMU group for all devices on the bus
-	 * if a DMA alias quirk is present on the bus.
+	 * Look for existing groups on device aliases.  If we alias another
+	 * device or another device aliases us, use the same group.
 	 */
-	tmp = NULL;
-	for_each_pci_dev(tmp) {
-		if (tmp->bus != pdev->bus ||
-		    !(tmp->dev_flags & PCI_DEV_FLAGS_DMA_ALIAS_DEVFN))
-			continue;
-
-		pci_dev_put(tmp);
-		tmp = NULL;
-
-		/* We have an alias quirk, search for an existing group */
-		for_each_pci_dev(tmp) {
-			struct iommu_group *group_tmp;
-
-			if (tmp->bus != pdev->bus)
-				continue;
-
-			group_tmp = iommu_group_get(&tmp->dev);
-			if (!group) {
-				group = group_tmp;
-				continue;
-			}
-
-			if (group_tmp) {
-				WARN_ON(group != group_tmp);
-				iommu_group_put(group_tmp);
-			}
-		}
-
-		return group ? group : iommu_group_alloc();
-	}
-
-	/*
-	 * Non-multifunction devices or multifunction devices supporting
-	 * ACS get their own group.
-	 */
-	if (!pdev->multifunction || pci_acs_enabled(pdev, REQ_ACS_FLAGS))
-		return iommu_group_alloc();
+	group = get_pci_alias_group(pdev, (unsigned long *)devfns);
+	if (group)
+		return group;
 
 	/*
-	 * Multifunction devices not supporting ACS share a group with other
-	 * similar devices in the same slot.
+	 * Look for existing groups on non-isolated functions on the same
+	 * slot and aliases of those funcions, if any.  No need to clear
+	 * the search bitmap, the tested devfns are still valid.
 	 */
-	tmp = NULL;
-	for_each_pci_dev(tmp) {
-		if (tmp == pdev || tmp->bus != pdev->bus ||
-		    PCI_SLOT(tmp->devfn) !=  PCI_SLOT(pdev->devfn) ||
-		    pci_acs_enabled(tmp, REQ_ACS_FLAGS))
-			continue;
-
-		group = iommu_group_get(&tmp->dev);
-		if (group) {
-			pci_dev_put(tmp);
-			return group;
-		}
-	}
+	group = get_pci_function_alias_group(pdev, (unsigned long *)devfns);
+	if (group)
+		return group;
 
 	/* No shared group found, allocate new */
 	return iommu_group_alloc();
@@ -708,7 +737,7 @@ static int add_iommu_group(struct device *dev, void *data)
 	const struct iommu_ops *ops = cb->ops;
 
 	if (!ops->add_device)
-		return -ENODEV;
+		return 0;
 
 	WARN_ON(dev->iommu_group);
 
@@ -770,18 +799,34 @@ static int iommu_bus_notifier(struct notifier_block *nb,
 	return 0;
 }
 
-static struct notifier_block iommu_bus_nb = {
-	.notifier_call = iommu_bus_notifier,
-};
-
-static void iommu_bus_init(struct bus_type *bus, const struct iommu_ops *ops)
+static int iommu_bus_init(struct bus_type *bus, const struct iommu_ops *ops)
 {
+	int err;
+	struct notifier_block *nb;
 	struct iommu_callback_data cb = {
 		.ops = ops,
 	};
 
-	bus_register_notifier(bus, &iommu_bus_nb);
-	bus_for_each_dev(bus, NULL, &cb, add_iommu_group);
+	nb = kzalloc(sizeof(struct notifier_block), GFP_KERNEL);
+	if (!nb)
+		return -ENOMEM;
+
+	nb->notifier_call = iommu_bus_notifier;
+
+	err = bus_register_notifier(bus, nb);
+	if (err) {
+		kfree(nb);
+		return err;
+	}
+
+	err = bus_for_each_dev(bus, NULL, &cb, add_iommu_group);
+	if (err) {
+		bus_unregister_notifier(bus, nb);
+		kfree(nb);
+		return err;
+	}
+
+	return 0;
 }
 
 /**
@@ -799,15 +844,19 @@ static void iommu_bus_init(struct bus_type *bus, const struct iommu_ops *ops)
  */
 int bus_set_iommu(struct bus_type *bus, const struct iommu_ops *ops)
 {
+	int err;
+
 	if (bus->iommu_ops != NULL)
 		return -EBUSY;
 
 	bus->iommu_ops = ops;
 
 	/* Do IOMMU specific setup for this bus-type */
-	iommu_bus_init(bus, ops);
+	err = iommu_bus_init(bus, ops);
+	if (err)
+		bus->iommu_ops = NULL;
 
-	return 0;
+	return err;
 }
 EXPORT_SYMBOL_GPL(bus_set_iommu);
 
@@ -817,6 +866,15 @@ bool iommu_present(struct bus_type *bus)
 }
 EXPORT_SYMBOL_GPL(iommu_present);
 
+bool iommu_capable(struct bus_type *bus, enum iommu_cap cap)
+{
+	if (!bus->iommu_ops || !bus->iommu_ops->capable)
+		return false;
+
+	return bus->iommu_ops->capable(cap);
+}
+EXPORT_SYMBOL_GPL(iommu_capable);
+
 /**
  * iommu_set_fault_handler() - set a fault handler for an iommu domain
  * @domain: iommu domain
@@ -947,16 +1005,6 @@ phys_addr_t iommu_iova_to_phys(struct iommu_domain *domain, dma_addr_t iova)
 }
 EXPORT_SYMBOL_GPL(iommu_iova_to_phys);
 
-int iommu_domain_has_cap(struct iommu_domain *domain,
-			 unsigned long cap)
-{
-	if (unlikely(domain->ops->domain_has_cap == NULL))
-		return 0;
-
-	return domain->ops->domain_has_cap(domain, cap);
-}
-EXPORT_SYMBOL_GPL(iommu_domain_has_cap);
-
 static size_t iommu_pgsize(struct iommu_domain *domain,
 			   unsigned long addr_merge, size_t size)
 {
@@ -1090,6 +1138,48 @@ size_t iommu_unmap(struct iommu_domain *domain, unsigned long iova, size_t size)
 }
 EXPORT_SYMBOL_GPL(iommu_unmap);
 
+size_t default_iommu_map_sg(struct iommu_domain *domain, unsigned long iova,
+			 struct scatterlist *sg, unsigned int nents, int prot)
+{
+	struct scatterlist *s;
+	size_t mapped = 0;
+	unsigned int i, min_pagesz;
+	int ret;
+
+	if (unlikely(domain->ops->pgsize_bitmap == 0UL))
+		return 0;
+
+	min_pagesz = 1 << __ffs(domain->ops->pgsize_bitmap);
+
+	for_each_sg(sg, s, nents, i) {
+		phys_addr_t phys = page_to_phys(sg_page(s)) + s->offset;
+
+		/*
+		 * We are mapping on IOMMU page boundaries, so offset within
+		 * the page must be 0. However, the IOMMU may support pages
+		 * smaller than PAGE_SIZE, so s->offset may still represent
+		 * an offset of that boundary within the CPU page.
+		 */
+		if (!IS_ALIGNED(s->offset, min_pagesz))
+			goto out_err;
+
+		ret = iommu_map(domain, iova + mapped, phys, s->length, prot);
+		if (ret)
+			goto out_err;
+
+		mapped += s->length;
+	}
+
+	return mapped;
+
+out_err:
+	/* undo mappings already done */
+	iommu_unmap(domain, iova, mapped);
+
+	return 0;
+
+}
+EXPORT_SYMBOL_GPL(default_iommu_map_sg);
 
 int iommu_domain_window_enable(struct iommu_domain *domain, u32 wnd_nr,
 			       phys_addr_t paddr, u64 size, int prot)
diff --git a/drivers/iommu/ipmmu-vmsa.c b/drivers/iommu/ipmmu-vmsa.c
index 7dab5cbcc775..748693192c20 100644
--- a/drivers/iommu/ipmmu-vmsa.c
+++ b/drivers/iommu/ipmmu-vmsa.c
@@ -558,7 +558,7 @@ static pmd_t *ipmmu_alloc_pmd(struct ipmmu_vmsa_device *mmu, pgd_t *pgd,
 
 static u64 ipmmu_page_prot(unsigned int prot, u64 type)
 {
-	u64 pgprot = ARM_VMSA_PTE_XN | ARM_VMSA_PTE_nG | ARM_VMSA_PTE_AF
+	u64 pgprot = ARM_VMSA_PTE_nG | ARM_VMSA_PTE_AF
 		   | ARM_VMSA_PTE_SH_IS | ARM_VMSA_PTE_AP_UNPRIV
 		   | ARM_VMSA_PTE_NS | type;
 
@@ -568,8 +568,8 @@ static u64 ipmmu_page_prot(unsigned int prot, u64 type)
 	if (prot & IOMMU_CACHE)
 		pgprot |= IMMAIR_ATTR_IDX_WBRWA << ARM_VMSA_PTE_ATTRINDX_SHIFT;
 
-	if (prot & IOMMU_EXEC)
-		pgprot &= ~ARM_VMSA_PTE_XN;
+	if (prot & IOMMU_NOEXEC)
+		pgprot |= ARM_VMSA_PTE_XN;
 	else if (!(prot & (IOMMU_READ | IOMMU_WRITE)))
 		/* If no access create a faulting entry to avoid TLB fills. */
 		pgprot &= ~ARM_VMSA_PTE_PAGE;
@@ -1127,6 +1127,7 @@ static const struct iommu_ops ipmmu_ops = {
 	.detach_dev = ipmmu_detach_device,
 	.map = ipmmu_map,
 	.unmap = ipmmu_unmap,
+	.map_sg = default_iommu_map_sg,
 	.iova_to_phys = ipmmu_iova_to_phys,
 	.add_device = ipmmu_add_device,
 	.remove_device = ipmmu_remove_device,
@@ -1184,7 +1185,7 @@ static int ipmmu_probe(struct platform_device *pdev)
 			       dev_name(&pdev->dev), mmu);
 	if (ret < 0) {
 		dev_err(&pdev->dev, "failed to request IRQ %d\n", irq);
-		return irq;
+		return ret;
 	}
 
 	ipmmu_device_reset(mmu);
@@ -1221,7 +1222,6 @@ static int ipmmu_remove(struct platform_device *pdev)
 
 static struct platform_driver ipmmu_driver = {
 	.driver = {
-		.owner = THIS_MODULE,
 		.name = "ipmmu-vmsa",
 	},
 	.probe = ipmmu_probe,
diff --git a/drivers/iommu/irq_remapping.c b/drivers/iommu/irq_remapping.c
index 33c439524080..89c4846683be 100644
--- a/drivers/iommu/irq_remapping.c
+++ b/drivers/iommu/irq_remapping.c
@@ -12,6 +12,7 @@
 #include <asm/processor.h>
 #include <asm/x86_init.h>
 #include <asm/apic.h>
+#include <asm/hpet.h>
 
 #include "irq_remapping.h"
 
@@ -55,19 +56,13 @@ static int do_setup_msi_irqs(struct pci_dev *dev, int nvec)
 	unsigned int irq;
 	struct msi_desc *msidesc;
 
-	WARN_ON(!list_is_singular(&dev->msi_list));
 	msidesc = list_entry(dev->msi_list.next, struct msi_desc, list);
-	WARN_ON(msidesc->irq);
-	WARN_ON(msidesc->msi_attrib.multiple);
-	WARN_ON(msidesc->nvec_used);
 
 	irq = irq_alloc_hwirqs(nvec, dev_to_node(&dev->dev));
 	if (irq == 0)
 		return -ENOSPC;
 
 	nvec_pow2 = __roundup_pow_of_two(nvec);
-	msidesc->nvec_used = nvec;
-	msidesc->msi_attrib.multiple = ilog2(nvec_pow2);
 	for (sub_handle = 0; sub_handle < nvec; sub_handle++) {
 		if (!sub_handle) {
 			index = msi_alloc_remapped_irq(dev, irq, nvec_pow2);
@@ -95,8 +90,6 @@ error:
 	 * IRQs from tearing down again in default_teardown_msi_irqs()
 	 */
 	msidesc->irq = 0;
-	msidesc->nvec_used = 0;
-	msidesc->msi_attrib.multiple = 0;
 
 	return ret;
 }
@@ -305,7 +298,7 @@ static int set_remapped_irq_affinity(struct irq_data *data,
 
 void free_remapped_irq(int irq)
 {
-	struct irq_cfg *cfg = irq_get_chip_data(irq);
+	struct irq_cfg *cfg = irq_cfg(irq);
 
 	if (!remap_ops || !remap_ops->free_irq)
 		return;
@@ -318,7 +311,7 @@ void compose_remapped_msi_msg(struct pci_dev *pdev,
 			      unsigned int irq, unsigned int dest,
 			      struct msi_msg *msg, u8 hpet_id)
 {
-	struct irq_cfg *cfg = irq_get_chip_data(irq);
+	struct irq_cfg *cfg = irq_cfg(irq);
 
 	if (!irq_remapped(cfg))
 		native_compose_msi_msg(pdev, irq, dest, msg, hpet_id);
@@ -345,10 +338,16 @@ static int msi_setup_remapped_irq(struct pci_dev *pdev, unsigned int irq,
 
 int setup_hpet_msi_remapped(unsigned int irq, unsigned int id)
 {
-	if (!remap_ops || !remap_ops->setup_hpet_msi)
+	int ret;
+
+	if (!remap_ops || !remap_ops->alloc_hpet_msi)
 		return -ENODEV;
 
-	return remap_ops->setup_hpet_msi(irq, id);
+	ret = remap_ops->alloc_hpet_msi(irq, id);
+	if (ret)
+		return -EINVAL;
+
+	return default_setup_hpet_msi(irq, id);
 }
 
 void panic_if_irq_remap(const char *msg)
@@ -365,7 +364,7 @@ static void ir_ack_apic_edge(struct irq_data *data)
 static void ir_ack_apic_level(struct irq_data *data)
 {
 	ack_APIC_irq();
-	eoi_ioapic_irq(data->irq, data->chip_data);
+	eoi_ioapic_irq(data->irq, irqd_cfg(data));
 }
 
 static void ir_print_prefix(struct irq_data *data, struct seq_file *p)
diff --git a/drivers/iommu/irq_remapping.h b/drivers/iommu/irq_remapping.h
index 90c4dae5a46b..fde250f86e60 100644
--- a/drivers/iommu/irq_remapping.h
+++ b/drivers/iommu/irq_remapping.h
@@ -80,7 +80,7 @@ struct irq_remap_ops {
 	int (*msi_setup_irq)(struct pci_dev *, unsigned int, int, int);
 
 	/* Setup interrupt remapping for an HPET MSI */
-	int (*setup_hpet_msi)(unsigned int, unsigned int);
+	int (*alloc_hpet_msi)(unsigned int, unsigned int);
 };
 
 extern struct irq_remap_ops intel_irq_remap_ops;
diff --git a/drivers/iommu/msm_iommu.c b/drivers/iommu/msm_iommu.c
index 49f41d6e02f1..e1b05379ca0e 100644
--- a/drivers/iommu/msm_iommu.c
+++ b/drivers/iommu/msm_iommu.c
@@ -73,8 +73,7 @@ fail:
 
 static void __disable_clocks(struct msm_iommu_drvdata *drvdata)
 {
-	if (drvdata->clk)
-		clk_disable(drvdata->clk);
+	clk_disable(drvdata->clk);
 	clk_disable(drvdata->pclk);
 }
 
@@ -603,10 +602,9 @@ fail:
 	return ret;
 }
 
-static int msm_iommu_domain_has_cap(struct iommu_domain *domain,
-				    unsigned long cap)
+static bool msm_iommu_capable(enum iommu_cap cap)
 {
-	return 0;
+	return false;
 }
 
 static void print_ctx_regs(void __iomem *base, int ctx)
@@ -675,14 +673,15 @@ fail:
 }
 
 static const struct iommu_ops msm_iommu_ops = {
+	.capable = msm_iommu_capable,
 	.domain_init = msm_iommu_domain_init,
 	.domain_destroy = msm_iommu_domain_destroy,
 	.attach_dev = msm_iommu_attach_dev,
 	.detach_dev = msm_iommu_detach_dev,
 	.map = msm_iommu_map,
 	.unmap = msm_iommu_unmap,
+	.map_sg = default_iommu_map_sg,
 	.iova_to_phys = msm_iommu_iova_to_phys,
-	.domain_has_cap = msm_iommu_domain_has_cap,
 	.pgsize_bitmap = MSM_IOMMU_PGSIZES,
 };
 
diff --git a/drivers/iommu/msm_iommu_dev.c b/drivers/iommu/msm_iommu_dev.c
index 61def7cb5263..b6d01f97e537 100644
--- a/drivers/iommu/msm_iommu_dev.c
+++ b/drivers/iommu/msm_iommu_dev.c
@@ -131,7 +131,7 @@ static int msm_iommu_probe(struct platform_device *pdev)
 	struct clk *iommu_clk;
 	struct clk *iommu_pclk;
 	struct msm_iommu_drvdata *drvdata;
-	struct msm_iommu_dev *iommu_dev = pdev->dev.platform_data;
+	struct msm_iommu_dev *iommu_dev = dev_get_platdata(&pdev->dev);
 	void __iomem *regs_base;
 	int ret, irq, par;
 
@@ -224,8 +224,7 @@ static int msm_iommu_probe(struct platform_device *pdev)
 
 	platform_set_drvdata(pdev, drvdata);
 
-	if (iommu_clk)
-		clk_disable(iommu_clk);
+	clk_disable(iommu_clk);
 
 	clk_disable(iommu_pclk);
 
@@ -264,7 +263,7 @@ static int msm_iommu_remove(struct platform_device *pdev)
 
 static int msm_iommu_ctx_probe(struct platform_device *pdev)
 {
-	struct msm_iommu_ctx_dev *c = pdev->dev.platform_data;
+	struct msm_iommu_ctx_dev *c = dev_get_platdata(&pdev->dev);
 	struct msm_iommu_drvdata *drvdata;
 	struct msm_iommu_ctx_drvdata *ctx_drvdata;
 	int i, ret;
@@ -323,8 +322,7 @@ static int msm_iommu_ctx_probe(struct platform_device *pdev)
 		SET_NSCFG(drvdata->base, mid, 3);
 	}
 
-	if (drvdata->clk)
-		clk_disable(drvdata->clk);
+	clk_disable(drvdata->clk);
 	clk_disable(drvdata->pclk);
 
 	dev_info(&pdev->dev, "context %s using bank %d\n", c->name, c->num);
diff --git a/drivers/iommu/of_iommu.c b/drivers/iommu/of_iommu.c
index e550ccb7634e..af1dc6a1c0a1 100644
--- a/drivers/iommu/of_iommu.c
+++ b/drivers/iommu/of_iommu.c
@@ -18,9 +18,14 @@
  */
 
 #include <linux/export.h>
+#include <linux/iommu.h>
 #include <linux/limits.h>
 #include <linux/of.h>
 #include <linux/of_iommu.h>
+#include <linux/slab.h>
+
+static const struct of_device_id __iommu_of_table_sentinel
+	__used __section(__iommu_of_table_end);
 
 /**
  * of_get_dma_window - Parse *dma-window property and returns 0 if found.
@@ -89,3 +94,87 @@ int of_get_dma_window(struct device_node *dn, const char *prefix, int index,
 	return 0;
 }
 EXPORT_SYMBOL_GPL(of_get_dma_window);
+
+struct of_iommu_node {
+	struct list_head list;
+	struct device_node *np;
+	struct iommu_ops *ops;
+};
+static LIST_HEAD(of_iommu_list);
+static DEFINE_SPINLOCK(of_iommu_lock);
+
+void of_iommu_set_ops(struct device_node *np, struct iommu_ops *ops)
+{
+	struct of_iommu_node *iommu = kzalloc(sizeof(*iommu), GFP_KERNEL);
+
+	if (WARN_ON(!iommu))
+		return;
+
+	INIT_LIST_HEAD(&iommu->list);
+	iommu->np = np;
+	iommu->ops = ops;
+	spin_lock(&of_iommu_lock);
+	list_add_tail(&iommu->list, &of_iommu_list);
+	spin_unlock(&of_iommu_lock);
+}
+
+struct iommu_ops *of_iommu_get_ops(struct device_node *np)
+{
+	struct of_iommu_node *node;
+	struct iommu_ops *ops = NULL;
+
+	spin_lock(&of_iommu_lock);
+	list_for_each_entry(node, &of_iommu_list, list)
+		if (node->np == np) {
+			ops = node->ops;
+			break;
+		}
+	spin_unlock(&of_iommu_lock);
+	return ops;
+}
+
+struct iommu_ops *of_iommu_configure(struct device *dev)
+{
+	struct of_phandle_args iommu_spec;
+	struct device_node *np;
+	struct iommu_ops *ops = NULL;
+	int idx = 0;
+
+	/*
+	 * We don't currently walk up the tree looking for a parent IOMMU.
+	 * See the `Notes:' section of
+	 * Documentation/devicetree/bindings/iommu/iommu.txt
+	 */
+	while (!of_parse_phandle_with_args(dev->of_node, "iommus",
+					   "#iommu-cells", idx,
+					   &iommu_spec)) {
+		np = iommu_spec.np;
+		ops = of_iommu_get_ops(np);
+
+		if (!ops || !ops->of_xlate || ops->of_xlate(dev, &iommu_spec))
+			goto err_put_node;
+
+		of_node_put(np);
+		idx++;
+	}
+
+	return ops;
+
+err_put_node:
+	of_node_put(np);
+	return NULL;
+}
+
+void __init of_iommu_init(void)
+{
+	struct device_node *np;
+	const struct of_device_id *match, *matches = &__iommu_of_table;
+
+	for_each_matching_node_and_match(np, matches, &match) {
+		const of_iommu_init_fn init_fn = match->data;
+
+		if (init_fn(np))
+			pr_err("Failed to initialise IOMMU %s\n",
+				of_node_full_name(np));
+	}
+}
diff --git a/drivers/iommu/omap-iommu-debug.c b/drivers/iommu/omap-iommu-debug.c
index 531658d17333..f3d20a2039d2 100644
--- a/drivers/iommu/omap-iommu-debug.c
+++ b/drivers/iommu/omap-iommu-debug.c
@@ -10,45 +10,35 @@
  * published by the Free Software Foundation.
  */
 
-#include <linux/module.h>
 #include <linux/err.h>
-#include <linux/clk.h>
 #include <linux/io.h>
 #include <linux/slab.h>
 #include <linux/uaccess.h>
-#include <linux/platform_device.h>
 #include <linux/debugfs.h>
-#include <linux/omap-iommu.h>
 #include <linux/platform_data/iommu-omap.h>
 
 #include "omap-iopgtable.h"
 #include "omap-iommu.h"
 
-#define MAXCOLUMN 100 /* for short messages */
-
 static DEFINE_MUTEX(iommu_debug_lock);
 
 static struct dentry *iommu_debug_root;
 
-static ssize_t debug_read_ver(struct file *file, char __user *userbuf,
-			      size_t count, loff_t *ppos)
+static inline bool is_omap_iommu_detached(struct omap_iommu *obj)
 {
-	u32 ver = omap_iommu_arch_version();
-	char buf[MAXCOLUMN], *p = buf;
-
-	p += sprintf(p, "H/W version: %d.%d\n", (ver >> 4) & 0xf , ver & 0xf);
-
-	return simple_read_from_buffer(userbuf, count, ppos, buf, p - buf);
+	return !obj->domain;
 }
 
 static ssize_t debug_read_regs(struct file *file, char __user *userbuf,
 			       size_t count, loff_t *ppos)
 {
-	struct device *dev = file->private_data;
-	struct omap_iommu *obj = dev_to_omap_iommu(dev);
+	struct omap_iommu *obj = file->private_data;
 	char *p, *buf;
 	ssize_t bytes;
 
+	if (is_omap_iommu_detached(obj))
+		return -EPERM;
+
 	buf = kmalloc(count, GFP_KERNEL);
 	if (!buf)
 		return -ENOMEM;
@@ -68,11 +58,13 @@ static ssize_t debug_read_regs(struct file *file, char __user *userbuf,
 static ssize_t debug_read_tlb(struct file *file, char __user *userbuf,
 			      size_t count, loff_t *ppos)
 {
-	struct device *dev = file->private_data;
-	struct omap_iommu *obj = dev_to_omap_iommu(dev);
+	struct omap_iommu *obj = file->private_data;
 	char *p, *buf;
 	ssize_t bytes, rest;
 
+	if (is_omap_iommu_detached(obj))
+		return -EPERM;
+
 	buf = kmalloc(count, GFP_KERNEL);
 	if (!buf)
 		return -ENOMEM;
@@ -93,133 +85,69 @@ static ssize_t debug_read_tlb(struct file *file, char __user *userbuf,
 	return bytes;
 }
 
-static ssize_t debug_write_pagetable(struct file *file,
-		     const char __user *userbuf, size_t count, loff_t *ppos)
+static void dump_ioptable(struct seq_file *s)
 {
-	struct iotlb_entry e;
-	struct cr_regs cr;
-	int err;
-	struct device *dev = file->private_data;
-	struct omap_iommu *obj = dev_to_omap_iommu(dev);
-	char buf[MAXCOLUMN], *p = buf;
-
-	count = min(count, sizeof(buf));
-
-	mutex_lock(&iommu_debug_lock);
-	if (copy_from_user(p, userbuf, count)) {
-		mutex_unlock(&iommu_debug_lock);
-		return -EFAULT;
-	}
-
-	sscanf(p, "%x %x", &cr.cam, &cr.ram);
-	if (!cr.cam || !cr.ram) {
-		mutex_unlock(&iommu_debug_lock);
-		return -EINVAL;
-	}
-
-	omap_iotlb_cr_to_e(&cr, &e);
-	err = omap_iopgtable_store_entry(obj, &e);
-	if (err)
-		dev_err(obj->dev, "%s: fail to store cr\n", __func__);
-
-	mutex_unlock(&iommu_debug_lock);
-	return count;
-}
-
-#define dump_ioptable_entry_one(lv, da, val)			\
-	({							\
-		int __err = 0;					\
-		ssize_t bytes;					\
-		const int maxcol = 22;				\
-		const char *str = "%d: %08x %08x\n";		\
-		bytes = snprintf(p, maxcol, str, lv, da, val);	\
-		p += bytes;					\
-		len -= bytes;					\
-		if (len < maxcol)				\
-			__err = -ENOMEM;			\
-		__err;						\
-	})
-
-static ssize_t dump_ioptable(struct omap_iommu *obj, char *buf, ssize_t len)
-{
-	int i;
-	u32 *iopgd;
-	char *p = buf;
+	int i, j;
+	u32 da;
+	u32 *iopgd, *iopte;
+	struct omap_iommu *obj = s->private;
 
 	spin_lock(&obj->page_table_lock);
 
 	iopgd = iopgd_offset(obj, 0);
 	for (i = 0; i < PTRS_PER_IOPGD; i++, iopgd++) {
-		int j, err;
-		u32 *iopte;
-		u32 da;
-
 		if (!*iopgd)
 			continue;
 
 		if (!(*iopgd & IOPGD_TABLE)) {
 			da = i << IOPGD_SHIFT;
-
-			err = dump_ioptable_entry_one(1, da, *iopgd);
-			if (err)
-				goto out;
+			seq_printf(s, "1: 0x%08x 0x%08x\n", da, *iopgd);
 			continue;
 		}
 
 		iopte = iopte_offset(iopgd, 0);
-
 		for (j = 0; j < PTRS_PER_IOPTE; j++, iopte++) {
 			if (!*iopte)
 				continue;
 
 			da = (i << IOPGD_SHIFT) + (j << IOPTE_SHIFT);
-			err = dump_ioptable_entry_one(2, da, *iopgd);
-			if (err)
-				goto out;
+			seq_printf(s, "2: 0x%08x 0x%08x\n", da, *iopte);
 		}
 	}
-out:
-	spin_unlock(&obj->page_table_lock);
 
-	return p - buf;
+	spin_unlock(&obj->page_table_lock);
 }
 
-static ssize_t debug_read_pagetable(struct file *file, char __user *userbuf,
-				    size_t count, loff_t *ppos)
+static int debug_read_pagetable(struct seq_file *s, void *data)
 {
-	struct device *dev = file->private_data;
-	struct omap_iommu *obj = dev_to_omap_iommu(dev);
-	char *p, *buf;
-	size_t bytes;
+	struct omap_iommu *obj = s->private;
 
-	buf = (char *)__get_free_page(GFP_KERNEL);
-	if (!buf)
-		return -ENOMEM;
-	p = buf;
-
-	p += sprintf(p, "L: %8s %8s\n", "da:", "pa:");
-	p += sprintf(p, "-----------------------------------------\n");
+	if (is_omap_iommu_detached(obj))
+		return -EPERM;
 
 	mutex_lock(&iommu_debug_lock);
 
-	bytes = PAGE_SIZE - (p - buf);
-	p += dump_ioptable(obj, p, bytes);
-
-	bytes = simple_read_from_buffer(userbuf, count, ppos, buf, p - buf);
+	seq_printf(s, "L: %8s %8s\n", "da:", "pte:");
+	seq_puts(s, "--------------------------\n");
+	dump_ioptable(s);
 
 	mutex_unlock(&iommu_debug_lock);
-	free_page((unsigned long)buf);
 
-	return bytes;
+	return 0;
 }
 
-#define DEBUG_FOPS(name)						\
-	static const struct file_operations debug_##name##_fops = {	\
-		.open = simple_open,					\
-		.read = debug_read_##name,				\
-		.write = debug_write_##name,				\
-		.llseek = generic_file_llseek,				\
-	};
+#define DEBUG_SEQ_FOPS_RO(name)						       \
+	static int debug_open_##name(struct inode *inode, struct file *file)   \
+	{								       \
+		return single_open(file, debug_read_##name, inode->i_private); \
+	}								       \
+									       \
+	static const struct file_operations debug_##name##_fops = {	       \
+		.open		= debug_open_##name,			       \
+		.read		= seq_read,				       \
+		.llseek		= seq_lseek,				       \
+		.release	= single_release,			       \
+	}
 
 #define DEBUG_FOPS_RO(name)						\
 	static const struct file_operations debug_##name##_fops = {	\
@@ -228,103 +156,63 @@ static ssize_t debug_read_pagetable(struct file *file, char __user *userbuf,
 		.llseek = generic_file_llseek,				\
 	};
 
-DEBUG_FOPS_RO(ver);
 DEBUG_FOPS_RO(regs);
 DEBUG_FOPS_RO(tlb);
-DEBUG_FOPS(pagetable);
+DEBUG_SEQ_FOPS_RO(pagetable);
 
 #define __DEBUG_ADD_FILE(attr, mode)					\
 	{								\
 		struct dentry *dent;					\
-		dent = debugfs_create_file(#attr, mode, parent,		\
-					   dev, &debug_##attr##_fops);	\
+		dent = debugfs_create_file(#attr, mode, obj->debug_dir,	\
+					   obj, &debug_##attr##_fops);	\
 		if (!dent)						\
-			return -ENOMEM;					\
+			goto err;					\
 	}
 
-#define DEBUG_ADD_FILE(name) __DEBUG_ADD_FILE(name, 0600)
 #define DEBUG_ADD_FILE_RO(name) __DEBUG_ADD_FILE(name, 0400)
 
-static int iommu_debug_register(struct device *dev, void *data)
+void omap_iommu_debugfs_add(struct omap_iommu *obj)
 {
-	struct platform_device *pdev = to_platform_device(dev);
-	struct omap_iommu *obj = platform_get_drvdata(pdev);
-	struct omap_iommu_arch_data *arch_data;
-	struct dentry *d, *parent;
-
-	if (!obj || !obj->dev)
-		return -EINVAL;
-
-	arch_data = kzalloc(sizeof(*arch_data), GFP_KERNEL);
-	if (!arch_data)
-		return -ENOMEM;
-
-	arch_data->iommu_dev = obj;
+	struct dentry *d;
 
-	dev->archdata.iommu = arch_data;
+	if (!iommu_debug_root)
+		return;
 
-	d = debugfs_create_dir(obj->name, iommu_debug_root);
-	if (!d)
-		goto nomem;
-	parent = d;
+	obj->debug_dir = debugfs_create_dir(obj->name, iommu_debug_root);
+	if (!obj->debug_dir)
+		return;
 
-	d = debugfs_create_u8("nr_tlb_entries", 400, parent,
+	d = debugfs_create_u8("nr_tlb_entries", 0400, obj->debug_dir,
 			      (u8 *)&obj->nr_tlb_entries);
 	if (!d)
-		goto nomem;
+		return;
 
-	DEBUG_ADD_FILE_RO(ver);
 	DEBUG_ADD_FILE_RO(regs);
 	DEBUG_ADD_FILE_RO(tlb);
-	DEBUG_ADD_FILE(pagetable);
+	DEBUG_ADD_FILE_RO(pagetable);
 
-	return 0;
+	return;
 
-nomem:
-	kfree(arch_data);
-	return -ENOMEM;
+err:
+	debugfs_remove_recursive(obj->debug_dir);
 }
 
-static int iommu_debug_unregister(struct device *dev, void *data)
+void omap_iommu_debugfs_remove(struct omap_iommu *obj)
 {
-	if (!dev->archdata.iommu)
-		return 0;
-
-	kfree(dev->archdata.iommu);
+	if (!obj->debug_dir)
+		return;
 
-	dev->archdata.iommu = NULL;
-
-	return 0;
+	debugfs_remove_recursive(obj->debug_dir);
 }
 
-static int __init iommu_debug_init(void)
+void __init omap_iommu_debugfs_init(void)
 {
-	struct dentry *d;
-	int err;
-
-	d = debugfs_create_dir("iommu", NULL);
-	if (!d)
-		return -ENOMEM;
-	iommu_debug_root = d;
-
-	err = omap_foreach_iommu_device(d, iommu_debug_register);
-	if (err)
-		goto err_out;
-	return 0;
-
-err_out:
-	debugfs_remove_recursive(iommu_debug_root);
-	return err;
+	iommu_debug_root = debugfs_create_dir("omap_iommu", NULL);
+	if (!iommu_debug_root)
+		pr_err("can't create debugfs dir\n");
 }
-module_init(iommu_debug_init)
 
-static void __exit iommu_debugfs_exit(void)
+void __exit omap_iommu_debugfs_exit(void)
 {
-	debugfs_remove_recursive(iommu_debug_root);
-	omap_foreach_iommu_device(NULL, iommu_debug_unregister);
+	debugfs_remove(iommu_debug_root);
 }
-module_exit(iommu_debugfs_exit)
-
-MODULE_DESCRIPTION("omap iommu: debugfs interface");
-MODULE_AUTHOR("Hiroshi DOYU <Hiroshi.DOYU@nokia.com>");
-MODULE_LICENSE("GPL v2");
diff --git a/drivers/iommu/omap-iommu.c b/drivers/iommu/omap-iommu.c
index e202b0c24120..bbb7dcef02d3 100644
--- a/drivers/iommu/omap-iommu.c
+++ b/drivers/iommu/omap-iommu.c
@@ -26,6 +26,7 @@
 #include <linux/of.h>
 #include <linux/of_iommu.h>
 #include <linux/of_irq.h>
+#include <linux/of_platform.h>
 
 #include <asm/cacheflush.h>
 
@@ -75,53 +76,23 @@ struct iotlb_lock {
 	short vict;
 };
 
-/* accommodate the difference between omap1 and omap2/3 */
-static const struct iommu_functions *arch_iommu;
-
 static struct platform_driver omap_iommu_driver;
 static struct kmem_cache *iopte_cachep;
 
 /**
- * omap_install_iommu_arch - Install archtecure specific iommu functions
- * @ops:	a pointer to architecture specific iommu functions
- *
- * There are several kind of iommu algorithm(tlb, pagetable) among
- * omap series. This interface installs such an iommu algorighm.
- **/
-int omap_install_iommu_arch(const struct iommu_functions *ops)
-{
-	if (arch_iommu)
-		return -EBUSY;
-
-	arch_iommu = ops;
-	return 0;
-}
-EXPORT_SYMBOL_GPL(omap_install_iommu_arch);
-
-/**
- * omap_uninstall_iommu_arch - Uninstall archtecure specific iommu functions
- * @ops:	a pointer to architecture specific iommu functions
- *
- * This interface uninstalls the iommu algorighm installed previously.
- **/
-void omap_uninstall_iommu_arch(const struct iommu_functions *ops)
-{
-	if (arch_iommu != ops)
-		pr_err("%s: not your arch\n", __func__);
-
-	arch_iommu = NULL;
-}
-EXPORT_SYMBOL_GPL(omap_uninstall_iommu_arch);
-
-/**
  * omap_iommu_save_ctx - Save registers for pm off-mode support
  * @dev:	client device
  **/
 void omap_iommu_save_ctx(struct device *dev)
 {
 	struct omap_iommu *obj = dev_to_omap_iommu(dev);
+	u32 *p = obj->ctx;
+	int i;
 
-	arch_iommu->save_ctx(obj);
+	for (i = 0; i < (MMU_REG_SIZE / sizeof(u32)); i++) {
+		p[i] = iommu_read_reg(obj, i * sizeof(u32));
+		dev_dbg(obj->dev, "%s\t[%02d] %08x\n", __func__, i, p[i]);
+	}
 }
 EXPORT_SYMBOL_GPL(omap_iommu_save_ctx);
 
@@ -132,28 +103,74 @@ EXPORT_SYMBOL_GPL(omap_iommu_save_ctx);
 void omap_iommu_restore_ctx(struct device *dev)
 {
 	struct omap_iommu *obj = dev_to_omap_iommu(dev);
+	u32 *p = obj->ctx;
+	int i;
 
-	arch_iommu->restore_ctx(obj);
+	for (i = 0; i < (MMU_REG_SIZE / sizeof(u32)); i++) {
+		iommu_write_reg(obj, p[i], i * sizeof(u32));
+		dev_dbg(obj->dev, "%s\t[%02d] %08x\n", __func__, i, p[i]);
+	}
 }
 EXPORT_SYMBOL_GPL(omap_iommu_restore_ctx);
 
-/**
- * omap_iommu_arch_version - Return running iommu arch version
- **/
-u32 omap_iommu_arch_version(void)
+static void __iommu_set_twl(struct omap_iommu *obj, bool on)
+{
+	u32 l = iommu_read_reg(obj, MMU_CNTL);
+
+	if (on)
+		iommu_write_reg(obj, MMU_IRQ_TWL_MASK, MMU_IRQENABLE);
+	else
+		iommu_write_reg(obj, MMU_IRQ_TLB_MISS_MASK, MMU_IRQENABLE);
+
+	l &= ~MMU_CNTL_MASK;
+	if (on)
+		l |= (MMU_CNTL_MMU_EN | MMU_CNTL_TWL_EN);
+	else
+		l |= (MMU_CNTL_MMU_EN);
+
+	iommu_write_reg(obj, l, MMU_CNTL);
+}
+
+static int omap2_iommu_enable(struct omap_iommu *obj)
+{
+	u32 l, pa;
+
+	if (!obj->iopgd || !IS_ALIGNED((u32)obj->iopgd,  SZ_16K))
+		return -EINVAL;
+
+	pa = virt_to_phys(obj->iopgd);
+	if (!IS_ALIGNED(pa, SZ_16K))
+		return -EINVAL;
+
+	l = iommu_read_reg(obj, MMU_REVISION);
+	dev_info(obj->dev, "%s: version %d.%d\n", obj->name,
+		 (l >> 4) & 0xf, l & 0xf);
+
+	iommu_write_reg(obj, pa, MMU_TTB);
+
+	if (obj->has_bus_err_back)
+		iommu_write_reg(obj, MMU_GP_REG_BUS_ERR_BACK_EN, MMU_GP_REG);
+
+	__iommu_set_twl(obj, true);
+
+	return 0;
+}
+
+static void omap2_iommu_disable(struct omap_iommu *obj)
 {
-	return arch_iommu->version;
+	u32 l = iommu_read_reg(obj, MMU_CNTL);
+
+	l &= ~MMU_CNTL_MASK;
+	iommu_write_reg(obj, l, MMU_CNTL);
+
+	dev_dbg(obj->dev, "%s is shutting down\n", obj->name);
 }
-EXPORT_SYMBOL_GPL(omap_iommu_arch_version);
 
 static int iommu_enable(struct omap_iommu *obj)
 {
 	int err;
 	struct platform_device *pdev = to_platform_device(obj->dev);
-	struct iommu_platform_data *pdata = pdev->dev.platform_data;
-
-	if (!arch_iommu)
-		return -ENODEV;
+	struct iommu_platform_data *pdata = dev_get_platdata(&pdev->dev);
 
 	if (pdata && pdata->deassert_reset) {
 		err = pdata->deassert_reset(pdev, pdata->reset_name);
@@ -165,7 +182,7 @@ static int iommu_enable(struct omap_iommu *obj)
 
 	pm_runtime_get_sync(obj->dev);
 
-	err = arch_iommu->enable(obj);
+	err = omap2_iommu_enable(obj);
 
 	return err;
 }
@@ -173,9 +190,9 @@ static int iommu_enable(struct omap_iommu *obj)
 static void iommu_disable(struct omap_iommu *obj)
 {
 	struct platform_device *pdev = to_platform_device(obj->dev);
-	struct iommu_platform_data *pdata = pdev->dev.platform_data;
+	struct iommu_platform_data *pdata = dev_get_platdata(&pdev->dev);
 
-	arch_iommu->disable(obj);
+	omap2_iommu_disable(obj);
 
 	pm_runtime_put_sync(obj->dev);
 
@@ -186,44 +203,51 @@ static void iommu_disable(struct omap_iommu *obj)
 /*
  *	TLB operations
  */
-void omap_iotlb_cr_to_e(struct cr_regs *cr, struct iotlb_entry *e)
-{
-	BUG_ON(!cr || !e);
-
-	arch_iommu->cr_to_e(cr, e);
-}
-EXPORT_SYMBOL_GPL(omap_iotlb_cr_to_e);
-
 static inline int iotlb_cr_valid(struct cr_regs *cr)
 {
 	if (!cr)
 		return -EINVAL;
 
-	return arch_iommu->cr_valid(cr);
-}
-
-static inline struct cr_regs *iotlb_alloc_cr(struct omap_iommu *obj,
-					     struct iotlb_entry *e)
-{
-	if (!e)
-		return NULL;
-
-	return arch_iommu->alloc_cr(obj, e);
+	return cr->cam & MMU_CAM_V;
 }
 
 static u32 iotlb_cr_to_virt(struct cr_regs *cr)
 {
-	return arch_iommu->cr_to_virt(cr);
+	u32 page_size = cr->cam & MMU_CAM_PGSZ_MASK;
+	u32 mask = get_cam_va_mask(cr->cam & page_size);
+
+	return cr->cam & mask;
 }
 
 static u32 get_iopte_attr(struct iotlb_entry *e)
 {
-	return arch_iommu->get_pte_attr(e);
+	u32 attr;
+
+	attr = e->mixed << 5;
+	attr |= e->endian;
+	attr |= e->elsz >> 3;
+	attr <<= (((e->pgsz == MMU_CAM_PGSZ_4K) ||
+			(e->pgsz == MMU_CAM_PGSZ_64K)) ? 0 : 6);
+	return attr;
 }
 
 static u32 iommu_report_fault(struct omap_iommu *obj, u32 *da)
 {
-	return arch_iommu->fault_isr(obj, da);
+	u32 status, fault_addr;
+
+	status = iommu_read_reg(obj, MMU_IRQSTATUS);
+	status &= MMU_IRQ_MASK;
+	if (!status) {
+		*da = 0;
+		return 0;
+	}
+
+	fault_addr = iommu_read_reg(obj, MMU_FAULT_AD);
+	*da = fault_addr;
+
+	iommu_write_reg(obj, status, MMU_IRQSTATUS);
+
+	return status;
 }
 
 static void iotlb_lock_get(struct omap_iommu *obj, struct iotlb_lock *l)
@@ -249,31 +273,19 @@ static void iotlb_lock_set(struct omap_iommu *obj, struct iotlb_lock *l)
 
 static void iotlb_read_cr(struct omap_iommu *obj, struct cr_regs *cr)
 {
-	arch_iommu->tlb_read_cr(obj, cr);
+	cr->cam = iommu_read_reg(obj, MMU_READ_CAM);
+	cr->ram = iommu_read_reg(obj, MMU_READ_RAM);
 }
 
 static void iotlb_load_cr(struct omap_iommu *obj, struct cr_regs *cr)
 {
-	arch_iommu->tlb_load_cr(obj, cr);
+	iommu_write_reg(obj, cr->cam | MMU_CAM_V, MMU_CAM);
+	iommu_write_reg(obj, cr->ram, MMU_RAM);
 
 	iommu_write_reg(obj, 1, MMU_FLUSH_ENTRY);
 	iommu_write_reg(obj, 1, MMU_LD_TLB);
 }
 
-/**
- * iotlb_dump_cr - Dump an iommu tlb entry into buf
- * @obj:	target iommu
- * @cr:		contents of cam and ram register
- * @buf:	output buffer
- **/
-static inline ssize_t iotlb_dump_cr(struct omap_iommu *obj, struct cr_regs *cr,
-				    char *buf)
-{
-	BUG_ON(!cr || !buf);
-
-	return arch_iommu->dump_cr(obj, cr, buf);
-}
-
 /* only used in iotlb iteration for-loop */
 static struct cr_regs __iotlb_read_cr(struct omap_iommu *obj, int n)
 {
@@ -288,12 +300,36 @@ static struct cr_regs __iotlb_read_cr(struct omap_iommu *obj, int n)
 	return cr;
 }
 
+#ifdef PREFETCH_IOTLB
+static struct cr_regs *iotlb_alloc_cr(struct omap_iommu *obj,
+				      struct iotlb_entry *e)
+{
+	struct cr_regs *cr;
+
+	if (!e)
+		return NULL;
+
+	if (e->da & ~(get_cam_va_mask(e->pgsz))) {
+		dev_err(obj->dev, "%s:\twrong alignment: %08x\n", __func__,
+			e->da);
+		return ERR_PTR(-EINVAL);
+	}
+
+	cr = kmalloc(sizeof(*cr), GFP_KERNEL);
+	if (!cr)
+		return ERR_PTR(-ENOMEM);
+
+	cr->cam = (e->da & MMU_CAM_VATAG_MASK) | e->prsvd | e->pgsz | e->valid;
+	cr->ram = e->pa | e->endian | e->elsz | e->mixed;
+
+	return cr;
+}
+
 /**
  * load_iotlb_entry - Set an iommu tlb entry
  * @obj:	target iommu
  * @e:		an iommu tlb entry info
  **/
-#ifdef PREFETCH_IOTLB
 static int load_iotlb_entry(struct omap_iommu *obj, struct iotlb_entry *e)
 {
 	int err = 0;
@@ -422,7 +458,45 @@ static void flush_iotlb_all(struct omap_iommu *obj)
 	pm_runtime_put_sync(obj->dev);
 }
 
-#if defined(CONFIG_OMAP_IOMMU_DEBUG) || defined(CONFIG_OMAP_IOMMU_DEBUG_MODULE)
+#ifdef CONFIG_OMAP_IOMMU_DEBUG
+
+#define pr_reg(name)							\
+	do {								\
+		ssize_t bytes;						\
+		const char *str = "%20s: %08x\n";			\
+		const int maxcol = 32;					\
+		bytes = snprintf(p, maxcol, str, __stringify(name),	\
+				 iommu_read_reg(obj, MMU_##name));	\
+		p += bytes;						\
+		len -= bytes;						\
+		if (len < maxcol)					\
+			goto out;					\
+	} while (0)
+
+static ssize_t
+omap2_iommu_dump_ctx(struct omap_iommu *obj, char *buf, ssize_t len)
+{
+	char *p = buf;
+
+	pr_reg(REVISION);
+	pr_reg(IRQSTATUS);
+	pr_reg(IRQENABLE);
+	pr_reg(WALKING_ST);
+	pr_reg(CNTL);
+	pr_reg(FAULT_AD);
+	pr_reg(TTB);
+	pr_reg(LOCK);
+	pr_reg(LD_TLB);
+	pr_reg(CAM);
+	pr_reg(RAM);
+	pr_reg(GFLUSH);
+	pr_reg(FLUSH_ENTRY);
+	pr_reg(READ_CAM);
+	pr_reg(READ_RAM);
+	pr_reg(EMU_FAULT_AD);
+out:
+	return p - buf;
+}
 
 ssize_t omap_iommu_dump_ctx(struct omap_iommu *obj, char *buf, ssize_t bytes)
 {
@@ -431,13 +505,12 @@ ssize_t omap_iommu_dump_ctx(struct omap_iommu *obj, char *buf, ssize_t bytes)
 
 	pm_runtime_get_sync(obj->dev);
 
-	bytes = arch_iommu->dump_ctx(obj, buf, bytes);
+	bytes = omap2_iommu_dump_ctx(obj, buf, bytes);
 
 	pm_runtime_put_sync(obj->dev);
 
 	return bytes;
 }
-EXPORT_SYMBOL_GPL(omap_iommu_dump_ctx);
 
 static int
 __dump_tlb_entries(struct omap_iommu *obj, struct cr_regs *crs, int num)
@@ -463,6 +536,24 @@ __dump_tlb_entries(struct omap_iommu *obj, struct cr_regs *crs, int num)
 }
 
 /**
+ * iotlb_dump_cr - Dump an iommu tlb entry into buf
+ * @obj:	target iommu
+ * @cr:		contents of cam and ram register
+ * @buf:	output buffer
+ **/
+static ssize_t iotlb_dump_cr(struct omap_iommu *obj, struct cr_regs *cr,
+			     char *buf)
+{
+	char *p = buf;
+
+	/* FIXME: Need more detail analysis of cam/ram */
+	p += sprintf(p, "%08x %08x %01x\n", cr->cam, cr->ram,
+					(cr->cam & MMU_CAM_P) ? 1 : 0);
+
+	return p - buf;
+}
+
+/**
  * omap_dump_tlb_entries - dump cr arrays to given buffer
  * @obj:	target iommu
  * @buf:	output buffer
@@ -487,16 +578,8 @@ size_t omap_dump_tlb_entries(struct omap_iommu *obj, char *buf, ssize_t bytes)
 
 	return p - buf;
 }
-EXPORT_SYMBOL_GPL(omap_dump_tlb_entries);
 
-int omap_foreach_iommu_device(void *data, int (*fn)(struct device *, void *))
-{
-	return driver_for_each_device(&omap_iommu_driver.driver,
-				      NULL, data, fn);
-}
-EXPORT_SYMBOL_GPL(omap_foreach_iommu_device);
-
-#endif /* CONFIG_OMAP_IOMMU_DEBUG_MODULE */
+#endif /* CONFIG_OMAP_IOMMU_DEBUG */
 
 /*
  *	H/W pagetable operations
@@ -679,7 +762,8 @@ iopgtable_store_entry_core(struct omap_iommu *obj, struct iotlb_entry *e)
  * @obj:	target iommu
  * @e:		an iommu tlb entry info
  **/
-int omap_iopgtable_store_entry(struct omap_iommu *obj, struct iotlb_entry *e)
+static int
+omap_iopgtable_store_entry(struct omap_iommu *obj, struct iotlb_entry *e)
 {
 	int err;
 
@@ -689,7 +773,6 @@ int omap_iopgtable_store_entry(struct omap_iommu *obj, struct iotlb_entry *e)
 		prefetch_iotlb_entry(obj, e);
 	return err;
 }
-EXPORT_SYMBOL_GPL(omap_iopgtable_store_entry);
 
 /**
  * iopgtable_lookup_entry - Lookup an iommu pte entry
@@ -818,8 +901,9 @@ static irqreturn_t iommu_fault_handler(int irq, void *data)
 	u32 *iopgd, *iopte;
 	struct omap_iommu *obj = data;
 	struct iommu_domain *domain = obj->domain;
+	struct omap_iommu_domain *omap_domain = domain->priv;
 
-	if (!obj->refcount)
+	if (!omap_domain->iommu_dev)
 		return IRQ_NONE;
 
 	errs = iommu_report_fault(obj, &da);
@@ -879,34 +963,18 @@ static struct omap_iommu *omap_iommu_attach(const char *name, u32 *iopgd)
 
 	spin_lock(&obj->iommu_lock);
 
-	/* an iommu device can only be attached once */
-	if (++obj->refcount > 1) {
-		dev_err(dev, "%s: already attached!\n", obj->name);
-		err = -EBUSY;
-		goto err_enable;
-	}
-
 	obj->iopgd = iopgd;
 	err = iommu_enable(obj);
 	if (err)
 		goto err_enable;
 	flush_iotlb_all(obj);
 
-	if (!try_module_get(obj->owner)) {
-		err = -ENODEV;
-		goto err_module;
-	}
-
 	spin_unlock(&obj->iommu_lock);
 
 	dev_dbg(obj->dev, "%s: %s\n", __func__, obj->name);
 	return obj;
 
-err_module:
-	if (obj->refcount == 1)
-		iommu_disable(obj);
 err_enable:
-	obj->refcount--;
 	spin_unlock(&obj->iommu_lock);
 	return ERR_PTR(err);
 }
@@ -922,11 +990,7 @@ static void omap_iommu_detach(struct omap_iommu *obj)
 
 	spin_lock(&obj->iommu_lock);
 
-	if (--obj->refcount == 0)
-		iommu_disable(obj);
-
-	module_put(obj->owner);
-
+	iommu_disable(obj);
 	obj->iopgd = NULL;
 
 	spin_unlock(&obj->iommu_lock);
@@ -943,7 +1007,7 @@ static int omap_iommu_probe(struct platform_device *pdev)
 	int irq;
 	struct omap_iommu *obj;
 	struct resource *res;
-	struct iommu_platform_data *pdata = pdev->dev.platform_data;
+	struct iommu_platform_data *pdata = dev_get_platdata(&pdev->dev);
 	struct device_node *of = pdev->dev.of_node;
 
 	obj = devm_kzalloc(&pdev->dev, sizeof(*obj) + MMU_REG_SIZE, GFP_KERNEL);
@@ -990,6 +1054,8 @@ static int omap_iommu_probe(struct platform_device *pdev)
 	pm_runtime_irq_safe(obj->dev);
 	pm_runtime_enable(obj->dev);
 
+	omap_iommu_debugfs_add(obj);
+
 	dev_info(&pdev->dev, "%s registered\n", obj->name);
 	return 0;
 }
@@ -999,6 +1065,7 @@ static int omap_iommu_remove(struct platform_device *pdev)
 	struct omap_iommu *obj = platform_get_drvdata(pdev);
 
 	iopgtable_clear_entry_all(obj);
+	omap_iommu_debugfs_remove(obj);
 
 	pm_runtime_disable(obj->dev);
 
@@ -1006,7 +1073,7 @@ static int omap_iommu_remove(struct platform_device *pdev)
 	return 0;
 }
 
-static struct of_device_id omap_iommu_of_match[] = {
+static const struct of_device_id omap_iommu_of_match[] = {
 	{ .compatible = "ti,omap2-iommu" },
 	{ .compatible = "ti,omap4-iommu" },
 	{ .compatible = "ti,dra7-iommu"	},
@@ -1035,7 +1102,6 @@ static u32 iotlb_init_entry(struct iotlb_entry *e, u32 da, u32 pa, int pgsz)
 	e->da		= da;
 	e->pa		= pa;
 	e->valid	= MMU_CAM_V;
-	/* FIXME: add OMAP1 support */
 	e->pgsz		= pgsz;
 	e->endian	= MMU_RAM_ENDIAN_LITTLE;
 	e->elsz		= MMU_RAM_ELSZ_8;
@@ -1091,6 +1157,11 @@ omap_iommu_attach_dev(struct iommu_domain *domain, struct device *dev)
 	struct omap_iommu_arch_data *arch_data = dev->archdata.iommu;
 	int ret = 0;
 
+	if (!arch_data || !arch_data->name) {
+		dev_err(dev, "device doesn't have an associated iommu\n");
+		return -EINVAL;
+	}
+
 	spin_lock(&omap_domain->lock);
 
 	/* only a single device is supported per domain for now */
@@ -1135,6 +1206,7 @@ static void _omap_iommu_detach_dev(struct omap_iommu_domain *omap_domain,
 
 	omap_domain->iommu_dev = arch_data->iommu_dev = NULL;
 	omap_domain->dev = NULL;
+	oiommu->domain = NULL;
 }
 
 static void omap_iommu_detach_dev(struct iommu_domain *domain,
@@ -1239,6 +1311,7 @@ static int omap_iommu_add_device(struct device *dev)
 {
 	struct omap_iommu_arch_data *arch_data;
 	struct device_node *np;
+	struct platform_device *pdev;
 
 	/*
 	 * Allocate the archdata iommu structure for DT-based devices.
@@ -1253,13 +1326,19 @@ static int omap_iommu_add_device(struct device *dev)
 	if (!np)
 		return 0;
 
+	pdev = of_find_device_by_node(np);
+	if (WARN_ON(!pdev)) {
+		of_node_put(np);
+		return -EINVAL;
+	}
+
 	arch_data = kzalloc(sizeof(*arch_data), GFP_KERNEL);
 	if (!arch_data) {
 		of_node_put(np);
 		return -ENOMEM;
 	}
 
-	arch_data->name = kstrdup(dev_name(dev), GFP_KERNEL);
+	arch_data->name = kstrdup(dev_name(&pdev->dev), GFP_KERNEL);
 	dev->archdata.iommu = arch_data;
 
 	of_node_put(np);
@@ -1285,6 +1364,7 @@ static const struct iommu_ops omap_iommu_ops = {
 	.detach_dev	= omap_iommu_detach_dev,
 	.map		= omap_iommu_map,
 	.unmap		= omap_iommu_unmap,
+	.map_sg		= default_iommu_map_sg,
 	.iova_to_phys	= omap_iommu_iova_to_phys,
 	.add_device	= omap_iommu_add_device,
 	.remove_device	= omap_iommu_remove_device,
@@ -1305,6 +1385,8 @@ static int __init omap_iommu_init(void)
 
 	bus_set_iommu(&platform_bus_type, &omap_iommu_ops);
 
+	omap_iommu_debugfs_init();
+
 	return platform_driver_register(&omap_iommu_driver);
 }
 /* must be ready before omap3isp is probed */
@@ -1315,6 +1397,8 @@ static void __exit omap_iommu_exit(void)
 	kmem_cache_destroy(iopte_cachep);
 
 	platform_driver_unregister(&omap_iommu_driver);
+
+	omap_iommu_debugfs_exit();
 }
 module_exit(omap_iommu_exit);
 
diff --git a/drivers/iommu/omap-iommu.h b/drivers/iommu/omap-iommu.h
index 1275a822934b..d736630df3c8 100644
--- a/drivers/iommu/omap-iommu.h
+++ b/drivers/iommu/omap-iommu.h
@@ -10,9 +10,8 @@
  * published by the Free Software Foundation.
  */
 
-#if defined(CONFIG_ARCH_OMAP1)
-#error "iommu for this processor not implemented yet"
-#endif
+#ifndef _OMAP_IOMMU_H
+#define _OMAP_IOMMU_H
 
 struct iotlb_entry {
 	u32 da;
@@ -28,13 +27,11 @@ struct iotlb_entry {
 
 struct omap_iommu {
 	const char	*name;
-	struct module	*owner;
 	void __iomem	*regbase;
 	struct device	*dev;
-	void		*isr_priv;
 	struct iommu_domain *domain;
+	struct dentry	*debug_dir;
 
-	unsigned int	refcount;
 	spinlock_t	iommu_lock;	/* global for this whole object */
 
 	/*
@@ -68,34 +65,6 @@ struct cr_regs {
 	};
 };
 
-/* architecture specific functions */
-struct iommu_functions {
-	unsigned long	version;
-
-	int (*enable)(struct omap_iommu *obj);
-	void (*disable)(struct omap_iommu *obj);
-	void (*set_twl)(struct omap_iommu *obj, bool on);
-	u32 (*fault_isr)(struct omap_iommu *obj, u32 *ra);
-
-	void (*tlb_read_cr)(struct omap_iommu *obj, struct cr_regs *cr);
-	void (*tlb_load_cr)(struct omap_iommu *obj, struct cr_regs *cr);
-
-	struct cr_regs *(*alloc_cr)(struct omap_iommu *obj,
-							struct iotlb_entry *e);
-	int (*cr_valid)(struct cr_regs *cr);
-	u32 (*cr_to_virt)(struct cr_regs *cr);
-	void (*cr_to_e)(struct cr_regs *cr, struct iotlb_entry *e);
-	ssize_t (*dump_cr)(struct omap_iommu *obj, struct cr_regs *cr,
-							char *buf);
-
-	u32 (*get_pte_attr)(struct iotlb_entry *e);
-
-	void (*save_ctx)(struct omap_iommu *obj);
-	void (*restore_ctx)(struct omap_iommu *obj);
-	ssize_t (*dump_ctx)(struct omap_iommu *obj, char *buf, ssize_t len);
-};
-
-#ifdef CONFIG_IOMMU_API
 /**
  * dev_to_omap_iommu() - retrieves an omap iommu object from a user device
  * @dev: iommu client device
@@ -106,7 +75,6 @@ static inline struct omap_iommu *dev_to_omap_iommu(struct device *dev)
 
 	return arch_data->iommu_dev;
 }
-#endif
 
 /*
  * MMU Register offsets
@@ -134,6 +102,28 @@ static inline struct omap_iommu *dev_to_omap_iommu(struct device *dev)
 /*
  * MMU Register bit definitions
  */
+/* IRQSTATUS & IRQENABLE */
+#define MMU_IRQ_MULTIHITFAULT	(1 << 4)
+#define MMU_IRQ_TABLEWALKFAULT	(1 << 3)
+#define MMU_IRQ_EMUMISS		(1 << 2)
+#define MMU_IRQ_TRANSLATIONFAULT	(1 << 1)
+#define MMU_IRQ_TLBMISS		(1 << 0)
+
+#define __MMU_IRQ_FAULT		\
+	(MMU_IRQ_MULTIHITFAULT | MMU_IRQ_EMUMISS | MMU_IRQ_TRANSLATIONFAULT)
+#define MMU_IRQ_MASK		\
+	(__MMU_IRQ_FAULT | MMU_IRQ_TABLEWALKFAULT | MMU_IRQ_TLBMISS)
+#define MMU_IRQ_TWL_MASK	(__MMU_IRQ_FAULT | MMU_IRQ_TABLEWALKFAULT)
+#define MMU_IRQ_TLB_MISS_MASK	(__MMU_IRQ_FAULT | MMU_IRQ_TLBMISS)
+
+/* MMU_CNTL */
+#define MMU_CNTL_SHIFT		1
+#define MMU_CNTL_MASK		(7 << MMU_CNTL_SHIFT)
+#define MMU_CNTL_EML_TLB	(1 << 3)
+#define MMU_CNTL_TWL_EN		(1 << 2)
+#define MMU_CNTL_MMU_EN		(1 << 1)
+
+/* CAM */
 #define MMU_CAM_VATAG_SHIFT	12
 #define MMU_CAM_VATAG_MASK \
 	((~0UL >> MMU_CAM_VATAG_SHIFT) << MMU_CAM_VATAG_SHIFT)
@@ -145,6 +135,7 @@ static inline struct omap_iommu *dev_to_omap_iommu(struct device *dev)
 #define MMU_CAM_PGSZ_4K		(2 << 0)
 #define MMU_CAM_PGSZ_16M	(3 << 0)
 
+/* RAM */
 #define MMU_RAM_PADDR_SHIFT	12
 #define MMU_RAM_PADDR_MASK \
 	((~0UL >> MMU_RAM_PADDR_SHIFT) << MMU_RAM_PADDR_SHIFT)
@@ -166,6 +157,12 @@ static inline struct omap_iommu *dev_to_omap_iommu(struct device *dev)
 
 #define MMU_GP_REG_BUS_ERR_BACK_EN	0x1
 
+#define get_cam_va_mask(pgsz)				\
+	(((pgsz) == MMU_CAM_PGSZ_16M) ? 0xff000000 :	\
+	 ((pgsz) == MMU_CAM_PGSZ_1M)  ? 0xfff00000 :	\
+	 ((pgsz) == MMU_CAM_PGSZ_64K) ? 0xffff0000 :	\
+	 ((pgsz) == MMU_CAM_PGSZ_4K)  ? 0xfffff000 : 0)
+
 /*
  * utilities for super page(16MB, 1MB, 64KB and 4KB)
  */
@@ -193,27 +190,25 @@ static inline struct omap_iommu *dev_to_omap_iommu(struct device *dev)
 /*
  * global functions
  */
-extern u32 omap_iommu_arch_version(void);
-
-extern void omap_iotlb_cr_to_e(struct cr_regs *cr, struct iotlb_entry *e);
-
-extern int
-omap_iopgtable_store_entry(struct omap_iommu *obj, struct iotlb_entry *e);
-
-extern void omap_iommu_save_ctx(struct device *dev);
-extern void omap_iommu_restore_ctx(struct device *dev);
-
-extern int omap_foreach_iommu_device(void *data,
-				int (*fn)(struct device *, void *));
-
-extern int omap_install_iommu_arch(const struct iommu_functions *ops);
-extern void omap_uninstall_iommu_arch(const struct iommu_functions *ops);
-
+#ifdef CONFIG_OMAP_IOMMU_DEBUG
 extern ssize_t
 omap_iommu_dump_ctx(struct omap_iommu *obj, char *buf, ssize_t len);
 extern size_t
 omap_dump_tlb_entries(struct omap_iommu *obj, char *buf, ssize_t len);
 
+void omap_iommu_debugfs_init(void);
+void omap_iommu_debugfs_exit(void);
+
+void omap_iommu_debugfs_add(struct omap_iommu *obj);
+void omap_iommu_debugfs_remove(struct omap_iommu *obj);
+#else
+static inline void omap_iommu_debugfs_init(void) { }
+static inline void omap_iommu_debugfs_exit(void) { }
+
+static inline void omap_iommu_debugfs_add(struct omap_iommu *obj) { }
+static inline void omap_iommu_debugfs_remove(struct omap_iommu *obj) { }
+#endif
+
 /*
  * register accessors
  */
@@ -226,3 +221,5 @@ static inline void iommu_write_reg(struct omap_iommu *obj, u32 val, size_t offs)
 {
 	__raw_writel(val, obj->regbase + offs);
 }
+
+#endif /* _OMAP_IOMMU_H */
diff --git a/drivers/iommu/omap-iommu2.c b/drivers/iommu/omap-iommu2.c
deleted file mode 100644
index 5e1ea3b0bf16..000000000000
--- a/drivers/iommu/omap-iommu2.c
+++ /dev/null
@@ -1,337 +0,0 @@
-/*
- * omap iommu: omap2/3 architecture specific functions
- *
- * Copyright (C) 2008-2009 Nokia Corporation
- *
- * Written by Hiroshi DOYU <Hiroshi.DOYU@nokia.com>,
- *		Paul Mundt and Toshihiro Kobayashi
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- */
-
-#include <linux/err.h>
-#include <linux/device.h>
-#include <linux/io.h>
-#include <linux/jiffies.h>
-#include <linux/module.h>
-#include <linux/omap-iommu.h>
-#include <linux/slab.h>
-#include <linux/stringify.h>
-#include <linux/platform_data/iommu-omap.h>
-
-#include "omap-iommu.h"
-
-/*
- * omap2 architecture specific register bit definitions
- */
-#define IOMMU_ARCH_VERSION	0x00000011
-
-/* IRQSTATUS & IRQENABLE */
-#define MMU_IRQ_MULTIHITFAULT	(1 << 4)
-#define MMU_IRQ_TABLEWALKFAULT	(1 << 3)
-#define MMU_IRQ_EMUMISS		(1 << 2)
-#define MMU_IRQ_TRANSLATIONFAULT	(1 << 1)
-#define MMU_IRQ_TLBMISS		(1 << 0)
-
-#define __MMU_IRQ_FAULT		\
-	(MMU_IRQ_MULTIHITFAULT | MMU_IRQ_EMUMISS | MMU_IRQ_TRANSLATIONFAULT)
-#define MMU_IRQ_MASK		\
-	(__MMU_IRQ_FAULT | MMU_IRQ_TABLEWALKFAULT | MMU_IRQ_TLBMISS)
-#define MMU_IRQ_TWL_MASK	(__MMU_IRQ_FAULT | MMU_IRQ_TABLEWALKFAULT)
-#define MMU_IRQ_TLB_MISS_MASK	(__MMU_IRQ_FAULT | MMU_IRQ_TLBMISS)
-
-/* MMU_CNTL */
-#define MMU_CNTL_SHIFT		1
-#define MMU_CNTL_MASK		(7 << MMU_CNTL_SHIFT)
-#define MMU_CNTL_EML_TLB	(1 << 3)
-#define MMU_CNTL_TWL_EN		(1 << 2)
-#define MMU_CNTL_MMU_EN		(1 << 1)
-
-#define get_cam_va_mask(pgsz)				\
-	(((pgsz) == MMU_CAM_PGSZ_16M) ? 0xff000000 :	\
-	 ((pgsz) == MMU_CAM_PGSZ_1M)  ? 0xfff00000 :	\
-	 ((pgsz) == MMU_CAM_PGSZ_64K) ? 0xffff0000 :	\
-	 ((pgsz) == MMU_CAM_PGSZ_4K)  ? 0xfffff000 : 0)
-
-/* IOMMU errors */
-#define OMAP_IOMMU_ERR_TLB_MISS		(1 << 0)
-#define OMAP_IOMMU_ERR_TRANS_FAULT	(1 << 1)
-#define OMAP_IOMMU_ERR_EMU_MISS		(1 << 2)
-#define OMAP_IOMMU_ERR_TBLWALK_FAULT	(1 << 3)
-#define OMAP_IOMMU_ERR_MULTIHIT_FAULT	(1 << 4)
-
-static void __iommu_set_twl(struct omap_iommu *obj, bool on)
-{
-	u32 l = iommu_read_reg(obj, MMU_CNTL);
-
-	if (on)
-		iommu_write_reg(obj, MMU_IRQ_TWL_MASK, MMU_IRQENABLE);
-	else
-		iommu_write_reg(obj, MMU_IRQ_TLB_MISS_MASK, MMU_IRQENABLE);
-
-	l &= ~MMU_CNTL_MASK;
-	if (on)
-		l |= (MMU_CNTL_MMU_EN | MMU_CNTL_TWL_EN);
-	else
-		l |= (MMU_CNTL_MMU_EN);
-
-	iommu_write_reg(obj, l, MMU_CNTL);
-}
-
-
-static int omap2_iommu_enable(struct omap_iommu *obj)
-{
-	u32 l, pa;
-
-	if (!obj->iopgd || !IS_ALIGNED((u32)obj->iopgd,  SZ_16K))
-		return -EINVAL;
-
-	pa = virt_to_phys(obj->iopgd);
-	if (!IS_ALIGNED(pa, SZ_16K))
-		return -EINVAL;
-
-	l = iommu_read_reg(obj, MMU_REVISION);
-	dev_info(obj->dev, "%s: version %d.%d\n", obj->name,
-		 (l >> 4) & 0xf, l & 0xf);
-
-	iommu_write_reg(obj, pa, MMU_TTB);
-
-	if (obj->has_bus_err_back)
-		iommu_write_reg(obj, MMU_GP_REG_BUS_ERR_BACK_EN, MMU_GP_REG);
-
-	__iommu_set_twl(obj, true);
-
-	return 0;
-}
-
-static void omap2_iommu_disable(struct omap_iommu *obj)
-{
-	u32 l = iommu_read_reg(obj, MMU_CNTL);
-
-	l &= ~MMU_CNTL_MASK;
-	iommu_write_reg(obj, l, MMU_CNTL);
-
-	dev_dbg(obj->dev, "%s is shutting down\n", obj->name);
-}
-
-static void omap2_iommu_set_twl(struct omap_iommu *obj, bool on)
-{
-	__iommu_set_twl(obj, false);
-}
-
-static u32 omap2_iommu_fault_isr(struct omap_iommu *obj, u32 *ra)
-{
-	u32 stat, da;
-	u32 errs = 0;
-
-	stat = iommu_read_reg(obj, MMU_IRQSTATUS);
-	stat &= MMU_IRQ_MASK;
-	if (!stat) {
-		*ra = 0;
-		return 0;
-	}
-
-	da = iommu_read_reg(obj, MMU_FAULT_AD);
-	*ra = da;
-
-	if (stat & MMU_IRQ_TLBMISS)
-		errs |= OMAP_IOMMU_ERR_TLB_MISS;
-	if (stat & MMU_IRQ_TRANSLATIONFAULT)
-		errs |= OMAP_IOMMU_ERR_TRANS_FAULT;
-	if (stat & MMU_IRQ_EMUMISS)
-		errs |= OMAP_IOMMU_ERR_EMU_MISS;
-	if (stat & MMU_IRQ_TABLEWALKFAULT)
-		errs |= OMAP_IOMMU_ERR_TBLWALK_FAULT;
-	if (stat & MMU_IRQ_MULTIHITFAULT)
-		errs |= OMAP_IOMMU_ERR_MULTIHIT_FAULT;
-	iommu_write_reg(obj, stat, MMU_IRQSTATUS);
-
-	return errs;
-}
-
-static void omap2_tlb_read_cr(struct omap_iommu *obj, struct cr_regs *cr)
-{
-	cr->cam = iommu_read_reg(obj, MMU_READ_CAM);
-	cr->ram = iommu_read_reg(obj, MMU_READ_RAM);
-}
-
-static void omap2_tlb_load_cr(struct omap_iommu *obj, struct cr_regs *cr)
-{
-	iommu_write_reg(obj, cr->cam | MMU_CAM_V, MMU_CAM);
-	iommu_write_reg(obj, cr->ram, MMU_RAM);
-}
-
-static u32 omap2_cr_to_virt(struct cr_regs *cr)
-{
-	u32 page_size = cr->cam & MMU_CAM_PGSZ_MASK;
-	u32 mask = get_cam_va_mask(cr->cam & page_size);
-
-	return cr->cam & mask;
-}
-
-static struct cr_regs *omap2_alloc_cr(struct omap_iommu *obj,
-						struct iotlb_entry *e)
-{
-	struct cr_regs *cr;
-
-	if (e->da & ~(get_cam_va_mask(e->pgsz))) {
-		dev_err(obj->dev, "%s:\twrong alignment: %08x\n", __func__,
-			e->da);
-		return ERR_PTR(-EINVAL);
-	}
-
-	cr = kmalloc(sizeof(*cr), GFP_KERNEL);
-	if (!cr)
-		return ERR_PTR(-ENOMEM);
-
-	cr->cam = (e->da & MMU_CAM_VATAG_MASK) | e->prsvd | e->pgsz | e->valid;
-	cr->ram = e->pa | e->endian | e->elsz | e->mixed;
-
-	return cr;
-}
-
-static inline int omap2_cr_valid(struct cr_regs *cr)
-{
-	return cr->cam & MMU_CAM_V;
-}
-
-static u32 omap2_get_pte_attr(struct iotlb_entry *e)
-{
-	u32 attr;
-
-	attr = e->mixed << 5;
-	attr |= e->endian;
-	attr |= e->elsz >> 3;
-	attr <<= (((e->pgsz == MMU_CAM_PGSZ_4K) ||
-			(e->pgsz == MMU_CAM_PGSZ_64K)) ? 0 : 6);
-	return attr;
-}
-
-static ssize_t
-omap2_dump_cr(struct omap_iommu *obj, struct cr_regs *cr, char *buf)
-{
-	char *p = buf;
-
-	/* FIXME: Need more detail analysis of cam/ram */
-	p += sprintf(p, "%08x %08x %01x\n", cr->cam, cr->ram,
-					(cr->cam & MMU_CAM_P) ? 1 : 0);
-
-	return p - buf;
-}
-
-#define pr_reg(name)							\
-	do {								\
-		ssize_t bytes;						\
-		const char *str = "%20s: %08x\n";			\
-		const int maxcol = 32;					\
-		bytes = snprintf(p, maxcol, str, __stringify(name),	\
-				 iommu_read_reg(obj, MMU_##name));	\
-		p += bytes;						\
-		len -= bytes;						\
-		if (len < maxcol)					\
-			goto out;					\
-	} while (0)
-
-static ssize_t
-omap2_iommu_dump_ctx(struct omap_iommu *obj, char *buf, ssize_t len)
-{
-	char *p = buf;
-
-	pr_reg(REVISION);
-	pr_reg(IRQSTATUS);
-	pr_reg(IRQENABLE);
-	pr_reg(WALKING_ST);
-	pr_reg(CNTL);
-	pr_reg(FAULT_AD);
-	pr_reg(TTB);
-	pr_reg(LOCK);
-	pr_reg(LD_TLB);
-	pr_reg(CAM);
-	pr_reg(RAM);
-	pr_reg(GFLUSH);
-	pr_reg(FLUSH_ENTRY);
-	pr_reg(READ_CAM);
-	pr_reg(READ_RAM);
-	pr_reg(EMU_FAULT_AD);
-out:
-	return p - buf;
-}
-
-static void omap2_iommu_save_ctx(struct omap_iommu *obj)
-{
-	int i;
-	u32 *p = obj->ctx;
-
-	for (i = 0; i < (MMU_REG_SIZE / sizeof(u32)); i++) {
-		p[i] = iommu_read_reg(obj, i * sizeof(u32));
-		dev_dbg(obj->dev, "%s\t[%02d] %08x\n", __func__, i, p[i]);
-	}
-
-	BUG_ON(p[0] != IOMMU_ARCH_VERSION);
-}
-
-static void omap2_iommu_restore_ctx(struct omap_iommu *obj)
-{
-	int i;
-	u32 *p = obj->ctx;
-
-	for (i = 0; i < (MMU_REG_SIZE / sizeof(u32)); i++) {
-		iommu_write_reg(obj, p[i], i * sizeof(u32));
-		dev_dbg(obj->dev, "%s\t[%02d] %08x\n", __func__, i, p[i]);
-	}
-
-	BUG_ON(p[0] != IOMMU_ARCH_VERSION);
-}
-
-static void omap2_cr_to_e(struct cr_regs *cr, struct iotlb_entry *e)
-{
-	e->da		= cr->cam & MMU_CAM_VATAG_MASK;
-	e->pa		= cr->ram & MMU_RAM_PADDR_MASK;
-	e->valid	= cr->cam & MMU_CAM_V;
-	e->pgsz		= cr->cam & MMU_CAM_PGSZ_MASK;
-	e->endian	= cr->ram & MMU_RAM_ENDIAN_MASK;
-	e->elsz		= cr->ram & MMU_RAM_ELSZ_MASK;
-	e->mixed	= cr->ram & MMU_RAM_MIXED;
-}
-
-static const struct iommu_functions omap2_iommu_ops = {
-	.version	= IOMMU_ARCH_VERSION,
-
-	.enable		= omap2_iommu_enable,
-	.disable	= omap2_iommu_disable,
-	.set_twl	= omap2_iommu_set_twl,
-	.fault_isr	= omap2_iommu_fault_isr,
-
-	.tlb_read_cr	= omap2_tlb_read_cr,
-	.tlb_load_cr	= omap2_tlb_load_cr,
-
-	.cr_to_e	= omap2_cr_to_e,
-	.cr_to_virt	= omap2_cr_to_virt,
-	.alloc_cr	= omap2_alloc_cr,
-	.cr_valid	= omap2_cr_valid,
-	.dump_cr	= omap2_dump_cr,
-
-	.get_pte_attr	= omap2_get_pte_attr,
-
-	.save_ctx	= omap2_iommu_save_ctx,
-	.restore_ctx	= omap2_iommu_restore_ctx,
-	.dump_ctx	= omap2_iommu_dump_ctx,
-};
-
-static int __init omap2_iommu_init(void)
-{
-	return omap_install_iommu_arch(&omap2_iommu_ops);
-}
-module_init(omap2_iommu_init);
-
-static void __exit omap2_iommu_exit(void)
-{
-	omap_uninstall_iommu_arch(&omap2_iommu_ops);
-}
-module_exit(omap2_iommu_exit);
-
-MODULE_AUTHOR("Hiroshi DOYU, Paul Mundt and Toshihiro Kobayashi");
-MODULE_DESCRIPTION("omap iommu: omap2/3 architecture specific functions");
-MODULE_LICENSE("GPL v2");
diff --git a/drivers/iommu/rockchip-iommu.c b/drivers/iommu/rockchip-iommu.c
new file mode 100644
index 000000000000..6a8b1ec4a48a
--- /dev/null
+++ b/drivers/iommu/rockchip-iommu.c
@@ -0,0 +1,1037 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <asm/cacheflush.h>
+#include <asm/pgtable.h>
+#include <linux/compiler.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <linux/errno.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/iommu.h>
+#include <linux/jiffies.h>
+#include <linux/list.h>
+#include <linux/mm.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_platform.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+
+/** MMU register offsets */
+#define RK_MMU_DTE_ADDR		0x00	/* Directory table address */
+#define RK_MMU_STATUS		0x04
+#define RK_MMU_COMMAND		0x08
+#define RK_MMU_PAGE_FAULT_ADDR	0x0C	/* IOVA of last page fault */
+#define RK_MMU_ZAP_ONE_LINE	0x10	/* Shootdown one IOTLB entry */
+#define RK_MMU_INT_RAWSTAT	0x14	/* IRQ status ignoring mask */
+#define RK_MMU_INT_CLEAR	0x18	/* Acknowledge and re-arm irq */
+#define RK_MMU_INT_MASK		0x1C	/* IRQ enable */
+#define RK_MMU_INT_STATUS	0x20	/* IRQ status after masking */
+#define RK_MMU_AUTO_GATING	0x24
+
+#define DTE_ADDR_DUMMY		0xCAFEBABE
+#define FORCE_RESET_TIMEOUT	100	/* ms */
+
+/* RK_MMU_STATUS fields */
+#define RK_MMU_STATUS_PAGING_ENABLED       BIT(0)
+#define RK_MMU_STATUS_PAGE_FAULT_ACTIVE    BIT(1)
+#define RK_MMU_STATUS_STALL_ACTIVE         BIT(2)
+#define RK_MMU_STATUS_IDLE                 BIT(3)
+#define RK_MMU_STATUS_REPLAY_BUFFER_EMPTY  BIT(4)
+#define RK_MMU_STATUS_PAGE_FAULT_IS_WRITE  BIT(5)
+#define RK_MMU_STATUS_STALL_NOT_ACTIVE     BIT(31)
+
+/* RK_MMU_COMMAND command values */
+#define RK_MMU_CMD_ENABLE_PAGING    0  /* Enable memory translation */
+#define RK_MMU_CMD_DISABLE_PAGING   1  /* Disable memory translation */
+#define RK_MMU_CMD_ENABLE_STALL     2  /* Stall paging to allow other cmds */
+#define RK_MMU_CMD_DISABLE_STALL    3  /* Stop stall re-enables paging */
+#define RK_MMU_CMD_ZAP_CACHE        4  /* Shoot down entire IOTLB */
+#define RK_MMU_CMD_PAGE_FAULT_DONE  5  /* Clear page fault */
+#define RK_MMU_CMD_FORCE_RESET      6  /* Reset all registers */
+
+/* RK_MMU_INT_* register fields */
+#define RK_MMU_IRQ_PAGE_FAULT    0x01  /* page fault */
+#define RK_MMU_IRQ_BUS_ERROR     0x02  /* bus read error */
+#define RK_MMU_IRQ_MASK          (RK_MMU_IRQ_PAGE_FAULT | RK_MMU_IRQ_BUS_ERROR)
+
+#define NUM_DT_ENTRIES 1024
+#define NUM_PT_ENTRIES 1024
+
+#define SPAGE_ORDER 12
+#define SPAGE_SIZE (1 << SPAGE_ORDER)
+
+ /*
+  * Support mapping any size that fits in one page table:
+  *   4 KiB to 4 MiB
+  */
+#define RK_IOMMU_PGSIZE_BITMAP 0x007ff000
+
+#define IOMMU_REG_POLL_COUNT_FAST 1000
+
+struct rk_iommu_domain {
+	struct list_head iommus;
+	u32 *dt; /* page directory table */
+	spinlock_t iommus_lock; /* lock for iommus list */
+	spinlock_t dt_lock; /* lock for modifying page directory table */
+};
+
+struct rk_iommu {
+	struct device *dev;
+	void __iomem *base;
+	int irq;
+	struct list_head node; /* entry in rk_iommu_domain.iommus */
+	struct iommu_domain *domain; /* domain to which iommu is attached */
+};
+
+static inline void rk_table_flush(u32 *va, unsigned int count)
+{
+	phys_addr_t pa_start = virt_to_phys(va);
+	phys_addr_t pa_end = virt_to_phys(va + count);
+	size_t size = pa_end - pa_start;
+
+	__cpuc_flush_dcache_area(va, size);
+	outer_flush_range(pa_start, pa_end);
+}
+
+/**
+ * Inspired by _wait_for in intel_drv.h
+ * This is NOT safe for use in interrupt context.
+ *
+ * Note that it's important that we check the condition again after having
+ * timed out, since the timeout could be due to preemption or similar and
+ * we've never had a chance to check the condition before the timeout.
+ */
+#define rk_wait_for(COND, MS) ({ \
+	unsigned long timeout__ = jiffies + msecs_to_jiffies(MS) + 1;	\
+	int ret__ = 0;							\
+	while (!(COND)) {						\
+		if (time_after(jiffies, timeout__)) {			\
+			ret__ = (COND) ? 0 : -ETIMEDOUT;		\
+			break;						\
+		}							\
+		usleep_range(50, 100);					\
+	}								\
+	ret__;								\
+})
+
+/*
+ * The Rockchip rk3288 iommu uses a 2-level page table.
+ * The first level is the "Directory Table" (DT).
+ * The DT consists of 1024 4-byte Directory Table Entries (DTEs), each pointing
+ * to a "Page Table".
+ * The second level is the 1024 Page Tables (PT).
+ * Each PT consists of 1024 4-byte Page Table Entries (PTEs), each pointing to
+ * a 4 KB page of physical memory.
+ *
+ * The DT and each PT fits in a single 4 KB page (4-bytes * 1024 entries).
+ * Each iommu device has a MMU_DTE_ADDR register that contains the physical
+ * address of the start of the DT page.
+ *
+ * The structure of the page table is as follows:
+ *
+ *                   DT
+ * MMU_DTE_ADDR -> +-----+
+ *                 |     |
+ *                 +-----+     PT
+ *                 | DTE | -> +-----+
+ *                 +-----+    |     |     Memory
+ *                 |     |    +-----+     Page
+ *                 |     |    | PTE | -> +-----+
+ *                 +-----+    +-----+    |     |
+ *                            |     |    |     |
+ *                            |     |    |     |
+ *                            +-----+    |     |
+ *                                       |     |
+ *                                       |     |
+ *                                       +-----+
+ */
+
+/*
+ * Each DTE has a PT address and a valid bit:
+ * +---------------------+-----------+-+
+ * | PT address          | Reserved  |V|
+ * +---------------------+-----------+-+
+ *  31:12 - PT address (PTs always starts on a 4 KB boundary)
+ *  11: 1 - Reserved
+ *      0 - 1 if PT @ PT address is valid
+ */
+#define RK_DTE_PT_ADDRESS_MASK    0xfffff000
+#define RK_DTE_PT_VALID           BIT(0)
+
+static inline phys_addr_t rk_dte_pt_address(u32 dte)
+{
+	return (phys_addr_t)dte & RK_DTE_PT_ADDRESS_MASK;
+}
+
+static inline bool rk_dte_is_pt_valid(u32 dte)
+{
+	return dte & RK_DTE_PT_VALID;
+}
+
+static u32 rk_mk_dte(u32 *pt)
+{
+	phys_addr_t pt_phys = virt_to_phys(pt);
+	return (pt_phys & RK_DTE_PT_ADDRESS_MASK) | RK_DTE_PT_VALID;
+}
+
+/*
+ * Each PTE has a Page address, some flags and a valid bit:
+ * +---------------------+---+-------+-+
+ * | Page address        |Rsv| Flags |V|
+ * +---------------------+---+-------+-+
+ *  31:12 - Page address (Pages always start on a 4 KB boundary)
+ *  11: 9 - Reserved
+ *   8: 1 - Flags
+ *      8 - Read allocate - allocate cache space on read misses
+ *      7 - Read cache - enable cache & prefetch of data
+ *      6 - Write buffer - enable delaying writes on their way to memory
+ *      5 - Write allocate - allocate cache space on write misses
+ *      4 - Write cache - different writes can be merged together
+ *      3 - Override cache attributes
+ *          if 1, bits 4-8 control cache attributes
+ *          if 0, the system bus defaults are used
+ *      2 - Writable
+ *      1 - Readable
+ *      0 - 1 if Page @ Page address is valid
+ */
+#define RK_PTE_PAGE_ADDRESS_MASK  0xfffff000
+#define RK_PTE_PAGE_FLAGS_MASK    0x000001fe
+#define RK_PTE_PAGE_WRITABLE      BIT(2)
+#define RK_PTE_PAGE_READABLE      BIT(1)
+#define RK_PTE_PAGE_VALID         BIT(0)
+
+static inline phys_addr_t rk_pte_page_address(u32 pte)
+{
+	return (phys_addr_t)pte & RK_PTE_PAGE_ADDRESS_MASK;
+}
+
+static inline bool rk_pte_is_page_valid(u32 pte)
+{
+	return pte & RK_PTE_PAGE_VALID;
+}
+
+/* TODO: set cache flags per prot IOMMU_CACHE */
+static u32 rk_mk_pte(phys_addr_t page, int prot)
+{
+	u32 flags = 0;
+	flags |= (prot & IOMMU_READ) ? RK_PTE_PAGE_READABLE : 0;
+	flags |= (prot & IOMMU_WRITE) ? RK_PTE_PAGE_WRITABLE : 0;
+	page &= RK_PTE_PAGE_ADDRESS_MASK;
+	return page | flags | RK_PTE_PAGE_VALID;
+}
+
+static u32 rk_mk_pte_invalid(u32 pte)
+{
+	return pte & ~RK_PTE_PAGE_VALID;
+}
+
+/*
+ * rk3288 iova (IOMMU Virtual Address) format
+ *  31       22.21       12.11          0
+ * +-----------+-----------+-------------+
+ * | DTE index | PTE index | Page offset |
+ * +-----------+-----------+-------------+
+ *  31:22 - DTE index   - index of DTE in DT
+ *  21:12 - PTE index   - index of PTE in PT @ DTE.pt_address
+ *  11: 0 - Page offset - offset into page @ PTE.page_address
+ */
+#define RK_IOVA_DTE_MASK    0xffc00000
+#define RK_IOVA_DTE_SHIFT   22
+#define RK_IOVA_PTE_MASK    0x003ff000
+#define RK_IOVA_PTE_SHIFT   12
+#define RK_IOVA_PAGE_MASK   0x00000fff
+#define RK_IOVA_PAGE_SHIFT  0
+
+static u32 rk_iova_dte_index(dma_addr_t iova)
+{
+	return (u32)(iova & RK_IOVA_DTE_MASK) >> RK_IOVA_DTE_SHIFT;
+}
+
+static u32 rk_iova_pte_index(dma_addr_t iova)
+{
+	return (u32)(iova & RK_IOVA_PTE_MASK) >> RK_IOVA_PTE_SHIFT;
+}
+
+static u32 rk_iova_page_offset(dma_addr_t iova)
+{
+	return (u32)(iova & RK_IOVA_PAGE_MASK) >> RK_IOVA_PAGE_SHIFT;
+}
+
+static u32 rk_iommu_read(struct rk_iommu *iommu, u32 offset)
+{
+	return readl(iommu->base + offset);
+}
+
+static void rk_iommu_write(struct rk_iommu *iommu, u32 offset, u32 value)
+{
+	writel(value, iommu->base + offset);
+}
+
+static void rk_iommu_command(struct rk_iommu *iommu, u32 command)
+{
+	writel(command, iommu->base + RK_MMU_COMMAND);
+}
+
+static void rk_iommu_zap_lines(struct rk_iommu *iommu, dma_addr_t iova,
+			       size_t size)
+{
+	dma_addr_t iova_end = iova + size;
+	/*
+	 * TODO(djkurtz): Figure out when it is more efficient to shootdown the
+	 * entire iotlb rather than iterate over individual iovas.
+	 */
+	for (; iova < iova_end; iova += SPAGE_SIZE)
+		rk_iommu_write(iommu, RK_MMU_ZAP_ONE_LINE, iova);
+}
+
+static bool rk_iommu_is_stall_active(struct rk_iommu *iommu)
+{
+	return rk_iommu_read(iommu, RK_MMU_STATUS) & RK_MMU_STATUS_STALL_ACTIVE;
+}
+
+static bool rk_iommu_is_paging_enabled(struct rk_iommu *iommu)
+{
+	return rk_iommu_read(iommu, RK_MMU_STATUS) &
+			     RK_MMU_STATUS_PAGING_ENABLED;
+}
+
+static int rk_iommu_enable_stall(struct rk_iommu *iommu)
+{
+	int ret;
+
+	if (rk_iommu_is_stall_active(iommu))
+		return 0;
+
+	/* Stall can only be enabled if paging is enabled */
+	if (!rk_iommu_is_paging_enabled(iommu))
+		return 0;
+
+	rk_iommu_command(iommu, RK_MMU_CMD_ENABLE_STALL);
+
+	ret = rk_wait_for(rk_iommu_is_stall_active(iommu), 1);
+	if (ret)
+		dev_err(iommu->dev, "Enable stall request timed out, status: %#08x\n",
+			rk_iommu_read(iommu, RK_MMU_STATUS));
+
+	return ret;
+}
+
+static int rk_iommu_disable_stall(struct rk_iommu *iommu)
+{
+	int ret;
+
+	if (!rk_iommu_is_stall_active(iommu))
+		return 0;
+
+	rk_iommu_command(iommu, RK_MMU_CMD_DISABLE_STALL);
+
+	ret = rk_wait_for(!rk_iommu_is_stall_active(iommu), 1);
+	if (ret)
+		dev_err(iommu->dev, "Disable stall request timed out, status: %#08x\n",
+			rk_iommu_read(iommu, RK_MMU_STATUS));
+
+	return ret;
+}
+
+static int rk_iommu_enable_paging(struct rk_iommu *iommu)
+{
+	int ret;
+
+	if (rk_iommu_is_paging_enabled(iommu))
+		return 0;
+
+	rk_iommu_command(iommu, RK_MMU_CMD_ENABLE_PAGING);
+
+	ret = rk_wait_for(rk_iommu_is_paging_enabled(iommu), 1);
+	if (ret)
+		dev_err(iommu->dev, "Enable paging request timed out, status: %#08x\n",
+			rk_iommu_read(iommu, RK_MMU_STATUS));
+
+	return ret;
+}
+
+static int rk_iommu_disable_paging(struct rk_iommu *iommu)
+{
+	int ret;
+
+	if (!rk_iommu_is_paging_enabled(iommu))
+		return 0;
+
+	rk_iommu_command(iommu, RK_MMU_CMD_DISABLE_PAGING);
+
+	ret = rk_wait_for(!rk_iommu_is_paging_enabled(iommu), 1);
+	if (ret)
+		dev_err(iommu->dev, "Disable paging request timed out, status: %#08x\n",
+			rk_iommu_read(iommu, RK_MMU_STATUS));
+
+	return ret;
+}
+
+static int rk_iommu_force_reset(struct rk_iommu *iommu)
+{
+	int ret;
+	u32 dte_addr;
+
+	/*
+	 * Check if register DTE_ADDR is working by writing DTE_ADDR_DUMMY
+	 * and verifying that upper 5 nybbles are read back.
+	 */
+	rk_iommu_write(iommu, RK_MMU_DTE_ADDR, DTE_ADDR_DUMMY);
+
+	dte_addr = rk_iommu_read(iommu, RK_MMU_DTE_ADDR);
+	if (dte_addr != (DTE_ADDR_DUMMY & RK_DTE_PT_ADDRESS_MASK)) {
+		dev_err(iommu->dev, "Error during raw reset. MMU_DTE_ADDR is not functioning\n");
+		return -EFAULT;
+	}
+
+	rk_iommu_command(iommu, RK_MMU_CMD_FORCE_RESET);
+
+	ret = rk_wait_for(rk_iommu_read(iommu, RK_MMU_DTE_ADDR) == 0x00000000,
+			  FORCE_RESET_TIMEOUT);
+	if (ret)
+		dev_err(iommu->dev, "FORCE_RESET command timed out\n");
+
+	return ret;
+}
+
+static void log_iova(struct rk_iommu *iommu, dma_addr_t iova)
+{
+	u32 dte_index, pte_index, page_offset;
+	u32 mmu_dte_addr;
+	phys_addr_t mmu_dte_addr_phys, dte_addr_phys;
+	u32 *dte_addr;
+	u32 dte;
+	phys_addr_t pte_addr_phys = 0;
+	u32 *pte_addr = NULL;
+	u32 pte = 0;
+	phys_addr_t page_addr_phys = 0;
+	u32 page_flags = 0;
+
+	dte_index = rk_iova_dte_index(iova);
+	pte_index = rk_iova_pte_index(iova);
+	page_offset = rk_iova_page_offset(iova);
+
+	mmu_dte_addr = rk_iommu_read(iommu, RK_MMU_DTE_ADDR);
+	mmu_dte_addr_phys = (phys_addr_t)mmu_dte_addr;
+
+	dte_addr_phys = mmu_dte_addr_phys + (4 * dte_index);
+	dte_addr = phys_to_virt(dte_addr_phys);
+	dte = *dte_addr;
+
+	if (!rk_dte_is_pt_valid(dte))
+		goto print_it;
+
+	pte_addr_phys = rk_dte_pt_address(dte) + (pte_index * 4);
+	pte_addr = phys_to_virt(pte_addr_phys);
+	pte = *pte_addr;
+
+	if (!rk_pte_is_page_valid(pte))
+		goto print_it;
+
+	page_addr_phys = rk_pte_page_address(pte) + page_offset;
+	page_flags = pte & RK_PTE_PAGE_FLAGS_MASK;
+
+print_it:
+	dev_err(iommu->dev, "iova = %pad: dte_index: %#03x pte_index: %#03x page_offset: %#03x\n",
+		&iova, dte_index, pte_index, page_offset);
+	dev_err(iommu->dev, "mmu_dte_addr: %pa dte@%pa: %#08x valid: %u pte@%pa: %#08x valid: %u page@%pa flags: %#03x\n",
+		&mmu_dte_addr_phys, &dte_addr_phys, dte,
+		rk_dte_is_pt_valid(dte), &pte_addr_phys, pte,
+		rk_pte_is_page_valid(pte), &page_addr_phys, page_flags);
+}
+
+static irqreturn_t rk_iommu_irq(int irq, void *dev_id)
+{
+	struct rk_iommu *iommu = dev_id;
+	u32 status;
+	u32 int_status;
+	dma_addr_t iova;
+
+	int_status = rk_iommu_read(iommu, RK_MMU_INT_STATUS);
+	if (int_status == 0)
+		return IRQ_NONE;
+
+	iova = rk_iommu_read(iommu, RK_MMU_PAGE_FAULT_ADDR);
+
+	if (int_status & RK_MMU_IRQ_PAGE_FAULT) {
+		int flags;
+
+		status = rk_iommu_read(iommu, RK_MMU_STATUS);
+		flags = (status & RK_MMU_STATUS_PAGE_FAULT_IS_WRITE) ?
+				IOMMU_FAULT_WRITE : IOMMU_FAULT_READ;
+
+		dev_err(iommu->dev, "Page fault at %pad of type %s\n",
+			&iova,
+			(flags == IOMMU_FAULT_WRITE) ? "write" : "read");
+
+		log_iova(iommu, iova);
+
+		/*
+		 * Report page fault to any installed handlers.
+		 * Ignore the return code, though, since we always zap cache
+		 * and clear the page fault anyway.
+		 */
+		if (iommu->domain)
+			report_iommu_fault(iommu->domain, iommu->dev, iova,
+					   flags);
+		else
+			dev_err(iommu->dev, "Page fault while iommu not attached to domain?\n");
+
+		rk_iommu_command(iommu, RK_MMU_CMD_ZAP_CACHE);
+		rk_iommu_command(iommu, RK_MMU_CMD_PAGE_FAULT_DONE);
+	}
+
+	if (int_status & RK_MMU_IRQ_BUS_ERROR)
+		dev_err(iommu->dev, "BUS_ERROR occurred at %pad\n", &iova);
+
+	if (int_status & ~RK_MMU_IRQ_MASK)
+		dev_err(iommu->dev, "unexpected int_status: %#08x\n",
+			int_status);
+
+	rk_iommu_write(iommu, RK_MMU_INT_CLEAR, int_status);
+
+	return IRQ_HANDLED;
+}
+
+static phys_addr_t rk_iommu_iova_to_phys(struct iommu_domain *domain,
+					 dma_addr_t iova)
+{
+	struct rk_iommu_domain *rk_domain = domain->priv;
+	unsigned long flags;
+	phys_addr_t pt_phys, phys = 0;
+	u32 dte, pte;
+	u32 *page_table;
+
+	spin_lock_irqsave(&rk_domain->dt_lock, flags);
+
+	dte = rk_domain->dt[rk_iova_dte_index(iova)];
+	if (!rk_dte_is_pt_valid(dte))
+		goto out;
+
+	pt_phys = rk_dte_pt_address(dte);
+	page_table = (u32 *)phys_to_virt(pt_phys);
+	pte = page_table[rk_iova_pte_index(iova)];
+	if (!rk_pte_is_page_valid(pte))
+		goto out;
+
+	phys = rk_pte_page_address(pte) + rk_iova_page_offset(iova);
+out:
+	spin_unlock_irqrestore(&rk_domain->dt_lock, flags);
+
+	return phys;
+}
+
+static void rk_iommu_zap_iova(struct rk_iommu_domain *rk_domain,
+			      dma_addr_t iova, size_t size)
+{
+	struct list_head *pos;
+	unsigned long flags;
+
+	/* shootdown these iova from all iommus using this domain */
+	spin_lock_irqsave(&rk_domain->iommus_lock, flags);
+	list_for_each(pos, &rk_domain->iommus) {
+		struct rk_iommu *iommu;
+		iommu = list_entry(pos, struct rk_iommu, node);
+		rk_iommu_zap_lines(iommu, iova, size);
+	}
+	spin_unlock_irqrestore(&rk_domain->iommus_lock, flags);
+}
+
+static u32 *rk_dte_get_page_table(struct rk_iommu_domain *rk_domain,
+				  dma_addr_t iova)
+{
+	u32 *page_table, *dte_addr;
+	u32 dte;
+	phys_addr_t pt_phys;
+
+	assert_spin_locked(&rk_domain->dt_lock);
+
+	dte_addr = &rk_domain->dt[rk_iova_dte_index(iova)];
+	dte = *dte_addr;
+	if (rk_dte_is_pt_valid(dte))
+		goto done;
+
+	page_table = (u32 *)get_zeroed_page(GFP_ATOMIC | GFP_DMA32);
+	if (!page_table)
+		return ERR_PTR(-ENOMEM);
+
+	dte = rk_mk_dte(page_table);
+	*dte_addr = dte;
+
+	rk_table_flush(page_table, NUM_PT_ENTRIES);
+	rk_table_flush(dte_addr, 1);
+
+	/*
+	 * Zap the first iova of newly allocated page table so iommu evicts
+	 * old cached value of new dte from the iotlb.
+	 */
+	rk_iommu_zap_iova(rk_domain, iova, SPAGE_SIZE);
+
+done:
+	pt_phys = rk_dte_pt_address(dte);
+	return (u32 *)phys_to_virt(pt_phys);
+}
+
+static size_t rk_iommu_unmap_iova(struct rk_iommu_domain *rk_domain,
+				  u32 *pte_addr, dma_addr_t iova, size_t size)
+{
+	unsigned int pte_count;
+	unsigned int pte_total = size / SPAGE_SIZE;
+
+	assert_spin_locked(&rk_domain->dt_lock);
+
+	for (pte_count = 0; pte_count < pte_total; pte_count++) {
+		u32 pte = pte_addr[pte_count];
+		if (!rk_pte_is_page_valid(pte))
+			break;
+
+		pte_addr[pte_count] = rk_mk_pte_invalid(pte);
+	}
+
+	rk_table_flush(pte_addr, pte_count);
+
+	return pte_count * SPAGE_SIZE;
+}
+
+static int rk_iommu_map_iova(struct rk_iommu_domain *rk_domain, u32 *pte_addr,
+			     dma_addr_t iova, phys_addr_t paddr, size_t size,
+			     int prot)
+{
+	unsigned int pte_count;
+	unsigned int pte_total = size / SPAGE_SIZE;
+	phys_addr_t page_phys;
+
+	assert_spin_locked(&rk_domain->dt_lock);
+
+	for (pte_count = 0; pte_count < pte_total; pte_count++) {
+		u32 pte = pte_addr[pte_count];
+
+		if (rk_pte_is_page_valid(pte))
+			goto unwind;
+
+		pte_addr[pte_count] = rk_mk_pte(paddr, prot);
+
+		paddr += SPAGE_SIZE;
+	}
+
+	rk_table_flush(pte_addr, pte_count);
+
+	return 0;
+unwind:
+	/* Unmap the range of iovas that we just mapped */
+	rk_iommu_unmap_iova(rk_domain, pte_addr, iova, pte_count * SPAGE_SIZE);
+
+	iova += pte_count * SPAGE_SIZE;
+	page_phys = rk_pte_page_address(pte_addr[pte_count]);
+	pr_err("iova: %pad already mapped to %pa cannot remap to phys: %pa prot: %#x\n",
+	       &iova, &page_phys, &paddr, prot);
+
+	return -EADDRINUSE;
+}
+
+static int rk_iommu_map(struct iommu_domain *domain, unsigned long _iova,
+			phys_addr_t paddr, size_t size, int prot)
+{
+	struct rk_iommu_domain *rk_domain = domain->priv;
+	unsigned long flags;
+	dma_addr_t iova = (dma_addr_t)_iova;
+	u32 *page_table, *pte_addr;
+	int ret;
+
+	spin_lock_irqsave(&rk_domain->dt_lock, flags);
+
+	/*
+	 * pgsize_bitmap specifies iova sizes that fit in one page table
+	 * (1024 4-KiB pages = 4 MiB).
+	 * So, size will always be 4096 <= size <= 4194304.
+	 * Since iommu_map() guarantees that both iova and size will be
+	 * aligned, we will always only be mapping from a single dte here.
+	 */
+	page_table = rk_dte_get_page_table(rk_domain, iova);
+	if (IS_ERR(page_table)) {
+		spin_unlock_irqrestore(&rk_domain->dt_lock, flags);
+		return PTR_ERR(page_table);
+	}
+
+	pte_addr = &page_table[rk_iova_pte_index(iova)];
+	ret = rk_iommu_map_iova(rk_domain, pte_addr, iova, paddr, size, prot);
+	spin_unlock_irqrestore(&rk_domain->dt_lock, flags);
+
+	return ret;
+}
+
+static size_t rk_iommu_unmap(struct iommu_domain *domain, unsigned long _iova,
+			     size_t size)
+{
+	struct rk_iommu_domain *rk_domain = domain->priv;
+	unsigned long flags;
+	dma_addr_t iova = (dma_addr_t)_iova;
+	phys_addr_t pt_phys;
+	u32 dte;
+	u32 *pte_addr;
+	size_t unmap_size;
+
+	spin_lock_irqsave(&rk_domain->dt_lock, flags);
+
+	/*
+	 * pgsize_bitmap specifies iova sizes that fit in one page table
+	 * (1024 4-KiB pages = 4 MiB).
+	 * So, size will always be 4096 <= size <= 4194304.
+	 * Since iommu_unmap() guarantees that both iova and size will be
+	 * aligned, we will always only be unmapping from a single dte here.
+	 */
+	dte = rk_domain->dt[rk_iova_dte_index(iova)];
+	/* Just return 0 if iova is unmapped */
+	if (!rk_dte_is_pt_valid(dte)) {
+		spin_unlock_irqrestore(&rk_domain->dt_lock, flags);
+		return 0;
+	}
+
+	pt_phys = rk_dte_pt_address(dte);
+	pte_addr = (u32 *)phys_to_virt(pt_phys) + rk_iova_pte_index(iova);
+	unmap_size = rk_iommu_unmap_iova(rk_domain, pte_addr, iova, size);
+
+	spin_unlock_irqrestore(&rk_domain->dt_lock, flags);
+
+	/* Shootdown iotlb entries for iova range that was just unmapped */
+	rk_iommu_zap_iova(rk_domain, iova, unmap_size);
+
+	return unmap_size;
+}
+
+static struct rk_iommu *rk_iommu_from_dev(struct device *dev)
+{
+	struct iommu_group *group;
+	struct device *iommu_dev;
+	struct rk_iommu *rk_iommu;
+
+	group = iommu_group_get(dev);
+	if (!group)
+		return NULL;
+	iommu_dev = iommu_group_get_iommudata(group);
+	rk_iommu = dev_get_drvdata(iommu_dev);
+	iommu_group_put(group);
+
+	return rk_iommu;
+}
+
+static int rk_iommu_attach_device(struct iommu_domain *domain,
+				  struct device *dev)
+{
+	struct rk_iommu *iommu;
+	struct rk_iommu_domain *rk_domain = domain->priv;
+	unsigned long flags;
+	int ret;
+	phys_addr_t dte_addr;
+
+	/*
+	 * Allow 'virtual devices' (e.g., drm) to attach to domain.
+	 * Such a device does not belong to an iommu group.
+	 */
+	iommu = rk_iommu_from_dev(dev);
+	if (!iommu)
+		return 0;
+
+	ret = rk_iommu_enable_stall(iommu);
+	if (ret)
+		return ret;
+
+	ret = rk_iommu_force_reset(iommu);
+	if (ret)
+		return ret;
+
+	iommu->domain = domain;
+
+	ret = devm_request_irq(dev, iommu->irq, rk_iommu_irq,
+			       IRQF_SHARED, dev_name(dev), iommu);
+	if (ret)
+		return ret;
+
+	dte_addr = virt_to_phys(rk_domain->dt);
+	rk_iommu_write(iommu, RK_MMU_DTE_ADDR, dte_addr);
+	rk_iommu_command(iommu, RK_MMU_CMD_ZAP_CACHE);
+	rk_iommu_write(iommu, RK_MMU_INT_MASK, RK_MMU_IRQ_MASK);
+
+	ret = rk_iommu_enable_paging(iommu);
+	if (ret)
+		return ret;
+
+	spin_lock_irqsave(&rk_domain->iommus_lock, flags);
+	list_add_tail(&iommu->node, &rk_domain->iommus);
+	spin_unlock_irqrestore(&rk_domain->iommus_lock, flags);
+
+	dev_info(dev, "Attached to iommu domain\n");
+
+	rk_iommu_disable_stall(iommu);
+
+	return 0;
+}
+
+static void rk_iommu_detach_device(struct iommu_domain *domain,
+				   struct device *dev)
+{
+	struct rk_iommu *iommu;
+	struct rk_iommu_domain *rk_domain = domain->priv;
+	unsigned long flags;
+
+	/* Allow 'virtual devices' (eg drm) to detach from domain */
+	iommu = rk_iommu_from_dev(dev);
+	if (!iommu)
+		return;
+
+	spin_lock_irqsave(&rk_domain->iommus_lock, flags);
+	list_del_init(&iommu->node);
+	spin_unlock_irqrestore(&rk_domain->iommus_lock, flags);
+
+	/* Ignore error while disabling, just keep going */
+	rk_iommu_enable_stall(iommu);
+	rk_iommu_disable_paging(iommu);
+	rk_iommu_write(iommu, RK_MMU_INT_MASK, 0);
+	rk_iommu_write(iommu, RK_MMU_DTE_ADDR, 0);
+	rk_iommu_disable_stall(iommu);
+
+	devm_free_irq(dev, iommu->irq, iommu);
+
+	iommu->domain = NULL;
+
+	dev_info(dev, "Detached from iommu domain\n");
+}
+
+static int rk_iommu_domain_init(struct iommu_domain *domain)
+{
+	struct rk_iommu_domain *rk_domain;
+
+	rk_domain = kzalloc(sizeof(*rk_domain), GFP_KERNEL);
+	if (!rk_domain)
+		return -ENOMEM;
+
+	/*
+	 * rk32xx iommus use a 2 level pagetable.
+	 * Each level1 (dt) and level2 (pt) table has 1024 4-byte entries.
+	 * Allocate one 4 KiB page for each table.
+	 */
+	rk_domain->dt = (u32 *)get_zeroed_page(GFP_KERNEL | GFP_DMA32);
+	if (!rk_domain->dt)
+		goto err_dt;
+
+	rk_table_flush(rk_domain->dt, NUM_DT_ENTRIES);
+
+	spin_lock_init(&rk_domain->iommus_lock);
+	spin_lock_init(&rk_domain->dt_lock);
+	INIT_LIST_HEAD(&rk_domain->iommus);
+
+	domain->priv = rk_domain;
+
+	return 0;
+err_dt:
+	kfree(rk_domain);
+	return -ENOMEM;
+}
+
+static void rk_iommu_domain_destroy(struct iommu_domain *domain)
+{
+	struct rk_iommu_domain *rk_domain = domain->priv;
+	int i;
+
+	WARN_ON(!list_empty(&rk_domain->iommus));
+
+	for (i = 0; i < NUM_DT_ENTRIES; i++) {
+		u32 dte = rk_domain->dt[i];
+		if (rk_dte_is_pt_valid(dte)) {
+			phys_addr_t pt_phys = rk_dte_pt_address(dte);
+			u32 *page_table = phys_to_virt(pt_phys);
+			free_page((unsigned long)page_table);
+		}
+	}
+
+	free_page((unsigned long)rk_domain->dt);
+	kfree(domain->priv);
+	domain->priv = NULL;
+}
+
+static bool rk_iommu_is_dev_iommu_master(struct device *dev)
+{
+	struct device_node *np = dev->of_node;
+	int ret;
+
+	/*
+	 * An iommu master has an iommus property containing a list of phandles
+	 * to iommu nodes, each with an #iommu-cells property with value 0.
+	 */
+	ret = of_count_phandle_with_args(np, "iommus", "#iommu-cells");
+	return (ret > 0);
+}
+
+static int rk_iommu_group_set_iommudata(struct iommu_group *group,
+					struct device *dev)
+{
+	struct device_node *np = dev->of_node;
+	struct platform_device *pd;
+	int ret;
+	struct of_phandle_args args;
+
+	/*
+	 * An iommu master has an iommus property containing a list of phandles
+	 * to iommu nodes, each with an #iommu-cells property with value 0.
+	 */
+	ret = of_parse_phandle_with_args(np, "iommus", "#iommu-cells", 0,
+					 &args);
+	if (ret) {
+		dev_err(dev, "of_parse_phandle_with_args(%s) => %d\n",
+			np->full_name, ret);
+		return ret;
+	}
+	if (args.args_count != 0) {
+		dev_err(dev, "incorrect number of iommu params found for %s (found %d, expected 0)\n",
+			args.np->full_name, args.args_count);
+		return -EINVAL;
+	}
+
+	pd = of_find_device_by_node(args.np);
+	of_node_put(args.np);
+	if (!pd) {
+		dev_err(dev, "iommu %s not found\n", args.np->full_name);
+		return -EPROBE_DEFER;
+	}
+
+	/* TODO(djkurtz): handle multiple slave iommus for a single master */
+	iommu_group_set_iommudata(group, &pd->dev, NULL);
+
+	return 0;
+}
+
+static int rk_iommu_add_device(struct device *dev)
+{
+	struct iommu_group *group;
+	int ret;
+
+	if (!rk_iommu_is_dev_iommu_master(dev))
+		return -ENODEV;
+
+	group = iommu_group_get(dev);
+	if (!group) {
+		group = iommu_group_alloc();
+		if (IS_ERR(group)) {
+			dev_err(dev, "Failed to allocate IOMMU group\n");
+			return PTR_ERR(group);
+		}
+	}
+
+	ret = iommu_group_add_device(group, dev);
+	if (ret)
+		goto err_put_group;
+
+	ret = rk_iommu_group_set_iommudata(group, dev);
+	if (ret)
+		goto err_remove_device;
+
+	iommu_group_put(group);
+
+	return 0;
+
+err_remove_device:
+	iommu_group_remove_device(dev);
+err_put_group:
+	iommu_group_put(group);
+	return ret;
+}
+
+static void rk_iommu_remove_device(struct device *dev)
+{
+	if (!rk_iommu_is_dev_iommu_master(dev))
+		return;
+
+	iommu_group_remove_device(dev);
+}
+
+static const struct iommu_ops rk_iommu_ops = {
+	.domain_init = rk_iommu_domain_init,
+	.domain_destroy = rk_iommu_domain_destroy,
+	.attach_dev = rk_iommu_attach_device,
+	.detach_dev = rk_iommu_detach_device,
+	.map = rk_iommu_map,
+	.unmap = rk_iommu_unmap,
+	.add_device = rk_iommu_add_device,
+	.remove_device = rk_iommu_remove_device,
+	.iova_to_phys = rk_iommu_iova_to_phys,
+	.pgsize_bitmap = RK_IOMMU_PGSIZE_BITMAP,
+};
+
+static int rk_iommu_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct rk_iommu *iommu;
+	struct resource *res;
+
+	iommu = devm_kzalloc(dev, sizeof(*iommu), GFP_KERNEL);
+	if (!iommu)
+		return -ENOMEM;
+
+	platform_set_drvdata(pdev, iommu);
+	iommu->dev = dev;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	iommu->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(iommu->base))
+		return PTR_ERR(iommu->base);
+
+	iommu->irq = platform_get_irq(pdev, 0);
+	if (iommu->irq < 0) {
+		dev_err(dev, "Failed to get IRQ, %d\n", iommu->irq);
+		return -ENXIO;
+	}
+
+	return 0;
+}
+
+static int rk_iommu_remove(struct platform_device *pdev)
+{
+	return 0;
+}
+
+#ifdef CONFIG_OF
+static const struct of_device_id rk_iommu_dt_ids[] = {
+	{ .compatible = "rockchip,iommu" },
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, rk_iommu_dt_ids);
+#endif
+
+static struct platform_driver rk_iommu_driver = {
+	.probe = rk_iommu_probe,
+	.remove = rk_iommu_remove,
+	.driver = {
+		   .name = "rk_iommu",
+		   .of_match_table = of_match_ptr(rk_iommu_dt_ids),
+	},
+};
+
+static int __init rk_iommu_init(void)
+{
+	int ret;
+
+	ret = bus_set_iommu(&platform_bus_type, &rk_iommu_ops);
+	if (ret)
+		return ret;
+
+	return platform_driver_register(&rk_iommu_driver);
+}
+static void __exit rk_iommu_exit(void)
+{
+	platform_driver_unregister(&rk_iommu_driver);
+}
+
+subsys_initcall(rk_iommu_init);
+module_exit(rk_iommu_exit);
+
+MODULE_DESCRIPTION("IOMMU API for Rockchip");
+MODULE_AUTHOR("Simon Xue <xxm@rock-chips.com> and Daniel Kurtz <djkurtz@chromium.org>");
+MODULE_ALIAS("platform:rockchip-iommu");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/iommu/shmobile-iommu.c b/drivers/iommu/shmobile-iommu.c
index 1333e6fb3405..f1b00774e4de 100644
--- a/drivers/iommu/shmobile-iommu.c
+++ b/drivers/iommu/shmobile-iommu.c
@@ -361,6 +361,7 @@ static const struct iommu_ops shmobile_iommu_ops = {
 	.detach_dev = shmobile_iommu_detach_device,
 	.map = shmobile_iommu_map,
 	.unmap = shmobile_iommu_unmap,
+	.map_sg = default_iommu_map_sg,
 	.iova_to_phys = shmobile_iommu_iova_to_phys,
 	.add_device = shmobile_iommu_add_device,
 	.pgsize_bitmap = SZ_1M | SZ_64K | SZ_4K,
diff --git a/drivers/iommu/shmobile-ipmmu.c b/drivers/iommu/shmobile-ipmmu.c
index bd97adecb1fd..951651a9746b 100644
--- a/drivers/iommu/shmobile-ipmmu.c
+++ b/drivers/iommu/shmobile-ipmmu.c
@@ -118,7 +118,6 @@ static int ipmmu_probe(struct platform_device *pdev)
 static struct platform_driver ipmmu_driver = {
 	.probe = ipmmu_probe,
 	.driver = {
-		.owner = THIS_MODULE,
 		.name = "ipmmu",
 	},
 };
diff --git a/drivers/iommu/tegra-gart.c b/drivers/iommu/tegra-gart.c
index b10a8ecede8e..c48da057dbb1 100644
--- a/drivers/iommu/tegra-gart.c
+++ b/drivers/iommu/tegra-gart.c
@@ -303,21 +303,21 @@ static phys_addr_t gart_iommu_iova_to_phys(struct iommu_domain *domain,
 	return pa;
 }
 
-static int gart_iommu_domain_has_cap(struct iommu_domain *domain,
-				     unsigned long cap)
+static bool gart_iommu_capable(enum iommu_cap cap)
 {
-	return 0;
+	return false;
 }
 
 static const struct iommu_ops gart_iommu_ops = {
+	.capable	= gart_iommu_capable,
 	.domain_init	= gart_iommu_domain_init,
 	.domain_destroy	= gart_iommu_domain_destroy,
 	.attach_dev	= gart_iommu_attach_dev,
 	.detach_dev	= gart_iommu_detach_dev,
 	.map		= gart_iommu_map,
+	.map_sg		= default_iommu_map_sg,
 	.unmap		= gart_iommu_unmap,
 	.iova_to_phys	= gart_iommu_iova_to_phys,
-	.domain_has_cap	= gart_iommu_domain_has_cap,
 	.pgsize_bitmap	= GART_IOMMU_PGSIZES,
 };
 
@@ -396,7 +396,7 @@ static int tegra_gart_probe(struct platform_device *pdev)
 	do_gart_setup(gart, NULL);
 
 	gart_handle = gart;
-	bus_set_iommu(&platform_bus_type, &gart_iommu_ops);
+
 	return 0;
 }
 
@@ -416,7 +416,7 @@ static const struct dev_pm_ops tegra_gart_pm_ops = {
 	.resume		= tegra_gart_resume,
 };
 
-static struct of_device_id tegra_gart_of_match[] = {
+static const struct of_device_id tegra_gart_of_match[] = {
 	{ .compatible = "nvidia,tegra20-gart", },
 	{ },
 };
@@ -426,7 +426,6 @@ static struct platform_driver tegra_gart_driver = {
 	.probe		= tegra_gart_probe,
 	.remove		= tegra_gart_remove,
 	.driver = {
-		.owner	= THIS_MODULE,
 		.name	= "tegra-gart",
 		.pm	= &tegra_gart_pm_ops,
 		.of_match_table = tegra_gart_of_match,
diff --git a/drivers/iommu/tegra-smmu.c b/drivers/iommu/tegra-smmu.c
index 3ded3894623c..6e134c7c227f 100644
--- a/drivers/iommu/tegra-smmu.c
+++ b/drivers/iommu/tegra-smmu.c
@@ -1,1296 +1,732 @@
 /*
- * IOMMU API for SMMU in Tegra30
+ * Copyright (C) 2011-2014 NVIDIA CORPORATION.  All rights reserved.
  *
- * Copyright (c) 2011-2013, NVIDIA CORPORATION.  All rights reserved.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms and conditions of the GNU General Public License,
- * version 2, as published by the Free Software Foundation.
- *
- * This program is distributed in the hope it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc.,
- * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
  */
 
-#define pr_fmt(fmt)	"%s(): " fmt, __func__
-
 #include <linux/err.h>
-#include <linux/module.h>
-#include <linux/platform_device.h>
-#include <linux/spinlock.h>
-#include <linux/slab.h>
-#include <linux/vmalloc.h>
-#include <linux/mm.h>
-#include <linux/pagemap.h>
-#include <linux/device.h>
-#include <linux/sched.h>
 #include <linux/iommu.h>
-#include <linux/io.h>
+#include <linux/kernel.h>
 #include <linux/of.h>
-#include <linux/of_iommu.h>
-#include <linux/debugfs.h>
-#include <linux/seq_file.h>
+#include <linux/of_device.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
 
 #include <soc/tegra/ahb.h>
+#include <soc/tegra/mc.h>
 
-#include <asm/page.h>
-#include <asm/cacheflush.h>
-
-enum smmu_hwgrp {
-	HWGRP_AFI,
-	HWGRP_AVPC,
-	HWGRP_DC,
-	HWGRP_DCB,
-	HWGRP_EPP,
-	HWGRP_G2,
-	HWGRP_HC,
-	HWGRP_HDA,
-	HWGRP_ISP,
-	HWGRP_MPE,
-	HWGRP_NV,
-	HWGRP_NV2,
-	HWGRP_PPCS,
-	HWGRP_SATA,
-	HWGRP_VDE,
-	HWGRP_VI,
-
-	HWGRP_COUNT,
-
-	HWGRP_END = ~0,
-};
+struct tegra_smmu {
+	void __iomem *regs;
+	struct device *dev;
 
-#define HWG_AFI		(1 << HWGRP_AFI)
-#define HWG_AVPC	(1 << HWGRP_AVPC)
-#define HWG_DC		(1 << HWGRP_DC)
-#define HWG_DCB		(1 << HWGRP_DCB)
-#define HWG_EPP		(1 << HWGRP_EPP)
-#define HWG_G2		(1 << HWGRP_G2)
-#define HWG_HC		(1 << HWGRP_HC)
-#define HWG_HDA		(1 << HWGRP_HDA)
-#define HWG_ISP		(1 << HWGRP_ISP)
-#define HWG_MPE		(1 << HWGRP_MPE)
-#define HWG_NV		(1 << HWGRP_NV)
-#define HWG_NV2		(1 << HWGRP_NV2)
-#define HWG_PPCS	(1 << HWGRP_PPCS)
-#define HWG_SATA	(1 << HWGRP_SATA)
-#define HWG_VDE		(1 << HWGRP_VDE)
-#define HWG_VI		(1 << HWGRP_VI)
-
-/* bitmap of the page sizes currently supported */
-#define SMMU_IOMMU_PGSIZES	(SZ_4K)
-
-#define SMMU_CONFIG				0x10
-#define SMMU_CONFIG_DISABLE			0
-#define SMMU_CONFIG_ENABLE			1
-
-/* REVISIT: To support multiple MCs */
-enum {
-	_MC = 0,
-};
+	struct tegra_mc *mc;
+	const struct tegra_smmu_soc *soc;
 
-enum {
-	_TLB = 0,
-	_PTC,
-};
+	unsigned long *asids;
+	struct mutex lock;
 
-#define SMMU_CACHE_CONFIG_BASE			0x14
-#define __SMMU_CACHE_CONFIG(mc, cache)		(SMMU_CACHE_CONFIG_BASE + 4 * cache)
-#define SMMU_CACHE_CONFIG(cache)		__SMMU_CACHE_CONFIG(_MC, cache)
-
-#define SMMU_CACHE_CONFIG_STATS_SHIFT		31
-#define SMMU_CACHE_CONFIG_STATS_ENABLE		(1 << SMMU_CACHE_CONFIG_STATS_SHIFT)
-#define SMMU_CACHE_CONFIG_STATS_TEST_SHIFT	30
-#define SMMU_CACHE_CONFIG_STATS_TEST		(1 << SMMU_CACHE_CONFIG_STATS_TEST_SHIFT)
-
-#define SMMU_TLB_CONFIG_HIT_UNDER_MISS__ENABLE	(1 << 29)
-#define SMMU_TLB_CONFIG_ACTIVE_LINES__VALUE	0x10
-#define SMMU_TLB_CONFIG_RESET_VAL		0x20000010
-
-#define SMMU_PTC_CONFIG_CACHE__ENABLE		(1 << 29)
-#define SMMU_PTC_CONFIG_INDEX_MAP__PATTERN	0x3f
-#define SMMU_PTC_CONFIG_RESET_VAL		0x2000003f
-
-#define SMMU_PTB_ASID				0x1c
-#define SMMU_PTB_ASID_CURRENT_SHIFT		0
-
-#define SMMU_PTB_DATA				0x20
-#define SMMU_PTB_DATA_RESET_VAL			0
-#define SMMU_PTB_DATA_ASID_NONSECURE_SHIFT	29
-#define SMMU_PTB_DATA_ASID_WRITABLE_SHIFT	30
-#define SMMU_PTB_DATA_ASID_READABLE_SHIFT	31
-
-#define SMMU_TLB_FLUSH				0x30
-#define SMMU_TLB_FLUSH_VA_MATCH_ALL		0
-#define SMMU_TLB_FLUSH_VA_MATCH_SECTION		2
-#define SMMU_TLB_FLUSH_VA_MATCH_GROUP		3
-#define SMMU_TLB_FLUSH_ASID_SHIFT		29
-#define SMMU_TLB_FLUSH_ASID_MATCH_DISABLE	0
-#define SMMU_TLB_FLUSH_ASID_MATCH_ENABLE	1
-#define SMMU_TLB_FLUSH_ASID_MATCH_SHIFT		31
-
-#define SMMU_PTC_FLUSH				0x34
-#define SMMU_PTC_FLUSH_TYPE_ALL			0
-#define SMMU_PTC_FLUSH_TYPE_ADR			1
-#define SMMU_PTC_FLUSH_ADR_SHIFT		4
-
-#define SMMU_ASID_SECURITY			0x38
-
-#define SMMU_STATS_CACHE_COUNT_BASE		0x1f0
-
-#define SMMU_STATS_CACHE_COUNT(mc, cache, hitmiss)		\
-	(SMMU_STATS_CACHE_COUNT_BASE + 8 * cache + 4 * hitmiss)
-
-#define SMMU_TRANSLATION_ENABLE_0		0x228
-#define SMMU_TRANSLATION_ENABLE_1		0x22c
-#define SMMU_TRANSLATION_ENABLE_2		0x230
-
-#define SMMU_AFI_ASID	0x238   /* PCIE */
-#define SMMU_AVPC_ASID	0x23c   /* AVP */
-#define SMMU_DC_ASID	0x240   /* Display controller */
-#define SMMU_DCB_ASID	0x244   /* Display controller B */
-#define SMMU_EPP_ASID	0x248   /* Encoder pre-processor */
-#define SMMU_G2_ASID	0x24c   /* 2D engine */
-#define SMMU_HC_ASID	0x250   /* Host1x */
-#define SMMU_HDA_ASID	0x254   /* High-def audio */
-#define SMMU_ISP_ASID	0x258   /* Image signal processor */
-#define SMMU_MPE_ASID	0x264   /* MPEG encoder */
-#define SMMU_NV_ASID	0x268   /* (3D) */
-#define SMMU_NV2_ASID	0x26c   /* (3D) */
-#define SMMU_PPCS_ASID	0x270   /* AHB */
-#define SMMU_SATA_ASID	0x278   /* SATA */
-#define SMMU_VDE_ASID	0x27c   /* Video decoder */
-#define SMMU_VI_ASID	0x280   /* Video input */
-
-#define SMMU_PDE_NEXT_SHIFT		28
-
-#define SMMU_TLB_FLUSH_VA_SECTION__MASK		0xffc00000
-#define SMMU_TLB_FLUSH_VA_SECTION__SHIFT	12 /* right shift */
-#define SMMU_TLB_FLUSH_VA_GROUP__MASK		0xffffc000
-#define SMMU_TLB_FLUSH_VA_GROUP__SHIFT		12 /* right shift */
-#define SMMU_TLB_FLUSH_VA(iova, which)	\
-	((((iova) & SMMU_TLB_FLUSH_VA_##which##__MASK) >> \
-		SMMU_TLB_FLUSH_VA_##which##__SHIFT) |	\
-	SMMU_TLB_FLUSH_VA_MATCH_##which)
-#define SMMU_PTB_ASID_CUR(n)	\
-		((n) << SMMU_PTB_ASID_CURRENT_SHIFT)
-#define SMMU_TLB_FLUSH_ASID_MATCH_disable		\
-		(SMMU_TLB_FLUSH_ASID_MATCH_DISABLE <<	\
-			SMMU_TLB_FLUSH_ASID_MATCH_SHIFT)
-#define SMMU_TLB_FLUSH_ASID_MATCH__ENABLE		\
-		(SMMU_TLB_FLUSH_ASID_MATCH_ENABLE <<	\
-			SMMU_TLB_FLUSH_ASID_MATCH_SHIFT)
-
-#define SMMU_PAGE_SHIFT 12
-#define SMMU_PAGE_SIZE	(1 << SMMU_PAGE_SHIFT)
-#define SMMU_PAGE_MASK	((1 << SMMU_PAGE_SHIFT) - 1)
-
-#define SMMU_PDIR_COUNT	1024
-#define SMMU_PDIR_SIZE	(sizeof(unsigned long) * SMMU_PDIR_COUNT)
-#define SMMU_PTBL_COUNT	1024
-#define SMMU_PTBL_SIZE	(sizeof(unsigned long) * SMMU_PTBL_COUNT)
-#define SMMU_PDIR_SHIFT	12
-#define SMMU_PDE_SHIFT	12
-#define SMMU_PTE_SHIFT	12
-#define SMMU_PFN_MASK	0x000fffff
-
-#define SMMU_ADDR_TO_PFN(addr)	((addr) >> 12)
-#define SMMU_ADDR_TO_PDN(addr)	((addr) >> 22)
-#define SMMU_PDN_TO_ADDR(pdn)	((pdn) << 22)
-
-#define _READABLE	(1 << SMMU_PTB_DATA_ASID_READABLE_SHIFT)
-#define _WRITABLE	(1 << SMMU_PTB_DATA_ASID_WRITABLE_SHIFT)
-#define _NONSECURE	(1 << SMMU_PTB_DATA_ASID_NONSECURE_SHIFT)
-#define _PDE_NEXT	(1 << SMMU_PDE_NEXT_SHIFT)
-#define _MASK_ATTR	(_READABLE | _WRITABLE | _NONSECURE)
-
-#define _PDIR_ATTR	(_READABLE | _WRITABLE | _NONSECURE)
-
-#define _PDE_ATTR	(_READABLE | _WRITABLE | _NONSECURE)
-#define _PDE_ATTR_N	(_PDE_ATTR | _PDE_NEXT)
-#define _PDE_VACANT(pdn)	(((pdn) << 10) | _PDE_ATTR)
-
-#define _PTE_ATTR	(_READABLE | _WRITABLE | _NONSECURE)
-#define _PTE_VACANT(addr)	(((addr) >> SMMU_PAGE_SHIFT) | _PTE_ATTR)
-
-#define SMMU_MK_PDIR(page, attr)	\
-		((page_to_phys(page) >> SMMU_PDIR_SHIFT) | (attr))
-#define SMMU_MK_PDE(page, attr)		\
-		(unsigned long)((page_to_phys(page) >> SMMU_PDE_SHIFT) | (attr))
-#define SMMU_EX_PTBL_PAGE(pde)		\
-		pfn_to_page((unsigned long)(pde) & SMMU_PFN_MASK)
-#define SMMU_PFN_TO_PTE(pfn, attr)	(unsigned long)((pfn) | (attr))
-
-#define SMMU_ASID_ENABLE(asid)	((asid) | (1 << 31))
-#define SMMU_ASID_DISABLE	0
-#define SMMU_ASID_ASID(n)	((n) & ~SMMU_ASID_ENABLE(0))
-
-#define NUM_SMMU_REG_BANKS	3
-
-#define smmu_client_enable_hwgrp(c, m)	smmu_client_set_hwgrp(c, m, 1)
-#define smmu_client_disable_hwgrp(c)	smmu_client_set_hwgrp(c, 0, 0)
-#define __smmu_client_enable_hwgrp(c, m) __smmu_client_set_hwgrp(c, m, 1)
-#define __smmu_client_disable_hwgrp(c)	__smmu_client_set_hwgrp(c, 0, 0)
-
-#define HWGRP_INIT(client) [HWGRP_##client] = SMMU_##client##_ASID
-
-static const u32 smmu_hwgrp_asid_reg[] = {
-	HWGRP_INIT(AFI),
-	HWGRP_INIT(AVPC),
-	HWGRP_INIT(DC),
-	HWGRP_INIT(DCB),
-	HWGRP_INIT(EPP),
-	HWGRP_INIT(G2),
-	HWGRP_INIT(HC),
-	HWGRP_INIT(HDA),
-	HWGRP_INIT(ISP),
-	HWGRP_INIT(MPE),
-	HWGRP_INIT(NV),
-	HWGRP_INIT(NV2),
-	HWGRP_INIT(PPCS),
-	HWGRP_INIT(SATA),
-	HWGRP_INIT(VDE),
-	HWGRP_INIT(VI),
+	struct list_head list;
 };
-#define HWGRP_ASID_REG(x) (smmu_hwgrp_asid_reg[x])
 
-/*
- * Per client for address space
- */
-struct smmu_client {
-	struct device		*dev;
-	struct list_head	list;
-	struct smmu_as		*as;
-	u32			hwgrp;
+struct tegra_smmu_as {
+	struct iommu_domain *domain;
+	struct tegra_smmu *smmu;
+	unsigned int use_count;
+	struct page *count;
+	struct page *pd;
+	unsigned id;
+	u32 attr;
 };
 
-/*
- * Per address space
- */
-struct smmu_as {
-	struct smmu_device	*smmu;	/* back pointer to container */
-	unsigned int		asid;
-	spinlock_t		lock;	/* for pagetable */
-	struct page		*pdir_page;
-	unsigned long		pdir_attr;
-	unsigned long		pde_attr;
-	unsigned long		pte_attr;
-	unsigned int		*pte_count;
-
-	struct list_head	client;
-	spinlock_t		client_lock; /* for client list */
-};
+static inline void smmu_writel(struct tegra_smmu *smmu, u32 value,
+			       unsigned long offset)
+{
+	writel(value, smmu->regs + offset);
+}
 
-struct smmu_debugfs_info {
-	struct smmu_device *smmu;
-	int mc;
-	int cache;
-};
+static inline u32 smmu_readl(struct tegra_smmu *smmu, unsigned long offset)
+{
+	return readl(smmu->regs + offset);
+}
 
-/*
- * Per SMMU device - IOMMU device
- */
-struct smmu_device {
-	void __iomem	*regbase;	/* register offset base */
-	void __iomem	**regs;		/* register block start address array */
-	void __iomem	**rege;		/* register block end address array */
-	int		nregs;		/* number of register blocks */
-
-	unsigned long	iovmm_base;	/* remappable base address */
-	unsigned long	page_count;	/* total remappable size */
-	spinlock_t	lock;
-	char		*name;
-	struct device	*dev;
-	struct page *avp_vector_page;	/* dummy page shared by all AS's */
+#define SMMU_CONFIG 0x010
+#define  SMMU_CONFIG_ENABLE (1 << 0)
 
-	/*
-	 * Register image savers for suspend/resume
-	 */
-	unsigned long translation_enable_0;
-	unsigned long translation_enable_1;
-	unsigned long translation_enable_2;
-	unsigned long asid_security;
+#define SMMU_TLB_CONFIG 0x14
+#define  SMMU_TLB_CONFIG_HIT_UNDER_MISS (1 << 29)
+#define  SMMU_TLB_CONFIG_ROUND_ROBIN_ARBITRATION (1 << 28)
+#define  SMMU_TLB_CONFIG_ACTIVE_LINES(x) ((x) & 0x3f)
 
-	struct dentry *debugfs_root;
-	struct smmu_debugfs_info *debugfs_info;
+#define SMMU_PTC_CONFIG 0x18
+#define  SMMU_PTC_CONFIG_ENABLE (1 << 29)
+#define  SMMU_PTC_CONFIG_REQ_LIMIT(x) (((x) & 0x0f) << 24)
+#define  SMMU_PTC_CONFIG_INDEX_MAP(x) ((x) & 0x3f)
 
-	struct device_node *ahb;
+#define SMMU_PTB_ASID 0x01c
+#define  SMMU_PTB_ASID_VALUE(x) ((x) & 0x7f)
 
-	int		num_as;
-	struct smmu_as	as[0];		/* Run-time allocated array */
-};
+#define SMMU_PTB_DATA 0x020
+#define  SMMU_PTB_DATA_VALUE(page, attr) (page_to_phys(page) >> 12 | (attr))
 
-static struct smmu_device *smmu_handle; /* unique for a system */
+#define SMMU_MK_PDE(page, attr) (page_to_phys(page) >> SMMU_PTE_SHIFT | (attr))
 
-/*
- *	SMMU register accessors
- */
-static bool inline smmu_valid_reg(struct smmu_device *smmu,
-				  void __iomem *addr)
-{
-	int i;
+#define SMMU_TLB_FLUSH 0x030
+#define  SMMU_TLB_FLUSH_VA_MATCH_ALL     (0 << 0)
+#define  SMMU_TLB_FLUSH_VA_MATCH_SECTION (2 << 0)
+#define  SMMU_TLB_FLUSH_VA_MATCH_GROUP   (3 << 0)
+#define  SMMU_TLB_FLUSH_ASID(x)          (((x) & 0x7f) << 24)
+#define  SMMU_TLB_FLUSH_VA_SECTION(addr) ((((addr) & 0xffc00000) >> 12) | \
+					  SMMU_TLB_FLUSH_VA_MATCH_SECTION)
+#define  SMMU_TLB_FLUSH_VA_GROUP(addr)   ((((addr) & 0xffffc000) >> 12) | \
+					  SMMU_TLB_FLUSH_VA_MATCH_GROUP)
+#define  SMMU_TLB_FLUSH_ASID_MATCH       (1 << 31)
 
-	for (i = 0; i < smmu->nregs; i++) {
-		if (addr < smmu->regs[i])
-			break;
-		if (addr <= smmu->rege[i])
-			return true;
-	}
+#define SMMU_PTC_FLUSH 0x034
+#define  SMMU_PTC_FLUSH_TYPE_ALL (0 << 0)
+#define  SMMU_PTC_FLUSH_TYPE_ADR (1 << 0)
 
-	return false;
-}
+#define SMMU_PTC_FLUSH_HI 0x9b8
+#define  SMMU_PTC_FLUSH_HI_MASK 0x3
 
-static inline u32 smmu_read(struct smmu_device *smmu, size_t offs)
-{
-	void __iomem *addr = smmu->regbase + offs;
+/* per-SWGROUP SMMU_*_ASID register */
+#define SMMU_ASID_ENABLE (1 << 31)
+#define SMMU_ASID_MASK 0x7f
+#define SMMU_ASID_VALUE(x) ((x) & SMMU_ASID_MASK)
 
-	BUG_ON(!smmu_valid_reg(smmu, addr));
+/* page table definitions */
+#define SMMU_NUM_PDE 1024
+#define SMMU_NUM_PTE 1024
 
-	return readl(addr);
-}
+#define SMMU_SIZE_PD (SMMU_NUM_PDE * 4)
+#define SMMU_SIZE_PT (SMMU_NUM_PTE * 4)
 
-static inline void smmu_write(struct smmu_device *smmu, u32 val, size_t offs)
-{
-	void __iomem *addr = smmu->regbase + offs;
+#define SMMU_PDE_SHIFT 22
+#define SMMU_PTE_SHIFT 12
 
-	BUG_ON(!smmu_valid_reg(smmu, addr));
+#define SMMU_PFN_MASK 0x000fffff
 
-	writel(val, addr);
-}
+#define SMMU_PD_READABLE	(1 << 31)
+#define SMMU_PD_WRITABLE	(1 << 30)
+#define SMMU_PD_NONSECURE	(1 << 29)
 
-#define VA_PAGE_TO_PA(va, page)	\
-	(page_to_phys(page) + ((unsigned long)(va) & ~PAGE_MASK))
+#define SMMU_PDE_READABLE	(1 << 31)
+#define SMMU_PDE_WRITABLE	(1 << 30)
+#define SMMU_PDE_NONSECURE	(1 << 29)
+#define SMMU_PDE_NEXT		(1 << 28)
 
-#define FLUSH_CPU_DCACHE(va, page, size)	\
-	do {	\
-		unsigned long _pa_ = VA_PAGE_TO_PA(va, page);		\
-		__cpuc_flush_dcache_area((void *)(va), (size_t)(size));	\
-		outer_flush_range(_pa_, _pa_+(size_t)(size));		\
-	} while (0)
+#define SMMU_PTE_READABLE	(1 << 31)
+#define SMMU_PTE_WRITABLE	(1 << 30)
+#define SMMU_PTE_NONSECURE	(1 << 29)
 
-/*
- * Any interaction between any block on PPSB and a block on APB or AHB
- * must have these read-back barriers to ensure the APB/AHB bus
- * transaction is complete before initiating activity on the PPSB
- * block.
- */
-#define FLUSH_SMMU_REGS(smmu)	smmu_read(smmu, SMMU_CONFIG)
+#define SMMU_PDE_ATTR		(SMMU_PDE_READABLE | SMMU_PDE_WRITABLE | \
+				 SMMU_PDE_NONSECURE)
+#define SMMU_PTE_ATTR		(SMMU_PTE_READABLE | SMMU_PTE_WRITABLE | \
+				 SMMU_PTE_NONSECURE)
 
-#define smmu_client_hwgrp(c) (u32)((c)->dev->platform_data)
-
-static int __smmu_client_set_hwgrp(struct smmu_client *c,
-				   unsigned long map, int on)
+static inline void smmu_flush_ptc(struct tegra_smmu *smmu, struct page *page,
+				  unsigned long offset)
 {
-	int i;
-	struct smmu_as *as = c->as;
-	u32 val, offs, mask = SMMU_ASID_ENABLE(as->asid);
-	struct smmu_device *smmu = as->smmu;
-
-	WARN_ON(!on && map);
-	if (on && !map)
-		return -EINVAL;
-	if (!on)
-		map = smmu_client_hwgrp(c);
-
-	for_each_set_bit(i, &map, HWGRP_COUNT) {
-		offs = HWGRP_ASID_REG(i);
-		val = smmu_read(smmu, offs);
-		if (on) {
-			if (WARN_ON(val & mask))
-				goto err_hw_busy;
-			val |= mask;
-		} else {
-			WARN_ON((val & mask) == mask);
-			val &= ~mask;
+	phys_addr_t phys = page ? page_to_phys(page) : 0;
+	u32 value;
+
+	if (page) {
+		offset &= ~(smmu->mc->soc->atom_size - 1);
+
+		if (smmu->mc->soc->num_address_bits > 32) {
+#ifdef CONFIG_PHYS_ADDR_T_64BIT
+			value = (phys >> 32) & SMMU_PTC_FLUSH_HI_MASK;
+#else
+			value = 0;
+#endif
+			smmu_writel(smmu, value, SMMU_PTC_FLUSH_HI);
 		}
-		smmu_write(smmu, val, offs);
-	}
-	FLUSH_SMMU_REGS(smmu);
-	c->hwgrp = map;
-	return 0;
 
-err_hw_busy:
-	for_each_set_bit(i, &map, HWGRP_COUNT) {
-		offs = HWGRP_ASID_REG(i);
-		val = smmu_read(smmu, offs);
-		val &= ~mask;
-		smmu_write(smmu, val, offs);
+		value = (phys + offset) | SMMU_PTC_FLUSH_TYPE_ADR;
+	} else {
+		value = SMMU_PTC_FLUSH_TYPE_ALL;
 	}
-	return -EBUSY;
+
+	smmu_writel(smmu, value, SMMU_PTC_FLUSH);
 }
 
-static int smmu_client_set_hwgrp(struct smmu_client *c, u32 map, int on)
+static inline void smmu_flush_tlb(struct tegra_smmu *smmu)
 {
-	u32 val;
-	unsigned long flags;
-	struct smmu_as *as = c->as;
-	struct smmu_device *smmu = as->smmu;
-
-	spin_lock_irqsave(&smmu->lock, flags);
-	val = __smmu_client_set_hwgrp(c, map, on);
-	spin_unlock_irqrestore(&smmu->lock, flags);
-	return val;
+	smmu_writel(smmu, SMMU_TLB_FLUSH_VA_MATCH_ALL, SMMU_TLB_FLUSH);
 }
 
-/*
- * Flush all TLB entries and all PTC entries
- * Caller must lock smmu
- */
-static void smmu_flush_regs(struct smmu_device *smmu, int enable)
+static inline void smmu_flush_tlb_asid(struct tegra_smmu *smmu,
+				       unsigned long asid)
 {
-	u32 val;
-
-	smmu_write(smmu, SMMU_PTC_FLUSH_TYPE_ALL, SMMU_PTC_FLUSH);
-	FLUSH_SMMU_REGS(smmu);
-	val = SMMU_TLB_FLUSH_VA_MATCH_ALL |
-		SMMU_TLB_FLUSH_ASID_MATCH_disable;
-	smmu_write(smmu, val, SMMU_TLB_FLUSH);
+	u32 value;
 
-	if (enable)
-		smmu_write(smmu, SMMU_CONFIG_ENABLE, SMMU_CONFIG);
-	FLUSH_SMMU_REGS(smmu);
+	value = SMMU_TLB_FLUSH_ASID_MATCH | SMMU_TLB_FLUSH_ASID(asid) |
+		SMMU_TLB_FLUSH_VA_MATCH_ALL;
+	smmu_writel(smmu, value, SMMU_TLB_FLUSH);
 }
 
-static int smmu_setup_regs(struct smmu_device *smmu)
+static inline void smmu_flush_tlb_section(struct tegra_smmu *smmu,
+					  unsigned long asid,
+					  unsigned long iova)
 {
-	int i;
-	u32 val;
+	u32 value;
 
-	for (i = 0; i < smmu->num_as; i++) {
-		struct smmu_as *as = &smmu->as[i];
-		struct smmu_client *c;
-
-		smmu_write(smmu, SMMU_PTB_ASID_CUR(as->asid), SMMU_PTB_ASID);
-		val = as->pdir_page ?
-			SMMU_MK_PDIR(as->pdir_page, as->pdir_attr) :
-			SMMU_PTB_DATA_RESET_VAL;
-		smmu_write(smmu, val, SMMU_PTB_DATA);
-
-		list_for_each_entry(c, &as->client, list)
-			__smmu_client_set_hwgrp(c, c->hwgrp, 1);
-	}
-
-	smmu_write(smmu, smmu->translation_enable_0, SMMU_TRANSLATION_ENABLE_0);
-	smmu_write(smmu, smmu->translation_enable_1, SMMU_TRANSLATION_ENABLE_1);
-	smmu_write(smmu, smmu->translation_enable_2, SMMU_TRANSLATION_ENABLE_2);
-	smmu_write(smmu, smmu->asid_security, SMMU_ASID_SECURITY);
-	smmu_write(smmu, SMMU_TLB_CONFIG_RESET_VAL, SMMU_CACHE_CONFIG(_TLB));
-	smmu_write(smmu, SMMU_PTC_CONFIG_RESET_VAL, SMMU_CACHE_CONFIG(_PTC));
-
-	smmu_flush_regs(smmu, 1);
-
-	return tegra_ahb_enable_smmu(smmu->ahb);
+	value = SMMU_TLB_FLUSH_ASID_MATCH | SMMU_TLB_FLUSH_ASID(asid) |
+		SMMU_TLB_FLUSH_VA_SECTION(iova);
+	smmu_writel(smmu, value, SMMU_TLB_FLUSH);
 }
 
-static void flush_ptc_and_tlb(struct smmu_device *smmu,
-		      struct smmu_as *as, dma_addr_t iova,
-		      unsigned long *pte, struct page *page, int is_pde)
+static inline void smmu_flush_tlb_group(struct tegra_smmu *smmu,
+					unsigned long asid,
+					unsigned long iova)
 {
-	u32 val;
-	unsigned long tlb_flush_va = is_pde
-		?  SMMU_TLB_FLUSH_VA(iova, SECTION)
-		:  SMMU_TLB_FLUSH_VA(iova, GROUP);
-
-	val = SMMU_PTC_FLUSH_TYPE_ADR | VA_PAGE_TO_PA(pte, page);
-	smmu_write(smmu, val, SMMU_PTC_FLUSH);
-	FLUSH_SMMU_REGS(smmu);
-	val = tlb_flush_va |
-		SMMU_TLB_FLUSH_ASID_MATCH__ENABLE |
-		(as->asid << SMMU_TLB_FLUSH_ASID_SHIFT);
-	smmu_write(smmu, val, SMMU_TLB_FLUSH);
-	FLUSH_SMMU_REGS(smmu);
-}
+	u32 value;
 
-static void free_ptbl(struct smmu_as *as, dma_addr_t iova)
-{
-	unsigned long pdn = SMMU_ADDR_TO_PDN(iova);
-	unsigned long *pdir = (unsigned long *)page_address(as->pdir_page);
-
-	if (pdir[pdn] != _PDE_VACANT(pdn)) {
-		dev_dbg(as->smmu->dev, "pdn: %lx\n", pdn);
-
-		ClearPageReserved(SMMU_EX_PTBL_PAGE(pdir[pdn]));
-		__free_page(SMMU_EX_PTBL_PAGE(pdir[pdn]));
-		pdir[pdn] = _PDE_VACANT(pdn);
-		FLUSH_CPU_DCACHE(&pdir[pdn], as->pdir_page, sizeof pdir[pdn]);
-		flush_ptc_and_tlb(as->smmu, as, iova, &pdir[pdn],
-				  as->pdir_page, 1);
-	}
+	value = SMMU_TLB_FLUSH_ASID_MATCH | SMMU_TLB_FLUSH_ASID(asid) |
+		SMMU_TLB_FLUSH_VA_GROUP(iova);
+	smmu_writel(smmu, value, SMMU_TLB_FLUSH);
 }
 
-static void free_pdir(struct smmu_as *as)
+static inline void smmu_flush(struct tegra_smmu *smmu)
 {
-	unsigned addr;
-	int count;
-	struct device *dev = as->smmu->dev;
-
-	if (!as->pdir_page)
-		return;
-
-	addr = as->smmu->iovmm_base;
-	count = as->smmu->page_count;
-	while (count-- > 0) {
-		free_ptbl(as, addr);
-		addr += SMMU_PAGE_SIZE * SMMU_PTBL_COUNT;
-	}
-	ClearPageReserved(as->pdir_page);
-	__free_page(as->pdir_page);
-	as->pdir_page = NULL;
-	devm_kfree(dev, as->pte_count);
-	as->pte_count = NULL;
+	smmu_readl(smmu, SMMU_CONFIG);
 }
 
-/*
- * Maps PTBL for given iova and returns the PTE address
- * Caller must unmap the mapped PTBL returned in *ptbl_page_p
- */
-static unsigned long *locate_pte(struct smmu_as *as,
-				 dma_addr_t iova, bool allocate,
-				 struct page **ptbl_page_p,
-				 unsigned int **count)
+static int tegra_smmu_alloc_asid(struct tegra_smmu *smmu, unsigned int *idp)
 {
-	unsigned long ptn = SMMU_ADDR_TO_PFN(iova);
-	unsigned long pdn = SMMU_ADDR_TO_PDN(iova);
-	unsigned long *pdir = page_address(as->pdir_page);
-	unsigned long *ptbl;
-
-	if (pdir[pdn] != _PDE_VACANT(pdn)) {
-		/* Mapped entry table already exists */
-		*ptbl_page_p = SMMU_EX_PTBL_PAGE(pdir[pdn]);
-		ptbl = page_address(*ptbl_page_p);
-	} else if (!allocate) {
-		return NULL;
-	} else {
-		int pn;
-		unsigned long addr = SMMU_PDN_TO_ADDR(pdn);
+	unsigned long id;
 
-		/* Vacant - allocate a new page table */
-		dev_dbg(as->smmu->dev, "New PTBL pdn: %lx\n", pdn);
+	mutex_lock(&smmu->lock);
 
-		*ptbl_page_p = alloc_page(GFP_ATOMIC);
-		if (!*ptbl_page_p) {
-			dev_err(as->smmu->dev,
-				"failed to allocate smmu_device page table\n");
-			return NULL;
-		}
-		SetPageReserved(*ptbl_page_p);
-		ptbl = (unsigned long *)page_address(*ptbl_page_p);
-		for (pn = 0; pn < SMMU_PTBL_COUNT;
-		     pn++, addr += SMMU_PAGE_SIZE) {
-			ptbl[pn] = _PTE_VACANT(addr);
-		}
-		FLUSH_CPU_DCACHE(ptbl, *ptbl_page_p, SMMU_PTBL_SIZE);
-		pdir[pdn] = SMMU_MK_PDE(*ptbl_page_p,
-					as->pde_attr | _PDE_NEXT);
-		FLUSH_CPU_DCACHE(&pdir[pdn], as->pdir_page, sizeof pdir[pdn]);
-		flush_ptc_and_tlb(as->smmu, as, iova, &pdir[pdn],
-				  as->pdir_page, 1);
+	id = find_first_zero_bit(smmu->asids, smmu->soc->num_asids);
+	if (id >= smmu->soc->num_asids) {
+		mutex_unlock(&smmu->lock);
+		return -ENOSPC;
 	}
-	*count = &as->pte_count[pdn];
 
-	return &ptbl[ptn % SMMU_PTBL_COUNT];
+	set_bit(id, smmu->asids);
+	*idp = id;
+
+	mutex_unlock(&smmu->lock);
+	return 0;
 }
 
-#ifdef CONFIG_SMMU_SIG_DEBUG
-static void put_signature(struct smmu_as *as,
-			  dma_addr_t iova, unsigned long pfn)
+static void tegra_smmu_free_asid(struct tegra_smmu *smmu, unsigned int id)
 {
-	struct page *page;
-	unsigned long *vaddr;
-
-	page = pfn_to_page(pfn);
-	vaddr = page_address(page);
-	if (!vaddr)
-		return;
-
-	vaddr[0] = iova;
-	vaddr[1] = pfn << PAGE_SHIFT;
-	FLUSH_CPU_DCACHE(vaddr, page, sizeof(vaddr[0]) * 2);
+	mutex_lock(&smmu->lock);
+	clear_bit(id, smmu->asids);
+	mutex_unlock(&smmu->lock);
 }
-#else
-static inline void put_signature(struct smmu_as *as,
-				 unsigned long addr, unsigned long pfn)
+
+static bool tegra_smmu_capable(enum iommu_cap cap)
 {
+	return false;
 }
-#endif
 
-/*
- * Caller must not hold as->lock
- */
-static int alloc_pdir(struct smmu_as *as)
+static int tegra_smmu_domain_init(struct iommu_domain *domain)
 {
-	unsigned long *pdir, flags;
-	int pdn, err = 0;
-	u32 val;
-	struct smmu_device *smmu = as->smmu;
-	struct page *page;
-	unsigned int *cnt;
+	struct tegra_smmu_as *as;
+	unsigned int i;
+	uint32_t *pd;
 
-	/*
-	 * do the allocation, then grab as->lock
-	 */
-	cnt = devm_kzalloc(smmu->dev,
-			   sizeof(cnt[0]) * SMMU_PDIR_COUNT,
-			   GFP_KERNEL);
-	page = alloc_page(GFP_KERNEL | __GFP_DMA);
+	as = kzalloc(sizeof(*as), GFP_KERNEL);
+	if (!as)
+		return -ENOMEM;
 
-	spin_lock_irqsave(&as->lock, flags);
+	as->attr = SMMU_PD_READABLE | SMMU_PD_WRITABLE | SMMU_PD_NONSECURE;
+	as->domain = domain;
 
-	if (as->pdir_page) {
-		/* We raced, free the redundant */
-		err = -EAGAIN;
-		goto err_out;
+	as->pd = alloc_page(GFP_KERNEL | __GFP_DMA);
+	if (!as->pd) {
+		kfree(as);
+		return -ENOMEM;
 	}
 
-	if (!page || !cnt) {
-		dev_err(smmu->dev, "failed to allocate at %s\n", __func__);
-		err = -ENOMEM;
-		goto err_out;
+	as->count = alloc_page(GFP_KERNEL);
+	if (!as->count) {
+		__free_page(as->pd);
+		kfree(as);
+		return -ENOMEM;
 	}
 
-	as->pdir_page = page;
-	as->pte_count = cnt;
+	/* clear PDEs */
+	pd = page_address(as->pd);
+	SetPageReserved(as->pd);
 
-	SetPageReserved(as->pdir_page);
-	pdir = page_address(as->pdir_page);
+	for (i = 0; i < SMMU_NUM_PDE; i++)
+		pd[i] = 0;
 
-	for (pdn = 0; pdn < SMMU_PDIR_COUNT; pdn++)
-		pdir[pdn] = _PDE_VACANT(pdn);
-	FLUSH_CPU_DCACHE(pdir, as->pdir_page, SMMU_PDIR_SIZE);
-	val = SMMU_PTC_FLUSH_TYPE_ADR | VA_PAGE_TO_PA(pdir, as->pdir_page);
-	smmu_write(smmu, val, SMMU_PTC_FLUSH);
-	FLUSH_SMMU_REGS(as->smmu);
-	val = SMMU_TLB_FLUSH_VA_MATCH_ALL |
-		SMMU_TLB_FLUSH_ASID_MATCH__ENABLE |
-		(as->asid << SMMU_TLB_FLUSH_ASID_SHIFT);
-	smmu_write(smmu, val, SMMU_TLB_FLUSH);
-	FLUSH_SMMU_REGS(as->smmu);
+	/* clear PDE usage counters */
+	pd = page_address(as->count);
+	SetPageReserved(as->count);
 
-	spin_unlock_irqrestore(&as->lock, flags);
-
-	return 0;
+	for (i = 0; i < SMMU_NUM_PDE; i++)
+		pd[i] = 0;
 
-err_out:
-	spin_unlock_irqrestore(&as->lock, flags);
+	domain->priv = as;
 
-	devm_kfree(smmu->dev, cnt);
-	if (page)
-		__free_page(page);
-	return err;
+	return 0;
 }
 
-static void __smmu_iommu_unmap(struct smmu_as *as, dma_addr_t iova)
+static void tegra_smmu_domain_destroy(struct iommu_domain *domain)
 {
-	unsigned long *pte;
-	struct page *page;
-	unsigned int *count;
+	struct tegra_smmu_as *as = domain->priv;
 
-	pte = locate_pte(as, iova, false, &page, &count);
-	if (WARN_ON(!pte))
-		return;
+	/* TODO: free page directory and page tables */
+	ClearPageReserved(as->pd);
 
-	if (WARN_ON(*pte == _PTE_VACANT(iova)))
-		return;
-
-	*pte = _PTE_VACANT(iova);
-	FLUSH_CPU_DCACHE(pte, page, sizeof(*pte));
-	flush_ptc_and_tlb(as->smmu, as, iova, pte, page, 0);
-	if (!--(*count))
-		free_ptbl(as, iova);
+	kfree(as);
 }
 
-static void __smmu_iommu_map_pfn(struct smmu_as *as, dma_addr_t iova,
-				 unsigned long pfn)
+static const struct tegra_smmu_swgroup *
+tegra_smmu_find_swgroup(struct tegra_smmu *smmu, unsigned int swgroup)
 {
-	struct smmu_device *smmu = as->smmu;
-	unsigned long *pte;
-	unsigned int *count;
-	struct page *page;
+	const struct tegra_smmu_swgroup *group = NULL;
+	unsigned int i;
 
-	pte = locate_pte(as, iova, true, &page, &count);
-	if (WARN_ON(!pte))
-		return;
+	for (i = 0; i < smmu->soc->num_swgroups; i++) {
+		if (smmu->soc->swgroups[i].swgroup == swgroup) {
+			group = &smmu->soc->swgroups[i];
+			break;
+		}
+	}
 
-	if (*pte == _PTE_VACANT(iova))
-		(*count)++;
-	*pte = SMMU_PFN_TO_PTE(pfn, as->pte_attr);
-	if (unlikely((*pte == _PTE_VACANT(iova))))
-		(*count)--;
-	FLUSH_CPU_DCACHE(pte, page, sizeof(*pte));
-	flush_ptc_and_tlb(smmu, as, iova, pte, page, 0);
-	put_signature(as, iova, pfn);
+	return group;
 }
 
-static int smmu_iommu_map(struct iommu_domain *domain, unsigned long iova,
-			  phys_addr_t pa, size_t bytes, int prot)
+static void tegra_smmu_enable(struct tegra_smmu *smmu, unsigned int swgroup,
+			      unsigned int asid)
 {
-	struct smmu_as *as = domain->priv;
-	unsigned long pfn = __phys_to_pfn(pa);
-	unsigned long flags;
+	const struct tegra_smmu_swgroup *group;
+	unsigned int i;
+	u32 value;
 
-	dev_dbg(as->smmu->dev, "[%d] %08lx:%pa\n", as->asid, iova, &pa);
+	for (i = 0; i < smmu->soc->num_clients; i++) {
+		const struct tegra_mc_client *client = &smmu->soc->clients[i];
 
-	if (!pfn_valid(pfn))
-		return -ENOMEM;
-
-	spin_lock_irqsave(&as->lock, flags);
-	__smmu_iommu_map_pfn(as, iova, pfn);
-	spin_unlock_irqrestore(&as->lock, flags);
-	return 0;
-}
-
-static size_t smmu_iommu_unmap(struct iommu_domain *domain, unsigned long iova,
-			       size_t bytes)
-{
-	struct smmu_as *as = domain->priv;
-	unsigned long flags;
+		if (client->swgroup != swgroup)
+			continue;
 
-	dev_dbg(as->smmu->dev, "[%d] %08lx\n", as->asid, iova);
+		value = smmu_readl(smmu, client->smmu.reg);
+		value |= BIT(client->smmu.bit);
+		smmu_writel(smmu, value, client->smmu.reg);
+	}
 
-	spin_lock_irqsave(&as->lock, flags);
-	__smmu_iommu_unmap(as, iova);
-	spin_unlock_irqrestore(&as->lock, flags);
-	return SMMU_PAGE_SIZE;
+	group = tegra_smmu_find_swgroup(smmu, swgroup);
+	if (group) {
+		value = smmu_readl(smmu, group->reg);
+		value &= ~SMMU_ASID_MASK;
+		value |= SMMU_ASID_VALUE(asid);
+		value |= SMMU_ASID_ENABLE;
+		smmu_writel(smmu, value, group->reg);
+	}
 }
 
-static phys_addr_t smmu_iommu_iova_to_phys(struct iommu_domain *domain,
-					   dma_addr_t iova)
+static void tegra_smmu_disable(struct tegra_smmu *smmu, unsigned int swgroup,
+			       unsigned int asid)
 {
-	struct smmu_as *as = domain->priv;
-	unsigned long *pte;
-	unsigned int *count;
-	struct page *page;
-	unsigned long pfn;
-	unsigned long flags;
+	const struct tegra_smmu_swgroup *group;
+	unsigned int i;
+	u32 value;
 
-	spin_lock_irqsave(&as->lock, flags);
+	group = tegra_smmu_find_swgroup(smmu, swgroup);
+	if (group) {
+		value = smmu_readl(smmu, group->reg);
+		value &= ~SMMU_ASID_MASK;
+		value |= SMMU_ASID_VALUE(asid);
+		value &= ~SMMU_ASID_ENABLE;
+		smmu_writel(smmu, value, group->reg);
+	}
 
-	pte = locate_pte(as, iova, true, &page, &count);
-	pfn = *pte & SMMU_PFN_MASK;
-	WARN_ON(!pfn_valid(pfn));
-	dev_dbg(as->smmu->dev,
-		"iova:%08llx pfn:%08lx asid:%d\n", (unsigned long long)iova,
-		 pfn, as->asid);
+	for (i = 0; i < smmu->soc->num_clients; i++) {
+		const struct tegra_mc_client *client = &smmu->soc->clients[i];
 
-	spin_unlock_irqrestore(&as->lock, flags);
-	return PFN_PHYS(pfn);
-}
+		if (client->swgroup != swgroup)
+			continue;
 
-static int smmu_iommu_domain_has_cap(struct iommu_domain *domain,
-				     unsigned long cap)
-{
-	return 0;
+		value = smmu_readl(smmu, client->smmu.reg);
+		value &= ~BIT(client->smmu.bit);
+		smmu_writel(smmu, value, client->smmu.reg);
+	}
 }
 
-static int smmu_iommu_attach_dev(struct iommu_domain *domain,
-				 struct device *dev)
+static int tegra_smmu_as_prepare(struct tegra_smmu *smmu,
+				 struct tegra_smmu_as *as)
 {
-	struct smmu_as *as = domain->priv;
-	struct smmu_device *smmu = as->smmu;
-	struct smmu_client *client, *c;
-	u32 map;
+	u32 value;
 	int err;
 
-	client = devm_kzalloc(smmu->dev, sizeof(*c), GFP_KERNEL);
-	if (!client)
-		return -ENOMEM;
-	client->dev = dev;
-	client->as = as;
-	map = (unsigned long)dev->platform_data;
-	if (!map)
-		return -EINVAL;
-
-	err = smmu_client_enable_hwgrp(client, map);
-	if (err)
-		goto err_hwgrp;
-
-	spin_lock(&as->client_lock);
-	list_for_each_entry(c, &as->client, list) {
-		if (c->dev == dev) {
-			dev_err(smmu->dev,
-				"%s is already attached\n", dev_name(c->dev));
-			err = -EINVAL;
-			goto err_client;
-		}
+	if (as->use_count > 0) {
+		as->use_count++;
+		return 0;
 	}
-	list_add(&client->list, &as->client);
-	spin_unlock(&as->client_lock);
 
-	/*
-	 * Reserve "page zero" for AVP vectors using a common dummy
-	 * page.
-	 */
-	if (map & HWG_AVPC) {
-		struct page *page;
+	err = tegra_smmu_alloc_asid(smmu, &as->id);
+	if (err < 0)
+		return err;
 
-		page = as->smmu->avp_vector_page;
-		__smmu_iommu_map_pfn(as, 0, page_to_pfn(page));
+	smmu->soc->ops->flush_dcache(as->pd, 0, SMMU_SIZE_PD);
+	smmu_flush_ptc(smmu, as->pd, 0);
+	smmu_flush_tlb_asid(smmu, as->id);
 
-		pr_info("Reserve \"page zero\" for AVP vectors using a common dummy\n");
-	}
+	smmu_writel(smmu, as->id & 0x7f, SMMU_PTB_ASID);
+	value = SMMU_PTB_DATA_VALUE(as->pd, as->attr);
+	smmu_writel(smmu, value, SMMU_PTB_DATA);
+	smmu_flush(smmu);
 
-	dev_dbg(smmu->dev, "%s is attached\n", dev_name(dev));
-	return 0;
+	as->smmu = smmu;
+	as->use_count++;
 
-err_client:
-	smmu_client_disable_hwgrp(client);
-	spin_unlock(&as->client_lock);
-err_hwgrp:
-	devm_kfree(smmu->dev, client);
-	return err;
+	return 0;
 }
 
-static void smmu_iommu_detach_dev(struct iommu_domain *domain,
-				  struct device *dev)
+static void tegra_smmu_as_unprepare(struct tegra_smmu *smmu,
+				    struct tegra_smmu_as *as)
 {
-	struct smmu_as *as = domain->priv;
-	struct smmu_device *smmu = as->smmu;
-	struct smmu_client *c;
-
-	spin_lock(&as->client_lock);
-
-	list_for_each_entry(c, &as->client, list) {
-		if (c->dev == dev) {
-			smmu_client_disable_hwgrp(c);
-			list_del(&c->list);
-			devm_kfree(smmu->dev, c);
-			c->as = NULL;
-			dev_dbg(smmu->dev,
-				"%s is detached\n", dev_name(c->dev));
-			goto out;
-		}
-	}
-	dev_err(smmu->dev, "Couldn't find %s\n", dev_name(dev));
-out:
-	spin_unlock(&as->client_lock);
+	if (--as->use_count > 0)
+		return;
+
+	tegra_smmu_free_asid(smmu, as->id);
+	as->smmu = NULL;
 }
 
-static int smmu_iommu_domain_init(struct iommu_domain *domain)
+static int tegra_smmu_attach_dev(struct iommu_domain *domain,
+				 struct device *dev)
 {
-	int i, err = -EAGAIN;
-	unsigned long flags;
-	struct smmu_as *as;
-	struct smmu_device *smmu = smmu_handle;
+	struct tegra_smmu *smmu = dev->archdata.iommu;
+	struct tegra_smmu_as *as = domain->priv;
+	struct device_node *np = dev->of_node;
+	struct of_phandle_args args;
+	unsigned int index = 0;
+	int err = 0;
 
-	/* Look for a free AS with lock held */
-	for  (i = 0; i < smmu->num_as; i++) {
-		as = &smmu->as[i];
+	while (!of_parse_phandle_with_args(np, "iommus", "#iommu-cells", index,
+					   &args)) {
+		unsigned int swgroup = args.args[0];
 
-		if (as->pdir_page)
+		if (args.np != smmu->dev->of_node) {
+			of_node_put(args.np);
 			continue;
+		}
 
-		err = alloc_pdir(as);
-		if (!err)
-			goto found;
+		of_node_put(args.np);
 
-		if (err != -EAGAIN)
-			break;
+		err = tegra_smmu_as_prepare(smmu, as);
+		if (err < 0)
+			return err;
+
+		tegra_smmu_enable(smmu, swgroup, as->id);
+		index++;
 	}
-	if (i == smmu->num_as)
-		dev_err(smmu->dev,  "no free AS\n");
-	return err;
 
-found:
-	spin_lock_irqsave(&smmu->lock, flags);
+	if (index == 0)
+		return -ENODEV;
 
-	/* Update PDIR register */
-	smmu_write(smmu, SMMU_PTB_ASID_CUR(as->asid), SMMU_PTB_ASID);
-	smmu_write(smmu,
-		   SMMU_MK_PDIR(as->pdir_page, as->pdir_attr), SMMU_PTB_DATA);
-	FLUSH_SMMU_REGS(smmu);
+	return 0;
+}
 
-	spin_unlock_irqrestore(&smmu->lock, flags);
+static void tegra_smmu_detach_dev(struct iommu_domain *domain, struct device *dev)
+{
+	struct tegra_smmu_as *as = domain->priv;
+	struct device_node *np = dev->of_node;
+	struct tegra_smmu *smmu = as->smmu;
+	struct of_phandle_args args;
+	unsigned int index = 0;
 
-	domain->priv = as;
+	while (!of_parse_phandle_with_args(np, "iommus", "#iommu-cells", index,
+					   &args)) {
+		unsigned int swgroup = args.args[0];
 
-	domain->geometry.aperture_start = smmu->iovmm_base;
-	domain->geometry.aperture_end   = smmu->iovmm_base +
-		smmu->page_count * SMMU_PAGE_SIZE - 1;
-	domain->geometry.force_aperture = true;
+		if (args.np != smmu->dev->of_node) {
+			of_node_put(args.np);
+			continue;
+		}
 
-	dev_dbg(smmu->dev, "smmu_as@%p\n", as);
+		of_node_put(args.np);
 
-	return 0;
+		tegra_smmu_disable(smmu, swgroup, as->id);
+		tegra_smmu_as_unprepare(smmu, as);
+		index++;
+	}
 }
 
-static void smmu_iommu_domain_destroy(struct iommu_domain *domain)
+static u32 *as_get_pte(struct tegra_smmu_as *as, dma_addr_t iova,
+		       struct page **pagep)
 {
-	struct smmu_as *as = domain->priv;
-	struct smmu_device *smmu = as->smmu;
-	unsigned long flags;
+	u32 *pd = page_address(as->pd), *pt, *count;
+	u32 pde = (iova >> SMMU_PDE_SHIFT) & 0x3ff;
+	u32 pte = (iova >> SMMU_PTE_SHIFT) & 0x3ff;
+	struct tegra_smmu *smmu = as->smmu;
+	struct page *page;
+	unsigned int i;
+
+	if (pd[pde] == 0) {
+		page = alloc_page(GFP_KERNEL | __GFP_DMA);
+		if (!page)
+			return NULL;
 
-	spin_lock_irqsave(&as->lock, flags);
+		pt = page_address(page);
+		SetPageReserved(page);
 
-	if (as->pdir_page) {
-		spin_lock(&smmu->lock);
-		smmu_write(smmu, SMMU_PTB_ASID_CUR(as->asid), SMMU_PTB_ASID);
-		smmu_write(smmu, SMMU_PTB_DATA_RESET_VAL, SMMU_PTB_DATA);
-		FLUSH_SMMU_REGS(smmu);
-		spin_unlock(&smmu->lock);
+		for (i = 0; i < SMMU_NUM_PTE; i++)
+			pt[i] = 0;
 
-		free_pdir(as);
-	}
+		smmu->soc->ops->flush_dcache(page, 0, SMMU_SIZE_PT);
 
-	if (!list_empty(&as->client)) {
-		struct smmu_client *c;
+		pd[pde] = SMMU_MK_PDE(page, SMMU_PDE_ATTR | SMMU_PDE_NEXT);
 
-		list_for_each_entry(c, &as->client, list)
-			smmu_iommu_detach_dev(domain, c->dev);
+		smmu->soc->ops->flush_dcache(as->pd, pde << 2, 4);
+		smmu_flush_ptc(smmu, as->pd, pde << 2);
+		smmu_flush_tlb_section(smmu, as->id, iova);
+		smmu_flush(smmu);
+	} else {
+		page = pfn_to_page(pd[pde] & SMMU_PFN_MASK);
+		pt = page_address(page);
 	}
 
-	spin_unlock_irqrestore(&as->lock, flags);
+	*pagep = page;
 
-	domain->priv = NULL;
-	dev_dbg(smmu->dev, "smmu_as@%p\n", as);
-}
+	/* Keep track of entries in this page table. */
+	count = page_address(as->count);
+	if (pt[pte] == 0)
+		count[pde]++;
 
-static const struct iommu_ops smmu_iommu_ops = {
-	.domain_init	= smmu_iommu_domain_init,
-	.domain_destroy	= smmu_iommu_domain_destroy,
-	.attach_dev	= smmu_iommu_attach_dev,
-	.detach_dev	= smmu_iommu_detach_dev,
-	.map		= smmu_iommu_map,
-	.unmap		= smmu_iommu_unmap,
-	.iova_to_phys	= smmu_iommu_iova_to_phys,
-	.domain_has_cap	= smmu_iommu_domain_has_cap,
-	.pgsize_bitmap	= SMMU_IOMMU_PGSIZES,
-};
-
-/* Should be in the order of enum */
-static const char * const smmu_debugfs_mc[] = { "mc", };
-static const char * const smmu_debugfs_cache[] = {  "tlb", "ptc", };
+	return &pt[pte];
+}
 
-static ssize_t smmu_debugfs_stats_write(struct file *file,
-					const char __user *buffer,
-					size_t count, loff_t *pos)
+static void as_put_pte(struct tegra_smmu_as *as, dma_addr_t iova)
 {
-	struct smmu_debugfs_info *info;
-	struct smmu_device *smmu;
-	int i;
-	enum {
-		_OFF = 0,
-		_ON,
-		_RESET,
-	};
-	const char * const command[] = {
-		[_OFF]		= "off",
-		[_ON]		= "on",
-		[_RESET]	= "reset",
-	};
-	char str[] = "reset";
-	u32 val;
-	size_t offs;
+	u32 pde = (iova >> SMMU_PDE_SHIFT) & 0x3ff;
+	u32 pte = (iova >> SMMU_PTE_SHIFT) & 0x3ff;
+	u32 *count = page_address(as->count);
+	u32 *pd = page_address(as->pd), *pt;
+	struct page *page;
 
-	count = min_t(size_t, count, sizeof(str));
-	if (copy_from_user(str, buffer, count))
-		return -EINVAL;
+	page = pfn_to_page(pd[pde] & SMMU_PFN_MASK);
+	pt = page_address(page);
 
-	for (i = 0; i < ARRAY_SIZE(command); i++)
-		if (strncmp(str, command[i],
-			    strlen(command[i])) == 0)
-			break;
+	/*
+	 * When no entries in this page table are used anymore, return the
+	 * memory page to the system.
+	 */
+	if (pt[pte] != 0) {
+		if (--count[pde] == 0) {
+			ClearPageReserved(page);
+			__free_page(page);
+			pd[pde] = 0;
+		}
 
-	if (i == ARRAY_SIZE(command))
-		return -EINVAL;
-
-	info = file_inode(file)->i_private;
-	smmu = info->smmu;
-
-	offs = SMMU_CACHE_CONFIG(info->cache);
-	val = smmu_read(smmu, offs);
-	switch (i) {
-	case _OFF:
-		val &= ~SMMU_CACHE_CONFIG_STATS_ENABLE;
-		val &= ~SMMU_CACHE_CONFIG_STATS_TEST;
-		smmu_write(smmu, val, offs);
-		break;
-	case _ON:
-		val |= SMMU_CACHE_CONFIG_STATS_ENABLE;
-		val &= ~SMMU_CACHE_CONFIG_STATS_TEST;
-		smmu_write(smmu, val, offs);
-		break;
-	case _RESET:
-		val |= SMMU_CACHE_CONFIG_STATS_TEST;
-		smmu_write(smmu, val, offs);
-		val &= ~SMMU_CACHE_CONFIG_STATS_TEST;
-		smmu_write(smmu, val, offs);
-		break;
-	default:
-		BUG();
-		break;
+		pt[pte] = 0;
 	}
-
-	dev_dbg(smmu->dev, "%s() %08x, %08x @%08x\n", __func__,
-		val, smmu_read(smmu, offs), offs);
-
-	return count;
 }
 
-static int smmu_debugfs_stats_show(struct seq_file *s, void *v)
+static int tegra_smmu_map(struct iommu_domain *domain, unsigned long iova,
+			  phys_addr_t paddr, size_t size, int prot)
 {
-	struct smmu_debugfs_info *info = s->private;
-	struct smmu_device *smmu = info->smmu;
-	int i;
-	const char * const stats[] = { "hit", "miss", };
+	struct tegra_smmu_as *as = domain->priv;
+	struct tegra_smmu *smmu = as->smmu;
+	unsigned long offset;
+	struct page *page;
+	u32 *pte;
 
+	pte = as_get_pte(as, iova, &page);
+	if (!pte)
+		return -ENOMEM;
 
-	for (i = 0; i < ARRAY_SIZE(stats); i++) {
-		u32 val;
-		size_t offs;
+	*pte = __phys_to_pfn(paddr) | SMMU_PTE_ATTR;
+	offset = offset_in_page(pte);
 
-		offs = SMMU_STATS_CACHE_COUNT(info->mc, info->cache, i);
-		val = smmu_read(smmu, offs);
-		seq_printf(s, "%s:%08x ", stats[i], val);
+	smmu->soc->ops->flush_dcache(page, offset, 4);
+	smmu_flush_ptc(smmu, page, offset);
+	smmu_flush_tlb_group(smmu, as->id, iova);
+	smmu_flush(smmu);
 
-		dev_dbg(smmu->dev, "%s() %s %08x @%08x\n", __func__,
-			stats[i], val, offs);
-	}
-	seq_printf(s, "\n");
 	return 0;
 }
 
-static int smmu_debugfs_stats_open(struct inode *inode, struct file *file)
+static size_t tegra_smmu_unmap(struct iommu_domain *domain, unsigned long iova,
+			       size_t size)
 {
-	return single_open(file, smmu_debugfs_stats_show, inode->i_private);
-}
+	struct tegra_smmu_as *as = domain->priv;
+	struct tegra_smmu *smmu = as->smmu;
+	unsigned long offset;
+	struct page *page;
+	u32 *pte;
 
-static const struct file_operations smmu_debugfs_stats_fops = {
-	.open		= smmu_debugfs_stats_open,
-	.read		= seq_read,
-	.llseek		= seq_lseek,
-	.release	= single_release,
-	.write		= smmu_debugfs_stats_write,
-};
+	pte = as_get_pte(as, iova, &page);
+	if (!pte)
+		return 0;
 
-static void smmu_debugfs_delete(struct smmu_device *smmu)
-{
-	debugfs_remove_recursive(smmu->debugfs_root);
-	kfree(smmu->debugfs_info);
+	offset = offset_in_page(pte);
+	as_put_pte(as, iova);
+
+	smmu->soc->ops->flush_dcache(page, offset, 4);
+	smmu_flush_ptc(smmu, page, offset);
+	smmu_flush_tlb_group(smmu, as->id, iova);
+	smmu_flush(smmu);
+
+	return size;
 }
 
-static void smmu_debugfs_create(struct smmu_device *smmu)
+static phys_addr_t tegra_smmu_iova_to_phys(struct iommu_domain *domain,
+					   dma_addr_t iova)
 {
-	int i;
-	size_t bytes;
-	struct dentry *root;
-
-	bytes = ARRAY_SIZE(smmu_debugfs_mc) * ARRAY_SIZE(smmu_debugfs_cache) *
-		sizeof(*smmu->debugfs_info);
-	smmu->debugfs_info = kmalloc(bytes, GFP_KERNEL);
-	if (!smmu->debugfs_info)
-		return;
-
-	root = debugfs_create_dir(dev_name(smmu->dev), NULL);
-	if (!root)
-		goto err_out;
-	smmu->debugfs_root = root;
-
-	for (i = 0; i < ARRAY_SIZE(smmu_debugfs_mc); i++) {
-		int j;
-		struct dentry *mc;
-
-		mc = debugfs_create_dir(smmu_debugfs_mc[i], root);
-		if (!mc)
-			goto err_out;
-
-		for (j = 0; j < ARRAY_SIZE(smmu_debugfs_cache); j++) {
-			struct dentry *cache;
-			struct smmu_debugfs_info *info;
-
-			info = smmu->debugfs_info;
-			info += i * ARRAY_SIZE(smmu_debugfs_mc) + j;
-			info->smmu = smmu;
-			info->mc = i;
-			info->cache = j;
-
-			cache = debugfs_create_file(smmu_debugfs_cache[j],
-						    S_IWUGO | S_IRUGO, mc,
-						    (void *)info,
-						    &smmu_debugfs_stats_fops);
-			if (!cache)
-				goto err_out;
-		}
-	}
+	struct tegra_smmu_as *as = domain->priv;
+	struct page *page;
+	unsigned long pfn;
+	u32 *pte;
 
-	return;
+	pte = as_get_pte(as, iova, &page);
+	pfn = *pte & SMMU_PFN_MASK;
 
-err_out:
-	smmu_debugfs_delete(smmu);
+	return PFN_PHYS(pfn);
 }
 
-static int tegra_smmu_suspend(struct device *dev)
+static struct tegra_smmu *tegra_smmu_find(struct device_node *np)
 {
-	struct smmu_device *smmu = dev_get_drvdata(dev);
+	struct platform_device *pdev;
+	struct tegra_mc *mc;
 
-	smmu->translation_enable_0 = smmu_read(smmu, SMMU_TRANSLATION_ENABLE_0);
-	smmu->translation_enable_1 = smmu_read(smmu, SMMU_TRANSLATION_ENABLE_1);
-	smmu->translation_enable_2 = smmu_read(smmu, SMMU_TRANSLATION_ENABLE_2);
-	smmu->asid_security = smmu_read(smmu, SMMU_ASID_SECURITY);
-	return 0;
+	pdev = of_find_device_by_node(np);
+	if (!pdev)
+		return NULL;
+
+	mc = platform_get_drvdata(pdev);
+	if (!mc)
+		return NULL;
+
+	return mc->smmu;
 }
 
-static int tegra_smmu_resume(struct device *dev)
+static int tegra_smmu_add_device(struct device *dev)
 {
-	struct smmu_device *smmu = dev_get_drvdata(dev);
-	unsigned long flags;
-	int err;
+	struct device_node *np = dev->of_node;
+	struct of_phandle_args args;
+	unsigned int index = 0;
 
-	spin_lock_irqsave(&smmu->lock, flags);
-	err = smmu_setup_regs(smmu);
-	spin_unlock_irqrestore(&smmu->lock, flags);
-	return err;
+	while (of_parse_phandle_with_args(np, "iommus", "#iommu-cells", index,
+					  &args) == 0) {
+		struct tegra_smmu *smmu;
+
+		smmu = tegra_smmu_find(args.np);
+		if (smmu) {
+			/*
+			 * Only a single IOMMU master interface is currently
+			 * supported by the Linux kernel, so abort after the
+			 * first match.
+			 */
+			dev->archdata.iommu = smmu;
+			break;
+		}
+
+		index++;
+	}
+
+	return 0;
 }
 
-static int tegra_smmu_probe(struct platform_device *pdev)
+static void tegra_smmu_remove_device(struct device *dev)
 {
-	struct smmu_device *smmu;
-	struct device *dev = &pdev->dev;
-	int i, asids, err = 0;
-	dma_addr_t uninitialized_var(base);
-	size_t bytes, uninitialized_var(size);
+	dev->archdata.iommu = NULL;
+}
 
-	if (smmu_handle)
-		return -EIO;
+static const struct iommu_ops tegra_smmu_ops = {
+	.capable = tegra_smmu_capable,
+	.domain_init = tegra_smmu_domain_init,
+	.domain_destroy = tegra_smmu_domain_destroy,
+	.attach_dev = tegra_smmu_attach_dev,
+	.detach_dev = tegra_smmu_detach_dev,
+	.add_device = tegra_smmu_add_device,
+	.remove_device = tegra_smmu_remove_device,
+	.map = tegra_smmu_map,
+	.unmap = tegra_smmu_unmap,
+	.map_sg = default_iommu_map_sg,
+	.iova_to_phys = tegra_smmu_iova_to_phys,
 
-	BUILD_BUG_ON(PAGE_SHIFT != SMMU_PAGE_SHIFT);
+	.pgsize_bitmap = SZ_4K,
+};
 
-	if (of_property_read_u32(dev->of_node, "nvidia,#asids", &asids))
-		return -ENODEV;
+static void tegra_smmu_ahb_enable(void)
+{
+	static const struct of_device_id ahb_match[] = {
+		{ .compatible = "nvidia,tegra30-ahb", },
+		{ }
+	};
+	struct device_node *ahb;
 
-	bytes = sizeof(*smmu) + asids * sizeof(*smmu->as);
-	smmu = devm_kzalloc(dev, bytes, GFP_KERNEL);
-	if (!smmu) {
-		dev_err(dev, "failed to allocate smmu_device\n");
-		return -ENOMEM;
+	ahb = of_find_matching_node(NULL, ahb_match);
+	if (ahb) {
+		tegra_ahb_enable_smmu(ahb);
+		of_node_put(ahb);
 	}
+}
 
-	smmu->nregs = pdev->num_resources;
-	smmu->regs = devm_kzalloc(dev, 2 * smmu->nregs * sizeof(*smmu->regs),
-				  GFP_KERNEL);
-	smmu->rege = smmu->regs + smmu->nregs;
-	if (!smmu->regs)
-		return -ENOMEM;
-	for (i = 0; i < smmu->nregs; i++) {
-		struct resource *res;
-
-		res = platform_get_resource(pdev, IORESOURCE_MEM, i);
-		smmu->regs[i] = devm_ioremap_resource(&pdev->dev, res);
-		if (IS_ERR(smmu->regs[i]))
-			return PTR_ERR(smmu->regs[i]);
-		smmu->rege[i] = smmu->regs[i] + resource_size(res) - 1;
-	}
-	/* Same as "mc" 1st regiter block start address */
-	smmu->regbase = (void __iomem *)((u32)smmu->regs[0] & PAGE_MASK);
+struct tegra_smmu *tegra_smmu_probe(struct device *dev,
+				    const struct tegra_smmu_soc *soc,
+				    struct tegra_mc *mc)
+{
+	struct tegra_smmu *smmu;
+	size_t size;
+	u32 value;
+	int err;
 
-	err = of_get_dma_window(dev->of_node, NULL, 0, NULL, &base, &size);
-	if (err)
-		return -ENODEV;
+	/* This can happen on Tegra20 which doesn't have an SMMU */
+	if (!soc)
+		return NULL;
 
-	if (size & SMMU_PAGE_MASK)
-		return -EINVAL;
+	smmu = devm_kzalloc(dev, sizeof(*smmu), GFP_KERNEL);
+	if (!smmu)
+		return ERR_PTR(-ENOMEM);
 
-	size >>= SMMU_PAGE_SHIFT;
-	if (!size)
-		return -EINVAL;
+	/*
+	 * This is a bit of a hack. Ideally we'd want to simply return this
+	 * value. However the IOMMU registration process will attempt to add
+	 * all devices to the IOMMU when bus_set_iommu() is called. In order
+	 * not to rely on global variables to track the IOMMU instance, we
+	 * set it here so that it can be looked up from the .add_device()
+	 * callback via the IOMMU device's .drvdata field.
+	 */
+	mc->smmu = smmu;
 
-	smmu->ahb = of_parse_phandle(dev->of_node, "nvidia,ahb", 0);
-	if (!smmu->ahb)
-		return -ENODEV;
+	size = BITS_TO_LONGS(soc->num_asids) * sizeof(long);
 
-	smmu->dev = dev;
-	smmu->num_as = asids;
-	smmu->iovmm_base = base;
-	smmu->page_count = size;
-
-	smmu->translation_enable_0 = ~0;
-	smmu->translation_enable_1 = ~0;
-	smmu->translation_enable_2 = ~0;
-	smmu->asid_security = 0;
-
-	for (i = 0; i < smmu->num_as; i++) {
-		struct smmu_as *as = &smmu->as[i];
-
-		as->smmu = smmu;
-		as->asid = i;
-		as->pdir_attr = _PDIR_ATTR;
-		as->pde_attr = _PDE_ATTR;
-		as->pte_attr = _PTE_ATTR;
-
-		spin_lock_init(&as->lock);
-		spin_lock_init(&as->client_lock);
-		INIT_LIST_HEAD(&as->client);
-	}
-	spin_lock_init(&smmu->lock);
-	err = smmu_setup_regs(smmu);
-	if (err)
-		return err;
-	platform_set_drvdata(pdev, smmu);
+	smmu->asids = devm_kzalloc(dev, size, GFP_KERNEL);
+	if (!smmu->asids)
+		return ERR_PTR(-ENOMEM);
 
-	smmu->avp_vector_page = alloc_page(GFP_KERNEL);
-	if (!smmu->avp_vector_page)
-		return -ENOMEM;
+	mutex_init(&smmu->lock);
 
-	smmu_debugfs_create(smmu);
-	smmu_handle = smmu;
-	bus_set_iommu(&platform_bus_type, &smmu_iommu_ops);
-	return 0;
-}
+	smmu->regs = mc->regs;
+	smmu->soc = soc;
+	smmu->dev = dev;
+	smmu->mc = mc;
 
-static int tegra_smmu_remove(struct platform_device *pdev)
-{
-	struct smmu_device *smmu = platform_get_drvdata(pdev);
-	int i;
+	value = SMMU_PTC_CONFIG_ENABLE | SMMU_PTC_CONFIG_INDEX_MAP(0x3f);
 
-	smmu_debugfs_delete(smmu);
+	if (soc->supports_request_limit)
+		value |= SMMU_PTC_CONFIG_REQ_LIMIT(8);
 
-	smmu_write(smmu, SMMU_CONFIG_DISABLE, SMMU_CONFIG);
-	for (i = 0; i < smmu->num_as; i++)
-		free_pdir(&smmu->as[i]);
-	__free_page(smmu->avp_vector_page);
-	smmu_handle = NULL;
-	return 0;
-}
+	smmu_writel(smmu, value, SMMU_PTC_CONFIG);
 
-static const struct dev_pm_ops tegra_smmu_pm_ops = {
-	.suspend	= tegra_smmu_suspend,
-	.resume		= tegra_smmu_resume,
-};
+	value = SMMU_TLB_CONFIG_HIT_UNDER_MISS |
+		SMMU_TLB_CONFIG_ACTIVE_LINES(0x20);
 
-static struct of_device_id tegra_smmu_of_match[] = {
-	{ .compatible = "nvidia,tegra30-smmu", },
-	{ },
-};
-MODULE_DEVICE_TABLE(of, tegra_smmu_of_match);
-
-static struct platform_driver tegra_smmu_driver = {
-	.probe		= tegra_smmu_probe,
-	.remove		= tegra_smmu_remove,
-	.driver = {
-		.owner	= THIS_MODULE,
-		.name	= "tegra-smmu",
-		.pm	= &tegra_smmu_pm_ops,
-		.of_match_table = tegra_smmu_of_match,
-	},
-};
+	if (soc->supports_round_robin_arbitration)
+		value |= SMMU_TLB_CONFIG_ROUND_ROBIN_ARBITRATION;
 
-static int tegra_smmu_init(void)
-{
-	return platform_driver_register(&tegra_smmu_driver);
-}
+	smmu_writel(smmu, value, SMMU_TLB_CONFIG);
 
-static void __exit tegra_smmu_exit(void)
-{
-	platform_driver_unregister(&tegra_smmu_driver);
-}
+	smmu_flush_ptc(smmu, NULL, 0);
+	smmu_flush_tlb(smmu);
+	smmu_writel(smmu, SMMU_CONFIG_ENABLE, SMMU_CONFIG);
+	smmu_flush(smmu);
+
+	tegra_smmu_ahb_enable();
 
-subsys_initcall(tegra_smmu_init);
-module_exit(tegra_smmu_exit);
+	err = bus_set_iommu(&platform_bus_type, &tegra_smmu_ops);
+	if (err < 0)
+		return ERR_PTR(err);
 
-MODULE_DESCRIPTION("IOMMU API for SMMU in Tegra30");
-MODULE_AUTHOR("Hiroshi DOYU <hdoyu@nvidia.com>");
-MODULE_ALIAS("platform:tegra-smmu");
-MODULE_LICENSE("GPL v2");
+	return smmu;
+}
diff --git a/drivers/ipack/carriers/tpci200.c b/drivers/ipack/carriers/tpci200.c
index de5e32151a1e..9b23843dcad4 100644
--- a/drivers/ipack/carriers/tpci200.c
+++ b/drivers/ipack/carriers/tpci200.c
@@ -572,7 +572,8 @@ static int tpci200_pci_probe(struct pci_dev *pdev,
 	/* Register the carrier in the industry pack bus driver */
 	tpci200->info->ipack_bus = ipack_bus_register(&pdev->dev,
 						      TPCI200_NB_SLOT,
-						      &tpci200_bus_ops);
+						      &tpci200_bus_ops,
+						      THIS_MODULE);
 	if (!tpci200->info->ipack_bus) {
 		dev_err(&pdev->dev,
 			"error registering the carrier on ipack driver\n");
diff --git a/drivers/ipack/devices/ipoctal.c b/drivers/ipack/devices/ipoctal.c
index e41bef048c23..035d5449227e 100644
--- a/drivers/ipack/devices/ipoctal.c
+++ b/drivers/ipack/devices/ipoctal.c
@@ -55,6 +55,22 @@ struct ipoctal {
 	u8 __iomem			*int_space;
 };
 
+static inline struct ipoctal *chan_to_ipoctal(struct ipoctal_channel *chan,
+					      unsigned int index)
+{
+	return container_of(chan, struct ipoctal, channel[index]);
+}
+
+static void ipoctal_reset_channel(struct ipoctal_channel *channel)
+{
+	iowrite8(CR_DISABLE_RX | CR_DISABLE_TX, &channel->regs->w.cr);
+	channel->rx_enable = 0;
+	iowrite8(CR_CMD_RESET_RX, &channel->regs->w.cr);
+	iowrite8(CR_CMD_RESET_TX, &channel->regs->w.cr);
+	iowrite8(CR_CMD_RESET_ERR_STATUS, &channel->regs->w.cr);
+	iowrite8(CR_CMD_RESET_MR, &channel->regs->w.cr);
+}
+
 static int ipoctal_port_activate(struct tty_port *port, struct tty_struct *tty)
 {
 	struct ipoctal_channel *channel;
@@ -72,12 +88,20 @@ static int ipoctal_port_activate(struct tty_port *port, struct tty_struct *tty)
 
 static int ipoctal_open(struct tty_struct *tty, struct file *file)
 {
-	struct ipoctal_channel *channel;
+	struct ipoctal_channel *channel = dev_get_drvdata(tty->dev);
+	struct ipoctal *ipoctal = chan_to_ipoctal(channel, tty->index);
+	int err;
 
-	channel = dev_get_drvdata(tty->dev);
 	tty->driver_data = channel;
 
-	return tty_port_open(&channel->tty_port, tty, file);
+	if (!ipack_get_carrier(ipoctal->dev))
+		return -EBUSY;
+
+	err = tty_port_open(&channel->tty_port, tty, file);
+	if (err)
+		ipack_put_carrier(ipoctal->dev);
+
+	return err;
 }
 
 static void ipoctal_reset_stats(struct ipoctal_stats *stats)
@@ -151,7 +175,6 @@ static void ipoctal_irq_rx(struct ipoctal_channel *channel, u8 sr)
 				flag = TTY_FRAME;
 			}
 			if (sr & SR_RECEIVED_BREAK) {
-				iowrite8(CR_CMD_RESET_BREAK_CHANGE, &channel->regs->w.cr);
 				channel->stats.rcv_break++;
 				flag = TTY_BREAK;
 			}
@@ -196,6 +219,9 @@ static void ipoctal_irq_channel(struct ipoctal_channel *channel)
 	isr = ioread8(&channel->block_regs->r.isr);
 	sr = ioread8(&channel->regs->r.sr);
 
+	if (isr & (IMR_DELTA_BREAK_A | IMR_DELTA_BREAK_B))
+		iowrite8(CR_CMD_RESET_BREAK_CHANGE, &channel->regs->w.cr);
+
 	if ((sr & SR_TX_EMPTY) && (channel->nb_bytes == 0)) {
 		iowrite8(CR_DISABLE_TX, &channel->regs->w.cr);
 		/* In case of RS-485, change from TX to RX when finishing TX.
@@ -304,10 +330,7 @@ static int ipoctal_inst_slot(struct ipoctal *ipoctal, unsigned int bus_nr,
 			channel->isr_rx_rdy_mask = ISR_RxRDY_FFULL_A;
 		}
 
-		iowrite8(CR_DISABLE_RX | CR_DISABLE_TX, &channel->regs->w.cr);
-		channel->rx_enable = 0;
-		iowrite8(CR_CMD_RESET_RX, &channel->regs->w.cr);
-		iowrite8(CR_CMD_RESET_TX, &channel->regs->w.cr);
+		ipoctal_reset_channel(channel);
 		iowrite8(MR1_CHRL_8_BITS | MR1_ERROR_CHAR | MR1_RxINT_RxRDY,
 			 &channel->regs->w.mr); /* mr1 */
 		iowrite8(0, &channel->regs->w.mr); /* mr2 */
@@ -467,11 +490,7 @@ static void ipoctal_set_termios(struct tty_struct *tty,
 	cflag = tty->termios.c_cflag;
 
 	/* Disable and reset everything before change the setup */
-	iowrite8(CR_DISABLE_RX | CR_DISABLE_TX, &channel->regs->w.cr);
-	iowrite8(CR_CMD_RESET_RX, &channel->regs->w.cr);
-	iowrite8(CR_CMD_RESET_TX, &channel->regs->w.cr);
-	iowrite8(CR_CMD_RESET_ERR_STATUS, &channel->regs->w.cr);
-	iowrite8(CR_CMD_RESET_MR, &channel->regs->w.cr);
+	ipoctal_reset_channel(channel);
 
 	/* Set Bits per chars */
 	switch (cflag & CSIZE) {
@@ -609,12 +628,7 @@ static void ipoctal_hangup(struct tty_struct *tty)
 
 	tty_port_hangup(&channel->tty_port);
 
-	iowrite8(CR_DISABLE_RX | CR_DISABLE_TX, &channel->regs->w.cr);
-	channel->rx_enable = 0;
-	iowrite8(CR_CMD_RESET_RX, &channel->regs->w.cr);
-	iowrite8(CR_CMD_RESET_TX, &channel->regs->w.cr);
-	iowrite8(CR_CMD_RESET_ERR_STATUS, &channel->regs->w.cr);
-	iowrite8(CR_CMD_RESET_MR, &channel->regs->w.cr);
+	ipoctal_reset_channel(channel);
 
 	clear_bit(ASYNCB_INITIALIZED, &channel->tty_port.flags);
 	wake_up_interruptible(&channel->tty_port.open_wait);
@@ -627,15 +641,19 @@ static void ipoctal_shutdown(struct tty_struct *tty)
 	if (channel == NULL)
 		return;
 
-	iowrite8(CR_DISABLE_RX | CR_DISABLE_TX, &channel->regs->w.cr);
-	channel->rx_enable = 0;
-	iowrite8(CR_CMD_RESET_RX, &channel->regs->w.cr);
-	iowrite8(CR_CMD_RESET_TX, &channel->regs->w.cr);
-	iowrite8(CR_CMD_RESET_ERR_STATUS, &channel->regs->w.cr);
-	iowrite8(CR_CMD_RESET_MR, &channel->regs->w.cr);
+	ipoctal_reset_channel(channel);
 	clear_bit(ASYNCB_INITIALIZED, &channel->tty_port.flags);
 }
 
+static void ipoctal_cleanup(struct tty_struct *tty)
+{
+	struct ipoctal_channel *channel = tty->driver_data;
+	struct ipoctal *ipoctal = chan_to_ipoctal(channel, tty->index);
+
+	/* release the carrier driver */
+	ipack_put_carrier(ipoctal->dev);
+}
+
 static const struct tty_operations ipoctal_fops = {
 	.ioctl =		NULL,
 	.open =			ipoctal_open,
@@ -647,6 +665,7 @@ static const struct tty_operations ipoctal_fops = {
 	.get_icount =		ipoctal_get_icount,
 	.hangup =		ipoctal_hangup,
 	.shutdown =		ipoctal_shutdown,
+	.cleanup =              ipoctal_cleanup,
 };
 
 static int ipoctal_probe(struct ipack_device *dev)
diff --git a/drivers/ipack/devices/ipoctal.h b/drivers/ipack/devices/ipoctal.h
index 28f1c4233154..7fede0eb6a0c 100644
--- a/drivers/ipack/devices/ipoctal.h
+++ b/drivers/ipack/devices/ipoctal.h
@@ -12,7 +12,7 @@
  * Software Foundation; version 2 of the License.
  */
 
-#ifndef _IPOCTAL_H
+#ifndef _IPOCTAL_H_
 #define _IPOCTAL_H_
 
 #define NR_CHANNELS		8
diff --git a/drivers/ipack/ipack.c b/drivers/ipack/ipack.c
index d0016ba469ed..c0e7b624ce54 100644
--- a/drivers/ipack/ipack.c
+++ b/drivers/ipack/ipack.c
@@ -206,7 +206,8 @@ static struct bus_type ipack_bus_type = {
 };
 
 struct ipack_bus_device *ipack_bus_register(struct device *parent, int slots,
-					    const struct ipack_bus_ops *ops)
+					    const struct ipack_bus_ops *ops,
+					    struct module *owner)
 {
 	int bus_nr;
 	struct ipack_bus_device *bus;
@@ -225,6 +226,7 @@ struct ipack_bus_device *ipack_bus_register(struct device *parent, int slots,
 	bus->parent = parent;
 	bus->slots = slots;
 	bus->ops = ops;
+	bus->owner = owner;
 	return bus;
 }
 EXPORT_SYMBOL_GPL(ipack_bus_register);
diff --git a/drivers/irqchip/Kconfig b/drivers/irqchip/Kconfig
index b8632bf9a7f3..cc79d2a5a8c2 100644
--- a/drivers/irqchip/Kconfig
+++ b/drivers/irqchip/Kconfig
@@ -5,8 +5,15 @@ config IRQCHIP
 config ARM_GIC
 	bool
 	select IRQ_DOMAIN
+	select IRQ_DOMAIN_HIERARCHY
 	select MULTI_IRQ_HANDLER
 
+config ARM_GIC_V2M
+	bool
+	depends on ARM_GIC
+	depends on PCI && PCI_MSI
+	select PCI_MSI_IRQ_DOMAIN
+
 config GIC_NON_BANKED
 	bool
 
@@ -14,6 +21,11 @@ config ARM_GIC_V3
 	bool
 	select IRQ_DOMAIN
 	select MULTI_IRQ_HANDLER
+	select IRQ_DOMAIN_HIERARCHY
+
+config ARM_GIC_V3_ITS
+	bool
+	select PCI_MSI_IRQ_DOMAIN
 
 config ARM_NVIC
 	bool
@@ -48,14 +60,19 @@ config ATMEL_AIC5_IRQ
 	select MULTI_IRQ_HANDLER
 	select SPARSE_IRQ
 
+config BCM7120_L2_IRQ
+	bool
+	select GENERIC_IRQ_CHIP
+	select IRQ_DOMAIN
+
 config BRCMSTB_L2_IRQ
 	bool
-	depends on ARM
 	select GENERIC_IRQ_CHIP
 	select IRQ_DOMAIN
 
 config DW_APB_ICTL
 	bool
+	select GENERIC_IRQ_CHIP
 	select IRQ_DOMAIN
 
 config IMGPDC_IRQ
@@ -75,6 +92,11 @@ config OR1K_PIC
 	bool
 	select IRQ_DOMAIN
 
+config OMAP_IRQCHIP
+	bool
+	select GENERIC_IRQ_CHIP
+	select IRQ_DOMAIN
+
 config ORION_IRQCHIP
 	bool
 	select IRQ_DOMAIN
@@ -109,7 +131,18 @@ config XTENSA_MX
 config IRQ_CROSSBAR
 	bool
 	help
-	  Support for a CROSSBAR ip that preceeds the main interrupt controller.
+	  Support for a CROSSBAR ip that precedes the main interrupt controller.
 	  The primary irqchip invokes the crossbar's callback which inturn allocates
 	  a free irq and configures the IP. Thus the peripheral interrupts are
 	  routed to one of the free irqchip interrupt lines.
+
+config KEYSTONE_IRQ
+	tristate "Keystone 2 IRQ controller IP"
+	depends on ARCH_KEYSTONE
+	help
+		Support for Texas Instruments Keystone 2 IRQ controller IP which
+		is part of the Keystone 2 IPC mechanism
+
+config MIPS_GIC
+	bool
+	select MIPS_CM
diff --git a/drivers/irqchip/Makefile b/drivers/irqchip/Makefile
index 73052ba9ca62..9516a324be6d 100644
--- a/drivers/irqchip/Makefile
+++ b/drivers/irqchip/Makefile
@@ -2,6 +2,7 @@ obj-$(CONFIG_IRQCHIP)			+= irqchip.o
 
 obj-$(CONFIG_ARCH_BCM2835)		+= irq-bcm2835.o
 obj-$(CONFIG_ARCH_EXYNOS)		+= exynos-combiner.o
+obj-$(CONFIG_ARCH_HIP04)		+= irq-hip04.o
 obj-$(CONFIG_ARCH_MMP)			+= irq-mmp.o
 obj-$(CONFIG_ARCH_MVEBU)		+= irq-armada-370-xp.o
 obj-$(CONFIG_ARCH_MXS)			+= irq-mxs.o
@@ -13,11 +14,14 @@ obj-$(CONFIG_ARCH_MOXART)		+= irq-moxart.o
 obj-$(CONFIG_CLPS711X_IRQCHIP)		+= irq-clps711x.o
 obj-$(CONFIG_OR1K_PIC)			+= irq-or1k-pic.o
 obj-$(CONFIG_ORION_IRQCHIP)		+= irq-orion.o
+obj-$(CONFIG_OMAP_IRQCHIP)		+= irq-omap-intc.o
 obj-$(CONFIG_ARCH_SUNXI)		+= irq-sun4i.o
 obj-$(CONFIG_ARCH_SUNXI)		+= irq-sunxi-nmi.o
 obj-$(CONFIG_ARCH_SPEAR3XX)		+= spear-shirq.o
 obj-$(CONFIG_ARM_GIC)			+= irq-gic.o irq-gic-common.o
+obj-$(CONFIG_ARM_GIC_V2M)		+= irq-gic-v2m.o
 obj-$(CONFIG_ARM_GIC_V3)		+= irq-gic-v3.o irq-gic-common.o
+obj-$(CONFIG_ARM_GIC_V3_ITS)		+= irq-gic-v3-its.o
 obj-$(CONFIG_ARM_NVIC)			+= irq-nvic.o
 obj-$(CONFIG_ARM_VIC)			+= irq-vic.o
 obj-$(CONFIG_ATMEL_AIC_IRQ)		+= irq-atmel-aic-common.o irq-atmel-aic.o
@@ -33,4 +37,8 @@ obj-$(CONFIG_TB10X_IRQC)		+= irq-tb10x.o
 obj-$(CONFIG_XTENSA)			+= irq-xtensa-pic.o
 obj-$(CONFIG_XTENSA_MX)			+= irq-xtensa-mx.o
 obj-$(CONFIG_IRQ_CROSSBAR)		+= irq-crossbar.o
+obj-$(CONFIG_BCM7120_L2_IRQ)		+= irq-bcm7120-l2.o
 obj-$(CONFIG_BRCMSTB_L2_IRQ)		+= irq-brcmstb-l2.o
+obj-$(CONFIG_KEYSTONE_IRQ)		+= irq-keystone.o
+obj-$(CONFIG_MIPS_GIC)			+= irq-mips-gic.o
+obj-$(CONFIG_ARCH_MEDIATEK)		+= irq-mtk-sysirq.o
diff --git a/drivers/irqchip/irq-armada-370-xp.c b/drivers/irqchip/irq-armada-370-xp.c
index 574aba0eba4e..463c235acbdc 100644
--- a/drivers/irqchip/irq-armada-370-xp.c
+++ b/drivers/irqchip/irq-armada-370-xp.c
@@ -26,6 +26,7 @@
 #include <linux/of_pci.h>
 #include <linux/irqdomain.h>
 #include <linux/slab.h>
+#include <linux/syscore_ops.h>
 #include <linux/msi.h>
 #include <asm/mach/arch.h>
 #include <asm/exception.h>
@@ -43,6 +44,7 @@
 #define ARMADA_370_XP_INT_CLEAR_ENABLE_OFFS	(0x34)
 #define ARMADA_370_XP_INT_SOURCE_CTL(irq)	(0x100 + irq*4)
 #define ARMADA_370_XP_INT_SOURCE_CPU_MASK	0xF
+#define ARMADA_370_XP_INT_IRQ_FIQ_MASK(cpuid)	((BIT(0) | BIT(8)) << cpuid)
 
 #define ARMADA_370_XP_CPU_INTACK_OFFS		(0x44)
 #define ARMADA_375_PPI_CAUSE			(0x10)
@@ -66,6 +68,7 @@
 static void __iomem *per_cpu_int_base;
 static void __iomem *main_int_base;
 static struct irq_domain *armada_370_xp_mpic_domain;
+static u32 doorbell_mask_reg;
 #ifdef CONFIG_PCI_MSI
 static struct irq_domain *armada_370_xp_msi_domain;
 static DECLARE_BITMAP(msi_used, PCI_MSI_DOORBELL_NR);
@@ -129,13 +132,17 @@ static void armada_370_xp_free_msi(int hwirq)
 	mutex_unlock(&msi_used_lock);
 }
 
-static int armada_370_xp_setup_msi_irq(struct msi_chip *chip,
+static int armada_370_xp_setup_msi_irq(struct msi_controller *chip,
 				       struct pci_dev *pdev,
 				       struct msi_desc *desc)
 {
 	struct msi_msg msg;
 	int virq, hwirq;
 
+	/* We support MSI, but not MSI-X */
+	if (desc->msi_attrib.is_msix)
+		return -EINVAL;
+
 	hwirq = armada_370_xp_alloc_msi();
 	if (hwirq < 0)
 		return hwirq;
@@ -152,11 +159,11 @@ static int armada_370_xp_setup_msi_irq(struct msi_chip *chip,
 	msg.address_hi = 0;
 	msg.data = 0xf00 | (hwirq + 16);
 
-	write_msi_msg(virq, &msg);
+	pci_write_msi_msg(virq, &msg);
 	return 0;
 }
 
-static void armada_370_xp_teardown_msi_irq(struct msi_chip *chip,
+static void armada_370_xp_teardown_msi_irq(struct msi_controller *chip,
 					   unsigned int irq)
 {
 	struct irq_data *d = irq_get_irq_data(irq);
@@ -166,21 +173,12 @@ static void armada_370_xp_teardown_msi_irq(struct msi_chip *chip,
 	armada_370_xp_free_msi(hwirq);
 }
 
-static int armada_370_xp_check_msi_device(struct msi_chip *chip, struct pci_dev *dev,
-					  int nvec, int type)
-{
-	/* We support MSI, but not MSI-X */
-	if (type == PCI_CAP_ID_MSI)
-		return 0;
-	return -EINVAL;
-}
-
 static struct irq_chip armada_370_xp_msi_irq_chip = {
 	.name = "armada_370_xp_msi_irq",
-	.irq_enable = unmask_msi_irq,
-	.irq_disable = mask_msi_irq,
-	.irq_mask = mask_msi_irq,
-	.irq_unmask = unmask_msi_irq,
+	.irq_enable = pci_msi_unmask_irq,
+	.irq_disable = pci_msi_mask_irq,
+	.irq_mask = pci_msi_mask_irq,
+	.irq_unmask = pci_msi_unmask_irq,
 };
 
 static int armada_370_xp_msi_map(struct irq_domain *domain, unsigned int virq,
@@ -200,7 +198,7 @@ static const struct irq_domain_ops armada_370_xp_msi_irq_ops = {
 static int armada_370_xp_msi_init(struct device_node *node,
 				  phys_addr_t main_int_phys_base)
 {
-	struct msi_chip *msi_chip;
+	struct msi_controller *msi_chip;
 	u32 reg;
 	int ret;
 
@@ -213,7 +211,6 @@ static int armada_370_xp_msi_init(struct device_node *node,
 
 	msi_chip->setup_irq = armada_370_xp_setup_msi_irq;
 	msi_chip->teardown_irq = armada_370_xp_teardown_msi_irq;
-	msi_chip->check_device = armada_370_xp_check_msi_device;
 	msi_chip->of_node = node;
 
 	armada_370_xp_msi_domain =
@@ -271,7 +268,7 @@ static int armada_xp_set_affinity(struct irq_data *d,
 	writel(reg, main_int_base + ARMADA_370_XP_INT_SOURCE_CTL(hwirq));
 	raw_spin_unlock(&irq_controller_lock);
 
-	return 0;
+	return IRQ_SET_MASK_OK;
 }
 #endif
 
@@ -393,13 +390,15 @@ static void armada_370_xp_handle_msi_irq(struct pt_regs *regs, bool is_chained)
 		if (!(msimask & BIT(msinr)))
 			continue;
 
-		irq = irq_find_mapping(armada_370_xp_msi_domain,
-				       msinr - 16);
-
-		if (is_chained)
+		if (is_chained) {
+			irq = irq_find_mapping(armada_370_xp_msi_domain,
+					       msinr - 16);
 			generic_handle_irq(irq);
-		else
-			handle_IRQ(irq, regs);
+		} else {
+			irq = msinr - 16;
+			handle_domain_irq(armada_370_xp_msi_domain,
+					  irq, regs);
+		}
 	}
 }
 #else
@@ -410,19 +409,29 @@ static void armada_370_xp_mpic_handle_cascade_irq(unsigned int irq,
 						  struct irq_desc *desc)
 {
 	struct irq_chip *chip = irq_get_chip(irq);
-	unsigned long irqmap, irqn;
+	unsigned long irqmap, irqn, irqsrc, cpuid;
 	unsigned int cascade_irq;
 
 	chained_irq_enter(chip, desc);
 
 	irqmap = readl_relaxed(per_cpu_int_base + ARMADA_375_PPI_CAUSE);
-
-	if (irqmap & BIT(0)) {
-		armada_370_xp_handle_msi_irq(NULL, true);
-		irqmap &= ~BIT(0);
-	}
+	cpuid = cpu_logical_map(smp_processor_id());
 
 	for_each_set_bit(irqn, &irqmap, BITS_PER_LONG) {
+		irqsrc = readl_relaxed(main_int_base +
+				       ARMADA_370_XP_INT_SOURCE_CTL(irqn));
+
+		/* Check if the interrupt is not masked on current CPU.
+		 * Test IRQ (0-1) and FIQ (8-9) mask bits.
+		 */
+		if (!(irqsrc & ARMADA_370_XP_INT_IRQ_FIQ_MASK(cpuid)))
+			continue;
+
+		if (irqn == 1) {
+			armada_370_xp_handle_msi_irq(NULL, true);
+			continue;
+		}
+
 		cascade_irq = irq_find_mapping(armada_370_xp_mpic_domain, irqn);
 		generic_handle_irq(cascade_irq);
 	}
@@ -444,9 +453,8 @@ armada_370_xp_handle_irq(struct pt_regs *regs)
 			break;
 
 		if (irqnr > 1) {
-			irqnr =	irq_find_mapping(armada_370_xp_mpic_domain,
-					irqnr);
-			handle_IRQ(irqnr, regs);
+			handle_domain_irq(armada_370_xp_mpic_domain,
+					  irqnr, regs);
 			continue;
 		}
 
@@ -479,6 +487,54 @@ armada_370_xp_handle_irq(struct pt_regs *regs)
 	} while (1);
 }
 
+static int armada_370_xp_mpic_suspend(void)
+{
+	doorbell_mask_reg = readl(per_cpu_int_base +
+				  ARMADA_370_XP_IN_DRBEL_MSK_OFFS);
+	return 0;
+}
+
+static void armada_370_xp_mpic_resume(void)
+{
+	int nirqs;
+	irq_hw_number_t irq;
+
+	/* Re-enable interrupts */
+	nirqs = (readl(main_int_base + ARMADA_370_XP_INT_CONTROL) >> 2) & 0x3ff;
+	for (irq = 0; irq < nirqs; irq++) {
+		struct irq_data *data;
+		int virq;
+
+		virq = irq_linear_revmap(armada_370_xp_mpic_domain, irq);
+		if (virq == 0)
+			continue;
+
+		if (irq != ARMADA_370_XP_TIMER0_PER_CPU_IRQ)
+			writel(irq, per_cpu_int_base +
+			       ARMADA_370_XP_INT_CLEAR_MASK_OFFS);
+		else
+			writel(irq, main_int_base +
+			       ARMADA_370_XP_INT_SET_ENABLE_OFFS);
+
+		data = irq_get_irq_data(virq);
+		if (!irqd_irq_disabled(data))
+			armada_370_xp_irq_unmask(data);
+	}
+
+	/* Reconfigure doorbells for IPIs and MSIs */
+	writel(doorbell_mask_reg,
+	       per_cpu_int_base + ARMADA_370_XP_IN_DRBEL_MSK_OFFS);
+	if (doorbell_mask_reg & IPI_DOORBELL_MASK)
+		writel(0, per_cpu_int_base + ARMADA_370_XP_INT_CLEAR_MASK_OFFS);
+	if (doorbell_mask_reg & PCI_MSI_DOORBELL_MASK)
+		writel(1, per_cpu_int_base + ARMADA_370_XP_INT_CLEAR_MASK_OFFS);
+}
+
+struct syscore_ops armada_370_xp_mpic_syscore_ops = {
+	.suspend	= armada_370_xp_mpic_suspend,
+	.resume		= armada_370_xp_mpic_resume,
+};
+
 static int __init armada_370_xp_mpic_of_init(struct device_node *node,
 					     struct device_node *parent)
 {
@@ -535,6 +591,8 @@ static int __init armada_370_xp_mpic_of_init(struct device_node *node,
 					armada_370_xp_mpic_handle_cascade_irq);
 	}
 
+	register_syscore_ops(&armada_370_xp_mpic_syscore_ops);
+
 	return 0;
 }
 
diff --git a/drivers/irqchip/irq-atmel-aic-common.c b/drivers/irqchip/irq-atmel-aic-common.c
index 6ae3cdee0681..63cd031b2c28 100644
--- a/drivers/irqchip/irq-atmel-aic-common.c
+++ b/drivers/irqchip/irq-atmel-aic-common.c
@@ -28,7 +28,7 @@
 #define AT91_AIC_IRQ_MIN_PRIORITY	0
 #define AT91_AIC_IRQ_MAX_PRIORITY	7
 
-#define AT91_AIC_SRCTYPE		GENMASK(7, 6)
+#define AT91_AIC_SRCTYPE		GENMASK(6, 5)
 #define AT91_AIC_SRCTYPE_LOW		(0 << 5)
 #define AT91_AIC_SRCTYPE_FALLING	(1 << 5)
 #define AT91_AIC_SRCTYPE_HIGH		(2 << 5)
@@ -74,7 +74,7 @@ int aic_common_set_type(struct irq_data *d, unsigned type, unsigned *val)
 		return -EINVAL;
 	}
 
-	*val &= AT91_AIC_SRCTYPE;
+	*val &= ~AT91_AIC_SRCTYPE;
 	*val |= aic_type;
 
 	return 0;
@@ -167,6 +167,32 @@ void __init aic_common_rtc_irq_fixup(struct device_node *root)
 	iounmap(regs);
 }
 
+#define AT91_RTT_MR		0x00			/* Real-time Mode Register */
+#define AT91_RTT_ALMIEN		(1 << 16)		/* Alarm Interrupt Enable */
+#define AT91_RTT_RTTINCIEN	(1 << 17)		/* Real Time Timer Increment Interrupt Enable */
+
+void __init aic_common_rtt_irq_fixup(struct device_node *root)
+{
+	struct device_node *np;
+	void __iomem *regs;
+
+	/*
+	 * The at91sam9263 SoC has 2 instances of the RTT block, hence we
+	 * iterate over the DT to find each occurrence.
+	 */
+	for_each_compatible_node(np, NULL, "atmel,at91sam9260-rtt") {
+		regs = of_iomap(np, 0);
+		if (!regs)
+			continue;
+
+		writel(readl(regs + AT91_RTT_MR) &
+		       ~(AT91_RTT_ALMIEN | AT91_RTT_RTTINCIEN),
+		       regs + AT91_RTT_MR);
+
+		iounmap(regs);
+	}
+}
+
 void __init aic_common_irq_fixup(const struct of_device_id *matches)
 {
 	struct device_node *root = of_find_node_by_path("/");
@@ -217,8 +243,9 @@ struct irq_domain *__init aic_common_of_init(struct device_node *node,
 	}
 
 	ret = irq_alloc_domain_generic_chips(domain, 32, 1, name,
-					     handle_level_irq, 0, 0,
-					     IRQCHIP_SKIP_SET_WAKE);
+					     handle_fasteoi_irq,
+					     IRQ_NOREQUEST | IRQ_NOPROBE |
+					     IRQ_NOAUTOEN, 0, 0);
 	if (ret)
 		goto err_domain_remove;
 
@@ -230,7 +257,6 @@ struct irq_domain *__init aic_common_of_init(struct device_node *node,
 		gc->unused = 0;
 		gc->wake_enabled = ~0;
 		gc->chip_types[0].type = IRQ_TYPE_SENSE_MASK;
-		gc->chip_types[0].handler = handle_fasteoi_irq;
 		gc->chip_types[0].chip.irq_eoi = irq_gc_eoi;
 		gc->chip_types[0].chip.irq_set_wake = irq_gc_set_wake;
 		gc->chip_types[0].chip.irq_shutdown = aic_common_shutdown;
diff --git a/drivers/irqchip/irq-atmel-aic-common.h b/drivers/irqchip/irq-atmel-aic-common.h
index 90aa00e918d6..603f0a9d5411 100644
--- a/drivers/irqchip/irq-atmel-aic-common.h
+++ b/drivers/irqchip/irq-atmel-aic-common.h
@@ -34,6 +34,8 @@ struct irq_domain *__init aic_common_of_init(struct device_node *node,
 
 void __init aic_common_rtc_irq_fixup(struct device_node *root);
 
+void __init aic_common_rtt_irq_fixup(struct device_node *root);
+
 void __init aic_common_irq_fixup(const struct of_device_id *matches);
 
 #endif /* __IRQ_ATMEL_AIC_COMMON_H */
diff --git a/drivers/irqchip/irq-atmel-aic.c b/drivers/irqchip/irq-atmel-aic.c
index a82869e9fb26..dae3604b32a9 100644
--- a/drivers/irqchip/irq-atmel-aic.c
+++ b/drivers/irqchip/irq-atmel-aic.c
@@ -65,15 +65,13 @@ aic_handle(struct pt_regs *regs)
 	u32 irqnr;
 	u32 irqstat;
 
-	irqnr = irq_reg_readl(gc->reg_base + AT91_AIC_IVR);
-	irqstat = irq_reg_readl(gc->reg_base + AT91_AIC_ISR);
-
-	irqnr = irq_find_mapping(aic_domain, irqnr);
+	irqnr = irq_reg_readl(gc, AT91_AIC_IVR);
+	irqstat = irq_reg_readl(gc, AT91_AIC_ISR);
 
 	if (!irqstat)
-		irq_reg_writel(0, gc->reg_base + AT91_AIC_EOICR);
+		irq_reg_writel(gc, 0, AT91_AIC_EOICR);
 	else
-		handle_IRQ(irqnr, regs);
+		handle_domain_irq(aic_domain, irqnr, regs);
 }
 
 static int aic_retrigger(struct irq_data *d)
@@ -82,7 +80,7 @@ static int aic_retrigger(struct irq_data *d)
 
 	/* Enable interrupt on AIC5 */
 	irq_gc_lock(gc);
-	irq_reg_writel(d->mask, gc->reg_base + AT91_AIC_ISCR);
+	irq_reg_writel(gc, d->mask, AT91_AIC_ISCR);
 	irq_gc_unlock(gc);
 
 	return 0;
@@ -94,12 +92,12 @@ static int aic_set_type(struct irq_data *d, unsigned type)
 	unsigned int smr;
 	int ret;
 
-	smr = irq_reg_readl(gc->reg_base + AT91_AIC_SMR(d->hwirq));
+	smr = irq_reg_readl(gc, AT91_AIC_SMR(d->hwirq));
 	ret = aic_common_set_type(d, type, &smr);
 	if (ret)
 		return ret;
 
-	irq_reg_writel(smr, gc->reg_base + AT91_AIC_SMR(d->hwirq));
+	irq_reg_writel(gc, smr, AT91_AIC_SMR(d->hwirq));
 
 	return 0;
 }
@@ -110,8 +108,8 @@ static void aic_suspend(struct irq_data *d)
 	struct irq_chip_generic *gc = irq_data_get_irq_chip_data(d);
 
 	irq_gc_lock(gc);
-	irq_reg_writel(gc->mask_cache, gc->reg_base + AT91_AIC_IDCR);
-	irq_reg_writel(gc->wake_active, gc->reg_base + AT91_AIC_IECR);
+	irq_reg_writel(gc, gc->mask_cache, AT91_AIC_IDCR);
+	irq_reg_writel(gc, gc->wake_active, AT91_AIC_IECR);
 	irq_gc_unlock(gc);
 }
 
@@ -120,8 +118,8 @@ static void aic_resume(struct irq_data *d)
 	struct irq_chip_generic *gc = irq_data_get_irq_chip_data(d);
 
 	irq_gc_lock(gc);
-	irq_reg_writel(gc->wake_active, gc->reg_base + AT91_AIC_IDCR);
-	irq_reg_writel(gc->mask_cache, gc->reg_base + AT91_AIC_IECR);
+	irq_reg_writel(gc, gc->wake_active, AT91_AIC_IDCR);
+	irq_reg_writel(gc, gc->mask_cache, AT91_AIC_IECR);
 	irq_gc_unlock(gc);
 }
 
@@ -130,8 +128,8 @@ static void aic_pm_shutdown(struct irq_data *d)
 	struct irq_chip_generic *gc = irq_data_get_irq_chip_data(d);
 
 	irq_gc_lock(gc);
-	irq_reg_writel(0xffffffff, gc->reg_base + AT91_AIC_IDCR);
-	irq_reg_writel(0xffffffff, gc->reg_base + AT91_AIC_ICCR);
+	irq_reg_writel(gc, 0xffffffff, AT91_AIC_IDCR);
+	irq_reg_writel(gc, 0xffffffff, AT91_AIC_ICCR);
 	irq_gc_unlock(gc);
 }
 #else
@@ -150,24 +148,24 @@ static void __init aic_hw_init(struct irq_domain *domain)
 	 * will not Lock out nIRQ
 	 */
 	for (i = 0; i < 8; i++)
-		irq_reg_writel(0, gc->reg_base + AT91_AIC_EOICR);
+		irq_reg_writel(gc, 0, AT91_AIC_EOICR);
 
 	/*
 	 * Spurious Interrupt ID in Spurious Vector Register.
 	 * When there is no current interrupt, the IRQ Vector Register
 	 * reads the value stored in AIC_SPU
 	 */
-	irq_reg_writel(0xffffffff, gc->reg_base + AT91_AIC_SPU);
+	irq_reg_writel(gc, 0xffffffff, AT91_AIC_SPU);
 
 	/* No debugging in AIC: Debug (Protect) Control Register */
-	irq_reg_writel(0, gc->reg_base + AT91_AIC_DCR);
+	irq_reg_writel(gc, 0, AT91_AIC_DCR);
 
 	/* Disable and clear all interrupts initially */
-	irq_reg_writel(0xffffffff, gc->reg_base + AT91_AIC_IDCR);
-	irq_reg_writel(0xffffffff, gc->reg_base + AT91_AIC_ICCR);
+	irq_reg_writel(gc, 0xffffffff, AT91_AIC_IDCR);
+	irq_reg_writel(gc, 0xffffffff, AT91_AIC_ICCR);
 
 	for (i = 0; i < 32; i++)
-		irq_reg_writel(i, gc->reg_base + AT91_AIC_SVR(i));
+		irq_reg_writel(gc, i, AT91_AIC_SVR(i));
 }
 
 static int aic_irq_domain_xlate(struct irq_domain *d,
@@ -197,10 +195,10 @@ static int aic_irq_domain_xlate(struct irq_domain *d,
 	gc = dgc->gc[idx];
 
 	irq_gc_lock(gc);
-	smr = irq_reg_readl(gc->reg_base + AT91_AIC_SMR(*out_hwirq));
+	smr = irq_reg_readl(gc, AT91_AIC_SMR(*out_hwirq));
 	ret = aic_common_set_priority(intspec[2], &smr);
 	if (!ret)
-		irq_reg_writel(smr, gc->reg_base + AT91_AIC_SMR(*out_hwirq));
+		irq_reg_writel(gc, smr, AT91_AIC_SMR(*out_hwirq));
 	irq_gc_unlock(gc);
 
 	return ret;
@@ -211,16 +209,32 @@ static const struct irq_domain_ops aic_irq_ops = {
 	.xlate	= aic_irq_domain_xlate,
 };
 
-static void __init at91sam9_aic_irq_fixup(struct device_node *root)
+static void __init at91rm9200_aic_irq_fixup(struct device_node *root)
+{
+	aic_common_rtc_irq_fixup(root);
+}
+
+static void __init at91sam9260_aic_irq_fixup(struct device_node *root)
+{
+	aic_common_rtt_irq_fixup(root);
+}
+
+static void __init at91sam9g45_aic_irq_fixup(struct device_node *root)
 {
 	aic_common_rtc_irq_fixup(root);
+	aic_common_rtt_irq_fixup(root);
 }
 
 static const struct of_device_id __initdata aic_irq_fixups[] = {
-	{ .compatible = "atmel,at91sam9g45", .data = at91sam9_aic_irq_fixup },
-	{ .compatible = "atmel,at91sam9n12", .data = at91sam9_aic_irq_fixup },
-	{ .compatible = "atmel,at91sam9rl", .data = at91sam9_aic_irq_fixup },
-	{ .compatible = "atmel,at91sam9x5", .data = at91sam9_aic_irq_fixup },
+	{ .compatible = "atmel,at91rm9200", .data = at91rm9200_aic_irq_fixup },
+	{ .compatible = "atmel,at91sam9g45", .data = at91sam9g45_aic_irq_fixup },
+	{ .compatible = "atmel,at91sam9n12", .data = at91rm9200_aic_irq_fixup },
+	{ .compatible = "atmel,at91sam9rl", .data = at91sam9g45_aic_irq_fixup },
+	{ .compatible = "atmel,at91sam9x5", .data = at91rm9200_aic_irq_fixup },
+	{ .compatible = "atmel,at91sam9260", .data = at91sam9260_aic_irq_fixup },
+	{ .compatible = "atmel,at91sam9261", .data = at91sam9260_aic_irq_fixup },
+	{ .compatible = "atmel,at91sam9263", .data = at91sam9260_aic_irq_fixup },
+	{ .compatible = "atmel,at91sam9g20", .data = at91sam9260_aic_irq_fixup },
 	{ /* sentinel */ },
 };
 
diff --git a/drivers/irqchip/irq-atmel-aic5.c b/drivers/irqchip/irq-atmel-aic5.c
index edb227081524..a2e8c3f876cb 100644
--- a/drivers/irqchip/irq-atmel-aic5.c
+++ b/drivers/irqchip/irq-atmel-aic5.c
@@ -75,15 +75,13 @@ aic5_handle(struct pt_regs *regs)
 	u32 irqnr;
 	u32 irqstat;
 
-	irqnr = irq_reg_readl(gc->reg_base + AT91_AIC5_IVR);
-	irqstat = irq_reg_readl(gc->reg_base + AT91_AIC5_ISR);
-
-	irqnr = irq_find_mapping(aic5_domain, irqnr);
+	irqnr = irq_reg_readl(gc, AT91_AIC5_IVR);
+	irqstat = irq_reg_readl(gc, AT91_AIC5_ISR);
 
 	if (!irqstat)
-		irq_reg_writel(0, gc->reg_base + AT91_AIC5_EOICR);
+		irq_reg_writel(gc, 0, AT91_AIC5_EOICR);
 	else
-		handle_IRQ(irqnr, regs);
+		handle_domain_irq(aic5_domain, irqnr, regs);
 }
 
 static void aic5_mask(struct irq_data *d)
@@ -94,8 +92,8 @@ static void aic5_mask(struct irq_data *d)
 
 	/* Disable interrupt on AIC5 */
 	irq_gc_lock(gc);
-	irq_reg_writel(d->hwirq, gc->reg_base + AT91_AIC5_SSR);
-	irq_reg_writel(1, gc->reg_base + AT91_AIC5_IDCR);
+	irq_reg_writel(gc, d->hwirq, AT91_AIC5_SSR);
+	irq_reg_writel(gc, 1, AT91_AIC5_IDCR);
 	gc->mask_cache &= ~d->mask;
 	irq_gc_unlock(gc);
 }
@@ -108,8 +106,8 @@ static void aic5_unmask(struct irq_data *d)
 
 	/* Enable interrupt on AIC5 */
 	irq_gc_lock(gc);
-	irq_reg_writel(d->hwirq, gc->reg_base + AT91_AIC5_SSR);
-	irq_reg_writel(1, gc->reg_base + AT91_AIC5_IECR);
+	irq_reg_writel(gc, d->hwirq, AT91_AIC5_SSR);
+	irq_reg_writel(gc, 1, AT91_AIC5_IECR);
 	gc->mask_cache |= d->mask;
 	irq_gc_unlock(gc);
 }
@@ -122,8 +120,8 @@ static int aic5_retrigger(struct irq_data *d)
 
 	/* Enable interrupt on AIC5 */
 	irq_gc_lock(gc);
-	irq_reg_writel(d->hwirq, gc->reg_base + AT91_AIC5_SSR);
-	irq_reg_writel(1, gc->reg_base + AT91_AIC5_ISCR);
+	irq_reg_writel(gc, d->hwirq, AT91_AIC5_SSR);
+	irq_reg_writel(gc, 1, AT91_AIC5_ISCR);
 	irq_gc_unlock(gc);
 
 	return 0;
@@ -138,11 +136,11 @@ static int aic5_set_type(struct irq_data *d, unsigned type)
 	int ret;
 
 	irq_gc_lock(gc);
-	irq_reg_writel(d->hwirq, gc->reg_base + AT91_AIC5_SSR);
-	smr = irq_reg_readl(gc->reg_base + AT91_AIC5_SMR);
+	irq_reg_writel(gc, d->hwirq, AT91_AIC5_SSR);
+	smr = irq_reg_readl(gc, AT91_AIC5_SMR);
 	ret = aic_common_set_type(d, type, &smr);
 	if (!ret)
-		irq_reg_writel(smr, gc->reg_base + AT91_AIC5_SMR);
+		irq_reg_writel(gc, smr, AT91_AIC5_SMR);
 	irq_gc_unlock(gc);
 
 	return ret;
@@ -164,12 +162,11 @@ static void aic5_suspend(struct irq_data *d)
 		if ((mask & gc->mask_cache) == (mask & gc->wake_active))
 			continue;
 
-		irq_reg_writel(i + gc->irq_base,
-			       bgc->reg_base + AT91_AIC5_SSR);
+		irq_reg_writel(bgc, i + gc->irq_base, AT91_AIC5_SSR);
 		if (mask & gc->wake_active)
-			irq_reg_writel(1, bgc->reg_base + AT91_AIC5_IECR);
+			irq_reg_writel(bgc, 1, AT91_AIC5_IECR);
 		else
-			irq_reg_writel(1, bgc->reg_base + AT91_AIC5_IDCR);
+			irq_reg_writel(bgc, 1, AT91_AIC5_IDCR);
 	}
 	irq_gc_unlock(bgc);
 }
@@ -189,12 +186,11 @@ static void aic5_resume(struct irq_data *d)
 		if ((mask & gc->mask_cache) == (mask & gc->wake_active))
 			continue;
 
-		irq_reg_writel(i + gc->irq_base,
-			       bgc->reg_base + AT91_AIC5_SSR);
+		irq_reg_writel(bgc, i + gc->irq_base, AT91_AIC5_SSR);
 		if (mask & gc->mask_cache)
-			irq_reg_writel(1, bgc->reg_base + AT91_AIC5_IECR);
+			irq_reg_writel(bgc, 1, AT91_AIC5_IECR);
 		else
-			irq_reg_writel(1, bgc->reg_base + AT91_AIC5_IDCR);
+			irq_reg_writel(bgc, 1, AT91_AIC5_IDCR);
 	}
 	irq_gc_unlock(bgc);
 }
@@ -209,10 +205,9 @@ static void aic5_pm_shutdown(struct irq_data *d)
 
 	irq_gc_lock(bgc);
 	for (i = 0; i < dgc->irqs_per_chip; i++) {
-		irq_reg_writel(i + gc->irq_base,
-			       bgc->reg_base + AT91_AIC5_SSR);
-		irq_reg_writel(1, bgc->reg_base + AT91_AIC5_IDCR);
-		irq_reg_writel(1, bgc->reg_base + AT91_AIC5_ICCR);
+		irq_reg_writel(bgc, i + gc->irq_base, AT91_AIC5_SSR);
+		irq_reg_writel(bgc, 1, AT91_AIC5_IDCR);
+		irq_reg_writel(bgc, 1, AT91_AIC5_ICCR);
 	}
 	irq_gc_unlock(bgc);
 }
@@ -232,24 +227,24 @@ static void __init aic5_hw_init(struct irq_domain *domain)
 	 * will not Lock out nIRQ
 	 */
 	for (i = 0; i < 8; i++)
-		irq_reg_writel(0, gc->reg_base + AT91_AIC5_EOICR);
+		irq_reg_writel(gc, 0, AT91_AIC5_EOICR);
 
 	/*
 	 * Spurious Interrupt ID in Spurious Vector Register.
 	 * When there is no current interrupt, the IRQ Vector Register
 	 * reads the value stored in AIC_SPU
 	 */
-	irq_reg_writel(0xffffffff, gc->reg_base + AT91_AIC5_SPU);
+	irq_reg_writel(gc, 0xffffffff, AT91_AIC5_SPU);
 
 	/* No debugging in AIC: Debug (Protect) Control Register */
-	irq_reg_writel(0, gc->reg_base + AT91_AIC5_DCR);
+	irq_reg_writel(gc, 0, AT91_AIC5_DCR);
 
 	/* Disable and clear all interrupts initially */
 	for (i = 0; i < domain->revmap_size; i++) {
-		irq_reg_writel(i, gc->reg_base + AT91_AIC5_SSR);
-		irq_reg_writel(i, gc->reg_base + AT91_AIC5_SVR);
-		irq_reg_writel(1, gc->reg_base + AT91_AIC5_IDCR);
-		irq_reg_writel(1, gc->reg_base + AT91_AIC5_ICCR);
+		irq_reg_writel(gc, i, AT91_AIC5_SSR);
+		irq_reg_writel(gc, i, AT91_AIC5_SVR);
+		irq_reg_writel(gc, 1, AT91_AIC5_IDCR);
+		irq_reg_writel(gc, 1, AT91_AIC5_ICCR);
 	}
 }
 
@@ -275,11 +270,11 @@ static int aic5_irq_domain_xlate(struct irq_domain *d,
 	gc = dgc->gc[0];
 
 	irq_gc_lock(gc);
-	irq_reg_writel(*out_hwirq, gc->reg_base + AT91_AIC5_SSR);
-	smr = irq_reg_readl(gc->reg_base + AT91_AIC5_SMR);
+	irq_reg_writel(gc, *out_hwirq, AT91_AIC5_SSR);
+	smr = irq_reg_readl(gc, AT91_AIC5_SMR);
 	ret = aic_common_set_priority(intspec[2], &smr);
 	if (!ret)
-		irq_reg_writel(intspec[2] | smr, gc->reg_base + AT91_AIC5_SMR);
+		irq_reg_writel(gc, intspec[2] | smr, AT91_AIC5_SMR);
 	irq_gc_unlock(gc);
 
 	return ret;
@@ -297,6 +292,7 @@ static void __init sama5d3_aic_irq_fixup(struct device_node *root)
 
 static const struct of_device_id __initdata aic5_irq_fixups[] = {
 	{ .compatible = "atmel,sama5d3", .data = sama5d3_aic_irq_fixup },
+	{ .compatible = "atmel,sama5d4", .data = sama5d3_aic_irq_fixup },
 	{ /* sentinel */ },
 };
 
@@ -343,7 +339,7 @@ static int __init aic5_of_init(struct device_node *node,
 	return 0;
 }
 
-#define NR_SAMA5D3_IRQS		50
+#define NR_SAMA5D3_IRQS		48
 
 static int __init sama5d3_aic5_of_init(struct device_node *node,
 				       struct device_node *parent)
@@ -351,3 +347,12 @@ static int __init sama5d3_aic5_of_init(struct device_node *node,
 	return aic5_of_init(node, parent, NR_SAMA5D3_IRQS);
 }
 IRQCHIP_DECLARE(sama5d3_aic5, "atmel,sama5d3-aic", sama5d3_aic5_of_init);
+
+#define NR_SAMA5D4_IRQS		68
+
+static int __init sama5d4_aic5_of_init(struct device_node *node,
+				       struct device_node *parent)
+{
+	return aic5_of_init(node, parent, NR_SAMA5D4_IRQS);
+}
+IRQCHIP_DECLARE(sama5d4_aic5, "atmel,sama5d4-aic", sama5d4_aic5_of_init);
diff --git a/drivers/irqchip/irq-bcm7120-l2.c b/drivers/irqchip/irq-bcm7120-l2.c
new file mode 100644
index 000000000000..8eec8e1201d9
--- /dev/null
+++ b/drivers/irqchip/irq-bcm7120-l2.c
@@ -0,0 +1,255 @@
+/*
+ * Broadcom BCM7120 style Level 2 interrupt controller driver
+ *
+ * Copyright (C) 2014 Broadcom Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#define pr_fmt(fmt)	KBUILD_MODNAME	": " fmt
+
+#include <linux/init.h>
+#include <linux/slab.h>
+#include <linux/module.h>
+#include <linux/kconfig.h>
+#include <linux/platform_device.h>
+#include <linux/of.h>
+#include <linux/of_irq.h>
+#include <linux/of_address.h>
+#include <linux/of_platform.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/io.h>
+#include <linux/irqdomain.h>
+#include <linux/reboot.h>
+#include <linux/bitops.h>
+#include <linux/irqchip/chained_irq.h>
+
+#include "irqchip.h"
+
+/* Register offset in the L2 interrupt controller */
+#define IRQEN		0x00
+#define IRQSTAT		0x04
+
+#define MAX_WORDS	4
+#define IRQS_PER_WORD	32
+
+struct bcm7120_l2_intc_data {
+	unsigned int n_words;
+	void __iomem *base[MAX_WORDS];
+	struct irq_domain *domain;
+	bool can_wake;
+	u32 irq_fwd_mask[MAX_WORDS];
+	u32 irq_map_mask[MAX_WORDS];
+};
+
+static void bcm7120_l2_intc_irq_handle(unsigned int irq, struct irq_desc *desc)
+{
+	struct bcm7120_l2_intc_data *b = irq_desc_get_handler_data(desc);
+	struct irq_chip *chip = irq_desc_get_chip(desc);
+	unsigned int idx;
+
+	chained_irq_enter(chip, desc);
+
+	for (idx = 0; idx < b->n_words; idx++) {
+		int base = idx * IRQS_PER_WORD;
+		struct irq_chip_generic *gc =
+			irq_get_domain_generic_chip(b->domain, base);
+		unsigned long pending;
+		int hwirq;
+
+		irq_gc_lock(gc);
+		pending = irq_reg_readl(gc, IRQSTAT) & gc->mask_cache;
+		irq_gc_unlock(gc);
+
+		for_each_set_bit(hwirq, &pending, IRQS_PER_WORD) {
+			generic_handle_irq(irq_find_mapping(b->domain,
+					   base + hwirq));
+		}
+	}
+
+	chained_irq_exit(chip, desc);
+}
+
+static void bcm7120_l2_intc_suspend(struct irq_data *d)
+{
+	struct irq_chip_generic *gc = irq_data_get_irq_chip_data(d);
+	struct bcm7120_l2_intc_data *b = gc->private;
+
+	irq_gc_lock(gc);
+	if (b->can_wake)
+		irq_reg_writel(gc, gc->mask_cache | gc->wake_active, IRQEN);
+	irq_gc_unlock(gc);
+}
+
+static void bcm7120_l2_intc_resume(struct irq_data *d)
+{
+	struct irq_chip_generic *gc = irq_data_get_irq_chip_data(d);
+
+	/* Restore the saved mask */
+	irq_gc_lock(gc);
+	irq_reg_writel(gc, gc->mask_cache, IRQEN);
+	irq_gc_unlock(gc);
+}
+
+static int bcm7120_l2_intc_init_one(struct device_node *dn,
+					struct bcm7120_l2_intc_data *data,
+					int irq, const __be32 *map_mask)
+{
+	int parent_irq;
+	unsigned int idx;
+
+	parent_irq = irq_of_parse_and_map(dn, irq);
+	if (!parent_irq) {
+		pr_err("failed to map interrupt %d\n", irq);
+		return -EINVAL;
+	}
+
+	/* For multiple parent IRQs with multiple words, this looks like:
+	 * <irq0_w0 irq0_w1 irq1_w0 irq1_w1 ...>
+	 */
+	for (idx = 0; idx < data->n_words; idx++)
+		data->irq_map_mask[idx] |=
+			be32_to_cpup(map_mask + irq * data->n_words + idx);
+
+	irq_set_handler_data(parent_irq, data);
+	irq_set_chained_handler(parent_irq, bcm7120_l2_intc_irq_handle);
+
+	return 0;
+}
+
+int __init bcm7120_l2_intc_of_init(struct device_node *dn,
+					struct device_node *parent)
+{
+	unsigned int clr = IRQ_NOREQUEST | IRQ_NOPROBE | IRQ_NOAUTOEN;
+	struct bcm7120_l2_intc_data *data;
+	struct irq_chip_generic *gc;
+	struct irq_chip_type *ct;
+	const __be32 *map_mask;
+	int num_parent_irqs;
+	int ret = 0, len;
+	unsigned int idx, irq, flags;
+
+	data = kzalloc(sizeof(*data), GFP_KERNEL);
+	if (!data)
+		return -ENOMEM;
+
+	for (idx = 0; idx < MAX_WORDS; idx++) {
+		data->base[idx] = of_iomap(dn, idx);
+		if (!data->base[idx])
+			break;
+		data->n_words = idx + 1;
+	}
+	if (!data->n_words) {
+		pr_err("failed to remap intc L2 registers\n");
+		ret = -ENOMEM;
+		goto out_unmap;
+	}
+
+	/* Enable all interrupts specified in the interrupt forward mask;
+	 * disable all others.  If the property doesn't exist (-EINVAL),
+	 * assume all zeroes.
+	 */
+	ret = of_property_read_u32_array(dn, "brcm,int-fwd-mask",
+					 data->irq_fwd_mask, data->n_words);
+	if (ret == 0 || ret == -EINVAL) {
+		for (idx = 0; idx < data->n_words; idx++)
+			__raw_writel(data->irq_fwd_mask[idx],
+				     data->base[idx] + IRQEN);
+	} else {
+		/* property exists but has the wrong number of words */
+		pr_err("invalid int-fwd-mask property\n");
+		ret = -EINVAL;
+		goto out_unmap;
+	}
+
+	num_parent_irqs = of_irq_count(dn);
+	if (num_parent_irqs <= 0) {
+		pr_err("invalid number of parent interrupts\n");
+		ret = -ENOMEM;
+		goto out_unmap;
+	}
+
+	map_mask = of_get_property(dn, "brcm,int-map-mask", &len);
+	if (!map_mask ||
+	    (len != (sizeof(*map_mask) * num_parent_irqs * data->n_words))) {
+		pr_err("invalid brcm,int-map-mask property\n");
+		ret = -EINVAL;
+		goto out_unmap;
+	}
+
+	for (irq = 0; irq < num_parent_irqs; irq++) {
+		ret = bcm7120_l2_intc_init_one(dn, data, irq, map_mask);
+		if (ret)
+			goto out_unmap;
+	}
+
+	data->domain = irq_domain_add_linear(dn, IRQS_PER_WORD * data->n_words,
+					     &irq_generic_chip_ops, NULL);
+	if (!data->domain) {
+		ret = -ENOMEM;
+		goto out_unmap;
+	}
+
+	/* MIPS chips strapped for BE will automagically configure the
+	 * peripheral registers for CPU-native byte order.
+	 */
+	flags = IRQ_GC_INIT_MASK_CACHE;
+	if (IS_ENABLED(CONFIG_MIPS) && IS_ENABLED(CONFIG_CPU_BIG_ENDIAN))
+		flags |= IRQ_GC_BE_IO;
+
+	ret = irq_alloc_domain_generic_chips(data->domain, IRQS_PER_WORD, 1,
+				dn->full_name, handle_level_irq, clr, 0, flags);
+	if (ret) {
+		pr_err("failed to allocate generic irq chip\n");
+		goto out_free_domain;
+	}
+
+	if (of_property_read_bool(dn, "brcm,irq-can-wake"))
+		data->can_wake = true;
+
+	for (idx = 0; idx < data->n_words; idx++) {
+		irq = idx * IRQS_PER_WORD;
+		gc = irq_get_domain_generic_chip(data->domain, irq);
+
+		gc->unused = 0xffffffff & ~data->irq_map_mask[idx];
+		gc->reg_base = data->base[idx];
+		gc->private = data;
+		ct = gc->chip_types;
+
+		ct->regs.mask = IRQEN;
+		ct->chip.irq_mask = irq_gc_mask_clr_bit;
+		ct->chip.irq_unmask = irq_gc_mask_set_bit;
+		ct->chip.irq_ack = irq_gc_noop;
+		ct->chip.irq_suspend = bcm7120_l2_intc_suspend;
+		ct->chip.irq_resume = bcm7120_l2_intc_resume;
+
+		if (data->can_wake) {
+			/* This IRQ chip can wake the system, set all
+			 * relevant child interupts in wake_enabled mask
+			 */
+			gc->wake_enabled = 0xffffffff;
+			gc->wake_enabled &= ~gc->unused;
+			ct->chip.irq_set_wake = irq_gc_set_wake;
+		}
+	}
+
+	pr_info("registered BCM7120 L2 intc (mem: 0x%p, parent IRQ(s): %d)\n",
+			data->base[0], num_parent_irqs);
+
+	return 0;
+
+out_free_domain:
+	irq_domain_remove(data->domain);
+out_unmap:
+	for (idx = 0; idx < MAX_WORDS; idx++) {
+		if (data->base[idx])
+			iounmap(data->base[idx]);
+	}
+	kfree(data);
+	return ret;
+}
+IRQCHIP_DECLARE(bcm7120_l2_intc, "brcm,bcm7120-l2-intc",
+		bcm7120_l2_intc_of_init);
diff --git a/drivers/irqchip/irq-brcmstb-l2.c b/drivers/irqchip/irq-brcmstb-l2.c
index c15c840987d2..313c2c64498a 100644
--- a/drivers/irqchip/irq-brcmstb-l2.c
+++ b/drivers/irqchip/irq-brcmstb-l2.c
@@ -18,7 +18,9 @@
 #include <linux/init.h>
 #include <linux/slab.h>
 #include <linux/module.h>
+#include <linux/kconfig.h>
 #include <linux/platform_device.h>
+#include <linux/spinlock.h>
 #include <linux/of.h>
 #include <linux/of_irq.h>
 #include <linux/of_address.h>
@@ -30,8 +32,6 @@
 #include <linux/irqchip.h>
 #include <linux/irqchip/chained_irq.h>
 
-#include <asm/mach/irq.h>
-
 #include "irqchip.h"
 
 /* Register offsets in the L2 interrupt controller */
@@ -54,23 +54,26 @@ struct brcmstb_l2_intc_data {
 static void brcmstb_l2_intc_irq_handle(unsigned int irq, struct irq_desc *desc)
 {
 	struct brcmstb_l2_intc_data *b = irq_desc_get_handler_data(desc);
+	struct irq_chip_generic *gc = irq_get_domain_generic_chip(b->domain, 0);
 	struct irq_chip *chip = irq_desc_get_chip(desc);
 	u32 status;
 
 	chained_irq_enter(chip, desc);
 
-	status = __raw_readl(b->base + CPU_STATUS) &
-		~(__raw_readl(b->base + CPU_MASK_STATUS));
+	status = irq_reg_readl(gc, CPU_STATUS) &
+		~(irq_reg_readl(gc, CPU_MASK_STATUS));
 
 	if (status == 0) {
-		do_bad_IRQ(irq, desc);
+		raw_spin_lock(&desc->lock);
+		handle_bad_irq(irq, desc);
+		raw_spin_unlock(&desc->lock);
 		goto out;
 	}
 
 	do {
 		irq = ffs(status) - 1;
 		/* ack at our level */
-		__raw_writel(1 << irq, b->base + CPU_CLEAR);
+		irq_reg_writel(gc, 1 << irq, CPU_CLEAR);
 		status &= ~(1 << irq);
 		generic_handle_irq(irq_find_mapping(b->domain, irq));
 	} while (status);
@@ -85,12 +88,12 @@ static void brcmstb_l2_intc_suspend(struct irq_data *d)
 
 	irq_gc_lock(gc);
 	/* Save the current mask */
-	b->saved_mask = __raw_readl(b->base + CPU_MASK_STATUS);
+	b->saved_mask = irq_reg_readl(gc, CPU_MASK_STATUS);
 
 	if (b->can_wake) {
 		/* Program the wakeup mask */
-		__raw_writel(~gc->wake_active, b->base + CPU_MASK_SET);
-		__raw_writel(gc->wake_active, b->base + CPU_MASK_CLEAR);
+		irq_reg_writel(gc, ~gc->wake_active, CPU_MASK_SET);
+		irq_reg_writel(gc, gc->wake_active, CPU_MASK_CLEAR);
 	}
 	irq_gc_unlock(gc);
 }
@@ -102,11 +105,11 @@ static void brcmstb_l2_intc_resume(struct irq_data *d)
 
 	irq_gc_lock(gc);
 	/* Clear unmasked non-wakeup interrupts */
-	__raw_writel(~b->saved_mask & ~gc->wake_active, b->base + CPU_CLEAR);
+	irq_reg_writel(gc, ~b->saved_mask & ~gc->wake_active, CPU_CLEAR);
 
 	/* Restore the saved mask */
-	__raw_writel(b->saved_mask, b->base + CPU_MASK_SET);
-	__raw_writel(~b->saved_mask, b->base + CPU_MASK_CLEAR);
+	irq_reg_writel(gc, b->saved_mask, CPU_MASK_SET);
+	irq_reg_writel(gc, ~b->saved_mask, CPU_MASK_CLEAR);
 	irq_gc_unlock(gc);
 }
 
@@ -118,6 +121,7 @@ int __init brcmstb_l2_intc_of_init(struct device_node *np,
 	struct irq_chip_generic *gc;
 	struct irq_chip_type *ct;
 	int ret;
+	unsigned int flags;
 
 	data = kzalloc(sizeof(*data), GFP_KERNEL);
 	if (!data)
@@ -131,13 +135,13 @@ int __init brcmstb_l2_intc_of_init(struct device_node *np,
 	}
 
 	/* Disable all interrupts by default */
-	__raw_writel(0xffffffff, data->base + CPU_MASK_SET);
-	__raw_writel(0xffffffff, data->base + CPU_CLEAR);
+	writel(0xffffffff, data->base + CPU_MASK_SET);
+	writel(0xffffffff, data->base + CPU_CLEAR);
 
 	data->parent_irq = irq_of_parse_and_map(np, 0);
-	if (data->parent_irq < 0) {
+	if (!data->parent_irq) {
 		pr_err("failed to find parent interrupt\n");
-		ret = data->parent_irq;
+		ret = -EINVAL;
 		goto out_unmap;
 	}
 
@@ -148,9 +152,16 @@ int __init brcmstb_l2_intc_of_init(struct device_node *np,
 		goto out_unmap;
 	}
 
+	/* MIPS chips strapped for BE will automagically configure the
+	 * peripheral registers for CPU-native byte order.
+	 */
+	flags = 0;
+	if (IS_ENABLED(CONFIG_MIPS) && IS_ENABLED(CONFIG_CPU_BIG_ENDIAN))
+		flags |= IRQ_GC_BE_IO;
+
 	/* Allocate a single Generic IRQ chip for this node */
 	ret = irq_alloc_domain_generic_chips(data->domain, 32, 1,
-				np->full_name, handle_edge_irq, clr, 0, 0);
+				np->full_name, handle_edge_irq, clr, 0, flags);
 	if (ret) {
 		pr_err("failed to allocate generic irq chip\n");
 		goto out_free_domain;
diff --git a/drivers/irqchip/irq-clps711x.c b/drivers/irqchip/irq-clps711x.c
index 33340dc97d1d..33127f131d78 100644
--- a/drivers/irqchip/irq-clps711x.c
+++ b/drivers/irqchip/irq-clps711x.c
@@ -76,24 +76,20 @@ static struct {
 
 static asmlinkage void __exception_irq_entry clps711x_irqh(struct pt_regs *regs)
 {
-	u32 irqnr, irqstat;
+	u32 irqstat;
 
 	do {
 		irqstat = readw_relaxed(clps711x_intc->intmr[0]) &
 			  readw_relaxed(clps711x_intc->intsr[0]);
-		if (irqstat) {
-			irqnr =	irq_find_mapping(clps711x_intc->domain,
-						 fls(irqstat) - 1);
-			handle_IRQ(irqnr, regs);
-		}
+		if (irqstat)
+			handle_domain_irq(clps711x_intc->domain,
+					  fls(irqstat) - 1, regs);
 
 		irqstat = readw_relaxed(clps711x_intc->intmr[1]) &
 			  readw_relaxed(clps711x_intc->intsr[1]);
-		if (irqstat) {
-			irqnr =	irq_find_mapping(clps711x_intc->domain,
-						 fls(irqstat) - 1 + 16);
-			handle_IRQ(irqnr, regs);
-		}
+		if (irqstat)
+			handle_domain_irq(clps711x_intc->domain,
+					  fls(irqstat) - 1 + 16, regs);
 	} while (irqstat);
 }
 
diff --git a/drivers/irqchip/irq-dw-apb-ictl.c b/drivers/irqchip/irq-dw-apb-ictl.c
index 31e231e1f566..53bb7326a60a 100644
--- a/drivers/irqchip/irq-dw-apb-ictl.c
+++ b/drivers/irqchip/irq-dw-apb-ictl.c
@@ -50,6 +50,21 @@ static void dw_apb_ictl_handler(unsigned int irq, struct irq_desc *desc)
 	chained_irq_exit(chip, desc);
 }
 
+#ifdef CONFIG_PM
+static void dw_apb_ictl_resume(struct irq_data *d)
+{
+	struct irq_chip_generic *gc = irq_data_get_irq_chip_data(d);
+	struct irq_chip_type *ct = irq_data_get_chip_type(d);
+
+	irq_gc_lock(gc);
+	writel_relaxed(~0, gc->reg_base + ct->regs.enable);
+	writel_relaxed(*ct->mask_cache, gc->reg_base + ct->regs.mask);
+	irq_gc_unlock(gc);
+}
+#else
+#define dw_apb_ictl_resume	NULL
+#endif /* CONFIG_PM */
+
 static int __init dw_apb_ictl_init(struct device_node *np,
 				   struct device_node *parent)
 {
@@ -94,16 +109,16 @@ static int __init dw_apb_ictl_init(struct device_node *np,
 	 */
 
 	/* mask and enable all interrupts */
-	writel(~0, iobase + APB_INT_MASK_L);
-	writel(~0, iobase + APB_INT_MASK_H);
-	writel(~0, iobase + APB_INT_ENABLE_L);
-	writel(~0, iobase + APB_INT_ENABLE_H);
+	writel_relaxed(~0, iobase + APB_INT_MASK_L);
+	writel_relaxed(~0, iobase + APB_INT_MASK_H);
+	writel_relaxed(~0, iobase + APB_INT_ENABLE_L);
+	writel_relaxed(~0, iobase + APB_INT_ENABLE_H);
 
-	reg = readl(iobase + APB_INT_ENABLE_H);
+	reg = readl_relaxed(iobase + APB_INT_ENABLE_H);
 	if (reg)
 		nrirqs = 32 + fls(reg);
 	else
-		nrirqs = fls(readl(iobase + APB_INT_ENABLE_L));
+		nrirqs = fls(readl_relaxed(iobase + APB_INT_ENABLE_L));
 
 	domain = irq_domain_add_linear(np, nrirqs,
 				       &irq_generic_chip_ops, NULL);
@@ -115,6 +130,7 @@ static int __init dw_apb_ictl_init(struct device_node *np,
 
 	ret = irq_alloc_domain_generic_chips(domain, 32, (nrirqs > 32) ? 2 : 1,
 					     np->name, handle_level_irq, clr, 0,
+					     IRQ_GC_MASK_CACHE_PER_TYPE |
 					     IRQ_GC_INIT_MASK_CACHE);
 	if (ret) {
 		pr_err("%s: unable to alloc irq domain gc\n", np->full_name);
@@ -126,13 +142,17 @@ static int __init dw_apb_ictl_init(struct device_node *np,
 	gc->reg_base = iobase;
 
 	gc->chip_types[0].regs.mask = APB_INT_MASK_L;
+	gc->chip_types[0].regs.enable = APB_INT_ENABLE_L;
 	gc->chip_types[0].chip.irq_mask = irq_gc_mask_set_bit;
 	gc->chip_types[0].chip.irq_unmask = irq_gc_mask_clr_bit;
+	gc->chip_types[0].chip.irq_resume = dw_apb_ictl_resume;
 
 	if (nrirqs > 32) {
 		gc->chip_types[1].regs.mask = APB_INT_MASK_H;
+		gc->chip_types[1].regs.enable = APB_INT_ENABLE_H;
 		gc->chip_types[1].chip.irq_mask = irq_gc_mask_set_bit;
 		gc->chip_types[1].chip.irq_unmask = irq_gc_mask_clr_bit;
+		gc->chip_types[1].chip.irq_resume = dw_apb_ictl_resume;
 	}
 
 	irq_set_handler_data(irq, gc);
diff --git a/drivers/irqchip/irq-gic-common.c b/drivers/irqchip/irq-gic-common.c
index 60ac704d2090..61541ff24397 100644
--- a/drivers/irqchip/irq-gic-common.c
+++ b/drivers/irqchip/irq-gic-common.c
@@ -74,20 +74,22 @@ void __init gic_dist_config(void __iomem *base, int gic_irqs,
 	 * Set all global interrupts to be level triggered, active low.
 	 */
 	for (i = 32; i < gic_irqs; i += 16)
-		writel_relaxed(0, base + GIC_DIST_CONFIG + i / 4);
+		writel_relaxed(GICD_INT_ACTLOW_LVLTRIG,
+					base + GIC_DIST_CONFIG + i / 4);
 
 	/*
 	 * Set priority on all global interrupts.
 	 */
 	for (i = 32; i < gic_irqs; i += 4)
-		writel_relaxed(0xa0a0a0a0, base + GIC_DIST_PRI + i);
+		writel_relaxed(GICD_INT_DEF_PRI_X4, base + GIC_DIST_PRI + i);
 
 	/*
 	 * Disable all interrupts.  Leave the PPI and SGIs alone
 	 * as they are enabled by redistributor registers.
 	 */
 	for (i = 32; i < gic_irqs; i += 32)
-		writel_relaxed(0xffffffff, base + GIC_DIST_ENABLE_CLEAR + i / 8);
+		writel_relaxed(GICD_INT_EN_CLR_X32,
+					base + GIC_DIST_ENABLE_CLEAR + i / 8);
 
 	if (sync_access)
 		sync_access();
@@ -101,14 +103,15 @@ void gic_cpu_config(void __iomem *base, void (*sync_access)(void))
 	 * Deal with the banked PPI and SGI interrupts - disable all
 	 * PPI interrupts, ensure all SGI interrupts are enabled.
 	 */
-	writel_relaxed(0xffff0000, base + GIC_DIST_ENABLE_CLEAR);
-	writel_relaxed(0x0000ffff, base + GIC_DIST_ENABLE_SET);
+	writel_relaxed(GICD_INT_EN_CLR_PPI, base + GIC_DIST_ENABLE_CLEAR);
+	writel_relaxed(GICD_INT_EN_SET_SGI, base + GIC_DIST_ENABLE_SET);
 
 	/*
 	 * Set priority on PPI and SGI interrupts
 	 */
 	for (i = 0; i < 32; i += 4)
-		writel_relaxed(0xa0a0a0a0, base + GIC_DIST_PRI + i * 4 / 4);
+		writel_relaxed(GICD_INT_DEF_PRI_X4,
+					base + GIC_DIST_PRI + i * 4 / 4);
 
 	if (sync_access)
 		sync_access();
diff --git a/drivers/irqchip/irq-gic-v2m.c b/drivers/irqchip/irq-gic-v2m.c
new file mode 100644
index 000000000000..fdf706555d72
--- /dev/null
+++ b/drivers/irqchip/irq-gic-v2m.c
@@ -0,0 +1,333 @@
+/*
+ * ARM GIC v2m MSI(-X) support
+ * Support for Message Signaled Interrupts for systems that
+ * implement ARM Generic Interrupt Controller: GICv2m.
+ *
+ * Copyright (C) 2014 Advanced Micro Devices, Inc.
+ * Authors: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
+ *	    Harish Kasiviswanathan <harish.kasiviswanathan@amd.com>
+ *	    Brandon Anderson <brandon.anderson@amd.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published
+ * by the Free Software Foundation.
+ */
+
+#define pr_fmt(fmt) "GICv2m: " fmt
+
+#include <linux/irq.h>
+#include <linux/irqdomain.h>
+#include <linux/kernel.h>
+#include <linux/of_address.h>
+#include <linux/of_pci.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+
+/*
+* MSI_TYPER:
+*     [31:26] Reserved
+*     [25:16] lowest SPI assigned to MSI
+*     [15:10] Reserved
+*     [9:0]   Numer of SPIs assigned to MSI
+*/
+#define V2M_MSI_TYPER		       0x008
+#define V2M_MSI_TYPER_BASE_SHIFT       16
+#define V2M_MSI_TYPER_BASE_MASK	       0x3FF
+#define V2M_MSI_TYPER_NUM_MASK	       0x3FF
+#define V2M_MSI_SETSPI_NS	       0x040
+#define V2M_MIN_SPI		       32
+#define V2M_MAX_SPI		       1019
+
+#define V2M_MSI_TYPER_BASE_SPI(x)      \
+	       (((x) >> V2M_MSI_TYPER_BASE_SHIFT) & V2M_MSI_TYPER_BASE_MASK)
+
+#define V2M_MSI_TYPER_NUM_SPI(x)       ((x) & V2M_MSI_TYPER_NUM_MASK)
+
+struct v2m_data {
+	spinlock_t msi_cnt_lock;
+	struct msi_controller mchip;
+	struct resource res;	/* GICv2m resource */
+	void __iomem *base;	/* GICv2m virt address */
+	u32 spi_start;		/* The SPI number that MSIs start */
+	u32 nr_spis;		/* The number of SPIs for MSIs */
+	unsigned long *bm;	/* MSI vector bitmap */
+	struct irq_domain *domain;
+};
+
+static void gicv2m_mask_msi_irq(struct irq_data *d)
+{
+	pci_msi_mask_irq(d);
+	irq_chip_mask_parent(d);
+}
+
+static void gicv2m_unmask_msi_irq(struct irq_data *d)
+{
+	pci_msi_unmask_irq(d);
+	irq_chip_unmask_parent(d);
+}
+
+static struct irq_chip gicv2m_msi_irq_chip = {
+	.name			= "MSI",
+	.irq_mask		= gicv2m_mask_msi_irq,
+	.irq_unmask		= gicv2m_unmask_msi_irq,
+	.irq_eoi		= irq_chip_eoi_parent,
+	.irq_write_msi_msg	= pci_msi_domain_write_msg,
+};
+
+static struct msi_domain_info gicv2m_msi_domain_info = {
+	.flags	= (MSI_FLAG_USE_DEF_DOM_OPS | MSI_FLAG_USE_DEF_CHIP_OPS |
+		   MSI_FLAG_PCI_MSIX),
+	.chip	= &gicv2m_msi_irq_chip,
+};
+
+static int gicv2m_set_affinity(struct irq_data *irq_data,
+			       const struct cpumask *mask, bool force)
+{
+	int ret;
+
+	ret = irq_chip_set_affinity_parent(irq_data, mask, force);
+	if (ret == IRQ_SET_MASK_OK)
+		ret = IRQ_SET_MASK_OK_DONE;
+
+	return ret;
+}
+
+static void gicv2m_compose_msi_msg(struct irq_data *data, struct msi_msg *msg)
+{
+	struct v2m_data *v2m = irq_data_get_irq_chip_data(data);
+	phys_addr_t addr = v2m->res.start + V2M_MSI_SETSPI_NS;
+
+	msg->address_hi = (u32) (addr >> 32);
+	msg->address_lo = (u32) (addr);
+	msg->data = data->hwirq;
+}
+
+static struct irq_chip gicv2m_irq_chip = {
+	.name			= "GICv2m",
+	.irq_mask		= irq_chip_mask_parent,
+	.irq_unmask		= irq_chip_unmask_parent,
+	.irq_eoi		= irq_chip_eoi_parent,
+	.irq_set_affinity	= gicv2m_set_affinity,
+	.irq_compose_msi_msg	= gicv2m_compose_msi_msg,
+};
+
+static int gicv2m_irq_gic_domain_alloc(struct irq_domain *domain,
+				       unsigned int virq,
+				       irq_hw_number_t hwirq)
+{
+	struct of_phandle_args args;
+	struct irq_data *d;
+	int err;
+
+	args.np = domain->parent->of_node;
+	args.args_count = 3;
+	args.args[0] = 0;
+	args.args[1] = hwirq - 32;
+	args.args[2] = IRQ_TYPE_EDGE_RISING;
+
+	err = irq_domain_alloc_irqs_parent(domain, virq, 1, &args);
+	if (err)
+		return err;
+
+	/* Configure the interrupt line to be edge */
+	d = irq_domain_get_irq_data(domain->parent, virq);
+	d->chip->irq_set_type(d, IRQ_TYPE_EDGE_RISING);
+	return 0;
+}
+
+static void gicv2m_unalloc_msi(struct v2m_data *v2m, unsigned int hwirq)
+{
+	int pos;
+
+	pos = hwirq - v2m->spi_start;
+	if (pos < 0 || pos >= v2m->nr_spis) {
+		pr_err("Failed to teardown msi. Invalid hwirq %d\n", hwirq);
+		return;
+	}
+
+	spin_lock(&v2m->msi_cnt_lock);
+	__clear_bit(pos, v2m->bm);
+	spin_unlock(&v2m->msi_cnt_lock);
+}
+
+static int gicv2m_irq_domain_alloc(struct irq_domain *domain, unsigned int virq,
+				   unsigned int nr_irqs, void *args)
+{
+	struct v2m_data *v2m = domain->host_data;
+	int hwirq, offset, err = 0;
+
+	spin_lock(&v2m->msi_cnt_lock);
+	offset = find_first_zero_bit(v2m->bm, v2m->nr_spis);
+	if (offset < v2m->nr_spis)
+		__set_bit(offset, v2m->bm);
+	else
+		err = -ENOSPC;
+	spin_unlock(&v2m->msi_cnt_lock);
+
+	if (err)
+		return err;
+
+	hwirq = v2m->spi_start + offset;
+
+	err = gicv2m_irq_gic_domain_alloc(domain, virq, hwirq);
+	if (err) {
+		gicv2m_unalloc_msi(v2m, hwirq);
+		return err;
+	}
+
+	irq_domain_set_hwirq_and_chip(domain, virq, hwirq,
+				      &gicv2m_irq_chip, v2m);
+
+	return 0;
+}
+
+static void gicv2m_irq_domain_free(struct irq_domain *domain,
+				   unsigned int virq, unsigned int nr_irqs)
+{
+	struct irq_data *d = irq_domain_get_irq_data(domain, virq);
+	struct v2m_data *v2m = irq_data_get_irq_chip_data(d);
+
+	BUG_ON(nr_irqs != 1);
+	gicv2m_unalloc_msi(v2m, d->hwirq);
+	irq_domain_free_irqs_parent(domain, virq, nr_irqs);
+}
+
+static const struct irq_domain_ops gicv2m_domain_ops = {
+	.alloc			= gicv2m_irq_domain_alloc,
+	.free			= gicv2m_irq_domain_free,
+};
+
+static bool is_msi_spi_valid(u32 base, u32 num)
+{
+	if (base < V2M_MIN_SPI) {
+		pr_err("Invalid MSI base SPI (base:%u)\n", base);
+		return false;
+	}
+
+	if ((num == 0) || (base + num > V2M_MAX_SPI)) {
+		pr_err("Number of SPIs (%u) exceed maximum (%u)\n",
+		       num, V2M_MAX_SPI - V2M_MIN_SPI + 1);
+		return false;
+	}
+
+	return true;
+}
+
+static int __init gicv2m_init_one(struct device_node *node,
+				  struct irq_domain *parent)
+{
+	int ret;
+	struct v2m_data *v2m;
+
+	v2m = kzalloc(sizeof(struct v2m_data), GFP_KERNEL);
+	if (!v2m) {
+		pr_err("Failed to allocate struct v2m_data.\n");
+		return -ENOMEM;
+	}
+
+	ret = of_address_to_resource(node, 0, &v2m->res);
+	if (ret) {
+		pr_err("Failed to allocate v2m resource.\n");
+		goto err_free_v2m;
+	}
+
+	v2m->base = ioremap(v2m->res.start, resource_size(&v2m->res));
+	if (!v2m->base) {
+		pr_err("Failed to map GICv2m resource\n");
+		ret = -ENOMEM;
+		goto err_free_v2m;
+	}
+
+	if (!of_property_read_u32(node, "arm,msi-base-spi", &v2m->spi_start) &&
+	    !of_property_read_u32(node, "arm,msi-num-spis", &v2m->nr_spis)) {
+		pr_info("Overriding V2M MSI_TYPER (base:%u, num:%u)\n",
+			v2m->spi_start, v2m->nr_spis);
+	} else {
+		u32 typer = readl_relaxed(v2m->base + V2M_MSI_TYPER);
+
+		v2m->spi_start = V2M_MSI_TYPER_BASE_SPI(typer);
+		v2m->nr_spis = V2M_MSI_TYPER_NUM_SPI(typer);
+	}
+
+	if (!is_msi_spi_valid(v2m->spi_start, v2m->nr_spis)) {
+		ret = -EINVAL;
+		goto err_iounmap;
+	}
+
+	v2m->bm = kzalloc(sizeof(long) * BITS_TO_LONGS(v2m->nr_spis),
+			  GFP_KERNEL);
+	if (!v2m->bm) {
+		ret = -ENOMEM;
+		goto err_iounmap;
+	}
+
+	v2m->domain = irq_domain_add_tree(NULL, &gicv2m_domain_ops, v2m);
+	if (!v2m->domain) {
+		pr_err("Failed to create GICv2m domain\n");
+		ret = -ENOMEM;
+		goto err_free_bm;
+	}
+
+	v2m->domain->parent = parent;
+	v2m->mchip.of_node = node;
+	v2m->mchip.domain = pci_msi_create_irq_domain(node,
+						      &gicv2m_msi_domain_info,
+						      v2m->domain);
+	if (!v2m->mchip.domain) {
+		pr_err("Failed to create MSI domain\n");
+		ret = -ENOMEM;
+		goto err_free_domains;
+	}
+
+	spin_lock_init(&v2m->msi_cnt_lock);
+
+	ret = of_pci_msi_chip_add(&v2m->mchip);
+	if (ret) {
+		pr_err("Failed to add msi_chip.\n");
+		goto err_free_domains;
+	}
+
+	pr_info("Node %s: range[%#lx:%#lx], SPI[%d:%d]\n", node->name,
+		(unsigned long)v2m->res.start, (unsigned long)v2m->res.end,
+		v2m->spi_start, (v2m->spi_start + v2m->nr_spis));
+
+	return 0;
+
+err_free_domains:
+	if (v2m->mchip.domain)
+		irq_domain_remove(v2m->mchip.domain);
+	if (v2m->domain)
+		irq_domain_remove(v2m->domain);
+err_free_bm:
+	kfree(v2m->bm);
+err_iounmap:
+	iounmap(v2m->base);
+err_free_v2m:
+	kfree(v2m);
+	return ret;
+}
+
+static struct of_device_id gicv2m_device_id[] = {
+	{	.compatible	= "arm,gic-v2m-frame",	},
+	{},
+};
+
+int __init gicv2m_of_init(struct device_node *node, struct irq_domain *parent)
+{
+	int ret = 0;
+	struct device_node *child;
+
+	for (child = of_find_matching_node(node, gicv2m_device_id); child;
+	     child = of_find_matching_node(child, gicv2m_device_id)) {
+		if (!of_find_property(child, "msi-controller", NULL))
+			continue;
+
+		ret = gicv2m_init_one(child, parent);
+		if (ret) {
+			of_node_put(node);
+			break;
+		}
+	}
+
+	return ret;
+}
diff --git a/drivers/irqchip/irq-gic-v3-its.c b/drivers/irqchip/irq-gic-v3-its.c
new file mode 100644
index 000000000000..d8996bdf0f61
--- /dev/null
+++ b/drivers/irqchip/irq-gic-v3-its.c
@@ -0,0 +1,1425 @@
+/*
+ * Copyright (C) 2013, 2014 ARM Limited, All Rights Reserved.
+ * Author: Marc Zyngier <marc.zyngier@arm.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/bitmap.h>
+#include <linux/cpu.h>
+#include <linux/delay.h>
+#include <linux/interrupt.h>
+#include <linux/log2.h>
+#include <linux/mm.h>
+#include <linux/msi.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/of_pci.h>
+#include <linux/of_platform.h>
+#include <linux/percpu.h>
+#include <linux/slab.h>
+
+#include <linux/irqchip/arm-gic-v3.h>
+
+#include <asm/cacheflush.h>
+#include <asm/cputype.h>
+#include <asm/exception.h>
+
+#include "irqchip.h"
+
+#define ITS_FLAGS_CMDQ_NEEDS_FLUSHING		(1 << 0)
+
+#define RDIST_FLAGS_PROPBASE_NEEDS_FLUSHING	(1 << 0)
+
+/*
+ * Collection structure - just an ID, and a redistributor address to
+ * ping. We use one per CPU as a bag of interrupts assigned to this
+ * CPU.
+ */
+struct its_collection {
+	u64			target_address;
+	u16			col_id;
+};
+
+/*
+ * The ITS structure - contains most of the infrastructure, with the
+ * msi_controller, the command queue, the collections, and the list of
+ * devices writing to it.
+ */
+struct its_node {
+	raw_spinlock_t		lock;
+	struct list_head	entry;
+	struct msi_controller	msi_chip;
+	struct irq_domain	*domain;
+	void __iomem		*base;
+	unsigned long		phys_base;
+	struct its_cmd_block	*cmd_base;
+	struct its_cmd_block	*cmd_write;
+	void			*tables[GITS_BASER_NR_REGS];
+	struct its_collection	*collections;
+	struct list_head	its_device_list;
+	u64			flags;
+	u32			ite_size;
+};
+
+#define ITS_ITT_ALIGN		SZ_256
+
+/*
+ * The ITS view of a device - belongs to an ITS, a collection, owns an
+ * interrupt translation table, and a list of interrupts.
+ */
+struct its_device {
+	struct list_head	entry;
+	struct its_node		*its;
+	struct its_collection	*collection;
+	void			*itt;
+	unsigned long		*lpi_map;
+	irq_hw_number_t		lpi_base;
+	int			nr_lpis;
+	u32			nr_ites;
+	u32			device_id;
+};
+
+static LIST_HEAD(its_nodes);
+static DEFINE_SPINLOCK(its_lock);
+static struct device_node *gic_root_node;
+static struct rdists *gic_rdists;
+
+#define gic_data_rdist()		(raw_cpu_ptr(gic_rdists->rdist))
+#define gic_data_rdist_rd_base()	(gic_data_rdist()->rd_base)
+
+/*
+ * ITS command descriptors - parameters to be encoded in a command
+ * block.
+ */
+struct its_cmd_desc {
+	union {
+		struct {
+			struct its_device *dev;
+			u32 event_id;
+		} its_inv_cmd;
+
+		struct {
+			struct its_device *dev;
+			u32 event_id;
+		} its_int_cmd;
+
+		struct {
+			struct its_device *dev;
+			int valid;
+		} its_mapd_cmd;
+
+		struct {
+			struct its_collection *col;
+			int valid;
+		} its_mapc_cmd;
+
+		struct {
+			struct its_device *dev;
+			u32 phys_id;
+			u32 event_id;
+		} its_mapvi_cmd;
+
+		struct {
+			struct its_device *dev;
+			struct its_collection *col;
+			u32 id;
+		} its_movi_cmd;
+
+		struct {
+			struct its_device *dev;
+			u32 event_id;
+		} its_discard_cmd;
+
+		struct {
+			struct its_collection *col;
+		} its_invall_cmd;
+	};
+};
+
+/*
+ * The ITS command block, which is what the ITS actually parses.
+ */
+struct its_cmd_block {
+	u64	raw_cmd[4];
+};
+
+#define ITS_CMD_QUEUE_SZ		SZ_64K
+#define ITS_CMD_QUEUE_NR_ENTRIES	(ITS_CMD_QUEUE_SZ / sizeof(struct its_cmd_block))
+
+typedef struct its_collection *(*its_cmd_builder_t)(struct its_cmd_block *,
+						    struct its_cmd_desc *);
+
+static void its_encode_cmd(struct its_cmd_block *cmd, u8 cmd_nr)
+{
+	cmd->raw_cmd[0] &= ~0xffUL;
+	cmd->raw_cmd[0] |= cmd_nr;
+}
+
+static void its_encode_devid(struct its_cmd_block *cmd, u32 devid)
+{
+	cmd->raw_cmd[0] &= ~(0xffffUL << 32);
+	cmd->raw_cmd[0] |= ((u64)devid) << 32;
+}
+
+static void its_encode_event_id(struct its_cmd_block *cmd, u32 id)
+{
+	cmd->raw_cmd[1] &= ~0xffffffffUL;
+	cmd->raw_cmd[1] |= id;
+}
+
+static void its_encode_phys_id(struct its_cmd_block *cmd, u32 phys_id)
+{
+	cmd->raw_cmd[1] &= 0xffffffffUL;
+	cmd->raw_cmd[1] |= ((u64)phys_id) << 32;
+}
+
+static void its_encode_size(struct its_cmd_block *cmd, u8 size)
+{
+	cmd->raw_cmd[1] &= ~0x1fUL;
+	cmd->raw_cmd[1] |= size & 0x1f;
+}
+
+static void its_encode_itt(struct its_cmd_block *cmd, u64 itt_addr)
+{
+	cmd->raw_cmd[2] &= ~0xffffffffffffUL;
+	cmd->raw_cmd[2] |= itt_addr & 0xffffffffff00UL;
+}
+
+static void its_encode_valid(struct its_cmd_block *cmd, int valid)
+{
+	cmd->raw_cmd[2] &= ~(1UL << 63);
+	cmd->raw_cmd[2] |= ((u64)!!valid) << 63;
+}
+
+static void its_encode_target(struct its_cmd_block *cmd, u64 target_addr)
+{
+	cmd->raw_cmd[2] &= ~(0xffffffffUL << 16);
+	cmd->raw_cmd[2] |= (target_addr & (0xffffffffUL << 16));
+}
+
+static void its_encode_collection(struct its_cmd_block *cmd, u16 col)
+{
+	cmd->raw_cmd[2] &= ~0xffffUL;
+	cmd->raw_cmd[2] |= col;
+}
+
+static inline void its_fixup_cmd(struct its_cmd_block *cmd)
+{
+	/* Let's fixup BE commands */
+	cmd->raw_cmd[0] = cpu_to_le64(cmd->raw_cmd[0]);
+	cmd->raw_cmd[1] = cpu_to_le64(cmd->raw_cmd[1]);
+	cmd->raw_cmd[2] = cpu_to_le64(cmd->raw_cmd[2]);
+	cmd->raw_cmd[3] = cpu_to_le64(cmd->raw_cmd[3]);
+}
+
+static struct its_collection *its_build_mapd_cmd(struct its_cmd_block *cmd,
+						 struct its_cmd_desc *desc)
+{
+	unsigned long itt_addr;
+	u8 size = ilog2(desc->its_mapd_cmd.dev->nr_ites);
+
+	itt_addr = virt_to_phys(desc->its_mapd_cmd.dev->itt);
+	itt_addr = ALIGN(itt_addr, ITS_ITT_ALIGN);
+
+	its_encode_cmd(cmd, GITS_CMD_MAPD);
+	its_encode_devid(cmd, desc->its_mapd_cmd.dev->device_id);
+	its_encode_size(cmd, size - 1);
+	its_encode_itt(cmd, itt_addr);
+	its_encode_valid(cmd, desc->its_mapd_cmd.valid);
+
+	its_fixup_cmd(cmd);
+
+	return desc->its_mapd_cmd.dev->collection;
+}
+
+static struct its_collection *its_build_mapc_cmd(struct its_cmd_block *cmd,
+						 struct its_cmd_desc *desc)
+{
+	its_encode_cmd(cmd, GITS_CMD_MAPC);
+	its_encode_collection(cmd, desc->its_mapc_cmd.col->col_id);
+	its_encode_target(cmd, desc->its_mapc_cmd.col->target_address);
+	its_encode_valid(cmd, desc->its_mapc_cmd.valid);
+
+	its_fixup_cmd(cmd);
+
+	return desc->its_mapc_cmd.col;
+}
+
+static struct its_collection *its_build_mapvi_cmd(struct its_cmd_block *cmd,
+						  struct its_cmd_desc *desc)
+{
+	its_encode_cmd(cmd, GITS_CMD_MAPVI);
+	its_encode_devid(cmd, desc->its_mapvi_cmd.dev->device_id);
+	its_encode_event_id(cmd, desc->its_mapvi_cmd.event_id);
+	its_encode_phys_id(cmd, desc->its_mapvi_cmd.phys_id);
+	its_encode_collection(cmd, desc->its_mapvi_cmd.dev->collection->col_id);
+
+	its_fixup_cmd(cmd);
+
+	return desc->its_mapvi_cmd.dev->collection;
+}
+
+static struct its_collection *its_build_movi_cmd(struct its_cmd_block *cmd,
+						 struct its_cmd_desc *desc)
+{
+	its_encode_cmd(cmd, GITS_CMD_MOVI);
+	its_encode_devid(cmd, desc->its_movi_cmd.dev->device_id);
+	its_encode_event_id(cmd, desc->its_movi_cmd.id);
+	its_encode_collection(cmd, desc->its_movi_cmd.col->col_id);
+
+	its_fixup_cmd(cmd);
+
+	return desc->its_movi_cmd.dev->collection;
+}
+
+static struct its_collection *its_build_discard_cmd(struct its_cmd_block *cmd,
+						    struct its_cmd_desc *desc)
+{
+	its_encode_cmd(cmd, GITS_CMD_DISCARD);
+	its_encode_devid(cmd, desc->its_discard_cmd.dev->device_id);
+	its_encode_event_id(cmd, desc->its_discard_cmd.event_id);
+
+	its_fixup_cmd(cmd);
+
+	return desc->its_discard_cmd.dev->collection;
+}
+
+static struct its_collection *its_build_inv_cmd(struct its_cmd_block *cmd,
+						struct its_cmd_desc *desc)
+{
+	its_encode_cmd(cmd, GITS_CMD_INV);
+	its_encode_devid(cmd, desc->its_inv_cmd.dev->device_id);
+	its_encode_event_id(cmd, desc->its_inv_cmd.event_id);
+
+	its_fixup_cmd(cmd);
+
+	return desc->its_inv_cmd.dev->collection;
+}
+
+static struct its_collection *its_build_invall_cmd(struct its_cmd_block *cmd,
+						   struct its_cmd_desc *desc)
+{
+	its_encode_cmd(cmd, GITS_CMD_INVALL);
+	its_encode_collection(cmd, desc->its_mapc_cmd.col->col_id);
+
+	its_fixup_cmd(cmd);
+
+	return NULL;
+}
+
+static u64 its_cmd_ptr_to_offset(struct its_node *its,
+				 struct its_cmd_block *ptr)
+{
+	return (ptr - its->cmd_base) * sizeof(*ptr);
+}
+
+static int its_queue_full(struct its_node *its)
+{
+	int widx;
+	int ridx;
+
+	widx = its->cmd_write - its->cmd_base;
+	ridx = readl_relaxed(its->base + GITS_CREADR) / sizeof(struct its_cmd_block);
+
+	/* This is incredibly unlikely to happen, unless the ITS locks up. */
+	if (((widx + 1) % ITS_CMD_QUEUE_NR_ENTRIES) == ridx)
+		return 1;
+
+	return 0;
+}
+
+static struct its_cmd_block *its_allocate_entry(struct its_node *its)
+{
+	struct its_cmd_block *cmd;
+	u32 count = 1000000;	/* 1s! */
+
+	while (its_queue_full(its)) {
+		count--;
+		if (!count) {
+			pr_err_ratelimited("ITS queue not draining\n");
+			return NULL;
+		}
+		cpu_relax();
+		udelay(1);
+	}
+
+	cmd = its->cmd_write++;
+
+	/* Handle queue wrapping */
+	if (its->cmd_write == (its->cmd_base + ITS_CMD_QUEUE_NR_ENTRIES))
+		its->cmd_write = its->cmd_base;
+
+	return cmd;
+}
+
+static struct its_cmd_block *its_post_commands(struct its_node *its)
+{
+	u64 wr = its_cmd_ptr_to_offset(its, its->cmd_write);
+
+	writel_relaxed(wr, its->base + GITS_CWRITER);
+
+	return its->cmd_write;
+}
+
+static void its_flush_cmd(struct its_node *its, struct its_cmd_block *cmd)
+{
+	/*
+	 * Make sure the commands written to memory are observable by
+	 * the ITS.
+	 */
+	if (its->flags & ITS_FLAGS_CMDQ_NEEDS_FLUSHING)
+		__flush_dcache_area(cmd, sizeof(*cmd));
+	else
+		dsb(ishst);
+}
+
+static void its_wait_for_range_completion(struct its_node *its,
+					  struct its_cmd_block *from,
+					  struct its_cmd_block *to)
+{
+	u64 rd_idx, from_idx, to_idx;
+	u32 count = 1000000;	/* 1s! */
+
+	from_idx = its_cmd_ptr_to_offset(its, from);
+	to_idx = its_cmd_ptr_to_offset(its, to);
+
+	while (1) {
+		rd_idx = readl_relaxed(its->base + GITS_CREADR);
+		if (rd_idx >= to_idx || rd_idx < from_idx)
+			break;
+
+		count--;
+		if (!count) {
+			pr_err_ratelimited("ITS queue timeout\n");
+			return;
+		}
+		cpu_relax();
+		udelay(1);
+	}
+}
+
+static void its_send_single_command(struct its_node *its,
+				    its_cmd_builder_t builder,
+				    struct its_cmd_desc *desc)
+{
+	struct its_cmd_block *cmd, *sync_cmd, *next_cmd;
+	struct its_collection *sync_col;
+
+	raw_spin_lock(&its->lock);
+
+	cmd = its_allocate_entry(its);
+	if (!cmd) {		/* We're soooooo screewed... */
+		pr_err_ratelimited("ITS can't allocate, dropping command\n");
+		raw_spin_unlock(&its->lock);
+		return;
+	}
+	sync_col = builder(cmd, desc);
+	its_flush_cmd(its, cmd);
+
+	if (sync_col) {
+		sync_cmd = its_allocate_entry(its);
+		if (!sync_cmd) {
+			pr_err_ratelimited("ITS can't SYNC, skipping\n");
+			goto post;
+		}
+		its_encode_cmd(sync_cmd, GITS_CMD_SYNC);
+		its_encode_target(sync_cmd, sync_col->target_address);
+		its_fixup_cmd(sync_cmd);
+		its_flush_cmd(its, sync_cmd);
+	}
+
+post:
+	next_cmd = its_post_commands(its);
+	raw_spin_unlock(&its->lock);
+
+	its_wait_for_range_completion(its, cmd, next_cmd);
+}
+
+static void its_send_inv(struct its_device *dev, u32 event_id)
+{
+	struct its_cmd_desc desc;
+
+	desc.its_inv_cmd.dev = dev;
+	desc.its_inv_cmd.event_id = event_id;
+
+	its_send_single_command(dev->its, its_build_inv_cmd, &desc);
+}
+
+static void its_send_mapd(struct its_device *dev, int valid)
+{
+	struct its_cmd_desc desc;
+
+	desc.its_mapd_cmd.dev = dev;
+	desc.its_mapd_cmd.valid = !!valid;
+
+	its_send_single_command(dev->its, its_build_mapd_cmd, &desc);
+}
+
+static void its_send_mapc(struct its_node *its, struct its_collection *col,
+			  int valid)
+{
+	struct its_cmd_desc desc;
+
+	desc.its_mapc_cmd.col = col;
+	desc.its_mapc_cmd.valid = !!valid;
+
+	its_send_single_command(its, its_build_mapc_cmd, &desc);
+}
+
+static void its_send_mapvi(struct its_device *dev, u32 irq_id, u32 id)
+{
+	struct its_cmd_desc desc;
+
+	desc.its_mapvi_cmd.dev = dev;
+	desc.its_mapvi_cmd.phys_id = irq_id;
+	desc.its_mapvi_cmd.event_id = id;
+
+	its_send_single_command(dev->its, its_build_mapvi_cmd, &desc);
+}
+
+static void its_send_movi(struct its_device *dev,
+			  struct its_collection *col, u32 id)
+{
+	struct its_cmd_desc desc;
+
+	desc.its_movi_cmd.dev = dev;
+	desc.its_movi_cmd.col = col;
+	desc.its_movi_cmd.id = id;
+
+	its_send_single_command(dev->its, its_build_movi_cmd, &desc);
+}
+
+static void its_send_discard(struct its_device *dev, u32 id)
+{
+	struct its_cmd_desc desc;
+
+	desc.its_discard_cmd.dev = dev;
+	desc.its_discard_cmd.event_id = id;
+
+	its_send_single_command(dev->its, its_build_discard_cmd, &desc);
+}
+
+static void its_send_invall(struct its_node *its, struct its_collection *col)
+{
+	struct its_cmd_desc desc;
+
+	desc.its_invall_cmd.col = col;
+
+	its_send_single_command(its, its_build_invall_cmd, &desc);
+}
+
+/*
+ * irqchip functions - assumes MSI, mostly.
+ */
+
+static inline u32 its_get_event_id(struct irq_data *d)
+{
+	struct its_device *its_dev = irq_data_get_irq_chip_data(d);
+	return d->hwirq - its_dev->lpi_base;
+}
+
+static void lpi_set_config(struct irq_data *d, bool enable)
+{
+	struct its_device *its_dev = irq_data_get_irq_chip_data(d);
+	irq_hw_number_t hwirq = d->hwirq;
+	u32 id = its_get_event_id(d);
+	u8 *cfg = page_address(gic_rdists->prop_page) + hwirq - 8192;
+
+	if (enable)
+		*cfg |= LPI_PROP_ENABLED;
+	else
+		*cfg &= ~LPI_PROP_ENABLED;
+
+	/*
+	 * Make the above write visible to the redistributors.
+	 * And yes, we're flushing exactly: One. Single. Byte.
+	 * Humpf...
+	 */
+	if (gic_rdists->flags & RDIST_FLAGS_PROPBASE_NEEDS_FLUSHING)
+		__flush_dcache_area(cfg, sizeof(*cfg));
+	else
+		dsb(ishst);
+	its_send_inv(its_dev, id);
+}
+
+static void its_mask_irq(struct irq_data *d)
+{
+	lpi_set_config(d, false);
+}
+
+static void its_unmask_irq(struct irq_data *d)
+{
+	lpi_set_config(d, true);
+}
+
+static void its_eoi_irq(struct irq_data *d)
+{
+	gic_write_eoir(d->hwirq);
+}
+
+static int its_set_affinity(struct irq_data *d, const struct cpumask *mask_val,
+			    bool force)
+{
+	unsigned int cpu = cpumask_any_and(mask_val, cpu_online_mask);
+	struct its_device *its_dev = irq_data_get_irq_chip_data(d);
+	struct its_collection *target_col;
+	u32 id = its_get_event_id(d);
+
+	if (cpu >= nr_cpu_ids)
+		return -EINVAL;
+
+	target_col = &its_dev->its->collections[cpu];
+	its_send_movi(its_dev, target_col, id);
+	its_dev->collection = target_col;
+
+	return IRQ_SET_MASK_OK_DONE;
+}
+
+static void its_irq_compose_msi_msg(struct irq_data *d, struct msi_msg *msg)
+{
+	struct its_device *its_dev = irq_data_get_irq_chip_data(d);
+	struct its_node *its;
+	u64 addr;
+
+	its = its_dev->its;
+	addr = its->phys_base + GITS_TRANSLATER;
+
+	msg->address_lo		= addr & ((1UL << 32) - 1);
+	msg->address_hi		= addr >> 32;
+	msg->data		= its_get_event_id(d);
+}
+
+static struct irq_chip its_irq_chip = {
+	.name			= "ITS",
+	.irq_mask		= its_mask_irq,
+	.irq_unmask		= its_unmask_irq,
+	.irq_eoi		= its_eoi_irq,
+	.irq_set_affinity	= its_set_affinity,
+	.irq_compose_msi_msg	= its_irq_compose_msi_msg,
+};
+
+static void its_mask_msi_irq(struct irq_data *d)
+{
+	pci_msi_mask_irq(d);
+	irq_chip_mask_parent(d);
+}
+
+static void its_unmask_msi_irq(struct irq_data *d)
+{
+	pci_msi_unmask_irq(d);
+	irq_chip_unmask_parent(d);
+}
+
+static struct irq_chip its_msi_irq_chip = {
+	.name			= "ITS-MSI",
+	.irq_unmask		= its_unmask_msi_irq,
+	.irq_mask		= its_mask_msi_irq,
+	.irq_eoi		= irq_chip_eoi_parent,
+	.irq_write_msi_msg	= pci_msi_domain_write_msg,
+};
+
+/*
+ * How we allocate LPIs:
+ *
+ * The GIC has id_bits bits for interrupt identifiers. From there, we
+ * must subtract 8192 which are reserved for SGIs/PPIs/SPIs. Then, as
+ * we allocate LPIs by chunks of 32, we can shift the whole thing by 5
+ * bits to the right.
+ *
+ * This gives us (((1UL << id_bits) - 8192) >> 5) possible allocations.
+ */
+#define IRQS_PER_CHUNK_SHIFT	5
+#define IRQS_PER_CHUNK		(1 << IRQS_PER_CHUNK_SHIFT)
+
+static unsigned long *lpi_bitmap;
+static u32 lpi_chunks;
+static DEFINE_SPINLOCK(lpi_lock);
+
+static int its_lpi_to_chunk(int lpi)
+{
+	return (lpi - 8192) >> IRQS_PER_CHUNK_SHIFT;
+}
+
+static int its_chunk_to_lpi(int chunk)
+{
+	return (chunk << IRQS_PER_CHUNK_SHIFT) + 8192;
+}
+
+static int its_lpi_init(u32 id_bits)
+{
+	lpi_chunks = its_lpi_to_chunk(1UL << id_bits);
+
+	lpi_bitmap = kzalloc(BITS_TO_LONGS(lpi_chunks) * sizeof(long),
+			     GFP_KERNEL);
+	if (!lpi_bitmap) {
+		lpi_chunks = 0;
+		return -ENOMEM;
+	}
+
+	pr_info("ITS: Allocated %d chunks for LPIs\n", (int)lpi_chunks);
+	return 0;
+}
+
+static unsigned long *its_lpi_alloc_chunks(int nr_irqs, int *base, int *nr_ids)
+{
+	unsigned long *bitmap = NULL;
+	int chunk_id;
+	int nr_chunks;
+	int i;
+
+	nr_chunks = DIV_ROUND_UP(nr_irqs, IRQS_PER_CHUNK);
+
+	spin_lock(&lpi_lock);
+
+	do {
+		chunk_id = bitmap_find_next_zero_area(lpi_bitmap, lpi_chunks,
+						      0, nr_chunks, 0);
+		if (chunk_id < lpi_chunks)
+			break;
+
+		nr_chunks--;
+	} while (nr_chunks > 0);
+
+	if (!nr_chunks)
+		goto out;
+
+	bitmap = kzalloc(BITS_TO_LONGS(nr_chunks * IRQS_PER_CHUNK) * sizeof (long),
+			 GFP_ATOMIC);
+	if (!bitmap)
+		goto out;
+
+	for (i = 0; i < nr_chunks; i++)
+		set_bit(chunk_id + i, lpi_bitmap);
+
+	*base = its_chunk_to_lpi(chunk_id);
+	*nr_ids = nr_chunks * IRQS_PER_CHUNK;
+
+out:
+	spin_unlock(&lpi_lock);
+
+	return bitmap;
+}
+
+static void its_lpi_free(unsigned long *bitmap, int base, int nr_ids)
+{
+	int lpi;
+
+	spin_lock(&lpi_lock);
+
+	for (lpi = base; lpi < (base + nr_ids); lpi += IRQS_PER_CHUNK) {
+		int chunk = its_lpi_to_chunk(lpi);
+		BUG_ON(chunk > lpi_chunks);
+		if (test_bit(chunk, lpi_bitmap)) {
+			clear_bit(chunk, lpi_bitmap);
+		} else {
+			pr_err("Bad LPI chunk %d\n", chunk);
+		}
+	}
+
+	spin_unlock(&lpi_lock);
+
+	kfree(bitmap);
+}
+
+/*
+ * We allocate 64kB for PROPBASE. That gives us at most 64K LPIs to
+ * deal with (one configuration byte per interrupt). PENDBASE has to
+ * be 64kB aligned (one bit per LPI, plus 8192 bits for SPI/PPI/SGI).
+ */
+#define LPI_PROPBASE_SZ		SZ_64K
+#define LPI_PENDBASE_SZ		(LPI_PROPBASE_SZ / 8 + SZ_1K)
+
+/*
+ * This is how many bits of ID we need, including the useless ones.
+ */
+#define LPI_NRBITS		ilog2(LPI_PROPBASE_SZ + SZ_8K)
+
+#define LPI_PROP_DEFAULT_PRIO	0xa0
+
+static int __init its_alloc_lpi_tables(void)
+{
+	phys_addr_t paddr;
+
+	gic_rdists->prop_page = alloc_pages(GFP_NOWAIT,
+					   get_order(LPI_PROPBASE_SZ));
+	if (!gic_rdists->prop_page) {
+		pr_err("Failed to allocate PROPBASE\n");
+		return -ENOMEM;
+	}
+
+	paddr = page_to_phys(gic_rdists->prop_page);
+	pr_info("GIC: using LPI property table @%pa\n", &paddr);
+
+	/* Priority 0xa0, Group-1, disabled */
+	memset(page_address(gic_rdists->prop_page),
+	       LPI_PROP_DEFAULT_PRIO | LPI_PROP_GROUP1,
+	       LPI_PROPBASE_SZ);
+
+	/* Make sure the GIC will observe the written configuration */
+	__flush_dcache_area(page_address(gic_rdists->prop_page), LPI_PROPBASE_SZ);
+
+	return 0;
+}
+
+static const char *its_base_type_string[] = {
+	[GITS_BASER_TYPE_DEVICE]	= "Devices",
+	[GITS_BASER_TYPE_VCPU]		= "Virtual CPUs",
+	[GITS_BASER_TYPE_CPU]		= "Physical CPUs",
+	[GITS_BASER_TYPE_COLLECTION]	= "Interrupt Collections",
+	[GITS_BASER_TYPE_RESERVED5] 	= "Reserved (5)",
+	[GITS_BASER_TYPE_RESERVED6] 	= "Reserved (6)",
+	[GITS_BASER_TYPE_RESERVED7] 	= "Reserved (7)",
+};
+
+static void its_free_tables(struct its_node *its)
+{
+	int i;
+
+	for (i = 0; i < GITS_BASER_NR_REGS; i++) {
+		if (its->tables[i]) {
+			free_page((unsigned long)its->tables[i]);
+			its->tables[i] = NULL;
+		}
+	}
+}
+
+static int its_alloc_tables(struct its_node *its)
+{
+	int err;
+	int i;
+	int psz = PAGE_SIZE;
+	u64 shr = GITS_BASER_InnerShareable;
+
+	for (i = 0; i < GITS_BASER_NR_REGS; i++) {
+		u64 val = readq_relaxed(its->base + GITS_BASER + i * 8);
+		u64 type = GITS_BASER_TYPE(val);
+		u64 entry_size = GITS_BASER_ENTRY_SIZE(val);
+		u64 tmp;
+		void *base;
+
+		if (type == GITS_BASER_TYPE_NONE)
+			continue;
+
+		/* We're lazy and only allocate a single page for now */
+		base = (void *)get_zeroed_page(GFP_KERNEL);
+		if (!base) {
+			err = -ENOMEM;
+			goto out_free;
+		}
+
+		its->tables[i] = base;
+
+retry_baser:
+		val = (virt_to_phys(base) 				 |
+		       (type << GITS_BASER_TYPE_SHIFT)			 |
+		       ((entry_size - 1) << GITS_BASER_ENTRY_SIZE_SHIFT) |
+		       GITS_BASER_WaWb					 |
+		       shr						 |
+		       GITS_BASER_VALID);
+
+		switch (psz) {
+		case SZ_4K:
+			val |= GITS_BASER_PAGE_SIZE_4K;
+			break;
+		case SZ_16K:
+			val |= GITS_BASER_PAGE_SIZE_16K;
+			break;
+		case SZ_64K:
+			val |= GITS_BASER_PAGE_SIZE_64K;
+			break;
+		}
+
+		val |= (PAGE_SIZE / psz) - 1;
+
+		writeq_relaxed(val, its->base + GITS_BASER + i * 8);
+		tmp = readq_relaxed(its->base + GITS_BASER + i * 8);
+
+		if ((val ^ tmp) & GITS_BASER_SHAREABILITY_MASK) {
+			/*
+			 * Shareability didn't stick. Just use
+			 * whatever the read reported, which is likely
+			 * to be the only thing this redistributor
+			 * supports.
+			 */
+			shr = tmp & GITS_BASER_SHAREABILITY_MASK;
+			goto retry_baser;
+		}
+
+		if ((val ^ tmp) & GITS_BASER_PAGE_SIZE_MASK) {
+			/*
+			 * Page size didn't stick. Let's try a smaller
+			 * size and retry. If we reach 4K, then
+			 * something is horribly wrong...
+			 */
+			switch (psz) {
+			case SZ_16K:
+				psz = SZ_4K;
+				goto retry_baser;
+			case SZ_64K:
+				psz = SZ_16K;
+				goto retry_baser;
+			}
+		}
+
+		if (val != tmp) {
+			pr_err("ITS: %s: GITS_BASER%d doesn't stick: %lx %lx\n",
+			       its->msi_chip.of_node->full_name, i,
+			       (unsigned long) val, (unsigned long) tmp);
+			err = -ENXIO;
+			goto out_free;
+		}
+
+		pr_info("ITS: allocated %d %s @%lx (psz %dK, shr %d)\n",
+			(int)(PAGE_SIZE / entry_size),
+			its_base_type_string[type],
+			(unsigned long)virt_to_phys(base),
+			psz / SZ_1K, (int)shr >> GITS_BASER_SHAREABILITY_SHIFT);
+	}
+
+	return 0;
+
+out_free:
+	its_free_tables(its);
+
+	return err;
+}
+
+static int its_alloc_collections(struct its_node *its)
+{
+	its->collections = kzalloc(nr_cpu_ids * sizeof(*its->collections),
+				   GFP_KERNEL);
+	if (!its->collections)
+		return -ENOMEM;
+
+	return 0;
+}
+
+static void its_cpu_init_lpis(void)
+{
+	void __iomem *rbase = gic_data_rdist_rd_base();
+	struct page *pend_page;
+	u64 val, tmp;
+
+	/* If we didn't allocate the pending table yet, do it now */
+	pend_page = gic_data_rdist()->pend_page;
+	if (!pend_page) {
+		phys_addr_t paddr;
+		/*
+		 * The pending pages have to be at least 64kB aligned,
+		 * hence the 'max(LPI_PENDBASE_SZ, SZ_64K)' below.
+		 */
+		pend_page = alloc_pages(GFP_NOWAIT | __GFP_ZERO,
+					get_order(max(LPI_PENDBASE_SZ, SZ_64K)));
+		if (!pend_page) {
+			pr_err("Failed to allocate PENDBASE for CPU%d\n",
+			       smp_processor_id());
+			return;
+		}
+
+		/* Make sure the GIC will observe the zero-ed page */
+		__flush_dcache_area(page_address(pend_page), LPI_PENDBASE_SZ);
+
+		paddr = page_to_phys(pend_page);
+		pr_info("CPU%d: using LPI pending table @%pa\n",
+			smp_processor_id(), &paddr);
+		gic_data_rdist()->pend_page = pend_page;
+	}
+
+	/* Disable LPIs */
+	val = readl_relaxed(rbase + GICR_CTLR);
+	val &= ~GICR_CTLR_ENABLE_LPIS;
+	writel_relaxed(val, rbase + GICR_CTLR);
+
+	/*
+	 * Make sure any change to the table is observable by the GIC.
+	 */
+	dsb(sy);
+
+	/* set PROPBASE */
+	val = (page_to_phys(gic_rdists->prop_page) |
+	       GICR_PROPBASER_InnerShareable |
+	       GICR_PROPBASER_WaWb |
+	       ((LPI_NRBITS - 1) & GICR_PROPBASER_IDBITS_MASK));
+
+	writeq_relaxed(val, rbase + GICR_PROPBASER);
+	tmp = readq_relaxed(rbase + GICR_PROPBASER);
+
+	if ((tmp ^ val) & GICR_PROPBASER_SHAREABILITY_MASK) {
+		pr_info_once("GIC: using cache flushing for LPI property table\n");
+		gic_rdists->flags |= RDIST_FLAGS_PROPBASE_NEEDS_FLUSHING;
+	}
+
+	/* set PENDBASE */
+	val = (page_to_phys(pend_page) |
+	       GICR_PROPBASER_InnerShareable |
+	       GICR_PROPBASER_WaWb);
+
+	writeq_relaxed(val, rbase + GICR_PENDBASER);
+
+	/* Enable LPIs */
+	val = readl_relaxed(rbase + GICR_CTLR);
+	val |= GICR_CTLR_ENABLE_LPIS;
+	writel_relaxed(val, rbase + GICR_CTLR);
+
+	/* Make sure the GIC has seen the above */
+	dsb(sy);
+}
+
+static void its_cpu_init_collection(void)
+{
+	struct its_node *its;
+	int cpu;
+
+	spin_lock(&its_lock);
+	cpu = smp_processor_id();
+
+	list_for_each_entry(its, &its_nodes, entry) {
+		u64 target;
+
+		/*
+		 * We now have to bind each collection to its target
+		 * redistributor.
+		 */
+		if (readq_relaxed(its->base + GITS_TYPER) & GITS_TYPER_PTA) {
+			/*
+			 * This ITS wants the physical address of the
+			 * redistributor.
+			 */
+			target = gic_data_rdist()->phys_base;
+		} else {
+			/*
+			 * This ITS wants a linear CPU number.
+			 */
+			target = readq_relaxed(gic_data_rdist_rd_base() + GICR_TYPER);
+			target = GICR_TYPER_CPU_NUMBER(target);
+		}
+
+		/* Perform collection mapping */
+		its->collections[cpu].target_address = target;
+		its->collections[cpu].col_id = cpu;
+
+		its_send_mapc(its, &its->collections[cpu], 1);
+		its_send_invall(its, &its->collections[cpu]);
+	}
+
+	spin_unlock(&its_lock);
+}
+
+static struct its_device *its_find_device(struct its_node *its, u32 dev_id)
+{
+	struct its_device *its_dev = NULL, *tmp;
+
+	raw_spin_lock(&its->lock);
+
+	list_for_each_entry(tmp, &its->its_device_list, entry) {
+		if (tmp->device_id == dev_id) {
+			its_dev = tmp;
+			break;
+		}
+	}
+
+	raw_spin_unlock(&its->lock);
+
+	return its_dev;
+}
+
+static struct its_device *its_create_device(struct its_node *its, u32 dev_id,
+					    int nvecs)
+{
+	struct its_device *dev;
+	unsigned long *lpi_map;
+	void *itt;
+	int lpi_base;
+	int nr_lpis;
+	int nr_ites;
+	int cpu;
+	int sz;
+
+	dev = kzalloc(sizeof(*dev), GFP_KERNEL);
+	/*
+	 * At least one bit of EventID is being used, hence a minimum
+	 * of two entries. No, the architecture doesn't let you
+	 * express an ITT with a single entry.
+	 */
+	nr_ites = max(2UL, roundup_pow_of_two(nvecs));
+	sz = nr_ites * its->ite_size;
+	sz = max(sz, ITS_ITT_ALIGN) + ITS_ITT_ALIGN - 1;
+	itt = kmalloc(sz, GFP_KERNEL);
+	lpi_map = its_lpi_alloc_chunks(nvecs, &lpi_base, &nr_lpis);
+
+	if (!dev || !itt || !lpi_map) {
+		kfree(dev);
+		kfree(itt);
+		kfree(lpi_map);
+		return NULL;
+	}
+
+	dev->its = its;
+	dev->itt = itt;
+	dev->nr_ites = nr_ites;
+	dev->lpi_map = lpi_map;
+	dev->lpi_base = lpi_base;
+	dev->nr_lpis = nr_lpis;
+	dev->device_id = dev_id;
+	INIT_LIST_HEAD(&dev->entry);
+
+	raw_spin_lock(&its->lock);
+	list_add(&dev->entry, &its->its_device_list);
+	raw_spin_unlock(&its->lock);
+
+	/* Bind the device to the first possible CPU */
+	cpu = cpumask_first(cpu_online_mask);
+	dev->collection = &its->collections[cpu];
+
+	/* Map device to its ITT */
+	its_send_mapd(dev, 1);
+
+	return dev;
+}
+
+static void its_free_device(struct its_device *its_dev)
+{
+	raw_spin_lock(&its_dev->its->lock);
+	list_del(&its_dev->entry);
+	raw_spin_unlock(&its_dev->its->lock);
+	kfree(its_dev->itt);
+	kfree(its_dev);
+}
+
+static int its_alloc_device_irq(struct its_device *dev, irq_hw_number_t *hwirq)
+{
+	int idx;
+
+	idx = find_first_zero_bit(dev->lpi_map, dev->nr_lpis);
+	if (idx == dev->nr_lpis)
+		return -ENOSPC;
+
+	*hwirq = dev->lpi_base + idx;
+	set_bit(idx, dev->lpi_map);
+
+	return 0;
+}
+
+static int its_msi_prepare(struct irq_domain *domain, struct device *dev,
+			   int nvec, msi_alloc_info_t *info)
+{
+	struct pci_dev *pdev;
+	struct its_node *its;
+	u32 dev_id;
+	struct its_device *its_dev;
+
+	if (!dev_is_pci(dev))
+		return -EINVAL;
+
+	pdev = to_pci_dev(dev);
+	dev_id = PCI_DEVID(pdev->bus->number, pdev->devfn);
+	its = domain->parent->host_data;
+
+	its_dev = its_find_device(its, dev_id);
+	if (WARN_ON(its_dev))
+		return -EINVAL;
+
+	its_dev = its_create_device(its, dev_id, nvec);
+	if (!its_dev)
+		return -ENOMEM;
+
+	dev_dbg(&pdev->dev, "ITT %d entries, %d bits\n", nvec, ilog2(nvec));
+
+	info->scratchpad[0].ptr = its_dev;
+	info->scratchpad[1].ptr = dev;
+	return 0;
+}
+
+static struct msi_domain_ops its_pci_msi_ops = {
+	.msi_prepare	= its_msi_prepare,
+};
+
+static struct msi_domain_info its_pci_msi_domain_info = {
+	.flags	= (MSI_FLAG_USE_DEF_DOM_OPS | MSI_FLAG_USE_DEF_CHIP_OPS |
+		   MSI_FLAG_MULTI_PCI_MSI | MSI_FLAG_PCI_MSIX),
+	.ops	= &its_pci_msi_ops,
+	.chip	= &its_msi_irq_chip,
+};
+
+static int its_irq_gic_domain_alloc(struct irq_domain *domain,
+				    unsigned int virq,
+				    irq_hw_number_t hwirq)
+{
+	struct of_phandle_args args;
+
+	args.np = domain->parent->of_node;
+	args.args_count = 3;
+	args.args[0] = GIC_IRQ_TYPE_LPI;
+	args.args[1] = hwirq;
+	args.args[2] = IRQ_TYPE_EDGE_RISING;
+
+	return irq_domain_alloc_irqs_parent(domain, virq, 1, &args);
+}
+
+static int its_irq_domain_alloc(struct irq_domain *domain, unsigned int virq,
+				unsigned int nr_irqs, void *args)
+{
+	msi_alloc_info_t *info = args;
+	struct its_device *its_dev = info->scratchpad[0].ptr;
+	irq_hw_number_t hwirq;
+	int err;
+	int i;
+
+	for (i = 0; i < nr_irqs; i++) {
+		err = its_alloc_device_irq(its_dev, &hwirq);
+		if (err)
+			return err;
+
+		err = its_irq_gic_domain_alloc(domain, virq + i, hwirq);
+		if (err)
+			return err;
+
+		irq_domain_set_hwirq_and_chip(domain, virq + i,
+					      hwirq, &its_irq_chip, its_dev);
+		dev_dbg(info->scratchpad[1].ptr, "ID:%d pID:%d vID:%d\n",
+			(int)(hwirq - its_dev->lpi_base), (int)hwirq, virq + i);
+	}
+
+	return 0;
+}
+
+static void its_irq_domain_activate(struct irq_domain *domain,
+				    struct irq_data *d)
+{
+	struct its_device *its_dev = irq_data_get_irq_chip_data(d);
+	u32 event = its_get_event_id(d);
+
+	/* Map the GIC IRQ and event to the device */
+	its_send_mapvi(its_dev, d->hwirq, event);
+}
+
+static void its_irq_domain_deactivate(struct irq_domain *domain,
+				      struct irq_data *d)
+{
+	struct its_device *its_dev = irq_data_get_irq_chip_data(d);
+	u32 event = its_get_event_id(d);
+
+	/* Stop the delivery of interrupts */
+	its_send_discard(its_dev, event);
+}
+
+static void its_irq_domain_free(struct irq_domain *domain, unsigned int virq,
+				unsigned int nr_irqs)
+{
+	struct irq_data *d = irq_domain_get_irq_data(domain, virq);
+	struct its_device *its_dev = irq_data_get_irq_chip_data(d);
+	int i;
+
+	for (i = 0; i < nr_irqs; i++) {
+		struct irq_data *data = irq_domain_get_irq_data(domain,
+								virq + i);
+		u32 event = its_get_event_id(data);
+
+		/* Mark interrupt index as unused */
+		clear_bit(event, its_dev->lpi_map);
+
+		/* Nuke the entry in the domain */
+		irq_domain_reset_irq_data(data);
+	}
+
+	/* If all interrupts have been freed, start mopping the floor */
+	if (bitmap_empty(its_dev->lpi_map, its_dev->nr_lpis)) {
+		its_lpi_free(its_dev->lpi_map,
+			     its_dev->lpi_base,
+			     its_dev->nr_lpis);
+
+		/* Unmap device/itt */
+		its_send_mapd(its_dev, 0);
+		its_free_device(its_dev);
+	}
+
+	irq_domain_free_irqs_parent(domain, virq, nr_irqs);
+}
+
+static const struct irq_domain_ops its_domain_ops = {
+	.alloc			= its_irq_domain_alloc,
+	.free			= its_irq_domain_free,
+	.activate		= its_irq_domain_activate,
+	.deactivate		= its_irq_domain_deactivate,
+};
+
+static int its_probe(struct device_node *node, struct irq_domain *parent)
+{
+	struct resource res;
+	struct its_node *its;
+	void __iomem *its_base;
+	u32 val;
+	u64 baser, tmp;
+	int err;
+
+	err = of_address_to_resource(node, 0, &res);
+	if (err) {
+		pr_warn("%s: no regs?\n", node->full_name);
+		return -ENXIO;
+	}
+
+	its_base = ioremap(res.start, resource_size(&res));
+	if (!its_base) {
+		pr_warn("%s: unable to map registers\n", node->full_name);
+		return -ENOMEM;
+	}
+
+	val = readl_relaxed(its_base + GITS_PIDR2) & GIC_PIDR2_ARCH_MASK;
+	if (val != 0x30 && val != 0x40) {
+		pr_warn("%s: no ITS detected, giving up\n", node->full_name);
+		err = -ENODEV;
+		goto out_unmap;
+	}
+
+	pr_info("ITS: %s\n", node->full_name);
+
+	its = kzalloc(sizeof(*its), GFP_KERNEL);
+	if (!its) {
+		err = -ENOMEM;
+		goto out_unmap;
+	}
+
+	raw_spin_lock_init(&its->lock);
+	INIT_LIST_HEAD(&its->entry);
+	INIT_LIST_HEAD(&its->its_device_list);
+	its->base = its_base;
+	its->phys_base = res.start;
+	its->msi_chip.of_node = node;
+	its->ite_size = ((readl_relaxed(its_base + GITS_TYPER) >> 4) & 0xf) + 1;
+
+	its->cmd_base = kzalloc(ITS_CMD_QUEUE_SZ, GFP_KERNEL);
+	if (!its->cmd_base) {
+		err = -ENOMEM;
+		goto out_free_its;
+	}
+	its->cmd_write = its->cmd_base;
+
+	err = its_alloc_tables(its);
+	if (err)
+		goto out_free_cmd;
+
+	err = its_alloc_collections(its);
+	if (err)
+		goto out_free_tables;
+
+	baser = (virt_to_phys(its->cmd_base)	|
+		 GITS_CBASER_WaWb		|
+		 GITS_CBASER_InnerShareable	|
+		 (ITS_CMD_QUEUE_SZ / SZ_4K - 1)	|
+		 GITS_CBASER_VALID);
+
+	writeq_relaxed(baser, its->base + GITS_CBASER);
+	tmp = readq_relaxed(its->base + GITS_CBASER);
+	writeq_relaxed(0, its->base + GITS_CWRITER);
+	writel_relaxed(1, its->base + GITS_CTLR);
+
+	if ((tmp ^ baser) & GITS_BASER_SHAREABILITY_MASK) {
+		pr_info("ITS: using cache flushing for cmd queue\n");
+		its->flags |= ITS_FLAGS_CMDQ_NEEDS_FLUSHING;
+	}
+
+	if (of_property_read_bool(its->msi_chip.of_node, "msi-controller")) {
+		its->domain = irq_domain_add_tree(NULL, &its_domain_ops, its);
+		if (!its->domain) {
+			err = -ENOMEM;
+			goto out_free_tables;
+		}
+
+		its->domain->parent = parent;
+
+		its->msi_chip.domain = pci_msi_create_irq_domain(node,
+								 &its_pci_msi_domain_info,
+								 its->domain);
+		if (!its->msi_chip.domain) {
+			err = -ENOMEM;
+			goto out_free_domains;
+		}
+
+		err = of_pci_msi_chip_add(&its->msi_chip);
+		if (err)
+			goto out_free_domains;
+	}
+
+	spin_lock(&its_lock);
+	list_add(&its->entry, &its_nodes);
+	spin_unlock(&its_lock);
+
+	return 0;
+
+out_free_domains:
+	if (its->msi_chip.domain)
+		irq_domain_remove(its->msi_chip.domain);
+	if (its->domain)
+		irq_domain_remove(its->domain);
+out_free_tables:
+	its_free_tables(its);
+out_free_cmd:
+	kfree(its->cmd_base);
+out_free_its:
+	kfree(its);
+out_unmap:
+	iounmap(its_base);
+	pr_err("ITS: failed probing %s (%d)\n", node->full_name, err);
+	return err;
+}
+
+static bool gic_rdists_supports_plpis(void)
+{
+	return !!(readl_relaxed(gic_data_rdist_rd_base() + GICR_TYPER) & GICR_TYPER_PLPIS);
+}
+
+int its_cpu_init(void)
+{
+	if (!gic_rdists_supports_plpis()) {
+		pr_info("CPU%d: LPIs not supported\n", smp_processor_id());
+		return -ENXIO;
+	}
+
+	if (!list_empty(&its_nodes)) {
+		its_cpu_init_lpis();
+		its_cpu_init_collection();
+	}
+
+	return 0;
+}
+
+static struct of_device_id its_device_id[] = {
+	{	.compatible	= "arm,gic-v3-its",	},
+	{},
+};
+
+int its_init(struct device_node *node, struct rdists *rdists,
+	     struct irq_domain *parent_domain)
+{
+	struct device_node *np;
+
+	for (np = of_find_matching_node(node, its_device_id); np;
+	     np = of_find_matching_node(np, its_device_id)) {
+		its_probe(np, parent_domain);
+	}
+
+	if (list_empty(&its_nodes)) {
+		pr_warn("ITS: No ITS available, not enabling LPIs\n");
+		return -ENXIO;
+	}
+
+	gic_rdists = rdists;
+	gic_root_node = node;
+
+	its_alloc_lpi_tables();
+	its_lpi_init(rdists->id_bits);
+
+	return 0;
+}
diff --git a/drivers/irqchip/irq-gic-v3.c b/drivers/irqchip/irq-gic-v3.c
index a0698b4f0303..1a146ccee701 100644
--- a/drivers/irqchip/irq-gic-v3.c
+++ b/drivers/irqchip/irq-gic-v3.c
@@ -16,6 +16,7 @@
  */
 
 #include <linux/cpu.h>
+#include <linux/cpu_pm.h>
 #include <linux/delay.h>
 #include <linux/interrupt.h>
 #include <linux/of.h>
@@ -33,20 +34,25 @@
 #include "irq-gic-common.h"
 #include "irqchip.h"
 
+struct redist_region {
+	void __iomem		*redist_base;
+	phys_addr_t		phys_base;
+};
+
 struct gic_chip_data {
 	void __iomem		*dist_base;
-	void __iomem		**redist_base;
-	void __iomem * __percpu	*rdist;
+	struct redist_region	*redist_regions;
+	struct rdists		rdists;
 	struct irq_domain	*domain;
 	u64			redist_stride;
-	u32			redist_regions;
+	u32			nr_redist_regions;
 	unsigned int		irq_nr;
 };
 
 static struct gic_chip_data gic_data __read_mostly;
 
-#define gic_data_rdist()		(this_cpu_ptr(gic_data.rdist))
-#define gic_data_rdist_rd_base()	(*gic_data_rdist())
+#define gic_data_rdist()		(this_cpu_ptr(gic_data.rdists.rdist))
+#define gic_data_rdist_rd_base()	(gic_data_rdist()->rd_base)
 #define gic_data_rdist_sgi_base()	(gic_data_rdist_rd_base() + SZ_64K)
 
 /* Our default, arbitrary priority value. Linux only uses one anyway. */
@@ -70,9 +76,6 @@ static inline void __iomem *gic_dist_base(struct irq_data *d)
 	if (d->hwirq <= 1023)		/* SPI -> dist_base */
 		return gic_data.dist_base;
 
-	if (d->hwirq >= 8192)
-		BUG();		/* LPI Detected!!! */
-
 	return NULL;
 }
 
@@ -155,7 +158,7 @@ static void gic_enable_sre(void)
 		pr_err("GIC: unable to set SRE (disabled at EL2), panic ahead\n");
 }
 
-static void gic_enable_redist(void)
+static void gic_enable_redist(bool enable)
 {
 	void __iomem *rbase;
 	u32 count = 1000000;	/* 1s! */
@@ -163,20 +166,30 @@ static void gic_enable_redist(void)
 
 	rbase = gic_data_rdist_rd_base();
 
-	/* Wake up this CPU redistributor */
 	val = readl_relaxed(rbase + GICR_WAKER);
-	val &= ~GICR_WAKER_ProcessorSleep;
+	if (enable)
+		/* Wake up this CPU redistributor */
+		val &= ~GICR_WAKER_ProcessorSleep;
+	else
+		val |= GICR_WAKER_ProcessorSleep;
 	writel_relaxed(val, rbase + GICR_WAKER);
 
-	while (readl_relaxed(rbase + GICR_WAKER) & GICR_WAKER_ChildrenAsleep) {
-		count--;
-		if (!count) {
-			pr_err_ratelimited("redist didn't wake up...\n");
-			return;
-		}
+	if (!enable) {		/* Check that GICR_WAKER is writeable */
+		val = readl_relaxed(rbase + GICR_WAKER);
+		if (!(val & GICR_WAKER_ProcessorSleep))
+			return;	/* No PM support in this redistributor */
+	}
+
+	while (count--) {
+		val = readl_relaxed(rbase + GICR_WAKER);
+		if (enable ^ (val & GICR_WAKER_ChildrenAsleep))
+			break;
 		cpu_relax();
 		udelay(1);
 	};
+	if (!count)
+		pr_err_ratelimited("redistributor failed to %s...\n",
+				   enable ? "wakeup" : "sleep");
 }
 
 /*
@@ -260,15 +273,14 @@ static asmlinkage void __exception_irq_entry gic_handle_irq(struct pt_regs *regs
 	do {
 		irqnr = gic_read_iar();
 
-		if (likely(irqnr > 15 && irqnr < 1020)) {
-			u64 irq = irq_find_mapping(gic_data.domain, irqnr);
-			if (likely(irq)) {
-				handle_IRQ(irq, regs);
-				continue;
+		if (likely(irqnr > 15 && irqnr < 1020) || irqnr >= 8192) {
+			int err;
+			err = handle_domain_irq(gic_data.domain, irqnr, regs);
+			if (err) {
+				WARN_ONCE(true, "Unexpected interrupt received!\n");
+				gic_write_eoir(irqnr);
 			}
-
-			WARN_ONCE(true, "Unexpected SPI received!\n");
-			gic_write_eoir(irqnr);
+			continue;
 		}
 		if (irqnr < 16) {
 			gic_write_eoir(irqnr);
@@ -323,8 +335,8 @@ static int gic_populate_rdist(void)
 	       MPIDR_AFFINITY_LEVEL(mpidr, 1) << 8 |
 	       MPIDR_AFFINITY_LEVEL(mpidr, 0));
 
-	for (i = 0; i < gic_data.redist_regions; i++) {
-		void __iomem *ptr = gic_data.redist_base[i];
+	for (i = 0; i < gic_data.nr_redist_regions; i++) {
+		void __iomem *ptr = gic_data.redist_regions[i].redist_base;
 		u32 reg;
 
 		reg = readl_relaxed(ptr + GICR_PIDR2) & GIC_PIDR2_ARCH_MASK;
@@ -337,10 +349,13 @@ static int gic_populate_rdist(void)
 		do {
 			typer = readq_relaxed(ptr + GICR_TYPER);
 			if ((typer >> 32) == aff) {
+				u64 offset = ptr - gic_data.redist_regions[i].redist_base;
 				gic_data_rdist_rd_base() = ptr;
-				pr_info("CPU%d: found redistributor %llx @%p\n",
+				gic_data_rdist()->phys_base = gic_data.redist_regions[i].phys_base + offset;
+				pr_info("CPU%d: found redistributor %llx region %d:%pa\n",
 					smp_processor_id(),
-					(unsigned long long)mpidr, ptr);
+					(unsigned long long)mpidr,
+					i, &gic_data_rdist()->phys_base);
 				return 0;
 			}
 
@@ -360,6 +375,26 @@ static int gic_populate_rdist(void)
 	return -ENODEV;
 }
 
+static void gic_cpu_sys_reg_init(void)
+{
+	/* Enable system registers */
+	gic_enable_sre();
+
+	/* Set priority mask register */
+	gic_write_pmr(DEFAULT_PMR_VALUE);
+
+	/* EOI deactivates interrupt too (mode 0) */
+	gic_write_ctlr(ICC_CTLR_EL1_EOImode_drop_dir);
+
+	/* ... and let's hit the road... */
+	gic_write_grpen1(1);
+}
+
+static int gic_dist_supports_lpis(void)
+{
+	return !!(readl_relaxed(gic_data.dist_base + GICD_TYPER) & GICD_TYPER_LPIS);
+}
+
 static void gic_cpu_init(void)
 {
 	void __iomem *rbase;
@@ -368,23 +403,18 @@ static void gic_cpu_init(void)
 	if (gic_populate_rdist())
 		return;
 
-	gic_enable_redist();
+	gic_enable_redist(true);
 
 	rbase = gic_data_rdist_sgi_base();
 
 	gic_cpu_config(rbase, gic_redist_wait_for_rwp);
 
-	/* Enable system registers */
-	gic_enable_sre();
+	/* Give LPIs a spin */
+	if (IS_ENABLED(CONFIG_ARM_GIC_V3_ITS) && gic_dist_supports_lpis())
+		its_cpu_init();
 
-	/* Set priority mask register */
-	gic_write_pmr(DEFAULT_PMR_VALUE);
-
-	/* EOI deactivates interrupt too (mode 0) */
-	gic_write_ctlr(ICC_CTLR_EL1_EOImode_drop_dir);
-
-	/* ... and let's hit the road... */
-	gic_write_grpen1(1);
+	/* initialise system registers */
+	gic_cpu_sys_reg_init();
 }
 
 #ifdef CONFIG_SMP
@@ -533,6 +563,33 @@ static int gic_set_affinity(struct irq_data *d, const struct cpumask *mask_val,
 #define gic_smp_init()		do { } while(0)
 #endif
 
+#ifdef CONFIG_CPU_PM
+static int gic_cpu_pm_notifier(struct notifier_block *self,
+			       unsigned long cmd, void *v)
+{
+	if (cmd == CPU_PM_EXIT) {
+		gic_enable_redist(true);
+		gic_cpu_sys_reg_init();
+	} else if (cmd == CPU_PM_ENTER) {
+		gic_write_grpen1(0);
+		gic_enable_redist(false);
+	}
+	return NOTIFY_OK;
+}
+
+static struct notifier_block gic_cpu_pm_notifier_block = {
+	.notifier_call = gic_cpu_pm_notifier,
+};
+
+static void gic_cpu_pm_init(void)
+{
+	cpu_pm_register_notifier(&gic_cpu_pm_notifier_block);
+}
+
+#else
+static inline void gic_cpu_pm_init(void) { }
+#endif /* CONFIG_CPU_PM */
+
 static struct irq_chip gic_chip = {
 	.name			= "GICv3",
 	.irq_mask		= gic_mask_irq,
@@ -542,26 +599,43 @@ static struct irq_chip gic_chip = {
 	.irq_set_affinity	= gic_set_affinity,
 };
 
+#define GIC_ID_NR		(1U << gic_data.rdists.id_bits)
+
 static int gic_irq_domain_map(struct irq_domain *d, unsigned int irq,
 			      irq_hw_number_t hw)
 {
 	/* SGIs are private to the core kernel */
 	if (hw < 16)
 		return -EPERM;
+	/* Nothing here */
+	if (hw >= gic_data.irq_nr && hw < 8192)
+		return -EPERM;
+	/* Off limits */
+	if (hw >= GIC_ID_NR)
+		return -EPERM;
+
 	/* PPIs */
 	if (hw < 32) {
 		irq_set_percpu_devid(irq);
-		irq_set_chip_and_handler(irq, &gic_chip,
-					 handle_percpu_devid_irq);
+		irq_domain_set_info(d, irq, hw, &gic_chip, d->host_data,
+				    handle_percpu_devid_irq, NULL, NULL);
 		set_irq_flags(irq, IRQF_VALID | IRQF_NOAUTOEN);
 	}
 	/* SPIs */
 	if (hw >= 32 && hw < gic_data.irq_nr) {
-		irq_set_chip_and_handler(irq, &gic_chip,
-					 handle_fasteoi_irq);
+		irq_domain_set_info(d, irq, hw, &gic_chip, d->host_data,
+				    handle_fasteoi_irq, NULL, NULL);
 		set_irq_flags(irq, IRQF_VALID | IRQF_PROBE);
 	}
-	irq_set_chip_data(irq, d->host_data);
+	/* LPIs */
+	if (hw >= 8192 && hw < GIC_ID_NR) {
+		if (!gic_dist_supports_lpis())
+			return -EPERM;
+		irq_domain_set_info(d, irq, hw, &gic_chip, d->host_data,
+				    handle_fasteoi_irq, NULL, NULL);
+		set_irq_flags(irq, IRQF_VALID);
+	}
+
 	return 0;
 }
 
@@ -582,6 +656,9 @@ static int gic_irq_domain_xlate(struct irq_domain *d,
 	case 1:			/* PPI */
 		*out_hwirq = intspec[1] + 16;
 		break;
+	case GIC_IRQ_TYPE_LPI:	/* LPI */
+		*out_hwirq = intspec[1];
+		break;
 	default:
 		return -EINVAL;
 	}
@@ -590,17 +667,50 @@ static int gic_irq_domain_xlate(struct irq_domain *d,
 	return 0;
 }
 
+static int gic_irq_domain_alloc(struct irq_domain *domain, unsigned int virq,
+				unsigned int nr_irqs, void *arg)
+{
+	int i, ret;
+	irq_hw_number_t hwirq;
+	unsigned int type = IRQ_TYPE_NONE;
+	struct of_phandle_args *irq_data = arg;
+
+	ret = gic_irq_domain_xlate(domain, irq_data->np, irq_data->args,
+				   irq_data->args_count, &hwirq, &type);
+	if (ret)
+		return ret;
+
+	for (i = 0; i < nr_irqs; i++)
+		gic_irq_domain_map(domain, virq + i, hwirq + i);
+
+	return 0;
+}
+
+static void gic_irq_domain_free(struct irq_domain *domain, unsigned int virq,
+				unsigned int nr_irqs)
+{
+	int i;
+
+	for (i = 0; i < nr_irqs; i++) {
+		struct irq_data *d = irq_domain_get_irq_data(domain, virq + i);
+		irq_set_handler(virq + i, NULL);
+		irq_domain_reset_irq_data(d);
+	}
+}
+
 static const struct irq_domain_ops gic_irq_domain_ops = {
-	.map = gic_irq_domain_map,
 	.xlate = gic_irq_domain_xlate,
+	.alloc = gic_irq_domain_alloc,
+	.free = gic_irq_domain_free,
 };
 
 static int __init gic_of_init(struct device_node *node, struct device_node *parent)
 {
 	void __iomem *dist_base;
-	void __iomem **redist_base;
+	struct redist_region *rdist_regs;
 	u64 redist_stride;
-	u32 redist_regions;
+	u32 nr_redist_regions;
+	u32 typer;
 	u32 reg;
 	int gic_irqs;
 	int err;
@@ -621,69 +731,79 @@ static int __init gic_of_init(struct device_node *node, struct device_node *pare
 		goto out_unmap_dist;
 	}
 
-	if (of_property_read_u32(node, "#redistributor-regions", &redist_regions))
-		redist_regions = 1;
+	if (of_property_read_u32(node, "#redistributor-regions", &nr_redist_regions))
+		nr_redist_regions = 1;
 
-	redist_base = kzalloc(sizeof(*redist_base) * redist_regions, GFP_KERNEL);
-	if (!redist_base) {
+	rdist_regs = kzalloc(sizeof(*rdist_regs) * nr_redist_regions, GFP_KERNEL);
+	if (!rdist_regs) {
 		err = -ENOMEM;
 		goto out_unmap_dist;
 	}
 
-	for (i = 0; i < redist_regions; i++) {
-		redist_base[i] = of_iomap(node, 1 + i);
-		if (!redist_base[i]) {
+	for (i = 0; i < nr_redist_regions; i++) {
+		struct resource res;
+		int ret;
+
+		ret = of_address_to_resource(node, 1 + i, &res);
+		rdist_regs[i].redist_base = of_iomap(node, 1 + i);
+		if (ret || !rdist_regs[i].redist_base) {
 			pr_err("%s: couldn't map region %d\n",
 			       node->full_name, i);
 			err = -ENODEV;
 			goto out_unmap_rdist;
 		}
+		rdist_regs[i].phys_base = res.start;
 	}
 
 	if (of_property_read_u64(node, "redistributor-stride", &redist_stride))
 		redist_stride = 0;
 
 	gic_data.dist_base = dist_base;
-	gic_data.redist_base = redist_base;
-	gic_data.redist_regions = redist_regions;
+	gic_data.redist_regions = rdist_regs;
+	gic_data.nr_redist_regions = nr_redist_regions;
 	gic_data.redist_stride = redist_stride;
 
 	/*
 	 * Find out how many interrupts are supported.
 	 * The GIC only supports up to 1020 interrupt sources (SGI+PPI+SPI)
 	 */
-	gic_irqs = readl_relaxed(gic_data.dist_base + GICD_TYPER) & 0x1f;
-	gic_irqs = (gic_irqs + 1) * 32;
+	typer = readl_relaxed(gic_data.dist_base + GICD_TYPER);
+	gic_data.rdists.id_bits = GICD_TYPER_ID_BITS(typer);
+	gic_irqs = GICD_TYPER_IRQS(typer);
 	if (gic_irqs > 1020)
 		gic_irqs = 1020;
 	gic_data.irq_nr = gic_irqs;
 
 	gic_data.domain = irq_domain_add_tree(node, &gic_irq_domain_ops,
 					      &gic_data);
-	gic_data.rdist = alloc_percpu(typeof(*gic_data.rdist));
+	gic_data.rdists.rdist = alloc_percpu(typeof(*gic_data.rdists.rdist));
 
-	if (WARN_ON(!gic_data.domain) || WARN_ON(!gic_data.rdist)) {
+	if (WARN_ON(!gic_data.domain) || WARN_ON(!gic_data.rdists.rdist)) {
 		err = -ENOMEM;
 		goto out_free;
 	}
 
 	set_handle_irq(gic_handle_irq);
 
+	if (IS_ENABLED(CONFIG_ARM_GIC_V3_ITS) && gic_dist_supports_lpis())
+		its_init(node, &gic_data.rdists, gic_data.domain);
+
 	gic_smp_init();
 	gic_dist_init();
 	gic_cpu_init();
+	gic_cpu_pm_init();
 
 	return 0;
 
 out_free:
 	if (gic_data.domain)
 		irq_domain_remove(gic_data.domain);
-	free_percpu(gic_data.rdist);
+	free_percpu(gic_data.rdists.rdist);
 out_unmap_rdist:
-	for (i = 0; i < redist_regions; i++)
-		if (redist_base[i])
-			iounmap(redist_base[i]);
-	kfree(redist_base);
+	for (i = 0; i < nr_redist_regions; i++)
+		if (rdist_regs[i].redist_base)
+			iounmap(rdist_regs[i].redist_base);
+	kfree(rdist_regs);
 out_unmap_dist:
 	iounmap(dist_base);
 	return err;
diff --git a/drivers/irqchip/irq-gic.c b/drivers/irqchip/irq-gic.c
index dda6dbc23565..d617ee5a3d8a 100644
--- a/drivers/irqchip/irq-gic.c
+++ b/drivers/irqchip/irq-gic.c
@@ -102,7 +102,7 @@ static struct gic_chip_data gic_data[MAX_GIC_NR] __read_mostly;
 #ifdef CONFIG_GIC_NON_BANKED
 static void __iomem *gic_get_percpu_base(union gic_base *base)
 {
-	return *__this_cpu_ptr(base->percpu_base);
+	return raw_cpu_read(*base->percpu_base);
 }
 
 static void __iomem *gic_get_common_base(union gic_base *base)
@@ -270,8 +270,7 @@ static void __exception_irq_entry gic_handle_irq(struct pt_regs *regs)
 		irqnr = irqstat & GICC_IAR_INT_ID_MASK;
 
 		if (likely(irqnr > 15 && irqnr < 1021)) {
-			irqnr = irq_find_mapping(gic->domain, irqnr);
-			handle_IRQ(irqnr, regs);
+			handle_domain_irq(gic->domain, irqnr, regs);
 			continue;
 		}
 		if (irqnr < 16) {
@@ -298,8 +297,8 @@ static void gic_handle_cascade_irq(unsigned int irq, struct irq_desc *desc)
 	status = readl_relaxed(gic_data_cpu_base(chip_data) + GIC_CPU_INTACK);
 	raw_spin_unlock(&irq_controller_lock);
 
-	gic_irq = (status & 0x3ff);
-	if (gic_irq == 1023)
+	gic_irq = (status & GICC_IAR_INT_ID_MASK);
+	if (gic_irq == GICC_INT_SPURIOUS)
 		goto out;
 
 	cascade_irq = irq_find_mapping(chip_data->domain, gic_irq);
@@ -353,6 +352,21 @@ static u8 gic_get_cpumask(struct gic_chip_data *gic)
 	return mask;
 }
 
+static void gic_cpu_if_up(void)
+{
+	void __iomem *cpu_base = gic_data_cpu_base(&gic_data[0]);
+	u32 bypass = 0;
+
+	/*
+	* Preserve bypass disable bits to be written back later
+	*/
+	bypass = readl(cpu_base + GIC_CPU_CTRL);
+	bypass &= GICC_DIS_BYPASS_MASK;
+
+	writel_relaxed(bypass | GICC_ENABLE, cpu_base + GIC_CPU_CTRL);
+}
+
+
 static void __init gic_dist_init(struct gic_chip_data *gic)
 {
 	unsigned int i;
@@ -360,7 +374,7 @@ static void __init gic_dist_init(struct gic_chip_data *gic)
 	unsigned int gic_irqs = gic->gic_irqs;
 	void __iomem *base = gic_data_dist_base(gic);
 
-	writel_relaxed(0, base + GIC_DIST_CTRL);
+	writel_relaxed(GICD_DISABLE, base + GIC_DIST_CTRL);
 
 	/*
 	 * Set all global interrupts to this CPU only.
@@ -373,7 +387,7 @@ static void __init gic_dist_init(struct gic_chip_data *gic)
 
 	gic_dist_config(base, gic_irqs, NULL);
 
-	writel_relaxed(1, base + GIC_DIST_CTRL);
+	writel_relaxed(GICD_ENABLE, base + GIC_DIST_CTRL);
 }
 
 static void gic_cpu_init(struct gic_chip_data *gic)
@@ -400,14 +414,18 @@ static void gic_cpu_init(struct gic_chip_data *gic)
 
 	gic_cpu_config(dist_base, NULL);
 
-	writel_relaxed(0xf0, base + GIC_CPU_PRIMASK);
-	writel_relaxed(1, base + GIC_CPU_CTRL);
+	writel_relaxed(GICC_INT_PRI_THRESHOLD, base + GIC_CPU_PRIMASK);
+	gic_cpu_if_up();
 }
 
 void gic_cpu_if_down(void)
 {
 	void __iomem *cpu_base = gic_data_cpu_base(&gic_data[0]);
-	writel_relaxed(0, cpu_base + GIC_CPU_CTRL);
+	u32 val = 0;
+
+	val = readl(cpu_base + GIC_CPU_CTRL);
+	val &= ~GICC_ENABLE;
+	writel_relaxed(val, cpu_base + GIC_CPU_CTRL);
 }
 
 #ifdef CONFIG_CPU_PM
@@ -467,14 +485,14 @@ static void gic_dist_restore(unsigned int gic_nr)
 	if (!dist_base)
 		return;
 
-	writel_relaxed(0, dist_base + GIC_DIST_CTRL);
+	writel_relaxed(GICD_DISABLE, dist_base + GIC_DIST_CTRL);
 
 	for (i = 0; i < DIV_ROUND_UP(gic_irqs, 16); i++)
 		writel_relaxed(gic_data[gic_nr].saved_spi_conf[i],
 			dist_base + GIC_DIST_CONFIG + i * 4);
 
 	for (i = 0; i < DIV_ROUND_UP(gic_irqs, 4); i++)
-		writel_relaxed(0xa0a0a0a0,
+		writel_relaxed(GICD_INT_DEF_PRI_X4,
 			dist_base + GIC_DIST_PRI + i * 4);
 
 	for (i = 0; i < DIV_ROUND_UP(gic_irqs, 4); i++)
@@ -485,7 +503,7 @@ static void gic_dist_restore(unsigned int gic_nr)
 		writel_relaxed(gic_data[gic_nr].saved_spi_enable[i],
 			dist_base + GIC_DIST_ENABLE_SET + i * 4);
 
-	writel_relaxed(1, dist_base + GIC_DIST_CTRL);
+	writel_relaxed(GICD_ENABLE, dist_base + GIC_DIST_CTRL);
 }
 
 static void gic_cpu_save(unsigned int gic_nr)
@@ -504,11 +522,11 @@ static void gic_cpu_save(unsigned int gic_nr)
 	if (!dist_base || !cpu_base)
 		return;
 
-	ptr = __this_cpu_ptr(gic_data[gic_nr].saved_ppi_enable);
+	ptr = raw_cpu_ptr(gic_data[gic_nr].saved_ppi_enable);
 	for (i = 0; i < DIV_ROUND_UP(32, 32); i++)
 		ptr[i] = readl_relaxed(dist_base + GIC_DIST_ENABLE_SET + i * 4);
 
-	ptr = __this_cpu_ptr(gic_data[gic_nr].saved_ppi_conf);
+	ptr = raw_cpu_ptr(gic_data[gic_nr].saved_ppi_conf);
 	for (i = 0; i < DIV_ROUND_UP(32, 16); i++)
 		ptr[i] = readl_relaxed(dist_base + GIC_DIST_CONFIG + i * 4);
 
@@ -530,19 +548,20 @@ static void gic_cpu_restore(unsigned int gic_nr)
 	if (!dist_base || !cpu_base)
 		return;
 
-	ptr = __this_cpu_ptr(gic_data[gic_nr].saved_ppi_enable);
+	ptr = raw_cpu_ptr(gic_data[gic_nr].saved_ppi_enable);
 	for (i = 0; i < DIV_ROUND_UP(32, 32); i++)
 		writel_relaxed(ptr[i], dist_base + GIC_DIST_ENABLE_SET + i * 4);
 
-	ptr = __this_cpu_ptr(gic_data[gic_nr].saved_ppi_conf);
+	ptr = raw_cpu_ptr(gic_data[gic_nr].saved_ppi_conf);
 	for (i = 0; i < DIV_ROUND_UP(32, 16); i++)
 		writel_relaxed(ptr[i], dist_base + GIC_DIST_CONFIG + i * 4);
 
 	for (i = 0; i < DIV_ROUND_UP(32, 4); i++)
-		writel_relaxed(0xa0a0a0a0, dist_base + GIC_DIST_PRI + i * 4);
+		writel_relaxed(GICD_INT_DEF_PRI_X4,
+					dist_base + GIC_DIST_PRI + i * 4);
 
-	writel_relaxed(0xf0, cpu_base + GIC_CPU_PRIMASK);
-	writel_relaxed(1, cpu_base + GIC_CPU_CTRL);
+	writel_relaxed(GICC_INT_PRI_THRESHOLD, cpu_base + GIC_CPU_PRIMASK);
+	gic_cpu_if_up();
 }
 
 static int gic_notifier(struct notifier_block *self, unsigned long cmd,	void *v)
@@ -769,17 +788,16 @@ static int gic_irq_domain_map(struct irq_domain *d, unsigned int irq,
 {
 	if (hw < 32) {
 		irq_set_percpu_devid(irq);
-		irq_set_chip_and_handler(irq, &gic_chip,
-					 handle_percpu_devid_irq);
+		irq_domain_set_info(d, irq, hw, &gic_chip, d->host_data,
+				    handle_percpu_devid_irq, NULL, NULL);
 		set_irq_flags(irq, IRQF_VALID | IRQF_NOAUTOEN);
 	} else {
-		irq_set_chip_and_handler(irq, &gic_chip,
-					 handle_fasteoi_irq);
+		irq_domain_set_info(d, irq, hw, &gic_chip, d->host_data,
+				    handle_fasteoi_irq, NULL, NULL);
 		set_irq_flags(irq, IRQF_VALID | IRQF_PROBE);
 
 		gic_routable_irq_domain_ops->map(d, irq, hw);
 	}
-	irq_set_chip_data(irq, d->host_data);
 	return 0;
 }
 
@@ -839,6 +857,31 @@ static struct notifier_block gic_cpu_notifier = {
 };
 #endif
 
+static int gic_irq_domain_alloc(struct irq_domain *domain, unsigned int virq,
+				unsigned int nr_irqs, void *arg)
+{
+	int i, ret;
+	irq_hw_number_t hwirq;
+	unsigned int type = IRQ_TYPE_NONE;
+	struct of_phandle_args *irq_data = arg;
+
+	ret = gic_irq_domain_xlate(domain, irq_data->np, irq_data->args,
+				   irq_data->args_count, &hwirq, &type);
+	if (ret)
+		return ret;
+
+	for (i = 0; i < nr_irqs; i++)
+		gic_irq_domain_map(domain, virq + i, hwirq + i);
+
+	return 0;
+}
+
+static const struct irq_domain_ops gic_irq_domain_hierarchy_ops = {
+	.xlate = gic_irq_domain_xlate,
+	.alloc = gic_irq_domain_alloc,
+	.free = irq_domain_free_irqs_top,
+};
+
 static const struct irq_domain_ops gic_irq_domain_ops = {
 	.map = gic_irq_domain_map,
 	.unmap = gic_irq_domain_unmap,
@@ -929,18 +972,6 @@ void __init gic_init_bases(unsigned int gic_nr, int irq_start,
 		gic_cpu_map[i] = 0xff;
 
 	/*
-	 * For primary GICs, skip over SGIs.
-	 * For secondary GICs, skip over PPIs, too.
-	 */
-	if (gic_nr == 0 && (irq_start & 31) > 0) {
-		hwirq_base = 16;
-		if (irq_start != -1)
-			irq_start = (irq_start & ~31) + 16;
-	} else {
-		hwirq_base = 32;
-	}
-
-	/*
 	 * Find out how many interrupts are supported.
 	 * The GIC only supports up to 1020 interrupt sources.
 	 */
@@ -950,10 +981,31 @@ void __init gic_init_bases(unsigned int gic_nr, int irq_start,
 		gic_irqs = 1020;
 	gic->gic_irqs = gic_irqs;
 
-	gic_irqs -= hwirq_base; /* calculate # of irqs to allocate */
+	if (node) {		/* DT case */
+		const struct irq_domain_ops *ops = &gic_irq_domain_hierarchy_ops;
+
+		if (!of_property_read_u32(node, "arm,routable-irqs",
+					  &nr_routable_irqs)) {
+			ops = &gic_irq_domain_ops;
+			gic_irqs = nr_routable_irqs;
+		}
+
+		gic->domain = irq_domain_add_linear(node, gic_irqs, ops, gic);
+	} else {		/* Non-DT case */
+		/*
+		 * For primary GICs, skip over SGIs.
+		 * For secondary GICs, skip over PPIs, too.
+		 */
+		if (gic_nr == 0 && (irq_start & 31) > 0) {
+			hwirq_base = 16;
+			if (irq_start != -1)
+				irq_start = (irq_start & ~31) + 16;
+		} else {
+			hwirq_base = 32;
+		}
+
+		gic_irqs -= hwirq_base; /* calculate # of irqs to allocate */
 
-	if (of_property_read_u32(node, "arm,routable-irqs",
-				 &nr_routable_irqs)) {
 		irq_base = irq_alloc_descs(irq_start, 16, gic_irqs,
 					   numa_node_id());
 		if (IS_ERR_VALUE(irq_base)) {
@@ -964,10 +1016,6 @@ void __init gic_init_bases(unsigned int gic_nr, int irq_start,
 
 		gic->domain = irq_domain_add_legacy(node, gic_irqs, irq_base,
 					hwirq_base, &gic_irq_domain_ops, gic);
-	} else {
-		gic->domain = irq_domain_add_linear(node, nr_routable_irqs,
-						    &gic_irq_domain_ops,
-						    gic);
 	}
 
 	if (WARN_ON(!gic->domain))
@@ -1018,10 +1066,16 @@ gic_of_init(struct device_node *node, struct device_node *parent)
 		irq = irq_of_parse_and_map(node, 0);
 		gic_cascade_irq(gic_cnt, irq);
 	}
+
+	if (IS_ENABLED(CONFIG_ARM_GIC_V2M))
+		gicv2m_of_init(node, gic_data[gic_cnt].domain);
+
 	gic_cnt++;
 	return 0;
 }
 IRQCHIP_DECLARE(gic_400, "arm,gic-400", gic_of_init);
+IRQCHIP_DECLARE(arm11mp_gic, "arm,arm11mp-gic", gic_of_init);
+IRQCHIP_DECLARE(arm1176jzf_dc_gic, "arm,arm1176jzf-devchip-gic", gic_of_init);
 IRQCHIP_DECLARE(cortex_a15_gic, "arm,cortex-a15-gic", gic_of_init);
 IRQCHIP_DECLARE(cortex_a9_gic, "arm,cortex-a9-gic", gic_of_init);
 IRQCHIP_DECLARE(cortex_a7_gic, "arm,cortex-a7-gic", gic_of_init);
diff --git a/drivers/irqchip/irq-hip04.c b/drivers/irqchip/irq-hip04.c
new file mode 100644
index 000000000000..6bc2deb73d53
--- /dev/null
+++ b/drivers/irqchip/irq-hip04.c
@@ -0,0 +1,423 @@
+/*
+ * Hisilicon HiP04 INTC
+ *
+ * Copyright (C) 2002-2014 ARM Limited.
+ * Copyright (c) 2013-2014 Hisilicon Ltd.
+ * Copyright (c) 2013-2014 Linaro Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * Interrupt architecture for the HIP04 INTC:
+ *
+ * o There is one Interrupt Distributor, which receives interrupts
+ *   from system devices and sends them to the Interrupt Controllers.
+ *
+ * o There is one CPU Interface per CPU, which sends interrupts sent
+ *   by the Distributor, and interrupts generated locally, to the
+ *   associated CPU. The base address of the CPU interface is usually
+ *   aliased so that the same address points to different chips depending
+ *   on the CPU it is accessed from.
+ *
+ * Note that IRQs 0-31 are special - they are local to each CPU.
+ * As such, the enable set/clear, pending set/clear and active bit
+ * registers are banked per-cpu for these sources.
+ */
+
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/err.h>
+#include <linux/module.h>
+#include <linux/list.h>
+#include <linux/smp.h>
+#include <linux/cpu.h>
+#include <linux/cpu_pm.h>
+#include <linux/cpumask.h>
+#include <linux/io.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/irqdomain.h>
+#include <linux/interrupt.h>
+#include <linux/slab.h>
+#include <linux/irqchip/arm-gic.h>
+
+#include <asm/irq.h>
+#include <asm/exception.h>
+#include <asm/smp_plat.h>
+
+#include "irq-gic-common.h"
+#include "irqchip.h"
+
+#define HIP04_MAX_IRQS		510
+
+struct hip04_irq_data {
+	void __iomem *dist_base;
+	void __iomem *cpu_base;
+	struct irq_domain *domain;
+	unsigned int nr_irqs;
+};
+
+static DEFINE_RAW_SPINLOCK(irq_controller_lock);
+
+/*
+ * The GIC mapping of CPU interfaces does not necessarily match
+ * the logical CPU numbering.  Let's use a mapping as returned
+ * by the GIC itself.
+ */
+#define NR_HIP04_CPU_IF 16
+static u16 hip04_cpu_map[NR_HIP04_CPU_IF] __read_mostly;
+
+static struct hip04_irq_data hip04_data __read_mostly;
+
+static inline void __iomem *hip04_dist_base(struct irq_data *d)
+{
+	struct hip04_irq_data *hip04_data = irq_data_get_irq_chip_data(d);
+	return hip04_data->dist_base;
+}
+
+static inline void __iomem *hip04_cpu_base(struct irq_data *d)
+{
+	struct hip04_irq_data *hip04_data = irq_data_get_irq_chip_data(d);
+	return hip04_data->cpu_base;
+}
+
+static inline unsigned int hip04_irq(struct irq_data *d)
+{
+	return d->hwirq;
+}
+
+/*
+ * Routines to acknowledge, disable and enable interrupts
+ */
+static void hip04_mask_irq(struct irq_data *d)
+{
+	u32 mask = 1 << (hip04_irq(d) % 32);
+
+	raw_spin_lock(&irq_controller_lock);
+	writel_relaxed(mask, hip04_dist_base(d) + GIC_DIST_ENABLE_CLEAR +
+		       (hip04_irq(d) / 32) * 4);
+	raw_spin_unlock(&irq_controller_lock);
+}
+
+static void hip04_unmask_irq(struct irq_data *d)
+{
+	u32 mask = 1 << (hip04_irq(d) % 32);
+
+	raw_spin_lock(&irq_controller_lock);
+	writel_relaxed(mask, hip04_dist_base(d) + GIC_DIST_ENABLE_SET +
+		       (hip04_irq(d) / 32) * 4);
+	raw_spin_unlock(&irq_controller_lock);
+}
+
+static void hip04_eoi_irq(struct irq_data *d)
+{
+	writel_relaxed(hip04_irq(d), hip04_cpu_base(d) + GIC_CPU_EOI);
+}
+
+static int hip04_irq_set_type(struct irq_data *d, unsigned int type)
+{
+	void __iomem *base = hip04_dist_base(d);
+	unsigned int irq = hip04_irq(d);
+
+	/* Interrupt configuration for SGIs can't be changed */
+	if (irq < 16)
+		return -EINVAL;
+
+	if (type != IRQ_TYPE_LEVEL_HIGH && type != IRQ_TYPE_EDGE_RISING)
+		return -EINVAL;
+
+	raw_spin_lock(&irq_controller_lock);
+
+	gic_configure_irq(irq, type, base, NULL);
+
+	raw_spin_unlock(&irq_controller_lock);
+
+	return 0;
+}
+
+#ifdef CONFIG_SMP
+static int hip04_irq_set_affinity(struct irq_data *d,
+				  const struct cpumask *mask_val,
+				  bool force)
+{
+	void __iomem *reg;
+	unsigned int cpu, shift = (hip04_irq(d) % 2) * 16;
+	u32 val, mask, bit;
+
+	if (!force)
+		cpu = cpumask_any_and(mask_val, cpu_online_mask);
+	else
+		cpu = cpumask_first(mask_val);
+
+	if (cpu >= NR_HIP04_CPU_IF || cpu >= nr_cpu_ids)
+		return -EINVAL;
+
+	raw_spin_lock(&irq_controller_lock);
+	reg = hip04_dist_base(d) + GIC_DIST_TARGET + ((hip04_irq(d) * 2) & ~3);
+	mask = 0xffff << shift;
+	bit = hip04_cpu_map[cpu] << shift;
+	val = readl_relaxed(reg) & ~mask;
+	writel_relaxed(val | bit, reg);
+	raw_spin_unlock(&irq_controller_lock);
+
+	return IRQ_SET_MASK_OK;
+}
+#endif
+
+static void __exception_irq_entry hip04_handle_irq(struct pt_regs *regs)
+{
+	u32 irqstat, irqnr;
+	void __iomem *cpu_base = hip04_data.cpu_base;
+
+	do {
+		irqstat = readl_relaxed(cpu_base + GIC_CPU_INTACK);
+		irqnr = irqstat & GICC_IAR_INT_ID_MASK;
+
+		if (likely(irqnr > 15 && irqnr <= HIP04_MAX_IRQS)) {
+			handle_domain_irq(hip04_data.domain, irqnr, regs);
+			continue;
+		}
+		if (irqnr < 16) {
+			writel_relaxed(irqstat, cpu_base + GIC_CPU_EOI);
+#ifdef CONFIG_SMP
+			handle_IPI(irqnr, regs);
+#endif
+			continue;
+		}
+		break;
+	} while (1);
+}
+
+static struct irq_chip hip04_irq_chip = {
+	.name			= "HIP04 INTC",
+	.irq_mask		= hip04_mask_irq,
+	.irq_unmask		= hip04_unmask_irq,
+	.irq_eoi		= hip04_eoi_irq,
+	.irq_set_type		= hip04_irq_set_type,
+#ifdef CONFIG_SMP
+	.irq_set_affinity	= hip04_irq_set_affinity,
+#endif
+};
+
+static u16 hip04_get_cpumask(struct hip04_irq_data *intc)
+{
+	void __iomem *base = intc->dist_base;
+	u32 mask, i;
+
+	for (i = mask = 0; i < 32; i += 2) {
+		mask = readl_relaxed(base + GIC_DIST_TARGET + i * 2);
+		mask |= mask >> 16;
+		if (mask)
+			break;
+	}
+
+	if (!mask)
+		pr_crit("GIC CPU mask not found - kernel will fail to boot.\n");
+
+	return mask;
+}
+
+static void __init hip04_irq_dist_init(struct hip04_irq_data *intc)
+{
+	unsigned int i;
+	u32 cpumask;
+	unsigned int nr_irqs = intc->nr_irqs;
+	void __iomem *base = intc->dist_base;
+
+	writel_relaxed(0, base + GIC_DIST_CTRL);
+
+	/*
+	 * Set all global interrupts to this CPU only.
+	 */
+	cpumask = hip04_get_cpumask(intc);
+	cpumask |= cpumask << 16;
+	for (i = 32; i < nr_irqs; i += 2)
+		writel_relaxed(cpumask, base + GIC_DIST_TARGET + ((i * 2) & ~3));
+
+	gic_dist_config(base, nr_irqs, NULL);
+
+	writel_relaxed(1, base + GIC_DIST_CTRL);
+}
+
+static void hip04_irq_cpu_init(struct hip04_irq_data *intc)
+{
+	void __iomem *dist_base = intc->dist_base;
+	void __iomem *base = intc->cpu_base;
+	unsigned int cpu_mask, cpu = smp_processor_id();
+	int i;
+
+	/*
+	 * Get what the GIC says our CPU mask is.
+	 */
+	BUG_ON(cpu >= NR_HIP04_CPU_IF);
+	cpu_mask = hip04_get_cpumask(intc);
+	hip04_cpu_map[cpu] = cpu_mask;
+
+	/*
+	 * Clear our mask from the other map entries in case they're
+	 * still undefined.
+	 */
+	for (i = 0; i < NR_HIP04_CPU_IF; i++)
+		if (i != cpu)
+			hip04_cpu_map[i] &= ~cpu_mask;
+
+	gic_cpu_config(dist_base, NULL);
+
+	writel_relaxed(0xf0, base + GIC_CPU_PRIMASK);
+	writel_relaxed(1, base + GIC_CPU_CTRL);
+}
+
+#ifdef CONFIG_SMP
+static void hip04_raise_softirq(const struct cpumask *mask, unsigned int irq)
+{
+	int cpu;
+	unsigned long flags, map = 0;
+
+	raw_spin_lock_irqsave(&irq_controller_lock, flags);
+
+	/* Convert our logical CPU mask into a physical one. */
+	for_each_cpu(cpu, mask)
+		map |= hip04_cpu_map[cpu];
+
+	/*
+	 * Ensure that stores to Normal memory are visible to the
+	 * other CPUs before they observe us issuing the IPI.
+	 */
+	dmb(ishst);
+
+	/* this always happens on GIC0 */
+	writel_relaxed(map << 8 | irq, hip04_data.dist_base + GIC_DIST_SOFTINT);
+
+	raw_spin_unlock_irqrestore(&irq_controller_lock, flags);
+}
+#endif
+
+static int hip04_irq_domain_map(struct irq_domain *d, unsigned int irq,
+				irq_hw_number_t hw)
+{
+	if (hw < 32) {
+		irq_set_percpu_devid(irq);
+		irq_set_chip_and_handler(irq, &hip04_irq_chip,
+					 handle_percpu_devid_irq);
+		set_irq_flags(irq, IRQF_VALID | IRQF_NOAUTOEN);
+	} else {
+		irq_set_chip_and_handler(irq, &hip04_irq_chip,
+					 handle_fasteoi_irq);
+		set_irq_flags(irq, IRQF_VALID | IRQF_PROBE);
+	}
+	irq_set_chip_data(irq, d->host_data);
+	return 0;
+}
+
+static int hip04_irq_domain_xlate(struct irq_domain *d,
+				  struct device_node *controller,
+				  const u32 *intspec, unsigned int intsize,
+				  unsigned long *out_hwirq,
+				  unsigned int *out_type)
+{
+	unsigned long ret = 0;
+
+	if (d->of_node != controller)
+		return -EINVAL;
+	if (intsize < 3)
+		return -EINVAL;
+
+	/* Get the interrupt number and add 16 to skip over SGIs */
+	*out_hwirq = intspec[1] + 16;
+
+	/* For SPIs, we need to add 16 more to get the irq ID number */
+	if (!intspec[0])
+		*out_hwirq += 16;
+
+	*out_type = intspec[2] & IRQ_TYPE_SENSE_MASK;
+
+	return ret;
+}
+
+#ifdef CONFIG_SMP
+static int hip04_irq_secondary_init(struct notifier_block *nfb,
+				    unsigned long action,
+				    void *hcpu)
+{
+	if (action == CPU_STARTING || action == CPU_STARTING_FROZEN)
+		hip04_irq_cpu_init(&hip04_data);
+	return NOTIFY_OK;
+}
+
+/*
+ * Notifier for enabling the INTC CPU interface. Set an arbitrarily high
+ * priority because the GIC needs to be up before the ARM generic timers.
+ */
+static struct notifier_block hip04_irq_cpu_notifier = {
+	.notifier_call	= hip04_irq_secondary_init,
+	.priority	= 100,
+};
+#endif
+
+static const struct irq_domain_ops hip04_irq_domain_ops = {
+	.map	= hip04_irq_domain_map,
+	.xlate	= hip04_irq_domain_xlate,
+};
+
+static int __init
+hip04_of_init(struct device_node *node, struct device_node *parent)
+{
+	irq_hw_number_t hwirq_base = 16;
+	int nr_irqs, irq_base, i;
+
+	if (WARN_ON(!node))
+		return -ENODEV;
+
+	hip04_data.dist_base = of_iomap(node, 0);
+	WARN(!hip04_data.dist_base, "fail to map hip04 intc dist registers\n");
+
+	hip04_data.cpu_base = of_iomap(node, 1);
+	WARN(!hip04_data.cpu_base, "unable to map hip04 intc cpu registers\n");
+
+	/*
+	 * Initialize the CPU interface map to all CPUs.
+	 * It will be refined as each CPU probes its ID.
+	 */
+	for (i = 0; i < NR_HIP04_CPU_IF; i++)
+		hip04_cpu_map[i] = 0xffff;
+
+	/*
+	 * Find out how many interrupts are supported.
+	 * The HIP04 INTC only supports up to 510 interrupt sources.
+	 */
+	nr_irqs = readl_relaxed(hip04_data.dist_base + GIC_DIST_CTR) & 0x1f;
+	nr_irqs = (nr_irqs + 1) * 32;
+	if (nr_irqs > HIP04_MAX_IRQS)
+		nr_irqs = HIP04_MAX_IRQS;
+	hip04_data.nr_irqs = nr_irqs;
+
+	nr_irqs -= hwirq_base; /* calculate # of irqs to allocate */
+
+	irq_base = irq_alloc_descs(-1, hwirq_base, nr_irqs, numa_node_id());
+	if (IS_ERR_VALUE(irq_base)) {
+		pr_err("failed to allocate IRQ numbers\n");
+		return -EINVAL;
+	}
+
+	hip04_data.domain = irq_domain_add_legacy(node, nr_irqs, irq_base,
+						  hwirq_base,
+						  &hip04_irq_domain_ops,
+						  &hip04_data);
+
+	if (WARN_ON(!hip04_data.domain))
+		return -EINVAL;
+
+#ifdef CONFIG_SMP
+	set_smp_cross_call(hip04_raise_softirq);
+	register_cpu_notifier(&hip04_irq_cpu_notifier);
+#endif
+	set_handle_irq(hip04_handle_irq);
+
+	hip04_irq_dist_init(&hip04_data);
+	hip04_irq_cpu_init(&hip04_data);
+
+	return 0;
+}
+IRQCHIP_DECLARE(hip04_intc, "hisilicon,hip04-intc", hip04_of_init);
diff --git a/drivers/irqchip/irq-keystone.c b/drivers/irqchip/irq-keystone.c
new file mode 100644
index 000000000000..78e8b3ce5252
--- /dev/null
+++ b/drivers/irqchip/irq-keystone.c
@@ -0,0 +1,231 @@
+/*
+ * Texas Instruments Keystone IRQ controller IP driver
+ *
+ * Copyright (C) 2014 Texas Instruments, Inc.
+ * Author: Sajesh Kumar Saran <sajesh@ti.com>
+ *	   Grygorii Strashko <grygorii.strashko@ti.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation version 2.
+ *
+ * This program is distributed "as is" WITHOUT ANY WARRANTY of any
+ * kind, whether express or implied; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/irq.h>
+#include <linux/bitops.h>
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/irqdomain.h>
+#include <linux/irqchip/chained_irq.h>
+#include <linux/of.h>
+#include <linux/of_platform.h>
+#include <linux/mfd/syscon.h>
+#include <linux/regmap.h>
+#include "irqchip.h"
+
+
+/* The source ID bits start from 4 to 31 (total 28 bits)*/
+#define BIT_OFS			4
+#define KEYSTONE_N_IRQ		(32 - BIT_OFS)
+
+struct keystone_irq_device {
+	struct device		*dev;
+	struct irq_chip		 chip;
+	u32			 mask;
+	int			 irq;
+	struct irq_domain	*irqd;
+	struct regmap		*devctrl_regs;
+	u32			devctrl_offset;
+};
+
+static inline u32 keystone_irq_readl(struct keystone_irq_device *kirq)
+{
+	int ret;
+	u32 val = 0;
+
+	ret = regmap_read(kirq->devctrl_regs, kirq->devctrl_offset, &val);
+	if (ret < 0)
+		dev_dbg(kirq->dev, "irq read failed ret(%d)\n", ret);
+	return val;
+}
+
+static inline void
+keystone_irq_writel(struct keystone_irq_device *kirq, u32 value)
+{
+	int ret;
+
+	ret = regmap_write(kirq->devctrl_regs, kirq->devctrl_offset, value);
+	if (ret < 0)
+		dev_dbg(kirq->dev, "irq write failed ret(%d)\n", ret);
+}
+
+static void keystone_irq_setmask(struct irq_data *d)
+{
+	struct keystone_irq_device *kirq = irq_data_get_irq_chip_data(d);
+
+	kirq->mask |= BIT(d->hwirq);
+	dev_dbg(kirq->dev, "mask %lu [%x]\n", d->hwirq, kirq->mask);
+}
+
+static void keystone_irq_unmask(struct irq_data *d)
+{
+	struct keystone_irq_device *kirq = irq_data_get_irq_chip_data(d);
+
+	kirq->mask &= ~BIT(d->hwirq);
+	dev_dbg(kirq->dev, "unmask %lu [%x]\n", d->hwirq, kirq->mask);
+}
+
+static void keystone_irq_ack(struct irq_data *d)
+{
+	/* nothing to do here */
+}
+
+static void keystone_irq_handler(unsigned irq, struct irq_desc *desc)
+{
+	struct keystone_irq_device *kirq = irq_desc_get_handler_data(desc);
+	unsigned long pending;
+	int src, virq;
+
+	dev_dbg(kirq->dev, "start irq %d\n", irq);
+
+	chained_irq_enter(irq_desc_get_chip(desc), desc);
+
+	pending = keystone_irq_readl(kirq);
+	keystone_irq_writel(kirq, pending);
+
+	dev_dbg(kirq->dev, "pending 0x%lx, mask 0x%x\n", pending, kirq->mask);
+
+	pending = (pending >> BIT_OFS) & ~kirq->mask;
+
+	dev_dbg(kirq->dev, "pending after mask 0x%lx\n", pending);
+
+	for (src = 0; src < KEYSTONE_N_IRQ; src++) {
+		if (BIT(src) & pending) {
+			virq = irq_find_mapping(kirq->irqd, src);
+			dev_dbg(kirq->dev, "dispatch bit %d, virq %d\n",
+				src, virq);
+			if (!virq)
+				dev_warn(kirq->dev, "sporious irq detected hwirq %d, virq %d\n",
+					 src, virq);
+			generic_handle_irq(virq);
+		}
+	}
+
+	chained_irq_exit(irq_desc_get_chip(desc), desc);
+
+	dev_dbg(kirq->dev, "end irq %d\n", irq);
+}
+
+static int keystone_irq_map(struct irq_domain *h, unsigned int virq,
+				irq_hw_number_t hw)
+{
+	struct keystone_irq_device *kirq = h->host_data;
+
+	irq_set_chip_data(virq, kirq);
+	irq_set_chip_and_handler(virq, &kirq->chip, handle_level_irq);
+	set_irq_flags(virq, IRQF_VALID | IRQF_PROBE);
+	return 0;
+}
+
+static struct irq_domain_ops keystone_irq_ops = {
+	.map	= keystone_irq_map,
+	.xlate	= irq_domain_xlate_onecell,
+};
+
+static int keystone_irq_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct device_node *np = dev->of_node;
+	struct keystone_irq_device *kirq;
+	int ret;
+
+	if (np == NULL)
+		return -EINVAL;
+
+	kirq = devm_kzalloc(dev, sizeof(*kirq), GFP_KERNEL);
+	if (!kirq)
+		return -ENOMEM;
+
+	kirq->devctrl_regs =
+		syscon_regmap_lookup_by_phandle(np, "ti,syscon-dev");
+	if (IS_ERR(kirq->devctrl_regs))
+		return PTR_ERR(kirq->devctrl_regs);
+
+	ret = of_property_read_u32_index(np, "ti,syscon-dev", 1,
+					 &kirq->devctrl_offset);
+	if (ret) {
+		dev_err(dev, "couldn't read the devctrl_offset offset!\n");
+		return ret;
+	}
+
+	kirq->irq = platform_get_irq(pdev, 0);
+	if (kirq->irq < 0) {
+		dev_err(dev, "no irq resource %d\n", kirq->irq);
+		return kirq->irq;
+	}
+
+	kirq->dev = dev;
+	kirq->mask = ~0x0;
+	kirq->chip.name		= "keystone-irq";
+	kirq->chip.irq_ack	= keystone_irq_ack;
+	kirq->chip.irq_mask	= keystone_irq_setmask;
+	kirq->chip.irq_unmask	= keystone_irq_unmask;
+
+	kirq->irqd = irq_domain_add_linear(np, KEYSTONE_N_IRQ,
+					   &keystone_irq_ops, kirq);
+	if (!kirq->irqd) {
+		dev_err(dev, "IRQ domain registration failed\n");
+		return -ENODEV;
+	}
+
+	platform_set_drvdata(pdev, kirq);
+
+	irq_set_chained_handler(kirq->irq, keystone_irq_handler);
+	irq_set_handler_data(kirq->irq, kirq);
+
+	/* clear all source bits */
+	keystone_irq_writel(kirq, ~0x0);
+
+	dev_info(dev, "irqchip registered, nr_irqs %u\n", KEYSTONE_N_IRQ);
+
+	return 0;
+}
+
+static int keystone_irq_remove(struct platform_device *pdev)
+{
+	struct keystone_irq_device *kirq = platform_get_drvdata(pdev);
+	int hwirq;
+
+	for (hwirq = 0; hwirq < KEYSTONE_N_IRQ; hwirq++)
+		irq_dispose_mapping(irq_find_mapping(kirq->irqd, hwirq));
+
+	irq_domain_remove(kirq->irqd);
+	return 0;
+}
+
+static const struct of_device_id keystone_irq_dt_ids[] = {
+	{ .compatible = "ti,keystone-irq", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, keystone_irq_dt_ids);
+
+static struct platform_driver keystone_irq_device_driver = {
+	.probe		= keystone_irq_probe,
+	.remove		= keystone_irq_remove,
+	.driver		= {
+		.name	= "keystone_irq",
+		.of_match_table	= of_match_ptr(keystone_irq_dt_ids),
+	}
+};
+
+module_platform_driver(keystone_irq_device_driver);
+
+MODULE_AUTHOR("Texas Instruments");
+MODULE_AUTHOR("Sajesh Kumar Saran");
+MODULE_AUTHOR("Grygorii Strashko");
+MODULE_DESCRIPTION("Keystone IRQ chip");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/irqchip/irq-mips-gic.c b/drivers/irqchip/irq-mips-gic.c
new file mode 100644
index 000000000000..2b0468e3df6a
--- /dev/null
+++ b/drivers/irqchip/irq-mips-gic.c
@@ -0,0 +1,789 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2008 Ralf Baechle (ralf@linux-mips.org)
+ * Copyright (C) 2012 MIPS Technologies, Inc.  All rights reserved.
+ */
+#include <linux/bitmap.h>
+#include <linux/clocksource.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/irqchip/mips-gic.h>
+#include <linux/of_address.h>
+#include <linux/sched.h>
+#include <linux/smp.h>
+
+#include <asm/mips-cm.h>
+#include <asm/setup.h>
+#include <asm/traps.h>
+
+#include <dt-bindings/interrupt-controller/mips-gic.h>
+
+#include "irqchip.h"
+
+unsigned int gic_present;
+
+struct gic_pcpu_mask {
+	DECLARE_BITMAP(pcpu_mask, GIC_MAX_INTRS);
+};
+
+static void __iomem *gic_base;
+static struct gic_pcpu_mask pcpu_masks[NR_CPUS];
+static DEFINE_SPINLOCK(gic_lock);
+static struct irq_domain *gic_irq_domain;
+static int gic_shared_intrs;
+static int gic_vpes;
+static unsigned int gic_cpu_pin;
+static struct irq_chip gic_level_irq_controller, gic_edge_irq_controller;
+
+static void __gic_irq_dispatch(void);
+
+static inline unsigned int gic_read(unsigned int reg)
+{
+	return __raw_readl(gic_base + reg);
+}
+
+static inline void gic_write(unsigned int reg, unsigned int val)
+{
+	__raw_writel(val, gic_base + reg);
+}
+
+static inline void gic_update_bits(unsigned int reg, unsigned int mask,
+				   unsigned int val)
+{
+	unsigned int regval;
+
+	regval = gic_read(reg);
+	regval &= ~mask;
+	regval |= val;
+	gic_write(reg, regval);
+}
+
+static inline void gic_reset_mask(unsigned int intr)
+{
+	gic_write(GIC_REG(SHARED, GIC_SH_RMASK) + GIC_INTR_OFS(intr),
+		  1 << GIC_INTR_BIT(intr));
+}
+
+static inline void gic_set_mask(unsigned int intr)
+{
+	gic_write(GIC_REG(SHARED, GIC_SH_SMASK) + GIC_INTR_OFS(intr),
+		  1 << GIC_INTR_BIT(intr));
+}
+
+static inline void gic_set_polarity(unsigned int intr, unsigned int pol)
+{
+	gic_update_bits(GIC_REG(SHARED, GIC_SH_SET_POLARITY) +
+			GIC_INTR_OFS(intr), 1 << GIC_INTR_BIT(intr),
+			pol << GIC_INTR_BIT(intr));
+}
+
+static inline void gic_set_trigger(unsigned int intr, unsigned int trig)
+{
+	gic_update_bits(GIC_REG(SHARED, GIC_SH_SET_TRIGGER) +
+			GIC_INTR_OFS(intr), 1 << GIC_INTR_BIT(intr),
+			trig << GIC_INTR_BIT(intr));
+}
+
+static inline void gic_set_dual_edge(unsigned int intr, unsigned int dual)
+{
+	gic_update_bits(GIC_REG(SHARED, GIC_SH_SET_DUAL) + GIC_INTR_OFS(intr),
+			1 << GIC_INTR_BIT(intr),
+			dual << GIC_INTR_BIT(intr));
+}
+
+static inline void gic_map_to_pin(unsigned int intr, unsigned int pin)
+{
+	gic_write(GIC_REG(SHARED, GIC_SH_INTR_MAP_TO_PIN_BASE) +
+		  GIC_SH_MAP_TO_PIN(intr), GIC_MAP_TO_PIN_MSK | pin);
+}
+
+static inline void gic_map_to_vpe(unsigned int intr, unsigned int vpe)
+{
+	gic_write(GIC_REG(SHARED, GIC_SH_INTR_MAP_TO_VPE_BASE) +
+		  GIC_SH_MAP_TO_VPE_REG_OFF(intr, vpe),
+		  GIC_SH_MAP_TO_VPE_REG_BIT(vpe));
+}
+
+#ifdef CONFIG_CLKSRC_MIPS_GIC
+cycle_t gic_read_count(void)
+{
+	unsigned int hi, hi2, lo;
+
+	do {
+		hi = gic_read(GIC_REG(SHARED, GIC_SH_COUNTER_63_32));
+		lo = gic_read(GIC_REG(SHARED, GIC_SH_COUNTER_31_00));
+		hi2 = gic_read(GIC_REG(SHARED, GIC_SH_COUNTER_63_32));
+	} while (hi2 != hi);
+
+	return (((cycle_t) hi) << 32) + lo;
+}
+
+unsigned int gic_get_count_width(void)
+{
+	unsigned int bits, config;
+
+	config = gic_read(GIC_REG(SHARED, GIC_SH_CONFIG));
+	bits = 32 + 4 * ((config & GIC_SH_CONFIG_COUNTBITS_MSK) >>
+			 GIC_SH_CONFIG_COUNTBITS_SHF);
+
+	return bits;
+}
+
+void gic_write_compare(cycle_t cnt)
+{
+	gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_COMPARE_HI),
+				(int)(cnt >> 32));
+	gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_COMPARE_LO),
+				(int)(cnt & 0xffffffff));
+}
+
+void gic_write_cpu_compare(cycle_t cnt, int cpu)
+{
+	unsigned long flags;
+
+	local_irq_save(flags);
+
+	gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_OTHER_ADDR), cpu);
+	gic_write(GIC_REG(VPE_OTHER, GIC_VPE_COMPARE_HI),
+				(int)(cnt >> 32));
+	gic_write(GIC_REG(VPE_OTHER, GIC_VPE_COMPARE_LO),
+				(int)(cnt & 0xffffffff));
+
+	local_irq_restore(flags);
+}
+
+cycle_t gic_read_compare(void)
+{
+	unsigned int hi, lo;
+
+	hi = gic_read(GIC_REG(VPE_LOCAL, GIC_VPE_COMPARE_HI));
+	lo = gic_read(GIC_REG(VPE_LOCAL, GIC_VPE_COMPARE_LO));
+
+	return (((cycle_t) hi) << 32) + lo;
+}
+#endif
+
+static bool gic_local_irq_is_routable(int intr)
+{
+	u32 vpe_ctl;
+
+	/* All local interrupts are routable in EIC mode. */
+	if (cpu_has_veic)
+		return true;
+
+	vpe_ctl = gic_read(GIC_REG(VPE_LOCAL, GIC_VPE_CTL));
+	switch (intr) {
+	case GIC_LOCAL_INT_TIMER:
+		return vpe_ctl & GIC_VPE_CTL_TIMER_RTBL_MSK;
+	case GIC_LOCAL_INT_PERFCTR:
+		return vpe_ctl & GIC_VPE_CTL_PERFCNT_RTBL_MSK;
+	case GIC_LOCAL_INT_FDC:
+		return vpe_ctl & GIC_VPE_CTL_FDC_RTBL_MSK;
+	case GIC_LOCAL_INT_SWINT0:
+	case GIC_LOCAL_INT_SWINT1:
+		return vpe_ctl & GIC_VPE_CTL_SWINT_RTBL_MSK;
+	default:
+		return true;
+	}
+}
+
+unsigned int gic_get_timer_pending(void)
+{
+	unsigned int vpe_pending;
+
+	vpe_pending = gic_read(GIC_REG(VPE_LOCAL, GIC_VPE_PEND));
+	return vpe_pending & GIC_VPE_PEND_TIMER_MSK;
+}
+
+static void gic_bind_eic_interrupt(int irq, int set)
+{
+	/* Convert irq vector # to hw int # */
+	irq -= GIC_PIN_TO_VEC_OFFSET;
+
+	/* Set irq to use shadow set */
+	gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_EIC_SHADOW_SET_BASE) +
+		  GIC_VPE_EIC_SS(irq), set);
+}
+
+void gic_send_ipi(unsigned int intr)
+{
+	gic_write(GIC_REG(SHARED, GIC_SH_WEDGE), GIC_SH_WEDGE_SET(intr));
+}
+
+int gic_get_c0_compare_int(void)
+{
+	if (!gic_local_irq_is_routable(GIC_LOCAL_INT_TIMER))
+		return MIPS_CPU_IRQ_BASE + cp0_compare_irq;
+	return irq_create_mapping(gic_irq_domain,
+				  GIC_LOCAL_TO_HWIRQ(GIC_LOCAL_INT_TIMER));
+}
+
+int gic_get_c0_perfcount_int(void)
+{
+	if (!gic_local_irq_is_routable(GIC_LOCAL_INT_PERFCTR)) {
+		/* Is the erformance counter shared with the timer? */
+		if (cp0_perfcount_irq < 0)
+			return -1;
+		return MIPS_CPU_IRQ_BASE + cp0_perfcount_irq;
+	}
+	return irq_create_mapping(gic_irq_domain,
+				  GIC_LOCAL_TO_HWIRQ(GIC_LOCAL_INT_PERFCTR));
+}
+
+static unsigned int gic_get_int(void)
+{
+	unsigned int i;
+	unsigned long *pcpu_mask;
+	unsigned long pending_reg, intrmask_reg;
+	DECLARE_BITMAP(pending, GIC_MAX_INTRS);
+	DECLARE_BITMAP(intrmask, GIC_MAX_INTRS);
+
+	/* Get per-cpu bitmaps */
+	pcpu_mask = pcpu_masks[smp_processor_id()].pcpu_mask;
+
+	pending_reg = GIC_REG(SHARED, GIC_SH_PEND);
+	intrmask_reg = GIC_REG(SHARED, GIC_SH_MASK);
+
+	for (i = 0; i < BITS_TO_LONGS(gic_shared_intrs); i++) {
+		pending[i] = gic_read(pending_reg);
+		intrmask[i] = gic_read(intrmask_reg);
+		pending_reg += 0x4;
+		intrmask_reg += 0x4;
+	}
+
+	bitmap_and(pending, pending, intrmask, gic_shared_intrs);
+	bitmap_and(pending, pending, pcpu_mask, gic_shared_intrs);
+
+	return find_first_bit(pending, gic_shared_intrs);
+}
+
+static void gic_mask_irq(struct irq_data *d)
+{
+	gic_reset_mask(GIC_HWIRQ_TO_SHARED(d->hwirq));
+}
+
+static void gic_unmask_irq(struct irq_data *d)
+{
+	gic_set_mask(GIC_HWIRQ_TO_SHARED(d->hwirq));
+}
+
+static void gic_ack_irq(struct irq_data *d)
+{
+	unsigned int irq = GIC_HWIRQ_TO_SHARED(d->hwirq);
+
+	gic_write(GIC_REG(SHARED, GIC_SH_WEDGE), GIC_SH_WEDGE_CLR(irq));
+}
+
+static int gic_set_type(struct irq_data *d, unsigned int type)
+{
+	unsigned int irq = GIC_HWIRQ_TO_SHARED(d->hwirq);
+	unsigned long flags;
+	bool is_edge;
+
+	spin_lock_irqsave(&gic_lock, flags);
+	switch (type & IRQ_TYPE_SENSE_MASK) {
+	case IRQ_TYPE_EDGE_FALLING:
+		gic_set_polarity(irq, GIC_POL_NEG);
+		gic_set_trigger(irq, GIC_TRIG_EDGE);
+		gic_set_dual_edge(irq, GIC_TRIG_DUAL_DISABLE);
+		is_edge = true;
+		break;
+	case IRQ_TYPE_EDGE_RISING:
+		gic_set_polarity(irq, GIC_POL_POS);
+		gic_set_trigger(irq, GIC_TRIG_EDGE);
+		gic_set_dual_edge(irq, GIC_TRIG_DUAL_DISABLE);
+		is_edge = true;
+		break;
+	case IRQ_TYPE_EDGE_BOTH:
+		/* polarity is irrelevant in this case */
+		gic_set_trigger(irq, GIC_TRIG_EDGE);
+		gic_set_dual_edge(irq, GIC_TRIG_DUAL_ENABLE);
+		is_edge = true;
+		break;
+	case IRQ_TYPE_LEVEL_LOW:
+		gic_set_polarity(irq, GIC_POL_NEG);
+		gic_set_trigger(irq, GIC_TRIG_LEVEL);
+		gic_set_dual_edge(irq, GIC_TRIG_DUAL_DISABLE);
+		is_edge = false;
+		break;
+	case IRQ_TYPE_LEVEL_HIGH:
+	default:
+		gic_set_polarity(irq, GIC_POL_POS);
+		gic_set_trigger(irq, GIC_TRIG_LEVEL);
+		gic_set_dual_edge(irq, GIC_TRIG_DUAL_DISABLE);
+		is_edge = false;
+		break;
+	}
+
+	if (is_edge) {
+		__irq_set_chip_handler_name_locked(d->irq,
+						   &gic_edge_irq_controller,
+						   handle_edge_irq, NULL);
+	} else {
+		__irq_set_chip_handler_name_locked(d->irq,
+						   &gic_level_irq_controller,
+						   handle_level_irq, NULL);
+	}
+	spin_unlock_irqrestore(&gic_lock, flags);
+
+	return 0;
+}
+
+#ifdef CONFIG_SMP
+static int gic_set_affinity(struct irq_data *d, const struct cpumask *cpumask,
+			    bool force)
+{
+	unsigned int irq = GIC_HWIRQ_TO_SHARED(d->hwirq);
+	cpumask_t	tmp = CPU_MASK_NONE;
+	unsigned long	flags;
+	int		i;
+
+	cpumask_and(&tmp, cpumask, cpu_online_mask);
+	if (cpus_empty(tmp))
+		return -EINVAL;
+
+	/* Assumption : cpumask refers to a single CPU */
+	spin_lock_irqsave(&gic_lock, flags);
+
+	/* Re-route this IRQ */
+	gic_map_to_vpe(irq, first_cpu(tmp));
+
+	/* Update the pcpu_masks */
+	for (i = 0; i < NR_CPUS; i++)
+		clear_bit(irq, pcpu_masks[i].pcpu_mask);
+	set_bit(irq, pcpu_masks[first_cpu(tmp)].pcpu_mask);
+
+	cpumask_copy(d->affinity, cpumask);
+	spin_unlock_irqrestore(&gic_lock, flags);
+
+	return IRQ_SET_MASK_OK_NOCOPY;
+}
+#endif
+
+static struct irq_chip gic_level_irq_controller = {
+	.name			=	"MIPS GIC",
+	.irq_mask		=	gic_mask_irq,
+	.irq_unmask		=	gic_unmask_irq,
+	.irq_set_type		=	gic_set_type,
+#ifdef CONFIG_SMP
+	.irq_set_affinity	=	gic_set_affinity,
+#endif
+};
+
+static struct irq_chip gic_edge_irq_controller = {
+	.name			=	"MIPS GIC",
+	.irq_ack		=	gic_ack_irq,
+	.irq_mask		=	gic_mask_irq,
+	.irq_unmask		=	gic_unmask_irq,
+	.irq_set_type		=	gic_set_type,
+#ifdef CONFIG_SMP
+	.irq_set_affinity	=	gic_set_affinity,
+#endif
+};
+
+static unsigned int gic_get_local_int(void)
+{
+	unsigned long pending, masked;
+
+	pending = gic_read(GIC_REG(VPE_LOCAL, GIC_VPE_PEND));
+	masked = gic_read(GIC_REG(VPE_LOCAL, GIC_VPE_MASK));
+
+	bitmap_and(&pending, &pending, &masked, GIC_NUM_LOCAL_INTRS);
+
+	return find_first_bit(&pending, GIC_NUM_LOCAL_INTRS);
+}
+
+static void gic_mask_local_irq(struct irq_data *d)
+{
+	int intr = GIC_HWIRQ_TO_LOCAL(d->hwirq);
+
+	gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_RMASK), 1 << intr);
+}
+
+static void gic_unmask_local_irq(struct irq_data *d)
+{
+	int intr = GIC_HWIRQ_TO_LOCAL(d->hwirq);
+
+	gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_SMASK), 1 << intr);
+}
+
+static struct irq_chip gic_local_irq_controller = {
+	.name			=	"MIPS GIC Local",
+	.irq_mask		=	gic_mask_local_irq,
+	.irq_unmask		=	gic_unmask_local_irq,
+};
+
+static void gic_mask_local_irq_all_vpes(struct irq_data *d)
+{
+	int intr = GIC_HWIRQ_TO_LOCAL(d->hwirq);
+	int i;
+	unsigned long flags;
+
+	spin_lock_irqsave(&gic_lock, flags);
+	for (i = 0; i < gic_vpes; i++) {
+		gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_OTHER_ADDR), i);
+		gic_write(GIC_REG(VPE_OTHER, GIC_VPE_RMASK), 1 << intr);
+	}
+	spin_unlock_irqrestore(&gic_lock, flags);
+}
+
+static void gic_unmask_local_irq_all_vpes(struct irq_data *d)
+{
+	int intr = GIC_HWIRQ_TO_LOCAL(d->hwirq);
+	int i;
+	unsigned long flags;
+
+	spin_lock_irqsave(&gic_lock, flags);
+	for (i = 0; i < gic_vpes; i++) {
+		gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_OTHER_ADDR), i);
+		gic_write(GIC_REG(VPE_OTHER, GIC_VPE_SMASK), 1 << intr);
+	}
+	spin_unlock_irqrestore(&gic_lock, flags);
+}
+
+static struct irq_chip gic_all_vpes_local_irq_controller = {
+	.name			=	"MIPS GIC Local",
+	.irq_mask		=	gic_mask_local_irq_all_vpes,
+	.irq_unmask		=	gic_unmask_local_irq_all_vpes,
+};
+
+static void __gic_irq_dispatch(void)
+{
+	unsigned int intr, virq;
+
+	while ((intr = gic_get_local_int()) != GIC_NUM_LOCAL_INTRS) {
+		virq = irq_linear_revmap(gic_irq_domain,
+					 GIC_LOCAL_TO_HWIRQ(intr));
+		do_IRQ(virq);
+	}
+
+	while ((intr = gic_get_int()) != gic_shared_intrs) {
+		virq = irq_linear_revmap(gic_irq_domain,
+					 GIC_SHARED_TO_HWIRQ(intr));
+		do_IRQ(virq);
+	}
+}
+
+static void gic_irq_dispatch(unsigned int irq, struct irq_desc *desc)
+{
+	__gic_irq_dispatch();
+}
+
+#ifdef CONFIG_MIPS_GIC_IPI
+static int gic_resched_int_base;
+static int gic_call_int_base;
+
+unsigned int plat_ipi_resched_int_xlate(unsigned int cpu)
+{
+	return gic_resched_int_base + cpu;
+}
+
+unsigned int plat_ipi_call_int_xlate(unsigned int cpu)
+{
+	return gic_call_int_base + cpu;
+}
+
+static irqreturn_t ipi_resched_interrupt(int irq, void *dev_id)
+{
+	scheduler_ipi();
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t ipi_call_interrupt(int irq, void *dev_id)
+{
+	smp_call_function_interrupt();
+
+	return IRQ_HANDLED;
+}
+
+static struct irqaction irq_resched = {
+	.handler	= ipi_resched_interrupt,
+	.flags		= IRQF_PERCPU,
+	.name		= "IPI resched"
+};
+
+static struct irqaction irq_call = {
+	.handler	= ipi_call_interrupt,
+	.flags		= IRQF_PERCPU,
+	.name		= "IPI call"
+};
+
+static __init void gic_ipi_init_one(unsigned int intr, int cpu,
+				    struct irqaction *action)
+{
+	int virq = irq_create_mapping(gic_irq_domain,
+				      GIC_SHARED_TO_HWIRQ(intr));
+	int i;
+
+	gic_map_to_vpe(intr, cpu);
+	for (i = 0; i < NR_CPUS; i++)
+		clear_bit(intr, pcpu_masks[i].pcpu_mask);
+	set_bit(intr, pcpu_masks[cpu].pcpu_mask);
+
+	irq_set_irq_type(virq, IRQ_TYPE_EDGE_RISING);
+
+	irq_set_handler(virq, handle_percpu_irq);
+	setup_irq(virq, action);
+}
+
+static __init void gic_ipi_init(void)
+{
+	int i;
+
+	/* Use last 2 * NR_CPUS interrupts as IPIs */
+	gic_resched_int_base = gic_shared_intrs - nr_cpu_ids;
+	gic_call_int_base = gic_resched_int_base - nr_cpu_ids;
+
+	for (i = 0; i < nr_cpu_ids; i++) {
+		gic_ipi_init_one(gic_call_int_base + i, i, &irq_call);
+		gic_ipi_init_one(gic_resched_int_base + i, i, &irq_resched);
+	}
+}
+#else
+static inline void gic_ipi_init(void)
+{
+}
+#endif
+
+static void __init gic_basic_init(void)
+{
+	unsigned int i;
+
+	board_bind_eic_interrupt = &gic_bind_eic_interrupt;
+
+	/* Setup defaults */
+	for (i = 0; i < gic_shared_intrs; i++) {
+		gic_set_polarity(i, GIC_POL_POS);
+		gic_set_trigger(i, GIC_TRIG_LEVEL);
+		gic_reset_mask(i);
+	}
+
+	for (i = 0; i < gic_vpes; i++) {
+		unsigned int j;
+
+		gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_OTHER_ADDR), i);
+		for (j = 0; j < GIC_NUM_LOCAL_INTRS; j++) {
+			if (!gic_local_irq_is_routable(j))
+				continue;
+			gic_write(GIC_REG(VPE_OTHER, GIC_VPE_RMASK), 1 << j);
+		}
+	}
+}
+
+static int gic_local_irq_domain_map(struct irq_domain *d, unsigned int virq,
+				    irq_hw_number_t hw)
+{
+	int intr = GIC_HWIRQ_TO_LOCAL(hw);
+	int ret = 0;
+	int i;
+	unsigned long flags;
+
+	if (!gic_local_irq_is_routable(intr))
+		return -EPERM;
+
+	/*
+	 * HACK: These are all really percpu interrupts, but the rest
+	 * of the MIPS kernel code does not use the percpu IRQ API for
+	 * the CP0 timer and performance counter interrupts.
+	 */
+	if (intr != GIC_LOCAL_INT_TIMER && intr != GIC_LOCAL_INT_PERFCTR) {
+		irq_set_chip_and_handler(virq,
+					 &gic_local_irq_controller,
+					 handle_percpu_devid_irq);
+		irq_set_percpu_devid(virq);
+	} else {
+		irq_set_chip_and_handler(virq,
+					 &gic_all_vpes_local_irq_controller,
+					 handle_percpu_irq);
+	}
+
+	spin_lock_irqsave(&gic_lock, flags);
+	for (i = 0; i < gic_vpes; i++) {
+		u32 val = GIC_MAP_TO_PIN_MSK | gic_cpu_pin;
+
+		gic_write(GIC_REG(VPE_LOCAL, GIC_VPE_OTHER_ADDR), i);
+
+		switch (intr) {
+		case GIC_LOCAL_INT_WD:
+			gic_write(GIC_REG(VPE_OTHER, GIC_VPE_WD_MAP), val);
+			break;
+		case GIC_LOCAL_INT_COMPARE:
+			gic_write(GIC_REG(VPE_OTHER, GIC_VPE_COMPARE_MAP), val);
+			break;
+		case GIC_LOCAL_INT_TIMER:
+			gic_write(GIC_REG(VPE_OTHER, GIC_VPE_TIMER_MAP), val);
+			break;
+		case GIC_LOCAL_INT_PERFCTR:
+			gic_write(GIC_REG(VPE_OTHER, GIC_VPE_PERFCTR_MAP), val);
+			break;
+		case GIC_LOCAL_INT_SWINT0:
+			gic_write(GIC_REG(VPE_OTHER, GIC_VPE_SWINT0_MAP), val);
+			break;
+		case GIC_LOCAL_INT_SWINT1:
+			gic_write(GIC_REG(VPE_OTHER, GIC_VPE_SWINT1_MAP), val);
+			break;
+		case GIC_LOCAL_INT_FDC:
+			gic_write(GIC_REG(VPE_OTHER, GIC_VPE_FDC_MAP), val);
+			break;
+		default:
+			pr_err("Invalid local IRQ %d\n", intr);
+			ret = -EINVAL;
+			break;
+		}
+	}
+	spin_unlock_irqrestore(&gic_lock, flags);
+
+	return ret;
+}
+
+static int gic_shared_irq_domain_map(struct irq_domain *d, unsigned int virq,
+				     irq_hw_number_t hw)
+{
+	int intr = GIC_HWIRQ_TO_SHARED(hw);
+	unsigned long flags;
+
+	irq_set_chip_and_handler(virq, &gic_level_irq_controller,
+				 handle_level_irq);
+
+	spin_lock_irqsave(&gic_lock, flags);
+	gic_map_to_pin(intr, gic_cpu_pin);
+	/* Map to VPE 0 by default */
+	gic_map_to_vpe(intr, 0);
+	set_bit(intr, pcpu_masks[0].pcpu_mask);
+	spin_unlock_irqrestore(&gic_lock, flags);
+
+	return 0;
+}
+
+static int gic_irq_domain_map(struct irq_domain *d, unsigned int virq,
+			      irq_hw_number_t hw)
+{
+	if (GIC_HWIRQ_TO_LOCAL(hw) < GIC_NUM_LOCAL_INTRS)
+		return gic_local_irq_domain_map(d, virq, hw);
+	return gic_shared_irq_domain_map(d, virq, hw);
+}
+
+static int gic_irq_domain_xlate(struct irq_domain *d, struct device_node *ctrlr,
+				const u32 *intspec, unsigned int intsize,
+				irq_hw_number_t *out_hwirq,
+				unsigned int *out_type)
+{
+	if (intsize != 3)
+		return -EINVAL;
+
+	if (intspec[0] == GIC_SHARED)
+		*out_hwirq = GIC_SHARED_TO_HWIRQ(intspec[1]);
+	else if (intspec[0] == GIC_LOCAL)
+		*out_hwirq = GIC_LOCAL_TO_HWIRQ(intspec[1]);
+	else
+		return -EINVAL;
+	*out_type = intspec[2] & IRQ_TYPE_SENSE_MASK;
+
+	return 0;
+}
+
+static struct irq_domain_ops gic_irq_domain_ops = {
+	.map = gic_irq_domain_map,
+	.xlate = gic_irq_domain_xlate,
+};
+
+static void __init __gic_init(unsigned long gic_base_addr,
+			      unsigned long gic_addrspace_size,
+			      unsigned int cpu_vec, unsigned int irqbase,
+			      struct device_node *node)
+{
+	unsigned int gicconfig;
+
+	gic_base = ioremap_nocache(gic_base_addr, gic_addrspace_size);
+
+	gicconfig = gic_read(GIC_REG(SHARED, GIC_SH_CONFIG));
+	gic_shared_intrs = (gicconfig & GIC_SH_CONFIG_NUMINTRS_MSK) >>
+		   GIC_SH_CONFIG_NUMINTRS_SHF;
+	gic_shared_intrs = ((gic_shared_intrs + 1) * 8);
+
+	gic_vpes = (gicconfig & GIC_SH_CONFIG_NUMVPES_MSK) >>
+		  GIC_SH_CONFIG_NUMVPES_SHF;
+	gic_vpes = gic_vpes + 1;
+
+	if (cpu_has_veic) {
+		/* Always use vector 1 in EIC mode */
+		gic_cpu_pin = 0;
+		set_vi_handler(gic_cpu_pin + GIC_PIN_TO_VEC_OFFSET,
+			       __gic_irq_dispatch);
+	} else {
+		gic_cpu_pin = cpu_vec - GIC_CPU_PIN_OFFSET;
+		irq_set_chained_handler(MIPS_CPU_IRQ_BASE + cpu_vec,
+					gic_irq_dispatch);
+	}
+
+	gic_irq_domain = irq_domain_add_simple(node, GIC_NUM_LOCAL_INTRS +
+					       gic_shared_intrs, irqbase,
+					       &gic_irq_domain_ops, NULL);
+	if (!gic_irq_domain)
+		panic("Failed to add GIC IRQ domain");
+
+	gic_basic_init();
+
+	gic_ipi_init();
+}
+
+void __init gic_init(unsigned long gic_base_addr,
+		     unsigned long gic_addrspace_size,
+		     unsigned int cpu_vec, unsigned int irqbase)
+{
+	__gic_init(gic_base_addr, gic_addrspace_size, cpu_vec, irqbase, NULL);
+}
+
+static int __init gic_of_init(struct device_node *node,
+			      struct device_node *parent)
+{
+	struct resource res;
+	unsigned int cpu_vec, i = 0, reserved = 0;
+	phys_addr_t gic_base;
+	size_t gic_len;
+
+	/* Find the first available CPU vector. */
+	while (!of_property_read_u32_index(node, "mti,reserved-cpu-vectors",
+					   i++, &cpu_vec))
+		reserved |= BIT(cpu_vec);
+	for (cpu_vec = 2; cpu_vec < 8; cpu_vec++) {
+		if (!(reserved & BIT(cpu_vec)))
+			break;
+	}
+	if (cpu_vec == 8) {
+		pr_err("No CPU vectors available for GIC\n");
+		return -ENODEV;
+	}
+
+	if (of_address_to_resource(node, 0, &res)) {
+		/*
+		 * Probe the CM for the GIC base address if not specified
+		 * in the device-tree.
+		 */
+		if (mips_cm_present()) {
+			gic_base = read_gcr_gic_base() &
+				~CM_GCR_GIC_BASE_GICEN_MSK;
+			gic_len = 0x20000;
+		} else {
+			pr_err("Failed to get GIC memory range\n");
+			return -ENODEV;
+		}
+	} else {
+		gic_base = res.start;
+		gic_len = resource_size(&res);
+	}
+
+	if (mips_cm_present())
+		write_gcr_gic_base(gic_base | CM_GCR_GIC_BASE_GICEN_MSK);
+	gic_present = true;
+
+	__gic_init(gic_base, gic_len, cpu_vec, 0, node);
+
+	return 0;
+}
+IRQCHIP_DECLARE(mips_gic, "mti,gic", gic_of_init);
diff --git a/drivers/irqchip/irq-mmp.c b/drivers/irqchip/irq-mmp.c
index 1c3e2c9b46ba..c0da57bdb89d 100644
--- a/drivers/irqchip/irq-mmp.c
+++ b/drivers/irqchip/irq-mmp.c
@@ -196,26 +196,24 @@ static struct mmp_intc_conf mmp2_conf = {
 
 static void __exception_irq_entry mmp_handle_irq(struct pt_regs *regs)
 {
-	int irq, hwirq;
+	int hwirq;
 
 	hwirq = readl_relaxed(mmp_icu_base + PJ1_INT_SEL);
 	if (!(hwirq & SEL_INT_PENDING))
 		return;
 	hwirq &= SEL_INT_NUM_MASK;
-	irq = irq_find_mapping(icu_data[0].domain, hwirq);
-	handle_IRQ(irq, regs);
+	handle_domain_irq(icu_data[0].domain, hwirq, regs);
 }
 
 static void __exception_irq_entry mmp2_handle_irq(struct pt_regs *regs)
 {
-	int irq, hwirq;
+	int hwirq;
 
 	hwirq = readl_relaxed(mmp_icu_base + PJ4_INT_SEL);
 	if (!(hwirq & SEL_INT_PENDING))
 		return;
 	hwirq &= SEL_INT_NUM_MASK;
-	irq = irq_find_mapping(icu_data[0].domain, hwirq);
-	handle_IRQ(irq, regs);
+	handle_domain_irq(icu_data[0].domain, hwirq, regs);
 }
 
 /* MMP (ARMv5) */
diff --git a/drivers/irqchip/irq-mtk-sysirq.c b/drivers/irqchip/irq-mtk-sysirq.c
new file mode 100644
index 000000000000..0b0d2c00a2df
--- /dev/null
+++ b/drivers/irqchip/irq-mtk-sysirq.c
@@ -0,0 +1,163 @@
+/*
+ * Copyright (c) 2014 MediaTek Inc.
+ * Author: Joe.C <yingjoe.chen@mediatek.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/irq.h>
+#include <linux/irqdomain.h>
+#include <linux/of.h>
+#include <linux/of_irq.h>
+#include <linux/of_address.h>
+#include <linux/io.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+
+#include "irqchip.h"
+
+#define MT6577_SYS_INTPOL_NUM	(224)
+
+struct mtk_sysirq_chip_data {
+	spinlock_t lock;
+	void __iomem *intpol_base;
+};
+
+static int mtk_sysirq_set_type(struct irq_data *data, unsigned int type)
+{
+	irq_hw_number_t hwirq = data->hwirq;
+	struct mtk_sysirq_chip_data *chip_data = data->chip_data;
+	u32 offset, reg_index, value;
+	unsigned long flags;
+	int ret;
+
+	offset = hwirq & 0x1f;
+	reg_index = hwirq >> 5;
+
+	spin_lock_irqsave(&chip_data->lock, flags);
+	value = readl_relaxed(chip_data->intpol_base + reg_index * 4);
+	if (type == IRQ_TYPE_LEVEL_LOW || type == IRQ_TYPE_EDGE_FALLING) {
+		if (type == IRQ_TYPE_LEVEL_LOW)
+			type = IRQ_TYPE_LEVEL_HIGH;
+		else
+			type = IRQ_TYPE_EDGE_RISING;
+		value |= (1 << offset);
+	} else {
+		value &= ~(1 << offset);
+	}
+	writel(value, chip_data->intpol_base + reg_index * 4);
+
+	data = data->parent_data;
+	ret = data->chip->irq_set_type(data, type);
+	spin_unlock_irqrestore(&chip_data->lock, flags);
+	return ret;
+}
+
+static struct irq_chip mtk_sysirq_chip = {
+	.name			= "MT_SYSIRQ",
+	.irq_mask		= irq_chip_mask_parent,
+	.irq_unmask		= irq_chip_unmask_parent,
+	.irq_eoi		= irq_chip_eoi_parent,
+	.irq_set_type		= mtk_sysirq_set_type,
+	.irq_retrigger		= irq_chip_retrigger_hierarchy,
+	.irq_set_affinity	= irq_chip_set_affinity_parent,
+};
+
+static int mtk_sysirq_domain_xlate(struct irq_domain *d,
+				   struct device_node *controller,
+				   const u32 *intspec, unsigned int intsize,
+				   unsigned long *out_hwirq,
+				   unsigned int *out_type)
+{
+	if (intsize != 3)
+		return -EINVAL;
+
+	/* sysirq doesn't support PPI */
+	if (intspec[0])
+		return -EINVAL;
+
+	*out_hwirq = intspec[1];
+	*out_type = intspec[2] & IRQ_TYPE_SENSE_MASK;
+	return 0;
+}
+
+static int mtk_sysirq_domain_alloc(struct irq_domain *domain, unsigned int virq,
+				   unsigned int nr_irqs, void *arg)
+{
+	int i;
+	irq_hw_number_t hwirq;
+	struct of_phandle_args *irq_data = arg;
+	struct of_phandle_args gic_data = *irq_data;
+
+	if (irq_data->args_count != 3)
+		return -EINVAL;
+
+	/* sysirq doesn't support PPI */
+	if (irq_data->args[0])
+		return -EINVAL;
+
+	hwirq = irq_data->args[1];
+	for (i = 0; i < nr_irqs; i++)
+		irq_domain_set_hwirq_and_chip(domain, virq + i, hwirq + i,
+					      &mtk_sysirq_chip,
+					      domain->host_data);
+
+	gic_data.np = domain->parent->of_node;
+	return irq_domain_alloc_irqs_parent(domain, virq, nr_irqs, &gic_data);
+}
+
+static struct irq_domain_ops sysirq_domain_ops = {
+	.xlate = mtk_sysirq_domain_xlate,
+	.alloc = mtk_sysirq_domain_alloc,
+	.free = irq_domain_free_irqs_common,
+};
+
+static int __init mtk_sysirq_of_init(struct device_node *node,
+				     struct device_node *parent)
+{
+	struct irq_domain *domain, *domain_parent;
+	struct mtk_sysirq_chip_data *chip_data;
+	int ret = 0;
+
+	domain_parent = irq_find_host(parent);
+	if (!domain_parent) {
+		pr_err("mtk_sysirq: interrupt-parent not found\n");
+		return -EINVAL;
+	}
+
+	chip_data = kzalloc(sizeof(*chip_data), GFP_KERNEL);
+	if (!chip_data)
+		return -ENOMEM;
+
+	chip_data->intpol_base = of_io_request_and_map(node, 0, "intpol");
+	if (IS_ERR(chip_data->intpol_base)) {
+		pr_err("mtk_sysirq: unable to map sysirq register\n");
+		ret = PTR_ERR(chip_data->intpol_base);
+		goto out_free;
+	}
+
+	domain = irq_domain_add_hierarchy(domain_parent, 0,
+					  MT6577_SYS_INTPOL_NUM, node,
+					  &sysirq_domain_ops, chip_data);
+	if (!domain) {
+		ret = -ENOMEM;
+		goto out_unmap;
+	}
+	spin_lock_init(&chip_data->lock);
+
+	return 0;
+
+out_unmap:
+	iounmap(chip_data->intpol_base);
+out_free:
+	kfree(chip_data);
+	return ret;
+}
+IRQCHIP_DECLARE(mtk_sysirq, "mediatek,mt6577-sysirq", mtk_sysirq_of_init);
diff --git a/drivers/irqchip/irq-mxs.c b/drivers/irqchip/irq-mxs.c
index 4044ff287663..e4acf1e3f8e3 100644
--- a/drivers/irqchip/irq-mxs.c
+++ b/drivers/irqchip/irq-mxs.c
@@ -78,8 +78,7 @@ asmlinkage void __exception_irq_entry icoll_handle_irq(struct pt_regs *regs)
 
 	irqnr = __raw_readl(icoll_base + HW_ICOLL_STAT_OFFSET);
 	__raw_writel(irqnr, icoll_base + HW_ICOLL_VECTOR);
-	irqnr = irq_find_mapping(icoll_domain, irqnr);
-	handle_IRQ(irqnr, regs);
+	handle_domain_irq(icoll_domain, irqnr, regs);
 }
 
 static int icoll_irq_domain_map(struct irq_domain *d, unsigned int virq,
diff --git a/drivers/irqchip/irq-omap-intc.c b/drivers/irqchip/irq-omap-intc.c
new file mode 100644
index 000000000000..c03f140acbae
--- /dev/null
+++ b/drivers/irqchip/irq-omap-intc.c
@@ -0,0 +1,418 @@
+/*
+ * linux/arch/arm/mach-omap2/irq.c
+ *
+ * Interrupt handler for OMAP2 boards.
+ *
+ * Copyright (C) 2005 Nokia Corporation
+ * Author: Paul Mundt <paul.mundt@nokia.com>
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License. See the file "COPYING" in the main directory of this archive
+ * for more details.
+ */
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+
+#include <asm/exception.h>
+#include <linux/irqdomain.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+
+#include "irqchip.h"
+
+/* Define these here for now until we drop all board-files */
+#define OMAP24XX_IC_BASE	0x480fe000
+#define OMAP34XX_IC_BASE	0x48200000
+
+/* selected INTC register offsets */
+
+#define INTC_REVISION		0x0000
+#define INTC_SYSCONFIG		0x0010
+#define INTC_SYSSTATUS		0x0014
+#define INTC_SIR		0x0040
+#define INTC_CONTROL		0x0048
+#define INTC_PROTECTION		0x004C
+#define INTC_IDLE		0x0050
+#define INTC_THRESHOLD		0x0068
+#define INTC_MIR0		0x0084
+#define INTC_MIR_CLEAR0		0x0088
+#define INTC_MIR_SET0		0x008c
+#define INTC_PENDING_IRQ0	0x0098
+#define INTC_PENDING_IRQ1	0x00b8
+#define INTC_PENDING_IRQ2	0x00d8
+#define INTC_PENDING_IRQ3	0x00f8
+#define INTC_ILR0		0x0100
+
+#define ACTIVEIRQ_MASK		0x7f	/* omap2/3 active interrupt bits */
+#define INTCPS_NR_ILR_REGS	128
+#define INTCPS_NR_MIR_REGS	4
+
+#define INTC_IDLE_FUNCIDLE	(1 << 0)
+#define INTC_IDLE_TURBO		(1 << 1)
+
+#define INTC_PROTECTION_ENABLE	(1 << 0)
+
+struct omap_intc_regs {
+	u32 sysconfig;
+	u32 protection;
+	u32 idle;
+	u32 threshold;
+	u32 ilr[INTCPS_NR_ILR_REGS];
+	u32 mir[INTCPS_NR_MIR_REGS];
+};
+static struct omap_intc_regs intc_context;
+
+static struct irq_domain *domain;
+static void __iomem *omap_irq_base;
+static int omap_nr_pending = 3;
+static int omap_nr_irqs = 96;
+
+static void intc_writel(u32 reg, u32 val)
+{
+	writel_relaxed(val, omap_irq_base + reg);
+}
+
+static u32 intc_readl(u32 reg)
+{
+	return readl_relaxed(omap_irq_base + reg);
+}
+
+void omap_intc_save_context(void)
+{
+	int i;
+
+	intc_context.sysconfig =
+		intc_readl(INTC_SYSCONFIG);
+	intc_context.protection =
+		intc_readl(INTC_PROTECTION);
+	intc_context.idle =
+		intc_readl(INTC_IDLE);
+	intc_context.threshold =
+		intc_readl(INTC_THRESHOLD);
+
+	for (i = 0; i < omap_nr_irqs; i++)
+		intc_context.ilr[i] =
+			intc_readl((INTC_ILR0 + 0x4 * i));
+	for (i = 0; i < INTCPS_NR_MIR_REGS; i++)
+		intc_context.mir[i] =
+			intc_readl(INTC_MIR0 + (0x20 * i));
+}
+
+void omap_intc_restore_context(void)
+{
+	int i;
+
+	intc_writel(INTC_SYSCONFIG, intc_context.sysconfig);
+	intc_writel(INTC_PROTECTION, intc_context.protection);
+	intc_writel(INTC_IDLE, intc_context.idle);
+	intc_writel(INTC_THRESHOLD, intc_context.threshold);
+
+	for (i = 0; i < omap_nr_irqs; i++)
+		intc_writel(INTC_ILR0 + 0x4 * i,
+				intc_context.ilr[i]);
+
+	for (i = 0; i < INTCPS_NR_MIR_REGS; i++)
+		intc_writel(INTC_MIR0 + 0x20 * i,
+			intc_context.mir[i]);
+	/* MIRs are saved and restore with other PRCM registers */
+}
+
+void omap3_intc_prepare_idle(void)
+{
+	/*
+	 * Disable autoidle as it can stall interrupt controller,
+	 * cf. errata ID i540 for 3430 (all revisions up to 3.1.x)
+	 */
+	intc_writel(INTC_SYSCONFIG, 0);
+	intc_writel(INTC_IDLE, INTC_IDLE_TURBO);
+}
+
+void omap3_intc_resume_idle(void)
+{
+	/* Re-enable autoidle */
+	intc_writel(INTC_SYSCONFIG, 1);
+	intc_writel(INTC_IDLE, 0);
+}
+
+/* XXX: FIQ and additional INTC support (only MPU at the moment) */
+static void omap_ack_irq(struct irq_data *d)
+{
+	intc_writel(INTC_CONTROL, 0x1);
+}
+
+static void omap_mask_ack_irq(struct irq_data *d)
+{
+	irq_gc_mask_disable_reg(d);
+	omap_ack_irq(d);
+}
+
+static void __init omap_irq_soft_reset(void)
+{
+	unsigned long tmp;
+
+	tmp = intc_readl(INTC_REVISION) & 0xff;
+
+	pr_info("IRQ: Found an INTC at 0x%p (revision %ld.%ld) with %d interrupts\n",
+		omap_irq_base, tmp >> 4, tmp & 0xf, omap_nr_irqs);
+
+	tmp = intc_readl(INTC_SYSCONFIG);
+	tmp |= 1 << 1;	/* soft reset */
+	intc_writel(INTC_SYSCONFIG, tmp);
+
+	while (!(intc_readl(INTC_SYSSTATUS) & 0x1))
+		/* Wait for reset to complete */;
+
+	/* Enable autoidle */
+	intc_writel(INTC_SYSCONFIG, 1 << 0);
+}
+
+int omap_irq_pending(void)
+{
+	int i;
+
+	for (i = 0; i < omap_nr_pending; i++)
+		if (intc_readl(INTC_PENDING_IRQ0 + (0x20 * i)))
+			return 1;
+	return 0;
+}
+
+void omap3_intc_suspend(void)
+{
+	/* A pending interrupt would prevent OMAP from entering suspend */
+	omap_ack_irq(NULL);
+}
+
+static int __init omap_alloc_gc_of(struct irq_domain *d, void __iomem *base)
+{
+	int ret;
+	int i;
+
+	ret = irq_alloc_domain_generic_chips(d, 32, 1, "INTC",
+			handle_level_irq, IRQ_NOREQUEST | IRQ_NOPROBE,
+			IRQ_LEVEL, 0);
+	if (ret) {
+		pr_warn("Failed to allocate irq chips\n");
+		return ret;
+	}
+
+	for (i = 0; i < omap_nr_pending; i++) {
+		struct irq_chip_generic *gc;
+		struct irq_chip_type *ct;
+
+		gc = irq_get_domain_generic_chip(d, 32 * i);
+		gc->reg_base = base;
+		ct = gc->chip_types;
+
+		ct->type = IRQ_TYPE_LEVEL_MASK;
+		ct->handler = handle_level_irq;
+
+		ct->chip.irq_ack = omap_mask_ack_irq;
+		ct->chip.irq_mask = irq_gc_mask_disable_reg;
+		ct->chip.irq_unmask = irq_gc_unmask_enable_reg;
+
+		ct->chip.flags |= IRQCHIP_SKIP_SET_WAKE;
+
+		ct->regs.enable = INTC_MIR_CLEAR0 + 32 * i;
+		ct->regs.disable = INTC_MIR_SET0 + 32 * i;
+	}
+
+	return 0;
+}
+
+static void __init omap_alloc_gc_legacy(void __iomem *base,
+		unsigned int irq_start, unsigned int num)
+{
+	struct irq_chip_generic *gc;
+	struct irq_chip_type *ct;
+
+	gc = irq_alloc_generic_chip("INTC", 1, irq_start, base,
+			handle_level_irq);
+	ct = gc->chip_types;
+	ct->chip.irq_ack = omap_mask_ack_irq;
+	ct->chip.irq_mask = irq_gc_mask_disable_reg;
+	ct->chip.irq_unmask = irq_gc_unmask_enable_reg;
+	ct->chip.flags |= IRQCHIP_SKIP_SET_WAKE;
+
+	ct->regs.enable = INTC_MIR_CLEAR0;
+	ct->regs.disable = INTC_MIR_SET0;
+	irq_setup_generic_chip(gc, IRQ_MSK(num), IRQ_GC_INIT_MASK_CACHE,
+			IRQ_NOREQUEST | IRQ_NOPROBE, 0);
+}
+
+static int __init omap_init_irq_of(struct device_node *node)
+{
+	int ret;
+
+	omap_irq_base = of_iomap(node, 0);
+	if (WARN_ON(!omap_irq_base))
+		return -ENOMEM;
+
+	domain = irq_domain_add_linear(node, omap_nr_irqs,
+			&irq_generic_chip_ops, NULL);
+
+	omap_irq_soft_reset();
+
+	ret = omap_alloc_gc_of(domain, omap_irq_base);
+	if (ret < 0)
+		irq_domain_remove(domain);
+
+	return ret;
+}
+
+static int __init omap_init_irq_legacy(u32 base, struct device_node *node)
+{
+	int j, irq_base;
+
+	omap_irq_base = ioremap(base, SZ_4K);
+	if (WARN_ON(!omap_irq_base))
+		return -ENOMEM;
+
+	irq_base = irq_alloc_descs(-1, 0, omap_nr_irqs, 0);
+	if (irq_base < 0) {
+		pr_warn("Couldn't allocate IRQ numbers\n");
+		irq_base = 0;
+	}
+
+	domain = irq_domain_add_legacy(node, omap_nr_irqs, irq_base, 0,
+			&irq_domain_simple_ops, NULL);
+
+	omap_irq_soft_reset();
+
+	for (j = 0; j < omap_nr_irqs; j += 32)
+		omap_alloc_gc_legacy(omap_irq_base + j, j + irq_base, 32);
+
+	return 0;
+}
+
+static void __init omap_irq_enable_protection(void)
+{
+	u32 reg;
+
+	reg = intc_readl(INTC_PROTECTION);
+	reg |= INTC_PROTECTION_ENABLE;
+	intc_writel(INTC_PROTECTION, reg);
+}
+
+static int __init omap_init_irq(u32 base, struct device_node *node)
+{
+	int ret;
+
+	/*
+	 * FIXME legacy OMAP DMA driver sitting under arch/arm/plat-omap/dma.c
+	 * depends is still not ready for linear IRQ domains; because of that
+	 * we need to temporarily "blacklist" OMAP2 and OMAP3 devices from using
+	 * linear IRQ Domain until that driver is finally fixed.
+	 */
+	if (of_device_is_compatible(node, "ti,omap2-intc") ||
+			of_device_is_compatible(node, "ti,omap3-intc")) {
+		struct resource res;
+
+		if (of_address_to_resource(node, 0, &res))
+			return -ENOMEM;
+
+		base = res.start;
+		ret = omap_init_irq_legacy(base, node);
+	} else if (node) {
+		ret = omap_init_irq_of(node);
+	} else {
+		ret = omap_init_irq_legacy(base, NULL);
+	}
+
+	if (ret == 0)
+		omap_irq_enable_protection();
+
+	return ret;
+}
+
+static asmlinkage void __exception_irq_entry
+omap_intc_handle_irq(struct pt_regs *regs)
+{
+	u32 irqnr = 0;
+	int handled_irq = 0;
+	int i;
+
+	do {
+		for (i = 0; i < omap_nr_pending; i++) {
+			irqnr = intc_readl(INTC_PENDING_IRQ0 + (0x20 * i));
+			if (irqnr)
+				goto out;
+		}
+
+out:
+		if (!irqnr)
+			break;
+
+		irqnr = intc_readl(INTC_SIR);
+		irqnr &= ACTIVEIRQ_MASK;
+
+		if (irqnr) {
+			handle_domain_irq(domain, irqnr, regs);
+			handled_irq = 1;
+		}
+	} while (irqnr);
+
+	/*
+	 * If an irq is masked or deasserted while active, we will
+	 * keep ending up here with no irq handled. So remove it from
+	 * the INTC with an ack.
+	 */
+	if (!handled_irq)
+		omap_ack_irq(NULL);
+}
+
+void __init omap2_init_irq(void)
+{
+	omap_nr_irqs = 96;
+	omap_nr_pending = 3;
+	omap_init_irq(OMAP24XX_IC_BASE, NULL);
+	set_handle_irq(omap_intc_handle_irq);
+}
+
+void __init omap3_init_irq(void)
+{
+	omap_nr_irqs = 96;
+	omap_nr_pending = 3;
+	omap_init_irq(OMAP34XX_IC_BASE, NULL);
+	set_handle_irq(omap_intc_handle_irq);
+}
+
+void __init ti81xx_init_irq(void)
+{
+	omap_nr_irqs = 96;
+	omap_nr_pending = 4;
+	omap_init_irq(OMAP34XX_IC_BASE, NULL);
+	set_handle_irq(omap_intc_handle_irq);
+}
+
+static int __init intc_of_init(struct device_node *node,
+			     struct device_node *parent)
+{
+	int ret;
+
+	omap_nr_pending = 3;
+	omap_nr_irqs = 96;
+
+	if (WARN_ON(!node))
+		return -ENODEV;
+
+	if (of_device_is_compatible(node, "ti,am33xx-intc")) {
+		omap_nr_irqs = 128;
+		omap_nr_pending = 4;
+	}
+
+	ret = omap_init_irq(-1, of_node_get(node));
+	if (ret < 0)
+		return ret;
+
+	set_handle_irq(omap_intc_handle_irq);
+
+	return 0;
+}
+
+IRQCHIP_DECLARE(omap2_intc, "ti,omap2-intc", intc_of_init);
+IRQCHIP_DECLARE(omap3_intc, "ti,omap3-intc", intc_of_init);
+IRQCHIP_DECLARE(am33xx_intc, "ti,am33xx-intc", intc_of_init);
diff --git a/drivers/irqchip/irq-or1k-pic.c b/drivers/irqchip/irq-or1k-pic.c
index 17ff033d9925..e93d079fe069 100644
--- a/drivers/irqchip/irq-or1k-pic.c
+++ b/drivers/irqchip/irq-or1k-pic.c
@@ -113,7 +113,7 @@ static inline int pic_get_irq(int first)
 	else
 		hwirq = hwirq + first - 1;
 
-	return irq_find_mapping(root_domain, hwirq);
+	return hwirq;
 }
 
 static void or1k_pic_handle_irq(struct pt_regs *regs)
@@ -121,7 +121,7 @@ static void or1k_pic_handle_irq(struct pt_regs *regs)
 	int irq = -1;
 
 	while ((irq = pic_get_irq(irq + 1)) != NO_IRQ)
-		handle_IRQ(irq, regs);
+		handle_domain_irq(root_domain, irq, regs);
 }
 
 static int or1k_map(struct irq_domain *d, unsigned int irq, irq_hw_number_t hw)
diff --git a/drivers/irqchip/irq-orion.c b/drivers/irqchip/irq-orion.c
index 34d18b48bb78..ad0c0f6f1d65 100644
--- a/drivers/irqchip/irq-orion.c
+++ b/drivers/irqchip/irq-orion.c
@@ -43,9 +43,8 @@ __exception_irq_entry orion_handle_irq(struct pt_regs *regs)
 			gc->mask_cache;
 		while (stat) {
 			u32 hwirq = __fls(stat);
-			u32 irq = irq_find_mapping(orion_irq_domain,
-						   gc->irq_base + hwirq);
-			handle_IRQ(irq, regs);
+			handle_domain_irq(orion_irq_domain,
+					  gc->irq_base + hwirq, regs);
 			stat &= ~(1 << hwirq);
 		}
 	}
diff --git a/drivers/irqchip/irq-renesas-intc-irqpin.c b/drivers/irqchip/irq-renesas-intc-irqpin.c
index 3ee78f02e5d7..078cac5e2d08 100644
--- a/drivers/irqchip/irq-renesas-intc-irqpin.c
+++ b/drivers/irqchip/irq-renesas-intc-irqpin.c
@@ -17,6 +17,7 @@
  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  */
 
+#include <linux/clk.h>
 #include <linux/init.h>
 #include <linux/of.h>
 #include <linux/platform_device.h>
@@ -30,6 +31,7 @@
 #include <linux/slab.h>
 #include <linux/module.h>
 #include <linux/platform_data/irq-renesas-intc-irqpin.h>
+#include <linux/pm_runtime.h>
 
 #define INTC_IRQPIN_MAX 8 /* maximum 8 interrupts per driver instance */
 
@@ -75,6 +77,7 @@ struct intc_irqpin_priv {
 	struct platform_device *pdev;
 	struct irq_chip irq_chip;
 	struct irq_domain *irq_domain;
+	struct clk *clk;
 	bool shared_irqs;
 	u8 shared_irq_mask;
 };
@@ -270,6 +273,21 @@ static int intc_irqpin_irq_set_type(struct irq_data *d, unsigned int type)
 				     value ^ INTC_IRQ_SENSE_VALID);
 }
 
+static int intc_irqpin_irq_set_wake(struct irq_data *d, unsigned int on)
+{
+	struct intc_irqpin_priv *p = irq_data_get_irq_chip_data(d);
+
+	if (!p->clk)
+		return 0;
+
+	if (on)
+		clk_enable(p->clk);
+	else
+		clk_disable(p->clk);
+
+	return 0;
+}
+
 static irqreturn_t intc_irqpin_irq_handler(int irq, void *dev_id)
 {
 	struct intc_irqpin_irq *i = dev_id;
@@ -329,7 +347,8 @@ static struct irq_domain_ops intc_irqpin_irq_domain_ops = {
 
 static int intc_irqpin_probe(struct platform_device *pdev)
 {
-	struct renesas_intc_irqpin_config *pdata = pdev->dev.platform_data;
+	struct device *dev = &pdev->dev;
+	struct renesas_intc_irqpin_config *pdata = dev->platform_data;
 	struct intc_irqpin_priv *p;
 	struct intc_irqpin_iomem *i;
 	struct resource *io[INTC_IRQPIN_REG_NR];
@@ -337,25 +356,24 @@ static int intc_irqpin_probe(struct platform_device *pdev)
 	struct irq_chip *irq_chip;
 	void (*enable_fn)(struct irq_data *d);
 	void (*disable_fn)(struct irq_data *d);
-	const char *name = dev_name(&pdev->dev);
+	const char *name = dev_name(dev);
 	int ref_irq;
 	int ret;
 	int k;
 
-	p = devm_kzalloc(&pdev->dev, sizeof(*p), GFP_KERNEL);
+	p = devm_kzalloc(dev, sizeof(*p), GFP_KERNEL);
 	if (!p) {
-		dev_err(&pdev->dev, "failed to allocate driver data\n");
-		ret = -ENOMEM;
-		goto err0;
+		dev_err(dev, "failed to allocate driver data\n");
+		return -ENOMEM;
 	}
 
 	/* deal with driver instance configuration */
 	if (pdata) {
 		memcpy(&p->config, pdata, sizeof(*pdata));
 	} else {
-		of_property_read_u32(pdev->dev.of_node, "sense-bitfield-width",
+		of_property_read_u32(dev->of_node, "sense-bitfield-width",
 				     &p->config.sense_bitfield_width);
-		p->config.control_parent = of_property_read_bool(pdev->dev.of_node,
+		p->config.control_parent = of_property_read_bool(dev->of_node,
 								 "control-parent");
 	}
 	if (!p->config.sense_bitfield_width)
@@ -364,11 +382,20 @@ static int intc_irqpin_probe(struct platform_device *pdev)
 	p->pdev = pdev;
 	platform_set_drvdata(pdev, p);
 
+	p->clk = devm_clk_get(dev, NULL);
+	if (IS_ERR(p->clk)) {
+		dev_warn(dev, "unable to get clock\n");
+		p->clk = NULL;
+	}
+
+	pm_runtime_enable(dev);
+	pm_runtime_get_sync(dev);
+
 	/* get hold of manadatory IOMEM */
 	for (k = 0; k < INTC_IRQPIN_REG_NR; k++) {
 		io[k] = platform_get_resource(pdev, IORESOURCE_MEM, k);
 		if (!io[k]) {
-			dev_err(&pdev->dev, "not enough IOMEM resources\n");
+			dev_err(dev, "not enough IOMEM resources\n");
 			ret = -EINVAL;
 			goto err0;
 		}
@@ -386,7 +413,7 @@ static int intc_irqpin_probe(struct platform_device *pdev)
 
 	p->number_of_irqs = k;
 	if (p->number_of_irqs < 1) {
-		dev_err(&pdev->dev, "not enough IRQ resources\n");
+		dev_err(dev, "not enough IRQ resources\n");
 		ret = -EINVAL;
 		goto err0;
 	}
@@ -407,15 +434,15 @@ static int intc_irqpin_probe(struct platform_device *pdev)
 			i->write = intc_irqpin_write32;
 			break;
 		default:
-			dev_err(&pdev->dev, "IOMEM size mismatch\n");
+			dev_err(dev, "IOMEM size mismatch\n");
 			ret = -EINVAL;
 			goto err0;
 		}
 
-		i->iomem = devm_ioremap_nocache(&pdev->dev, io[k]->start,
+		i->iomem = devm_ioremap_nocache(dev, io[k]->start,
 						resource_size(io[k]));
 		if (!i->iomem) {
-			dev_err(&pdev->dev, "failed to remap IOMEM\n");
+			dev_err(dev, "failed to remap IOMEM\n");
 			ret = -ENXIO;
 			goto err0;
 		}
@@ -454,39 +481,36 @@ static int intc_irqpin_probe(struct platform_device *pdev)
 	irq_chip->name = name;
 	irq_chip->irq_mask = disable_fn;
 	irq_chip->irq_unmask = enable_fn;
-	irq_chip->irq_enable = enable_fn;
-	irq_chip->irq_disable = disable_fn;
 	irq_chip->irq_set_type = intc_irqpin_irq_set_type;
-	irq_chip->flags	= IRQCHIP_SKIP_SET_WAKE;
+	irq_chip->irq_set_wake = intc_irqpin_irq_set_wake;
+	irq_chip->flags	= IRQCHIP_MASK_ON_SUSPEND;
 
-	p->irq_domain = irq_domain_add_simple(pdev->dev.of_node,
+	p->irq_domain = irq_domain_add_simple(dev->of_node,
 					      p->number_of_irqs,
 					      p->config.irq_base,
 					      &intc_irqpin_irq_domain_ops, p);
 	if (!p->irq_domain) {
 		ret = -ENXIO;
-		dev_err(&pdev->dev, "cannot initialize irq domain\n");
+		dev_err(dev, "cannot initialize irq domain\n");
 		goto err0;
 	}
 
 	if (p->shared_irqs) {
 		/* request one shared interrupt */
-		if (devm_request_irq(&pdev->dev, p->irq[0].requested_irq,
+		if (devm_request_irq(dev, p->irq[0].requested_irq,
 				intc_irqpin_shared_irq_handler,
 				IRQF_SHARED, name, p)) {
-			dev_err(&pdev->dev, "failed to request low IRQ\n");
+			dev_err(dev, "failed to request low IRQ\n");
 			ret = -ENOENT;
 			goto err1;
 		}
 	} else {
 		/* request interrupts one by one */
 		for (k = 0; k < p->number_of_irqs; k++) {
-			if (devm_request_irq(&pdev->dev,
-					p->irq[k].requested_irq,
-					intc_irqpin_irq_handler,
-					0, name, &p->irq[k])) {
-				dev_err(&pdev->dev,
-					"failed to request low IRQ\n");
+			if (devm_request_irq(dev, p->irq[k].requested_irq,
+					     intc_irqpin_irq_handler, 0, name,
+					     &p->irq[k])) {
+				dev_err(dev, "failed to request low IRQ\n");
 				ret = -ENOENT;
 				goto err1;
 			}
@@ -497,12 +521,12 @@ static int intc_irqpin_probe(struct platform_device *pdev)
 	for (k = 0; k < p->number_of_irqs; k++)
 		intc_irqpin_mask_unmask_prio(p, k, 0);
 
-	dev_info(&pdev->dev, "driving %d irqs\n", p->number_of_irqs);
+	dev_info(dev, "driving %d irqs\n", p->number_of_irqs);
 
 	/* warn in case of mismatch if irq base is specified */
 	if (p->config.irq_base) {
 		if (p->config.irq_base != p->irq[0].domain_irq)
-			dev_warn(&pdev->dev, "irq base mismatch (%d/%d)\n",
+			dev_warn(dev, "irq base mismatch (%d/%d)\n",
 				 p->config.irq_base, p->irq[0].domain_irq);
 	}
 
@@ -511,6 +535,8 @@ static int intc_irqpin_probe(struct platform_device *pdev)
 err1:
 	irq_domain_remove(p->irq_domain);
 err0:
+	pm_runtime_put(dev);
+	pm_runtime_disable(dev);
 	return ret;
 }
 
@@ -519,7 +545,8 @@ static int intc_irqpin_remove(struct platform_device *pdev)
 	struct intc_irqpin_priv *p = platform_get_drvdata(pdev);
 
 	irq_domain_remove(p->irq_domain);
-
+	pm_runtime_put(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
 	return 0;
 }
 
@@ -535,7 +562,6 @@ static struct platform_driver intc_irqpin_device_driver = {
 	.driver		= {
 		.name	= "renesas_intc_irqpin",
 		.of_match_table = intc_irqpin_dt_ids,
-		.owner  = THIS_MODULE,
 	}
 };
 
diff --git a/drivers/irqchip/irq-renesas-irqc.c b/drivers/irqchip/irq-renesas-irqc.c
index 8777065012a5..384e6ed61d7c 100644
--- a/drivers/irqchip/irq-renesas-irqc.c
+++ b/drivers/irqchip/irq-renesas-irqc.c
@@ -281,7 +281,6 @@ static struct platform_driver irqc_device_driver = {
 	.driver		= {
 		.name	= "renesas_irqc",
 		.of_match_table	= irqc_dt_ids,
-		.owner	= THIS_MODULE,
 	}
 };
 
diff --git a/drivers/irqchip/irq-s3c24xx.c b/drivers/irqchip/irq-s3c24xx.c
index 78a6accd205f..c8d373fcd823 100644
--- a/drivers/irqchip/irq-s3c24xx.c
+++ b/drivers/irqchip/irq-s3c24xx.c
@@ -339,7 +339,6 @@ static inline int s3c24xx_handle_intc(struct s3c_irq_intc *intc,
 {
 	int pnd;
 	int offset;
-	int irq;
 
 	pnd = __raw_readl(intc->reg_intpnd);
 	if (!pnd)
@@ -365,8 +364,7 @@ static inline int s3c24xx_handle_intc(struct s3c_irq_intc *intc,
 	if (!(pnd & (1 << offset)))
 		offset =  __ffs(pnd);
 
-	irq = irq_find_mapping(intc->domain, intc_offset + offset);
-	handle_IRQ(irq, regs);
+	handle_domain_irq(intc->domain, intc_offset + offset, regs);
 	return true;
 }
 
diff --git a/drivers/irqchip/irq-sirfsoc.c b/drivers/irqchip/irq-sirfsoc.c
index 5e54f6d71e77..a469355df352 100644
--- a/drivers/irqchip/irq-sirfsoc.c
+++ b/drivers/irqchip/irq-sirfsoc.c
@@ -50,12 +50,10 @@ sirfsoc_alloc_gc(void __iomem *base, unsigned int irq_start, unsigned int num)
 static void __exception_irq_entry sirfsoc_handle_irq(struct pt_regs *regs)
 {
 	void __iomem *base = sirfsoc_irqdomain->host_data;
-	u32 irqstat, irqnr;
+	u32 irqstat;
 
 	irqstat = readl_relaxed(base + SIRFSOC_INIT_IRQ_ID);
-	irqnr = irq_find_mapping(sirfsoc_irqdomain, irqstat & 0xff);
-
-	handle_IRQ(irqnr, regs);
+	handle_domain_irq(sirfsoc_irqdomain, irqstat & 0xff, regs);
 }
 
 static int __init sirfsoc_irq_init(struct device_node *np,
diff --git a/drivers/irqchip/irq-sun4i.c b/drivers/irqchip/irq-sun4i.c
index 6fcef4a95a18..64155b686081 100644
--- a/drivers/irqchip/irq-sun4i.c
+++ b/drivers/irqchip/irq-sun4i.c
@@ -136,7 +136,7 @@ IRQCHIP_DECLARE(allwinner_sun4i_ic, "allwinner,sun4i-a10-ic", sun4i_of_init);
 
 static void __exception_irq_entry sun4i_handle_irq(struct pt_regs *regs)
 {
-	u32 irq, hwirq;
+	u32 hwirq;
 
 	/*
 	 * hwirq == 0 can mean one of 3 things:
@@ -154,8 +154,7 @@ static void __exception_irq_entry sun4i_handle_irq(struct pt_regs *regs)
 		return;
 
 	do {
-		irq = irq_find_mapping(sun4i_irq_domain, hwirq);
-		handle_IRQ(irq, regs);
+		handle_domain_irq(sun4i_irq_domain, hwirq, regs);
 		hwirq = readl(sun4i_irq_base + SUN4I_IRQ_VECTOR_REG) >> 2;
 	} while (hwirq != 0);
 }
diff --git a/drivers/irqchip/irq-sunxi-nmi.c b/drivers/irqchip/irq-sunxi-nmi.c
index 12f547a44ae4..4a9ce5b50c5b 100644
--- a/drivers/irqchip/irq-sunxi-nmi.c
+++ b/drivers/irqchip/irq-sunxi-nmi.c
@@ -50,12 +50,12 @@ static struct sunxi_sc_nmi_reg_offs sun6i_reg_offs = {
 static inline void sunxi_sc_nmi_write(struct irq_chip_generic *gc, u32 off,
 				      u32 val)
 {
-	irq_reg_writel(val, gc->reg_base + off);
+	irq_reg_writel(gc, val, off);
 }
 
 static inline u32 sunxi_sc_nmi_read(struct irq_chip_generic *gc, u32 off)
 {
-	return irq_reg_readl(gc->reg_base + off);
+	return irq_reg_readl(gc, off);
 }
 
 static void sunxi_sc_nmi_handle_irq(unsigned int irq, struct irq_desc *desc)
diff --git a/drivers/irqchip/irq-tb10x.c b/drivers/irqchip/irq-tb10x.c
index 7c44c99bf1f2..accc20036a3c 100644
--- a/drivers/irqchip/irq-tb10x.c
+++ b/drivers/irqchip/irq-tb10x.c
@@ -43,12 +43,12 @@
 static inline void ab_irqctl_writereg(struct irq_chip_generic *gc, u32 reg,
 	u32 val)
 {
-	irq_reg_writel(val, gc->reg_base + reg);
+	irq_reg_writel(gc, val, reg);
 }
 
 static inline u32 ab_irqctl_readreg(struct irq_chip_generic *gc, u32 reg)
 {
-	return irq_reg_readl(gc->reg_base + reg);
+	return irq_reg_readl(gc, reg);
 }
 
 static int tb10x_irq_set_type(struct irq_data *data, unsigned int flow_type)
diff --git a/drivers/irqchip/irq-versatile-fpga.c b/drivers/irqchip/irq-versatile-fpga.c
index ccf58548b161..1ab451729a5c 100644
--- a/drivers/irqchip/irq-versatile-fpga.c
+++ b/drivers/irqchip/irq-versatile-fpga.c
@@ -96,7 +96,7 @@ static int handle_one_fpga(struct fpga_irq_data *f, struct pt_regs *regs)
 
 	while ((status  = readl(f->base + IRQ_STATUS))) {
 		irq = ffs(status) - 1;
-		handle_IRQ(irq_find_mapping(f->domain, irq), regs);
+		handle_domain_irq(f->domain, irq, regs);
 		handled = 1;
 	}
 
diff --git a/drivers/irqchip/irq-vic.c b/drivers/irqchip/irq-vic.c
index 7d35287f9e90..54089debf2dc 100644
--- a/drivers/irqchip/irq-vic.c
+++ b/drivers/irqchip/irq-vic.c
@@ -219,7 +219,7 @@ static int handle_one_vic(struct vic_device *vic, struct pt_regs *regs)
 
 	while ((stat = readl_relaxed(vic->base + VIC_IRQ_STATUS))) {
 		irq = ffs(stat) - 1;
-		handle_IRQ(irq_find_mapping(vic->domain, irq), regs);
+		handle_domain_irq(vic->domain, irq, regs);
 		handled = 1;
 	}
 
diff --git a/drivers/irqchip/irq-vt8500.c b/drivers/irqchip/irq-vt8500.c
index eb6e91efdec8..b7af816f2769 100644
--- a/drivers/irqchip/irq-vt8500.c
+++ b/drivers/irqchip/irq-vt8500.c
@@ -181,7 +181,7 @@ static struct irq_domain_ops vt8500_irq_domain_ops = {
 static void __exception_irq_entry vt8500_handle_irq(struct pt_regs *regs)
 {
 	u32 stat, i;
-	int irqnr, virq;
+	int irqnr;
 	void __iomem *base;
 
 	/* Loop through each active controller */
@@ -198,8 +198,7 @@ static void __exception_irq_entry vt8500_handle_irq(struct pt_regs *regs)
 				continue;
 		}
 
-		virq = irq_find_mapping(intc[i].domain, irqnr);
-		handle_IRQ(virq, regs);
+		handle_domain_irq(intc[i].domain, irqnr, regs);
 	}
 }
 
diff --git a/drivers/irqchip/irq-zevio.c b/drivers/irqchip/irq-zevio.c
index ceb3a4318f73..e4ef74ed454a 100644
--- a/drivers/irqchip/irq-zevio.c
+++ b/drivers/irqchip/irq-zevio.c
@@ -56,8 +56,7 @@ static void __exception_irq_entry zevio_handle_irq(struct pt_regs *regs)
 
 	while (readl(zevio_irq_io + IO_STATUS)) {
 		irqnr = readl(zevio_irq_io + IO_CURRENT);
-		irqnr = irq_find_mapping(zevio_irq_domain, irqnr);
-		handle_IRQ(irqnr, regs);
+		handle_domain_irq(zevio_irq_domain, irqnr, regs);
 	};
 }
 
diff --git a/drivers/isdn/capi/capidrv.c b/drivers/isdn/capi/capidrv.c
index fd6d28f3fc36..1cc6ca8bfbda 100644
--- a/drivers/isdn/capi/capidrv.c
+++ b/drivers/isdn/capi/capidrv.c
@@ -506,7 +506,10 @@ static void send_message(capidrv_contr *card, _cmsg *cmsg)
 	struct sk_buff *skb;
 	size_t len;
 
-	capi_cmsg2message(cmsg, cmsg->buf);
+	if (capi_cmsg2message(cmsg, cmsg->buf)) {
+		printk(KERN_ERR "capidrv::send_message: parser failure\n");
+		return;
+	}
 	len = CAPIMSG_LEN(cmsg->buf);
 	skb = alloc_skb(len, GFP_ATOMIC);
 	if (!skb) {
@@ -1578,7 +1581,12 @@ static _cmsg s_cmsg;
 
 static void capidrv_recv_message(struct capi20_appl *ap, struct sk_buff *skb)
 {
-	capi_message2cmsg(&s_cmsg, skb->data);
+	if (capi_message2cmsg(&s_cmsg, skb->data)) {
+		printk(KERN_ERR "capidrv: applid=%d: received invalid message\n",
+		       ap->applid);
+		kfree_skb(skb);
+		return;
+	}
 	if (debugmode > 3) {
 		_cdebbuf *cdb = capi_cmsg2str(&s_cmsg);
 
@@ -1903,7 +1911,11 @@ static int capidrv_command(isdn_ctrl *c, capidrv_contr *card)
 				       NULL,	/* Useruserdata */
 				       NULL	/* Facilitydataarray */
 			);
-		capi_cmsg2message(&cmdcmsg, cmdcmsg.buf);
+		if (capi_cmsg2message(&cmdcmsg, cmdcmsg.buf)) {
+			printk(KERN_ERR "capidrv-%d: capidrv_command: parser failure\n",
+			       card->contrnr);
+			return -EINVAL;
+		}
 		plci_change_state(card, bchan->plcip, EV_PLCI_CONNECT_RESP);
 		send_message(card, &cmdcmsg);
 		return 0;
@@ -2090,7 +2102,11 @@ static int if_sendbuf(int id, int channel, int doack, struct sk_buff *skb)
 	if (capidrv_add_ack(nccip, datahandle, doack ? (int)skb->len : -1) < 0)
 		return 0;
 
-	capi_cmsg2message(&sendcmsg, sendcmsg.buf);
+	if (capi_cmsg2message(&sendcmsg, sendcmsg.buf)) {
+		printk(KERN_ERR "capidrv-%d: if_sendbuf: parser failure\n",
+		       card->contrnr);
+		return -EINVAL;
+	}
 	msglen = CAPIMSG_LEN(sendcmsg.buf);
 	if (skb_headroom(skb) < msglen) {
 		struct sk_buff *nskb = skb_realloc_headroom(skb, msglen);
diff --git a/drivers/isdn/capi/capiutil.c b/drivers/isdn/capi/capiutil.c
index 6e797e502cfa..9846d82eb097 100644
--- a/drivers/isdn/capi/capiutil.c
+++ b/drivers/isdn/capi/capiutil.c
@@ -201,18 +201,30 @@ static unsigned char *cpars[] =
 #define structTRcpyovl(x, y, l) memmove(y, x, l)
 
 /*-------------------------------------------------------*/
-static unsigned command_2_index(unsigned c, unsigned sc)
+static unsigned command_2_index(u8 c, u8 sc)
 {
 	if (c & 0x80)
 		c = 0x9 + (c & 0x0f);
-	else if (c <= 0x0f);
 	else if (c == 0x41)
 		c = 0x9 + 0x1;
-	else if (c == 0xff)
+	if (c > 0x18)
 		c = 0x00;
 	return (sc & 3) * (0x9 + 0x9) + c;
 }
 
+/**
+ * capi_cmd2par() - find parameter string for CAPI 2.0 command/subcommand
+ * @cmd:	command number
+ * @subcmd:	subcommand number
+ *
+ * Return value: static string, NULL if command/subcommand unknown
+ */
+
+static unsigned char *capi_cmd2par(u8 cmd, u8 subcmd)
+{
+	return cpars[command_2_index(cmd, subcmd)];
+}
+
 /*-------------------------------------------------------*/
 #define TYP (cdef[cmsg->par[cmsg->p]].typ)
 #define OFF (((u8 *)cmsg) + cdef[cmsg->par[cmsg->p]].off)
@@ -305,7 +317,9 @@ unsigned capi_cmsg2message(_cmsg *cmsg, u8 *msg)
 	cmsg->m = msg;
 	cmsg->l = 8;
 	cmsg->p = 0;
-	cmsg->par = cpars[command_2_index(cmsg->Command, cmsg->Subcommand)];
+	cmsg->par = capi_cmd2par(cmsg->Command, cmsg->Subcommand);
+	if (!cmsg->par)
+		return 1;	/* invalid command/subcommand */
 
 	pars_2_message(cmsg);
 
@@ -378,7 +392,9 @@ unsigned capi_message2cmsg(_cmsg *cmsg, u8 *msg)
 	cmsg->p = 0;
 	byteTRcpy(cmsg->m + 4, &cmsg->Command);
 	byteTRcpy(cmsg->m + 5, &cmsg->Subcommand);
-	cmsg->par = cpars[command_2_index(cmsg->Command, cmsg->Subcommand)];
+	cmsg->par = capi_cmd2par(cmsg->Command, cmsg->Subcommand);
+	if (!cmsg->par)
+		return 1;	/* invalid command/subcommand */
 
 	message_2_pars(cmsg);
 
@@ -473,12 +489,17 @@ static char *mnames[] =
  * @cmd:	command number
  * @subcmd:	subcommand number
  *
- * Return value: static string, NULL if command/subcommand unknown
+ * Return value: static string
  */
 
 char *capi_cmd2str(u8 cmd, u8 subcmd)
 {
-	return mnames[command_2_index(cmd, subcmd)];
+	char *result;
+
+	result = mnames[command_2_index(cmd, subcmd)];
+	if (result == NULL)
+		result = "INVALID_COMMAND";
+	return result;
 }
 
 
@@ -628,6 +649,9 @@ static _cdebbuf *printstruct(_cdebbuf *cdb, u8 *m)
 
 static _cdebbuf *protocol_message_2_pars(_cdebbuf *cdb, _cmsg *cmsg, int level)
 {
+	if (!cmsg->par)
+		return NULL;	/* invalid command/subcommand */
+
 	for (; TYP != _CEND; cmsg->p++) {
 		int slen = 29 + 3 - level;
 		int i;
@@ -762,10 +786,10 @@ _cdebbuf *capi_message2str(u8 *msg)
 	cmsg->p = 0;
 	byteTRcpy(cmsg->m + 4, &cmsg->Command);
 	byteTRcpy(cmsg->m + 5, &cmsg->Subcommand);
-	cmsg->par = cpars[command_2_index(cmsg->Command, cmsg->Subcommand)];
+	cmsg->par = capi_cmd2par(cmsg->Command, cmsg->Subcommand);
 
 	cdb = bufprint(cdb, "%-26s ID=%03d #0x%04x LEN=%04d\n",
-		       mnames[command_2_index(cmsg->Command, cmsg->Subcommand)],
+		       capi_cmd2str(cmsg->Command, cmsg->Subcommand),
 		       ((unsigned short *) msg)[1],
 		       ((unsigned short *) msg)[3],
 		       ((unsigned short *) msg)[0]);
@@ -799,7 +823,7 @@ _cdebbuf *capi_cmsg2str(_cmsg *cmsg)
 	cmsg->l = 8;
 	cmsg->p = 0;
 	cdb = bufprint(cdb, "%s ID=%03d #0x%04x LEN=%04d\n",
-		       mnames[command_2_index(cmsg->Command, cmsg->Subcommand)],
+		       capi_cmd2str(cmsg->Command, cmsg->Subcommand),
 		       ((u16 *) cmsg->m)[1],
 		       ((u16 *) cmsg->m)[3],
 		       ((u16 *) cmsg->m)[0]);
diff --git a/drivers/isdn/capi/kcapi.c b/drivers/isdn/capi/kcapi.c
index c123709acf82..823f6985b260 100644
--- a/drivers/isdn/capi/kcapi.c
+++ b/drivers/isdn/capi/kcapi.c
@@ -1184,7 +1184,7 @@ static int old_capi_manufacturer(unsigned int cmd, void __user *data)
  * Return value: CAPI result code
  */
 
-int capi20_manufacturer(unsigned int cmd, void __user *data)
+int capi20_manufacturer(unsigned long cmd, void __user *data)
 {
 	struct capi_ctr *ctr;
 	int retval;
@@ -1259,7 +1259,7 @@ int capi20_manufacturer(unsigned int cmd, void __user *data)
 	}
 
 	default:
-		printk(KERN_ERR "kcapi: manufacturer command %d unknown.\n",
+		printk(KERN_ERR "kcapi: manufacturer command %lu unknown.\n",
 		       cmd);
 		break;
 
diff --git a/drivers/isdn/gigaset/Kconfig b/drivers/isdn/gigaset/Kconfig
index dde5e09e6267..83f62b8d82b5 100644
--- a/drivers/isdn/gigaset/Kconfig
+++ b/drivers/isdn/gigaset/Kconfig
@@ -20,7 +20,7 @@ if ISDN_DRV_GIGASET
 config GIGASET_CAPI
 	bool "Gigaset CAPI support"
 	depends on ISDN_CAPI='y'||(ISDN_CAPI='m'&&ISDN_DRV_GIGASET='m')
-	default ISDN_I4L='n'
+	default 'y'
 	help
 	  Build the Gigaset driver as a CAPI 2.0 driver interfacing with
 	  the Kernel CAPI subsystem. To use it with the old ISDN4Linux
diff --git a/drivers/isdn/gigaset/bas-gigaset.c b/drivers/isdn/gigaset/bas-gigaset.c
index b7ae0a0dd5b6..aecec6d32463 100644
--- a/drivers/isdn/gigaset/bas-gigaset.c
+++ b/drivers/isdn/gigaset/bas-gigaset.c
@@ -2365,7 +2365,7 @@ static int gigaset_probe(struct usb_interface *interface,
 	endpoint = &hostif->endpoint[0].desc;
 	usb_fill_int_urb(ucs->urb_int_in, udev,
 			 usb_rcvintpipe(udev,
-					(endpoint->bEndpointAddress) & 0x0f),
+					usb_endpoint_num(endpoint)),
 			 ucs->int_in_buf, IP_MSGSIZE, read_int_callback, cs,
 			 endpoint->bInterval);
 	rc = usb_submit_urb(ucs->urb_int_in, GFP_KERNEL);
diff --git a/drivers/isdn/gigaset/capi.c b/drivers/isdn/gigaset/capi.c
index 3286903a95d2..ccec7778cad2 100644
--- a/drivers/isdn/gigaset/capi.c
+++ b/drivers/isdn/gigaset/capi.c
@@ -250,6 +250,8 @@ static inline void dump_rawmsg(enum debuglevel level, const char *tag,
 	l -= 12;
 	if (l <= 0)
 		return;
+	if (l > 64)
+		l = 64; /* arbitrary limit */
 	dbgline = kmalloc(3 * l, GFP_ATOMIC);
 	if (!dbgline)
 		return;
@@ -645,7 +647,13 @@ int gigaset_isdn_icall(struct at_state_t *at_state)
 					__func__);
 				break;
 			}
-			capi_cmsg2message(&iif->hcmsg, __skb_put(skb, msgsize));
+			if (capi_cmsg2message(&iif->hcmsg,
+					      __skb_put(skb, msgsize))) {
+				dev_err(cs->dev, "%s: message parser failure\n",
+					__func__);
+				dev_kfree_skb_any(skb);
+				break;
+			}
 			dump_cmsg(DEBUG_CMD, __func__, &iif->hcmsg);
 
 			/* add to listeners on this B channel, update state */
@@ -691,7 +699,12 @@ static void send_disconnect_ind(struct bc_state *bcs,
 		dev_err(cs->dev, "%s: out of memory\n", __func__);
 		return;
 	}
-	capi_cmsg2message(&iif->hcmsg, __skb_put(skb, CAPI_DISCONNECT_IND_LEN));
+	if (capi_cmsg2message(&iif->hcmsg,
+			      __skb_put(skb, CAPI_DISCONNECT_IND_LEN))) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, &iif->hcmsg);
 	capi_ctr_handle_message(&iif->ctr, ap->id, skb);
 }
@@ -721,8 +734,12 @@ static void send_disconnect_b3_ind(struct bc_state *bcs,
 		dev_err(cs->dev, "%s: out of memory\n", __func__);
 		return;
 	}
-	capi_cmsg2message(&iif->hcmsg,
-			  __skb_put(skb, CAPI_DISCONNECT_B3_IND_BASELEN));
+	if (capi_cmsg2message(&iif->hcmsg,
+			  __skb_put(skb, CAPI_DISCONNECT_B3_IND_BASELEN))) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, &iif->hcmsg);
 	capi_ctr_handle_message(&iif->ctr, ap->id, skb);
 }
@@ -787,7 +804,11 @@ void gigaset_isdn_connD(struct bc_state *bcs)
 		dev_err(cs->dev, "%s: out of memory\n", __func__);
 		return;
 	}
-	capi_cmsg2message(&iif->hcmsg, __skb_put(skb, msgsize));
+	if (capi_cmsg2message(&iif->hcmsg, __skb_put(skb, msgsize))) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, &iif->hcmsg);
 	capi_ctr_handle_message(&iif->ctr, ap->id, skb);
 }
@@ -887,7 +908,11 @@ void gigaset_isdn_connB(struct bc_state *bcs)
 		dev_err(cs->dev, "%s: out of memory\n", __func__);
 		return;
 	}
-	capi_cmsg2message(&iif->hcmsg, __skb_put(skb, msgsize));
+	if (capi_cmsg2message(&iif->hcmsg, __skb_put(skb, msgsize))) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, &iif->hcmsg);
 	capi_ctr_handle_message(&iif->ctr, ap->id, skb);
 }
@@ -1094,13 +1119,19 @@ static void send_conf(struct gigaset_capi_ctr *iif,
 		      struct sk_buff *skb,
 		      u16 info)
 {
+	struct cardstate *cs = iif->ctr.driverdata;
+
 	/*
 	 * _CONF replies always only have NCCI and Info parameters
 	 * so they'll fit into the _REQ message skb
 	 */
 	capi_cmsg_answer(&iif->acmsg);
 	iif->acmsg.Info = info;
-	capi_cmsg2message(&iif->acmsg, skb->data);
+	if (capi_cmsg2message(&iif->acmsg, skb->data)) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	__skb_trim(skb, CAPI_STDCONF_LEN);
 	dump_cmsg(DEBUG_CMD, __func__, &iif->acmsg);
 	capi_ctr_handle_message(&iif->ctr, ap->id, skb);
@@ -1122,7 +1153,11 @@ static void do_facility_req(struct gigaset_capi_ctr *iif,
 	static u8 confparam[10];	/* max. 9 octets + length byte */
 
 	/* decode message */
-	capi_message2cmsg(cmsg, skb->data);
+	if (capi_message2cmsg(cmsg, skb->data)) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, cmsg);
 
 	/*
@@ -1180,6 +1215,7 @@ static void do_facility_req(struct gigaset_capi_ctr *iif,
 				confparam[3] = 2;	/* length */
 				capimsg_setu16(confparam, 4,
 					       CapiSupplementaryServiceNotSupported);
+				break;
 			}
 			info = CapiSuccess;
 			confparam[3] = 2;	/* length */
@@ -1220,6 +1256,7 @@ static void do_facility_req(struct gigaset_capi_ctr *iif,
 	}
 
 	/* send FACILITY_CONF with given Info and confirmation parameter */
+	dev_kfree_skb_any(skb);
 	capi_cmsg_answer(cmsg);
 	cmsg->Info = info;
 	cmsg->FacilityConfirmationParameter = confparam;
@@ -1229,7 +1266,11 @@ static void do_facility_req(struct gigaset_capi_ctr *iif,
 		dev_err(cs->dev, "%s: out of memory\n", __func__);
 		return;
 	}
-	capi_cmsg2message(cmsg, __skb_put(cskb, msgsize));
+	if (capi_cmsg2message(cmsg, __skb_put(cskb, msgsize))) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(cskb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, cmsg);
 	capi_ctr_handle_message(&iif->ctr, ap->id, cskb);
 }
@@ -1243,8 +1284,14 @@ static void do_listen_req(struct gigaset_capi_ctr *iif,
 			  struct gigaset_capi_appl *ap,
 			  struct sk_buff *skb)
 {
+	struct cardstate *cs = iif->ctr.driverdata;
+
 	/* decode message */
-	capi_message2cmsg(&iif->acmsg, skb->data);
+	if (capi_message2cmsg(&iif->acmsg, skb->data)) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, &iif->acmsg);
 
 	/* store listening parameters */
@@ -1261,8 +1308,14 @@ static void do_alert_req(struct gigaset_capi_ctr *iif,
 			 struct gigaset_capi_appl *ap,
 			 struct sk_buff *skb)
 {
+	struct cardstate *cs = iif->ctr.driverdata;
+
 	/* decode message */
-	capi_message2cmsg(&iif->acmsg, skb->data);
+	if (capi_message2cmsg(&iif->acmsg, skb->data)) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, &iif->acmsg);
 	send_conf(iif, ap, skb, CapiAlertAlreadySent);
 }
@@ -1287,7 +1340,11 @@ static void do_connect_req(struct gigaset_capi_ctr *iif,
 	u16 info;
 
 	/* decode message */
-	capi_message2cmsg(cmsg, skb->data);
+	if (capi_message2cmsg(cmsg, skb->data)) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, cmsg);
 
 	/* get free B channel & construct PLCI */
@@ -1574,7 +1631,11 @@ static void do_connect_resp(struct gigaset_capi_ctr *iif,
 	int channel;
 
 	/* decode message */
-	capi_message2cmsg(cmsg, skb->data);
+	if (capi_message2cmsg(cmsg, skb->data)) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, cmsg);
 	dev_kfree_skb_any(skb);
 
@@ -1740,7 +1801,11 @@ static void do_connect_b3_req(struct gigaset_capi_ctr *iif,
 	int channel;
 
 	/* decode message */
-	capi_message2cmsg(cmsg, skb->data);
+	if (capi_message2cmsg(cmsg, skb->data)) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, cmsg);
 
 	/* extract and check channel number from PLCI */
@@ -1785,7 +1850,11 @@ static void do_connect_b3_resp(struct gigaset_capi_ctr *iif,
 	u8 command;
 
 	/* decode message */
-	capi_message2cmsg(cmsg, skb->data);
+	if (capi_message2cmsg(cmsg, skb->data)) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, cmsg);
 
 	/* extract and check channel number and NCCI */
@@ -1825,7 +1894,11 @@ static void do_connect_b3_resp(struct gigaset_capi_ctr *iif,
 	capi_cmsg_header(cmsg, ap->id, command, CAPI_IND,
 			 ap->nextMessageNumber++, cmsg->adr.adrNCCI);
 	__skb_trim(skb, msgsize);
-	capi_cmsg2message(cmsg, skb->data);
+	if (capi_cmsg2message(cmsg, skb->data)) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, cmsg);
 	capi_ctr_handle_message(&iif->ctr, ap->id, skb);
 }
@@ -1847,7 +1920,11 @@ static void do_disconnect_req(struct gigaset_capi_ctr *iif,
 	int channel;
 
 	/* decode message */
-	capi_message2cmsg(cmsg, skb->data);
+	if (capi_message2cmsg(cmsg, skb->data)) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, cmsg);
 
 	/* extract and check channel number from PLCI */
@@ -1903,8 +1980,14 @@ static void do_disconnect_req(struct gigaset_capi_ctr *iif,
 			kfree(b3cmsg);
 			return;
 		}
-		capi_cmsg2message(b3cmsg,
-				  __skb_put(b3skb, CAPI_DISCONNECT_B3_IND_BASELEN));
+		if (capi_cmsg2message(b3cmsg,
+				      __skb_put(b3skb, CAPI_DISCONNECT_B3_IND_BASELEN))) {
+			dev_err(cs->dev, "%s: message parser failure\n",
+				__func__);
+			kfree(b3cmsg);
+			dev_kfree_skb_any(b3skb);
+			return;
+		}
 		dump_cmsg(DEBUG_CMD, __func__, b3cmsg);
 		kfree(b3cmsg);
 		capi_ctr_handle_message(&iif->ctr, ap->id, b3skb);
@@ -1935,7 +2018,11 @@ static void do_disconnect_b3_req(struct gigaset_capi_ctr *iif,
 	int channel;
 
 	/* decode message */
-	capi_message2cmsg(cmsg, skb->data);
+	if (capi_message2cmsg(cmsg, skb->data)) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, cmsg);
 
 	/* extract and check channel number and NCCI */
@@ -2052,8 +2139,14 @@ static void do_reset_b3_req(struct gigaset_capi_ctr *iif,
 			    struct gigaset_capi_appl *ap,
 			    struct sk_buff *skb)
 {
+	struct cardstate *cs = iif->ctr.driverdata;
+
 	/* decode message */
-	capi_message2cmsg(&iif->acmsg, skb->data);
+	if (capi_message2cmsg(&iif->acmsg, skb->data)) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, &iif->acmsg);
 	send_conf(iif, ap, skb,
 		  CapiResetProcedureNotSupportedByCurrentProtocol);
@@ -2066,8 +2159,14 @@ static void do_unsupported(struct gigaset_capi_ctr *iif,
 			   struct gigaset_capi_appl *ap,
 			   struct sk_buff *skb)
 {
+	struct cardstate *cs = iif->ctr.driverdata;
+
 	/* decode message */
-	capi_message2cmsg(&iif->acmsg, skb->data);
+	if (capi_message2cmsg(&iif->acmsg, skb->data)) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, &iif->acmsg);
 	send_conf(iif, ap, skb, CapiMessageNotSupportedInCurrentState);
 }
@@ -2079,8 +2178,14 @@ static void do_nothing(struct gigaset_capi_ctr *iif,
 		       struct gigaset_capi_appl *ap,
 		       struct sk_buff *skb)
 {
+	struct cardstate *cs = iif->ctr.driverdata;
+
 	/* decode message */
-	capi_message2cmsg(&iif->acmsg, skb->data);
+	if (capi_message2cmsg(&iif->acmsg, skb->data)) {
+		dev_err(cs->dev, "%s: message parser failure\n", __func__);
+		dev_kfree_skb_any(skb);
+		return;
+	}
 	dump_cmsg(DEBUG_CMD, __func__, &iif->acmsg);
 	dev_kfree_skb_any(skb);
 }
@@ -2357,7 +2462,7 @@ int gigaset_isdn_regdev(struct cardstate *cs, const char *isdnid)
 	struct gigaset_capi_ctr *iif;
 	int rc;
 
-	iif = kmalloc(sizeof(*iif), GFP_KERNEL);
+	iif = kzalloc(sizeof(*iif), GFP_KERNEL);
 	if (!iif) {
 		pr_err("%s: out of memory\n", __func__);
 		return -ENOMEM;
@@ -2366,7 +2471,7 @@ int gigaset_isdn_regdev(struct cardstate *cs, const char *isdnid)
 	/* prepare controller structure */
 	iif->ctr.owner         = THIS_MODULE;
 	iif->ctr.driverdata    = cs;
-	strncpy(iif->ctr.name, isdnid, sizeof(iif->ctr.name));
+	strncpy(iif->ctr.name, isdnid, sizeof(iif->ctr.name) - 1);
 	iif->ctr.driver_name   = "gigaset";
 	iif->ctr.load_firmware = NULL;
 	iif->ctr.reset_ctr     = NULL;
diff --git a/drivers/isdn/gigaset/ev-layer.c b/drivers/isdn/gigaset/ev-layer.c
index 7459b127ddd5..c8ced12fa452 100644
--- a/drivers/isdn/gigaset/ev-layer.c
+++ b/drivers/isdn/gigaset/ev-layer.c
@@ -604,14 +604,14 @@ void gigaset_handle_modem_response(struct cardstate *cs)
 }
 EXPORT_SYMBOL_GPL(gigaset_handle_modem_response);
 
-/* disconnect
+/* disconnect_nobc
  * process closing of connection associated with given AT state structure
+ * without B channel
  */
-static void disconnect(struct at_state_t **at_state_p)
+static void disconnect_nobc(struct at_state_t **at_state_p,
+			    struct cardstate *cs)
 {
 	unsigned long flags;
-	struct bc_state *bcs = (*at_state_p)->bcs;
-	struct cardstate *cs = (*at_state_p)->cs;
 
 	spin_lock_irqsave(&cs->lock, flags);
 	++(*at_state_p)->seq_index;
@@ -622,23 +622,44 @@ static void disconnect(struct at_state_t **at_state_p)
 		gig_dbg(DEBUG_EVENT, "Scheduling PC_UMMODE");
 		cs->commands_pending = 1;
 	}
-	spin_unlock_irqrestore(&cs->lock, flags);
 
-	if (bcs) {
-		/* B channel assigned: invoke hardware specific handler */
-		cs->ops->close_bchannel(bcs);
-		/* notify LL */
-		if (bcs->chstate & (CHS_D_UP | CHS_NOTIFY_LL)) {
-			bcs->chstate &= ~(CHS_D_UP | CHS_NOTIFY_LL);
-			gigaset_isdn_hupD(bcs);
-		}
-	} else {
-		/* no B channel assigned: just deallocate */
-		spin_lock_irqsave(&cs->lock, flags);
+	/* check for and deallocate temporary AT state */
+	if (!list_empty(&(*at_state_p)->list)) {
 		list_del(&(*at_state_p)->list);
 		kfree(*at_state_p);
 		*at_state_p = NULL;
-		spin_unlock_irqrestore(&cs->lock, flags);
+	}
+
+	spin_unlock_irqrestore(&cs->lock, flags);
+}
+
+/* disconnect_bc
+ * process closing of connection associated with given AT state structure
+ * and B channel
+ */
+static void disconnect_bc(struct at_state_t *at_state,
+			  struct cardstate *cs, struct bc_state *bcs)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&cs->lock, flags);
+	++at_state->seq_index;
+
+	/* revert to selected idle mode */
+	if (!cs->cidmode) {
+		cs->at_state.pending_commands |= PC_UMMODE;
+		gig_dbg(DEBUG_EVENT, "Scheduling PC_UMMODE");
+		cs->commands_pending = 1;
+	}
+	spin_unlock_irqrestore(&cs->lock, flags);
+
+	/* invoke hardware specific handler */
+	cs->ops->close_bchannel(bcs);
+
+	/* notify LL */
+	if (bcs->chstate & (CHS_D_UP | CHS_NOTIFY_LL)) {
+		bcs->chstate &= ~(CHS_D_UP | CHS_NOTIFY_LL);
+		gigaset_isdn_hupD(bcs);
 	}
 }
 
@@ -646,7 +667,7 @@ static void disconnect(struct at_state_t **at_state_p)
  * get a free AT state structure: either one of those associated with the
  * B channels of the Gigaset device, or if none of those is available,
  * a newly allocated one with bcs=NULL
- * The structure should be freed by calling disconnect() after use.
+ * The structure should be freed by calling disconnect_nobc() after use.
  */
 static inline struct at_state_t *get_free_channel(struct cardstate *cs,
 						  int cid)
@@ -1057,7 +1078,7 @@ static void do_action(int action, struct cardstate *cs,
 		      struct event_t *ev)
 {
 	struct at_state_t *at_state = *p_at_state;
-	struct at_state_t *at_state2;
+	struct bc_state *bcs2;
 	unsigned long flags;
 
 	int channel;
@@ -1156,8 +1177,8 @@ static void do_action(int action, struct cardstate *cs,
 		break;
 	case ACT_RING:
 		/* get fresh AT state structure for new CID */
-		at_state2 = get_free_channel(cs, ev->parameter);
-		if (!at_state2) {
+		at_state = get_free_channel(cs, ev->parameter);
+		if (!at_state) {
 			dev_warn(cs->dev,
 				 "RING ignored: could not allocate channel structure\n");
 			break;
@@ -1166,16 +1187,16 @@ static void do_action(int action, struct cardstate *cs,
 		/* initialize AT state structure
 		 * note that bcs may be NULL if no B channel is free
 		 */
-		at_state2->ConState = 700;
+		at_state->ConState = 700;
 		for (i = 0; i < STR_NUM; ++i) {
-			kfree(at_state2->str_var[i]);
-			at_state2->str_var[i] = NULL;
+			kfree(at_state->str_var[i]);
+			at_state->str_var[i] = NULL;
 		}
-		at_state2->int_var[VAR_ZCTP] = -1;
+		at_state->int_var[VAR_ZCTP] = -1;
 
 		spin_lock_irqsave(&cs->lock, flags);
-		at_state2->timer_expires = RING_TIMEOUT;
-		at_state2->timer_active = 1;
+		at_state->timer_expires = RING_TIMEOUT;
+		at_state->timer_active = 1;
 		spin_unlock_irqrestore(&cs->lock, flags);
 		break;
 	case ACT_ICALL:
@@ -1213,14 +1234,17 @@ static void do_action(int action, struct cardstate *cs,
 	case ACT_DISCONNECT:
 		cs->cur_at_seq = SEQ_NONE;
 		at_state->cid = -1;
-		if (bcs && cs->onechannel && cs->dle) {
+		if (!bcs) {
+			disconnect_nobc(p_at_state, cs);
+		} else if (cs->onechannel && cs->dle) {
 			/* Check for other open channels not needed:
 			 * DLE only used for M10x with one B channel.
 			 */
 			at_state->pending_commands |= PC_DLE0;
 			cs->commands_pending = 1;
-		} else
-			disconnect(p_at_state);
+		} else {
+			disconnect_bc(at_state, cs, bcs);
+		}
 		break;
 	case ACT_FAKEDLE0:
 		at_state->int_var[VAR_ZDLE] = 0;
@@ -1228,24 +1252,27 @@ static void do_action(int action, struct cardstate *cs,
 		/* fall through */
 	case ACT_DLE0:
 		cs->cur_at_seq = SEQ_NONE;
-		at_state2 = &cs->bcs[cs->curchannel].at_state;
-		disconnect(&at_state2);
+		bcs2 = cs->bcs + cs->curchannel;
+		disconnect_bc(&bcs2->at_state, cs, bcs2);
 		break;
 	case ACT_ABORTHUP:
 		cs->cur_at_seq = SEQ_NONE;
 		dev_warn(cs->dev, "Could not hang up.\n");
 		at_state->cid = -1;
-		if (bcs && cs->onechannel)
+		if (!bcs)
+			disconnect_nobc(p_at_state, cs);
+		else if (cs->onechannel)
 			at_state->pending_commands |= PC_DLE0;
 		else
-			disconnect(p_at_state);
+			disconnect_bc(at_state, cs, bcs);
 		schedule_init(cs, MS_RECOVER);
 		break;
 	case ACT_FAILDLE0:
 		cs->cur_at_seq = SEQ_NONE;
-		dev_warn(cs->dev, "Could not leave DLE mode.\n");
-		at_state2 = &cs->bcs[cs->curchannel].at_state;
-		disconnect(&at_state2);
+		dev_warn(cs->dev, "Error leaving DLE mode.\n");
+		cs->dle = 0;
+		bcs2 = cs->bcs + cs->curchannel;
+		disconnect_bc(&bcs2->at_state, cs, bcs2);
 		schedule_init(cs, MS_RECOVER);
 		break;
 	case ACT_FAILDLE1:
@@ -1274,14 +1301,14 @@ static void do_action(int action, struct cardstate *cs,
 		if (reinit_and_retry(cs, channel) < 0) {
 			dev_warn(cs->dev,
 				 "Could not get a call ID. Cannot dial.\n");
-			at_state2 = &cs->bcs[channel].at_state;
-			disconnect(&at_state2);
+			bcs2 = cs->bcs + channel;
+			disconnect_bc(&bcs2->at_state, cs, bcs2);
 		}
 		break;
 	case ACT_ABORTCID:
 		cs->cur_at_seq = SEQ_NONE;
-		at_state2 = &cs->bcs[cs->curchannel].at_state;
-		disconnect(&at_state2);
+		bcs2 = cs->bcs + cs->curchannel;
+		disconnect_bc(&bcs2->at_state, cs, bcs2);
 		break;
 
 	case ACT_DIALING:
@@ -1290,7 +1317,10 @@ static void do_action(int action, struct cardstate *cs,
 		break;
 
 	case ACT_ABORTACCEPT:	/* hangup/error/timeout during ICALL procssng */
-		disconnect(p_at_state);
+		if (bcs)
+			disconnect_bc(at_state, cs, bcs);
+		else
+			disconnect_nobc(p_at_state, cs);
 		break;
 
 	case ACT_ABORTDIAL:	/* error/timeout during dial preparation */
@@ -1379,6 +1409,11 @@ static void do_action(int action, struct cardstate *cs,
 	/* events from the LL */
 
 	case ACT_DIAL:
+		if (!ev->ptr) {
+			*p_genresp = 1;
+			*p_resp_code = RSP_ERROR;
+			break;
+		}
 		start_dial(at_state, ev->ptr, ev->parameter);
 		break;
 	case ACT_ACCEPT:
diff --git a/drivers/isdn/gigaset/gigaset.h b/drivers/isdn/gigaset/gigaset.h
index eb63a0f7a02a..166537e2dfca 100644
--- a/drivers/isdn/gigaset/gigaset.h
+++ b/drivers/isdn/gigaset/gigaset.h
@@ -751,9 +751,6 @@ void gigaset_stop(struct cardstate *cs);
 /* Tell common.c that the driver is being unloaded. */
 int gigaset_shutdown(struct cardstate *cs);
 
-/* Tell common.c that an skb has been sent. */
-void gigaset_skb_sent(struct bc_state *bcs, struct sk_buff *skb);
-
 /* Append event to the queue.
  * Returns NULL on failure or a pointer to the event on success.
  * ptr must be kmalloc()ed (and not be freed by the caller).
diff --git a/drivers/isdn/gigaset/usb-gigaset.c b/drivers/isdn/gigaset/usb-gigaset.c
index d0a41cb0cf62..5f306e2eece5 100644
--- a/drivers/isdn/gigaset/usb-gigaset.c
+++ b/drivers/isdn/gigaset/usb-gigaset.c
@@ -135,14 +135,13 @@ struct usb_cardstate {
 	/* Output buffer */
 	unsigned char		*bulk_out_buffer;
 	int			bulk_out_size;
-	__u8			bulk_out_endpointAddr;
+	int			bulk_out_epnum;
 	struct urb		*bulk_out_urb;
 
 	/* Input buffer */
 	unsigned char		*rcvbuf;
 	int			rcvbuf_size;
 	struct urb		*read_urb;
-	__u8			int_in_endpointAddr;
 
 	char			bchars[6];		/* for request 0x19 */
 };
@@ -294,7 +293,7 @@ static int gigaset_close_bchannel(struct bc_state *bcs)
 }
 
 static int write_modem(struct cardstate *cs);
-static int send_cb(struct cardstate *cs, struct cmdbuf_t *cb);
+static int send_cb(struct cardstate *cs);
 
 
 /* Write tasklet handler: Continue sending current skb, or send command, or
@@ -304,8 +303,6 @@ static void gigaset_modem_fill(unsigned long data)
 {
 	struct cardstate *cs = (struct cardstate *) data;
 	struct bc_state *bcs = &cs->bcs[0]; /* only one channel */
-	struct cmdbuf_t *cb;
-	int again;
 
 	gig_dbg(DEBUG_OUTPUT, "modem_fill");
 
@@ -314,36 +311,32 @@ static void gigaset_modem_fill(unsigned long data)
 		return;
 	}
 
-	do {
-		again = 0;
-		if (!bcs->tx_skb) { /* no skb is being sent */
-			cb = cs->cmdbuf;
-			if (cb) { /* commands to send? */
-				gig_dbg(DEBUG_OUTPUT, "modem_fill: cb");
-				if (send_cb(cs, cb) < 0) {
-					gig_dbg(DEBUG_OUTPUT,
-						"modem_fill: send_cb failed");
-					again = 1; /* no callback will be
-						      called! */
-				}
-			} else { /* skbs to send? */
-				bcs->tx_skb = skb_dequeue(&bcs->squeue);
-				if (bcs->tx_skb)
-					gig_dbg(DEBUG_INTR,
-						"Dequeued skb (Adr: %lx)!",
-						(unsigned long) bcs->tx_skb);
-			}
-		}
-
-		if (bcs->tx_skb) {
-			gig_dbg(DEBUG_OUTPUT, "modem_fill: tx_skb");
-			if (write_modem(cs) < 0) {
+again:
+	if (!bcs->tx_skb) {	/* no skb is being sent */
+		if (cs->cmdbuf) {	/* commands to send? */
+			gig_dbg(DEBUG_OUTPUT, "modem_fill: cb");
+			if (send_cb(cs) < 0) {
 				gig_dbg(DEBUG_OUTPUT,
-					"modem_fill: write_modem failed");
-				again = 1; /* no callback will be called! */
+					"modem_fill: send_cb failed");
+				goto again; /* no callback will be called! */
 			}
+			return;
 		}
-	} while (again);
+
+		/* skbs to send? */
+		bcs->tx_skb = skb_dequeue(&bcs->squeue);
+		if (!bcs->tx_skb)
+			return;
+
+		gig_dbg(DEBUG_INTR, "Dequeued skb (Adr: %lx)!",
+			(unsigned long) bcs->tx_skb);
+	}
+
+	gig_dbg(DEBUG_OUTPUT, "modem_fill: tx_skb");
+	if (write_modem(cs) < 0) {
+		gig_dbg(DEBUG_OUTPUT, "modem_fill: write_modem failed");
+		goto again;	/* no callback will be called! */
+	}
 }
 
 /*
@@ -430,9 +423,9 @@ static void gigaset_write_bulk_callback(struct urb *urb)
 	spin_unlock_irqrestore(&cs->lock, flags);
 }
 
-static int send_cb(struct cardstate *cs, struct cmdbuf_t *cb)
+static int send_cb(struct cardstate *cs)
 {
-	struct cmdbuf_t *tcb;
+	struct cmdbuf_t *cb = cs->cmdbuf;
 	unsigned long flags;
 	int count;
 	int status = -ENOENT;
@@ -440,33 +433,34 @@ static int send_cb(struct cardstate *cs, struct cmdbuf_t *cb)
 
 	do {
 		if (!cb->len) {
-			tcb = cb;
-
 			spin_lock_irqsave(&cs->cmdlock, flags);
 			cs->cmdbytes -= cs->curlen;
 			gig_dbg(DEBUG_OUTPUT, "send_cb: sent %u bytes, %u left",
 				cs->curlen, cs->cmdbytes);
-			cs->cmdbuf = cb = cb->next;
-			if (cb) {
-				cb->prev = NULL;
-				cs->curlen = cb->len;
+			cs->cmdbuf = cb->next;
+			if (cs->cmdbuf) {
+				cs->cmdbuf->prev = NULL;
+				cs->curlen = cs->cmdbuf->len;
 			} else {
 				cs->lastcmdbuf = NULL;
 				cs->curlen = 0;
 			}
 			spin_unlock_irqrestore(&cs->cmdlock, flags);
 
-			if (tcb->wake_tasklet)
-				tasklet_schedule(tcb->wake_tasklet);
-			kfree(tcb);
+			if (cb->wake_tasklet)
+				tasklet_schedule(cb->wake_tasklet);
+			kfree(cb);
+
+			cb = cs->cmdbuf;
 		}
+
 		if (cb) {
 			count = min(cb->len, ucs->bulk_out_size);
 			gig_dbg(DEBUG_OUTPUT, "send_cb: send %d bytes", count);
 
 			usb_fill_bulk_urb(ucs->bulk_out_urb, ucs->udev,
 					  usb_sndbulkpipe(ucs->udev,
-							  ucs->bulk_out_endpointAddr & 0x0f),
+							  ucs->bulk_out_epnum),
 					  cb->buf + cb->offset, count,
 					  gigaset_write_bulk_callback, cs);
 
@@ -498,6 +492,7 @@ static int send_cb(struct cardstate *cs, struct cmdbuf_t *cb)
 static int gigaset_write_cmd(struct cardstate *cs, struct cmdbuf_t *cb)
 {
 	unsigned long flags;
+	int len;
 
 	gigaset_dbg_buffer(cs->mstate != MS_LOCKED ?
 			   DEBUG_TRANSCMD : DEBUG_LOCKCMD,
@@ -516,10 +511,11 @@ static int gigaset_write_cmd(struct cardstate *cs, struct cmdbuf_t *cb)
 	spin_unlock_irqrestore(&cs->cmdlock, flags);
 
 	spin_lock_irqsave(&cs->lock, flags);
+	len = cb->len;
 	if (cs->connected)
 		tasklet_schedule(&cs->write_tasklet);
 	spin_unlock_irqrestore(&cs->lock, flags);
-	return cb->len;
+	return len;
 }
 
 static int gigaset_write_room(struct cardstate *cs)
@@ -628,8 +624,7 @@ static int write_modem(struct cardstate *cs)
 	if (cs->connected) {
 		usb_fill_bulk_urb(ucs->bulk_out_urb, ucs->udev,
 				  usb_sndbulkpipe(ucs->udev,
-						  ucs->bulk_out_endpointAddr &
-						  0x0f),
+						  ucs->bulk_out_epnum),
 				  ucs->bulk_out_buffer, count,
 				  gigaset_write_bulk_callback, cs);
 		ret = usb_submit_urb(ucs->bulk_out_urb, GFP_ATOMIC);
@@ -714,7 +709,7 @@ static int gigaset_probe(struct usb_interface *interface,
 
 	buffer_size = le16_to_cpu(endpoint->wMaxPacketSize);
 	ucs->bulk_out_size = buffer_size;
-	ucs->bulk_out_endpointAddr = endpoint->bEndpointAddress;
+	ucs->bulk_out_epnum = usb_endpoint_num(endpoint);
 	ucs->bulk_out_buffer = kmalloc(buffer_size, GFP_KERNEL);
 	if (!ucs->bulk_out_buffer) {
 		dev_err(cs->dev, "Couldn't allocate bulk_out_buffer\n");
@@ -741,7 +736,6 @@ static int gigaset_probe(struct usb_interface *interface,
 	}
 	buffer_size = le16_to_cpu(endpoint->wMaxPacketSize);
 	ucs->rcvbuf_size = buffer_size;
-	ucs->int_in_endpointAddr = endpoint->bEndpointAddress;
 	ucs->rcvbuf = kmalloc(buffer_size, GFP_KERNEL);
 	if (!ucs->rcvbuf) {
 		dev_err(cs->dev, "Couldn't allocate rcvbuf\n");
@@ -750,8 +744,7 @@ static int gigaset_probe(struct usb_interface *interface,
 	}
 	/* Fill the interrupt urb and send it to the core */
 	usb_fill_int_urb(ucs->read_urb, udev,
-			 usb_rcvintpipe(udev,
-					endpoint->bEndpointAddress & 0x0f),
+			 usb_rcvintpipe(udev, usb_endpoint_num(endpoint)),
 			 ucs->rcvbuf, buffer_size,
 			 gigaset_read_int_callback,
 			 cs, endpoint->bInterval);
diff --git a/drivers/isdn/hardware/eicon/message.c b/drivers/isdn/hardware/eicon/message.c
index a82e542ffc21..0b380603a578 100644
--- a/drivers/isdn/hardware/eicon/message.c
+++ b/drivers/isdn/hardware/eicon/message.c
@@ -4880,7 +4880,7 @@ static void sig_ind(PLCI *plci)
 	byte SS_Ind[] = "\x05\x02\x00\x02\x00\x00"; /* Hold_Ind struct*/
 	byte CF_Ind[] = "\x09\x02\x00\x06\x00\x00\x00\x00\x00\x00";
 	byte Interr_Err_Ind[] = "\x0a\x02\x00\x07\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00";
-	byte CONF_Ind[] = "\x09\x16\x00\x06\x00\x00\0x00\0x00\0x00\0x00";
+	byte CONF_Ind[] = "\x09\x16\x00\x06\x00\x00\x00\x00\x00\x00";
 	byte force_mt_info = false;
 	byte dir;
 	dword d;
diff --git a/drivers/isdn/hisax/hfc_2bs0.c b/drivers/isdn/hisax/hfc_2bs0.c
index 838531b6a60e..14dada42874e 100644
--- a/drivers/isdn/hisax/hfc_2bs0.c
+++ b/drivers/isdn/hisax/hfc_2bs0.c
@@ -31,7 +31,7 @@ WaitForBusy(struct IsdnCardState *cs)
 		to--;
 	}
 	if (!to) {
-		printk(KERN_WARNING "HiSax: waitforBusy timeout\n");
+		printk(KERN_WARNING "HiSax: %s timeout\n", __func__);
 		return (0);
 	} else
 		return (to);
diff --git a/drivers/isdn/hisax/hfc_sx.c b/drivers/isdn/hisax/hfc_sx.c
index fa1fefd711cd..b1fad81f0722 100644
--- a/drivers/isdn/hisax/hfc_sx.c
+++ b/drivers/isdn/hisax/hfc_sx.c
@@ -1159,7 +1159,8 @@ hfcsx_l2l1(struct PStack *st, int pr, void *arg)
 	case (PH_PULL | INDICATION):
 		spin_lock_irqsave(&bcs->cs->lock, flags);
 		if (bcs->tx_skb) {
-			printk(KERN_WARNING "hfc_l2l1: this shouldn't happen\n");
+			printk(KERN_WARNING "%s: this shouldn't happen\n",
+			       __func__);
 		} else {
 //				test_and_set_bit(BC_FLG_BUSY, &bcs->Flag);
 			bcs->tx_skb = skb;
diff --git a/drivers/isdn/hisax/hfc_usb.c b/drivers/isdn/hisax/hfc_usb.c
index 849a80752685..678bd5224bc3 100644
--- a/drivers/isdn/hisax/hfc_usb.c
+++ b/drivers/isdn/hisax/hfc_usb.c
@@ -927,9 +927,8 @@ start_int_fifo(usb_fifo *fifo)
 	fifo->active = 1;	/* must be marked active */
 	errcode = usb_submit_urb(fifo->urb, GFP_KERNEL);
 	if (errcode) {
-		printk(KERN_ERR
-		       "HFC-S USB: submit URB error(start_int_info): status:%i\n",
-		       errcode);
+		printk(KERN_ERR "HFC-S USB: submit URB error(%s): status:%i\n",
+		       __func__, errcode);
 		fifo->active = 0;
 		fifo->skbuff = NULL;
 	}
diff --git a/drivers/isdn/hisax/ipacx.c b/drivers/isdn/hisax/ipacx.c
index 5faa5de24305..9cc26b40a437 100644
--- a/drivers/isdn/hisax/ipacx.c
+++ b/drivers/isdn/hisax/ipacx.c
@@ -580,7 +580,7 @@ bch_fill_fifo(struct BCState *bcs)
 	if (cs->debug & L1_DEB_HSCX_FIFO) {
 		char *t = bcs->blog;
 
-		t += sprintf(t, "chb_fill_fifo() B-%d cnt %d", hscx, count);
+		t += sprintf(t, "%s() B-%d cnt %d", __func__, hscx, count);
 		QuickHex(t, ptr, count);
 		debugl1(cs, "%s", bcs->blog);
 	}
diff --git a/drivers/isdn/hisax/isdnl1.c b/drivers/isdn/hisax/isdnl1.c
index 800095781bfb..a560842c0e48 100644
--- a/drivers/isdn/hisax/isdnl1.c
+++ b/drivers/isdn/hisax/isdnl1.c
@@ -867,7 +867,7 @@ l1_msg(struct IsdnCardState *cs, int pr, void *arg) {
 			break;
 		default:
 			if (cs->debug)
-				debugl1(cs, "l1msg %04X unhandled", pr);
+				debugl1(cs, "%s %04X unhandled", __func__, pr);
 			break;
 		}
 		st = st->next;
diff --git a/drivers/isdn/hisax/isdnl3.c b/drivers/isdn/hisax/isdnl3.c
index 45b03840f716..c754706f83cd 100644
--- a/drivers/isdn/hisax/isdnl3.c
+++ b/drivers/isdn/hisax/isdnl3.c
@@ -153,7 +153,7 @@ void
 newl3state(struct l3_process *pc, int state)
 {
 	if (pc->debug & L3_DEB_STATE)
-		l3_debug(pc->st, "newstate cr %d %d --> %d",
+		l3_debug(pc->st, "%s cr %d %d --> %d", __func__,
 			 pc->callref & 0x7F,
 			 pc->state, state);
 	pc->state = state;
diff --git a/drivers/isdn/hysdn/hycapi.c b/drivers/isdn/hysdn/hycapi.c
index 00aad10507d8..93bae94314a6 100644
--- a/drivers/isdn/hysdn/hycapi.c
+++ b/drivers/isdn/hysdn/hycapi.c
@@ -501,7 +501,7 @@ static char *hycapi_procinfo(struct capi_ctr *ctrl)
 {
 	hycapictrl_info *cinfo = (hycapictrl_info *)(ctrl->driverdata);
 #ifdef HYCAPI_PRINTFNAMES
-	printk(KERN_NOTICE "hycapi_proc_info\n");
+	printk(KERN_NOTICE "%s\n", __func__);
 #endif
 	if (!cinfo)
 		return "";
diff --git a/drivers/isdn/i4l/isdn_tty.c b/drivers/isdn/i4l/isdn_tty.c
index 3c5f2491a16f..bc912611fe09 100644
--- a/drivers/isdn/i4l/isdn_tty.c
+++ b/drivers/isdn/i4l/isdn_tty.c
@@ -1043,11 +1043,6 @@ isdn_tty_change_speed(modem_info *info)
 	if (!(cflag & PARODD))
 		cval |= UART_LCR_EPAR;
 
-	/* CTS flow control flag and modem status interrupts */
-	if (cflag & CRTSCTS) {
-		port->flags |= ASYNC_CTS_FLOW;
-	} else
-		port->flags &= ~ASYNC_CTS_FLOW;
 	if (cflag & CLOCAL)
 		port->flags &= ~ASYNC_CHECK_CD;
 	else {
diff --git a/drivers/isdn/mISDN/dsp_cmx.c b/drivers/isdn/mISDN/dsp_cmx.c
index a4f05c54c32b..87f7dff20ff6 100644
--- a/drivers/isdn/mISDN/dsp_cmx.c
+++ b/drivers/isdn/mISDN/dsp_cmx.c
@@ -1454,66 +1454,63 @@ dsp_cmx_send_member(struct dsp *dsp, int len, s32 *c, int members)
 #ifdef CMX_CONF_DEBUG
 	if (0) {
 #else
-		if (members == 2) {
+	if (members == 2) {
 #endif
-			/* "other" becomes other party */
-			other = (list_entry(conf->mlist.next,
-					    struct dsp_conf_member, list))->dsp;
-			if (other == member)
-				other = (list_entry(conf->mlist.prev,
-						    struct dsp_conf_member, list))->dsp;
-			o_q = other->rx_buff; /* received data */
-			o_rr = (other->rx_R + len) & CMX_BUFF_MASK;
-			/* end of rx-pointer */
-			o_r = (o_rr - rr + r) & CMX_BUFF_MASK;
-			/* start rx-pointer at current read position*/
-			/* -> if echo is NOT enabled */
-			if (!dsp->echo.software) {
-				/*
-				 * -> copy other member's rx-data,
-				 * if tx-data is available, mix
-				 */
-				while (o_r != o_rr && t != tt) {
-					*d++ = dsp_audio_mix_law[(p[t] << 8) | o_q[o_r]];
-					t = (t + 1) & CMX_BUFF_MASK;
-					o_r = (o_r + 1) & CMX_BUFF_MASK;
-				}
-				while (o_r != o_rr) {
-					*d++ = o_q[o_r];
-					o_r = (o_r + 1) & CMX_BUFF_MASK;
-				}
-				/* -> if echo is enabled */
-			} else {
-				/*
-				 * -> mix other member's rx-data with echo,
-				 * if tx-data is available, mix
-				 */
-				while (r != rr && t != tt) {
-					sample = dsp_audio_law_to_s32[p[t]] +
-						dsp_audio_law_to_s32[q[r]] +
-						dsp_audio_law_to_s32[o_q[o_r]];
-					if (sample < -32768)
-						sample = -32768;
-					else if (sample > 32767)
-						sample = 32767;
-					*d++ = dsp_audio_s16_to_law[sample & 0xffff];
-					/* tx-data + rx_data + echo */
-					t = (t + 1) & CMX_BUFF_MASK;
-					r = (r + 1) & CMX_BUFF_MASK;
-					o_r = (o_r + 1) & CMX_BUFF_MASK;
-				}
-				while (r != rr) {
-					*d++ = dsp_audio_mix_law[(q[r] << 8) | o_q[o_r]];
-					r = (r + 1) & CMX_BUFF_MASK;
-					o_r = (o_r + 1) & CMX_BUFF_MASK;
-				}
+		/* "other" becomes other party */
+		other = (list_entry(conf->mlist.next,
+				    struct dsp_conf_member, list))->dsp;
+		if (other == member)
+			other = (list_entry(conf->mlist.prev,
+				    struct dsp_conf_member, list))->dsp;
+		o_q = other->rx_buff; /* received data */
+		o_rr = (other->rx_R + len) & CMX_BUFF_MASK;
+		/* end of rx-pointer */
+		o_r = (o_rr - rr + r) & CMX_BUFF_MASK;
+		/* start rx-pointer at current read position*/
+		/* -> if echo is NOT enabled */
+		if (!dsp->echo.software) {
+			/*
+			 * -> copy other member's rx-data,
+			 * if tx-data is available, mix
+			 */
+			while (o_r != o_rr && t != tt) {
+				*d++ = dsp_audio_mix_law[(p[t] << 8) | o_q[o_r]];
+				t = (t + 1) & CMX_BUFF_MASK;
+				o_r = (o_r + 1) & CMX_BUFF_MASK;
+			}
+			while (o_r != o_rr) {
+				*d++ = o_q[o_r];
+				o_r = (o_r + 1) & CMX_BUFF_MASK;
+			}
+			/* -> if echo is enabled */
+		} else {
+			/*
+			 * -> mix other member's rx-data with echo,
+			 * if tx-data is available, mix
+			 */
+			while (r != rr && t != tt) {
+				sample = dsp_audio_law_to_s32[p[t]] +
+					dsp_audio_law_to_s32[q[r]] +
+					dsp_audio_law_to_s32[o_q[o_r]];
+				if (sample < -32768)
+					sample = -32768;
+				else if (sample > 32767)
+					sample = 32767;
+				*d++ = dsp_audio_s16_to_law[sample & 0xffff];
+				/* tx-data + rx_data + echo */
+				t = (t + 1) & CMX_BUFF_MASK;
+				r = (r + 1) & CMX_BUFF_MASK;
+				o_r = (o_r + 1) & CMX_BUFF_MASK;
+			}
+			while (r != rr) {
+				*d++ = dsp_audio_mix_law[(q[r] << 8) | o_q[o_r]];
+				r = (r + 1) & CMX_BUFF_MASK;
+				o_r = (o_r + 1) & CMX_BUFF_MASK;
 			}
-			dsp->tx_R = t;
-			goto send_packet;
 		}
-#ifdef DSP_NEVER_DEFINED
+		dsp->tx_R = t;
+		goto send_packet;
 	}
-#endif
 	/* PROCESS DATA (three or more members) */
 	/* -> if echo is NOT enabled */
 	if (!dsp->echo.software) {
diff --git a/drivers/isdn/mISDN/l1oip_codec.c b/drivers/isdn/mISDN/l1oip_codec.c
index a601c8472220..9b033be11a5f 100644
--- a/drivers/isdn/mISDN/l1oip_codec.c
+++ b/drivers/isdn/mISDN/l1oip_codec.c
@@ -312,10 +312,8 @@ l1oip_ulaw_to_alaw(u8 *data, int len, u8 *result)
 void
 l1oip_4bit_free(void)
 {
-	if (table_dec)
-		vfree(table_dec);
-	if (table_com)
-		vfree(table_com);
+	vfree(table_dec);
+	vfree(table_com);
 	table_com = NULL;
 	table_dec = NULL;
 }
diff --git a/drivers/isdn/mISDN/l1oip_core.c b/drivers/isdn/mISDN/l1oip_core.c
index 9f454d76cc06..67c21876c35f 100644
--- a/drivers/isdn/mISDN/l1oip_core.c
+++ b/drivers/isdn/mISDN/l1oip_core.c
@@ -1334,7 +1334,7 @@ init_card(struct l1oip *hc, int pri, int bundle)
 	if (id[l1oip_cnt] == 0) {
 		printk(KERN_WARNING "Warning: No 'id' value given or "
 		       "0, this is highly unsecure. Please use 32 "
-		       "bit randmom number 0x...\n");
+		       "bit random number 0x...\n");
 	}
 	hc->id = id[l1oip_cnt];
 	if (debug & DEBUG_L1OIP_INIT)
diff --git a/drivers/isdn/mISDN/socket.c b/drivers/isdn/mISDN/socket.c
index 1be82284cf9d..84b35925ee4d 100644
--- a/drivers/isdn/mISDN/socket.c
+++ b/drivers/isdn/mISDN/socket.c
@@ -163,7 +163,7 @@ mISDN_sock_recvmsg(struct kiocb *iocb, struct socket *sock,
 	memcpy(skb_push(skb, MISDN_HEADER_LEN), mISDN_HEAD_P(skb),
 	       MISDN_HEADER_LEN);
 
-	err = skb_copy_datagram_iovec(skb, 0, msg->msg_iov, copied);
+	err = skb_copy_datagram_msg(skb, 0, msg, copied);
 
 	mISDN_sock_cmsg(sk, msg, skb);
 
@@ -203,7 +203,7 @@ mISDN_sock_sendmsg(struct kiocb *iocb, struct socket *sock,
 	if (!skb)
 		goto done;
 
-	if (memcpy_fromiovec(skb_put(skb, len), msg->msg_iov, len)) {
+	if (memcpy_from_msg(skb_put(skb, len), msg, len)) {
 		err = -EFAULT;
 		goto done;
 	}
diff --git a/drivers/isdn/pcbit/layer2.c b/drivers/isdn/pcbit/layer2.c
index 42ecfef80132..46e1240ae074 100644
--- a/drivers/isdn/pcbit/layer2.c
+++ b/drivers/isdn/pcbit/layer2.c
@@ -85,7 +85,6 @@ pcbit_l2_write(struct pcbit_dev *dev, ulong msg, ushort refnum,
 	}
 	if ((frame = kmalloc(sizeof(struct frame_buf),
 			     GFP_ATOMIC)) == NULL) {
-		printk(KERN_WARNING "pcbit_2_write: kmalloc failed\n");
 		dev_kfree_skb(skb);
 		return -1;
 	}
diff --git a/drivers/leds/Kconfig b/drivers/leds/Kconfig
index 8c96e2ddf43b..a6c3d2f153f3 100644
--- a/drivers/leds/Kconfig
+++ b/drivers/leds/Kconfig
@@ -250,6 +250,17 @@ config LEDS_LP8788
 	help
 	  This option enables support for the Keyboard LEDs on the LP8788 PMIC.
 
+config LEDS_LP8860
+	tristate "LED support for the TI LP8860 4 channel LED driver"
+	depends on LEDS_CLASS && I2C
+	select REGMAP_I2C
+	help
+	  If you say yes here you get support for the TI LP8860 4 channel
+	  LED driver.
+	  This option enables support for the display cluster LEDs
+	  on the LP8860 4 channel LED driver using the I2C communication
+	  bus.
+
 config LEDS_CLEVO_MAIL
 	tristate "Mail LED on Clevo notebook"
 	depends on LEDS_CLASS
@@ -410,7 +421,7 @@ config LEDS_MC13783
 config LEDS_NS2
 	tristate "LED support for Network Space v2 GPIO LEDs"
 	depends on LEDS_CLASS
-	depends on ARCH_KIRKWOOD || MACH_KIRKWOOD
+	depends on MACH_KIRKWOOD
 	default y
 	help
 	  This option enable support for the dual-GPIO LED found on the
@@ -420,7 +431,7 @@ config LEDS_NS2
 config LEDS_NETXBIG
 	tristate "LED support for Big Network series LEDs"
 	depends on LEDS_CLASS
-	depends on ARCH_KIRKWOOD || MACH_KIRKWOOD
+	depends on MACH_KIRKWOOD
 	default y
 	help
 	  This option enable support for LEDs found on the LaCie 2Big
@@ -468,6 +479,15 @@ config LEDS_OT200
 	  This option enables support for the LEDs on the Bachmann OT200.
 	  Say Y to enable LEDs on the Bachmann OT200.
 
+config LEDS_MENF21BMC
+	tristate "LED support for the MEN 14F021P00 BMC"
+	depends on LEDS_CLASS && MFD_MENF21BMC
+	help
+	  Say Y here to include support for the MEN 14F021P00 BMC LEDs.
+
+	  This driver can also be built as a module. If so the module
+	  will be called leds-menf21bmc.
+
 comment "LED driver for blink(1) USB RGB LED is under Special HID drivers (HID_THINGM)"
 
 config LEDS_BLINKM
@@ -478,6 +498,16 @@ config LEDS_BLINKM
 	  This option enables support for the BlinkM RGB LED connected
 	  through I2C. Say Y to enable support for the BlinkM LED.
 
+config LEDS_SYSCON
+	bool "LED support for LEDs on system controllers"
+	depends on LEDS_CLASS=y
+	depends on MFD_SYSCON
+	depends on OF
+	help
+	  This option enabled support for the LEDs on syscon type
+	  devices. This will only work with device tree enabled
+	  devices.
+
 config LEDS_VERSATILE
 	tristate "LED support for the ARM Versatile and RealView"
 	depends on ARCH_REALVIEW || ARCH_VERSATILE
diff --git a/drivers/leds/Makefile b/drivers/leds/Makefile
index d8cc5f2777de..1c65a191d907 100644
--- a/drivers/leds/Makefile
+++ b/drivers/leds/Makefile
@@ -28,6 +28,7 @@ obj-$(CONFIG_LEDS_LP5523)		+= leds-lp5523.o
 obj-$(CONFIG_LEDS_LP5562)		+= leds-lp5562.o
 obj-$(CONFIG_LEDS_LP8501)		+= leds-lp8501.o
 obj-$(CONFIG_LEDS_LP8788)		+= leds-lp8788.o
+obj-$(CONFIG_LEDS_LP8860)		+= leds-lp8860.o
 obj-$(CONFIG_LEDS_TCA6507)		+= leds-tca6507.o
 obj-$(CONFIG_LEDS_CLEVO_MAIL)		+= leds-clevo-mail.o
 obj-$(CONFIG_LEDS_IPAQ_MICRO)		+= leds-ipaq-micro.o
@@ -53,7 +54,9 @@ obj-$(CONFIG_LEDS_ASIC3)		+= leds-asic3.o
 obj-$(CONFIG_LEDS_MAX8997)		+= leds-max8997.o
 obj-$(CONFIG_LEDS_LM355x)		+= leds-lm355x.o
 obj-$(CONFIG_LEDS_BLINKM)		+= leds-blinkm.o
+obj-$(CONFIG_LEDS_SYSCON)		+= leds-syscon.o
 obj-$(CONFIG_LEDS_VERSATILE)		+= leds-versatile.o
+obj-$(CONFIG_LEDS_MENF21BMC)		+= leds-menf21bmc.o
 
 # LED SPI Drivers
 obj-$(CONFIG_LEDS_DAC124S085)		+= leds-dac124s085.o
diff --git a/drivers/leds/led-class.c b/drivers/leds/led-class.c
index aa29198fca3e..dbeebac38d31 100644
--- a/drivers/leds/led-class.c
+++ b/drivers/leds/led-class.c
@@ -9,26 +9,21 @@
  * published by the Free Software Foundation.
  */
 
-#include <linux/module.h>
-#include <linux/kernel.h>
+#include <linux/ctype.h>
+#include <linux/device.h>
+#include <linux/err.h>
 #include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/leds.h>
 #include <linux/list.h>
+#include <linux/module.h>
+#include <linux/slab.h>
 #include <linux/spinlock.h>
-#include <linux/device.h>
 #include <linux/timer.h>
-#include <linux/err.h>
-#include <linux/ctype.h>
-#include <linux/leds.h>
 #include "leds.h"
 
 static struct class *leds_class;
 
-static void led_update_brightness(struct led_classdev *led_cdev)
-{
-	if (led_cdev->brightness_get)
-		led_cdev->brightness = led_cdev->brightness_get(led_cdev);
-}
-
 static ssize_t brightness_show(struct device *dev,
 		struct device_attribute *attr, char *buf)
 {
@@ -45,28 +40,38 @@ static ssize_t brightness_store(struct device *dev,
 {
 	struct led_classdev *led_cdev = dev_get_drvdata(dev);
 	unsigned long state;
-	ssize_t ret = -EINVAL;
+	ssize_t ret;
+
+	mutex_lock(&led_cdev->led_access);
+
+	if (led_sysfs_is_disabled(led_cdev)) {
+		ret = -EBUSY;
+		goto unlock;
+	}
 
 	ret = kstrtoul(buf, 10, &state);
 	if (ret)
-		return ret;
+		goto unlock;
 
 	if (state == LED_OFF)
 		led_trigger_remove(led_cdev);
-	__led_set_brightness(led_cdev, state);
+	led_set_brightness(led_cdev, state);
 
-	return size;
+	ret = size;
+unlock:
+	mutex_unlock(&led_cdev->led_access);
+	return ret;
 }
 static DEVICE_ATTR_RW(brightness);
 
-static ssize_t led_max_brightness_show(struct device *dev,
+static ssize_t max_brightness_show(struct device *dev,
 		struct device_attribute *attr, char *buf)
 {
 	struct led_classdev *led_cdev = dev_get_drvdata(dev);
 
 	return sprintf(buf, "%u\n", led_cdev->max_brightness);
 }
-static DEVICE_ATTR(max_brightness, 0444, led_max_brightness_show, NULL);
+static DEVICE_ATTR_RO(max_brightness);
 
 #ifdef CONFIG_LEDS_TRIGGERS
 static DEVICE_ATTR(trigger, 0644, led_trigger_show, led_trigger_store);
@@ -104,7 +109,7 @@ static void led_timer_function(unsigned long data)
 	unsigned long delay;
 
 	if (!led_cdev->blink_delay_on || !led_cdev->blink_delay_off) {
-		__led_set_brightness(led_cdev, LED_OFF);
+		led_set_brightness_async(led_cdev, LED_OFF);
 		return;
 	}
 
@@ -127,7 +132,7 @@ static void led_timer_function(unsigned long data)
 		delay = led_cdev->blink_delay_off;
 	}
 
-	__led_set_brightness(led_cdev, brightness);
+	led_set_brightness_async(led_cdev, brightness);
 
 	/* Return in next iteration if led is in one-shot mode and we are in
 	 * the final blink state so that the led is toggled each delay_on +
@@ -153,7 +158,7 @@ static void set_brightness_delayed(struct work_struct *ws)
 
 	led_stop_software_blink(led_cdev);
 
-	__led_set_brightness(led_cdev, led_cdev->delayed_set_value);
+	led_set_brightness_async(led_cdev, led_cdev->delayed_set_value);
 }
 
 /**
@@ -219,6 +224,7 @@ int led_classdev_register(struct device *parent, struct led_classdev *led_cdev)
 #ifdef CONFIG_LEDS_TRIGGERS
 	init_rwsem(&led_cdev->trigger_lock);
 #endif
+	mutex_init(&led_cdev->led_access);
 	/* add to the list of leds */
 	down_write(&leds_list_lock);
 	list_add_tail(&led_cdev->node, &leds_list);
@@ -227,6 +233,8 @@ int led_classdev_register(struct device *parent, struct led_classdev *led_cdev)
 	if (!led_cdev->max_brightness)
 		led_cdev->max_brightness = LED_FULL;
 
+	led_cdev->flags |= SET_BRIGHTNESS_ASYNC;
+
 	led_update_brightness(led_cdev);
 
 	INIT_WORK(&led_cdev->set_brightness_work, set_brightness_delayed);
@@ -272,6 +280,8 @@ void led_classdev_unregister(struct led_classdev *led_cdev)
 	down_write(&leds_list_lock);
 	list_del(&led_cdev->node);
 	up_write(&leds_list_lock);
+
+	mutex_destroy(&led_cdev->led_access);
 }
 EXPORT_SYMBOL_GPL(led_classdev_unregister);
 
diff --git a/drivers/leds/led-core.c b/drivers/leds/led-core.c
index 71b40d3bf776..9886dace5ad2 100644
--- a/drivers/leds/led-core.c
+++ b/drivers/leds/led-core.c
@@ -12,10 +12,11 @@
  */
 
 #include <linux/kernel.h>
+#include <linux/leds.h>
 #include <linux/list.h>
 #include <linux/module.h>
+#include <linux/mutex.h>
 #include <linux/rwsem.h>
-#include <linux/leds.h>
 #include "leds.h"
 
 DECLARE_RWSEM(leds_list_lock);
@@ -41,13 +42,13 @@ static void led_set_software_blink(struct led_classdev *led_cdev,
 
 	/* never on - just set to off */
 	if (!delay_on) {
-		__led_set_brightness(led_cdev, LED_OFF);
+		led_set_brightness_async(led_cdev, LED_OFF);
 		return;
 	}
 
 	/* never off - just set to brightness */
 	if (!delay_off) {
-		__led_set_brightness(led_cdev, led_cdev->blink_brightness);
+		led_set_brightness_async(led_cdev, led_cdev->blink_brightness);
 		return;
 	}
 
@@ -116,6 +117,8 @@ EXPORT_SYMBOL_GPL(led_stop_software_blink);
 void led_set_brightness(struct led_classdev *led_cdev,
 			enum led_brightness brightness)
 {
+	int ret = 0;
+
 	/* delay brightness setting if need to stop soft-blink timer */
 	if (led_cdev->blink_delay_on || led_cdev->blink_delay_off) {
 		led_cdev->delayed_set_value = brightness;
@@ -123,6 +126,50 @@ void led_set_brightness(struct led_classdev *led_cdev,
 		return;
 	}
 
-	__led_set_brightness(led_cdev, brightness);
+	if (led_cdev->flags & SET_BRIGHTNESS_ASYNC) {
+		led_set_brightness_async(led_cdev, brightness);
+		return;
+	} else if (led_cdev->flags & SET_BRIGHTNESS_SYNC)
+		ret = led_set_brightness_sync(led_cdev, brightness);
+	else
+		ret = -EINVAL;
+
+	if (ret < 0)
+		dev_dbg(led_cdev->dev, "Setting LED brightness failed (%d)\n",
+			ret);
 }
 EXPORT_SYMBOL(led_set_brightness);
+
+int led_update_brightness(struct led_classdev *led_cdev)
+{
+	int ret = 0;
+
+	if (led_cdev->brightness_get) {
+		ret = led_cdev->brightness_get(led_cdev);
+		if (ret >= 0) {
+			led_cdev->brightness = ret;
+			return 0;
+		}
+	}
+
+	return ret;
+}
+EXPORT_SYMBOL(led_update_brightness);
+
+/* Caller must ensure led_cdev->led_access held */
+void led_sysfs_disable(struct led_classdev *led_cdev)
+{
+	lockdep_assert_held(&led_cdev->led_access);
+
+	led_cdev->flags |= LED_SYSFS_DISABLE;
+}
+EXPORT_SYMBOL_GPL(led_sysfs_disable);
+
+/* Caller must ensure led_cdev->led_access held */
+void led_sysfs_enable(struct led_classdev *led_cdev)
+{
+	lockdep_assert_held(&led_cdev->led_access);
+
+	led_cdev->flags &= ~LED_SYSFS_DISABLE;
+}
+EXPORT_SYMBOL_GPL(led_sysfs_enable);
diff --git a/drivers/leds/led-triggers.c b/drivers/leds/led-triggers.c
index c3734f10fdd5..e8b1120f486d 100644
--- a/drivers/leds/led-triggers.c
+++ b/drivers/leds/led-triggers.c
@@ -37,6 +37,14 @@ ssize_t led_trigger_store(struct device *dev, struct device_attribute *attr,
 	char trigger_name[TRIG_NAME_MAX];
 	struct led_trigger *trig;
 	size_t len;
+	int ret = count;
+
+	mutex_lock(&led_cdev->led_access);
+
+	if (led_sysfs_is_disabled(led_cdev)) {
+		ret = -EBUSY;
+		goto unlock;
+	}
 
 	trigger_name[sizeof(trigger_name) - 1] = '\0';
 	strncpy(trigger_name, buf, sizeof(trigger_name) - 1);
@@ -47,7 +55,7 @@ ssize_t led_trigger_store(struct device *dev, struct device_attribute *attr,
 
 	if (!strcmp(trigger_name, "none")) {
 		led_trigger_remove(led_cdev);
-		return count;
+		goto unlock;
 	}
 
 	down_read(&triggers_list_lock);
@@ -58,12 +66,14 @@ ssize_t led_trigger_store(struct device *dev, struct device_attribute *attr,
 			up_write(&led_cdev->trigger_lock);
 
 			up_read(&triggers_list_lock);
-			return count;
+			goto unlock;
 		}
 	}
 	up_read(&triggers_list_lock);
 
-	return -EINVAL;
+unlock:
+	mutex_unlock(&led_cdev->led_access);
+	return ret;
 }
 EXPORT_SYMBOL_GPL(led_trigger_store);
 
diff --git a/drivers/leds/leds-88pm860x.c b/drivers/leds/leds-88pm860x.c
index c2def5551ce1..1497a09166d6 100644
--- a/drivers/leds/leds-88pm860x.c
+++ b/drivers/leds/leds-88pm860x.c
@@ -237,7 +237,6 @@ static int pm860x_led_remove(struct platform_device *pdev)
 static struct platform_driver pm860x_led_driver = {
 	.driver	= {
 		.name	= "88pm860x-led",
-		.owner	= THIS_MODULE,
 	},
 	.probe	= pm860x_led_probe,
 	.remove	= pm860x_led_remove,
diff --git a/drivers/leds/leds-adp5520.c b/drivers/leds/leds-adp5520.c
index 5036d7b4f82e..07e66cae32d3 100644
--- a/drivers/leds/leds-adp5520.c
+++ b/drivers/leds/leds-adp5520.c
@@ -201,7 +201,6 @@ static int adp5520_led_remove(struct platform_device *pdev)
 static struct platform_driver adp5520_led_driver = {
 	.driver	= {
 		.name	= "adp5520-led",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= adp5520_led_probe,
 	.remove		= adp5520_led_remove,
diff --git a/drivers/leds/leds-asic3.c b/drivers/leds/leds-asic3.c
index 70c74a7f0dfe..1b71eac639f0 100644
--- a/drivers/leds/leds-asic3.c
+++ b/drivers/leds/leds-asic3.c
@@ -168,7 +168,6 @@ static struct platform_driver asic3_led_driver = {
 	.remove		= asic3_led_remove,
 	.driver		= {
 		.name	= "leds-asic3",
-		.owner	= THIS_MODULE,
 		.pm	= &asic3_led_pm_ops,
 	},
 };
diff --git a/drivers/leds/leds-clevo-mail.c b/drivers/leds/leds-clevo-mail.c
index f58a354428e3..0f9ed1ea0e89 100644
--- a/drivers/leds/leds-clevo-mail.c
+++ b/drivers/leds/leds-clevo-mail.c
@@ -168,7 +168,6 @@ static struct platform_driver clevo_mail_led_driver = {
 	.remove		= clevo_mail_led_remove,
 	.driver		= {
 		.name		= KBUILD_MODNAME,
-		.owner		= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/leds/leds-cobalt-qube.c b/drivers/leds/leds-cobalt-qube.c
index 910339d86edf..d97522080491 100644
--- a/drivers/leds/leds-cobalt-qube.c
+++ b/drivers/leds/leds-cobalt-qube.c
@@ -76,7 +76,6 @@ static struct platform_driver cobalt_qube_led_driver = {
 	.remove	= cobalt_qube_led_remove,
 	.driver	= {
 		.name	= "cobalt-qube-leds",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/leds/leds-cobalt-raq.c b/drivers/leds/leds-cobalt-raq.c
index 001088b31373..06dbe18a2065 100644
--- a/drivers/leds/leds-cobalt-raq.c
+++ b/drivers/leds/leds-cobalt-raq.c
@@ -124,7 +124,6 @@ static struct platform_driver cobalt_raq_led_driver = {
 	.remove	= cobalt_raq_led_remove,
 	.driver = {
 		.name	= "cobalt-raq-leds",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/leds/leds-da903x.c b/drivers/leds/leds-da903x.c
index 54b8b5216b8b..952ba96e5b38 100644
--- a/drivers/leds/leds-da903x.c
+++ b/drivers/leds/leds-da903x.c
@@ -144,7 +144,6 @@ static int da903x_led_remove(struct platform_device *pdev)
 static struct platform_driver da903x_led_driver = {
 	.driver	= {
 		.name	= "da903x-led",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= da903x_led_probe,
 	.remove		= da903x_led_remove,
diff --git a/drivers/leds/leds-da9052.c b/drivers/leds/leds-da9052.c
index e4da1f460ac5..28291b6acc8e 100644
--- a/drivers/leds/leds-da9052.c
+++ b/drivers/leds/leds-da9052.c
@@ -199,7 +199,6 @@ static int da9052_led_remove(struct platform_device *pdev)
 static struct platform_driver da9052_led_driver = {
 	.driver		= {
 		.name	= "da9052-leds",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= da9052_led_probe,
 	.remove		= da9052_led_remove,
diff --git a/drivers/leds/leds-gpio-register.c b/drivers/leds/leds-gpio-register.c
index 1c4ed5510f35..75717ba68ae0 100644
--- a/drivers/leds/leds-gpio-register.c
+++ b/drivers/leds/leds-gpio-register.c
@@ -7,9 +7,9 @@
  * Free Software Foundation.
  */
 #include <linux/err.h>
+#include <linux/leds.h>
 #include <linux/platform_device.h>
 #include <linux/slab.h>
-#include <linux/leds.h>
 
 /**
  * gpio_led_register_device - register a gpio-led device
@@ -28,6 +28,9 @@ struct platform_device *__init gpio_led_register_device(
 	struct platform_device *ret;
 	struct gpio_led_platform_data _pdata = *pdata;
 
+	if (!pdata->num_leds)
+		return ERR_PTR(-EINVAL);
+
 	_pdata.leds = kmemdup(pdata->leds,
 			pdata->num_leds * sizeof(*pdata->leds), GFP_KERNEL);
 	if (!_pdata.leds)
diff --git a/drivers/leds/leds-gpio.c b/drivers/leds/leds-gpio.c
index 57ff20fecf57..7ea1ea42c2d2 100644
--- a/drivers/leds/leds-gpio.c
+++ b/drivers/leds/leds-gpio.c
@@ -10,42 +10,39 @@
  * published by the Free Software Foundation.
  *
  */
-#include <linux/kernel.h>
-#include <linux/platform_device.h>
+#include <linux/err.h>
 #include <linux/gpio.h>
+#include <linux/gpio/consumer.h>
+#include <linux/kernel.h>
 #include <linux/leds.h>
-#include <linux/of.h>
-#include <linux/of_platform.h>
-#include <linux/of_gpio.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/property.h>
 #include <linux/slab.h>
 #include <linux/workqueue.h>
-#include <linux/module.h>
-#include <linux/err.h>
 
 struct gpio_led_data {
 	struct led_classdev cdev;
-	unsigned gpio;
+	struct gpio_desc *gpiod;
 	struct work_struct work;
 	u8 new_level;
 	u8 can_sleep;
-	u8 active_low;
 	u8 blinking;
-	int (*platform_gpio_blink_set)(unsigned gpio, int state,
+	int (*platform_gpio_blink_set)(struct gpio_desc *desc, int state,
 			unsigned long *delay_on, unsigned long *delay_off);
 };
 
 static void gpio_led_work(struct work_struct *work)
 {
-	struct gpio_led_data	*led_dat =
+	struct gpio_led_data *led_dat =
 		container_of(work, struct gpio_led_data, work);
 
 	if (led_dat->blinking) {
-		led_dat->platform_gpio_blink_set(led_dat->gpio,
-						 led_dat->new_level,
-						 NULL, NULL);
+		led_dat->platform_gpio_blink_set(led_dat->gpiod,
+					led_dat->new_level, NULL, NULL);
 		led_dat->blinking = 0;
 	} else
-		gpio_set_value_cansleep(led_dat->gpio, led_dat->new_level);
+		gpiod_set_value_cansleep(led_dat->gpiod, led_dat->new_level);
 }
 
 static void gpio_led_set(struct led_classdev *led_cdev,
@@ -60,9 +57,6 @@ static void gpio_led_set(struct led_classdev *led_cdev,
 	else
 		level = 1;
 
-	if (led_dat->active_low)
-		level = !level;
-
 	/* Setting GPIOs with I2C/etc requires a task context, and we don't
 	 * seem to have a reliable way to know if we're already in one; so
 	 * let's just assume the worst.
@@ -72,11 +66,11 @@ static void gpio_led_set(struct led_classdev *led_cdev,
 		schedule_work(&led_dat->work);
 	} else {
 		if (led_dat->blinking) {
-			led_dat->platform_gpio_blink_set(led_dat->gpio, level,
+			led_dat->platform_gpio_blink_set(led_dat->gpiod, level,
 							 NULL, NULL);
 			led_dat->blinking = 0;
 		} else
-			gpio_set_value(led_dat->gpio, level);
+			gpiod_set_value(led_dat->gpiod, level);
 	}
 }
 
@@ -87,34 +81,49 @@ static int gpio_blink_set(struct led_classdev *led_cdev,
 		container_of(led_cdev, struct gpio_led_data, cdev);
 
 	led_dat->blinking = 1;
-	return led_dat->platform_gpio_blink_set(led_dat->gpio, GPIO_LED_BLINK,
+	return led_dat->platform_gpio_blink_set(led_dat->gpiod, GPIO_LED_BLINK,
 						delay_on, delay_off);
 }
 
 static int create_gpio_led(const struct gpio_led *template,
 	struct gpio_led_data *led_dat, struct device *parent,
-	int (*blink_set)(unsigned, int, unsigned long *, unsigned long *))
+	int (*blink_set)(struct gpio_desc *, int, unsigned long *,
+			 unsigned long *))
 {
 	int ret, state;
 
-	led_dat->gpio = -1;
+	led_dat->gpiod = template->gpiod;
+	if (!led_dat->gpiod) {
+		/*
+		 * This is the legacy code path for platform code that
+		 * still uses GPIO numbers. Ultimately we would like to get
+		 * rid of this block completely.
+		 */
+		unsigned long flags = 0;
+
+		/* skip leds that aren't available */
+		if (!gpio_is_valid(template->gpio)) {
+			dev_info(parent, "Skipping unavailable LED gpio %d (%s)\n",
+					template->gpio, template->name);
+			return 0;
+		}
 
-	/* skip leds that aren't available */
-	if (!gpio_is_valid(template->gpio)) {
-		dev_info(parent, "Skipping unavailable LED gpio %d (%s)\n",
-				template->gpio, template->name);
-		return 0;
-	}
+		if (template->active_low)
+			flags |= GPIOF_ACTIVE_LOW;
 
-	ret = devm_gpio_request(parent, template->gpio, template->name);
-	if (ret < 0)
-		return ret;
+		ret = devm_gpio_request_one(parent, template->gpio, flags,
+					    template->name);
+		if (ret < 0)
+			return ret;
+
+		led_dat->gpiod = gpio_to_desc(template->gpio);
+		if (IS_ERR(led_dat->gpiod))
+			return PTR_ERR(led_dat->gpiod);
+	}
 
 	led_dat->cdev.name = template->name;
 	led_dat->cdev.default_trigger = template->default_trigger;
-	led_dat->gpio = template->gpio;
-	led_dat->can_sleep = gpio_cansleep(template->gpio);
-	led_dat->active_low = template->active_low;
+	led_dat->can_sleep = gpiod_cansleep(led_dat->gpiod);
 	led_dat->blinking = 0;
 	if (blink_set) {
 		led_dat->platform_gpio_blink_set = blink_set;
@@ -122,30 +131,24 @@ static int create_gpio_led(const struct gpio_led *template,
 	}
 	led_dat->cdev.brightness_set = gpio_led_set;
 	if (template->default_state == LEDS_GPIO_DEFSTATE_KEEP)
-		state = !!gpio_get_value_cansleep(led_dat->gpio) ^ led_dat->active_low;
+		state = !!gpiod_get_value_cansleep(led_dat->gpiod);
 	else
 		state = (template->default_state == LEDS_GPIO_DEFSTATE_ON);
 	led_dat->cdev.brightness = state ? LED_FULL : LED_OFF;
 	if (!template->retain_state_suspended)
 		led_dat->cdev.flags |= LED_CORE_SUSPENDRESUME;
 
-	ret = gpio_direction_output(led_dat->gpio, led_dat->active_low ^ state);
+	ret = gpiod_direction_output(led_dat->gpiod, state);
 	if (ret < 0)
 		return ret;
 
 	INIT_WORK(&led_dat->work, gpio_led_work);
 
-	ret = led_classdev_register(parent, &led_dat->cdev);
-	if (ret < 0)
-		return ret;
-
-	return 0;
+	return led_classdev_register(parent, &led_dat->cdev);
 }
 
 static void delete_gpio_led(struct gpio_led_data *led)
 {
-	if (!gpio_is_valid(led->gpio))
-		return;
 	led_classdev_unregister(&led->cdev);
 	cancel_work_sync(&led->work);
 }
@@ -161,40 +164,47 @@ static inline int sizeof_gpio_leds_priv(int num_leds)
 		(sizeof(struct gpio_led_data) * num_leds);
 }
 
-/* Code to create from OpenFirmware platform devices */
-#ifdef CONFIG_OF_GPIO
-static struct gpio_leds_priv *gpio_leds_create_of(struct platform_device *pdev)
+static struct gpio_leds_priv *gpio_leds_create(struct platform_device *pdev)
 {
-	struct device_node *np = pdev->dev.of_node, *child;
+	struct device *dev = &pdev->dev;
+	struct fwnode_handle *child;
 	struct gpio_leds_priv *priv;
 	int count, ret;
+	struct device_node *np;
 
-	/* count LEDs in this device, so we know how much to allocate */
-	count = of_get_available_child_count(np);
+	count = device_get_child_node_count(dev);
 	if (!count)
 		return ERR_PTR(-ENODEV);
 
-	for_each_available_child_of_node(np, child)
-		if (of_get_gpio(child, 0) == -EPROBE_DEFER)
-			return ERR_PTR(-EPROBE_DEFER);
-
-	priv = devm_kzalloc(&pdev->dev, sizeof_gpio_leds_priv(count),
-			GFP_KERNEL);
+	priv = devm_kzalloc(dev, sizeof_gpio_leds_priv(count), GFP_KERNEL);
 	if (!priv)
 		return ERR_PTR(-ENOMEM);
 
-	for_each_available_child_of_node(np, child) {
+	device_for_each_child_node(dev, child) {
 		struct gpio_led led = {};
-		enum of_gpio_flags flags;
-		const char *state;
-
-		led.gpio = of_get_gpio_flags(child, 0, &flags);
-		led.active_low = flags & OF_GPIO_ACTIVE_LOW;
-		led.name = of_get_property(child, "label", NULL) ? : child->name;
-		led.default_trigger =
-			of_get_property(child, "linux,default-trigger", NULL);
-		state = of_get_property(child, "default-state", NULL);
-		if (state) {
+		const char *state = NULL;
+
+		led.gpiod = devm_get_gpiod_from_child(dev, child);
+		if (IS_ERR(led.gpiod)) {
+			fwnode_handle_put(child);
+			goto err;
+		}
+
+		np = of_node(child);
+
+		if (fwnode_property_present(child, "label")) {
+			fwnode_property_read_string(child, "label", &led.name);
+		} else {
+			if (IS_ENABLED(CONFIG_OF) && !led.name && np)
+				led.name = np->name;
+			if (!led.name)
+				return ERR_PTR(-EINVAL);
+		}
+		fwnode_property_read_string(child, "linux,default-trigger",
+					    &led.default_trigger);
+
+		if (!fwnode_property_read_string(child, "default-state",
+						 &state)) {
 			if (!strcmp(state, "keep"))
 				led.default_state = LEDS_GPIO_DEFSTATE_KEEP;
 			else if (!strcmp(state, "on"))
@@ -203,13 +213,13 @@ static struct gpio_leds_priv *gpio_leds_create_of(struct platform_device *pdev)
 				led.default_state = LEDS_GPIO_DEFSTATE_OFF;
 		}
 
-		if (of_get_property(child, "retain-state-suspended", NULL))
+		if (fwnode_property_present(child, "retain-state-suspended"))
 			led.retain_state_suspended = 1;
 
 		ret = create_gpio_led(&led, &priv->leds[priv->num_leds++],
-				      &pdev->dev, NULL);
+				      dev, NULL);
 		if (ret < 0) {
-			of_node_put(child);
+			fwnode_handle_put(child);
 			goto err;
 		}
 	}
@@ -228,13 +238,6 @@ static const struct of_device_id of_gpio_leds_match[] = {
 };
 
 MODULE_DEVICE_TABLE(of, of_gpio_leds_match);
-#else /* CONFIG_OF_GPIO */
-static struct gpio_leds_priv *gpio_leds_create_of(struct platform_device *pdev)
-{
-	return ERR_PTR(-ENODEV);
-}
-#endif /* CONFIG_OF_GPIO */
-
 
 static int gpio_led_probe(struct platform_device *pdev)
 {
@@ -242,7 +245,6 @@ static int gpio_led_probe(struct platform_device *pdev)
 	struct gpio_leds_priv *priv;
 	int i, ret = 0;
 
-
 	if (pdata && pdata->num_leds) {
 		priv = devm_kzalloc(&pdev->dev,
 				sizeof_gpio_leds_priv(pdata->num_leds),
@@ -263,7 +265,7 @@ static int gpio_led_probe(struct platform_device *pdev)
 			}
 		}
 	} else {
-		priv = gpio_leds_create_of(pdev);
+		priv = gpio_leds_create(pdev);
 		if (IS_ERR(priv))
 			return PTR_ERR(priv);
 	}
@@ -289,8 +291,7 @@ static struct platform_driver gpio_led_driver = {
 	.remove		= gpio_led_remove,
 	.driver		= {
 		.name	= "leds-gpio",
-		.owner	= THIS_MODULE,
-		.of_match_table = of_match_ptr(of_gpio_leds_match),
+		.of_match_table = of_gpio_leds_match,
 	},
 };
 
diff --git a/drivers/leds/leds-hp6xx.c b/drivers/leds/leds-hp6xx.c
index d61a98896c71..0b84c0113126 100644
--- a/drivers/leds/leds-hp6xx.c
+++ b/drivers/leds/leds-hp6xx.c
@@ -83,7 +83,6 @@ static struct platform_driver hp6xxled_driver = {
 	.remove		= hp6xxled_remove,
 	.driver		= {
 		.name		= "hp6xx-led",
-		.owner		= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/leds/leds-lm3533.c b/drivers/leds/leds-lm3533.c
index cbf61a40137d..6e2e02035dd7 100644
--- a/drivers/leds/leds-lm3533.c
+++ b/drivers/leds/leds-lm3533.c
@@ -766,7 +766,6 @@ static void lm3533_led_shutdown(struct platform_device *pdev)
 static struct platform_driver lm3533_led_driver = {
 	.driver = {
 		.name = "lm3533-leds",
-		.owner = THIS_MODULE,
 	},
 	.probe		= lm3533_led_probe,
 	.remove		= lm3533_led_remove,
diff --git a/drivers/leds/leds-lp3944.c b/drivers/leds/leds-lp3944.c
index 8e1abdcd4c9d..53144fb96167 100644
--- a/drivers/leds/leds-lp3944.c
+++ b/drivers/leds/leds-lp3944.c
@@ -335,7 +335,8 @@ static int lp3944_configure(struct i2c_client *client,
 			}
 
 			/* to expose the default value to userspace */
-			led->ldev.brightness = led->status;
+			led->ldev.brightness =
+					(enum led_brightness) led->status;
 
 			/* Set the default led status */
 			err = lp3944_led_set(led, led->status);
diff --git a/drivers/leds/leds-lp8788.c b/drivers/leds/leds-lp8788.c
index 7c2cb384e7ae..3409f03c1fa8 100644
--- a/drivers/leds/leds-lp8788.c
+++ b/drivers/leds/leds-lp8788.c
@@ -183,7 +183,6 @@ static struct platform_driver lp8788_led_driver = {
 	.remove = lp8788_led_remove,
 	.driver = {
 		.name = LP8788_DEV_KEYLED,
-		.owner = THIS_MODULE,
 	},
 };
 module_platform_driver(lp8788_led_driver);
diff --git a/drivers/leds/leds-lp8860.c b/drivers/leds/leds-lp8860.c
new file mode 100644
index 000000000000..840e93f3ab3e
--- /dev/null
+++ b/drivers/leds/leds-lp8860.c
@@ -0,0 +1,491 @@
+/*
+ * TI LP8860 4-Channel LED Driver
+ *
+ * Copyright (C) 2014 Texas Instruments
+ *
+ * Author: Dan Murphy <dmurphy@ti.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * version 2 as published by the Free Software Foundation.
+ *
+ */
+
+#include <linux/i2c.h>
+#include <linux/init.h>
+#include <linux/leds.h>
+#include <linux/regmap.h>
+#include <linux/regulator/consumer.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/of.h>
+#include <linux/of_gpio.h>
+#include <linux/gpio/consumer.h>
+#include <linux/slab.h>
+
+#define LP8860_DISP_CL1_BRT_MSB		0x00
+#define LP8860_DISP_CL1_BRT_LSB		0x01
+#define LP8860_DISP_CL1_CURR_MSB	0x02
+#define LP8860_DISP_CL1_CURR_LSB	0x03
+#define LP8860_CL2_BRT_MSB		0x04
+#define LP8860_CL2_BRT_LSB		0x05
+#define LP8860_CL2_CURRENT		0x06
+#define LP8860_CL3_BRT_MSB		0x07
+#define LP8860_CL3_BRT_LSB		0x08
+#define LP8860_CL3_CURRENT		0x09
+#define LP8860_CL4_BRT_MSB		0x0a
+#define LP8860_CL4_BRT_LSB		0x0b
+#define LP8860_CL4_CURRENT		0x0c
+#define LP8860_CONFIG			0x0d
+#define LP8860_STATUS			0x0e
+#define LP8860_FAULT			0x0f
+#define LP8860_LED_FAULT		0x10
+#define LP8860_FAULT_CLEAR		0x11
+#define LP8860_ID			0x12
+#define LP8860_TEMP_MSB			0x13
+#define LP8860_TEMP_LSB			0x14
+#define LP8860_DISP_LED_CURR_MSB	0x15
+#define LP8860_DISP_LED_CURR_LSB	0x16
+#define LP8860_DISP_LED_PWM_MSB		0x17
+#define LP8860_DISP_LED_PWM_LSB		0x18
+#define LP8860_EEPROM_CNTRL		0x19
+#define LP8860_EEPROM_UNLOCK		0x1a
+
+#define LP8860_EEPROM_REG_0		0x60
+#define LP8860_EEPROM_REG_1		0x61
+#define LP8860_EEPROM_REG_2		0x62
+#define LP8860_EEPROM_REG_3		0x63
+#define LP8860_EEPROM_REG_4		0x64
+#define LP8860_EEPROM_REG_5		0x65
+#define LP8860_EEPROM_REG_6		0x66
+#define LP8860_EEPROM_REG_7		0x67
+#define LP8860_EEPROM_REG_8		0x68
+#define LP8860_EEPROM_REG_9		0x69
+#define LP8860_EEPROM_REG_10		0x6a
+#define LP8860_EEPROM_REG_11		0x6b
+#define LP8860_EEPROM_REG_12		0x6c
+#define LP8860_EEPROM_REG_13		0x6d
+#define LP8860_EEPROM_REG_14		0x6e
+#define LP8860_EEPROM_REG_15		0x6f
+#define LP8860_EEPROM_REG_16		0x70
+#define LP8860_EEPROM_REG_17		0x71
+#define LP8860_EEPROM_REG_18		0x72
+#define LP8860_EEPROM_REG_19		0x73
+#define LP8860_EEPROM_REG_20		0x74
+#define LP8860_EEPROM_REG_21		0x75
+#define LP8860_EEPROM_REG_22		0x76
+#define LP8860_EEPROM_REG_23		0x77
+#define LP8860_EEPROM_REG_24		0x78
+
+#define LP8860_LOCK_EEPROM		0x00
+#define LP8860_UNLOCK_EEPROM		0x01
+#define LP8860_PROGRAM_EEPROM		0x02
+#define LP8860_EEPROM_CODE_1		0x08
+#define LP8860_EEPROM_CODE_2		0xba
+#define LP8860_EEPROM_CODE_3		0xef
+
+#define LP8860_CLEAR_FAULTS		0x01
+
+#define LP8860_DISP_LED_NAME		"display_cluster"
+
+/**
+ * struct lp8860_led -
+ * @lock - Lock for reading/writing the device
+ * @work - Work item used to off load the brightness register writes
+ * @client - Pointer to the I2C client
+ * @led_dev - led class device pointer
+ * @regmap - Devices register map
+ * @eeprom_regmap - EEPROM register map
+ * @enable_gpio - VDDIO/EN gpio to enable communication interface
+ * @regulator - LED supply regulator pointer
+ * @brightness - Current brightness value requested
+ * @label - LED label
+**/
+struct lp8860_led {
+	struct mutex lock;
+	struct work_struct work;
+	struct i2c_client *client;
+	struct led_classdev led_dev;
+	struct regmap *regmap;
+	struct regmap *eeprom_regmap;
+	struct gpio_desc *enable_gpio;
+	struct regulator *regulator;
+	enum led_brightness brightness;
+	const char *label;
+};
+
+struct lp8860_eeprom_reg {
+	uint8_t reg;
+	uint8_t value;
+};
+
+static struct lp8860_eeprom_reg lp8860_eeprom_disp_regs[] = {
+	{ LP8860_EEPROM_REG_0, 0xed },
+	{ LP8860_EEPROM_REG_1, 0xdf },
+	{ LP8860_EEPROM_REG_2, 0xdc },
+	{ LP8860_EEPROM_REG_3, 0xf0 },
+	{ LP8860_EEPROM_REG_4, 0xdf },
+	{ LP8860_EEPROM_REG_5, 0xe5 },
+	{ LP8860_EEPROM_REG_6, 0xf2 },
+	{ LP8860_EEPROM_REG_7, 0x77 },
+	{ LP8860_EEPROM_REG_8, 0x77 },
+	{ LP8860_EEPROM_REG_9, 0x71 },
+	{ LP8860_EEPROM_REG_10, 0x3f },
+	{ LP8860_EEPROM_REG_11, 0xb7 },
+	{ LP8860_EEPROM_REG_12, 0x17 },
+	{ LP8860_EEPROM_REG_13, 0xef },
+	{ LP8860_EEPROM_REG_14, 0xb0 },
+	{ LP8860_EEPROM_REG_15, 0x87 },
+	{ LP8860_EEPROM_REG_16, 0xce },
+	{ LP8860_EEPROM_REG_17, 0x72 },
+	{ LP8860_EEPROM_REG_18, 0xe5 },
+	{ LP8860_EEPROM_REG_19, 0xdf },
+	{ LP8860_EEPROM_REG_20, 0x35 },
+	{ LP8860_EEPROM_REG_21, 0x06 },
+	{ LP8860_EEPROM_REG_22, 0xdc },
+	{ LP8860_EEPROM_REG_23, 0x88 },
+	{ LP8860_EEPROM_REG_24, 0x3E },
+};
+
+static int lp8860_unlock_eeprom(struct lp8860_led *led, int lock)
+{
+	int ret;
+
+	mutex_lock(&led->lock);
+
+	if (lock == LP8860_UNLOCK_EEPROM) {
+		ret = regmap_write(led->regmap,
+			LP8860_EEPROM_UNLOCK,
+			LP8860_EEPROM_CODE_1);
+		if (ret) {
+			dev_err(&led->client->dev, "EEPROM Unlock failed\n");
+			goto out;
+		}
+
+		ret = regmap_write(led->regmap,
+			LP8860_EEPROM_UNLOCK,
+			LP8860_EEPROM_CODE_2);
+		if (ret) {
+			dev_err(&led->client->dev, "EEPROM Unlock failed\n");
+			goto out;
+		}
+		ret = regmap_write(led->regmap,
+			LP8860_EEPROM_UNLOCK,
+			LP8860_EEPROM_CODE_3);
+		if (ret) {
+			dev_err(&led->client->dev, "EEPROM Unlock failed\n");
+			goto out;
+		}
+	} else {
+		ret = regmap_write(led->regmap,
+			LP8860_EEPROM_UNLOCK,
+			LP8860_LOCK_EEPROM);
+	}
+
+out:
+	mutex_unlock(&led->lock);
+	return ret;
+}
+
+static int lp8860_fault_check(struct lp8860_led *led)
+{
+	int ret, fault;
+	unsigned int read_buf;
+
+	ret = regmap_read(led->regmap, LP8860_LED_FAULT, &read_buf);
+	if (ret)
+		goto out;
+
+	fault = read_buf;
+
+	ret = regmap_read(led->regmap, LP8860_FAULT, &read_buf);
+	if (ret)
+		goto out;
+
+	fault |= read_buf;
+
+	/* Attempt to clear any faults */
+	if (fault)
+		ret = regmap_write(led->regmap, LP8860_FAULT_CLEAR,
+			LP8860_CLEAR_FAULTS);
+out:
+	return ret;
+}
+
+static void lp8860_led_brightness_work(struct work_struct *work)
+{
+	struct lp8860_led *led = container_of(work, struct lp8860_led, work);
+	int ret;
+	int disp_brightness = led->brightness * 255;
+
+	mutex_lock(&led->lock);
+
+	ret = lp8860_fault_check(led);
+	if (ret) {
+		dev_err(&led->client->dev, "Cannot read/clear faults\n");
+		goto out;
+	}
+
+	ret = regmap_write(led->regmap, LP8860_DISP_CL1_BRT_MSB,
+			(disp_brightness & 0xff00) >> 8);
+	if (ret) {
+		dev_err(&led->client->dev, "Cannot write CL1 MSB\n");
+		goto out;
+	}
+
+	ret = regmap_write(led->regmap, LP8860_DISP_CL1_BRT_LSB,
+			disp_brightness & 0xff);
+	if (ret) {
+		dev_err(&led->client->dev, "Cannot write CL1 LSB\n");
+		goto out;
+	}
+out:
+	mutex_unlock(&led->lock);
+}
+
+static void lp8860_brightness_set(struct led_classdev *led_cdev,
+				enum led_brightness brt_val)
+{
+	struct lp8860_led *led =
+			container_of(led_cdev, struct lp8860_led, led_dev);
+
+	led->brightness = brt_val;
+	schedule_work(&led->work);
+}
+
+static int lp8860_init(struct lp8860_led *led)
+{
+	unsigned int read_buf;
+	int ret, i, reg_count;
+
+	if (led->enable_gpio)
+		gpiod_direction_output(led->enable_gpio, 1);
+
+	ret = lp8860_fault_check(led);
+	if (ret)
+		goto out;
+
+	ret = regmap_read(led->regmap, LP8860_STATUS, &read_buf);
+	if (ret)
+		goto out;
+
+	ret = lp8860_unlock_eeprom(led, LP8860_UNLOCK_EEPROM);
+	if (ret) {
+		dev_err(&led->client->dev, "Failed unlocking EEPROM\n");
+		goto out;
+	}
+
+	reg_count = ARRAY_SIZE(lp8860_eeprom_disp_regs) / sizeof(lp8860_eeprom_disp_regs[0]);
+	for (i = 0; i < reg_count; i++) {
+		ret = regmap_write(led->eeprom_regmap,
+				lp8860_eeprom_disp_regs[i].reg,
+				lp8860_eeprom_disp_regs[i].value);
+		if (ret) {
+			dev_err(&led->client->dev, "Failed writing EEPROM\n");
+			goto out;
+		}
+	}
+
+	ret = lp8860_unlock_eeprom(led, LP8860_LOCK_EEPROM);
+	if (ret)
+		goto out;
+
+	ret = regmap_write(led->regmap,
+			LP8860_EEPROM_CNTRL,
+			LP8860_PROGRAM_EEPROM);
+	if (ret)
+		dev_err(&led->client->dev, "Failed programming EEPROM\n");
+out:
+	if (ret)
+		if (led->enable_gpio)
+			gpiod_direction_output(led->enable_gpio, 0);
+	return ret;
+}
+
+static struct reg_default lp8860_reg_defs[] = {
+	{ LP8860_DISP_CL1_BRT_MSB, 0x00},
+	{ LP8860_DISP_CL1_BRT_LSB, 0x00},
+	{ LP8860_DISP_CL1_CURR_MSB, 0x00},
+	{ LP8860_DISP_CL1_CURR_LSB, 0x00},
+	{ LP8860_CL2_BRT_MSB, 0x00},
+	{ LP8860_CL2_BRT_LSB, 0x00},
+	{ LP8860_CL2_CURRENT, 0x00},
+	{ LP8860_CL3_BRT_MSB, 0x00},
+	{ LP8860_CL3_BRT_LSB, 0x00},
+	{ LP8860_CL3_CURRENT, 0x00},
+	{ LP8860_CL4_BRT_MSB, 0x00},
+	{ LP8860_CL4_BRT_LSB, 0x00},
+	{ LP8860_CL4_CURRENT, 0x00},
+	{ LP8860_CONFIG, 0x00},
+	{ LP8860_FAULT_CLEAR, 0x00},
+	{ LP8860_EEPROM_CNTRL, 0x80},
+	{ LP8860_EEPROM_UNLOCK, 0x00},
+};
+
+static const struct regmap_config lp8860_regmap_config = {
+	.reg_bits = 8,
+	.val_bits = 8,
+
+	.max_register = LP8860_EEPROM_UNLOCK,
+	.reg_defaults = lp8860_reg_defs,
+	.num_reg_defaults = ARRAY_SIZE(lp8860_reg_defs),
+	.cache_type = REGCACHE_NONE,
+};
+
+static struct reg_default lp8860_eeprom_defs[] = {
+	{ LP8860_EEPROM_REG_0, 0x00 },
+	{ LP8860_EEPROM_REG_1, 0x00 },
+	{ LP8860_EEPROM_REG_2, 0x00 },
+	{ LP8860_EEPROM_REG_3, 0x00 },
+	{ LP8860_EEPROM_REG_4, 0x00 },
+	{ LP8860_EEPROM_REG_5, 0x00 },
+	{ LP8860_EEPROM_REG_6, 0x00 },
+	{ LP8860_EEPROM_REG_7, 0x00 },
+	{ LP8860_EEPROM_REG_8, 0x00 },
+	{ LP8860_EEPROM_REG_9, 0x00 },
+	{ LP8860_EEPROM_REG_10, 0x00 },
+	{ LP8860_EEPROM_REG_11, 0x00 },
+	{ LP8860_EEPROM_REG_12, 0x00 },
+	{ LP8860_EEPROM_REG_13, 0x00 },
+	{ LP8860_EEPROM_REG_14, 0x00 },
+	{ LP8860_EEPROM_REG_15, 0x00 },
+	{ LP8860_EEPROM_REG_16, 0x00 },
+	{ LP8860_EEPROM_REG_17, 0x00 },
+	{ LP8860_EEPROM_REG_18, 0x00 },
+	{ LP8860_EEPROM_REG_19, 0x00 },
+	{ LP8860_EEPROM_REG_20, 0x00 },
+	{ LP8860_EEPROM_REG_21, 0x00 },
+	{ LP8860_EEPROM_REG_22, 0x00 },
+	{ LP8860_EEPROM_REG_23, 0x00 },
+	{ LP8860_EEPROM_REG_24, 0x00 },
+};
+
+static const struct regmap_config lp8860_eeprom_regmap_config = {
+	.reg_bits = 8,
+	.val_bits = 8,
+
+	.max_register = LP8860_EEPROM_REG_24,
+	.reg_defaults = lp8860_eeprom_defs,
+	.num_reg_defaults = ARRAY_SIZE(lp8860_eeprom_defs),
+	.cache_type = REGCACHE_NONE,
+};
+
+static int lp8860_probe(struct i2c_client *client,
+			const struct i2c_device_id *id)
+{
+	int ret;
+	struct lp8860_led *led;
+	struct device_node *np = client->dev.of_node;
+
+	led = devm_kzalloc(&client->dev, sizeof(*led), GFP_KERNEL);
+	if (!led)
+		return -ENOMEM;
+
+	led->label = LP8860_DISP_LED_NAME;
+
+	if (client->dev.of_node) {
+		ret = of_property_read_string(np, "label", &led->label);
+		if (ret) {
+			dev_err(&client->dev, "Missing label in dt\n");
+			return -EINVAL;
+		}
+	}
+
+	led->enable_gpio = devm_gpiod_get(&client->dev, "enable");
+	if (IS_ERR(led->enable_gpio))
+		led->enable_gpio = NULL;
+	else
+		gpiod_direction_output(led->enable_gpio, 0);
+
+	led->regulator = devm_regulator_get(&client->dev, "vled");
+	if (IS_ERR(led->regulator))
+		led->regulator = NULL;
+
+	led->client = client;
+	led->led_dev.name = led->label;
+	led->led_dev.max_brightness = LED_FULL;
+	led->led_dev.brightness_set = lp8860_brightness_set;
+
+	mutex_init(&led->lock);
+	INIT_WORK(&led->work, lp8860_led_brightness_work);
+
+	i2c_set_clientdata(client, led);
+
+	led->regmap = devm_regmap_init_i2c(client, &lp8860_regmap_config);
+	if (IS_ERR(led->regmap)) {
+		ret = PTR_ERR(led->regmap);
+		dev_err(&client->dev, "Failed to allocate register map: %d\n",
+			ret);
+		return ret;
+	}
+
+	led->eeprom_regmap = devm_regmap_init_i2c(client, &lp8860_eeprom_regmap_config);
+	if (IS_ERR(led->eeprom_regmap)) {
+		ret = PTR_ERR(led->eeprom_regmap);
+		dev_err(&client->dev, "Failed to allocate register map: %d\n",
+			ret);
+		return ret;
+	}
+
+	ret = lp8860_init(led);
+	if (ret)
+		return ret;
+
+	ret = led_classdev_register(&client->dev, &led->led_dev);
+	if (ret) {
+		dev_err(&client->dev, "led register err: %d\n", ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+static int lp8860_remove(struct i2c_client *client)
+{
+	struct lp8860_led *led = i2c_get_clientdata(client);
+	int ret;
+
+	led_classdev_unregister(&led->led_dev);
+	cancel_work_sync(&led->work);
+
+	if (led->enable_gpio)
+		gpiod_direction_output(led->enable_gpio, 0);
+
+	if (led->regulator) {
+		ret = regulator_disable(led->regulator);
+		if (ret)
+			dev_err(&led->client->dev,
+				"Failed to disable regulator\n");
+	}
+
+	return 0;
+}
+
+static const struct i2c_device_id lp8860_id[] = {
+	{ "lp8860", 0 },
+	{ }
+};
+MODULE_DEVICE_TABLE(i2c, lp8860_id);
+
+#ifdef CONFIG_OF
+static const struct of_device_id of_lp8860_leds_match[] = {
+	{ .compatible = "ti,lp8860", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, of_lp8860_leds_match);
+#endif
+
+static struct i2c_driver lp8860_driver = {
+	.driver = {
+		.name	= "lp8860",
+		.of_match_table = of_match_ptr(of_lp8860_leds_match),
+	},
+	.probe		= lp8860_probe,
+	.remove		= lp8860_remove,
+	.id_table	= lp8860_id,
+};
+module_i2c_driver(lp8860_driver);
+
+MODULE_DESCRIPTION("Texas Instruments LP8860 LED drvier");
+MODULE_AUTHOR("Dan Murphy <dmurphy@ti.com>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/leds/leds-lt3593.c b/drivers/leds/leds-lt3593.c
index 059f5b1f3553..9f41124765cc 100644
--- a/drivers/leds/leds-lt3593.c
+++ b/drivers/leds/leds-lt3593.c
@@ -184,7 +184,6 @@ static struct platform_driver lt3593_led_driver = {
 	.remove		= lt3593_led_remove,
 	.driver		= {
 		.name	= "leds-lt3593",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/leds/leds-max8997.c b/drivers/leds/leds-max8997.c
index 607bc2755aba..c592aa5662bb 100644
--- a/drivers/leds/leds-max8997.c
+++ b/drivers/leds/leds-max8997.c
@@ -303,7 +303,6 @@ static int max8997_led_remove(struct platform_device *pdev)
 static struct platform_driver max8997_led_driver = {
 	.driver = {
 		.name  = "max8997-led",
-		.owner = THIS_MODULE,
 	},
 	.probe  = max8997_led_probe,
 	.remove = max8997_led_remove,
diff --git a/drivers/leds/leds-mc13783.c b/drivers/leds/leds-mc13783.c
index f1db88e25138..85c3714e1b5a 100644
--- a/drivers/leds/leds-mc13783.c
+++ b/drivers/leds/leds-mc13783.c
@@ -325,7 +325,6 @@ MODULE_DEVICE_TABLE(platform, mc13xxx_led_id_table);
 static struct platform_driver mc13xxx_led_driver = {
 	.driver	= {
 		.name	= "mc13xxx-led",
-		.owner	= THIS_MODULE,
 	},
 	.remove		= mc13xxx_led_remove,
 	.id_table	= mc13xxx_led_id_table,
diff --git a/drivers/leds/leds-menf21bmc.c b/drivers/leds/leds-menf21bmc.c
new file mode 100644
index 000000000000..4b9eea815b1a
--- /dev/null
+++ b/drivers/leds/leds-menf21bmc.c
@@ -0,0 +1,130 @@
+/*
+ *  MEN 14F021P00 Board Management Controller (BMC) LEDs Driver.
+ *
+ *  This is the core LED driver of the MEN 14F021P00 BMC.
+ *  There are four LEDs available which can be switched on and off.
+ *  STATUS LED, HOT SWAP LED, USER LED 1, USER LED 2
+ *
+ *  Copyright (C) 2014 MEN Mikro Elektronik Nuernberg GmbH
+ *
+ *  This program is free software; you can redistribute  it and/or modify it
+ *  under  the terms of  the GNU General  Public License as published by the
+ *  Free Software Foundation;  either version 2 of the  License, or (at your
+ *  option) any later version.
+ */
+
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/platform_device.h>
+#include <linux/leds.h>
+#include <linux/i2c.h>
+
+#define BMC_CMD_LED_GET_SET	0xA0
+#define BMC_BIT_LED_STATUS	BIT(0)
+#define BMC_BIT_LED_HOTSWAP	BIT(1)
+#define BMC_BIT_LED_USER1	BIT(2)
+#define BMC_BIT_LED_USER2	BIT(3)
+
+struct menf21bmc_led {
+	struct led_classdev cdev;
+	u8 led_bit;
+	const char *name;
+	struct i2c_client *i2c_client;
+};
+
+static struct menf21bmc_led leds[] = {
+	{
+		.name = "menf21bmc:led_status",
+		.led_bit = BMC_BIT_LED_STATUS,
+	},
+	{
+		.name = "menf21bmc:led_hotswap",
+		.led_bit = BMC_BIT_LED_HOTSWAP,
+	},
+	{
+		.name = "menf21bmc:led_user1",
+		.led_bit = BMC_BIT_LED_USER1,
+	},
+	{
+		.name = "menf21bmc:led_user2",
+		.led_bit = BMC_BIT_LED_USER2,
+	}
+};
+
+static DEFINE_MUTEX(led_lock);
+
+static void
+menf21bmc_led_set(struct led_classdev *led_cdev, enum led_brightness value)
+{
+	int led_val;
+	struct menf21bmc_led *led = container_of(led_cdev,
+					struct menf21bmc_led, cdev);
+
+	mutex_lock(&led_lock);
+	led_val = i2c_smbus_read_byte_data(led->i2c_client,
+					   BMC_CMD_LED_GET_SET);
+	if (led_val < 0)
+		goto err_out;
+
+	if (value == LED_OFF)
+		led_val &= ~led->led_bit;
+	else
+		led_val |= led->led_bit;
+
+	i2c_smbus_write_byte_data(led->i2c_client,
+				  BMC_CMD_LED_GET_SET, led_val);
+err_out:
+	mutex_unlock(&led_lock);
+}
+
+static int menf21bmc_led_probe(struct platform_device *pdev)
+{
+	int i;
+	int ret;
+	struct i2c_client *i2c_client = to_i2c_client(pdev->dev.parent);
+
+	for (i = 0; i < ARRAY_SIZE(leds); i++) {
+		leds[i].cdev.name = leds[i].name;
+		leds[i].cdev.brightness_set = menf21bmc_led_set;
+		leds[i].i2c_client = i2c_client;
+		ret = led_classdev_register(&pdev->dev, &leds[i].cdev);
+		if (ret < 0)
+			goto err_free_leds;
+	}
+	dev_info(&pdev->dev, "MEN 140F21P00 BMC LED device enabled\n");
+
+	return 0;
+
+err_free_leds:
+	dev_err(&pdev->dev, "failed to register LED device\n");
+
+	for (i = i - 1; i >= 0; i--)
+		led_classdev_unregister(&leds[i].cdev);
+
+	return ret;
+}
+
+static int menf21bmc_led_remove(struct platform_device *pdev)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(leds); i++)
+		led_classdev_unregister(&leds[i].cdev);
+
+	return 0;
+}
+
+static struct platform_driver menf21bmc_led = {
+	.probe		= menf21bmc_led_probe,
+	.remove		= menf21bmc_led_remove,
+	.driver		= {
+		.name		= "menf21bmc_led",
+	},
+};
+
+module_platform_driver(menf21bmc_led);
+
+MODULE_AUTHOR("Andreas Werner <andreas.werner@men.de>");
+MODULE_DESCRIPTION("MEN 14F021P00 BMC led driver");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS("platform:menf21bmc_led");
diff --git a/drivers/leds/leds-net48xx.c b/drivers/leds/leds-net48xx.c
index 27d06c528246..ec3a2e8adcae 100644
--- a/drivers/leds/leds-net48xx.c
+++ b/drivers/leds/leds-net48xx.c
@@ -53,7 +53,6 @@ static struct platform_driver net48xx_led_driver = {
 	.remove		= net48xx_led_remove,
 	.driver		= {
 		.name		= DRVNAME,
-		.owner		= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/leds/leds-netxbig.c b/drivers/leds/leds-netxbig.c
index 64fde485dcaa..25e419752a7b 100644
--- a/drivers/leds/leds-netxbig.c
+++ b/drivers/leds/leds-netxbig.c
@@ -330,18 +330,18 @@ create_netxbig_led(struct platform_device *pdev,
 	led_dat->sata = 0;
 	led_dat->cdev.brightness = LED_OFF;
 	led_dat->cdev.flags |= LED_CORE_SUSPENDRESUME;
-	/*
-	 * If available, expose the SATA activity blink capability through
-	 * a "sata" sysfs attribute.
-	 */
-	if (led_dat->mode_val[NETXBIG_LED_SATA] != NETXBIG_LED_INVALID_MODE)
-		led_dat->cdev.groups = netxbig_led_groups;
 	led_dat->mode_addr = template->mode_addr;
 	led_dat->mode_val = template->mode_val;
 	led_dat->bright_addr = template->bright_addr;
 	led_dat->bright_max = (1 << pdata->gpio_ext->num_data) - 1;
 	led_dat->timer = pdata->timer;
 	led_dat->num_timer = pdata->num_timer;
+	/*
+	 * If available, expose the SATA activity blink capability through
+	 * a "sata" sysfs attribute.
+	 */
+	if (led_dat->mode_val[NETXBIG_LED_SATA] != NETXBIG_LED_INVALID_MODE)
+		led_dat->cdev.groups = netxbig_led_groups;
 
 	return led_classdev_register(&pdev->dev, &led_dat->cdev);
 }
@@ -404,7 +404,6 @@ static struct platform_driver netxbig_led_driver = {
 	.remove		= netxbig_led_remove,
 	.driver		= {
 		.name	= "leds-netxbig",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/leds/leds-ns2.c b/drivers/leds/leds-ns2.c
index 231993d1fe21..1fd6adbb43b7 100644
--- a/drivers/leds/leds-ns2.c
+++ b/drivers/leds/leds-ns2.c
@@ -380,7 +380,6 @@ static struct platform_driver ns2_led_driver = {
 	.remove		= ns2_led_remove,
 	.driver		= {
 		.name		= "leds-ns2",
-		.owner		= THIS_MODULE,
 		.of_match_table	= of_match_ptr(of_ns2_leds_match),
 	},
 };
diff --git a/drivers/leds/leds-ot200.c b/drivers/leds/leds-ot200.c
index c9d906098466..39870de20a26 100644
--- a/drivers/leds/leds-ot200.c
+++ b/drivers/leds/leds-ot200.c
@@ -158,7 +158,6 @@ static struct platform_driver ot200_led_driver = {
 	.remove		= ot200_led_remove,
 	.driver		= {
 		.name	= "leds-ot200",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/leds/leds-pca9532.c b/drivers/leds/leds-pca9532.c
index 4a0e786b7832..5a6363d161a2 100644
--- a/drivers/leds/leds-pca9532.c
+++ b/drivers/leds/leds-pca9532.c
@@ -319,14 +319,8 @@ static int pca9532_destroy_devices(struct pca9532_data *data, int n_devs)
 	}
 
 #ifdef CONFIG_LEDS_PCA9532_GPIO
-	if (data->gpio.dev) {
-		int err = gpiochip_remove(&data->gpio);
-		if (err) {
-			dev_err(&data->client->dev, "%s failed, %d\n",
-						"gpiochip_remove()", err);
-			return err;
-		}
-	}
+	if (data->gpio.dev)
+		gpiochip_remove(&data->gpio);
 #endif
 
 	return 0;
diff --git a/drivers/leds/leds-pwm.c b/drivers/leds/leds-pwm.c
index d672bb4480f6..f668500a2157 100644
--- a/drivers/leds/leds-pwm.c
+++ b/drivers/leds/leds-pwm.c
@@ -232,7 +232,6 @@ static struct platform_driver led_pwm_driver = {
 	.remove		= led_pwm_remove,
 	.driver		= {
 		.name	= "leds_pwm",
-		.owner	= THIS_MODULE,
 		.of_match_table = of_pwm_leds_match,
 	},
 };
diff --git a/drivers/leds/leds-rb532.c b/drivers/leds/leds-rb532.c
index 2e746d257b02..fcd1215b64a2 100644
--- a/drivers/leds/leds-rb532.c
+++ b/drivers/leds/leds-rb532.c
@@ -53,7 +53,6 @@ static struct platform_driver rb532_led_driver = {
 	.remove = rb532_led_remove,
 	.driver = {
 		.name = "rb532-led",
-		.owner = THIS_MODULE,
 	},
 };
 
diff --git a/drivers/leds/leds-regulator.c b/drivers/leds/leds-regulator.c
index 358430db6e66..ffc21397a675 100644
--- a/drivers/leds/leds-regulator.c
+++ b/drivers/leds/leds-regulator.c
@@ -153,24 +153,21 @@ static int regulator_led_probe(struct platform_device *pdev)
 		return -ENODEV;
 	}
 
-	vcc = regulator_get_exclusive(&pdev->dev, "vled");
+	vcc = devm_regulator_get_exclusive(&pdev->dev, "vled");
 	if (IS_ERR(vcc)) {
 		dev_err(&pdev->dev, "Cannot get vcc for %s\n", pdata->name);
 		return PTR_ERR(vcc);
 	}
 
 	led = devm_kzalloc(&pdev->dev, sizeof(*led), GFP_KERNEL);
-	if (led == NULL) {
-		ret = -ENOMEM;
-		goto err_vcc;
-	}
+	if (led == NULL)
+		return -ENOMEM;
 
 	led->cdev.max_brightness = led_regulator_get_max_brightness(vcc);
 	if (pdata->brightness > led->cdev.max_brightness) {
 		dev_err(&pdev->dev, "Invalid default brightness %d\n",
 				pdata->brightness);
-		ret = -EINVAL;
-		goto err_vcc;
+		return -EINVAL;
 	}
 	led->value = pdata->brightness;
 
@@ -191,7 +188,7 @@ static int regulator_led_probe(struct platform_device *pdev)
 	ret = led_classdev_register(&pdev->dev, &led->cdev);
 	if (ret < 0) {
 		cancel_work_sync(&led->work);
-		goto err_vcc;
+		return ret;
 	}
 
 	/* to expose the default value to userspace */
@@ -201,10 +198,6 @@ static int regulator_led_probe(struct platform_device *pdev)
 	regulator_led_set_value(led);
 
 	return 0;
-
-err_vcc:
-	regulator_put(vcc);
-	return ret;
 }
 
 static int regulator_led_remove(struct platform_device *pdev)
@@ -214,14 +207,12 @@ static int regulator_led_remove(struct platform_device *pdev)
 	led_classdev_unregister(&led->cdev);
 	cancel_work_sync(&led->work);
 	regulator_led_disable(led);
-	regulator_put(led->vcc);
 	return 0;
 }
 
 static struct platform_driver regulator_led_driver = {
 	.driver = {
 		   .name  = "leds-regulator",
-		   .owner = THIS_MODULE,
 		   },
 	.probe  = regulator_led_probe,
 	.remove = regulator_led_remove,
diff --git a/drivers/leds/leds-s3c24xx.c b/drivers/leds/leds-s3c24xx.c
index 785eb53a87fc..83641a7b299a 100644
--- a/drivers/leds/leds-s3c24xx.c
+++ b/drivers/leds/leds-s3c24xx.c
@@ -116,7 +116,6 @@ static struct platform_driver s3c24xx_led_driver = {
 	.remove		= s3c24xx_led_remove,
 	.driver		= {
 		.name		= "s3c24xx_led",
-		.owner		= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/leds/leds-sunfire.c b/drivers/leds/leds-sunfire.c
index 0b8cc4a021a6..c2553c54f2cf 100644
--- a/drivers/leds/leds-sunfire.c
+++ b/drivers/leds/leds-sunfire.c
@@ -223,7 +223,6 @@ static struct platform_driver sunfire_clockboard_led_driver = {
 	.remove		= sunfire_led_generic_remove,
 	.driver		= {
 		.name	= "sunfire-clockboard-leds",
-		.owner	= THIS_MODULE,
 	},
 };
 
@@ -232,7 +231,6 @@ static struct platform_driver sunfire_fhc_led_driver = {
 	.remove		= sunfire_led_generic_remove,
 	.driver		= {
 		.name	= "sunfire-fhc-leds",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/leds/leds-syscon.c b/drivers/leds/leds-syscon.c
new file mode 100644
index 000000000000..6896e2d9ba58
--- /dev/null
+++ b/drivers/leds/leds-syscon.c
@@ -0,0 +1,167 @@
+/*
+ * Generic Syscon LEDs Driver
+ *
+ * Copyright (c) 2014, Linaro Limited
+ * Author: Linus Walleij <linus.walleij@linaro.org>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+ * MA 02111-1307 USA
+ */
+#include <linux/io.h>
+#include <linux/of_device.h>
+#include <linux/of_address.h>
+#include <linux/platform_device.h>
+#include <linux/stat.h>
+#include <linux/slab.h>
+#include <linux/mfd/syscon.h>
+#include <linux/regmap.h>
+#include <linux/leds.h>
+
+/**
+ * struct syscon_led - state container for syscon based LEDs
+ * @cdev: LED class device for this LED
+ * @map: regmap to access the syscon device backing this LED
+ * @offset: the offset into the syscon regmap for the LED register
+ * @mask: the bit in the register corresponding to the LED
+ * @state: current state of the LED
+ */
+struct syscon_led {
+	struct led_classdev cdev;
+	struct regmap *map;
+	u32 offset;
+	u32 mask;
+	bool state;
+};
+
+static void syscon_led_set(struct led_classdev *led_cdev,
+	enum led_brightness value)
+{
+	struct syscon_led *sled =
+		container_of(led_cdev, struct syscon_led, cdev);
+	u32 val;
+	int ret;
+
+	if (value == LED_OFF) {
+		val = 0;
+		sled->state = false;
+	} else {
+		val = sled->mask;
+		sled->state = true;
+	}
+
+	ret = regmap_update_bits(sled->map, sled->offset, sled->mask, val);
+	if (ret < 0)
+		dev_err(sled->cdev.dev, "error updating LED status\n");
+}
+
+static int __init syscon_leds_spawn(struct device_node *np,
+				    struct device *dev,
+				    struct regmap *map)
+{
+	struct device_node *child;
+	int ret;
+
+	for_each_available_child_of_node(np, child) {
+		struct syscon_led *sled;
+		const char *state;
+
+		/* Only check for register-bit-leds */
+		if (of_property_match_string(child, "compatible",
+					     "register-bit-led") < 0)
+			continue;
+
+		sled = devm_kzalloc(dev, sizeof(*sled), GFP_KERNEL);
+		if (!sled)
+			return -ENOMEM;
+
+		sled->map = map;
+
+		if (of_property_read_u32(child, "offset", &sled->offset))
+			return -EINVAL;
+		if (of_property_read_u32(child, "mask", &sled->mask))
+			return -EINVAL;
+		sled->cdev.name =
+			of_get_property(child, "label", NULL) ? : child->name;
+		sled->cdev.default_trigger =
+			of_get_property(child, "linux,default-trigger", NULL);
+
+		state = of_get_property(child, "default-state", NULL);
+		if (state) {
+			if (!strcmp(state, "keep")) {
+				u32 val;
+
+				ret = regmap_read(map, sled->offset, &val);
+				if (ret < 0)
+					return ret;
+				sled->state = !!(val & sled->mask);
+			} else if (!strcmp(state, "on")) {
+				sled->state = true;
+				ret = regmap_update_bits(map, sled->offset,
+							 sled->mask,
+							 sled->mask);
+				if (ret < 0)
+					return ret;
+			} else {
+				sled->state = false;
+				ret = regmap_update_bits(map, sled->offset,
+							 sled->mask, 0);
+				if (ret < 0)
+					return ret;
+			}
+		}
+		sled->cdev.brightness_set = syscon_led_set;
+
+		ret = led_classdev_register(dev, &sled->cdev);
+		if (ret < 0)
+			return ret;
+
+		dev_info(dev, "registered LED %s\n", sled->cdev.name);
+	}
+	return 0;
+}
+
+static int __init syscon_leds_init(void)
+{
+	struct device_node *np;
+
+	for_each_of_allnodes(np) {
+		struct platform_device *pdev;
+		struct regmap *map;
+		int ret;
+
+		if (!of_device_is_compatible(np, "syscon"))
+			continue;
+
+		map = syscon_node_to_regmap(np);
+		if (IS_ERR(map)) {
+			pr_err("error getting regmap for syscon LEDs\n");
+			continue;
+		}
+
+		/*
+		 * If the map is there, the device should be there, we allocate
+		 * memory on the syscon device's behalf here.
+		 */
+		pdev = of_find_device_by_node(np);
+		if (!pdev)
+			return -ENODEV;
+		ret = syscon_leds_spawn(np, &pdev->dev, map);
+		if (ret)
+			dev_err(&pdev->dev, "could not spawn syscon LEDs\n");
+	}
+
+	return 0;
+}
+device_initcall(syscon_leds_init);
diff --git a/drivers/leds/leds-tca6507.c b/drivers/leds/leds-tca6507.c
index 3d9e267a56c4..20fa8e77f186 100644
--- a/drivers/leds/leds-tca6507.c
+++ b/drivers/leds/leds-tca6507.c
@@ -667,11 +667,8 @@ static int tca6507_probe_gpios(struct i2c_client *client,
 
 static void tca6507_remove_gpio(struct tca6507_chip *tca)
 {
-	if (tca->gpio.ngpio) {
-		int err = gpiochip_remove(&tca->gpio);
-		dev_err(&tca->client->dev, "%s failed, %d\n",
-			"gpiochip_remove()", err);
-	}
+	if (tca->gpio.ngpio)
+		gpiochip_remove(&tca->gpio);
 }
 #else /* CONFIG_GPIOLIB */
 static int tca6507_probe_gpios(struct i2c_client *client,
diff --git a/drivers/leds/leds-wm831x-status.c b/drivers/leds/leds-wm831x-status.c
index 1b71e0701002..56027ef7c7e8 100644
--- a/drivers/leds/leds-wm831x-status.c
+++ b/drivers/leds/leds-wm831x-status.c
@@ -312,7 +312,6 @@ static int wm831x_status_remove(struct platform_device *pdev)
 static struct platform_driver wm831x_status_driver = {
 	.driver = {
 		   .name = "wm831x-status",
-		   .owner = THIS_MODULE,
 		   },
 	.probe = wm831x_status_probe,
 	.remove = wm831x_status_remove,
diff --git a/drivers/leds/leds-wm8350.c b/drivers/leds/leds-wm8350.c
index 4133ffe29015..0d121835673f 100644
--- a/drivers/leds/leds-wm8350.c
+++ b/drivers/leds/leds-wm8350.c
@@ -272,7 +272,6 @@ static int wm8350_led_remove(struct platform_device *pdev)
 static struct platform_driver wm8350_led_driver = {
 	.driver = {
 		   .name = "wm8350-led",
-		   .owner = THIS_MODULE,
 		   },
 	.probe = wm8350_led_probe,
 	.remove = wm8350_led_remove,
diff --git a/drivers/leds/leds-wrap.c b/drivers/leds/leds-wrap.c
index b358cc05eff5..1ba3defdd460 100644
--- a/drivers/leds/leds-wrap.c
+++ b/drivers/leds/leds-wrap.c
@@ -111,7 +111,6 @@ static struct platform_driver wrap_led_driver = {
 	.remove		= wrap_led_remove,
 	.driver		= {
 		.name		= DRVNAME,
-		.owner		= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/leds/leds.h b/drivers/leds/leds.h
index 4c50365344a9..2348dbda5269 100644
--- a/drivers/leds/leds.h
+++ b/drivers/leds/leds.h
@@ -17,16 +17,28 @@
 #include <linux/rwsem.h>
 #include <linux/leds.h>
 
-static inline void __led_set_brightness(struct led_classdev *led_cdev,
+static inline void led_set_brightness_async(struct led_classdev *led_cdev,
 					enum led_brightness value)
 {
-	if (value > led_cdev->max_brightness)
-		value = led_cdev->max_brightness;
-	led_cdev->brightness = value;
+	led_cdev->brightness = min(value, led_cdev->max_brightness);
+
 	if (!(led_cdev->flags & LED_SUSPENDED))
 		led_cdev->brightness_set(led_cdev, value);
 }
 
+static inline int led_set_brightness_sync(struct led_classdev *led_cdev,
+					enum led_brightness value)
+{
+	int ret = 0;
+
+	led_cdev->brightness = min(value, led_cdev->max_brightness);
+
+	if (!(led_cdev->flags & LED_SUSPENDED))
+		ret = led_cdev->brightness_set_sync(led_cdev,
+						led_cdev->brightness);
+	return ret;
+}
+
 static inline int led_get_brightness(struct led_classdev *led_cdev)
 {
 	return led_cdev->brightness;
diff --git a/drivers/leds/trigger/ledtrig-backlight.c b/drivers/leds/trigger/ledtrig-backlight.c
index 47e55aa9eefa..59eca17d9661 100644
--- a/drivers/leds/trigger/ledtrig-backlight.c
+++ b/drivers/leds/trigger/ledtrig-backlight.c
@@ -51,9 +51,9 @@ static int fb_notifier_callback(struct notifier_block *p,
 
 	if ((n->old_status == UNBLANK) ^ n->invert) {
 		n->brightness = led->brightness;
-		__led_set_brightness(led, LED_OFF);
+		led_set_brightness_async(led, LED_OFF);
 	} else {
-		__led_set_brightness(led, n->brightness);
+		led_set_brightness_async(led, n->brightness);
 	}
 
 	n->old_status = new_status;
@@ -89,9 +89,9 @@ static ssize_t bl_trig_invert_store(struct device *dev,
 
 	/* After inverting, we need to update the LED. */
 	if ((n->old_status == BLANK) ^ n->invert)
-		__led_set_brightness(led, LED_OFF);
+		led_set_brightness_async(led, LED_OFF);
 	else
-		__led_set_brightness(led, n->brightness);
+		led_set_brightness_async(led, n->brightness);
 
 	return num;
 }
diff --git a/drivers/leds/trigger/ledtrig-default-on.c b/drivers/leds/trigger/ledtrig-default-on.c
index 81a91be8e18d..6f38f883aaf1 100644
--- a/drivers/leds/trigger/ledtrig-default-on.c
+++ b/drivers/leds/trigger/ledtrig-default-on.c
@@ -19,7 +19,7 @@
 
 static void defon_trig_activate(struct led_classdev *led_cdev)
 {
-	__led_set_brightness(led_cdev, led_cdev->max_brightness);
+	led_set_brightness_async(led_cdev, led_cdev->max_brightness);
 }
 
 static struct led_trigger defon_led_trigger = {
diff --git a/drivers/leds/trigger/ledtrig-gpio.c b/drivers/leds/trigger/ledtrig-gpio.c
index 35812e3a37f2..4cc7040746c6 100644
--- a/drivers/leds/trigger/ledtrig-gpio.c
+++ b/drivers/leds/trigger/ledtrig-gpio.c
@@ -48,18 +48,18 @@ static void gpio_trig_work(struct work_struct *work)
 	if (!gpio_data->gpio)
 		return;
 
-	tmp = gpio_get_value(gpio_data->gpio);
+	tmp = gpio_get_value_cansleep(gpio_data->gpio);
 	if (gpio_data->inverted)
 		tmp = !tmp;
 
 	if (tmp) {
 		if (gpio_data->desired_brightness)
-			__led_set_brightness(gpio_data->led,
+			led_set_brightness_async(gpio_data->led,
 					   gpio_data->desired_brightness);
 		else
-			__led_set_brightness(gpio_data->led, LED_FULL);
+			led_set_brightness_async(gpio_data->led, LED_FULL);
 	} else {
-		__led_set_brightness(gpio_data->led, LED_OFF);
+		led_set_brightness_async(gpio_data->led, LED_OFF);
 	}
 }
 
diff --git a/drivers/leds/trigger/ledtrig-heartbeat.c b/drivers/leds/trigger/ledtrig-heartbeat.c
index 5c8464a33172..fea6871d2609 100644
--- a/drivers/leds/trigger/ledtrig-heartbeat.c
+++ b/drivers/leds/trigger/ledtrig-heartbeat.c
@@ -74,7 +74,7 @@ static void led_heartbeat_function(unsigned long data)
 		break;
 	}
 
-	__led_set_brightness(led_cdev, brightness);
+	led_set_brightness_async(led_cdev, brightness);
 	mod_timer(&heartbeat_data->timer, jiffies + delay);
 }
 
diff --git a/drivers/leds/trigger/ledtrig-oneshot.c b/drivers/leds/trigger/ledtrig-oneshot.c
index cb4c7466692a..fbd02cdc3ad7 100644
--- a/drivers/leds/trigger/ledtrig-oneshot.c
+++ b/drivers/leds/trigger/ledtrig-oneshot.c
@@ -63,9 +63,9 @@ static ssize_t led_invert_store(struct device *dev,
 	oneshot_data->invert = !!state;
 
 	if (oneshot_data->invert)
-		__led_set_brightness(led_cdev, LED_FULL);
+		led_set_brightness_async(led_cdev, LED_FULL);
 	else
-		__led_set_brightness(led_cdev, LED_OFF);
+		led_set_brightness_async(led_cdev, LED_OFF);
 
 	return size;
 }
diff --git a/drivers/leds/trigger/ledtrig-transient.c b/drivers/leds/trigger/ledtrig-transient.c
index e5abc00bb00c..3c34de404d18 100644
--- a/drivers/leds/trigger/ledtrig-transient.c
+++ b/drivers/leds/trigger/ledtrig-transient.c
@@ -41,7 +41,7 @@ static void transient_timer_function(unsigned long data)
 	struct transient_trig_data *transient_data = led_cdev->trigger_data;
 
 	transient_data->activate = 0;
-	__led_set_brightness(led_cdev, transient_data->restore_state);
+	led_set_brightness_async(led_cdev, transient_data->restore_state);
 }
 
 static ssize_t transient_activate_show(struct device *dev,
@@ -72,7 +72,8 @@ static ssize_t transient_activate_store(struct device *dev,
 	if (state == 0 && transient_data->activate == 1) {
 		del_timer(&transient_data->timer);
 		transient_data->activate = state;
-		__led_set_brightness(led_cdev, transient_data->restore_state);
+		led_set_brightness_async(led_cdev,
+					transient_data->restore_state);
 		return size;
 	}
 
@@ -80,7 +81,7 @@ static ssize_t transient_activate_store(struct device *dev,
 	if (state == 1 && transient_data->activate == 0 &&
 	    transient_data->duration != 0) {
 		transient_data->activate = state;
-		__led_set_brightness(led_cdev, transient_data->state);
+		led_set_brightness_async(led_cdev, transient_data->state);
 		transient_data->restore_state =
 		    (transient_data->state == LED_FULL) ? LED_OFF : LED_FULL;
 		mod_timer(&transient_data->timer,
@@ -203,7 +204,8 @@ static void transient_trig_deactivate(struct led_classdev *led_cdev)
 
 	if (led_cdev->activated) {
 		del_timer_sync(&transient_data->timer);
-		__led_set_brightness(led_cdev, transient_data->restore_state);
+		led_set_brightness_async(led_cdev,
+					transient_data->restore_state);
 		device_remove_file(led_cdev->dev, &dev_attr_activate);
 		device_remove_file(led_cdev->dev, &dev_attr_duration);
 		device_remove_file(led_cdev->dev, &dev_attr_state);
diff --git a/drivers/lguest/lguest_device.c b/drivers/lguest/lguest_device.c
index d0a1d8a45c81..89088d6538fd 100644
--- a/drivers/lguest/lguest_device.c
+++ b/drivers/lguest/lguest_device.c
@@ -94,7 +94,7 @@ static unsigned desc_size(const struct lguest_device_desc *desc)
 }
 
 /* This gets the device's feature bits. */
-static u32 lg_get_features(struct virtio_device *vdev)
+static u64 lg_get_features(struct virtio_device *vdev)
 {
 	unsigned int i;
 	u32 features = 0;
@@ -126,7 +126,7 @@ static void status_notify(struct virtio_device *vdev)
  * sorted out, this routine is called so we can tell the Host which features we
  * understand and accept.
  */
-static void lg_finalize_features(struct virtio_device *vdev)
+static int lg_finalize_features(struct virtio_device *vdev)
 {
 	unsigned int i, bits;
 	struct lguest_device_desc *desc = to_lgdev(vdev)->desc;
@@ -136,20 +136,25 @@ static void lg_finalize_features(struct virtio_device *vdev)
 	/* Give virtio_ring a chance to accept features. */
 	vring_transport_features(vdev);
 
+	/* Make sure we don't have any features > 32 bits! */
+	BUG_ON((u32)vdev->features != vdev->features);
+
 	/*
-	 * The vdev->feature array is a Linux bitmask: this isn't the same as a
-	 * the simple array of bits used by lguest devices for features.  So we
-	 * do this slow, manual conversion which is completely general.
+	 * Since lguest is currently x86-only, we're little-endian.  That
+	 * means we could just memcpy.  But it's not time critical, and in
+	 * case someone copies this code, we do it the slow, obvious way.
 	 */
 	memset(out_features, 0, desc->feature_len);
 	bits = min_t(unsigned, desc->feature_len, sizeof(vdev->features)) * 8;
 	for (i = 0; i < bits; i++) {
-		if (test_bit(i, vdev->features))
+		if (__virtio_test_bit(vdev, i))
 			out_features[i / 8] |= (1 << (i % 8));
 	}
 
 	/* Tell Host we've finished with this device's feature negotiation */
 	status_notify(vdev);
+
+	return 0;
 }
 
 /* Once they've found a field, getting a copy of it is easy. */
diff --git a/drivers/macintosh/Kconfig b/drivers/macintosh/Kconfig
index 3067d56b11a6..5844b80bd90e 100644
--- a/drivers/macintosh/Kconfig
+++ b/drivers/macintosh/Kconfig
@@ -204,16 +204,6 @@ config THERM_ADT746X
           iBook G4, and the ATI based aluminium PowerBooks, allowing slightly
 	  better fan behaviour by default, and some manual control.
 
-config THERM_PM72
-	tristate "Support for thermal management on PowerMac G5 (AGP)"
-	depends on I2C && I2C_POWERMAC && PPC_PMAC64
-	default n
-	help
-	  This driver provides thermostat and fan control for the desktop
-	  G5 machines.
-
-	  This is deprecated, use windfarm instead.
-
 config WINDFARM
 	tristate "New PowerMac thermal control infrastructure"
 	depends on PPC
diff --git a/drivers/macintosh/Makefile b/drivers/macintosh/Makefile
index d2f0120bc878..383ba920085b 100644
--- a/drivers/macintosh/Makefile
+++ b/drivers/macintosh/Makefile
@@ -25,7 +25,6 @@ obj-$(CONFIG_ADB_IOP)		+= adb-iop.o
 obj-$(CONFIG_ADB_PMU68K)	+= via-pmu68k.o
 obj-$(CONFIG_ADB_MACIO)		+= macio-adb.o
 
-obj-$(CONFIG_THERM_PM72)	+= therm_pm72.o
 obj-$(CONFIG_THERM_WINDTUNNEL)	+= therm_windtunnel.o
 obj-$(CONFIG_THERM_ADT746X)	+= therm_adt746x.o
 obj-$(CONFIG_WINDFARM)	        += windfarm_core.o
diff --git a/drivers/macintosh/adb.c b/drivers/macintosh/adb.c
index 9e9c56758a08..226179b975a0 100644
--- a/drivers/macintosh/adb.c
+++ b/drivers/macintosh/adb.c
@@ -411,6 +411,7 @@ adb_poll(void)
 		return;
 	adb_controller->poll();
 }
+EXPORT_SYMBOL(adb_poll);
 
 static void adb_sync_req_done(struct adb_request *req)
 {
@@ -460,6 +461,7 @@ adb_request(struct adb_request *req, void (*done)(struct adb_request *),
 
 	return rc;
 }
+EXPORT_SYMBOL(adb_request);
 
  /* Ultimately this should return the number of devices with
     the given default id.
@@ -495,6 +497,7 @@ adb_register(int default_id, int handler_id, struct adb_ids *ids,
 	mutex_unlock(&adb_handler_mutex);
 	return ids->nids;
 }
+EXPORT_SYMBOL(adb_register);
 
 int
 adb_unregister(int index)
@@ -516,6 +519,7 @@ adb_unregister(int index)
 	mutex_unlock(&adb_handler_mutex);
 	return ret;
 }
+EXPORT_SYMBOL(adb_unregister);
 
 void
 adb_input(unsigned char *buf, int nb, int autopoll)
@@ -582,6 +586,7 @@ adb_try_handler_change(int address, int new_id)
 	mutex_unlock(&adb_handler_mutex);
 	return ret;
 }
+EXPORT_SYMBOL(adb_try_handler_change);
 
 int
 adb_get_infos(int address, int *original_address, int *handler_id)
diff --git a/drivers/macintosh/smu.c b/drivers/macintosh/smu.c
index 4eab93aa570b..10ae69bcbbd2 100644
--- a/drivers/macintosh/smu.c
+++ b/drivers/macintosh/smu.c
@@ -667,7 +667,6 @@ static struct platform_driver smu_of_platform_driver =
 {
 	.driver = {
 		.name = "smu",
-		.owner = THIS_MODULE,
 		.of_match_table = smu_platform_match,
 	},
 	.probe		= smu_platform_probe,
diff --git a/drivers/macintosh/therm_pm72.c b/drivers/macintosh/therm_pm72.c
deleted file mode 100644
index 97cfc5ac9fd0..000000000000
--- a/drivers/macintosh/therm_pm72.c
+++ /dev/null
@@ -1,2279 +0,0 @@
-/*
- * Device driver for the thermostats & fan controller of  the
- * Apple G5 "PowerMac7,2" desktop machines.
- *
- * (c) Copyright IBM Corp. 2003-2004
- *
- * Maintained by: Benjamin Herrenschmidt
- *                <benh@kernel.crashing.org>
- * 
- *
- * The algorithm used is the PID control algorithm, used the same
- * way the published Darwin code does, using the same values that
- * are present in the Darwin 7.0 snapshot property lists.
- *
- * As far as the CPUs control loops are concerned, I use the
- * calibration & PID constants provided by the EEPROM,
- * I do _not_ embed any value from the property lists, as the ones
- * provided by Darwin 7.0 seem to always have an older version that
- * what I've seen on the actual computers.
- * It would be interesting to verify that though. Darwin has a
- * version code of 1.0.0d11 for all control loops it seems, while
- * so far, the machines EEPROMs contain a dataset versioned 1.0.0f
- *
- * Darwin doesn't provide source to all parts, some missing
- * bits like the AppleFCU driver or the actual scale of some
- * of the values returned by sensors had to be "guessed" some
- * way... or based on what Open Firmware does.
- *
- * I didn't yet figure out how to get the slots power consumption
- * out of the FCU, so that part has not been implemented yet and
- * the slots fan is set to a fixed 50% PWM, hoping this value is
- * safe enough ...
- *
- * Note: I have observed strange oscillations of the CPU control
- * loop on a dual G5 here. When idle, the CPU exhaust fan tend to
- * oscillates slowly (over several minutes) between the minimum
- * of 300RPMs and approx. 1000 RPMs. I don't know what is causing
- * this, it could be some incorrect constant or an error in the
- * way I ported the algorithm, or it could be just normal. I
- * don't have full understanding on the way Apple tweaked the PID
- * algorithm for the CPU control, it is definitely not a standard
- * implementation...
- *
- * TODO:  - Check MPU structure version/signature
- *        - Add things like /sbin/overtemp for non-critical
- *          overtemp conditions so userland can take some policy
- *          decisions, like slowing down CPUs
- *	  - Deal with fan and i2c failures in a better way
- *	  - Maybe do a generic PID based on params used for
- *	    U3 and Drives ? Definitely need to factor code a bit
- *          better... also make sensor detection more robust using
- *          the device-tree to probe for them
- *        - Figure out how to get the slots consumption and set the
- *          slots fan accordingly
- *
- * History:
- *
- *  Nov. 13, 2003 : 0.5
- *	- First release
- *
- *  Nov. 14, 2003 : 0.6
- *	- Read fan speed from FCU, low level fan routines now deal
- *	  with errors & check fan status, though higher level don't
- *	  do much.
- *	- Move a bunch of definitions to .h file
- *
- *  Nov. 18, 2003 : 0.7
- *	- Fix build on ppc64 kernel
- *	- Move back statics definitions to .c file
- *	- Avoid calling schedule_timeout with a negative number
- *
- *  Dec. 18, 2003 : 0.8
- *	- Fix typo when reading back fan speed on 2 CPU machines
- *
- *  Mar. 11, 2004 : 0.9
- *	- Rework code accessing the ADC chips, make it more robust and
- *	  closer to the chip spec. Also make sure it is configured properly,
- *        I've seen yet unexplained cases where on startup, I would have stale
- *        values in the configuration register
- *	- Switch back to use of target fan speed for PID, thus lowering
- *        pressure on i2c
- *
- *  Oct. 20, 2004 : 1.1
- *	- Add device-tree lookup for fan IDs, should detect liquid cooling
- *        pumps when present
- *	- Enable driver for PowerMac7,3 machines
- *	- Split the U3/Backside cooling on U3 & U3H versions as Darwin does
- *	- Add new CPU cooling algorithm for machines with liquid cooling
- *	- Workaround for some PowerMac7,3 with empty "fan" node in the devtree
- *	- Fix a signed/unsigned compare issue in some PID loops
- *
- *  Mar. 10, 2005 : 1.2
- *	- Add basic support for Xserve G5
- *	- Retrieve pumps min/max from EEPROM image in device-tree (broken)
- *	- Use min/max macros here or there
- *	- Latest darwin updated U3H min fan speed to 20% PWM
- *
- *  July. 06, 2006 : 1.3
- *	- Fix setting of RPM fans on Xserve G5 (they were going too fast)
- *      - Add missing slots fan control loop for Xserve G5
- *	- Lower fixed slots fan speed from 50% to 40% on desktop G5s. We
- *        still can't properly implement the control loop for these, so let's
- *        reduce the noise a little bit, it appears that 40% still gives us
- *        a pretty good air flow
- *	- Add code to "tickle" the FCU regulary so it doesn't think that
- *        we are gone while in fact, the machine just didn't need any fan
- *        speed change lately
- *
- */
-
-#include <linux/types.h>
-#include <linux/module.h>
-#include <linux/errno.h>
-#include <linux/kernel.h>
-#include <linux/delay.h>
-#include <linux/sched.h>
-#include <linux/init.h>
-#include <linux/spinlock.h>
-#include <linux/wait.h>
-#include <linux/reboot.h>
-#include <linux/kmod.h>
-#include <linux/i2c.h>
-#include <linux/kthread.h>
-#include <linux/mutex.h>
-#include <linux/of_device.h>
-#include <linux/of_platform.h>
-#include <asm/prom.h>
-#include <asm/machdep.h>
-#include <asm/io.h>
-#include <asm/sections.h>
-#include <asm/macio.h>
-
-#include "therm_pm72.h"
-
-#define VERSION "1.3"
-
-#undef DEBUG
-
-#ifdef DEBUG
-#define DBG(args...)	printk(args)
-#else
-#define DBG(args...)	do { } while(0)
-#endif
-
-
-/*
- * Driver statics
- */
-
-static struct platform_device *		of_dev;
-static struct i2c_adapter *		u3_0;
-static struct i2c_adapter *		u3_1;
-static struct i2c_adapter *		k2;
-static struct i2c_client *		fcu;
-static struct cpu_pid_state		processor_state[2];
-static struct basckside_pid_params	backside_params;
-static struct backside_pid_state	backside_state;
-static struct drives_pid_state		drives_state;
-static struct dimm_pid_state		dimms_state;
-static struct slots_pid_state		slots_state;
-static int				state;
-static int				cpu_count;
-static int				cpu_pid_type;
-static struct task_struct		*ctrl_task;
-static struct completion		ctrl_complete;
-static int				critical_state;
-static int				rackmac;
-static s32				dimm_output_clamp;
-static int 				fcu_rpm_shift;
-static int				fcu_tickle_ticks;
-static DEFINE_MUTEX(driver_lock);
-
-/*
- * We have 3 types of CPU PID control. One is "split" old style control
- * for intake & exhaust fans, the other is "combined" control for both
- * CPUs that also deals with the pumps when present. To be "compatible"
- * with OS X at this point, we only use "COMBINED" on the machines that
- * are identified as having the pumps (though that identification is at
- * least dodgy). Ultimately, we could probably switch completely to this
- * algorithm provided we hack it to deal with the UP case
- */
-#define CPU_PID_TYPE_SPLIT	0
-#define CPU_PID_TYPE_COMBINED	1
-#define CPU_PID_TYPE_RACKMAC	2
-
-/*
- * This table describes all fans in the FCU. The "id" and "type" values
- * are defaults valid for all earlier machines. Newer machines will
- * eventually override the table content based on the device-tree
- */
-struct fcu_fan_table
-{
-	char*	loc;	/* location code */
-	int	type;	/* 0 = rpm, 1 = pwm, 2 = pump */
-	int	id;	/* id or -1 */
-};
-
-#define FCU_FAN_RPM		0
-#define FCU_FAN_PWM		1
-
-#define FCU_FAN_ABSENT_ID	-1
-
-#define FCU_FAN_COUNT		ARRAY_SIZE(fcu_fans)
-
-struct fcu_fan_table	fcu_fans[] = {
-	[BACKSIDE_FAN_PWM_INDEX] = {
-		.loc	= "BACKSIDE,SYS CTRLR FAN",
-		.type	= FCU_FAN_PWM,
-		.id	= BACKSIDE_FAN_PWM_DEFAULT_ID,
-	},
-	[DRIVES_FAN_RPM_INDEX] = {
-		.loc	= "DRIVE BAY",
-		.type	= FCU_FAN_RPM,
-		.id	= DRIVES_FAN_RPM_DEFAULT_ID,
-	},
-	[SLOTS_FAN_PWM_INDEX] = {
-		.loc	= "SLOT,PCI FAN",
-		.type	= FCU_FAN_PWM,
-		.id	= SLOTS_FAN_PWM_DEFAULT_ID,
-	},
-	[CPUA_INTAKE_FAN_RPM_INDEX] = {
-		.loc	= "CPU A INTAKE",
-		.type	= FCU_FAN_RPM,
-		.id	= CPUA_INTAKE_FAN_RPM_DEFAULT_ID,
-	},
-	[CPUA_EXHAUST_FAN_RPM_INDEX] = {
-		.loc	= "CPU A EXHAUST",
-		.type	= FCU_FAN_RPM,
-		.id	= CPUA_EXHAUST_FAN_RPM_DEFAULT_ID,
-	},
-	[CPUB_INTAKE_FAN_RPM_INDEX] = {
-		.loc	= "CPU B INTAKE",
-		.type	= FCU_FAN_RPM,
-		.id	= CPUB_INTAKE_FAN_RPM_DEFAULT_ID,
-	},
-	[CPUB_EXHAUST_FAN_RPM_INDEX] = {
-		.loc	= "CPU B EXHAUST",
-		.type	= FCU_FAN_RPM,
-		.id	= CPUB_EXHAUST_FAN_RPM_DEFAULT_ID,
-	},
-	/* pumps aren't present by default, have to be looked up in the
-	 * device-tree
-	 */
-	[CPUA_PUMP_RPM_INDEX] = {
-		.loc	= "CPU A PUMP",
-		.type	= FCU_FAN_RPM,		
-		.id	= FCU_FAN_ABSENT_ID,
-	},
-	[CPUB_PUMP_RPM_INDEX] = {
-		.loc	= "CPU B PUMP",
-		.type	= FCU_FAN_RPM,
-		.id	= FCU_FAN_ABSENT_ID,
-	},
-	/* Xserve fans */
-	[CPU_A1_FAN_RPM_INDEX] = {
-		.loc	= "CPU A 1",
-		.type	= FCU_FAN_RPM,
-		.id	= FCU_FAN_ABSENT_ID,
-	},
-	[CPU_A2_FAN_RPM_INDEX] = {
-		.loc	= "CPU A 2",
-		.type	= FCU_FAN_RPM,
-		.id	= FCU_FAN_ABSENT_ID,
-	},
-	[CPU_A3_FAN_RPM_INDEX] = {
-		.loc	= "CPU A 3",
-		.type	= FCU_FAN_RPM,
-		.id	= FCU_FAN_ABSENT_ID,
-	},
-	[CPU_B1_FAN_RPM_INDEX] = {
-		.loc	= "CPU B 1",
-		.type	= FCU_FAN_RPM,
-		.id	= FCU_FAN_ABSENT_ID,
-	},
-	[CPU_B2_FAN_RPM_INDEX] = {
-		.loc	= "CPU B 2",
-		.type	= FCU_FAN_RPM,
-		.id	= FCU_FAN_ABSENT_ID,
-	},
-	[CPU_B3_FAN_RPM_INDEX] = {
-		.loc	= "CPU B 3",
-		.type	= FCU_FAN_RPM,
-		.id	= FCU_FAN_ABSENT_ID,
-	},
-};
-
-static struct i2c_driver therm_pm72_driver;
-
-/*
- * Utility function to create an i2c_client structure and
- * attach it to one of u3 adapters
- */
-static struct i2c_client *attach_i2c_chip(int id, const char *name)
-{
-	struct i2c_client *clt;
-	struct i2c_adapter *adap;
-	struct i2c_board_info info;
-
-	if (id & 0x200)
-		adap = k2;
-	else if (id & 0x100)
-		adap = u3_1;
-	else
-		adap = u3_0;
-	if (adap == NULL)
-		return NULL;
-
-	memset(&info, 0, sizeof(struct i2c_board_info));
-	info.addr = (id >> 1) & 0x7f;
-	strlcpy(info.type, "therm_pm72", I2C_NAME_SIZE);
-	clt = i2c_new_device(adap, &info);
-	if (!clt) {
-		printk(KERN_ERR "therm_pm72: Failed to attach to i2c ID 0x%x\n", id);
-		return NULL;
-	}
-
-	/*
-	 * Let i2c-core delete that device on driver removal.
-	 * This is safe because i2c-core holds the core_lock mutex for us.
-	 */
-	list_add_tail(&clt->detected, &therm_pm72_driver.clients);
-	return clt;
-}
-
-/*
- * Here are the i2c chip access wrappers
- */
-
-static void initialize_adc(struct cpu_pid_state *state)
-{
-	int rc;
-	u8 buf[2];
-
-	/* Read ADC the configuration register and cache it. We
-	 * also make sure Config2 contains proper values, I've seen
-	 * cases where we got stale grabage in there, thus preventing
-	 * proper reading of conv. values
-	 */
-
-	/* Clear Config2 */
-	buf[0] = 5;
-	buf[1] = 0;
-	i2c_master_send(state->monitor, buf, 2);
-
-	/* Read & cache Config1 */
-	buf[0] = 1;
-	rc = i2c_master_send(state->monitor, buf, 1);
-	if (rc > 0) {
-		rc = i2c_master_recv(state->monitor, buf, 1);
-		if (rc > 0) {
-			state->adc_config = buf[0];
-			DBG("ADC config reg: %02x\n", state->adc_config);
-			/* Disable shutdown mode */
-		       	state->adc_config &= 0xfe;
-			buf[0] = 1;
-			buf[1] = state->adc_config;
-			rc = i2c_master_send(state->monitor, buf, 2);
-		}
-	}
-	if (rc <= 0)
-		printk(KERN_ERR "therm_pm72: Error reading ADC config"
-		       " register !\n");
-}
-
-static int read_smon_adc(struct cpu_pid_state *state, int chan)
-{
-	int rc, data, tries = 0;
-	u8 buf[2];
-
-	for (;;) {
-		/* Set channel */
-		buf[0] = 1;
-		buf[1] = (state->adc_config & 0x1f) | (chan << 5);
-		rc = i2c_master_send(state->monitor, buf, 2);
-		if (rc <= 0)
-			goto error;
-		/* Wait for conversion */
-		msleep(1);
-		/* Switch to data register */
-		buf[0] = 4;
-		rc = i2c_master_send(state->monitor, buf, 1);
-		if (rc <= 0)
-			goto error;
-		/* Read result */
-		rc = i2c_master_recv(state->monitor, buf, 2);
-		if (rc < 0)
-			goto error;
-		data = ((u16)buf[0]) << 8 | (u16)buf[1];
-		return data >> 6;
-	error:
-		DBG("Error reading ADC, retrying...\n");
-		if (++tries > 10) {
-			printk(KERN_ERR "therm_pm72: Error reading ADC !\n");
-			return -1;
-		}
-		msleep(10);
-	}
-}
-
-static int read_lm87_reg(struct i2c_client * chip, int reg)
-{
-	int rc, tries = 0;
-	u8 buf;
-
-	for (;;) {
-		/* Set address */
-		buf = (u8)reg;
-		rc = i2c_master_send(chip, &buf, 1);
-		if (rc <= 0)
-			goto error;
-		rc = i2c_master_recv(chip, &buf, 1);
-		if (rc <= 0)
-			goto error;
-		return (int)buf;
-	error:
-		DBG("Error reading LM87, retrying...\n");
-		if (++tries > 10) {
-			printk(KERN_ERR "therm_pm72: Error reading LM87 !\n");
-			return -1;
-		}
-		msleep(10);
-	}
-}
-
-static int fan_read_reg(int reg, unsigned char *buf, int nb)
-{
-	int tries, nr, nw;
-
-	buf[0] = reg;
-	tries = 0;
-	for (;;) {
-		nw = i2c_master_send(fcu, buf, 1);
-		if (nw > 0 || (nw < 0 && nw != -EIO) || tries >= 100)
-			break;
-		msleep(10);
-		++tries;
-	}
-	if (nw <= 0) {
-		printk(KERN_ERR "Failure writing address to FCU: %d", nw);
-		return -EIO;
-	}
-	tries = 0;
-	for (;;) {
-		nr = i2c_master_recv(fcu, buf, nb);
-		if (nr > 0 || (nr < 0 && nr != -ENODEV) || tries >= 100)
-			break;
-		msleep(10);
-		++tries;
-	}
-	if (nr <= 0)
-		printk(KERN_ERR "Failure reading data from FCU: %d", nw);
-	return nr;
-}
-
-static int fan_write_reg(int reg, const unsigned char *ptr, int nb)
-{
-	int tries, nw;
-	unsigned char buf[16];
-
-	buf[0] = reg;
-	memcpy(buf+1, ptr, nb);
-	++nb;
-	tries = 0;
-	for (;;) {
-		nw = i2c_master_send(fcu, buf, nb);
-		if (nw > 0 || (nw < 0 && nw != -EIO) || tries >= 100)
-			break;
-		msleep(10);
-		++tries;
-	}
-	if (nw < 0)
-		printk(KERN_ERR "Failure writing to FCU: %d", nw);
-	return nw;
-}
-
-static int start_fcu(void)
-{
-	unsigned char buf = 0xff;
-	int rc;
-
-	rc = fan_write_reg(0xe, &buf, 1);
-	if (rc < 0)
-		return -EIO;
-	rc = fan_write_reg(0x2e, &buf, 1);
-	if (rc < 0)
-		return -EIO;
-	rc = fan_read_reg(0, &buf, 1);
-	if (rc < 0)
-		return -EIO;
-	fcu_rpm_shift = (buf == 1) ? 2 : 3;
-	printk(KERN_DEBUG "FCU Initialized, RPM fan shift is %d\n",
-	       fcu_rpm_shift);
-
-	return 0;
-}
-
-static int set_rpm_fan(int fan_index, int rpm)
-{
-	unsigned char buf[2];
-	int rc, id, min, max;
-
-	if (fcu_fans[fan_index].type != FCU_FAN_RPM)
-		return -EINVAL;
-	id = fcu_fans[fan_index].id; 
-	if (id == FCU_FAN_ABSENT_ID)
-		return -EINVAL;
-
-	min = 2400 >> fcu_rpm_shift;
-	max = 56000 >> fcu_rpm_shift;
-
-	if (rpm < min)
-		rpm = min;
-	else if (rpm > max)
-		rpm = max;
-	buf[0] = rpm >> (8 - fcu_rpm_shift);
-	buf[1] = rpm << fcu_rpm_shift;
-	rc = fan_write_reg(0x10 + (id * 2), buf, 2);
-	if (rc < 0)
-		return -EIO;
-	return 0;
-}
-
-static int get_rpm_fan(int fan_index, int programmed)
-{
-	unsigned char failure;
-	unsigned char active;
-	unsigned char buf[2];
-	int rc, id, reg_base;
-
-	if (fcu_fans[fan_index].type != FCU_FAN_RPM)
-		return -EINVAL;
-	id = fcu_fans[fan_index].id; 
-	if (id == FCU_FAN_ABSENT_ID)
-		return -EINVAL;
-
-	rc = fan_read_reg(0xb, &failure, 1);
-	if (rc != 1)
-		return -EIO;
-	if ((failure & (1 << id)) != 0)
-		return -EFAULT;
-	rc = fan_read_reg(0xd, &active, 1);
-	if (rc != 1)
-		return -EIO;
-	if ((active & (1 << id)) == 0)
-		return -ENXIO;
-
-	/* Programmed value or real current speed */
-	reg_base = programmed ? 0x10 : 0x11;
-	rc = fan_read_reg(reg_base + (id * 2), buf, 2);
-	if (rc != 2)
-		return -EIO;
-
-	return (buf[0] << (8 - fcu_rpm_shift)) | buf[1] >> fcu_rpm_shift;
-}
-
-static int set_pwm_fan(int fan_index, int pwm)
-{
-	unsigned char buf[2];
-	int rc, id;
-
-	if (fcu_fans[fan_index].type != FCU_FAN_PWM)
-		return -EINVAL;
-	id = fcu_fans[fan_index].id; 
-	if (id == FCU_FAN_ABSENT_ID)
-		return -EINVAL;
-
-	if (pwm < 10)
-		pwm = 10;
-	else if (pwm > 100)
-		pwm = 100;
-	pwm = (pwm * 2559) / 1000;
-	buf[0] = pwm;
-	rc = fan_write_reg(0x30 + (id * 2), buf, 1);
-	if (rc < 0)
-		return rc;
-	return 0;
-}
-
-static int get_pwm_fan(int fan_index)
-{
-	unsigned char failure;
-	unsigned char active;
-	unsigned char buf[2];
-	int rc, id;
-
-	if (fcu_fans[fan_index].type != FCU_FAN_PWM)
-		return -EINVAL;
-	id = fcu_fans[fan_index].id; 
-	if (id == FCU_FAN_ABSENT_ID)
-		return -EINVAL;
-
-	rc = fan_read_reg(0x2b, &failure, 1);
-	if (rc != 1)
-		return -EIO;
-	if ((failure & (1 << id)) != 0)
-		return -EFAULT;
-	rc = fan_read_reg(0x2d, &active, 1);
-	if (rc != 1)
-		return -EIO;
-	if ((active & (1 << id)) == 0)
-		return -ENXIO;
-
-	/* Programmed value or real current speed */
-	rc = fan_read_reg(0x30 + (id * 2), buf, 1);
-	if (rc != 1)
-		return -EIO;
-
-	return (buf[0] * 1000) / 2559;
-}
-
-static void tickle_fcu(void)
-{
-	int pwm;
-
-	pwm = get_pwm_fan(SLOTS_FAN_PWM_INDEX);
-
-	DBG("FCU Tickle, slots fan is: %d\n", pwm);
-	if (pwm < 0)
-		pwm = 100;
-
-	if (!rackmac) {
-		pwm = SLOTS_FAN_DEFAULT_PWM;
-	} else if (pwm < SLOTS_PID_OUTPUT_MIN)
-		pwm = SLOTS_PID_OUTPUT_MIN;
-
-	/* That is hopefully enough to make the FCU happy */
-	set_pwm_fan(SLOTS_FAN_PWM_INDEX, pwm);
-}
-
-
-/*
- * Utility routine to read the CPU calibration EEPROM data
- * from the device-tree
- */
-static int read_eeprom(int cpu, struct mpu_data *out)
-{
-	struct device_node *np;
-	char nodename[64];
-	const u8 *data;
-	int len;
-
-	/* prom.c routine for finding a node by path is a bit brain dead
-	 * and requires exact @xxx unit numbers. This is a bit ugly but
-	 * will work for these machines
-	 */
-	sprintf(nodename, "/u3@0,f8000000/i2c@f8001000/cpuid@a%d", cpu ? 2 : 0);
-	np = of_find_node_by_path(nodename);
-	if (np == NULL) {
-		printk(KERN_ERR "therm_pm72: Failed to retrieve cpuid node from device-tree\n");
-		return -ENODEV;
-	}
-	data = of_get_property(np, "cpuid", &len);
-	if (data == NULL) {
-		printk(KERN_ERR "therm_pm72: Failed to retrieve cpuid property from device-tree\n");
-		of_node_put(np);
-		return -ENODEV;
-	}
-	memcpy(out, data, sizeof(struct mpu_data));
-	of_node_put(np);
-	
-	return 0;
-}
-
-static void fetch_cpu_pumps_minmax(void)
-{
-	struct cpu_pid_state *state0 = &processor_state[0];
-	struct cpu_pid_state *state1 = &processor_state[1];
-	u16 pump_min = 0, pump_max = 0xffff;
-	u16 tmp[4];
-
-	/* Try to fetch pumps min/max infos from eeprom */
-
-	memcpy(&tmp, &state0->mpu.processor_part_num, 8);
-	if (tmp[0] != 0xffff && tmp[1] != 0xffff) {
-		pump_min = max(pump_min, tmp[0]);
-		pump_max = min(pump_max, tmp[1]);
-	}
-	if (tmp[2] != 0xffff && tmp[3] != 0xffff) {
-		pump_min = max(pump_min, tmp[2]);
-		pump_max = min(pump_max, tmp[3]);
-	}
-
-	/* Double check the values, this _IS_ needed as the EEPROM on
-	 * some dual 2.5Ghz G5s seem, at least, to have both min & max
-	 * same to the same value ... (grrrr)
-	 */
-	if (pump_min == pump_max || pump_min == 0 || pump_max == 0xffff) {
-		pump_min = CPU_PUMP_OUTPUT_MIN;
-		pump_max = CPU_PUMP_OUTPUT_MAX;
-	}
-
-	state0->pump_min = state1->pump_min = pump_min;
-	state0->pump_max = state1->pump_max = pump_max;
-}
-
-/* 
- * Now, unfortunately, sysfs doesn't give us a nice void * we could
- * pass around to the attribute functions, so we don't really have
- * choice but implement a bunch of them...
- *
- * That sucks a bit, we take the lock because FIX32TOPRINT evaluates
- * the input twice... I accept patches :)
- */
-#define BUILD_SHOW_FUNC_FIX(name, data)				\
-static ssize_t show_##name(struct device *dev, struct device_attribute *attr, char *buf)	\
-{								\
-	ssize_t r;						\
-	mutex_lock(&driver_lock);					\
-	r = sprintf(buf, "%d.%03d", FIX32TOPRINT(data));	\
-	mutex_unlock(&driver_lock);					\
-	return r;						\
-}
-#define BUILD_SHOW_FUNC_INT(name, data)				\
-static ssize_t show_##name(struct device *dev, struct device_attribute *attr, char *buf)	\
-{								\
-	return sprintf(buf, "%d", data);			\
-}
-
-BUILD_SHOW_FUNC_FIX(cpu0_temperature, processor_state[0].last_temp)
-BUILD_SHOW_FUNC_FIX(cpu0_voltage, processor_state[0].voltage)
-BUILD_SHOW_FUNC_FIX(cpu0_current, processor_state[0].current_a)
-BUILD_SHOW_FUNC_INT(cpu0_exhaust_fan_rpm, processor_state[0].rpm)
-BUILD_SHOW_FUNC_INT(cpu0_intake_fan_rpm, processor_state[0].intake_rpm)
-
-BUILD_SHOW_FUNC_FIX(cpu1_temperature, processor_state[1].last_temp)
-BUILD_SHOW_FUNC_FIX(cpu1_voltage, processor_state[1].voltage)
-BUILD_SHOW_FUNC_FIX(cpu1_current, processor_state[1].current_a)
-BUILD_SHOW_FUNC_INT(cpu1_exhaust_fan_rpm, processor_state[1].rpm)
-BUILD_SHOW_FUNC_INT(cpu1_intake_fan_rpm, processor_state[1].intake_rpm)
-
-BUILD_SHOW_FUNC_FIX(backside_temperature, backside_state.last_temp)
-BUILD_SHOW_FUNC_INT(backside_fan_pwm, backside_state.pwm)
-
-BUILD_SHOW_FUNC_FIX(drives_temperature, drives_state.last_temp)
-BUILD_SHOW_FUNC_INT(drives_fan_rpm, drives_state.rpm)
-
-BUILD_SHOW_FUNC_FIX(slots_temperature, slots_state.last_temp)
-BUILD_SHOW_FUNC_INT(slots_fan_pwm, slots_state.pwm)
-
-BUILD_SHOW_FUNC_FIX(dimms_temperature, dimms_state.last_temp)
-
-static DEVICE_ATTR(cpu0_temperature,S_IRUGO,show_cpu0_temperature,NULL);
-static DEVICE_ATTR(cpu0_voltage,S_IRUGO,show_cpu0_voltage,NULL);
-static DEVICE_ATTR(cpu0_current,S_IRUGO,show_cpu0_current,NULL);
-static DEVICE_ATTR(cpu0_exhaust_fan_rpm,S_IRUGO,show_cpu0_exhaust_fan_rpm,NULL);
-static DEVICE_ATTR(cpu0_intake_fan_rpm,S_IRUGO,show_cpu0_intake_fan_rpm,NULL);
-
-static DEVICE_ATTR(cpu1_temperature,S_IRUGO,show_cpu1_temperature,NULL);
-static DEVICE_ATTR(cpu1_voltage,S_IRUGO,show_cpu1_voltage,NULL);
-static DEVICE_ATTR(cpu1_current,S_IRUGO,show_cpu1_current,NULL);
-static DEVICE_ATTR(cpu1_exhaust_fan_rpm,S_IRUGO,show_cpu1_exhaust_fan_rpm,NULL);
-static DEVICE_ATTR(cpu1_intake_fan_rpm,S_IRUGO,show_cpu1_intake_fan_rpm,NULL);
-
-static DEVICE_ATTR(backside_temperature,S_IRUGO,show_backside_temperature,NULL);
-static DEVICE_ATTR(backside_fan_pwm,S_IRUGO,show_backside_fan_pwm,NULL);
-
-static DEVICE_ATTR(drives_temperature,S_IRUGO,show_drives_temperature,NULL);
-static DEVICE_ATTR(drives_fan_rpm,S_IRUGO,show_drives_fan_rpm,NULL);
-
-static DEVICE_ATTR(slots_temperature,S_IRUGO,show_slots_temperature,NULL);
-static DEVICE_ATTR(slots_fan_pwm,S_IRUGO,show_slots_fan_pwm,NULL);
-
-static DEVICE_ATTR(dimms_temperature,S_IRUGO,show_dimms_temperature,NULL);
-
-/*
- * CPUs fans control loop
- */
-
-static int do_read_one_cpu_values(struct cpu_pid_state *state, s32 *temp, s32 *power)
-{
-	s32 ltemp, volts, amps;
-	int index, rc = 0;
-
-	/* Default (in case of error) */
-	*temp = state->cur_temp;
-	*power = state->cur_power;
-
-	if (cpu_pid_type == CPU_PID_TYPE_RACKMAC)
-		index = (state->index == 0) ?
-			CPU_A1_FAN_RPM_INDEX : CPU_B1_FAN_RPM_INDEX;
-	else
-		index = (state->index == 0) ?
-			CPUA_EXHAUST_FAN_RPM_INDEX : CPUB_EXHAUST_FAN_RPM_INDEX;
-
-	/* Read current fan status */
-	rc = get_rpm_fan(index, !RPM_PID_USE_ACTUAL_SPEED);
-	if (rc < 0) {
-		/* XXX What do we do now ? Nothing for now, keep old value, but
-		 * return error upstream
-		 */
-		DBG("  cpu %d, fan reading error !\n", state->index);
-	} else {
-		state->rpm = rc;
-		DBG("  cpu %d, exhaust RPM: %d\n", state->index, state->rpm);
-	}
-
-	/* Get some sensor readings and scale it */
-	ltemp = read_smon_adc(state, 1);
-	if (ltemp == -1) {
-		/* XXX What do we do now ? */
-		state->overtemp++;
-		if (rc == 0)
-			rc = -EIO;
-		DBG("  cpu %d, temp reading error !\n", state->index);
-	} else {
-		/* Fixup temperature according to diode calibration
-		 */
-		DBG("  cpu %d, temp raw: %04x, m_diode: %04x, b_diode: %04x\n",
-		    state->index,
-		    ltemp, state->mpu.mdiode, state->mpu.bdiode);
-		*temp = ((s32)ltemp * (s32)state->mpu.mdiode + ((s32)state->mpu.bdiode << 12)) >> 2;
-		state->last_temp = *temp;
-		DBG("  temp: %d.%03d\n", FIX32TOPRINT((*temp)));
-	}
-
-	/*
-	 * Read voltage & current and calculate power
-	 */
-	volts = read_smon_adc(state, 3);
-	amps = read_smon_adc(state, 4);
-
-	/* Scale voltage and current raw sensor values according to fixed scales
-	 * obtained in Darwin and calculate power from I and V
-	 */
-	volts *= ADC_CPU_VOLTAGE_SCALE;
-	amps *= ADC_CPU_CURRENT_SCALE;
-	*power = (((u64)volts) * ((u64)amps)) >> 16;
-	state->voltage = volts;
-	state->current_a = amps;
-	state->last_power = *power;
-
-	DBG("  cpu %d, current: %d.%03d, voltage: %d.%03d, power: %d.%03d W\n",
-	    state->index, FIX32TOPRINT(state->current_a),
-	    FIX32TOPRINT(state->voltage), FIX32TOPRINT(*power));
-
-	return 0;
-}
-
-static void do_cpu_pid(struct cpu_pid_state *state, s32 temp, s32 power)
-{
-	s32 power_target, integral, derivative, proportional, adj_in_target, sval;
-	s64 integ_p, deriv_p, prop_p, sum; 
-	int i;
-
-	/* Calculate power target value (could be done once for all)
-	 * and convert to a 16.16 fp number
-	 */
-	power_target = ((u32)(state->mpu.pmaxh - state->mpu.padjmax)) << 16;
-	DBG("  power target: %d.%03d, error: %d.%03d\n",
-	    FIX32TOPRINT(power_target), FIX32TOPRINT(power_target - power));
-
-	/* Store temperature and power in history array */
-	state->cur_temp = (state->cur_temp + 1) % CPU_TEMP_HISTORY_SIZE;
-	state->temp_history[state->cur_temp] = temp;
-	state->cur_power = (state->cur_power + 1) % state->count_power;
-	state->power_history[state->cur_power] = power;
-	state->error_history[state->cur_power] = power_target - power;
-	
-	/* If first loop, fill the history table */
-	if (state->first) {
-		for (i = 0; i < (state->count_power - 1); i++) {
-			state->cur_power = (state->cur_power + 1) % state->count_power;
-			state->power_history[state->cur_power] = power;
-			state->error_history[state->cur_power] = power_target - power;
-		}
-		for (i = 0; i < (CPU_TEMP_HISTORY_SIZE - 1); i++) {
-			state->cur_temp = (state->cur_temp + 1) % CPU_TEMP_HISTORY_SIZE;
-			state->temp_history[state->cur_temp] = temp;			
-		}
-		state->first = 0;
-	}
-
-	/* Calculate the integral term normally based on the "power" values */
-	sum = 0;
-	integral = 0;
-	for (i = 0; i < state->count_power; i++)
-		integral += state->error_history[i];
-	integral *= CPU_PID_INTERVAL;
-	DBG("  integral: %08x\n", integral);
-
-	/* Calculate the adjusted input (sense value).
-	 *   G_r is 12.20
-	 *   integ is 16.16
-	 *   so the result is 28.36
-	 *
-	 * input target is mpu.ttarget, input max is mpu.tmax
-	 */
-	integ_p = ((s64)state->mpu.pid_gr) * (s64)integral;
-	DBG("   integ_p: %d\n", (int)(integ_p >> 36));
-	sval = (state->mpu.tmax << 16) - ((integ_p >> 20) & 0xffffffff);
-	adj_in_target = (state->mpu.ttarget << 16);
-	if (adj_in_target > sval)
-		adj_in_target = sval;
-	DBG("   adj_in_target: %d.%03d, ttarget: %d\n", FIX32TOPRINT(adj_in_target),
-	    state->mpu.ttarget);
-
-	/* Calculate the derivative term */
-	derivative = state->temp_history[state->cur_temp] -
-		state->temp_history[(state->cur_temp + CPU_TEMP_HISTORY_SIZE - 1)
-				    % CPU_TEMP_HISTORY_SIZE];
-	derivative /= CPU_PID_INTERVAL;
-	deriv_p = ((s64)state->mpu.pid_gd) * (s64)derivative;
-	DBG("   deriv_p: %d\n", (int)(deriv_p >> 36));
-	sum += deriv_p;
-
-	/* Calculate the proportional term */
-	proportional = temp - adj_in_target;
-	prop_p = ((s64)state->mpu.pid_gp) * (s64)proportional;
-	DBG("   prop_p: %d\n", (int)(prop_p >> 36));
-	sum += prop_p;
-
-	/* Scale sum */
-	sum >>= 36;
-
-	DBG("   sum: %d\n", (int)sum);
-	state->rpm += (s32)sum;
-}
-
-static void do_monitor_cpu_combined(void)
-{
-	struct cpu_pid_state *state0 = &processor_state[0];
-	struct cpu_pid_state *state1 = &processor_state[1];
-	s32 temp0, power0, temp1, power1;
-	s32 temp_combi, power_combi;
-	int rc, intake, pump;
-
-	rc = do_read_one_cpu_values(state0, &temp0, &power0);
-	if (rc < 0) {
-		/* XXX What do we do now ? */
-	}
-	state1->overtemp = 0;
-	rc = do_read_one_cpu_values(state1, &temp1, &power1);
-	if (rc < 0) {
-		/* XXX What do we do now ? */
-	}
-	if (state1->overtemp)
-		state0->overtemp++;
-
-	temp_combi = max(temp0, temp1);
-	power_combi = max(power0, power1);
-
-	/* Check tmax, increment overtemp if we are there. At tmax+8, we go
-	 * full blown immediately and try to trigger a shutdown
-	 */
-	if (temp_combi >= ((state0->mpu.tmax + 8) << 16)) {
-		printk(KERN_WARNING "Warning ! Temperature way above maximum (%d) !\n",
-		       temp_combi >> 16);
-		state0->overtemp += CPU_MAX_OVERTEMP / 4;
-	} else if (temp_combi > (state0->mpu.tmax << 16)) {
-		state0->overtemp++;
-		printk(KERN_WARNING "Temperature %d above max %d. overtemp %d\n",
-		       temp_combi >> 16, state0->mpu.tmax, state0->overtemp);
-	} else {
-		if (state0->overtemp)
-			printk(KERN_WARNING "Temperature back down to %d\n",
-			       temp_combi >> 16);
-		state0->overtemp = 0;
-	}
-	if (state0->overtemp >= CPU_MAX_OVERTEMP)
-		critical_state = 1;
-	if (state0->overtemp > 0) {
-		state0->rpm = state0->mpu.rmaxn_exhaust_fan;
-		state0->intake_rpm = intake = state0->mpu.rmaxn_intake_fan;
-		pump = state0->pump_max;
-		goto do_set_fans;
-	}
-
-	/* Do the PID */
-	do_cpu_pid(state0, temp_combi, power_combi);
-
-	/* Range check */
-	state0->rpm = max(state0->rpm, (int)state0->mpu.rminn_exhaust_fan);
-	state0->rpm = min(state0->rpm, (int)state0->mpu.rmaxn_exhaust_fan);
-
-	/* Calculate intake fan speed */
-	intake = (state0->rpm * CPU_INTAKE_SCALE) >> 16;
-	intake = max(intake, (int)state0->mpu.rminn_intake_fan);
-	intake = min(intake, (int)state0->mpu.rmaxn_intake_fan);
-	state0->intake_rpm = intake;
-
-	/* Calculate pump speed */
-	pump = (state0->rpm * state0->pump_max) /
-		state0->mpu.rmaxn_exhaust_fan;
-	pump = min(pump, state0->pump_max);
-	pump = max(pump, state0->pump_min);
-	
- do_set_fans:
-	/* We copy values from state 0 to state 1 for /sysfs */
-	state1->rpm = state0->rpm;
-	state1->intake_rpm = state0->intake_rpm;
-
-	DBG("** CPU %d RPM: %d Ex, %d, Pump: %d, In, overtemp: %d\n",
-	    state1->index, (int)state1->rpm, intake, pump, state1->overtemp);
-
-	/* We should check for errors, shouldn't we ? But then, what
-	 * do we do once the error occurs ? For FCU notified fan
-	 * failures (-EFAULT) we probably want to notify userland
-	 * some way...
-	 */
-	set_rpm_fan(CPUA_INTAKE_FAN_RPM_INDEX, intake);
-	set_rpm_fan(CPUA_EXHAUST_FAN_RPM_INDEX, state0->rpm);
-	set_rpm_fan(CPUB_INTAKE_FAN_RPM_INDEX, intake);
-	set_rpm_fan(CPUB_EXHAUST_FAN_RPM_INDEX, state0->rpm);
-
-	if (fcu_fans[CPUA_PUMP_RPM_INDEX].id != FCU_FAN_ABSENT_ID)
-		set_rpm_fan(CPUA_PUMP_RPM_INDEX, pump);
-	if (fcu_fans[CPUB_PUMP_RPM_INDEX].id != FCU_FAN_ABSENT_ID)
-		set_rpm_fan(CPUB_PUMP_RPM_INDEX, pump);
-}
-
-static void do_monitor_cpu_split(struct cpu_pid_state *state)
-{
-	s32 temp, power;
-	int rc, intake;
-
-	/* Read current fan status */
-	rc = do_read_one_cpu_values(state, &temp, &power);
-	if (rc < 0) {
-		/* XXX What do we do now ? */
-	}
-
-	/* Check tmax, increment overtemp if we are there. At tmax+8, we go
-	 * full blown immediately and try to trigger a shutdown
-	 */
-	if (temp >= ((state->mpu.tmax + 8) << 16)) {
-		printk(KERN_WARNING "Warning ! CPU %d temperature way above maximum"
-		       " (%d) !\n",
-		       state->index, temp >> 16);
-		state->overtemp += CPU_MAX_OVERTEMP / 4;
-	} else if (temp > (state->mpu.tmax << 16)) {
-		state->overtemp++;
-		printk(KERN_WARNING "CPU %d temperature %d above max %d. overtemp %d\n",
-		       state->index, temp >> 16, state->mpu.tmax, state->overtemp);
-	} else {
-		if (state->overtemp)
-			printk(KERN_WARNING "CPU %d temperature back down to %d\n",
-			       state->index, temp >> 16);
-		state->overtemp = 0;
-	}
-	if (state->overtemp >= CPU_MAX_OVERTEMP)
-		critical_state = 1;
-	if (state->overtemp > 0) {
-		state->rpm = state->mpu.rmaxn_exhaust_fan;
-		state->intake_rpm = intake = state->mpu.rmaxn_intake_fan;
-		goto do_set_fans;
-	}
-
-	/* Do the PID */
-	do_cpu_pid(state, temp, power);
-
-	/* Range check */
-	state->rpm = max(state->rpm, (int)state->mpu.rminn_exhaust_fan);
-	state->rpm = min(state->rpm, (int)state->mpu.rmaxn_exhaust_fan);
-
-	/* Calculate intake fan */
-	intake = (state->rpm * CPU_INTAKE_SCALE) >> 16;
-	intake = max(intake, (int)state->mpu.rminn_intake_fan);
-	intake = min(intake, (int)state->mpu.rmaxn_intake_fan);
-	state->intake_rpm = intake;
-
- do_set_fans:
-	DBG("** CPU %d RPM: %d Ex, %d In, overtemp: %d\n",
-	    state->index, (int)state->rpm, intake, state->overtemp);
-
-	/* We should check for errors, shouldn't we ? But then, what
-	 * do we do once the error occurs ? For FCU notified fan
-	 * failures (-EFAULT) we probably want to notify userland
-	 * some way...
-	 */
-	if (state->index == 0) {
-		set_rpm_fan(CPUA_INTAKE_FAN_RPM_INDEX, intake);
-		set_rpm_fan(CPUA_EXHAUST_FAN_RPM_INDEX, state->rpm);
-	} else {
-		set_rpm_fan(CPUB_INTAKE_FAN_RPM_INDEX, intake);
-		set_rpm_fan(CPUB_EXHAUST_FAN_RPM_INDEX, state->rpm);
-	}
-}
-
-static void do_monitor_cpu_rack(struct cpu_pid_state *state)
-{
-	s32 temp, power, fan_min;
-	int rc;
-
-	/* Read current fan status */
-	rc = do_read_one_cpu_values(state, &temp, &power);
-	if (rc < 0) {
-		/* XXX What do we do now ? */
-	}
-
-	/* Check tmax, increment overtemp if we are there. At tmax+8, we go
-	 * full blown immediately and try to trigger a shutdown
-	 */
-	if (temp >= ((state->mpu.tmax + 8) << 16)) {
-		printk(KERN_WARNING "Warning ! CPU %d temperature way above maximum"
-		       " (%d) !\n",
-		       state->index, temp >> 16);
-		state->overtemp = CPU_MAX_OVERTEMP / 4;
-	} else if (temp > (state->mpu.tmax << 16)) {
-		state->overtemp++;
-		printk(KERN_WARNING "CPU %d temperature %d above max %d. overtemp %d\n",
-		       state->index, temp >> 16, state->mpu.tmax, state->overtemp);
-	} else {
-		if (state->overtemp)
-			printk(KERN_WARNING "CPU %d temperature back down to %d\n",
-			       state->index, temp >> 16);
-		state->overtemp = 0;
-	}
-	if (state->overtemp >= CPU_MAX_OVERTEMP)
-		critical_state = 1;
-	if (state->overtemp > 0) {
-		state->rpm = state->intake_rpm = state->mpu.rmaxn_intake_fan;
-		goto do_set_fans;
-	}
-
-	/* Do the PID */
-	do_cpu_pid(state, temp, power);
-
-	/* Check clamp from dimms */
-	fan_min = dimm_output_clamp;
-	fan_min = max(fan_min, (int)state->mpu.rminn_intake_fan);
-
-	DBG(" CPU min mpu = %d, min dimm = %d\n",
-	    state->mpu.rminn_intake_fan, dimm_output_clamp);
-
-	state->rpm = max(state->rpm, (int)fan_min);
-	state->rpm = min(state->rpm, (int)state->mpu.rmaxn_intake_fan);
-	state->intake_rpm = state->rpm;
-
- do_set_fans:
-	DBG("** CPU %d RPM: %d overtemp: %d\n",
-	    state->index, (int)state->rpm, state->overtemp);
-
-	/* We should check for errors, shouldn't we ? But then, what
-	 * do we do once the error occurs ? For FCU notified fan
-	 * failures (-EFAULT) we probably want to notify userland
-	 * some way...
-	 */
-	if (state->index == 0) {
-		set_rpm_fan(CPU_A1_FAN_RPM_INDEX, state->rpm);
-		set_rpm_fan(CPU_A2_FAN_RPM_INDEX, state->rpm);
-		set_rpm_fan(CPU_A3_FAN_RPM_INDEX, state->rpm);
-	} else {
-		set_rpm_fan(CPU_B1_FAN_RPM_INDEX, state->rpm);
-		set_rpm_fan(CPU_B2_FAN_RPM_INDEX, state->rpm);
-		set_rpm_fan(CPU_B3_FAN_RPM_INDEX, state->rpm);
-	}
-}
-
-/*
- * Initialize the state structure for one CPU control loop
- */
-static int init_processor_state(struct cpu_pid_state *state, int index)
-{
-	int err;
-
-	state->index = index;
-	state->first = 1;
-	state->rpm = (cpu_pid_type == CPU_PID_TYPE_RACKMAC) ? 4000 : 1000;
-	state->overtemp = 0;
-	state->adc_config = 0x00;
-
-
-	if (index == 0)
-		state->monitor = attach_i2c_chip(SUPPLY_MONITOR_ID, "CPU0_monitor");
-	else if (index == 1)
-		state->monitor = attach_i2c_chip(SUPPLY_MONITORB_ID, "CPU1_monitor");
-	if (state->monitor == NULL)
-		goto fail;
-
-	if (read_eeprom(index, &state->mpu))
-		goto fail;
-
-	state->count_power = state->mpu.tguardband;
-	if (state->count_power > CPU_POWER_HISTORY_SIZE) {
-		printk(KERN_WARNING "Warning ! too many power history slots\n");
-		state->count_power = CPU_POWER_HISTORY_SIZE;
-	}
-	DBG("CPU %d Using %d power history entries\n", index, state->count_power);
-
-	if (index == 0) {
-		err = device_create_file(&of_dev->dev, &dev_attr_cpu0_temperature);
-		err |= device_create_file(&of_dev->dev, &dev_attr_cpu0_voltage);
-		err |= device_create_file(&of_dev->dev, &dev_attr_cpu0_current);
-		err |= device_create_file(&of_dev->dev, &dev_attr_cpu0_exhaust_fan_rpm);
-		err |= device_create_file(&of_dev->dev, &dev_attr_cpu0_intake_fan_rpm);
-	} else {
-		err = device_create_file(&of_dev->dev, &dev_attr_cpu1_temperature);
-		err |= device_create_file(&of_dev->dev, &dev_attr_cpu1_voltage);
-		err |= device_create_file(&of_dev->dev, &dev_attr_cpu1_current);
-		err |= device_create_file(&of_dev->dev, &dev_attr_cpu1_exhaust_fan_rpm);
-		err |= device_create_file(&of_dev->dev, &dev_attr_cpu1_intake_fan_rpm);
-	}
-	if (err)
-		printk(KERN_WARNING "Failed to create some of the attribute"
-			"files for CPU %d\n", index);
-
-	return 0;
- fail:
-	state->monitor = NULL;
-	
-	return -ENODEV;
-}
-
-/*
- * Dispose of the state data for one CPU control loop
- */
-static void dispose_processor_state(struct cpu_pid_state *state)
-{
-	if (state->monitor == NULL)
-		return;
-
-	if (state->index == 0) {
-		device_remove_file(&of_dev->dev, &dev_attr_cpu0_temperature);
-		device_remove_file(&of_dev->dev, &dev_attr_cpu0_voltage);
-		device_remove_file(&of_dev->dev, &dev_attr_cpu0_current);
-		device_remove_file(&of_dev->dev, &dev_attr_cpu0_exhaust_fan_rpm);
-		device_remove_file(&of_dev->dev, &dev_attr_cpu0_intake_fan_rpm);
-	} else {
-		device_remove_file(&of_dev->dev, &dev_attr_cpu1_temperature);
-		device_remove_file(&of_dev->dev, &dev_attr_cpu1_voltage);
-		device_remove_file(&of_dev->dev, &dev_attr_cpu1_current);
-		device_remove_file(&of_dev->dev, &dev_attr_cpu1_exhaust_fan_rpm);
-		device_remove_file(&of_dev->dev, &dev_attr_cpu1_intake_fan_rpm);
-	}
-
-	state->monitor = NULL;
-}
-
-/*
- * Motherboard backside & U3 heatsink fan control loop
- */
-static void do_monitor_backside(struct backside_pid_state *state)
-{
-	s32 temp, integral, derivative, fan_min;
-	s64 integ_p, deriv_p, prop_p, sum; 
-	int i, rc;
-
-	if (--state->ticks != 0)
-		return;
-	state->ticks = backside_params.interval;
-
-	DBG("backside:\n");
-
-	/* Check fan status */
-	rc = get_pwm_fan(BACKSIDE_FAN_PWM_INDEX);
-	if (rc < 0) {
-		printk(KERN_WARNING "Error %d reading backside fan !\n", rc);
-		/* XXX What do we do now ? */
-	} else
-		state->pwm = rc;
-	DBG("  current pwm: %d\n", state->pwm);
-
-	/* Get some sensor readings */
-	temp = i2c_smbus_read_byte_data(state->monitor, MAX6690_EXT_TEMP) << 16;
-	state->last_temp = temp;
-	DBG("  temp: %d.%03d, target: %d.%03d\n", FIX32TOPRINT(temp),
-	    FIX32TOPRINT(backside_params.input_target));
-
-	/* Store temperature and error in history array */
-	state->cur_sample = (state->cur_sample + 1) % BACKSIDE_PID_HISTORY_SIZE;
-	state->sample_history[state->cur_sample] = temp;
-	state->error_history[state->cur_sample] = temp - backside_params.input_target;
-	
-	/* If first loop, fill the history table */
-	if (state->first) {
-		for (i = 0; i < (BACKSIDE_PID_HISTORY_SIZE - 1); i++) {
-			state->cur_sample = (state->cur_sample + 1) %
-				BACKSIDE_PID_HISTORY_SIZE;
-			state->sample_history[state->cur_sample] = temp;
-			state->error_history[state->cur_sample] =
-				temp - backside_params.input_target;
-		}
-		state->first = 0;
-	}
-
-	/* Calculate the integral term */
-	sum = 0;
-	integral = 0;
-	for (i = 0; i < BACKSIDE_PID_HISTORY_SIZE; i++)
-		integral += state->error_history[i];
-	integral *= backside_params.interval;
-	DBG("  integral: %08x\n", integral);
-	integ_p = ((s64)backside_params.G_r) * (s64)integral;
-	DBG("   integ_p: %d\n", (int)(integ_p >> 36));
-	sum += integ_p;
-
-	/* Calculate the derivative term */
-	derivative = state->error_history[state->cur_sample] -
-		state->error_history[(state->cur_sample + BACKSIDE_PID_HISTORY_SIZE - 1)
-				    % BACKSIDE_PID_HISTORY_SIZE];
-	derivative /= backside_params.interval;
-	deriv_p = ((s64)backside_params.G_d) * (s64)derivative;
-	DBG("   deriv_p: %d\n", (int)(deriv_p >> 36));
-	sum += deriv_p;
-
-	/* Calculate the proportional term */
-	prop_p = ((s64)backside_params.G_p) * (s64)(state->error_history[state->cur_sample]);
-	DBG("   prop_p: %d\n", (int)(prop_p >> 36));
-	sum += prop_p;
-
-	/* Scale sum */
-	sum >>= 36;
-
-	DBG("   sum: %d\n", (int)sum);
-	if (backside_params.additive)
-		state->pwm += (s32)sum;
-	else
-		state->pwm = sum;
-
-	/* Check for clamp */
-	fan_min = (dimm_output_clamp * 100) / 14000;
-	fan_min = max(fan_min, backside_params.output_min);
-
-	state->pwm = max(state->pwm, fan_min);
-	state->pwm = min(state->pwm, backside_params.output_max);
-
-	DBG("** BACKSIDE PWM: %d\n", (int)state->pwm);
-	set_pwm_fan(BACKSIDE_FAN_PWM_INDEX, state->pwm);
-}
-
-/*
- * Initialize the state structure for the backside fan control loop
- */
-static int init_backside_state(struct backside_pid_state *state)
-{
-	struct device_node *u3;
-	int u3h = 1; /* conservative by default */
-	int err;
-
-	/*
-	 * There are different PID params for machines with U3 and machines
-	 * with U3H, pick the right ones now
-	 */
-	u3 = of_find_node_by_path("/u3@0,f8000000");
-	if (u3 != NULL) {
-		const u32 *vers = of_get_property(u3, "device-rev", NULL);
-		if (vers)
-			if (((*vers) & 0x3f) < 0x34)
-				u3h = 0;
-		of_node_put(u3);
-	}
-
-	if (rackmac) {
-		backside_params.G_d = BACKSIDE_PID_RACK_G_d;
-		backside_params.input_target = BACKSIDE_PID_RACK_INPUT_TARGET;
-		backside_params.output_min = BACKSIDE_PID_U3H_OUTPUT_MIN;
-		backside_params.interval = BACKSIDE_PID_RACK_INTERVAL;
-		backside_params.G_p = BACKSIDE_PID_RACK_G_p;
-		backside_params.G_r = BACKSIDE_PID_G_r;
-		backside_params.output_max = BACKSIDE_PID_OUTPUT_MAX;
-		backside_params.additive = 0;
-	} else if (u3h) {
-		backside_params.G_d = BACKSIDE_PID_U3H_G_d;
-		backside_params.input_target = BACKSIDE_PID_U3H_INPUT_TARGET;
-		backside_params.output_min = BACKSIDE_PID_U3H_OUTPUT_MIN;
-		backside_params.interval = BACKSIDE_PID_INTERVAL;
-		backside_params.G_p = BACKSIDE_PID_G_p;
-		backside_params.G_r = BACKSIDE_PID_G_r;
-		backside_params.output_max = BACKSIDE_PID_OUTPUT_MAX;
-		backside_params.additive = 1;
-	} else {
-		backside_params.G_d = BACKSIDE_PID_U3_G_d;
-		backside_params.input_target = BACKSIDE_PID_U3_INPUT_TARGET;
-		backside_params.output_min = BACKSIDE_PID_U3_OUTPUT_MIN;
-		backside_params.interval = BACKSIDE_PID_INTERVAL;
-		backside_params.G_p = BACKSIDE_PID_G_p;
-		backside_params.G_r = BACKSIDE_PID_G_r;
-		backside_params.output_max = BACKSIDE_PID_OUTPUT_MAX;
-		backside_params.additive = 1;
-	}
-
-	state->ticks = 1;
-	state->first = 1;
-	state->pwm = 50;
-
-	state->monitor = attach_i2c_chip(BACKSIDE_MAX_ID, "backside_temp");
-	if (state->monitor == NULL)
-		return -ENODEV;
-
-	err = device_create_file(&of_dev->dev, &dev_attr_backside_temperature);
-	err |= device_create_file(&of_dev->dev, &dev_attr_backside_fan_pwm);
-	if (err)
-		printk(KERN_WARNING "Failed to create attribute file(s)"
-			" for backside fan\n");
-
-	return 0;
-}
-
-/*
- * Dispose of the state data for the backside control loop
- */
-static void dispose_backside_state(struct backside_pid_state *state)
-{
-	if (state->monitor == NULL)
-		return;
-
-	device_remove_file(&of_dev->dev, &dev_attr_backside_temperature);
-	device_remove_file(&of_dev->dev, &dev_attr_backside_fan_pwm);
-
-	state->monitor = NULL;
-}
- 
-/*
- * Drives bay fan control loop
- */
-static void do_monitor_drives(struct drives_pid_state *state)
-{
-	s32 temp, integral, derivative;
-	s64 integ_p, deriv_p, prop_p, sum; 
-	int i, rc;
-
-	if (--state->ticks != 0)
-		return;
-	state->ticks = DRIVES_PID_INTERVAL;
-
-	DBG("drives:\n");
-
-	/* Check fan status */
-	rc = get_rpm_fan(DRIVES_FAN_RPM_INDEX, !RPM_PID_USE_ACTUAL_SPEED);
-	if (rc < 0) {
-		printk(KERN_WARNING "Error %d reading drives fan !\n", rc);
-		/* XXX What do we do now ? */
-	} else
-		state->rpm = rc;
-	DBG("  current rpm: %d\n", state->rpm);
-
-	/* Get some sensor readings */
-	temp = le16_to_cpu(i2c_smbus_read_word_data(state->monitor,
-						    DS1775_TEMP)) << 8;
-	state->last_temp = temp;
-	DBG("  temp: %d.%03d, target: %d.%03d\n", FIX32TOPRINT(temp),
-	    FIX32TOPRINT(DRIVES_PID_INPUT_TARGET));
-
-	/* Store temperature and error in history array */
-	state->cur_sample = (state->cur_sample + 1) % DRIVES_PID_HISTORY_SIZE;
-	state->sample_history[state->cur_sample] = temp;
-	state->error_history[state->cur_sample] = temp - DRIVES_PID_INPUT_TARGET;
-	
-	/* If first loop, fill the history table */
-	if (state->first) {
-		for (i = 0; i < (DRIVES_PID_HISTORY_SIZE - 1); i++) {
-			state->cur_sample = (state->cur_sample + 1) %
-				DRIVES_PID_HISTORY_SIZE;
-			state->sample_history[state->cur_sample] = temp;
-			state->error_history[state->cur_sample] =
-				temp - DRIVES_PID_INPUT_TARGET;
-		}
-		state->first = 0;
-	}
-
-	/* Calculate the integral term */
-	sum = 0;
-	integral = 0;
-	for (i = 0; i < DRIVES_PID_HISTORY_SIZE; i++)
-		integral += state->error_history[i];
-	integral *= DRIVES_PID_INTERVAL;
-	DBG("  integral: %08x\n", integral);
-	integ_p = ((s64)DRIVES_PID_G_r) * (s64)integral;
-	DBG("   integ_p: %d\n", (int)(integ_p >> 36));
-	sum += integ_p;
-
-	/* Calculate the derivative term */
-	derivative = state->error_history[state->cur_sample] -
-		state->error_history[(state->cur_sample + DRIVES_PID_HISTORY_SIZE - 1)
-				    % DRIVES_PID_HISTORY_SIZE];
-	derivative /= DRIVES_PID_INTERVAL;
-	deriv_p = ((s64)DRIVES_PID_G_d) * (s64)derivative;
-	DBG("   deriv_p: %d\n", (int)(deriv_p >> 36));
-	sum += deriv_p;
-
-	/* Calculate the proportional term */
-	prop_p = ((s64)DRIVES_PID_G_p) * (s64)(state->error_history[state->cur_sample]);
-	DBG("   prop_p: %d\n", (int)(prop_p >> 36));
-	sum += prop_p;
-
-	/* Scale sum */
-	sum >>= 36;
-
-	DBG("   sum: %d\n", (int)sum);
-	state->rpm += (s32)sum;
-
-	state->rpm = max(state->rpm, DRIVES_PID_OUTPUT_MIN);
-	state->rpm = min(state->rpm, DRIVES_PID_OUTPUT_MAX);
-
-	DBG("** DRIVES RPM: %d\n", (int)state->rpm);
-	set_rpm_fan(DRIVES_FAN_RPM_INDEX, state->rpm);
-}
-
-/*
- * Initialize the state structure for the drives bay fan control loop
- */
-static int init_drives_state(struct drives_pid_state *state)
-{
-	int err;
-
-	state->ticks = 1;
-	state->first = 1;
-	state->rpm = 1000;
-
-	state->monitor = attach_i2c_chip(DRIVES_DALLAS_ID, "drives_temp");
-	if (state->monitor == NULL)
-		return -ENODEV;
-
-	err = device_create_file(&of_dev->dev, &dev_attr_drives_temperature);
-	err |= device_create_file(&of_dev->dev, &dev_attr_drives_fan_rpm);
-	if (err)
-		printk(KERN_WARNING "Failed to create attribute file(s)"
-			" for drives bay fan\n");
-
-	return 0;
-}
-
-/*
- * Dispose of the state data for the drives control loop
- */
-static void dispose_drives_state(struct drives_pid_state *state)
-{
-	if (state->monitor == NULL)
-		return;
-
-	device_remove_file(&of_dev->dev, &dev_attr_drives_temperature);
-	device_remove_file(&of_dev->dev, &dev_attr_drives_fan_rpm);
-
-	state->monitor = NULL;
-}
-
-/*
- * DIMMs temp control loop
- */
-static void do_monitor_dimms(struct dimm_pid_state *state)
-{
-	s32 temp, integral, derivative, fan_min;
-	s64 integ_p, deriv_p, prop_p, sum;
-	int i;
-
-	if (--state->ticks != 0)
-		return;
-	state->ticks = DIMM_PID_INTERVAL;
-
-	DBG("DIMM:\n");
-
-	DBG("  current value: %d\n", state->output);
-
-	temp = read_lm87_reg(state->monitor, LM87_INT_TEMP);
-	if (temp < 0)
-		return;
-	temp <<= 16;
-	state->last_temp = temp;
-	DBG("  temp: %d.%03d, target: %d.%03d\n", FIX32TOPRINT(temp),
-	    FIX32TOPRINT(DIMM_PID_INPUT_TARGET));
-
-	/* Store temperature and error in history array */
-	state->cur_sample = (state->cur_sample + 1) % DIMM_PID_HISTORY_SIZE;
-	state->sample_history[state->cur_sample] = temp;
-	state->error_history[state->cur_sample] = temp - DIMM_PID_INPUT_TARGET;
-
-	/* If first loop, fill the history table */
-	if (state->first) {
-		for (i = 0; i < (DIMM_PID_HISTORY_SIZE - 1); i++) {
-			state->cur_sample = (state->cur_sample + 1) %
-				DIMM_PID_HISTORY_SIZE;
-			state->sample_history[state->cur_sample] = temp;
-			state->error_history[state->cur_sample] =
-				temp - DIMM_PID_INPUT_TARGET;
-		}
-		state->first = 0;
-	}
-
-	/* Calculate the integral term */
-	sum = 0;
-	integral = 0;
-	for (i = 0; i < DIMM_PID_HISTORY_SIZE; i++)
-		integral += state->error_history[i];
-	integral *= DIMM_PID_INTERVAL;
-	DBG("  integral: %08x\n", integral);
-	integ_p = ((s64)DIMM_PID_G_r) * (s64)integral;
-	DBG("   integ_p: %d\n", (int)(integ_p >> 36));
-	sum += integ_p;
-
-	/* Calculate the derivative term */
-	derivative = state->error_history[state->cur_sample] -
-		state->error_history[(state->cur_sample + DIMM_PID_HISTORY_SIZE - 1)
-				    % DIMM_PID_HISTORY_SIZE];
-	derivative /= DIMM_PID_INTERVAL;
-	deriv_p = ((s64)DIMM_PID_G_d) * (s64)derivative;
-	DBG("   deriv_p: %d\n", (int)(deriv_p >> 36));
-	sum += deriv_p;
-
-	/* Calculate the proportional term */
-	prop_p = ((s64)DIMM_PID_G_p) * (s64)(state->error_history[state->cur_sample]);
-	DBG("   prop_p: %d\n", (int)(prop_p >> 36));
-	sum += prop_p;
-
-	/* Scale sum */
-	sum >>= 36;
-
-	DBG("   sum: %d\n", (int)sum);
-	state->output = (s32)sum;
-	state->output = max(state->output, DIMM_PID_OUTPUT_MIN);
-	state->output = min(state->output, DIMM_PID_OUTPUT_MAX);
-	dimm_output_clamp = state->output;
-
-	DBG("** DIMM clamp value: %d\n", (int)state->output);
-
-	/* Backside PID is only every 5 seconds, force backside fan clamping now */
-	fan_min = (dimm_output_clamp * 100) / 14000;
-	fan_min = max(fan_min, backside_params.output_min);
-	if (backside_state.pwm < fan_min) {
-		backside_state.pwm = fan_min;
-		DBG(" -> applying clamp to backside fan now: %d  !\n", fan_min);
-		set_pwm_fan(BACKSIDE_FAN_PWM_INDEX, fan_min);
-	}
-}
-
-/*
- * Initialize the state structure for the DIMM temp control loop
- */
-static int init_dimms_state(struct dimm_pid_state *state)
-{
-	state->ticks = 1;
-	state->first = 1;
-	state->output = 4000;
-
-	state->monitor = attach_i2c_chip(XSERVE_DIMMS_LM87, "dimms_temp");
-	if (state->monitor == NULL)
-		return -ENODEV;
-
-	if (device_create_file(&of_dev->dev, &dev_attr_dimms_temperature))
-		printk(KERN_WARNING "Failed to create attribute file"
-			" for DIMM temperature\n");
-
-	return 0;
-}
-
-/*
- * Dispose of the state data for the DIMM control loop
- */
-static void dispose_dimms_state(struct dimm_pid_state *state)
-{
-	if (state->monitor == NULL)
-		return;
-
-	device_remove_file(&of_dev->dev, &dev_attr_dimms_temperature);
-
-	state->monitor = NULL;
-}
-
-/*
- * Slots fan control loop
- */
-static void do_monitor_slots(struct slots_pid_state *state)
-{
-	s32 temp, integral, derivative;
-	s64 integ_p, deriv_p, prop_p, sum;
-	int i, rc;
-
-	if (--state->ticks != 0)
-		return;
-	state->ticks = SLOTS_PID_INTERVAL;
-
-	DBG("slots:\n");
-
-	/* Check fan status */
-	rc = get_pwm_fan(SLOTS_FAN_PWM_INDEX);
-	if (rc < 0) {
-		printk(KERN_WARNING "Error %d reading slots fan !\n", rc);
-		/* XXX What do we do now ? */
-	} else
-		state->pwm = rc;
-	DBG("  current pwm: %d\n", state->pwm);
-
-	/* Get some sensor readings */
-	temp = le16_to_cpu(i2c_smbus_read_word_data(state->monitor,
-						    DS1775_TEMP)) << 8;
-	state->last_temp = temp;
-	DBG("  temp: %d.%03d, target: %d.%03d\n", FIX32TOPRINT(temp),
-	    FIX32TOPRINT(SLOTS_PID_INPUT_TARGET));
-
-	/* Store temperature and error in history array */
-	state->cur_sample = (state->cur_sample + 1) % SLOTS_PID_HISTORY_SIZE;
-	state->sample_history[state->cur_sample] = temp;
-	state->error_history[state->cur_sample] = temp - SLOTS_PID_INPUT_TARGET;
-
-	/* If first loop, fill the history table */
-	if (state->first) {
-		for (i = 0; i < (SLOTS_PID_HISTORY_SIZE - 1); i++) {
-			state->cur_sample = (state->cur_sample + 1) %
-				SLOTS_PID_HISTORY_SIZE;
-			state->sample_history[state->cur_sample] = temp;
-			state->error_history[state->cur_sample] =
-				temp - SLOTS_PID_INPUT_TARGET;
-		}
-		state->first = 0;
-	}
-
-	/* Calculate the integral term */
-	sum = 0;
-	integral = 0;
-	for (i = 0; i < SLOTS_PID_HISTORY_SIZE; i++)
-		integral += state->error_history[i];
-	integral *= SLOTS_PID_INTERVAL;
-	DBG("  integral: %08x\n", integral);
-	integ_p = ((s64)SLOTS_PID_G_r) * (s64)integral;
-	DBG("   integ_p: %d\n", (int)(integ_p >> 36));
-	sum += integ_p;
-
-	/* Calculate the derivative term */
-	derivative = state->error_history[state->cur_sample] -
-		state->error_history[(state->cur_sample + SLOTS_PID_HISTORY_SIZE - 1)
-				    % SLOTS_PID_HISTORY_SIZE];
-	derivative /= SLOTS_PID_INTERVAL;
-	deriv_p = ((s64)SLOTS_PID_G_d) * (s64)derivative;
-	DBG("   deriv_p: %d\n", (int)(deriv_p >> 36));
-	sum += deriv_p;
-
-	/* Calculate the proportional term */
-	prop_p = ((s64)SLOTS_PID_G_p) * (s64)(state->error_history[state->cur_sample]);
-	DBG("   prop_p: %d\n", (int)(prop_p >> 36));
-	sum += prop_p;
-
-	/* Scale sum */
-	sum >>= 36;
-
-	DBG("   sum: %d\n", (int)sum);
-	state->pwm = (s32)sum;
-
-	state->pwm = max(state->pwm, SLOTS_PID_OUTPUT_MIN);
-	state->pwm = min(state->pwm, SLOTS_PID_OUTPUT_MAX);
-
-	DBG("** DRIVES PWM: %d\n", (int)state->pwm);
-	set_pwm_fan(SLOTS_FAN_PWM_INDEX, state->pwm);
-}
-
-/*
- * Initialize the state structure for the slots bay fan control loop
- */
-static int init_slots_state(struct slots_pid_state *state)
-{
-	int err;
-
-	state->ticks = 1;
-	state->first = 1;
-	state->pwm = 50;
-
-	state->monitor = attach_i2c_chip(XSERVE_SLOTS_LM75, "slots_temp");
-	if (state->monitor == NULL)
-		return -ENODEV;
-
-	err = device_create_file(&of_dev->dev, &dev_attr_slots_temperature);
-	err |= device_create_file(&of_dev->dev, &dev_attr_slots_fan_pwm);
-	if (err)
-		printk(KERN_WARNING "Failed to create attribute file(s)"
-			" for slots bay fan\n");
-
-	return 0;
-}
-
-/*
- * Dispose of the state data for the slots control loop
- */
-static void dispose_slots_state(struct slots_pid_state *state)
-{
-	if (state->monitor == NULL)
-		return;
-
-	device_remove_file(&of_dev->dev, &dev_attr_slots_temperature);
-	device_remove_file(&of_dev->dev, &dev_attr_slots_fan_pwm);
-
-	state->monitor = NULL;
-}
-
-
-static int call_critical_overtemp(void)
-{
-	char *argv[] = { critical_overtemp_path, NULL };
-	static char *envp[] = { "HOME=/",
-				"TERM=linux",
-				"PATH=/sbin:/usr/sbin:/bin:/usr/bin",
-				NULL };
-
-	return call_usermodehelper(critical_overtemp_path,
-				   argv, envp, UMH_WAIT_EXEC);
-}
-
-
-/*
- * Here's the kernel thread that calls the various control loops
- */
-static int main_control_loop(void *x)
-{
-	DBG("main_control_loop started\n");
-
-	mutex_lock(&driver_lock);
-
-	if (start_fcu() < 0) {
-		printk(KERN_ERR "kfand: failed to start FCU\n");
-		mutex_unlock(&driver_lock);
-		goto out;
-	}
-
-	/* Set the PCI fan once for now on non-RackMac */
-	if (!rackmac)
-		set_pwm_fan(SLOTS_FAN_PWM_INDEX, SLOTS_FAN_DEFAULT_PWM);
-
-	/* Initialize ADCs */
-	initialize_adc(&processor_state[0]);
-	if (processor_state[1].monitor != NULL)
-		initialize_adc(&processor_state[1]);
-
-	fcu_tickle_ticks = FCU_TICKLE_TICKS;
-
-	mutex_unlock(&driver_lock);
-
-	while (state == state_attached) {
-		unsigned long elapsed, start;
-
-		start = jiffies;
-
-		mutex_lock(&driver_lock);
-
-		/* Tickle the FCU just in case */
-		if (--fcu_tickle_ticks < 0) {
-			fcu_tickle_ticks = FCU_TICKLE_TICKS;
-			tickle_fcu();
-		}
-
-		/* First, we always calculate the new DIMMs state on an Xserve */
-		if (rackmac)
-			do_monitor_dimms(&dimms_state);
-
-		/* Then, the CPUs */
-		if (cpu_pid_type == CPU_PID_TYPE_COMBINED)
-			do_monitor_cpu_combined();
-		else if (cpu_pid_type == CPU_PID_TYPE_RACKMAC) {
-			do_monitor_cpu_rack(&processor_state[0]);
-			if (processor_state[1].monitor != NULL)
-				do_monitor_cpu_rack(&processor_state[1]);
-			// better deal with UP
-		} else {
-			do_monitor_cpu_split(&processor_state[0]);
-			if (processor_state[1].monitor != NULL)
-				do_monitor_cpu_split(&processor_state[1]);
-			// better deal with UP
-		}
-		/* Then, the rest */
-		do_monitor_backside(&backside_state);
-		if (rackmac)
-			do_monitor_slots(&slots_state);
-		else
-			do_monitor_drives(&drives_state);
-		mutex_unlock(&driver_lock);
-
-		if (critical_state == 1) {
-			printk(KERN_WARNING "Temperature control detected a critical condition\n");
-			printk(KERN_WARNING "Attempting to shut down...\n");
-			if (call_critical_overtemp()) {
-				printk(KERN_WARNING "Can't call %s, power off now!\n",
-				       critical_overtemp_path);
-				machine_power_off();
-			}
-		}
-		if (critical_state > 0)
-			critical_state++;
-		if (critical_state > MAX_CRITICAL_STATE) {
-			printk(KERN_WARNING "Shutdown timed out, power off now !\n");
-			machine_power_off();
-		}
-
-		// FIXME: Deal with signals
-		elapsed = jiffies - start;
-		if (elapsed < HZ)
-			schedule_timeout_interruptible(HZ - elapsed);
-	}
-
- out:
-	DBG("main_control_loop ended\n");
-
-	ctrl_task = 0;
-	complete_and_exit(&ctrl_complete, 0);
-}
-
-/*
- * Dispose the control loops when tearing down
- */
-static void dispose_control_loops(void)
-{
-	dispose_processor_state(&processor_state[0]);
-	dispose_processor_state(&processor_state[1]);
-	dispose_backside_state(&backside_state);
-	dispose_drives_state(&drives_state);
-	dispose_slots_state(&slots_state);
-	dispose_dimms_state(&dimms_state);
-}
-
-/*
- * Create the control loops. U3-0 i2c bus is up, so we can now
- * get to the various sensors
- */
-static int create_control_loops(void)
-{
-	struct device_node *np;
-
-	/* Count CPUs from the device-tree, we don't care how many are
-	 * actually used by Linux
-	 */
-	cpu_count = 0;
-	for (np = NULL; NULL != (np = of_find_node_by_type(np, "cpu"));)
-		cpu_count++;
-
-	DBG("counted %d CPUs in the device-tree\n", cpu_count);
-
-	/* Decide the type of PID algorithm to use based on the presence of
-	 * the pumps, though that may not be the best way, that is good enough
-	 * for now
-	 */
-	if (rackmac)
-		cpu_pid_type = CPU_PID_TYPE_RACKMAC;
-	else if (of_machine_is_compatible("PowerMac7,3")
-	    && (cpu_count > 1)
-	    && fcu_fans[CPUA_PUMP_RPM_INDEX].id != FCU_FAN_ABSENT_ID
-	    && fcu_fans[CPUB_PUMP_RPM_INDEX].id != FCU_FAN_ABSENT_ID) {
-		printk(KERN_INFO "Liquid cooling pumps detected, using new algorithm !\n");
-		cpu_pid_type = CPU_PID_TYPE_COMBINED;
-	} else
-		cpu_pid_type = CPU_PID_TYPE_SPLIT;
-
-	/* Create control loops for everything. If any fail, everything
-	 * fails
-	 */
-	if (init_processor_state(&processor_state[0], 0))
-		goto fail;
-	if (cpu_pid_type == CPU_PID_TYPE_COMBINED)
-		fetch_cpu_pumps_minmax();
-
-	if (cpu_count > 1 && init_processor_state(&processor_state[1], 1))
-		goto fail;
-	if (init_backside_state(&backside_state))
-		goto fail;
-	if (rackmac && init_dimms_state(&dimms_state))
-		goto fail;
-	if (rackmac && init_slots_state(&slots_state))
-		goto fail;
-	if (!rackmac && init_drives_state(&drives_state))
-		goto fail;
-
-	DBG("all control loops up !\n");
-
-	return 0;
-	
- fail:
-	DBG("failure creating control loops, disposing\n");
-
-	dispose_control_loops();
-
-	return -ENODEV;
-}
-
-/*
- * Start the control loops after everything is up, that is create
- * the thread that will make them run
- */
-static void start_control_loops(void)
-{
-	init_completion(&ctrl_complete);
-
-	ctrl_task = kthread_run(main_control_loop, NULL, "kfand");
-}
-
-/*
- * Stop the control loops when tearing down
- */
-static void stop_control_loops(void)
-{
-	if (ctrl_task)
-		wait_for_completion(&ctrl_complete);
-}
-
-/*
- * Attach to the i2c FCU after detecting U3-1 bus
- */
-static int attach_fcu(void)
-{
-	fcu = attach_i2c_chip(FAN_CTRLER_ID, "fcu");
-	if (fcu == NULL)
-		return -ENODEV;
-
-	DBG("FCU attached\n");
-
-	return 0;
-}
-
-/*
- * Detach from the i2c FCU when tearing down
- */
-static void detach_fcu(void)
-{
-	fcu = NULL;
-}
-
-/*
- * Attach to the i2c controller. We probe the various chips based
- * on the device-tree nodes and build everything for the driver to
- * run, we then kick the driver monitoring thread
- */
-static int therm_pm72_attach(struct i2c_adapter *adapter)
-{
-	mutex_lock(&driver_lock);
-
-	/* Check state */
-	if (state == state_detached)
-		state = state_attaching;
-	if (state != state_attaching) {
-		mutex_unlock(&driver_lock);
-		return 0;
-	}
-
-	/* Check if we are looking for one of these */
-	if (u3_0 == NULL && !strcmp(adapter->name, "u3 0")) {
-		u3_0 = adapter;
-		DBG("found U3-0\n");
-		if (k2 || !rackmac)
-			if (create_control_loops())
-				u3_0 = NULL;
-	} else if (u3_1 == NULL && !strcmp(adapter->name, "u3 1")) {
-		u3_1 = adapter;
-		DBG("found U3-1, attaching FCU\n");
-		if (attach_fcu())
-			u3_1 = NULL;
-	} else if (k2 == NULL && !strcmp(adapter->name, "mac-io 0")) {
-		k2 = adapter;
-		DBG("Found K2\n");
-		if (u3_0 && rackmac)
-			if (create_control_loops())
-				k2 = NULL;
-	}
-	/* We got all we need, start control loops */
-	if (u3_0 != NULL && u3_1 != NULL && (k2 || !rackmac)) {
-		DBG("everything up, starting control loops\n");
-		state = state_attached;
-		start_control_loops();
-	}
-	mutex_unlock(&driver_lock);
-
-	return 0;
-}
-
-static int therm_pm72_probe(struct i2c_client *client,
-			    const struct i2c_device_id *id)
-{
-	/* Always succeed, the real work was done in therm_pm72_attach() */
-	return 0;
-}
-
-/*
- * Called when any of the devices which participates into thermal management
- * is going away.
- */
-static int therm_pm72_remove(struct i2c_client *client)
-{
-	struct i2c_adapter *adapter = client->adapter;
-
-	mutex_lock(&driver_lock);
-
-	if (state != state_detached)
-		state = state_detaching;
-
-	/* Stop control loops if any */
-	DBG("stopping control loops\n");
-	mutex_unlock(&driver_lock);
-	stop_control_loops();
-	mutex_lock(&driver_lock);
-
-	if (u3_0 != NULL && !strcmp(adapter->name, "u3 0")) {
-		DBG("lost U3-0, disposing control loops\n");
-		dispose_control_loops();
-		u3_0 = NULL;
-	}
-	
-	if (u3_1 != NULL && !strcmp(adapter->name, "u3 1")) {
-		DBG("lost U3-1, detaching FCU\n");
-		detach_fcu();
-		u3_1 = NULL;
-	}
-	if (u3_0 == NULL && u3_1 == NULL)
-		state = state_detached;
-
-	mutex_unlock(&driver_lock);
-
-	return 0;
-}
-
-/*
- * i2c_driver structure to attach to the host i2c controller
- */
-
-static const struct i2c_device_id therm_pm72_id[] = {
-	/*
-	 * Fake device name, thermal management is done by several
-	 * chips but we don't need to differentiate between them at
-	 * this point.
-	 */
-	{ "therm_pm72", 0 },
-	{ }
-};
-
-static struct i2c_driver therm_pm72_driver = {
-	.driver = {
-		.name	= "therm_pm72",
-	},
-	.attach_adapter	= therm_pm72_attach,
-	.probe		= therm_pm72_probe,
-	.remove		= therm_pm72_remove,
-	.id_table	= therm_pm72_id,
-};
-
-static int fan_check_loc_match(const char *loc, int fan)
-{
-	char	tmp[64];
-	char	*c, *e;
-
-	strlcpy(tmp, fcu_fans[fan].loc, 64);
-
-	c = tmp;
-	for (;;) {
-		e = strchr(c, ',');
-		if (e)
-			*e = 0;
-		if (strcmp(loc, c) == 0)
-			return 1;
-		if (e == NULL)
-			break;
-		c = e + 1;
-	}
-	return 0;
-}
-
-static void fcu_lookup_fans(struct device_node *fcu_node)
-{
-	struct device_node *np = NULL;
-	int i;
-
-	/* The table is filled by default with values that are suitable
-	 * for the old machines without device-tree informations. We scan
-	 * the device-tree and override those values with whatever is
-	 * there
-	 */
-
-	DBG("Looking up FCU controls in device-tree...\n");
-
-	while ((np = of_get_next_child(fcu_node, np)) != NULL) {
-		int type = -1;
-		const char *loc;
-		const u32 *reg;
-
-		DBG(" control: %s, type: %s\n", np->name, np->type);
-
-		/* Detect control type */
-		if (!strcmp(np->type, "fan-rpm-control") ||
-		    !strcmp(np->type, "fan-rpm"))
-			type = FCU_FAN_RPM;
-		if (!strcmp(np->type, "fan-pwm-control") ||
-		    !strcmp(np->type, "fan-pwm"))
-			type = FCU_FAN_PWM;
-		/* Only care about fans for now */
-		if (type == -1)
-			continue;
-
-		/* Lookup for a matching location */
-		loc = of_get_property(np, "location", NULL);
-		reg = of_get_property(np, "reg", NULL);
-		if (loc == NULL || reg == NULL)
-			continue;
-		DBG(" matching location: %s, reg: 0x%08x\n", loc, *reg);
-
-		for (i = 0; i < FCU_FAN_COUNT; i++) {
-			int fan_id;
-
-			if (!fan_check_loc_match(loc, i))
-				continue;
-			DBG(" location match, index: %d\n", i);
-			fcu_fans[i].id = FCU_FAN_ABSENT_ID;
-			if (type != fcu_fans[i].type) {
-				printk(KERN_WARNING "therm_pm72: Fan type mismatch "
-				       "in device-tree for %s\n", np->full_name);
-				break;
-			}
-			if (type == FCU_FAN_RPM)
-				fan_id = ((*reg) - 0x10) / 2;
-			else
-				fan_id = ((*reg) - 0x30) / 2;
-			if (fan_id > 7) {
-				printk(KERN_WARNING "therm_pm72: Can't parse "
-				       "fan ID in device-tree for %s\n", np->full_name);
-				break;
-			}
-			DBG(" fan id -> %d, type -> %d\n", fan_id, type);
-			fcu_fans[i].id = fan_id;
-		}
-	}
-
-	/* Now dump the array */
-	printk(KERN_INFO "Detected fan controls:\n");
-	for (i = 0; i < FCU_FAN_COUNT; i++) {
-		if (fcu_fans[i].id == FCU_FAN_ABSENT_ID)
-			continue;
-		printk(KERN_INFO "  %d: %s fan, id %d, location: %s\n", i,
-		       fcu_fans[i].type == FCU_FAN_RPM ? "RPM" : "PWM",
-		       fcu_fans[i].id, fcu_fans[i].loc);
-	}
-}
-
-static int fcu_of_probe(struct platform_device* dev)
-{
-	state = state_detached;
-	of_dev = dev;
-
-	dev_info(&dev->dev, "PowerMac G5 Thermal control driver %s\n", VERSION);
-
-	/* Lookup the fans in the device tree */
-	fcu_lookup_fans(dev->dev.of_node);
-
-	/* Add the driver */
-	return i2c_add_driver(&therm_pm72_driver);
-}
-
-static int fcu_of_remove(struct platform_device* dev)
-{
-	i2c_del_driver(&therm_pm72_driver);
-
-	return 0;
-}
-
-static const struct of_device_id fcu_match[] = 
-{
-	{
-	.type		= "fcu",
-	},
-	{},
-};
-MODULE_DEVICE_TABLE(of, fcu_match);
-
-static struct platform_driver fcu_of_platform_driver = 
-{
-	.driver = {
-		.name = "temperature",
-		.owner = THIS_MODULE,
-		.of_match_table = fcu_match,
-	},
-	.probe		= fcu_of_probe,
-	.remove		= fcu_of_remove
-};
-
-/*
- * Check machine type, attach to i2c controller
- */
-static int __init therm_pm72_init(void)
-{
-	rackmac = of_machine_is_compatible("RackMac3,1");
-
-	if (!of_machine_is_compatible("PowerMac7,2") &&
-	    !of_machine_is_compatible("PowerMac7,3") &&
-	    !rackmac)
-	    	return -ENODEV;
-
-	return platform_driver_register(&fcu_of_platform_driver);
-}
-
-static void __exit therm_pm72_exit(void)
-{
-	platform_driver_unregister(&fcu_of_platform_driver);
-}
-
-module_init(therm_pm72_init);
-module_exit(therm_pm72_exit);
-
-MODULE_AUTHOR("Benjamin Herrenschmidt <benh@kernel.crashing.org>");
-MODULE_DESCRIPTION("Driver for Apple's PowerMac G5 thermal control");
-MODULE_LICENSE("GPL");
-
diff --git a/drivers/macintosh/therm_pm72.h b/drivers/macintosh/therm_pm72.h
deleted file mode 100644
index df3680e2a22f..000000000000
--- a/drivers/macintosh/therm_pm72.h
+++ /dev/null
@@ -1,326 +0,0 @@
-#ifndef __THERM_PMAC_7_2_H__
-#define __THERM_PMAC_7_2_H__
-
-typedef unsigned short fu16;
-typedef int fs32;
-typedef short fs16;
-
-struct mpu_data
-{
-	u8	signature;		/* 0x00 - EEPROM sig. */
-	u8	bytes_used;		/* 0x01 - Bytes used in eeprom (160 ?) */
-	u8	size;			/* 0x02 - EEPROM size (256 ?) */
-	u8	version;		/* 0x03 - EEPROM version */
-	u32	data_revision;		/* 0x04 - Dataset revision */
-	u8	processor_bin_code[3];	/* 0x08 - Processor BIN code */
-	u8	bin_code_expansion;	/* 0x0b - ??? (padding ?) */
-	u8	processor_num;		/* 0x0c - Number of CPUs on this MPU */
-	u8	input_mul_bus_div;	/* 0x0d - Clock input multiplier/bus divider */
-	u8	reserved1[2];		/* 0x0e - */
-	u32	input_clk_freq_high;	/* 0x10 - Input clock frequency high */
-	u8	cpu_nb_target_cycles;	/* 0x14 - ??? */
-	u8	cpu_statlat;		/* 0x15 - ??? */
-	u8	cpu_snooplat;		/* 0x16 - ??? */
-	u8	cpu_snoopacc;		/* 0x17 - ??? */
-	u8	nb_paamwin;		/* 0x18 - ??? */
-	u8	nb_statlat;		/* 0x19 - ??? */
-	u8	nb_snooplat;		/* 0x1a - ??? */
-	u8	nb_snoopwin;		/* 0x1b - ??? */
-	u8	api_bus_mode;		/* 0x1c - ??? */
-	u8	reserved2[3];		/* 0x1d - */
-	u32	input_clk_freq_low;	/* 0x20 - Input clock frequency low */
-	u8	processor_card_slot;	/* 0x24 - Processor card slot number */
-	u8	reserved3[2];		/* 0x25 - */
-	u8	padjmax;       		/* 0x27 - Max power adjustment (Not in OF!) */
-	u8	ttarget;		/* 0x28 - Target temperature */
-	u8	tmax;			/* 0x29 - Max temperature */
-	u8	pmaxh;			/* 0x2a - Max power */
-	u8	tguardband;		/* 0x2b - Guardband temp ??? Hist. len in OSX */
-	fs32	pid_gp;			/* 0x2c - PID proportional gain */
-	fs32	pid_gr;			/* 0x30 - PID reset gain */
-	fs32	pid_gd;			/* 0x34 - PID derivative gain */
-	fu16	voph;			/* 0x38 - Vop High */
-	fu16	vopl;			/* 0x3a - Vop Low */
-	fs16	nactual_die;		/* 0x3c - nActual Die */
-	fs16	nactual_heatsink;	/* 0x3e - nActual Heatsink */
-	fs16	nactual_system;		/* 0x40 - nActual System */
-	u16	calibration_flags;	/* 0x42 - Calibration flags */
-	fu16	mdiode;			/* 0x44 - Diode M value (scaling factor) */
-	fs16	bdiode;			/* 0x46 - Diode B value (offset) */
-	fs32	theta_heat_sink;	/* 0x48 - Theta heat sink */
-	u16	rminn_intake_fan;	/* 0x4c - Intake fan min RPM */
-	u16	rmaxn_intake_fan;	/* 0x4e - Intake fan max RPM */
-	u16	rminn_exhaust_fan;	/* 0x50 - Exhaust fan min RPM */
-	u16	rmaxn_exhaust_fan;	/* 0x52 - Exhaust fan max RPM */
-	u8	processor_part_num[8];	/* 0x54 - Processor part number XX pumps min/max */
-	u32	processor_lot_num;	/* 0x5c - Processor lot number */
-	u8	orig_card_sernum[0x10];	/* 0x60 - Card original serial number */
-	u8	curr_card_sernum[0x10];	/* 0x70 - Card current serial number */
-	u8	mlb_sernum[0x18];	/* 0x80 - MLB serial number */
-	u32	checksum1;		/* 0x98 - */
-	u32	checksum2;		/* 0x9c - */	
-}; /* Total size = 0xa0 */
-
-/* Display a 16.16 fixed point value */
-#define FIX32TOPRINT(f)	((f) >> 16),((((f) & 0xffff) * 1000) >> 16)
-
-/*
- * Maximum number of seconds to be in critical state (after a
- * normal shutdown attempt). If the machine isn't down after
- * this counter elapses, we force an immediate machine power
- * off.
- */
-#define MAX_CRITICAL_STATE			30
-static char * critical_overtemp_path = "/sbin/critical_overtemp";
-
-/*
- * This option is "weird" :) Basically, if you define this to 1
- * the control loop for the RPMs fans (not PWMs) will apply the
- * correction factor obtained from the PID to the _actual_ RPM
- * speed read from the FCU.
- * If you define the below constant to 0, then it will be
- * applied to the setpoint RPM speed, that is basically the
- * speed we proviously "asked" for.
- *
- * I'm not sure which of these Apple's algorithm is supposed
- * to use
- */
-#define RPM_PID_USE_ACTUAL_SPEED		0
-
-/*
- * i2c IDs. Currently, we hard code those and assume that
- * the FCU is on U3 bus 1 while all sensors are on U3 bus
- * 0. This appear to be safe enough for this first version
- * of the driver, though I would accept any clean patch
- * doing a better use of the device-tree without turning the
- * while i2c registration mechanism into a racy mess
- *
- * Note: Xserve changed this. We have some bits on the K2 bus,
- * which I arbitrarily set to 0x200. Ultimately, we really want
- * too lookup these in the device-tree though
- */
-#define FAN_CTRLER_ID		0x15e
-#define SUPPLY_MONITOR_ID      	0x58
-#define SUPPLY_MONITORB_ID     	0x5a
-#define DRIVES_DALLAS_ID	0x94
-#define BACKSIDE_MAX_ID		0x98
-#define XSERVE_DIMMS_LM87	0x25a
-#define XSERVE_SLOTS_LM75	0x290
-
-/*
- * Some MAX6690, DS1775, LM87 register definitions
- */
-#define MAX6690_INT_TEMP	0
-#define MAX6690_EXT_TEMP	1
-#define DS1775_TEMP		0
-#define LM87_INT_TEMP		0x27
-
-/*
- * Scaling factors for the AD7417 ADC converters (except
- * for the CPU diode which is obtained from the EEPROM).
- * Those values are obtained from the property list of
- * the darwin driver
- */
-#define ADC_12V_CURRENT_SCALE	0x0320	/* _AD2 */
-#define ADC_CPU_VOLTAGE_SCALE	0x00a0	/* _AD3 */
-#define ADC_CPU_CURRENT_SCALE	0x1f40	/* _AD4 */
-
-/*
- * PID factors for the U3/Backside fan control loop. We have 2 sets
- * of values here, one set for U3 and one set for U3H
- */
-#define BACKSIDE_FAN_PWM_DEFAULT_ID	1
-#define BACKSIDE_FAN_PWM_INDEX		0
-#define BACKSIDE_PID_U3_G_d		0x02800000
-#define BACKSIDE_PID_U3H_G_d		0x01400000
-#define BACKSIDE_PID_RACK_G_d		0x00500000
-#define BACKSIDE_PID_G_p		0x00500000
-#define BACKSIDE_PID_RACK_G_p		0x0004cccc
-#define BACKSIDE_PID_G_r		0x00000000
-#define BACKSIDE_PID_U3_INPUT_TARGET	0x00410000
-#define BACKSIDE_PID_U3H_INPUT_TARGET	0x004b0000
-#define BACKSIDE_PID_RACK_INPUT_TARGET	0x00460000
-#define BACKSIDE_PID_INTERVAL		5
-#define BACKSIDE_PID_RACK_INTERVAL	1
-#define BACKSIDE_PID_OUTPUT_MAX		100
-#define BACKSIDE_PID_U3_OUTPUT_MIN	20
-#define BACKSIDE_PID_U3H_OUTPUT_MIN	20
-#define BACKSIDE_PID_HISTORY_SIZE	2
-
-struct basckside_pid_params
-{
-	s32			G_d;
-	s32			G_p;
-	s32			G_r;
-	s32			input_target;
-	s32			output_min;
-	s32			output_max;
-	s32			interval;
-	int			additive;
-};
-
-struct backside_pid_state
-{
-	int			ticks;
-	struct i2c_client *	monitor;
-	s32		       	sample_history[BACKSIDE_PID_HISTORY_SIZE];
-	s32			error_history[BACKSIDE_PID_HISTORY_SIZE];
-	int			cur_sample;
-	s32			last_temp;
-	int			pwm;
-	int			first;
-};
-
-/*
- * PID factors for the Drive Bay fan control loop
- */
-#define DRIVES_FAN_RPM_DEFAULT_ID	2
-#define DRIVES_FAN_RPM_INDEX		1
-#define DRIVES_PID_G_d			0x01e00000
-#define DRIVES_PID_G_p			0x00500000
-#define DRIVES_PID_G_r			0x00000000
-#define DRIVES_PID_INPUT_TARGET		0x00280000
-#define DRIVES_PID_INTERVAL    		5
-#define DRIVES_PID_OUTPUT_MAX		4000
-#define DRIVES_PID_OUTPUT_MIN		300
-#define DRIVES_PID_HISTORY_SIZE		2
-
-struct drives_pid_state
-{
-	int			ticks;
-	struct i2c_client *	monitor;
-	s32	       		sample_history[BACKSIDE_PID_HISTORY_SIZE];
-	s32			error_history[BACKSIDE_PID_HISTORY_SIZE];
-	int			cur_sample;
-	s32			last_temp;
-	int			rpm;
-	int			first;
-};
-
-#define SLOTS_FAN_PWM_DEFAULT_ID	2
-#define SLOTS_FAN_PWM_INDEX		2
-#define	SLOTS_FAN_DEFAULT_PWM		40 /* Do better here ! */
-
-
-/*
- * PID factors for the Xserve DIMM control loop
- */
-#define DIMM_PID_G_d			0
-#define DIMM_PID_G_p			0
-#define DIMM_PID_G_r			0x06553600
-#define DIMM_PID_INPUT_TARGET		3276800
-#define DIMM_PID_INTERVAL    		1
-#define DIMM_PID_OUTPUT_MAX		14000
-#define DIMM_PID_OUTPUT_MIN		4000
-#define DIMM_PID_HISTORY_SIZE		20
-
-struct dimm_pid_state
-{
-	int			ticks;
-	struct i2c_client *	monitor;
-	s32	       		sample_history[DIMM_PID_HISTORY_SIZE];
-	s32			error_history[DIMM_PID_HISTORY_SIZE];
-	int			cur_sample;
-	s32			last_temp;
-	int			first;
-	int			output;
-};
-
-
-/*
- * PID factors for the Xserve Slots control loop
- */
-#define SLOTS_PID_G_d			0
-#define SLOTS_PID_G_p			0
-#define SLOTS_PID_G_r			0x00100000
-#define SLOTS_PID_INPUT_TARGET		3200000
-#define SLOTS_PID_INTERVAL    		1
-#define SLOTS_PID_OUTPUT_MAX		100
-#define SLOTS_PID_OUTPUT_MIN		20
-#define SLOTS_PID_HISTORY_SIZE		20
-
-struct slots_pid_state
-{
-	int			ticks;
-	struct i2c_client *	monitor;
-	s32	       		sample_history[SLOTS_PID_HISTORY_SIZE];
-	s32			error_history[SLOTS_PID_HISTORY_SIZE];
-	int			cur_sample;
-	s32			last_temp;
-	int			first;
-	int			pwm;
-};
-
-
-
-/* Desktops */
-
-#define CPUA_INTAKE_FAN_RPM_DEFAULT_ID	3
-#define CPUA_EXHAUST_FAN_RPM_DEFAULT_ID	4
-#define CPUB_INTAKE_FAN_RPM_DEFAULT_ID	5
-#define CPUB_EXHAUST_FAN_RPM_DEFAULT_ID	6
-
-#define CPUA_INTAKE_FAN_RPM_INDEX	3
-#define CPUA_EXHAUST_FAN_RPM_INDEX	4
-#define CPUB_INTAKE_FAN_RPM_INDEX	5
-#define CPUB_EXHAUST_FAN_RPM_INDEX	6
-
-#define CPU_INTAKE_SCALE		0x0000f852
-#define CPU_TEMP_HISTORY_SIZE		2
-#define CPU_POWER_HISTORY_SIZE		10
-#define CPU_PID_INTERVAL		1
-#define CPU_MAX_OVERTEMP		90
-
-#define CPUA_PUMP_RPM_INDEX		7
-#define CPUB_PUMP_RPM_INDEX		8
-#define CPU_PUMP_OUTPUT_MAX		3200
-#define CPU_PUMP_OUTPUT_MIN		1250
-
-/* Xserve */
-#define CPU_A1_FAN_RPM_INDEX		9
-#define CPU_A2_FAN_RPM_INDEX		10
-#define CPU_A3_FAN_RPM_INDEX		11
-#define CPU_B1_FAN_RPM_INDEX		12
-#define CPU_B2_FAN_RPM_INDEX		13
-#define CPU_B3_FAN_RPM_INDEX		14
-
-
-struct cpu_pid_state
-{
-	int			index;
-	struct i2c_client *	monitor;
-	struct mpu_data		mpu;
-	int			overtemp;
-	s32	       		temp_history[CPU_TEMP_HISTORY_SIZE];
-	int			cur_temp;
-	s32			power_history[CPU_POWER_HISTORY_SIZE];
-	s32			error_history[CPU_POWER_HISTORY_SIZE];
-	int			cur_power;
-	int			count_power;
-	int			rpm;
-	int			intake_rpm;
-	s32			voltage;
-	s32			current_a;
-	s32			last_temp;
-	s32			last_power;
-	int			first;
-	u8			adc_config;
-	s32			pump_min;
-	s32			pump_max;
-};
-
-/* Tickle FCU every 10 seconds */
-#define FCU_TICKLE_TICKS	10
-
-/*
- * Driver state
- */
-enum {
-	state_detached,
-	state_attaching,
-	state_attached,
-	state_detaching,
-};
-
-
-#endif /* __THERM_PMAC_7_2_H__ */
diff --git a/drivers/macintosh/therm_windtunnel.c b/drivers/macintosh/therm_windtunnel.c
index 3b4a157714b1..109dcaa15934 100644
--- a/drivers/macintosh/therm_windtunnel.c
+++ b/drivers/macintosh/therm_windtunnel.c
@@ -463,7 +463,6 @@ static const struct of_device_id therm_of_match[] = {{
 static struct platform_driver therm_of_driver = {
 	.driver = {
 		.name = "temperature",
-		.owner = THIS_MODULE,
 		.of_match_table = therm_of_match,
 	},
 	.probe		= therm_of_probe,
diff --git a/drivers/macintosh/via-cuda.c b/drivers/macintosh/via-cuda.c
index d61f271d2207..bad18130f125 100644
--- a/drivers/macintosh/via-cuda.c
+++ b/drivers/macintosh/via-cuda.c
@@ -379,6 +379,7 @@ cuda_request(struct adb_request *req, void (*done)(struct adb_request *),
     req->reply_expected = 1;
     return cuda_write(req);
 }
+EXPORT_SYMBOL(cuda_request);
 
 static int
 cuda_write(struct adb_request *req)
@@ -441,6 +442,7 @@ cuda_poll(void)
     if (cuda_irq)
 	enable_irq(cuda_irq);
 }
+EXPORT_SYMBOL(cuda_poll);
 
 static irqreturn_t
 cuda_interrupt(int irq, void *arg)
diff --git a/drivers/macintosh/windfarm_pm81.c b/drivers/macintosh/windfarm_pm81.c
index 2a5e1b15b1d2..93faf298a3c5 100644
--- a/drivers/macintosh/windfarm_pm81.c
+++ b/drivers/macintosh/windfarm_pm81.c
@@ -770,7 +770,6 @@ static struct platform_driver wf_smu_driver = {
         .remove = wf_smu_remove,
 	.driver = {
 		.name = "windfarm",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/macintosh/windfarm_pm91.c b/drivers/macintosh/windfarm_pm91.c
index a8ac66cd3b13..81fdf40c5b82 100644
--- a/drivers/macintosh/windfarm_pm91.c
+++ b/drivers/macintosh/windfarm_pm91.c
@@ -699,7 +699,6 @@ static struct platform_driver wf_smu_driver = {
         .remove = wf_smu_remove,
 	.driver = {
 		.name = "windfarm",
-		.owner	= THIS_MODULE,
 	},
 };
 
diff --git a/drivers/mailbox/Kconfig b/drivers/mailbox/Kconfig
index 9fd9c6717e0c..c04fed9eb15d 100644
--- a/drivers/mailbox/Kconfig
+++ b/drivers/mailbox/Kconfig
@@ -33,4 +33,16 @@ config OMAP_MBOX_KFIFO_SIZE
 	  Specify the default size of mailbox's kfifo buffers (bytes).
 	  This can also be changed at runtime (via the mbox_kfifo_size
 	  module parameter).
+
+config PCC
+	bool "Platform Communication Channel Driver"
+	depends on ACPI
+	help
+	  ACPI 5.0+ spec defines a generic mode of communication
+	  between the OS and a platform such as the BMC. This medium
+	  (PCC) is typically used by CPPC (ACPI CPU Performance management),
+	  RAS (ACPI reliability protocol) and MPST (ACPI Memory power
+	  states). Select this driver if your platform implements the
+	  PCC clients mentioned above.
+
 endif
diff --git a/drivers/mailbox/Makefile b/drivers/mailbox/Makefile
index 6d184dbcaca8..dd412c22208b 100644
--- a/drivers/mailbox/Makefile
+++ b/drivers/mailbox/Makefile
@@ -1,3 +1,9 @@
+# Generic MAILBOX API
+
+obj-$(CONFIG_MAILBOX)		+= mailbox.o
+
 obj-$(CONFIG_PL320_MBOX)	+= pl320-ipc.o
 
 obj-$(CONFIG_OMAP2PLUS_MBOX)	+= omap-mailbox.o
+
+obj-$(CONFIG_PCC)		+= pcc.o
diff --git a/drivers/mailbox/mailbox.c b/drivers/mailbox/mailbox.c
new file mode 100644
index 000000000000..59aad4d5da53
--- /dev/null
+++ b/drivers/mailbox/mailbox.c
@@ -0,0 +1,468 @@
+/*
+ * Mailbox: Common code for Mailbox controllers and users
+ *
+ * Copyright (C) 2013-2014 Linaro Ltd.
+ * Author: Jassi Brar <jassisinghbrar@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/interrupt.h>
+#include <linux/spinlock.h>
+#include <linux/mutex.h>
+#include <linux/delay.h>
+#include <linux/slab.h>
+#include <linux/err.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/bitops.h>
+#include <linux/mailbox_client.h>
+#include <linux/mailbox_controller.h>
+
+#include "mailbox.h"
+
+static LIST_HEAD(mbox_cons);
+static DEFINE_MUTEX(con_mutex);
+
+static void poll_txdone(unsigned long data);
+
+static int add_to_rbuf(struct mbox_chan *chan, void *mssg)
+{
+	int idx;
+	unsigned long flags;
+
+	spin_lock_irqsave(&chan->lock, flags);
+
+	/* See if there is any space left */
+	if (chan->msg_count == MBOX_TX_QUEUE_LEN) {
+		spin_unlock_irqrestore(&chan->lock, flags);
+		return -ENOBUFS;
+	}
+
+	idx = chan->msg_free;
+	chan->msg_data[idx] = mssg;
+	chan->msg_count++;
+
+	if (idx == MBOX_TX_QUEUE_LEN - 1)
+		chan->msg_free = 0;
+	else
+		chan->msg_free++;
+
+	spin_unlock_irqrestore(&chan->lock, flags);
+
+	return idx;
+}
+
+static void msg_submit(struct mbox_chan *chan)
+{
+	unsigned count, idx;
+	unsigned long flags;
+	void *data;
+	int err = -EBUSY;
+
+	spin_lock_irqsave(&chan->lock, flags);
+
+	if (!chan->msg_count || chan->active_req)
+		goto exit;
+
+	count = chan->msg_count;
+	idx = chan->msg_free;
+	if (idx >= count)
+		idx -= count;
+	else
+		idx += MBOX_TX_QUEUE_LEN - count;
+
+	data = chan->msg_data[idx];
+
+	if (chan->cl->tx_prepare)
+		chan->cl->tx_prepare(chan->cl, data);
+	/* Try to submit a message to the MBOX controller */
+	err = chan->mbox->ops->send_data(chan, data);
+	if (!err) {
+		chan->active_req = data;
+		chan->msg_count--;
+	}
+exit:
+	spin_unlock_irqrestore(&chan->lock, flags);
+
+	if (!err && chan->txdone_method == TXDONE_BY_POLL)
+		poll_txdone((unsigned long)chan->mbox);
+}
+
+static void tx_tick(struct mbox_chan *chan, int r)
+{
+	unsigned long flags;
+	void *mssg;
+
+	spin_lock_irqsave(&chan->lock, flags);
+	mssg = chan->active_req;
+	chan->active_req = NULL;
+	spin_unlock_irqrestore(&chan->lock, flags);
+
+	/* Submit next message */
+	msg_submit(chan);
+
+	/* Notify the client */
+	if (mssg && chan->cl->tx_done)
+		chan->cl->tx_done(chan->cl, mssg, r);
+
+	if (chan->cl->tx_block)
+		complete(&chan->tx_complete);
+}
+
+static void poll_txdone(unsigned long data)
+{
+	struct mbox_controller *mbox = (struct mbox_controller *)data;
+	bool txdone, resched = false;
+	int i;
+
+	for (i = 0; i < mbox->num_chans; i++) {
+		struct mbox_chan *chan = &mbox->chans[i];
+
+		if (chan->active_req && chan->cl) {
+			txdone = chan->mbox->ops->last_tx_done(chan);
+			if (txdone)
+				tx_tick(chan, 0);
+			else
+				resched = true;
+		}
+	}
+
+	if (resched)
+		mod_timer(&mbox->poll, jiffies +
+				msecs_to_jiffies(mbox->txpoll_period));
+}
+
+/**
+ * mbox_chan_received_data - A way for controller driver to push data
+ *				received from remote to the upper layer.
+ * @chan: Pointer to the mailbox channel on which RX happened.
+ * @mssg: Client specific message typecasted as void *
+ *
+ * After startup and before shutdown any data received on the chan
+ * is passed on to the API via atomic mbox_chan_received_data().
+ * The controller should ACK the RX only after this call returns.
+ */
+void mbox_chan_received_data(struct mbox_chan *chan, void *mssg)
+{
+	/* No buffering the received data */
+	if (chan->cl->rx_callback)
+		chan->cl->rx_callback(chan->cl, mssg);
+}
+EXPORT_SYMBOL_GPL(mbox_chan_received_data);
+
+/**
+ * mbox_chan_txdone - A way for controller driver to notify the
+ *			framework that the last TX has completed.
+ * @chan: Pointer to the mailbox chan on which TX happened.
+ * @r: Status of last TX - OK or ERROR
+ *
+ * The controller that has IRQ for TX ACK calls this atomic API
+ * to tick the TX state machine. It works only if txdone_irq
+ * is set by the controller.
+ */
+void mbox_chan_txdone(struct mbox_chan *chan, int r)
+{
+	if (unlikely(!(chan->txdone_method & TXDONE_BY_IRQ))) {
+		dev_err(chan->mbox->dev,
+		       "Controller can't run the TX ticker\n");
+		return;
+	}
+
+	tx_tick(chan, r);
+}
+EXPORT_SYMBOL_GPL(mbox_chan_txdone);
+
+/**
+ * mbox_client_txdone - The way for a client to run the TX state machine.
+ * @chan: Mailbox channel assigned to this client.
+ * @r: Success status of last transmission.
+ *
+ * The client/protocol had received some 'ACK' packet and it notifies
+ * the API that the last packet was sent successfully. This only works
+ * if the controller can't sense TX-Done.
+ */
+void mbox_client_txdone(struct mbox_chan *chan, int r)
+{
+	if (unlikely(!(chan->txdone_method & TXDONE_BY_ACK))) {
+		dev_err(chan->mbox->dev, "Client can't run the TX ticker\n");
+		return;
+	}
+
+	tx_tick(chan, r);
+}
+EXPORT_SYMBOL_GPL(mbox_client_txdone);
+
+/**
+ * mbox_client_peek_data - A way for client driver to pull data
+ *			received from remote by the controller.
+ * @chan: Mailbox channel assigned to this client.
+ *
+ * A poke to controller driver for any received data.
+ * The data is actually passed onto client via the
+ * mbox_chan_received_data()
+ * The call can be made from atomic context, so the controller's
+ * implementation of peek_data() must not sleep.
+ *
+ * Return: True, if controller has, and is going to push after this,
+ *          some data.
+ *         False, if controller doesn't have any data to be read.
+ */
+bool mbox_client_peek_data(struct mbox_chan *chan)
+{
+	if (chan->mbox->ops->peek_data)
+		return chan->mbox->ops->peek_data(chan);
+
+	return false;
+}
+EXPORT_SYMBOL_GPL(mbox_client_peek_data);
+
+/**
+ * mbox_send_message -	For client to submit a message to be
+ *				sent to the remote.
+ * @chan: Mailbox channel assigned to this client.
+ * @mssg: Client specific message typecasted.
+ *
+ * For client to submit data to the controller destined for a remote
+ * processor. If the client had set 'tx_block', the call will return
+ * either when the remote receives the data or when 'tx_tout' millisecs
+ * run out.
+ *  In non-blocking mode, the requests are buffered by the API and a
+ * non-negative token is returned for each queued request. If the request
+ * is not queued, a negative token is returned. Upon failure or successful
+ * TX, the API calls 'tx_done' from atomic context, from which the client
+ * could submit yet another request.
+ * The pointer to message should be preserved until it is sent
+ * over the chan, i.e, tx_done() is made.
+ * This function could be called from atomic context as it simply
+ * queues the data and returns a token against the request.
+ *
+ * Return: Non-negative integer for successful submission (non-blocking mode)
+ *	or transmission over chan (blocking mode).
+ *	Negative value denotes failure.
+ */
+int mbox_send_message(struct mbox_chan *chan, void *mssg)
+{
+	int t;
+
+	if (!chan || !chan->cl)
+		return -EINVAL;
+
+	t = add_to_rbuf(chan, mssg);
+	if (t < 0) {
+		dev_err(chan->mbox->dev, "Try increasing MBOX_TX_QUEUE_LEN\n");
+		return t;
+	}
+
+	msg_submit(chan);
+
+	if (chan->cl->tx_block && chan->active_req) {
+		unsigned long wait;
+		int ret;
+
+		if (!chan->cl->tx_tout) /* wait forever */
+			wait = msecs_to_jiffies(3600000);
+		else
+			wait = msecs_to_jiffies(chan->cl->tx_tout);
+
+		ret = wait_for_completion_timeout(&chan->tx_complete, wait);
+		if (ret == 0) {
+			t = -EIO;
+			tx_tick(chan, -EIO);
+		}
+	}
+
+	return t;
+}
+EXPORT_SYMBOL_GPL(mbox_send_message);
+
+/**
+ * mbox_request_channel - Request a mailbox channel.
+ * @cl: Identity of the client requesting the channel.
+ * @index: Index of mailbox specifier in 'mboxes' property.
+ *
+ * The Client specifies its requirements and capabilities while asking for
+ * a mailbox channel. It can't be called from atomic context.
+ * The channel is exclusively allocated and can't be used by another
+ * client before the owner calls mbox_free_channel.
+ * After assignment, any packet received on this channel will be
+ * handed over to the client via the 'rx_callback'.
+ * The framework holds reference to the client, so the mbox_client
+ * structure shouldn't be modified until the mbox_free_channel returns.
+ *
+ * Return: Pointer to the channel assigned to the client if successful.
+ *		ERR_PTR for request failure.
+ */
+struct mbox_chan *mbox_request_channel(struct mbox_client *cl, int index)
+{
+	struct device *dev = cl->dev;
+	struct mbox_controller *mbox;
+	struct of_phandle_args spec;
+	struct mbox_chan *chan;
+	unsigned long flags;
+	int ret;
+
+	if (!dev || !dev->of_node) {
+		pr_debug("%s: No owner device node\n", __func__);
+		return ERR_PTR(-ENODEV);
+	}
+
+	mutex_lock(&con_mutex);
+
+	if (of_parse_phandle_with_args(dev->of_node, "mboxes",
+				       "#mbox-cells", index, &spec)) {
+		dev_dbg(dev, "%s: can't parse \"mboxes\" property\n", __func__);
+		mutex_unlock(&con_mutex);
+		return ERR_PTR(-ENODEV);
+	}
+
+	chan = NULL;
+	list_for_each_entry(mbox, &mbox_cons, node)
+		if (mbox->dev->of_node == spec.np) {
+			chan = mbox->of_xlate(mbox, &spec);
+			break;
+		}
+
+	of_node_put(spec.np);
+
+	if (!chan || chan->cl || !try_module_get(mbox->dev->driver->owner)) {
+		dev_dbg(dev, "%s: mailbox not free\n", __func__);
+		mutex_unlock(&con_mutex);
+		return ERR_PTR(-EBUSY);
+	}
+
+	spin_lock_irqsave(&chan->lock, flags);
+	chan->msg_free = 0;
+	chan->msg_count = 0;
+	chan->active_req = NULL;
+	chan->cl = cl;
+	init_completion(&chan->tx_complete);
+
+	if (chan->txdone_method	== TXDONE_BY_POLL && cl->knows_txdone)
+		chan->txdone_method |= TXDONE_BY_ACK;
+
+	spin_unlock_irqrestore(&chan->lock, flags);
+
+	ret = chan->mbox->ops->startup(chan);
+	if (ret) {
+		dev_err(dev, "Unable to startup the chan (%d)\n", ret);
+		mbox_free_channel(chan);
+		chan = ERR_PTR(ret);
+	}
+
+	mutex_unlock(&con_mutex);
+	return chan;
+}
+EXPORT_SYMBOL_GPL(mbox_request_channel);
+
+/**
+ * mbox_free_channel - The client relinquishes control of a mailbox
+ *			channel by this call.
+ * @chan: The mailbox channel to be freed.
+ */
+void mbox_free_channel(struct mbox_chan *chan)
+{
+	unsigned long flags;
+
+	if (!chan || !chan->cl)
+		return;
+
+	chan->mbox->ops->shutdown(chan);
+
+	/* The queued TX requests are simply aborted, no callbacks are made */
+	spin_lock_irqsave(&chan->lock, flags);
+	chan->cl = NULL;
+	chan->active_req = NULL;
+	if (chan->txdone_method == (TXDONE_BY_POLL | TXDONE_BY_ACK))
+		chan->txdone_method = TXDONE_BY_POLL;
+
+	module_put(chan->mbox->dev->driver->owner);
+	spin_unlock_irqrestore(&chan->lock, flags);
+}
+EXPORT_SYMBOL_GPL(mbox_free_channel);
+
+static struct mbox_chan *
+of_mbox_index_xlate(struct mbox_controller *mbox,
+		    const struct of_phandle_args *sp)
+{
+	int ind = sp->args[0];
+
+	if (ind >= mbox->num_chans)
+		return NULL;
+
+	return &mbox->chans[ind];
+}
+
+/**
+ * mbox_controller_register - Register the mailbox controller
+ * @mbox:	Pointer to the mailbox controller.
+ *
+ * The controller driver registers its communication channels
+ */
+int mbox_controller_register(struct mbox_controller *mbox)
+{
+	int i, txdone;
+
+	/* Sanity check */
+	if (!mbox || !mbox->dev || !mbox->ops || !mbox->num_chans)
+		return -EINVAL;
+
+	if (mbox->txdone_irq)
+		txdone = TXDONE_BY_IRQ;
+	else if (mbox->txdone_poll)
+		txdone = TXDONE_BY_POLL;
+	else /* It has to be ACK then */
+		txdone = TXDONE_BY_ACK;
+
+	if (txdone == TXDONE_BY_POLL) {
+		mbox->poll.function = &poll_txdone;
+		mbox->poll.data = (unsigned long)mbox;
+		init_timer(&mbox->poll);
+	}
+
+	for (i = 0; i < mbox->num_chans; i++) {
+		struct mbox_chan *chan = &mbox->chans[i];
+
+		chan->cl = NULL;
+		chan->mbox = mbox;
+		chan->txdone_method = txdone;
+		spin_lock_init(&chan->lock);
+	}
+
+	if (!mbox->of_xlate)
+		mbox->of_xlate = of_mbox_index_xlate;
+
+	mutex_lock(&con_mutex);
+	list_add_tail(&mbox->node, &mbox_cons);
+	mutex_unlock(&con_mutex);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(mbox_controller_register);
+
+/**
+ * mbox_controller_unregister - Unregister the mailbox controller
+ * @mbox:	Pointer to the mailbox controller.
+ */
+void mbox_controller_unregister(struct mbox_controller *mbox)
+{
+	int i;
+
+	if (!mbox)
+		return;
+
+	mutex_lock(&con_mutex);
+
+	list_del(&mbox->node);
+
+	for (i = 0; i < mbox->num_chans; i++)
+		mbox_free_channel(&mbox->chans[i]);
+
+	if (mbox->txdone_poll)
+		del_timer_sync(&mbox->poll);
+
+	mutex_unlock(&con_mutex);
+}
+EXPORT_SYMBOL_GPL(mbox_controller_unregister);
diff --git a/drivers/mailbox/mailbox.h b/drivers/mailbox/mailbox.h
new file mode 100644
index 000000000000..456ba68513bb
--- /dev/null
+++ b/drivers/mailbox/mailbox.h
@@ -0,0 +1,14 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef __MAILBOX_H
+#define __MAILBOX_H
+
+#define TXDONE_BY_IRQ	BIT(0) /* controller has remote RTR irq */
+#define TXDONE_BY_POLL	BIT(1) /* controller can read status of last TX */
+#define TXDONE_BY_ACK	BIT(2) /* S/W ACK recevied by Client ticks the TX */
+
+#endif /* __MAILBOX_H */
diff --git a/drivers/mailbox/omap-mailbox.c b/drivers/mailbox/omap-mailbox.c
index a27e00e63a8a..0f332c178b07 100644
--- a/drivers/mailbox/omap-mailbox.c
+++ b/drivers/mailbox/omap-mailbox.c
@@ -29,12 +29,14 @@
 #include <linux/slab.h>
 #include <linux/kfifo.h>
 #include <linux/err.h>
-#include <linux/notifier.h>
 #include <linux/module.h>
+#include <linux/of_device.h>
 #include <linux/platform_device.h>
 #include <linux/pm_runtime.h>
 #include <linux/platform_data/mailbox-omap.h>
 #include <linux/omap-mailbox.h>
+#include <linux/mailbox_controller.h>
+#include <linux/mailbox_client.h>
 
 #define MAILBOX_REVISION		0x000
 #define MAILBOX_MESSAGE(m)		(0x040 + 4 * (m))
@@ -79,7 +81,6 @@ struct omap_mbox_queue {
 	spinlock_t		lock;
 	struct kfifo		fifo;
 	struct work_struct	work;
-	struct tasklet_struct	tasklet;
 	struct omap_mbox	*mbox;
 	bool full;
 };
@@ -91,21 +92,33 @@ struct omap_mbox_device {
 	u32 num_users;
 	u32 num_fifos;
 	struct omap_mbox **mboxes;
+	struct mbox_controller controller;
 	struct list_head elem;
 };
 
+struct omap_mbox_fifo_info {
+	int tx_id;
+	int tx_usr;
+	int tx_irq;
+
+	int rx_id;
+	int rx_usr;
+	int rx_irq;
+
+	const char *name;
+};
+
 struct omap_mbox {
 	const char		*name;
 	int			irq;
-	struct omap_mbox_queue	*txq, *rxq;
+	struct omap_mbox_queue	*rxq;
 	struct device		*dev;
 	struct omap_mbox_device *parent;
 	struct omap_mbox_fifo	tx_fifo;
 	struct omap_mbox_fifo	rx_fifo;
 	u32			ctx[OMAP4_MBOX_NR_REGS];
 	u32			intr_type;
-	int			use_count;
-	struct blocking_notifier_head	notifier;
+	struct mbox_chan	*chan;
 };
 
 /* global variables for the mailbox devices */
@@ -116,6 +129,14 @@ static unsigned int mbox_kfifo_size = CONFIG_OMAP_MBOX_KFIFO_SIZE;
 module_param(mbox_kfifo_size, uint, S_IRUGO);
 MODULE_PARM_DESC(mbox_kfifo_size, "Size of omap's mailbox kfifo (bytes)");
 
+static struct omap_mbox *mbox_chan_to_omap_mbox(struct mbox_chan *chan)
+{
+	if (!chan || !chan->con_priv)
+		return NULL;
+
+	return (struct omap_mbox *)chan->con_priv;
+}
+
 static inline
 unsigned int mbox_read_reg(struct omap_mbox_device *mdev, size_t ofs)
 {
@@ -181,41 +202,14 @@ static int is_mbox_irq(struct omap_mbox *mbox, omap_mbox_irq_t irq)
 	return (int)(enable & status & bit);
 }
 
-/*
- * message sender
- */
-int omap_mbox_msg_send(struct omap_mbox *mbox, mbox_msg_t msg)
-{
-	struct omap_mbox_queue *mq = mbox->txq;
-	int ret = 0, len;
-
-	spin_lock_bh(&mq->lock);
-
-	if (kfifo_avail(&mq->fifo) < sizeof(msg)) {
-		ret = -ENOMEM;
-		goto out;
-	}
-
-	if (kfifo_is_empty(&mq->fifo) && !mbox_fifo_full(mbox)) {
-		mbox_fifo_write(mbox, msg);
-		goto out;
-	}
-
-	len = kfifo_in(&mq->fifo, (unsigned char *)&msg, sizeof(msg));
-	WARN_ON(len != sizeof(msg));
-
-	tasklet_schedule(&mbox->txq->tasklet);
-
-out:
-	spin_unlock_bh(&mq->lock);
-	return ret;
-}
-EXPORT_SYMBOL(omap_mbox_msg_send);
-
-void omap_mbox_save_ctx(struct omap_mbox *mbox)
+void omap_mbox_save_ctx(struct mbox_chan *chan)
 {
 	int i;
 	int nr_regs;
+	struct omap_mbox *mbox = mbox_chan_to_omap_mbox(chan);
+
+	if (WARN_ON(!mbox))
+		return;
 
 	if (mbox->intr_type)
 		nr_regs = OMAP4_MBOX_NR_REGS;
@@ -230,10 +224,14 @@ void omap_mbox_save_ctx(struct omap_mbox *mbox)
 }
 EXPORT_SYMBOL(omap_mbox_save_ctx);
 
-void omap_mbox_restore_ctx(struct omap_mbox *mbox)
+void omap_mbox_restore_ctx(struct mbox_chan *chan)
 {
 	int i;
 	int nr_regs;
+	struct omap_mbox *mbox = mbox_chan_to_omap_mbox(chan);
+
+	if (WARN_ON(!mbox))
+		return;
 
 	if (mbox->intr_type)
 		nr_regs = OMAP4_MBOX_NR_REGS;
@@ -241,14 +239,13 @@ void omap_mbox_restore_ctx(struct omap_mbox *mbox)
 		nr_regs = MBOX_NR_REGS;
 	for (i = 0; i < nr_regs; i++) {
 		mbox_write_reg(mbox->parent, mbox->ctx[i], i * sizeof(u32));
-
 		dev_dbg(mbox->dev, "%s: [%02x] %08x\n", __func__,
 			i, mbox->ctx[i]);
 	}
 }
 EXPORT_SYMBOL(omap_mbox_restore_ctx);
 
-void omap_mbox_enable_irq(struct omap_mbox *mbox, omap_mbox_irq_t irq)
+static void _omap_mbox_enable_irq(struct omap_mbox *mbox, omap_mbox_irq_t irq)
 {
 	u32 l;
 	struct omap_mbox_fifo *fifo = (irq == IRQ_TX) ?
@@ -260,9 +257,8 @@ void omap_mbox_enable_irq(struct omap_mbox *mbox, omap_mbox_irq_t irq)
 	l |= bit;
 	mbox_write_reg(mbox->parent, l, irqenable);
 }
-EXPORT_SYMBOL(omap_mbox_enable_irq);
 
-void omap_mbox_disable_irq(struct omap_mbox *mbox, omap_mbox_irq_t irq)
+static void _omap_mbox_disable_irq(struct omap_mbox *mbox, omap_mbox_irq_t irq)
 {
 	struct omap_mbox_fifo *fifo = (irq == IRQ_TX) ?
 				&mbox->tx_fifo : &mbox->rx_fifo;
@@ -278,28 +274,28 @@ void omap_mbox_disable_irq(struct omap_mbox *mbox, omap_mbox_irq_t irq)
 
 	mbox_write_reg(mbox->parent, bit, irqdisable);
 }
-EXPORT_SYMBOL(omap_mbox_disable_irq);
 
-static void mbox_tx_tasklet(unsigned long tx_data)
+void omap_mbox_enable_irq(struct mbox_chan *chan, omap_mbox_irq_t irq)
 {
-	struct omap_mbox *mbox = (struct omap_mbox *)tx_data;
-	struct omap_mbox_queue *mq = mbox->txq;
-	mbox_msg_t msg;
-	int ret;
+	struct omap_mbox *mbox = mbox_chan_to_omap_mbox(chan);
 
-	while (kfifo_len(&mq->fifo)) {
-		if (mbox_fifo_full(mbox)) {
-			omap_mbox_enable_irq(mbox, IRQ_TX);
-			break;
-		}
+	if (WARN_ON(!mbox))
+		return;
 
-		ret = kfifo_out(&mq->fifo, (unsigned char *)&msg,
-								sizeof(msg));
-		WARN_ON(ret != sizeof(msg));
+	_omap_mbox_enable_irq(mbox, irq);
+}
+EXPORT_SYMBOL(omap_mbox_enable_irq);
 
-		mbox_fifo_write(mbox, msg);
-	}
+void omap_mbox_disable_irq(struct mbox_chan *chan, omap_mbox_irq_t irq)
+{
+	struct omap_mbox *mbox = mbox_chan_to_omap_mbox(chan);
+
+	if (WARN_ON(!mbox))
+		return;
+
+	_omap_mbox_disable_irq(mbox, irq);
 }
+EXPORT_SYMBOL(omap_mbox_disable_irq);
 
 /*
  * Message receiver(workqueue)
@@ -315,12 +311,11 @@ static void mbox_rx_work(struct work_struct *work)
 		len = kfifo_out(&mq->fifo, (unsigned char *)&msg, sizeof(msg));
 		WARN_ON(len != sizeof(msg));
 
-		blocking_notifier_call_chain(&mq->mbox->notifier, len,
-								(void *)msg);
+		mbox_chan_received_data(mq->mbox->chan, (void *)msg);
 		spin_lock_irq(&mq->lock);
 		if (mq->full) {
 			mq->full = false;
-			omap_mbox_enable_irq(mq->mbox, IRQ_RX);
+			_omap_mbox_enable_irq(mq->mbox, IRQ_RX);
 		}
 		spin_unlock_irq(&mq->lock);
 	}
@@ -331,9 +326,9 @@ static void mbox_rx_work(struct work_struct *work)
  */
 static void __mbox_tx_interrupt(struct omap_mbox *mbox)
 {
-	omap_mbox_disable_irq(mbox, IRQ_TX);
+	_omap_mbox_disable_irq(mbox, IRQ_TX);
 	ack_mbox_irq(mbox, IRQ_TX);
-	tasklet_schedule(&mbox->txq->tasklet);
+	mbox_chan_txdone(mbox->chan, 0);
 }
 
 static void __mbox_rx_interrupt(struct omap_mbox *mbox)
@@ -344,7 +339,7 @@ static void __mbox_rx_interrupt(struct omap_mbox *mbox)
 
 	while (!mbox_fifo_empty(mbox)) {
 		if (unlikely(kfifo_avail(&mq->fifo) < sizeof(msg))) {
-			omap_mbox_disable_irq(mbox, IRQ_RX);
+			_omap_mbox_disable_irq(mbox, IRQ_RX);
 			mq->full = true;
 			goto nomem;
 		}
@@ -375,11 +370,13 @@ static irqreturn_t mbox_interrupt(int irq, void *p)
 }
 
 static struct omap_mbox_queue *mbox_queue_alloc(struct omap_mbox *mbox,
-					void (*work) (struct work_struct *),
-					void (*tasklet)(unsigned long))
+					void (*work)(struct work_struct *))
 {
 	struct omap_mbox_queue *mq;
 
+	if (!work)
+		return NULL;
+
 	mq = kzalloc(sizeof(struct omap_mbox_queue), GFP_KERNEL);
 	if (!mq)
 		return NULL;
@@ -389,12 +386,9 @@ static struct omap_mbox_queue *mbox_queue_alloc(struct omap_mbox *mbox,
 	if (kfifo_alloc(&mq->fifo, mbox_kfifo_size, GFP_KERNEL))
 		goto error;
 
-	if (work)
-		INIT_WORK(&mq->work, work);
-
-	if (tasklet)
-		tasklet_init(&mq->tasklet, tasklet, (unsigned long)mbox);
+	INIT_WORK(&mq->work, work);
 	return mq;
+
 error:
 	kfree(mq);
 	return NULL;
@@ -410,71 +404,35 @@ static int omap_mbox_startup(struct omap_mbox *mbox)
 {
 	int ret = 0;
 	struct omap_mbox_queue *mq;
-	struct omap_mbox_device *mdev = mbox->parent;
 
-	mutex_lock(&mdev->cfg_lock);
-	ret = pm_runtime_get_sync(mdev->dev);
-	if (unlikely(ret < 0))
-		goto fail_startup;
-
-	if (!mbox->use_count++) {
-		mq = mbox_queue_alloc(mbox, NULL, mbox_tx_tasklet);
-		if (!mq) {
-			ret = -ENOMEM;
-			goto fail_alloc_txq;
-		}
-		mbox->txq = mq;
+	mq = mbox_queue_alloc(mbox, mbox_rx_work);
+	if (!mq)
+		return -ENOMEM;
+	mbox->rxq = mq;
+	mq->mbox = mbox;
+
+	ret = request_irq(mbox->irq, mbox_interrupt, IRQF_SHARED,
+			  mbox->name, mbox);
+	if (unlikely(ret)) {
+		pr_err("failed to register mailbox interrupt:%d\n", ret);
+		goto fail_request_irq;
+	}
 
-		mq = mbox_queue_alloc(mbox, mbox_rx_work, NULL);
-		if (!mq) {
-			ret = -ENOMEM;
-			goto fail_alloc_rxq;
-		}
-		mbox->rxq = mq;
-		mq->mbox = mbox;
-		ret = request_irq(mbox->irq, mbox_interrupt, IRQF_SHARED,
-							mbox->name, mbox);
-		if (unlikely(ret)) {
-			pr_err("failed to register mailbox interrupt:%d\n",
-									ret);
-			goto fail_request_irq;
-		}
+	_omap_mbox_enable_irq(mbox, IRQ_RX);
 
-		omap_mbox_enable_irq(mbox, IRQ_RX);
-	}
-	mutex_unlock(&mdev->cfg_lock);
 	return 0;
 
 fail_request_irq:
 	mbox_queue_free(mbox->rxq);
-fail_alloc_rxq:
-	mbox_queue_free(mbox->txq);
-fail_alloc_txq:
-	pm_runtime_put_sync(mdev->dev);
-	mbox->use_count--;
-fail_startup:
-	mutex_unlock(&mdev->cfg_lock);
 	return ret;
 }
 
 static void omap_mbox_fini(struct omap_mbox *mbox)
 {
-	struct omap_mbox_device *mdev = mbox->parent;
-
-	mutex_lock(&mdev->cfg_lock);
-
-	if (!--mbox->use_count) {
-		omap_mbox_disable_irq(mbox, IRQ_RX);
-		free_irq(mbox->irq, mbox);
-		tasklet_kill(&mbox->txq->tasklet);
-		flush_work(&mbox->rxq->work);
-		mbox_queue_free(mbox->txq);
-		mbox_queue_free(mbox->rxq);
-	}
-
-	pm_runtime_put_sync(mdev->dev);
-
-	mutex_unlock(&mdev->cfg_lock);
+	_omap_mbox_disable_irq(mbox, IRQ_RX);
+	free_irq(mbox->irq, mbox);
+	flush_work(&mbox->rxq->work);
+	mbox_queue_free(mbox->rxq);
 }
 
 static struct omap_mbox *omap_mbox_device_find(struct omap_mbox_device *mdev,
@@ -496,42 +454,55 @@ static struct omap_mbox *omap_mbox_device_find(struct omap_mbox_device *mdev,
 	return mbox;
 }
 
-struct omap_mbox *omap_mbox_get(const char *name, struct notifier_block *nb)
+struct mbox_chan *omap_mbox_request_channel(struct mbox_client *cl,
+					    const char *chan_name)
 {
+	struct device *dev = cl->dev;
 	struct omap_mbox *mbox = NULL;
 	struct omap_mbox_device *mdev;
+	struct mbox_chan *chan;
+	unsigned long flags;
 	int ret;
 
+	if (!dev)
+		return ERR_PTR(-ENODEV);
+
+	if (dev->of_node) {
+		pr_err("%s: please use mbox_request_channel(), this API is supported only for OMAP non-DT usage\n",
+		       __func__);
+		return ERR_PTR(-ENODEV);
+	}
+
 	mutex_lock(&omap_mbox_devices_lock);
 	list_for_each_entry(mdev, &omap_mbox_devices, elem) {
-		mbox = omap_mbox_device_find(mdev, name);
+		mbox = omap_mbox_device_find(mdev, chan_name);
 		if (mbox)
 			break;
 	}
 	mutex_unlock(&omap_mbox_devices_lock);
 
-	if (!mbox)
+	if (!mbox || !mbox->chan)
 		return ERR_PTR(-ENOENT);
 
-	if (nb)
-		blocking_notifier_chain_register(&mbox->notifier, nb);
+	chan = mbox->chan;
+	spin_lock_irqsave(&chan->lock, flags);
+	chan->msg_free = 0;
+	chan->msg_count = 0;
+	chan->active_req = NULL;
+	chan->cl = cl;
+	init_completion(&chan->tx_complete);
+	spin_unlock_irqrestore(&chan->lock, flags);
 
-	ret = omap_mbox_startup(mbox);
+	ret = chan->mbox->ops->startup(chan);
 	if (ret) {
-		blocking_notifier_chain_unregister(&mbox->notifier, nb);
-		return ERR_PTR(-ENODEV);
+		pr_err("Unable to startup the chan (%d)\n", ret);
+		mbox_free_channel(chan);
+		chan = ERR_PTR(ret);
 	}
 
-	return mbox;
-}
-EXPORT_SYMBOL(omap_mbox_get);
-
-void omap_mbox_put(struct omap_mbox *mbox, struct notifier_block *nb)
-{
-	blocking_notifier_chain_unregister(&mbox->notifier, nb);
-	omap_mbox_fini(mbox);
+	return chan;
 }
-EXPORT_SYMBOL(omap_mbox_put);
+EXPORT_SYMBOL(omap_mbox_request_channel);
 
 static struct class omap_mbox_class = { .name = "mbox", };
 
@@ -547,25 +518,25 @@ static int omap_mbox_register(struct omap_mbox_device *mdev)
 	mboxes = mdev->mboxes;
 	for (i = 0; mboxes[i]; i++) {
 		struct omap_mbox *mbox = mboxes[i];
-		mbox->dev = device_create(&omap_mbox_class,
-				mdev->dev, 0, mbox, "%s", mbox->name);
+		mbox->dev = device_create(&omap_mbox_class, mdev->dev,
+					0, mbox, "%s", mbox->name);
 		if (IS_ERR(mbox->dev)) {
 			ret = PTR_ERR(mbox->dev);
 			goto err_out;
 		}
-
-		BLOCKING_INIT_NOTIFIER_HEAD(&mbox->notifier);
 	}
 
 	mutex_lock(&omap_mbox_devices_lock);
 	list_add(&mdev->elem, &omap_mbox_devices);
 	mutex_unlock(&omap_mbox_devices_lock);
 
-	return 0;
+	ret = mbox_controller_register(&mdev->controller);
 
 err_out:
-	while (i--)
-		device_unregister(mboxes[i]->dev);
+	if (ret) {
+		while (i--)
+			device_unregister(mboxes[i]->dev);
+	}
 	return ret;
 }
 
@@ -581,30 +552,201 @@ static int omap_mbox_unregister(struct omap_mbox_device *mdev)
 	list_del(&mdev->elem);
 	mutex_unlock(&omap_mbox_devices_lock);
 
+	mbox_controller_unregister(&mdev->controller);
+
 	mboxes = mdev->mboxes;
 	for (i = 0; mboxes[i]; i++)
 		device_unregister(mboxes[i]->dev);
 	return 0;
 }
 
+static int omap_mbox_chan_startup(struct mbox_chan *chan)
+{
+	struct omap_mbox *mbox = mbox_chan_to_omap_mbox(chan);
+	struct omap_mbox_device *mdev = mbox->parent;
+	int ret = 0;
+
+	mutex_lock(&mdev->cfg_lock);
+	pm_runtime_get_sync(mdev->dev);
+	ret = omap_mbox_startup(mbox);
+	if (ret)
+		pm_runtime_put_sync(mdev->dev);
+	mutex_unlock(&mdev->cfg_lock);
+	return ret;
+}
+
+static void omap_mbox_chan_shutdown(struct mbox_chan *chan)
+{
+	struct omap_mbox *mbox = mbox_chan_to_omap_mbox(chan);
+	struct omap_mbox_device *mdev = mbox->parent;
+
+	mutex_lock(&mdev->cfg_lock);
+	omap_mbox_fini(mbox);
+	pm_runtime_put_sync(mdev->dev);
+	mutex_unlock(&mdev->cfg_lock);
+}
+
+static int omap_mbox_chan_send_data(struct mbox_chan *chan, void *data)
+{
+	struct omap_mbox *mbox = mbox_chan_to_omap_mbox(chan);
+	int ret = -EBUSY;
+
+	if (!mbox)
+		return -EINVAL;
+
+	if (!mbox_fifo_full(mbox)) {
+		mbox_fifo_write(mbox, (mbox_msg_t)data);
+		ret = 0;
+	}
+
+	/* always enable the interrupt */
+	_omap_mbox_enable_irq(mbox, IRQ_TX);
+	return ret;
+}
+
+static struct mbox_chan_ops omap_mbox_chan_ops = {
+	.startup        = omap_mbox_chan_startup,
+	.send_data      = omap_mbox_chan_send_data,
+	.shutdown       = omap_mbox_chan_shutdown,
+};
+
+static const struct of_device_id omap_mailbox_of_match[] = {
+	{
+		.compatible	= "ti,omap2-mailbox",
+		.data		= (void *)MBOX_INTR_CFG_TYPE1,
+	},
+	{
+		.compatible	= "ti,omap3-mailbox",
+		.data		= (void *)MBOX_INTR_CFG_TYPE1,
+	},
+	{
+		.compatible	= "ti,omap4-mailbox",
+		.data		= (void *)MBOX_INTR_CFG_TYPE2,
+	},
+	{
+		/* end */
+	},
+};
+MODULE_DEVICE_TABLE(of, omap_mailbox_of_match);
+
+static struct mbox_chan *omap_mbox_of_xlate(struct mbox_controller *controller,
+					    const struct of_phandle_args *sp)
+{
+	phandle phandle = sp->args[0];
+	struct device_node *node;
+	struct omap_mbox_device *mdev;
+	struct omap_mbox *mbox;
+
+	mdev = container_of(controller, struct omap_mbox_device, controller);
+	if (WARN_ON(!mdev))
+		return NULL;
+
+	node = of_find_node_by_phandle(phandle);
+	if (!node) {
+		pr_err("%s: could not find node phandle 0x%x\n",
+		       __func__, phandle);
+		return NULL;
+	}
+
+	mbox = omap_mbox_device_find(mdev, node->name);
+	of_node_put(node);
+	return mbox ? mbox->chan : NULL;
+}
+
 static int omap_mbox_probe(struct platform_device *pdev)
 {
 	struct resource *mem;
 	int ret;
+	struct mbox_chan *chnls;
 	struct omap_mbox **list, *mbox, *mboxblk;
 	struct omap_mbox_pdata *pdata = pdev->dev.platform_data;
-	struct omap_mbox_dev_info *info;
+	struct omap_mbox_dev_info *info = NULL;
+	struct omap_mbox_fifo_info *finfo, *finfoblk;
 	struct omap_mbox_device *mdev;
 	struct omap_mbox_fifo *fifo;
-	u32 intr_type;
+	struct device_node *node = pdev->dev.of_node;
+	struct device_node *child;
+	const struct of_device_id *match;
+	u32 intr_type, info_count;
+	u32 num_users, num_fifos;
+	u32 tmp[3];
 	u32 l;
 	int i;
 
-	if (!pdata || !pdata->info_cnt || !pdata->info) {
+	if (!node && (!pdata || !pdata->info_cnt || !pdata->info)) {
 		pr_err("%s: platform not supported\n", __func__);
 		return -ENODEV;
 	}
 
+	if (node) {
+		match = of_match_device(omap_mailbox_of_match, &pdev->dev);
+		if (!match)
+			return -ENODEV;
+		intr_type = (u32)match->data;
+
+		if (of_property_read_u32(node, "ti,mbox-num-users",
+					 &num_users))
+			return -ENODEV;
+
+		if (of_property_read_u32(node, "ti,mbox-num-fifos",
+					 &num_fifos))
+			return -ENODEV;
+
+		info_count = of_get_available_child_count(node);
+		if (!info_count) {
+			dev_err(&pdev->dev, "no available mbox devices found\n");
+			return -ENODEV;
+		}
+	} else { /* non-DT device creation */
+		info_count = pdata->info_cnt;
+		info = pdata->info;
+		intr_type = pdata->intr_type;
+		num_users = pdata->num_users;
+		num_fifos = pdata->num_fifos;
+	}
+
+	finfoblk = devm_kzalloc(&pdev->dev, info_count * sizeof(*finfoblk),
+				GFP_KERNEL);
+	if (!finfoblk)
+		return -ENOMEM;
+
+	finfo = finfoblk;
+	child = NULL;
+	for (i = 0; i < info_count; i++, finfo++) {
+		if (node) {
+			child = of_get_next_available_child(node, child);
+			ret = of_property_read_u32_array(child, "ti,mbox-tx",
+							 tmp, ARRAY_SIZE(tmp));
+			if (ret)
+				return ret;
+			finfo->tx_id = tmp[0];
+			finfo->tx_irq = tmp[1];
+			finfo->tx_usr = tmp[2];
+
+			ret = of_property_read_u32_array(child, "ti,mbox-rx",
+							 tmp, ARRAY_SIZE(tmp));
+			if (ret)
+				return ret;
+			finfo->rx_id = tmp[0];
+			finfo->rx_irq = tmp[1];
+			finfo->rx_usr = tmp[2];
+
+			finfo->name = child->name;
+		} else {
+			finfo->tx_id = info->tx_id;
+			finfo->rx_id = info->rx_id;
+			finfo->tx_usr = info->usr_id;
+			finfo->tx_irq = info->irq_id;
+			finfo->rx_usr = info->usr_id;
+			finfo->rx_irq = info->irq_id;
+			finfo->name = info->name;
+			info++;
+		}
+		if (finfo->tx_id >= num_fifos || finfo->rx_id >= num_fifos ||
+		    finfo->tx_usr >= num_users || finfo->rx_usr >= num_users)
+			return -EINVAL;
+	}
+
 	mdev = devm_kzalloc(&pdev->dev, sizeof(*mdev), GFP_KERNEL);
 	if (!mdev)
 		return -ENOMEM;
@@ -615,51 +757,65 @@ static int omap_mbox_probe(struct platform_device *pdev)
 		return PTR_ERR(mdev->mbox_base);
 
 	/* allocate one extra for marking end of list */
-	list = devm_kzalloc(&pdev->dev, (pdata->info_cnt + 1) * sizeof(*list),
+	list = devm_kzalloc(&pdev->dev, (info_count + 1) * sizeof(*list),
 			    GFP_KERNEL);
 	if (!list)
 		return -ENOMEM;
 
-	mboxblk = devm_kzalloc(&pdev->dev, pdata->info_cnt * sizeof(*mbox),
+	chnls = devm_kzalloc(&pdev->dev, (info_count + 1) * sizeof(*chnls),
+			     GFP_KERNEL);
+	if (!chnls)
+		return -ENOMEM;
+
+	mboxblk = devm_kzalloc(&pdev->dev, info_count * sizeof(*mbox),
 			       GFP_KERNEL);
 	if (!mboxblk)
 		return -ENOMEM;
 
-	info = pdata->info;
-	intr_type = pdata->intr_type;
 	mbox = mboxblk;
-	for (i = 0; i < pdata->info_cnt; i++, info++) {
+	finfo = finfoblk;
+	for (i = 0; i < info_count; i++, finfo++) {
 		fifo = &mbox->tx_fifo;
-		fifo->msg = MAILBOX_MESSAGE(info->tx_id);
-		fifo->fifo_stat = MAILBOX_FIFOSTATUS(info->tx_id);
-		fifo->intr_bit = MAILBOX_IRQ_NOTFULL(info->tx_id);
-		fifo->irqenable = MAILBOX_IRQENABLE(intr_type, info->usr_id);
-		fifo->irqstatus = MAILBOX_IRQSTATUS(intr_type, info->usr_id);
-		fifo->irqdisable = MAILBOX_IRQDISABLE(intr_type, info->usr_id);
+		fifo->msg = MAILBOX_MESSAGE(finfo->tx_id);
+		fifo->fifo_stat = MAILBOX_FIFOSTATUS(finfo->tx_id);
+		fifo->intr_bit = MAILBOX_IRQ_NOTFULL(finfo->tx_id);
+		fifo->irqenable = MAILBOX_IRQENABLE(intr_type, finfo->tx_usr);
+		fifo->irqstatus = MAILBOX_IRQSTATUS(intr_type, finfo->tx_usr);
+		fifo->irqdisable = MAILBOX_IRQDISABLE(intr_type, finfo->tx_usr);
 
 		fifo = &mbox->rx_fifo;
-		fifo->msg =  MAILBOX_MESSAGE(info->rx_id);
-		fifo->msg_stat =  MAILBOX_MSGSTATUS(info->rx_id);
-		fifo->intr_bit = MAILBOX_IRQ_NEWMSG(info->rx_id);
-		fifo->irqenable = MAILBOX_IRQENABLE(intr_type, info->usr_id);
-		fifo->irqstatus = MAILBOX_IRQSTATUS(intr_type, info->usr_id);
-		fifo->irqdisable = MAILBOX_IRQDISABLE(intr_type, info->usr_id);
+		fifo->msg = MAILBOX_MESSAGE(finfo->rx_id);
+		fifo->msg_stat =  MAILBOX_MSGSTATUS(finfo->rx_id);
+		fifo->intr_bit = MAILBOX_IRQ_NEWMSG(finfo->rx_id);
+		fifo->irqenable = MAILBOX_IRQENABLE(intr_type, finfo->rx_usr);
+		fifo->irqstatus = MAILBOX_IRQSTATUS(intr_type, finfo->rx_usr);
+		fifo->irqdisable = MAILBOX_IRQDISABLE(intr_type, finfo->rx_usr);
 
 		mbox->intr_type = intr_type;
 
 		mbox->parent = mdev;
-		mbox->name = info->name;
-		mbox->irq = platform_get_irq(pdev, info->irq_id);
+		mbox->name = finfo->name;
+		mbox->irq = platform_get_irq(pdev, finfo->tx_irq);
 		if (mbox->irq < 0)
 			return mbox->irq;
+		mbox->chan = &chnls[i];
+		chnls[i].con_priv = mbox;
 		list[i] = mbox++;
 	}
 
 	mutex_init(&mdev->cfg_lock);
 	mdev->dev = &pdev->dev;
-	mdev->num_users = pdata->num_users;
-	mdev->num_fifos = pdata->num_fifos;
+	mdev->num_users = num_users;
+	mdev->num_fifos = num_fifos;
 	mdev->mboxes = list;
+
+	/* OMAP does not have a Tx-Done IRQ, but rather a Tx-Ready IRQ */
+	mdev->controller.txdone_irq = true;
+	mdev->controller.dev = mdev->dev;
+	mdev->controller.ops = &omap_mbox_chan_ops;
+	mdev->controller.chans = chnls;
+	mdev->controller.num_chans = info_count;
+	mdev->controller.of_xlate = omap_mbox_of_xlate;
 	ret = omap_mbox_register(mdev);
 	if (ret)
 		return ret;
@@ -684,6 +840,7 @@ static int omap_mbox_probe(struct platform_device *pdev)
 	if (ret < 0)
 		goto unregister;
 
+	devm_kfree(&pdev->dev, finfoblk);
 	return 0;
 
 unregister:
@@ -707,7 +864,7 @@ static struct platform_driver omap_mbox_driver = {
 	.remove	= omap_mbox_remove,
 	.driver	= {
 		.name = "omap-mailbox",
-		.owner = THIS_MODULE,
+		.of_match_table = of_match_ptr(omap_mailbox_of_match),
 	},
 };
 
diff --git a/drivers/mailbox/pcc.c b/drivers/mailbox/pcc.c
new file mode 100644
index 000000000000..6dbf6fcbdfaf
--- /dev/null
+++ b/drivers/mailbox/pcc.c
@@ -0,0 +1,403 @@
+/*
+ *	Copyright (C) 2014 Linaro Ltd.
+ *	Author:	Ashwin Chaugule <ashwin.chaugule@linaro.org>
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  PCC (Platform Communication Channel) is defined in the ACPI 5.0+
+ *  specification. It is a mailbox like mechanism to allow clients
+ *  such as CPPC (Collaborative Processor Performance Control), RAS
+ *  (Reliability, Availability and Serviceability) and MPST (Memory
+ *  Node Power State Table) to talk to the platform (e.g. BMC) through
+ *  shared memory regions as defined in the PCC table entries. The PCC
+ *  specification supports a Doorbell mechanism for the PCC clients
+ *  to notify the platform about new data. This Doorbell information
+ *  is also specified in each PCC table entry. See pcc_send_data()
+ *  and pcc_tx_done() for basic mode of operation.
+ *
+ *  For more details about PCC, please see the ACPI specification from
+ *  http://www.uefi.org/ACPIv5.1 Section 14.
+ *
+ *  This file implements PCC as a Mailbox controller and allows for PCC
+ *  clients to be implemented as its Mailbox Client Channels.
+ */
+
+#include <linux/acpi.h>
+#include <linux/delay.h>
+#include <linux/io.h>
+#include <linux/init.h>
+#include <linux/list.h>
+#include <linux/platform_device.h>
+#include <linux/mailbox_controller.h>
+#include <linux/mailbox_client.h>
+
+#include "mailbox.h"
+
+#define MAX_PCC_SUBSPACES	256
+#define PCCS_SS_SIG_MAGIC	0x50434300
+#define PCC_CMD_COMPLETE	0x1
+
+static struct mbox_chan *pcc_mbox_channels;
+
+static struct mbox_controller pcc_mbox_ctrl = {};
+/**
+ * get_pcc_channel - Given a PCC subspace idx, get
+ *	the respective mbox_channel.
+ * @id: PCC subspace index.
+ *
+ * Return: ERR_PTR(errno) if error, else pointer
+ *	to mbox channel.
+ */
+static struct mbox_chan *get_pcc_channel(int id)
+{
+	struct mbox_chan *pcc_chan;
+
+	if (id < 0 || id > pcc_mbox_ctrl.num_chans)
+		return ERR_PTR(-ENOENT);
+
+	pcc_chan = (struct mbox_chan *)
+		(unsigned long) pcc_mbox_channels +
+		(id * sizeof(*pcc_chan));
+
+	return pcc_chan;
+}
+
+/**
+ * get_subspace_id - Given a Mailbox channel, find out the
+ *		PCC subspace id.
+ * @chan: Pointer to Mailbox Channel from which we want
+ *		the index.
+ * Return: Errno if not found, else positive index number.
+ */
+static int get_subspace_id(struct mbox_chan *chan)
+{
+	unsigned int id = chan - pcc_mbox_channels;
+
+	if (id < 0 || id > pcc_mbox_ctrl.num_chans)
+		return -ENOENT;
+
+	return id;
+}
+
+/**
+ * pcc_mbox_request_channel - PCC clients call this function to
+ *		request a pointer to their PCC subspace, from which they
+ *		can get the details of communicating with the remote.
+ * @cl: Pointer to Mailbox client, so we know where to bind the
+ *		Channel.
+ * @subspace_id: The PCC Subspace index as parsed in the PCC client
+ *		ACPI package. This is used to lookup the array of PCC
+ *		subspaces as parsed by the PCC Mailbox controller.
+ *
+ * Return: Pointer to the Mailbox Channel if successful or
+ *		ERR_PTR.
+ */
+struct mbox_chan *pcc_mbox_request_channel(struct mbox_client *cl,
+		int subspace_id)
+{
+	struct device *dev = pcc_mbox_ctrl.dev;
+	struct mbox_chan *chan;
+	unsigned long flags;
+
+	/*
+	 * Each PCC Subspace is a Mailbox Channel.
+	 * The PCC Clients get their PCC Subspace ID
+	 * from their own tables and pass it here.
+	 * This returns a pointer to the PCC subspace
+	 * for the Client to operate on.
+	 */
+	chan = get_pcc_channel(subspace_id);
+
+	if (!chan || chan->cl) {
+		dev_err(dev, "%s: PCC mailbox not free\n", __func__);
+		return ERR_PTR(-EBUSY);
+	}
+
+	spin_lock_irqsave(&chan->lock, flags);
+	chan->msg_free = 0;
+	chan->msg_count = 0;
+	chan->active_req = NULL;
+	chan->cl = cl;
+	init_completion(&chan->tx_complete);
+
+	if (chan->txdone_method == TXDONE_BY_POLL && cl->knows_txdone)
+		chan->txdone_method |= TXDONE_BY_ACK;
+
+	spin_unlock_irqrestore(&chan->lock, flags);
+
+	return chan;
+}
+EXPORT_SYMBOL_GPL(pcc_mbox_request_channel);
+
+/**
+ * pcc_mbox_free_channel - Clients call this to free their Channel.
+ *
+ * @chan: Pointer to the mailbox channel as returned by
+ *		pcc_mbox_request_channel()
+ */
+void pcc_mbox_free_channel(struct mbox_chan *chan)
+{
+	unsigned long flags;
+
+	if (!chan || !chan->cl)
+		return;
+
+	spin_lock_irqsave(&chan->lock, flags);
+	chan->cl = NULL;
+	chan->active_req = NULL;
+	if (chan->txdone_method == (TXDONE_BY_POLL | TXDONE_BY_ACK))
+		chan->txdone_method = TXDONE_BY_POLL;
+
+	spin_unlock_irqrestore(&chan->lock, flags);
+}
+EXPORT_SYMBOL_GPL(pcc_mbox_free_channel);
+
+/**
+ * pcc_tx_done - Callback from Mailbox controller code to
+ *		check if PCC message transmission completed.
+ * @chan: Pointer to Mailbox channel on which previous
+ *		transmission occurred.
+ *
+ * Return: TRUE if succeeded.
+ */
+static bool pcc_tx_done(struct mbox_chan *chan)
+{
+	struct acpi_pcct_hw_reduced *pcct_ss = chan->con_priv;
+	struct acpi_pcct_shared_memory *generic_comm_base =
+		(struct acpi_pcct_shared_memory *) pcct_ss->base_address;
+	u16 cmd_delay = pcct_ss->latency;
+	unsigned int retries = 0;
+
+	/* Try a few times while waiting for platform to consume */
+	while (!(readw_relaxed(&generic_comm_base->status)
+		    & PCC_CMD_COMPLETE)) {
+
+		if (retries++ < 5)
+			udelay(cmd_delay);
+		else {
+			/*
+			 * If the remote is dead, this will cause the Mbox
+			 * controller to timeout after mbox client.tx_tout
+			 * msecs.
+			 */
+			pr_err("PCC platform did not respond.\n");
+			return false;
+		}
+	}
+	return true;
+}
+
+/**
+ * pcc_send_data - Called from Mailbox Controller code to finally
+ *	transmit data over channel.
+ * @chan: Pointer to Mailbox channel over which to send data.
+ * @data: Actual data to be written over channel.
+ *
+ * Return: Err if something failed else 0 for success.
+ */
+static int pcc_send_data(struct mbox_chan *chan, void *data)
+{
+	struct acpi_pcct_hw_reduced *pcct_ss = chan->con_priv;
+	struct acpi_pcct_shared_memory *generic_comm_base =
+		(struct acpi_pcct_shared_memory *) pcct_ss->base_address;
+	struct acpi_generic_address doorbell;
+	u64 doorbell_preserve;
+	u64 doorbell_val;
+	u64 doorbell_write;
+	u16 cmd = *(u16 *) data;
+	u16 ss_idx = -1;
+
+	ss_idx = get_subspace_id(chan);
+
+	if (ss_idx < 0) {
+		pr_err("Invalid Subspace ID from PCC client\n");
+		return -EINVAL;
+	}
+
+	doorbell = pcct_ss->doorbell_register;
+	doorbell_preserve = pcct_ss->preserve_mask;
+	doorbell_write = pcct_ss->write_mask;
+
+	/* Write to the shared comm region. */
+	writew(cmd, &generic_comm_base->command);
+
+	/* Write Subspace MAGIC value so platform can identify destination. */
+	writel((PCCS_SS_SIG_MAGIC | ss_idx), &generic_comm_base->signature);
+
+	/* Flip CMD COMPLETE bit */
+	writew(0, &generic_comm_base->status);
+
+	/* Sync notification from OSPM to Platform. */
+	acpi_read(&doorbell_val, &doorbell);
+	acpi_write((doorbell_val & doorbell_preserve) | doorbell_write,
+			&doorbell);
+
+	return 0;
+}
+
+static struct mbox_chan_ops pcc_chan_ops = {
+	.send_data = pcc_send_data,
+	.last_tx_done = pcc_tx_done,
+};
+
+/**
+ * parse_pcc_subspace - Parse the PCC table and verify PCC subspace
+ *		entries. There should be one entry per PCC client.
+ * @header: Pointer to the ACPI subtable header under the PCCT.
+ * @end: End of subtable entry.
+ *
+ * Return: 0 for Success, else errno.
+ *
+ * This gets called for each entry in the PCC table.
+ */
+static int parse_pcc_subspace(struct acpi_subtable_header *header,
+		const unsigned long end)
+{
+	struct acpi_pcct_hw_reduced *pcct_ss;
+
+	if (pcc_mbox_ctrl.num_chans <= MAX_PCC_SUBSPACES) {
+		pcct_ss = (struct acpi_pcct_hw_reduced *) header;
+
+		if (pcct_ss->header.type !=
+				ACPI_PCCT_TYPE_HW_REDUCED_SUBSPACE) {
+			pr_err("Incorrect PCC Subspace type detected\n");
+			return -EINVAL;
+		}
+	}
+
+	return 0;
+}
+
+/**
+ * acpi_pcc_probe - Parse the ACPI tree for the PCCT.
+ *
+ * Return: 0 for Success, else errno.
+ */
+static int __init acpi_pcc_probe(void)
+{
+	acpi_size pcct_tbl_header_size;
+	struct acpi_table_header *pcct_tbl;
+	struct acpi_subtable_header *pcct_entry;
+	int count, i;
+	acpi_status status = AE_OK;
+
+	/* Search for PCCT */
+	status = acpi_get_table_with_size(ACPI_SIG_PCCT, 0,
+			&pcct_tbl,
+			&pcct_tbl_header_size);
+
+	if (ACPI_FAILURE(status) || !pcct_tbl) {
+		pr_warn("PCCT header not found.\n");
+		return -ENODEV;
+	}
+
+	count = acpi_table_parse_entries(ACPI_SIG_PCCT,
+			sizeof(struct acpi_table_pcct),
+			ACPI_PCCT_TYPE_HW_REDUCED_SUBSPACE,
+			parse_pcc_subspace, MAX_PCC_SUBSPACES);
+
+	if (count <= 0) {
+		pr_err("Error parsing PCC subspaces from PCCT\n");
+		return -EINVAL;
+	}
+
+	pcc_mbox_channels = kzalloc(sizeof(struct mbox_chan) *
+			count, GFP_KERNEL);
+
+	if (!pcc_mbox_channels) {
+		pr_err("Could not allocate space for PCC mbox channels\n");
+		return -ENOMEM;
+	}
+
+	/* Point to the first PCC subspace entry */
+	pcct_entry = (struct acpi_subtable_header *) (
+		(unsigned long) pcct_tbl + sizeof(struct acpi_table_pcct));
+
+	for (i = 0; i < count; i++) {
+		pcc_mbox_channels[i].con_priv = pcct_entry;
+		pcct_entry = (struct acpi_subtable_header *)
+			((unsigned long) pcct_entry + pcct_entry->length);
+	}
+
+	pcc_mbox_ctrl.num_chans = count;
+
+	pr_info("Detected %d PCC Subspaces\n", pcc_mbox_ctrl.num_chans);
+
+	return 0;
+}
+
+/**
+ * pcc_mbox_probe - Called when we find a match for the
+ *	PCCT platform device. This is purely used to represent
+ *	the PCCT as a virtual device for registering with the
+ *	generic Mailbox framework.
+ *
+ * @pdev: Pointer to platform device returned when a match
+ *	is found.
+ *
+ *	Return: 0 for Success, else errno.
+ */
+static int pcc_mbox_probe(struct platform_device *pdev)
+{
+	int ret = 0;
+
+	pcc_mbox_ctrl.chans = pcc_mbox_channels;
+	pcc_mbox_ctrl.ops = &pcc_chan_ops;
+	pcc_mbox_ctrl.txdone_poll = true;
+	pcc_mbox_ctrl.txpoll_period = 10;
+	pcc_mbox_ctrl.dev = &pdev->dev;
+
+	pr_info("Registering PCC driver as Mailbox controller\n");
+	ret = mbox_controller_register(&pcc_mbox_ctrl);
+
+	if (ret) {
+		pr_err("Err registering PCC as Mailbox controller: %d\n", ret);
+		ret = -ENODEV;
+	}
+
+	return ret;
+}
+
+struct platform_driver pcc_mbox_driver = {
+	.probe = pcc_mbox_probe,
+	.driver = {
+		.name = "PCCT",
+		.owner = THIS_MODULE,
+	},
+};
+
+static int __init pcc_init(void)
+{
+	int ret;
+	struct platform_device *pcc_pdev;
+
+	if (acpi_disabled)
+		return -ENODEV;
+
+	/* Check if PCC support is available. */
+	ret = acpi_pcc_probe();
+
+	if (ret) {
+		pr_err("ACPI PCC probe failed.\n");
+		return -ENODEV;
+	}
+
+	pcc_pdev = platform_create_bundle(&pcc_mbox_driver,
+			pcc_mbox_probe, NULL, 0, NULL, 0);
+
+	if (!pcc_pdev) {
+		pr_err("Err creating PCC platform bundle\n");
+		return -ENODEV;
+	}
+
+	return 0;
+}
+device_initcall(pcc_init);
diff --git a/drivers/mailbox/pl320-ipc.c b/drivers/mailbox/pl320-ipc.c
index d873cbae2fbb..f3755e0aa935 100644
--- a/drivers/mailbox/pl320-ipc.c
+++ b/drivers/mailbox/pl320-ipc.c
@@ -26,7 +26,7 @@
 #include <linux/device.h>
 #include <linux/amba/bus.h>
 
-#include <linux/mailbox.h>
+#include <linux/pl320-ipc.h>
 
 #define IPCMxSOURCE(m)		((m) * 0x40)
 #define IPCMxDSET(m)		(((m) * 0x40) + 0x004)
diff --git a/drivers/mcb/mcb-internal.h b/drivers/mcb/mcb-internal.h
index f956ef26c0ce..fb7493dcfb79 100644
--- a/drivers/mcb/mcb-internal.h
+++ b/drivers/mcb/mcb-internal.h
@@ -7,6 +7,7 @@
 #define PCI_DEVICE_ID_MEN_CHAMELEON	0x4d45
 #define CHAMELEON_FILENAME_LEN		12
 #define CHAMELEONV2_MAGIC		0xabce
+#define CHAM_HEADER_SIZE		0x200
 
 enum chameleon_descriptor_type {
 	CHAMELEON_DTYPE_GENERAL = 0x0,
diff --git a/drivers/mcb/mcb-pci.c b/drivers/mcb/mcb-pci.c
index b59181965643..5e1bd5db02c8 100644
--- a/drivers/mcb/mcb-pci.c
+++ b/drivers/mcb/mcb-pci.c
@@ -17,6 +17,7 @@
 
 struct priv {
 	struct mcb_bus *bus;
+	phys_addr_t mapbase;
 	void __iomem *base;
 };
 
@@ -31,8 +32,8 @@ static int mcb_pci_get_irq(struct mcb_device *mdev)
 
 static int mcb_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 {
+	struct resource *res;
 	struct priv *priv;
-	phys_addr_t mapbase;
 	int ret;
 	int num_cells;
 	unsigned long flags;
@@ -47,19 +48,21 @@ static int mcb_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 		return -ENODEV;
 	}
 
-	mapbase = pci_resource_start(pdev, 0);
-	if (!mapbase) {
+	priv->mapbase = pci_resource_start(pdev, 0);
+	if (!priv->mapbase) {
 		dev_err(&pdev->dev, "No PCI resource\n");
 		goto err_start;
 	}
 
-	ret = pci_request_region(pdev, 0, KBUILD_MODNAME);
-	if (ret) {
-		dev_err(&pdev->dev, "Failed to request PCI BARs\n");
+	res = request_mem_region(priv->mapbase, CHAM_HEADER_SIZE,
+				 KBUILD_MODNAME);
+	if (IS_ERR(res)) {
+		dev_err(&pdev->dev, "Failed to request PCI memory\n");
+		ret = PTR_ERR(res);
 		goto err_start;
 	}
 
-	priv->base = pci_iomap(pdev, 0, 0);
+	priv->base = ioremap(priv->mapbase, CHAM_HEADER_SIZE);
 	if (!priv->base) {
 		dev_err(&pdev->dev, "Cannot ioremap\n");
 		ret = -ENOMEM;
@@ -84,7 +87,7 @@ static int mcb_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 
 	priv->bus->get_irq = mcb_pci_get_irq;
 
-	ret = chameleon_parse_cells(priv->bus, mapbase, priv->base);
+	ret = chameleon_parse_cells(priv->bus, priv->mapbase, priv->base);
 	if (ret < 0)
 		goto err_drvdata;
 	num_cells = ret;
@@ -93,8 +96,10 @@ static int mcb_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 
 	mcb_bus_add_devices(priv->bus);
 
+	return 0;
+
 err_drvdata:
-	pci_iounmap(pdev, priv->base);
+	iounmap(priv->base);
 err_ioremap:
 	pci_release_region(pdev, 0);
 err_start:
@@ -107,6 +112,10 @@ static void mcb_pci_remove(struct pci_dev *pdev)
 	struct priv *priv = pci_get_drvdata(pdev);
 
 	mcb_release_bus(priv->bus);
+
+	iounmap(priv->base);
+	release_region(priv->mapbase, CHAM_HEADER_SIZE);
+	pci_disable_device(pdev);
 }
 
 static const struct pci_device_id mcb_pci_tbl[] = {
diff --git a/drivers/md/bcache/request.c b/drivers/md/bcache/request.c
index 62e6e98186b5..ab43faddb447 100644
--- a/drivers/md/bcache/request.c
+++ b/drivers/md/bcache/request.c
@@ -601,13 +601,8 @@ static void request_endio(struct bio *bio, int error)
 static void bio_complete(struct search *s)
 {
 	if (s->orig_bio) {
-		int cpu, rw = bio_data_dir(s->orig_bio);
-		unsigned long duration = jiffies - s->start_time;
-
-		cpu = part_stat_lock();
-		part_round_stats(cpu, &s->d->disk->part0);
-		part_stat_add(cpu, &s->d->disk->part0, ticks[rw], duration);
-		part_stat_unlock();
+		generic_end_io_acct(bio_data_dir(s->orig_bio),
+				    &s->d->disk->part0, s->start_time);
 
 		trace_bcache_request_end(s->d, s->orig_bio);
 		bio_endio(s->orig_bio, s->iop.error);
@@ -959,12 +954,9 @@ static void cached_dev_make_request(struct request_queue *q, struct bio *bio)
 	struct search *s;
 	struct bcache_device *d = bio->bi_bdev->bd_disk->private_data;
 	struct cached_dev *dc = container_of(d, struct cached_dev, disk);
-	int cpu, rw = bio_data_dir(bio);
+	int rw = bio_data_dir(bio);
 
-	cpu = part_stat_lock();
-	part_stat_inc(cpu, &d->disk->part0, ios[rw]);
-	part_stat_add(cpu, &d->disk->part0, sectors[rw], bio_sectors(bio));
-	part_stat_unlock();
+	generic_start_io_acct(rw, bio_sectors(bio), &d->disk->part0);
 
 	bio->bi_bdev = dc->bdev;
 	bio->bi_iter.bi_sector += dc->sb.data_offset;
@@ -1074,12 +1066,9 @@ static void flash_dev_make_request(struct request_queue *q, struct bio *bio)
 	struct search *s;
 	struct closure *cl;
 	struct bcache_device *d = bio->bi_bdev->bd_disk->private_data;
-	int cpu, rw = bio_data_dir(bio);
+	int rw = bio_data_dir(bio);
 
-	cpu = part_stat_lock();
-	part_stat_inc(cpu, &d->disk->part0, ios[rw]);
-	part_stat_add(cpu, &d->disk->part0, sectors[rw], bio_sectors(bio));
-	part_stat_unlock();
+	generic_start_io_acct(rw, bio_sectors(bio), &d->disk->part0);
 
 	s = search_alloc(bio, d);
 	cl = &s->cl;
diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index d4713d098a39..4dd2bb7167f0 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -842,6 +842,7 @@ static int bcache_device_init(struct bcache_device *d, unsigned block_size,
 	q->limits.logical_block_size	= block_size;
 	q->limits.physical_block_size	= block_size;
 	set_bit(QUEUE_FLAG_NONROT,	&d->disk->queue->queue_flags);
+	clear_bit(QUEUE_FLAG_ADD_RANDOM, &d->disk->queue->queue_flags);
 	set_bit(QUEUE_FLAG_DISCARD,	&d->disk->queue->queue_flags);
 
 	blk_queue_flush(q, REQ_FLUSH|REQ_FUA);
diff --git a/drivers/md/bitmap.c b/drivers/md/bitmap.c
index 67f8b31e2054..da3604e73e8a 100644
--- a/drivers/md/bitmap.c
+++ b/drivers/md/bitmap.c
@@ -879,7 +879,6 @@ void bitmap_unplug(struct bitmap *bitmap)
 {
 	unsigned long i;
 	int dirty, need_write;
-	int wait = 0;
 
 	if (!bitmap || !bitmap->storage.filemap ||
 	    test_bit(BITMAP_STALE, &bitmap->flags))
@@ -897,16 +896,13 @@ void bitmap_unplug(struct bitmap *bitmap)
 			clear_page_attr(bitmap, i, BITMAP_PAGE_PENDING);
 			write_page(bitmap, bitmap->storage.filemap[i], 0);
 		}
-		if (dirty)
-			wait = 1;
-	}
-	if (wait) { /* if any writes were performed, we need to wait on them */
-		if (bitmap->storage.file)
-			wait_event(bitmap->write_wait,
-				   atomic_read(&bitmap->pending_writes)==0);
-		else
-			md_super_wait(bitmap->mddev);
 	}
+	if (bitmap->storage.file)
+		wait_event(bitmap->write_wait,
+			   atomic_read(&bitmap->pending_writes)==0);
+	else
+		md_super_wait(bitmap->mddev);
+
 	if (test_bit(BITMAP_WRITE_ERROR, &bitmap->flags))
 		bitmap_file_kick(bitmap);
 }
diff --git a/drivers/md/dm-bio-prison.c b/drivers/md/dm-bio-prison.c
index f752d12081ff..be065300e93c 100644
--- a/drivers/md/dm-bio-prison.c
+++ b/drivers/md/dm-bio-prison.c
@@ -14,68 +14,38 @@
 
 /*----------------------------------------------------------------*/
 
-struct bucket {
-	spinlock_t lock;
-	struct hlist_head cells;
-};
+#define MIN_CELLS 1024
 
 struct dm_bio_prison {
+	spinlock_t lock;
 	mempool_t *cell_pool;
-
-	unsigned nr_buckets;
-	unsigned hash_mask;
-	struct bucket *buckets;
+	struct rb_root cells;
 };
 
-/*----------------------------------------------------------------*/
-
-static uint32_t calc_nr_buckets(unsigned nr_cells)
-{
-	uint32_t n = 128;
-
-	nr_cells /= 4;
-	nr_cells = min(nr_cells, 8192u);
-
-	while (n < nr_cells)
-		n <<= 1;
-
-	return n;
-}
-
 static struct kmem_cache *_cell_cache;
 
-static void init_bucket(struct bucket *b)
-{
-	spin_lock_init(&b->lock);
-	INIT_HLIST_HEAD(&b->cells);
-}
+/*----------------------------------------------------------------*/
 
 /*
  * @nr_cells should be the number of cells you want in use _concurrently_.
  * Don't confuse it with the number of distinct keys.
  */
-struct dm_bio_prison *dm_bio_prison_create(unsigned nr_cells)
+struct dm_bio_prison *dm_bio_prison_create(void)
 {
-	unsigned i;
-	uint32_t nr_buckets = calc_nr_buckets(nr_cells);
-	size_t len = sizeof(struct dm_bio_prison) +
-		(sizeof(struct bucket) * nr_buckets);
-	struct dm_bio_prison *prison = kmalloc(len, GFP_KERNEL);
+	struct dm_bio_prison *prison = kmalloc(sizeof(*prison), GFP_KERNEL);
 
 	if (!prison)
 		return NULL;
 
-	prison->cell_pool = mempool_create_slab_pool(nr_cells, _cell_cache);
+	spin_lock_init(&prison->lock);
+
+	prison->cell_pool = mempool_create_slab_pool(MIN_CELLS, _cell_cache);
 	if (!prison->cell_pool) {
 		kfree(prison);
 		return NULL;
 	}
 
-	prison->nr_buckets = nr_buckets;
-	prison->hash_mask = nr_buckets - 1;
-	prison->buckets = (struct bucket *) (prison + 1);
-	for (i = 0; i < nr_buckets; i++)
-		init_bucket(prison->buckets + i);
+	prison->cells = RB_ROOT;
 
 	return prison;
 }
@@ -101,68 +71,73 @@ void dm_bio_prison_free_cell(struct dm_bio_prison *prison,
 }
 EXPORT_SYMBOL_GPL(dm_bio_prison_free_cell);
 
-static uint32_t hash_key(struct dm_bio_prison *prison, struct dm_cell_key *key)
+static void __setup_new_cell(struct dm_cell_key *key,
+			     struct bio *holder,
+			     struct dm_bio_prison_cell *cell)
 {
-	const unsigned long BIG_PRIME = 4294967291UL;
-	uint64_t hash = key->block * BIG_PRIME;
-
-	return (uint32_t) (hash & prison->hash_mask);
+       memcpy(&cell->key, key, sizeof(cell->key));
+       cell->holder = holder;
+       bio_list_init(&cell->bios);
 }
 
-static int keys_equal(struct dm_cell_key *lhs, struct dm_cell_key *rhs)
+static int cmp_keys(struct dm_cell_key *lhs,
+		    struct dm_cell_key *rhs)
 {
-	       return (lhs->virtual == rhs->virtual) &&
-		       (lhs->dev == rhs->dev) &&
-		       (lhs->block == rhs->block);
-}
+	if (lhs->virtual < rhs->virtual)
+		return -1;
 
-static struct bucket *get_bucket(struct dm_bio_prison *prison,
-				 struct dm_cell_key *key)
-{
-	return prison->buckets + hash_key(prison, key);
-}
+	if (lhs->virtual > rhs->virtual)
+		return 1;
 
-static struct dm_bio_prison_cell *__search_bucket(struct bucket *b,
-						  struct dm_cell_key *key)
-{
-	struct dm_bio_prison_cell *cell;
+	if (lhs->dev < rhs->dev)
+		return -1;
 
-	hlist_for_each_entry(cell, &b->cells, list)
-		if (keys_equal(&cell->key, key))
-			return cell;
+	if (lhs->dev > rhs->dev)
+		return 1;
 
-	return NULL;
-}
+	if (lhs->block_end <= rhs->block_begin)
+		return -1;
 
-static void __setup_new_cell(struct bucket *b,
-			     struct dm_cell_key *key,
-			     struct bio *holder,
-			     struct dm_bio_prison_cell *cell)
-{
-	memcpy(&cell->key, key, sizeof(cell->key));
-	cell->holder = holder;
-	bio_list_init(&cell->bios);
-	hlist_add_head(&cell->list, &b->cells);
+	if (lhs->block_begin >= rhs->block_end)
+		return 1;
+
+	return 0;
 }
 
-static int __bio_detain(struct bucket *b,
+static int __bio_detain(struct dm_bio_prison *prison,
 			struct dm_cell_key *key,
 			struct bio *inmate,
 			struct dm_bio_prison_cell *cell_prealloc,
 			struct dm_bio_prison_cell **cell_result)
 {
-	struct dm_bio_prison_cell *cell;
-
-	cell = __search_bucket(b, key);
-	if (cell) {
-		if (inmate)
-			bio_list_add(&cell->bios, inmate);
-		*cell_result = cell;
-		return 1;
+	int r;
+	struct rb_node **new = &prison->cells.rb_node, *parent = NULL;
+
+	while (*new) {
+		struct dm_bio_prison_cell *cell =
+			container_of(*new, struct dm_bio_prison_cell, node);
+
+		r = cmp_keys(key, &cell->key);
+
+		parent = *new;
+		if (r < 0)
+			new = &((*new)->rb_left);
+		else if (r > 0)
+			new = &((*new)->rb_right);
+		else {
+			if (inmate)
+				bio_list_add(&cell->bios, inmate);
+			*cell_result = cell;
+			return 1;
+		}
 	}
 
-	__setup_new_cell(b, key, inmate, cell_prealloc);
+	__setup_new_cell(key, inmate, cell_prealloc);
 	*cell_result = cell_prealloc;
+
+	rb_link_node(&cell_prealloc->node, parent, new);
+	rb_insert_color(&cell_prealloc->node, &prison->cells);
+
 	return 0;
 }
 
@@ -174,11 +149,10 @@ static int bio_detain(struct dm_bio_prison *prison,
 {
 	int r;
 	unsigned long flags;
-	struct bucket *b = get_bucket(prison, key);
 
-	spin_lock_irqsave(&b->lock, flags);
-	r = __bio_detain(b, key, inmate, cell_prealloc, cell_result);
-	spin_unlock_irqrestore(&b->lock, flags);
+	spin_lock_irqsave(&prison->lock, flags);
+	r = __bio_detain(prison, key, inmate, cell_prealloc, cell_result);
+	spin_unlock_irqrestore(&prison->lock, flags);
 
 	return r;
 }
@@ -205,10 +179,11 @@ EXPORT_SYMBOL_GPL(dm_get_cell);
 /*
  * @inmates must have been initialised prior to this call
  */
-static void __cell_release(struct dm_bio_prison_cell *cell,
+static void __cell_release(struct dm_bio_prison *prison,
+			   struct dm_bio_prison_cell *cell,
 			   struct bio_list *inmates)
 {
-	hlist_del(&cell->list);
+	rb_erase(&cell->node, &prison->cells);
 
 	if (inmates) {
 		if (cell->holder)
@@ -222,21 +197,21 @@ void dm_cell_release(struct dm_bio_prison *prison,
 		     struct bio_list *bios)
 {
 	unsigned long flags;
-	struct bucket *b = get_bucket(prison, &cell->key);
 
-	spin_lock_irqsave(&b->lock, flags);
-	__cell_release(cell, bios);
-	spin_unlock_irqrestore(&b->lock, flags);
+	spin_lock_irqsave(&prison->lock, flags);
+	__cell_release(prison, cell, bios);
+	spin_unlock_irqrestore(&prison->lock, flags);
 }
 EXPORT_SYMBOL_GPL(dm_cell_release);
 
 /*
  * Sometimes we don't want the holder, just the additional bios.
  */
-static void __cell_release_no_holder(struct dm_bio_prison_cell *cell,
+static void __cell_release_no_holder(struct dm_bio_prison *prison,
+				     struct dm_bio_prison_cell *cell,
 				     struct bio_list *inmates)
 {
-	hlist_del(&cell->list);
+	rb_erase(&cell->node, &prison->cells);
 	bio_list_merge(inmates, &cell->bios);
 }
 
@@ -245,11 +220,10 @@ void dm_cell_release_no_holder(struct dm_bio_prison *prison,
 			       struct bio_list *inmates)
 {
 	unsigned long flags;
-	struct bucket *b = get_bucket(prison, &cell->key);
 
-	spin_lock_irqsave(&b->lock, flags);
-	__cell_release_no_holder(cell, inmates);
-	spin_unlock_irqrestore(&b->lock, flags);
+	spin_lock_irqsave(&prison->lock, flags);
+	__cell_release_no_holder(prison, cell, inmates);
+	spin_unlock_irqrestore(&prison->lock, flags);
 }
 EXPORT_SYMBOL_GPL(dm_cell_release_no_holder);
 
@@ -267,6 +241,20 @@ void dm_cell_error(struct dm_bio_prison *prison,
 }
 EXPORT_SYMBOL_GPL(dm_cell_error);
 
+void dm_cell_visit_release(struct dm_bio_prison *prison,
+			   void (*visit_fn)(void *, struct dm_bio_prison_cell *),
+			   void *context,
+			   struct dm_bio_prison_cell *cell)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&prison->lock, flags);
+	visit_fn(context, cell);
+	rb_erase(&cell->node, &prison->cells);
+	spin_unlock_irqrestore(&prison->lock, flags);
+}
+EXPORT_SYMBOL_GPL(dm_cell_visit_release);
+
 /*----------------------------------------------------------------*/
 
 #define DEFERRED_SET_SIZE 64
diff --git a/drivers/md/dm-bio-prison.h b/drivers/md/dm-bio-prison.h
index 6805a142b750..74cf01144b1f 100644
--- a/drivers/md/dm-bio-prison.h
+++ b/drivers/md/dm-bio-prison.h
@@ -10,8 +10,8 @@
 #include "persistent-data/dm-block-manager.h" /* FIXME: for dm_block_t */
 #include "dm-thin-metadata.h" /* FIXME: for dm_thin_id */
 
-#include <linux/list.h>
 #include <linux/bio.h>
+#include <linux/rbtree.h>
 
 /*----------------------------------------------------------------*/
 
@@ -23,11 +23,14 @@
  */
 struct dm_bio_prison;
 
-/* FIXME: this needs to be more abstract */
+/*
+ * Keys define a range of blocks within either a virtual or physical
+ * device.
+ */
 struct dm_cell_key {
 	int virtual;
 	dm_thin_id dev;
-	dm_block_t block;
+	dm_block_t block_begin, block_end;
 };
 
 /*
@@ -35,13 +38,15 @@ struct dm_cell_key {
  * themselves.
  */
 struct dm_bio_prison_cell {
-	struct hlist_node list;
+	struct list_head user_list;	/* for client use */
+	struct rb_node node;
+
 	struct dm_cell_key key;
 	struct bio *holder;
 	struct bio_list bios;
 };
 
-struct dm_bio_prison *dm_bio_prison_create(unsigned nr_cells);
+struct dm_bio_prison *dm_bio_prison_create(void);
 void dm_bio_prison_destroy(struct dm_bio_prison *prison);
 
 /*
@@ -57,7 +62,7 @@ void dm_bio_prison_free_cell(struct dm_bio_prison *prison,
 			     struct dm_bio_prison_cell *cell);
 
 /*
- * Creates, or retrieves a cell for the given key.
+ * Creates, or retrieves a cell that overlaps the given key.
  *
  * Returns 1 if pre-existing cell returned, zero if new cell created using
  * @cell_prealloc.
@@ -68,7 +73,8 @@ int dm_get_cell(struct dm_bio_prison *prison,
 		struct dm_bio_prison_cell **cell_result);
 
 /*
- * An atomic op that combines retrieving a cell, and adding a bio to it.
+ * An atomic op that combines retrieving or creating a cell, and adding a
+ * bio to it.
  *
  * Returns 1 if the cell was already held, 0 if @inmate is the new holder.
  */
@@ -87,6 +93,14 @@ void dm_cell_release_no_holder(struct dm_bio_prison *prison,
 void dm_cell_error(struct dm_bio_prison *prison,
 		   struct dm_bio_prison_cell *cell, int error);
 
+/*
+ * Visits the cell and then releases.  Guarantees no new inmates are
+ * inserted between the visit and release.
+ */
+void dm_cell_visit_release(struct dm_bio_prison *prison,
+			   void (*visit_fn)(void *, struct dm_bio_prison_cell *),
+			   void *context, struct dm_bio_prison_cell *cell);
+
 /*----------------------------------------------------------------*/
 
 /*
diff --git a/drivers/md/dm-bufio.c b/drivers/md/dm-bufio.c
index ab472c557d18..c33b49792b87 100644
--- a/drivers/md/dm-bufio.c
+++ b/drivers/md/dm-bufio.c
@@ -14,6 +14,7 @@
 #include <linux/vmalloc.h>
 #include <linux/shrinker.h>
 #include <linux/module.h>
+#include <linux/rbtree.h>
 
 #define DM_MSG_PREFIX "bufio"
 
@@ -34,26 +35,23 @@
 /*
  * Check buffer ages in this interval (seconds)
  */
-#define DM_BUFIO_WORK_TIMER_SECS	10
+#define DM_BUFIO_WORK_TIMER_SECS	30
 
 /*
  * Free buffers when they are older than this (seconds)
  */
-#define DM_BUFIO_DEFAULT_AGE_SECS	60
+#define DM_BUFIO_DEFAULT_AGE_SECS	300
 
 /*
- * The number of bvec entries that are embedded directly in the buffer.
- * If the chunk size is larger, dm-io is used to do the io.
+ * The nr of bytes of cached data to keep around.
  */
-#define DM_BUFIO_INLINE_VECS		16
+#define DM_BUFIO_DEFAULT_RETAIN_BYTES   (256 * 1024)
 
 /*
- * Buffer hash
+ * The number of bvec entries that are embedded directly in the buffer.
+ * If the chunk size is larger, dm-io is used to do the io.
  */
-#define DM_BUFIO_HASH_BITS	20
-#define DM_BUFIO_HASH(block) \
-	((((block) >> DM_BUFIO_HASH_BITS) ^ (block)) & \
-	 ((1 << DM_BUFIO_HASH_BITS) - 1))
+#define DM_BUFIO_INLINE_VECS		16
 
 /*
  * Don't try to use kmem_cache_alloc for blocks larger than this.
@@ -106,7 +104,7 @@ struct dm_bufio_client {
 
 	unsigned minimum_buffers;
 
-	struct hlist_head *cache_hash;
+	struct rb_root buffer_tree;
 	wait_queue_head_t free_buffer_wait;
 
 	int async_write_error;
@@ -135,7 +133,7 @@ enum data_mode {
 };
 
 struct dm_buffer {
-	struct hlist_node hash_list;
+	struct rb_node node;
 	struct list_head lru_list;
 	sector_t block;
 	void *data;
@@ -223,6 +221,7 @@ static DEFINE_SPINLOCK(param_spinlock);
  * Buffers are freed after this timeout
  */
 static unsigned dm_bufio_max_age = DM_BUFIO_DEFAULT_AGE_SECS;
+static unsigned dm_bufio_retain_bytes = DM_BUFIO_DEFAULT_RETAIN_BYTES;
 
 static unsigned long dm_bufio_peak_allocated;
 static unsigned long dm_bufio_allocated_kmem_cache;
@@ -253,6 +252,53 @@ static LIST_HEAD(dm_bufio_all_clients);
  */
 static DEFINE_MUTEX(dm_bufio_clients_lock);
 
+/*----------------------------------------------------------------
+ * A red/black tree acts as an index for all the buffers.
+ *--------------------------------------------------------------*/
+static struct dm_buffer *__find(struct dm_bufio_client *c, sector_t block)
+{
+	struct rb_node *n = c->buffer_tree.rb_node;
+	struct dm_buffer *b;
+
+	while (n) {
+		b = container_of(n, struct dm_buffer, node);
+
+		if (b->block == block)
+			return b;
+
+		n = (b->block < block) ? n->rb_left : n->rb_right;
+	}
+
+	return NULL;
+}
+
+static void __insert(struct dm_bufio_client *c, struct dm_buffer *b)
+{
+	struct rb_node **new = &c->buffer_tree.rb_node, *parent = NULL;
+	struct dm_buffer *found;
+
+	while (*new) {
+		found = container_of(*new, struct dm_buffer, node);
+
+		if (found->block == b->block) {
+			BUG_ON(found != b);
+			return;
+		}
+
+		parent = *new;
+		new = (found->block < b->block) ?
+			&((*new)->rb_left) : &((*new)->rb_right);
+	}
+
+	rb_link_node(&b->node, parent, new);
+	rb_insert_color(&b->node, &c->buffer_tree);
+}
+
+static void __remove(struct dm_bufio_client *c, struct dm_buffer *b)
+{
+	rb_erase(&b->node, &c->buffer_tree);
+}
+
 /*----------------------------------------------------------------*/
 
 static void adjust_total_allocated(enum data_mode data_mode, long diff)
@@ -434,7 +480,7 @@ static void __link_buffer(struct dm_buffer *b, sector_t block, int dirty)
 	b->block = block;
 	b->list_mode = dirty;
 	list_add(&b->lru_list, &c->lru[dirty]);
-	hlist_add_head(&b->hash_list, &c->cache_hash[DM_BUFIO_HASH(block)]);
+	__insert(b->c, b);
 	b->last_accessed = jiffies;
 }
 
@@ -448,7 +494,7 @@ static void __unlink_buffer(struct dm_buffer *b)
 	BUG_ON(!c->n_buffers[b->list_mode]);
 
 	c->n_buffers[b->list_mode]--;
-	hlist_del(&b->hash_list);
+	__remove(b->c, b);
 	list_del(&b->lru_list);
 }
 
@@ -465,6 +511,7 @@ static void __relink_lru(struct dm_buffer *b, int dirty)
 	c->n_buffers[dirty]++;
 	b->list_mode = dirty;
 	list_move(&b->lru_list, &c->lru[dirty]);
+	b->last_accessed = jiffies;
 }
 
 /*----------------------------------------------------------------
@@ -531,6 +578,19 @@ static void use_dmio(struct dm_buffer *b, int rw, sector_t block,
 		end_io(&b->bio, r);
 }
 
+static void inline_endio(struct bio *bio, int error)
+{
+	bio_end_io_t *end_fn = bio->bi_private;
+
+	/*
+	 * Reset the bio to free any attached resources
+	 * (e.g. bio integrity profiles).
+	 */
+	bio_reset(bio);
+
+	end_fn(bio, error);
+}
+
 static void use_inline_bio(struct dm_buffer *b, int rw, sector_t block,
 			   bio_end_io_t *end_io)
 {
@@ -542,7 +602,12 @@ static void use_inline_bio(struct dm_buffer *b, int rw, sector_t block,
 	b->bio.bi_max_vecs = DM_BUFIO_INLINE_VECS;
 	b->bio.bi_iter.bi_sector = block << b->c->sectors_per_block_bits;
 	b->bio.bi_bdev = b->c->bdev;
-	b->bio.bi_end_io = end_io;
+	b->bio.bi_end_io = inline_endio;
+	/*
+	 * Use of .bi_private isn't a problem here because
+	 * the dm_buffer's inline bio is local to bufio.
+	 */
+	b->bio.bi_private = end_io;
 
 	/*
 	 * We assume that if len >= PAGE_SIZE ptr is page-aligned.
@@ -720,7 +785,6 @@ static void __wait_for_free_buffer(struct dm_bufio_client *c)
 
 	io_schedule();
 
-	set_task_state(current, TASK_RUNNING);
 	remove_wait_queue(&c->free_buffer_wait, &wait);
 
 	dm_bufio_lock(c);
@@ -887,23 +951,6 @@ static void __check_watermark(struct dm_bufio_client *c,
 		__write_dirty_buffers_async(c, 1, write_list);
 }
 
-/*
- * Find a buffer in the hash.
- */
-static struct dm_buffer *__find(struct dm_bufio_client *c, sector_t block)
-{
-	struct dm_buffer *b;
-
-	hlist_for_each_entry(b, &c->cache_hash[DM_BUFIO_HASH(block)],
-			     hash_list) {
-		dm_bufio_cond_resched();
-		if (b->block == block)
-			return b;
-	}
-
-	return NULL;
-}
-
 /*----------------------------------------------------------------
  * Getting a buffer
  *--------------------------------------------------------------*/
@@ -1433,48 +1480,55 @@ static void drop_buffers(struct dm_bufio_client *c)
 }
 
 /*
- * Test if the buffer is unused and too old, and commit it.
- * At if noio is set, we must not do any I/O because we hold
- * dm_bufio_clients_lock and we would risk deadlock if the I/O gets rerouted to
- * different bufio client.
+ * We may not be able to evict this buffer if IO pending or the client
+ * is still using it.  Caller is expected to know buffer is too old.
+ *
+ * And if GFP_NOFS is used, we must not do any I/O because we hold
+ * dm_bufio_clients_lock and we would risk deadlock if the I/O gets
+ * rerouted to different bufio client.
  */
-static int __cleanup_old_buffer(struct dm_buffer *b, gfp_t gfp,
-				unsigned long max_jiffies)
+static bool __try_evict_buffer(struct dm_buffer *b, gfp_t gfp)
 {
-	if (jiffies - b->last_accessed < max_jiffies)
-		return 0;
-
-	if (!(gfp & __GFP_IO)) {
+	if (!(gfp & __GFP_FS)) {
 		if (test_bit(B_READING, &b->state) ||
 		    test_bit(B_WRITING, &b->state) ||
 		    test_bit(B_DIRTY, &b->state))
-			return 0;
+			return false;
 	}
 
 	if (b->hold_count)
-		return 0;
+		return false;
 
 	__make_buffer_clean(b);
 	__unlink_buffer(b);
 	__free_buffer_wake(b);
 
-	return 1;
+	return true;
+}
+
+static unsigned get_retain_buffers(struct dm_bufio_client *c)
+{
+        unsigned retain_bytes = ACCESS_ONCE(dm_bufio_retain_bytes);
+        return retain_bytes / c->block_size;
 }
 
-static long __scan(struct dm_bufio_client *c, unsigned long nr_to_scan,
-		   gfp_t gfp_mask)
+static unsigned long __scan(struct dm_bufio_client *c, unsigned long nr_to_scan,
+			    gfp_t gfp_mask)
 {
 	int l;
 	struct dm_buffer *b, *tmp;
-	long freed = 0;
+	unsigned long freed = 0;
+	unsigned long count = nr_to_scan;
+	unsigned retain_target = get_retain_buffers(c);
 
 	for (l = 0; l < LIST_SIZE; l++) {
 		list_for_each_entry_safe_reverse(b, tmp, &c->lru[l], lru_list) {
-			freed += __cleanup_old_buffer(b, gfp_mask, 0);
-			if (!--nr_to_scan)
-				break;
+			if (__try_evict_buffer(b, gfp_mask))
+				freed++;
+			if (!--nr_to_scan || ((count - freed) <= retain_target))
+				return freed;
+			dm_bufio_cond_resched();
 		}
-		dm_bufio_cond_resched();
 	}
 	return freed;
 }
@@ -1486,7 +1540,7 @@ dm_bufio_shrink_scan(struct shrinker *shrink, struct shrink_control *sc)
 	unsigned long freed;
 
 	c = container_of(shrink, struct dm_bufio_client, shrinker);
-	if (sc->gfp_mask & __GFP_IO)
+	if (sc->gfp_mask & __GFP_FS)
 		dm_bufio_lock(c);
 	else if (!dm_bufio_trylock(c))
 		return SHRINK_STOP;
@@ -1503,7 +1557,7 @@ dm_bufio_shrink_count(struct shrinker *shrink, struct shrink_control *sc)
 	unsigned long count;
 
 	c = container_of(shrink, struct dm_bufio_client, shrinker);
-	if (sc->gfp_mask & __GFP_IO)
+	if (sc->gfp_mask & __GFP_FS)
 		dm_bufio_lock(c);
 	else if (!dm_bufio_trylock(c))
 		return 0;
@@ -1533,11 +1587,7 @@ struct dm_bufio_client *dm_bufio_client_create(struct block_device *bdev, unsign
 		r = -ENOMEM;
 		goto bad_client;
 	}
-	c->cache_hash = vmalloc(sizeof(struct hlist_head) << DM_BUFIO_HASH_BITS);
-	if (!c->cache_hash) {
-		r = -ENOMEM;
-		goto bad_hash;
-	}
+	c->buffer_tree = RB_ROOT;
 
 	c->bdev = bdev;
 	c->block_size = block_size;
@@ -1556,9 +1606,6 @@ struct dm_bufio_client *dm_bufio_client_create(struct block_device *bdev, unsign
 		c->n_buffers[i] = 0;
 	}
 
-	for (i = 0; i < 1 << DM_BUFIO_HASH_BITS; i++)
-		INIT_HLIST_HEAD(&c->cache_hash[i]);
-
 	mutex_init(&c->lock);
 	INIT_LIST_HEAD(&c->reserved_buffers);
 	c->need_reserved_buffers = reserved_buffers;
@@ -1632,8 +1679,6 @@ bad_cache:
 	}
 	dm_io_client_destroy(c->dm_io);
 bad_dm_io:
-	vfree(c->cache_hash);
-bad_hash:
 	kfree(c);
 bad_client:
 	return ERR_PTR(r);
@@ -1660,9 +1705,7 @@ void dm_bufio_client_destroy(struct dm_bufio_client *c)
 
 	mutex_unlock(&dm_bufio_clients_lock);
 
-	for (i = 0; i < 1 << DM_BUFIO_HASH_BITS; i++)
-		BUG_ON(!hlist_empty(&c->cache_hash[i]));
-
+	BUG_ON(!RB_EMPTY_ROOT(&c->buffer_tree));
 	BUG_ON(c->need_reserved_buffers);
 
 	while (!list_empty(&c->reserved_buffers)) {
@@ -1680,36 +1723,60 @@ void dm_bufio_client_destroy(struct dm_bufio_client *c)
 		BUG_ON(c->n_buffers[i]);
 
 	dm_io_client_destroy(c->dm_io);
-	vfree(c->cache_hash);
 	kfree(c);
 }
 EXPORT_SYMBOL_GPL(dm_bufio_client_destroy);
 
-static void cleanup_old_buffers(void)
+static unsigned get_max_age_hz(void)
 {
-	unsigned long max_age = ACCESS_ONCE(dm_bufio_max_age);
-	struct dm_bufio_client *c;
+	unsigned max_age = ACCESS_ONCE(dm_bufio_max_age);
 
-	if (max_age > ULONG_MAX / HZ)
-		max_age = ULONG_MAX / HZ;
+	if (max_age > UINT_MAX / HZ)
+		max_age = UINT_MAX / HZ;
 
-	mutex_lock(&dm_bufio_clients_lock);
-	list_for_each_entry(c, &dm_bufio_all_clients, client_list) {
-		if (!dm_bufio_trylock(c))
-			continue;
+	return max_age * HZ;
+}
 
-		while (!list_empty(&c->lru[LIST_CLEAN])) {
-			struct dm_buffer *b;
-			b = list_entry(c->lru[LIST_CLEAN].prev,
-				       struct dm_buffer, lru_list);
-			if (!__cleanup_old_buffer(b, 0, max_age * HZ))
-				break;
-			dm_bufio_cond_resched();
-		}
+static bool older_than(struct dm_buffer *b, unsigned long age_hz)
+{
+	return (jiffies - b->last_accessed) >= age_hz;
+}
+
+static void __evict_old_buffers(struct dm_bufio_client *c, unsigned long age_hz)
+{
+	struct dm_buffer *b, *tmp;
+	unsigned retain_target = get_retain_buffers(c);
+	unsigned count;
+
+	dm_bufio_lock(c);
+
+	count = c->n_buffers[LIST_CLEAN] + c->n_buffers[LIST_DIRTY];
+	list_for_each_entry_safe_reverse(b, tmp, &c->lru[LIST_CLEAN], lru_list) {
+		if (count <= retain_target)
+			break;
+
+		if (!older_than(b, age_hz))
+			break;
+
+		if (__try_evict_buffer(b, 0))
+			count--;
 
-		dm_bufio_unlock(c);
 		dm_bufio_cond_resched();
 	}
+
+	dm_bufio_unlock(c);
+}
+
+static void cleanup_old_buffers(void)
+{
+	unsigned long max_age_hz = get_max_age_hz();
+	struct dm_bufio_client *c;
+
+	mutex_lock(&dm_bufio_clients_lock);
+
+	list_for_each_entry(c, &dm_bufio_all_clients, client_list)
+		__evict_old_buffers(c, max_age_hz);
+
 	mutex_unlock(&dm_bufio_clients_lock);
 }
 
@@ -1834,6 +1901,9 @@ MODULE_PARM_DESC(max_cache_size_bytes, "Size of metadata cache");
 module_param_named(max_age_seconds, dm_bufio_max_age, uint, S_IRUGO | S_IWUSR);
 MODULE_PARM_DESC(max_age_seconds, "Max age of a buffer in seconds");
 
+module_param_named(retain_bytes, dm_bufio_retain_bytes, uint, S_IRUGO | S_IWUSR);
+MODULE_PARM_DESC(retain_bytes, "Try to keep at least this many bytes cached in memory");
+
 module_param_named(peak_allocated_bytes, dm_bufio_peak_allocated, ulong, S_IRUGO | S_IWUSR);
 MODULE_PARM_DESC(peak_allocated_bytes, "Tracks the maximum allocated memory");
 
diff --git a/drivers/md/dm-cache-block-types.h b/drivers/md/dm-cache-block-types.h
index aac0e2df06be..bed4ad4e1b7c 100644
--- a/drivers/md/dm-cache-block-types.h
+++ b/drivers/md/dm-cache-block-types.h
@@ -19,6 +19,7 @@
 
 typedef dm_block_t __bitwise__ dm_oblock_t;
 typedef uint32_t __bitwise__ dm_cblock_t;
+typedef dm_block_t __bitwise__ dm_dblock_t;
 
 static inline dm_oblock_t to_oblock(dm_block_t b)
 {
@@ -40,4 +41,14 @@ static inline uint32_t from_cblock(dm_cblock_t b)
 	return (__force uint32_t) b;
 }
 
+static inline dm_dblock_t to_dblock(dm_block_t b)
+{
+	return (__force dm_dblock_t) b;
+}
+
+static inline dm_block_t from_dblock(dm_dblock_t b)
+{
+	return (__force dm_block_t) b;
+}
+
 #endif /* DM_CACHE_BLOCK_TYPES_H */
diff --git a/drivers/md/dm-cache-metadata.c b/drivers/md/dm-cache-metadata.c
index 06709257adde..c1c010498a21 100644
--- a/drivers/md/dm-cache-metadata.c
+++ b/drivers/md/dm-cache-metadata.c
@@ -94,6 +94,9 @@ struct cache_disk_superblock {
 } __packed;
 
 struct dm_cache_metadata {
+	atomic_t ref_count;
+	struct list_head list;
+
 	struct block_device *bdev;
 	struct dm_block_manager *bm;
 	struct dm_space_map *metadata_sm;
@@ -109,7 +112,7 @@ struct dm_cache_metadata {
 	dm_block_t discard_root;
 
 	sector_t discard_block_size;
-	dm_oblock_t discard_nr_blocks;
+	dm_dblock_t discard_nr_blocks;
 
 	sector_t data_block_size;
 	dm_cblock_t cache_blocks;
@@ -329,7 +332,7 @@ static int __write_initial_superblock(struct dm_cache_metadata *cmd)
 	disk_super->hint_root = cpu_to_le64(cmd->hint_root);
 	disk_super->discard_root = cpu_to_le64(cmd->discard_root);
 	disk_super->discard_block_size = cpu_to_le64(cmd->discard_block_size);
-	disk_super->discard_nr_blocks = cpu_to_le64(from_oblock(cmd->discard_nr_blocks));
+	disk_super->discard_nr_blocks = cpu_to_le64(from_dblock(cmd->discard_nr_blocks));
 	disk_super->metadata_block_size = cpu_to_le32(DM_CACHE_METADATA_BLOCK_SIZE);
 	disk_super->data_block_size = cpu_to_le32(cmd->data_block_size);
 	disk_super->cache_blocks = cpu_to_le32(0);
@@ -528,7 +531,7 @@ static void read_superblock_fields(struct dm_cache_metadata *cmd,
 	cmd->hint_root = le64_to_cpu(disk_super->hint_root);
 	cmd->discard_root = le64_to_cpu(disk_super->discard_root);
 	cmd->discard_block_size = le64_to_cpu(disk_super->discard_block_size);
-	cmd->discard_nr_blocks = to_oblock(le64_to_cpu(disk_super->discard_nr_blocks));
+	cmd->discard_nr_blocks = to_dblock(le64_to_cpu(disk_super->discard_nr_blocks));
 	cmd->data_block_size = le32_to_cpu(disk_super->data_block_size);
 	cmd->cache_blocks = to_cblock(le32_to_cpu(disk_super->cache_blocks));
 	strncpy(cmd->policy_name, disk_super->policy_name, sizeof(cmd->policy_name));
@@ -626,7 +629,7 @@ static int __commit_transaction(struct dm_cache_metadata *cmd,
 	disk_super->hint_root = cpu_to_le64(cmd->hint_root);
 	disk_super->discard_root = cpu_to_le64(cmd->discard_root);
 	disk_super->discard_block_size = cpu_to_le64(cmd->discard_block_size);
-	disk_super->discard_nr_blocks = cpu_to_le64(from_oblock(cmd->discard_nr_blocks));
+	disk_super->discard_nr_blocks = cpu_to_le64(from_dblock(cmd->discard_nr_blocks));
 	disk_super->cache_blocks = cpu_to_le32(from_cblock(cmd->cache_blocks));
 	strncpy(disk_super->policy_name, cmd->policy_name, sizeof(disk_super->policy_name));
 	disk_super->policy_version[0] = cpu_to_le32(cmd->policy_version[0]);
@@ -669,10 +672,10 @@ static void unpack_value(__le64 value_le, dm_oblock_t *block, unsigned *flags)
 
 /*----------------------------------------------------------------*/
 
-struct dm_cache_metadata *dm_cache_metadata_open(struct block_device *bdev,
-						 sector_t data_block_size,
-						 bool may_format_device,
-						 size_t policy_hint_size)
+static struct dm_cache_metadata *metadata_open(struct block_device *bdev,
+					       sector_t data_block_size,
+					       bool may_format_device,
+					       size_t policy_hint_size)
 {
 	int r;
 	struct dm_cache_metadata *cmd;
@@ -680,9 +683,10 @@ struct dm_cache_metadata *dm_cache_metadata_open(struct block_device *bdev,
 	cmd = kzalloc(sizeof(*cmd), GFP_KERNEL);
 	if (!cmd) {
 		DMERR("could not allocate metadata struct");
-		return NULL;
+		return ERR_PTR(-ENOMEM);
 	}
 
+	atomic_set(&cmd->ref_count, 1);
 	init_rwsem(&cmd->root_lock);
 	cmd->bdev = bdev;
 	cmd->data_block_size = data_block_size;
@@ -705,10 +709,96 @@ struct dm_cache_metadata *dm_cache_metadata_open(struct block_device *bdev,
 	return cmd;
 }
 
+/*
+ * We keep a little list of ref counted metadata objects to prevent two
+ * different target instances creating separate bufio instances.  This is
+ * an issue if a table is reloaded before the suspend.
+ */
+static DEFINE_MUTEX(table_lock);
+static LIST_HEAD(table);
+
+static struct dm_cache_metadata *lookup(struct block_device *bdev)
+{
+	struct dm_cache_metadata *cmd;
+
+	list_for_each_entry(cmd, &table, list)
+		if (cmd->bdev == bdev) {
+			atomic_inc(&cmd->ref_count);
+			return cmd;
+		}
+
+	return NULL;
+}
+
+static struct dm_cache_metadata *lookup_or_open(struct block_device *bdev,
+						sector_t data_block_size,
+						bool may_format_device,
+						size_t policy_hint_size)
+{
+	struct dm_cache_metadata *cmd, *cmd2;
+
+	mutex_lock(&table_lock);
+	cmd = lookup(bdev);
+	mutex_unlock(&table_lock);
+
+	if (cmd)
+		return cmd;
+
+	cmd = metadata_open(bdev, data_block_size, may_format_device, policy_hint_size);
+	if (!IS_ERR(cmd)) {
+		mutex_lock(&table_lock);
+		cmd2 = lookup(bdev);
+		if (cmd2) {
+			mutex_unlock(&table_lock);
+			__destroy_persistent_data_objects(cmd);
+			kfree(cmd);
+			return cmd2;
+		}
+		list_add(&cmd->list, &table);
+		mutex_unlock(&table_lock);
+	}
+
+	return cmd;
+}
+
+static bool same_params(struct dm_cache_metadata *cmd, sector_t data_block_size)
+{
+	if (cmd->data_block_size != data_block_size) {
+		DMERR("data_block_size (%llu) different from that in metadata (%llu)\n",
+		      (unsigned long long) data_block_size,
+		      (unsigned long long) cmd->data_block_size);
+		return false;
+	}
+
+	return true;
+}
+
+struct dm_cache_metadata *dm_cache_metadata_open(struct block_device *bdev,
+						 sector_t data_block_size,
+						 bool may_format_device,
+						 size_t policy_hint_size)
+{
+	struct dm_cache_metadata *cmd = lookup_or_open(bdev, data_block_size,
+						       may_format_device, policy_hint_size);
+
+	if (!IS_ERR(cmd) && !same_params(cmd, data_block_size)) {
+		dm_cache_metadata_close(cmd);
+		return ERR_PTR(-EINVAL);
+	}
+
+	return cmd;
+}
+
 void dm_cache_metadata_close(struct dm_cache_metadata *cmd)
 {
-	__destroy_persistent_data_objects(cmd);
-	kfree(cmd);
+	if (atomic_dec_and_test(&cmd->ref_count)) {
+		mutex_lock(&table_lock);
+		list_del(&cmd->list);
+		mutex_unlock(&table_lock);
+
+		__destroy_persistent_data_objects(cmd);
+		kfree(cmd);
+	}
 }
 
 /*
@@ -797,15 +887,15 @@ out:
 
 int dm_cache_discard_bitset_resize(struct dm_cache_metadata *cmd,
 				   sector_t discard_block_size,
-				   dm_oblock_t new_nr_entries)
+				   dm_dblock_t new_nr_entries)
 {
 	int r;
 
 	down_write(&cmd->root_lock);
 	r = dm_bitset_resize(&cmd->discard_info,
 			     cmd->discard_root,
-			     from_oblock(cmd->discard_nr_blocks),
-			     from_oblock(new_nr_entries),
+			     from_dblock(cmd->discard_nr_blocks),
+			     from_dblock(new_nr_entries),
 			     false, &cmd->discard_root);
 	if (!r) {
 		cmd->discard_block_size = discard_block_size;
@@ -818,28 +908,28 @@ int dm_cache_discard_bitset_resize(struct dm_cache_metadata *cmd,
 	return r;
 }
 
-static int __set_discard(struct dm_cache_metadata *cmd, dm_oblock_t b)
+static int __set_discard(struct dm_cache_metadata *cmd, dm_dblock_t b)
 {
 	return dm_bitset_set_bit(&cmd->discard_info, cmd->discard_root,
-				 from_oblock(b), &cmd->discard_root);
+				 from_dblock(b), &cmd->discard_root);
 }
 
-static int __clear_discard(struct dm_cache_metadata *cmd, dm_oblock_t b)
+static int __clear_discard(struct dm_cache_metadata *cmd, dm_dblock_t b)
 {
 	return dm_bitset_clear_bit(&cmd->discard_info, cmd->discard_root,
-				   from_oblock(b), &cmd->discard_root);
+				   from_dblock(b), &cmd->discard_root);
 }
 
-static int __is_discarded(struct dm_cache_metadata *cmd, dm_oblock_t b,
+static int __is_discarded(struct dm_cache_metadata *cmd, dm_dblock_t b,
 			  bool *is_discarded)
 {
 	return dm_bitset_test_bit(&cmd->discard_info, cmd->discard_root,
-				  from_oblock(b), &cmd->discard_root,
+				  from_dblock(b), &cmd->discard_root,
 				  is_discarded);
 }
 
 static int __discard(struct dm_cache_metadata *cmd,
-		     dm_oblock_t dblock, bool discard)
+		     dm_dblock_t dblock, bool discard)
 {
 	int r;
 
@@ -852,7 +942,7 @@ static int __discard(struct dm_cache_metadata *cmd,
 }
 
 int dm_cache_set_discard(struct dm_cache_metadata *cmd,
-			 dm_oblock_t dblock, bool discard)
+			 dm_dblock_t dblock, bool discard)
 {
 	int r;
 
@@ -870,8 +960,8 @@ static int __load_discards(struct dm_cache_metadata *cmd,
 	dm_block_t b;
 	bool discard;
 
-	for (b = 0; b < from_oblock(cmd->discard_nr_blocks); b++) {
-		dm_oblock_t dblock = to_oblock(b);
+	for (b = 0; b < from_dblock(cmd->discard_nr_blocks); b++) {
+		dm_dblock_t dblock = to_dblock(b);
 
 		if (cmd->clean_when_opened) {
 			r = __is_discarded(cmd, dblock, &discard);
diff --git a/drivers/md/dm-cache-metadata.h b/drivers/md/dm-cache-metadata.h
index 7383c90ccdb8..4ecc403be283 100644
--- a/drivers/md/dm-cache-metadata.h
+++ b/drivers/md/dm-cache-metadata.h
@@ -70,14 +70,14 @@ dm_cblock_t dm_cache_size(struct dm_cache_metadata *cmd);
 
 int dm_cache_discard_bitset_resize(struct dm_cache_metadata *cmd,
 				   sector_t discard_block_size,
-				   dm_oblock_t new_nr_entries);
+				   dm_dblock_t new_nr_entries);
 
 typedef int (*load_discard_fn)(void *context, sector_t discard_block_size,
-			       dm_oblock_t dblock, bool discarded);
+			       dm_dblock_t dblock, bool discarded);
 int dm_cache_load_discards(struct dm_cache_metadata *cmd,
 			   load_discard_fn fn, void *context);
 
-int dm_cache_set_discard(struct dm_cache_metadata *cmd, dm_oblock_t dblock, bool discard);
+int dm_cache_set_discard(struct dm_cache_metadata *cmd, dm_dblock_t dblock, bool discard);
 
 int dm_cache_remove_mapping(struct dm_cache_metadata *cmd, dm_cblock_t cblock);
 int dm_cache_insert_mapping(struct dm_cache_metadata *cmd, dm_cblock_t cblock, dm_oblock_t oblock);
diff --git a/drivers/md/dm-cache-policy-mq.c b/drivers/md/dm-cache-policy-mq.c
index 0e385e40909e..13f547a4eeb6 100644
--- a/drivers/md/dm-cache-policy-mq.c
+++ b/drivers/md/dm-cache-policy-mq.c
@@ -181,24 +181,30 @@ static void queue_shift_down(struct queue *q)
  * Gives us the oldest entry of the lowest popoulated level.  If the first
  * level is emptied then we shift down one level.
  */
-static struct list_head *queue_pop(struct queue *q)
+static struct list_head *queue_peek(struct queue *q)
 {
 	unsigned level;
-	struct list_head *r;
 
 	for (level = 0; level < NR_QUEUE_LEVELS; level++)
-		if (!list_empty(q->qs + level)) {
-			r = q->qs[level].next;
-			list_del(r);
+		if (!list_empty(q->qs + level))
+			return q->qs[level].next;
 
-			/* have we just emptied the bottom level? */
-			if (level == 0 && list_empty(q->qs))
-				queue_shift_down(q);
+	return NULL;
+}
 
-			return r;
-		}
+static struct list_head *queue_pop(struct queue *q)
+{
+	struct list_head *r = queue_peek(q);
 
-	return NULL;
+	if (r) {
+		list_del(r);
+
+		/* have we just emptied the bottom level? */
+		if (list_empty(q->qs))
+			queue_shift_down(q);
+	}
+
+	return r;
 }
 
 static struct list_head *list_pop(struct list_head *lh)
@@ -383,13 +389,6 @@ struct mq_policy {
 	unsigned generation;
 	unsigned generation_period; /* in lookups (will probably change) */
 
-	/*
-	 * Entries in the pre_cache whose hit count passes the promotion
-	 * threshold move to the cache proper.  Working out the correct
-	 * value for the promotion_threshold is crucial to this policy.
-	 */
-	unsigned promote_threshold;
-
 	unsigned discard_promote_adjustment;
 	unsigned read_promote_adjustment;
 	unsigned write_promote_adjustment;
@@ -406,6 +405,7 @@ struct mq_policy {
 #define DEFAULT_DISCARD_PROMOTE_ADJUSTMENT 1
 #define DEFAULT_READ_PROMOTE_ADJUSTMENT 4
 #define DEFAULT_WRITE_PROMOTE_ADJUSTMENT 8
+#define DISCOURAGE_DEMOTING_DIRTY_THRESHOLD 128
 
 /*----------------------------------------------------------------*/
 
@@ -518,6 +518,12 @@ static struct entry *pop(struct mq_policy *mq, struct queue *q)
 	return e;
 }
 
+static struct entry *peek(struct queue *q)
+{
+	struct list_head *h = queue_peek(q);
+	return h ? container_of(h, struct entry, list) : NULL;
+}
+
 /*
  * Has this entry already been updated?
  */
@@ -570,10 +576,6 @@ static void check_generation(struct mq_policy *mq)
 					break;
 			}
 		}
-
-		mq->promote_threshold = nr ? total / nr : 1;
-		if (mq->promote_threshold * nr < total)
-			mq->promote_threshold++;
 	}
 }
 
@@ -641,6 +643,30 @@ static int demote_cblock(struct mq_policy *mq, dm_oblock_t *oblock)
 }
 
 /*
+ * Entries in the pre_cache whose hit count passes the promotion
+ * threshold move to the cache proper.  Working out the correct
+ * value for the promotion_threshold is crucial to this policy.
+ */
+static unsigned promote_threshold(struct mq_policy *mq)
+{
+	struct entry *e;
+
+	if (any_free_cblocks(mq))
+		return 0;
+
+	e = peek(&mq->cache_clean);
+	if (e)
+		return e->hit_count;
+
+	e = peek(&mq->cache_dirty);
+	if (e)
+		return e->hit_count + DISCOURAGE_DEMOTING_DIRTY_THRESHOLD;
+
+	/* This should never happen */
+	return 0;
+}
+
+/*
  * We modify the basic promotion_threshold depending on the specific io.
  *
  * If the origin block has been discarded then there's no cost to copy it
@@ -653,7 +679,7 @@ static unsigned adjusted_promote_threshold(struct mq_policy *mq,
 					   bool discarded_oblock, int data_dir)
 {
 	if (data_dir == READ)
-		return mq->promote_threshold + mq->read_promote_adjustment;
+		return promote_threshold(mq) + mq->read_promote_adjustment;
 
 	if (discarded_oblock && (any_free_cblocks(mq) || any_clean_cblocks(mq))) {
 		/*
@@ -663,7 +689,7 @@ static unsigned adjusted_promote_threshold(struct mq_policy *mq,
 		return mq->discard_promote_adjustment;
 	}
 
-	return mq->promote_threshold + mq->write_promote_adjustment;
+	return promote_threshold(mq) + mq->write_promote_adjustment;
 }
 
 static bool should_promote(struct mq_policy *mq, struct entry *e,
@@ -839,7 +865,8 @@ static int map(struct mq_policy *mq, dm_oblock_t oblock,
 	if (e && in_cache(mq, e))
 		r = cache_entry_found(mq, e, result);
 
-	else if (iot_pattern(&mq->tracker) == PATTERN_SEQUENTIAL)
+	else if (mq->tracker.thresholds[PATTERN_SEQUENTIAL] &&
+		 iot_pattern(&mq->tracker) == PATTERN_SEQUENTIAL)
 		result->op = POLICY_MISS;
 
 	else if (e)
@@ -1230,7 +1257,6 @@ static struct dm_cache_policy *mq_create(dm_cblock_t cache_size,
 	mq->tick = 0;
 	mq->hit_count = 0;
 	mq->generation = 0;
-	mq->promote_threshold = 0;
 	mq->discard_promote_adjustment = DEFAULT_DISCARD_PROMOTE_ADJUSTMENT;
 	mq->read_promote_adjustment = DEFAULT_READ_PROMOTE_ADJUSTMENT;
 	mq->write_promote_adjustment = DEFAULT_WRITE_PROMOTE_ADJUSTMENT;
@@ -1265,7 +1291,7 @@ bad_pre_cache_init:
 
 static struct dm_cache_policy_type mq_policy_type = {
 	.name = "mq",
-	.version = {1, 2, 0},
+	.version = {1, 3, 0},
 	.hint_size = 4,
 	.owner = THIS_MODULE,
 	.create = mq_create
@@ -1273,7 +1299,7 @@ static struct dm_cache_policy_type mq_policy_type = {
 
 static struct dm_cache_policy_type default_policy_type = {
 	.name = "default",
-	.version = {1, 2, 0},
+	.version = {1, 3, 0},
 	.hint_size = 4,
 	.owner = THIS_MODULE,
 	.create = mq_create,
diff --git a/drivers/md/dm-cache-target.c b/drivers/md/dm-cache-target.c
index 7130505c2425..e1650539cc2f 100644
--- a/drivers/md/dm-cache-target.c
+++ b/drivers/md/dm-cache-target.c
@@ -95,7 +95,6 @@ static void dm_unhook_bio(struct dm_hook_info *h, struct bio *bio)
 
 /*----------------------------------------------------------------*/
 
-#define PRISON_CELLS 1024
 #define MIGRATION_POOL_SIZE 128
 #define COMMIT_PERIOD HZ
 #define MIGRATION_COUNT_WINDOW 10
@@ -222,7 +221,13 @@ struct cache {
 	struct list_head need_commit_migrations;
 	sector_t migration_threshold;
 	wait_queue_head_t migration_wait;
-	atomic_t nr_migrations;
+	atomic_t nr_allocated_migrations;
+
+	/*
+	 * The number of in flight migrations that are performing
+	 * background io. eg, promotion, writeback.
+	 */
+	atomic_t nr_io_migrations;
 
 	wait_queue_head_t quiescing_wait;
 	atomic_t quiescing;
@@ -237,8 +242,9 @@ struct cache {
 	/*
 	 * origin_blocks entries, discarded if set.
 	 */
-	dm_oblock_t discard_nr_blocks;
+	dm_dblock_t discard_nr_blocks;
 	unsigned long *discard_bitset;
+	uint32_t discard_block_size; /* a power of 2 times sectors per block */
 
 	/*
 	 * Rather than reconstructing the table line for the status we just
@@ -258,7 +264,6 @@ struct cache {
 	struct dm_deferred_set *all_io_ds;
 
 	mempool_t *migration_pool;
-	struct dm_cache_migration *next_migration;
 
 	struct dm_cache_policy *policy;
 	unsigned policy_nr_args;
@@ -310,6 +315,7 @@ struct dm_cache_migration {
 	dm_cblock_t cblock;
 
 	bool err:1;
+	bool discard:1;
 	bool writeback:1;
 	bool demote:1;
 	bool promote:1;
@@ -349,10 +355,31 @@ static void free_prison_cell(struct cache *cache, struct dm_bio_prison_cell *cel
 	dm_bio_prison_free_cell(cache->prison, cell);
 }
 
+static struct dm_cache_migration *alloc_migration(struct cache *cache)
+{
+	struct dm_cache_migration *mg;
+
+	mg = mempool_alloc(cache->migration_pool, GFP_NOWAIT);
+	if (mg) {
+		mg->cache = cache;
+		atomic_inc(&mg->cache->nr_allocated_migrations);
+	}
+
+	return mg;
+}
+
+static void free_migration(struct dm_cache_migration *mg)
+{
+	if (atomic_dec_and_test(&mg->cache->nr_allocated_migrations))
+		wake_up(&mg->cache->migration_wait);
+
+	mempool_free(mg, mg->cache->migration_pool);
+}
+
 static int prealloc_data_structs(struct cache *cache, struct prealloc *p)
 {
 	if (!p->mg) {
-		p->mg = mempool_alloc(cache->migration_pool, GFP_NOWAIT);
+		p->mg = alloc_migration(cache);
 		if (!p->mg)
 			return -ENOMEM;
 	}
@@ -381,7 +408,7 @@ static void prealloc_free_structs(struct cache *cache, struct prealloc *p)
 		free_prison_cell(cache, p->cell1);
 
 	if (p->mg)
-		mempool_free(p->mg, cache->migration_pool);
+		free_migration(p->mg);
 }
 
 static struct dm_cache_migration *prealloc_get_migration(struct prealloc *p)
@@ -433,11 +460,12 @@ static void prealloc_put_cell(struct prealloc *p, struct dm_bio_prison_cell *cel
 
 /*----------------------------------------------------------------*/
 
-static void build_key(dm_oblock_t oblock, struct dm_cell_key *key)
+static void build_key(dm_oblock_t begin, dm_oblock_t end, struct dm_cell_key *key)
 {
 	key->virtual = 0;
 	key->dev = 0;
-	key->block = from_oblock(oblock);
+	key->block_begin = from_oblock(begin);
+	key->block_end = from_oblock(end);
 }
 
 /*
@@ -447,15 +475,15 @@ static void build_key(dm_oblock_t oblock, struct dm_cell_key *key)
  */
 typedef void (*cell_free_fn)(void *context, struct dm_bio_prison_cell *cell);
 
-static int bio_detain(struct cache *cache, dm_oblock_t oblock,
-		      struct bio *bio, struct dm_bio_prison_cell *cell_prealloc,
-		      cell_free_fn free_fn, void *free_context,
-		      struct dm_bio_prison_cell **cell_result)
+static int bio_detain_range(struct cache *cache, dm_oblock_t oblock_begin, dm_oblock_t oblock_end,
+			    struct bio *bio, struct dm_bio_prison_cell *cell_prealloc,
+			    cell_free_fn free_fn, void *free_context,
+			    struct dm_bio_prison_cell **cell_result)
 {
 	int r;
 	struct dm_cell_key key;
 
-	build_key(oblock, &key);
+	build_key(oblock_begin, oblock_end, &key);
 	r = dm_bio_detain(cache->prison, &key, bio, cell_prealloc, cell_result);
 	if (r)
 		free_fn(free_context, cell_prealloc);
@@ -463,6 +491,16 @@ static int bio_detain(struct cache *cache, dm_oblock_t oblock,
 	return r;
 }
 
+static int bio_detain(struct cache *cache, dm_oblock_t oblock,
+		      struct bio *bio, struct dm_bio_prison_cell *cell_prealloc,
+		      cell_free_fn free_fn, void *free_context,
+		      struct dm_bio_prison_cell **cell_result)
+{
+	dm_oblock_t end = to_oblock(from_oblock(oblock) + 1ULL);
+	return bio_detain_range(cache, oblock, end, bio,
+				cell_prealloc, free_fn, free_context, cell_result);
+}
+
 static int get_cell(struct cache *cache,
 		    dm_oblock_t oblock,
 		    struct prealloc *structs,
@@ -474,7 +512,7 @@ static int get_cell(struct cache *cache,
 
 	cell_prealloc = prealloc_get_cell(structs);
 
-	build_key(oblock, &key);
+	build_key(oblock, to_oblock(from_oblock(oblock) + 1ULL), &key);
 	r = dm_get_cell(cache->prison, &key, cell_prealloc, cell_result);
 	if (r)
 		prealloc_put_cell(structs, cell_prealloc);
@@ -524,33 +562,57 @@ static dm_block_t block_div(dm_block_t b, uint32_t n)
 	return b;
 }
 
-static void set_discard(struct cache *cache, dm_oblock_t b)
+static dm_block_t oblocks_per_dblock(struct cache *cache)
+{
+	dm_block_t oblocks = cache->discard_block_size;
+
+	if (block_size_is_power_of_two(cache))
+		oblocks >>= cache->sectors_per_block_shift;
+	else
+		oblocks = block_div(oblocks, cache->sectors_per_block);
+
+	return oblocks;
+}
+
+static dm_dblock_t oblock_to_dblock(struct cache *cache, dm_oblock_t oblock)
+{
+	return to_dblock(block_div(from_oblock(oblock),
+				   oblocks_per_dblock(cache)));
+}
+
+static dm_oblock_t dblock_to_oblock(struct cache *cache, dm_dblock_t dblock)
+{
+	return to_oblock(from_dblock(dblock) * oblocks_per_dblock(cache));
+}
+
+static void set_discard(struct cache *cache, dm_dblock_t b)
 {
 	unsigned long flags;
 
+	BUG_ON(from_dblock(b) >= from_dblock(cache->discard_nr_blocks));
 	atomic_inc(&cache->stats.discard_count);
 
 	spin_lock_irqsave(&cache->lock, flags);
-	set_bit(from_oblock(b), cache->discard_bitset);
+	set_bit(from_dblock(b), cache->discard_bitset);
 	spin_unlock_irqrestore(&cache->lock, flags);
 }
 
-static void clear_discard(struct cache *cache, dm_oblock_t b)
+static void clear_discard(struct cache *cache, dm_dblock_t b)
 {
 	unsigned long flags;
 
 	spin_lock_irqsave(&cache->lock, flags);
-	clear_bit(from_oblock(b), cache->discard_bitset);
+	clear_bit(from_dblock(b), cache->discard_bitset);
 	spin_unlock_irqrestore(&cache->lock, flags);
 }
 
-static bool is_discarded(struct cache *cache, dm_oblock_t b)
+static bool is_discarded(struct cache *cache, dm_dblock_t b)
 {
 	int r;
 	unsigned long flags;
 
 	spin_lock_irqsave(&cache->lock, flags);
-	r = test_bit(from_oblock(b), cache->discard_bitset);
+	r = test_bit(from_dblock(b), cache->discard_bitset);
 	spin_unlock_irqrestore(&cache->lock, flags);
 
 	return r;
@@ -562,7 +624,8 @@ static bool is_discarded_oblock(struct cache *cache, dm_oblock_t b)
 	unsigned long flags;
 
 	spin_lock_irqsave(&cache->lock, flags);
-	r = test_bit(from_oblock(b), cache->discard_bitset);
+	r = test_bit(from_dblock(oblock_to_dblock(cache, b)),
+		     cache->discard_bitset);
 	spin_unlock_irqrestore(&cache->lock, flags);
 
 	return r;
@@ -687,7 +750,7 @@ static void remap_to_origin_clear_discard(struct cache *cache, struct bio *bio,
 	check_if_tick_bio_needed(cache, bio);
 	remap_to_origin(cache, bio);
 	if (bio_data_dir(bio) == WRITE)
-		clear_discard(cache, oblock);
+		clear_discard(cache, oblock_to_dblock(cache, oblock));
 }
 
 static void remap_to_cache_dirty(struct cache *cache, struct bio *bio,
@@ -697,7 +760,7 @@ static void remap_to_cache_dirty(struct cache *cache, struct bio *bio,
 	remap_to_cache(cache, bio, cblock);
 	if (bio_data_dir(bio) == WRITE) {
 		set_dirty(cache, oblock, cblock);
-		clear_discard(cache, oblock);
+		clear_discard(cache, oblock_to_dblock(cache, oblock));
 	}
 }
 
@@ -817,24 +880,14 @@ static void remap_to_origin_then_cache(struct cache *cache, struct bio *bio,
  * Migration covers moving data from the origin device to the cache, or
  * vice versa.
  *--------------------------------------------------------------*/
-static void free_migration(struct dm_cache_migration *mg)
+static void inc_io_migrations(struct cache *cache)
 {
-	mempool_free(mg, mg->cache->migration_pool);
+	atomic_inc(&cache->nr_io_migrations);
 }
 
-static void inc_nr_migrations(struct cache *cache)
+static void dec_io_migrations(struct cache *cache)
 {
-	atomic_inc(&cache->nr_migrations);
-}
-
-static void dec_nr_migrations(struct cache *cache)
-{
-	atomic_dec(&cache->nr_migrations);
-
-	/*
-	 * Wake the worker in case we're suspending the target.
-	 */
-	wake_up(&cache->migration_wait);
+	atomic_dec(&cache->nr_io_migrations);
 }
 
 static void __cell_defer(struct cache *cache, struct dm_bio_prison_cell *cell,
@@ -857,11 +910,10 @@ static void cell_defer(struct cache *cache, struct dm_bio_prison_cell *cell,
 	wake_worker(cache);
 }
 
-static void cleanup_migration(struct dm_cache_migration *mg)
+static void free_io_migration(struct dm_cache_migration *mg)
 {
-	struct cache *cache = mg->cache;
+	dec_io_migrations(mg->cache);
 	free_migration(mg);
-	dec_nr_migrations(cache);
 }
 
 static void migration_failure(struct dm_cache_migration *mg)
@@ -886,7 +938,7 @@ static void migration_failure(struct dm_cache_migration *mg)
 		cell_defer(cache, mg->new_ocell, true);
 	}
 
-	cleanup_migration(mg);
+	free_io_migration(mg);
 }
 
 static void migration_success_pre_commit(struct dm_cache_migration *mg)
@@ -897,7 +949,7 @@ static void migration_success_pre_commit(struct dm_cache_migration *mg)
 	if (mg->writeback) {
 		clear_dirty(cache, mg->old_oblock, mg->cblock);
 		cell_defer(cache, mg->old_ocell, false);
-		cleanup_migration(mg);
+		free_io_migration(mg);
 		return;
 
 	} else if (mg->demote) {
@@ -907,14 +959,14 @@ static void migration_success_pre_commit(struct dm_cache_migration *mg)
 					     mg->old_oblock);
 			if (mg->promote)
 				cell_defer(cache, mg->new_ocell, true);
-			cleanup_migration(mg);
+			free_io_migration(mg);
 			return;
 		}
 	} else {
 		if (dm_cache_insert_mapping(cache->cmd, mg->cblock, mg->new_oblock)) {
 			DMWARN_LIMIT("promotion failed; couldn't update on disk metadata");
 			policy_remove_mapping(cache->policy, mg->new_oblock);
-			cleanup_migration(mg);
+			free_io_migration(mg);
 			return;
 		}
 	}
@@ -947,18 +999,22 @@ static void migration_success_post_commit(struct dm_cache_migration *mg)
 		} else {
 			if (mg->invalidate)
 				policy_remove_mapping(cache->policy, mg->old_oblock);
-			cleanup_migration(mg);
+			free_io_migration(mg);
 		}
 
 	} else {
-		clear_dirty(cache, mg->new_oblock, mg->cblock);
-		if (mg->requeue_holder)
+		if (mg->requeue_holder) {
+			clear_dirty(cache, mg->new_oblock, mg->cblock);
 			cell_defer(cache, mg->new_ocell, true);
-		else {
+		} else {
+			/*
+			 * The block was promoted via an overwrite, so it's dirty.
+			 */
+			set_dirty(cache, mg->new_oblock, mg->cblock);
 			bio_endio(mg->new_ocell->holder, 0);
 			cell_defer(cache, mg->new_ocell, false);
 		}
-		cleanup_migration(mg);
+		free_io_migration(mg);
 	}
 }
 
@@ -978,7 +1034,7 @@ static void copy_complete(int read_err, unsigned long write_err, void *context)
 	wake_worker(cache);
 }
 
-static void issue_copy_real(struct dm_cache_migration *mg)
+static void issue_copy(struct dm_cache_migration *mg)
 {
 	int r;
 	struct dm_io_region o_region, c_region;
@@ -1057,11 +1113,46 @@ static void avoid_copy(struct dm_cache_migration *mg)
 	migration_success_pre_commit(mg);
 }
 
-static void issue_copy(struct dm_cache_migration *mg)
+static void calc_discard_block_range(struct cache *cache, struct bio *bio,
+				     dm_dblock_t *b, dm_dblock_t *e)
+{
+	sector_t sb = bio->bi_iter.bi_sector;
+	sector_t se = bio_end_sector(bio);
+
+	*b = to_dblock(dm_sector_div_up(sb, cache->discard_block_size));
+
+	if (se - sb < cache->discard_block_size)
+		*e = *b;
+	else
+		*e = to_dblock(block_div(se, cache->discard_block_size));
+}
+
+static void issue_discard(struct dm_cache_migration *mg)
+{
+	dm_dblock_t b, e;
+	struct bio *bio = mg->new_ocell->holder;
+
+	calc_discard_block_range(mg->cache, bio, &b, &e);
+	while (b != e) {
+		set_discard(mg->cache, b);
+		b = to_dblock(from_dblock(b) + 1);
+	}
+
+	bio_endio(bio, 0);
+	cell_defer(mg->cache, mg->new_ocell, false);
+	free_migration(mg);
+}
+
+static void issue_copy_or_discard(struct dm_cache_migration *mg)
 {
 	bool avoid;
 	struct cache *cache = mg->cache;
 
+	if (mg->discard) {
+		issue_discard(mg);
+		return;
+	}
+
 	if (mg->writeback || mg->demote)
 		avoid = !is_dirty(cache, mg->cblock) ||
 			is_discarded_oblock(cache, mg->old_oblock);
@@ -1070,13 +1161,14 @@ static void issue_copy(struct dm_cache_migration *mg)
 
 		avoid = is_discarded_oblock(cache, mg->new_oblock);
 
-		if (!avoid && bio_writes_complete_block(cache, bio)) {
+		if (writeback_mode(&cache->features) &&
+		    !avoid && bio_writes_complete_block(cache, bio)) {
 			issue_overwrite(mg, bio);
 			return;
 		}
 	}
 
-	avoid ? avoid_copy(mg) : issue_copy_real(mg);
+	avoid ? avoid_copy(mg) : issue_copy(mg);
 }
 
 static void complete_migration(struct dm_cache_migration *mg)
@@ -1161,6 +1253,7 @@ static void promote(struct cache *cache, struct prealloc *structs,
 	struct dm_cache_migration *mg = prealloc_get_migration(structs);
 
 	mg->err = false;
+	mg->discard = false;
 	mg->writeback = false;
 	mg->demote = false;
 	mg->promote = true;
@@ -1173,7 +1266,7 @@ static void promote(struct cache *cache, struct prealloc *structs,
 	mg->new_ocell = cell;
 	mg->start_jiffies = jiffies;
 
-	inc_nr_migrations(cache);
+	inc_io_migrations(cache);
 	quiesce_migration(mg);
 }
 
@@ -1184,6 +1277,7 @@ static void writeback(struct cache *cache, struct prealloc *structs,
 	struct dm_cache_migration *mg = prealloc_get_migration(structs);
 
 	mg->err = false;
+	mg->discard = false;
 	mg->writeback = true;
 	mg->demote = false;
 	mg->promote = false;
@@ -1196,7 +1290,7 @@ static void writeback(struct cache *cache, struct prealloc *structs,
 	mg->new_ocell = NULL;
 	mg->start_jiffies = jiffies;
 
-	inc_nr_migrations(cache);
+	inc_io_migrations(cache);
 	quiesce_migration(mg);
 }
 
@@ -1209,6 +1303,7 @@ static void demote_then_promote(struct cache *cache, struct prealloc *structs,
 	struct dm_cache_migration *mg = prealloc_get_migration(structs);
 
 	mg->err = false;
+	mg->discard = false;
 	mg->writeback = false;
 	mg->demote = true;
 	mg->promote = true;
@@ -1222,7 +1317,7 @@ static void demote_then_promote(struct cache *cache, struct prealloc *structs,
 	mg->new_ocell = new_ocell;
 	mg->start_jiffies = jiffies;
 
-	inc_nr_migrations(cache);
+	inc_io_migrations(cache);
 	quiesce_migration(mg);
 }
 
@@ -1237,6 +1332,7 @@ static void invalidate(struct cache *cache, struct prealloc *structs,
 	struct dm_cache_migration *mg = prealloc_get_migration(structs);
 
 	mg->err = false;
+	mg->discard = false;
 	mg->writeback = false;
 	mg->demote = true;
 	mg->promote = false;
@@ -1249,7 +1345,27 @@ static void invalidate(struct cache *cache, struct prealloc *structs,
 	mg->new_ocell = NULL;
 	mg->start_jiffies = jiffies;
 
-	inc_nr_migrations(cache);
+	inc_io_migrations(cache);
+	quiesce_migration(mg);
+}
+
+static void discard(struct cache *cache, struct prealloc *structs,
+		    struct dm_bio_prison_cell *cell)
+{
+	struct dm_cache_migration *mg = prealloc_get_migration(structs);
+
+	mg->err = false;
+	mg->discard = true;
+	mg->writeback = false;
+	mg->demote = false;
+	mg->promote = false;
+	mg->requeue_holder = false;
+	mg->invalidate = false;
+	mg->cache = cache;
+	mg->old_ocell = NULL;
+	mg->new_ocell = cell;
+	mg->start_jiffies = jiffies;
+
 	quiesce_migration(mg);
 }
 
@@ -1286,36 +1402,32 @@ static void process_flush_bio(struct cache *cache, struct bio *bio)
 	issue(cache, bio);
 }
 
-/*
- * People generally discard large parts of a device, eg, the whole device
- * when formatting.  Splitting these large discards up into cache block
- * sized ios and then quiescing (always neccessary for discard) takes too
- * long.
- *
- * We keep it simple, and allow any size of discard to come in, and just
- * mark off blocks on the discard bitset.  No passdown occurs!
- *
- * To implement passdown we need to change the bio_prison such that a cell
- * can have a key that spans many blocks.
- */
-static void process_discard_bio(struct cache *cache, struct bio *bio)
+static void process_discard_bio(struct cache *cache, struct prealloc *structs,
+				struct bio *bio)
 {
-	dm_block_t start_block = dm_sector_div_up(bio->bi_iter.bi_sector,
-						  cache->sectors_per_block);
-	dm_block_t end_block = bio_end_sector(bio);
-	dm_block_t b;
+	int r;
+	dm_dblock_t b, e;
+	struct dm_bio_prison_cell *cell_prealloc, *new_ocell;
 
-	end_block = block_div(end_block, cache->sectors_per_block);
+	calc_discard_block_range(cache, bio, &b, &e);
+	if (b == e) {
+		bio_endio(bio, 0);
+		return;
+	}
 
-	for (b = start_block; b < end_block; b++)
-		set_discard(cache, to_oblock(b));
+	cell_prealloc = prealloc_get_cell(structs);
+	r = bio_detain_range(cache, dblock_to_oblock(cache, b), dblock_to_oblock(cache, e), bio, cell_prealloc,
+			     (cell_free_fn) prealloc_put_cell,
+			     structs, &new_ocell);
+	if (r > 0)
+		return;
 
-	bio_endio(bio, 0);
+	discard(cache, structs, new_ocell);
 }
 
 static bool spare_migration_bandwidth(struct cache *cache)
 {
-	sector_t current_volume = (atomic_read(&cache->nr_migrations) + 1) *
+	sector_t current_volume = (atomic_read(&cache->nr_io_migrations) + 1) *
 		cache->sectors_per_block;
 	return current_volume < cache->migration_threshold;
 }
@@ -1340,9 +1452,8 @@ static void process_bio(struct cache *cache, struct prealloc *structs,
 	dm_oblock_t block = get_bio_block(cache, bio);
 	struct dm_bio_prison_cell *cell_prealloc, *old_ocell, *new_ocell;
 	struct policy_result lookup_result;
-	bool discarded_block = is_discarded_oblock(cache, block);
 	bool passthrough = passthrough_mode(&cache->features);
-	bool can_migrate = !passthrough && (discarded_block || spare_migration_bandwidth(cache));
+	bool discarded_block, can_migrate;
 
 	/*
 	 * Check to see if that block is currently migrating.
@@ -1354,6 +1465,9 @@ static void process_bio(struct cache *cache, struct prealloc *structs,
 	if (r > 0)
 		return;
 
+	discarded_block = is_discarded_oblock(cache, block);
+	can_migrate = !passthrough && (discarded_block || spare_migration_bandwidth(cache));
+
 	r = policy_map(cache->policy, block, true, can_migrate, discarded_block,
 		       bio, &lookup_result);
 
@@ -1500,7 +1614,7 @@ static void process_deferred_bios(struct cache *cache)
 		if (bio->bi_rw & REQ_FLUSH)
 			process_flush_bio(cache, bio);
 		else if (bio->bi_rw & REQ_DISCARD)
-			process_discard_bio(cache, bio);
+			process_discard_bio(cache, &structs, bio);
 		else
 			process_bio(cache, &structs, bio);
 	}
@@ -1665,7 +1779,7 @@ static void stop_quiescing(struct cache *cache)
 
 static void wait_for_migrations(struct cache *cache)
 {
-	wait_event(cache->migration_wait, !atomic_read(&cache->nr_migrations));
+	wait_event(cache->migration_wait, !atomic_read(&cache->nr_allocated_migrations));
 }
 
 static void stop_worker(struct cache *cache)
@@ -1715,7 +1829,7 @@ static void do_worker(struct work_struct *ws)
 			process_invalidation_requests(cache);
 		}
 
-		process_migrations(cache, &cache->quiesced_migrations, issue_copy);
+		process_migrations(cache, &cache->quiesced_migrations, issue_copy_or_discard);
 		process_migrations(cache, &cache->completed_migrations, complete_migration);
 
 		if (commit_if_needed(cache)) {
@@ -1777,9 +1891,6 @@ static void destroy(struct cache *cache)
 {
 	unsigned i;
 
-	if (cache->next_migration)
-		mempool_free(cache->next_migration, cache->migration_pool);
-
 	if (cache->migration_pool)
 		mempool_destroy(cache->migration_pool);
 
@@ -2180,6 +2291,45 @@ static int create_cache_policy(struct cache *cache, struct cache_args *ca,
 	return 0;
 }
 
+/*
+ * We want the discard block size to be at least the size of the cache
+ * block size and have no more than 2^14 discard blocks across the origin.
+ */
+#define MAX_DISCARD_BLOCKS (1 << 14)
+
+static bool too_many_discard_blocks(sector_t discard_block_size,
+				    sector_t origin_size)
+{
+	(void) sector_div(origin_size, discard_block_size);
+
+	return origin_size > MAX_DISCARD_BLOCKS;
+}
+
+static sector_t calculate_discard_block_size(sector_t cache_block_size,
+					     sector_t origin_size)
+{
+	sector_t discard_block_size = cache_block_size;
+
+	if (origin_size)
+		while (too_many_discard_blocks(discard_block_size, origin_size))
+			discard_block_size *= 2;
+
+	return discard_block_size;
+}
+
+static void set_cache_size(struct cache *cache, dm_cblock_t size)
+{
+	dm_block_t nr_blocks = from_cblock(size);
+
+	if (nr_blocks > (1 << 20) && cache->cache_size != size)
+		DMWARN_LIMIT("You have created a cache device with a lot of individual cache blocks (%llu)\n"
+			     "All these mappings can consume a lot of kernel memory, and take some time to read/write.\n"
+			     "Please consider increasing the cache block size to reduce the overall cache block count.",
+			     (unsigned long long) nr_blocks);
+
+	cache->cache_size = size;
+}
+
 #define DEFAULT_MIGRATION_THRESHOLD 2048
 
 static int cache_create(struct cache_args *ca, struct cache **result)
@@ -2204,8 +2354,7 @@ static int cache_create(struct cache_args *ca, struct cache **result)
 	ti->num_discard_bios = 1;
 	ti->discards_supported = true;
 	ti->discard_zeroes_data_unsupported = true;
-	/* Discard bios must be split on a block boundary */
-	ti->split_discard_bios = true;
+	ti->split_discard_bios = false;
 
 	cache->features = ca->features;
 	ti->per_bio_data_size = get_per_bio_data_size(cache);
@@ -2235,10 +2384,10 @@ static int cache_create(struct cache_args *ca, struct cache **result)
 
 		cache->sectors_per_block_shift = -1;
 		cache_size = block_div(cache_size, ca->block_size);
-		cache->cache_size = to_cblock(cache_size);
+		set_cache_size(cache, to_cblock(cache_size));
 	} else {
 		cache->sectors_per_block_shift = __ffs(ca->block_size);
-		cache->cache_size = to_cblock(ca->cache_sectors >> cache->sectors_per_block_shift);
+		set_cache_size(cache, to_cblock(ca->cache_sectors >> cache->sectors_per_block_shift));
 	}
 
 	r = create_cache_policy(cache, ca, error);
@@ -2287,7 +2436,8 @@ static int cache_create(struct cache_args *ca, struct cache **result)
 	INIT_LIST_HEAD(&cache->quiesced_migrations);
 	INIT_LIST_HEAD(&cache->completed_migrations);
 	INIT_LIST_HEAD(&cache->need_commit_migrations);
-	atomic_set(&cache->nr_migrations, 0);
+	atomic_set(&cache->nr_allocated_migrations, 0);
+	atomic_set(&cache->nr_io_migrations, 0);
 	init_waitqueue_head(&cache->migration_wait);
 
 	init_waitqueue_head(&cache->quiescing_wait);
@@ -2303,13 +2453,17 @@ static int cache_create(struct cache_args *ca, struct cache **result)
 	}
 	clear_bitset(cache->dirty_bitset, from_cblock(cache->cache_size));
 
-	cache->discard_nr_blocks = cache->origin_blocks;
-	cache->discard_bitset = alloc_bitset(from_oblock(cache->discard_nr_blocks));
+	cache->discard_block_size =
+		calculate_discard_block_size(cache->sectors_per_block,
+					     cache->origin_sectors);
+	cache->discard_nr_blocks = to_dblock(dm_sector_div_up(cache->origin_sectors,
+							      cache->discard_block_size));
+	cache->discard_bitset = alloc_bitset(from_dblock(cache->discard_nr_blocks));
 	if (!cache->discard_bitset) {
 		*error = "could not allocate discard bitset";
 		goto bad;
 	}
-	clear_bitset(cache->discard_bitset, from_oblock(cache->discard_nr_blocks));
+	clear_bitset(cache->discard_bitset, from_dblock(cache->discard_nr_blocks));
 
 	cache->copier = dm_kcopyd_client_create(&dm_kcopyd_throttle);
 	if (IS_ERR(cache->copier)) {
@@ -2327,7 +2481,7 @@ static int cache_create(struct cache_args *ca, struct cache **result)
 	INIT_DELAYED_WORK(&cache->waker, do_waker);
 	cache->last_commit_jiffies = jiffies;
 
-	cache->prison = dm_bio_prison_create(PRISON_CELLS);
+	cache->prison = dm_bio_prison_create();
 	if (!cache->prison) {
 		*error = "could not create bio prison";
 		goto bad;
@@ -2346,8 +2500,6 @@ static int cache_create(struct cache_args *ca, struct cache **result)
 		goto bad;
 	}
 
-	cache->next_migration = NULL;
-
 	cache->need_tick_bio = true;
 	cache->sized = false;
 	cache->invalidate = false;
@@ -2549,11 +2701,11 @@ static int __cache_map(struct cache *cache, struct bio *bio, struct dm_bio_priso
 static int cache_map(struct dm_target *ti, struct bio *bio)
 {
 	int r;
-	struct dm_bio_prison_cell *cell;
+	struct dm_bio_prison_cell *cell = NULL;
 	struct cache *cache = ti->private;
 
 	r = __cache_map(cache, bio, &cell);
-	if (r == DM_MAPIO_REMAPPED) {
+	if (r == DM_MAPIO_REMAPPED && cell) {
 		inc_ds(cache, bio, cell);
 		cell_defer(cache, cell, false);
 	}
@@ -2599,16 +2751,16 @@ static int write_discard_bitset(struct cache *cache)
 {
 	unsigned i, r;
 
-	r = dm_cache_discard_bitset_resize(cache->cmd, cache->sectors_per_block,
-					   cache->origin_blocks);
+	r = dm_cache_discard_bitset_resize(cache->cmd, cache->discard_block_size,
+					   cache->discard_nr_blocks);
 	if (r) {
 		DMERR("could not resize on-disk discard bitset");
 		return r;
 	}
 
-	for (i = 0; i < from_oblock(cache->discard_nr_blocks); i++) {
-		r = dm_cache_set_discard(cache->cmd, to_oblock(i),
-					 is_discarded(cache, to_oblock(i)));
+	for (i = 0; i < from_dblock(cache->discard_nr_blocks); i++) {
+		r = dm_cache_set_discard(cache->cmd, to_dblock(i),
+					 is_discarded(cache, to_dblock(i)));
 		if (r)
 			return r;
 	}
@@ -2680,15 +2832,86 @@ static int load_mapping(void *context, dm_oblock_t oblock, dm_cblock_t cblock,
 	return 0;
 }
 
+/*
+ * The discard block size in the on disk metadata is not
+ * neccessarily the same as we're currently using.  So we have to
+ * be careful to only set the discarded attribute if we know it
+ * covers a complete block of the new size.
+ */
+struct discard_load_info {
+	struct cache *cache;
+
+	/*
+	 * These blocks are sized using the on disk dblock size, rather
+	 * than the current one.
+	 */
+	dm_block_t block_size;
+	dm_block_t discard_begin, discard_end;
+};
+
+static void discard_load_info_init(struct cache *cache,
+				   struct discard_load_info *li)
+{
+	li->cache = cache;
+	li->discard_begin = li->discard_end = 0;
+}
+
+static void set_discard_range(struct discard_load_info *li)
+{
+	sector_t b, e;
+
+	if (li->discard_begin == li->discard_end)
+		return;
+
+	/*
+	 * Convert to sectors.
+	 */
+	b = li->discard_begin * li->block_size;
+	e = li->discard_end * li->block_size;
+
+	/*
+	 * Then convert back to the current dblock size.
+	 */
+	b = dm_sector_div_up(b, li->cache->discard_block_size);
+	sector_div(e, li->cache->discard_block_size);
+
+	/*
+	 * The origin may have shrunk, so we need to check we're still in
+	 * bounds.
+	 */
+	if (e > from_dblock(li->cache->discard_nr_blocks))
+		e = from_dblock(li->cache->discard_nr_blocks);
+
+	for (; b < e; b++)
+		set_discard(li->cache, to_dblock(b));
+}
+
 static int load_discard(void *context, sector_t discard_block_size,
-			dm_oblock_t oblock, bool discard)
+			dm_dblock_t dblock, bool discard)
 {
-	struct cache *cache = context;
+	struct discard_load_info *li = context;
 
-	if (discard)
-		set_discard(cache, oblock);
-	else
-		clear_discard(cache, oblock);
+	li->block_size = discard_block_size;
+
+	if (discard) {
+		if (from_dblock(dblock) == li->discard_end)
+			/*
+			 * We're already in a discard range, just extend it.
+			 */
+			li->discard_end = li->discard_end + 1ULL;
+
+		else {
+			/*
+			 * Emit the old range and start a new one.
+			 */
+			set_discard_range(li);
+			li->discard_begin = from_dblock(dblock);
+			li->discard_end = li->discard_begin + 1ULL;
+		}
+	} else {
+		set_discard_range(li);
+		li->discard_begin = li->discard_end = 0;
+	}
 
 	return 0;
 }
@@ -2730,7 +2953,7 @@ static int resize_cache_dev(struct cache *cache, dm_cblock_t new_size)
 		return r;
 	}
 
-	cache->cache_size = new_size;
+	set_cache_size(cache, new_size);
 
 	return 0;
 }
@@ -2772,11 +2995,22 @@ static int cache_preresume(struct dm_target *ti)
 	}
 
 	if (!cache->loaded_discards) {
-		r = dm_cache_load_discards(cache->cmd, load_discard, cache);
+		struct discard_load_info li;
+
+		/*
+		 * The discard bitset could have been resized, or the
+		 * discard block size changed.  To be safe we start by
+		 * setting every dblock to not discarded.
+		 */
+		clear_bitset(cache->discard_bitset, from_dblock(cache->discard_nr_blocks));
+
+		discard_load_info_init(cache, &li);
+		r = dm_cache_load_discards(cache->cmd, load_discard, &li);
 		if (r) {
 			DMERR("could not load origin discards");
 			return r;
 		}
+		set_discard_range(&li);
 
 		cache->loaded_discards = true;
 	}
@@ -3079,8 +3313,9 @@ static void set_discard_limits(struct cache *cache, struct queue_limits *limits)
 	/*
 	 * FIXME: these limits may be incompatible with the cache device
 	 */
-	limits->max_discard_sectors = cache->sectors_per_block;
-	limits->discard_granularity = cache->sectors_per_block << SECTOR_SHIFT;
+	limits->max_discard_sectors = min_t(sector_t, cache->discard_block_size * 1024,
+					    cache->origin_sectors);
+	limits->discard_granularity = cache->discard_block_size << SECTOR_SHIFT;
 }
 
 static void cache_io_hints(struct dm_target *ti, struct queue_limits *limits)
@@ -3104,7 +3339,7 @@ static void cache_io_hints(struct dm_target *ti, struct queue_limits *limits)
 
 static struct target_type cache_target = {
 	.name = "cache",
-	.version = {1, 5, 0},
+	.version = {1, 6, 0},
 	.module = THIS_MODULE,
 	.ctr = cache_ctr,
 	.dtr = cache_dtr,
diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c
index cd15e0801228..08981be7baa1 100644
--- a/drivers/md/dm-crypt.c
+++ b/drivers/md/dm-crypt.c
@@ -526,29 +526,26 @@ static int crypt_iv_lmk_one(struct crypt_config *cc, u8 *iv,
 			    u8 *data)
 {
 	struct iv_lmk_private *lmk = &cc->iv_gen_private.lmk;
-	struct {
-		struct shash_desc desc;
-		char ctx[crypto_shash_descsize(lmk->hash_tfm)];
-	} sdesc;
+	SHASH_DESC_ON_STACK(desc, lmk->hash_tfm);
 	struct md5_state md5state;
 	__le32 buf[4];
 	int i, r;
 
-	sdesc.desc.tfm = lmk->hash_tfm;
-	sdesc.desc.flags = CRYPTO_TFM_REQ_MAY_SLEEP;
+	desc->tfm = lmk->hash_tfm;
+	desc->flags = CRYPTO_TFM_REQ_MAY_SLEEP;
 
-	r = crypto_shash_init(&sdesc.desc);
+	r = crypto_shash_init(desc);
 	if (r)
 		return r;
 
 	if (lmk->seed) {
-		r = crypto_shash_update(&sdesc.desc, lmk->seed, LMK_SEED_SIZE);
+		r = crypto_shash_update(desc, lmk->seed, LMK_SEED_SIZE);
 		if (r)
 			return r;
 	}
 
 	/* Sector is always 512B, block size 16, add data of blocks 1-31 */
-	r = crypto_shash_update(&sdesc.desc, data + 16, 16 * 31);
+	r = crypto_shash_update(desc, data + 16, 16 * 31);
 	if (r)
 		return r;
 
@@ -557,12 +554,12 @@ static int crypt_iv_lmk_one(struct crypt_config *cc, u8 *iv,
 	buf[1] = cpu_to_le32((((u64)dmreq->iv_sector >> 32) & 0x00FFFFFF) | 0x80000000);
 	buf[2] = cpu_to_le32(4024);
 	buf[3] = 0;
-	r = crypto_shash_update(&sdesc.desc, (u8 *)buf, sizeof(buf));
+	r = crypto_shash_update(desc, (u8 *)buf, sizeof(buf));
 	if (r)
 		return r;
 
 	/* No MD5 padding here */
-	r = crypto_shash_export(&sdesc.desc, &md5state);
+	r = crypto_shash_export(desc, &md5state);
 	if (r)
 		return r;
 
@@ -679,10 +676,7 @@ static int crypt_iv_tcw_whitening(struct crypt_config *cc,
 	struct iv_tcw_private *tcw = &cc->iv_gen_private.tcw;
 	u64 sector = cpu_to_le64((u64)dmreq->iv_sector);
 	u8 buf[TCW_WHITENING_SIZE];
-	struct {
-		struct shash_desc desc;
-		char ctx[crypto_shash_descsize(tcw->crc32_tfm)];
-	} sdesc;
+	SHASH_DESC_ON_STACK(desc, tcw->crc32_tfm);
 	int i, r;
 
 	/* xor whitening with sector number */
@@ -691,16 +685,16 @@ static int crypt_iv_tcw_whitening(struct crypt_config *cc,
 	crypto_xor(&buf[8], (u8 *)&sector, 8);
 
 	/* calculate crc32 for every 32bit part and xor it */
-	sdesc.desc.tfm = tcw->crc32_tfm;
-	sdesc.desc.flags = CRYPTO_TFM_REQ_MAY_SLEEP;
+	desc->tfm = tcw->crc32_tfm;
+	desc->flags = CRYPTO_TFM_REQ_MAY_SLEEP;
 	for (i = 0; i < 4; i++) {
-		r = crypto_shash_init(&sdesc.desc);
+		r = crypto_shash_init(desc);
 		if (r)
 			goto out;
-		r = crypto_shash_update(&sdesc.desc, &buf[i * 4], 4);
+		r = crypto_shash_update(desc, &buf[i * 4], 4);
 		if (r)
 			goto out;
-		r = crypto_shash_final(&sdesc.desc, &buf[i * 4]);
+		r = crypto_shash_final(desc, &buf[i * 4]);
 		if (r)
 			goto out;
 	}
@@ -711,7 +705,7 @@ static int crypt_iv_tcw_whitening(struct crypt_config *cc,
 	for (i = 0; i < ((1 << SECTOR_SHIFT) / 8); i++)
 		crypto_xor(data + i * 8, buf, 8);
 out:
-	memset(buf, 0, sizeof(buf));
+	memzero_explicit(buf, sizeof(buf));
 	return r;
 }
 
diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c
index 51521429fb59..73f791bb9ea4 100644
--- a/drivers/md/dm-ioctl.c
+++ b/drivers/md/dm-ioctl.c
@@ -684,11 +684,14 @@ static void __dev_status(struct mapped_device *md, struct dm_ioctl *param)
 	int srcu_idx;
 
 	param->flags &= ~(DM_SUSPEND_FLAG | DM_READONLY_FLAG |
-			  DM_ACTIVE_PRESENT_FLAG);
+			  DM_ACTIVE_PRESENT_FLAG | DM_INTERNAL_SUSPEND_FLAG);
 
 	if (dm_suspended_md(md))
 		param->flags |= DM_SUSPEND_FLAG;
 
+	if (dm_suspended_internally_md(md))
+		param->flags |= DM_INTERNAL_SUSPEND_FLAG;
+
 	if (dm_test_deferred_remove_flag(md))
 		param->flags |= DM_DEFERRED_REMOVE;
 
@@ -1418,7 +1421,7 @@ static void retrieve_deps(struct dm_table *table,
 	deps->count = count;
 	count = 0;
 	list_for_each_entry (dd, dm_table_get_devices(table), list)
-		deps->dev[count++] = huge_encode_dev(dd->dm_dev.bdev->bd_dev);
+		deps->dev[count++] = huge_encode_dev(dd->dm_dev->bdev->bd_dev);
 
 	param->data_size = param->data_start + needed;
 }
diff --git a/drivers/md/dm-log-userspace-transfer.c b/drivers/md/dm-log-userspace-transfer.c
index b428c0ae63d5..39ad9664d397 100644
--- a/drivers/md/dm-log-userspace-transfer.c
+++ b/drivers/md/dm-log-userspace-transfer.c
@@ -272,7 +272,7 @@ int dm_ulog_tfr_init(void)
 
 	r = cn_add_callback(&ulog_cn_id, "dmlogusr", cn_ulog_callback);
 	if (r) {
-		cn_del_callback(&ulog_cn_id);
+		kfree(prealloced_cn_msg);
 		return r;
 	}
 
diff --git a/drivers/md/dm-mpath.c b/drivers/md/dm-mpath.c
index 833d7e752f06..7b6b0f0f831a 100644
--- a/drivers/md/dm-mpath.c
+++ b/drivers/md/dm-mpath.c
@@ -317,8 +317,10 @@ static void __choose_pgpath(struct multipath *m, size_t nr_bytes)
 	struct priority_group *pg;
 	unsigned bypassed = 1;
 
-	if (!m->nr_valid_paths)
+	if (!m->nr_valid_paths) {
+		m->queue_io = 0;
 		goto failed;
+	}
 
 	/* Were we instructed to switch PG? */
 	if (m->next_pg) {
diff --git a/drivers/md/dm-raid.c b/drivers/md/dm-raid.c
index 4880b69e2e9e..07c0fa0fa284 100644
--- a/drivers/md/dm-raid.c
+++ b/drivers/md/dm-raid.c
@@ -1,6 +1,6 @@
 /*
  * Copyright (C) 2010-2011 Neil Brown
- * Copyright (C) 2010-2011 Red Hat, Inc. All rights reserved.
+ * Copyright (C) 2010-2014 Red Hat, Inc. All rights reserved.
  *
  * This file is released under the GPL.
  */
@@ -18,6 +18,8 @@
 
 #define DM_MSG_PREFIX "raid"
 
+static bool devices_handle_discard_safely = false;
+
 /*
  * The following flags are used by dm-raid.c to set up the array state.
  * They must be cleared before md_run is called.
@@ -475,6 +477,8 @@ too_many:
  *                                      will form the "stripe"
  *    [[no]sync]			Force or prevent recovery of the
  *                                      entire array
+ *    [devices_handle_discard_safely]	Allow discards on RAID4/5/6; useful if RAID
+ *					member device(s) properly support TRIM/UNMAP
  *    [rebuild <idx>]			Rebuild the drive indicated by the index
  *    [daemon_sleep <ms>]		Time between bitmap daemon work to
  *                                      clear bits
@@ -785,8 +789,7 @@ struct dm_raid_superblock {
 	__le32 layout;
 	__le32 stripe_sectors;
 
-	__u8 pad[452];		/* Round struct to 512 bytes. */
-				/* Always set to 0 when writing. */
+	/* Remainder of a logical block is zero-filled when writing (see super_sync()). */
 } __packed;
 
 static int read_disk_sb(struct md_rdev *rdev, int size)
@@ -823,7 +826,7 @@ static void super_sync(struct mddev *mddev, struct md_rdev *rdev)
 		    test_bit(Faulty, &(rs->dev[i].rdev.flags)))
 			failed_devices |= (1ULL << i);
 
-	memset(sb, 0, sizeof(*sb));
+	memset(sb + 1, 0, rdev->sb_size - sizeof(*sb));
 
 	sb->magic = cpu_to_le32(DM_RAID_MAGIC);
 	sb->features = cpu_to_le32(0);	/* No features yet */
@@ -858,7 +861,11 @@ static int super_load(struct md_rdev *rdev, struct md_rdev *refdev)
 	uint64_t events_sb, events_refsb;
 
 	rdev->sb_start = 0;
-	rdev->sb_size = sizeof(*sb);
+	rdev->sb_size = bdev_logical_block_size(rdev->meta_bdev);
+	if (rdev->sb_size < sizeof(*sb) || rdev->sb_size > PAGE_SIZE) {
+		DMERR("superblock size of a logical block is no longer valid");
+		return -EINVAL;
+	}
 
 	ret = read_disk_sb(rdev, rdev->sb_size);
 	if (ret)
@@ -1150,6 +1157,53 @@ static int analyse_superblocks(struct dm_target *ti, struct raid_set *rs)
 }
 
 /*
+ * Enable/disable discard support on RAID set depending on
+ * RAID level and discard properties of underlying RAID members.
+ */
+static void configure_discard_support(struct dm_target *ti, struct raid_set *rs)
+{
+	int i;
+	bool raid456;
+
+	/* Assume discards not supported until after checks below. */
+	ti->discards_supported = false;
+
+	/* RAID level 4,5,6 require discard_zeroes_data for data integrity! */
+	raid456 = (rs->md.level == 4 || rs->md.level == 5 || rs->md.level == 6);
+
+	for (i = 0; i < rs->md.raid_disks; i++) {
+		struct request_queue *q;
+
+		if (!rs->dev[i].rdev.bdev)
+			continue;
+
+		q = bdev_get_queue(rs->dev[i].rdev.bdev);
+		if (!q || !blk_queue_discard(q))
+			return;
+
+		if (raid456) {
+			if (!q->limits.discard_zeroes_data)
+				return;
+			if (!devices_handle_discard_safely) {
+				DMERR("raid456 discard support disabled due to discard_zeroes_data uncertainty.");
+				DMERR("Set dm-raid.devices_handle_discard_safely=Y to override.");
+				return;
+			}
+		}
+	}
+
+	/* All RAID members properly support discards */
+	ti->discards_supported = true;
+
+	/*
+	 * RAID1 and RAID10 personalities require bio splitting,
+	 * RAID0/4/5/6 don't and process large discard bios properly.
+	 */
+	ti->split_discard_bios = !!(rs->md.level == 1 || rs->md.level == 10);
+	ti->num_discard_bios = 1;
+}
+
+/*
  * Construct a RAID4/5/6 mapping:
  * Args:
  *	<raid_type> <#raid_params> <raid_params>		\
@@ -1231,6 +1285,11 @@ static int raid_ctr(struct dm_target *ti, unsigned argc, char **argv)
 	ti->private = rs;
 	ti->num_flush_bios = 1;
 
+	/*
+	 * Disable/enable discard support on RAID set.
+	 */
+	configure_discard_support(ti, rs);
+
 	mutex_lock(&rs->md.reconfig_mutex);
 	ret = md_run(&rs->md);
 	rs->md.in_sync = 0; /* Assume already marked dirty */
@@ -1652,7 +1711,7 @@ static void raid_resume(struct dm_target *ti)
 
 static struct target_type raid_target = {
 	.name = "raid",
-	.version = {1, 5, 2},
+	.version = {1, 6, 0},
 	.module = THIS_MODULE,
 	.ctr = raid_ctr,
 	.dtr = raid_dtr,
@@ -1683,6 +1742,10 @@ static void __exit dm_raid_exit(void)
 module_init(dm_raid_init);
 module_exit(dm_raid_exit);
 
+module_param(devices_handle_discard_safely, bool, 0644);
+MODULE_PARM_DESC(devices_handle_discard_safely,
+		 "Set to Y if all devices in each array reliably return zeroes on reads from discarded regions");
+
 MODULE_DESCRIPTION(DM_NAME " raid4/5/6 target");
 MODULE_ALIAS("dm-raid1");
 MODULE_ALIAS("dm-raid10");
diff --git a/drivers/md/dm-stats.c b/drivers/md/dm-stats.c
index 28a90122a5a8..f478a4c96d2f 100644
--- a/drivers/md/dm-stats.c
+++ b/drivers/md/dm-stats.c
@@ -548,7 +548,7 @@ void dm_stats_account_io(struct dm_stats *stats, unsigned long bi_rw,
 		 * A race condition can at worst result in the merged flag being
 		 * misrepresented, so we don't have to disable preemption here.
 		 */
-		last = __this_cpu_ptr(stats->last);
+		last = raw_cpu_ptr(stats->last);
 		stats_aux->merged =
 			(bi_sector == (ACCESS_ONCE(last->last_sector) &&
 				       ((bi_rw & (REQ_WRITE | REQ_DISCARD)) ==
@@ -824,7 +824,7 @@ static int message_stats_create(struct mapped_device *md,
 		return 1;
 
 	id = dm_stats_create(dm_get_stats(md), start, end, step, program_id, aux_data,
-			     dm_internal_suspend, dm_internal_resume, md);
+			     dm_internal_suspend_fast, dm_internal_resume_fast, md);
 	if (id < 0)
 		return id;
 
diff --git a/drivers/md/dm-stripe.c b/drivers/md/dm-stripe.c
index d1600d2aa2e2..f8b37d4c05d8 100644
--- a/drivers/md/dm-stripe.c
+++ b/drivers/md/dm-stripe.c
@@ -159,8 +159,10 @@ static int stripe_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 		sc->stripes_shift = __ffs(stripes);
 
 	r = dm_set_target_max_io_len(ti, chunk_size);
-	if (r)
+	if (r) {
+		kfree(sc);
 		return r;
+	}
 
 	ti->num_flush_bios = stripes;
 	ti->num_discard_bios = stripes;
diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c
index f9c6cb8dbcf8..3afae9e062f8 100644
--- a/drivers/md/dm-table.c
+++ b/drivers/md/dm-table.c
@@ -210,15 +210,16 @@ int dm_table_create(struct dm_table **result, fmode_t mode,
 	return 0;
 }
 
-static void free_devices(struct list_head *devices)
+static void free_devices(struct list_head *devices, struct mapped_device *md)
 {
 	struct list_head *tmp, *next;
 
 	list_for_each_safe(tmp, next, devices) {
 		struct dm_dev_internal *dd =
 		    list_entry(tmp, struct dm_dev_internal, list);
-		DMWARN("dm_table_destroy: dm_put_device call missing for %s",
-		       dd->dm_dev.name);
+		DMWARN("%s: dm_table_destroy: dm_put_device call missing for %s",
+		       dm_device_name(md), dd->dm_dev->name);
+		dm_put_table_device(md, dd->dm_dev);
 		kfree(dd);
 	}
 }
@@ -247,7 +248,7 @@ void dm_table_destroy(struct dm_table *t)
 	vfree(t->highs);
 
 	/* free the device list */
-	free_devices(&t->devices);
+	free_devices(&t->devices, t->md);
 
 	dm_free_md_mempools(t->mempools);
 
@@ -262,53 +263,13 @@ static struct dm_dev_internal *find_device(struct list_head *l, dev_t dev)
 	struct dm_dev_internal *dd;
 
 	list_for_each_entry (dd, l, list)
-		if (dd->dm_dev.bdev->bd_dev == dev)
+		if (dd->dm_dev->bdev->bd_dev == dev)
 			return dd;
 
 	return NULL;
 }
 
 /*
- * Open a device so we can use it as a map destination.
- */
-static int open_dev(struct dm_dev_internal *d, dev_t dev,
-		    struct mapped_device *md)
-{
-	static char *_claim_ptr = "I belong to device-mapper";
-	struct block_device *bdev;
-
-	int r;
-
-	BUG_ON(d->dm_dev.bdev);
-
-	bdev = blkdev_get_by_dev(dev, d->dm_dev.mode | FMODE_EXCL, _claim_ptr);
-	if (IS_ERR(bdev))
-		return PTR_ERR(bdev);
-
-	r = bd_link_disk_holder(bdev, dm_disk(md));
-	if (r) {
-		blkdev_put(bdev, d->dm_dev.mode | FMODE_EXCL);
-		return r;
-	}
-
-	d->dm_dev.bdev = bdev;
-	return 0;
-}
-
-/*
- * Close a device that we've been using.
- */
-static void close_dev(struct dm_dev_internal *d, struct mapped_device *md)
-{
-	if (!d->dm_dev.bdev)
-		return;
-
-	bd_unlink_disk_holder(d->dm_dev.bdev, dm_disk(md));
-	blkdev_put(d->dm_dev.bdev, d->dm_dev.mode | FMODE_EXCL);
-	d->dm_dev.bdev = NULL;
-}
-
-/*
  * If possible, this checks an area of a destination device is invalid.
  */
 static int device_area_is_invalid(struct dm_target *ti, struct dm_dev *dev,
@@ -386,19 +347,17 @@ static int upgrade_mode(struct dm_dev_internal *dd, fmode_t new_mode,
 			struct mapped_device *md)
 {
 	int r;
-	struct dm_dev_internal dd_new, dd_old;
-
-	dd_new = dd_old = *dd;
+	struct dm_dev *old_dev, *new_dev;
 
-	dd_new.dm_dev.mode |= new_mode;
-	dd_new.dm_dev.bdev = NULL;
+	old_dev = dd->dm_dev;
 
-	r = open_dev(&dd_new, dd->dm_dev.bdev->bd_dev, md);
+	r = dm_get_table_device(md, dd->dm_dev->bdev->bd_dev,
+				dd->dm_dev->mode | new_mode, &new_dev);
 	if (r)
 		return r;
 
-	dd->dm_dev.mode |= new_mode;
-	close_dev(&dd_old, md);
+	dd->dm_dev = new_dev;
+	dm_put_table_device(md, old_dev);
 
 	return 0;
 }
@@ -440,27 +399,22 @@ int dm_get_device(struct dm_target *ti, const char *path, fmode_t mode,
 		if (!dd)
 			return -ENOMEM;
 
-		dd->dm_dev.mode = mode;
-		dd->dm_dev.bdev = NULL;
-
-		if ((r = open_dev(dd, dev, t->md))) {
+		if ((r = dm_get_table_device(t->md, dev, mode, &dd->dm_dev))) {
 			kfree(dd);
 			return r;
 		}
 
-		format_dev_t(dd->dm_dev.name, dev);
-
 		atomic_set(&dd->count, 0);
 		list_add(&dd->list, &t->devices);
 
-	} else if (dd->dm_dev.mode != (mode | dd->dm_dev.mode)) {
+	} else if (dd->dm_dev->mode != (mode | dd->dm_dev->mode)) {
 		r = upgrade_mode(dd, mode, t->md);
 		if (r)
 			return r;
 	}
 	atomic_inc(&dd->count);
 
-	*result = &dd->dm_dev;
+	*result = dd->dm_dev;
 	return 0;
 }
 EXPORT_SYMBOL(dm_get_device);
@@ -505,11 +459,23 @@ static int dm_set_device_limits(struct dm_target *ti, struct dm_dev *dev,
  */
 void dm_put_device(struct dm_target *ti, struct dm_dev *d)
 {
-	struct dm_dev_internal *dd = container_of(d, struct dm_dev_internal,
-						  dm_dev);
+	int found = 0;
+	struct list_head *devices = &ti->table->devices;
+	struct dm_dev_internal *dd;
 
+	list_for_each_entry(dd, devices, list) {
+		if (dd->dm_dev == d) {
+			found = 1;
+			break;
+		}
+	}
+	if (!found) {
+		DMWARN("%s: device %s not in table devices list",
+		       dm_device_name(ti->table->md), d->name);
+		return;
+	}
 	if (atomic_dec_and_test(&dd->count)) {
-		close_dev(dd, ti->table->md);
+		dm_put_table_device(ti->table->md, d);
 		list_del(&dd->list);
 		kfree(dd);
 	}
@@ -906,7 +872,7 @@ static int dm_table_set_type(struct dm_table *t)
 	/* Non-request-stackable devices can't be used for request-based dm */
 	devices = dm_table_get_devices(t);
 	list_for_each_entry(dd, devices, list) {
-		if (!blk_queue_stackable(bdev_get_queue(dd->dm_dev.bdev))) {
+		if (!blk_queue_stackable(bdev_get_queue(dd->dm_dev->bdev))) {
 			DMWARN("table load rejected: including"
 			       " non-request-stackable devices");
 			return -EINVAL;
@@ -1043,7 +1009,7 @@ static struct gendisk * dm_table_get_integrity_disk(struct dm_table *t,
 	struct gendisk *prev_disk = NULL, *template_disk = NULL;
 
 	list_for_each_entry(dd, devices, list) {
-		template_disk = dd->dm_dev.bdev->bd_disk;
+		template_disk = dd->dm_dev->bdev->bd_disk;
 		if (!blk_get_integrity(template_disk))
 			goto no_integrity;
 		if (!match_all && !blk_integrity_is_initialized(template_disk))
@@ -1555,18 +1521,32 @@ fmode_t dm_table_get_mode(struct dm_table *t)
 }
 EXPORT_SYMBOL(dm_table_get_mode);
 
-static void suspend_targets(struct dm_table *t, unsigned postsuspend)
+enum suspend_mode {
+	PRESUSPEND,
+	PRESUSPEND_UNDO,
+	POSTSUSPEND,
+};
+
+static void suspend_targets(struct dm_table *t, enum suspend_mode mode)
 {
 	int i = t->num_targets;
 	struct dm_target *ti = t->targets;
 
 	while (i--) {
-		if (postsuspend) {
+		switch (mode) {
+		case PRESUSPEND:
+			if (ti->type->presuspend)
+				ti->type->presuspend(ti);
+			break;
+		case PRESUSPEND_UNDO:
+			if (ti->type->presuspend_undo)
+				ti->type->presuspend_undo(ti);
+			break;
+		case POSTSUSPEND:
 			if (ti->type->postsuspend)
 				ti->type->postsuspend(ti);
-		} else if (ti->type->presuspend)
-			ti->type->presuspend(ti);
-
+			break;
+		}
 		ti++;
 	}
 }
@@ -1576,7 +1556,15 @@ void dm_table_presuspend_targets(struct dm_table *t)
 	if (!t)
 		return;
 
-	suspend_targets(t, 0);
+	suspend_targets(t, PRESUSPEND);
+}
+
+void dm_table_presuspend_undo_targets(struct dm_table *t)
+{
+	if (!t)
+		return;
+
+	suspend_targets(t, PRESUSPEND_UNDO);
 }
 
 void dm_table_postsuspend_targets(struct dm_table *t)
@@ -1584,7 +1572,7 @@ void dm_table_postsuspend_targets(struct dm_table *t)
 	if (!t)
 		return;
 
-	suspend_targets(t, 1);
+	suspend_targets(t, POSTSUSPEND);
 }
 
 int dm_table_resume_targets(struct dm_table *t)
@@ -1629,7 +1617,7 @@ int dm_table_any_congested(struct dm_table *t, int bdi_bits)
 	int r = 0;
 
 	list_for_each_entry(dd, devices, list) {
-		struct request_queue *q = bdev_get_queue(dd->dm_dev.bdev);
+		struct request_queue *q = bdev_get_queue(dd->dm_dev->bdev);
 		char b[BDEVNAME_SIZE];
 
 		if (likely(q))
@@ -1637,7 +1625,7 @@ int dm_table_any_congested(struct dm_table *t, int bdi_bits)
 		else
 			DMWARN_LIMIT("%s: any_congested: nonexistent device %s",
 				     dm_device_name(t->md),
-				     bdevname(dd->dm_dev.bdev, b));
+				     bdevname(dd->dm_dev->bdev, b));
 	}
 
 	list_for_each_entry(cb, &t->target_callbacks, list)
diff --git a/drivers/md/dm-thin-metadata.c b/drivers/md/dm-thin-metadata.c
index e9d33ad59df5..43adbb863f5a 100644
--- a/drivers/md/dm-thin-metadata.c
+++ b/drivers/md/dm-thin-metadata.c
@@ -1384,42 +1384,38 @@ static bool __snapshotted_since(struct dm_thin_device *td, uint32_t time)
 }
 
 int dm_thin_find_block(struct dm_thin_device *td, dm_block_t block,
-		       int can_block, struct dm_thin_lookup_result *result)
+		       int can_issue_io, struct dm_thin_lookup_result *result)
 {
-	int r = -EINVAL;
-	uint64_t block_time = 0;
+	int r;
 	__le64 value;
 	struct dm_pool_metadata *pmd = td->pmd;
 	dm_block_t keys[2] = { td->id, block };
 	struct dm_btree_info *info;
 
-	if (can_block) {
-		down_read(&pmd->root_lock);
-		info = &pmd->info;
-	} else if (down_read_trylock(&pmd->root_lock))
-		info = &pmd->nb_info;
-	else
-		return -EWOULDBLOCK;
-
 	if (pmd->fail_io)
-		goto out;
+		return -EINVAL;
 
-	r = dm_btree_lookup(info, pmd->root, keys, &value);
-	if (!r)
-		block_time = le64_to_cpu(value);
+	down_read(&pmd->root_lock);
 
-out:
-	up_read(&pmd->root_lock);
+	if (can_issue_io) {
+		info = &pmd->info;
+	} else
+		info = &pmd->nb_info;
 
+	r = dm_btree_lookup(info, pmd->root, keys, &value);
 	if (!r) {
+		uint64_t block_time = 0;
 		dm_block_t exception_block;
 		uint32_t exception_time;
+
+		block_time = le64_to_cpu(value);
 		unpack_block_time(block_time, &exception_block,
 				  &exception_time);
 		result->block = exception_block;
 		result->shared = __snapshotted_since(td, exception_time);
 	}
 
+	up_read(&pmd->root_lock);
 	return r;
 }
 
@@ -1813,3 +1809,8 @@ bool dm_pool_metadata_needs_check(struct dm_pool_metadata *pmd)
 
 	return needs_check;
 }
+
+void dm_pool_issue_prefetches(struct dm_pool_metadata *pmd)
+{
+	dm_tm_issue_prefetches(pmd->tm);
+}
diff --git a/drivers/md/dm-thin-metadata.h b/drivers/md/dm-thin-metadata.h
index e3c857db195a..921d15ee56a0 100644
--- a/drivers/md/dm-thin-metadata.h
+++ b/drivers/md/dm-thin-metadata.h
@@ -139,12 +139,12 @@ struct dm_thin_lookup_result {
 
 /*
  * Returns:
- *   -EWOULDBLOCK iff @can_block is set and would block.
+ *   -EWOULDBLOCK iff @can_issue_io is set and would issue IO
  *   -ENODATA iff that mapping is not present.
  *   0 success
  */
 int dm_thin_find_block(struct dm_thin_device *td, dm_block_t block,
-		       int can_block, struct dm_thin_lookup_result *result);
+		       int can_issue_io, struct dm_thin_lookup_result *result);
 
 /*
  * Obtain an unused block.
@@ -213,6 +213,11 @@ int dm_pool_register_metadata_threshold(struct dm_pool_metadata *pmd,
 int dm_pool_metadata_set_needs_check(struct dm_pool_metadata *pmd);
 bool dm_pool_metadata_needs_check(struct dm_pool_metadata *pmd);
 
+/*
+ * Issue any prefetches that may be useful.
+ */
+void dm_pool_issue_prefetches(struct dm_pool_metadata *pmd);
+
 /*----------------------------------------------------------------*/
 
 #endif
diff --git a/drivers/md/dm-thin.c b/drivers/md/dm-thin.c
index 4843801173fe..07705ee181e3 100644
--- a/drivers/md/dm-thin.c
+++ b/drivers/md/dm-thin.c
@@ -11,11 +11,13 @@
 #include <linux/device-mapper.h>
 #include <linux/dm-io.h>
 #include <linux/dm-kcopyd.h>
+#include <linux/log2.h>
 #include <linux/list.h>
 #include <linux/rculist.h>
 #include <linux/init.h>
 #include <linux/module.h>
 #include <linux/slab.h>
+#include <linux/sort.h>
 #include <linux/rbtree.h>
 
 #define	DM_MSG_PREFIX	"thin"
@@ -25,7 +27,6 @@
  */
 #define ENDIO_HOOK_POOL_SIZE 1024
 #define MAPPING_POOL_SIZE 1024
-#define PRISON_CELLS 1024
 #define COMMIT_PERIOD HZ
 #define NO_SPACE_TIMEOUT_SECS 60
 
@@ -114,7 +115,8 @@ static void build_data_key(struct dm_thin_device *td,
 {
 	key->virtual = 0;
 	key->dev = dm_thin_dev_id(td);
-	key->block = b;
+	key->block_begin = b;
+	key->block_end = b + 1ULL;
 }
 
 static void build_virtual_key(struct dm_thin_device *td, dm_block_t b,
@@ -122,7 +124,55 @@ static void build_virtual_key(struct dm_thin_device *td, dm_block_t b,
 {
 	key->virtual = 1;
 	key->dev = dm_thin_dev_id(td);
-	key->block = b;
+	key->block_begin = b;
+	key->block_end = b + 1ULL;
+}
+
+/*----------------------------------------------------------------*/
+
+#define THROTTLE_THRESHOLD (1 * HZ)
+
+struct throttle {
+	struct rw_semaphore lock;
+	unsigned long threshold;
+	bool throttle_applied;
+};
+
+static void throttle_init(struct throttle *t)
+{
+	init_rwsem(&t->lock);
+	t->throttle_applied = false;
+}
+
+static void throttle_work_start(struct throttle *t)
+{
+	t->threshold = jiffies + THROTTLE_THRESHOLD;
+}
+
+static void throttle_work_update(struct throttle *t)
+{
+	if (!t->throttle_applied && jiffies > t->threshold) {
+		down_write(&t->lock);
+		t->throttle_applied = true;
+	}
+}
+
+static void throttle_work_complete(struct throttle *t)
+{
+	if (t->throttle_applied) {
+		t->throttle_applied = false;
+		up_write(&t->lock);
+	}
+}
+
+static void throttle_lock(struct throttle *t)
+{
+	down_read(&t->lock);
+}
+
+static void throttle_unlock(struct throttle *t)
+{
+	up_read(&t->lock);
 }
 
 /*----------------------------------------------------------------*/
@@ -155,8 +205,11 @@ struct pool_features {
 
 struct thin_c;
 typedef void (*process_bio_fn)(struct thin_c *tc, struct bio *bio);
+typedef void (*process_cell_fn)(struct thin_c *tc, struct dm_bio_prison_cell *cell);
 typedef void (*process_mapping_fn)(struct dm_thin_new_mapping *m);
 
+#define CELL_SORT_ARRAY_SIZE 8192
+
 struct pool {
 	struct list_head list;
 	struct dm_target *ti;	/* Only set if a pool target is bound */
@@ -171,11 +224,13 @@ struct pool {
 
 	struct pool_features pf;
 	bool low_water_triggered:1;	/* A dm event has been sent */
+	bool suspended:1;
 
 	struct dm_bio_prison *prison;
 	struct dm_kcopyd_client *copier;
 
 	struct workqueue_struct *wq;
+	struct throttle throttle;
 	struct work_struct worker;
 	struct delayed_work waker;
 	struct delayed_work no_space_timeout;
@@ -198,8 +253,13 @@ struct pool {
 	process_bio_fn process_bio;
 	process_bio_fn process_discard;
 
+	process_cell_fn process_cell;
+	process_cell_fn process_discard_cell;
+
 	process_mapping_fn process_prepared_mapping;
 	process_mapping_fn process_prepared_discard;
+
+	struct dm_bio_prison_cell *cell_sort_array[CELL_SORT_ARRAY_SIZE];
 };
 
 static enum pool_mode get_pool_mode(struct pool *pool);
@@ -232,8 +292,11 @@ struct thin_c {
 
 	struct pool *pool;
 	struct dm_thin_device *td;
+	struct mapped_device *thin_md;
+
 	bool requeue_mode:1;
 	spinlock_t lock;
+	struct list_head deferred_cells;
 	struct bio_list deferred_bio_list;
 	struct bio_list retry_on_resume_list;
 	struct rb_root sort_bio_list; /* sorted list of deferred bios */
@@ -290,6 +353,15 @@ static void cell_release(struct pool *pool,
 	dm_bio_prison_free_cell(pool->prison, cell);
 }
 
+static void cell_visit_release(struct pool *pool,
+			       void (*fn)(void *, struct dm_bio_prison_cell *),
+			       void *context,
+			       struct dm_bio_prison_cell *cell)
+{
+	dm_cell_visit_release(pool->prison, fn, context, cell);
+	dm_bio_prison_free_cell(pool->prison, cell);
+}
+
 static void cell_release_no_holder(struct pool *pool,
 				   struct dm_bio_prison_cell *cell,
 				   struct bio_list *bios)
@@ -298,19 +370,6 @@ static void cell_release_no_holder(struct pool *pool,
 	dm_bio_prison_free_cell(pool->prison, cell);
 }
 
-static void cell_defer_no_holder_no_free(struct thin_c *tc,
-					 struct dm_bio_prison_cell *cell)
-{
-	struct pool *pool = tc->pool;
-	unsigned long flags;
-
-	spin_lock_irqsave(&tc->lock, flags);
-	dm_cell_release_no_holder(pool->prison, cell, &tc->deferred_bio_list);
-	spin_unlock_irqrestore(&tc->lock, flags);
-
-	wake_worker(pool);
-}
-
 static void cell_error_with_code(struct pool *pool,
 				 struct dm_bio_prison_cell *cell, int error_code)
 {
@@ -323,6 +382,16 @@ static void cell_error(struct pool *pool, struct dm_bio_prison_cell *cell)
 	cell_error_with_code(pool, cell, -EIO);
 }
 
+static void cell_success(struct pool *pool, struct dm_bio_prison_cell *cell)
+{
+	cell_error_with_code(pool, cell, 0);
+}
+
+static void cell_requeue(struct pool *pool, struct dm_bio_prison_cell *cell)
+{
+	cell_error_with_code(pool, cell, DM_ENDIO_REQUEUE);
+}
+
 /*----------------------------------------------------------------*/
 
 /*
@@ -393,44 +462,65 @@ struct dm_thin_endio_hook {
 	struct rb_node rb_node;
 };
 
-static void requeue_bio_list(struct thin_c *tc, struct bio_list *master)
+static void __merge_bio_list(struct bio_list *bios, struct bio_list *master)
+{
+	bio_list_merge(bios, master);
+	bio_list_init(master);
+}
+
+static void error_bio_list(struct bio_list *bios, int error)
 {
 	struct bio *bio;
+
+	while ((bio = bio_list_pop(bios)))
+		bio_endio(bio, error);
+}
+
+static void error_thin_bio_list(struct thin_c *tc, struct bio_list *master, int error)
+{
 	struct bio_list bios;
 	unsigned long flags;
 
 	bio_list_init(&bios);
 
 	spin_lock_irqsave(&tc->lock, flags);
-	bio_list_merge(&bios, master);
-	bio_list_init(master);
+	__merge_bio_list(&bios, master);
 	spin_unlock_irqrestore(&tc->lock, flags);
 
-	while ((bio = bio_list_pop(&bios)))
-		bio_endio(bio, DM_ENDIO_REQUEUE);
+	error_bio_list(&bios, error);
 }
 
-static void requeue_io(struct thin_c *tc)
+static void requeue_deferred_cells(struct thin_c *tc)
 {
-	requeue_bio_list(tc, &tc->deferred_bio_list);
-	requeue_bio_list(tc, &tc->retry_on_resume_list);
+	struct pool *pool = tc->pool;
+	unsigned long flags;
+	struct list_head cells;
+	struct dm_bio_prison_cell *cell, *tmp;
+
+	INIT_LIST_HEAD(&cells);
+
+	spin_lock_irqsave(&tc->lock, flags);
+	list_splice_init(&tc->deferred_cells, &cells);
+	spin_unlock_irqrestore(&tc->lock, flags);
+
+	list_for_each_entry_safe(cell, tmp, &cells, user_list)
+		cell_requeue(pool, cell);
 }
 
-static void error_thin_retry_list(struct thin_c *tc)
+static void requeue_io(struct thin_c *tc)
 {
-	struct bio *bio;
-	unsigned long flags;
 	struct bio_list bios;
+	unsigned long flags;
 
 	bio_list_init(&bios);
 
 	spin_lock_irqsave(&tc->lock, flags);
-	bio_list_merge(&bios, &tc->retry_on_resume_list);
-	bio_list_init(&tc->retry_on_resume_list);
+	__merge_bio_list(&bios, &tc->deferred_bio_list);
+	__merge_bio_list(&bios, &tc->retry_on_resume_list);
 	spin_unlock_irqrestore(&tc->lock, flags);
 
-	while ((bio = bio_list_pop(&bios)))
-		bio_io_error(bio);
+	error_bio_list(&bios, DM_ENDIO_REQUEUE);
+	requeue_deferred_cells(tc);
 }
 
 static void error_retry_list(struct pool *pool)
@@ -439,7 +529,7 @@ static void error_retry_list(struct pool *pool)
 
 	rcu_read_lock();
 	list_for_each_entry_rcu(tc, &pool->active_thins, list)
-		error_thin_retry_list(tc);
+		error_thin_bio_list(tc, &tc->retry_on_resume_list, -EIO);
 	rcu_read_unlock();
 }
 
@@ -629,33 +719,75 @@ static void overwrite_endio(struct bio *bio, int err)
  */
 
 /*
- * This sends the bios in the cell back to the deferred_bios list.
+ * This sends the bios in the cell, except the original holder, back
+ * to the deferred_bios list.
  */
-static void cell_defer(struct thin_c *tc, struct dm_bio_prison_cell *cell)
+static void cell_defer_no_holder(struct thin_c *tc, struct dm_bio_prison_cell *cell)
 {
 	struct pool *pool = tc->pool;
 	unsigned long flags;
 
 	spin_lock_irqsave(&tc->lock, flags);
-	cell_release(pool, cell, &tc->deferred_bio_list);
+	cell_release_no_holder(pool, cell, &tc->deferred_bio_list);
 	spin_unlock_irqrestore(&tc->lock, flags);
 
 	wake_worker(pool);
 }
 
-/*
- * Same as cell_defer above, except it omits the original holder of the cell.
- */
-static void cell_defer_no_holder(struct thin_c *tc, struct dm_bio_prison_cell *cell)
+static void thin_defer_bio(struct thin_c *tc, struct bio *bio);
+
+struct remap_info {
+	struct thin_c *tc;
+	struct bio_list defer_bios;
+	struct bio_list issue_bios;
+};
+
+static void __inc_remap_and_issue_cell(void *context,
+				       struct dm_bio_prison_cell *cell)
 {
-	struct pool *pool = tc->pool;
-	unsigned long flags;
+	struct remap_info *info = context;
+	struct bio *bio;
 
-	spin_lock_irqsave(&tc->lock, flags);
-	cell_release_no_holder(pool, cell, &tc->deferred_bio_list);
-	spin_unlock_irqrestore(&tc->lock, flags);
+	while ((bio = bio_list_pop(&cell->bios))) {
+		if (bio->bi_rw & (REQ_DISCARD | REQ_FLUSH | REQ_FUA))
+			bio_list_add(&info->defer_bios, bio);
+		else {
+			inc_all_io_entry(info->tc->pool, bio);
 
-	wake_worker(pool);
+			/*
+			 * We can't issue the bios with the bio prison lock
+			 * held, so we add them to a list to issue on
+			 * return from this function.
+			 */
+			bio_list_add(&info->issue_bios, bio);
+		}
+	}
+}
+
+static void inc_remap_and_issue_cell(struct thin_c *tc,
+				     struct dm_bio_prison_cell *cell,
+				     dm_block_t block)
+{
+	struct bio *bio;
+	struct remap_info info;
+
+	info.tc = tc;
+	bio_list_init(&info.defer_bios);
+	bio_list_init(&info.issue_bios);
+
+	/*
+	 * We have to be careful to inc any bios we're about to issue
+	 * before the cell is released, and avoid a race with new bios
+	 * being added to the cell.
+	 */
+	cell_visit_release(tc->pool, __inc_remap_and_issue_cell,
+			   &info, cell);
+
+	while ((bio = bio_list_pop(&info.defer_bios)))
+		thin_defer_bio(tc, bio);
+
+	while ((bio = bio_list_pop(&info.issue_bios)))
+		remap_and_issue(info.tc, bio, block);
 }
 
 static void process_prepared_mapping_fail(struct dm_thin_new_mapping *m)
@@ -706,10 +838,13 @@ static void process_prepared_mapping(struct dm_thin_new_mapping *m)
 	 * the bios in the cell.
 	 */
 	if (bio) {
-		cell_defer_no_holder(tc, m->cell);
+		inc_remap_and_issue_cell(tc, m->cell, m->data_block);
 		bio_endio(bio, 0);
-	} else
-		cell_defer(tc, m->cell);
+	} else {
+		inc_all_io_entry(tc->pool, m->cell->holder);
+		remap_and_issue(tc, m->cell->holder, m->data_block);
+		inc_remap_and_issue_cell(tc, m->cell, m->data_block);
+	}
 
 out:
 	list_del(&m->list);
@@ -842,6 +977,20 @@ static void ll_zero(struct thin_c *tc, struct dm_thin_new_mapping *m,
 	}
 }
 
+static void remap_and_issue_overwrite(struct thin_c *tc, struct bio *bio,
+				      dm_block_t data_block,
+				      struct dm_thin_new_mapping *m)
+{
+	struct pool *pool = tc->pool;
+	struct dm_thin_endio_hook *h = dm_per_bio_data(bio, sizeof(struct dm_thin_endio_hook));
+
+	h->overwrite_mapping = m;
+	m->bio = bio;
+	save_and_set_endio(bio, &m->saved_bi_end_io, overwrite_endio);
+	inc_all_io_entry(pool, bio);
+	remap_and_issue(tc, bio, data_block);
+}
+
 /*
  * A partial copy also needs to zero the uncopied region.
  */
@@ -876,15 +1025,9 @@ static void schedule_copy(struct thin_c *tc, dm_block_t virt_block,
 	 * If the whole block of data is being overwritten, we can issue the
 	 * bio immediately. Otherwise we use kcopyd to clone the data first.
 	 */
-	if (io_overwrites_block(pool, bio)) {
-		struct dm_thin_endio_hook *h = dm_per_bio_data(bio, sizeof(struct dm_thin_endio_hook));
-
-		h->overwrite_mapping = m;
-		m->bio = bio;
-		save_and_set_endio(bio, &m->saved_bi_end_io, overwrite_endio);
-		inc_all_io_entry(pool, bio);
-		remap_and_issue(tc, bio, data_dest);
-	} else {
+	if (io_overwrites_block(pool, bio))
+		remap_and_issue_overwrite(tc, bio, data_dest, m);
+	else {
 		struct dm_io_region from, to;
 
 		from.bdev = origin->bdev;
@@ -953,16 +1096,10 @@ static void schedule_zero(struct thin_c *tc, dm_block_t virt_block,
 	if (!pool->pf.zero_new_blocks)
 		process_prepared_mapping(m);
 
-	else if (io_overwrites_block(pool, bio)) {
-		struct dm_thin_endio_hook *h = dm_per_bio_data(bio, sizeof(struct dm_thin_endio_hook));
-
-		h->overwrite_mapping = m;
-		m->bio = bio;
-		save_and_set_endio(bio, &m->saved_bi_end_io, overwrite_endio);
-		inc_all_io_entry(pool, bio);
-		remap_and_issue(tc, bio, data_block);
+	else if (io_overwrites_block(pool, bio))
+		remap_and_issue_overwrite(tc, bio, data_block, m);
 
-	} else
+	else
 		ll_zero(tc, m,
 			data_block * pool->sectors_per_block,
 			(data_block + 1) * pool->sectors_per_block);
@@ -990,6 +1127,24 @@ static void schedule_external_copy(struct thin_c *tc, dm_block_t virt_block,
 		schedule_zero(tc, virt_block, data_dest, cell, bio);
 }
 
+static void set_pool_mode(struct pool *pool, enum pool_mode new_mode);
+
+static void check_for_space(struct pool *pool)
+{
+	int r;
+	dm_block_t nr_free;
+
+	if (get_pool_mode(pool) != PM_OUT_OF_DATA_SPACE)
+		return;
+
+	r = dm_pool_get_free_block_count(pool->pmd, &nr_free);
+	if (r)
+		return;
+
+	if (nr_free)
+		set_pool_mode(pool, PM_WRITE);
+}
+
 /*
  * A non-zero return indicates read_only or fail_io mode.
  * Many callers don't care about the return value.
@@ -1004,6 +1159,8 @@ static int commit(struct pool *pool)
 	r = dm_pool_commit_metadata(pool->pmd);
 	if (r)
 		metadata_operation_failed(pool, "dm_pool_commit_metadata", r);
+	else
+		check_for_space(pool);
 
 	return r;
 }
@@ -1022,8 +1179,6 @@ static void check_low_water_mark(struct pool *pool, dm_block_t free_blocks)
 	}
 }
 
-static void set_pool_mode(struct pool *pool, enum pool_mode new_mode);
-
 static int alloc_data_block(struct thin_c *tc, dm_block_t *result)
 {
 	int r;
@@ -1134,29 +1289,25 @@ static void retry_bios_on_resume(struct pool *pool, struct dm_bio_prison_cell *c
 	bio_list_init(&bios);
 	cell_release(pool, cell, &bios);
 
-	error = should_error_unserviceable_bio(pool);
-	if (error)
-		while ((bio = bio_list_pop(&bios)))
-			bio_endio(bio, error);
-	else
-		while ((bio = bio_list_pop(&bios)))
-			retry_on_resume(bio);
+	while ((bio = bio_list_pop(&bios)))
+		retry_on_resume(bio);
 }
 
-static void process_discard(struct thin_c *tc, struct bio *bio)
+static void process_discard_cell(struct thin_c *tc, struct dm_bio_prison_cell *cell)
 {
 	int r;
-	unsigned long flags;
+	struct bio *bio = cell->holder;
 	struct pool *pool = tc->pool;
-	struct dm_bio_prison_cell *cell, *cell2;
-	struct dm_cell_key key, key2;
+	struct dm_bio_prison_cell *cell2;
+	struct dm_cell_key key2;
 	dm_block_t block = get_bio_block(tc, bio);
 	struct dm_thin_lookup_result lookup_result;
 	struct dm_thin_new_mapping *m;
 
-	build_virtual_key(tc->td, block, &key);
-	if (bio_detain(tc->pool, &key, bio, &cell))
+	if (tc->requeue_mode) {
+		cell_requeue(pool, cell);
 		return;
+	}
 
 	r = dm_thin_find_block(tc->td, block, 1, &lookup_result);
 	switch (r) {
@@ -1187,12 +1338,9 @@ static void process_discard(struct thin_c *tc, struct bio *bio)
 			m->cell2 = cell2;
 			m->bio = bio;
 
-			if (!dm_deferred_set_add_work(pool->all_io_ds, &m->list)) {
-				spin_lock_irqsave(&pool->lock, flags);
-				list_add_tail(&m->list, &pool->prepared_discards);
-				spin_unlock_irqrestore(&pool->lock, flags);
-				wake_worker(pool);
-			}
+			if (!dm_deferred_set_add_work(pool->all_io_ds, &m->list))
+				pool->process_prepared_discard(m);
+
 		} else {
 			inc_all_io_entry(pool, bio);
 			cell_defer_no_holder(tc, cell);
@@ -1227,6 +1375,19 @@ static void process_discard(struct thin_c *tc, struct bio *bio)
 	}
 }
 
+static void process_discard_bio(struct thin_c *tc, struct bio *bio)
+{
+	struct dm_bio_prison_cell *cell;
+	struct dm_cell_key key;
+	dm_block_t block = get_bio_block(tc, bio);
+
+	build_virtual_key(tc->td, block, &key);
+	if (bio_detain(tc->pool, &key, bio, &cell))
+		return;
+
+	process_discard_cell(tc, cell);
+}
+
 static void break_sharing(struct thin_c *tc, struct bio *bio, dm_block_t block,
 			  struct dm_cell_key *key,
 			  struct dm_thin_lookup_result *lookup_result,
@@ -1255,11 +1416,53 @@ static void break_sharing(struct thin_c *tc, struct bio *bio, dm_block_t block,
 	}
 }
 
+static void __remap_and_issue_shared_cell(void *context,
+					  struct dm_bio_prison_cell *cell)
+{
+	struct remap_info *info = context;
+	struct bio *bio;
+
+	while ((bio = bio_list_pop(&cell->bios))) {
+		if ((bio_data_dir(bio) == WRITE) ||
+		    (bio->bi_rw & (REQ_DISCARD | REQ_FLUSH | REQ_FUA)))
+			bio_list_add(&info->defer_bios, bio);
+		else {
+			struct dm_thin_endio_hook *h = dm_per_bio_data(bio, sizeof(struct dm_thin_endio_hook));;
+
+			h->shared_read_entry = dm_deferred_entry_inc(info->tc->pool->shared_read_ds);
+			inc_all_io_entry(info->tc->pool, bio);
+			bio_list_add(&info->issue_bios, bio);
+		}
+	}
+}
+
+static void remap_and_issue_shared_cell(struct thin_c *tc,
+					struct dm_bio_prison_cell *cell,
+					dm_block_t block)
+{
+	struct bio *bio;
+	struct remap_info info;
+
+	info.tc = tc;
+	bio_list_init(&info.defer_bios);
+	bio_list_init(&info.issue_bios);
+
+	cell_visit_release(tc->pool, __remap_and_issue_shared_cell,
+			   &info, cell);
+
+	while ((bio = bio_list_pop(&info.defer_bios)))
+		thin_defer_bio(tc, bio);
+
+	while ((bio = bio_list_pop(&info.issue_bios)))
+		remap_and_issue(tc, bio, block);
+}
+
 static void process_shared_bio(struct thin_c *tc, struct bio *bio,
 			       dm_block_t block,
-			       struct dm_thin_lookup_result *lookup_result)
+			       struct dm_thin_lookup_result *lookup_result,
+			       struct dm_bio_prison_cell *virt_cell)
 {
-	struct dm_bio_prison_cell *cell;
+	struct dm_bio_prison_cell *data_cell;
 	struct pool *pool = tc->pool;
 	struct dm_cell_key key;
 
@@ -1268,19 +1471,23 @@ static void process_shared_bio(struct thin_c *tc, struct bio *bio,
 	 * of being broken so we have nothing further to do here.
 	 */
 	build_data_key(tc->td, lookup_result->block, &key);
-	if (bio_detain(pool, &key, bio, &cell))
+	if (bio_detain(pool, &key, bio, &data_cell)) {
+		cell_defer_no_holder(tc, virt_cell);
 		return;
+	}
 
-	if (bio_data_dir(bio) == WRITE && bio->bi_iter.bi_size)
-		break_sharing(tc, bio, block, &key, lookup_result, cell);
-	else {
+	if (bio_data_dir(bio) == WRITE && bio->bi_iter.bi_size) {
+		break_sharing(tc, bio, block, &key, lookup_result, data_cell);
+		cell_defer_no_holder(tc, virt_cell);
+	} else {
 		struct dm_thin_endio_hook *h = dm_per_bio_data(bio, sizeof(struct dm_thin_endio_hook));
 
 		h->shared_read_entry = dm_deferred_entry_inc(pool->shared_read_ds);
 		inc_all_io_entry(pool, bio);
-		cell_defer_no_holder(tc, cell);
-
 		remap_and_issue(tc, bio, lookup_result->block);
+
+		remap_and_issue_shared_cell(tc, data_cell, lookup_result->block);
+		remap_and_issue_shared_cell(tc, virt_cell, lookup_result->block);
 	}
 }
 
@@ -1333,34 +1540,28 @@ static void provision_block(struct thin_c *tc, struct bio *bio, dm_block_t block
 	}
 }
 
-static void process_bio(struct thin_c *tc, struct bio *bio)
+static void process_cell(struct thin_c *tc, struct dm_bio_prison_cell *cell)
 {
 	int r;
 	struct pool *pool = tc->pool;
+	struct bio *bio = cell->holder;
 	dm_block_t block = get_bio_block(tc, bio);
-	struct dm_bio_prison_cell *cell;
-	struct dm_cell_key key;
 	struct dm_thin_lookup_result lookup_result;
 
-	/*
-	 * If cell is already occupied, then the block is already
-	 * being provisioned so we have nothing further to do here.
-	 */
-	build_virtual_key(tc->td, block, &key);
-	if (bio_detain(pool, &key, bio, &cell))
+	if (tc->requeue_mode) {
+		cell_requeue(pool, cell);
 		return;
+	}
 
 	r = dm_thin_find_block(tc->td, block, 1, &lookup_result);
 	switch (r) {
 	case 0:
-		if (lookup_result.shared) {
-			process_shared_bio(tc, bio, block, &lookup_result);
-			cell_defer_no_holder(tc, cell); /* FIXME: pass this cell into process_shared? */
-		} else {
+		if (lookup_result.shared)
+			process_shared_bio(tc, bio, block, &lookup_result, cell);
+		else {
 			inc_all_io_entry(pool, bio);
-			cell_defer_no_holder(tc, cell);
-
 			remap_and_issue(tc, bio, lookup_result.block);
+			inc_remap_and_issue_cell(tc, cell, lookup_result.block);
 		}
 		break;
 
@@ -1394,7 +1595,26 @@ static void process_bio(struct thin_c *tc, struct bio *bio)
 	}
 }
 
-static void process_bio_read_only(struct thin_c *tc, struct bio *bio)
+static void process_bio(struct thin_c *tc, struct bio *bio)
+{
+	struct pool *pool = tc->pool;
+	dm_block_t block = get_bio_block(tc, bio);
+	struct dm_bio_prison_cell *cell;
+	struct dm_cell_key key;
+
+	/*
+	 * If cell is already occupied, then the block is already
+	 * being provisioned so we have nothing further to do here.
+	 */
+	build_virtual_key(tc->td, block, &key);
+	if (bio_detain(pool, &key, bio, &cell))
+		return;
+
+	process_cell(tc, cell);
+}
+
+static void __process_bio_read_only(struct thin_c *tc, struct bio *bio,
+				    struct dm_bio_prison_cell *cell)
 {
 	int r;
 	int rw = bio_data_dir(bio);
@@ -1404,15 +1624,21 @@ static void process_bio_read_only(struct thin_c *tc, struct bio *bio)
 	r = dm_thin_find_block(tc->td, block, 1, &lookup_result);
 	switch (r) {
 	case 0:
-		if (lookup_result.shared && (rw == WRITE) && bio->bi_iter.bi_size)
+		if (lookup_result.shared && (rw == WRITE) && bio->bi_iter.bi_size) {
 			handle_unserviceable_bio(tc->pool, bio);
-		else {
+			if (cell)
+				cell_defer_no_holder(tc, cell);
+		} else {
 			inc_all_io_entry(tc->pool, bio);
 			remap_and_issue(tc, bio, lookup_result.block);
+			if (cell)
+				inc_remap_and_issue_cell(tc, cell, lookup_result.block);
 		}
 		break;
 
 	case -ENODATA:
+		if (cell)
+			cell_defer_no_holder(tc, cell);
 		if (rw != READ) {
 			handle_unserviceable_bio(tc->pool, bio);
 			break;
@@ -1431,11 +1657,23 @@ static void process_bio_read_only(struct thin_c *tc, struct bio *bio)
 	default:
 		DMERR_LIMIT("%s: dm_thin_find_block() failed: error = %d",
 			    __func__, r);
+		if (cell)
+			cell_defer_no_holder(tc, cell);
 		bio_io_error(bio);
 		break;
 	}
 }
 
+static void process_bio_read_only(struct thin_c *tc, struct bio *bio)
+{
+	__process_bio_read_only(tc, bio, NULL);
+}
+
+static void process_cell_read_only(struct thin_c *tc, struct dm_bio_prison_cell *cell)
+{
+	__process_bio_read_only(tc, cell->holder, cell);
+}
+
 static void process_bio_success(struct thin_c *tc, struct bio *bio)
 {
 	bio_endio(bio, 0);
@@ -1446,6 +1684,16 @@ static void process_bio_fail(struct thin_c *tc, struct bio *bio)
 	bio_io_error(bio);
 }
 
+static void process_cell_success(struct thin_c *tc, struct dm_bio_prison_cell *cell)
+{
+	cell_success(tc->pool, cell);
+}
+
+static void process_cell_fail(struct thin_c *tc, struct dm_bio_prison_cell *cell)
+{
+	cell_error(tc->pool, cell);
+}
+
 /*
  * FIXME: should we also commit due to size of transaction, measured in
  * metadata blocks?
@@ -1527,9 +1775,10 @@ static void process_thin_deferred_bios(struct thin_c *tc)
 	struct bio *bio;
 	struct bio_list bios;
 	struct blk_plug plug;
+	unsigned count = 0;
 
 	if (tc->requeue_mode) {
-		requeue_bio_list(tc, &tc->deferred_bio_list);
+		error_thin_bio_list(tc, &tc->deferred_bio_list, DM_ENDIO_REQUEUE);
 		return;
 	}
 
@@ -1568,10 +1817,97 @@ static void process_thin_deferred_bios(struct thin_c *tc)
 			pool->process_discard(tc, bio);
 		else
 			pool->process_bio(tc, bio);
+
+		if ((count++ & 127) == 0) {
+			throttle_work_update(&pool->throttle);
+			dm_pool_issue_prefetches(pool->pmd);
+		}
 	}
 	blk_finish_plug(&plug);
 }
 
+static int cmp_cells(const void *lhs, const void *rhs)
+{
+	struct dm_bio_prison_cell *lhs_cell = *((struct dm_bio_prison_cell **) lhs);
+	struct dm_bio_prison_cell *rhs_cell = *((struct dm_bio_prison_cell **) rhs);
+
+	BUG_ON(!lhs_cell->holder);
+	BUG_ON(!rhs_cell->holder);
+
+	if (lhs_cell->holder->bi_iter.bi_sector < rhs_cell->holder->bi_iter.bi_sector)
+		return -1;
+
+	if (lhs_cell->holder->bi_iter.bi_sector > rhs_cell->holder->bi_iter.bi_sector)
+		return 1;
+
+	return 0;
+}
+
+static unsigned sort_cells(struct pool *pool, struct list_head *cells)
+{
+	unsigned count = 0;
+	struct dm_bio_prison_cell *cell, *tmp;
+
+	list_for_each_entry_safe(cell, tmp, cells, user_list) {
+		if (count >= CELL_SORT_ARRAY_SIZE)
+			break;
+
+		pool->cell_sort_array[count++] = cell;
+		list_del(&cell->user_list);
+	}
+
+	sort(pool->cell_sort_array, count, sizeof(cell), cmp_cells, NULL);
+
+	return count;
+}
+
+static void process_thin_deferred_cells(struct thin_c *tc)
+{
+	struct pool *pool = tc->pool;
+	unsigned long flags;
+	struct list_head cells;
+	struct dm_bio_prison_cell *cell;
+	unsigned i, j, count;
+
+	INIT_LIST_HEAD(&cells);
+
+	spin_lock_irqsave(&tc->lock, flags);
+	list_splice_init(&tc->deferred_cells, &cells);
+	spin_unlock_irqrestore(&tc->lock, flags);
+
+	if (list_empty(&cells))
+		return;
+
+	do {
+		count = sort_cells(tc->pool, &cells);
+
+		for (i = 0; i < count; i++) {
+			cell = pool->cell_sort_array[i];
+			BUG_ON(!cell->holder);
+
+			/*
+			 * If we've got no free new_mapping structs, and processing
+			 * this bio might require one, we pause until there are some
+			 * prepared mappings to process.
+			 */
+			if (ensure_next_mapping(pool)) {
+				for (j = i; j < count; j++)
+					list_add(&pool->cell_sort_array[j]->user_list, &cells);
+
+				spin_lock_irqsave(&tc->lock, flags);
+				list_splice(&cells, &tc->deferred_cells);
+				spin_unlock_irqrestore(&tc->lock, flags);
+				return;
+			}
+
+			if (cell->holder->bi_rw & REQ_DISCARD)
+				pool->process_discard_cell(tc, cell);
+			else
+				pool->process_cell(tc, cell);
+		}
+	} while (!list_empty(&cells));
+}
+
 static void thin_get(struct thin_c *tc);
 static void thin_put(struct thin_c *tc);
 
@@ -1620,6 +1956,7 @@ static void process_deferred_bios(struct pool *pool)
 
 	tc = get_first_thin(pool);
 	while (tc) {
+		process_thin_deferred_cells(tc);
 		process_thin_deferred_bios(tc);
 		tc = get_next_thin(pool, tc);
 	}
@@ -1653,9 +1990,15 @@ static void do_worker(struct work_struct *ws)
 {
 	struct pool *pool = container_of(ws, struct pool, worker);
 
+	throttle_work_start(&pool->throttle);
+	dm_pool_issue_prefetches(pool->pmd);
+	throttle_work_update(&pool->throttle);
 	process_prepared(pool, &pool->prepared_mappings, &pool->process_prepared_mapping);
+	throttle_work_update(&pool->throttle);
 	process_prepared(pool, &pool->prepared_discards, &pool->process_prepared_discard);
+	throttle_work_update(&pool->throttle);
 	process_deferred_bios(pool);
+	throttle_work_complete(&pool->throttle);
 }
 
 /*
@@ -1792,6 +2135,8 @@ static void set_pool_mode(struct pool *pool, enum pool_mode new_mode)
 		dm_pool_metadata_read_only(pool->pmd);
 		pool->process_bio = process_bio_fail;
 		pool->process_discard = process_bio_fail;
+		pool->process_cell = process_cell_fail;
+		pool->process_discard_cell = process_cell_fail;
 		pool->process_prepared_mapping = process_prepared_mapping_fail;
 		pool->process_prepared_discard = process_prepared_discard_fail;
 
@@ -1804,6 +2149,8 @@ static void set_pool_mode(struct pool *pool, enum pool_mode new_mode)
 		dm_pool_metadata_read_only(pool->pmd);
 		pool->process_bio = process_bio_read_only;
 		pool->process_discard = process_bio_success;
+		pool->process_cell = process_cell_read_only;
+		pool->process_discard_cell = process_cell_success;
 		pool->process_prepared_mapping = process_prepared_mapping_fail;
 		pool->process_prepared_discard = process_prepared_discard_passdown;
 
@@ -1822,9 +2169,11 @@ static void set_pool_mode(struct pool *pool, enum pool_mode new_mode)
 		if (old_mode != new_mode)
 			notify_of_pool_mode_change(pool, "out-of-data-space");
 		pool->process_bio = process_bio_read_only;
-		pool->process_discard = process_discard;
+		pool->process_discard = process_discard_bio;
+		pool->process_cell = process_cell_read_only;
+		pool->process_discard_cell = process_discard_cell;
 		pool->process_prepared_mapping = process_prepared_mapping;
-		pool->process_prepared_discard = process_prepared_discard_passdown;
+		pool->process_prepared_discard = process_prepared_discard;
 
 		if (!pool->pf.error_if_no_space && no_space_timeout)
 			queue_delayed_work(pool->wq, &pool->no_space_timeout, no_space_timeout);
@@ -1835,7 +2184,9 @@ static void set_pool_mode(struct pool *pool, enum pool_mode new_mode)
 			notify_of_pool_mode_change(pool, "write");
 		dm_pool_metadata_read_write(pool->pmd);
 		pool->process_bio = process_bio;
-		pool->process_discard = process_discard;
+		pool->process_discard = process_discard_bio;
+		pool->process_cell = process_cell;
+		pool->process_discard_cell = process_discard_cell;
 		pool->process_prepared_mapping = process_prepared_mapping;
 		pool->process_prepared_discard = process_prepared_discard;
 		break;
@@ -1895,6 +2246,29 @@ static void thin_defer_bio(struct thin_c *tc, struct bio *bio)
 	wake_worker(pool);
 }
 
+static void thin_defer_bio_with_throttle(struct thin_c *tc, struct bio *bio)
+{
+	struct pool *pool = tc->pool;
+
+	throttle_lock(&pool->throttle);
+	thin_defer_bio(tc, bio);
+	throttle_unlock(&pool->throttle);
+}
+
+static void thin_defer_cell(struct thin_c *tc, struct dm_bio_prison_cell *cell)
+{
+	unsigned long flags;
+	struct pool *pool = tc->pool;
+
+	throttle_lock(&pool->throttle);
+	spin_lock_irqsave(&tc->lock, flags);
+	list_add_tail(&cell->user_list, &tc->deferred_cells);
+	spin_unlock_irqrestore(&tc->lock, flags);
+	throttle_unlock(&pool->throttle);
+
+	wake_worker(pool);
+}
+
 static void thin_hook_bio(struct thin_c *tc, struct bio *bio)
 {
 	struct dm_thin_endio_hook *h = dm_per_bio_data(bio, sizeof(struct dm_thin_endio_hook));
@@ -1915,8 +2289,7 @@ static int thin_bio_map(struct dm_target *ti, struct bio *bio)
 	dm_block_t block = get_bio_block(tc, bio);
 	struct dm_thin_device *td = tc->td;
 	struct dm_thin_lookup_result result;
-	struct dm_bio_prison_cell cell1, cell2;
-	struct dm_bio_prison_cell *cell_result;
+	struct dm_bio_prison_cell *virt_cell, *data_cell;
 	struct dm_cell_key key;
 
 	thin_hook_bio(tc, bio);
@@ -1932,10 +2305,18 @@ static int thin_bio_map(struct dm_target *ti, struct bio *bio)
 	}
 
 	if (bio->bi_rw & (REQ_DISCARD | REQ_FLUSH | REQ_FUA)) {
-		thin_defer_bio(tc, bio);
+		thin_defer_bio_with_throttle(tc, bio);
 		return DM_MAPIO_SUBMITTED;
 	}
 
+	/*
+	 * We must hold the virtual cell before doing the lookup, otherwise
+	 * there's a race with discard.
+	 */
+	build_virtual_key(tc->td, block, &key);
+	if (bio_detain(tc->pool, &key, bio, &virt_cell))
+		return DM_MAPIO_SUBMITTED;
+
 	r = dm_thin_find_block(td, block, 0, &result);
 
 	/*
@@ -1958,23 +2339,19 @@ static int thin_bio_map(struct dm_target *ti, struct bio *bio)
 			 * More distant ancestors are irrelevant. The
 			 * shared flag will be set in their case.
 			 */
-			thin_defer_bio(tc, bio);
+			thin_defer_cell(tc, virt_cell);
 			return DM_MAPIO_SUBMITTED;
 		}
 
-		build_virtual_key(tc->td, block, &key);
-		if (dm_bio_detain(tc->pool->prison, &key, bio, &cell1, &cell_result))
-			return DM_MAPIO_SUBMITTED;
-
 		build_data_key(tc->td, result.block, &key);
-		if (dm_bio_detain(tc->pool->prison, &key, bio, &cell2, &cell_result)) {
-			cell_defer_no_holder_no_free(tc, &cell1);
+		if (bio_detain(tc->pool, &key, bio, &data_cell)) {
+			cell_defer_no_holder(tc, virt_cell);
 			return DM_MAPIO_SUBMITTED;
 		}
 
 		inc_all_io_entry(tc->pool, bio);
-		cell_defer_no_holder_no_free(tc, &cell2);
-		cell_defer_no_holder_no_free(tc, &cell1);
+		cell_defer_no_holder(tc, data_cell);
+		cell_defer_no_holder(tc, virt_cell);
 
 		remap(tc, bio, result.block);
 		return DM_MAPIO_REMAPPED;
@@ -1986,16 +2363,13 @@ static int thin_bio_map(struct dm_target *ti, struct bio *bio)
 			 * of doing so.
 			 */
 			handle_unserviceable_bio(tc->pool, bio);
+			cell_defer_no_holder(tc, virt_cell);
 			return DM_MAPIO_SUBMITTED;
 		}
 		/* fall through */
 
 	case -EWOULDBLOCK:
-		/*
-		 * In future, the failed dm_thin_find_block above could
-		 * provide the hint to load the metadata into cache.
-		 */
-		thin_defer_bio(tc, bio);
+		thin_defer_cell(tc, virt_cell);
 		return DM_MAPIO_SUBMITTED;
 
 	default:
@@ -2005,6 +2379,7 @@ static int thin_bio_map(struct dm_target *ti, struct bio *bio)
 		 * pool is switched to fail-io mode.
 		 */
 		bio_io_error(bio);
+		cell_defer_no_holder(tc, virt_cell);
 		return DM_MAPIO_SUBMITTED;
 	}
 }
@@ -2185,7 +2560,7 @@ static struct pool *pool_create(struct mapped_device *pool_md,
 		pool->sectors_per_block_shift = __ffs(block_size);
 	pool->low_water_blocks = 0;
 	pool_features_init(&pool->pf);
-	pool->prison = dm_bio_prison_create(PRISON_CELLS);
+	pool->prison = dm_bio_prison_create();
 	if (!pool->prison) {
 		*error = "Error creating pool's bio prison";
 		err_p = ERR_PTR(-ENOMEM);
@@ -2211,6 +2586,7 @@ static struct pool *pool_create(struct mapped_device *pool_md,
 		goto bad_wq;
 	}
 
+	throttle_init(&pool->throttle);
 	INIT_WORK(&pool->worker, do_worker);
 	INIT_DELAYED_WORK(&pool->waker, do_waker);
 	INIT_DELAYED_WORK(&pool->no_space_timeout, do_no_space_timeout);
@@ -2220,6 +2596,7 @@ static struct pool *pool_create(struct mapped_device *pool_md,
 	INIT_LIST_HEAD(&pool->prepared_discards);
 	INIT_LIST_HEAD(&pool->active_thins);
 	pool->low_water_triggered = false;
+	pool->suspended = true;
 
 	pool->shared_read_ds = dm_deferred_set_create();
 	if (!pool->shared_read_ds) {
@@ -2756,20 +3133,77 @@ static int pool_preresume(struct dm_target *ti)
 	return 0;
 }
 
+static void pool_suspend_active_thins(struct pool *pool)
+{
+	struct thin_c *tc;
+
+	/* Suspend all active thin devices */
+	tc = get_first_thin(pool);
+	while (tc) {
+		dm_internal_suspend_noflush(tc->thin_md);
+		tc = get_next_thin(pool, tc);
+	}
+}
+
+static void pool_resume_active_thins(struct pool *pool)
+{
+	struct thin_c *tc;
+
+	/* Resume all active thin devices */
+	tc = get_first_thin(pool);
+	while (tc) {
+		dm_internal_resume(tc->thin_md);
+		tc = get_next_thin(pool, tc);
+	}
+}
+
 static void pool_resume(struct dm_target *ti)
 {
 	struct pool_c *pt = ti->private;
 	struct pool *pool = pt->pool;
 	unsigned long flags;
 
+	/*
+	 * Must requeue active_thins' bios and then resume
+	 * active_thins _before_ clearing 'suspend' flag.
+	 */
+	requeue_bios(pool);
+	pool_resume_active_thins(pool);
+
 	spin_lock_irqsave(&pool->lock, flags);
 	pool->low_water_triggered = false;
+	pool->suspended = false;
 	spin_unlock_irqrestore(&pool->lock, flags);
-	requeue_bios(pool);
 
 	do_waker(&pool->waker.work);
 }
 
+static void pool_presuspend(struct dm_target *ti)
+{
+	struct pool_c *pt = ti->private;
+	struct pool *pool = pt->pool;
+	unsigned long flags;
+
+	spin_lock_irqsave(&pool->lock, flags);
+	pool->suspended = true;
+	spin_unlock_irqrestore(&pool->lock, flags);
+
+	pool_suspend_active_thins(pool);
+}
+
+static void pool_presuspend_undo(struct dm_target *ti)
+{
+	struct pool_c *pt = ti->private;
+	struct pool *pool = pt->pool;
+	unsigned long flags;
+
+	pool_resume_active_thins(pool);
+
+	spin_lock_irqsave(&pool->lock, flags);
+	pool->suspended = false;
+	spin_unlock_irqrestore(&pool->lock, flags);
+}
+
 static void pool_postsuspend(struct dm_target *ti)
 {
 	struct pool_c *pt = ti->private;
@@ -2941,7 +3375,6 @@ static int process_release_metadata_snap_mesg(unsigned argc, char **argv, struct
  *   create_thin	<dev_id>
  *   create_snap	<dev_id> <origin_id>
  *   delete		<dev_id>
- *   trim		<dev_id> <new_size_in_sectors>
  *   set_transaction_id <current_trans_id> <new_trans_id>
  *   reserve_metadata_snap
  *   release_metadata_snap
@@ -2952,6 +3385,12 @@ static int pool_message(struct dm_target *ti, unsigned argc, char **argv)
 	struct pool_c *pt = ti->private;
 	struct pool *pool = pt->pool;
 
+	if (get_pool_mode(pool) >= PM_READ_ONLY) {
+		DMERR("%s: unable to service pool target messages in READ_ONLY or FAIL mode",
+		      dm_device_name(pool->pool_md));
+		return -EINVAL;
+	}
+
 	if (!strcasecmp(argv[0], "create_thin"))
 		r = process_create_thin_mesg(argc, argv, pool);
 
@@ -3169,15 +3608,35 @@ static void pool_io_hints(struct dm_target *ti, struct queue_limits *limits)
 {
 	struct pool_c *pt = ti->private;
 	struct pool *pool = pt->pool;
-	uint64_t io_opt_sectors = limits->io_opt >> SECTOR_SHIFT;
+	sector_t io_opt_sectors = limits->io_opt >> SECTOR_SHIFT;
+
+	/*
+	 * If max_sectors is smaller than pool->sectors_per_block adjust it
+	 * to the highest possible power-of-2 factor of pool->sectors_per_block.
+	 * This is especially beneficial when the pool's data device is a RAID
+	 * device that has a full stripe width that matches pool->sectors_per_block
+	 * -- because even though partial RAID stripe-sized IOs will be issued to a
+	 *    single RAID stripe; when aggregated they will end on a full RAID stripe
+	 *    boundary.. which avoids additional partial RAID stripe writes cascading
+	 */
+	if (limits->max_sectors < pool->sectors_per_block) {
+		while (!is_factor(pool->sectors_per_block, limits->max_sectors)) {
+			if ((limits->max_sectors & (limits->max_sectors - 1)) == 0)
+				limits->max_sectors--;
+			limits->max_sectors = rounddown_pow_of_two(limits->max_sectors);
+		}
+	}
 
 	/*
 	 * If the system-determined stacked limits are compatible with the
 	 * pool's blocksize (io_opt is a factor) do not override them.
 	 */
 	if (io_opt_sectors < pool->sectors_per_block ||
-	    do_div(io_opt_sectors, pool->sectors_per_block)) {
-		blk_limits_io_min(limits, pool->sectors_per_block << SECTOR_SHIFT);
+	    !is_factor(io_opt_sectors, pool->sectors_per_block)) {
+		if (is_factor(pool->sectors_per_block, limits->max_sectors))
+			blk_limits_io_min(limits, limits->max_sectors << SECTOR_SHIFT);
+		else
+			blk_limits_io_min(limits, pool->sectors_per_block << SECTOR_SHIFT);
 		blk_limits_io_opt(limits, pool->sectors_per_block << SECTOR_SHIFT);
 	}
 
@@ -3206,11 +3665,13 @@ static struct target_type pool_target = {
 	.name = "thin-pool",
 	.features = DM_TARGET_SINGLETON | DM_TARGET_ALWAYS_WRITEABLE |
 		    DM_TARGET_IMMUTABLE,
-	.version = {1, 13, 0},
+	.version = {1, 14, 0},
 	.module = THIS_MODULE,
 	.ctr = pool_ctr,
 	.dtr = pool_dtr,
 	.map = pool_map,
+	.presuspend = pool_presuspend,
+	.presuspend_undo = pool_presuspend_undo,
 	.postsuspend = pool_postsuspend,
 	.preresume = pool_preresume,
 	.resume = pool_resume,
@@ -3240,14 +3701,14 @@ static void thin_dtr(struct dm_target *ti)
 	struct thin_c *tc = ti->private;
 	unsigned long flags;
 
-	thin_put(tc);
-	wait_for_completion(&tc->can_destroy);
-
 	spin_lock_irqsave(&tc->pool->lock, flags);
 	list_del_rcu(&tc->list);
 	spin_unlock_irqrestore(&tc->pool->lock, flags);
 	synchronize_rcu();
 
+	thin_put(tc);
+	wait_for_completion(&tc->can_destroy);
+
 	mutex_lock(&dm_thin_pool_table.mutex);
 
 	__pool_dec(tc->pool);
@@ -3294,7 +3755,9 @@ static int thin_ctr(struct dm_target *ti, unsigned argc, char **argv)
 		r = -ENOMEM;
 		goto out_unlock;
 	}
+	tc->thin_md = dm_table_get_md(ti->table);
 	spin_lock_init(&tc->lock);
+	INIT_LIST_HEAD(&tc->deferred_cells);
 	bio_list_init(&tc->deferred_bio_list);
 	bio_list_init(&tc->retry_on_resume_list);
 	tc->sort_bio_list = RB_ROOT;
@@ -3339,18 +3802,18 @@ static int thin_ctr(struct dm_target *ti, unsigned argc, char **argv)
 	if (get_pool_mode(tc->pool) == PM_FAIL) {
 		ti->error = "Couldn't open thin device, Pool is in fail mode";
 		r = -EINVAL;
-		goto bad_thin_open;
+		goto bad_pool;
 	}
 
 	r = dm_pool_open_thin_device(tc->pool->pmd, tc->dev_id, &tc->td);
 	if (r) {
 		ti->error = "Couldn't open thin internal device";
-		goto bad_thin_open;
+		goto bad_pool;
 	}
 
 	r = dm_set_target_max_io_len(ti, tc->pool->sectors_per_block);
 	if (r)
-		goto bad_target_max_io_len;
+		goto bad;
 
 	ti->num_flush_bios = 1;
 	ti->flush_supported = true;
@@ -3365,14 +3828,18 @@ static int thin_ctr(struct dm_target *ti, unsigned argc, char **argv)
 		ti->split_discard_bios = true;
 	}
 
-	dm_put(pool_md);
-
 	mutex_unlock(&dm_thin_pool_table.mutex);
 
+	spin_lock_irqsave(&tc->pool->lock, flags);
+	if (tc->pool->suspended) {
+		spin_unlock_irqrestore(&tc->pool->lock, flags);
+		mutex_lock(&dm_thin_pool_table.mutex); /* reacquire for __pool_dec */
+		ti->error = "Unable to activate thin device while pool is suspended";
+		r = -EINVAL;
+		goto bad;
+	}
 	atomic_set(&tc->refcount, 1);
 	init_completion(&tc->can_destroy);
-
-	spin_lock_irqsave(&tc->pool->lock, flags);
 	list_add_tail_rcu(&tc->list, &tc->pool->active_thins);
 	spin_unlock_irqrestore(&tc->pool->lock, flags);
 	/*
@@ -3383,11 +3850,13 @@ static int thin_ctr(struct dm_target *ti, unsigned argc, char **argv)
 	 */
 	synchronize_rcu();
 
+	dm_put(pool_md);
+
 	return 0;
 
-bad_target_max_io_len:
+bad:
 	dm_pool_close_thin_device(tc->td);
-bad_thin_open:
+bad_pool:
 	__pool_dec(tc->pool);
 bad_pool_lookup:
 	dm_put(pool_md);
@@ -3533,6 +4002,21 @@ err:
 	DMEMIT("Error");
 }
 
+static int thin_merge(struct dm_target *ti, struct bvec_merge_data *bvm,
+		      struct bio_vec *biovec, int max_size)
+{
+	struct thin_c *tc = ti->private;
+	struct request_queue *q = bdev_get_queue(tc->pool_dev->bdev);
+
+	if (!q->merge_bvec_fn)
+		return max_size;
+
+	bvm->bi_bdev = tc->pool_dev->bdev;
+	bvm->bi_sector = dm_target_offset(ti, bvm->bi_sector);
+
+	return min(max_size, q->merge_bvec_fn(q, bvm, biovec));
+}
+
 static int thin_iterate_devices(struct dm_target *ti,
 				iterate_devices_callout_fn fn, void *data)
 {
@@ -3557,7 +4041,7 @@ static int thin_iterate_devices(struct dm_target *ti,
 
 static struct target_type thin_target = {
 	.name = "thin",
-	.version = {1, 13, 0},
+	.version = {1, 14, 0},
 	.module	= THIS_MODULE,
 	.ctr = thin_ctr,
 	.dtr = thin_dtr,
@@ -3567,6 +4051,7 @@ static struct target_type thin_target = {
 	.presuspend = thin_presuspend,
 	.postsuspend = thin_postsuspend,
 	.status = thin_status,
+	.merge = thin_merge,
 	.iterate_devices = thin_iterate_devices,
 };
 
diff --git a/drivers/md/dm.c b/drivers/md/dm.c
index 32b958dbc499..2caf5b374649 100644
--- a/drivers/md/dm.c
+++ b/drivers/md/dm.c
@@ -19,6 +19,7 @@
 #include <linux/idr.h>
 #include <linux/hdreg.h>
 #include <linux/delay.h>
+#include <linux/wait.h>
 
 #include <trace/events/block.h>
 
@@ -117,6 +118,7 @@ EXPORT_SYMBOL_GPL(dm_get_rq_mapinfo);
 #define DMF_NOFLUSH_SUSPENDING 5
 #define DMF_MERGE_IS_OPTIONAL 6
 #define DMF_DEFERRED_REMOVE 7
+#define DMF_SUSPENDED_INTERNALLY 8
 
 /*
  * A dummy definition to make RCU happy.
@@ -140,7 +142,10 @@ struct mapped_device {
 	 * Use dm_get_live_table{_fast} or take suspend_lock for
 	 * dereference.
 	 */
-	struct dm_table *map;
+	struct dm_table __rcu *map;
+
+	struct list_head table_devices;
+	struct mutex table_devices_lock;
 
 	unsigned long flags;
 
@@ -201,6 +206,9 @@ struct mapped_device {
 	/* zero-length flush that will be cloned and submitted to targets */
 	struct bio flush_bio;
 
+	/* the number of internal suspends */
+	unsigned internal_suspend_count;
+
 	struct dm_stats stats;
 };
 
@@ -212,6 +220,12 @@ struct dm_md_mempools {
 	struct bio_set *bs;
 };
 
+struct table_device {
+	struct list_head list;
+	atomic_t count;
+	struct dm_dev dm_dev;
+};
+
 #define RESERVED_BIO_BASED_IOS		16
 #define RESERVED_REQUEST_BASED_IOS	256
 #define RESERVED_MAX_IOS		1024
@@ -516,14 +530,15 @@ retry:
 		goto out;
 
 	tgt = dm_table_get_target(map, 0);
+	if (!tgt->type->ioctl)
+		goto out;
 
 	if (dm_suspended_md(md)) {
 		r = -EAGAIN;
 		goto out;
 	}
 
-	if (tgt->type->ioctl)
-		r = tgt->type->ioctl(tgt, cmd, arg);
+	r = tgt->type->ioctl(tgt, cmd, arg);
 
 out:
 	dm_put_live_table(md, srcu_idx);
@@ -593,13 +608,10 @@ static void end_io_acct(struct dm_io *io)
 	struct mapped_device *md = io->md;
 	struct bio *bio = io->bio;
 	unsigned long duration = jiffies - io->start_time;
-	int pending, cpu;
+	int pending;
 	int rw = bio_data_dir(bio);
 
-	cpu = part_stat_lock();
-	part_round_stats(cpu, &dm_disk(md)->part0);
-	part_stat_add(cpu, &dm_disk(md)->part0, ticks[rw], duration);
-	part_stat_unlock();
+	generic_end_io_acct(rw, &dm_disk(md)->part0, io->start_time);
 
 	if (unlikely(dm_stats_used(&md->stats)))
 		dm_stats_account_io(&md->stats, bio->bi_rw, bio->bi_iter.bi_sector,
@@ -670,6 +682,120 @@ static void dm_put_live_table_fast(struct mapped_device *md) __releases(RCU)
 }
 
 /*
+ * Open a table device so we can use it as a map destination.
+ */
+static int open_table_device(struct table_device *td, dev_t dev,
+			     struct mapped_device *md)
+{
+	static char *_claim_ptr = "I belong to device-mapper";
+	struct block_device *bdev;
+
+	int r;
+
+	BUG_ON(td->dm_dev.bdev);
+
+	bdev = blkdev_get_by_dev(dev, td->dm_dev.mode | FMODE_EXCL, _claim_ptr);
+	if (IS_ERR(bdev))
+		return PTR_ERR(bdev);
+
+	r = bd_link_disk_holder(bdev, dm_disk(md));
+	if (r) {
+		blkdev_put(bdev, td->dm_dev.mode | FMODE_EXCL);
+		return r;
+	}
+
+	td->dm_dev.bdev = bdev;
+	return 0;
+}
+
+/*
+ * Close a table device that we've been using.
+ */
+static void close_table_device(struct table_device *td, struct mapped_device *md)
+{
+	if (!td->dm_dev.bdev)
+		return;
+
+	bd_unlink_disk_holder(td->dm_dev.bdev, dm_disk(md));
+	blkdev_put(td->dm_dev.bdev, td->dm_dev.mode | FMODE_EXCL);
+	td->dm_dev.bdev = NULL;
+}
+
+static struct table_device *find_table_device(struct list_head *l, dev_t dev,
+					      fmode_t mode) {
+	struct table_device *td;
+
+	list_for_each_entry(td, l, list)
+		if (td->dm_dev.bdev->bd_dev == dev && td->dm_dev.mode == mode)
+			return td;
+
+	return NULL;
+}
+
+int dm_get_table_device(struct mapped_device *md, dev_t dev, fmode_t mode,
+			struct dm_dev **result) {
+	int r;
+	struct table_device *td;
+
+	mutex_lock(&md->table_devices_lock);
+	td = find_table_device(&md->table_devices, dev, mode);
+	if (!td) {
+		td = kmalloc(sizeof(*td), GFP_KERNEL);
+		if (!td) {
+			mutex_unlock(&md->table_devices_lock);
+			return -ENOMEM;
+		}
+
+		td->dm_dev.mode = mode;
+		td->dm_dev.bdev = NULL;
+
+		if ((r = open_table_device(td, dev, md))) {
+			mutex_unlock(&md->table_devices_lock);
+			kfree(td);
+			return r;
+		}
+
+		format_dev_t(td->dm_dev.name, dev);
+
+		atomic_set(&td->count, 0);
+		list_add(&td->list, &md->table_devices);
+	}
+	atomic_inc(&td->count);
+	mutex_unlock(&md->table_devices_lock);
+
+	*result = &td->dm_dev;
+	return 0;
+}
+EXPORT_SYMBOL_GPL(dm_get_table_device);
+
+void dm_put_table_device(struct mapped_device *md, struct dm_dev *d)
+{
+	struct table_device *td = container_of(d, struct table_device, dm_dev);
+
+	mutex_lock(&md->table_devices_lock);
+	if (atomic_dec_and_test(&td->count)) {
+		close_table_device(td, md);
+		list_del(&td->list);
+		kfree(td);
+	}
+	mutex_unlock(&md->table_devices_lock);
+}
+EXPORT_SYMBOL(dm_put_table_device);
+
+static void free_table_devices(struct list_head *devices)
+{
+	struct list_head *tmp, *next;
+
+	list_for_each_safe(tmp, next, devices) {
+		struct table_device *td = list_entry(tmp, struct table_device, list);
+
+		DMWARN("dm_destroy: %s still exists with %d references",
+		       td->dm_dev.name, atomic_read(&td->count));
+		kfree(td);
+	}
+}
+
+/*
  * Get the geometry associated with a dm device
  */
 int dm_get_geometry(struct mapped_device *md, struct hd_geometry *geo)
@@ -776,7 +902,7 @@ static void disable_write_same(struct mapped_device *md)
 
 static void clone_endio(struct bio *bio, int error)
 {
-	int r = 0;
+	int r = error;
 	struct dm_target_io *tio = container_of(bio, struct dm_target_io, clone);
 	struct dm_io *io = tio->io;
 	struct mapped_device *md = tio->io->md;
@@ -1249,13 +1375,13 @@ static void clone_bio(struct dm_target_io *tio, struct bio *bio,
 }
 
 static struct dm_target_io *alloc_tio(struct clone_info *ci,
-				      struct dm_target *ti, int nr_iovecs,
+				      struct dm_target *ti,
 				      unsigned target_bio_nr)
 {
 	struct dm_target_io *tio;
 	struct bio *clone;
 
-	clone = bio_alloc_bioset(GFP_NOIO, nr_iovecs, ci->md->bs);
+	clone = bio_alloc_bioset(GFP_NOIO, 0, ci->md->bs);
 	tio = container_of(clone, struct dm_target_io, clone);
 
 	tio->io = ci->io;
@@ -1269,17 +1395,12 @@ static void __clone_and_map_simple_bio(struct clone_info *ci,
 				       struct dm_target *ti,
 				       unsigned target_bio_nr, unsigned *len)
 {
-	struct dm_target_io *tio = alloc_tio(ci, ti, ci->bio->bi_max_vecs, target_bio_nr);
+	struct dm_target_io *tio = alloc_tio(ci, ti, target_bio_nr);
 	struct bio *clone = &tio->clone;
 
 	tio->len_ptr = len;
 
-	/*
-	 * Discard requests require the bio's inline iovecs be initialized.
-	 * ci->bio->bi_max_vecs is BIO_INLINE_VECS anyway, for both flush
-	 * and discard, so no need for concern about wasted bvec allocations.
-	 */
-	 __bio_clone_fast(clone, ci->bio);
+	__bio_clone_fast(clone, ci->bio);
 	if (len)
 		bio_setup_sector(clone, ci->sector, *len);
 
@@ -1322,7 +1443,7 @@ static void __clone_and_map_data_bio(struct clone_info *ci, struct dm_target *ti
 		num_target_bios = ti->num_write_bios(ti, bio);
 
 	for (target_bio_nr = 0; target_bio_nr < num_target_bios; target_bio_nr++) {
-		tio = alloc_tio(ci, ti, 0, target_bio_nr);
+		tio = alloc_tio(ci, ti, target_bio_nr);
 		tio->len_ptr = len;
 		clone_bio(tio, bio, sector, *len);
 		__map_bio(tio);
@@ -1489,9 +1610,9 @@ static int dm_merge_bvec(struct request_queue *q,
 	 * Find maximum amount of I/O that won't need splitting
 	 */
 	max_sectors = min(max_io_len(bvm->bi_sector, ti),
-			  (sector_t) BIO_MAX_SECTORS);
+			  (sector_t) queue_max_sectors(q));
 	max_size = (max_sectors << SECTOR_SHIFT) - bvm->bi_size;
-	if (max_size < 0)
+	if (unlikely(max_size < 0)) /* this shouldn't _ever_ happen */
 		max_size = 0;
 
 	/*
@@ -1503,10 +1624,10 @@ static int dm_merge_bvec(struct request_queue *q,
 		max_size = ti->type->merge(ti, bvm, biovec, max_size);
 	/*
 	 * If the target doesn't support merge method and some of the devices
-	 * provided their merge_bvec method (we know this by looking at
-	 * queue_max_hw_sectors), then we can't allow bios with multiple vector
-	 * entries.  So always set max_size to 0, and the code below allows
-	 * just one page.
+	 * provided their merge_bvec method (we know this by looking for the
+	 * max_hw_sectors that dm_set_device_limits may set), then we can't
+	 * allow bios with multiple vector entries.  So always set max_size
+	 * to 0, and the code below allows just one page.
 	 */
 	else if (queue_max_hw_sectors(q) <= PAGE_SIZE >> 9)
 		max_size = 0;
@@ -1530,16 +1651,12 @@ static void _dm_request(struct request_queue *q, struct bio *bio)
 {
 	int rw = bio_data_dir(bio);
 	struct mapped_device *md = q->queuedata;
-	int cpu;
 	int srcu_idx;
 	struct dm_table *map;
 
 	map = dm_get_live_table(md, &srcu_idx);
 
-	cpu = part_stat_lock();
-	part_stat_inc(cpu, &dm_disk(md)->part0, ios[rw]);
-	part_stat_add(cpu, &dm_disk(md)->part0, sectors[rw], bio_sectors(bio));
-	part_stat_unlock();
+	generic_start_io_acct(rw, bio_sectors(bio), &dm_disk(md)->part0);
 
 	/* if we're suspended, we have to queue this io for later */
 	if (unlikely(test_bit(DMF_BLOCK_IO_FOR_SUSPEND, &md->flags))) {
@@ -1949,12 +2066,14 @@ static struct mapped_device *alloc_dev(int minor)
 	md->type = DM_TYPE_NONE;
 	mutex_init(&md->suspend_lock);
 	mutex_init(&md->type_lock);
+	mutex_init(&md->table_devices_lock);
 	spin_lock_init(&md->deferred_lock);
 	atomic_set(&md->holders, 1);
 	atomic_set(&md->open_count, 0);
 	atomic_set(&md->event_nr, 0);
 	atomic_set(&md->uevent_seq, 0);
 	INIT_LIST_HEAD(&md->uevent_list);
+	INIT_LIST_HEAD(&md->table_devices);
 	spin_lock_init(&md->uevent_lock);
 
 	md->queue = blk_alloc_queue(GFP_KERNEL);
@@ -2040,6 +2159,7 @@ static void free_dev(struct mapped_device *md)
 	blk_integrity_unregister(md->disk);
 	del_gendisk(md->disk);
 	cleanup_srcu_struct(&md->io_barrier);
+	free_table_devices(&md->table_devices);
 	free_minor(minor);
 
 	spin_lock(&_minor_lock);
@@ -2211,7 +2331,7 @@ static struct dm_table *__bind(struct mapped_device *md, struct dm_table *t,
 
 	merge_is_optional = dm_table_merge_is_optional(t);
 
-	old_map = md->map;
+	old_map = rcu_dereference_protected(md->map, lockdep_is_held(&md->suspend_lock));
 	rcu_assign_pointer(md->map, t);
 	md->immutable_target_type = dm_table_get_immutable_target_type(t);
 
@@ -2220,7 +2340,8 @@ static struct dm_table *__bind(struct mapped_device *md, struct dm_table *t,
 		set_bit(DMF_MERGE_IS_OPTIONAL, &md->flags);
 	else
 		clear_bit(DMF_MERGE_IS_OPTIONAL, &md->flags);
-	dm_sync_table(md);
+	if (old_map)
+		dm_sync_table(md);
 
 	return old_map;
 }
@@ -2230,7 +2351,7 @@ static struct dm_table *__bind(struct mapped_device *md, struct dm_table *t,
  */
 static struct dm_table *__unbind(struct mapped_device *md)
 {
-	struct dm_table *map = md->map;
+	struct dm_table *map = rcu_dereference_protected(md->map, 1);
 
 	if (!map)
 		return NULL;
@@ -2595,36 +2716,18 @@ static void unlock_fs(struct mapped_device *md)
 }
 
 /*
- * We need to be able to change a mapping table under a mounted
- * filesystem.  For example we might want to move some data in
- * the background.  Before the table can be swapped with
- * dm_bind_table, dm_suspend must be called to flush any in
- * flight bios and ensure that any further io gets deferred.
- */
-/*
- * Suspend mechanism in request-based dm.
- *
- * 1. Flush all I/Os by lock_fs() if needed.
- * 2. Stop dispatching any I/O by stopping the request_queue.
- * 3. Wait for all in-flight I/Os to be completed or requeued.
+ * If __dm_suspend returns 0, the device is completely quiescent
+ * now. There is no request-processing activity. All new requests
+ * are being added to md->deferred list.
  *
- * To abort suspend, start the request_queue.
+ * Caller must hold md->suspend_lock
  */
-int dm_suspend(struct mapped_device *md, unsigned suspend_flags)
+static int __dm_suspend(struct mapped_device *md, struct dm_table *map,
+			unsigned suspend_flags, int interruptible)
 {
-	struct dm_table *map = NULL;
-	int r = 0;
-	int do_lockfs = suspend_flags & DM_SUSPEND_LOCKFS_FLAG ? 1 : 0;
-	int noflush = suspend_flags & DM_SUSPEND_NOFLUSH_FLAG ? 1 : 0;
-
-	mutex_lock(&md->suspend_lock);
-
-	if (dm_suspended_md(md)) {
-		r = -EINVAL;
-		goto out_unlock;
-	}
-
-	map = md->map;
+	bool do_lockfs = suspend_flags & DM_SUSPEND_LOCKFS_FLAG;
+	bool noflush = suspend_flags & DM_SUSPEND_NOFLUSH_FLAG;
+	int r;
 
 	/*
 	 * DMF_NOFLUSH_SUSPENDING must be set before presuspend.
@@ -2633,7 +2736,10 @@ int dm_suspend(struct mapped_device *md, unsigned suspend_flags)
 	if (noflush)
 		set_bit(DMF_NOFLUSH_SUSPENDING, &md->flags);
 
-	/* This does not get reverted if there's an error later. */
+	/*
+	 * This gets reverted if there's an error later and the targets
+	 * provide the .presuspend_undo hook.
+	 */
 	dm_table_presuspend_targets(map);
 
 	/*
@@ -2644,8 +2750,10 @@ int dm_suspend(struct mapped_device *md, unsigned suspend_flags)
 	 */
 	if (!noflush && do_lockfs) {
 		r = lock_fs(md);
-		if (r)
-			goto out_unlock;
+		if (r) {
+			dm_table_presuspend_undo_targets(map);
+			return r;
+		}
 	}
 
 	/*
@@ -2661,7 +2769,8 @@ int dm_suspend(struct mapped_device *md, unsigned suspend_flags)
 	 * flush_workqueue(md->wq).
 	 */
 	set_bit(DMF_BLOCK_IO_FOR_SUSPEND, &md->flags);
-	synchronize_srcu(&md->io_barrier);
+	if (map)
+		synchronize_srcu(&md->io_barrier);
 
 	/*
 	 * Stop md->queue before flushing md->wq in case request-based
@@ -2677,11 +2786,12 @@ int dm_suspend(struct mapped_device *md, unsigned suspend_flags)
 	 * We call dm_wait_for_completion to wait for all existing requests
 	 * to finish.
 	 */
-	r = dm_wait_for_completion(md, TASK_INTERRUPTIBLE);
+	r = dm_wait_for_completion(md, interruptible);
 
 	if (noflush)
 		clear_bit(DMF_NOFLUSH_SUSPENDING, &md->flags);
-	synchronize_srcu(&md->io_barrier);
+	if (map)
+		synchronize_srcu(&md->io_barrier);
 
 	/* were we interrupted ? */
 	if (r < 0) {
@@ -2691,14 +2801,56 @@ int dm_suspend(struct mapped_device *md, unsigned suspend_flags)
 			start_queue(md->queue);
 
 		unlock_fs(md);
-		goto out_unlock; /* pushback list is already flushed, so skip flush */
+		dm_table_presuspend_undo_targets(map);
+		/* pushback list is already flushed, so skip flush */
 	}
 
-	/*
-	 * If dm_wait_for_completion returned 0, the device is completely
-	 * quiescent now. There is no request-processing activity. All new
-	 * requests are being added to md->deferred list.
-	 */
+	return r;
+}
+
+/*
+ * We need to be able to change a mapping table under a mounted
+ * filesystem.  For example we might want to move some data in
+ * the background.  Before the table can be swapped with
+ * dm_bind_table, dm_suspend must be called to flush any in
+ * flight bios and ensure that any further io gets deferred.
+ */
+/*
+ * Suspend mechanism in request-based dm.
+ *
+ * 1. Flush all I/Os by lock_fs() if needed.
+ * 2. Stop dispatching any I/O by stopping the request_queue.
+ * 3. Wait for all in-flight I/Os to be completed or requeued.
+ *
+ * To abort suspend, start the request_queue.
+ */
+int dm_suspend(struct mapped_device *md, unsigned suspend_flags)
+{
+	struct dm_table *map = NULL;
+	int r = 0;
+
+retry:
+	mutex_lock_nested(&md->suspend_lock, SINGLE_DEPTH_NESTING);
+
+	if (dm_suspended_md(md)) {
+		r = -EINVAL;
+		goto out_unlock;
+	}
+
+	if (dm_suspended_internally_md(md)) {
+		/* already internally suspended, wait for internal resume */
+		mutex_unlock(&md->suspend_lock);
+		r = wait_on_bit(&md->flags, DMF_SUSPENDED_INTERNALLY, TASK_INTERRUPTIBLE);
+		if (r)
+			return r;
+		goto retry;
+	}
+
+	map = rcu_dereference_protected(md->map, lockdep_is_held(&md->suspend_lock));
+
+	r = __dm_suspend(md, map, suspend_flags, TASK_INTERRUPTIBLE);
+	if (r)
+		goto out_unlock;
 
 	set_bit(DMF_SUSPENDED, &md->flags);
 
@@ -2709,22 +2861,13 @@ out_unlock:
 	return r;
 }
 
-int dm_resume(struct mapped_device *md)
+static int __dm_resume(struct mapped_device *md, struct dm_table *map)
 {
-	int r = -EINVAL;
-	struct dm_table *map = NULL;
-
-	mutex_lock(&md->suspend_lock);
-	if (!dm_suspended_md(md))
-		goto out;
-
-	map = md->map;
-	if (!map || !dm_table_get_size(map))
-		goto out;
-
-	r = dm_table_resume_targets(map);
-	if (r)
-		goto out;
+	if (map) {
+		int r = dm_table_resume_targets(map);
+		if (r)
+			return r;
+	}
 
 	dm_queue_flush(md);
 
@@ -2738,6 +2881,37 @@ int dm_resume(struct mapped_device *md)
 
 	unlock_fs(md);
 
+	return 0;
+}
+
+int dm_resume(struct mapped_device *md)
+{
+	int r = -EINVAL;
+	struct dm_table *map = NULL;
+
+retry:
+	mutex_lock_nested(&md->suspend_lock, SINGLE_DEPTH_NESTING);
+
+	if (!dm_suspended_md(md))
+		goto out;
+
+	if (dm_suspended_internally_md(md)) {
+		/* already internally suspended, wait for internal resume */
+		mutex_unlock(&md->suspend_lock);
+		r = wait_on_bit(&md->flags, DMF_SUSPENDED_INTERNALLY, TASK_INTERRUPTIBLE);
+		if (r)
+			return r;
+		goto retry;
+	}
+
+	map = rcu_dereference_protected(md->map, lockdep_is_held(&md->suspend_lock));
+	if (!map || !dm_table_get_size(map))
+		goto out;
+
+	r = __dm_resume(md, map);
+	if (r)
+		goto out;
+
 	clear_bit(DMF_SUSPENDED, &md->flags);
 
 	r = 0;
@@ -2751,15 +2925,82 @@ out:
  * Internal suspend/resume works like userspace-driven suspend. It waits
  * until all bios finish and prevents issuing new bios to the target drivers.
  * It may be used only from the kernel.
- *
- * Internal suspend holds md->suspend_lock, which prevents interaction with
- * userspace-driven suspend.
  */
 
-void dm_internal_suspend(struct mapped_device *md)
+static void __dm_internal_suspend(struct mapped_device *md, unsigned suspend_flags)
 {
-	mutex_lock(&md->suspend_lock);
+	struct dm_table *map = NULL;
+
+	if (md->internal_suspend_count++)
+		return; /* nested internal suspend */
+
+	if (dm_suspended_md(md)) {
+		set_bit(DMF_SUSPENDED_INTERNALLY, &md->flags);
+		return; /* nest suspend */
+	}
+
+	map = rcu_dereference_protected(md->map, lockdep_is_held(&md->suspend_lock));
+
+	/*
+	 * Using TASK_UNINTERRUPTIBLE because only NOFLUSH internal suspend is
+	 * supported.  Properly supporting a TASK_INTERRUPTIBLE internal suspend
+	 * would require changing .presuspend to return an error -- avoid this
+	 * until there is a need for more elaborate variants of internal suspend.
+	 */
+	(void) __dm_suspend(md, map, suspend_flags, TASK_UNINTERRUPTIBLE);
+
+	set_bit(DMF_SUSPENDED_INTERNALLY, &md->flags);
+
+	dm_table_postsuspend_targets(map);
+}
+
+static void __dm_internal_resume(struct mapped_device *md)
+{
+	BUG_ON(!md->internal_suspend_count);
+
+	if (--md->internal_suspend_count)
+		return; /* resume from nested internal suspend */
+
 	if (dm_suspended_md(md))
+		goto done; /* resume from nested suspend */
+
+	/*
+	 * NOTE: existing callers don't need to call dm_table_resume_targets
+	 * (which may fail -- so best to avoid it for now by passing NULL map)
+	 */
+	(void) __dm_resume(md, NULL);
+
+done:
+	clear_bit(DMF_SUSPENDED_INTERNALLY, &md->flags);
+	smp_mb__after_atomic();
+	wake_up_bit(&md->flags, DMF_SUSPENDED_INTERNALLY);
+}
+
+void dm_internal_suspend_noflush(struct mapped_device *md)
+{
+	mutex_lock(&md->suspend_lock);
+	__dm_internal_suspend(md, DM_SUSPEND_NOFLUSH_FLAG);
+	mutex_unlock(&md->suspend_lock);
+}
+EXPORT_SYMBOL_GPL(dm_internal_suspend_noflush);
+
+void dm_internal_resume(struct mapped_device *md)
+{
+	mutex_lock(&md->suspend_lock);
+	__dm_internal_resume(md);
+	mutex_unlock(&md->suspend_lock);
+}
+EXPORT_SYMBOL_GPL(dm_internal_resume);
+
+/*
+ * Fast variants of internal suspend/resume hold md->suspend_lock,
+ * which prevents interaction with userspace-driven suspend.
+ */
+
+void dm_internal_suspend_fast(struct mapped_device *md)
+{
+	mutex_lock(&md->suspend_lock);
+	if (dm_suspended_md(md) || dm_suspended_internally_md(md))
 		return;
 
 	set_bit(DMF_BLOCK_IO_FOR_SUSPEND, &md->flags);
@@ -2768,9 +3009,9 @@ void dm_internal_suspend(struct mapped_device *md)
 	dm_wait_for_completion(md, TASK_UNINTERRUPTIBLE);
 }
 
-void dm_internal_resume(struct mapped_device *md)
+void dm_internal_resume_fast(struct mapped_device *md)
 {
-	if (dm_suspended_md(md))
+	if (dm_suspended_md(md) || dm_suspended_internally_md(md))
 		goto done;
 
 	dm_queue_flush(md);
@@ -2856,6 +3097,11 @@ int dm_suspended_md(struct mapped_device *md)
 	return test_bit(DMF_SUSPENDED, &md->flags);
 }
 
+int dm_suspended_internally_md(struct mapped_device *md)
+{
+	return test_bit(DMF_SUSPENDED_INTERNALLY, &md->flags);
+}
+
 int dm_test_deferred_remove_flag(struct mapped_device *md)
 {
 	return test_bit(DMF_DEFERRED_REMOVE, &md->flags);
@@ -2900,7 +3146,7 @@ struct dm_md_mempools *dm_alloc_md_mempools(unsigned type, unsigned integrity, u
 	if (!pools->io_pool)
 		goto out;
 
-	pools->bs = bioset_create(pool_size, front_pad);
+	pools->bs = bioset_create_nobvec(pool_size, front_pad);
 	if (!pools->bs)
 		goto out;
 
diff --git a/drivers/md/dm.h b/drivers/md/dm.h
index e81d2152fa68..84b0f9e4ba6c 100644
--- a/drivers/md/dm.h
+++ b/drivers/md/dm.h
@@ -44,7 +44,7 @@
 struct dm_dev_internal {
 	struct list_head list;
 	atomic_t count;
-	struct dm_dev dm_dev;
+	struct dm_dev *dm_dev;
 };
 
 struct dm_table;
@@ -65,6 +65,7 @@ void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q,
 			       struct queue_limits *limits);
 struct list_head *dm_table_get_devices(struct dm_table *t);
 void dm_table_presuspend_targets(struct dm_table *t);
+void dm_table_presuspend_undo_targets(struct dm_table *t);
 void dm_table_postsuspend_targets(struct dm_table *t);
 int dm_table_resume_targets(struct dm_table *t);
 int dm_table_any_congested(struct dm_table *t, int bdi_bits);
@@ -129,6 +130,15 @@ int dm_deleting_md(struct mapped_device *md);
 int dm_suspended_md(struct mapped_device *md);
 
 /*
+ * Internal suspend and resume methods.
+ */
+int dm_suspended_internally_md(struct mapped_device *md);
+void dm_internal_suspend_fast(struct mapped_device *md);
+void dm_internal_resume_fast(struct mapped_device *md);
+void dm_internal_suspend_noflush(struct mapped_device *md);
+void dm_internal_resume(struct mapped_device *md);
+
+/*
  * Test if the device is scheduled for deferred remove.
  */
 int dm_test_deferred_remove_flag(struct mapped_device *md);
@@ -188,6 +198,9 @@ int dm_cancel_deferred_remove(struct mapped_device *md);
 int dm_request_based(struct mapped_device *md);
 sector_t dm_get_size(struct mapped_device *md);
 struct request_queue *dm_get_md_queue(struct mapped_device *md);
+int dm_get_table_device(struct mapped_device *md, dev_t dev, fmode_t mode,
+			struct dm_dev **result);
+void dm_put_table_device(struct mapped_device *md, struct dm_dev *d);
 struct dm_stats *dm_get_stats(struct mapped_device *md);
 
 int dm_kobject_uevent(struct mapped_device *md, enum kobject_action action,
diff --git a/drivers/md/linear.c b/drivers/md/linear.c
index 56f534b4a2d2..64713b77df1c 100644
--- a/drivers/md/linear.c
+++ b/drivers/md/linear.c
@@ -10,10 +10,10 @@
    it under the terms of the GNU General Public License as published by
    the Free Software Foundation; either version 2, or (at your option)
    any later version.
-   
+
    You should have received a copy of the GNU General Public License
    (for example /usr/src/linux/COPYING); if not, write to the Free
-   Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  
+   Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 #include <linux/blkdev.h>
@@ -25,7 +25,7 @@
 #include "linear.h"
 
 /*
- * find which device holds a particular offset 
+ * find which device holds a particular offset
  */
 static inline struct dev_info *which_dev(struct mddev *mddev, sector_t sector)
 {
@@ -355,7 +355,6 @@ static void linear_status (struct seq_file *seq, struct mddev *mddev)
 	seq_printf(seq, " %dk rounding", mddev->chunk_sectors / 2);
 }
 
-
 static struct md_personality linear_personality =
 {
 	.name		= "linear",
@@ -379,7 +378,6 @@ static void linear_exit (void)
 	unregister_md_personality (&linear_personality);
 }
 
-
 module_init(linear_init);
 module_exit(linear_exit);
 MODULE_LICENSE("GPL");
diff --git a/drivers/md/md.c b/drivers/md/md.c
index 1294238610df..709755fb6d7b 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -1,6 +1,6 @@
 /*
    md.c : Multiple Devices driver for Linux
-	  Copyright (C) 1998, 1999, 2000 Ingo Molnar
+     Copyright (C) 1998, 1999, 2000 Ingo Molnar
 
      completely rewritten, based on the MD driver code from Marc Zyngier
 
@@ -66,8 +66,6 @@ static void autostart_arrays(int part);
 static LIST_HEAD(pers_list);
 static DEFINE_SPINLOCK(pers_lock);
 
-static void md_print_devices(void);
-
 static DECLARE_WAIT_QUEUE_HEAD(resync_wait);
 static struct workqueue_struct *md_wq;
 static struct workqueue_struct *md_misc_wq;
@@ -75,8 +73,6 @@ static struct workqueue_struct *md_misc_wq;
 static int remove_and_add_spares(struct mddev *mddev,
 				 struct md_rdev *this);
 
-#define MD_BUG(x...) { printk("md: bug in file %s, line %d\n", __FILE__, __LINE__); md_print_devices(); }
-
 /*
  * Default number of read corrections we'll attempt on an rdev
  * before ejecting it from the array. We divide the read error
@@ -218,7 +214,6 @@ static void md_new_event_inintr(struct mddev *mddev)
 static LIST_HEAD(all_mddevs);
 static DEFINE_SPINLOCK(all_mddevs_lock);
 
-
 /*
  * iterates through all used mddevs in the system.
  * We take care to grab the all_mddevs_lock whenever navigating
@@ -228,7 +223,7 @@ static DEFINE_SPINLOCK(all_mddevs_lock);
  */
 #define for_each_mddev(_mddev,_tmp)					\
 									\
-	for (({ spin_lock(&all_mddevs_lock); 				\
+	for (({ spin_lock(&all_mddevs_lock);				\
 		_tmp = all_mddevs.next;					\
 		_mddev = NULL;});					\
 	     ({ if (_tmp != &all_mddevs)				\
@@ -241,7 +236,6 @@ static DEFINE_SPINLOCK(all_mddevs_lock);
 		_tmp = _tmp->next;})					\
 		)
 
-
 /* Rather than calling directly into the personality make_request function,
  * IO requests come here first so that we can check if the device is
  * being suspended pending a reconfiguration.
@@ -253,7 +247,6 @@ static void md_make_request(struct request_queue *q, struct bio *bio)
 {
 	const int rw = bio_data_dir(bio);
 	struct mddev *mddev = q->queuedata;
-	int cpu;
 	unsigned int sectors;
 
 	if (mddev == NULL || mddev->pers == NULL
@@ -290,10 +283,7 @@ static void md_make_request(struct request_queue *q, struct bio *bio)
 	sectors = bio_sectors(bio);
 	mddev->pers->make_request(mddev, bio);
 
-	cpu = part_stat_lock();
-	part_stat_inc(cpu, &mddev->gendisk->part0, ios[rw]);
-	part_stat_add(cpu, &mddev->gendisk->part0, sectors[rw], sectors);
-	part_stat_unlock();
+	generic_start_io_acct(rw, sectors, &mddev->gendisk->part0);
 
 	if (atomic_dec_and_test(&mddev->active_io) && mddev->suspended)
 		wake_up(&mddev->sb_wait);
@@ -488,7 +478,7 @@ void mddev_init(struct mddev *mddev)
 }
 EXPORT_SYMBOL_GPL(mddev_init);
 
-static struct mddev * mddev_find(dev_t unit)
+static struct mddev *mddev_find(dev_t unit)
 {
 	struct mddev *mddev, *new = NULL;
 
@@ -530,7 +520,7 @@ static struct mddev * mddev_find(dev_t unit)
 				kfree(new);
 				return NULL;
 			}
-				
+
 			is_free = 1;
 			list_for_each_entry(mddev, &all_mddevs, all_mddevs)
 				if (mddev->unit == dev) {
@@ -562,7 +552,7 @@ static struct mddev * mddev_find(dev_t unit)
 	goto retry;
 }
 
-static inline int __must_check mddev_lock(struct mddev * mddev)
+static inline int __must_check mddev_lock(struct mddev *mddev)
 {
 	return mutex_lock_interruptible(&mddev->reconfig_mutex);
 }
@@ -570,7 +560,7 @@ static inline int __must_check mddev_lock(struct mddev * mddev)
 /* Sometimes we need to take the lock in a situation where
  * failure due to interrupts is not acceptable.
  */
-static inline void mddev_lock_nointr(struct mddev * mddev)
+static inline void mddev_lock_nointr(struct mddev *mddev)
 {
 	mutex_lock(&mddev->reconfig_mutex);
 }
@@ -580,14 +570,14 @@ static inline int mddev_is_locked(struct mddev *mddev)
 	return mutex_is_locked(&mddev->reconfig_mutex);
 }
 
-static inline int mddev_trylock(struct mddev * mddev)
+static inline int mddev_trylock(struct mddev *mddev)
 {
 	return mutex_trylock(&mddev->reconfig_mutex);
 }
 
 static struct attribute_group md_redundancy_group;
 
-static void mddev_unlock(struct mddev * mddev)
+static void mddev_unlock(struct mddev *mddev)
 {
 	if (mddev->to_remove) {
 		/* These cannot be removed under reconfig_mutex as
@@ -630,17 +620,6 @@ static void mddev_unlock(struct mddev * mddev)
 	spin_unlock(&pers_lock);
 }
 
-static struct md_rdev * find_rdev_nr(struct mddev *mddev, int nr)
-{
-	struct md_rdev *rdev;
-
-	rdev_for_each(rdev, mddev)
-		if (rdev->desc_nr == nr)
-			return rdev;
-
-	return NULL;
-}
-
 static struct md_rdev *find_rdev_nr_rcu(struct mddev *mddev, int nr)
 {
 	struct md_rdev *rdev;
@@ -693,11 +672,8 @@ static inline sector_t calc_dev_sboffset(struct md_rdev *rdev)
 	return MD_NEW_SIZE_SECTORS(num_sectors);
 }
 
-static int alloc_disk_sb(struct md_rdev * rdev)
+static int alloc_disk_sb(struct md_rdev *rdev)
 {
-	if (rdev->sb_page)
-		MD_BUG();
-
 	rdev->sb_page = alloc_page(GFP_KERNEL);
 	if (!rdev->sb_page) {
 		printk(KERN_ALERT "md: out of memory.\n");
@@ -766,14 +742,7 @@ void md_super_write(struct mddev *mddev, struct md_rdev *rdev,
 void md_super_wait(struct mddev *mddev)
 {
 	/* wait for all superblock writes that were scheduled to complete */
-	DEFINE_WAIT(wq);
-	for(;;) {
-		prepare_to_wait(&mddev->sb_wait, &wq, TASK_UNINTERRUPTIBLE);
-		if (atomic_read(&mddev->pending_writes)==0)
-			break;
-		schedule();
-	}
-	finish_wait(&mddev->sb_wait, &wq);
+	wait_event(mddev->sb_wait, atomic_read(&mddev->pending_writes)==0);
 }
 
 int sync_page_io(struct md_rdev *rdev, sector_t sector, int size,
@@ -801,17 +770,13 @@ int sync_page_io(struct md_rdev *rdev, sector_t sector, int size,
 }
 EXPORT_SYMBOL_GPL(sync_page_io);
 
-static int read_disk_sb(struct md_rdev * rdev, int size)
+static int read_disk_sb(struct md_rdev *rdev, int size)
 {
 	char b[BDEVNAME_SIZE];
-	if (!rdev->sb_page) {
-		MD_BUG();
-		return -EINVAL;
-	}
+
 	if (rdev->sb_loaded)
 		return 0;
 
-
 	if (!sync_page_io(rdev, 0, size, rdev->sb_page, READ, true))
 		goto fail;
 	rdev->sb_loaded = 1;
@@ -825,7 +790,7 @@ fail:
 
 static int uuid_equal(mdp_super_t *sb1, mdp_super_t *sb2)
 {
-	return 	sb1->set_uuid0 == sb2->set_uuid0 &&
+	return	sb1->set_uuid0 == sb2->set_uuid0 &&
 		sb1->set_uuid1 == sb2->set_uuid1 &&
 		sb1->set_uuid2 == sb2->set_uuid2 &&
 		sb1->set_uuid3 == sb2->set_uuid3;
@@ -861,14 +826,13 @@ abort:
 	return ret;
 }
 
-
 static u32 md_csum_fold(u32 csum)
 {
 	csum = (csum & 0xffff) + (csum >> 16);
 	return (csum & 0xffff) + (csum >> 16);
 }
 
-static unsigned int calc_sb_csum(mdp_super_t * sb)
+static unsigned int calc_sb_csum(mdp_super_t *sb)
 {
 	u64 newcsum = 0;
 	u32 *sb32 = (u32*)sb;
@@ -882,7 +846,6 @@ static unsigned int calc_sb_csum(mdp_super_t * sb)
 		newcsum += sb32[i];
 	csum = (newcsum & 0xffffffff) + (newcsum>>32);
 
-
 #ifdef CONFIG_ALPHA
 	/* This used to use csum_partial, which was wrong for several
 	 * reasons including that different results are returned on
@@ -899,7 +862,6 @@ static unsigned int calc_sb_csum(mdp_super_t * sb)
 	return csum;
 }
 
-
 /*
  * Handle superblock details.
  * We want to be able to handle multiple superblock formats
@@ -965,7 +927,7 @@ int md_check_no_bitmap(struct mddev *mddev)
 EXPORT_SYMBOL(md_check_no_bitmap);
 
 /*
- * load_super for 0.90.0 
+ * load_super for 0.90.0
  */
 static int super_90_load(struct md_rdev *rdev, struct md_rdev *refdev, int minor_version)
 {
@@ -1044,7 +1006,7 @@ static int super_90_load(struct md_rdev *rdev, struct md_rdev *refdev, int minor
 		ev2 = md_event(refsb);
 		if (ev1 > ev2)
 			ret = 1;
-		else 
+		else
 			ret = 0;
 	}
 	rdev->sectors = rdev->sb_start;
@@ -1118,7 +1080,7 @@ static int super_90_validate(struct mddev *mddev, struct md_rdev *rdev)
 		if (sb->state & (1<<MD_SB_CLEAN))
 			mddev->recovery_cp = MaxSector;
 		else {
-			if (sb->events_hi == sb->cp_events_hi && 
+			if (sb->events_hi == sb->cp_events_hi &&
 				sb->events_lo == sb->cp_events_lo) {
 				mddev->recovery_cp = sb->recovery_cp;
 			} else
@@ -1146,7 +1108,7 @@ static int super_90_validate(struct mddev *mddev, struct md_rdev *rdev)
 		++ev1;
 		if (sb->disks[rdev->desc_nr].state & (
 			    (1<<MD_DISK_SYNC) | (1 << MD_DISK_ACTIVE)))
-			if (ev1 < mddev->events) 
+			if (ev1 < mddev->events)
 				return -EINVAL;
 	} else if (mddev->bitmap) {
 		/* if adding to array with a bitmap, then we can accept an
@@ -1197,7 +1159,6 @@ static void super_90_sync(struct mddev *mddev, struct md_rdev *rdev)
 	struct md_rdev *rdev2;
 	int next_spare = mddev->raid_disks;
 
-
 	/* make rdev->sb match mddev data..
 	 *
 	 * 1/ zero out disks
@@ -1366,7 +1327,7 @@ super_90_allow_new_offset(struct md_rdev *rdev, unsigned long long new_offset)
  * version 1 superblock
  */
 
-static __le32 calc_sb_1_csum(struct mdp_superblock_1 * sb)
+static __le32 calc_sb_1_csum(struct mdp_superblock_1 *sb)
 {
 	__le32 disk_csum;
 	u32 csum;
@@ -1430,7 +1391,6 @@ static int super_1_load(struct md_rdev *rdev, struct md_rdev *refdev, int minor_
 	ret = read_disk_sb(rdev, 4096);
 	if (ret) return ret;
 
-
 	sb = page_address(rdev->sb_page);
 
 	if (sb->magic != cpu_to_le32(MD_SB_MAGIC) ||
@@ -1817,7 +1777,7 @@ retry:
 
 	for (i=0; i<max_dev;i++)
 		sb->dev_roles[i] = cpu_to_le16(0xfffe);
-	
+
 	rdev_for_each(rdev2, mddev) {
 		i = rdev2->desc_nr;
 		if (test_bit(Faulty, &rdev2->flags))
@@ -2033,18 +1993,13 @@ void md_integrity_add_rdev(struct md_rdev *rdev, struct mddev *mddev)
 }
 EXPORT_SYMBOL(md_integrity_add_rdev);
 
-static int bind_rdev_to_array(struct md_rdev * rdev, struct mddev * mddev)
+static int bind_rdev_to_array(struct md_rdev *rdev, struct mddev *mddev)
 {
 	char b[BDEVNAME_SIZE];
 	struct kobject *ko;
 	char *s;
 	int err;
 
-	if (rdev->mddev) {
-		MD_BUG();
-		return -EINVAL;
-	}
-
 	/* prevent duplicates */
 	if (find_rdev(mddev, rdev->bdev->bd_dev))
 		return -EEXIST;
@@ -2067,16 +2022,21 @@ static int bind_rdev_to_array(struct md_rdev * rdev, struct mddev * mddev)
 	 * If it is -1, assign a free number, else
 	 * check number is not in use
 	 */
+	rcu_read_lock();
 	if (rdev->desc_nr < 0) {
 		int choice = 0;
-		if (mddev->pers) choice = mddev->raid_disks;
-		while (find_rdev_nr(mddev, choice))
+		if (mddev->pers)
+			choice = mddev->raid_disks;
+		while (find_rdev_nr_rcu(mddev, choice))
 			choice++;
 		rdev->desc_nr = choice;
 	} else {
-		if (find_rdev_nr(mddev, rdev->desc_nr))
+		if (find_rdev_nr_rcu(mddev, rdev->desc_nr)) {
+			rcu_read_unlock();
 			return -EBUSY;
+		}
 	}
+	rcu_read_unlock();
 	if (mddev->max_disks && rdev->desc_nr >= mddev->max_disks) {
 		printk(KERN_WARNING "md: %s: array is limited to %d devices\n",
 		       mdname(mddev), mddev->max_disks);
@@ -2118,13 +2078,10 @@ static void md_delayed_delete(struct work_struct *ws)
 	kobject_put(&rdev->kobj);
 }
 
-static void unbind_rdev_from_array(struct md_rdev * rdev)
+static void unbind_rdev_from_array(struct md_rdev *rdev)
 {
 	char b[BDEVNAME_SIZE];
-	if (!rdev->mddev) {
-		MD_BUG();
-		return;
-	}
+
 	bd_unlink_disk_holder(rdev->bdev, rdev->mddev->gendisk);
 	list_del_rcu(&rdev->same_set);
 	printk(KERN_INFO "md: unbind<%s>\n", bdevname(rdev->bdev,b));
@@ -2169,20 +2126,17 @@ static void unlock_rdev(struct md_rdev *rdev)
 {
 	struct block_device *bdev = rdev->bdev;
 	rdev->bdev = NULL;
-	if (!bdev)
-		MD_BUG();
 	blkdev_put(bdev, FMODE_READ|FMODE_WRITE|FMODE_EXCL);
 }
 
 void md_autodetect_dev(dev_t dev);
 
-static void export_rdev(struct md_rdev * rdev)
+static void export_rdev(struct md_rdev *rdev)
 {
 	char b[BDEVNAME_SIZE];
+
 	printk(KERN_INFO "md: export_rdev(%s)\n",
 		bdevname(rdev->bdev,b));
-	if (rdev->mddev)
-		MD_BUG();
 	md_rdev_clear(rdev);
 #ifndef MODULE
 	if (test_bit(AutoDetected, &rdev->flags))
@@ -2192,7 +2146,7 @@ static void export_rdev(struct md_rdev * rdev)
 	kobject_put(&rdev->kobj);
 }
 
-static void kick_rdev_from_array(struct md_rdev * rdev)
+static void kick_rdev_from_array(struct md_rdev *rdev)
 {
 	unbind_rdev_from_array(rdev);
 	export_rdev(rdev);
@@ -2200,153 +2154,18 @@ static void kick_rdev_from_array(struct md_rdev * rdev)
 
 static void export_array(struct mddev *mddev)
 {
-	struct md_rdev *rdev, *tmp;
+	struct md_rdev *rdev;
 
-	rdev_for_each_safe(rdev, tmp, mddev) {
-		if (!rdev->mddev) {
-			MD_BUG();
-			continue;
-		}
+	while (!list_empty(&mddev->disks)) {
+		rdev = list_first_entry(&mddev->disks, struct md_rdev,
+					same_set);
 		kick_rdev_from_array(rdev);
 	}
-	if (!list_empty(&mddev->disks))
-		MD_BUG();
 	mddev->raid_disks = 0;
 	mddev->major_version = 0;
 }
 
-static void print_desc(mdp_disk_t *desc)
-{
-	printk(" DISK<N:%d,(%d,%d),R:%d,S:%d>\n", desc->number,
-		desc->major,desc->minor,desc->raid_disk,desc->state);
-}
-
-static void print_sb_90(mdp_super_t *sb)
-{
-	int i;
-
-	printk(KERN_INFO 
-		"md:  SB: (V:%d.%d.%d) ID:<%08x.%08x.%08x.%08x> CT:%08x\n",
-		sb->major_version, sb->minor_version, sb->patch_version,
-		sb->set_uuid0, sb->set_uuid1, sb->set_uuid2, sb->set_uuid3,
-		sb->ctime);
-	printk(KERN_INFO "md:     L%d S%08d ND:%d RD:%d md%d LO:%d CS:%d\n",
-		sb->level, sb->size, sb->nr_disks, sb->raid_disks,
-		sb->md_minor, sb->layout, sb->chunk_size);
-	printk(KERN_INFO "md:     UT:%08x ST:%d AD:%d WD:%d"
-		" FD:%d SD:%d CSUM:%08x E:%08lx\n",
-		sb->utime, sb->state, sb->active_disks, sb->working_disks,
-		sb->failed_disks, sb->spare_disks,
-		sb->sb_csum, (unsigned long)sb->events_lo);
-
-	printk(KERN_INFO);
-	for (i = 0; i < MD_SB_DISKS; i++) {
-		mdp_disk_t *desc;
-
-		desc = sb->disks + i;
-		if (desc->number || desc->major || desc->minor ||
-		    desc->raid_disk || (desc->state && (desc->state != 4))) {
-			printk("     D %2d: ", i);
-			print_desc(desc);
-		}
-	}
-	printk(KERN_INFO "md:     THIS: ");
-	print_desc(&sb->this_disk);
-}
-
-static void print_sb_1(struct mdp_superblock_1 *sb)
-{
-	__u8 *uuid;
-
-	uuid = sb->set_uuid;
-	printk(KERN_INFO
-	       "md:  SB: (V:%u) (F:0x%08x) Array-ID:<%pU>\n"
-	       "md:    Name: \"%s\" CT:%llu\n",
-		le32_to_cpu(sb->major_version),
-		le32_to_cpu(sb->feature_map),
-		uuid,
-		sb->set_name,
-		(unsigned long long)le64_to_cpu(sb->ctime)
-		       & MD_SUPERBLOCK_1_TIME_SEC_MASK);
-
-	uuid = sb->device_uuid;
-	printk(KERN_INFO
-	       "md:       L%u SZ%llu RD:%u LO:%u CS:%u DO:%llu DS:%llu SO:%llu"
-			" RO:%llu\n"
-	       "md:     Dev:%08x UUID: %pU\n"
-	       "md:       (F:0x%08x) UT:%llu Events:%llu ResyncOffset:%llu CSUM:0x%08x\n"
-	       "md:         (MaxDev:%u) \n",
-		le32_to_cpu(sb->level),
-		(unsigned long long)le64_to_cpu(sb->size),
-		le32_to_cpu(sb->raid_disks),
-		le32_to_cpu(sb->layout),
-		le32_to_cpu(sb->chunksize),
-		(unsigned long long)le64_to_cpu(sb->data_offset),
-		(unsigned long long)le64_to_cpu(sb->data_size),
-		(unsigned long long)le64_to_cpu(sb->super_offset),
-		(unsigned long long)le64_to_cpu(sb->recovery_offset),
-		le32_to_cpu(sb->dev_number),
-		uuid,
-		sb->devflags,
-		(unsigned long long)le64_to_cpu(sb->utime) & MD_SUPERBLOCK_1_TIME_SEC_MASK,
-		(unsigned long long)le64_to_cpu(sb->events),
-		(unsigned long long)le64_to_cpu(sb->resync_offset),
-		le32_to_cpu(sb->sb_csum),
-		le32_to_cpu(sb->max_dev)
-		);
-}
-
-static void print_rdev(struct md_rdev *rdev, int major_version)
-{
-	char b[BDEVNAME_SIZE];
-	printk(KERN_INFO "md: rdev %s, Sect:%08llu F:%d S:%d DN:%u\n",
-		bdevname(rdev->bdev, b), (unsigned long long)rdev->sectors,
-	        test_bit(Faulty, &rdev->flags), test_bit(In_sync, &rdev->flags),
-	        rdev->desc_nr);
-	if (rdev->sb_loaded) {
-		printk(KERN_INFO "md: rdev superblock (MJ:%d):\n", major_version);
-		switch (major_version) {
-		case 0:
-			print_sb_90(page_address(rdev->sb_page));
-			break;
-		case 1:
-			print_sb_1(page_address(rdev->sb_page));
-			break;
-		}
-	} else
-		printk(KERN_INFO "md: no rdev superblock!\n");
-}
-
-static void md_print_devices(void)
-{
-	struct list_head *tmp;
-	struct md_rdev *rdev;
-	struct mddev *mddev;
-	char b[BDEVNAME_SIZE];
-
-	printk("\n");
-	printk("md:	**********************************\n");
-	printk("md:	* <COMPLETE RAID STATE PRINTOUT> *\n");
-	printk("md:	**********************************\n");
-	for_each_mddev(mddev, tmp) {
-
-		if (mddev->bitmap)
-			bitmap_print_sb(mddev->bitmap);
-		else
-			printk("%s: ", mdname(mddev));
-		rdev_for_each(rdev, mddev)
-			printk("<%s>", bdevname(rdev->bdev,b));
-		printk("\n");
-
-		rdev_for_each(rdev, mddev)
-			print_rdev(rdev, mddev->major_version);
-	}
-	printk("md:	**********************************\n");
-	printk("\n");
-}
-
-
-static void sync_sbs(struct mddev * mddev, int nospares)
+static void sync_sbs(struct mddev *mddev, int nospares)
 {
 	/* Update each superblock (in-memory image), but
 	 * if we are allowed to, skip spares which already
@@ -2369,7 +2188,7 @@ static void sync_sbs(struct mddev * mddev, int nospares)
 	}
 }
 
-static void md_update_sb(struct mddev * mddev, int force_change)
+static void md_update_sb(struct mddev *mddev, int force_change)
 {
 	struct md_rdev *rdev;
 	int sync_req;
@@ -2390,7 +2209,7 @@ repeat:
 		    mddev->curr_resync_completed > rdev->recovery_offset)
 				rdev->recovery_offset = mddev->curr_resync_completed;
 
-	}	
+	}
 	if (!mddev->persistent) {
 		clear_bit(MD_CHANGE_CLEAN, &mddev->flags);
 		clear_bit(MD_CHANGE_DEVS, &mddev->flags);
@@ -2453,15 +2272,12 @@ repeat:
 		mddev->can_decrease_events = nospares;
 	}
 
-	if (!mddev->events) {
-		/*
-		 * oops, this 64-bit counter should never wrap.
-		 * Either we are in around ~1 trillion A.C., assuming
-		 * 1 reboot per second, or we have a bug:
-		 */
-		MD_BUG();
-		mddev->events --;
-	}
+	/*
+	 * This 64-bit counter should never wrap.
+	 * Either we are in around ~1 trillion A.C., assuming
+	 * 1 reboot per second, or we have a bug...
+	 */
+	WARN_ON(mddev->events == 0);
 
 	rdev_for_each(rdev, mddev) {
 		if (rdev->badblocks.changed)
@@ -2668,10 +2484,12 @@ state_store(struct md_rdev *rdev, const char *buf, size_t len)
 		set_bit(In_sync, &rdev->flags);
 		err = 0;
 	} else if (cmd_match(buf, "-insync") && rdev->raid_disk >= 0) {
-		clear_bit(In_sync, &rdev->flags);
-		rdev->saved_raid_disk = rdev->raid_disk;
-		rdev->raid_disk = -1;
-		err = 0;
+		if (rdev->mddev->pers == NULL) {
+			clear_bit(In_sync, &rdev->flags);
+			rdev->saved_raid_disk = rdev->raid_disk;
+			rdev->raid_disk = -1;
+			err = 0;
+		}
 	} else if (cmd_match(buf, "write_error")) {
 		set_bit(WriteErrorSeen, &rdev->flags);
 		err = 0;
@@ -2829,7 +2647,6 @@ slot_store(struct md_rdev *rdev, const char *buf, size_t len)
 	return len;
 }
 
-
 static struct rdev_sysfs_entry rdev_slot =
 __ATTR(slot, S_IRUGO|S_IWUSR, slot_show, slot_store);
 
@@ -2874,7 +2691,8 @@ static ssize_t new_offset_store(struct md_rdev *rdev,
 	if (kstrtoull(buf, 10, &new_offset) < 0)
 		return -EINVAL;
 
-	if (mddev->sync_thread)
+	if (mddev->sync_thread ||
+	    test_bit(MD_RECOVERY_RUNNING,&mddev->recovery))
 		return -EBUSY;
 	if (new_offset == rdev->data_offset)
 		/* reset is always permitted */
@@ -2980,20 +2798,20 @@ rdev_size_store(struct md_rdev *rdev, const char *buf, size_t len)
 
 	rdev->sectors = sectors;
 	if (sectors > oldsectors && my_mddev->external) {
-		/* need to check that all other rdevs with the same ->bdev
-		 * do not overlap.  We need to unlock the mddev to avoid
-		 * a deadlock.  We have already changed rdev->sectors, and if
-		 * we have to change it back, we will have the lock again.
+		/* Need to check that all other rdevs with the same
+		 * ->bdev do not overlap.  'rcu' is sufficient to walk
+		 * the rdev lists safely.
+		 * This check does not provide a hard guarantee, it
+		 * just helps avoid dangerous mistakes.
 		 */
 		struct mddev *mddev;
 		int overlap = 0;
 		struct list_head *tmp;
 
-		mddev_unlock(my_mddev);
+		rcu_read_lock();
 		for_each_mddev(mddev, tmp) {
 			struct md_rdev *rdev2;
 
-			mddev_lock_nointr(mddev);
 			rdev_for_each(rdev2, mddev)
 				if (rdev->bdev == rdev2->bdev &&
 				    rdev != rdev2 &&
@@ -3003,13 +2821,12 @@ rdev_size_store(struct md_rdev *rdev, const char *buf, size_t len)
 					overlap = 1;
 					break;
 				}
-			mddev_unlock(mddev);
 			if (overlap) {
 				mddev_put(mddev);
 				break;
 			}
 		}
-		mddev_lock_nointr(my_mddev);
+		rcu_read_unlock();
 		if (overlap) {
 			/* Someone else could have slipped in a size
 			 * change here, but doing so is just silly.
@@ -3027,7 +2844,6 @@ rdev_size_store(struct md_rdev *rdev, const char *buf, size_t len)
 static struct rdev_sysfs_entry rdev_size =
 __ATTR(size, S_IRUGO|S_IWUSR, rdev_size_show, rdev_size_store);
 
-
 static ssize_t recovery_start_show(struct md_rdev *rdev, char *page)
 {
 	unsigned long long recovery_start = rdev->recovery_offset;
@@ -3063,7 +2879,6 @@ static ssize_t recovery_start_store(struct md_rdev *rdev, const char *buf, size_
 static struct rdev_sysfs_entry rdev_recovery_start =
 __ATTR(recovery_start, S_IRUGO|S_IWUSR, recovery_start_show, recovery_start_store);
 
-
 static ssize_t
 badblocks_show(struct badblocks *bb, char *page, int unack);
 static ssize_t
@@ -3084,7 +2899,6 @@ static ssize_t bb_store(struct md_rdev *rdev, const char *page, size_t len)
 static struct rdev_sysfs_entry rdev_bad_blocks =
 __ATTR(bad_blocks, S_IRUGO|S_IWUSR, bb_show, bb_store);
 
-
 static ssize_t ubb_show(struct md_rdev *rdev, char *page)
 {
 	return badblocks_show(&rdev->badblocks, page, 1);
@@ -3241,7 +3055,7 @@ static struct md_rdev *md_import_device(dev_t newdev, int super_format, int supe
 
 	size = i_size_read(rdev->bdev->bd_inode) >> BLOCK_SIZE_BITS;
 	if (!size) {
-		printk(KERN_WARNING 
+		printk(KERN_WARNING
 			"md: %s has zero or unknown size, marking faulty!\n",
 			bdevname(rdev->bdev,b));
 		err = -EINVAL;
@@ -3260,7 +3074,7 @@ static struct md_rdev *md_import_device(dev_t newdev, int super_format, int supe
 			goto abort_free;
 		}
 		if (err < 0) {
-			printk(KERN_WARNING 
+			printk(KERN_WARNING
 				"md: could not read %s's sb, not importing!\n",
 				bdevname(rdev->bdev,b));
 			goto abort_free;
@@ -3281,8 +3095,7 @@ abort_free:
  * Check a full RAID array for plausibility
  */
 
-
-static void analyze_sbs(struct mddev * mddev)
+static void analyze_sbs(struct mddev *mddev)
 {
 	int i;
 	struct md_rdev *rdev, *freshest, *tmp;
@@ -3300,12 +3113,11 @@ static void analyze_sbs(struct mddev * mddev)
 		default:
 			printk( KERN_ERR \
 				"md: fatal superblock inconsistency in %s"
-				" -- removing from array\n", 
+				" -- removing from array\n",
 				bdevname(rdev->bdev,b));
 			kick_rdev_from_array(rdev);
 		}
 
-
 	super_types[mddev->major_version].
 		validate_super(mddev, freshest);
 
@@ -3344,7 +3156,7 @@ static void analyze_sbs(struct mddev * mddev)
 /* Read a fixed-point number.
  * Numbers in sysfs attributes should be in "standard" units where
  * possible, so time should be in seconds.
- * However we internally use a a much smaller unit such as 
+ * However we internally use a a much smaller unit such as
  * milliseconds or jiffies.
  * This function takes a decimal number with a possible fractional
  * component, and produces an integer which is the result of
@@ -3381,7 +3193,6 @@ int strict_strtoul_scaled(const char *cp, unsigned long *res, int scale)
 	return 0;
 }
 
-
 static void md_safemode_timeout(unsigned long data);
 
 static ssize_t
@@ -3458,6 +3269,7 @@ level_store(struct mddev *mddev, const char *buf, size_t len)
 	 */
 
 	if (mddev->sync_thread ||
+	    test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) ||
 	    mddev->reshape_position != MaxSector ||
 	    mddev->sysfs_active)
 		return -EBUSY;
@@ -3524,7 +3336,7 @@ level_store(struct mddev *mddev, const char *buf, size_t len)
 	/* Looks like we have a winner */
 	mddev_suspend(mddev);
 	mddev->pers->stop(mddev);
-	
+
 	if (mddev->pers->sync_request == NULL &&
 	    pers->sync_request != NULL) {
 		/* need to add the md_redundancy_group */
@@ -3533,7 +3345,7 @@ level_store(struct mddev *mddev, const char *buf, size_t len)
 			       "md: cannot register extra attributes for %s\n",
 			       mdname(mddev));
 		mddev->sysfs_action = sysfs_get_dirent(mddev->kobj.sd, "sync_action");
-	}		
+	}
 	if (mddev->pers->sync_request != NULL &&
 	    pers->sync_request == NULL) {
 		/* need to remove the md_redundancy_group */
@@ -3611,7 +3423,6 @@ level_store(struct mddev *mddev, const char *buf, size_t len)
 static struct md_sysfs_entry md_level =
 __ATTR(level, S_IRUGO|S_IWUSR, level_show, level_store);
 
-
 static ssize_t
 layout_show(struct mddev *mddev, char *page)
 {
@@ -3654,7 +3465,6 @@ layout_store(struct mddev *mddev, const char *buf, size_t len)
 static struct md_sysfs_entry md_layout =
 __ATTR(layout, S_IRUGO|S_IWUSR, layout_show, layout_store);
 
-
 static ssize_t
 raid_disks_show(struct mddev *mddev, char *page)
 {
@@ -3859,9 +3669,9 @@ array_state_show(struct mddev *mddev, char *page)
 	return sprintf(page, "%s\n", array_states[st]);
 }
 
-static int do_md_stop(struct mddev * mddev, int ro, struct block_device *bdev);
-static int md_set_readonly(struct mddev * mddev, struct block_device *bdev);
-static int do_md_run(struct mddev * mddev);
+static int do_md_stop(struct mddev *mddev, int ro, struct block_device *bdev);
+static int md_set_readonly(struct mddev *mddev, struct block_device *bdev);
+static int do_md_run(struct mddev *mddev);
 static int restart_array(struct mddev *mddev);
 
 static ssize_t
@@ -4012,7 +3822,6 @@ new_dev_store(struct mddev *mddev, const char *buf, size_t len)
 	    minor != MINOR(dev))
 		return -EOVERFLOW;
 
-
 	if (mddev->persistent) {
 		rdev = md_import_device(dev, mddev->major_version,
 					mddev->minor_version);
@@ -4108,7 +3917,6 @@ size_store(struct mddev *mddev, const char *buf, size_t len)
 static struct md_sysfs_entry md_size =
 __ATTR(component_size, S_IRUGO|S_IWUSR, size_show, size_store);
 
-
 /* Metadata version.
  * This is one of
  *   'none' for arrays with no metadata (good luck...)
@@ -4216,6 +4024,7 @@ action_store(struct mddev *mddev, const char *page, size_t len)
 		clear_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
 
 	if (cmd_match(page, "idle") || cmd_match(page, "frozen")) {
+		flush_workqueue(md_misc_wq);
 		if (mddev->sync_thread) {
 			set_bit(MD_RECOVERY_INTR, &mddev->recovery);
 			md_reap_sync_thread(mddev);
@@ -4490,7 +4299,7 @@ suspend_lo_store(struct mddev *mddev, const char *buf, size_t len)
 	unsigned long long new = simple_strtoull(buf, &e, 10);
 	unsigned long long old = mddev->suspend_lo;
 
-	if (mddev->pers == NULL || 
+	if (mddev->pers == NULL ||
 	    mddev->pers->quiesce == NULL)
 		return -EINVAL;
 	if (buf == e || (*e && *e != '\n'))
@@ -4510,7 +4319,6 @@ suspend_lo_store(struct mddev *mddev, const char *buf, size_t len)
 static struct md_sysfs_entry md_suspend_lo =
 __ATTR(suspend_lo, S_IRUGO|S_IWUSR, suspend_lo_show, suspend_lo_store);
 
-
 static ssize_t
 suspend_hi_show(struct mddev *mddev, char *page)
 {
@@ -4698,7 +4506,6 @@ static struct attribute_group md_redundancy_group = {
 	.attrs = md_redundancy_attrs,
 };
 
-
 static ssize_t
 md_attr_show(struct kobject *kobj, struct attribute *attr, char *page)
 {
@@ -5111,7 +4918,7 @@ int md_run(struct mddev *mddev)
 	} else if (mddev->ro == 2) /* auto-readonly not meaningful */
 		mddev->ro = 0;
 
- 	atomic_set(&mddev->writes_pending,0);
+	atomic_set(&mddev->writes_pending,0);
 	atomic_set(&mddev->max_corr_read_errors,
 		   MD_DEFAULT_MAX_CORRECTED_READ_ERRORS);
 	mddev->safemode = 0;
@@ -5125,9 +4932,9 @@ int md_run(struct mddev *mddev)
 		if (rdev->raid_disk >= 0)
 			if (sysfs_link_rdev(mddev, rdev))
 				/* failure here is OK */;
-	
+
 	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
-	
+
 	if (mddev->flags & MD_UPDATE_SB_FLAGS)
 		md_update_sb(mddev, 0);
 
@@ -5236,6 +5043,7 @@ static void md_clean(struct mddev *mddev)
 static void __md_stop_writes(struct mddev *mddev)
 {
 	set_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
+	flush_workqueue(md_misc_wq);
 	if (mddev->sync_thread) {
 		set_bit(MD_RECOVERY_INTR, &mddev->recovery);
 		md_reap_sync_thread(mddev);
@@ -5296,23 +5104,27 @@ static int md_set_readonly(struct mddev *mddev, struct block_device *bdev)
 		set_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
 		md_wakeup_thread(mddev->thread);
 	}
-	if (mddev->sync_thread) {
+	if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery))
 		set_bit(MD_RECOVERY_INTR, &mddev->recovery);
+	if (mddev->sync_thread)
 		/* Thread might be blocked waiting for metadata update
 		 * which will now never happen */
 		wake_up_process(mddev->sync_thread->tsk);
-	}
+
 	mddev_unlock(mddev);
-	wait_event(resync_wait, mddev->sync_thread == NULL);
+	wait_event(resync_wait, !test_bit(MD_RECOVERY_RUNNING,
+					  &mddev->recovery));
 	mddev_lock_nointr(mddev);
 
 	mutex_lock(&mddev->open_mutex);
-	if (atomic_read(&mddev->openers) > !!bdev ||
+	if ((mddev->pers && atomic_read(&mddev->openers) > !!bdev) ||
 	    mddev->sync_thread ||
+	    test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) ||
 	    (bdev && !test_bit(MD_STILL_CLOSED, &mddev->flags))) {
 		printk("md: %s still in use.\n",mdname(mddev));
 		if (did_freeze) {
 			clear_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
+			set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
 			md_wakeup_thread(mddev->thread);
 		}
 		err = -EBUSY;
@@ -5327,6 +5139,8 @@ static int md_set_readonly(struct mddev *mddev, struct block_device *bdev)
 		mddev->ro = 1;
 		set_disk_ro(mddev->gendisk, 1);
 		clear_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
+		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
+		md_wakeup_thread(mddev->thread);
 		sysfs_notify_dirent_safe(mddev->sysfs_state);
 		err = 0;
 	}
@@ -5339,7 +5153,7 @@ out:
  *   0 - completely stop and dis-assemble array
  *   2 - stop but do not disassemble array
  */
-static int do_md_stop(struct mddev * mddev, int mode,
+static int do_md_stop(struct mddev *mddev, int mode,
 		      struct block_device *bdev)
 {
 	struct gendisk *disk = mddev->gendisk;
@@ -5351,25 +5165,30 @@ static int do_md_stop(struct mddev * mddev, int mode,
 		set_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
 		md_wakeup_thread(mddev->thread);
 	}
-	if (mddev->sync_thread) {
+	if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery))
 		set_bit(MD_RECOVERY_INTR, &mddev->recovery);
+	if (mddev->sync_thread)
 		/* Thread might be blocked waiting for metadata update
 		 * which will now never happen */
 		wake_up_process(mddev->sync_thread->tsk);
-	}
+
 	mddev_unlock(mddev);
-	wait_event(resync_wait, mddev->sync_thread == NULL);
+	wait_event(resync_wait, (mddev->sync_thread == NULL &&
+				 !test_bit(MD_RECOVERY_RUNNING,
+					   &mddev->recovery)));
 	mddev_lock_nointr(mddev);
 
 	mutex_lock(&mddev->open_mutex);
-	if (atomic_read(&mddev->openers) > !!bdev ||
+	if ((mddev->pers && atomic_read(&mddev->openers) > !!bdev) ||
 	    mddev->sysfs_active ||
 	    mddev->sync_thread ||
+	    test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) ||
 	    (bdev && !test_bit(MD_STILL_CLOSED, &mddev->flags))) {
 		printk("md: %s still in use.\n",mdname(mddev));
 		mutex_unlock(&mddev->open_mutex);
 		if (did_freeze) {
 			clear_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
+			set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
 			md_wakeup_thread(mddev->thread);
 		}
 		return -EBUSY;
@@ -5512,12 +5331,12 @@ static void autorun_devices(int part)
 				"md: cannot allocate memory for md drive.\n");
 			break;
 		}
-		if (mddev_lock(mddev)) 
+		if (mddev_lock(mddev))
 			printk(KERN_WARNING "md: %s locked, cannot run\n",
 			       mdname(mddev));
 		else if (mddev->raid_disks || mddev->major_version
 			 || !list_empty(&mddev->disks)) {
-			printk(KERN_WARNING 
+			printk(KERN_WARNING
 				"md: %s already running, cannot run %s\n",
 				mdname(mddev), bdevname(rdev0->bdev,b));
 			mddev_unlock(mddev);
@@ -5545,7 +5364,7 @@ static void autorun_devices(int part)
 }
 #endif /* !MODULE */
 
-static int get_version(void __user * arg)
+static int get_version(void __user *arg)
 {
 	mdu_version_t ver;
 
@@ -5559,7 +5378,7 @@ static int get_version(void __user * arg)
 	return 0;
 }
 
-static int get_array_info(struct mddev * mddev, void __user * arg)
+static int get_array_info(struct mddev *mddev, void __user *arg)
 {
 	mdu_array_info_t info;
 	int nr,working,insync,failed,spare;
@@ -5574,7 +5393,7 @@ static int get_array_info(struct mddev * mddev, void __user * arg)
 		else {
 			working++;
 			if (test_bit(In_sync, &rdev->flags))
-				insync++;	
+				insync++;
 			else
 				spare++;
 		}
@@ -5614,7 +5433,7 @@ static int get_array_info(struct mddev * mddev, void __user * arg)
 	return 0;
 }
 
-static int get_bitmap_file(struct mddev * mddev, void __user * arg)
+static int get_bitmap_file(struct mddev *mddev, void __user * arg)
 {
 	mdu_bitmap_file_t *file = NULL; /* too big for stack allocation */
 	char *ptr, *buf = NULL;
@@ -5652,7 +5471,7 @@ out:
 	return err;
 }
 
-static int get_disk_info(struct mddev * mddev, void __user * arg)
+static int get_disk_info(struct mddev *mddev, void __user * arg)
 {
 	mdu_disk_info_t info;
 	struct md_rdev *rdev;
@@ -5688,7 +5507,7 @@ static int get_disk_info(struct mddev * mddev, void __user * arg)
 	return 0;
 }
 
-static int add_new_disk(struct mddev * mddev, mdu_disk_info_t *info)
+static int add_new_disk(struct mddev *mddev, mdu_disk_info_t *info)
 {
 	char b[BDEVNAME_SIZE], b2[BDEVNAME_SIZE];
 	struct md_rdev *rdev;
@@ -5702,7 +5521,7 @@ static int add_new_disk(struct mddev * mddev, mdu_disk_info_t *info)
 		/* expecting a device which has a superblock */
 		rdev = md_import_device(dev, mddev->major_version, mddev->minor_version);
 		if (IS_ERR(rdev)) {
-			printk(KERN_WARNING 
+			printk(KERN_WARNING
 				"md: md_import_device returned %ld\n",
 				PTR_ERR(rdev));
 			return PTR_ERR(rdev);
@@ -5714,9 +5533,9 @@ static int add_new_disk(struct mddev * mddev, mdu_disk_info_t *info)
 			err = super_types[mddev->major_version]
 				.load_super(rdev, rdev0, mddev->minor_version);
 			if (err < 0) {
-				printk(KERN_WARNING 
+				printk(KERN_WARNING
 					"md: %s has different UUID to %s\n",
-					bdevname(rdev->bdev,b), 
+					bdevname(rdev->bdev,b),
 					bdevname(rdev0->bdev,b2));
 				export_rdev(rdev);
 				return -EINVAL;
@@ -5736,7 +5555,7 @@ static int add_new_disk(struct mddev * mddev, mdu_disk_info_t *info)
 	if (mddev->pers) {
 		int err;
 		if (!mddev->pers->hot_add_disk) {
-			printk(KERN_WARNING 
+			printk(KERN_WARNING
 				"%s: personality does not support diskops!\n",
 			       mdname(mddev));
 			return -EINVAL;
@@ -5747,7 +5566,7 @@ static int add_new_disk(struct mddev * mddev, mdu_disk_info_t *info)
 		else
 			rdev = md_import_device(dev, -1, -1);
 		if (IS_ERR(rdev)) {
-			printk(KERN_WARNING 
+			printk(KERN_WARNING
 				"md: md_import_device returned %ld\n",
 				PTR_ERR(rdev));
 			return PTR_ERR(rdev);
@@ -5821,7 +5640,7 @@ static int add_new_disk(struct mddev * mddev, mdu_disk_info_t *info)
 		int err;
 		rdev = md_import_device(dev, -1, 0);
 		if (IS_ERR(rdev)) {
-			printk(KERN_WARNING 
+			printk(KERN_WARNING
 				"md: error, md_import_device() returned %ld\n",
 				PTR_ERR(rdev));
 			return PTR_ERR(rdev);
@@ -5856,7 +5675,7 @@ static int add_new_disk(struct mddev * mddev, mdu_disk_info_t *info)
 	return 0;
 }
 
-static int hot_remove_disk(struct mddev * mddev, dev_t dev)
+static int hot_remove_disk(struct mddev *mddev, dev_t dev)
 {
 	char b[BDEVNAME_SIZE];
 	struct md_rdev *rdev;
@@ -5882,7 +5701,7 @@ busy:
 	return -EBUSY;
 }
 
-static int hot_add_disk(struct mddev * mddev, dev_t dev)
+static int hot_add_disk(struct mddev *mddev, dev_t dev)
 {
 	char b[BDEVNAME_SIZE];
 	int err;
@@ -5898,7 +5717,7 @@ static int hot_add_disk(struct mddev * mddev, dev_t dev)
 		return -EINVAL;
 	}
 	if (!mddev->pers->hot_add_disk) {
-		printk(KERN_WARNING 
+		printk(KERN_WARNING
 			"%s: personality does not support diskops!\n",
 			mdname(mddev));
 		return -EINVAL;
@@ -5906,7 +5725,7 @@ static int hot_add_disk(struct mddev * mddev, dev_t dev)
 
 	rdev = md_import_device(dev, -1, 0);
 	if (IS_ERR(rdev)) {
-		printk(KERN_WARNING 
+		printk(KERN_WARNING
 			"md: error, md_import_device() returned %ld\n",
 			PTR_ERR(rdev));
 		return -EINVAL;
@@ -5920,7 +5739,7 @@ static int hot_add_disk(struct mddev * mddev, dev_t dev)
 	rdev->sectors = rdev->sb_start;
 
 	if (test_bit(Faulty, &rdev->flags)) {
-		printk(KERN_WARNING 
+		printk(KERN_WARNING
 			"md: can not hot-add faulty %s disk to %s!\n",
 			bdevname(rdev->bdev,b), mdname(mddev));
 		err = -EINVAL;
@@ -5968,7 +5787,6 @@ static int set_bitmap_file(struct mddev *mddev, int fd)
 		/* we should be able to change the bitmap.. */
 	}
 
-
 	if (fd >= 0) {
 		struct inode *inode;
 		if (mddev->bitmap)
@@ -6039,7 +5857,7 @@ static int set_bitmap_file(struct mddev *mddev, int fd)
  *  The minor and patch _version numbers are also kept incase the
  *  super_block handler wishes to interpret them.
  */
-static int set_array_info(struct mddev * mddev, mdu_array_info_t *info)
+static int set_array_info(struct mddev *mddev, mdu_array_info_t *info)
 {
 
 	if (info->raid_disks == 0) {
@@ -6048,7 +5866,7 @@ static int set_array_info(struct mddev * mddev, mdu_array_info_t *info)
 		    info->major_version >= ARRAY_SIZE(super_types) ||
 		    super_types[info->major_version].name == NULL) {
 			/* maybe try to auto-load a module? */
-			printk(KERN_INFO 
+			printk(KERN_INFO
 				"md: superblock version %d not known\n",
 				info->major_version);
 			return -EINVAL;
@@ -6139,7 +5957,8 @@ static int update_size(struct mddev *mddev, sector_t num_sectors)
 	 * of each device.  If num_sectors is zero, we find the largest size
 	 * that fits.
 	 */
-	if (mddev->sync_thread)
+	if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) ||
+	    mddev->sync_thread)
 		return -EBUSY;
 	if (mddev->ro)
 		return -EROFS;
@@ -6170,7 +5989,9 @@ static int update_raid_disks(struct mddev *mddev, int raid_disks)
 	if (raid_disks <= 0 ||
 	    (mddev->max_disks && raid_disks >= mddev->max_disks))
 		return -EINVAL;
-	if (mddev->sync_thread || mddev->reshape_position != MaxSector)
+	if (mddev->sync_thread ||
+	    test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) ||
+	    mddev->reshape_position != MaxSector)
 		return -EBUSY;
 
 	rdev_for_each(rdev, mddev) {
@@ -6196,7 +6017,6 @@ static int update_raid_disks(struct mddev *mddev, int raid_disks)
 	return rv;
 }
 
-
 /*
  * update_array_info is used to change the configuration of an
  * on-line array.
@@ -6347,7 +6167,6 @@ static inline bool md_ioctl_valid(unsigned int cmd)
 	case GET_DISK_INFO:
 	case HOT_ADD_DISK:
 	case HOT_REMOVE_DISK:
-	case PRINT_RAID_DEBUG:
 	case RAID_AUTORUN:
 	case RAID_VERSION:
 	case RESTART_ARRAY_RW:
@@ -6391,18 +6210,13 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode,
 	switch (cmd) {
 	case RAID_VERSION:
 		err = get_version(argp);
-		goto done;
-
-	case PRINT_RAID_DEBUG:
-		err = 0;
-		md_print_devices();
-		goto done;
+		goto out;
 
 #ifndef MODULE
 	case RAID_AUTORUN:
 		err = 0;
 		autostart_arrays(arg);
-		goto done;
+		goto out;
 #endif
 	default:;
 	}
@@ -6415,7 +6229,7 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode,
 
 	if (!mddev) {
 		BUG();
-		goto abort;
+		goto out;
 	}
 
 	/* Some actions do not requires the mutex */
@@ -6425,18 +6239,18 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode,
 			err = -ENODEV;
 		else
 			err = get_array_info(mddev, argp);
-		goto abort;
+		goto out;
 
 	case GET_DISK_INFO:
 		if (!mddev->raid_disks && !mddev->external)
 			err = -ENODEV;
 		else
 			err = get_disk_info(mddev, argp);
-		goto abort;
+		goto out;
 
 	case SET_DISK_FAULTY:
 		err = set_disk_faulty(mddev, new_decode_dev(arg));
-		goto abort;
+		goto out;
 	}
 
 	if (cmd == ADD_NEW_DISK)
@@ -6454,10 +6268,10 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode,
 		 * and writes
 		 */
 		mutex_lock(&mddev->open_mutex);
-		if (atomic_read(&mddev->openers) > 1) {
+		if (mddev->pers && atomic_read(&mddev->openers) > 1) {
 			mutex_unlock(&mddev->open_mutex);
 			err = -EBUSY;
-			goto abort;
+			goto out;
 		}
 		set_bit(MD_STILL_CLOSED, &mddev->flags);
 		mutex_unlock(&mddev->open_mutex);
@@ -6465,10 +6279,10 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode,
 	}
 	err = mddev_lock(mddev);
 	if (err) {
-		printk(KERN_INFO 
+		printk(KERN_INFO
 			"md: ioctl lock interrupted, reason %d, cmd %d\n",
 			err, cmd);
-		goto abort;
+		goto out;
 	}
 
 	if (cmd == SET_ARRAY_INFO) {
@@ -6477,38 +6291,38 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode,
 			memset(&info, 0, sizeof(info));
 		else if (copy_from_user(&info, argp, sizeof(info))) {
 			err = -EFAULT;
-			goto abort_unlock;
+			goto unlock;
 		}
 		if (mddev->pers) {
 			err = update_array_info(mddev, &info);
 			if (err) {
 				printk(KERN_WARNING "md: couldn't update"
 				       " array info. %d\n", err);
-				goto abort_unlock;
+				goto unlock;
 			}
-			goto done_unlock;
+			goto unlock;
 		}
 		if (!list_empty(&mddev->disks)) {
 			printk(KERN_WARNING
 			       "md: array %s already has disks!\n",
 			       mdname(mddev));
 			err = -EBUSY;
-			goto abort_unlock;
+			goto unlock;
 		}
 		if (mddev->raid_disks) {
 			printk(KERN_WARNING
 			       "md: array %s already initialised!\n",
 			       mdname(mddev));
 			err = -EBUSY;
-			goto abort_unlock;
+			goto unlock;
 		}
 		err = set_array_info(mddev, &info);
 		if (err) {
 			printk(KERN_WARNING "md: couldn't set"
 			       " array info. %d\n", err);
-			goto abort_unlock;
+			goto unlock;
 		}
-		goto done_unlock;
+		goto unlock;
 	}
 
 	/*
@@ -6521,7 +6335,7 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode,
 	    && cmd != RUN_ARRAY && cmd != SET_BITMAP_FILE
 	    && cmd != GET_BITMAP_FILE) {
 		err = -ENODEV;
-		goto abort_unlock;
+		goto unlock;
 	}
 
 	/*
@@ -6530,23 +6344,23 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode,
 	switch (cmd) {
 	case GET_BITMAP_FILE:
 		err = get_bitmap_file(mddev, argp);
-		goto done_unlock;
+		goto unlock;
 
 	case RESTART_ARRAY_RW:
 		err = restart_array(mddev);
-		goto done_unlock;
+		goto unlock;
 
 	case STOP_ARRAY:
 		err = do_md_stop(mddev, 0, bdev);
-		goto done_unlock;
+		goto unlock;
 
 	case STOP_ARRAY_RO:
 		err = md_set_readonly(mddev, bdev);
-		goto done_unlock;
+		goto unlock;
 
 	case HOT_REMOVE_DISK:
 		err = hot_remove_disk(mddev, new_decode_dev(arg));
-		goto done_unlock;
+		goto unlock;
 
 	case ADD_NEW_DISK:
 		/* We can support ADD_NEW_DISK on read-only arrays
@@ -6562,14 +6376,14 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode,
 				break;
 			else
 				err = add_new_disk(mddev, &info);
-			goto done_unlock;
+			goto unlock;
 		}
 		break;
 
 	case BLKROSET:
 		if (get_user(ro, (int __user *)(arg))) {
 			err = -EFAULT;
-			goto done_unlock;
+			goto unlock;
 		}
 		err = -EINVAL;
 
@@ -6577,11 +6391,11 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode,
 		 * does not matter, no writes are coming
 		 */
 		if (ro)
-			goto done_unlock;
+			goto unlock;
 
 		/* are we are already prepared for writes? */
 		if (mddev->ro != 1)
-			goto done_unlock;
+			goto unlock;
 
 		/* transitioning to readauto need only happen for
 		 * arrays that call md_write_start
@@ -6593,17 +6407,14 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode,
 				set_disk_ro(mddev->gendisk, 0);
 			}
 		}
-		goto done_unlock;
+		goto unlock;
 	}
 
 	/*
 	 * The remaining ioctls are changing the state of the
 	 * superblock, so we do not allow them on read-only arrays.
-	 * However non-MD ioctls (e.g. get-size) will still come through
-	 * here and hit the 'default' below, so only disallow
-	 * 'md' ioctls, and switch to rw mode if started auto-readonly.
 	 */
-	if (_IOC_TYPE(cmd) == MD_MAJOR && mddev->ro && mddev->pers) {
+	if (mddev->ro && mddev->pers) {
 		if (mddev->ro == 2) {
 			mddev->ro = 0;
 			sysfs_notify_dirent_safe(mddev->sysfs_state);
@@ -6621,7 +6432,7 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode,
 			}
 		} else {
 			err = -EROFS;
-			goto abort_unlock;
+			goto unlock;
 		}
 	}
 
@@ -6633,38 +6444,32 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode,
 			err = -EFAULT;
 		else
 			err = add_new_disk(mddev, &info);
-		goto done_unlock;
+		goto unlock;
 	}
 
 	case HOT_ADD_DISK:
 		err = hot_add_disk(mddev, new_decode_dev(arg));
-		goto done_unlock;
+		goto unlock;
 
 	case RUN_ARRAY:
 		err = do_md_run(mddev);
-		goto done_unlock;
+		goto unlock;
 
 	case SET_BITMAP_FILE:
 		err = set_bitmap_file(mddev, (int)arg);
-		goto done_unlock;
+		goto unlock;
 
 	default:
 		err = -EINVAL;
-		goto abort_unlock;
+		goto unlock;
 	}
 
-done_unlock:
-abort_unlock:
+unlock:
 	if (mddev->hold_active == UNTIL_IOCTL &&
 	    err != -EINVAL)
 		mddev->hold_active = 0;
 	mddev_unlock(mddev);
-
-	return err;
-done:
-	if (err)
-		MD_BUG();
-abort:
+out:
 	return err;
 }
 #ifdef CONFIG_COMPAT
@@ -6726,7 +6531,7 @@ static int md_open(struct block_device *bdev, fmode_t mode)
 
 static void md_release(struct gendisk *disk, fmode_t mode)
 {
- 	struct mddev *mddev = disk->private_data;
+	struct mddev *mddev = disk->private_data;
 
 	BUG_ON(!mddev);
 	atomic_dec(&mddev->openers);
@@ -6761,7 +6566,7 @@ static const struct block_device_operations md_fops =
 	.revalidate_disk= md_revalidate,
 };
 
-static int md_thread(void * arg)
+static int md_thread(void *arg)
 {
 	struct md_thread *thread = arg;
 
@@ -6810,6 +6615,7 @@ void md_wakeup_thread(struct md_thread *thread)
 		wake_up(&thread->wqueue);
 	}
 }
+EXPORT_SYMBOL(md_wakeup_thread);
 
 struct md_thread *md_register_thread(void (*run) (struct md_thread *),
 		struct mddev *mddev, const char *name)
@@ -6835,6 +6641,7 @@ struct md_thread *md_register_thread(void (*run) (struct md_thread *),
 	}
 	return thread;
 }
+EXPORT_SYMBOL(md_register_thread);
 
 void md_unregister_thread(struct md_thread **threadp)
 {
@@ -6852,14 +6659,10 @@ void md_unregister_thread(struct md_thread **threadp)
 	kthread_stop(thread->tsk);
 	kfree(thread);
 }
+EXPORT_SYMBOL(md_unregister_thread);
 
 void md_error(struct mddev *mddev, struct md_rdev *rdev)
 {
-	if (!mddev) {
-		MD_BUG();
-		return;
-	}
-
 	if (!rdev || test_bit(Faulty, &rdev->flags))
 		return;
 
@@ -6876,6 +6679,7 @@ void md_error(struct mddev *mddev, struct md_rdev *rdev)
 		queue_work(md_misc_wq, &mddev->event_work);
 	md_new_event_inintr(mddev);
 }
+EXPORT_SYMBOL(md_error);
 
 /* seq_file implementation /proc/mdstat */
 
@@ -6898,8 +6702,7 @@ static void status_unused(struct seq_file *seq)
 	seq_printf(seq, "\n");
 }
 
-
-static void status_resync(struct seq_file *seq, struct mddev * mddev)
+static void status_resync(struct seq_file *seq, struct mddev *mddev)
 {
 	sector_t max_sectors, resync, res;
 	unsigned long dt, db;
@@ -6919,13 +6722,7 @@ static void status_resync(struct seq_file *seq, struct mddev * mddev)
 	else
 		max_sectors = mddev->dev_sectors;
 
-	/*
-	 * Should not happen.
-	 */
-	if (!max_sectors) {
-		MD_BUG();
-		return;
-	}
+	WARN_ON(max_sectors == 0);
 	/* Pick 'scale' such that (resync>>scale)*1000 will fit
 	 * in a sector_t, and (max_sectors>>scale) will fit in a
 	 * u32, as those are the requirements for sector_div.
@@ -7021,7 +6818,7 @@ static void *md_seq_next(struct seq_file *seq, void *v, loff_t *pos)
 {
 	struct list_head *tmp;
 	struct mddev *next_mddev, *mddev = v;
-	
+
 	++*pos;
 	if (v == (void*)2)
 		return NULL;
@@ -7036,7 +6833,7 @@ static void *md_seq_next(struct seq_file *seq, void *v, loff_t *pos)
 	else {
 		next_mddev = (void*)2;
 		*pos = 0x10000;
-	}		
+	}
 	spin_unlock(&all_mddevs_lock);
 
 	if (v != (void*)1)
@@ -7132,7 +6929,7 @@ static int md_seq_show(struct seq_file *seq, void *v)
 
 		if (mddev->pers) {
 			mddev->pers->status(seq, mddev);
-	 		seq_printf(seq, "\n      ");
+			seq_printf(seq, "\n      ");
 			if (mddev->pers->sync_request) {
 				if (mddev->curr_resync > 2) {
 					status_resync(seq, mddev);
@@ -7150,7 +6947,7 @@ static int md_seq_show(struct seq_file *seq, void *v)
 		seq_printf(seq, "\n");
 	}
 	mddev_unlock(mddev);
-	
+
 	return 0;
 }
 
@@ -7182,7 +6979,7 @@ static unsigned int mdstat_poll(struct file *filp, poll_table *wait)
 	int mask;
 
 	if (md_unloading)
-		return POLLIN|POLLRDNORM|POLLERR|POLLPRI;;
+		return POLLIN|POLLRDNORM|POLLERR|POLLPRI;
 	poll_wait(filp, &md_event_waiters, wait);
 
 	/* always allow read */
@@ -7204,12 +7001,14 @@ static const struct file_operations md_seq_fops = {
 
 int register_md_personality(struct md_personality *p)
 {
+	printk(KERN_INFO "md: %s personality registered for level %d\n",
+						p->name, p->level);
 	spin_lock(&pers_lock);
 	list_add_tail(&p->list, &pers_list);
-	printk(KERN_INFO "md: %s personality registered for level %d\n", p->name, p->level);
 	spin_unlock(&pers_lock);
 	return 0;
 }
+EXPORT_SYMBOL(register_md_personality);
 
 int unregister_md_personality(struct md_personality *p)
 {
@@ -7219,10 +7018,11 @@ int unregister_md_personality(struct md_personality *p)
 	spin_unlock(&pers_lock);
 	return 0;
 }
+EXPORT_SYMBOL(unregister_md_personality);
 
 static int is_mddev_idle(struct mddev *mddev, int init)
 {
-	struct md_rdev * rdev;
+	struct md_rdev *rdev;
 	int idle;
 	int curr_events;
 
@@ -7276,7 +7076,7 @@ void md_done_sync(struct mddev *mddev, int blocks, int ok)
 		// stop recovery, signal do_sync ....
 	}
 }
-
+EXPORT_SYMBOL(md_done_sync);
 
 /* md_write_start(mddev, bi)
  * If we need to update some array metadata (e.g. 'active' flag
@@ -7317,6 +7117,7 @@ void md_write_start(struct mddev *mddev, struct bio *bi)
 	wait_event(mddev->sb_wait,
 		   !test_bit(MD_CHANGE_PENDING, &mddev->flags));
 }
+EXPORT_SYMBOL(md_write_start);
 
 void md_write_end(struct mddev *mddev)
 {
@@ -7327,6 +7128,7 @@ void md_write_end(struct mddev *mddev)
 			mod_timer(&mddev->safemode_timer, jiffies + mddev->safemode_delay);
 	}
 }
+EXPORT_SYMBOL(md_write_end);
 
 /* md_allow_write(mddev)
  * Calling this ensures that the array is marked 'active' so that writes
@@ -7784,6 +7586,34 @@ no_add:
 	return spares;
 }
 
+static void md_start_sync(struct work_struct *ws)
+{
+	struct mddev *mddev = container_of(ws, struct mddev, del_work);
+
+	mddev->sync_thread = md_register_thread(md_do_sync,
+						mddev,
+						"resync");
+	if (!mddev->sync_thread) {
+		printk(KERN_ERR "%s: could not start resync"
+		       " thread...\n",
+		       mdname(mddev));
+		/* leave the spares where they are, it shouldn't hurt */
+		clear_bit(MD_RECOVERY_SYNC, &mddev->recovery);
+		clear_bit(MD_RECOVERY_RESHAPE, &mddev->recovery);
+		clear_bit(MD_RECOVERY_REQUESTED, &mddev->recovery);
+		clear_bit(MD_RECOVERY_CHECK, &mddev->recovery);
+		clear_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
+		wake_up(&resync_wait);
+		if (test_and_clear_bit(MD_RECOVERY_RECOVER,
+				       &mddev->recovery))
+			if (mddev->sysfs_action)
+				sysfs_notify_dirent_safe(mddev->sysfs_action);
+	} else
+		md_wakeup_thread(mddev->sync_thread);
+	sysfs_notify_dirent_safe(mddev->sysfs_action);
+	md_new_event(mddev);
+}
+
 /*
  * This routine is regularly called by all per-raid-array threads to
  * deal with generic issues like resync and super-block update.
@@ -7900,7 +7730,7 @@ void md_check_recovery(struct mddev *mddev)
 
 		if (!test_and_clear_bit(MD_RECOVERY_NEEDED, &mddev->recovery) ||
 		    test_bit(MD_RECOVERY_FROZEN, &mddev->recovery))
-			goto unlock;
+			goto not_running;
 		/* no recovery is running.
 		 * remove any failed drives, then
 		 * add spares if possible.
@@ -7912,7 +7742,7 @@ void md_check_recovery(struct mddev *mddev)
 			if (mddev->pers->check_reshape == NULL ||
 			    mddev->pers->check_reshape(mddev) != 0)
 				/* Cannot proceed */
-				goto unlock;
+				goto not_running;
 			set_bit(MD_RECOVERY_RESHAPE, &mddev->recovery);
 			clear_bit(MD_RECOVERY_RECOVER, &mddev->recovery);
 		} else if ((spares = remove_and_add_spares(mddev, NULL))) {
@@ -7925,7 +7755,7 @@ void md_check_recovery(struct mddev *mddev)
 			clear_bit(MD_RECOVERY_RECOVER, &mddev->recovery);
 		} else if (!test_bit(MD_RECOVERY_SYNC, &mddev->recovery))
 			/* nothing to be done ... */
-			goto unlock;
+			goto not_running;
 
 		if (mddev->pers->sync_request) {
 			if (spares) {
@@ -7935,37 +7765,25 @@ void md_check_recovery(struct mddev *mddev)
 				 */
 				bitmap_write_all(mddev->bitmap);
 			}
-			mddev->sync_thread = md_register_thread(md_do_sync,
-								mddev,
-								"resync");
-			if (!mddev->sync_thread) {
-				printk(KERN_ERR "%s: could not start resync"
-					" thread...\n", 
-					mdname(mddev));
-				/* leave the spares where they are, it shouldn't hurt */
-				clear_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
-				clear_bit(MD_RECOVERY_SYNC, &mddev->recovery);
-				clear_bit(MD_RECOVERY_RESHAPE, &mddev->recovery);
-				clear_bit(MD_RECOVERY_REQUESTED, &mddev->recovery);
-				clear_bit(MD_RECOVERY_CHECK, &mddev->recovery);
-			} else
-				md_wakeup_thread(mddev->sync_thread);
-			sysfs_notify_dirent_safe(mddev->sysfs_action);
-			md_new_event(mddev);
+			INIT_WORK(&mddev->del_work, md_start_sync);
+			queue_work(md_misc_wq, &mddev->del_work);
+			goto unlock;
 		}
-	unlock:
-		wake_up(&mddev->sb_wait);
-
+	not_running:
 		if (!mddev->sync_thread) {
 			clear_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
+			wake_up(&resync_wait);
 			if (test_and_clear_bit(MD_RECOVERY_RECOVER,
 					       &mddev->recovery))
 				if (mddev->sysfs_action)
 					sysfs_notify_dirent_safe(mddev->sysfs_action);
 		}
+	unlock:
+		wake_up(&mddev->sb_wait);
 		mddev_unlock(mddev);
 	}
 }
+EXPORT_SYMBOL(md_check_recovery);
 
 void md_reap_sync_thread(struct mddev *mddev)
 {
@@ -7973,7 +7791,6 @@ void md_reap_sync_thread(struct mddev *mddev)
 
 	/* resync has finished, collect result */
 	md_unregister_thread(&mddev->sync_thread);
-	wake_up(&resync_wait);
 	if (!test_bit(MD_RECOVERY_INTR, &mddev->recovery) &&
 	    !test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) {
 		/* success...*/
@@ -8001,6 +7818,7 @@ void md_reap_sync_thread(struct mddev *mddev)
 	clear_bit(MD_RECOVERY_RESHAPE, &mddev->recovery);
 	clear_bit(MD_RECOVERY_REQUESTED, &mddev->recovery);
 	clear_bit(MD_RECOVERY_CHECK, &mddev->recovery);
+	wake_up(&resync_wait);
 	/* flag recovery needed just to double check */
 	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
 	sysfs_notify_dirent_safe(mddev->sysfs_action);
@@ -8008,6 +7826,7 @@ void md_reap_sync_thread(struct mddev *mddev)
 	if (mddev->event_work.func)
 		queue_work(md_misc_wq, &mddev->event_work);
 }
+EXPORT_SYMBOL(md_reap_sync_thread);
 
 void md_wait_for_blocked_rdev(struct md_rdev *rdev, struct mddev *mddev)
 {
@@ -8641,7 +8460,6 @@ void md_autodetect_dev(dev_t dev)
 	}
 }
 
-
 static void autostart_arrays(int part)
 {
 	struct md_rdev *rdev;
@@ -8665,10 +8483,9 @@ static void autostart_arrays(int part)
 		if (IS_ERR(rdev))
 			continue;
 
-		if (test_bit(Faulty, &rdev->flags)) {
-			MD_BUG();
+		if (test_bit(Faulty, &rdev->flags))
 			continue;
-		}
+
 		set_bit(AutoDetected, &rdev->flags);
 		list_add(&rdev->same_set, &pending_raid_disks);
 		i_passed++;
@@ -8736,20 +8553,8 @@ static int set_ro(const char *val, struct kernel_param *kp)
 
 module_param_call(start_ro, set_ro, get_ro, NULL, S_IRUSR|S_IWUSR);
 module_param(start_dirty_degraded, int, S_IRUGO|S_IWUSR);
-
 module_param_call(new_array, add_named_array, NULL, NULL, S_IWUSR);
 
-EXPORT_SYMBOL(register_md_personality);
-EXPORT_SYMBOL(unregister_md_personality);
-EXPORT_SYMBOL(md_error);
-EXPORT_SYMBOL(md_done_sync);
-EXPORT_SYMBOL(md_write_start);
-EXPORT_SYMBOL(md_write_end);
-EXPORT_SYMBOL(md_register_thread);
-EXPORT_SYMBOL(md_unregister_thread);
-EXPORT_SYMBOL(md_wakeup_thread);
-EXPORT_SYMBOL(md_check_recovery);
-EXPORT_SYMBOL(md_reap_sync_thread);
 MODULE_LICENSE("GPL");
 MODULE_DESCRIPTION("MD RAID framework");
 MODULE_ALIAS("md");
diff --git a/drivers/md/md.h b/drivers/md/md.h
index a49d991f3fe1..03cec5bdcaae 100644
--- a/drivers/md/md.h
+++ b/drivers/md/md.h
@@ -1,15 +1,15 @@
 /*
    md.h : kernel internal structure of the Linux MD driver
           Copyright (C) 1996-98 Ingo Molnar, Gadi Oxman
-	  
+
    This program is free software; you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
    the Free Software Foundation; either version 2, or (at your option)
    any later version.
-   
+
    You should have received a copy of the GNU General Public License
    (for example /usr/src/linux/COPYING); if not, write to the Free
-   Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  
+   Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 #ifndef _MD_MD_H
@@ -56,7 +56,7 @@ struct md_rdev {
 	__u64		sb_events;
 	sector_t	data_offset;	/* start of data in array */
 	sector_t	new_data_offset;/* only relevant while reshaping */
-	sector_t 	sb_start;	/* offset of the super block (in 512byte sectors) */
+	sector_t	sb_start;	/* offset of the super block (in 512byte sectors) */
 	int		sb_size;	/* bytes in the superblock */
 	int		preferred_minor;	/* autorun support */
 
@@ -239,7 +239,7 @@ struct mddev {
 					minor_version,
 					patch_version;
 	int				persistent;
-	int 				external;	/* metadata is
+	int				external;	/* metadata is
 							 * managed externally */
 	char				metadata_type[17]; /* externally set*/
 	int				chunk_sectors;
@@ -248,7 +248,7 @@ struct mddev {
 	char				clevel[16];
 	int				raid_disks;
 	int				max_disks;
-	sector_t			dev_sectors; 	/* used size of
+	sector_t			dev_sectors;	/* used size of
 							 * component devices */
 	sector_t			array_sectors; /* exported array size */
 	int				external_size; /* size managed
@@ -312,7 +312,7 @@ struct mddev {
 	int				parallel_resync;
 
 	int				ok_start_degraded;
-	/* recovery/resync flags 
+	/* recovery/resync flags
 	 * NEEDED:   we might need to start a resync/recover
 	 * RUNNING:  a thread is running, or about to be started
 	 * SYNC:     actually doing a resync, not a recovery
@@ -392,20 +392,20 @@ struct mddev {
 
 	unsigned int			safemode;	/* if set, update "clean" superblock
 							 * when no writes pending.
-							 */ 
+							 */
 	unsigned int			safemode_delay;
 	struct timer_list		safemode_timer;
-	atomic_t			writes_pending; 
+	atomic_t			writes_pending;
 	struct request_queue		*queue;	/* for plugging ... */
 
-	struct bitmap                   *bitmap; /* the bitmap for the device */
+	struct bitmap			*bitmap; /* the bitmap for the device */
 	struct {
 		struct file		*file; /* the bitmap file */
 		loff_t			offset; /* offset from superblock of
 						 * start of bitmap. May be
 						 * negative, but not '0'
 						 * For external metadata, offset
-						 * from start of device. 
+						 * from start of device.
 						 */
 		unsigned long		space; /* space available at this offset */
 		loff_t			default_offset; /* this is the offset to use when
@@ -421,7 +421,7 @@ struct mddev {
 		int			external;
 	} bitmap_info;
 
-	atomic_t 			max_corr_read_errors; /* max read retries */
+	atomic_t			max_corr_read_errors; /* max read retries */
 	struct list_head		all_mddevs;
 
 	struct attribute_group		*to_remove;
@@ -439,7 +439,6 @@ struct mddev {
 	void (*sync_super)(struct mddev *mddev, struct md_rdev *rdev);
 };
 
-
 static inline void rdev_dec_pending(struct md_rdev *rdev, struct mddev *mddev)
 {
 	int faulty = test_bit(Faulty, &rdev->flags);
@@ -449,7 +448,7 @@ static inline void rdev_dec_pending(struct md_rdev *rdev, struct mddev *mddev)
 
 static inline void md_sync_acct(struct block_device *bdev, unsigned long nr_sectors)
 {
-        atomic_add(nr_sectors, &bdev->bd_contains->bd_disk->sync_io);
+	atomic_add(nr_sectors, &bdev->bd_contains->bd_disk->sync_io);
 }
 
 struct md_personality
@@ -463,7 +462,7 @@ struct md_personality
 	int (*stop)(struct mddev *mddev);
 	void (*status)(struct seq_file *seq, struct mddev *mddev);
 	/* error_handler must set ->faulty and clear ->in_sync
-	 * if appropriate, and should abort recovery if needed 
+	 * if appropriate, and should abort recovery if needed
 	 */
 	void (*error_handler)(struct mddev *mddev, struct md_rdev *rdev);
 	int (*hot_add_disk) (struct mddev *mddev, struct md_rdev *rdev);
@@ -493,7 +492,6 @@ struct md_personality
 	void *(*takeover) (struct mddev *mddev);
 };
 
-
 struct md_sysfs_entry {
 	struct attribute attr;
 	ssize_t (*show)(struct mddev *, char *);
@@ -560,7 +558,7 @@ struct md_thread {
 	void			(*run) (struct md_thread *thread);
 	struct mddev		*mddev;
 	wait_queue_head_t	wqueue;
-	unsigned long           flags;
+	unsigned long		flags;
 	struct task_struct	*tsk;
 	unsigned long		timeout;
 	void			*private;
@@ -594,7 +592,7 @@ extern void md_flush_request(struct mddev *mddev, struct bio *bio);
 extern void md_super_write(struct mddev *mddev, struct md_rdev *rdev,
 			   sector_t sector, int size, struct page *page);
 extern void md_super_wait(struct mddev *mddev);
-extern int sync_page_io(struct md_rdev *rdev, sector_t sector, int size, 
+extern int sync_page_io(struct md_rdev *rdev, sector_t sector, int size,
 			struct page *page, int rw, bool metadata_op);
 extern void md_do_sync(struct md_thread *thread);
 extern void md_new_event(struct mddev *mddev);
diff --git a/drivers/md/multipath.c b/drivers/md/multipath.c
index 849ad39f547b..399272f9c042 100644
--- a/drivers/md/multipath.c
+++ b/drivers/md/multipath.c
@@ -31,13 +31,12 @@
 
 #define	NR_RESERVED_BUFS	32
 
-
 static int multipath_map (struct mpconf *conf)
 {
 	int i, disks = conf->raid_disks;
 
 	/*
-	 * Later we do read balancing on the read side 
+	 * Later we do read balancing on the read side
 	 * now we use the first available disk.
 	 */
 
@@ -68,7 +67,6 @@ static void multipath_reschedule_retry (struct multipath_bh *mp_bh)
 	md_wakeup_thread(mddev->thread);
 }
 
-
 /*
  * multipath_end_bh_io() is called when we have finished servicing a multipathed
  * operation and are ready to return a success/failure code to the buffer
@@ -98,8 +96,8 @@ static void multipath_end_request(struct bio *bio, int error)
 		 */
 		char b[BDEVNAME_SIZE];
 		md_error (mp_bh->mddev, rdev);
-		printk(KERN_ERR "multipath: %s: rescheduling sector %llu\n", 
-		       bdevname(rdev->bdev,b), 
+		printk(KERN_ERR "multipath: %s: rescheduling sector %llu\n",
+		       bdevname(rdev->bdev,b),
 		       (unsigned long long)bio->bi_iter.bi_sector);
 		multipath_reschedule_retry(mp_bh);
 	} else
@@ -145,12 +143,12 @@ static void multipath_status (struct seq_file *seq, struct mddev *mddev)
 {
 	struct mpconf *conf = mddev->private;
 	int i;
-	
+
 	seq_printf (seq, " [%d/%d] [", conf->raid_disks,
 		    conf->raid_disks - mddev->degraded);
 	for (i = 0; i < conf->raid_disks; i++)
 		seq_printf (seq, "%s",
-			       conf->multipaths[i].rdev && 
+			       conf->multipaths[i].rdev &&
 			       test_bit(In_sync, &conf->multipaths[i].rdev->flags) ? "U" : "_");
 	seq_printf (seq, "]");
 }
@@ -195,7 +193,7 @@ static void multipath_error (struct mddev *mddev, struct md_rdev *rdev)
 		 * first check if this is a queued request for a device
 		 * which has just failed.
 		 */
-		printk(KERN_ALERT 
+		printk(KERN_ALERT
 		       "multipath: only one IO path left and IO error.\n");
 		/* leave it active... it's all we have */
 		return;
@@ -242,7 +240,6 @@ static void print_multipath_conf (struct mpconf *conf)
 	}
 }
 
-
 static int multipath_add_disk(struct mddev *mddev, struct md_rdev *rdev)
 {
 	struct mpconf *conf = mddev->private;
@@ -325,8 +322,6 @@ abort:
 	return err;
 }
 
-
-
 /*
  * This is a kernel thread which:
  *
@@ -356,7 +351,7 @@ static void multipathd(struct md_thread *thread)
 
 		bio = &mp_bh->bio;
 		bio->bi_iter.bi_sector = mp_bh->master_bio->bi_iter.bi_sector;
-		
+
 		if ((mp_bh->path = multipath_map (conf))<0) {
 			printk(KERN_ALERT "multipath: %s: unrecoverable IO read"
 				" error for block %llu\n",
@@ -414,7 +409,7 @@ static int multipath_run (struct mddev *mddev)
 	conf = kzalloc(sizeof(struct mpconf), GFP_KERNEL);
 	mddev->private = conf;
 	if (!conf) {
-		printk(KERN_ERR 
+		printk(KERN_ERR
 			"multipath: couldn't allocate memory for %s\n",
 			mdname(mddev));
 		goto out;
@@ -423,7 +418,7 @@ static int multipath_run (struct mddev *mddev)
 	conf->multipaths = kzalloc(sizeof(struct multipath_info)*mddev->raid_disks,
 				   GFP_KERNEL);
 	if (!conf->multipaths) {
-		printk(KERN_ERR 
+		printk(KERN_ERR
 			"multipath: couldn't allocate memory for %s\n",
 			mdname(mddev));
 		goto out_free_conf;
@@ -469,7 +464,7 @@ static int multipath_run (struct mddev *mddev)
 	conf->pool = mempool_create_kmalloc_pool(NR_RESERVED_BUFS,
 						 sizeof(struct multipath_bh));
 	if (conf->pool == NULL) {
-		printk(KERN_ERR 
+		printk(KERN_ERR
 			"multipath: couldn't allocate memory for %s\n",
 			mdname(mddev));
 		goto out_free_conf;
@@ -485,7 +480,7 @@ static int multipath_run (struct mddev *mddev)
 		}
 	}
 
-	printk(KERN_INFO 
+	printk(KERN_INFO
 		"multipath: array %s active with %d out of %d IO paths\n",
 		mdname(mddev), conf->raid_disks - mddev->degraded,
 	       mddev->raid_disks);
@@ -512,7 +507,6 @@ out:
 	return -EIO;
 }
 
-
 static int multipath_stop (struct mddev *mddev)
 {
 	struct mpconf *conf = mddev->private;
diff --git a/drivers/md/persistent-data/dm-array.c b/drivers/md/persistent-data/dm-array.c
index 1d75b1dc1e2e..e64b61ad0ef3 100644
--- a/drivers/md/persistent-data/dm-array.c
+++ b/drivers/md/persistent-data/dm-array.c
@@ -645,8 +645,10 @@ static int array_resize(struct dm_array_info *info, dm_block_t root,
 	int r;
 	struct resize resize;
 
-	if (old_size == new_size)
+	if (old_size == new_size) {
+		*new_root = root;
 		return 0;
+	}
 
 	resize.info = info;
 	resize.root = root;
diff --git a/drivers/md/persistent-data/dm-btree-internal.h b/drivers/md/persistent-data/dm-btree-internal.h
index 37d367bb9aa8..bf2b80d5c470 100644
--- a/drivers/md/persistent-data/dm-btree-internal.h
+++ b/drivers/md/persistent-data/dm-btree-internal.h
@@ -42,6 +42,12 @@ struct btree_node {
 } __packed;
 
 
+/*
+ * Locks a block using the btree node validator.
+ */
+int bn_read_lock(struct dm_btree_info *info, dm_block_t b,
+		 struct dm_block **result);
+
 void inc_children(struct dm_transaction_manager *tm, struct btree_node *n,
 		  struct dm_btree_value_type *vt);
 
diff --git a/drivers/md/persistent-data/dm-btree-spine.c b/drivers/md/persistent-data/dm-btree-spine.c
index cf9fd676ae44..1b5e13ec7f96 100644
--- a/drivers/md/persistent-data/dm-btree-spine.c
+++ b/drivers/md/persistent-data/dm-btree-spine.c
@@ -92,7 +92,7 @@ struct dm_block_validator btree_node_validator = {
 
 /*----------------------------------------------------------------*/
 
-static int bn_read_lock(struct dm_btree_info *info, dm_block_t b,
+int bn_read_lock(struct dm_btree_info *info, dm_block_t b,
 		 struct dm_block **result)
 {
 	return dm_tm_read_lock(info->tm, b, &btree_node_validator, result);
diff --git a/drivers/md/persistent-data/dm-btree.c b/drivers/md/persistent-data/dm-btree.c
index 416060c25709..200ac12a1d40 100644
--- a/drivers/md/persistent-data/dm-btree.c
+++ b/drivers/md/persistent-data/dm-btree.c
@@ -847,22 +847,26 @@ EXPORT_SYMBOL_GPL(dm_btree_find_lowest_key);
  * FIXME: We shouldn't use a recursive algorithm when we have limited stack
  * space.  Also this only works for single level trees.
  */
-static int walk_node(struct ro_spine *s, dm_block_t block,
+static int walk_node(struct dm_btree_info *info, dm_block_t block,
 		     int (*fn)(void *context, uint64_t *keys, void *leaf),
 		     void *context)
 {
 	int r;
 	unsigned i, nr;
+	struct dm_block *node;
 	struct btree_node *n;
 	uint64_t keys;
 
-	r = ro_step(s, block);
-	n = ro_node(s);
+	r = bn_read_lock(info, block, &node);
+	if (r)
+		return r;
+
+	n = dm_block_data(node);
 
 	nr = le32_to_cpu(n->header.nr_entries);
 	for (i = 0; i < nr; i++) {
 		if (le32_to_cpu(n->header.flags) & INTERNAL_NODE) {
-			r = walk_node(s, value64(n, i), fn, context);
+			r = walk_node(info, value64(n, i), fn, context);
 			if (r)
 				goto out;
 		} else {
@@ -874,7 +878,7 @@ static int walk_node(struct ro_spine *s, dm_block_t block,
 	}
 
 out:
-	ro_pop(s);
+	dm_tm_unlock(info->tm, node);
 	return r;
 }
 
@@ -882,15 +886,7 @@ int dm_btree_walk(struct dm_btree_info *info, dm_block_t root,
 		  int (*fn)(void *context, uint64_t *keys, void *leaf),
 		  void *context)
 {
-	int r;
-	struct ro_spine spine;
-
 	BUG_ON(info->levels > 1);
-
-	init_ro_spine(&spine, info);
-	r = walk_node(&spine, root, fn, context);
-	exit_ro_spine(&spine);
-
-	return r;
+	return walk_node(info, root, fn, context);
 }
 EXPORT_SYMBOL_GPL(dm_btree_walk);
diff --git a/drivers/md/persistent-data/dm-space-map-metadata.c b/drivers/md/persistent-data/dm-space-map-metadata.c
index 786b689bdfc7..e8a904298887 100644
--- a/drivers/md/persistent-data/dm-space-map-metadata.c
+++ b/drivers/md/persistent-data/dm-space-map-metadata.c
@@ -564,7 +564,9 @@ static int sm_bootstrap_get_nr_blocks(struct dm_space_map *sm, dm_block_t *count
 {
 	struct sm_metadata *smm = container_of(sm, struct sm_metadata, sm);
 
-	return smm->ll.nr_blocks;
+	*count = smm->ll.nr_blocks;
+
+	return 0;
 }
 
 static int sm_bootstrap_get_nr_free(struct dm_space_map *sm, dm_block_t *count)
@@ -581,7 +583,9 @@ static int sm_bootstrap_get_count(struct dm_space_map *sm, dm_block_t b,
 {
 	struct sm_metadata *smm = container_of(sm, struct sm_metadata, sm);
 
-	return b < smm->begin ? 1 : 0;
+	*result = (b < smm->begin) ? 1 : 0;
+
+	return 0;
 }
 
 static int sm_bootstrap_count_is_more_than_one(struct dm_space_map *sm,
diff --git a/drivers/md/persistent-data/dm-transaction-manager.c b/drivers/md/persistent-data/dm-transaction-manager.c
index 3bc30a0ae3d6..9cb797d800cf 100644
--- a/drivers/md/persistent-data/dm-transaction-manager.c
+++ b/drivers/md/persistent-data/dm-transaction-manager.c
@@ -10,6 +10,8 @@
 #include "dm-persistent-data-internal.h"
 
 #include <linux/export.h>
+#include <linux/mutex.h>
+#include <linux/hash.h>
 #include <linux/slab.h>
 #include <linux/device-mapper.h>
 
@@ -17,6 +19,61 @@
 
 /*----------------------------------------------------------------*/
 
+#define PREFETCH_SIZE 128
+#define PREFETCH_BITS 7
+#define PREFETCH_SENTINEL ((dm_block_t) -1ULL)
+
+struct prefetch_set {
+	struct mutex lock;
+	dm_block_t blocks[PREFETCH_SIZE];
+};
+
+static unsigned prefetch_hash(dm_block_t b)
+{
+	return hash_64(b, PREFETCH_BITS);
+}
+
+static void prefetch_wipe(struct prefetch_set *p)
+{
+	unsigned i;
+	for (i = 0; i < PREFETCH_SIZE; i++)
+		p->blocks[i] = PREFETCH_SENTINEL;
+}
+
+static void prefetch_init(struct prefetch_set *p)
+{
+	mutex_init(&p->lock);
+	prefetch_wipe(p);
+}
+
+static void prefetch_add(struct prefetch_set *p, dm_block_t b)
+{
+	unsigned h = prefetch_hash(b);
+
+	mutex_lock(&p->lock);
+	if (p->blocks[h] == PREFETCH_SENTINEL)
+		p->blocks[h] = b;
+
+	mutex_unlock(&p->lock);
+}
+
+static void prefetch_issue(struct prefetch_set *p, struct dm_block_manager *bm)
+{
+	unsigned i;
+
+	mutex_lock(&p->lock);
+
+	for (i = 0; i < PREFETCH_SIZE; i++)
+		if (p->blocks[i] != PREFETCH_SENTINEL) {
+			dm_bm_prefetch(bm, p->blocks[i]);
+			p->blocks[i] = PREFETCH_SENTINEL;
+		}
+
+	mutex_unlock(&p->lock);
+}
+
+/*----------------------------------------------------------------*/
+
 struct shadow_info {
 	struct hlist_node hlist;
 	dm_block_t where;
@@ -37,6 +94,8 @@ struct dm_transaction_manager {
 
 	spinlock_t lock;
 	struct hlist_head buckets[DM_HASH_SIZE];
+
+	struct prefetch_set prefetches;
 };
 
 /*----------------------------------------------------------------*/
@@ -117,6 +176,8 @@ static struct dm_transaction_manager *dm_tm_create(struct dm_block_manager *bm,
 	for (i = 0; i < DM_HASH_SIZE; i++)
 		INIT_HLIST_HEAD(tm->buckets + i);
 
+	prefetch_init(&tm->prefetches);
+
 	return tm;
 }
 
@@ -268,8 +329,14 @@ int dm_tm_read_lock(struct dm_transaction_manager *tm, dm_block_t b,
 		    struct dm_block_validator *v,
 		    struct dm_block **blk)
 {
-	if (tm->is_clone)
-		return dm_bm_read_try_lock(tm->real->bm, b, v, blk);
+	if (tm->is_clone) {
+		int r = dm_bm_read_try_lock(tm->real->bm, b, v, blk);
+
+		if (r == -EWOULDBLOCK)
+			prefetch_add(&tm->real->prefetches, b);
+
+		return r;
+	}
 
 	return dm_bm_read_lock(tm->bm, b, v, blk);
 }
@@ -317,6 +384,12 @@ struct dm_block_manager *dm_tm_get_bm(struct dm_transaction_manager *tm)
 	return tm->bm;
 }
 
+void dm_tm_issue_prefetches(struct dm_transaction_manager *tm)
+{
+	prefetch_issue(&tm->prefetches, tm->bm);
+}
+EXPORT_SYMBOL_GPL(dm_tm_issue_prefetches);
+
 /*----------------------------------------------------------------*/
 
 static int dm_tm_create_internal(struct dm_block_manager *bm,
diff --git a/drivers/md/persistent-data/dm-transaction-manager.h b/drivers/md/persistent-data/dm-transaction-manager.h
index 2772ed2a781a..2e0d4d66fb1b 100644
--- a/drivers/md/persistent-data/dm-transaction-manager.h
+++ b/drivers/md/persistent-data/dm-transaction-manager.h
@@ -109,6 +109,13 @@ int dm_tm_ref(struct dm_transaction_manager *tm, dm_block_t b,
 struct dm_block_manager *dm_tm_get_bm(struct dm_transaction_manager *tm);
 
 /*
+ * If you're using a non-blocking clone the tm will build up a list of
+ * requested blocks that weren't in core.  This call will request those
+ * blocks to be prefetched.
+ */
+void dm_tm_issue_prefetches(struct dm_transaction_manager *tm);
+
+/*
  * A little utility that ties the knot by producing a transaction manager
  * that has a space map managed by the transaction manager...
  *
diff --git a/drivers/md/raid0.c b/drivers/md/raid0.c
index cf91f5910c7c..ba6b85de96d2 100644
--- a/drivers/md/raid0.c
+++ b/drivers/md/raid0.c
@@ -1,10 +1,9 @@
 /*
    raid0.c : Multiple Devices driver for Linux
-             Copyright (C) 1994-96 Marc ZYNGIER
+	     Copyright (C) 1994-96 Marc ZYNGIER
 	     <zyngier@ufr-info-p7.ibp.fr> or
 	     <maz@gloups.fdn.fr>
-             Copyright (C) 1999, 2000 Ingo Molnar, Red Hat
-
+	     Copyright (C) 1999, 2000 Ingo Molnar, Red Hat
 
    RAID-0 management functions.
 
@@ -12,10 +11,10 @@
    it under the terms of the GNU General Public License as published by
    the Free Software Foundation; either version 2, or (at your option)
    any later version.
-   
+
    You should have received a copy of the GNU General Public License
    (for example /usr/src/linux/COPYING); if not, write to the Free
-   Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  
+   Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
 
 #include <linux/blkdev.h>
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index 55de4f6f7eaf..40b35be34f8d 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -494,7 +494,6 @@ static void raid1_end_write_request(struct bio *bio, int error)
 		bio_put(to_put);
 }
 
-
 /*
  * This routine returns the disk from which the requested read should
  * be done. There is a per-array 'next expected sequential IO' sector
@@ -901,18 +900,18 @@ static sector_t wait_barrier(struct r1conf *conf, struct bio *bio)
 		 * However if there are already pending
 		 * requests (preventing the barrier from
 		 * rising completely), and the
-		 * pre-process bio queue isn't empty,
+		 * per-process bio queue isn't empty,
 		 * then don't wait, as we need to empty
-		 * that queue to get the nr_pending
-		 * count down.
+		 * that queue to allow conf->start_next_window
+		 * to increase.
 		 */
 		wait_event_lock_irq(conf->wait_barrier,
 				    !conf->array_frozen &&
 				    (!conf->barrier ||
-				    ((conf->start_next_window <
-				      conf->next_resync + RESYNC_SECTORS) &&
-				     current->bio_list &&
-				     !bio_list_empty(current->bio_list))),
+				     ((conf->start_next_window <
+				       conf->next_resync + RESYNC_SECTORS) &&
+				      current->bio_list &&
+				      !bio_list_empty(current->bio_list))),
 				    conf->resync_lock);
 		conf->nr_waiting--;
 	}
@@ -1001,8 +1000,7 @@ static void unfreeze_array(struct r1conf *conf)
 	spin_unlock_irq(&conf->resync_lock);
 }
 
-
-/* duplicate the data pages for behind I/O 
+/* duplicate the data pages for behind I/O
  */
 static void alloc_behind_pages(struct bio *bio, struct r1bio *r1_bio)
 {
@@ -1471,7 +1469,6 @@ static void status(struct seq_file *seq, struct mddev *mddev)
 	seq_printf(seq, "]");
 }
 
-
 static void error(struct mddev *mddev, struct md_rdev *rdev)
 {
 	char b[BDEVNAME_SIZE];
@@ -1565,7 +1562,7 @@ static int raid1_spare_active(struct mddev *mddev)
 	unsigned long flags;
 
 	/*
-	 * Find all failed disks within the RAID1 configuration 
+	 * Find all failed disks within the RAID1 configuration
 	 * and mark them readable.
 	 * Called under mddev lock, so rcu protection not needed.
 	 */
@@ -1606,7 +1603,6 @@ static int raid1_spare_active(struct mddev *mddev)
 	return count;
 }
 
-
 static int raid1_add_disk(struct mddev *mddev, struct md_rdev *rdev)
 {
 	struct r1conf *conf = mddev->private;
@@ -1735,7 +1731,6 @@ abort:
 	return err;
 }
 
-
 static void end_sync_read(struct bio *bio, int error)
 {
 	struct r1bio *r1_bio = bio->bi_private;
@@ -1947,7 +1942,7 @@ static int fix_sync_read_error(struct r1bio *r1_bio)
 	return 1;
 }
 
-static int process_checks(struct r1bio *r1_bio)
+static void process_checks(struct r1bio *r1_bio)
 {
 	/* We have read all readable devices.  If we haven't
 	 * got the block, then there is no hope left.
@@ -2039,7 +2034,6 @@ static int process_checks(struct r1bio *r1_bio)
 
 		bio_copy_data(sbio, pbio);
 	}
-	return 0;
 }
 
 static void sync_request_write(struct mddev *mddev, struct r1bio *r1_bio)
@@ -2057,8 +2051,8 @@ static void sync_request_write(struct mddev *mddev, struct r1bio *r1_bio)
 			return;
 
 	if (test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery))
-		if (process_checks(r1_bio) < 0)
-			return;
+		process_checks(r1_bio);
+
 	/*
 	 * schedule writes
 	 */
@@ -2458,7 +2452,6 @@ static void raid1d(struct md_thread *thread)
 	blk_finish_plug(&plug);
 }
 
-
 static int init_resync(struct r1conf *conf)
 {
 	int buffs;
@@ -2722,7 +2715,7 @@ static sector_t sync_request(struct mddev *mddev, sector_t sector_nr, int *skipp
 						/* remove last page from this bio */
 						bio->bi_vcnt--;
 						bio->bi_iter.bi_size -= len;
-						bio->bi_flags &= ~(1<< BIO_SEG_VALID);
+						__clear_bit(BIO_SEG_VALID, &bio->bi_flags);
 					}
 					goto bio_full;
 				}
@@ -2947,9 +2940,9 @@ static int run(struct mddev *mddev)
 		printk(KERN_NOTICE "md/raid1:%s: not clean"
 		       " -- starting background reconstruction\n",
 		       mdname(mddev));
-	printk(KERN_INFO 
+	printk(KERN_INFO
 		"md/raid1:%s: active with %d out of %d mirrors\n",
-		mdname(mddev), mddev->raid_disks - mddev->degraded, 
+		mdname(mddev), mddev->raid_disks - mddev->degraded,
 		mddev->raid_disks);
 
 	/*
diff --git a/drivers/md/raid1.h b/drivers/md/raid1.h
index 9bebca7bff2f..33bda55ef9f7 100644
--- a/drivers/md/raid1.h
+++ b/drivers/md/raid1.h
@@ -90,7 +90,6 @@ struct r1conf {
 	 */
 	int			recovery_disabled;
 
-
 	/* poolinfo contains information about the content of the
 	 * mempools - it changes when the array grows or shrinks
 	 */
@@ -103,7 +102,6 @@ struct r1conf {
 	 */
 	struct page		*tmppage;
 
-
 	/* When taking over an array from a different personality, we store
 	 * the new thread here until we fully activate the array.
 	 */
diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index 6703751d87d7..32e282f4c83c 100644
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -366,7 +366,6 @@ static void raid10_end_read_request(struct bio *bio, int error)
 	struct md_rdev *rdev;
 	struct r10conf *conf = r10_bio->mddev->private;
 
-
 	slot = r10_bio->read_slot;
 	dev = r10_bio->devs[slot].devnum;
 	rdev = r10_bio->devs[slot].rdev;
@@ -1559,7 +1558,6 @@ static void make_request(struct mddev *mddev, struct bio *bio)
 
 	md_write_start(mddev, bio);
 
-
 	do {
 
 		/*
@@ -1782,7 +1780,6 @@ static int raid10_spare_active(struct mddev *mddev)
 	return count;
 }
 
-
 static int raid10_add_disk(struct mddev *mddev, struct md_rdev *rdev)
 {
 	struct r10conf *conf = mddev->private;
@@ -1929,7 +1926,6 @@ abort:
 	return err;
 }
 
-
 static void end_sync_read(struct bio *bio, int error)
 {
 	struct r10bio *r10_bio = bio->bi_private;
@@ -2295,7 +2291,6 @@ static void recovery_request_write(struct mddev *mddev, struct r10bio *r10_bio)
 	}
 }
 
-
 /*
  * Used by fix_read_error() to decay the per rdev read_errors.
  * We halve the read error count for every hour that has elapsed
@@ -2852,7 +2847,6 @@ static void raid10d(struct md_thread *thread)
 	blk_finish_plug(&plug);
 }
 
-
 static int init_resync(struct r10conf *conf)
 {
 	int buffs;
@@ -3388,7 +3382,7 @@ static sector_t sync_request(struct mddev *mddev, sector_t sector_nr,
 				/* remove last page from this bio */
 				bio2->bi_vcnt--;
 				bio2->bi_iter.bi_size -= len;
-				bio2->bi_flags &= ~(1<< BIO_SEG_VALID);
+				__clear_bit(BIO_SEG_VALID, &bio2->bi_flags);
 			}
 			goto bio_full;
 		}
@@ -3776,7 +3770,6 @@ static int run(struct mddev *mddev)
 		blk_queue_merge_bvec(mddev->queue, raid10_mergeable_bvec);
 	}
 
-
 	if (md_integrity_register(mddev))
 		goto out_free_conf;
 
@@ -3834,6 +3827,8 @@ static int stop(struct mddev *mddev)
 		mempool_destroy(conf->r10bio_pool);
 	safe_put_page(conf->tmppage);
 	kfree(conf->mirrors);
+	kfree(conf->mirrors_old);
+	kfree(conf->mirrors_new);
 	kfree(conf);
 	mddev->private = NULL;
 	return 0;
@@ -4121,7 +4116,7 @@ static int raid10_start_reshape(struct mddev *mddev)
 		memcpy(conf->mirrors_new, conf->mirrors,
 		       sizeof(struct raid10_info)*conf->prev.raid_disks);
 		smp_mb();
-		kfree(conf->mirrors_old); /* FIXME and elsewhere */
+		kfree(conf->mirrors_old);
 		conf->mirrors_old = conf->mirrors;
 		conf->mirrors = conf->mirrors_new;
 		conf->mirrors_new = NULL;
@@ -4416,7 +4411,7 @@ read_more:
 	read_bio->bi_end_io = end_sync_read;
 	read_bio->bi_rw = READ;
 	read_bio->bi_flags &= (~0UL << BIO_RESET_BITS);
-	read_bio->bi_flags |= 1 << BIO_UPTODATE;
+	__set_bit(BIO_UPTODATE, &read_bio->bi_flags);
 	read_bio->bi_vcnt = 0;
 	read_bio->bi_iter.bi_size = 0;
 	r10_bio->master_bio = read_bio;
@@ -4473,7 +4468,7 @@ read_more:
 				/* Remove last page from this bio */
 				bio2->bi_vcnt--;
 				bio2->bi_iter.bi_size -= len;
-				bio2->bi_flags &= ~(1<<BIO_SEG_VALID);
+				__clear_bit(BIO_SEG_VALID, &bio2->bi_flags);
 			}
 			goto bio_full;
 		}
@@ -4575,7 +4570,6 @@ static void end_reshape(struct r10conf *conf)
 	conf->fullsync = 0;
 }
 
-
 static int handle_reshape_read_error(struct mddev *mddev,
 				     struct r10bio *r10_bio)
 {
diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index 9f0fbecd1eb5..c1b0d52bfcb0 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -463,7 +463,6 @@ static inline void insert_hash(struct r5conf *conf, struct stripe_head *sh)
 	hlist_add_head(&sh->hash, hp);
 }
 
-
 /* find an idle stripe, make sure it is unhashed, and return it. */
 static struct stripe_head *get_free_stripe(struct r5conf *conf, int hash)
 {
@@ -531,9 +530,7 @@ static void init_stripe(struct stripe_head *sh, sector_t sector, int previous)
 	BUG_ON(stripe_operations_active(sh));
 
 	pr_debug("init_stripe called, stripe %llu\n",
-		(unsigned long long)sh->sector);
-
-	remove_hash(sh);
+		(unsigned long long)sector);
 retry:
 	seq = read_seqcount_begin(&conf->gen_lock);
 	sh->generation = conf->generation - previous;
@@ -542,7 +539,6 @@ retry:
 	stripe_set_idx(sector, conf, previous, sh);
 	sh->state = 0;
 
-
 	for (i = sh->disks; i--; ) {
 		struct r5dev *dev = &sh->dev[i];
 
@@ -1350,7 +1346,6 @@ ops_run_compute6_2(struct stripe_head *sh, struct raid5_percpu *percpu)
 	}
 }
 
-
 static void ops_complete_prexor(void *stripe_head_ref)
 {
 	struct stripe_head *sh = stripe_head_ref;
@@ -2419,7 +2414,6 @@ static sector_t raid5_compute_sector(struct r5conf *conf, sector_t r_sector,
 	return new_sector;
 }
 
-
 static sector_t compute_blocknr(struct stripe_head *sh, int i, int previous)
 {
 	struct r5conf *conf = sh->raid_conf;
@@ -2437,7 +2431,6 @@ static sector_t compute_blocknr(struct stripe_head *sh, int i, int previous)
 	sector_t r_sector;
 	struct stripe_head sh2;
 
-
 	chunk_offset = sector_div(new_sector, sectors_per_chunk);
 	stripe = new_sector;
 
@@ -2541,7 +2534,6 @@ static sector_t compute_blocknr(struct stripe_head *sh, int i, int previous)
 	return r_sector;
 }
 
-
 static void
 schedule_reconstruction(struct stripe_head *sh, struct stripe_head_state *s,
 			 int rcw, int expand)
@@ -2925,8 +2917,11 @@ static int fetch_block(struct stripe_head *sh, struct stripe_head_state *s,
 	     (sh->raid_conf->level <= 5 && s->failed && fdev[0]->towrite &&
 	      (!test_bit(R5_Insync, &dev->flags) || test_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) &&
 	      !test_bit(R5_OVERWRITE, &fdev[0]->flags)) ||
-	     (sh->raid_conf->level == 6 && s->failed && s->to_write &&
-	      s->to_write - s->non_overwrite < sh->raid_conf->raid_disks - 2 &&
+	     ((sh->raid_conf->level == 6 ||
+	       sh->sector >= sh->raid_conf->mddev->recovery_cp)
+	      && s->failed && s->to_write &&
+	      (s->to_write - s->non_overwrite <
+	       sh->raid_conf->raid_disks - sh->raid_conf->max_degraded) &&
 	      (!test_bit(R5_Insync, &dev->flags) || test_bit(STRIPE_PREREAD_ACTIVE, &sh->state))))) {
 		/* we would like to get this block, possibly by computing it,
 		 * otherwise read it if the backing disk is insync
@@ -3013,7 +3008,6 @@ static void handle_stripe_fill(struct stripe_head *sh,
 	set_bit(STRIPE_HANDLE, &sh->state);
 }
 
-
 /* handle_stripe_clean_event
  * any written block on an uptodate or failed drive can be returned.
  * Note that if we 'wrote' to a failed drive, it will be UPTODATE, but
@@ -3304,7 +3298,6 @@ static void handle_parity_checks5(struct r5conf *conf, struct stripe_head *sh,
 	}
 }
 
-
 static void handle_parity_checks6(struct r5conf *conf, struct stripe_head *sh,
 				  struct stripe_head_state *s,
 				  int disks)
@@ -3939,7 +3932,6 @@ static void handle_stripe(struct stripe_head *sh)
 			}
 		}
 
-
 	/* Finish reconstruct operations initiated by the expansion process */
 	if (sh->reconstruct_state == reconstruct_state_result) {
 		struct stripe_head *sh_src
@@ -4137,7 +4129,6 @@ static int raid5_mergeable_bvec(struct request_queue *q,
 		return max;
 }
 
-
 static int in_chunk_boundary(struct mddev *mddev, struct bio *bio)
 {
 	sector_t sector = bio->bi_iter.bi_sector + get_start_sect(bio->bi_bdev);
@@ -4167,7 +4158,6 @@ static void add_bio_to_retry(struct bio *bi,struct r5conf *conf)
 	md_wakeup_thread(conf->mddev->thread);
 }
 
-
 static struct bio *remove_bio_from_retry(struct r5conf *conf)
 {
 	struct bio *bi;
@@ -4191,7 +4181,6 @@ static struct bio *remove_bio_from_retry(struct r5conf *conf)
 	return bi;
 }
 
-
 /*
  *  The "raid5_align_endio" should check if the read succeeded and if it
  *  did, call bio_endio on the original bio (having bio_put the new bio
@@ -4224,7 +4213,6 @@ static void raid5_align_endio(struct bio *bi, int error)
 		return;
 	}
 
-
 	pr_debug("raid5_align_endio : io error...handing IO for a retry\n");
 
 	add_bio_to_retry(raid_bi, conf);
@@ -4249,7 +4237,6 @@ static int bio_fits_rdev(struct bio *bi)
 	return 1;
 }
 
-
 static int chunk_aligned_read(struct mddev *mddev, struct bio * raid_bio)
 {
 	struct r5conf *conf = mddev->private;
@@ -4301,7 +4288,7 @@ static int chunk_aligned_read(struct mddev *mddev, struct bio * raid_bio)
 		rcu_read_unlock();
 		raid_bio->bi_next = (void*)rdev;
 		align_bi->bi_bdev =  rdev->bdev;
-		align_bi->bi_flags &= ~(1 << BIO_SEG_VALID);
+		__clear_bit(BIO_SEG_VALID, &align_bi->bi_flags);
 
 		if (!bio_fits_rdev(align_bi) ||
 		    is_badblock(rdev, align_bi->bi_iter.bi_sector,
@@ -5446,7 +5433,6 @@ raid5_skip_copy = __ATTR(skip_copy, S_IRUGO | S_IWUSR,
 					raid5_show_skip_copy,
 					raid5_store_skip_copy);
 
-
 static ssize_t
 stripe_cache_active_show(struct mddev *mddev, char *page)
 {
@@ -5898,7 +5884,6 @@ static struct r5conf *setup_conf(struct mddev *mddev)
 		return ERR_PTR(-ENOMEM);
 }
 
-
 static int only_parity(int raid_disk, int algo, int raid_disks, int max_degraded)
 {
 	switch (algo) {
@@ -5911,7 +5896,7 @@ static int only_parity(int raid_disk, int algo, int raid_disks, int max_degraded
 			return 1;
 		break;
 	case ALGORITHM_PARITY_0_6:
-		if (raid_disk == 0 || 
+		if (raid_disk == 0 ||
 		    raid_disk == raid_disks - 1)
 			return 1;
 		break;
@@ -6165,7 +6150,6 @@ static int run(struct mddev *mddev)
 							"reshape");
 	}
 
-
 	/* Ok, everything is just fine now */
 	if (mddev->to_remove == &raid5_attrs_group)
 		mddev->to_remove = NULL;
@@ -6814,7 +6798,6 @@ static void raid5_quiesce(struct mddev *mddev, int state)
 	}
 }
 
-
 static void *raid45_takeover_raid0(struct mddev *mddev, int level)
 {
 	struct r0conf *raid0_conf = mddev->private;
@@ -6841,7 +6824,6 @@ static void *raid45_takeover_raid0(struct mddev *mddev, int level)
 	return setup_conf(mddev);
 }
 
-
 static void *raid5_takeover_raid1(struct mddev *mddev)
 {
 	int chunksect;
@@ -6902,7 +6884,6 @@ static void *raid5_takeover_raid6(struct mddev *mddev)
 	return setup_conf(mddev);
 }
 
-
 static int raid5_check_reshape(struct mddev *mddev)
 {
 	/* For a 2-drive array, the layout and chunk size can be changed
@@ -7051,7 +7032,6 @@ static void *raid6_takeover(struct mddev *mddev)
 	return setup_conf(mddev);
 }
 
-
 static struct md_personality raid6_personality =
 {
 	.name		= "raid6",
diff --git a/drivers/md/raid5.h b/drivers/md/raid5.h
index bc72cd4be5f8..d59f5ca743cd 100644
--- a/drivers/md/raid5.h
+++ b/drivers/md/raid5.h
@@ -155,7 +155,7 @@
  */
 
 /*
- * Operations state - intermediate states that are visible outside of 
+ * Operations state - intermediate states that are visible outside of
  *   STRIPE_ACTIVE.
  * In general _idle indicates nothing is running, _run indicates a data
  * processing operation is active, and _result means the data processing result
@@ -364,7 +364,6 @@ enum {
  * HANDLE gets cleared if stripe_handle leaves nothing locked.
  */
 
-
 struct disk_info {
 	struct md_rdev	*rdev, *replacement;
 };
@@ -528,7 +527,6 @@ struct r5conf {
 #define ALGORITHM_ROTATING_N_RESTART	9 /* DDF PRL=6 RLQ=2 */
 #define ALGORITHM_ROTATING_N_CONTINUE	10 /*DDF PRL=6 RLQ=3 */
 
-
 /* For every RAID5 algorithm we define a RAID6 algorithm
  * with exactly the same layout for data and parity, and
  * with the Q block always on the last device (N-1).
diff --git a/drivers/media/Kconfig b/drivers/media/Kconfig
index 3c89fcbc621e..49cd30870e0d 100644
--- a/drivers/media/Kconfig
+++ b/drivers/media/Kconfig
@@ -160,7 +160,6 @@ source "drivers/media/usb/Kconfig"
 source "drivers/media/pci/Kconfig"
 source "drivers/media/platform/Kconfig"
 source "drivers/media/mmc/Kconfig"
-source "drivers/media/parport/Kconfig"
 source "drivers/media/radio/Kconfig"
 
 comment "Supported FireWire (IEEE 1394) Adapters"
diff --git a/drivers/media/Makefile b/drivers/media/Makefile
index 620f275a45c9..e608bbce0c35 100644
--- a/drivers/media/Makefile
+++ b/drivers/media/Makefile
@@ -28,6 +28,6 @@ obj-y += rc/
 # Finally, merge the drivers that require the core
 #
 
-obj-y += common/ platform/ pci/ usb/ mmc/ firewire/ parport/
+obj-y += common/ platform/ pci/ usb/ mmc/ firewire/
 obj-$(CONFIG_VIDEO_DEV) += radio/
 
diff --git a/drivers/media/common/b2c2/flexcop.h b/drivers/media/common/b2c2/flexcop.h
index 897b10c85ad9..8942bdacbf61 100644
--- a/